diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7983 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1131,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002652519893899204,
+      "grad_norm": 968.0919232146924,
+      "learning_rate": 8.771929824561404e-08,
+      "loss": 12.4211,
+      "step": 1
+    },
+    {
+      "epoch": 0.005305039787798408,
+      "grad_norm": 1026.2449443021458,
+      "learning_rate": 1.7543859649122808e-07,
+      "loss": 12.452,
+      "step": 2
+    },
+    {
+      "epoch": 0.007957559681697613,
+      "grad_norm": 1017.251056225472,
+      "learning_rate": 2.6315789473684213e-07,
+      "loss": 12.264,
+      "step": 3
+    },
+    {
+      "epoch": 0.010610079575596816,
+      "grad_norm": 1065.85515059101,
+      "learning_rate": 3.5087719298245616e-07,
+      "loss": 12.3229,
+      "step": 4
+    },
+    {
+      "epoch": 0.013262599469496022,
+      "grad_norm": 952.5383153380988,
+      "learning_rate": 4.385964912280702e-07,
+      "loss": 12.3374,
+      "step": 5
+    },
+    {
+      "epoch": 0.015915119363395226,
+      "grad_norm": 1013.6977210971099,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 12.316,
+      "step": 6
+    },
+    {
+      "epoch": 0.01856763925729443,
+      "grad_norm": 1060.4701992924765,
+      "learning_rate": 6.140350877192982e-07,
+      "loss": 12.1177,
+      "step": 7
+    },
+    {
+      "epoch": 0.021220159151193633,
+      "grad_norm": 874.2817791313958,
+      "learning_rate": 7.017543859649123e-07,
+      "loss": 12.059,
+      "step": 8
+    },
+    {
+      "epoch": 0.023872679045092837,
+      "grad_norm": 1034.117575834617,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 11.6855,
+      "step": 9
+    },
+    {
+      "epoch": 0.026525198938992044,
+      "grad_norm": 958.1913728158022,
+      "learning_rate": 8.771929824561404e-07,
+      "loss": 11.4163,
+      "step": 10
+    },
+    {
+      "epoch": 0.029177718832891247,
+      "grad_norm": 1220.2117771599096,
+      "learning_rate": 9.649122807017545e-07,
+      "loss": 10.6914,
+      "step": 11
+    },
+    {
+      "epoch": 0.03183023872679045,
+      "grad_norm": 1048.1213049240255,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 10.1105,
+      "step": 12
+    },
+    {
+      "epoch": 0.034482758620689655,
+      "grad_norm": 789.9306134938697,
+      "learning_rate": 1.1403508771929824e-06,
+      "loss": 10.0797,
+      "step": 13
+    },
+    {
+      "epoch": 0.03713527851458886,
+      "grad_norm": 949.2047562477611,
+      "learning_rate": 1.2280701754385965e-06,
+      "loss": 9.7583,
+      "step": 14
+    },
+    {
+      "epoch": 0.03978779840848806,
+      "grad_norm": 754.369190005223,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 8.4975,
+      "step": 15
+    },
+    {
+      "epoch": 0.042440318302387266,
+      "grad_norm": 824.0571350755927,
+      "learning_rate": 1.4035087719298246e-06,
+      "loss": 8.4441,
+      "step": 16
+    },
+    {
+      "epoch": 0.04509283819628647,
+      "grad_norm": 1037.9890686137453,
+      "learning_rate": 1.4912280701754387e-06,
+      "loss": 7.9298,
+      "step": 17
+    },
+    {
+      "epoch": 0.04774535809018567,
+      "grad_norm": 765.6957994118179,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 7.7697,
+      "step": 18
+    },
+    {
+      "epoch": 0.050397877984084884,
+      "grad_norm": 718.3426984808572,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 7.3831,
+      "step": 19
+    },
+    {
+      "epoch": 0.05305039787798409,
+      "grad_norm": 553.427473727637,
+      "learning_rate": 1.7543859649122807e-06,
+      "loss": 6.9709,
+      "step": 20
+    },
+    {
+      "epoch": 0.05570291777188329,
+      "grad_norm": 483.40199996798475,
+      "learning_rate": 1.8421052631578948e-06,
+      "loss": 5.811,
+      "step": 21
+    },
+    {
+      "epoch": 0.058355437665782495,
+      "grad_norm": 707.095094504861,
+      "learning_rate": 1.929824561403509e-06,
+      "loss": 5.5033,
+      "step": 22
+    },
+    {
+      "epoch": 0.0610079575596817,
+      "grad_norm": 553.1469775776469,
+      "learning_rate": 2.017543859649123e-06,
+      "loss": 5.1927,
+      "step": 23
+    },
+    {
+      "epoch": 0.0636604774535809,
+      "grad_norm": 419.1819226196746,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 4.874,
+      "step": 24
+    },
+    {
+      "epoch": 0.06631299734748011,
+      "grad_norm": 608.8428174720044,
+      "learning_rate": 2.192982456140351e-06,
+      "loss": 4.7575,
+      "step": 25
+    },
+    {
+      "epoch": 0.06896551724137931,
+      "grad_norm": 333.80789191200347,
+      "learning_rate": 2.280701754385965e-06,
+      "loss": 4.5398,
+      "step": 26
+    },
+    {
+      "epoch": 0.07161803713527852,
+      "grad_norm": 368.1717826168138,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 4.4244,
+      "step": 27
+    },
+    {
+      "epoch": 0.07427055702917772,
+      "grad_norm": 393.3469206961459,
+      "learning_rate": 2.456140350877193e-06,
+      "loss": 4.2777,
+      "step": 28
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 456.3624347928031,
+      "learning_rate": 2.5438596491228075e-06,
+      "loss": 4.0362,
+      "step": 29
+    },
+    {
+      "epoch": 0.07957559681697612,
+      "grad_norm": 424.8598900805099,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 3.932,
+      "step": 30
+    },
+    {
+      "epoch": 0.08222811671087533,
+      "grad_norm": 260.5563303150786,
+      "learning_rate": 2.7192982456140356e-06,
+      "loss": 3.648,
+      "step": 31
+    },
+    {
+      "epoch": 0.08488063660477453,
+      "grad_norm": 268.7488965609003,
+      "learning_rate": 2.8070175438596493e-06,
+      "loss": 3.5179,
+      "step": 32
+    },
+    {
+      "epoch": 0.08753315649867374,
+      "grad_norm": 280.7842985515052,
+      "learning_rate": 2.8947368421052634e-06,
+      "loss": 3.409,
+      "step": 33
+    },
+    {
+      "epoch": 0.09018567639257294,
+      "grad_norm": 352.12688319511284,
+      "learning_rate": 2.9824561403508774e-06,
+      "loss": 3.3254,
+      "step": 34
+    },
+    {
+      "epoch": 0.09283819628647215,
+      "grad_norm": 288.5902403747857,
+      "learning_rate": 3.0701754385964915e-06,
+      "loss": 3.1845,
+      "step": 35
+    },
+    {
+      "epoch": 0.09549071618037135,
+      "grad_norm": 304.1109830674416,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 3.0876,
+      "step": 36
+    },
+    {
+      "epoch": 0.09814323607427056,
+      "grad_norm": 232.21896852625596,
+      "learning_rate": 3.2456140350877197e-06,
+      "loss": 3.0052,
+      "step": 37
+    },
+    {
+      "epoch": 0.10079575596816977,
+      "grad_norm": 255.80964498093417,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 2.9565,
+      "step": 38
+    },
+    {
+      "epoch": 0.10344827586206896,
+      "grad_norm": 222.66022371200634,
+      "learning_rate": 3.421052631578948e-06,
+      "loss": 2.9059,
+      "step": 39
+    },
+    {
+      "epoch": 0.10610079575596817,
+      "grad_norm": 237.06732274651333,
+      "learning_rate": 3.5087719298245615e-06,
+      "loss": 2.8504,
+      "step": 40
+    },
+    {
+      "epoch": 0.10875331564986737,
+      "grad_norm": 298.7469595360898,
+      "learning_rate": 3.596491228070176e-06,
+      "loss": 2.8541,
+      "step": 41
+    },
+    {
+      "epoch": 0.11140583554376658,
+      "grad_norm": 229.14920993336642,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 2.7648,
+      "step": 42
+    },
+    {
+      "epoch": 0.11405835543766578,
+      "grad_norm": 436.82119269000754,
+      "learning_rate": 3.7719298245614037e-06,
+      "loss": 2.8093,
+      "step": 43
+    },
+    {
+      "epoch": 0.11671087533156499,
+      "grad_norm": 256.71171046255853,
+      "learning_rate": 3.859649122807018e-06,
+      "loss": 2.6412,
+      "step": 44
+    },
+    {
+      "epoch": 0.11936339522546419,
+      "grad_norm": 371.9802996469374,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 2.7262,
+      "step": 45
+    },
+    {
+      "epoch": 0.1220159151193634,
+      "grad_norm": 261.98684757391896,
+      "learning_rate": 4.035087719298246e-06,
+      "loss": 2.6309,
+      "step": 46
+    },
+    {
+      "epoch": 0.1246684350132626,
+      "grad_norm": 217.98738126104192,
+      "learning_rate": 4.12280701754386e-06,
+      "loss": 2.6587,
+      "step": 47
+    },
+    {
+      "epoch": 0.1273209549071618,
+      "grad_norm": 262.62076176117773,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 2.728,
+      "step": 48
+    },
+    {
+      "epoch": 0.129973474801061,
+      "grad_norm": 206.72393702517692,
+      "learning_rate": 4.298245614035088e-06,
+      "loss": 2.5768,
+      "step": 49
+    },
+    {
+      "epoch": 0.13262599469496023,
+      "grad_norm": 240.11903830674424,
+      "learning_rate": 4.385964912280702e-06,
+      "loss": 2.6957,
+      "step": 50
+    },
+    {
+      "epoch": 0.13527851458885942,
+      "grad_norm": 363.69145331696933,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 2.9907,
+      "step": 51
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 210.8252588976506,
+      "learning_rate": 4.56140350877193e-06,
+      "loss": 2.4764,
+      "step": 52
+    },
+    {
+      "epoch": 0.14058355437665782,
+      "grad_norm": 235.74742130809076,
+      "learning_rate": 4.649122807017544e-06,
+      "loss": 2.5137,
+      "step": 53
+    },
+    {
+      "epoch": 0.14323607427055704,
+      "grad_norm": 328.4498421369927,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 2.6422,
+      "step": 54
+    },
+    {
+      "epoch": 0.14588859416445624,
+      "grad_norm": 248.58638479316332,
+      "learning_rate": 4.824561403508772e-06,
+      "loss": 2.5274,
+      "step": 55
+    },
+    {
+      "epoch": 0.14854111405835543,
+      "grad_norm": 261.9775254925213,
+      "learning_rate": 4.912280701754386e-06,
+      "loss": 2.5519,
+      "step": 56
+    },
+    {
+      "epoch": 0.15119363395225463,
+      "grad_norm": 281.2075117591644,
+      "learning_rate": 5e-06,
+      "loss": 2.4874,
+      "step": 57
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 253.54349985340127,
+      "learning_rate": 5.087719298245615e-06,
+      "loss": 2.5132,
+      "step": 58
+    },
+    {
+      "epoch": 0.15649867374005305,
+      "grad_norm": 220.72491071710908,
+      "learning_rate": 5.175438596491229e-06,
+      "loss": 2.4212,
+      "step": 59
+    },
+    {
+      "epoch": 0.15915119363395225,
+      "grad_norm": 252.66943722144052,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 2.5385,
+      "step": 60
+    },
+    {
+      "epoch": 0.16180371352785147,
+      "grad_norm": 224.39089747364216,
+      "learning_rate": 5.350877192982457e-06,
+      "loss": 2.4101,
+      "step": 61
+    },
+    {
+      "epoch": 0.16445623342175067,
+      "grad_norm": 262.76676290957175,
+      "learning_rate": 5.438596491228071e-06,
+      "loss": 2.409,
+      "step": 62
+    },
+    {
+      "epoch": 0.16710875331564987,
+      "grad_norm": 323.6668556817878,
+      "learning_rate": 5.526315789473685e-06,
+      "loss": 2.5888,
+      "step": 63
+    },
+    {
+      "epoch": 0.16976127320954906,
+      "grad_norm": 241.96152699479308,
+      "learning_rate": 5.6140350877192985e-06,
+      "loss": 2.4001,
+      "step": 64
+    },
+    {
+      "epoch": 0.1724137931034483,
+      "grad_norm": 239.23277265884937,
+      "learning_rate": 5.701754385964913e-06,
+      "loss": 2.4135,
+      "step": 65
+    },
+    {
+      "epoch": 0.17506631299734748,
+      "grad_norm": 179.6461482025435,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 2.3741,
+      "step": 66
+    },
+    {
+      "epoch": 0.17771883289124668,
+      "grad_norm": 175.576493404712,
+      "learning_rate": 5.877192982456141e-06,
+      "loss": 2.38,
+      "step": 67
+    },
+    {
+      "epoch": 0.18037135278514588,
+      "grad_norm": 223.99940759589632,
+      "learning_rate": 5.964912280701755e-06,
+      "loss": 2.4092,
+      "step": 68
+    },
+    {
+      "epoch": 0.1830238726790451,
+      "grad_norm": 176.1394420449391,
+      "learning_rate": 6.0526315789473685e-06,
+      "loss": 2.3259,
+      "step": 69
+    },
+    {
+      "epoch": 0.1856763925729443,
+      "grad_norm": 196.34227933609677,
+      "learning_rate": 6.140350877192983e-06,
+      "loss": 2.3665,
+      "step": 70
+    },
+    {
+      "epoch": 0.1883289124668435,
+      "grad_norm": 186.26080573377942,
+      "learning_rate": 6.2280701754385975e-06,
+      "loss": 2.3239,
+      "step": 71
+    },
+    {
+      "epoch": 0.1909814323607427,
+      "grad_norm": 250.63759991206928,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 2.4194,
+      "step": 72
+    },
+    {
+      "epoch": 0.19363395225464192,
+      "grad_norm": 184.67595443945896,
+      "learning_rate": 6.403508771929825e-06,
+      "loss": 2.2825,
+      "step": 73
+    },
+    {
+      "epoch": 0.1962864721485411,
+      "grad_norm": 315.4699350961708,
+      "learning_rate": 6.491228070175439e-06,
+      "loss": 2.6249,
+      "step": 74
+    },
+    {
+      "epoch": 0.1989389920424403,
+      "grad_norm": 196.94429675165398,
+      "learning_rate": 6.578947368421054e-06,
+      "loss": 2.2925,
+      "step": 75
+    },
+    {
+      "epoch": 0.20159151193633953,
+      "grad_norm": 232.42841343000674,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 2.5517,
+      "step": 76
+    },
+    {
+      "epoch": 0.20424403183023873,
+      "grad_norm": 169.87354614948592,
+      "learning_rate": 6.754385964912281e-06,
+      "loss": 2.2481,
+      "step": 77
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 178.528334398205,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 2.247,
+      "step": 78
+    },
+    {
+      "epoch": 0.20954907161803712,
+      "grad_norm": 197.92006440702107,
+      "learning_rate": 6.92982456140351e-06,
+      "loss": 2.4433,
+      "step": 79
+    },
+    {
+      "epoch": 0.21220159151193635,
+      "grad_norm": 197.10177112042496,
+      "learning_rate": 7.017543859649123e-06,
+      "loss": 2.2713,
+      "step": 80
+    },
+    {
+      "epoch": 0.21485411140583555,
+      "grad_norm": 172.54053286188068,
+      "learning_rate": 7.1052631578947375e-06,
+      "loss": 2.3019,
+      "step": 81
+    },
+    {
+      "epoch": 0.21750663129973474,
+      "grad_norm": 228.60506906514024,
+      "learning_rate": 7.192982456140352e-06,
+      "loss": 2.1584,
+      "step": 82
+    },
+    {
+      "epoch": 0.22015915119363394,
+      "grad_norm": 137.3028809321962,
+      "learning_rate": 7.280701754385966e-06,
+      "loss": 2.1627,
+      "step": 83
+    },
+    {
+      "epoch": 0.22281167108753316,
+      "grad_norm": 138.40608701456674,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 2.1745,
+      "step": 84
+    },
+    {
+      "epoch": 0.22546419098143236,
+      "grad_norm": 162.95787474014585,
+      "learning_rate": 7.456140350877194e-06,
+      "loss": 2.1807,
+      "step": 85
+    },
+    {
+      "epoch": 0.22811671087533156,
+      "grad_norm": 137.23258451294802,
+      "learning_rate": 7.5438596491228074e-06,
+      "loss": 2.2449,
+      "step": 86
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 136.69569023050238,
+      "learning_rate": 7.631578947368423e-06,
+      "loss": 2.0813,
+      "step": 87
+    },
+    {
+      "epoch": 0.23342175066312998,
+      "grad_norm": 150.51052121600267,
+      "learning_rate": 7.719298245614036e-06,
+      "loss": 2.2483,
+      "step": 88
+    },
+    {
+      "epoch": 0.23607427055702918,
+      "grad_norm": 130.82973054024674,
+      "learning_rate": 7.80701754385965e-06,
+      "loss": 2.0624,
+      "step": 89
+    },
+    {
+      "epoch": 0.23872679045092837,
+      "grad_norm": 197.9266835754521,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 2.5235,
+      "step": 90
+    },
+    {
+      "epoch": 0.2413793103448276,
+      "grad_norm": 139.32851502743648,
+      "learning_rate": 7.982456140350877e-06,
+      "loss": 2.4074,
+      "step": 91
+    },
+    {
+      "epoch": 0.2440318302387268,
+      "grad_norm": 161.62586817939928,
+      "learning_rate": 8.070175438596492e-06,
+      "loss": 2.6251,
+      "step": 92
+    },
+    {
+      "epoch": 0.246684350132626,
+      "grad_norm": 130.369540296953,
+      "learning_rate": 8.157894736842106e-06,
+      "loss": 2.218,
+      "step": 93
+    },
+    {
+      "epoch": 0.2493368700265252,
+      "grad_norm": 172.2161586170502,
+      "learning_rate": 8.24561403508772e-06,
+      "loss": 2.4513,
+      "step": 94
+    },
+    {
+      "epoch": 0.2519893899204244,
+      "grad_norm": 133.83426479043686,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 2.1488,
+      "step": 95
+    },
+    {
+      "epoch": 0.2546419098143236,
+      "grad_norm": 162.54679516690865,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 2.1986,
+      "step": 96
+    },
+    {
+      "epoch": 0.2572944297082228,
+      "grad_norm": 117.31357015065828,
+      "learning_rate": 8.508771929824563e-06,
+      "loss": 2.1464,
+      "step": 97
+    },
+    {
+      "epoch": 0.259946949602122,
+      "grad_norm": 95.62016778990143,
+      "learning_rate": 8.596491228070176e-06,
+      "loss": 1.9437,
+      "step": 98
+    },
+    {
+      "epoch": 0.2625994694960212,
+      "grad_norm": 112.95580192337754,
+      "learning_rate": 8.68421052631579e-06,
+      "loss": 2.0702,
+      "step": 99
+    },
+    {
+      "epoch": 0.26525198938992045,
+      "grad_norm": 133.0945606777901,
+      "learning_rate": 8.771929824561405e-06,
+      "loss": 1.9378,
+      "step": 100
+    },
+    {
+      "epoch": 0.26790450928381965,
+      "grad_norm": 146.2661005058808,
+      "learning_rate": 8.859649122807017e-06,
+      "loss": 1.9705,
+      "step": 101
+    },
+    {
+      "epoch": 0.27055702917771884,
+      "grad_norm": 140.71789603854634,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 2.0788,
+      "step": 102
+    },
+    {
+      "epoch": 0.27320954907161804,
+      "grad_norm": 127.48171170499167,
+      "learning_rate": 9.035087719298246e-06,
+      "loss": 1.8274,
+      "step": 103
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 131.29016341680753,
+      "learning_rate": 9.12280701754386e-06,
+      "loss": 2.2052,
+      "step": 104
+    },
+    {
+      "epoch": 0.27851458885941643,
+      "grad_norm": 83.38173666635986,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 1.8772,
+      "step": 105
+    },
+    {
+      "epoch": 0.28116710875331563,
+      "grad_norm": 95.4005640169517,
+      "learning_rate": 9.298245614035088e-06,
+      "loss": 1.8585,
+      "step": 106
+    },
+    {
+      "epoch": 0.2838196286472148,
+      "grad_norm": 81.6868222976919,
+      "learning_rate": 9.385964912280703e-06,
+      "loss": 1.7842,
+      "step": 107
+    },
+    {
+      "epoch": 0.2864721485411141,
+      "grad_norm": 101.3668892091329,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 1.955,
+      "step": 108
+    },
+    {
+      "epoch": 0.2891246684350133,
+      "grad_norm": 90.10488002182764,
+      "learning_rate": 9.56140350877193e-06,
+      "loss": 1.9475,
+      "step": 109
+    },
+    {
+      "epoch": 0.2917771883289125,
+      "grad_norm": 75.64098223185391,
+      "learning_rate": 9.649122807017545e-06,
+      "loss": 1.7069,
+      "step": 110
+    },
+    {
+      "epoch": 0.29442970822281167,
+      "grad_norm": 70.02492521413883,
+      "learning_rate": 9.736842105263159e-06,
+      "loss": 1.9182,
+      "step": 111
+    },
+    {
+      "epoch": 0.29708222811671087,
+      "grad_norm": 113.78219869230792,
+      "learning_rate": 9.824561403508772e-06,
+      "loss": 2.3653,
+      "step": 112
+    },
+    {
+      "epoch": 0.29973474801061006,
+      "grad_norm": 78.94175240520144,
+      "learning_rate": 9.912280701754386e-06,
+      "loss": 1.9443,
+      "step": 113
+    },
+    {
+      "epoch": 0.30238726790450926,
+      "grad_norm": 78.49120181281637,
+      "learning_rate": 1e-05,
+      "loss": 1.8674,
+      "step": 114
+    },
+    {
+      "epoch": 0.3050397877984085,
+      "grad_norm": 86.72103695721908,
+      "learning_rate": 9.999976144006772e-06,
+      "loss": 2.2703,
+      "step": 115
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 47.848839063498325,
+      "learning_rate": 9.999904576254723e-06,
+      "loss": 1.572,
+      "step": 116
+    },
+    {
+      "epoch": 0.3103448275862069,
+      "grad_norm": 79.61015049499017,
+      "learning_rate": 9.999785297426788e-06,
+      "loss": 1.7971,
+      "step": 117
+    },
+    {
+      "epoch": 0.3129973474801061,
+      "grad_norm": 68.38718933568896,
+      "learning_rate": 9.99961830866117e-06,
+      "loss": 1.9313,
+      "step": 118
+    },
+    {
+      "epoch": 0.3156498673740053,
+      "grad_norm": 51.179876920034914,
+      "learning_rate": 9.999403611551341e-06,
+      "loss": 1.5565,
+      "step": 119
+    },
+    {
+      "epoch": 0.3183023872679045,
+      "grad_norm": 64.29725797705285,
+      "learning_rate": 9.999141208146029e-06,
+      "loss": 1.5536,
+      "step": 120
+    },
+    {
+      "epoch": 0.3209549071618037,
+      "grad_norm": 49.96518518478552,
+      "learning_rate": 9.998831100949188e-06,
+      "loss": 1.4645,
+      "step": 121
+    },
+    {
+      "epoch": 0.32360742705570295,
+      "grad_norm": 49.25606054644458,
+      "learning_rate": 9.998473292919987e-06,
+      "loss": 1.5135,
+      "step": 122
+    },
+    {
+      "epoch": 0.32625994694960214,
+      "grad_norm": 38.39002954508541,
+      "learning_rate": 9.998067787472772e-06,
+      "loss": 1.3848,
+      "step": 123
+    },
+    {
+      "epoch": 0.32891246684350134,
+      "grad_norm": 51.46260995748639,
+      "learning_rate": 9.997614588477033e-06,
+      "loss": 1.5078,
+      "step": 124
+    },
+    {
+      "epoch": 0.33156498673740054,
+      "grad_norm": 47.707086918124624,
+      "learning_rate": 9.997113700257383e-06,
+      "loss": 1.3766,
+      "step": 125
+    },
+    {
+      "epoch": 0.33421750663129973,
+      "grad_norm": 64.04590461927296,
+      "learning_rate": 9.99656512759349e-06,
+      "loss": 1.4349,
+      "step": 126
+    },
+    {
+      "epoch": 0.33687002652519893,
+      "grad_norm": 53.447600347293225,
+      "learning_rate": 9.995968875720052e-06,
+      "loss": 1.482,
+      "step": 127
+    },
+    {
+      "epoch": 0.3395225464190981,
+      "grad_norm": 48.31587410813423,
+      "learning_rate": 9.995324950326746e-06,
+      "loss": 1.3262,
+      "step": 128
+    },
+    {
+      "epoch": 0.3421750663129973,
+      "grad_norm": 69.41825947491212,
+      "learning_rate": 9.994633357558158e-06,
+      "loss": 1.4316,
+      "step": 129
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 57.775756049681775,
+      "learning_rate": 9.993894104013748e-06,
+      "loss": 1.4574,
+      "step": 130
+    },
+    {
+      "epoch": 0.34748010610079577,
+      "grad_norm": 41.27089386693434,
+      "learning_rate": 9.99310719674776e-06,
+      "loss": 1.2503,
+      "step": 131
+    },
+    {
+      "epoch": 0.35013262599469497,
+      "grad_norm": 34.4370860027977,
+      "learning_rate": 9.992272643269181e-06,
+      "loss": 1.0917,
+      "step": 132
+    },
+    {
+      "epoch": 0.35278514588859416,
+      "grad_norm": 47.92471790261903,
+      "learning_rate": 9.99139045154165e-06,
+      "loss": 1.2382,
+      "step": 133
+    },
+    {
+      "epoch": 0.35543766578249336,
+      "grad_norm": 49.9081128353915,
+      "learning_rate": 9.99046062998339e-06,
+      "loss": 1.3388,
+      "step": 134
+    },
+    {
+      "epoch": 0.35809018567639256,
+      "grad_norm": 45.548101311222744,
+      "learning_rate": 9.989483187467128e-06,
+      "loss": 1.127,
+      "step": 135
+    },
+    {
+      "epoch": 0.36074270557029176,
+      "grad_norm": 56.784478228249924,
+      "learning_rate": 9.988458133320009e-06,
+      "loss": 1.29,
+      "step": 136
+    },
+    {
+      "epoch": 0.363395225464191,
+      "grad_norm": 45.43238811368123,
+      "learning_rate": 9.987385477323507e-06,
+      "loss": 1.2792,
+      "step": 137
+    },
+    {
+      "epoch": 0.3660477453580902,
+      "grad_norm": 51.89382859053669,
+      "learning_rate": 9.986265229713332e-06,
+      "loss": 1.3534,
+      "step": 138
+    },
+    {
+      "epoch": 0.3687002652519894,
+      "grad_norm": 46.33340281664256,
+      "learning_rate": 9.985097401179333e-06,
+      "loss": 1.3419,
+      "step": 139
+    },
+    {
+      "epoch": 0.3713527851458886,
+      "grad_norm": 35.94198985151234,
+      "learning_rate": 9.983882002865392e-06,
+      "loss": 0.9718,
+      "step": 140
+    },
+    {
+      "epoch": 0.3740053050397878,
+      "grad_norm": 35.294849184766264,
+      "learning_rate": 9.982619046369321e-06,
+      "loss": 1.003,
+      "step": 141
+    },
+    {
+      "epoch": 0.376657824933687,
+      "grad_norm": 42.16976850569041,
+      "learning_rate": 9.981308543742759e-06,
+      "loss": 1.2051,
+      "step": 142
+    },
+    {
+      "epoch": 0.3793103448275862,
+      "grad_norm": 47.66629555583024,
+      "learning_rate": 9.979950507491035e-06,
+      "loss": 1.499,
+      "step": 143
+    },
+    {
+      "epoch": 0.3819628647214854,
+      "grad_norm": 31.799357696770404,
+      "learning_rate": 9.978544950573075e-06,
+      "loss": 0.9198,
+      "step": 144
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 31.399975284772456,
+      "learning_rate": 9.97709188640126e-06,
+      "loss": 0.8313,
+      "step": 145
+    },
+    {
+      "epoch": 0.38726790450928383,
+      "grad_norm": 39.60233683027201,
+      "learning_rate": 9.975591328841306e-06,
+      "loss": 0.9907,
+      "step": 146
+    },
+    {
+      "epoch": 0.38992042440318303,
+      "grad_norm": 50.72888144941861,
+      "learning_rate": 9.974043292212129e-06,
+      "loss": 0.9659,
+      "step": 147
+    },
+    {
+      "epoch": 0.3925729442970822,
+      "grad_norm": 34.2382037770928,
+      "learning_rate": 9.97244779128571e-06,
+      "loss": 0.8095,
+      "step": 148
+    },
+    {
+      "epoch": 0.3952254641909814,
+      "grad_norm": 30.53623445439956,
+      "learning_rate": 9.970804841286954e-06,
+      "loss": 0.7986,
+      "step": 149
+    },
+    {
+      "epoch": 0.3978779840848806,
+      "grad_norm": 21.32608698047807,
+      "learning_rate": 9.96911445789354e-06,
+      "loss": 0.656,
+      "step": 150
+    },
+    {
+      "epoch": 0.4005305039787798,
+      "grad_norm": 27.679120212238363,
+      "learning_rate": 9.96737665723578e-06,
+      "loss": 0.7913,
+      "step": 151
+    },
+    {
+      "epoch": 0.40318302387267907,
+      "grad_norm": 32.969472057658585,
+      "learning_rate": 9.965591455896456e-06,
+      "loss": 0.841,
+      "step": 152
+    },
+    {
+      "epoch": 0.40583554376657827,
+      "grad_norm": 412.28241091884604,
+      "learning_rate": 9.963758870910672e-06,
+      "loss": 1.2917,
+      "step": 153
+    },
+    {
+      "epoch": 0.40848806366047746,
+      "grad_norm": 49.04456362529717,
+      "learning_rate": 9.961878919765678e-06,
+      "loss": 0.8898,
+      "step": 154
+    },
+    {
+      "epoch": 0.41114058355437666,
+      "grad_norm": 39.36253019431454,
+      "learning_rate": 9.95995162040072e-06,
+      "loss": 0.7099,
+      "step": 155
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 34.63935013627246,
+      "learning_rate": 9.957976991206847e-06,
+      "loss": 0.7765,
+      "step": 156
+    },
+    {
+      "epoch": 0.41644562334217505,
+      "grad_norm": 26.840735102660002,
+      "learning_rate": 9.95595505102676e-06,
+      "loss": 0.6499,
+      "step": 157
+    },
+    {
+      "epoch": 0.41909814323607425,
+      "grad_norm": 63.18942895204754,
+      "learning_rate": 9.953885819154615e-06,
+      "loss": 1.0861,
+      "step": 158
+    },
+    {
+      "epoch": 0.4217506631299735,
+      "grad_norm": 25.445198024275147,
+      "learning_rate": 9.951769315335843e-06,
+      "loss": 0.7166,
+      "step": 159
+    },
+    {
+      "epoch": 0.4244031830238727,
+      "grad_norm": 26.58185371356998,
+      "learning_rate": 9.949605559766969e-06,
+      "loss": 0.5657,
+      "step": 160
+    },
+    {
+      "epoch": 0.4270557029177719,
+      "grad_norm": 27.958321235093127,
+      "learning_rate": 9.947394573095403e-06,
+      "loss": 0.5804,
+      "step": 161
+    },
+    {
+      "epoch": 0.4297082228116711,
+      "grad_norm": 29.962873219114176,
+      "learning_rate": 9.94513637641926e-06,
+      "loss": 0.5276,
+      "step": 162
+    },
+    {
+      "epoch": 0.4323607427055703,
+      "grad_norm": 25.88658622108772,
+      "learning_rate": 9.942830991287149e-06,
+      "loss": 0.5638,
+      "step": 163
+    },
+    {
+      "epoch": 0.4350132625994695,
+      "grad_norm": 28.58328032342846,
+      "learning_rate": 9.940478439697973e-06,
+      "loss": 0.5453,
+      "step": 164
+    },
+    {
+      "epoch": 0.4376657824933687,
+      "grad_norm": 31.677528986906,
+      "learning_rate": 9.938078744100713e-06,
+      "loss": 0.5716,
+      "step": 165
+    },
+    {
+      "epoch": 0.4403183023872679,
+      "grad_norm": 26.616435831422383,
+      "learning_rate": 9.935631927394216e-06,
+      "loss": 0.5028,
+      "step": 166
+    },
+    {
+      "epoch": 0.44297082228116713,
+      "grad_norm": 34.69859278391251,
+      "learning_rate": 9.933138012926982e-06,
+      "loss": 0.5672,
+      "step": 167
+    },
+    {
+      "epoch": 0.44562334217506633,
+      "grad_norm": 24.597680749706683,
+      "learning_rate": 9.930597024496933e-06,
+      "loss": 0.4989,
+      "step": 168
+    },
+    {
+      "epoch": 0.4482758620689655,
+      "grad_norm": 47.44734871590014,
+      "learning_rate": 9.928008986351187e-06,
+      "loss": 0.5661,
+      "step": 169
+    },
+    {
+      "epoch": 0.4509283819628647,
+      "grad_norm": 27.688296544663764,
+      "learning_rate": 9.925373923185835e-06,
+      "loss": 0.5102,
+      "step": 170
+    },
+    {
+      "epoch": 0.4535809018567639,
+      "grad_norm": 28.913477864873013,
+      "learning_rate": 9.922691860145696e-06,
+      "loss": 0.527,
+      "step": 171
+    },
+    {
+      "epoch": 0.4562334217506631,
+      "grad_norm": 23.76336979943884,
+      "learning_rate": 9.919962822824083e-06,
+      "loss": 0.4584,
+      "step": 172
+    },
+    {
+      "epoch": 0.4588859416445623,
+      "grad_norm": 27.428221662294682,
+      "learning_rate": 9.917186837262552e-06,
+      "loss": 0.5161,
+      "step": 173
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 24.816760913180836,
+      "learning_rate": 9.91436392995066e-06,
+      "loss": 0.4373,
+      "step": 174
+    },
+    {
+      "epoch": 0.46419098143236076,
+      "grad_norm": 20.042908299766527,
+      "learning_rate": 9.91149412782571e-06,
+      "loss": 0.3862,
+      "step": 175
+    },
+    {
+      "epoch": 0.46684350132625996,
+      "grad_norm": 22.52969310546478,
+      "learning_rate": 9.908577458272496e-06,
+      "loss": 0.4759,
+      "step": 176
+    },
+    {
+      "epoch": 0.46949602122015915,
+      "grad_norm": 35.08325684519957,
+      "learning_rate": 9.905613949123036e-06,
+      "loss": 0.443,
+      "step": 177
+    },
+    {
+      "epoch": 0.47214854111405835,
+      "grad_norm": 27.25235345890855,
+      "learning_rate": 9.902603628656312e-06,
+      "loss": 0.3758,
+      "step": 178
+    },
+    {
+      "epoch": 0.47480106100795755,
+      "grad_norm": 45.22071533665291,
+      "learning_rate": 9.899546525597998e-06,
+      "loss": 0.5609,
+      "step": 179
+    },
+    {
+      "epoch": 0.47745358090185674,
+      "grad_norm": 37.26690294810362,
+      "learning_rate": 9.896442669120188e-06,
+      "loss": 0.4942,
+      "step": 180
+    },
+    {
+      "epoch": 0.48010610079575594,
+      "grad_norm": 29.928428373271707,
+      "learning_rate": 9.893292088841109e-06,
+      "loss": 0.4503,
+      "step": 181
+    },
+    {
+      "epoch": 0.4827586206896552,
+      "grad_norm": 21.722116105223737,
+      "learning_rate": 9.890094814824854e-06,
+      "loss": 0.3151,
+      "step": 182
+    },
+    {
+      "epoch": 0.4854111405835544,
+      "grad_norm": 24.607735893749748,
+      "learning_rate": 9.886850877581079e-06,
+      "loss": 0.3298,
+      "step": 183
+    },
+    {
+      "epoch": 0.4880636604774536,
+      "grad_norm": 33.43828584668856,
+      "learning_rate": 9.883560308064723e-06,
+      "loss": 0.3457,
+      "step": 184
+    },
+    {
+      "epoch": 0.4907161803713528,
+      "grad_norm": 31.065576895170462,
+      "learning_rate": 9.880223137675709e-06,
+      "loss": 0.3693,
+      "step": 185
+    },
+    {
+      "epoch": 0.493368700265252,
+      "grad_norm": 37.49242844056875,
+      "learning_rate": 9.87683939825864e-06,
+      "loss": 0.4888,
+      "step": 186
+    },
+    {
+      "epoch": 0.4960212201591512,
+      "grad_norm": 38.74998441973847,
+      "learning_rate": 9.873409122102505e-06,
+      "loss": 0.468,
+      "step": 187
+    },
+    {
+      "epoch": 0.4986737400530504,
+      "grad_norm": 35.008093098415614,
+      "learning_rate": 9.86993234194036e-06,
+      "loss": 0.4246,
+      "step": 188
+    },
+    {
+      "epoch": 0.5013262599469496,
+      "grad_norm": 27.804082583361943,
+      "learning_rate": 9.866409090949023e-06,
+      "loss": 0.3621,
+      "step": 189
+    },
+    {
+      "epoch": 0.5039787798408488,
+      "grad_norm": 25.629862496043533,
+      "learning_rate": 9.862839402748754e-06,
+      "loss": 0.4143,
+      "step": 190
+    },
+    {
+      "epoch": 0.506631299734748,
+      "grad_norm": 41.827303798404294,
+      "learning_rate": 9.859223311402937e-06,
+      "loss": 0.4188,
+      "step": 191
+    },
+    {
+      "epoch": 0.5092838196286472,
+      "grad_norm": 24.462284819105808,
+      "learning_rate": 9.855560851417752e-06,
+      "loss": 0.3295,
+      "step": 192
+    },
+    {
+      "epoch": 0.5119363395225465,
+      "grad_norm": 22.79091787353534,
+      "learning_rate": 9.851852057741846e-06,
+      "loss": 0.2292,
+      "step": 193
+    },
+    {
+      "epoch": 0.5145888594164456,
+      "grad_norm": 31.885324600524022,
+      "learning_rate": 9.848096965766005e-06,
+      "loss": 0.401,
+      "step": 194
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 30.710948037812585,
+      "learning_rate": 9.844295611322804e-06,
+      "loss": 0.323,
+      "step": 195
+    },
+    {
+      "epoch": 0.519893899204244,
+      "grad_norm": 41.79890030360689,
+      "learning_rate": 9.84044803068628e-06,
+      "loss": 0.3313,
+      "step": 196
+    },
+    {
+      "epoch": 0.5225464190981433,
+      "grad_norm": 22.535018094427123,
+      "learning_rate": 9.836554260571577e-06,
+      "loss": 0.3267,
+      "step": 197
+    },
+    {
+      "epoch": 0.5251989389920424,
+      "grad_norm": 34.92359618277551,
+      "learning_rate": 9.832614338134595e-06,
+      "loss": 0.4107,
+      "step": 198
+    },
+    {
+      "epoch": 0.5278514588859416,
+      "grad_norm": 38.47784571433333,
+      "learning_rate": 9.828628300971639e-06,
+      "loss": 0.3746,
+      "step": 199
+    },
+    {
+      "epoch": 0.5305039787798409,
+      "grad_norm": 51.09946363547108,
+      "learning_rate": 9.82459618711906e-06,
+      "loss": 0.425,
+      "step": 200
+    },
+    {
+      "epoch": 0.53315649867374,
+      "grad_norm": 31.35834407033397,
+      "learning_rate": 9.82051803505289e-06,
+      "loss": 0.3532,
+      "step": 201
+    },
+    {
+      "epoch": 0.5358090185676393,
+      "grad_norm": 26.935093809636733,
+      "learning_rate": 9.816393883688475e-06,
+      "loss": 0.2851,
+      "step": 202
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 22.52622655874006,
+      "learning_rate": 9.812223772380107e-06,
+      "loss": 0.2908,
+      "step": 203
+    },
+    {
+      "epoch": 0.5411140583554377,
+      "grad_norm": 51.489094847599844,
+      "learning_rate": 9.808007740920647e-06,
+      "loss": 0.5435,
+      "step": 204
+    },
+    {
+      "epoch": 0.5437665782493368,
+      "grad_norm": 59.633888635972156,
+      "learning_rate": 9.803745829541138e-06,
+      "loss": 0.3766,
+      "step": 205
+    },
+    {
+      "epoch": 0.5464190981432361,
+      "grad_norm": 38.02033428452659,
+      "learning_rate": 9.799438078910433e-06,
+      "loss": 0.2728,
+      "step": 206
+    },
+    {
+      "epoch": 0.5490716180371353,
+      "grad_norm": 51.48954310514882,
+      "learning_rate": 9.795084530134801e-06,
+      "loss": 0.2818,
+      "step": 207
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 53.46148414443766,
+      "learning_rate": 9.790685224757534e-06,
+      "loss": 0.4253,
+      "step": 208
+    },
+    {
+      "epoch": 0.5543766578249337,
+      "grad_norm": 43.35134143380862,
+      "learning_rate": 9.786240204758552e-06,
+      "loss": 0.3331,
+      "step": 209
+    },
+    {
+      "epoch": 0.5570291777188329,
+      "grad_norm": 57.29778955186939,
+      "learning_rate": 9.781749512554e-06,
+      "loss": 0.3571,
+      "step": 210
+    },
+    {
+      "epoch": 0.5596816976127321,
+      "grad_norm": 49.20361109100751,
+      "learning_rate": 9.777213190995849e-06,
+      "loss": 0.4044,
+      "step": 211
+    },
+    {
+      "epoch": 0.5623342175066313,
+      "grad_norm": 38.82482405915261,
+      "learning_rate": 9.772631283371481e-06,
+      "loss": 0.3374,
+      "step": 212
+    },
+    {
+      "epoch": 0.5649867374005305,
+      "grad_norm": 48.021781844446835,
+      "learning_rate": 9.768003833403278e-06,
+      "loss": 0.3962,
+      "step": 213
+    },
+    {
+      "epoch": 0.5676392572944297,
+      "grad_norm": 52.28939094494638,
+      "learning_rate": 9.763330885248206e-06,
+      "loss": 0.2568,
+      "step": 214
+    },
+    {
+      "epoch": 0.5702917771883289,
+      "grad_norm": 58.18428330420366,
+      "learning_rate": 9.758612483497395e-06,
+      "loss": 0.4075,
+      "step": 215
+    },
+    {
+      "epoch": 0.5729442970822282,
+      "grad_norm": 65.91359809789627,
+      "learning_rate": 9.753848673175707e-06,
+      "loss": 0.3884,
+      "step": 216
+    },
+    {
+      "epoch": 0.5755968169761273,
+      "grad_norm": 85.87965148754081,
+      "learning_rate": 9.749039499741313e-06,
+      "loss": 0.3771,
+      "step": 217
+    },
+    {
+      "epoch": 0.5782493368700266,
+      "grad_norm": 173.13163239297148,
+      "learning_rate": 9.744185009085258e-06,
+      "loss": 0.4767,
+      "step": 218
+    },
+    {
+      "epoch": 0.5809018567639257,
+      "grad_norm": 74.19989440496742,
+      "learning_rate": 9.739285247531019e-06,
+      "loss": 0.5952,
+      "step": 219
+    },
+    {
+      "epoch": 0.583554376657825,
+      "grad_norm": 67.181613202375,
+      "learning_rate": 9.734340261834068e-06,
+      "loss": 0.4403,
+      "step": 220
+    },
+    {
+      "epoch": 0.5862068965517241,
+      "grad_norm": 87.56458995512162,
+      "learning_rate": 9.72935009918142e-06,
+      "loss": 0.3334,
+      "step": 221
+    },
+    {
+      "epoch": 0.5888594164456233,
+      "grad_norm": 98.26463421494698,
+      "learning_rate": 9.724314807191197e-06,
+      "loss": 0.5103,
+      "step": 222
+    },
+    {
+      "epoch": 0.5915119363395226,
+      "grad_norm": 371.4672878296043,
+      "learning_rate": 9.719234433912148e-06,
+      "loss": 0.6872,
+      "step": 223
+    },
+    {
+      "epoch": 0.5941644562334217,
+      "grad_norm": 87.40317227687746,
+      "learning_rate": 9.714109027823218e-06,
+      "loss": 0.4359,
+      "step": 224
+    },
+    {
+      "epoch": 0.596816976127321,
+      "grad_norm": 141.95076404109608,
+      "learning_rate": 9.708938637833065e-06,
+      "loss": 0.4145,
+      "step": 225
+    },
+    {
+      "epoch": 0.5994694960212201,
+      "grad_norm": 64.57421523606523,
+      "learning_rate": 9.703723313279607e-06,
+      "loss": 0.3707,
+      "step": 226
+    },
+    {
+      "epoch": 0.6021220159151194,
+      "grad_norm": 79.25888498842731,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.3337,
+      "step": 227
+    },
+    {
+      "epoch": 0.6047745358090185,
+      "grad_norm": 138.38119877155972,
+      "learning_rate": 9.693158059977879e-06,
+      "loss": 0.3935,
+      "step": 228
+    },
+    {
+      "epoch": 0.6074270557029178,
+      "grad_norm": 62.52742051706753,
+      "learning_rate": 9.687808232047452e-06,
+      "loss": 0.4193,
+      "step": 229
+    },
+    {
+      "epoch": 0.610079575596817,
+      "grad_norm": 73.92852240551677,
+      "learning_rate": 9.682413671188444e-06,
+      "loss": 0.3691,
+      "step": 230
+    },
+    {
+      "epoch": 0.6127320954907162,
+      "grad_norm": 103.53554748059803,
+      "learning_rate": 9.6769744288779e-06,
+      "loss": 0.3535,
+      "step": 231
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 42.7125473553319,
+      "learning_rate": 9.671490557019234e-06,
+      "loss": 0.3155,
+      "step": 232
+    },
+    {
+      "epoch": 0.6180371352785146,
+      "grad_norm": 54.829192160820845,
+      "learning_rate": 9.665962107941725e-06,
+      "loss": 0.2673,
+      "step": 233
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 305.3150318928213,
+      "learning_rate": 9.660389134400034e-06,
+      "loss": 0.2952,
+      "step": 234
+    },
+    {
+      "epoch": 0.623342175066313,
+      "grad_norm": 212.30409368555127,
+      "learning_rate": 9.654771689573685e-06,
+      "loss": 0.3867,
+      "step": 235
+    },
+    {
+      "epoch": 0.6259946949602122,
+      "grad_norm": 105.43931388039499,
+      "learning_rate": 9.649109827066572e-06,
+      "loss": 0.3817,
+      "step": 236
+    },
+    {
+      "epoch": 0.6286472148541115,
+      "grad_norm": 81.59794116200509,
+      "learning_rate": 9.643403600906433e-06,
+      "loss": 0.4814,
+      "step": 237
+    },
+    {
+      "epoch": 0.6312997347480106,
+      "grad_norm": 214.11174329135335,
+      "learning_rate": 9.637653065544349e-06,
+      "loss": 0.371,
+      "step": 238
+    },
+    {
+      "epoch": 0.6339522546419099,
+      "grad_norm": 118.89467666014248,
+      "learning_rate": 9.63185827585421e-06,
+      "loss": 0.3935,
+      "step": 239
+    },
+    {
+      "epoch": 0.636604774535809,
+      "grad_norm": 74.98460803487606,
+      "learning_rate": 9.626019287132202e-06,
+      "loss": 0.4457,
+      "step": 240
+    },
+    {
+      "epoch": 0.6392572944297082,
+      "grad_norm": 41.472870074642465,
+      "learning_rate": 9.620136155096276e-06,
+      "loss": 0.2736,
+      "step": 241
+    },
+    {
+      "epoch": 0.6419098143236074,
+      "grad_norm": 74.2956558934981,
+      "learning_rate": 9.614208935885615e-06,
+      "loss": 0.5428,
+      "step": 242
+    },
+    {
+      "epoch": 0.6445623342175066,
+      "grad_norm": 127.0632945401361,
+      "learning_rate": 9.608237686060099e-06,
+      "loss": 0.4162,
+      "step": 243
+    },
+    {
+      "epoch": 0.6472148541114059,
+      "grad_norm": 103.70620760268216,
+      "learning_rate": 9.602222462599768e-06,
+      "loss": 0.5392,
+      "step": 244
+    },
+    {
+      "epoch": 0.649867374005305,
+      "grad_norm": 90.43287095478969,
+      "learning_rate": 9.59616332290427e-06,
+      "loss": 0.3134,
+      "step": 245
+    },
+    {
+      "epoch": 0.6525198938992043,
+      "grad_norm": 254.77023599890552,
+      "learning_rate": 9.590060324792328e-06,
+      "loss": 0.329,
+      "step": 246
+    },
+    {
+      "epoch": 0.6551724137931034,
+      "grad_norm": 58.742322679359965,
+      "learning_rate": 9.58391352650117e-06,
+      "loss": 0.3592,
+      "step": 247
+    },
+    {
+      "epoch": 0.6578249336870027,
+      "grad_norm": 98.73104949830449,
+      "learning_rate": 9.577722986685992e-06,
+      "loss": 0.331,
+      "step": 248
+    },
+    {
+      "epoch": 0.6604774535809018,
+      "grad_norm": 109.9795313881565,
+      "learning_rate": 9.571488764419381e-06,
+      "loss": 0.3457,
+      "step": 249
+    },
+    {
+      "epoch": 0.6631299734748011,
+      "grad_norm": 60.5343482414627,
+      "learning_rate": 9.565210919190764e-06,
+      "loss": 0.3269,
+      "step": 250
+    },
+    {
+      "epoch": 0.6657824933687002,
+      "grad_norm": 35.54885592977575,
+      "learning_rate": 9.558889510905836e-06,
+      "loss": 0.2986,
+      "step": 251
+    },
+    {
+      "epoch": 0.6684350132625995,
+      "grad_norm": 36.55925854943744,
+      "learning_rate": 9.552524599885982e-06,
+      "loss": 0.2985,
+      "step": 252
+    },
+    {
+      "epoch": 0.6710875331564987,
+      "grad_norm": 54.695899509794316,
+      "learning_rate": 9.546116246867716e-06,
+      "loss": 0.2343,
+      "step": 253
+    },
+    {
+      "epoch": 0.6737400530503979,
+      "grad_norm": 60.49903065247721,
+      "learning_rate": 9.539664513002085e-06,
+      "loss": 0.2674,
+      "step": 254
+    },
+    {
+      "epoch": 0.6763925729442971,
+      "grad_norm": 62.93317082598356,
+      "learning_rate": 9.5331694598541e-06,
+      "loss": 0.3461,
+      "step": 255
+    },
+    {
+      "epoch": 0.6790450928381963,
+      "grad_norm": 116.04475232722392,
+      "learning_rate": 9.526631149402135e-06,
+      "loss": 0.3104,
+      "step": 256
+    },
+    {
+      "epoch": 0.6816976127320955,
+      "grad_norm": 67.25880114731598,
+      "learning_rate": 9.520049644037349e-06,
+      "loss": 0.3375,
+      "step": 257
+    },
+    {
+      "epoch": 0.6843501326259946,
+      "grad_norm": 71.72910226758626,
+      "learning_rate": 9.51342500656308e-06,
+      "loss": 0.3797,
+      "step": 258
+    },
+    {
+      "epoch": 0.6870026525198939,
+      "grad_norm": 46.44975219408069,
+      "learning_rate": 9.506757300194249e-06,
+      "loss": 0.2771,
+      "step": 259
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 106.56600861392955,
+      "learning_rate": 9.500046588556762e-06,
+      "loss": 0.3469,
+      "step": 260
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 60.83080447251247,
+      "learning_rate": 9.493292935686896e-06,
+      "loss": 0.3156,
+      "step": 261
+    },
+    {
+      "epoch": 0.6949602122015915,
+      "grad_norm": 33.135336113479035,
+      "learning_rate": 9.486496406030687e-06,
+      "loss": 0.3143,
+      "step": 262
+    },
+    {
+      "epoch": 0.6976127320954907,
+      "grad_norm": 75.21806103795748,
+      "learning_rate": 9.479657064443321e-06,
+      "loss": 0.5503,
+      "step": 263
+    },
+    {
+      "epoch": 0.7002652519893899,
+      "grad_norm": 53.35765416421378,
+      "learning_rate": 9.472774976188515e-06,
+      "loss": 0.335,
+      "step": 264
+    },
+    {
+      "epoch": 0.7029177718832891,
+      "grad_norm": 23.697336562644033,
+      "learning_rate": 9.46585020693789e-06,
+      "loss": 0.2394,
+      "step": 265
+    },
+    {
+      "epoch": 0.7055702917771883,
+      "grad_norm": 29.870651851658987,
+      "learning_rate": 9.458882822770342e-06,
+      "loss": 0.3299,
+      "step": 266
+    },
+    {
+      "epoch": 0.7082228116710876,
+      "grad_norm": 41.11458931120419,
+      "learning_rate": 9.451872890171419e-06,
+      "loss": 0.3131,
+      "step": 267
+    },
+    {
+      "epoch": 0.7108753315649867,
+      "grad_norm": 28.525249893009217,
+      "learning_rate": 9.444820476032687e-06,
+      "loss": 0.3914,
+      "step": 268
+    },
+    {
+      "epoch": 0.713527851458886,
+      "grad_norm": 32.78558988940678,
+      "learning_rate": 9.43772564765108e-06,
+      "loss": 0.3668,
+      "step": 269
+    },
+    {
+      "epoch": 0.7161803713527851,
+      "grad_norm": 32.87361891444267,
+      "learning_rate": 9.430588472728271e-06,
+      "loss": 0.4,
+      "step": 270
+    },
+    {
+      "epoch": 0.7188328912466844,
+      "grad_norm": 34.95801441555627,
+      "learning_rate": 9.423409019370015e-06,
+      "loss": 0.3718,
+      "step": 271
+    },
+    {
+      "epoch": 0.7214854111405835,
+      "grad_norm": 33.71518005464951,
+      "learning_rate": 9.416187356085513e-06,
+      "loss": 0.3528,
+      "step": 272
+    },
+    {
+      "epoch": 0.7241379310344828,
+      "grad_norm": 24.622857735134367,
+      "learning_rate": 9.408923551786742e-06,
+      "loss": 0.2989,
+      "step": 273
+    },
+    {
+      "epoch": 0.726790450928382,
+      "grad_norm": 20.513102908220276,
+      "learning_rate": 9.401617675787812e-06,
+      "loss": 0.2849,
+      "step": 274
+    },
+    {
+      "epoch": 0.7294429708222812,
+      "grad_norm": 49.62935372844656,
+      "learning_rate": 9.39426979780429e-06,
+      "loss": 0.3088,
+      "step": 275
+    },
+    {
+      "epoch": 0.7320954907161804,
+      "grad_norm": 99.4036166043459,
+      "learning_rate": 9.386879987952549e-06,
+      "loss": 0.3753,
+      "step": 276
+    },
+    {
+      "epoch": 0.7347480106100795,
+      "grad_norm": 37.958768445614425,
+      "learning_rate": 9.379448316749092e-06,
+      "loss": 0.4278,
+      "step": 277
+    },
+    {
+      "epoch": 0.7374005305039788,
+      "grad_norm": 22.952448667294956,
+      "learning_rate": 9.371974855109876e-06,
+      "loss": 0.3055,
+      "step": 278
+    },
+    {
+      "epoch": 0.7400530503978779,
+      "grad_norm": 32.39914140080223,
+      "learning_rate": 9.364459674349642e-06,
+      "loss": 0.39,
+      "step": 279
+    },
+    {
+      "epoch": 0.7427055702917772,
+      "grad_norm": 262.64318764193524,
+      "learning_rate": 9.356902846181229e-06,
+      "loss": 0.3677,
+      "step": 280
+    },
+    {
+      "epoch": 0.7453580901856764,
+      "grad_norm": 22.215532388982677,
+      "learning_rate": 9.349304442714895e-06,
+      "loss": 0.2969,
+      "step": 281
+    },
+    {
+      "epoch": 0.7480106100795756,
+      "grad_norm": 17.778109461565176,
+      "learning_rate": 9.341664536457626e-06,
+      "loss": 0.271,
+      "step": 282
+    },
+    {
+      "epoch": 0.7506631299734748,
+      "grad_norm": 29.2037598726396,
+      "learning_rate": 9.33398320031244e-06,
+      "loss": 0.3859,
+      "step": 283
+    },
+    {
+      "epoch": 0.753315649867374,
+      "grad_norm": 22.099580190625854,
+      "learning_rate": 9.326260507577702e-06,
+      "loss": 0.2939,
+      "step": 284
+    },
+    {
+      "epoch": 0.7559681697612732,
+      "grad_norm": 35.83350243227712,
+      "learning_rate": 9.318496531946411e-06,
+      "loss": 0.3866,
+      "step": 285
+    },
+    {
+      "epoch": 0.7586206896551724,
+      "grad_norm": 23.56156299043329,
+      "learning_rate": 9.310691347505506e-06,
+      "loss": 0.2483,
+      "step": 286
+    },
+    {
+      "epoch": 0.7612732095490716,
+      "grad_norm": 19.983478638726627,
+      "learning_rate": 9.30284502873516e-06,
+      "loss": 0.3167,
+      "step": 287
+    },
+    {
+      "epoch": 0.7639257294429708,
+      "grad_norm": 48.94714254686104,
+      "learning_rate": 9.294957650508065e-06,
+      "loss": 0.3196,
+      "step": 288
+    },
+    {
+      "epoch": 0.76657824933687,
+      "grad_norm": 16.30906262917392,
+      "learning_rate": 9.287029288088716e-06,
+      "loss": 0.2335,
+      "step": 289
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 25.475699429435473,
+      "learning_rate": 9.279060017132698e-06,
+      "loss": 0.2862,
+      "step": 290
+    },
+    {
+      "epoch": 0.7718832891246684,
+      "grad_norm": 28.384146206526353,
+      "learning_rate": 9.27104991368596e-06,
+      "loss": 0.2908,
+      "step": 291
+    },
+    {
+      "epoch": 0.7745358090185677,
+      "grad_norm": 48.12691541658247,
+      "learning_rate": 9.262999054184093e-06,
+      "loss": 0.2833,
+      "step": 292
+    },
+    {
+      "epoch": 0.7771883289124668,
+      "grad_norm": 44.07215651045462,
+      "learning_rate": 9.254907515451593e-06,
+      "loss": 0.2849,
+      "step": 293
+    },
+    {
+      "epoch": 0.7798408488063661,
+      "grad_norm": 39.591136153451586,
+      "learning_rate": 9.246775374701139e-06,
+      "loss": 0.2239,
+      "step": 294
+    },
+    {
+      "epoch": 0.7824933687002652,
+      "grad_norm": 25.79331189444745,
+      "learning_rate": 9.238602709532851e-06,
+      "loss": 0.1937,
+      "step": 295
+    },
+    {
+      "epoch": 0.7851458885941645,
+      "grad_norm": 17.94060621935374,
+      "learning_rate": 9.230389597933545e-06,
+      "loss": 0.2606,
+      "step": 296
+    },
+    {
+      "epoch": 0.7877984084880637,
+      "grad_norm": 24.95938848211813,
+      "learning_rate": 9.222136118275996e-06,
+      "loss": 0.3266,
+      "step": 297
+    },
+    {
+      "epoch": 0.7904509283819628,
+      "grad_norm": 23.220036606199297,
+      "learning_rate": 9.213842349318185e-06,
+      "loss": 0.2797,
+      "step": 298
+    },
+    {
+      "epoch": 0.7931034482758621,
+      "grad_norm": 22.330136839566496,
+      "learning_rate": 9.205508370202552e-06,
+      "loss": 0.3015,
+      "step": 299
+    },
+    {
+      "epoch": 0.7957559681697612,
+      "grad_norm": 32.818473260764726,
+      "learning_rate": 9.197134260455233e-06,
+      "loss": 0.3396,
+      "step": 300
+    },
+    {
+      "epoch": 0.7984084880636605,
+      "grad_norm": 36.20294272608413,
+      "learning_rate": 9.188720099985316e-06,
+      "loss": 0.3389,
+      "step": 301
+    },
+    {
+      "epoch": 0.8010610079575596,
+      "grad_norm": 43.50941855978113,
+      "learning_rate": 9.180265969084058e-06,
+      "loss": 0.3658,
+      "step": 302
+    },
+    {
+      "epoch": 0.8037135278514589,
+      "grad_norm": 27.768953084611958,
+      "learning_rate": 9.171771948424138e-06,
+      "loss": 0.2811,
+      "step": 303
+    },
+    {
+      "epoch": 0.8063660477453581,
+      "grad_norm": 19.52078292934591,
+      "learning_rate": 9.163238119058873e-06,
+      "loss": 0.2006,
+      "step": 304
+    },
+    {
+      "epoch": 0.8090185676392573,
+      "grad_norm": 29.981050417557388,
+      "learning_rate": 9.154664562421453e-06,
+      "loss": 0.4175,
+      "step": 305
+    },
+    {
+      "epoch": 0.8116710875331565,
+      "grad_norm": 40.14474678185231,
+      "learning_rate": 9.146051360324166e-06,
+      "loss": 0.355,
+      "step": 306
+    },
+    {
+      "epoch": 0.8143236074270557,
+      "grad_norm": 33.53880491765791,
+      "learning_rate": 9.137398594957605e-06,
+      "loss": 0.3023,
+      "step": 307
+    },
+    {
+      "epoch": 0.8169761273209549,
+      "grad_norm": 27.011343413853503,
+      "learning_rate": 9.128706348889895e-06,
+      "loss": 0.3198,
+      "step": 308
+    },
+    {
+      "epoch": 0.8196286472148541,
+      "grad_norm": 31.239861629198902,
+      "learning_rate": 9.119974705065902e-06,
+      "loss": 0.2698,
+      "step": 309
+    },
+    {
+      "epoch": 0.8222811671087533,
+      "grad_norm": 160.25189491094991,
+      "learning_rate": 9.111203746806439e-06,
+      "loss": 0.2455,
+      "step": 310
+    },
+    {
+      "epoch": 0.8249336870026526,
+      "grad_norm": 19.517004117689925,
+      "learning_rate": 9.102393557807476e-06,
+      "loss": 0.2269,
+      "step": 311
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 31.404118966904665,
+      "learning_rate": 9.093544222139338e-06,
+      "loss": 0.3547,
+      "step": 312
+    },
+    {
+      "epoch": 0.830238726790451,
+      "grad_norm": 23.04361811420225,
+      "learning_rate": 9.084655824245899e-06,
+      "loss": 0.2616,
+      "step": 313
+    },
+    {
+      "epoch": 0.8328912466843501,
+      "grad_norm": 19.480645791210744,
+      "learning_rate": 9.075728448943783e-06,
+      "loss": 0.226,
+      "step": 314
+    },
+    {
+      "epoch": 0.8355437665782494,
+      "grad_norm": 30.320644718264788,
+      "learning_rate": 9.066762181421552e-06,
+      "loss": 0.2461,
+      "step": 315
+    },
+    {
+      "epoch": 0.8381962864721485,
+      "grad_norm": 34.8700120321434,
+      "learning_rate": 9.057757107238897e-06,
+      "loss": 0.3387,
+      "step": 316
+    },
+    {
+      "epoch": 0.8408488063660478,
+      "grad_norm": 52.98866951359855,
+      "learning_rate": 9.048713312325806e-06,
+      "loss": 0.372,
+      "step": 317
+    },
+    {
+      "epoch": 0.843501326259947,
+      "grad_norm": 33.12902942097088,
+      "learning_rate": 9.039630882981769e-06,
+      "loss": 0.2694,
+      "step": 318
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 49.80785762572212,
+      "learning_rate": 9.030509905874934e-06,
+      "loss": 0.4448,
+      "step": 319
+    },
+    {
+      "epoch": 0.8488063660477454,
+      "grad_norm": 48.26008386302785,
+      "learning_rate": 9.021350468041287e-06,
+      "loss": 0.4231,
+      "step": 320
+    },
+    {
+      "epoch": 0.8514588859416445,
+      "grad_norm": 47.68623752795164,
+      "learning_rate": 9.012152656883824e-06,
+      "loss": 0.2507,
+      "step": 321
+    },
+    {
+      "epoch": 0.8541114058355438,
+      "grad_norm": 29.364366147300288,
+      "learning_rate": 9.002916560171713e-06,
+      "loss": 0.3595,
+      "step": 322
+    },
+    {
+      "epoch": 0.8567639257294429,
+      "grad_norm": 20.836229758537232,
+      "learning_rate": 8.993642266039457e-06,
+      "loss": 0.2587,
+      "step": 323
+    },
+    {
+      "epoch": 0.8594164456233422,
+      "grad_norm": 42.67691753843682,
+      "learning_rate": 8.984329862986056e-06,
+      "loss": 0.4073,
+      "step": 324
+    },
+    {
+      "epoch": 0.8620689655172413,
+      "grad_norm": 27.185261724802686,
+      "learning_rate": 8.974979439874161e-06,
+      "loss": 0.3171,
+      "step": 325
+    },
+    {
+      "epoch": 0.8647214854111406,
+      "grad_norm": 19.41663669746958,
+      "learning_rate": 8.965591085929222e-06,
+      "loss": 0.2436,
+      "step": 326
+    },
+    {
+      "epoch": 0.8673740053050398,
+      "grad_norm": 24.66264338874257,
+      "learning_rate": 8.956164890738643e-06,
+      "loss": 0.1981,
+      "step": 327
+    },
+    {
+      "epoch": 0.870026525198939,
+      "grad_norm": 47.7679626658194,
+      "learning_rate": 8.946700944250925e-06,
+      "loss": 0.4011,
+      "step": 328
+    },
+    {
+      "epoch": 0.8726790450928382,
+      "grad_norm": 27.493035563524856,
+      "learning_rate": 8.937199336774805e-06,
+      "loss": 0.378,
+      "step": 329
+    },
+    {
+      "epoch": 0.8753315649867374,
+      "grad_norm": 33.915658193431625,
+      "learning_rate": 8.927660158978392e-06,
+      "loss": 0.2557,
+      "step": 330
+    },
+    {
+      "epoch": 0.8779840848806366,
+      "grad_norm": 34.75387246545008,
+      "learning_rate": 8.918083501888318e-06,
+      "loss": 0.2806,
+      "step": 331
+    },
+    {
+      "epoch": 0.8806366047745358,
+      "grad_norm": 47.864807509921775,
+      "learning_rate": 8.908469456888845e-06,
+      "loss": 0.2115,
+      "step": 332
+    },
+    {
+      "epoch": 0.883289124668435,
+      "grad_norm": 23.19477545972193,
+      "learning_rate": 8.898818115721009e-06,
+      "loss": 0.2153,
+      "step": 333
+    },
+    {
+      "epoch": 0.8859416445623343,
+      "grad_norm": 29.645775504694704,
+      "learning_rate": 8.889129570481742e-06,
+      "loss": 0.1996,
+      "step": 334
+    },
+    {
+      "epoch": 0.8885941644562334,
+      "grad_norm": 33.88979624484442,
+      "learning_rate": 8.879403913622996e-06,
+      "loss": 0.44,
+      "step": 335
+    },
+    {
+      "epoch": 0.8912466843501327,
+      "grad_norm": 108.03362140730333,
+      "learning_rate": 8.86964123795085e-06,
+      "loss": 0.315,
+      "step": 336
+    },
+    {
+      "epoch": 0.8938992042440318,
+      "grad_norm": 26.650695163554616,
+      "learning_rate": 8.859841636624632e-06,
+      "loss": 0.3892,
+      "step": 337
+    },
+    {
+      "epoch": 0.896551724137931,
+      "grad_norm": 99.55841306535598,
+      "learning_rate": 8.850005203156035e-06,
+      "loss": 0.3036,
+      "step": 338
+    },
+    {
+      "epoch": 0.8992042440318302,
+      "grad_norm": 48.33341336310134,
+      "learning_rate": 8.84013203140821e-06,
+      "loss": 0.2906,
+      "step": 339
+    },
+    {
+      "epoch": 0.9018567639257294,
+      "grad_norm": 32.637651201360114,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.2077,
+      "step": 340
+    },
+    {
+      "epoch": 0.9045092838196287,
+      "grad_norm": 25.534166507812365,
+      "learning_rate": 8.820275850279473e-06,
+      "loss": 0.1985,
+      "step": 341
+    },
+    {
+      "epoch": 0.9071618037135278,
+      "grad_norm": 19.29296435033772,
+      "learning_rate": 8.810293030374126e-06,
+      "loss": 0.15,
+      "step": 342
+    },
+    {
+      "epoch": 0.9098143236074271,
+      "grad_norm": 27.61534726493821,
+      "learning_rate": 8.800273851138882e-06,
+      "loss": 0.2247,
+      "step": 343
+    },
+    {
+      "epoch": 0.9124668435013262,
+      "grad_norm": 25.349121736508582,
+      "learning_rate": 8.790218408180736e-06,
+      "loss": 0.1389,
+      "step": 344
+    },
+    {
+      "epoch": 0.9151193633952255,
+      "grad_norm": 19.504694513734783,
+      "learning_rate": 8.780126797452713e-06,
+      "loss": 0.1762,
+      "step": 345
+    },
+    {
+      "epoch": 0.9177718832891246,
+      "grad_norm": 30.50787858362382,
+      "learning_rate": 8.769999115252976e-06,
+      "loss": 0.1956,
+      "step": 346
+    },
+    {
+      "epoch": 0.9204244031830239,
+      "grad_norm": 20.104267575195813,
+      "learning_rate": 8.759835458223889e-06,
+      "loss": 0.1143,
+      "step": 347
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 13.594728975157095,
+      "learning_rate": 8.749635923351108e-06,
+      "loss": 0.0834,
+      "step": 348
+    },
+    {
+      "epoch": 0.9257294429708223,
+      "grad_norm": 29.496882045130082,
+      "learning_rate": 8.739400607962644e-06,
+      "loss": 0.2157,
+      "step": 349
+    },
+    {
+      "epoch": 0.9283819628647215,
+      "grad_norm": 35.773188660484834,
+      "learning_rate": 8.729129609727948e-06,
+      "loss": 0.3571,
+      "step": 350
+    },
+    {
+      "epoch": 0.9310344827586207,
+      "grad_norm": 13.875590781519865,
+      "learning_rate": 8.71882302665696e-06,
+      "loss": 0.0751,
+      "step": 351
+    },
+    {
+      "epoch": 0.9336870026525199,
+      "grad_norm": 44.46404138029754,
+      "learning_rate": 8.708480957099195e-06,
+      "loss": 0.1722,
+      "step": 352
+    },
+    {
+      "epoch": 0.9363395225464191,
+      "grad_norm": 45.19632637637808,
+      "learning_rate": 8.698103499742785e-06,
+      "loss": 0.3458,
+      "step": 353
+    },
+    {
+      "epoch": 0.9389920424403183,
+      "grad_norm": 26.55265455350115,
+      "learning_rate": 8.687690753613554e-06,
+      "loss": 0.294,
+      "step": 354
+    },
+    {
+      "epoch": 0.9416445623342176,
+      "grad_norm": 28.758380748351815,
+      "learning_rate": 8.677242818074064e-06,
+      "loss": 0.167,
+      "step": 355
+    },
+    {
+      "epoch": 0.9442970822281167,
+      "grad_norm": 28.435166312141938,
+      "learning_rate": 8.666759792822662e-06,
+      "loss": 0.2465,
+      "step": 356
+    },
+    {
+      "epoch": 0.946949602122016,
+      "grad_norm": 43.70197021130316,
+      "learning_rate": 8.656241777892544e-06,
+      "loss": 0.2928,
+      "step": 357
+    },
+    {
+      "epoch": 0.9496021220159151,
+      "grad_norm": 32.384482978840225,
+      "learning_rate": 8.645688873650785e-06,
+      "loss": 0.23,
+      "step": 358
+    },
+    {
+      "epoch": 0.9522546419098143,
+      "grad_norm": 29.50639836832452,
+      "learning_rate": 8.635101180797391e-06,
+      "loss": 0.0831,
+      "step": 359
+    },
+    {
+      "epoch": 0.9549071618037135,
+      "grad_norm": 27.98334313472022,
+      "learning_rate": 8.624478800364332e-06,
+      "loss": 0.3275,
+      "step": 360
+    },
+    {
+      "epoch": 0.9575596816976127,
+      "grad_norm": 28.204464260730454,
+      "learning_rate": 8.613821833714584e-06,
+      "loss": 0.1508,
+      "step": 361
+    },
+    {
+      "epoch": 0.9602122015915119,
+      "grad_norm": 25.1275557521092,
+      "learning_rate": 8.603130382541156e-06,
+      "loss": 0.2585,
+      "step": 362
+    },
+    {
+      "epoch": 0.9628647214854111,
+      "grad_norm": 14.538485052093858,
+      "learning_rate": 8.592404548866123e-06,
+      "loss": 0.1698,
+      "step": 363
+    },
+    {
+      "epoch": 0.9655172413793104,
+      "grad_norm": 38.7193182526393,
+      "learning_rate": 8.581644435039652e-06,
+      "loss": 0.2151,
+      "step": 364
+    },
+    {
+      "epoch": 0.9681697612732095,
+      "grad_norm": 43.83895037233051,
+      "learning_rate": 8.570850143739022e-06,
+      "loss": 0.2332,
+      "step": 365
+    },
+    {
+      "epoch": 0.9708222811671088,
+      "grad_norm": 30.371526624539523,
+      "learning_rate": 8.56002177796765e-06,
+      "loss": 0.4292,
+      "step": 366
+    },
+    {
+      "epoch": 0.9734748010610079,
+      "grad_norm": 28.866647333264957,
+      "learning_rate": 8.549159441054105e-06,
+      "loss": 0.3052,
+      "step": 367
+    },
+    {
+      "epoch": 0.9761273209549072,
+      "grad_norm": 17.088822690787353,
+      "learning_rate": 8.538263236651119e-06,
+      "loss": 0.0924,
+      "step": 368
+    },
+    {
+      "epoch": 0.9787798408488063,
+      "grad_norm": 8.296167808657815,
+      "learning_rate": 8.527333268734607e-06,
+      "loss": 0.0425,
+      "step": 369
+    },
+    {
+      "epoch": 0.9814323607427056,
+      "grad_norm": 139.90331496389769,
+      "learning_rate": 8.516369641602662e-06,
+      "loss": 0.3744,
+      "step": 370
+    },
+    {
+      "epoch": 0.9840848806366048,
+      "grad_norm": 35.92284504373762,
+      "learning_rate": 8.505372459874572e-06,
+      "loss": 0.4011,
+      "step": 371
+    },
+    {
+      "epoch": 0.986737400530504,
+      "grad_norm": 42.55708188098206,
+      "learning_rate": 8.494341828489812e-06,
+      "loss": 0.3883,
+      "step": 372
+    },
+    {
+      "epoch": 0.9893899204244032,
+      "grad_norm": 26.194586839898157,
+      "learning_rate": 8.483277852707053e-06,
+      "loss": 0.2485,
+      "step": 373
+    },
+    {
+      "epoch": 0.9920424403183024,
+      "grad_norm": 17.83869232418558,
+      "learning_rate": 8.472180638103143e-06,
+      "loss": 0.1974,
+      "step": 374
+    },
+    {
+      "epoch": 0.9946949602122016,
+      "grad_norm": 34.735683009320994,
+      "learning_rate": 8.461050290572114e-06,
+      "loss": 0.13,
+      "step": 375
+    },
+    {
+      "epoch": 0.9973474801061007,
+      "grad_norm": 26.532678956391074,
+      "learning_rate": 8.449886916324168e-06,
+      "loss": 0.3265,
+      "step": 376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 77.79014811596227,
+      "learning_rate": 8.43869062188465e-06,
+      "loss": 0.3241,
+      "step": 377
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.5891069173812866,
+      "eval_runtime": 176.0707,
+      "eval_samples_per_second": 12.012,
+      "eval_steps_per_second": 1.505,
+      "step": 377
+    },
+    {
+      "epoch": 1.0026525198938991,
+      "grad_norm": 36.470611529759154,
+      "learning_rate": 8.427461514093056e-06,
+      "loss": 0.1684,
+      "step": 378
+    },
+    {
+      "epoch": 1.0053050397877985,
+      "grad_norm": 17.706888163127573,
+      "learning_rate": 8.41619970010199e-06,
+      "loss": 0.2765,
+      "step": 379
+    },
+    {
+      "epoch": 1.0079575596816976,
+      "grad_norm": 13.031570699168565,
+      "learning_rate": 8.404905287376158e-06,
+      "loss": 0.1101,
+      "step": 380
+    },
+    {
+      "epoch": 1.0106100795755968,
+      "grad_norm": 19.552662198077687,
+      "learning_rate": 8.39357838369133e-06,
+      "loss": 0.4447,
+      "step": 381
+    },
+    {
+      "epoch": 1.013262599469496,
+      "grad_norm": 17.815095847838503,
+      "learning_rate": 8.382219097133323e-06,
+      "loss": 0.2052,
+      "step": 382
+    },
+    {
+      "epoch": 1.0159151193633953,
+      "grad_norm": 12.505092725567371,
+      "learning_rate": 8.370827536096966e-06,
+      "loss": 0.0624,
+      "step": 383
+    },
+    {
+      "epoch": 1.0185676392572944,
+      "grad_norm": 11.478473259709562,
+      "learning_rate": 8.359403809285054e-06,
+      "loss": 0.1261,
+      "step": 384
+    },
+    {
+      "epoch": 1.0212201591511936,
+      "grad_norm": 8.521853145549802,
+      "learning_rate": 8.347948025707331e-06,
+      "loss": 0.0629,
+      "step": 385
+    },
+    {
+      "epoch": 1.023872679045093,
+      "grad_norm": 18.02070920605987,
+      "learning_rate": 8.336460294679431e-06,
+      "loss": 0.1441,
+      "step": 386
+    },
+    {
+      "epoch": 1.026525198938992,
+      "grad_norm": 17.00091095428731,
+      "learning_rate": 8.324940725821853e-06,
+      "loss": 0.1839,
+      "step": 387
+    },
+    {
+      "epoch": 1.0291777188328912,
+      "grad_norm": 7.130845145762619,
+      "learning_rate": 8.313389429058895e-06,
+      "loss": 0.0586,
+      "step": 388
+    },
+    {
+      "epoch": 1.0318302387267904,
+      "grad_norm": 9.12548545838751,
+      "learning_rate": 8.301806514617622e-06,
+      "loss": 0.1166,
+      "step": 389
+    },
+    {
+      "epoch": 1.0344827586206897,
+      "grad_norm": 36.64605899376915,
+      "learning_rate": 8.290192093026805e-06,
+      "loss": 0.2631,
+      "step": 390
+    },
+    {
+      "epoch": 1.0371352785145889,
+      "grad_norm": 30.0558772429006,
+      "learning_rate": 8.27854627511587e-06,
+      "loss": 0.1365,
+      "step": 391
+    },
+    {
+      "epoch": 1.039787798408488,
+      "grad_norm": 13.606408087389894,
+      "learning_rate": 8.266869172013835e-06,
+      "loss": 0.0601,
+      "step": 392
+    },
+    {
+      "epoch": 1.0424403183023874,
+      "grad_norm": 36.25256069833946,
+      "learning_rate": 8.255160895148263e-06,
+      "loss": 0.2423,
+      "step": 393
+    },
+    {
+      "epoch": 1.0450928381962865,
+      "grad_norm": 33.67722940279395,
+      "learning_rate": 8.243421556244179e-06,
+      "loss": 0.2654,
+      "step": 394
+    },
+    {
+      "epoch": 1.0477453580901857,
+      "grad_norm": 40.52545842308433,
+      "learning_rate": 8.23165126732302e-06,
+      "loss": 0.2066,
+      "step": 395
+    },
+    {
+      "epoch": 1.0503978779840848,
+      "grad_norm": 25.669838018256662,
+      "learning_rate": 8.219850140701557e-06,
+      "loss": 0.2277,
+      "step": 396
+    },
+    {
+      "epoch": 1.0530503978779842,
+      "grad_norm": 43.15769165903028,
+      "learning_rate": 8.208018288990832e-06,
+      "loss": 0.2268,
+      "step": 397
+    },
+    {
+      "epoch": 1.0557029177718833,
+      "grad_norm": 16.94140330545107,
+      "learning_rate": 8.196155825095073e-06,
+      "loss": 0.0841,
+      "step": 398
+    },
+    {
+      "epoch": 1.0583554376657824,
+      "grad_norm": 9.529747502579328,
+      "learning_rate": 8.184262862210624e-06,
+      "loss": 0.0631,
+      "step": 399
+    },
+    {
+      "epoch": 1.0610079575596818,
+      "grad_norm": 13.150162864905353,
+      "learning_rate": 8.172339513824863e-06,
+      "loss": 0.0842,
+      "step": 400
+    },
+    {
+      "epoch": 1.063660477453581,
+      "grad_norm": 11.45721931237483,
+      "learning_rate": 8.160385893715113e-06,
+      "loss": 0.1061,
+      "step": 401
+    },
+    {
+      "epoch": 1.06631299734748,
+      "grad_norm": 35.43358074136856,
+      "learning_rate": 8.14840211594757e-06,
+      "loss": 0.2425,
+      "step": 402
+    },
+    {
+      "epoch": 1.0689655172413792,
+      "grad_norm": 27.428682315233864,
+      "learning_rate": 8.136388294876204e-06,
+      "loss": 0.1924,
+      "step": 403
+    },
+    {
+      "epoch": 1.0716180371352786,
+      "grad_norm": 8.171222022680151,
+      "learning_rate": 8.124344545141663e-06,
+      "loss": 0.1284,
+      "step": 404
+    },
+    {
+      "epoch": 1.0742705570291777,
+      "grad_norm": 39.06536575690631,
+      "learning_rate": 8.112270981670196e-06,
+      "loss": 0.232,
+      "step": 405
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 21.32963011576174,
+      "learning_rate": 8.10016771967254e-06,
+      "loss": 0.1055,
+      "step": 406
+    },
+    {
+      "epoch": 1.079575596816976,
+      "grad_norm": 27.22475799075306,
+      "learning_rate": 8.088034874642834e-06,
+      "loss": 0.1489,
+      "step": 407
+    },
+    {
+      "epoch": 1.0822281167108754,
+      "grad_norm": 14.3882842008674,
+      "learning_rate": 8.075872562357502e-06,
+      "loss": 0.0829,
+      "step": 408
+    },
+    {
+      "epoch": 1.0848806366047745,
+      "grad_norm": 58.56202943213565,
+      "learning_rate": 8.063680898874158e-06,
+      "loss": 0.3128,
+      "step": 409
+    },
+    {
+      "epoch": 1.0875331564986737,
+      "grad_norm": 68.2761310554643,
+      "learning_rate": 8.051460000530501e-06,
+      "loss": 0.1267,
+      "step": 410
+    },
+    {
+      "epoch": 1.090185676392573,
+      "grad_norm": 43.810216329803836,
+      "learning_rate": 8.039209983943201e-06,
+      "loss": 0.2736,
+      "step": 411
+    },
+    {
+      "epoch": 1.0928381962864722,
+      "grad_norm": 60.97125301465749,
+      "learning_rate": 8.026930966006778e-06,
+      "loss": 0.1616,
+      "step": 412
+    },
+    {
+      "epoch": 1.0954907161803713,
+      "grad_norm": 45.630973756286146,
+      "learning_rate": 8.014623063892504e-06,
+      "loss": 0.2629,
+      "step": 413
+    },
+    {
+      "epoch": 1.0981432360742707,
+      "grad_norm": 3.457495328921145,
+      "learning_rate": 8.002286395047267e-06,
+      "loss": 0.0236,
+      "step": 414
+    },
+    {
+      "epoch": 1.1007957559681698,
+      "grad_norm": 62.65821806193905,
+      "learning_rate": 7.989921077192464e-06,
+      "loss": 0.4217,
+      "step": 415
+    },
+    {
+      "epoch": 1.103448275862069,
+      "grad_norm": 22.66117796728126,
+      "learning_rate": 7.97752722832287e-06,
+      "loss": 0.1689,
+      "step": 416
+    },
+    {
+      "epoch": 1.106100795755968,
+      "grad_norm": 11.75462762962119,
+      "learning_rate": 7.965104966705518e-06,
+      "loss": 0.0972,
+      "step": 417
+    },
+    {
+      "epoch": 1.1087533156498675,
+      "grad_norm": 41.913827864061176,
+      "learning_rate": 7.95265441087856e-06,
+      "loss": 0.2382,
+      "step": 418
+    },
+    {
+      "epoch": 1.1114058355437666,
+      "grad_norm": 47.459409982780066,
+      "learning_rate": 7.940175679650145e-06,
+      "loss": 0.4864,
+      "step": 419
+    },
+    {
+      "epoch": 1.1140583554376657,
+      "grad_norm": 19.367649648717457,
+      "learning_rate": 7.927668892097288e-06,
+      "loss": 0.0742,
+      "step": 420
+    },
+    {
+      "epoch": 1.1167108753315649,
+      "grad_norm": 34.18374747743834,
+      "learning_rate": 7.915134167564724e-06,
+      "loss": 0.1962,
+      "step": 421
+    },
+    {
+      "epoch": 1.1193633952254642,
+      "grad_norm": 47.150789672019954,
+      "learning_rate": 7.902571625663773e-06,
+      "loss": 0.2737,
+      "step": 422
+    },
+    {
+      "epoch": 1.1220159151193634,
+      "grad_norm": 51.86857593447088,
+      "learning_rate": 7.889981386271202e-06,
+      "loss": 0.0715,
+      "step": 423
+    },
+    {
+      "epoch": 1.1246684350132625,
+      "grad_norm": 37.26299672015612,
+      "learning_rate": 7.877363569528076e-06,
+      "loss": 0.2618,
+      "step": 424
+    },
+    {
+      "epoch": 1.1273209549071619,
+      "grad_norm": 55.28707481699974,
+      "learning_rate": 7.864718295838615e-06,
+      "loss": 0.1249,
+      "step": 425
+    },
+    {
+      "epoch": 1.129973474801061,
+      "grad_norm": 84.16566896954009,
+      "learning_rate": 7.852045685869046e-06,
+      "loss": 0.4398,
+      "step": 426
+    },
+    {
+      "epoch": 1.1326259946949602,
+      "grad_norm": 55.8899032446642,
+      "learning_rate": 7.839345860546448e-06,
+      "loss": 0.3044,
+      "step": 427
+    },
+    {
+      "epoch": 1.1352785145888595,
+      "grad_norm": 10.659192422147509,
+      "learning_rate": 7.826618941057597e-06,
+      "loss": 0.0611,
+      "step": 428
+    },
+    {
+      "epoch": 1.1379310344827587,
+      "grad_norm": 42.57054449009807,
+      "learning_rate": 7.81386504884782e-06,
+      "loss": 0.1664,
+      "step": 429
+    },
+    {
+      "epoch": 1.1405835543766578,
+      "grad_norm": 27.750046981174442,
+      "learning_rate": 7.80108430561982e-06,
+      "loss": 0.1654,
+      "step": 430
+    },
+    {
+      "epoch": 1.143236074270557,
+      "grad_norm": 24.9396201345219,
+      "learning_rate": 7.788276833332527e-06,
+      "loss": 0.1293,
+      "step": 431
+    },
+    {
+      "epoch": 1.1458885941644563,
+      "grad_norm": 20.34868157498092,
+      "learning_rate": 7.775442754199929e-06,
+      "loss": 0.1705,
+      "step": 432
+    },
+    {
+      "epoch": 1.1485411140583555,
+      "grad_norm": 58.399543578541476,
+      "learning_rate": 7.762582190689912e-06,
+      "loss": 0.156,
+      "step": 433
+    },
+    {
+      "epoch": 1.1511936339522546,
+      "grad_norm": 25.459151955023966,
+      "learning_rate": 7.749695265523076e-06,
+      "loss": 0.2423,
+      "step": 434
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 23.22794672451964,
+      "learning_rate": 7.736782101671587e-06,
+      "loss": 0.1002,
+      "step": 435
+    },
+    {
+      "epoch": 1.156498673740053,
+      "grad_norm": 19.872411993502894,
+      "learning_rate": 7.723842822357982e-06,
+      "loss": 0.078,
+      "step": 436
+    },
+    {
+      "epoch": 1.1591511936339522,
+      "grad_norm": 6.868162120751368,
+      "learning_rate": 7.710877551054004e-06,
+      "loss": 0.0298,
+      "step": 437
+    },
+    {
+      "epoch": 1.1618037135278514,
+      "grad_norm": 26.292541068653883,
+      "learning_rate": 7.697886411479422e-06,
+      "loss": 0.1934,
+      "step": 438
+    },
+    {
+      "epoch": 1.1644562334217508,
+      "grad_norm": 21.43672508582472,
+      "learning_rate": 7.684869527600856e-06,
+      "loss": 0.1535,
+      "step": 439
+    },
+    {
+      "epoch": 1.16710875331565,
+      "grad_norm": 16.62769259399857,
+      "learning_rate": 7.67182702363058e-06,
+      "loss": 0.1433,
+      "step": 440
+    },
+    {
+      "epoch": 1.169761273209549,
+      "grad_norm": 16.685896603459977,
+      "learning_rate": 7.658759024025349e-06,
+      "loss": 0.0708,
+      "step": 441
+    },
+    {
+      "epoch": 1.1724137931034484,
+      "grad_norm": 26.005011847487346,
+      "learning_rate": 7.645665653485205e-06,
+      "loss": 0.199,
+      "step": 442
+    },
+    {
+      "epoch": 1.1750663129973475,
+      "grad_norm": 5.716604464026939,
+      "learning_rate": 7.632547036952296e-06,
+      "loss": 0.0217,
+      "step": 443
+    },
+    {
+      "epoch": 1.1777188328912467,
+      "grad_norm": 36.897628289818144,
+      "learning_rate": 7.6194032996096685e-06,
+      "loss": 0.2061,
+      "step": 444
+    },
+    {
+      "epoch": 1.1803713527851458,
+      "grad_norm": 28.876741113162478,
+      "learning_rate": 7.606234566880089e-06,
+      "loss": 0.13,
+      "step": 445
+    },
+    {
+      "epoch": 1.1830238726790452,
+      "grad_norm": 19.080592510468772,
+      "learning_rate": 7.593040964424836e-06,
+      "loss": 0.1685,
+      "step": 446
+    },
+    {
+      "epoch": 1.1856763925729443,
+      "grad_norm": 29.96186333981723,
+      "learning_rate": 7.579822618142505e-06,
+      "loss": 0.1566,
+      "step": 447
+    },
+    {
+      "epoch": 1.1883289124668435,
+      "grad_norm": 32.8034743799736,
+      "learning_rate": 7.5665796541678106e-06,
+      "loss": 0.1605,
+      "step": 448
+    },
+    {
+      "epoch": 1.1909814323607426,
+      "grad_norm": 27.661953801989608,
+      "learning_rate": 7.553312198870373e-06,
+      "loss": 0.0956,
+      "step": 449
+    },
+    {
+      "epoch": 1.193633952254642,
+      "grad_norm": 12.306782289979477,
+      "learning_rate": 7.540020378853523e-06,
+      "loss": 0.0777,
+      "step": 450
+    },
+    {
+      "epoch": 1.1962864721485411,
+      "grad_norm": 16.700054834521364,
+      "learning_rate": 7.526704320953091e-06,
+      "loss": 0.2292,
+      "step": 451
+    },
+    {
+      "epoch": 1.1989389920424403,
+      "grad_norm": 18.34021244065456,
+      "learning_rate": 7.513364152236185e-06,
+      "loss": 0.1332,
+      "step": 452
+    },
+    {
+      "epoch": 1.2015915119363396,
+      "grad_norm": 17.67473447987808,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.0685,
+      "step": 453
+    },
+    {
+      "epoch": 1.2042440318302388,
+      "grad_norm": 18.160624034169683,
+      "learning_rate": 7.486611991770586e-06,
+      "loss": 0.0866,
+      "step": 454
+    },
+    {
+      "epoch": 1.206896551724138,
+      "grad_norm": 30.183878085653877,
+      "learning_rate": 7.473200255301635e-06,
+      "loss": 0.1514,
+      "step": 455
+    },
+    {
+      "epoch": 1.209549071618037,
+      "grad_norm": 11.560796164676075,
+      "learning_rate": 7.459764918573264e-06,
+      "loss": 0.0702,
+      "step": 456
+    },
+    {
+      "epoch": 1.2122015915119364,
+      "grad_norm": 110.73822732726704,
+      "learning_rate": 7.446306109790798e-06,
+      "loss": 0.0581,
+      "step": 457
+    },
+    {
+      "epoch": 1.2148541114058355,
+      "grad_norm": 87.5160478741724,
+      "learning_rate": 7.432823957383533e-06,
+      "loss": 0.2227,
+      "step": 458
+    },
+    {
+      "epoch": 1.2175066312997347,
+      "grad_norm": 60.30264082510386,
+      "learning_rate": 7.419318590003524e-06,
+      "loss": 0.1127,
+      "step": 459
+    },
+    {
+      "epoch": 1.2201591511936338,
+      "grad_norm": 21.619284863772585,
+      "learning_rate": 7.405790136524353e-06,
+      "loss": 0.0821,
+      "step": 460
+    },
+    {
+      "epoch": 1.2228116710875332,
+      "grad_norm": 56.712783723095264,
+      "learning_rate": 7.392238726039897e-06,
+      "loss": 0.1583,
+      "step": 461
+    },
+    {
+      "epoch": 1.2254641909814323,
+      "grad_norm": 24.170384444203012,
+      "learning_rate": 7.3786644878631035e-06,
+      "loss": 0.1189,
+      "step": 462
+    },
+    {
+      "epoch": 1.2281167108753315,
+      "grad_norm": 10.562962951144847,
+      "learning_rate": 7.365067551524739e-06,
+      "loss": 0.0449,
+      "step": 463
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 99.528184771969,
+      "learning_rate": 7.3514480467721786e-06,
+      "loss": 0.0788,
+      "step": 464
+    },
+    {
+      "epoch": 1.23342175066313,
+      "grad_norm": 12.34676892311528,
+      "learning_rate": 7.3378061035681415e-06,
+      "loss": 0.0469,
+      "step": 465
+    },
+    {
+      "epoch": 1.2360742705570291,
+      "grad_norm": 68.2832497273848,
+      "learning_rate": 7.324141852089473e-06,
+      "loss": 0.1851,
+      "step": 466
+    },
+    {
+      "epoch": 1.2387267904509285,
+      "grad_norm": 11.108520534758835,
+      "learning_rate": 7.3104554227258895e-06,
+      "loss": 0.0882,
+      "step": 467
+    },
+    {
+      "epoch": 1.2413793103448276,
+      "grad_norm": 42.060698163435674,
+      "learning_rate": 7.296746946078737e-06,
+      "loss": 0.2728,
+      "step": 468
+    },
+    {
+      "epoch": 1.2440318302387268,
+      "grad_norm": 17.56387098625625,
+      "learning_rate": 7.283016552959745e-06,
+      "loss": 0.0696,
+      "step": 469
+    },
+    {
+      "epoch": 1.246684350132626,
+      "grad_norm": 30.23208035695815,
+      "learning_rate": 7.269264374389781e-06,
+      "loss": 0.1132,
+      "step": 470
+    },
+    {
+      "epoch": 1.2493368700265253,
+      "grad_norm": 26.412011305182677,
+      "learning_rate": 7.255490541597594e-06,
+      "loss": 0.1062,
+      "step": 471
+    },
+    {
+      "epoch": 1.2519893899204244,
+      "grad_norm": 42.983237948628,
+      "learning_rate": 7.2416951860185735e-06,
+      "loss": 0.1112,
+      "step": 472
+    },
+    {
+      "epoch": 1.2546419098143236,
+      "grad_norm": 5.851594403240968,
+      "learning_rate": 7.2278784392934775e-06,
+      "loss": 0.0579,
+      "step": 473
+    },
+    {
+      "epoch": 1.2572944297082227,
+      "grad_norm": 42.07578121579949,
+      "learning_rate": 7.2140404332671986e-06,
+      "loss": 0.0693,
+      "step": 474
+    },
+    {
+      "epoch": 1.259946949602122,
+      "grad_norm": 46.7174671935105,
+      "learning_rate": 7.200181299987483e-06,
+      "loss": 0.1177,
+      "step": 475
+    },
+    {
+      "epoch": 1.2625994694960212,
+      "grad_norm": 27.379743656894938,
+      "learning_rate": 7.186301171703689e-06,
+      "loss": 0.1629,
+      "step": 476
+    },
+    {
+      "epoch": 1.2652519893899203,
+      "grad_norm": 30.216270759458997,
+      "learning_rate": 7.172400180865514e-06,
+      "loss": 0.0605,
+      "step": 477
+    },
+    {
+      "epoch": 1.2679045092838197,
+      "grad_norm": 20.806360193025363,
+      "learning_rate": 7.158478460121735e-06,
+      "loss": 0.0519,
+      "step": 478
+    },
+    {
+      "epoch": 1.2705570291777188,
+      "grad_norm": 10.651496274071377,
+      "learning_rate": 7.144536142318945e-06,
+      "loss": 0.052,
+      "step": 479
+    },
+    {
+      "epoch": 1.273209549071618,
+      "grad_norm": 23.494134794931725,
+      "learning_rate": 7.130573360500277e-06,
+      "loss": 0.1685,
+      "step": 480
+    },
+    {
+      "epoch": 1.2758620689655173,
+      "grad_norm": 24.791552622223954,
+      "learning_rate": 7.116590247904144e-06,
+      "loss": 0.0613,
+      "step": 481
+    },
+    {
+      "epoch": 1.2785145888594165,
+      "grad_norm": 6.621282957067961,
+      "learning_rate": 7.102586937962961e-06,
+      "loss": 0.0403,
+      "step": 482
+    },
+    {
+      "epoch": 1.2811671087533156,
+      "grad_norm": 23.984601170750345,
+      "learning_rate": 7.088563564301874e-06,
+      "loss": 0.07,
+      "step": 483
+    },
+    {
+      "epoch": 1.2838196286472148,
+      "grad_norm": 9.929247414366557,
+      "learning_rate": 7.074520260737487e-06,
+      "loss": 0.0464,
+      "step": 484
+    },
+    {
+      "epoch": 1.2864721485411141,
+      "grad_norm": 12.867645506911478,
+      "learning_rate": 7.060457161276581e-06,
+      "loss": 0.0985,
+      "step": 485
+    },
+    {
+      "epoch": 1.2891246684350133,
+      "grad_norm": 7.624514156491183,
+      "learning_rate": 7.046374400114842e-06,
+      "loss": 0.0249,
+      "step": 486
+    },
+    {
+      "epoch": 1.2917771883289124,
+      "grad_norm": 30.32207231686849,
+      "learning_rate": 7.032272111635565e-06,
+      "loss": 0.1042,
+      "step": 487
+    },
+    {
+      "epoch": 1.2944297082228116,
+      "grad_norm": 57.229453179332594,
+      "learning_rate": 7.018150430408394e-06,
+      "loss": 0.4131,
+      "step": 488
+    },
+    {
+      "epoch": 1.297082228116711,
+      "grad_norm": 17.842806157709017,
+      "learning_rate": 7.004009491188023e-06,
+      "loss": 0.0452,
+      "step": 489
+    },
+    {
+      "epoch": 1.29973474801061,
+      "grad_norm": 21.44800192709656,
+      "learning_rate": 6.989849428912908e-06,
+      "loss": 0.037,
+      "step": 490
+    },
+    {
+      "epoch": 1.3023872679045092,
+      "grad_norm": 34.16602355273998,
+      "learning_rate": 6.975670378703993e-06,
+      "loss": 0.2023,
+      "step": 491
+    },
+    {
+      "epoch": 1.3050397877984086,
+      "grad_norm": 26.85872386445106,
+      "learning_rate": 6.961472475863406e-06,
+      "loss": 0.1479,
+      "step": 492
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 27.914655387462602,
+      "learning_rate": 6.947255855873176e-06,
+      "loss": 0.1027,
+      "step": 493
+    },
+    {
+      "epoch": 1.3103448275862069,
+      "grad_norm": 8.47195867510813,
+      "learning_rate": 6.933020654393941e-06,
+      "loss": 0.0626,
+      "step": 494
+    },
+    {
+      "epoch": 1.3129973474801062,
+      "grad_norm": 15.786375616446254,
+      "learning_rate": 6.918767007263646e-06,
+      "loss": 0.1095,
+      "step": 495
+    },
+    {
+      "epoch": 1.3156498673740054,
+      "grad_norm": 14.966235305898794,
+      "learning_rate": 6.904495050496258e-06,
+      "loss": 0.086,
+      "step": 496
+    },
+    {
+      "epoch": 1.3183023872679045,
+      "grad_norm": 8.971458193639767,
+      "learning_rate": 6.8902049202804574e-06,
+      "loss": 0.0561,
+      "step": 497
+    },
+    {
+      "epoch": 1.3209549071618036,
+      "grad_norm": 17.554889727273423,
+      "learning_rate": 6.875896752978345e-06,
+      "loss": 0.0503,
+      "step": 498
+    },
+    {
+      "epoch": 1.323607427055703,
+      "grad_norm": 15.147440419341597,
+      "learning_rate": 6.861570685124135e-06,
+      "loss": 0.0494,
+      "step": 499
+    },
+    {
+      "epoch": 1.3262599469496021,
+      "grad_norm": 27.20170910120345,
+      "learning_rate": 6.847226853422863e-06,
+      "loss": 0.0252,
+      "step": 500
+    },
+    {
+      "epoch": 1.3289124668435013,
+      "grad_norm": 22.553693373205064,
+      "learning_rate": 6.832865394749065e-06,
+      "loss": 0.0806,
+      "step": 501
+    },
+    {
+      "epoch": 1.3315649867374004,
+      "grad_norm": 44.035317078004645,
+      "learning_rate": 6.8184864461454866e-06,
+      "loss": 0.0409,
+      "step": 502
+    },
+    {
+      "epoch": 1.3342175066312998,
+      "grad_norm": 45.74768341442066,
+      "learning_rate": 6.804090144821772e-06,
+      "loss": 0.1278,
+      "step": 503
+    },
+    {
+      "epoch": 1.336870026525199,
+      "grad_norm": 63.72486362138629,
+      "learning_rate": 6.7896766281531435e-06,
+      "loss": 0.3504,
+      "step": 504
+    },
+    {
+      "epoch": 1.339522546419098,
+      "grad_norm": 14.659860419565552,
+      "learning_rate": 6.775246033679105e-06,
+      "loss": 0.1378,
+      "step": 505
+    },
+    {
+      "epoch": 1.3421750663129974,
+      "grad_norm": 15.094010123566344,
+      "learning_rate": 6.760798499102121e-06,
+      "loss": 0.0443,
+      "step": 506
+    },
+    {
+      "epoch": 1.3448275862068966,
+      "grad_norm": 12.088460229768618,
+      "learning_rate": 6.7463341622863074e-06,
+      "loss": 0.0737,
+      "step": 507
+    },
+    {
+      "epoch": 1.3474801061007957,
+      "grad_norm": 7.743245120977951,
+      "learning_rate": 6.7318531612561145e-06,
+      "loss": 0.0345,
+      "step": 508
+    },
+    {
+      "epoch": 1.350132625994695,
+      "grad_norm": 28.851196219475575,
+      "learning_rate": 6.717355634195004e-06,
+      "loss": 0.1229,
+      "step": 509
+    },
+    {
+      "epoch": 1.3527851458885942,
+      "grad_norm": 39.8940905052454,
+      "learning_rate": 6.702841719444141e-06,
+      "loss": 0.1008,
+      "step": 510
+    },
+    {
+      "epoch": 1.3554376657824934,
+      "grad_norm": 45.13925991061865,
+      "learning_rate": 6.688311555501064e-06,
+      "loss": 0.269,
+      "step": 511
+    },
+    {
+      "epoch": 1.3580901856763925,
+      "grad_norm": 16.570082019867765,
+      "learning_rate": 6.673765281018373e-06,
+      "loss": 0.1264,
+      "step": 512
+    },
+    {
+      "epoch": 1.3607427055702916,
+      "grad_norm": 9.936276076523223,
+      "learning_rate": 6.659203034802397e-06,
+      "loss": 0.0442,
+      "step": 513
+    },
+    {
+      "epoch": 1.363395225464191,
+      "grad_norm": 491.03239020588376,
+      "learning_rate": 6.644624955811873e-06,
+      "loss": 0.2389,
+      "step": 514
+    },
+    {
+      "epoch": 1.3660477453580901,
+      "grad_norm": 15.199004520494972,
+      "learning_rate": 6.630031183156628e-06,
+      "loss": 0.0862,
+      "step": 515
+    },
+    {
+      "epoch": 1.3687002652519893,
+      "grad_norm": 11.166070002874784,
+      "learning_rate": 6.615421856096231e-06,
+      "loss": 0.0523,
+      "step": 516
+    },
+    {
+      "epoch": 1.3713527851458887,
+      "grad_norm": 21.807652878549746,
+      "learning_rate": 6.6007971140386915e-06,
+      "loss": 0.1845,
+      "step": 517
+    },
+    {
+      "epoch": 1.3740053050397878,
+      "grad_norm": 9.085575610594937,
+      "learning_rate": 6.586157096539105e-06,
+      "loss": 0.0187,
+      "step": 518
+    },
+    {
+      "epoch": 1.376657824933687,
+      "grad_norm": 7.831782583878153,
+      "learning_rate": 6.571501943298335e-06,
+      "loss": 0.0404,
+      "step": 519
+    },
+    {
+      "epoch": 1.3793103448275863,
+      "grad_norm": 9.70973898788606,
+      "learning_rate": 6.556831794161678e-06,
+      "loss": 0.0448,
+      "step": 520
+    },
+    {
+      "epoch": 1.3819628647214854,
+      "grad_norm": 68.24554091396725,
+      "learning_rate": 6.542146789117524e-06,
+      "loss": 0.1233,
+      "step": 521
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 10.500883042336119,
+      "learning_rate": 6.527447068296026e-06,
+      "loss": 0.039,
+      "step": 522
+    },
+    {
+      "epoch": 1.387267904509284,
+      "grad_norm": 25.24639428355656,
+      "learning_rate": 6.512732771967758e-06,
+      "loss": 0.0529,
+      "step": 523
+    },
+    {
+      "epoch": 1.389920424403183,
+      "grad_norm": 3.132482062228408,
+      "learning_rate": 6.498004040542385e-06,
+      "loss": 0.0161,
+      "step": 524
+    },
+    {
+      "epoch": 1.3925729442970822,
+      "grad_norm": 38.50562742710699,
+      "learning_rate": 6.483261014567311e-06,
+      "loss": 0.0607,
+      "step": 525
+    },
+    {
+      "epoch": 1.3952254641909814,
+      "grad_norm": 13.901422622952236,
+      "learning_rate": 6.4685038347263495e-06,
+      "loss": 0.046,
+      "step": 526
+    },
+    {
+      "epoch": 1.3978779840848805,
+      "grad_norm": 38.43373398443968,
+      "learning_rate": 6.453732641838372e-06,
+      "loss": 0.1998,
+      "step": 527
+    },
+    {
+      "epoch": 1.4005305039787799,
+      "grad_norm": 28.880004159481086,
+      "learning_rate": 6.4389475768559675e-06,
+      "loss": 0.0993,
+      "step": 528
+    },
+    {
+      "epoch": 1.403183023872679,
+      "grad_norm": 41.90220673902449,
+      "learning_rate": 6.4241487808641044e-06,
+      "loss": 0.0416,
+      "step": 529
+    },
+    {
+      "epoch": 1.4058355437665782,
+      "grad_norm": 7.0622226239396735,
+      "learning_rate": 6.409336395078771e-06,
+      "loss": 0.0223,
+      "step": 530
+    },
+    {
+      "epoch": 1.4084880636604775,
+      "grad_norm": 37.18705225070335,
+      "learning_rate": 6.394510560845637e-06,
+      "loss": 0.1225,
+      "step": 531
+    },
+    {
+      "epoch": 1.4111405835543767,
+      "grad_norm": 12.461676647059019,
+      "learning_rate": 6.379671419638703e-06,
+      "loss": 0.0449,
+      "step": 532
+    },
+    {
+      "epoch": 1.4137931034482758,
+      "grad_norm": 25.69067972611454,
+      "learning_rate": 6.3648191130589524e-06,
+      "loss": 0.1461,
+      "step": 533
+    },
+    {
+      "epoch": 1.4164456233421752,
+      "grad_norm": 4.433735117691165,
+      "learning_rate": 6.349953782832991e-06,
+      "loss": 0.0163,
+      "step": 534
+    },
+    {
+      "epoch": 1.4190981432360743,
+      "grad_norm": 35.5915224640076,
+      "learning_rate": 6.335075570811708e-06,
+      "loss": 0.0421,
+      "step": 535
+    },
+    {
+      "epoch": 1.4217506631299734,
+      "grad_norm": 12.883444328072557,
+      "learning_rate": 6.320184618968915e-06,
+      "loss": 0.0533,
+      "step": 536
+    },
+    {
+      "epoch": 1.4244031830238728,
+      "grad_norm": 24.55048357158165,
+      "learning_rate": 6.305281069399989e-06,
+      "loss": 0.0583,
+      "step": 537
+    },
+    {
+      "epoch": 1.427055702917772,
+      "grad_norm": 33.72346507980671,
+      "learning_rate": 6.290365064320521e-06,
+      "loss": 0.2696,
+      "step": 538
+    },
+    {
+      "epoch": 1.429708222811671,
+      "grad_norm": 106.39331911829048,
+      "learning_rate": 6.275436746064957e-06,
+      "loss": 0.301,
+      "step": 539
+    },
+    {
+      "epoch": 1.4323607427055702,
+      "grad_norm": 15.979137251612,
+      "learning_rate": 6.26049625708524e-06,
+      "loss": 0.0331,
+      "step": 540
+    },
+    {
+      "epoch": 1.4350132625994694,
+      "grad_norm": 28.11729203908229,
+      "learning_rate": 6.245543739949455e-06,
+      "loss": 0.2345,
+      "step": 541
+    },
+    {
+      "epoch": 1.4376657824933687,
+      "grad_norm": 13.077394458293497,
+      "learning_rate": 6.2305793373404564e-06,
+      "loss": 0.0358,
+      "step": 542
+    },
+    {
+      "epoch": 1.4403183023872679,
+      "grad_norm": 5.554488097306693,
+      "learning_rate": 6.215603192054523e-06,
+      "loss": 0.0144,
+      "step": 543
+    },
+    {
+      "epoch": 1.442970822281167,
+      "grad_norm": 3.1244872050868495,
+      "learning_rate": 6.2006154469999824e-06,
+      "loss": 0.0097,
+      "step": 544
+    },
+    {
+      "epoch": 1.4456233421750664,
+      "grad_norm": 8.489049690971624,
+      "learning_rate": 6.185616245195849e-06,
+      "loss": 0.0272,
+      "step": 545
+    },
+    {
+      "epoch": 1.4482758620689655,
+      "grad_norm": 39.45660661182102,
+      "learning_rate": 6.17060572977047e-06,
+      "loss": 0.0626,
+      "step": 546
+    },
+    {
+      "epoch": 1.4509283819628647,
+      "grad_norm": 9.639199657748263,
+      "learning_rate": 6.155584043960145e-06,
+      "loss": 0.1411,
+      "step": 547
+    },
+    {
+      "epoch": 1.453580901856764,
+      "grad_norm": 27.845194387729055,
+      "learning_rate": 6.140551331107767e-06,
+      "loss": 0.0311,
+      "step": 548
+    },
+    {
+      "epoch": 1.4562334217506632,
+      "grad_norm": 43.69139166324222,
+      "learning_rate": 6.125507734661458e-06,
+      "loss": 0.0922,
+      "step": 549
+    },
+    {
+      "epoch": 1.4588859416445623,
+      "grad_norm": 39.59031227302006,
+      "learning_rate": 6.110453398173188e-06,
+      "loss": 0.0969,
+      "step": 550
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 46.21688165038216,
+      "learning_rate": 6.095388465297418e-06,
+      "loss": 0.0548,
+      "step": 551
+    },
+    {
+      "epoch": 1.4641909814323608,
+      "grad_norm": 25.52563185215739,
+      "learning_rate": 6.080313079789723e-06,
+      "loss": 0.0626,
+      "step": 552
+    },
+    {
+      "epoch": 1.46684350132626,
+      "grad_norm": 15.010437110226805,
+      "learning_rate": 6.0652273855054225e-06,
+      "loss": 0.0799,
+      "step": 553
+    },
+    {
+      "epoch": 1.469496021220159,
+      "grad_norm": 0.6240187668769939,
+      "learning_rate": 6.050131526398202e-06,
+      "loss": 0.0071,
+      "step": 554
+    },
+    {
+      "epoch": 1.4721485411140582,
+      "grad_norm": 24.41151697104724,
+      "learning_rate": 6.035025646518747e-06,
+      "loss": 0.0566,
+      "step": 555
+    },
+    {
+      "epoch": 1.4748010610079576,
+      "grad_norm": 43.88869277127279,
+      "learning_rate": 6.019909890013367e-06,
+      "loss": 0.1289,
+      "step": 556
+    },
+    {
+      "epoch": 1.4774535809018567,
+      "grad_norm": 1.8126838533443241,
+      "learning_rate": 6.004784401122613e-06,
+      "loss": 0.0123,
+      "step": 557
+    },
+    {
+      "epoch": 1.4801061007957559,
+      "grad_norm": 50.60535466560564,
+      "learning_rate": 5.9896493241799115e-06,
+      "loss": 0.2817,
+      "step": 558
+    },
+    {
+      "epoch": 1.4827586206896552,
+      "grad_norm": 31.769450904461493,
+      "learning_rate": 5.974504803610178e-06,
+      "loss": 0.038,
+      "step": 559
+    },
+    {
+      "epoch": 1.4854111405835544,
+      "grad_norm": 36.41160490344729,
+      "learning_rate": 5.959350983928446e-06,
+      "loss": 0.2033,
+      "step": 560
+    },
+    {
+      "epoch": 1.4880636604774535,
+      "grad_norm": 1.0570484873390567,
+      "learning_rate": 5.944188009738483e-06,
+      "loss": 0.007,
+      "step": 561
+    },
+    {
+      "epoch": 1.490716180371353,
+      "grad_norm": 7.600018984649344,
+      "learning_rate": 5.929016025731413e-06,
+      "loss": 0.0117,
+      "step": 562
+    },
+    {
+      "epoch": 1.493368700265252,
+      "grad_norm": 46.11036080527606,
+      "learning_rate": 5.913835176684335e-06,
+      "loss": 0.0896,
+      "step": 563
+    },
+    {
+      "epoch": 1.4960212201591512,
+      "grad_norm": 6.801838928839265,
+      "learning_rate": 5.898645607458941e-06,
+      "loss": 0.018,
+      "step": 564
+    },
+    {
+      "epoch": 1.4986737400530503,
+      "grad_norm": 33.05399188153241,
+      "learning_rate": 5.883447463000136e-06,
+      "loss": 0.0144,
+      "step": 565
+    },
+    {
+      "epoch": 1.5013262599469495,
+      "grad_norm": 23.00632593695278,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.0361,
+      "step": 566
+    },
+    {
+      "epoch": 1.5039787798408488,
+      "grad_norm": 3.5405959675882896,
+      "learning_rate": 5.8530260285696674e-06,
+      "loss": 0.0108,
+      "step": 567
+    },
+    {
+      "epoch": 1.506631299734748,
+      "grad_norm": 11.97736850552611,
+      "learning_rate": 5.837803028891418e-06,
+      "loss": 0.0239,
+      "step": 568
+    },
+    {
+      "epoch": 1.509283819628647,
+      "grad_norm": 6.472410570187345,
+      "learning_rate": 5.822572034563812e-06,
+      "loss": 0.0167,
+      "step": 569
+    },
+    {
+      "epoch": 1.5119363395225465,
+      "grad_norm": 4.641359282270338,
+      "learning_rate": 5.807333190927054e-06,
+      "loss": 0.0081,
+      "step": 570
+    },
+    {
+      "epoch": 1.5145888594164456,
+      "grad_norm": 119.51677620720801,
+      "learning_rate": 5.792086643396238e-06,
+      "loss": 0.0296,
+      "step": 571
+    },
+    {
+      "epoch": 1.5172413793103448,
+      "grad_norm": 19.160987194114814,
+      "learning_rate": 5.776832537459983e-06,
+      "loss": 0.2339,
+      "step": 572
+    },
+    {
+      "epoch": 1.5198938992042441,
+      "grad_norm": 64.96051480014378,
+      "learning_rate": 5.761571018679025e-06,
+      "loss": 0.2978,
+      "step": 573
+    },
+    {
+      "epoch": 1.5225464190981433,
+      "grad_norm": 22.520538143186805,
+      "learning_rate": 5.746302232684843e-06,
+      "loss": 0.0431,
+      "step": 574
+    },
+    {
+      "epoch": 1.5251989389920424,
+      "grad_norm": 8.49094605957141,
+      "learning_rate": 5.731026325178255e-06,
+      "loss": 0.0211,
+      "step": 575
+    },
+    {
+      "epoch": 1.5278514588859418,
+      "grad_norm": 45.70005511753745,
+      "learning_rate": 5.715743441928041e-06,
+      "loss": 0.098,
+      "step": 576
+    },
+    {
+      "epoch": 1.530503978779841,
+      "grad_norm": 14.782978428535925,
+      "learning_rate": 5.700453728769545e-06,
+      "loss": 0.0165,
+      "step": 577
+    },
+    {
+      "epoch": 1.53315649867374,
+      "grad_norm": 13.046864605038222,
+      "learning_rate": 5.6851573316032845e-06,
+      "loss": 0.0241,
+      "step": 578
+    },
+    {
+      "epoch": 1.5358090185676394,
+      "grad_norm": 64.59802552950447,
+      "learning_rate": 5.669854396393559e-06,
+      "loss": 0.0738,
+      "step": 579
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 166.49480105447577,
+      "learning_rate": 5.654545069167056e-06,
+      "loss": 0.1385,
+      "step": 580
+    },
+    {
+      "epoch": 1.5411140583554377,
+      "grad_norm": 17.008871192463243,
+      "learning_rate": 5.639229496011456e-06,
+      "loss": 0.0833,
+      "step": 581
+    },
+    {
+      "epoch": 1.5437665782493368,
+      "grad_norm": 1.8492882406831308,
+      "learning_rate": 5.623907823074044e-06,
+      "loss": 0.0073,
+      "step": 582
+    },
+    {
+      "epoch": 1.546419098143236,
+      "grad_norm": 30.430767357234902,
+      "learning_rate": 5.60858019656031e-06,
+      "loss": 0.0909,
+      "step": 583
+    },
+    {
+      "epoch": 1.5490716180371353,
+      "grad_norm": 13.266355913015387,
+      "learning_rate": 5.593246762732558e-06,
+      "loss": 0.0483,
+      "step": 584
+    },
+    {
+      "epoch": 1.5517241379310345,
+      "grad_norm": 14.567204624660986,
+      "learning_rate": 5.577907667908505e-06,
+      "loss": 0.0256,
+      "step": 585
+    },
+    {
+      "epoch": 1.5543766578249336,
+      "grad_norm": 34.19788288920179,
+      "learning_rate": 5.562563058459884e-06,
+      "loss": 0.1541,
+      "step": 586
+    },
+    {
+      "epoch": 1.557029177718833,
+      "grad_norm": 54.14509214775431,
+      "learning_rate": 5.5472130808110595e-06,
+      "loss": 0.1065,
+      "step": 587
+    },
+    {
+      "epoch": 1.5596816976127321,
+      "grad_norm": 28.62720507192516,
+      "learning_rate": 5.531857881437612e-06,
+      "loss": 0.0893,
+      "step": 588
+    },
+    {
+      "epoch": 1.5623342175066313,
+      "grad_norm": 28.054311802068526,
+      "learning_rate": 5.516497606864959e-06,
+      "loss": 0.0565,
+      "step": 589
+    },
+    {
+      "epoch": 1.5649867374005306,
+      "grad_norm": 31.379538762961293,
+      "learning_rate": 5.50113240366694e-06,
+      "loss": 0.1069,
+      "step": 590
+    },
+    {
+      "epoch": 1.5676392572944295,
+      "grad_norm": 10.37843513767325,
+      "learning_rate": 5.48576241846443e-06,
+      "loss": 0.0262,
+      "step": 591
+    },
+    {
+      "epoch": 1.570291777188329,
+      "grad_norm": 15.483387797663624,
+      "learning_rate": 5.470387797923934e-06,
+      "loss": 0.0173,
+      "step": 592
+    },
+    {
+      "epoch": 1.5729442970822283,
+      "grad_norm": 15.29920880392035,
+      "learning_rate": 5.4550086887561874e-06,
+      "loss": 0.0319,
+      "step": 593
+    },
+    {
+      "epoch": 1.5755968169761272,
+      "grad_norm": 32.908373060607694,
+      "learning_rate": 5.4396252377147615e-06,
+      "loss": 0.0486,
+      "step": 594
+    },
+    {
+      "epoch": 1.5782493368700266,
+      "grad_norm": 4.944210731487496,
+      "learning_rate": 5.424237591594658e-06,
+      "loss": 0.0122,
+      "step": 595
+    },
+    {
+      "epoch": 1.5809018567639257,
+      "grad_norm": 14.2873911899618,
+      "learning_rate": 5.4088458972309085e-06,
+      "loss": 0.029,
+      "step": 596
+    },
+    {
+      "epoch": 1.5835543766578248,
+      "grad_norm": 19.811484884375027,
+      "learning_rate": 5.39345030149718e-06,
+      "loss": 0.0395,
+      "step": 597
+    },
+    {
+      "epoch": 1.5862068965517242,
+      "grad_norm": 25.622053063552784,
+      "learning_rate": 5.378050951304356e-06,
+      "loss": 0.0648,
+      "step": 598
+    },
+    {
+      "epoch": 1.5888594164456233,
+      "grad_norm": 17.43025816114162,
+      "learning_rate": 5.362647993599159e-06,
+      "loss": 0.0193,
+      "step": 599
+    },
+    {
+      "epoch": 1.5915119363395225,
+      "grad_norm": 5.31676256926054,
+      "learning_rate": 5.347241575362729e-06,
+      "loss": 0.0124,
+      "step": 600
+    },
+    {
+      "epoch": 1.5941644562334218,
+      "grad_norm": 1.1389079557738029,
+      "learning_rate": 5.3318318436092335e-06,
+      "loss": 0.0069,
+      "step": 601
+    },
+    {
+      "epoch": 1.596816976127321,
+      "grad_norm": 10.627284322705297,
+      "learning_rate": 5.31641894538445e-06,
+      "loss": 0.0151,
+      "step": 602
+    },
+    {
+      "epoch": 1.5994694960212201,
+      "grad_norm": 98.99497329825951,
+      "learning_rate": 5.30100302776438e-06,
+      "loss": 0.0967,
+      "step": 603
+    },
+    {
+      "epoch": 1.6021220159151195,
+      "grad_norm": 24.27158846293466,
+      "learning_rate": 5.285584237853832e-06,
+      "loss": 0.0324,
+      "step": 604
+    },
+    {
+      "epoch": 1.6047745358090184,
+      "grad_norm": 28.948767908864784,
+      "learning_rate": 5.270162722785026e-06,
+      "loss": 0.033,
+      "step": 605
+    },
+    {
+      "epoch": 1.6074270557029178,
+      "grad_norm": 26.33013336955424,
+      "learning_rate": 5.254738629716186e-06,
+      "loss": 0.1195,
+      "step": 606
+    },
+    {
+      "epoch": 1.6100795755968171,
+      "grad_norm": 36.50352169617263,
+      "learning_rate": 5.239312105830135e-06,
+      "loss": 0.0208,
+      "step": 607
+    },
+    {
+      "epoch": 1.612732095490716,
+      "grad_norm": 19.431136725192903,
+      "learning_rate": 5.223883298332894e-06,
+      "loss": 0.0259,
+      "step": 608
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 34.3112525202407,
+      "learning_rate": 5.208452354452275e-06,
+      "loss": 0.0651,
+      "step": 609
+    },
+    {
+      "epoch": 1.6180371352785146,
+      "grad_norm": 37.8345226955255,
+      "learning_rate": 5.19301942143647e-06,
+      "loss": 0.0347,
+      "step": 610
+    },
+    {
+      "epoch": 1.6206896551724137,
+      "grad_norm": 68.3527326674225,
+      "learning_rate": 5.1775846465526625e-06,
+      "loss": 0.1516,
+      "step": 611
+    },
+    {
+      "epoch": 1.623342175066313,
+      "grad_norm": 66.62188237442813,
+      "learning_rate": 5.162148177085604e-06,
+      "loss": 0.1933,
+      "step": 612
+    },
+    {
+      "epoch": 1.6259946949602122,
+      "grad_norm": 65.94874176325419,
+      "learning_rate": 5.14671016033622e-06,
+      "loss": 0.0999,
+      "step": 613
+    },
+    {
+      "epoch": 1.6286472148541113,
+      "grad_norm": 4.319664540685107,
+      "learning_rate": 5.131270743620201e-06,
+      "loss": 0.0104,
+      "step": 614
+    },
+    {
+      "epoch": 1.6312997347480107,
+      "grad_norm": 5.488574522745108,
+      "learning_rate": 5.115830074266592e-06,
+      "loss": 0.0123,
+      "step": 615
+    },
+    {
+      "epoch": 1.6339522546419099,
+      "grad_norm": 47.41871483146368,
+      "learning_rate": 5.100388299616395e-06,
+      "loss": 0.0305,
+      "step": 616
+    },
+    {
+      "epoch": 1.636604774535809,
+      "grad_norm": 5.2412860000731225,
+      "learning_rate": 5.084945567021159e-06,
+      "loss": 0.0118,
+      "step": 617
+    },
+    {
+      "epoch": 1.6392572944297084,
+      "grad_norm": 20.43067020592513,
+      "learning_rate": 5.069502023841576e-06,
+      "loss": 0.0136,
+      "step": 618
+    },
+    {
+      "epoch": 1.6419098143236073,
+      "grad_norm": 55.13620630926507,
+      "learning_rate": 5.054057817446067e-06,
+      "loss": 0.027,
+      "step": 619
+    },
+    {
+      "epoch": 1.6445623342175066,
+      "grad_norm": 24.012783309520152,
+      "learning_rate": 5.038613095209392e-06,
+      "loss": 0.0297,
+      "step": 620
+    },
+    {
+      "epoch": 1.647214854111406,
+      "grad_norm": 246.41322304845076,
+      "learning_rate": 5.0231680045112174e-06,
+      "loss": 0.0122,
+      "step": 621
+    },
+    {
+      "epoch": 1.649867374005305,
+      "grad_norm": 1.908263943570197,
+      "learning_rate": 5.00772269273474e-06,
+      "loss": 0.0055,
+      "step": 622
+    },
+    {
+      "epoch": 1.6525198938992043,
+      "grad_norm": 7.9188343410709,
+      "learning_rate": 4.9922773072652615e-06,
+      "loss": 0.0118,
+      "step": 623
+    },
+    {
+      "epoch": 1.6551724137931034,
+      "grad_norm": 3.9827474382276122,
+      "learning_rate": 4.976831995488784e-06,
+      "loss": 0.007,
+      "step": 624
+    },
+    {
+      "epoch": 1.6578249336870026,
+      "grad_norm": 35.85065017259431,
+      "learning_rate": 4.961386904790611e-06,
+      "loss": 0.0291,
+      "step": 625
+    },
+    {
+      "epoch": 1.660477453580902,
+      "grad_norm": 1.8833465914635488,
+      "learning_rate": 4.945942182553932e-06,
+      "loss": 0.0061,
+      "step": 626
+    },
+    {
+      "epoch": 1.663129973474801,
+      "grad_norm": 2.0782864974444752,
+      "learning_rate": 4.9304979761584256e-06,
+      "loss": 0.0072,
+      "step": 627
+    },
+    {
+      "epoch": 1.6657824933687002,
+      "grad_norm": 0.9498564062083344,
+      "learning_rate": 4.915054432978842e-06,
+      "loss": 0.0054,
+      "step": 628
+    },
+    {
+      "epoch": 1.6684350132625996,
+      "grad_norm": 39.36865575587075,
+      "learning_rate": 4.899611700383608e-06,
+      "loss": 0.0337,
+      "step": 629
+    },
+    {
+      "epoch": 1.6710875331564987,
+      "grad_norm": 44.4769789219546,
+      "learning_rate": 4.884169925733409e-06,
+      "loss": 0.1049,
+      "step": 630
+    },
+    {
+      "epoch": 1.6737400530503979,
+      "grad_norm": 3.166712761063399,
+      "learning_rate": 4.868729256379802e-06,
+      "loss": 0.0059,
+      "step": 631
+    },
+    {
+      "epoch": 1.6763925729442972,
+      "grad_norm": 0.5167268015197257,
+      "learning_rate": 4.8532898396637815e-06,
+      "loss": 0.0045,
+      "step": 632
+    },
+    {
+      "epoch": 1.6790450928381961,
+      "grad_norm": 0.4315587098083175,
+      "learning_rate": 4.837851822914397e-06,
+      "loss": 0.0045,
+      "step": 633
+    },
+    {
+      "epoch": 1.6816976127320955,
+      "grad_norm": 34.489146105146105,
+      "learning_rate": 4.822415353447339e-06,
+      "loss": 0.1604,
+      "step": 634
+    },
+    {
+      "epoch": 1.6843501326259946,
+      "grad_norm": 0.347847431409771,
+      "learning_rate": 4.80698057856353e-06,
+      "loss": 0.0041,
+      "step": 635
+    },
+    {
+      "epoch": 1.6870026525198938,
+      "grad_norm": 0.4165775687885535,
+      "learning_rate": 4.791547645547727e-06,
+      "loss": 0.0041,
+      "step": 636
+    },
+    {
+      "epoch": 1.6896551724137931,
+      "grad_norm": 117.93594590550879,
+      "learning_rate": 4.7761167016671064e-06,
+      "loss": 0.1003,
+      "step": 637
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 66.349687103629,
+      "learning_rate": 4.760687894169867e-06,
+      "loss": 0.1888,
+      "step": 638
+    },
+    {
+      "epoch": 1.6949602122015914,
+      "grad_norm": 2.5640604180704947,
+      "learning_rate": 4.7452613702838166e-06,
+      "loss": 0.0056,
+      "step": 639
+    },
+    {
+      "epoch": 1.6976127320954908,
+      "grad_norm": 243.72962362462474,
+      "learning_rate": 4.729837277214975e-06,
+      "loss": 0.0926,
+      "step": 640
+    },
+    {
+      "epoch": 1.70026525198939,
+      "grad_norm": 7.350167948874671,
+      "learning_rate": 4.7144157621461694e-06,
+      "loss": 0.0095,
+      "step": 641
+    },
+    {
+      "epoch": 1.702917771883289,
+      "grad_norm": 0.6606315834117484,
+      "learning_rate": 4.698996972235622e-06,
+      "loss": 0.0043,
+      "step": 642
+    },
+    {
+      "epoch": 1.7055702917771884,
+      "grad_norm": 6.7420487073193005,
+      "learning_rate": 4.683581054615551e-06,
+      "loss": 0.0095,
+      "step": 643
+    },
+    {
+      "epoch": 1.7082228116710876,
+      "grad_norm": 5.880007837879827,
+      "learning_rate": 4.668168156390769e-06,
+      "loss": 0.0061,
+      "step": 644
+    },
+    {
+      "epoch": 1.7108753315649867,
+      "grad_norm": 5.405586207286194,
+      "learning_rate": 4.652758424637271e-06,
+      "loss": 0.0074,
+      "step": 645
+    },
+    {
+      "epoch": 1.713527851458886,
+      "grad_norm": 13.642408752250786,
+      "learning_rate": 4.637352006400842e-06,
+      "loss": 0.015,
+      "step": 646
+    },
+    {
+      "epoch": 1.716180371352785,
+      "grad_norm": 21.078816092264656,
+      "learning_rate": 4.621949048695646e-06,
+      "loss": 0.0198,
+      "step": 647
+    },
+    {
+      "epoch": 1.7188328912466844,
+      "grad_norm": 53.552715766878414,
+      "learning_rate": 4.606549698502824e-06,
+      "loss": 0.0687,
+      "step": 648
+    },
+    {
+      "epoch": 1.7214854111405835,
+      "grad_norm": 65.69163142454755,
+      "learning_rate": 4.5911541027690914e-06,
+      "loss": 0.1032,
+      "step": 649
+    },
+    {
+      "epoch": 1.7241379310344827,
+      "grad_norm": 6.4577164762694785,
+      "learning_rate": 4.575762408405343e-06,
+      "loss": 0.0054,
+      "step": 650
+    },
+    {
+      "epoch": 1.726790450928382,
+      "grad_norm": 22.799875878261123,
+      "learning_rate": 4.56037476228524e-06,
+      "loss": 0.051,
+      "step": 651
+    },
+    {
+      "epoch": 1.7294429708222812,
+      "grad_norm": 9.888082550044638,
+      "learning_rate": 4.544991311243815e-06,
+      "loss": 0.0087,
+      "step": 652
+    },
+    {
+      "epoch": 1.7320954907161803,
+      "grad_norm": 30.204004753927336,
+      "learning_rate": 4.529612202076069e-06,
+      "loss": 0.193,
+      "step": 653
+    },
+    {
+      "epoch": 1.7347480106100797,
+      "grad_norm": 0.20191470860720032,
+      "learning_rate": 4.514237581535571e-06,
+      "loss": 0.0033,
+      "step": 654
+    },
+    {
+      "epoch": 1.7374005305039788,
+      "grad_norm": 5.33579782789315,
+      "learning_rate": 4.498867596333061e-06,
+      "loss": 0.0055,
+      "step": 655
+    },
+    {
+      "epoch": 1.740053050397878,
+      "grad_norm": 0.4949743432127943,
+      "learning_rate": 4.4835023931350425e-06,
+      "loss": 0.0037,
+      "step": 656
+    },
+    {
+      "epoch": 1.7427055702917773,
+      "grad_norm": 26.415507427778962,
+      "learning_rate": 4.468142118562389e-06,
+      "loss": 0.0243,
+      "step": 657
+    },
+    {
+      "epoch": 1.7453580901856764,
+      "grad_norm": 23.92557305475512,
+      "learning_rate": 4.452786919188943e-06,
+      "loss": 0.0094,
+      "step": 658
+    },
+    {
+      "epoch": 1.7480106100795756,
+      "grad_norm": 1.34416749285188,
+      "learning_rate": 4.437436941540116e-06,
+      "loss": 0.0042,
+      "step": 659
+    },
+    {
+      "epoch": 1.750663129973475,
+      "grad_norm": 35.23422292881152,
+      "learning_rate": 4.422092332091497e-06,
+      "loss": 0.1418,
+      "step": 660
+    },
+    {
+      "epoch": 1.7533156498673739,
+      "grad_norm": 21.82973318934849,
+      "learning_rate": 4.4067532372674434e-06,
+      "loss": 0.0118,
+      "step": 661
+    },
+    {
+      "epoch": 1.7559681697612732,
+      "grad_norm": 10.25652325742471,
+      "learning_rate": 4.391419803439691e-06,
+      "loss": 0.0098,
+      "step": 662
+    },
+    {
+      "epoch": 1.7586206896551724,
+      "grad_norm": 1.0092521511671273,
+      "learning_rate": 4.3760921769259585e-06,
+      "loss": 0.0042,
+      "step": 663
+    },
+    {
+      "epoch": 1.7612732095490715,
+      "grad_norm": 0.18022257900606575,
+      "learning_rate": 4.360770503988545e-06,
+      "loss": 0.0032,
+      "step": 664
+    },
+    {
+      "epoch": 1.7639257294429709,
+      "grad_norm": 0.18362886938020564,
+      "learning_rate": 4.345454930832946e-06,
+      "loss": 0.0032,
+      "step": 665
+    },
+    {
+      "epoch": 1.76657824933687,
+      "grad_norm": 52.59486973274669,
+      "learning_rate": 4.3301456036064415e-06,
+      "loss": 0.0188,
+      "step": 666
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.17663417348622706,
+      "learning_rate": 4.314842668396716e-06,
+      "loss": 0.003,
+      "step": 667
+    },
+    {
+      "epoch": 1.7718832891246685,
+      "grad_norm": 34.55861746298538,
+      "learning_rate": 4.299546271230457e-06,
+      "loss": 0.1371,
+      "step": 668
+    },
+    {
+      "epoch": 1.7745358090185677,
+      "grad_norm": 3.00901581132572,
+      "learning_rate": 4.28425655807196e-06,
+      "loss": 0.0058,
+      "step": 669
+    },
+    {
+      "epoch": 1.7771883289124668,
+      "grad_norm": 1.472166462932808,
+      "learning_rate": 4.268973674821747e-06,
+      "loss": 0.0036,
+      "step": 670
+    },
+    {
+      "epoch": 1.7798408488063662,
+      "grad_norm": 0.30438323731414085,
+      "learning_rate": 4.2536977673151594e-06,
+      "loss": 0.0029,
+      "step": 671
+    },
+    {
+      "epoch": 1.782493368700265,
+      "grad_norm": 0.8798922827717847,
+      "learning_rate": 4.2384289813209754e-06,
+      "loss": 0.0034,
+      "step": 672
+    },
+    {
+      "epoch": 1.7851458885941645,
+      "grad_norm": 2.200407200049607,
+      "learning_rate": 4.223167462540018e-06,
+      "loss": 0.0038,
+      "step": 673
+    },
+    {
+      "epoch": 1.7877984084880638,
+      "grad_norm": 0.8436364418418659,
+      "learning_rate": 4.207913356603762e-06,
+      "loss": 0.0031,
+      "step": 674
+    },
+    {
+      "epoch": 1.7904509283819627,
+      "grad_norm": 0.34783958925230496,
+      "learning_rate": 4.192666809072948e-06,
+      "loss": 0.0029,
+      "step": 675
+    },
+    {
+      "epoch": 1.793103448275862,
+      "grad_norm": 8.706721085918481,
+      "learning_rate": 4.1774279654361895e-06,
+      "loss": 0.0093,
+      "step": 676
+    },
+    {
+      "epoch": 1.7957559681697612,
+      "grad_norm": 46.6268725543766,
+      "learning_rate": 4.162196971108584e-06,
+      "loss": 0.0428,
+      "step": 677
+    },
+    {
+      "epoch": 1.7984084880636604,
+      "grad_norm": 21.44766513368348,
+      "learning_rate": 4.146973971430333e-06,
+      "loss": 0.0105,
+      "step": 678
+    },
+    {
+      "epoch": 1.8010610079575597,
+      "grad_norm": 17.24328013019805,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.1565,
+      "step": 679
+    },
+    {
+      "epoch": 1.8037135278514589,
+      "grad_norm": 0.1531046540956738,
+      "learning_rate": 4.116552536999865e-06,
+      "loss": 0.0027,
+      "step": 680
+    },
+    {
+      "epoch": 1.806366047745358,
+      "grad_norm": 0.34748754074866783,
+      "learning_rate": 4.101354392541061e-06,
+      "loss": 0.0028,
+      "step": 681
+    },
+    {
+      "epoch": 1.8090185676392574,
+      "grad_norm": 6.766920577207019,
+      "learning_rate": 4.086164823315667e-06,
+      "loss": 0.004,
+      "step": 682
+    },
+    {
+      "epoch": 1.8116710875331565,
+      "grad_norm": 44.46819293351116,
+      "learning_rate": 4.070983974268588e-06,
+      "loss": 0.031,
+      "step": 683
+    },
+    {
+      "epoch": 1.8143236074270557,
+      "grad_norm": 11.603635061420663,
+      "learning_rate": 4.055811990261518e-06,
+      "loss": 0.0099,
+      "step": 684
+    },
+    {
+      "epoch": 1.816976127320955,
+      "grad_norm": 0.4777433788159791,
+      "learning_rate": 4.040649016071555e-06,
+      "loss": 0.0029,
+      "step": 685
+    },
+    {
+      "epoch": 1.819628647214854,
+      "grad_norm": 40.475341750138206,
+      "learning_rate": 4.025495196389824e-06,
+      "loss": 0.0787,
+      "step": 686
+    },
+    {
+      "epoch": 1.8222811671087533,
+      "grad_norm": 0.2714511227984912,
+      "learning_rate": 4.010350675820091e-06,
+      "loss": 0.0027,
+      "step": 687
+    },
+    {
+      "epoch": 1.8249336870026527,
+      "grad_norm": 0.5834117283731054,
+      "learning_rate": 3.9952155988773876e-06,
+      "loss": 0.0031,
+      "step": 688
+    },
+    {
+      "epoch": 1.8275862068965516,
+      "grad_norm": 0.6449148052226085,
+      "learning_rate": 3.980090109986634e-06,
+      "loss": 0.0029,
+      "step": 689
+    },
+    {
+      "epoch": 1.830238726790451,
+      "grad_norm": 17.141551314961788,
+      "learning_rate": 3.964974353481254e-06,
+      "loss": 0.0299,
+      "step": 690
+    },
+    {
+      "epoch": 1.83289124668435,
+      "grad_norm": 25.433319764821093,
+      "learning_rate": 3.949868473601801e-06,
+      "loss": 0.0427,
+      "step": 691
+    },
+    {
+      "epoch": 1.8355437665782492,
+      "grad_norm": 0.6583579876602191,
+      "learning_rate": 3.934772614494581e-06,
+      "loss": 0.0032,
+      "step": 692
+    },
+    {
+      "epoch": 1.8381962864721486,
+      "grad_norm": 5.5504711429288545,
+      "learning_rate": 3.9196869202102775e-06,
+      "loss": 0.0052,
+      "step": 693
+    },
+    {
+      "epoch": 1.8408488063660478,
+      "grad_norm": 0.2578200579248058,
+      "learning_rate": 3.904611534702583e-06,
+      "loss": 0.0025,
+      "step": 694
+    },
+    {
+      "epoch": 1.843501326259947,
+      "grad_norm": 4.66905475011612,
+      "learning_rate": 3.889546601826813e-06,
+      "loss": 0.005,
+      "step": 695
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 50.6553420715786,
+      "learning_rate": 3.874492265338544e-06,
+      "loss": 0.0401,
+      "step": 696
+    },
+    {
+      "epoch": 1.8488063660477454,
+      "grad_norm": 0.155879357954373,
+      "learning_rate": 3.859448668892233e-06,
+      "loss": 0.0024,
+      "step": 697
+    },
+    {
+      "epoch": 1.8514588859416445,
+      "grad_norm": 5.118868015888596,
+      "learning_rate": 3.844415956039856e-06,
+      "loss": 0.0031,
+      "step": 698
+    },
+    {
+      "epoch": 1.854111405835544,
+      "grad_norm": 0.1458255118431307,
+      "learning_rate": 3.829394270229531e-06,
+      "loss": 0.0024,
+      "step": 699
+    },
+    {
+      "epoch": 1.8567639257294428,
+      "grad_norm": 0.15503449645311218,
+      "learning_rate": 3.814383754804152e-06,
+      "loss": 0.0024,
+      "step": 700
+    },
+    {
+      "epoch": 1.8594164456233422,
+      "grad_norm": 13.511687384914936,
+      "learning_rate": 3.79938455300002e-06,
+      "loss": 0.0537,
+      "step": 701
+    },
+    {
+      "epoch": 1.8620689655172413,
+      "grad_norm": 37.51104666572117,
+      "learning_rate": 3.7843968079454773e-06,
+      "loss": 0.0821,
+      "step": 702
+    },
+    {
+      "epoch": 1.8647214854111405,
+      "grad_norm": 0.13563308608550503,
+      "learning_rate": 3.7694206626595444e-06,
+      "loss": 0.0023,
+      "step": 703
+    },
+    {
+      "epoch": 1.8673740053050398,
+      "grad_norm": 0.18848861798425623,
+      "learning_rate": 3.7544562600505475e-06,
+      "loss": 0.0023,
+      "step": 704
+    },
+    {
+      "epoch": 1.870026525198939,
+      "grad_norm": 0.21091030708084085,
+      "learning_rate": 3.7395037429147615e-06,
+      "loss": 0.0023,
+      "step": 705
+    },
+    {
+      "epoch": 1.8726790450928381,
+      "grad_norm": 15.378659773256478,
+      "learning_rate": 3.724563253935045e-06,
+      "loss": 0.0182,
+      "step": 706
+    },
+    {
+      "epoch": 1.8753315649867375,
+      "grad_norm": 75.53306694832403,
+      "learning_rate": 3.7096349356794803e-06,
+      "loss": 0.1587,
+      "step": 707
+    },
+    {
+      "epoch": 1.8779840848806366,
+      "grad_norm": 125.70443954838983,
+      "learning_rate": 3.694718930600012e-06,
+      "loss": 0.1481,
+      "step": 708
+    },
+    {
+      "epoch": 1.8806366047745358,
+      "grad_norm": 0.1516937714071399,
+      "learning_rate": 3.6798153810310854e-06,
+      "loss": 0.0023,
+      "step": 709
+    },
+    {
+      "epoch": 1.8832891246684351,
+      "grad_norm": 106.63520750539608,
+      "learning_rate": 3.6649244291882923e-06,
+      "loss": 0.0725,
+      "step": 710
+    },
+    {
+      "epoch": 1.8859416445623343,
+      "grad_norm": 16.272773685544387,
+      "learning_rate": 3.6500462171670104e-06,
+      "loss": 0.0155,
+      "step": 711
+    },
+    {
+      "epoch": 1.8885941644562334,
+      "grad_norm": 2.0319296678156427,
+      "learning_rate": 3.6351808869410484e-06,
+      "loss": 0.0028,
+      "step": 712
+    },
+    {
+      "epoch": 1.8912466843501328,
+      "grad_norm": 173.5411500891954,
+      "learning_rate": 3.6203285803612975e-06,
+      "loss": 0.0279,
+      "step": 713
+    },
+    {
+      "epoch": 1.8938992042440317,
+      "grad_norm": 2.234030900991913,
+      "learning_rate": 3.605489439154365e-06,
+      "loss": 0.0045,
+      "step": 714
+    },
+    {
+      "epoch": 1.896551724137931,
+      "grad_norm": 29.878514877041937,
+      "learning_rate": 3.5906636049212316e-06,
+      "loss": 0.0518,
+      "step": 715
+    },
+    {
+      "epoch": 1.8992042440318302,
+      "grad_norm": 23.612513450372212,
+      "learning_rate": 3.575851219135898e-06,
+      "loss": 0.0225,
+      "step": 716
+    },
+    {
+      "epoch": 1.9018567639257293,
+      "grad_norm": 13.40219828568539,
+      "learning_rate": 3.5610524231440324e-06,
+      "loss": 0.0051,
+      "step": 717
+    },
+    {
+      "epoch": 1.9045092838196287,
+      "grad_norm": 0.7617183910694865,
+      "learning_rate": 3.5462673581616298e-06,
+      "loss": 0.0027,
+      "step": 718
+    },
+    {
+      "epoch": 1.9071618037135278,
+      "grad_norm": 76.01576611393702,
+      "learning_rate": 3.5314961652736517e-06,
+      "loss": 0.054,
+      "step": 719
+    },
+    {
+      "epoch": 1.909814323607427,
+      "grad_norm": 4.686784862497009,
+      "learning_rate": 3.5167389854326907e-06,
+      "loss": 0.0032,
+      "step": 720
+    },
+    {
+      "epoch": 1.9124668435013263,
+      "grad_norm": 3.3635617826020194,
+      "learning_rate": 3.501995959457616e-06,
+      "loss": 0.0033,
+      "step": 721
+    },
+    {
+      "epoch": 1.9151193633952255,
+      "grad_norm": 0.14187419174072188,
+      "learning_rate": 3.487267228032242e-06,
+      "loss": 0.0021,
+      "step": 722
+    },
+    {
+      "epoch": 1.9177718832891246,
+      "grad_norm": 22.366611259381504,
+      "learning_rate": 3.472552931703975e-06,
+      "loss": 0.0104,
+      "step": 723
+    },
+    {
+      "epoch": 1.920424403183024,
+      "grad_norm": 0.14420645288299855,
+      "learning_rate": 3.457853210882477e-06,
+      "loss": 0.0021,
+      "step": 724
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 23.81159747484969,
+      "learning_rate": 3.443168205838323e-06,
+      "loss": 0.0144,
+      "step": 725
+    },
+    {
+      "epoch": 1.9257294429708223,
+      "grad_norm": 3.577888638195551,
+      "learning_rate": 3.428498056701665e-06,
+      "loss": 0.0071,
+      "step": 726
+    },
+    {
+      "epoch": 1.9283819628647216,
+      "grad_norm": 0.3401435550413983,
+      "learning_rate": 3.413842903460896e-06,
+      "loss": 0.0024,
+      "step": 727
+    },
+    {
+      "epoch": 1.9310344827586206,
+      "grad_norm": 0.1764668896414946,
+      "learning_rate": 3.39920288596131e-06,
+      "loss": 0.0022,
+      "step": 728
+    },
+    {
+      "epoch": 1.93368700265252,
+      "grad_norm": 0.11599443588337781,
+      "learning_rate": 3.3845781439037695e-06,
+      "loss": 0.002,
+      "step": 729
+    },
+    {
+      "epoch": 1.936339522546419,
+      "grad_norm": 0.26998040037052706,
+      "learning_rate": 3.369968816843375e-06,
+      "loss": 0.0021,
+      "step": 730
+    },
+    {
+      "epoch": 1.9389920424403182,
+      "grad_norm": 0.11246222321274224,
+      "learning_rate": 3.3553750441881266e-06,
+      "loss": 0.002,
+      "step": 731
+    },
+    {
+      "epoch": 1.9416445623342176,
+      "grad_norm": 14.149523916959522,
+      "learning_rate": 3.3407969651976045e-06,
+      "loss": 0.0107,
+      "step": 732
+    },
+    {
+      "epoch": 1.9442970822281167,
+      "grad_norm": 0.11957540686765213,
+      "learning_rate": 3.326234718981628e-06,
+      "loss": 0.002,
+      "step": 733
+    },
+    {
+      "epoch": 1.9469496021220158,
+      "grad_norm": 12.058904025993744,
+      "learning_rate": 3.311688444498937e-06,
+      "loss": 0.0124,
+      "step": 734
+    },
+    {
+      "epoch": 1.9496021220159152,
+      "grad_norm": 13.396218874482619,
+      "learning_rate": 3.2971582805558622e-06,
+      "loss": 0.0081,
+      "step": 735
+    },
+    {
+      "epoch": 1.9522546419098143,
+      "grad_norm": 29.29268777424113,
+      "learning_rate": 3.2826443658049977e-06,
+      "loss": 0.0158,
+      "step": 736
+    },
+    {
+      "epoch": 1.9549071618037135,
+      "grad_norm": 0.46579167700659485,
+      "learning_rate": 3.2681468387438876e-06,
+      "loss": 0.0021,
+      "step": 737
+    },
+    {
+      "epoch": 1.9575596816976129,
+      "grad_norm": 0.3536695125051511,
+      "learning_rate": 3.253665837713694e-06,
+      "loss": 0.0022,
+      "step": 738
+    },
+    {
+      "epoch": 1.9602122015915118,
+      "grad_norm": 6.891295718169065,
+      "learning_rate": 3.239201500897881e-06,
+      "loss": 0.0033,
+      "step": 739
+    },
+    {
+      "epoch": 1.9628647214854111,
+      "grad_norm": 0.7698901035663716,
+      "learning_rate": 3.224753966320898e-06,
+      "loss": 0.0024,
+      "step": 740
+    },
+    {
+      "epoch": 1.9655172413793105,
+      "grad_norm": 0.11484925883131841,
+      "learning_rate": 3.2103233718468574e-06,
+      "loss": 0.0019,
+      "step": 741
+    },
+    {
+      "epoch": 1.9681697612732094,
+      "grad_norm": 0.11269632332694055,
+      "learning_rate": 3.1959098551782285e-06,
+      "loss": 0.0019,
+      "step": 742
+    },
+    {
+      "epoch": 1.9708222811671088,
+      "grad_norm": 0.10750371446519882,
+      "learning_rate": 3.181513553854514e-06,
+      "loss": 0.0018,
+      "step": 743
+    },
+    {
+      "epoch": 1.973474801061008,
+      "grad_norm": 27.91274099515536,
+      "learning_rate": 3.167134605250938e-06,
+      "loss": 0.0236,
+      "step": 744
+    },
+    {
+      "epoch": 1.976127320954907,
+      "grad_norm": 4.635156938141786,
+      "learning_rate": 3.152773146577138e-06,
+      "loss": 0.0028,
+      "step": 745
+    },
+    {
+      "epoch": 1.9787798408488064,
+      "grad_norm": 0.10837590253402313,
+      "learning_rate": 3.138429314875865e-06,
+      "loss": 0.0018,
+      "step": 746
+    },
+    {
+      "epoch": 1.9814323607427056,
+      "grad_norm": 0.10512317169385756,
+      "learning_rate": 3.1241032470216564e-06,
+      "loss": 0.0018,
+      "step": 747
+    },
+    {
+      "epoch": 1.9840848806366047,
+      "grad_norm": 0.15713070471420854,
+      "learning_rate": 3.109795079719544e-06,
+      "loss": 0.0018,
+      "step": 748
+    },
+    {
+      "epoch": 1.986737400530504,
+      "grad_norm": 0.1117481175631464,
+      "learning_rate": 3.0955049495037435e-06,
+      "loss": 0.0018,
+      "step": 749
+    },
+    {
+      "epoch": 1.9893899204244032,
+      "grad_norm": 0.8320658742184063,
+      "learning_rate": 3.081232992736355e-06,
+      "loss": 0.0022,
+      "step": 750
+    },
+    {
+      "epoch": 1.9920424403183024,
+      "grad_norm": 1.1226609993016856,
+      "learning_rate": 3.0669793456060613e-06,
+      "loss": 0.002,
+      "step": 751
+    },
+    {
+      "epoch": 1.9946949602122017,
+      "grad_norm": 9.229479885535163,
+      "learning_rate": 3.052744144126826e-06,
+      "loss": 0.0042,
+      "step": 752
+    },
+    {
+      "epoch": 1.9973474801061006,
+      "grad_norm": 18.484575713900153,
+      "learning_rate": 3.0385275241365965e-06,
+      "loss": 0.0055,
+      "step": 753
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.1049135207644034,
+      "learning_rate": 3.024329621296008e-06,
+      "loss": 0.0017,
+      "step": 754
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.2834593057632446,
+      "eval_runtime": 175.5899,
+      "eval_samples_per_second": 12.045,
+      "eval_steps_per_second": 1.509,
+      "step": 754
+    },
+    {
+      "epoch": 2.0026525198938994,
+      "grad_norm": 0.10101414623833138,
+      "learning_rate": 3.0101505710870914e-06,
+      "loss": 0.0017,
+      "step": 755
+    },
+    {
+      "epoch": 2.0053050397877983,
+      "grad_norm": 121.10117336213098,
+      "learning_rate": 2.9959905088119777e-06,
+      "loss": 0.017,
+      "step": 756
+    },
+    {
+      "epoch": 2.0079575596816976,
+      "grad_norm": 0.09676049678711801,
+      "learning_rate": 2.981849569591606e-06,
+      "loss": 0.0017,
+      "step": 757
+    },
+    {
+      "epoch": 2.010610079575597,
+      "grad_norm": 0.09764992907138446,
+      "learning_rate": 2.9677278883644367e-06,
+      "loss": 0.0017,
+      "step": 758
+    },
+    {
+      "epoch": 2.013262599469496,
+      "grad_norm": 0.12161105471130512,
+      "learning_rate": 2.9536255998851615e-06,
+      "loss": 0.0017,
+      "step": 759
+    },
+    {
+      "epoch": 2.0159151193633953,
+      "grad_norm": 0.3552964778774677,
+      "learning_rate": 2.9395428387234192e-06,
+      "loss": 0.0018,
+      "step": 760
+    },
+    {
+      "epoch": 2.018567639257294,
+      "grad_norm": 18.783349816477077,
+      "learning_rate": 2.9254797392625146e-06,
+      "loss": 0.0059,
+      "step": 761
+    },
+    {
+      "epoch": 2.0212201591511936,
+      "grad_norm": 0.09852900541452761,
+      "learning_rate": 2.9114364356981274e-06,
+      "loss": 0.0017,
+      "step": 762
+    },
+    {
+      "epoch": 2.023872679045093,
+      "grad_norm": 0.1438724087105207,
+      "learning_rate": 2.8974130620370405e-06,
+      "loss": 0.0017,
+      "step": 763
+    },
+    {
+      "epoch": 2.026525198938992,
+      "grad_norm": 29.762842374389095,
+      "learning_rate": 2.883409752095857e-06,
+      "loss": 0.1009,
+      "step": 764
+    },
+    {
+      "epoch": 2.029177718832891,
+      "grad_norm": 0.17818985776038995,
+      "learning_rate": 2.8694266394997238e-06,
+      "loss": 0.0017,
+      "step": 765
+    },
+    {
+      "epoch": 2.0318302387267906,
+      "grad_norm": 3.859943692060382,
+      "learning_rate": 2.8554638576810565e-06,
+      "loss": 0.0021,
+      "step": 766
+    },
+    {
+      "epoch": 2.0344827586206895,
+      "grad_norm": 0.625503230759066,
+      "learning_rate": 2.8415215398782657e-06,
+      "loss": 0.0017,
+      "step": 767
+    },
+    {
+      "epoch": 2.037135278514589,
+      "grad_norm": 0.4145565521734179,
+      "learning_rate": 2.827599819134489e-06,
+      "loss": 0.002,
+      "step": 768
+    },
+    {
+      "epoch": 2.0397877984084882,
+      "grad_norm": 0.24554217445213175,
+      "learning_rate": 2.813698828296312e-06,
+      "loss": 0.0017,
+      "step": 769
+    },
+    {
+      "epoch": 2.042440318302387,
+      "grad_norm": 11.827792784868253,
+      "learning_rate": 2.799818700012518e-06,
+      "loss": 0.0031,
+      "step": 770
+    },
+    {
+      "epoch": 2.0450928381962865,
+      "grad_norm": 0.09099465142665557,
+      "learning_rate": 2.7859595667328027e-06,
+      "loss": 0.0016,
+      "step": 771
+    },
+    {
+      "epoch": 2.047745358090186,
+      "grad_norm": 0.14639851471248677,
+      "learning_rate": 2.772121560706522e-06,
+      "loss": 0.0016,
+      "step": 772
+    },
+    {
+      "epoch": 2.050397877984085,
+      "grad_norm": 0.08792165154659169,
+      "learning_rate": 2.758304813981428e-06,
+      "loss": 0.0015,
+      "step": 773
+    },
+    {
+      "epoch": 2.053050397877984,
+      "grad_norm": 1.5842864989485481,
+      "learning_rate": 2.7445094584024067e-06,
+      "loss": 0.0019,
+      "step": 774
+    },
+    {
+      "epoch": 2.055702917771883,
+      "grad_norm": 0.09802686427150523,
+      "learning_rate": 2.7307356256102215e-06,
+      "loss": 0.0015,
+      "step": 775
+    },
+    {
+      "epoch": 2.0583554376657824,
+      "grad_norm": 26.05241910155726,
+      "learning_rate": 2.716983447040257e-06,
+      "loss": 0.0528,
+      "step": 776
+    },
+    {
+      "epoch": 2.061007957559682,
+      "grad_norm": 58.20790352116913,
+      "learning_rate": 2.703253053921266e-06,
+      "loss": 0.126,
+      "step": 777
+    },
+    {
+      "epoch": 2.0636604774535807,
+      "grad_norm": 1.0111648799278619,
+      "learning_rate": 2.689544577274113e-06,
+      "loss": 0.0017,
+      "step": 778
+    },
+    {
+      "epoch": 2.06631299734748,
+      "grad_norm": 0.23489459557522469,
+      "learning_rate": 2.6758581479105274e-06,
+      "loss": 0.0016,
+      "step": 779
+    },
+    {
+      "epoch": 2.0689655172413794,
+      "grad_norm": 53.804694742394936,
+      "learning_rate": 2.6621938964318593e-06,
+      "loss": 0.2099,
+      "step": 780
+    },
+    {
+      "epoch": 2.0716180371352784,
+      "grad_norm": 0.5405793593391763,
+      "learning_rate": 2.6485519532278235e-06,
+      "loss": 0.0017,
+      "step": 781
+    },
+    {
+      "epoch": 2.0742705570291777,
+      "grad_norm": 0.14108840494858138,
+      "learning_rate": 2.6349324484752612e-06,
+      "loss": 0.0015,
+      "step": 782
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.08587398205977333,
+      "learning_rate": 2.621335512136899e-06,
+      "loss": 0.0015,
+      "step": 783
+    },
+    {
+      "epoch": 2.079575596816976,
+      "grad_norm": 0.08756504123370183,
+      "learning_rate": 2.6077612739601015e-06,
+      "loss": 0.0015,
+      "step": 784
+    },
+    {
+      "epoch": 2.0822281167108754,
+      "grad_norm": 0.39557920489381565,
+      "learning_rate": 2.5942098634756475e-06,
+      "loss": 0.0017,
+      "step": 785
+    },
+    {
+      "epoch": 2.0848806366047747,
+      "grad_norm": 2.011210517395444,
+      "learning_rate": 2.580681409996477e-06,
+      "loss": 0.0026,
+      "step": 786
+    },
+    {
+      "epoch": 2.0875331564986737,
+      "grad_norm": 11.55596230506205,
+      "learning_rate": 2.567176042616471e-06,
+      "loss": 0.0047,
+      "step": 787
+    },
+    {
+      "epoch": 2.090185676392573,
+      "grad_norm": 2.4156310012643103,
+      "learning_rate": 2.5536938902092056e-06,
+      "loss": 0.0021,
+      "step": 788
+    },
+    {
+      "epoch": 2.092838196286472,
+      "grad_norm": 0.18764992825176918,
+      "learning_rate": 2.5402350814267364e-06,
+      "loss": 0.0017,
+      "step": 789
+    },
+    {
+      "epoch": 2.0954907161803713,
+      "grad_norm": 2.2919461648589556,
+      "learning_rate": 2.526799744698366e-06,
+      "loss": 0.0025,
+      "step": 790
+    },
+    {
+      "epoch": 2.0981432360742707,
+      "grad_norm": 7.710549965295524,
+      "learning_rate": 2.5133880082294155e-06,
+      "loss": 0.005,
+      "step": 791
+    },
+    {
+      "epoch": 2.1007957559681696,
+      "grad_norm": 0.15012400497585224,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.0016,
+      "step": 792
+    },
+    {
+      "epoch": 2.103448275862069,
+      "grad_norm": 0.08220096963803952,
+      "learning_rate": 2.486635847763815e-06,
+      "loss": 0.0014,
+      "step": 793
+    },
+    {
+      "epoch": 2.1061007957559683,
+      "grad_norm": 0.7783894450675292,
+      "learning_rate": 2.473295679046911e-06,
+      "loss": 0.0018,
+      "step": 794
+    },
+    {
+      "epoch": 2.1087533156498672,
+      "grad_norm": 0.12477741474784955,
+      "learning_rate": 2.4599796211464772e-06,
+      "loss": 0.0015,
+      "step": 795
+    },
+    {
+      "epoch": 2.1114058355437666,
+      "grad_norm": 0.08144394193328244,
+      "learning_rate": 2.446687801129628e-06,
+      "loss": 0.0014,
+      "step": 796
+    },
+    {
+      "epoch": 2.114058355437666,
+      "grad_norm": 31.61865013891495,
+      "learning_rate": 2.433420345832191e-06,
+      "loss": 0.0235,
+      "step": 797
+    },
+    {
+      "epoch": 2.116710875331565,
+      "grad_norm": 0.16630727907217205,
+      "learning_rate": 2.4201773818574956e-06,
+      "loss": 0.0015,
+      "step": 798
+    },
+    {
+      "epoch": 2.1193633952254642,
+      "grad_norm": 0.08241735796809634,
+      "learning_rate": 2.406959035575166e-06,
+      "loss": 0.0014,
+      "step": 799
+    },
+    {
+      "epoch": 2.1220159151193636,
+      "grad_norm": 3.062034039881723,
+      "learning_rate": 2.393765433119913e-06,
+      "loss": 0.0021,
+      "step": 800
+    },
+    {
+      "epoch": 2.1246684350132625,
+      "grad_norm": 0.11937503096401074,
+      "learning_rate": 2.3805967003903336e-06,
+      "loss": 0.0014,
+      "step": 801
+    },
+    {
+      "epoch": 2.127320954907162,
+      "grad_norm": 20.773342722596187,
+      "learning_rate": 2.3674529630477074e-06,
+      "loss": 0.0073,
+      "step": 802
+    },
+    {
+      "epoch": 2.129973474801061,
+      "grad_norm": 0.46381492457559487,
+      "learning_rate": 2.3543343465147956e-06,
+      "loss": 0.0019,
+      "step": 803
+    },
+    {
+      "epoch": 2.13262599469496,
+      "grad_norm": 0.0804988750293357,
+      "learning_rate": 2.341240975974653e-06,
+      "loss": 0.0014,
+      "step": 804
+    },
+    {
+      "epoch": 2.1352785145888595,
+      "grad_norm": 0.3497581554814395,
+      "learning_rate": 2.328172976369421e-06,
+      "loss": 0.0016,
+      "step": 805
+    },
+    {
+      "epoch": 2.1379310344827585,
+      "grad_norm": 0.07679137336103294,
+      "learning_rate": 2.315130472399145e-06,
+      "loss": 0.0013,
+      "step": 806
+    },
+    {
+      "epoch": 2.140583554376658,
+      "grad_norm": 0.07466622329861163,
+      "learning_rate": 2.302113588520578e-06,
+      "loss": 0.0013,
+      "step": 807
+    },
+    {
+      "epoch": 2.143236074270557,
+      "grad_norm": 0.07758030739160703,
+      "learning_rate": 2.289122448945997e-06,
+      "loss": 0.0013,
+      "step": 808
+    },
+    {
+      "epoch": 2.145888594164456,
+      "grad_norm": 0.16681181140025278,
+      "learning_rate": 2.2761571776420187e-06,
+      "loss": 0.0014,
+      "step": 809
+    },
+    {
+      "epoch": 2.1485411140583555,
+      "grad_norm": 7.399139547650065,
+      "learning_rate": 2.263217898328415e-06,
+      "loss": 0.0049,
+      "step": 810
+    },
+    {
+      "epoch": 2.151193633952255,
+      "grad_norm": 0.5942035002264717,
+      "learning_rate": 2.2503047344769256e-06,
+      "loss": 0.0014,
+      "step": 811
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.08081617536253838,
+      "learning_rate": 2.23741780931009e-06,
+      "loss": 0.0013,
+      "step": 812
+    },
+    {
+      "epoch": 2.156498673740053,
+      "grad_norm": 0.7462752578343561,
+      "learning_rate": 2.2245572458000714e-06,
+      "loss": 0.0014,
+      "step": 813
+    },
+    {
+      "epoch": 2.159151193633952,
+      "grad_norm": 19.74892285398778,
+      "learning_rate": 2.211723166667475e-06,
+      "loss": 0.006,
+      "step": 814
+    },
+    {
+      "epoch": 2.1618037135278514,
+      "grad_norm": 0.08005713357681245,
+      "learning_rate": 2.198915694380182e-06,
+      "loss": 0.0013,
+      "step": 815
+    },
+    {
+      "epoch": 2.1644562334217508,
+      "grad_norm": 0.6071057574038246,
+      "learning_rate": 2.1861349511521817e-06,
+      "loss": 0.0014,
+      "step": 816
+    },
+    {
+      "epoch": 2.1671087533156497,
+      "grad_norm": 0.0752186839187155,
+      "learning_rate": 2.173381058942402e-06,
+      "loss": 0.0013,
+      "step": 817
+    },
+    {
+      "epoch": 2.169761273209549,
+      "grad_norm": 0.1305584374266705,
+      "learning_rate": 2.1606541394535528e-06,
+      "loss": 0.0013,
+      "step": 818
+    },
+    {
+      "epoch": 2.1724137931034484,
+      "grad_norm": 0.07375217036909364,
+      "learning_rate": 2.147954314130955e-06,
+      "loss": 0.0013,
+      "step": 819
+    },
+    {
+      "epoch": 2.1750663129973473,
+      "grad_norm": 0.08838482083846465,
+      "learning_rate": 2.135281704161386e-06,
+      "loss": 0.0013,
+      "step": 820
+    },
+    {
+      "epoch": 2.1777188328912467,
+      "grad_norm": 0.07360478242627387,
+      "learning_rate": 2.122636430471926e-06,
+      "loss": 0.0013,
+      "step": 821
+    },
+    {
+      "epoch": 2.180371352785146,
+      "grad_norm": 0.08514009164619807,
+      "learning_rate": 2.1100186137288005e-06,
+      "loss": 0.0013,
+      "step": 822
+    },
+    {
+      "epoch": 2.183023872679045,
+      "grad_norm": 125.92010206347204,
+      "learning_rate": 2.0974283743362283e-06,
+      "loss": 0.0039,
+      "step": 823
+    },
+    {
+      "epoch": 2.1856763925729443,
+      "grad_norm": 0.07215971233954403,
+      "learning_rate": 2.084865832435278e-06,
+      "loss": 0.0012,
+      "step": 824
+    },
+    {
+      "epoch": 2.1883289124668437,
+      "grad_norm": 0.07285977518445678,
+      "learning_rate": 2.072331107902713e-06,
+      "loss": 0.0013,
+      "step": 825
+    },
+    {
+      "epoch": 2.1909814323607426,
+      "grad_norm": 0.07234080373019404,
+      "learning_rate": 2.0598243203498562e-06,
+      "loss": 0.0013,
+      "step": 826
+    },
+    {
+      "epoch": 2.193633952254642,
+      "grad_norm": 0.0886750422890646,
+      "learning_rate": 2.0473455891214416e-06,
+      "loss": 0.0013,
+      "step": 827
+    },
+    {
+      "epoch": 2.1962864721485413,
+      "grad_norm": 0.07148362497663868,
+      "learning_rate": 2.034895033294483e-06,
+      "loss": 0.0012,
+      "step": 828
+    },
+    {
+      "epoch": 2.1989389920424403,
+      "grad_norm": 0.09041579908666639,
+      "learning_rate": 2.0224727716771297e-06,
+      "loss": 0.0013,
+      "step": 829
+    },
+    {
+      "epoch": 2.2015915119363396,
+      "grad_norm": 0.26201814358060044,
+      "learning_rate": 2.0100789228075375e-06,
+      "loss": 0.0013,
+      "step": 830
+    },
+    {
+      "epoch": 2.2042440318302385,
+      "grad_norm": 0.07092692566307471,
+      "learning_rate": 1.9977136049527348e-06,
+      "loss": 0.0012,
+      "step": 831
+    },
+    {
+      "epoch": 2.206896551724138,
+      "grad_norm": 0.7680618172089979,
+      "learning_rate": 1.9853769361074964e-06,
+      "loss": 0.0013,
+      "step": 832
+    },
+    {
+      "epoch": 2.2095490716180373,
+      "grad_norm": 0.08149814922086594,
+      "learning_rate": 1.973069033993223e-06,
+      "loss": 0.0012,
+      "step": 833
+    },
+    {
+      "epoch": 2.212201591511936,
+      "grad_norm": 0.06918207979942406,
+      "learning_rate": 1.960790016056801e-06,
+      "loss": 0.0012,
+      "step": 834
+    },
+    {
+      "epoch": 2.2148541114058355,
+      "grad_norm": 0.06941092834274308,
+      "learning_rate": 1.9485399994694998e-06,
+      "loss": 0.0012,
+      "step": 835
+    },
+    {
+      "epoch": 2.217506631299735,
+      "grad_norm": 0.11042988652448069,
+      "learning_rate": 1.9363191011258426e-06,
+      "loss": 0.0013,
+      "step": 836
+    },
+    {
+      "epoch": 2.220159151193634,
+      "grad_norm": 0.06929264033253076,
+      "learning_rate": 1.9241274376425e-06,
+      "loss": 0.0012,
+      "step": 837
+    },
+    {
+      "epoch": 2.222811671087533,
+      "grad_norm": 0.07191532035559675,
+      "learning_rate": 1.9119651253571676e-06,
+      "loss": 0.0012,
+      "step": 838
+    },
+    {
+      "epoch": 2.2254641909814326,
+      "grad_norm": 0.19378968639900365,
+      "learning_rate": 1.89983228032746e-06,
+      "loss": 0.0013,
+      "step": 839
+    },
+    {
+      "epoch": 2.2281167108753315,
+      "grad_norm": 32.12461736878749,
+      "learning_rate": 1.8877290183298058e-06,
+      "loss": 0.0479,
+      "step": 840
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.43830491728757237,
+      "learning_rate": 1.8756554548583377e-06,
+      "loss": 0.0013,
+      "step": 841
+    },
+    {
+      "epoch": 2.2334217506631298,
+      "grad_norm": 0.2187235621150115,
+      "learning_rate": 1.863611705123798e-06,
+      "loss": 0.0013,
+      "step": 842
+    },
+    {
+      "epoch": 2.236074270557029,
+      "grad_norm": 0.6143192857944316,
+      "learning_rate": 1.8515978840524302e-06,
+      "loss": 0.0014,
+      "step": 843
+    },
+    {
+      "epoch": 2.2387267904509285,
+      "grad_norm": 0.11313871454305385,
+      "learning_rate": 1.8396141062848877e-06,
+      "loss": 0.0012,
+      "step": 844
+    },
+    {
+      "epoch": 2.2413793103448274,
+      "grad_norm": 45.11241620874413,
+      "learning_rate": 1.827660486175139e-06,
+      "loss": 0.0059,
+      "step": 845
+    },
+    {
+      "epoch": 2.2440318302387268,
+      "grad_norm": 0.06623676243698338,
+      "learning_rate": 1.8157371377893769e-06,
+      "loss": 0.0012,
+      "step": 846
+    },
+    {
+      "epoch": 2.246684350132626,
+      "grad_norm": 0.15303018443046673,
+      "learning_rate": 1.803844174904928e-06,
+      "loss": 0.0012,
+      "step": 847
+    },
+    {
+      "epoch": 2.249336870026525,
+      "grad_norm": 0.0670464354914666,
+      "learning_rate": 1.7919817110091691e-06,
+      "loss": 0.0012,
+      "step": 848
+    },
+    {
+      "epoch": 2.2519893899204244,
+      "grad_norm": 0.06709139284552051,
+      "learning_rate": 1.7801498592984445e-06,
+      "loss": 0.0012,
+      "step": 849
+    },
+    {
+      "epoch": 2.2546419098143238,
+      "grad_norm": 1.8527488282182363,
+      "learning_rate": 1.7683487326769826e-06,
+      "loss": 0.0017,
+      "step": 850
+    },
+    {
+      "epoch": 2.2572944297082227,
+      "grad_norm": 0.0722005269953955,
+      "learning_rate": 1.756578443755822e-06,
+      "loss": 0.0011,
+      "step": 851
+    },
+    {
+      "epoch": 2.259946949602122,
+      "grad_norm": 0.06645532711286986,
+      "learning_rate": 1.7448391048517378e-06,
+      "loss": 0.0012,
+      "step": 852
+    },
+    {
+      "epoch": 2.2625994694960214,
+      "grad_norm": 0.06561899843988953,
+      "learning_rate": 1.7331308279861641e-06,
+      "loss": 0.0011,
+      "step": 853
+    },
+    {
+      "epoch": 2.2652519893899203,
+      "grad_norm": 0.0661860205744516,
+      "learning_rate": 1.7214537248841317e-06,
+      "loss": 0.0012,
+      "step": 854
+    },
+    {
+      "epoch": 2.2679045092838197,
+      "grad_norm": 0.6079953038884391,
+      "learning_rate": 1.709807906973196e-06,
+      "loss": 0.0014,
+      "step": 855
+    },
+    {
+      "epoch": 2.270557029177719,
+      "grad_norm": 0.06488881068760848,
+      "learning_rate": 1.6981934853823796e-06,
+      "loss": 0.0011,
+      "step": 856
+    },
+    {
+      "epoch": 2.273209549071618,
+      "grad_norm": 0.06797124970633543,
+      "learning_rate": 1.6866105709411069e-06,
+      "loss": 0.0011,
+      "step": 857
+    },
+    {
+      "epoch": 2.2758620689655173,
+      "grad_norm": 76.16571947400874,
+      "learning_rate": 1.6750592741781496e-06,
+      "loss": 0.3501,
+      "step": 858
+    },
+    {
+      "epoch": 2.2785145888594163,
+      "grad_norm": 0.06558683917842384,
+      "learning_rate": 1.6635397053205704e-06,
+      "loss": 0.0011,
+      "step": 859
+    },
+    {
+      "epoch": 2.2811671087533156,
+      "grad_norm": 0.09655686844705165,
+      "learning_rate": 1.6520519742926705e-06,
+      "loss": 0.0011,
+      "step": 860
+    },
+    {
+      "epoch": 2.283819628647215,
+      "grad_norm": 0.0686245128786292,
+      "learning_rate": 1.640596190714947e-06,
+      "loss": 0.0012,
+      "step": 861
+    },
+    {
+      "epoch": 2.286472148541114,
+      "grad_norm": 0.0650546943127916,
+      "learning_rate": 1.6291724639030353e-06,
+      "loss": 0.0011,
+      "step": 862
+    },
+    {
+      "epoch": 2.2891246684350133,
+      "grad_norm": 208.7702949244788,
+      "learning_rate": 1.6177809028666769e-06,
+      "loss": 0.012,
+      "step": 863
+    },
+    {
+      "epoch": 2.2917771883289126,
+      "grad_norm": 29.89517410626624,
+      "learning_rate": 1.6064216163086716e-06,
+      "loss": 0.0623,
+      "step": 864
+    },
+    {
+      "epoch": 2.2944297082228116,
+      "grad_norm": 0.0653400164041259,
+      "learning_rate": 1.595094712623843e-06,
+      "loss": 0.0011,
+      "step": 865
+    },
+    {
+      "epoch": 2.297082228116711,
+      "grad_norm": 0.06671777147106228,
+      "learning_rate": 1.5838002998980107e-06,
+      "loss": 0.0011,
+      "step": 866
+    },
+    {
+      "epoch": 2.29973474801061,
+      "grad_norm": 0.08549683500973079,
+      "learning_rate": 1.5725384859069454e-06,
+      "loss": 0.0011,
+      "step": 867
+    },
+    {
+      "epoch": 2.302387267904509,
+      "grad_norm": 1.6239494853557073,
+      "learning_rate": 1.5613093781153503e-06,
+      "loss": 0.0015,
+      "step": 868
+    },
+    {
+      "epoch": 2.3050397877984086,
+      "grad_norm": 0.06268027521119812,
+      "learning_rate": 1.550113083675836e-06,
+      "loss": 0.0011,
+      "step": 869
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 21.69711575041162,
+      "learning_rate": 1.5389497094278861e-06,
+      "loss": 0.007,
+      "step": 870
+    },
+    {
+      "epoch": 2.310344827586207,
+      "grad_norm": 0.06701049997320575,
+      "learning_rate": 1.5278193618968584e-06,
+      "loss": 0.0011,
+      "step": 871
+    },
+    {
+      "epoch": 2.312997347480106,
+      "grad_norm": 0.07382333923573142,
+      "learning_rate": 1.5167221472929489e-06,
+      "loss": 0.0011,
+      "step": 872
+    },
+    {
+      "epoch": 2.315649867374005,
+      "grad_norm": 76.7596193056003,
+      "learning_rate": 1.5056581715101887e-06,
+      "loss": 0.0647,
+      "step": 873
+    },
+    {
+      "epoch": 2.3183023872679045,
+      "grad_norm": 15.64396853251376,
+      "learning_rate": 1.4946275401254301e-06,
+      "loss": 0.0063,
+      "step": 874
+    },
+    {
+      "epoch": 2.320954907161804,
+      "grad_norm": 0.1214125458833792,
+      "learning_rate": 1.4836303583973384e-06,
+      "loss": 0.0011,
+      "step": 875
+    },
+    {
+      "epoch": 2.323607427055703,
+      "grad_norm": 122.35872430226972,
+      "learning_rate": 1.472666731265394e-06,
+      "loss": 0.0435,
+      "step": 876
+    },
+    {
+      "epoch": 2.326259946949602,
+      "grad_norm": 0.06264186932736737,
+      "learning_rate": 1.4617367633488816e-06,
+      "loss": 0.0011,
+      "step": 877
+    },
+    {
+      "epoch": 2.3289124668435015,
+      "grad_norm": 0.06412223063437182,
+      "learning_rate": 1.4508405589458968e-06,
+      "loss": 0.0011,
+      "step": 878
+    },
+    {
+      "epoch": 2.3315649867374004,
+      "grad_norm": 0.06418332438027141,
+      "learning_rate": 1.4399782220323515e-06,
+      "loss": 0.0011,
+      "step": 879
+    },
+    {
+      "epoch": 2.3342175066313,
+      "grad_norm": 0.06185564601567612,
+      "learning_rate": 1.4291498562609802e-06,
+      "loss": 0.0011,
+      "step": 880
+    },
+    {
+      "epoch": 2.336870026525199,
+      "grad_norm": 0.06336656034657022,
+      "learning_rate": 1.4183555649603503e-06,
+      "loss": 0.0011,
+      "step": 881
+    },
+    {
+      "epoch": 2.339522546419098,
+      "grad_norm": 0.06146985918004954,
+      "learning_rate": 1.4075954511338784e-06,
+      "loss": 0.0011,
+      "step": 882
+    },
+    {
+      "epoch": 2.3421750663129974,
+      "grad_norm": 0.08053700437359687,
+      "learning_rate": 1.396869617458846e-06,
+      "loss": 0.0011,
+      "step": 883
+    },
+    {
+      "epoch": 2.344827586206897,
+      "grad_norm": 0.5218052672812636,
+      "learning_rate": 1.3861781662854162e-06,
+      "loss": 0.0012,
+      "step": 884
+    },
+    {
+      "epoch": 2.3474801061007957,
+      "grad_norm": 1.5216995484373934,
+      "learning_rate": 1.3755211996356687e-06,
+      "loss": 0.0015,
+      "step": 885
+    },
+    {
+      "epoch": 2.350132625994695,
+      "grad_norm": 0.0621162072022446,
+      "learning_rate": 1.3648988192026108e-06,
+      "loss": 0.0011,
+      "step": 886
+    },
+    {
+      "epoch": 2.352785145888594,
+      "grad_norm": 0.061884930132394454,
+      "learning_rate": 1.3543111263492165e-06,
+      "loss": 0.0011,
+      "step": 887
+    },
+    {
+      "epoch": 2.3554376657824934,
+      "grad_norm": 0.06028055456747343,
+      "learning_rate": 1.3437582221074574e-06,
+      "loss": 0.001,
+      "step": 888
+    },
+    {
+      "epoch": 2.3580901856763927,
+      "grad_norm": 0.08255518588823123,
+      "learning_rate": 1.3332402071773376e-06,
+      "loss": 0.0011,
+      "step": 889
+    },
+    {
+      "epoch": 2.3607427055702916,
+      "grad_norm": 0.0660571988137438,
+      "learning_rate": 1.322757181925937e-06,
+      "loss": 0.0011,
+      "step": 890
+    },
+    {
+      "epoch": 2.363395225464191,
+      "grad_norm": 5.704097339238524,
+      "learning_rate": 1.3123092463864456e-06,
+      "loss": 0.0013,
+      "step": 891
+    },
+    {
+      "epoch": 2.3660477453580904,
+      "grad_norm": 139.48648832304622,
+      "learning_rate": 1.301896500257217e-06,
+      "loss": 0.1495,
+      "step": 892
+    },
+    {
+      "epoch": 2.3687002652519893,
+      "grad_norm": 0.06562913191383259,
+      "learning_rate": 1.2915190429008084e-06,
+      "loss": 0.0011,
+      "step": 893
+    },
+    {
+      "epoch": 2.3713527851458887,
+      "grad_norm": 0.06094523998371494,
+      "learning_rate": 1.2811769733430406e-06,
+      "loss": 0.0011,
+      "step": 894
+    },
+    {
+      "epoch": 2.3740053050397876,
+      "grad_norm": 0.06227692335984811,
+      "learning_rate": 1.2708703902720538e-06,
+      "loss": 0.001,
+      "step": 895
+    },
+    {
+      "epoch": 2.376657824933687,
+      "grad_norm": 0.06695392791086113,
+      "learning_rate": 1.260599392037356e-06,
+      "loss": 0.0011,
+      "step": 896
+    },
+    {
+      "epoch": 2.3793103448275863,
+      "grad_norm": 0.21611121807915867,
+      "learning_rate": 1.250364076648894e-06,
+      "loss": 0.0012,
+      "step": 897
+    },
+    {
+      "epoch": 2.381962864721485,
+      "grad_norm": 1.8698349335731828,
+      "learning_rate": 1.2401645417761126e-06,
+      "loss": 0.0014,
+      "step": 898
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.5713724389512606,
+      "learning_rate": 1.2300008847470252e-06,
+      "loss": 0.0012,
+      "step": 899
+    },
+    {
+      "epoch": 2.387267904509284,
+      "grad_norm": 47.11357878755216,
+      "learning_rate": 1.2198732025472876e-06,
+      "loss": 0.1677,
+      "step": 900
+    },
+    {
+      "epoch": 2.389920424403183,
+      "grad_norm": 0.06090183448028676,
+      "learning_rate": 1.2097815918192652e-06,
+      "loss": 0.001,
+      "step": 901
+    },
+    {
+      "epoch": 2.3925729442970822,
+      "grad_norm": 0.060156871686836734,
+      "learning_rate": 1.1997261488611173e-06,
+      "loss": 0.001,
+      "step": 902
+    },
+    {
+      "epoch": 2.3952254641909816,
+      "grad_norm": 0.06807379303255552,
+      "learning_rate": 1.1897069696258756e-06,
+      "loss": 0.001,
+      "step": 903
+    },
+    {
+      "epoch": 2.3978779840848805,
+      "grad_norm": 0.07237555713720957,
+      "learning_rate": 1.1797241497205285e-06,
+      "loss": 0.0011,
+      "step": 904
+    },
+    {
+      "epoch": 2.40053050397878,
+      "grad_norm": 0.06004611220653036,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.001,
+      "step": 905
+    },
+    {
+      "epoch": 2.4031830238726792,
+      "grad_norm": 0.05909328032696501,
+      "learning_rate": 1.1598679685917901e-06,
+      "loss": 0.001,
+      "step": 906
+    },
+    {
+      "epoch": 2.405835543766578,
+      "grad_norm": 0.0611763251688093,
+      "learning_rate": 1.1499947968439673e-06,
+      "loss": 0.001,
+      "step": 907
+    },
+    {
+      "epoch": 2.4084880636604775,
+      "grad_norm": 0.07785505410015264,
+      "learning_rate": 1.1401583633753683e-06,
+      "loss": 0.001,
+      "step": 908
+    },
+    {
+      "epoch": 2.411140583554377,
+      "grad_norm": 0.07389775729807248,
+      "learning_rate": 1.1303587620491513e-06,
+      "loss": 0.001,
+      "step": 909
+    },
+    {
+      "epoch": 2.413793103448276,
+      "grad_norm": 0.07855752742958044,
+      "learning_rate": 1.120596086377005e-06,
+      "loss": 0.001,
+      "step": 910
+    },
+    {
+      "epoch": 2.416445623342175,
+      "grad_norm": 0.4890171657386728,
+      "learning_rate": 1.1108704295182582e-06,
+      "loss": 0.0012,
+      "step": 911
+    },
+    {
+      "epoch": 2.419098143236074,
+      "grad_norm": 41.63509191196258,
+      "learning_rate": 1.1011818842789928e-06,
+      "loss": 0.0063,
+      "step": 912
+    },
+    {
+      "epoch": 2.4217506631299734,
+      "grad_norm": 0.06306195603906277,
+      "learning_rate": 1.0915305431111561e-06,
+      "loss": 0.001,
+      "step": 913
+    },
+    {
+      "epoch": 2.424403183023873,
+      "grad_norm": 0.9707371949916439,
+      "learning_rate": 1.0819164981116825e-06,
+      "loss": 0.0012,
+      "step": 914
+    },
+    {
+      "epoch": 2.4270557029177717,
+      "grad_norm": 0.0811953522176445,
+      "learning_rate": 1.0723398410216085e-06,
+      "loss": 0.001,
+      "step": 915
+    },
+    {
+      "epoch": 2.429708222811671,
+      "grad_norm": 0.06384966998157086,
+      "learning_rate": 1.0628006632251975e-06,
+      "loss": 0.001,
+      "step": 916
+    },
+    {
+      "epoch": 2.4323607427055705,
+      "grad_norm": 0.05995114120465396,
+      "learning_rate": 1.0532990557490768e-06,
+      "loss": 0.001,
+      "step": 917
+    },
+    {
+      "epoch": 2.4350132625994694,
+      "grad_norm": 2.4610924632978253,
+      "learning_rate": 1.043835109261357e-06,
+      "loss": 0.0016,
+      "step": 918
+    },
+    {
+      "epoch": 2.4376657824933687,
+      "grad_norm": 0.05878584207452931,
+      "learning_rate": 1.034408914070779e-06,
+      "loss": 0.001,
+      "step": 919
+    },
+    {
+      "epoch": 2.4403183023872677,
+      "grad_norm": 0.0578918748475833,
+      "learning_rate": 1.0250205601258407e-06,
+      "loss": 0.001,
+      "step": 920
+    },
+    {
+      "epoch": 2.442970822281167,
+      "grad_norm": 126.67531388600968,
+      "learning_rate": 1.0156701370139454e-06,
+      "loss": 0.0323,
+      "step": 921
+    },
+    {
+      "epoch": 2.4456233421750664,
+      "grad_norm": 0.08277846289252522,
+      "learning_rate": 1.0063577339605452e-06,
+      "loss": 0.001,
+      "step": 922
+    },
+    {
+      "epoch": 2.4482758620689653,
+      "grad_norm": 0.05811273002395958,
+      "learning_rate": 9.970834398282887e-07,
+      "loss": 0.001,
+      "step": 923
+    },
+    {
+      "epoch": 2.4509283819628647,
+      "grad_norm": 0.06580267260863659,
+      "learning_rate": 9.878473431161767e-07,
+      "loss": 0.001,
+      "step": 924
+    },
+    {
+      "epoch": 2.453580901856764,
+      "grad_norm": 0.1232830259114401,
+      "learning_rate": 9.786495319587136e-07,
+      "loss": 0.001,
+      "step": 925
+    },
+    {
+      "epoch": 2.456233421750663,
+      "grad_norm": 0.28685021836803104,
+      "learning_rate": 9.694900941250674e-07,
+      "loss": 0.0011,
+      "step": 926
+    },
+    {
+      "epoch": 2.4588859416445623,
+      "grad_norm": 0.0628208200867451,
+      "learning_rate": 9.603691170182316e-07,
+      "loss": 0.001,
+      "step": 927
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.05767179007096357,
+      "learning_rate": 9.512866876741949e-07,
+      "loss": 0.001,
+      "step": 928
+    },
+    {
+      "epoch": 2.4641909814323606,
+      "grad_norm": 0.05651101689159219,
+      "learning_rate": 9.42242892761106e-07,
+      "loss": 0.001,
+      "step": 929
+    },
+    {
+      "epoch": 2.46684350132626,
+      "grad_norm": 0.1312083202582577,
+      "learning_rate": 9.332378185784491e-07,
+      "loss": 0.001,
+      "step": 930
+    },
+    {
+      "epoch": 2.4694960212201593,
+      "grad_norm": 0.0794531209151415,
+      "learning_rate": 9.242715510562195e-07,
+      "loss": 0.001,
+      "step": 931
+    },
+    {
+      "epoch": 2.4721485411140582,
+      "grad_norm": 0.06758011176979471,
+      "learning_rate": 9.153441757541026e-07,
+      "loss": 0.001,
+      "step": 932
+    },
+    {
+      "epoch": 2.4748010610079576,
+      "grad_norm": 0.05628258012555615,
+      "learning_rate": 9.064557778606631e-07,
+      "loss": 0.001,
+      "step": 933
+    },
+    {
+      "epoch": 2.477453580901857,
+      "grad_norm": 0.8798413491531053,
+      "learning_rate": 8.97606442192524e-07,
+      "loss": 0.0013,
+      "step": 934
+    },
+    {
+      "epoch": 2.480106100795756,
+      "grad_norm": 0.058598663584796984,
+      "learning_rate": 8.887962531935612e-07,
+      "loss": 0.001,
+      "step": 935
+    },
+    {
+      "epoch": 2.4827586206896552,
+      "grad_norm": 0.09406748583021195,
+      "learning_rate": 8.800252949340998e-07,
+      "loss": 0.001,
+      "step": 936
+    },
+    {
+      "epoch": 2.4854111405835546,
+      "grad_norm": 0.11316186423995639,
+      "learning_rate": 8.712936511101056e-07,
+      "loss": 0.001,
+      "step": 937
+    },
+    {
+      "epoch": 2.4880636604774535,
+      "grad_norm": 0.05689090074692826,
+      "learning_rate": 8.62601405042397e-07,
+      "loss": 0.001,
+      "step": 938
+    },
+    {
+      "epoch": 2.490716180371353,
+      "grad_norm": 0.05648519376070825,
+      "learning_rate": 8.539486396758357e-07,
+      "loss": 0.001,
+      "step": 939
+    },
+    {
+      "epoch": 2.493368700265252,
+      "grad_norm": 16.679551063134323,
+      "learning_rate": 8.453354375785477e-07,
+      "loss": 0.0074,
+      "step": 940
+    },
+    {
+      "epoch": 2.496021220159151,
+      "grad_norm": 3.5045054099899935,
+      "learning_rate": 8.367618809411299e-07,
+      "loss": 0.0014,
+      "step": 941
+    },
+    {
+      "epoch": 2.4986737400530505,
+      "grad_norm": 0.0629968232537991,
+      "learning_rate": 8.282280515758639e-07,
+      "loss": 0.001,
+      "step": 942
+    },
+    {
+      "epoch": 2.5013262599469495,
+      "grad_norm": 0.056459018233367006,
+      "learning_rate": 8.197340309159429e-07,
+      "loss": 0.001,
+      "step": 943
+    },
+    {
+      "epoch": 2.503978779840849,
+      "grad_norm": 0.057128628619581345,
+      "learning_rate": 8.112799000146853e-07,
+      "loss": 0.001,
+      "step": 944
+    },
+    {
+      "epoch": 2.5066312997347477,
+      "grad_norm": 9.50454174227497,
+      "learning_rate": 8.02865739544767e-07,
+      "loss": 0.0016,
+      "step": 945
+    },
+    {
+      "epoch": 2.509283819628647,
+      "grad_norm": 5.4634210425161855,
+      "learning_rate": 7.944916297974498e-07,
+      "loss": 0.0052,
+      "step": 946
+    },
+    {
+      "epoch": 2.5119363395225465,
+      "grad_norm": 0.05572183180594424,
+      "learning_rate": 7.861576506818147e-07,
+      "loss": 0.001,
+      "step": 947
+    },
+    {
+      "epoch": 2.5145888594164454,
+      "grad_norm": 0.05649628277552702,
+      "learning_rate": 7.778638817240042e-07,
+      "loss": 0.001,
+      "step": 948
+    },
+    {
+      "epoch": 2.5172413793103448,
+      "grad_norm": 0.05583993949613943,
+      "learning_rate": 7.696104020664552e-07,
+      "loss": 0.001,
+      "step": 949
+    },
+    {
+      "epoch": 2.519893899204244,
+      "grad_norm": 0.06926676295906672,
+      "learning_rate": 7.613972904671496e-07,
+      "loss": 0.001,
+      "step": 950
+    },
+    {
+      "epoch": 2.522546419098143,
+      "grad_norm": 0.33919856089089423,
+      "learning_rate": 7.532246252988617e-07,
+      "loss": 0.0011,
+      "step": 951
+    },
+    {
+      "epoch": 2.5251989389920424,
+      "grad_norm": 0.0568461133850681,
+      "learning_rate": 7.450924845484092e-07,
+      "loss": 0.001,
+      "step": 952
+    },
+    {
+      "epoch": 2.5278514588859418,
+      "grad_norm": 0.05561015970931866,
+      "learning_rate": 7.370009458159099e-07,
+      "loss": 0.001,
+      "step": 953
+    },
+    {
+      "epoch": 2.5305039787798407,
+      "grad_norm": 193.0480706166829,
+      "learning_rate": 7.289500863140414e-07,
+      "loss": 0.0609,
+      "step": 954
+    },
+    {
+      "epoch": 2.53315649867374,
+      "grad_norm": 0.05494039154416162,
+      "learning_rate": 7.20939982867303e-07,
+      "loss": 0.001,
+      "step": 955
+    },
+    {
+      "epoch": 2.5358090185676394,
+      "grad_norm": 5.829073488339818,
+      "learning_rate": 7.129707119112838e-07,
+      "loss": 0.0017,
+      "step": 956
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.05435253762937014,
+      "learning_rate": 7.05042349491935e-07,
+      "loss": 0.0009,
+      "step": 957
+    },
+    {
+      "epoch": 2.5411140583554377,
+      "grad_norm": 0.055234367394669366,
+      "learning_rate": 6.971549712648401e-07,
+      "loss": 0.001,
+      "step": 958
+    },
+    {
+      "epoch": 2.543766578249337,
+      "grad_norm": 0.06126906513727519,
+      "learning_rate": 6.893086524944953e-07,
+      "loss": 0.001,
+      "step": 959
+    },
+    {
+      "epoch": 2.546419098143236,
+      "grad_norm": 0.055051354699559824,
+      "learning_rate": 6.815034680535915e-07,
+      "loss": 0.001,
+      "step": 960
+    },
+    {
+      "epoch": 2.5490716180371353,
+      "grad_norm": 0.054625158391608,
+      "learning_rate": 6.737394924223e-07,
+      "loss": 0.001,
+      "step": 961
+    },
+    {
+      "epoch": 2.5517241379310347,
+      "grad_norm": 0.05462583302647743,
+      "learning_rate": 6.660167996875605e-07,
+      "loss": 0.001,
+      "step": 962
+    },
+    {
+      "epoch": 2.5543766578249336,
+      "grad_norm": 0.0539700747311138,
+      "learning_rate": 6.583354635423755e-07,
+      "loss": 0.0009,
+      "step": 963
+    },
+    {
+      "epoch": 2.557029177718833,
+      "grad_norm": 0.06025288814861885,
+      "learning_rate": 6.506955572851059e-07,
+      "loss": 0.0009,
+      "step": 964
+    },
+    {
+      "epoch": 2.5596816976127323,
+      "grad_norm": 2.8245737115251823,
+      "learning_rate": 6.430971538187725e-07,
+      "loss": 0.0015,
+      "step": 965
+    },
+    {
+      "epoch": 2.5623342175066313,
+      "grad_norm": 0.11963108172214902,
+      "learning_rate": 6.355403256503595e-07,
+      "loss": 0.001,
+      "step": 966
+    },
+    {
+      "epoch": 2.5649867374005306,
+      "grad_norm": 0.09183752042364275,
+      "learning_rate": 6.280251448901253e-07,
+      "loss": 0.001,
+      "step": 967
+    },
+    {
+      "epoch": 2.5676392572944295,
+      "grad_norm": 0.054134332400518646,
+      "learning_rate": 6.205516832509089e-07,
+      "loss": 0.0009,
+      "step": 968
+    },
+    {
+      "epoch": 2.570291777188329,
+      "grad_norm": 0.05426296615344519,
+      "learning_rate": 6.131200120474512e-07,
+      "loss": 0.0009,
+      "step": 969
+    },
+    {
+      "epoch": 2.5729442970822283,
+      "grad_norm": 0.05419212137654628,
+      "learning_rate": 6.057302021957113e-07,
+      "loss": 0.0009,
+      "step": 970
+    },
+    {
+      "epoch": 2.575596816976127,
+      "grad_norm": 0.09188409544827594,
+      "learning_rate": 5.983823242121888e-07,
+      "loss": 0.001,
+      "step": 971
+    },
+    {
+      "epoch": 2.5782493368700266,
+      "grad_norm": 0.05510727973349538,
+      "learning_rate": 5.910764482132575e-07,
+      "loss": 0.001,
+      "step": 972
+    },
+    {
+      "epoch": 2.5809018567639255,
+      "grad_norm": 0.05667808363822762,
+      "learning_rate": 5.838126439144875e-07,
+      "loss": 0.001,
+      "step": 973
+    },
+    {
+      "epoch": 2.583554376657825,
+      "grad_norm": 0.05447871452441066,
+      "learning_rate": 5.765909806299863e-07,
+      "loss": 0.0009,
+      "step": 974
+    },
+    {
+      "epoch": 2.586206896551724,
+      "grad_norm": 0.05427308536099464,
+      "learning_rate": 5.694115272717326e-07,
+      "loss": 0.0009,
+      "step": 975
+    },
+    {
+      "epoch": 2.588859416445623,
+      "grad_norm": 2.240544163367182,
+      "learning_rate": 5.622743523489216e-07,
+      "loss": 0.0015,
+      "step": 976
+    },
+    {
+      "epoch": 2.5915119363395225,
+      "grad_norm": 0.055255049579976116,
+      "learning_rate": 5.551795239673146e-07,
+      "loss": 0.0009,
+      "step": 977
+    },
+    {
+      "epoch": 2.594164456233422,
+      "grad_norm": 0.055423684292052425,
+      "learning_rate": 5.481271098285818e-07,
+      "loss": 0.001,
+      "step": 978
+    },
+    {
+      "epoch": 2.5968169761273208,
+      "grad_norm": 0.054617622358565654,
+      "learning_rate": 5.411171772296609e-07,
+      "loss": 0.001,
+      "step": 979
+    },
+    {
+      "epoch": 2.59946949602122,
+      "grad_norm": 0.14999271213831958,
+      "learning_rate": 5.34149793062112e-07,
+      "loss": 0.001,
+      "step": 980
+    },
+    {
+      "epoch": 2.6021220159151195,
+      "grad_norm": 1.167126402354413,
+      "learning_rate": 5.272250238114857e-07,
+      "loss": 0.0011,
+      "step": 981
+    },
+    {
+      "epoch": 2.6047745358090184,
+      "grad_norm": 0.05627867368251513,
+      "learning_rate": 5.203429355566797e-07,
+      "loss": 0.0009,
+      "step": 982
+    },
+    {
+      "epoch": 2.6074270557029178,
+      "grad_norm": 84.0806645690484,
+      "learning_rate": 5.13503593969315e-07,
+      "loss": 0.1598,
+      "step": 983
+    },
+    {
+      "epoch": 2.610079575596817,
+      "grad_norm": 1.1328372073341084,
+      "learning_rate": 5.067070643131056e-07,
+      "loss": 0.0013,
+      "step": 984
+    },
+    {
+      "epoch": 2.612732095490716,
+      "grad_norm": 0.0796073800123807,
+      "learning_rate": 4.999534114432386e-07,
+      "loss": 0.001,
+      "step": 985
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.05328800469497512,
+      "learning_rate": 4.932426998057516e-07,
+      "loss": 0.0009,
+      "step": 986
+    },
+    {
+      "epoch": 2.618037135278515,
+      "grad_norm": 0.05451165857633103,
+      "learning_rate": 4.865749934369224e-07,
+      "loss": 0.0009,
+      "step": 987
+    },
+    {
+      "epoch": 2.6206896551724137,
+      "grad_norm": 0.05664990591190814,
+      "learning_rate": 4.799503559626528e-07,
+      "loss": 0.0009,
+      "step": 988
+    },
+    {
+      "epoch": 2.623342175066313,
+      "grad_norm": 0.3250507408971192,
+      "learning_rate": 4.733688505978673e-07,
+      "loss": 0.001,
+      "step": 989
+    },
+    {
+      "epoch": 2.6259946949602124,
+      "grad_norm": 0.05434258563183708,
+      "learning_rate": 4.668305401459022e-07,
+      "loss": 0.0009,
+      "step": 990
+    },
+    {
+      "epoch": 2.6286472148541113,
+      "grad_norm": 0.061283545784569245,
+      "learning_rate": 4.603354869979165e-07,
+      "loss": 0.0009,
+      "step": 991
+    },
+    {
+      "epoch": 2.6312997347480107,
+      "grad_norm": 0.053706180235464154,
+      "learning_rate": 4.5388375313228595e-07,
+      "loss": 0.0009,
+      "step": 992
+    },
+    {
+      "epoch": 2.63395225464191,
+      "grad_norm": 0.05378315775577338,
+      "learning_rate": 4.4747540011401913e-07,
+      "loss": 0.0009,
+      "step": 993
+    },
+    {
+      "epoch": 2.636604774535809,
+      "grad_norm": 0.22382515301532738,
+      "learning_rate": 4.4111048909416644e-07,
+      "loss": 0.001,
+      "step": 994
+    },
+    {
+      "epoch": 2.6392572944297084,
+      "grad_norm": 0.05566658839487468,
+      "learning_rate": 4.347890808092359e-07,
+      "loss": 0.0009,
+      "step": 995
+    },
+    {
+      "epoch": 2.6419098143236073,
+      "grad_norm": 0.08617493762139619,
+      "learning_rate": 4.2851123558061927e-07,
+      "loss": 0.0009,
+      "step": 996
+    },
+    {
+      "epoch": 2.6445623342175066,
+      "grad_norm": 0.14979392407078365,
+      "learning_rate": 4.2227701331400974e-07,
+      "loss": 0.0009,
+      "step": 997
+    },
+    {
+      "epoch": 2.647214854111406,
+      "grad_norm": 0.2293056663517138,
+      "learning_rate": 4.1608647349883123e-07,
+      "loss": 0.001,
+      "step": 998
+    },
+    {
+      "epoch": 2.649867374005305,
+      "grad_norm": 0.05731999235079868,
+      "learning_rate": 4.0993967520767455e-07,
+      "loss": 0.0009,
+      "step": 999
+    },
+    {
+      "epoch": 2.6525198938992043,
+      "grad_norm": 0.07648364037970812,
+      "learning_rate": 4.0383667709573083e-07,
+      "loss": 0.0009,
+      "step": 1000
+    },
+    {
+      "epoch": 2.655172413793103,
+      "grad_norm": 1.7607923873864444,
+      "learning_rate": 3.9777753740023404e-07,
+      "loss": 0.0021,
+      "step": 1001
+    },
+    {
+      "epoch": 2.6578249336870026,
+      "grad_norm": 0.0775365755949617,
+      "learning_rate": 3.9176231393990183e-07,
+      "loss": 0.0009,
+      "step": 1002
+    },
+    {
+      "epoch": 2.660477453580902,
+      "grad_norm": 0.05311175699372582,
+      "learning_rate": 3.8579106411438636e-07,
+      "loss": 0.0009,
+      "step": 1003
+    },
+    {
+      "epoch": 2.663129973474801,
+      "grad_norm": 0.11581146565084956,
+      "learning_rate": 3.7986384490372395e-07,
+      "loss": 0.0009,
+      "step": 1004
+    },
+    {
+      "epoch": 2.6657824933687,
+      "grad_norm": 0.21597509677221616,
+      "learning_rate": 3.739807128677986e-07,
+      "loss": 0.001,
+      "step": 1005
+    },
+    {
+      "epoch": 2.6684350132625996,
+      "grad_norm": 0.1286558762259199,
+      "learning_rate": 3.6814172414579075e-07,
+      "loss": 0.001,
+      "step": 1006
+    },
+    {
+      "epoch": 2.6710875331564985,
+      "grad_norm": 0.5351619199712117,
+      "learning_rate": 3.6234693445565185e-07,
+      "loss": 0.001,
+      "step": 1007
+    },
+    {
+      "epoch": 2.673740053050398,
+      "grad_norm": 0.05331786914670257,
+      "learning_rate": 3.5659639909356725e-07,
+      "loss": 0.0009,
+      "step": 1008
+    },
+    {
+      "epoch": 2.676392572944297,
+      "grad_norm": 0.06103172326099002,
+      "learning_rate": 3.5089017293342965e-07,
+      "loss": 0.0009,
+      "step": 1009
+    },
+    {
+      "epoch": 2.679045092838196,
+      "grad_norm": 0.06613660996010433,
+      "learning_rate": 3.45228310426316e-07,
+      "loss": 0.0009,
+      "step": 1010
+    },
+    {
+      "epoch": 2.6816976127320955,
+      "grad_norm": 0.0536713030282571,
+      "learning_rate": 3.39610865599968e-07,
+      "loss": 0.0009,
+      "step": 1011
+    },
+    {
+      "epoch": 2.684350132625995,
+      "grad_norm": 1.483789679555465,
+      "learning_rate": 3.34037892058276e-07,
+      "loss": 0.0016,
+      "step": 1012
+    },
+    {
+      "epoch": 2.687002652519894,
+      "grad_norm": 0.12782926147594997,
+      "learning_rate": 3.285094429807673e-07,
+      "loss": 0.0009,
+      "step": 1013
+    },
+    {
+      "epoch": 2.689655172413793,
+      "grad_norm": 0.06823393720694111,
+      "learning_rate": 3.230255711220992e-07,
+      "loss": 0.0009,
+      "step": 1014
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.05299592095410251,
+      "learning_rate": 3.175863288115566e-07,
+      "loss": 0.0009,
+      "step": 1015
+    },
+    {
+      "epoch": 2.6949602122015914,
+      "grad_norm": 0.057694825533013465,
+      "learning_rate": 3.121917679525505e-07,
+      "loss": 0.0009,
+      "step": 1016
+    },
+    {
+      "epoch": 2.697612732095491,
+      "grad_norm": 0.053518700931445305,
+      "learning_rate": 3.0684194002212287e-07,
+      "loss": 0.0009,
+      "step": 1017
+    },
+    {
+      "epoch": 2.70026525198939,
+      "grad_norm": 3.195820329025052,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.0017,
+      "step": 1018
+    },
+    {
+      "epoch": 2.702917771883289,
+      "grad_norm": 0.05446527273973968,
+      "learning_rate": 2.962766867203926e-07,
+      "loss": 0.0009,
+      "step": 1019
+    },
+    {
+      "epoch": 2.7055702917771884,
+      "grad_norm": 1.4468641306749286,
+      "learning_rate": 2.910613621669356e-07,
+      "loss": 0.0013,
+      "step": 1020
+    },
+    {
+      "epoch": 2.708222811671088,
+      "grad_norm": 0.052517541806575904,
+      "learning_rate": 2.8589097217678383e-07,
+      "loss": 0.0009,
+      "step": 1021
+    },
+    {
+      "epoch": 2.7108753315649867,
+      "grad_norm": 0.05317169513735476,
+      "learning_rate": 2.807655660878533e-07,
+      "loss": 0.0009,
+      "step": 1022
+    },
+    {
+      "epoch": 2.713527851458886,
+      "grad_norm": 0.06655482238956349,
+      "learning_rate": 2.756851928088056e-07,
+      "loss": 0.0009,
+      "step": 1023
+    },
+    {
+      "epoch": 2.716180371352785,
+      "grad_norm": 0.053219968452906045,
+      "learning_rate": 2.706499008185798e-07,
+      "loss": 0.0009,
+      "step": 1024
+    },
+    {
+      "epoch": 2.7188328912466844,
+      "grad_norm": 1.2989529905000294,
+      "learning_rate": 2.6565973816593424e-07,
+      "loss": 0.0013,
+      "step": 1025
+    },
+    {
+      "epoch": 2.7214854111405833,
+      "grad_norm": 0.05269752151551146,
+      "learning_rate": 2.607147524689829e-07,
+      "loss": 0.0009,
+      "step": 1026
+    },
+    {
+      "epoch": 2.7241379310344827,
+      "grad_norm": 4.045459700751683,
+      "learning_rate": 2.558149909147434e-07,
+      "loss": 0.0032,
+      "step": 1027
+    },
+    {
+      "epoch": 2.726790450928382,
+      "grad_norm": 0.05492134016720336,
+      "learning_rate": 2.5096050025868734e-07,
+      "loss": 0.0009,
+      "step": 1028
+    },
+    {
+      "epoch": 2.729442970822281,
+      "grad_norm": 0.053646721730707204,
+      "learning_rate": 2.461513268242938e-07,
+      "loss": 0.0009,
+      "step": 1029
+    },
+    {
+      "epoch": 2.7320954907161803,
+      "grad_norm": 0.052914877380319573,
+      "learning_rate": 2.4138751650260585e-07,
+      "loss": 0.0009,
+      "step": 1030
+    },
+    {
+      "epoch": 2.7347480106100797,
+      "grad_norm": 0.07124315478777281,
+      "learning_rate": 2.366691147517941e-07,
+      "loss": 0.0009,
+      "step": 1031
+    },
+    {
+      "epoch": 2.7374005305039786,
+      "grad_norm": 0.05595575110643659,
+      "learning_rate": 2.3199616659672352e-07,
+      "loss": 0.0009,
+      "step": 1032
+    },
+    {
+      "epoch": 2.740053050397878,
+      "grad_norm": 0.05263344966511017,
+      "learning_rate": 2.2736871662852045e-07,
+      "loss": 0.0009,
+      "step": 1033
+    },
+    {
+      "epoch": 2.7427055702917773,
+      "grad_norm": 0.07728000008037111,
+      "learning_rate": 2.2278680900415183e-07,
+      "loss": 0.0009,
+      "step": 1034
+    },
+    {
+      "epoch": 2.7453580901856762,
+      "grad_norm": 12.460780762688533,
+      "learning_rate": 2.1825048744600062e-07,
+      "loss": 0.0047,
+      "step": 1035
+    },
+    {
+      "epoch": 2.7480106100795756,
+      "grad_norm": 0.05353882363770759,
+      "learning_rate": 2.1375979524144942e-07,
+      "loss": 0.0009,
+      "step": 1036
+    },
+    {
+      "epoch": 2.750663129973475,
+      "grad_norm": 0.05303639510725005,
+      "learning_rate": 2.093147752424668e-07,
+      "loss": 0.0009,
+      "step": 1037
+    },
+    {
+      "epoch": 2.753315649867374,
+      "grad_norm": 0.0535875347321289,
+      "learning_rate": 2.0491546986519896e-07,
+      "loss": 0.0009,
+      "step": 1038
+    },
+    {
+      "epoch": 2.7559681697612732,
+      "grad_norm": 0.25982418588180967,
+      "learning_rate": 2.0056192108956762e-07,
+      "loss": 0.001,
+      "step": 1039
+    },
+    {
+      "epoch": 2.7586206896551726,
+      "grad_norm": 3.4179988636523007,
+      "learning_rate": 1.962541704588633e-07,
+      "loss": 0.0013,
+      "step": 1040
+    },
+    {
+      "epoch": 2.7612732095490715,
+      "grad_norm": 0.0531298644994267,
+      "learning_rate": 1.9199225907935492e-07,
+      "loss": 0.0009,
+      "step": 1041
+    },
+    {
+      "epoch": 2.763925729442971,
+      "grad_norm": 0.19402946963760118,
+      "learning_rate": 1.8777622761989355e-07,
+      "loss": 0.0009,
+      "step": 1042
+    },
+    {
+      "epoch": 2.7665782493368702,
+      "grad_norm": 0.05266794784629391,
+      "learning_rate": 1.8360611631152602e-07,
+      "loss": 0.0009,
+      "step": 1043
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.11251991386409466,
+      "learning_rate": 1.794819649471119e-07,
+      "loss": 0.0009,
+      "step": 1044
+    },
+    {
+      "epoch": 2.7718832891246685,
+      "grad_norm": 0.07494694405757234,
+      "learning_rate": 1.7540381288094154e-07,
+      "loss": 0.0009,
+      "step": 1045
+    },
+    {
+      "epoch": 2.774535809018568,
+      "grad_norm": 0.05866658843140271,
+      "learning_rate": 1.7137169902836203e-07,
+      "loss": 0.0009,
+      "step": 1046
+    },
+    {
+      "epoch": 2.777188328912467,
+      "grad_norm": 0.6591494820631454,
+      "learning_rate": 1.6738566186540628e-07,
+      "loss": 0.001,
+      "step": 1047
+    },
+    {
+      "epoch": 2.779840848806366,
+      "grad_norm": 0.053120743227235555,
+      "learning_rate": 1.6344573942842333e-07,
+      "loss": 0.0009,
+      "step": 1048
+    },
+    {
+      "epoch": 2.782493368700265,
+      "grad_norm": 0.05209145348520484,
+      "learning_rate": 1.5955196931371985e-07,
+      "loss": 0.0009,
+      "step": 1049
+    },
+    {
+      "epoch": 2.7851458885941645,
+      "grad_norm": 0.05305398481377058,
+      "learning_rate": 1.5570438867719695e-07,
+      "loss": 0.0009,
+      "step": 1050
+    },
+    {
+      "epoch": 2.787798408488064,
+      "grad_norm": 0.05171852149864696,
+      "learning_rate": 1.5190303423399722e-07,
+      "loss": 0.0009,
+      "step": 1051
+    },
+    {
+      "epoch": 2.7904509283819627,
+      "grad_norm": 0.07918897798668659,
+      "learning_rate": 1.4814794225815443e-07,
+      "loss": 0.0009,
+      "step": 1052
+    },
+    {
+      "epoch": 2.793103448275862,
+      "grad_norm": 93.58069225282338,
+      "learning_rate": 1.4443914858224938e-07,
+      "loss": 0.2561,
+      "step": 1053
+    },
+    {
+      "epoch": 2.795755968169761,
+      "grad_norm": 0.053878150878539866,
+      "learning_rate": 1.4077668859706407e-07,
+      "loss": 0.0009,
+      "step": 1054
+    },
+    {
+      "epoch": 2.7984084880636604,
+      "grad_norm": 0.05257860875770833,
+      "learning_rate": 1.3716059725124687e-07,
+      "loss": 0.0009,
+      "step": 1055
+    },
+    {
+      "epoch": 2.8010610079575597,
+      "grad_norm": 0.08796784767244706,
+      "learning_rate": 1.335909090509785e-07,
+      "loss": 0.0009,
+      "step": 1056
+    },
+    {
+      "epoch": 2.8037135278514587,
+      "grad_norm": 0.055315059528931385,
+      "learning_rate": 1.300676580596405e-07,
+      "loss": 0.0009,
+      "step": 1057
+    },
+    {
+      "epoch": 2.806366047745358,
+      "grad_norm": 0.052615718274800716,
+      "learning_rate": 1.2659087789749557e-07,
+      "loss": 0.0009,
+      "step": 1058
+    },
+    {
+      "epoch": 2.8090185676392574,
+      "grad_norm": 0.8818547244938975,
+      "learning_rate": 1.2316060174136e-07,
+      "loss": 0.0013,
+      "step": 1059
+    },
+    {
+      "epoch": 2.8116710875331563,
+      "grad_norm": 0.05258280180456069,
+      "learning_rate": 1.197768623242923e-07,
+      "loss": 0.0009,
+      "step": 1060
+    },
+    {
+      "epoch": 2.8143236074270557,
+      "grad_norm": 0.05269544838709277,
+      "learning_rate": 1.1643969193527783e-07,
+      "loss": 0.0009,
+      "step": 1061
+    },
+    {
+      "epoch": 2.816976127320955,
+      "grad_norm": 0.055977552732450954,
+      "learning_rate": 1.1314912241892184e-07,
+      "loss": 0.0009,
+      "step": 1062
+    },
+    {
+      "epoch": 2.819628647214854,
+      "grad_norm": 0.4988239402449866,
+      "learning_rate": 1.0990518517514759e-07,
+      "loss": 0.0011,
+      "step": 1063
+    },
+    {
+      "epoch": 2.8222811671087533,
+      "grad_norm": 0.05136956191107987,
+      "learning_rate": 1.0670791115889146e-07,
+      "loss": 0.0009,
+      "step": 1064
+    },
+    {
+      "epoch": 2.8249336870026527,
+      "grad_norm": 0.05252309276464186,
+      "learning_rate": 1.035573308798138e-07,
+      "loss": 0.0009,
+      "step": 1065
+    },
+    {
+      "epoch": 2.8275862068965516,
+      "grad_norm": 0.053164137556096884,
+      "learning_rate": 1.0045347440200192e-07,
+      "loss": 0.0009,
+      "step": 1066
+    },
+    {
+      "epoch": 2.830238726790451,
+      "grad_norm": 0.0520539785498693,
+      "learning_rate": 9.739637134368817e-08,
+      "loss": 0.0009,
+      "step": 1067
+    },
+    {
+      "epoch": 2.8328912466843503,
+      "grad_norm": 0.05131132586881724,
+      "learning_rate": 9.43860508769645e-08,
+      "loss": 0.0009,
+      "step": 1068
+    },
+    {
+      "epoch": 2.8355437665782492,
+      "grad_norm": 0.052547889240969016,
+      "learning_rate": 9.142254172750498e-08,
+      "loss": 0.0009,
+      "step": 1069
+    },
+    {
+      "epoch": 2.8381962864721486,
+      "grad_norm": 0.05356818930213062,
+      "learning_rate": 8.850587217429096e-08,
+      "loss": 0.0009,
+      "step": 1070
+    },
+    {
+      "epoch": 2.840848806366048,
+      "grad_norm": 0.05281190171358196,
+      "learning_rate": 8.563607004934193e-08,
+      "loss": 0.0009,
+      "step": 1071
+    },
+    {
+      "epoch": 2.843501326259947,
+      "grad_norm": 0.21186282791791453,
+      "learning_rate": 8.281316273744955e-08,
+      "loss": 0.001,
+      "step": 1072
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.05324233857203547,
+      "learning_rate": 8.003717717591786e-08,
+      "loss": 0.0009,
+      "step": 1073
+    },
+    {
+      "epoch": 2.8488063660477456,
+      "grad_norm": 0.05163807966434934,
+      "learning_rate": 7.730813985430407e-08,
+      "loss": 0.0009,
+      "step": 1074
+    },
+    {
+      "epoch": 2.8514588859416445,
+      "grad_norm": 0.0890978599720558,
+      "learning_rate": 7.4626076814166e-08,
+      "loss": 0.001,
+      "step": 1075
+    },
+    {
+      "epoch": 2.854111405835544,
+      "grad_norm": 0.053259104192480915,
+      "learning_rate": 7.199101364881389e-08,
+      "loss": 0.0009,
+      "step": 1076
+    },
+    {
+      "epoch": 2.856763925729443,
+      "grad_norm": 0.05231435936161711,
+      "learning_rate": 6.940297550306895e-08,
+      "loss": 0.0009,
+      "step": 1077
+    },
+    {
+      "epoch": 2.859416445623342,
+      "grad_norm": 0.19533030749474406,
+      "learning_rate": 6.686198707301861e-08,
+      "loss": 0.001,
+      "step": 1078
+    },
+    {
+      "epoch": 2.862068965517241,
+      "grad_norm": 18.458776333269714,
+      "learning_rate": 6.436807260578437e-08,
+      "loss": 0.0047,
+      "step": 1079
+    },
+    {
+      "epoch": 2.8647214854111405,
+      "grad_norm": 0.05118343115403007,
+      "learning_rate": 6.192125589928821e-08,
+      "loss": 0.0009,
+      "step": 1080
+    },
+    {
+      "epoch": 2.86737400530504,
+      "grad_norm": 0.05541616336363455,
+      "learning_rate": 5.952156030202716e-08,
+      "loss": 0.0009,
+      "step": 1081
+    },
+    {
+      "epoch": 2.8700265251989387,
+      "grad_norm": 0.056049761657979084,
+      "learning_rate": 5.7169008712851245e-08,
+      "loss": 0.0009,
+      "step": 1082
+    },
+    {
+      "epoch": 2.872679045092838,
+      "grad_norm": 0.06719700505364218,
+      "learning_rate": 5.486362358074093e-08,
+      "loss": 0.0009,
+      "step": 1083
+    },
+    {
+      "epoch": 2.8753315649867375,
+      "grad_norm": 0.05192255208597775,
+      "learning_rate": 5.2605426904598356e-08,
+      "loss": 0.0009,
+      "step": 1084
+    },
+    {
+      "epoch": 2.8779840848806364,
+      "grad_norm": 0.0642612322291816,
+      "learning_rate": 5.0394440233031975e-08,
+      "loss": 0.0009,
+      "step": 1085
+    },
+    {
+      "epoch": 2.8806366047745358,
+      "grad_norm": 0.05383635370381051,
+      "learning_rate": 4.823068466415615e-08,
+      "loss": 0.0009,
+      "step": 1086
+    },
+    {
+      "epoch": 2.883289124668435,
+      "grad_norm": 0.06186926525340373,
+      "learning_rate": 4.611418084538577e-08,
+      "loss": 0.0009,
+      "step": 1087
+    },
+    {
+      "epoch": 2.885941644562334,
+      "grad_norm": 0.05261394296745666,
+      "learning_rate": 4.4044948973240855e-08,
+      "loss": 0.0009,
+      "step": 1088
+    },
+    {
+      "epoch": 2.8885941644562334,
+      "grad_norm": 0.051805840583547105,
+      "learning_rate": 4.202300879315446e-08,
+      "loss": 0.0009,
+      "step": 1089
+    },
+    {
+      "epoch": 2.8912466843501328,
+      "grad_norm": 0.05127305205775271,
+      "learning_rate": 4.004837959928287e-08,
+      "loss": 0.0009,
+      "step": 1090
+    },
+    {
+      "epoch": 2.8938992042440317,
+      "grad_norm": 0.17569307264006073,
+      "learning_rate": 3.8121080234322374e-08,
+      "loss": 0.001,
+      "step": 1091
+    },
+    {
+      "epoch": 2.896551724137931,
+      "grad_norm": 2.0806965078994732,
+      "learning_rate": 3.6241129089329416e-08,
+      "loss": 0.0012,
+      "step": 1092
+    },
+    {
+      "epoch": 2.8992042440318304,
+      "grad_norm": 0.06461716501522924,
+      "learning_rate": 3.4408544103544663e-08,
+      "loss": 0.0009,
+      "step": 1093
+    },
+    {
+      "epoch": 2.9018567639257293,
+      "grad_norm": 0.05172591688297936,
+      "learning_rate": 3.262334276422141e-08,
+      "loss": 0.0009,
+      "step": 1094
+    },
+    {
+      "epoch": 2.9045092838196287,
+      "grad_norm": 0.05436450940189997,
+      "learning_rate": 3.088554210646133e-08,
+      "loss": 0.0009,
+      "step": 1095
+    },
+    {
+      "epoch": 2.907161803713528,
+      "grad_norm": 0.056837362432379836,
+      "learning_rate": 2.9195158713047345e-08,
+      "loss": 0.0009,
+      "step": 1096
+    },
+    {
+      "epoch": 2.909814323607427,
+      "grad_norm": 0.05587010599540712,
+      "learning_rate": 2.7552208714290428e-08,
+      "loss": 0.0009,
+      "step": 1097
+    },
+    {
+      "epoch": 2.9124668435013263,
+      "grad_norm": 0.1892861189366528,
+      "learning_rate": 2.595670778787196e-08,
+      "loss": 0.0009,
+      "step": 1098
+    },
+    {
+      "epoch": 2.9151193633952257,
+      "grad_norm": 0.052539471946618024,
+      "learning_rate": 2.4408671158695495e-08,
+      "loss": 0.0009,
+      "step": 1099
+    },
+    {
+      "epoch": 2.9177718832891246,
+      "grad_norm": 0.0538284361460942,
+      "learning_rate": 2.2908113598741344e-08,
+      "loss": 0.0009,
+      "step": 1100
+    },
+    {
+      "epoch": 2.920424403183024,
+      "grad_norm": 0.05244144456334917,
+      "learning_rate": 2.1455049426926666e-08,
+      "loss": 0.0009,
+      "step": 1101
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.05124729513708506,
+      "learning_rate": 2.004949250896615e-08,
+      "loss": 0.0009,
+      "step": 1102
+    },
+    {
+      "epoch": 2.9257294429708223,
+      "grad_norm": 0.05183879368776961,
+      "learning_rate": 1.8691456257243223e-08,
+      "loss": 0.0009,
+      "step": 1103
+    },
+    {
+      "epoch": 2.9283819628647216,
+      "grad_norm": 0.053332712370574915,
+      "learning_rate": 1.7380953630678488e-08,
+      "loss": 0.0009,
+      "step": 1104
+    },
+    {
+      "epoch": 2.9310344827586206,
+      "grad_norm": 0.05274454609932001,
+      "learning_rate": 1.6117997134609263e-08,
+      "loss": 0.0009,
+      "step": 1105
+    },
+    {
+      "epoch": 2.93368700265252,
+      "grad_norm": 0.051714962632025734,
+      "learning_rate": 1.4902598820668023e-08,
+      "loss": 0.0009,
+      "step": 1106
+    },
+    {
+      "epoch": 2.936339522546419,
+      "grad_norm": 0.05202589931146708,
+      "learning_rate": 1.373477028666803e-08,
+      "loss": 0.0009,
+      "step": 1107
+    },
+    {
+      "epoch": 2.938992042440318,
+      "grad_norm": 0.051866404556823766,
+      "learning_rate": 1.2614522676493435e-08,
+      "loss": 0.0009,
+      "step": 1108
+    },
+    {
+      "epoch": 2.9416445623342176,
+      "grad_norm": 0.08369474899970679,
+      "learning_rate": 1.1541866679992131e-08,
+      "loss": 0.0009,
+      "step": 1109
+    },
+    {
+      "epoch": 2.9442970822281165,
+      "grad_norm": 0.4772825657639168,
+      "learning_rate": 1.0516812532873622e-08,
+      "loss": 0.0009,
+      "step": 1110
+    },
+    {
+      "epoch": 2.946949602122016,
+      "grad_norm": 0.2513499011538888,
+      "learning_rate": 9.53937001661187e-09,
+      "loss": 0.001,
+      "step": 1111
+    },
+    {
+      "epoch": 2.949602122015915,
+      "grad_norm": 0.0737504415082238,
+      "learning_rate": 8.609548458351492e-09,
+      "loss": 0.0009,
+      "step": 1112
+    },
+    {
+      "epoch": 2.952254641909814,
+      "grad_norm": 5.681724126406856,
+      "learning_rate": 7.727356730820035e-09,
+      "loss": 0.0025,
+      "step": 1113
+    },
+    {
+      "epoch": 2.9549071618037135,
+      "grad_norm": 0.052377456451629,
+      "learning_rate": 6.892803252240287e-09,
+      "loss": 0.0009,
+      "step": 1114
+    },
+    {
+      "epoch": 2.957559681697613,
+      "grad_norm": 0.05184413140922385,
+      "learning_rate": 6.105895986253108e-09,
+      "loss": 0.0009,
+      "step": 1115
+    },
+    {
+      "epoch": 2.9602122015915118,
+      "grad_norm": 0.05192498288489938,
+      "learning_rate": 5.366642441841374e-09,
+      "loss": 0.0009,
+      "step": 1116
+    },
+    {
+      "epoch": 2.962864721485411,
+      "grad_norm": 0.05167782759169226,
+      "learning_rate": 4.675049673255605e-09,
+      "loss": 0.0009,
+      "step": 1117
+    },
+    {
+      "epoch": 2.9655172413793105,
+      "grad_norm": 0.051963406241927336,
+      "learning_rate": 4.031124279948451e-09,
+      "loss": 0.0009,
+      "step": 1118
+    },
+    {
+      "epoch": 2.9681697612732094,
+      "grad_norm": 0.05196275371314866,
+      "learning_rate": 3.4348724065119687e-09,
+      "loss": 0.0009,
+      "step": 1119
+    },
+    {
+      "epoch": 2.970822281167109,
+      "grad_norm": 0.05320377195057068,
+      "learning_rate": 2.886299742618226e-09,
+      "loss": 0.0009,
+      "step": 1120
+    },
+    {
+      "epoch": 2.973474801061008,
+      "grad_norm": 0.05179371657100488,
+      "learning_rate": 2.385411522966563e-09,
+      "loss": 0.0009,
+      "step": 1121
+    },
+    {
+      "epoch": 2.976127320954907,
+      "grad_norm": 0.052686275251922716,
+      "learning_rate": 1.9322125272297488e-09,
+      "loss": 0.0009,
+      "step": 1122
+    },
+    {
+      "epoch": 2.9787798408488064,
+      "grad_norm": 0.10193440229554045,
+      "learning_rate": 1.5267070800140116e-09,
+      "loss": 0.0009,
+      "step": 1123
+    },
+    {
+      "epoch": 2.981432360742706,
+      "grad_norm": 0.05257306758022402,
+      "learning_rate": 1.168899050812966e-09,
+      "loss": 0.0009,
+      "step": 1124
+    },
+    {
+      "epoch": 2.9840848806366047,
+      "grad_norm": 0.05177475358407255,
+      "learning_rate": 8.587918539726403e-10,
+      "loss": 0.0009,
+      "step": 1125
+    },
+    {
+      "epoch": 2.986737400530504,
+      "grad_norm": 0.055400243939752354,
+      "learning_rate": 5.963884486598348e-10,
+      "loss": 0.0009,
+      "step": 1126
+    },
+    {
+      "epoch": 2.9893899204244034,
+      "grad_norm": 0.05514716417865198,
+      "learning_rate": 3.816913388315913e-10,
+      "loss": 0.0009,
+      "step": 1127
+    },
+    {
+      "epoch": 2.9920424403183024,
+      "grad_norm": 0.1732782167812435,
+      "learning_rate": 2.1470257321298815e-10,
+      "loss": 0.001,
+      "step": 1128
+    },
+    {
+      "epoch": 2.9946949602122017,
+      "grad_norm": 0.05178211393320526,
+      "learning_rate": 9.54237452771567e-11,
+      "loss": 0.0009,
+      "step": 1129
+    },
+    {
+      "epoch": 2.9973474801061006,
+      "grad_norm": 0.05235516499580651,
+      "learning_rate": 2.3855993230292862e-11,
+      "loss": 0.0009,
+      "step": 1130
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.052436056499771545,
+      "learning_rate": 0.0,
+      "loss": 0.0009,
+      "step": 1131
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.031876802444458,
+      "eval_runtime": 175.8179,
+      "eval_samples_per_second": 12.029,
+      "eval_steps_per_second": 1.507,
+      "step": 1131
+    },
+    {
+      "epoch": 3.0,
+      "step": 1131,
+      "total_flos": 33123426287616.0,
+      "train_loss": 0.5493979175771367,
+      "train_runtime": 4957.8152,
+      "train_samples_per_second": 3.649,
+      "train_steps_per_second": 0.228
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1131,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 33123426287616.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}