diff --git "a/metallama3_8b/limo/trainer_state.json" "b/metallama3_8b/limo/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/metallama3_8b/limo/trainer_state.json"
@@ -0,0 +1,14393 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 2050,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004878048780487805,
+      "grad_norm": 27.79998016357422,
+      "learning_rate": 5e-06,
+      "loss": 1.4179,
+      "step": 1
+    },
+    {
+      "epoch": 0.00975609756097561,
+      "grad_norm": 4.086409091949463,
+      "learning_rate": 4.999997064365715e-06,
+      "loss": 1.1405,
+      "step": 2
+    },
+    {
+      "epoch": 0.014634146341463415,
+      "grad_norm": 4.499151229858398,
+      "learning_rate": 4.999988257469751e-06,
+      "loss": 0.8682,
+      "step": 3
+    },
+    {
+      "epoch": 0.01951219512195122,
+      "grad_norm": 4.555822849273682,
+      "learning_rate": 4.999973579332793e-06,
+      "loss": 0.9961,
+      "step": 4
+    },
+    {
+      "epoch": 0.024390243902439025,
+      "grad_norm": 5.6235246658325195,
+      "learning_rate": 4.999953029989312e-06,
+      "loss": 1.0173,
+      "step": 5
+    },
+    {
+      "epoch": 0.02926829268292683,
+      "grad_norm": 3.9943182468414307,
+      "learning_rate": 4.999926609487568e-06,
+      "loss": 1.1083,
+      "step": 6
+    },
+    {
+      "epoch": 0.03414634146341464,
+      "grad_norm": 5.685941219329834,
+      "learning_rate": 4.9998943178896106e-06,
+      "loss": 1.1109,
+      "step": 7
+    },
+    {
+      "epoch": 0.03902439024390244,
+      "grad_norm": 15.914257049560547,
+      "learning_rate": 4.999856155271276e-06,
+      "loss": 1.821,
+      "step": 8
+    },
+    {
+      "epoch": 0.04390243902439024,
+      "grad_norm": 4.147185325622559,
+      "learning_rate": 4.999812121722191e-06,
+      "loss": 1.0417,
+      "step": 9
+    },
+    {
+      "epoch": 0.04878048780487805,
+      "grad_norm": 11.123332977294922,
+      "learning_rate": 4.999762217345766e-06,
+      "loss": 1.5672,
+      "step": 10
+    },
+    {
+      "epoch": 0.05365853658536585,
+      "grad_norm": 2.842331886291504,
+      "learning_rate": 4.999706442259205e-06,
+      "loss": 0.7297,
+      "step": 11
+    },
+    {
+      "epoch": 0.05853658536585366,
+      "grad_norm": 37.685062408447266,
+      "learning_rate": 4.999644796593492e-06,
+      "loss": 0.9112,
+      "step": 12
+    },
+    {
+      "epoch": 0.06341463414634146,
+      "grad_norm": 11.214252471923828,
+      "learning_rate": 4.999577280493407e-06,
+      "loss": 0.7854,
+      "step": 13
+    },
+    {
+      "epoch": 0.06829268292682927,
+      "grad_norm": 5.10387659072876,
+      "learning_rate": 4.99950389411751e-06,
+      "loss": 1.1317,
+      "step": 14
+    },
+    {
+      "epoch": 0.07317073170731707,
+      "grad_norm": 3.685403347015381,
+      "learning_rate": 4.999424637638148e-06,
+      "loss": 0.7864,
+      "step": 15
+    },
+    {
+      "epoch": 0.07804878048780488,
+      "grad_norm": 2.9567184448242188,
+      "learning_rate": 4.999339511241458e-06,
+      "loss": 0.8494,
+      "step": 16
+    },
+    {
+      "epoch": 0.08292682926829269,
+      "grad_norm": 11.396956443786621,
+      "learning_rate": 4.9992485151273584e-06,
+      "loss": 1.2189,
+      "step": 17
+    },
+    {
+      "epoch": 0.08780487804878048,
+      "grad_norm": 7.007385730743408,
+      "learning_rate": 4.999151649509554e-06,
+      "loss": 1.0532,
+      "step": 18
+    },
+    {
+      "epoch": 0.09268292682926829,
+      "grad_norm": 3.4347329139709473,
+      "learning_rate": 4.9990489146155356e-06,
+      "loss": 1.088,
+      "step": 19
+    },
+    {
+      "epoch": 0.0975609756097561,
+      "grad_norm": 3.1865031719207764,
+      "learning_rate": 4.9989403106865765e-06,
+      "loss": 1.0414,
+      "step": 20
+    },
+    {
+      "epoch": 0.1024390243902439,
+      "grad_norm": 3.4605791568756104,
+      "learning_rate": 4.9988258379777334e-06,
+      "loss": 0.8878,
+      "step": 21
+    },
+    {
+      "epoch": 0.1073170731707317,
+      "grad_norm": 2.860478639602661,
+      "learning_rate": 4.998705496757846e-06,
+      "loss": 0.9151,
+      "step": 22
+    },
+    {
+      "epoch": 0.11219512195121951,
+      "grad_norm": 9.101946830749512,
+      "learning_rate": 4.998579287309538e-06,
+      "loss": 1.4304,
+      "step": 23
+    },
+    {
+      "epoch": 0.11707317073170732,
+      "grad_norm": 24.21122169494629,
+      "learning_rate": 4.998447209929211e-06,
+      "loss": 1.0858,
+      "step": 24
+    },
+    {
+      "epoch": 0.12195121951219512,
+      "grad_norm": 3.286980152130127,
+      "learning_rate": 4.998309264927053e-06,
+      "loss": 0.6571,
+      "step": 25
+    },
+    {
+      "epoch": 0.12682926829268293,
+      "grad_norm": 4.0232062339782715,
+      "learning_rate": 4.998165452627025e-06,
+      "loss": 0.8493,
+      "step": 26
+    },
+    {
+      "epoch": 0.13170731707317074,
+      "grad_norm": 3.7688663005828857,
+      "learning_rate": 4.998015773366874e-06,
+      "loss": 0.9224,
+      "step": 27
+    },
+    {
+      "epoch": 0.13658536585365855,
+      "grad_norm": 2.9382026195526123,
+      "learning_rate": 4.997860227498122e-06,
+      "loss": 0.7588,
+      "step": 28
+    },
+    {
+      "epoch": 0.14146341463414633,
+      "grad_norm": 4.327457904815674,
+      "learning_rate": 4.99769881538607e-06,
+      "loss": 1.1817,
+      "step": 29
+    },
+    {
+      "epoch": 0.14634146341463414,
+      "grad_norm": 3.47487735748291,
+      "learning_rate": 4.997531537409794e-06,
+      "loss": 1.0737,
+      "step": 30
+    },
+    {
+      "epoch": 0.15121951219512195,
+      "grad_norm": 3.0616214275360107,
+      "learning_rate": 4.99735839396215e-06,
+      "loss": 0.7899,
+      "step": 31
+    },
+    {
+      "epoch": 0.15609756097560976,
+      "grad_norm": 3.065070152282715,
+      "learning_rate": 4.9971793854497655e-06,
+      "loss": 0.7745,
+      "step": 32
+    },
+    {
+      "epoch": 0.16097560975609757,
+      "grad_norm": 3.5202279090881348,
+      "learning_rate": 4.996994512293042e-06,
+      "loss": 0.984,
+      "step": 33
+    },
+    {
+      "epoch": 0.16585365853658537,
+      "grad_norm": 3.421769142150879,
+      "learning_rate": 4.996803774926157e-06,
+      "loss": 0.8235,
+      "step": 34
+    },
+    {
+      "epoch": 0.17073170731707318,
+      "grad_norm": 4.6582207679748535,
+      "learning_rate": 4.996607173797059e-06,
+      "loss": 1.3227,
+      "step": 35
+    },
+    {
+      "epoch": 0.17560975609756097,
+      "grad_norm": 2.9829282760620117,
+      "learning_rate": 4.996404709367466e-06,
+      "loss": 0.8854,
+      "step": 36
+    },
+    {
+      "epoch": 0.18048780487804877,
+      "grad_norm": 2.5982632637023926,
+      "learning_rate": 4.996196382112868e-06,
+      "loss": 0.6786,
+      "step": 37
+    },
+    {
+      "epoch": 0.18536585365853658,
+      "grad_norm": 2.9807393550872803,
+      "learning_rate": 4.9959821925225235e-06,
+      "loss": 0.9344,
+      "step": 38
+    },
+    {
+      "epoch": 0.1902439024390244,
+      "grad_norm": 2.7364351749420166,
+      "learning_rate": 4.995762141099456e-06,
+      "loss": 0.814,
+      "step": 39
+    },
+    {
+      "epoch": 0.1951219512195122,
+      "grad_norm": 3.4324638843536377,
+      "learning_rate": 4.995536228360461e-06,
+      "loss": 1.0276,
+      "step": 40
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.911834716796875,
+      "learning_rate": 4.995304454836095e-06,
+      "loss": 0.9291,
+      "step": 41
+    },
+    {
+      "epoch": 0.2048780487804878,
+      "grad_norm": 3.0294723510742188,
+      "learning_rate": 4.9950668210706795e-06,
+      "loss": 0.8145,
+      "step": 42
+    },
+    {
+      "epoch": 0.2097560975609756,
+      "grad_norm": 4.681829452514648,
+      "learning_rate": 4.994823327622299e-06,
+      "loss": 0.8779,
+      "step": 43
+    },
+    {
+      "epoch": 0.2146341463414634,
+      "grad_norm": 3.643914222717285,
+      "learning_rate": 4.9945739750628e-06,
+      "loss": 0.8196,
+      "step": 44
+    },
+    {
+      "epoch": 0.21951219512195122,
+      "grad_norm": 2.7542076110839844,
+      "learning_rate": 4.994318763977789e-06,
+      "loss": 0.8443,
+      "step": 45
+    },
+    {
+      "epoch": 0.22439024390243903,
+      "grad_norm": 6.873605728149414,
+      "learning_rate": 4.994057694966632e-06,
+      "loss": 1.0328,
+      "step": 46
+    },
+    {
+      "epoch": 0.22926829268292684,
+      "grad_norm": 3.11810040473938,
+      "learning_rate": 4.993790768642449e-06,
+      "loss": 1.0673,
+      "step": 47
+    },
+    {
+      "epoch": 0.23414634146341465,
+      "grad_norm": 4.360548496246338,
+      "learning_rate": 4.99351798563212e-06,
+      "loss": 1.3198,
+      "step": 48
+    },
+    {
+      "epoch": 0.23902439024390243,
+      "grad_norm": 2.6894314289093018,
+      "learning_rate": 4.993239346576278e-06,
+      "loss": 0.8743,
+      "step": 49
+    },
+    {
+      "epoch": 0.24390243902439024,
+      "grad_norm": 3.2640421390533447,
+      "learning_rate": 4.99295485212931e-06,
+      "loss": 1.109,
+      "step": 50
+    },
+    {
+      "epoch": 0.24878048780487805,
+      "grad_norm": 3.1565866470336914,
+      "learning_rate": 4.992664502959351e-06,
+      "loss": 0.9291,
+      "step": 51
+    },
+    {
+      "epoch": 0.25365853658536586,
+      "grad_norm": 3.4829447269439697,
+      "learning_rate": 4.99236829974829e-06,
+      "loss": 0.8159,
+      "step": 52
+    },
+    {
+      "epoch": 0.25853658536585367,
+      "grad_norm": 2.7535626888275146,
+      "learning_rate": 4.992066243191762e-06,
+      "loss": 1.0359,
+      "step": 53
+    },
+    {
+      "epoch": 0.2634146341463415,
+      "grad_norm": 2.482935905456543,
+      "learning_rate": 4.991758333999148e-06,
+      "loss": 0.8091,
+      "step": 54
+    },
+    {
+      "epoch": 0.2682926829268293,
+      "grad_norm": 2.917445659637451,
+      "learning_rate": 4.991444572893575e-06,
+      "loss": 0.6925,
+      "step": 55
+    },
+    {
+      "epoch": 0.2731707317073171,
+      "grad_norm": 2.9802236557006836,
+      "learning_rate": 4.991124960611916e-06,
+      "loss": 0.6329,
+      "step": 56
+    },
+    {
+      "epoch": 0.2780487804878049,
+      "grad_norm": 2.9677224159240723,
+      "learning_rate": 4.99079949790478e-06,
+      "loss": 0.8069,
+      "step": 57
+    },
+    {
+      "epoch": 0.28292682926829266,
+      "grad_norm": 2.8304293155670166,
+      "learning_rate": 4.99046818553652e-06,
+      "loss": 0.8682,
+      "step": 58
+    },
+    {
+      "epoch": 0.28780487804878047,
+      "grad_norm": 5.253443717956543,
+      "learning_rate": 4.9901310242852246e-06,
+      "loss": 1.1069,
+      "step": 59
+    },
+    {
+      "epoch": 0.2926829268292683,
+      "grad_norm": 3.686016082763672,
+      "learning_rate": 4.9897880149427206e-06,
+      "loss": 0.9465,
+      "step": 60
+    },
+    {
+      "epoch": 0.2975609756097561,
+      "grad_norm": 3.6372263431549072,
+      "learning_rate": 4.989439158314566e-06,
+      "loss": 0.9738,
+      "step": 61
+    },
+    {
+      "epoch": 0.3024390243902439,
+      "grad_norm": 3.0756819248199463,
+      "learning_rate": 4.989084455220056e-06,
+      "loss": 0.6417,
+      "step": 62
+    },
+    {
+      "epoch": 0.3073170731707317,
+      "grad_norm": 3.379222869873047,
+      "learning_rate": 4.988723906492212e-06,
+      "loss": 1.0092,
+      "step": 63
+    },
+    {
+      "epoch": 0.3121951219512195,
+      "grad_norm": 3.4571032524108887,
+      "learning_rate": 4.988357512977785e-06,
+      "loss": 0.6691,
+      "step": 64
+    },
+    {
+      "epoch": 0.3170731707317073,
+      "grad_norm": 3.1982104778289795,
+      "learning_rate": 4.987985275537252e-06,
+      "loss": 0.6651,
+      "step": 65
+    },
+    {
+      "epoch": 0.32195121951219513,
+      "grad_norm": 2.9723124504089355,
+      "learning_rate": 4.9876071950448185e-06,
+      "loss": 0.9227,
+      "step": 66
+    },
+    {
+      "epoch": 0.32682926829268294,
+      "grad_norm": 2.5521399974823,
+      "learning_rate": 4.987223272388407e-06,
+      "loss": 0.6664,
+      "step": 67
+    },
+    {
+      "epoch": 0.33170731707317075,
+      "grad_norm": 2.8934121131896973,
+      "learning_rate": 4.986833508469663e-06,
+      "loss": 0.997,
+      "step": 68
+    },
+    {
+      "epoch": 0.33658536585365856,
+      "grad_norm": 4.7546586990356445,
+      "learning_rate": 4.98643790420395e-06,
+      "loss": 0.8551,
+      "step": 69
+    },
+    {
+      "epoch": 0.34146341463414637,
+      "grad_norm": 3.091616153717041,
+      "learning_rate": 4.986036460520348e-06,
+      "loss": 0.8874,
+      "step": 70
+    },
+    {
+      "epoch": 0.3463414634146341,
+      "grad_norm": 4.1724677085876465,
+      "learning_rate": 4.98562917836165e-06,
+      "loss": 1.1393,
+      "step": 71
+    },
+    {
+      "epoch": 0.35121951219512193,
+      "grad_norm": 2.6568572521209717,
+      "learning_rate": 4.985216058684362e-06,
+      "loss": 0.6379,
+      "step": 72
+    },
+    {
+      "epoch": 0.35609756097560974,
+      "grad_norm": 2.396416187286377,
+      "learning_rate": 4.984797102458697e-06,
+      "loss": 1.0292,
+      "step": 73
+    },
+    {
+      "epoch": 0.36097560975609755,
+      "grad_norm": 3.0667319297790527,
+      "learning_rate": 4.984372310668579e-06,
+      "loss": 0.7048,
+      "step": 74
+    },
+    {
+      "epoch": 0.36585365853658536,
+      "grad_norm": 2.4820518493652344,
+      "learning_rate": 4.983941684311633e-06,
+      "loss": 1.2353,
+      "step": 75
+    },
+    {
+      "epoch": 0.37073170731707317,
+      "grad_norm": 4.062836647033691,
+      "learning_rate": 4.983505224399188e-06,
+      "loss": 0.8933,
+      "step": 76
+    },
+    {
+      "epoch": 0.375609756097561,
+      "grad_norm": 2.4480767250061035,
+      "learning_rate": 4.983062931956275e-06,
+      "loss": 0.8221,
+      "step": 77
+    },
+    {
+      "epoch": 0.3804878048780488,
+      "grad_norm": 3.134138822555542,
+      "learning_rate": 4.9826148080216195e-06,
+      "loss": 0.8899,
+      "step": 78
+    },
+    {
+      "epoch": 0.3853658536585366,
+      "grad_norm": 2.8165836334228516,
+      "learning_rate": 4.9821608536476445e-06,
+      "loss": 1.2451,
+      "step": 79
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 3.734433650970459,
+      "learning_rate": 4.981701069900465e-06,
+      "loss": 0.8536,
+      "step": 80
+    },
+    {
+      "epoch": 0.3951219512195122,
+      "grad_norm": 2.853421449661255,
+      "learning_rate": 4.9812354578598876e-06,
+      "loss": 0.7857,
+      "step": 81
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.541687250137329,
+      "learning_rate": 4.980764018619405e-06,
+      "loss": 0.8332,
+      "step": 82
+    },
+    {
+      "epoch": 0.40487804878048783,
+      "grad_norm": 4.405911445617676,
+      "learning_rate": 4.980286753286196e-06,
+      "loss": 0.9927,
+      "step": 83
+    },
+    {
+      "epoch": 0.4097560975609756,
+      "grad_norm": 3.3034985065460205,
+      "learning_rate": 4.97980366298112e-06,
+      "loss": 0.8161,
+      "step": 84
+    },
+    {
+      "epoch": 0.4146341463414634,
+      "grad_norm": 2.6678085327148438,
+      "learning_rate": 4.97931474883872e-06,
+      "loss": 0.8017,
+      "step": 85
+    },
+    {
+      "epoch": 0.4195121951219512,
+      "grad_norm": 2.58524227142334,
+      "learning_rate": 4.978820012007213e-06,
+      "loss": 0.8811,
+      "step": 86
+    },
+    {
+      "epoch": 0.424390243902439,
+      "grad_norm": 2.482597827911377,
+      "learning_rate": 4.978319453648495e-06,
+      "loss": 0.9461,
+      "step": 87
+    },
+    {
+      "epoch": 0.4292682926829268,
+      "grad_norm": 2.5731301307678223,
+      "learning_rate": 4.977813074938128e-06,
+      "loss": 0.8835,
+      "step": 88
+    },
+    {
+      "epoch": 0.43414634146341463,
+      "grad_norm": 2.7914488315582275,
+      "learning_rate": 4.977300877065347e-06,
+      "loss": 0.8466,
+      "step": 89
+    },
+    {
+      "epoch": 0.43902439024390244,
+      "grad_norm": 2.416043758392334,
+      "learning_rate": 4.976782861233053e-06,
+      "loss": 0.7132,
+      "step": 90
+    },
+    {
+      "epoch": 0.44390243902439025,
+      "grad_norm": 3.7616264820098877,
+      "learning_rate": 4.976259028657812e-06,
+      "loss": 0.7639,
+      "step": 91
+    },
+    {
+      "epoch": 0.44878048780487806,
+      "grad_norm": 2.6081621646881104,
+      "learning_rate": 4.975729380569845e-06,
+      "loss": 0.8055,
+      "step": 92
+    },
+    {
+      "epoch": 0.45365853658536587,
+      "grad_norm": 3.3343570232391357,
+      "learning_rate": 4.975193918213035e-06,
+      "loss": 0.6042,
+      "step": 93
+    },
+    {
+      "epoch": 0.4585365853658537,
+      "grad_norm": 2.517544746398926,
+      "learning_rate": 4.974652642844921e-06,
+      "loss": 0.7672,
+      "step": 94
+    },
+    {
+      "epoch": 0.4634146341463415,
+      "grad_norm": 4.173468589782715,
+      "learning_rate": 4.974105555736693e-06,
+      "loss": 1.0682,
+      "step": 95
+    },
+    {
+      "epoch": 0.4682926829268293,
+      "grad_norm": 2.8422317504882812,
+      "learning_rate": 4.973552658173186e-06,
+      "loss": 0.7841,
+      "step": 96
+    },
+    {
+      "epoch": 0.47317073170731705,
+      "grad_norm": 5.042182445526123,
+      "learning_rate": 4.972993951452887e-06,
+      "loss": 0.8851,
+      "step": 97
+    },
+    {
+      "epoch": 0.47804878048780486,
+      "grad_norm": 5.977590560913086,
+      "learning_rate": 4.9724294368879214e-06,
+      "loss": 0.9059,
+      "step": 98
+    },
+    {
+      "epoch": 0.48292682926829267,
+      "grad_norm": 4.227641582489014,
+      "learning_rate": 4.971859115804055e-06,
+      "loss": 1.0152,
+      "step": 99
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 3.180952548980713,
+      "learning_rate": 4.9712829895406935e-06,
+      "loss": 0.8092,
+      "step": 100
+    },
+    {
+      "epoch": 0.4926829268292683,
+      "grad_norm": 11.220394134521484,
+      "learning_rate": 4.970701059450872e-06,
+      "loss": 0.8239,
+      "step": 101
+    },
+    {
+      "epoch": 0.4975609756097561,
+      "grad_norm": 2.346975088119507,
+      "learning_rate": 4.970113326901258e-06,
+      "loss": 0.9283,
+      "step": 102
+    },
+    {
+      "epoch": 0.5024390243902439,
+      "grad_norm": 2.9470982551574707,
+      "learning_rate": 4.9695197932721455e-06,
+      "loss": 0.9429,
+      "step": 103
+    },
+    {
+      "epoch": 0.5073170731707317,
+      "grad_norm": 3.6048219203948975,
+      "learning_rate": 4.968920459957453e-06,
+      "loss": 0.9231,
+      "step": 104
+    },
+    {
+      "epoch": 0.5121951219512195,
+      "grad_norm": 2.8181886672973633,
+      "learning_rate": 4.968315328364719e-06,
+      "loss": 1.0005,
+      "step": 105
+    },
+    {
+      "epoch": 0.5170731707317073,
+      "grad_norm": 3.114147424697876,
+      "learning_rate": 4.9677043999151e-06,
+      "loss": 1.1326,
+      "step": 106
+    },
+    {
+      "epoch": 0.5219512195121951,
+      "grad_norm": 2.965885639190674,
+      "learning_rate": 4.967087676043366e-06,
+      "loss": 0.541,
+      "step": 107
+    },
+    {
+      "epoch": 0.526829268292683,
+      "grad_norm": 3.098677635192871,
+      "learning_rate": 4.966465158197897e-06,
+      "loss": 0.9473,
+      "step": 108
+    },
+    {
+      "epoch": 0.5317073170731708,
+      "grad_norm": 2.8640191555023193,
+      "learning_rate": 4.965836847840681e-06,
+      "loss": 0.6678,
+      "step": 109
+    },
+    {
+      "epoch": 0.5365853658536586,
+      "grad_norm": 3.0950934886932373,
+      "learning_rate": 4.96520274644731e-06,
+      "loss": 0.9251,
+      "step": 110
+    },
+    {
+      "epoch": 0.5414634146341464,
+      "grad_norm": 2.99444317817688,
+      "learning_rate": 4.964562855506976e-06,
+      "loss": 0.7807,
+      "step": 111
+    },
+    {
+      "epoch": 0.5463414634146342,
+      "grad_norm": 2.348639726638794,
+      "learning_rate": 4.963917176522466e-06,
+      "loss": 0.6395,
+      "step": 112
+    },
+    {
+      "epoch": 0.551219512195122,
+      "grad_norm": 3.5988354682922363,
+      "learning_rate": 4.963265711010164e-06,
+      "loss": 1.0658,
+      "step": 113
+    },
+    {
+      "epoch": 0.5560975609756098,
+      "grad_norm": 3.3423564434051514,
+      "learning_rate": 4.9626084605000395e-06,
+      "loss": 0.8974,
+      "step": 114
+    },
+    {
+      "epoch": 0.5609756097560976,
+      "grad_norm": 2.8353331089019775,
+      "learning_rate": 4.961945426535652e-06,
+      "loss": 0.6144,
+      "step": 115
+    },
+    {
+      "epoch": 0.5658536585365853,
+      "grad_norm": 2.752387046813965,
+      "learning_rate": 4.961276610674141e-06,
+      "loss": 0.9083,
+      "step": 116
+    },
+    {
+      "epoch": 0.5707317073170731,
+      "grad_norm": 2.2654404640197754,
+      "learning_rate": 4.960602014486225e-06,
+      "loss": 1.0101,
+      "step": 117
+    },
+    {
+      "epoch": 0.5756097560975609,
+      "grad_norm": 3.344377040863037,
+      "learning_rate": 4.959921639556199e-06,
+      "loss": 0.8391,
+      "step": 118
+    },
+    {
+      "epoch": 0.5804878048780487,
+      "grad_norm": 3.1620500087738037,
+      "learning_rate": 4.959235487481928e-06,
+      "loss": 1.0431,
+      "step": 119
+    },
+    {
+      "epoch": 0.5853658536585366,
+      "grad_norm": 2.857048273086548,
+      "learning_rate": 4.958543559874846e-06,
+      "loss": 0.5864,
+      "step": 120
+    },
+    {
+      "epoch": 0.5902439024390244,
+      "grad_norm": 3.1736063957214355,
+      "learning_rate": 4.9578458583599495e-06,
+      "loss": 0.7868,
+      "step": 121
+    },
+    {
+      "epoch": 0.5951219512195122,
+      "grad_norm": 3.5520827770233154,
+      "learning_rate": 4.957142384575795e-06,
+      "loss": 0.7901,
+      "step": 122
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.265103578567505,
+      "learning_rate": 4.956433140174498e-06,
+      "loss": 0.9067,
+      "step": 123
+    },
+    {
+      "epoch": 0.6048780487804878,
+      "grad_norm": 3.1181187629699707,
+      "learning_rate": 4.9557181268217225e-06,
+      "loss": 0.8971,
+      "step": 124
+    },
+    {
+      "epoch": 0.6097560975609756,
+      "grad_norm": 2.4123694896698,
+      "learning_rate": 4.954997346196683e-06,
+      "loss": 1.2123,
+      "step": 125
+    },
+    {
+      "epoch": 0.6146341463414634,
+      "grad_norm": 2.9646875858306885,
+      "learning_rate": 4.954270799992138e-06,
+      "loss": 0.7696,
+      "step": 126
+    },
+    {
+      "epoch": 0.6195121951219512,
+      "grad_norm": 2.7457995414733887,
+      "learning_rate": 4.953538489914387e-06,
+      "loss": 0.7919,
+      "step": 127
+    },
+    {
+      "epoch": 0.624390243902439,
+      "grad_norm": 5.096850395202637,
+      "learning_rate": 4.9528004176832654e-06,
+      "loss": 0.6494,
+      "step": 128
+    },
+    {
+      "epoch": 0.6292682926829268,
+      "grad_norm": 3.124955177307129,
+      "learning_rate": 4.952056585032142e-06,
+      "loss": 1.0546,
+      "step": 129
+    },
+    {
+      "epoch": 0.6341463414634146,
+      "grad_norm": 2.4860167503356934,
+      "learning_rate": 4.951306993707913e-06,
+      "loss": 0.7907,
+      "step": 130
+    },
+    {
+      "epoch": 0.6390243902439025,
+      "grad_norm": 2.3380239009857178,
+      "learning_rate": 4.950551645470998e-06,
+      "loss": 0.7433,
+      "step": 131
+    },
+    {
+      "epoch": 0.6439024390243903,
+      "grad_norm": 2.8945236206054688,
+      "learning_rate": 4.9497905420953406e-06,
+      "loss": 0.7682,
+      "step": 132
+    },
+    {
+      "epoch": 0.6487804878048781,
+      "grad_norm": 3.429776430130005,
+      "learning_rate": 4.949023685368395e-06,
+      "loss": 0.8411,
+      "step": 133
+    },
+    {
+      "epoch": 0.6536585365853659,
+      "grad_norm": 2.8853516578674316,
+      "learning_rate": 4.948251077091131e-06,
+      "loss": 1.0792,
+      "step": 134
+    },
+    {
+      "epoch": 0.6585365853658537,
+      "grad_norm": 2.145598888397217,
+      "learning_rate": 4.947472719078025e-06,
+      "loss": 0.8033,
+      "step": 135
+    },
+    {
+      "epoch": 0.6634146341463415,
+      "grad_norm": 2.5064377784729004,
+      "learning_rate": 4.9466886131570565e-06,
+      "loss": 0.939,
+      "step": 136
+    },
+    {
+      "epoch": 0.6682926829268293,
+      "grad_norm": 2.5700225830078125,
+      "learning_rate": 4.945898761169704e-06,
+      "loss": 1.0418,
+      "step": 137
+    },
+    {
+      "epoch": 0.6731707317073171,
+      "grad_norm": 2.3390917778015137,
+      "learning_rate": 4.945103164970941e-06,
+      "loss": 0.6158,
+      "step": 138
+    },
+    {
+      "epoch": 0.6780487804878049,
+      "grad_norm": 2.1538751125335693,
+      "learning_rate": 4.9443018264292304e-06,
+      "loss": 0.6995,
+      "step": 139
+    },
+    {
+      "epoch": 0.6829268292682927,
+      "grad_norm": 5.255710601806641,
+      "learning_rate": 4.9434947474265225e-06,
+      "loss": 1.0382,
+      "step": 140
+    },
+    {
+      "epoch": 0.6878048780487804,
+      "grad_norm": 2.5547356605529785,
+      "learning_rate": 4.942681929858249e-06,
+      "loss": 1.037,
+      "step": 141
+    },
+    {
+      "epoch": 0.6926829268292682,
+      "grad_norm": 2.613280773162842,
+      "learning_rate": 4.941863375633315e-06,
+      "loss": 0.9071,
+      "step": 142
+    },
+    {
+      "epoch": 0.697560975609756,
+      "grad_norm": 2.9957327842712402,
+      "learning_rate": 4.9410390866741056e-06,
+      "loss": 0.7908,
+      "step": 143
+    },
+    {
+      "epoch": 0.7024390243902439,
+      "grad_norm": 2.410107374191284,
+      "learning_rate": 4.9402090649164655e-06,
+      "loss": 0.7739,
+      "step": 144
+    },
+    {
+      "epoch": 0.7073170731707317,
+      "grad_norm": 2.352013349533081,
+      "learning_rate": 4.9393733123097085e-06,
+      "loss": 0.939,
+      "step": 145
+    },
+    {
+      "epoch": 0.7121951219512195,
+      "grad_norm": 2.5164194107055664,
+      "learning_rate": 4.9385318308166065e-06,
+      "loss": 0.8729,
+      "step": 146
+    },
+    {
+      "epoch": 0.7170731707317073,
+      "grad_norm": 4.213881015777588,
+      "learning_rate": 4.937684622413385e-06,
+      "loss": 0.6124,
+      "step": 147
+    },
+    {
+      "epoch": 0.7219512195121951,
+      "grad_norm": 2.7950191497802734,
+      "learning_rate": 4.9368316890897185e-06,
+      "loss": 0.975,
+      "step": 148
+    },
+    {
+      "epoch": 0.7268292682926829,
+      "grad_norm": 2.8618874549865723,
+      "learning_rate": 4.9359730328487264e-06,
+      "loss": 0.5832,
+      "step": 149
+    },
+    {
+      "epoch": 0.7317073170731707,
+      "grad_norm": 2.6943812370300293,
+      "learning_rate": 4.935108655706972e-06,
+      "loss": 0.8124,
+      "step": 150
+    },
+    {
+      "epoch": 0.7365853658536585,
+      "grad_norm": 3.2164082527160645,
+      "learning_rate": 4.934238559694448e-06,
+      "loss": 1.1446,
+      "step": 151
+    },
+    {
+      "epoch": 0.7414634146341463,
+      "grad_norm": 3.05002498626709,
+      "learning_rate": 4.9333627468545845e-06,
+      "loss": 0.7884,
+      "step": 152
+    },
+    {
+      "epoch": 0.7463414634146341,
+      "grad_norm": 2.863351583480835,
+      "learning_rate": 4.932481219244231e-06,
+      "loss": 0.7918,
+      "step": 153
+    },
+    {
+      "epoch": 0.751219512195122,
+      "grad_norm": 2.4947102069854736,
+      "learning_rate": 4.931593978933666e-06,
+      "loss": 0.775,
+      "step": 154
+    },
+    {
+      "epoch": 0.7560975609756098,
+      "grad_norm": 2.918886184692383,
+      "learning_rate": 4.930701028006577e-06,
+      "loss": 0.993,
+      "step": 155
+    },
+    {
+      "epoch": 0.7609756097560976,
+      "grad_norm": 2.835956573486328,
+      "learning_rate": 4.929802368560066e-06,
+      "loss": 0.7911,
+      "step": 156
+    },
+    {
+      "epoch": 0.7658536585365854,
+      "grad_norm": 3.3073575496673584,
+      "learning_rate": 4.928898002704642e-06,
+      "loss": 0.9346,
+      "step": 157
+    },
+    {
+      "epoch": 0.7707317073170732,
+      "grad_norm": 3.086146354675293,
+      "learning_rate": 4.927987932564215e-06,
+      "loss": 0.817,
+      "step": 158
+    },
+    {
+      "epoch": 0.775609756097561,
+      "grad_norm": 2.5419743061065674,
+      "learning_rate": 4.927072160276092e-06,
+      "loss": 0.7918,
+      "step": 159
+    },
+    {
+      "epoch": 0.7804878048780488,
+      "grad_norm": 3.984297275543213,
+      "learning_rate": 4.926150687990969e-06,
+      "loss": 0.7153,
+      "step": 160
+    },
+    {
+      "epoch": 0.7853658536585366,
+      "grad_norm": 2.4703335762023926,
+      "learning_rate": 4.925223517872934e-06,
+      "loss": 0.8982,
+      "step": 161
+    },
+    {
+      "epoch": 0.7902439024390244,
+      "grad_norm": 2.81785249710083,
+      "learning_rate": 4.9242906520994484e-06,
+      "loss": 0.9839,
+      "step": 162
+    },
+    {
+      "epoch": 0.7951219512195122,
+      "grad_norm": 2.3304924964904785,
+      "learning_rate": 4.923352092861358e-06,
+      "loss": 0.8406,
+      "step": 163
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.339498519897461,
+      "learning_rate": 4.922407842362875e-06,
+      "loss": 0.6602,
+      "step": 164
+    },
+    {
+      "epoch": 0.8048780487804879,
+      "grad_norm": 3.488255262374878,
+      "learning_rate": 4.921457902821578e-06,
+      "loss": 0.9779,
+      "step": 165
+    },
+    {
+      "epoch": 0.8097560975609757,
+      "grad_norm": 2.8528945446014404,
+      "learning_rate": 4.920502276468408e-06,
+      "loss": 0.8821,
+      "step": 166
+    },
+    {
+      "epoch": 0.8146341463414634,
+      "grad_norm": 3.4649784564971924,
+      "learning_rate": 4.9195409655476605e-06,
+      "loss": 0.7539,
+      "step": 167
+    },
+    {
+      "epoch": 0.8195121951219512,
+      "grad_norm": 2.3109042644500732,
+      "learning_rate": 4.918573972316982e-06,
+      "loss": 0.9807,
+      "step": 168
+    },
+    {
+      "epoch": 0.824390243902439,
+      "grad_norm": 2.678666353225708,
+      "learning_rate": 4.917601299047361e-06,
+      "loss": 0.8318,
+      "step": 169
+    },
+    {
+      "epoch": 0.8292682926829268,
+      "grad_norm": 2.730614185333252,
+      "learning_rate": 4.916622948023129e-06,
+      "loss": 0.7816,
+      "step": 170
+    },
+    {
+      "epoch": 0.8341463414634146,
+      "grad_norm": 2.9835665225982666,
+      "learning_rate": 4.915638921541952e-06,
+      "loss": 0.6633,
+      "step": 171
+    },
+    {
+      "epoch": 0.8390243902439024,
+      "grad_norm": 3.31217360496521,
+      "learning_rate": 4.914649221914822e-06,
+      "loss": 0.9296,
+      "step": 172
+    },
+    {
+      "epoch": 0.8439024390243902,
+      "grad_norm": 2.9021658897399902,
+      "learning_rate": 4.913653851466057e-06,
+      "loss": 0.6864,
+      "step": 173
+    },
+    {
+      "epoch": 0.848780487804878,
+      "grad_norm": 3.3672914505004883,
+      "learning_rate": 4.912652812533291e-06,
+      "loss": 0.8599,
+      "step": 174
+    },
+    {
+      "epoch": 0.8536585365853658,
+      "grad_norm": 2.4871644973754883,
+      "learning_rate": 4.911646107467472e-06,
+      "loss": 0.8949,
+      "step": 175
+    },
+    {
+      "epoch": 0.8585365853658536,
+      "grad_norm": 2.728022813796997,
+      "learning_rate": 4.9106337386328524e-06,
+      "loss": 0.9758,
+      "step": 176
+    },
+    {
+      "epoch": 0.8634146341463415,
+      "grad_norm": 2.704252243041992,
+      "learning_rate": 4.909615708406991e-06,
+      "loss": 0.8954,
+      "step": 177
+    },
+    {
+      "epoch": 0.8682926829268293,
+      "grad_norm": 2.4002223014831543,
+      "learning_rate": 4.908592019180738e-06,
+      "loss": 0.7157,
+      "step": 178
+    },
+    {
+      "epoch": 0.8731707317073171,
+      "grad_norm": 2.1927788257598877,
+      "learning_rate": 4.907562673358234e-06,
+      "loss": 0.6358,
+      "step": 179
+    },
+    {
+      "epoch": 0.8780487804878049,
+      "grad_norm": 2.458500623703003,
+      "learning_rate": 4.906527673356907e-06,
+      "loss": 0.6685,
+      "step": 180
+    },
+    {
+      "epoch": 0.8829268292682927,
+      "grad_norm": 2.5924787521362305,
+      "learning_rate": 4.905487021607462e-06,
+      "loss": 0.5686,
+      "step": 181
+    },
+    {
+      "epoch": 0.8878048780487805,
+      "grad_norm": 3.0923380851745605,
+      "learning_rate": 4.904440720553876e-06,
+      "loss": 0.8538,
+      "step": 182
+    },
+    {
+      "epoch": 0.8926829268292683,
+      "grad_norm": 2.8001527786254883,
+      "learning_rate": 4.903388772653396e-06,
+      "loss": 0.8292,
+      "step": 183
+    },
+    {
+      "epoch": 0.8975609756097561,
+      "grad_norm": 2.4344072341918945,
+      "learning_rate": 4.902331180376529e-06,
+      "loss": 0.7946,
+      "step": 184
+    },
+    {
+      "epoch": 0.9024390243902439,
+      "grad_norm": 2.6313226222991943,
+      "learning_rate": 4.901267946207038e-06,
+      "loss": 0.9269,
+      "step": 185
+    },
+    {
+      "epoch": 0.9073170731707317,
+      "grad_norm": 2.4776692390441895,
+      "learning_rate": 4.900199072641937e-06,
+      "loss": 0.7433,
+      "step": 186
+    },
+    {
+      "epoch": 0.9121951219512195,
+      "grad_norm": 2.339869260787964,
+      "learning_rate": 4.899124562191484e-06,
+      "loss": 0.6577,
+      "step": 187
+    },
+    {
+      "epoch": 0.9170731707317074,
+      "grad_norm": 3.076890468597412,
+      "learning_rate": 4.8980444173791735e-06,
+      "loss": 0.5989,
+      "step": 188
+    },
+    {
+      "epoch": 0.9219512195121952,
+      "grad_norm": 2.83957839012146,
+      "learning_rate": 4.896958640741735e-06,
+      "loss": 0.9364,
+      "step": 189
+    },
+    {
+      "epoch": 0.926829268292683,
+      "grad_norm": 2.770867347717285,
+      "learning_rate": 4.895867234829121e-06,
+      "loss": 1.0328,
+      "step": 190
+    },
+    {
+      "epoch": 0.9317073170731708,
+      "grad_norm": 2.7819619178771973,
+      "learning_rate": 4.894770202204509e-06,
+      "loss": 0.772,
+      "step": 191
+    },
+    {
+      "epoch": 0.9365853658536586,
+      "grad_norm": 3.925703763961792,
+      "learning_rate": 4.893667545444285e-06,
+      "loss": 0.8128,
+      "step": 192
+    },
+    {
+      "epoch": 0.9414634146341463,
+      "grad_norm": 3.034944534301758,
+      "learning_rate": 4.8925592671380495e-06,
+      "loss": 0.7418,
+      "step": 193
+    },
+    {
+      "epoch": 0.9463414634146341,
+      "grad_norm": 2.3350143432617188,
+      "learning_rate": 4.891445369888601e-06,
+      "loss": 0.5979,
+      "step": 194
+    },
+    {
+      "epoch": 0.9512195121951219,
+      "grad_norm": 2.6433160305023193,
+      "learning_rate": 4.890325856311936e-06,
+      "loss": 0.9664,
+      "step": 195
+    },
+    {
+      "epoch": 0.9560975609756097,
+      "grad_norm": 2.715142011642456,
+      "learning_rate": 4.889200729037241e-06,
+      "loss": 0.8482,
+      "step": 196
+    },
+    {
+      "epoch": 0.9609756097560975,
+      "grad_norm": 2.6157352924346924,
+      "learning_rate": 4.888069990706884e-06,
+      "loss": 0.7173,
+      "step": 197
+    },
+    {
+      "epoch": 0.9658536585365853,
+      "grad_norm": 3.7308952808380127,
+      "learning_rate": 4.886933643976414e-06,
+      "loss": 0.5433,
+      "step": 198
+    },
+    {
+      "epoch": 0.9707317073170731,
+      "grad_norm": 3.1134045124053955,
+      "learning_rate": 4.885791691514548e-06,
+      "loss": 0.5997,
+      "step": 199
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 2.421365976333618,
+      "learning_rate": 4.884644136003172e-06,
+      "loss": 0.6477,
+      "step": 200
+    },
+    {
+      "epoch": 0.9804878048780488,
+      "grad_norm": 2.8676180839538574,
+      "learning_rate": 4.883490980137327e-06,
+      "loss": 1.3465,
+      "step": 201
+    },
+    {
+      "epoch": 0.9853658536585366,
+      "grad_norm": 2.236189603805542,
+      "learning_rate": 4.882332226625208e-06,
+      "loss": 0.7533,
+      "step": 202
+    },
+    {
+      "epoch": 0.9902439024390244,
+      "grad_norm": 2.2514970302581787,
+      "learning_rate": 4.881167878188158e-06,
+      "loss": 0.8555,
+      "step": 203
+    },
+    {
+      "epoch": 0.9951219512195122,
+      "grad_norm": 2.6856095790863037,
+      "learning_rate": 4.8799979375606565e-06,
+      "loss": 0.7634,
+      "step": 204
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5563852787017822,
+      "learning_rate": 4.878822407490319e-06,
+      "loss": 0.66,
+      "step": 205
+    },
+    {
+      "epoch": 1.0048780487804878,
+      "grad_norm": 4.7092814445495605,
+      "learning_rate": 4.8776412907378845e-06,
+      "loss": 0.7429,
+      "step": 206
+    },
+    {
+      "epoch": 1.0097560975609756,
+      "grad_norm": 2.9133448600769043,
+      "learning_rate": 4.876454590077216e-06,
+      "loss": 0.5735,
+      "step": 207
+    },
+    {
+      "epoch": 1.0146341463414634,
+      "grad_norm": 2.7012641429901123,
+      "learning_rate": 4.875262308295289e-06,
+      "loss": 0.8065,
+      "step": 208
+    },
+    {
+      "epoch": 1.0195121951219512,
+      "grad_norm": 3.703998327255249,
+      "learning_rate": 4.874064448192185e-06,
+      "loss": 0.7148,
+      "step": 209
+    },
+    {
+      "epoch": 1.024390243902439,
+      "grad_norm": 3.044930934906006,
+      "learning_rate": 4.872861012581088e-06,
+      "loss": 0.5606,
+      "step": 210
+    },
+    {
+      "epoch": 1.0292682926829269,
+      "grad_norm": 3.661381244659424,
+      "learning_rate": 4.871652004288275e-06,
+      "loss": 0.6492,
+      "step": 211
+    },
+    {
+      "epoch": 1.0341463414634147,
+      "grad_norm": 3.18344783782959,
+      "learning_rate": 4.870437426153113e-06,
+      "loss": 0.633,
+      "step": 212
+    },
+    {
+      "epoch": 1.0390243902439025,
+      "grad_norm": 4.596707820892334,
+      "learning_rate": 4.869217281028045e-06,
+      "loss": 0.842,
+      "step": 213
+    },
+    {
+      "epoch": 1.0439024390243903,
+      "grad_norm": 4.116331577301025,
+      "learning_rate": 4.867991571778592e-06,
+      "loss": 0.8371,
+      "step": 214
+    },
+    {
+      "epoch": 1.048780487804878,
+      "grad_norm": 3.152939558029175,
+      "learning_rate": 4.866760301283342e-06,
+      "loss": 0.4728,
+      "step": 215
+    },
+    {
+      "epoch": 1.053658536585366,
+      "grad_norm": 2.8732805252075195,
+      "learning_rate": 4.865523472433942e-06,
+      "loss": 0.651,
+      "step": 216
+    },
+    {
+      "epoch": 1.0585365853658537,
+      "grad_norm": 2.967480421066284,
+      "learning_rate": 4.8642810881350935e-06,
+      "loss": 0.6361,
+      "step": 217
+    },
+    {
+      "epoch": 1.0634146341463415,
+      "grad_norm": 2.816798210144043,
+      "learning_rate": 4.863033151304546e-06,
+      "loss": 0.6206,
+      "step": 218
+    },
+    {
+      "epoch": 1.0682926829268293,
+      "grad_norm": 3.168349027633667,
+      "learning_rate": 4.861779664873088e-06,
+      "loss": 0.7782,
+      "step": 219
+    },
+    {
+      "epoch": 1.0731707317073171,
+      "grad_norm": 3.7496471405029297,
+      "learning_rate": 4.8605206317845425e-06,
+      "loss": 0.8504,
+      "step": 220
+    },
+    {
+      "epoch": 1.078048780487805,
+      "grad_norm": 2.7087056636810303,
+      "learning_rate": 4.859256054995758e-06,
+      "loss": 0.7771,
+      "step": 221
+    },
+    {
+      "epoch": 1.0829268292682928,
+      "grad_norm": 2.803703546524048,
+      "learning_rate": 4.8579859374766e-06,
+      "loss": 0.4308,
+      "step": 222
+    },
+    {
+      "epoch": 1.0878048780487806,
+      "grad_norm": 2.4199142456054688,
+      "learning_rate": 4.856710282209952e-06,
+      "loss": 0.3739,
+      "step": 223
+    },
+    {
+      "epoch": 1.0926829268292684,
+      "grad_norm": 2.384037494659424,
+      "learning_rate": 4.855429092191698e-06,
+      "loss": 0.6548,
+      "step": 224
+    },
+    {
+      "epoch": 1.0975609756097562,
+      "grad_norm": 3.0230021476745605,
+      "learning_rate": 4.854142370430725e-06,
+      "loss": 0.6932,
+      "step": 225
+    },
+    {
+      "epoch": 1.102439024390244,
+      "grad_norm": 3.0248661041259766,
+      "learning_rate": 4.8528501199489045e-06,
+      "loss": 0.6491,
+      "step": 226
+    },
+    {
+      "epoch": 1.1073170731707318,
+      "grad_norm": 4.046666145324707,
+      "learning_rate": 4.851552343781099e-06,
+      "loss": 0.7946,
+      "step": 227
+    },
+    {
+      "epoch": 1.1121951219512196,
+      "grad_norm": 2.8751168251037598,
+      "learning_rate": 4.850249044975145e-06,
+      "loss": 0.7629,
+      "step": 228
+    },
+    {
+      "epoch": 1.1170731707317074,
+      "grad_norm": 2.8649816513061523,
+      "learning_rate": 4.848940226591849e-06,
+      "loss": 0.9114,
+      "step": 229
+    },
+    {
+      "epoch": 1.1219512195121952,
+      "grad_norm": 3.2590744495391846,
+      "learning_rate": 4.847625891704982e-06,
+      "loss": 0.535,
+      "step": 230
+    },
+    {
+      "epoch": 1.126829268292683,
+      "grad_norm": 3.230659008026123,
+      "learning_rate": 4.846306043401268e-06,
+      "loss": 0.7134,
+      "step": 231
+    },
+    {
+      "epoch": 1.1317073170731708,
+      "grad_norm": 3.5220088958740234,
+      "learning_rate": 4.844980684780381e-06,
+      "loss": 0.5375,
+      "step": 232
+    },
+    {
+      "epoch": 1.1365853658536587,
+      "grad_norm": 3.074052095413208,
+      "learning_rate": 4.8436498189549345e-06,
+      "loss": 0.5486,
+      "step": 233
+    },
+    {
+      "epoch": 1.1414634146341462,
+      "grad_norm": 2.511216163635254,
+      "learning_rate": 4.842313449050477e-06,
+      "loss": 0.5203,
+      "step": 234
+    },
+    {
+      "epoch": 1.146341463414634,
+      "grad_norm": 2.6082136631011963,
+      "learning_rate": 4.840971578205486e-06,
+      "loss": 0.4978,
+      "step": 235
+    },
+    {
+      "epoch": 1.1512195121951219,
+      "grad_norm": 2.4481778144836426,
+      "learning_rate": 4.839624209571352e-06,
+      "loss": 0.348,
+      "step": 236
+    },
+    {
+      "epoch": 1.1560975609756097,
+      "grad_norm": 2.7532148361206055,
+      "learning_rate": 4.838271346312381e-06,
+      "loss": 0.8068,
+      "step": 237
+    },
+    {
+      "epoch": 1.1609756097560975,
+      "grad_norm": 2.6562349796295166,
+      "learning_rate": 4.836912991605782e-06,
+      "loss": 0.8823,
+      "step": 238
+    },
+    {
+      "epoch": 1.1658536585365853,
+      "grad_norm": 3.032168388366699,
+      "learning_rate": 4.835549148641663e-06,
+      "loss": 0.501,
+      "step": 239
+    },
+    {
+      "epoch": 1.170731707317073,
+      "grad_norm": 3.4816956520080566,
+      "learning_rate": 4.834179820623018e-06,
+      "loss": 0.6406,
+      "step": 240
+    },
+    {
+      "epoch": 1.175609756097561,
+      "grad_norm": 2.480642318725586,
+      "learning_rate": 4.832805010765724e-06,
+      "loss": 0.537,
+      "step": 241
+    },
+    {
+      "epoch": 1.1804878048780487,
+      "grad_norm": 2.7662222385406494,
+      "learning_rate": 4.831424722298531e-06,
+      "loss": 0.6464,
+      "step": 242
+    },
+    {
+      "epoch": 1.1853658536585365,
+      "grad_norm": 3.2929866313934326,
+      "learning_rate": 4.830038958463061e-06,
+      "loss": 0.6888,
+      "step": 243
+    },
+    {
+      "epoch": 1.1902439024390243,
+      "grad_norm": 5.094089031219482,
+      "learning_rate": 4.828647722513785e-06,
+      "loss": 0.8342,
+      "step": 244
+    },
+    {
+      "epoch": 1.1951219512195121,
+      "grad_norm": 3.6679818630218506,
+      "learning_rate": 4.827251017718034e-06,
+      "loss": 0.7849,
+      "step": 245
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.97290301322937,
+      "learning_rate": 4.8258488473559794e-06,
+      "loss": 0.7995,
+      "step": 246
+    },
+    {
+      "epoch": 1.2048780487804878,
+      "grad_norm": 3.3555023670196533,
+      "learning_rate": 4.824441214720629e-06,
+      "loss": 0.8718,
+      "step": 247
+    },
+    {
+      "epoch": 1.2097560975609756,
+      "grad_norm": 2.309361219406128,
+      "learning_rate": 4.823028123117818e-06,
+      "loss": 0.3731,
+      "step": 248
+    },
+    {
+      "epoch": 1.2146341463414634,
+      "grad_norm": 2.607269763946533,
+      "learning_rate": 4.8216095758662015e-06,
+      "loss": 0.7321,
+      "step": 249
+    },
+    {
+      "epoch": 1.2195121951219512,
+      "grad_norm": 2.5667428970336914,
+      "learning_rate": 4.82018557629725e-06,
+      "loss": 0.7561,
+      "step": 250
+    },
+    {
+      "epoch": 1.224390243902439,
+      "grad_norm": 2.7664871215820312,
+      "learning_rate": 4.8187561277552376e-06,
+      "loss": 0.638,
+      "step": 251
+    },
+    {
+      "epoch": 1.2292682926829268,
+      "grad_norm": 2.2880401611328125,
+      "learning_rate": 4.817321233597232e-06,
+      "loss": 0.6996,
+      "step": 252
+    },
+    {
+      "epoch": 1.2341463414634146,
+      "grad_norm": 2.7615559101104736,
+      "learning_rate": 4.815880897193095e-06,
+      "loss": 0.5432,
+      "step": 253
+    },
+    {
+      "epoch": 1.2390243902439024,
+      "grad_norm": 2.9052155017852783,
+      "learning_rate": 4.814435121925466e-06,
+      "loss": 0.781,
+      "step": 254
+    },
+    {
+      "epoch": 1.2439024390243902,
+      "grad_norm": 3.2035205364227295,
+      "learning_rate": 4.812983911189761e-06,
+      "loss": 0.6884,
+      "step": 255
+    },
+    {
+      "epoch": 1.248780487804878,
+      "grad_norm": 2.8139917850494385,
+      "learning_rate": 4.811527268394157e-06,
+      "loss": 0.4984,
+      "step": 256
+    },
+    {
+      "epoch": 1.2536585365853659,
+      "grad_norm": 2.849602699279785,
+      "learning_rate": 4.810065196959591e-06,
+      "loss": 0.553,
+      "step": 257
+    },
+    {
+      "epoch": 1.2585365853658537,
+      "grad_norm": 2.8745057582855225,
+      "learning_rate": 4.8085977003197496e-06,
+      "loss": 0.7955,
+      "step": 258
+    },
+    {
+      "epoch": 1.2634146341463415,
+      "grad_norm": 3.4053122997283936,
+      "learning_rate": 4.807124781921059e-06,
+      "loss": 0.9715,
+      "step": 259
+    },
+    {
+      "epoch": 1.2682926829268293,
+      "grad_norm": 3.1741702556610107,
+      "learning_rate": 4.805646445222679e-06,
+      "loss": 0.6306,
+      "step": 260
+    },
+    {
+      "epoch": 1.273170731707317,
+      "grad_norm": 2.5348331928253174,
+      "learning_rate": 4.804162693696494e-06,
+      "loss": 0.5192,
+      "step": 261
+    },
+    {
+      "epoch": 1.278048780487805,
+      "grad_norm": 3.2491304874420166,
+      "learning_rate": 4.802673530827105e-06,
+      "loss": 0.5369,
+      "step": 262
+    },
+    {
+      "epoch": 1.2829268292682927,
+      "grad_norm": 2.670273780822754,
+      "learning_rate": 4.801178960111823e-06,
+      "loss": 0.5864,
+      "step": 263
+    },
+    {
+      "epoch": 1.2878048780487805,
+      "grad_norm": 2.5655579566955566,
+      "learning_rate": 4.799678985060658e-06,
+      "loss": 0.7864,
+      "step": 264
+    },
+    {
+      "epoch": 1.2926829268292683,
+      "grad_norm": 2.6352531909942627,
+      "learning_rate": 4.798173609196314e-06,
+      "loss": 0.8198,
+      "step": 265
+    },
+    {
+      "epoch": 1.2975609756097561,
+      "grad_norm": 3.028343677520752,
+      "learning_rate": 4.796662836054176e-06,
+      "loss": 0.4621,
+      "step": 266
+    },
+    {
+      "epoch": 1.302439024390244,
+      "grad_norm": 2.757690191268921,
+      "learning_rate": 4.795146669182304e-06,
+      "loss": 0.6237,
+      "step": 267
+    },
+    {
+      "epoch": 1.3073170731707318,
+      "grad_norm": 2.564842462539673,
+      "learning_rate": 4.793625112141431e-06,
+      "loss": 0.4981,
+      "step": 268
+    },
+    {
+      "epoch": 1.3121951219512196,
+      "grad_norm": 2.69234299659729,
+      "learning_rate": 4.792098168504943e-06,
+      "loss": 0.5384,
+      "step": 269
+    },
+    {
+      "epoch": 1.3170731707317074,
+      "grad_norm": 2.794144868850708,
+      "learning_rate": 4.790565841858879e-06,
+      "loss": 0.5535,
+      "step": 270
+    },
+    {
+      "epoch": 1.3219512195121952,
+      "grad_norm": 2.850296974182129,
+      "learning_rate": 4.789028135801919e-06,
+      "loss": 0.7492,
+      "step": 271
+    },
+    {
+      "epoch": 1.326829268292683,
+      "grad_norm": 3.287806987762451,
+      "learning_rate": 4.787485053945377e-06,
+      "loss": 0.8367,
+      "step": 272
+    },
+    {
+      "epoch": 1.3317073170731708,
+      "grad_norm": 2.479343891143799,
+      "learning_rate": 4.785936599913193e-06,
+      "loss": 0.6875,
+      "step": 273
+    },
+    {
+      "epoch": 1.3365853658536586,
+      "grad_norm": 3.171198844909668,
+      "learning_rate": 4.784382777341922e-06,
+      "loss": 0.733,
+      "step": 274
+    },
+    {
+      "epoch": 1.3414634146341464,
+      "grad_norm": 2.866610050201416,
+      "learning_rate": 4.782823589880729e-06,
+      "loss": 0.9719,
+      "step": 275
+    },
+    {
+      "epoch": 1.346341463414634,
+      "grad_norm": 2.3714404106140137,
+      "learning_rate": 4.7812590411913755e-06,
+      "loss": 0.6979,
+      "step": 276
+    },
+    {
+      "epoch": 1.3512195121951218,
+      "grad_norm": 2.3838706016540527,
+      "learning_rate": 4.779689134948217e-06,
+      "loss": 0.9697,
+      "step": 277
+    },
+    {
+      "epoch": 1.3560975609756096,
+      "grad_norm": 3.2992005348205566,
+      "learning_rate": 4.77811387483819e-06,
+      "loss": 0.4799,
+      "step": 278
+    },
+    {
+      "epoch": 1.3609756097560974,
+      "grad_norm": 3.403024435043335,
+      "learning_rate": 4.776533264560804e-06,
+      "loss": 0.7478,
+      "step": 279
+    },
+    {
+      "epoch": 1.3658536585365852,
+      "grad_norm": 2.669820785522461,
+      "learning_rate": 4.774947307828134e-06,
+      "loss": 0.8622,
+      "step": 280
+    },
+    {
+      "epoch": 1.370731707317073,
+      "grad_norm": 2.4695041179656982,
+      "learning_rate": 4.773356008364812e-06,
+      "loss": 0.5792,
+      "step": 281
+    },
+    {
+      "epoch": 1.3756097560975609,
+      "grad_norm": 3.1744325160980225,
+      "learning_rate": 4.771759369908017e-06,
+      "loss": 0.4368,
+      "step": 282
+    },
+    {
+      "epoch": 1.3804878048780487,
+      "grad_norm": 2.8564929962158203,
+      "learning_rate": 4.7701573962074635e-06,
+      "loss": 0.6337,
+      "step": 283
+    },
+    {
+      "epoch": 1.3853658536585365,
+      "grad_norm": 2.4109890460968018,
+      "learning_rate": 4.7685500910254015e-06,
+      "loss": 0.5042,
+      "step": 284
+    },
+    {
+      "epoch": 1.3902439024390243,
+      "grad_norm": 2.389765977859497,
+      "learning_rate": 4.766937458136598e-06,
+      "loss": 0.7427,
+      "step": 285
+    },
+    {
+      "epoch": 1.395121951219512,
+      "grad_norm": 2.412153720855713,
+      "learning_rate": 4.765319501328332e-06,
+      "loss": 0.6956,
+      "step": 286
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.6756227016448975,
+      "learning_rate": 4.763696224400391e-06,
+      "loss": 0.5152,
+      "step": 287
+    },
+    {
+      "epoch": 1.4048780487804877,
+      "grad_norm": 2.4644389152526855,
+      "learning_rate": 4.762067631165049e-06,
+      "loss": 0.5583,
+      "step": 288
+    },
+    {
+      "epoch": 1.4097560975609755,
+      "grad_norm": 2.6496896743774414,
+      "learning_rate": 4.760433725447071e-06,
+      "loss": 0.6824,
+      "step": 289
+    },
+    {
+      "epoch": 1.4146341463414633,
+      "grad_norm": 2.9843268394470215,
+      "learning_rate": 4.758794511083697e-06,
+      "loss": 0.7914,
+      "step": 290
+    },
+    {
+      "epoch": 1.4195121951219511,
+      "grad_norm": 3.639101266860962,
+      "learning_rate": 4.757149991924633e-06,
+      "loss": 0.6827,
+      "step": 291
+    },
+    {
+      "epoch": 1.424390243902439,
+      "grad_norm": 3.2047319412231445,
+      "learning_rate": 4.755500171832045e-06,
+      "loss": 0.5908,
+      "step": 292
+    },
+    {
+      "epoch": 1.4292682926829268,
+      "grad_norm": 2.463202953338623,
+      "learning_rate": 4.753845054680548e-06,
+      "loss": 0.6469,
+      "step": 293
+    },
+    {
+      "epoch": 1.4341463414634146,
+      "grad_norm": 2.711195945739746,
+      "learning_rate": 4.752184644357197e-06,
+      "loss": 0.5412,
+      "step": 294
+    },
+    {
+      "epoch": 1.4390243902439024,
+      "grad_norm": 2.239082098007202,
+      "learning_rate": 4.750518944761477e-06,
+      "loss": 0.5324,
+      "step": 295
+    },
+    {
+      "epoch": 1.4439024390243902,
+      "grad_norm": 2.711050271987915,
+      "learning_rate": 4.748847959805297e-06,
+      "loss": 0.5317,
+      "step": 296
+    },
+    {
+      "epoch": 1.448780487804878,
+      "grad_norm": 2.4389946460723877,
+      "learning_rate": 4.7471716934129774e-06,
+      "loss": 0.5199,
+      "step": 297
+    },
+    {
+      "epoch": 1.4536585365853658,
+      "grad_norm": 2.6532390117645264,
+      "learning_rate": 4.745490149521242e-06,
+      "loss": 0.4874,
+      "step": 298
+    },
+    {
+      "epoch": 1.4585365853658536,
+      "grad_norm": 2.2970616817474365,
+      "learning_rate": 4.743803332079209e-06,
+      "loss": 0.5416,
+      "step": 299
+    },
+    {
+      "epoch": 1.4634146341463414,
+      "grad_norm": 2.4206762313842773,
+      "learning_rate": 4.742111245048382e-06,
+      "loss": 0.5628,
+      "step": 300
+    },
+    {
+      "epoch": 1.4682926829268292,
+      "grad_norm": 2.7086844444274902,
+      "learning_rate": 4.740413892402639e-06,
+      "loss": 0.5847,
+      "step": 301
+    },
+    {
+      "epoch": 1.473170731707317,
+      "grad_norm": 2.848602771759033,
+      "learning_rate": 4.738711278128228e-06,
+      "loss": 0.5889,
+      "step": 302
+    },
+    {
+      "epoch": 1.4780487804878049,
+      "grad_norm": 3.5257909297943115,
+      "learning_rate": 4.7370034062237476e-06,
+      "loss": 0.3917,
+      "step": 303
+    },
+    {
+      "epoch": 1.4829268292682927,
+      "grad_norm": 6.47664213180542,
+      "learning_rate": 4.73529028070015e-06,
+      "loss": 0.5592,
+      "step": 304
+    },
+    {
+      "epoch": 1.4878048780487805,
+      "grad_norm": 2.8833930492401123,
+      "learning_rate": 4.733571905580723e-06,
+      "loss": 0.843,
+      "step": 305
+    },
+    {
+      "epoch": 1.4926829268292683,
+      "grad_norm": 2.9924156665802,
+      "learning_rate": 4.731848284901082e-06,
+      "loss": 0.7041,
+      "step": 306
+    },
+    {
+      "epoch": 1.497560975609756,
+      "grad_norm": 2.9858405590057373,
+      "learning_rate": 4.730119422709165e-06,
+      "loss": 0.4914,
+      "step": 307
+    },
+    {
+      "epoch": 1.502439024390244,
+      "grad_norm": 3.4032366275787354,
+      "learning_rate": 4.728385323065215e-06,
+      "loss": 0.644,
+      "step": 308
+    },
+    {
+      "epoch": 1.5073170731707317,
+      "grad_norm": 2.86360502243042,
+      "learning_rate": 4.7266459900417815e-06,
+      "loss": 0.5335,
+      "step": 309
+    },
+    {
+      "epoch": 1.5121951219512195,
+      "grad_norm": 3.183012008666992,
+      "learning_rate": 4.724901427723698e-06,
+      "loss": 0.8275,
+      "step": 310
+    },
+    {
+      "epoch": 1.5170731707317073,
+      "grad_norm": 3.4128706455230713,
+      "learning_rate": 4.723151640208084e-06,
+      "loss": 0.4091,
+      "step": 311
+    },
+    {
+      "epoch": 1.5219512195121951,
+      "grad_norm": 2.765897512435913,
+      "learning_rate": 4.721396631604327e-06,
+      "loss": 0.4414,
+      "step": 312
+    },
+    {
+      "epoch": 1.526829268292683,
+      "grad_norm": 3.2348268032073975,
+      "learning_rate": 4.7196364060340785e-06,
+      "loss": 0.5423,
+      "step": 313
+    },
+    {
+      "epoch": 1.5317073170731708,
+      "grad_norm": 2.7270045280456543,
+      "learning_rate": 4.7178709676312416e-06,
+      "loss": 0.8072,
+      "step": 314
+    },
+    {
+      "epoch": 1.5365853658536586,
+      "grad_norm": 2.525298833847046,
+      "learning_rate": 4.716100320541961e-06,
+      "loss": 1.0254,
+      "step": 315
+    },
+    {
+      "epoch": 1.5414634146341464,
+      "grad_norm": 2.371321678161621,
+      "learning_rate": 4.714324468924614e-06,
+      "loss": 0.6541,
+      "step": 316
+    },
+    {
+      "epoch": 1.5463414634146342,
+      "grad_norm": 3.0820438861846924,
+      "learning_rate": 4.712543416949803e-06,
+      "loss": 0.7519,
+      "step": 317
+    },
+    {
+      "epoch": 1.551219512195122,
+      "grad_norm": 2.710369348526001,
+      "learning_rate": 4.71075716880034e-06,
+      "loss": 0.7232,
+      "step": 318
+    },
+    {
+      "epoch": 1.5560975609756098,
+      "grad_norm": 2.4568352699279785,
+      "learning_rate": 4.708965728671243e-06,
+      "loss": 0.8059,
+      "step": 319
+    },
+    {
+      "epoch": 1.5609756097560976,
+      "grad_norm": 2.7511191368103027,
+      "learning_rate": 4.7071691007697214e-06,
+      "loss": 0.6579,
+      "step": 320
+    },
+    {
+      "epoch": 1.5658536585365854,
+      "grad_norm": 2.6519858837127686,
+      "learning_rate": 4.705367289315172e-06,
+      "loss": 0.6989,
+      "step": 321
+    },
+    {
+      "epoch": 1.5707317073170732,
+      "grad_norm": 2.763019323348999,
+      "learning_rate": 4.703560298539158e-06,
+      "loss": 0.4916,
+      "step": 322
+    },
+    {
+      "epoch": 1.575609756097561,
+      "grad_norm": 2.6480252742767334,
+      "learning_rate": 4.701748132685415e-06,
+      "loss": 0.5076,
+      "step": 323
+    },
+    {
+      "epoch": 1.5804878048780489,
+      "grad_norm": 2.4289543628692627,
+      "learning_rate": 4.699930796009825e-06,
+      "loss": 0.559,
+      "step": 324
+    },
+    {
+      "epoch": 1.5853658536585367,
+      "grad_norm": 4.0515899658203125,
+      "learning_rate": 4.698108292780418e-06,
+      "loss": 0.7388,
+      "step": 325
+    },
+    {
+      "epoch": 1.5902439024390245,
+      "grad_norm": 2.5959129333496094,
+      "learning_rate": 4.696280627277356e-06,
+      "loss": 0.5469,
+      "step": 326
+    },
+    {
+      "epoch": 1.5951219512195123,
+      "grad_norm": 2.3453526496887207,
+      "learning_rate": 4.6944478037929255e-06,
+      "loss": 0.5494,
+      "step": 327
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.7527170181274414,
+      "learning_rate": 4.692609826631525e-06,
+      "loss": 0.7536,
+      "step": 328
+    },
+    {
+      "epoch": 1.604878048780488,
+      "grad_norm": 3.423588275909424,
+      "learning_rate": 4.690766700109659e-06,
+      "loss": 0.4586,
+      "step": 329
+    },
+    {
+      "epoch": 1.6097560975609757,
+      "grad_norm": 2.620429754257202,
+      "learning_rate": 4.6889184285559234e-06,
+      "loss": 0.4799,
+      "step": 330
+    },
+    {
+      "epoch": 1.6146341463414635,
+      "grad_norm": 6.416718006134033,
+      "learning_rate": 4.687065016310996e-06,
+      "loss": 0.7502,
+      "step": 331
+    },
+    {
+      "epoch": 1.6195121951219513,
+      "grad_norm": 2.7324717044830322,
+      "learning_rate": 4.685206467727631e-06,
+      "loss": 0.5923,
+      "step": 332
+    },
+    {
+      "epoch": 1.6243902439024391,
+      "grad_norm": 2.582935333251953,
+      "learning_rate": 4.683342787170644e-06,
+      "loss": 0.5619,
+      "step": 333
+    },
+    {
+      "epoch": 1.629268292682927,
+      "grad_norm": 2.8339877128601074,
+      "learning_rate": 4.6814739790169006e-06,
+      "loss": 0.55,
+      "step": 334
+    },
+    {
+      "epoch": 1.6341463414634148,
+      "grad_norm": 2.733982563018799,
+      "learning_rate": 4.679600047655313e-06,
+      "loss": 0.7243,
+      "step": 335
+    },
+    {
+      "epoch": 1.6390243902439026,
+      "grad_norm": 3.192747116088867,
+      "learning_rate": 4.6777209974868194e-06,
+      "loss": 1.132,
+      "step": 336
+    },
+    {
+      "epoch": 1.6439024390243904,
+      "grad_norm": 2.5185582637786865,
+      "learning_rate": 4.675836832924387e-06,
+      "loss": 0.55,
+      "step": 337
+    },
+    {
+      "epoch": 1.6487804878048782,
+      "grad_norm": 2.7306225299835205,
+      "learning_rate": 4.673947558392989e-06,
+      "loss": 0.4418,
+      "step": 338
+    },
+    {
+      "epoch": 1.653658536585366,
+      "grad_norm": 2.7026166915893555,
+      "learning_rate": 4.6720531783296e-06,
+      "loss": 0.5897,
+      "step": 339
+    },
+    {
+      "epoch": 1.6585365853658538,
+      "grad_norm": 2.5981674194335938,
+      "learning_rate": 4.670153697183185e-06,
+      "loss": 0.5889,
+      "step": 340
+    },
+    {
+      "epoch": 1.6634146341463416,
+      "grad_norm": 3.0985405445098877,
+      "learning_rate": 4.668249119414692e-06,
+      "loss": 0.5607,
+      "step": 341
+    },
+    {
+      "epoch": 1.6682926829268294,
+      "grad_norm": 2.7609124183654785,
+      "learning_rate": 4.666339449497033e-06,
+      "loss": 0.6284,
+      "step": 342
+    },
+    {
+      "epoch": 1.6731707317073172,
+      "grad_norm": 3.186077356338501,
+      "learning_rate": 4.664424691915084e-06,
+      "loss": 0.5751,
+      "step": 343
+    },
+    {
+      "epoch": 1.678048780487805,
+      "grad_norm": 3.644227981567383,
+      "learning_rate": 4.6625048511656675e-06,
+      "loss": 0.586,
+      "step": 344
+    },
+    {
+      "epoch": 1.6829268292682928,
+      "grad_norm": 3.196373462677002,
+      "learning_rate": 4.660579931757543e-06,
+      "loss": 0.5086,
+      "step": 345
+    },
+    {
+      "epoch": 1.6878048780487804,
+      "grad_norm": 2.7773900032043457,
+      "learning_rate": 4.6586499382113985e-06,
+      "loss": 0.5934,
+      "step": 346
+    },
+    {
+      "epoch": 1.6926829268292682,
+      "grad_norm": 2.3397631645202637,
+      "learning_rate": 4.6567148750598375e-06,
+      "loss": 0.7654,
+      "step": 347
+    },
+    {
+      "epoch": 1.697560975609756,
+      "grad_norm": 2.5567805767059326,
+      "learning_rate": 4.6547747468473705e-06,
+      "loss": 0.8908,
+      "step": 348
+    },
+    {
+      "epoch": 1.7024390243902439,
+      "grad_norm": 2.9218900203704834,
+      "learning_rate": 4.652829558130404e-06,
+      "loss": 0.4383,
+      "step": 349
+    },
+    {
+      "epoch": 1.7073170731707317,
+      "grad_norm": 2.962965250015259,
+      "learning_rate": 4.6508793134772265e-06,
+      "loss": 0.6031,
+      "step": 350
+    },
+    {
+      "epoch": 1.7121951219512195,
+      "grad_norm": 2.487739324569702,
+      "learning_rate": 4.648924017468003e-06,
+      "loss": 0.533,
+      "step": 351
+    },
+    {
+      "epoch": 1.7170731707317073,
+      "grad_norm": 2.769474506378174,
+      "learning_rate": 4.646963674694761e-06,
+      "loss": 0.8125,
+      "step": 352
+    },
+    {
+      "epoch": 1.721951219512195,
+      "grad_norm": 2.678243398666382,
+      "learning_rate": 4.64499828976138e-06,
+      "loss": 0.386,
+      "step": 353
+    },
+    {
+      "epoch": 1.726829268292683,
+      "grad_norm": 3.2764477729797363,
+      "learning_rate": 4.64302786728358e-06,
+      "loss": 0.4792,
+      "step": 354
+    },
+    {
+      "epoch": 1.7317073170731707,
+      "grad_norm": 2.6092708110809326,
+      "learning_rate": 4.641052411888913e-06,
+      "loss": 0.5031,
+      "step": 355
+    },
+    {
+      "epoch": 1.7365853658536585,
+      "grad_norm": 3.4002952575683594,
+      "learning_rate": 4.6390719282167515e-06,
+      "loss": 0.4726,
+      "step": 356
+    },
+    {
+      "epoch": 1.7414634146341463,
+      "grad_norm": 2.7558157444000244,
+      "learning_rate": 4.637086420918276e-06,
+      "loss": 0.7794,
+      "step": 357
+    },
+    {
+      "epoch": 1.7463414634146341,
+      "grad_norm": 2.239021062850952,
+      "learning_rate": 4.635095894656465e-06,
+      "loss": 0.6202,
+      "step": 358
+    },
+    {
+      "epoch": 1.751219512195122,
+      "grad_norm": 2.0502119064331055,
+      "learning_rate": 4.633100354106085e-06,
+      "loss": 0.3743,
+      "step": 359
+    },
+    {
+      "epoch": 1.7560975609756098,
+      "grad_norm": 2.842203140258789,
+      "learning_rate": 4.631099803953677e-06,
+      "loss": 0.8143,
+      "step": 360
+    },
+    {
+      "epoch": 1.7609756097560976,
+      "grad_norm": 2.8408772945404053,
+      "learning_rate": 4.629094248897546e-06,
+      "loss": 0.4986,
+      "step": 361
+    },
+    {
+      "epoch": 1.7658536585365854,
+      "grad_norm": 2.755530595779419,
+      "learning_rate": 4.627083693647757e-06,
+      "loss": 0.5833,
+      "step": 362
+    },
+    {
+      "epoch": 1.7707317073170732,
+      "grad_norm": 2.717116355895996,
+      "learning_rate": 4.625068142926111e-06,
+      "loss": 0.885,
+      "step": 363
+    },
+    {
+      "epoch": 1.775609756097561,
+      "grad_norm": 2.2784435749053955,
+      "learning_rate": 4.623047601466144e-06,
+      "loss": 0.7351,
+      "step": 364
+    },
+    {
+      "epoch": 1.7804878048780488,
+      "grad_norm": 2.3133914470672607,
+      "learning_rate": 4.621022074013114e-06,
+      "loss": 0.6426,
+      "step": 365
+    },
+    {
+      "epoch": 1.7853658536585366,
+      "grad_norm": 3.13562273979187,
+      "learning_rate": 4.618991565323987e-06,
+      "loss": 0.5588,
+      "step": 366
+    },
+    {
+      "epoch": 1.7902439024390244,
+      "grad_norm": 2.458186388015747,
+      "learning_rate": 4.616956080167426e-06,
+      "loss": 0.5424,
+      "step": 367
+    },
+    {
+      "epoch": 1.7951219512195122,
+      "grad_norm": 2.4780080318450928,
+      "learning_rate": 4.614915623323786e-06,
+      "loss": 0.8664,
+      "step": 368
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.623966932296753,
+      "learning_rate": 4.612870199585092e-06,
+      "loss": 0.4495,
+      "step": 369
+    },
+    {
+      "epoch": 1.8048780487804879,
+      "grad_norm": 2.7326242923736572,
+      "learning_rate": 4.610819813755038e-06,
+      "loss": 0.5099,
+      "step": 370
+    },
+    {
+      "epoch": 1.8097560975609757,
+      "grad_norm": 2.951014757156372,
+      "learning_rate": 4.608764470648971e-06,
+      "loss": 0.4322,
+      "step": 371
+    },
+    {
+      "epoch": 1.8146341463414632,
+      "grad_norm": 2.869870185852051,
+      "learning_rate": 4.606704175093879e-06,
+      "loss": 0.4744,
+      "step": 372
+    },
+    {
+      "epoch": 1.819512195121951,
+      "grad_norm": 2.686054229736328,
+      "learning_rate": 4.604638931928383e-06,
+      "loss": 0.797,
+      "step": 373
+    },
+    {
+      "epoch": 1.8243902439024389,
+      "grad_norm": 2.6421749591827393,
+      "learning_rate": 4.602568746002718e-06,
+      "loss": 0.4904,
+      "step": 374
+    },
+    {
+      "epoch": 1.8292682926829267,
+      "grad_norm": 2.949144124984741,
+      "learning_rate": 4.600493622178734e-06,
+      "loss": 0.8682,
+      "step": 375
+    },
+    {
+      "epoch": 1.8341463414634145,
+      "grad_norm": 2.554733991622925,
+      "learning_rate": 4.598413565329876e-06,
+      "loss": 0.5426,
+      "step": 376
+    },
+    {
+      "epoch": 1.8390243902439023,
+      "grad_norm": 2.3334367275238037,
+      "learning_rate": 4.596328580341169e-06,
+      "loss": 0.5628,
+      "step": 377
+    },
+    {
+      "epoch": 1.84390243902439,
+      "grad_norm": 2.577664613723755,
+      "learning_rate": 4.5942386721092195e-06,
+      "loss": 0.7073,
+      "step": 378
+    },
+    {
+      "epoch": 1.848780487804878,
+      "grad_norm": 3.1247141361236572,
+      "learning_rate": 4.592143845542189e-06,
+      "loss": 0.6526,
+      "step": 379
+    },
+    {
+      "epoch": 1.8536585365853657,
+      "grad_norm": 2.7015256881713867,
+      "learning_rate": 4.590044105559797e-06,
+      "loss": 0.8377,
+      "step": 380
+    },
+    {
+      "epoch": 1.8585365853658535,
+      "grad_norm": 2.573819398880005,
+      "learning_rate": 4.587939457093296e-06,
+      "loss": 0.5485,
+      "step": 381
+    },
+    {
+      "epoch": 1.8634146341463413,
+      "grad_norm": 2.8607687950134277,
+      "learning_rate": 4.585829905085468e-06,
+      "loss": 0.6065,
+      "step": 382
+    },
+    {
+      "epoch": 1.8682926829268292,
+      "grad_norm": 2.526625394821167,
+      "learning_rate": 4.5837154544906135e-06,
+      "loss": 0.7812,
+      "step": 383
+    },
+    {
+      "epoch": 1.873170731707317,
+      "grad_norm": 2.4161314964294434,
+      "learning_rate": 4.581596110274535e-06,
+      "loss": 0.7061,
+      "step": 384
+    },
+    {
+      "epoch": 1.8780487804878048,
+      "grad_norm": 2.34195876121521,
+      "learning_rate": 4.579471877414527e-06,
+      "loss": 0.9446,
+      "step": 385
+    },
+    {
+      "epoch": 1.8829268292682926,
+      "grad_norm": 3.7710156440734863,
+      "learning_rate": 4.577342760899368e-06,
+      "loss": 0.78,
+      "step": 386
+    },
+    {
+      "epoch": 1.8878048780487804,
+      "grad_norm": 2.5192313194274902,
+      "learning_rate": 4.575208765729302e-06,
+      "loss": 0.5205,
+      "step": 387
+    },
+    {
+      "epoch": 1.8926829268292682,
+      "grad_norm": 2.467484951019287,
+      "learning_rate": 4.573069896916035e-06,
+      "loss": 0.7827,
+      "step": 388
+    },
+    {
+      "epoch": 1.897560975609756,
+      "grad_norm": 2.640676259994507,
+      "learning_rate": 4.5709261594827125e-06,
+      "loss": 0.6512,
+      "step": 389
+    },
+    {
+      "epoch": 1.9024390243902438,
+      "grad_norm": 2.976623296737671,
+      "learning_rate": 4.568777558463922e-06,
+      "loss": 0.5548,
+      "step": 390
+    },
+    {
+      "epoch": 1.9073170731707316,
+      "grad_norm": 2.289722442626953,
+      "learning_rate": 4.566624098905665e-06,
+      "loss": 0.7038,
+      "step": 391
+    },
+    {
+      "epoch": 1.9121951219512194,
+      "grad_norm": 2.9512040615081787,
+      "learning_rate": 4.564465785865359e-06,
+      "loss": 0.5416,
+      "step": 392
+    },
+    {
+      "epoch": 1.9170731707317072,
+      "grad_norm": 2.394874095916748,
+      "learning_rate": 4.56230262441182e-06,
+      "loss": 0.4068,
+      "step": 393
+    },
+    {
+      "epoch": 1.921951219512195,
+      "grad_norm": 6.885486602783203,
+      "learning_rate": 4.560134619625247e-06,
+      "loss": 0.6197,
+      "step": 394
+    },
+    {
+      "epoch": 1.9268292682926829,
+      "grad_norm": 2.311272144317627,
+      "learning_rate": 4.5579617765972155e-06,
+      "loss": 0.5692,
+      "step": 395
+    },
+    {
+      "epoch": 1.9317073170731707,
+      "grad_norm": 2.4662933349609375,
+      "learning_rate": 4.555784100430662e-06,
+      "loss": 0.4836,
+      "step": 396
+    },
+    {
+      "epoch": 1.9365853658536585,
+      "grad_norm": 2.602741241455078,
+      "learning_rate": 4.553601596239877e-06,
+      "loss": 0.4594,
+      "step": 397
+    },
+    {
+      "epoch": 1.9414634146341463,
+      "grad_norm": 3.443909168243408,
+      "learning_rate": 4.551414269150489e-06,
+      "loss": 0.6053,
+      "step": 398
+    },
+    {
+      "epoch": 1.946341463414634,
+      "grad_norm": 2.5391502380371094,
+      "learning_rate": 4.54922212429945e-06,
+      "loss": 0.5133,
+      "step": 399
+    },
+    {
+      "epoch": 1.951219512195122,
+      "grad_norm": 2.7105700969696045,
+      "learning_rate": 4.547025166835027e-06,
+      "loss": 0.6984,
+      "step": 400
+    },
+    {
+      "epoch": 1.9560975609756097,
+      "grad_norm": 2.6098098754882812,
+      "learning_rate": 4.544823401916794e-06,
+      "loss": 0.7944,
+      "step": 401
+    },
+    {
+      "epoch": 1.9609756097560975,
+      "grad_norm": 2.7527425289154053,
+      "learning_rate": 4.542616834715612e-06,
+      "loss": 0.639,
+      "step": 402
+    },
+    {
+      "epoch": 1.9658536585365853,
+      "grad_norm": 2.760303258895874,
+      "learning_rate": 4.540405470413618e-06,
+      "loss": 0.4229,
+      "step": 403
+    },
+    {
+      "epoch": 1.9707317073170731,
+      "grad_norm": 2.4989006519317627,
+      "learning_rate": 4.53818931420422e-06,
+      "loss": 0.7482,
+      "step": 404
+    },
+    {
+      "epoch": 1.975609756097561,
+      "grad_norm": 2.3687169551849365,
+      "learning_rate": 4.535968371292076e-06,
+      "loss": 0.6146,
+      "step": 405
+    },
+    {
+      "epoch": 1.9804878048780488,
+      "grad_norm": 2.4285244941711426,
+      "learning_rate": 4.533742646893086e-06,
+      "loss": 0.6964,
+      "step": 406
+    },
+    {
+      "epoch": 1.9853658536585366,
+      "grad_norm": 2.337266206741333,
+      "learning_rate": 4.531512146234383e-06,
+      "loss": 0.6248,
+      "step": 407
+    },
+    {
+      "epoch": 1.9902439024390244,
+      "grad_norm": 2.704972743988037,
+      "learning_rate": 4.529276874554312e-06,
+      "loss": 0.8715,
+      "step": 408
+    },
+    {
+      "epoch": 1.9951219512195122,
+      "grad_norm": 2.2151944637298584,
+      "learning_rate": 4.527036837102426e-06,
+      "loss": 0.4945,
+      "step": 409
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.691330671310425,
+      "learning_rate": 4.524792039139471e-06,
+      "loss": 0.7085,
+      "step": 410
+    },
+    {
+      "epoch": 2.004878048780488,
+      "grad_norm": 2.9423086643218994,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.3178,
+      "step": 411
+    },
+    {
+      "epoch": 2.0097560975609756,
+      "grad_norm": 2.860677719116211,
+      "learning_rate": 4.520288182779214e-06,
+      "loss": 0.5092,
+      "step": 412
+    },
+    {
+      "epoch": 2.0146341463414634,
+      "grad_norm": 2.7503843307495117,
+      "learning_rate": 4.518029134959253e-06,
+      "loss": 0.314,
+      "step": 413
+    },
+    {
+      "epoch": 2.0195121951219512,
+      "grad_norm": 4.541809558868408,
+      "learning_rate": 4.515765347782878e-06,
+      "loss": 0.5287,
+      "step": 414
+    },
+    {
+      "epoch": 2.024390243902439,
+      "grad_norm": 9.126826286315918,
+      "learning_rate": 4.5134968265666085e-06,
+      "loss": 0.8221,
+      "step": 415
+    },
+    {
+      "epoch": 2.029268292682927,
+      "grad_norm": 4.4358229637146,
+      "learning_rate": 4.511223576638084e-06,
+      "loss": 0.5402,
+      "step": 416
+    },
+    {
+      "epoch": 2.0341463414634147,
+      "grad_norm": 3.1090731620788574,
+      "learning_rate": 4.508945603336049e-06,
+      "loss": 0.617,
+      "step": 417
+    },
+    {
+      "epoch": 2.0390243902439025,
+      "grad_norm": 2.6933369636535645,
+      "learning_rate": 4.50666291201034e-06,
+      "loss": 0.3541,
+      "step": 418
+    },
+    {
+      "epoch": 2.0439024390243903,
+      "grad_norm": 5.898099899291992,
+      "learning_rate": 4.504375508021876e-06,
+      "loss": 0.4842,
+      "step": 419
+    },
+    {
+      "epoch": 2.048780487804878,
+      "grad_norm": 2.950939178466797,
+      "learning_rate": 4.50208339674264e-06,
+      "loss": 0.6168,
+      "step": 420
+    },
+    {
+      "epoch": 2.053658536585366,
+      "grad_norm": 3.2513322830200195,
+      "learning_rate": 4.499786583555675e-06,
+      "loss": 0.6425,
+      "step": 421
+    },
+    {
+      "epoch": 2.0585365853658537,
+      "grad_norm": 2.911562442779541,
+      "learning_rate": 4.497485073855061e-06,
+      "loss": 0.364,
+      "step": 422
+    },
+    {
+      "epoch": 2.0634146341463415,
+      "grad_norm": 4.2179274559021,
+      "learning_rate": 4.495178873045913e-06,
+      "loss": 0.3687,
+      "step": 423
+    },
+    {
+      "epoch": 2.0682926829268293,
+      "grad_norm": 3.2010395526885986,
+      "learning_rate": 4.4928679865443605e-06,
+      "loss": 0.4068,
+      "step": 424
+    },
+    {
+      "epoch": 2.073170731707317,
+      "grad_norm": 3.2425589561462402,
+      "learning_rate": 4.4905524197775366e-06,
+      "loss": 0.4759,
+      "step": 425
+    },
+    {
+      "epoch": 2.078048780487805,
+      "grad_norm": 2.9252519607543945,
+      "learning_rate": 4.4882321781835666e-06,
+      "loss": 0.4197,
+      "step": 426
+    },
+    {
+      "epoch": 2.0829268292682928,
+      "grad_norm": 2.7859911918640137,
+      "learning_rate": 4.4859072672115565e-06,
+      "loss": 0.2294,
+      "step": 427
+    },
+    {
+      "epoch": 2.0878048780487806,
+      "grad_norm": 3.138796091079712,
+      "learning_rate": 4.483577692321577e-06,
+      "loss": 0.7572,
+      "step": 428
+    },
+    {
+      "epoch": 2.0926829268292684,
+      "grad_norm": 3.1447339057922363,
+      "learning_rate": 4.481243458984651e-06,
+      "loss": 0.4035,
+      "step": 429
+    },
+    {
+      "epoch": 2.097560975609756,
+      "grad_norm": 3.1876862049102783,
+      "learning_rate": 4.478904572682743e-06,
+      "loss": 0.5776,
+      "step": 430
+    },
+    {
+      "epoch": 2.102439024390244,
+      "grad_norm": 2.934257745742798,
+      "learning_rate": 4.476561038908745e-06,
+      "loss": 0.4005,
+      "step": 431
+    },
+    {
+      "epoch": 2.107317073170732,
+      "grad_norm": 2.904954433441162,
+      "learning_rate": 4.474212863166464e-06,
+      "loss": 0.5689,
+      "step": 432
+    },
+    {
+      "epoch": 2.1121951219512196,
+      "grad_norm": 3.6023731231689453,
+      "learning_rate": 4.471860050970608e-06,
+      "loss": 0.5068,
+      "step": 433
+    },
+    {
+      "epoch": 2.1170731707317074,
+      "grad_norm": 4.073422431945801,
+      "learning_rate": 4.469502607846774e-06,
+      "loss": 0.8349,
+      "step": 434
+    },
+    {
+      "epoch": 2.1219512195121952,
+      "grad_norm": 2.813789129257202,
+      "learning_rate": 4.467140539331434e-06,
+      "loss": 0.3641,
+      "step": 435
+    },
+    {
+      "epoch": 2.126829268292683,
+      "grad_norm": 3.874516248703003,
+      "learning_rate": 4.464773850971924e-06,
+      "loss": 0.222,
+      "step": 436
+    },
+    {
+      "epoch": 2.131707317073171,
+      "grad_norm": 3.1221084594726562,
+      "learning_rate": 4.46240254832643e-06,
+      "loss": 0.3799,
+      "step": 437
+    },
+    {
+      "epoch": 2.1365853658536587,
+      "grad_norm": 3.298933267593384,
+      "learning_rate": 4.460026636963971e-06,
+      "loss": 0.4759,
+      "step": 438
+    },
+    {
+      "epoch": 2.1414634146341465,
+      "grad_norm": 2.456233024597168,
+      "learning_rate": 4.4576461224643965e-06,
+      "loss": 0.384,
+      "step": 439
+    },
+    {
+      "epoch": 2.1463414634146343,
+      "grad_norm": 2.8427460193634033,
+      "learning_rate": 4.455261010418359e-06,
+      "loss": 0.391,
+      "step": 440
+    },
+    {
+      "epoch": 2.151219512195122,
+      "grad_norm": 3.0267624855041504,
+      "learning_rate": 4.452871306427314e-06,
+      "loss": 0.6177,
+      "step": 441
+    },
+    {
+      "epoch": 2.15609756097561,
+      "grad_norm": 3.437302827835083,
+      "learning_rate": 4.450477016103498e-06,
+      "loss": 0.5143,
+      "step": 442
+    },
+    {
+      "epoch": 2.1609756097560977,
+      "grad_norm": 3.152210235595703,
+      "learning_rate": 4.4480781450699205e-06,
+      "loss": 0.3783,
+      "step": 443
+    },
+    {
+      "epoch": 2.1658536585365855,
+      "grad_norm": 3.507753372192383,
+      "learning_rate": 4.4456746989603464e-06,
+      "loss": 0.3574,
+      "step": 444
+    },
+    {
+      "epoch": 2.1707317073170733,
+      "grad_norm": 2.8855366706848145,
+      "learning_rate": 4.443266683419289e-06,
+      "loss": 0.5088,
+      "step": 445
+    },
+    {
+      "epoch": 2.175609756097561,
+      "grad_norm": 2.7776072025299072,
+      "learning_rate": 4.440854104101988e-06,
+      "loss": 0.3773,
+      "step": 446
+    },
+    {
+      "epoch": 2.180487804878049,
+      "grad_norm": 3.019484281539917,
+      "learning_rate": 4.438436966674406e-06,
+      "loss": 0.5002,
+      "step": 447
+    },
+    {
+      "epoch": 2.1853658536585368,
+      "grad_norm": 3.6962451934814453,
+      "learning_rate": 4.436015276813208e-06,
+      "loss": 0.4601,
+      "step": 448
+    },
+    {
+      "epoch": 2.1902439024390246,
+      "grad_norm": 3.1288888454437256,
+      "learning_rate": 4.4335890402057505e-06,
+      "loss": 0.5422,
+      "step": 449
+    },
+    {
+      "epoch": 2.1951219512195124,
+      "grad_norm": 3.7083234786987305,
+      "learning_rate": 4.431158262550067e-06,
+      "loss": 0.4684,
+      "step": 450
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.1714789867401123,
+      "learning_rate": 4.428722949554858e-06,
+      "loss": 0.2528,
+      "step": 451
+    },
+    {
+      "epoch": 2.204878048780488,
+      "grad_norm": 3.0773637294769287,
+      "learning_rate": 4.426283106939474e-06,
+      "loss": 0.4061,
+      "step": 452
+    },
+    {
+      "epoch": 2.209756097560976,
+      "grad_norm": 2.604093551635742,
+      "learning_rate": 4.423838740433903e-06,
+      "loss": 0.4779,
+      "step": 453
+    },
+    {
+      "epoch": 2.2146341463414636,
+      "grad_norm": 2.9293880462646484,
+      "learning_rate": 4.4213898557787586e-06,
+      "loss": 0.233,
+      "step": 454
+    },
+    {
+      "epoch": 2.2195121951219514,
+      "grad_norm": 2.9195125102996826,
+      "learning_rate": 4.4189364587252636e-06,
+      "loss": 0.7756,
+      "step": 455
+    },
+    {
+      "epoch": 2.2243902439024392,
+      "grad_norm": 3.2263920307159424,
+      "learning_rate": 4.416478555035241e-06,
+      "loss": 0.2806,
+      "step": 456
+    },
+    {
+      "epoch": 2.229268292682927,
+      "grad_norm": 2.8109211921691895,
+      "learning_rate": 4.4140161504810935e-06,
+      "loss": 0.3923,
+      "step": 457
+    },
+    {
+      "epoch": 2.234146341463415,
+      "grad_norm": 2.645853281021118,
+      "learning_rate": 4.4115492508457986e-06,
+      "loss": 0.289,
+      "step": 458
+    },
+    {
+      "epoch": 2.2390243902439027,
+      "grad_norm": 3.3712451457977295,
+      "learning_rate": 4.409077861922887e-06,
+      "loss": 0.5053,
+      "step": 459
+    },
+    {
+      "epoch": 2.2439024390243905,
+      "grad_norm": 2.6892387866973877,
+      "learning_rate": 4.406601989516435e-06,
+      "loss": 0.3363,
+      "step": 460
+    },
+    {
+      "epoch": 2.2487804878048783,
+      "grad_norm": 2.3195693492889404,
+      "learning_rate": 4.404121639441047e-06,
+      "loss": 0.2367,
+      "step": 461
+    },
+    {
+      "epoch": 2.253658536585366,
+      "grad_norm": 3.0115339756011963,
+      "learning_rate": 4.401636817521843e-06,
+      "loss": 0.4942,
+      "step": 462
+    },
+    {
+      "epoch": 2.258536585365854,
+      "grad_norm": 2.9528865814208984,
+      "learning_rate": 4.399147529594447e-06,
+      "loss": 0.3328,
+      "step": 463
+    },
+    {
+      "epoch": 2.2634146341463417,
+      "grad_norm": 3.110799551010132,
+      "learning_rate": 4.3966537815049686e-06,
+      "loss": 0.3917,
+      "step": 464
+    },
+    {
+      "epoch": 2.2682926829268295,
+      "grad_norm": 3.2973792552948,
+      "learning_rate": 4.394155579109994e-06,
+      "loss": 0.5203,
+      "step": 465
+    },
+    {
+      "epoch": 2.2731707317073173,
+      "grad_norm": 4.7184038162231445,
+      "learning_rate": 4.391652928276572e-06,
+      "loss": 0.729,
+      "step": 466
+    },
+    {
+      "epoch": 2.278048780487805,
+      "grad_norm": 3.1992053985595703,
+      "learning_rate": 4.389145834882195e-06,
+      "loss": 0.4822,
+      "step": 467
+    },
+    {
+      "epoch": 2.2829268292682925,
+      "grad_norm": 4.320055961608887,
+      "learning_rate": 4.386634304814789e-06,
+      "loss": 0.3962,
+      "step": 468
+    },
+    {
+      "epoch": 2.2878048780487803,
+      "grad_norm": 3.704524517059326,
+      "learning_rate": 4.384118343972704e-06,
+      "loss": 0.5996,
+      "step": 469
+    },
+    {
+      "epoch": 2.292682926829268,
+      "grad_norm": 2.8172974586486816,
+      "learning_rate": 4.381597958264692e-06,
+      "loss": 0.6328,
+      "step": 470
+    },
+    {
+      "epoch": 2.297560975609756,
+      "grad_norm": 2.7418763637542725,
+      "learning_rate": 4.379073153609896e-06,
+      "loss": 0.6254,
+      "step": 471
+    },
+    {
+      "epoch": 2.3024390243902437,
+      "grad_norm": 5.364504337310791,
+      "learning_rate": 4.37654393593784e-06,
+      "loss": 0.6793,
+      "step": 472
+    },
+    {
+      "epoch": 2.3073170731707315,
+      "grad_norm": 2.935291290283203,
+      "learning_rate": 4.3740103111884096e-06,
+      "loss": 0.4161,
+      "step": 473
+    },
+    {
+      "epoch": 2.3121951219512193,
+      "grad_norm": 3.085155963897705,
+      "learning_rate": 4.371472285311842e-06,
+      "loss": 0.3329,
+      "step": 474
+    },
+    {
+      "epoch": 2.317073170731707,
+      "grad_norm": 2.2218778133392334,
+      "learning_rate": 4.368929864268709e-06,
+      "loss": 0.2687,
+      "step": 475
+    },
+    {
+      "epoch": 2.321951219512195,
+      "grad_norm": 3.3985276222229004,
+      "learning_rate": 4.366383054029907e-06,
+      "loss": 0.5934,
+      "step": 476
+    },
+    {
+      "epoch": 2.3268292682926828,
+      "grad_norm": 3.0726048946380615,
+      "learning_rate": 4.363831860576638e-06,
+      "loss": 0.5033,
+      "step": 477
+    },
+    {
+      "epoch": 2.3317073170731706,
+      "grad_norm": 2.728628635406494,
+      "learning_rate": 4.361276289900396e-06,
+      "loss": 0.4492,
+      "step": 478
+    },
+    {
+      "epoch": 2.3365853658536584,
+      "grad_norm": 3.1294424533843994,
+      "learning_rate": 4.358716348002962e-06,
+      "loss": 0.619,
+      "step": 479
+    },
+    {
+      "epoch": 2.341463414634146,
+      "grad_norm": 3.5564961433410645,
+      "learning_rate": 4.356152040896376e-06,
+      "loss": 0.4018,
+      "step": 480
+    },
+    {
+      "epoch": 2.346341463414634,
+      "grad_norm": 2.9329910278320312,
+      "learning_rate": 4.3535833746029335e-06,
+      "loss": 0.3062,
+      "step": 481
+    },
+    {
+      "epoch": 2.351219512195122,
+      "grad_norm": 3.744480848312378,
+      "learning_rate": 4.351010355155165e-06,
+      "loss": 0.3387,
+      "step": 482
+    },
+    {
+      "epoch": 2.3560975609756096,
+      "grad_norm": 2.537912130355835,
+      "learning_rate": 4.348432988595828e-06,
+      "loss": 0.3103,
+      "step": 483
+    },
+    {
+      "epoch": 2.3609756097560974,
+      "grad_norm": 3.232128858566284,
+      "learning_rate": 4.345851280977885e-06,
+      "loss": 0.6782,
+      "step": 484
+    },
+    {
+      "epoch": 2.3658536585365852,
+      "grad_norm": 3.601463794708252,
+      "learning_rate": 4.343265238364496e-06,
+      "loss": 0.3195,
+      "step": 485
+    },
+    {
+      "epoch": 2.370731707317073,
+      "grad_norm": 4.05529260635376,
+      "learning_rate": 4.340674866829001e-06,
+      "loss": 0.4639,
+      "step": 486
+    },
+    {
+      "epoch": 2.375609756097561,
+      "grad_norm": 4.128161430358887,
+      "learning_rate": 4.338080172454908e-06,
+      "loss": 0.7229,
+      "step": 487
+    },
+    {
+      "epoch": 2.3804878048780487,
+      "grad_norm": 2.665430784225464,
+      "learning_rate": 4.335481161335875e-06,
+      "loss": 0.4334,
+      "step": 488
+    },
+    {
+      "epoch": 2.3853658536585365,
+      "grad_norm": 3.777899742126465,
+      "learning_rate": 4.332877839575699e-06,
+      "loss": 0.3409,
+      "step": 489
+    },
+    {
+      "epoch": 2.3902439024390243,
+      "grad_norm": 2.9942116737365723,
+      "learning_rate": 4.330270213288301e-06,
+      "loss": 0.5221,
+      "step": 490
+    },
+    {
+      "epoch": 2.395121951219512,
+      "grad_norm": 3.518601417541504,
+      "learning_rate": 4.32765828859771e-06,
+      "loss": 0.7078,
+      "step": 491
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.452350378036499,
+      "learning_rate": 4.325042071638051e-06,
+      "loss": 0.5902,
+      "step": 492
+    },
+    {
+      "epoch": 2.4048780487804877,
+      "grad_norm": 3.072655200958252,
+      "learning_rate": 4.322421568553529e-06,
+      "loss": 0.3746,
+      "step": 493
+    },
+    {
+      "epoch": 2.4097560975609755,
+      "grad_norm": 2.8621394634246826,
+      "learning_rate": 4.319796785498416e-06,
+      "loss": 0.3474,
+      "step": 494
+    },
+    {
+      "epoch": 2.4146341463414633,
+      "grad_norm": 3.3891537189483643,
+      "learning_rate": 4.317167728637032e-06,
+      "loss": 0.5171,
+      "step": 495
+    },
+    {
+      "epoch": 2.419512195121951,
+      "grad_norm": 2.505720376968384,
+      "learning_rate": 4.314534404143738e-06,
+      "loss": 0.4263,
+      "step": 496
+    },
+    {
+      "epoch": 2.424390243902439,
+      "grad_norm": 2.6280455589294434,
+      "learning_rate": 4.3118968182029155e-06,
+      "loss": 0.5072,
+      "step": 497
+    },
+    {
+      "epoch": 2.4292682926829268,
+      "grad_norm": 2.703711748123169,
+      "learning_rate": 4.3092549770089566e-06,
+      "loss": 0.2742,
+      "step": 498
+    },
+    {
+      "epoch": 2.4341463414634146,
+      "grad_norm": 3.0358169078826904,
+      "learning_rate": 4.306608886766243e-06,
+      "loss": 0.4814,
+      "step": 499
+    },
+    {
+      "epoch": 2.4390243902439024,
+      "grad_norm": 3.263326406478882,
+      "learning_rate": 4.303958553689137e-06,
+      "loss": 0.4188,
+      "step": 500
+    },
+    {
+      "epoch": 2.44390243902439,
+      "grad_norm": 2.833951950073242,
+      "learning_rate": 4.3013039840019675e-06,
+      "loss": 0.6436,
+      "step": 501
+    },
+    {
+      "epoch": 2.448780487804878,
+      "grad_norm": 3.6790921688079834,
+      "learning_rate": 4.2986451839390105e-06,
+      "loss": 0.2862,
+      "step": 502
+    },
+    {
+      "epoch": 2.453658536585366,
+      "grad_norm": 2.7376418113708496,
+      "learning_rate": 4.295982159744476e-06,
+      "loss": 0.4926,
+      "step": 503
+    },
+    {
+      "epoch": 2.4585365853658536,
+      "grad_norm": 3.575244665145874,
+      "learning_rate": 4.293314917672498e-06,
+      "loss": 0.5717,
+      "step": 504
+    },
+    {
+      "epoch": 2.4634146341463414,
+      "grad_norm": 2.8722269535064697,
+      "learning_rate": 4.290643463987114e-06,
+      "loss": 0.2707,
+      "step": 505
+    },
+    {
+      "epoch": 2.4682926829268292,
+      "grad_norm": 2.8118090629577637,
+      "learning_rate": 4.287967804962252e-06,
+      "loss": 0.347,
+      "step": 506
+    },
+    {
+      "epoch": 2.473170731707317,
+      "grad_norm": 3.345698356628418,
+      "learning_rate": 4.285287946881718e-06,
+      "loss": 0.2103,
+      "step": 507
+    },
+    {
+      "epoch": 2.478048780487805,
+      "grad_norm": 3.0156590938568115,
+      "learning_rate": 4.282603896039178e-06,
+      "loss": 0.6405,
+      "step": 508
+    },
+    {
+      "epoch": 2.4829268292682927,
+      "grad_norm": 3.102205753326416,
+      "learning_rate": 4.279915658738145e-06,
+      "loss": 0.4027,
+      "step": 509
+    },
+    {
+      "epoch": 2.4878048780487805,
+      "grad_norm": 2.8665261268615723,
+      "learning_rate": 4.277223241291966e-06,
+      "loss": 0.6503,
+      "step": 510
+    },
+    {
+      "epoch": 2.4926829268292683,
+      "grad_norm": 2.5396728515625,
+      "learning_rate": 4.274526650023801e-06,
+      "loss": 0.5006,
+      "step": 511
+    },
+    {
+      "epoch": 2.497560975609756,
+      "grad_norm": 3.4846577644348145,
+      "learning_rate": 4.271825891266617e-06,
+      "loss": 0.479,
+      "step": 512
+    },
+    {
+      "epoch": 2.502439024390244,
+      "grad_norm": 4.5995612144470215,
+      "learning_rate": 4.269120971363164e-06,
+      "loss": 0.6667,
+      "step": 513
+    },
+    {
+      "epoch": 2.5073170731707317,
+      "grad_norm": 3.2117559909820557,
+      "learning_rate": 4.266411896665967e-06,
+      "loss": 0.2977,
+      "step": 514
+    },
+    {
+      "epoch": 2.5121951219512195,
+      "grad_norm": 2.798161268234253,
+      "learning_rate": 4.263698673537309e-06,
+      "loss": 0.3912,
+      "step": 515
+    },
+    {
+      "epoch": 2.5170731707317073,
+      "grad_norm": 3.593287944793701,
+      "learning_rate": 4.260981308349214e-06,
+      "loss": 0.615,
+      "step": 516
+    },
+    {
+      "epoch": 2.521951219512195,
+      "grad_norm": 3.06075119972229,
+      "learning_rate": 4.258259807483434e-06,
+      "loss": 0.4559,
+      "step": 517
+    },
+    {
+      "epoch": 2.526829268292683,
+      "grad_norm": 2.893202543258667,
+      "learning_rate": 4.255534177331435e-06,
+      "loss": 0.4993,
+      "step": 518
+    },
+    {
+      "epoch": 2.5317073170731708,
+      "grad_norm": 3.613308906555176,
+      "learning_rate": 4.252804424294378e-06,
+      "loss": 0.4581,
+      "step": 519
+    },
+    {
+      "epoch": 2.5365853658536586,
+      "grad_norm": 3.1191842555999756,
+      "learning_rate": 4.25007055478311e-06,
+      "loss": 0.5403,
+      "step": 520
+    },
+    {
+      "epoch": 2.5414634146341464,
+      "grad_norm": 3.653355836868286,
+      "learning_rate": 4.247332575218144e-06,
+      "loss": 0.3658,
+      "step": 521
+    },
+    {
+      "epoch": 2.546341463414634,
+      "grad_norm": 3.1386306285858154,
+      "learning_rate": 4.244590492029643e-06,
+      "loss": 0.6342,
+      "step": 522
+    },
+    {
+      "epoch": 2.551219512195122,
+      "grad_norm": 3.0894742012023926,
+      "learning_rate": 4.241844311657411e-06,
+      "loss": 0.3411,
+      "step": 523
+    },
+    {
+      "epoch": 2.55609756097561,
+      "grad_norm": 3.205916404724121,
+      "learning_rate": 4.239094040550875e-06,
+      "loss": 0.2829,
+      "step": 524
+    },
+    {
+      "epoch": 2.5609756097560976,
+      "grad_norm": 2.378857374191284,
+      "learning_rate": 4.236339685169065e-06,
+      "loss": 0.4749,
+      "step": 525
+    },
+    {
+      "epoch": 2.5658536585365854,
+      "grad_norm": 3.8657875061035156,
+      "learning_rate": 4.233581251980604e-06,
+      "loss": 0.2485,
+      "step": 526
+    },
+    {
+      "epoch": 2.5707317073170732,
+      "grad_norm": 3.565807580947876,
+      "learning_rate": 4.230818747463696e-06,
+      "loss": 0.4488,
+      "step": 527
+    },
+    {
+      "epoch": 2.575609756097561,
+      "grad_norm": 2.6909685134887695,
+      "learning_rate": 4.228052178106101e-06,
+      "loss": 0.4495,
+      "step": 528
+    },
+    {
+      "epoch": 2.580487804878049,
+      "grad_norm": 2.937680244445801,
+      "learning_rate": 4.2252815504051285e-06,
+      "loss": 0.2396,
+      "step": 529
+    },
+    {
+      "epoch": 2.5853658536585367,
+      "grad_norm": 5.55731201171875,
+      "learning_rate": 4.222506870867618e-06,
+      "loss": 0.6784,
+      "step": 530
+    },
+    {
+      "epoch": 2.5902439024390245,
+      "grad_norm": 2.7388782501220703,
+      "learning_rate": 4.2197281460099245e-06,
+      "loss": 0.5543,
+      "step": 531
+    },
+    {
+      "epoch": 2.5951219512195123,
+      "grad_norm": 3.311134099960327,
+      "learning_rate": 4.216945382357905e-06,
+      "loss": 0.5281,
+      "step": 532
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.511232376098633,
+      "learning_rate": 4.214158586446901e-06,
+      "loss": 0.8019,
+      "step": 533
+    },
+    {
+      "epoch": 2.604878048780488,
+      "grad_norm": 4.416641712188721,
+      "learning_rate": 4.211367764821722e-06,
+      "loss": 0.7769,
+      "step": 534
+    },
+    {
+      "epoch": 2.6097560975609757,
+      "grad_norm": 2.9849908351898193,
+      "learning_rate": 4.208572924036634e-06,
+      "loss": 0.4077,
+      "step": 535
+    },
+    {
+      "epoch": 2.6146341463414635,
+      "grad_norm": 2.8512160778045654,
+      "learning_rate": 4.2057740706553415e-06,
+      "loss": 0.433,
+      "step": 536
+    },
+    {
+      "epoch": 2.6195121951219513,
+      "grad_norm": 2.6729629039764404,
+      "learning_rate": 4.202971211250971e-06,
+      "loss": 0.5957,
+      "step": 537
+    },
+    {
+      "epoch": 2.624390243902439,
+      "grad_norm": 2.4570281505584717,
+      "learning_rate": 4.200164352406061e-06,
+      "loss": 0.3013,
+      "step": 538
+    },
+    {
+      "epoch": 2.629268292682927,
+      "grad_norm": 3.3771679401397705,
+      "learning_rate": 4.197353500712539e-06,
+      "loss": 0.5646,
+      "step": 539
+    },
+    {
+      "epoch": 2.6341463414634148,
+      "grad_norm": 3.163496494293213,
+      "learning_rate": 4.1945386627717115e-06,
+      "loss": 0.4529,
+      "step": 540
+    },
+    {
+      "epoch": 2.6390243902439026,
+      "grad_norm": 8.32056713104248,
+      "learning_rate": 4.191719845194246e-06,
+      "loss": 0.6076,
+      "step": 541
+    },
+    {
+      "epoch": 2.6439024390243904,
+      "grad_norm": 2.7657363414764404,
+      "learning_rate": 4.188897054600156e-06,
+      "loss": 0.4855,
+      "step": 542
+    },
+    {
+      "epoch": 2.648780487804878,
+      "grad_norm": 3.299283504486084,
+      "learning_rate": 4.186070297618787e-06,
+      "loss": 0.5836,
+      "step": 543
+    },
+    {
+      "epoch": 2.653658536585366,
+      "grad_norm": 2.3928205966949463,
+      "learning_rate": 4.183239580888799e-06,
+      "loss": 0.6266,
+      "step": 544
+    },
+    {
+      "epoch": 2.658536585365854,
+      "grad_norm": 3.395251750946045,
+      "learning_rate": 4.18040491105815e-06,
+      "loss": 0.429,
+      "step": 545
+    },
+    {
+      "epoch": 2.6634146341463416,
+      "grad_norm": 2.690936803817749,
+      "learning_rate": 4.177566294784085e-06,
+      "loss": 0.391,
+      "step": 546
+    },
+    {
+      "epoch": 2.6682926829268294,
+      "grad_norm": 3.7687628269195557,
+      "learning_rate": 4.174723738733114e-06,
+      "loss": 0.6548,
+      "step": 547
+    },
+    {
+      "epoch": 2.6731707317073172,
+      "grad_norm": 2.7884976863861084,
+      "learning_rate": 4.171877249581001e-06,
+      "loss": 0.5188,
+      "step": 548
+    },
+    {
+      "epoch": 2.678048780487805,
+      "grad_norm": 3.0811641216278076,
+      "learning_rate": 4.169026834012748e-06,
+      "loss": 0.3494,
+      "step": 549
+    },
+    {
+      "epoch": 2.682926829268293,
+      "grad_norm": 3.090078592300415,
+      "learning_rate": 4.166172498722577e-06,
+      "loss": 0.3621,
+      "step": 550
+    },
+    {
+      "epoch": 2.68780487804878,
+      "grad_norm": 3.925424098968506,
+      "learning_rate": 4.163314250413913e-06,
+      "loss": 0.7187,
+      "step": 551
+    },
+    {
+      "epoch": 2.692682926829268,
+      "grad_norm": 3.3590312004089355,
+      "learning_rate": 4.160452095799378e-06,
+      "loss": 0.428,
+      "step": 552
+    },
+    {
+      "epoch": 2.697560975609756,
+      "grad_norm": 3.08093523979187,
+      "learning_rate": 4.157586041600759e-06,
+      "loss": 0.202,
+      "step": 553
+    },
+    {
+      "epoch": 2.7024390243902436,
+      "grad_norm": 2.9391448497772217,
+      "learning_rate": 4.154716094549008e-06,
+      "loss": 0.5238,
+      "step": 554
+    },
+    {
+      "epoch": 2.7073170731707314,
+      "grad_norm": 2.9869461059570312,
+      "learning_rate": 4.151842261384217e-06,
+      "loss": 0.3073,
+      "step": 555
+    },
+    {
+      "epoch": 2.7121951219512193,
+      "grad_norm": 3.8973608016967773,
+      "learning_rate": 4.148964548855603e-06,
+      "loss": 0.8435,
+      "step": 556
+    },
+    {
+      "epoch": 2.717073170731707,
+      "grad_norm": 2.3596479892730713,
+      "learning_rate": 4.146082963721496e-06,
+      "loss": 0.2562,
+      "step": 557
+    },
+    {
+      "epoch": 2.721951219512195,
+      "grad_norm": 3.4964873790740967,
+      "learning_rate": 4.143197512749322e-06,
+      "loss": 1.0144,
+      "step": 558
+    },
+    {
+      "epoch": 2.7268292682926827,
+      "grad_norm": 2.8925280570983887,
+      "learning_rate": 4.140308202715581e-06,
+      "loss": 0.7581,
+      "step": 559
+    },
+    {
+      "epoch": 2.7317073170731705,
+      "grad_norm": 2.622724771499634,
+      "learning_rate": 4.13741504040584e-06,
+      "loss": 0.3114,
+      "step": 560
+    },
+    {
+      "epoch": 2.7365853658536583,
+      "grad_norm": 3.775834321975708,
+      "learning_rate": 4.134518032614713e-06,
+      "loss": 0.4384,
+      "step": 561
+    },
+    {
+      "epoch": 2.741463414634146,
+      "grad_norm": 2.691236972808838,
+      "learning_rate": 4.1316171861458445e-06,
+      "loss": 0.3141,
+      "step": 562
+    },
+    {
+      "epoch": 2.746341463414634,
+      "grad_norm": 3.059152841567993,
+      "learning_rate": 4.128712507811893e-06,
+      "loss": 0.5777,
+      "step": 563
+    },
+    {
+      "epoch": 2.7512195121951217,
+      "grad_norm": 2.867432117462158,
+      "learning_rate": 4.125804004434517e-06,
+      "loss": 0.5542,
+      "step": 564
+    },
+    {
+      "epoch": 2.7560975609756095,
+      "grad_norm": 2.796438694000244,
+      "learning_rate": 4.12289168284436e-06,
+      "loss": 0.3442,
+      "step": 565
+    },
+    {
+      "epoch": 2.7609756097560973,
+      "grad_norm": 3.052199125289917,
+      "learning_rate": 4.119975549881029e-06,
+      "loss": 0.4754,
+      "step": 566
+    },
+    {
+      "epoch": 2.765853658536585,
+      "grad_norm": 2.5463602542877197,
+      "learning_rate": 4.1170556123930846e-06,
+      "loss": 0.2988,
+      "step": 567
+    },
+    {
+      "epoch": 2.770731707317073,
+      "grad_norm": 3.003124475479126,
+      "learning_rate": 4.114131877238021e-06,
+      "loss": 0.4642,
+      "step": 568
+    },
+    {
+      "epoch": 2.7756097560975608,
+      "grad_norm": 2.4988298416137695,
+      "learning_rate": 4.111204351282254e-06,
+      "loss": 0.3493,
+      "step": 569
+    },
+    {
+      "epoch": 2.7804878048780486,
+      "grad_norm": 2.7403693199157715,
+      "learning_rate": 4.108273041401098e-06,
+      "loss": 0.4007,
+      "step": 570
+    },
+    {
+      "epoch": 2.7853658536585364,
+      "grad_norm": 4.101940155029297,
+      "learning_rate": 4.105337954478756e-06,
+      "loss": 0.7815,
+      "step": 571
+    },
+    {
+      "epoch": 2.790243902439024,
+      "grad_norm": 3.229969024658203,
+      "learning_rate": 4.102399097408304e-06,
+      "loss": 0.6099,
+      "step": 572
+    },
+    {
+      "epoch": 2.795121951219512,
+      "grad_norm": 3.234693765640259,
+      "learning_rate": 4.099456477091667e-06,
+      "loss": 0.2478,
+      "step": 573
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.9824702739715576,
+      "learning_rate": 4.096510100439611e-06,
+      "loss": 0.6403,
+      "step": 574
+    },
+    {
+      "epoch": 2.8048780487804876,
+      "grad_norm": 2.8012478351593018,
+      "learning_rate": 4.093559974371725e-06,
+      "loss": 0.2509,
+      "step": 575
+    },
+    {
+      "epoch": 2.8097560975609754,
+      "grad_norm": 2.915400743484497,
+      "learning_rate": 4.0906061058164e-06,
+      "loss": 0.7552,
+      "step": 576
+    },
+    {
+      "epoch": 2.8146341463414632,
+      "grad_norm": 3.467665672302246,
+      "learning_rate": 4.087648501710819e-06,
+      "loss": 0.3146,
+      "step": 577
+    },
+    {
+      "epoch": 2.819512195121951,
+      "grad_norm": 3.1628401279449463,
+      "learning_rate": 4.084687169000938e-06,
+      "loss": 0.507,
+      "step": 578
+    },
+    {
+      "epoch": 2.824390243902439,
+      "grad_norm": 2.4069066047668457,
+      "learning_rate": 4.081722114641469e-06,
+      "loss": 0.4116,
+      "step": 579
+    },
+    {
+      "epoch": 2.8292682926829267,
+      "grad_norm": 3.698174238204956,
+      "learning_rate": 4.0787533455958626e-06,
+      "loss": 0.2264,
+      "step": 580
+    },
+    {
+      "epoch": 2.8341463414634145,
+      "grad_norm": 3.0896191596984863,
+      "learning_rate": 4.075780868836296e-06,
+      "loss": 0.3197,
+      "step": 581
+    },
+    {
+      "epoch": 2.8390243902439023,
+      "grad_norm": 3.098562240600586,
+      "learning_rate": 4.072804691343653e-06,
+      "loss": 0.4045,
+      "step": 582
+    },
+    {
+      "epoch": 2.84390243902439,
+      "grad_norm": 3.9232118129730225,
+      "learning_rate": 4.069824820107507e-06,
+      "loss": 0.9564,
+      "step": 583
+    },
+    {
+      "epoch": 2.848780487804878,
+      "grad_norm": 2.7176268100738525,
+      "learning_rate": 4.06684126212611e-06,
+      "loss": 0.2703,
+      "step": 584
+    },
+    {
+      "epoch": 2.8536585365853657,
+      "grad_norm": 2.4905827045440674,
+      "learning_rate": 4.063854024406369e-06,
+      "loss": 0.4828,
+      "step": 585
+    },
+    {
+      "epoch": 2.8585365853658535,
+      "grad_norm": 2.848784923553467,
+      "learning_rate": 4.060863113963835e-06,
+      "loss": 0.4131,
+      "step": 586
+    },
+    {
+      "epoch": 2.8634146341463413,
+      "grad_norm": 2.599665403366089,
+      "learning_rate": 4.057868537822683e-06,
+      "loss": 0.4464,
+      "step": 587
+    },
+    {
+      "epoch": 2.868292682926829,
+      "grad_norm": 3.1770827770233154,
+      "learning_rate": 4.054870303015695e-06,
+      "loss": 0.2825,
+      "step": 588
+    },
+    {
+      "epoch": 2.873170731707317,
+      "grad_norm": 3.18332839012146,
+      "learning_rate": 4.05186841658425e-06,
+      "loss": 0.4438,
+      "step": 589
+    },
+    {
+      "epoch": 2.8780487804878048,
+      "grad_norm": 2.7485718727111816,
+      "learning_rate": 4.048862885578301e-06,
+      "loss": 0.4817,
+      "step": 590
+    },
+    {
+      "epoch": 2.8829268292682926,
+      "grad_norm": 2.9712934494018555,
+      "learning_rate": 4.045853717056358e-06,
+      "loss": 0.5157,
+      "step": 591
+    },
+    {
+      "epoch": 2.8878048780487804,
+      "grad_norm": 2.246858835220337,
+      "learning_rate": 4.0428409180854775e-06,
+      "loss": 0.4029,
+      "step": 592
+    },
+    {
+      "epoch": 2.892682926829268,
+      "grad_norm": 2.683434247970581,
+      "learning_rate": 4.039824495741238e-06,
+      "loss": 0.3796,
+      "step": 593
+    },
+    {
+      "epoch": 2.897560975609756,
+      "grad_norm": 2.6297569274902344,
+      "learning_rate": 4.036804457107733e-06,
+      "loss": 0.4467,
+      "step": 594
+    },
+    {
+      "epoch": 2.902439024390244,
+      "grad_norm": 5.318776607513428,
+      "learning_rate": 4.0337808092775435e-06,
+      "loss": 0.7007,
+      "step": 595
+    },
+    {
+      "epoch": 2.9073170731707316,
+      "grad_norm": 3.069889783859253,
+      "learning_rate": 4.030753559351728e-06,
+      "loss": 0.3219,
+      "step": 596
+    },
+    {
+      "epoch": 2.9121951219512194,
+      "grad_norm": 1.9730123281478882,
+      "learning_rate": 4.027722714439808e-06,
+      "loss": 0.3038,
+      "step": 597
+    },
+    {
+      "epoch": 2.9170731707317072,
+      "grad_norm": 3.7959916591644287,
+      "learning_rate": 4.024688281659743e-06,
+      "loss": 0.7768,
+      "step": 598
+    },
+    {
+      "epoch": 2.921951219512195,
+      "grad_norm": 3.900886297225952,
+      "learning_rate": 4.021650268137924e-06,
+      "loss": 0.4667,
+      "step": 599
+    },
+    {
+      "epoch": 2.926829268292683,
+      "grad_norm": 2.6155691146850586,
+      "learning_rate": 4.018608681009143e-06,
+      "loss": 0.3852,
+      "step": 600
+    },
+    {
+      "epoch": 2.9317073170731707,
+      "grad_norm": 3.2715704441070557,
+      "learning_rate": 4.015563527416596e-06,
+      "loss": 0.4804,
+      "step": 601
+    },
+    {
+      "epoch": 2.9365853658536585,
+      "grad_norm": 3.001425266265869,
+      "learning_rate": 4.012514814511844e-06,
+      "loss": 0.4152,
+      "step": 602
+    },
+    {
+      "epoch": 2.9414634146341463,
+      "grad_norm": 2.685360908508301,
+      "learning_rate": 4.009462549454816e-06,
+      "loss": 0.5029,
+      "step": 603
+    },
+    {
+      "epoch": 2.946341463414634,
+      "grad_norm": 3.4670183658599854,
+      "learning_rate": 4.006406739413775e-06,
+      "loss": 0.4857,
+      "step": 604
+    },
+    {
+      "epoch": 2.951219512195122,
+      "grad_norm": 3.0613298416137695,
+      "learning_rate": 4.003347391565317e-06,
+      "loss": 0.4449,
+      "step": 605
+    },
+    {
+      "epoch": 2.9560975609756097,
+      "grad_norm": 3.207186698913574,
+      "learning_rate": 4.000284513094342e-06,
+      "loss": 0.4808,
+      "step": 606
+    },
+    {
+      "epoch": 2.9609756097560975,
+      "grad_norm": 2.910578727722168,
+      "learning_rate": 3.997218111194042e-06,
+      "loss": 0.4395,
+      "step": 607
+    },
+    {
+      "epoch": 2.9658536585365853,
+      "grad_norm": 2.581918954849243,
+      "learning_rate": 3.994148193065886e-06,
+      "loss": 0.3264,
+      "step": 608
+    },
+    {
+      "epoch": 2.970731707317073,
+      "grad_norm": 2.6517748832702637,
+      "learning_rate": 3.991074765919598e-06,
+      "loss": 0.3285,
+      "step": 609
+    },
+    {
+      "epoch": 2.975609756097561,
+      "grad_norm": 3.509756088256836,
+      "learning_rate": 3.987997836973147e-06,
+      "loss": 0.3638,
+      "step": 610
+    },
+    {
+      "epoch": 2.9804878048780488,
+      "grad_norm": 2.7382352352142334,
+      "learning_rate": 3.984917413452721e-06,
+      "loss": 0.3853,
+      "step": 611
+    },
+    {
+      "epoch": 2.9853658536585366,
+      "grad_norm": 3.998974323272705,
+      "learning_rate": 3.981833502592717e-06,
+      "loss": 0.6411,
+      "step": 612
+    },
+    {
+      "epoch": 2.9902439024390244,
+      "grad_norm": 3.305126428604126,
+      "learning_rate": 3.978746111635725e-06,
+      "loss": 0.2759,
+      "step": 613
+    },
+    {
+      "epoch": 2.995121951219512,
+      "grad_norm": 3.137300968170166,
+      "learning_rate": 3.9756552478325045e-06,
+      "loss": 0.4566,
+      "step": 614
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.617291212081909,
+      "learning_rate": 3.972560918441972e-06,
+      "loss": 0.2221,
+      "step": 615
+    },
+    {
+      "epoch": 3.004878048780488,
+      "grad_norm": 2.787429094314575,
+      "learning_rate": 3.969463130731183e-06,
+      "loss": 0.2403,
+      "step": 616
+    },
+    {
+      "epoch": 3.0097560975609756,
+      "grad_norm": 3.0412075519561768,
+      "learning_rate": 3.966361891975316e-06,
+      "loss": 0.2635,
+      "step": 617
+    },
+    {
+      "epoch": 3.0146341463414634,
+      "grad_norm": 2.9949851036071777,
+      "learning_rate": 3.963257209457652e-06,
+      "loss": 0.3294,
+      "step": 618
+    },
+    {
+      "epoch": 3.0195121951219512,
+      "grad_norm": 3.0510809421539307,
+      "learning_rate": 3.960149090469561e-06,
+      "loss": 0.1338,
+      "step": 619
+    },
+    {
+      "epoch": 3.024390243902439,
+      "grad_norm": 3.669482707977295,
+      "learning_rate": 3.957037542310484e-06,
+      "loss": 0.1469,
+      "step": 620
+    },
+    {
+      "epoch": 3.029268292682927,
+      "grad_norm": 4.677116870880127,
+      "learning_rate": 3.953922572287915e-06,
+      "loss": 0.2788,
+      "step": 621
+    },
+    {
+      "epoch": 3.0341463414634147,
+      "grad_norm": 4.33144474029541,
+      "learning_rate": 3.950804187717384e-06,
+      "loss": 0.4521,
+      "step": 622
+    },
+    {
+      "epoch": 3.0390243902439025,
+      "grad_norm": 3.466639757156372,
+      "learning_rate": 3.947682395922439e-06,
+      "loss": 0.5113,
+      "step": 623
+    },
+    {
+      "epoch": 3.0439024390243903,
+      "grad_norm": 3.2332122325897217,
+      "learning_rate": 3.9445572042346346e-06,
+      "loss": 0.0968,
+      "step": 624
+    },
+    {
+      "epoch": 3.048780487804878,
+      "grad_norm": 2.6108055114746094,
+      "learning_rate": 3.941428619993505e-06,
+      "loss": 0.2462,
+      "step": 625
+    },
+    {
+      "epoch": 3.053658536585366,
+      "grad_norm": 3.2512595653533936,
+      "learning_rate": 3.938296650546552e-06,
+      "loss": 0.1782,
+      "step": 626
+    },
+    {
+      "epoch": 3.0585365853658537,
+      "grad_norm": 3.4350366592407227,
+      "learning_rate": 3.935161303249231e-06,
+      "loss": 0.2955,
+      "step": 627
+    },
+    {
+      "epoch": 3.0634146341463415,
+      "grad_norm": 3.42012619972229,
+      "learning_rate": 3.932022585464928e-06,
+      "loss": 0.3259,
+      "step": 628
+    },
+    {
+      "epoch": 3.0682926829268293,
+      "grad_norm": 3.458043336868286,
+      "learning_rate": 3.928880504564943e-06,
+      "loss": 0.2306,
+      "step": 629
+    },
+    {
+      "epoch": 3.073170731707317,
+      "grad_norm": 2.646616220474243,
+      "learning_rate": 3.92573506792848e-06,
+      "loss": 0.2197,
+      "step": 630
+    },
+    {
+      "epoch": 3.078048780487805,
+      "grad_norm": 3.5558857917785645,
+      "learning_rate": 3.9225862829426184e-06,
+      "loss": 0.1607,
+      "step": 631
+    },
+    {
+      "epoch": 3.0829268292682928,
+      "grad_norm": 3.6011338233947754,
+      "learning_rate": 3.919434157002303e-06,
+      "loss": 0.3087,
+      "step": 632
+    },
+    {
+      "epoch": 3.0878048780487806,
+      "grad_norm": 2.339879035949707,
+      "learning_rate": 3.916278697510325e-06,
+      "loss": 0.2213,
+      "step": 633
+    },
+    {
+      "epoch": 3.0926829268292684,
+      "grad_norm": 3.268162488937378,
+      "learning_rate": 3.913119911877305e-06,
+      "loss": 0.318,
+      "step": 634
+    },
+    {
+      "epoch": 3.097560975609756,
+      "grad_norm": 4.062571048736572,
+      "learning_rate": 3.909957807521674e-06,
+      "loss": 0.1757,
+      "step": 635
+    },
+    {
+      "epoch": 3.102439024390244,
+      "grad_norm": 2.997659683227539,
+      "learning_rate": 3.906792391869657e-06,
+      "loss": 0.2391,
+      "step": 636
+    },
+    {
+      "epoch": 3.107317073170732,
+      "grad_norm": 3.7037394046783447,
+      "learning_rate": 3.903623672355258e-06,
+      "loss": 0.2548,
+      "step": 637
+    },
+    {
+      "epoch": 3.1121951219512196,
+      "grad_norm": 3.110579252243042,
+      "learning_rate": 3.900451656420237e-06,
+      "loss": 0.2389,
+      "step": 638
+    },
+    {
+      "epoch": 3.1170731707317074,
+      "grad_norm": 3.3332321643829346,
+      "learning_rate": 3.897276351514097e-06,
+      "loss": 0.1371,
+      "step": 639
+    },
+    {
+      "epoch": 3.1219512195121952,
+      "grad_norm": 3.8275935649871826,
+      "learning_rate": 3.894097765094065e-06,
+      "loss": 0.3363,
+      "step": 640
+    },
+    {
+      "epoch": 3.126829268292683,
+      "grad_norm": 2.3731374740600586,
+      "learning_rate": 3.890915904625075e-06,
+      "loss": 0.1314,
+      "step": 641
+    },
+    {
+      "epoch": 3.131707317073171,
+      "grad_norm": 3.1511282920837402,
+      "learning_rate": 3.887730777579751e-06,
+      "loss": 0.3563,
+      "step": 642
+    },
+    {
+      "epoch": 3.1365853658536587,
+      "grad_norm": 4.2254862785339355,
+      "learning_rate": 3.884542391438387e-06,
+      "loss": 0.5053,
+      "step": 643
+    },
+    {
+      "epoch": 3.1414634146341465,
+      "grad_norm": 4.579670429229736,
+      "learning_rate": 3.88135075368893e-06,
+      "loss": 0.6259,
+      "step": 644
+    },
+    {
+      "epoch": 3.1463414634146343,
+      "grad_norm": 3.2102746963500977,
+      "learning_rate": 3.878155871826968e-06,
+      "loss": 0.2599,
+      "step": 645
+    },
+    {
+      "epoch": 3.151219512195122,
+      "grad_norm": 2.5569686889648438,
+      "learning_rate": 3.874957753355701e-06,
+      "loss": 0.2075,
+      "step": 646
+    },
+    {
+      "epoch": 3.15609756097561,
+      "grad_norm": 3.588925838470459,
+      "learning_rate": 3.8717564057859365e-06,
+      "loss": 0.4577,
+      "step": 647
+    },
+    {
+      "epoch": 3.1609756097560977,
+      "grad_norm": 3.6163878440856934,
+      "learning_rate": 3.868551836636063e-06,
+      "loss": 0.4023,
+      "step": 648
+    },
+    {
+      "epoch": 3.1658536585365855,
+      "grad_norm": 3.8688390254974365,
+      "learning_rate": 3.865344053432035e-06,
+      "loss": 0.1669,
+      "step": 649
+    },
+    {
+      "epoch": 3.1707317073170733,
+      "grad_norm": 3.419734001159668,
+      "learning_rate": 3.862133063707353e-06,
+      "loss": 0.2766,
+      "step": 650
+    },
+    {
+      "epoch": 3.175609756097561,
+      "grad_norm": 2.9860243797302246,
+      "learning_rate": 3.858918875003053e-06,
+      "loss": 0.1788,
+      "step": 651
+    },
+    {
+      "epoch": 3.180487804878049,
+      "grad_norm": 3.0619022846221924,
+      "learning_rate": 3.855701494867679e-06,
+      "loss": 0.224,
+      "step": 652
+    },
+    {
+      "epoch": 3.1853658536585368,
+      "grad_norm": 3.3668978214263916,
+      "learning_rate": 3.852480930857275e-06,
+      "loss": 0.4029,
+      "step": 653
+    },
+    {
+      "epoch": 3.1902439024390246,
+      "grad_norm": 3.543147563934326,
+      "learning_rate": 3.849257190535356e-06,
+      "loss": 0.2096,
+      "step": 654
+    },
+    {
+      "epoch": 3.1951219512195124,
+      "grad_norm": 3.793619155883789,
+      "learning_rate": 3.846030281472902e-06,
+      "loss": 0.5574,
+      "step": 655
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.021289110183716,
+      "learning_rate": 3.842800211248333e-06,
+      "loss": 0.2233,
+      "step": 656
+    },
+    {
+      "epoch": 3.204878048780488,
+      "grad_norm": 4.582934856414795,
+      "learning_rate": 3.839566987447492e-06,
+      "loss": 0.3871,
+      "step": 657
+    },
+    {
+      "epoch": 3.209756097560976,
+      "grad_norm": 2.996340274810791,
+      "learning_rate": 3.8363306176636296e-06,
+      "loss": 0.4325,
+      "step": 658
+    },
+    {
+      "epoch": 3.2146341463414636,
+      "grad_norm": 3.3190877437591553,
+      "learning_rate": 3.833091109497384e-06,
+      "loss": 0.5321,
+      "step": 659
+    },
+    {
+      "epoch": 3.2195121951219514,
+      "grad_norm": 3.2532856464385986,
+      "learning_rate": 3.829848470556765e-06,
+      "loss": 0.1359,
+      "step": 660
+    },
+    {
+      "epoch": 3.2243902439024392,
+      "grad_norm": 2.7875044345855713,
+      "learning_rate": 3.8266027084571335e-06,
+      "loss": 0.3145,
+      "step": 661
+    },
+    {
+      "epoch": 3.229268292682927,
+      "grad_norm": 3.748253583908081,
+      "learning_rate": 3.823353830821187e-06,
+      "loss": 0.1252,
+      "step": 662
+    },
+    {
+      "epoch": 3.234146341463415,
+      "grad_norm": 2.858293294906616,
+      "learning_rate": 3.820101845278937e-06,
+      "loss": 0.2589,
+      "step": 663
+    },
+    {
+      "epoch": 3.2390243902439027,
+      "grad_norm": 3.7470967769622803,
+      "learning_rate": 3.816846759467696e-06,
+      "loss": 0.2594,
+      "step": 664
+    },
+    {
+      "epoch": 3.2439024390243905,
+      "grad_norm": 3.676196813583374,
+      "learning_rate": 3.8135885810320587e-06,
+      "loss": 0.2998,
+      "step": 665
+    },
+    {
+      "epoch": 3.2487804878048783,
+      "grad_norm": 3.0943140983581543,
+      "learning_rate": 3.810327317623881e-06,
+      "loss": 0.2238,
+      "step": 666
+    },
+    {
+      "epoch": 3.253658536585366,
+      "grad_norm": 3.5907349586486816,
+      "learning_rate": 3.8070629769022628e-06,
+      "loss": 0.3381,
+      "step": 667
+    },
+    {
+      "epoch": 3.258536585365854,
+      "grad_norm": 3.1195285320281982,
+      "learning_rate": 3.8037955665335335e-06,
+      "loss": 0.2407,
+      "step": 668
+    },
+    {
+      "epoch": 3.2634146341463417,
+      "grad_norm": 3.422292947769165,
+      "learning_rate": 3.800525094191231e-06,
+      "loss": 0.2957,
+      "step": 669
+    },
+    {
+      "epoch": 3.2682926829268295,
+      "grad_norm": 2.5264663696289062,
+      "learning_rate": 3.797251567556083e-06,
+      "loss": 0.2493,
+      "step": 670
+    },
+    {
+      "epoch": 3.2731707317073173,
+      "grad_norm": 3.350219964981079,
+      "learning_rate": 3.793974994315991e-06,
+      "loss": 0.1186,
+      "step": 671
+    },
+    {
+      "epoch": 3.278048780487805,
+      "grad_norm": 4.175906181335449,
+      "learning_rate": 3.790695382166013e-06,
+      "loss": 0.3453,
+      "step": 672
+    },
+    {
+      "epoch": 3.2829268292682925,
+      "grad_norm": 3.006072521209717,
+      "learning_rate": 3.7874127388083415e-06,
+      "loss": 0.1981,
+      "step": 673
+    },
+    {
+      "epoch": 3.2878048780487803,
+      "grad_norm": 3.368561029434204,
+      "learning_rate": 3.7841270719522895e-06,
+      "loss": 0.2934,
+      "step": 674
+    },
+    {
+      "epoch": 3.292682926829268,
+      "grad_norm": 4.374331951141357,
+      "learning_rate": 3.7808383893142692e-06,
+      "loss": 0.1359,
+      "step": 675
+    },
+    {
+      "epoch": 3.297560975609756,
+      "grad_norm": 3.297102451324463,
+      "learning_rate": 3.7775466986177763e-06,
+      "loss": 0.2498,
+      "step": 676
+    },
+    {
+      "epoch": 3.3024390243902437,
+      "grad_norm": 2.8914761543273926,
+      "learning_rate": 3.774252007593371e-06,
+      "loss": 0.1308,
+      "step": 677
+    },
+    {
+      "epoch": 3.3073170731707315,
+      "grad_norm": 3.1550722122192383,
+      "learning_rate": 3.7709543239786593e-06,
+      "loss": 0.3915,
+      "step": 678
+    },
+    {
+      "epoch": 3.3121951219512193,
+      "grad_norm": 3.2302658557891846,
+      "learning_rate": 3.767653655518277e-06,
+      "loss": 0.2558,
+      "step": 679
+    },
+    {
+      "epoch": 3.317073170731707,
+      "grad_norm": 4.4321770668029785,
+      "learning_rate": 3.7643500099638673e-06,
+      "loss": 0.1988,
+      "step": 680
+    },
+    {
+      "epoch": 3.321951219512195,
+      "grad_norm": 2.970566749572754,
+      "learning_rate": 3.7610433950740667e-06,
+      "loss": 0.4908,
+      "step": 681
+    },
+    {
+      "epoch": 3.3268292682926828,
+      "grad_norm": 3.5516228675842285,
+      "learning_rate": 3.757733818614485e-06,
+      "loss": 0.304,
+      "step": 682
+    },
+    {
+      "epoch": 3.3317073170731706,
+      "grad_norm": 2.7555387020111084,
+      "learning_rate": 3.7544212883576856e-06,
+      "loss": 0.2533,
+      "step": 683
+    },
+    {
+      "epoch": 3.3365853658536584,
+      "grad_norm": 3.61226749420166,
+      "learning_rate": 3.751105812083172e-06,
+      "loss": 0.1771,
+      "step": 684
+    },
+    {
+      "epoch": 3.341463414634146,
+      "grad_norm": 3.0466206073760986,
+      "learning_rate": 3.7477873975773655e-06,
+      "loss": 0.4213,
+      "step": 685
+    },
+    {
+      "epoch": 3.346341463414634,
+      "grad_norm": 3.6091527938842773,
+      "learning_rate": 3.7444660526335853e-06,
+      "loss": 0.3808,
+      "step": 686
+    },
+    {
+      "epoch": 3.351219512195122,
+      "grad_norm": 3.8443002700805664,
+      "learning_rate": 3.741141785052036e-06,
+      "loss": 0.6438,
+      "step": 687
+    },
+    {
+      "epoch": 3.3560975609756096,
+      "grad_norm": 3.845909833908081,
+      "learning_rate": 3.737814602639784e-06,
+      "loss": 0.3686,
+      "step": 688
+    },
+    {
+      "epoch": 3.3609756097560974,
+      "grad_norm": 2.904892921447754,
+      "learning_rate": 3.7344845132107427e-06,
+      "loss": 0.2934,
+      "step": 689
+    },
+    {
+      "epoch": 3.3658536585365852,
+      "grad_norm": 3.4766387939453125,
+      "learning_rate": 3.731151524585651e-06,
+      "loss": 0.3299,
+      "step": 690
+    },
+    {
+      "epoch": 3.370731707317073,
+      "grad_norm": 4.236767768859863,
+      "learning_rate": 3.7278156445920584e-06,
+      "loss": 0.6303,
+      "step": 691
+    },
+    {
+      "epoch": 3.375609756097561,
+      "grad_norm": 3.1122591495513916,
+      "learning_rate": 3.724476881064303e-06,
+      "loss": 0.2432,
+      "step": 692
+    },
+    {
+      "epoch": 3.3804878048780487,
+      "grad_norm": 3.0971457958221436,
+      "learning_rate": 3.721135241843496e-06,
+      "loss": 0.3131,
+      "step": 693
+    },
+    {
+      "epoch": 3.3853658536585365,
+      "grad_norm": 3.9365804195404053,
+      "learning_rate": 3.7177907347775016e-06,
+      "loss": 0.3372,
+      "step": 694
+    },
+    {
+      "epoch": 3.3902439024390243,
+      "grad_norm": 3.760373115539551,
+      "learning_rate": 3.71444336772092e-06,
+      "loss": 0.5055,
+      "step": 695
+    },
+    {
+      "epoch": 3.395121951219512,
+      "grad_norm": 4.360848426818848,
+      "learning_rate": 3.711093148535068e-06,
+      "loss": 0.6183,
+      "step": 696
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 3.7713537216186523,
+      "learning_rate": 3.707740085087959e-06,
+      "loss": 0.1568,
+      "step": 697
+    },
+    {
+      "epoch": 3.4048780487804877,
+      "grad_norm": 3.8532230854034424,
+      "learning_rate": 3.7043841852542884e-06,
+      "loss": 0.2826,
+      "step": 698
+    },
+    {
+      "epoch": 3.4097560975609755,
+      "grad_norm": 3.0548605918884277,
+      "learning_rate": 3.701025456915411e-06,
+      "loss": 0.1918,
+      "step": 699
+    },
+    {
+      "epoch": 3.4146341463414633,
+      "grad_norm": 3.2431821823120117,
+      "learning_rate": 3.697663907959327e-06,
+      "loss": 0.2493,
+      "step": 700
+    },
+    {
+      "epoch": 3.419512195121951,
+      "grad_norm": 3.7301864624023438,
+      "learning_rate": 3.6942995462806574e-06,
+      "loss": 0.4913,
+      "step": 701
+    },
+    {
+      "epoch": 3.424390243902439,
+      "grad_norm": 2.5468900203704834,
+      "learning_rate": 3.6909323797806314e-06,
+      "loss": 0.1788,
+      "step": 702
+    },
+    {
+      "epoch": 3.4292682926829268,
+      "grad_norm": 3.3719515800476074,
+      "learning_rate": 3.6875624163670635e-06,
+      "loss": 0.4162,
+      "step": 703
+    },
+    {
+      "epoch": 3.4341463414634146,
+      "grad_norm": 3.528010368347168,
+      "learning_rate": 3.6841896639543394e-06,
+      "loss": 0.1924,
+      "step": 704
+    },
+    {
+      "epoch": 3.4390243902439024,
+      "grad_norm": 3.3636631965637207,
+      "learning_rate": 3.6808141304633924e-06,
+      "loss": 0.3177,
+      "step": 705
+    },
+    {
+      "epoch": 3.44390243902439,
+      "grad_norm": 3.418705463409424,
+      "learning_rate": 3.6774358238216878e-06,
+      "loss": 0.2301,
+      "step": 706
+    },
+    {
+      "epoch": 3.448780487804878,
+      "grad_norm": 4.720373630523682,
+      "learning_rate": 3.6740547519632048e-06,
+      "loss": 0.1894,
+      "step": 707
+    },
+    {
+      "epoch": 3.453658536585366,
+      "grad_norm": 2.9635703563690186,
+      "learning_rate": 3.670670922828414e-06,
+      "loss": 0.2642,
+      "step": 708
+    },
+    {
+      "epoch": 3.4585365853658536,
+      "grad_norm": 4.934754371643066,
+      "learning_rate": 3.667284344364264e-06,
+      "loss": 0.2275,
+      "step": 709
+    },
+    {
+      "epoch": 3.4634146341463414,
+      "grad_norm": 3.090585231781006,
+      "learning_rate": 3.6638950245241604e-06,
+      "loss": 0.4447,
+      "step": 710
+    },
+    {
+      "epoch": 3.4682926829268292,
+      "grad_norm": 4.360495090484619,
+      "learning_rate": 3.660502971267945e-06,
+      "loss": 0.2415,
+      "step": 711
+    },
+    {
+      "epoch": 3.473170731707317,
+      "grad_norm": 3.4893476963043213,
+      "learning_rate": 3.65710819256188e-06,
+      "loss": 0.0921,
+      "step": 712
+    },
+    {
+      "epoch": 3.478048780487805,
+      "grad_norm": 3.2423770427703857,
+      "learning_rate": 3.65371069637863e-06,
+      "loss": 0.2371,
+      "step": 713
+    },
+    {
+      "epoch": 3.4829268292682927,
+      "grad_norm": 3.0775890350341797,
+      "learning_rate": 3.650310490697238e-06,
+      "loss": 0.4026,
+      "step": 714
+    },
+    {
+      "epoch": 3.4878048780487805,
+      "grad_norm": 3.906625270843506,
+      "learning_rate": 3.646907583503114e-06,
+      "loss": 0.4312,
+      "step": 715
+    },
+    {
+      "epoch": 3.4926829268292683,
+      "grad_norm": 3.2140414714813232,
+      "learning_rate": 3.6435019827880093e-06,
+      "loss": 0.2309,
+      "step": 716
+    },
+    {
+      "epoch": 3.497560975609756,
+      "grad_norm": 3.048523426055908,
+      "learning_rate": 3.640093696550003e-06,
+      "loss": 0.296,
+      "step": 717
+    },
+    {
+      "epoch": 3.502439024390244,
+      "grad_norm": 2.9669039249420166,
+      "learning_rate": 3.6366827327934817e-06,
+      "loss": 0.2723,
+      "step": 718
+    },
+    {
+      "epoch": 3.5073170731707317,
+      "grad_norm": 3.6941726207733154,
+      "learning_rate": 3.6332690995291176e-06,
+      "loss": 0.3797,
+      "step": 719
+    },
+    {
+      "epoch": 3.5121951219512195,
+      "grad_norm": 5.135766506195068,
+      "learning_rate": 3.6298528047738545e-06,
+      "loss": 0.9868,
+      "step": 720
+    },
+    {
+      "epoch": 3.5170731707317073,
+      "grad_norm": 3.2021052837371826,
+      "learning_rate": 3.626433856550886e-06,
+      "loss": 0.4069,
+      "step": 721
+    },
+    {
+      "epoch": 3.521951219512195,
+      "grad_norm": 3.094444513320923,
+      "learning_rate": 3.623012262889637e-06,
+      "loss": 0.3368,
+      "step": 722
+    },
+    {
+      "epoch": 3.526829268292683,
+      "grad_norm": 3.609285354614258,
+      "learning_rate": 3.6195880318257465e-06,
+      "loss": 0.3972,
+      "step": 723
+    },
+    {
+      "epoch": 3.5317073170731708,
+      "grad_norm": 4.236501216888428,
+      "learning_rate": 3.616161171401046e-06,
+      "loss": 0.52,
+      "step": 724
+    },
+    {
+      "epoch": 3.5365853658536586,
+      "grad_norm": 3.504526376724243,
+      "learning_rate": 3.612731689663542e-06,
+      "loss": 0.23,
+      "step": 725
+    },
+    {
+      "epoch": 3.5414634146341464,
+      "grad_norm": 3.233591079711914,
+      "learning_rate": 3.6092995946673996e-06,
+      "loss": 0.4151,
+      "step": 726
+    },
+    {
+      "epoch": 3.546341463414634,
+      "grad_norm": 3.6701886653900146,
+      "learning_rate": 3.605864894472918e-06,
+      "loss": 0.2798,
+      "step": 727
+    },
+    {
+      "epoch": 3.551219512195122,
+      "grad_norm": 3.8713181018829346,
+      "learning_rate": 3.602427597146516e-06,
+      "loss": 0.4336,
+      "step": 728
+    },
+    {
+      "epoch": 3.55609756097561,
+      "grad_norm": 5.49612283706665,
+      "learning_rate": 3.5989877107607134e-06,
+      "loss": 0.4803,
+      "step": 729
+    },
+    {
+      "epoch": 3.5609756097560976,
+      "grad_norm": 3.771005392074585,
+      "learning_rate": 3.5955452433941075e-06,
+      "loss": 0.3698,
+      "step": 730
+    },
+    {
+      "epoch": 3.5658536585365854,
+      "grad_norm": 2.970822334289551,
+      "learning_rate": 3.5921002031313586e-06,
+      "loss": 0.2373,
+      "step": 731
+    },
+    {
+      "epoch": 3.5707317073170732,
+      "grad_norm": 3.517249584197998,
+      "learning_rate": 3.58865259806317e-06,
+      "loss": 0.1908,
+      "step": 732
+    },
+    {
+      "epoch": 3.575609756097561,
+      "grad_norm": 3.6825428009033203,
+      "learning_rate": 3.585202436286267e-06,
+      "loss": 0.3993,
+      "step": 733
+    },
+    {
+      "epoch": 3.580487804878049,
+      "grad_norm": 3.387479066848755,
+      "learning_rate": 3.581749725903381e-06,
+      "loss": 0.4237,
+      "step": 734
+    },
+    {
+      "epoch": 3.5853658536585367,
+      "grad_norm": 3.5004806518554688,
+      "learning_rate": 3.5782944750232274e-06,
+      "loss": 0.3011,
+      "step": 735
+    },
+    {
+      "epoch": 3.5902439024390245,
+      "grad_norm": 3.461731433868408,
+      "learning_rate": 3.574836691760489e-06,
+      "loss": 0.0896,
+      "step": 736
+    },
+    {
+      "epoch": 3.5951219512195123,
+      "grad_norm": 3.9598381519317627,
+      "learning_rate": 3.571376384235795e-06,
+      "loss": 0.2751,
+      "step": 737
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 4.053933143615723,
+      "learning_rate": 3.5679135605757035e-06,
+      "loss": 0.2086,
+      "step": 738
+    },
+    {
+      "epoch": 3.604878048780488,
+      "grad_norm": 2.9683544635772705,
+      "learning_rate": 3.564448228912682e-06,
+      "loss": 0.1659,
+      "step": 739
+    },
+    {
+      "epoch": 3.6097560975609757,
+      "grad_norm": 3.6598448753356934,
+      "learning_rate": 3.5609803973850877e-06,
+      "loss": 0.2469,
+      "step": 740
+    },
+    {
+      "epoch": 3.6146341463414635,
+      "grad_norm": 3.449335813522339,
+      "learning_rate": 3.557510074137147e-06,
+      "loss": 0.375,
+      "step": 741
+    },
+    {
+      "epoch": 3.6195121951219513,
+      "grad_norm": 2.7666923999786377,
+      "learning_rate": 3.554037267318942e-06,
+      "loss": 0.3133,
+      "step": 742
+    },
+    {
+      "epoch": 3.624390243902439,
+      "grad_norm": 2.8951869010925293,
+      "learning_rate": 3.5505619850863847e-06,
+      "loss": 0.2243,
+      "step": 743
+    },
+    {
+      "epoch": 3.629268292682927,
+      "grad_norm": 3.477747678756714,
+      "learning_rate": 3.5470842356012007e-06,
+      "loss": 0.1321,
+      "step": 744
+    },
+    {
+      "epoch": 3.6341463414634148,
+      "grad_norm": 3.810480833053589,
+      "learning_rate": 3.5436040270309113e-06,
+      "loss": 0.361,
+      "step": 745
+    },
+    {
+      "epoch": 3.6390243902439026,
+      "grad_norm": 3.0730793476104736,
+      "learning_rate": 3.540121367548811e-06,
+      "loss": 0.1523,
+      "step": 746
+    },
+    {
+      "epoch": 3.6439024390243904,
+      "grad_norm": 3.6878390312194824,
+      "learning_rate": 3.5366362653339524e-06,
+      "loss": 0.4898,
+      "step": 747
+    },
+    {
+      "epoch": 3.648780487804878,
+      "grad_norm": 3.6432242393493652,
+      "learning_rate": 3.533148728571124e-06,
+      "loss": 0.1397,
+      "step": 748
+    },
+    {
+      "epoch": 3.653658536585366,
+      "grad_norm": 3.7047760486602783,
+      "learning_rate": 3.5296587654508317e-06,
+      "loss": 0.323,
+      "step": 749
+    },
+    {
+      "epoch": 3.658536585365854,
+      "grad_norm": 3.777132749557495,
+      "learning_rate": 3.526166384169279e-06,
+      "loss": 0.5577,
+      "step": 750
+    },
+    {
+      "epoch": 3.6634146341463416,
+      "grad_norm": 3.7970924377441406,
+      "learning_rate": 3.5226715929283507e-06,
+      "loss": 0.245,
+      "step": 751
+    },
+    {
+      "epoch": 3.6682926829268294,
+      "grad_norm": 2.8203537464141846,
+      "learning_rate": 3.519174399935588e-06,
+      "loss": 0.1619,
+      "step": 752
+    },
+    {
+      "epoch": 3.6731707317073172,
+      "grad_norm": 3.4040987491607666,
+      "learning_rate": 3.5156748134041767e-06,
+      "loss": 0.1047,
+      "step": 753
+    },
+    {
+      "epoch": 3.678048780487805,
+      "grad_norm": 3.927960157394409,
+      "learning_rate": 3.5121728415529203e-06,
+      "loss": 0.5713,
+      "step": 754
+    },
+    {
+      "epoch": 3.682926829268293,
+      "grad_norm": 3.3833277225494385,
+      "learning_rate": 3.5086684926062266e-06,
+      "loss": 0.2174,
+      "step": 755
+    },
+    {
+      "epoch": 3.68780487804878,
+      "grad_norm": 3.989307403564453,
+      "learning_rate": 3.505161774794085e-06,
+      "loss": 0.285,
+      "step": 756
+    },
+    {
+      "epoch": 3.692682926829268,
+      "grad_norm": 2.742429494857788,
+      "learning_rate": 3.5016526963520474e-06,
+      "loss": 0.1602,
+      "step": 757
+    },
+    {
+      "epoch": 3.697560975609756,
+      "grad_norm": 3.7082698345184326,
+      "learning_rate": 3.498141265521212e-06,
+      "loss": 0.666,
+      "step": 758
+    },
+    {
+      "epoch": 3.7024390243902436,
+      "grad_norm": 3.033196210861206,
+      "learning_rate": 3.4946274905481997e-06,
+      "loss": 0.2024,
+      "step": 759
+    },
+    {
+      "epoch": 3.7073170731707314,
+      "grad_norm": 3.7145371437072754,
+      "learning_rate": 3.4911113796851364e-06,
+      "loss": 0.2719,
+      "step": 760
+    },
+    {
+      "epoch": 3.7121951219512193,
+      "grad_norm": 3.580298900604248,
+      "learning_rate": 3.487592941189636e-06,
+      "loss": 0.1537,
+      "step": 761
+    },
+    {
+      "epoch": 3.717073170731707,
+      "grad_norm": 4.753757953643799,
+      "learning_rate": 3.484072183324776e-06,
+      "loss": 0.6149,
+      "step": 762
+    },
+    {
+      "epoch": 3.721951219512195,
+      "grad_norm": 3.5575687885284424,
+      "learning_rate": 3.4805491143590823e-06,
+      "loss": 0.4241,
+      "step": 763
+    },
+    {
+      "epoch": 3.7268292682926827,
+      "grad_norm": 3.215224266052246,
+      "learning_rate": 3.4770237425665103e-06,
+      "loss": 0.3037,
+      "step": 764
+    },
+    {
+      "epoch": 3.7317073170731705,
+      "grad_norm": 2.9899685382843018,
+      "learning_rate": 3.4734960762264204e-06,
+      "loss": 0.4854,
+      "step": 765
+    },
+    {
+      "epoch": 3.7365853658536583,
+      "grad_norm": 3.5880227088928223,
+      "learning_rate": 3.469966123623563e-06,
+      "loss": 0.3849,
+      "step": 766
+    },
+    {
+      "epoch": 3.741463414634146,
+      "grad_norm": 3.472750186920166,
+      "learning_rate": 3.46643389304806e-06,
+      "loss": 0.3159,
+      "step": 767
+    },
+    {
+      "epoch": 3.746341463414634,
+      "grad_norm": 4.355650901794434,
+      "learning_rate": 3.4628993927953786e-06,
+      "loss": 0.7527,
+      "step": 768
+    },
+    {
+      "epoch": 3.7512195121951217,
+      "grad_norm": 2.94575834274292,
+      "learning_rate": 3.45936263116632e-06,
+      "loss": 0.1716,
+      "step": 769
+    },
+    {
+      "epoch": 3.7560975609756095,
+      "grad_norm": 2.991525173187256,
+      "learning_rate": 3.4558236164669957e-06,
+      "loss": 0.2061,
+      "step": 770
+    },
+    {
+      "epoch": 3.7609756097560973,
+      "grad_norm": 3.134000301361084,
+      "learning_rate": 3.4522823570088073e-06,
+      "loss": 0.1338,
+      "step": 771
+    },
+    {
+      "epoch": 3.765853658536585,
+      "grad_norm": 3.722140312194824,
+      "learning_rate": 3.4487388611084295e-06,
+      "loss": 0.2615,
+      "step": 772
+    },
+    {
+      "epoch": 3.770731707317073,
+      "grad_norm": 3.7941153049468994,
+      "learning_rate": 3.445193137087788e-06,
+      "loss": 0.1401,
+      "step": 773
+    },
+    {
+      "epoch": 3.7756097560975608,
+      "grad_norm": 2.872941732406616,
+      "learning_rate": 3.4416451932740424e-06,
+      "loss": 0.2934,
+      "step": 774
+    },
+    {
+      "epoch": 3.7804878048780486,
+      "grad_norm": 4.5019941329956055,
+      "learning_rate": 3.4380950379995652e-06,
+      "loss": 0.4579,
+      "step": 775
+    },
+    {
+      "epoch": 3.7853658536585364,
+      "grad_norm": 2.682884931564331,
+      "learning_rate": 3.434542679601922e-06,
+      "loss": 0.2979,
+      "step": 776
+    },
+    {
+      "epoch": 3.790243902439024,
+      "grad_norm": 3.3044273853302,
+      "learning_rate": 3.4309881264238538e-06,
+      "loss": 0.1196,
+      "step": 777
+    },
+    {
+      "epoch": 3.795121951219512,
+      "grad_norm": 3.102760076522827,
+      "learning_rate": 3.4274313868132547e-06,
+      "loss": 0.2026,
+      "step": 778
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 3.3304500579833984,
+      "learning_rate": 3.4238724691231534e-06,
+      "loss": 0.2135,
+      "step": 779
+    },
+    {
+      "epoch": 3.8048780487804876,
+      "grad_norm": 3.295119047164917,
+      "learning_rate": 3.4203113817116955e-06,
+      "loss": 0.4418,
+      "step": 780
+    },
+    {
+      "epoch": 3.8097560975609754,
+      "grad_norm": 3.6655640602111816,
+      "learning_rate": 3.4167481329421204e-06,
+      "loss": 0.203,
+      "step": 781
+    },
+    {
+      "epoch": 3.8146341463414632,
+      "grad_norm": 3.387830972671509,
+      "learning_rate": 3.4131827311827447e-06,
+      "loss": 0.3225,
+      "step": 782
+    },
+    {
+      "epoch": 3.819512195121951,
+      "grad_norm": 2.621633529663086,
+      "learning_rate": 3.4096151848069416e-06,
+      "loss": 0.1704,
+      "step": 783
+    },
+    {
+      "epoch": 3.824390243902439,
+      "grad_norm": 2.974344491958618,
+      "learning_rate": 3.4060455021931195e-06,
+      "loss": 0.2785,
+      "step": 784
+    },
+    {
+      "epoch": 3.8292682926829267,
+      "grad_norm": 3.452131748199463,
+      "learning_rate": 3.402473691724704e-06,
+      "loss": 0.223,
+      "step": 785
+    },
+    {
+      "epoch": 3.8341463414634145,
+      "grad_norm": 2.6373705863952637,
+      "learning_rate": 3.39889976179012e-06,
+      "loss": 0.2368,
+      "step": 786
+    },
+    {
+      "epoch": 3.8390243902439023,
+      "grad_norm": 2.863184928894043,
+      "learning_rate": 3.3953237207827673e-06,
+      "loss": 0.3294,
+      "step": 787
+    },
+    {
+      "epoch": 3.84390243902439,
+      "grad_norm": 5.104704856872559,
+      "learning_rate": 3.391745577101005e-06,
+      "loss": 0.5431,
+      "step": 788
+    },
+    {
+      "epoch": 3.848780487804878,
+      "grad_norm": 3.951310634613037,
+      "learning_rate": 3.3881653391481306e-06,
+      "loss": 0.2546,
+      "step": 789
+    },
+    {
+      "epoch": 3.8536585365853657,
+      "grad_norm": 3.9903225898742676,
+      "learning_rate": 3.384583015332359e-06,
+      "loss": 0.3293,
+      "step": 790
+    },
+    {
+      "epoch": 3.8585365853658535,
+      "grad_norm": 3.3149220943450928,
+      "learning_rate": 3.380998614066805e-06,
+      "loss": 0.1861,
+      "step": 791
+    },
+    {
+      "epoch": 3.8634146341463413,
+      "grad_norm": 3.6755223274230957,
+      "learning_rate": 3.3774121437694606e-06,
+      "loss": 0.2498,
+      "step": 792
+    },
+    {
+      "epoch": 3.868292682926829,
+      "grad_norm": 3.192918300628662,
+      "learning_rate": 3.3738236128631786e-06,
+      "loss": 0.1525,
+      "step": 793
+    },
+    {
+      "epoch": 3.873170731707317,
+      "grad_norm": 3.5358777046203613,
+      "learning_rate": 3.3702330297756503e-06,
+      "loss": 0.3622,
+      "step": 794
+    },
+    {
+      "epoch": 3.8780487804878048,
+      "grad_norm": 3.619878053665161,
+      "learning_rate": 3.366640402939387e-06,
+      "loss": 0.1051,
+      "step": 795
+    },
+    {
+      "epoch": 3.8829268292682926,
+      "grad_norm": 7.085352420806885,
+      "learning_rate": 3.363045740791698e-06,
+      "loss": 0.4606,
+      "step": 796
+    },
+    {
+      "epoch": 3.8878048780487804,
+      "grad_norm": 2.523165464401245,
+      "learning_rate": 3.3594490517746774e-06,
+      "loss": 0.2267,
+      "step": 797
+    },
+    {
+      "epoch": 3.892682926829268,
+      "grad_norm": 2.7026922702789307,
+      "learning_rate": 3.3558503443351733e-06,
+      "loss": 0.2792,
+      "step": 798
+    },
+    {
+      "epoch": 3.897560975609756,
+      "grad_norm": 2.9232428073883057,
+      "learning_rate": 3.352249626924777e-06,
+      "loss": 0.2579,
+      "step": 799
+    },
+    {
+      "epoch": 3.902439024390244,
+      "grad_norm": 4.760788440704346,
+      "learning_rate": 3.348646907999801e-06,
+      "loss": 0.6983,
+      "step": 800
+    },
+    {
+      "epoch": 3.9073170731707316,
+      "grad_norm": 3.198249578475952,
+      "learning_rate": 3.345042196021257e-06,
+      "loss": 0.3265,
+      "step": 801
+    },
+    {
+      "epoch": 3.9121951219512194,
+      "grad_norm": 4.069286823272705,
+      "learning_rate": 3.3414354994548385e-06,
+      "loss": 0.497,
+      "step": 802
+    },
+    {
+      "epoch": 3.9170731707317072,
+      "grad_norm": 3.4435410499572754,
+      "learning_rate": 3.337826826770898e-06,
+      "loss": 0.2812,
+      "step": 803
+    },
+    {
+      "epoch": 3.921951219512195,
+      "grad_norm": 3.9805212020874023,
+      "learning_rate": 3.3342161864444312e-06,
+      "loss": 0.2277,
+      "step": 804
+    },
+    {
+      "epoch": 3.926829268292683,
+      "grad_norm": 3.348925828933716,
+      "learning_rate": 3.3306035869550534e-06,
+      "loss": 0.1614,
+      "step": 805
+    },
+    {
+      "epoch": 3.9317073170731707,
+      "grad_norm": 4.7613701820373535,
+      "learning_rate": 3.326989036786981e-06,
+      "loss": 0.3269,
+      "step": 806
+    },
+    {
+      "epoch": 3.9365853658536585,
+      "grad_norm": 3.807502508163452,
+      "learning_rate": 3.3233725444290126e-06,
+      "loss": 0.2619,
+      "step": 807
+    },
+    {
+      "epoch": 3.9414634146341463,
+      "grad_norm": 3.2690203189849854,
+      "learning_rate": 3.3197541183745065e-06,
+      "loss": 0.4334,
+      "step": 808
+    },
+    {
+      "epoch": 3.946341463414634,
+      "grad_norm": 3.396993398666382,
+      "learning_rate": 3.3161337671213634e-06,
+      "loss": 0.2738,
+      "step": 809
+    },
+    {
+      "epoch": 3.951219512195122,
+      "grad_norm": 3.086669921875,
+      "learning_rate": 3.312511499172006e-06,
+      "loss": 0.1597,
+      "step": 810
+    },
+    {
+      "epoch": 3.9560975609756097,
+      "grad_norm": 3.5688745975494385,
+      "learning_rate": 3.3088873230333562e-06,
+      "loss": 0.3195,
+      "step": 811
+    },
+    {
+      "epoch": 3.9609756097560975,
+      "grad_norm": 3.4843621253967285,
+      "learning_rate": 3.3052612472168193e-06,
+      "loss": 0.1865,
+      "step": 812
+    },
+    {
+      "epoch": 3.9658536585365853,
+      "grad_norm": 2.8479580879211426,
+      "learning_rate": 3.3016332802382618e-06,
+      "loss": 0.3108,
+      "step": 813
+    },
+    {
+      "epoch": 3.970731707317073,
+      "grad_norm": 3.3241543769836426,
+      "learning_rate": 3.2980034306179897e-06,
+      "loss": 0.2099,
+      "step": 814
+    },
+    {
+      "epoch": 3.975609756097561,
+      "grad_norm": 2.817675828933716,
+      "learning_rate": 3.294371706880733e-06,
+      "loss": 0.3073,
+      "step": 815
+    },
+    {
+      "epoch": 3.9804878048780488,
+      "grad_norm": 2.9535388946533203,
+      "learning_rate": 3.290738117555622e-06,
+      "loss": 0.2024,
+      "step": 816
+    },
+    {
+      "epoch": 3.9853658536585366,
+      "grad_norm": 5.021281719207764,
+      "learning_rate": 3.2871026711761666e-06,
+      "loss": 0.508,
+      "step": 817
+    },
+    {
+      "epoch": 3.9902439024390244,
+      "grad_norm": 3.3377649784088135,
+      "learning_rate": 3.2834653762802414e-06,
+      "loss": 0.2116,
+      "step": 818
+    },
+    {
+      "epoch": 3.995121951219512,
+      "grad_norm": 4.412073135375977,
+      "learning_rate": 3.2798262414100594e-06,
+      "loss": 0.2177,
+      "step": 819
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.174323797225952,
+      "learning_rate": 3.2761852751121566e-06,
+      "loss": 0.1737,
+      "step": 820
+    },
+    {
+      "epoch": 4.004878048780488,
+      "grad_norm": 2.921494960784912,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.2569,
+      "step": 821
+    },
+    {
+      "epoch": 4.009756097560976,
+      "grad_norm": 2.693495512008667,
+      "learning_rate": 3.2688978824408136e-06,
+      "loss": 0.1621,
+      "step": 822
+    },
+    {
+      "epoch": 4.014634146341463,
+      "grad_norm": 2.705796718597412,
+      "learning_rate": 3.2652514731818698e-06,
+      "loss": 0.1121,
+      "step": 823
+    },
+    {
+      "epoch": 4.019512195121951,
+      "grad_norm": 3.2621448040008545,
+      "learning_rate": 3.2616032667241564e-06,
+      "loss": 0.0835,
+      "step": 824
+    },
+    {
+      "epoch": 4.024390243902439,
+      "grad_norm": 3.6205084323883057,
+      "learning_rate": 3.257953271635513e-06,
+      "loss": 0.3731,
+      "step": 825
+    },
+    {
+      "epoch": 4.029268292682927,
+      "grad_norm": 3.2600371837615967,
+      "learning_rate": 3.2543014964879814e-06,
+      "loss": 0.1051,
+      "step": 826
+    },
+    {
+      "epoch": 4.034146341463415,
+      "grad_norm": 3.865178346633911,
+      "learning_rate": 3.250647949857781e-06,
+      "loss": 0.0916,
+      "step": 827
+    },
+    {
+      "epoch": 4.0390243902439025,
+      "grad_norm": 6.9700927734375,
+      "learning_rate": 3.2469926403252932e-06,
+      "loss": 0.4037,
+      "step": 828
+    },
+    {
+      "epoch": 4.04390243902439,
+      "grad_norm": 3.658712148666382,
+      "learning_rate": 3.2433355764750417e-06,
+      "loss": 0.0523,
+      "step": 829
+    },
+    {
+      "epoch": 4.048780487804878,
+      "grad_norm": 4.911301612854004,
+      "learning_rate": 3.2396767668956656e-06,
+      "loss": 0.2616,
+      "step": 830
+    },
+    {
+      "epoch": 4.053658536585366,
+      "grad_norm": 5.019360542297363,
+      "learning_rate": 3.2360162201799085e-06,
+      "loss": 0.195,
+      "step": 831
+    },
+    {
+      "epoch": 4.058536585365854,
+      "grad_norm": 3.493767261505127,
+      "learning_rate": 3.2323539449245906e-06,
+      "loss": 0.1245,
+      "step": 832
+    },
+    {
+      "epoch": 4.0634146341463415,
+      "grad_norm": 4.246248722076416,
+      "learning_rate": 3.2286899497305917e-06,
+      "loss": 0.1147,
+      "step": 833
+    },
+    {
+      "epoch": 4.068292682926829,
+      "grad_norm": 2.993704319000244,
+      "learning_rate": 3.2250242432028335e-06,
+      "loss": 0.2189,
+      "step": 834
+    },
+    {
+      "epoch": 4.073170731707317,
+      "grad_norm": 4.695023059844971,
+      "learning_rate": 3.221356833950254e-06,
+      "loss": 0.4685,
+      "step": 835
+    },
+    {
+      "epoch": 4.078048780487805,
+      "grad_norm": 2.777644634246826,
+      "learning_rate": 3.21768773058579e-06,
+      "loss": 0.1245,
+      "step": 836
+    },
+    {
+      "epoch": 4.082926829268293,
+      "grad_norm": 3.3545901775360107,
+      "learning_rate": 3.21401694172636e-06,
+      "loss": 0.1342,
+      "step": 837
+    },
+    {
+      "epoch": 4.087804878048781,
+      "grad_norm": 2.2222652435302734,
+      "learning_rate": 3.2103444759928383e-06,
+      "loss": 0.0484,
+      "step": 838
+    },
+    {
+      "epoch": 4.092682926829268,
+      "grad_norm": 2.580345630645752,
+      "learning_rate": 3.2066703420100377e-06,
+      "loss": 0.0592,
+      "step": 839
+    },
+    {
+      "epoch": 4.097560975609756,
+      "grad_norm": 3.8652923107147217,
+      "learning_rate": 3.2029945484066883e-06,
+      "loss": 0.2536,
+      "step": 840
+    },
+    {
+      "epoch": 4.102439024390244,
+      "grad_norm": 3.0441582202911377,
+      "learning_rate": 3.1993171038154203e-06,
+      "loss": 0.1221,
+      "step": 841
+    },
+    {
+      "epoch": 4.107317073170732,
+      "grad_norm": 2.2795114517211914,
+      "learning_rate": 3.1956380168727385e-06,
+      "loss": 0.1231,
+      "step": 842
+    },
+    {
+      "epoch": 4.11219512195122,
+      "grad_norm": 3.701009750366211,
+      "learning_rate": 3.191957296219007e-06,
+      "loss": 0.2144,
+      "step": 843
+    },
+    {
+      "epoch": 4.117073170731707,
+      "grad_norm": 3.452637195587158,
+      "learning_rate": 3.1882749504984247e-06,
+      "loss": 0.1026,
+      "step": 844
+    },
+    {
+      "epoch": 4.121951219512195,
+      "grad_norm": 2.4208810329437256,
+      "learning_rate": 3.1845909883590076e-06,
+      "loss": 0.1124,
+      "step": 845
+    },
+    {
+      "epoch": 4.126829268292683,
+      "grad_norm": 4.353063583374023,
+      "learning_rate": 3.180905418452569e-06,
+      "loss": 0.2804,
+      "step": 846
+    },
+    {
+      "epoch": 4.131707317073171,
+      "grad_norm": 3.1151084899902344,
+      "learning_rate": 3.1772182494346963e-06,
+      "loss": 0.1748,
+      "step": 847
+    },
+    {
+      "epoch": 4.136585365853659,
+      "grad_norm": 3.457940101623535,
+      "learning_rate": 3.1735294899647344e-06,
+      "loss": 0.1984,
+      "step": 848
+    },
+    {
+      "epoch": 4.1414634146341465,
+      "grad_norm": 3.3556935787200928,
+      "learning_rate": 3.169839148705762e-06,
+      "loss": 0.1332,
+      "step": 849
+    },
+    {
+      "epoch": 4.146341463414634,
+      "grad_norm": 3.5510823726654053,
+      "learning_rate": 3.1661472343245725e-06,
+      "loss": 0.4788,
+      "step": 850
+    },
+    {
+      "epoch": 4.151219512195122,
+      "grad_norm": 4.036712646484375,
+      "learning_rate": 3.162453755491655e-06,
+      "loss": 0.2437,
+      "step": 851
+    },
+    {
+      "epoch": 4.15609756097561,
+      "grad_norm": 4.417062282562256,
+      "learning_rate": 3.158758720881171e-06,
+      "loss": 0.203,
+      "step": 852
+    },
+    {
+      "epoch": 4.160975609756098,
+      "grad_norm": 3.920558214187622,
+      "learning_rate": 3.155062139170937e-06,
+      "loss": 0.1462,
+      "step": 853
+    },
+    {
+      "epoch": 4.1658536585365855,
+      "grad_norm": 6.472081661224365,
+      "learning_rate": 3.1513640190424034e-06,
+      "loss": 0.0972,
+      "step": 854
+    },
+    {
+      "epoch": 4.170731707317073,
+      "grad_norm": 3.975947141647339,
+      "learning_rate": 3.147664369180632e-06,
+      "loss": 0.1092,
+      "step": 855
+    },
+    {
+      "epoch": 4.175609756097561,
+      "grad_norm": 4.977376937866211,
+      "learning_rate": 3.143963198274278e-06,
+      "loss": 0.2215,
+      "step": 856
+    },
+    {
+      "epoch": 4.180487804878049,
+      "grad_norm": 3.595460891723633,
+      "learning_rate": 3.140260515015569e-06,
+      "loss": 0.1771,
+      "step": 857
+    },
+    {
+      "epoch": 4.185365853658537,
+      "grad_norm": 3.1085658073425293,
+      "learning_rate": 3.136556328100284e-06,
+      "loss": 0.1995,
+      "step": 858
+    },
+    {
+      "epoch": 4.190243902439025,
+      "grad_norm": 4.355626583099365,
+      "learning_rate": 3.132850646227734e-06,
+      "loss": 0.4048,
+      "step": 859
+    },
+    {
+      "epoch": 4.195121951219512,
+      "grad_norm": 3.8079614639282227,
+      "learning_rate": 3.12914347810074e-06,
+      "loss": 0.1914,
+      "step": 860
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 3.725804328918457,
+      "learning_rate": 3.125434832425613e-06,
+      "loss": 0.1579,
+      "step": 861
+    },
+    {
+      "epoch": 4.204878048780488,
+      "grad_norm": 2.974649667739868,
+      "learning_rate": 3.121724717912138e-06,
+      "loss": 0.1814,
+      "step": 862
+    },
+    {
+      "epoch": 4.209756097560976,
+      "grad_norm": 3.6391279697418213,
+      "learning_rate": 3.118013143273542e-06,
+      "loss": 0.1481,
+      "step": 863
+    },
+    {
+      "epoch": 4.214634146341464,
+      "grad_norm": 3.216643810272217,
+      "learning_rate": 3.1143001172264893e-06,
+      "loss": 0.113,
+      "step": 864
+    },
+    {
+      "epoch": 4.219512195121951,
+      "grad_norm": 3.605855941772461,
+      "learning_rate": 3.1105856484910474e-06,
+      "loss": 0.1405,
+      "step": 865
+    },
+    {
+      "epoch": 4.224390243902439,
+      "grad_norm": 2.7186765670776367,
+      "learning_rate": 3.1068697457906736e-06,
+      "loss": 0.097,
+      "step": 866
+    },
+    {
+      "epoch": 4.229268292682927,
+      "grad_norm": 3.980973243713379,
+      "learning_rate": 3.1031524178521938e-06,
+      "loss": 0.2207,
+      "step": 867
+    },
+    {
+      "epoch": 4.234146341463415,
+      "grad_norm": 3.4623806476593018,
+      "learning_rate": 3.0994336734057804e-06,
+      "loss": 0.0552,
+      "step": 868
+    },
+    {
+      "epoch": 4.239024390243903,
+      "grad_norm": 3.7556748390197754,
+      "learning_rate": 3.0957135211849315e-06,
+      "loss": 0.1743,
+      "step": 869
+    },
+    {
+      "epoch": 4.2439024390243905,
+      "grad_norm": 3.3547914028167725,
+      "learning_rate": 3.0919919699264535e-06,
+      "loss": 0.1195,
+      "step": 870
+    },
+    {
+      "epoch": 4.248780487804878,
+      "grad_norm": 4.392014503479004,
+      "learning_rate": 3.0882690283704355e-06,
+      "loss": 0.6174,
+      "step": 871
+    },
+    {
+      "epoch": 4.253658536585366,
+      "grad_norm": 2.7031409740448,
+      "learning_rate": 3.084544705260234e-06,
+      "loss": 0.1359,
+      "step": 872
+    },
+    {
+      "epoch": 4.258536585365854,
+      "grad_norm": 2.3518481254577637,
+      "learning_rate": 3.080819009342451e-06,
+      "loss": 0.0786,
+      "step": 873
+    },
+    {
+      "epoch": 4.263414634146342,
+      "grad_norm": 2.636204481124878,
+      "learning_rate": 3.077091949366908e-06,
+      "loss": 0.0677,
+      "step": 874
+    },
+    {
+      "epoch": 4.2682926829268295,
+      "grad_norm": 2.8670942783355713,
+      "learning_rate": 3.073363534086636e-06,
+      "loss": 0.1084,
+      "step": 875
+    },
+    {
+      "epoch": 4.273170731707317,
+      "grad_norm": 2.7044737339019775,
+      "learning_rate": 3.0696337722578444e-06,
+      "loss": 0.0681,
+      "step": 876
+    },
+    {
+      "epoch": 4.278048780487805,
+      "grad_norm": 3.481539487838745,
+      "learning_rate": 3.0659026726399072e-06,
+      "loss": 0.2262,
+      "step": 877
+    },
+    {
+      "epoch": 4.282926829268293,
+      "grad_norm": 3.7746224403381348,
+      "learning_rate": 3.0621702439953393e-06,
+      "loss": 0.2169,
+      "step": 878
+    },
+    {
+      "epoch": 4.287804878048781,
+      "grad_norm": 3.6386263370513916,
+      "learning_rate": 3.0584364950897768e-06,
+      "loss": 0.0581,
+      "step": 879
+    },
+    {
+      "epoch": 4.2926829268292686,
+      "grad_norm": 3.389408588409424,
+      "learning_rate": 3.0547014346919574e-06,
+      "loss": 0.1687,
+      "step": 880
+    },
+    {
+      "epoch": 4.297560975609756,
+      "grad_norm": 3.6510157585144043,
+      "learning_rate": 3.0509650715736977e-06,
+      "loss": 0.1362,
+      "step": 881
+    },
+    {
+      "epoch": 4.302439024390244,
+      "grad_norm": 3.334210157394409,
+      "learning_rate": 3.0472274145098744e-06,
+      "loss": 0.1865,
+      "step": 882
+    },
+    {
+      "epoch": 4.307317073170732,
+      "grad_norm": 4.747341632843018,
+      "learning_rate": 3.0434884722784026e-06,
+      "loss": 0.2385,
+      "step": 883
+    },
+    {
+      "epoch": 4.31219512195122,
+      "grad_norm": 3.9266858100891113,
+      "learning_rate": 3.0397482536602168e-06,
+      "loss": 0.1004,
+      "step": 884
+    },
+    {
+      "epoch": 4.317073170731708,
+      "grad_norm": 2.984821081161499,
+      "learning_rate": 3.0360067674392475e-06,
+      "loss": 0.1469,
+      "step": 885
+    },
+    {
+      "epoch": 4.321951219512195,
+      "grad_norm": 2.6379380226135254,
+      "learning_rate": 3.0322640224024024e-06,
+      "loss": 0.0829,
+      "step": 886
+    },
+    {
+      "epoch": 4.326829268292683,
+      "grad_norm": 3.885495185852051,
+      "learning_rate": 3.0285200273395478e-06,
+      "loss": 0.2256,
+      "step": 887
+    },
+    {
+      "epoch": 4.331707317073171,
+      "grad_norm": 3.950394868850708,
+      "learning_rate": 3.024774791043481e-06,
+      "loss": 0.2402,
+      "step": 888
+    },
+    {
+      "epoch": 4.336585365853659,
+      "grad_norm": 4.147830963134766,
+      "learning_rate": 3.021028322309921e-06,
+      "loss": 0.2198,
+      "step": 889
+    },
+    {
+      "epoch": 4.341463414634147,
+      "grad_norm": 4.0821638107299805,
+      "learning_rate": 3.0172806299374734e-06,
+      "loss": 0.2304,
+      "step": 890
+    },
+    {
+      "epoch": 4.3463414634146345,
+      "grad_norm": 4.142312049865723,
+      "learning_rate": 3.0135317227276247e-06,
+      "loss": 0.2864,
+      "step": 891
+    },
+    {
+      "epoch": 4.351219512195122,
+      "grad_norm": 3.008504867553711,
+      "learning_rate": 3.0097816094847104e-06,
+      "loss": 0.2045,
+      "step": 892
+    },
+    {
+      "epoch": 4.35609756097561,
+      "grad_norm": 3.1674623489379883,
+      "learning_rate": 3.0060302990158984e-06,
+      "loss": 0.0864,
+      "step": 893
+    },
+    {
+      "epoch": 4.360975609756098,
+      "grad_norm": 3.3412492275238037,
+      "learning_rate": 3.002277800131171e-06,
+      "loss": 0.076,
+      "step": 894
+    },
+    {
+      "epoch": 4.365853658536586,
+      "grad_norm": 3.067330837249756,
+      "learning_rate": 2.998524121643298e-06,
+      "loss": 0.1724,
+      "step": 895
+    },
+    {
+      "epoch": 4.3707317073170735,
+      "grad_norm": 3.9015982151031494,
+      "learning_rate": 2.994769272367822e-06,
+      "loss": 0.2,
+      "step": 896
+    },
+    {
+      "epoch": 4.375609756097561,
+      "grad_norm": 3.0136911869049072,
+      "learning_rate": 2.991013261123035e-06,
+      "loss": 0.0852,
+      "step": 897
+    },
+    {
+      "epoch": 4.380487804878049,
+      "grad_norm": 3.6834237575531006,
+      "learning_rate": 2.9872560967299554e-06,
+      "loss": 0.1449,
+      "step": 898
+    },
+    {
+      "epoch": 4.385365853658537,
+      "grad_norm": 3.3486039638519287,
+      "learning_rate": 2.9834977880123132e-06,
+      "loss": 0.0659,
+      "step": 899
+    },
+    {
+      "epoch": 4.390243902439025,
+      "grad_norm": 2.971315622329712,
+      "learning_rate": 2.9797383437965243e-06,
+      "loss": 0.1114,
+      "step": 900
+    },
+    {
+      "epoch": 4.3951219512195125,
+      "grad_norm": 2.683359146118164,
+      "learning_rate": 2.975977772911671e-06,
+      "loss": 0.0822,
+      "step": 901
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 2.9941935539245605,
+      "learning_rate": 2.972216084189482e-06,
+      "loss": 0.0858,
+      "step": 902
+    },
+    {
+      "epoch": 4.404878048780488,
+      "grad_norm": 2.4938626289367676,
+      "learning_rate": 2.9684532864643123e-06,
+      "loss": 0.1162,
+      "step": 903
+    },
+    {
+      "epoch": 4.409756097560976,
+      "grad_norm": 2.9364712238311768,
+      "learning_rate": 2.964689388573118e-06,
+      "loss": 0.0821,
+      "step": 904
+    },
+    {
+      "epoch": 4.414634146341464,
+      "grad_norm": 3.3638134002685547,
+      "learning_rate": 2.9609243993554434e-06,
+      "loss": 0.25,
+      "step": 905
+    },
+    {
+      "epoch": 4.419512195121952,
+      "grad_norm": 3.657277822494507,
+      "learning_rate": 2.9571583276533923e-06,
+      "loss": 0.0852,
+      "step": 906
+    },
+    {
+      "epoch": 4.424390243902439,
+      "grad_norm": 5.486263275146484,
+      "learning_rate": 2.9533911823116124e-06,
+      "loss": 0.5123,
+      "step": 907
+    },
+    {
+      "epoch": 4.429268292682927,
+      "grad_norm": 5.194574356079102,
+      "learning_rate": 2.9496229721772734e-06,
+      "loss": 0.1854,
+      "step": 908
+    },
+    {
+      "epoch": 4.434146341463415,
+      "grad_norm": 3.520110845565796,
+      "learning_rate": 2.9458537061000435e-06,
+      "loss": 0.1785,
+      "step": 909
+    },
+    {
+      "epoch": 4.439024390243903,
+      "grad_norm": 3.417991876602173,
+      "learning_rate": 2.9420833929320726e-06,
+      "loss": 0.1603,
+      "step": 910
+    },
+    {
+      "epoch": 4.443902439024391,
+      "grad_norm": 5.225805282592773,
+      "learning_rate": 2.93831204152797e-06,
+      "loss": 0.3046,
+      "step": 911
+    },
+    {
+      "epoch": 4.4487804878048784,
+      "grad_norm": 3.541433572769165,
+      "learning_rate": 2.9345396607447807e-06,
+      "loss": 0.0631,
+      "step": 912
+    },
+    {
+      "epoch": 4.453658536585366,
+      "grad_norm": 3.909377098083496,
+      "learning_rate": 2.9307662594419704e-06,
+      "loss": 0.125,
+      "step": 913
+    },
+    {
+      "epoch": 4.458536585365854,
+      "grad_norm": 3.6604416370391846,
+      "learning_rate": 2.9269918464814e-06,
+      "loss": 0.156,
+      "step": 914
+    },
+    {
+      "epoch": 4.463414634146342,
+      "grad_norm": 3.7413833141326904,
+      "learning_rate": 2.923216430727306e-06,
+      "loss": 0.3334,
+      "step": 915
+    },
+    {
+      "epoch": 4.46829268292683,
+      "grad_norm": 3.531996011734009,
+      "learning_rate": 2.9194400210462808e-06,
+      "loss": 0.2534,
+      "step": 916
+    },
+    {
+      "epoch": 4.473170731707317,
+      "grad_norm": 4.163621425628662,
+      "learning_rate": 2.91566262630725e-06,
+      "loss": 0.352,
+      "step": 917
+    },
+    {
+      "epoch": 4.478048780487805,
+      "grad_norm": 3.923635482788086,
+      "learning_rate": 2.9118842553814526e-06,
+      "loss": 0.1132,
+      "step": 918
+    },
+    {
+      "epoch": 4.482926829268292,
+      "grad_norm": 2.833768844604492,
+      "learning_rate": 2.9081049171424223e-06,
+      "loss": 0.086,
+      "step": 919
+    },
+    {
+      "epoch": 4.487804878048781,
+      "grad_norm": 2.9006292819976807,
+      "learning_rate": 2.9043246204659624e-06,
+      "loss": 0.0693,
+      "step": 920
+    },
+    {
+      "epoch": 4.492682926829268,
+      "grad_norm": 3.699376344680786,
+      "learning_rate": 2.9005433742301274e-06,
+      "loss": 0.2463,
+      "step": 921
+    },
+    {
+      "epoch": 4.4975609756097565,
+      "grad_norm": 4.882141590118408,
+      "learning_rate": 2.8967611873152037e-06,
+      "loss": 0.2275,
+      "step": 922
+    },
+    {
+      "epoch": 4.5024390243902435,
+      "grad_norm": 3.0554678440093994,
+      "learning_rate": 2.892978068603683e-06,
+      "loss": 0.0752,
+      "step": 923
+    },
+    {
+      "epoch": 4.507317073170732,
+      "grad_norm": 3.1225268840789795,
+      "learning_rate": 2.889194026980249e-06,
+      "loss": 0.1649,
+      "step": 924
+    },
+    {
+      "epoch": 4.512195121951219,
+      "grad_norm": 17.75234031677246,
+      "learning_rate": 2.8854090713317514e-06,
+      "loss": 0.0437,
+      "step": 925
+    },
+    {
+      "epoch": 4.517073170731708,
+      "grad_norm": 3.011223554611206,
+      "learning_rate": 2.8816232105471864e-06,
+      "loss": 0.0747,
+      "step": 926
+    },
+    {
+      "epoch": 4.521951219512195,
+      "grad_norm": 4.327573299407959,
+      "learning_rate": 2.877836453517677e-06,
+      "loss": 0.3884,
+      "step": 927
+    },
+    {
+      "epoch": 4.526829268292683,
+      "grad_norm": 3.8694965839385986,
+      "learning_rate": 2.8740488091364492e-06,
+      "loss": 0.2741,
+      "step": 928
+    },
+    {
+      "epoch": 4.53170731707317,
+      "grad_norm": 5.375877380371094,
+      "learning_rate": 2.870260286298814e-06,
+      "loss": 0.364,
+      "step": 929
+    },
+    {
+      "epoch": 4.536585365853659,
+      "grad_norm": 3.380891799926758,
+      "learning_rate": 2.866470893902147e-06,
+      "loss": 0.1495,
+      "step": 930
+    },
+    {
+      "epoch": 4.541463414634146,
+      "grad_norm": 3.723992109298706,
+      "learning_rate": 2.8626806408458626e-06,
+      "loss": 0.1403,
+      "step": 931
+    },
+    {
+      "epoch": 4.546341463414635,
+      "grad_norm": 3.0534417629241943,
+      "learning_rate": 2.8588895360313983e-06,
+      "loss": 0.0946,
+      "step": 932
+    },
+    {
+      "epoch": 4.5512195121951216,
+      "grad_norm": 2.8875234127044678,
+      "learning_rate": 2.8550975883621935e-06,
+      "loss": 0.1851,
+      "step": 933
+    },
+    {
+      "epoch": 4.55609756097561,
+      "grad_norm": 3.532166004180908,
+      "learning_rate": 2.8513048067436644e-06,
+      "loss": 0.178,
+      "step": 934
+    },
+    {
+      "epoch": 4.560975609756097,
+      "grad_norm": 2.942798376083374,
+      "learning_rate": 2.847511200083187e-06,
+      "loss": 0.1131,
+      "step": 935
+    },
+    {
+      "epoch": 4.565853658536585,
+      "grad_norm": 2.926874876022339,
+      "learning_rate": 2.843716777290074e-06,
+      "loss": 0.1251,
+      "step": 936
+    },
+    {
+      "epoch": 4.570731707317073,
+      "grad_norm": 3.525895357131958,
+      "learning_rate": 2.839921547275556e-06,
+      "loss": 0.0946,
+      "step": 937
+    },
+    {
+      "epoch": 4.575609756097561,
+      "grad_norm": 3.7033681869506836,
+      "learning_rate": 2.836125518952759e-06,
+      "loss": 0.1529,
+      "step": 938
+    },
+    {
+      "epoch": 4.580487804878048,
+      "grad_norm": 3.235154867172241,
+      "learning_rate": 2.8323287012366845e-06,
+      "loss": 0.2511,
+      "step": 939
+    },
+    {
+      "epoch": 4.585365853658536,
+      "grad_norm": 3.5275583267211914,
+      "learning_rate": 2.828531103044186e-06,
+      "loss": 0.1474,
+      "step": 940
+    },
+    {
+      "epoch": 4.590243902439024,
+      "grad_norm": 3.1356353759765625,
+      "learning_rate": 2.8247327332939512e-06,
+      "loss": 0.2249,
+      "step": 941
+    },
+    {
+      "epoch": 4.595121951219512,
+      "grad_norm": 3.789210081100464,
+      "learning_rate": 2.82093360090648e-06,
+      "loss": 0.2258,
+      "step": 942
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 4.841623306274414,
+      "learning_rate": 2.8171337148040636e-06,
+      "loss": 0.2235,
+      "step": 943
+    },
+    {
+      "epoch": 4.6048780487804875,
+      "grad_norm": 3.161630630493164,
+      "learning_rate": 2.813333083910761e-06,
+      "loss": 0.1562,
+      "step": 944
+    },
+    {
+      "epoch": 4.609756097560975,
+      "grad_norm": 2.8718132972717285,
+      "learning_rate": 2.8095317171523835e-06,
+      "loss": 0.0625,
+      "step": 945
+    },
+    {
+      "epoch": 4.614634146341463,
+      "grad_norm": 3.6432454586029053,
+      "learning_rate": 2.805729623456469e-06,
+      "loss": 0.2205,
+      "step": 946
+    },
+    {
+      "epoch": 4.619512195121951,
+      "grad_norm": 4.382034778594971,
+      "learning_rate": 2.8019268117522624e-06,
+      "loss": 0.3241,
+      "step": 947
+    },
+    {
+      "epoch": 4.624390243902439,
+      "grad_norm": 3.2998175621032715,
+      "learning_rate": 2.798123290970695e-06,
+      "loss": 0.1983,
+      "step": 948
+    },
+    {
+      "epoch": 4.6292682926829265,
+      "grad_norm": 3.8665990829467773,
+      "learning_rate": 2.794319070044365e-06,
+      "loss": 0.3391,
+      "step": 949
+    },
+    {
+      "epoch": 4.634146341463414,
+      "grad_norm": 3.628403425216675,
+      "learning_rate": 2.790514157907512e-06,
+      "loss": 0.1329,
+      "step": 950
+    },
+    {
+      "epoch": 4.639024390243902,
+      "grad_norm": 2.8889615535736084,
+      "learning_rate": 2.786708563496002e-06,
+      "loss": 0.141,
+      "step": 951
+    },
+    {
+      "epoch": 4.64390243902439,
+      "grad_norm": 4.07351541519165,
+      "learning_rate": 2.782902295747299e-06,
+      "loss": 0.2935,
+      "step": 952
+    },
+    {
+      "epoch": 4.648780487804878,
+      "grad_norm": 4.220067024230957,
+      "learning_rate": 2.7790953636004536e-06,
+      "loss": 0.318,
+      "step": 953
+    },
+    {
+      "epoch": 4.6536585365853655,
+      "grad_norm": 3.8444325923919678,
+      "learning_rate": 2.775287775996074e-06,
+      "loss": 0.3388,
+      "step": 954
+    },
+    {
+      "epoch": 4.658536585365853,
+      "grad_norm": 3.197313070297241,
+      "learning_rate": 2.7714795418763067e-06,
+      "loss": 0.0925,
+      "step": 955
+    },
+    {
+      "epoch": 4.663414634146341,
+      "grad_norm": 4.0050811767578125,
+      "learning_rate": 2.7676706701848187e-06,
+      "loss": 0.2811,
+      "step": 956
+    },
+    {
+      "epoch": 4.668292682926829,
+      "grad_norm": 3.217160224914551,
+      "learning_rate": 2.763861169866774e-06,
+      "loss": 0.311,
+      "step": 957
+    },
+    {
+      "epoch": 4.673170731707317,
+      "grad_norm": 2.9892494678497314,
+      "learning_rate": 2.7600510498688104e-06,
+      "loss": 0.0582,
+      "step": 958
+    },
+    {
+      "epoch": 4.678048780487805,
+      "grad_norm": 3.954805374145508,
+      "learning_rate": 2.7562403191390246e-06,
+      "loss": 0.1238,
+      "step": 959
+    },
+    {
+      "epoch": 4.682926829268292,
+      "grad_norm": 2.9582695960998535,
+      "learning_rate": 2.7524289866269467e-06,
+      "loss": 0.1243,
+      "step": 960
+    },
+    {
+      "epoch": 4.68780487804878,
+      "grad_norm": 2.807002544403076,
+      "learning_rate": 2.748617061283518e-06,
+      "loss": 0.1388,
+      "step": 961
+    },
+    {
+      "epoch": 4.692682926829268,
+      "grad_norm": 3.980499505996704,
+      "learning_rate": 2.744804552061074e-06,
+      "loss": 0.1144,
+      "step": 962
+    },
+    {
+      "epoch": 4.697560975609756,
+      "grad_norm": 3.6389007568359375,
+      "learning_rate": 2.740991467913321e-06,
+      "loss": 0.2155,
+      "step": 963
+    },
+    {
+      "epoch": 4.702439024390244,
+      "grad_norm": 3.0950801372528076,
+      "learning_rate": 2.737177817795315e-06,
+      "loss": 0.0983,
+      "step": 964
+    },
+    {
+      "epoch": 4.7073170731707314,
+      "grad_norm": 3.1723053455352783,
+      "learning_rate": 2.7333636106634414e-06,
+      "loss": 0.1365,
+      "step": 965
+    },
+    {
+      "epoch": 4.712195121951219,
+      "grad_norm": 3.83921217918396,
+      "learning_rate": 2.7295488554753957e-06,
+      "loss": 0.1977,
+      "step": 966
+    },
+    {
+      "epoch": 4.717073170731707,
+      "grad_norm": 3.348057746887207,
+      "learning_rate": 2.725733561190157e-06,
+      "loss": 0.1311,
+      "step": 967
+    },
+    {
+      "epoch": 4.721951219512195,
+      "grad_norm": 3.828483819961548,
+      "learning_rate": 2.721917736767973e-06,
+      "loss": 0.2464,
+      "step": 968
+    },
+    {
+      "epoch": 4.726829268292683,
+      "grad_norm": 2.6004624366760254,
+      "learning_rate": 2.7181013911703357e-06,
+      "loss": 0.1088,
+      "step": 969
+    },
+    {
+      "epoch": 4.7317073170731705,
+      "grad_norm": 3.316990852355957,
+      "learning_rate": 2.714284533359961e-06,
+      "loss": 0.1492,
+      "step": 970
+    },
+    {
+      "epoch": 4.736585365853658,
+      "grad_norm": 3.8770010471343994,
+      "learning_rate": 2.710467172300768e-06,
+      "loss": 0.218,
+      "step": 971
+    },
+    {
+      "epoch": 4.741463414634146,
+      "grad_norm": 4.456376552581787,
+      "learning_rate": 2.706649316957857e-06,
+      "loss": 0.2199,
+      "step": 972
+    },
+    {
+      "epoch": 4.746341463414634,
+      "grad_norm": 3.3376309871673584,
+      "learning_rate": 2.7028309762974897e-06,
+      "loss": 0.0595,
+      "step": 973
+    },
+    {
+      "epoch": 4.751219512195122,
+      "grad_norm": 3.6755495071411133,
+      "learning_rate": 2.699012159287069e-06,
+      "loss": 0.1653,
+      "step": 974
+    },
+    {
+      "epoch": 4.7560975609756095,
+      "grad_norm": 2.939887046813965,
+      "learning_rate": 2.6951928748951125e-06,
+      "loss": 0.0681,
+      "step": 975
+    },
+    {
+      "epoch": 4.760975609756097,
+      "grad_norm": 3.4101195335388184,
+      "learning_rate": 2.69137313209124e-06,
+      "loss": 0.2046,
+      "step": 976
+    },
+    {
+      "epoch": 4.765853658536585,
+      "grad_norm": 3.9811208248138428,
+      "learning_rate": 2.687552939846145e-06,
+      "loss": 0.2255,
+      "step": 977
+    },
+    {
+      "epoch": 4.770731707317073,
+      "grad_norm": 3.484255313873291,
+      "learning_rate": 2.6837323071315766e-06,
+      "loss": 0.0512,
+      "step": 978
+    },
+    {
+      "epoch": 4.775609756097561,
+      "grad_norm": 3.9005143642425537,
+      "learning_rate": 2.679911242920321e-06,
+      "loss": 0.162,
+      "step": 979
+    },
+    {
+      "epoch": 4.780487804878049,
+      "grad_norm": 4.933374881744385,
+      "learning_rate": 2.6760897561861742e-06,
+      "loss": 0.398,
+      "step": 980
+    },
+    {
+      "epoch": 4.785365853658536,
+      "grad_norm": 3.0741539001464844,
+      "learning_rate": 2.672267855903927e-06,
+      "loss": 0.0507,
+      "step": 981
+    },
+    {
+      "epoch": 4.790243902439024,
+      "grad_norm": 3.023772716522217,
+      "learning_rate": 2.6684455510493413e-06,
+      "loss": 0.2066,
+      "step": 982
+    },
+    {
+      "epoch": 4.795121951219512,
+      "grad_norm": 3.0102407932281494,
+      "learning_rate": 2.6646228505991267e-06,
+      "loss": 0.2296,
+      "step": 983
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 3.902200222015381,
+      "learning_rate": 2.6607997635309246e-06,
+      "loss": 0.14,
+      "step": 984
+    },
+    {
+      "epoch": 4.804878048780488,
+      "grad_norm": 3.836185932159424,
+      "learning_rate": 2.6569762988232838e-06,
+      "loss": 0.1583,
+      "step": 985
+    },
+    {
+      "epoch": 4.809756097560975,
+      "grad_norm": 3.539628744125366,
+      "learning_rate": 2.653152465455639e-06,
+      "loss": 0.2619,
+      "step": 986
+    },
+    {
+      "epoch": 4.814634146341463,
+      "grad_norm": 4.716914653778076,
+      "learning_rate": 2.6493282724082913e-06,
+      "loss": 0.3029,
+      "step": 987
+    },
+    {
+      "epoch": 4.819512195121951,
+      "grad_norm": 3.466914176940918,
+      "learning_rate": 2.6455037286623864e-06,
+      "loss": 0.095,
+      "step": 988
+    },
+    {
+      "epoch": 4.824390243902439,
+      "grad_norm": 2.1798667907714844,
+      "learning_rate": 2.6416788431998935e-06,
+      "loss": 0.1232,
+      "step": 989
+    },
+    {
+      "epoch": 4.829268292682927,
+      "grad_norm": 3.309039354324341,
+      "learning_rate": 2.637853625003585e-06,
+      "loss": 0.3671,
+      "step": 990
+    },
+    {
+      "epoch": 4.8341463414634145,
+      "grad_norm": 3.2619435787200928,
+      "learning_rate": 2.6340280830570142e-06,
+      "loss": 0.194,
+      "step": 991
+    },
+    {
+      "epoch": 4.839024390243902,
+      "grad_norm": 3.601161003112793,
+      "learning_rate": 2.6302022263444947e-06,
+      "loss": 0.1214,
+      "step": 992
+    },
+    {
+      "epoch": 4.84390243902439,
+      "grad_norm": 4.13787841796875,
+      "learning_rate": 2.6263760638510793e-06,
+      "loss": 0.311,
+      "step": 993
+    },
+    {
+      "epoch": 4.848780487804878,
+      "grad_norm": 3.0474166870117188,
+      "learning_rate": 2.6225496045625394e-06,
+      "loss": 0.1853,
+      "step": 994
+    },
+    {
+      "epoch": 4.853658536585366,
+      "grad_norm": 4.481237411499023,
+      "learning_rate": 2.6187228574653428e-06,
+      "loss": 0.2088,
+      "step": 995
+    },
+    {
+      "epoch": 4.8585365853658535,
+      "grad_norm": 3.235966444015503,
+      "learning_rate": 2.614895831546633e-06,
+      "loss": 0.1439,
+      "step": 996
+    },
+    {
+      "epoch": 4.863414634146341,
+      "grad_norm": 4.103270053863525,
+      "learning_rate": 2.6110685357942096e-06,
+      "loss": 0.2823,
+      "step": 997
+    },
+    {
+      "epoch": 4.868292682926829,
+      "grad_norm": 4.134536266326904,
+      "learning_rate": 2.6072409791965048e-06,
+      "loss": 0.2963,
+      "step": 998
+    },
+    {
+      "epoch": 4.873170731707317,
+      "grad_norm": 4.124892711639404,
+      "learning_rate": 2.6034131707425638e-06,
+      "loss": 0.4127,
+      "step": 999
+    },
+    {
+      "epoch": 4.878048780487805,
+      "grad_norm": 3.565139055252075,
+      "learning_rate": 2.5995851194220223e-06,
+      "loss": 0.1601,
+      "step": 1000
+    },
+    {
+      "epoch": 4.882926829268293,
+      "grad_norm": 2.7548017501831055,
+      "learning_rate": 2.595756834225089e-06,
+      "loss": 0.161,
+      "step": 1001
+    },
+    {
+      "epoch": 4.88780487804878,
+      "grad_norm": 3.9297611713409424,
+      "learning_rate": 2.5919283241425188e-06,
+      "loss": 0.1013,
+      "step": 1002
+    },
+    {
+      "epoch": 4.892682926829268,
+      "grad_norm": 2.4904236793518066,
+      "learning_rate": 2.5880995981655965e-06,
+      "loss": 0.1177,
+      "step": 1003
+    },
+    {
+      "epoch": 4.897560975609756,
+      "grad_norm": 3.513308048248291,
+      "learning_rate": 2.584270665286113e-06,
+      "loss": 0.0682,
+      "step": 1004
+    },
+    {
+      "epoch": 4.902439024390244,
+      "grad_norm": 4.221067428588867,
+      "learning_rate": 2.580441534496346e-06,
+      "loss": 0.1502,
+      "step": 1005
+    },
+    {
+      "epoch": 4.907317073170732,
+      "grad_norm": 3.4298903942108154,
+      "learning_rate": 2.576612214789039e-06,
+      "loss": 0.1772,
+      "step": 1006
+    },
+    {
+      "epoch": 4.912195121951219,
+      "grad_norm": 4.402887344360352,
+      "learning_rate": 2.5727827151573747e-06,
+      "loss": 0.2029,
+      "step": 1007
+    },
+    {
+      "epoch": 4.917073170731707,
+      "grad_norm": 4.194999694824219,
+      "learning_rate": 2.568953044594964e-06,
+      "loss": 0.1269,
+      "step": 1008
+    },
+    {
+      "epoch": 4.921951219512195,
+      "grad_norm": 3.657607078552246,
+      "learning_rate": 2.5651232120958157e-06,
+      "loss": 0.1311,
+      "step": 1009
+    },
+    {
+      "epoch": 4.926829268292683,
+      "grad_norm": 4.092184543609619,
+      "learning_rate": 2.56129322665432e-06,
+      "loss": 0.1085,
+      "step": 1010
+    },
+    {
+      "epoch": 4.931707317073171,
+      "grad_norm": 3.3648242950439453,
+      "learning_rate": 2.5574630972652263e-06,
+      "loss": 0.0782,
+      "step": 1011
+    },
+    {
+      "epoch": 4.9365853658536585,
+      "grad_norm": 3.7215166091918945,
+      "learning_rate": 2.553632832923622e-06,
+      "loss": 0.1391,
+      "step": 1012
+    },
+    {
+      "epoch": 4.941463414634146,
+      "grad_norm": 4.045740127563477,
+      "learning_rate": 2.5498024426249107e-06,
+      "loss": 0.3141,
+      "step": 1013
+    },
+    {
+      "epoch": 4.946341463414634,
+      "grad_norm": 3.2363107204437256,
+      "learning_rate": 2.545971935364794e-06,
+      "loss": 0.0679,
+      "step": 1014
+    },
+    {
+      "epoch": 4.951219512195122,
+      "grad_norm": 3.057283639907837,
+      "learning_rate": 2.5421413201392443e-06,
+      "loss": 0.1382,
+      "step": 1015
+    },
+    {
+      "epoch": 4.95609756097561,
+      "grad_norm": 3.591535806655884,
+      "learning_rate": 2.538310605944491e-06,
+      "loss": 0.112,
+      "step": 1016
+    },
+    {
+      "epoch": 4.9609756097560975,
+      "grad_norm": 3.1629281044006348,
+      "learning_rate": 2.534479801776996e-06,
+      "loss": 0.1261,
+      "step": 1017
+    },
+    {
+      "epoch": 4.965853658536585,
+      "grad_norm": 2.691740036010742,
+      "learning_rate": 2.53064891663343e-06,
+      "loss": 0.2328,
+      "step": 1018
+    },
+    {
+      "epoch": 4.970731707317073,
+      "grad_norm": 3.2620503902435303,
+      "learning_rate": 2.526817959510655e-06,
+      "loss": 0.193,
+      "step": 1019
+    },
+    {
+      "epoch": 4.975609756097561,
+      "grad_norm": 3.0721535682678223,
+      "learning_rate": 2.5229869394057038e-06,
+      "loss": 0.2444,
+      "step": 1020
+    },
+    {
+      "epoch": 4.980487804878049,
+      "grad_norm": 2.6279208660125732,
+      "learning_rate": 2.5191558653157542e-06,
+      "loss": 0.1103,
+      "step": 1021
+    },
+    {
+      "epoch": 4.985365853658537,
+      "grad_norm": 2.9295670986175537,
+      "learning_rate": 2.515324746238113e-06,
+      "loss": 0.0553,
+      "step": 1022
+    },
+    {
+      "epoch": 4.990243902439024,
+      "grad_norm": 3.3960084915161133,
+      "learning_rate": 2.511493591170191e-06,
+      "loss": 0.1686,
+      "step": 1023
+    },
+    {
+      "epoch": 4.995121951219512,
+      "grad_norm": 4.138705253601074,
+      "learning_rate": 2.5076624091094846e-06,
+      "loss": 0.1208,
+      "step": 1024
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 2.603870391845703,
+      "learning_rate": 2.503831209053554e-06,
+      "loss": 0.1216,
+      "step": 1025
+    },
+    {
+      "epoch": 5.004878048780488,
+      "grad_norm": 2.525205612182617,
+      "learning_rate": 2.5e-06,
+      "loss": 0.0984,
+      "step": 1026
+    },
+    {
+      "epoch": 5.009756097560976,
+      "grad_norm": 3.2502501010894775,
+      "learning_rate": 2.4961687909464462e-06,
+      "loss": 0.1323,
+      "step": 1027
+    },
+    {
+      "epoch": 5.014634146341463,
+      "grad_norm": 5.363409519195557,
+      "learning_rate": 2.492337590890516e-06,
+      "loss": 0.3516,
+      "step": 1028
+    },
+    {
+      "epoch": 5.019512195121951,
+      "grad_norm": 2.887723445892334,
+      "learning_rate": 2.4885064088298097e-06,
+      "loss": 0.1931,
+      "step": 1029
+    },
+    {
+      "epoch": 5.024390243902439,
+      "grad_norm": 3.4529435634613037,
+      "learning_rate": 2.4846752537618875e-06,
+      "loss": 0.0675,
+      "step": 1030
+    },
+    {
+      "epoch": 5.029268292682927,
+      "grad_norm": 4.202361106872559,
+      "learning_rate": 2.480844134684246e-06,
+      "loss": 0.1643,
+      "step": 1031
+    },
+    {
+      "epoch": 5.034146341463415,
+      "grad_norm": 2.910275459289551,
+      "learning_rate": 2.4770130605942966e-06,
+      "loss": 0.11,
+      "step": 1032
+    },
+    {
+      "epoch": 5.0390243902439025,
+      "grad_norm": 3.5430362224578857,
+      "learning_rate": 2.4731820404893457e-06,
+      "loss": 0.0614,
+      "step": 1033
+    },
+    {
+      "epoch": 5.04390243902439,
+      "grad_norm": 4.501879692077637,
+      "learning_rate": 2.469351083366571e-06,
+      "loss": 0.0954,
+      "step": 1034
+    },
+    {
+      "epoch": 5.048780487804878,
+      "grad_norm": 2.732261896133423,
+      "learning_rate": 2.4655201982230044e-06,
+      "loss": 0.0275,
+      "step": 1035
+    },
+    {
+      "epoch": 5.053658536585366,
+      "grad_norm": 3.5926437377929688,
+      "learning_rate": 2.4616893940555094e-06,
+      "loss": 0.0661,
+      "step": 1036
+    },
+    {
+      "epoch": 5.058536585365854,
+      "grad_norm": 4.790312767028809,
+      "learning_rate": 2.457858679860757e-06,
+      "loss": 0.2976,
+      "step": 1037
+    },
+    {
+      "epoch": 5.0634146341463415,
+      "grad_norm": 4.453246116638184,
+      "learning_rate": 2.4540280646352072e-06,
+      "loss": 0.1216,
+      "step": 1038
+    },
+    {
+      "epoch": 5.068292682926829,
+      "grad_norm": 3.288011074066162,
+      "learning_rate": 2.45019755737509e-06,
+      "loss": 0.0877,
+      "step": 1039
+    },
+    {
+      "epoch": 5.073170731707317,
+      "grad_norm": 3.566927671432495,
+      "learning_rate": 2.4463671670763787e-06,
+      "loss": 0.1661,
+      "step": 1040
+    },
+    {
+      "epoch": 5.078048780487805,
+      "grad_norm": 3.250047206878662,
+      "learning_rate": 2.4425369027347746e-06,
+      "loss": 0.211,
+      "step": 1041
+    },
+    {
+      "epoch": 5.082926829268293,
+      "grad_norm": 3.0214977264404297,
+      "learning_rate": 2.4387067733456804e-06,
+      "loss": 0.093,
+      "step": 1042
+    },
+    {
+      "epoch": 5.087804878048781,
+      "grad_norm": 3.8162097930908203,
+      "learning_rate": 2.4348767879041847e-06,
+      "loss": 0.0777,
+      "step": 1043
+    },
+    {
+      "epoch": 5.092682926829268,
+      "grad_norm": 3.8071560859680176,
+      "learning_rate": 2.4310469554050366e-06,
+      "loss": 0.087,
+      "step": 1044
+    },
+    {
+      "epoch": 5.097560975609756,
+      "grad_norm": 3.1032073497772217,
+      "learning_rate": 2.4272172848426257e-06,
+      "loss": 0.1105,
+      "step": 1045
+    },
+    {
+      "epoch": 5.102439024390244,
+      "grad_norm": 2.8980185985565186,
+      "learning_rate": 2.423387785210962e-06,
+      "loss": 0.0704,
+      "step": 1046
+    },
+    {
+      "epoch": 5.107317073170732,
+      "grad_norm": 3.9110755920410156,
+      "learning_rate": 2.4195584655036544e-06,
+      "loss": 0.2118,
+      "step": 1047
+    },
+    {
+      "epoch": 5.11219512195122,
+      "grad_norm": 2.678884506225586,
+      "learning_rate": 2.4157293347138877e-06,
+      "loss": 0.0664,
+      "step": 1048
+    },
+    {
+      "epoch": 5.117073170731707,
+      "grad_norm": 3.183046340942383,
+      "learning_rate": 2.4119004018344043e-06,
+      "loss": 0.1767,
+      "step": 1049
+    },
+    {
+      "epoch": 5.121951219512195,
+      "grad_norm": 3.9198925495147705,
+      "learning_rate": 2.408071675857482e-06,
+      "loss": 0.1288,
+      "step": 1050
+    },
+    {
+      "epoch": 5.126829268292683,
+      "grad_norm": 4.378621578216553,
+      "learning_rate": 2.404243165774912e-06,
+      "loss": 0.1724,
+      "step": 1051
+    },
+    {
+      "epoch": 5.131707317073171,
+      "grad_norm": 2.5509133338928223,
+      "learning_rate": 2.4004148805779785e-06,
+      "loss": 0.0382,
+      "step": 1052
+    },
+    {
+      "epoch": 5.136585365853659,
+      "grad_norm": 3.692396402359009,
+      "learning_rate": 2.3965868292574375e-06,
+      "loss": 0.0942,
+      "step": 1053
+    },
+    {
+      "epoch": 5.1414634146341465,
+      "grad_norm": 3.8537800312042236,
+      "learning_rate": 2.392759020803496e-06,
+      "loss": 0.0819,
+      "step": 1054
+    },
+    {
+      "epoch": 5.146341463414634,
+      "grad_norm": 4.02876091003418,
+      "learning_rate": 2.3889314642057916e-06,
+      "loss": 0.0866,
+      "step": 1055
+    },
+    {
+      "epoch": 5.151219512195122,
+      "grad_norm": 3.531857490539551,
+      "learning_rate": 2.3851041684533677e-06,
+      "loss": 0.1557,
+      "step": 1056
+    },
+    {
+      "epoch": 5.15609756097561,
+      "grad_norm": 2.231265068054199,
+      "learning_rate": 2.381277142534658e-06,
+      "loss": 0.0421,
+      "step": 1057
+    },
+    {
+      "epoch": 5.160975609756098,
+      "grad_norm": 3.159226894378662,
+      "learning_rate": 2.3774503954374614e-06,
+      "loss": 0.0395,
+      "step": 1058
+    },
+    {
+      "epoch": 5.1658536585365855,
+      "grad_norm": 3.0375123023986816,
+      "learning_rate": 2.373623936148921e-06,
+      "loss": 0.1869,
+      "step": 1059
+    },
+    {
+      "epoch": 5.170731707317073,
+      "grad_norm": 5.4905900955200195,
+      "learning_rate": 2.369797773655506e-06,
+      "loss": 0.1426,
+      "step": 1060
+    },
+    {
+      "epoch": 5.175609756097561,
+      "grad_norm": 2.8739638328552246,
+      "learning_rate": 2.3659719169429866e-06,
+      "loss": 0.0788,
+      "step": 1061
+    },
+    {
+      "epoch": 5.180487804878049,
+      "grad_norm": 2.612183094024658,
+      "learning_rate": 2.3621463749964153e-06,
+      "loss": 0.0449,
+      "step": 1062
+    },
+    {
+      "epoch": 5.185365853658537,
+      "grad_norm": 2.0573198795318604,
+      "learning_rate": 2.3583211568001073e-06,
+      "loss": 0.0264,
+      "step": 1063
+    },
+    {
+      "epoch": 5.190243902439025,
+      "grad_norm": 2.3667244911193848,
+      "learning_rate": 2.3544962713376144e-06,
+      "loss": 0.0507,
+      "step": 1064
+    },
+    {
+      "epoch": 5.195121951219512,
+      "grad_norm": 2.1223740577697754,
+      "learning_rate": 2.3506717275917095e-06,
+      "loss": 0.0576,
+      "step": 1065
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 2.2630319595336914,
+      "learning_rate": 2.346847534544362e-06,
+      "loss": 0.0523,
+      "step": 1066
+    },
+    {
+      "epoch": 5.204878048780488,
+      "grad_norm": 3.201913595199585,
+      "learning_rate": 2.3430237011767166e-06,
+      "loss": 0.0847,
+      "step": 1067
+    },
+    {
+      "epoch": 5.209756097560976,
+      "grad_norm": 2.2149481773376465,
+      "learning_rate": 2.3392002364690762e-06,
+      "loss": 0.0215,
+      "step": 1068
+    },
+    {
+      "epoch": 5.214634146341464,
+      "grad_norm": 4.425244331359863,
+      "learning_rate": 2.335377149400874e-06,
+      "loss": 0.1018,
+      "step": 1069
+    },
+    {
+      "epoch": 5.219512195121951,
+      "grad_norm": 4.548358917236328,
+      "learning_rate": 2.3315544489506596e-06,
+      "loss": 0.1485,
+      "step": 1070
+    },
+    {
+      "epoch": 5.224390243902439,
+      "grad_norm": 3.635796546936035,
+      "learning_rate": 2.3277321440960733e-06,
+      "loss": 0.111,
+      "step": 1071
+    },
+    {
+      "epoch": 5.229268292682927,
+      "grad_norm": 2.3180043697357178,
+      "learning_rate": 2.323910243813826e-06,
+      "loss": 0.0267,
+      "step": 1072
+    },
+    {
+      "epoch": 5.234146341463415,
+      "grad_norm": 3.675490379333496,
+      "learning_rate": 2.3200887570796798e-06,
+      "loss": 0.153,
+      "step": 1073
+    },
+    {
+      "epoch": 5.239024390243903,
+      "grad_norm": 2.883225202560425,
+      "learning_rate": 2.316267692868424e-06,
+      "loss": 0.0968,
+      "step": 1074
+    },
+    {
+      "epoch": 5.2439024390243905,
+      "grad_norm": 3.0320188999176025,
+      "learning_rate": 2.312447060153856e-06,
+      "loss": 0.0786,
+      "step": 1075
+    },
+    {
+      "epoch": 5.248780487804878,
+      "grad_norm": 2.682695150375366,
+      "learning_rate": 2.308626867908761e-06,
+      "loss": 0.0677,
+      "step": 1076
+    },
+    {
+      "epoch": 5.253658536585366,
+      "grad_norm": 3.941967010498047,
+      "learning_rate": 2.3048071251048884e-06,
+      "loss": 0.1059,
+      "step": 1077
+    },
+    {
+      "epoch": 5.258536585365854,
+      "grad_norm": 6.485599517822266,
+      "learning_rate": 2.300987840712932e-06,
+      "loss": 0.1331,
+      "step": 1078
+    },
+    {
+      "epoch": 5.263414634146342,
+      "grad_norm": 3.809269905090332,
+      "learning_rate": 2.297169023702511e-06,
+      "loss": 0.169,
+      "step": 1079
+    },
+    {
+      "epoch": 5.2682926829268295,
+      "grad_norm": 3.115626573562622,
+      "learning_rate": 2.2933506830421436e-06,
+      "loss": 0.1349,
+      "step": 1080
+    },
+    {
+      "epoch": 5.273170731707317,
+      "grad_norm": 2.2234909534454346,
+      "learning_rate": 2.2895328276992325e-06,
+      "loss": 0.0191,
+      "step": 1081
+    },
+    {
+      "epoch": 5.278048780487805,
+      "grad_norm": 3.896925926208496,
+      "learning_rate": 2.28571546664004e-06,
+      "loss": 0.1961,
+      "step": 1082
+    },
+    {
+      "epoch": 5.282926829268293,
+      "grad_norm": 2.4134509563446045,
+      "learning_rate": 2.281898608829665e-06,
+      "loss": 0.02,
+      "step": 1083
+    },
+    {
+      "epoch": 5.287804878048781,
+      "grad_norm": 2.7599191665649414,
+      "learning_rate": 2.2780822632320273e-06,
+      "loss": 0.0763,
+      "step": 1084
+    },
+    {
+      "epoch": 5.2926829268292686,
+      "grad_norm": 2.465637683868408,
+      "learning_rate": 2.2742664388098435e-06,
+      "loss": 0.0403,
+      "step": 1085
+    },
+    {
+      "epoch": 5.297560975609756,
+      "grad_norm": 2.4026618003845215,
+      "learning_rate": 2.270451144524605e-06,
+      "loss": 0.0982,
+      "step": 1086
+    },
+    {
+      "epoch": 5.302439024390244,
+      "grad_norm": 3.3339459896087646,
+      "learning_rate": 2.266636389336559e-06,
+      "loss": 0.09,
+      "step": 1087
+    },
+    {
+      "epoch": 5.307317073170732,
+      "grad_norm": 2.113255023956299,
+      "learning_rate": 2.262822182204686e-06,
+      "loss": 0.0267,
+      "step": 1088
+    },
+    {
+      "epoch": 5.31219512195122,
+      "grad_norm": 3.1760852336883545,
+      "learning_rate": 2.2590085320866798e-06,
+      "loss": 0.0295,
+      "step": 1089
+    },
+    {
+      "epoch": 5.317073170731708,
+      "grad_norm": 2.9674434661865234,
+      "learning_rate": 2.255195447938927e-06,
+      "loss": 0.0261,
+      "step": 1090
+    },
+    {
+      "epoch": 5.321951219512195,
+      "grad_norm": 3.4384074211120605,
+      "learning_rate": 2.251382938716482e-06,
+      "loss": 0.0936,
+      "step": 1091
+    },
+    {
+      "epoch": 5.326829268292683,
+      "grad_norm": 3.3814568519592285,
+      "learning_rate": 2.2475710133730533e-06,
+      "loss": 0.0426,
+      "step": 1092
+    },
+    {
+      "epoch": 5.331707317073171,
+      "grad_norm": 3.081317663192749,
+      "learning_rate": 2.243759680860975e-06,
+      "loss": 0.0799,
+      "step": 1093
+    },
+    {
+      "epoch": 5.336585365853659,
+      "grad_norm": 3.5608482360839844,
+      "learning_rate": 2.2399489501311896e-06,
+      "loss": 0.0906,
+      "step": 1094
+    },
+    {
+      "epoch": 5.341463414634147,
+      "grad_norm": 3.7886314392089844,
+      "learning_rate": 2.2361388301332265e-06,
+      "loss": 0.2152,
+      "step": 1095
+    },
+    {
+      "epoch": 5.3463414634146345,
+      "grad_norm": 1.9531102180480957,
+      "learning_rate": 2.2323293298151817e-06,
+      "loss": 0.0359,
+      "step": 1096
+    },
+    {
+      "epoch": 5.351219512195122,
+      "grad_norm": 2.2828023433685303,
+      "learning_rate": 2.2285204581236937e-06,
+      "loss": 0.0368,
+      "step": 1097
+    },
+    {
+      "epoch": 5.35609756097561,
+      "grad_norm": 3.110262870788574,
+      "learning_rate": 2.2247122240039268e-06,
+      "loss": 0.0426,
+      "step": 1098
+    },
+    {
+      "epoch": 5.360975609756098,
+      "grad_norm": 2.3293566703796387,
+      "learning_rate": 2.2209046363995464e-06,
+      "loss": 0.0223,
+      "step": 1099
+    },
+    {
+      "epoch": 5.365853658536586,
+      "grad_norm": 2.990884780883789,
+      "learning_rate": 2.217097704252701e-06,
+      "loss": 0.1276,
+      "step": 1100
+    },
+    {
+      "epoch": 5.3707317073170735,
+      "grad_norm": 2.568014144897461,
+      "learning_rate": 2.2132914365039993e-06,
+      "loss": 0.0639,
+      "step": 1101
+    },
+    {
+      "epoch": 5.375609756097561,
+      "grad_norm": 2.618478536605835,
+      "learning_rate": 2.2094858420924882e-06,
+      "loss": 0.0166,
+      "step": 1102
+    },
+    {
+      "epoch": 5.380487804878049,
+      "grad_norm": 4.526919364929199,
+      "learning_rate": 2.205680929955635e-06,
+      "loss": 0.144,
+      "step": 1103
+    },
+    {
+      "epoch": 5.385365853658537,
+      "grad_norm": 2.7236886024475098,
+      "learning_rate": 2.201876709029305e-06,
+      "loss": 0.1004,
+      "step": 1104
+    },
+    {
+      "epoch": 5.390243902439025,
+      "grad_norm": 2.1577632427215576,
+      "learning_rate": 2.198073188247738e-06,
+      "loss": 0.0453,
+      "step": 1105
+    },
+    {
+      "epoch": 5.3951219512195125,
+      "grad_norm": 2.5170321464538574,
+      "learning_rate": 2.1942703765435317e-06,
+      "loss": 0.0195,
+      "step": 1106
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 3.962658643722534,
+      "learning_rate": 2.190468282847617e-06,
+      "loss": 0.1512,
+      "step": 1107
+    },
+    {
+      "epoch": 5.404878048780488,
+      "grad_norm": 4.297860622406006,
+      "learning_rate": 2.186666916089239e-06,
+      "loss": 0.2572,
+      "step": 1108
+    },
+    {
+      "epoch": 5.409756097560976,
+      "grad_norm": 2.8933565616607666,
+      "learning_rate": 2.1828662851959377e-06,
+      "loss": 0.0536,
+      "step": 1109
+    },
+    {
+      "epoch": 5.414634146341464,
+      "grad_norm": 2.9397451877593994,
+      "learning_rate": 2.1790663990935203e-06,
+      "loss": 0.0778,
+      "step": 1110
+    },
+    {
+      "epoch": 5.419512195121952,
+      "grad_norm": 3.5210094451904297,
+      "learning_rate": 2.1752672667060488e-06,
+      "loss": 0.0558,
+      "step": 1111
+    },
+    {
+      "epoch": 5.424390243902439,
+      "grad_norm": 2.9027626514434814,
+      "learning_rate": 2.1714688969558146e-06,
+      "loss": 0.041,
+      "step": 1112
+    },
+    {
+      "epoch": 5.429268292682927,
+      "grad_norm": 3.7691168785095215,
+      "learning_rate": 2.167671298763316e-06,
+      "loss": 0.1644,
+      "step": 1113
+    },
+    {
+      "epoch": 5.434146341463415,
+      "grad_norm": 3.493008852005005,
+      "learning_rate": 2.1638744810472414e-06,
+      "loss": 0.1587,
+      "step": 1114
+    },
+    {
+      "epoch": 5.439024390243903,
+      "grad_norm": 2.711196184158325,
+      "learning_rate": 2.1600784527244445e-06,
+      "loss": 0.0605,
+      "step": 1115
+    },
+    {
+      "epoch": 5.443902439024391,
+      "grad_norm": 4.365038871765137,
+      "learning_rate": 2.1562832227099266e-06,
+      "loss": 0.1897,
+      "step": 1116
+    },
+    {
+      "epoch": 5.4487804878048784,
+      "grad_norm": 4.621466159820557,
+      "learning_rate": 2.152488799916814e-06,
+      "loss": 0.1525,
+      "step": 1117
+    },
+    {
+      "epoch": 5.453658536585366,
+      "grad_norm": 4.8721089363098145,
+      "learning_rate": 2.148695193256336e-06,
+      "loss": 0.189,
+      "step": 1118
+    },
+    {
+      "epoch": 5.458536585365854,
+      "grad_norm": 2.8999173641204834,
+      "learning_rate": 2.1449024116378064e-06,
+      "loss": 0.095,
+      "step": 1119
+    },
+    {
+      "epoch": 5.463414634146342,
+      "grad_norm": 2.4865314960479736,
+      "learning_rate": 2.1411104639686013e-06,
+      "loss": 0.0432,
+      "step": 1120
+    },
+    {
+      "epoch": 5.46829268292683,
+      "grad_norm": 3.8497228622436523,
+      "learning_rate": 2.137319359154138e-06,
+      "loss": 0.0954,
+      "step": 1121
+    },
+    {
+      "epoch": 5.473170731707317,
+      "grad_norm": 2.3643507957458496,
+      "learning_rate": 2.133529106097853e-06,
+      "loss": 0.0362,
+      "step": 1122
+    },
+    {
+      "epoch": 5.478048780487805,
+      "grad_norm": 3.017826795578003,
+      "learning_rate": 2.1297397137011862e-06,
+      "loss": 0.0875,
+      "step": 1123
+    },
+    {
+      "epoch": 5.482926829268292,
+      "grad_norm": 3.239320755004883,
+      "learning_rate": 2.125951190863551e-06,
+      "loss": 0.0758,
+      "step": 1124
+    },
+    {
+      "epoch": 5.487804878048781,
+      "grad_norm": 2.566241979598999,
+      "learning_rate": 2.1221635464823237e-06,
+      "loss": 0.0605,
+      "step": 1125
+    },
+    {
+      "epoch": 5.492682926829268,
+      "grad_norm": 4.810088157653809,
+      "learning_rate": 2.1183767894528135e-06,
+      "loss": 0.2403,
+      "step": 1126
+    },
+    {
+      "epoch": 5.4975609756097565,
+      "grad_norm": 2.083263397216797,
+      "learning_rate": 2.114590928668249e-06,
+      "loss": 0.0223,
+      "step": 1127
+    },
+    {
+      "epoch": 5.5024390243902435,
+      "grad_norm": 2.6812374591827393,
+      "learning_rate": 2.1108059730197517e-06,
+      "loss": 0.0617,
+      "step": 1128
+    },
+    {
+      "epoch": 5.507317073170732,
+      "grad_norm": 3.196735143661499,
+      "learning_rate": 2.1070219313963173e-06,
+      "loss": 0.043,
+      "step": 1129
+    },
+    {
+      "epoch": 5.512195121951219,
+      "grad_norm": 2.775470495223999,
+      "learning_rate": 2.1032388126847967e-06,
+      "loss": 0.0595,
+      "step": 1130
+    },
+    {
+      "epoch": 5.517073170731708,
+      "grad_norm": 2.8632407188415527,
+      "learning_rate": 2.099456625769872e-06,
+      "loss": 0.0186,
+      "step": 1131
+    },
+    {
+      "epoch": 5.521951219512195,
+      "grad_norm": 4.075018405914307,
+      "learning_rate": 2.0956753795340376e-06,
+      "loss": 0.0616,
+      "step": 1132
+    },
+    {
+      "epoch": 5.526829268292683,
+      "grad_norm": 3.206327199935913,
+      "learning_rate": 2.091895082857578e-06,
+      "loss": 0.1895,
+      "step": 1133
+    },
+    {
+      "epoch": 5.53170731707317,
+      "grad_norm": 2.967588186264038,
+      "learning_rate": 2.0881157446185474e-06,
+      "loss": 0.0484,
+      "step": 1134
+    },
+    {
+      "epoch": 5.536585365853659,
+      "grad_norm": 2.850929021835327,
+      "learning_rate": 2.0843373736927506e-06,
+      "loss": 0.037,
+      "step": 1135
+    },
+    {
+      "epoch": 5.541463414634146,
+      "grad_norm": 2.2505147457122803,
+      "learning_rate": 2.08055997895372e-06,
+      "loss": 0.0227,
+      "step": 1136
+    },
+    {
+      "epoch": 5.546341463414635,
+      "grad_norm": 2.5258476734161377,
+      "learning_rate": 2.0767835692726944e-06,
+      "loss": 0.0296,
+      "step": 1137
+    },
+    {
+      "epoch": 5.5512195121951216,
+      "grad_norm": 3.498741388320923,
+      "learning_rate": 2.0730081535186e-06,
+      "loss": 0.16,
+      "step": 1138
+    },
+    {
+      "epoch": 5.55609756097561,
+      "grad_norm": 2.8635222911834717,
+      "learning_rate": 2.06923374055803e-06,
+      "loss": 0.0725,
+      "step": 1139
+    },
+    {
+      "epoch": 5.560975609756097,
+      "grad_norm": 2.2779290676116943,
+      "learning_rate": 2.0654603392552193e-06,
+      "loss": 0.0198,
+      "step": 1140
+    },
+    {
+      "epoch": 5.565853658536585,
+      "grad_norm": 3.1651058197021484,
+      "learning_rate": 2.0616879584720305e-06,
+      "loss": 0.1144,
+      "step": 1141
+    },
+    {
+      "epoch": 5.570731707317073,
+      "grad_norm": 2.4238595962524414,
+      "learning_rate": 2.057916607067928e-06,
+      "loss": 0.0491,
+      "step": 1142
+    },
+    {
+      "epoch": 5.575609756097561,
+      "grad_norm": 2.3248515129089355,
+      "learning_rate": 2.054146293899957e-06,
+      "loss": 0.035,
+      "step": 1143
+    },
+    {
+      "epoch": 5.580487804878048,
+      "grad_norm": 2.9506516456604004,
+      "learning_rate": 2.0503770278227274e-06,
+      "loss": 0.0639,
+      "step": 1144
+    },
+    {
+      "epoch": 5.585365853658536,
+      "grad_norm": 2.6403958797454834,
+      "learning_rate": 2.0466088176883876e-06,
+      "loss": 0.0258,
+      "step": 1145
+    },
+    {
+      "epoch": 5.590243902439024,
+      "grad_norm": 3.150115728378296,
+      "learning_rate": 2.042841672346608e-06,
+      "loss": 0.0634,
+      "step": 1146
+    },
+    {
+      "epoch": 5.595121951219512,
+      "grad_norm": 2.742691993713379,
+      "learning_rate": 2.039075600644557e-06,
+      "loss": 0.0464,
+      "step": 1147
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 2.733694076538086,
+      "learning_rate": 2.0353106114268824e-06,
+      "loss": 0.0829,
+      "step": 1148
+    },
+    {
+      "epoch": 5.6048780487804875,
+      "grad_norm": 2.511229991912842,
+      "learning_rate": 2.031546713535688e-06,
+      "loss": 0.0321,
+      "step": 1149
+    },
+    {
+      "epoch": 5.609756097560975,
+      "grad_norm": 3.019669532775879,
+      "learning_rate": 2.027783915810518e-06,
+      "loss": 0.05,
+      "step": 1150
+    },
+    {
+      "epoch": 5.614634146341463,
+      "grad_norm": 3.497159242630005,
+      "learning_rate": 2.024022227088329e-06,
+      "loss": 0.1984,
+      "step": 1151
+    },
+    {
+      "epoch": 5.619512195121951,
+      "grad_norm": 3.4637508392333984,
+      "learning_rate": 2.020261656203476e-06,
+      "loss": 0.1673,
+      "step": 1152
+    },
+    {
+      "epoch": 5.624390243902439,
+      "grad_norm": 2.4312477111816406,
+      "learning_rate": 2.016502211987687e-06,
+      "loss": 0.1106,
+      "step": 1153
+    },
+    {
+      "epoch": 5.6292682926829265,
+      "grad_norm": 2.7801673412323,
+      "learning_rate": 2.0127439032700446e-06,
+      "loss": 0.0374,
+      "step": 1154
+    },
+    {
+      "epoch": 5.634146341463414,
+      "grad_norm": 2.9346680641174316,
+      "learning_rate": 2.0089867388769664e-06,
+      "loss": 0.0674,
+      "step": 1155
+    },
+    {
+      "epoch": 5.639024390243902,
+      "grad_norm": 2.274888277053833,
+      "learning_rate": 2.0052307276321793e-06,
+      "loss": 0.0365,
+      "step": 1156
+    },
+    {
+      "epoch": 5.64390243902439,
+      "grad_norm": 3.069890022277832,
+      "learning_rate": 2.001475878356703e-06,
+      "loss": 0.0758,
+      "step": 1157
+    },
+    {
+      "epoch": 5.648780487804878,
+      "grad_norm": 3.8594915866851807,
+      "learning_rate": 1.99772219986883e-06,
+      "loss": 0.176,
+      "step": 1158
+    },
+    {
+      "epoch": 5.6536585365853655,
+      "grad_norm": 3.4886410236358643,
+      "learning_rate": 1.9939697009841024e-06,
+      "loss": 0.0491,
+      "step": 1159
+    },
+    {
+      "epoch": 5.658536585365853,
+      "grad_norm": 2.697946786880493,
+      "learning_rate": 1.990218390515291e-06,
+      "loss": 0.0741,
+      "step": 1160
+    },
+    {
+      "epoch": 5.663414634146341,
+      "grad_norm": 3.5290887355804443,
+      "learning_rate": 1.9864682772723757e-06,
+      "loss": 0.0826,
+      "step": 1161
+    },
+    {
+      "epoch": 5.668292682926829,
+      "grad_norm": 2.0601298809051514,
+      "learning_rate": 1.9827193700625274e-06,
+      "loss": 0.0378,
+      "step": 1162
+    },
+    {
+      "epoch": 5.673170731707317,
+      "grad_norm": 3.8458635807037354,
+      "learning_rate": 1.978971677690081e-06,
+      "loss": 0.2466,
+      "step": 1163
+    },
+    {
+      "epoch": 5.678048780487805,
+      "grad_norm": 2.788210153579712,
+      "learning_rate": 1.97522520895652e-06,
+      "loss": 0.0205,
+      "step": 1164
+    },
+    {
+      "epoch": 5.682926829268292,
+      "grad_norm": 3.1904587745666504,
+      "learning_rate": 1.971479972660454e-06,
+      "loss": 0.0998,
+      "step": 1165
+    },
+    {
+      "epoch": 5.68780487804878,
+      "grad_norm": 2.4664318561553955,
+      "learning_rate": 1.967735977597598e-06,
+      "loss": 0.0217,
+      "step": 1166
+    },
+    {
+      "epoch": 5.692682926829268,
+      "grad_norm": 2.1392667293548584,
+      "learning_rate": 1.9639932325607538e-06,
+      "loss": 0.048,
+      "step": 1167
+    },
+    {
+      "epoch": 5.697560975609756,
+      "grad_norm": 3.7127058506011963,
+      "learning_rate": 1.9602517463397845e-06,
+      "loss": 0.0302,
+      "step": 1168
+    },
+    {
+      "epoch": 5.702439024390244,
+      "grad_norm": 2.916168689727783,
+      "learning_rate": 1.9565115277215978e-06,
+      "loss": 0.0724,
+      "step": 1169
+    },
+    {
+      "epoch": 5.7073170731707314,
+      "grad_norm": 2.4352428913116455,
+      "learning_rate": 1.952772585490127e-06,
+      "loss": 0.0464,
+      "step": 1170
+    },
+    {
+      "epoch": 5.712195121951219,
+      "grad_norm": 2.8311455249786377,
+      "learning_rate": 1.9490349284263036e-06,
+      "loss": 0.0239,
+      "step": 1171
+    },
+    {
+      "epoch": 5.717073170731707,
+      "grad_norm": 3.3592801094055176,
+      "learning_rate": 1.9452985653080443e-06,
+      "loss": 0.0719,
+      "step": 1172
+    },
+    {
+      "epoch": 5.721951219512195,
+      "grad_norm": 2.450922966003418,
+      "learning_rate": 1.9415635049102245e-06,
+      "loss": 0.0408,
+      "step": 1173
+    },
+    {
+      "epoch": 5.726829268292683,
+      "grad_norm": 4.750118255615234,
+      "learning_rate": 1.937829756004662e-06,
+      "loss": 0.2049,
+      "step": 1174
+    },
+    {
+      "epoch": 5.7317073170731705,
+      "grad_norm": 3.0643811225891113,
+      "learning_rate": 1.9340973273600944e-06,
+      "loss": 0.0636,
+      "step": 1175
+    },
+    {
+      "epoch": 5.736585365853658,
+      "grad_norm": 3.313904047012329,
+      "learning_rate": 1.930366227742157e-06,
+      "loss": 0.1252,
+      "step": 1176
+    },
+    {
+      "epoch": 5.741463414634146,
+      "grad_norm": 3.8996808528900146,
+      "learning_rate": 1.9266364659133653e-06,
+      "loss": 0.0687,
+      "step": 1177
+    },
+    {
+      "epoch": 5.746341463414634,
+      "grad_norm": 2.727555274963379,
+      "learning_rate": 1.922908050633093e-06,
+      "loss": 0.0333,
+      "step": 1178
+    },
+    {
+      "epoch": 5.751219512195122,
+      "grad_norm": 3.270087718963623,
+      "learning_rate": 1.919180990657551e-06,
+      "loss": 0.0792,
+      "step": 1179
+    },
+    {
+      "epoch": 5.7560975609756095,
+      "grad_norm": 2.6631274223327637,
+      "learning_rate": 1.9154552947397668e-06,
+      "loss": 0.069,
+      "step": 1180
+    },
+    {
+      "epoch": 5.760975609756097,
+      "grad_norm": 4.4460554122924805,
+      "learning_rate": 1.9117309716295658e-06,
+      "loss": 0.115,
+      "step": 1181
+    },
+    {
+      "epoch": 5.765853658536585,
+      "grad_norm": 2.5652341842651367,
+      "learning_rate": 1.9080080300735478e-06,
+      "loss": 0.0537,
+      "step": 1182
+    },
+    {
+      "epoch": 5.770731707317073,
+      "grad_norm": 3.046436071395874,
+      "learning_rate": 1.9042864788150695e-06,
+      "loss": 0.0817,
+      "step": 1183
+    },
+    {
+      "epoch": 5.775609756097561,
+      "grad_norm": 2.121629238128662,
+      "learning_rate": 1.9005663265942206e-06,
+      "loss": 0.0289,
+      "step": 1184
+    },
+    {
+      "epoch": 5.780487804878049,
+      "grad_norm": 2.271918535232544,
+      "learning_rate": 1.8968475821478066e-06,
+      "loss": 0.0357,
+      "step": 1185
+    },
+    {
+      "epoch": 5.785365853658536,
+      "grad_norm": 2.582473039627075,
+      "learning_rate": 1.8931302542093274e-06,
+      "loss": 0.0584,
+      "step": 1186
+    },
+    {
+      "epoch": 5.790243902439024,
+      "grad_norm": 2.502952814102173,
+      "learning_rate": 1.8894143515089539e-06,
+      "loss": 0.0324,
+      "step": 1187
+    },
+    {
+      "epoch": 5.795121951219512,
+      "grad_norm": 1.9735453128814697,
+      "learning_rate": 1.8856998827735118e-06,
+      "loss": 0.0338,
+      "step": 1188
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 4.441845893859863,
+      "learning_rate": 1.8819868567264588e-06,
+      "loss": 0.1706,
+      "step": 1189
+    },
+    {
+      "epoch": 5.804878048780488,
+      "grad_norm": 2.5450692176818848,
+      "learning_rate": 1.8782752820878636e-06,
+      "loss": 0.0463,
+      "step": 1190
+    },
+    {
+      "epoch": 5.809756097560975,
+      "grad_norm": 3.718183755874634,
+      "learning_rate": 1.8745651675743876e-06,
+      "loss": 0.1188,
+      "step": 1191
+    },
+    {
+      "epoch": 5.814634146341463,
+      "grad_norm": 3.246532678604126,
+      "learning_rate": 1.870856521899261e-06,
+      "loss": 0.0984,
+      "step": 1192
+    },
+    {
+      "epoch": 5.819512195121951,
+      "grad_norm": 2.9522783756256104,
+      "learning_rate": 1.867149353772267e-06,
+      "loss": 0.0195,
+      "step": 1193
+    },
+    {
+      "epoch": 5.824390243902439,
+      "grad_norm": 2.3266429901123047,
+      "learning_rate": 1.863443671899717e-06,
+      "loss": 0.0236,
+      "step": 1194
+    },
+    {
+      "epoch": 5.829268292682927,
+      "grad_norm": 3.696749448776245,
+      "learning_rate": 1.8597394849844319e-06,
+      "loss": 0.1108,
+      "step": 1195
+    },
+    {
+      "epoch": 5.8341463414634145,
+      "grad_norm": 2.375624179840088,
+      "learning_rate": 1.8560368017257229e-06,
+      "loss": 0.0388,
+      "step": 1196
+    },
+    {
+      "epoch": 5.839024390243902,
+      "grad_norm": 4.0437092781066895,
+      "learning_rate": 1.8523356308193696e-06,
+      "loss": 0.3098,
+      "step": 1197
+    },
+    {
+      "epoch": 5.84390243902439,
+      "grad_norm": 3.165165424346924,
+      "learning_rate": 1.8486359809575977e-06,
+      "loss": 0.0775,
+      "step": 1198
+    },
+    {
+      "epoch": 5.848780487804878,
+      "grad_norm": 4.1991190910339355,
+      "learning_rate": 1.8449378608290638e-06,
+      "loss": 0.1222,
+      "step": 1199
+    },
+    {
+      "epoch": 5.853658536585366,
+      "grad_norm": 4.6657819747924805,
+      "learning_rate": 1.8412412791188306e-06,
+      "loss": 0.1146,
+      "step": 1200
+    },
+    {
+      "epoch": 5.8585365853658535,
+      "grad_norm": 4.569516181945801,
+      "learning_rate": 1.8375462445083464e-06,
+      "loss": 0.1113,
+      "step": 1201
+    },
+    {
+      "epoch": 5.863414634146341,
+      "grad_norm": 3.1565654277801514,
+      "learning_rate": 1.8338527656754285e-06,
+      "loss": 0.0416,
+      "step": 1202
+    },
+    {
+      "epoch": 5.868292682926829,
+      "grad_norm": 3.3474619388580322,
+      "learning_rate": 1.830160851294239e-06,
+      "loss": 0.0613,
+      "step": 1203
+    },
+    {
+      "epoch": 5.873170731707317,
+      "grad_norm": 4.30797004699707,
+      "learning_rate": 1.8264705100352662e-06,
+      "loss": 0.197,
+      "step": 1204
+    },
+    {
+      "epoch": 5.878048780487805,
+      "grad_norm": 2.7259573936462402,
+      "learning_rate": 1.8227817505653045e-06,
+      "loss": 0.0821,
+      "step": 1205
+    },
+    {
+      "epoch": 5.882926829268293,
+      "grad_norm": 3.515812873840332,
+      "learning_rate": 1.8190945815474323e-06,
+      "loss": 0.1246,
+      "step": 1206
+    },
+    {
+      "epoch": 5.88780487804878,
+      "grad_norm": 2.9223313331604004,
+      "learning_rate": 1.8154090116409934e-06,
+      "loss": 0.0703,
+      "step": 1207
+    },
+    {
+      "epoch": 5.892682926829268,
+      "grad_norm": 3.9529640674591064,
+      "learning_rate": 1.811725049501577e-06,
+      "loss": 0.1078,
+      "step": 1208
+    },
+    {
+      "epoch": 5.897560975609756,
+      "grad_norm": 4.1674580574035645,
+      "learning_rate": 1.8080427037809941e-06,
+      "loss": 0.1648,
+      "step": 1209
+    },
+    {
+      "epoch": 5.902439024390244,
+      "grad_norm": 3.1308021545410156,
+      "learning_rate": 1.8043619831272623e-06,
+      "loss": 0.061,
+      "step": 1210
+    },
+    {
+      "epoch": 5.907317073170732,
+      "grad_norm": 3.9667179584503174,
+      "learning_rate": 1.8006828961845807e-06,
+      "loss": 0.1863,
+      "step": 1211
+    },
+    {
+      "epoch": 5.912195121951219,
+      "grad_norm": 5.438168048858643,
+      "learning_rate": 1.7970054515933124e-06,
+      "loss": 0.2387,
+      "step": 1212
+    },
+    {
+      "epoch": 5.917073170731707,
+      "grad_norm": 5.505797863006592,
+      "learning_rate": 1.793329657989964e-06,
+      "loss": 0.2053,
+      "step": 1213
+    },
+    {
+      "epoch": 5.921951219512195,
+      "grad_norm": 2.8043150901794434,
+      "learning_rate": 1.7896555240071627e-06,
+      "loss": 0.026,
+      "step": 1214
+    },
+    {
+      "epoch": 5.926829268292683,
+      "grad_norm": 2.836164712905884,
+      "learning_rate": 1.7859830582736406e-06,
+      "loss": 0.0735,
+      "step": 1215
+    },
+    {
+      "epoch": 5.931707317073171,
+      "grad_norm": 2.8286306858062744,
+      "learning_rate": 1.782312269414211e-06,
+      "loss": 0.0586,
+      "step": 1216
+    },
+    {
+      "epoch": 5.9365853658536585,
+      "grad_norm": 4.4354329109191895,
+      "learning_rate": 1.7786431660497474e-06,
+      "loss": 0.3086,
+      "step": 1217
+    },
+    {
+      "epoch": 5.941463414634146,
+      "grad_norm": 4.0963640213012695,
+      "learning_rate": 1.7749757567971678e-06,
+      "loss": 0.0978,
+      "step": 1218
+    },
+    {
+      "epoch": 5.946341463414634,
+      "grad_norm": 2.726062536239624,
+      "learning_rate": 1.7713100502694091e-06,
+      "loss": 0.0976,
+      "step": 1219
+    },
+    {
+      "epoch": 5.951219512195122,
+      "grad_norm": 2.6566951274871826,
+      "learning_rate": 1.7676460550754104e-06,
+      "loss": 0.02,
+      "step": 1220
+    },
+    {
+      "epoch": 5.95609756097561,
+      "grad_norm": 2.7710952758789062,
+      "learning_rate": 1.7639837798200923e-06,
+      "loss": 0.0741,
+      "step": 1221
+    },
+    {
+      "epoch": 5.9609756097560975,
+      "grad_norm": 2.3678600788116455,
+      "learning_rate": 1.7603232331043346e-06,
+      "loss": 0.0542,
+      "step": 1222
+    },
+    {
+      "epoch": 5.965853658536585,
+      "grad_norm": 6.45259428024292,
+      "learning_rate": 1.7566644235249591e-06,
+      "loss": 0.3552,
+      "step": 1223
+    },
+    {
+      "epoch": 5.970731707317073,
+      "grad_norm": 1.8916475772857666,
+      "learning_rate": 1.7530073596747072e-06,
+      "loss": 0.0405,
+      "step": 1224
+    },
+    {
+      "epoch": 5.975609756097561,
+      "grad_norm": 2.1637566089630127,
+      "learning_rate": 1.74935205014222e-06,
+      "loss": 0.0178,
+      "step": 1225
+    },
+    {
+      "epoch": 5.980487804878049,
+      "grad_norm": 2.5959200859069824,
+      "learning_rate": 1.7456985035120194e-06,
+      "loss": 0.0264,
+      "step": 1226
+    },
+    {
+      "epoch": 5.985365853658537,
+      "grad_norm": 2.50264573097229,
+      "learning_rate": 1.7420467283644877e-06,
+      "loss": 0.0555,
+      "step": 1227
+    },
+    {
+      "epoch": 5.990243902439024,
+      "grad_norm": 2.4692020416259766,
+      "learning_rate": 1.738396733275844e-06,
+      "loss": 0.0546,
+      "step": 1228
+    },
+    {
+      "epoch": 5.995121951219512,
+      "grad_norm": 5.540846824645996,
+      "learning_rate": 1.7347485268181309e-06,
+      "loss": 0.1967,
+      "step": 1229
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.8322839736938477,
+      "learning_rate": 1.7311021175591868e-06,
+      "loss": 0.0491,
+      "step": 1230
+    },
+    {
+      "epoch": 6.004878048780488,
+      "grad_norm": 2.719622850418091,
+      "learning_rate": 1.7274575140626318e-06,
+      "loss": 0.0359,
+      "step": 1231
+    },
+    {
+      "epoch": 6.009756097560976,
+      "grad_norm": 2.859675884246826,
+      "learning_rate": 1.7238147248878444e-06,
+      "loss": 0.0585,
+      "step": 1232
+    },
+    {
+      "epoch": 6.014634146341463,
+      "grad_norm": 1.6761114597320557,
+      "learning_rate": 1.7201737585899415e-06,
+      "loss": 0.0188,
+      "step": 1233
+    },
+    {
+      "epoch": 6.019512195121951,
+      "grad_norm": 2.1588776111602783,
+      "learning_rate": 1.7165346237197594e-06,
+      "loss": 0.0484,
+      "step": 1234
+    },
+    {
+      "epoch": 6.024390243902439,
+      "grad_norm": 4.209983825683594,
+      "learning_rate": 1.7128973288238344e-06,
+      "loss": 0.0776,
+      "step": 1235
+    },
+    {
+      "epoch": 6.029268292682927,
+      "grad_norm": 2.3979365825653076,
+      "learning_rate": 1.709261882444379e-06,
+      "loss": 0.0338,
+      "step": 1236
+    },
+    {
+      "epoch": 6.034146341463415,
+      "grad_norm": 3.0030531883239746,
+      "learning_rate": 1.705628293119268e-06,
+      "loss": 0.0385,
+      "step": 1237
+    },
+    {
+      "epoch": 6.0390243902439025,
+      "grad_norm": 9.65616512298584,
+      "learning_rate": 1.701996569382011e-06,
+      "loss": 0.2601,
+      "step": 1238
+    },
+    {
+      "epoch": 6.04390243902439,
+      "grad_norm": 3.0590052604675293,
+      "learning_rate": 1.6983667197617386e-06,
+      "loss": 0.034,
+      "step": 1239
+    },
+    {
+      "epoch": 6.048780487804878,
+      "grad_norm": 3.6949822902679443,
+      "learning_rate": 1.6947387527831813e-06,
+      "loss": 0.0155,
+      "step": 1240
+    },
+    {
+      "epoch": 6.053658536585366,
+      "grad_norm": 1.2870460748672485,
+      "learning_rate": 1.6911126769666442e-06,
+      "loss": 0.0078,
+      "step": 1241
+    },
+    {
+      "epoch": 6.058536585365854,
+      "grad_norm": 4.307460784912109,
+      "learning_rate": 1.6874885008279945e-06,
+      "loss": 0.1429,
+      "step": 1242
+    },
+    {
+      "epoch": 6.0634146341463415,
+      "grad_norm": 2.334972858428955,
+      "learning_rate": 1.683866232878637e-06,
+      "loss": 0.0123,
+      "step": 1243
+    },
+    {
+      "epoch": 6.068292682926829,
+      "grad_norm": 2.4121835231781006,
+      "learning_rate": 1.6802458816254941e-06,
+      "loss": 0.0139,
+      "step": 1244
+    },
+    {
+      "epoch": 6.073170731707317,
+      "grad_norm": 1.9224514961242676,
+      "learning_rate": 1.676627455570988e-06,
+      "loss": 0.0312,
+      "step": 1245
+    },
+    {
+      "epoch": 6.078048780487805,
+      "grad_norm": 2.8293309211730957,
+      "learning_rate": 1.6730109632130199e-06,
+      "loss": 0.0464,
+      "step": 1246
+    },
+    {
+      "epoch": 6.082926829268293,
+      "grad_norm": 1.6368179321289062,
+      "learning_rate": 1.6693964130449472e-06,
+      "loss": 0.0085,
+      "step": 1247
+    },
+    {
+      "epoch": 6.087804878048781,
+      "grad_norm": 2.5535073280334473,
+      "learning_rate": 1.6657838135555696e-06,
+      "loss": 0.0482,
+      "step": 1248
+    },
+    {
+      "epoch": 6.092682926829268,
+      "grad_norm": 3.7743096351623535,
+      "learning_rate": 1.6621731732291024e-06,
+      "loss": 0.0235,
+      "step": 1249
+    },
+    {
+      "epoch": 6.097560975609756,
+      "grad_norm": 2.9921820163726807,
+      "learning_rate": 1.6585645005451623e-06,
+      "loss": 0.0455,
+      "step": 1250
+    },
+    {
+      "epoch": 6.102439024390244,
+      "grad_norm": 2.369581937789917,
+      "learning_rate": 1.6549578039787436e-06,
+      "loss": 0.0499,
+      "step": 1251
+    },
+    {
+      "epoch": 6.107317073170732,
+      "grad_norm": 2.163815498352051,
+      "learning_rate": 1.6513530920001998e-06,
+      "loss": 0.0118,
+      "step": 1252
+    },
+    {
+      "epoch": 6.11219512195122,
+      "grad_norm": 2.034928560256958,
+      "learning_rate": 1.6477503730752237e-06,
+      "loss": 0.0189,
+      "step": 1253
+    },
+    {
+      "epoch": 6.117073170731707,
+      "grad_norm": 2.7306160926818848,
+      "learning_rate": 1.6441496556648278e-06,
+      "loss": 0.0492,
+      "step": 1254
+    },
+    {
+      "epoch": 6.121951219512195,
+      "grad_norm": 3.7521040439605713,
+      "learning_rate": 1.6405509482253234e-06,
+      "loss": 0.1717,
+      "step": 1255
+    },
+    {
+      "epoch": 6.126829268292683,
+      "grad_norm": 1.8965831995010376,
+      "learning_rate": 1.636954259208302e-06,
+      "loss": 0.0194,
+      "step": 1256
+    },
+    {
+      "epoch": 6.131707317073171,
+      "grad_norm": 3.010024070739746,
+      "learning_rate": 1.6333595970606143e-06,
+      "loss": 0.0334,
+      "step": 1257
+    },
+    {
+      "epoch": 6.136585365853659,
+      "grad_norm": 3.7091450691223145,
+      "learning_rate": 1.62976697022435e-06,
+      "loss": 0.0705,
+      "step": 1258
+    },
+    {
+      "epoch": 6.1414634146341465,
+      "grad_norm": 3.5719785690307617,
+      "learning_rate": 1.6261763871368225e-06,
+      "loss": 0.0322,
+      "step": 1259
+    },
+    {
+      "epoch": 6.146341463414634,
+      "grad_norm": 3.3224213123321533,
+      "learning_rate": 1.6225878562305403e-06,
+      "loss": 0.0653,
+      "step": 1260
+    },
+    {
+      "epoch": 6.151219512195122,
+      "grad_norm": 3.78924822807312,
+      "learning_rate": 1.6190013859331958e-06,
+      "loss": 0.0557,
+      "step": 1261
+    },
+    {
+      "epoch": 6.15609756097561,
+      "grad_norm": 2.429412841796875,
+      "learning_rate": 1.6154169846676415e-06,
+      "loss": 0.0277,
+      "step": 1262
+    },
+    {
+      "epoch": 6.160975609756098,
+      "grad_norm": 2.626167058944702,
+      "learning_rate": 1.6118346608518698e-06,
+      "loss": 0.0305,
+      "step": 1263
+    },
+    {
+      "epoch": 6.1658536585365855,
+      "grad_norm": 2.44846248626709,
+      "learning_rate": 1.6082544228989958e-06,
+      "loss": 0.0093,
+      "step": 1264
+    },
+    {
+      "epoch": 6.170731707317073,
+      "grad_norm": 2.9345643520355225,
+      "learning_rate": 1.6046762792172336e-06,
+      "loss": 0.0198,
+      "step": 1265
+    },
+    {
+      "epoch": 6.175609756097561,
+      "grad_norm": 3.224313497543335,
+      "learning_rate": 1.6011002382098806e-06,
+      "loss": 0.0673,
+      "step": 1266
+    },
+    {
+      "epoch": 6.180487804878049,
+      "grad_norm": 1.9066869020462036,
+      "learning_rate": 1.5975263082752968e-06,
+      "loss": 0.0115,
+      "step": 1267
+    },
+    {
+      "epoch": 6.185365853658537,
+      "grad_norm": 2.7153308391571045,
+      "learning_rate": 1.5939544978068816e-06,
+      "loss": 0.0529,
+      "step": 1268
+    },
+    {
+      "epoch": 6.190243902439025,
+      "grad_norm": 2.2173709869384766,
+      "learning_rate": 1.590384815193059e-06,
+      "loss": 0.0643,
+      "step": 1269
+    },
+    {
+      "epoch": 6.195121951219512,
+      "grad_norm": 3.1238555908203125,
+      "learning_rate": 1.5868172688172559e-06,
+      "loss": 0.064,
+      "step": 1270
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 2.7765870094299316,
+      "learning_rate": 1.5832518670578802e-06,
+      "loss": 0.0676,
+      "step": 1271
+    },
+    {
+      "epoch": 6.204878048780488,
+      "grad_norm": 2.9892525672912598,
+      "learning_rate": 1.5796886182883053e-06,
+      "loss": 0.074,
+      "step": 1272
+    },
+    {
+      "epoch": 6.209756097560976,
+      "grad_norm": 2.0955512523651123,
+      "learning_rate": 1.5761275308768476e-06,
+      "loss": 0.0311,
+      "step": 1273
+    },
+    {
+      "epoch": 6.214634146341464,
+      "grad_norm": 1.8085861206054688,
+      "learning_rate": 1.5725686131867462e-06,
+      "loss": 0.0108,
+      "step": 1274
+    },
+    {
+      "epoch": 6.219512195121951,
+      "grad_norm": 3.026421308517456,
+      "learning_rate": 1.569011873576147e-06,
+      "loss": 0.0464,
+      "step": 1275
+    },
+    {
+      "epoch": 6.224390243902439,
+      "grad_norm": 2.3395111560821533,
+      "learning_rate": 1.5654573203980782e-06,
+      "loss": 0.0221,
+      "step": 1276
+    },
+    {
+      "epoch": 6.229268292682927,
+      "grad_norm": 3.6158692836761475,
+      "learning_rate": 1.5619049620004354e-06,
+      "loss": 0.0693,
+      "step": 1277
+    },
+    {
+      "epoch": 6.234146341463415,
+      "grad_norm": 1.6186567544937134,
+      "learning_rate": 1.5583548067259584e-06,
+      "loss": 0.0198,
+      "step": 1278
+    },
+    {
+      "epoch": 6.239024390243903,
+      "grad_norm": 2.7193195819854736,
+      "learning_rate": 1.5548068629122126e-06,
+      "loss": 0.0687,
+      "step": 1279
+    },
+    {
+      "epoch": 6.2439024390243905,
+      "grad_norm": 2.7472658157348633,
+      "learning_rate": 1.5512611388915711e-06,
+      "loss": 0.053,
+      "step": 1280
+    },
+    {
+      "epoch": 6.248780487804878,
+      "grad_norm": 4.694706439971924,
+      "learning_rate": 1.5477176429911934e-06,
+      "loss": 0.2076,
+      "step": 1281
+    },
+    {
+      "epoch": 6.253658536585366,
+      "grad_norm": 1.609309434890747,
+      "learning_rate": 1.5441763835330048e-06,
+      "loss": 0.0108,
+      "step": 1282
+    },
+    {
+      "epoch": 6.258536585365854,
+      "grad_norm": 1.7064504623413086,
+      "learning_rate": 1.5406373688336807e-06,
+      "loss": 0.0114,
+      "step": 1283
+    },
+    {
+      "epoch": 6.263414634146342,
+      "grad_norm": 1.967726469039917,
+      "learning_rate": 1.5371006072046225e-06,
+      "loss": 0.0209,
+      "step": 1284
+    },
+    {
+      "epoch": 6.2682926829268295,
+      "grad_norm": 2.4065544605255127,
+      "learning_rate": 1.5335661069519408e-06,
+      "loss": 0.0741,
+      "step": 1285
+    },
+    {
+      "epoch": 6.273170731707317,
+      "grad_norm": 2.2167603969573975,
+      "learning_rate": 1.5300338763764371e-06,
+      "loss": 0.0121,
+      "step": 1286
+    },
+    {
+      "epoch": 6.278048780487805,
+      "grad_norm": 3.229228973388672,
+      "learning_rate": 1.5265039237735804e-06,
+      "loss": 0.0226,
+      "step": 1287
+    },
+    {
+      "epoch": 6.282926829268293,
+      "grad_norm": 1.889419674873352,
+      "learning_rate": 1.5229762574334903e-06,
+      "loss": 0.0116,
+      "step": 1288
+    },
+    {
+      "epoch": 6.287804878048781,
+      "grad_norm": 3.7595815658569336,
+      "learning_rate": 1.5194508856409181e-06,
+      "loss": 0.0775,
+      "step": 1289
+    },
+    {
+      "epoch": 6.2926829268292686,
+      "grad_norm": 2.527560234069824,
+      "learning_rate": 1.515927816675225e-06,
+      "loss": 0.0355,
+      "step": 1290
+    },
+    {
+      "epoch": 6.297560975609756,
+      "grad_norm": 1.9718955755233765,
+      "learning_rate": 1.5124070588103648e-06,
+      "loss": 0.0127,
+      "step": 1291
+    },
+    {
+      "epoch": 6.302439024390244,
+      "grad_norm": 1.9010120630264282,
+      "learning_rate": 1.5088886203148643e-06,
+      "loss": 0.0188,
+      "step": 1292
+    },
+    {
+      "epoch": 6.307317073170732,
+      "grad_norm": 3.2093472480773926,
+      "learning_rate": 1.505372509451801e-06,
+      "loss": 0.0845,
+      "step": 1293
+    },
+    {
+      "epoch": 6.31219512195122,
+      "grad_norm": 1.6723257303237915,
+      "learning_rate": 1.5018587344787888e-06,
+      "loss": 0.0265,
+      "step": 1294
+    },
+    {
+      "epoch": 6.317073170731708,
+      "grad_norm": 3.246812343597412,
+      "learning_rate": 1.498347303647953e-06,
+      "loss": 0.0833,
+      "step": 1295
+    },
+    {
+      "epoch": 6.321951219512195,
+      "grad_norm": 2.887834072113037,
+      "learning_rate": 1.4948382252059158e-06,
+      "loss": 0.0416,
+      "step": 1296
+    },
+    {
+      "epoch": 6.326829268292683,
+      "grad_norm": 2.5762557983398438,
+      "learning_rate": 1.4913315073937742e-06,
+      "loss": 0.0614,
+      "step": 1297
+    },
+    {
+      "epoch": 6.331707317073171,
+      "grad_norm": 3.3746497631073,
+      "learning_rate": 1.4878271584470805e-06,
+      "loss": 0.0601,
+      "step": 1298
+    },
+    {
+      "epoch": 6.336585365853659,
+      "grad_norm": 2.4984664916992188,
+      "learning_rate": 1.4843251865958242e-06,
+      "loss": 0.0189,
+      "step": 1299
+    },
+    {
+      "epoch": 6.341463414634147,
+      "grad_norm": 3.178300619125366,
+      "learning_rate": 1.4808256000644128e-06,
+      "loss": 0.038,
+      "step": 1300
+    },
+    {
+      "epoch": 6.3463414634146345,
+      "grad_norm": 2.6362273693084717,
+      "learning_rate": 1.4773284070716504e-06,
+      "loss": 0.041,
+      "step": 1301
+    },
+    {
+      "epoch": 6.351219512195122,
+      "grad_norm": 2.1512129306793213,
+      "learning_rate": 1.473833615830722e-06,
+      "loss": 0.0227,
+      "step": 1302
+    },
+    {
+      "epoch": 6.35609756097561,
+      "grad_norm": 2.2898178100585938,
+      "learning_rate": 1.4703412345491692e-06,
+      "loss": 0.039,
+      "step": 1303
+    },
+    {
+      "epoch": 6.360975609756098,
+      "grad_norm": 2.6641080379486084,
+      "learning_rate": 1.4668512714288763e-06,
+      "loss": 0.0431,
+      "step": 1304
+    },
+    {
+      "epoch": 6.365853658536586,
+      "grad_norm": 1.7466667890548706,
+      "learning_rate": 1.4633637346660478e-06,
+      "loss": 0.013,
+      "step": 1305
+    },
+    {
+      "epoch": 6.3707317073170735,
+      "grad_norm": 2.437889575958252,
+      "learning_rate": 1.4598786324511892e-06,
+      "loss": 0.0181,
+      "step": 1306
+    },
+    {
+      "epoch": 6.375609756097561,
+      "grad_norm": 2.5054142475128174,
+      "learning_rate": 1.456395972969089e-06,
+      "loss": 0.0248,
+      "step": 1307
+    },
+    {
+      "epoch": 6.380487804878049,
+      "grad_norm": 3.2294511795043945,
+      "learning_rate": 1.4529157643987995e-06,
+      "loss": 0.0561,
+      "step": 1308
+    },
+    {
+      "epoch": 6.385365853658537,
+      "grad_norm": 2.260188341140747,
+      "learning_rate": 1.4494380149136162e-06,
+      "loss": 0.0593,
+      "step": 1309
+    },
+    {
+      "epoch": 6.390243902439025,
+      "grad_norm": 2.4961163997650146,
+      "learning_rate": 1.4459627326810576e-06,
+      "loss": 0.0257,
+      "step": 1310
+    },
+    {
+      "epoch": 6.3951219512195125,
+      "grad_norm": 3.4153239727020264,
+      "learning_rate": 1.4424899258628533e-06,
+      "loss": 0.0223,
+      "step": 1311
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 2.6308839321136475,
+      "learning_rate": 1.439019602614914e-06,
+      "loss": 0.0112,
+      "step": 1312
+    },
+    {
+      "epoch": 6.404878048780488,
+      "grad_norm": 2.754530191421509,
+      "learning_rate": 1.4355517710873184e-06,
+      "loss": 0.068,
+      "step": 1313
+    },
+    {
+      "epoch": 6.409756097560976,
+      "grad_norm": 4.473151683807373,
+      "learning_rate": 1.432086439424297e-06,
+      "loss": 0.0825,
+      "step": 1314
+    },
+    {
+      "epoch": 6.414634146341464,
+      "grad_norm": 4.85701322555542,
+      "learning_rate": 1.428623615764206e-06,
+      "loss": 0.1812,
+      "step": 1315
+    },
+    {
+      "epoch": 6.419512195121952,
+      "grad_norm": 1.6678224802017212,
+      "learning_rate": 1.4251633082395117e-06,
+      "loss": 0.0207,
+      "step": 1316
+    },
+    {
+      "epoch": 6.424390243902439,
+      "grad_norm": 2.9730937480926514,
+      "learning_rate": 1.4217055249767734e-06,
+      "loss": 0.0617,
+      "step": 1317
+    },
+    {
+      "epoch": 6.429268292682927,
+      "grad_norm": 2.503786563873291,
+      "learning_rate": 1.4182502740966203e-06,
+      "loss": 0.0137,
+      "step": 1318
+    },
+    {
+      "epoch": 6.434146341463415,
+      "grad_norm": 3.0798017978668213,
+      "learning_rate": 1.4147975637137334e-06,
+      "loss": 0.0329,
+      "step": 1319
+    },
+    {
+      "epoch": 6.439024390243903,
+      "grad_norm": 3.008155345916748,
+      "learning_rate": 1.411347401936831e-06,
+      "loss": 0.0487,
+      "step": 1320
+    },
+    {
+      "epoch": 6.443902439024391,
+      "grad_norm": 2.5451765060424805,
+      "learning_rate": 1.4078997968686425e-06,
+      "loss": 0.0582,
+      "step": 1321
+    },
+    {
+      "epoch": 6.4487804878048784,
+      "grad_norm": 2.042696475982666,
+      "learning_rate": 1.404454756605893e-06,
+      "loss": 0.0336,
+      "step": 1322
+    },
+    {
+      "epoch": 6.453658536585366,
+      "grad_norm": 3.0421411991119385,
+      "learning_rate": 1.4010122892392872e-06,
+      "loss": 0.1372,
+      "step": 1323
+    },
+    {
+      "epoch": 6.458536585365854,
+      "grad_norm": 2.0793251991271973,
+      "learning_rate": 1.3975724028534842e-06,
+      "loss": 0.0452,
+      "step": 1324
+    },
+    {
+      "epoch": 6.463414634146342,
+      "grad_norm": 2.6149914264678955,
+      "learning_rate": 1.394135105527083e-06,
+      "loss": 0.0431,
+      "step": 1325
+    },
+    {
+      "epoch": 6.46829268292683,
+      "grad_norm": 2.818507671356201,
+      "learning_rate": 1.3907004053326006e-06,
+      "loss": 0.0242,
+      "step": 1326
+    },
+    {
+      "epoch": 6.473170731707317,
+      "grad_norm": 2.328993558883667,
+      "learning_rate": 1.387268310336458e-06,
+      "loss": 0.0293,
+      "step": 1327
+    },
+    {
+      "epoch": 6.478048780487805,
+      "grad_norm": 2.2032642364501953,
+      "learning_rate": 1.3838388285989552e-06,
+      "loss": 0.0232,
+      "step": 1328
+    },
+    {
+      "epoch": 6.482926829268292,
+      "grad_norm": 2.039983034133911,
+      "learning_rate": 1.380411968174254e-06,
+      "loss": 0.0256,
+      "step": 1329
+    },
+    {
+      "epoch": 6.487804878048781,
+      "grad_norm": 3.7261271476745605,
+      "learning_rate": 1.3769877371103635e-06,
+      "loss": 0.1285,
+      "step": 1330
+    },
+    {
+      "epoch": 6.492682926829268,
+      "grad_norm": 3.7156264781951904,
+      "learning_rate": 1.373566143449115e-06,
+      "loss": 0.1621,
+      "step": 1331
+    },
+    {
+      "epoch": 6.4975609756097565,
+      "grad_norm": 1.5905455350875854,
+      "learning_rate": 1.3701471952261457e-06,
+      "loss": 0.0126,
+      "step": 1332
+    },
+    {
+      "epoch": 6.5024390243902435,
+      "grad_norm": 2.8808465003967285,
+      "learning_rate": 1.3667309004708832e-06,
+      "loss": 0.0211,
+      "step": 1333
+    },
+    {
+      "epoch": 6.507317073170732,
+      "grad_norm": 3.9190757274627686,
+      "learning_rate": 1.3633172672065195e-06,
+      "loss": 0.062,
+      "step": 1334
+    },
+    {
+      "epoch": 6.512195121951219,
+      "grad_norm": 1.6948635578155518,
+      "learning_rate": 1.359906303449997e-06,
+      "loss": 0.0126,
+      "step": 1335
+    },
+    {
+      "epoch": 6.517073170731708,
+      "grad_norm": 2.3967642784118652,
+      "learning_rate": 1.3564980172119913e-06,
+      "loss": 0.0111,
+      "step": 1336
+    },
+    {
+      "epoch": 6.521951219512195,
+      "grad_norm": 3.5275399684906006,
+      "learning_rate": 1.3530924164968873e-06,
+      "loss": 0.1024,
+      "step": 1337
+    },
+    {
+      "epoch": 6.526829268292683,
+      "grad_norm": 2.0768814086914062,
+      "learning_rate": 1.3496895093027617e-06,
+      "loss": 0.0254,
+      "step": 1338
+    },
+    {
+      "epoch": 6.53170731707317,
+      "grad_norm": 1.8964029550552368,
+      "learning_rate": 1.3462893036213706e-06,
+      "loss": 0.0188,
+      "step": 1339
+    },
+    {
+      "epoch": 6.536585365853659,
+      "grad_norm": 1.679545283317566,
+      "learning_rate": 1.3428918074381203e-06,
+      "loss": 0.0195,
+      "step": 1340
+    },
+    {
+      "epoch": 6.541463414634146,
+      "grad_norm": 2.204637050628662,
+      "learning_rate": 1.3394970287320553e-06,
+      "loss": 0.0317,
+      "step": 1341
+    },
+    {
+      "epoch": 6.546341463414635,
+      "grad_norm": 2.014052629470825,
+      "learning_rate": 1.3361049754758404e-06,
+      "loss": 0.0191,
+      "step": 1342
+    },
+    {
+      "epoch": 6.5512195121951216,
+      "grad_norm": 1.4630589485168457,
+      "learning_rate": 1.3327156556357369e-06,
+      "loss": 0.0079,
+      "step": 1343
+    },
+    {
+      "epoch": 6.55609756097561,
+      "grad_norm": 2.876132011413574,
+      "learning_rate": 1.3293290771715875e-06,
+      "loss": 0.0345,
+      "step": 1344
+    },
+    {
+      "epoch": 6.560975609756097,
+      "grad_norm": 1.793338656425476,
+      "learning_rate": 1.3259452480367963e-06,
+      "loss": 0.0409,
+      "step": 1345
+    },
+    {
+      "epoch": 6.565853658536585,
+      "grad_norm": 2.2791552543640137,
+      "learning_rate": 1.3225641761783126e-06,
+      "loss": 0.0494,
+      "step": 1346
+    },
+    {
+      "epoch": 6.570731707317073,
+      "grad_norm": 4.255206108093262,
+      "learning_rate": 1.3191858695366084e-06,
+      "loss": 0.0842,
+      "step": 1347
+    },
+    {
+      "epoch": 6.575609756097561,
+      "grad_norm": 2.449460506439209,
+      "learning_rate": 1.3158103360456603e-06,
+      "loss": 0.0399,
+      "step": 1348
+    },
+    {
+      "epoch": 6.580487804878048,
+      "grad_norm": 2.780730724334717,
+      "learning_rate": 1.3124375836329362e-06,
+      "loss": 0.0272,
+      "step": 1349
+    },
+    {
+      "epoch": 6.585365853658536,
+      "grad_norm": 1.925681233406067,
+      "learning_rate": 1.3090676202193692e-06,
+      "loss": 0.007,
+      "step": 1350
+    },
+    {
+      "epoch": 6.590243902439024,
+      "grad_norm": 2.069791555404663,
+      "learning_rate": 1.3057004537193424e-06,
+      "loss": 0.016,
+      "step": 1351
+    },
+    {
+      "epoch": 6.595121951219512,
+      "grad_norm": 1.863872766494751,
+      "learning_rate": 1.302336092040673e-06,
+      "loss": 0.016,
+      "step": 1352
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 2.351259231567383,
+      "learning_rate": 1.298974543084589e-06,
+      "loss": 0.0172,
+      "step": 1353
+    },
+    {
+      "epoch": 6.6048780487804875,
+      "grad_norm": 1.848115086555481,
+      "learning_rate": 1.2956158147457116e-06,
+      "loss": 0.0412,
+      "step": 1354
+    },
+    {
+      "epoch": 6.609756097560975,
+      "grad_norm": 1.6395928859710693,
+      "learning_rate": 1.2922599149120412e-06,
+      "loss": 0.0181,
+      "step": 1355
+    },
+    {
+      "epoch": 6.614634146341463,
+      "grad_norm": 2.1267426013946533,
+      "learning_rate": 1.2889068514649328e-06,
+      "loss": 0.04,
+      "step": 1356
+    },
+    {
+      "epoch": 6.619512195121951,
+      "grad_norm": 1.6603496074676514,
+      "learning_rate": 1.2855566322790796e-06,
+      "loss": 0.0108,
+      "step": 1357
+    },
+    {
+      "epoch": 6.624390243902439,
+      "grad_norm": 2.2724838256835938,
+      "learning_rate": 1.2822092652224989e-06,
+      "loss": 0.0284,
+      "step": 1358
+    },
+    {
+      "epoch": 6.6292682926829265,
+      "grad_norm": 2.222623825073242,
+      "learning_rate": 1.2788647581565048e-06,
+      "loss": 0.0128,
+      "step": 1359
+    },
+    {
+      "epoch": 6.634146341463414,
+      "grad_norm": 2.710681676864624,
+      "learning_rate": 1.275523118935697e-06,
+      "loss": 0.0184,
+      "step": 1360
+    },
+    {
+      "epoch": 6.639024390243902,
+      "grad_norm": 2.354264736175537,
+      "learning_rate": 1.2721843554079418e-06,
+      "loss": 0.0313,
+      "step": 1361
+    },
+    {
+      "epoch": 6.64390243902439,
+      "grad_norm": 3.886909008026123,
+      "learning_rate": 1.2688484754143493e-06,
+      "loss": 0.1184,
+      "step": 1362
+    },
+    {
+      "epoch": 6.648780487804878,
+      "grad_norm": 3.088468313217163,
+      "learning_rate": 1.2655154867892577e-06,
+      "loss": 0.0353,
+      "step": 1363
+    },
+    {
+      "epoch": 6.6536585365853655,
+      "grad_norm": 2.987576484680176,
+      "learning_rate": 1.2621853973602158e-06,
+      "loss": 0.0349,
+      "step": 1364
+    },
+    {
+      "epoch": 6.658536585365853,
+      "grad_norm": 1.719212293624878,
+      "learning_rate": 1.2588582149479645e-06,
+      "loss": 0.0081,
+      "step": 1365
+    },
+    {
+      "epoch": 6.663414634146341,
+      "grad_norm": 2.1641178131103516,
+      "learning_rate": 1.2555339473664151e-06,
+      "loss": 0.0279,
+      "step": 1366
+    },
+    {
+      "epoch": 6.668292682926829,
+      "grad_norm": 2.9424984455108643,
+      "learning_rate": 1.2522126024226347e-06,
+      "loss": 0.0492,
+      "step": 1367
+    },
+    {
+      "epoch": 6.673170731707317,
+      "grad_norm": 1.961077332496643,
+      "learning_rate": 1.2488941879168278e-06,
+      "loss": 0.0084,
+      "step": 1368
+    },
+    {
+      "epoch": 6.678048780487805,
+      "grad_norm": 2.302565097808838,
+      "learning_rate": 1.2455787116423148e-06,
+      "loss": 0.0486,
+      "step": 1369
+    },
+    {
+      "epoch": 6.682926829268292,
+      "grad_norm": 2.187194347381592,
+      "learning_rate": 1.2422661813855158e-06,
+      "loss": 0.0319,
+      "step": 1370
+    },
+    {
+      "epoch": 6.68780487804878,
+      "grad_norm": 2.0076377391815186,
+      "learning_rate": 1.238956604925934e-06,
+      "loss": 0.016,
+      "step": 1371
+    },
+    {
+      "epoch": 6.692682926829268,
+      "grad_norm": 4.137681484222412,
+      "learning_rate": 1.2356499900361333e-06,
+      "loss": 0.0557,
+      "step": 1372
+    },
+    {
+      "epoch": 6.697560975609756,
+      "grad_norm": 2.0039637088775635,
+      "learning_rate": 1.2323463444817227e-06,
+      "loss": 0.0219,
+      "step": 1373
+    },
+    {
+      "epoch": 6.702439024390244,
+      "grad_norm": 2.943314552307129,
+      "learning_rate": 1.2290456760213405e-06,
+      "loss": 0.0849,
+      "step": 1374
+    },
+    {
+      "epoch": 6.7073170731707314,
+      "grad_norm": 2.715120553970337,
+      "learning_rate": 1.2257479924066296e-06,
+      "loss": 0.0857,
+      "step": 1375
+    },
+    {
+      "epoch": 6.712195121951219,
+      "grad_norm": 3.144104480743408,
+      "learning_rate": 1.2224533013822237e-06,
+      "loss": 0.0648,
+      "step": 1376
+    },
+    {
+      "epoch": 6.717073170731707,
+      "grad_norm": 2.830066680908203,
+      "learning_rate": 1.2191616106857312e-06,
+      "loss": 0.0426,
+      "step": 1377
+    },
+    {
+      "epoch": 6.721951219512195,
+      "grad_norm": 3.1005899906158447,
+      "learning_rate": 1.2158729280477112e-06,
+      "loss": 0.0478,
+      "step": 1378
+    },
+    {
+      "epoch": 6.726829268292683,
+      "grad_norm": 2.2102460861206055,
+      "learning_rate": 1.2125872611916578e-06,
+      "loss": 0.0273,
+      "step": 1379
+    },
+    {
+      "epoch": 6.7317073170731705,
+      "grad_norm": 2.860288619995117,
+      "learning_rate": 1.2093046178339869e-06,
+      "loss": 0.0201,
+      "step": 1380
+    },
+    {
+      "epoch": 6.736585365853658,
+      "grad_norm": 1.5914067029953003,
+      "learning_rate": 1.206025005684009e-06,
+      "loss": 0.0148,
+      "step": 1381
+    },
+    {
+      "epoch": 6.741463414634146,
+      "grad_norm": 1.8609223365783691,
+      "learning_rate": 1.202748432443918e-06,
+      "loss": 0.0073,
+      "step": 1382
+    },
+    {
+      "epoch": 6.746341463414634,
+      "grad_norm": 3.0532407760620117,
+      "learning_rate": 1.1994749058087695e-06,
+      "loss": 0.0344,
+      "step": 1383
+    },
+    {
+      "epoch": 6.751219512195122,
+      "grad_norm": 4.0601677894592285,
+      "learning_rate": 1.196204433466467e-06,
+      "loss": 0.0837,
+      "step": 1384
+    },
+    {
+      "epoch": 6.7560975609756095,
+      "grad_norm": 2.6982672214508057,
+      "learning_rate": 1.192937023097738e-06,
+      "loss": 0.0425,
+      "step": 1385
+    },
+    {
+      "epoch": 6.760975609756097,
+      "grad_norm": 1.431360125541687,
+      "learning_rate": 1.1896726823761195e-06,
+      "loss": 0.0065,
+      "step": 1386
+    },
+    {
+      "epoch": 6.765853658536585,
+      "grad_norm": 2.116907835006714,
+      "learning_rate": 1.1864114189679413e-06,
+      "loss": 0.0133,
+      "step": 1387
+    },
+    {
+      "epoch": 6.770731707317073,
+      "grad_norm": 2.6869874000549316,
+      "learning_rate": 1.183153240532304e-06,
+      "loss": 0.0188,
+      "step": 1388
+    },
+    {
+      "epoch": 6.775609756097561,
+      "grad_norm": 2.0294089317321777,
+      "learning_rate": 1.179898154721063e-06,
+      "loss": 0.0234,
+      "step": 1389
+    },
+    {
+      "epoch": 6.780487804878049,
+      "grad_norm": 2.3081958293914795,
+      "learning_rate": 1.1766461691788137e-06,
+      "loss": 0.0208,
+      "step": 1390
+    },
+    {
+      "epoch": 6.785365853658536,
+      "grad_norm": 3.4795000553131104,
+      "learning_rate": 1.1733972915428665e-06,
+      "loss": 0.0728,
+      "step": 1391
+    },
+    {
+      "epoch": 6.790243902439024,
+      "grad_norm": 2.5121219158172607,
+      "learning_rate": 1.1701515294432348e-06,
+      "loss": 0.0291,
+      "step": 1392
+    },
+    {
+      "epoch": 6.795121951219512,
+      "grad_norm": 5.1100172996521,
+      "learning_rate": 1.1669088905026156e-06,
+      "loss": 0.0988,
+      "step": 1393
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 2.5434396266937256,
+      "learning_rate": 1.163669382336371e-06,
+      "loss": 0.0399,
+      "step": 1394
+    },
+    {
+      "epoch": 6.804878048780488,
+      "grad_norm": 2.7811660766601562,
+      "learning_rate": 1.160433012552508e-06,
+      "loss": 0.0134,
+      "step": 1395
+    },
+    {
+      "epoch": 6.809756097560975,
+      "grad_norm": 3.2409870624542236,
+      "learning_rate": 1.1571997887516672e-06,
+      "loss": 0.0795,
+      "step": 1396
+    },
+    {
+      "epoch": 6.814634146341463,
+      "grad_norm": 2.5300986766815186,
+      "learning_rate": 1.1539697185270982e-06,
+      "loss": 0.0329,
+      "step": 1397
+    },
+    {
+      "epoch": 6.819512195121951,
+      "grad_norm": 1.8510549068450928,
+      "learning_rate": 1.1507428094646448e-06,
+      "loss": 0.0213,
+      "step": 1398
+    },
+    {
+      "epoch": 6.824390243902439,
+      "grad_norm": 1.8820618391036987,
+      "learning_rate": 1.1475190691427255e-06,
+      "loss": 0.0172,
+      "step": 1399
+    },
+    {
+      "epoch": 6.829268292682927,
+      "grad_norm": 1.3415460586547852,
+      "learning_rate": 1.1442985051323205e-06,
+      "loss": 0.0029,
+      "step": 1400
+    },
+    {
+      "epoch": 6.8341463414634145,
+      "grad_norm": 6.033786296844482,
+      "learning_rate": 1.1410811249969475e-06,
+      "loss": 0.1638,
+      "step": 1401
+    },
+    {
+      "epoch": 6.839024390243902,
+      "grad_norm": 2.990328311920166,
+      "learning_rate": 1.1378669362926468e-06,
+      "loss": 0.0779,
+      "step": 1402
+    },
+    {
+      "epoch": 6.84390243902439,
+      "grad_norm": 3.2766308784484863,
+      "learning_rate": 1.1346559465679656e-06,
+      "loss": 0.0528,
+      "step": 1403
+    },
+    {
+      "epoch": 6.848780487804878,
+      "grad_norm": 1.266032338142395,
+      "learning_rate": 1.1314481633639374e-06,
+      "loss": 0.0057,
+      "step": 1404
+    },
+    {
+      "epoch": 6.853658536585366,
+      "grad_norm": 3.1048431396484375,
+      "learning_rate": 1.1282435942140632e-06,
+      "loss": 0.1772,
+      "step": 1405
+    },
+    {
+      "epoch": 6.8585365853658535,
+      "grad_norm": 2.264822483062744,
+      "learning_rate": 1.1250422466442992e-06,
+      "loss": 0.0176,
+      "step": 1406
+    },
+    {
+      "epoch": 6.863414634146341,
+      "grad_norm": 2.0890846252441406,
+      "learning_rate": 1.1218441281730334e-06,
+      "loss": 0.0184,
+      "step": 1407
+    },
+    {
+      "epoch": 6.868292682926829,
+      "grad_norm": 1.8351202011108398,
+      "learning_rate": 1.1186492463110696e-06,
+      "loss": 0.0127,
+      "step": 1408
+    },
+    {
+      "epoch": 6.873170731707317,
+      "grad_norm": 1.447196125984192,
+      "learning_rate": 1.1154576085616135e-06,
+      "loss": 0.0094,
+      "step": 1409
+    },
+    {
+      "epoch": 6.878048780487805,
+      "grad_norm": 1.6414039134979248,
+      "learning_rate": 1.1122692224202491e-06,
+      "loss": 0.0138,
+      "step": 1410
+    },
+    {
+      "epoch": 6.882926829268293,
+      "grad_norm": 2.87068772315979,
+      "learning_rate": 1.1090840953749253e-06,
+      "loss": 0.0821,
+      "step": 1411
+    },
+    {
+      "epoch": 6.88780487804878,
+      "grad_norm": 2.0476415157318115,
+      "learning_rate": 1.1059022349059362e-06,
+      "loss": 0.0222,
+      "step": 1412
+    },
+    {
+      "epoch": 6.892682926829268,
+      "grad_norm": 4.169386863708496,
+      "learning_rate": 1.102723648485905e-06,
+      "loss": 0.1183,
+      "step": 1413
+    },
+    {
+      "epoch": 6.897560975609756,
+      "grad_norm": 4.47883415222168,
+      "learning_rate": 1.0995483435797643e-06,
+      "loss": 0.0528,
+      "step": 1414
+    },
+    {
+      "epoch": 6.902439024390244,
+      "grad_norm": 2.0025508403778076,
+      "learning_rate": 1.0963763276447435e-06,
+      "loss": 0.0106,
+      "step": 1415
+    },
+    {
+      "epoch": 6.907317073170732,
+      "grad_norm": 2.4212136268615723,
+      "learning_rate": 1.0932076081303442e-06,
+      "loss": 0.0454,
+      "step": 1416
+    },
+    {
+      "epoch": 6.912195121951219,
+      "grad_norm": 1.7873961925506592,
+      "learning_rate": 1.0900421924783272e-06,
+      "loss": 0.022,
+      "step": 1417
+    },
+    {
+      "epoch": 6.917073170731707,
+      "grad_norm": 2.0345218181610107,
+      "learning_rate": 1.0868800881226962e-06,
+      "loss": 0.0261,
+      "step": 1418
+    },
+    {
+      "epoch": 6.921951219512195,
+      "grad_norm": 3.086538314819336,
+      "learning_rate": 1.0837213024896764e-06,
+      "loss": 0.0257,
+      "step": 1419
+    },
+    {
+      "epoch": 6.926829268292683,
+      "grad_norm": 2.9401397705078125,
+      "learning_rate": 1.080565842997698e-06,
+      "loss": 0.087,
+      "step": 1420
+    },
+    {
+      "epoch": 6.931707317073171,
+      "grad_norm": 1.305415153503418,
+      "learning_rate": 1.0774137170573826e-06,
+      "loss": 0.0147,
+      "step": 1421
+    },
+    {
+      "epoch": 6.9365853658536585,
+      "grad_norm": 3.0256683826446533,
+      "learning_rate": 1.074264932071521e-06,
+      "loss": 0.1183,
+      "step": 1422
+    },
+    {
+      "epoch": 6.941463414634146,
+      "grad_norm": 2.3618743419647217,
+      "learning_rate": 1.0711194954350568e-06,
+      "loss": 0.0186,
+      "step": 1423
+    },
+    {
+      "epoch": 6.946341463414634,
+      "grad_norm": 2.004451036453247,
+      "learning_rate": 1.0679774145350735e-06,
+      "loss": 0.0222,
+      "step": 1424
+    },
+    {
+      "epoch": 6.951219512195122,
+      "grad_norm": 3.089723587036133,
+      "learning_rate": 1.0648386967507703e-06,
+      "loss": 0.0824,
+      "step": 1425
+    },
+    {
+      "epoch": 6.95609756097561,
+      "grad_norm": 1.9310235977172852,
+      "learning_rate": 1.0617033494534486e-06,
+      "loss": 0.0247,
+      "step": 1426
+    },
+    {
+      "epoch": 6.9609756097560975,
+      "grad_norm": 1.973836898803711,
+      "learning_rate": 1.0585713800064964e-06,
+      "loss": 0.0142,
+      "step": 1427
+    },
+    {
+      "epoch": 6.965853658536585,
+      "grad_norm": 2.9914112091064453,
+      "learning_rate": 1.0554427957653663e-06,
+      "loss": 0.0681,
+      "step": 1428
+    },
+    {
+      "epoch": 6.970731707317073,
+      "grad_norm": 3.356689691543579,
+      "learning_rate": 1.0523176040775615e-06,
+      "loss": 0.0916,
+      "step": 1429
+    },
+    {
+      "epoch": 6.975609756097561,
+      "grad_norm": 2.3305246829986572,
+      "learning_rate": 1.0491958122826173e-06,
+      "loss": 0.0611,
+      "step": 1430
+    },
+    {
+      "epoch": 6.980487804878049,
+      "grad_norm": 1.7383835315704346,
+      "learning_rate": 1.0460774277120866e-06,
+      "loss": 0.0182,
+      "step": 1431
+    },
+    {
+      "epoch": 6.985365853658537,
+      "grad_norm": 2.585674524307251,
+      "learning_rate": 1.0429624576895177e-06,
+      "loss": 0.0084,
+      "step": 1432
+    },
+    {
+      "epoch": 6.990243902439024,
+      "grad_norm": 3.023864269256592,
+      "learning_rate": 1.03985090953044e-06,
+      "loss": 0.0411,
+      "step": 1433
+    },
+    {
+      "epoch": 6.995121951219512,
+      "grad_norm": 2.281674861907959,
+      "learning_rate": 1.0367427905423497e-06,
+      "loss": 0.0464,
+      "step": 1434
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 1.4372339248657227,
+      "learning_rate": 1.0336381080246858e-06,
+      "loss": 0.0124,
+      "step": 1435
+    },
+    {
+      "epoch": 7.004878048780488,
+      "grad_norm": 1.9526969194412231,
+      "learning_rate": 1.0305368692688175e-06,
+      "loss": 0.0179,
+      "step": 1436
+    },
+    {
+      "epoch": 7.009756097560976,
+      "grad_norm": 1.7297903299331665,
+      "learning_rate": 1.027439081558029e-06,
+      "loss": 0.0119,
+      "step": 1437
+    },
+    {
+      "epoch": 7.014634146341463,
+      "grad_norm": 2.2754275798797607,
+      "learning_rate": 1.0243447521674967e-06,
+      "loss": 0.0278,
+      "step": 1438
+    },
+    {
+      "epoch": 7.019512195121951,
+      "grad_norm": 5.485769271850586,
+      "learning_rate": 1.021253888364276e-06,
+      "loss": 0.1259,
+      "step": 1439
+    },
+    {
+      "epoch": 7.024390243902439,
+      "grad_norm": 0.9085121750831604,
+      "learning_rate": 1.018166497407284e-06,
+      "loss": 0.0047,
+      "step": 1440
+    },
+    {
+      "epoch": 7.029268292682927,
+      "grad_norm": 1.0291047096252441,
+      "learning_rate": 1.0150825865472813e-06,
+      "loss": 0.0044,
+      "step": 1441
+    },
+    {
+      "epoch": 7.034146341463415,
+      "grad_norm": 0.8040009140968323,
+      "learning_rate": 1.0120021630268542e-06,
+      "loss": 0.0044,
+      "step": 1442
+    },
+    {
+      "epoch": 7.0390243902439025,
+      "grad_norm": 1.3701342344284058,
+      "learning_rate": 1.0089252340804025e-06,
+      "loss": 0.0081,
+      "step": 1443
+    },
+    {
+      "epoch": 7.04390243902439,
+      "grad_norm": 2.89591646194458,
+      "learning_rate": 1.0058518069341152e-06,
+      "loss": 0.0318,
+      "step": 1444
+    },
+    {
+      "epoch": 7.048780487804878,
+      "grad_norm": 1.3153692483901978,
+      "learning_rate": 1.002781888805958e-06,
+      "loss": 0.0067,
+      "step": 1445
+    },
+    {
+      "epoch": 7.053658536585366,
+      "grad_norm": 1.4490022659301758,
+      "learning_rate": 9.997154869056588e-07,
+      "loss": 0.0064,
+      "step": 1446
+    },
+    {
+      "epoch": 7.058536585365854,
+      "grad_norm": 1.7938638925552368,
+      "learning_rate": 9.966526084346837e-07,
+      "loss": 0.0057,
+      "step": 1447
+    },
+    {
+      "epoch": 7.0634146341463415,
+      "grad_norm": 3.7182836532592773,
+      "learning_rate": 9.935932605862258e-07,
+      "loss": 0.0365,
+      "step": 1448
+    },
+    {
+      "epoch": 7.068292682926829,
+      "grad_norm": 1.7843579053878784,
+      "learning_rate": 9.905374505451853e-07,
+      "loss": 0.0345,
+      "step": 1449
+    },
+    {
+      "epoch": 7.073170731707317,
+      "grad_norm": 2.9557483196258545,
+      "learning_rate": 9.874851854881565e-07,
+      "loss": 0.0384,
+      "step": 1450
+    },
+    {
+      "epoch": 7.078048780487805,
+      "grad_norm": 1.6237356662750244,
+      "learning_rate": 9.844364725834058e-07,
+      "loss": 0.0116,
+      "step": 1451
+    },
+    {
+      "epoch": 7.082926829268293,
+      "grad_norm": 3.7120912075042725,
+      "learning_rate": 9.813913189908571e-07,
+      "loss": 0.0267,
+      "step": 1452
+    },
+    {
+      "epoch": 7.087804878048781,
+      "grad_norm": 1.9991087913513184,
+      "learning_rate": 9.783497318620783e-07,
+      "loss": 0.0376,
+      "step": 1453
+    },
+    {
+      "epoch": 7.092682926829268,
+      "grad_norm": 1.5474026203155518,
+      "learning_rate": 9.75311718340258e-07,
+      "loss": 0.0057,
+      "step": 1454
+    },
+    {
+      "epoch": 7.097560975609756,
+      "grad_norm": 2.060807943344116,
+      "learning_rate": 9.722772855601927e-07,
+      "loss": 0.0386,
+      "step": 1455
+    },
+    {
+      "epoch": 7.102439024390244,
+      "grad_norm": 1.1991411447525024,
+      "learning_rate": 9.692464406482727e-07,
+      "loss": 0.006,
+      "step": 1456
+    },
+    {
+      "epoch": 7.107317073170732,
+      "grad_norm": 1.8907703161239624,
+      "learning_rate": 9.662191907224582e-07,
+      "loss": 0.0066,
+      "step": 1457
+    },
+    {
+      "epoch": 7.11219512195122,
+      "grad_norm": 2.0351309776306152,
+      "learning_rate": 9.63195542892268e-07,
+      "loss": 0.0201,
+      "step": 1458
+    },
+    {
+      "epoch": 7.117073170731707,
+      "grad_norm": 1.3973944187164307,
+      "learning_rate": 9.601755042587624e-07,
+      "loss": 0.0112,
+      "step": 1459
+    },
+    {
+      "epoch": 7.121951219512195,
+      "grad_norm": 1.3639394044876099,
+      "learning_rate": 9.571590819145244e-07,
+      "loss": 0.0066,
+      "step": 1460
+    },
+    {
+      "epoch": 7.126829268292683,
+      "grad_norm": 1.7362885475158691,
+      "learning_rate": 9.541462829436426e-07,
+      "loss": 0.0136,
+      "step": 1461
+    },
+    {
+      "epoch": 7.131707317073171,
+      "grad_norm": 2.9414384365081787,
+      "learning_rate": 9.511371144217005e-07,
+      "loss": 0.0228,
+      "step": 1462
+    },
+    {
+      "epoch": 7.136585365853659,
+      "grad_norm": 2.944575548171997,
+      "learning_rate": 9.481315834157512e-07,
+      "loss": 0.027,
+      "step": 1463
+    },
+    {
+      "epoch": 7.1414634146341465,
+      "grad_norm": 2.4692747592926025,
+      "learning_rate": 9.451296969843058e-07,
+      "loss": 0.0152,
+      "step": 1464
+    },
+    {
+      "epoch": 7.146341463414634,
+      "grad_norm": 1.804129719734192,
+      "learning_rate": 9.42131462177319e-07,
+      "loss": 0.0071,
+      "step": 1465
+    },
+    {
+      "epoch": 7.151219512195122,
+      "grad_norm": 1.8012168407440186,
+      "learning_rate": 9.39136886036166e-07,
+      "loss": 0.0054,
+      "step": 1466
+    },
+    {
+      "epoch": 7.15609756097561,
+      "grad_norm": 1.9471648931503296,
+      "learning_rate": 9.361459755936316e-07,
+      "loss": 0.0067,
+      "step": 1467
+    },
+    {
+      "epoch": 7.160975609756098,
+      "grad_norm": 1.8837870359420776,
+      "learning_rate": 9.331587378738902e-07,
+      "loss": 0.0105,
+      "step": 1468
+    },
+    {
+      "epoch": 7.1658536585365855,
+      "grad_norm": 2.358891487121582,
+      "learning_rate": 9.301751798924935e-07,
+      "loss": 0.0331,
+      "step": 1469
+    },
+    {
+      "epoch": 7.170731707317073,
+      "grad_norm": 1.1501671075820923,
+      "learning_rate": 9.27195308656349e-07,
+      "loss": 0.0076,
+      "step": 1470
+    },
+    {
+      "epoch": 7.175609756097561,
+      "grad_norm": 2.3329083919525146,
+      "learning_rate": 9.24219131163705e-07,
+      "loss": 0.0243,
+      "step": 1471
+    },
+    {
+      "epoch": 7.180487804878049,
+      "grad_norm": 1.6030691862106323,
+      "learning_rate": 9.212466544041385e-07,
+      "loss": 0.0051,
+      "step": 1472
+    },
+    {
+      "epoch": 7.185365853658537,
+      "grad_norm": 2.005582094192505,
+      "learning_rate": 9.182778853585325e-07,
+      "loss": 0.0146,
+      "step": 1473
+    },
+    {
+      "epoch": 7.190243902439025,
+      "grad_norm": 1.86012601852417,
+      "learning_rate": 9.153128309990622e-07,
+      "loss": 0.0273,
+      "step": 1474
+    },
+    {
+      "epoch": 7.195121951219512,
+      "grad_norm": 2.218923568725586,
+      "learning_rate": 9.123514982891813e-07,
+      "loss": 0.0225,
+      "step": 1475
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 1.9950376749038696,
+      "learning_rate": 9.093938941836012e-07,
+      "loss": 0.0156,
+      "step": 1476
+    },
+    {
+      "epoch": 7.204878048780488,
+      "grad_norm": 1.6428661346435547,
+      "learning_rate": 9.064400256282757e-07,
+      "loss": 0.0158,
+      "step": 1477
+    },
+    {
+      "epoch": 7.209756097560976,
+      "grad_norm": 1.7983390092849731,
+      "learning_rate": 9.034898995603894e-07,
+      "loss": 0.0138,
+      "step": 1478
+    },
+    {
+      "epoch": 7.214634146341464,
+      "grad_norm": 2.2069218158721924,
+      "learning_rate": 9.00543522908334e-07,
+      "loss": 0.0308,
+      "step": 1479
+    },
+    {
+      "epoch": 7.219512195121951,
+      "grad_norm": 1.4668920040130615,
+      "learning_rate": 8.976009025916962e-07,
+      "loss": 0.006,
+      "step": 1480
+    },
+    {
+      "epoch": 7.224390243902439,
+      "grad_norm": 1.8956354856491089,
+      "learning_rate": 8.946620455212438e-07,
+      "loss": 0.0121,
+      "step": 1481
+    },
+    {
+      "epoch": 7.229268292682927,
+      "grad_norm": 2.5479676723480225,
+      "learning_rate": 8.917269585989027e-07,
+      "loss": 0.0424,
+      "step": 1482
+    },
+    {
+      "epoch": 7.234146341463415,
+      "grad_norm": 1.7482987642288208,
+      "learning_rate": 8.887956487177462e-07,
+      "loss": 0.0189,
+      "step": 1483
+    },
+    {
+      "epoch": 7.239024390243903,
+      "grad_norm": 1.5023657083511353,
+      "learning_rate": 8.858681227619789e-07,
+      "loss": 0.0118,
+      "step": 1484
+    },
+    {
+      "epoch": 7.2439024390243905,
+      "grad_norm": 1.2069121599197388,
+      "learning_rate": 8.829443876069163e-07,
+      "loss": 0.0043,
+      "step": 1485
+    },
+    {
+      "epoch": 7.248780487804878,
+      "grad_norm": 1.5843572616577148,
+      "learning_rate": 8.800244501189722e-07,
+      "loss": 0.0111,
+      "step": 1486
+    },
+    {
+      "epoch": 7.253658536585366,
+      "grad_norm": 2.541588544845581,
+      "learning_rate": 8.771083171556407e-07,
+      "loss": 0.0582,
+      "step": 1487
+    },
+    {
+      "epoch": 7.258536585365854,
+      "grad_norm": 0.9306992292404175,
+      "learning_rate": 8.741959955654833e-07,
+      "loss": 0.0051,
+      "step": 1488
+    },
+    {
+      "epoch": 7.263414634146342,
+      "grad_norm": 1.4105901718139648,
+      "learning_rate": 8.712874921881082e-07,
+      "loss": 0.0175,
+      "step": 1489
+    },
+    {
+      "epoch": 7.2682926829268295,
+      "grad_norm": 2.8943028450012207,
+      "learning_rate": 8.683828138541559e-07,
+      "loss": 0.0827,
+      "step": 1490
+    },
+    {
+      "epoch": 7.273170731707317,
+      "grad_norm": 2.512991428375244,
+      "learning_rate": 8.654819673852874e-07,
+      "loss": 0.0347,
+      "step": 1491
+    },
+    {
+      "epoch": 7.278048780487805,
+      "grad_norm": 1.6571681499481201,
+      "learning_rate": 8.625849595941608e-07,
+      "loss": 0.0055,
+      "step": 1492
+    },
+    {
+      "epoch": 7.282926829268293,
+      "grad_norm": 1.3162294626235962,
+      "learning_rate": 8.596917972844199e-07,
+      "loss": 0.0043,
+      "step": 1493
+    },
+    {
+      "epoch": 7.287804878048781,
+      "grad_norm": 1.761405110359192,
+      "learning_rate": 8.568024872506792e-07,
+      "loss": 0.0176,
+      "step": 1494
+    },
+    {
+      "epoch": 7.2926829268292686,
+      "grad_norm": 0.7546011805534363,
+      "learning_rate": 8.539170362785043e-07,
+      "loss": 0.0025,
+      "step": 1495
+    },
+    {
+      "epoch": 7.297560975609756,
+      "grad_norm": 1.6910885572433472,
+      "learning_rate": 8.510354511443975e-07,
+      "loss": 0.0093,
+      "step": 1496
+    },
+    {
+      "epoch": 7.302439024390244,
+      "grad_norm": 1.6627765893936157,
+      "learning_rate": 8.48157738615784e-07,
+      "loss": 0.0066,
+      "step": 1497
+    },
+    {
+      "epoch": 7.307317073170732,
+      "grad_norm": 0.8881242871284485,
+      "learning_rate": 8.452839054509926e-07,
+      "loss": 0.0055,
+      "step": 1498
+    },
+    {
+      "epoch": 7.31219512195122,
+      "grad_norm": 1.0791494846343994,
+      "learning_rate": 8.42413958399241e-07,
+      "loss": 0.0059,
+      "step": 1499
+    },
+    {
+      "epoch": 7.317073170731708,
+      "grad_norm": 1.5198945999145508,
+      "learning_rate": 8.39547904200623e-07,
+      "loss": 0.0049,
+      "step": 1500
+    },
+    {
+      "epoch": 7.321951219512195,
+      "grad_norm": 1.7168906927108765,
+      "learning_rate": 8.366857495860869e-07,
+      "loss": 0.0204,
+      "step": 1501
+    },
+    {
+      "epoch": 7.326829268292683,
+      "grad_norm": 1.70030677318573,
+      "learning_rate": 8.338275012774247e-07,
+      "loss": 0.0161,
+      "step": 1502
+    },
+    {
+      "epoch": 7.331707317073171,
+      "grad_norm": 2.1044130325317383,
+      "learning_rate": 8.309731659872522e-07,
+      "loss": 0.0088,
+      "step": 1503
+    },
+    {
+      "epoch": 7.336585365853659,
+      "grad_norm": 1.5040123462677002,
+      "learning_rate": 8.281227504189992e-07,
+      "loss": 0.0204,
+      "step": 1504
+    },
+    {
+      "epoch": 7.341463414634147,
+      "grad_norm": 1.6814212799072266,
+      "learning_rate": 8.252762612668869e-07,
+      "loss": 0.0238,
+      "step": 1505
+    },
+    {
+      "epoch": 7.3463414634146345,
+      "grad_norm": 2.2541606426239014,
+      "learning_rate": 8.224337052159154e-07,
+      "loss": 0.0063,
+      "step": 1506
+    },
+    {
+      "epoch": 7.351219512195122,
+      "grad_norm": 2.3999500274658203,
+      "learning_rate": 8.195950889418503e-07,
+      "loss": 0.0123,
+      "step": 1507
+    },
+    {
+      "epoch": 7.35609756097561,
+      "grad_norm": 2.8464221954345703,
+      "learning_rate": 8.167604191112021e-07,
+      "loss": 0.0296,
+      "step": 1508
+    },
+    {
+      "epoch": 7.360975609756098,
+      "grad_norm": 2.178104877471924,
+      "learning_rate": 8.139297023812131e-07,
+      "loss": 0.0148,
+      "step": 1509
+    },
+    {
+      "epoch": 7.365853658536586,
+      "grad_norm": 1.6489804983139038,
+      "learning_rate": 8.111029453998448e-07,
+      "loss": 0.0057,
+      "step": 1510
+    },
+    {
+      "epoch": 7.3707317073170735,
+      "grad_norm": 1.9705169200897217,
+      "learning_rate": 8.082801548057553e-07,
+      "loss": 0.0098,
+      "step": 1511
+    },
+    {
+      "epoch": 7.375609756097561,
+      "grad_norm": 1.2231075763702393,
+      "learning_rate": 8.05461337228289e-07,
+      "loss": 0.007,
+      "step": 1512
+    },
+    {
+      "epoch": 7.380487804878049,
+      "grad_norm": 1.5212552547454834,
+      "learning_rate": 8.026464992874617e-07,
+      "loss": 0.0058,
+      "step": 1513
+    },
+    {
+      "epoch": 7.385365853658537,
+      "grad_norm": 0.5752282738685608,
+      "learning_rate": 7.998356475939398e-07,
+      "loss": 0.0011,
+      "step": 1514
+    },
+    {
+      "epoch": 7.390243902439025,
+      "grad_norm": 1.3227447271347046,
+      "learning_rate": 7.970287887490289e-07,
+      "loss": 0.0041,
+      "step": 1515
+    },
+    {
+      "epoch": 7.3951219512195125,
+      "grad_norm": 1.2051570415496826,
+      "learning_rate": 7.942259293446594e-07,
+      "loss": 0.0027,
+      "step": 1516
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 1.4740777015686035,
+      "learning_rate": 7.914270759633669e-07,
+      "loss": 0.006,
+      "step": 1517
+    },
+    {
+      "epoch": 7.404878048780488,
+      "grad_norm": 1.8853001594543457,
+      "learning_rate": 7.886322351782782e-07,
+      "loss": 0.0066,
+      "step": 1518
+    },
+    {
+      "epoch": 7.409756097560976,
+      "grad_norm": 1.907251238822937,
+      "learning_rate": 7.858414135530995e-07,
+      "loss": 0.0133,
+      "step": 1519
+    },
+    {
+      "epoch": 7.414634146341464,
+      "grad_norm": 1.3397895097732544,
+      "learning_rate": 7.83054617642095e-07,
+      "loss": 0.0092,
+      "step": 1520
+    },
+    {
+      "epoch": 7.419512195121952,
+      "grad_norm": 2.878927707672119,
+      "learning_rate": 7.802718539900761e-07,
+      "loss": 0.0113,
+      "step": 1521
+    },
+    {
+      "epoch": 7.424390243902439,
+      "grad_norm": 1.0312106609344482,
+      "learning_rate": 7.774931291323826e-07,
+      "loss": 0.0045,
+      "step": 1522
+    },
+    {
+      "epoch": 7.429268292682927,
+      "grad_norm": 2.2703888416290283,
+      "learning_rate": 7.747184495948723e-07,
+      "loss": 0.0692,
+      "step": 1523
+    },
+    {
+      "epoch": 7.434146341463415,
+      "grad_norm": 3.0323078632354736,
+      "learning_rate": 7.719478218939e-07,
+      "loss": 0.0462,
+      "step": 1524
+    },
+    {
+      "epoch": 7.439024390243903,
+      "grad_norm": 1.4211952686309814,
+      "learning_rate": 7.691812525363044e-07,
+      "loss": 0.008,
+      "step": 1525
+    },
+    {
+      "epoch": 7.443902439024391,
+      "grad_norm": 0.9588236808776855,
+      "learning_rate": 7.66418748019396e-07,
+      "loss": 0.0042,
+      "step": 1526
+    },
+    {
+      "epoch": 7.4487804878048784,
+      "grad_norm": 2.837219476699829,
+      "learning_rate": 7.636603148309363e-07,
+      "loss": 0.0033,
+      "step": 1527
+    },
+    {
+      "epoch": 7.453658536585366,
+      "grad_norm": 1.8552638292312622,
+      "learning_rate": 7.609059594491253e-07,
+      "loss": 0.0181,
+      "step": 1528
+    },
+    {
+      "epoch": 7.458536585365854,
+      "grad_norm": 4.836069583892822,
+      "learning_rate": 7.581556883425886e-07,
+      "loss": 0.1868,
+      "step": 1529
+    },
+    {
+      "epoch": 7.463414634146342,
+      "grad_norm": 2.180760622024536,
+      "learning_rate": 7.55409507970358e-07,
+      "loss": 0.0305,
+      "step": 1530
+    },
+    {
+      "epoch": 7.46829268292683,
+      "grad_norm": 1.0799378156661987,
+      "learning_rate": 7.526674247818569e-07,
+      "loss": 0.0027,
+      "step": 1531
+    },
+    {
+      "epoch": 7.473170731707317,
+      "grad_norm": 2.1196658611297607,
+      "learning_rate": 7.499294452168904e-07,
+      "loss": 0.019,
+      "step": 1532
+    },
+    {
+      "epoch": 7.478048780487805,
+      "grad_norm": 1.6932553052902222,
+      "learning_rate": 7.471955757056227e-07,
+      "loss": 0.0101,
+      "step": 1533
+    },
+    {
+      "epoch": 7.482926829268292,
+      "grad_norm": 1.3473751544952393,
+      "learning_rate": 7.444658226685656e-07,
+      "loss": 0.0066,
+      "step": 1534
+    },
+    {
+      "epoch": 7.487804878048781,
+      "grad_norm": 2.3404016494750977,
+      "learning_rate": 7.417401925165666e-07,
+      "loss": 0.0139,
+      "step": 1535
+    },
+    {
+      "epoch": 7.492682926829268,
+      "grad_norm": 1.2845433950424194,
+      "learning_rate": 7.390186916507869e-07,
+      "loss": 0.0053,
+      "step": 1536
+    },
+    {
+      "epoch": 7.4975609756097565,
+      "grad_norm": 1.0809649229049683,
+      "learning_rate": 7.363013264626914e-07,
+      "loss": 0.0031,
+      "step": 1537
+    },
+    {
+      "epoch": 7.5024390243902435,
+      "grad_norm": 2.2649292945861816,
+      "learning_rate": 7.335881033340334e-07,
+      "loss": 0.0257,
+      "step": 1538
+    },
+    {
+      "epoch": 7.507317073170732,
+      "grad_norm": 1.3488918542861938,
+      "learning_rate": 7.308790286368373e-07,
+      "loss": 0.0092,
+      "step": 1539
+    },
+    {
+      "epoch": 7.512195121951219,
+      "grad_norm": 2.239190101623535,
+      "learning_rate": 7.281741087333846e-07,
+      "loss": 0.024,
+      "step": 1540
+    },
+    {
+      "epoch": 7.517073170731708,
+      "grad_norm": 1.9454522132873535,
+      "learning_rate": 7.254733499761993e-07,
+      "loss": 0.0177,
+      "step": 1541
+    },
+    {
+      "epoch": 7.521951219512195,
+      "grad_norm": 1.9299415349960327,
+      "learning_rate": 7.22776758708035e-07,
+      "loss": 0.0439,
+      "step": 1542
+    },
+    {
+      "epoch": 7.526829268292683,
+      "grad_norm": 2.2676074504852295,
+      "learning_rate": 7.200843412618555e-07,
+      "loss": 0.0387,
+      "step": 1543
+    },
+    {
+      "epoch": 7.53170731707317,
+      "grad_norm": 1.2385426759719849,
+      "learning_rate": 7.173961039608227e-07,
+      "loss": 0.0082,
+      "step": 1544
+    },
+    {
+      "epoch": 7.536585365853659,
+      "grad_norm": 1.8637615442276,
+      "learning_rate": 7.147120531182828e-07,
+      "loss": 0.0194,
+      "step": 1545
+    },
+    {
+      "epoch": 7.541463414634146,
+      "grad_norm": 1.6695958375930786,
+      "learning_rate": 7.120321950377487e-07,
+      "loss": 0.006,
+      "step": 1546
+    },
+    {
+      "epoch": 7.546341463414635,
+      "grad_norm": 1.916746735572815,
+      "learning_rate": 7.093565360128863e-07,
+      "loss": 0.0104,
+      "step": 1547
+    },
+    {
+      "epoch": 7.5512195121951216,
+      "grad_norm": 1.6002378463745117,
+      "learning_rate": 7.066850823275024e-07,
+      "loss": 0.0173,
+      "step": 1548
+    },
+    {
+      "epoch": 7.55609756097561,
+      "grad_norm": 1.5249438285827637,
+      "learning_rate": 7.040178402555245e-07,
+      "loss": 0.0088,
+      "step": 1549
+    },
+    {
+      "epoch": 7.560975609756097,
+      "grad_norm": 2.1726534366607666,
+      "learning_rate": 7.013548160609901e-07,
+      "loss": 0.0098,
+      "step": 1550
+    },
+    {
+      "epoch": 7.565853658536585,
+      "grad_norm": 1.901904582977295,
+      "learning_rate": 6.986960159980327e-07,
+      "loss": 0.0196,
+      "step": 1551
+    },
+    {
+      "epoch": 7.570731707317073,
+      "grad_norm": 2.577242136001587,
+      "learning_rate": 6.960414463108631e-07,
+      "loss": 0.021,
+      "step": 1552
+    },
+    {
+      "epoch": 7.575609756097561,
+      "grad_norm": 1.4463082551956177,
+      "learning_rate": 6.933911132337575e-07,
+      "loss": 0.0076,
+      "step": 1553
+    },
+    {
+      "epoch": 7.580487804878048,
+      "grad_norm": 2.5811946392059326,
+      "learning_rate": 6.907450229910443e-07,
+      "loss": 0.0204,
+      "step": 1554
+    },
+    {
+      "epoch": 7.585365853658536,
+      "grad_norm": 1.0530297756195068,
+      "learning_rate": 6.881031817970848e-07,
+      "loss": 0.0046,
+      "step": 1555
+    },
+    {
+      "epoch": 7.590243902439024,
+      "grad_norm": 2.995915651321411,
+      "learning_rate": 6.854655958562625e-07,
+      "loss": 0.0566,
+      "step": 1556
+    },
+    {
+      "epoch": 7.595121951219512,
+      "grad_norm": 1.253089189529419,
+      "learning_rate": 6.82832271362969e-07,
+      "loss": 0.0048,
+      "step": 1557
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 2.830667495727539,
+      "learning_rate": 6.802032145015855e-07,
+      "loss": 0.0351,
+      "step": 1558
+    },
+    {
+      "epoch": 7.6048780487804875,
+      "grad_norm": 2.8280539512634277,
+      "learning_rate": 6.775784314464717e-07,
+      "loss": 0.0171,
+      "step": 1559
+    },
+    {
+      "epoch": 7.609756097560975,
+      "grad_norm": 1.7876580953598022,
+      "learning_rate": 6.749579283619492e-07,
+      "loss": 0.01,
+      "step": 1560
+    },
+    {
+      "epoch": 7.614634146341463,
+      "grad_norm": 1.540212869644165,
+      "learning_rate": 6.723417114022907e-07,
+      "loss": 0.0162,
+      "step": 1561
+    },
+    {
+      "epoch": 7.619512195121951,
+      "grad_norm": 2.5126969814300537,
+      "learning_rate": 6.697297867117e-07,
+      "loss": 0.0237,
+      "step": 1562
+    },
+    {
+      "epoch": 7.624390243902439,
+      "grad_norm": 1.5419458150863647,
+      "learning_rate": 6.671221604243014e-07,
+      "loss": 0.0116,
+      "step": 1563
+    },
+    {
+      "epoch": 7.6292682926829265,
+      "grad_norm": 3.469961404800415,
+      "learning_rate": 6.645188386641257e-07,
+      "loss": 0.0506,
+      "step": 1564
+    },
+    {
+      "epoch": 7.634146341463414,
+      "grad_norm": 0.8771130442619324,
+      "learning_rate": 6.61919827545093e-07,
+      "loss": 0.002,
+      "step": 1565
+    },
+    {
+      "epoch": 7.639024390243902,
+      "grad_norm": 3.036559820175171,
+      "learning_rate": 6.593251331709993e-07,
+      "loss": 0.0673,
+      "step": 1566
+    },
+    {
+      "epoch": 7.64390243902439,
+      "grad_norm": 3.379220724105835,
+      "learning_rate": 6.567347616355049e-07,
+      "loss": 0.063,
+      "step": 1567
+    },
+    {
+      "epoch": 7.648780487804878,
+      "grad_norm": 0.7666990756988525,
+      "learning_rate": 6.541487190221163e-07,
+      "loss": 0.003,
+      "step": 1568
+    },
+    {
+      "epoch": 7.6536585365853655,
+      "grad_norm": 1.2181665897369385,
+      "learning_rate": 6.515670114041725e-07,
+      "loss": 0.0037,
+      "step": 1569
+    },
+    {
+      "epoch": 7.658536585365853,
+      "grad_norm": 1.0194541215896606,
+      "learning_rate": 6.489896448448349e-07,
+      "loss": 0.0043,
+      "step": 1570
+    },
+    {
+      "epoch": 7.663414634146341,
+      "grad_norm": 2.2625741958618164,
+      "learning_rate": 6.464166253970672e-07,
+      "loss": 0.0144,
+      "step": 1571
+    },
+    {
+      "epoch": 7.668292682926829,
+      "grad_norm": 1.0256692171096802,
+      "learning_rate": 6.43847959103624e-07,
+      "loss": 0.0029,
+      "step": 1572
+    },
+    {
+      "epoch": 7.673170731707317,
+      "grad_norm": 2.0418128967285156,
+      "learning_rate": 6.412836519970383e-07,
+      "loss": 0.0144,
+      "step": 1573
+    },
+    {
+      "epoch": 7.678048780487805,
+      "grad_norm": 0.8498746752738953,
+      "learning_rate": 6.387237100996041e-07,
+      "loss": 0.0026,
+      "step": 1574
+    },
+    {
+      "epoch": 7.682926829268292,
+      "grad_norm": 1.1043775081634521,
+      "learning_rate": 6.361681394233631e-07,
+      "loss": 0.0093,
+      "step": 1575
+    },
+    {
+      "epoch": 7.68780487804878,
+      "grad_norm": 1.064835786819458,
+      "learning_rate": 6.336169459700933e-07,
+      "loss": 0.0081,
+      "step": 1576
+    },
+    {
+      "epoch": 7.692682926829268,
+      "grad_norm": 1.2024056911468506,
+      "learning_rate": 6.310701357312909e-07,
+      "loss": 0.0054,
+      "step": 1577
+    },
+    {
+      "epoch": 7.697560975609756,
+      "grad_norm": 1.9509804248809814,
+      "learning_rate": 6.285277146881588e-07,
+      "loss": 0.0051,
+      "step": 1578
+    },
+    {
+      "epoch": 7.702439024390244,
+      "grad_norm": 1.8738386631011963,
+      "learning_rate": 6.259896888115904e-07,
+      "loss": 0.0118,
+      "step": 1579
+    },
+    {
+      "epoch": 7.7073170731707314,
+      "grad_norm": 1.356726884841919,
+      "learning_rate": 6.234560640621606e-07,
+      "loss": 0.009,
+      "step": 1580
+    },
+    {
+      "epoch": 7.712195121951219,
+      "grad_norm": 0.6530736684799194,
+      "learning_rate": 6.209268463901047e-07,
+      "loss": 0.0015,
+      "step": 1581
+    },
+    {
+      "epoch": 7.717073170731707,
+      "grad_norm": 1.3714262247085571,
+      "learning_rate": 6.184020417353084e-07,
+      "loss": 0.0051,
+      "step": 1582
+    },
+    {
+      "epoch": 7.721951219512195,
+      "grad_norm": 3.015583038330078,
+      "learning_rate": 6.158816560272962e-07,
+      "loss": 0.0383,
+      "step": 1583
+    },
+    {
+      "epoch": 7.726829268292683,
+      "grad_norm": 3.2355704307556152,
+      "learning_rate": 6.133656951852113e-07,
+      "loss": 0.0422,
+      "step": 1584
+    },
+    {
+      "epoch": 7.7317073170731705,
+      "grad_norm": 1.2933087348937988,
+      "learning_rate": 6.10854165117806e-07,
+      "loss": 0.0082,
+      "step": 1585
+    },
+    {
+      "epoch": 7.736585365853658,
+      "grad_norm": 1.6866157054901123,
+      "learning_rate": 6.083470717234285e-07,
+      "loss": 0.0052,
+      "step": 1586
+    },
+    {
+      "epoch": 7.741463414634146,
+      "grad_norm": 1.4597362279891968,
+      "learning_rate": 6.058444208900061e-07,
+      "loss": 0.0094,
+      "step": 1587
+    },
+    {
+      "epoch": 7.746341463414634,
+      "grad_norm": 0.9200596213340759,
+      "learning_rate": 6.033462184950317e-07,
+      "loss": 0.0034,
+      "step": 1588
+    },
+    {
+      "epoch": 7.751219512195122,
+      "grad_norm": 1.707422137260437,
+      "learning_rate": 6.008524704055535e-07,
+      "loss": 0.0141,
+      "step": 1589
+    },
+    {
+      "epoch": 7.7560975609756095,
+      "grad_norm": 1.8554565906524658,
+      "learning_rate": 5.983631824781572e-07,
+      "loss": 0.0108,
+      "step": 1590
+    },
+    {
+      "epoch": 7.760975609756097,
+      "grad_norm": 1.5421279668807983,
+      "learning_rate": 5.95878360558953e-07,
+      "loss": 0.0075,
+      "step": 1591
+    },
+    {
+      "epoch": 7.765853658536585,
+      "grad_norm": 1.5643326044082642,
+      "learning_rate": 5.933980104835652e-07,
+      "loss": 0.018,
+      "step": 1592
+    },
+    {
+      "epoch": 7.770731707317073,
+      "grad_norm": 1.7024025917053223,
+      "learning_rate": 5.909221380771132e-07,
+      "loss": 0.0207,
+      "step": 1593
+    },
+    {
+      "epoch": 7.775609756097561,
+      "grad_norm": 1.820544719696045,
+      "learning_rate": 5.884507491542024e-07,
+      "loss": 0.0217,
+      "step": 1594
+    },
+    {
+      "epoch": 7.780487804878049,
+      "grad_norm": 1.6761897802352905,
+      "learning_rate": 5.859838495189068e-07,
+      "loss": 0.0055,
+      "step": 1595
+    },
+    {
+      "epoch": 7.785365853658536,
+      "grad_norm": 2.3035616874694824,
+      "learning_rate": 5.835214449647602e-07,
+      "loss": 0.0147,
+      "step": 1596
+    },
+    {
+      "epoch": 7.790243902439024,
+      "grad_norm": 2.0507681369781494,
+      "learning_rate": 5.810635412747373e-07,
+      "loss": 0.0065,
+      "step": 1597
+    },
+    {
+      "epoch": 7.795121951219512,
+      "grad_norm": 1.3789564371109009,
+      "learning_rate": 5.786101442212422e-07,
+      "loss": 0.0077,
+      "step": 1598
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 3.313107490539551,
+      "learning_rate": 5.761612595660979e-07,
+      "loss": 0.0699,
+      "step": 1599
+    },
+    {
+      "epoch": 7.804878048780488,
+      "grad_norm": 1.2391237020492554,
+      "learning_rate": 5.737168930605272e-07,
+      "loss": 0.0017,
+      "step": 1600
+    },
+    {
+      "epoch": 7.809756097560975,
+      "grad_norm": 1.1187714338302612,
+      "learning_rate": 5.712770504451426e-07,
+      "loss": 0.0101,
+      "step": 1601
+    },
+    {
+      "epoch": 7.814634146341463,
+      "grad_norm": 2.7611069679260254,
+      "learning_rate": 5.688417374499336e-07,
+      "loss": 0.0143,
+      "step": 1602
+    },
+    {
+      "epoch": 7.819512195121951,
+      "grad_norm": 1.627295732498169,
+      "learning_rate": 5.664109597942504e-07,
+      "loss": 0.0062,
+      "step": 1603
+    },
+    {
+      "epoch": 7.824390243902439,
+      "grad_norm": 4.538354396820068,
+      "learning_rate": 5.639847231867917e-07,
+      "loss": 0.1058,
+      "step": 1604
+    },
+    {
+      "epoch": 7.829268292682927,
+      "grad_norm": 1.783469319343567,
+      "learning_rate": 5.61563033325594e-07,
+      "loss": 0.0178,
+      "step": 1605
+    },
+    {
+      "epoch": 7.8341463414634145,
+      "grad_norm": 2.259584665298462,
+      "learning_rate": 5.591458958980123e-07,
+      "loss": 0.0204,
+      "step": 1606
+    },
+    {
+      "epoch": 7.839024390243902,
+      "grad_norm": 2.0741965770721436,
+      "learning_rate": 5.567333165807115e-07,
+      "loss": 0.0201,
+      "step": 1607
+    },
+    {
+      "epoch": 7.84390243902439,
+      "grad_norm": 0.8751707077026367,
+      "learning_rate": 5.543253010396538e-07,
+      "loss": 0.0077,
+      "step": 1608
+    },
+    {
+      "epoch": 7.848780487804878,
+      "grad_norm": 1.7383732795715332,
+      "learning_rate": 5.519218549300806e-07,
+      "loss": 0.0176,
+      "step": 1609
+    },
+    {
+      "epoch": 7.853658536585366,
+      "grad_norm": 2.0462191104888916,
+      "learning_rate": 5.495229838965021e-07,
+      "loss": 0.031,
+      "step": 1610
+    },
+    {
+      "epoch": 7.8585365853658535,
+      "grad_norm": 1.3201459646224976,
+      "learning_rate": 5.471286935726866e-07,
+      "loss": 0.0062,
+      "step": 1611
+    },
+    {
+      "epoch": 7.863414634146341,
+      "grad_norm": 2.9285616874694824,
+      "learning_rate": 5.447389895816416e-07,
+      "loss": 0.0615,
+      "step": 1612
+    },
+    {
+      "epoch": 7.868292682926829,
+      "grad_norm": 3.1918647289276123,
+      "learning_rate": 5.423538775356049e-07,
+      "loss": 0.0377,
+      "step": 1613
+    },
+    {
+      "epoch": 7.873170731707317,
+      "grad_norm": 1.406246542930603,
+      "learning_rate": 5.399733630360287e-07,
+      "loss": 0.0122,
+      "step": 1614
+    },
+    {
+      "epoch": 7.878048780487805,
+      "grad_norm": 1.7651537656784058,
+      "learning_rate": 5.375974516735713e-07,
+      "loss": 0.015,
+      "step": 1615
+    },
+    {
+      "epoch": 7.882926829268293,
+      "grad_norm": 1.9614673852920532,
+      "learning_rate": 5.352261490280767e-07,
+      "loss": 0.0058,
+      "step": 1616
+    },
+    {
+      "epoch": 7.88780487804878,
+      "grad_norm": 1.6031639575958252,
+      "learning_rate": 5.328594606685661e-07,
+      "loss": 0.0041,
+      "step": 1617
+    },
+    {
+      "epoch": 7.892682926829268,
+      "grad_norm": 0.9787303805351257,
+      "learning_rate": 5.304973921532264e-07,
+      "loss": 0.0067,
+      "step": 1618
+    },
+    {
+      "epoch": 7.897560975609756,
+      "grad_norm": 1.2693779468536377,
+      "learning_rate": 5.281399490293923e-07,
+      "loss": 0.0064,
+      "step": 1619
+    },
+    {
+      "epoch": 7.902439024390244,
+      "grad_norm": 1.8421361446380615,
+      "learning_rate": 5.257871368335357e-07,
+      "loss": 0.0182,
+      "step": 1620
+    },
+    {
+      "epoch": 7.907317073170732,
+      "grad_norm": 0.9667096138000488,
+      "learning_rate": 5.234389610912552e-07,
+      "loss": 0.0024,
+      "step": 1621
+    },
+    {
+      "epoch": 7.912195121951219,
+      "grad_norm": 3.2266018390655518,
+      "learning_rate": 5.210954273172578e-07,
+      "loss": 0.02,
+      "step": 1622
+    },
+    {
+      "epoch": 7.917073170731707,
+      "grad_norm": 1.5821634531021118,
+      "learning_rate": 5.187565410153497e-07,
+      "loss": 0.024,
+      "step": 1623
+    },
+    {
+      "epoch": 7.921951219512195,
+      "grad_norm": 1.9864275455474854,
+      "learning_rate": 5.164223076784239e-07,
+      "loss": 0.0103,
+      "step": 1624
+    },
+    {
+      "epoch": 7.926829268292683,
+      "grad_norm": 1.866466999053955,
+      "learning_rate": 5.14092732788444e-07,
+      "loss": 0.0268,
+      "step": 1625
+    },
+    {
+      "epoch": 7.931707317073171,
+      "grad_norm": 1.165686011314392,
+      "learning_rate": 5.117678218164337e-07,
+      "loss": 0.0085,
+      "step": 1626
+    },
+    {
+      "epoch": 7.9365853658536585,
+      "grad_norm": 1.1883208751678467,
+      "learning_rate": 5.094475802224644e-07,
+      "loss": 0.006,
+      "step": 1627
+    },
+    {
+      "epoch": 7.941463414634146,
+      "grad_norm": 1.5121057033538818,
+      "learning_rate": 5.071320134556404e-07,
+      "loss": 0.003,
+      "step": 1628
+    },
+    {
+      "epoch": 7.946341463414634,
+      "grad_norm": 1.1923614740371704,
+      "learning_rate": 5.048211269540868e-07,
+      "loss": 0.0064,
+      "step": 1629
+    },
+    {
+      "epoch": 7.951219512195122,
+      "grad_norm": 1.33751380443573,
+      "learning_rate": 5.025149261449391e-07,
+      "loss": 0.0082,
+      "step": 1630
+    },
+    {
+      "epoch": 7.95609756097561,
+      "grad_norm": 1.9143925905227661,
+      "learning_rate": 5.002134164443262e-07,
+      "loss": 0.0202,
+      "step": 1631
+    },
+    {
+      "epoch": 7.9609756097560975,
+      "grad_norm": 1.2547078132629395,
+      "learning_rate": 4.979166032573607e-07,
+      "loss": 0.0033,
+      "step": 1632
+    },
+    {
+      "epoch": 7.965853658536585,
+      "grad_norm": 2.3050332069396973,
+      "learning_rate": 4.956244919781247e-07,
+      "loss": 0.052,
+      "step": 1633
+    },
+    {
+      "epoch": 7.970731707317073,
+      "grad_norm": 1.4462478160858154,
+      "learning_rate": 4.933370879896604e-07,
+      "loss": 0.0049,
+      "step": 1634
+    },
+    {
+      "epoch": 7.975609756097561,
+      "grad_norm": 1.519913911819458,
+      "learning_rate": 4.91054396663952e-07,
+      "loss": 0.0102,
+      "step": 1635
+    },
+    {
+      "epoch": 7.980487804878049,
+      "grad_norm": 2.9544193744659424,
+      "learning_rate": 4.887764233619163e-07,
+      "loss": 0.0112,
+      "step": 1636
+    },
+    {
+      "epoch": 7.985365853658537,
+      "grad_norm": 0.9778392314910889,
+      "learning_rate": 4.865031734333919e-07,
+      "loss": 0.0032,
+      "step": 1637
+    },
+    {
+      "epoch": 7.990243902439024,
+      "grad_norm": 2.783501386642456,
+      "learning_rate": 4.842346522171226e-07,
+      "loss": 0.012,
+      "step": 1638
+    },
+    {
+      "epoch": 7.995121951219512,
+      "grad_norm": 1.5644093751907349,
+      "learning_rate": 4.819708650407467e-07,
+      "loss": 0.0184,
+      "step": 1639
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.5741018056869507,
+      "learning_rate": 4.797118172207863e-07,
+      "loss": 0.0112,
+      "step": 1640
+    },
+    {
+      "epoch": 8.004878048780487,
+      "grad_norm": 0.9010241031646729,
+      "learning_rate": 4.774575140626317e-07,
+      "loss": 0.0064,
+      "step": 1641
+    },
+    {
+      "epoch": 8.009756097560976,
+      "grad_norm": 0.8204272985458374,
+      "learning_rate": 4.752079608605295e-07,
+      "loss": 0.003,
+      "step": 1642
+    },
+    {
+      "epoch": 8.014634146341463,
+      "grad_norm": 1.8131763935089111,
+      "learning_rate": 4.7296316289757366e-07,
+      "loss": 0.0063,
+      "step": 1643
+    },
+    {
+      "epoch": 8.019512195121951,
+      "grad_norm": 0.9918075799942017,
+      "learning_rate": 4.7072312544568844e-07,
+      "loss": 0.0039,
+      "step": 1644
+    },
+    {
+      "epoch": 8.024390243902438,
+      "grad_norm": 0.5097177028656006,
+      "learning_rate": 4.6848785376561733e-07,
+      "loss": 0.0028,
+      "step": 1645
+    },
+    {
+      "epoch": 8.029268292682927,
+      "grad_norm": 0.3497299253940582,
+      "learning_rate": 4.6625735310691396e-07,
+      "loss": 0.0021,
+      "step": 1646
+    },
+    {
+      "epoch": 8.034146341463414,
+      "grad_norm": 0.9271900057792664,
+      "learning_rate": 4.6403162870792524e-07,
+      "loss": 0.005,
+      "step": 1647
+    },
+    {
+      "epoch": 8.039024390243902,
+      "grad_norm": 0.951755940914154,
+      "learning_rate": 4.618106857957805e-07,
+      "loss": 0.0042,
+      "step": 1648
+    },
+    {
+      "epoch": 8.04390243902439,
+      "grad_norm": 0.6863508820533752,
+      "learning_rate": 4.5959452958638213e-07,
+      "loss": 0.0014,
+      "step": 1649
+    },
+    {
+      "epoch": 8.048780487804878,
+      "grad_norm": 0.45382270216941833,
+      "learning_rate": 4.573831652843888e-07,
+      "loss": 0.0012,
+      "step": 1650
+    },
+    {
+      "epoch": 8.053658536585365,
+      "grad_norm": 1.8319289684295654,
+      "learning_rate": 4.55176598083206e-07,
+      "loss": 0.0234,
+      "step": 1651
+    },
+    {
+      "epoch": 8.058536585365854,
+      "grad_norm": 1.2312507629394531,
+      "learning_rate": 4.5297483316497276e-07,
+      "loss": 0.0042,
+      "step": 1652
+    },
+    {
+      "epoch": 8.06341463414634,
+      "grad_norm": 1.4057971239089966,
+      "learning_rate": 4.5077787570055097e-07,
+      "loss": 0.0085,
+      "step": 1653
+    },
+    {
+      "epoch": 8.06829268292683,
+      "grad_norm": 3.3510940074920654,
+      "learning_rate": 4.4858573084951173e-07,
+      "loss": 0.0628,
+      "step": 1654
+    },
+    {
+      "epoch": 8.073170731707316,
+      "grad_norm": 0.6469231247901917,
+      "learning_rate": 4.463984037601224e-07,
+      "loss": 0.0026,
+      "step": 1655
+    },
+    {
+      "epoch": 8.078048780487805,
+      "grad_norm": 0.9491491317749023,
+      "learning_rate": 4.4421589956933827e-07,
+      "loss": 0.0021,
+      "step": 1656
+    },
+    {
+      "epoch": 8.082926829268292,
+      "grad_norm": 1.0847301483154297,
+      "learning_rate": 4.420382234027859e-07,
+      "loss": 0.0042,
+      "step": 1657
+    },
+    {
+      "epoch": 8.08780487804878,
+      "grad_norm": 0.5364987254142761,
+      "learning_rate": 4.398653803747532e-07,
+      "loss": 0.0045,
+      "step": 1658
+    },
+    {
+      "epoch": 8.092682926829267,
+      "grad_norm": 1.057804822921753,
+      "learning_rate": 4.3769737558817996e-07,
+      "loss": 0.0015,
+      "step": 1659
+    },
+    {
+      "epoch": 8.097560975609756,
+      "grad_norm": 1.2050957679748535,
+      "learning_rate": 4.355342141346405e-07,
+      "loss": 0.0124,
+      "step": 1660
+    },
+    {
+      "epoch": 8.102439024390243,
+      "grad_norm": 0.2821386754512787,
+      "learning_rate": 4.3337590109433505e-07,
+      "loss": 0.002,
+      "step": 1661
+    },
+    {
+      "epoch": 8.107317073170732,
+      "grad_norm": 0.7883970141410828,
+      "learning_rate": 4.3122244153607914e-07,
+      "loss": 0.0013,
+      "step": 1662
+    },
+    {
+      "epoch": 8.112195121951219,
+      "grad_norm": 1.1907166242599487,
+      "learning_rate": 4.2907384051728754e-07,
+      "loss": 0.0201,
+      "step": 1663
+    },
+    {
+      "epoch": 8.117073170731707,
+      "grad_norm": 1.3646314144134521,
+      "learning_rate": 4.2693010308396566e-07,
+      "loss": 0.0039,
+      "step": 1664
+    },
+    {
+      "epoch": 8.121951219512194,
+      "grad_norm": 2.0689423084259033,
+      "learning_rate": 4.247912342706975e-07,
+      "loss": 0.0035,
+      "step": 1665
+    },
+    {
+      "epoch": 8.126829268292683,
+      "grad_norm": 0.4086499810218811,
+      "learning_rate": 4.22657239100632e-07,
+      "loss": 0.0009,
+      "step": 1666
+    },
+    {
+      "epoch": 8.13170731707317,
+      "grad_norm": 0.9431869387626648,
+      "learning_rate": 4.2052812258547265e-07,
+      "loss": 0.0018,
+      "step": 1667
+    },
+    {
+      "epoch": 8.136585365853659,
+      "grad_norm": 0.9063575863838196,
+      "learning_rate": 4.184038897254655e-07,
+      "loss": 0.0021,
+      "step": 1668
+    },
+    {
+      "epoch": 8.141463414634146,
+      "grad_norm": 2.707298517227173,
+      "learning_rate": 4.1628454550938697e-07,
+      "loss": 0.019,
+      "step": 1669
+    },
+    {
+      "epoch": 8.146341463414634,
+      "grad_norm": 1.687988042831421,
+      "learning_rate": 4.141700949145322e-07,
+      "loss": 0.0144,
+      "step": 1670
+    },
+    {
+      "epoch": 8.151219512195121,
+      "grad_norm": 0.8905831575393677,
+      "learning_rate": 4.1206054290670537e-07,
+      "loss": 0.0088,
+      "step": 1671
+    },
+    {
+      "epoch": 8.15609756097561,
+      "grad_norm": 1.418512225151062,
+      "learning_rate": 4.0995589444020433e-07,
+      "loss": 0.0083,
+      "step": 1672
+    },
+    {
+      "epoch": 8.160975609756097,
+      "grad_norm": 1.1676236391067505,
+      "learning_rate": 4.0785615445781106e-07,
+      "loss": 0.0027,
+      "step": 1673
+    },
+    {
+      "epoch": 8.165853658536586,
+      "grad_norm": 1.5615407228469849,
+      "learning_rate": 4.057613278907818e-07,
+      "loss": 0.0089,
+      "step": 1674
+    },
+    {
+      "epoch": 8.170731707317072,
+      "grad_norm": 1.0604172945022583,
+      "learning_rate": 4.036714196588318e-07,
+      "loss": 0.0034,
+      "step": 1675
+    },
+    {
+      "epoch": 8.175609756097561,
+      "grad_norm": 1.3175733089447021,
+      "learning_rate": 4.015864346701251e-07,
+      "loss": 0.0021,
+      "step": 1676
+    },
+    {
+      "epoch": 8.180487804878048,
+      "grad_norm": 0.2539370059967041,
+      "learning_rate": 3.99506377821266e-07,
+      "loss": 0.0005,
+      "step": 1677
+    },
+    {
+      "epoch": 8.185365853658537,
+      "grad_norm": 0.8106228113174438,
+      "learning_rate": 3.97431253997283e-07,
+      "loss": 0.003,
+      "step": 1678
+    },
+    {
+      "epoch": 8.190243902439024,
+      "grad_norm": 0.6703351140022278,
+      "learning_rate": 3.9536106807161857e-07,
+      "loss": 0.0028,
+      "step": 1679
+    },
+    {
+      "epoch": 8.195121951219512,
+      "grad_norm": 1.2921632528305054,
+      "learning_rate": 3.932958249061214e-07,
+      "loss": 0.0097,
+      "step": 1680
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.7795253992080688,
+      "learning_rate": 3.9123552935102976e-07,
+      "loss": 0.004,
+      "step": 1681
+    },
+    {
+      "epoch": 8.204878048780488,
+      "grad_norm": 1.3402642011642456,
+      "learning_rate": 3.891801862449629e-07,
+      "loss": 0.0189,
+      "step": 1682
+    },
+    {
+      "epoch": 8.209756097560975,
+      "grad_norm": 0.6951391696929932,
+      "learning_rate": 3.8712980041490905e-07,
+      "loss": 0.0038,
+      "step": 1683
+    },
+    {
+      "epoch": 8.214634146341464,
+      "grad_norm": 0.8145114183425903,
+      "learning_rate": 3.850843766762155e-07,
+      "loss": 0.0038,
+      "step": 1684
+    },
+    {
+      "epoch": 8.21951219512195,
+      "grad_norm": 0.30702775716781616,
+      "learning_rate": 3.830439198325747e-07,
+      "loss": 0.0008,
+      "step": 1685
+    },
+    {
+      "epoch": 8.22439024390244,
+      "grad_norm": 0.45050138235092163,
+      "learning_rate": 3.81008434676014e-07,
+      "loss": 0.0013,
+      "step": 1686
+    },
+    {
+      "epoch": 8.229268292682926,
+      "grad_norm": 0.7875486612319946,
+      "learning_rate": 3.789779259868864e-07,
+      "loss": 0.0016,
+      "step": 1687
+    },
+    {
+      "epoch": 8.234146341463415,
+      "grad_norm": 0.9437265396118164,
+      "learning_rate": 3.769523985338566e-07,
+      "loss": 0.0045,
+      "step": 1688
+    },
+    {
+      "epoch": 8.239024390243902,
+      "grad_norm": 1.2928845882415771,
+      "learning_rate": 3.749318570738897e-07,
+      "loss": 0.0057,
+      "step": 1689
+    },
+    {
+      "epoch": 8.24390243902439,
+      "grad_norm": 0.9615103006362915,
+      "learning_rate": 3.7291630635224397e-07,
+      "loss": 0.0026,
+      "step": 1690
+    },
+    {
+      "epoch": 8.248780487804877,
+      "grad_norm": 0.8654932975769043,
+      "learning_rate": 3.709057511024541e-07,
+      "loss": 0.0056,
+      "step": 1691
+    },
+    {
+      "epoch": 8.253658536585366,
+      "grad_norm": 1.1101908683776855,
+      "learning_rate": 3.689001960463243e-07,
+      "loss": 0.0019,
+      "step": 1692
+    },
+    {
+      "epoch": 8.258536585365853,
+      "grad_norm": 0.9586653709411621,
+      "learning_rate": 3.668996458939156e-07,
+      "loss": 0.003,
+      "step": 1693
+    },
+    {
+      "epoch": 8.263414634146342,
+      "grad_norm": 1.1638360023498535,
+      "learning_rate": 3.649041053435354e-07,
+      "loss": 0.0031,
+      "step": 1694
+    },
+    {
+      "epoch": 8.268292682926829,
+      "grad_norm": 0.41364336013793945,
+      "learning_rate": 3.62913579081724e-07,
+      "loss": 0.0012,
+      "step": 1695
+    },
+    {
+      "epoch": 8.273170731707317,
+      "grad_norm": 1.1794198751449585,
+      "learning_rate": 3.609280717832489e-07,
+      "loss": 0.0067,
+      "step": 1696
+    },
+    {
+      "epoch": 8.278048780487804,
+      "grad_norm": 0.7281041741371155,
+      "learning_rate": 3.5894758811108795e-07,
+      "loss": 0.002,
+      "step": 1697
+    },
+    {
+      "epoch": 8.282926829268293,
+      "grad_norm": 0.42419376969337463,
+      "learning_rate": 3.5697213271642164e-07,
+      "loss": 0.0008,
+      "step": 1698
+    },
+    {
+      "epoch": 8.28780487804878,
+      "grad_norm": 0.6596572995185852,
+      "learning_rate": 3.5500171023862136e-07,
+      "loss": 0.0028,
+      "step": 1699
+    },
+    {
+      "epoch": 8.292682926829269,
+      "grad_norm": 1.236666202545166,
+      "learning_rate": 3.530363253052399e-07,
+      "loss": 0.0032,
+      "step": 1700
+    },
+    {
+      "epoch": 8.297560975609755,
+      "grad_norm": 0.977694571018219,
+      "learning_rate": 3.510759825319976e-07,
+      "loss": 0.0068,
+      "step": 1701
+    },
+    {
+      "epoch": 8.302439024390244,
+      "grad_norm": 1.0168365240097046,
+      "learning_rate": 3.491206865227739e-07,
+      "loss": 0.0017,
+      "step": 1702
+    },
+    {
+      "epoch": 8.307317073170731,
+      "grad_norm": 2.269639253616333,
+      "learning_rate": 3.4717044186959676e-07,
+      "loss": 0.0398,
+      "step": 1703
+    },
+    {
+      "epoch": 8.31219512195122,
+      "grad_norm": 1.0657192468643188,
+      "learning_rate": 3.452252531526301e-07,
+      "loss": 0.0049,
+      "step": 1704
+    },
+    {
+      "epoch": 8.317073170731707,
+      "grad_norm": 1.50715970993042,
+      "learning_rate": 3.432851249401628e-07,
+      "loss": 0.0164,
+      "step": 1705
+    },
+    {
+      "epoch": 8.321951219512195,
+      "grad_norm": 0.701214611530304,
+      "learning_rate": 3.413500617886023e-07,
+      "loss": 0.0038,
+      "step": 1706
+    },
+    {
+      "epoch": 8.326829268292682,
+      "grad_norm": 1.6810981035232544,
+      "learning_rate": 3.394200682424578e-07,
+      "loss": 0.0118,
+      "step": 1707
+    },
+    {
+      "epoch": 8.331707317073171,
+      "grad_norm": 1.4712997674942017,
+      "learning_rate": 3.374951488343328e-07,
+      "loss": 0.006,
+      "step": 1708
+    },
+    {
+      "epoch": 8.336585365853658,
+      "grad_norm": 0.6115317940711975,
+      "learning_rate": 3.355753080849164e-07,
+      "loss": 0.0011,
+      "step": 1709
+    },
+    {
+      "epoch": 8.341463414634147,
+      "grad_norm": 0.8171163201332092,
+      "learning_rate": 3.3366055050296776e-07,
+      "loss": 0.0024,
+      "step": 1710
+    },
+    {
+      "epoch": 8.346341463414634,
+      "grad_norm": 0.7722201943397522,
+      "learning_rate": 3.3175088058530925e-07,
+      "loss": 0.0028,
+      "step": 1711
+    },
+    {
+      "epoch": 8.351219512195122,
+      "grad_norm": 3.0709617137908936,
+      "learning_rate": 3.2984630281681556e-07,
+      "loss": 0.0109,
+      "step": 1712
+    },
+    {
+      "epoch": 8.35609756097561,
+      "grad_norm": 1.7634369134902954,
+      "learning_rate": 3.2794682167040125e-07,
+      "loss": 0.0031,
+      "step": 1713
+    },
+    {
+      "epoch": 8.360975609756098,
+      "grad_norm": 1.7657215595245361,
+      "learning_rate": 3.2605244160701155e-07,
+      "loss": 0.01,
+      "step": 1714
+    },
+    {
+      "epoch": 8.365853658536585,
+      "grad_norm": 1.432230830192566,
+      "learning_rate": 3.2416316707561316e-07,
+      "loss": 0.0042,
+      "step": 1715
+    },
+    {
+      "epoch": 8.370731707317074,
+      "grad_norm": 0.465900719165802,
+      "learning_rate": 3.2227900251318055e-07,
+      "loss": 0.0021,
+      "step": 1716
+    },
+    {
+      "epoch": 8.37560975609756,
+      "grad_norm": 1.3770387172698975,
+      "learning_rate": 3.2039995234468854e-07,
+      "loss": 0.0031,
+      "step": 1717
+    },
+    {
+      "epoch": 8.38048780487805,
+      "grad_norm": 0.4842236638069153,
+      "learning_rate": 3.1852602098309984e-07,
+      "loss": 0.0009,
+      "step": 1718
+    },
+    {
+      "epoch": 8.385365853658536,
+      "grad_norm": 0.6840565204620361,
+      "learning_rate": 3.1665721282935683e-07,
+      "loss": 0.0047,
+      "step": 1719
+    },
+    {
+      "epoch": 8.390243902439025,
+      "grad_norm": 0.5206313729286194,
+      "learning_rate": 3.147935322723694e-07,
+      "loss": 0.0026,
+      "step": 1720
+    },
+    {
+      "epoch": 8.395121951219512,
+      "grad_norm": 1.131412386894226,
+      "learning_rate": 3.1293498368900414e-07,
+      "loss": 0.0019,
+      "step": 1721
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.5872076153755188,
+      "learning_rate": 3.1108157144407765e-07,
+      "loss": 0.0009,
+      "step": 1722
+    },
+    {
+      "epoch": 8.404878048780487,
+      "grad_norm": 1.1455132961273193,
+      "learning_rate": 3.092332998903416e-07,
+      "loss": 0.0047,
+      "step": 1723
+    },
+    {
+      "epoch": 8.409756097560976,
+      "grad_norm": 1.4331532716751099,
+      "learning_rate": 3.073901733684748e-07,
+      "loss": 0.0162,
+      "step": 1724
+    },
+    {
+      "epoch": 8.414634146341463,
+      "grad_norm": 0.8186633586883545,
+      "learning_rate": 3.055521962070751e-07,
+      "loss": 0.0078,
+      "step": 1725
+    },
+    {
+      "epoch": 8.419512195121952,
+      "grad_norm": 0.9004407525062561,
+      "learning_rate": 3.0371937272264454e-07,
+      "loss": 0.0035,
+      "step": 1726
+    },
+    {
+      "epoch": 8.424390243902439,
+      "grad_norm": 0.8009728789329529,
+      "learning_rate": 3.0189170721958234e-07,
+      "loss": 0.0011,
+      "step": 1727
+    },
+    {
+      "epoch": 8.429268292682927,
+      "grad_norm": 0.7846589088439941,
+      "learning_rate": 3.000692039901756e-07,
+      "loss": 0.0042,
+      "step": 1728
+    },
+    {
+      "epoch": 8.434146341463414,
+      "grad_norm": 1.2301117181777954,
+      "learning_rate": 2.982518673145862e-07,
+      "loss": 0.0159,
+      "step": 1729
+    },
+    {
+      "epoch": 8.439024390243903,
+      "grad_norm": 0.8503583073616028,
+      "learning_rate": 2.9643970146084193e-07,
+      "loss": 0.0021,
+      "step": 1730
+    },
+    {
+      "epoch": 8.44390243902439,
+      "grad_norm": 1.661842942237854,
+      "learning_rate": 2.9463271068482955e-07,
+      "loss": 0.0124,
+      "step": 1731
+    },
+    {
+      "epoch": 8.448780487804878,
+      "grad_norm": 0.7799263000488281,
+      "learning_rate": 2.928308992302792e-07,
+      "loss": 0.0038,
+      "step": 1732
+    },
+    {
+      "epoch": 8.453658536585365,
+      "grad_norm": 0.6021434664726257,
+      "learning_rate": 2.9103427132875785e-07,
+      "loss": 0.0013,
+      "step": 1733
+    },
+    {
+      "epoch": 8.458536585365854,
+      "grad_norm": 1.430431604385376,
+      "learning_rate": 2.892428311996609e-07,
+      "loss": 0.0151,
+      "step": 1734
+    },
+    {
+      "epoch": 8.463414634146341,
+      "grad_norm": 1.1589592695236206,
+      "learning_rate": 2.8745658305019824e-07,
+      "loss": 0.0037,
+      "step": 1735
+    },
+    {
+      "epoch": 8.46829268292683,
+      "grad_norm": 0.7232568860054016,
+      "learning_rate": 2.856755310753867e-07,
+      "loss": 0.0046,
+      "step": 1736
+    },
+    {
+      "epoch": 8.473170731707317,
+      "grad_norm": 0.6265125274658203,
+      "learning_rate": 2.8389967945803984e-07,
+      "loss": 0.0014,
+      "step": 1737
+    },
+    {
+      "epoch": 8.478048780487805,
+      "grad_norm": 0.7115193009376526,
+      "learning_rate": 2.821290323687592e-07,
+      "loss": 0.0036,
+      "step": 1738
+    },
+    {
+      "epoch": 8.482926829268292,
+      "grad_norm": 0.5157519578933716,
+      "learning_rate": 2.803635939659222e-07,
+      "loss": 0.0016,
+      "step": 1739
+    },
+    {
+      "epoch": 8.487804878048781,
+      "grad_norm": 0.9217156767845154,
+      "learning_rate": 2.786033683956732e-07,
+      "loss": 0.0052,
+      "step": 1740
+    },
+    {
+      "epoch": 8.492682926829268,
+      "grad_norm": 4.063957691192627,
+      "learning_rate": 2.7684835979191664e-07,
+      "loss": 0.0999,
+      "step": 1741
+    },
+    {
+      "epoch": 8.497560975609757,
+      "grad_norm": 0.38870275020599365,
+      "learning_rate": 2.7509857227630223e-07,
+      "loss": 0.0009,
+      "step": 1742
+    },
+    {
+      "epoch": 8.502439024390243,
+      "grad_norm": 0.8282430768013,
+      "learning_rate": 2.733540099582188e-07,
+      "loss": 0.0026,
+      "step": 1743
+    },
+    {
+      "epoch": 8.507317073170732,
+      "grad_norm": 1.7269257307052612,
+      "learning_rate": 2.7161467693478493e-07,
+      "loss": 0.0094,
+      "step": 1744
+    },
+    {
+      "epoch": 8.512195121951219,
+      "grad_norm": 1.4464598894119263,
+      "learning_rate": 2.6988057729083613e-07,
+      "loss": 0.006,
+      "step": 1745
+    },
+    {
+      "epoch": 8.517073170731708,
+      "grad_norm": 0.9648481011390686,
+      "learning_rate": 2.681517150989185e-07,
+      "loss": 0.0043,
+      "step": 1746
+    },
+    {
+      "epoch": 8.521951219512195,
+      "grad_norm": 0.7762707471847534,
+      "learning_rate": 2.664280944192782e-07,
+      "loss": 0.0026,
+      "step": 1747
+    },
+    {
+      "epoch": 8.526829268292683,
+      "grad_norm": 0.9751222133636475,
+      "learning_rate": 2.64709719299851e-07,
+      "loss": 0.0044,
+      "step": 1748
+    },
+    {
+      "epoch": 8.53170731707317,
+      "grad_norm": 0.5906254053115845,
+      "learning_rate": 2.6299659377625296e-07,
+      "loss": 0.0008,
+      "step": 1749
+    },
+    {
+      "epoch": 8.536585365853659,
+      "grad_norm": 1.9417753219604492,
+      "learning_rate": 2.612887218717733e-07,
+      "loss": 0.0324,
+      "step": 1750
+    },
+    {
+      "epoch": 8.541463414634146,
+      "grad_norm": 0.6434907913208008,
+      "learning_rate": 2.5958610759736133e-07,
+      "loss": 0.0028,
+      "step": 1751
+    },
+    {
+      "epoch": 8.546341463414635,
+      "grad_norm": 0.8546578884124756,
+      "learning_rate": 2.5788875495161846e-07,
+      "loss": 0.0019,
+      "step": 1752
+    },
+    {
+      "epoch": 8.551219512195122,
+      "grad_norm": 0.8363909721374512,
+      "learning_rate": 2.561966679207917e-07,
+      "loss": 0.0028,
+      "step": 1753
+    },
+    {
+      "epoch": 8.55609756097561,
+      "grad_norm": 1.4901739358901978,
+      "learning_rate": 2.545098504787588e-07,
+      "loss": 0.0266,
+      "step": 1754
+    },
+    {
+      "epoch": 8.560975609756097,
+      "grad_norm": 0.6730532646179199,
+      "learning_rate": 2.5282830658702323e-07,
+      "loss": 0.0009,
+      "step": 1755
+    },
+    {
+      "epoch": 8.565853658536586,
+      "grad_norm": 0.7190845608711243,
+      "learning_rate": 2.511520401947032e-07,
+      "loss": 0.0056,
+      "step": 1756
+    },
+    {
+      "epoch": 8.570731707317073,
+      "grad_norm": 0.441381573677063,
+      "learning_rate": 2.494810552385232e-07,
+      "loss": 0.0009,
+      "step": 1757
+    },
+    {
+      "epoch": 8.575609756097561,
+      "grad_norm": 1.103507399559021,
+      "learning_rate": 2.47815355642804e-07,
+      "loss": 0.0023,
+      "step": 1758
+    },
+    {
+      "epoch": 8.580487804878048,
+      "grad_norm": 1.994994878768921,
+      "learning_rate": 2.461549453194523e-07,
+      "loss": 0.0454,
+      "step": 1759
+    },
+    {
+      "epoch": 8.585365853658537,
+      "grad_norm": 2.3645970821380615,
+      "learning_rate": 2.444998281679553e-07,
+      "loss": 0.0204,
+      "step": 1760
+    },
+    {
+      "epoch": 8.590243902439024,
+      "grad_norm": 1.7933200597763062,
+      "learning_rate": 2.428500080753676e-07,
+      "loss": 0.0387,
+      "step": 1761
+    },
+    {
+      "epoch": 8.595121951219513,
+      "grad_norm": 1.6070597171783447,
+      "learning_rate": 2.412054889163035e-07,
+      "loss": 0.0014,
+      "step": 1762
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.2842216193675995,
+      "learning_rate": 2.3956627455292924e-07,
+      "loss": 0.0011,
+      "step": 1763
+    },
+    {
+      "epoch": 8.604878048780488,
+      "grad_norm": 0.8213078379631042,
+      "learning_rate": 2.3793236883495164e-07,
+      "loss": 0.003,
+      "step": 1764
+    },
+    {
+      "epoch": 8.609756097560975,
+      "grad_norm": 0.9147091507911682,
+      "learning_rate": 2.363037755996095e-07,
+      "loss": 0.0032,
+      "step": 1765
+    },
+    {
+      "epoch": 8.614634146341464,
+      "grad_norm": 1.4246805906295776,
+      "learning_rate": 2.3468049867166747e-07,
+      "loss": 0.0037,
+      "step": 1766
+    },
+    {
+      "epoch": 8.61951219512195,
+      "grad_norm": 0.5553964376449585,
+      "learning_rate": 2.3306254186340305e-07,
+      "loss": 0.0014,
+      "step": 1767
+    },
+    {
+      "epoch": 8.62439024390244,
+      "grad_norm": 1.6941331624984741,
+      "learning_rate": 2.314499089745989e-07,
+      "loss": 0.0125,
+      "step": 1768
+    },
+    {
+      "epoch": 8.629268292682926,
+      "grad_norm": 2.965517520904541,
+      "learning_rate": 2.2984260379253693e-07,
+      "loss": 0.0855,
+      "step": 1769
+    },
+    {
+      "epoch": 8.634146341463415,
+      "grad_norm": 0.9295977354049683,
+      "learning_rate": 2.2824063009198428e-07,
+      "loss": 0.0031,
+      "step": 1770
+    },
+    {
+      "epoch": 8.639024390243902,
+      "grad_norm": 0.990189254283905,
+      "learning_rate": 2.2664399163518786e-07,
+      "loss": 0.0056,
+      "step": 1771
+    },
+    {
+      "epoch": 8.64390243902439,
+      "grad_norm": 1.7282871007919312,
+      "learning_rate": 2.25052692171866e-07,
+      "loss": 0.022,
+      "step": 1772
+    },
+    {
+      "epoch": 8.648780487804878,
+      "grad_norm": 1.2093932628631592,
+      "learning_rate": 2.2346673543919645e-07,
+      "loss": 0.0025,
+      "step": 1773
+    },
+    {
+      "epoch": 8.653658536585366,
+      "grad_norm": 0.9555385112762451,
+      "learning_rate": 2.2188612516181067e-07,
+      "loss": 0.0081,
+      "step": 1774
+    },
+    {
+      "epoch": 8.658536585365853,
+      "grad_norm": 0.7467104196548462,
+      "learning_rate": 2.203108650517835e-07,
+      "loss": 0.0015,
+      "step": 1775
+    },
+    {
+      "epoch": 8.663414634146342,
+      "grad_norm": 0.893450140953064,
+      "learning_rate": 2.1874095880862505e-07,
+      "loss": 0.0023,
+      "step": 1776
+    },
+    {
+      "epoch": 8.668292682926829,
+      "grad_norm": 1.0488923788070679,
+      "learning_rate": 2.171764101192722e-07,
+      "loss": 0.002,
+      "step": 1777
+    },
+    {
+      "epoch": 8.673170731707318,
+      "grad_norm": 1.1046003103256226,
+      "learning_rate": 2.1561722265807827e-07,
+      "loss": 0.002,
+      "step": 1778
+    },
+    {
+      "epoch": 8.678048780487805,
+      "grad_norm": 0.38860198855400085,
+      "learning_rate": 2.1406340008680748e-07,
+      "loss": 0.0015,
+      "step": 1779
+    },
+    {
+      "epoch": 8.682926829268293,
+      "grad_norm": 0.9970881938934326,
+      "learning_rate": 2.1251494605462358e-07,
+      "loss": 0.0028,
+      "step": 1780
+    },
+    {
+      "epoch": 8.68780487804878,
+      "grad_norm": 0.32808956503868103,
+      "learning_rate": 2.1097186419808151e-07,
+      "loss": 0.0008,
+      "step": 1781
+    },
+    {
+      "epoch": 8.692682926829269,
+      "grad_norm": 0.25458696484565735,
+      "learning_rate": 2.094341581411216e-07,
+      "loss": 0.0012,
+      "step": 1782
+    },
+    {
+      "epoch": 8.697560975609756,
+      "grad_norm": 0.3530316948890686,
+      "learning_rate": 2.0790183149505733e-07,
+      "loss": 0.0021,
+      "step": 1783
+    },
+    {
+      "epoch": 8.702439024390245,
+      "grad_norm": 0.6706930994987488,
+      "learning_rate": 2.063748878585689e-07,
+      "loss": 0.0028,
+      "step": 1784
+    },
+    {
+      "epoch": 8.707317073170731,
+      "grad_norm": 0.9568914175033569,
+      "learning_rate": 2.0485333081769588e-07,
+      "loss": 0.0018,
+      "step": 1785
+    },
+    {
+      "epoch": 8.71219512195122,
+      "grad_norm": 1.2713409662246704,
+      "learning_rate": 2.0333716394582536e-07,
+      "loss": 0.0142,
+      "step": 1786
+    },
+    {
+      "epoch": 8.717073170731707,
+      "grad_norm": 1.7427871227264404,
+      "learning_rate": 2.0182639080368634e-07,
+      "loss": 0.0135,
+      "step": 1787
+    },
+    {
+      "epoch": 8.721951219512196,
+      "grad_norm": 0.8939143419265747,
+      "learning_rate": 2.003210149393417e-07,
+      "loss": 0.0078,
+      "step": 1788
+    },
+    {
+      "epoch": 8.726829268292683,
+      "grad_norm": 1.1459598541259766,
+      "learning_rate": 1.9882103988817735e-07,
+      "loss": 0.0066,
+      "step": 1789
+    },
+    {
+      "epoch": 8.731707317073171,
+      "grad_norm": 0.875706672668457,
+      "learning_rate": 1.9732646917289545e-07,
+      "loss": 0.0051,
+      "step": 1790
+    },
+    {
+      "epoch": 8.736585365853658,
+      "grad_norm": 0.2884235084056854,
+      "learning_rate": 1.958373063035071e-07,
+      "loss": 0.001,
+      "step": 1791
+    },
+    {
+      "epoch": 8.741463414634147,
+      "grad_norm": 1.3679368495941162,
+      "learning_rate": 1.9435355477732205e-07,
+      "loss": 0.0057,
+      "step": 1792
+    },
+    {
+      "epoch": 8.746341463414634,
+      "grad_norm": 0.5913633108139038,
+      "learning_rate": 1.928752180789417e-07,
+      "loss": 0.0023,
+      "step": 1793
+    },
+    {
+      "epoch": 8.751219512195123,
+      "grad_norm": 1.565428376197815,
+      "learning_rate": 1.9140229968025058e-07,
+      "loss": 0.0191,
+      "step": 1794
+    },
+    {
+      "epoch": 8.75609756097561,
+      "grad_norm": 1.4710811376571655,
+      "learning_rate": 1.8993480304040912e-07,
+      "loss": 0.0114,
+      "step": 1795
+    },
+    {
+      "epoch": 8.760975609756098,
+      "grad_norm": 1.803842306137085,
+      "learning_rate": 1.8847273160584378e-07,
+      "loss": 0.0046,
+      "step": 1796
+    },
+    {
+      "epoch": 8.765853658536585,
+      "grad_norm": 0.694587230682373,
+      "learning_rate": 1.8701608881023957e-07,
+      "loss": 0.0014,
+      "step": 1797
+    },
+    {
+      "epoch": 8.770731707317074,
+      "grad_norm": 0.7563489675521851,
+      "learning_rate": 1.855648780745342e-07,
+      "loss": 0.0085,
+      "step": 1798
+    },
+    {
+      "epoch": 8.77560975609756,
+      "grad_norm": 1.1587045192718506,
+      "learning_rate": 1.8411910280690588e-07,
+      "loss": 0.0034,
+      "step": 1799
+    },
+    {
+      "epoch": 8.78048780487805,
+      "grad_norm": 1.7251181602478027,
+      "learning_rate": 1.826787664027685e-07,
+      "loss": 0.0119,
+      "step": 1800
+    },
+    {
+      "epoch": 8.785365853658536,
+      "grad_norm": 1.3170053958892822,
+      "learning_rate": 1.8124387224476347e-07,
+      "loss": 0.0059,
+      "step": 1801
+    },
+    {
+      "epoch": 8.790243902439025,
+      "grad_norm": 0.927018940448761,
+      "learning_rate": 1.7981442370274993e-07,
+      "loss": 0.0021,
+      "step": 1802
+    },
+    {
+      "epoch": 8.795121951219512,
+      "grad_norm": 2.3129045963287354,
+      "learning_rate": 1.783904241337983e-07,
+      "loss": 0.0085,
+      "step": 1803
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 1.1010651588439941,
+      "learning_rate": 1.7697187688218291e-07,
+      "loss": 0.0037,
+      "step": 1804
+    },
+    {
+      "epoch": 8.804878048780488,
+      "grad_norm": 0.3990725576877594,
+      "learning_rate": 1.7555878527937164e-07,
+      "loss": 0.0008,
+      "step": 1805
+    },
+    {
+      "epoch": 8.809756097560976,
+      "grad_norm": 1.022905707359314,
+      "learning_rate": 1.7415115264402065e-07,
+      "loss": 0.0092,
+      "step": 1806
+    },
+    {
+      "epoch": 8.814634146341463,
+      "grad_norm": 0.7391730546951294,
+      "learning_rate": 1.727489822819664e-07,
+      "loss": 0.0016,
+      "step": 1807
+    },
+    {
+      "epoch": 8.819512195121952,
+      "grad_norm": 0.5859627723693848,
+      "learning_rate": 1.7135227748621585e-07,
+      "loss": 0.0012,
+      "step": 1808
+    },
+    {
+      "epoch": 8.824390243902439,
+      "grad_norm": 1.5222235918045044,
+      "learning_rate": 1.699610415369407e-07,
+      "loss": 0.0126,
+      "step": 1809
+    },
+    {
+      "epoch": 8.829268292682928,
+      "grad_norm": 0.8635048270225525,
+      "learning_rate": 1.6857527770146876e-07,
+      "loss": 0.0086,
+      "step": 1810
+    },
+    {
+      "epoch": 8.834146341463414,
+      "grad_norm": 0.8385710120201111,
+      "learning_rate": 1.6719498923427697e-07,
+      "loss": 0.0031,
+      "step": 1811
+    },
+    {
+      "epoch": 8.839024390243903,
+      "grad_norm": 1.0619077682495117,
+      "learning_rate": 1.6582017937698287e-07,
+      "loss": 0.0083,
+      "step": 1812
+    },
+    {
+      "epoch": 8.84390243902439,
+      "grad_norm": 0.6677606701850891,
+      "learning_rate": 1.6445085135833732e-07,
+      "loss": 0.002,
+      "step": 1813
+    },
+    {
+      "epoch": 8.848780487804879,
+      "grad_norm": 0.703705370426178,
+      "learning_rate": 1.6308700839421793e-07,
+      "loss": 0.0027,
+      "step": 1814
+    },
+    {
+      "epoch": 8.853658536585366,
+      "grad_norm": 0.7628077864646912,
+      "learning_rate": 1.6172865368762004e-07,
+      "loss": 0.0028,
+      "step": 1815
+    },
+    {
+      "epoch": 8.858536585365854,
+      "grad_norm": 0.7577258348464966,
+      "learning_rate": 1.6037579042864876e-07,
+      "loss": 0.0011,
+      "step": 1816
+    },
+    {
+      "epoch": 8.863414634146341,
+      "grad_norm": 1.2882269620895386,
+      "learning_rate": 1.5902842179451482e-07,
+      "loss": 0.0082,
+      "step": 1817
+    },
+    {
+      "epoch": 8.86829268292683,
+      "grad_norm": 1.030044436454773,
+      "learning_rate": 1.576865509495229e-07,
+      "loss": 0.0068,
+      "step": 1818
+    },
+    {
+      "epoch": 8.873170731707317,
+      "grad_norm": 1.9678841829299927,
+      "learning_rate": 1.5635018104506627e-07,
+      "loss": 0.0085,
+      "step": 1819
+    },
+    {
+      "epoch": 8.878048780487806,
+      "grad_norm": 0.756213366985321,
+      "learning_rate": 1.5501931521962055e-07,
+      "loss": 0.0062,
+      "step": 1820
+    },
+    {
+      "epoch": 8.882926829268293,
+      "grad_norm": 1.1753418445587158,
+      "learning_rate": 1.5369395659873305e-07,
+      "loss": 0.0043,
+      "step": 1821
+    },
+    {
+      "epoch": 8.887804878048781,
+      "grad_norm": 0.8144367933273315,
+      "learning_rate": 1.5237410829501864e-07,
+      "loss": 0.0042,
+      "step": 1822
+    },
+    {
+      "epoch": 8.892682926829268,
+      "grad_norm": 1.0879873037338257,
+      "learning_rate": 1.510597734081512e-07,
+      "loss": 0.0077,
+      "step": 1823
+    },
+    {
+      "epoch": 8.897560975609757,
+      "grad_norm": 1.7992119789123535,
+      "learning_rate": 1.497509550248555e-07,
+      "loss": 0.0013,
+      "step": 1824
+    },
+    {
+      "epoch": 8.902439024390244,
+      "grad_norm": 1.0460071563720703,
+      "learning_rate": 1.4844765621890135e-07,
+      "loss": 0.0091,
+      "step": 1825
+    },
+    {
+      "epoch": 8.907317073170733,
+      "grad_norm": 1.5372941493988037,
+      "learning_rate": 1.471498800510962e-07,
+      "loss": 0.005,
+      "step": 1826
+    },
+    {
+      "epoch": 8.91219512195122,
+      "grad_norm": 0.3672512173652649,
+      "learning_rate": 1.4585762956927624e-07,
+      "loss": 0.0014,
+      "step": 1827
+    },
+    {
+      "epoch": 8.917073170731708,
+      "grad_norm": 1.0456454753875732,
+      "learning_rate": 1.4457090780830185e-07,
+      "loss": 0.0063,
+      "step": 1828
+    },
+    {
+      "epoch": 8.921951219512195,
+      "grad_norm": 0.9190329909324646,
+      "learning_rate": 1.432897177900483e-07,
+      "loss": 0.0065,
+      "step": 1829
+    },
+    {
+      "epoch": 8.926829268292684,
+      "grad_norm": 1.8261685371398926,
+      "learning_rate": 1.4201406252340038e-07,
+      "loss": 0.0099,
+      "step": 1830
+    },
+    {
+      "epoch": 8.93170731707317,
+      "grad_norm": 1.1341190338134766,
+      "learning_rate": 1.407439450042433e-07,
+      "loss": 0.0042,
+      "step": 1831
+    },
+    {
+      "epoch": 8.93658536585366,
+      "grad_norm": 11.465933799743652,
+      "learning_rate": 1.3947936821545772e-07,
+      "loss": 0.004,
+      "step": 1832
+    },
+    {
+      "epoch": 8.941463414634146,
+      "grad_norm": 0.5747786164283752,
+      "learning_rate": 1.3822033512691209e-07,
+      "loss": 0.0009,
+      "step": 1833
+    },
+    {
+      "epoch": 8.946341463414633,
+      "grad_norm": 1.1908810138702393,
+      "learning_rate": 1.369668486954545e-07,
+      "loss": 0.0028,
+      "step": 1834
+    },
+    {
+      "epoch": 8.951219512195122,
+      "grad_norm": 0.2560107111930847,
+      "learning_rate": 1.3571891186490687e-07,
+      "loss": 0.001,
+      "step": 1835
+    },
+    {
+      "epoch": 8.95609756097561,
+      "grad_norm": 0.5070216059684753,
+      "learning_rate": 1.3447652756605894e-07,
+      "loss": 0.0024,
+      "step": 1836
+    },
+    {
+      "epoch": 8.960975609756098,
+      "grad_norm": 0.507199227809906,
+      "learning_rate": 1.3323969871665897e-07,
+      "loss": 0.0015,
+      "step": 1837
+    },
+    {
+      "epoch": 8.965853658536584,
+      "grad_norm": 0.29779553413391113,
+      "learning_rate": 1.3200842822140818e-07,
+      "loss": 0.0007,
+      "step": 1838
+    },
+    {
+      "epoch": 8.970731707317073,
+      "grad_norm": 0.4603523015975952,
+      "learning_rate": 1.3078271897195572e-07,
+      "loss": 0.0018,
+      "step": 1839
+    },
+    {
+      "epoch": 8.975609756097562,
+      "grad_norm": 1.0771223306655884,
+      "learning_rate": 1.2956257384688807e-07,
+      "loss": 0.0063,
+      "step": 1840
+    },
+    {
+      "epoch": 8.980487804878049,
+      "grad_norm": 0.798372745513916,
+      "learning_rate": 1.283479957117248e-07,
+      "loss": 0.002,
+      "step": 1841
+    },
+    {
+      "epoch": 8.985365853658536,
+      "grad_norm": 2.3283369541168213,
+      "learning_rate": 1.2713898741891244e-07,
+      "loss": 0.0398,
+      "step": 1842
+    },
+    {
+      "epoch": 8.990243902439024,
+      "grad_norm": 0.18683794140815735,
+      "learning_rate": 1.2593555180781591e-07,
+      "loss": 0.0004,
+      "step": 1843
+    },
+    {
+      "epoch": 8.995121951219513,
+      "grad_norm": 2.2289419174194336,
+      "learning_rate": 1.2473769170471188e-07,
+      "loss": 0.0713,
+      "step": 1844
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.9360214471817017,
+      "learning_rate": 1.2354540992278452e-07,
+      "loss": 0.002,
+      "step": 1845
+    },
+    {
+      "epoch": 9.004878048780487,
+      "grad_norm": 0.11728485673666,
+      "learning_rate": 1.223587092621162e-07,
+      "loss": 0.0004,
+      "step": 1846
+    },
+    {
+      "epoch": 9.009756097560976,
+      "grad_norm": 2.8439087867736816,
+      "learning_rate": 1.2117759250968225e-07,
+      "loss": 0.0791,
+      "step": 1847
+    },
+    {
+      "epoch": 9.014634146341463,
+      "grad_norm": 0.3048456311225891,
+      "learning_rate": 1.2000206243934358e-07,
+      "loss": 0.0021,
+      "step": 1848
+    },
+    {
+      "epoch": 9.019512195121951,
+      "grad_norm": 0.35457128286361694,
+      "learning_rate": 1.1883212181184212e-07,
+      "loss": 0.0014,
+      "step": 1849
+    },
+    {
+      "epoch": 9.024390243902438,
+      "grad_norm": 0.4256647527217865,
+      "learning_rate": 1.176677733747919e-07,
+      "loss": 0.003,
+      "step": 1850
+    },
+    {
+      "epoch": 9.029268292682927,
+      "grad_norm": 0.14073246717453003,
+      "learning_rate": 1.1650901986267365e-07,
+      "loss": 0.0009,
+      "step": 1851
+    },
+    {
+      "epoch": 9.034146341463414,
+      "grad_norm": 0.2287226915359497,
+      "learning_rate": 1.1535586399682885e-07,
+      "loss": 0.001,
+      "step": 1852
+    },
+    {
+      "epoch": 9.039024390243902,
+      "grad_norm": 0.1520719975233078,
+      "learning_rate": 1.1420830848545256e-07,
+      "loss": 0.0008,
+      "step": 1853
+    },
+    {
+      "epoch": 9.04390243902439,
+      "grad_norm": 0.7066623568534851,
+      "learning_rate": 1.1306635602358673e-07,
+      "loss": 0.0086,
+      "step": 1854
+    },
+    {
+      "epoch": 9.048780487804878,
+      "grad_norm": 0.5992008447647095,
+      "learning_rate": 1.1193000929311638e-07,
+      "loss": 0.0023,
+      "step": 1855
+    },
+    {
+      "epoch": 9.053658536585365,
+      "grad_norm": 1.6487441062927246,
+      "learning_rate": 1.1079927096275978e-07,
+      "loss": 0.0235,
+      "step": 1856
+    },
+    {
+      "epoch": 9.058536585365854,
+      "grad_norm": 0.2044752836227417,
+      "learning_rate": 1.0967414368806384e-07,
+      "loss": 0.0013,
+      "step": 1857
+    },
+    {
+      "epoch": 9.06341463414634,
+      "grad_norm": 0.13774175941944122,
+      "learning_rate": 1.0855463011139905e-07,
+      "loss": 0.0005,
+      "step": 1858
+    },
+    {
+      "epoch": 9.06829268292683,
+      "grad_norm": 0.1757974475622177,
+      "learning_rate": 1.0744073286195089e-07,
+      "loss": 0.0011,
+      "step": 1859
+    },
+    {
+      "epoch": 9.073170731707316,
+      "grad_norm": 0.3755623698234558,
+      "learning_rate": 1.0633245455571511e-07,
+      "loss": 0.0015,
+      "step": 1860
+    },
+    {
+      "epoch": 9.078048780487805,
+      "grad_norm": 0.1744445413351059,
+      "learning_rate": 1.052297977954922e-07,
+      "loss": 0.001,
+      "step": 1861
+    },
+    {
+      "epoch": 9.082926829268292,
+      "grad_norm": 0.23634043335914612,
+      "learning_rate": 1.0413276517087956e-07,
+      "loss": 0.0005,
+      "step": 1862
+    },
+    {
+      "epoch": 9.08780487804878,
+      "grad_norm": 0.31559380888938904,
+      "learning_rate": 1.0304135925826603e-07,
+      "loss": 0.002,
+      "step": 1863
+    },
+    {
+      "epoch": 9.092682926829267,
+      "grad_norm": 0.44957175850868225,
+      "learning_rate": 1.0195558262082683e-07,
+      "loss": 0.0052,
+      "step": 1864
+    },
+    {
+      "epoch": 9.097560975609756,
+      "grad_norm": 0.23585057258605957,
+      "learning_rate": 1.0087543780851666e-07,
+      "loss": 0.0009,
+      "step": 1865
+    },
+    {
+      "epoch": 9.102439024390243,
+      "grad_norm": 0.26482903957366943,
+      "learning_rate": 9.98009273580633e-08,
+      "loss": 0.0017,
+      "step": 1866
+    },
+    {
+      "epoch": 9.107317073170732,
+      "grad_norm": 0.21670401096343994,
+      "learning_rate": 9.87320537929623e-08,
+      "loss": 0.0006,
+      "step": 1867
+    },
+    {
+      "epoch": 9.112195121951219,
+      "grad_norm": 0.11546074599027634,
+      "learning_rate": 9.766881962347208e-08,
+      "loss": 0.0006,
+      "step": 1868
+    },
+    {
+      "epoch": 9.117073170731707,
+      "grad_norm": 0.35039573907852173,
+      "learning_rate": 9.661122734660521e-08,
+      "loss": 0.0017,
+      "step": 1869
+    },
+    {
+      "epoch": 9.121951219512194,
+      "grad_norm": 0.25725650787353516,
+      "learning_rate": 9.555927944612492e-08,
+      "loss": 0.0015,
+      "step": 1870
+    },
+    {
+      "epoch": 9.126829268292683,
+      "grad_norm": 1.0865508317947388,
+      "learning_rate": 9.451297839253915e-08,
+      "loss": 0.0179,
+      "step": 1871
+    },
+    {
+      "epoch": 9.13170731707317,
+      "grad_norm": 0.29501980543136597,
+      "learning_rate": 9.34723266430937e-08,
+      "loss": 0.0013,
+      "step": 1872
+    },
+    {
+      "epoch": 9.136585365853659,
+      "grad_norm": 0.3127771019935608,
+      "learning_rate": 9.243732664176636e-08,
+      "loss": 0.0013,
+      "step": 1873
+    },
+    {
+      "epoch": 9.141463414634146,
+      "grad_norm": 0.47584185004234314,
+      "learning_rate": 9.140798081926277e-08,
+      "loss": 0.0042,
+      "step": 1874
+    },
+    {
+      "epoch": 9.146341463414634,
+      "grad_norm": 0.22509703040122986,
+      "learning_rate": 9.03842915930095e-08,
+      "loss": 0.0014,
+      "step": 1875
+    },
+    {
+      "epoch": 9.151219512195121,
+      "grad_norm": 0.2254130244255066,
+      "learning_rate": 8.936626136714754e-08,
+      "loss": 0.0014,
+      "step": 1876
+    },
+    {
+      "epoch": 9.15609756097561,
+      "grad_norm": 0.4184035360813141,
+      "learning_rate": 8.835389253252918e-08,
+      "loss": 0.0036,
+      "step": 1877
+    },
+    {
+      "epoch": 9.160975609756097,
+      "grad_norm": 0.8849661946296692,
+      "learning_rate": 8.734718746670978e-08,
+      "loss": 0.0195,
+      "step": 1878
+    },
+    {
+      "epoch": 9.165853658536586,
+      "grad_norm": 0.3465995192527771,
+      "learning_rate": 8.634614853394341e-08,
+      "loss": 0.0009,
+      "step": 1879
+    },
+    {
+      "epoch": 9.170731707317072,
+      "grad_norm": 0.5498316884040833,
+      "learning_rate": 8.53507780851781e-08,
+      "loss": 0.0028,
+      "step": 1880
+    },
+    {
+      "epoch": 9.175609756097561,
+      "grad_norm": 0.4553240239620209,
+      "learning_rate": 8.436107845804842e-08,
+      "loss": 0.0023,
+      "step": 1881
+    },
+    {
+      "epoch": 9.180487804878048,
+      "grad_norm": 0.3339614272117615,
+      "learning_rate": 8.33770519768709e-08,
+      "loss": 0.0008,
+      "step": 1882
+    },
+    {
+      "epoch": 9.185365853658537,
+      "grad_norm": 0.3212447762489319,
+      "learning_rate": 8.239870095263974e-08,
+      "loss": 0.0018,
+      "step": 1883
+    },
+    {
+      "epoch": 9.190243902439024,
+      "grad_norm": 0.2665475904941559,
+      "learning_rate": 8.142602768301921e-08,
+      "loss": 0.0016,
+      "step": 1884
+    },
+    {
+      "epoch": 9.195121951219512,
+      "grad_norm": 0.8188057541847229,
+      "learning_rate": 8.045903445233982e-08,
+      "loss": 0.0063,
+      "step": 1885
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 0.458200603723526,
+      "learning_rate": 7.949772353159191e-08,
+      "loss": 0.0049,
+      "step": 1886
+    },
+    {
+      "epoch": 9.204878048780488,
+      "grad_norm": 0.50230872631073,
+      "learning_rate": 7.854209717842231e-08,
+      "loss": 0.0006,
+      "step": 1887
+    },
+    {
+      "epoch": 9.209756097560975,
+      "grad_norm": 0.12954330444335938,
+      "learning_rate": 7.759215763712579e-08,
+      "loss": 0.0007,
+      "step": 1888
+    },
+    {
+      "epoch": 9.214634146341464,
+      "grad_norm": 0.3889886438846588,
+      "learning_rate": 7.664790713864223e-08,
+      "loss": 0.0038,
+      "step": 1889
+    },
+    {
+      "epoch": 9.21951219512195,
+      "grad_norm": 0.8406491875648499,
+      "learning_rate": 7.57093479005519e-08,
+      "loss": 0.0059,
+      "step": 1890
+    },
+    {
+      "epoch": 9.22439024390244,
+      "grad_norm": 0.27930590510368347,
+      "learning_rate": 7.477648212706746e-08,
+      "loss": 0.0009,
+      "step": 1891
+    },
+    {
+      "epoch": 9.229268292682926,
+      "grad_norm": 0.2927345037460327,
+      "learning_rate": 7.384931200903084e-08,
+      "loss": 0.0019,
+      "step": 1892
+    },
+    {
+      "epoch": 9.234146341463415,
+      "grad_norm": 0.5030691027641296,
+      "learning_rate": 7.29278397239086e-08,
+      "loss": 0.0036,
+      "step": 1893
+    },
+    {
+      "epoch": 9.239024390243902,
+      "grad_norm": 0.14574876427650452,
+      "learning_rate": 7.20120674357852e-08,
+      "loss": 0.0005,
+      "step": 1894
+    },
+    {
+      "epoch": 9.24390243902439,
+      "grad_norm": 0.286927729845047,
+      "learning_rate": 7.110199729535805e-08,
+      "loss": 0.0009,
+      "step": 1895
+    },
+    {
+      "epoch": 9.248780487804877,
+      "grad_norm": 0.44844964146614075,
+      "learning_rate": 7.019763143993441e-08,
+      "loss": 0.0047,
+      "step": 1896
+    },
+    {
+      "epoch": 9.253658536585366,
+      "grad_norm": 0.16901901364326477,
+      "learning_rate": 6.929897199342395e-08,
+      "loss": 0.0006,
+      "step": 1897
+    },
+    {
+      "epoch": 9.258536585365853,
+      "grad_norm": 0.19660663604736328,
+      "learning_rate": 6.840602106633425e-08,
+      "loss": 0.0005,
+      "step": 1898
+    },
+    {
+      "epoch": 9.263414634146342,
+      "grad_norm": 0.2517840564250946,
+      "learning_rate": 6.751878075576867e-08,
+      "loss": 0.001,
+      "step": 1899
+    },
+    {
+      "epoch": 9.268292682926829,
+      "grad_norm": 0.6886439323425293,
+      "learning_rate": 6.663725314541652e-08,
+      "loss": 0.0046,
+      "step": 1900
+    },
+    {
+      "epoch": 9.273170731707317,
+      "grad_norm": 0.2044619917869568,
+      "learning_rate": 6.576144030555259e-08,
+      "loss": 0.0009,
+      "step": 1901
+    },
+    {
+      "epoch": 9.278048780487804,
+      "grad_norm": 0.5199993848800659,
+      "learning_rate": 6.489134429302906e-08,
+      "loss": 0.0038,
+      "step": 1902
+    },
+    {
+      "epoch": 9.282926829268293,
+      "grad_norm": 0.20676910877227783,
+      "learning_rate": 6.402696715127387e-08,
+      "loss": 0.0007,
+      "step": 1903
+    },
+    {
+      "epoch": 9.28780487804878,
+      "grad_norm": 0.13005101680755615,
+      "learning_rate": 6.316831091028237e-08,
+      "loss": 0.0005,
+      "step": 1904
+    },
+    {
+      "epoch": 9.292682926829269,
+      "grad_norm": 0.12870948016643524,
+      "learning_rate": 6.23153775866156e-08,
+      "loss": 0.0004,
+      "step": 1905
+    },
+    {
+      "epoch": 9.297560975609755,
+      "grad_norm": 0.4530372619628906,
+      "learning_rate": 6.14681691833935e-08,
+      "loss": 0.0027,
+      "step": 1906
+    },
+    {
+      "epoch": 9.302439024390244,
+      "grad_norm": 0.14936240017414093,
+      "learning_rate": 6.062668769029168e-08,
+      "loss": 0.0006,
+      "step": 1907
+    },
+    {
+      "epoch": 9.307317073170731,
+      "grad_norm": 1.0447592735290527,
+      "learning_rate": 5.979093508353489e-08,
+      "loss": 0.0033,
+      "step": 1908
+    },
+    {
+      "epoch": 9.31219512195122,
+      "grad_norm": 0.9839334487915039,
+      "learning_rate": 5.896091332589532e-08,
+      "loss": 0.0143,
+      "step": 1909
+    },
+    {
+      "epoch": 9.317073170731707,
+      "grad_norm": 0.13809092342853546,
+      "learning_rate": 5.813662436668477e-08,
+      "loss": 0.0006,
+      "step": 1910
+    },
+    {
+      "epoch": 9.321951219512195,
+      "grad_norm": 0.2679869830608368,
+      "learning_rate": 5.731807014175195e-08,
+      "loss": 0.0009,
+      "step": 1911
+    },
+    {
+      "epoch": 9.326829268292682,
+      "grad_norm": 0.09745966643095016,
+      "learning_rate": 5.650525257347744e-08,
+      "loss": 0.0004,
+      "step": 1912
+    },
+    {
+      "epoch": 9.331707317073171,
+      "grad_norm": 0.15892420709133148,
+      "learning_rate": 5.569817357076984e-08,
+      "loss": 0.0012,
+      "step": 1913
+    },
+    {
+      "epoch": 9.336585365853658,
+      "grad_norm": 0.9430788159370422,
+      "learning_rate": 5.489683502905935e-08,
+      "loss": 0.0039,
+      "step": 1914
+    },
+    {
+      "epoch": 9.341463414634147,
+      "grad_norm": 0.15283145010471344,
+      "learning_rate": 5.410123883029639e-08,
+      "loss": 0.001,
+      "step": 1915
+    },
+    {
+      "epoch": 9.346341463414634,
+      "grad_norm": 0.3713572025299072,
+      "learning_rate": 5.3311386842944125e-08,
+      "loss": 0.0017,
+      "step": 1916
+    },
+    {
+      "epoch": 9.351219512195122,
+      "grad_norm": 0.29313772916793823,
+      "learning_rate": 5.25272809219754e-08,
+      "loss": 0.0021,
+      "step": 1917
+    },
+    {
+      "epoch": 9.35609756097561,
+      "grad_norm": 0.219829261302948,
+      "learning_rate": 5.17489229088694e-08,
+      "loss": 0.001,
+      "step": 1918
+    },
+    {
+      "epoch": 9.360975609756098,
+      "grad_norm": 0.35704150795936584,
+      "learning_rate": 5.097631463160585e-08,
+      "loss": 0.002,
+      "step": 1919
+    },
+    {
+      "epoch": 9.365853658536585,
+      "grad_norm": 0.44924139976501465,
+      "learning_rate": 5.020945790466025e-08,
+      "loss": 0.0007,
+      "step": 1920
+    },
+    {
+      "epoch": 9.370731707317074,
+      "grad_norm": 0.10656553506851196,
+      "learning_rate": 4.944835452900199e-08,
+      "loss": 0.0005,
+      "step": 1921
+    },
+    {
+      "epoch": 9.37560975609756,
+      "grad_norm": 0.8902695178985596,
+      "learning_rate": 4.869300629208762e-08,
+      "loss": 0.0084,
+      "step": 1922
+    },
+    {
+      "epoch": 9.38048780487805,
+      "grad_norm": 0.3544962406158447,
+      "learning_rate": 4.7943414967858426e-08,
+      "loss": 0.0024,
+      "step": 1923
+    },
+    {
+      "epoch": 9.385365853658536,
+      "grad_norm": 0.3950733244419098,
+      "learning_rate": 4.7199582316734827e-08,
+      "loss": 0.0006,
+      "step": 1924
+    },
+    {
+      "epoch": 9.390243902439025,
+      "grad_norm": 0.45972177386283875,
+      "learning_rate": 4.6461510085613616e-08,
+      "loss": 0.0012,
+      "step": 1925
+    },
+    {
+      "epoch": 9.395121951219512,
+      "grad_norm": 0.12690195441246033,
+      "learning_rate": 4.5729200007862686e-08,
+      "loss": 0.0004,
+      "step": 1926
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 0.2692466974258423,
+      "learning_rate": 4.5002653803317975e-08,
+      "loss": 0.0016,
+      "step": 1927
+    },
+    {
+      "epoch": 9.404878048780487,
+      "grad_norm": 0.5844394564628601,
+      "learning_rate": 4.428187317827848e-08,
+      "loss": 0.001,
+      "step": 1928
+    },
+    {
+      "epoch": 9.409756097560976,
+      "grad_norm": 0.5482091307640076,
+      "learning_rate": 4.356685982550263e-08,
+      "loss": 0.0016,
+      "step": 1929
+    },
+    {
+      "epoch": 9.414634146341463,
+      "grad_norm": 0.06951025128364563,
+      "learning_rate": 4.285761542420497e-08,
+      "loss": 0.0004,
+      "step": 1930
+    },
+    {
+      "epoch": 9.419512195121952,
+      "grad_norm": 0.1519978791475296,
+      "learning_rate": 4.215414164005116e-08,
+      "loss": 0.0007,
+      "step": 1931
+    },
+    {
+      "epoch": 9.424390243902439,
+      "grad_norm": 0.3855389654636383,
+      "learning_rate": 4.145644012515465e-08,
+      "loss": 0.0012,
+      "step": 1932
+    },
+    {
+      "epoch": 9.429268292682927,
+      "grad_norm": 0.28962311148643494,
+      "learning_rate": 4.076451251807223e-08,
+      "loss": 0.0024,
+      "step": 1933
+    },
+    {
+      "epoch": 9.434146341463414,
+      "grad_norm": 0.12305665761232376,
+      "learning_rate": 4.0078360443801535e-08,
+      "loss": 0.0005,
+      "step": 1934
+    },
+    {
+      "epoch": 9.439024390243903,
+      "grad_norm": 0.5113069415092468,
+      "learning_rate": 3.9397985513775495e-08,
+      "loss": 0.0007,
+      "step": 1935
+    },
+    {
+      "epoch": 9.44390243902439,
+      "grad_norm": 0.11020799726247787,
+      "learning_rate": 3.872338932585984e-08,
+      "loss": 0.0006,
+      "step": 1936
+    },
+    {
+      "epoch": 9.448780487804878,
+      "grad_norm": 0.24607239663600922,
+      "learning_rate": 3.8054573464348655e-08,
+      "loss": 0.0012,
+      "step": 1937
+    },
+    {
+      "epoch": 9.453658536585365,
+      "grad_norm": 0.09522794187068939,
+      "learning_rate": 3.739153949996105e-08,
+      "loss": 0.0004,
+      "step": 1938
+    },
+    {
+      "epoch": 9.458536585365854,
+      "grad_norm": 0.3217187523841858,
+      "learning_rate": 3.6734288989836994e-08,
+      "loss": 0.0018,
+      "step": 1939
+    },
+    {
+      "epoch": 9.463414634146341,
+      "grad_norm": 0.10770946741104126,
+      "learning_rate": 3.608282347753428e-08,
+      "loss": 0.0005,
+      "step": 1940
+    },
+    {
+      "epoch": 9.46829268292683,
+      "grad_norm": 0.18529640138149261,
+      "learning_rate": 3.543714449302488e-08,
+      "loss": 0.0014,
+      "step": 1941
+    },
+    {
+      "epoch": 9.473170731707317,
+      "grad_norm": 0.3584231436252594,
+      "learning_rate": 3.479725355268998e-08,
+      "loss": 0.0011,
+      "step": 1942
+    },
+    {
+      "epoch": 9.478048780487805,
+      "grad_norm": 0.3854292035102844,
+      "learning_rate": 3.4163152159318866e-08,
+      "loss": 0.0019,
+      "step": 1943
+    },
+    {
+      "epoch": 9.482926829268292,
+      "grad_norm": 0.08858831971883774,
+      "learning_rate": 3.353484180210337e-08,
+      "loss": 0.0003,
+      "step": 1944
+    },
+    {
+      "epoch": 9.487804878048781,
+      "grad_norm": 0.5076143741607666,
+      "learning_rate": 3.291232395663424e-08,
+      "loss": 0.0057,
+      "step": 1945
+    },
+    {
+      "epoch": 9.492682926829268,
+      "grad_norm": 0.38053473830223083,
+      "learning_rate": 3.229560008490007e-08,
+      "loss": 0.0021,
+      "step": 1946
+    },
+    {
+      "epoch": 9.497560975609757,
+      "grad_norm": 1.4997718334197998,
+      "learning_rate": 3.168467163528116e-08,
+      "loss": 0.0079,
+      "step": 1947
+    },
+    {
+      "epoch": 9.502439024390243,
+      "grad_norm": 0.7466314435005188,
+      "learning_rate": 3.1079540042547315e-08,
+      "loss": 0.0077,
+      "step": 1948
+    },
+    {
+      "epoch": 9.507317073170732,
+      "grad_norm": 0.11087851971387863,
+      "learning_rate": 3.0480206727855066e-08,
+      "loss": 0.0004,
+      "step": 1949
+    },
+    {
+      "epoch": 9.512195121951219,
+      "grad_norm": 0.2965907156467438,
+      "learning_rate": 2.988667309874294e-08,
+      "loss": 0.0015,
+      "step": 1950
+    },
+    {
+      "epoch": 9.517073170731708,
+      "grad_norm": 1.4327231645584106,
+      "learning_rate": 2.9298940549128962e-08,
+      "loss": 0.0132,
+      "step": 1951
+    },
+    {
+      "epoch": 9.521951219512195,
+      "grad_norm": 0.9336621165275574,
+      "learning_rate": 2.871701045930708e-08,
+      "loss": 0.0019,
+      "step": 1952
+    },
+    {
+      "epoch": 9.526829268292683,
+      "grad_norm": 1.2587624788284302,
+      "learning_rate": 2.8140884195945184e-08,
+      "loss": 0.0024,
+      "step": 1953
+    },
+    {
+      "epoch": 9.53170731707317,
+      "grad_norm": 0.13109427690505981,
+      "learning_rate": 2.7570563112079564e-08,
+      "loss": 0.001,
+      "step": 1954
+    },
+    {
+      "epoch": 9.536585365853659,
+      "grad_norm": 0.2514895796775818,
+      "learning_rate": 2.700604854711353e-08,
+      "loss": 0.0013,
+      "step": 1955
+    },
+    {
+      "epoch": 9.541463414634146,
+      "grad_norm": 0.3432636857032776,
+      "learning_rate": 2.6447341826814077e-08,
+      "loss": 0.0005,
+      "step": 1956
+    },
+    {
+      "epoch": 9.546341463414635,
+      "grad_norm": 0.4550987780094147,
+      "learning_rate": 2.5894444263307728e-08,
+      "loss": 0.001,
+      "step": 1957
+    },
+    {
+      "epoch": 9.551219512195122,
+      "grad_norm": 1.2675397396087646,
+      "learning_rate": 2.5347357155078577e-08,
+      "loss": 0.0103,
+      "step": 1958
+    },
+    {
+      "epoch": 9.55609756097561,
+      "grad_norm": 0.1289552003145218,
+      "learning_rate": 2.4806081786964974e-08,
+      "loss": 0.0006,
+      "step": 1959
+    },
+    {
+      "epoch": 9.560975609756097,
+      "grad_norm": 0.6298596858978271,
+      "learning_rate": 2.4270619430156183e-08,
+      "loss": 0.0019,
+      "step": 1960
+    },
+    {
+      "epoch": 9.565853658536586,
+      "grad_norm": 0.538487434387207,
+      "learning_rate": 2.3740971342189056e-08,
+      "loss": 0.0027,
+      "step": 1961
+    },
+    {
+      "epoch": 9.570731707317073,
+      "grad_norm": 0.8478948473930359,
+      "learning_rate": 2.321713876694637e-08,
+      "loss": 0.0133,
+      "step": 1962
+    },
+    {
+      "epoch": 9.575609756097561,
+      "grad_norm": 1.0609294176101685,
+      "learning_rate": 2.269912293465293e-08,
+      "loss": 0.008,
+      "step": 1963
+    },
+    {
+      "epoch": 9.580487804878048,
+      "grad_norm": 0.634739100933075,
+      "learning_rate": 2.2186925061872532e-08,
+      "loss": 0.0025,
+      "step": 1964
+    },
+    {
+      "epoch": 9.585365853658537,
+      "grad_norm": 0.43630343675613403,
+      "learning_rate": 2.1680546351506016e-08,
+      "loss": 0.003,
+      "step": 1965
+    },
+    {
+      "epoch": 9.590243902439024,
+      "grad_norm": 0.3712899684906006,
+      "learning_rate": 2.117998799278709e-08,
+      "loss": 0.0036,
+      "step": 1966
+    },
+    {
+      "epoch": 9.595121951219513,
+      "grad_norm": 0.13679739832878113,
+      "learning_rate": 2.068525116128095e-08,
+      "loss": 0.0006,
+      "step": 1967
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 1.8157588243484497,
+      "learning_rate": 2.0196337018880962e-08,
+      "loss": 0.0659,
+      "step": 1968
+    },
+    {
+      "epoch": 9.604878048780488,
+      "grad_norm": 0.07176486402750015,
+      "learning_rate": 1.9713246713805588e-08,
+      "loss": 0.0003,
+      "step": 1969
+    },
+    {
+      "epoch": 9.609756097560975,
+      "grad_norm": 0.33367958664894104,
+      "learning_rate": 1.9235981380595625e-08,
+      "loss": 0.0013,
+      "step": 1970
+    },
+    {
+      "epoch": 9.614634146341464,
+      "grad_norm": 0.08895006775856018,
+      "learning_rate": 1.876454214011253e-08,
+      "loss": 0.0004,
+      "step": 1971
+    },
+    {
+      "epoch": 9.61951219512195,
+      "grad_norm": 0.2062547653913498,
+      "learning_rate": 1.8298930099534817e-08,
+      "loss": 0.001,
+      "step": 1972
+    },
+    {
+      "epoch": 9.62439024390244,
+      "grad_norm": 0.1351477950811386,
+      "learning_rate": 1.783914635235584e-08,
+      "loss": 0.0006,
+      "step": 1973
+    },
+    {
+      "epoch": 9.629268292682926,
+      "grad_norm": 0.5446783304214478,
+      "learning_rate": 1.738519197838101e-08,
+      "loss": 0.0038,
+      "step": 1974
+    },
+    {
+      "epoch": 9.634146341463415,
+      "grad_norm": 0.12655134499073029,
+      "learning_rate": 1.6937068043725856e-08,
+      "loss": 0.0006,
+      "step": 1975
+    },
+    {
+      "epoch": 9.639024390243902,
+      "grad_norm": 0.7479956150054932,
+      "learning_rate": 1.6494775600812418e-08,
+      "loss": 0.0026,
+      "step": 1976
+    },
+    {
+      "epoch": 9.64390243902439,
+      "grad_norm": 0.39983221888542175,
+      "learning_rate": 1.6058315688367852e-08,
+      "loss": 0.003,
+      "step": 1977
+    },
+    {
+      "epoch": 9.648780487804878,
+      "grad_norm": 0.2727876305580139,
+      "learning_rate": 1.5627689331421946e-08,
+      "loss": 0.0015,
+      "step": 1978
+    },
+    {
+      "epoch": 9.653658536585366,
+      "grad_norm": 0.17525868117809296,
+      "learning_rate": 1.520289754130322e-08,
+      "loss": 0.001,
+      "step": 1979
+    },
+    {
+      "epoch": 9.658536585365853,
+      "grad_norm": 0.2446790337562561,
+      "learning_rate": 1.478394131563865e-08,
+      "loss": 0.0011,
+      "step": 1980
+    },
+    {
+      "epoch": 9.663414634146342,
+      "grad_norm": 0.37458178400993347,
+      "learning_rate": 1.4370821638350353e-08,
+      "loss": 0.0022,
+      "step": 1981
+    },
+    {
+      "epoch": 9.668292682926829,
+      "grad_norm": 0.1664375215768814,
+      "learning_rate": 1.396353947965251e-08,
+      "loss": 0.0006,
+      "step": 1982
+    },
+    {
+      "epoch": 9.673170731707318,
+      "grad_norm": 0.08668441325426102,
+      "learning_rate": 1.3562095796050279e-08,
+      "loss": 0.0003,
+      "step": 1983
+    },
+    {
+      "epoch": 9.678048780487805,
+      "grad_norm": 0.2897089719772339,
+      "learning_rate": 1.3166491530337555e-08,
+      "loss": 0.001,
+      "step": 1984
+    },
+    {
+      "epoch": 9.682926829268293,
+      "grad_norm": 0.21582652628421783,
+      "learning_rate": 1.2776727611593653e-08,
+      "loss": 0.0007,
+      "step": 1985
+    },
+    {
+      "epoch": 9.68780487804878,
+      "grad_norm": 0.3643123507499695,
+      "learning_rate": 1.2392804955181915e-08,
+      "loss": 0.002,
+      "step": 1986
+    },
+    {
+      "epoch": 9.692682926829269,
+      "grad_norm": 0.5870813131332397,
+      "learning_rate": 1.2014724462747763e-08,
+      "loss": 0.0016,
+      "step": 1987
+    },
+    {
+      "epoch": 9.697560975609756,
+      "grad_norm": 0.19344697892665863,
+      "learning_rate": 1.1642487022215931e-08,
+      "loss": 0.0008,
+      "step": 1988
+    },
+    {
+      "epoch": 9.702439024390245,
+      "grad_norm": 0.15417703986167908,
+      "learning_rate": 1.1276093507788798e-08,
+      "loss": 0.001,
+      "step": 1989
+    },
+    {
+      "epoch": 9.707317073170731,
+      "grad_norm": 0.2714616358280182,
+      "learning_rate": 1.0915544779944164e-08,
+      "loss": 0.0022,
+      "step": 1990
+    },
+    {
+      "epoch": 9.71219512195122,
+      "grad_norm": 0.14375440776348114,
+      "learning_rate": 1.0560841685433864e-08,
+      "loss": 0.0008,
+      "step": 1991
+    },
+    {
+      "epoch": 9.717073170731707,
+      "grad_norm": 0.19977939128875732,
+      "learning_rate": 1.021198505728016e-08,
+      "loss": 0.0011,
+      "step": 1992
+    },
+    {
+      "epoch": 9.721951219512196,
+      "grad_norm": 0.20787867903709412,
+      "learning_rate": 9.868975714775741e-09,
+      "loss": 0.0012,
+      "step": 1993
+    },
+    {
+      "epoch": 9.726829268292683,
+      "grad_norm": 0.3878643810749054,
+      "learning_rate": 9.531814463480394e-09,
+      "loss": 0.0008,
+      "step": 1994
+    },
+    {
+      "epoch": 9.731707317073171,
+      "grad_norm": 0.5140596032142639,
+      "learning_rate": 9.200502095220166e-09,
+      "loss": 0.0034,
+      "step": 1995
+    },
+    {
+      "epoch": 9.736585365853658,
+      "grad_norm": 0.2106190174818039,
+      "learning_rate": 8.875039388084317e-09,
+      "loss": 0.0008,
+      "step": 1996
+    },
+    {
+      "epoch": 9.741463414634147,
+      "grad_norm": 0.09516038745641708,
+      "learning_rate": 8.555427106424485e-09,
+      "loss": 0.0005,
+      "step": 1997
+    },
+    {
+      "epoch": 9.746341463414634,
+      "grad_norm": 1.439642310142517,
+      "learning_rate": 8.241666000852466e-09,
+      "loss": 0.0314,
+      "step": 1998
+    },
+    {
+      "epoch": 9.751219512195123,
+      "grad_norm": 0.10020413249731064,
+      "learning_rate": 7.933756808238823e-09,
+      "loss": 0.0004,
+      "step": 1999
+    },
+    {
+      "epoch": 9.75609756097561,
+      "grad_norm": 0.4296906888484955,
+      "learning_rate": 7.631700251710116e-09,
+      "loss": 0.0022,
+      "step": 2000
+    },
+    {
+      "epoch": 9.760975609756098,
+      "grad_norm": 0.4867343604564667,
+      "learning_rate": 7.335497040648898e-09,
+      "loss": 0.0024,
+      "step": 2001
+    },
+    {
+      "epoch": 9.765853658536585,
+      "grad_norm": 3.0838112831115723,
+      "learning_rate": 7.045147870690105e-09,
+      "loss": 0.0796,
+      "step": 2002
+    },
+    {
+      "epoch": 9.770731707317074,
+      "grad_norm": 0.26949402689933777,
+      "learning_rate": 6.760653423721619e-09,
+      "loss": 0.0012,
+      "step": 2003
+    },
+    {
+      "epoch": 9.77560975609756,
+      "grad_norm": 0.854682445526123,
+      "learning_rate": 6.4820143678800964e-09,
+      "loss": 0.0059,
+      "step": 2004
+    },
+    {
+      "epoch": 9.78048780487805,
+      "grad_norm": 0.06472957879304886,
+      "learning_rate": 6.209231357551526e-09,
+      "loss": 0.0003,
+      "step": 2005
+    },
+    {
+      "epoch": 9.785365853658536,
+      "grad_norm": 0.9941632747650146,
+      "learning_rate": 5.942305033369289e-09,
+      "loss": 0.0113,
+      "step": 2006
+    },
+    {
+      "epoch": 9.790243902439025,
+      "grad_norm": 0.08150490373373032,
+      "learning_rate": 5.681236022211378e-09,
+      "loss": 0.0003,
+      "step": 2007
+    },
+    {
+      "epoch": 9.795121951219512,
+      "grad_norm": 0.37303054332733154,
+      "learning_rate": 5.426024937200402e-09,
+      "loss": 0.0021,
+      "step": 2008
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 0.12861268222332,
+      "learning_rate": 5.176672377701364e-09,
+      "loss": 0.0004,
+      "step": 2009
+    },
+    {
+      "epoch": 9.804878048780488,
+      "grad_norm": 0.13954521715641022,
+      "learning_rate": 4.933178929321103e-09,
+      "loss": 0.0006,
+      "step": 2010
+    },
+    {
+      "epoch": 9.809756097560976,
+      "grad_norm": 0.8102789521217346,
+      "learning_rate": 4.695545163905524e-09,
+      "loss": 0.0047,
+      "step": 2011
+    },
+    {
+      "epoch": 9.814634146341463,
+      "grad_norm": 0.8437443971633911,
+      "learning_rate": 4.463771639539038e-09,
+      "loss": 0.0013,
+      "step": 2012
+    },
+    {
+      "epoch": 9.819512195121952,
+      "grad_norm": 0.3098134994506836,
+      "learning_rate": 4.237858900543734e-09,
+      "loss": 0.0025,
+      "step": 2013
+    },
+    {
+      "epoch": 9.824390243902439,
+      "grad_norm": 0.7686973214149475,
+      "learning_rate": 4.017807477477154e-09,
+      "loss": 0.0045,
+      "step": 2014
+    },
+    {
+      "epoch": 9.829268292682928,
+      "grad_norm": 0.45219677686691284,
+      "learning_rate": 3.803617887132016e-09,
+      "loss": 0.0017,
+      "step": 2015
+    },
+    {
+      "epoch": 9.834146341463414,
+      "grad_norm": 0.529446542263031,
+      "learning_rate": 3.5952906325339988e-09,
+      "loss": 0.0043,
+      "step": 2016
+    },
+    {
+      "epoch": 9.839024390243903,
+      "grad_norm": 0.35920700430870056,
+      "learning_rate": 3.3928262029411794e-09,
+      "loss": 0.0025,
+      "step": 2017
+    },
+    {
+      "epoch": 9.84390243902439,
+      "grad_norm": 0.3075787127017975,
+      "learning_rate": 3.196225073842929e-09,
+      "loss": 0.0025,
+      "step": 2018
+    },
+    {
+      "epoch": 9.848780487804879,
+      "grad_norm": 0.1374140977859497,
+      "learning_rate": 3.005487706958243e-09,
+      "loss": 0.0005,
+      "step": 2019
+    },
+    {
+      "epoch": 9.853658536585366,
+      "grad_norm": 0.5697541236877441,
+      "learning_rate": 2.8206145502354678e-09,
+      "loss": 0.0026,
+      "step": 2020
+    },
+    {
+      "epoch": 9.858536585365854,
+      "grad_norm": 1.0206952095031738,
+      "learning_rate": 2.641606037850353e-09,
+      "loss": 0.0101,
+      "step": 2021
+    },
+    {
+      "epoch": 9.863414634146341,
+      "grad_norm": 0.29209089279174805,
+      "learning_rate": 2.468462590205778e-09,
+      "loss": 0.0022,
+      "step": 2022
+    },
+    {
+      "epoch": 9.86829268292683,
+      "grad_norm": 0.13821417093276978,
+      "learning_rate": 2.3011846139306404e-09,
+      "loss": 0.0006,
+      "step": 2023
+    },
+    {
+      "epoch": 9.873170731707317,
+      "grad_norm": 0.4531463086605072,
+      "learning_rate": 2.13977250187819e-09,
+      "loss": 0.0018,
+      "step": 2024
+    },
+    {
+      "epoch": 9.878048780487806,
+      "grad_norm": 0.4380701184272766,
+      "learning_rate": 1.9842266331260296e-09,
+      "loss": 0.0031,
+      "step": 2025
+    },
+    {
+      "epoch": 9.882926829268293,
+      "grad_norm": 0.33851730823516846,
+      "learning_rate": 1.834547372975004e-09,
+      "loss": 0.0013,
+      "step": 2026
+    },
+    {
+      "epoch": 9.887804878048781,
+      "grad_norm": 0.4231720566749573,
+      "learning_rate": 1.6907350729478133e-09,
+      "loss": 0.0024,
+      "step": 2027
+    },
+    {
+      "epoch": 9.892682926829268,
+      "grad_norm": 0.4602144658565521,
+      "learning_rate": 1.5527900707887344e-09,
+      "loss": 0.004,
+      "step": 2028
+    },
+    {
+      "epoch": 9.897560975609757,
+      "grad_norm": 0.9638814330101013,
+      "learning_rate": 1.4207126904625114e-09,
+      "loss": 0.0097,
+      "step": 2029
+    },
+    {
+      "epoch": 9.902439024390244,
+      "grad_norm": 0.1374921053647995,
+      "learning_rate": 1.2945032421540771e-09,
+      "loss": 0.0005,
+      "step": 2030
+    },
+    {
+      "epoch": 9.907317073170733,
+      "grad_norm": 0.45432549715042114,
+      "learning_rate": 1.1741620222671667e-09,
+      "loss": 0.0041,
+      "step": 2031
+    },
+    {
+      "epoch": 9.91219512195122,
+      "grad_norm": 0.1905360370874405,
+      "learning_rate": 1.0596893134240394e-09,
+      "loss": 0.0007,
+      "step": 2032
+    },
+    {
+      "epoch": 9.917073170731708,
+      "grad_norm": 0.41532090306282043,
+      "learning_rate": 9.51085384464645e-10,
+      "loss": 0.007,
+      "step": 2033
+    },
+    {
+      "epoch": 9.921951219512195,
+      "grad_norm": 0.5167479515075684,
+      "learning_rate": 8.48350490446348e-10,
+      "loss": 0.0018,
+      "step": 2034
+    },
+    {
+      "epoch": 9.926829268292684,
+      "grad_norm": 0.4736052453517914,
+      "learning_rate": 7.514848726422608e-10,
+      "loss": 0.0028,
+      "step": 2035
+    },
+    {
+      "epoch": 9.93170731707317,
+      "grad_norm": 0.19710786640644073,
+      "learning_rate": 6.604887585426323e-10,
+      "loss": 0.0009,
+      "step": 2036
+    },
+    {
+      "epoch": 9.93658536585366,
+      "grad_norm": 0.4432890713214874,
+      "learning_rate": 5.753623618520721e-10,
+      "loss": 0.0049,
+      "step": 2037
+    },
+    {
+      "epoch": 9.941463414634146,
+      "grad_norm": 1.073500394821167,
+      "learning_rate": 4.961058824909382e-10,
+      "loss": 0.0044,
+      "step": 2038
+    },
+    {
+      "epoch": 9.946341463414633,
+      "grad_norm": 0.17539291083812714,
+      "learning_rate": 4.2271950659311665e-10,
+      "loss": 0.0005,
+      "step": 2039
+    },
+    {
+      "epoch": 9.951219512195122,
+      "grad_norm": 0.3674123287200928,
+      "learning_rate": 3.5520340650768705e-10,
+      "loss": 0.0012,
+      "step": 2040
+    },
+    {
+      "epoch": 9.95609756097561,
+      "grad_norm": 0.20704206824302673,
+      "learning_rate": 2.9355774079614653e-10,
+      "loss": 0.0007,
+      "step": 2041
+    },
+    {
+      "epoch": 9.960975609756098,
+      "grad_norm": 0.07566344738006592,
+      "learning_rate": 2.377826542343531e-10,
+      "loss": 0.0003,
+      "step": 2042
+    },
+    {
+      "epoch": 9.965853658536584,
+      "grad_norm": 0.1342095583677292,
+      "learning_rate": 1.8787827781002743e-10,
+      "loss": 0.0005,
+      "step": 2043
+    },
+    {
+      "epoch": 9.970731707317073,
+      "grad_norm": 0.7139898538589478,
+      "learning_rate": 1.4384472872414067e-10,
+      "loss": 0.0013,
+      "step": 2044
+    },
+    {
+      "epoch": 9.975609756097562,
+      "grad_norm": 0.11133516579866409,
+      "learning_rate": 1.056821103900818e-10,
+      "loss": 0.0006,
+      "step": 2045
+    },
+    {
+      "epoch": 9.980487804878049,
+      "grad_norm": 0.16971242427825928,
+      "learning_rate": 7.339051243254735e-11,
+      "loss": 0.0011,
+      "step": 2046
+    },
+    {
+      "epoch": 9.985365853658536,
+      "grad_norm": 1.3622301816940308,
+      "learning_rate": 4.697001068892926e-11,
+      "loss": 0.0202,
+      "step": 2047
+    },
+    {
+      "epoch": 9.990243902439024,
+      "grad_norm": 0.10895299166440964,
+      "learning_rate": 2.642066720792702e-11,
+      "loss": 0.0005,
+      "step": 2048
+    },
+    {
+      "epoch": 9.995121951219513,
+      "grad_norm": 0.32567188143730164,
+      "learning_rate": 1.1742530249547745e-11,
+      "loss": 0.0017,
+      "step": 2049
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.06682642549276352,
+      "learning_rate": 2.9356342859387933e-12,
+      "loss": 0.0002,
+      "step": 2050
+    },
+    {
+      "epoch": 10.0,
+      "step": 2050,
+      "total_flos": 5.892331269877924e+17,
+      "train_loss": 0.2632100873960966,
+      "train_runtime": 9760.8256,
+      "train_samples_per_second": 0.837,
+      "train_steps_per_second": 0.21
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2050,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 206,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.892331269877924e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}