diff --git "a/adapter_info/trainer_state.json" "b/adapter_info/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/adapter_info/trainer_state.json"
@@ -0,0 +1,10766 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1532,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008071025020177562,
+      "grad_norm": 0.09123530238866806,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 2.2797,
+      "step": 1
+    },
+    {
+      "epoch": 0.0016142050040355124,
+      "grad_norm": 0.09823115170001984,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 2.2867,
+      "step": 2
+    },
+    {
+      "epoch": 0.002421307506053269,
+      "grad_norm": 0.09137655049562454,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 2.2867,
+      "step": 3
+    },
+    {
+      "epoch": 0.003228410008071025,
+      "grad_norm": 0.10211791098117828,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 2.4699,
+      "step": 4
+    },
+    {
+      "epoch": 0.004035512510088781,
+      "grad_norm": 0.09525001049041748,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 2.4167,
+      "step": 5
+    },
+    {
+      "epoch": 0.004842615012106538,
+      "grad_norm": 0.10257679224014282,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 2.505,
+      "step": 6
+    },
+    {
+      "epoch": 0.005649717514124294,
+      "grad_norm": 0.10152263939380646,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 2.2791,
+      "step": 7
+    },
+    {
+      "epoch": 0.00645682001614205,
+      "grad_norm": 0.10086437314748764,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 2.3387,
+      "step": 8
+    },
+    {
+      "epoch": 0.007263922518159807,
+      "grad_norm": 0.10246654599905014,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 2.4762,
+      "step": 9
+    },
+    {
+      "epoch": 0.008071025020177562,
+      "grad_norm": 0.09224504977464676,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 2.1899,
+      "step": 10
+    },
+    {
+      "epoch": 0.00887812752219532,
+      "grad_norm": 0.09916076809167862,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 2.3367,
+      "step": 11
+    },
+    {
+      "epoch": 0.009685230024213076,
+      "grad_norm": 0.09999067336320877,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 2.3378,
+      "step": 12
+    },
+    {
+      "epoch": 0.010492332526230832,
+      "grad_norm": 0.0847872793674469,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 2.1473,
+      "step": 13
+    },
+    {
+      "epoch": 0.011299435028248588,
+      "grad_norm": 0.09087894856929779,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 2.2322,
+      "step": 14
+    },
+    {
+      "epoch": 0.012106537530266344,
+      "grad_norm": 0.09416953474283218,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 2.2214,
+      "step": 15
+    },
+    {
+      "epoch": 0.0129136400322841,
+      "grad_norm": 0.0900375097990036,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 2.0069,
+      "step": 16
+    },
+    {
+      "epoch": 0.013720742534301856,
+      "grad_norm": 0.09621618688106537,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 2.0504,
+      "step": 17
+    },
+    {
+      "epoch": 0.014527845036319613,
+      "grad_norm": 0.08563732355833054,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 1.8789,
+      "step": 18
+    },
+    {
+      "epoch": 0.01533494753833737,
+      "grad_norm": 0.09274489432573318,
+      "learning_rate": 1e-05,
+      "loss": 1.8559,
+      "step": 19
+    },
+    {
+      "epoch": 0.016142050040355124,
+      "grad_norm": 0.08004423975944519,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 1.7343,
+      "step": 20
+    },
+    {
+      "epoch": 0.01694915254237288,
+      "grad_norm": 0.07666365802288055,
+      "learning_rate": 1.105263157894737e-05,
+      "loss": 1.6508,
+      "step": 21
+    },
+    {
+      "epoch": 0.01775625504439064,
+      "grad_norm": 0.07832133024930954,
+      "learning_rate": 1.1578947368421053e-05,
+      "loss": 1.4855,
+      "step": 22
+    },
+    {
+      "epoch": 0.018563357546408393,
+      "grad_norm": 0.09107667207717896,
+      "learning_rate": 1.2105263157894737e-05,
+      "loss": 1.8072,
+      "step": 23
+    },
+    {
+      "epoch": 0.01937046004842615,
+      "grad_norm": 0.07449404150247574,
+      "learning_rate": 1.263157894736842e-05,
+      "loss": 1.4041,
+      "step": 24
+    },
+    {
+      "epoch": 0.020177562550443905,
+      "grad_norm": 0.0739971473813057,
+      "learning_rate": 1.3157894736842108e-05,
+      "loss": 1.4815,
+      "step": 25
+    },
+    {
+      "epoch": 0.020984665052461663,
+      "grad_norm": 0.06932147592306137,
+      "learning_rate": 1.3684210526315791e-05,
+      "loss": 1.2558,
+      "step": 26
+    },
+    {
+      "epoch": 0.021791767554479417,
+      "grad_norm": 0.07036187499761581,
+      "learning_rate": 1.4210526315789475e-05,
+      "loss": 1.1721,
+      "step": 27
+    },
+    {
+      "epoch": 0.022598870056497175,
+      "grad_norm": 0.0687093511223793,
+      "learning_rate": 1.4736842105263159e-05,
+      "loss": 1.2257,
+      "step": 28
+    },
+    {
+      "epoch": 0.023405972558514933,
+      "grad_norm": 0.07659783214330673,
+      "learning_rate": 1.5263157894736846e-05,
+      "loss": 1.1707,
+      "step": 29
+    },
+    {
+      "epoch": 0.024213075060532687,
+      "grad_norm": 0.06630632281303406,
+      "learning_rate": 1.578947368421053e-05,
+      "loss": 1.0853,
+      "step": 30
+    },
+    {
+      "epoch": 0.025020177562550445,
+      "grad_norm": 0.08422867953777313,
+      "learning_rate": 1.6315789473684213e-05,
+      "loss": 1.1875,
+      "step": 31
+    },
+    {
+      "epoch": 0.0258272800645682,
+      "grad_norm": 0.06504432111978531,
+      "learning_rate": 1.6842105263157896e-05,
+      "loss": 1.0025,
+      "step": 32
+    },
+    {
+      "epoch": 0.026634382566585957,
+      "grad_norm": 0.06000803783535957,
+      "learning_rate": 1.736842105263158e-05,
+      "loss": 0.9418,
+      "step": 33
+    },
+    {
+      "epoch": 0.02744148506860371,
+      "grad_norm": 0.06218458712100983,
+      "learning_rate": 1.7894736842105264e-05,
+      "loss": 0.9268,
+      "step": 34
+    },
+    {
+      "epoch": 0.02824858757062147,
+      "grad_norm": 0.05695388466119766,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 0.9855,
+      "step": 35
+    },
+    {
+      "epoch": 0.029055690072639227,
+      "grad_norm": 0.05278893560171127,
+      "learning_rate": 1.894736842105263e-05,
+      "loss": 0.9472,
+      "step": 36
+    },
+    {
+      "epoch": 0.02986279257465698,
+      "grad_norm": 0.05200784653425217,
+      "learning_rate": 1.9473684210526318e-05,
+      "loss": 0.8339,
+      "step": 37
+    },
+    {
+      "epoch": 0.03066989507667474,
+      "grad_norm": 0.05492912605404854,
+      "learning_rate": 2e-05,
+      "loss": 0.9994,
+      "step": 38
+    },
+    {
+      "epoch": 0.031476997578692496,
+      "grad_norm": 0.049852605909109116,
+      "learning_rate": 1.9999965787604234e-05,
+      "loss": 0.8292,
+      "step": 39
+    },
+    {
+      "epoch": 0.03228410008071025,
+      "grad_norm": 0.05501826852560043,
+      "learning_rate": 1.999986315065103e-05,
+      "loss": 0.8816,
+      "step": 40
+    },
+    {
+      "epoch": 0.033091202582728005,
+      "grad_norm": 0.055790241807699203,
+      "learning_rate": 1.9999692089842683e-05,
+      "loss": 0.8772,
+      "step": 41
+    },
+    {
+      "epoch": 0.03389830508474576,
+      "grad_norm": 0.05849922075867653,
+      "learning_rate": 1.9999452606349666e-05,
+      "loss": 0.8292,
+      "step": 42
+    },
+    {
+      "epoch": 0.03470540758676352,
+      "grad_norm": 0.04576743021607399,
+      "learning_rate": 1.999914470181065e-05,
+      "loss": 0.8168,
+      "step": 43
+    },
+    {
+      "epoch": 0.03551251008878128,
+      "grad_norm": 0.0529194213449955,
+      "learning_rate": 1.9998768378332455e-05,
+      "loss": 0.8172,
+      "step": 44
+    },
+    {
+      "epoch": 0.03631961259079903,
+      "grad_norm": 0.08031924813985825,
+      "learning_rate": 1.9998323638490072e-05,
+      "loss": 0.7076,
+      "step": 45
+    },
+    {
+      "epoch": 0.03712671509281679,
+      "grad_norm": 0.05479391664266586,
+      "learning_rate": 1.9997810485326627e-05,
+      "loss": 0.8023,
+      "step": 46
+    },
+    {
+      "epoch": 0.037933817594834544,
+      "grad_norm": 0.04970843344926834,
+      "learning_rate": 1.9997228922353352e-05,
+      "loss": 0.779,
+      "step": 47
+    },
+    {
+      "epoch": 0.0387409200968523,
+      "grad_norm": 0.03704531863331795,
+      "learning_rate": 1.9996578953549584e-05,
+      "loss": 0.7064,
+      "step": 48
+    },
+    {
+      "epoch": 0.03954802259887006,
+      "grad_norm": 0.04975225776433945,
+      "learning_rate": 1.9995860583362724e-05,
+      "loss": 0.7547,
+      "step": 49
+    },
+    {
+      "epoch": 0.04035512510088781,
+      "grad_norm": 0.043880075216293335,
+      "learning_rate": 1.9995073816708203e-05,
+      "loss": 0.6839,
+      "step": 50
+    },
+    {
+      "epoch": 0.04116222760290557,
+      "grad_norm": 0.053028006106615067,
+      "learning_rate": 1.999421865896945e-05,
+      "loss": 0.8114,
+      "step": 51
+    },
+    {
+      "epoch": 0.041969330104923326,
+      "grad_norm": 0.0408385694026947,
+      "learning_rate": 1.999329511599787e-05,
+      "loss": 0.72,
+      "step": 52
+    },
+    {
+      "epoch": 0.042776432606941084,
+      "grad_norm": 0.040051594376564026,
+      "learning_rate": 1.9992303194112782e-05,
+      "loss": 0.7465,
+      "step": 53
+    },
+    {
+      "epoch": 0.043583535108958835,
+      "grad_norm": 0.03900357335805893,
+      "learning_rate": 1.99912429001014e-05,
+      "loss": 0.8,
+      "step": 54
+    },
+    {
+      "epoch": 0.04439063761097659,
+      "grad_norm": 0.03680373355746269,
+      "learning_rate": 1.9990114241218753e-05,
+      "loss": 0.6754,
+      "step": 55
+    },
+    {
+      "epoch": 0.04519774011299435,
+      "grad_norm": 0.039670154452323914,
+      "learning_rate": 1.998891722518767e-05,
+      "loss": 0.744,
+      "step": 56
+    },
+    {
+      "epoch": 0.04600484261501211,
+      "grad_norm": 0.04500587657094002,
+      "learning_rate": 1.998765186019871e-05,
+      "loss": 0.7246,
+      "step": 57
+    },
+    {
+      "epoch": 0.046811945117029866,
+      "grad_norm": 0.035050880163908005,
+      "learning_rate": 1.9986318154910106e-05,
+      "loss": 0.7131,
+      "step": 58
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 0.03428708761930466,
+      "learning_rate": 1.9984916118447704e-05,
+      "loss": 0.718,
+      "step": 59
+    },
+    {
+      "epoch": 0.048426150121065374,
+      "grad_norm": 0.03644852712750435,
+      "learning_rate": 1.9983445760404915e-05,
+      "loss": 0.7542,
+      "step": 60
+    },
+    {
+      "epoch": 0.04923325262308313,
+      "grad_norm": 0.032885193824768066,
+      "learning_rate": 1.998190709084263e-05,
+      "loss": 0.6672,
+      "step": 61
+    },
+    {
+      "epoch": 0.05004035512510089,
+      "grad_norm": 0.037364475429058075,
+      "learning_rate": 1.9980300120289164e-05,
+      "loss": 0.7374,
+      "step": 62
+    },
+    {
+      "epoch": 0.05084745762711865,
+      "grad_norm": 0.0386485680937767,
+      "learning_rate": 1.997862485974018e-05,
+      "loss": 0.7086,
+      "step": 63
+    },
+    {
+      "epoch": 0.0516545601291364,
+      "grad_norm": 0.04049423336982727,
+      "learning_rate": 1.9976881320658617e-05,
+      "loss": 0.7577,
+      "step": 64
+    },
+    {
+      "epoch": 0.052461662631154156,
+      "grad_norm": 0.03641749545931816,
+      "learning_rate": 1.99750695149746e-05,
+      "loss": 0.6173,
+      "step": 65
+    },
+    {
+      "epoch": 0.053268765133171914,
+      "grad_norm": 0.04646143317222595,
+      "learning_rate": 1.9973189455085368e-05,
+      "loss": 0.682,
+      "step": 66
+    },
+    {
+      "epoch": 0.05407586763518967,
+      "grad_norm": 0.03728644922375679,
+      "learning_rate": 1.9971241153855202e-05,
+      "loss": 0.6916,
+      "step": 67
+    },
+    {
+      "epoch": 0.05488297013720742,
+      "grad_norm": 0.03582463040947914,
+      "learning_rate": 1.9969224624615306e-05,
+      "loss": 0.5859,
+      "step": 68
+    },
+    {
+      "epoch": 0.05569007263922518,
+      "grad_norm": 0.038537513464689255,
+      "learning_rate": 1.996713988116374e-05,
+      "loss": 0.6288,
+      "step": 69
+    },
+    {
+      "epoch": 0.05649717514124294,
+      "grad_norm": 0.04025536775588989,
+      "learning_rate": 1.996498693776532e-05,
+      "loss": 0.6208,
+      "step": 70
+    },
+    {
+      "epoch": 0.057304277643260695,
+      "grad_norm": 0.04114150255918503,
+      "learning_rate": 1.996276580915151e-05,
+      "loss": 0.669,
+      "step": 71
+    },
+    {
+      "epoch": 0.05811138014527845,
+      "grad_norm": 0.03860090300440788,
+      "learning_rate": 1.996047651052034e-05,
+      "loss": 0.603,
+      "step": 72
+    },
+    {
+      "epoch": 0.058918482647296204,
+      "grad_norm": 0.041348908096551895,
+      "learning_rate": 1.995811905753629e-05,
+      "loss": 0.648,
+      "step": 73
+    },
+    {
+      "epoch": 0.05972558514931396,
+      "grad_norm": 0.048039790242910385,
+      "learning_rate": 1.9955693466330184e-05,
+      "loss": 0.7023,
+      "step": 74
+    },
+    {
+      "epoch": 0.06053268765133172,
+      "grad_norm": 0.037045303732156754,
+      "learning_rate": 1.9953199753499074e-05,
+      "loss": 0.6015,
+      "step": 75
+    },
+    {
+      "epoch": 0.06133979015334948,
+      "grad_norm": 0.042891182005405426,
+      "learning_rate": 1.9950637936106142e-05,
+      "loss": 0.6587,
+      "step": 76
+    },
+    {
+      "epoch": 0.062146892655367235,
+      "grad_norm": 0.03565942123532295,
+      "learning_rate": 1.994800803168057e-05,
+      "loss": 0.6205,
+      "step": 77
+    },
+    {
+      "epoch": 0.06295399515738499,
+      "grad_norm": 0.03730818256735802,
+      "learning_rate": 1.9945310058217422e-05,
+      "loss": 0.6147,
+      "step": 78
+    },
+    {
+      "epoch": 0.06376109765940274,
+      "grad_norm": 0.04279996454715729,
+      "learning_rate": 1.9942544034177523e-05,
+      "loss": 0.6845,
+      "step": 79
+    },
+    {
+      "epoch": 0.0645682001614205,
+      "grad_norm": 0.03682834282517433,
+      "learning_rate": 1.9939709978487344e-05,
+      "loss": 0.6147,
+      "step": 80
+    },
+    {
+      "epoch": 0.06537530266343826,
+      "grad_norm": 0.04337656497955322,
+      "learning_rate": 1.9936807910538844e-05,
+      "loss": 0.681,
+      "step": 81
+    },
+    {
+      "epoch": 0.06618240516545601,
+      "grad_norm": 0.0396178774535656,
+      "learning_rate": 1.9933837850189365e-05,
+      "loss": 0.5515,
+      "step": 82
+    },
+    {
+      "epoch": 0.06698950766747377,
+      "grad_norm": 0.04068434238433838,
+      "learning_rate": 1.993079981776148e-05,
+      "loss": 0.6166,
+      "step": 83
+    },
+    {
+      "epoch": 0.06779661016949153,
+      "grad_norm": 0.04743474721908569,
+      "learning_rate": 1.9927693834042867e-05,
+      "loss": 0.6416,
+      "step": 84
+    },
+    {
+      "epoch": 0.06860371267150928,
+      "grad_norm": 0.04356124997138977,
+      "learning_rate": 1.9924519920286153e-05,
+      "loss": 0.5854,
+      "step": 85
+    },
+    {
+      "epoch": 0.06941081517352704,
+      "grad_norm": 0.039751145988702774,
+      "learning_rate": 1.992127809820878e-05,
+      "loss": 0.5932,
+      "step": 86
+    },
+    {
+      "epoch": 0.07021791767554479,
+      "grad_norm": 0.038546741008758545,
+      "learning_rate": 1.9917968389992838e-05,
+      "loss": 0.6388,
+      "step": 87
+    },
+    {
+      "epoch": 0.07102502017756256,
+      "grad_norm": 0.0414685420691967,
+      "learning_rate": 1.9914590818284947e-05,
+      "loss": 0.6816,
+      "step": 88
+    },
+    {
+      "epoch": 0.07183212267958031,
+      "grad_norm": 0.038934748619794846,
+      "learning_rate": 1.991114540619607e-05,
+      "loss": 0.6178,
+      "step": 89
+    },
+    {
+      "epoch": 0.07263922518159806,
+      "grad_norm": 0.037200141698122025,
+      "learning_rate": 1.990763217730136e-05,
+      "loss": 0.5509,
+      "step": 90
+    },
+    {
+      "epoch": 0.07344632768361582,
+      "grad_norm": 0.03767746314406395,
+      "learning_rate": 1.9904051155640018e-05,
+      "loss": 0.6043,
+      "step": 91
+    },
+    {
+      "epoch": 0.07425343018563357,
+      "grad_norm": 0.03699817880988121,
+      "learning_rate": 1.990040236571511e-05,
+      "loss": 0.518,
+      "step": 92
+    },
+    {
+      "epoch": 0.07506053268765134,
+      "grad_norm": 0.04023592919111252,
+      "learning_rate": 1.9896685832493404e-05,
+      "loss": 0.5704,
+      "step": 93
+    },
+    {
+      "epoch": 0.07586763518966909,
+      "grad_norm": 0.03699226304888725,
+      "learning_rate": 1.98929015814052e-05,
+      "loss": 0.5447,
+      "step": 94
+    },
+    {
+      "epoch": 0.07667473769168684,
+      "grad_norm": 0.04335412383079529,
+      "learning_rate": 1.988904963834416e-05,
+      "loss": 0.5441,
+      "step": 95
+    },
+    {
+      "epoch": 0.0774818401937046,
+      "grad_norm": 0.04365302622318268,
+      "learning_rate": 1.988513002966712e-05,
+      "loss": 0.597,
+      "step": 96
+    },
+    {
+      "epoch": 0.07828894269572235,
+      "grad_norm": 0.040038615465164185,
+      "learning_rate": 1.9881142782193924e-05,
+      "loss": 0.5233,
+      "step": 97
+    },
+    {
+      "epoch": 0.07909604519774012,
+      "grad_norm": 0.03829483687877655,
+      "learning_rate": 1.987708792320723e-05,
+      "loss": 0.5462,
+      "step": 98
+    },
+    {
+      "epoch": 0.07990314769975787,
+      "grad_norm": 0.03859161585569382,
+      "learning_rate": 1.9872965480452326e-05,
+      "loss": 0.5514,
+      "step": 99
+    },
+    {
+      "epoch": 0.08071025020177562,
+      "grad_norm": 0.043581772595644,
+      "learning_rate": 1.986877548213694e-05,
+      "loss": 0.5961,
+      "step": 100
+    },
+    {
+      "epoch": 0.08151735270379339,
+      "grad_norm": 0.045486994087696075,
+      "learning_rate": 1.9864517956931046e-05,
+      "loss": 0.5943,
+      "step": 101
+    },
+    {
+      "epoch": 0.08232445520581114,
+      "grad_norm": 0.04024514555931091,
+      "learning_rate": 1.9860192933966674e-05,
+      "loss": 0.5931,
+      "step": 102
+    },
+    {
+      "epoch": 0.08313155770782889,
+      "grad_norm": 0.040237389504909515,
+      "learning_rate": 1.98558004428377e-05,
+      "loss": 0.5814,
+      "step": 103
+    },
+    {
+      "epoch": 0.08393866020984665,
+      "grad_norm": 0.045958422124385834,
+      "learning_rate": 1.9851340513599658e-05,
+      "loss": 0.6151,
+      "step": 104
+    },
+    {
+      "epoch": 0.0847457627118644,
+      "grad_norm": 0.04062805697321892,
+      "learning_rate": 1.9846813176769518e-05,
+      "loss": 0.539,
+      "step": 105
+    },
+    {
+      "epoch": 0.08555286521388217,
+      "grad_norm": 0.03963112458586693,
+      "learning_rate": 1.9842218463325488e-05,
+      "loss": 0.5272,
+      "step": 106
+    },
+    {
+      "epoch": 0.08635996771589992,
+      "grad_norm": 0.041944652795791626,
+      "learning_rate": 1.98375564047068e-05,
+      "loss": 0.6423,
+      "step": 107
+    },
+    {
+      "epoch": 0.08716707021791767,
+      "grad_norm": 0.03997455909848213,
+      "learning_rate": 1.9832827032813492e-05,
+      "loss": 0.6484,
+      "step": 108
+    },
+    {
+      "epoch": 0.08797417271993543,
+      "grad_norm": 0.03940528258681297,
+      "learning_rate": 1.9828030380006193e-05,
+      "loss": 0.5202,
+      "step": 109
+    },
+    {
+      "epoch": 0.08878127522195318,
+      "grad_norm": 0.03852485492825508,
+      "learning_rate": 1.9823166479105898e-05,
+      "loss": 0.4952,
+      "step": 110
+    },
+    {
+      "epoch": 0.08958837772397095,
+      "grad_norm": 0.04255465418100357,
+      "learning_rate": 1.9818235363393748e-05,
+      "loss": 0.5827,
+      "step": 111
+    },
+    {
+      "epoch": 0.0903954802259887,
+      "grad_norm": 0.038469936698675156,
+      "learning_rate": 1.9813237066610803e-05,
+      "loss": 0.5568,
+      "step": 112
+    },
+    {
+      "epoch": 0.09120258272800645,
+      "grad_norm": 0.04193663224577904,
+      "learning_rate": 1.9808171622957803e-05,
+      "loss": 0.5405,
+      "step": 113
+    },
+    {
+      "epoch": 0.09200968523002422,
+      "grad_norm": 0.04573473334312439,
+      "learning_rate": 1.9803039067094937e-05,
+      "loss": 0.5158,
+      "step": 114
+    },
+    {
+      "epoch": 0.09281678773204197,
+      "grad_norm": 0.04189401865005493,
+      "learning_rate": 1.9797839434141616e-05,
+      "loss": 0.5968,
+      "step": 115
+    },
+    {
+      "epoch": 0.09362389023405973,
+      "grad_norm": 0.04444560781121254,
+      "learning_rate": 1.9792572759676218e-05,
+      "loss": 0.5191,
+      "step": 116
+    },
+    {
+      "epoch": 0.09443099273607748,
+      "grad_norm": 0.04850579798221588,
+      "learning_rate": 1.9787239079735852e-05,
+      "loss": 0.563,
+      "step": 117
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 0.03827857971191406,
+      "learning_rate": 1.9781838430816117e-05,
+      "loss": 0.4721,
+      "step": 118
+    },
+    {
+      "epoch": 0.096045197740113,
+      "grad_norm": 0.041206639260053635,
+      "learning_rate": 1.9776370849870837e-05,
+      "loss": 0.5149,
+      "step": 119
+    },
+    {
+      "epoch": 0.09685230024213075,
+      "grad_norm": 0.041229840368032455,
+      "learning_rate": 1.9770836374311817e-05,
+      "loss": 0.5404,
+      "step": 120
+    },
+    {
+      "epoch": 0.09765940274414851,
+      "grad_norm": 0.046815477311611176,
+      "learning_rate": 1.9765235042008596e-05,
+      "loss": 0.5211,
+      "step": 121
+    },
+    {
+      "epoch": 0.09846650524616626,
+      "grad_norm": 0.043892793357372284,
+      "learning_rate": 1.9759566891288172e-05,
+      "loss": 0.5504,
+      "step": 122
+    },
+    {
+      "epoch": 0.09927360774818401,
+      "grad_norm": 0.041209544986486435,
+      "learning_rate": 1.9753831960934747e-05,
+      "loss": 0.531,
+      "step": 123
+    },
+    {
+      "epoch": 0.10008071025020178,
+      "grad_norm": 0.03901658579707146,
+      "learning_rate": 1.9748030290189466e-05,
+      "loss": 0.5191,
+      "step": 124
+    },
+    {
+      "epoch": 0.10088781275221953,
+      "grad_norm": 0.03842341527342796,
+      "learning_rate": 1.9742161918750135e-05,
+      "loss": 0.5228,
+      "step": 125
+    },
+    {
+      "epoch": 0.1016949152542373,
+      "grad_norm": 0.049568090587854385,
+      "learning_rate": 1.9736226886770966e-05,
+      "loss": 0.4958,
+      "step": 126
+    },
+    {
+      "epoch": 0.10250201775625505,
+      "grad_norm": 0.04007592052221298,
+      "learning_rate": 1.973022523486229e-05,
+      "loss": 0.5539,
+      "step": 127
+    },
+    {
+      "epoch": 0.1033091202582728,
+      "grad_norm": 0.0374893881380558,
+      "learning_rate": 1.972415700409029e-05,
+      "loss": 0.5441,
+      "step": 128
+    },
+    {
+      "epoch": 0.10411622276029056,
+      "grad_norm": 0.046349216252565384,
+      "learning_rate": 1.97180222359767e-05,
+      "loss": 0.5518,
+      "step": 129
+    },
+    {
+      "epoch": 0.10492332526230831,
+      "grad_norm": 0.04060063138604164,
+      "learning_rate": 1.9711820972498552e-05,
+      "loss": 0.4912,
+      "step": 130
+    },
+    {
+      "epoch": 0.10573042776432606,
+      "grad_norm": 0.03931838274002075,
+      "learning_rate": 1.9705553256087856e-05,
+      "loss": 0.4939,
+      "step": 131
+    },
+    {
+      "epoch": 0.10653753026634383,
+      "grad_norm": 0.04112625494599342,
+      "learning_rate": 1.9699219129631336e-05,
+      "loss": 0.557,
+      "step": 132
+    },
+    {
+      "epoch": 0.10734463276836158,
+      "grad_norm": 0.046433791518211365,
+      "learning_rate": 1.9692818636470113e-05,
+      "loss": 0.621,
+      "step": 133
+    },
+    {
+      "epoch": 0.10815173527037934,
+      "grad_norm": 0.04291560500860214,
+      "learning_rate": 1.968635182039943e-05,
+      "loss": 0.4939,
+      "step": 134
+    },
+    {
+      "epoch": 0.1089588377723971,
+      "grad_norm": 0.044043783098459244,
+      "learning_rate": 1.967981872566835e-05,
+      "loss": 0.5557,
+      "step": 135
+    },
+    {
+      "epoch": 0.10976594027441484,
+      "grad_norm": 0.043234121054410934,
+      "learning_rate": 1.9673219396979428e-05,
+      "loss": 0.477,
+      "step": 136
+    },
+    {
+      "epoch": 0.11057304277643261,
+      "grad_norm": 0.037246223539114,
+      "learning_rate": 1.9666553879488435e-05,
+      "loss": 0.4863,
+      "step": 137
+    },
+    {
+      "epoch": 0.11138014527845036,
+      "grad_norm": 0.04165706783533096,
+      "learning_rate": 1.9659822218804034e-05,
+      "loss": 0.5986,
+      "step": 138
+    },
+    {
+      "epoch": 0.11218724778046812,
+      "grad_norm": 0.037833321839571,
+      "learning_rate": 1.965302446098748e-05,
+      "loss": 0.5649,
+      "step": 139
+    },
+    {
+      "epoch": 0.11299435028248588,
+      "grad_norm": 0.04190123453736305,
+      "learning_rate": 1.964616065255228e-05,
+      "loss": 0.5044,
+      "step": 140
+    },
+    {
+      "epoch": 0.11380145278450363,
+      "grad_norm": 0.04529823362827301,
+      "learning_rate": 1.9639230840463907e-05,
+      "loss": 0.627,
+      "step": 141
+    },
+    {
+      "epoch": 0.11460855528652139,
+      "grad_norm": 0.03832566738128662,
+      "learning_rate": 1.963223507213945e-05,
+      "loss": 0.5422,
+      "step": 142
+    },
+    {
+      "epoch": 0.11541565778853914,
+      "grad_norm": 0.04111456498503685,
+      "learning_rate": 1.9625173395447315e-05,
+      "loss": 0.548,
+      "step": 143
+    },
+    {
+      "epoch": 0.1162227602905569,
+      "grad_norm": 0.039997123181819916,
+      "learning_rate": 1.961804585870687e-05,
+      "loss": 0.5015,
+      "step": 144
+    },
+    {
+      "epoch": 0.11702986279257466,
+      "grad_norm": 0.041700903326272964,
+      "learning_rate": 1.9610852510688142e-05,
+      "loss": 0.5675,
+      "step": 145
+    },
+    {
+      "epoch": 0.11783696529459241,
+      "grad_norm": 0.03820529580116272,
+      "learning_rate": 1.9603593400611463e-05,
+      "loss": 0.5111,
+      "step": 146
+    },
+    {
+      "epoch": 0.11864406779661017,
+      "grad_norm": 0.043333180248737335,
+      "learning_rate": 1.9596268578147142e-05,
+      "loss": 0.584,
+      "step": 147
+    },
+    {
+      "epoch": 0.11945117029862792,
+      "grad_norm": 0.03793581947684288,
+      "learning_rate": 1.9588878093415125e-05,
+      "loss": 0.5232,
+      "step": 148
+    },
+    {
+      "epoch": 0.12025827280064569,
+      "grad_norm": 0.04295019060373306,
+      "learning_rate": 1.958142199698465e-05,
+      "loss": 0.533,
+      "step": 149
+    },
+    {
+      "epoch": 0.12106537530266344,
+      "grad_norm": 0.04009397327899933,
+      "learning_rate": 1.9573900339873897e-05,
+      "loss": 0.5095,
+      "step": 150
+    },
+    {
+      "epoch": 0.12187247780468119,
+      "grad_norm": 0.04774903878569603,
+      "learning_rate": 1.9566313173549653e-05,
+      "loss": 0.5519,
+      "step": 151
+    },
+    {
+      "epoch": 0.12267958030669895,
+      "grad_norm": 0.0475604310631752,
+      "learning_rate": 1.9558660549926946e-05,
+      "loss": 0.5511,
+      "step": 152
+    },
+    {
+      "epoch": 0.1234866828087167,
+      "grad_norm": 0.037444114685058594,
+      "learning_rate": 1.9550942521368692e-05,
+      "loss": 0.5134,
+      "step": 153
+    },
+    {
+      "epoch": 0.12429378531073447,
+      "grad_norm": 0.0871221199631691,
+      "learning_rate": 1.9543159140685336e-05,
+      "loss": 0.5256,
+      "step": 154
+    },
+    {
+      "epoch": 0.12510088781275222,
+      "grad_norm": 0.03306428715586662,
+      "learning_rate": 1.9535310461134504e-05,
+      "loss": 0.4591,
+      "step": 155
+    },
+    {
+      "epoch": 0.12590799031476999,
+      "grad_norm": 0.03849384933710098,
+      "learning_rate": 1.9527396536420624e-05,
+      "loss": 0.536,
+      "step": 156
+    },
+    {
+      "epoch": 0.12671509281678772,
+      "grad_norm": 0.03767150267958641,
+      "learning_rate": 1.951941742069455e-05,
+      "loss": 0.5601,
+      "step": 157
+    },
+    {
+      "epoch": 0.1275221953188055,
+      "grad_norm": 0.03878191486001015,
+      "learning_rate": 1.9511373168553228e-05,
+      "loss": 0.4948,
+      "step": 158
+    },
+    {
+      "epoch": 0.12832929782082325,
+      "grad_norm": 0.038342781364917755,
+      "learning_rate": 1.9503263835039275e-05,
+      "loss": 0.5839,
+      "step": 159
+    },
+    {
+      "epoch": 0.129136400322841,
+      "grad_norm": 0.04736362025141716,
+      "learning_rate": 1.9495089475640644e-05,
+      "loss": 0.5203,
+      "step": 160
+    },
+    {
+      "epoch": 0.12994350282485875,
+      "grad_norm": 0.04264182969927788,
+      "learning_rate": 1.9486850146290218e-05,
+      "loss": 0.5562,
+      "step": 161
+    },
+    {
+      "epoch": 0.13075060532687652,
+      "grad_norm": 0.036652226001024246,
+      "learning_rate": 1.9478545903365432e-05,
+      "loss": 0.5224,
+      "step": 162
+    },
+    {
+      "epoch": 0.13155770782889428,
+      "grad_norm": 0.037250448018312454,
+      "learning_rate": 1.9470176803687896e-05,
+      "loss": 0.4876,
+      "step": 163
+    },
+    {
+      "epoch": 0.13236481033091202,
+      "grad_norm": 0.05088205635547638,
+      "learning_rate": 1.9461742904523003e-05,
+      "loss": 0.487,
+      "step": 164
+    },
+    {
+      "epoch": 0.13317191283292978,
+      "grad_norm": 0.04238941892981529,
+      "learning_rate": 1.9453244263579532e-05,
+      "loss": 0.4689,
+      "step": 165
+    },
+    {
+      "epoch": 0.13397901533494755,
+      "grad_norm": 0.04009280726313591,
+      "learning_rate": 1.9444680939009255e-05,
+      "loss": 0.5043,
+      "step": 166
+    },
+    {
+      "epoch": 0.13478611783696529,
+      "grad_norm": 0.04117881506681442,
+      "learning_rate": 1.9436052989406538e-05,
+      "loss": 0.5367,
+      "step": 167
+    },
+    {
+      "epoch": 0.13559322033898305,
+      "grad_norm": 0.04030587896704674,
+      "learning_rate": 1.9427360473807957e-05,
+      "loss": 0.4733,
+      "step": 168
+    },
+    {
+      "epoch": 0.13640032284100082,
+      "grad_norm": 0.04532231390476227,
+      "learning_rate": 1.941860345169186e-05,
+      "loss": 0.4976,
+      "step": 169
+    },
+    {
+      "epoch": 0.13720742534301855,
+      "grad_norm": 0.03877172991633415,
+      "learning_rate": 1.9409781982977988e-05,
+      "loss": 0.5249,
+      "step": 170
+    },
+    {
+      "epoch": 0.13801452784503632,
+      "grad_norm": 0.039575811475515366,
+      "learning_rate": 1.940089612802706e-05,
+      "loss": 0.534,
+      "step": 171
+    },
+    {
+      "epoch": 0.13882163034705408,
+      "grad_norm": 0.04503751918673515,
+      "learning_rate": 1.9391945947640352e-05,
+      "loss": 0.552,
+      "step": 172
+    },
+    {
+      "epoch": 0.13962873284907182,
+      "grad_norm": 0.041354332119226456,
+      "learning_rate": 1.938293150305929e-05,
+      "loss": 0.521,
+      "step": 173
+    },
+    {
+      "epoch": 0.14043583535108958,
+      "grad_norm": 0.03722424805164337,
+      "learning_rate": 1.937385285596502e-05,
+      "loss": 0.3982,
+      "step": 174
+    },
+    {
+      "epoch": 0.14124293785310735,
+      "grad_norm": 0.044452231377363205,
+      "learning_rate": 1.9364710068477994e-05,
+      "loss": 0.4525,
+      "step": 175
+    },
+    {
+      "epoch": 0.1420500403551251,
+      "grad_norm": 0.0437353290617466,
+      "learning_rate": 1.9355503203157545e-05,
+      "loss": 0.5436,
+      "step": 176
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.036104194819927216,
+      "learning_rate": 1.934623232300146e-05,
+      "loss": 0.463,
+      "step": 177
+    },
+    {
+      "epoch": 0.14366424535916061,
+      "grad_norm": 0.03944510221481323,
+      "learning_rate": 1.9336897491445544e-05,
+      "loss": 0.5538,
+      "step": 178
+    },
+    {
+      "epoch": 0.14447134786117838,
+      "grad_norm": 0.039386969059705734,
+      "learning_rate": 1.9327498772363187e-05,
+      "loss": 0.5193,
+      "step": 179
+    },
+    {
+      "epoch": 0.14527845036319612,
+      "grad_norm": 0.04991620406508446,
+      "learning_rate": 1.931803623006492e-05,
+      "loss": 0.5429,
+      "step": 180
+    },
+    {
+      "epoch": 0.14608555286521388,
+      "grad_norm": 0.03600313887000084,
+      "learning_rate": 1.9308509929298e-05,
+      "loss": 0.506,
+      "step": 181
+    },
+    {
+      "epoch": 0.14689265536723164,
+      "grad_norm": 0.03998609259724617,
+      "learning_rate": 1.9298919935245944e-05,
+      "loss": 0.5426,
+      "step": 182
+    },
+    {
+      "epoch": 0.14769975786924938,
+      "grad_norm": 0.040966298431158066,
+      "learning_rate": 1.9289266313528075e-05,
+      "loss": 0.5069,
+      "step": 183
+    },
+    {
+      "epoch": 0.14850686037126715,
+      "grad_norm": 0.04375036060810089,
+      "learning_rate": 1.927954913019911e-05,
+      "loss": 0.505,
+      "step": 184
+    },
+    {
+      "epoch": 0.1493139628732849,
+      "grad_norm": 0.038769107311964035,
+      "learning_rate": 1.926976845174867e-05,
+      "loss": 0.5215,
+      "step": 185
+    },
+    {
+      "epoch": 0.15012106537530268,
+      "grad_norm": 0.03845265135169029,
+      "learning_rate": 1.9259924345100837e-05,
+      "loss": 0.5043,
+      "step": 186
+    },
+    {
+      "epoch": 0.1509281678773204,
+      "grad_norm": 0.03340338170528412,
+      "learning_rate": 1.9250016877613714e-05,
+      "loss": 0.4641,
+      "step": 187
+    },
+    {
+      "epoch": 0.15173527037933818,
+      "grad_norm": 0.042416080832481384,
+      "learning_rate": 1.9240046117078935e-05,
+      "loss": 0.528,
+      "step": 188
+    },
+    {
+      "epoch": 0.15254237288135594,
+      "grad_norm": 0.04447726532816887,
+      "learning_rate": 1.9230012131721223e-05,
+      "loss": 0.541,
+      "step": 189
+    },
+    {
+      "epoch": 0.15334947538337368,
+      "grad_norm": 0.0382893830537796,
+      "learning_rate": 1.9219914990197914e-05,
+      "loss": 0.5028,
+      "step": 190
+    },
+    {
+      "epoch": 0.15415657788539144,
+      "grad_norm": 0.039434779435396194,
+      "learning_rate": 1.9209754761598488e-05,
+      "loss": 0.5406,
+      "step": 191
+    },
+    {
+      "epoch": 0.1549636803874092,
+      "grad_norm": 0.03525169566273689,
+      "learning_rate": 1.91995315154441e-05,
+      "loss": 0.4859,
+      "step": 192
+    },
+    {
+      "epoch": 0.15577078288942695,
+      "grad_norm": 0.0399198979139328,
+      "learning_rate": 1.9189245321687093e-05,
+      "loss": 0.563,
+      "step": 193
+    },
+    {
+      "epoch": 0.1565778853914447,
+      "grad_norm": 0.040046997368335724,
+      "learning_rate": 1.917889625071054e-05,
+      "loss": 0.5029,
+      "step": 194
+    },
+    {
+      "epoch": 0.15738498789346247,
+      "grad_norm": 0.04070790857076645,
+      "learning_rate": 1.916848437332774e-05,
+      "loss": 0.4785,
+      "step": 195
+    },
+    {
+      "epoch": 0.15819209039548024,
+      "grad_norm": 0.03891390189528465,
+      "learning_rate": 1.9158009760781744e-05,
+      "loss": 0.4973,
+      "step": 196
+    },
+    {
+      "epoch": 0.15899919289749798,
+      "grad_norm": 0.03878718987107277,
+      "learning_rate": 1.9147472484744873e-05,
+      "loss": 0.5465,
+      "step": 197
+    },
+    {
+      "epoch": 0.15980629539951574,
+      "grad_norm": 0.03733861818909645,
+      "learning_rate": 1.913687261731822e-05,
+      "loss": 0.5212,
+      "step": 198
+    },
+    {
+      "epoch": 0.1606133979015335,
+      "grad_norm": 0.03874503821134567,
+      "learning_rate": 1.9126210231031158e-05,
+      "loss": 0.4533,
+      "step": 199
+    },
+    {
+      "epoch": 0.16142050040355124,
+      "grad_norm": 0.03465745225548744,
+      "learning_rate": 1.9115485398840838e-05,
+      "loss": 0.445,
+      "step": 200
+    },
+    {
+      "epoch": 0.162227602905569,
+      "grad_norm": 0.04122417792677879,
+      "learning_rate": 1.9104698194131703e-05,
+      "loss": 0.5264,
+      "step": 201
+    },
+    {
+      "epoch": 0.16303470540758677,
+      "grad_norm": 0.04614901915192604,
+      "learning_rate": 1.9093848690714977e-05,
+      "loss": 0.5669,
+      "step": 202
+    },
+    {
+      "epoch": 0.1638418079096045,
+      "grad_norm": 0.03678954392671585,
+      "learning_rate": 1.908293696282816e-05,
+      "loss": 0.4039,
+      "step": 203
+    },
+    {
+      "epoch": 0.16464891041162227,
+      "grad_norm": 0.03784651309251785,
+      "learning_rate": 1.907196308513452e-05,
+      "loss": 0.4659,
+      "step": 204
+    },
+    {
+      "epoch": 0.16545601291364004,
+      "grad_norm": 0.04650195315480232,
+      "learning_rate": 1.906092713272259e-05,
+      "loss": 0.4867,
+      "step": 205
+    },
+    {
+      "epoch": 0.16626311541565778,
+      "grad_norm": 0.04255447909235954,
+      "learning_rate": 1.904982918110565e-05,
+      "loss": 0.44,
+      "step": 206
+    },
+    {
+      "epoch": 0.16707021791767554,
+      "grad_norm": 0.0433458648622036,
+      "learning_rate": 1.9038669306221194e-05,
+      "loss": 0.4586,
+      "step": 207
+    },
+    {
+      "epoch": 0.1678773204196933,
+      "grad_norm": 0.04138747602701187,
+      "learning_rate": 1.9027447584430435e-05,
+      "loss": 0.5412,
+      "step": 208
+    },
+    {
+      "epoch": 0.16868442292171107,
+      "grad_norm": 0.05272183194756508,
+      "learning_rate": 1.9016164092517767e-05,
+      "loss": 0.514,
+      "step": 209
+    },
+    {
+      "epoch": 0.1694915254237288,
+      "grad_norm": 0.04436493664979935,
+      "learning_rate": 1.9004818907690255e-05,
+      "loss": 0.4419,
+      "step": 210
+    },
+    {
+      "epoch": 0.17029862792574657,
+      "grad_norm": 0.03898143395781517,
+      "learning_rate": 1.899341210757709e-05,
+      "loss": 0.4747,
+      "step": 211
+    },
+    {
+      "epoch": 0.17110573042776434,
+      "grad_norm": 0.04165184497833252,
+      "learning_rate": 1.8981943770229058e-05,
+      "loss": 0.5105,
+      "step": 212
+    },
+    {
+      "epoch": 0.17191283292978207,
+      "grad_norm": 0.03903055191040039,
+      "learning_rate": 1.897041397411802e-05,
+      "loss": 0.4984,
+      "step": 213
+    },
+    {
+      "epoch": 0.17271993543179984,
+      "grad_norm": 0.04989544302225113,
+      "learning_rate": 1.8958822798136367e-05,
+      "loss": 0.4596,
+      "step": 214
+    },
+    {
+      "epoch": 0.1735270379338176,
+      "grad_norm": 0.040278587490320206,
+      "learning_rate": 1.8947170321596477e-05,
+      "loss": 0.5342,
+      "step": 215
+    },
+    {
+      "epoch": 0.17433414043583534,
+      "grad_norm": 0.04341874271631241,
+      "learning_rate": 1.893545662423018e-05,
+      "loss": 0.4712,
+      "step": 216
+    },
+    {
+      "epoch": 0.1751412429378531,
+      "grad_norm": 0.04516846686601639,
+      "learning_rate": 1.8923681786188207e-05,
+      "loss": 0.476,
+      "step": 217
+    },
+    {
+      "epoch": 0.17594834543987087,
+      "grad_norm": 0.039414409548044205,
+      "learning_rate": 1.891184588803964e-05,
+      "loss": 0.4865,
+      "step": 218
+    },
+    {
+      "epoch": 0.17675544794188863,
+      "grad_norm": 0.042853113263845444,
+      "learning_rate": 1.8899949010771365e-05,
+      "loss": 0.5758,
+      "step": 219
+    },
+    {
+      "epoch": 0.17756255044390637,
+      "grad_norm": 0.04502491280436516,
+      "learning_rate": 1.888799123578752e-05,
+      "loss": 0.5365,
+      "step": 220
+    },
+    {
+      "epoch": 0.17836965294592413,
+      "grad_norm": 0.03714486211538315,
+      "learning_rate": 1.8875972644908924e-05,
+      "loss": 0.5428,
+      "step": 221
+    },
+    {
+      "epoch": 0.1791767554479419,
+      "grad_norm": 0.0374261736869812,
+      "learning_rate": 1.886389332037254e-05,
+      "loss": 0.4923,
+      "step": 222
+    },
+    {
+      "epoch": 0.17998385794995964,
+      "grad_norm": 0.04436585307121277,
+      "learning_rate": 1.8851753344830897e-05,
+      "loss": 0.5548,
+      "step": 223
+    },
+    {
+      "epoch": 0.1807909604519774,
+      "grad_norm": 0.037220027297735214,
+      "learning_rate": 1.8839552801351515e-05,
+      "loss": 0.4781,
+      "step": 224
+    },
+    {
+      "epoch": 0.18159806295399517,
+      "grad_norm": 0.03513636440038681,
+      "learning_rate": 1.882729177341637e-05,
+      "loss": 0.4831,
+      "step": 225
+    },
+    {
+      "epoch": 0.1824051654560129,
+      "grad_norm": 0.04300914704799652,
+      "learning_rate": 1.881497034492128e-05,
+      "loss": 0.4867,
+      "step": 226
+    },
+    {
+      "epoch": 0.18321226795803067,
+      "grad_norm": 0.03557848930358887,
+      "learning_rate": 1.8802588600175365e-05,
+      "loss": 0.4856,
+      "step": 227
+    },
+    {
+      "epoch": 0.18401937046004843,
+      "grad_norm": 0.03743339329957962,
+      "learning_rate": 1.879014662390046e-05,
+      "loss": 0.5118,
+      "step": 228
+    },
+    {
+      "epoch": 0.1848264729620662,
+      "grad_norm": 0.037632618099451065,
+      "learning_rate": 1.877764450123053e-05,
+      "loss": 0.5256,
+      "step": 229
+    },
+    {
+      "epoch": 0.18563357546408393,
+      "grad_norm": 0.03996562957763672,
+      "learning_rate": 1.8765082317711076e-05,
+      "loss": 0.4748,
+      "step": 230
+    },
+    {
+      "epoch": 0.1864406779661017,
+      "grad_norm": 0.03809904307126999,
+      "learning_rate": 1.875246015929859e-05,
+      "loss": 0.5161,
+      "step": 231
+    },
+    {
+      "epoch": 0.18724778046811946,
+      "grad_norm": 0.037256382405757904,
+      "learning_rate": 1.8739778112359923e-05,
+      "loss": 0.4922,
+      "step": 232
+    },
+    {
+      "epoch": 0.1880548829701372,
+      "grad_norm": 0.03946573659777641,
+      "learning_rate": 1.8727036263671714e-05,
+      "loss": 0.4865,
+      "step": 233
+    },
+    {
+      "epoch": 0.18886198547215496,
+      "grad_norm": 0.03847169876098633,
+      "learning_rate": 1.8714234700419802e-05,
+      "loss": 0.4714,
+      "step": 234
+    },
+    {
+      "epoch": 0.18966908797417273,
+      "grad_norm": 0.04428127035498619,
+      "learning_rate": 1.8701373510198613e-05,
+      "loss": 0.5162,
+      "step": 235
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 0.03315884992480278,
+      "learning_rate": 1.8688452781010576e-05,
+      "loss": 0.4288,
+      "step": 236
+    },
+    {
+      "epoch": 0.19128329297820823,
+      "grad_norm": 0.044353123754262924,
+      "learning_rate": 1.867547260126551e-05,
+      "loss": 0.5049,
+      "step": 237
+    },
+    {
+      "epoch": 0.192090395480226,
+      "grad_norm": 0.041326623409986496,
+      "learning_rate": 1.8662433059780017e-05,
+      "loss": 0.5128,
+      "step": 238
+    },
+    {
+      "epoch": 0.19289749798224373,
+      "grad_norm": 0.04205748811364174,
+      "learning_rate": 1.8649334245776904e-05,
+      "loss": 0.4981,
+      "step": 239
+    },
+    {
+      "epoch": 0.1937046004842615,
+      "grad_norm": 0.04195690155029297,
+      "learning_rate": 1.8636176248884518e-05,
+      "loss": 0.4683,
+      "step": 240
+    },
+    {
+      "epoch": 0.19451170298627926,
+      "grad_norm": 0.0392894521355629,
+      "learning_rate": 1.8622959159136182e-05,
+      "loss": 0.4387,
+      "step": 241
+    },
+    {
+      "epoch": 0.19531880548829703,
+      "grad_norm": 0.04331105947494507,
+      "learning_rate": 1.860968306696956e-05,
+      "loss": 0.5649,
+      "step": 242
+    },
+    {
+      "epoch": 0.19612590799031476,
+      "grad_norm": 0.03682456538081169,
+      "learning_rate": 1.8596348063226037e-05,
+      "loss": 0.5376,
+      "step": 243
+    },
+    {
+      "epoch": 0.19693301049233253,
+      "grad_norm": 0.042542796581983566,
+      "learning_rate": 1.8582954239150095e-05,
+      "loss": 0.4451,
+      "step": 244
+    },
+    {
+      "epoch": 0.1977401129943503,
+      "grad_norm": 0.038883309811353683,
+      "learning_rate": 1.85695016863887e-05,
+      "loss": 0.4738,
+      "step": 245
+    },
+    {
+      "epoch": 0.19854721549636803,
+      "grad_norm": 0.040823884308338165,
+      "learning_rate": 1.8555990496990655e-05,
+      "loss": 0.4148,
+      "step": 246
+    },
+    {
+      "epoch": 0.1993543179983858,
+      "grad_norm": 0.03684435412287712,
+      "learning_rate": 1.8542420763406e-05,
+      "loss": 0.5052,
+      "step": 247
+    },
+    {
+      "epoch": 0.20016142050040356,
+      "grad_norm": 0.04214322566986084,
+      "learning_rate": 1.8528792578485352e-05,
+      "loss": 0.5138,
+      "step": 248
+    },
+    {
+      "epoch": 0.2009685230024213,
+      "grad_norm": 0.042726051062345505,
+      "learning_rate": 1.851510603547928e-05,
+      "loss": 0.4969,
+      "step": 249
+    },
+    {
+      "epoch": 0.20177562550443906,
+      "grad_norm": 0.034282345324754715,
+      "learning_rate": 1.8501361228037673e-05,
+      "loss": 0.4237,
+      "step": 250
+    },
+    {
+      "epoch": 0.20258272800645682,
+      "grad_norm": 0.038208913058042526,
+      "learning_rate": 1.8487558250209088e-05,
+      "loss": 0.4203,
+      "step": 251
+    },
+    {
+      "epoch": 0.2033898305084746,
+      "grad_norm": 0.04460599273443222,
+      "learning_rate": 1.847369719644011e-05,
+      "loss": 0.5293,
+      "step": 252
+    },
+    {
+      "epoch": 0.20419693301049233,
+      "grad_norm": 0.04102775827050209,
+      "learning_rate": 1.8459778161574715e-05,
+      "loss": 0.3967,
+      "step": 253
+    },
+    {
+      "epoch": 0.2050040355125101,
+      "grad_norm": 0.04214366525411606,
+      "learning_rate": 1.8445801240853603e-05,
+      "loss": 0.5049,
+      "step": 254
+    },
+    {
+      "epoch": 0.20581113801452786,
+      "grad_norm": 0.04023770987987518,
+      "learning_rate": 1.8431766529913572e-05,
+      "loss": 0.4902,
+      "step": 255
+    },
+    {
+      "epoch": 0.2066182405165456,
+      "grad_norm": 0.04067148268222809,
+      "learning_rate": 1.8417674124786832e-05,
+      "loss": 0.4539,
+      "step": 256
+    },
+    {
+      "epoch": 0.20742534301856336,
+      "grad_norm": 0.05563601106405258,
+      "learning_rate": 1.840352412190037e-05,
+      "loss": 0.5169,
+      "step": 257
+    },
+    {
+      "epoch": 0.20823244552058112,
+      "grad_norm": 0.03937901183962822,
+      "learning_rate": 1.8389316618075294e-05,
+      "loss": 0.5009,
+      "step": 258
+    },
+    {
+      "epoch": 0.20903954802259886,
+      "grad_norm": 0.04437664896249771,
+      "learning_rate": 1.8375051710526143e-05,
+      "loss": 0.499,
+      "step": 259
+    },
+    {
+      "epoch": 0.20984665052461662,
+      "grad_norm": 0.04327564314007759,
+      "learning_rate": 1.8360729496860255e-05,
+      "loss": 0.5357,
+      "step": 260
+    },
+    {
+      "epoch": 0.2106537530266344,
+      "grad_norm": 0.0612650103867054,
+      "learning_rate": 1.8346350075077078e-05,
+      "loss": 0.5325,
+      "step": 261
+    },
+    {
+      "epoch": 0.21146085552865213,
+      "grad_norm": 0.04520949721336365,
+      "learning_rate": 1.8331913543567502e-05,
+      "loss": 0.428,
+      "step": 262
+    },
+    {
+      "epoch": 0.2122679580306699,
+      "grad_norm": 0.04141777381300926,
+      "learning_rate": 1.83174200011132e-05,
+      "loss": 0.45,
+      "step": 263
+    },
+    {
+      "epoch": 0.21307506053268765,
+      "grad_norm": 0.03921473026275635,
+      "learning_rate": 1.8302869546885927e-05,
+      "loss": 0.4547,
+      "step": 264
+    },
+    {
+      "epoch": 0.21388216303470542,
+      "grad_norm": 0.040408920496702194,
+      "learning_rate": 1.8288262280446868e-05,
+      "loss": 0.5301,
+      "step": 265
+    },
+    {
+      "epoch": 0.21468926553672316,
+      "grad_norm": 0.038676533848047256,
+      "learning_rate": 1.8273598301745935e-05,
+      "loss": 0.4751,
+      "step": 266
+    },
+    {
+      "epoch": 0.21549636803874092,
+      "grad_norm": 0.04012975096702576,
+      "learning_rate": 1.8258877711121103e-05,
+      "loss": 0.428,
+      "step": 267
+    },
+    {
+      "epoch": 0.21630347054075869,
+      "grad_norm": 0.035383693873882294,
+      "learning_rate": 1.8244100609297698e-05,
+      "loss": 0.462,
+      "step": 268
+    },
+    {
+      "epoch": 0.21711057304277642,
+      "grad_norm": 0.034487176686525345,
+      "learning_rate": 1.8229267097387737e-05,
+      "loss": 0.4795,
+      "step": 269
+    },
+    {
+      "epoch": 0.2179176755447942,
+      "grad_norm": 0.03842625394463539,
+      "learning_rate": 1.8214377276889212e-05,
+      "loss": 0.5321,
+      "step": 270
+    },
+    {
+      "epoch": 0.21872477804681195,
+      "grad_norm": 0.03516872972249985,
+      "learning_rate": 1.8199431249685413e-05,
+      "loss": 0.4355,
+      "step": 271
+    },
+    {
+      "epoch": 0.2195318805488297,
+      "grad_norm": 0.03586851432919502,
+      "learning_rate": 1.818442911804422e-05,
+      "loss": 0.4521,
+      "step": 272
+    },
+    {
+      "epoch": 0.22033898305084745,
+      "grad_norm": 0.03677456080913544,
+      "learning_rate": 1.81693709846174e-05,
+      "loss": 0.4225,
+      "step": 273
+    },
+    {
+      "epoch": 0.22114608555286522,
+      "grad_norm": 0.04374263435602188,
+      "learning_rate": 1.8154256952439926e-05,
+      "loss": 0.529,
+      "step": 274
+    },
+    {
+      "epoch": 0.22195318805488298,
+      "grad_norm": 0.04119790717959404,
+      "learning_rate": 1.813908712492924e-05,
+      "loss": 0.5778,
+      "step": 275
+    },
+    {
+      "epoch": 0.22276029055690072,
+      "grad_norm": 0.03965405747294426,
+      "learning_rate": 1.8123861605884572e-05,
+      "loss": 0.4413,
+      "step": 276
+    },
+    {
+      "epoch": 0.22356739305891848,
+      "grad_norm": 0.038712743669748306,
+      "learning_rate": 1.8108580499486225e-05,
+      "loss": 0.4621,
+      "step": 277
+    },
+    {
+      "epoch": 0.22437449556093625,
+      "grad_norm": 0.03554076701402664,
+      "learning_rate": 1.8093243910294845e-05,
+      "loss": 0.4627,
+      "step": 278
+    },
+    {
+      "epoch": 0.22518159806295399,
+      "grad_norm": 0.03951288387179375,
+      "learning_rate": 1.807785194325072e-05,
+      "loss": 0.4631,
+      "step": 279
+    },
+    {
+      "epoch": 0.22598870056497175,
+      "grad_norm": 0.04191136360168457,
+      "learning_rate": 1.8062404703673078e-05,
+      "loss": 0.4426,
+      "step": 280
+    },
+    {
+      "epoch": 0.22679580306698952,
+      "grad_norm": 0.04739841818809509,
+      "learning_rate": 1.804690229725932e-05,
+      "loss": 0.4568,
+      "step": 281
+    },
+    {
+      "epoch": 0.22760290556900725,
+      "grad_norm": 0.03690197318792343,
+      "learning_rate": 1.803134483008434e-05,
+      "loss": 0.3947,
+      "step": 282
+    },
+    {
+      "epoch": 0.22841000807102502,
+      "grad_norm": 0.039233967661857605,
+      "learning_rate": 1.801573240859979e-05,
+      "loss": 0.4882,
+      "step": 283
+    },
+    {
+      "epoch": 0.22921711057304278,
+      "grad_norm": 0.04171024635434151,
+      "learning_rate": 1.8000065139633336e-05,
+      "loss": 0.4821,
+      "step": 284
+    },
+    {
+      "epoch": 0.23002421307506055,
+      "grad_norm": 0.0385863296687603,
+      "learning_rate": 1.7984343130387936e-05,
+      "loss": 0.4326,
+      "step": 285
+    },
+    {
+      "epoch": 0.23083131557707828,
+      "grad_norm": 0.042283158749341965,
+      "learning_rate": 1.7968566488441115e-05,
+      "loss": 0.4834,
+      "step": 286
+    },
+    {
+      "epoch": 0.23163841807909605,
+      "grad_norm": 0.033981919288635254,
+      "learning_rate": 1.795273532174421e-05,
+      "loss": 0.4169,
+      "step": 287
+    },
+    {
+      "epoch": 0.2324455205811138,
+      "grad_norm": 0.034205395728349686,
+      "learning_rate": 1.7936849738621656e-05,
+      "loss": 0.4283,
+      "step": 288
+    },
+    {
+      "epoch": 0.23325262308313155,
+      "grad_norm": 0.03619087114930153,
+      "learning_rate": 1.7920909847770223e-05,
+      "loss": 0.4406,
+      "step": 289
+    },
+    {
+      "epoch": 0.23405972558514931,
+      "grad_norm": 0.04226299375295639,
+      "learning_rate": 1.790491575825828e-05,
+      "loss": 0.5038,
+      "step": 290
+    },
+    {
+      "epoch": 0.23486682808716708,
+      "grad_norm": 0.04093988239765167,
+      "learning_rate": 1.7888867579525045e-05,
+      "loss": 0.508,
+      "step": 291
+    },
+    {
+      "epoch": 0.23567393058918482,
+      "grad_norm": 0.040389351546764374,
+      "learning_rate": 1.787276542137986e-05,
+      "loss": 0.5127,
+      "step": 292
+    },
+    {
+      "epoch": 0.23648103309120258,
+      "grad_norm": 0.03728820011019707,
+      "learning_rate": 1.78566093940014e-05,
+      "loss": 0.4442,
+      "step": 293
+    },
+    {
+      "epoch": 0.23728813559322035,
+      "grad_norm": 0.0355042964220047,
+      "learning_rate": 1.784039960793694e-05,
+      "loss": 0.4448,
+      "step": 294
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 0.039183955639600754,
+      "learning_rate": 1.7824136174101615e-05,
+      "loss": 0.4753,
+      "step": 295
+    },
+    {
+      "epoch": 0.23890234059725585,
+      "grad_norm": 0.039957690984010696,
+      "learning_rate": 1.780781920377762e-05,
+      "loss": 0.4083,
+      "step": 296
+    },
+    {
+      "epoch": 0.2397094430992736,
+      "grad_norm": 0.03782647103071213,
+      "learning_rate": 1.779144880861349e-05,
+      "loss": 0.4701,
+      "step": 297
+    },
+    {
+      "epoch": 0.24051654560129138,
+      "grad_norm": 0.04110096022486687,
+      "learning_rate": 1.777502510062331e-05,
+      "loss": 0.3975,
+      "step": 298
+    },
+    {
+      "epoch": 0.2413236481033091,
+      "grad_norm": 0.039122823625802994,
+      "learning_rate": 1.7758548192185966e-05,
+      "loss": 0.4414,
+      "step": 299
+    },
+    {
+      "epoch": 0.24213075060532688,
+      "grad_norm": 0.041810158640146255,
+      "learning_rate": 1.7742018196044352e-05,
+      "loss": 0.4951,
+      "step": 300
+    },
+    {
+      "epoch": 0.24293785310734464,
+      "grad_norm": 0.04400990530848503,
+      "learning_rate": 1.7725435225304632e-05,
+      "loss": 0.5081,
+      "step": 301
+    },
+    {
+      "epoch": 0.24374495560936238,
+      "grad_norm": 0.036244601011276245,
+      "learning_rate": 1.7708799393435426e-05,
+      "loss": 0.4776,
+      "step": 302
+    },
+    {
+      "epoch": 0.24455205811138014,
+      "grad_norm": 0.04246852546930313,
+      "learning_rate": 1.7692110814267077e-05,
+      "loss": 0.409,
+      "step": 303
+    },
+    {
+      "epoch": 0.2453591606133979,
+      "grad_norm": 0.03885439783334732,
+      "learning_rate": 1.7675369601990834e-05,
+      "loss": 0.4556,
+      "step": 304
+    },
+    {
+      "epoch": 0.24616626311541565,
+      "grad_norm": 0.04391565918922424,
+      "learning_rate": 1.765857587115809e-05,
+      "loss": 0.5051,
+      "step": 305
+    },
+    {
+      "epoch": 0.2469733656174334,
+      "grad_norm": 0.044587794691324234,
+      "learning_rate": 1.7641729736679608e-05,
+      "loss": 0.525,
+      "step": 306
+    },
+    {
+      "epoch": 0.24778046811945117,
+      "grad_norm": 0.03848964348435402,
+      "learning_rate": 1.7624831313824707e-05,
+      "loss": 0.4648,
+      "step": 307
+    },
+    {
+      "epoch": 0.24858757062146894,
+      "grad_norm": 0.04077491536736488,
+      "learning_rate": 1.7607880718220493e-05,
+      "loss": 0.4774,
+      "step": 308
+    },
+    {
+      "epoch": 0.24939467312348668,
+      "grad_norm": 0.03753182664513588,
+      "learning_rate": 1.759087806585106e-05,
+      "loss": 0.4698,
+      "step": 309
+    },
+    {
+      "epoch": 0.25020177562550444,
+      "grad_norm": 0.038406360894441605,
+      "learning_rate": 1.7573823473056704e-05,
+      "loss": 0.4519,
+      "step": 310
+    },
+    {
+      "epoch": 0.2510088781275222,
+      "grad_norm": 0.03918207436800003,
+      "learning_rate": 1.7556717056533124e-05,
+      "loss": 0.4447,
+      "step": 311
+    },
+    {
+      "epoch": 0.25181598062953997,
+      "grad_norm": 0.04120977967977524,
+      "learning_rate": 1.7539558933330618e-05,
+      "loss": 0.4791,
+      "step": 312
+    },
+    {
+      "epoch": 0.25262308313155774,
+      "grad_norm": 0.043005820363759995,
+      "learning_rate": 1.7522349220853284e-05,
+      "loss": 0.4385,
+      "step": 313
+    },
+    {
+      "epoch": 0.25343018563357544,
+      "grad_norm": 0.043990328907966614,
+      "learning_rate": 1.750508803685822e-05,
+      "loss": 0.4184,
+      "step": 314
+    },
+    {
+      "epoch": 0.2542372881355932,
+      "grad_norm": 0.0447077676653862,
+      "learning_rate": 1.748777549945472e-05,
+      "loss": 0.4524,
+      "step": 315
+    },
+    {
+      "epoch": 0.255044390637611,
+      "grad_norm": 0.046113114804029465,
+      "learning_rate": 1.747041172710346e-05,
+      "loss": 0.4221,
+      "step": 316
+    },
+    {
+      "epoch": 0.25585149313962874,
+      "grad_norm": 0.0372101292014122,
+      "learning_rate": 1.745299683861569e-05,
+      "loss": 0.439,
+      "step": 317
+    },
+    {
+      "epoch": 0.2566585956416465,
+      "grad_norm": 0.040746718645095825,
+      "learning_rate": 1.743553095315242e-05,
+      "loss": 0.4499,
+      "step": 318
+    },
+    {
+      "epoch": 0.25746569814366427,
+      "grad_norm": 0.03624715283513069,
+      "learning_rate": 1.741801419022361e-05,
+      "loss": 0.4495,
+      "step": 319
+    },
+    {
+      "epoch": 0.258272800645682,
+      "grad_norm": 0.04276509955525398,
+      "learning_rate": 1.740044666968734e-05,
+      "loss": 0.4393,
+      "step": 320
+    },
+    {
+      "epoch": 0.25907990314769974,
+      "grad_norm": 0.041294898837804794,
+      "learning_rate": 1.738282851174901e-05,
+      "loss": 0.4413,
+      "step": 321
+    },
+    {
+      "epoch": 0.2598870056497175,
+      "grad_norm": 0.042043715715408325,
+      "learning_rate": 1.7365159836960494e-05,
+      "loss": 0.5303,
+      "step": 322
+    },
+    {
+      "epoch": 0.26069410815173527,
+      "grad_norm": 0.04050707444548607,
+      "learning_rate": 1.734744076621933e-05,
+      "loss": 0.5062,
+      "step": 323
+    },
+    {
+      "epoch": 0.26150121065375304,
+      "grad_norm": 0.03927404060959816,
+      "learning_rate": 1.7329671420767898e-05,
+      "loss": 0.4519,
+      "step": 324
+    },
+    {
+      "epoch": 0.2623083131557708,
+      "grad_norm": 0.04352013021707535,
+      "learning_rate": 1.7311851922192568e-05,
+      "loss": 0.4835,
+      "step": 325
+    },
+    {
+      "epoch": 0.26311541565778856,
+      "grad_norm": 0.04132748767733574,
+      "learning_rate": 1.729398239242288e-05,
+      "loss": 0.4179,
+      "step": 326
+    },
+    {
+      "epoch": 0.2639225181598063,
+      "grad_norm": 0.04499141499400139,
+      "learning_rate": 1.727606295373073e-05,
+      "loss": 0.4907,
+      "step": 327
+    },
+    {
+      "epoch": 0.26472962066182404,
+      "grad_norm": 0.0411851666867733,
+      "learning_rate": 1.7258093728729503e-05,
+      "loss": 0.4438,
+      "step": 328
+    },
+    {
+      "epoch": 0.2655367231638418,
+      "grad_norm": 0.045802779495716095,
+      "learning_rate": 1.724007484037324e-05,
+      "loss": 0.4276,
+      "step": 329
+    },
+    {
+      "epoch": 0.26634382566585957,
+      "grad_norm": 0.04081873223185539,
+      "learning_rate": 1.722200641195581e-05,
+      "loss": 0.4514,
+      "step": 330
+    },
+    {
+      "epoch": 0.26715092816787733,
+      "grad_norm": 0.04145602881908417,
+      "learning_rate": 1.7203888567110066e-05,
+      "loss": 0.5665,
+      "step": 331
+    },
+    {
+      "epoch": 0.2679580306698951,
+      "grad_norm": 0.03948275372385979,
+      "learning_rate": 1.7185721429806975e-05,
+      "loss": 0.4196,
+      "step": 332
+    },
+    {
+      "epoch": 0.2687651331719128,
+      "grad_norm": 0.04905907064676285,
+      "learning_rate": 1.7167505124354803e-05,
+      "loss": 0.4886,
+      "step": 333
+    },
+    {
+      "epoch": 0.26957223567393057,
+      "grad_norm": 0.04699491709470749,
+      "learning_rate": 1.714923977539823e-05,
+      "loss": 0.4767,
+      "step": 334
+    },
+    {
+      "epoch": 0.27037933817594834,
+      "grad_norm": 0.04403752461075783,
+      "learning_rate": 1.7130925507917534e-05,
+      "loss": 0.5452,
+      "step": 335
+    },
+    {
+      "epoch": 0.2711864406779661,
+      "grad_norm": 0.03853682428598404,
+      "learning_rate": 1.7112562447227703e-05,
+      "loss": 0.4439,
+      "step": 336
+    },
+    {
+      "epoch": 0.27199354317998387,
+      "grad_norm": 0.040106069296598434,
+      "learning_rate": 1.70941507189776e-05,
+      "loss": 0.4426,
+      "step": 337
+    },
+    {
+      "epoch": 0.27280064568200163,
+      "grad_norm": 0.04464896023273468,
+      "learning_rate": 1.7075690449149092e-05,
+      "loss": 0.4732,
+      "step": 338
+    },
+    {
+      "epoch": 0.2736077481840194,
+      "grad_norm": 0.0417223796248436,
+      "learning_rate": 1.7057181764056188e-05,
+      "loss": 0.4296,
+      "step": 339
+    },
+    {
+      "epoch": 0.2744148506860371,
+      "grad_norm": 0.03915628045797348,
+      "learning_rate": 1.7038624790344185e-05,
+      "loss": 0.4623,
+      "step": 340
+    },
+    {
+      "epoch": 0.27522195318805487,
+      "grad_norm": 0.035377517342567444,
+      "learning_rate": 1.7020019654988784e-05,
+      "loss": 0.3977,
+      "step": 341
+    },
+    {
+      "epoch": 0.27602905569007263,
+      "grad_norm": 0.04356204345822334,
+      "learning_rate": 1.7001366485295235e-05,
+      "loss": 0.4423,
+      "step": 342
+    },
+    {
+      "epoch": 0.2768361581920904,
+      "grad_norm": 0.03903350234031677,
+      "learning_rate": 1.6982665408897468e-05,
+      "loss": 0.4568,
+      "step": 343
+    },
+    {
+      "epoch": 0.27764326069410816,
+      "grad_norm": 0.045728739351034164,
+      "learning_rate": 1.6963916553757204e-05,
+      "loss": 0.4824,
+      "step": 344
+    },
+    {
+      "epoch": 0.2784503631961259,
+      "grad_norm": 0.03853655606508255,
+      "learning_rate": 1.6945120048163092e-05,
+      "loss": 0.4289,
+      "step": 345
+    },
+    {
+      "epoch": 0.27925746569814364,
+      "grad_norm": 0.037414539605379105,
+      "learning_rate": 1.6926276020729836e-05,
+      "loss": 0.4266,
+      "step": 346
+    },
+    {
+      "epoch": 0.2800645682001614,
+      "grad_norm": 0.042327504605054855,
+      "learning_rate": 1.6907384600397295e-05,
+      "loss": 0.472,
+      "step": 347
+    },
+    {
+      "epoch": 0.28087167070217917,
+      "grad_norm": 0.039189938455820084,
+      "learning_rate": 1.6888445916429624e-05,
+      "loss": 0.5246,
+      "step": 348
+    },
+    {
+      "epoch": 0.28167877320419693,
+      "grad_norm": 0.04332740232348442,
+      "learning_rate": 1.686946009841437e-05,
+      "loss": 0.4247,
+      "step": 349
+    },
+    {
+      "epoch": 0.2824858757062147,
+      "grad_norm": 0.03479583561420441,
+      "learning_rate": 1.6850427276261596e-05,
+      "loss": 0.4145,
+      "step": 350
+    },
+    {
+      "epoch": 0.28329297820823246,
+      "grad_norm": 0.03648196533322334,
+      "learning_rate": 1.683134758020299e-05,
+      "loss": 0.4035,
+      "step": 351
+    },
+    {
+      "epoch": 0.2841000807102502,
+      "grad_norm": 0.03672903776168823,
+      "learning_rate": 1.681222114079098e-05,
+      "loss": 0.4395,
+      "step": 352
+    },
+    {
+      "epoch": 0.28490718321226793,
+      "grad_norm": 0.04137750715017319,
+      "learning_rate": 1.6793048088897827e-05,
+      "loss": 0.4235,
+      "step": 353
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.03800482675433159,
+      "learning_rate": 1.677382855571473e-05,
+      "loss": 0.4711,
+      "step": 354
+    },
+    {
+      "epoch": 0.28652138821630346,
+      "grad_norm": 0.03724050521850586,
+      "learning_rate": 1.675456267275096e-05,
+      "loss": 0.474,
+      "step": 355
+    },
+    {
+      "epoch": 0.28732849071832123,
+      "grad_norm": 0.040064744651317596,
+      "learning_rate": 1.6735250571832907e-05,
+      "loss": 0.467,
+      "step": 356
+    },
+    {
+      "epoch": 0.288135593220339,
+      "grad_norm": 0.038956187665462494,
+      "learning_rate": 1.6715892385103223e-05,
+      "loss": 0.4088,
+      "step": 357
+    },
+    {
+      "epoch": 0.28894269572235676,
+      "grad_norm": 0.03888452798128128,
+      "learning_rate": 1.6696488245019898e-05,
+      "loss": 0.4535,
+      "step": 358
+    },
+    {
+      "epoch": 0.2897497982243745,
+      "grad_norm": 0.040293607860803604,
+      "learning_rate": 1.6677038284355358e-05,
+      "loss": 0.4577,
+      "step": 359
+    },
+    {
+      "epoch": 0.29055690072639223,
+      "grad_norm": 0.03922054544091225,
+      "learning_rate": 1.6657542636195546e-05,
+      "loss": 0.4294,
+      "step": 360
+    },
+    {
+      "epoch": 0.29136400322841,
+      "grad_norm": 0.039267491549253464,
+      "learning_rate": 1.6638001433939036e-05,
+      "loss": 0.4304,
+      "step": 361
+    },
+    {
+      "epoch": 0.29217110573042776,
+      "grad_norm": 0.03777645528316498,
+      "learning_rate": 1.6618414811296094e-05,
+      "loss": 0.4243,
+      "step": 362
+    },
+    {
+      "epoch": 0.2929782082324455,
+      "grad_norm": 0.038406968116760254,
+      "learning_rate": 1.6598782902287775e-05,
+      "loss": 0.4831,
+      "step": 363
+    },
+    {
+      "epoch": 0.2937853107344633,
+      "grad_norm": 0.039704568684101105,
+      "learning_rate": 1.657910584124501e-05,
+      "loss": 0.4482,
+      "step": 364
+    },
+    {
+      "epoch": 0.29459241323648105,
+      "grad_norm": 0.04083388298749924,
+      "learning_rate": 1.655938376280768e-05,
+      "loss": 0.4234,
+      "step": 365
+    },
+    {
+      "epoch": 0.29539951573849876,
+      "grad_norm": 0.037917107343673706,
+      "learning_rate": 1.653961680192369e-05,
+      "loss": 0.3928,
+      "step": 366
+    },
+    {
+      "epoch": 0.29620661824051653,
+      "grad_norm": 0.0368562787771225,
+      "learning_rate": 1.6519805093848063e-05,
+      "loss": 0.439,
+      "step": 367
+    },
+    {
+      "epoch": 0.2970137207425343,
+      "grad_norm": 0.040610842406749725,
+      "learning_rate": 1.6499948774142e-05,
+      "loss": 0.4654,
+      "step": 368
+    },
+    {
+      "epoch": 0.29782082324455206,
+      "grad_norm": 0.040698591619729996,
+      "learning_rate": 1.648004797867195e-05,
+      "loss": 0.4465,
+      "step": 369
+    },
+    {
+      "epoch": 0.2986279257465698,
+      "grad_norm": 0.0374905951321125,
+      "learning_rate": 1.6460102843608692e-05,
+      "loss": 0.4665,
+      "step": 370
+    },
+    {
+      "epoch": 0.2994350282485876,
+      "grad_norm": 0.04037656635046005,
+      "learning_rate": 1.6440113505426397e-05,
+      "loss": 0.4234,
+      "step": 371
+    },
+    {
+      "epoch": 0.30024213075060535,
+      "grad_norm": 0.038281265646219254,
+      "learning_rate": 1.6420080100901693e-05,
+      "loss": 0.4879,
+      "step": 372
+    },
+    {
+      "epoch": 0.30104923325262306,
+      "grad_norm": 0.043232664465904236,
+      "learning_rate": 1.640000276711274e-05,
+      "loss": 0.4534,
+      "step": 373
+    },
+    {
+      "epoch": 0.3018563357546408,
+      "grad_norm": 0.039622269570827484,
+      "learning_rate": 1.637988164143827e-05,
+      "loss": 0.4077,
+      "step": 374
+    },
+    {
+      "epoch": 0.3026634382566586,
+      "grad_norm": 0.03671490401029587,
+      "learning_rate": 1.6359716861556668e-05,
+      "loss": 0.4214,
+      "step": 375
+    },
+    {
+      "epoch": 0.30347054075867635,
+      "grad_norm": 0.051524847745895386,
+      "learning_rate": 1.6339508565445016e-05,
+      "loss": 0.4914,
+      "step": 376
+    },
+    {
+      "epoch": 0.3042776432606941,
+      "grad_norm": 0.04071008041501045,
+      "learning_rate": 1.6319256891378166e-05,
+      "loss": 0.4077,
+      "step": 377
+    },
+    {
+      "epoch": 0.3050847457627119,
+      "grad_norm": 0.03799833729863167,
+      "learning_rate": 1.629896197792777e-05,
+      "loss": 0.3997,
+      "step": 378
+    },
+    {
+      "epoch": 0.3058918482647296,
+      "grad_norm": 0.03393981233239174,
+      "learning_rate": 1.627862396396135e-05,
+      "loss": 0.3734,
+      "step": 379
+    },
+    {
+      "epoch": 0.30669895076674736,
+      "grad_norm": 0.04095921292901039,
+      "learning_rate": 1.6258242988641346e-05,
+      "loss": 0.5031,
+      "step": 380
+    },
+    {
+      "epoch": 0.3075060532687651,
+      "grad_norm": 0.04322320595383644,
+      "learning_rate": 1.6237819191424155e-05,
+      "loss": 0.3861,
+      "step": 381
+    },
+    {
+      "epoch": 0.3083131557707829,
+      "grad_norm": 0.039408132433891296,
+      "learning_rate": 1.6217352712059186e-05,
+      "loss": 0.4431,
+      "step": 382
+    },
+    {
+      "epoch": 0.30912025827280065,
+      "grad_norm": 0.04281940311193466,
+      "learning_rate": 1.619684369058789e-05,
+      "loss": 0.4247,
+      "step": 383
+    },
+    {
+      "epoch": 0.3099273607748184,
+      "grad_norm": 0.03588446229696274,
+      "learning_rate": 1.617629226734283e-05,
+      "loss": 0.4107,
+      "step": 384
+    },
+    {
+      "epoch": 0.3107344632768362,
+      "grad_norm": 0.04118945822119713,
+      "learning_rate": 1.6155698582946678e-05,
+      "loss": 0.3958,
+      "step": 385
+    },
+    {
+      "epoch": 0.3115415657788539,
+      "grad_norm": 0.039396982640028,
+      "learning_rate": 1.6135062778311304e-05,
+      "loss": 0.4786,
+      "step": 386
+    },
+    {
+      "epoch": 0.31234866828087166,
+      "grad_norm": 0.03626495227217674,
+      "learning_rate": 1.6114384994636765e-05,
+      "loss": 0.3843,
+      "step": 387
+    },
+    {
+      "epoch": 0.3131557707828894,
+      "grad_norm": 0.04274310916662216,
+      "learning_rate": 1.6093665373410357e-05,
+      "loss": 0.488,
+      "step": 388
+    },
+    {
+      "epoch": 0.3139628732849072,
+      "grad_norm": 0.040091730654239655,
+      "learning_rate": 1.6072904056405673e-05,
+      "loss": 0.4844,
+      "step": 389
+    },
+    {
+      "epoch": 0.31476997578692495,
+      "grad_norm": 0.041715774685144424,
+      "learning_rate": 1.6052101185681576e-05,
+      "loss": 0.457,
+      "step": 390
+    },
+    {
+      "epoch": 0.3155770782889427,
+      "grad_norm": 0.03776617348194122,
+      "learning_rate": 1.6031256903581283e-05,
+      "loss": 0.4419,
+      "step": 391
+    },
+    {
+      "epoch": 0.3163841807909605,
+      "grad_norm": 0.03789287060499191,
+      "learning_rate": 1.601037135273136e-05,
+      "loss": 0.4598,
+      "step": 392
+    },
+    {
+      "epoch": 0.3171912832929782,
+      "grad_norm": 0.036457955837249756,
+      "learning_rate": 1.598944467604075e-05,
+      "loss": 0.3921,
+      "step": 393
+    },
+    {
+      "epoch": 0.31799838579499595,
+      "grad_norm": 0.03638232499361038,
+      "learning_rate": 1.5968477016699802e-05,
+      "loss": 0.4188,
+      "step": 394
+    },
+    {
+      "epoch": 0.3188054882970137,
+      "grad_norm": 0.03561306744813919,
+      "learning_rate": 1.594746851817929e-05,
+      "loss": 0.3882,
+      "step": 395
+    },
+    {
+      "epoch": 0.3196125907990315,
+      "grad_norm": 0.037340566515922546,
+      "learning_rate": 1.592641932422943e-05,
+      "loss": 0.4307,
+      "step": 396
+    },
+    {
+      "epoch": 0.32041969330104925,
+      "grad_norm": 0.037597768008708954,
+      "learning_rate": 1.5905329578878885e-05,
+      "loss": 0.4076,
+      "step": 397
+    },
+    {
+      "epoch": 0.321226795803067,
+      "grad_norm": 0.03577461093664169,
+      "learning_rate": 1.5884199426433804e-05,
+      "loss": 0.404,
+      "step": 398
+    },
+    {
+      "epoch": 0.3220338983050847,
+      "grad_norm": 0.05212230980396271,
+      "learning_rate": 1.5863029011476815e-05,
+      "loss": 0.5017,
+      "step": 399
+    },
+    {
+      "epoch": 0.3228410008071025,
+      "grad_norm": 0.03799346089363098,
+      "learning_rate": 1.584181847886604e-05,
+      "loss": 0.4341,
+      "step": 400
+    },
+    {
+      "epoch": 0.32364810330912025,
+      "grad_norm": 0.042107388377189636,
+      "learning_rate": 1.5820567973734103e-05,
+      "loss": 0.4576,
+      "step": 401
+    },
+    {
+      "epoch": 0.324455205811138,
+      "grad_norm": 0.03835953027009964,
+      "learning_rate": 1.579927764148714e-05,
+      "loss": 0.4717,
+      "step": 402
+    },
+    {
+      "epoch": 0.3252623083131558,
+      "grad_norm": 0.04675079137086868,
+      "learning_rate": 1.5777947627803818e-05,
+      "loss": 0.4853,
+      "step": 403
+    },
+    {
+      "epoch": 0.32606941081517354,
+      "grad_norm": 0.03846751153469086,
+      "learning_rate": 1.57565780786343e-05,
+      "loss": 0.4379,
+      "step": 404
+    },
+    {
+      "epoch": 0.3268765133171913,
+      "grad_norm": 0.04810432344675064,
+      "learning_rate": 1.5735169140199284e-05,
+      "loss": 0.4313,
+      "step": 405
+    },
+    {
+      "epoch": 0.327683615819209,
+      "grad_norm": 0.04163535684347153,
+      "learning_rate": 1.5713720958988985e-05,
+      "loss": 0.4422,
+      "step": 406
+    },
+    {
+      "epoch": 0.3284907183212268,
+      "grad_norm": 0.042189642786979675,
+      "learning_rate": 1.5692233681762137e-05,
+      "loss": 0.4573,
+      "step": 407
+    },
+    {
+      "epoch": 0.32929782082324455,
+      "grad_norm": 0.038153890520334244,
+      "learning_rate": 1.5670707455544983e-05,
+      "loss": 0.4377,
+      "step": 408
+    },
+    {
+      "epoch": 0.3301049233252623,
+      "grad_norm": 0.040447358042001724,
+      "learning_rate": 1.564914242763028e-05,
+      "loss": 0.4143,
+      "step": 409
+    },
+    {
+      "epoch": 0.3309120258272801,
+      "grad_norm": 0.03961403667926788,
+      "learning_rate": 1.562753874557628e-05,
+      "loss": 0.4291,
+      "step": 410
+    },
+    {
+      "epoch": 0.33171912832929784,
+      "grad_norm": 0.04563172534108162,
+      "learning_rate": 1.560589655720573e-05,
+      "loss": 0.4549,
+      "step": 411
+    },
+    {
+      "epoch": 0.33252623083131555,
+      "grad_norm": 0.04365930333733559,
+      "learning_rate": 1.5584216010604852e-05,
+      "loss": 0.4165,
+      "step": 412
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.03692984953522682,
+      "learning_rate": 1.5562497254122333e-05,
+      "loss": 0.3979,
+      "step": 413
+    },
+    {
+      "epoch": 0.3341404358353511,
+      "grad_norm": 0.039555348455905914,
+      "learning_rate": 1.5540740436368313e-05,
+      "loss": 0.4813,
+      "step": 414
+    },
+    {
+      "epoch": 0.33494753833736884,
+      "grad_norm": 0.03751777485013008,
+      "learning_rate": 1.551894570621336e-05,
+      "loss": 0.4167,
+      "step": 415
+    },
+    {
+      "epoch": 0.3357546408393866,
+      "grad_norm": 0.03755151107907295,
+      "learning_rate": 1.5497113212787463e-05,
+      "loss": 0.4185,
+      "step": 416
+    },
+    {
+      "epoch": 0.3365617433414044,
+      "grad_norm": 0.0399174690246582,
+      "learning_rate": 1.5475243105479002e-05,
+      "loss": 0.3922,
+      "step": 417
+    },
+    {
+      "epoch": 0.33736884584342214,
+      "grad_norm": 0.03419012576341629,
+      "learning_rate": 1.5453335533933733e-05,
+      "loss": 0.4386,
+      "step": 418
+    },
+    {
+      "epoch": 0.33817594834543985,
+      "grad_norm": 0.04388118535280228,
+      "learning_rate": 1.543139064805376e-05,
+      "loss": 0.5092,
+      "step": 419
+    },
+    {
+      "epoch": 0.3389830508474576,
+      "grad_norm": 0.04407074674963951,
+      "learning_rate": 1.54094085979965e-05,
+      "loss": 0.4565,
+      "step": 420
+    },
+    {
+      "epoch": 0.3397901533494754,
+      "grad_norm": 0.04343196749687195,
+      "learning_rate": 1.538738953417368e-05,
+      "loss": 0.4412,
+      "step": 421
+    },
+    {
+      "epoch": 0.34059725585149314,
+      "grad_norm": 0.03644689917564392,
+      "learning_rate": 1.5365333607250277e-05,
+      "loss": 0.4272,
+      "step": 422
+    },
+    {
+      "epoch": 0.3414043583535109,
+      "grad_norm": 0.04378410428762436,
+      "learning_rate": 1.5343240968143516e-05,
+      "loss": 0.4221,
+      "step": 423
+    },
+    {
+      "epoch": 0.34221146085552867,
+      "grad_norm": 0.03766017034649849,
+      "learning_rate": 1.532111176802182e-05,
+      "loss": 0.4441,
+      "step": 424
+    },
+    {
+      "epoch": 0.34301856335754644,
+      "grad_norm": 0.03888798505067825,
+      "learning_rate": 1.5298946158303776e-05,
+      "loss": 0.4014,
+      "step": 425
+    },
+    {
+      "epoch": 0.34382566585956414,
+      "grad_norm": 0.042303815484046936,
+      "learning_rate": 1.527674429065711e-05,
+      "loss": 0.4548,
+      "step": 426
+    },
+    {
+      "epoch": 0.3446327683615819,
+      "grad_norm": 0.040103986859321594,
+      "learning_rate": 1.5254506316997638e-05,
+      "loss": 0.4539,
+      "step": 427
+    },
+    {
+      "epoch": 0.3454398708635997,
+      "grad_norm": 0.04512703791260719,
+      "learning_rate": 1.5232232389488228e-05,
+      "loss": 0.4667,
+      "step": 428
+    },
+    {
+      "epoch": 0.34624697336561744,
+      "grad_norm": 0.0437651164829731,
+      "learning_rate": 1.520992266053777e-05,
+      "loss": 0.4823,
+      "step": 429
+    },
+    {
+      "epoch": 0.3470540758676352,
+      "grad_norm": 0.044296592473983765,
+      "learning_rate": 1.5187577282800114e-05,
+      "loss": 0.5237,
+      "step": 430
+    },
+    {
+      "epoch": 0.34786117836965297,
+      "grad_norm": 0.048969656229019165,
+      "learning_rate": 1.5165196409173044e-05,
+      "loss": 0.4724,
+      "step": 431
+    },
+    {
+      "epoch": 0.3486682808716707,
+      "grad_norm": 0.04471556469798088,
+      "learning_rate": 1.5142780192797222e-05,
+      "loss": 0.4207,
+      "step": 432
+    },
+    {
+      "epoch": 0.34947538337368844,
+      "grad_norm": 0.04149989038705826,
+      "learning_rate": 1.512032878705514e-05,
+      "loss": 0.483,
+      "step": 433
+    },
+    {
+      "epoch": 0.3502824858757062,
+      "grad_norm": 0.041714008897542953,
+      "learning_rate": 1.5097842345570075e-05,
+      "loss": 0.4601,
+      "step": 434
+    },
+    {
+      "epoch": 0.35108958837772397,
+      "grad_norm": 0.03949635103344917,
+      "learning_rate": 1.5075321022205032e-05,
+      "loss": 0.3904,
+      "step": 435
+    },
+    {
+      "epoch": 0.35189669087974174,
+      "grad_norm": 0.03717697039246559,
+      "learning_rate": 1.5052764971061696e-05,
+      "loss": 0.3682,
+      "step": 436
+    },
+    {
+      "epoch": 0.3527037933817595,
+      "grad_norm": 0.042058829218149185,
+      "learning_rate": 1.503017434647938e-05,
+      "loss": 0.4627,
+      "step": 437
+    },
+    {
+      "epoch": 0.35351089588377727,
+      "grad_norm": 0.03921369090676308,
+      "learning_rate": 1.5007549303033959e-05,
+      "loss": 0.4001,
+      "step": 438
+    },
+    {
+      "epoch": 0.354317998385795,
+      "grad_norm": 0.06680601090192795,
+      "learning_rate": 1.4984889995536825e-05,
+      "loss": 0.4442,
+      "step": 439
+    },
+    {
+      "epoch": 0.35512510088781274,
+      "grad_norm": 0.048763904720544815,
+      "learning_rate": 1.4962196579033813e-05,
+      "loss": 0.4623,
+      "step": 440
+    },
+    {
+      "epoch": 0.3559322033898305,
+      "grad_norm": 0.042305104434490204,
+      "learning_rate": 1.4939469208804152e-05,
+      "loss": 0.4807,
+      "step": 441
+    },
+    {
+      "epoch": 0.35673930589184827,
+      "grad_norm": 0.039217207580804825,
+      "learning_rate": 1.4916708040359403e-05,
+      "loss": 0.4262,
+      "step": 442
+    },
+    {
+      "epoch": 0.35754640839386603,
+      "grad_norm": 0.04164797067642212,
+      "learning_rate": 1.4893913229442383e-05,
+      "loss": 0.4513,
+      "step": 443
+    },
+    {
+      "epoch": 0.3583535108958838,
+      "grad_norm": 0.04174966737627983,
+      "learning_rate": 1.4871084932026114e-05,
+      "loss": 0.4474,
+      "step": 444
+    },
+    {
+      "epoch": 0.3591606133979015,
+      "grad_norm": 0.04466141760349274,
+      "learning_rate": 1.4848223304312738e-05,
+      "loss": 0.4859,
+      "step": 445
+    },
+    {
+      "epoch": 0.35996771589991927,
+      "grad_norm": 0.04579891636967659,
+      "learning_rate": 1.4825328502732477e-05,
+      "loss": 0.4946,
+      "step": 446
+    },
+    {
+      "epoch": 0.36077481840193704,
+      "grad_norm": 0.04515714570879936,
+      "learning_rate": 1.4802400683942526e-05,
+      "loss": 0.3916,
+      "step": 447
+    },
+    {
+      "epoch": 0.3615819209039548,
+      "grad_norm": 0.0440857857465744,
+      "learning_rate": 1.4779440004826006e-05,
+      "loss": 0.4027,
+      "step": 448
+    },
+    {
+      "epoch": 0.36238902340597257,
+      "grad_norm": 0.044258441776037216,
+      "learning_rate": 1.4756446622490894e-05,
+      "loss": 0.4781,
+      "step": 449
+    },
+    {
+      "epoch": 0.36319612590799033,
+      "grad_norm": 0.0432918481528759,
+      "learning_rate": 1.4733420694268917e-05,
+      "loss": 0.3359,
+      "step": 450
+    },
+    {
+      "epoch": 0.3640032284100081,
+      "grad_norm": 0.03601154685020447,
+      "learning_rate": 1.4710362377714516e-05,
+      "loss": 0.3822,
+      "step": 451
+    },
+    {
+      "epoch": 0.3648103309120258,
+      "grad_norm": 0.04147439822554588,
+      "learning_rate": 1.4687271830603744e-05,
+      "loss": 0.4054,
+      "step": 452
+    },
+    {
+      "epoch": 0.36561743341404357,
+      "grad_norm": 0.03952731937170029,
+      "learning_rate": 1.466414921093318e-05,
+      "loss": 0.4001,
+      "step": 453
+    },
+    {
+      "epoch": 0.36642453591606133,
+      "grad_norm": 0.04244634881615639,
+      "learning_rate": 1.4640994676918873e-05,
+      "loss": 0.4463,
+      "step": 454
+    },
+    {
+      "epoch": 0.3672316384180791,
+      "grad_norm": 0.03925754129886627,
+      "learning_rate": 1.4617808386995243e-05,
+      "loss": 0.3938,
+      "step": 455
+    },
+    {
+      "epoch": 0.36803874092009686,
+      "grad_norm": 0.03985142707824707,
+      "learning_rate": 1.4594590499813985e-05,
+      "loss": 0.4186,
+      "step": 456
+    },
+    {
+      "epoch": 0.3688458434221146,
+      "grad_norm": 0.04152598977088928,
+      "learning_rate": 1.4571341174243017e-05,
+      "loss": 0.4217,
+      "step": 457
+    },
+    {
+      "epoch": 0.3696529459241324,
+      "grad_norm": 0.043532952666282654,
+      "learning_rate": 1.4548060569365358e-05,
+      "loss": 0.4714,
+      "step": 458
+    },
+    {
+      "epoch": 0.3704600484261501,
+      "grad_norm": 0.04257139936089516,
+      "learning_rate": 1.4524748844478065e-05,
+      "loss": 0.4244,
+      "step": 459
+    },
+    {
+      "epoch": 0.37126715092816787,
+      "grad_norm": 0.04078930243849754,
+      "learning_rate": 1.4501406159091132e-05,
+      "loss": 0.3825,
+      "step": 460
+    },
+    {
+      "epoch": 0.37207425343018563,
+      "grad_norm": 0.04355788230895996,
+      "learning_rate": 1.4478032672926389e-05,
+      "loss": 0.4495,
+      "step": 461
+    },
+    {
+      "epoch": 0.3728813559322034,
+      "grad_norm": 0.04207145422697067,
+      "learning_rate": 1.445462854591644e-05,
+      "loss": 0.4162,
+      "step": 462
+    },
+    {
+      "epoch": 0.37368845843422116,
+      "grad_norm": 0.04463491588830948,
+      "learning_rate": 1.4431193938203524e-05,
+      "loss": 0.4389,
+      "step": 463
+    },
+    {
+      "epoch": 0.3744955609362389,
+      "grad_norm": 0.045574747025966644,
+      "learning_rate": 1.4407729010138463e-05,
+      "loss": 0.4531,
+      "step": 464
+    },
+    {
+      "epoch": 0.37530266343825663,
+      "grad_norm": 0.0455825999379158,
+      "learning_rate": 1.4384233922279538e-05,
+      "loss": 0.4312,
+      "step": 465
+    },
+    {
+      "epoch": 0.3761097659402744,
+      "grad_norm": 0.04084734618663788,
+      "learning_rate": 1.4360708835391392e-05,
+      "loss": 0.4295,
+      "step": 466
+    },
+    {
+      "epoch": 0.37691686844229216,
+      "grad_norm": 0.04858178272843361,
+      "learning_rate": 1.433715391044395e-05,
+      "loss": 0.519,
+      "step": 467
+    },
+    {
+      "epoch": 0.37772397094430993,
+      "grad_norm": 0.03454524278640747,
+      "learning_rate": 1.4313569308611289e-05,
+      "loss": 0.3733,
+      "step": 468
+    },
+    {
+      "epoch": 0.3785310734463277,
+      "grad_norm": 0.03903524577617645,
+      "learning_rate": 1.4289955191270555e-05,
+      "loss": 0.3699,
+      "step": 469
+    },
+    {
+      "epoch": 0.37933817594834546,
+      "grad_norm": 0.05262704938650131,
+      "learning_rate": 1.4266311720000857e-05,
+      "loss": 0.5171,
+      "step": 470
+    },
+    {
+      "epoch": 0.3801452784503632,
+      "grad_norm": 0.0390932559967041,
+      "learning_rate": 1.4242639056582155e-05,
+      "loss": 0.3953,
+      "step": 471
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 0.04563722386956215,
+      "learning_rate": 1.4218937362994152e-05,
+      "loss": 0.502,
+      "step": 472
+    },
+    {
+      "epoch": 0.3817594834543987,
+      "grad_norm": 0.03626189008355141,
+      "learning_rate": 1.419520680141519e-05,
+      "loss": 0.4058,
+      "step": 473
+    },
+    {
+      "epoch": 0.38256658595641646,
+      "grad_norm": 0.04210507124662399,
+      "learning_rate": 1.4171447534221147e-05,
+      "loss": 0.4653,
+      "step": 474
+    },
+    {
+      "epoch": 0.3833736884584342,
+      "grad_norm": 0.046014368534088135,
+      "learning_rate": 1.4147659723984313e-05,
+      "loss": 0.4791,
+      "step": 475
+    },
+    {
+      "epoch": 0.384180790960452,
+      "grad_norm": 0.041533246636390686,
+      "learning_rate": 1.412384353347228e-05,
+      "loss": 0.402,
+      "step": 476
+    },
+    {
+      "epoch": 0.38498789346246975,
+      "grad_norm": 0.041039034724235535,
+      "learning_rate": 1.4099999125646839e-05,
+      "loss": 0.4235,
+      "step": 477
+    },
+    {
+      "epoch": 0.38579499596448746,
+      "grad_norm": 0.04687739536166191,
+      "learning_rate": 1.407612666366285e-05,
+      "loss": 0.4187,
+      "step": 478
+    },
+    {
+      "epoch": 0.38660209846650523,
+      "grad_norm": 0.04226434975862503,
+      "learning_rate": 1.4052226310867138e-05,
+      "loss": 0.437,
+      "step": 479
+    },
+    {
+      "epoch": 0.387409200968523,
+      "grad_norm": 0.03916895389556885,
+      "learning_rate": 1.4028298230797371e-05,
+      "loss": 0.3641,
+      "step": 480
+    },
+    {
+      "epoch": 0.38821630347054076,
+      "grad_norm": 0.03749072924256325,
+      "learning_rate": 1.4004342587180934e-05,
+      "loss": 0.4067,
+      "step": 481
+    },
+    {
+      "epoch": 0.3890234059725585,
+      "grad_norm": 0.03737153112888336,
+      "learning_rate": 1.3980359543933822e-05,
+      "loss": 0.3473,
+      "step": 482
+    },
+    {
+      "epoch": 0.3898305084745763,
+      "grad_norm": 0.04256528615951538,
+      "learning_rate": 1.3956349265159508e-05,
+      "loss": 0.3997,
+      "step": 483
+    },
+    {
+      "epoch": 0.39063761097659405,
+      "grad_norm": 0.0409972257912159,
+      "learning_rate": 1.3932311915147822e-05,
+      "loss": 0.4516,
+      "step": 484
+    },
+    {
+      "epoch": 0.39144471347861176,
+      "grad_norm": 0.04119865968823433,
+      "learning_rate": 1.3908247658373835e-05,
+      "loss": 0.4282,
+      "step": 485
+    },
+    {
+      "epoch": 0.3922518159806295,
+      "grad_norm": 0.041595458984375,
+      "learning_rate": 1.3884156659496715e-05,
+      "loss": 0.4043,
+      "step": 486
+    },
+    {
+      "epoch": 0.3930589184826473,
+      "grad_norm": 0.035765454173088074,
+      "learning_rate": 1.3860039083358628e-05,
+      "loss": 0.4139,
+      "step": 487
+    },
+    {
+      "epoch": 0.39386602098466506,
+      "grad_norm": 0.03818071261048317,
+      "learning_rate": 1.3835895094983578e-05,
+      "loss": 0.3724,
+      "step": 488
+    },
+    {
+      "epoch": 0.3946731234866828,
+      "grad_norm": 0.03883254528045654,
+      "learning_rate": 1.3811724859576305e-05,
+      "loss": 0.446,
+      "step": 489
+    },
+    {
+      "epoch": 0.3954802259887006,
+      "grad_norm": 0.044025968760252,
+      "learning_rate": 1.3787528542521145e-05,
+      "loss": 0.4644,
+      "step": 490
+    },
+    {
+      "epoch": 0.3962873284907183,
+      "grad_norm": 0.03932966664433479,
+      "learning_rate": 1.376330630938089e-05,
+      "loss": 0.4721,
+      "step": 491
+    },
+    {
+      "epoch": 0.39709443099273606,
+      "grad_norm": 0.039546482264995575,
+      "learning_rate": 1.3739058325895663e-05,
+      "loss": 0.4095,
+      "step": 492
+    },
+    {
+      "epoch": 0.3979015334947538,
+      "grad_norm": 0.0403420627117157,
+      "learning_rate": 1.371478475798179e-05,
+      "loss": 0.4073,
+      "step": 493
+    },
+    {
+      "epoch": 0.3987086359967716,
+      "grad_norm": 0.042821839451789856,
+      "learning_rate": 1.3690485771730649e-05,
+      "loss": 0.5032,
+      "step": 494
+    },
+    {
+      "epoch": 0.39951573849878935,
+      "grad_norm": 0.04345512017607689,
+      "learning_rate": 1.3666161533407551e-05,
+      "loss": 0.4399,
+      "step": 495
+    },
+    {
+      "epoch": 0.4003228410008071,
+      "grad_norm": 0.04715168476104736,
+      "learning_rate": 1.3641812209450585e-05,
+      "loss": 0.4391,
+      "step": 496
+    },
+    {
+      "epoch": 0.4011299435028249,
+      "grad_norm": 0.04834922403097153,
+      "learning_rate": 1.3617437966469496e-05,
+      "loss": 0.4231,
+      "step": 497
+    },
+    {
+      "epoch": 0.4019370460048426,
+      "grad_norm": 0.04127978906035423,
+      "learning_rate": 1.3593038971244534e-05,
+      "loss": 0.4376,
+      "step": 498
+    },
+    {
+      "epoch": 0.40274414850686036,
+      "grad_norm": 0.049879882484674454,
+      "learning_rate": 1.3568615390725311e-05,
+      "loss": 0.4751,
+      "step": 499
+    },
+    {
+      "epoch": 0.4035512510088781,
+      "grad_norm": 0.04311080649495125,
+      "learning_rate": 1.3544167392029675e-05,
+      "loss": 0.4184,
+      "step": 500
+    },
+    {
+      "epoch": 0.4043583535108959,
+      "grad_norm": 0.04278581589460373,
+      "learning_rate": 1.3519695142442539e-05,
+      "loss": 0.3598,
+      "step": 501
+    },
+    {
+      "epoch": 0.40516545601291365,
+      "grad_norm": 0.03896728530526161,
+      "learning_rate": 1.3495198809414764e-05,
+      "loss": 0.3933,
+      "step": 502
+    },
+    {
+      "epoch": 0.4059725585149314,
+      "grad_norm": 0.03846316412091255,
+      "learning_rate": 1.3470678560562003e-05,
+      "loss": 0.3793,
+      "step": 503
+    },
+    {
+      "epoch": 0.4067796610169492,
+      "grad_norm": 0.04504529759287834,
+      "learning_rate": 1.344613456366354e-05,
+      "loss": 0.4258,
+      "step": 504
+    },
+    {
+      "epoch": 0.4075867635189669,
+      "grad_norm": 0.04523047059774399,
+      "learning_rate": 1.3421566986661166e-05,
+      "loss": 0.382,
+      "step": 505
+    },
+    {
+      "epoch": 0.40839386602098465,
+      "grad_norm": 0.0470433384180069,
+      "learning_rate": 1.3396975997658013e-05,
+      "loss": 0.4093,
+      "step": 506
+    },
+    {
+      "epoch": 0.4092009685230024,
+      "grad_norm": 0.041309718042612076,
+      "learning_rate": 1.337236176491741e-05,
+      "loss": 0.4026,
+      "step": 507
+    },
+    {
+      "epoch": 0.4100080710250202,
+      "grad_norm": 0.04193636029958725,
+      "learning_rate": 1.3347724456861735e-05,
+      "loss": 0.384,
+      "step": 508
+    },
+    {
+      "epoch": 0.41081517352703795,
+      "grad_norm": 0.04549890384078026,
+      "learning_rate": 1.3323064242071248e-05,
+      "loss": 0.4228,
+      "step": 509
+    },
+    {
+      "epoch": 0.4116222760290557,
+      "grad_norm": 0.0428113155066967,
+      "learning_rate": 1.3298381289282964e-05,
+      "loss": 0.4309,
+      "step": 510
+    },
+    {
+      "epoch": 0.4124293785310734,
+      "grad_norm": 0.04721170291304588,
+      "learning_rate": 1.327367576738947e-05,
+      "loss": 0.41,
+      "step": 511
+    },
+    {
+      "epoch": 0.4132364810330912,
+      "grad_norm": 0.047412823885679245,
+      "learning_rate": 1.3248947845437776e-05,
+      "loss": 0.5009,
+      "step": 512
+    },
+    {
+      "epoch": 0.41404358353510895,
+      "grad_norm": 0.04840272292494774,
+      "learning_rate": 1.3224197692628185e-05,
+      "loss": 0.4325,
+      "step": 513
+    },
+    {
+      "epoch": 0.4148506860371267,
+      "grad_norm": 0.04272138699889183,
+      "learning_rate": 1.3199425478313093e-05,
+      "loss": 0.3978,
+      "step": 514
+    },
+    {
+      "epoch": 0.4156577885391445,
+      "grad_norm": 0.04556628689169884,
+      "learning_rate": 1.3174631371995861e-05,
+      "loss": 0.447,
+      "step": 515
+    },
+    {
+      "epoch": 0.41646489104116224,
+      "grad_norm": 0.040329255163669586,
+      "learning_rate": 1.3149815543329652e-05,
+      "loss": 0.4256,
+      "step": 516
+    },
+    {
+      "epoch": 0.41727199354318,
+      "grad_norm": 0.039529670029878616,
+      "learning_rate": 1.3124978162116245e-05,
+      "loss": 0.4201,
+      "step": 517
+    },
+    {
+      "epoch": 0.4180790960451977,
+      "grad_norm": 0.03994109109044075,
+      "learning_rate": 1.3100119398304911e-05,
+      "loss": 0.3687,
+      "step": 518
+    },
+    {
+      "epoch": 0.4188861985472155,
+      "grad_norm": 0.04244811832904816,
+      "learning_rate": 1.3075239421991221e-05,
+      "loss": 0.4158,
+      "step": 519
+    },
+    {
+      "epoch": 0.41969330104923325,
+      "grad_norm": 0.03751133754849434,
+      "learning_rate": 1.3050338403415892e-05,
+      "loss": 0.4107,
+      "step": 520
+    },
+    {
+      "epoch": 0.420500403551251,
+      "grad_norm": 0.043357253074645996,
+      "learning_rate": 1.3025416512963628e-05,
+      "loss": 0.4352,
+      "step": 521
+    },
+    {
+      "epoch": 0.4213075060532688,
+      "grad_norm": 0.038147930055856705,
+      "learning_rate": 1.3000473921161947e-05,
+      "loss": 0.3219,
+      "step": 522
+    },
+    {
+      "epoch": 0.42211460855528654,
+      "grad_norm": 0.04303905740380287,
+      "learning_rate": 1.297551079868001e-05,
+      "loss": 0.5032,
+      "step": 523
+    },
+    {
+      "epoch": 0.42292171105730425,
+      "grad_norm": 0.042353641241788864,
+      "learning_rate": 1.2950527316327457e-05,
+      "loss": 0.4332,
+      "step": 524
+    },
+    {
+      "epoch": 0.423728813559322,
+      "grad_norm": 0.04321341589093208,
+      "learning_rate": 1.2925523645053253e-05,
+      "loss": 0.4201,
+      "step": 525
+    },
+    {
+      "epoch": 0.4245359160613398,
+      "grad_norm": 0.04340353608131409,
+      "learning_rate": 1.2900499955944495e-05,
+      "loss": 0.4224,
+      "step": 526
+    },
+    {
+      "epoch": 0.42534301856335754,
+      "grad_norm": 0.04402042180299759,
+      "learning_rate": 1.2875456420225251e-05,
+      "loss": 0.4612,
+      "step": 527
+    },
+    {
+      "epoch": 0.4261501210653753,
+      "grad_norm": 0.0438569076359272,
+      "learning_rate": 1.28503932092554e-05,
+      "loss": 0.3924,
+      "step": 528
+    },
+    {
+      "epoch": 0.4269572235673931,
+      "grad_norm": 0.042652543634176254,
+      "learning_rate": 1.2825310494529435e-05,
+      "loss": 0.4295,
+      "step": 529
+    },
+    {
+      "epoch": 0.42776432606941084,
+      "grad_norm": 0.03889980539679527,
+      "learning_rate": 1.2800208447675307e-05,
+      "loss": 0.4072,
+      "step": 530
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.039839811623096466,
+      "learning_rate": 1.2775087240453252e-05,
+      "loss": 0.4294,
+      "step": 531
+    },
+    {
+      "epoch": 0.4293785310734463,
+      "grad_norm": 0.042306624352931976,
+      "learning_rate": 1.2749947044754605e-05,
+      "loss": 0.3922,
+      "step": 532
+    },
+    {
+      "epoch": 0.4301856335754641,
+      "grad_norm": 0.045233551412820816,
+      "learning_rate": 1.272478803260063e-05,
+      "loss": 0.3839,
+      "step": 533
+    },
+    {
+      "epoch": 0.43099273607748184,
+      "grad_norm": 0.04435190185904503,
+      "learning_rate": 1.2699610376141345e-05,
+      "loss": 0.4093,
+      "step": 534
+    },
+    {
+      "epoch": 0.4317998385794996,
+      "grad_norm": 0.04015007242560387,
+      "learning_rate": 1.2674414247654338e-05,
+      "loss": 0.3753,
+      "step": 535
+    },
+    {
+      "epoch": 0.43260694108151737,
+      "grad_norm": 0.04248462989926338,
+      "learning_rate": 1.2649199819543595e-05,
+      "loss": 0.416,
+      "step": 536
+    },
+    {
+      "epoch": 0.43341404358353514,
+      "grad_norm": 0.04055071994662285,
+      "learning_rate": 1.2623967264338315e-05,
+      "loss": 0.3541,
+      "step": 537
+    },
+    {
+      "epoch": 0.43422114608555284,
+      "grad_norm": 0.04720915108919144,
+      "learning_rate": 1.2598716754691728e-05,
+      "loss": 0.4218,
+      "step": 538
+    },
+    {
+      "epoch": 0.4350282485875706,
+      "grad_norm": 0.0534139946103096,
+      "learning_rate": 1.2573448463379923e-05,
+      "loss": 0.42,
+      "step": 539
+    },
+    {
+      "epoch": 0.4358353510895884,
+      "grad_norm": 0.04554564505815506,
+      "learning_rate": 1.2548162563300653e-05,
+      "loss": 0.4565,
+      "step": 540
+    },
+    {
+      "epoch": 0.43664245359160614,
+      "grad_norm": 0.04438880831003189,
+      "learning_rate": 1.2522859227472165e-05,
+      "loss": 0.4409,
+      "step": 541
+    },
+    {
+      "epoch": 0.4374495560936239,
+      "grad_norm": 0.046392470598220825,
+      "learning_rate": 1.2497538629032004e-05,
+      "loss": 0.3867,
+      "step": 542
+    },
+    {
+      "epoch": 0.43825665859564167,
+      "grad_norm": 0.045731160789728165,
+      "learning_rate": 1.2472200941235839e-05,
+      "loss": 0.428,
+      "step": 543
+    },
+    {
+      "epoch": 0.4390637610976594,
+      "grad_norm": 0.04255305975675583,
+      "learning_rate": 1.2446846337456273e-05,
+      "loss": 0.4047,
+      "step": 544
+    },
+    {
+      "epoch": 0.43987086359967714,
+      "grad_norm": 0.03939209133386612,
+      "learning_rate": 1.242147499118165e-05,
+      "loss": 0.3873,
+      "step": 545
+    },
+    {
+      "epoch": 0.4406779661016949,
+      "grad_norm": 0.045134663581848145,
+      "learning_rate": 1.2396087076014882e-05,
+      "loss": 0.4079,
+      "step": 546
+    },
+    {
+      "epoch": 0.44148506860371267,
+      "grad_norm": 0.04972568526864052,
+      "learning_rate": 1.2370682765672242e-05,
+      "loss": 0.418,
+      "step": 547
+    },
+    {
+      "epoch": 0.44229217110573044,
+      "grad_norm": 0.041813429445028305,
+      "learning_rate": 1.23452622339822e-05,
+      "loss": 0.388,
+      "step": 548
+    },
+    {
+      "epoch": 0.4430992736077482,
+      "grad_norm": 0.040732573717832565,
+      "learning_rate": 1.2319825654884215e-05,
+      "loss": 0.4179,
+      "step": 549
+    },
+    {
+      "epoch": 0.44390637610976597,
+      "grad_norm": 0.047029539942741394,
+      "learning_rate": 1.2294373202427545e-05,
+      "loss": 0.4755,
+      "step": 550
+    },
+    {
+      "epoch": 0.4447134786117837,
+      "grad_norm": 0.042257849127054214,
+      "learning_rate": 1.226890505077007e-05,
+      "loss": 0.411,
+      "step": 551
+    },
+    {
+      "epoch": 0.44552058111380144,
+      "grad_norm": 0.042869675904512405,
+      "learning_rate": 1.224342137417708e-05,
+      "loss": 0.4055,
+      "step": 552
+    },
+    {
+      "epoch": 0.4463276836158192,
+      "grad_norm": 0.049996279180049896,
+      "learning_rate": 1.2217922347020104e-05,
+      "loss": 0.4589,
+      "step": 553
+    },
+    {
+      "epoch": 0.44713478611783697,
+      "grad_norm": 0.045039620250463486,
+      "learning_rate": 1.2192408143775708e-05,
+      "loss": 0.3789,
+      "step": 554
+    },
+    {
+      "epoch": 0.44794188861985473,
+      "grad_norm": 0.045354608446359634,
+      "learning_rate": 1.2166878939024292e-05,
+      "loss": 0.4314,
+      "step": 555
+    },
+    {
+      "epoch": 0.4487489911218725,
+      "grad_norm": 0.043889421969652176,
+      "learning_rate": 1.2141334907448907e-05,
+      "loss": 0.4507,
+      "step": 556
+    },
+    {
+      "epoch": 0.4495560936238902,
+      "grad_norm": 0.0434400737285614,
+      "learning_rate": 1.2115776223834056e-05,
+      "loss": 0.3997,
+      "step": 557
+    },
+    {
+      "epoch": 0.45036319612590797,
+      "grad_norm": 0.04490221291780472,
+      "learning_rate": 1.2090203063064503e-05,
+      "loss": 0.4415,
+      "step": 558
+    },
+    {
+      "epoch": 0.45117029862792574,
+      "grad_norm": 0.0410829521715641,
+      "learning_rate": 1.2064615600124064e-05,
+      "loss": 0.4138,
+      "step": 559
+    },
+    {
+      "epoch": 0.4519774011299435,
+      "grad_norm": 0.04448908194899559,
+      "learning_rate": 1.2039014010094418e-05,
+      "loss": 0.3875,
+      "step": 560
+    },
+    {
+      "epoch": 0.45278450363196127,
+      "grad_norm": 0.04140767827630043,
+      "learning_rate": 1.2013398468153921e-05,
+      "loss": 0.3939,
+      "step": 561
+    },
+    {
+      "epoch": 0.45359160613397903,
+      "grad_norm": 0.04112355411052704,
+      "learning_rate": 1.1987769149576372e-05,
+      "loss": 0.3972,
+      "step": 562
+    },
+    {
+      "epoch": 0.4543987086359968,
+      "grad_norm": 0.04010258987545967,
+      "learning_rate": 1.1962126229729856e-05,
+      "loss": 0.4379,
+      "step": 563
+    },
+    {
+      "epoch": 0.4552058111380145,
+      "grad_norm": 0.04502004757523537,
+      "learning_rate": 1.193646988407552e-05,
+      "loss": 0.3902,
+      "step": 564
+    },
+    {
+      "epoch": 0.45601291364003227,
+      "grad_norm": 0.04058624431490898,
+      "learning_rate": 1.1910800288166366e-05,
+      "loss": 0.3886,
+      "step": 565
+    },
+    {
+      "epoch": 0.45682001614205003,
+      "grad_norm": 0.04509864002466202,
+      "learning_rate": 1.1885117617646078e-05,
+      "loss": 0.4397,
+      "step": 566
+    },
+    {
+      "epoch": 0.4576271186440678,
+      "grad_norm": 0.04298222064971924,
+      "learning_rate": 1.1859422048247788e-05,
+      "loss": 0.4143,
+      "step": 567
+    },
+    {
+      "epoch": 0.45843422114608556,
+      "grad_norm": 0.04298047721385956,
+      "learning_rate": 1.1833713755792895e-05,
+      "loss": 0.459,
+      "step": 568
+    },
+    {
+      "epoch": 0.45924132364810333,
+      "grad_norm": 0.046091094613075256,
+      "learning_rate": 1.1807992916189856e-05,
+      "loss": 0.4352,
+      "step": 569
+    },
+    {
+      "epoch": 0.4600484261501211,
+      "grad_norm": 0.0470380075275898,
+      "learning_rate": 1.178225970543298e-05,
+      "loss": 0.4522,
+      "step": 570
+    },
+    {
+      "epoch": 0.4608555286521388,
+      "grad_norm": 0.04680854454636574,
+      "learning_rate": 1.1756514299601218e-05,
+      "loss": 0.4675,
+      "step": 571
+    },
+    {
+      "epoch": 0.46166263115415657,
+      "grad_norm": 0.04647412523627281,
+      "learning_rate": 1.1730756874856983e-05,
+      "loss": 0.4437,
+      "step": 572
+    },
+    {
+      "epoch": 0.46246973365617433,
+      "grad_norm": 0.041261084377765656,
+      "learning_rate": 1.1704987607444911e-05,
+      "loss": 0.4302,
+      "step": 573
+    },
+    {
+      "epoch": 0.4632768361581921,
+      "grad_norm": 0.04350990429520607,
+      "learning_rate": 1.167920667369068e-05,
+      "loss": 0.3811,
+      "step": 574
+    },
+    {
+      "epoch": 0.46408393866020986,
+      "grad_norm": 0.045072976499795914,
+      "learning_rate": 1.1653414249999786e-05,
+      "loss": 0.3791,
+      "step": 575
+    },
+    {
+      "epoch": 0.4648910411622276,
+      "grad_norm": 0.039517756551504135,
+      "learning_rate": 1.1627610512856355e-05,
+      "loss": 0.3528,
+      "step": 576
+    },
+    {
+      "epoch": 0.46569814366424533,
+      "grad_norm": 0.0447382777929306,
+      "learning_rate": 1.1601795638821923e-05,
+      "loss": 0.408,
+      "step": 577
+    },
+    {
+      "epoch": 0.4665052461662631,
+      "grad_norm": 0.041687775403261185,
+      "learning_rate": 1.1575969804534222e-05,
+      "loss": 0.3981,
+      "step": 578
+    },
+    {
+      "epoch": 0.46731234866828086,
+      "grad_norm": 0.041715968400239944,
+      "learning_rate": 1.1550133186705987e-05,
+      "loss": 0.3914,
+      "step": 579
+    },
+    {
+      "epoch": 0.46811945117029863,
+      "grad_norm": 0.04036974161863327,
+      "learning_rate": 1.1524285962123738e-05,
+      "loss": 0.3387,
+      "step": 580
+    },
+    {
+      "epoch": 0.4689265536723164,
+      "grad_norm": 0.04490208625793457,
+      "learning_rate": 1.1498428307646568e-05,
+      "loss": 0.4202,
+      "step": 581
+    },
+    {
+      "epoch": 0.46973365617433416,
+      "grad_norm": 0.041817329823970795,
+      "learning_rate": 1.1472560400204943e-05,
+      "loss": 0.3995,
+      "step": 582
+    },
+    {
+      "epoch": 0.4705407586763519,
+      "grad_norm": 0.042354173958301544,
+      "learning_rate": 1.1446682416799474e-05,
+      "loss": 0.4135,
+      "step": 583
+    },
+    {
+      "epoch": 0.47134786117836963,
+      "grad_norm": 0.04079662635922432,
+      "learning_rate": 1.1420794534499729e-05,
+      "loss": 0.3493,
+      "step": 584
+    },
+    {
+      "epoch": 0.4721549636803874,
+      "grad_norm": 0.039188142865896225,
+      "learning_rate": 1.1394896930442998e-05,
+      "loss": 0.3186,
+      "step": 585
+    },
+    {
+      "epoch": 0.47296206618240516,
+      "grad_norm": 0.04252387955784798,
+      "learning_rate": 1.1368989781833102e-05,
+      "loss": 0.3874,
+      "step": 586
+    },
+    {
+      "epoch": 0.4737691686844229,
+      "grad_norm": 0.041413117200136185,
+      "learning_rate": 1.1343073265939161e-05,
+      "loss": 0.4033,
+      "step": 587
+    },
+    {
+      "epoch": 0.4745762711864407,
+      "grad_norm": 0.04359855130314827,
+      "learning_rate": 1.1317147560094394e-05,
+      "loss": 0.4151,
+      "step": 588
+    },
+    {
+      "epoch": 0.47538337368845845,
+      "grad_norm": 0.04194974526762962,
+      "learning_rate": 1.1291212841694907e-05,
+      "loss": 0.3966,
+      "step": 589
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.051467522978782654,
+      "learning_rate": 1.126526928819847e-05,
+      "loss": 0.4678,
+      "step": 590
+    },
+    {
+      "epoch": 0.47699757869249393,
+      "grad_norm": 0.047667261213064194,
+      "learning_rate": 1.1239317077123301e-05,
+      "loss": 0.4456,
+      "step": 591
+    },
+    {
+      "epoch": 0.4778046811945117,
+      "grad_norm": 0.044262733310461044,
+      "learning_rate": 1.1213356386046871e-05,
+      "loss": 0.4349,
+      "step": 592
+    },
+    {
+      "epoch": 0.47861178369652946,
+      "grad_norm": 0.04664575308561325,
+      "learning_rate": 1.1187387392604665e-05,
+      "loss": 0.4318,
+      "step": 593
+    },
+    {
+      "epoch": 0.4794188861985472,
+      "grad_norm": 0.043433092534542084,
+      "learning_rate": 1.1161410274488975e-05,
+      "loss": 0.4258,
+      "step": 594
+    },
+    {
+      "epoch": 0.480225988700565,
+      "grad_norm": 0.04289019852876663,
+      "learning_rate": 1.1135425209447692e-05,
+      "loss": 0.4328,
+      "step": 595
+    },
+    {
+      "epoch": 0.48103309120258275,
+      "grad_norm": 0.04451136663556099,
+      "learning_rate": 1.1109432375283087e-05,
+      "loss": 0.4602,
+      "step": 596
+    },
+    {
+      "epoch": 0.48184019370460046,
+      "grad_norm": 0.04060007631778717,
+      "learning_rate": 1.108343194985058e-05,
+      "loss": 0.3685,
+      "step": 597
+    },
+    {
+      "epoch": 0.4826472962066182,
+      "grad_norm": 0.043877966701984406,
+      "learning_rate": 1.105742411105754e-05,
+      "loss": 0.4037,
+      "step": 598
+    },
+    {
+      "epoch": 0.483454398708636,
+      "grad_norm": 0.04378955066204071,
+      "learning_rate": 1.1031409036862066e-05,
+      "loss": 0.4241,
+      "step": 599
+    },
+    {
+      "epoch": 0.48426150121065376,
+      "grad_norm": 0.04598834738135338,
+      "learning_rate": 1.100538690527176e-05,
+      "loss": 0.4278,
+      "step": 600
+    },
+    {
+      "epoch": 0.4850686037126715,
+      "grad_norm": 0.04028361290693283,
+      "learning_rate": 1.097935789434251e-05,
+      "loss": 0.3873,
+      "step": 601
+    },
+    {
+      "epoch": 0.4858757062146893,
+      "grad_norm": 0.0447799488902092,
+      "learning_rate": 1.0953322182177289e-05,
+      "loss": 0.4208,
+      "step": 602
+    },
+    {
+      "epoch": 0.48668280871670705,
+      "grad_norm": 0.0401802696287632,
+      "learning_rate": 1.0927279946924907e-05,
+      "loss": 0.3745,
+      "step": 603
+    },
+    {
+      "epoch": 0.48748991121872476,
+      "grad_norm": 0.049999117851257324,
+      "learning_rate": 1.0901231366778817e-05,
+      "loss": 0.4353,
+      "step": 604
+    },
+    {
+      "epoch": 0.4882970137207425,
+      "grad_norm": 0.04383356496691704,
+      "learning_rate": 1.087517661997589e-05,
+      "loss": 0.3881,
+      "step": 605
+    },
+    {
+      "epoch": 0.4891041162227603,
+      "grad_norm": 0.039079997688531876,
+      "learning_rate": 1.0849115884795186e-05,
+      "loss": 0.3434,
+      "step": 606
+    },
+    {
+      "epoch": 0.48991121872477805,
+      "grad_norm": 0.04208359494805336,
+      "learning_rate": 1.0823049339556739e-05,
+      "loss": 0.4092,
+      "step": 607
+    },
+    {
+      "epoch": 0.4907183212267958,
+      "grad_norm": 0.041914671659469604,
+      "learning_rate": 1.0796977162620343e-05,
+      "loss": 0.3685,
+      "step": 608
+    },
+    {
+      "epoch": 0.4915254237288136,
+      "grad_norm": 0.0390278585255146,
+      "learning_rate": 1.0770899532384329e-05,
+      "loss": 0.3857,
+      "step": 609
+    },
+    {
+      "epoch": 0.4923325262308313,
+      "grad_norm": 0.04671277478337288,
+      "learning_rate": 1.0744816627284337e-05,
+      "loss": 0.4768,
+      "step": 610
+    },
+    {
+      "epoch": 0.49313962873284906,
+      "grad_norm": 0.045115381479263306,
+      "learning_rate": 1.0718728625792095e-05,
+      "loss": 0.4087,
+      "step": 611
+    },
+    {
+      "epoch": 0.4939467312348668,
+      "grad_norm": 0.052573543041944504,
+      "learning_rate": 1.0692635706414217e-05,
+      "loss": 0.4743,
+      "step": 612
+    },
+    {
+      "epoch": 0.4947538337368846,
+      "grad_norm": 0.045512620359659195,
+      "learning_rate": 1.0666538047690956e-05,
+      "loss": 0.3968,
+      "step": 613
+    },
+    {
+      "epoch": 0.49556093623890235,
+      "grad_norm": 0.04206302762031555,
+      "learning_rate": 1.0640435828194995e-05,
+      "loss": 0.4065,
+      "step": 614
+    },
+    {
+      "epoch": 0.4963680387409201,
+      "grad_norm": 0.04784831777215004,
+      "learning_rate": 1.0614329226530238e-05,
+      "loss": 0.4117,
+      "step": 615
+    },
+    {
+      "epoch": 0.4971751412429379,
+      "grad_norm": 0.0442197360098362,
+      "learning_rate": 1.058821842133055e-05,
+      "loss": 0.387,
+      "step": 616
+    },
+    {
+      "epoch": 0.4979822437449556,
+      "grad_norm": 0.04519395902752876,
+      "learning_rate": 1.0562103591258577e-05,
+      "loss": 0.3652,
+      "step": 617
+    },
+    {
+      "epoch": 0.49878934624697335,
+      "grad_norm": 0.044820185750722885,
+      "learning_rate": 1.05359849150045e-05,
+      "loss": 0.4579,
+      "step": 618
+    },
+    {
+      "epoch": 0.4995964487489911,
+      "grad_norm": 0.04219202697277069,
+      "learning_rate": 1.0509862571284812e-05,
+      "loss": 0.4186,
+      "step": 619
+    },
+    {
+      "epoch": 0.5004035512510089,
+      "grad_norm": 0.03920457139611244,
+      "learning_rate": 1.0483736738841115e-05,
+      "loss": 0.3337,
+      "step": 620
+    },
+    {
+      "epoch": 0.5012106537530266,
+      "grad_norm": 0.04564947262406349,
+      "learning_rate": 1.0457607596438864e-05,
+      "loss": 0.4153,
+      "step": 621
+    },
+    {
+      "epoch": 0.5020177562550444,
+      "grad_norm": 0.04112798348069191,
+      "learning_rate": 1.0431475322866174e-05,
+      "loss": 0.4401,
+      "step": 622
+    },
+    {
+      "epoch": 0.5028248587570622,
+      "grad_norm": 0.0465184785425663,
+      "learning_rate": 1.040534009693258e-05,
+      "loss": 0.4448,
+      "step": 623
+    },
+    {
+      "epoch": 0.5036319612590799,
+      "grad_norm": 0.03798345848917961,
+      "learning_rate": 1.0379202097467825e-05,
+      "loss": 0.3517,
+      "step": 624
+    },
+    {
+      "epoch": 0.5044390637610977,
+      "grad_norm": 0.04183805361390114,
+      "learning_rate": 1.035306150332062e-05,
+      "loss": 0.4508,
+      "step": 625
+    },
+    {
+      "epoch": 0.5052461662631155,
+      "grad_norm": 0.04639915004372597,
+      "learning_rate": 1.032691849335744e-05,
+      "loss": 0.3802,
+      "step": 626
+    },
+    {
+      "epoch": 0.5060532687651331,
+      "grad_norm": 0.044055815786123276,
+      "learning_rate": 1.0300773246461286e-05,
+      "loss": 0.4868,
+      "step": 627
+    },
+    {
+      "epoch": 0.5068603712671509,
+      "grad_norm": 0.03922991082072258,
+      "learning_rate": 1.0274625941530464e-05,
+      "loss": 0.3864,
+      "step": 628
+    },
+    {
+      "epoch": 0.5076674737691687,
+      "grad_norm": 0.04367651045322418,
+      "learning_rate": 1.024847675747736e-05,
+      "loss": 0.4214,
+      "step": 629
+    },
+    {
+      "epoch": 0.5084745762711864,
+      "grad_norm": 0.04028239846229553,
+      "learning_rate": 1.0222325873227223e-05,
+      "loss": 0.3886,
+      "step": 630
+    },
+    {
+      "epoch": 0.5092816787732042,
+      "grad_norm": 0.04197577387094498,
+      "learning_rate": 1.0196173467716935e-05,
+      "loss": 0.3542,
+      "step": 631
+    },
+    {
+      "epoch": 0.510088781275222,
+      "grad_norm": 0.0527801550924778,
+      "learning_rate": 1.017001971989378e-05,
+      "loss": 0.4785,
+      "step": 632
+    },
+    {
+      "epoch": 0.5108958837772397,
+      "grad_norm": 0.04306379333138466,
+      "learning_rate": 1.0143864808714238e-05,
+      "loss": 0.4334,
+      "step": 633
+    },
+    {
+      "epoch": 0.5117029862792575,
+      "grad_norm": 0.03770716115832329,
+      "learning_rate": 1.0117708913142744e-05,
+      "loss": 0.3406,
+      "step": 634
+    },
+    {
+      "epoch": 0.5125100887812752,
+      "grad_norm": 0.04461565613746643,
+      "learning_rate": 1.0091552212150464e-05,
+      "loss": 0.3929,
+      "step": 635
+    },
+    {
+      "epoch": 0.513317191283293,
+      "grad_norm": 0.04541825130581856,
+      "learning_rate": 1.0065394884714077e-05,
+      "loss": 0.4846,
+      "step": 636
+    },
+    {
+      "epoch": 0.5141242937853108,
+      "grad_norm": 0.04394349828362465,
+      "learning_rate": 1.0039237109814559e-05,
+      "loss": 0.4096,
+      "step": 637
+    },
+    {
+      "epoch": 0.5149313962873285,
+      "grad_norm": 0.04815034940838814,
+      "learning_rate": 1.0013079066435933e-05,
+      "loss": 0.4932,
+      "step": 638
+    },
+    {
+      "epoch": 0.5157384987893463,
+      "grad_norm": 0.04465879872441292,
+      "learning_rate": 9.986920933564069e-06,
+      "loss": 0.4452,
+      "step": 639
+    },
+    {
+      "epoch": 0.516545601291364,
+      "grad_norm": 0.046694882214069366,
+      "learning_rate": 9.960762890185443e-06,
+      "loss": 0.4554,
+      "step": 640
+    },
+    {
+      "epoch": 0.5173527037933817,
+      "grad_norm": 0.039234355092048645,
+      "learning_rate": 9.934605115285924e-06,
+      "loss": 0.3147,
+      "step": 641
+    },
+    {
+      "epoch": 0.5181598062953995,
+      "grad_norm": 0.0512404665350914,
+      "learning_rate": 9.908447787849542e-06,
+      "loss": 0.4644,
+      "step": 642
+    },
+    {
+      "epoch": 0.5189669087974172,
+      "grad_norm": 0.04253727197647095,
+      "learning_rate": 9.88229108685726e-06,
+      "loss": 0.3808,
+      "step": 643
+    },
+    {
+      "epoch": 0.519774011299435,
+      "grad_norm": 0.03961801528930664,
+      "learning_rate": 9.856135191285763e-06,
+      "loss": 0.3379,
+      "step": 644
+    },
+    {
+      "epoch": 0.5205811138014528,
+      "grad_norm": 0.04661339893937111,
+      "learning_rate": 9.82998028010622e-06,
+      "loss": 0.3963,
+      "step": 645
+    },
+    {
+      "epoch": 0.5213882163034705,
+      "grad_norm": 0.04584723711013794,
+      "learning_rate": 9.803826532283068e-06,
+      "loss": 0.4102,
+      "step": 646
+    },
+    {
+      "epoch": 0.5221953188054883,
+      "grad_norm": 0.043115437030792236,
+      "learning_rate": 9.777674126772782e-06,
+      "loss": 0.4193,
+      "step": 647
+    },
+    {
+      "epoch": 0.5230024213075061,
+      "grad_norm": 0.0466901957988739,
+      "learning_rate": 9.751523242522644e-06,
+      "loss": 0.4218,
+      "step": 648
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 0.04835409298539162,
+      "learning_rate": 9.725374058469538e-06,
+      "loss": 0.4366,
+      "step": 649
+    },
+    {
+      "epoch": 0.5246166263115416,
+      "grad_norm": 0.04258834943175316,
+      "learning_rate": 9.699226753538716e-06,
+      "loss": 0.3321,
+      "step": 650
+    },
+    {
+      "epoch": 0.5254237288135594,
+      "grad_norm": 0.04763965308666229,
+      "learning_rate": 9.67308150664256e-06,
+      "loss": 0.3591,
+      "step": 651
+    },
+    {
+      "epoch": 0.5262308313155771,
+      "grad_norm": 0.04827116057276726,
+      "learning_rate": 9.646938496679383e-06,
+      "loss": 0.4874,
+      "step": 652
+    },
+    {
+      "epoch": 0.5270379338175948,
+      "grad_norm": 0.04092332348227501,
+      "learning_rate": 9.62079790253218e-06,
+      "loss": 0.4025,
+      "step": 653
+    },
+    {
+      "epoch": 0.5278450363196125,
+      "grad_norm": 0.04576878994703293,
+      "learning_rate": 9.594659903067424e-06,
+      "loss": 0.463,
+      "step": 654
+    },
+    {
+      "epoch": 0.5286521388216303,
+      "grad_norm": 0.044727567583322525,
+      "learning_rate": 9.56852467713383e-06,
+      "loss": 0.3952,
+      "step": 655
+    },
+    {
+      "epoch": 0.5294592413236481,
+      "grad_norm": 0.0517955981194973,
+      "learning_rate": 9.54239240356114e-06,
+      "loss": 0.4618,
+      "step": 656
+    },
+    {
+      "epoch": 0.5302663438256658,
+      "grad_norm": 0.046104107052087784,
+      "learning_rate": 9.516263261158892e-06,
+      "loss": 0.4143,
+      "step": 657
+    },
+    {
+      "epoch": 0.5310734463276836,
+      "grad_norm": 0.048121001571416855,
+      "learning_rate": 9.49013742871519e-06,
+      "loss": 0.4183,
+      "step": 658
+    },
+    {
+      "epoch": 0.5318805488297014,
+      "grad_norm": 0.0463542602956295,
+      "learning_rate": 9.464015084995503e-06,
+      "loss": 0.4279,
+      "step": 659
+    },
+    {
+      "epoch": 0.5326876513317191,
+      "grad_norm": 0.04264969006180763,
+      "learning_rate": 9.437896408741426e-06,
+      "loss": 0.3497,
+      "step": 660
+    },
+    {
+      "epoch": 0.5334947538337369,
+      "grad_norm": 0.04616716876626015,
+      "learning_rate": 9.411781578669452e-06,
+      "loss": 0.3653,
+      "step": 661
+    },
+    {
+      "epoch": 0.5343018563357547,
+      "grad_norm": 0.04366273805499077,
+      "learning_rate": 9.385670773469767e-06,
+      "loss": 0.3905,
+      "step": 662
+    },
+    {
+      "epoch": 0.5351089588377724,
+      "grad_norm": 0.04917929694056511,
+      "learning_rate": 9.359564171805006e-06,
+      "loss": 0.4314,
+      "step": 663
+    },
+    {
+      "epoch": 0.5359160613397902,
+      "grad_norm": 0.042748548090457916,
+      "learning_rate": 9.333461952309049e-06,
+      "loss": 0.3719,
+      "step": 664
+    },
+    {
+      "epoch": 0.536723163841808,
+      "grad_norm": 0.042014606297016144,
+      "learning_rate": 9.307364293585785e-06,
+      "loss": 0.377,
+      "step": 665
+    },
+    {
+      "epoch": 0.5375302663438256,
+      "grad_norm": 0.047551706433296204,
+      "learning_rate": 9.281271374207907e-06,
+      "loss": 0.4342,
+      "step": 666
+    },
+    {
+      "epoch": 0.5383373688458434,
+      "grad_norm": 0.04877143353223801,
+      "learning_rate": 9.25518337271567e-06,
+      "loss": 0.4899,
+      "step": 667
+    },
+    {
+      "epoch": 0.5391444713478611,
+      "grad_norm": 0.043122414499521255,
+      "learning_rate": 9.229100467615673e-06,
+      "loss": 0.4008,
+      "step": 668
+    },
+    {
+      "epoch": 0.5399515738498789,
+      "grad_norm": 0.04934109374880791,
+      "learning_rate": 9.20302283737966e-06,
+      "loss": 0.4822,
+      "step": 669
+    },
+    {
+      "epoch": 0.5407586763518967,
+      "grad_norm": 0.04138447344303131,
+      "learning_rate": 9.176950660443263e-06,
+      "loss": 0.3791,
+      "step": 670
+    },
+    {
+      "epoch": 0.5415657788539144,
+      "grad_norm": 0.050151653587818146,
+      "learning_rate": 9.150884115204816e-06,
+      "loss": 0.3879,
+      "step": 671
+    },
+    {
+      "epoch": 0.5423728813559322,
+      "grad_norm": 0.03810346871614456,
+      "learning_rate": 9.124823380024114e-06,
+      "loss": 0.3253,
+      "step": 672
+    },
+    {
+      "epoch": 0.54317998385795,
+      "grad_norm": 0.04531979188323021,
+      "learning_rate": 9.098768633221187e-06,
+      "loss": 0.4136,
+      "step": 673
+    },
+    {
+      "epoch": 0.5439870863599677,
+      "grad_norm": 0.04374850168824196,
+      "learning_rate": 9.072720053075096e-06,
+      "loss": 0.413,
+      "step": 674
+    },
+    {
+      "epoch": 0.5447941888619855,
+      "grad_norm": 0.049856919795274734,
+      "learning_rate": 9.046677817822716e-06,
+      "loss": 0.4459,
+      "step": 675
+    },
+    {
+      "epoch": 0.5456012913640033,
+      "grad_norm": 0.04950174689292908,
+      "learning_rate": 9.02064210565749e-06,
+      "loss": 0.4244,
+      "step": 676
+    },
+    {
+      "epoch": 0.546408393866021,
+      "grad_norm": 0.04497706890106201,
+      "learning_rate": 8.994613094728246e-06,
+      "loss": 0.3811,
+      "step": 677
+    },
+    {
+      "epoch": 0.5472154963680388,
+      "grad_norm": 0.04163501039147377,
+      "learning_rate": 8.968590963137935e-06,
+      "loss": 0.3594,
+      "step": 678
+    },
+    {
+      "epoch": 0.5480225988700564,
+      "grad_norm": 0.050817154347896576,
+      "learning_rate": 8.942575888942462e-06,
+      "loss": 0.3757,
+      "step": 679
+    },
+    {
+      "epoch": 0.5488297013720742,
+      "grad_norm": 0.050537239760160446,
+      "learning_rate": 8.916568050149422e-06,
+      "loss": 0.4263,
+      "step": 680
+    },
+    {
+      "epoch": 0.549636803874092,
+      "grad_norm": 0.0474698431789875,
+      "learning_rate": 8.890567624716913e-06,
+      "loss": 0.357,
+      "step": 681
+    },
+    {
+      "epoch": 0.5504439063761097,
+      "grad_norm": 0.04469543695449829,
+      "learning_rate": 8.86457479055231e-06,
+      "loss": 0.3911,
+      "step": 682
+    },
+    {
+      "epoch": 0.5512510088781275,
+      "grad_norm": 0.044582296162843704,
+      "learning_rate": 8.83858972551103e-06,
+      "loss": 0.3657,
+      "step": 683
+    },
+    {
+      "epoch": 0.5520581113801453,
+      "grad_norm": 0.04360214248299599,
+      "learning_rate": 8.812612607395338e-06,
+      "loss": 0.4512,
+      "step": 684
+    },
+    {
+      "epoch": 0.552865213882163,
+      "grad_norm": 0.03942641615867615,
+      "learning_rate": 8.78664361395313e-06,
+      "loss": 0.3448,
+      "step": 685
+    },
+    {
+      "epoch": 0.5536723163841808,
+      "grad_norm": 0.04751434177160263,
+      "learning_rate": 8.760682922876697e-06,
+      "loss": 0.4069,
+      "step": 686
+    },
+    {
+      "epoch": 0.5544794188861986,
+      "grad_norm": 0.039195213466882706,
+      "learning_rate": 8.734730711801535e-06,
+      "loss": 0.3701,
+      "step": 687
+    },
+    {
+      "epoch": 0.5552865213882163,
+      "grad_norm": 0.049299318343400955,
+      "learning_rate": 8.708787158305094e-06,
+      "loss": 0.4405,
+      "step": 688
+    },
+    {
+      "epoch": 0.5560936238902341,
+      "grad_norm": 0.04242517799139023,
+      "learning_rate": 8.682852439905609e-06,
+      "loss": 0.3831,
+      "step": 689
+    },
+    {
+      "epoch": 0.5569007263922519,
+      "grad_norm": 0.039667051285505295,
+      "learning_rate": 8.656926734060842e-06,
+      "loss": 0.3735,
+      "step": 690
+    },
+    {
+      "epoch": 0.5577078288942696,
+      "grad_norm": 0.04560346528887749,
+      "learning_rate": 8.631010218166902e-06,
+      "loss": 0.3369,
+      "step": 691
+    },
+    {
+      "epoch": 0.5585149313962873,
+      "grad_norm": 0.04885312542319298,
+      "learning_rate": 8.605103069557007e-06,
+      "loss": 0.3539,
+      "step": 692
+    },
+    {
+      "epoch": 0.559322033898305,
+      "grad_norm": 0.04993152990937233,
+      "learning_rate": 8.579205465500276e-06,
+      "loss": 0.3958,
+      "step": 693
+    },
+    {
+      "epoch": 0.5601291364003228,
+      "grad_norm": 0.04083723947405815,
+      "learning_rate": 8.553317583200527e-06,
+      "loss": 0.4112,
+      "step": 694
+    },
+    {
+      "epoch": 0.5609362389023406,
+      "grad_norm": 0.04566939175128937,
+      "learning_rate": 8.52743959979506e-06,
+      "loss": 0.413,
+      "step": 695
+    },
+    {
+      "epoch": 0.5617433414043583,
+      "grad_norm": 0.043536778539419174,
+      "learning_rate": 8.501571692353432e-06,
+      "loss": 0.3832,
+      "step": 696
+    },
+    {
+      "epoch": 0.5625504439063761,
+      "grad_norm": 0.04263819009065628,
+      "learning_rate": 8.475714037876263e-06,
+      "loss": 0.4028,
+      "step": 697
+    },
+    {
+      "epoch": 0.5633575464083939,
+      "grad_norm": 0.046215664595365524,
+      "learning_rate": 8.449866813294016e-06,
+      "loss": 0.3721,
+      "step": 698
+    },
+    {
+      "epoch": 0.5641646489104116,
+      "grad_norm": 0.04530710726976395,
+      "learning_rate": 8.424030195465781e-06,
+      "loss": 0.3834,
+      "step": 699
+    },
+    {
+      "epoch": 0.5649717514124294,
+      "grad_norm": 0.048315443098545074,
+      "learning_rate": 8.398204361178079e-06,
+      "loss": 0.4508,
+      "step": 700
+    },
+    {
+      "epoch": 0.5657788539144472,
+      "grad_norm": 0.04589349031448364,
+      "learning_rate": 8.372389487143647e-06,
+      "loss": 0.3801,
+      "step": 701
+    },
+    {
+      "epoch": 0.5665859564164649,
+      "grad_norm": 0.047280244529247284,
+      "learning_rate": 8.346585750000215e-06,
+      "loss": 0.4046,
+      "step": 702
+    },
+    {
+      "epoch": 0.5673930589184827,
+      "grad_norm": 0.043632179498672485,
+      "learning_rate": 8.320793326309327e-06,
+      "loss": 0.4125,
+      "step": 703
+    },
+    {
+      "epoch": 0.5682001614205004,
+      "grad_norm": 0.042842645198106766,
+      "learning_rate": 8.295012392555092e-06,
+      "loss": 0.3482,
+      "step": 704
+    },
+    {
+      "epoch": 0.5690072639225182,
+      "grad_norm": 0.04746723547577858,
+      "learning_rate": 8.26924312514302e-06,
+      "loss": 0.4518,
+      "step": 705
+    },
+    {
+      "epoch": 0.5698143664245359,
+      "grad_norm": 0.050085753202438354,
+      "learning_rate": 8.243485700398782e-06,
+      "loss": 0.4467,
+      "step": 706
+    },
+    {
+      "epoch": 0.5706214689265536,
+      "grad_norm": 0.049086228013038635,
+      "learning_rate": 8.217740294567024e-06,
+      "loss": 0.4415,
+      "step": 707
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.04610542580485344,
+      "learning_rate": 8.192007083810149e-06,
+      "loss": 0.3963,
+      "step": 708
+    },
+    {
+      "epoch": 0.5722356739305892,
+      "grad_norm": 0.04760580509901047,
+      "learning_rate": 8.166286244207109e-06,
+      "loss": 0.445,
+      "step": 709
+    },
+    {
+      "epoch": 0.5730427764326069,
+      "grad_norm": 0.04431818798184395,
+      "learning_rate": 8.140577951752213e-06,
+      "loss": 0.358,
+      "step": 710
+    },
+    {
+      "epoch": 0.5738498789346247,
+      "grad_norm": 0.04857053607702255,
+      "learning_rate": 8.114882382353925e-06,
+      "loss": 0.3933,
+      "step": 711
+    },
+    {
+      "epoch": 0.5746569814366425,
+      "grad_norm": 0.04636574909090996,
+      "learning_rate": 8.089199711833634e-06,
+      "loss": 0.4183,
+      "step": 712
+    },
+    {
+      "epoch": 0.5754640839386602,
+      "grad_norm": 0.04742918536067009,
+      "learning_rate": 8.063530115924486e-06,
+      "loss": 0.3524,
+      "step": 713
+    },
+    {
+      "epoch": 0.576271186440678,
+      "grad_norm": 0.043312814086675644,
+      "learning_rate": 8.037873770270149e-06,
+      "loss": 0.3452,
+      "step": 714
+    },
+    {
+      "epoch": 0.5770782889426957,
+      "grad_norm": 0.047119125723838806,
+      "learning_rate": 8.012230850423632e-06,
+      "loss": 0.3949,
+      "step": 715
+    },
+    {
+      "epoch": 0.5778853914447135,
+      "grad_norm": 0.04490087553858757,
+      "learning_rate": 7.986601531846084e-06,
+      "loss": 0.3413,
+      "step": 716
+    },
+    {
+      "epoch": 0.5786924939467313,
+      "grad_norm": 0.041443049907684326,
+      "learning_rate": 7.960985989905584e-06,
+      "loss": 0.3568,
+      "step": 717
+    },
+    {
+      "epoch": 0.579499596448749,
+      "grad_norm": 0.043406542390584946,
+      "learning_rate": 7.935384399875941e-06,
+      "loss": 0.33,
+      "step": 718
+    },
+    {
+      "epoch": 0.5803066989507667,
+      "grad_norm": 0.04098028689622879,
+      "learning_rate": 7.9097969369355e-06,
+      "loss": 0.3882,
+      "step": 719
+    },
+    {
+      "epoch": 0.5811138014527845,
+      "grad_norm": 0.04382213577628136,
+      "learning_rate": 7.884223776165947e-06,
+      "loss": 0.3984,
+      "step": 720
+    },
+    {
+      "epoch": 0.5819209039548022,
+      "grad_norm": 0.03980768844485283,
+      "learning_rate": 7.858665092551095e-06,
+      "loss": 0.355,
+      "step": 721
+    },
+    {
+      "epoch": 0.58272800645682,
+      "grad_norm": 0.04712940379977226,
+      "learning_rate": 7.833121060975708e-06,
+      "loss": 0.4196,
+      "step": 722
+    },
+    {
+      "epoch": 0.5835351089588378,
+      "grad_norm": 0.04403112828731537,
+      "learning_rate": 7.807591856224295e-06,
+      "loss": 0.3674,
+      "step": 723
+    },
+    {
+      "epoch": 0.5843422114608555,
+      "grad_norm": 0.04864363372325897,
+      "learning_rate": 7.782077652979898e-06,
+      "loss": 0.4704,
+      "step": 724
+    },
+    {
+      "epoch": 0.5851493139628733,
+      "grad_norm": 0.04410781338810921,
+      "learning_rate": 7.756578625822923e-06,
+      "loss": 0.4068,
+      "step": 725
+    },
+    {
+      "epoch": 0.585956416464891,
+      "grad_norm": 0.039659466594457626,
+      "learning_rate": 7.731094949229934e-06,
+      "loss": 0.3709,
+      "step": 726
+    },
+    {
+      "epoch": 0.5867635189669088,
+      "grad_norm": 0.041823580861091614,
+      "learning_rate": 7.705626797572456e-06,
+      "loss": 0.3826,
+      "step": 727
+    },
+    {
+      "epoch": 0.5875706214689266,
+      "grad_norm": 0.048205725848674774,
+      "learning_rate": 7.680174345115789e-06,
+      "loss": 0.4217,
+      "step": 728
+    },
+    {
+      "epoch": 0.5883777239709443,
+      "grad_norm": 0.04867953807115555,
+      "learning_rate": 7.654737766017801e-06,
+      "loss": 0.3801,
+      "step": 729
+    },
+    {
+      "epoch": 0.5891848264729621,
+      "grad_norm": 0.04925863444805145,
+      "learning_rate": 7.629317234327761e-06,
+      "loss": 0.3971,
+      "step": 730
+    },
+    {
+      "epoch": 0.5899919289749799,
+      "grad_norm": 0.04729580134153366,
+      "learning_rate": 7.603912923985122e-06,
+      "loss": 0.4304,
+      "step": 731
+    },
+    {
+      "epoch": 0.5907990314769975,
+      "grad_norm": 0.0465373694896698,
+      "learning_rate": 7.5785250088183495e-06,
+      "loss": 0.3699,
+      "step": 732
+    },
+    {
+      "epoch": 0.5916061339790153,
+      "grad_norm": 0.045501019805669785,
+      "learning_rate": 7.55315366254373e-06,
+      "loss": 0.3363,
+      "step": 733
+    },
+    {
+      "epoch": 0.5924132364810331,
+      "grad_norm": 0.05061681196093559,
+      "learning_rate": 7.527799058764163e-06,
+      "loss": 0.4148,
+      "step": 734
+    },
+    {
+      "epoch": 0.5932203389830508,
+      "grad_norm": 0.046648457646369934,
+      "learning_rate": 7.502461370967999e-06,
+      "loss": 0.389,
+      "step": 735
+    },
+    {
+      "epoch": 0.5940274414850686,
+      "grad_norm": 0.04624919220805168,
+      "learning_rate": 7.47714077252784e-06,
+      "loss": 0.3875,
+      "step": 736
+    },
+    {
+      "epoch": 0.5948345439870864,
+      "grad_norm": 0.04941301420331001,
+      "learning_rate": 7.45183743669935e-06,
+      "loss": 0.4064,
+      "step": 737
+    },
+    {
+      "epoch": 0.5956416464891041,
+      "grad_norm": 0.04847235605120659,
+      "learning_rate": 7.426551536620082e-06,
+      "loss": 0.3604,
+      "step": 738
+    },
+    {
+      "epoch": 0.5964487489911219,
+      "grad_norm": 0.04394169896841049,
+      "learning_rate": 7.401283245308275e-06,
+      "loss": 0.3765,
+      "step": 739
+    },
+    {
+      "epoch": 0.5972558514931396,
+      "grad_norm": 0.048312414437532425,
+      "learning_rate": 7.376032735661689e-06,
+      "loss": 0.4014,
+      "step": 740
+    },
+    {
+      "epoch": 0.5980629539951574,
+      "grad_norm": 0.045725490897893906,
+      "learning_rate": 7.350800180456405e-06,
+      "loss": 0.4155,
+      "step": 741
+    },
+    {
+      "epoch": 0.5988700564971752,
+      "grad_norm": 0.050338540226221085,
+      "learning_rate": 7.325585752345663e-06,
+      "loss": 0.4415,
+      "step": 742
+    },
+    {
+      "epoch": 0.5996771589991929,
+      "grad_norm": 0.04538509622216225,
+      "learning_rate": 7.30038962385866e-06,
+      "loss": 0.3874,
+      "step": 743
+    },
+    {
+      "epoch": 0.6004842615012107,
+      "grad_norm": 0.05074746534228325,
+      "learning_rate": 7.275211967399374e-06,
+      "loss": 0.3919,
+      "step": 744
+    },
+    {
+      "epoch": 0.6012913640032284,
+      "grad_norm": 0.04364052787423134,
+      "learning_rate": 7.250052955245399e-06,
+      "loss": 0.3886,
+      "step": 745
+    },
+    {
+      "epoch": 0.6020984665052461,
+      "grad_norm": 0.049053583294153214,
+      "learning_rate": 7.224912759546752e-06,
+      "loss": 0.4,
+      "step": 746
+    },
+    {
+      "epoch": 0.6029055690072639,
+      "grad_norm": 0.05290306359529495,
+      "learning_rate": 7.199791552324694e-06,
+      "loss": 0.4192,
+      "step": 747
+    },
+    {
+      "epoch": 0.6037126715092817,
+      "grad_norm": 0.05066477879881859,
+      "learning_rate": 7.1746895054705705e-06,
+      "loss": 0.4287,
+      "step": 748
+    },
+    {
+      "epoch": 0.6045197740112994,
+      "grad_norm": 0.04861418902873993,
+      "learning_rate": 7.149606790744602e-06,
+      "loss": 0.412,
+      "step": 749
+    },
+    {
+      "epoch": 0.6053268765133172,
+      "grad_norm": 0.04922877252101898,
+      "learning_rate": 7.12454357977475e-06,
+      "loss": 0.4198,
+      "step": 750
+    },
+    {
+      "epoch": 0.606133979015335,
+      "grad_norm": 0.039801307022571564,
+      "learning_rate": 7.099500044055507e-06,
+      "loss": 0.3603,
+      "step": 751
+    },
+    {
+      "epoch": 0.6069410815173527,
+      "grad_norm": 0.0413752906024456,
+      "learning_rate": 7.07447635494675e-06,
+      "loss": 0.3778,
+      "step": 752
+    },
+    {
+      "epoch": 0.6077481840193705,
+      "grad_norm": 0.05113828927278519,
+      "learning_rate": 7.049472683672544e-06,
+      "loss": 0.4032,
+      "step": 753
+    },
+    {
+      "epoch": 0.6085552865213882,
+      "grad_norm": 0.04697990417480469,
+      "learning_rate": 7.0244892013199976e-06,
+      "loss": 0.3934,
+      "step": 754
+    },
+    {
+      "epoch": 0.609362389023406,
+      "grad_norm": 0.04355325922369957,
+      "learning_rate": 6.999526078838056e-06,
+      "loss": 0.3863,
+      "step": 755
+    },
+    {
+      "epoch": 0.6101694915254238,
+      "grad_norm": 0.04471498355269432,
+      "learning_rate": 6.9745834870363725e-06,
+      "loss": 0.3622,
+      "step": 756
+    },
+    {
+      "epoch": 0.6109765940274415,
+      "grad_norm": 0.04962782561779022,
+      "learning_rate": 6.949661596584108e-06,
+      "loss": 0.402,
+      "step": 757
+    },
+    {
+      "epoch": 0.6117836965294592,
+      "grad_norm": 0.04287125542759895,
+      "learning_rate": 6.924760578008782e-06,
+      "loss": 0.3331,
+      "step": 758
+    },
+    {
+      "epoch": 0.612590799031477,
+      "grad_norm": 0.048510193824768066,
+      "learning_rate": 6.899880601695095e-06,
+      "loss": 0.4074,
+      "step": 759
+    },
+    {
+      "epoch": 0.6133979015334947,
+      "grad_norm": 0.045081108808517456,
+      "learning_rate": 6.875021837883758e-06,
+      "loss": 0.4115,
+      "step": 760
+    },
+    {
+      "epoch": 0.6142050040355125,
+      "grad_norm": 0.053174663335084915,
+      "learning_rate": 6.8501844566703525e-06,
+      "loss": 0.4534,
+      "step": 761
+    },
+    {
+      "epoch": 0.6150121065375302,
+      "grad_norm": 0.04487127065658569,
+      "learning_rate": 6.8253686280041395e-06,
+      "loss": 0.4223,
+      "step": 762
+    },
+    {
+      "epoch": 0.615819209039548,
+      "grad_norm": 0.050555501133203506,
+      "learning_rate": 6.800574521686908e-06,
+      "loss": 0.4536,
+      "step": 763
+    },
+    {
+      "epoch": 0.6166263115415658,
+      "grad_norm": 0.047019895166158676,
+      "learning_rate": 6.775802307371819e-06,
+      "loss": 0.3819,
+      "step": 764
+    },
+    {
+      "epoch": 0.6174334140435835,
+      "grad_norm": 0.04500943049788475,
+      "learning_rate": 6.751052154562227e-06,
+      "loss": 0.3923,
+      "step": 765
+    },
+    {
+      "epoch": 0.6182405165456013,
+      "grad_norm": 0.04650492966175079,
+      "learning_rate": 6.726324232610535e-06,
+      "loss": 0.3725,
+      "step": 766
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 0.04661106690764427,
+      "learning_rate": 6.701618710717036e-06,
+      "loss": 0.3785,
+      "step": 767
+    },
+    {
+      "epoch": 0.6198547215496368,
+      "grad_norm": 0.044705361127853394,
+      "learning_rate": 6.676935757928751e-06,
+      "loss": 0.3442,
+      "step": 768
+    },
+    {
+      "epoch": 0.6206618240516546,
+      "grad_norm": 0.04856501892209053,
+      "learning_rate": 6.652275543138271e-06,
+      "loss": 0.4079,
+      "step": 769
+    },
+    {
+      "epoch": 0.6214689265536724,
+      "grad_norm": 0.0447048656642437,
+      "learning_rate": 6.627638235082594e-06,
+      "loss": 0.3875,
+      "step": 770
+    },
+    {
+      "epoch": 0.6222760290556901,
+      "grad_norm": 0.04440969228744507,
+      "learning_rate": 6.603024002341991e-06,
+      "loss": 0.3647,
+      "step": 771
+    },
+    {
+      "epoch": 0.6230831315577078,
+      "grad_norm": 0.046474307775497437,
+      "learning_rate": 6.578433013338838e-06,
+      "loss": 0.4085,
+      "step": 772
+    },
+    {
+      "epoch": 0.6238902340597255,
+      "grad_norm": 0.04039683938026428,
+      "learning_rate": 6.553865436336461e-06,
+      "loss": 0.3918,
+      "step": 773
+    },
+    {
+      "epoch": 0.6246973365617433,
+      "grad_norm": 0.041126858443021774,
+      "learning_rate": 6.529321439438002e-06,
+      "loss": 0.3198,
+      "step": 774
+    },
+    {
+      "epoch": 0.6255044390637611,
+      "grad_norm": 0.047061994671821594,
+      "learning_rate": 6.504801190585239e-06,
+      "loss": 0.3589,
+      "step": 775
+    },
+    {
+      "epoch": 0.6263115415657788,
+      "grad_norm": 0.0462849922478199,
+      "learning_rate": 6.4803048575574635e-06,
+      "loss": 0.3766,
+      "step": 776
+    },
+    {
+      "epoch": 0.6271186440677966,
+      "grad_norm": 0.038880471140146255,
+      "learning_rate": 6.455832607970327e-06,
+      "loss": 0.338,
+      "step": 777
+    },
+    {
+      "epoch": 0.6279257465698144,
+      "grad_norm": 0.04801465570926666,
+      "learning_rate": 6.431384609274689e-06,
+      "loss": 0.3689,
+      "step": 778
+    },
+    {
+      "epoch": 0.6287328490718321,
+      "grad_norm": 0.04502476006746292,
+      "learning_rate": 6.406961028755471e-06,
+      "loss": 0.4114,
+      "step": 779
+    },
+    {
+      "epoch": 0.6295399515738499,
+      "grad_norm": 0.050989337265491486,
+      "learning_rate": 6.382562033530506e-06,
+      "loss": 0.4461,
+      "step": 780
+    },
+    {
+      "epoch": 0.6303470540758677,
+      "grad_norm": 0.04601307213306427,
+      "learning_rate": 6.358187790549418e-06,
+      "loss": 0.3419,
+      "step": 781
+    },
+    {
+      "epoch": 0.6311541565778854,
+      "grad_norm": 0.04786008596420288,
+      "learning_rate": 6.333838466592453e-06,
+      "loss": 0.4029,
+      "step": 782
+    },
+    {
+      "epoch": 0.6319612590799032,
+      "grad_norm": 0.049381598830223083,
+      "learning_rate": 6.309514228269351e-06,
+      "loss": 0.3908,
+      "step": 783
+    },
+    {
+      "epoch": 0.632768361581921,
+      "grad_norm": 0.04913535341620445,
+      "learning_rate": 6.285215242018214e-06,
+      "loss": 0.3768,
+      "step": 784
+    },
+    {
+      "epoch": 0.6335754640839386,
+      "grad_norm": 0.048182886093854904,
+      "learning_rate": 6.260941674104339e-06,
+      "loss": 0.3727,
+      "step": 785
+    },
+    {
+      "epoch": 0.6343825665859564,
+      "grad_norm": 0.04583975300192833,
+      "learning_rate": 6.236693690619112e-06,
+      "loss": 0.3501,
+      "step": 786
+    },
+    {
+      "epoch": 0.6351896690879741,
+      "grad_norm": 0.049178920686244965,
+      "learning_rate": 6.212471457478857e-06,
+      "loss": 0.4163,
+      "step": 787
+    },
+    {
+      "epoch": 0.6359967715899919,
+      "grad_norm": 0.057297833263874054,
+      "learning_rate": 6.188275140423694e-06,
+      "loss": 0.4326,
+      "step": 788
+    },
+    {
+      "epoch": 0.6368038740920097,
+      "grad_norm": 0.04626166820526123,
+      "learning_rate": 6.164104905016426e-06,
+      "loss": 0.4021,
+      "step": 789
+    },
+    {
+      "epoch": 0.6376109765940274,
+      "grad_norm": 0.04393228515982628,
+      "learning_rate": 6.1399609166413765e-06,
+      "loss": 0.3317,
+      "step": 790
+    },
+    {
+      "epoch": 0.6384180790960452,
+      "grad_norm": 0.041127268224954605,
+      "learning_rate": 6.115843340503288e-06,
+      "loss": 0.3409,
+      "step": 791
+    },
+    {
+      "epoch": 0.639225181598063,
+      "grad_norm": 0.04137645289301872,
+      "learning_rate": 6.091752341626168e-06,
+      "loss": 0.3631,
+      "step": 792
+    },
+    {
+      "epoch": 0.6400322841000807,
+      "grad_norm": 0.04440583661198616,
+      "learning_rate": 6.0676880848521794e-06,
+      "loss": 0.4272,
+      "step": 793
+    },
+    {
+      "epoch": 0.6408393866020985,
+      "grad_norm": 0.05209174379706383,
+      "learning_rate": 6.043650734840496e-06,
+      "loss": 0.4008,
+      "step": 794
+    },
+    {
+      "epoch": 0.6416464891041163,
+      "grad_norm": 0.045219723135232925,
+      "learning_rate": 6.019640456066181e-06,
+      "loss": 0.3482,
+      "step": 795
+    },
+    {
+      "epoch": 0.642453591606134,
+      "grad_norm": 0.05367840453982353,
+      "learning_rate": 5.995657412819068e-06,
+      "loss": 0.4266,
+      "step": 796
+    },
+    {
+      "epoch": 0.6432606941081518,
+      "grad_norm": 0.0441439189016819,
+      "learning_rate": 5.971701769202632e-06,
+      "loss": 0.3938,
+      "step": 797
+    },
+    {
+      "epoch": 0.6440677966101694,
+      "grad_norm": 0.046317096799612045,
+      "learning_rate": 5.947773689132863e-06,
+      "loss": 0.3681,
+      "step": 798
+    },
+    {
+      "epoch": 0.6448748991121872,
+      "grad_norm": 0.045697569847106934,
+      "learning_rate": 5.923873336337154e-06,
+      "loss": 0.4577,
+      "step": 799
+    },
+    {
+      "epoch": 0.645682001614205,
+      "grad_norm": 0.044494569301605225,
+      "learning_rate": 5.900000874353164e-06,
+      "loss": 0.4019,
+      "step": 800
+    },
+    {
+      "epoch": 0.6464891041162227,
+      "grad_norm": 0.048398248851299286,
+      "learning_rate": 5.876156466527723e-06,
+      "loss": 0.3927,
+      "step": 801
+    },
+    {
+      "epoch": 0.6472962066182405,
+      "grad_norm": 0.05004660412669182,
+      "learning_rate": 5.852340276015689e-06,
+      "loss": 0.3581,
+      "step": 802
+    },
+    {
+      "epoch": 0.6481033091202583,
+      "grad_norm": 0.04907209798693657,
+      "learning_rate": 5.828552465778855e-06,
+      "loss": 0.3857,
+      "step": 803
+    },
+    {
+      "epoch": 0.648910411622276,
+      "grad_norm": 0.045290783047676086,
+      "learning_rate": 5.804793198584815e-06,
+      "loss": 0.35,
+      "step": 804
+    },
+    {
+      "epoch": 0.6497175141242938,
+      "grad_norm": 0.049314212054014206,
+      "learning_rate": 5.781062637005853e-06,
+      "loss": 0.4207,
+      "step": 805
+    },
+    {
+      "epoch": 0.6505246166263116,
+      "grad_norm": 0.04956541955471039,
+      "learning_rate": 5.757360943417848e-06,
+      "loss": 0.4034,
+      "step": 806
+    },
+    {
+      "epoch": 0.6513317191283293,
+      "grad_norm": 0.047677140682935715,
+      "learning_rate": 5.7336882799991425e-06,
+      "loss": 0.3735,
+      "step": 807
+    },
+    {
+      "epoch": 0.6521388216303471,
+      "grad_norm": 0.04479856416583061,
+      "learning_rate": 5.710044808729447e-06,
+      "loss": 0.3895,
+      "step": 808
+    },
+    {
+      "epoch": 0.6529459241323649,
+      "grad_norm": 0.051077935844659805,
+      "learning_rate": 5.686430691388718e-06,
+      "loss": 0.4304,
+      "step": 809
+    },
+    {
+      "epoch": 0.6537530266343826,
+      "grad_norm": 0.04743655025959015,
+      "learning_rate": 5.662846089556054e-06,
+      "loss": 0.3646,
+      "step": 810
+    },
+    {
+      "epoch": 0.6545601291364003,
+      "grad_norm": 0.04958782345056534,
+      "learning_rate": 5.639291164608609e-06,
+      "loss": 0.3504,
+      "step": 811
+    },
+    {
+      "epoch": 0.655367231638418,
+      "grad_norm": 0.04840300232172012,
+      "learning_rate": 5.6157660777204635e-06,
+      "loss": 0.3677,
+      "step": 812
+    },
+    {
+      "epoch": 0.6561743341404358,
+      "grad_norm": 0.04449933022260666,
+      "learning_rate": 5.592270989861535e-06,
+      "loss": 0.3586,
+      "step": 813
+    },
+    {
+      "epoch": 0.6569814366424536,
+      "grad_norm": 0.048872679471969604,
+      "learning_rate": 5.568806061796478e-06,
+      "loss": 0.4055,
+      "step": 814
+    },
+    {
+      "epoch": 0.6577885391444713,
+      "grad_norm": 0.05295393988490105,
+      "learning_rate": 5.545371454083566e-06,
+      "loss": 0.4166,
+      "step": 815
+    },
+    {
+      "epoch": 0.6585956416464891,
+      "grad_norm": 0.050493236631155014,
+      "learning_rate": 5.521967327073612e-06,
+      "loss": 0.3925,
+      "step": 816
+    },
+    {
+      "epoch": 0.6594027441485069,
+      "grad_norm": 0.05255108326673508,
+      "learning_rate": 5.498593840908871e-06,
+      "loss": 0.4376,
+      "step": 817
+    },
+    {
+      "epoch": 0.6602098466505246,
+      "grad_norm": 0.04289624094963074,
+      "learning_rate": 5.475251155521935e-06,
+      "loss": 0.4391,
+      "step": 818
+    },
+    {
+      "epoch": 0.6610169491525424,
+      "grad_norm": 0.04436655342578888,
+      "learning_rate": 5.4519394306346415e-06,
+      "loss": 0.3855,
+      "step": 819
+    },
+    {
+      "epoch": 0.6618240516545602,
+      "grad_norm": 0.045538369566202164,
+      "learning_rate": 5.428658825756987e-06,
+      "loss": 0.383,
+      "step": 820
+    },
+    {
+      "epoch": 0.6626311541565779,
+      "grad_norm": 0.051048438996076584,
+      "learning_rate": 5.4054095001860196e-06,
+      "loss": 0.3938,
+      "step": 821
+    },
+    {
+      "epoch": 0.6634382566585957,
+      "grad_norm": 0.042728327214717865,
+      "learning_rate": 5.382191613004761e-06,
+      "loss": 0.3949,
+      "step": 822
+    },
+    {
+      "epoch": 0.6642453591606134,
+      "grad_norm": 0.0470639243721962,
+      "learning_rate": 5.359005323081127e-06,
+      "loss": 0.4268,
+      "step": 823
+    },
+    {
+      "epoch": 0.6650524616626311,
+      "grad_norm": 0.049070440232753754,
+      "learning_rate": 5.335850789066819e-06,
+      "loss": 0.3786,
+      "step": 824
+    },
+    {
+      "epoch": 0.6658595641646489,
+      "grad_norm": 0.05124800652265549,
+      "learning_rate": 5.31272816939626e-06,
+      "loss": 0.3951,
+      "step": 825
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.045142266899347305,
+      "learning_rate": 5.289637622285485e-06,
+      "loss": 0.3774,
+      "step": 826
+    },
+    {
+      "epoch": 0.6674737691686844,
+      "grad_norm": 0.04465753585100174,
+      "learning_rate": 5.266579305731088e-06,
+      "loss": 0.3376,
+      "step": 827
+    },
+    {
+      "epoch": 0.6682808716707022,
+      "grad_norm": 0.046953234821558,
+      "learning_rate": 5.2435533775091115e-06,
+      "loss": 0.3808,
+      "step": 828
+    },
+    {
+      "epoch": 0.6690879741727199,
+      "grad_norm": 0.048814088106155396,
+      "learning_rate": 5.220559995173995e-06,
+      "loss": 0.3921,
+      "step": 829
+    },
+    {
+      "epoch": 0.6698950766747377,
+      "grad_norm": 0.05050293728709221,
+      "learning_rate": 5.197599316057479e-06,
+      "loss": 0.4176,
+      "step": 830
+    },
+    {
+      "epoch": 0.6707021791767555,
+      "grad_norm": 0.041133660823106766,
+      "learning_rate": 5.174671497267525e-06,
+      "loss": 0.3142,
+      "step": 831
+    },
+    {
+      "epoch": 0.6715092816787732,
+      "grad_norm": 0.04762447252869606,
+      "learning_rate": 5.151776695687261e-06,
+      "loss": 0.3962,
+      "step": 832
+    },
+    {
+      "epoch": 0.672316384180791,
+      "grad_norm": 0.049833763390779495,
+      "learning_rate": 5.128915067973888e-06,
+      "loss": 0.3544,
+      "step": 833
+    },
+    {
+      "epoch": 0.6731234866828087,
+      "grad_norm": 0.04340628907084465,
+      "learning_rate": 5.106086770557619e-06,
+      "loss": 0.3583,
+      "step": 834
+    },
+    {
+      "epoch": 0.6739305891848265,
+      "grad_norm": 0.05325612425804138,
+      "learning_rate": 5.083291959640602e-06,
+      "loss": 0.4229,
+      "step": 835
+    },
+    {
+      "epoch": 0.6747376916868443,
+      "grad_norm": 0.04724089801311493,
+      "learning_rate": 5.0605307911958515e-06,
+      "loss": 0.3477,
+      "step": 836
+    },
+    {
+      "epoch": 0.6755447941888619,
+      "grad_norm": 0.04456155002117157,
+      "learning_rate": 5.03780342096619e-06,
+      "loss": 0.3177,
+      "step": 837
+    },
+    {
+      "epoch": 0.6763518966908797,
+      "grad_norm": 0.053026482462882996,
+      "learning_rate": 5.015110004463177e-06,
+      "loss": 0.4088,
+      "step": 838
+    },
+    {
+      "epoch": 0.6771589991928975,
+      "grad_norm": 0.040615472942590714,
+      "learning_rate": 4.992450696966039e-06,
+      "loss": 0.349,
+      "step": 839
+    },
+    {
+      "epoch": 0.6779661016949152,
+      "grad_norm": 0.051405586302280426,
+      "learning_rate": 4.969825653520621e-06,
+      "loss": 0.37,
+      "step": 840
+    },
+    {
+      "epoch": 0.678773204196933,
+      "grad_norm": 0.04546589031815529,
+      "learning_rate": 4.947235028938309e-06,
+      "loss": 0.3746,
+      "step": 841
+    },
+    {
+      "epoch": 0.6795803066989508,
+      "grad_norm": 0.04085971787571907,
+      "learning_rate": 4.9246789777949725e-06,
+      "loss": 0.3511,
+      "step": 842
+    },
+    {
+      "epoch": 0.6803874092009685,
+      "grad_norm": 0.050242677330970764,
+      "learning_rate": 4.902157654429928e-06,
+      "loss": 0.4374,
+      "step": 843
+    },
+    {
+      "epoch": 0.6811945117029863,
+      "grad_norm": 0.052028290927410126,
+      "learning_rate": 4.8796712129448596e-06,
+      "loss": 0.3637,
+      "step": 844
+    },
+    {
+      "epoch": 0.682001614205004,
+      "grad_norm": 0.04794751852750778,
+      "learning_rate": 4.857219807202781e-06,
+      "loss": 0.3421,
+      "step": 845
+    },
+    {
+      "epoch": 0.6828087167070218,
+      "grad_norm": 0.04805922508239746,
+      "learning_rate": 4.834803590826956e-06,
+      "loss": 0.3942,
+      "step": 846
+    },
+    {
+      "epoch": 0.6836158192090396,
+      "grad_norm": 0.05546120926737785,
+      "learning_rate": 4.8124227171998905e-06,
+      "loss": 0.4107,
+      "step": 847
+    },
+    {
+      "epoch": 0.6844229217110573,
+      "grad_norm": 0.052278514951467514,
+      "learning_rate": 4.790077339462234e-06,
+      "loss": 0.4613,
+      "step": 848
+    },
+    {
+      "epoch": 0.6852300242130751,
+      "grad_norm": 0.04965787008404732,
+      "learning_rate": 4.7677676105117734e-06,
+      "loss": 0.406,
+      "step": 849
+    },
+    {
+      "epoch": 0.6860371267150929,
+      "grad_norm": 0.04963855817914009,
+      "learning_rate": 4.745493683002368e-06,
+      "loss": 0.403,
+      "step": 850
+    },
+    {
+      "epoch": 0.6868442292171105,
+      "grad_norm": 0.03866651654243469,
+      "learning_rate": 4.723255709342893e-06,
+      "loss": 0.292,
+      "step": 851
+    },
+    {
+      "epoch": 0.6876513317191283,
+      "grad_norm": 0.04614201933145523,
+      "learning_rate": 4.701053841696226e-06,
+      "loss": 0.3667,
+      "step": 852
+    },
+    {
+      "epoch": 0.688458434221146,
+      "grad_norm": 0.051998917013406754,
+      "learning_rate": 4.678888231978186e-06,
+      "loss": 0.429,
+      "step": 853
+    },
+    {
+      "epoch": 0.6892655367231638,
+      "grad_norm": 0.044506240636110306,
+      "learning_rate": 4.6567590318564886e-06,
+      "loss": 0.3222,
+      "step": 854
+    },
+    {
+      "epoch": 0.6900726392251816,
+      "grad_norm": 0.047115735709667206,
+      "learning_rate": 4.634666392749729e-06,
+      "loss": 0.3772,
+      "step": 855
+    },
+    {
+      "epoch": 0.6908797417271993,
+      "grad_norm": 0.046469517052173615,
+      "learning_rate": 4.612610465826326e-06,
+      "loss": 0.3457,
+      "step": 856
+    },
+    {
+      "epoch": 0.6916868442292171,
+      "grad_norm": 0.05154654383659363,
+      "learning_rate": 4.590591402003501e-06,
+      "loss": 0.4502,
+      "step": 857
+    },
+    {
+      "epoch": 0.6924939467312349,
+      "grad_norm": 0.05142826586961746,
+      "learning_rate": 4.568609351946241e-06,
+      "loss": 0.3652,
+      "step": 858
+    },
+    {
+      "epoch": 0.6933010492332526,
+      "grad_norm": 0.056190911680459976,
+      "learning_rate": 4.546664466066268e-06,
+      "loss": 0.4137,
+      "step": 859
+    },
+    {
+      "epoch": 0.6941081517352704,
+      "grad_norm": 0.04742627218365669,
+      "learning_rate": 4.524756894521003e-06,
+      "loss": 0.3578,
+      "step": 860
+    },
+    {
+      "epoch": 0.6949152542372882,
+      "grad_norm": 0.053221605718135834,
+      "learning_rate": 4.502886787212543e-06,
+      "loss": 0.413,
+      "step": 861
+    },
+    {
+      "epoch": 0.6957223567393059,
+      "grad_norm": 0.05199446156620979,
+      "learning_rate": 4.481054293786645e-06,
+      "loss": 0.372,
+      "step": 862
+    },
+    {
+      "epoch": 0.6965294592413237,
+      "grad_norm": 0.050332196056842804,
+      "learning_rate": 4.4592595636316915e-06,
+      "loss": 0.365,
+      "step": 863
+    },
+    {
+      "epoch": 0.6973365617433414,
+      "grad_norm": 0.0501842126250267,
+      "learning_rate": 4.437502745877668e-06,
+      "loss": 0.4525,
+      "step": 864
+    },
+    {
+      "epoch": 0.6981436642453591,
+      "grad_norm": 0.049436986446380615,
+      "learning_rate": 4.415783989395151e-06,
+      "loss": 0.3934,
+      "step": 865
+    },
+    {
+      "epoch": 0.6989507667473769,
+      "grad_norm": 0.051288872957229614,
+      "learning_rate": 4.394103442794273e-06,
+      "loss": 0.316,
+      "step": 866
+    },
+    {
+      "epoch": 0.6997578692493946,
+      "grad_norm": 0.04671122506260872,
+      "learning_rate": 4.372461254423722e-06,
+      "loss": 0.3919,
+      "step": 867
+    },
+    {
+      "epoch": 0.7005649717514124,
+      "grad_norm": 0.0548025481402874,
+      "learning_rate": 4.350857572369722e-06,
+      "loss": 0.4481,
+      "step": 868
+    },
+    {
+      "epoch": 0.7013720742534302,
+      "grad_norm": 0.047147806733846664,
+      "learning_rate": 4.329292544455017e-06,
+      "loss": 0.3677,
+      "step": 869
+    },
+    {
+      "epoch": 0.7021791767554479,
+      "grad_norm": 0.04897439107298851,
+      "learning_rate": 4.3077663182378634e-06,
+      "loss": 0.3749,
+      "step": 870
+    },
+    {
+      "epoch": 0.7029862792574657,
+      "grad_norm": 0.05925610661506653,
+      "learning_rate": 4.286279041011017e-06,
+      "loss": 0.4326,
+      "step": 871
+    },
+    {
+      "epoch": 0.7037933817594835,
+      "grad_norm": 0.05080854520201683,
+      "learning_rate": 4.264830859800719e-06,
+      "loss": 0.3676,
+      "step": 872
+    },
+    {
+      "epoch": 0.7046004842615012,
+      "grad_norm": 0.05124906823039055,
+      "learning_rate": 4.243421921365702e-06,
+      "loss": 0.4365,
+      "step": 873
+    },
+    {
+      "epoch": 0.705407586763519,
+      "grad_norm": 0.046696025878190994,
+      "learning_rate": 4.222052372196184e-06,
+      "loss": 0.3863,
+      "step": 874
+    },
+    {
+      "epoch": 0.7062146892655368,
+      "grad_norm": 0.0472051203250885,
+      "learning_rate": 4.2007223585128576e-06,
+      "loss": 0.3862,
+      "step": 875
+    },
+    {
+      "epoch": 0.7070217917675545,
+      "grad_norm": 0.04693121090531349,
+      "learning_rate": 4.1794320262659015e-06,
+      "loss": 0.3594,
+      "step": 876
+    },
+    {
+      "epoch": 0.7078288942695722,
+      "grad_norm": 0.052231140434741974,
+      "learning_rate": 4.158181521133964e-06,
+      "loss": 0.38,
+      "step": 877
+    },
+    {
+      "epoch": 0.70863599677159,
+      "grad_norm": 0.04406065121293068,
+      "learning_rate": 4.136970988523189e-06,
+      "loss": 0.3882,
+      "step": 878
+    },
+    {
+      "epoch": 0.7094430992736077,
+      "grad_norm": 0.04371519759297371,
+      "learning_rate": 4.115800573566197e-06,
+      "loss": 0.3596,
+      "step": 879
+    },
+    {
+      "epoch": 0.7102502017756255,
+      "grad_norm": 0.042410142719745636,
+      "learning_rate": 4.0946704211211165e-06,
+      "loss": 0.3507,
+      "step": 880
+    },
+    {
+      "epoch": 0.7110573042776432,
+      "grad_norm": 0.04645742103457451,
+      "learning_rate": 4.073580675770575e-06,
+      "loss": 0.3602,
+      "step": 881
+    },
+    {
+      "epoch": 0.711864406779661,
+      "grad_norm": 0.05483156070113182,
+      "learning_rate": 4.052531481820713e-06,
+      "loss": 0.4491,
+      "step": 882
+    },
+    {
+      "epoch": 0.7126715092816788,
+      "grad_norm": 0.04915105551481247,
+      "learning_rate": 4.0315229833002e-06,
+      "loss": 0.3792,
+      "step": 883
+    },
+    {
+      "epoch": 0.7134786117836965,
+      "grad_norm": 0.05002758279442787,
+      "learning_rate": 4.010555323959253e-06,
+      "loss": 0.4236,
+      "step": 884
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.04625044763088226,
+      "learning_rate": 3.989628647268645e-06,
+      "loss": 0.4021,
+      "step": 885
+    },
+    {
+      "epoch": 0.7150928167877321,
+      "grad_norm": 0.050767894834280014,
+      "learning_rate": 3.9687430964187225e-06,
+      "loss": 0.4123,
+      "step": 886
+    },
+    {
+      "epoch": 0.7158999192897498,
+      "grad_norm": 0.04956914111971855,
+      "learning_rate": 3.9478988143184286e-06,
+      "loss": 0.4182,
+      "step": 887
+    },
+    {
+      "epoch": 0.7167070217917676,
+      "grad_norm": 0.044376108795404434,
+      "learning_rate": 3.927095943594331e-06,
+      "loss": 0.3106,
+      "step": 888
+    },
+    {
+      "epoch": 0.7175141242937854,
+      "grad_norm": 0.044406965374946594,
+      "learning_rate": 3.906334626589642e-06,
+      "loss": 0.3655,
+      "step": 889
+    },
+    {
+      "epoch": 0.718321226795803,
+      "grad_norm": 0.04712330549955368,
+      "learning_rate": 3.885615005363237e-06,
+      "loss": 0.3986,
+      "step": 890
+    },
+    {
+      "epoch": 0.7191283292978208,
+      "grad_norm": 0.04551446810364723,
+      "learning_rate": 3.864937221688697e-06,
+      "loss": 0.3642,
+      "step": 891
+    },
+    {
+      "epoch": 0.7199354317998385,
+      "grad_norm": 0.04825610667467117,
+      "learning_rate": 3.844301417053323e-06,
+      "loss": 0.4192,
+      "step": 892
+    },
+    {
+      "epoch": 0.7207425343018563,
+      "grad_norm": 0.05064934864640236,
+      "learning_rate": 3.823707732657175e-06,
+      "loss": 0.3854,
+      "step": 893
+    },
+    {
+      "epoch": 0.7215496368038741,
+      "grad_norm": 0.05061221495270729,
+      "learning_rate": 3.8031563094121104e-06,
+      "loss": 0.4301,
+      "step": 894
+    },
+    {
+      "epoch": 0.7223567393058918,
+      "grad_norm": 0.04634290561079979,
+      "learning_rate": 3.782647287940817e-06,
+      "loss": 0.3964,
+      "step": 895
+    },
+    {
+      "epoch": 0.7231638418079096,
+      "grad_norm": 0.049565527588129044,
+      "learning_rate": 3.762180808575848e-06,
+      "loss": 0.4221,
+      "step": 896
+    },
+    {
+      "epoch": 0.7239709443099274,
+      "grad_norm": 0.0462944395840168,
+      "learning_rate": 3.7417570113586553e-06,
+      "loss": 0.3673,
+      "step": 897
+    },
+    {
+      "epoch": 0.7247780468119451,
+      "grad_norm": 0.04157343879342079,
+      "learning_rate": 3.721376036038653e-06,
+      "loss": 0.3436,
+      "step": 898
+    },
+    {
+      "epoch": 0.7255851493139629,
+      "grad_norm": 0.05127215012907982,
+      "learning_rate": 3.701038022072234e-06,
+      "loss": 0.384,
+      "step": 899
+    },
+    {
+      "epoch": 0.7263922518159807,
+      "grad_norm": 0.0550130195915699,
+      "learning_rate": 3.680743108621836e-06,
+      "loss": 0.4338,
+      "step": 900
+    },
+    {
+      "epoch": 0.7271993543179984,
+      "grad_norm": 0.048612404614686966,
+      "learning_rate": 3.6604914345549857e-06,
+      "loss": 0.3708,
+      "step": 901
+    },
+    {
+      "epoch": 0.7280064568200162,
+      "grad_norm": 0.04515503719449043,
+      "learning_rate": 3.6402831384433346e-06,
+      "loss": 0.3574,
+      "step": 902
+    },
+    {
+      "epoch": 0.7288135593220338,
+      "grad_norm": 0.048535142093896866,
+      "learning_rate": 3.6201183585617304e-06,
+      "loss": 0.3614,
+      "step": 903
+    },
+    {
+      "epoch": 0.7296206618240516,
+      "grad_norm": 0.05240613594651222,
+      "learning_rate": 3.5999972328872623e-06,
+      "loss": 0.3595,
+      "step": 904
+    },
+    {
+      "epoch": 0.7304277643260694,
+      "grad_norm": 0.04590952768921852,
+      "learning_rate": 3.5799198990983063e-06,
+      "loss": 0.4138,
+      "step": 905
+    },
+    {
+      "epoch": 0.7312348668280871,
+      "grad_norm": 0.04791981354355812,
+      "learning_rate": 3.5598864945736077e-06,
+      "loss": 0.4026,
+      "step": 906
+    },
+    {
+      "epoch": 0.7320419693301049,
+      "grad_norm": 0.05078763887286186,
+      "learning_rate": 3.539897156391312e-06,
+      "loss": 0.3849,
+      "step": 907
+    },
+    {
+      "epoch": 0.7328490718321227,
+      "grad_norm": 0.04727828875184059,
+      "learning_rate": 3.5199520213280526e-06,
+      "loss": 0.3521,
+      "step": 908
+    },
+    {
+      "epoch": 0.7336561743341404,
+      "grad_norm": 0.045900169759988785,
+      "learning_rate": 3.500051225858001e-06,
+      "loss": 0.3558,
+      "step": 909
+    },
+    {
+      "epoch": 0.7344632768361582,
+      "grad_norm": 0.05100518465042114,
+      "learning_rate": 3.480194906151938e-06,
+      "loss": 0.3983,
+      "step": 910
+    },
+    {
+      "epoch": 0.735270379338176,
+      "grad_norm": 0.049115296453237534,
+      "learning_rate": 3.4603831980763136e-06,
+      "loss": 0.3813,
+      "step": 911
+    },
+    {
+      "epoch": 0.7360774818401937,
+      "grad_norm": 0.05351780727505684,
+      "learning_rate": 3.440616237192325e-06,
+      "loss": 0.4019,
+      "step": 912
+    },
+    {
+      "epoch": 0.7368845843422115,
+      "grad_norm": 0.04854034259915352,
+      "learning_rate": 3.4208941587549917e-06,
+      "loss": 0.4318,
+      "step": 913
+    },
+    {
+      "epoch": 0.7376916868442293,
+      "grad_norm": 0.04963256046175957,
+      "learning_rate": 3.401217097712226e-06,
+      "loss": 0.3967,
+      "step": 914
+    },
+    {
+      "epoch": 0.738498789346247,
+      "grad_norm": 0.048572007566690445,
+      "learning_rate": 3.3815851887039064e-06,
+      "loss": 0.3585,
+      "step": 915
+    },
+    {
+      "epoch": 0.7393058918482648,
+      "grad_norm": 0.05005783215165138,
+      "learning_rate": 3.3619985660609655e-06,
+      "loss": 0.3355,
+      "step": 916
+    },
+    {
+      "epoch": 0.7401129943502824,
+      "grad_norm": 0.04688713699579239,
+      "learning_rate": 3.342457363804458e-06,
+      "loss": 0.3541,
+      "step": 917
+    },
+    {
+      "epoch": 0.7409200968523002,
+      "grad_norm": 0.04656829312443733,
+      "learning_rate": 3.322961715644647e-06,
+      "loss": 0.3095,
+      "step": 918
+    },
+    {
+      "epoch": 0.741727199354318,
+      "grad_norm": 0.05018598213791847,
+      "learning_rate": 3.3035117549801034e-06,
+      "loss": 0.3648,
+      "step": 919
+    },
+    {
+      "epoch": 0.7425343018563357,
+      "grad_norm": 0.05106114596128464,
+      "learning_rate": 3.284107614896778e-06,
+      "loss": 0.4279,
+      "step": 920
+    },
+    {
+      "epoch": 0.7433414043583535,
+      "grad_norm": 0.045182712376117706,
+      "learning_rate": 3.2647494281670967e-06,
+      "loss": 0.394,
+      "step": 921
+    },
+    {
+      "epoch": 0.7441485068603713,
+      "grad_norm": 0.04775145649909973,
+      "learning_rate": 3.2454373272490436e-06,
+      "loss": 0.337,
+      "step": 922
+    },
+    {
+      "epoch": 0.744955609362389,
+      "grad_norm": 0.045995697379112244,
+      "learning_rate": 3.226171444285272e-06,
+      "loss": 0.3459,
+      "step": 923
+    },
+    {
+      "epoch": 0.7457627118644068,
+      "grad_norm": 0.0462590754032135,
+      "learning_rate": 3.20695191110218e-06,
+      "loss": 0.3684,
+      "step": 924
+    },
+    {
+      "epoch": 0.7465698143664246,
+      "grad_norm": 0.0479058213531971,
+      "learning_rate": 3.187778859209022e-06,
+      "loss": 0.366,
+      "step": 925
+    },
+    {
+      "epoch": 0.7473769168684423,
+      "grad_norm": 0.050555307418107986,
+      "learning_rate": 3.1686524197970103e-06,
+      "loss": 0.4058,
+      "step": 926
+    },
+    {
+      "epoch": 0.7481840193704601,
+      "grad_norm": 0.048881128430366516,
+      "learning_rate": 3.1495727237384087e-06,
+      "loss": 0.3504,
+      "step": 927
+    },
+    {
+      "epoch": 0.7489911218724778,
+      "grad_norm": 0.05036919191479683,
+      "learning_rate": 3.1305399015856332e-06,
+      "loss": 0.3647,
+      "step": 928
+    },
+    {
+      "epoch": 0.7497982243744956,
+      "grad_norm": 0.04874005541205406,
+      "learning_rate": 3.111554083570377e-06,
+      "loss": 0.3649,
+      "step": 929
+    },
+    {
+      "epoch": 0.7506053268765133,
+      "grad_norm": 0.05192394554615021,
+      "learning_rate": 3.0926153996027066e-06,
+      "loss": 0.3635,
+      "step": 930
+    },
+    {
+      "epoch": 0.751412429378531,
+      "grad_norm": 0.05109355226159096,
+      "learning_rate": 3.0737239792701656e-06,
+      "loss": 0.3651,
+      "step": 931
+    },
+    {
+      "epoch": 0.7522195318805488,
+      "grad_norm": 0.04810253158211708,
+      "learning_rate": 3.05487995183691e-06,
+      "loss": 0.3923,
+      "step": 932
+    },
+    {
+      "epoch": 0.7530266343825666,
+      "grad_norm": 0.044694848358631134,
+      "learning_rate": 3.0360834462428002e-06,
+      "loss": 0.3712,
+      "step": 933
+    },
+    {
+      "epoch": 0.7538337368845843,
+      "grad_norm": 0.04288363456726074,
+      "learning_rate": 3.0173345911025343e-06,
+      "loss": 0.3793,
+      "step": 934
+    },
+    {
+      "epoch": 0.7546408393866021,
+      "grad_norm": 0.05430062487721443,
+      "learning_rate": 2.9986335147047642e-06,
+      "loss": 0.3893,
+      "step": 935
+    },
+    {
+      "epoch": 0.7554479418886199,
+      "grad_norm": 0.04797254130244255,
+      "learning_rate": 2.9799803450112185e-06,
+      "loss": 0.3265,
+      "step": 936
+    },
+    {
+      "epoch": 0.7562550443906376,
+      "grad_norm": 0.0462198331952095,
+      "learning_rate": 2.961375209655819e-06,
+      "loss": 0.3466,
+      "step": 937
+    },
+    {
+      "epoch": 0.7570621468926554,
+      "grad_norm": 0.05069499462842941,
+      "learning_rate": 2.942818235943814e-06,
+      "loss": 0.4104,
+      "step": 938
+    },
+    {
+      "epoch": 0.7578692493946732,
+      "grad_norm": 0.05128340423107147,
+      "learning_rate": 2.92430955085091e-06,
+      "loss": 0.3697,
+      "step": 939
+    },
+    {
+      "epoch": 0.7586763518966909,
+      "grad_norm": 0.05035511404275894,
+      "learning_rate": 2.9058492810224003e-06,
+      "loss": 0.3698,
+      "step": 940
+    },
+    {
+      "epoch": 0.7594834543987087,
+      "grad_norm": 0.05182954668998718,
+      "learning_rate": 2.8874375527722975e-06,
+      "loss": 0.3887,
+      "step": 941
+    },
+    {
+      "epoch": 0.7602905569007264,
+      "grad_norm": 0.047854945063591,
+      "learning_rate": 2.8690744920824697e-06,
+      "loss": 0.3705,
+      "step": 942
+    },
+    {
+      "epoch": 0.7610976594027441,
+      "grad_norm": 0.05556565895676613,
+      "learning_rate": 2.850760224601774e-06,
+      "loss": 0.4228,
+      "step": 943
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 0.05028011277318001,
+      "learning_rate": 2.832494875645202e-06,
+      "loss": 0.399,
+      "step": 944
+    },
+    {
+      "epoch": 0.7627118644067796,
+      "grad_norm": 0.04954536259174347,
+      "learning_rate": 2.814278570193025e-06,
+      "loss": 0.3334,
+      "step": 945
+    },
+    {
+      "epoch": 0.7635189669087974,
+      "grad_norm": 0.0548592135310173,
+      "learning_rate": 2.796111432889934e-06,
+      "loss": 0.415,
+      "step": 946
+    },
+    {
+      "epoch": 0.7643260694108152,
+      "grad_norm": 0.043614424765110016,
+      "learning_rate": 2.77799358804419e-06,
+      "loss": 0.3448,
+      "step": 947
+    },
+    {
+      "epoch": 0.7651331719128329,
+      "grad_norm": 0.04365172237157822,
+      "learning_rate": 2.759925159626763e-06,
+      "loss": 0.3924,
+      "step": 948
+    },
+    {
+      "epoch": 0.7659402744148507,
+      "grad_norm": 0.050896864384412766,
+      "learning_rate": 2.7419062712705025e-06,
+      "loss": 0.3812,
+      "step": 949
+    },
+    {
+      "epoch": 0.7667473769168685,
+      "grad_norm": 0.05392688512802124,
+      "learning_rate": 2.7239370462692716e-06,
+      "loss": 0.4221,
+      "step": 950
+    },
+    {
+      "epoch": 0.7675544794188862,
+      "grad_norm": 0.050148848444223404,
+      "learning_rate": 2.7060176075771217e-06,
+      "loss": 0.3669,
+      "step": 951
+    },
+    {
+      "epoch": 0.768361581920904,
+      "grad_norm": 0.04388059303164482,
+      "learning_rate": 2.6881480778074396e-06,
+      "loss": 0.3798,
+      "step": 952
+    },
+    {
+      "epoch": 0.7691686844229217,
+      "grad_norm": 0.048120852559804916,
+      "learning_rate": 2.6703285792321045e-06,
+      "loss": 0.373,
+      "step": 953
+    },
+    {
+      "epoch": 0.7699757869249395,
+      "grad_norm": 0.04943971335887909,
+      "learning_rate": 2.6525592337806684e-06,
+      "loss": 0.3712,
+      "step": 954
+    },
+    {
+      "epoch": 0.7707828894269573,
+      "grad_norm": 0.04997150972485542,
+      "learning_rate": 2.634840163039508e-06,
+      "loss": 0.3958,
+      "step": 955
+    },
+    {
+      "epoch": 0.7715899919289749,
+      "grad_norm": 0.06133696436882019,
+      "learning_rate": 2.617171488250991e-06,
+      "loss": 0.4321,
+      "step": 956
+    },
+    {
+      "epoch": 0.7723970944309927,
+      "grad_norm": 0.05077798292040825,
+      "learning_rate": 2.599553330312663e-06,
+      "loss": 0.3904,
+      "step": 957
+    },
+    {
+      "epoch": 0.7732041969330105,
+      "grad_norm": 0.05077609792351723,
+      "learning_rate": 2.581985809776395e-06,
+      "loss": 0.361,
+      "step": 958
+    },
+    {
+      "epoch": 0.7740112994350282,
+      "grad_norm": 0.04809306561946869,
+      "learning_rate": 2.564469046847583e-06,
+      "loss": 0.3769,
+      "step": 959
+    },
+    {
+      "epoch": 0.774818401937046,
+      "grad_norm": 0.04891866073012352,
+      "learning_rate": 2.5470031613843127e-06,
+      "loss": 0.3927,
+      "step": 960
+    },
+    {
+      "epoch": 0.7756255044390638,
+      "grad_norm": 0.04797545447945595,
+      "learning_rate": 2.529588272896544e-06,
+      "loss": 0.3339,
+      "step": 961
+    },
+    {
+      "epoch": 0.7764326069410815,
+      "grad_norm": 0.05068304389715195,
+      "learning_rate": 2.512224500545285e-06,
+      "loss": 0.4047,
+      "step": 962
+    },
+    {
+      "epoch": 0.7772397094430993,
+      "grad_norm": 0.05368385463953018,
+      "learning_rate": 2.494911963141784e-06,
+      "loss": 0.3864,
+      "step": 963
+    },
+    {
+      "epoch": 0.778046811945117,
+      "grad_norm": 0.05426260456442833,
+      "learning_rate": 2.4776507791467195e-06,
+      "loss": 0.3241,
+      "step": 964
+    },
+    {
+      "epoch": 0.7788539144471348,
+      "grad_norm": 0.052247244864702225,
+      "learning_rate": 2.4604410666693846e-06,
+      "loss": 0.3536,
+      "step": 965
+    },
+    {
+      "epoch": 0.7796610169491526,
+      "grad_norm": 0.04787401109933853,
+      "learning_rate": 2.4432829434668757e-06,
+      "loss": 0.3588,
+      "step": 966
+    },
+    {
+      "epoch": 0.7804681194511703,
+      "grad_norm": 0.050289399921894073,
+      "learning_rate": 2.426176526943298e-06,
+      "loss": 0.4559,
+      "step": 967
+    },
+    {
+      "epoch": 0.7812752219531881,
+      "grad_norm": 0.048816096037626266,
+      "learning_rate": 2.4091219341489458e-06,
+      "loss": 0.3831,
+      "step": 968
+    },
+    {
+      "epoch": 0.7820823244552058,
+      "grad_norm": 0.05117584764957428,
+      "learning_rate": 2.3921192817795125e-06,
+      "loss": 0.3452,
+      "step": 969
+    },
+    {
+      "epoch": 0.7828894269572235,
+      "grad_norm": 0.04862622544169426,
+      "learning_rate": 2.3751686861752967e-06,
+      "loss": 0.4173,
+      "step": 970
+    },
+    {
+      "epoch": 0.7836965294592413,
+      "grad_norm": 0.0534110851585865,
+      "learning_rate": 2.3582702633203925e-06,
+      "loss": 0.4199,
+      "step": 971
+    },
+    {
+      "epoch": 0.784503631961259,
+      "grad_norm": 0.05284995958209038,
+      "learning_rate": 2.3414241288419114e-06,
+      "loss": 0.4122,
+      "step": 972
+    },
+    {
+      "epoch": 0.7853107344632768,
+      "grad_norm": 0.049155883491039276,
+      "learning_rate": 2.32463039800917e-06,
+      "loss": 0.4517,
+      "step": 973
+    },
+    {
+      "epoch": 0.7861178369652946,
+      "grad_norm": 0.050774216651916504,
+      "learning_rate": 2.307889185732928e-06,
+      "loss": 0.4045,
+      "step": 974
+    },
+    {
+      "epoch": 0.7869249394673123,
+      "grad_norm": 0.055596962571144104,
+      "learning_rate": 2.2912006065645763e-06,
+      "loss": 0.4097,
+      "step": 975
+    },
+    {
+      "epoch": 0.7877320419693301,
+      "grad_norm": 0.04935078322887421,
+      "learning_rate": 2.2745647746953714e-06,
+      "loss": 0.3764,
+      "step": 976
+    },
+    {
+      "epoch": 0.7885391444713479,
+      "grad_norm": 0.05432211980223656,
+      "learning_rate": 2.25798180395565e-06,
+      "loss": 0.3792,
+      "step": 977
+    },
+    {
+      "epoch": 0.7893462469733656,
+      "grad_norm": 0.05181829631328583,
+      "learning_rate": 2.2414518078140366e-06,
+      "loss": 0.3573,
+      "step": 978
+    },
+    {
+      "epoch": 0.7901533494753834,
+      "grad_norm": 0.04890815168619156,
+      "learning_rate": 2.2249748993766908e-06,
+      "loss": 0.3253,
+      "step": 979
+    },
+    {
+      "epoch": 0.7909604519774012,
+      "grad_norm": 0.051907122135162354,
+      "learning_rate": 2.208551191386512e-06,
+      "loss": 0.4315,
+      "step": 980
+    },
+    {
+      "epoch": 0.7917675544794189,
+      "grad_norm": 0.05192062631249428,
+      "learning_rate": 2.192180796222384e-06,
+      "loss": 0.3794,
+      "step": 981
+    },
+    {
+      "epoch": 0.7925746569814366,
+      "grad_norm": 0.05353526771068573,
+      "learning_rate": 2.17586382589839e-06,
+      "loss": 0.3847,
+      "step": 982
+    },
+    {
+      "epoch": 0.7933817594834544,
+      "grad_norm": 0.04733423516154289,
+      "learning_rate": 2.1596003920630626e-06,
+      "loss": 0.3336,
+      "step": 983
+    },
+    {
+      "epoch": 0.7941888619854721,
+      "grad_norm": 0.046822912991046906,
+      "learning_rate": 2.143390605998604e-06,
+      "loss": 0.3655,
+      "step": 984
+    },
+    {
+      "epoch": 0.7949959644874899,
+      "grad_norm": 0.05472195893526077,
+      "learning_rate": 2.1272345786201408e-06,
+      "loss": 0.4253,
+      "step": 985
+    },
+    {
+      "epoch": 0.7958030669895076,
+      "grad_norm": 0.04976249858736992,
+      "learning_rate": 2.111132420474953e-06,
+      "loss": 0.32,
+      "step": 986
+    },
+    {
+      "epoch": 0.7966101694915254,
+      "grad_norm": 0.04802384972572327,
+      "learning_rate": 2.095084241741725e-06,
+      "loss": 0.3397,
+      "step": 987
+    },
+    {
+      "epoch": 0.7974172719935432,
+      "grad_norm": 0.047920335084199905,
+      "learning_rate": 2.0790901522297822e-06,
+      "loss": 0.3472,
+      "step": 988
+    },
+    {
+      "epoch": 0.7982243744955609,
+      "grad_norm": 0.04877651855349541,
+      "learning_rate": 2.063150261378346e-06,
+      "loss": 0.3795,
+      "step": 989
+    },
+    {
+      "epoch": 0.7990314769975787,
+      "grad_norm": 0.05073659494519234,
+      "learning_rate": 2.0472646782557915e-06,
+      "loss": 0.378,
+      "step": 990
+    },
+    {
+      "epoch": 0.7998385794995965,
+      "grad_norm": 0.053008127957582474,
+      "learning_rate": 2.0314335115588866e-06,
+      "loss": 0.4238,
+      "step": 991
+    },
+    {
+      "epoch": 0.8006456820016142,
+      "grad_norm": 0.052719682455062866,
+      "learning_rate": 2.0156568696120636e-06,
+      "loss": 0.4257,
+      "step": 992
+    },
+    {
+      "epoch": 0.801452784503632,
+      "grad_norm": 0.048593856394290924,
+      "learning_rate": 1.9999348603666657e-06,
+      "loss": 0.3694,
+      "step": 993
+    },
+    {
+      "epoch": 0.8022598870056498,
+      "grad_norm": 0.04972432553768158,
+      "learning_rate": 1.9842675914002117e-06,
+      "loss": 0.3558,
+      "step": 994
+    },
+    {
+      "epoch": 0.8030669895076675,
+      "grad_norm": 0.04915817454457283,
+      "learning_rate": 1.968655169915661e-06,
+      "loss": 0.3779,
+      "step": 995
+    },
+    {
+      "epoch": 0.8038740920096852,
+      "grad_norm": 0.055468250066041946,
+      "learning_rate": 1.9530977027406828e-06,
+      "loss": 0.4878,
+      "step": 996
+    },
+    {
+      "epoch": 0.804681194511703,
+      "grad_norm": 0.052601367235183716,
+      "learning_rate": 1.937595296326924e-06,
+      "loss": 0.4093,
+      "step": 997
+    },
+    {
+      "epoch": 0.8054882970137207,
+      "grad_norm": 0.05105729401111603,
+      "learning_rate": 1.922148056749279e-06,
+      "loss": 0.4446,
+      "step": 998
+    },
+    {
+      "epoch": 0.8062953995157385,
+      "grad_norm": 0.05593395605683327,
+      "learning_rate": 1.9067560897051585e-06,
+      "loss": 0.3759,
+      "step": 999
+    },
+    {
+      "epoch": 0.8071025020177562,
+      "grad_norm": 0.048278238624334335,
+      "learning_rate": 1.8914195005137793e-06,
+      "loss": 0.4164,
+      "step": 1000
+    },
+    {
+      "epoch": 0.807909604519774,
+      "grad_norm": 0.04513949155807495,
+      "learning_rate": 1.8761383941154288e-06,
+      "loss": 0.336,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8087167070217918,
+      "grad_norm": 0.051345016807317734,
+      "learning_rate": 1.8609128750707629e-06,
+      "loss": 0.3387,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 0.05032331123948097,
+      "learning_rate": 1.845743047560079e-06,
+      "loss": 0.3358,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8103309120258273,
+      "grad_norm": 0.05130847543478012,
+      "learning_rate": 1.8306290153826012e-06,
+      "loss": 0.385,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8111380145278451,
+      "grad_norm": 0.04798781871795654,
+      "learning_rate": 1.8155708819557826e-06,
+      "loss": 0.3949,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8119451170298628,
+      "grad_norm": 0.0486295223236084,
+      "learning_rate": 1.8005687503145886e-06,
+      "loss": 0.3852,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8127522195318806,
+      "grad_norm": 0.04694712907075882,
+      "learning_rate": 1.7856227231107892e-06,
+      "loss": 0.3438,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8135593220338984,
+      "grad_norm": 0.05250077322125435,
+      "learning_rate": 1.7707329026122665e-06,
+      "loss": 0.4018,
+      "step": 1008
+    },
+    {
+      "epoch": 0.814366424535916,
+      "grad_norm": 0.04679161310195923,
+      "learning_rate": 1.7558993907023037e-06,
+      "loss": 0.3884,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8151735270379338,
+      "grad_norm": 0.055237215012311935,
+      "learning_rate": 1.7411222888788993e-06,
+      "loss": 0.3961,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8159806295399515,
+      "grad_norm": 0.05706053599715233,
+      "learning_rate": 1.7264016982540632e-06,
+      "loss": 0.3956,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8167877320419693,
+      "grad_norm": 0.05254807323217392,
+      "learning_rate": 1.7117377195531337e-06,
+      "loss": 0.3668,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8175948345439871,
+      "grad_norm": 0.050519250333309174,
+      "learning_rate": 1.6971304531140753e-06,
+      "loss": 0.4352,
+      "step": 1013
+    },
+    {
+      "epoch": 0.8184019370460048,
+      "grad_norm": 0.04889274388551712,
+      "learning_rate": 1.682579998886803e-06,
+      "loss": 0.3763,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8192090395480226,
+      "grad_norm": 0.04828520119190216,
+      "learning_rate": 1.6680864564324984e-06,
+      "loss": 0.382,
+      "step": 1015
+    },
+    {
+      "epoch": 0.8200161420500404,
+      "grad_norm": 0.04728667438030243,
+      "learning_rate": 1.6536499249229243e-06,
+      "loss": 0.3385,
+      "step": 1016
+    },
+    {
+      "epoch": 0.8208232445520581,
+      "grad_norm": 0.0539923720061779,
+      "learning_rate": 1.639270503139746e-06,
+      "loss": 0.4052,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8216303470540759,
+      "grad_norm": 0.04704679921269417,
+      "learning_rate": 1.62494828947386e-06,
+      "loss": 0.3622,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8224374495560937,
+      "grad_norm": 0.054417211562395096,
+      "learning_rate": 1.6106833819247114e-06,
+      "loss": 0.3984,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8232445520581114,
+      "grad_norm": 0.05438614636659622,
+      "learning_rate": 1.5964758780996314e-06,
+      "loss": 0.3658,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8240516545601292,
+      "grad_norm": 0.04959606006741524,
+      "learning_rate": 1.5823258752131698e-06,
+      "loss": 0.351,
+      "step": 1021
+    },
+    {
+      "epoch": 0.8248587570621468,
+      "grad_norm": 0.04936448112130165,
+      "learning_rate": 1.5682334700864289e-06,
+      "loss": 0.3469,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8256658595641646,
+      "grad_norm": 0.04924401640892029,
+      "learning_rate": 1.5541987591463969e-06,
+      "loss": 0.3566,
+      "step": 1023
+    },
+    {
+      "epoch": 0.8264729620661824,
+      "grad_norm": 0.051505424082279205,
+      "learning_rate": 1.5402218384252875e-06,
+      "loss": 0.3495,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8272800645682001,
+      "grad_norm": 0.050205498933792114,
+      "learning_rate": 1.5263028035598903e-06,
+      "loss": 0.4093,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8280871670702179,
+      "grad_norm": 0.04906272143125534,
+      "learning_rate": 1.512441749790916e-06,
+      "loss": 0.3906,
+      "step": 1026
+    },
+    {
+      "epoch": 0.8288942695722357,
+      "grad_norm": 0.05131586641073227,
+      "learning_rate": 1.4986387719623286e-06,
+      "loss": 0.4302,
+      "step": 1027
+    },
+    {
+      "epoch": 0.8297013720742534,
+      "grad_norm": 0.05708414316177368,
+      "learning_rate": 1.4848939645207227e-06,
+      "loss": 0.3953,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8305084745762712,
+      "grad_norm": 0.05040213093161583,
+      "learning_rate": 1.471207421514651e-06,
+      "loss": 0.3571,
+      "step": 1029
+    },
+    {
+      "epoch": 0.831315577078289,
+      "grad_norm": 0.046975214034318924,
+      "learning_rate": 1.4575792365940023e-06,
+      "loss": 0.3597,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8321226795803067,
+      "grad_norm": 0.04971659183502197,
+      "learning_rate": 1.4440095030093459e-06,
+      "loss": 0.3662,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8329297820823245,
+      "grad_norm": 0.05251837521791458,
+      "learning_rate": 1.4304983136113048e-06,
+      "loss": 0.3641,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8337368845843423,
+      "grad_norm": 0.058566804975271225,
+      "learning_rate": 1.4170457608499077e-06,
+      "loss": 0.3912,
+      "step": 1033
+    },
+    {
+      "epoch": 0.83454398708636,
+      "grad_norm": 0.052699752151966095,
+      "learning_rate": 1.4036519367739655e-06,
+      "loss": 0.3913,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8353510895883777,
+      "grad_norm": 0.05296236276626587,
+      "learning_rate": 1.39031693303044e-06,
+      "loss": 0.3577,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8361581920903954,
+      "grad_norm": 0.04882199317216873,
+      "learning_rate": 1.3770408408638193e-06,
+      "loss": 0.3931,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8369652945924132,
+      "grad_norm": 0.049222007393836975,
+      "learning_rate": 1.3638237511154839e-06,
+      "loss": 0.3793,
+      "step": 1037
+    },
+    {
+      "epoch": 0.837772397094431,
+      "grad_norm": 0.046755850315093994,
+      "learning_rate": 1.3506657542231006e-06,
+      "loss": 0.3604,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8385794995964487,
+      "grad_norm": 0.052193090319633484,
+      "learning_rate": 1.3375669402199842e-06,
+      "loss": 0.3871,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8393866020984665,
+      "grad_norm": 0.052387386560440063,
+      "learning_rate": 1.3245273987344954e-06,
+      "loss": 0.3727,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8401937046004843,
+      "grad_norm": 0.04449461027979851,
+      "learning_rate": 1.3115472189894275e-06,
+      "loss": 0.3227,
+      "step": 1041
+    },
+    {
+      "epoch": 0.841000807102502,
+      "grad_norm": 0.04822150245308876,
+      "learning_rate": 1.2986264898013878e-06,
+      "loss": 0.3887,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8418079096045198,
+      "grad_norm": 0.049816031008958817,
+      "learning_rate": 1.2857652995801984e-06,
+      "loss": 0.3654,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8426150121065376,
+      "grad_norm": 0.053092245012521744,
+      "learning_rate": 1.2729637363282876e-06,
+      "loss": 0.396,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8434221146085553,
+      "grad_norm": 0.05084899067878723,
+      "learning_rate": 1.2602218876400807e-06,
+      "loss": 0.3845,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8442292171105731,
+      "grad_norm": 0.052148766815662384,
+      "learning_rate": 1.2475398407014117e-06,
+      "loss": 0.4181,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8450363196125908,
+      "grad_norm": 0.050065118819475174,
+      "learning_rate": 1.2349176822889242e-06,
+      "loss": 0.3919,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8458434221146085,
+      "grad_norm": 0.04557253420352936,
+      "learning_rate": 1.222355498769473e-06,
+      "loss": 0.3882,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8466505246166263,
+      "grad_norm": 0.05082574486732483,
+      "learning_rate": 1.20985337609954e-06,
+      "loss": 0.3537,
+      "step": 1049
+    },
+    {
+      "epoch": 0.847457627118644,
+      "grad_norm": 0.05619888752698898,
+      "learning_rate": 1.197411399824635e-06,
+      "loss": 0.4135,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8482647296206618,
+      "grad_norm": 0.04628870263695717,
+      "learning_rate": 1.1850296550787243e-06,
+      "loss": 0.3351,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8490718321226796,
+      "grad_norm": 0.04656703397631645,
+      "learning_rate": 1.1727082265836343e-06,
+      "loss": 0.392,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8498789346246973,
+      "grad_norm": 0.046244919300079346,
+      "learning_rate": 1.1604471986484855e-06,
+      "loss": 0.3328,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8506860371267151,
+      "grad_norm": 0.049990214407444,
+      "learning_rate": 1.1482466551691074e-06,
+      "loss": 0.3703,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8514931396287329,
+      "grad_norm": 0.05336960032582283,
+      "learning_rate": 1.136106679627461e-06,
+      "loss": 0.4406,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8523002421307506,
+      "grad_norm": 0.04921855032444,
+      "learning_rate": 1.1240273550910774e-06,
+      "loss": 0.3643,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8531073446327684,
+      "grad_norm": 0.05234561115503311,
+      "learning_rate": 1.1120087642124843e-06,
+      "loss": 0.3505,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8539144471347861,
+      "grad_norm": 0.04735950753092766,
+      "learning_rate": 1.100050989228636e-06,
+      "loss": 0.386,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8547215496368039,
+      "grad_norm": 0.04945006221532822,
+      "learning_rate": 1.0881541119603633e-06,
+      "loss": 0.3505,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8555286521388217,
+      "grad_norm": 0.04880194738507271,
+      "learning_rate": 1.0763182138117945e-06,
+      "loss": 0.3159,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8563357546408394,
+      "grad_norm": 0.04775940626859665,
+      "learning_rate": 1.0645433757698209e-06,
+      "loss": 0.3387,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.05463748052716255,
+      "learning_rate": 1.0528296784035241e-06,
+      "loss": 0.4155,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8579499596448749,
+      "grad_norm": 0.049621522426605225,
+      "learning_rate": 1.0411772018636368e-06,
+      "loss": 0.3434,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8587570621468926,
+      "grad_norm": 0.051833219826221466,
+      "learning_rate": 1.0295860258819856e-06,
+      "loss": 0.4296,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8595641646489104,
+      "grad_norm": 0.049275219440460205,
+      "learning_rate": 1.0180562297709473e-06,
+      "loss": 0.3551,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8603712671509282,
+      "grad_norm": 0.051726870238780975,
+      "learning_rate": 1.0065878924229144e-06,
+      "loss": 0.3456,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8611783696529459,
+      "grad_norm": 0.05007703974843025,
+      "learning_rate": 9.95181092309745e-07,
+      "loss": 0.3393,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8619854721549637,
+      "grad_norm": 0.05635494366288185,
+      "learning_rate": 9.838359074822323e-07,
+      "loss": 0.4061,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8627925746569814,
+      "grad_norm": 0.05516330897808075,
+      "learning_rate": 9.725524155695688e-07,
+      "loss": 0.4037,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8635996771589992,
+      "grad_norm": 0.0452413484454155,
+      "learning_rate": 9.613306937788102e-07,
+      "loss": 0.3627,
+      "step": 1070
+    },
+    {
+      "epoch": 0.864406779661017,
+      "grad_norm": 0.044155534356832504,
+      "learning_rate": 9.501708188943514e-07,
+      "loss": 0.3338,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8652138821630347,
+      "grad_norm": 0.05689741298556328,
+      "learning_rate": 9.390728672774074e-07,
+      "loss": 0.4419,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8660209846650525,
+      "grad_norm": 0.06184333935379982,
+      "learning_rate": 9.280369148654789e-07,
+      "loss": 0.45,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8668280871670703,
+      "grad_norm": 0.05039246752858162,
+      "learning_rate": 9.170630371718436e-07,
+      "loss": 0.3764,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8676351896690879,
+      "grad_norm": 0.052961479872465134,
+      "learning_rate": 9.06151309285026e-07,
+      "loss": 0.4372,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8684422921711057,
+      "grad_norm": 0.054415419697761536,
+      "learning_rate": 8.953018058682994e-07,
+      "loss": 0.4019,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8692493946731235,
+      "grad_norm": 0.0524618923664093,
+      "learning_rate": 8.845146011591654e-07,
+      "loss": 0.3817,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8700564971751412,
+      "grad_norm": 0.05585518106818199,
+      "learning_rate": 8.737897689688446e-07,
+      "loss": 0.3745,
+      "step": 1078
+    },
+    {
+      "epoch": 0.870863599677159,
+      "grad_norm": 0.04939413070678711,
+      "learning_rate": 8.631273826817821e-07,
+      "loss": 0.3667,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8716707021791767,
+      "grad_norm": 0.05061611533164978,
+      "learning_rate": 8.525275152551282e-07,
+      "loss": 0.3755,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8724778046811945,
+      "grad_norm": 0.051909856498241425,
+      "learning_rate": 8.419902392182588e-07,
+      "loss": 0.3544,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8732849071832123,
+      "grad_norm": 0.0464894101023674,
+      "learning_rate": 8.315156266722635e-07,
+      "loss": 0.2752,
+      "step": 1082
+    },
+    {
+      "epoch": 0.87409200968523,
+      "grad_norm": 0.04502680152654648,
+      "learning_rate": 8.211037492894625e-07,
+      "loss": 0.3677,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8748991121872478,
+      "grad_norm": 0.04734842851758003,
+      "learning_rate": 8.107546783129095e-07,
+      "loss": 0.298,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8757062146892656,
+      "grad_norm": 0.04617322236299515,
+      "learning_rate": 8.00468484555904e-07,
+      "loss": 0.3164,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8765133171912833,
+      "grad_norm": 0.052291661500930786,
+      "learning_rate": 7.902452384015136e-07,
+      "loss": 0.3937,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8773204196933011,
+      "grad_norm": 0.047588132321834564,
+      "learning_rate": 7.800850098020874e-07,
+      "loss": 0.3512,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8781275221953188,
+      "grad_norm": 0.052274979650974274,
+      "learning_rate": 7.699878682787787e-07,
+      "loss": 0.3795,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8789346246973365,
+      "grad_norm": 0.052272357046604156,
+      "learning_rate": 7.599538829210684e-07,
+      "loss": 0.4079,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8797417271993543,
+      "grad_norm": 0.05046515539288521,
+      "learning_rate": 7.499831223862907e-07,
+      "loss": 0.3683,
+      "step": 1090
+    },
+    {
+      "epoch": 0.880548829701372,
+      "grad_norm": 0.054393284022808075,
+      "learning_rate": 7.40075654899165e-07,
+      "loss": 0.3961,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8813559322033898,
+      "grad_norm": 0.055753178894519806,
+      "learning_rate": 7.302315482513333e-07,
+      "loss": 0.4332,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8821630347054076,
+      "grad_norm": 0.054667022079229355,
+      "learning_rate": 7.204508698008894e-07,
+      "loss": 0.3557,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8829701372074253,
+      "grad_norm": 0.056577954441308975,
+      "learning_rate": 7.10733686471925e-07,
+      "loss": 0.4301,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8837772397094431,
+      "grad_norm": 0.048220470547676086,
+      "learning_rate": 7.010800647540583e-07,
+      "loss": 0.3505,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8845843422114609,
+      "grad_norm": 0.05123896524310112,
+      "learning_rate": 6.91490070702e-07,
+      "loss": 0.3813,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8853914447134786,
+      "grad_norm": 0.05035271868109703,
+      "learning_rate": 6.819637699350811e-07,
+      "loss": 0.3497,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8861985472154964,
+      "grad_norm": 0.05092376098036766,
+      "learning_rate": 6.72501227636817e-07,
+      "loss": 0.3873,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8870056497175142,
+      "grad_norm": 0.047779954969882965,
+      "learning_rate": 6.631025085544563e-07,
+      "loss": 0.3953,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8878127522195319,
+      "grad_norm": 0.05742986127734184,
+      "learning_rate": 6.53767676998539e-07,
+      "loss": 0.3918,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8886198547215496,
+      "grad_norm": 0.04611702263355255,
+      "learning_rate": 6.444967968424543e-07,
+      "loss": 0.3518,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8894269572235673,
+      "grad_norm": 0.04918667674064636,
+      "learning_rate": 6.352899315220085e-07,
+      "loss": 0.3576,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8902340597255851,
+      "grad_norm": 0.05174895375967026,
+      "learning_rate": 6.261471440349831e-07,
+      "loss": 0.3991,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8910411622276029,
+      "grad_norm": 0.05327307805418968,
+      "learning_rate": 6.170684969407115e-07,
+      "loss": 0.4019,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8918482647296206,
+      "grad_norm": 0.05138659477233887,
+      "learning_rate": 6.08054052359649e-07,
+      "loss": 0.3321,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8926553672316384,
+      "grad_norm": 0.053043756633996964,
+      "learning_rate": 5.991038719729414e-07,
+      "loss": 0.3577,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8934624697336562,
+      "grad_norm": 0.05240008980035782,
+      "learning_rate": 5.902180170220129e-07,
+      "loss": 0.3996,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8942695722356739,
+      "grad_norm": 0.04946550726890564,
+      "learning_rate": 5.81396548308144e-07,
+      "loss": 0.415,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8950766747376917,
+      "grad_norm": 0.045649804174900055,
+      "learning_rate": 5.726395261920448e-07,
+      "loss": 0.3409,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8958837772397095,
+      "grad_norm": 0.05500829964876175,
+      "learning_rate": 5.63947010593462e-07,
+      "loss": 0.4325,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8966908797417272,
+      "grad_norm": 0.049944207072257996,
+      "learning_rate": 5.553190609907478e-07,
+      "loss": 0.3422,
+      "step": 1111
+    },
+    {
+      "epoch": 0.897497982243745,
+      "grad_norm": 0.049136772751808167,
+      "learning_rate": 5.467557364204701e-07,
+      "loss": 0.4173,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8983050847457628,
+      "grad_norm": 0.04535511136054993,
+      "learning_rate": 5.382570954769972e-07,
+      "loss": 0.3397,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8991121872477804,
+      "grad_norm": 0.048276886343955994,
+      "learning_rate": 5.298231963121059e-07,
+      "loss": 0.3849,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8999192897497982,
+      "grad_norm": 0.05075502023100853,
+      "learning_rate": 5.214540966345727e-07,
+      "loss": 0.3805,
+      "step": 1115
+    },
+    {
+      "epoch": 0.9007263922518159,
+      "grad_norm": 0.049204349517822266,
+      "learning_rate": 5.131498537097857e-07,
+      "loss": 0.3482,
+      "step": 1116
+    },
+    {
+      "epoch": 0.9015334947538337,
+      "grad_norm": 0.04519697278738022,
+      "learning_rate": 5.049105243593577e-07,
+      "loss": 0.3455,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9023405972558515,
+      "grad_norm": 0.04939783364534378,
+      "learning_rate": 4.967361649607261e-07,
+      "loss": 0.3688,
+      "step": 1118
+    },
+    {
+      "epoch": 0.9031476997578692,
+      "grad_norm": 0.04733356088399887,
+      "learning_rate": 4.886268314467746e-07,
+      "loss": 0.3951,
+      "step": 1119
+    },
+    {
+      "epoch": 0.903954802259887,
+      "grad_norm": 0.04528392478823662,
+      "learning_rate": 4.80582579305452e-07,
+      "loss": 0.3545,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 0.05273212119936943,
+      "learning_rate": 4.726034635793808e-07,
+      "loss": 0.4049,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9055690072639225,
+      "grad_norm": 0.05001245066523552,
+      "learning_rate": 4.646895388654982e-07,
+      "loss": 0.3526,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9063761097659403,
+      "grad_norm": 0.05269649997353554,
+      "learning_rate": 4.568408593146656e-07,
+      "loss": 0.3427,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9071832122679581,
+      "grad_norm": 0.053200993686914444,
+      "learning_rate": 4.490574786313107e-07,
+      "loss": 0.3871,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9079903147699758,
+      "grad_norm": 0.05451187118887901,
+      "learning_rate": 4.4133945007305614e-07,
+      "loss": 0.3682,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9087974172719936,
+      "grad_norm": 0.05771080031991005,
+      "learning_rate": 4.3368682645034687e-07,
+      "loss": 0.3663,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9096045197740112,
+      "grad_norm": 0.05234134569764137,
+      "learning_rate": 4.260996601261036e-07,
+      "loss": 0.3683,
+      "step": 1127
+    },
+    {
+      "epoch": 0.910411622276029,
+      "grad_norm": 0.05028250813484192,
+      "learning_rate": 4.18578003015353e-07,
+      "loss": 0.3965,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9112187247780468,
+      "grad_norm": 0.053080759942531586,
+      "learning_rate": 4.1112190658487707e-07,
+      "loss": 0.4066,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9120258272800645,
+      "grad_norm": 0.04984816908836365,
+      "learning_rate": 4.0373142185285984e-07,
+      "loss": 0.3632,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9128329297820823,
+      "grad_norm": 0.04705459997057915,
+      "learning_rate": 3.964065993885391e-07,
+      "loss": 0.3467,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9136400322841001,
+      "grad_norm": 0.054150789976119995,
+      "learning_rate": 3.8914748931185854e-07,
+      "loss": 0.3637,
+      "step": 1132
+    },
+    {
+      "epoch": 0.9144471347861178,
+      "grad_norm": 0.05248238146305084,
+      "learning_rate": 3.8195414129313046e-07,
+      "loss": 0.4039,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9152542372881356,
+      "grad_norm": 0.047117725014686584,
+      "learning_rate": 3.7482660455268825e-07,
+      "loss": 0.3589,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9160613397901534,
+      "grad_norm": 0.054256290197372437,
+      "learning_rate": 3.6776492786055105e-07,
+      "loss": 0.4435,
+      "step": 1135
+    },
+    {
+      "epoch": 0.9168684422921711,
+      "grad_norm": 0.054758455604314804,
+      "learning_rate": 3.6076915953609625e-07,
+      "loss": 0.3956,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9176755447941889,
+      "grad_norm": 0.048094723373651505,
+      "learning_rate": 3.53839347447722e-07,
+      "loss": 0.3926,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9184826472962067,
+      "grad_norm": 0.049094658344984055,
+      "learning_rate": 3.469755390125229e-07,
+      "loss": 0.3614,
+      "step": 1138
+    },
+    {
+      "epoch": 0.9192897497982244,
+      "grad_norm": 0.04998333007097244,
+      "learning_rate": 3.401777811959661e-07,
+      "loss": 0.3428,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9200968523002422,
+      "grad_norm": 0.046912651509046555,
+      "learning_rate": 3.33446120511568e-07,
+      "loss": 0.318,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9209039548022598,
+      "grad_norm": 0.049167439341545105,
+      "learning_rate": 3.267806030205756e-07,
+      "loss": 0.3459,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9217110573042776,
+      "grad_norm": 0.05661160126328468,
+      "learning_rate": 3.201812743316524e-07,
+      "loss": 0.4099,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9225181598062954,
+      "grad_norm": 0.05792146548628807,
+      "learning_rate": 3.136481796005686e-07,
+      "loss": 0.4039,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9233252623083131,
+      "grad_norm": 0.05346622318029404,
+      "learning_rate": 3.0718136352988925e-07,
+      "loss": 0.352,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9241323648103309,
+      "grad_norm": 0.05379043519496918,
+      "learning_rate": 3.0078087036866765e-07,
+      "loss": 0.364,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9249394673123487,
+      "grad_norm": 0.04992745816707611,
+      "learning_rate": 2.9444674391214457e-07,
+      "loss": 0.4011,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9257465698143664,
+      "grad_norm": 0.050753600895404816,
+      "learning_rate": 2.8817902750144953e-07,
+      "loss": 0.4096,
+      "step": 1147
+    },
+    {
+      "epoch": 0.9265536723163842,
+      "grad_norm": 0.045072466135025024,
+      "learning_rate": 2.819777640233001e-07,
+      "loss": 0.3485,
+      "step": 1148
+    },
+    {
+      "epoch": 0.927360774818402,
+      "grad_norm": 0.04854454845190048,
+      "learning_rate": 2.758429959097131e-07,
+      "loss": 0.3396,
+      "step": 1149
+    },
+    {
+      "epoch": 0.9281678773204197,
+      "grad_norm": 0.05049902945756912,
+      "learning_rate": 2.697747651377114e-07,
+      "loss": 0.37,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9289749798224375,
+      "grad_norm": 0.05854698270559311,
+      "learning_rate": 2.6377311322903665e-07,
+      "loss": 0.4093,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9297820823244553,
+      "grad_norm": 0.054229289293289185,
+      "learning_rate": 2.578380812498671e-07,
+      "loss": 0.401,
+      "step": 1152
+    },
+    {
+      "epoch": 0.930589184826473,
+      "grad_norm": 0.05177401751279831,
+      "learning_rate": 2.519697098105378e-07,
+      "loss": 0.3644,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9313962873284907,
+      "grad_norm": 0.056719232350587845,
+      "learning_rate": 2.4616803906525433e-07,
+      "loss": 0.3965,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9322033898305084,
+      "grad_norm": 0.050419602543115616,
+      "learning_rate": 2.4043310871182943e-07,
+      "loss": 0.4054,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9330104923325262,
+      "grad_norm": 0.044479768723249435,
+      "learning_rate": 2.3476495799140574e-07,
+      "loss": 0.3401,
+      "step": 1156
+    },
+    {
+      "epoch": 0.933817594834544,
+      "grad_norm": 0.056795384734869,
+      "learning_rate": 2.291636256881846e-07,
+      "loss": 0.4476,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9346246973365617,
+      "grad_norm": 0.05207651108503342,
+      "learning_rate": 2.2362915012916542e-07,
+      "loss": 0.3511,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9354317998385795,
+      "grad_norm": 0.05466938018798828,
+      "learning_rate": 2.1816156918388454e-07,
+      "loss": 0.4056,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9362389023405973,
+      "grad_norm": 0.050367191433906555,
+      "learning_rate": 2.1276092026414673e-07,
+      "loss": 0.434,
+      "step": 1160
+    },
+    {
+      "epoch": 0.937046004842615,
+      "grad_norm": 0.05178472772240639,
+      "learning_rate": 2.0742724032378315e-07,
+      "loss": 0.3642,
+      "step": 1161
+    },
+    {
+      "epoch": 0.9378531073446328,
+      "grad_norm": 0.05684222653508186,
+      "learning_rate": 2.0216056585838474e-07,
+      "loss": 0.399,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9386602098466506,
+      "grad_norm": 0.05073128268122673,
+      "learning_rate": 1.9696093290506368e-07,
+      "loss": 0.4132,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9394673123486683,
+      "grad_norm": 0.054002247750759125,
+      "learning_rate": 1.9182837704219914e-07,
+      "loss": 0.4296,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9402744148506861,
+      "grad_norm": 0.048514340072870255,
+      "learning_rate": 1.867629333891985e-07,
+      "loss": 0.3644,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9410815173527038,
+      "grad_norm": 0.05368981882929802,
+      "learning_rate": 1.817646366062531e-07,
+      "loss": 0.3647,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9418886198547215,
+      "grad_norm": 0.05559367686510086,
+      "learning_rate": 1.7683352089410523e-07,
+      "loss": 0.3993,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9426957223567393,
+      "grad_norm": 0.05872499570250511,
+      "learning_rate": 1.7196961999381147e-07,
+      "loss": 0.3889,
+      "step": 1168
+    },
+    {
+      "epoch": 0.943502824858757,
+      "grad_norm": 0.051862310618162155,
+      "learning_rate": 1.6717296718651078e-07,
+      "loss": 0.397,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9443099273607748,
+      "grad_norm": 0.0534013956785202,
+      "learning_rate": 1.6244359529320242e-07,
+      "loss": 0.3978,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9451170298627926,
+      "grad_norm": 0.05433657020330429,
+      "learning_rate": 1.577815366745139e-07,
+      "loss": 0.3812,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9459241323648103,
+      "grad_norm": 0.0524270161986351,
+      "learning_rate": 1.531868232304845e-07,
+      "loss": 0.3357,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9467312348668281,
+      "grad_norm": 0.054285310208797455,
+      "learning_rate": 1.4865948640034432e-07,
+      "loss": 0.4095,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9475383373688459,
+      "grad_norm": 0.062304362654685974,
+      "learning_rate": 1.4419955716230116e-07,
+      "loss": 0.4376,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9483454398708636,
+      "grad_norm": 0.05329901725053787,
+      "learning_rate": 1.398070660333295e-07,
+      "loss": 0.3448,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9491525423728814,
+      "grad_norm": 0.04291558638215065,
+      "learning_rate": 1.3548204306895628e-07,
+      "loss": 0.3197,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9499596448748991,
+      "grad_norm": 0.058099936693906784,
+      "learning_rate": 1.3122451786306223e-07,
+      "loss": 0.4031,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9507667473769169,
+      "grad_norm": 0.05085314065217972,
+      "learning_rate": 1.270345195476741e-07,
+      "loss": 0.3729,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9515738498789347,
+      "grad_norm": 0.048932112753391266,
+      "learning_rate": 1.229120767927694e-07,
+      "loss": 0.3084,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.05446094647049904,
+      "learning_rate": 1.188572178060754e-07,
+      "loss": 0.3768,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9531880548829701,
+      "grad_norm": 0.057141583412885666,
+      "learning_rate": 1.1486997033288038e-07,
+      "loss": 0.3922,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9539951573849879,
+      "grad_norm": 0.05461932346224785,
+      "learning_rate": 1.1095036165584272e-07,
+      "loss": 0.3816,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9548022598870056,
+      "grad_norm": 0.052471328526735306,
+      "learning_rate": 1.0709841859480097e-07,
+      "loss": 0.3956,
+      "step": 1183
+    },
+    {
+      "epoch": 0.9556093623890234,
+      "grad_norm": 0.043329499661922455,
+      "learning_rate": 1.0331416750659851e-07,
+      "loss": 0.3205,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9564164648910412,
+      "grad_norm": 0.04855290800333023,
+      "learning_rate": 9.959763428489256e-08,
+      "loss": 0.3544,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9572235673930589,
+      "grad_norm": 0.05016739293932915,
+      "learning_rate": 9.59488443599843e-08,
+      "loss": 0.3534,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9580306698950767,
+      "grad_norm": 0.05878947302699089,
+      "learning_rate": 9.236782269864242e-08,
+      "loss": 0.3834,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9588377723970944,
+      "grad_norm": 0.051569633185863495,
+      "learning_rate": 8.885459380393313e-08,
+      "loss": 0.3874,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9596448748991122,
+      "grad_norm": 0.05080967769026756,
+      "learning_rate": 8.540918171505374e-08,
+      "loss": 0.3476,
+      "step": 1189
+    },
+    {
+      "epoch": 0.96045197740113,
+      "grad_norm": 0.04376181587576866,
+      "learning_rate": 8.203161000716164e-08,
+      "loss": 0.3373,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9612590799031477,
+      "grad_norm": 0.055172570049762726,
+      "learning_rate": 7.87219017912233e-08,
+      "loss": 0.3961,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9620661824051655,
+      "grad_norm": 0.04730622097849846,
+      "learning_rate": 7.548007971384774e-08,
+      "loss": 0.3841,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9628732849071832,
+      "grad_norm": 0.06012022867798805,
+      "learning_rate": 7.230616595713336e-08,
+      "loss": 0.4317,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9636803874092009,
+      "grad_norm": 0.048860322684049606,
+      "learning_rate": 6.920018223852021e-08,
+      "loss": 0.3674,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9644874899112187,
+      "grad_norm": 0.05125834792852402,
+      "learning_rate": 6.616214981063685e-08,
+      "loss": 0.361,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9652945924132365,
+      "grad_norm": 0.048129141330718994,
+      "learning_rate": 6.319208946115817e-08,
+      "loss": 0.4051,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9661016949152542,
+      "grad_norm": 0.05246327072381973,
+      "learning_rate": 6.029002151265673e-08,
+      "loss": 0.3963,
+      "step": 1197
+    },
+    {
+      "epoch": 0.966908797417272,
+      "grad_norm": 0.04887653514742851,
+      "learning_rate": 5.745596582247603e-08,
+      "loss": 0.4077,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9677158999192897,
+      "grad_norm": 0.04455339536070824,
+      "learning_rate": 5.46899417825808e-08,
+      "loss": 0.2854,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9685230024213075,
+      "grad_norm": 0.04976993426680565,
+      "learning_rate": 5.199196831943254e-08,
+      "loss": 0.3683,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9693301049233253,
+      "grad_norm": 0.050257232040166855,
+      "learning_rate": 4.936206389385967e-08,
+      "loss": 0.3282,
+      "step": 1201
+    },
+    {
+      "epoch": 0.970137207425343,
+      "grad_norm": 0.05376894772052765,
+      "learning_rate": 4.680024650092652e-08,
+      "loss": 0.3623,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9709443099273608,
+      "grad_norm": 0.048789188265800476,
+      "learning_rate": 4.430653366981785e-08,
+      "loss": 0.3376,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9717514124293786,
+      "grad_norm": 0.04705846682190895,
+      "learning_rate": 4.1880942463710104e-08,
+      "loss": 0.3704,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9725585149313963,
+      "grad_norm": 0.049898602068424225,
+      "learning_rate": 3.9523489479661447e-08,
+      "loss": 0.3926,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9733656174334141,
+      "grad_norm": 0.05981026589870453,
+      "learning_rate": 3.7234190848493e-08,
+      "loss": 0.3665,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9741727199354318,
+      "grad_norm": 0.04264838621020317,
+      "learning_rate": 3.501306223468448e-08,
+      "loss": 0.3608,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9749798224374495,
+      "grad_norm": 0.052027132362127304,
+      "learning_rate": 3.286011883626206e-08,
+      "loss": 0.3859,
+      "step": 1208
+    },
+    {
+      "epoch": 0.9757869249394673,
+      "grad_norm": 0.05197995901107788,
+      "learning_rate": 3.077537538469622e-08,
+      "loss": 0.4028,
+      "step": 1209
+    },
+    {
+      "epoch": 0.976594027441485,
+      "grad_norm": 0.05613158270716667,
+      "learning_rate": 2.8758846144799624e-08,
+      "loss": 0.4116,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9774011299435028,
+      "grad_norm": 0.05353546887636185,
+      "learning_rate": 2.6810544914631644e-08,
+      "loss": 0.3928,
+      "step": 1211
+    },
+    {
+      "epoch": 0.9782082324455206,
+      "grad_norm": 0.051475610584020615,
+      "learning_rate": 2.493048502540285e-08,
+      "loss": 0.3398,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9790153349475383,
+      "grad_norm": 0.05530683323740959,
+      "learning_rate": 2.311867934138512e-08,
+      "loss": 0.4251,
+      "step": 1213
+    },
+    {
+      "epoch": 0.9798224374495561,
+      "grad_norm": 0.05283194035291672,
+      "learning_rate": 2.137514025981946e-08,
+      "loss": 0.3988,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9806295399515739,
+      "grad_norm": 0.054351918399333954,
+      "learning_rate": 1.96998797108372e-08,
+      "loss": 0.4202,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9814366424535916,
+      "grad_norm": 0.047337841242551804,
+      "learning_rate": 1.8092909157372272e-08,
+      "loss": 0.3381,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9822437449556094,
+      "grad_norm": 0.05288921296596527,
+      "learning_rate": 1.655423959508795e-08,
+      "loss": 0.3813,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9830508474576272,
+      "grad_norm": 0.056825585663318634,
+      "learning_rate": 1.5083881552299118e-08,
+      "loss": 0.4856,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9838579499596449,
+      "grad_norm": 0.05358508601784706,
+      "learning_rate": 1.3681845089897895e-08,
+      "loss": 0.3537,
+      "step": 1219
+    },
+    {
+      "epoch": 0.9846650524616626,
+      "grad_norm": 0.052583999931812286,
+      "learning_rate": 1.234813980129257e-08,
+      "loss": 0.4317,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9854721549636803,
+      "grad_norm": 0.04891924932599068,
+      "learning_rate": 1.1082774812330998e-08,
+      "loss": 0.3668,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9862792574656981,
+      "grad_norm": 0.0513627864420414,
+      "learning_rate": 9.885758781249532e-09,
+      "loss": 0.3821,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9870863599677159,
+      "grad_norm": 0.048382095992565155,
+      "learning_rate": 8.757099898603071e-09,
+      "loss": 0.326,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9878934624697336,
+      "grad_norm": 0.04995129629969597,
+      "learning_rate": 7.696805887217329e-09,
+      "loss": 0.3843,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9887005649717514,
+      "grad_norm": 0.04516911506652832,
+      "learning_rate": 6.704884002132206e-09,
+      "loss": 0.3553,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9895076674737692,
+      "grad_norm": 0.05033162236213684,
+      "learning_rate": 5.781341030551835e-09,
+      "loss": 0.3908,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9903147699757869,
+      "grad_norm": 0.05497167259454727,
+      "learning_rate": 4.926183291800169e-09,
+      "loss": 0.4237,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9911218724778047,
+      "grad_norm": 0.050287820398807526,
+      "learning_rate": 4.139416637276572e-09,
+      "loss": 0.3525,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9919289749798225,
+      "grad_norm": 0.054579704999923706,
+      "learning_rate": 3.4210464504158547e-09,
+      "loss": 0.4267,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9927360774818402,
+      "grad_norm": 0.057175084948539734,
+      "learning_rate": 2.7710776466494114e-09,
+      "loss": 0.3782,
+      "step": 1230
+    },
+    {
+      "epoch": 0.993543179983858,
+      "grad_norm": 0.052831992506980896,
+      "learning_rate": 2.1895146733763583e-09,
+      "loss": 0.3787,
+      "step": 1231
+    },
+    {
+      "epoch": 0.9943502824858758,
+      "grad_norm": 0.04933362454175949,
+      "learning_rate": 1.6763615099291142e-09,
+      "loss": 0.3536,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9951573849878934,
+      "grad_norm": 0.05213562771677971,
+      "learning_rate": 1.2316216675467563e-09,
+      "loss": 0.3125,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9959644874899112,
+      "grad_norm": 0.051006123423576355,
+      "learning_rate": 8.552981893539259e-10,
+      "loss": 0.3783,
+      "step": 1234
+    },
+    {
+      "epoch": 0.9967715899919289,
+      "grad_norm": 0.05340098217129707,
+      "learning_rate": 5.473936503341826e-10,
+      "loss": 0.4189,
+      "step": 1235
+    },
+    {
+      "epoch": 0.9975786924939467,
+      "grad_norm": 0.05506238713860512,
+      "learning_rate": 3.0791015732001274e-10,
+      "loss": 0.3906,
+      "step": 1236
+    },
+    {
+      "epoch": 0.9983857949959645,
+      "grad_norm": 0.05131802707910538,
+      "learning_rate": 1.3684934897062462e-10,
+      "loss": 0.3888,
+      "step": 1237
+    },
+    {
+      "epoch": 0.9991928974979822,
+      "grad_norm": 0.051829464733600616,
+      "learning_rate": 3.421239576750779e-11,
+      "loss": 0.3569,
+      "step": 1238
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.04889125004410744,
+      "learning_rate": 0.0,
+      "loss": 0.364,
+      "step": 1239
+    },
+    {
+      "epoch": 0.8093994778067886,
+      "grad_norm": 0.05489634722471237,
+      "learning_rate": 1.8457022888197873e-06,
+      "loss": 0.5747,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8100522193211488,
+      "grad_norm": 0.05567055940628052,
+      "learning_rate": 1.8334827698003644e-06,
+      "loss": 0.5034,
+      "step": 1241
+    },
+    {
+      "epoch": 0.8107049608355091,
+      "grad_norm": 0.052047327160835266,
+      "learning_rate": 1.8212997512837815e-06,
+      "loss": 0.569,
+      "step": 1242
+    },
+    {
+      "epoch": 0.8113577023498695,
+      "grad_norm": 0.053265251219272614,
+      "learning_rate": 1.8091532877224194e-06,
+      "loss": 0.5673,
+      "step": 1243
+    },
+    {
+      "epoch": 0.8120104438642297,
+      "grad_norm": 0.05723145231604576,
+      "learning_rate": 1.7970434334052822e-06,
+      "loss": 0.5226,
+      "step": 1244
+    },
+    {
+      "epoch": 0.8126631853785901,
+      "grad_norm": 0.056260693818330765,
+      "learning_rate": 1.784970242457732e-06,
+      "loss": 0.5682,
+      "step": 1245
+    },
+    {
+      "epoch": 0.8133159268929504,
+      "grad_norm": 0.06144409254193306,
+      "learning_rate": 1.7729337688412772e-06,
+      "loss": 0.5818,
+      "step": 1246
+    },
+    {
+      "epoch": 0.8139686684073107,
+      "grad_norm": 0.05530359968543053,
+      "learning_rate": 1.7609340663533115e-06,
+      "loss": 0.5372,
+      "step": 1247
+    },
+    {
+      "epoch": 0.814621409921671,
+      "grad_norm": 0.052408602088689804,
+      "learning_rate": 1.7489711886268713e-06,
+      "loss": 0.5431,
+      "step": 1248
+    },
+    {
+      "epoch": 0.8152741514360313,
+      "grad_norm": 0.04931781440973282,
+      "learning_rate": 1.7370451891304208e-06,
+      "loss": 0.5209,
+      "step": 1249
+    },
+    {
+      "epoch": 0.8159268929503917,
+      "grad_norm": 0.052259039133787155,
+      "learning_rate": 1.725156121167576e-06,
+      "loss": 0.5618,
+      "step": 1250
+    },
+    {
+      "epoch": 0.816579634464752,
+      "grad_norm": 0.054095637053251266,
+      "learning_rate": 1.7133040378769039e-06,
+      "loss": 0.4872,
+      "step": 1251
+    },
+    {
+      "epoch": 0.8172323759791122,
+      "grad_norm": 0.05096259340643883,
+      "learning_rate": 1.701488992231658e-06,
+      "loss": 0.5254,
+      "step": 1252
+    },
+    {
+      "epoch": 0.8178851174934726,
+      "grad_norm": 0.04998144507408142,
+      "learning_rate": 1.6897110370395452e-06,
+      "loss": 0.4692,
+      "step": 1253
+    },
+    {
+      "epoch": 0.8185378590078329,
+      "grad_norm": 0.05502447485923767,
+      "learning_rate": 1.6779702249425122e-06,
+      "loss": 0.4983,
+      "step": 1254
+    },
+    {
+      "epoch": 0.8191906005221932,
+      "grad_norm": 0.05895658954977989,
+      "learning_rate": 1.6662666084164791e-06,
+      "loss": 0.5427,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8198433420365535,
+      "grad_norm": 0.05542084947228432,
+      "learning_rate": 1.6546002397711247e-06,
+      "loss": 0.5164,
+      "step": 1256
+    },
+    {
+      "epoch": 0.8204960835509139,
+      "grad_norm": 0.053591158241033554,
+      "learning_rate": 1.6429711711496499e-06,
+      "loss": 0.491,
+      "step": 1257
+    },
+    {
+      "epoch": 0.8211488250652742,
+      "grad_norm": 0.051953479647636414,
+      "learning_rate": 1.6313794545285312e-06,
+      "loss": 0.5057,
+      "step": 1258
+    },
+    {
+      "epoch": 0.8218015665796344,
+      "grad_norm": 0.05032678321003914,
+      "learning_rate": 1.6198251417173138e-06,
+      "loss": 0.4555,
+      "step": 1259
+    },
+    {
+      "epoch": 0.8224543080939948,
+      "grad_norm": 0.054147832095623016,
+      "learning_rate": 1.6083082843583552e-06,
+      "loss": 0.5148,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8231070496083551,
+      "grad_norm": 0.056531306356191635,
+      "learning_rate": 1.5968289339266084e-06,
+      "loss": 0.5749,
+      "step": 1261
+    },
+    {
+      "epoch": 0.8237597911227154,
+      "grad_norm": 0.04894256219267845,
+      "learning_rate": 1.5853871417293876e-06,
+      "loss": 0.4717,
+      "step": 1262
+    },
+    {
+      "epoch": 0.8244125326370757,
+      "grad_norm": 0.057365089654922485,
+      "learning_rate": 1.5739829589061384e-06,
+      "loss": 0.5384,
+      "step": 1263
+    },
+    {
+      "epoch": 0.825065274151436,
+      "grad_norm": 0.05126437544822693,
+      "learning_rate": 1.5626164364282103e-06,
+      "loss": 0.4903,
+      "step": 1264
+    },
+    {
+      "epoch": 0.8257180156657964,
+      "grad_norm": 0.06285741925239563,
+      "learning_rate": 1.5512876250986308e-06,
+      "loss": 0.6048,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8263707571801566,
+      "grad_norm": 0.05392901971936226,
+      "learning_rate": 1.539996575551872e-06,
+      "loss": 0.5462,
+      "step": 1266
+    },
+    {
+      "epoch": 0.827023498694517,
+      "grad_norm": 0.053971629589796066,
+      "learning_rate": 1.5287433382536342e-06,
+      "loss": 0.5608,
+      "step": 1267
+    },
+    {
+      "epoch": 0.8276762402088773,
+      "grad_norm": 0.046124279499053955,
+      "learning_rate": 1.5175279635006102e-06,
+      "loss": 0.4852,
+      "step": 1268
+    },
+    {
+      "epoch": 0.8283289817232375,
+      "grad_norm": 0.056769512593746185,
+      "learning_rate": 1.5063505014202651e-06,
+      "loss": 0.5403,
+      "step": 1269
+    },
+    {
+      "epoch": 0.8289817232375979,
+      "grad_norm": 0.05687325447797775,
+      "learning_rate": 1.4952110019706156e-06,
+      "loss": 0.5591,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8296344647519582,
+      "grad_norm": 0.05294881388545036,
+      "learning_rate": 1.4841095149399998e-06,
+      "loss": 0.4884,
+      "step": 1271
+    },
+    {
+      "epoch": 0.8302872062663186,
+      "grad_norm": 0.05597991868853569,
+      "learning_rate": 1.4730460899468601e-06,
+      "loss": 0.5337,
+      "step": 1272
+    },
+    {
+      "epoch": 0.8309399477806788,
+      "grad_norm": 0.05722576007246971,
+      "learning_rate": 1.4620207764395177e-06,
+      "loss": 0.4941,
+      "step": 1273
+    },
+    {
+      "epoch": 0.8315926892950392,
+      "grad_norm": 0.05292990803718567,
+      "learning_rate": 1.4510336236959554e-06,
+      "loss": 0.474,
+      "step": 1274
+    },
+    {
+      "epoch": 0.8322454308093995,
+      "grad_norm": 0.05963551253080368,
+      "learning_rate": 1.4400846808235946e-06,
+      "loss": 0.6022,
+      "step": 1275
+    },
+    {
+      "epoch": 0.8328981723237598,
+      "grad_norm": 0.0566948726773262,
+      "learning_rate": 1.4291739967590746e-06,
+      "loss": 0.5265,
+      "step": 1276
+    },
+    {
+      "epoch": 0.8335509138381201,
+      "grad_norm": 0.05305665358901024,
+      "learning_rate": 1.4183016202680378e-06,
+      "loss": 0.5592,
+      "step": 1277
+    },
+    {
+      "epoch": 0.8342036553524804,
+      "grad_norm": 0.058979522436857224,
+      "learning_rate": 1.4074675999449095e-06,
+      "loss": 0.5232,
+      "step": 1278
+    },
+    {
+      "epoch": 0.8348563968668408,
+      "grad_norm": 0.05601240321993828,
+      "learning_rate": 1.3966719842126808e-06,
+      "loss": 0.5869,
+      "step": 1279
+    },
+    {
+      "epoch": 0.835509138381201,
+      "grad_norm": 0.05249645933508873,
+      "learning_rate": 1.3859148213226903e-06,
+      "loss": 0.4805,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8361618798955613,
+      "grad_norm": 0.058439917862415314,
+      "learning_rate": 1.3751961593544171e-06,
+      "loss": 0.5394,
+      "step": 1281
+    },
+    {
+      "epoch": 0.8368146214099217,
+      "grad_norm": 0.055136680603027344,
+      "learning_rate": 1.3645160462152495e-06,
+      "loss": 0.5072,
+      "step": 1282
+    },
+    {
+      "epoch": 0.837467362924282,
+      "grad_norm": 0.059607915580272675,
+      "learning_rate": 1.3538745296402867e-06,
+      "loss": 0.5557,
+      "step": 1283
+    },
+    {
+      "epoch": 0.8381201044386423,
+      "grad_norm": 0.0572134405374527,
+      "learning_rate": 1.3432716571921178e-06,
+      "loss": 0.51,
+      "step": 1284
+    },
+    {
+      "epoch": 0.8387728459530026,
+      "grad_norm": 0.053891804069280624,
+      "learning_rate": 1.3327074762606096e-06,
+      "loss": 0.5298,
+      "step": 1285
+    },
+    {
+      "epoch": 0.839425587467363,
+      "grad_norm": 0.0589805468916893,
+      "learning_rate": 1.3221820340627044e-06,
+      "loss": 0.6048,
+      "step": 1286
+    },
+    {
+      "epoch": 0.8400783289817232,
+      "grad_norm": 0.05279667675495148,
+      "learning_rate": 1.311695377642187e-06,
+      "loss": 0.4724,
+      "step": 1287
+    },
+    {
+      "epoch": 0.8407310704960835,
+      "grad_norm": 0.052640657871961594,
+      "learning_rate": 1.3012475538694935e-06,
+      "loss": 0.4407,
+      "step": 1288
+    },
+    {
+      "epoch": 0.8413838120104439,
+      "grad_norm": 0.0526045598089695,
+      "learning_rate": 1.2908386094415049e-06,
+      "loss": 0.5229,
+      "step": 1289
+    },
+    {
+      "epoch": 0.8420365535248042,
+      "grad_norm": 0.0544457770884037,
+      "learning_rate": 1.280468590881312e-06,
+      "loss": 0.4922,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8426892950391645,
+      "grad_norm": 0.054909445345401764,
+      "learning_rate": 1.2701375445380459e-06,
+      "loss": 0.5311,
+      "step": 1291
+    },
+    {
+      "epoch": 0.8433420365535248,
+      "grad_norm": 0.056963834911584854,
+      "learning_rate": 1.2598455165866319e-06,
+      "loss": 0.5472,
+      "step": 1292
+    },
+    {
+      "epoch": 0.8439947780678851,
+      "grad_norm": 0.05432257801294327,
+      "learning_rate": 1.24959255302761e-06,
+      "loss": 0.4912,
+      "step": 1293
+    },
+    {
+      "epoch": 0.8446475195822454,
+      "grad_norm": 0.0577394962310791,
+      "learning_rate": 1.2393786996869262e-06,
+      "loss": 0.4972,
+      "step": 1294
+    },
+    {
+      "epoch": 0.8453002610966057,
+      "grad_norm": 0.05427301302552223,
+      "learning_rate": 1.229204002215706e-06,
+      "loss": 0.5027,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8459530026109661,
+      "grad_norm": 0.052240319550037384,
+      "learning_rate": 1.2190685060900843e-06,
+      "loss": 0.5267,
+      "step": 1296
+    },
+    {
+      "epoch": 0.8466057441253264,
+      "grad_norm": 0.053541723638772964,
+      "learning_rate": 1.2089722566109752e-06,
+      "loss": 0.5539,
+      "step": 1297
+    },
+    {
+      "epoch": 0.8472584856396866,
+      "grad_norm": 0.054370395839214325,
+      "learning_rate": 1.1989152989038745e-06,
+      "loss": 0.5465,
+      "step": 1298
+    },
+    {
+      "epoch": 0.847911227154047,
+      "grad_norm": 0.052129361778497696,
+      "learning_rate": 1.1888976779186745e-06,
+      "loss": 0.5215,
+      "step": 1299
+    },
+    {
+      "epoch": 0.8485639686684073,
+      "grad_norm": 0.0562812015414238,
+      "learning_rate": 1.1789194384294377e-06,
+      "loss": 0.4823,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8492167101827677,
+      "grad_norm": 0.05429472029209137,
+      "learning_rate": 1.1689806250342196e-06,
+      "loss": 0.5207,
+      "step": 1301
+    },
+    {
+      "epoch": 0.8498694516971279,
+      "grad_norm": 0.05133857950568199,
+      "learning_rate": 1.159081282154858e-06,
+      "loss": 0.4531,
+      "step": 1302
+    },
+    {
+      "epoch": 0.8505221932114883,
+      "grad_norm": 0.05682549253106117,
+      "learning_rate": 1.1492214540367674e-06,
+      "loss": 0.5547,
+      "step": 1303
+    },
+    {
+      "epoch": 0.8511749347258486,
+      "grad_norm": 0.05034679174423218,
+      "learning_rate": 1.1394011847487618e-06,
+      "loss": 0.5243,
+      "step": 1304
+    },
+    {
+      "epoch": 0.8518276762402088,
+      "grad_norm": 0.056760791689157486,
+      "learning_rate": 1.1296205181828378e-06,
+      "loss": 0.5835,
+      "step": 1305
+    },
+    {
+      "epoch": 0.8524804177545692,
+      "grad_norm": 0.054008033126592636,
+      "learning_rate": 1.1198794980539908e-06,
+      "loss": 0.4983,
+      "step": 1306
+    },
+    {
+      "epoch": 0.8531331592689295,
+      "grad_norm": 0.05667291581630707,
+      "learning_rate": 1.1101781679000135e-06,
+      "loss": 0.5211,
+      "step": 1307
+    },
+    {
+      "epoch": 0.8537859007832899,
+      "grad_norm": 0.05850447714328766,
+      "learning_rate": 1.1005165710812982e-06,
+      "loss": 0.5183,
+      "step": 1308
+    },
+    {
+      "epoch": 0.8544386422976501,
+      "grad_norm": 0.0570049025118351,
+      "learning_rate": 1.0908947507806567e-06,
+      "loss": 0.4721,
+      "step": 1309
+    },
+    {
+      "epoch": 0.8550913838120104,
+      "grad_norm": 0.05443601682782173,
+      "learning_rate": 1.081312750003114e-06,
+      "loss": 0.5579,
+      "step": 1310
+    },
+    {
+      "epoch": 0.8557441253263708,
+      "grad_norm": 0.05229130759835243,
+      "learning_rate": 1.071770611575721e-06,
+      "loss": 0.4809,
+      "step": 1311
+    },
+    {
+      "epoch": 0.856396866840731,
+      "grad_norm": 0.05782579630613327,
+      "learning_rate": 1.0622683781473596e-06,
+      "loss": 0.4737,
+      "step": 1312
+    },
+    {
+      "epoch": 0.8570496083550914,
+      "grad_norm": 0.05550665035843849,
+      "learning_rate": 1.0528060921885607e-06,
+      "loss": 0.5707,
+      "step": 1313
+    },
+    {
+      "epoch": 0.8577023498694517,
+      "grad_norm": 0.05818123370409012,
+      "learning_rate": 1.043383795991304e-06,
+      "loss": 0.5846,
+      "step": 1314
+    },
+    {
+      "epoch": 0.858355091383812,
+      "grad_norm": 0.05427548661828041,
+      "learning_rate": 1.0340015316688358e-06,
+      "loss": 0.5087,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8590078328981723,
+      "grad_norm": 0.05885540693998337,
+      "learning_rate": 1.0246593411554796e-06,
+      "loss": 0.4413,
+      "step": 1316
+    },
+    {
+      "epoch": 0.8596605744125326,
+      "grad_norm": 0.05844886973500252,
+      "learning_rate": 1.0153572662064448e-06,
+      "loss": 0.5454,
+      "step": 1317
+    },
+    {
+      "epoch": 0.860313315926893,
+      "grad_norm": 0.054623618721961975,
+      "learning_rate": 1.0060953483976454e-06,
+      "loss": 0.5637,
+      "step": 1318
+    },
+    {
+      "epoch": 0.8609660574412533,
+      "grad_norm": 0.064220130443573,
+      "learning_rate": 9.968736291255122e-07,
+      "loss": 0.5217,
+      "step": 1319
+    },
+    {
+      "epoch": 0.8616187989556136,
+      "grad_norm": 0.05971851199865341,
+      "learning_rate": 9.87692149606806e-07,
+      "loss": 0.4555,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8622715404699739,
+      "grad_norm": 0.053942758589982986,
+      "learning_rate": 9.78550950878433e-07,
+      "loss": 0.5939,
+      "step": 1321
+    },
+    {
+      "epoch": 0.8629242819843342,
+      "grad_norm": 0.05949598550796509,
+      "learning_rate": 9.69450073797269e-07,
+      "loss": 0.4948,
+      "step": 1322
+    },
+    {
+      "epoch": 0.8635770234986945,
+      "grad_norm": 0.05386824905872345,
+      "learning_rate": 9.603895590399648e-07,
+      "loss": 0.4731,
+      "step": 1323
+    },
+    {
+      "epoch": 0.8642297650130548,
+      "grad_norm": 0.05269710347056389,
+      "learning_rate": 9.513694471027735e-07,
+      "loss": 0.4558,
+      "step": 1324
+    },
+    {
+      "epoch": 0.8648825065274152,
+      "grad_norm": 0.05899108573794365,
+      "learning_rate": 9.423897783013658e-07,
+      "loss": 0.5939,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8655352480417755,
+      "grad_norm": 0.055841196328401566,
+      "learning_rate": 9.334505927706516e-07,
+      "loss": 0.4845,
+      "step": 1326
+    },
+    {
+      "epoch": 0.8661879895561357,
+      "grad_norm": 0.055123697966337204,
+      "learning_rate": 9.245519304645978e-07,
+      "loss": 0.4934,
+      "step": 1327
+    },
+    {
+      "epoch": 0.8668407310704961,
+      "grad_norm": 0.053594909608364105,
+      "learning_rate": 9.156938311560526e-07,
+      "loss": 0.5103,
+      "step": 1328
+    },
+    {
+      "epoch": 0.8674934725848564,
+      "grad_norm": 0.053080540150403976,
+      "learning_rate": 9.068763344365683e-07,
+      "loss": 0.428,
+      "step": 1329
+    },
+    {
+      "epoch": 0.8681462140992167,
+      "grad_norm": 0.04971330985426903,
+      "learning_rate": 8.980994797162157e-07,
+      "loss": 0.4763,
+      "step": 1330
+    },
+    {
+      "epoch": 0.868798955613577,
+      "grad_norm": 0.056616924703121185,
+      "learning_rate": 8.893633062234285e-07,
+      "loss": 0.5279,
+      "step": 1331
+    },
+    {
+      "epoch": 0.8694516971279374,
+      "grad_norm": 0.05742453783750534,
+      "learning_rate": 8.806678530047985e-07,
+      "loss": 0.4932,
+      "step": 1332
+    },
+    {
+      "epoch": 0.8701044386422977,
+      "grad_norm": 0.05315115302801132,
+      "learning_rate": 8.720131589249281e-07,
+      "loss": 0.4897,
+      "step": 1333
+    },
+    {
+      "epoch": 0.8707571801566579,
+      "grad_norm": 0.057442694902420044,
+      "learning_rate": 8.633992626662402e-07,
+      "loss": 0.5205,
+      "step": 1334
+    },
+    {
+      "epoch": 0.8714099216710183,
+      "grad_norm": 0.05508160963654518,
+      "learning_rate": 8.548262027288101e-07,
+      "loss": 0.4926,
+      "step": 1335
+    },
+    {
+      "epoch": 0.8720626631853786,
+      "grad_norm": 0.05157114192843437,
+      "learning_rate": 8.462940174302026e-07,
+      "loss": 0.5125,
+      "step": 1336
+    },
+    {
+      "epoch": 0.8727154046997389,
+      "grad_norm": 0.05958451330661774,
+      "learning_rate": 8.378027449052784e-07,
+      "loss": 0.5025,
+      "step": 1337
+    },
+    {
+      "epoch": 0.8733681462140992,
+      "grad_norm": 0.055542390793561935,
+      "learning_rate": 8.293524231060468e-07,
+      "loss": 0.4989,
+      "step": 1338
+    },
+    {
+      "epoch": 0.8740208877284595,
+      "grad_norm": 0.06172487139701843,
+      "learning_rate": 8.20943089801487e-07,
+      "loss": 0.537,
+      "step": 1339
+    },
+    {
+      "epoch": 0.8746736292428199,
+      "grad_norm": 0.05982014536857605,
+      "learning_rate": 8.125747825773689e-07,
+      "loss": 0.5673,
+      "step": 1340
+    },
+    {
+      "epoch": 0.8753263707571801,
+      "grad_norm": 0.05253879725933075,
+      "learning_rate": 8.042475388361104e-07,
+      "loss": 0.4503,
+      "step": 1341
+    },
+    {
+      "epoch": 0.8759791122715405,
+      "grad_norm": 0.0581713505089283,
+      "learning_rate": 7.959613957965794e-07,
+      "loss": 0.5507,
+      "step": 1342
+    },
+    {
+      "epoch": 0.8766318537859008,
+      "grad_norm": 0.05596986040472984,
+      "learning_rate": 7.877163904939522e-07,
+      "loss": 0.5085,
+      "step": 1343
+    },
+    {
+      "epoch": 0.8772845953002611,
+      "grad_norm": 0.054406262934207916,
+      "learning_rate": 7.795125597795406e-07,
+      "loss": 0.5056,
+      "step": 1344
+    },
+    {
+      "epoch": 0.8779373368146214,
+      "grad_norm": 0.05836174637079239,
+      "learning_rate": 7.713499403206138e-07,
+      "loss": 0.5326,
+      "step": 1345
+    },
+    {
+      "epoch": 0.8785900783289817,
+      "grad_norm": 0.05304201692342758,
+      "learning_rate": 7.632285686002594e-07,
+      "loss": 0.4556,
+      "step": 1346
+    },
+    {
+      "epoch": 0.8792428198433421,
+      "grad_norm": 0.05944960191845894,
+      "learning_rate": 7.551484809172005e-07,
+      "loss": 0.5656,
+      "step": 1347
+    },
+    {
+      "epoch": 0.8798955613577023,
+      "grad_norm": 0.062106695026159286,
+      "learning_rate": 7.471097133856353e-07,
+      "loss": 0.562,
+      "step": 1348
+    },
+    {
+      "epoch": 0.8805483028720626,
+      "grad_norm": 0.05966848507523537,
+      "learning_rate": 7.391123019350932e-07,
+      "loss": 0.5701,
+      "step": 1349
+    },
+    {
+      "epoch": 0.881201044386423,
+      "grad_norm": 0.05123576894402504,
+      "learning_rate": 7.311562823102469e-07,
+      "loss": 0.5516,
+      "step": 1350
+    },
+    {
+      "epoch": 0.8818537859007833,
+      "grad_norm": 0.056662097573280334,
+      "learning_rate": 7.232416900707739e-07,
+      "loss": 0.5026,
+      "step": 1351
+    },
+    {
+      "epoch": 0.8825065274151436,
+      "grad_norm": 0.05869949981570244,
+      "learning_rate": 7.153685605911964e-07,
+      "loss": 0.5144,
+      "step": 1352
+    },
+    {
+      "epoch": 0.8831592689295039,
+      "grad_norm": 0.05815920606255531,
+      "learning_rate": 7.075369290607049e-07,
+      "loss": 0.5185,
+      "step": 1353
+    },
+    {
+      "epoch": 0.8838120104438643,
+      "grad_norm": 0.05506941303610802,
+      "learning_rate": 6.997468304830247e-07,
+      "loss": 0.579,
+      "step": 1354
+    },
+    {
+      "epoch": 0.8844647519582245,
+      "grad_norm": 0.058971211314201355,
+      "learning_rate": 6.919982996762431e-07,
+      "loss": 0.5332,
+      "step": 1355
+    },
+    {
+      "epoch": 0.8851174934725848,
+      "grad_norm": 0.05661641061306,
+      "learning_rate": 6.842913712726551e-07,
+      "loss": 0.5755,
+      "step": 1356
+    },
+    {
+      "epoch": 0.8857702349869452,
+      "grad_norm": 0.057749152183532715,
+      "learning_rate": 6.766260797186242e-07,
+      "loss": 0.4955,
+      "step": 1357
+    },
+    {
+      "epoch": 0.8864229765013055,
+      "grad_norm": 0.05507257953286171,
+      "learning_rate": 6.690024592744027e-07,
+      "loss": 0.4629,
+      "step": 1358
+    },
+    {
+      "epoch": 0.8870757180156658,
+      "grad_norm": 0.06188316270709038,
+      "learning_rate": 6.614205440140042e-07,
+      "loss": 0.5789,
+      "step": 1359
+    },
+    {
+      "epoch": 0.8877284595300261,
+      "grad_norm": 0.05622296407818794,
+      "learning_rate": 6.538803678250338e-07,
+      "loss": 0.5255,
+      "step": 1360
+    },
+    {
+      "epoch": 0.8883812010443864,
+      "grad_norm": 0.053286001086235046,
+      "learning_rate": 6.463819644085412e-07,
+      "loss": 0.4926,
+      "step": 1361
+    },
+    {
+      "epoch": 0.8890339425587467,
+      "grad_norm": 0.05363309010863304,
+      "learning_rate": 6.389253672788754e-07,
+      "loss": 0.5245,
+      "step": 1362
+    },
+    {
+      "epoch": 0.889686684073107,
+      "grad_norm": 0.05182503163814545,
+      "learning_rate": 6.315106097635304e-07,
+      "loss": 0.5641,
+      "step": 1363
+    },
+    {
+      "epoch": 0.8903394255874674,
+      "grad_norm": 0.054177477955818176,
+      "learning_rate": 6.241377250029934e-07,
+      "loss": 0.537,
+      "step": 1364
+    },
+    {
+      "epoch": 0.8909921671018277,
+      "grad_norm": 0.05669175460934639,
+      "learning_rate": 6.168067459506066e-07,
+      "loss": 0.4641,
+      "step": 1365
+    },
+    {
+      "epoch": 0.891644908616188,
+      "grad_norm": 0.058600328862667084,
+      "learning_rate": 6.095177053724011e-07,
+      "loss": 0.5456,
+      "step": 1366
+    },
+    {
+      "epoch": 0.8922976501305483,
+      "grad_norm": 0.05741265043616295,
+      "learning_rate": 6.022706358469776e-07,
+      "loss": 0.4892,
+      "step": 1367
+    },
+    {
+      "epoch": 0.8929503916449086,
+      "grad_norm": 0.06110072880983353,
+      "learning_rate": 5.950655697653363e-07,
+      "loss": 0.5218,
+      "step": 1368
+    },
+    {
+      "epoch": 0.893603133159269,
+      "grad_norm": 0.061403632164001465,
+      "learning_rate": 5.879025393307436e-07,
+      "loss": 0.5595,
+      "step": 1369
+    },
+    {
+      "epoch": 0.8942558746736292,
+      "grad_norm": 0.05400973930954933,
+      "learning_rate": 5.807815765585878e-07,
+      "loss": 0.5213,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8949086161879896,
+      "grad_norm": 0.055990613996982574,
+      "learning_rate": 5.737027132762341e-07,
+      "loss": 0.5479,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8955613577023499,
+      "grad_norm": 0.05797429010272026,
+      "learning_rate": 5.666659811228803e-07,
+      "loss": 0.5287,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8962140992167101,
+      "grad_norm": 0.05289648100733757,
+      "learning_rate": 5.596714115494217e-07,
+      "loss": 0.4631,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8968668407310705,
+      "grad_norm": 0.05488736554980278,
+      "learning_rate": 5.527190358183032e-07,
+      "loss": 0.5012,
+      "step": 1374
+    },
+    {
+      "epoch": 0.8975195822454308,
+      "grad_norm": 0.05579644441604614,
+      "learning_rate": 5.458088850033849e-07,
+      "loss": 0.488,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8981723237597912,
+      "grad_norm": 0.06353001296520233,
+      "learning_rate": 5.389409899898013e-07,
+      "loss": 0.5669,
+      "step": 1376
+    },
+    {
+      "epoch": 0.8988250652741514,
+      "grad_norm": 0.06207989528775215,
+      "learning_rate": 5.321153814738222e-07,
+      "loss": 0.5379,
+      "step": 1377
+    },
+    {
+      "epoch": 0.8994778067885117,
+      "grad_norm": 0.04865063726902008,
+      "learning_rate": 5.253320899627179e-07,
+      "loss": 0.4628,
+      "step": 1378
+    },
+    {
+      "epoch": 0.9001305483028721,
+      "grad_norm": 0.05397767573595047,
+      "learning_rate": 5.185911457746207e-07,
+      "loss": 0.4546,
+      "step": 1379
+    },
+    {
+      "epoch": 0.9007832898172323,
+      "grad_norm": 0.0543336383998394,
+      "learning_rate": 5.1189257903839e-07,
+      "loss": 0.4474,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9014360313315927,
+      "grad_norm": 0.055923014879226685,
+      "learning_rate": 5.052364196934779e-07,
+      "loss": 0.4864,
+      "step": 1381
+    },
+    {
+      "epoch": 0.902088772845953,
+      "grad_norm": 0.060891781002283096,
+      "learning_rate": 4.986226974897967e-07,
+      "loss": 0.5346,
+      "step": 1382
+    },
+    {
+      "epoch": 0.9027415143603134,
+      "grad_norm": 0.05753452330827713,
+      "learning_rate": 4.920514419875821e-07,
+      "loss": 0.5118,
+      "step": 1383
+    },
+    {
+      "epoch": 0.9033942558746736,
+      "grad_norm": 0.050467561930418015,
+      "learning_rate": 4.855226825572667e-07,
+      "loss": 0.4379,
+      "step": 1384
+    },
+    {
+      "epoch": 0.9040469973890339,
+      "grad_norm": 0.05872552469372749,
+      "learning_rate": 4.79036448379343e-07,
+      "loss": 0.4756,
+      "step": 1385
+    },
+    {
+      "epoch": 0.9046997389033943,
+      "grad_norm": 0.05431582033634186,
+      "learning_rate": 4.725927684442366e-07,
+      "loss": 0.4922,
+      "step": 1386
+    },
+    {
+      "epoch": 0.9053524804177546,
+      "grad_norm": 0.0574292428791523,
+      "learning_rate": 4.661916715521764e-07,
+      "loss": 0.5334,
+      "step": 1387
+    },
+    {
+      "epoch": 0.9060052219321149,
+      "grad_norm": 0.050154268741607666,
+      "learning_rate": 4.598331863130612e-07,
+      "loss": 0.4558,
+      "step": 1388
+    },
+    {
+      "epoch": 0.9066579634464752,
+      "grad_norm": 0.053724534809589386,
+      "learning_rate": 4.535173411463423e-07,
+      "loss": 0.4891,
+      "step": 1389
+    },
+    {
+      "epoch": 0.9073107049608355,
+      "grad_norm": 0.05234300717711449,
+      "learning_rate": 4.472441642808845e-07,
+      "loss": 0.4858,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9079634464751958,
+      "grad_norm": 0.062360599637031555,
+      "learning_rate": 4.410136837548462e-07,
+      "loss": 0.5657,
+      "step": 1391
+    },
+    {
+      "epoch": 0.9086161879895561,
+      "grad_norm": 0.06348732858896255,
+      "learning_rate": 4.348259274155542e-07,
+      "loss": 0.5722,
+      "step": 1392
+    },
+    {
+      "epoch": 0.9092689295039165,
+      "grad_norm": 0.05668312683701515,
+      "learning_rate": 4.286809229193778e-07,
+      "loss": 0.4447,
+      "step": 1393
+    },
+    {
+      "epoch": 0.9099216710182768,
+      "grad_norm": 0.06089024245738983,
+      "learning_rate": 4.225786977316093e-07,
+      "loss": 0.4826,
+      "step": 1394
+    },
+    {
+      "epoch": 0.910574412532637,
+      "grad_norm": 0.05654263123869896,
+      "learning_rate": 4.165192791263295e-07,
+      "loss": 0.4942,
+      "step": 1395
+    },
+    {
+      "epoch": 0.9112271540469974,
+      "grad_norm": 0.058262892067432404,
+      "learning_rate": 4.1050269418629887e-07,
+      "loss": 0.5348,
+      "step": 1396
+    },
+    {
+      "epoch": 0.9118798955613577,
+      "grad_norm": 0.05608157068490982,
+      "learning_rate": 4.045289698028343e-07,
+      "loss": 0.5433,
+      "step": 1397
+    },
+    {
+      "epoch": 0.912532637075718,
+      "grad_norm": 0.05260564386844635,
+      "learning_rate": 3.985981326756794e-07,
+      "loss": 0.4732,
+      "step": 1398
+    },
+    {
+      "epoch": 0.9131853785900783,
+      "grad_norm": 0.056355442851781845,
+      "learning_rate": 3.927102093128976e-07,
+      "loss": 0.512,
+      "step": 1399
+    },
+    {
+      "epoch": 0.9138381201044387,
+      "grad_norm": 0.06535626202821732,
+      "learning_rate": 3.868652260307437e-07,
+      "loss": 0.6469,
+      "step": 1400
+    },
+    {
+      "epoch": 0.914490861618799,
+      "grad_norm": 0.05038583278656006,
+      "learning_rate": 3.810632089535526e-07,
+      "loss": 0.5062,
+      "step": 1401
+    },
+    {
+      "epoch": 0.9151436031331592,
+      "grad_norm": 0.05900431424379349,
+      "learning_rate": 3.753041840136218e-07,
+      "loss": 0.4974,
+      "step": 1402
+    },
+    {
+      "epoch": 0.9157963446475196,
+      "grad_norm": 0.05980895459651947,
+      "learning_rate": 3.6958817695109006e-07,
+      "loss": 0.5418,
+      "step": 1403
+    },
+    {
+      "epoch": 0.9164490861618799,
+      "grad_norm": 0.061525844037532806,
+      "learning_rate": 3.6391521331383126e-07,
+      "loss": 0.6017,
+      "step": 1404
+    },
+    {
+      "epoch": 0.9171018276762402,
+      "grad_norm": 0.053908564150333405,
+      "learning_rate": 3.5828531845733206e-07,
+      "loss": 0.4829,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9177545691906005,
+      "grad_norm": 0.05936963111162186,
+      "learning_rate": 3.526985175445796e-07,
+      "loss": 0.6039,
+      "step": 1406
+    },
+    {
+      "epoch": 0.9184073107049608,
+      "grad_norm": 0.05012655258178711,
+      "learning_rate": 3.4715483554595974e-07,
+      "loss": 0.5029,
+      "step": 1407
+    },
+    {
+      "epoch": 0.9190600522193212,
+      "grad_norm": 0.05771825835108757,
+      "learning_rate": 3.416542972391268e-07,
+      "loss": 0.5213,
+      "step": 1408
+    },
+    {
+      "epoch": 0.9197127937336814,
+      "grad_norm": 0.05283350124955177,
+      "learning_rate": 3.361969272089116e-07,
+      "loss": 0.555,
+      "step": 1409
+    },
+    {
+      "epoch": 0.9203655352480418,
+      "grad_norm": 0.053370680660009384,
+      "learning_rate": 3.3078274984719825e-07,
+      "loss": 0.5161,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9210182767624021,
+      "grad_norm": 0.054678335785865784,
+      "learning_rate": 3.254117893528186e-07,
+      "loss": 0.5444,
+      "step": 1411
+    },
+    {
+      "epoch": 0.9216710182767625,
+      "grad_norm": 0.05304795876145363,
+      "learning_rate": 3.2008406973145e-07,
+      "loss": 0.4347,
+      "step": 1412
+    },
+    {
+      "epoch": 0.9223237597911227,
+      "grad_norm": 0.052384186536073685,
+      "learning_rate": 3.147996147955012e-07,
+      "loss": 0.5218,
+      "step": 1413
+    },
+    {
+      "epoch": 0.922976501305483,
+      "grad_norm": 0.05954265594482422,
+      "learning_rate": 3.095584481640068e-07,
+      "loss": 0.5493,
+      "step": 1414
+    },
+    {
+      "epoch": 0.9236292428198434,
+      "grad_norm": 0.055753324180841446,
+      "learning_rate": 3.04360593262526e-07,
+      "loss": 0.4741,
+      "step": 1415
+    },
+    {
+      "epoch": 0.9242819843342036,
+      "grad_norm": 0.06900015473365784,
+      "learning_rate": 2.9920607332302844e-07,
+      "loss": 0.5573,
+      "step": 1416
+    },
+    {
+      "epoch": 0.924934725848564,
+      "grad_norm": 0.05286920815706253,
+      "learning_rate": 2.9409491138380655e-07,
+      "loss": 0.5286,
+      "step": 1417
+    },
+    {
+      "epoch": 0.9255874673629243,
+      "grad_norm": 0.055355388671159744,
+      "learning_rate": 2.8902713028935546e-07,
+      "loss": 0.5284,
+      "step": 1418
+    },
+    {
+      "epoch": 0.9262402088772846,
+      "grad_norm": 0.05673305317759514,
+      "learning_rate": 2.840027526902811e-07,
+      "loss": 0.5191,
+      "step": 1419
+    },
+    {
+      "epoch": 0.9268929503916449,
+      "grad_norm": 0.05061788111925125,
+      "learning_rate": 2.7902180104319443e-07,
+      "loss": 0.4535,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9275456919060052,
+      "grad_norm": 0.059594254940748215,
+      "learning_rate": 2.7408429761061393e-07,
+      "loss": 0.5076,
+      "step": 1421
+    },
+    {
+      "epoch": 0.9281984334203656,
+      "grad_norm": 0.05334387719631195,
+      "learning_rate": 2.691902644608657e-07,
+      "loss": 0.5151,
+      "step": 1422
+    },
+    {
+      "epoch": 0.9288511749347258,
+      "grad_norm": 0.05792006850242615,
+      "learning_rate": 2.643397234679823e-07,
+      "loss": 0.5472,
+      "step": 1423
+    },
+    {
+      "epoch": 0.9295039164490861,
+      "grad_norm": 0.05704867094755173,
+      "learning_rate": 2.5953269631160847e-07,
+      "loss": 0.4783,
+      "step": 1424
+    },
+    {
+      "epoch": 0.9301566579634465,
+      "grad_norm": 0.05878971889615059,
+      "learning_rate": 2.547692044769012e-07,
+      "loss": 0.5099,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9308093994778068,
+      "grad_norm": 0.05621445178985596,
+      "learning_rate": 2.500492692544354e-07,
+      "loss": 0.5338,
+      "step": 1426
+    },
+    {
+      "epoch": 0.9314621409921671,
+      "grad_norm": 0.058878008276224136,
+      "learning_rate": 2.453729117401082e-07,
+      "loss": 0.5617,
+      "step": 1427
+    },
+    {
+      "epoch": 0.9321148825065274,
+      "grad_norm": 0.061294734477996826,
+      "learning_rate": 2.4074015283504504e-07,
+      "loss": 0.5835,
+      "step": 1428
+    },
+    {
+      "epoch": 0.9327676240208878,
+      "grad_norm": 0.05842115730047226,
+      "learning_rate": 2.3615101324550694e-07,
+      "loss": 0.5393,
+      "step": 1429
+    },
+    {
+      "epoch": 0.933420365535248,
+      "grad_norm": 0.055717576295137405,
+      "learning_rate": 2.3160551348279438e-07,
+      "loss": 0.5517,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9340731070496083,
+      "grad_norm": 0.054703693836927414,
+      "learning_rate": 2.2710367386316156e-07,
+      "loss": 0.5908,
+      "step": 1431
+    },
+    {
+      "epoch": 0.9347258485639687,
+      "grad_norm": 0.05739562585949898,
+      "learning_rate": 2.2264551450772e-07,
+      "loss": 0.5172,
+      "step": 1432
+    },
+    {
+      "epoch": 0.935378590078329,
+      "grad_norm": 0.05853067338466644,
+      "learning_rate": 2.1823105534235166e-07,
+      "loss": 0.6115,
+      "step": 1433
+    },
+    {
+      "epoch": 0.9360313315926893,
+      "grad_norm": 0.05493749678134918,
+      "learning_rate": 2.1386031609761937e-07,
+      "loss": 0.4931,
+      "step": 1434
+    },
+    {
+      "epoch": 0.9366840731070496,
+      "grad_norm": 0.056509993970394135,
+      "learning_rate": 2.095333163086777e-07,
+      "loss": 0.5375,
+      "step": 1435
+    },
+    {
+      "epoch": 0.9373368146214099,
+      "grad_norm": 0.05576891824603081,
+      "learning_rate": 2.052500753151876e-07,
+      "loss": 0.4706,
+      "step": 1436
+    },
+    {
+      "epoch": 0.9379895561357703,
+      "grad_norm": 0.05480458587408066,
+      "learning_rate": 2.0101061226122654e-07,
+      "loss": 0.5043,
+      "step": 1437
+    },
+    {
+      "epoch": 0.9386422976501305,
+      "grad_norm": 0.06164632365107536,
+      "learning_rate": 1.9681494609520735e-07,
+      "loss": 0.5588,
+      "step": 1438
+    },
+    {
+      "epoch": 0.9392950391644909,
+      "grad_norm": 0.05780002474784851,
+      "learning_rate": 1.926630955697917e-07,
+      "loss": 0.4877,
+      "step": 1439
+    },
+    {
+      "epoch": 0.9399477806788512,
+      "grad_norm": 0.059480633586645126,
+      "learning_rate": 1.8855507924180337e-07,
+      "loss": 0.5192,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9406005221932114,
+      "grad_norm": 0.06673169136047363,
+      "learning_rate": 1.844909154721497e-07,
+      "loss": 0.5979,
+      "step": 1441
+    },
+    {
+      "epoch": 0.9412532637075718,
+      "grad_norm": 0.058235302567481995,
+      "learning_rate": 1.8047062242573576e-07,
+      "loss": 0.5195,
+      "step": 1442
+    },
+    {
+      "epoch": 0.9419060052219321,
+      "grad_norm": 0.057730354368686676,
+      "learning_rate": 1.764942180713869e-07,
+      "loss": 0.556,
+      "step": 1443
+    },
+    {
+      "epoch": 0.9425587467362925,
+      "grad_norm": 0.054570272564888,
+      "learning_rate": 1.7256172018176864e-07,
+      "loss": 0.5287,
+      "step": 1444
+    },
+    {
+      "epoch": 0.9432114882506527,
+      "grad_norm": 0.06531942635774612,
+      "learning_rate": 1.6867314633330023e-07,
+      "loss": 0.5772,
+      "step": 1445
+    },
+    {
+      "epoch": 0.943864229765013,
+      "grad_norm": 0.0553065724670887,
+      "learning_rate": 1.6482851390608235e-07,
+      "loss": 0.5504,
+      "step": 1446
+    },
+    {
+      "epoch": 0.9445169712793734,
+      "grad_norm": 0.057470377534627914,
+      "learning_rate": 1.6102784008382278e-07,
+      "loss": 0.5519,
+      "step": 1447
+    },
+    {
+      "epoch": 0.9451697127937336,
+      "grad_norm": 0.06070839986205101,
+      "learning_rate": 1.5727114185374758e-07,
+      "loss": 0.5528,
+      "step": 1448
+    },
+    {
+      "epoch": 0.945822454308094,
+      "grad_norm": 0.05690762773156166,
+      "learning_rate": 1.5355843600653896e-07,
+      "loss": 0.5234,
+      "step": 1449
+    },
+    {
+      "epoch": 0.9464751958224543,
+      "grad_norm": 0.05993395298719406,
+      "learning_rate": 1.4988973913625082e-07,
+      "loss": 0.5453,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9471279373368147,
+      "grad_norm": 0.05665702000260353,
+      "learning_rate": 1.4626506764023663e-07,
+      "loss": 0.5268,
+      "step": 1451
+    },
+    {
+      "epoch": 0.9477806788511749,
+      "grad_norm": 0.05812077969312668,
+      "learning_rate": 1.4268443771908058e-07,
+      "loss": 0.5163,
+      "step": 1452
+    },
+    {
+      "epoch": 0.9484334203655352,
+      "grad_norm": 0.06277963519096375,
+      "learning_rate": 1.391478653765177e-07,
+      "loss": 0.5589,
+      "step": 1453
+    },
+    {
+      "epoch": 0.9490861618798956,
+      "grad_norm": 0.055003996938467026,
+      "learning_rate": 1.3565536641936827e-07,
+      "loss": 0.5688,
+      "step": 1454
+    },
+    {
+      "epoch": 0.9497389033942559,
+      "grad_norm": 0.055438101291656494,
+      "learning_rate": 1.3220695645746684e-07,
+      "loss": 0.5138,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9503916449086162,
+      "grad_norm": 0.051719021052122116,
+      "learning_rate": 1.2880265090358668e-07,
+      "loss": 0.4777,
+      "step": 1456
+    },
+    {
+      "epoch": 0.9510443864229765,
+      "grad_norm": 0.05389731004834175,
+      "learning_rate": 1.2544246497337986e-07,
+      "loss": 0.4902,
+      "step": 1457
+    },
+    {
+      "epoch": 0.9516971279373369,
+      "grad_norm": 0.061336226761341095,
+      "learning_rate": 1.2212641368529842e-07,
+      "loss": 0.5617,
+      "step": 1458
+    },
+    {
+      "epoch": 0.9523498694516971,
+      "grad_norm": 0.05890428647398949,
+      "learning_rate": 1.1885451186053886e-07,
+      "loss": 0.5077,
+      "step": 1459
+    },
+    {
+      "epoch": 0.9530026109660574,
+      "grad_norm": 0.058837950229644775,
+      "learning_rate": 1.1562677412296996e-07,
+      "loss": 0.5116,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9536553524804178,
+      "grad_norm": 0.05983196198940277,
+      "learning_rate": 1.1244321489906285e-07,
+      "loss": 0.5378,
+      "step": 1461
+    },
+    {
+      "epoch": 0.9543080939947781,
+      "grad_norm": 0.058248184621334076,
+      "learning_rate": 1.0930384841783548e-07,
+      "loss": 0.5436,
+      "step": 1462
+    },
+    {
+      "epoch": 0.9549608355091384,
+      "grad_norm": 0.05723857879638672,
+      "learning_rate": 1.0620868871078493e-07,
+      "loss": 0.5372,
+      "step": 1463
+    },
+    {
+      "epoch": 0.9556135770234987,
+      "grad_norm": 0.059107016772031784,
+      "learning_rate": 1.0315774961182412e-07,
+      "loss": 0.4748,
+      "step": 1464
+    },
+    {
+      "epoch": 0.956266318537859,
+      "grad_norm": 0.05467505007982254,
+      "learning_rate": 1.0015104475721848e-07,
+      "loss": 0.5476,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9569190600522193,
+      "grad_norm": 0.060881637036800385,
+      "learning_rate": 9.71885875855294e-08,
+      "loss": 0.564,
+      "step": 1466
+    },
+    {
+      "epoch": 0.9575718015665796,
+      "grad_norm": 0.05629203841090202,
+      "learning_rate": 9.4270391337552e-08,
+      "loss": 0.5343,
+      "step": 1467
+    },
+    {
+      "epoch": 0.95822454308094,
+      "grad_norm": 0.055595796555280685,
+      "learning_rate": 9.13964690562552e-08,
+      "loss": 0.535,
+      "step": 1468
+    },
+    {
+      "epoch": 0.9588772845953003,
+      "grad_norm": 0.053988490253686905,
+      "learning_rate": 8.856683358672402e-08,
+      "loss": 0.5259,
+      "step": 1469
+    },
+    {
+      "epoch": 0.9595300261096605,
+      "grad_norm": 0.05572625994682312,
+      "learning_rate": 8.578149757610176e-08,
+      "loss": 0.5388,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9601827676240209,
+      "grad_norm": 0.05242741480469704,
+      "learning_rate": 8.30404734735346e-08,
+      "loss": 0.4877,
+      "step": 1471
+    },
+    {
+      "epoch": 0.9608355091383812,
+      "grad_norm": 0.05756179615855217,
+      "learning_rate": 8.034377353011603e-08,
+      "loss": 0.5024,
+      "step": 1472
+    },
+    {
+      "epoch": 0.9614882506527415,
+      "grad_norm": 0.0551273413002491,
+      "learning_rate": 7.769140979882905e-08,
+      "loss": 0.558,
+      "step": 1473
+    },
+    {
+      "epoch": 0.9621409921671018,
+      "grad_norm": 0.056963566690683365,
+      "learning_rate": 7.508339413449528e-08,
+      "loss": 0.4535,
+      "step": 1474
+    },
+    {
+      "epoch": 0.9627937336814621,
+      "grad_norm": 0.05817368999123573,
+      "learning_rate": 7.251973819372371e-08,
+      "loss": 0.5771,
+      "step": 1475
+    },
+    {
+      "epoch": 0.9634464751958225,
+      "grad_norm": 0.057898059487342834,
+      "learning_rate": 7.000045343485306e-08,
+      "loss": 0.5658,
+      "step": 1476
+    },
+    {
+      "epoch": 0.9640992167101827,
+      "grad_norm": 0.05078176409006119,
+      "learning_rate": 6.752555111790515e-08,
+      "loss": 0.5092,
+      "step": 1477
+    },
+    {
+      "epoch": 0.9647519582245431,
+      "grad_norm": 0.05763321742415428,
+      "learning_rate": 6.509504230453379e-08,
+      "loss": 0.5376,
+      "step": 1478
+    },
+    {
+      "epoch": 0.9654046997389034,
+      "grad_norm": 0.055951688438653946,
+      "learning_rate": 6.270893785797261e-08,
+      "loss": 0.5519,
+      "step": 1479
+    },
+    {
+      "epoch": 0.9660574412532638,
+      "grad_norm": 0.055584829300642014,
+      "learning_rate": 6.036724844299069e-08,
+      "loss": 0.5564,
+      "step": 1480
+    },
+    {
+      "epoch": 0.966710182767624,
+      "grad_norm": 0.05595998466014862,
+      "learning_rate": 5.806998452584034e-08,
+      "loss": 0.3988,
+      "step": 1481
+    },
+    {
+      "epoch": 0.9673629242819843,
+      "grad_norm": 0.05466803163290024,
+      "learning_rate": 5.581715637421492e-08,
+      "loss": 0.5162,
+      "step": 1482
+    },
+    {
+      "epoch": 0.9680156657963447,
+      "grad_norm": 0.05414329841732979,
+      "learning_rate": 5.360877405720111e-08,
+      "loss": 0.4885,
+      "step": 1483
+    },
+    {
+      "epoch": 0.9686684073107049,
+      "grad_norm": 0.06734245270490646,
+      "learning_rate": 5.144484744523004e-08,
+      "loss": 0.6203,
+      "step": 1484
+    },
+    {
+      "epoch": 0.9693211488250653,
+      "grad_norm": 0.05423906818032265,
+      "learning_rate": 4.932538621004068e-08,
+      "loss": 0.4688,
+      "step": 1485
+    },
+    {
+      "epoch": 0.9699738903394256,
+      "grad_norm": 0.05275886133313179,
+      "learning_rate": 4.7250399824629867e-08,
+      "loss": 0.4565,
+      "step": 1486
+    },
+    {
+      "epoch": 0.970626631853786,
+      "grad_norm": 0.059688255190849304,
+      "learning_rate": 4.521989756321565e-08,
+      "loss": 0.5684,
+      "step": 1487
+    },
+    {
+      "epoch": 0.9712793733681462,
+      "grad_norm": 0.05637152120471001,
+      "learning_rate": 4.323388850118848e-08,
+      "loss": 0.5786,
+      "step": 1488
+    },
+    {
+      "epoch": 0.9719321148825065,
+      "grad_norm": 0.06485294550657272,
+      "learning_rate": 4.129238151508008e-08,
+      "loss": 0.592,
+      "step": 1489
+    },
+    {
+      "epoch": 0.9725848563968669,
+      "grad_norm": 0.04909993335604668,
+      "learning_rate": 3.939538528251463e-08,
+      "loss": 0.4946,
+      "step": 1490
+    },
+    {
+      "epoch": 0.9732375979112271,
+      "grad_norm": 0.058074407279491425,
+      "learning_rate": 3.754290828217655e-08,
+      "loss": 0.5951,
+      "step": 1491
+    },
+    {
+      "epoch": 0.9738903394255874,
+      "grad_norm": 0.056894298642873764,
+      "learning_rate": 3.5734958793769426e-08,
+      "loss": 0.4916,
+      "step": 1492
+    },
+    {
+      "epoch": 0.9745430809399478,
+      "grad_norm": 0.0597614087164402,
+      "learning_rate": 3.397154489798049e-08,
+      "loss": 0.5783,
+      "step": 1493
+    },
+    {
+      "epoch": 0.9751958224543081,
+      "grad_norm": 0.05714694410562515,
+      "learning_rate": 3.225267447644065e-08,
+      "loss": 0.5015,
+      "step": 1494
+    },
+    {
+      "epoch": 0.9758485639686684,
+      "grad_norm": 0.05466402322053909,
+      "learning_rate": 3.057835521169783e-08,
+      "loss": 0.5161,
+      "step": 1495
+    },
+    {
+      "epoch": 0.9765013054830287,
+      "grad_norm": 0.06516537070274353,
+      "learning_rate": 2.8948594587170366e-08,
+      "loss": 0.5422,
+      "step": 1496
+    },
+    {
+      "epoch": 0.9771540469973891,
+      "grad_norm": 0.055320855230093,
+      "learning_rate": 2.7363399887128105e-08,
+      "loss": 0.4968,
+      "step": 1497
+    },
+    {
+      "epoch": 0.9778067885117493,
+      "grad_norm": 0.062362801283597946,
+      "learning_rate": 2.5822778196645804e-08,
+      "loss": 0.6666,
+      "step": 1498
+    },
+    {
+      "epoch": 0.9784595300261096,
+      "grad_norm": 0.0544227734208107,
+      "learning_rate": 2.4326736401579788e-08,
+      "loss": 0.5162,
+      "step": 1499
+    },
+    {
+      "epoch": 0.97911227154047,
+      "grad_norm": 0.061307311058044434,
+      "learning_rate": 2.287528118853688e-08,
+      "loss": 0.5837,
+      "step": 1500
+    },
+    {
+      "epoch": 0.9797650130548303,
+      "grad_norm": 0.06065037474036217,
+      "learning_rate": 2.1468419044839984e-08,
+      "loss": 0.5936,
+      "step": 1501
+    },
+    {
+      "epoch": 0.9804177545691906,
+      "grad_norm": 0.05570589005947113,
+      "learning_rate": 2.010615625850365e-08,
+      "loss": 0.5298,
+      "step": 1502
+    },
+    {
+      "epoch": 0.9810704960835509,
+      "grad_norm": 0.06140119209885597,
+      "learning_rate": 1.8788498918204112e-08,
+      "loss": 0.5371,
+      "step": 1503
+    },
+    {
+      "epoch": 0.9817232375979112,
+      "grad_norm": 0.05893724039196968,
+      "learning_rate": 1.7515452913250407e-08,
+      "loss": 0.5344,
+      "step": 1504
+    },
+    {
+      "epoch": 0.9823759791122716,
+      "grad_norm": 0.05744578316807747,
+      "learning_rate": 1.6287023933564407e-08,
+      "loss": 0.4739,
+      "step": 1505
+    },
+    {
+      "epoch": 0.9830287206266318,
+      "grad_norm": 0.05734417587518692,
+      "learning_rate": 1.510321746964416e-08,
+      "loss": 0.5262,
+      "step": 1506
+    },
+    {
+      "epoch": 0.9836814621409922,
+      "grad_norm": 0.06392619013786316,
+      "learning_rate": 1.3964038812551706e-08,
+      "loss": 0.62,
+      "step": 1507
+    },
+    {
+      "epoch": 0.9843342036553525,
+      "grad_norm": 0.04519110545516014,
+      "learning_rate": 1.2869493053880855e-08,
+      "loss": 0.432,
+      "step": 1508
+    },
+    {
+      "epoch": 0.9849869451697127,
+      "grad_norm": 0.052288834005594254,
+      "learning_rate": 1.1819585085737217e-08,
+      "loss": 0.4645,
+      "step": 1509
+    },
+    {
+      "epoch": 0.9856396866840731,
+      "grad_norm": 0.06511126458644867,
+      "learning_rate": 1.0814319600718216e-08,
+      "loss": 0.5786,
+      "step": 1510
+    },
+    {
+      "epoch": 0.9862924281984334,
+      "grad_norm": 0.055628255009651184,
+      "learning_rate": 9.853701091888656e-09,
+      "loss": 0.4908,
+      "step": 1511
+    },
+    {
+      "epoch": 0.9869451697127938,
+      "grad_norm": 0.05760030820965767,
+      "learning_rate": 8.937733852764086e-09,
+      "loss": 0.6044,
+      "step": 1512
+    },
+    {
+      "epoch": 0.987597911227154,
+      "grad_norm": 0.048606064170598984,
+      "learning_rate": 8.066421977286355e-09,
+      "loss": 0.4827,
+      "step": 1513
+    },
+    {
+      "epoch": 0.9882506527415144,
+      "grad_norm": 0.05585504323244095,
+      "learning_rate": 7.239769359811411e-09,
+      "loss": 0.5,
+      "step": 1514
+    },
+    {
+      "epoch": 0.9889033942558747,
+      "grad_norm": 0.05974135175347328,
+      "learning_rate": 6.457779695090427e-09,
+      "loss": 0.5583,
+      "step": 1515
+    },
+    {
+      "epoch": 0.9895561357702349,
+      "grad_norm": 0.05551174655556679,
+      "learning_rate": 5.7204564782498136e-09,
+      "loss": 0.5376,
+      "step": 1516
+    },
+    {
+      "epoch": 0.9902088772845953,
+      "grad_norm": 0.05592876672744751,
+      "learning_rate": 5.027803004779008e-09,
+      "loss": 0.5015,
+      "step": 1517
+    },
+    {
+      "epoch": 0.9908616187989556,
+      "grad_norm": 0.058642178773880005,
+      "learning_rate": 4.37982237051271e-09,
+      "loss": 0.6064,
+      "step": 1518
+    },
+    {
+      "epoch": 0.991514360313316,
+      "grad_norm": 0.0588836707174778,
+      "learning_rate": 3.776517471621999e-09,
+      "loss": 0.4826,
+      "step": 1519
+    },
+    {
+      "epoch": 0.9921671018276762,
+      "grad_norm": 0.05897413566708565,
+      "learning_rate": 3.217891004596574e-09,
+      "loss": 0.5246,
+      "step": 1520
+    },
+    {
+      "epoch": 0.9928198433420365,
+      "grad_norm": 0.058347541838884354,
+      "learning_rate": 2.7039454662336484e-09,
+      "loss": 0.5419,
+      "step": 1521
+    },
+    {
+      "epoch": 0.9934725848563969,
+      "grad_norm": 0.06031772866845131,
+      "learning_rate": 2.2346831536312895e-09,
+      "loss": 0.5225,
+      "step": 1522
+    },
+    {
+      "epoch": 0.9941253263707572,
+      "grad_norm": 0.05832480639219284,
+      "learning_rate": 1.810106164169545e-09,
+      "loss": 0.5059,
+      "step": 1523
+    },
+    {
+      "epoch": 0.9947780678851175,
+      "grad_norm": 0.05937207490205765,
+      "learning_rate": 1.4302163955093317e-09,
+      "loss": 0.5684,
+      "step": 1524
+    },
+    {
+      "epoch": 0.9954308093994778,
+      "grad_norm": 0.05283202975988388,
+      "learning_rate": 1.0950155455802247e-09,
+      "loss": 0.514,
+      "step": 1525
+    },
+    {
+      "epoch": 0.9960835509138382,
+      "grad_norm": 0.061174239963293076,
+      "learning_rate": 8.045051125726844e-10,
+      "loss": 0.5251,
+      "step": 1526
+    },
+    {
+      "epoch": 0.9967362924281984,
+      "grad_norm": 0.06188638508319855,
+      "learning_rate": 5.586863949325061e-10,
+      "loss": 0.5051,
+      "step": 1527
+    },
+    {
+      "epoch": 0.9973890339425587,
+      "grad_norm": 0.054148390889167786,
+      "learning_rate": 3.575604913530484e-10,
+      "loss": 0.5091,
+      "step": 1528
+    },
+    {
+      "epoch": 0.9980417754569191,
+      "grad_norm": 0.0549580417573452,
+      "learning_rate": 2.011283007730125e-10,
+      "loss": 0.5188,
+      "step": 1529
+    },
+    {
+      "epoch": 0.9986945169712794,
+      "grad_norm": 0.05601628124713898,
+      "learning_rate": 8.93905223708913e-11,
+      "loss": 0.5068,
+      "step": 1530
+    },
+    {
+      "epoch": 0.9993472584856397,
+      "grad_norm": 0.056755177676677704,
+      "learning_rate": 2.2347655563859096e-11,
+      "loss": 0.5351,
+      "step": 1531
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.06067736819386482,
+      "learning_rate": 0.0,
+      "loss": 0.5493,
+      "step": 1532
+    },
+    {
+      "epoch": 1.0,
+      "step": 1532,
+      "total_flos": 0.0,
+      "train_loss": 0.10018481526222303,
+      "train_runtime": 5590.0623,
+      "train_samples_per_second": 280.582,
+      "train_steps_per_second": 0.274
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1532,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "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": 0.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}