diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15961 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9967040210942653,
+  "eval_steps": 500,
+  "global_step": 2274,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0013183915622940012,
+      "grad_norm": 51.917762451625315,
+      "learning_rate": 0.0,
+      "loss": 11.2575,
+      "step": 1
+    },
+    {
+      "epoch": 0.0026367831245880024,
+      "grad_norm": 48.27282286585122,
+      "learning_rate": 2.192982456140351e-07,
+      "loss": 11.4033,
+      "step": 2
+    },
+    {
+      "epoch": 0.003955174686882004,
+      "grad_norm": 49.93126516903941,
+      "learning_rate": 4.385964912280702e-07,
+      "loss": 11.3734,
+      "step": 3
+    },
+    {
+      "epoch": 0.005273566249176005,
+      "grad_norm": 49.25140499115234,
+      "learning_rate": 6.578947368421053e-07,
+      "loss": 11.3809,
+      "step": 4
+    },
+    {
+      "epoch": 0.0065919578114700065,
+      "grad_norm": 49.660885359671525,
+      "learning_rate": 8.771929824561404e-07,
+      "loss": 11.3691,
+      "step": 5
+    },
+    {
+      "epoch": 0.007910349373764008,
+      "grad_norm": 51.69146961221111,
+      "learning_rate": 1.0964912280701754e-06,
+      "loss": 11.2251,
+      "step": 6
+    },
+    {
+      "epoch": 0.00922874093605801,
+      "grad_norm": 52.959938195758845,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 11.2523,
+      "step": 7
+    },
+    {
+      "epoch": 0.01054713249835201,
+      "grad_norm": 54.91199516011705,
+      "learning_rate": 1.5350877192982455e-06,
+      "loss": 10.97,
+      "step": 8
+    },
+    {
+      "epoch": 0.011865524060646011,
+      "grad_norm": 55.542874835577,
+      "learning_rate": 1.7543859649122807e-06,
+      "loss": 10.9778,
+      "step": 9
+    },
+    {
+      "epoch": 0.013183915622940013,
+      "grad_norm": 73.77501967808887,
+      "learning_rate": 1.9736842105263157e-06,
+      "loss": 9.9818,
+      "step": 10
+    },
+    {
+      "epoch": 0.014502307185234015,
+      "grad_norm": 76.77946087915366,
+      "learning_rate": 2.1929824561403507e-06,
+      "loss": 9.7545,
+      "step": 11
+    },
+    {
+      "epoch": 0.015820698747528016,
+      "grad_norm": 86.28421917106388,
+      "learning_rate": 2.412280701754386e-06,
+      "loss": 9.5841,
+      "step": 12
+    },
+    {
+      "epoch": 0.017139090309822018,
+      "grad_norm": 90.48301349345185,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 9.2563,
+      "step": 13
+    },
+    {
+      "epoch": 0.01845748187211602,
+      "grad_norm": 66.7233733044322,
+      "learning_rate": 2.850877192982456e-06,
+      "loss": 4.2705,
+      "step": 14
+    },
+    {
+      "epoch": 0.01977587343441002,
+      "grad_norm": 61.21956059545141,
+      "learning_rate": 3.070175438596491e-06,
+      "loss": 4.0206,
+      "step": 15
+    },
+    {
+      "epoch": 0.02109426499670402,
+      "grad_norm": 45.858758438793615,
+      "learning_rate": 3.2894736842105265e-06,
+      "loss": 3.5524,
+      "step": 16
+    },
+    {
+      "epoch": 0.02241265655899802,
+      "grad_norm": 38.19463256862833,
+      "learning_rate": 3.5087719298245615e-06,
+      "loss": 3.2381,
+      "step": 17
+    },
+    {
+      "epoch": 0.023731048121292023,
+      "grad_norm": 15.130237110415795,
+      "learning_rate": 3.7280701754385965e-06,
+      "loss": 2.2201,
+      "step": 18
+    },
+    {
+      "epoch": 0.025049439683586024,
+      "grad_norm": 6.977563717453175,
+      "learning_rate": 3.9473684210526315e-06,
+      "loss": 1.7599,
+      "step": 19
+    },
+    {
+      "epoch": 0.026367831245880026,
+      "grad_norm": 5.449166141636689,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 1.7067,
+      "step": 20
+    },
+    {
+      "epoch": 0.027686222808174028,
+      "grad_norm": 4.5079724238509,
+      "learning_rate": 4.3859649122807014e-06,
+      "loss": 1.6136,
+      "step": 21
+    },
+    {
+      "epoch": 0.02900461437046803,
+      "grad_norm": 3.4832764289049427,
+      "learning_rate": 4.605263157894737e-06,
+      "loss": 1.519,
+      "step": 22
+    },
+    {
+      "epoch": 0.03032300593276203,
+      "grad_norm": 2.8100903737826477,
+      "learning_rate": 4.824561403508772e-06,
+      "loss": 1.4868,
+      "step": 23
+    },
+    {
+      "epoch": 0.03164139749505603,
+      "grad_norm": 2.371954189541486,
+      "learning_rate": 5.043859649122807e-06,
+      "loss": 1.3671,
+      "step": 24
+    },
+    {
+      "epoch": 0.03295978905735003,
+      "grad_norm": 1.646860836065967,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 1.2102,
+      "step": 25
+    },
+    {
+      "epoch": 0.034278180619644036,
+      "grad_norm": 10.726308831094054,
+      "learning_rate": 5.482456140350877e-06,
+      "loss": 1.2526,
+      "step": 26
+    },
+    {
+      "epoch": 0.035596572181938034,
+      "grad_norm": 14.336715923179828,
+      "learning_rate": 5.701754385964912e-06,
+      "loss": 1.2459,
+      "step": 27
+    },
+    {
+      "epoch": 0.03691496374423204,
+      "grad_norm": 1.7776645392838684,
+      "learning_rate": 5.921052631578948e-06,
+      "loss": 1.2479,
+      "step": 28
+    },
+    {
+      "epoch": 0.03823335530652604,
+      "grad_norm": 1.3045854738024645,
+      "learning_rate": 6.140350877192982e-06,
+      "loss": 1.1512,
+      "step": 29
+    },
+    {
+      "epoch": 0.03955174686882004,
+      "grad_norm": 1.0278540664372422,
+      "learning_rate": 6.3596491228070184e-06,
+      "loss": 1.1011,
+      "step": 30
+    },
+    {
+      "epoch": 0.04087013843111404,
+      "grad_norm": 0.9434958429229233,
+      "learning_rate": 6.578947368421053e-06,
+      "loss": 1.138,
+      "step": 31
+    },
+    {
+      "epoch": 0.04218852999340804,
+      "grad_norm": 0.8415859303115572,
+      "learning_rate": 6.798245614035088e-06,
+      "loss": 1.0371,
+      "step": 32
+    },
+    {
+      "epoch": 0.043506921555702044,
+      "grad_norm": 0.80256278699881,
+      "learning_rate": 7.017543859649123e-06,
+      "loss": 1.0294,
+      "step": 33
+    },
+    {
+      "epoch": 0.04482531311799604,
+      "grad_norm": 0.7830554844922408,
+      "learning_rate": 7.236842105263158e-06,
+      "loss": 1.0712,
+      "step": 34
+    },
+    {
+      "epoch": 0.04614370468029005,
+      "grad_norm": 0.7702876033679092,
+      "learning_rate": 7.456140350877193e-06,
+      "loss": 1.002,
+      "step": 35
+    },
+    {
+      "epoch": 0.047462096242584045,
+      "grad_norm": 0.6773421769456159,
+      "learning_rate": 7.675438596491228e-06,
+      "loss": 1.0493,
+      "step": 36
+    },
+    {
+      "epoch": 0.04878048780487805,
+      "grad_norm": 0.6322340231808778,
+      "learning_rate": 7.894736842105263e-06,
+      "loss": 1.003,
+      "step": 37
+    },
+    {
+      "epoch": 0.05009887936717205,
+      "grad_norm": 0.6446056910237059,
+      "learning_rate": 8.1140350877193e-06,
+      "loss": 0.9894,
+      "step": 38
+    },
+    {
+      "epoch": 0.051417270929466054,
+      "grad_norm": 0.6186145777928085,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.9969,
+      "step": 39
+    },
+    {
+      "epoch": 0.05273566249176005,
+      "grad_norm": 0.5896139376675595,
+      "learning_rate": 8.552631578947368e-06,
+      "loss": 0.9269,
+      "step": 40
+    },
+    {
+      "epoch": 0.05405405405405406,
+      "grad_norm": 0.5636159709768188,
+      "learning_rate": 8.771929824561403e-06,
+      "loss": 1.0484,
+      "step": 41
+    },
+    {
+      "epoch": 0.055372445616348055,
+      "grad_norm": 0.48237357902187306,
+      "learning_rate": 8.991228070175439e-06,
+      "loss": 0.9732,
+      "step": 42
+    },
+    {
+      "epoch": 0.05669083717864205,
+      "grad_norm": 0.47605040759333966,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 1.0071,
+      "step": 43
+    },
+    {
+      "epoch": 0.05800922874093606,
+      "grad_norm": 0.4854589254214662,
+      "learning_rate": 9.42982456140351e-06,
+      "loss": 0.9632,
+      "step": 44
+    },
+    {
+      "epoch": 0.05932762030323006,
+      "grad_norm": 0.5041771749343741,
+      "learning_rate": 9.649122807017545e-06,
+      "loss": 0.9361,
+      "step": 45
+    },
+    {
+      "epoch": 0.06064601186552406,
+      "grad_norm": 0.5051788446713545,
+      "learning_rate": 9.868421052631579e-06,
+      "loss": 0.9346,
+      "step": 46
+    },
+    {
+      "epoch": 0.06196440342781806,
+      "grad_norm": 0.44434185446969715,
+      "learning_rate": 1.0087719298245614e-05,
+      "loss": 0.8725,
+      "step": 47
+    },
+    {
+      "epoch": 0.06328279499011207,
+      "grad_norm": 0.3667243214721247,
+      "learning_rate": 1.030701754385965e-05,
+      "loss": 0.912,
+      "step": 48
+    },
+    {
+      "epoch": 0.06460118655240607,
+      "grad_norm": 0.36415699241774235,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 0.8648,
+      "step": 49
+    },
+    {
+      "epoch": 0.06591957811470006,
+      "grad_norm": 0.4139924774088768,
+      "learning_rate": 1.074561403508772e-05,
+      "loss": 0.8764,
+      "step": 50
+    },
+    {
+      "epoch": 0.06723796967699407,
+      "grad_norm": 0.4258434085232776,
+      "learning_rate": 1.0964912280701754e-05,
+      "loss": 0.8451,
+      "step": 51
+    },
+    {
+      "epoch": 0.06855636123928807,
+      "grad_norm": 0.34637780485998043,
+      "learning_rate": 1.118421052631579e-05,
+      "loss": 0.8584,
+      "step": 52
+    },
+    {
+      "epoch": 0.06987475280158208,
+      "grad_norm": 0.3494801564321333,
+      "learning_rate": 1.1403508771929824e-05,
+      "loss": 0.9093,
+      "step": 53
+    },
+    {
+      "epoch": 0.07119314436387607,
+      "grad_norm": 0.30449136854875564,
+      "learning_rate": 1.162280701754386e-05,
+      "loss": 0.8164,
+      "step": 54
+    },
+    {
+      "epoch": 0.07251153592617007,
+      "grad_norm": 0.33318893460411936,
+      "learning_rate": 1.1842105263157895e-05,
+      "loss": 0.8611,
+      "step": 55
+    },
+    {
+      "epoch": 0.07382992748846408,
+      "grad_norm": 0.3619090495716855,
+      "learning_rate": 1.206140350877193e-05,
+      "loss": 0.8217,
+      "step": 56
+    },
+    {
+      "epoch": 0.07514831905075807,
+      "grad_norm": 0.3046841203096698,
+      "learning_rate": 1.2280701754385964e-05,
+      "loss": 0.7681,
+      "step": 57
+    },
+    {
+      "epoch": 0.07646671061305207,
+      "grad_norm": 0.3139391061176445,
+      "learning_rate": 1.25e-05,
+      "loss": 0.8443,
+      "step": 58
+    },
+    {
+      "epoch": 0.07778510217534608,
+      "grad_norm": 0.2958733736097859,
+      "learning_rate": 1.2719298245614037e-05,
+      "loss": 0.8374,
+      "step": 59
+    },
+    {
+      "epoch": 0.07910349373764008,
+      "grad_norm": 0.3288606161970216,
+      "learning_rate": 1.2938596491228071e-05,
+      "loss": 0.8543,
+      "step": 60
+    },
+    {
+      "epoch": 0.08042188529993408,
+      "grad_norm": 0.2850484662362131,
+      "learning_rate": 1.3157894736842106e-05,
+      "loss": 0.8166,
+      "step": 61
+    },
+    {
+      "epoch": 0.08174027686222808,
+      "grad_norm": 0.27750998498933777,
+      "learning_rate": 1.337719298245614e-05,
+      "loss": 0.8246,
+      "step": 62
+    },
+    {
+      "epoch": 0.08305866842452209,
+      "grad_norm": 0.2933158639357168,
+      "learning_rate": 1.3596491228070177e-05,
+      "loss": 0.8886,
+      "step": 63
+    },
+    {
+      "epoch": 0.08437705998681608,
+      "grad_norm": 0.30903699440452614,
+      "learning_rate": 1.3815789473684213e-05,
+      "loss": 0.7879,
+      "step": 64
+    },
+    {
+      "epoch": 0.08569545154911008,
+      "grad_norm": 0.28108524487440917,
+      "learning_rate": 1.4035087719298246e-05,
+      "loss": 0.7815,
+      "step": 65
+    },
+    {
+      "epoch": 0.08701384311140409,
+      "grad_norm": 0.2873884366248471,
+      "learning_rate": 1.425438596491228e-05,
+      "loss": 0.7908,
+      "step": 66
+    },
+    {
+      "epoch": 0.08833223467369809,
+      "grad_norm": 0.2796383587009447,
+      "learning_rate": 1.4473684210526317e-05,
+      "loss": 0.7616,
+      "step": 67
+    },
+    {
+      "epoch": 0.08965062623599208,
+      "grad_norm": 0.2737275439298298,
+      "learning_rate": 1.4692982456140353e-05,
+      "loss": 0.8174,
+      "step": 68
+    },
+    {
+      "epoch": 0.09096901779828609,
+      "grad_norm": 0.29050853142796473,
+      "learning_rate": 1.4912280701754386e-05,
+      "loss": 0.8243,
+      "step": 69
+    },
+    {
+      "epoch": 0.0922874093605801,
+      "grad_norm": 0.2566322956472508,
+      "learning_rate": 1.5131578947368422e-05,
+      "loss": 0.8283,
+      "step": 70
+    },
+    {
+      "epoch": 0.0936058009228741,
+      "grad_norm": 0.26247896091259876,
+      "learning_rate": 1.5350877192982457e-05,
+      "loss": 0.7806,
+      "step": 71
+    },
+    {
+      "epoch": 0.09492419248516809,
+      "grad_norm": 0.2652893104885863,
+      "learning_rate": 1.5570175438596493e-05,
+      "loss": 0.7658,
+      "step": 72
+    },
+    {
+      "epoch": 0.0962425840474621,
+      "grad_norm": 0.27823159495633376,
+      "learning_rate": 1.5789473684210526e-05,
+      "loss": 0.7991,
+      "step": 73
+    },
+    {
+      "epoch": 0.0975609756097561,
+      "grad_norm": 0.2645532722313751,
+      "learning_rate": 1.6008771929824562e-05,
+      "loss": 0.7789,
+      "step": 74
+    },
+    {
+      "epoch": 0.09887936717205009,
+      "grad_norm": 0.26779041018213473,
+      "learning_rate": 1.62280701754386e-05,
+      "loss": 0.7923,
+      "step": 75
+    },
+    {
+      "epoch": 0.1001977587343441,
+      "grad_norm": 0.2772430918345152,
+      "learning_rate": 1.6447368421052635e-05,
+      "loss": 0.8018,
+      "step": 76
+    },
+    {
+      "epoch": 0.1015161502966381,
+      "grad_norm": 0.30553134960328576,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.781,
+      "step": 77
+    },
+    {
+      "epoch": 0.10283454185893211,
+      "grad_norm": 0.2664983732334821,
+      "learning_rate": 1.6885964912280704e-05,
+      "loss": 0.7319,
+      "step": 78
+    },
+    {
+      "epoch": 0.1041529334212261,
+      "grad_norm": 0.3043163465655231,
+      "learning_rate": 1.7105263157894737e-05,
+      "loss": 0.8196,
+      "step": 79
+    },
+    {
+      "epoch": 0.1054713249835201,
+      "grad_norm": 0.3303164115908353,
+      "learning_rate": 1.7324561403508773e-05,
+      "loss": 0.7623,
+      "step": 80
+    },
+    {
+      "epoch": 0.10678971654581411,
+      "grad_norm": 0.2791448000341634,
+      "learning_rate": 1.7543859649122806e-05,
+      "loss": 0.78,
+      "step": 81
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 0.3411525360938291,
+      "learning_rate": 1.7763157894736842e-05,
+      "loss": 0.7624,
+      "step": 82
+    },
+    {
+      "epoch": 0.1094264996704021,
+      "grad_norm": 0.2929421028107074,
+      "learning_rate": 1.7982456140350878e-05,
+      "loss": 0.7694,
+      "step": 83
+    },
+    {
+      "epoch": 0.11074489123269611,
+      "grad_norm": 0.2596128950556148,
+      "learning_rate": 1.8201754385964914e-05,
+      "loss": 0.7467,
+      "step": 84
+    },
+    {
+      "epoch": 0.11206328279499012,
+      "grad_norm": 0.28993343568654295,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 0.7629,
+      "step": 85
+    },
+    {
+      "epoch": 0.1133816743572841,
+      "grad_norm": 0.26626515034039433,
+      "learning_rate": 1.8640350877192984e-05,
+      "loss": 0.7801,
+      "step": 86
+    },
+    {
+      "epoch": 0.11470006591957811,
+      "grad_norm": 0.2676331602265469,
+      "learning_rate": 1.885964912280702e-05,
+      "loss": 0.7553,
+      "step": 87
+    },
+    {
+      "epoch": 0.11601845748187212,
+      "grad_norm": 0.2864191366265554,
+      "learning_rate": 1.9078947368421056e-05,
+      "loss": 0.7935,
+      "step": 88
+    },
+    {
+      "epoch": 0.11733684904416612,
+      "grad_norm": 0.24953342540815326,
+      "learning_rate": 1.929824561403509e-05,
+      "loss": 0.7462,
+      "step": 89
+    },
+    {
+      "epoch": 0.11865524060646011,
+      "grad_norm": 0.26204895676021545,
+      "learning_rate": 1.9517543859649122e-05,
+      "loss": 0.7309,
+      "step": 90
+    },
+    {
+      "epoch": 0.11997363216875412,
+      "grad_norm": 0.2481497520434085,
+      "learning_rate": 1.9736842105263158e-05,
+      "loss": 0.7554,
+      "step": 91
+    },
+    {
+      "epoch": 0.12129202373104812,
+      "grad_norm": 0.2670427820880772,
+      "learning_rate": 1.9956140350877194e-05,
+      "loss": 0.7661,
+      "step": 92
+    },
+    {
+      "epoch": 0.12261041529334213,
+      "grad_norm": 0.2612350851954564,
+      "learning_rate": 2.0175438596491227e-05,
+      "loss": 0.7251,
+      "step": 93
+    },
+    {
+      "epoch": 0.12392880685563612,
+      "grad_norm": 0.2566275203465054,
+      "learning_rate": 2.0394736842105264e-05,
+      "loss": 0.7467,
+      "step": 94
+    },
+    {
+      "epoch": 0.12524719841793014,
+      "grad_norm": 0.2704741489035136,
+      "learning_rate": 2.06140350877193e-05,
+      "loss": 0.7509,
+      "step": 95
+    },
+    {
+      "epoch": 0.12656558998022413,
+      "grad_norm": 0.2784542834393439,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.7557,
+      "step": 96
+    },
+    {
+      "epoch": 0.12788398154251812,
+      "grad_norm": 0.2595732878666499,
+      "learning_rate": 2.105263157894737e-05,
+      "loss": 0.7507,
+      "step": 97
+    },
+    {
+      "epoch": 0.12920237310481214,
+      "grad_norm": 0.3071708817711585,
+      "learning_rate": 2.1271929824561405e-05,
+      "loss": 0.7676,
+      "step": 98
+    },
+    {
+      "epoch": 0.13052076466710613,
+      "grad_norm": 0.31415868331964025,
+      "learning_rate": 2.149122807017544e-05,
+      "loss": 0.7908,
+      "step": 99
+    },
+    {
+      "epoch": 0.13183915622940012,
+      "grad_norm": 0.2848712525734455,
+      "learning_rate": 2.1710526315789474e-05,
+      "loss": 0.7055,
+      "step": 100
+    },
+    {
+      "epoch": 0.13315754779169414,
+      "grad_norm": 0.28064730711349845,
+      "learning_rate": 2.1929824561403507e-05,
+      "loss": 0.7529,
+      "step": 101
+    },
+    {
+      "epoch": 0.13447593935398813,
+      "grad_norm": 0.30303191082429065,
+      "learning_rate": 2.2149122807017543e-05,
+      "loss": 0.7592,
+      "step": 102
+    },
+    {
+      "epoch": 0.13579433091628212,
+      "grad_norm": 0.26966642906552274,
+      "learning_rate": 2.236842105263158e-05,
+      "loss": 0.7557,
+      "step": 103
+    },
+    {
+      "epoch": 0.13711272247857614,
+      "grad_norm": 0.2826858419201188,
+      "learning_rate": 2.2587719298245616e-05,
+      "loss": 0.7847,
+      "step": 104
+    },
+    {
+      "epoch": 0.13843111404087013,
+      "grad_norm": 0.3315473771209985,
+      "learning_rate": 2.280701754385965e-05,
+      "loss": 0.7512,
+      "step": 105
+    },
+    {
+      "epoch": 0.13974950560316415,
+      "grad_norm": 0.3116243798638302,
+      "learning_rate": 2.3026315789473685e-05,
+      "loss": 0.7331,
+      "step": 106
+    },
+    {
+      "epoch": 0.14106789716545814,
+      "grad_norm": 0.3038980439145983,
+      "learning_rate": 2.324561403508772e-05,
+      "loss": 0.7647,
+      "step": 107
+    },
+    {
+      "epoch": 0.14238628872775214,
+      "grad_norm": 0.2822963245428743,
+      "learning_rate": 2.3464912280701758e-05,
+      "loss": 0.7079,
+      "step": 108
+    },
+    {
+      "epoch": 0.14370468029004616,
+      "grad_norm": 0.3344110537993316,
+      "learning_rate": 2.368421052631579e-05,
+      "loss": 0.7086,
+      "step": 109
+    },
+    {
+      "epoch": 0.14502307185234015,
+      "grad_norm": 0.28446147130288724,
+      "learning_rate": 2.3903508771929827e-05,
+      "loss": 0.7297,
+      "step": 110
+    },
+    {
+      "epoch": 0.14634146341463414,
+      "grad_norm": 0.28903954325158066,
+      "learning_rate": 2.412280701754386e-05,
+      "loss": 0.7001,
+      "step": 111
+    },
+    {
+      "epoch": 0.14765985497692816,
+      "grad_norm": 0.28877659477718887,
+      "learning_rate": 2.4342105263157896e-05,
+      "loss": 0.6864,
+      "step": 112
+    },
+    {
+      "epoch": 0.14897824653922215,
+      "grad_norm": 0.28982759393694424,
+      "learning_rate": 2.456140350877193e-05,
+      "loss": 0.7335,
+      "step": 113
+    },
+    {
+      "epoch": 0.15029663810151614,
+      "grad_norm": 0.33287077682562793,
+      "learning_rate": 2.4780701754385965e-05,
+      "loss": 0.741,
+      "step": 114
+    },
+    {
+      "epoch": 0.15161502966381016,
+      "grad_norm": 0.29591262938379675,
+      "learning_rate": 2.5e-05,
+      "loss": 0.7704,
+      "step": 115
+    },
+    {
+      "epoch": 0.15293342122610415,
+      "grad_norm": 0.3322716853285586,
+      "learning_rate": 2.5219298245614037e-05,
+      "loss": 0.7471,
+      "step": 116
+    },
+    {
+      "epoch": 0.15425181278839814,
+      "grad_norm": 0.314898709201311,
+      "learning_rate": 2.5438596491228074e-05,
+      "loss": 0.6898,
+      "step": 117
+    },
+    {
+      "epoch": 0.15557020435069216,
+      "grad_norm": 0.3012940295191572,
+      "learning_rate": 2.565789473684211e-05,
+      "loss": 0.6659,
+      "step": 118
+    },
+    {
+      "epoch": 0.15688859591298615,
+      "grad_norm": 0.5072769463753124,
+      "learning_rate": 2.5877192982456143e-05,
+      "loss": 0.7172,
+      "step": 119
+    },
+    {
+      "epoch": 0.15820698747528017,
+      "grad_norm": 0.35026232918974776,
+      "learning_rate": 2.6096491228070176e-05,
+      "loss": 0.7278,
+      "step": 120
+    },
+    {
+      "epoch": 0.15952537903757416,
+      "grad_norm": 0.2772696338872679,
+      "learning_rate": 2.6315789473684212e-05,
+      "loss": 0.7223,
+      "step": 121
+    },
+    {
+      "epoch": 0.16084377059986815,
+      "grad_norm": 0.336813473792303,
+      "learning_rate": 2.6535087719298245e-05,
+      "loss": 0.6939,
+      "step": 122
+    },
+    {
+      "epoch": 0.16216216216216217,
+      "grad_norm": 0.3482870571526658,
+      "learning_rate": 2.675438596491228e-05,
+      "loss": 0.7061,
+      "step": 123
+    },
+    {
+      "epoch": 0.16348055372445616,
+      "grad_norm": 0.35567250163928854,
+      "learning_rate": 2.6973684210526317e-05,
+      "loss": 0.6892,
+      "step": 124
+    },
+    {
+      "epoch": 0.16479894528675015,
+      "grad_norm": 0.2870781081028409,
+      "learning_rate": 2.7192982456140354e-05,
+      "loss": 0.7363,
+      "step": 125
+    },
+    {
+      "epoch": 0.16611733684904417,
+      "grad_norm": 0.31657686449047395,
+      "learning_rate": 2.741228070175439e-05,
+      "loss": 0.7318,
+      "step": 126
+    },
+    {
+      "epoch": 0.16743572841133816,
+      "grad_norm": 0.264005635258038,
+      "learning_rate": 2.7631578947368426e-05,
+      "loss": 0.7097,
+      "step": 127
+    },
+    {
+      "epoch": 0.16875411997363216,
+      "grad_norm": 0.31908566652881637,
+      "learning_rate": 2.7850877192982456e-05,
+      "loss": 0.7415,
+      "step": 128
+    },
+    {
+      "epoch": 0.17007251153592617,
+      "grad_norm": 0.30079890757201655,
+      "learning_rate": 2.8070175438596492e-05,
+      "loss": 0.7017,
+      "step": 129
+    },
+    {
+      "epoch": 0.17139090309822017,
+      "grad_norm": 0.30675084063253305,
+      "learning_rate": 2.8289473684210528e-05,
+      "loss": 0.7111,
+      "step": 130
+    },
+    {
+      "epoch": 0.17270929466051418,
+      "grad_norm": 0.28710431879456716,
+      "learning_rate": 2.850877192982456e-05,
+      "loss": 0.6629,
+      "step": 131
+    },
+    {
+      "epoch": 0.17402768622280818,
+      "grad_norm": 0.3063935953018529,
+      "learning_rate": 2.8728070175438597e-05,
+      "loss": 0.6755,
+      "step": 132
+    },
+    {
+      "epoch": 0.17534607778510217,
+      "grad_norm": 0.31312576456861385,
+      "learning_rate": 2.8947368421052634e-05,
+      "loss": 0.7233,
+      "step": 133
+    },
+    {
+      "epoch": 0.17666446934739619,
+      "grad_norm": 0.2862750708539955,
+      "learning_rate": 2.916666666666667e-05,
+      "loss": 0.6783,
+      "step": 134
+    },
+    {
+      "epoch": 0.17798286090969018,
+      "grad_norm": 0.3396704488409864,
+      "learning_rate": 2.9385964912280706e-05,
+      "loss": 0.7028,
+      "step": 135
+    },
+    {
+      "epoch": 0.17930125247198417,
+      "grad_norm": 0.34312913305988835,
+      "learning_rate": 2.9605263157894735e-05,
+      "loss": 0.7324,
+      "step": 136
+    },
+    {
+      "epoch": 0.1806196440342782,
+      "grad_norm": 0.35208504672050817,
+      "learning_rate": 2.9824561403508772e-05,
+      "loss": 0.7214,
+      "step": 137
+    },
+    {
+      "epoch": 0.18193803559657218,
+      "grad_norm": 0.31624061227645844,
+      "learning_rate": 3.0043859649122808e-05,
+      "loss": 0.7172,
+      "step": 138
+    },
+    {
+      "epoch": 0.18325642715886617,
+      "grad_norm": 0.35176040300659606,
+      "learning_rate": 3.0263157894736844e-05,
+      "loss": 0.7257,
+      "step": 139
+    },
+    {
+      "epoch": 0.1845748187211602,
+      "grad_norm": 0.3001080417890375,
+      "learning_rate": 3.048245614035088e-05,
+      "loss": 0.6637,
+      "step": 140
+    },
+    {
+      "epoch": 0.18589321028345418,
+      "grad_norm": 0.3008060591817919,
+      "learning_rate": 3.0701754385964913e-05,
+      "loss": 0.6985,
+      "step": 141
+    },
+    {
+      "epoch": 0.1872116018457482,
+      "grad_norm": 0.30851029103879685,
+      "learning_rate": 3.092105263157895e-05,
+      "loss": 0.6987,
+      "step": 142
+    },
+    {
+      "epoch": 0.1885299934080422,
+      "grad_norm": 0.2791199480873937,
+      "learning_rate": 3.1140350877192986e-05,
+      "loss": 0.7428,
+      "step": 143
+    },
+    {
+      "epoch": 0.18984838497033618,
+      "grad_norm": 0.2904327709561921,
+      "learning_rate": 3.1359649122807015e-05,
+      "loss": 0.6625,
+      "step": 144
+    },
+    {
+      "epoch": 0.1911667765326302,
+      "grad_norm": 0.30488266671437875,
+      "learning_rate": 3.157894736842105e-05,
+      "loss": 0.7153,
+      "step": 145
+    },
+    {
+      "epoch": 0.1924851680949242,
+      "grad_norm": 0.29797734610936893,
+      "learning_rate": 3.179824561403509e-05,
+      "loss": 0.6613,
+      "step": 146
+    },
+    {
+      "epoch": 0.19380355965721818,
+      "grad_norm": 0.33121627643851576,
+      "learning_rate": 3.2017543859649124e-05,
+      "loss": 0.6487,
+      "step": 147
+    },
+    {
+      "epoch": 0.1951219512195122,
+      "grad_norm": 0.293768032488189,
+      "learning_rate": 3.223684210526316e-05,
+      "loss": 0.7182,
+      "step": 148
+    },
+    {
+      "epoch": 0.1964403427818062,
+      "grad_norm": 0.3789278668959811,
+      "learning_rate": 3.24561403508772e-05,
+      "loss": 0.7097,
+      "step": 149
+    },
+    {
+      "epoch": 0.19775873434410018,
+      "grad_norm": 0.320769206921991,
+      "learning_rate": 3.267543859649123e-05,
+      "loss": 0.7172,
+      "step": 150
+    },
+    {
+      "epoch": 0.1990771259063942,
+      "grad_norm": 0.371376045308909,
+      "learning_rate": 3.289473684210527e-05,
+      "loss": 0.732,
+      "step": 151
+    },
+    {
+      "epoch": 0.2003955174686882,
+      "grad_norm": 0.29574716061069645,
+      "learning_rate": 3.31140350877193e-05,
+      "loss": 0.6822,
+      "step": 152
+    },
+    {
+      "epoch": 0.2017139090309822,
+      "grad_norm": 0.38552453403821096,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7199,
+      "step": 153
+    },
+    {
+      "epoch": 0.2030323005932762,
+      "grad_norm": 0.3181086449374998,
+      "learning_rate": 3.355263157894737e-05,
+      "loss": 0.6381,
+      "step": 154
+    },
+    {
+      "epoch": 0.2043506921555702,
+      "grad_norm": 0.3170409384833641,
+      "learning_rate": 3.377192982456141e-05,
+      "loss": 0.6913,
+      "step": 155
+    },
+    {
+      "epoch": 0.20566908371786422,
+      "grad_norm": 0.33044934169764306,
+      "learning_rate": 3.3991228070175444e-05,
+      "loss": 0.6349,
+      "step": 156
+    },
+    {
+      "epoch": 0.2069874752801582,
+      "grad_norm": 0.3073736986293781,
+      "learning_rate": 3.421052631578947e-05,
+      "loss": 0.6389,
+      "step": 157
+    },
+    {
+      "epoch": 0.2083058668424522,
+      "grad_norm": 0.38259863027798596,
+      "learning_rate": 3.442982456140351e-05,
+      "loss": 0.6886,
+      "step": 158
+    },
+    {
+      "epoch": 0.20962425840474622,
+      "grad_norm": 0.4512852528449524,
+      "learning_rate": 3.4649122807017546e-05,
+      "loss": 0.6689,
+      "step": 159
+    },
+    {
+      "epoch": 0.2109426499670402,
+      "grad_norm": 0.32443826810550475,
+      "learning_rate": 3.4868421052631575e-05,
+      "loss": 0.6879,
+      "step": 160
+    },
+    {
+      "epoch": 0.2122610415293342,
+      "grad_norm": 0.4127416465918417,
+      "learning_rate": 3.508771929824561e-05,
+      "loss": 0.6415,
+      "step": 161
+    },
+    {
+      "epoch": 0.21357943309162822,
+      "grad_norm": 0.3281659547133054,
+      "learning_rate": 3.530701754385965e-05,
+      "loss": 0.662,
+      "step": 162
+    },
+    {
+      "epoch": 0.2148978246539222,
+      "grad_norm": 0.3765753883258304,
+      "learning_rate": 3.5526315789473684e-05,
+      "loss": 0.6539,
+      "step": 163
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 0.3935555713964011,
+      "learning_rate": 3.574561403508772e-05,
+      "loss": 0.6967,
+      "step": 164
+    },
+    {
+      "epoch": 0.21753460777851022,
+      "grad_norm": 0.3718670008626901,
+      "learning_rate": 3.5964912280701756e-05,
+      "loss": 0.6787,
+      "step": 165
+    },
+    {
+      "epoch": 0.2188529993408042,
+      "grad_norm": 0.33157307519941503,
+      "learning_rate": 3.618421052631579e-05,
+      "loss": 0.6896,
+      "step": 166
+    },
+    {
+      "epoch": 0.22017139090309823,
+      "grad_norm": 0.3607621082857418,
+      "learning_rate": 3.640350877192983e-05,
+      "loss": 0.6859,
+      "step": 167
+    },
+    {
+      "epoch": 0.22148978246539222,
+      "grad_norm": 0.31777525120993216,
+      "learning_rate": 3.662280701754386e-05,
+      "loss": 0.7103,
+      "step": 168
+    },
+    {
+      "epoch": 0.2228081740276862,
+      "grad_norm": 0.3450356914129178,
+      "learning_rate": 3.6842105263157895e-05,
+      "loss": 0.7071,
+      "step": 169
+    },
+    {
+      "epoch": 0.22412656558998023,
+      "grad_norm": 0.28635022573331004,
+      "learning_rate": 3.706140350877193e-05,
+      "loss": 0.6639,
+      "step": 170
+    },
+    {
+      "epoch": 0.22544495715227422,
+      "grad_norm": 0.3032290256115864,
+      "learning_rate": 3.728070175438597e-05,
+      "loss": 0.7201,
+      "step": 171
+    },
+    {
+      "epoch": 0.2267633487145682,
+      "grad_norm": 0.30825314045412605,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.6988,
+      "step": 172
+    },
+    {
+      "epoch": 0.22808174027686223,
+      "grad_norm": 0.2926820508113059,
+      "learning_rate": 3.771929824561404e-05,
+      "loss": 0.662,
+      "step": 173
+    },
+    {
+      "epoch": 0.22940013183915622,
+      "grad_norm": 0.3236284040629951,
+      "learning_rate": 3.7938596491228076e-05,
+      "loss": 0.6709,
+      "step": 174
+    },
+    {
+      "epoch": 0.23071852340145024,
+      "grad_norm": 0.35481661266442,
+      "learning_rate": 3.815789473684211e-05,
+      "loss": 0.6673,
+      "step": 175
+    },
+    {
+      "epoch": 0.23203691496374423,
+      "grad_norm": 0.409480842839928,
+      "learning_rate": 3.837719298245614e-05,
+      "loss": 0.7176,
+      "step": 176
+    },
+    {
+      "epoch": 0.23335530652603823,
+      "grad_norm": 0.32083237549024607,
+      "learning_rate": 3.859649122807018e-05,
+      "loss": 0.6963,
+      "step": 177
+    },
+    {
+      "epoch": 0.23467369808833224,
+      "grad_norm": 0.39214624315241386,
+      "learning_rate": 3.8815789473684214e-05,
+      "loss": 0.6846,
+      "step": 178
+    },
+    {
+      "epoch": 0.23599208965062624,
+      "grad_norm": 0.36493569097049877,
+      "learning_rate": 3.9035087719298244e-05,
+      "loss": 0.6545,
+      "step": 179
+    },
+    {
+      "epoch": 0.23731048121292023,
+      "grad_norm": 0.3881088061682472,
+      "learning_rate": 3.925438596491228e-05,
+      "loss": 0.6349,
+      "step": 180
+    },
+    {
+      "epoch": 0.23862887277521425,
+      "grad_norm": 0.3771037490508129,
+      "learning_rate": 3.9473684210526316e-05,
+      "loss": 0.6453,
+      "step": 181
+    },
+    {
+      "epoch": 0.23994726433750824,
+      "grad_norm": 0.3522056909155721,
+      "learning_rate": 3.969298245614035e-05,
+      "loss": 0.6642,
+      "step": 182
+    },
+    {
+      "epoch": 0.24126565589980223,
+      "grad_norm": 0.3684585531529165,
+      "learning_rate": 3.991228070175439e-05,
+      "loss": 0.6406,
+      "step": 183
+    },
+    {
+      "epoch": 0.24258404746209625,
+      "grad_norm": 0.3380669723498605,
+      "learning_rate": 4.0131578947368425e-05,
+      "loss": 0.6495,
+      "step": 184
+    },
+    {
+      "epoch": 0.24390243902439024,
+      "grad_norm": 0.3731107054415869,
+      "learning_rate": 4.0350877192982455e-05,
+      "loss": 0.6639,
+      "step": 185
+    },
+    {
+      "epoch": 0.24522083058668426,
+      "grad_norm": 0.390490016715261,
+      "learning_rate": 4.057017543859649e-05,
+      "loss": 0.6759,
+      "step": 186
+    },
+    {
+      "epoch": 0.24653922214897825,
+      "grad_norm": 0.3272806551311441,
+      "learning_rate": 4.078947368421053e-05,
+      "loss": 0.6621,
+      "step": 187
+    },
+    {
+      "epoch": 0.24785761371127224,
+      "grad_norm": 0.38046993249613237,
+      "learning_rate": 4.100877192982456e-05,
+      "loss": 0.6842,
+      "step": 188
+    },
+    {
+      "epoch": 0.24917600527356626,
+      "grad_norm": 0.35983172042880973,
+      "learning_rate": 4.12280701754386e-05,
+      "loss": 0.6485,
+      "step": 189
+    },
+    {
+      "epoch": 0.2504943968358603,
+      "grad_norm": 0.37385854689668807,
+      "learning_rate": 4.1447368421052636e-05,
+      "loss": 0.6733,
+      "step": 190
+    },
+    {
+      "epoch": 0.25181278839815424,
+      "grad_norm": 0.3588531615513882,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.6506,
+      "step": 191
+    },
+    {
+      "epoch": 0.25313117996044826,
+      "grad_norm": 0.34081338072549205,
+      "learning_rate": 4.188596491228071e-05,
+      "loss": 0.6618,
+      "step": 192
+    },
+    {
+      "epoch": 0.2544495715227423,
+      "grad_norm": 0.3575428492908049,
+      "learning_rate": 4.210526315789474e-05,
+      "loss": 0.6927,
+      "step": 193
+    },
+    {
+      "epoch": 0.25576796308503624,
+      "grad_norm": 0.37302640113732083,
+      "learning_rate": 4.2324561403508774e-05,
+      "loss": 0.6347,
+      "step": 194
+    },
+    {
+      "epoch": 0.25708635464733026,
+      "grad_norm": 0.3484666967368402,
+      "learning_rate": 4.254385964912281e-05,
+      "loss": 0.6263,
+      "step": 195
+    },
+    {
+      "epoch": 0.2584047462096243,
+      "grad_norm": 0.3975034819231687,
+      "learning_rate": 4.2763157894736847e-05,
+      "loss": 0.6409,
+      "step": 196
+    },
+    {
+      "epoch": 0.25972313777191824,
+      "grad_norm": 0.30834451103423716,
+      "learning_rate": 4.298245614035088e-05,
+      "loss": 0.6625,
+      "step": 197
+    },
+    {
+      "epoch": 0.26104152933421226,
+      "grad_norm": 0.45594057338387944,
+      "learning_rate": 4.320175438596491e-05,
+      "loss": 0.6361,
+      "step": 198
+    },
+    {
+      "epoch": 0.2623599208965063,
+      "grad_norm": 0.3927356797335999,
+      "learning_rate": 4.342105263157895e-05,
+      "loss": 0.6959,
+      "step": 199
+    },
+    {
+      "epoch": 0.26367831245880025,
+      "grad_norm": 0.4457914720499941,
+      "learning_rate": 4.3640350877192985e-05,
+      "loss": 0.6791,
+      "step": 200
+    },
+    {
+      "epoch": 0.26499670402109426,
+      "grad_norm": 0.43321045717935375,
+      "learning_rate": 4.3859649122807014e-05,
+      "loss": 0.6448,
+      "step": 201
+    },
+    {
+      "epoch": 0.2663150955833883,
+      "grad_norm": 0.36405780308442487,
+      "learning_rate": 4.407894736842105e-05,
+      "loss": 0.6562,
+      "step": 202
+    },
+    {
+      "epoch": 0.26763348714568225,
+      "grad_norm": 0.4317233908101476,
+      "learning_rate": 4.429824561403509e-05,
+      "loss": 0.6919,
+      "step": 203
+    },
+    {
+      "epoch": 0.26895187870797627,
+      "grad_norm": 0.41242091693454097,
+      "learning_rate": 4.451754385964912e-05,
+      "loss": 0.7028,
+      "step": 204
+    },
+    {
+      "epoch": 0.2702702702702703,
+      "grad_norm": 0.377762382767973,
+      "learning_rate": 4.473684210526316e-05,
+      "loss": 0.6481,
+      "step": 205
+    },
+    {
+      "epoch": 0.27158866183256425,
+      "grad_norm": 0.3784630050802653,
+      "learning_rate": 4.4956140350877196e-05,
+      "loss": 0.661,
+      "step": 206
+    },
+    {
+      "epoch": 0.27290705339485827,
+      "grad_norm": 0.35625027908688556,
+      "learning_rate": 4.517543859649123e-05,
+      "loss": 0.655,
+      "step": 207
+    },
+    {
+      "epoch": 0.2742254449571523,
+      "grad_norm": 0.45149809372130356,
+      "learning_rate": 4.539473684210527e-05,
+      "loss": 0.6688,
+      "step": 208
+    },
+    {
+      "epoch": 0.27554383651944625,
+      "grad_norm": 0.3652984402811423,
+      "learning_rate": 4.56140350877193e-05,
+      "loss": 0.6196,
+      "step": 209
+    },
+    {
+      "epoch": 0.27686222808174027,
+      "grad_norm": 0.3116903283264657,
+      "learning_rate": 4.5833333333333334e-05,
+      "loss": 0.6305,
+      "step": 210
+    },
+    {
+      "epoch": 0.2781806196440343,
+      "grad_norm": 0.3397050146291172,
+      "learning_rate": 4.605263157894737e-05,
+      "loss": 0.6768,
+      "step": 211
+    },
+    {
+      "epoch": 0.2794990112063283,
+      "grad_norm": 0.3282738260111823,
+      "learning_rate": 4.6271929824561406e-05,
+      "loss": 0.67,
+      "step": 212
+    },
+    {
+      "epoch": 0.28081740276862227,
+      "grad_norm": 0.34644478022881225,
+      "learning_rate": 4.649122807017544e-05,
+      "loss": 0.6892,
+      "step": 213
+    },
+    {
+      "epoch": 0.2821357943309163,
+      "grad_norm": 0.3427887768532096,
+      "learning_rate": 4.671052631578948e-05,
+      "loss": 0.6437,
+      "step": 214
+    },
+    {
+      "epoch": 0.2834541858932103,
+      "grad_norm": 0.30786868350900637,
+      "learning_rate": 4.6929824561403515e-05,
+      "loss": 0.6481,
+      "step": 215
+    },
+    {
+      "epoch": 0.28477257745550427,
+      "grad_norm": 0.3317024192595864,
+      "learning_rate": 4.714912280701755e-05,
+      "loss": 0.6601,
+      "step": 216
+    },
+    {
+      "epoch": 0.2860909690177983,
+      "grad_norm": 0.3245860895443998,
+      "learning_rate": 4.736842105263158e-05,
+      "loss": 0.6288,
+      "step": 217
+    },
+    {
+      "epoch": 0.2874093605800923,
+      "grad_norm": 0.33639000507748024,
+      "learning_rate": 4.758771929824562e-05,
+      "loss": 0.6981,
+      "step": 218
+    },
+    {
+      "epoch": 0.2887277521423863,
+      "grad_norm": 0.387188243754734,
+      "learning_rate": 4.780701754385965e-05,
+      "loss": 0.6809,
+      "step": 219
+    },
+    {
+      "epoch": 0.2900461437046803,
+      "grad_norm": 0.32191442622367583,
+      "learning_rate": 4.802631578947368e-05,
+      "loss": 0.642,
+      "step": 220
+    },
+    {
+      "epoch": 0.2913645352669743,
+      "grad_norm": 0.3712280340333288,
+      "learning_rate": 4.824561403508772e-05,
+      "loss": 0.6416,
+      "step": 221
+    },
+    {
+      "epoch": 0.2926829268292683,
+      "grad_norm": 0.33673789446967484,
+      "learning_rate": 4.8464912280701755e-05,
+      "loss": 0.6326,
+      "step": 222
+    },
+    {
+      "epoch": 0.2940013183915623,
+      "grad_norm": 0.3858528606638668,
+      "learning_rate": 4.868421052631579e-05,
+      "loss": 0.6724,
+      "step": 223
+    },
+    {
+      "epoch": 0.2953197099538563,
+      "grad_norm": 0.3218983392289052,
+      "learning_rate": 4.890350877192983e-05,
+      "loss": 0.6592,
+      "step": 224
+    },
+    {
+      "epoch": 0.2966381015161503,
+      "grad_norm": 0.4030239959994596,
+      "learning_rate": 4.912280701754386e-05,
+      "loss": 0.6459,
+      "step": 225
+    },
+    {
+      "epoch": 0.2979564930784443,
+      "grad_norm": 0.3436354584521706,
+      "learning_rate": 4.9342105263157894e-05,
+      "loss": 0.6453,
+      "step": 226
+    },
+    {
+      "epoch": 0.2992748846407383,
+      "grad_norm": 0.3922357886432537,
+      "learning_rate": 4.956140350877193e-05,
+      "loss": 0.6496,
+      "step": 227
+    },
+    {
+      "epoch": 0.3005932762030323,
+      "grad_norm": 0.3000705395466455,
+      "learning_rate": 4.9780701754385966e-05,
+      "loss": 0.6702,
+      "step": 228
+    },
+    {
+      "epoch": 0.3019116677653263,
+      "grad_norm": 0.35570863466250346,
+      "learning_rate": 5e-05,
+      "loss": 0.6534,
+      "step": 229
+    },
+    {
+      "epoch": 0.3032300593276203,
+      "grad_norm": 0.37054918468731984,
+      "learning_rate": 4.997556207233627e-05,
+      "loss": 0.6751,
+      "step": 230
+    },
+    {
+      "epoch": 0.3045484508899143,
+      "grad_norm": 0.34574085850512243,
+      "learning_rate": 4.995112414467253e-05,
+      "loss": 0.6547,
+      "step": 231
+    },
+    {
+      "epoch": 0.3058668424522083,
+      "grad_norm": 0.45257663852241703,
+      "learning_rate": 4.99266862170088e-05,
+      "loss": 0.6744,
+      "step": 232
+    },
+    {
+      "epoch": 0.3071852340145023,
+      "grad_norm": 0.39824848519352374,
+      "learning_rate": 4.990224828934507e-05,
+      "loss": 0.6722,
+      "step": 233
+    },
+    {
+      "epoch": 0.3085036255767963,
+      "grad_norm": 0.3918338722934089,
+      "learning_rate": 4.987781036168133e-05,
+      "loss": 0.6549,
+      "step": 234
+    },
+    {
+      "epoch": 0.3098220171390903,
+      "grad_norm": 0.3689363510913978,
+      "learning_rate": 4.98533724340176e-05,
+      "loss": 0.6321,
+      "step": 235
+    },
+    {
+      "epoch": 0.3111404087013843,
+      "grad_norm": 0.5358597181616104,
+      "learning_rate": 4.982893450635386e-05,
+      "loss": 0.6372,
+      "step": 236
+    },
+    {
+      "epoch": 0.31245880026367834,
+      "grad_norm": 0.30191895032886684,
+      "learning_rate": 4.980449657869013e-05,
+      "loss": 0.6498,
+      "step": 237
+    },
+    {
+      "epoch": 0.3137771918259723,
+      "grad_norm": 0.5424150864896571,
+      "learning_rate": 4.978005865102639e-05,
+      "loss": 0.6451,
+      "step": 238
+    },
+    {
+      "epoch": 0.3150955833882663,
+      "grad_norm": 0.47796635978214974,
+      "learning_rate": 4.9755620723362664e-05,
+      "loss": 0.7334,
+      "step": 239
+    },
+    {
+      "epoch": 0.31641397495056034,
+      "grad_norm": 0.4154301925440886,
+      "learning_rate": 4.973118279569893e-05,
+      "loss": 0.5773,
+      "step": 240
+    },
+    {
+      "epoch": 0.3177323665128543,
+      "grad_norm": 0.4839757586293214,
+      "learning_rate": 4.9706744868035194e-05,
+      "loss": 0.618,
+      "step": 241
+    },
+    {
+      "epoch": 0.3190507580751483,
+      "grad_norm": 0.39998537160269343,
+      "learning_rate": 4.968230694037146e-05,
+      "loss": 0.5687,
+      "step": 242
+    },
+    {
+      "epoch": 0.32036914963744234,
+      "grad_norm": 0.4230065850600772,
+      "learning_rate": 4.965786901270772e-05,
+      "loss": 0.655,
+      "step": 243
+    },
+    {
+      "epoch": 0.3216875411997363,
+      "grad_norm": 0.3947654772449055,
+      "learning_rate": 4.963343108504399e-05,
+      "loss": 0.6387,
+      "step": 244
+    },
+    {
+      "epoch": 0.3230059327620303,
+      "grad_norm": 0.46225246149402366,
+      "learning_rate": 4.960899315738026e-05,
+      "loss": 0.6359,
+      "step": 245
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 0.42156532794163354,
+      "learning_rate": 4.9584555229716524e-05,
+      "loss": 0.6483,
+      "step": 246
+    },
+    {
+      "epoch": 0.3256427158866183,
+      "grad_norm": 0.441903594033409,
+      "learning_rate": 4.956011730205279e-05,
+      "loss": 0.6222,
+      "step": 247
+    },
+    {
+      "epoch": 0.3269611074489123,
+      "grad_norm": 0.4144427958738356,
+      "learning_rate": 4.9535679374389054e-05,
+      "loss": 0.6214,
+      "step": 248
+    },
+    {
+      "epoch": 0.32827949901120634,
+      "grad_norm": 0.41761868001679375,
+      "learning_rate": 4.951124144672532e-05,
+      "loss": 0.6526,
+      "step": 249
+    },
+    {
+      "epoch": 0.3295978905735003,
+      "grad_norm": 0.3737469534739662,
+      "learning_rate": 4.9486803519061584e-05,
+      "loss": 0.6822,
+      "step": 250
+    },
+    {
+      "epoch": 0.3309162821357943,
+      "grad_norm": 0.31665076410555587,
+      "learning_rate": 4.9462365591397855e-05,
+      "loss": 0.6182,
+      "step": 251
+    },
+    {
+      "epoch": 0.33223467369808835,
+      "grad_norm": 0.453341003680116,
+      "learning_rate": 4.943792766373412e-05,
+      "loss": 0.6429,
+      "step": 252
+    },
+    {
+      "epoch": 0.3335530652603823,
+      "grad_norm": 0.32471269385768686,
+      "learning_rate": 4.9413489736070385e-05,
+      "loss": 0.6549,
+      "step": 253
+    },
+    {
+      "epoch": 0.3348714568226763,
+      "grad_norm": 0.39268429117058606,
+      "learning_rate": 4.938905180840665e-05,
+      "loss": 0.6565,
+      "step": 254
+    },
+    {
+      "epoch": 0.33618984838497035,
+      "grad_norm": 0.41743456240010696,
+      "learning_rate": 4.9364613880742914e-05,
+      "loss": 0.694,
+      "step": 255
+    },
+    {
+      "epoch": 0.3375082399472643,
+      "grad_norm": 0.3225047208532587,
+      "learning_rate": 4.934017595307918e-05,
+      "loss": 0.6189,
+      "step": 256
+    },
+    {
+      "epoch": 0.33882663150955833,
+      "grad_norm": 0.3927683333615166,
+      "learning_rate": 4.931573802541545e-05,
+      "loss": 0.5813,
+      "step": 257
+    },
+    {
+      "epoch": 0.34014502307185235,
+      "grad_norm": 0.3521363885770207,
+      "learning_rate": 4.9291300097751715e-05,
+      "loss": 0.65,
+      "step": 258
+    },
+    {
+      "epoch": 0.34146341463414637,
+      "grad_norm": 0.4015046684272929,
+      "learning_rate": 4.926686217008798e-05,
+      "loss": 0.624,
+      "step": 259
+    },
+    {
+      "epoch": 0.34278180619644033,
+      "grad_norm": 0.3362618076133002,
+      "learning_rate": 4.9242424242424245e-05,
+      "loss": 0.6576,
+      "step": 260
+    },
+    {
+      "epoch": 0.34410019775873435,
+      "grad_norm": 0.3609445243286373,
+      "learning_rate": 4.921798631476051e-05,
+      "loss": 0.6326,
+      "step": 261
+    },
+    {
+      "epoch": 0.34541858932102837,
+      "grad_norm": 0.3476615550655693,
+      "learning_rate": 4.9193548387096775e-05,
+      "loss": 0.6126,
+      "step": 262
+    },
+    {
+      "epoch": 0.34673698088332233,
+      "grad_norm": 0.3831031228584112,
+      "learning_rate": 4.9169110459433046e-05,
+      "loss": 0.6251,
+      "step": 263
+    },
+    {
+      "epoch": 0.34805537244561635,
+      "grad_norm": 0.33671595107748487,
+      "learning_rate": 4.914467253176931e-05,
+      "loss": 0.6933,
+      "step": 264
+    },
+    {
+      "epoch": 0.34937376400791037,
+      "grad_norm": 0.33404806376666996,
+      "learning_rate": 4.9120234604105576e-05,
+      "loss": 0.6603,
+      "step": 265
+    },
+    {
+      "epoch": 0.35069215557020433,
+      "grad_norm": 0.3657852166192865,
+      "learning_rate": 4.909579667644184e-05,
+      "loss": 0.6582,
+      "step": 266
+    },
+    {
+      "epoch": 0.35201054713249835,
+      "grad_norm": 0.34352476433011964,
+      "learning_rate": 4.9071358748778105e-05,
+      "loss": 0.6689,
+      "step": 267
+    },
+    {
+      "epoch": 0.35332893869479237,
+      "grad_norm": 0.34503692178402884,
+      "learning_rate": 4.904692082111437e-05,
+      "loss": 0.6186,
+      "step": 268
+    },
+    {
+      "epoch": 0.35464733025708634,
+      "grad_norm": 0.3583764765666181,
+      "learning_rate": 4.902248289345064e-05,
+      "loss": 0.6037,
+      "step": 269
+    },
+    {
+      "epoch": 0.35596572181938035,
+      "grad_norm": 0.3303984105427765,
+      "learning_rate": 4.899804496578691e-05,
+      "loss": 0.6395,
+      "step": 270
+    },
+    {
+      "epoch": 0.3572841133816744,
+      "grad_norm": 0.39130859497529386,
+      "learning_rate": 4.897360703812317e-05,
+      "loss": 0.621,
+      "step": 271
+    },
+    {
+      "epoch": 0.35860250494396834,
+      "grad_norm": 0.3273479678070148,
+      "learning_rate": 4.8949169110459436e-05,
+      "loss": 0.6769,
+      "step": 272
+    },
+    {
+      "epoch": 0.35992089650626236,
+      "grad_norm": 0.32671765024985316,
+      "learning_rate": 4.89247311827957e-05,
+      "loss": 0.656,
+      "step": 273
+    },
+    {
+      "epoch": 0.3612392880685564,
+      "grad_norm": 0.3076440047967864,
+      "learning_rate": 4.8900293255131966e-05,
+      "loss": 0.6608,
+      "step": 274
+    },
+    {
+      "epoch": 0.36255767963085034,
+      "grad_norm": 0.33296561851527123,
+      "learning_rate": 4.887585532746823e-05,
+      "loss": 0.6109,
+      "step": 275
+    },
+    {
+      "epoch": 0.36387607119314436,
+      "grad_norm": 0.38779998995661497,
+      "learning_rate": 4.88514173998045e-05,
+      "loss": 0.6132,
+      "step": 276
+    },
+    {
+      "epoch": 0.3651944627554384,
+      "grad_norm": 0.31541140536988005,
+      "learning_rate": 4.882697947214077e-05,
+      "loss": 0.6382,
+      "step": 277
+    },
+    {
+      "epoch": 0.36651285431773234,
+      "grad_norm": 0.3572360053908412,
+      "learning_rate": 4.880254154447703e-05,
+      "loss": 0.6305,
+      "step": 278
+    },
+    {
+      "epoch": 0.36783124588002636,
+      "grad_norm": 0.3678601031448029,
+      "learning_rate": 4.87781036168133e-05,
+      "loss": 0.6441,
+      "step": 279
+    },
+    {
+      "epoch": 0.3691496374423204,
+      "grad_norm": 0.32755296619295676,
+      "learning_rate": 4.875366568914956e-05,
+      "loss": 0.6162,
+      "step": 280
+    },
+    {
+      "epoch": 0.3704680290046144,
+      "grad_norm": 0.2855109494775668,
+      "learning_rate": 4.8729227761485826e-05,
+      "loss": 0.6513,
+      "step": 281
+    },
+    {
+      "epoch": 0.37178642056690836,
+      "grad_norm": 0.29991521446930225,
+      "learning_rate": 4.87047898338221e-05,
+      "loss": 0.6041,
+      "step": 282
+    },
+    {
+      "epoch": 0.3731048121292024,
+      "grad_norm": 0.29264036142383154,
+      "learning_rate": 4.868035190615836e-05,
+      "loss": 0.5943,
+      "step": 283
+    },
+    {
+      "epoch": 0.3744232036914964,
+      "grad_norm": 0.3082407653847592,
+      "learning_rate": 4.865591397849463e-05,
+      "loss": 0.6503,
+      "step": 284
+    },
+    {
+      "epoch": 0.37574159525379036,
+      "grad_norm": 0.3386867584145979,
+      "learning_rate": 4.863147605083089e-05,
+      "loss": 0.6503,
+      "step": 285
+    },
+    {
+      "epoch": 0.3770599868160844,
+      "grad_norm": 0.32495476490905595,
+      "learning_rate": 4.860703812316716e-05,
+      "loss": 0.6877,
+      "step": 286
+    },
+    {
+      "epoch": 0.3783783783783784,
+      "grad_norm": 0.36901281809729264,
+      "learning_rate": 4.858260019550342e-05,
+      "loss": 0.684,
+      "step": 287
+    },
+    {
+      "epoch": 0.37969676994067236,
+      "grad_norm": 0.30556314298352727,
+      "learning_rate": 4.855816226783969e-05,
+      "loss": 0.6454,
+      "step": 288
+    },
+    {
+      "epoch": 0.3810151615029664,
+      "grad_norm": 0.38958756266058914,
+      "learning_rate": 4.853372434017596e-05,
+      "loss": 0.6463,
+      "step": 289
+    },
+    {
+      "epoch": 0.3823335530652604,
+      "grad_norm": 0.3409899755114748,
+      "learning_rate": 4.850928641251222e-05,
+      "loss": 0.6178,
+      "step": 290
+    },
+    {
+      "epoch": 0.38365194462755436,
+      "grad_norm": 0.3249444057573833,
+      "learning_rate": 4.848484848484849e-05,
+      "loss": 0.6259,
+      "step": 291
+    },
+    {
+      "epoch": 0.3849703361898484,
+      "grad_norm": 0.40481582066620236,
+      "learning_rate": 4.846041055718475e-05,
+      "loss": 0.6224,
+      "step": 292
+    },
+    {
+      "epoch": 0.3862887277521424,
+      "grad_norm": 0.3266613346159151,
+      "learning_rate": 4.843597262952102e-05,
+      "loss": 0.6451,
+      "step": 293
+    },
+    {
+      "epoch": 0.38760711931443637,
+      "grad_norm": 0.31918730585613925,
+      "learning_rate": 4.841153470185729e-05,
+      "loss": 0.6074,
+      "step": 294
+    },
+    {
+      "epoch": 0.3889255108767304,
+      "grad_norm": 0.2960238380926744,
+      "learning_rate": 4.8387096774193554e-05,
+      "loss": 0.6113,
+      "step": 295
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 0.3183668455148114,
+      "learning_rate": 4.836265884652982e-05,
+      "loss": 0.6289,
+      "step": 296
+    },
+    {
+      "epoch": 0.39156229400131837,
+      "grad_norm": 0.3538570104478579,
+      "learning_rate": 4.833822091886608e-05,
+      "loss": 0.6267,
+      "step": 297
+    },
+    {
+      "epoch": 0.3928806855636124,
+      "grad_norm": 0.2932596824068987,
+      "learning_rate": 4.831378299120235e-05,
+      "loss": 0.6358,
+      "step": 298
+    },
+    {
+      "epoch": 0.3941990771259064,
+      "grad_norm": 0.35000634461967195,
+      "learning_rate": 4.828934506353861e-05,
+      "loss": 0.6414,
+      "step": 299
+    },
+    {
+      "epoch": 0.39551746868820037,
+      "grad_norm": 0.28016526306765527,
+      "learning_rate": 4.8264907135874885e-05,
+      "loss": 0.6337,
+      "step": 300
+    },
+    {
+      "epoch": 0.3968358602504944,
+      "grad_norm": 0.29463684490408093,
+      "learning_rate": 4.824046920821115e-05,
+      "loss": 0.637,
+      "step": 301
+    },
+    {
+      "epoch": 0.3981542518127884,
+      "grad_norm": 0.3038903563771035,
+      "learning_rate": 4.8216031280547414e-05,
+      "loss": 0.6014,
+      "step": 302
+    },
+    {
+      "epoch": 0.3994726433750824,
+      "grad_norm": 0.27681340028644763,
+      "learning_rate": 4.819159335288368e-05,
+      "loss": 0.6515,
+      "step": 303
+    },
+    {
+      "epoch": 0.4007910349373764,
+      "grad_norm": 0.30680102002193416,
+      "learning_rate": 4.8167155425219944e-05,
+      "loss": 0.6231,
+      "step": 304
+    },
+    {
+      "epoch": 0.4021094264996704,
+      "grad_norm": 0.3091537573795029,
+      "learning_rate": 4.814271749755621e-05,
+      "loss": 0.6263,
+      "step": 305
+    },
+    {
+      "epoch": 0.4034278180619644,
+      "grad_norm": 0.31813875232140304,
+      "learning_rate": 4.811827956989248e-05,
+      "loss": 0.6007,
+      "step": 306
+    },
+    {
+      "epoch": 0.4047462096242584,
+      "grad_norm": 0.2861921002784733,
+      "learning_rate": 4.8093841642228745e-05,
+      "loss": 0.6207,
+      "step": 307
+    },
+    {
+      "epoch": 0.4060646011865524,
+      "grad_norm": 0.2938154272545149,
+      "learning_rate": 4.806940371456501e-05,
+      "loss": 0.5903,
+      "step": 308
+    },
+    {
+      "epoch": 0.40738299274884643,
+      "grad_norm": 0.2975278360974395,
+      "learning_rate": 4.8044965786901275e-05,
+      "loss": 0.6567,
+      "step": 309
+    },
+    {
+      "epoch": 0.4087013843111404,
+      "grad_norm": 0.30543247834345727,
+      "learning_rate": 4.802052785923754e-05,
+      "loss": 0.6179,
+      "step": 310
+    },
+    {
+      "epoch": 0.4100197758734344,
+      "grad_norm": 0.3065856680178293,
+      "learning_rate": 4.7996089931573804e-05,
+      "loss": 0.6115,
+      "step": 311
+    },
+    {
+      "epoch": 0.41133816743572843,
+      "grad_norm": 0.2673870122127383,
+      "learning_rate": 4.7971652003910076e-05,
+      "loss": 0.6178,
+      "step": 312
+    },
+    {
+      "epoch": 0.4126565589980224,
+      "grad_norm": 0.3173084863444237,
+      "learning_rate": 4.794721407624634e-05,
+      "loss": 0.6502,
+      "step": 313
+    },
+    {
+      "epoch": 0.4139749505603164,
+      "grad_norm": 0.3165225479513198,
+      "learning_rate": 4.7922776148582605e-05,
+      "loss": 0.6121,
+      "step": 314
+    },
+    {
+      "epoch": 0.41529334212261043,
+      "grad_norm": 0.312730900133789,
+      "learning_rate": 4.789833822091887e-05,
+      "loss": 0.593,
+      "step": 315
+    },
+    {
+      "epoch": 0.4166117336849044,
+      "grad_norm": 0.31972902742359965,
+      "learning_rate": 4.7873900293255135e-05,
+      "loss": 0.6641,
+      "step": 316
+    },
+    {
+      "epoch": 0.4179301252471984,
+      "grad_norm": 0.3991938707378862,
+      "learning_rate": 4.78494623655914e-05,
+      "loss": 0.6346,
+      "step": 317
+    },
+    {
+      "epoch": 0.41924851680949243,
+      "grad_norm": 0.30624623977599336,
+      "learning_rate": 4.782502443792767e-05,
+      "loss": 0.6153,
+      "step": 318
+    },
+    {
+      "epoch": 0.4205669083717864,
+      "grad_norm": 0.3524265551103345,
+      "learning_rate": 4.7800586510263936e-05,
+      "loss": 0.6076,
+      "step": 319
+    },
+    {
+      "epoch": 0.4218852999340804,
+      "grad_norm": 0.3396523309922421,
+      "learning_rate": 4.77761485826002e-05,
+      "loss": 0.6397,
+      "step": 320
+    },
+    {
+      "epoch": 0.42320369149637443,
+      "grad_norm": 0.29868375700539923,
+      "learning_rate": 4.7751710654936466e-05,
+      "loss": 0.6056,
+      "step": 321
+    },
+    {
+      "epoch": 0.4245220830586684,
+      "grad_norm": 0.27995213522625895,
+      "learning_rate": 4.772727272727273e-05,
+      "loss": 0.5877,
+      "step": 322
+    },
+    {
+      "epoch": 0.4258404746209624,
+      "grad_norm": 0.291125033984549,
+      "learning_rate": 4.7702834799608995e-05,
+      "loss": 0.6218,
+      "step": 323
+    },
+    {
+      "epoch": 0.42715886618325644,
+      "grad_norm": 0.30623451322331496,
+      "learning_rate": 4.767839687194526e-05,
+      "loss": 0.631,
+      "step": 324
+    },
+    {
+      "epoch": 0.42847725774555045,
+      "grad_norm": 0.29793096690786974,
+      "learning_rate": 4.765395894428153e-05,
+      "loss": 0.5881,
+      "step": 325
+    },
+    {
+      "epoch": 0.4297956493078444,
+      "grad_norm": 0.2867268134601255,
+      "learning_rate": 4.7629521016617796e-05,
+      "loss": 0.6081,
+      "step": 326
+    },
+    {
+      "epoch": 0.43111404087013844,
+      "grad_norm": 0.3107121560212918,
+      "learning_rate": 4.760508308895406e-05,
+      "loss": 0.639,
+      "step": 327
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 0.27902176688112956,
+      "learning_rate": 4.7580645161290326e-05,
+      "loss": 0.5996,
+      "step": 328
+    },
+    {
+      "epoch": 0.4337508239947264,
+      "grad_norm": 0.3345016157470328,
+      "learning_rate": 4.755620723362659e-05,
+      "loss": 0.6346,
+      "step": 329
+    },
+    {
+      "epoch": 0.43506921555702044,
+      "grad_norm": 0.29843993845987427,
+      "learning_rate": 4.7531769305962856e-05,
+      "loss": 0.6388,
+      "step": 330
+    },
+    {
+      "epoch": 0.43638760711931446,
+      "grad_norm": 0.32118818143223987,
+      "learning_rate": 4.750733137829913e-05,
+      "loss": 0.6391,
+      "step": 331
+    },
+    {
+      "epoch": 0.4377059986816084,
+      "grad_norm": 0.3095390435894711,
+      "learning_rate": 4.748289345063539e-05,
+      "loss": 0.6497,
+      "step": 332
+    },
+    {
+      "epoch": 0.43902439024390244,
+      "grad_norm": 0.31282718717401103,
+      "learning_rate": 4.745845552297166e-05,
+      "loss": 0.6504,
+      "step": 333
+    },
+    {
+      "epoch": 0.44034278180619646,
+      "grad_norm": 0.3160538633226805,
+      "learning_rate": 4.743401759530792e-05,
+      "loss": 0.6062,
+      "step": 334
+    },
+    {
+      "epoch": 0.4416611733684904,
+      "grad_norm": 0.29495362341526915,
+      "learning_rate": 4.7409579667644186e-05,
+      "loss": 0.6094,
+      "step": 335
+    },
+    {
+      "epoch": 0.44297956493078444,
+      "grad_norm": 0.3243277511393967,
+      "learning_rate": 4.738514173998045e-05,
+      "loss": 0.5893,
+      "step": 336
+    },
+    {
+      "epoch": 0.44429795649307846,
+      "grad_norm": 0.317297447285423,
+      "learning_rate": 4.736070381231672e-05,
+      "loss": 0.6475,
+      "step": 337
+    },
+    {
+      "epoch": 0.4456163480553724,
+      "grad_norm": 0.2973094358704073,
+      "learning_rate": 4.733626588465299e-05,
+      "loss": 0.6105,
+      "step": 338
+    },
+    {
+      "epoch": 0.44693473961766644,
+      "grad_norm": 0.31062398164815275,
+      "learning_rate": 4.731182795698925e-05,
+      "loss": 0.6358,
+      "step": 339
+    },
+    {
+      "epoch": 0.44825313117996046,
+      "grad_norm": 0.3192943748321918,
+      "learning_rate": 4.728739002932552e-05,
+      "loss": 0.6069,
+      "step": 340
+    },
+    {
+      "epoch": 0.4495715227422544,
+      "grad_norm": 0.2841609670035806,
+      "learning_rate": 4.726295210166178e-05,
+      "loss": 0.6044,
+      "step": 341
+    },
+    {
+      "epoch": 0.45088991430454844,
+      "grad_norm": 0.33134100483470275,
+      "learning_rate": 4.723851417399805e-05,
+      "loss": 0.6191,
+      "step": 342
+    },
+    {
+      "epoch": 0.45220830586684246,
+      "grad_norm": 0.2828345031096051,
+      "learning_rate": 4.721407624633432e-05,
+      "loss": 0.6298,
+      "step": 343
+    },
+    {
+      "epoch": 0.4535266974291364,
+      "grad_norm": 0.2861618649650625,
+      "learning_rate": 4.718963831867058e-05,
+      "loss": 0.6089,
+      "step": 344
+    },
+    {
+      "epoch": 0.45484508899143045,
+      "grad_norm": 0.3184132095013165,
+      "learning_rate": 4.716520039100685e-05,
+      "loss": 0.6531,
+      "step": 345
+    },
+    {
+      "epoch": 0.45616348055372447,
+      "grad_norm": 0.28455933190125676,
+      "learning_rate": 4.714076246334311e-05,
+      "loss": 0.6146,
+      "step": 346
+    },
+    {
+      "epoch": 0.4574818721160185,
+      "grad_norm": 0.29187047094507085,
+      "learning_rate": 4.711632453567938e-05,
+      "loss": 0.5924,
+      "step": 347
+    },
+    {
+      "epoch": 0.45880026367831245,
+      "grad_norm": 0.2876678036212224,
+      "learning_rate": 4.709188660801564e-05,
+      "loss": 0.582,
+      "step": 348
+    },
+    {
+      "epoch": 0.46011865524060647,
+      "grad_norm": 0.3073991982230297,
+      "learning_rate": 4.7067448680351914e-05,
+      "loss": 0.6168,
+      "step": 349
+    },
+    {
+      "epoch": 0.4614370468029005,
+      "grad_norm": 0.26420866044133307,
+      "learning_rate": 4.704301075268818e-05,
+      "loss": 0.6435,
+      "step": 350
+    },
+    {
+      "epoch": 0.46275543836519445,
+      "grad_norm": 0.31681086165711847,
+      "learning_rate": 4.7018572825024444e-05,
+      "loss": 0.6096,
+      "step": 351
+    },
+    {
+      "epoch": 0.46407382992748847,
+      "grad_norm": 0.27612292418802065,
+      "learning_rate": 4.699413489736071e-05,
+      "loss": 0.6446,
+      "step": 352
+    },
+    {
+      "epoch": 0.4653922214897825,
+      "grad_norm": 0.34564810406255386,
+      "learning_rate": 4.696969696969697e-05,
+      "loss": 0.6378,
+      "step": 353
+    },
+    {
+      "epoch": 0.46671061305207645,
+      "grad_norm": 0.26031202601937503,
+      "learning_rate": 4.694525904203324e-05,
+      "loss": 0.6022,
+      "step": 354
+    },
+    {
+      "epoch": 0.46802900461437047,
+      "grad_norm": 0.32808999621351986,
+      "learning_rate": 4.692082111436951e-05,
+      "loss": 0.6325,
+      "step": 355
+    },
+    {
+      "epoch": 0.4693473961766645,
+      "grad_norm": 0.3072293192939072,
+      "learning_rate": 4.6896383186705774e-05,
+      "loss": 0.6479,
+      "step": 356
+    },
+    {
+      "epoch": 0.47066578773895845,
+      "grad_norm": 0.30867978024641207,
+      "learning_rate": 4.687194525904203e-05,
+      "loss": 0.6307,
+      "step": 357
+    },
+    {
+      "epoch": 0.47198417930125247,
+      "grad_norm": 0.33867108195042434,
+      "learning_rate": 4.68475073313783e-05,
+      "loss": 0.6279,
+      "step": 358
+    },
+    {
+      "epoch": 0.4733025708635465,
+      "grad_norm": 0.3061040351873903,
+      "learning_rate": 4.682306940371456e-05,
+      "loss": 0.6517,
+      "step": 359
+    },
+    {
+      "epoch": 0.47462096242584045,
+      "grad_norm": 0.36095916426267216,
+      "learning_rate": 4.6798631476050834e-05,
+      "loss": 0.64,
+      "step": 360
+    },
+    {
+      "epoch": 0.4759393539881345,
+      "grad_norm": 0.2866120132598102,
+      "learning_rate": 4.67741935483871e-05,
+      "loss": 0.6048,
+      "step": 361
+    },
+    {
+      "epoch": 0.4772577455504285,
+      "grad_norm": 0.31542016726240635,
+      "learning_rate": 4.674975562072336e-05,
+      "loss": 0.6009,
+      "step": 362
+    },
+    {
+      "epoch": 0.47857613711272246,
+      "grad_norm": 0.31722878577965813,
+      "learning_rate": 4.672531769305963e-05,
+      "loss": 0.6303,
+      "step": 363
+    },
+    {
+      "epoch": 0.4798945286750165,
+      "grad_norm": 0.2959197888550148,
+      "learning_rate": 4.670087976539589e-05,
+      "loss": 0.6295,
+      "step": 364
+    },
+    {
+      "epoch": 0.4812129202373105,
+      "grad_norm": 0.36873416717635554,
+      "learning_rate": 4.667644183773216e-05,
+      "loss": 0.5919,
+      "step": 365
+    },
+    {
+      "epoch": 0.48253131179960446,
+      "grad_norm": 0.33716476096096903,
+      "learning_rate": 4.665200391006843e-05,
+      "loss": 0.6319,
+      "step": 366
+    },
+    {
+      "epoch": 0.4838497033618985,
+      "grad_norm": 0.40293872239112494,
+      "learning_rate": 4.6627565982404694e-05,
+      "loss": 0.6461,
+      "step": 367
+    },
+    {
+      "epoch": 0.4851680949241925,
+      "grad_norm": 0.33156589146812687,
+      "learning_rate": 4.660312805474096e-05,
+      "loss": 0.5849,
+      "step": 368
+    },
+    {
+      "epoch": 0.4864864864864865,
+      "grad_norm": 0.34924885777761533,
+      "learning_rate": 4.6578690127077224e-05,
+      "loss": 0.625,
+      "step": 369
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 0.3369034527922939,
+      "learning_rate": 4.655425219941349e-05,
+      "loss": 0.5957,
+      "step": 370
+    },
+    {
+      "epoch": 0.4891232696110745,
+      "grad_norm": 0.289771280961373,
+      "learning_rate": 4.652981427174975e-05,
+      "loss": 0.5588,
+      "step": 371
+    },
+    {
+      "epoch": 0.4904416611733685,
+      "grad_norm": 0.3552760359361117,
+      "learning_rate": 4.650537634408602e-05,
+      "loss": 0.6172,
+      "step": 372
+    },
+    {
+      "epoch": 0.4917600527356625,
+      "grad_norm": 0.292421577234314,
+      "learning_rate": 4.648093841642229e-05,
+      "loss": 0.6127,
+      "step": 373
+    },
+    {
+      "epoch": 0.4930784442979565,
+      "grad_norm": 0.3215779853091097,
+      "learning_rate": 4.6456500488758554e-05,
+      "loss": 0.6007,
+      "step": 374
+    },
+    {
+      "epoch": 0.4943968358602505,
+      "grad_norm": 0.2695619980476224,
+      "learning_rate": 4.643206256109482e-05,
+      "loss": 0.5936,
+      "step": 375
+    },
+    {
+      "epoch": 0.4957152274225445,
+      "grad_norm": 0.32699778761946613,
+      "learning_rate": 4.6407624633431084e-05,
+      "loss": 0.5667,
+      "step": 376
+    },
+    {
+      "epoch": 0.4970336189848385,
+      "grad_norm": 0.26515828218043735,
+      "learning_rate": 4.638318670576735e-05,
+      "loss": 0.6194,
+      "step": 377
+    },
+    {
+      "epoch": 0.4983520105471325,
+      "grad_norm": 0.33452550831366734,
+      "learning_rate": 4.6358748778103614e-05,
+      "loss": 0.6041,
+      "step": 378
+    },
+    {
+      "epoch": 0.4996704021094265,
+      "grad_norm": 0.28731841401766944,
+      "learning_rate": 4.6334310850439885e-05,
+      "loss": 0.6503,
+      "step": 379
+    },
+    {
+      "epoch": 0.5009887936717206,
+      "grad_norm": 0.35017245327731433,
+      "learning_rate": 4.630987292277615e-05,
+      "loss": 0.5719,
+      "step": 380
+    },
+    {
+      "epoch": 0.5023071852340145,
+      "grad_norm": 0.29647495955123276,
+      "learning_rate": 4.6285434995112415e-05,
+      "loss": 0.6197,
+      "step": 381
+    },
+    {
+      "epoch": 0.5036255767963085,
+      "grad_norm": 0.3653018304538099,
+      "learning_rate": 4.626099706744868e-05,
+      "loss": 0.6124,
+      "step": 382
+    },
+    {
+      "epoch": 0.5049439683586026,
+      "grad_norm": 0.27313090198727347,
+      "learning_rate": 4.6236559139784944e-05,
+      "loss": 0.6193,
+      "step": 383
+    },
+    {
+      "epoch": 0.5062623599208965,
+      "grad_norm": 0.3398704492172276,
+      "learning_rate": 4.621212121212121e-05,
+      "loss": 0.5929,
+      "step": 384
+    },
+    {
+      "epoch": 0.5075807514831905,
+      "grad_norm": 0.26904337569418946,
+      "learning_rate": 4.618768328445748e-05,
+      "loss": 0.6306,
+      "step": 385
+    },
+    {
+      "epoch": 0.5088991430454846,
+      "grad_norm": 0.32147013215705333,
+      "learning_rate": 4.6163245356793745e-05,
+      "loss": 0.5958,
+      "step": 386
+    },
+    {
+      "epoch": 0.5102175346077785,
+      "grad_norm": 0.318130488091003,
+      "learning_rate": 4.613880742913001e-05,
+      "loss": 0.6252,
+      "step": 387
+    },
+    {
+      "epoch": 0.5115359261700725,
+      "grad_norm": 0.3410270792639676,
+      "learning_rate": 4.6114369501466275e-05,
+      "loss": 0.5962,
+      "step": 388
+    },
+    {
+      "epoch": 0.5128543177323666,
+      "grad_norm": 0.2835238729892702,
+      "learning_rate": 4.608993157380254e-05,
+      "loss": 0.6039,
+      "step": 389
+    },
+    {
+      "epoch": 0.5141727092946605,
+      "grad_norm": 0.41396060462021156,
+      "learning_rate": 4.6065493646138805e-05,
+      "loss": 0.5775,
+      "step": 390
+    },
+    {
+      "epoch": 0.5154911008569545,
+      "grad_norm": 0.28547706296773334,
+      "learning_rate": 4.6041055718475076e-05,
+      "loss": 0.5851,
+      "step": 391
+    },
+    {
+      "epoch": 0.5168094924192486,
+      "grad_norm": 0.31138725263262984,
+      "learning_rate": 4.601661779081134e-05,
+      "loss": 0.5886,
+      "step": 392
+    },
+    {
+      "epoch": 0.5181278839815425,
+      "grad_norm": 0.27887781712437537,
+      "learning_rate": 4.5992179863147606e-05,
+      "loss": 0.5884,
+      "step": 393
+    },
+    {
+      "epoch": 0.5194462755438365,
+      "grad_norm": 0.3244759607904912,
+      "learning_rate": 4.596774193548387e-05,
+      "loss": 0.6128,
+      "step": 394
+    },
+    {
+      "epoch": 0.5207646671061306,
+      "grad_norm": 0.2829228243295781,
+      "learning_rate": 4.5943304007820135e-05,
+      "loss": 0.5897,
+      "step": 395
+    },
+    {
+      "epoch": 0.5220830586684245,
+      "grad_norm": 0.3143553132058307,
+      "learning_rate": 4.59188660801564e-05,
+      "loss": 0.6023,
+      "step": 396
+    },
+    {
+      "epoch": 0.5234014502307185,
+      "grad_norm": 0.2799380945838576,
+      "learning_rate": 4.589442815249267e-05,
+      "loss": 0.6071,
+      "step": 397
+    },
+    {
+      "epoch": 0.5247198417930126,
+      "grad_norm": 0.2917620565594732,
+      "learning_rate": 4.586999022482894e-05,
+      "loss": 0.6035,
+      "step": 398
+    },
+    {
+      "epoch": 0.5260382333553065,
+      "grad_norm": 0.26943339176710984,
+      "learning_rate": 4.58455522971652e-05,
+      "loss": 0.625,
+      "step": 399
+    },
+    {
+      "epoch": 0.5273566249176005,
+      "grad_norm": 0.31157727442367095,
+      "learning_rate": 4.5821114369501466e-05,
+      "loss": 0.5913,
+      "step": 400
+    },
+    {
+      "epoch": 0.5286750164798946,
+      "grad_norm": 0.2782944562071961,
+      "learning_rate": 4.579667644183773e-05,
+      "loss": 0.6441,
+      "step": 401
+    },
+    {
+      "epoch": 0.5299934080421885,
+      "grad_norm": 0.342351607983682,
+      "learning_rate": 4.5772238514173996e-05,
+      "loss": 0.61,
+      "step": 402
+    },
+    {
+      "epoch": 0.5313117996044825,
+      "grad_norm": 0.27419664599331467,
+      "learning_rate": 4.574780058651027e-05,
+      "loss": 0.6362,
+      "step": 403
+    },
+    {
+      "epoch": 0.5326301911667766,
+      "grad_norm": 0.32713304068046967,
+      "learning_rate": 4.572336265884653e-05,
+      "loss": 0.6366,
+      "step": 404
+    },
+    {
+      "epoch": 0.5339485827290705,
+      "grad_norm": 0.28107918247545055,
+      "learning_rate": 4.56989247311828e-05,
+      "loss": 0.5771,
+      "step": 405
+    },
+    {
+      "epoch": 0.5352669742913645,
+      "grad_norm": 0.287364420723218,
+      "learning_rate": 4.567448680351906e-05,
+      "loss": 0.6366,
+      "step": 406
+    },
+    {
+      "epoch": 0.5365853658536586,
+      "grad_norm": 0.3339642455055895,
+      "learning_rate": 4.565004887585533e-05,
+      "loss": 0.6146,
+      "step": 407
+    },
+    {
+      "epoch": 0.5379037574159525,
+      "grad_norm": 0.25668048327157955,
+      "learning_rate": 4.562561094819159e-05,
+      "loss": 0.5871,
+      "step": 408
+    },
+    {
+      "epoch": 0.5392221489782465,
+      "grad_norm": 0.3047313315993508,
+      "learning_rate": 4.560117302052786e-05,
+      "loss": 0.6031,
+      "step": 409
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 0.24660853509931008,
+      "learning_rate": 4.557673509286413e-05,
+      "loss": 0.5746,
+      "step": 410
+    },
+    {
+      "epoch": 0.5418589321028345,
+      "grad_norm": 0.2691443207201147,
+      "learning_rate": 4.555229716520039e-05,
+      "loss": 0.6432,
+      "step": 411
+    },
+    {
+      "epoch": 0.5431773236651285,
+      "grad_norm": 0.28544779772944384,
+      "learning_rate": 4.552785923753666e-05,
+      "loss": 0.6322,
+      "step": 412
+    },
+    {
+      "epoch": 0.5444957152274226,
+      "grad_norm": 0.28051603921325746,
+      "learning_rate": 4.550342130987292e-05,
+      "loss": 0.5842,
+      "step": 413
+    },
+    {
+      "epoch": 0.5458141067897165,
+      "grad_norm": 0.3251214806204222,
+      "learning_rate": 4.547898338220919e-05,
+      "loss": 0.5734,
+      "step": 414
+    },
+    {
+      "epoch": 0.5471324983520105,
+      "grad_norm": 0.2811551250657306,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.5872,
+      "step": 415
+    },
+    {
+      "epoch": 0.5484508899143046,
+      "grad_norm": 0.2743557405759694,
+      "learning_rate": 4.543010752688172e-05,
+      "loss": 0.6372,
+      "step": 416
+    },
+    {
+      "epoch": 0.5497692814765985,
+      "grad_norm": 0.3613511779607068,
+      "learning_rate": 4.540566959921799e-05,
+      "loss": 0.6001,
+      "step": 417
+    },
+    {
+      "epoch": 0.5510876730388925,
+      "grad_norm": 0.26559237111029127,
+      "learning_rate": 4.538123167155425e-05,
+      "loss": 0.5639,
+      "step": 418
+    },
+    {
+      "epoch": 0.5524060646011866,
+      "grad_norm": 0.3289124185281232,
+      "learning_rate": 4.535679374389052e-05,
+      "loss": 0.6031,
+      "step": 419
+    },
+    {
+      "epoch": 0.5537244561634805,
+      "grad_norm": 0.295162431436294,
+      "learning_rate": 4.533235581622678e-05,
+      "loss": 0.6286,
+      "step": 420
+    },
+    {
+      "epoch": 0.5550428477257745,
+      "grad_norm": 0.30427665006140364,
+      "learning_rate": 4.530791788856305e-05,
+      "loss": 0.6293,
+      "step": 421
+    },
+    {
+      "epoch": 0.5563612392880686,
+      "grad_norm": 0.26273159065097074,
+      "learning_rate": 4.528347996089932e-05,
+      "loss": 0.6097,
+      "step": 422
+    },
+    {
+      "epoch": 0.5576796308503625,
+      "grad_norm": 0.27723472588875814,
+      "learning_rate": 4.5259042033235584e-05,
+      "loss": 0.5947,
+      "step": 423
+    },
+    {
+      "epoch": 0.5589980224126566,
+      "grad_norm": 0.2521131422658628,
+      "learning_rate": 4.523460410557185e-05,
+      "loss": 0.5973,
+      "step": 424
+    },
+    {
+      "epoch": 0.5603164139749506,
+      "grad_norm": 0.2490211155854374,
+      "learning_rate": 4.521016617790811e-05,
+      "loss": 0.6437,
+      "step": 425
+    },
+    {
+      "epoch": 0.5616348055372445,
+      "grad_norm": 0.31744592637596547,
+      "learning_rate": 4.518572825024438e-05,
+      "loss": 0.6456,
+      "step": 426
+    },
+    {
+      "epoch": 0.5629531970995386,
+      "grad_norm": 0.29674893557631676,
+      "learning_rate": 4.516129032258064e-05,
+      "loss": 0.5757,
+      "step": 427
+    },
+    {
+      "epoch": 0.5642715886618326,
+      "grad_norm": 0.2902025205452769,
+      "learning_rate": 4.5136852394916915e-05,
+      "loss": 0.5806,
+      "step": 428
+    },
+    {
+      "epoch": 0.5655899802241265,
+      "grad_norm": 0.2920089638710051,
+      "learning_rate": 4.511241446725318e-05,
+      "loss": 0.5893,
+      "step": 429
+    },
+    {
+      "epoch": 0.5669083717864206,
+      "grad_norm": 0.3389934580486735,
+      "learning_rate": 4.5087976539589444e-05,
+      "loss": 0.6261,
+      "step": 430
+    },
+    {
+      "epoch": 0.5682267633487146,
+      "grad_norm": 0.27527802267856866,
+      "learning_rate": 4.506353861192571e-05,
+      "loss": 0.6018,
+      "step": 431
+    },
+    {
+      "epoch": 0.5695451549110085,
+      "grad_norm": 0.328235204253521,
+      "learning_rate": 4.5039100684261974e-05,
+      "loss": 0.6067,
+      "step": 432
+    },
+    {
+      "epoch": 0.5708635464733026,
+      "grad_norm": 0.27922568549890703,
+      "learning_rate": 4.501466275659824e-05,
+      "loss": 0.5551,
+      "step": 433
+    },
+    {
+      "epoch": 0.5721819380355966,
+      "grad_norm": 0.31522035418376626,
+      "learning_rate": 4.499022482893451e-05,
+      "loss": 0.609,
+      "step": 434
+    },
+    {
+      "epoch": 0.5735003295978905,
+      "grad_norm": 0.3457316153117814,
+      "learning_rate": 4.4965786901270775e-05,
+      "loss": 0.6062,
+      "step": 435
+    },
+    {
+      "epoch": 0.5748187211601846,
+      "grad_norm": 0.279500172487973,
+      "learning_rate": 4.494134897360704e-05,
+      "loss": 0.5895,
+      "step": 436
+    },
+    {
+      "epoch": 0.5761371127224786,
+      "grad_norm": 0.27711874171281825,
+      "learning_rate": 4.4916911045943305e-05,
+      "loss": 0.6191,
+      "step": 437
+    },
+    {
+      "epoch": 0.5774555042847725,
+      "grad_norm": 0.30035058666849346,
+      "learning_rate": 4.489247311827957e-05,
+      "loss": 0.6017,
+      "step": 438
+    },
+    {
+      "epoch": 0.5787738958470666,
+      "grad_norm": 0.2897126775071078,
+      "learning_rate": 4.4868035190615834e-05,
+      "loss": 0.5967,
+      "step": 439
+    },
+    {
+      "epoch": 0.5800922874093606,
+      "grad_norm": 0.28341910316161856,
+      "learning_rate": 4.4843597262952106e-05,
+      "loss": 0.6125,
+      "step": 440
+    },
+    {
+      "epoch": 0.5814106789716545,
+      "grad_norm": 0.2883626760940507,
+      "learning_rate": 4.481915933528837e-05,
+      "loss": 0.5811,
+      "step": 441
+    },
+    {
+      "epoch": 0.5827290705339486,
+      "grad_norm": 0.34695204988281697,
+      "learning_rate": 4.4794721407624635e-05,
+      "loss": 0.6517,
+      "step": 442
+    },
+    {
+      "epoch": 0.5840474620962426,
+      "grad_norm": 0.29442886324313,
+      "learning_rate": 4.47702834799609e-05,
+      "loss": 0.6097,
+      "step": 443
+    },
+    {
+      "epoch": 0.5853658536585366,
+      "grad_norm": 0.29903934601883125,
+      "learning_rate": 4.4745845552297165e-05,
+      "loss": 0.5876,
+      "step": 444
+    },
+    {
+      "epoch": 0.5866842452208306,
+      "grad_norm": 0.2915316286446714,
+      "learning_rate": 4.472140762463343e-05,
+      "loss": 0.5956,
+      "step": 445
+    },
+    {
+      "epoch": 0.5880026367831246,
+      "grad_norm": 0.25240343633357704,
+      "learning_rate": 4.46969696969697e-05,
+      "loss": 0.6093,
+      "step": 446
+    },
+    {
+      "epoch": 0.5893210283454186,
+      "grad_norm": 0.29646519890526646,
+      "learning_rate": 4.4672531769305966e-05,
+      "loss": 0.5877,
+      "step": 447
+    },
+    {
+      "epoch": 0.5906394199077126,
+      "grad_norm": 0.254550273812172,
+      "learning_rate": 4.464809384164223e-05,
+      "loss": 0.5639,
+      "step": 448
+    },
+    {
+      "epoch": 0.5919578114700066,
+      "grad_norm": 0.3122261613114902,
+      "learning_rate": 4.4623655913978496e-05,
+      "loss": 0.6127,
+      "step": 449
+    },
+    {
+      "epoch": 0.5932762030323006,
+      "grad_norm": 0.2430042040399562,
+      "learning_rate": 4.459921798631476e-05,
+      "loss": 0.5647,
+      "step": 450
+    },
+    {
+      "epoch": 0.5945945945945946,
+      "grad_norm": 0.26753232524835774,
+      "learning_rate": 4.4574780058651025e-05,
+      "loss": 0.5842,
+      "step": 451
+    },
+    {
+      "epoch": 0.5959129861568886,
+      "grad_norm": 0.25017208738283403,
+      "learning_rate": 4.45503421309873e-05,
+      "loss": 0.6153,
+      "step": 452
+    },
+    {
+      "epoch": 0.5972313777191826,
+      "grad_norm": 0.2606399638431493,
+      "learning_rate": 4.452590420332356e-05,
+      "loss": 0.6102,
+      "step": 453
+    },
+    {
+      "epoch": 0.5985497692814766,
+      "grad_norm": 0.29527176263378324,
+      "learning_rate": 4.4501466275659826e-05,
+      "loss": 0.5988,
+      "step": 454
+    },
+    {
+      "epoch": 0.5998681608437706,
+      "grad_norm": 0.25431395483908786,
+      "learning_rate": 4.447702834799609e-05,
+      "loss": 0.5827,
+      "step": 455
+    },
+    {
+      "epoch": 0.6011865524060646,
+      "grad_norm": 0.27827479495772617,
+      "learning_rate": 4.4452590420332356e-05,
+      "loss": 0.6034,
+      "step": 456
+    },
+    {
+      "epoch": 0.6025049439683586,
+      "grad_norm": 0.2605036085953959,
+      "learning_rate": 4.442815249266862e-05,
+      "loss": 0.5768,
+      "step": 457
+    },
+    {
+      "epoch": 0.6038233355306526,
+      "grad_norm": 0.2834242318025845,
+      "learning_rate": 4.440371456500489e-05,
+      "loss": 0.6003,
+      "step": 458
+    },
+    {
+      "epoch": 0.6051417270929466,
+      "grad_norm": 0.26290576050271974,
+      "learning_rate": 4.437927663734116e-05,
+      "loss": 0.6102,
+      "step": 459
+    },
+    {
+      "epoch": 0.6064601186552406,
+      "grad_norm": 0.27403059659613005,
+      "learning_rate": 4.435483870967742e-05,
+      "loss": 0.5565,
+      "step": 460
+    },
+    {
+      "epoch": 0.6077785102175346,
+      "grad_norm": 0.3226752056408468,
+      "learning_rate": 4.433040078201369e-05,
+      "loss": 0.6023,
+      "step": 461
+    },
+    {
+      "epoch": 0.6090969017798286,
+      "grad_norm": 0.2627466089838883,
+      "learning_rate": 4.430596285434995e-05,
+      "loss": 0.6105,
+      "step": 462
+    },
+    {
+      "epoch": 0.6104152933421226,
+      "grad_norm": 0.2763884227770848,
+      "learning_rate": 4.4281524926686216e-05,
+      "loss": 0.6065,
+      "step": 463
+    },
+    {
+      "epoch": 0.6117336849044166,
+      "grad_norm": 0.2585331798801367,
+      "learning_rate": 4.425708699902248e-05,
+      "loss": 0.5806,
+      "step": 464
+    },
+    {
+      "epoch": 0.6130520764667106,
+      "grad_norm": 0.26679274724959645,
+      "learning_rate": 4.423264907135875e-05,
+      "loss": 0.6047,
+      "step": 465
+    },
+    {
+      "epoch": 0.6143704680290046,
+      "grad_norm": 0.28550278905043663,
+      "learning_rate": 4.420821114369502e-05,
+      "loss": 0.6059,
+      "step": 466
+    },
+    {
+      "epoch": 0.6156888595912986,
+      "grad_norm": 0.2979338022635888,
+      "learning_rate": 4.418377321603128e-05,
+      "loss": 0.6167,
+      "step": 467
+    },
+    {
+      "epoch": 0.6170072511535926,
+      "grad_norm": 0.30737667984673916,
+      "learning_rate": 4.415933528836755e-05,
+      "loss": 0.6193,
+      "step": 468
+    },
+    {
+      "epoch": 0.6183256427158866,
+      "grad_norm": 0.2914111483062786,
+      "learning_rate": 4.413489736070381e-05,
+      "loss": 0.5723,
+      "step": 469
+    },
+    {
+      "epoch": 0.6196440342781806,
+      "grad_norm": 0.2959826891409274,
+      "learning_rate": 4.411045943304008e-05,
+      "loss": 0.6378,
+      "step": 470
+    },
+    {
+      "epoch": 0.6209624258404747,
+      "grad_norm": 0.28925978665901353,
+      "learning_rate": 4.408602150537635e-05,
+      "loss": 0.5858,
+      "step": 471
+    },
+    {
+      "epoch": 0.6222808174027686,
+      "grad_norm": 0.2828496771653559,
+      "learning_rate": 4.406158357771261e-05,
+      "loss": 0.5628,
+      "step": 472
+    },
+    {
+      "epoch": 0.6235992089650626,
+      "grad_norm": 0.26522030363953464,
+      "learning_rate": 4.403714565004888e-05,
+      "loss": 0.5821,
+      "step": 473
+    },
+    {
+      "epoch": 0.6249176005273567,
+      "grad_norm": 0.31868057232354713,
+      "learning_rate": 4.401270772238514e-05,
+      "loss": 0.6104,
+      "step": 474
+    },
+    {
+      "epoch": 0.6262359920896506,
+      "grad_norm": 0.25124114493266986,
+      "learning_rate": 4.398826979472141e-05,
+      "loss": 0.6002,
+      "step": 475
+    },
+    {
+      "epoch": 0.6275543836519446,
+      "grad_norm": 0.28820921958498946,
+      "learning_rate": 4.396383186705767e-05,
+      "loss": 0.5667,
+      "step": 476
+    },
+    {
+      "epoch": 0.6288727752142387,
+      "grad_norm": 0.2739869691921316,
+      "learning_rate": 4.3939393939393944e-05,
+      "loss": 0.6289,
+      "step": 477
+    },
+    {
+      "epoch": 0.6301911667765326,
+      "grad_norm": 0.3091516468879867,
+      "learning_rate": 4.391495601173021e-05,
+      "loss": 0.6114,
+      "step": 478
+    },
+    {
+      "epoch": 0.6315095583388266,
+      "grad_norm": 0.2942019946709481,
+      "learning_rate": 4.3890518084066474e-05,
+      "loss": 0.5972,
+      "step": 479
+    },
+    {
+      "epoch": 0.6328279499011207,
+      "grad_norm": 0.3127915168695264,
+      "learning_rate": 4.386608015640274e-05,
+      "loss": 0.6181,
+      "step": 480
+    },
+    {
+      "epoch": 0.6341463414634146,
+      "grad_norm": 0.24444854061550297,
+      "learning_rate": 4.3841642228739e-05,
+      "loss": 0.6091,
+      "step": 481
+    },
+    {
+      "epoch": 0.6354647330257086,
+      "grad_norm": 0.27125473910312803,
+      "learning_rate": 4.381720430107527e-05,
+      "loss": 0.5547,
+      "step": 482
+    },
+    {
+      "epoch": 0.6367831245880027,
+      "grad_norm": 0.2430699490525948,
+      "learning_rate": 4.379276637341154e-05,
+      "loss": 0.5661,
+      "step": 483
+    },
+    {
+      "epoch": 0.6381015161502966,
+      "grad_norm": 0.30357253978210214,
+      "learning_rate": 4.3768328445747804e-05,
+      "loss": 0.5698,
+      "step": 484
+    },
+    {
+      "epoch": 0.6394199077125906,
+      "grad_norm": 0.25197907830801214,
+      "learning_rate": 4.374389051808407e-05,
+      "loss": 0.5768,
+      "step": 485
+    },
+    {
+      "epoch": 0.6407382992748847,
+      "grad_norm": 0.3144840915170465,
+      "learning_rate": 4.3719452590420334e-05,
+      "loss": 0.6016,
+      "step": 486
+    },
+    {
+      "epoch": 0.6420566908371786,
+      "grad_norm": 0.25165356629802105,
+      "learning_rate": 4.36950146627566e-05,
+      "loss": 0.5684,
+      "step": 487
+    },
+    {
+      "epoch": 0.6433750823994726,
+      "grad_norm": 0.3335945781853816,
+      "learning_rate": 4.3670576735092864e-05,
+      "loss": 0.5734,
+      "step": 488
+    },
+    {
+      "epoch": 0.6446934739617667,
+      "grad_norm": 0.24399989662908764,
+      "learning_rate": 4.3646138807429135e-05,
+      "loss": 0.6034,
+      "step": 489
+    },
+    {
+      "epoch": 0.6460118655240606,
+      "grad_norm": 0.3530505700798496,
+      "learning_rate": 4.36217008797654e-05,
+      "loss": 0.6429,
+      "step": 490
+    },
+    {
+      "epoch": 0.6473302570863546,
+      "grad_norm": 0.2621243965994534,
+      "learning_rate": 4.3597262952101665e-05,
+      "loss": 0.5936,
+      "step": 491
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 0.3058710158273273,
+      "learning_rate": 4.357282502443793e-05,
+      "loss": 0.5767,
+      "step": 492
+    },
+    {
+      "epoch": 0.6499670402109426,
+      "grad_norm": 0.27882695494372167,
+      "learning_rate": 4.3548387096774194e-05,
+      "loss": 0.566,
+      "step": 493
+    },
+    {
+      "epoch": 0.6512854317732366,
+      "grad_norm": 0.36448142507036324,
+      "learning_rate": 4.352394916911046e-05,
+      "loss": 0.5782,
+      "step": 494
+    },
+    {
+      "epoch": 0.6526038233355307,
+      "grad_norm": 0.2937612059395994,
+      "learning_rate": 4.349951124144673e-05,
+      "loss": 0.5627,
+      "step": 495
+    },
+    {
+      "epoch": 0.6539222148978246,
+      "grad_norm": 0.4392617137310561,
+      "learning_rate": 4.3475073313782996e-05,
+      "loss": 0.5812,
+      "step": 496
+    },
+    {
+      "epoch": 0.6552406064601186,
+      "grad_norm": 0.322385855673909,
+      "learning_rate": 4.345063538611926e-05,
+      "loss": 0.6142,
+      "step": 497
+    },
+    {
+      "epoch": 0.6565589980224127,
+      "grad_norm": 0.26325362550263126,
+      "learning_rate": 4.3426197458455525e-05,
+      "loss": 0.5688,
+      "step": 498
+    },
+    {
+      "epoch": 0.6578773895847067,
+      "grad_norm": 0.30890495972102266,
+      "learning_rate": 4.340175953079179e-05,
+      "loss": 0.6067,
+      "step": 499
+    },
+    {
+      "epoch": 0.6591957811470006,
+      "grad_norm": 0.26027184553573895,
+      "learning_rate": 4.3377321603128055e-05,
+      "loss": 0.6087,
+      "step": 500
+    },
+    {
+      "epoch": 0.6605141727092947,
+      "grad_norm": 0.2832167294414089,
+      "learning_rate": 4.3352883675464326e-05,
+      "loss": 0.5878,
+      "step": 501
+    },
+    {
+      "epoch": 0.6618325642715887,
+      "grad_norm": 0.2488612079914232,
+      "learning_rate": 4.332844574780059e-05,
+      "loss": 0.5962,
+      "step": 502
+    },
+    {
+      "epoch": 0.6631509558338826,
+      "grad_norm": 0.2867469304330498,
+      "learning_rate": 4.3304007820136856e-05,
+      "loss": 0.6597,
+      "step": 503
+    },
+    {
+      "epoch": 0.6644693473961767,
+      "grad_norm": 0.26175107800633013,
+      "learning_rate": 4.327956989247312e-05,
+      "loss": 0.5688,
+      "step": 504
+    },
+    {
+      "epoch": 0.6657877389584707,
+      "grad_norm": 0.28328870877004936,
+      "learning_rate": 4.3255131964809385e-05,
+      "loss": 0.6165,
+      "step": 505
+    },
+    {
+      "epoch": 0.6671061305207646,
+      "grad_norm": 0.25558330009857266,
+      "learning_rate": 4.323069403714565e-05,
+      "loss": 0.5763,
+      "step": 506
+    },
+    {
+      "epoch": 0.6684245220830587,
+      "grad_norm": 0.2682816142351484,
+      "learning_rate": 4.320625610948192e-05,
+      "loss": 0.6084,
+      "step": 507
+    },
+    {
+      "epoch": 0.6697429136453527,
+      "grad_norm": 0.25125076626842047,
+      "learning_rate": 4.318181818181819e-05,
+      "loss": 0.5614,
+      "step": 508
+    },
+    {
+      "epoch": 0.6710613052076466,
+      "grad_norm": 0.2578243420389455,
+      "learning_rate": 4.315738025415445e-05,
+      "loss": 0.5737,
+      "step": 509
+    },
+    {
+      "epoch": 0.6723796967699407,
+      "grad_norm": 0.28430400511855536,
+      "learning_rate": 4.3132942326490716e-05,
+      "loss": 0.5945,
+      "step": 510
+    },
+    {
+      "epoch": 0.6736980883322347,
+      "grad_norm": 0.2647917524812557,
+      "learning_rate": 4.310850439882698e-05,
+      "loss": 0.5583,
+      "step": 511
+    },
+    {
+      "epoch": 0.6750164798945286,
+      "grad_norm": 0.3011049271778911,
+      "learning_rate": 4.3084066471163246e-05,
+      "loss": 0.5751,
+      "step": 512
+    },
+    {
+      "epoch": 0.6763348714568227,
+      "grad_norm": 0.2742729256857888,
+      "learning_rate": 4.305962854349951e-05,
+      "loss": 0.5916,
+      "step": 513
+    },
+    {
+      "epoch": 0.6776532630191167,
+      "grad_norm": 0.24514061459827793,
+      "learning_rate": 4.303519061583578e-05,
+      "loss": 0.5398,
+      "step": 514
+    },
+    {
+      "epoch": 0.6789716545814107,
+      "grad_norm": 0.3039649017723097,
+      "learning_rate": 4.301075268817205e-05,
+      "loss": 0.6119,
+      "step": 515
+    },
+    {
+      "epoch": 0.6802900461437047,
+      "grad_norm": 0.30484298311882746,
+      "learning_rate": 4.298631476050831e-05,
+      "loss": 0.5785,
+      "step": 516
+    },
+    {
+      "epoch": 0.6816084377059987,
+      "grad_norm": 0.2570172310840409,
+      "learning_rate": 4.296187683284458e-05,
+      "loss": 0.6032,
+      "step": 517
+    },
+    {
+      "epoch": 0.6829268292682927,
+      "grad_norm": 0.28256463703953677,
+      "learning_rate": 4.293743890518084e-05,
+      "loss": 0.5812,
+      "step": 518
+    },
+    {
+      "epoch": 0.6842452208305867,
+      "grad_norm": 0.2392710922458081,
+      "learning_rate": 4.2913000977517106e-05,
+      "loss": 0.5858,
+      "step": 519
+    },
+    {
+      "epoch": 0.6855636123928807,
+      "grad_norm": 0.27471314818636816,
+      "learning_rate": 4.288856304985338e-05,
+      "loss": 0.6028,
+      "step": 520
+    },
+    {
+      "epoch": 0.6868820039551747,
+      "grad_norm": 0.23218594079244867,
+      "learning_rate": 4.286412512218964e-05,
+      "loss": 0.5971,
+      "step": 521
+    },
+    {
+      "epoch": 0.6882003955174687,
+      "grad_norm": 0.2630721123556713,
+      "learning_rate": 4.283968719452591e-05,
+      "loss": 0.6091,
+      "step": 522
+    },
+    {
+      "epoch": 0.6895187870797627,
+      "grad_norm": 0.25306601684210844,
+      "learning_rate": 4.281524926686217e-05,
+      "loss": 0.5578,
+      "step": 523
+    },
+    {
+      "epoch": 0.6908371786420567,
+      "grad_norm": 0.28168684730148397,
+      "learning_rate": 4.279081133919844e-05,
+      "loss": 0.5814,
+      "step": 524
+    },
+    {
+      "epoch": 0.6921555702043507,
+      "grad_norm": 0.2552584616712125,
+      "learning_rate": 4.27663734115347e-05,
+      "loss": 0.5621,
+      "step": 525
+    },
+    {
+      "epoch": 0.6934739617666447,
+      "grad_norm": 0.29066710902945647,
+      "learning_rate": 4.2741935483870973e-05,
+      "loss": 0.6153,
+      "step": 526
+    },
+    {
+      "epoch": 0.6947923533289387,
+      "grad_norm": 0.2533688188908539,
+      "learning_rate": 4.271749755620724e-05,
+      "loss": 0.5937,
+      "step": 527
+    },
+    {
+      "epoch": 0.6961107448912327,
+      "grad_norm": 0.25228539308193443,
+      "learning_rate": 4.26930596285435e-05,
+      "loss": 0.5944,
+      "step": 528
+    },
+    {
+      "epoch": 0.6974291364535267,
+      "grad_norm": 0.24272791200942767,
+      "learning_rate": 4.266862170087977e-05,
+      "loss": 0.5688,
+      "step": 529
+    },
+    {
+      "epoch": 0.6987475280158207,
+      "grad_norm": 0.2398912844258825,
+      "learning_rate": 4.264418377321603e-05,
+      "loss": 0.5707,
+      "step": 530
+    },
+    {
+      "epoch": 0.7000659195781147,
+      "grad_norm": 0.25492604181661643,
+      "learning_rate": 4.26197458455523e-05,
+      "loss": 0.6039,
+      "step": 531
+    },
+    {
+      "epoch": 0.7013843111404087,
+      "grad_norm": 0.2637197360682849,
+      "learning_rate": 4.259530791788857e-05,
+      "loss": 0.5844,
+      "step": 532
+    },
+    {
+      "epoch": 0.7027027027027027,
+      "grad_norm": 0.22774697422030357,
+      "learning_rate": 4.2570869990224834e-05,
+      "loss": 0.5868,
+      "step": 533
+    },
+    {
+      "epoch": 0.7040210942649967,
+      "grad_norm": 0.3014981568859656,
+      "learning_rate": 4.25464320625611e-05,
+      "loss": 0.5762,
+      "step": 534
+    },
+    {
+      "epoch": 0.7053394858272907,
+      "grad_norm": 0.27806470019859497,
+      "learning_rate": 4.252199413489736e-05,
+      "loss": 0.6198,
+      "step": 535
+    },
+    {
+      "epoch": 0.7066578773895847,
+      "grad_norm": 0.269919792185068,
+      "learning_rate": 4.249755620723363e-05,
+      "loss": 0.5449,
+      "step": 536
+    },
+    {
+      "epoch": 0.7079762689518787,
+      "grad_norm": 0.27272241747866655,
+      "learning_rate": 4.247311827956989e-05,
+      "loss": 0.6018,
+      "step": 537
+    },
+    {
+      "epoch": 0.7092946605141727,
+      "grad_norm": 0.2534543193586306,
+      "learning_rate": 4.2448680351906165e-05,
+      "loss": 0.5967,
+      "step": 538
+    },
+    {
+      "epoch": 0.7106130520764667,
+      "grad_norm": 0.2545568293376221,
+      "learning_rate": 4.242424242424243e-05,
+      "loss": 0.6063,
+      "step": 539
+    },
+    {
+      "epoch": 0.7119314436387607,
+      "grad_norm": 0.27438117225765923,
+      "learning_rate": 4.2399804496578694e-05,
+      "loss": 0.5963,
+      "step": 540
+    },
+    {
+      "epoch": 0.7132498352010547,
+      "grad_norm": 0.26843071567586513,
+      "learning_rate": 4.237536656891496e-05,
+      "loss": 0.5989,
+      "step": 541
+    },
+    {
+      "epoch": 0.7145682267633487,
+      "grad_norm": 0.24571404037122097,
+      "learning_rate": 4.2350928641251224e-05,
+      "loss": 0.5527,
+      "step": 542
+    },
+    {
+      "epoch": 0.7158866183256427,
+      "grad_norm": 0.24457481376052473,
+      "learning_rate": 4.232649071358749e-05,
+      "loss": 0.5738,
+      "step": 543
+    },
+    {
+      "epoch": 0.7172050098879367,
+      "grad_norm": 0.24703802759089602,
+      "learning_rate": 4.230205278592376e-05,
+      "loss": 0.5585,
+      "step": 544
+    },
+    {
+      "epoch": 0.7185234014502307,
+      "grad_norm": 0.26606679567061164,
+      "learning_rate": 4.2277614858260025e-05,
+      "loss": 0.5646,
+      "step": 545
+    },
+    {
+      "epoch": 0.7198417930125247,
+      "grad_norm": 0.2635747113986744,
+      "learning_rate": 4.225317693059629e-05,
+      "loss": 0.5767,
+      "step": 546
+    },
+    {
+      "epoch": 0.7211601845748187,
+      "grad_norm": 0.2530850973015238,
+      "learning_rate": 4.2228739002932555e-05,
+      "loss": 0.5779,
+      "step": 547
+    },
+    {
+      "epoch": 0.7224785761371127,
+      "grad_norm": 0.26879996451722626,
+      "learning_rate": 4.220430107526882e-05,
+      "loss": 0.5761,
+      "step": 548
+    },
+    {
+      "epoch": 0.7237969676994067,
+      "grad_norm": 0.2554259770579154,
+      "learning_rate": 4.2179863147605084e-05,
+      "loss": 0.596,
+      "step": 549
+    },
+    {
+      "epoch": 0.7251153592617007,
+      "grad_norm": 0.24333252328203567,
+      "learning_rate": 4.2155425219941356e-05,
+      "loss": 0.6081,
+      "step": 550
+    },
+    {
+      "epoch": 0.7264337508239948,
+      "grad_norm": 0.28018068232527626,
+      "learning_rate": 4.213098729227762e-05,
+      "loss": 0.5864,
+      "step": 551
+    },
+    {
+      "epoch": 0.7277521423862887,
+      "grad_norm": 0.24567895271662646,
+      "learning_rate": 4.2106549364613885e-05,
+      "loss": 0.5914,
+      "step": 552
+    },
+    {
+      "epoch": 0.7290705339485827,
+      "grad_norm": 0.28738113551106464,
+      "learning_rate": 4.208211143695015e-05,
+      "loss": 0.583,
+      "step": 553
+    },
+    {
+      "epoch": 0.7303889255108768,
+      "grad_norm": 0.2450724032252225,
+      "learning_rate": 4.2057673509286415e-05,
+      "loss": 0.6322,
+      "step": 554
+    },
+    {
+      "epoch": 0.7317073170731707,
+      "grad_norm": 0.27157940003915915,
+      "learning_rate": 4.203323558162268e-05,
+      "loss": 0.5904,
+      "step": 555
+    },
+    {
+      "epoch": 0.7330257086354647,
+      "grad_norm": 0.2858236598345372,
+      "learning_rate": 4.2008797653958945e-05,
+      "loss": 0.5197,
+      "step": 556
+    },
+    {
+      "epoch": 0.7343441001977588,
+      "grad_norm": 0.27202127130916154,
+      "learning_rate": 4.1984359726295216e-05,
+      "loss": 0.6013,
+      "step": 557
+    },
+    {
+      "epoch": 0.7356624917600527,
+      "grad_norm": 0.2739332019527236,
+      "learning_rate": 4.195992179863148e-05,
+      "loss": 0.5829,
+      "step": 558
+    },
+    {
+      "epoch": 0.7369808833223468,
+      "grad_norm": 0.27448927634680226,
+      "learning_rate": 4.1935483870967746e-05,
+      "loss": 0.5515,
+      "step": 559
+    },
+    {
+      "epoch": 0.7382992748846408,
+      "grad_norm": 0.27504907971493653,
+      "learning_rate": 4.191104594330401e-05,
+      "loss": 0.597,
+      "step": 560
+    },
+    {
+      "epoch": 0.7396176664469347,
+      "grad_norm": 0.2880746883291735,
+      "learning_rate": 4.1886608015640275e-05,
+      "loss": 0.5659,
+      "step": 561
+    },
+    {
+      "epoch": 0.7409360580092288,
+      "grad_norm": 0.2533126411660386,
+      "learning_rate": 4.186217008797654e-05,
+      "loss": 0.5578,
+      "step": 562
+    },
+    {
+      "epoch": 0.7422544495715228,
+      "grad_norm": 0.30390533937156955,
+      "learning_rate": 4.183773216031281e-05,
+      "loss": 0.5978,
+      "step": 563
+    },
+    {
+      "epoch": 0.7435728411338167,
+      "grad_norm": 0.24847143841038347,
+      "learning_rate": 4.1813294232649076e-05,
+      "loss": 0.5531,
+      "step": 564
+    },
+    {
+      "epoch": 0.7448912326961108,
+      "grad_norm": 0.3059979177317098,
+      "learning_rate": 4.178885630498534e-05,
+      "loss": 0.5791,
+      "step": 565
+    },
+    {
+      "epoch": 0.7462096242584048,
+      "grad_norm": 0.28446679403220976,
+      "learning_rate": 4.1764418377321606e-05,
+      "loss": 0.5808,
+      "step": 566
+    },
+    {
+      "epoch": 0.7475280158206987,
+      "grad_norm": 0.31963069778977743,
+      "learning_rate": 4.173998044965787e-05,
+      "loss": 0.5932,
+      "step": 567
+    },
+    {
+      "epoch": 0.7488464073829928,
+      "grad_norm": 0.3089193694834029,
+      "learning_rate": 4.1715542521994136e-05,
+      "loss": 0.5714,
+      "step": 568
+    },
+    {
+      "epoch": 0.7501647989452868,
+      "grad_norm": 0.2915679040436759,
+      "learning_rate": 4.169110459433041e-05,
+      "loss": 0.5802,
+      "step": 569
+    },
+    {
+      "epoch": 0.7514831905075807,
+      "grad_norm": 0.31653319957808224,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.6131,
+      "step": 570
+    },
+    {
+      "epoch": 0.7528015820698748,
+      "grad_norm": 0.27193849306080103,
+      "learning_rate": 4.164222873900294e-05,
+      "loss": 0.6006,
+      "step": 571
+    },
+    {
+      "epoch": 0.7541199736321688,
+      "grad_norm": 0.28076181088213387,
+      "learning_rate": 4.16177908113392e-05,
+      "loss": 0.5685,
+      "step": 572
+    },
+    {
+      "epoch": 0.7554383651944627,
+      "grad_norm": 0.2660766115895722,
+      "learning_rate": 4.1593352883675466e-05,
+      "loss": 0.5896,
+      "step": 573
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 0.28957420513300675,
+      "learning_rate": 4.156891495601173e-05,
+      "loss": 0.5873,
+      "step": 574
+    },
+    {
+      "epoch": 0.7580751483190508,
+      "grad_norm": 0.2850682216741874,
+      "learning_rate": 4.1544477028348e-05,
+      "loss": 0.5792,
+      "step": 575
+    },
+    {
+      "epoch": 0.7593935398813447,
+      "grad_norm": 0.2687641625806662,
+      "learning_rate": 4.152003910068427e-05,
+      "loss": 0.5876,
+      "step": 576
+    },
+    {
+      "epoch": 0.7607119314436388,
+      "grad_norm": 0.30029382927612913,
+      "learning_rate": 4.149560117302053e-05,
+      "loss": 0.5589,
+      "step": 577
+    },
+    {
+      "epoch": 0.7620303230059328,
+      "grad_norm": 0.25977703105187067,
+      "learning_rate": 4.14711632453568e-05,
+      "loss": 0.5943,
+      "step": 578
+    },
+    {
+      "epoch": 0.7633487145682267,
+      "grad_norm": 0.2804078182050432,
+      "learning_rate": 4.144672531769306e-05,
+      "loss": 0.5729,
+      "step": 579
+    },
+    {
+      "epoch": 0.7646671061305208,
+      "grad_norm": 0.27111890005175887,
+      "learning_rate": 4.142228739002933e-05,
+      "loss": 0.5768,
+      "step": 580
+    },
+    {
+      "epoch": 0.7659854976928148,
+      "grad_norm": 0.2989140460792668,
+      "learning_rate": 4.13978494623656e-05,
+      "loss": 0.5526,
+      "step": 581
+    },
+    {
+      "epoch": 0.7673038892551087,
+      "grad_norm": 0.23446275004001751,
+      "learning_rate": 4.137341153470186e-05,
+      "loss": 0.5713,
+      "step": 582
+    },
+    {
+      "epoch": 0.7686222808174028,
+      "grad_norm": 0.2626681473279307,
+      "learning_rate": 4.134897360703813e-05,
+      "loss": 0.6062,
+      "step": 583
+    },
+    {
+      "epoch": 0.7699406723796968,
+      "grad_norm": 0.286247316667868,
+      "learning_rate": 4.132453567937439e-05,
+      "loss": 0.5832,
+      "step": 584
+    },
+    {
+      "epoch": 0.7712590639419907,
+      "grad_norm": 0.2851779101982015,
+      "learning_rate": 4.130009775171066e-05,
+      "loss": 0.6051,
+      "step": 585
+    },
+    {
+      "epoch": 0.7725774555042848,
+      "grad_norm": 0.2475933836262059,
+      "learning_rate": 4.127565982404692e-05,
+      "loss": 0.5477,
+      "step": 586
+    },
+    {
+      "epoch": 0.7738958470665788,
+      "grad_norm": 0.27972587868202436,
+      "learning_rate": 4.1251221896383194e-05,
+      "loss": 0.6317,
+      "step": 587
+    },
+    {
+      "epoch": 0.7752142386288727,
+      "grad_norm": 0.2782060408098243,
+      "learning_rate": 4.122678396871946e-05,
+      "loss": 0.59,
+      "step": 588
+    },
+    {
+      "epoch": 0.7765326301911668,
+      "grad_norm": 0.2942990056969366,
+      "learning_rate": 4.1202346041055724e-05,
+      "loss": 0.5856,
+      "step": 589
+    },
+    {
+      "epoch": 0.7778510217534608,
+      "grad_norm": 0.24647019970055226,
+      "learning_rate": 4.117790811339199e-05,
+      "loss": 0.5778,
+      "step": 590
+    },
+    {
+      "epoch": 0.7791694133157547,
+      "grad_norm": 0.2565175439187913,
+      "learning_rate": 4.115347018572825e-05,
+      "loss": 0.5525,
+      "step": 591
+    },
+    {
+      "epoch": 0.7804878048780488,
+      "grad_norm": 0.33472935029434553,
+      "learning_rate": 4.112903225806452e-05,
+      "loss": 0.5615,
+      "step": 592
+    },
+    {
+      "epoch": 0.7818061964403428,
+      "grad_norm": 0.2525474451666082,
+      "learning_rate": 4.110459433040079e-05,
+      "loss": 0.5809,
+      "step": 593
+    },
+    {
+      "epoch": 0.7831245880026367,
+      "grad_norm": 0.2721471710439492,
+      "learning_rate": 4.1080156402737054e-05,
+      "loss": 0.5616,
+      "step": 594
+    },
+    {
+      "epoch": 0.7844429795649308,
+      "grad_norm": 0.25614445025884547,
+      "learning_rate": 4.105571847507332e-05,
+      "loss": 0.5915,
+      "step": 595
+    },
+    {
+      "epoch": 0.7857613711272248,
+      "grad_norm": 0.2352072541251627,
+      "learning_rate": 4.1031280547409584e-05,
+      "loss": 0.5304,
+      "step": 596
+    },
+    {
+      "epoch": 0.7870797626895187,
+      "grad_norm": 0.24918291912978888,
+      "learning_rate": 4.100684261974585e-05,
+      "loss": 0.5879,
+      "step": 597
+    },
+    {
+      "epoch": 0.7883981542518128,
+      "grad_norm": 0.2605431347759016,
+      "learning_rate": 4.0982404692082114e-05,
+      "loss": 0.5625,
+      "step": 598
+    },
+    {
+      "epoch": 0.7897165458141068,
+      "grad_norm": 0.2556091252826518,
+      "learning_rate": 4.0957966764418385e-05,
+      "loss": 0.5809,
+      "step": 599
+    },
+    {
+      "epoch": 0.7910349373764007,
+      "grad_norm": 0.27730054080546096,
+      "learning_rate": 4.093352883675465e-05,
+      "loss": 0.5785,
+      "step": 600
+    },
+    {
+      "epoch": 0.7923533289386948,
+      "grad_norm": 0.31688837260965635,
+      "learning_rate": 4.0909090909090915e-05,
+      "loss": 0.5768,
+      "step": 601
+    },
+    {
+      "epoch": 0.7936717205009888,
+      "grad_norm": 0.25511458288302774,
+      "learning_rate": 4.088465298142718e-05,
+      "loss": 0.59,
+      "step": 602
+    },
+    {
+      "epoch": 0.7949901120632828,
+      "grad_norm": 0.27568375393526295,
+      "learning_rate": 4.0860215053763444e-05,
+      "loss": 0.5655,
+      "step": 603
+    },
+    {
+      "epoch": 0.7963085036255768,
+      "grad_norm": 0.24531498341411834,
+      "learning_rate": 4.083577712609971e-05,
+      "loss": 0.5763,
+      "step": 604
+    },
+    {
+      "epoch": 0.7976268951878708,
+      "grad_norm": 0.27737389124937584,
+      "learning_rate": 4.0811339198435974e-05,
+      "loss": 0.6049,
+      "step": 605
+    },
+    {
+      "epoch": 0.7989452867501649,
+      "grad_norm": 0.2665381526208649,
+      "learning_rate": 4.0786901270772246e-05,
+      "loss": 0.5911,
+      "step": 606
+    },
+    {
+      "epoch": 0.8002636783124588,
+      "grad_norm": 0.3004312699442013,
+      "learning_rate": 4.076246334310851e-05,
+      "loss": 0.5701,
+      "step": 607
+    },
+    {
+      "epoch": 0.8015820698747528,
+      "grad_norm": 0.259406450402484,
+      "learning_rate": 4.0738025415444775e-05,
+      "loss": 0.5788,
+      "step": 608
+    },
+    {
+      "epoch": 0.8029004614370469,
+      "grad_norm": 0.25388247567477534,
+      "learning_rate": 4.071358748778104e-05,
+      "loss": 0.5744,
+      "step": 609
+    },
+    {
+      "epoch": 0.8042188529993408,
+      "grad_norm": 0.2804898488362729,
+      "learning_rate": 4.0689149560117305e-05,
+      "loss": 0.572,
+      "step": 610
+    },
+    {
+      "epoch": 0.8055372445616348,
+      "grad_norm": 0.26827548217579805,
+      "learning_rate": 4.066471163245357e-05,
+      "loss": 0.5764,
+      "step": 611
+    },
+    {
+      "epoch": 0.8068556361239289,
+      "grad_norm": 0.2472992485976218,
+      "learning_rate": 4.064027370478984e-05,
+      "loss": 0.5911,
+      "step": 612
+    },
+    {
+      "epoch": 0.8081740276862228,
+      "grad_norm": 0.2772544059452531,
+      "learning_rate": 4.06158357771261e-05,
+      "loss": 0.5838,
+      "step": 613
+    },
+    {
+      "epoch": 0.8094924192485168,
+      "grad_norm": 0.23476969544254697,
+      "learning_rate": 4.0591397849462364e-05,
+      "loss": 0.6063,
+      "step": 614
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 0.27613003917257456,
+      "learning_rate": 4.056695992179863e-05,
+      "loss": 0.5735,
+      "step": 615
+    },
+    {
+      "epoch": 0.8121292023731048,
+      "grad_norm": 0.23819051975607636,
+      "learning_rate": 4.0542521994134894e-05,
+      "loss": 0.5555,
+      "step": 616
+    },
+    {
+      "epoch": 0.8134475939353988,
+      "grad_norm": 0.25441514328928067,
+      "learning_rate": 4.0518084066471165e-05,
+      "loss": 0.5818,
+      "step": 617
+    },
+    {
+      "epoch": 0.8147659854976929,
+      "grad_norm": 0.286505321906632,
+      "learning_rate": 4.049364613880743e-05,
+      "loss": 0.6137,
+      "step": 618
+    },
+    {
+      "epoch": 0.8160843770599868,
+      "grad_norm": 0.22405947143028673,
+      "learning_rate": 4.0469208211143695e-05,
+      "loss": 0.596,
+      "step": 619
+    },
+    {
+      "epoch": 0.8174027686222808,
+      "grad_norm": 0.27215956122999413,
+      "learning_rate": 4.044477028347996e-05,
+      "loss": 0.5903,
+      "step": 620
+    },
+    {
+      "epoch": 0.8187211601845749,
+      "grad_norm": 0.25126604847350664,
+      "learning_rate": 4.0420332355816224e-05,
+      "loss": 0.5915,
+      "step": 621
+    },
+    {
+      "epoch": 0.8200395517468688,
+      "grad_norm": 0.23752875054756378,
+      "learning_rate": 4.039589442815249e-05,
+      "loss": 0.5672,
+      "step": 622
+    },
+    {
+      "epoch": 0.8213579433091628,
+      "grad_norm": 0.31346179443011246,
+      "learning_rate": 4.037145650048876e-05,
+      "loss": 0.5894,
+      "step": 623
+    },
+    {
+      "epoch": 0.8226763348714569,
+      "grad_norm": 0.23918367431513135,
+      "learning_rate": 4.0347018572825025e-05,
+      "loss": 0.5932,
+      "step": 624
+    },
+    {
+      "epoch": 0.8239947264337508,
+      "grad_norm": 0.32454555758690956,
+      "learning_rate": 4.032258064516129e-05,
+      "loss": 0.5404,
+      "step": 625
+    },
+    {
+      "epoch": 0.8253131179960448,
+      "grad_norm": 0.26194551529324944,
+      "learning_rate": 4.0298142717497555e-05,
+      "loss": 0.5445,
+      "step": 626
+    },
+    {
+      "epoch": 0.8266315095583389,
+      "grad_norm": 0.30753622735264785,
+      "learning_rate": 4.027370478983382e-05,
+      "loss": 0.5915,
+      "step": 627
+    },
+    {
+      "epoch": 0.8279499011206328,
+      "grad_norm": 0.2707063917970171,
+      "learning_rate": 4.0249266862170085e-05,
+      "loss": 0.5456,
+      "step": 628
+    },
+    {
+      "epoch": 0.8292682926829268,
+      "grad_norm": 0.29332919158487597,
+      "learning_rate": 4.0224828934506356e-05,
+      "loss": 0.5873,
+      "step": 629
+    },
+    {
+      "epoch": 0.8305866842452209,
+      "grad_norm": 0.2997230878996013,
+      "learning_rate": 4.020039100684262e-05,
+      "loss": 0.5884,
+      "step": 630
+    },
+    {
+      "epoch": 0.8319050758075148,
+      "grad_norm": 0.27574457040937517,
+      "learning_rate": 4.0175953079178886e-05,
+      "loss": 0.6042,
+      "step": 631
+    },
+    {
+      "epoch": 0.8332234673698088,
+      "grad_norm": 0.32309655855533437,
+      "learning_rate": 4.015151515151515e-05,
+      "loss": 0.553,
+      "step": 632
+    },
+    {
+      "epoch": 0.8345418589321029,
+      "grad_norm": 0.260408410960416,
+      "learning_rate": 4.0127077223851415e-05,
+      "loss": 0.568,
+      "step": 633
+    },
+    {
+      "epoch": 0.8358602504943968,
+      "grad_norm": 0.2861236084639323,
+      "learning_rate": 4.010263929618768e-05,
+      "loss": 0.5931,
+      "step": 634
+    },
+    {
+      "epoch": 0.8371786420566908,
+      "grad_norm": 0.2847060901193924,
+      "learning_rate": 4.007820136852395e-05,
+      "loss": 0.5812,
+      "step": 635
+    },
+    {
+      "epoch": 0.8384970336189849,
+      "grad_norm": 0.2667817782589144,
+      "learning_rate": 4.005376344086022e-05,
+      "loss": 0.5589,
+      "step": 636
+    },
+    {
+      "epoch": 0.8398154251812788,
+      "grad_norm": 0.27161768067042885,
+      "learning_rate": 4.002932551319648e-05,
+      "loss": 0.5718,
+      "step": 637
+    },
+    {
+      "epoch": 0.8411338167435728,
+      "grad_norm": 0.2581137909748153,
+      "learning_rate": 4.0004887585532746e-05,
+      "loss": 0.5692,
+      "step": 638
+    },
+    {
+      "epoch": 0.8424522083058669,
+      "grad_norm": 0.22932264311939074,
+      "learning_rate": 3.998044965786901e-05,
+      "loss": 0.5624,
+      "step": 639
+    },
+    {
+      "epoch": 0.8437705998681608,
+      "grad_norm": 0.28920344563872724,
+      "learning_rate": 3.9956011730205276e-05,
+      "loss": 0.5576,
+      "step": 640
+    },
+    {
+      "epoch": 0.8450889914304548,
+      "grad_norm": 0.22970784577397887,
+      "learning_rate": 3.993157380254155e-05,
+      "loss": 0.5576,
+      "step": 641
+    },
+    {
+      "epoch": 0.8464073829927489,
+      "grad_norm": 0.2909485578032682,
+      "learning_rate": 3.990713587487781e-05,
+      "loss": 0.5814,
+      "step": 642
+    },
+    {
+      "epoch": 0.8477257745550428,
+      "grad_norm": 0.23972791189571147,
+      "learning_rate": 3.988269794721408e-05,
+      "loss": 0.5523,
+      "step": 643
+    },
+    {
+      "epoch": 0.8490441661173368,
+      "grad_norm": 0.2772054078403191,
+      "learning_rate": 3.985826001955034e-05,
+      "loss": 0.608,
+      "step": 644
+    },
+    {
+      "epoch": 0.8503625576796309,
+      "grad_norm": 0.2704863225310378,
+      "learning_rate": 3.983382209188661e-05,
+      "loss": 0.5441,
+      "step": 645
+    },
+    {
+      "epoch": 0.8516809492419248,
+      "grad_norm": 0.2783402514020125,
+      "learning_rate": 3.980938416422287e-05,
+      "loss": 0.6073,
+      "step": 646
+    },
+    {
+      "epoch": 0.8529993408042189,
+      "grad_norm": 0.2580558613883854,
+      "learning_rate": 3.978494623655914e-05,
+      "loss": 0.5547,
+      "step": 647
+    },
+    {
+      "epoch": 0.8543177323665129,
+      "grad_norm": 0.2398749358333043,
+      "learning_rate": 3.976050830889541e-05,
+      "loss": 0.5646,
+      "step": 648
+    },
+    {
+      "epoch": 0.8556361239288068,
+      "grad_norm": 0.27859020004891927,
+      "learning_rate": 3.973607038123167e-05,
+      "loss": 0.5639,
+      "step": 649
+    },
+    {
+      "epoch": 0.8569545154911009,
+      "grad_norm": 0.2432703530519651,
+      "learning_rate": 3.971163245356794e-05,
+      "loss": 0.5886,
+      "step": 650
+    },
+    {
+      "epoch": 0.8582729070533949,
+      "grad_norm": 0.2630806971272542,
+      "learning_rate": 3.96871945259042e-05,
+      "loss": 0.5905,
+      "step": 651
+    },
+    {
+      "epoch": 0.8595912986156888,
+      "grad_norm": 0.24639274802412442,
+      "learning_rate": 3.966275659824047e-05,
+      "loss": 0.5044,
+      "step": 652
+    },
+    {
+      "epoch": 0.8609096901779829,
+      "grad_norm": 0.2688654710389458,
+      "learning_rate": 3.963831867057673e-05,
+      "loss": 0.5893,
+      "step": 653
+    },
+    {
+      "epoch": 0.8622280817402769,
+      "grad_norm": 0.24572379653240067,
+      "learning_rate": 3.9613880742913e-05,
+      "loss": 0.5869,
+      "step": 654
+    },
+    {
+      "epoch": 0.8635464733025708,
+      "grad_norm": 0.2516823093204936,
+      "learning_rate": 3.958944281524927e-05,
+      "loss": 0.5409,
+      "step": 655
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 0.26012154450158637,
+      "learning_rate": 3.956500488758553e-05,
+      "loss": 0.6212,
+      "step": 656
+    },
+    {
+      "epoch": 0.8661832564271589,
+      "grad_norm": 0.2495456819951276,
+      "learning_rate": 3.95405669599218e-05,
+      "loss": 0.5352,
+      "step": 657
+    },
+    {
+      "epoch": 0.8675016479894528,
+      "grad_norm": 0.22323323511733179,
+      "learning_rate": 3.951612903225806e-05,
+      "loss": 0.5748,
+      "step": 658
+    },
+    {
+      "epoch": 0.8688200395517469,
+      "grad_norm": 0.25747628886130935,
+      "learning_rate": 3.949169110459433e-05,
+      "loss": 0.5734,
+      "step": 659
+    },
+    {
+      "epoch": 0.8701384311140409,
+      "grad_norm": 0.23633666511075382,
+      "learning_rate": 3.94672531769306e-05,
+      "loss": 0.5729,
+      "step": 660
+    },
+    {
+      "epoch": 0.8714568226763348,
+      "grad_norm": 0.2327230833872904,
+      "learning_rate": 3.9442815249266864e-05,
+      "loss": 0.5468,
+      "step": 661
+    },
+    {
+      "epoch": 0.8727752142386289,
+      "grad_norm": 0.2911396131826174,
+      "learning_rate": 3.941837732160313e-05,
+      "loss": 0.5438,
+      "step": 662
+    },
+    {
+      "epoch": 0.8740936058009229,
+      "grad_norm": 0.23809647038334042,
+      "learning_rate": 3.939393939393939e-05,
+      "loss": 0.5522,
+      "step": 663
+    },
+    {
+      "epoch": 0.8754119973632168,
+      "grad_norm": 0.2675095137100887,
+      "learning_rate": 3.936950146627566e-05,
+      "loss": 0.5386,
+      "step": 664
+    },
+    {
+      "epoch": 0.8767303889255109,
+      "grad_norm": 0.23957921139365,
+      "learning_rate": 3.934506353861192e-05,
+      "loss": 0.5701,
+      "step": 665
+    },
+    {
+      "epoch": 0.8780487804878049,
+      "grad_norm": 0.2374625500844515,
+      "learning_rate": 3.9320625610948195e-05,
+      "loss": 0.5903,
+      "step": 666
+    },
+    {
+      "epoch": 0.8793671720500988,
+      "grad_norm": 0.26215460408061897,
+      "learning_rate": 3.929618768328446e-05,
+      "loss": 0.5417,
+      "step": 667
+    },
+    {
+      "epoch": 0.8806855636123929,
+      "grad_norm": 0.244583057548248,
+      "learning_rate": 3.9271749755620724e-05,
+      "loss": 0.6029,
+      "step": 668
+    },
+    {
+      "epoch": 0.8820039551746869,
+      "grad_norm": 0.25591608366862945,
+      "learning_rate": 3.924731182795699e-05,
+      "loss": 0.6095,
+      "step": 669
+    },
+    {
+      "epoch": 0.8833223467369808,
+      "grad_norm": 0.2773412412422752,
+      "learning_rate": 3.9222873900293254e-05,
+      "loss": 0.5987,
+      "step": 670
+    },
+    {
+      "epoch": 0.8846407382992749,
+      "grad_norm": 0.24359255635955676,
+      "learning_rate": 3.919843597262952e-05,
+      "loss": 0.5487,
+      "step": 671
+    },
+    {
+      "epoch": 0.8859591298615689,
+      "grad_norm": 0.30677599295742874,
+      "learning_rate": 3.917399804496579e-05,
+      "loss": 0.6028,
+      "step": 672
+    },
+    {
+      "epoch": 0.8872775214238628,
+      "grad_norm": 0.2486521614907155,
+      "learning_rate": 3.9149560117302055e-05,
+      "loss": 0.5837,
+      "step": 673
+    },
+    {
+      "epoch": 0.8885959129861569,
+      "grad_norm": 0.2747985237355314,
+      "learning_rate": 3.912512218963832e-05,
+      "loss": 0.5437,
+      "step": 674
+    },
+    {
+      "epoch": 0.8899143045484509,
+      "grad_norm": 0.24615676321613958,
+      "learning_rate": 3.9100684261974585e-05,
+      "loss": 0.5619,
+      "step": 675
+    },
+    {
+      "epoch": 0.8912326961107448,
+      "grad_norm": 0.2805176182290872,
+      "learning_rate": 3.907624633431085e-05,
+      "loss": 0.602,
+      "step": 676
+    },
+    {
+      "epoch": 0.8925510876730389,
+      "grad_norm": 0.25073657230318663,
+      "learning_rate": 3.9051808406647114e-05,
+      "loss": 0.5903,
+      "step": 677
+    },
+    {
+      "epoch": 0.8938694792353329,
+      "grad_norm": 0.2401829368331181,
+      "learning_rate": 3.9027370478983386e-05,
+      "loss": 0.5417,
+      "step": 678
+    },
+    {
+      "epoch": 0.8951878707976269,
+      "grad_norm": 0.2725945087984761,
+      "learning_rate": 3.900293255131965e-05,
+      "loss": 0.5853,
+      "step": 679
+    },
+    {
+      "epoch": 0.8965062623599209,
+      "grad_norm": 0.25013512227140855,
+      "learning_rate": 3.8978494623655915e-05,
+      "loss": 0.5679,
+      "step": 680
+    },
+    {
+      "epoch": 0.8978246539222149,
+      "grad_norm": 0.24393064218492416,
+      "learning_rate": 3.895405669599218e-05,
+      "loss": 0.5345,
+      "step": 681
+    },
+    {
+      "epoch": 0.8991430454845089,
+      "grad_norm": 0.2701753442958192,
+      "learning_rate": 3.8929618768328445e-05,
+      "loss": 0.5293,
+      "step": 682
+    },
+    {
+      "epoch": 0.9004614370468029,
+      "grad_norm": 0.2541513002002444,
+      "learning_rate": 3.890518084066471e-05,
+      "loss": 0.5848,
+      "step": 683
+    },
+    {
+      "epoch": 0.9017798286090969,
+      "grad_norm": 0.2401244987353776,
+      "learning_rate": 3.888074291300098e-05,
+      "loss": 0.5468,
+      "step": 684
+    },
+    {
+      "epoch": 0.9030982201713909,
+      "grad_norm": 0.2517490433721017,
+      "learning_rate": 3.8856304985337246e-05,
+      "loss": 0.5688,
+      "step": 685
+    },
+    {
+      "epoch": 0.9044166117336849,
+      "grad_norm": 0.28685884791036226,
+      "learning_rate": 3.883186705767351e-05,
+      "loss": 0.571,
+      "step": 686
+    },
+    {
+      "epoch": 0.9057350032959789,
+      "grad_norm": 0.23502400413910268,
+      "learning_rate": 3.8807429130009776e-05,
+      "loss": 0.5646,
+      "step": 687
+    },
+    {
+      "epoch": 0.9070533948582729,
+      "grad_norm": 0.2863436735403986,
+      "learning_rate": 3.878299120234604e-05,
+      "loss": 0.5404,
+      "step": 688
+    },
+    {
+      "epoch": 0.9083717864205669,
+      "grad_norm": 0.26322386192859415,
+      "learning_rate": 3.8758553274682305e-05,
+      "loss": 0.5499,
+      "step": 689
+    },
+    {
+      "epoch": 0.9096901779828609,
+      "grad_norm": 0.23118858037566792,
+      "learning_rate": 3.873411534701858e-05,
+      "loss": 0.5672,
+      "step": 690
+    },
+    {
+      "epoch": 0.9110085695451549,
+      "grad_norm": 0.25359274068783,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.5764,
+      "step": 691
+    },
+    {
+      "epoch": 0.9123269611074489,
+      "grad_norm": 0.2732012552157185,
+      "learning_rate": 3.8685239491691106e-05,
+      "loss": 0.5928,
+      "step": 692
+    },
+    {
+      "epoch": 0.9136453526697429,
+      "grad_norm": 0.2386449116227092,
+      "learning_rate": 3.866080156402737e-05,
+      "loss": 0.5417,
+      "step": 693
+    },
+    {
+      "epoch": 0.914963744232037,
+      "grad_norm": 0.24930572847265844,
+      "learning_rate": 3.8636363636363636e-05,
+      "loss": 0.567,
+      "step": 694
+    },
+    {
+      "epoch": 0.9162821357943309,
+      "grad_norm": 0.2400916144935081,
+      "learning_rate": 3.86119257086999e-05,
+      "loss": 0.5128,
+      "step": 695
+    },
+    {
+      "epoch": 0.9176005273566249,
+      "grad_norm": 0.21228899660684725,
+      "learning_rate": 3.8587487781036166e-05,
+      "loss": 0.5513,
+      "step": 696
+    },
+    {
+      "epoch": 0.918918918918919,
+      "grad_norm": 0.24612804983070322,
+      "learning_rate": 3.856304985337244e-05,
+      "loss": 0.5553,
+      "step": 697
+    },
+    {
+      "epoch": 0.9202373104812129,
+      "grad_norm": 0.278841063860232,
+      "learning_rate": 3.85386119257087e-05,
+      "loss": 0.5845,
+      "step": 698
+    },
+    {
+      "epoch": 0.9215557020435069,
+      "grad_norm": 0.2548769552867638,
+      "learning_rate": 3.851417399804497e-05,
+      "loss": 0.5598,
+      "step": 699
+    },
+    {
+      "epoch": 0.922874093605801,
+      "grad_norm": 0.2864763865687837,
+      "learning_rate": 3.848973607038123e-05,
+      "loss": 0.548,
+      "step": 700
+    },
+    {
+      "epoch": 0.9241924851680949,
+      "grad_norm": 0.26161333432960604,
+      "learning_rate": 3.8465298142717496e-05,
+      "loss": 0.5783,
+      "step": 701
+    },
+    {
+      "epoch": 0.9255108767303889,
+      "grad_norm": 0.2949738268123761,
+      "learning_rate": 3.844086021505376e-05,
+      "loss": 0.6147,
+      "step": 702
+    },
+    {
+      "epoch": 0.926829268292683,
+      "grad_norm": 0.24832499055386176,
+      "learning_rate": 3.841642228739003e-05,
+      "loss": 0.5428,
+      "step": 703
+    },
+    {
+      "epoch": 0.9281476598549769,
+      "grad_norm": 0.277018541177385,
+      "learning_rate": 3.83919843597263e-05,
+      "loss": 0.5316,
+      "step": 704
+    },
+    {
+      "epoch": 0.9294660514172709,
+      "grad_norm": 0.2666200091608695,
+      "learning_rate": 3.836754643206256e-05,
+      "loss": 0.5874,
+      "step": 705
+    },
+    {
+      "epoch": 0.930784442979565,
+      "grad_norm": 0.26157956967651835,
+      "learning_rate": 3.834310850439883e-05,
+      "loss": 0.5751,
+      "step": 706
+    },
+    {
+      "epoch": 0.9321028345418589,
+      "grad_norm": 0.2874491200216373,
+      "learning_rate": 3.831867057673509e-05,
+      "loss": 0.5724,
+      "step": 707
+    },
+    {
+      "epoch": 0.9334212261041529,
+      "grad_norm": 0.27053108228012923,
+      "learning_rate": 3.829423264907136e-05,
+      "loss": 0.5769,
+      "step": 708
+    },
+    {
+      "epoch": 0.934739617666447,
+      "grad_norm": 0.2641118065860094,
+      "learning_rate": 3.826979472140763e-05,
+      "loss": 0.5595,
+      "step": 709
+    },
+    {
+      "epoch": 0.9360580092287409,
+      "grad_norm": 0.27215404493922674,
+      "learning_rate": 3.824535679374389e-05,
+      "loss": 0.6009,
+      "step": 710
+    },
+    {
+      "epoch": 0.9373764007910349,
+      "grad_norm": 0.26062062908049943,
+      "learning_rate": 3.822091886608016e-05,
+      "loss": 0.573,
+      "step": 711
+    },
+    {
+      "epoch": 0.938694792353329,
+      "grad_norm": 0.2744658306552036,
+      "learning_rate": 3.819648093841642e-05,
+      "loss": 0.578,
+      "step": 712
+    },
+    {
+      "epoch": 0.9400131839156229,
+      "grad_norm": 0.2458348468271122,
+      "learning_rate": 3.817204301075269e-05,
+      "loss": 0.5562,
+      "step": 713
+    },
+    {
+      "epoch": 0.9413315754779169,
+      "grad_norm": 0.2201252511286915,
+      "learning_rate": 3.814760508308895e-05,
+      "loss": 0.5833,
+      "step": 714
+    },
+    {
+      "epoch": 0.942649967040211,
+      "grad_norm": 0.2438588566528336,
+      "learning_rate": 3.8123167155425224e-05,
+      "loss": 0.5436,
+      "step": 715
+    },
+    {
+      "epoch": 0.9439683586025049,
+      "grad_norm": 0.21439513748560987,
+      "learning_rate": 3.809872922776149e-05,
+      "loss": 0.517,
+      "step": 716
+    },
+    {
+      "epoch": 0.9452867501647989,
+      "grad_norm": 0.2529144711246782,
+      "learning_rate": 3.8074291300097754e-05,
+      "loss": 0.5411,
+      "step": 717
+    },
+    {
+      "epoch": 0.946605141727093,
+      "grad_norm": 0.25679989732406483,
+      "learning_rate": 3.804985337243402e-05,
+      "loss": 0.5597,
+      "step": 718
+    },
+    {
+      "epoch": 0.9479235332893869,
+      "grad_norm": 0.2313872661865403,
+      "learning_rate": 3.802541544477028e-05,
+      "loss": 0.5517,
+      "step": 719
+    },
+    {
+      "epoch": 0.9492419248516809,
+      "grad_norm": 0.2319746665118741,
+      "learning_rate": 3.800097751710655e-05,
+      "loss": 0.5569,
+      "step": 720
+    },
+    {
+      "epoch": 0.950560316413975,
+      "grad_norm": 0.23679775956105037,
+      "learning_rate": 3.797653958944282e-05,
+      "loss": 0.5628,
+      "step": 721
+    },
+    {
+      "epoch": 0.951878707976269,
+      "grad_norm": 0.2373286969171995,
+      "learning_rate": 3.7952101661779084e-05,
+      "loss": 0.5637,
+      "step": 722
+    },
+    {
+      "epoch": 0.9531970995385629,
+      "grad_norm": 0.23086575927122563,
+      "learning_rate": 3.792766373411535e-05,
+      "loss": 0.5504,
+      "step": 723
+    },
+    {
+      "epoch": 0.954515491100857,
+      "grad_norm": 0.2328705906551139,
+      "learning_rate": 3.7903225806451614e-05,
+      "loss": 0.5684,
+      "step": 724
+    },
+    {
+      "epoch": 0.955833882663151,
+      "grad_norm": 0.23166380225332392,
+      "learning_rate": 3.787878787878788e-05,
+      "loss": 0.5579,
+      "step": 725
+    },
+    {
+      "epoch": 0.9571522742254449,
+      "grad_norm": 0.2389554169877941,
+      "learning_rate": 3.7854349951124144e-05,
+      "loss": 0.5743,
+      "step": 726
+    },
+    {
+      "epoch": 0.958470665787739,
+      "grad_norm": 0.244862355542489,
+      "learning_rate": 3.7829912023460415e-05,
+      "loss": 0.5501,
+      "step": 727
+    },
+    {
+      "epoch": 0.959789057350033,
+      "grad_norm": 0.23473121982996262,
+      "learning_rate": 3.780547409579668e-05,
+      "loss": 0.5245,
+      "step": 728
+    },
+    {
+      "epoch": 0.9611074489123269,
+      "grad_norm": 0.27406216507174075,
+      "learning_rate": 3.7781036168132945e-05,
+      "loss": 0.5369,
+      "step": 729
+    },
+    {
+      "epoch": 0.962425840474621,
+      "grad_norm": 0.255912726891568,
+      "learning_rate": 3.775659824046921e-05,
+      "loss": 0.5756,
+      "step": 730
+    },
+    {
+      "epoch": 0.963744232036915,
+      "grad_norm": 0.2505268350758867,
+      "learning_rate": 3.7732160312805474e-05,
+      "loss": 0.5538,
+      "step": 731
+    },
+    {
+      "epoch": 0.9650626235992089,
+      "grad_norm": 0.26135188015717015,
+      "learning_rate": 3.770772238514174e-05,
+      "loss": 0.6015,
+      "step": 732
+    },
+    {
+      "epoch": 0.966381015161503,
+      "grad_norm": 0.24657420977469607,
+      "learning_rate": 3.768328445747801e-05,
+      "loss": 0.5866,
+      "step": 733
+    },
+    {
+      "epoch": 0.967699406723797,
+      "grad_norm": 0.2657945958578638,
+      "learning_rate": 3.7658846529814276e-05,
+      "loss": 0.5463,
+      "step": 734
+    },
+    {
+      "epoch": 0.9690177982860909,
+      "grad_norm": 0.26146017578128394,
+      "learning_rate": 3.763440860215054e-05,
+      "loss": 0.5599,
+      "step": 735
+    },
+    {
+      "epoch": 0.970336189848385,
+      "grad_norm": 0.2354031694925808,
+      "learning_rate": 3.7609970674486805e-05,
+      "loss": 0.5517,
+      "step": 736
+    },
+    {
+      "epoch": 0.971654581410679,
+      "grad_norm": 0.22189333635774294,
+      "learning_rate": 3.758553274682307e-05,
+      "loss": 0.5345,
+      "step": 737
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 0.24912682256290014,
+      "learning_rate": 3.7561094819159335e-05,
+      "loss": 0.5574,
+      "step": 738
+    },
+    {
+      "epoch": 0.974291364535267,
+      "grad_norm": 0.2325956875017011,
+      "learning_rate": 3.7536656891495606e-05,
+      "loss": 0.5424,
+      "step": 739
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 0.23003568167062688,
+      "learning_rate": 3.751221896383187e-05,
+      "loss": 0.5342,
+      "step": 740
+    },
+    {
+      "epoch": 0.976928147659855,
+      "grad_norm": 0.2256145108857715,
+      "learning_rate": 3.7487781036168136e-05,
+      "loss": 0.5704,
+      "step": 741
+    },
+    {
+      "epoch": 0.978246539222149,
+      "grad_norm": 0.23431569670656302,
+      "learning_rate": 3.74633431085044e-05,
+      "loss": 0.5434,
+      "step": 742
+    },
+    {
+      "epoch": 0.979564930784443,
+      "grad_norm": 0.24132668758915002,
+      "learning_rate": 3.7438905180840665e-05,
+      "loss": 0.5638,
+      "step": 743
+    },
+    {
+      "epoch": 0.980883322346737,
+      "grad_norm": 0.2387737370641021,
+      "learning_rate": 3.741446725317693e-05,
+      "loss": 0.5369,
+      "step": 744
+    },
+    {
+      "epoch": 0.982201713909031,
+      "grad_norm": 0.2406908117690978,
+      "learning_rate": 3.7390029325513195e-05,
+      "loss": 0.5714,
+      "step": 745
+    },
+    {
+      "epoch": 0.983520105471325,
+      "grad_norm": 0.2426417597061762,
+      "learning_rate": 3.736559139784947e-05,
+      "loss": 0.5535,
+      "step": 746
+    },
+    {
+      "epoch": 0.984838497033619,
+      "grad_norm": 0.2743483739324856,
+      "learning_rate": 3.734115347018573e-05,
+      "loss": 0.5394,
+      "step": 747
+    },
+    {
+      "epoch": 0.986156888595913,
+      "grad_norm": 0.23952145639956465,
+      "learning_rate": 3.7316715542521996e-05,
+      "loss": 0.5652,
+      "step": 748
+    },
+    {
+      "epoch": 0.987475280158207,
+      "grad_norm": 0.23484248601713406,
+      "learning_rate": 3.729227761485826e-05,
+      "loss": 0.5535,
+      "step": 749
+    },
+    {
+      "epoch": 0.988793671720501,
+      "grad_norm": 0.27719500782981377,
+      "learning_rate": 3.7267839687194526e-05,
+      "loss": 0.5758,
+      "step": 750
+    },
+    {
+      "epoch": 0.990112063282795,
+      "grad_norm": 0.2634739454894618,
+      "learning_rate": 3.724340175953079e-05,
+      "loss": 0.5783,
+      "step": 751
+    },
+    {
+      "epoch": 0.991430454845089,
+      "grad_norm": 0.2688217523423705,
+      "learning_rate": 3.721896383186706e-05,
+      "loss": 0.6054,
+      "step": 752
+    },
+    {
+      "epoch": 0.992748846407383,
+      "grad_norm": 0.262893447510181,
+      "learning_rate": 3.719452590420333e-05,
+      "loss": 0.5609,
+      "step": 753
+    },
+    {
+      "epoch": 0.994067237969677,
+      "grad_norm": 0.2186218898444037,
+      "learning_rate": 3.717008797653959e-05,
+      "loss": 0.5236,
+      "step": 754
+    },
+    {
+      "epoch": 0.995385629531971,
+      "grad_norm": 0.262545743630205,
+      "learning_rate": 3.714565004887586e-05,
+      "loss": 0.5388,
+      "step": 755
+    },
+    {
+      "epoch": 0.996704021094265,
+      "grad_norm": 0.24371455011280926,
+      "learning_rate": 3.712121212121212e-05,
+      "loss": 0.5836,
+      "step": 756
+    },
+    {
+      "epoch": 0.998022412656559,
+      "grad_norm": 0.23961062101680805,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.5579,
+      "step": 757
+    },
+    {
+      "epoch": 0.999340804218853,
+      "grad_norm": 0.2542922514087688,
+      "learning_rate": 3.707233626588466e-05,
+      "loss": 0.5584,
+      "step": 758
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2542922514087688,
+      "learning_rate": 3.704789833822092e-05,
+      "loss": 0.5026,
+      "step": 759
+    },
+    {
+      "epoch": 1.001318391562294,
+      "grad_norm": 0.35227998468050725,
+      "learning_rate": 3.702346041055719e-05,
+      "loss": 0.4793,
+      "step": 760
+    },
+    {
+      "epoch": 1.002636783124588,
+      "grad_norm": 0.3171156924195792,
+      "learning_rate": 3.699902248289345e-05,
+      "loss": 0.4739,
+      "step": 761
+    },
+    {
+      "epoch": 1.003955174686882,
+      "grad_norm": 0.2797789945491188,
+      "learning_rate": 3.697458455522972e-05,
+      "loss": 0.4995,
+      "step": 762
+    },
+    {
+      "epoch": 1.005273566249176,
+      "grad_norm": 0.3087106361987999,
+      "learning_rate": 3.695014662756598e-05,
+      "loss": 0.4882,
+      "step": 763
+    },
+    {
+      "epoch": 1.00659195781147,
+      "grad_norm": 0.28716003063255724,
+      "learning_rate": 3.6925708699902253e-05,
+      "loss": 0.4699,
+      "step": 764
+    },
+    {
+      "epoch": 1.007910349373764,
+      "grad_norm": 0.2883156122542544,
+      "learning_rate": 3.690127077223852e-05,
+      "loss": 0.5007,
+      "step": 765
+    },
+    {
+      "epoch": 1.009228740936058,
+      "grad_norm": 0.2833285717579239,
+      "learning_rate": 3.687683284457478e-05,
+      "loss": 0.4745,
+      "step": 766
+    },
+    {
+      "epoch": 1.010547132498352,
+      "grad_norm": 0.2775788623126738,
+      "learning_rate": 3.685239491691105e-05,
+      "loss": 0.4757,
+      "step": 767
+    },
+    {
+      "epoch": 1.011865524060646,
+      "grad_norm": 0.2651467221689858,
+      "learning_rate": 3.682795698924731e-05,
+      "loss": 0.4514,
+      "step": 768
+    },
+    {
+      "epoch": 1.01318391562294,
+      "grad_norm": 0.29839235934220965,
+      "learning_rate": 3.680351906158358e-05,
+      "loss": 0.4767,
+      "step": 769
+    },
+    {
+      "epoch": 1.014502307185234,
+      "grad_norm": 0.29488950737931585,
+      "learning_rate": 3.677908113391985e-05,
+      "loss": 0.4691,
+      "step": 770
+    },
+    {
+      "epoch": 1.015820698747528,
+      "grad_norm": 0.27367786180264003,
+      "learning_rate": 3.6754643206256114e-05,
+      "loss": 0.467,
+      "step": 771
+    },
+    {
+      "epoch": 1.017139090309822,
+      "grad_norm": 0.29417272000046657,
+      "learning_rate": 3.673020527859238e-05,
+      "loss": 0.4934,
+      "step": 772
+    },
+    {
+      "epoch": 1.018457481872116,
+      "grad_norm": 0.26926469947835735,
+      "learning_rate": 3.670576735092864e-05,
+      "loss": 0.492,
+      "step": 773
+    },
+    {
+      "epoch": 1.01977587343441,
+      "grad_norm": 1.093544910745387,
+      "learning_rate": 3.668132942326491e-05,
+      "loss": 0.482,
+      "step": 774
+    },
+    {
+      "epoch": 1.021094264996704,
+      "grad_norm": 0.28743908490664805,
+      "learning_rate": 3.665689149560117e-05,
+      "loss": 0.4951,
+      "step": 775
+    },
+    {
+      "epoch": 1.022412656558998,
+      "grad_norm": 0.26331091296141146,
+      "learning_rate": 3.6632453567937445e-05,
+      "loss": 0.519,
+      "step": 776
+    },
+    {
+      "epoch": 1.023731048121292,
+      "grad_norm": 0.2543506493517187,
+      "learning_rate": 3.660801564027371e-05,
+      "loss": 0.5238,
+      "step": 777
+    },
+    {
+      "epoch": 1.025049439683586,
+      "grad_norm": 0.26167901787823467,
+      "learning_rate": 3.6583577712609974e-05,
+      "loss": 0.4782,
+      "step": 778
+    },
+    {
+      "epoch": 1.02636783124588,
+      "grad_norm": 0.28560366301674545,
+      "learning_rate": 3.655913978494624e-05,
+      "loss": 0.473,
+      "step": 779
+    },
+    {
+      "epoch": 1.027686222808174,
+      "grad_norm": 0.23388119737485777,
+      "learning_rate": 3.6534701857282504e-05,
+      "loss": 0.4856,
+      "step": 780
+    },
+    {
+      "epoch": 1.029004614370468,
+      "grad_norm": 0.24655330710779644,
+      "learning_rate": 3.651026392961877e-05,
+      "loss": 0.5076,
+      "step": 781
+    },
+    {
+      "epoch": 1.030323005932762,
+      "grad_norm": 0.2575926808260151,
+      "learning_rate": 3.648582600195504e-05,
+      "loss": 0.476,
+      "step": 782
+    },
+    {
+      "epoch": 1.031641397495056,
+      "grad_norm": 0.24379287719822396,
+      "learning_rate": 3.6461388074291305e-05,
+      "loss": 0.5196,
+      "step": 783
+    },
+    {
+      "epoch": 1.03295978905735,
+      "grad_norm": 0.22067290124006286,
+      "learning_rate": 3.643695014662757e-05,
+      "loss": 0.4691,
+      "step": 784
+    },
+    {
+      "epoch": 1.034278180619644,
+      "grad_norm": 0.2825611100816932,
+      "learning_rate": 3.6412512218963835e-05,
+      "loss": 0.4802,
+      "step": 785
+    },
+    {
+      "epoch": 1.035596572181938,
+      "grad_norm": 0.2201826902464278,
+      "learning_rate": 3.63880742913001e-05,
+      "loss": 0.4359,
+      "step": 786
+    },
+    {
+      "epoch": 1.036914963744232,
+      "grad_norm": 0.24304361960984122,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 0.4844,
+      "step": 787
+    },
+    {
+      "epoch": 1.038233355306526,
+      "grad_norm": 0.2612328701403773,
+      "learning_rate": 3.633919843597263e-05,
+      "loss": 0.4884,
+      "step": 788
+    },
+    {
+      "epoch": 1.03955174686882,
+      "grad_norm": 0.28372181328058116,
+      "learning_rate": 3.63147605083089e-05,
+      "loss": 0.4517,
+      "step": 789
+    },
+    {
+      "epoch": 1.040870138431114,
+      "grad_norm": 0.2501403024747959,
+      "learning_rate": 3.6290322580645165e-05,
+      "loss": 0.4794,
+      "step": 790
+    },
+    {
+      "epoch": 1.042188529993408,
+      "grad_norm": 0.27687203135368715,
+      "learning_rate": 3.626588465298143e-05,
+      "loss": 0.502,
+      "step": 791
+    },
+    {
+      "epoch": 1.043506921555702,
+      "grad_norm": 0.26132241213905255,
+      "learning_rate": 3.6241446725317695e-05,
+      "loss": 0.4849,
+      "step": 792
+    },
+    {
+      "epoch": 1.044825313117996,
+      "grad_norm": 0.2855853356790974,
+      "learning_rate": 3.621700879765396e-05,
+      "loss": 0.4581,
+      "step": 793
+    },
+    {
+      "epoch": 1.04614370468029,
+      "grad_norm": 0.27555486723433936,
+      "learning_rate": 3.6192570869990225e-05,
+      "loss": 0.4455,
+      "step": 794
+    },
+    {
+      "epoch": 1.047462096242584,
+      "grad_norm": 0.27267375700585395,
+      "learning_rate": 3.6168132942326496e-05,
+      "loss": 0.4795,
+      "step": 795
+    },
+    {
+      "epoch": 1.048780487804878,
+      "grad_norm": 0.2805149238014777,
+      "learning_rate": 3.614369501466276e-05,
+      "loss": 0.4525,
+      "step": 796
+    },
+    {
+      "epoch": 1.050098879367172,
+      "grad_norm": 0.2443118652121828,
+      "learning_rate": 3.6119257086999026e-05,
+      "loss": 0.4876,
+      "step": 797
+    },
+    {
+      "epoch": 1.051417270929466,
+      "grad_norm": 0.2937661758277491,
+      "learning_rate": 3.609481915933529e-05,
+      "loss": 0.4999,
+      "step": 798
+    },
+    {
+      "epoch": 1.05273566249176,
+      "grad_norm": 0.24305401053436632,
+      "learning_rate": 3.6070381231671555e-05,
+      "loss": 0.4855,
+      "step": 799
+    },
+    {
+      "epoch": 1.054054054054054,
+      "grad_norm": 0.2606953932514111,
+      "learning_rate": 3.604594330400782e-05,
+      "loss": 0.4515,
+      "step": 800
+    },
+    {
+      "epoch": 1.055372445616348,
+      "grad_norm": 0.23813492926647362,
+      "learning_rate": 3.602150537634409e-05,
+      "loss": 0.4823,
+      "step": 801
+    },
+    {
+      "epoch": 1.056690837178642,
+      "grad_norm": 0.2682935617860416,
+      "learning_rate": 3.5997067448680356e-05,
+      "loss": 0.4525,
+      "step": 802
+    },
+    {
+      "epoch": 1.0580092287409362,
+      "grad_norm": 0.22757782184643785,
+      "learning_rate": 3.597262952101662e-05,
+      "loss": 0.4612,
+      "step": 803
+    },
+    {
+      "epoch": 1.05932762030323,
+      "grad_norm": 0.22193727019158713,
+      "learning_rate": 3.5948191593352886e-05,
+      "loss": 0.4684,
+      "step": 804
+    },
+    {
+      "epoch": 1.060646011865524,
+      "grad_norm": 0.28435407510931454,
+      "learning_rate": 3.592375366568915e-05,
+      "loss": 0.4864,
+      "step": 805
+    },
+    {
+      "epoch": 1.061964403427818,
+      "grad_norm": 0.23905823474210838,
+      "learning_rate": 3.5899315738025416e-05,
+      "loss": 0.4787,
+      "step": 806
+    },
+    {
+      "epoch": 1.063282794990112,
+      "grad_norm": 0.29167782086254906,
+      "learning_rate": 3.587487781036169e-05,
+      "loss": 0.5056,
+      "step": 807
+    },
+    {
+      "epoch": 1.064601186552406,
+      "grad_norm": 0.22426631107649136,
+      "learning_rate": 3.585043988269795e-05,
+      "loss": 0.4969,
+      "step": 808
+    },
+    {
+      "epoch": 1.0659195781147002,
+      "grad_norm": 0.24956939791705882,
+      "learning_rate": 3.582600195503422e-05,
+      "loss": 0.463,
+      "step": 809
+    },
+    {
+      "epoch": 1.067237969676994,
+      "grad_norm": 0.24437206339621906,
+      "learning_rate": 3.580156402737048e-05,
+      "loss": 0.4867,
+      "step": 810
+    },
+    {
+      "epoch": 1.068556361239288,
+      "grad_norm": 0.26829271276977296,
+      "learning_rate": 3.5777126099706746e-05,
+      "loss": 0.4926,
+      "step": 811
+    },
+    {
+      "epoch": 1.0698747528015822,
+      "grad_norm": 0.2694486399496288,
+      "learning_rate": 3.575268817204301e-05,
+      "loss": 0.4581,
+      "step": 812
+    },
+    {
+      "epoch": 1.071193144363876,
+      "grad_norm": 0.2314407432265082,
+      "learning_rate": 3.572825024437928e-05,
+      "loss": 0.4641,
+      "step": 813
+    },
+    {
+      "epoch": 1.07251153592617,
+      "grad_norm": 0.26290930348716496,
+      "learning_rate": 3.570381231671555e-05,
+      "loss": 0.4507,
+      "step": 814
+    },
+    {
+      "epoch": 1.0738299274884642,
+      "grad_norm": 0.27464211019231866,
+      "learning_rate": 3.567937438905181e-05,
+      "loss": 0.47,
+      "step": 815
+    },
+    {
+      "epoch": 1.075148319050758,
+      "grad_norm": 0.24392177528982625,
+      "learning_rate": 3.565493646138808e-05,
+      "loss": 0.4713,
+      "step": 816
+    },
+    {
+      "epoch": 1.076466710613052,
+      "grad_norm": 0.2560556934611447,
+      "learning_rate": 3.563049853372434e-05,
+      "loss": 0.4672,
+      "step": 817
+    },
+    {
+      "epoch": 1.0777851021753462,
+      "grad_norm": 0.28913783375399005,
+      "learning_rate": 3.560606060606061e-05,
+      "loss": 0.484,
+      "step": 818
+    },
+    {
+      "epoch": 1.07910349373764,
+      "grad_norm": 0.24544843565837038,
+      "learning_rate": 3.558162267839688e-05,
+      "loss": 0.4634,
+      "step": 819
+    },
+    {
+      "epoch": 1.080421885299934,
+      "grad_norm": 0.24983267701204717,
+      "learning_rate": 3.555718475073314e-05,
+      "loss": 0.4795,
+      "step": 820
+    },
+    {
+      "epoch": 1.0817402768622282,
+      "grad_norm": 0.24657234808961817,
+      "learning_rate": 3.553274682306941e-05,
+      "loss": 0.4413,
+      "step": 821
+    },
+    {
+      "epoch": 1.083058668424522,
+      "grad_norm": 0.3070453037823855,
+      "learning_rate": 3.550830889540567e-05,
+      "loss": 0.4652,
+      "step": 822
+    },
+    {
+      "epoch": 1.084377059986816,
+      "grad_norm": 0.21918138380181393,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.4466,
+      "step": 823
+    },
+    {
+      "epoch": 1.0856954515491102,
+      "grad_norm": 0.26751804780305344,
+      "learning_rate": 3.54594330400782e-05,
+      "loss": 0.5063,
+      "step": 824
+    },
+    {
+      "epoch": 1.087013843111404,
+      "grad_norm": 0.24704146774317437,
+      "learning_rate": 3.5434995112414474e-05,
+      "loss": 0.4604,
+      "step": 825
+    },
+    {
+      "epoch": 1.088332234673698,
+      "grad_norm": 0.24733471230027887,
+      "learning_rate": 3.541055718475074e-05,
+      "loss": 0.481,
+      "step": 826
+    },
+    {
+      "epoch": 1.0896506262359922,
+      "grad_norm": 0.24778802879808934,
+      "learning_rate": 3.5386119257087004e-05,
+      "loss": 0.488,
+      "step": 827
+    },
+    {
+      "epoch": 1.090969017798286,
+      "grad_norm": 0.2401309674284125,
+      "learning_rate": 3.536168132942327e-05,
+      "loss": 0.491,
+      "step": 828
+    },
+    {
+      "epoch": 1.09228740936058,
+      "grad_norm": 0.23562704291147082,
+      "learning_rate": 3.533724340175953e-05,
+      "loss": 0.4421,
+      "step": 829
+    },
+    {
+      "epoch": 1.0936058009228742,
+      "grad_norm": 0.2319868655040486,
+      "learning_rate": 3.53128054740958e-05,
+      "loss": 0.5261,
+      "step": 830
+    },
+    {
+      "epoch": 1.094924192485168,
+      "grad_norm": 0.26346128294007864,
+      "learning_rate": 3.528836754643207e-05,
+      "loss": 0.4406,
+      "step": 831
+    },
+    {
+      "epoch": 1.096242584047462,
+      "grad_norm": 0.21951855515760837,
+      "learning_rate": 3.5263929618768334e-05,
+      "loss": 0.4718,
+      "step": 832
+    },
+    {
+      "epoch": 1.0975609756097562,
+      "grad_norm": 0.284551935257363,
+      "learning_rate": 3.52394916911046e-05,
+      "loss": 0.4944,
+      "step": 833
+    },
+    {
+      "epoch": 1.09887936717205,
+      "grad_norm": 0.21292637020743369,
+      "learning_rate": 3.5215053763440864e-05,
+      "loss": 0.4804,
+      "step": 834
+    },
+    {
+      "epoch": 1.1001977587343441,
+      "grad_norm": 0.27688281712619295,
+      "learning_rate": 3.519061583577713e-05,
+      "loss": 0.4724,
+      "step": 835
+    },
+    {
+      "epoch": 1.1015161502966382,
+      "grad_norm": 0.2444249940727895,
+      "learning_rate": 3.5166177908113394e-05,
+      "loss": 0.5235,
+      "step": 836
+    },
+    {
+      "epoch": 1.102834541858932,
+      "grad_norm": 0.23513157535578333,
+      "learning_rate": 3.514173998044966e-05,
+      "loss": 0.4432,
+      "step": 837
+    },
+    {
+      "epoch": 1.1041529334212261,
+      "grad_norm": 0.23869288544878528,
+      "learning_rate": 3.511730205278593e-05,
+      "loss": 0.4656,
+      "step": 838
+    },
+    {
+      "epoch": 1.1054713249835202,
+      "grad_norm": 0.26665874658701433,
+      "learning_rate": 3.5092864125122195e-05,
+      "loss": 0.4694,
+      "step": 839
+    },
+    {
+      "epoch": 1.106789716545814,
+      "grad_norm": 0.24767781066831773,
+      "learning_rate": 3.506842619745846e-05,
+      "loss": 0.4813,
+      "step": 840
+    },
+    {
+      "epoch": 1.1081081081081081,
+      "grad_norm": 0.22735984096588405,
+      "learning_rate": 3.5043988269794724e-05,
+      "loss": 0.4579,
+      "step": 841
+    },
+    {
+      "epoch": 1.1094264996704022,
+      "grad_norm": 0.23362897732543694,
+      "learning_rate": 3.501955034213099e-05,
+      "loss": 0.4719,
+      "step": 842
+    },
+    {
+      "epoch": 1.110744891232696,
+      "grad_norm": 0.257684065345504,
+      "learning_rate": 3.4995112414467254e-05,
+      "loss": 0.4433,
+      "step": 843
+    },
+    {
+      "epoch": 1.1120632827949901,
+      "grad_norm": 0.2153642646483253,
+      "learning_rate": 3.4970674486803526e-05,
+      "loss": 0.4674,
+      "step": 844
+    },
+    {
+      "epoch": 1.1133816743572842,
+      "grad_norm": 0.2498091498251156,
+      "learning_rate": 3.494623655913979e-05,
+      "loss": 0.4527,
+      "step": 845
+    },
+    {
+      "epoch": 1.114700065919578,
+      "grad_norm": 0.2406987479572704,
+      "learning_rate": 3.4921798631476055e-05,
+      "loss": 0.4611,
+      "step": 846
+    },
+    {
+      "epoch": 1.1160184574818721,
+      "grad_norm": 0.24013024931793053,
+      "learning_rate": 3.489736070381232e-05,
+      "loss": 0.4615,
+      "step": 847
+    },
+    {
+      "epoch": 1.1173368490441662,
+      "grad_norm": 0.24467806920645313,
+      "learning_rate": 3.4872922776148585e-05,
+      "loss": 0.5241,
+      "step": 848
+    },
+    {
+      "epoch": 1.11865524060646,
+      "grad_norm": 0.2605027525397413,
+      "learning_rate": 3.484848484848485e-05,
+      "loss": 0.4831,
+      "step": 849
+    },
+    {
+      "epoch": 1.1199736321687541,
+      "grad_norm": 0.23064607324880076,
+      "learning_rate": 3.482404692082112e-05,
+      "loss": 0.495,
+      "step": 850
+    },
+    {
+      "epoch": 1.1212920237310482,
+      "grad_norm": 0.25604658757475063,
+      "learning_rate": 3.4799608993157386e-05,
+      "loss": 0.4902,
+      "step": 851
+    },
+    {
+      "epoch": 1.122610415293342,
+      "grad_norm": 0.22850023698876337,
+      "learning_rate": 3.477517106549365e-05,
+      "loss": 0.4798,
+      "step": 852
+    },
+    {
+      "epoch": 1.1239288068556361,
+      "grad_norm": 0.2389703758119579,
+      "learning_rate": 3.4750733137829916e-05,
+      "loss": 0.4601,
+      "step": 853
+    },
+    {
+      "epoch": 1.1252471984179302,
+      "grad_norm": 0.23482866919395196,
+      "learning_rate": 3.472629521016618e-05,
+      "loss": 0.4833,
+      "step": 854
+    },
+    {
+      "epoch": 1.126565589980224,
+      "grad_norm": 0.2409779462702182,
+      "learning_rate": 3.4701857282502445e-05,
+      "loss": 0.4555,
+      "step": 855
+    },
+    {
+      "epoch": 1.1278839815425181,
+      "grad_norm": 0.2638885491545315,
+      "learning_rate": 3.467741935483872e-05,
+      "loss": 0.4845,
+      "step": 856
+    },
+    {
+      "epoch": 1.1292023731048122,
+      "grad_norm": 0.2517445569986723,
+      "learning_rate": 3.465298142717498e-05,
+      "loss": 0.4666,
+      "step": 857
+    },
+    {
+      "epoch": 1.130520764667106,
+      "grad_norm": 0.25014758403433646,
+      "learning_rate": 3.4628543499511246e-05,
+      "loss": 0.5026,
+      "step": 858
+    },
+    {
+      "epoch": 1.1318391562294001,
+      "grad_norm": 0.2521997874025587,
+      "learning_rate": 3.460410557184751e-05,
+      "loss": 0.4806,
+      "step": 859
+    },
+    {
+      "epoch": 1.1331575477916942,
+      "grad_norm": 0.2555206213317847,
+      "learning_rate": 3.4579667644183776e-05,
+      "loss": 0.4782,
+      "step": 860
+    },
+    {
+      "epoch": 1.134475939353988,
+      "grad_norm": 0.25789243207049745,
+      "learning_rate": 3.455522971652004e-05,
+      "loss": 0.4677,
+      "step": 861
+    },
+    {
+      "epoch": 1.1357943309162821,
+      "grad_norm": 0.24701710379989944,
+      "learning_rate": 3.453079178885631e-05,
+      "loss": 0.469,
+      "step": 862
+    },
+    {
+      "epoch": 1.1371127224785762,
+      "grad_norm": 0.27324472243494113,
+      "learning_rate": 3.450635386119258e-05,
+      "loss": 0.4828,
+      "step": 863
+    },
+    {
+      "epoch": 1.13843111404087,
+      "grad_norm": 0.26681972667964243,
+      "learning_rate": 3.448191593352884e-05,
+      "loss": 0.4622,
+      "step": 864
+    },
+    {
+      "epoch": 1.1397495056031641,
+      "grad_norm": 0.23890499836772744,
+      "learning_rate": 3.445747800586511e-05,
+      "loss": 0.4908,
+      "step": 865
+    },
+    {
+      "epoch": 1.1410678971654582,
+      "grad_norm": 0.2465061161395173,
+      "learning_rate": 3.443304007820137e-05,
+      "loss": 0.5078,
+      "step": 866
+    },
+    {
+      "epoch": 1.142386288727752,
+      "grad_norm": 0.273818191439379,
+      "learning_rate": 3.4408602150537636e-05,
+      "loss": 0.4694,
+      "step": 867
+    },
+    {
+      "epoch": 1.1437046802900461,
+      "grad_norm": 0.23665491337437958,
+      "learning_rate": 3.438416422287391e-05,
+      "loss": 0.4758,
+      "step": 868
+    },
+    {
+      "epoch": 1.1450230718523402,
+      "grad_norm": 0.2316790867141602,
+      "learning_rate": 3.4359726295210166e-05,
+      "loss": 0.4536,
+      "step": 869
+    },
+    {
+      "epoch": 1.146341463414634,
+      "grad_norm": 0.2557885535837887,
+      "learning_rate": 3.433528836754643e-05,
+      "loss": 0.4672,
+      "step": 870
+    },
+    {
+      "epoch": 1.1476598549769281,
+      "grad_norm": 0.2209031285912192,
+      "learning_rate": 3.4310850439882695e-05,
+      "loss": 0.478,
+      "step": 871
+    },
+    {
+      "epoch": 1.1489782465392222,
+      "grad_norm": 0.22345402029112227,
+      "learning_rate": 3.428641251221896e-05,
+      "loss": 0.4721,
+      "step": 872
+    },
+    {
+      "epoch": 1.150296638101516,
+      "grad_norm": 0.2435593848276839,
+      "learning_rate": 3.426197458455523e-05,
+      "loss": 0.4504,
+      "step": 873
+    },
+    {
+      "epoch": 1.1516150296638101,
+      "grad_norm": 0.23463165934348576,
+      "learning_rate": 3.42375366568915e-05,
+      "loss": 0.4731,
+      "step": 874
+    },
+    {
+      "epoch": 1.1529334212261042,
+      "grad_norm": 0.22413373189817168,
+      "learning_rate": 3.421309872922776e-05,
+      "loss": 0.5119,
+      "step": 875
+    },
+    {
+      "epoch": 1.154251812788398,
+      "grad_norm": 0.2514651509172245,
+      "learning_rate": 3.4188660801564026e-05,
+      "loss": 0.4756,
+      "step": 876
+    },
+    {
+      "epoch": 1.1555702043506921,
+      "grad_norm": 0.21320205785518553,
+      "learning_rate": 3.416422287390029e-05,
+      "loss": 0.4669,
+      "step": 877
+    },
+    {
+      "epoch": 1.1568885959129862,
+      "grad_norm": 0.23119764810086663,
+      "learning_rate": 3.4139784946236556e-05,
+      "loss": 0.4623,
+      "step": 878
+    },
+    {
+      "epoch": 1.15820698747528,
+      "grad_norm": 0.28812218362116415,
+      "learning_rate": 3.411534701857283e-05,
+      "loss": 0.4929,
+      "step": 879
+    },
+    {
+      "epoch": 1.1595253790375741,
+      "grad_norm": 0.2541819685947355,
+      "learning_rate": 3.409090909090909e-05,
+      "loss": 0.4722,
+      "step": 880
+    },
+    {
+      "epoch": 1.1608437705998682,
+      "grad_norm": 0.21719520531863404,
+      "learning_rate": 3.406647116324536e-05,
+      "loss": 0.4784,
+      "step": 881
+    },
+    {
+      "epoch": 1.1621621621621623,
+      "grad_norm": 0.25831153050776723,
+      "learning_rate": 3.404203323558162e-05,
+      "loss": 0.4922,
+      "step": 882
+    },
+    {
+      "epoch": 1.1634805537244561,
+      "grad_norm": 0.26761150786265914,
+      "learning_rate": 3.401759530791789e-05,
+      "loss": 0.456,
+      "step": 883
+    },
+    {
+      "epoch": 1.1647989452867502,
+      "grad_norm": 0.2434666942170409,
+      "learning_rate": 3.399315738025415e-05,
+      "loss": 0.4959,
+      "step": 884
+    },
+    {
+      "epoch": 1.166117336849044,
+      "grad_norm": 0.239658569056963,
+      "learning_rate": 3.3968719452590416e-05,
+      "loss": 0.4472,
+      "step": 885
+    },
+    {
+      "epoch": 1.1674357284113381,
+      "grad_norm": 0.23574148366265477,
+      "learning_rate": 3.394428152492669e-05,
+      "loss": 0.4709,
+      "step": 886
+    },
+    {
+      "epoch": 1.1687541199736322,
+      "grad_norm": 0.27540558094612266,
+      "learning_rate": 3.391984359726295e-05,
+      "loss": 0.4687,
+      "step": 887
+    },
+    {
+      "epoch": 1.1700725115359263,
+      "grad_norm": 0.23235237437861023,
+      "learning_rate": 3.389540566959922e-05,
+      "loss": 0.4808,
+      "step": 888
+    },
+    {
+      "epoch": 1.1713909030982201,
+      "grad_norm": 0.29004863649636103,
+      "learning_rate": 3.387096774193548e-05,
+      "loss": 0.487,
+      "step": 889
+    },
+    {
+      "epoch": 1.1727092946605142,
+      "grad_norm": 0.25492282815207773,
+      "learning_rate": 3.384652981427175e-05,
+      "loss": 0.4719,
+      "step": 890
+    },
+    {
+      "epoch": 1.174027686222808,
+      "grad_norm": 0.25339544964659144,
+      "learning_rate": 3.382209188660801e-05,
+      "loss": 0.4655,
+      "step": 891
+    },
+    {
+      "epoch": 1.1753460777851021,
+      "grad_norm": 0.2923622086246815,
+      "learning_rate": 3.3797653958944283e-05,
+      "loss": 0.4512,
+      "step": 892
+    },
+    {
+      "epoch": 1.1766644693473962,
+      "grad_norm": 0.23485997501839004,
+      "learning_rate": 3.377321603128055e-05,
+      "loss": 0.4905,
+      "step": 893
+    },
+    {
+      "epoch": 1.1779828609096903,
+      "grad_norm": 0.2668589551578169,
+      "learning_rate": 3.374877810361681e-05,
+      "loss": 0.4562,
+      "step": 894
+    },
+    {
+      "epoch": 1.1793012524719841,
+      "grad_norm": 0.29876498970775006,
+      "learning_rate": 3.372434017595308e-05,
+      "loss": 0.4737,
+      "step": 895
+    },
+    {
+      "epoch": 1.1806196440342782,
+      "grad_norm": 0.25034217325235375,
+      "learning_rate": 3.369990224828934e-05,
+      "loss": 0.4702,
+      "step": 896
+    },
+    {
+      "epoch": 1.181938035596572,
+      "grad_norm": 0.2397418839678866,
+      "learning_rate": 3.367546432062561e-05,
+      "loss": 0.5135,
+      "step": 897
+    },
+    {
+      "epoch": 1.1832564271588661,
+      "grad_norm": 0.3081421438906284,
+      "learning_rate": 3.365102639296188e-05,
+      "loss": 0.4653,
+      "step": 898
+    },
+    {
+      "epoch": 1.1845748187211602,
+      "grad_norm": 0.24727707242485086,
+      "learning_rate": 3.3626588465298144e-05,
+      "loss": 0.4783,
+      "step": 899
+    },
+    {
+      "epoch": 1.1858932102834543,
+      "grad_norm": 0.22826905107327558,
+      "learning_rate": 3.360215053763441e-05,
+      "loss": 0.4727,
+      "step": 900
+    },
+    {
+      "epoch": 1.1872116018457481,
+      "grad_norm": 0.27541041382128484,
+      "learning_rate": 3.357771260997067e-05,
+      "loss": 0.4832,
+      "step": 901
+    },
+    {
+      "epoch": 1.1885299934080422,
+      "grad_norm": 0.248586646875075,
+      "learning_rate": 3.355327468230694e-05,
+      "loss": 0.4598,
+      "step": 902
+    },
+    {
+      "epoch": 1.189848384970336,
+      "grad_norm": 0.24184065854446118,
+      "learning_rate": 3.35288367546432e-05,
+      "loss": 0.4533,
+      "step": 903
+    },
+    {
+      "epoch": 1.1911667765326301,
+      "grad_norm": 0.25570654011306804,
+      "learning_rate": 3.3504398826979475e-05,
+      "loss": 0.4899,
+      "step": 904
+    },
+    {
+      "epoch": 1.1924851680949242,
+      "grad_norm": 0.2719461199275369,
+      "learning_rate": 3.347996089931574e-05,
+      "loss": 0.47,
+      "step": 905
+    },
+    {
+      "epoch": 1.1938035596572183,
+      "grad_norm": 0.22499434599949472,
+      "learning_rate": 3.3455522971652004e-05,
+      "loss": 0.4776,
+      "step": 906
+    },
+    {
+      "epoch": 1.1951219512195121,
+      "grad_norm": 0.23070571525177377,
+      "learning_rate": 3.343108504398827e-05,
+      "loss": 0.4688,
+      "step": 907
+    },
+    {
+      "epoch": 1.1964403427818062,
+      "grad_norm": 0.23455211571368678,
+      "learning_rate": 3.3406647116324534e-05,
+      "loss": 0.4462,
+      "step": 908
+    },
+    {
+      "epoch": 1.1977587343441,
+      "grad_norm": 0.24942772991204318,
+      "learning_rate": 3.33822091886608e-05,
+      "loss": 0.4632,
+      "step": 909
+    },
+    {
+      "epoch": 1.1990771259063941,
+      "grad_norm": 0.23062971413845482,
+      "learning_rate": 3.335777126099707e-05,
+      "loss": 0.4611,
+      "step": 910
+    },
+    {
+      "epoch": 1.2003955174686882,
+      "grad_norm": 0.2567728879795924,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.4818,
+      "step": 911
+    },
+    {
+      "epoch": 1.2017139090309823,
+      "grad_norm": 0.24677823338045987,
+      "learning_rate": 3.33088954056696e-05,
+      "loss": 0.4965,
+      "step": 912
+    },
+    {
+      "epoch": 1.2030323005932761,
+      "grad_norm": 0.2607245273008426,
+      "learning_rate": 3.3284457478005865e-05,
+      "loss": 0.4681,
+      "step": 913
+    },
+    {
+      "epoch": 1.2043506921555702,
+      "grad_norm": 0.2471871025402272,
+      "learning_rate": 3.326001955034213e-05,
+      "loss": 0.4949,
+      "step": 914
+    },
+    {
+      "epoch": 1.2056690837178643,
+      "grad_norm": 0.23572188149499484,
+      "learning_rate": 3.3235581622678394e-05,
+      "loss": 0.487,
+      "step": 915
+    },
+    {
+      "epoch": 1.2069874752801582,
+      "grad_norm": 0.2856763925724387,
+      "learning_rate": 3.3211143695014666e-05,
+      "loss": 0.457,
+      "step": 916
+    },
+    {
+      "epoch": 1.2083058668424522,
+      "grad_norm": 0.24482061852995912,
+      "learning_rate": 3.318670576735093e-05,
+      "loss": 0.4647,
+      "step": 917
+    },
+    {
+      "epoch": 1.2096242584047463,
+      "grad_norm": 0.23647933523853154,
+      "learning_rate": 3.3162267839687195e-05,
+      "loss": 0.4601,
+      "step": 918
+    },
+    {
+      "epoch": 1.2109426499670402,
+      "grad_norm": 0.22916761127062124,
+      "learning_rate": 3.313782991202346e-05,
+      "loss": 0.4647,
+      "step": 919
+    },
+    {
+      "epoch": 1.2122610415293342,
+      "grad_norm": 0.24263841837024533,
+      "learning_rate": 3.3113391984359725e-05,
+      "loss": 0.4649,
+      "step": 920
+    },
+    {
+      "epoch": 1.2135794330916283,
+      "grad_norm": 0.2530235079916965,
+      "learning_rate": 3.308895405669599e-05,
+      "loss": 0.4796,
+      "step": 921
+    },
+    {
+      "epoch": 1.2148978246539222,
+      "grad_norm": 0.2171283222503539,
+      "learning_rate": 3.306451612903226e-05,
+      "loss": 0.474,
+      "step": 922
+    },
+    {
+      "epoch": 1.2162162162162162,
+      "grad_norm": 0.24469578103938108,
+      "learning_rate": 3.3040078201368526e-05,
+      "loss": 0.4892,
+      "step": 923
+    },
+    {
+      "epoch": 1.2175346077785103,
+      "grad_norm": 0.27623197253164145,
+      "learning_rate": 3.301564027370479e-05,
+      "loss": 0.4851,
+      "step": 924
+    },
+    {
+      "epoch": 1.2188529993408042,
+      "grad_norm": 0.22633308662784898,
+      "learning_rate": 3.2991202346041056e-05,
+      "loss": 0.5,
+      "step": 925
+    },
+    {
+      "epoch": 1.2201713909030982,
+      "grad_norm": 0.28411619408181704,
+      "learning_rate": 3.296676441837732e-05,
+      "loss": 0.4714,
+      "step": 926
+    },
+    {
+      "epoch": 1.2214897824653923,
+      "grad_norm": 0.253849391747479,
+      "learning_rate": 3.2942326490713585e-05,
+      "loss": 0.455,
+      "step": 927
+    },
+    {
+      "epoch": 1.2228081740276862,
+      "grad_norm": 0.2567023616117178,
+      "learning_rate": 3.291788856304985e-05,
+      "loss": 0.4724,
+      "step": 928
+    },
+    {
+      "epoch": 1.2241265655899802,
+      "grad_norm": 0.2390636162764636,
+      "learning_rate": 3.289345063538612e-05,
+      "loss": 0.44,
+      "step": 929
+    },
+    {
+      "epoch": 1.2254449571522743,
+      "grad_norm": 0.27329971520642554,
+      "learning_rate": 3.2869012707722386e-05,
+      "loss": 0.4893,
+      "step": 930
+    },
+    {
+      "epoch": 1.2267633487145682,
+      "grad_norm": 0.2404579946326346,
+      "learning_rate": 3.284457478005865e-05,
+      "loss": 0.4983,
+      "step": 931
+    },
+    {
+      "epoch": 1.2280817402768622,
+      "grad_norm": 0.4801158277514549,
+      "learning_rate": 3.2820136852394916e-05,
+      "loss": 0.4691,
+      "step": 932
+    },
+    {
+      "epoch": 1.2294001318391563,
+      "grad_norm": 0.2427313487185624,
+      "learning_rate": 3.279569892473118e-05,
+      "loss": 0.4635,
+      "step": 933
+    },
+    {
+      "epoch": 1.2307185234014502,
+      "grad_norm": 0.2754623695894105,
+      "learning_rate": 3.2771260997067446e-05,
+      "loss": 0.4615,
+      "step": 934
+    },
+    {
+      "epoch": 1.2320369149637442,
+      "grad_norm": 0.22313077938378398,
+      "learning_rate": 3.274682306940372e-05,
+      "loss": 0.4468,
+      "step": 935
+    },
+    {
+      "epoch": 1.2333553065260383,
+      "grad_norm": 0.270314243820246,
+      "learning_rate": 3.272238514173998e-05,
+      "loss": 0.4604,
+      "step": 936
+    },
+    {
+      "epoch": 1.2346736980883322,
+      "grad_norm": 0.2481220171956302,
+      "learning_rate": 3.269794721407625e-05,
+      "loss": 0.4499,
+      "step": 937
+    },
+    {
+      "epoch": 1.2359920896506262,
+      "grad_norm": 0.22573113776710904,
+      "learning_rate": 3.267350928641251e-05,
+      "loss": 0.4853,
+      "step": 938
+    },
+    {
+      "epoch": 1.2373104812129203,
+      "grad_norm": 0.2535195003369712,
+      "learning_rate": 3.2649071358748776e-05,
+      "loss": 0.4887,
+      "step": 939
+    },
+    {
+      "epoch": 1.2386288727752142,
+      "grad_norm": 0.24291151033848415,
+      "learning_rate": 3.262463343108504e-05,
+      "loss": 0.4582,
+      "step": 940
+    },
+    {
+      "epoch": 1.2399472643375082,
+      "grad_norm": 0.23530981888180083,
+      "learning_rate": 3.260019550342131e-05,
+      "loss": 0.4498,
+      "step": 941
+    },
+    {
+      "epoch": 1.2412656558998023,
+      "grad_norm": 0.2370216902585665,
+      "learning_rate": 3.257575757575758e-05,
+      "loss": 0.4482,
+      "step": 942
+    },
+    {
+      "epoch": 1.2425840474620962,
+      "grad_norm": 0.2559657946317417,
+      "learning_rate": 3.255131964809384e-05,
+      "loss": 0.4885,
+      "step": 943
+    },
+    {
+      "epoch": 1.2439024390243902,
+      "grad_norm": 0.2611093922310556,
+      "learning_rate": 3.252688172043011e-05,
+      "loss": 0.4557,
+      "step": 944
+    },
+    {
+      "epoch": 1.2452208305866843,
+      "grad_norm": 0.2616220168786616,
+      "learning_rate": 3.250244379276637e-05,
+      "loss": 0.4578,
+      "step": 945
+    },
+    {
+      "epoch": 1.2465392221489782,
+      "grad_norm": 0.2659677449379407,
+      "learning_rate": 3.247800586510264e-05,
+      "loss": 0.4606,
+      "step": 946
+    },
+    {
+      "epoch": 1.2478576137112722,
+      "grad_norm": 0.24950637964635644,
+      "learning_rate": 3.245356793743891e-05,
+      "loss": 0.4929,
+      "step": 947
+    },
+    {
+      "epoch": 1.2491760052735663,
+      "grad_norm": 0.2738213408133406,
+      "learning_rate": 3.242913000977517e-05,
+      "loss": 0.4571,
+      "step": 948
+    },
+    {
+      "epoch": 1.2504943968358604,
+      "grad_norm": 0.22073244772814132,
+      "learning_rate": 3.240469208211144e-05,
+      "loss": 0.4621,
+      "step": 949
+    },
+    {
+      "epoch": 1.2518127883981542,
+      "grad_norm": 0.23073884250777044,
+      "learning_rate": 3.23802541544477e-05,
+      "loss": 0.4649,
+      "step": 950
+    },
+    {
+      "epoch": 1.2531311799604483,
+      "grad_norm": 0.2558917822063307,
+      "learning_rate": 3.235581622678397e-05,
+      "loss": 0.4661,
+      "step": 951
+    },
+    {
+      "epoch": 1.2544495715227422,
+      "grad_norm": 0.258410394168111,
+      "learning_rate": 3.233137829912023e-05,
+      "loss": 0.4628,
+      "step": 952
+    },
+    {
+      "epoch": 1.2557679630850362,
+      "grad_norm": 0.2301689989142098,
+      "learning_rate": 3.2306940371456504e-05,
+      "loss": 0.4931,
+      "step": 953
+    },
+    {
+      "epoch": 1.2570863546473303,
+      "grad_norm": 0.25640297099084286,
+      "learning_rate": 3.228250244379277e-05,
+      "loss": 0.4735,
+      "step": 954
+    },
+    {
+      "epoch": 1.2584047462096244,
+      "grad_norm": 0.24466781035264648,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.4491,
+      "step": 955
+    },
+    {
+      "epoch": 1.2597231377719182,
+      "grad_norm": 0.21488373833660676,
+      "learning_rate": 3.22336265884653e-05,
+      "loss": 0.4761,
+      "step": 956
+    },
+    {
+      "epoch": 1.2610415293342123,
+      "grad_norm": 0.23753452386168794,
+      "learning_rate": 3.220918866080156e-05,
+      "loss": 0.5066,
+      "step": 957
+    },
+    {
+      "epoch": 1.2623599208965062,
+      "grad_norm": 0.24573345370474287,
+      "learning_rate": 3.218475073313783e-05,
+      "loss": 0.5041,
+      "step": 958
+    },
+    {
+      "epoch": 1.2636783124588002,
+      "grad_norm": 0.22823559630130183,
+      "learning_rate": 3.21603128054741e-05,
+      "loss": 0.497,
+      "step": 959
+    },
+    {
+      "epoch": 1.2649967040210943,
+      "grad_norm": 0.2433453405771188,
+      "learning_rate": 3.2135874877810364e-05,
+      "loss": 0.4626,
+      "step": 960
+    },
+    {
+      "epoch": 1.2663150955833884,
+      "grad_norm": 0.22096207557491,
+      "learning_rate": 3.211143695014663e-05,
+      "loss": 0.4594,
+      "step": 961
+    },
+    {
+      "epoch": 1.2676334871456822,
+      "grad_norm": 0.29745101391569545,
+      "learning_rate": 3.2086999022482894e-05,
+      "loss": 0.5017,
+      "step": 962
+    },
+    {
+      "epoch": 1.2689518787079763,
+      "grad_norm": 0.23858300900582768,
+      "learning_rate": 3.206256109481916e-05,
+      "loss": 0.5098,
+      "step": 963
+    },
+    {
+      "epoch": 1.2702702702702702,
+      "grad_norm": 0.27259340565508544,
+      "learning_rate": 3.2038123167155424e-05,
+      "loss": 0.5216,
+      "step": 964
+    },
+    {
+      "epoch": 1.2715886618325642,
+      "grad_norm": 0.2559155314795055,
+      "learning_rate": 3.2013685239491695e-05,
+      "loss": 0.4673,
+      "step": 965
+    },
+    {
+      "epoch": 1.2729070533948583,
+      "grad_norm": 0.23936290739010796,
+      "learning_rate": 3.198924731182796e-05,
+      "loss": 0.4803,
+      "step": 966
+    },
+    {
+      "epoch": 1.2742254449571524,
+      "grad_norm": 0.27390790836444107,
+      "learning_rate": 3.1964809384164225e-05,
+      "loss": 0.4845,
+      "step": 967
+    },
+    {
+      "epoch": 1.2755438365194463,
+      "grad_norm": 0.24265895828946646,
+      "learning_rate": 3.194037145650049e-05,
+      "loss": 0.4505,
+      "step": 968
+    },
+    {
+      "epoch": 1.2768622280817403,
+      "grad_norm": 0.2472810994791956,
+      "learning_rate": 3.1915933528836754e-05,
+      "loss": 0.4384,
+      "step": 969
+    },
+    {
+      "epoch": 1.2781806196440342,
+      "grad_norm": 0.2417362133420201,
+      "learning_rate": 3.189149560117302e-05,
+      "loss": 0.499,
+      "step": 970
+    },
+    {
+      "epoch": 1.2794990112063283,
+      "grad_norm": 0.23314289652670495,
+      "learning_rate": 3.186705767350929e-05,
+      "loss": 0.4791,
+      "step": 971
+    },
+    {
+      "epoch": 1.2808174027686223,
+      "grad_norm": 0.251388682776699,
+      "learning_rate": 3.1842619745845556e-05,
+      "loss": 0.4691,
+      "step": 972
+    },
+    {
+      "epoch": 1.2821357943309164,
+      "grad_norm": 0.2433330116125256,
+      "learning_rate": 3.181818181818182e-05,
+      "loss": 0.4854,
+      "step": 973
+    },
+    {
+      "epoch": 1.2834541858932103,
+      "grad_norm": 0.2193990719243341,
+      "learning_rate": 3.1793743890518085e-05,
+      "loss": 0.459,
+      "step": 974
+    },
+    {
+      "epoch": 1.2847725774555043,
+      "grad_norm": 0.24284358003363354,
+      "learning_rate": 3.176930596285435e-05,
+      "loss": 0.5079,
+      "step": 975
+    },
+    {
+      "epoch": 1.2860909690177982,
+      "grad_norm": 0.24238144963866398,
+      "learning_rate": 3.1744868035190615e-05,
+      "loss": 0.4823,
+      "step": 976
+    },
+    {
+      "epoch": 1.2874093605800923,
+      "grad_norm": 0.2480630017861259,
+      "learning_rate": 3.172043010752688e-05,
+      "loss": 0.4916,
+      "step": 977
+    },
+    {
+      "epoch": 1.2887277521423863,
+      "grad_norm": 0.24071645139526218,
+      "learning_rate": 3.169599217986315e-05,
+      "loss": 0.4613,
+      "step": 978
+    },
+    {
+      "epoch": 1.2900461437046804,
+      "grad_norm": 0.2388767150759139,
+      "learning_rate": 3.1671554252199416e-05,
+      "loss": 0.4701,
+      "step": 979
+    },
+    {
+      "epoch": 1.2913645352669743,
+      "grad_norm": 0.27241744939550955,
+      "learning_rate": 3.164711632453568e-05,
+      "loss": 0.493,
+      "step": 980
+    },
+    {
+      "epoch": 1.2926829268292683,
+      "grad_norm": 0.2411486003530689,
+      "learning_rate": 3.1622678396871946e-05,
+      "loss": 0.4652,
+      "step": 981
+    },
+    {
+      "epoch": 1.2940013183915622,
+      "grad_norm": 0.230230669962328,
+      "learning_rate": 3.159824046920821e-05,
+      "loss": 0.4507,
+      "step": 982
+    },
+    {
+      "epoch": 1.2953197099538563,
+      "grad_norm": 0.22360825287604283,
+      "learning_rate": 3.1573802541544475e-05,
+      "loss": 0.4715,
+      "step": 983
+    },
+    {
+      "epoch": 1.2966381015161503,
+      "grad_norm": 0.29809175282286804,
+      "learning_rate": 3.154936461388075e-05,
+      "loss": 0.4859,
+      "step": 984
+    },
+    {
+      "epoch": 1.2979564930784444,
+      "grad_norm": 0.23745202886697742,
+      "learning_rate": 3.152492668621701e-05,
+      "loss": 0.4504,
+      "step": 985
+    },
+    {
+      "epoch": 1.2992748846407383,
+      "grad_norm": 0.2800715640919627,
+      "learning_rate": 3.1500488758553276e-05,
+      "loss": 0.4628,
+      "step": 986
+    },
+    {
+      "epoch": 1.3005932762030323,
+      "grad_norm": 0.26111360338480233,
+      "learning_rate": 3.147605083088954e-05,
+      "loss": 0.4968,
+      "step": 987
+    },
+    {
+      "epoch": 1.3019116677653262,
+      "grad_norm": 0.257239561318979,
+      "learning_rate": 3.1451612903225806e-05,
+      "loss": 0.4817,
+      "step": 988
+    },
+    {
+      "epoch": 1.3032300593276203,
+      "grad_norm": 0.2756340691212747,
+      "learning_rate": 3.142717497556207e-05,
+      "loss": 0.4719,
+      "step": 989
+    },
+    {
+      "epoch": 1.3045484508899143,
+      "grad_norm": 0.24329721309738064,
+      "learning_rate": 3.140273704789834e-05,
+      "loss": 0.466,
+      "step": 990
+    },
+    {
+      "epoch": 1.3058668424522084,
+      "grad_norm": 0.21489512038025352,
+      "learning_rate": 3.137829912023461e-05,
+      "loss": 0.4501,
+      "step": 991
+    },
+    {
+      "epoch": 1.3071852340145023,
+      "grad_norm": 0.2546218052835097,
+      "learning_rate": 3.135386119257087e-05,
+      "loss": 0.4828,
+      "step": 992
+    },
+    {
+      "epoch": 1.3085036255767963,
+      "grad_norm": 0.2582492162943645,
+      "learning_rate": 3.132942326490714e-05,
+      "loss": 0.4631,
+      "step": 993
+    },
+    {
+      "epoch": 1.3098220171390902,
+      "grad_norm": 0.21681176091689894,
+      "learning_rate": 3.13049853372434e-05,
+      "loss": 0.4588,
+      "step": 994
+    },
+    {
+      "epoch": 1.3111404087013843,
+      "grad_norm": 0.3044401222047412,
+      "learning_rate": 3.1280547409579666e-05,
+      "loss": 0.4743,
+      "step": 995
+    },
+    {
+      "epoch": 1.3124588002636783,
+      "grad_norm": 0.21713452076421264,
+      "learning_rate": 3.125610948191594e-05,
+      "loss": 0.4638,
+      "step": 996
+    },
+    {
+      "epoch": 1.3137771918259724,
+      "grad_norm": 0.22646715187038555,
+      "learning_rate": 3.12316715542522e-05,
+      "loss": 0.4422,
+      "step": 997
+    },
+    {
+      "epoch": 1.3150955833882663,
+      "grad_norm": 0.23500386747807173,
+      "learning_rate": 3.120723362658847e-05,
+      "loss": 0.4674,
+      "step": 998
+    },
+    {
+      "epoch": 1.3164139749505603,
+      "grad_norm": 0.2234328639229824,
+      "learning_rate": 3.118279569892473e-05,
+      "loss": 0.4623,
+      "step": 999
+    },
+    {
+      "epoch": 1.3177323665128542,
+      "grad_norm": 0.23031029860642577,
+      "learning_rate": 3.1158357771261e-05,
+      "loss": 0.4556,
+      "step": 1000
+    },
+    {
+      "epoch": 1.3190507580751483,
+      "grad_norm": 0.2192143731437018,
+      "learning_rate": 3.113391984359726e-05,
+      "loss": 0.468,
+      "step": 1001
+    },
+    {
+      "epoch": 1.3203691496374423,
+      "grad_norm": 0.2321028596848366,
+      "learning_rate": 3.1109481915933533e-05,
+      "loss": 0.4609,
+      "step": 1002
+    },
+    {
+      "epoch": 1.3216875411997364,
+      "grad_norm": 0.23013535258273385,
+      "learning_rate": 3.10850439882698e-05,
+      "loss": 0.4634,
+      "step": 1003
+    },
+    {
+      "epoch": 1.3230059327620303,
+      "grad_norm": 0.21962172306510855,
+      "learning_rate": 3.106060606060606e-05,
+      "loss": 0.4566,
+      "step": 1004
+    },
+    {
+      "epoch": 1.3243243243243243,
+      "grad_norm": 0.21669012038593483,
+      "learning_rate": 3.103616813294233e-05,
+      "loss": 0.4553,
+      "step": 1005
+    },
+    {
+      "epoch": 1.3256427158866182,
+      "grad_norm": 0.24558968050521374,
+      "learning_rate": 3.101173020527859e-05,
+      "loss": 0.4909,
+      "step": 1006
+    },
+    {
+      "epoch": 1.3269611074489123,
+      "grad_norm": 0.2215683315961168,
+      "learning_rate": 3.098729227761486e-05,
+      "loss": 0.4598,
+      "step": 1007
+    },
+    {
+      "epoch": 1.3282794990112063,
+      "grad_norm": 0.22074336466323455,
+      "learning_rate": 3.096285434995113e-05,
+      "loss": 0.4777,
+      "step": 1008
+    },
+    {
+      "epoch": 1.3295978905735004,
+      "grad_norm": 0.22887169040407446,
+      "learning_rate": 3.0938416422287394e-05,
+      "loss": 0.4965,
+      "step": 1009
+    },
+    {
+      "epoch": 1.3309162821357943,
+      "grad_norm": 0.22057051930818672,
+      "learning_rate": 3.091397849462366e-05,
+      "loss": 0.4567,
+      "step": 1010
+    },
+    {
+      "epoch": 1.3322346736980883,
+      "grad_norm": 0.2308639894028987,
+      "learning_rate": 3.0889540566959923e-05,
+      "loss": 0.4779,
+      "step": 1011
+    },
+    {
+      "epoch": 1.3335530652603822,
+      "grad_norm": 0.2477342764670805,
+      "learning_rate": 3.086510263929619e-05,
+      "loss": 0.4835,
+      "step": 1012
+    },
+    {
+      "epoch": 1.3348714568226763,
+      "grad_norm": 0.2216820986580052,
+      "learning_rate": 3.084066471163245e-05,
+      "loss": 0.4488,
+      "step": 1013
+    },
+    {
+      "epoch": 1.3361898483849703,
+      "grad_norm": 0.2285081539838011,
+      "learning_rate": 3.0816226783968725e-05,
+      "loss": 0.4637,
+      "step": 1014
+    },
+    {
+      "epoch": 1.3375082399472644,
+      "grad_norm": 0.24429475146712654,
+      "learning_rate": 3.079178885630499e-05,
+      "loss": 0.4683,
+      "step": 1015
+    },
+    {
+      "epoch": 1.3388266315095583,
+      "grad_norm": 0.2536657686666498,
+      "learning_rate": 3.0767350928641254e-05,
+      "loss": 0.4507,
+      "step": 1016
+    },
+    {
+      "epoch": 1.3401450230718523,
+      "grad_norm": 0.2433392185295441,
+      "learning_rate": 3.074291300097752e-05,
+      "loss": 0.4575,
+      "step": 1017
+    },
+    {
+      "epoch": 1.3414634146341464,
+      "grad_norm": 0.2540213444673301,
+      "learning_rate": 3.0718475073313784e-05,
+      "loss": 0.4648,
+      "step": 1018
+    },
+    {
+      "epoch": 1.3427818061964403,
+      "grad_norm": 0.2427250377561981,
+      "learning_rate": 3.069403714565005e-05,
+      "loss": 0.4708,
+      "step": 1019
+    },
+    {
+      "epoch": 1.3441001977587343,
+      "grad_norm": 0.23710895274772445,
+      "learning_rate": 3.066959921798631e-05,
+      "loss": 0.4857,
+      "step": 1020
+    },
+    {
+      "epoch": 1.3454185893210284,
+      "grad_norm": 0.2359640980640193,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.4847,
+      "step": 1021
+    },
+    {
+      "epoch": 1.3467369808833223,
+      "grad_norm": 0.22047818694684976,
+      "learning_rate": 3.062072336265885e-05,
+      "loss": 0.4304,
+      "step": 1022
+    },
+    {
+      "epoch": 1.3480553724456164,
+      "grad_norm": 0.21281511694628147,
+      "learning_rate": 3.0596285434995115e-05,
+      "loss": 0.4628,
+      "step": 1023
+    },
+    {
+      "epoch": 1.3493737640079104,
+      "grad_norm": 0.25031346510190444,
+      "learning_rate": 3.057184750733138e-05,
+      "loss": 0.4989,
+      "step": 1024
+    },
+    {
+      "epoch": 1.3506921555702043,
+      "grad_norm": 0.23415535441046903,
+      "learning_rate": 3.0547409579667644e-05,
+      "loss": 0.5286,
+      "step": 1025
+    },
+    {
+      "epoch": 1.3520105471324984,
+      "grad_norm": 0.2310742694389797,
+      "learning_rate": 3.052297165200391e-05,
+      "loss": 0.4766,
+      "step": 1026
+    },
+    {
+      "epoch": 1.3533289386947924,
+      "grad_norm": 0.21789441895223274,
+      "learning_rate": 3.0498533724340177e-05,
+      "loss": 0.4766,
+      "step": 1027
+    },
+    {
+      "epoch": 1.3546473302570863,
+      "grad_norm": 0.22314214959044645,
+      "learning_rate": 3.0474095796676445e-05,
+      "loss": 0.4765,
+      "step": 1028
+    },
+    {
+      "epoch": 1.3559657218193804,
+      "grad_norm": 0.216777902683528,
+      "learning_rate": 3.044965786901271e-05,
+      "loss": 0.4592,
+      "step": 1029
+    },
+    {
+      "epoch": 1.3572841133816744,
+      "grad_norm": 0.22550628180394322,
+      "learning_rate": 3.0425219941348975e-05,
+      "loss": 0.4644,
+      "step": 1030
+    },
+    {
+      "epoch": 1.3586025049439683,
+      "grad_norm": 0.22159073633397622,
+      "learning_rate": 3.0400782013685243e-05,
+      "loss": 0.4777,
+      "step": 1031
+    },
+    {
+      "epoch": 1.3599208965062624,
+      "grad_norm": 0.23036911880283145,
+      "learning_rate": 3.0376344086021508e-05,
+      "loss": 0.4477,
+      "step": 1032
+    },
+    {
+      "epoch": 1.3612392880685564,
+      "grad_norm": 0.22639332347514984,
+      "learning_rate": 3.0351906158357773e-05,
+      "loss": 0.4553,
+      "step": 1033
+    },
+    {
+      "epoch": 1.3625576796308503,
+      "grad_norm": 0.2234678674368293,
+      "learning_rate": 3.032746823069404e-05,
+      "loss": 0.4586,
+      "step": 1034
+    },
+    {
+      "epoch": 1.3638760711931444,
+      "grad_norm": 0.2199776346927498,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 0.4571,
+      "step": 1035
+    },
+    {
+      "epoch": 1.3651944627554384,
+      "grad_norm": 0.2471432092682019,
+      "learning_rate": 3.027859237536657e-05,
+      "loss": 0.4484,
+      "step": 1036
+    },
+    {
+      "epoch": 1.3665128543177323,
+      "grad_norm": 0.23616969841282145,
+      "learning_rate": 3.025415444770284e-05,
+      "loss": 0.4703,
+      "step": 1037
+    },
+    {
+      "epoch": 1.3678312458800264,
+      "grad_norm": 0.25263087647582344,
+      "learning_rate": 3.0229716520039104e-05,
+      "loss": 0.4681,
+      "step": 1038
+    },
+    {
+      "epoch": 1.3691496374423204,
+      "grad_norm": 0.28181694411782354,
+      "learning_rate": 3.0205278592375368e-05,
+      "loss": 0.4771,
+      "step": 1039
+    },
+    {
+      "epoch": 1.3704680290046145,
+      "grad_norm": 0.25039789029537196,
+      "learning_rate": 3.0180840664711636e-05,
+      "loss": 0.4569,
+      "step": 1040
+    },
+    {
+      "epoch": 1.3717864205669084,
+      "grad_norm": 0.21755837566893146,
+      "learning_rate": 3.01564027370479e-05,
+      "loss": 0.4642,
+      "step": 1041
+    },
+    {
+      "epoch": 1.3731048121292024,
+      "grad_norm": 0.2524205207466487,
+      "learning_rate": 3.0131964809384166e-05,
+      "loss": 0.4821,
+      "step": 1042
+    },
+    {
+      "epoch": 1.3744232036914963,
+      "grad_norm": 0.2202691265848731,
+      "learning_rate": 3.010752688172043e-05,
+      "loss": 0.4699,
+      "step": 1043
+    },
+    {
+      "epoch": 1.3757415952537904,
+      "grad_norm": 0.2480643933473554,
+      "learning_rate": 3.00830889540567e-05,
+      "loss": 0.4691,
+      "step": 1044
+    },
+    {
+      "epoch": 1.3770599868160844,
+      "grad_norm": 0.22881318812372609,
+      "learning_rate": 3.0058651026392964e-05,
+      "loss": 0.462,
+      "step": 1045
+    },
+    {
+      "epoch": 1.3783783783783785,
+      "grad_norm": 0.23186536545526948,
+      "learning_rate": 3.003421309872923e-05,
+      "loss": 0.4322,
+      "step": 1046
+    },
+    {
+      "epoch": 1.3796967699406724,
+      "grad_norm": 0.24917382996636497,
+      "learning_rate": 3.0009775171065497e-05,
+      "loss": 0.4771,
+      "step": 1047
+    },
+    {
+      "epoch": 1.3810151615029664,
+      "grad_norm": 0.22858352584976105,
+      "learning_rate": 2.998533724340176e-05,
+      "loss": 0.4746,
+      "step": 1048
+    },
+    {
+      "epoch": 1.3823335530652603,
+      "grad_norm": 0.2492883579171899,
+      "learning_rate": 2.9960899315738026e-05,
+      "loss": 0.4443,
+      "step": 1049
+    },
+    {
+      "epoch": 1.3836519446275544,
+      "grad_norm": 0.23076552655675953,
+      "learning_rate": 2.9936461388074295e-05,
+      "loss": 0.4451,
+      "step": 1050
+    },
+    {
+      "epoch": 1.3849703361898484,
+      "grad_norm": 0.23444866093149344,
+      "learning_rate": 2.991202346041056e-05,
+      "loss": 0.4748,
+      "step": 1051
+    },
+    {
+      "epoch": 1.3862887277521425,
+      "grad_norm": 0.24408209569369557,
+      "learning_rate": 2.9887585532746824e-05,
+      "loss": 0.4836,
+      "step": 1052
+    },
+    {
+      "epoch": 1.3876071193144364,
+      "grad_norm": 0.24027858260294485,
+      "learning_rate": 2.9863147605083092e-05,
+      "loss": 0.4604,
+      "step": 1053
+    },
+    {
+      "epoch": 1.3889255108767304,
+      "grad_norm": 0.2445621976312017,
+      "learning_rate": 2.9838709677419357e-05,
+      "loss": 0.4795,
+      "step": 1054
+    },
+    {
+      "epoch": 1.3902439024390243,
+      "grad_norm": 0.222618123796317,
+      "learning_rate": 2.9814271749755622e-05,
+      "loss": 0.4824,
+      "step": 1055
+    },
+    {
+      "epoch": 1.3915622940013184,
+      "grad_norm": 0.2349124594795995,
+      "learning_rate": 2.978983382209189e-05,
+      "loss": 0.4476,
+      "step": 1056
+    },
+    {
+      "epoch": 1.3928806855636124,
+      "grad_norm": 0.2278729896450846,
+      "learning_rate": 2.9765395894428155e-05,
+      "loss": 0.4463,
+      "step": 1057
+    },
+    {
+      "epoch": 1.3941990771259065,
+      "grad_norm": 0.20347663019822745,
+      "learning_rate": 2.974095796676442e-05,
+      "loss": 0.4819,
+      "step": 1058
+    },
+    {
+      "epoch": 1.3955174686882004,
+      "grad_norm": 0.24356500499282363,
+      "learning_rate": 2.9716520039100688e-05,
+      "loss": 0.4741,
+      "step": 1059
+    },
+    {
+      "epoch": 1.3968358602504944,
+      "grad_norm": 0.24837406507739276,
+      "learning_rate": 2.9692082111436953e-05,
+      "loss": 0.4662,
+      "step": 1060
+    },
+    {
+      "epoch": 1.3981542518127883,
+      "grad_norm": 0.21923354257935287,
+      "learning_rate": 2.9667644183773218e-05,
+      "loss": 0.4423,
+      "step": 1061
+    },
+    {
+      "epoch": 1.3994726433750824,
+      "grad_norm": 0.22371253057085944,
+      "learning_rate": 2.9643206256109486e-05,
+      "loss": 0.4389,
+      "step": 1062
+    },
+    {
+      "epoch": 1.4007910349373764,
+      "grad_norm": 0.21493620961645665,
+      "learning_rate": 2.961876832844575e-05,
+      "loss": 0.4532,
+      "step": 1063
+    },
+    {
+      "epoch": 1.4021094264996705,
+      "grad_norm": 0.2348119454358976,
+      "learning_rate": 2.9594330400782015e-05,
+      "loss": 0.4592,
+      "step": 1064
+    },
+    {
+      "epoch": 1.4034278180619644,
+      "grad_norm": 0.22304887645465973,
+      "learning_rate": 2.9569892473118284e-05,
+      "loss": 0.4626,
+      "step": 1065
+    },
+    {
+      "epoch": 1.4047462096242584,
+      "grad_norm": 0.22696051345300566,
+      "learning_rate": 2.954545454545455e-05,
+      "loss": 0.426,
+      "step": 1066
+    },
+    {
+      "epoch": 1.4060646011865523,
+      "grad_norm": 0.2149662719648378,
+      "learning_rate": 2.9521016617790813e-05,
+      "loss": 0.5063,
+      "step": 1067
+    },
+    {
+      "epoch": 1.4073829927488464,
+      "grad_norm": 0.31565704953149476,
+      "learning_rate": 2.949657869012708e-05,
+      "loss": 0.4839,
+      "step": 1068
+    },
+    {
+      "epoch": 1.4087013843111404,
+      "grad_norm": 0.23079873225031464,
+      "learning_rate": 2.9472140762463346e-05,
+      "loss": 0.4655,
+      "step": 1069
+    },
+    {
+      "epoch": 1.4100197758734345,
+      "grad_norm": 0.23789819505067117,
+      "learning_rate": 2.944770283479961e-05,
+      "loss": 0.4909,
+      "step": 1070
+    },
+    {
+      "epoch": 1.4113381674357284,
+      "grad_norm": 0.24054671619973703,
+      "learning_rate": 2.942326490713588e-05,
+      "loss": 0.4535,
+      "step": 1071
+    },
+    {
+      "epoch": 1.4126565589980224,
+      "grad_norm": 0.21808712828434718,
+      "learning_rate": 2.9398826979472144e-05,
+      "loss": 0.4574,
+      "step": 1072
+    },
+    {
+      "epoch": 1.4139749505603163,
+      "grad_norm": 0.23909109004158255,
+      "learning_rate": 2.937438905180841e-05,
+      "loss": 0.4608,
+      "step": 1073
+    },
+    {
+      "epoch": 1.4152933421226104,
+      "grad_norm": 0.2605716333306736,
+      "learning_rate": 2.9349951124144677e-05,
+      "loss": 0.4765,
+      "step": 1074
+    },
+    {
+      "epoch": 1.4166117336849045,
+      "grad_norm": 0.23159981028163287,
+      "learning_rate": 2.9325513196480942e-05,
+      "loss": 0.4563,
+      "step": 1075
+    },
+    {
+      "epoch": 1.4179301252471985,
+      "grad_norm": 0.2375717496350306,
+      "learning_rate": 2.9301075268817207e-05,
+      "loss": 0.4599,
+      "step": 1076
+    },
+    {
+      "epoch": 1.4192485168094924,
+      "grad_norm": 0.24547174091756602,
+      "learning_rate": 2.9276637341153475e-05,
+      "loss": 0.4583,
+      "step": 1077
+    },
+    {
+      "epoch": 1.4205669083717865,
+      "grad_norm": 0.21395600723826017,
+      "learning_rate": 2.925219941348974e-05,
+      "loss": 0.4574,
+      "step": 1078
+    },
+    {
+      "epoch": 1.4218852999340803,
+      "grad_norm": 0.24566872943520876,
+      "learning_rate": 2.9227761485826004e-05,
+      "loss": 0.444,
+      "step": 1079
+    },
+    {
+      "epoch": 1.4232036914963744,
+      "grad_norm": 0.21439267555610966,
+      "learning_rate": 2.9203323558162273e-05,
+      "loss": 0.4458,
+      "step": 1080
+    },
+    {
+      "epoch": 1.4245220830586685,
+      "grad_norm": 0.22012724731311947,
+      "learning_rate": 2.9178885630498537e-05,
+      "loss": 0.4739,
+      "step": 1081
+    },
+    {
+      "epoch": 1.4258404746209625,
+      "grad_norm": 0.24971446034433706,
+      "learning_rate": 2.9154447702834802e-05,
+      "loss": 0.4716,
+      "step": 1082
+    },
+    {
+      "epoch": 1.4271588661832564,
+      "grad_norm": 0.2137789100158401,
+      "learning_rate": 2.913000977517107e-05,
+      "loss": 0.4391,
+      "step": 1083
+    },
+    {
+      "epoch": 1.4284772577455505,
+      "grad_norm": 0.24185329907937356,
+      "learning_rate": 2.9105571847507335e-05,
+      "loss": 0.4909,
+      "step": 1084
+    },
+    {
+      "epoch": 1.4297956493078443,
+      "grad_norm": 0.22025418989382062,
+      "learning_rate": 2.90811339198436e-05,
+      "loss": 0.4808,
+      "step": 1085
+    },
+    {
+      "epoch": 1.4311140408701384,
+      "grad_norm": 0.22516180026772328,
+      "learning_rate": 2.9056695992179868e-05,
+      "loss": 0.4359,
+      "step": 1086
+    },
+    {
+      "epoch": 1.4324324324324325,
+      "grad_norm": 0.23601577117218894,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.4798,
+      "step": 1087
+    },
+    {
+      "epoch": 1.4337508239947265,
+      "grad_norm": 0.22993422596109767,
+      "learning_rate": 2.9007820136852398e-05,
+      "loss": 0.4694,
+      "step": 1088
+    },
+    {
+      "epoch": 1.4350692155570204,
+      "grad_norm": 0.22362975229307816,
+      "learning_rate": 2.8983382209188663e-05,
+      "loss": 0.474,
+      "step": 1089
+    },
+    {
+      "epoch": 1.4363876071193145,
+      "grad_norm": 0.21358081407368504,
+      "learning_rate": 2.895894428152493e-05,
+      "loss": 0.4567,
+      "step": 1090
+    },
+    {
+      "epoch": 1.4377059986816083,
+      "grad_norm": 0.24587814791221144,
+      "learning_rate": 2.8934506353861196e-05,
+      "loss": 0.4635,
+      "step": 1091
+    },
+    {
+      "epoch": 1.4390243902439024,
+      "grad_norm": 0.21832617723073494,
+      "learning_rate": 2.891006842619746e-05,
+      "loss": 0.4761,
+      "step": 1092
+    },
+    {
+      "epoch": 1.4403427818061965,
+      "grad_norm": 0.22570057071672217,
+      "learning_rate": 2.888563049853373e-05,
+      "loss": 0.4763,
+      "step": 1093
+    },
+    {
+      "epoch": 1.4416611733684905,
+      "grad_norm": 0.24732392430910616,
+      "learning_rate": 2.8861192570869993e-05,
+      "loss": 0.4721,
+      "step": 1094
+    },
+    {
+      "epoch": 1.4429795649307844,
+      "grad_norm": 0.2439929492144735,
+      "learning_rate": 2.8836754643206258e-05,
+      "loss": 0.4844,
+      "step": 1095
+    },
+    {
+      "epoch": 1.4442979564930785,
+      "grad_norm": 0.2253671245961365,
+      "learning_rate": 2.8812316715542526e-05,
+      "loss": 0.4576,
+      "step": 1096
+    },
+    {
+      "epoch": 1.4456163480553723,
+      "grad_norm": 0.28246458558074544,
+      "learning_rate": 2.878787878787879e-05,
+      "loss": 0.4799,
+      "step": 1097
+    },
+    {
+      "epoch": 1.4469347396176664,
+      "grad_norm": 0.21127688156249302,
+      "learning_rate": 2.8763440860215056e-05,
+      "loss": 0.4516,
+      "step": 1098
+    },
+    {
+      "epoch": 1.4482531311799605,
+      "grad_norm": 0.5780320185291061,
+      "learning_rate": 2.8739002932551324e-05,
+      "loss": 0.4938,
+      "step": 1099
+    },
+    {
+      "epoch": 1.4495715227422545,
+      "grad_norm": 0.23880876388572025,
+      "learning_rate": 2.871456500488759e-05,
+      "loss": 0.4849,
+      "step": 1100
+    },
+    {
+      "epoch": 1.4508899143045484,
+      "grad_norm": 0.2363360773165683,
+      "learning_rate": 2.8690127077223854e-05,
+      "loss": 0.4564,
+      "step": 1101
+    },
+    {
+      "epoch": 1.4522083058668425,
+      "grad_norm": 0.24773342075344842,
+      "learning_rate": 2.8665689149560122e-05,
+      "loss": 0.4788,
+      "step": 1102
+    },
+    {
+      "epoch": 1.4535266974291363,
+      "grad_norm": 0.24431503764857307,
+      "learning_rate": 2.8641251221896387e-05,
+      "loss": 0.4424,
+      "step": 1103
+    },
+    {
+      "epoch": 1.4548450889914304,
+      "grad_norm": 0.24006718785726563,
+      "learning_rate": 2.861681329423265e-05,
+      "loss": 0.4858,
+      "step": 1104
+    },
+    {
+      "epoch": 1.4561634805537245,
+      "grad_norm": 0.2740532633925799,
+      "learning_rate": 2.859237536656892e-05,
+      "loss": 0.456,
+      "step": 1105
+    },
+    {
+      "epoch": 1.4574818721160185,
+      "grad_norm": 0.23381135907597766,
+      "learning_rate": 2.8567937438905184e-05,
+      "loss": 0.4379,
+      "step": 1106
+    },
+    {
+      "epoch": 1.4588002636783124,
+      "grad_norm": 0.24120265757834322,
+      "learning_rate": 2.854349951124145e-05,
+      "loss": 0.4234,
+      "step": 1107
+    },
+    {
+      "epoch": 1.4601186552406065,
+      "grad_norm": 0.2461700761255084,
+      "learning_rate": 2.8519061583577717e-05,
+      "loss": 0.4422,
+      "step": 1108
+    },
+    {
+      "epoch": 1.4614370468029005,
+      "grad_norm": 0.24386888245025193,
+      "learning_rate": 2.8494623655913982e-05,
+      "loss": 0.4573,
+      "step": 1109
+    },
+    {
+      "epoch": 1.4627554383651944,
+      "grad_norm": 0.22143117981811625,
+      "learning_rate": 2.8470185728250247e-05,
+      "loss": 0.4623,
+      "step": 1110
+    },
+    {
+      "epoch": 1.4640738299274885,
+      "grad_norm": 0.24562074922482396,
+      "learning_rate": 2.8445747800586515e-05,
+      "loss": 0.4552,
+      "step": 1111
+    },
+    {
+      "epoch": 1.4653922214897825,
+      "grad_norm": 0.24390862648839062,
+      "learning_rate": 2.842130987292278e-05,
+      "loss": 0.4897,
+      "step": 1112
+    },
+    {
+      "epoch": 1.4667106130520764,
+      "grad_norm": 0.25942356347348167,
+      "learning_rate": 2.8396871945259045e-05,
+      "loss": 0.486,
+      "step": 1113
+    },
+    {
+      "epoch": 1.4680290046143705,
+      "grad_norm": 0.24799941151894503,
+      "learning_rate": 2.8372434017595313e-05,
+      "loss": 0.4562,
+      "step": 1114
+    },
+    {
+      "epoch": 1.4693473961766645,
+      "grad_norm": 0.24253985586191293,
+      "learning_rate": 2.8347996089931578e-05,
+      "loss": 0.4541,
+      "step": 1115
+    },
+    {
+      "epoch": 1.4706657877389584,
+      "grad_norm": 0.24040927413057658,
+      "learning_rate": 2.8323558162267843e-05,
+      "loss": 0.4897,
+      "step": 1116
+    },
+    {
+      "epoch": 1.4719841793012525,
+      "grad_norm": 0.22797024944944477,
+      "learning_rate": 2.829912023460411e-05,
+      "loss": 0.4724,
+      "step": 1117
+    },
+    {
+      "epoch": 1.4733025708635465,
+      "grad_norm": 0.22420455307567194,
+      "learning_rate": 2.8274682306940376e-05,
+      "loss": 0.5096,
+      "step": 1118
+    },
+    {
+      "epoch": 1.4746209624258404,
+      "grad_norm": 0.22682649592286236,
+      "learning_rate": 2.825024437927664e-05,
+      "loss": 0.4663,
+      "step": 1119
+    },
+    {
+      "epoch": 1.4759393539881345,
+      "grad_norm": 0.24143440156099719,
+      "learning_rate": 2.822580645161291e-05,
+      "loss": 0.464,
+      "step": 1120
+    },
+    {
+      "epoch": 1.4772577455504285,
+      "grad_norm": 0.24229264032099856,
+      "learning_rate": 2.8201368523949173e-05,
+      "loss": 0.4736,
+      "step": 1121
+    },
+    {
+      "epoch": 1.4785761371127224,
+      "grad_norm": 0.2444280302965182,
+      "learning_rate": 2.8176930596285438e-05,
+      "loss": 0.4417,
+      "step": 1122
+    },
+    {
+      "epoch": 1.4798945286750165,
+      "grad_norm": 0.2410914362671272,
+      "learning_rate": 2.8152492668621706e-05,
+      "loss": 0.4418,
+      "step": 1123
+    },
+    {
+      "epoch": 1.4812129202373105,
+      "grad_norm": 0.2406142967016478,
+      "learning_rate": 2.812805474095797e-05,
+      "loss": 0.4802,
+      "step": 1124
+    },
+    {
+      "epoch": 1.4825313117996044,
+      "grad_norm": 0.2452794352009678,
+      "learning_rate": 2.8103616813294233e-05,
+      "loss": 0.4644,
+      "step": 1125
+    },
+    {
+      "epoch": 1.4838497033618985,
+      "grad_norm": 0.252531967736398,
+      "learning_rate": 2.8079178885630497e-05,
+      "loss": 0.4549,
+      "step": 1126
+    },
+    {
+      "epoch": 1.4851680949241926,
+      "grad_norm": 0.22890528440087674,
+      "learning_rate": 2.8054740957966762e-05,
+      "loss": 0.4818,
+      "step": 1127
+    },
+    {
+      "epoch": 1.4864864864864864,
+      "grad_norm": 0.22737261790798738,
+      "learning_rate": 2.803030303030303e-05,
+      "loss": 0.4507,
+      "step": 1128
+    },
+    {
+      "epoch": 1.4878048780487805,
+      "grad_norm": 0.22146836036277223,
+      "learning_rate": 2.8005865102639295e-05,
+      "loss": 0.4455,
+      "step": 1129
+    },
+    {
+      "epoch": 1.4891232696110746,
+      "grad_norm": 0.23315792952800324,
+      "learning_rate": 2.798142717497556e-05,
+      "loss": 0.4073,
+      "step": 1130
+    },
+    {
+      "epoch": 1.4904416611733686,
+      "grad_norm": 0.21424175956864866,
+      "learning_rate": 2.7956989247311828e-05,
+      "loss": 0.4557,
+      "step": 1131
+    },
+    {
+      "epoch": 1.4917600527356625,
+      "grad_norm": 0.2327031251450455,
+      "learning_rate": 2.7932551319648093e-05,
+      "loss": 0.4576,
+      "step": 1132
+    },
+    {
+      "epoch": 1.4930784442979566,
+      "grad_norm": 0.2297204614410037,
+      "learning_rate": 2.7908113391984358e-05,
+      "loss": 0.4428,
+      "step": 1133
+    },
+    {
+      "epoch": 1.4943968358602504,
+      "grad_norm": 0.21873971757048355,
+      "learning_rate": 2.7883675464320626e-05,
+      "loss": 0.4796,
+      "step": 1134
+    },
+    {
+      "epoch": 1.4957152274225445,
+      "grad_norm": 0.23652012665394856,
+      "learning_rate": 2.785923753665689e-05,
+      "loss": 0.4513,
+      "step": 1135
+    },
+    {
+      "epoch": 1.4970336189848386,
+      "grad_norm": 0.28634539924486596,
+      "learning_rate": 2.7834799608993156e-05,
+      "loss": 0.5047,
+      "step": 1136
+    },
+    {
+      "epoch": 1.4983520105471326,
+      "grad_norm": 0.2372561077832039,
+      "learning_rate": 2.7810361681329424e-05,
+      "loss": 0.4323,
+      "step": 1137
+    },
+    {
+      "epoch": 1.4996704021094265,
+      "grad_norm": 0.2601398290925287,
+      "learning_rate": 2.778592375366569e-05,
+      "loss": 0.4522,
+      "step": 1138
+    },
+    {
+      "epoch": 1.5009887936717206,
+      "grad_norm": 0.24425352457068444,
+      "learning_rate": 2.7761485826001953e-05,
+      "loss": 0.4598,
+      "step": 1139
+    },
+    {
+      "epoch": 1.5023071852340144,
+      "grad_norm": 0.2351892816869354,
+      "learning_rate": 2.7737047898338218e-05,
+      "loss": 0.467,
+      "step": 1140
+    },
+    {
+      "epoch": 1.5036255767963085,
+      "grad_norm": 0.2610310309891974,
+      "learning_rate": 2.7712609970674486e-05,
+      "loss": 0.4576,
+      "step": 1141
+    },
+    {
+      "epoch": 1.5049439683586026,
+      "grad_norm": 0.24224728227788167,
+      "learning_rate": 2.768817204301075e-05,
+      "loss": 0.465,
+      "step": 1142
+    },
+    {
+      "epoch": 1.5062623599208966,
+      "grad_norm": 0.2441934109025499,
+      "learning_rate": 2.7663734115347016e-05,
+      "loss": 0.44,
+      "step": 1143
+    },
+    {
+      "epoch": 1.5075807514831905,
+      "grad_norm": 0.2394759207418652,
+      "learning_rate": 2.7639296187683284e-05,
+      "loss": 0.4652,
+      "step": 1144
+    },
+    {
+      "epoch": 1.5088991430454846,
+      "grad_norm": 0.23547804358132082,
+      "learning_rate": 2.761485826001955e-05,
+      "loss": 0.4594,
+      "step": 1145
+    },
+    {
+      "epoch": 1.5102175346077784,
+      "grad_norm": 0.230948382942716,
+      "learning_rate": 2.7590420332355814e-05,
+      "loss": 0.4822,
+      "step": 1146
+    },
+    {
+      "epoch": 1.5115359261700725,
+      "grad_norm": 0.271285859306579,
+      "learning_rate": 2.7565982404692082e-05,
+      "loss": 0.4936,
+      "step": 1147
+    },
+    {
+      "epoch": 1.5128543177323666,
+      "grad_norm": 0.2533061892620308,
+      "learning_rate": 2.7541544477028347e-05,
+      "loss": 0.4369,
+      "step": 1148
+    },
+    {
+      "epoch": 1.5141727092946606,
+      "grad_norm": 0.2381870416058091,
+      "learning_rate": 2.751710654936461e-05,
+      "loss": 0.4459,
+      "step": 1149
+    },
+    {
+      "epoch": 1.5154911008569545,
+      "grad_norm": 0.22503570884778645,
+      "learning_rate": 2.749266862170088e-05,
+      "loss": 0.4433,
+      "step": 1150
+    },
+    {
+      "epoch": 1.5168094924192486,
+      "grad_norm": 0.21491527520767673,
+      "learning_rate": 2.7468230694037145e-05,
+      "loss": 0.4623,
+      "step": 1151
+    },
+    {
+      "epoch": 1.5181278839815424,
+      "grad_norm": 0.24129485471702142,
+      "learning_rate": 2.744379276637341e-05,
+      "loss": 0.4809,
+      "step": 1152
+    },
+    {
+      "epoch": 1.5194462755438365,
+      "grad_norm": 0.23031552758735344,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.4371,
+      "step": 1153
+    },
+    {
+      "epoch": 1.5207646671061306,
+      "grad_norm": 0.22771123707879085,
+      "learning_rate": 2.7394916911045942e-05,
+      "loss": 0.4551,
+      "step": 1154
+    },
+    {
+      "epoch": 1.5220830586684246,
+      "grad_norm": 0.24985607931456297,
+      "learning_rate": 2.7370478983382207e-05,
+      "loss": 0.4999,
+      "step": 1155
+    },
+    {
+      "epoch": 1.5234014502307185,
+      "grad_norm": 0.2352358272903499,
+      "learning_rate": 2.7346041055718475e-05,
+      "loss": 0.4372,
+      "step": 1156
+    },
+    {
+      "epoch": 1.5247198417930126,
+      "grad_norm": 0.23883049192235517,
+      "learning_rate": 2.732160312805474e-05,
+      "loss": 0.4793,
+      "step": 1157
+    },
+    {
+      "epoch": 1.5260382333553064,
+      "grad_norm": 0.23244884048534203,
+      "learning_rate": 2.7297165200391005e-05,
+      "loss": 0.458,
+      "step": 1158
+    },
+    {
+      "epoch": 1.5273566249176005,
+      "grad_norm": 0.22929555529817197,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.4377,
+      "step": 1159
+    },
+    {
+      "epoch": 1.5286750164798946,
+      "grad_norm": 0.24765094994046205,
+      "learning_rate": 2.7248289345063538e-05,
+      "loss": 0.4863,
+      "step": 1160
+    },
+    {
+      "epoch": 1.5299934080421886,
+      "grad_norm": 0.25123918827727293,
+      "learning_rate": 2.7223851417399803e-05,
+      "loss": 0.4668,
+      "step": 1161
+    },
+    {
+      "epoch": 1.5313117996044825,
+      "grad_norm": 0.24028129109897486,
+      "learning_rate": 2.719941348973607e-05,
+      "loss": 0.4528,
+      "step": 1162
+    },
+    {
+      "epoch": 1.5326301911667766,
+      "grad_norm": 0.2412021157817711,
+      "learning_rate": 2.7174975562072336e-05,
+      "loss": 0.4667,
+      "step": 1163
+    },
+    {
+      "epoch": 1.5339485827290704,
+      "grad_norm": 0.23073397111213045,
+      "learning_rate": 2.71505376344086e-05,
+      "loss": 0.4565,
+      "step": 1164
+    },
+    {
+      "epoch": 1.5352669742913645,
+      "grad_norm": 0.23269210309828756,
+      "learning_rate": 2.712609970674487e-05,
+      "loss": 0.4552,
+      "step": 1165
+    },
+    {
+      "epoch": 1.5365853658536586,
+      "grad_norm": 0.24978794579214766,
+      "learning_rate": 2.7101661779081133e-05,
+      "loss": 0.4491,
+      "step": 1166
+    },
+    {
+      "epoch": 1.5379037574159526,
+      "grad_norm": 0.2770836783502218,
+      "learning_rate": 2.7077223851417398e-05,
+      "loss": 0.4592,
+      "step": 1167
+    },
+    {
+      "epoch": 1.5392221489782465,
+      "grad_norm": 0.23426069970710242,
+      "learning_rate": 2.7052785923753666e-05,
+      "loss": 0.4427,
+      "step": 1168
+    },
+    {
+      "epoch": 1.5405405405405406,
+      "grad_norm": 0.27261339123123574,
+      "learning_rate": 2.702834799608993e-05,
+      "loss": 0.4698,
+      "step": 1169
+    },
+    {
+      "epoch": 1.5418589321028344,
+      "grad_norm": 0.2600079452834312,
+      "learning_rate": 2.7003910068426196e-05,
+      "loss": 0.4824,
+      "step": 1170
+    },
+    {
+      "epoch": 1.5431773236651285,
+      "grad_norm": 0.25232105708221836,
+      "learning_rate": 2.6979472140762464e-05,
+      "loss": 0.4651,
+      "step": 1171
+    },
+    {
+      "epoch": 1.5444957152274226,
+      "grad_norm": 0.24988665227560306,
+      "learning_rate": 2.695503421309873e-05,
+      "loss": 0.4726,
+      "step": 1172
+    },
+    {
+      "epoch": 1.5458141067897166,
+      "grad_norm": 0.24355747709898115,
+      "learning_rate": 2.6930596285434994e-05,
+      "loss": 0.4554,
+      "step": 1173
+    },
+    {
+      "epoch": 1.5471324983520105,
+      "grad_norm": 0.25939290127295483,
+      "learning_rate": 2.6906158357771262e-05,
+      "loss": 0.4739,
+      "step": 1174
+    },
+    {
+      "epoch": 1.5484508899143046,
+      "grad_norm": 0.25764108587444484,
+      "learning_rate": 2.6881720430107527e-05,
+      "loss": 0.4476,
+      "step": 1175
+    },
+    {
+      "epoch": 1.5497692814765984,
+      "grad_norm": 0.2284419672917885,
+      "learning_rate": 2.685728250244379e-05,
+      "loss": 0.4471,
+      "step": 1176
+    },
+    {
+      "epoch": 1.5510876730388925,
+      "grad_norm": 0.23929546231664098,
+      "learning_rate": 2.683284457478006e-05,
+      "loss": 0.4875,
+      "step": 1177
+    },
+    {
+      "epoch": 1.5524060646011866,
+      "grad_norm": 0.219347615998279,
+      "learning_rate": 2.6808406647116325e-05,
+      "loss": 0.4551,
+      "step": 1178
+    },
+    {
+      "epoch": 1.5537244561634806,
+      "grad_norm": 0.24954947670868521,
+      "learning_rate": 2.678396871945259e-05,
+      "loss": 0.455,
+      "step": 1179
+    },
+    {
+      "epoch": 1.5550428477257745,
+      "grad_norm": 0.2145470095323426,
+      "learning_rate": 2.6759530791788858e-05,
+      "loss": 0.438,
+      "step": 1180
+    },
+    {
+      "epoch": 1.5563612392880686,
+      "grad_norm": 0.23638319734804977,
+      "learning_rate": 2.6735092864125122e-05,
+      "loss": 0.4766,
+      "step": 1181
+    },
+    {
+      "epoch": 1.5576796308503624,
+      "grad_norm": 0.22081846489106097,
+      "learning_rate": 2.6710654936461387e-05,
+      "loss": 0.4503,
+      "step": 1182
+    },
+    {
+      "epoch": 1.5589980224126565,
+      "grad_norm": 0.25904707052437215,
+      "learning_rate": 2.6686217008797655e-05,
+      "loss": 0.4425,
+      "step": 1183
+    },
+    {
+      "epoch": 1.5603164139749506,
+      "grad_norm": 0.22015155781992762,
+      "learning_rate": 2.666177908113392e-05,
+      "loss": 0.4426,
+      "step": 1184
+    },
+    {
+      "epoch": 1.5616348055372447,
+      "grad_norm": 0.2177652470886103,
+      "learning_rate": 2.6637341153470185e-05,
+      "loss": 0.4598,
+      "step": 1185
+    },
+    {
+      "epoch": 1.5629531970995387,
+      "grad_norm": 0.2662187293713749,
+      "learning_rate": 2.661290322580645e-05,
+      "loss": 0.459,
+      "step": 1186
+    },
+    {
+      "epoch": 1.5642715886618326,
+      "grad_norm": 0.22435156598439465,
+      "learning_rate": 2.6588465298142718e-05,
+      "loss": 0.4317,
+      "step": 1187
+    },
+    {
+      "epoch": 1.5655899802241264,
+      "grad_norm": 0.2219660718544592,
+      "learning_rate": 2.6564027370478983e-05,
+      "loss": 0.4757,
+      "step": 1188
+    },
+    {
+      "epoch": 1.5669083717864205,
+      "grad_norm": 0.26594525708561306,
+      "learning_rate": 2.6539589442815248e-05,
+      "loss": 0.4591,
+      "step": 1189
+    },
+    {
+      "epoch": 1.5682267633487146,
+      "grad_norm": 0.22303761040369458,
+      "learning_rate": 2.6515151515151516e-05,
+      "loss": 0.4304,
+      "step": 1190
+    },
+    {
+      "epoch": 1.5695451549110087,
+      "grad_norm": 0.22435722280164005,
+      "learning_rate": 2.649071358748778e-05,
+      "loss": 0.4265,
+      "step": 1191
+    },
+    {
+      "epoch": 1.5708635464733027,
+      "grad_norm": 0.2223353902284494,
+      "learning_rate": 2.6466275659824045e-05,
+      "loss": 0.4495,
+      "step": 1192
+    },
+    {
+      "epoch": 1.5721819380355966,
+      "grad_norm": 0.24749315212471315,
+      "learning_rate": 2.6441837732160314e-05,
+      "loss": 0.4848,
+      "step": 1193
+    },
+    {
+      "epoch": 1.5735003295978904,
+      "grad_norm": 0.21837795927323062,
+      "learning_rate": 2.641739980449658e-05,
+      "loss": 0.4554,
+      "step": 1194
+    },
+    {
+      "epoch": 1.5748187211601845,
+      "grad_norm": 0.22727972330413612,
+      "learning_rate": 2.6392961876832843e-05,
+      "loss": 0.4558,
+      "step": 1195
+    },
+    {
+      "epoch": 1.5761371127224786,
+      "grad_norm": 0.25729905347443716,
+      "learning_rate": 2.636852394916911e-05,
+      "loss": 0.4437,
+      "step": 1196
+    },
+    {
+      "epoch": 1.5774555042847727,
+      "grad_norm": 0.22201823807479265,
+      "learning_rate": 2.6344086021505376e-05,
+      "loss": 0.4983,
+      "step": 1197
+    },
+    {
+      "epoch": 1.5787738958470667,
+      "grad_norm": 0.2175516903336271,
+      "learning_rate": 2.631964809384164e-05,
+      "loss": 0.423,
+      "step": 1198
+    },
+    {
+      "epoch": 1.5800922874093606,
+      "grad_norm": 0.22138407133156857,
+      "learning_rate": 2.629521016617791e-05,
+      "loss": 0.4429,
+      "step": 1199
+    },
+    {
+      "epoch": 1.5814106789716544,
+      "grad_norm": 0.24294787306946383,
+      "learning_rate": 2.6270772238514174e-05,
+      "loss": 0.4464,
+      "step": 1200
+    },
+    {
+      "epoch": 1.5827290705339485,
+      "grad_norm": 0.22682647401044156,
+      "learning_rate": 2.624633431085044e-05,
+      "loss": 0.4526,
+      "step": 1201
+    },
+    {
+      "epoch": 1.5840474620962426,
+      "grad_norm": 0.22520463647521247,
+      "learning_rate": 2.6221896383186707e-05,
+      "loss": 0.469,
+      "step": 1202
+    },
+    {
+      "epoch": 1.5853658536585367,
+      "grad_norm": 0.22054934931871384,
+      "learning_rate": 2.6197458455522972e-05,
+      "loss": 0.4535,
+      "step": 1203
+    },
+    {
+      "epoch": 1.5866842452208307,
+      "grad_norm": 0.2038839366800306,
+      "learning_rate": 2.6173020527859237e-05,
+      "loss": 0.4722,
+      "step": 1204
+    },
+    {
+      "epoch": 1.5880026367831246,
+      "grad_norm": 0.24637393225775095,
+      "learning_rate": 2.6148582600195505e-05,
+      "loss": 0.4438,
+      "step": 1205
+    },
+    {
+      "epoch": 1.5893210283454184,
+      "grad_norm": 0.23449695408182278,
+      "learning_rate": 2.612414467253177e-05,
+      "loss": 0.4521,
+      "step": 1206
+    },
+    {
+      "epoch": 1.5906394199077125,
+      "grad_norm": 0.20735626843241153,
+      "learning_rate": 2.6099706744868034e-05,
+      "loss": 0.4511,
+      "step": 1207
+    },
+    {
+      "epoch": 1.5919578114700066,
+      "grad_norm": 0.21233490122973153,
+      "learning_rate": 2.6075268817204303e-05,
+      "loss": 0.4585,
+      "step": 1208
+    },
+    {
+      "epoch": 1.5932762030323007,
+      "grad_norm": 0.25519761485387216,
+      "learning_rate": 2.6050830889540567e-05,
+      "loss": 0.4547,
+      "step": 1209
+    },
+    {
+      "epoch": 1.5945945945945947,
+      "grad_norm": 0.22922587196312985,
+      "learning_rate": 2.6026392961876832e-05,
+      "loss": 0.4847,
+      "step": 1210
+    },
+    {
+      "epoch": 1.5959129861568886,
+      "grad_norm": 0.2302972759920558,
+      "learning_rate": 2.60019550342131e-05,
+      "loss": 0.446,
+      "step": 1211
+    },
+    {
+      "epoch": 1.5972313777191824,
+      "grad_norm": 0.2442904022520796,
+      "learning_rate": 2.5977517106549365e-05,
+      "loss": 0.4402,
+      "step": 1212
+    },
+    {
+      "epoch": 1.5985497692814765,
+      "grad_norm": 0.21565674098834503,
+      "learning_rate": 2.595307917888563e-05,
+      "loss": 0.4154,
+      "step": 1213
+    },
+    {
+      "epoch": 1.5998681608437706,
+      "grad_norm": 0.22882621816428245,
+      "learning_rate": 2.5928641251221898e-05,
+      "loss": 0.4423,
+      "step": 1214
+    },
+    {
+      "epoch": 1.6011865524060647,
+      "grad_norm": 0.24481645032809152,
+      "learning_rate": 2.5904203323558163e-05,
+      "loss": 0.4438,
+      "step": 1215
+    },
+    {
+      "epoch": 1.6025049439683587,
+      "grad_norm": 0.2296906710052342,
+      "learning_rate": 2.5879765395894428e-05,
+      "loss": 0.4708,
+      "step": 1216
+    },
+    {
+      "epoch": 1.6038233355306526,
+      "grad_norm": 0.23657724997405047,
+      "learning_rate": 2.5855327468230696e-05,
+      "loss": 0.4756,
+      "step": 1217
+    },
+    {
+      "epoch": 1.6051417270929464,
+      "grad_norm": 0.23652380652568888,
+      "learning_rate": 2.583088954056696e-05,
+      "loss": 0.4691,
+      "step": 1218
+    },
+    {
+      "epoch": 1.6064601186552405,
+      "grad_norm": 0.21761482095795945,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.4831,
+      "step": 1219
+    },
+    {
+      "epoch": 1.6077785102175346,
+      "grad_norm": 0.23740035881099092,
+      "learning_rate": 2.5782013685239494e-05,
+      "loss": 0.4641,
+      "step": 1220
+    },
+    {
+      "epoch": 1.6090969017798287,
+      "grad_norm": 0.22099353710604266,
+      "learning_rate": 2.575757575757576e-05,
+      "loss": 0.4726,
+      "step": 1221
+    },
+    {
+      "epoch": 1.6104152933421227,
+      "grad_norm": 0.2631623893208033,
+      "learning_rate": 2.5733137829912023e-05,
+      "loss": 0.4723,
+      "step": 1222
+    },
+    {
+      "epoch": 1.6117336849044166,
+      "grad_norm": 0.24148175552018708,
+      "learning_rate": 2.570869990224829e-05,
+      "loss": 0.4338,
+      "step": 1223
+    },
+    {
+      "epoch": 1.6130520764667104,
+      "grad_norm": 0.2271905112711295,
+      "learning_rate": 2.5684261974584556e-05,
+      "loss": 0.4528,
+      "step": 1224
+    },
+    {
+      "epoch": 1.6143704680290045,
+      "grad_norm": 0.27703185381466877,
+      "learning_rate": 2.565982404692082e-05,
+      "loss": 0.443,
+      "step": 1225
+    },
+    {
+      "epoch": 1.6156888595912986,
+      "grad_norm": 0.21459211588226904,
+      "learning_rate": 2.563538611925709e-05,
+      "loss": 0.4627,
+      "step": 1226
+    },
+    {
+      "epoch": 1.6170072511535927,
+      "grad_norm": 0.24920583016736422,
+      "learning_rate": 2.5610948191593354e-05,
+      "loss": 0.4539,
+      "step": 1227
+    },
+    {
+      "epoch": 1.6183256427158867,
+      "grad_norm": 0.22549936745779775,
+      "learning_rate": 2.558651026392962e-05,
+      "loss": 0.4521,
+      "step": 1228
+    },
+    {
+      "epoch": 1.6196440342781806,
+      "grad_norm": 0.23462128499716542,
+      "learning_rate": 2.5562072336265887e-05,
+      "loss": 0.4587,
+      "step": 1229
+    },
+    {
+      "epoch": 1.6209624258404747,
+      "grad_norm": 0.22540407460216183,
+      "learning_rate": 2.5537634408602152e-05,
+      "loss": 0.4757,
+      "step": 1230
+    },
+    {
+      "epoch": 1.6222808174027685,
+      "grad_norm": 0.24705655844820928,
+      "learning_rate": 2.5513196480938417e-05,
+      "loss": 0.4426,
+      "step": 1231
+    },
+    {
+      "epoch": 1.6235992089650626,
+      "grad_norm": 0.21855560736231033,
+      "learning_rate": 2.548875855327468e-05,
+      "loss": 0.4712,
+      "step": 1232
+    },
+    {
+      "epoch": 1.6249176005273567,
+      "grad_norm": 0.23212939992828677,
+      "learning_rate": 2.546432062561095e-05,
+      "loss": 0.4493,
+      "step": 1233
+    },
+    {
+      "epoch": 1.6262359920896508,
+      "grad_norm": 0.2366108780194849,
+      "learning_rate": 2.5439882697947214e-05,
+      "loss": 0.4729,
+      "step": 1234
+    },
+    {
+      "epoch": 1.6275543836519446,
+      "grad_norm": 0.2458399015335568,
+      "learning_rate": 2.541544477028348e-05,
+      "loss": 0.4647,
+      "step": 1235
+    },
+    {
+      "epoch": 1.6288727752142387,
+      "grad_norm": 0.23392218367428577,
+      "learning_rate": 2.5391006842619747e-05,
+      "loss": 0.4251,
+      "step": 1236
+    },
+    {
+      "epoch": 1.6301911667765325,
+      "grad_norm": 0.23095764731650026,
+      "learning_rate": 2.5366568914956012e-05,
+      "loss": 0.4456,
+      "step": 1237
+    },
+    {
+      "epoch": 1.6315095583388266,
+      "grad_norm": 0.23244226972368318,
+      "learning_rate": 2.5342130987292277e-05,
+      "loss": 0.418,
+      "step": 1238
+    },
+    {
+      "epoch": 1.6328279499011207,
+      "grad_norm": 0.22166744095624297,
+      "learning_rate": 2.5317693059628545e-05,
+      "loss": 0.4834,
+      "step": 1239
+    },
+    {
+      "epoch": 1.6341463414634148,
+      "grad_norm": 0.22635644083419576,
+      "learning_rate": 2.529325513196481e-05,
+      "loss": 0.4596,
+      "step": 1240
+    },
+    {
+      "epoch": 1.6354647330257086,
+      "grad_norm": 0.22203036048401,
+      "learning_rate": 2.5268817204301075e-05,
+      "loss": 0.4323,
+      "step": 1241
+    },
+    {
+      "epoch": 1.6367831245880027,
+      "grad_norm": 0.21024675209463323,
+      "learning_rate": 2.5244379276637343e-05,
+      "loss": 0.4375,
+      "step": 1242
+    },
+    {
+      "epoch": 1.6381015161502965,
+      "grad_norm": 0.2500329147072633,
+      "learning_rate": 2.5219941348973608e-05,
+      "loss": 0.4488,
+      "step": 1243
+    },
+    {
+      "epoch": 1.6394199077125906,
+      "grad_norm": 0.21070284254695207,
+      "learning_rate": 2.5195503421309873e-05,
+      "loss": 0.4446,
+      "step": 1244
+    },
+    {
+      "epoch": 1.6407382992748847,
+      "grad_norm": 0.2017501894180166,
+      "learning_rate": 2.517106549364614e-05,
+      "loss": 0.4534,
+      "step": 1245
+    },
+    {
+      "epoch": 1.6420566908371788,
+      "grad_norm": 0.23162238460925844,
+      "learning_rate": 2.5146627565982406e-05,
+      "loss": 0.4496,
+      "step": 1246
+    },
+    {
+      "epoch": 1.6433750823994726,
+      "grad_norm": 0.20943436543581134,
+      "learning_rate": 2.512218963831867e-05,
+      "loss": 0.4455,
+      "step": 1247
+    },
+    {
+      "epoch": 1.6446934739617667,
+      "grad_norm": 0.20675968879505105,
+      "learning_rate": 2.509775171065494e-05,
+      "loss": 0.4505,
+      "step": 1248
+    },
+    {
+      "epoch": 1.6460118655240605,
+      "grad_norm": 0.2170417236595002,
+      "learning_rate": 2.5073313782991203e-05,
+      "loss": 0.4542,
+      "step": 1249
+    },
+    {
+      "epoch": 1.6473302570863546,
+      "grad_norm": 0.21425720333931475,
+      "learning_rate": 2.5048875855327468e-05,
+      "loss": 0.4725,
+      "step": 1250
+    },
+    {
+      "epoch": 1.6486486486486487,
+      "grad_norm": 0.23268281621604509,
+      "learning_rate": 2.5024437927663736e-05,
+      "loss": 0.4461,
+      "step": 1251
+    },
+    {
+      "epoch": 1.6499670402109428,
+      "grad_norm": 0.21267698755988104,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4345,
+      "step": 1252
+    },
+    {
+      "epoch": 1.6512854317732366,
+      "grad_norm": 0.2149995720472877,
+      "learning_rate": 2.4975562072336266e-05,
+      "loss": 0.4554,
+      "step": 1253
+    },
+    {
+      "epoch": 1.6526038233355307,
+      "grad_norm": 0.2291270319941916,
+      "learning_rate": 2.4951124144672534e-05,
+      "loss": 0.458,
+      "step": 1254
+    },
+    {
+      "epoch": 1.6539222148978245,
+      "grad_norm": 0.23347070720561566,
+      "learning_rate": 2.49266862170088e-05,
+      "loss": 0.4371,
+      "step": 1255
+    },
+    {
+      "epoch": 1.6552406064601186,
+      "grad_norm": 0.2253607287174239,
+      "learning_rate": 2.4902248289345064e-05,
+      "loss": 0.4456,
+      "step": 1256
+    },
+    {
+      "epoch": 1.6565589980224127,
+      "grad_norm": 0.22877116722734273,
+      "learning_rate": 2.4877810361681332e-05,
+      "loss": 0.4452,
+      "step": 1257
+    },
+    {
+      "epoch": 1.6578773895847068,
+      "grad_norm": 0.2333735997946014,
+      "learning_rate": 2.4853372434017597e-05,
+      "loss": 0.4647,
+      "step": 1258
+    },
+    {
+      "epoch": 1.6591957811470006,
+      "grad_norm": 0.22649523652509274,
+      "learning_rate": 2.482893450635386e-05,
+      "loss": 0.439,
+      "step": 1259
+    },
+    {
+      "epoch": 1.6605141727092947,
+      "grad_norm": 0.21969432802211775,
+      "learning_rate": 2.480449657869013e-05,
+      "loss": 0.438,
+      "step": 1260
+    },
+    {
+      "epoch": 1.6618325642715885,
+      "grad_norm": 0.23813416163718962,
+      "learning_rate": 2.4780058651026395e-05,
+      "loss": 0.4468,
+      "step": 1261
+    },
+    {
+      "epoch": 1.6631509558338826,
+      "grad_norm": 0.23025685058621526,
+      "learning_rate": 2.475562072336266e-05,
+      "loss": 0.4252,
+      "step": 1262
+    },
+    {
+      "epoch": 1.6644693473961767,
+      "grad_norm": 0.23122140508768563,
+      "learning_rate": 2.4731182795698928e-05,
+      "loss": 0.44,
+      "step": 1263
+    },
+    {
+      "epoch": 1.6657877389584708,
+      "grad_norm": 0.2445987004542158,
+      "learning_rate": 2.4706744868035192e-05,
+      "loss": 0.4709,
+      "step": 1264
+    },
+    {
+      "epoch": 1.6671061305207646,
+      "grad_norm": 0.21233979256680444,
+      "learning_rate": 2.4682306940371457e-05,
+      "loss": 0.4433,
+      "step": 1265
+    },
+    {
+      "epoch": 1.6684245220830587,
+      "grad_norm": 0.23146314437126664,
+      "learning_rate": 2.4657869012707725e-05,
+      "loss": 0.4694,
+      "step": 1266
+    },
+    {
+      "epoch": 1.6697429136453525,
+      "grad_norm": 0.22995885972371477,
+      "learning_rate": 2.463343108504399e-05,
+      "loss": 0.4252,
+      "step": 1267
+    },
+    {
+      "epoch": 1.6710613052076466,
+      "grad_norm": 0.24700497298356397,
+      "learning_rate": 2.4608993157380255e-05,
+      "loss": 0.4646,
+      "step": 1268
+    },
+    {
+      "epoch": 1.6723796967699407,
+      "grad_norm": 0.23489806990571735,
+      "learning_rate": 2.4584555229716523e-05,
+      "loss": 0.4457,
+      "step": 1269
+    },
+    {
+      "epoch": 1.6736980883322348,
+      "grad_norm": 0.21611693440735655,
+      "learning_rate": 2.4560117302052788e-05,
+      "loss": 0.4615,
+      "step": 1270
+    },
+    {
+      "epoch": 1.6750164798945286,
+      "grad_norm": 0.23087913580244657,
+      "learning_rate": 2.4535679374389053e-05,
+      "loss": 0.4533,
+      "step": 1271
+    },
+    {
+      "epoch": 1.6763348714568227,
+      "grad_norm": 0.2184471652855449,
+      "learning_rate": 2.451124144672532e-05,
+      "loss": 0.4444,
+      "step": 1272
+    },
+    {
+      "epoch": 1.6776532630191165,
+      "grad_norm": 0.22120509023464946,
+      "learning_rate": 2.4486803519061586e-05,
+      "loss": 0.4564,
+      "step": 1273
+    },
+    {
+      "epoch": 1.6789716545814106,
+      "grad_norm": 0.2457304216166486,
+      "learning_rate": 2.446236559139785e-05,
+      "loss": 0.4636,
+      "step": 1274
+    },
+    {
+      "epoch": 1.6802900461437047,
+      "grad_norm": 0.21984907709303192,
+      "learning_rate": 2.4437927663734115e-05,
+      "loss": 0.4603,
+      "step": 1275
+    },
+    {
+      "epoch": 1.6816084377059988,
+      "grad_norm": 0.2250862284344557,
+      "learning_rate": 2.4413489736070384e-05,
+      "loss": 0.4492,
+      "step": 1276
+    },
+    {
+      "epoch": 1.6829268292682928,
+      "grad_norm": 0.24893414445142859,
+      "learning_rate": 2.438905180840665e-05,
+      "loss": 0.4336,
+      "step": 1277
+    },
+    {
+      "epoch": 1.6842452208305867,
+      "grad_norm": 0.22284842500006352,
+      "learning_rate": 2.4364613880742913e-05,
+      "loss": 0.4636,
+      "step": 1278
+    },
+    {
+      "epoch": 1.6855636123928806,
+      "grad_norm": 0.23349136256097167,
+      "learning_rate": 2.434017595307918e-05,
+      "loss": 0.4698,
+      "step": 1279
+    },
+    {
+      "epoch": 1.6868820039551746,
+      "grad_norm": 0.22463096515712372,
+      "learning_rate": 2.4315738025415446e-05,
+      "loss": 0.4703,
+      "step": 1280
+    },
+    {
+      "epoch": 1.6882003955174687,
+      "grad_norm": 0.20444239312517656,
+      "learning_rate": 2.429130009775171e-05,
+      "loss": 0.4539,
+      "step": 1281
+    },
+    {
+      "epoch": 1.6895187870797628,
+      "grad_norm": 0.22487183866266974,
+      "learning_rate": 2.426686217008798e-05,
+      "loss": 0.4141,
+      "step": 1282
+    },
+    {
+      "epoch": 1.6908371786420568,
+      "grad_norm": 0.22117602175446957,
+      "learning_rate": 2.4242424242424244e-05,
+      "loss": 0.4303,
+      "step": 1283
+    },
+    {
+      "epoch": 1.6921555702043507,
+      "grad_norm": 0.21963564879986763,
+      "learning_rate": 2.421798631476051e-05,
+      "loss": 0.4798,
+      "step": 1284
+    },
+    {
+      "epoch": 1.6934739617666446,
+      "grad_norm": 0.24835156908949316,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.4446,
+      "step": 1285
+    },
+    {
+      "epoch": 1.6947923533289386,
+      "grad_norm": 0.24081391906665406,
+      "learning_rate": 2.416911045943304e-05,
+      "loss": 0.4511,
+      "step": 1286
+    },
+    {
+      "epoch": 1.6961107448912327,
+      "grad_norm": 0.2108367880285926,
+      "learning_rate": 2.4144672531769306e-05,
+      "loss": 0.4482,
+      "step": 1287
+    },
+    {
+      "epoch": 1.6974291364535268,
+      "grad_norm": 0.239592606845252,
+      "learning_rate": 2.4120234604105575e-05,
+      "loss": 0.4457,
+      "step": 1288
+    },
+    {
+      "epoch": 1.6987475280158209,
+      "grad_norm": 0.23073211268003435,
+      "learning_rate": 2.409579667644184e-05,
+      "loss": 0.4415,
+      "step": 1289
+    },
+    {
+      "epoch": 1.7000659195781147,
+      "grad_norm": 0.23044746043228354,
+      "learning_rate": 2.4071358748778104e-05,
+      "loss": 0.4314,
+      "step": 1290
+    },
+    {
+      "epoch": 1.7013843111404086,
+      "grad_norm": 0.22103451824305184,
+      "learning_rate": 2.4046920821114372e-05,
+      "loss": 0.4229,
+      "step": 1291
+    },
+    {
+      "epoch": 1.7027027027027026,
+      "grad_norm": 0.23837305642003417,
+      "learning_rate": 2.4022482893450637e-05,
+      "loss": 0.4922,
+      "step": 1292
+    },
+    {
+      "epoch": 1.7040210942649967,
+      "grad_norm": 0.31167538781617893,
+      "learning_rate": 2.3998044965786902e-05,
+      "loss": 0.4796,
+      "step": 1293
+    },
+    {
+      "epoch": 1.7053394858272908,
+      "grad_norm": 0.21817695531587913,
+      "learning_rate": 2.397360703812317e-05,
+      "loss": 0.4328,
+      "step": 1294
+    },
+    {
+      "epoch": 1.7066578773895849,
+      "grad_norm": 0.23089547311480316,
+      "learning_rate": 2.3949169110459435e-05,
+      "loss": 0.4569,
+      "step": 1295
+    },
+    {
+      "epoch": 1.7079762689518787,
+      "grad_norm": 0.24201006348543067,
+      "learning_rate": 2.39247311827957e-05,
+      "loss": 0.4266,
+      "step": 1296
+    },
+    {
+      "epoch": 1.7092946605141726,
+      "grad_norm": 0.2317854180826016,
+      "learning_rate": 2.3900293255131968e-05,
+      "loss": 0.4657,
+      "step": 1297
+    },
+    {
+      "epoch": 1.7106130520764666,
+      "grad_norm": 0.21481838477644827,
+      "learning_rate": 2.3875855327468233e-05,
+      "loss": 0.4586,
+      "step": 1298
+    },
+    {
+      "epoch": 1.7119314436387607,
+      "grad_norm": 0.22501440290134972,
+      "learning_rate": 2.3851417399804498e-05,
+      "loss": 0.4645,
+      "step": 1299
+    },
+    {
+      "epoch": 1.7132498352010548,
+      "grad_norm": 0.2178923958590864,
+      "learning_rate": 2.3826979472140766e-05,
+      "loss": 0.4296,
+      "step": 1300
+    },
+    {
+      "epoch": 1.7145682267633489,
+      "grad_norm": 0.20740774934211023,
+      "learning_rate": 2.380254154447703e-05,
+      "loss": 0.4437,
+      "step": 1301
+    },
+    {
+      "epoch": 1.7158866183256427,
+      "grad_norm": 0.21325368863665742,
+      "learning_rate": 2.3778103616813295e-05,
+      "loss": 0.4302,
+      "step": 1302
+    },
+    {
+      "epoch": 1.7172050098879366,
+      "grad_norm": 0.21731707500641012,
+      "learning_rate": 2.3753665689149564e-05,
+      "loss": 0.4431,
+      "step": 1303
+    },
+    {
+      "epoch": 1.7185234014502306,
+      "grad_norm": 0.21637297652706527,
+      "learning_rate": 2.372922776148583e-05,
+      "loss": 0.4395,
+      "step": 1304
+    },
+    {
+      "epoch": 1.7198417930125247,
+      "grad_norm": 0.2262704966596932,
+      "learning_rate": 2.3704789833822093e-05,
+      "loss": 0.4195,
+      "step": 1305
+    },
+    {
+      "epoch": 1.7211601845748188,
+      "grad_norm": 0.21876749736263493,
+      "learning_rate": 2.368035190615836e-05,
+      "loss": 0.4093,
+      "step": 1306
+    },
+    {
+      "epoch": 1.7224785761371129,
+      "grad_norm": 0.22856619169624617,
+      "learning_rate": 2.3655913978494626e-05,
+      "loss": 0.452,
+      "step": 1307
+    },
+    {
+      "epoch": 1.7237969676994067,
+      "grad_norm": 0.2433346628027909,
+      "learning_rate": 2.363147605083089e-05,
+      "loss": 0.4694,
+      "step": 1308
+    },
+    {
+      "epoch": 1.7251153592617006,
+      "grad_norm": 0.21434661728045573,
+      "learning_rate": 2.360703812316716e-05,
+      "loss": 0.4626,
+      "step": 1309
+    },
+    {
+      "epoch": 1.7264337508239946,
+      "grad_norm": 0.2293513088045069,
+      "learning_rate": 2.3582600195503424e-05,
+      "loss": 0.4365,
+      "step": 1310
+    },
+    {
+      "epoch": 1.7277521423862887,
+      "grad_norm": 0.24428926118478997,
+      "learning_rate": 2.355816226783969e-05,
+      "loss": 0.4743,
+      "step": 1311
+    },
+    {
+      "epoch": 1.7290705339485828,
+      "grad_norm": 0.2287615764401124,
+      "learning_rate": 2.3533724340175957e-05,
+      "loss": 0.4262,
+      "step": 1312
+    },
+    {
+      "epoch": 1.7303889255108769,
+      "grad_norm": 0.2137720982360762,
+      "learning_rate": 2.3509286412512222e-05,
+      "loss": 0.4422,
+      "step": 1313
+    },
+    {
+      "epoch": 1.7317073170731707,
+      "grad_norm": 0.2172646709179954,
+      "learning_rate": 2.3484848484848487e-05,
+      "loss": 0.4369,
+      "step": 1314
+    },
+    {
+      "epoch": 1.7330257086354646,
+      "grad_norm": 0.23481114526676852,
+      "learning_rate": 2.3460410557184755e-05,
+      "loss": 0.4379,
+      "step": 1315
+    },
+    {
+      "epoch": 1.7343441001977586,
+      "grad_norm": 0.22494156411555105,
+      "learning_rate": 2.3435972629521016e-05,
+      "loss": 0.4627,
+      "step": 1316
+    },
+    {
+      "epoch": 1.7356624917600527,
+      "grad_norm": 0.232305054590693,
+      "learning_rate": 2.341153470185728e-05,
+      "loss": 0.4778,
+      "step": 1317
+    },
+    {
+      "epoch": 1.7369808833223468,
+      "grad_norm": 0.2241232571591852,
+      "learning_rate": 2.338709677419355e-05,
+      "loss": 0.4555,
+      "step": 1318
+    },
+    {
+      "epoch": 1.7382992748846409,
+      "grad_norm": 0.2297129599006242,
+      "learning_rate": 2.3362658846529814e-05,
+      "loss": 0.4676,
+      "step": 1319
+    },
+    {
+      "epoch": 1.7396176664469347,
+      "grad_norm": 0.20981392935025628,
+      "learning_rate": 2.333822091886608e-05,
+      "loss": 0.4481,
+      "step": 1320
+    },
+    {
+      "epoch": 1.7409360580092288,
+      "grad_norm": 0.22737136718565307,
+      "learning_rate": 2.3313782991202347e-05,
+      "loss": 0.4328,
+      "step": 1321
+    },
+    {
+      "epoch": 1.7422544495715226,
+      "grad_norm": 0.24472193131787728,
+      "learning_rate": 2.3289345063538612e-05,
+      "loss": 0.4443,
+      "step": 1322
+    },
+    {
+      "epoch": 1.7435728411338167,
+      "grad_norm": 0.20702576868902212,
+      "learning_rate": 2.3264907135874877e-05,
+      "loss": 0.4185,
+      "step": 1323
+    },
+    {
+      "epoch": 1.7448912326961108,
+      "grad_norm": 0.23089623806478482,
+      "learning_rate": 2.3240469208211145e-05,
+      "loss": 0.438,
+      "step": 1324
+    },
+    {
+      "epoch": 1.7462096242584049,
+      "grad_norm": 0.25529748116520273,
+      "learning_rate": 2.321603128054741e-05,
+      "loss": 0.4694,
+      "step": 1325
+    },
+    {
+      "epoch": 1.7475280158206987,
+      "grad_norm": 0.2179899465010871,
+      "learning_rate": 2.3191593352883674e-05,
+      "loss": 0.4875,
+      "step": 1326
+    },
+    {
+      "epoch": 1.7488464073829928,
+      "grad_norm": 0.22557046515612472,
+      "learning_rate": 2.3167155425219943e-05,
+      "loss": 0.4745,
+      "step": 1327
+    },
+    {
+      "epoch": 1.7501647989452866,
+      "grad_norm": 0.22051811530571566,
+      "learning_rate": 2.3142717497556207e-05,
+      "loss": 0.439,
+      "step": 1328
+    },
+    {
+      "epoch": 1.7514831905075807,
+      "grad_norm": 0.2561866543932515,
+      "learning_rate": 2.3118279569892472e-05,
+      "loss": 0.4626,
+      "step": 1329
+    },
+    {
+      "epoch": 1.7528015820698748,
+      "grad_norm": 0.2579665810833442,
+      "learning_rate": 2.309384164222874e-05,
+      "loss": 0.4659,
+      "step": 1330
+    },
+    {
+      "epoch": 1.7541199736321689,
+      "grad_norm": 0.23165280912787573,
+      "learning_rate": 2.3069403714565005e-05,
+      "loss": 0.4584,
+      "step": 1331
+    },
+    {
+      "epoch": 1.7554383651944627,
+      "grad_norm": 0.26782028339364883,
+      "learning_rate": 2.304496578690127e-05,
+      "loss": 0.4316,
+      "step": 1332
+    },
+    {
+      "epoch": 1.7567567567567568,
+      "grad_norm": 0.22073900296688956,
+      "learning_rate": 2.3020527859237538e-05,
+      "loss": 0.4683,
+      "step": 1333
+    },
+    {
+      "epoch": 1.7580751483190507,
+      "grad_norm": 0.21569757180242494,
+      "learning_rate": 2.2996089931573803e-05,
+      "loss": 0.4169,
+      "step": 1334
+    },
+    {
+      "epoch": 1.7593935398813447,
+      "grad_norm": 0.27958462391295685,
+      "learning_rate": 2.2971652003910068e-05,
+      "loss": 0.4844,
+      "step": 1335
+    },
+    {
+      "epoch": 1.7607119314436388,
+      "grad_norm": 0.23671863055323814,
+      "learning_rate": 2.2947214076246336e-05,
+      "loss": 0.4556,
+      "step": 1336
+    },
+    {
+      "epoch": 1.7620303230059329,
+      "grad_norm": 0.235274174394857,
+      "learning_rate": 2.29227761485826e-05,
+      "loss": 0.4446,
+      "step": 1337
+    },
+    {
+      "epoch": 1.7633487145682267,
+      "grad_norm": 0.23716249112884563,
+      "learning_rate": 2.2898338220918866e-05,
+      "loss": 0.4777,
+      "step": 1338
+    },
+    {
+      "epoch": 1.7646671061305208,
+      "grad_norm": 0.223649541813464,
+      "learning_rate": 2.2873900293255134e-05,
+      "loss": 0.4373,
+      "step": 1339
+    },
+    {
+      "epoch": 1.7659854976928147,
+      "grad_norm": 0.21981968777565494,
+      "learning_rate": 2.28494623655914e-05,
+      "loss": 0.4263,
+      "step": 1340
+    },
+    {
+      "epoch": 1.7673038892551087,
+      "grad_norm": 0.22895233760285355,
+      "learning_rate": 2.2825024437927663e-05,
+      "loss": 0.451,
+      "step": 1341
+    },
+    {
+      "epoch": 1.7686222808174028,
+      "grad_norm": 0.22187270516126548,
+      "learning_rate": 2.280058651026393e-05,
+      "loss": 0.4428,
+      "step": 1342
+    },
+    {
+      "epoch": 1.7699406723796969,
+      "grad_norm": 0.2264016018955621,
+      "learning_rate": 2.2776148582600196e-05,
+      "loss": 0.4699,
+      "step": 1343
+    },
+    {
+      "epoch": 1.7712590639419907,
+      "grad_norm": 0.24782811620734982,
+      "learning_rate": 2.275171065493646e-05,
+      "loss": 0.4498,
+      "step": 1344
+    },
+    {
+      "epoch": 1.7725774555042848,
+      "grad_norm": 0.2141177615292286,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 0.4536,
+      "step": 1345
+    },
+    {
+      "epoch": 1.7738958470665787,
+      "grad_norm": 0.24220196828496443,
+      "learning_rate": 2.2702834799608994e-05,
+      "loss": 0.4184,
+      "step": 1346
+    },
+    {
+      "epoch": 1.7752142386288727,
+      "grad_norm": 0.256770041421616,
+      "learning_rate": 2.267839687194526e-05,
+      "loss": 0.4947,
+      "step": 1347
+    },
+    {
+      "epoch": 1.7765326301911668,
+      "grad_norm": 0.23637849914384568,
+      "learning_rate": 2.2653958944281524e-05,
+      "loss": 0.438,
+      "step": 1348
+    },
+    {
+      "epoch": 1.7778510217534609,
+      "grad_norm": 0.2428007538150315,
+      "learning_rate": 2.2629521016617792e-05,
+      "loss": 0.4274,
+      "step": 1349
+    },
+    {
+      "epoch": 1.7791694133157547,
+      "grad_norm": 0.22854729030045823,
+      "learning_rate": 2.2605083088954057e-05,
+      "loss": 0.466,
+      "step": 1350
+    },
+    {
+      "epoch": 1.7804878048780488,
+      "grad_norm": 0.2341226837853732,
+      "learning_rate": 2.258064516129032e-05,
+      "loss": 0.4517,
+      "step": 1351
+    },
+    {
+      "epoch": 1.7818061964403427,
+      "grad_norm": 0.2526369348134983,
+      "learning_rate": 2.255620723362659e-05,
+      "loss": 0.4769,
+      "step": 1352
+    },
+    {
+      "epoch": 1.7831245880026367,
+      "grad_norm": 0.23040489930014424,
+      "learning_rate": 2.2531769305962854e-05,
+      "loss": 0.4444,
+      "step": 1353
+    },
+    {
+      "epoch": 1.7844429795649308,
+      "grad_norm": 0.25277894502411674,
+      "learning_rate": 2.250733137829912e-05,
+      "loss": 0.4584,
+      "step": 1354
+    },
+    {
+      "epoch": 1.7857613711272249,
+      "grad_norm": 0.2419935891975454,
+      "learning_rate": 2.2482893450635387e-05,
+      "loss": 0.4389,
+      "step": 1355
+    },
+    {
+      "epoch": 1.7870797626895187,
+      "grad_norm": 0.24130822824806356,
+      "learning_rate": 2.2458455522971652e-05,
+      "loss": 0.4429,
+      "step": 1356
+    },
+    {
+      "epoch": 1.7883981542518128,
+      "grad_norm": 0.2447651773200519,
+      "learning_rate": 2.2434017595307917e-05,
+      "loss": 0.4392,
+      "step": 1357
+    },
+    {
+      "epoch": 1.7897165458141067,
+      "grad_norm": 0.2622004011919672,
+      "learning_rate": 2.2409579667644185e-05,
+      "loss": 0.4372,
+      "step": 1358
+    },
+    {
+      "epoch": 1.7910349373764007,
+      "grad_norm": 0.2516879903766481,
+      "learning_rate": 2.238514173998045e-05,
+      "loss": 0.4617,
+      "step": 1359
+    },
+    {
+      "epoch": 1.7923533289386948,
+      "grad_norm": 0.254172539072606,
+      "learning_rate": 2.2360703812316715e-05,
+      "loss": 0.4721,
+      "step": 1360
+    },
+    {
+      "epoch": 1.7936717205009889,
+      "grad_norm": 0.26124419131789506,
+      "learning_rate": 2.2336265884652983e-05,
+      "loss": 0.4281,
+      "step": 1361
+    },
+    {
+      "epoch": 1.794990112063283,
+      "grad_norm": 0.23337321060270946,
+      "learning_rate": 2.2311827956989248e-05,
+      "loss": 0.444,
+      "step": 1362
+    },
+    {
+      "epoch": 1.7963085036255768,
+      "grad_norm": 0.2735398685729882,
+      "learning_rate": 2.2287390029325513e-05,
+      "loss": 0.4513,
+      "step": 1363
+    },
+    {
+      "epoch": 1.7976268951878707,
+      "grad_norm": 0.26189564562505036,
+      "learning_rate": 2.226295210166178e-05,
+      "loss": 0.4455,
+      "step": 1364
+    },
+    {
+      "epoch": 1.7989452867501647,
+      "grad_norm": 0.2154634640484005,
+      "learning_rate": 2.2238514173998046e-05,
+      "loss": 0.435,
+      "step": 1365
+    },
+    {
+      "epoch": 1.8002636783124588,
+      "grad_norm": 0.2834993924008965,
+      "learning_rate": 2.221407624633431e-05,
+      "loss": 0.46,
+      "step": 1366
+    },
+    {
+      "epoch": 1.801582069874753,
+      "grad_norm": 0.2508261989060455,
+      "learning_rate": 2.218963831867058e-05,
+      "loss": 0.4462,
+      "step": 1367
+    },
+    {
+      "epoch": 1.802900461437047,
+      "grad_norm": 0.21522683376842433,
+      "learning_rate": 2.2165200391006843e-05,
+      "loss": 0.4397,
+      "step": 1368
+    },
+    {
+      "epoch": 1.8042188529993408,
+      "grad_norm": 0.25393837605258157,
+      "learning_rate": 2.2140762463343108e-05,
+      "loss": 0.4246,
+      "step": 1369
+    },
+    {
+      "epoch": 1.8055372445616347,
+      "grad_norm": 0.25457545464896003,
+      "learning_rate": 2.2116324535679376e-05,
+      "loss": 0.4429,
+      "step": 1370
+    },
+    {
+      "epoch": 1.8068556361239287,
+      "grad_norm": 0.2063821861981881,
+      "learning_rate": 2.209188660801564e-05,
+      "loss": 0.4492,
+      "step": 1371
+    },
+    {
+      "epoch": 1.8081740276862228,
+      "grad_norm": 0.2804726061827568,
+      "learning_rate": 2.2067448680351906e-05,
+      "loss": 0.4359,
+      "step": 1372
+    },
+    {
+      "epoch": 1.809492419248517,
+      "grad_norm": 0.26694672598382835,
+      "learning_rate": 2.2043010752688174e-05,
+      "loss": 0.4531,
+      "step": 1373
+    },
+    {
+      "epoch": 1.810810810810811,
+      "grad_norm": 0.2506760002737737,
+      "learning_rate": 2.201857282502444e-05,
+      "loss": 0.4733,
+      "step": 1374
+    },
+    {
+      "epoch": 1.8121292023731048,
+      "grad_norm": 0.2904307102961233,
+      "learning_rate": 2.1994134897360704e-05,
+      "loss": 0.4498,
+      "step": 1375
+    },
+    {
+      "epoch": 1.8134475939353987,
+      "grad_norm": 0.22514423487678403,
+      "learning_rate": 2.1969696969696972e-05,
+      "loss": 0.4866,
+      "step": 1376
+    },
+    {
+      "epoch": 1.8147659854976927,
+      "grad_norm": 0.2245004561001592,
+      "learning_rate": 2.1945259042033237e-05,
+      "loss": 0.4535,
+      "step": 1377
+    },
+    {
+      "epoch": 1.8160843770599868,
+      "grad_norm": 0.2342642759445307,
+      "learning_rate": 2.19208211143695e-05,
+      "loss": 0.4414,
+      "step": 1378
+    },
+    {
+      "epoch": 1.817402768622281,
+      "grad_norm": 0.23572898527997294,
+      "learning_rate": 2.189638318670577e-05,
+      "loss": 0.4494,
+      "step": 1379
+    },
+    {
+      "epoch": 1.818721160184575,
+      "grad_norm": 0.2090041796382176,
+      "learning_rate": 2.1871945259042035e-05,
+      "loss": 0.4649,
+      "step": 1380
+    },
+    {
+      "epoch": 1.8200395517468688,
+      "grad_norm": 0.2220688911524869,
+      "learning_rate": 2.18475073313783e-05,
+      "loss": 0.4386,
+      "step": 1381
+    },
+    {
+      "epoch": 1.8213579433091627,
+      "grad_norm": 0.20031419624597685,
+      "learning_rate": 2.1823069403714568e-05,
+      "loss": 0.4125,
+      "step": 1382
+    },
+    {
+      "epoch": 1.8226763348714567,
+      "grad_norm": 0.2158482449569417,
+      "learning_rate": 2.1798631476050832e-05,
+      "loss": 0.4265,
+      "step": 1383
+    },
+    {
+      "epoch": 1.8239947264337508,
+      "grad_norm": 0.21558160683532104,
+      "learning_rate": 2.1774193548387097e-05,
+      "loss": 0.4445,
+      "step": 1384
+    },
+    {
+      "epoch": 1.825313117996045,
+      "grad_norm": 0.2150254725946213,
+      "learning_rate": 2.1749755620723365e-05,
+      "loss": 0.4514,
+      "step": 1385
+    },
+    {
+      "epoch": 1.826631509558339,
+      "grad_norm": 0.24043326762601872,
+      "learning_rate": 2.172531769305963e-05,
+      "loss": 0.4525,
+      "step": 1386
+    },
+    {
+      "epoch": 1.8279499011206328,
+      "grad_norm": 0.2228895774787868,
+      "learning_rate": 2.1700879765395895e-05,
+      "loss": 0.4134,
+      "step": 1387
+    },
+    {
+      "epoch": 1.8292682926829267,
+      "grad_norm": 0.2184531932038404,
+      "learning_rate": 2.1676441837732163e-05,
+      "loss": 0.446,
+      "step": 1388
+    },
+    {
+      "epoch": 1.8305866842452208,
+      "grad_norm": 0.216509313078702,
+      "learning_rate": 2.1652003910068428e-05,
+      "loss": 0.4279,
+      "step": 1389
+    },
+    {
+      "epoch": 1.8319050758075148,
+      "grad_norm": 0.21902511225836463,
+      "learning_rate": 2.1627565982404693e-05,
+      "loss": 0.458,
+      "step": 1390
+    },
+    {
+      "epoch": 1.833223467369809,
+      "grad_norm": 0.23485995489901926,
+      "learning_rate": 2.160312805474096e-05,
+      "loss": 0.432,
+      "step": 1391
+    },
+    {
+      "epoch": 1.834541858932103,
+      "grad_norm": 0.228532082051778,
+      "learning_rate": 2.1578690127077226e-05,
+      "loss": 0.4519,
+      "step": 1392
+    },
+    {
+      "epoch": 1.8358602504943968,
+      "grad_norm": 0.2369538756231286,
+      "learning_rate": 2.155425219941349e-05,
+      "loss": 0.4624,
+      "step": 1393
+    },
+    {
+      "epoch": 1.8371786420566907,
+      "grad_norm": 0.2195942906885903,
+      "learning_rate": 2.1529814271749755e-05,
+      "loss": 0.4712,
+      "step": 1394
+    },
+    {
+      "epoch": 1.8384970336189848,
+      "grad_norm": 0.24501854675609572,
+      "learning_rate": 2.1505376344086024e-05,
+      "loss": 0.4512,
+      "step": 1395
+    },
+    {
+      "epoch": 1.8398154251812788,
+      "grad_norm": 0.24123397748149958,
+      "learning_rate": 2.148093841642229e-05,
+      "loss": 0.4288,
+      "step": 1396
+    },
+    {
+      "epoch": 1.841133816743573,
+      "grad_norm": 0.21632039858086205,
+      "learning_rate": 2.1456500488758553e-05,
+      "loss": 0.4347,
+      "step": 1397
+    },
+    {
+      "epoch": 1.842452208305867,
+      "grad_norm": 0.21812591328997558,
+      "learning_rate": 2.143206256109482e-05,
+      "loss": 0.4386,
+      "step": 1398
+    },
+    {
+      "epoch": 1.8437705998681608,
+      "grad_norm": 0.24464804420946357,
+      "learning_rate": 2.1407624633431086e-05,
+      "loss": 0.4394,
+      "step": 1399
+    },
+    {
+      "epoch": 1.8450889914304547,
+      "grad_norm": 0.2294531926833806,
+      "learning_rate": 2.138318670576735e-05,
+      "loss": 0.4526,
+      "step": 1400
+    },
+    {
+      "epoch": 1.8464073829927488,
+      "grad_norm": 0.23180517646910126,
+      "learning_rate": 2.135874877810362e-05,
+      "loss": 0.4517,
+      "step": 1401
+    },
+    {
+      "epoch": 1.8477257745550428,
+      "grad_norm": 0.27563734102397836,
+      "learning_rate": 2.1334310850439884e-05,
+      "loss": 0.4226,
+      "step": 1402
+    },
+    {
+      "epoch": 1.849044166117337,
+      "grad_norm": 0.26895530099591525,
+      "learning_rate": 2.130987292277615e-05,
+      "loss": 0.4675,
+      "step": 1403
+    },
+    {
+      "epoch": 1.850362557679631,
+      "grad_norm": 0.26056967279489024,
+      "learning_rate": 2.1285434995112417e-05,
+      "loss": 0.4518,
+      "step": 1404
+    },
+    {
+      "epoch": 1.8516809492419248,
+      "grad_norm": 0.251794388672966,
+      "learning_rate": 2.126099706744868e-05,
+      "loss": 0.4678,
+      "step": 1405
+    },
+    {
+      "epoch": 1.852999340804219,
+      "grad_norm": 0.24434435325231865,
+      "learning_rate": 2.1236559139784946e-05,
+      "loss": 0.4399,
+      "step": 1406
+    },
+    {
+      "epoch": 1.8543177323665128,
+      "grad_norm": 0.2540188275404348,
+      "learning_rate": 2.1212121212121215e-05,
+      "loss": 0.4336,
+      "step": 1407
+    },
+    {
+      "epoch": 1.8556361239288068,
+      "grad_norm": 0.24991274588197274,
+      "learning_rate": 2.118768328445748e-05,
+      "loss": 0.4239,
+      "step": 1408
+    },
+    {
+      "epoch": 1.856954515491101,
+      "grad_norm": 0.20094206946741394,
+      "learning_rate": 2.1163245356793744e-05,
+      "loss": 0.4228,
+      "step": 1409
+    },
+    {
+      "epoch": 1.858272907053395,
+      "grad_norm": 0.22651875217478484,
+      "learning_rate": 2.1138807429130012e-05,
+      "loss": 0.4437,
+      "step": 1410
+    },
+    {
+      "epoch": 1.8595912986156888,
+      "grad_norm": 0.2644510314668226,
+      "learning_rate": 2.1114369501466277e-05,
+      "loss": 0.4055,
+      "step": 1411
+    },
+    {
+      "epoch": 1.860909690177983,
+      "grad_norm": 0.21155431435888983,
+      "learning_rate": 2.1089931573802542e-05,
+      "loss": 0.4374,
+      "step": 1412
+    },
+    {
+      "epoch": 1.8622280817402768,
+      "grad_norm": 0.24390523530952146,
+      "learning_rate": 2.106549364613881e-05,
+      "loss": 0.4724,
+      "step": 1413
+    },
+    {
+      "epoch": 1.8635464733025708,
+      "grad_norm": 0.23689126529657917,
+      "learning_rate": 2.1041055718475075e-05,
+      "loss": 0.4408,
+      "step": 1414
+    },
+    {
+      "epoch": 1.864864864864865,
+      "grad_norm": 0.23149979394587072,
+      "learning_rate": 2.101661779081134e-05,
+      "loss": 0.4753,
+      "step": 1415
+    },
+    {
+      "epoch": 1.866183256427159,
+      "grad_norm": 0.26660937210824354,
+      "learning_rate": 2.0992179863147608e-05,
+      "loss": 0.4581,
+      "step": 1416
+    },
+    {
+      "epoch": 1.8675016479894528,
+      "grad_norm": 0.2506263369288428,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.4417,
+      "step": 1417
+    },
+    {
+      "epoch": 1.868820039551747,
+      "grad_norm": 0.2045229210860702,
+      "learning_rate": 2.0943304007820138e-05,
+      "loss": 0.4412,
+      "step": 1418
+    },
+    {
+      "epoch": 1.8701384311140408,
+      "grad_norm": 0.22455859343981052,
+      "learning_rate": 2.0918866080156406e-05,
+      "loss": 0.4255,
+      "step": 1419
+    },
+    {
+      "epoch": 1.8714568226763348,
+      "grad_norm": 0.2299753780638211,
+      "learning_rate": 2.089442815249267e-05,
+      "loss": 0.4336,
+      "step": 1420
+    },
+    {
+      "epoch": 1.872775214238629,
+      "grad_norm": 0.21763195257063978,
+      "learning_rate": 2.0869990224828935e-05,
+      "loss": 0.4549,
+      "step": 1421
+    },
+    {
+      "epoch": 1.874093605800923,
+      "grad_norm": 0.21072894760577596,
+      "learning_rate": 2.0845552297165204e-05,
+      "loss": 0.4435,
+      "step": 1422
+    },
+    {
+      "epoch": 1.8754119973632168,
+      "grad_norm": 0.24131733852001216,
+      "learning_rate": 2.082111436950147e-05,
+      "loss": 0.4351,
+      "step": 1423
+    },
+    {
+      "epoch": 1.876730388925511,
+      "grad_norm": 0.2234317303780953,
+      "learning_rate": 2.0796676441837733e-05,
+      "loss": 0.4655,
+      "step": 1424
+    },
+    {
+      "epoch": 1.8780487804878048,
+      "grad_norm": 0.22295381356244265,
+      "learning_rate": 2.0772238514174e-05,
+      "loss": 0.4478,
+      "step": 1425
+    },
+    {
+      "epoch": 1.8793671720500988,
+      "grad_norm": 0.23782400940349355,
+      "learning_rate": 2.0747800586510266e-05,
+      "loss": 0.4387,
+      "step": 1426
+    },
+    {
+      "epoch": 1.880685563612393,
+      "grad_norm": 0.22863237943489276,
+      "learning_rate": 2.072336265884653e-05,
+      "loss": 0.4413,
+      "step": 1427
+    },
+    {
+      "epoch": 1.882003955174687,
+      "grad_norm": 0.2118748712008781,
+      "learning_rate": 2.06989247311828e-05,
+      "loss": 0.4588,
+      "step": 1428
+    },
+    {
+      "epoch": 1.8833223467369808,
+      "grad_norm": 0.22082714819650182,
+      "learning_rate": 2.0674486803519064e-05,
+      "loss": 0.4293,
+      "step": 1429
+    },
+    {
+      "epoch": 1.884640738299275,
+      "grad_norm": 0.21518579207366526,
+      "learning_rate": 2.065004887585533e-05,
+      "loss": 0.4526,
+      "step": 1430
+    },
+    {
+      "epoch": 1.8859591298615688,
+      "grad_norm": 0.22099523077921776,
+      "learning_rate": 2.0625610948191597e-05,
+      "loss": 0.424,
+      "step": 1431
+    },
+    {
+      "epoch": 1.8872775214238628,
+      "grad_norm": 0.204599396684548,
+      "learning_rate": 2.0601173020527862e-05,
+      "loss": 0.428,
+      "step": 1432
+    },
+    {
+      "epoch": 1.888595912986157,
+      "grad_norm": 0.21977682949521715,
+      "learning_rate": 2.0576735092864127e-05,
+      "loss": 0.4484,
+      "step": 1433
+    },
+    {
+      "epoch": 1.889914304548451,
+      "grad_norm": 0.20855874123970927,
+      "learning_rate": 2.0552297165200395e-05,
+      "loss": 0.4696,
+      "step": 1434
+    },
+    {
+      "epoch": 1.8912326961107448,
+      "grad_norm": 0.21881259557661692,
+      "learning_rate": 2.052785923753666e-05,
+      "loss": 0.4567,
+      "step": 1435
+    },
+    {
+      "epoch": 1.892551087673039,
+      "grad_norm": 0.2237771936113666,
+      "learning_rate": 2.0503421309872924e-05,
+      "loss": 0.4454,
+      "step": 1436
+    },
+    {
+      "epoch": 1.8938694792353328,
+      "grad_norm": 0.20085980661207817,
+      "learning_rate": 2.0478983382209193e-05,
+      "loss": 0.4406,
+      "step": 1437
+    },
+    {
+      "epoch": 1.8951878707976269,
+      "grad_norm": 0.21762236178076172,
+      "learning_rate": 2.0454545454545457e-05,
+      "loss": 0.4386,
+      "step": 1438
+    },
+    {
+      "epoch": 1.896506262359921,
+      "grad_norm": 0.21658612596048446,
+      "learning_rate": 2.0430107526881722e-05,
+      "loss": 0.4724,
+      "step": 1439
+    },
+    {
+      "epoch": 1.897824653922215,
+      "grad_norm": 0.21850630859622155,
+      "learning_rate": 2.0405669599217987e-05,
+      "loss": 0.431,
+      "step": 1440
+    },
+    {
+      "epoch": 1.8991430454845089,
+      "grad_norm": 0.23028290145741412,
+      "learning_rate": 2.0381231671554255e-05,
+      "loss": 0.4375,
+      "step": 1441
+    },
+    {
+      "epoch": 1.900461437046803,
+      "grad_norm": 0.651605294714195,
+      "learning_rate": 2.035679374389052e-05,
+      "loss": 0.477,
+      "step": 1442
+    },
+    {
+      "epoch": 1.9017798286090968,
+      "grad_norm": 0.21214005309700326,
+      "learning_rate": 2.0332355816226785e-05,
+      "loss": 0.4316,
+      "step": 1443
+    },
+    {
+      "epoch": 1.9030982201713909,
+      "grad_norm": 0.24009145385556596,
+      "learning_rate": 2.030791788856305e-05,
+      "loss": 0.4297,
+      "step": 1444
+    },
+    {
+      "epoch": 1.904416611733685,
+      "grad_norm": 0.21621597025689643,
+      "learning_rate": 2.0283479960899314e-05,
+      "loss": 0.4298,
+      "step": 1445
+    },
+    {
+      "epoch": 1.905735003295979,
+      "grad_norm": 0.2022455838015858,
+      "learning_rate": 2.0259042033235583e-05,
+      "loss": 0.4207,
+      "step": 1446
+    },
+    {
+      "epoch": 1.9070533948582729,
+      "grad_norm": 0.20915315275765248,
+      "learning_rate": 2.0234604105571847e-05,
+      "loss": 0.4745,
+      "step": 1447
+    },
+    {
+      "epoch": 1.908371786420567,
+      "grad_norm": 0.22393522122350662,
+      "learning_rate": 2.0210166177908112e-05,
+      "loss": 0.4395,
+      "step": 1448
+    },
+    {
+      "epoch": 1.9096901779828608,
+      "grad_norm": 0.2181302981685662,
+      "learning_rate": 2.018572825024438e-05,
+      "loss": 0.4407,
+      "step": 1449
+    },
+    {
+      "epoch": 1.9110085695451549,
+      "grad_norm": 0.2155156493300112,
+      "learning_rate": 2.0161290322580645e-05,
+      "loss": 0.4339,
+      "step": 1450
+    },
+    {
+      "epoch": 1.912326961107449,
+      "grad_norm": 0.22378088719685169,
+      "learning_rate": 2.013685239491691e-05,
+      "loss": 0.4637,
+      "step": 1451
+    },
+    {
+      "epoch": 1.913645352669743,
+      "grad_norm": 0.21434327715534415,
+      "learning_rate": 2.0112414467253178e-05,
+      "loss": 0.4564,
+      "step": 1452
+    },
+    {
+      "epoch": 1.914963744232037,
+      "grad_norm": 0.22687657236352482,
+      "learning_rate": 2.0087976539589443e-05,
+      "loss": 0.4213,
+      "step": 1453
+    },
+    {
+      "epoch": 1.916282135794331,
+      "grad_norm": 0.20999783626081867,
+      "learning_rate": 2.0063538611925708e-05,
+      "loss": 0.4158,
+      "step": 1454
+    },
+    {
+      "epoch": 1.9176005273566248,
+      "grad_norm": 0.20221328784747078,
+      "learning_rate": 2.0039100684261976e-05,
+      "loss": 0.4369,
+      "step": 1455
+    },
+    {
+      "epoch": 1.9189189189189189,
+      "grad_norm": 0.2512923765683733,
+      "learning_rate": 2.001466275659824e-05,
+      "loss": 0.4246,
+      "step": 1456
+    },
+    {
+      "epoch": 1.920237310481213,
+      "grad_norm": 0.21030934220744288,
+      "learning_rate": 1.9990224828934506e-05,
+      "loss": 0.445,
+      "step": 1457
+    },
+    {
+      "epoch": 1.921555702043507,
+      "grad_norm": 0.21969150327660414,
+      "learning_rate": 1.9965786901270774e-05,
+      "loss": 0.4185,
+      "step": 1458
+    },
+    {
+      "epoch": 1.922874093605801,
+      "grad_norm": 0.23745382348913452,
+      "learning_rate": 1.994134897360704e-05,
+      "loss": 0.4619,
+      "step": 1459
+    },
+    {
+      "epoch": 1.924192485168095,
+      "grad_norm": 0.24313137777583116,
+      "learning_rate": 1.9916911045943303e-05,
+      "loss": 0.4342,
+      "step": 1460
+    },
+    {
+      "epoch": 1.9255108767303888,
+      "grad_norm": 0.22186628246677192,
+      "learning_rate": 1.989247311827957e-05,
+      "loss": 0.441,
+      "step": 1461
+    },
+    {
+      "epoch": 1.9268292682926829,
+      "grad_norm": 0.22903136826529763,
+      "learning_rate": 1.9868035190615836e-05,
+      "loss": 0.4836,
+      "step": 1462
+    },
+    {
+      "epoch": 1.928147659854977,
+      "grad_norm": 0.27330156080316836,
+      "learning_rate": 1.98435972629521e-05,
+      "loss": 0.4468,
+      "step": 1463
+    },
+    {
+      "epoch": 1.929466051417271,
+      "grad_norm": 0.23868711145859867,
+      "learning_rate": 1.9819159335288366e-05,
+      "loss": 0.4379,
+      "step": 1464
+    },
+    {
+      "epoch": 1.930784442979565,
+      "grad_norm": 0.2142721419620576,
+      "learning_rate": 1.9794721407624634e-05,
+      "loss": 0.4435,
+      "step": 1465
+    },
+    {
+      "epoch": 1.932102834541859,
+      "grad_norm": 0.23487918947296993,
+      "learning_rate": 1.97702834799609e-05,
+      "loss": 0.4639,
+      "step": 1466
+    },
+    {
+      "epoch": 1.9334212261041528,
+      "grad_norm": 0.24842392925116136,
+      "learning_rate": 1.9745845552297164e-05,
+      "loss": 0.4466,
+      "step": 1467
+    },
+    {
+      "epoch": 1.9347396176664469,
+      "grad_norm": 0.21410046623785542,
+      "learning_rate": 1.9721407624633432e-05,
+      "loss": 0.4346,
+      "step": 1468
+    },
+    {
+      "epoch": 1.936058009228741,
+      "grad_norm": 0.20621705939695184,
+      "learning_rate": 1.9696969696969697e-05,
+      "loss": 0.4555,
+      "step": 1469
+    },
+    {
+      "epoch": 1.937376400791035,
+      "grad_norm": 0.22640394596820923,
+      "learning_rate": 1.967253176930596e-05,
+      "loss": 0.4296,
+      "step": 1470
+    },
+    {
+      "epoch": 1.938694792353329,
+      "grad_norm": 0.22718325411602744,
+      "learning_rate": 1.964809384164223e-05,
+      "loss": 0.4476,
+      "step": 1471
+    },
+    {
+      "epoch": 1.940013183915623,
+      "grad_norm": 0.22963487216079229,
+      "learning_rate": 1.9623655913978494e-05,
+      "loss": 0.4373,
+      "step": 1472
+    },
+    {
+      "epoch": 1.9413315754779168,
+      "grad_norm": 0.2118205979882543,
+      "learning_rate": 1.959921798631476e-05,
+      "loss": 0.4603,
+      "step": 1473
+    },
+    {
+      "epoch": 1.9426499670402109,
+      "grad_norm": 0.20979825775255637,
+      "learning_rate": 1.9574780058651027e-05,
+      "loss": 0.4447,
+      "step": 1474
+    },
+    {
+      "epoch": 1.943968358602505,
+      "grad_norm": 0.24312425271676502,
+      "learning_rate": 1.9550342130987292e-05,
+      "loss": 0.4874,
+      "step": 1475
+    },
+    {
+      "epoch": 1.945286750164799,
+      "grad_norm": 0.22427772249329014,
+      "learning_rate": 1.9525904203323557e-05,
+      "loss": 0.4061,
+      "step": 1476
+    },
+    {
+      "epoch": 1.946605141727093,
+      "grad_norm": 0.20076378844600026,
+      "learning_rate": 1.9501466275659825e-05,
+      "loss": 0.454,
+      "step": 1477
+    },
+    {
+      "epoch": 1.947923533289387,
+      "grad_norm": 0.21183711553873094,
+      "learning_rate": 1.947702834799609e-05,
+      "loss": 0.4467,
+      "step": 1478
+    },
+    {
+      "epoch": 1.9492419248516808,
+      "grad_norm": 0.22805784685853756,
+      "learning_rate": 1.9452590420332355e-05,
+      "loss": 0.4332,
+      "step": 1479
+    },
+    {
+      "epoch": 1.9505603164139749,
+      "grad_norm": 0.21328478729698744,
+      "learning_rate": 1.9428152492668623e-05,
+      "loss": 0.4221,
+      "step": 1480
+    },
+    {
+      "epoch": 1.951878707976269,
+      "grad_norm": 0.19668837616994597,
+      "learning_rate": 1.9403714565004888e-05,
+      "loss": 0.4464,
+      "step": 1481
+    },
+    {
+      "epoch": 1.953197099538563,
+      "grad_norm": 0.21672450862134104,
+      "learning_rate": 1.9379276637341153e-05,
+      "loss": 0.4376,
+      "step": 1482
+    },
+    {
+      "epoch": 1.954515491100857,
+      "grad_norm": 0.2119311802091252,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.4482,
+      "step": 1483
+    },
+    {
+      "epoch": 1.955833882663151,
+      "grad_norm": 0.20966362163217572,
+      "learning_rate": 1.9330400782013686e-05,
+      "loss": 0.4486,
+      "step": 1484
+    },
+    {
+      "epoch": 1.9571522742254448,
+      "grad_norm": 0.23601566461457632,
+      "learning_rate": 1.930596285434995e-05,
+      "loss": 0.4307,
+      "step": 1485
+    },
+    {
+      "epoch": 1.9584706657877389,
+      "grad_norm": 0.242919410429755,
+      "learning_rate": 1.928152492668622e-05,
+      "loss": 0.4109,
+      "step": 1486
+    },
+    {
+      "epoch": 1.959789057350033,
+      "grad_norm": 0.21297920693271927,
+      "learning_rate": 1.9257086999022483e-05,
+      "loss": 0.4173,
+      "step": 1487
+    },
+    {
+      "epoch": 1.961107448912327,
+      "grad_norm": 0.23605887796773778,
+      "learning_rate": 1.9232649071358748e-05,
+      "loss": 0.4681,
+      "step": 1488
+    },
+    {
+      "epoch": 1.962425840474621,
+      "grad_norm": 0.22809605678961353,
+      "learning_rate": 1.9208211143695016e-05,
+      "loss": 0.4275,
+      "step": 1489
+    },
+    {
+      "epoch": 1.963744232036915,
+      "grad_norm": 0.2081003149704621,
+      "learning_rate": 1.918377321603128e-05,
+      "loss": 0.443,
+      "step": 1490
+    },
+    {
+      "epoch": 1.9650626235992088,
+      "grad_norm": 0.20808158134265395,
+      "learning_rate": 1.9159335288367546e-05,
+      "loss": 0.4365,
+      "step": 1491
+    },
+    {
+      "epoch": 1.9663810151615029,
+      "grad_norm": 0.2195524871931144,
+      "learning_rate": 1.9134897360703814e-05,
+      "loss": 0.4458,
+      "step": 1492
+    },
+    {
+      "epoch": 1.967699406723797,
+      "grad_norm": 0.2372060595706366,
+      "learning_rate": 1.911045943304008e-05,
+      "loss": 0.4157,
+      "step": 1493
+    },
+    {
+      "epoch": 1.969017798286091,
+      "grad_norm": 0.21416295817541797,
+      "learning_rate": 1.9086021505376344e-05,
+      "loss": 0.4096,
+      "step": 1494
+    },
+    {
+      "epoch": 1.970336189848385,
+      "grad_norm": 0.21702948327191374,
+      "learning_rate": 1.9061583577712612e-05,
+      "loss": 0.4353,
+      "step": 1495
+    },
+    {
+      "epoch": 1.971654581410679,
+      "grad_norm": 0.22949098827183775,
+      "learning_rate": 1.9037145650048877e-05,
+      "loss": 0.432,
+      "step": 1496
+    },
+    {
+      "epoch": 1.972972972972973,
+      "grad_norm": 0.21527541035095082,
+      "learning_rate": 1.901270772238514e-05,
+      "loss": 0.4566,
+      "step": 1497
+    },
+    {
+      "epoch": 1.9742913645352669,
+      "grad_norm": 0.230143811216392,
+      "learning_rate": 1.898826979472141e-05,
+      "loss": 0.4288,
+      "step": 1498
+    },
+    {
+      "epoch": 1.975609756097561,
+      "grad_norm": 0.21800755383763673,
+      "learning_rate": 1.8963831867057675e-05,
+      "loss": 0.4298,
+      "step": 1499
+    },
+    {
+      "epoch": 1.976928147659855,
+      "grad_norm": 0.216754782658239,
+      "learning_rate": 1.893939393939394e-05,
+      "loss": 0.4133,
+      "step": 1500
+    },
+    {
+      "epoch": 1.978246539222149,
+      "grad_norm": 0.2101385356997487,
+      "learning_rate": 1.8914956011730208e-05,
+      "loss": 0.4299,
+      "step": 1501
+    },
+    {
+      "epoch": 1.979564930784443,
+      "grad_norm": 0.22512919783608212,
+      "learning_rate": 1.8890518084066472e-05,
+      "loss": 0.4451,
+      "step": 1502
+    },
+    {
+      "epoch": 1.980883322346737,
+      "grad_norm": 0.21532685655891295,
+      "learning_rate": 1.8866080156402737e-05,
+      "loss": 0.4321,
+      "step": 1503
+    },
+    {
+      "epoch": 1.9822017139090309,
+      "grad_norm": 0.2098681106165605,
+      "learning_rate": 1.8841642228739005e-05,
+      "loss": 0.4186,
+      "step": 1504
+    },
+    {
+      "epoch": 1.983520105471325,
+      "grad_norm": 0.23820514848823854,
+      "learning_rate": 1.881720430107527e-05,
+      "loss": 0.4445,
+      "step": 1505
+    },
+    {
+      "epoch": 1.984838497033619,
+      "grad_norm": 0.2047311022722257,
+      "learning_rate": 1.8792766373411535e-05,
+      "loss": 0.4332,
+      "step": 1506
+    },
+    {
+      "epoch": 1.986156888595913,
+      "grad_norm": 0.22029706339432087,
+      "learning_rate": 1.8768328445747803e-05,
+      "loss": 0.4123,
+      "step": 1507
+    },
+    {
+      "epoch": 1.987475280158207,
+      "grad_norm": 0.22512253879854907,
+      "learning_rate": 1.8743890518084068e-05,
+      "loss": 0.4669,
+      "step": 1508
+    },
+    {
+      "epoch": 1.988793671720501,
+      "grad_norm": 0.22006169903318604,
+      "learning_rate": 1.8719452590420333e-05,
+      "loss": 0.4653,
+      "step": 1509
+    },
+    {
+      "epoch": 1.9901120632827949,
+      "grad_norm": 0.23383612631155792,
+      "learning_rate": 1.8695014662756598e-05,
+      "loss": 0.4611,
+      "step": 1510
+    },
+    {
+      "epoch": 1.991430454845089,
+      "grad_norm": 0.21327370842564972,
+      "learning_rate": 1.8670576735092866e-05,
+      "loss": 0.448,
+      "step": 1511
+    },
+    {
+      "epoch": 1.992748846407383,
+      "grad_norm": 0.21988146295772387,
+      "learning_rate": 1.864613880742913e-05,
+      "loss": 0.4547,
+      "step": 1512
+    },
+    {
+      "epoch": 1.994067237969677,
+      "grad_norm": 0.22480678379842745,
+      "learning_rate": 1.8621700879765395e-05,
+      "loss": 0.4408,
+      "step": 1513
+    },
+    {
+      "epoch": 1.995385629531971,
+      "grad_norm": 0.23156290570895854,
+      "learning_rate": 1.8597262952101664e-05,
+      "loss": 0.453,
+      "step": 1514
+    },
+    {
+      "epoch": 1.996704021094265,
+      "grad_norm": 0.21773730922305692,
+      "learning_rate": 1.857282502443793e-05,
+      "loss": 0.4468,
+      "step": 1515
+    },
+    {
+      "epoch": 1.9980224126565589,
+      "grad_norm": 0.22542282442310663,
+      "learning_rate": 1.8548387096774193e-05,
+      "loss": 0.4171,
+      "step": 1516
+    },
+    {
+      "epoch": 1.999340804218853,
+      "grad_norm": 0.21189153066526817,
+      "learning_rate": 1.852394916911046e-05,
+      "loss": 0.466,
+      "step": 1517
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.33068820516119557,
+      "learning_rate": 1.8499511241446726e-05,
+      "loss": 0.3964,
+      "step": 1518
+    },
+    {
+      "epoch": 2.001318391562294,
+      "grad_norm": 0.26741756144851037,
+      "learning_rate": 1.847507331378299e-05,
+      "loss": 0.3272,
+      "step": 1519
+    },
+    {
+      "epoch": 2.002636783124588,
+      "grad_norm": 0.2291393520239885,
+      "learning_rate": 1.845063538611926e-05,
+      "loss": 0.3408,
+      "step": 1520
+    },
+    {
+      "epoch": 2.0039551746868822,
+      "grad_norm": 0.2427773076436279,
+      "learning_rate": 1.8426197458455524e-05,
+      "loss": 0.3238,
+      "step": 1521
+    },
+    {
+      "epoch": 2.005273566249176,
+      "grad_norm": 0.32589000033626786,
+      "learning_rate": 1.840175953079179e-05,
+      "loss": 0.3686,
+      "step": 1522
+    },
+    {
+      "epoch": 2.00659195781147,
+      "grad_norm": 0.22965311915001102,
+      "learning_rate": 1.8377321603128057e-05,
+      "loss": 0.3706,
+      "step": 1523
+    },
+    {
+      "epoch": 2.007910349373764,
+      "grad_norm": 0.2988233162129117,
+      "learning_rate": 1.835288367546432e-05,
+      "loss": 0.365,
+      "step": 1524
+    },
+    {
+      "epoch": 2.009228740936058,
+      "grad_norm": 0.27946338326750164,
+      "learning_rate": 1.8328445747800586e-05,
+      "loss": 0.3346,
+      "step": 1525
+    },
+    {
+      "epoch": 2.010547132498352,
+      "grad_norm": 0.23981556056191242,
+      "learning_rate": 1.8304007820136855e-05,
+      "loss": 0.3509,
+      "step": 1526
+    },
+    {
+      "epoch": 2.0118655240606462,
+      "grad_norm": 0.23863701429399445,
+      "learning_rate": 1.827956989247312e-05,
+      "loss": 0.3372,
+      "step": 1527
+    },
+    {
+      "epoch": 2.01318391562294,
+      "grad_norm": 0.28676461141256815,
+      "learning_rate": 1.8255131964809384e-05,
+      "loss": 0.3538,
+      "step": 1528
+    },
+    {
+      "epoch": 2.014502307185234,
+      "grad_norm": 0.260924249647913,
+      "learning_rate": 1.8230694037145652e-05,
+      "loss": 0.3519,
+      "step": 1529
+    },
+    {
+      "epoch": 2.015820698747528,
+      "grad_norm": 0.2797078677924895,
+      "learning_rate": 1.8206256109481917e-05,
+      "loss": 0.3344,
+      "step": 1530
+    },
+    {
+      "epoch": 2.017139090309822,
+      "grad_norm": 0.28853269334672765,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.3489,
+      "step": 1531
+    },
+    {
+      "epoch": 2.018457481872116,
+      "grad_norm": 0.2558466695951493,
+      "learning_rate": 1.815738025415445e-05,
+      "loss": 0.3378,
+      "step": 1532
+    },
+    {
+      "epoch": 2.0197758734344102,
+      "grad_norm": 0.2643834575852086,
+      "learning_rate": 1.8132942326490715e-05,
+      "loss": 0.3221,
+      "step": 1533
+    },
+    {
+      "epoch": 2.021094264996704,
+      "grad_norm": 0.23596818622623297,
+      "learning_rate": 1.810850439882698e-05,
+      "loss": 0.3413,
+      "step": 1534
+    },
+    {
+      "epoch": 2.022412656558998,
+      "grad_norm": 0.2417277783013071,
+      "learning_rate": 1.8084066471163248e-05,
+      "loss": 0.328,
+      "step": 1535
+    },
+    {
+      "epoch": 2.023731048121292,
+      "grad_norm": 0.2663144372224886,
+      "learning_rate": 1.8059628543499513e-05,
+      "loss": 0.3598,
+      "step": 1536
+    },
+    {
+      "epoch": 2.025049439683586,
+      "grad_norm": 0.20760554097363196,
+      "learning_rate": 1.8035190615835778e-05,
+      "loss": 0.3273,
+      "step": 1537
+    },
+    {
+      "epoch": 2.02636783124588,
+      "grad_norm": 0.207671580060574,
+      "learning_rate": 1.8010752688172046e-05,
+      "loss": 0.3241,
+      "step": 1538
+    },
+    {
+      "epoch": 2.0276862228081742,
+      "grad_norm": 0.241394202637982,
+      "learning_rate": 1.798631476050831e-05,
+      "loss": 0.3175,
+      "step": 1539
+    },
+    {
+      "epoch": 2.029004614370468,
+      "grad_norm": 0.21749932733315322,
+      "learning_rate": 1.7961876832844575e-05,
+      "loss": 0.3422,
+      "step": 1540
+    },
+    {
+      "epoch": 2.030323005932762,
+      "grad_norm": 0.23364428958665948,
+      "learning_rate": 1.7937438905180844e-05,
+      "loss": 0.3558,
+      "step": 1541
+    },
+    {
+      "epoch": 2.031641397495056,
+      "grad_norm": 0.23173389239004274,
+      "learning_rate": 1.791300097751711e-05,
+      "loss": 0.334,
+      "step": 1542
+    },
+    {
+      "epoch": 2.03295978905735,
+      "grad_norm": 0.2316312705611543,
+      "learning_rate": 1.7888563049853373e-05,
+      "loss": 0.3382,
+      "step": 1543
+    },
+    {
+      "epoch": 2.034278180619644,
+      "grad_norm": 0.23842943745254225,
+      "learning_rate": 1.786412512218964e-05,
+      "loss": 0.3388,
+      "step": 1544
+    },
+    {
+      "epoch": 2.0355965721819382,
+      "grad_norm": 0.22014444632437336,
+      "learning_rate": 1.7839687194525906e-05,
+      "loss": 0.3383,
+      "step": 1545
+    },
+    {
+      "epoch": 2.036914963744232,
+      "grad_norm": 0.21093903240293832,
+      "learning_rate": 1.781524926686217e-05,
+      "loss": 0.3436,
+      "step": 1546
+    },
+    {
+      "epoch": 2.038233355306526,
+      "grad_norm": 0.22328218438247663,
+      "learning_rate": 1.779081133919844e-05,
+      "loss": 0.3499,
+      "step": 1547
+    },
+    {
+      "epoch": 2.03955174686882,
+      "grad_norm": 0.23432365794338406,
+      "learning_rate": 1.7766373411534704e-05,
+      "loss": 0.3476,
+      "step": 1548
+    },
+    {
+      "epoch": 2.040870138431114,
+      "grad_norm": 0.21467555000789068,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.3592,
+      "step": 1549
+    },
+    {
+      "epoch": 2.042188529993408,
+      "grad_norm": 0.20194199123712744,
+      "learning_rate": 1.7717497556207237e-05,
+      "loss": 0.3428,
+      "step": 1550
+    },
+    {
+      "epoch": 2.0435069215557022,
+      "grad_norm": 0.22090989347034615,
+      "learning_rate": 1.7693059628543502e-05,
+      "loss": 0.3401,
+      "step": 1551
+    },
+    {
+      "epoch": 2.044825313117996,
+      "grad_norm": 0.20473928426270788,
+      "learning_rate": 1.7668621700879767e-05,
+      "loss": 0.3444,
+      "step": 1552
+    },
+    {
+      "epoch": 2.04614370468029,
+      "grad_norm": 0.21233277872827783,
+      "learning_rate": 1.7644183773216035e-05,
+      "loss": 0.33,
+      "step": 1553
+    },
+    {
+      "epoch": 2.047462096242584,
+      "grad_norm": 0.27511740350749786,
+      "learning_rate": 1.76197458455523e-05,
+      "loss": 0.3808,
+      "step": 1554
+    },
+    {
+      "epoch": 2.048780487804878,
+      "grad_norm": 0.22300157532667347,
+      "learning_rate": 1.7595307917888564e-05,
+      "loss": 0.3501,
+      "step": 1555
+    },
+    {
+      "epoch": 2.050098879367172,
+      "grad_norm": 0.3321674792938774,
+      "learning_rate": 1.757086999022483e-05,
+      "loss": 0.35,
+      "step": 1556
+    },
+    {
+      "epoch": 2.0514172709294662,
+      "grad_norm": 0.22203418160088811,
+      "learning_rate": 1.7546432062561097e-05,
+      "loss": 0.3407,
+      "step": 1557
+    },
+    {
+      "epoch": 2.05273566249176,
+      "grad_norm": 0.21599255767847284,
+      "learning_rate": 1.7521994134897362e-05,
+      "loss": 0.3248,
+      "step": 1558
+    },
+    {
+      "epoch": 2.054054054054054,
+      "grad_norm": 0.20125786196268797,
+      "learning_rate": 1.7497556207233627e-05,
+      "loss": 0.3357,
+      "step": 1559
+    },
+    {
+      "epoch": 2.055372445616348,
+      "grad_norm": 0.22801850432040444,
+      "learning_rate": 1.7473118279569895e-05,
+      "loss": 0.3604,
+      "step": 1560
+    },
+    {
+      "epoch": 2.056690837178642,
+      "grad_norm": 0.21686577030051837,
+      "learning_rate": 1.744868035190616e-05,
+      "loss": 0.3418,
+      "step": 1561
+    },
+    {
+      "epoch": 2.058009228740936,
+      "grad_norm": 0.22156283645510197,
+      "learning_rate": 1.7424242424242425e-05,
+      "loss": 0.3474,
+      "step": 1562
+    },
+    {
+      "epoch": 2.0593276203032302,
+      "grad_norm": 0.2111976471531977,
+      "learning_rate": 1.7399804496578693e-05,
+      "loss": 0.3295,
+      "step": 1563
+    },
+    {
+      "epoch": 2.060646011865524,
+      "grad_norm": 0.20572110835938678,
+      "learning_rate": 1.7375366568914958e-05,
+      "loss": 0.3305,
+      "step": 1564
+    },
+    {
+      "epoch": 2.061964403427818,
+      "grad_norm": 0.2262503953758642,
+      "learning_rate": 1.7350928641251223e-05,
+      "loss": 0.3494,
+      "step": 1565
+    },
+    {
+      "epoch": 2.063282794990112,
+      "grad_norm": 0.18813099492730037,
+      "learning_rate": 1.732649071358749e-05,
+      "loss": 0.3118,
+      "step": 1566
+    },
+    {
+      "epoch": 2.064601186552406,
+      "grad_norm": 0.22426415919093753,
+      "learning_rate": 1.7302052785923756e-05,
+      "loss": 0.346,
+      "step": 1567
+    },
+    {
+      "epoch": 2.0659195781147,
+      "grad_norm": 0.21490083991443093,
+      "learning_rate": 1.727761485826002e-05,
+      "loss": 0.3339,
+      "step": 1568
+    },
+    {
+      "epoch": 2.0672379696769942,
+      "grad_norm": 0.20632854953633037,
+      "learning_rate": 1.725317693059629e-05,
+      "loss": 0.3379,
+      "step": 1569
+    },
+    {
+      "epoch": 2.068556361239288,
+      "grad_norm": 0.20285010802958353,
+      "learning_rate": 1.7228739002932553e-05,
+      "loss": 0.3278,
+      "step": 1570
+    },
+    {
+      "epoch": 2.069874752801582,
+      "grad_norm": 0.21081546550529287,
+      "learning_rate": 1.7204301075268818e-05,
+      "loss": 0.3364,
+      "step": 1571
+    },
+    {
+      "epoch": 2.071193144363876,
+      "grad_norm": 0.20642019946268803,
+      "learning_rate": 1.7179863147605083e-05,
+      "loss": 0.3359,
+      "step": 1572
+    },
+    {
+      "epoch": 2.07251153592617,
+      "grad_norm": 0.19187211969221923,
+      "learning_rate": 1.7155425219941348e-05,
+      "loss": 0.3345,
+      "step": 1573
+    },
+    {
+      "epoch": 2.073829927488464,
+      "grad_norm": 0.19191648039814707,
+      "learning_rate": 1.7130987292277616e-05,
+      "loss": 0.3212,
+      "step": 1574
+    },
+    {
+      "epoch": 2.0751483190507582,
+      "grad_norm": 0.20193539086574935,
+      "learning_rate": 1.710654936461388e-05,
+      "loss": 0.3273,
+      "step": 1575
+    },
+    {
+      "epoch": 2.076466710613052,
+      "grad_norm": 0.1924647828917233,
+      "learning_rate": 1.7082111436950146e-05,
+      "loss": 0.3201,
+      "step": 1576
+    },
+    {
+      "epoch": 2.077785102175346,
+      "grad_norm": 0.21037100528111702,
+      "learning_rate": 1.7057673509286414e-05,
+      "loss": 0.3387,
+      "step": 1577
+    },
+    {
+      "epoch": 2.07910349373764,
+      "grad_norm": 0.1951166937705083,
+      "learning_rate": 1.703323558162268e-05,
+      "loss": 0.3348,
+      "step": 1578
+    },
+    {
+      "epoch": 2.080421885299934,
+      "grad_norm": 0.19969186425421434,
+      "learning_rate": 1.7008797653958943e-05,
+      "loss": 0.321,
+      "step": 1579
+    },
+    {
+      "epoch": 2.081740276862228,
+      "grad_norm": 0.18769484350037763,
+      "learning_rate": 1.6984359726295208e-05,
+      "loss": 0.3195,
+      "step": 1580
+    },
+    {
+      "epoch": 2.0830586684245223,
+      "grad_norm": 0.21641505436703395,
+      "learning_rate": 1.6959921798631476e-05,
+      "loss": 0.3456,
+      "step": 1581
+    },
+    {
+      "epoch": 2.084377059986816,
+      "grad_norm": 0.22836162203053878,
+      "learning_rate": 1.693548387096774e-05,
+      "loss": 0.3514,
+      "step": 1582
+    },
+    {
+      "epoch": 2.08569545154911,
+      "grad_norm": 0.20404225914810695,
+      "learning_rate": 1.6911045943304006e-05,
+      "loss": 0.3502,
+      "step": 1583
+    },
+    {
+      "epoch": 2.087013843111404,
+      "grad_norm": 0.22123417168258377,
+      "learning_rate": 1.6886608015640274e-05,
+      "loss": 0.3305,
+      "step": 1584
+    },
+    {
+      "epoch": 2.088332234673698,
+      "grad_norm": 0.2099938637082831,
+      "learning_rate": 1.686217008797654e-05,
+      "loss": 0.3332,
+      "step": 1585
+    },
+    {
+      "epoch": 2.089650626235992,
+      "grad_norm": 0.23769371892901311,
+      "learning_rate": 1.6837732160312804e-05,
+      "loss": 0.3437,
+      "step": 1586
+    },
+    {
+      "epoch": 2.0909690177982863,
+      "grad_norm": 0.212770752324405,
+      "learning_rate": 1.6813294232649072e-05,
+      "loss": 0.3151,
+      "step": 1587
+    },
+    {
+      "epoch": 2.09228740936058,
+      "grad_norm": 0.2226714081835481,
+      "learning_rate": 1.6788856304985337e-05,
+      "loss": 0.3673,
+      "step": 1588
+    },
+    {
+      "epoch": 2.093605800922874,
+      "grad_norm": 0.21309869758247063,
+      "learning_rate": 1.67644183773216e-05,
+      "loss": 0.3363,
+      "step": 1589
+    },
+    {
+      "epoch": 2.094924192485168,
+      "grad_norm": 0.22127819615216315,
+      "learning_rate": 1.673998044965787e-05,
+      "loss": 0.3639,
+      "step": 1590
+    },
+    {
+      "epoch": 2.096242584047462,
+      "grad_norm": 0.22393625446449328,
+      "learning_rate": 1.6715542521994134e-05,
+      "loss": 0.3396,
+      "step": 1591
+    },
+    {
+      "epoch": 2.097560975609756,
+      "grad_norm": 0.19381257024359874,
+      "learning_rate": 1.66911045943304e-05,
+      "loss": 0.3332,
+      "step": 1592
+    },
+    {
+      "epoch": 2.0988793671720503,
+      "grad_norm": 0.22491848743943751,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.3591,
+      "step": 1593
+    },
+    {
+      "epoch": 2.100197758734344,
+      "grad_norm": 0.21070360078624428,
+      "learning_rate": 1.6642228739002932e-05,
+      "loss": 0.3402,
+      "step": 1594
+    },
+    {
+      "epoch": 2.101516150296638,
+      "grad_norm": 0.20583282090934538,
+      "learning_rate": 1.6617790811339197e-05,
+      "loss": 0.32,
+      "step": 1595
+    },
+    {
+      "epoch": 2.102834541858932,
+      "grad_norm": 0.21917037541809928,
+      "learning_rate": 1.6593352883675465e-05,
+      "loss": 0.3431,
+      "step": 1596
+    },
+    {
+      "epoch": 2.104152933421226,
+      "grad_norm": 0.20532050538704952,
+      "learning_rate": 1.656891495601173e-05,
+      "loss": 0.3316,
+      "step": 1597
+    },
+    {
+      "epoch": 2.10547132498352,
+      "grad_norm": 0.19692340691074428,
+      "learning_rate": 1.6544477028347995e-05,
+      "loss": 0.3314,
+      "step": 1598
+    },
+    {
+      "epoch": 2.1067897165458143,
+      "grad_norm": 0.20054744300296357,
+      "learning_rate": 1.6520039100684263e-05,
+      "loss": 0.3265,
+      "step": 1599
+    },
+    {
+      "epoch": 2.108108108108108,
+      "grad_norm": 0.2153994006917322,
+      "learning_rate": 1.6495601173020528e-05,
+      "loss": 0.3266,
+      "step": 1600
+    },
+    {
+      "epoch": 2.109426499670402,
+      "grad_norm": 0.21140769599199216,
+      "learning_rate": 1.6471163245356793e-05,
+      "loss": 0.3487,
+      "step": 1601
+    },
+    {
+      "epoch": 2.110744891232696,
+      "grad_norm": 0.2001351259216344,
+      "learning_rate": 1.644672531769306e-05,
+      "loss": 0.3441,
+      "step": 1602
+    },
+    {
+      "epoch": 2.11206328279499,
+      "grad_norm": 0.21151484940797988,
+      "learning_rate": 1.6422287390029326e-05,
+      "loss": 0.3331,
+      "step": 1603
+    },
+    {
+      "epoch": 2.113381674357284,
+      "grad_norm": 0.19712714154905991,
+      "learning_rate": 1.639784946236559e-05,
+      "loss": 0.3347,
+      "step": 1604
+    },
+    {
+      "epoch": 2.1147000659195783,
+      "grad_norm": 0.2135614323068103,
+      "learning_rate": 1.637341153470186e-05,
+      "loss": 0.3409,
+      "step": 1605
+    },
+    {
+      "epoch": 2.1160184574818723,
+      "grad_norm": 0.20760396916854149,
+      "learning_rate": 1.6348973607038123e-05,
+      "loss": 0.3445,
+      "step": 1606
+    },
+    {
+      "epoch": 2.117336849044166,
+      "grad_norm": 0.21904002533133163,
+      "learning_rate": 1.6324535679374388e-05,
+      "loss": 0.3315,
+      "step": 1607
+    },
+    {
+      "epoch": 2.11865524060646,
+      "grad_norm": 0.21406841678365124,
+      "learning_rate": 1.6300097751710656e-05,
+      "loss": 0.3473,
+      "step": 1608
+    },
+    {
+      "epoch": 2.119973632168754,
+      "grad_norm": 0.2071159162963191,
+      "learning_rate": 1.627565982404692e-05,
+      "loss": 0.353,
+      "step": 1609
+    },
+    {
+      "epoch": 2.121292023731048,
+      "grad_norm": 0.21263747140652609,
+      "learning_rate": 1.6251221896383186e-05,
+      "loss": 0.3521,
+      "step": 1610
+    },
+    {
+      "epoch": 2.1226104152933423,
+      "grad_norm": 0.22477917520494242,
+      "learning_rate": 1.6226783968719454e-05,
+      "loss": 0.351,
+      "step": 1611
+    },
+    {
+      "epoch": 2.123928806855636,
+      "grad_norm": 0.20050531730028265,
+      "learning_rate": 1.620234604105572e-05,
+      "loss": 0.3192,
+      "step": 1612
+    },
+    {
+      "epoch": 2.12524719841793,
+      "grad_norm": 0.20714616547758516,
+      "learning_rate": 1.6177908113391984e-05,
+      "loss": 0.3118,
+      "step": 1613
+    },
+    {
+      "epoch": 2.126565589980224,
+      "grad_norm": 0.21263615749470666,
+      "learning_rate": 1.6153470185728252e-05,
+      "loss": 0.3379,
+      "step": 1614
+    },
+    {
+      "epoch": 2.127883981542518,
+      "grad_norm": 0.21502397150483116,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.3595,
+      "step": 1615
+    },
+    {
+      "epoch": 2.129202373104812,
+      "grad_norm": 0.21166748067532268,
+      "learning_rate": 1.610459433040078e-05,
+      "loss": 0.3334,
+      "step": 1616
+    },
+    {
+      "epoch": 2.1305207646671063,
+      "grad_norm": 0.20495512874557437,
+      "learning_rate": 1.608015640273705e-05,
+      "loss": 0.3252,
+      "step": 1617
+    },
+    {
+      "epoch": 2.1318391562294003,
+      "grad_norm": 0.2211451508390733,
+      "learning_rate": 1.6055718475073315e-05,
+      "loss": 0.333,
+      "step": 1618
+    },
+    {
+      "epoch": 2.133157547791694,
+      "grad_norm": 0.20200594347589806,
+      "learning_rate": 1.603128054740958e-05,
+      "loss": 0.3328,
+      "step": 1619
+    },
+    {
+      "epoch": 2.134475939353988,
+      "grad_norm": 0.22452203415434516,
+      "learning_rate": 1.6006842619745848e-05,
+      "loss": 0.3522,
+      "step": 1620
+    },
+    {
+      "epoch": 2.135794330916282,
+      "grad_norm": 0.22221212072019925,
+      "learning_rate": 1.5982404692082112e-05,
+      "loss": 0.3397,
+      "step": 1621
+    },
+    {
+      "epoch": 2.137112722478576,
+      "grad_norm": 0.24233741729970867,
+      "learning_rate": 1.5957966764418377e-05,
+      "loss": 0.371,
+      "step": 1622
+    },
+    {
+      "epoch": 2.1384311140408703,
+      "grad_norm": 0.21824332281362213,
+      "learning_rate": 1.5933528836754645e-05,
+      "loss": 0.3413,
+      "step": 1623
+    },
+    {
+      "epoch": 2.1397495056031643,
+      "grad_norm": 0.21752465216573502,
+      "learning_rate": 1.590909090909091e-05,
+      "loss": 0.3435,
+      "step": 1624
+    },
+    {
+      "epoch": 2.141067897165458,
+      "grad_norm": 0.21644572164872308,
+      "learning_rate": 1.5884652981427175e-05,
+      "loss": 0.3451,
+      "step": 1625
+    },
+    {
+      "epoch": 2.142386288727752,
+      "grad_norm": 0.283041196812053,
+      "learning_rate": 1.586021505376344e-05,
+      "loss": 0.3427,
+      "step": 1626
+    },
+    {
+      "epoch": 2.143704680290046,
+      "grad_norm": 0.21739806257496513,
+      "learning_rate": 1.5835777126099708e-05,
+      "loss": 0.3317,
+      "step": 1627
+    },
+    {
+      "epoch": 2.14502307185234,
+      "grad_norm": 0.21295082266864496,
+      "learning_rate": 1.5811339198435973e-05,
+      "loss": 0.3345,
+      "step": 1628
+    },
+    {
+      "epoch": 2.1463414634146343,
+      "grad_norm": 0.214760147288853,
+      "learning_rate": 1.5786901270772238e-05,
+      "loss": 0.3351,
+      "step": 1629
+    },
+    {
+      "epoch": 2.1476598549769284,
+      "grad_norm": 0.23225847911179873,
+      "learning_rate": 1.5762463343108506e-05,
+      "loss": 0.3369,
+      "step": 1630
+    },
+    {
+      "epoch": 2.148978246539222,
+      "grad_norm": 0.2076230789955957,
+      "learning_rate": 1.573802541544477e-05,
+      "loss": 0.3196,
+      "step": 1631
+    },
+    {
+      "epoch": 2.150296638101516,
+      "grad_norm": 0.22910397589023618,
+      "learning_rate": 1.5713587487781035e-05,
+      "loss": 0.3491,
+      "step": 1632
+    },
+    {
+      "epoch": 2.15161502966381,
+      "grad_norm": 0.2345266890640023,
+      "learning_rate": 1.5689149560117304e-05,
+      "loss": 0.3392,
+      "step": 1633
+    },
+    {
+      "epoch": 2.152933421226104,
+      "grad_norm": 0.2112010913804191,
+      "learning_rate": 1.566471163245357e-05,
+      "loss": 0.351,
+      "step": 1634
+    },
+    {
+      "epoch": 2.1542518127883983,
+      "grad_norm": 0.2073478064365827,
+      "learning_rate": 1.5640273704789833e-05,
+      "loss": 0.3244,
+      "step": 1635
+    },
+    {
+      "epoch": 2.1555702043506924,
+      "grad_norm": 0.2295766821375366,
+      "learning_rate": 1.56158357771261e-05,
+      "loss": 0.336,
+      "step": 1636
+    },
+    {
+      "epoch": 2.156888595912986,
+      "grad_norm": 0.22497247817361962,
+      "learning_rate": 1.5591397849462366e-05,
+      "loss": 0.3604,
+      "step": 1637
+    },
+    {
+      "epoch": 2.15820698747528,
+      "grad_norm": 0.20860267748050423,
+      "learning_rate": 1.556695992179863e-05,
+      "loss": 0.3497,
+      "step": 1638
+    },
+    {
+      "epoch": 2.159525379037574,
+      "grad_norm": 0.21216712534827575,
+      "learning_rate": 1.55425219941349e-05,
+      "loss": 0.3301,
+      "step": 1639
+    },
+    {
+      "epoch": 2.160843770599868,
+      "grad_norm": 0.2017133530965416,
+      "learning_rate": 1.5518084066471164e-05,
+      "loss": 0.3116,
+      "step": 1640
+    },
+    {
+      "epoch": 2.1621621621621623,
+      "grad_norm": 0.21113535143230308,
+      "learning_rate": 1.549364613880743e-05,
+      "loss": 0.3552,
+      "step": 1641
+    },
+    {
+      "epoch": 2.1634805537244564,
+      "grad_norm": 0.20516386472949807,
+      "learning_rate": 1.5469208211143697e-05,
+      "loss": 0.3428,
+      "step": 1642
+    },
+    {
+      "epoch": 2.16479894528675,
+      "grad_norm": 0.21506878461173973,
+      "learning_rate": 1.5444770283479962e-05,
+      "loss": 0.3588,
+      "step": 1643
+    },
+    {
+      "epoch": 2.166117336849044,
+      "grad_norm": 0.20909771684925144,
+      "learning_rate": 1.5420332355816226e-05,
+      "loss": 0.3508,
+      "step": 1644
+    },
+    {
+      "epoch": 2.167435728411338,
+      "grad_norm": 0.20998044907558178,
+      "learning_rate": 1.5395894428152495e-05,
+      "loss": 0.3244,
+      "step": 1645
+    },
+    {
+      "epoch": 2.168754119973632,
+      "grad_norm": 0.21257827844619856,
+      "learning_rate": 1.537145650048876e-05,
+      "loss": 0.3505,
+      "step": 1646
+    },
+    {
+      "epoch": 2.1700725115359263,
+      "grad_norm": 0.19861876424956496,
+      "learning_rate": 1.5347018572825024e-05,
+      "loss": 0.321,
+      "step": 1647
+    },
+    {
+      "epoch": 2.1713909030982204,
+      "grad_norm": 0.21339470472419347,
+      "learning_rate": 1.5322580645161292e-05,
+      "loss": 0.3297,
+      "step": 1648
+    },
+    {
+      "epoch": 2.172709294660514,
+      "grad_norm": 0.2128801435501254,
+      "learning_rate": 1.5298142717497557e-05,
+      "loss": 0.3561,
+      "step": 1649
+    },
+    {
+      "epoch": 2.174027686222808,
+      "grad_norm": 0.20015025342181988,
+      "learning_rate": 1.5273704789833822e-05,
+      "loss": 0.3178,
+      "step": 1650
+    },
+    {
+      "epoch": 2.175346077785102,
+      "grad_norm": 0.20604530754617761,
+      "learning_rate": 1.5249266862170089e-05,
+      "loss": 0.3185,
+      "step": 1651
+    },
+    {
+      "epoch": 2.176664469347396,
+      "grad_norm": 0.22996089438445513,
+      "learning_rate": 1.5224828934506355e-05,
+      "loss": 0.3469,
+      "step": 1652
+    },
+    {
+      "epoch": 2.1779828609096903,
+      "grad_norm": 0.21993819837887282,
+      "learning_rate": 1.5200391006842622e-05,
+      "loss": 0.337,
+      "step": 1653
+    },
+    {
+      "epoch": 2.1793012524719844,
+      "grad_norm": 0.21875385076262294,
+      "learning_rate": 1.5175953079178886e-05,
+      "loss": 0.3541,
+      "step": 1654
+    },
+    {
+      "epoch": 2.180619644034278,
+      "grad_norm": 0.20796021045494945,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.3527,
+      "step": 1655
+    },
+    {
+      "epoch": 2.181938035596572,
+      "grad_norm": 0.2191742713237783,
+      "learning_rate": 1.512707722385142e-05,
+      "loss": 0.3464,
+      "step": 1656
+    },
+    {
+      "epoch": 2.183256427158866,
+      "grad_norm": 0.22161889221531966,
+      "learning_rate": 1.5102639296187684e-05,
+      "loss": 0.3594,
+      "step": 1657
+    },
+    {
+      "epoch": 2.18457481872116,
+      "grad_norm": 0.2129016895619246,
+      "learning_rate": 1.507820136852395e-05,
+      "loss": 0.3416,
+      "step": 1658
+    },
+    {
+      "epoch": 2.1858932102834543,
+      "grad_norm": 0.21266720826735983,
+      "learning_rate": 1.5053763440860215e-05,
+      "loss": 0.3438,
+      "step": 1659
+    },
+    {
+      "epoch": 2.1872116018457484,
+      "grad_norm": 0.2249284129643314,
+      "learning_rate": 1.5029325513196482e-05,
+      "loss": 0.3435,
+      "step": 1660
+    },
+    {
+      "epoch": 2.188529993408042,
+      "grad_norm": 0.21156079607066397,
+      "learning_rate": 1.5004887585532748e-05,
+      "loss": 0.3565,
+      "step": 1661
+    },
+    {
+      "epoch": 2.189848384970336,
+      "grad_norm": 0.21041412202896906,
+      "learning_rate": 1.4980449657869013e-05,
+      "loss": 0.3392,
+      "step": 1662
+    },
+    {
+      "epoch": 2.19116677653263,
+      "grad_norm": 0.21220364447921142,
+      "learning_rate": 1.495601173020528e-05,
+      "loss": 0.332,
+      "step": 1663
+    },
+    {
+      "epoch": 2.192485168094924,
+      "grad_norm": 0.21608547563970246,
+      "learning_rate": 1.4931573802541546e-05,
+      "loss": 0.3489,
+      "step": 1664
+    },
+    {
+      "epoch": 2.1938035596572183,
+      "grad_norm": 0.22838932643197485,
+      "learning_rate": 1.4907135874877811e-05,
+      "loss": 0.3506,
+      "step": 1665
+    },
+    {
+      "epoch": 2.1951219512195124,
+      "grad_norm": 0.22523693266803446,
+      "learning_rate": 1.4882697947214078e-05,
+      "loss": 0.362,
+      "step": 1666
+    },
+    {
+      "epoch": 2.196440342781806,
+      "grad_norm": 0.20926431015747549,
+      "learning_rate": 1.4858260019550344e-05,
+      "loss": 0.3209,
+      "step": 1667
+    },
+    {
+      "epoch": 2.1977587343441,
+      "grad_norm": 0.2265244504800887,
+      "learning_rate": 1.4833822091886609e-05,
+      "loss": 0.3583,
+      "step": 1668
+    },
+    {
+      "epoch": 2.199077125906394,
+      "grad_norm": 0.21511957275211438,
+      "learning_rate": 1.4809384164222875e-05,
+      "loss": 0.3589,
+      "step": 1669
+    },
+    {
+      "epoch": 2.2003955174686882,
+      "grad_norm": 0.21425511348649134,
+      "learning_rate": 1.4784946236559142e-05,
+      "loss": 0.33,
+      "step": 1670
+    },
+    {
+      "epoch": 2.2017139090309823,
+      "grad_norm": 0.20739375259616588,
+      "learning_rate": 1.4760508308895407e-05,
+      "loss": 0.3273,
+      "step": 1671
+    },
+    {
+      "epoch": 2.2030323005932764,
+      "grad_norm": 0.21472801013794351,
+      "learning_rate": 1.4736070381231673e-05,
+      "loss": 0.331,
+      "step": 1672
+    },
+    {
+      "epoch": 2.20435069215557,
+      "grad_norm": 0.20717490268666386,
+      "learning_rate": 1.471163245356794e-05,
+      "loss": 0.3517,
+      "step": 1673
+    },
+    {
+      "epoch": 2.205669083717864,
+      "grad_norm": 0.2213172223834101,
+      "learning_rate": 1.4687194525904204e-05,
+      "loss": 0.3495,
+      "step": 1674
+    },
+    {
+      "epoch": 2.206987475280158,
+      "grad_norm": 0.2146774420779988,
+      "learning_rate": 1.4662756598240471e-05,
+      "loss": 0.3485,
+      "step": 1675
+    },
+    {
+      "epoch": 2.2083058668424522,
+      "grad_norm": 0.20685697747167162,
+      "learning_rate": 1.4638318670576737e-05,
+      "loss": 0.3451,
+      "step": 1676
+    },
+    {
+      "epoch": 2.2096242584047463,
+      "grad_norm": 0.2049021326485404,
+      "learning_rate": 1.4613880742913002e-05,
+      "loss": 0.3292,
+      "step": 1677
+    },
+    {
+      "epoch": 2.2109426499670404,
+      "grad_norm": 0.21964603327331608,
+      "learning_rate": 1.4589442815249269e-05,
+      "loss": 0.3456,
+      "step": 1678
+    },
+    {
+      "epoch": 2.212261041529334,
+      "grad_norm": 0.20703935888843578,
+      "learning_rate": 1.4565004887585535e-05,
+      "loss": 0.3547,
+      "step": 1679
+    },
+    {
+      "epoch": 2.213579433091628,
+      "grad_norm": 0.20307447173542123,
+      "learning_rate": 1.45405669599218e-05,
+      "loss": 0.3341,
+      "step": 1680
+    },
+    {
+      "epoch": 2.214897824653922,
+      "grad_norm": 0.2133529798579718,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.3621,
+      "step": 1681
+    },
+    {
+      "epoch": 2.2162162162162162,
+      "grad_norm": 0.20315201896191615,
+      "learning_rate": 1.4491691104594331e-05,
+      "loss": 0.3304,
+      "step": 1682
+    },
+    {
+      "epoch": 2.2175346077785103,
+      "grad_norm": 0.20243057791344077,
+      "learning_rate": 1.4467253176930598e-05,
+      "loss": 0.3471,
+      "step": 1683
+    },
+    {
+      "epoch": 2.2188529993408044,
+      "grad_norm": 0.20439450394189443,
+      "learning_rate": 1.4442815249266864e-05,
+      "loss": 0.3486,
+      "step": 1684
+    },
+    {
+      "epoch": 2.220171390903098,
+      "grad_norm": 0.20529580383989204,
+      "learning_rate": 1.4418377321603129e-05,
+      "loss": 0.3281,
+      "step": 1685
+    },
+    {
+      "epoch": 2.221489782465392,
+      "grad_norm": 0.21024989755789758,
+      "learning_rate": 1.4393939393939396e-05,
+      "loss": 0.3365,
+      "step": 1686
+    },
+    {
+      "epoch": 2.222808174027686,
+      "grad_norm": 0.21565706638606572,
+      "learning_rate": 1.4369501466275662e-05,
+      "loss": 0.3566,
+      "step": 1687
+    },
+    {
+      "epoch": 2.2241265655899802,
+      "grad_norm": 0.19644777136768937,
+      "learning_rate": 1.4345063538611927e-05,
+      "loss": 0.3248,
+      "step": 1688
+    },
+    {
+      "epoch": 2.2254449571522743,
+      "grad_norm": 0.20167680835317145,
+      "learning_rate": 1.4320625610948193e-05,
+      "loss": 0.3368,
+      "step": 1689
+    },
+    {
+      "epoch": 2.2267633487145684,
+      "grad_norm": 0.206279253871997,
+      "learning_rate": 1.429618768328446e-05,
+      "loss": 0.3267,
+      "step": 1690
+    },
+    {
+      "epoch": 2.2280817402768625,
+      "grad_norm": 0.21268697961228766,
+      "learning_rate": 1.4271749755620725e-05,
+      "loss": 0.342,
+      "step": 1691
+    },
+    {
+      "epoch": 2.229400131839156,
+      "grad_norm": 0.20453208356975233,
+      "learning_rate": 1.4247311827956991e-05,
+      "loss": 0.3321,
+      "step": 1692
+    },
+    {
+      "epoch": 2.23071852340145,
+      "grad_norm": 0.20114238379346494,
+      "learning_rate": 1.4222873900293258e-05,
+      "loss": 0.3263,
+      "step": 1693
+    },
+    {
+      "epoch": 2.2320369149637442,
+      "grad_norm": 0.2066017852458238,
+      "learning_rate": 1.4198435972629522e-05,
+      "loss": 0.3408,
+      "step": 1694
+    },
+    {
+      "epoch": 2.2333553065260383,
+      "grad_norm": 0.2190379402777761,
+      "learning_rate": 1.4173998044965789e-05,
+      "loss": 0.3382,
+      "step": 1695
+    },
+    {
+      "epoch": 2.2346736980883324,
+      "grad_norm": 0.20315334320215997,
+      "learning_rate": 1.4149560117302055e-05,
+      "loss": 0.3328,
+      "step": 1696
+    },
+    {
+      "epoch": 2.235992089650626,
+      "grad_norm": 0.21417568696623773,
+      "learning_rate": 1.412512218963832e-05,
+      "loss": 0.3521,
+      "step": 1697
+    },
+    {
+      "epoch": 2.23731048121292,
+      "grad_norm": 0.19507797474695562,
+      "learning_rate": 1.4100684261974587e-05,
+      "loss": 0.3342,
+      "step": 1698
+    },
+    {
+      "epoch": 2.238628872775214,
+      "grad_norm": 0.21209226278415202,
+      "learning_rate": 1.4076246334310853e-05,
+      "loss": 0.349,
+      "step": 1699
+    },
+    {
+      "epoch": 2.2399472643375082,
+      "grad_norm": 0.20487865109959824,
+      "learning_rate": 1.4051808406647116e-05,
+      "loss": 0.3359,
+      "step": 1700
+    },
+    {
+      "epoch": 2.2412656558998023,
+      "grad_norm": 0.20354218731394577,
+      "learning_rate": 1.4027370478983381e-05,
+      "loss": 0.362,
+      "step": 1701
+    },
+    {
+      "epoch": 2.2425840474620964,
+      "grad_norm": 0.20881801269297076,
+      "learning_rate": 1.4002932551319648e-05,
+      "loss": 0.3433,
+      "step": 1702
+    },
+    {
+      "epoch": 2.2439024390243905,
+      "grad_norm": 0.1937124692972319,
+      "learning_rate": 1.3978494623655914e-05,
+      "loss": 0.3327,
+      "step": 1703
+    },
+    {
+      "epoch": 2.245220830586684,
+      "grad_norm": 0.2028759547535333,
+      "learning_rate": 1.3954056695992179e-05,
+      "loss": 0.3491,
+      "step": 1704
+    },
+    {
+      "epoch": 2.246539222148978,
+      "grad_norm": 0.2060474945271629,
+      "learning_rate": 1.3929618768328445e-05,
+      "loss": 0.3484,
+      "step": 1705
+    },
+    {
+      "epoch": 2.2478576137112722,
+      "grad_norm": 0.20419137871476467,
+      "learning_rate": 1.3905180840664712e-05,
+      "loss": 0.3513,
+      "step": 1706
+    },
+    {
+      "epoch": 2.2491760052735663,
+      "grad_norm": 0.19572249222208107,
+      "learning_rate": 1.3880742913000977e-05,
+      "loss": 0.3254,
+      "step": 1707
+    },
+    {
+      "epoch": 2.2504943968358604,
+      "grad_norm": 0.20793278030271248,
+      "learning_rate": 1.3856304985337243e-05,
+      "loss": 0.3546,
+      "step": 1708
+    },
+    {
+      "epoch": 2.251812788398154,
+      "grad_norm": 0.19861420276839953,
+      "learning_rate": 1.3831867057673508e-05,
+      "loss": 0.3135,
+      "step": 1709
+    },
+    {
+      "epoch": 2.253131179960448,
+      "grad_norm": 0.2141781011558213,
+      "learning_rate": 1.3807429130009774e-05,
+      "loss": 0.3424,
+      "step": 1710
+    },
+    {
+      "epoch": 2.254449571522742,
+      "grad_norm": 0.20923274663807012,
+      "learning_rate": 1.3782991202346041e-05,
+      "loss": 0.3279,
+      "step": 1711
+    },
+    {
+      "epoch": 2.2557679630850362,
+      "grad_norm": 0.206412598723254,
+      "learning_rate": 1.3758553274682306e-05,
+      "loss": 0.341,
+      "step": 1712
+    },
+    {
+      "epoch": 2.2570863546473303,
+      "grad_norm": 0.21030073016878567,
+      "learning_rate": 1.3734115347018572e-05,
+      "loss": 0.3399,
+      "step": 1713
+    },
+    {
+      "epoch": 2.2584047462096244,
+      "grad_norm": 0.21399382507400633,
+      "learning_rate": 1.3709677419354839e-05,
+      "loss": 0.3307,
+      "step": 1714
+    },
+    {
+      "epoch": 2.2597231377719185,
+      "grad_norm": 0.19905050256510406,
+      "learning_rate": 1.3685239491691104e-05,
+      "loss": 0.3388,
+      "step": 1715
+    },
+    {
+      "epoch": 2.261041529334212,
+      "grad_norm": 0.21450465957274062,
+      "learning_rate": 1.366080156402737e-05,
+      "loss": 0.3295,
+      "step": 1716
+    },
+    {
+      "epoch": 2.262359920896506,
+      "grad_norm": 0.1992749681636228,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.3532,
+      "step": 1717
+    },
+    {
+      "epoch": 2.2636783124588002,
+      "grad_norm": 0.21096681288627595,
+      "learning_rate": 1.3611925708699901e-05,
+      "loss": 0.3546,
+      "step": 1718
+    },
+    {
+      "epoch": 2.2649967040210943,
+      "grad_norm": 0.2087394851694881,
+      "learning_rate": 1.3587487781036168e-05,
+      "loss": 0.3413,
+      "step": 1719
+    },
+    {
+      "epoch": 2.2663150955833884,
+      "grad_norm": 0.2083497030448489,
+      "learning_rate": 1.3563049853372434e-05,
+      "loss": 0.3481,
+      "step": 1720
+    },
+    {
+      "epoch": 2.267633487145682,
+      "grad_norm": 0.2040134359803411,
+      "learning_rate": 1.3538611925708699e-05,
+      "loss": 0.3334,
+      "step": 1721
+    },
+    {
+      "epoch": 2.268951878707976,
+      "grad_norm": 0.2251252442428985,
+      "learning_rate": 1.3514173998044966e-05,
+      "loss": 0.356,
+      "step": 1722
+    },
+    {
+      "epoch": 2.27027027027027,
+      "grad_norm": 0.21498303741773667,
+      "learning_rate": 1.3489736070381232e-05,
+      "loss": 0.3317,
+      "step": 1723
+    },
+    {
+      "epoch": 2.2715886618325642,
+      "grad_norm": 0.20011271568090516,
+      "learning_rate": 1.3465298142717497e-05,
+      "loss": 0.3398,
+      "step": 1724
+    },
+    {
+      "epoch": 2.2729070533948583,
+      "grad_norm": 0.20971775549004593,
+      "learning_rate": 1.3440860215053763e-05,
+      "loss": 0.337,
+      "step": 1725
+    },
+    {
+      "epoch": 2.2742254449571524,
+      "grad_norm": 0.20034056489170637,
+      "learning_rate": 1.341642228739003e-05,
+      "loss": 0.3234,
+      "step": 1726
+    },
+    {
+      "epoch": 2.2755438365194465,
+      "grad_norm": 0.21657469616865016,
+      "learning_rate": 1.3391984359726295e-05,
+      "loss": 0.3383,
+      "step": 1727
+    },
+    {
+      "epoch": 2.27686222808174,
+      "grad_norm": 0.20428345931450306,
+      "learning_rate": 1.3367546432062561e-05,
+      "loss": 0.3305,
+      "step": 1728
+    },
+    {
+      "epoch": 2.278180619644034,
+      "grad_norm": 0.20056327644969432,
+      "learning_rate": 1.3343108504398828e-05,
+      "loss": 0.3566,
+      "step": 1729
+    },
+    {
+      "epoch": 2.2794990112063283,
+      "grad_norm": 0.21225451829274686,
+      "learning_rate": 1.3318670576735093e-05,
+      "loss": 0.3454,
+      "step": 1730
+    },
+    {
+      "epoch": 2.2808174027686223,
+      "grad_norm": 0.22420332604088758,
+      "learning_rate": 1.3294232649071359e-05,
+      "loss": 0.3509,
+      "step": 1731
+    },
+    {
+      "epoch": 2.2821357943309164,
+      "grad_norm": 0.19975574759348566,
+      "learning_rate": 1.3269794721407624e-05,
+      "loss": 0.3382,
+      "step": 1732
+    },
+    {
+      "epoch": 2.2834541858932105,
+      "grad_norm": 0.19571220260495392,
+      "learning_rate": 1.324535679374389e-05,
+      "loss": 0.3473,
+      "step": 1733
+    },
+    {
+      "epoch": 2.284772577455504,
+      "grad_norm": 0.20461995802568392,
+      "learning_rate": 1.3220918866080157e-05,
+      "loss": 0.3427,
+      "step": 1734
+    },
+    {
+      "epoch": 2.286090969017798,
+      "grad_norm": 0.2112581148141922,
+      "learning_rate": 1.3196480938416422e-05,
+      "loss": 0.346,
+      "step": 1735
+    },
+    {
+      "epoch": 2.2874093605800923,
+      "grad_norm": 0.20911305471189612,
+      "learning_rate": 1.3172043010752688e-05,
+      "loss": 0.3342,
+      "step": 1736
+    },
+    {
+      "epoch": 2.2887277521423863,
+      "grad_norm": 0.20140525313421775,
+      "learning_rate": 1.3147605083088955e-05,
+      "loss": 0.3188,
+      "step": 1737
+    },
+    {
+      "epoch": 2.2900461437046804,
+      "grad_norm": 0.20977206599901155,
+      "learning_rate": 1.312316715542522e-05,
+      "loss": 0.3319,
+      "step": 1738
+    },
+    {
+      "epoch": 2.2913645352669745,
+      "grad_norm": 0.21184058677172052,
+      "learning_rate": 1.3098729227761486e-05,
+      "loss": 0.3513,
+      "step": 1739
+    },
+    {
+      "epoch": 2.292682926829268,
+      "grad_norm": 0.20742845215088096,
+      "learning_rate": 1.3074291300097752e-05,
+      "loss": 0.3336,
+      "step": 1740
+    },
+    {
+      "epoch": 2.294001318391562,
+      "grad_norm": 0.21419207350419364,
+      "learning_rate": 1.3049853372434017e-05,
+      "loss": 0.3633,
+      "step": 1741
+    },
+    {
+      "epoch": 2.2953197099538563,
+      "grad_norm": 0.19226251672055594,
+      "learning_rate": 1.3025415444770284e-05,
+      "loss": 0.3261,
+      "step": 1742
+    },
+    {
+      "epoch": 2.2966381015161503,
+      "grad_norm": 0.2119172031075403,
+      "learning_rate": 1.300097751710655e-05,
+      "loss": 0.3447,
+      "step": 1743
+    },
+    {
+      "epoch": 2.2979564930784444,
+      "grad_norm": 0.20895860391890378,
+      "learning_rate": 1.2976539589442815e-05,
+      "loss": 0.3253,
+      "step": 1744
+    },
+    {
+      "epoch": 2.2992748846407385,
+      "grad_norm": 0.21335536678207814,
+      "learning_rate": 1.2952101661779081e-05,
+      "loss": 0.3331,
+      "step": 1745
+    },
+    {
+      "epoch": 2.300593276203032,
+      "grad_norm": 0.20052679455127692,
+      "learning_rate": 1.2927663734115348e-05,
+      "loss": 0.3321,
+      "step": 1746
+    },
+    {
+      "epoch": 2.301911667765326,
+      "grad_norm": 0.19988225027155432,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.3369,
+      "step": 1747
+    },
+    {
+      "epoch": 2.3032300593276203,
+      "grad_norm": 0.20781574716152487,
+      "learning_rate": 1.287878787878788e-05,
+      "loss": 0.3662,
+      "step": 1748
+    },
+    {
+      "epoch": 2.3045484508899143,
+      "grad_norm": 0.1982323336994524,
+      "learning_rate": 1.2854349951124146e-05,
+      "loss": 0.3145,
+      "step": 1749
+    },
+    {
+      "epoch": 2.3058668424522084,
+      "grad_norm": 0.24406328091500387,
+      "learning_rate": 1.282991202346041e-05,
+      "loss": 0.338,
+      "step": 1750
+    },
+    {
+      "epoch": 2.3071852340145025,
+      "grad_norm": 0.20193551331588375,
+      "learning_rate": 1.2805474095796677e-05,
+      "loss": 0.3419,
+      "step": 1751
+    },
+    {
+      "epoch": 2.308503625576796,
+      "grad_norm": 0.19973746717841245,
+      "learning_rate": 1.2781036168132944e-05,
+      "loss": 0.3322,
+      "step": 1752
+    },
+    {
+      "epoch": 2.30982201713909,
+      "grad_norm": 0.2003092216912326,
+      "learning_rate": 1.2756598240469208e-05,
+      "loss": 0.3217,
+      "step": 1753
+    },
+    {
+      "epoch": 2.3111404087013843,
+      "grad_norm": 0.22689889123968077,
+      "learning_rate": 1.2732160312805475e-05,
+      "loss": 0.3427,
+      "step": 1754
+    },
+    {
+      "epoch": 2.3124588002636783,
+      "grad_norm": 0.21871140981916795,
+      "learning_rate": 1.270772238514174e-05,
+      "loss": 0.3356,
+      "step": 1755
+    },
+    {
+      "epoch": 2.3137771918259724,
+      "grad_norm": 0.19920869798335794,
+      "learning_rate": 1.2683284457478006e-05,
+      "loss": 0.3421,
+      "step": 1756
+    },
+    {
+      "epoch": 2.3150955833882665,
+      "grad_norm": 0.20549675595342548,
+      "learning_rate": 1.2658846529814273e-05,
+      "loss": 0.3295,
+      "step": 1757
+    },
+    {
+      "epoch": 2.31641397495056,
+      "grad_norm": 0.2143019254957071,
+      "learning_rate": 1.2634408602150537e-05,
+      "loss": 0.3421,
+      "step": 1758
+    },
+    {
+      "epoch": 2.317732366512854,
+      "grad_norm": 0.2130568136640732,
+      "learning_rate": 1.2609970674486804e-05,
+      "loss": 0.3086,
+      "step": 1759
+    },
+    {
+      "epoch": 2.3190507580751483,
+      "grad_norm": 0.22822655450237608,
+      "learning_rate": 1.258553274682307e-05,
+      "loss": 0.3325,
+      "step": 1760
+    },
+    {
+      "epoch": 2.3203691496374423,
+      "grad_norm": 0.2041317371579539,
+      "learning_rate": 1.2561094819159335e-05,
+      "loss": 0.3473,
+      "step": 1761
+    },
+    {
+      "epoch": 2.3216875411997364,
+      "grad_norm": 0.20351381770785218,
+      "learning_rate": 1.2536656891495602e-05,
+      "loss": 0.3412,
+      "step": 1762
+    },
+    {
+      "epoch": 2.3230059327620305,
+      "grad_norm": 0.22066984637839612,
+      "learning_rate": 1.2512218963831868e-05,
+      "loss": 0.3248,
+      "step": 1763
+    },
+    {
+      "epoch": 2.3243243243243246,
+      "grad_norm": 0.22549959785537105,
+      "learning_rate": 1.2487781036168133e-05,
+      "loss": 0.3329,
+      "step": 1764
+    },
+    {
+      "epoch": 2.325642715886618,
+      "grad_norm": 0.20723878508062174,
+      "learning_rate": 1.24633431085044e-05,
+      "loss": 0.3354,
+      "step": 1765
+    },
+    {
+      "epoch": 2.3269611074489123,
+      "grad_norm": 0.1995185809776796,
+      "learning_rate": 1.2438905180840666e-05,
+      "loss": 0.3333,
+      "step": 1766
+    },
+    {
+      "epoch": 2.3282794990112063,
+      "grad_norm": 0.21655713743748561,
+      "learning_rate": 1.241446725317693e-05,
+      "loss": 0.3387,
+      "step": 1767
+    },
+    {
+      "epoch": 2.3295978905735004,
+      "grad_norm": 0.22940690473396555,
+      "learning_rate": 1.2390029325513197e-05,
+      "loss": 0.3282,
+      "step": 1768
+    },
+    {
+      "epoch": 2.3309162821357945,
+      "grad_norm": 0.2108028367874812,
+      "learning_rate": 1.2365591397849464e-05,
+      "loss": 0.3481,
+      "step": 1769
+    },
+    {
+      "epoch": 2.332234673698088,
+      "grad_norm": 0.20904822797020764,
+      "learning_rate": 1.2341153470185729e-05,
+      "loss": 0.3148,
+      "step": 1770
+    },
+    {
+      "epoch": 2.333553065260382,
+      "grad_norm": 0.1940393855584944,
+      "learning_rate": 1.2316715542521995e-05,
+      "loss": 0.3154,
+      "step": 1771
+    },
+    {
+      "epoch": 2.3348714568226763,
+      "grad_norm": 0.23929278848086188,
+      "learning_rate": 1.2292277614858262e-05,
+      "loss": 0.3465,
+      "step": 1772
+    },
+    {
+      "epoch": 2.3361898483849703,
+      "grad_norm": 0.21131890485248134,
+      "learning_rate": 1.2267839687194526e-05,
+      "loss": 0.3511,
+      "step": 1773
+    },
+    {
+      "epoch": 2.3375082399472644,
+      "grad_norm": 0.19973865566627355,
+      "learning_rate": 1.2243401759530793e-05,
+      "loss": 0.3392,
+      "step": 1774
+    },
+    {
+      "epoch": 2.3388266315095585,
+      "grad_norm": 0.20359872773900337,
+      "learning_rate": 1.2218963831867058e-05,
+      "loss": 0.3334,
+      "step": 1775
+    },
+    {
+      "epoch": 2.3401450230718526,
+      "grad_norm": 0.20202339460263055,
+      "learning_rate": 1.2194525904203324e-05,
+      "loss": 0.3348,
+      "step": 1776
+    },
+    {
+      "epoch": 2.341463414634146,
+      "grad_norm": 0.20385972706305908,
+      "learning_rate": 1.217008797653959e-05,
+      "loss": 0.3455,
+      "step": 1777
+    },
+    {
+      "epoch": 2.3427818061964403,
+      "grad_norm": 0.21705660635334917,
+      "learning_rate": 1.2145650048875855e-05,
+      "loss": 0.3455,
+      "step": 1778
+    },
+    {
+      "epoch": 2.3441001977587343,
+      "grad_norm": 0.1957139629713652,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.3295,
+      "step": 1779
+    },
+    {
+      "epoch": 2.3454185893210284,
+      "grad_norm": 0.20144137790984198,
+      "learning_rate": 1.2096774193548388e-05,
+      "loss": 0.338,
+      "step": 1780
+    },
+    {
+      "epoch": 2.3467369808833225,
+      "grad_norm": 0.2109156021258922,
+      "learning_rate": 1.2072336265884653e-05,
+      "loss": 0.3106,
+      "step": 1781
+    },
+    {
+      "epoch": 2.348055372445616,
+      "grad_norm": 0.2078624849426534,
+      "learning_rate": 1.204789833822092e-05,
+      "loss": 0.3493,
+      "step": 1782
+    },
+    {
+      "epoch": 2.34937376400791,
+      "grad_norm": 0.20663002150479617,
+      "learning_rate": 1.2023460410557186e-05,
+      "loss": 0.3574,
+      "step": 1783
+    },
+    {
+      "epoch": 2.3506921555702043,
+      "grad_norm": 0.2070979047155682,
+      "learning_rate": 1.1999022482893451e-05,
+      "loss": 0.3439,
+      "step": 1784
+    },
+    {
+      "epoch": 2.3520105471324984,
+      "grad_norm": 0.2034120294634911,
+      "learning_rate": 1.1974584555229718e-05,
+      "loss": 0.3399,
+      "step": 1785
+    },
+    {
+      "epoch": 2.3533289386947924,
+      "grad_norm": 0.20054112327917367,
+      "learning_rate": 1.1950146627565984e-05,
+      "loss": 0.3247,
+      "step": 1786
+    },
+    {
+      "epoch": 2.3546473302570865,
+      "grad_norm": 0.19892672394489122,
+      "learning_rate": 1.1925708699902249e-05,
+      "loss": 0.332,
+      "step": 1787
+    },
+    {
+      "epoch": 2.3559657218193806,
+      "grad_norm": 0.22347051307606078,
+      "learning_rate": 1.1901270772238515e-05,
+      "loss": 0.3406,
+      "step": 1788
+    },
+    {
+      "epoch": 2.357284113381674,
+      "grad_norm": 0.20742046931599514,
+      "learning_rate": 1.1876832844574782e-05,
+      "loss": 0.343,
+      "step": 1789
+    },
+    {
+      "epoch": 2.3586025049439683,
+      "grad_norm": 0.19510072366970097,
+      "learning_rate": 1.1852394916911047e-05,
+      "loss": 0.3317,
+      "step": 1790
+    },
+    {
+      "epoch": 2.3599208965062624,
+      "grad_norm": 0.20687603538139934,
+      "learning_rate": 1.1827956989247313e-05,
+      "loss": 0.3421,
+      "step": 1791
+    },
+    {
+      "epoch": 2.3612392880685564,
+      "grad_norm": 0.22291637028488068,
+      "learning_rate": 1.180351906158358e-05,
+      "loss": 0.3446,
+      "step": 1792
+    },
+    {
+      "epoch": 2.3625576796308505,
+      "grad_norm": 0.21101174205036988,
+      "learning_rate": 1.1779081133919844e-05,
+      "loss": 0.3507,
+      "step": 1793
+    },
+    {
+      "epoch": 2.363876071193144,
+      "grad_norm": 0.22371448912052133,
+      "learning_rate": 1.1754643206256111e-05,
+      "loss": 0.3243,
+      "step": 1794
+    },
+    {
+      "epoch": 2.365194462755438,
+      "grad_norm": 0.21172622696665075,
+      "learning_rate": 1.1730205278592377e-05,
+      "loss": 0.3523,
+      "step": 1795
+    },
+    {
+      "epoch": 2.3665128543177323,
+      "grad_norm": 0.20550015869306612,
+      "learning_rate": 1.170576735092864e-05,
+      "loss": 0.3404,
+      "step": 1796
+    },
+    {
+      "epoch": 2.3678312458800264,
+      "grad_norm": 0.19852165957932813,
+      "learning_rate": 1.1681329423264907e-05,
+      "loss": 0.3273,
+      "step": 1797
+    },
+    {
+      "epoch": 2.3691496374423204,
+      "grad_norm": 0.20949300946677088,
+      "learning_rate": 1.1656891495601173e-05,
+      "loss": 0.3482,
+      "step": 1798
+    },
+    {
+      "epoch": 2.3704680290046145,
+      "grad_norm": 0.220797109307791,
+      "learning_rate": 1.1632453567937438e-05,
+      "loss": 0.3533,
+      "step": 1799
+    },
+    {
+      "epoch": 2.3717864205669086,
+      "grad_norm": 0.22483201249058712,
+      "learning_rate": 1.1608015640273705e-05,
+      "loss": 0.3524,
+      "step": 1800
+    },
+    {
+      "epoch": 2.373104812129202,
+      "grad_norm": 0.20409424143880087,
+      "learning_rate": 1.1583577712609971e-05,
+      "loss": 0.3358,
+      "step": 1801
+    },
+    {
+      "epoch": 2.3744232036914963,
+      "grad_norm": 0.19962613987864913,
+      "learning_rate": 1.1559139784946236e-05,
+      "loss": 0.3445,
+      "step": 1802
+    },
+    {
+      "epoch": 2.3757415952537904,
+      "grad_norm": 0.19919512702903153,
+      "learning_rate": 1.1534701857282503e-05,
+      "loss": 0.3348,
+      "step": 1803
+    },
+    {
+      "epoch": 2.3770599868160844,
+      "grad_norm": 0.21852321022982774,
+      "learning_rate": 1.1510263929618769e-05,
+      "loss": 0.3359,
+      "step": 1804
+    },
+    {
+      "epoch": 2.3783783783783785,
+      "grad_norm": 0.2126122758479001,
+      "learning_rate": 1.1485826001955034e-05,
+      "loss": 0.3409,
+      "step": 1805
+    },
+    {
+      "epoch": 2.379696769940672,
+      "grad_norm": 0.21246858896732151,
+      "learning_rate": 1.14613880742913e-05,
+      "loss": 0.3502,
+      "step": 1806
+    },
+    {
+      "epoch": 2.381015161502966,
+      "grad_norm": 0.1988393353477164,
+      "learning_rate": 1.1436950146627567e-05,
+      "loss": 0.3369,
+      "step": 1807
+    },
+    {
+      "epoch": 2.3823335530652603,
+      "grad_norm": 0.20209023885018204,
+      "learning_rate": 1.1412512218963832e-05,
+      "loss": 0.3256,
+      "step": 1808
+    },
+    {
+      "epoch": 2.3836519446275544,
+      "grad_norm": 0.2032918552296078,
+      "learning_rate": 1.1388074291300098e-05,
+      "loss": 0.3311,
+      "step": 1809
+    },
+    {
+      "epoch": 2.3849703361898484,
+      "grad_norm": 0.19985119840434012,
+      "learning_rate": 1.1363636363636365e-05,
+      "loss": 0.3346,
+      "step": 1810
+    },
+    {
+      "epoch": 2.3862887277521425,
+      "grad_norm": 0.20403058656692058,
+      "learning_rate": 1.133919843597263e-05,
+      "loss": 0.3551,
+      "step": 1811
+    },
+    {
+      "epoch": 2.3876071193144366,
+      "grad_norm": 0.19729532743459258,
+      "learning_rate": 1.1314760508308896e-05,
+      "loss": 0.34,
+      "step": 1812
+    },
+    {
+      "epoch": 2.38892551087673,
+      "grad_norm": 0.20985203685113657,
+      "learning_rate": 1.129032258064516e-05,
+      "loss": 0.3519,
+      "step": 1813
+    },
+    {
+      "epoch": 2.3902439024390243,
+      "grad_norm": 0.2001603426332322,
+      "learning_rate": 1.1265884652981427e-05,
+      "loss": 0.3297,
+      "step": 1814
+    },
+    {
+      "epoch": 2.3915622940013184,
+      "grad_norm": 0.21407876316714988,
+      "learning_rate": 1.1241446725317694e-05,
+      "loss": 0.3455,
+      "step": 1815
+    },
+    {
+      "epoch": 2.3928806855636124,
+      "grad_norm": 0.19993394161114,
+      "learning_rate": 1.1217008797653959e-05,
+      "loss": 0.3363,
+      "step": 1816
+    },
+    {
+      "epoch": 2.3941990771259065,
+      "grad_norm": 0.20285053268327816,
+      "learning_rate": 1.1192570869990225e-05,
+      "loss": 0.348,
+      "step": 1817
+    },
+    {
+      "epoch": 2.3955174686882,
+      "grad_norm": 0.21381973900148576,
+      "learning_rate": 1.1168132942326492e-05,
+      "loss": 0.3465,
+      "step": 1818
+    },
+    {
+      "epoch": 2.396835860250494,
+      "grad_norm": 0.19287722132086518,
+      "learning_rate": 1.1143695014662756e-05,
+      "loss": 0.3202,
+      "step": 1819
+    },
+    {
+      "epoch": 2.3981542518127883,
+      "grad_norm": 0.19679281231063395,
+      "learning_rate": 1.1119257086999023e-05,
+      "loss": 0.3342,
+      "step": 1820
+    },
+    {
+      "epoch": 2.3994726433750824,
+      "grad_norm": 0.19885758031947923,
+      "learning_rate": 1.109481915933529e-05,
+      "loss": 0.3472,
+      "step": 1821
+    },
+    {
+      "epoch": 2.4007910349373764,
+      "grad_norm": 0.20502906331457119,
+      "learning_rate": 1.1070381231671554e-05,
+      "loss": 0.3434,
+      "step": 1822
+    },
+    {
+      "epoch": 2.4021094264996705,
+      "grad_norm": 0.20443246573072918,
+      "learning_rate": 1.104594330400782e-05,
+      "loss": 0.3355,
+      "step": 1823
+    },
+    {
+      "epoch": 2.4034278180619646,
+      "grad_norm": 0.21531659184774476,
+      "learning_rate": 1.1021505376344087e-05,
+      "loss": 0.3446,
+      "step": 1824
+    },
+    {
+      "epoch": 2.4047462096242582,
+      "grad_norm": 0.21429754792639025,
+      "learning_rate": 1.0997067448680352e-05,
+      "loss": 0.362,
+      "step": 1825
+    },
+    {
+      "epoch": 2.4060646011865523,
+      "grad_norm": 0.19649136286781108,
+      "learning_rate": 1.0972629521016618e-05,
+      "loss": 0.3442,
+      "step": 1826
+    },
+    {
+      "epoch": 2.4073829927488464,
+      "grad_norm": 0.19758739298444156,
+      "learning_rate": 1.0948191593352885e-05,
+      "loss": 0.3267,
+      "step": 1827
+    },
+    {
+      "epoch": 2.4087013843111404,
+      "grad_norm": 0.21263291648809513,
+      "learning_rate": 1.092375366568915e-05,
+      "loss": 0.3528,
+      "step": 1828
+    },
+    {
+      "epoch": 2.4100197758734345,
+      "grad_norm": 0.2103401642040274,
+      "learning_rate": 1.0899315738025416e-05,
+      "loss": 0.3373,
+      "step": 1829
+    },
+    {
+      "epoch": 2.4113381674357286,
+      "grad_norm": 0.19207927443934522,
+      "learning_rate": 1.0874877810361683e-05,
+      "loss": 0.3245,
+      "step": 1830
+    },
+    {
+      "epoch": 2.4126565589980222,
+      "grad_norm": 0.20649070592284213,
+      "learning_rate": 1.0850439882697947e-05,
+      "loss": 0.3371,
+      "step": 1831
+    },
+    {
+      "epoch": 2.4139749505603163,
+      "grad_norm": 0.20571319251791165,
+      "learning_rate": 1.0826001955034214e-05,
+      "loss": 0.3455,
+      "step": 1832
+    },
+    {
+      "epoch": 2.4152933421226104,
+      "grad_norm": 0.20319712912439378,
+      "learning_rate": 1.080156402737048e-05,
+      "loss": 0.332,
+      "step": 1833
+    },
+    {
+      "epoch": 2.4166117336849045,
+      "grad_norm": 0.20558267994309726,
+      "learning_rate": 1.0777126099706745e-05,
+      "loss": 0.3406,
+      "step": 1834
+    },
+    {
+      "epoch": 2.4179301252471985,
+      "grad_norm": 0.20843472364180787,
+      "learning_rate": 1.0752688172043012e-05,
+      "loss": 0.3322,
+      "step": 1835
+    },
+    {
+      "epoch": 2.4192485168094926,
+      "grad_norm": 0.20228872203877912,
+      "learning_rate": 1.0728250244379277e-05,
+      "loss": 0.3281,
+      "step": 1836
+    },
+    {
+      "epoch": 2.4205669083717862,
+      "grad_norm": 0.19909269996873166,
+      "learning_rate": 1.0703812316715543e-05,
+      "loss": 0.3465,
+      "step": 1837
+    },
+    {
+      "epoch": 2.4218852999340803,
+      "grad_norm": 0.20762726021927963,
+      "learning_rate": 1.067937438905181e-05,
+      "loss": 0.3341,
+      "step": 1838
+    },
+    {
+      "epoch": 2.4232036914963744,
+      "grad_norm": 0.2171839490947941,
+      "learning_rate": 1.0654936461388074e-05,
+      "loss": 0.3382,
+      "step": 1839
+    },
+    {
+      "epoch": 2.4245220830586685,
+      "grad_norm": 0.2100559507160311,
+      "learning_rate": 1.063049853372434e-05,
+      "loss": 0.3417,
+      "step": 1840
+    },
+    {
+      "epoch": 2.4258404746209625,
+      "grad_norm": 0.20908502285799938,
+      "learning_rate": 1.0606060606060607e-05,
+      "loss": 0.3292,
+      "step": 1841
+    },
+    {
+      "epoch": 2.4271588661832566,
+      "grad_norm": 0.24487298573939842,
+      "learning_rate": 1.0581622678396872e-05,
+      "loss": 0.3265,
+      "step": 1842
+    },
+    {
+      "epoch": 2.4284772577455502,
+      "grad_norm": 0.19148212831370584,
+      "learning_rate": 1.0557184750733139e-05,
+      "loss": 0.3182,
+      "step": 1843
+    },
+    {
+      "epoch": 2.4297956493078443,
+      "grad_norm": 0.1926753995916553,
+      "learning_rate": 1.0532746823069405e-05,
+      "loss": 0.3231,
+      "step": 1844
+    },
+    {
+      "epoch": 2.4311140408701384,
+      "grad_norm": 0.20172273073541447,
+      "learning_rate": 1.050830889540567e-05,
+      "loss": 0.3313,
+      "step": 1845
+    },
+    {
+      "epoch": 2.4324324324324325,
+      "grad_norm": 0.2133444129785141,
+      "learning_rate": 1.0483870967741936e-05,
+      "loss": 0.349,
+      "step": 1846
+    },
+    {
+      "epoch": 2.4337508239947265,
+      "grad_norm": 0.1947005023872194,
+      "learning_rate": 1.0459433040078203e-05,
+      "loss": 0.3377,
+      "step": 1847
+    },
+    {
+      "epoch": 2.4350692155570206,
+      "grad_norm": 0.21435775415426336,
+      "learning_rate": 1.0434995112414468e-05,
+      "loss": 0.3354,
+      "step": 1848
+    },
+    {
+      "epoch": 2.4363876071193147,
+      "grad_norm": 0.21120426082675378,
+      "learning_rate": 1.0410557184750734e-05,
+      "loss": 0.3464,
+      "step": 1849
+    },
+    {
+      "epoch": 2.4377059986816083,
+      "grad_norm": 0.19258691717377502,
+      "learning_rate": 1.0386119257087e-05,
+      "loss": 0.3305,
+      "step": 1850
+    },
+    {
+      "epoch": 2.4390243902439024,
+      "grad_norm": 0.1949914355570352,
+      "learning_rate": 1.0361681329423266e-05,
+      "loss": 0.3259,
+      "step": 1851
+    },
+    {
+      "epoch": 2.4403427818061965,
+      "grad_norm": 0.19601764380687742,
+      "learning_rate": 1.0337243401759532e-05,
+      "loss": 0.3248,
+      "step": 1852
+    },
+    {
+      "epoch": 2.4416611733684905,
+      "grad_norm": 0.19501667615178156,
+      "learning_rate": 1.0312805474095798e-05,
+      "loss": 0.3374,
+      "step": 1853
+    },
+    {
+      "epoch": 2.4429795649307846,
+      "grad_norm": 0.19155883720006905,
+      "learning_rate": 1.0288367546432063e-05,
+      "loss": 0.3136,
+      "step": 1854
+    },
+    {
+      "epoch": 2.4442979564930782,
+      "grad_norm": 0.20567830477864812,
+      "learning_rate": 1.026392961876833e-05,
+      "loss": 0.3453,
+      "step": 1855
+    },
+    {
+      "epoch": 2.4456163480553723,
+      "grad_norm": 0.20199818154374763,
+      "learning_rate": 1.0239491691104596e-05,
+      "loss": 0.3296,
+      "step": 1856
+    },
+    {
+      "epoch": 2.4469347396176664,
+      "grad_norm": 0.22341733326603788,
+      "learning_rate": 1.0215053763440861e-05,
+      "loss": 0.3489,
+      "step": 1857
+    },
+    {
+      "epoch": 2.4482531311799605,
+      "grad_norm": 0.18827753689950466,
+      "learning_rate": 1.0190615835777128e-05,
+      "loss": 0.3468,
+      "step": 1858
+    },
+    {
+      "epoch": 2.4495715227422545,
+      "grad_norm": 0.20555891729362233,
+      "learning_rate": 1.0166177908113392e-05,
+      "loss": 0.3414,
+      "step": 1859
+    },
+    {
+      "epoch": 2.4508899143045486,
+      "grad_norm": 0.20644662585277576,
+      "learning_rate": 1.0141739980449657e-05,
+      "loss": 0.3659,
+      "step": 1860
+    },
+    {
+      "epoch": 2.4522083058668427,
+      "grad_norm": 0.20611883987882682,
+      "learning_rate": 1.0117302052785924e-05,
+      "loss": 0.3514,
+      "step": 1861
+    },
+    {
+      "epoch": 2.4535266974291363,
+      "grad_norm": 0.20022746417033002,
+      "learning_rate": 1.009286412512219e-05,
+      "loss": 0.3421,
+      "step": 1862
+    },
+    {
+      "epoch": 2.4548450889914304,
+      "grad_norm": 0.21069394789927798,
+      "learning_rate": 1.0068426197458455e-05,
+      "loss": 0.3443,
+      "step": 1863
+    },
+    {
+      "epoch": 2.4561634805537245,
+      "grad_norm": 0.20099751129539206,
+      "learning_rate": 1.0043988269794721e-05,
+      "loss": 0.3419,
+      "step": 1864
+    },
+    {
+      "epoch": 2.4574818721160185,
+      "grad_norm": 0.1969797435937931,
+      "learning_rate": 1.0019550342130988e-05,
+      "loss": 0.3365,
+      "step": 1865
+    },
+    {
+      "epoch": 2.4588002636783126,
+      "grad_norm": 0.22292373236189686,
+      "learning_rate": 9.995112414467253e-06,
+      "loss": 0.3404,
+      "step": 1866
+    },
+    {
+      "epoch": 2.4601186552406062,
+      "grad_norm": 0.2186449317555616,
+      "learning_rate": 9.97067448680352e-06,
+      "loss": 0.3335,
+      "step": 1867
+    },
+    {
+      "epoch": 2.4614370468029003,
+      "grad_norm": 0.19586685301430345,
+      "learning_rate": 9.946236559139786e-06,
+      "loss": 0.3304,
+      "step": 1868
+    },
+    {
+      "epoch": 2.4627554383651944,
+      "grad_norm": 0.20358893968985348,
+      "learning_rate": 9.92179863147605e-06,
+      "loss": 0.3344,
+      "step": 1869
+    },
+    {
+      "epoch": 2.4640738299274885,
+      "grad_norm": 0.2132482732845485,
+      "learning_rate": 9.897360703812317e-06,
+      "loss": 0.3611,
+      "step": 1870
+    },
+    {
+      "epoch": 2.4653922214897825,
+      "grad_norm": 0.20694813045880717,
+      "learning_rate": 9.872922776148582e-06,
+      "loss": 0.3252,
+      "step": 1871
+    },
+    {
+      "epoch": 2.4667106130520766,
+      "grad_norm": 0.19701083592305912,
+      "learning_rate": 9.848484848484848e-06,
+      "loss": 0.3408,
+      "step": 1872
+    },
+    {
+      "epoch": 2.4680290046143707,
+      "grad_norm": 0.19382897809681027,
+      "learning_rate": 9.824046920821115e-06,
+      "loss": 0.3284,
+      "step": 1873
+    },
+    {
+      "epoch": 2.4693473961766643,
+      "grad_norm": 0.20346229560027032,
+      "learning_rate": 9.79960899315738e-06,
+      "loss": 0.3348,
+      "step": 1874
+    },
+    {
+      "epoch": 2.4706657877389584,
+      "grad_norm": 0.20289730350661503,
+      "learning_rate": 9.775171065493646e-06,
+      "loss": 0.331,
+      "step": 1875
+    },
+    {
+      "epoch": 2.4719841793012525,
+      "grad_norm": 0.1899916212094589,
+      "learning_rate": 9.750733137829913e-06,
+      "loss": 0.3144,
+      "step": 1876
+    },
+    {
+      "epoch": 2.4733025708635465,
+      "grad_norm": 0.21550199256213734,
+      "learning_rate": 9.726295210166177e-06,
+      "loss": 0.3488,
+      "step": 1877
+    },
+    {
+      "epoch": 2.4746209624258406,
+      "grad_norm": 0.19043266839368833,
+      "learning_rate": 9.701857282502444e-06,
+      "loss": 0.3331,
+      "step": 1878
+    },
+    {
+      "epoch": 2.4759393539881343,
+      "grad_norm": 0.19773568145621453,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.3403,
+      "step": 1879
+    },
+    {
+      "epoch": 2.4772577455504283,
+      "grad_norm": 0.2012094686003171,
+      "learning_rate": 9.652981427174975e-06,
+      "loss": 0.3456,
+      "step": 1880
+    },
+    {
+      "epoch": 2.4785761371127224,
+      "grad_norm": 0.19662739104929602,
+      "learning_rate": 9.628543499511242e-06,
+      "loss": 0.3311,
+      "step": 1881
+    },
+    {
+      "epoch": 2.4798945286750165,
+      "grad_norm": 0.19181597355535301,
+      "learning_rate": 9.604105571847508e-06,
+      "loss": 0.3328,
+      "step": 1882
+    },
+    {
+      "epoch": 2.4812129202373105,
+      "grad_norm": 0.20230919858050925,
+      "learning_rate": 9.579667644183773e-06,
+      "loss": 0.3333,
+      "step": 1883
+    },
+    {
+      "epoch": 2.4825313117996046,
+      "grad_norm": 0.20097512127823092,
+      "learning_rate": 9.55522971652004e-06,
+      "loss": 0.3363,
+      "step": 1884
+    },
+    {
+      "epoch": 2.4838497033618987,
+      "grad_norm": 0.19399621681073403,
+      "learning_rate": 9.530791788856306e-06,
+      "loss": 0.3398,
+      "step": 1885
+    },
+    {
+      "epoch": 2.4851680949241923,
+      "grad_norm": 0.20109507894004533,
+      "learning_rate": 9.50635386119257e-06,
+      "loss": 0.3463,
+      "step": 1886
+    },
+    {
+      "epoch": 2.4864864864864864,
+      "grad_norm": 0.18889041141536034,
+      "learning_rate": 9.481915933528837e-06,
+      "loss": 0.3276,
+      "step": 1887
+    },
+    {
+      "epoch": 2.4878048780487805,
+      "grad_norm": 0.19985333078176526,
+      "learning_rate": 9.457478005865104e-06,
+      "loss": 0.3312,
+      "step": 1888
+    },
+    {
+      "epoch": 2.4891232696110746,
+      "grad_norm": 0.20537327592079485,
+      "learning_rate": 9.433040078201369e-06,
+      "loss": 0.3297,
+      "step": 1889
+    },
+    {
+      "epoch": 2.4904416611733686,
+      "grad_norm": 0.20001564892488127,
+      "learning_rate": 9.408602150537635e-06,
+      "loss": 0.3367,
+      "step": 1890
+    },
+    {
+      "epoch": 2.4917600527356623,
+      "grad_norm": 0.1995292177615265,
+      "learning_rate": 9.384164222873902e-06,
+      "loss": 0.3276,
+      "step": 1891
+    },
+    {
+      "epoch": 2.4930784442979563,
+      "grad_norm": 0.21346942711010877,
+      "learning_rate": 9.359726295210166e-06,
+      "loss": 0.3319,
+      "step": 1892
+    },
+    {
+      "epoch": 2.4943968358602504,
+      "grad_norm": 0.209552975025353,
+      "learning_rate": 9.335288367546433e-06,
+      "loss": 0.3415,
+      "step": 1893
+    },
+    {
+      "epoch": 2.4957152274225445,
+      "grad_norm": 0.19601881886242692,
+      "learning_rate": 9.310850439882698e-06,
+      "loss": 0.3237,
+      "step": 1894
+    },
+    {
+      "epoch": 2.4970336189848386,
+      "grad_norm": 0.1976131192614562,
+      "learning_rate": 9.286412512218964e-06,
+      "loss": 0.3341,
+      "step": 1895
+    },
+    {
+      "epoch": 2.4983520105471326,
+      "grad_norm": 0.21117275264279317,
+      "learning_rate": 9.26197458455523e-06,
+      "loss": 0.3562,
+      "step": 1896
+    },
+    {
+      "epoch": 2.4996704021094267,
+      "grad_norm": 0.2004099911220413,
+      "learning_rate": 9.237536656891495e-06,
+      "loss": 0.323,
+      "step": 1897
+    },
+    {
+      "epoch": 2.5009887936717208,
+      "grad_norm": 0.20495850721595624,
+      "learning_rate": 9.213098729227762e-06,
+      "loss": 0.3383,
+      "step": 1898
+    },
+    {
+      "epoch": 2.5023071852340144,
+      "grad_norm": 0.19557074700739227,
+      "learning_rate": 9.188660801564028e-06,
+      "loss": 0.3454,
+      "step": 1899
+    },
+    {
+      "epoch": 2.5036255767963085,
+      "grad_norm": 0.1946302944514409,
+      "learning_rate": 9.164222873900293e-06,
+      "loss": 0.336,
+      "step": 1900
+    },
+    {
+      "epoch": 2.5049439683586026,
+      "grad_norm": 0.20709575620831122,
+      "learning_rate": 9.13978494623656e-06,
+      "loss": 0.3457,
+      "step": 1901
+    },
+    {
+      "epoch": 2.5062623599208966,
+      "grad_norm": 0.202292080309933,
+      "learning_rate": 9.115347018572826e-06,
+      "loss": 0.3396,
+      "step": 1902
+    },
+    {
+      "epoch": 2.5075807514831903,
+      "grad_norm": 0.2047631443598355,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.338,
+      "step": 1903
+    },
+    {
+      "epoch": 2.5088991430454843,
+      "grad_norm": 0.20396816651465863,
+      "learning_rate": 9.066471163245358e-06,
+      "loss": 0.3488,
+      "step": 1904
+    },
+    {
+      "epoch": 2.5102175346077784,
+      "grad_norm": 0.20001805732217035,
+      "learning_rate": 9.042033235581624e-06,
+      "loss": 0.3378,
+      "step": 1905
+    },
+    {
+      "epoch": 2.5115359261700725,
+      "grad_norm": 0.19289690957102237,
+      "learning_rate": 9.017595307917889e-06,
+      "loss": 0.3245,
+      "step": 1906
+    },
+    {
+      "epoch": 2.5128543177323666,
+      "grad_norm": 0.20255819083855,
+      "learning_rate": 8.993157380254155e-06,
+      "loss": 0.355,
+      "step": 1907
+    },
+    {
+      "epoch": 2.5141727092946606,
+      "grad_norm": 0.20958254799687123,
+      "learning_rate": 8.968719452590422e-06,
+      "loss": 0.3338,
+      "step": 1908
+    },
+    {
+      "epoch": 2.5154911008569547,
+      "grad_norm": 0.20047441611289485,
+      "learning_rate": 8.944281524926687e-06,
+      "loss": 0.3518,
+      "step": 1909
+    },
+    {
+      "epoch": 2.516809492419249,
+      "grad_norm": 0.20044947364700563,
+      "learning_rate": 8.919843597262953e-06,
+      "loss": 0.3285,
+      "step": 1910
+    },
+    {
+      "epoch": 2.5181278839815424,
+      "grad_norm": 0.19360785510201522,
+      "learning_rate": 8.89540566959922e-06,
+      "loss": 0.3216,
+      "step": 1911
+    },
+    {
+      "epoch": 2.5194462755438365,
+      "grad_norm": 0.19903358847732458,
+      "learning_rate": 8.870967741935484e-06,
+      "loss": 0.3474,
+      "step": 1912
+    },
+    {
+      "epoch": 2.5207646671061306,
+      "grad_norm": 0.21350378606029605,
+      "learning_rate": 8.846529814271751e-06,
+      "loss": 0.3484,
+      "step": 1913
+    },
+    {
+      "epoch": 2.5220830586684246,
+      "grad_norm": 0.19217168227798245,
+      "learning_rate": 8.822091886608017e-06,
+      "loss": 0.338,
+      "step": 1914
+    },
+    {
+      "epoch": 2.5234014502307183,
+      "grad_norm": 0.1936732743020969,
+      "learning_rate": 8.797653958944282e-06,
+      "loss": 0.3239,
+      "step": 1915
+    },
+    {
+      "epoch": 2.5247198417930123,
+      "grad_norm": 0.21349609802703332,
+      "learning_rate": 8.773216031280549e-06,
+      "loss": 0.3566,
+      "step": 1916
+    },
+    {
+      "epoch": 2.5260382333553064,
+      "grad_norm": 0.2105968572907688,
+      "learning_rate": 8.748778103616813e-06,
+      "loss": 0.3518,
+      "step": 1917
+    },
+    {
+      "epoch": 2.5273566249176005,
+      "grad_norm": 0.2123781078095922,
+      "learning_rate": 8.72434017595308e-06,
+      "loss": 0.351,
+      "step": 1918
+    },
+    {
+      "epoch": 2.5286750164798946,
+      "grad_norm": 0.19651118354520547,
+      "learning_rate": 8.699902248289346e-06,
+      "loss": 0.339,
+      "step": 1919
+    },
+    {
+      "epoch": 2.5299934080421886,
+      "grad_norm": 0.2050652270121521,
+      "learning_rate": 8.675464320625611e-06,
+      "loss": 0.352,
+      "step": 1920
+    },
+    {
+      "epoch": 2.5313117996044827,
+      "grad_norm": 0.20369387421859933,
+      "learning_rate": 8.651026392961878e-06,
+      "loss": 0.3364,
+      "step": 1921
+    },
+    {
+      "epoch": 2.532630191166777,
+      "grad_norm": 0.20173424451502656,
+      "learning_rate": 8.626588465298144e-06,
+      "loss": 0.3282,
+      "step": 1922
+    },
+    {
+      "epoch": 2.5339485827290704,
+      "grad_norm": 0.20415137378486098,
+      "learning_rate": 8.602150537634409e-06,
+      "loss": 0.3386,
+      "step": 1923
+    },
+    {
+      "epoch": 2.5352669742913645,
+      "grad_norm": 0.19200808217413734,
+      "learning_rate": 8.577712609970674e-06,
+      "loss": 0.343,
+      "step": 1924
+    },
+    {
+      "epoch": 2.5365853658536586,
+      "grad_norm": 0.19507984103866605,
+      "learning_rate": 8.55327468230694e-06,
+      "loss": 0.332,
+      "step": 1925
+    },
+    {
+      "epoch": 2.5379037574159526,
+      "grad_norm": 0.20922584084585288,
+      "learning_rate": 8.528836754643207e-06,
+      "loss": 0.3549,
+      "step": 1926
+    },
+    {
+      "epoch": 2.5392221489782463,
+      "grad_norm": 0.22147635456177805,
+      "learning_rate": 8.504398826979472e-06,
+      "loss": 0.333,
+      "step": 1927
+    },
+    {
+      "epoch": 2.5405405405405403,
+      "grad_norm": 0.20487408765943432,
+      "learning_rate": 8.479960899315738e-06,
+      "loss": 0.3231,
+      "step": 1928
+    },
+    {
+      "epoch": 2.5418589321028344,
+      "grad_norm": 0.20175381478706247,
+      "learning_rate": 8.455522971652003e-06,
+      "loss": 0.3537,
+      "step": 1929
+    },
+    {
+      "epoch": 2.5431773236651285,
+      "grad_norm": 0.20556532111619483,
+      "learning_rate": 8.43108504398827e-06,
+      "loss": 0.3436,
+      "step": 1930
+    },
+    {
+      "epoch": 2.5444957152274226,
+      "grad_norm": 0.8372536013447391,
+      "learning_rate": 8.406647116324536e-06,
+      "loss": 0.3528,
+      "step": 1931
+    },
+    {
+      "epoch": 2.5458141067897166,
+      "grad_norm": 0.20969083245940057,
+      "learning_rate": 8.3822091886608e-06,
+      "loss": 0.3166,
+      "step": 1932
+    },
+    {
+      "epoch": 2.5471324983520107,
+      "grad_norm": 0.20921184914710955,
+      "learning_rate": 8.357771260997067e-06,
+      "loss": 0.3613,
+      "step": 1933
+    },
+    {
+      "epoch": 2.548450889914305,
+      "grad_norm": 0.22447640807804134,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.3332,
+      "step": 1934
+    },
+    {
+      "epoch": 2.5497692814765984,
+      "grad_norm": 0.20205346440317343,
+      "learning_rate": 8.308895405669599e-06,
+      "loss": 0.3418,
+      "step": 1935
+    },
+    {
+      "epoch": 2.5510876730388925,
+      "grad_norm": 0.21641330670067627,
+      "learning_rate": 8.284457478005865e-06,
+      "loss": 0.3342,
+      "step": 1936
+    },
+    {
+      "epoch": 2.5524060646011866,
+      "grad_norm": 0.2189445690171485,
+      "learning_rate": 8.260019550342132e-06,
+      "loss": 0.3271,
+      "step": 1937
+    },
+    {
+      "epoch": 2.5537244561634806,
+      "grad_norm": 0.20204395324936886,
+      "learning_rate": 8.235581622678396e-06,
+      "loss": 0.344,
+      "step": 1938
+    },
+    {
+      "epoch": 2.5550428477257743,
+      "grad_norm": 0.19556589134725533,
+      "learning_rate": 8.211143695014663e-06,
+      "loss": 0.3276,
+      "step": 1939
+    },
+    {
+      "epoch": 2.5563612392880684,
+      "grad_norm": 0.20489950258524364,
+      "learning_rate": 8.18670576735093e-06,
+      "loss": 0.3402,
+      "step": 1940
+    },
+    {
+      "epoch": 2.5576796308503624,
+      "grad_norm": 0.19874102045943082,
+      "learning_rate": 8.162267839687194e-06,
+      "loss": 0.332,
+      "step": 1941
+    },
+    {
+      "epoch": 2.5589980224126565,
+      "grad_norm": 0.20073498446758986,
+      "learning_rate": 8.13782991202346e-06,
+      "loss": 0.3259,
+      "step": 1942
+    },
+    {
+      "epoch": 2.5603164139749506,
+      "grad_norm": 0.2181210147980304,
+      "learning_rate": 8.113391984359727e-06,
+      "loss": 0.3435,
+      "step": 1943
+    },
+    {
+      "epoch": 2.5616348055372447,
+      "grad_norm": 0.2080648975449887,
+      "learning_rate": 8.088954056695992e-06,
+      "loss": 0.3374,
+      "step": 1944
+    },
+    {
+      "epoch": 2.5629531970995387,
+      "grad_norm": 0.19366765018139098,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.3477,
+      "step": 1945
+    },
+    {
+      "epoch": 2.564271588661833,
+      "grad_norm": 0.20107553972068903,
+      "learning_rate": 8.040078201368525e-06,
+      "loss": 0.3262,
+      "step": 1946
+    },
+    {
+      "epoch": 2.5655899802241264,
+      "grad_norm": 0.20307089190149993,
+      "learning_rate": 8.01564027370479e-06,
+      "loss": 0.3264,
+      "step": 1947
+    },
+    {
+      "epoch": 2.5669083717864205,
+      "grad_norm": 0.2273572963267233,
+      "learning_rate": 7.991202346041056e-06,
+      "loss": 0.3378,
+      "step": 1948
+    },
+    {
+      "epoch": 2.5682267633487146,
+      "grad_norm": 0.19309035991039558,
+      "learning_rate": 7.966764418377323e-06,
+      "loss": 0.3347,
+      "step": 1949
+    },
+    {
+      "epoch": 2.5695451549110087,
+      "grad_norm": 0.19320094121587061,
+      "learning_rate": 7.942326490713587e-06,
+      "loss": 0.3451,
+      "step": 1950
+    },
+    {
+      "epoch": 2.5708635464733027,
+      "grad_norm": 0.20856733593972984,
+      "learning_rate": 7.917888563049854e-06,
+      "loss": 0.3486,
+      "step": 1951
+    },
+    {
+      "epoch": 2.5721819380355964,
+      "grad_norm": 0.20381751761792144,
+      "learning_rate": 7.893450635386119e-06,
+      "loss": 0.3342,
+      "step": 1952
+    },
+    {
+      "epoch": 2.5735003295978904,
+      "grad_norm": 0.19077867191666764,
+      "learning_rate": 7.869012707722385e-06,
+      "loss": 0.3223,
+      "step": 1953
+    },
+    {
+      "epoch": 2.5748187211601845,
+      "grad_norm": 0.2051262546108451,
+      "learning_rate": 7.844574780058652e-06,
+      "loss": 0.3394,
+      "step": 1954
+    },
+    {
+      "epoch": 2.5761371127224786,
+      "grad_norm": 0.1891302134927005,
+      "learning_rate": 7.820136852394917e-06,
+      "loss": 0.3209,
+      "step": 1955
+    },
+    {
+      "epoch": 2.5774555042847727,
+      "grad_norm": 0.18865766818050161,
+      "learning_rate": 7.795698924731183e-06,
+      "loss": 0.3301,
+      "step": 1956
+    },
+    {
+      "epoch": 2.5787738958470667,
+      "grad_norm": 0.20215895323304742,
+      "learning_rate": 7.77126099706745e-06,
+      "loss": 0.3657,
+      "step": 1957
+    },
+    {
+      "epoch": 2.580092287409361,
+      "grad_norm": 0.19255865470192393,
+      "learning_rate": 7.746823069403714e-06,
+      "loss": 0.3427,
+      "step": 1958
+    },
+    {
+      "epoch": 2.5814106789716544,
+      "grad_norm": 0.1907480352013244,
+      "learning_rate": 7.722385141739981e-06,
+      "loss": 0.3399,
+      "step": 1959
+    },
+    {
+      "epoch": 2.5827290705339485,
+      "grad_norm": 0.20329287701731064,
+      "learning_rate": 7.697947214076247e-06,
+      "loss": 0.355,
+      "step": 1960
+    },
+    {
+      "epoch": 2.5840474620962426,
+      "grad_norm": 0.20128444478181343,
+      "learning_rate": 7.673509286412512e-06,
+      "loss": 0.3298,
+      "step": 1961
+    },
+    {
+      "epoch": 2.5853658536585367,
+      "grad_norm": 0.1971281981039563,
+      "learning_rate": 7.649071358748779e-06,
+      "loss": 0.3442,
+      "step": 1962
+    },
+    {
+      "epoch": 2.5866842452208307,
+      "grad_norm": 0.19153696519884056,
+      "learning_rate": 7.624633431085044e-06,
+      "loss": 0.3281,
+      "step": 1963
+    },
+    {
+      "epoch": 2.5880026367831244,
+      "grad_norm": 0.19560208142258093,
+      "learning_rate": 7.600195503421311e-06,
+      "loss": 0.3352,
+      "step": 1964
+    },
+    {
+      "epoch": 2.5893210283454184,
+      "grad_norm": 0.18492626069698953,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.3069,
+      "step": 1965
+    },
+    {
+      "epoch": 2.5906394199077125,
+      "grad_norm": 0.20319793336394923,
+      "learning_rate": 7.551319648093842e-06,
+      "loss": 0.3416,
+      "step": 1966
+    },
+    {
+      "epoch": 2.5919578114700066,
+      "grad_norm": 0.20350217181137578,
+      "learning_rate": 7.526881720430108e-06,
+      "loss": 0.3251,
+      "step": 1967
+    },
+    {
+      "epoch": 2.5932762030323007,
+      "grad_norm": 0.20736511876981786,
+      "learning_rate": 7.502443792766374e-06,
+      "loss": 0.3477,
+      "step": 1968
+    },
+    {
+      "epoch": 2.5945945945945947,
+      "grad_norm": 0.1954204766424121,
+      "learning_rate": 7.47800586510264e-06,
+      "loss": 0.3366,
+      "step": 1969
+    },
+    {
+      "epoch": 2.595912986156889,
+      "grad_norm": 0.19785979359627473,
+      "learning_rate": 7.4535679374389055e-06,
+      "loss": 0.3331,
+      "step": 1970
+    },
+    {
+      "epoch": 2.5972313777191824,
+      "grad_norm": 0.20456012350172212,
+      "learning_rate": 7.429130009775172e-06,
+      "loss": 0.3375,
+      "step": 1971
+    },
+    {
+      "epoch": 2.5985497692814765,
+      "grad_norm": 0.20010332653230492,
+      "learning_rate": 7.404692082111438e-06,
+      "loss": 0.3441,
+      "step": 1972
+    },
+    {
+      "epoch": 2.5998681608437706,
+      "grad_norm": 0.19980468336506466,
+      "learning_rate": 7.380254154447703e-06,
+      "loss": 0.347,
+      "step": 1973
+    },
+    {
+      "epoch": 2.6011865524060647,
+      "grad_norm": 0.22302327213945006,
+      "learning_rate": 7.35581622678397e-06,
+      "loss": 0.3478,
+      "step": 1974
+    },
+    {
+      "epoch": 2.6025049439683587,
+      "grad_norm": 0.2013099925635949,
+      "learning_rate": 7.3313782991202354e-06,
+      "loss": 0.3287,
+      "step": 1975
+    },
+    {
+      "epoch": 2.6038233355306524,
+      "grad_norm": 0.19592774386081113,
+      "learning_rate": 7.306940371456501e-06,
+      "loss": 0.3261,
+      "step": 1976
+    },
+    {
+      "epoch": 2.6051417270929464,
+      "grad_norm": 0.20981822688197466,
+      "learning_rate": 7.282502443792768e-06,
+      "loss": 0.3555,
+      "step": 1977
+    },
+    {
+      "epoch": 2.6064601186552405,
+      "grad_norm": 0.18400977550948508,
+      "learning_rate": 7.258064516129033e-06,
+      "loss": 0.3254,
+      "step": 1978
+    },
+    {
+      "epoch": 2.6077785102175346,
+      "grad_norm": 0.20176380529464624,
+      "learning_rate": 7.233626588465299e-06,
+      "loss": 0.3443,
+      "step": 1979
+    },
+    {
+      "epoch": 2.6090969017798287,
+      "grad_norm": 0.2060332520842171,
+      "learning_rate": 7.2091886608015645e-06,
+      "loss": 0.3292,
+      "step": 1980
+    },
+    {
+      "epoch": 2.6104152933421227,
+      "grad_norm": 0.2067731231292038,
+      "learning_rate": 7.184750733137831e-06,
+      "loss": 0.3473,
+      "step": 1981
+    },
+    {
+      "epoch": 2.611733684904417,
+      "grad_norm": 0.19977226325287675,
+      "learning_rate": 7.160312805474097e-06,
+      "loss": 0.3504,
+      "step": 1982
+    },
+    {
+      "epoch": 2.6130520764667104,
+      "grad_norm": 0.18846240653474325,
+      "learning_rate": 7.135874877810362e-06,
+      "loss": 0.321,
+      "step": 1983
+    },
+    {
+      "epoch": 2.6143704680290045,
+      "grad_norm": 0.20245842804753172,
+      "learning_rate": 7.111436950146629e-06,
+      "loss": 0.3501,
+      "step": 1984
+    },
+    {
+      "epoch": 2.6156888595912986,
+      "grad_norm": 0.1945413861924033,
+      "learning_rate": 7.0869990224828945e-06,
+      "loss": 0.3313,
+      "step": 1985
+    },
+    {
+      "epoch": 2.6170072511535927,
+      "grad_norm": 0.1970000665715431,
+      "learning_rate": 7.06256109481916e-06,
+      "loss": 0.3335,
+      "step": 1986
+    },
+    {
+      "epoch": 2.6183256427158867,
+      "grad_norm": 0.20357537064333797,
+      "learning_rate": 7.038123167155427e-06,
+      "loss": 0.3322,
+      "step": 1987
+    },
+    {
+      "epoch": 2.6196440342781804,
+      "grad_norm": 0.2140759379523475,
+      "learning_rate": 7.0136852394916906e-06,
+      "loss": 0.361,
+      "step": 1988
+    },
+    {
+      "epoch": 2.6209624258404745,
+      "grad_norm": 0.20874180864789343,
+      "learning_rate": 6.989247311827957e-06,
+      "loss": 0.3247,
+      "step": 1989
+    },
+    {
+      "epoch": 2.6222808174027685,
+      "grad_norm": 0.20072548261971024,
+      "learning_rate": 6.964809384164223e-06,
+      "loss": 0.3354,
+      "step": 1990
+    },
+    {
+      "epoch": 2.6235992089650626,
+      "grad_norm": 0.1973627054796603,
+      "learning_rate": 6.940371456500488e-06,
+      "loss": 0.3255,
+      "step": 1991
+    },
+    {
+      "epoch": 2.6249176005273567,
+      "grad_norm": 0.19179683414868592,
+      "learning_rate": 6.915933528836754e-06,
+      "loss": 0.3313,
+      "step": 1992
+    },
+    {
+      "epoch": 2.6262359920896508,
+      "grad_norm": 0.19887822318537124,
+      "learning_rate": 6.8914956011730205e-06,
+      "loss": 0.3316,
+      "step": 1993
+    },
+    {
+      "epoch": 2.627554383651945,
+      "grad_norm": 0.19058230977685417,
+      "learning_rate": 6.867057673509286e-06,
+      "loss": 0.328,
+      "step": 1994
+    },
+    {
+      "epoch": 2.628872775214239,
+      "grad_norm": 0.19397700147330335,
+      "learning_rate": 6.842619745845552e-06,
+      "loss": 0.3262,
+      "step": 1995
+    },
+    {
+      "epoch": 2.6301911667765325,
+      "grad_norm": 0.19539525523210186,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 0.3365,
+      "step": 1996
+    },
+    {
+      "epoch": 2.6315095583388266,
+      "grad_norm": 0.2007737390073539,
+      "learning_rate": 6.793743890518084e-06,
+      "loss": 0.3214,
+      "step": 1997
+    },
+    {
+      "epoch": 2.6328279499011207,
+      "grad_norm": 0.19798605900261598,
+      "learning_rate": 6.7693059628543496e-06,
+      "loss": 0.3543,
+      "step": 1998
+    },
+    {
+      "epoch": 2.6341463414634148,
+      "grad_norm": 0.1992955255767942,
+      "learning_rate": 6.744868035190616e-06,
+      "loss": 0.3377,
+      "step": 1999
+    },
+    {
+      "epoch": 2.6354647330257084,
+      "grad_norm": 0.20754333461020452,
+      "learning_rate": 6.720430107526882e-06,
+      "loss": 0.3397,
+      "step": 2000
+    },
+    {
+      "epoch": 2.6367831245880025,
+      "grad_norm": 0.1970810618209529,
+      "learning_rate": 6.695992179863147e-06,
+      "loss": 0.3275,
+      "step": 2001
+    },
+    {
+      "epoch": 2.6381015161502965,
+      "grad_norm": 0.19341097036814817,
+      "learning_rate": 6.671554252199414e-06,
+      "loss": 0.3138,
+      "step": 2002
+    },
+    {
+      "epoch": 2.6394199077125906,
+      "grad_norm": 0.19347986090356536,
+      "learning_rate": 6.6471163245356795e-06,
+      "loss": 0.3238,
+      "step": 2003
+    },
+    {
+      "epoch": 2.6407382992748847,
+      "grad_norm": 0.1922772034661779,
+      "learning_rate": 6.622678396871945e-06,
+      "loss": 0.31,
+      "step": 2004
+    },
+    {
+      "epoch": 2.6420566908371788,
+      "grad_norm": 0.19865275506782712,
+      "learning_rate": 6.598240469208211e-06,
+      "loss": 0.3487,
+      "step": 2005
+    },
+    {
+      "epoch": 2.643375082399473,
+      "grad_norm": 0.20230491725147687,
+      "learning_rate": 6.573802541544477e-06,
+      "loss": 0.3368,
+      "step": 2006
+    },
+    {
+      "epoch": 2.644693473961767,
+      "grad_norm": 0.23330300245909633,
+      "learning_rate": 6.549364613880743e-06,
+      "loss": 0.3428,
+      "step": 2007
+    },
+    {
+      "epoch": 2.6460118655240605,
+      "grad_norm": 0.2032280947402933,
+      "learning_rate": 6.524926686217009e-06,
+      "loss": 0.3535,
+      "step": 2008
+    },
+    {
+      "epoch": 2.6473302570863546,
+      "grad_norm": 0.18835333010398456,
+      "learning_rate": 6.500488758553275e-06,
+      "loss": 0.3051,
+      "step": 2009
+    },
+    {
+      "epoch": 2.6486486486486487,
+      "grad_norm": 0.19571031104902534,
+      "learning_rate": 6.476050830889541e-06,
+      "loss": 0.3306,
+      "step": 2010
+    },
+    {
+      "epoch": 2.6499670402109428,
+      "grad_norm": 0.20687228797888116,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.3509,
+      "step": 2011
+    },
+    {
+      "epoch": 2.6512854317732364,
+      "grad_norm": 0.19365512451294264,
+      "learning_rate": 6.427174975562073e-06,
+      "loss": 0.333,
+      "step": 2012
+    },
+    {
+      "epoch": 2.6526038233355305,
+      "grad_norm": 0.20046103345376834,
+      "learning_rate": 6.4027370478983385e-06,
+      "loss": 0.3361,
+      "step": 2013
+    },
+    {
+      "epoch": 2.6539222148978245,
+      "grad_norm": 0.20733857843267475,
+      "learning_rate": 6.378299120234604e-06,
+      "loss": 0.3566,
+      "step": 2014
+    },
+    {
+      "epoch": 2.6552406064601186,
+      "grad_norm": 0.20040317842059957,
+      "learning_rate": 6.35386119257087e-06,
+      "loss": 0.3381,
+      "step": 2015
+    },
+    {
+      "epoch": 2.6565589980224127,
+      "grad_norm": 0.20817273828628485,
+      "learning_rate": 6.329423264907136e-06,
+      "loss": 0.3266,
+      "step": 2016
+    },
+    {
+      "epoch": 2.6578773895847068,
+      "grad_norm": 0.19812342256052914,
+      "learning_rate": 6.304985337243402e-06,
+      "loss": 0.3505,
+      "step": 2017
+    },
+    {
+      "epoch": 2.659195781147001,
+      "grad_norm": 0.22034686150435662,
+      "learning_rate": 6.280547409579668e-06,
+      "loss": 0.3544,
+      "step": 2018
+    },
+    {
+      "epoch": 2.660514172709295,
+      "grad_norm": 0.19501506849787847,
+      "learning_rate": 6.256109481915934e-06,
+      "loss": 0.3327,
+      "step": 2019
+    },
+    {
+      "epoch": 2.6618325642715885,
+      "grad_norm": 0.19515117832330553,
+      "learning_rate": 6.2316715542522e-06,
+      "loss": 0.3363,
+      "step": 2020
+    },
+    {
+      "epoch": 2.6631509558338826,
+      "grad_norm": 0.20320840525049483,
+      "learning_rate": 6.207233626588465e-06,
+      "loss": 0.3487,
+      "step": 2021
+    },
+    {
+      "epoch": 2.6644693473961767,
+      "grad_norm": 0.19819592541711292,
+      "learning_rate": 6.182795698924732e-06,
+      "loss": 0.3415,
+      "step": 2022
+    },
+    {
+      "epoch": 2.6657877389584708,
+      "grad_norm": 0.2061364579198658,
+      "learning_rate": 6.1583577712609975e-06,
+      "loss": 0.3344,
+      "step": 2023
+    },
+    {
+      "epoch": 2.6671061305207644,
+      "grad_norm": 0.19147458539382373,
+      "learning_rate": 6.133919843597263e-06,
+      "loss": 0.3341,
+      "step": 2024
+    },
+    {
+      "epoch": 2.6684245220830585,
+      "grad_norm": 0.19979013944393628,
+      "learning_rate": 6.109481915933529e-06,
+      "loss": 0.3288,
+      "step": 2025
+    },
+    {
+      "epoch": 2.6697429136453525,
+      "grad_norm": 0.19329038284882966,
+      "learning_rate": 6.085043988269795e-06,
+      "loss": 0.3223,
+      "step": 2026
+    },
+    {
+      "epoch": 2.6710613052076466,
+      "grad_norm": 0.19303188734544718,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.3416,
+      "step": 2027
+    },
+    {
+      "epoch": 2.6723796967699407,
+      "grad_norm": 0.21068774891696337,
+      "learning_rate": 6.036168132942327e-06,
+      "loss": 0.3353,
+      "step": 2028
+    },
+    {
+      "epoch": 2.6736980883322348,
+      "grad_norm": 0.18628886289174187,
+      "learning_rate": 6.011730205278593e-06,
+      "loss": 0.3161,
+      "step": 2029
+    },
+    {
+      "epoch": 2.675016479894529,
+      "grad_norm": 0.18794148898306237,
+      "learning_rate": 5.987292277614859e-06,
+      "loss": 0.3286,
+      "step": 2030
+    },
+    {
+      "epoch": 2.676334871456823,
+      "grad_norm": 0.19904906076453327,
+      "learning_rate": 5.962854349951124e-06,
+      "loss": 0.3243,
+      "step": 2031
+    },
+    {
+      "epoch": 2.6776532630191165,
+      "grad_norm": 0.19953976010887278,
+      "learning_rate": 5.938416422287391e-06,
+      "loss": 0.3435,
+      "step": 2032
+    },
+    {
+      "epoch": 2.6789716545814106,
+      "grad_norm": 0.2006765515174467,
+      "learning_rate": 5.9139784946236566e-06,
+      "loss": 0.3518,
+      "step": 2033
+    },
+    {
+      "epoch": 2.6802900461437047,
+      "grad_norm": 0.1926027028431472,
+      "learning_rate": 5.889540566959922e-06,
+      "loss": 0.3329,
+      "step": 2034
+    },
+    {
+      "epoch": 2.6816084377059988,
+      "grad_norm": 0.19275061291715584,
+      "learning_rate": 5.865102639296189e-06,
+      "loss": 0.333,
+      "step": 2035
+    },
+    {
+      "epoch": 2.682926829268293,
+      "grad_norm": 0.1948314507567309,
+      "learning_rate": 5.8406647116324535e-06,
+      "loss": 0.3152,
+      "step": 2036
+    },
+    {
+      "epoch": 2.6842452208305865,
+      "grad_norm": 0.189532379159927,
+      "learning_rate": 5.816226783968719e-06,
+      "loss": 0.329,
+      "step": 2037
+    },
+    {
+      "epoch": 2.6855636123928806,
+      "grad_norm": 0.19625727744316665,
+      "learning_rate": 5.791788856304986e-06,
+      "loss": 0.3622,
+      "step": 2038
+    },
+    {
+      "epoch": 2.6868820039551746,
+      "grad_norm": 0.20748184083535912,
+      "learning_rate": 5.767350928641251e-06,
+      "loss": 0.3471,
+      "step": 2039
+    },
+    {
+      "epoch": 2.6882003955174687,
+      "grad_norm": 0.19313880039273804,
+      "learning_rate": 5.742913000977517e-06,
+      "loss": 0.3498,
+      "step": 2040
+    },
+    {
+      "epoch": 2.6895187870797628,
+      "grad_norm": 0.20043820709994997,
+      "learning_rate": 5.7184750733137834e-06,
+      "loss": 0.3353,
+      "step": 2041
+    },
+    {
+      "epoch": 2.690837178642057,
+      "grad_norm": 0.19144582864647716,
+      "learning_rate": 5.694037145650049e-06,
+      "loss": 0.3399,
+      "step": 2042
+    },
+    {
+      "epoch": 2.692155570204351,
+      "grad_norm": 0.1985032489942447,
+      "learning_rate": 5.669599217986315e-06,
+      "loss": 0.3483,
+      "step": 2043
+    },
+    {
+      "epoch": 2.6934739617666446,
+      "grad_norm": 0.21339002475642693,
+      "learning_rate": 5.64516129032258e-06,
+      "loss": 0.3483,
+      "step": 2044
+    },
+    {
+      "epoch": 2.6947923533289386,
+      "grad_norm": 0.19731285095456808,
+      "learning_rate": 5.620723362658847e-06,
+      "loss": 0.3439,
+      "step": 2045
+    },
+    {
+      "epoch": 2.6961107448912327,
+      "grad_norm": 0.19151266508411882,
+      "learning_rate": 5.5962854349951125e-06,
+      "loss": 0.3214,
+      "step": 2046
+    },
+    {
+      "epoch": 2.6974291364535268,
+      "grad_norm": 0.19531125837460966,
+      "learning_rate": 5.571847507331378e-06,
+      "loss": 0.3297,
+      "step": 2047
+    },
+    {
+      "epoch": 2.698747528015821,
+      "grad_norm": 0.2040876165893838,
+      "learning_rate": 5.547409579667645e-06,
+      "loss": 0.3723,
+      "step": 2048
+    },
+    {
+      "epoch": 2.7000659195781145,
+      "grad_norm": 0.19266760947444342,
+      "learning_rate": 5.52297165200391e-06,
+      "loss": 0.3152,
+      "step": 2049
+    },
+    {
+      "epoch": 2.7013843111404086,
+      "grad_norm": 0.19211452352467057,
+      "learning_rate": 5.498533724340176e-06,
+      "loss": 0.3342,
+      "step": 2050
+    },
+    {
+      "epoch": 2.7027027027027026,
+      "grad_norm": 0.18718079948746566,
+      "learning_rate": 5.4740957966764424e-06,
+      "loss": 0.3208,
+      "step": 2051
+    },
+    {
+      "epoch": 2.7040210942649967,
+      "grad_norm": 0.18992815020822054,
+      "learning_rate": 5.449657869012708e-06,
+      "loss": 0.34,
+      "step": 2052
+    },
+    {
+      "epoch": 2.705339485827291,
+      "grad_norm": 0.19615835086277775,
+      "learning_rate": 5.425219941348974e-06,
+      "loss": 0.3435,
+      "step": 2053
+    },
+    {
+      "epoch": 2.706657877389585,
+      "grad_norm": 0.1937842849322194,
+      "learning_rate": 5.40078201368524e-06,
+      "loss": 0.3305,
+      "step": 2054
+    },
+    {
+      "epoch": 2.707976268951879,
+      "grad_norm": 0.19573228417208113,
+      "learning_rate": 5.376344086021506e-06,
+      "loss": 0.3529,
+      "step": 2055
+    },
+    {
+      "epoch": 2.7092946605141726,
+      "grad_norm": 0.1992104277821638,
+      "learning_rate": 5.3519061583577715e-06,
+      "loss": 0.3497,
+      "step": 2056
+    },
+    {
+      "epoch": 2.7106130520764666,
+      "grad_norm": 0.20993821189830347,
+      "learning_rate": 5.327468230694037e-06,
+      "loss": 0.3506,
+      "step": 2057
+    },
+    {
+      "epoch": 2.7119314436387607,
+      "grad_norm": 0.1940839956980311,
+      "learning_rate": 5.303030303030304e-06,
+      "loss": 0.343,
+      "step": 2058
+    },
+    {
+      "epoch": 2.713249835201055,
+      "grad_norm": 0.20297715952936174,
+      "learning_rate": 5.278592375366569e-06,
+      "loss": 0.3494,
+      "step": 2059
+    },
+    {
+      "epoch": 2.714568226763349,
+      "grad_norm": 0.18868033632582276,
+      "learning_rate": 5.254154447702835e-06,
+      "loss": 0.3432,
+      "step": 2060
+    },
+    {
+      "epoch": 2.7158866183256425,
+      "grad_norm": 0.20223095182743353,
+      "learning_rate": 5.2297165200391015e-06,
+      "loss": 0.3337,
+      "step": 2061
+    },
+    {
+      "epoch": 2.7172050098879366,
+      "grad_norm": 0.19331658014090553,
+      "learning_rate": 5.205278592375367e-06,
+      "loss": 0.3315,
+      "step": 2062
+    },
+    {
+      "epoch": 2.7185234014502306,
+      "grad_norm": 0.2066808102957745,
+      "learning_rate": 5.180840664711633e-06,
+      "loss": 0.3225,
+      "step": 2063
+    },
+    {
+      "epoch": 2.7198417930125247,
+      "grad_norm": 0.18600868915737956,
+      "learning_rate": 5.156402737047899e-06,
+      "loss": 0.3229,
+      "step": 2064
+    },
+    {
+      "epoch": 2.721160184574819,
+      "grad_norm": 0.19580613504367286,
+      "learning_rate": 5.131964809384165e-06,
+      "loss": 0.3495,
+      "step": 2065
+    },
+    {
+      "epoch": 2.722478576137113,
+      "grad_norm": 0.1948653182453331,
+      "learning_rate": 5.1075268817204305e-06,
+      "loss": 0.3293,
+      "step": 2066
+    },
+    {
+      "epoch": 2.723796967699407,
+      "grad_norm": 0.19614675861979547,
+      "learning_rate": 5.083088954056696e-06,
+      "loss": 0.3218,
+      "step": 2067
+    },
+    {
+      "epoch": 2.7251153592617006,
+      "grad_norm": 0.19510032425701251,
+      "learning_rate": 5.058651026392962e-06,
+      "loss": 0.3468,
+      "step": 2068
+    },
+    {
+      "epoch": 2.7264337508239946,
+      "grad_norm": 0.19879743049180232,
+      "learning_rate": 5.0342130987292275e-06,
+      "loss": 0.3156,
+      "step": 2069
+    },
+    {
+      "epoch": 2.7277521423862887,
+      "grad_norm": 0.1907468501153064,
+      "learning_rate": 5.009775171065494e-06,
+      "loss": 0.3314,
+      "step": 2070
+    },
+    {
+      "epoch": 2.729070533948583,
+      "grad_norm": 0.19473629694975167,
+      "learning_rate": 4.98533724340176e-06,
+      "loss": 0.3423,
+      "step": 2071
+    },
+    {
+      "epoch": 2.730388925510877,
+      "grad_norm": 0.2031826090872409,
+      "learning_rate": 4.960899315738025e-06,
+      "loss": 0.3547,
+      "step": 2072
+    },
+    {
+      "epoch": 2.7317073170731705,
+      "grad_norm": 0.1941324729932609,
+      "learning_rate": 4.936461388074291e-06,
+      "loss": 0.3388,
+      "step": 2073
+    },
+    {
+      "epoch": 2.7330257086354646,
+      "grad_norm": 0.19224625675054066,
+      "learning_rate": 4.912023460410557e-06,
+      "loss": 0.3246,
+      "step": 2074
+    },
+    {
+      "epoch": 2.7343441001977586,
+      "grad_norm": 0.20572730441188697,
+      "learning_rate": 4.887585532746823e-06,
+      "loss": 0.3587,
+      "step": 2075
+    },
+    {
+      "epoch": 2.7356624917600527,
+      "grad_norm": 0.19911446416483222,
+      "learning_rate": 4.863147605083089e-06,
+      "loss": 0.341,
+      "step": 2076
+    },
+    {
+      "epoch": 2.736980883322347,
+      "grad_norm": 0.1870819339149579,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.3353,
+      "step": 2077
+    },
+    {
+      "epoch": 2.738299274884641,
+      "grad_norm": 0.20624512961830685,
+      "learning_rate": 4.814271749755621e-06,
+      "loss": 0.3307,
+      "step": 2078
+    },
+    {
+      "epoch": 2.739617666446935,
+      "grad_norm": 0.19829936656785999,
+      "learning_rate": 4.7898338220918865e-06,
+      "loss": 0.3317,
+      "step": 2079
+    },
+    {
+      "epoch": 2.740936058009229,
+      "grad_norm": 0.19581791489124453,
+      "learning_rate": 4.765395894428153e-06,
+      "loss": 0.3334,
+      "step": 2080
+    },
+    {
+      "epoch": 2.7422544495715226,
+      "grad_norm": 0.19054040546528875,
+      "learning_rate": 4.740957966764419e-06,
+      "loss": 0.3355,
+      "step": 2081
+    },
+    {
+      "epoch": 2.7435728411338167,
+      "grad_norm": 0.1863631655971733,
+      "learning_rate": 4.716520039100684e-06,
+      "loss": 0.3278,
+      "step": 2082
+    },
+    {
+      "epoch": 2.744891232696111,
+      "grad_norm": 0.200997338711363,
+      "learning_rate": 4.692082111436951e-06,
+      "loss": 0.3462,
+      "step": 2083
+    },
+    {
+      "epoch": 2.746209624258405,
+      "grad_norm": 0.19057622680622172,
+      "learning_rate": 4.6676441837732164e-06,
+      "loss": 0.3435,
+      "step": 2084
+    },
+    {
+      "epoch": 2.7475280158206985,
+      "grad_norm": 0.1950290992406957,
+      "learning_rate": 4.643206256109482e-06,
+      "loss": 0.33,
+      "step": 2085
+    },
+    {
+      "epoch": 2.7488464073829926,
+      "grad_norm": 0.2005991092795198,
+      "learning_rate": 4.618768328445748e-06,
+      "loss": 0.3383,
+      "step": 2086
+    },
+    {
+      "epoch": 2.7501647989452866,
+      "grad_norm": 0.19886301482113788,
+      "learning_rate": 4.594330400782014e-06,
+      "loss": 0.3251,
+      "step": 2087
+    },
+    {
+      "epoch": 2.7514831905075807,
+      "grad_norm": 0.2015389997755768,
+      "learning_rate": 4.56989247311828e-06,
+      "loss": 0.3453,
+      "step": 2088
+    },
+    {
+      "epoch": 2.752801582069875,
+      "grad_norm": 0.19046593810802795,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.3192,
+      "step": 2089
+    },
+    {
+      "epoch": 2.754119973632169,
+      "grad_norm": 0.19407542359190486,
+      "learning_rate": 4.521016617790812e-06,
+      "loss": 0.3332,
+      "step": 2090
+    },
+    {
+      "epoch": 2.755438365194463,
+      "grad_norm": 0.19121077363307534,
+      "learning_rate": 4.496578690127078e-06,
+      "loss": 0.3279,
+      "step": 2091
+    },
+    {
+      "epoch": 2.756756756756757,
+      "grad_norm": 0.19261910633540463,
+      "learning_rate": 4.472140762463343e-06,
+      "loss": 0.331,
+      "step": 2092
+    },
+    {
+      "epoch": 2.7580751483190507,
+      "grad_norm": 0.1872575449870302,
+      "learning_rate": 4.44770283479961e-06,
+      "loss": 0.3267,
+      "step": 2093
+    },
+    {
+      "epoch": 2.7593935398813447,
+      "grad_norm": 0.18989567062775445,
+      "learning_rate": 4.4232649071358754e-06,
+      "loss": 0.3411,
+      "step": 2094
+    },
+    {
+      "epoch": 2.760711931443639,
+      "grad_norm": 0.18481316356971866,
+      "learning_rate": 4.398826979472141e-06,
+      "loss": 0.3326,
+      "step": 2095
+    },
+    {
+      "epoch": 2.762030323005933,
+      "grad_norm": 0.20181049926404268,
+      "learning_rate": 4.374389051808407e-06,
+      "loss": 0.3266,
+      "step": 2096
+    },
+    {
+      "epoch": 2.7633487145682265,
+      "grad_norm": 0.20032100452158455,
+      "learning_rate": 4.349951124144673e-06,
+      "loss": 0.3291,
+      "step": 2097
+    },
+    {
+      "epoch": 2.7646671061305206,
+      "grad_norm": 0.19377780034441922,
+      "learning_rate": 4.325513196480939e-06,
+      "loss": 0.3459,
+      "step": 2098
+    },
+    {
+      "epoch": 2.7659854976928147,
+      "grad_norm": 0.1908243170145121,
+      "learning_rate": 4.3010752688172045e-06,
+      "loss": 0.3318,
+      "step": 2099
+    },
+    {
+      "epoch": 2.7673038892551087,
+      "grad_norm": 0.18910242374735092,
+      "learning_rate": 4.27663734115347e-06,
+      "loss": 0.3293,
+      "step": 2100
+    },
+    {
+      "epoch": 2.768622280817403,
+      "grad_norm": 0.19675577080950357,
+      "learning_rate": 4.252199413489736e-06,
+      "loss": 0.3407,
+      "step": 2101
+    },
+    {
+      "epoch": 2.769940672379697,
+      "grad_norm": 0.1837722444949726,
+      "learning_rate": 4.2277614858260015e-06,
+      "loss": 0.3194,
+      "step": 2102
+    },
+    {
+      "epoch": 2.771259063941991,
+      "grad_norm": 0.19189585539728862,
+      "learning_rate": 4.203323558162268e-06,
+      "loss": 0.3333,
+      "step": 2103
+    },
+    {
+      "epoch": 2.772577455504285,
+      "grad_norm": 0.19282412882813152,
+      "learning_rate": 4.178885630498534e-06,
+      "loss": 0.3339,
+      "step": 2104
+    },
+    {
+      "epoch": 2.7738958470665787,
+      "grad_norm": 0.1925822502664143,
+      "learning_rate": 4.154447702834799e-06,
+      "loss": 0.3359,
+      "step": 2105
+    },
+    {
+      "epoch": 2.7752142386288727,
+      "grad_norm": 0.18472455996833684,
+      "learning_rate": 4.130009775171066e-06,
+      "loss": 0.3175,
+      "step": 2106
+    },
+    {
+      "epoch": 2.776532630191167,
+      "grad_norm": 0.19938046868579282,
+      "learning_rate": 4.105571847507331e-06,
+      "loss": 0.3374,
+      "step": 2107
+    },
+    {
+      "epoch": 2.777851021753461,
+      "grad_norm": 0.19921703839707253,
+      "learning_rate": 4.081133919843597e-06,
+      "loss": 0.349,
+      "step": 2108
+    },
+    {
+      "epoch": 2.7791694133157545,
+      "grad_norm": 0.19031899853037068,
+      "learning_rate": 4.0566959921798636e-06,
+      "loss": 0.3322,
+      "step": 2109
+    },
+    {
+      "epoch": 2.7804878048780486,
+      "grad_norm": 0.19184301091768105,
+      "learning_rate": 4.032258064516129e-06,
+      "loss": 0.3466,
+      "step": 2110
+    },
+    {
+      "epoch": 2.7818061964403427,
+      "grad_norm": 0.19371359187277096,
+      "learning_rate": 4.007820136852395e-06,
+      "loss": 0.3426,
+      "step": 2111
+    },
+    {
+      "epoch": 2.7831245880026367,
+      "grad_norm": 0.19815390533729657,
+      "learning_rate": 3.983382209188661e-06,
+      "loss": 0.3405,
+      "step": 2112
+    },
+    {
+      "epoch": 2.784442979564931,
+      "grad_norm": 0.20452648638368856,
+      "learning_rate": 3.958944281524927e-06,
+      "loss": 0.3661,
+      "step": 2113
+    },
+    {
+      "epoch": 2.785761371127225,
+      "grad_norm": 0.2102472373404867,
+      "learning_rate": 3.934506353861193e-06,
+      "loss": 0.3476,
+      "step": 2114
+    },
+    {
+      "epoch": 2.787079762689519,
+      "grad_norm": 0.1924463215304557,
+      "learning_rate": 3.910068426197458e-06,
+      "loss": 0.3223,
+      "step": 2115
+    },
+    {
+      "epoch": 2.788398154251813,
+      "grad_norm": 0.19650839660020433,
+      "learning_rate": 3.885630498533725e-06,
+      "loss": 0.3175,
+      "step": 2116
+    },
+    {
+      "epoch": 2.7897165458141067,
+      "grad_norm": 0.19485558279465567,
+      "learning_rate": 3.8611925708699904e-06,
+      "loss": 0.3376,
+      "step": 2117
+    },
+    {
+      "epoch": 2.7910349373764007,
+      "grad_norm": 0.19130881866331642,
+      "learning_rate": 3.836754643206256e-06,
+      "loss": 0.339,
+      "step": 2118
+    },
+    {
+      "epoch": 2.792353328938695,
+      "grad_norm": 0.1894774268374077,
+      "learning_rate": 3.812316715542522e-06,
+      "loss": 0.3367,
+      "step": 2119
+    },
+    {
+      "epoch": 2.793671720500989,
+      "grad_norm": 0.19174458937444114,
+      "learning_rate": 3.7878787878787882e-06,
+      "loss": 0.3312,
+      "step": 2120
+    },
+    {
+      "epoch": 2.794990112063283,
+      "grad_norm": 0.2121580649478023,
+      "learning_rate": 3.763440860215054e-06,
+      "loss": 0.3562,
+      "step": 2121
+    },
+    {
+      "epoch": 2.7963085036255766,
+      "grad_norm": 0.1879505914778147,
+      "learning_rate": 3.73900293255132e-06,
+      "loss": 0.3222,
+      "step": 2122
+    },
+    {
+      "epoch": 2.7976268951878707,
+      "grad_norm": 0.19100819635223004,
+      "learning_rate": 3.714565004887586e-06,
+      "loss": 0.3226,
+      "step": 2123
+    },
+    {
+      "epoch": 2.7989452867501647,
+      "grad_norm": 0.204405373876338,
+      "learning_rate": 3.6901270772238517e-06,
+      "loss": 0.347,
+      "step": 2124
+    },
+    {
+      "epoch": 2.800263678312459,
+      "grad_norm": 0.19744741948390865,
+      "learning_rate": 3.6656891495601177e-06,
+      "loss": 0.3306,
+      "step": 2125
+    },
+    {
+      "epoch": 2.801582069874753,
+      "grad_norm": 0.19622362458741707,
+      "learning_rate": 3.641251221896384e-06,
+      "loss": 0.345,
+      "step": 2126
+    },
+    {
+      "epoch": 2.802900461437047,
+      "grad_norm": 0.178048849522325,
+      "learning_rate": 3.6168132942326494e-06,
+      "loss": 0.3079,
+      "step": 2127
+    },
+    {
+      "epoch": 2.804218852999341,
+      "grad_norm": 0.1926019981659532,
+      "learning_rate": 3.5923753665689155e-06,
+      "loss": 0.3269,
+      "step": 2128
+    },
+    {
+      "epoch": 2.8055372445616347,
+      "grad_norm": 0.19090093759852486,
+      "learning_rate": 3.567937438905181e-06,
+      "loss": 0.3268,
+      "step": 2129
+    },
+    {
+      "epoch": 2.8068556361239287,
+      "grad_norm": 0.19242608877625508,
+      "learning_rate": 3.5434995112414472e-06,
+      "loss": 0.3202,
+      "step": 2130
+    },
+    {
+      "epoch": 2.808174027686223,
+      "grad_norm": 0.2025644564586471,
+      "learning_rate": 3.5190615835777133e-06,
+      "loss": 0.3442,
+      "step": 2131
+    },
+    {
+      "epoch": 2.809492419248517,
+      "grad_norm": 0.1929920834079015,
+      "learning_rate": 3.4946236559139785e-06,
+      "loss": 0.3337,
+      "step": 2132
+    },
+    {
+      "epoch": 2.810810810810811,
+      "grad_norm": 0.18598283318342665,
+      "learning_rate": 3.470185728250244e-06,
+      "loss": 0.3359,
+      "step": 2133
+    },
+    {
+      "epoch": 2.8121292023731046,
+      "grad_norm": 0.18709037279742347,
+      "learning_rate": 3.4457478005865102e-06,
+      "loss": 0.3163,
+      "step": 2134
+    },
+    {
+      "epoch": 2.8134475939353987,
+      "grad_norm": 0.19671151873192397,
+      "learning_rate": 3.421309872922776e-06,
+      "loss": 0.3396,
+      "step": 2135
+    },
+    {
+      "epoch": 2.8147659854976927,
+      "grad_norm": 0.1963072428058073,
+      "learning_rate": 3.396871945259042e-06,
+      "loss": 0.3493,
+      "step": 2136
+    },
+    {
+      "epoch": 2.816084377059987,
+      "grad_norm": 0.1873365601008704,
+      "learning_rate": 3.372434017595308e-06,
+      "loss": 0.3339,
+      "step": 2137
+    },
+    {
+      "epoch": 2.817402768622281,
+      "grad_norm": 0.19892061671488317,
+      "learning_rate": 3.3479960899315737e-06,
+      "loss": 0.3411,
+      "step": 2138
+    },
+    {
+      "epoch": 2.818721160184575,
+      "grad_norm": 0.18966364663782573,
+      "learning_rate": 3.3235581622678398e-06,
+      "loss": 0.3308,
+      "step": 2139
+    },
+    {
+      "epoch": 2.820039551746869,
+      "grad_norm": 0.19111494019444392,
+      "learning_rate": 3.2991202346041054e-06,
+      "loss": 0.326,
+      "step": 2140
+    },
+    {
+      "epoch": 2.8213579433091627,
+      "grad_norm": 0.19058206961483667,
+      "learning_rate": 3.2746823069403715e-06,
+      "loss": 0.3288,
+      "step": 2141
+    },
+    {
+      "epoch": 2.8226763348714567,
+      "grad_norm": 0.1789953907967342,
+      "learning_rate": 3.2502443792766375e-06,
+      "loss": 0.3219,
+      "step": 2142
+    },
+    {
+      "epoch": 2.823994726433751,
+      "grad_norm": 0.19421950300574983,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.327,
+      "step": 2143
+    },
+    {
+      "epoch": 2.825313117996045,
+      "grad_norm": 0.19434339781436663,
+      "learning_rate": 3.2013685239491693e-06,
+      "loss": 0.3257,
+      "step": 2144
+    },
+    {
+      "epoch": 2.826631509558339,
+      "grad_norm": 0.18936351004004753,
+      "learning_rate": 3.176930596285435e-06,
+      "loss": 0.3353,
+      "step": 2145
+    },
+    {
+      "epoch": 2.8279499011206326,
+      "grad_norm": 0.19698498983314916,
+      "learning_rate": 3.152492668621701e-06,
+      "loss": 0.3444,
+      "step": 2146
+    },
+    {
+      "epoch": 2.8292682926829267,
+      "grad_norm": 0.1989629721632402,
+      "learning_rate": 3.128054740957967e-06,
+      "loss": 0.3387,
+      "step": 2147
+    },
+    {
+      "epoch": 2.8305866842452208,
+      "grad_norm": 0.18816343369755775,
+      "learning_rate": 3.1036168132942327e-06,
+      "loss": 0.3216,
+      "step": 2148
+    },
+    {
+      "epoch": 2.831905075807515,
+      "grad_norm": 0.18811987971680927,
+      "learning_rate": 3.0791788856304988e-06,
+      "loss": 0.3171,
+      "step": 2149
+    },
+    {
+      "epoch": 2.833223467369809,
+      "grad_norm": 0.18908532303431305,
+      "learning_rate": 3.0547409579667644e-06,
+      "loss": 0.3284,
+      "step": 2150
+    },
+    {
+      "epoch": 2.834541858932103,
+      "grad_norm": 0.19569774842923737,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.3559,
+      "step": 2151
+    },
+    {
+      "epoch": 2.835860250494397,
+      "grad_norm": 0.1949032243751212,
+      "learning_rate": 3.0058651026392966e-06,
+      "loss": 0.3353,
+      "step": 2152
+    },
+    {
+      "epoch": 2.8371786420566907,
+      "grad_norm": 0.18845388646122846,
+      "learning_rate": 2.981427174975562e-06,
+      "loss": 0.3383,
+      "step": 2153
+    },
+    {
+      "epoch": 2.8384970336189848,
+      "grad_norm": 0.19930605479223554,
+      "learning_rate": 2.9569892473118283e-06,
+      "loss": 0.3356,
+      "step": 2154
+    },
+    {
+      "epoch": 2.839815425181279,
+      "grad_norm": 0.20446211404398068,
+      "learning_rate": 2.9325513196480943e-06,
+      "loss": 0.3552,
+      "step": 2155
+    },
+    {
+      "epoch": 2.841133816743573,
+      "grad_norm": 0.19362140448930634,
+      "learning_rate": 2.9081133919843596e-06,
+      "loss": 0.3477,
+      "step": 2156
+    },
+    {
+      "epoch": 2.842452208305867,
+      "grad_norm": 0.1922226174978375,
+      "learning_rate": 2.8836754643206256e-06,
+      "loss": 0.3316,
+      "step": 2157
+    },
+    {
+      "epoch": 2.8437705998681606,
+      "grad_norm": 0.19157264482803996,
+      "learning_rate": 2.8592375366568917e-06,
+      "loss": 0.3232,
+      "step": 2158
+    },
+    {
+      "epoch": 2.8450889914304547,
+      "grad_norm": 0.18553872136340369,
+      "learning_rate": 2.8347996089931574e-06,
+      "loss": 0.3314,
+      "step": 2159
+    },
+    {
+      "epoch": 2.8464073829927488,
+      "grad_norm": 0.18757908190236655,
+      "learning_rate": 2.8103616813294234e-06,
+      "loss": 0.326,
+      "step": 2160
+    },
+    {
+      "epoch": 2.847725774555043,
+      "grad_norm": 0.19080575097158475,
+      "learning_rate": 2.785923753665689e-06,
+      "loss": 0.3092,
+      "step": 2161
+    },
+    {
+      "epoch": 2.849044166117337,
+      "grad_norm": 0.19075661301420238,
+      "learning_rate": 2.761485826001955e-06,
+      "loss": 0.3367,
+      "step": 2162
+    },
+    {
+      "epoch": 2.850362557679631,
+      "grad_norm": 0.23161400613315627,
+      "learning_rate": 2.7370478983382212e-06,
+      "loss": 0.3336,
+      "step": 2163
+    },
+    {
+      "epoch": 2.851680949241925,
+      "grad_norm": 0.19061705883834523,
+      "learning_rate": 2.712609970674487e-06,
+      "loss": 0.3402,
+      "step": 2164
+    },
+    {
+      "epoch": 2.852999340804219,
+      "grad_norm": 0.19716027726493054,
+      "learning_rate": 2.688172043010753e-06,
+      "loss": 0.3271,
+      "step": 2165
+    },
+    {
+      "epoch": 2.8543177323665128,
+      "grad_norm": 0.19041981276719772,
+      "learning_rate": 2.6637341153470186e-06,
+      "loss": 0.3328,
+      "step": 2166
+    },
+    {
+      "epoch": 2.855636123928807,
+      "grad_norm": 0.18771827080040457,
+      "learning_rate": 2.6392961876832847e-06,
+      "loss": 0.3222,
+      "step": 2167
+    },
+    {
+      "epoch": 2.856954515491101,
+      "grad_norm": 0.1955227597232197,
+      "learning_rate": 2.6148582600195507e-06,
+      "loss": 0.3386,
+      "step": 2168
+    },
+    {
+      "epoch": 2.858272907053395,
+      "grad_norm": 0.18605589896417826,
+      "learning_rate": 2.5904203323558164e-06,
+      "loss": 0.3389,
+      "step": 2169
+    },
+    {
+      "epoch": 2.8595912986156886,
+      "grad_norm": 0.19120111498301257,
+      "learning_rate": 2.5659824046920824e-06,
+      "loss": 0.3287,
+      "step": 2170
+    },
+    {
+      "epoch": 2.8609096901779827,
+      "grad_norm": 0.19531114317421822,
+      "learning_rate": 2.541544477028348e-06,
+      "loss": 0.3378,
+      "step": 2171
+    },
+    {
+      "epoch": 2.8622280817402768,
+      "grad_norm": 0.18867634578018425,
+      "learning_rate": 2.5171065493646137e-06,
+      "loss": 0.33,
+      "step": 2172
+    },
+    {
+      "epoch": 2.863546473302571,
+      "grad_norm": 0.18961809380822722,
+      "learning_rate": 2.49266862170088e-06,
+      "loss": 0.3309,
+      "step": 2173
+    },
+    {
+      "epoch": 2.864864864864865,
+      "grad_norm": 0.1843064745059568,
+      "learning_rate": 2.4682306940371455e-06,
+      "loss": 0.3278,
+      "step": 2174
+    },
+    {
+      "epoch": 2.866183256427159,
+      "grad_norm": 0.1934028206696591,
+      "learning_rate": 2.4437927663734115e-06,
+      "loss": 0.3348,
+      "step": 2175
+    },
+    {
+      "epoch": 2.867501647989453,
+      "grad_norm": 0.19738623216056772,
+      "learning_rate": 2.4193548387096776e-06,
+      "loss": 0.3578,
+      "step": 2176
+    },
+    {
+      "epoch": 2.868820039551747,
+      "grad_norm": 0.178633171919283,
+      "learning_rate": 2.3949169110459433e-06,
+      "loss": 0.3074,
+      "step": 2177
+    },
+    {
+      "epoch": 2.8701384311140408,
+      "grad_norm": 0.19019908693682677,
+      "learning_rate": 2.3704789833822093e-06,
+      "loss": 0.3315,
+      "step": 2178
+    },
+    {
+      "epoch": 2.871456822676335,
+      "grad_norm": 0.19716212037124078,
+      "learning_rate": 2.3460410557184754e-06,
+      "loss": 0.3354,
+      "step": 2179
+    },
+    {
+      "epoch": 2.872775214238629,
+      "grad_norm": 0.1958698108655413,
+      "learning_rate": 2.321603128054741e-06,
+      "loss": 0.3351,
+      "step": 2180
+    },
+    {
+      "epoch": 2.874093605800923,
+      "grad_norm": 0.19056023858701332,
+      "learning_rate": 2.297165200391007e-06,
+      "loss": 0.3319,
+      "step": 2181
+    },
+    {
+      "epoch": 2.8754119973632166,
+      "grad_norm": 0.19161409314601097,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 0.3417,
+      "step": 2182
+    },
+    {
+      "epoch": 2.8767303889255107,
+      "grad_norm": 0.18438143018644826,
+      "learning_rate": 2.248289345063539e-06,
+      "loss": 0.3179,
+      "step": 2183
+    },
+    {
+      "epoch": 2.8780487804878048,
+      "grad_norm": 0.6689029333333235,
+      "learning_rate": 2.223851417399805e-06,
+      "loss": 0.318,
+      "step": 2184
+    },
+    {
+      "epoch": 2.879367172050099,
+      "grad_norm": 0.18477973160684033,
+      "learning_rate": 2.1994134897360705e-06,
+      "loss": 0.3229,
+      "step": 2185
+    },
+    {
+      "epoch": 2.880685563612393,
+      "grad_norm": 0.1884274433487631,
+      "learning_rate": 2.1749755620723366e-06,
+      "loss": 0.3386,
+      "step": 2186
+    },
+    {
+      "epoch": 2.882003955174687,
+      "grad_norm": 0.20630423423851857,
+      "learning_rate": 2.1505376344086023e-06,
+      "loss": 0.3476,
+      "step": 2187
+    },
+    {
+      "epoch": 2.883322346736981,
+      "grad_norm": 0.18614179305840556,
+      "learning_rate": 2.126099706744868e-06,
+      "loss": 0.3192,
+      "step": 2188
+    },
+    {
+      "epoch": 2.884640738299275,
+      "grad_norm": 0.18680914808772175,
+      "learning_rate": 2.101661779081134e-06,
+      "loss": 0.3338,
+      "step": 2189
+    },
+    {
+      "epoch": 2.8859591298615688,
+      "grad_norm": 0.199192581930322,
+      "learning_rate": 2.0772238514173996e-06,
+      "loss": 0.3311,
+      "step": 2190
+    },
+    {
+      "epoch": 2.887277521423863,
+      "grad_norm": 0.18352074878779923,
+      "learning_rate": 2.0527859237536657e-06,
+      "loss": 0.3224,
+      "step": 2191
+    },
+    {
+      "epoch": 2.888595912986157,
+      "grad_norm": 0.18932127201513124,
+      "learning_rate": 2.0283479960899318e-06,
+      "loss": 0.3321,
+      "step": 2192
+    },
+    {
+      "epoch": 2.889914304548451,
+      "grad_norm": 0.19094772226045478,
+      "learning_rate": 2.0039100684261974e-06,
+      "loss": 0.3294,
+      "step": 2193
+    },
+    {
+      "epoch": 2.8912326961107446,
+      "grad_norm": 0.1867504359627165,
+      "learning_rate": 1.9794721407624635e-06,
+      "loss": 0.3301,
+      "step": 2194
+    },
+    {
+      "epoch": 2.8925510876730387,
+      "grad_norm": 0.18945708331220618,
+      "learning_rate": 1.955034213098729e-06,
+      "loss": 0.3314,
+      "step": 2195
+    },
+    {
+      "epoch": 2.8938694792353328,
+      "grad_norm": 0.2010827815855261,
+      "learning_rate": 1.9305962854349952e-06,
+      "loss": 0.3483,
+      "step": 2196
+    },
+    {
+      "epoch": 2.895187870797627,
+      "grad_norm": 0.19129046977578942,
+      "learning_rate": 1.906158357771261e-06,
+      "loss": 0.3453,
+      "step": 2197
+    },
+    {
+      "epoch": 2.896506262359921,
+      "grad_norm": 0.1858865628065424,
+      "learning_rate": 1.881720430107527e-06,
+      "loss": 0.3287,
+      "step": 2198
+    },
+    {
+      "epoch": 2.897824653922215,
+      "grad_norm": 0.1937357760715379,
+      "learning_rate": 1.857282502443793e-06,
+      "loss": 0.3362,
+      "step": 2199
+    },
+    {
+      "epoch": 2.899143045484509,
+      "grad_norm": 0.18672337117579027,
+      "learning_rate": 1.8328445747800589e-06,
+      "loss": 0.3303,
+      "step": 2200
+    },
+    {
+      "epoch": 2.900461437046803,
+      "grad_norm": 0.19972482035537656,
+      "learning_rate": 1.8084066471163247e-06,
+      "loss": 0.3558,
+      "step": 2201
+    },
+    {
+      "epoch": 2.901779828609097,
+      "grad_norm": 0.2143530123432217,
+      "learning_rate": 1.7839687194525906e-06,
+      "loss": 0.334,
+      "step": 2202
+    },
+    {
+      "epoch": 2.903098220171391,
+      "grad_norm": 0.19290532685452746,
+      "learning_rate": 1.7595307917888567e-06,
+      "loss": 0.3249,
+      "step": 2203
+    },
+    {
+      "epoch": 2.904416611733685,
+      "grad_norm": 0.19181620611320752,
+      "learning_rate": 1.735092864125122e-06,
+      "loss": 0.3293,
+      "step": 2204
+    },
+    {
+      "epoch": 2.905735003295979,
+      "grad_norm": 0.18912359310043722,
+      "learning_rate": 1.710654936461388e-06,
+      "loss": 0.3374,
+      "step": 2205
+    },
+    {
+      "epoch": 2.9070533948582726,
+      "grad_norm": 0.19112480588104505,
+      "learning_rate": 1.686217008797654e-06,
+      "loss": 0.3186,
+      "step": 2206
+    },
+    {
+      "epoch": 2.9083717864205667,
+      "grad_norm": 0.1883043907200599,
+      "learning_rate": 1.6617790811339199e-06,
+      "loss": 0.3334,
+      "step": 2207
+    },
+    {
+      "epoch": 2.909690177982861,
+      "grad_norm": 0.18806671239260117,
+      "learning_rate": 1.6373411534701857e-06,
+      "loss": 0.3238,
+      "step": 2208
+    },
+    {
+      "epoch": 2.911008569545155,
+      "grad_norm": 0.19753318311910928,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 0.3381,
+      "step": 2209
+    },
+    {
+      "epoch": 2.912326961107449,
+      "grad_norm": 0.19190350104812826,
+      "learning_rate": 1.5884652981427175e-06,
+      "loss": 0.3387,
+      "step": 2210
+    },
+    {
+      "epoch": 2.913645352669743,
+      "grad_norm": 0.19215843076182873,
+      "learning_rate": 1.5640273704789835e-06,
+      "loss": 0.3434,
+      "step": 2211
+    },
+    {
+      "epoch": 2.914963744232037,
+      "grad_norm": 0.18968106545515034,
+      "learning_rate": 1.5395894428152494e-06,
+      "loss": 0.3339,
+      "step": 2212
+    },
+    {
+      "epoch": 2.916282135794331,
+      "grad_norm": 0.1858259276620435,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 0.3335,
+      "step": 2213
+    },
+    {
+      "epoch": 2.917600527356625,
+      "grad_norm": 0.18981708618049276,
+      "learning_rate": 1.490713587487781e-06,
+      "loss": 0.3282,
+      "step": 2214
+    },
+    {
+      "epoch": 2.918918918918919,
+      "grad_norm": 0.1839769265905639,
+      "learning_rate": 1.4662756598240472e-06,
+      "loss": 0.3274,
+      "step": 2215
+    },
+    {
+      "epoch": 2.920237310481213,
+      "grad_norm": 0.18552911683989365,
+      "learning_rate": 1.4418377321603128e-06,
+      "loss": 0.3337,
+      "step": 2216
+    },
+    {
+      "epoch": 2.921555702043507,
+      "grad_norm": 0.187029140710389,
+      "learning_rate": 1.4173998044965787e-06,
+      "loss": 0.3221,
+      "step": 2217
+    },
+    {
+      "epoch": 2.922874093605801,
+      "grad_norm": 0.19811259612688467,
+      "learning_rate": 1.3929618768328445e-06,
+      "loss": 0.3435,
+      "step": 2218
+    },
+    {
+      "epoch": 2.9241924851680947,
+      "grad_norm": 0.18466205327234514,
+      "learning_rate": 1.3685239491691106e-06,
+      "loss": 0.3197,
+      "step": 2219
+    },
+    {
+      "epoch": 2.925510876730389,
+      "grad_norm": 0.192881515797869,
+      "learning_rate": 1.3440860215053765e-06,
+      "loss": 0.3436,
+      "step": 2220
+    },
+    {
+      "epoch": 2.926829268292683,
+      "grad_norm": 0.1928857699911267,
+      "learning_rate": 1.3196480938416423e-06,
+      "loss": 0.347,
+      "step": 2221
+    },
+    {
+      "epoch": 2.928147659854977,
+      "grad_norm": 0.20687379212822307,
+      "learning_rate": 1.2952101661779082e-06,
+      "loss": 0.3465,
+      "step": 2222
+    },
+    {
+      "epoch": 2.929466051417271,
+      "grad_norm": 0.20621409294889367,
+      "learning_rate": 1.270772238514174e-06,
+      "loss": 0.3311,
+      "step": 2223
+    },
+    {
+      "epoch": 2.930784442979565,
+      "grad_norm": 0.18405562442775913,
+      "learning_rate": 1.24633431085044e-06,
+      "loss": 0.3278,
+      "step": 2224
+    },
+    {
+      "epoch": 2.932102834541859,
+      "grad_norm": 0.18742244442807207,
+      "learning_rate": 1.2218963831867058e-06,
+      "loss": 0.3296,
+      "step": 2225
+    },
+    {
+      "epoch": 2.933421226104153,
+      "grad_norm": 0.20457576924428175,
+      "learning_rate": 1.1974584555229716e-06,
+      "loss": 0.3397,
+      "step": 2226
+    },
+    {
+      "epoch": 2.934739617666447,
+      "grad_norm": 0.19487384710463235,
+      "learning_rate": 1.1730205278592377e-06,
+      "loss": 0.3467,
+      "step": 2227
+    },
+    {
+      "epoch": 2.936058009228741,
+      "grad_norm": 0.1865150467862171,
+      "learning_rate": 1.1485826001955036e-06,
+      "loss": 0.305,
+      "step": 2228
+    },
+    {
+      "epoch": 2.937376400791035,
+      "grad_norm": 0.19676714121965183,
+      "learning_rate": 1.1241446725317694e-06,
+      "loss": 0.3251,
+      "step": 2229
+    },
+    {
+      "epoch": 2.938694792353329,
+      "grad_norm": 0.19336964436174725,
+      "learning_rate": 1.0997067448680353e-06,
+      "loss": 0.3299,
+      "step": 2230
+    },
+    {
+      "epoch": 2.9400131839156227,
+      "grad_norm": 0.1944381061324559,
+      "learning_rate": 1.0752688172043011e-06,
+      "loss": 0.3529,
+      "step": 2231
+    },
+    {
+      "epoch": 2.941331575477917,
+      "grad_norm": 0.19494393167163127,
+      "learning_rate": 1.050830889540567e-06,
+      "loss": 0.3408,
+      "step": 2232
+    },
+    {
+      "epoch": 2.942649967040211,
+      "grad_norm": 0.18315685734559586,
+      "learning_rate": 1.0263929618768329e-06,
+      "loss": 0.3335,
+      "step": 2233
+    },
+    {
+      "epoch": 2.943968358602505,
+      "grad_norm": 0.17892067571618472,
+      "learning_rate": 1.0019550342130987e-06,
+      "loss": 0.3223,
+      "step": 2234
+    },
+    {
+      "epoch": 2.945286750164799,
+      "grad_norm": 0.19219176167844915,
+      "learning_rate": 9.775171065493646e-07,
+      "loss": 0.3335,
+      "step": 2235
+    },
+    {
+      "epoch": 2.946605141727093,
+      "grad_norm": 0.18936754484636978,
+      "learning_rate": 9.530791788856305e-07,
+      "loss": 0.3393,
+      "step": 2236
+    },
+    {
+      "epoch": 2.947923533289387,
+      "grad_norm": 0.19362999883778786,
+      "learning_rate": 9.286412512218965e-07,
+      "loss": 0.3383,
+      "step": 2237
+    },
+    {
+      "epoch": 2.949241924851681,
+      "grad_norm": 0.1818710473371544,
+      "learning_rate": 9.042033235581624e-07,
+      "loss": 0.3243,
+      "step": 2238
+    },
+    {
+      "epoch": 2.950560316413975,
+      "grad_norm": 0.1868066591510131,
+      "learning_rate": 8.797653958944283e-07,
+      "loss": 0.3294,
+      "step": 2239
+    },
+    {
+      "epoch": 2.951878707976269,
+      "grad_norm": 0.18543916050146084,
+      "learning_rate": 8.55327468230694e-07,
+      "loss": 0.3108,
+      "step": 2240
+    },
+    {
+      "epoch": 2.953197099538563,
+      "grad_norm": 0.18615153053875813,
+      "learning_rate": 8.308895405669599e-07,
+      "loss": 0.3271,
+      "step": 2241
+    },
+    {
+      "epoch": 2.954515491100857,
+      "grad_norm": 0.1861853343179116,
+      "learning_rate": 8.064516129032258e-07,
+      "loss": 0.3299,
+      "step": 2242
+    },
+    {
+      "epoch": 2.9558338826631507,
+      "grad_norm": 0.1847176459957543,
+      "learning_rate": 7.820136852394918e-07,
+      "loss": 0.319,
+      "step": 2243
+    },
+    {
+      "epoch": 2.957152274225445,
+      "grad_norm": 0.17757771524982666,
+      "learning_rate": 7.575757575757576e-07,
+      "loss": 0.3127,
+      "step": 2244
+    },
+    {
+      "epoch": 2.958470665787739,
+      "grad_norm": 0.18456392682554573,
+      "learning_rate": 7.331378299120236e-07,
+      "loss": 0.3279,
+      "step": 2245
+    },
+    {
+      "epoch": 2.959789057350033,
+      "grad_norm": 0.19309473069088814,
+      "learning_rate": 7.086999022482893e-07,
+      "loss": 0.3291,
+      "step": 2246
+    },
+    {
+      "epoch": 2.961107448912327,
+      "grad_norm": 0.1903092877285198,
+      "learning_rate": 6.842619745845553e-07,
+      "loss": 0.3316,
+      "step": 2247
+    },
+    {
+      "epoch": 2.962425840474621,
+      "grad_norm": 0.18245210779414223,
+      "learning_rate": 6.598240469208212e-07,
+      "loss": 0.3119,
+      "step": 2248
+    },
+    {
+      "epoch": 2.963744232036915,
+      "grad_norm": 0.1841277283244122,
+      "learning_rate": 6.35386119257087e-07,
+      "loss": 0.3262,
+      "step": 2249
+    },
+    {
+      "epoch": 2.965062623599209,
+      "grad_norm": 0.1979446079768614,
+      "learning_rate": 6.109481915933529e-07,
+      "loss": 0.3546,
+      "step": 2250
+    },
+    {
+      "epoch": 2.966381015161503,
+      "grad_norm": 0.19320870858658562,
+      "learning_rate": 5.865102639296188e-07,
+      "loss": 0.3379,
+      "step": 2251
+    },
+    {
+      "epoch": 2.967699406723797,
+      "grad_norm": 0.18926690574164307,
+      "learning_rate": 5.620723362658847e-07,
+      "loss": 0.3415,
+      "step": 2252
+    },
+    {
+      "epoch": 2.969017798286091,
+      "grad_norm": 0.1924139382173674,
+      "learning_rate": 5.376344086021506e-07,
+      "loss": 0.3458,
+      "step": 2253
+    },
+    {
+      "epoch": 2.970336189848385,
+      "grad_norm": 0.18944742250280996,
+      "learning_rate": 5.131964809384164e-07,
+      "loss": 0.3557,
+      "step": 2254
+    },
+    {
+      "epoch": 2.9716545814106787,
+      "grad_norm": 0.18537515114044686,
+      "learning_rate": 4.887585532746823e-07,
+      "loss": 0.3194,
+      "step": 2255
+    },
+    {
+      "epoch": 2.972972972972973,
+      "grad_norm": 0.18652706496355523,
+      "learning_rate": 4.6432062561094825e-07,
+      "loss": 0.3259,
+      "step": 2256
+    },
+    {
+      "epoch": 2.974291364535267,
+      "grad_norm": 0.18585687966435793,
+      "learning_rate": 4.3988269794721416e-07,
+      "loss": 0.3228,
+      "step": 2257
+    },
+    {
+      "epoch": 2.975609756097561,
+      "grad_norm": 0.18051257152000585,
+      "learning_rate": 4.1544477028347997e-07,
+      "loss": 0.3164,
+      "step": 2258
+    },
+    {
+      "epoch": 2.976928147659855,
+      "grad_norm": 0.18813256826149558,
+      "learning_rate": 3.910068426197459e-07,
+      "loss": 0.3158,
+      "step": 2259
+    },
+    {
+      "epoch": 2.978246539222149,
+      "grad_norm": 0.19041665401318006,
+      "learning_rate": 3.665689149560118e-07,
+      "loss": 0.3401,
+      "step": 2260
+    },
+    {
+      "epoch": 2.979564930784443,
+      "grad_norm": 0.19103558126952788,
+      "learning_rate": 3.4213098729227765e-07,
+      "loss": 0.3521,
+      "step": 2261
+    },
+    {
+      "epoch": 2.9808833223467373,
+      "grad_norm": 0.18572400663468236,
+      "learning_rate": 3.176930596285435e-07,
+      "loss": 0.3252,
+      "step": 2262
+    },
+    {
+      "epoch": 2.982201713909031,
+      "grad_norm": 0.1877888728390521,
+      "learning_rate": 2.932551319648094e-07,
+      "loss": 0.3318,
+      "step": 2263
+    },
+    {
+      "epoch": 2.983520105471325,
+      "grad_norm": 0.19193845854430874,
+      "learning_rate": 2.688172043010753e-07,
+      "loss": 0.348,
+      "step": 2264
+    },
+    {
+      "epoch": 2.984838497033619,
+      "grad_norm": 0.18512373622893824,
+      "learning_rate": 2.4437927663734114e-07,
+      "loss": 0.3383,
+      "step": 2265
+    },
+    {
+      "epoch": 2.986156888595913,
+      "grad_norm": 0.18794082347188792,
+      "learning_rate": 2.1994134897360708e-07,
+      "loss": 0.3331,
+      "step": 2266
+    },
+    {
+      "epoch": 2.9874752801582067,
+      "grad_norm": 0.18484321334450296,
+      "learning_rate": 1.9550342130987294e-07,
+      "loss": 0.3173,
+      "step": 2267
+    },
+    {
+      "epoch": 2.988793671720501,
+      "grad_norm": 0.17841389219419607,
+      "learning_rate": 1.7106549364613883e-07,
+      "loss": 0.3181,
+      "step": 2268
+    },
+    {
+      "epoch": 2.990112063282795,
+      "grad_norm": 0.18568243227376235,
+      "learning_rate": 1.466275659824047e-07,
+      "loss": 0.3381,
+      "step": 2269
+    },
+    {
+      "epoch": 2.991430454845089,
+      "grad_norm": 0.1839480056737462,
+      "learning_rate": 1.2218963831867057e-07,
+      "loss": 0.3315,
+      "step": 2270
+    },
+    {
+      "epoch": 2.992748846407383,
+      "grad_norm": 0.1949162135658014,
+      "learning_rate": 9.775171065493647e-08,
+      "loss": 0.3574,
+      "step": 2271
+    },
+    {
+      "epoch": 2.994067237969677,
+      "grad_norm": 0.18485910166129382,
+      "learning_rate": 7.331378299120236e-08,
+      "loss": 0.336,
+      "step": 2272
+    },
+    {
+      "epoch": 2.995385629531971,
+      "grad_norm": 0.18845276045973033,
+      "learning_rate": 4.8875855327468235e-08,
+      "loss": 0.3518,
+      "step": 2273
+    },
+    {
+      "epoch": 2.9967040210942653,
+      "grad_norm": 0.18753144828161455,
+      "learning_rate": 2.4437927663734118e-08,
+      "loss": 0.3447,
+      "step": 2274
+    },
+    {
+      "epoch": 2.9967040210942653,
+      "step": 2274,
+      "total_flos": 2.527950138540399e+19,
+      "train_loss": 0.5455541882922802,
+      "train_runtime": 66596.6337,
+      "train_samples_per_second": 0.547,
+      "train_steps_per_second": 0.034
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2274,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "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": 2.527950138540399e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}