diff --git "a/checkpoint-1818/trainer_state.json" "b/checkpoint-1818/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1818/trainer_state.json"
@@ -0,0 +1,12759 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.10006054268259122,
+  "eval_steps": 500,
+  "global_step": 1818,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 5.503880235566074e-05,
+      "grad_norm": 459.8753356933594,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 3.303,
+      "step": 1
+    },
+    {
+      "epoch": 0.00011007760471132149,
+      "grad_norm": 314.2561950683594,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 2.8226,
+      "step": 2
+    },
+    {
+      "epoch": 0.0001651164070669822,
+      "grad_norm": 314.1292419433594,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 2.8517,
+      "step": 3
+    },
+    {
+      "epoch": 0.00022015520942264297,
+      "grad_norm": 312.4049072265625,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 2.6248,
+      "step": 4
+    },
+    {
+      "epoch": 0.0002751940117783037,
+      "grad_norm": 353.7213134765625,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 2.7883,
+      "step": 5
+    },
+    {
+      "epoch": 0.0003302328141339644,
+      "grad_norm": 278.41668701171875,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 2.5468,
+      "step": 6
+    },
+    {
+      "epoch": 0.0003852716164896252,
+      "grad_norm": 336.14532470703125,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 2.7721,
+      "step": 7
+    },
+    {
+      "epoch": 0.00044031041884528595,
+      "grad_norm": 201.19374084472656,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.4873,
+      "step": 8
+    },
+    {
+      "epoch": 0.0004953492212009466,
+      "grad_norm": 184.7027587890625,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 2.6647,
+      "step": 9
+    },
+    {
+      "epoch": 0.0005503880235566074,
+      "grad_norm": 154.597412109375,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.602,
+      "step": 10
+    },
+    {
+      "epoch": 0.0006054268259122681,
+      "grad_norm": 40.47785568237305,
+      "learning_rate": 1.1e-06,
+      "loss": 2.6716,
+      "step": 11
+    },
+    {
+      "epoch": 0.0006604656282679288,
+      "grad_norm": 25.338607788085938,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 2.2631,
+      "step": 12
+    },
+    {
+      "epoch": 0.0007155044306235897,
+      "grad_norm": 24.976919174194336,
+      "learning_rate": 1.3e-06,
+      "loss": 2.3564,
+      "step": 13
+    },
+    {
+      "epoch": 0.0007705432329792504,
+      "grad_norm": 15.239912033081055,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 2.3295,
+      "step": 14
+    },
+    {
+      "epoch": 0.0008255820353349112,
+      "grad_norm": 14.125042915344238,
+      "learning_rate": 1.5e-06,
+      "loss": 2.307,
+      "step": 15
+    },
+    {
+      "epoch": 0.0008806208376905719,
+      "grad_norm": 13.163726806640625,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 2.1493,
+      "step": 16
+    },
+    {
+      "epoch": 0.0009356596400462326,
+      "grad_norm": 8.726515769958496,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 2.0333,
+      "step": 17
+    },
+    {
+      "epoch": 0.0009906984424018933,
+      "grad_norm": 9.072502136230469,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 2.2046,
+      "step": 18
+    },
+    {
+      "epoch": 0.001045737244757554,
+      "grad_norm": 9.412588119506836,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 2.2001,
+      "step": 19
+    },
+    {
+      "epoch": 0.0011007760471132147,
+      "grad_norm": 8.67534065246582,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.7679,
+      "step": 20
+    },
+    {
+      "epoch": 0.0011558148494688755,
+      "grad_norm": 14.015918731689453,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 1.9566,
+      "step": 21
+    },
+    {
+      "epoch": 0.0012108536518245362,
+      "grad_norm": 7.9474687576293945,
+      "learning_rate": 2.2e-06,
+      "loss": 1.9085,
+      "step": 22
+    },
+    {
+      "epoch": 0.001265892454180197,
+      "grad_norm": 6.806368350982666,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 1.7918,
+      "step": 23
+    },
+    {
+      "epoch": 0.0013209312565358577,
+      "grad_norm": 5.3452582359313965,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.8321,
+      "step": 24
+    },
+    {
+      "epoch": 0.0013759700588915184,
+      "grad_norm": 8.744244575500488,
+      "learning_rate": 2.5e-06,
+      "loss": 1.6317,
+      "step": 25
+    },
+    {
+      "epoch": 0.0014310088612471794,
+      "grad_norm": 5.304683685302734,
+      "learning_rate": 2.6e-06,
+      "loss": 1.6846,
+      "step": 26
+    },
+    {
+      "epoch": 0.00148604766360284,
+      "grad_norm": 5.650127410888672,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 1.7449,
+      "step": 27
+    },
+    {
+      "epoch": 0.0015410864659585008,
+      "grad_norm": 5.479269504547119,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.8158,
+      "step": 28
+    },
+    {
+      "epoch": 0.0015961252683141616,
+      "grad_norm": 4.873537063598633,
+      "learning_rate": 2.9e-06,
+      "loss": 1.8015,
+      "step": 29
+    },
+    {
+      "epoch": 0.0016511640706698223,
+      "grad_norm": 4.971101760864258,
+      "learning_rate": 3e-06,
+      "loss": 1.9034,
+      "step": 30
+    },
+    {
+      "epoch": 0.001706202873025483,
+      "grad_norm": 4.407571315765381,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 1.9037,
+      "step": 31
+    },
+    {
+      "epoch": 0.0017612416753811438,
+      "grad_norm": 4.429073810577393,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.6812,
+      "step": 32
+    },
+    {
+      "epoch": 0.0018162804777368045,
+      "grad_norm": 5.16085147857666,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 1.7627,
+      "step": 33
+    },
+    {
+      "epoch": 0.0018713192800924653,
+      "grad_norm": 4.0805768966674805,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.6799,
+      "step": 34
+    },
+    {
+      "epoch": 0.001926358082448126,
+      "grad_norm": 4.548702239990234,
+      "learning_rate": 3.5e-06,
+      "loss": 1.7799,
+      "step": 35
+    },
+    {
+      "epoch": 0.0019813968848037865,
+      "grad_norm": 5.181888580322266,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.8235,
+      "step": 36
+    },
+    {
+      "epoch": 0.0020364356871594475,
+      "grad_norm": 3.9876129627227783,
+      "learning_rate": 3.7e-06,
+      "loss": 1.5999,
+      "step": 37
+    },
+    {
+      "epoch": 0.002091474489515108,
+      "grad_norm": 6.325051307678223,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 1.7499,
+      "step": 38
+    },
+    {
+      "epoch": 0.002146513291870769,
+      "grad_norm": 6.199049949645996,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 1.784,
+      "step": 39
+    },
+    {
+      "epoch": 0.0022015520942264295,
+      "grad_norm": 4.83912992477417,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.8895,
+      "step": 40
+    },
+    {
+      "epoch": 0.0022565908965820904,
+      "grad_norm": 4.515626907348633,
+      "learning_rate": 4.1e-06,
+      "loss": 1.4887,
+      "step": 41
+    },
+    {
+      "epoch": 0.002311629698937751,
+      "grad_norm": 5.032265663146973,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.7324,
+      "step": 42
+    },
+    {
+      "epoch": 0.002366668501293412,
+      "grad_norm": 4.1879048347473145,
+      "learning_rate": 4.3e-06,
+      "loss": 1.4912,
+      "step": 43
+    },
+    {
+      "epoch": 0.0024217073036490724,
+      "grad_norm": 4.128026485443115,
+      "learning_rate": 4.4e-06,
+      "loss": 1.554,
+      "step": 44
+    },
+    {
+      "epoch": 0.0024767461060047334,
+      "grad_norm": 4.527958393096924,
+      "learning_rate": 4.5e-06,
+      "loss": 1.652,
+      "step": 45
+    },
+    {
+      "epoch": 0.002531784908360394,
+      "grad_norm": 4.8388190269470215,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.6696,
+      "step": 46
+    },
+    {
+      "epoch": 0.002586823710716055,
+      "grad_norm": 4.2088541984558105,
+      "learning_rate": 4.7e-06,
+      "loss": 1.568,
+      "step": 47
+    },
+    {
+      "epoch": 0.0026418625130717154,
+      "grad_norm": 4.789997577667236,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.642,
+      "step": 48
+    },
+    {
+      "epoch": 0.0026969013154273763,
+      "grad_norm": 4.408346652984619,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 1.5181,
+      "step": 49
+    },
+    {
+      "epoch": 0.002751940117783037,
+      "grad_norm": 4.572340488433838,
+      "learning_rate": 5e-06,
+      "loss": 1.6698,
+      "step": 50
+    },
+    {
+      "epoch": 0.0028069789201386978,
+      "grad_norm": 4.728564739227295,
+      "learning_rate": 5.1e-06,
+      "loss": 1.5785,
+      "step": 51
+    },
+    {
+      "epoch": 0.0028620177224943587,
+      "grad_norm": 4.449855327606201,
+      "learning_rate": 5.2e-06,
+      "loss": 1.4624,
+      "step": 52
+    },
+    {
+      "epoch": 0.0029170565248500193,
+      "grad_norm": 4.127189636230469,
+      "learning_rate": 5.300000000000001e-06,
+      "loss": 1.6061,
+      "step": 53
+    },
+    {
+      "epoch": 0.00297209532720568,
+      "grad_norm": 4.244532108306885,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 1.491,
+      "step": 54
+    },
+    {
+      "epoch": 0.0030271341295613407,
+      "grad_norm": 3.437682628631592,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.1967,
+      "step": 55
+    },
+    {
+      "epoch": 0.0030821729319170017,
+      "grad_norm": 3.83516788482666,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.4731,
+      "step": 56
+    },
+    {
+      "epoch": 0.003137211734272662,
+      "grad_norm": 3.9108972549438477,
+      "learning_rate": 5.7e-06,
+      "loss": 1.4393,
+      "step": 57
+    },
+    {
+      "epoch": 0.003192250536628323,
+      "grad_norm": 3.5258419513702393,
+      "learning_rate": 5.8e-06,
+      "loss": 1.4206,
+      "step": 58
+    },
+    {
+      "epoch": 0.0032472893389839837,
+      "grad_norm": 4.124903678894043,
+      "learning_rate": 5.9e-06,
+      "loss": 1.4747,
+      "step": 59
+    },
+    {
+      "epoch": 0.0033023281413396446,
+      "grad_norm": 4.055769920349121,
+      "learning_rate": 6e-06,
+      "loss": 1.4655,
+      "step": 60
+    },
+    {
+      "epoch": 0.003357366943695305,
+      "grad_norm": 3.904837131500244,
+      "learning_rate": 6.1e-06,
+      "loss": 1.5125,
+      "step": 61
+    },
+    {
+      "epoch": 0.003412405746050966,
+      "grad_norm": 3.2904794216156006,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 1.4596,
+      "step": 62
+    },
+    {
+      "epoch": 0.0034674445484066266,
+      "grad_norm": 3.24053692817688,
+      "learning_rate": 6.300000000000001e-06,
+      "loss": 1.3851,
+      "step": 63
+    },
+    {
+      "epoch": 0.0035224833507622876,
+      "grad_norm": 3.457639217376709,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.4019,
+      "step": 64
+    },
+    {
+      "epoch": 0.003577522153117948,
+      "grad_norm": 3.073054790496826,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.2872,
+      "step": 65
+    },
+    {
+      "epoch": 0.003632560955473609,
+      "grad_norm": 2.6726694107055664,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 1.2361,
+      "step": 66
+    },
+    {
+      "epoch": 0.0036875997578292696,
+      "grad_norm": 2.9378459453582764,
+      "learning_rate": 6.700000000000001e-06,
+      "loss": 1.4452,
+      "step": 67
+    },
+    {
+      "epoch": 0.0037426385601849305,
+      "grad_norm": 2.81107234954834,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.4804,
+      "step": 68
+    },
+    {
+      "epoch": 0.003797677362540591,
+      "grad_norm": 2.60062313079834,
+      "learning_rate": 6.9e-06,
+      "loss": 1.3263,
+      "step": 69
+    },
+    {
+      "epoch": 0.003852716164896252,
+      "grad_norm": 2.5642921924591064,
+      "learning_rate": 7e-06,
+      "loss": 1.2751,
+      "step": 70
+    },
+    {
+      "epoch": 0.0039077549672519125,
+      "grad_norm": 2.3608031272888184,
+      "learning_rate": 7.100000000000001e-06,
+      "loss": 1.2614,
+      "step": 71
+    },
+    {
+      "epoch": 0.003962793769607573,
+      "grad_norm": 2.7201738357543945,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.5018,
+      "step": 72
+    },
+    {
+      "epoch": 0.004017832571963234,
+      "grad_norm": 2.584726095199585,
+      "learning_rate": 7.3e-06,
+      "loss": 1.3519,
+      "step": 73
+    },
+    {
+      "epoch": 0.004072871374318895,
+      "grad_norm": 1.9693044424057007,
+      "learning_rate": 7.4e-06,
+      "loss": 1.0934,
+      "step": 74
+    },
+    {
+      "epoch": 0.0041279101766745555,
+      "grad_norm": 2.220736503601074,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.4687,
+      "step": 75
+    },
+    {
+      "epoch": 0.004182948979030216,
+      "grad_norm": 2.2629456520080566,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.3328,
+      "step": 76
+    },
+    {
+      "epoch": 0.004237987781385877,
+      "grad_norm": 2.051820993423462,
+      "learning_rate": 7.7e-06,
+      "loss": 1.3058,
+      "step": 77
+    },
+    {
+      "epoch": 0.004293026583741538,
+      "grad_norm": 2.2451820373535156,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 1.3556,
+      "step": 78
+    },
+    {
+      "epoch": 0.004348065386097198,
+      "grad_norm": 3.13584303855896,
+      "learning_rate": 7.9e-06,
+      "loss": 1.3262,
+      "step": 79
+    },
+    {
+      "epoch": 0.004403104188452859,
+      "grad_norm": 5.024479866027832,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.2103,
+      "step": 80
+    },
+    {
+      "epoch": 0.00445814299080852,
+      "grad_norm": 2.070889711380005,
+      "learning_rate": 8.1e-06,
+      "loss": 1.1994,
+      "step": 81
+    },
+    {
+      "epoch": 0.004513181793164181,
+      "grad_norm": 2.797286033630371,
+      "learning_rate": 8.2e-06,
+      "loss": 1.3075,
+      "step": 82
+    },
+    {
+      "epoch": 0.004568220595519841,
+      "grad_norm": 2.11370849609375,
+      "learning_rate": 8.3e-06,
+      "loss": 1.36,
+      "step": 83
+    },
+    {
+      "epoch": 0.004623259397875502,
+      "grad_norm": 2.5416152477264404,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.3484,
+      "step": 84
+    },
+    {
+      "epoch": 0.004678298200231163,
+      "grad_norm": 2.4702343940734863,
+      "learning_rate": 8.5e-06,
+      "loss": 1.3677,
+      "step": 85
+    },
+    {
+      "epoch": 0.004733337002586824,
+      "grad_norm": 3.670365333557129,
+      "learning_rate": 8.6e-06,
+      "loss": 1.2192,
+      "step": 86
+    },
+    {
+      "epoch": 0.004788375804942484,
+      "grad_norm": 2.282954692840576,
+      "learning_rate": 8.700000000000001e-06,
+      "loss": 1.2982,
+      "step": 87
+    },
+    {
+      "epoch": 0.004843414607298145,
+      "grad_norm": 2.3659238815307617,
+      "learning_rate": 8.8e-06,
+      "loss": 1.3206,
+      "step": 88
+    },
+    {
+      "epoch": 0.004898453409653806,
+      "grad_norm": 4.939981460571289,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 1.4328,
+      "step": 89
+    },
+    {
+      "epoch": 0.004953492212009467,
+      "grad_norm": 2.335858106613159,
+      "learning_rate": 9e-06,
+      "loss": 1.2603,
+      "step": 90
+    },
+    {
+      "epoch": 0.005008531014365127,
+      "grad_norm": 2.2165043354034424,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 1.3141,
+      "step": 91
+    },
+    {
+      "epoch": 0.005063569816720788,
+      "grad_norm": 2.7872185707092285,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 1.3314,
+      "step": 92
+    },
+    {
+      "epoch": 0.005118608619076449,
+      "grad_norm": 2.6353912353515625,
+      "learning_rate": 9.3e-06,
+      "loss": 1.2027,
+      "step": 93
+    },
+    {
+      "epoch": 0.00517364742143211,
+      "grad_norm": 3.2509102821350098,
+      "learning_rate": 9.4e-06,
+      "loss": 1.2316,
+      "step": 94
+    },
+    {
+      "epoch": 0.00522868622378777,
+      "grad_norm": 2.4560611248016357,
+      "learning_rate": 9.5e-06,
+      "loss": 1.1848,
+      "step": 95
+    },
+    {
+      "epoch": 0.005283725026143431,
+      "grad_norm": 2.338151216506958,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.2392,
+      "step": 96
+    },
+    {
+      "epoch": 0.005338763828499092,
+      "grad_norm": 2.231065034866333,
+      "learning_rate": 9.7e-06,
+      "loss": 1.2089,
+      "step": 97
+    },
+    {
+      "epoch": 0.005393802630854753,
+      "grad_norm": 2.278428077697754,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.2267,
+      "step": 98
+    },
+    {
+      "epoch": 0.005448841433210413,
+      "grad_norm": 2.4422810077667236,
+      "learning_rate": 9.9e-06,
+      "loss": 1.2041,
+      "step": 99
+    },
+    {
+      "epoch": 0.005503880235566074,
+      "grad_norm": 2.216248035430908,
+      "learning_rate": 1e-05,
+      "loss": 1.0798,
+      "step": 100
+    },
+    {
+      "epoch": 0.005558919037921735,
+      "grad_norm": 2.3301615715026855,
+      "learning_rate": 9.99999998121067e-06,
+      "loss": 1.3069,
+      "step": 101
+    },
+    {
+      "epoch": 0.0056139578402773956,
+      "grad_norm": 2.315436363220215,
+      "learning_rate": 9.999999924842678e-06,
+      "loss": 1.1589,
+      "step": 102
+    },
+    {
+      "epoch": 0.005668996642633056,
+      "grad_norm": 2.3522140979766846,
+      "learning_rate": 9.999999830896024e-06,
+      "loss": 1.0978,
+      "step": 103
+    },
+    {
+      "epoch": 0.0057240354449887175,
+      "grad_norm": 2.5798308849334717,
+      "learning_rate": 9.99999969937071e-06,
+      "loss": 1.0599,
+      "step": 104
+    },
+    {
+      "epoch": 0.005779074247344378,
+      "grad_norm": 2.456644058227539,
+      "learning_rate": 9.999999530266738e-06,
+      "loss": 1.1682,
+      "step": 105
+    },
+    {
+      "epoch": 0.0058341130497000385,
+      "grad_norm": 2.1559031009674072,
+      "learning_rate": 9.999999323584106e-06,
+      "loss": 1.0631,
+      "step": 106
+    },
+    {
+      "epoch": 0.005889151852055699,
+      "grad_norm": 2.2985048294067383,
+      "learning_rate": 9.99999907932282e-06,
+      "loss": 1.1455,
+      "step": 107
+    },
+    {
+      "epoch": 0.00594419065441136,
+      "grad_norm": 2.596167802810669,
+      "learning_rate": 9.999998797482877e-06,
+      "loss": 1.1686,
+      "step": 108
+    },
+    {
+      "epoch": 0.005999229456767021,
+      "grad_norm": 2.378618001937866,
+      "learning_rate": 9.999998478064283e-06,
+      "loss": 1.2226,
+      "step": 109
+    },
+    {
+      "epoch": 0.0060542682591226814,
+      "grad_norm": 2.228116750717163,
+      "learning_rate": 9.999998121067038e-06,
+      "loss": 1.1396,
+      "step": 110
+    },
+    {
+      "epoch": 0.006109307061478342,
+      "grad_norm": 2.4419472217559814,
+      "learning_rate": 9.999997726491146e-06,
+      "loss": 1.1401,
+      "step": 111
+    },
+    {
+      "epoch": 0.006164345863834003,
+      "grad_norm": 2.0695526599884033,
+      "learning_rate": 9.999997294336608e-06,
+      "loss": 1.1868,
+      "step": 112
+    },
+    {
+      "epoch": 0.006219384666189664,
+      "grad_norm": 2.3170363903045654,
+      "learning_rate": 9.99999682460343e-06,
+      "loss": 1.1172,
+      "step": 113
+    },
+    {
+      "epoch": 0.006274423468545324,
+      "grad_norm": 2.670466184616089,
+      "learning_rate": 9.999996317291615e-06,
+      "loss": 1.2481,
+      "step": 114
+    },
+    {
+      "epoch": 0.006329462270900985,
+      "grad_norm": 2.1214540004730225,
+      "learning_rate": 9.999995772401166e-06,
+      "loss": 0.9994,
+      "step": 115
+    },
+    {
+      "epoch": 0.006384501073256646,
+      "grad_norm": 1.9283969402313232,
+      "learning_rate": 9.999995189932085e-06,
+      "loss": 1.0692,
+      "step": 116
+    },
+    {
+      "epoch": 0.006439539875612307,
+      "grad_norm": 2.2620882987976074,
+      "learning_rate": 9.99999456988438e-06,
+      "loss": 1.0725,
+      "step": 117
+    },
+    {
+      "epoch": 0.006494578677967967,
+      "grad_norm": 2.2121341228485107,
+      "learning_rate": 9.999993912258055e-06,
+      "loss": 1.1328,
+      "step": 118
+    },
+    {
+      "epoch": 0.006549617480323628,
+      "grad_norm": 2.298126220703125,
+      "learning_rate": 9.999993217053113e-06,
+      "loss": 1.1272,
+      "step": 119
+    },
+    {
+      "epoch": 0.006604656282679289,
+      "grad_norm": 1.81593656539917,
+      "learning_rate": 9.99999248426956e-06,
+      "loss": 1.017,
+      "step": 120
+    },
+    {
+      "epoch": 0.00665969508503495,
+      "grad_norm": 2.1174378395080566,
+      "learning_rate": 9.999991713907403e-06,
+      "loss": 1.0557,
+      "step": 121
+    },
+    {
+      "epoch": 0.00671473388739061,
+      "grad_norm": 1.9061017036437988,
+      "learning_rate": 9.999990905966647e-06,
+      "loss": 1.0379,
+      "step": 122
+    },
+    {
+      "epoch": 0.006769772689746271,
+      "grad_norm": 1.912500023841858,
+      "learning_rate": 9.999990060447297e-06,
+      "loss": 1.104,
+      "step": 123
+    },
+    {
+      "epoch": 0.006824811492101932,
+      "grad_norm": 1.9249529838562012,
+      "learning_rate": 9.99998917734936e-06,
+      "loss": 1.0136,
+      "step": 124
+    },
+    {
+      "epoch": 0.006879850294457593,
+      "grad_norm": 1.8504948616027832,
+      "learning_rate": 9.999988256672843e-06,
+      "loss": 0.99,
+      "step": 125
+    },
+    {
+      "epoch": 0.006934889096813253,
+      "grad_norm": 1.720042109489441,
+      "learning_rate": 9.999987298417753e-06,
+      "loss": 1.0666,
+      "step": 126
+    },
+    {
+      "epoch": 0.006989927899168914,
+      "grad_norm": 1.778251051902771,
+      "learning_rate": 9.999986302584097e-06,
+      "loss": 1.0424,
+      "step": 127
+    },
+    {
+      "epoch": 0.007044966701524575,
+      "grad_norm": 1.9485961198806763,
+      "learning_rate": 9.999985269171881e-06,
+      "loss": 1.105,
+      "step": 128
+    },
+    {
+      "epoch": 0.007100005503880236,
+      "grad_norm": 3.0802104473114014,
+      "learning_rate": 9.999984198181114e-06,
+      "loss": 1.1081,
+      "step": 129
+    },
+    {
+      "epoch": 0.007155044306235896,
+      "grad_norm": 1.7476954460144043,
+      "learning_rate": 9.999983089611806e-06,
+      "loss": 0.9677,
+      "step": 130
+    },
+    {
+      "epoch": 0.007210083108591557,
+      "grad_norm": 1.6127299070358276,
+      "learning_rate": 9.999981943463963e-06,
+      "loss": 0.9937,
+      "step": 131
+    },
+    {
+      "epoch": 0.007265121910947218,
+      "grad_norm": 2.1477208137512207,
+      "learning_rate": 9.999980759737594e-06,
+      "loss": 1.0319,
+      "step": 132
+    },
+    {
+      "epoch": 0.007320160713302879,
+      "grad_norm": 1.531163215637207,
+      "learning_rate": 9.999979538432707e-06,
+      "loss": 0.8696,
+      "step": 133
+    },
+    {
+      "epoch": 0.007375199515658539,
+      "grad_norm": 1.8226820230484009,
+      "learning_rate": 9.999978279549313e-06,
+      "loss": 1.2061,
+      "step": 134
+    },
+    {
+      "epoch": 0.0074302383180142,
+      "grad_norm": 1.481895923614502,
+      "learning_rate": 9.99997698308742e-06,
+      "loss": 0.949,
+      "step": 135
+    },
+    {
+      "epoch": 0.007485277120369861,
+      "grad_norm": 1.6715927124023438,
+      "learning_rate": 9.99997564904704e-06,
+      "loss": 1.1579,
+      "step": 136
+    },
+    {
+      "epoch": 0.0075403159227255215,
+      "grad_norm": 1.4235272407531738,
+      "learning_rate": 9.999974277428179e-06,
+      "loss": 1.064,
+      "step": 137
+    },
+    {
+      "epoch": 0.007595354725081182,
+      "grad_norm": 1.3524872064590454,
+      "learning_rate": 9.999972868230852e-06,
+      "loss": 0.9141,
+      "step": 138
+    },
+    {
+      "epoch": 0.007650393527436843,
+      "grad_norm": 1.3741765022277832,
+      "learning_rate": 9.999971421455066e-06,
+      "loss": 1.0256,
+      "step": 139
+    },
+    {
+      "epoch": 0.007705432329792504,
+      "grad_norm": 1.9869598150253296,
+      "learning_rate": 9.999969937100835e-06,
+      "loss": 0.9489,
+      "step": 140
+    },
+    {
+      "epoch": 0.0077604711321481645,
+      "grad_norm": 1.4785465002059937,
+      "learning_rate": 9.999968415168166e-06,
+      "loss": 0.9243,
+      "step": 141
+    },
+    {
+      "epoch": 0.007815509934503825,
+      "grad_norm": 1.5476176738739014,
+      "learning_rate": 9.999966855657074e-06,
+      "loss": 1.178,
+      "step": 142
+    },
+    {
+      "epoch": 0.007870548736859486,
+      "grad_norm": 1.500401258468628,
+      "learning_rate": 9.99996525856757e-06,
+      "loss": 0.9837,
+      "step": 143
+    },
+    {
+      "epoch": 0.007925587539215146,
+      "grad_norm": 1.3777157068252563,
+      "learning_rate": 9.999963623899664e-06,
+      "loss": 1.0732,
+      "step": 144
+    },
+    {
+      "epoch": 0.007980626341570807,
+      "grad_norm": 1.4466841220855713,
+      "learning_rate": 9.99996195165337e-06,
+      "loss": 0.9779,
+      "step": 145
+    },
+    {
+      "epoch": 0.008035665143926469,
+      "grad_norm": 1.5304051637649536,
+      "learning_rate": 9.9999602418287e-06,
+      "loss": 1.196,
+      "step": 146
+    },
+    {
+      "epoch": 0.008090703946282128,
+      "grad_norm": 1.9012362957000732,
+      "learning_rate": 9.99995849442567e-06,
+      "loss": 0.9797,
+      "step": 147
+    },
+    {
+      "epoch": 0.00814574274863779,
+      "grad_norm": 1.430679202079773,
+      "learning_rate": 9.999956709444289e-06,
+      "loss": 0.9869,
+      "step": 148
+    },
+    {
+      "epoch": 0.00820078155099345,
+      "grad_norm": 1.3489817380905151,
+      "learning_rate": 9.99995488688457e-06,
+      "loss": 1.0137,
+      "step": 149
+    },
+    {
+      "epoch": 0.008255820353349111,
+      "grad_norm": 1.1878125667572021,
+      "learning_rate": 9.999953026746531e-06,
+      "loss": 0.9355,
+      "step": 150
+    },
+    {
+      "epoch": 0.008310859155704772,
+      "grad_norm": 1.3481942415237427,
+      "learning_rate": 9.999951129030182e-06,
+      "loss": 1.1235,
+      "step": 151
+    },
+    {
+      "epoch": 0.008365897958060432,
+      "grad_norm": 1.7335314750671387,
+      "learning_rate": 9.999949193735539e-06,
+      "loss": 0.9382,
+      "step": 152
+    },
+    {
+      "epoch": 0.008420936760416093,
+      "grad_norm": 1.2029480934143066,
+      "learning_rate": 9.999947220862615e-06,
+      "loss": 0.9419,
+      "step": 153
+    },
+    {
+      "epoch": 0.008475975562771755,
+      "grad_norm": 1.2104203701019287,
+      "learning_rate": 9.999945210411428e-06,
+      "loss": 0.9196,
+      "step": 154
+    },
+    {
+      "epoch": 0.008531014365127414,
+      "grad_norm": 1.1857126951217651,
+      "learning_rate": 9.999943162381991e-06,
+      "loss": 0.9421,
+      "step": 155
+    },
+    {
+      "epoch": 0.008586053167483076,
+      "grad_norm": 1.115027904510498,
+      "learning_rate": 9.999941076774319e-06,
+      "loss": 0.9634,
+      "step": 156
+    },
+    {
+      "epoch": 0.008641091969838737,
+      "grad_norm": 1.4227553606033325,
+      "learning_rate": 9.999938953588428e-06,
+      "loss": 1.0036,
+      "step": 157
+    },
+    {
+      "epoch": 0.008696130772194397,
+      "grad_norm": 1.2913776636123657,
+      "learning_rate": 9.999936792824334e-06,
+      "loss": 0.9232,
+      "step": 158
+    },
+    {
+      "epoch": 0.008751169574550058,
+      "grad_norm": 1.2817318439483643,
+      "learning_rate": 9.999934594482055e-06,
+      "loss": 0.9691,
+      "step": 159
+    },
+    {
+      "epoch": 0.008806208376905718,
+      "grad_norm": 1.5647841691970825,
+      "learning_rate": 9.999932358561604e-06,
+      "loss": 1.1842,
+      "step": 160
+    },
+    {
+      "epoch": 0.00886124717926138,
+      "grad_norm": 1.368135929107666,
+      "learning_rate": 9.999930085063002e-06,
+      "loss": 1.0873,
+      "step": 161
+    },
+    {
+      "epoch": 0.00891628598161704,
+      "grad_norm": 1.2297240495681763,
+      "learning_rate": 9.999927773986262e-06,
+      "loss": 1.0778,
+      "step": 162
+    },
+    {
+      "epoch": 0.0089713247839727,
+      "grad_norm": 1.0658279657363892,
+      "learning_rate": 9.999925425331405e-06,
+      "loss": 0.9008,
+      "step": 163
+    },
+    {
+      "epoch": 0.009026363586328362,
+      "grad_norm": 1.3484326601028442,
+      "learning_rate": 9.999923039098445e-06,
+      "loss": 1.0664,
+      "step": 164
+    },
+    {
+      "epoch": 0.009081402388684023,
+      "grad_norm": 1.1839075088500977,
+      "learning_rate": 9.999920615287401e-06,
+      "loss": 0.9257,
+      "step": 165
+    },
+    {
+      "epoch": 0.009136441191039683,
+      "grad_norm": 1.2757254838943481,
+      "learning_rate": 9.999918153898295e-06,
+      "loss": 0.9473,
+      "step": 166
+    },
+    {
+      "epoch": 0.009191479993395344,
+      "grad_norm": 1.2414579391479492,
+      "learning_rate": 9.99991565493114e-06,
+      "loss": 1.1091,
+      "step": 167
+    },
+    {
+      "epoch": 0.009246518795751004,
+      "grad_norm": 1.2802611589431763,
+      "learning_rate": 9.999913118385959e-06,
+      "loss": 1.063,
+      "step": 168
+    },
+    {
+      "epoch": 0.009301557598106665,
+      "grad_norm": 1.2055327892303467,
+      "learning_rate": 9.99991054426277e-06,
+      "loss": 0.8,
+      "step": 169
+    },
+    {
+      "epoch": 0.009356596400462327,
+      "grad_norm": 1.0391098260879517,
+      "learning_rate": 9.99990793256159e-06,
+      "loss": 0.8672,
+      "step": 170
+    },
+    {
+      "epoch": 0.009411635202817986,
+      "grad_norm": 1.131536602973938,
+      "learning_rate": 9.99990528328244e-06,
+      "loss": 0.9569,
+      "step": 171
+    },
+    {
+      "epoch": 0.009466674005173648,
+      "grad_norm": 1.164307951927185,
+      "learning_rate": 9.999902596425342e-06,
+      "loss": 0.9999,
+      "step": 172
+    },
+    {
+      "epoch": 0.009521712807529309,
+      "grad_norm": 1.2099504470825195,
+      "learning_rate": 9.999899871990313e-06,
+      "loss": 0.9994,
+      "step": 173
+    },
+    {
+      "epoch": 0.009576751609884969,
+      "grad_norm": 1.7294539213180542,
+      "learning_rate": 9.999897109977376e-06,
+      "loss": 1.0265,
+      "step": 174
+    },
+    {
+      "epoch": 0.00963179041224063,
+      "grad_norm": 1.3009883165359497,
+      "learning_rate": 9.99989431038655e-06,
+      "loss": 0.9022,
+      "step": 175
+    },
+    {
+      "epoch": 0.00968682921459629,
+      "grad_norm": 1.1014611721038818,
+      "learning_rate": 9.999891473217857e-06,
+      "loss": 0.8476,
+      "step": 176
+    },
+    {
+      "epoch": 0.009741868016951951,
+      "grad_norm": 1.2410900592803955,
+      "learning_rate": 9.99988859847132e-06,
+      "loss": 1.0272,
+      "step": 177
+    },
+    {
+      "epoch": 0.009796906819307612,
+      "grad_norm": 1.336348295211792,
+      "learning_rate": 9.999885686146957e-06,
+      "loss": 0.9456,
+      "step": 178
+    },
+    {
+      "epoch": 0.009851945621663272,
+      "grad_norm": 1.2931095361709595,
+      "learning_rate": 9.99988273624479e-06,
+      "loss": 0.9554,
+      "step": 179
+    },
+    {
+      "epoch": 0.009906984424018933,
+      "grad_norm": 1.2647838592529297,
+      "learning_rate": 9.999879748764845e-06,
+      "loss": 1.0394,
+      "step": 180
+    },
+    {
+      "epoch": 0.009962023226374595,
+      "grad_norm": 1.3485127687454224,
+      "learning_rate": 9.99987672370714e-06,
+      "loss": 1.1016,
+      "step": 181
+    },
+    {
+      "epoch": 0.010017062028730254,
+      "grad_norm": 1.110187292098999,
+      "learning_rate": 9.999873661071702e-06,
+      "loss": 0.946,
+      "step": 182
+    },
+    {
+      "epoch": 0.010072100831085916,
+      "grad_norm": 1.0991623401641846,
+      "learning_rate": 9.999870560858551e-06,
+      "loss": 1.0084,
+      "step": 183
+    },
+    {
+      "epoch": 0.010127139633441576,
+      "grad_norm": 1.049804449081421,
+      "learning_rate": 9.999867423067713e-06,
+      "loss": 0.8264,
+      "step": 184
+    },
+    {
+      "epoch": 0.010182178435797237,
+      "grad_norm": 1.0947058200836182,
+      "learning_rate": 9.999864247699207e-06,
+      "loss": 0.8884,
+      "step": 185
+    },
+    {
+      "epoch": 0.010237217238152898,
+      "grad_norm": 1.1147902011871338,
+      "learning_rate": 9.999861034753061e-06,
+      "loss": 0.9657,
+      "step": 186
+    },
+    {
+      "epoch": 0.010292256040508558,
+      "grad_norm": 1.260027527809143,
+      "learning_rate": 9.999857784229298e-06,
+      "loss": 1.0102,
+      "step": 187
+    },
+    {
+      "epoch": 0.01034729484286422,
+      "grad_norm": 1.1275582313537598,
+      "learning_rate": 9.999854496127942e-06,
+      "loss": 1.028,
+      "step": 188
+    },
+    {
+      "epoch": 0.01040233364521988,
+      "grad_norm": 1.1377174854278564,
+      "learning_rate": 9.999851170449018e-06,
+      "loss": 1.032,
+      "step": 189
+    },
+    {
+      "epoch": 0.01045737244757554,
+      "grad_norm": 1.1734225749969482,
+      "learning_rate": 9.999847807192552e-06,
+      "loss": 1.0009,
+      "step": 190
+    },
+    {
+      "epoch": 0.010512411249931202,
+      "grad_norm": 1.1934596300125122,
+      "learning_rate": 9.999844406358565e-06,
+      "loss": 1.0432,
+      "step": 191
+    },
+    {
+      "epoch": 0.010567450052286861,
+      "grad_norm": 1.0638024806976318,
+      "learning_rate": 9.99984096794709e-06,
+      "loss": 0.8651,
+      "step": 192
+    },
+    {
+      "epoch": 0.010622488854642523,
+      "grad_norm": 1.2381829023361206,
+      "learning_rate": 9.999837491958147e-06,
+      "loss": 1.0088,
+      "step": 193
+    },
+    {
+      "epoch": 0.010677527656998184,
+      "grad_norm": 1.030246615409851,
+      "learning_rate": 9.999833978391763e-06,
+      "loss": 0.9488,
+      "step": 194
+    },
+    {
+      "epoch": 0.010732566459353844,
+      "grad_norm": 1.1640657186508179,
+      "learning_rate": 9.999830427247965e-06,
+      "loss": 1.0588,
+      "step": 195
+    },
+    {
+      "epoch": 0.010787605261709505,
+      "grad_norm": 1.0431616306304932,
+      "learning_rate": 9.99982683852678e-06,
+      "loss": 0.8728,
+      "step": 196
+    },
+    {
+      "epoch": 0.010842644064065167,
+      "grad_norm": 1.032263159751892,
+      "learning_rate": 9.999823212228235e-06,
+      "loss": 0.9498,
+      "step": 197
+    },
+    {
+      "epoch": 0.010897682866420826,
+      "grad_norm": 1.1383745670318604,
+      "learning_rate": 9.999819548352358e-06,
+      "loss": 0.9498,
+      "step": 198
+    },
+    {
+      "epoch": 0.010952721668776488,
+      "grad_norm": 1.1324639320373535,
+      "learning_rate": 9.999815846899175e-06,
+      "loss": 1.0432,
+      "step": 199
+    },
+    {
+      "epoch": 0.011007760471132147,
+      "grad_norm": 1.188672661781311,
+      "learning_rate": 9.999812107868714e-06,
+      "loss": 0.982,
+      "step": 200
+    },
+    {
+      "epoch": 0.011062799273487809,
+      "grad_norm": 1.1011098623275757,
+      "learning_rate": 9.999808331261005e-06,
+      "loss": 0.9587,
+      "step": 201
+    },
+    {
+      "epoch": 0.01111783807584347,
+      "grad_norm": 1.1782938241958618,
+      "learning_rate": 9.999804517076073e-06,
+      "loss": 1.0659,
+      "step": 202
+    },
+    {
+      "epoch": 0.01117287687819913,
+      "grad_norm": 1.0520117282867432,
+      "learning_rate": 9.99980066531395e-06,
+      "loss": 1.0056,
+      "step": 203
+    },
+    {
+      "epoch": 0.011227915680554791,
+      "grad_norm": 1.1584919691085815,
+      "learning_rate": 9.999796775974663e-06,
+      "loss": 0.9435,
+      "step": 204
+    },
+    {
+      "epoch": 0.011282954482910452,
+      "grad_norm": 1.2201849222183228,
+      "learning_rate": 9.999792849058242e-06,
+      "loss": 1.0562,
+      "step": 205
+    },
+    {
+      "epoch": 0.011337993285266112,
+      "grad_norm": 1.2985976934432983,
+      "learning_rate": 9.999788884564715e-06,
+      "loss": 1.0126,
+      "step": 206
+    },
+    {
+      "epoch": 0.011393032087621774,
+      "grad_norm": 0.9926307201385498,
+      "learning_rate": 9.999784882494115e-06,
+      "loss": 0.7875,
+      "step": 207
+    },
+    {
+      "epoch": 0.011448070889977435,
+      "grad_norm": 1.103365182876587,
+      "learning_rate": 9.99978084284647e-06,
+      "loss": 0.9833,
+      "step": 208
+    },
+    {
+      "epoch": 0.011503109692333095,
+      "grad_norm": 1.1798462867736816,
+      "learning_rate": 9.99977676562181e-06,
+      "loss": 0.8479,
+      "step": 209
+    },
+    {
+      "epoch": 0.011558148494688756,
+      "grad_norm": 1.2887194156646729,
+      "learning_rate": 9.999772650820168e-06,
+      "loss": 0.9606,
+      "step": 210
+    },
+    {
+      "epoch": 0.011613187297044416,
+      "grad_norm": 1.1120634078979492,
+      "learning_rate": 9.99976849844157e-06,
+      "loss": 0.9604,
+      "step": 211
+    },
+    {
+      "epoch": 0.011668226099400077,
+      "grad_norm": 1.1248979568481445,
+      "learning_rate": 9.999764308486052e-06,
+      "loss": 0.9428,
+      "step": 212
+    },
+    {
+      "epoch": 0.011723264901755738,
+      "grad_norm": 1.274610161781311,
+      "learning_rate": 9.999760080953643e-06,
+      "loss": 0.9044,
+      "step": 213
+    },
+    {
+      "epoch": 0.011778303704111398,
+      "grad_norm": 1.1746865510940552,
+      "learning_rate": 9.999755815844377e-06,
+      "loss": 0.9114,
+      "step": 214
+    },
+    {
+      "epoch": 0.01183334250646706,
+      "grad_norm": 1.2531086206436157,
+      "learning_rate": 9.999751513158282e-06,
+      "loss": 1.0785,
+      "step": 215
+    },
+    {
+      "epoch": 0.01188838130882272,
+      "grad_norm": 1.0789539813995361,
+      "learning_rate": 9.999747172895395e-06,
+      "loss": 0.9794,
+      "step": 216
+    },
+    {
+      "epoch": 0.01194342011117838,
+      "grad_norm": 1.1805329322814941,
+      "learning_rate": 9.999742795055746e-06,
+      "loss": 0.9602,
+      "step": 217
+    },
+    {
+      "epoch": 0.011998458913534042,
+      "grad_norm": 2.309329032897949,
+      "learning_rate": 9.99973837963937e-06,
+      "loss": 0.9482,
+      "step": 218
+    },
+    {
+      "epoch": 0.012053497715889702,
+      "grad_norm": 1.2379088401794434,
+      "learning_rate": 9.999733926646296e-06,
+      "loss": 1.0237,
+      "step": 219
+    },
+    {
+      "epoch": 0.012108536518245363,
+      "grad_norm": 1.1581377983093262,
+      "learning_rate": 9.999729436076562e-06,
+      "loss": 1.0583,
+      "step": 220
+    },
+    {
+      "epoch": 0.012163575320601024,
+      "grad_norm": 1.3006727695465088,
+      "learning_rate": 9.999724907930199e-06,
+      "loss": 0.9581,
+      "step": 221
+    },
+    {
+      "epoch": 0.012218614122956684,
+      "grad_norm": 1.3215982913970947,
+      "learning_rate": 9.999720342207243e-06,
+      "loss": 0.9438,
+      "step": 222
+    },
+    {
+      "epoch": 0.012273652925312345,
+      "grad_norm": 1.1107337474822998,
+      "learning_rate": 9.999715738907727e-06,
+      "loss": 0.9987,
+      "step": 223
+    },
+    {
+      "epoch": 0.012328691727668007,
+      "grad_norm": 1.0745457410812378,
+      "learning_rate": 9.999711098031685e-06,
+      "loss": 0.9637,
+      "step": 224
+    },
+    {
+      "epoch": 0.012383730530023666,
+      "grad_norm": 1.110861897468567,
+      "learning_rate": 9.999706419579154e-06,
+      "loss": 1.0225,
+      "step": 225
+    },
+    {
+      "epoch": 0.012438769332379328,
+      "grad_norm": 1.0755527019500732,
+      "learning_rate": 9.999701703550167e-06,
+      "loss": 1.0204,
+      "step": 226
+    },
+    {
+      "epoch": 0.012493808134734987,
+      "grad_norm": 1.1694976091384888,
+      "learning_rate": 9.99969694994476e-06,
+      "loss": 1.0566,
+      "step": 227
+    },
+    {
+      "epoch": 0.012548846937090649,
+      "grad_norm": 1.455856442451477,
+      "learning_rate": 9.99969215876297e-06,
+      "loss": 0.9397,
+      "step": 228
+    },
+    {
+      "epoch": 0.01260388573944631,
+      "grad_norm": 1.0707073211669922,
+      "learning_rate": 9.99968733000483e-06,
+      "loss": 0.8286,
+      "step": 229
+    },
+    {
+      "epoch": 0.01265892454180197,
+      "grad_norm": 1.189548134803772,
+      "learning_rate": 9.99968246367038e-06,
+      "loss": 0.8762,
+      "step": 230
+    },
+    {
+      "epoch": 0.012713963344157631,
+      "grad_norm": 1.1439214944839478,
+      "learning_rate": 9.999677559759655e-06,
+      "loss": 0.9187,
+      "step": 231
+    },
+    {
+      "epoch": 0.012769002146513293,
+      "grad_norm": 1.2329761981964111,
+      "learning_rate": 9.999672618272691e-06,
+      "loss": 1.0374,
+      "step": 232
+    },
+    {
+      "epoch": 0.012824040948868952,
+      "grad_norm": 1.1545134782791138,
+      "learning_rate": 9.999667639209527e-06,
+      "loss": 0.9343,
+      "step": 233
+    },
+    {
+      "epoch": 0.012879079751224614,
+      "grad_norm": 1.0946775674819946,
+      "learning_rate": 9.999662622570198e-06,
+      "loss": 0.9568,
+      "step": 234
+    },
+    {
+      "epoch": 0.012934118553580273,
+      "grad_norm": 1.2099589109420776,
+      "learning_rate": 9.999657568354743e-06,
+      "loss": 1.0364,
+      "step": 235
+    },
+    {
+      "epoch": 0.012989157355935935,
+      "grad_norm": 1.09062922000885,
+      "learning_rate": 9.999652476563202e-06,
+      "loss": 1.0289,
+      "step": 236
+    },
+    {
+      "epoch": 0.013044196158291596,
+      "grad_norm": 1.154557228088379,
+      "learning_rate": 9.999647347195612e-06,
+      "loss": 0.9925,
+      "step": 237
+    },
+    {
+      "epoch": 0.013099234960647256,
+      "grad_norm": 1.025374174118042,
+      "learning_rate": 9.999642180252008e-06,
+      "loss": 0.9346,
+      "step": 238
+    },
+    {
+      "epoch": 0.013154273763002917,
+      "grad_norm": 1.1473641395568848,
+      "learning_rate": 9.999636975732433e-06,
+      "loss": 1.0244,
+      "step": 239
+    },
+    {
+      "epoch": 0.013209312565358578,
+      "grad_norm": 1.0421240329742432,
+      "learning_rate": 9.999631733636923e-06,
+      "loss": 0.9368,
+      "step": 240
+    },
+    {
+      "epoch": 0.013264351367714238,
+      "grad_norm": 1.1076610088348389,
+      "learning_rate": 9.99962645396552e-06,
+      "loss": 1.0276,
+      "step": 241
+    },
+    {
+      "epoch": 0.0133193901700699,
+      "grad_norm": 1.143559455871582,
+      "learning_rate": 9.999621136718266e-06,
+      "loss": 0.9626,
+      "step": 242
+    },
+    {
+      "epoch": 0.01337442897242556,
+      "grad_norm": 1.0958378314971924,
+      "learning_rate": 9.999615781895195e-06,
+      "loss": 1.0254,
+      "step": 243
+    },
+    {
+      "epoch": 0.01342946777478122,
+      "grad_norm": 1.117688536643982,
+      "learning_rate": 9.99961038949635e-06,
+      "loss": 0.9685,
+      "step": 244
+    },
+    {
+      "epoch": 0.013484506577136882,
+      "grad_norm": 1.1645647287368774,
+      "learning_rate": 9.999604959521771e-06,
+      "loss": 1.0666,
+      "step": 245
+    },
+    {
+      "epoch": 0.013539545379492542,
+      "grad_norm": 1.1238516569137573,
+      "learning_rate": 9.999599491971502e-06,
+      "loss": 1.0252,
+      "step": 246
+    },
+    {
+      "epoch": 0.013594584181848203,
+      "grad_norm": 1.0196914672851562,
+      "learning_rate": 9.999593986845579e-06,
+      "loss": 0.9389,
+      "step": 247
+    },
+    {
+      "epoch": 0.013649622984203864,
+      "grad_norm": 1.0231372117996216,
+      "learning_rate": 9.999588444144049e-06,
+      "loss": 0.8786,
+      "step": 248
+    },
+    {
+      "epoch": 0.013704661786559524,
+      "grad_norm": 1.2504147291183472,
+      "learning_rate": 9.999582863866947e-06,
+      "loss": 1.0969,
+      "step": 249
+    },
+    {
+      "epoch": 0.013759700588915185,
+      "grad_norm": 1.1123549938201904,
+      "learning_rate": 9.99957724601432e-06,
+      "loss": 0.8833,
+      "step": 250
+    },
+    {
+      "epoch": 0.013814739391270847,
+      "grad_norm": 1.1068202257156372,
+      "learning_rate": 9.999571590586208e-06,
+      "loss": 0.9709,
+      "step": 251
+    },
+    {
+      "epoch": 0.013869778193626506,
+      "grad_norm": 0.9891651272773743,
+      "learning_rate": 9.999565897582655e-06,
+      "loss": 0.8598,
+      "step": 252
+    },
+    {
+      "epoch": 0.013924816995982168,
+      "grad_norm": 0.9866491556167603,
+      "learning_rate": 9.999560167003703e-06,
+      "loss": 0.8101,
+      "step": 253
+    },
+    {
+      "epoch": 0.013979855798337828,
+      "grad_norm": 1.0862594842910767,
+      "learning_rate": 9.999554398849396e-06,
+      "loss": 0.9411,
+      "step": 254
+    },
+    {
+      "epoch": 0.014034894600693489,
+      "grad_norm": 1.1898949146270752,
+      "learning_rate": 9.999548593119774e-06,
+      "loss": 0.9548,
+      "step": 255
+    },
+    {
+      "epoch": 0.01408993340304915,
+      "grad_norm": 1.2167880535125732,
+      "learning_rate": 9.999542749814886e-06,
+      "loss": 1.0302,
+      "step": 256
+    },
+    {
+      "epoch": 0.01414497220540481,
+      "grad_norm": 1.0784146785736084,
+      "learning_rate": 9.999536868934771e-06,
+      "loss": 0.8875,
+      "step": 257
+    },
+    {
+      "epoch": 0.014200011007760471,
+      "grad_norm": 1.1128027439117432,
+      "learning_rate": 9.999530950479475e-06,
+      "loss": 0.9498,
+      "step": 258
+    },
+    {
+      "epoch": 0.014255049810116133,
+      "grad_norm": 1.1311595439910889,
+      "learning_rate": 9.999524994449044e-06,
+      "loss": 0.9035,
+      "step": 259
+    },
+    {
+      "epoch": 0.014310088612471792,
+      "grad_norm": 1.225615382194519,
+      "learning_rate": 9.999519000843521e-06,
+      "loss": 1.0104,
+      "step": 260
+    },
+    {
+      "epoch": 0.014365127414827454,
+      "grad_norm": 1.2347793579101562,
+      "learning_rate": 9.99951296966295e-06,
+      "loss": 1.0288,
+      "step": 261
+    },
+    {
+      "epoch": 0.014420166217183113,
+      "grad_norm": 1.1837103366851807,
+      "learning_rate": 9.99950690090738e-06,
+      "loss": 0.9553,
+      "step": 262
+    },
+    {
+      "epoch": 0.014475205019538775,
+      "grad_norm": 1.1985397338867188,
+      "learning_rate": 9.999500794576852e-06,
+      "loss": 0.9561,
+      "step": 263
+    },
+    {
+      "epoch": 0.014530243821894436,
+      "grad_norm": 1.036928415298462,
+      "learning_rate": 9.999494650671418e-06,
+      "loss": 0.8906,
+      "step": 264
+    },
+    {
+      "epoch": 0.014585282624250096,
+      "grad_norm": 1.0797842741012573,
+      "learning_rate": 9.999488469191116e-06,
+      "loss": 0.8975,
+      "step": 265
+    },
+    {
+      "epoch": 0.014640321426605757,
+      "grad_norm": 1.0571156740188599,
+      "learning_rate": 9.999482250136e-06,
+      "loss": 0.9334,
+      "step": 266
+    },
+    {
+      "epoch": 0.014695360228961419,
+      "grad_norm": 1.2065023183822632,
+      "learning_rate": 9.999475993506114e-06,
+      "loss": 0.8986,
+      "step": 267
+    },
+    {
+      "epoch": 0.014750399031317078,
+      "grad_norm": 1.201586127281189,
+      "learning_rate": 9.999469699301502e-06,
+      "loss": 0.9192,
+      "step": 268
+    },
+    {
+      "epoch": 0.01480543783367274,
+      "grad_norm": 1.0470168590545654,
+      "learning_rate": 9.999463367522216e-06,
+      "loss": 0.8604,
+      "step": 269
+    },
+    {
+      "epoch": 0.0148604766360284,
+      "grad_norm": 1.1142147779464722,
+      "learning_rate": 9.9994569981683e-06,
+      "loss": 0.9847,
+      "step": 270
+    },
+    {
+      "epoch": 0.01491551543838406,
+      "grad_norm": 1.0352061986923218,
+      "learning_rate": 9.999450591239805e-06,
+      "loss": 0.8927,
+      "step": 271
+    },
+    {
+      "epoch": 0.014970554240739722,
+      "grad_norm": 1.0353184938430786,
+      "learning_rate": 9.999444146736779e-06,
+      "loss": 0.8435,
+      "step": 272
+    },
+    {
+      "epoch": 0.015025593043095382,
+      "grad_norm": 1.2091951370239258,
+      "learning_rate": 9.999437664659267e-06,
+      "loss": 0.8959,
+      "step": 273
+    },
+    {
+      "epoch": 0.015080631845451043,
+      "grad_norm": 1.006361722946167,
+      "learning_rate": 9.999431145007319e-06,
+      "loss": 0.8579,
+      "step": 274
+    },
+    {
+      "epoch": 0.015135670647806704,
+      "grad_norm": 1.1265509128570557,
+      "learning_rate": 9.999424587780985e-06,
+      "loss": 0.8808,
+      "step": 275
+    },
+    {
+      "epoch": 0.015190709450162364,
+      "grad_norm": 1.060882568359375,
+      "learning_rate": 9.999417992980317e-06,
+      "loss": 1.044,
+      "step": 276
+    },
+    {
+      "epoch": 0.015245748252518026,
+      "grad_norm": 1.0216747522354126,
+      "learning_rate": 9.999411360605358e-06,
+      "loss": 0.7773,
+      "step": 277
+    },
+    {
+      "epoch": 0.015300787054873685,
+      "grad_norm": 1.1382462978363037,
+      "learning_rate": 9.999404690656163e-06,
+      "loss": 0.8954,
+      "step": 278
+    },
+    {
+      "epoch": 0.015355825857229347,
+      "grad_norm": 1.113815188407898,
+      "learning_rate": 9.99939798313278e-06,
+      "loss": 0.8143,
+      "step": 279
+    },
+    {
+      "epoch": 0.015410864659585008,
+      "grad_norm": 1.123530387878418,
+      "learning_rate": 9.99939123803526e-06,
+      "loss": 0.8872,
+      "step": 280
+    },
+    {
+      "epoch": 0.015465903461940668,
+      "grad_norm": 1.0873669385910034,
+      "learning_rate": 9.999384455363656e-06,
+      "loss": 1.008,
+      "step": 281
+    },
+    {
+      "epoch": 0.015520942264296329,
+      "grad_norm": 1.5956637859344482,
+      "learning_rate": 9.999377635118014e-06,
+      "loss": 0.9456,
+      "step": 282
+    },
+    {
+      "epoch": 0.01557598106665199,
+      "grad_norm": 1.1471425294876099,
+      "learning_rate": 9.999370777298389e-06,
+      "loss": 0.9897,
+      "step": 283
+    },
+    {
+      "epoch": 0.01563101986900765,
+      "grad_norm": 0.9960193634033203,
+      "learning_rate": 9.999363881904831e-06,
+      "loss": 0.8196,
+      "step": 284
+    },
+    {
+      "epoch": 0.01568605867136331,
+      "grad_norm": 1.1033951044082642,
+      "learning_rate": 9.999356948937393e-06,
+      "loss": 0.879,
+      "step": 285
+    },
+    {
+      "epoch": 0.015741097473718973,
+      "grad_norm": 1.157765507698059,
+      "learning_rate": 9.999349978396126e-06,
+      "loss": 1.0116,
+      "step": 286
+    },
+    {
+      "epoch": 0.015796136276074634,
+      "grad_norm": 1.0472352504730225,
+      "learning_rate": 9.999342970281084e-06,
+      "loss": 0.8657,
+      "step": 287
+    },
+    {
+      "epoch": 0.015851175078430292,
+      "grad_norm": 1.1346659660339355,
+      "learning_rate": 9.999335924592315e-06,
+      "loss": 0.8482,
+      "step": 288
+    },
+    {
+      "epoch": 0.015906213880785953,
+      "grad_norm": 1.1164487600326538,
+      "learning_rate": 9.999328841329879e-06,
+      "loss": 1.0542,
+      "step": 289
+    },
+    {
+      "epoch": 0.015961252683141615,
+      "grad_norm": 1.1890591382980347,
+      "learning_rate": 9.999321720493825e-06,
+      "loss": 0.9598,
+      "step": 290
+    },
+    {
+      "epoch": 0.016016291485497276,
+      "grad_norm": 1.0419867038726807,
+      "learning_rate": 9.999314562084205e-06,
+      "loss": 0.9548,
+      "step": 291
+    },
+    {
+      "epoch": 0.016071330287852938,
+      "grad_norm": 1.0652042627334595,
+      "learning_rate": 9.999307366101077e-06,
+      "loss": 0.9359,
+      "step": 292
+    },
+    {
+      "epoch": 0.016126369090208596,
+      "grad_norm": 1.0166404247283936,
+      "learning_rate": 9.999300132544492e-06,
+      "loss": 0.9276,
+      "step": 293
+    },
+    {
+      "epoch": 0.016181407892564257,
+      "grad_norm": 1.1638866662979126,
+      "learning_rate": 9.999292861414507e-06,
+      "loss": 0.957,
+      "step": 294
+    },
+    {
+      "epoch": 0.01623644669491992,
+      "grad_norm": 1.5505993366241455,
+      "learning_rate": 9.999285552711173e-06,
+      "loss": 0.9878,
+      "step": 295
+    },
+    {
+      "epoch": 0.01629148549727558,
+      "grad_norm": 1.177262783050537,
+      "learning_rate": 9.999278206434549e-06,
+      "loss": 0.8631,
+      "step": 296
+    },
+    {
+      "epoch": 0.01634652429963124,
+      "grad_norm": 1.8578168153762817,
+      "learning_rate": 9.999270822584687e-06,
+      "loss": 0.9684,
+      "step": 297
+    },
+    {
+      "epoch": 0.0164015631019869,
+      "grad_norm": 1.2617360353469849,
+      "learning_rate": 9.999263401161643e-06,
+      "loss": 1.014,
+      "step": 298
+    },
+    {
+      "epoch": 0.01645660190434256,
+      "grad_norm": 0.9740132689476013,
+      "learning_rate": 9.999255942165475e-06,
+      "loss": 0.8606,
+      "step": 299
+    },
+    {
+      "epoch": 0.016511640706698222,
+      "grad_norm": 0.9821745753288269,
+      "learning_rate": 9.999248445596238e-06,
+      "loss": 0.8241,
+      "step": 300
+    },
+    {
+      "epoch": 0.016566679509053883,
+      "grad_norm": 1.0200445652008057,
+      "learning_rate": 9.999240911453986e-06,
+      "loss": 0.8256,
+      "step": 301
+    },
+    {
+      "epoch": 0.016621718311409545,
+      "grad_norm": 1.4100390672683716,
+      "learning_rate": 9.999233339738779e-06,
+      "loss": 0.9057,
+      "step": 302
+    },
+    {
+      "epoch": 0.016676757113765206,
+      "grad_norm": 1.056544303894043,
+      "learning_rate": 9.99922573045067e-06,
+      "loss": 1.0808,
+      "step": 303
+    },
+    {
+      "epoch": 0.016731795916120864,
+      "grad_norm": 0.9271026253700256,
+      "learning_rate": 9.99921808358972e-06,
+      "loss": 0.878,
+      "step": 304
+    },
+    {
+      "epoch": 0.016786834718476525,
+      "grad_norm": 0.9864157438278198,
+      "learning_rate": 9.999210399155987e-06,
+      "loss": 0.9198,
+      "step": 305
+    },
+    {
+      "epoch": 0.016841873520832187,
+      "grad_norm": 1.093995451927185,
+      "learning_rate": 9.999202677149525e-06,
+      "loss": 0.9794,
+      "step": 306
+    },
+    {
+      "epoch": 0.016896912323187848,
+      "grad_norm": 0.9717912077903748,
+      "learning_rate": 9.999194917570395e-06,
+      "loss": 0.8764,
+      "step": 307
+    },
+    {
+      "epoch": 0.01695195112554351,
+      "grad_norm": 1.0026428699493408,
+      "learning_rate": 9.999187120418653e-06,
+      "loss": 0.8526,
+      "step": 308
+    },
+    {
+      "epoch": 0.017006989927899167,
+      "grad_norm": 1.122870922088623,
+      "learning_rate": 9.999179285694359e-06,
+      "loss": 0.9773,
+      "step": 309
+    },
+    {
+      "epoch": 0.01706202873025483,
+      "grad_norm": 1.0522836446762085,
+      "learning_rate": 9.999171413397572e-06,
+      "loss": 1.0183,
+      "step": 310
+    },
+    {
+      "epoch": 0.01711706753261049,
+      "grad_norm": 0.9303658604621887,
+      "learning_rate": 9.99916350352835e-06,
+      "loss": 0.8402,
+      "step": 311
+    },
+    {
+      "epoch": 0.01717210633496615,
+      "grad_norm": 0.9606096148490906,
+      "learning_rate": 9.999155556086755e-06,
+      "loss": 0.9692,
+      "step": 312
+    },
+    {
+      "epoch": 0.017227145137321813,
+      "grad_norm": 1.176992416381836,
+      "learning_rate": 9.999147571072844e-06,
+      "loss": 0.8172,
+      "step": 313
+    },
+    {
+      "epoch": 0.017282183939677474,
+      "grad_norm": 1.1948801279067993,
+      "learning_rate": 9.999139548486678e-06,
+      "loss": 1.0205,
+      "step": 314
+    },
+    {
+      "epoch": 0.017337222742033132,
+      "grad_norm": 1.0064897537231445,
+      "learning_rate": 9.999131488328318e-06,
+      "loss": 0.9479,
+      "step": 315
+    },
+    {
+      "epoch": 0.017392261544388794,
+      "grad_norm": 1.048242449760437,
+      "learning_rate": 9.999123390597822e-06,
+      "loss": 0.9862,
+      "step": 316
+    },
+    {
+      "epoch": 0.017447300346744455,
+      "grad_norm": 1.12875497341156,
+      "learning_rate": 9.999115255295256e-06,
+      "loss": 0.9743,
+      "step": 317
+    },
+    {
+      "epoch": 0.017502339149100116,
+      "grad_norm": 1.0607460737228394,
+      "learning_rate": 9.999107082420674e-06,
+      "loss": 0.8878,
+      "step": 318
+    },
+    {
+      "epoch": 0.017557377951455778,
+      "grad_norm": 1.1480191946029663,
+      "learning_rate": 9.999098871974144e-06,
+      "loss": 0.8769,
+      "step": 319
+    },
+    {
+      "epoch": 0.017612416753811436,
+      "grad_norm": 1.1150004863739014,
+      "learning_rate": 9.999090623955724e-06,
+      "loss": 0.8615,
+      "step": 320
+    },
+    {
+      "epoch": 0.017667455556167097,
+      "grad_norm": 1.137839913368225,
+      "learning_rate": 9.999082338365478e-06,
+      "loss": 0.9703,
+      "step": 321
+    },
+    {
+      "epoch": 0.01772249435852276,
+      "grad_norm": 1.0883489847183228,
+      "learning_rate": 9.999074015203467e-06,
+      "loss": 0.9273,
+      "step": 322
+    },
+    {
+      "epoch": 0.01777753316087842,
+      "grad_norm": 1.0999557971954346,
+      "learning_rate": 9.999065654469752e-06,
+      "loss": 0.9605,
+      "step": 323
+    },
+    {
+      "epoch": 0.01783257196323408,
+      "grad_norm": 0.9911689758300781,
+      "learning_rate": 9.999057256164401e-06,
+      "loss": 0.9117,
+      "step": 324
+    },
+    {
+      "epoch": 0.01788761076558974,
+      "grad_norm": 1.040933609008789,
+      "learning_rate": 9.999048820287472e-06,
+      "loss": 0.9229,
+      "step": 325
+    },
+    {
+      "epoch": 0.0179426495679454,
+      "grad_norm": 1.4341392517089844,
+      "learning_rate": 9.999040346839031e-06,
+      "loss": 1.0718,
+      "step": 326
+    },
+    {
+      "epoch": 0.017997688370301062,
+      "grad_norm": 1.0246332883834839,
+      "learning_rate": 9.99903183581914e-06,
+      "loss": 0.9617,
+      "step": 327
+    },
+    {
+      "epoch": 0.018052727172656723,
+      "grad_norm": 10.162322998046875,
+      "learning_rate": 9.999023287227863e-06,
+      "loss": 1.0391,
+      "step": 328
+    },
+    {
+      "epoch": 0.018107765975012385,
+      "grad_norm": 1.3370027542114258,
+      "learning_rate": 9.999014701065266e-06,
+      "loss": 1.0211,
+      "step": 329
+    },
+    {
+      "epoch": 0.018162804777368046,
+      "grad_norm": 1.0146219730377197,
+      "learning_rate": 9.999006077331413e-06,
+      "loss": 0.8611,
+      "step": 330
+    },
+    {
+      "epoch": 0.018217843579723704,
+      "grad_norm": 1.0899269580841064,
+      "learning_rate": 9.998997416026368e-06,
+      "loss": 0.9209,
+      "step": 331
+    },
+    {
+      "epoch": 0.018272882382079365,
+      "grad_norm": 1.1343204975128174,
+      "learning_rate": 9.998988717150198e-06,
+      "loss": 0.9405,
+      "step": 332
+    },
+    {
+      "epoch": 0.018327921184435027,
+      "grad_norm": 1.2308380603790283,
+      "learning_rate": 9.998979980702965e-06,
+      "loss": 0.9579,
+      "step": 333
+    },
+    {
+      "epoch": 0.018382959986790688,
+      "grad_norm": 1.1433519124984741,
+      "learning_rate": 9.998971206684737e-06,
+      "loss": 1.0045,
+      "step": 334
+    },
+    {
+      "epoch": 0.01843799878914635,
+      "grad_norm": 1.0585781335830688,
+      "learning_rate": 9.99896239509558e-06,
+      "loss": 0.9171,
+      "step": 335
+    },
+    {
+      "epoch": 0.018493037591502007,
+      "grad_norm": 1.2735164165496826,
+      "learning_rate": 9.99895354593556e-06,
+      "loss": 1.1001,
+      "step": 336
+    },
+    {
+      "epoch": 0.01854807639385767,
+      "grad_norm": 1.2905755043029785,
+      "learning_rate": 9.998944659204744e-06,
+      "loss": 1.0294,
+      "step": 337
+    },
+    {
+      "epoch": 0.01860311519621333,
+      "grad_norm": 1.1442075967788696,
+      "learning_rate": 9.998935734903198e-06,
+      "loss": 0.9385,
+      "step": 338
+    },
+    {
+      "epoch": 0.01865815399856899,
+      "grad_norm": 1.1005232334136963,
+      "learning_rate": 9.998926773030987e-06,
+      "loss": 1.026,
+      "step": 339
+    },
+    {
+      "epoch": 0.018713192800924653,
+      "grad_norm": 1.2770785093307495,
+      "learning_rate": 9.998917773588182e-06,
+      "loss": 1.0015,
+      "step": 340
+    },
+    {
+      "epoch": 0.01876823160328031,
+      "grad_norm": 1.0963070392608643,
+      "learning_rate": 9.998908736574849e-06,
+      "loss": 0.9347,
+      "step": 341
+    },
+    {
+      "epoch": 0.018823270405635972,
+      "grad_norm": 1.10364830493927,
+      "learning_rate": 9.998899661991055e-06,
+      "loss": 0.869,
+      "step": 342
+    },
+    {
+      "epoch": 0.018878309207991634,
+      "grad_norm": 1.0364975929260254,
+      "learning_rate": 9.99889054983687e-06,
+      "loss": 0.9855,
+      "step": 343
+    },
+    {
+      "epoch": 0.018933348010347295,
+      "grad_norm": 1.104702115058899,
+      "learning_rate": 9.998881400112362e-06,
+      "loss": 0.9555,
+      "step": 344
+    },
+    {
+      "epoch": 0.018988386812702956,
+      "grad_norm": 0.9957441687583923,
+      "learning_rate": 9.998872212817599e-06,
+      "loss": 0.9634,
+      "step": 345
+    },
+    {
+      "epoch": 0.019043425615058618,
+      "grad_norm": 1.262271523475647,
+      "learning_rate": 9.998862987952651e-06,
+      "loss": 1.0133,
+      "step": 346
+    },
+    {
+      "epoch": 0.019098464417414276,
+      "grad_norm": 1.2075226306915283,
+      "learning_rate": 9.998853725517587e-06,
+      "loss": 1.0588,
+      "step": 347
+    },
+    {
+      "epoch": 0.019153503219769937,
+      "grad_norm": 1.0609898567199707,
+      "learning_rate": 9.998844425512477e-06,
+      "loss": 0.9952,
+      "step": 348
+    },
+    {
+      "epoch": 0.0192085420221256,
+      "grad_norm": 1.1930195093154907,
+      "learning_rate": 9.998835087937389e-06,
+      "loss": 0.9617,
+      "step": 349
+    },
+    {
+      "epoch": 0.01926358082448126,
+      "grad_norm": 1.2359932661056519,
+      "learning_rate": 9.998825712792396e-06,
+      "loss": 0.8768,
+      "step": 350
+    },
+    {
+      "epoch": 0.01931861962683692,
+      "grad_norm": 0.9984115362167358,
+      "learning_rate": 9.998816300077566e-06,
+      "loss": 0.8205,
+      "step": 351
+    },
+    {
+      "epoch": 0.01937365842919258,
+      "grad_norm": 1.6853677034378052,
+      "learning_rate": 9.998806849792972e-06,
+      "loss": 0.9066,
+      "step": 352
+    },
+    {
+      "epoch": 0.01942869723154824,
+      "grad_norm": 1.2869856357574463,
+      "learning_rate": 9.998797361938683e-06,
+      "loss": 1.0054,
+      "step": 353
+    },
+    {
+      "epoch": 0.019483736033903902,
+      "grad_norm": 1.2791584730148315,
+      "learning_rate": 9.99878783651477e-06,
+      "loss": 0.7627,
+      "step": 354
+    },
+    {
+      "epoch": 0.019538774836259563,
+      "grad_norm": 1.0795867443084717,
+      "learning_rate": 9.998778273521307e-06,
+      "loss": 0.9343,
+      "step": 355
+    },
+    {
+      "epoch": 0.019593813638615225,
+      "grad_norm": 1.0926088094711304,
+      "learning_rate": 9.998768672958365e-06,
+      "loss": 0.943,
+      "step": 356
+    },
+    {
+      "epoch": 0.019648852440970886,
+      "grad_norm": 1.0530847311019897,
+      "learning_rate": 9.998759034826015e-06,
+      "loss": 0.9656,
+      "step": 357
+    },
+    {
+      "epoch": 0.019703891243326544,
+      "grad_norm": 1.1793400049209595,
+      "learning_rate": 9.99874935912433e-06,
+      "loss": 0.9799,
+      "step": 358
+    },
+    {
+      "epoch": 0.019758930045682205,
+      "grad_norm": 1.0726191997528076,
+      "learning_rate": 9.998739645853383e-06,
+      "loss": 0.8739,
+      "step": 359
+    },
+    {
+      "epoch": 0.019813968848037867,
+      "grad_norm": 1.0488981008529663,
+      "learning_rate": 9.998729895013246e-06,
+      "loss": 0.8986,
+      "step": 360
+    },
+    {
+      "epoch": 0.019869007650393528,
+      "grad_norm": 1.8267477750778198,
+      "learning_rate": 9.998720106603993e-06,
+      "loss": 0.9175,
+      "step": 361
+    },
+    {
+      "epoch": 0.01992404645274919,
+      "grad_norm": 0.9868306517601013,
+      "learning_rate": 9.9987102806257e-06,
+      "loss": 0.9609,
+      "step": 362
+    },
+    {
+      "epoch": 0.019979085255104848,
+      "grad_norm": 1.0171183347702026,
+      "learning_rate": 9.998700417078438e-06,
+      "loss": 0.8904,
+      "step": 363
+    },
+    {
+      "epoch": 0.02003412405746051,
+      "grad_norm": 0.9800812602043152,
+      "learning_rate": 9.998690515962282e-06,
+      "loss": 0.8344,
+      "step": 364
+    },
+    {
+      "epoch": 0.02008916285981617,
+      "grad_norm": 1.024707317352295,
+      "learning_rate": 9.998680577277304e-06,
+      "loss": 0.9026,
+      "step": 365
+    },
+    {
+      "epoch": 0.02014420166217183,
+      "grad_norm": 1.1056619882583618,
+      "learning_rate": 9.998670601023584e-06,
+      "loss": 1.017,
+      "step": 366
+    },
+    {
+      "epoch": 0.020199240464527493,
+      "grad_norm": 1.0555908679962158,
+      "learning_rate": 9.998660587201191e-06,
+      "loss": 0.9627,
+      "step": 367
+    },
+    {
+      "epoch": 0.02025427926688315,
+      "grad_norm": 0.9502031803131104,
+      "learning_rate": 9.998650535810204e-06,
+      "loss": 0.935,
+      "step": 368
+    },
+    {
+      "epoch": 0.020309318069238812,
+      "grad_norm": 1.0355613231658936,
+      "learning_rate": 9.998640446850699e-06,
+      "loss": 0.9946,
+      "step": 369
+    },
+    {
+      "epoch": 0.020364356871594474,
+      "grad_norm": 0.9906355142593384,
+      "learning_rate": 9.99863032032275e-06,
+      "loss": 0.9389,
+      "step": 370
+    },
+    {
+      "epoch": 0.020419395673950135,
+      "grad_norm": 0.9483911395072937,
+      "learning_rate": 9.99862015622643e-06,
+      "loss": 0.979,
+      "step": 371
+    },
+    {
+      "epoch": 0.020474434476305797,
+      "grad_norm": 0.9769986271858215,
+      "learning_rate": 9.998609954561822e-06,
+      "loss": 0.8972,
+      "step": 372
+    },
+    {
+      "epoch": 0.020529473278661458,
+      "grad_norm": 1.1682699918746948,
+      "learning_rate": 9.998599715329e-06,
+      "loss": 0.943,
+      "step": 373
+    },
+    {
+      "epoch": 0.020584512081017116,
+      "grad_norm": 1.007912516593933,
+      "learning_rate": 9.99858943852804e-06,
+      "loss": 0.8825,
+      "step": 374
+    },
+    {
+      "epoch": 0.020639550883372777,
+      "grad_norm": 0.9788785576820374,
+      "learning_rate": 9.99857912415902e-06,
+      "loss": 0.9667,
+      "step": 375
+    },
+    {
+      "epoch": 0.02069458968572844,
+      "grad_norm": 1.0804275274276733,
+      "learning_rate": 9.998568772222017e-06,
+      "loss": 1.0026,
+      "step": 376
+    },
+    {
+      "epoch": 0.0207496284880841,
+      "grad_norm": 1.0859237909317017,
+      "learning_rate": 9.998558382717109e-06,
+      "loss": 0.9592,
+      "step": 377
+    },
+    {
+      "epoch": 0.02080466729043976,
+      "grad_norm": 1.2925337553024292,
+      "learning_rate": 9.998547955644373e-06,
+      "loss": 0.9067,
+      "step": 378
+    },
+    {
+      "epoch": 0.02085970609279542,
+      "grad_norm": 0.9853373765945435,
+      "learning_rate": 9.99853749100389e-06,
+      "loss": 0.9538,
+      "step": 379
+    },
+    {
+      "epoch": 0.02091474489515108,
+      "grad_norm": 1.0461076498031616,
+      "learning_rate": 9.998526988795738e-06,
+      "loss": 0.9261,
+      "step": 380
+    },
+    {
+      "epoch": 0.020969783697506742,
+      "grad_norm": 1.024559497833252,
+      "learning_rate": 9.998516449019995e-06,
+      "loss": 0.9117,
+      "step": 381
+    },
+    {
+      "epoch": 0.021024822499862404,
+      "grad_norm": 1.1474825143814087,
+      "learning_rate": 9.998505871676739e-06,
+      "loss": 1.0177,
+      "step": 382
+    },
+    {
+      "epoch": 0.021079861302218065,
+      "grad_norm": 0.9587596654891968,
+      "learning_rate": 9.998495256766051e-06,
+      "loss": 0.8809,
+      "step": 383
+    },
+    {
+      "epoch": 0.021134900104573723,
+      "grad_norm": 0.9505122303962708,
+      "learning_rate": 9.998484604288013e-06,
+      "loss": 0.9266,
+      "step": 384
+    },
+    {
+      "epoch": 0.021189938906929384,
+      "grad_norm": 0.9625647664070129,
+      "learning_rate": 9.9984739142427e-06,
+      "loss": 0.9073,
+      "step": 385
+    },
+    {
+      "epoch": 0.021244977709285046,
+      "grad_norm": 0.9650934338569641,
+      "learning_rate": 9.998463186630196e-06,
+      "loss": 0.9042,
+      "step": 386
+    },
+    {
+      "epoch": 0.021300016511640707,
+      "grad_norm": 1.0289491415023804,
+      "learning_rate": 9.99845242145058e-06,
+      "loss": 0.929,
+      "step": 387
+    },
+    {
+      "epoch": 0.02135505531399637,
+      "grad_norm": 0.9543869495391846,
+      "learning_rate": 9.998441618703935e-06,
+      "loss": 0.9406,
+      "step": 388
+    },
+    {
+      "epoch": 0.02141009411635203,
+      "grad_norm": 0.9276942610740662,
+      "learning_rate": 9.99843077839034e-06,
+      "loss": 0.8982,
+      "step": 389
+    },
+    {
+      "epoch": 0.021465132918707688,
+      "grad_norm": 0.9264664053916931,
+      "learning_rate": 9.998419900509877e-06,
+      "loss": 0.7255,
+      "step": 390
+    },
+    {
+      "epoch": 0.02152017172106335,
+      "grad_norm": 0.9961187243461609,
+      "learning_rate": 9.998408985062628e-06,
+      "loss": 0.9826,
+      "step": 391
+    },
+    {
+      "epoch": 0.02157521052341901,
+      "grad_norm": 0.966596245765686,
+      "learning_rate": 9.998398032048676e-06,
+      "loss": 0.8159,
+      "step": 392
+    },
+    {
+      "epoch": 0.021630249325774672,
+      "grad_norm": 1.1336095333099365,
+      "learning_rate": 9.998387041468102e-06,
+      "loss": 0.9289,
+      "step": 393
+    },
+    {
+      "epoch": 0.021685288128130333,
+      "grad_norm": 1.0453619956970215,
+      "learning_rate": 9.998376013320989e-06,
+      "loss": 0.8816,
+      "step": 394
+    },
+    {
+      "epoch": 0.02174032693048599,
+      "grad_norm": 0.8961821794509888,
+      "learning_rate": 9.998364947607419e-06,
+      "loss": 0.871,
+      "step": 395
+    },
+    {
+      "epoch": 0.021795365732841653,
+      "grad_norm": 1.3420332670211792,
+      "learning_rate": 9.998353844327477e-06,
+      "loss": 0.9338,
+      "step": 396
+    },
+    {
+      "epoch": 0.021850404535197314,
+      "grad_norm": 0.9635335206985474,
+      "learning_rate": 9.998342703481246e-06,
+      "loss": 0.9592,
+      "step": 397
+    },
+    {
+      "epoch": 0.021905443337552975,
+      "grad_norm": 1.3322341442108154,
+      "learning_rate": 9.998331525068807e-06,
+      "loss": 1.0974,
+      "step": 398
+    },
+    {
+      "epoch": 0.021960482139908637,
+      "grad_norm": 1.017220377922058,
+      "learning_rate": 9.998320309090247e-06,
+      "loss": 0.9827,
+      "step": 399
+    },
+    {
+      "epoch": 0.022015520942264295,
+      "grad_norm": 1.0080329179763794,
+      "learning_rate": 9.99830905554565e-06,
+      "loss": 0.877,
+      "step": 400
+    },
+    {
+      "epoch": 0.022070559744619956,
+      "grad_norm": 0.9883211255073547,
+      "learning_rate": 9.998297764435101e-06,
+      "loss": 0.9625,
+      "step": 401
+    },
+    {
+      "epoch": 0.022125598546975617,
+      "grad_norm": 1.0948412418365479,
+      "learning_rate": 9.998286435758684e-06,
+      "loss": 0.9058,
+      "step": 402
+    },
+    {
+      "epoch": 0.02218063734933128,
+      "grad_norm": 0.9402000308036804,
+      "learning_rate": 9.998275069516482e-06,
+      "loss": 0.8882,
+      "step": 403
+    },
+    {
+      "epoch": 0.02223567615168694,
+      "grad_norm": 0.9858806133270264,
+      "learning_rate": 9.998263665708583e-06,
+      "loss": 0.9086,
+      "step": 404
+    },
+    {
+      "epoch": 0.0222907149540426,
+      "grad_norm": 1.0556131601333618,
+      "learning_rate": 9.998252224335073e-06,
+      "loss": 0.9583,
+      "step": 405
+    },
+    {
+      "epoch": 0.02234575375639826,
+      "grad_norm": 1.092766284942627,
+      "learning_rate": 9.998240745396037e-06,
+      "loss": 0.9124,
+      "step": 406
+    },
+    {
+      "epoch": 0.02240079255875392,
+      "grad_norm": 1.1902250051498413,
+      "learning_rate": 9.998229228891563e-06,
+      "loss": 1.0566,
+      "step": 407
+    },
+    {
+      "epoch": 0.022455831361109582,
+      "grad_norm": 1.067906141281128,
+      "learning_rate": 9.998217674821734e-06,
+      "loss": 0.9823,
+      "step": 408
+    },
+    {
+      "epoch": 0.022510870163465244,
+      "grad_norm": 1.0051710605621338,
+      "learning_rate": 9.998206083186638e-06,
+      "loss": 0.9141,
+      "step": 409
+    },
+    {
+      "epoch": 0.022565908965820905,
+      "grad_norm": 1.046412467956543,
+      "learning_rate": 9.998194453986367e-06,
+      "loss": 0.9439,
+      "step": 410
+    },
+    {
+      "epoch": 0.022620947768176563,
+      "grad_norm": 1.1103553771972656,
+      "learning_rate": 9.998182787221e-06,
+      "loss": 0.9494,
+      "step": 411
+    },
+    {
+      "epoch": 0.022675986570532224,
+      "grad_norm": 1.0508466958999634,
+      "learning_rate": 9.998171082890632e-06,
+      "loss": 0.9202,
+      "step": 412
+    },
+    {
+      "epoch": 0.022731025372887886,
+      "grad_norm": 1.1364226341247559,
+      "learning_rate": 9.998159340995347e-06,
+      "loss": 0.9859,
+      "step": 413
+    },
+    {
+      "epoch": 0.022786064175243547,
+      "grad_norm": 1.2073607444763184,
+      "learning_rate": 9.998147561535234e-06,
+      "loss": 0.8883,
+      "step": 414
+    },
+    {
+      "epoch": 0.02284110297759921,
+      "grad_norm": 1.0657012462615967,
+      "learning_rate": 9.998135744510384e-06,
+      "loss": 0.8321,
+      "step": 415
+    },
+    {
+      "epoch": 0.02289614177995487,
+      "grad_norm": 1.0101548433303833,
+      "learning_rate": 9.998123889920881e-06,
+      "loss": 0.9374,
+      "step": 416
+    },
+    {
+      "epoch": 0.022951180582310528,
+      "grad_norm": 1.057455062866211,
+      "learning_rate": 9.998111997766817e-06,
+      "loss": 0.8831,
+      "step": 417
+    },
+    {
+      "epoch": 0.02300621938466619,
+      "grad_norm": 1.206092357635498,
+      "learning_rate": 9.998100068048282e-06,
+      "loss": 0.8812,
+      "step": 418
+    },
+    {
+      "epoch": 0.02306125818702185,
+      "grad_norm": 1.0709773302078247,
+      "learning_rate": 9.998088100765366e-06,
+      "loss": 0.9486,
+      "step": 419
+    },
+    {
+      "epoch": 0.023116296989377512,
+      "grad_norm": 1.066469669342041,
+      "learning_rate": 9.998076095918156e-06,
+      "loss": 1.0229,
+      "step": 420
+    },
+    {
+      "epoch": 0.023171335791733173,
+      "grad_norm": 1.0443583726882935,
+      "learning_rate": 9.998064053506744e-06,
+      "loss": 0.8615,
+      "step": 421
+    },
+    {
+      "epoch": 0.02322637459408883,
+      "grad_norm": 1.103096842765808,
+      "learning_rate": 9.99805197353122e-06,
+      "loss": 0.9909,
+      "step": 422
+    },
+    {
+      "epoch": 0.023281413396444493,
+      "grad_norm": 0.9804643392562866,
+      "learning_rate": 9.998039855991677e-06,
+      "loss": 0.9214,
+      "step": 423
+    },
+    {
+      "epoch": 0.023336452198800154,
+      "grad_norm": 0.9880676865577698,
+      "learning_rate": 9.998027700888202e-06,
+      "loss": 0.9345,
+      "step": 424
+    },
+    {
+      "epoch": 0.023391491001155815,
+      "grad_norm": 0.9633826017379761,
+      "learning_rate": 9.99801550822089e-06,
+      "loss": 0.9897,
+      "step": 425
+    },
+    {
+      "epoch": 0.023446529803511477,
+      "grad_norm": 1.0159331560134888,
+      "learning_rate": 9.998003277989831e-06,
+      "loss": 0.9385,
+      "step": 426
+    },
+    {
+      "epoch": 0.023501568605867135,
+      "grad_norm": 1.009667158126831,
+      "learning_rate": 9.99799101019512e-06,
+      "loss": 0.9013,
+      "step": 427
+    },
+    {
+      "epoch": 0.023556607408222796,
+      "grad_norm": 0.9478578567504883,
+      "learning_rate": 9.997978704836842e-06,
+      "loss": 0.8775,
+      "step": 428
+    },
+    {
+      "epoch": 0.023611646210578457,
+      "grad_norm": 1.013181447982788,
+      "learning_rate": 9.997966361915096e-06,
+      "loss": 0.8797,
+      "step": 429
+    },
+    {
+      "epoch": 0.02366668501293412,
+      "grad_norm": 1.0337481498718262,
+      "learning_rate": 9.997953981429974e-06,
+      "loss": 1.0047,
+      "step": 430
+    },
+    {
+      "epoch": 0.02372172381528978,
+      "grad_norm": 0.9423721432685852,
+      "learning_rate": 9.997941563381566e-06,
+      "loss": 0.8639,
+      "step": 431
+    },
+    {
+      "epoch": 0.02377676261764544,
+      "grad_norm": 1.100492000579834,
+      "learning_rate": 9.997929107769968e-06,
+      "loss": 1.0022,
+      "step": 432
+    },
+    {
+      "epoch": 0.0238318014200011,
+      "grad_norm": 1.1232364177703857,
+      "learning_rate": 9.997916614595272e-06,
+      "loss": 0.9145,
+      "step": 433
+    },
+    {
+      "epoch": 0.02388684022235676,
+      "grad_norm": 0.9466833472251892,
+      "learning_rate": 9.997904083857572e-06,
+      "loss": 0.9397,
+      "step": 434
+    },
+    {
+      "epoch": 0.023941879024712422,
+      "grad_norm": 0.9514566659927368,
+      "learning_rate": 9.997891515556963e-06,
+      "loss": 0.8025,
+      "step": 435
+    },
+    {
+      "epoch": 0.023996917827068084,
+      "grad_norm": 0.9292222261428833,
+      "learning_rate": 9.997878909693539e-06,
+      "loss": 0.7739,
+      "step": 436
+    },
+    {
+      "epoch": 0.024051956629423745,
+      "grad_norm": 1.1049963235855103,
+      "learning_rate": 9.997866266267397e-06,
+      "loss": 0.9439,
+      "step": 437
+    },
+    {
+      "epoch": 0.024106995431779403,
+      "grad_norm": 1.0938019752502441,
+      "learning_rate": 9.997853585278627e-06,
+      "loss": 0.9479,
+      "step": 438
+    },
+    {
+      "epoch": 0.024162034234135064,
+      "grad_norm": 1.0423611402511597,
+      "learning_rate": 9.997840866727331e-06,
+      "loss": 0.9309,
+      "step": 439
+    },
+    {
+      "epoch": 0.024217073036490726,
+      "grad_norm": 1.0584756135940552,
+      "learning_rate": 9.997828110613598e-06,
+      "loss": 1.0218,
+      "step": 440
+    },
+    {
+      "epoch": 0.024272111838846387,
+      "grad_norm": 0.9986408948898315,
+      "learning_rate": 9.997815316937527e-06,
+      "loss": 0.9734,
+      "step": 441
+    },
+    {
+      "epoch": 0.02432715064120205,
+      "grad_norm": 0.9680983424186707,
+      "learning_rate": 9.997802485699215e-06,
+      "loss": 0.9286,
+      "step": 442
+    },
+    {
+      "epoch": 0.024382189443557706,
+      "grad_norm": 1.2231700420379639,
+      "learning_rate": 9.997789616898757e-06,
+      "loss": 0.8083,
+      "step": 443
+    },
+    {
+      "epoch": 0.024437228245913368,
+      "grad_norm": 1.0064021348953247,
+      "learning_rate": 9.99777671053625e-06,
+      "loss": 0.9161,
+      "step": 444
+    },
+    {
+      "epoch": 0.02449226704826903,
+      "grad_norm": 0.9658541679382324,
+      "learning_rate": 9.99776376661179e-06,
+      "loss": 0.8027,
+      "step": 445
+    },
+    {
+      "epoch": 0.02454730585062469,
+      "grad_norm": 0.9440343379974365,
+      "learning_rate": 9.997750785125477e-06,
+      "loss": 0.9124,
+      "step": 446
+    },
+    {
+      "epoch": 0.024602344652980352,
+      "grad_norm": 0.998792827129364,
+      "learning_rate": 9.997737766077404e-06,
+      "loss": 0.8699,
+      "step": 447
+    },
+    {
+      "epoch": 0.024657383455336013,
+      "grad_norm": 1.430880069732666,
+      "learning_rate": 9.997724709467676e-06,
+      "loss": 0.9158,
+      "step": 448
+    },
+    {
+      "epoch": 0.02471242225769167,
+      "grad_norm": 0.9737820029258728,
+      "learning_rate": 9.997711615296384e-06,
+      "loss": 0.9496,
+      "step": 449
+    },
+    {
+      "epoch": 0.024767461060047333,
+      "grad_norm": 0.9710075855255127,
+      "learning_rate": 9.997698483563629e-06,
+      "loss": 0.8714,
+      "step": 450
+    },
+    {
+      "epoch": 0.024822499862402994,
+      "grad_norm": 1.5286253690719604,
+      "learning_rate": 9.997685314269511e-06,
+      "loss": 0.8421,
+      "step": 451
+    },
+    {
+      "epoch": 0.024877538664758655,
+      "grad_norm": 1.0269445180892944,
+      "learning_rate": 9.99767210741413e-06,
+      "loss": 1.0131,
+      "step": 452
+    },
+    {
+      "epoch": 0.024932577467114317,
+      "grad_norm": 0.9780508279800415,
+      "learning_rate": 9.99765886299758e-06,
+      "loss": 0.9897,
+      "step": 453
+    },
+    {
+      "epoch": 0.024987616269469975,
+      "grad_norm": 0.998332679271698,
+      "learning_rate": 9.997645581019965e-06,
+      "loss": 0.9647,
+      "step": 454
+    },
+    {
+      "epoch": 0.025042655071825636,
+      "grad_norm": 1.7062602043151855,
+      "learning_rate": 9.997632261481383e-06,
+      "loss": 1.0729,
+      "step": 455
+    },
+    {
+      "epoch": 0.025097693874181298,
+      "grad_norm": 0.9793694615364075,
+      "learning_rate": 9.997618904381936e-06,
+      "loss": 0.9556,
+      "step": 456
+    },
+    {
+      "epoch": 0.02515273267653696,
+      "grad_norm": 1.0183895826339722,
+      "learning_rate": 9.997605509721721e-06,
+      "loss": 0.9194,
+      "step": 457
+    },
+    {
+      "epoch": 0.02520777147889262,
+      "grad_norm": 1.0288400650024414,
+      "learning_rate": 9.997592077500844e-06,
+      "loss": 0.955,
+      "step": 458
+    },
+    {
+      "epoch": 0.025262810281248282,
+      "grad_norm": 0.9551253914833069,
+      "learning_rate": 9.997578607719401e-06,
+      "loss": 0.8498,
+      "step": 459
+    },
+    {
+      "epoch": 0.02531784908360394,
+      "grad_norm": 0.9648008942604065,
+      "learning_rate": 9.997565100377494e-06,
+      "loss": 0.9306,
+      "step": 460
+    },
+    {
+      "epoch": 0.0253728878859596,
+      "grad_norm": 0.9206677675247192,
+      "learning_rate": 9.997551555475225e-06,
+      "loss": 0.7874,
+      "step": 461
+    },
+    {
+      "epoch": 0.025427926688315262,
+      "grad_norm": 1.0479545593261719,
+      "learning_rate": 9.997537973012698e-06,
+      "loss": 0.9201,
+      "step": 462
+    },
+    {
+      "epoch": 0.025482965490670924,
+      "grad_norm": 1.0329946279525757,
+      "learning_rate": 9.997524352990013e-06,
+      "loss": 0.9577,
+      "step": 463
+    },
+    {
+      "epoch": 0.025538004293026585,
+      "grad_norm": 1.1177828311920166,
+      "learning_rate": 9.997510695407273e-06,
+      "loss": 1.0041,
+      "step": 464
+    },
+    {
+      "epoch": 0.025593043095382243,
+      "grad_norm": 1.0351577997207642,
+      "learning_rate": 9.99749700026458e-06,
+      "loss": 0.9952,
+      "step": 465
+    },
+    {
+      "epoch": 0.025648081897737905,
+      "grad_norm": 0.905274510383606,
+      "learning_rate": 9.997483267562035e-06,
+      "loss": 0.8185,
+      "step": 466
+    },
+    {
+      "epoch": 0.025703120700093566,
+      "grad_norm": 1.0749776363372803,
+      "learning_rate": 9.997469497299747e-06,
+      "loss": 1.0611,
+      "step": 467
+    },
+    {
+      "epoch": 0.025758159502449227,
+      "grad_norm": 0.8972223401069641,
+      "learning_rate": 9.997455689477815e-06,
+      "loss": 0.8994,
+      "step": 468
+    },
+    {
+      "epoch": 0.02581319830480489,
+      "grad_norm": 1.0669914484024048,
+      "learning_rate": 9.997441844096342e-06,
+      "loss": 1.06,
+      "step": 469
+    },
+    {
+      "epoch": 0.025868237107160547,
+      "grad_norm": 1.0431914329528809,
+      "learning_rate": 9.997427961155435e-06,
+      "loss": 0.8657,
+      "step": 470
+    },
+    {
+      "epoch": 0.025923275909516208,
+      "grad_norm": 0.9609962701797485,
+      "learning_rate": 9.997414040655198e-06,
+      "loss": 0.8864,
+      "step": 471
+    },
+    {
+      "epoch": 0.02597831471187187,
+      "grad_norm": 1.0829721689224243,
+      "learning_rate": 9.997400082595735e-06,
+      "loss": 0.9221,
+      "step": 472
+    },
+    {
+      "epoch": 0.02603335351422753,
+      "grad_norm": 0.992082953453064,
+      "learning_rate": 9.99738608697715e-06,
+      "loss": 0.8455,
+      "step": 473
+    },
+    {
+      "epoch": 0.026088392316583192,
+      "grad_norm": 1.0486301183700562,
+      "learning_rate": 9.997372053799547e-06,
+      "loss": 0.8729,
+      "step": 474
+    },
+    {
+      "epoch": 0.026143431118938854,
+      "grad_norm": 1.0328491926193237,
+      "learning_rate": 9.997357983063036e-06,
+      "loss": 0.8788,
+      "step": 475
+    },
+    {
+      "epoch": 0.02619846992129451,
+      "grad_norm": 0.963333249092102,
+      "learning_rate": 9.997343874767719e-06,
+      "loss": 0.892,
+      "step": 476
+    },
+    {
+      "epoch": 0.026253508723650173,
+      "grad_norm": 1.1606497764587402,
+      "learning_rate": 9.997329728913704e-06,
+      "loss": 0.9984,
+      "step": 477
+    },
+    {
+      "epoch": 0.026308547526005834,
+      "grad_norm": 1.241650104522705,
+      "learning_rate": 9.997315545501096e-06,
+      "loss": 0.946,
+      "step": 478
+    },
+    {
+      "epoch": 0.026363586328361496,
+      "grad_norm": 1.008004069328308,
+      "learning_rate": 9.99730132453e-06,
+      "loss": 0.849,
+      "step": 479
+    },
+    {
+      "epoch": 0.026418625130717157,
+      "grad_norm": 0.9883478879928589,
+      "learning_rate": 9.997287066000527e-06,
+      "loss": 0.9478,
+      "step": 480
+    },
+    {
+      "epoch": 0.026473663933072815,
+      "grad_norm": 1.0224446058273315,
+      "learning_rate": 9.997272769912783e-06,
+      "loss": 1.0318,
+      "step": 481
+    },
+    {
+      "epoch": 0.026528702735428476,
+      "grad_norm": 0.9412569403648376,
+      "learning_rate": 9.997258436266874e-06,
+      "loss": 0.9119,
+      "step": 482
+    },
+    {
+      "epoch": 0.026583741537784138,
+      "grad_norm": 0.9214537739753723,
+      "learning_rate": 9.997244065062906e-06,
+      "loss": 0.8785,
+      "step": 483
+    },
+    {
+      "epoch": 0.0266387803401398,
+      "grad_norm": 1.0015628337860107,
+      "learning_rate": 9.997229656300991e-06,
+      "loss": 0.8869,
+      "step": 484
+    },
+    {
+      "epoch": 0.02669381914249546,
+      "grad_norm": 0.8965190052986145,
+      "learning_rate": 9.997215209981237e-06,
+      "loss": 0.7009,
+      "step": 485
+    },
+    {
+      "epoch": 0.02674885794485112,
+      "grad_norm": 1.1976135969161987,
+      "learning_rate": 9.997200726103749e-06,
+      "loss": 0.9795,
+      "step": 486
+    },
+    {
+      "epoch": 0.02680389674720678,
+      "grad_norm": 0.864780843257904,
+      "learning_rate": 9.997186204668639e-06,
+      "loss": 0.7687,
+      "step": 487
+    },
+    {
+      "epoch": 0.02685893554956244,
+      "grad_norm": 0.9946566820144653,
+      "learning_rate": 9.997171645676013e-06,
+      "loss": 0.9672,
+      "step": 488
+    },
+    {
+      "epoch": 0.026913974351918103,
+      "grad_norm": 1.043835997581482,
+      "learning_rate": 9.997157049125985e-06,
+      "loss": 0.862,
+      "step": 489
+    },
+    {
+      "epoch": 0.026969013154273764,
+      "grad_norm": 0.9697456955909729,
+      "learning_rate": 9.99714241501866e-06,
+      "loss": 0.8368,
+      "step": 490
+    },
+    {
+      "epoch": 0.027024051956629425,
+      "grad_norm": 0.9975618124008179,
+      "learning_rate": 9.997127743354153e-06,
+      "loss": 0.8739,
+      "step": 491
+    },
+    {
+      "epoch": 0.027079090758985083,
+      "grad_norm": 1.0055313110351562,
+      "learning_rate": 9.99711303413257e-06,
+      "loss": 0.9227,
+      "step": 492
+    },
+    {
+      "epoch": 0.027134129561340745,
+      "grad_norm": 1.0418384075164795,
+      "learning_rate": 9.997098287354024e-06,
+      "loss": 0.9978,
+      "step": 493
+    },
+    {
+      "epoch": 0.027189168363696406,
+      "grad_norm": 0.8648970723152161,
+      "learning_rate": 9.997083503018625e-06,
+      "loss": 0.8363,
+      "step": 494
+    },
+    {
+      "epoch": 0.027244207166052067,
+      "grad_norm": 1.13506019115448,
+      "learning_rate": 9.997068681126483e-06,
+      "loss": 0.8851,
+      "step": 495
+    },
+    {
+      "epoch": 0.02729924596840773,
+      "grad_norm": 0.974400520324707,
+      "learning_rate": 9.997053821677712e-06,
+      "loss": 0.8533,
+      "step": 496
+    },
+    {
+      "epoch": 0.027354284770763387,
+      "grad_norm": 1.226507544517517,
+      "learning_rate": 9.997038924672419e-06,
+      "loss": 0.8586,
+      "step": 497
+    },
+    {
+      "epoch": 0.027409323573119048,
+      "grad_norm": 1.004753589630127,
+      "learning_rate": 9.997023990110721e-06,
+      "loss": 0.8974,
+      "step": 498
+    },
+    {
+      "epoch": 0.02746436237547471,
+      "grad_norm": 1.0492571592330933,
+      "learning_rate": 9.997009017992729e-06,
+      "loss": 0.8457,
+      "step": 499
+    },
+    {
+      "epoch": 0.02751940117783037,
+      "grad_norm": 1.0068167448043823,
+      "learning_rate": 9.996994008318554e-06,
+      "loss": 0.9608,
+      "step": 500
+    },
+    {
+      "epoch": 0.027574439980186032,
+      "grad_norm": 0.9686044454574585,
+      "learning_rate": 9.996978961088311e-06,
+      "loss": 0.9041,
+      "step": 501
+    },
+    {
+      "epoch": 0.027629478782541694,
+      "grad_norm": 1.281728744506836,
+      "learning_rate": 9.99696387630211e-06,
+      "loss": 0.9739,
+      "step": 502
+    },
+    {
+      "epoch": 0.02768451758489735,
+      "grad_norm": 0.9069758653640747,
+      "learning_rate": 9.996948753960065e-06,
+      "loss": 0.8467,
+      "step": 503
+    },
+    {
+      "epoch": 0.027739556387253013,
+      "grad_norm": 1.0337222814559937,
+      "learning_rate": 9.996933594062293e-06,
+      "loss": 0.9638,
+      "step": 504
+    },
+    {
+      "epoch": 0.027794595189608674,
+      "grad_norm": 0.9695359468460083,
+      "learning_rate": 9.996918396608905e-06,
+      "loss": 0.8986,
+      "step": 505
+    },
+    {
+      "epoch": 0.027849633991964336,
+      "grad_norm": 0.9120615124702454,
+      "learning_rate": 9.996903161600016e-06,
+      "loss": 0.9103,
+      "step": 506
+    },
+    {
+      "epoch": 0.027904672794319997,
+      "grad_norm": 0.9736546874046326,
+      "learning_rate": 9.996887889035741e-06,
+      "loss": 0.9308,
+      "step": 507
+    },
+    {
+      "epoch": 0.027959711596675655,
+      "grad_norm": 1.0184897184371948,
+      "learning_rate": 9.996872578916192e-06,
+      "loss": 0.8978,
+      "step": 508
+    },
+    {
+      "epoch": 0.028014750399031316,
+      "grad_norm": 0.9791838526725769,
+      "learning_rate": 9.996857231241489e-06,
+      "loss": 0.8639,
+      "step": 509
+    },
+    {
+      "epoch": 0.028069789201386978,
+      "grad_norm": 1.2985681295394897,
+      "learning_rate": 9.996841846011742e-06,
+      "loss": 0.9581,
+      "step": 510
+    },
+    {
+      "epoch": 0.02812482800374264,
+      "grad_norm": 1.0647368431091309,
+      "learning_rate": 9.996826423227071e-06,
+      "loss": 1.0565,
+      "step": 511
+    },
+    {
+      "epoch": 0.0281798668060983,
+      "grad_norm": 1.0336421728134155,
+      "learning_rate": 9.996810962887591e-06,
+      "loss": 1.008,
+      "step": 512
+    },
+    {
+      "epoch": 0.02823490560845396,
+      "grad_norm": 1.1838933229446411,
+      "learning_rate": 9.996795464993416e-06,
+      "loss": 0.8359,
+      "step": 513
+    },
+    {
+      "epoch": 0.02828994441080962,
+      "grad_norm": 0.9898360371589661,
+      "learning_rate": 9.996779929544663e-06,
+      "loss": 0.8501,
+      "step": 514
+    },
+    {
+      "epoch": 0.02834498321316528,
+      "grad_norm": 0.9836066365242004,
+      "learning_rate": 9.99676435654145e-06,
+      "loss": 0.8795,
+      "step": 515
+    },
+    {
+      "epoch": 0.028400022015520943,
+      "grad_norm": 1.0621601343154907,
+      "learning_rate": 9.996748745983895e-06,
+      "loss": 0.8746,
+      "step": 516
+    },
+    {
+      "epoch": 0.028455060817876604,
+      "grad_norm": 1.0082437992095947,
+      "learning_rate": 9.996733097872113e-06,
+      "loss": 0.9278,
+      "step": 517
+    },
+    {
+      "epoch": 0.028510099620232265,
+      "grad_norm": 0.9903931617736816,
+      "learning_rate": 9.996717412206222e-06,
+      "loss": 0.8264,
+      "step": 518
+    },
+    {
+      "epoch": 0.028565138422587923,
+      "grad_norm": 1.0797243118286133,
+      "learning_rate": 9.996701688986342e-06,
+      "loss": 1.0077,
+      "step": 519
+    },
+    {
+      "epoch": 0.028620177224943585,
+      "grad_norm": 1.147133231163025,
+      "learning_rate": 9.99668592821259e-06,
+      "loss": 0.9374,
+      "step": 520
+    },
+    {
+      "epoch": 0.028675216027299246,
+      "grad_norm": 0.9993947744369507,
+      "learning_rate": 9.996670129885082e-06,
+      "loss": 0.9562,
+      "step": 521
+    },
+    {
+      "epoch": 0.028730254829654907,
+      "grad_norm": 0.8580895066261292,
+      "learning_rate": 9.99665429400394e-06,
+      "loss": 0.7985,
+      "step": 522
+    },
+    {
+      "epoch": 0.02878529363201057,
+      "grad_norm": 0.9251388907432556,
+      "learning_rate": 9.996638420569281e-06,
+      "loss": 0.7323,
+      "step": 523
+    },
+    {
+      "epoch": 0.028840332434366227,
+      "grad_norm": 1.0010193586349487,
+      "learning_rate": 9.996622509581227e-06,
+      "loss": 0.9316,
+      "step": 524
+    },
+    {
+      "epoch": 0.028895371236721888,
+      "grad_norm": 0.9822579026222229,
+      "learning_rate": 9.996606561039894e-06,
+      "loss": 0.8978,
+      "step": 525
+    },
+    {
+      "epoch": 0.02895041003907755,
+      "grad_norm": 1.0760595798492432,
+      "learning_rate": 9.996590574945403e-06,
+      "loss": 0.9125,
+      "step": 526
+    },
+    {
+      "epoch": 0.02900544884143321,
+      "grad_norm": 1.138869285583496,
+      "learning_rate": 9.996574551297876e-06,
+      "loss": 0.8185,
+      "step": 527
+    },
+    {
+      "epoch": 0.029060487643788872,
+      "grad_norm": 1.002994179725647,
+      "learning_rate": 9.996558490097433e-06,
+      "loss": 0.9404,
+      "step": 528
+    },
+    {
+      "epoch": 0.02911552644614453,
+      "grad_norm": 0.9550611972808838,
+      "learning_rate": 9.996542391344194e-06,
+      "loss": 0.859,
+      "step": 529
+    },
+    {
+      "epoch": 0.02917056524850019,
+      "grad_norm": 0.9236055612564087,
+      "learning_rate": 9.996526255038277e-06,
+      "loss": 0.7758,
+      "step": 530
+    },
+    {
+      "epoch": 0.029225604050855853,
+      "grad_norm": 1.103966474533081,
+      "learning_rate": 9.996510081179808e-06,
+      "loss": 1.0147,
+      "step": 531
+    },
+    {
+      "epoch": 0.029280642853211514,
+      "grad_norm": 0.9884665012359619,
+      "learning_rate": 9.996493869768906e-06,
+      "loss": 0.8784,
+      "step": 532
+    },
+    {
+      "epoch": 0.029335681655567176,
+      "grad_norm": 0.9173223376274109,
+      "learning_rate": 9.996477620805694e-06,
+      "loss": 0.8741,
+      "step": 533
+    },
+    {
+      "epoch": 0.029390720457922837,
+      "grad_norm": 0.965548574924469,
+      "learning_rate": 9.996461334290294e-06,
+      "loss": 0.8989,
+      "step": 534
+    },
+    {
+      "epoch": 0.029445759260278495,
+      "grad_norm": 0.9939296245574951,
+      "learning_rate": 9.996445010222828e-06,
+      "loss": 0.8552,
+      "step": 535
+    },
+    {
+      "epoch": 0.029500798062634156,
+      "grad_norm": 1.0081578493118286,
+      "learning_rate": 9.996428648603417e-06,
+      "loss": 0.9138,
+      "step": 536
+    },
+    {
+      "epoch": 0.029555836864989818,
+      "grad_norm": 1.0139487981796265,
+      "learning_rate": 9.996412249432188e-06,
+      "loss": 0.9452,
+      "step": 537
+    },
+    {
+      "epoch": 0.02961087566734548,
+      "grad_norm": 0.9463647603988647,
+      "learning_rate": 9.996395812709262e-06,
+      "loss": 0.8721,
+      "step": 538
+    },
+    {
+      "epoch": 0.02966591446970114,
+      "grad_norm": 0.9981473684310913,
+      "learning_rate": 9.99637933843476e-06,
+      "loss": 0.7791,
+      "step": 539
+    },
+    {
+      "epoch": 0.0297209532720568,
+      "grad_norm": 1.1637190580368042,
+      "learning_rate": 9.996362826608812e-06,
+      "loss": 0.8798,
+      "step": 540
+    },
+    {
+      "epoch": 0.02977599207441246,
+      "grad_norm": 2.2887051105499268,
+      "learning_rate": 9.996346277231536e-06,
+      "loss": 0.9303,
+      "step": 541
+    },
+    {
+      "epoch": 0.02983103087676812,
+      "grad_norm": 0.9173391461372375,
+      "learning_rate": 9.99632969030306e-06,
+      "loss": 0.8627,
+      "step": 542
+    },
+    {
+      "epoch": 0.029886069679123783,
+      "grad_norm": 1.033355474472046,
+      "learning_rate": 9.996313065823506e-06,
+      "loss": 0.9906,
+      "step": 543
+    },
+    {
+      "epoch": 0.029941108481479444,
+      "grad_norm": 0.9286639094352722,
+      "learning_rate": 9.996296403793002e-06,
+      "loss": 0.7043,
+      "step": 544
+    },
+    {
+      "epoch": 0.029996147283835102,
+      "grad_norm": 0.963238000869751,
+      "learning_rate": 9.996279704211671e-06,
+      "loss": 1.0236,
+      "step": 545
+    },
+    {
+      "epoch": 0.030051186086190763,
+      "grad_norm": 1.0275089740753174,
+      "learning_rate": 9.99626296707964e-06,
+      "loss": 0.976,
+      "step": 546
+    },
+    {
+      "epoch": 0.030106224888546425,
+      "grad_norm": 1.0944674015045166,
+      "learning_rate": 9.996246192397032e-06,
+      "loss": 0.9209,
+      "step": 547
+    },
+    {
+      "epoch": 0.030161263690902086,
+      "grad_norm": 0.9620945453643799,
+      "learning_rate": 9.996229380163976e-06,
+      "loss": 0.8973,
+      "step": 548
+    },
+    {
+      "epoch": 0.030216302493257748,
+      "grad_norm": 1.032549500465393,
+      "learning_rate": 9.996212530380597e-06,
+      "loss": 0.892,
+      "step": 549
+    },
+    {
+      "epoch": 0.03027134129561341,
+      "grad_norm": 1.0433719158172607,
+      "learning_rate": 9.996195643047023e-06,
+      "loss": 0.8428,
+      "step": 550
+    },
+    {
+      "epoch": 0.030326380097969067,
+      "grad_norm": 1.1541085243225098,
+      "learning_rate": 9.996178718163378e-06,
+      "loss": 0.9084,
+      "step": 551
+    },
+    {
+      "epoch": 0.03038141890032473,
+      "grad_norm": 0.9386873245239258,
+      "learning_rate": 9.996161755729793e-06,
+      "loss": 0.9246,
+      "step": 552
+    },
+    {
+      "epoch": 0.03043645770268039,
+      "grad_norm": 1.092236042022705,
+      "learning_rate": 9.996144755746393e-06,
+      "loss": 0.8419,
+      "step": 553
+    },
+    {
+      "epoch": 0.03049149650503605,
+      "grad_norm": 0.9517606496810913,
+      "learning_rate": 9.996127718213306e-06,
+      "loss": 0.9002,
+      "step": 554
+    },
+    {
+      "epoch": 0.030546535307391712,
+      "grad_norm": 0.965972900390625,
+      "learning_rate": 9.996110643130661e-06,
+      "loss": 0.9197,
+      "step": 555
+    },
+    {
+      "epoch": 0.03060157410974737,
+      "grad_norm": 0.9396095275878906,
+      "learning_rate": 9.996093530498586e-06,
+      "loss": 0.8686,
+      "step": 556
+    },
+    {
+      "epoch": 0.030656612912103032,
+      "grad_norm": 1.0154120922088623,
+      "learning_rate": 9.99607638031721e-06,
+      "loss": 0.9773,
+      "step": 557
+    },
+    {
+      "epoch": 0.030711651714458693,
+      "grad_norm": 1.3572301864624023,
+      "learning_rate": 9.99605919258666e-06,
+      "loss": 0.911,
+      "step": 558
+    },
+    {
+      "epoch": 0.030766690516814355,
+      "grad_norm": 0.968278169631958,
+      "learning_rate": 9.996041967307066e-06,
+      "loss": 0.7704,
+      "step": 559
+    },
+    {
+      "epoch": 0.030821729319170016,
+      "grad_norm": 0.9867869019508362,
+      "learning_rate": 9.99602470447856e-06,
+      "loss": 0.873,
+      "step": 560
+    },
+    {
+      "epoch": 0.030876768121525677,
+      "grad_norm": 1.056450605392456,
+      "learning_rate": 9.996007404101269e-06,
+      "loss": 0.941,
+      "step": 561
+    },
+    {
+      "epoch": 0.030931806923881335,
+      "grad_norm": 1.0419799089431763,
+      "learning_rate": 9.995990066175321e-06,
+      "loss": 0.957,
+      "step": 562
+    },
+    {
+      "epoch": 0.030986845726236997,
+      "grad_norm": 0.9789314866065979,
+      "learning_rate": 9.995972690700852e-06,
+      "loss": 0.9229,
+      "step": 563
+    },
+    {
+      "epoch": 0.031041884528592658,
+      "grad_norm": 0.917783796787262,
+      "learning_rate": 9.995955277677989e-06,
+      "loss": 0.8186,
+      "step": 564
+    },
+    {
+      "epoch": 0.03109692333094832,
+      "grad_norm": 1.0231432914733887,
+      "learning_rate": 9.995937827106863e-06,
+      "loss": 0.8624,
+      "step": 565
+    },
+    {
+      "epoch": 0.03115196213330398,
+      "grad_norm": 0.9552083015441895,
+      "learning_rate": 9.995920338987605e-06,
+      "loss": 0.7967,
+      "step": 566
+    },
+    {
+      "epoch": 0.03120700093565964,
+      "grad_norm": 0.9441083669662476,
+      "learning_rate": 9.995902813320349e-06,
+      "loss": 0.8471,
+      "step": 567
+    },
+    {
+      "epoch": 0.0312620397380153,
+      "grad_norm": 1.0025299787521362,
+      "learning_rate": 9.995885250105223e-06,
+      "loss": 0.8646,
+      "step": 568
+    },
+    {
+      "epoch": 0.03131707854037096,
+      "grad_norm": 0.8997280597686768,
+      "learning_rate": 9.99586764934236e-06,
+      "loss": 0.8736,
+      "step": 569
+    },
+    {
+      "epoch": 0.03137211734272662,
+      "grad_norm": 0.9090663194656372,
+      "learning_rate": 9.995850011031896e-06,
+      "loss": 0.8548,
+      "step": 570
+    },
+    {
+      "epoch": 0.031427156145082284,
+      "grad_norm": 0.9641294479370117,
+      "learning_rate": 9.995832335173959e-06,
+      "loss": 0.8667,
+      "step": 571
+    },
+    {
+      "epoch": 0.031482194947437946,
+      "grad_norm": 0.9165804982185364,
+      "learning_rate": 9.995814621768682e-06,
+      "loss": 0.803,
+      "step": 572
+    },
+    {
+      "epoch": 0.03153723374979361,
+      "grad_norm": 0.9672492742538452,
+      "learning_rate": 9.995796870816202e-06,
+      "loss": 0.8335,
+      "step": 573
+    },
+    {
+      "epoch": 0.03159227255214927,
+      "grad_norm": 0.9359404444694519,
+      "learning_rate": 9.995779082316648e-06,
+      "loss": 0.8294,
+      "step": 574
+    },
+    {
+      "epoch": 0.03164731135450492,
+      "grad_norm": 0.926925003528595,
+      "learning_rate": 9.995761256270157e-06,
+      "loss": 0.7714,
+      "step": 575
+    },
+    {
+      "epoch": 0.031702350156860584,
+      "grad_norm": 1.1848629713058472,
+      "learning_rate": 9.995743392676862e-06,
+      "loss": 0.8925,
+      "step": 576
+    },
+    {
+      "epoch": 0.031757388959216246,
+      "grad_norm": 0.9624786972999573,
+      "learning_rate": 9.995725491536897e-06,
+      "loss": 0.9292,
+      "step": 577
+    },
+    {
+      "epoch": 0.03181242776157191,
+      "grad_norm": 0.9479736089706421,
+      "learning_rate": 9.995707552850396e-06,
+      "loss": 0.8797,
+      "step": 578
+    },
+    {
+      "epoch": 0.03186746656392757,
+      "grad_norm": 0.9551546573638916,
+      "learning_rate": 9.995689576617494e-06,
+      "loss": 0.8793,
+      "step": 579
+    },
+    {
+      "epoch": 0.03192250536628323,
+      "grad_norm": 0.9210056662559509,
+      "learning_rate": 9.995671562838325e-06,
+      "loss": 0.9714,
+      "step": 580
+    },
+    {
+      "epoch": 0.03197754416863889,
+      "grad_norm": 1.063117504119873,
+      "learning_rate": 9.995653511513029e-06,
+      "loss": 0.9608,
+      "step": 581
+    },
+    {
+      "epoch": 0.03203258297099455,
+      "grad_norm": 0.9426459670066833,
+      "learning_rate": 9.995635422641736e-06,
+      "loss": 0.9102,
+      "step": 582
+    },
+    {
+      "epoch": 0.032087621773350214,
+      "grad_norm": 1.0176693201065063,
+      "learning_rate": 9.995617296224584e-06,
+      "loss": 0.9109,
+      "step": 583
+    },
+    {
+      "epoch": 0.032142660575705875,
+      "grad_norm": 0.9457042217254639,
+      "learning_rate": 9.995599132261711e-06,
+      "loss": 0.9017,
+      "step": 584
+    },
+    {
+      "epoch": 0.03219769937806154,
+      "grad_norm": 1.5851638317108154,
+      "learning_rate": 9.995580930753252e-06,
+      "loss": 0.967,
+      "step": 585
+    },
+    {
+      "epoch": 0.03225273818041719,
+      "grad_norm": 0.9961487054824829,
+      "learning_rate": 9.995562691699345e-06,
+      "loss": 0.9396,
+      "step": 586
+    },
+    {
+      "epoch": 0.03230777698277285,
+      "grad_norm": 0.9892112016677856,
+      "learning_rate": 9.995544415100125e-06,
+      "loss": 0.9058,
+      "step": 587
+    },
+    {
+      "epoch": 0.032362815785128514,
+      "grad_norm": 0.9052272439002991,
+      "learning_rate": 9.99552610095573e-06,
+      "loss": 0.9194,
+      "step": 588
+    },
+    {
+      "epoch": 0.032417854587484175,
+      "grad_norm": 0.8381399512290955,
+      "learning_rate": 9.995507749266297e-06,
+      "loss": 0.7465,
+      "step": 589
+    },
+    {
+      "epoch": 0.03247289338983984,
+      "grad_norm": 1.018964171409607,
+      "learning_rate": 9.995489360031969e-06,
+      "loss": 0.841,
+      "step": 590
+    },
+    {
+      "epoch": 0.0325279321921955,
+      "grad_norm": 0.908311128616333,
+      "learning_rate": 9.995470933252876e-06,
+      "loss": 0.8592,
+      "step": 591
+    },
+    {
+      "epoch": 0.03258297099455116,
+      "grad_norm": 1.2986040115356445,
+      "learning_rate": 9.995452468929162e-06,
+      "loss": 0.8341,
+      "step": 592
+    },
+    {
+      "epoch": 0.03263800979690682,
+      "grad_norm": 1.6565190553665161,
+      "learning_rate": 9.995433967060966e-06,
+      "loss": 0.8681,
+      "step": 593
+    },
+    {
+      "epoch": 0.03269304859926248,
+      "grad_norm": 0.9725674390792847,
+      "learning_rate": 9.995415427648423e-06,
+      "loss": 0.8449,
+      "step": 594
+    },
+    {
+      "epoch": 0.032748087401618144,
+      "grad_norm": 0.8683852553367615,
+      "learning_rate": 9.995396850691677e-06,
+      "loss": 0.8478,
+      "step": 595
+    },
+    {
+      "epoch": 0.0328031262039738,
+      "grad_norm": 0.9912856817245483,
+      "learning_rate": 9.995378236190862e-06,
+      "loss": 0.8912,
+      "step": 596
+    },
+    {
+      "epoch": 0.03285816500632946,
+      "grad_norm": 0.9396800398826599,
+      "learning_rate": 9.995359584146125e-06,
+      "loss": 0.856,
+      "step": 597
+    },
+    {
+      "epoch": 0.03291320380868512,
+      "grad_norm": 1.385006308555603,
+      "learning_rate": 9.995340894557601e-06,
+      "loss": 0.9633,
+      "step": 598
+    },
+    {
+      "epoch": 0.03296824261104078,
+      "grad_norm": 0.8982875943183899,
+      "learning_rate": 9.995322167425433e-06,
+      "loss": 0.9244,
+      "step": 599
+    },
+    {
+      "epoch": 0.033023281413396444,
+      "grad_norm": 0.8981022834777832,
+      "learning_rate": 9.995303402749759e-06,
+      "loss": 0.8854,
+      "step": 600
+    },
+    {
+      "epoch": 0.033078320215752105,
+      "grad_norm": 0.9917197227478027,
+      "learning_rate": 9.995284600530724e-06,
+      "loss": 1.0086,
+      "step": 601
+    },
+    {
+      "epoch": 0.033133359018107766,
+      "grad_norm": 1.0540626049041748,
+      "learning_rate": 9.995265760768464e-06,
+      "loss": 1.0022,
+      "step": 602
+    },
+    {
+      "epoch": 0.03318839782046343,
+      "grad_norm": 0.9523479342460632,
+      "learning_rate": 9.995246883463126e-06,
+      "loss": 0.9893,
+      "step": 603
+    },
+    {
+      "epoch": 0.03324343662281909,
+      "grad_norm": 0.9824770092964172,
+      "learning_rate": 9.99522796861485e-06,
+      "loss": 0.8385,
+      "step": 604
+    },
+    {
+      "epoch": 0.03329847542517475,
+      "grad_norm": 1.0968893766403198,
+      "learning_rate": 9.995209016223776e-06,
+      "loss": 1.0109,
+      "step": 605
+    },
+    {
+      "epoch": 0.03335351422753041,
+      "grad_norm": 0.9115625023841858,
+      "learning_rate": 9.995190026290049e-06,
+      "loss": 0.8656,
+      "step": 606
+    },
+    {
+      "epoch": 0.033408553029886066,
+      "grad_norm": 0.9795814156532288,
+      "learning_rate": 9.99517099881381e-06,
+      "loss": 0.8941,
+      "step": 607
+    },
+    {
+      "epoch": 0.03346359183224173,
+      "grad_norm": 0.9317291378974915,
+      "learning_rate": 9.995151933795204e-06,
+      "loss": 0.7819,
+      "step": 608
+    },
+    {
+      "epoch": 0.03351863063459739,
+      "grad_norm": 0.9936283230781555,
+      "learning_rate": 9.995132831234373e-06,
+      "loss": 0.8674,
+      "step": 609
+    },
+    {
+      "epoch": 0.03357366943695305,
+      "grad_norm": 0.9872812032699585,
+      "learning_rate": 9.995113691131462e-06,
+      "loss": 0.9038,
+      "step": 610
+    },
+    {
+      "epoch": 0.03362870823930871,
+      "grad_norm": 0.9516895413398743,
+      "learning_rate": 9.995094513486611e-06,
+      "loss": 0.9038,
+      "step": 611
+    },
+    {
+      "epoch": 0.03368374704166437,
+      "grad_norm": 1.090579867362976,
+      "learning_rate": 9.995075298299968e-06,
+      "loss": 0.9587,
+      "step": 612
+    },
+    {
+      "epoch": 0.033738785844020035,
+      "grad_norm": 1.021398663520813,
+      "learning_rate": 9.995056045571677e-06,
+      "loss": 0.9569,
+      "step": 613
+    },
+    {
+      "epoch": 0.033793824646375696,
+      "grad_norm": 1.009657382965088,
+      "learning_rate": 9.99503675530188e-06,
+      "loss": 0.8346,
+      "step": 614
+    },
+    {
+      "epoch": 0.03384886344873136,
+      "grad_norm": 1.0478712320327759,
+      "learning_rate": 9.995017427490725e-06,
+      "loss": 1.0566,
+      "step": 615
+    },
+    {
+      "epoch": 0.03390390225108702,
+      "grad_norm": 1.1391830444335938,
+      "learning_rate": 9.994998062138355e-06,
+      "loss": 1.0727,
+      "step": 616
+    },
+    {
+      "epoch": 0.03395894105344268,
+      "grad_norm": 1.0172302722930908,
+      "learning_rate": 9.994978659244918e-06,
+      "loss": 0.7869,
+      "step": 617
+    },
+    {
+      "epoch": 0.034013979855798335,
+      "grad_norm": 1.0532630681991577,
+      "learning_rate": 9.994959218810558e-06,
+      "loss": 0.8626,
+      "step": 618
+    },
+    {
+      "epoch": 0.034069018658153996,
+      "grad_norm": 0.8300478458404541,
+      "learning_rate": 9.99493974083542e-06,
+      "loss": 0.8166,
+      "step": 619
+    },
+    {
+      "epoch": 0.03412405746050966,
+      "grad_norm": 1.0613664388656616,
+      "learning_rate": 9.994920225319656e-06,
+      "loss": 0.8899,
+      "step": 620
+    },
+    {
+      "epoch": 0.03417909626286532,
+      "grad_norm": 0.9827042818069458,
+      "learning_rate": 9.994900672263406e-06,
+      "loss": 0.8243,
+      "step": 621
+    },
+    {
+      "epoch": 0.03423413506522098,
+      "grad_norm": 0.8790082931518555,
+      "learning_rate": 9.994881081666818e-06,
+      "loss": 0.8153,
+      "step": 622
+    },
+    {
+      "epoch": 0.03428917386757664,
+      "grad_norm": 1.033378005027771,
+      "learning_rate": 9.994861453530044e-06,
+      "loss": 0.8916,
+      "step": 623
+    },
+    {
+      "epoch": 0.0343442126699323,
+      "grad_norm": 0.9547238349914551,
+      "learning_rate": 9.994841787853227e-06,
+      "loss": 0.9141,
+      "step": 624
+    },
+    {
+      "epoch": 0.034399251472287964,
+      "grad_norm": 0.9606438279151917,
+      "learning_rate": 9.994822084636514e-06,
+      "loss": 0.9435,
+      "step": 625
+    },
+    {
+      "epoch": 0.034454290274643626,
+      "grad_norm": 0.8461503982543945,
+      "learning_rate": 9.994802343880059e-06,
+      "loss": 0.7914,
+      "step": 626
+    },
+    {
+      "epoch": 0.03450932907699929,
+      "grad_norm": 1.144538402557373,
+      "learning_rate": 9.994782565584004e-06,
+      "loss": 0.8025,
+      "step": 627
+    },
+    {
+      "epoch": 0.03456436787935495,
+      "grad_norm": 1.0099962949752808,
+      "learning_rate": 9.994762749748502e-06,
+      "loss": 0.9607,
+      "step": 628
+    },
+    {
+      "epoch": 0.0346194066817106,
+      "grad_norm": 0.9822041988372803,
+      "learning_rate": 9.9947428963737e-06,
+      "loss": 0.9216,
+      "step": 629
+    },
+    {
+      "epoch": 0.034674445484066264,
+      "grad_norm": 0.9056866765022278,
+      "learning_rate": 9.994723005459746e-06,
+      "loss": 0.7913,
+      "step": 630
+    },
+    {
+      "epoch": 0.034729484286421926,
+      "grad_norm": 1.0099287033081055,
+      "learning_rate": 9.994703077006792e-06,
+      "loss": 0.9937,
+      "step": 631
+    },
+    {
+      "epoch": 0.03478452308877759,
+      "grad_norm": 0.9559167623519897,
+      "learning_rate": 9.994683111014984e-06,
+      "loss": 0.9774,
+      "step": 632
+    },
+    {
+      "epoch": 0.03483956189113325,
+      "grad_norm": 1.0359059572219849,
+      "learning_rate": 9.994663107484478e-06,
+      "loss": 0.9062,
+      "step": 633
+    },
+    {
+      "epoch": 0.03489460069348891,
+      "grad_norm": 0.8803057074546814,
+      "learning_rate": 9.99464306641542e-06,
+      "loss": 0.9638,
+      "step": 634
+    },
+    {
+      "epoch": 0.03494963949584457,
+      "grad_norm": 1.0926579236984253,
+      "learning_rate": 9.994622987807962e-06,
+      "loss": 1.0467,
+      "step": 635
+    },
+    {
+      "epoch": 0.03500467829820023,
+      "grad_norm": 1.0051401853561401,
+      "learning_rate": 9.994602871662253e-06,
+      "loss": 0.8717,
+      "step": 636
+    },
+    {
+      "epoch": 0.035059717100555894,
+      "grad_norm": 1.2007508277893066,
+      "learning_rate": 9.994582717978448e-06,
+      "loss": 0.8004,
+      "step": 637
+    },
+    {
+      "epoch": 0.035114755902911556,
+      "grad_norm": 0.8826266527175903,
+      "learning_rate": 9.994562526756695e-06,
+      "loss": 0.8888,
+      "step": 638
+    },
+    {
+      "epoch": 0.03516979470526721,
+      "grad_norm": 0.9953717589378357,
+      "learning_rate": 9.994542297997147e-06,
+      "loss": 0.8999,
+      "step": 639
+    },
+    {
+      "epoch": 0.03522483350762287,
+      "grad_norm": 1.0203614234924316,
+      "learning_rate": 9.994522031699958e-06,
+      "loss": 0.8241,
+      "step": 640
+    },
+    {
+      "epoch": 0.03527987230997853,
+      "grad_norm": 0.8760203719139099,
+      "learning_rate": 9.994501727865276e-06,
+      "loss": 0.7893,
+      "step": 641
+    },
+    {
+      "epoch": 0.035334911112334194,
+      "grad_norm": 1.024888277053833,
+      "learning_rate": 9.994481386493257e-06,
+      "loss": 0.9865,
+      "step": 642
+    },
+    {
+      "epoch": 0.035389949914689856,
+      "grad_norm": 0.907454788684845,
+      "learning_rate": 9.994461007584052e-06,
+      "loss": 0.891,
+      "step": 643
+    },
+    {
+      "epoch": 0.03544498871704552,
+      "grad_norm": 1.0400965213775635,
+      "learning_rate": 9.994440591137816e-06,
+      "loss": 0.9345,
+      "step": 644
+    },
+    {
+      "epoch": 0.03550002751940118,
+      "grad_norm": 0.9816616177558899,
+      "learning_rate": 9.9944201371547e-06,
+      "loss": 0.91,
+      "step": 645
+    },
+    {
+      "epoch": 0.03555506632175684,
+      "grad_norm": 1.0528117418289185,
+      "learning_rate": 9.99439964563486e-06,
+      "loss": 0.952,
+      "step": 646
+    },
+    {
+      "epoch": 0.0356101051241125,
+      "grad_norm": 0.9802080988883972,
+      "learning_rate": 9.99437911657845e-06,
+      "loss": 0.9392,
+      "step": 647
+    },
+    {
+      "epoch": 0.03566514392646816,
+      "grad_norm": 0.9580393433570862,
+      "learning_rate": 9.994358549985623e-06,
+      "loss": 0.874,
+      "step": 648
+    },
+    {
+      "epoch": 0.035720182728823824,
+      "grad_norm": 0.8935576677322388,
+      "learning_rate": 9.994337945856533e-06,
+      "loss": 0.8435,
+      "step": 649
+    },
+    {
+      "epoch": 0.03577522153117948,
+      "grad_norm": 1.009699821472168,
+      "learning_rate": 9.994317304191337e-06,
+      "loss": 0.9436,
+      "step": 650
+    },
+    {
+      "epoch": 0.03583026033353514,
+      "grad_norm": 0.9126121401786804,
+      "learning_rate": 9.994296624990188e-06,
+      "loss": 0.8424,
+      "step": 651
+    },
+    {
+      "epoch": 0.0358852991358908,
+      "grad_norm": 0.9555553197860718,
+      "learning_rate": 9.994275908253243e-06,
+      "loss": 0.93,
+      "step": 652
+    },
+    {
+      "epoch": 0.03594033793824646,
+      "grad_norm": 0.8359857797622681,
+      "learning_rate": 9.994255153980658e-06,
+      "loss": 0.6326,
+      "step": 653
+    },
+    {
+      "epoch": 0.035995376740602124,
+      "grad_norm": 0.8918783664703369,
+      "learning_rate": 9.994234362172587e-06,
+      "loss": 0.8287,
+      "step": 654
+    },
+    {
+      "epoch": 0.036050415542957785,
+      "grad_norm": 0.9878549575805664,
+      "learning_rate": 9.994213532829188e-06,
+      "loss": 0.8841,
+      "step": 655
+    },
+    {
+      "epoch": 0.03610545434531345,
+      "grad_norm": 0.9504040479660034,
+      "learning_rate": 9.994192665950617e-06,
+      "loss": 1.0182,
+      "step": 656
+    },
+    {
+      "epoch": 0.03616049314766911,
+      "grad_norm": 0.9531422257423401,
+      "learning_rate": 9.99417176153703e-06,
+      "loss": 0.8504,
+      "step": 657
+    },
+    {
+      "epoch": 0.03621553195002477,
+      "grad_norm": 0.9580292105674744,
+      "learning_rate": 9.994150819588587e-06,
+      "loss": 0.8048,
+      "step": 658
+    },
+    {
+      "epoch": 0.03627057075238043,
+      "grad_norm": 0.9786819815635681,
+      "learning_rate": 9.99412984010544e-06,
+      "loss": 0.9124,
+      "step": 659
+    },
+    {
+      "epoch": 0.03632560955473609,
+      "grad_norm": 0.9733422994613647,
+      "learning_rate": 9.994108823087751e-06,
+      "loss": 0.8868,
+      "step": 660
+    },
+    {
+      "epoch": 0.03638064835709175,
+      "grad_norm": 1.093173623085022,
+      "learning_rate": 9.994087768535679e-06,
+      "loss": 0.9428,
+      "step": 661
+    },
+    {
+      "epoch": 0.03643568715944741,
+      "grad_norm": 0.9067148566246033,
+      "learning_rate": 9.994066676449378e-06,
+      "loss": 0.8838,
+      "step": 662
+    },
+    {
+      "epoch": 0.03649072596180307,
+      "grad_norm": 0.9509521722793579,
+      "learning_rate": 9.99404554682901e-06,
+      "loss": 0.9034,
+      "step": 663
+    },
+    {
+      "epoch": 0.03654576476415873,
+      "grad_norm": 0.9523824453353882,
+      "learning_rate": 9.994024379674731e-06,
+      "loss": 0.9623,
+      "step": 664
+    },
+    {
+      "epoch": 0.03660080356651439,
+      "grad_norm": 0.987276554107666,
+      "learning_rate": 9.994003174986703e-06,
+      "loss": 0.8817,
+      "step": 665
+    },
+    {
+      "epoch": 0.036655842368870054,
+      "grad_norm": 0.9500744342803955,
+      "learning_rate": 9.993981932765083e-06,
+      "loss": 0.9742,
+      "step": 666
+    },
+    {
+      "epoch": 0.036710881171225715,
+      "grad_norm": 0.9420705437660217,
+      "learning_rate": 9.993960653010034e-06,
+      "loss": 0.9657,
+      "step": 667
+    },
+    {
+      "epoch": 0.036765919973581376,
+      "grad_norm": 0.9443248510360718,
+      "learning_rate": 9.99393933572171e-06,
+      "loss": 0.8468,
+      "step": 668
+    },
+    {
+      "epoch": 0.03682095877593704,
+      "grad_norm": 0.9666558504104614,
+      "learning_rate": 9.993917980900276e-06,
+      "loss": 0.9871,
+      "step": 669
+    },
+    {
+      "epoch": 0.0368759975782927,
+      "grad_norm": 1.0236201286315918,
+      "learning_rate": 9.993896588545892e-06,
+      "loss": 0.9814,
+      "step": 670
+    },
+    {
+      "epoch": 0.03693103638064836,
+      "grad_norm": 1.016190528869629,
+      "learning_rate": 9.993875158658716e-06,
+      "loss": 1.0156,
+      "step": 671
+    },
+    {
+      "epoch": 0.036986075183004015,
+      "grad_norm": 0.9296661019325256,
+      "learning_rate": 9.993853691238913e-06,
+      "loss": 0.7956,
+      "step": 672
+    },
+    {
+      "epoch": 0.037041113985359676,
+      "grad_norm": 0.9276684522628784,
+      "learning_rate": 9.993832186286643e-06,
+      "loss": 0.9253,
+      "step": 673
+    },
+    {
+      "epoch": 0.03709615278771534,
+      "grad_norm": 0.8588787913322449,
+      "learning_rate": 9.993810643802065e-06,
+      "loss": 0.7878,
+      "step": 674
+    },
+    {
+      "epoch": 0.037151191590071,
+      "grad_norm": 0.9955212473869324,
+      "learning_rate": 9.993789063785344e-06,
+      "loss": 0.8711,
+      "step": 675
+    },
+    {
+      "epoch": 0.03720623039242666,
+      "grad_norm": 0.925578236579895,
+      "learning_rate": 9.993767446236642e-06,
+      "loss": 0.9431,
+      "step": 676
+    },
+    {
+      "epoch": 0.03726126919478232,
+      "grad_norm": 0.9610552787780762,
+      "learning_rate": 9.99374579115612e-06,
+      "loss": 0.887,
+      "step": 677
+    },
+    {
+      "epoch": 0.03731630799713798,
+      "grad_norm": 1.0052428245544434,
+      "learning_rate": 9.99372409854394e-06,
+      "loss": 0.8751,
+      "step": 678
+    },
+    {
+      "epoch": 0.037371346799493645,
+      "grad_norm": 0.9503066539764404,
+      "learning_rate": 9.99370236840027e-06,
+      "loss": 0.8556,
+      "step": 679
+    },
+    {
+      "epoch": 0.037426385601849306,
+      "grad_norm": 2.426232099533081,
+      "learning_rate": 9.993680600725266e-06,
+      "loss": 0.9077,
+      "step": 680
+    },
+    {
+      "epoch": 0.03748142440420497,
+      "grad_norm": 0.9119723439216614,
+      "learning_rate": 9.993658795519096e-06,
+      "loss": 0.8575,
+      "step": 681
+    },
+    {
+      "epoch": 0.03753646320656062,
+      "grad_norm": 0.9688286781311035,
+      "learning_rate": 9.993636952781923e-06,
+      "loss": 0.8921,
+      "step": 682
+    },
+    {
+      "epoch": 0.03759150200891628,
+      "grad_norm": 1.030013084411621,
+      "learning_rate": 9.993615072513913e-06,
+      "loss": 0.8622,
+      "step": 683
+    },
+    {
+      "epoch": 0.037646540811271945,
+      "grad_norm": 1.055187463760376,
+      "learning_rate": 9.993593154715228e-06,
+      "loss": 0.9251,
+      "step": 684
+    },
+    {
+      "epoch": 0.037701579613627606,
+      "grad_norm": 1.0518591403961182,
+      "learning_rate": 9.993571199386032e-06,
+      "loss": 0.9575,
+      "step": 685
+    },
+    {
+      "epoch": 0.03775661841598327,
+      "grad_norm": 0.9232666492462158,
+      "learning_rate": 9.993549206526495e-06,
+      "loss": 0.8522,
+      "step": 686
+    },
+    {
+      "epoch": 0.03781165721833893,
+      "grad_norm": 1.0212332010269165,
+      "learning_rate": 9.993527176136775e-06,
+      "loss": 0.9358,
+      "step": 687
+    },
+    {
+      "epoch": 0.03786669602069459,
+      "grad_norm": 0.9137141108512878,
+      "learning_rate": 9.993505108217045e-06,
+      "loss": 0.8561,
+      "step": 688
+    },
+    {
+      "epoch": 0.03792173482305025,
+      "grad_norm": 1.0069375038146973,
+      "learning_rate": 9.993483002767465e-06,
+      "loss": 0.8274,
+      "step": 689
+    },
+    {
+      "epoch": 0.03797677362540591,
+      "grad_norm": 0.9820672869682312,
+      "learning_rate": 9.993460859788204e-06,
+      "loss": 0.907,
+      "step": 690
+    },
+    {
+      "epoch": 0.038031812427761574,
+      "grad_norm": 1.0042002201080322,
+      "learning_rate": 9.993438679279428e-06,
+      "loss": 0.9263,
+      "step": 691
+    },
+    {
+      "epoch": 0.038086851230117236,
+      "grad_norm": 0.9733695983886719,
+      "learning_rate": 9.993416461241304e-06,
+      "loss": 0.8455,
+      "step": 692
+    },
+    {
+      "epoch": 0.03814189003247289,
+      "grad_norm": 0.9106015563011169,
+      "learning_rate": 9.993394205673996e-06,
+      "loss": 0.8469,
+      "step": 693
+    },
+    {
+      "epoch": 0.03819692883482855,
+      "grad_norm": 0.9802660346031189,
+      "learning_rate": 9.993371912577677e-06,
+      "loss": 0.8662,
+      "step": 694
+    },
+    {
+      "epoch": 0.03825196763718421,
+      "grad_norm": 0.9183964729309082,
+      "learning_rate": 9.99334958195251e-06,
+      "loss": 0.8968,
+      "step": 695
+    },
+    {
+      "epoch": 0.038307006439539874,
+      "grad_norm": 0.9572185277938843,
+      "learning_rate": 9.993327213798663e-06,
+      "loss": 0.953,
+      "step": 696
+    },
+    {
+      "epoch": 0.038362045241895536,
+      "grad_norm": 1.4480071067810059,
+      "learning_rate": 9.993304808116307e-06,
+      "loss": 1.1131,
+      "step": 697
+    },
+    {
+      "epoch": 0.0384170840442512,
+      "grad_norm": 0.9297361969947815,
+      "learning_rate": 9.993282364905607e-06,
+      "loss": 0.884,
+      "step": 698
+    },
+    {
+      "epoch": 0.03847212284660686,
+      "grad_norm": 0.9400073885917664,
+      "learning_rate": 9.993259884166735e-06,
+      "loss": 0.932,
+      "step": 699
+    },
+    {
+      "epoch": 0.03852716164896252,
+      "grad_norm": 0.9231798052787781,
+      "learning_rate": 9.993237365899858e-06,
+      "loss": 0.8981,
+      "step": 700
+    },
+    {
+      "epoch": 0.03858220045131818,
+      "grad_norm": 0.8233712911605835,
+      "learning_rate": 9.993214810105144e-06,
+      "loss": 0.8218,
+      "step": 701
+    },
+    {
+      "epoch": 0.03863723925367384,
+      "grad_norm": 1.0997854471206665,
+      "learning_rate": 9.993192216782768e-06,
+      "loss": 0.9298,
+      "step": 702
+    },
+    {
+      "epoch": 0.038692278056029504,
+      "grad_norm": 0.9570802450180054,
+      "learning_rate": 9.993169585932893e-06,
+      "loss": 0.7815,
+      "step": 703
+    },
+    {
+      "epoch": 0.03874731685838516,
+      "grad_norm": 0.9913730025291443,
+      "learning_rate": 9.993146917555692e-06,
+      "loss": 0.9621,
+      "step": 704
+    },
+    {
+      "epoch": 0.03880235566074082,
+      "grad_norm": 1.088767409324646,
+      "learning_rate": 9.993124211651334e-06,
+      "loss": 0.9295,
+      "step": 705
+    },
+    {
+      "epoch": 0.03885739446309648,
+      "grad_norm": 0.8199124336242676,
+      "learning_rate": 9.993101468219995e-06,
+      "loss": 0.7613,
+      "step": 706
+    },
+    {
+      "epoch": 0.03891243326545214,
+      "grad_norm": 1.112566351890564,
+      "learning_rate": 9.99307868726184e-06,
+      "loss": 0.791,
+      "step": 707
+    },
+    {
+      "epoch": 0.038967472067807804,
+      "grad_norm": 0.9372578859329224,
+      "learning_rate": 9.99305586877704e-06,
+      "loss": 0.8567,
+      "step": 708
+    },
+    {
+      "epoch": 0.039022510870163465,
+      "grad_norm": 1.0167721509933472,
+      "learning_rate": 9.99303301276577e-06,
+      "loss": 0.9787,
+      "step": 709
+    },
+    {
+      "epoch": 0.03907754967251913,
+      "grad_norm": 1.3526856899261475,
+      "learning_rate": 9.993010119228202e-06,
+      "loss": 1.2215,
+      "step": 710
+    },
+    {
+      "epoch": 0.03913258847487479,
+      "grad_norm": 0.8819016814231873,
+      "learning_rate": 9.992987188164505e-06,
+      "loss": 0.7736,
+      "step": 711
+    },
+    {
+      "epoch": 0.03918762727723045,
+      "grad_norm": 1.0033677816390991,
+      "learning_rate": 9.992964219574852e-06,
+      "loss": 0.9919,
+      "step": 712
+    },
+    {
+      "epoch": 0.03924266607958611,
+      "grad_norm": 0.894926130771637,
+      "learning_rate": 9.992941213459417e-06,
+      "loss": 0.9058,
+      "step": 713
+    },
+    {
+      "epoch": 0.03929770488194177,
+      "grad_norm": 0.9481377005577087,
+      "learning_rate": 9.992918169818373e-06,
+      "loss": 0.8436,
+      "step": 714
+    },
+    {
+      "epoch": 0.03935274368429743,
+      "grad_norm": 0.9312933087348938,
+      "learning_rate": 9.992895088651893e-06,
+      "loss": 0.8869,
+      "step": 715
+    },
+    {
+      "epoch": 0.03940778248665309,
+      "grad_norm": 0.9765705466270447,
+      "learning_rate": 9.99287196996015e-06,
+      "loss": 0.9512,
+      "step": 716
+    },
+    {
+      "epoch": 0.03946282128900875,
+      "grad_norm": 0.9610235691070557,
+      "learning_rate": 9.992848813743317e-06,
+      "loss": 0.8005,
+      "step": 717
+    },
+    {
+      "epoch": 0.03951786009136441,
+      "grad_norm": 1.102995753288269,
+      "learning_rate": 9.99282562000157e-06,
+      "loss": 0.8017,
+      "step": 718
+    },
+    {
+      "epoch": 0.03957289889372007,
+      "grad_norm": 1.023317575454712,
+      "learning_rate": 9.99280238873508e-06,
+      "loss": 0.911,
+      "step": 719
+    },
+    {
+      "epoch": 0.039627937696075734,
+      "grad_norm": 1.0531049966812134,
+      "learning_rate": 9.992779119944025e-06,
+      "loss": 0.8562,
+      "step": 720
+    },
+    {
+      "epoch": 0.039682976498431395,
+      "grad_norm": 0.918250322341919,
+      "learning_rate": 9.992755813628579e-06,
+      "loss": 0.92,
+      "step": 721
+    },
+    {
+      "epoch": 0.039738015300787057,
+      "grad_norm": 0.8508251309394836,
+      "learning_rate": 9.992732469788915e-06,
+      "loss": 0.7347,
+      "step": 722
+    },
+    {
+      "epoch": 0.03979305410314272,
+      "grad_norm": 0.9184926152229309,
+      "learning_rate": 9.992709088425211e-06,
+      "loss": 0.8732,
+      "step": 723
+    },
+    {
+      "epoch": 0.03984809290549838,
+      "grad_norm": 1.1613929271697998,
+      "learning_rate": 9.992685669537643e-06,
+      "loss": 0.9522,
+      "step": 724
+    },
+    {
+      "epoch": 0.039903131707854034,
+      "grad_norm": 1.091513752937317,
+      "learning_rate": 9.992662213126386e-06,
+      "loss": 0.9646,
+      "step": 725
+    },
+    {
+      "epoch": 0.039958170510209695,
+      "grad_norm": 1.057803750038147,
+      "learning_rate": 9.992638719191615e-06,
+      "loss": 0.7032,
+      "step": 726
+    },
+    {
+      "epoch": 0.040013209312565357,
+      "grad_norm": 0.8771823644638062,
+      "learning_rate": 9.992615187733508e-06,
+      "loss": 0.8577,
+      "step": 727
+    },
+    {
+      "epoch": 0.04006824811492102,
+      "grad_norm": 0.9471028447151184,
+      "learning_rate": 9.992591618752244e-06,
+      "loss": 0.9057,
+      "step": 728
+    },
+    {
+      "epoch": 0.04012328691727668,
+      "grad_norm": 0.9547705054283142,
+      "learning_rate": 9.992568012247995e-06,
+      "loss": 0.9549,
+      "step": 729
+    },
+    {
+      "epoch": 0.04017832571963234,
+      "grad_norm": 0.8862974047660828,
+      "learning_rate": 9.992544368220941e-06,
+      "loss": 0.8593,
+      "step": 730
+    },
+    {
+      "epoch": 0.040233364521988,
+      "grad_norm": 0.906334400177002,
+      "learning_rate": 9.992520686671261e-06,
+      "loss": 0.8832,
+      "step": 731
+    },
+    {
+      "epoch": 0.04028840332434366,
+      "grad_norm": 1.07270085811615,
+      "learning_rate": 9.992496967599133e-06,
+      "loss": 0.9409,
+      "step": 732
+    },
+    {
+      "epoch": 0.040343442126699325,
+      "grad_norm": 0.9026005268096924,
+      "learning_rate": 9.992473211004734e-06,
+      "loss": 0.8326,
+      "step": 733
+    },
+    {
+      "epoch": 0.040398480929054986,
+      "grad_norm": 0.9762942790985107,
+      "learning_rate": 9.992449416888241e-06,
+      "loss": 0.9048,
+      "step": 734
+    },
+    {
+      "epoch": 0.04045351973141065,
+      "grad_norm": 0.9658033847808838,
+      "learning_rate": 9.992425585249837e-06,
+      "loss": 0.9219,
+      "step": 735
+    },
+    {
+      "epoch": 0.0405085585337663,
+      "grad_norm": 0.8909044861793518,
+      "learning_rate": 9.992401716089698e-06,
+      "loss": 0.8564,
+      "step": 736
+    },
+    {
+      "epoch": 0.04056359733612196,
+      "grad_norm": 1.0387929677963257,
+      "learning_rate": 9.992377809408001e-06,
+      "loss": 0.9533,
+      "step": 737
+    },
+    {
+      "epoch": 0.040618636138477625,
+      "grad_norm": 0.9044275879859924,
+      "learning_rate": 9.99235386520493e-06,
+      "loss": 0.8508,
+      "step": 738
+    },
+    {
+      "epoch": 0.040673674940833286,
+      "grad_norm": 1.019377589225769,
+      "learning_rate": 9.992329883480667e-06,
+      "loss": 0.8684,
+      "step": 739
+    },
+    {
+      "epoch": 0.04072871374318895,
+      "grad_norm": 0.9394627213478088,
+      "learning_rate": 9.992305864235385e-06,
+      "loss": 0.7665,
+      "step": 740
+    },
+    {
+      "epoch": 0.04078375254554461,
+      "grad_norm": 0.8652323484420776,
+      "learning_rate": 9.99228180746927e-06,
+      "loss": 0.8576,
+      "step": 741
+    },
+    {
+      "epoch": 0.04083879134790027,
+      "grad_norm": 0.9347619414329529,
+      "learning_rate": 9.992257713182502e-06,
+      "loss": 0.9586,
+      "step": 742
+    },
+    {
+      "epoch": 0.04089383015025593,
+      "grad_norm": 0.9510203003883362,
+      "learning_rate": 9.99223358137526e-06,
+      "loss": 0.9092,
+      "step": 743
+    },
+    {
+      "epoch": 0.04094886895261159,
+      "grad_norm": 0.8242866396903992,
+      "learning_rate": 9.992209412047729e-06,
+      "loss": 0.6997,
+      "step": 744
+    },
+    {
+      "epoch": 0.041003907754967255,
+      "grad_norm": 0.8842730522155762,
+      "learning_rate": 9.992185205200087e-06,
+      "loss": 0.8873,
+      "step": 745
+    },
+    {
+      "epoch": 0.041058946557322916,
+      "grad_norm": 1.0813730955123901,
+      "learning_rate": 9.992160960832518e-06,
+      "loss": 1.0162,
+      "step": 746
+    },
+    {
+      "epoch": 0.04111398535967857,
+      "grad_norm": 1.1276283264160156,
+      "learning_rate": 9.9921366789452e-06,
+      "loss": 1.0004,
+      "step": 747
+    },
+    {
+      "epoch": 0.04116902416203423,
+      "grad_norm": 0.8810326457023621,
+      "learning_rate": 9.992112359538323e-06,
+      "loss": 0.7823,
+      "step": 748
+    },
+    {
+      "epoch": 0.04122406296438989,
+      "grad_norm": 0.9939407110214233,
+      "learning_rate": 9.992088002612066e-06,
+      "loss": 1.0016,
+      "step": 749
+    },
+    {
+      "epoch": 0.041279101766745555,
+      "grad_norm": 1.0963523387908936,
+      "learning_rate": 9.99206360816661e-06,
+      "loss": 0.9252,
+      "step": 750
+    },
+    {
+      "epoch": 0.041334140569101216,
+      "grad_norm": 1.1346478462219238,
+      "learning_rate": 9.99203917620214e-06,
+      "loss": 0.9608,
+      "step": 751
+    },
+    {
+      "epoch": 0.04138917937145688,
+      "grad_norm": 1.0108580589294434,
+      "learning_rate": 9.992014706718841e-06,
+      "loss": 0.9179,
+      "step": 752
+    },
+    {
+      "epoch": 0.04144421817381254,
+      "grad_norm": 0.897293210029602,
+      "learning_rate": 9.991990199716894e-06,
+      "loss": 0.9295,
+      "step": 753
+    },
+    {
+      "epoch": 0.0414992569761682,
+      "grad_norm": 1.0152363777160645,
+      "learning_rate": 9.991965655196488e-06,
+      "loss": 0.8467,
+      "step": 754
+    },
+    {
+      "epoch": 0.04155429577852386,
+      "grad_norm": 0.8655388355255127,
+      "learning_rate": 9.9919410731578e-06,
+      "loss": 0.796,
+      "step": 755
+    },
+    {
+      "epoch": 0.04160933458087952,
+      "grad_norm": 1.0140331983566284,
+      "learning_rate": 9.991916453601023e-06,
+      "loss": 0.8444,
+      "step": 756
+    },
+    {
+      "epoch": 0.041664373383235184,
+      "grad_norm": 0.9387341141700745,
+      "learning_rate": 9.991891796526338e-06,
+      "loss": 0.8669,
+      "step": 757
+    },
+    {
+      "epoch": 0.04171941218559084,
+      "grad_norm": 0.9395696520805359,
+      "learning_rate": 9.991867101933928e-06,
+      "loss": 0.8376,
+      "step": 758
+    },
+    {
+      "epoch": 0.0417744509879465,
+      "grad_norm": 1.0856634378433228,
+      "learning_rate": 9.991842369823983e-06,
+      "loss": 0.9271,
+      "step": 759
+    },
+    {
+      "epoch": 0.04182948979030216,
+      "grad_norm": 0.8777190446853638,
+      "learning_rate": 9.991817600196687e-06,
+      "loss": 0.9197,
+      "step": 760
+    },
+    {
+      "epoch": 0.04188452859265782,
+      "grad_norm": 0.9639917016029358,
+      "learning_rate": 9.991792793052225e-06,
+      "loss": 0.8835,
+      "step": 761
+    },
+    {
+      "epoch": 0.041939567395013484,
+      "grad_norm": 0.9384773969650269,
+      "learning_rate": 9.991767948390785e-06,
+      "loss": 0.8403,
+      "step": 762
+    },
+    {
+      "epoch": 0.041994606197369146,
+      "grad_norm": 0.8987650275230408,
+      "learning_rate": 9.991743066212554e-06,
+      "loss": 0.7948,
+      "step": 763
+    },
+    {
+      "epoch": 0.04204964499972481,
+      "grad_norm": 1.0545049905776978,
+      "learning_rate": 9.991718146517717e-06,
+      "loss": 0.9359,
+      "step": 764
+    },
+    {
+      "epoch": 0.04210468380208047,
+      "grad_norm": 0.9840022325515747,
+      "learning_rate": 9.991693189306463e-06,
+      "loss": 0.9188,
+      "step": 765
+    },
+    {
+      "epoch": 0.04215972260443613,
+      "grad_norm": 0.8769927620887756,
+      "learning_rate": 9.991668194578981e-06,
+      "loss": 0.8647,
+      "step": 766
+    },
+    {
+      "epoch": 0.04221476140679179,
+      "grad_norm": 0.9268791675567627,
+      "learning_rate": 9.991643162335455e-06,
+      "loss": 0.897,
+      "step": 767
+    },
+    {
+      "epoch": 0.042269800209147446,
+      "grad_norm": 0.9316747784614563,
+      "learning_rate": 9.991618092576075e-06,
+      "loss": 0.9341,
+      "step": 768
+    },
+    {
+      "epoch": 0.04232483901150311,
+      "grad_norm": 0.8348364233970642,
+      "learning_rate": 9.991592985301031e-06,
+      "loss": 0.7528,
+      "step": 769
+    },
+    {
+      "epoch": 0.04237987781385877,
+      "grad_norm": 0.9139068126678467,
+      "learning_rate": 9.99156784051051e-06,
+      "loss": 0.8596,
+      "step": 770
+    },
+    {
+      "epoch": 0.04243491661621443,
+      "grad_norm": 0.9403928518295288,
+      "learning_rate": 9.991542658204701e-06,
+      "loss": 0.974,
+      "step": 771
+    },
+    {
+      "epoch": 0.04248995541857009,
+      "grad_norm": 0.993549108505249,
+      "learning_rate": 9.991517438383793e-06,
+      "loss": 0.9479,
+      "step": 772
+    },
+    {
+      "epoch": 0.04254499422092575,
+      "grad_norm": 0.8494916558265686,
+      "learning_rate": 9.991492181047975e-06,
+      "loss": 0.9149,
+      "step": 773
+    },
+    {
+      "epoch": 0.042600033023281414,
+      "grad_norm": 1.0351910591125488,
+      "learning_rate": 9.991466886197441e-06,
+      "loss": 0.9552,
+      "step": 774
+    },
+    {
+      "epoch": 0.042655071825637075,
+      "grad_norm": 0.916829526424408,
+      "learning_rate": 9.991441553832375e-06,
+      "loss": 0.8781,
+      "step": 775
+    },
+    {
+      "epoch": 0.04271011062799274,
+      "grad_norm": 1.113476276397705,
+      "learning_rate": 9.991416183952972e-06,
+      "loss": 0.8137,
+      "step": 776
+    },
+    {
+      "epoch": 0.0427651494303484,
+      "grad_norm": 1.1608171463012695,
+      "learning_rate": 9.991390776559421e-06,
+      "loss": 1.0045,
+      "step": 777
+    },
+    {
+      "epoch": 0.04282018823270406,
+      "grad_norm": 1.0045493841171265,
+      "learning_rate": 9.991365331651913e-06,
+      "loss": 0.8813,
+      "step": 778
+    },
+    {
+      "epoch": 0.042875227035059714,
+      "grad_norm": 0.918820858001709,
+      "learning_rate": 9.991339849230639e-06,
+      "loss": 0.9198,
+      "step": 779
+    },
+    {
+      "epoch": 0.042930265837415375,
+      "grad_norm": 0.9875735640525818,
+      "learning_rate": 9.991314329295792e-06,
+      "loss": 0.8665,
+      "step": 780
+    },
+    {
+      "epoch": 0.04298530463977104,
+      "grad_norm": 0.873768150806427,
+      "learning_rate": 9.991288771847561e-06,
+      "loss": 0.8606,
+      "step": 781
+    },
+    {
+      "epoch": 0.0430403434421267,
+      "grad_norm": 0.8892746567726135,
+      "learning_rate": 9.991263176886139e-06,
+      "loss": 0.9011,
+      "step": 782
+    },
+    {
+      "epoch": 0.04309538224448236,
+      "grad_norm": 1.097734808921814,
+      "learning_rate": 9.99123754441172e-06,
+      "loss": 1.009,
+      "step": 783
+    },
+    {
+      "epoch": 0.04315042104683802,
+      "grad_norm": 1.0065964460372925,
+      "learning_rate": 9.991211874424497e-06,
+      "loss": 0.9492,
+      "step": 784
+    },
+    {
+      "epoch": 0.04320545984919368,
+      "grad_norm": 1.0791678428649902,
+      "learning_rate": 9.99118616692466e-06,
+      "loss": 1.0142,
+      "step": 785
+    },
+    {
+      "epoch": 0.043260498651549344,
+      "grad_norm": 0.9454777836799622,
+      "learning_rate": 9.991160421912404e-06,
+      "loss": 0.8058,
+      "step": 786
+    },
+    {
+      "epoch": 0.043315537453905005,
+      "grad_norm": 0.9448156952857971,
+      "learning_rate": 9.991134639387922e-06,
+      "loss": 0.8184,
+      "step": 787
+    },
+    {
+      "epoch": 0.043370576256260666,
+      "grad_norm": 0.9636550545692444,
+      "learning_rate": 9.99110881935141e-06,
+      "loss": 0.8606,
+      "step": 788
+    },
+    {
+      "epoch": 0.04342561505861633,
+      "grad_norm": 0.9933613538742065,
+      "learning_rate": 9.991082961803058e-06,
+      "loss": 0.9449,
+      "step": 789
+    },
+    {
+      "epoch": 0.04348065386097198,
+      "grad_norm": 0.8906797170639038,
+      "learning_rate": 9.991057066743065e-06,
+      "loss": 0.8053,
+      "step": 790
+    },
+    {
+      "epoch": 0.043535692663327644,
+      "grad_norm": 1.0393906831741333,
+      "learning_rate": 9.991031134171621e-06,
+      "loss": 0.8487,
+      "step": 791
+    },
+    {
+      "epoch": 0.043590731465683305,
+      "grad_norm": 1.0618231296539307,
+      "learning_rate": 9.991005164088923e-06,
+      "loss": 0.9847,
+      "step": 792
+    },
+    {
+      "epoch": 0.043645770268038966,
+      "grad_norm": 0.9525149464607239,
+      "learning_rate": 9.990979156495167e-06,
+      "loss": 0.9318,
+      "step": 793
+    },
+    {
+      "epoch": 0.04370080907039463,
+      "grad_norm": 0.9430851936340332,
+      "learning_rate": 9.990953111390546e-06,
+      "loss": 0.8483,
+      "step": 794
+    },
+    {
+      "epoch": 0.04375584787275029,
+      "grad_norm": 0.9259672164916992,
+      "learning_rate": 9.99092702877526e-06,
+      "loss": 0.9365,
+      "step": 795
+    },
+    {
+      "epoch": 0.04381088667510595,
+      "grad_norm": 0.942609965801239,
+      "learning_rate": 9.9909009086495e-06,
+      "loss": 0.8408,
+      "step": 796
+    },
+    {
+      "epoch": 0.04386592547746161,
+      "grad_norm": 0.939255952835083,
+      "learning_rate": 9.990874751013467e-06,
+      "loss": 0.8749,
+      "step": 797
+    },
+    {
+      "epoch": 0.04392096427981727,
+      "grad_norm": 1.1701711416244507,
+      "learning_rate": 9.990848555867353e-06,
+      "loss": 0.9312,
+      "step": 798
+    },
+    {
+      "epoch": 0.043976003082172935,
+      "grad_norm": 1.0441124439239502,
+      "learning_rate": 9.990822323211358e-06,
+      "loss": 0.8618,
+      "step": 799
+    },
+    {
+      "epoch": 0.04403104188452859,
+      "grad_norm": 0.9601489305496216,
+      "learning_rate": 9.990796053045679e-06,
+      "loss": 0.9569,
+      "step": 800
+    },
+    {
+      "epoch": 0.04408608068688425,
+      "grad_norm": 0.9394032955169678,
+      "learning_rate": 9.990769745370513e-06,
+      "loss": 0.846,
+      "step": 801
+    },
+    {
+      "epoch": 0.04414111948923991,
+      "grad_norm": 0.9631348252296448,
+      "learning_rate": 9.990743400186056e-06,
+      "loss": 0.8754,
+      "step": 802
+    },
+    {
+      "epoch": 0.04419615829159557,
+      "grad_norm": 0.9234963059425354,
+      "learning_rate": 9.990717017492508e-06,
+      "loss": 0.8613,
+      "step": 803
+    },
+    {
+      "epoch": 0.044251197093951235,
+      "grad_norm": 0.9169090390205383,
+      "learning_rate": 9.990690597290069e-06,
+      "loss": 0.8867,
+      "step": 804
+    },
+    {
+      "epoch": 0.044306235896306896,
+      "grad_norm": 1.0194867849349976,
+      "learning_rate": 9.990664139578933e-06,
+      "loss": 0.8675,
+      "step": 805
+    },
+    {
+      "epoch": 0.04436127469866256,
+      "grad_norm": 1.3226114511489868,
+      "learning_rate": 9.990637644359302e-06,
+      "loss": 0.997,
+      "step": 806
+    },
+    {
+      "epoch": 0.04441631350101822,
+      "grad_norm": 0.8904317617416382,
+      "learning_rate": 9.990611111631374e-06,
+      "loss": 0.7274,
+      "step": 807
+    },
+    {
+      "epoch": 0.04447135230337388,
+      "grad_norm": 0.8909007906913757,
+      "learning_rate": 9.99058454139535e-06,
+      "loss": 0.8141,
+      "step": 808
+    },
+    {
+      "epoch": 0.04452639110572954,
+      "grad_norm": 1.004015564918518,
+      "learning_rate": 9.990557933651429e-06,
+      "loss": 0.9883,
+      "step": 809
+    },
+    {
+      "epoch": 0.0445814299080852,
+      "grad_norm": 1.1215732097625732,
+      "learning_rate": 9.990531288399807e-06,
+      "loss": 0.9355,
+      "step": 810
+    },
+    {
+      "epoch": 0.04463646871044086,
+      "grad_norm": 1.0545012950897217,
+      "learning_rate": 9.99050460564069e-06,
+      "loss": 0.9532,
+      "step": 811
+    },
+    {
+      "epoch": 0.04469150751279652,
+      "grad_norm": 0.9608867168426514,
+      "learning_rate": 9.990477885374277e-06,
+      "loss": 0.9363,
+      "step": 812
+    },
+    {
+      "epoch": 0.04474654631515218,
+      "grad_norm": 0.8750461935997009,
+      "learning_rate": 9.990451127600766e-06,
+      "loss": 0.7343,
+      "step": 813
+    },
+    {
+      "epoch": 0.04480158511750784,
+      "grad_norm": 0.891740620136261,
+      "learning_rate": 9.99042433232036e-06,
+      "loss": 0.8541,
+      "step": 814
+    },
+    {
+      "epoch": 0.0448566239198635,
+      "grad_norm": 1.1520029306411743,
+      "learning_rate": 9.990397499533264e-06,
+      "loss": 0.7696,
+      "step": 815
+    },
+    {
+      "epoch": 0.044911662722219164,
+      "grad_norm": 0.9526278972625732,
+      "learning_rate": 9.990370629239673e-06,
+      "loss": 0.8953,
+      "step": 816
+    },
+    {
+      "epoch": 0.044966701524574826,
+      "grad_norm": 0.9218434691429138,
+      "learning_rate": 9.990343721439795e-06,
+      "loss": 0.8198,
+      "step": 817
+    },
+    {
+      "epoch": 0.04502174032693049,
+      "grad_norm": 0.8502745628356934,
+      "learning_rate": 9.990316776133827e-06,
+      "loss": 0.8035,
+      "step": 818
+    },
+    {
+      "epoch": 0.04507677912928615,
+      "grad_norm": 0.8861565589904785,
+      "learning_rate": 9.990289793321975e-06,
+      "loss": 0.8626,
+      "step": 819
+    },
+    {
+      "epoch": 0.04513181793164181,
+      "grad_norm": 1.1113256216049194,
+      "learning_rate": 9.99026277300444e-06,
+      "loss": 0.9363,
+      "step": 820
+    },
+    {
+      "epoch": 0.04518685673399747,
+      "grad_norm": 0.9984708428382874,
+      "learning_rate": 9.990235715181426e-06,
+      "loss": 1.0376,
+      "step": 821
+    },
+    {
+      "epoch": 0.045241895536353126,
+      "grad_norm": 0.9026711583137512,
+      "learning_rate": 9.990208619853137e-06,
+      "loss": 0.9079,
+      "step": 822
+    },
+    {
+      "epoch": 0.04529693433870879,
+      "grad_norm": 0.8724965453147888,
+      "learning_rate": 9.990181487019775e-06,
+      "loss": 0.8665,
+      "step": 823
+    },
+    {
+      "epoch": 0.04535197314106445,
+      "grad_norm": 0.8923047780990601,
+      "learning_rate": 9.990154316681543e-06,
+      "loss": 0.7779,
+      "step": 824
+    },
+    {
+      "epoch": 0.04540701194342011,
+      "grad_norm": 0.9024640321731567,
+      "learning_rate": 9.99012710883865e-06,
+      "loss": 0.8859,
+      "step": 825
+    },
+    {
+      "epoch": 0.04546205074577577,
+      "grad_norm": 0.9245888590812683,
+      "learning_rate": 9.990099863491296e-06,
+      "loss": 0.8501,
+      "step": 826
+    },
+    {
+      "epoch": 0.04551708954813143,
+      "grad_norm": 0.9257050156593323,
+      "learning_rate": 9.990072580639687e-06,
+      "loss": 0.9561,
+      "step": 827
+    },
+    {
+      "epoch": 0.045572128350487094,
+      "grad_norm": 0.995610773563385,
+      "learning_rate": 9.99004526028403e-06,
+      "loss": 0.917,
+      "step": 828
+    },
+    {
+      "epoch": 0.045627167152842756,
+      "grad_norm": 0.9524009823799133,
+      "learning_rate": 9.990017902424525e-06,
+      "loss": 0.9184,
+      "step": 829
+    },
+    {
+      "epoch": 0.04568220595519842,
+      "grad_norm": 0.9264503121376038,
+      "learning_rate": 9.989990507061385e-06,
+      "loss": 0.8615,
+      "step": 830
+    },
+    {
+      "epoch": 0.04573724475755408,
+      "grad_norm": 1.0068570375442505,
+      "learning_rate": 9.989963074194809e-06,
+      "loss": 0.8331,
+      "step": 831
+    },
+    {
+      "epoch": 0.04579228355990974,
+      "grad_norm": 0.9295952320098877,
+      "learning_rate": 9.989935603825009e-06,
+      "loss": 0.8387,
+      "step": 832
+    },
+    {
+      "epoch": 0.045847322362265394,
+      "grad_norm": 1.0408827066421509,
+      "learning_rate": 9.989908095952186e-06,
+      "loss": 0.9686,
+      "step": 833
+    },
+    {
+      "epoch": 0.045902361164621056,
+      "grad_norm": 0.8874136209487915,
+      "learning_rate": 9.989880550576551e-06,
+      "loss": 0.815,
+      "step": 834
+    },
+    {
+      "epoch": 0.04595739996697672,
+      "grad_norm": 0.9898836016654968,
+      "learning_rate": 9.989852967698311e-06,
+      "loss": 0.9458,
+      "step": 835
+    },
+    {
+      "epoch": 0.04601243876933238,
+      "grad_norm": 0.9828970432281494,
+      "learning_rate": 9.989825347317668e-06,
+      "loss": 0.7922,
+      "step": 836
+    },
+    {
+      "epoch": 0.04606747757168804,
+      "grad_norm": 1.025447964668274,
+      "learning_rate": 9.989797689434836e-06,
+      "loss": 0.9349,
+      "step": 837
+    },
+    {
+      "epoch": 0.0461225163740437,
+      "grad_norm": 0.8623831272125244,
+      "learning_rate": 9.98976999405002e-06,
+      "loss": 0.8786,
+      "step": 838
+    },
+    {
+      "epoch": 0.04617755517639936,
+      "grad_norm": 0.9614997506141663,
+      "learning_rate": 9.98974226116343e-06,
+      "loss": 0.7885,
+      "step": 839
+    },
+    {
+      "epoch": 0.046232593978755024,
+      "grad_norm": 1.0207616090774536,
+      "learning_rate": 9.989714490775269e-06,
+      "loss": 0.9786,
+      "step": 840
+    },
+    {
+      "epoch": 0.046287632781110685,
+      "grad_norm": 0.8509595990180969,
+      "learning_rate": 9.98968668288575e-06,
+      "loss": 0.7312,
+      "step": 841
+    },
+    {
+      "epoch": 0.04634267158346635,
+      "grad_norm": 0.9822607040405273,
+      "learning_rate": 9.989658837495084e-06,
+      "loss": 0.952,
+      "step": 842
+    },
+    {
+      "epoch": 0.046397710385822,
+      "grad_norm": 1.0058252811431885,
+      "learning_rate": 9.989630954603477e-06,
+      "loss": 0.8811,
+      "step": 843
+    },
+    {
+      "epoch": 0.04645274918817766,
+      "grad_norm": 1.0146985054016113,
+      "learning_rate": 9.989603034211139e-06,
+      "loss": 0.9051,
+      "step": 844
+    },
+    {
+      "epoch": 0.046507787990533324,
+      "grad_norm": 0.8976503610610962,
+      "learning_rate": 9.98957507631828e-06,
+      "loss": 0.879,
+      "step": 845
+    },
+    {
+      "epoch": 0.046562826792888985,
+      "grad_norm": 0.8791939616203308,
+      "learning_rate": 9.989547080925111e-06,
+      "loss": 0.8944,
+      "step": 846
+    },
+    {
+      "epoch": 0.04661786559524465,
+      "grad_norm": 0.8530884981155396,
+      "learning_rate": 9.989519048031842e-06,
+      "loss": 0.9029,
+      "step": 847
+    },
+    {
+      "epoch": 0.04667290439760031,
+      "grad_norm": 0.9621617197990417,
+      "learning_rate": 9.989490977638683e-06,
+      "loss": 0.8374,
+      "step": 848
+    },
+    {
+      "epoch": 0.04672794319995597,
+      "grad_norm": 0.9629075527191162,
+      "learning_rate": 9.989462869745845e-06,
+      "loss": 0.9032,
+      "step": 849
+    },
+    {
+      "epoch": 0.04678298200231163,
+      "grad_norm": 1.3256126642227173,
+      "learning_rate": 9.989434724353541e-06,
+      "loss": 0.9748,
+      "step": 850
+    },
+    {
+      "epoch": 0.04683802080466729,
+      "grad_norm": 1.0230494737625122,
+      "learning_rate": 9.989406541461979e-06,
+      "loss": 0.9752,
+      "step": 851
+    },
+    {
+      "epoch": 0.046893059607022954,
+      "grad_norm": 0.8454533219337463,
+      "learning_rate": 9.989378321071375e-06,
+      "loss": 0.8426,
+      "step": 852
+    },
+    {
+      "epoch": 0.046948098409378615,
+      "grad_norm": 0.9995863437652588,
+      "learning_rate": 9.989350063181939e-06,
+      "loss": 0.9955,
+      "step": 853
+    },
+    {
+      "epoch": 0.04700313721173427,
+      "grad_norm": 0.8956604599952698,
+      "learning_rate": 9.989321767793883e-06,
+      "loss": 0.9024,
+      "step": 854
+    },
+    {
+      "epoch": 0.04705817601408993,
+      "grad_norm": 1.0123292207717896,
+      "learning_rate": 9.989293434907419e-06,
+      "loss": 0.7856,
+      "step": 855
+    },
+    {
+      "epoch": 0.04711321481644559,
+      "grad_norm": 0.814577043056488,
+      "learning_rate": 9.989265064522762e-06,
+      "loss": 0.8377,
+      "step": 856
+    },
+    {
+      "epoch": 0.047168253618801254,
+      "grad_norm": 1.1571552753448486,
+      "learning_rate": 9.989236656640125e-06,
+      "loss": 0.8562,
+      "step": 857
+    },
+    {
+      "epoch": 0.047223292421156915,
+      "grad_norm": 0.9681577682495117,
+      "learning_rate": 9.98920821125972e-06,
+      "loss": 0.8473,
+      "step": 858
+    },
+    {
+      "epoch": 0.047278331223512576,
+      "grad_norm": 0.9680121541023254,
+      "learning_rate": 9.989179728381761e-06,
+      "loss": 0.9811,
+      "step": 859
+    },
+    {
+      "epoch": 0.04733337002586824,
+      "grad_norm": 0.985477089881897,
+      "learning_rate": 9.989151208006464e-06,
+      "loss": 0.6994,
+      "step": 860
+    },
+    {
+      "epoch": 0.0473884088282239,
+      "grad_norm": 0.8612962365150452,
+      "learning_rate": 9.98912265013404e-06,
+      "loss": 0.7667,
+      "step": 861
+    },
+    {
+      "epoch": 0.04744344763057956,
+      "grad_norm": 0.8884604573249817,
+      "learning_rate": 9.989094054764708e-06,
+      "loss": 0.8382,
+      "step": 862
+    },
+    {
+      "epoch": 0.04749848643293522,
+      "grad_norm": 1.036881923675537,
+      "learning_rate": 9.989065421898681e-06,
+      "loss": 0.8748,
+      "step": 863
+    },
+    {
+      "epoch": 0.04755352523529088,
+      "grad_norm": 0.9954493045806885,
+      "learning_rate": 9.989036751536171e-06,
+      "loss": 0.9174,
+      "step": 864
+    },
+    {
+      "epoch": 0.04760856403764654,
+      "grad_norm": 0.9984694123268127,
+      "learning_rate": 9.989008043677399e-06,
+      "loss": 0.7636,
+      "step": 865
+    },
+    {
+      "epoch": 0.0476636028400022,
+      "grad_norm": 1.0412588119506836,
+      "learning_rate": 9.988979298322576e-06,
+      "loss": 0.773,
+      "step": 866
+    },
+    {
+      "epoch": 0.04771864164235786,
+      "grad_norm": 0.8034874796867371,
+      "learning_rate": 9.98895051547192e-06,
+      "loss": 0.7914,
+      "step": 867
+    },
+    {
+      "epoch": 0.04777368044471352,
+      "grad_norm": 0.8983979225158691,
+      "learning_rate": 9.988921695125648e-06,
+      "loss": 0.7292,
+      "step": 868
+    },
+    {
+      "epoch": 0.04782871924706918,
+      "grad_norm": 0.9445077776908875,
+      "learning_rate": 9.988892837283976e-06,
+      "loss": 0.8263,
+      "step": 869
+    },
+    {
+      "epoch": 0.047883758049424845,
+      "grad_norm": 1.0753306150436401,
+      "learning_rate": 9.988863941947121e-06,
+      "loss": 1.1122,
+      "step": 870
+    },
+    {
+      "epoch": 0.047938796851780506,
+      "grad_norm": 1.0091484785079956,
+      "learning_rate": 9.9888350091153e-06,
+      "loss": 0.9276,
+      "step": 871
+    },
+    {
+      "epoch": 0.04799383565413617,
+      "grad_norm": 1.0977306365966797,
+      "learning_rate": 9.988806038788732e-06,
+      "loss": 0.854,
+      "step": 872
+    },
+    {
+      "epoch": 0.04804887445649183,
+      "grad_norm": 1.0285007953643799,
+      "learning_rate": 9.988777030967632e-06,
+      "loss": 0.9441,
+      "step": 873
+    },
+    {
+      "epoch": 0.04810391325884749,
+      "grad_norm": 0.8973976373672485,
+      "learning_rate": 9.988747985652218e-06,
+      "loss": 0.786,
+      "step": 874
+    },
+    {
+      "epoch": 0.04815895206120315,
+      "grad_norm": 0.9809553623199463,
+      "learning_rate": 9.98871890284271e-06,
+      "loss": 0.9042,
+      "step": 875
+    },
+    {
+      "epoch": 0.048213990863558806,
+      "grad_norm": 0.8514279723167419,
+      "learning_rate": 9.988689782539326e-06,
+      "loss": 0.7874,
+      "step": 876
+    },
+    {
+      "epoch": 0.04826902966591447,
+      "grad_norm": 0.8299674391746521,
+      "learning_rate": 9.988660624742286e-06,
+      "loss": 0.8704,
+      "step": 877
+    },
+    {
+      "epoch": 0.04832406846827013,
+      "grad_norm": 0.9862462282180786,
+      "learning_rate": 9.988631429451809e-06,
+      "loss": 0.9963,
+      "step": 878
+    },
+    {
+      "epoch": 0.04837910727062579,
+      "grad_norm": 0.9041131734848022,
+      "learning_rate": 9.988602196668111e-06,
+      "loss": 0.9207,
+      "step": 879
+    },
+    {
+      "epoch": 0.04843414607298145,
+      "grad_norm": 0.8597276210784912,
+      "learning_rate": 9.988572926391416e-06,
+      "loss": 0.8226,
+      "step": 880
+    },
+    {
+      "epoch": 0.04848918487533711,
+      "grad_norm": 0.9494329690933228,
+      "learning_rate": 9.988543618621941e-06,
+      "loss": 0.8834,
+      "step": 881
+    },
+    {
+      "epoch": 0.048544223677692774,
+      "grad_norm": 0.9129118323326111,
+      "learning_rate": 9.98851427335991e-06,
+      "loss": 0.7819,
+      "step": 882
+    },
+    {
+      "epoch": 0.048599262480048436,
+      "grad_norm": 0.9145999550819397,
+      "learning_rate": 9.988484890605539e-06,
+      "loss": 0.885,
+      "step": 883
+    },
+    {
+      "epoch": 0.0486543012824041,
+      "grad_norm": 1.0115307569503784,
+      "learning_rate": 9.98845547035905e-06,
+      "loss": 0.8347,
+      "step": 884
+    },
+    {
+      "epoch": 0.04870934008475976,
+      "grad_norm": 1.1372706890106201,
+      "learning_rate": 9.988426012620667e-06,
+      "loss": 0.944,
+      "step": 885
+    },
+    {
+      "epoch": 0.04876437888711541,
+      "grad_norm": 0.9502811431884766,
+      "learning_rate": 9.98839651739061e-06,
+      "loss": 0.9054,
+      "step": 886
+    },
+    {
+      "epoch": 0.048819417689471074,
+      "grad_norm": 0.9612823128700256,
+      "learning_rate": 9.988366984669097e-06,
+      "loss": 0.8796,
+      "step": 887
+    },
+    {
+      "epoch": 0.048874456491826736,
+      "grad_norm": 0.9551461935043335,
+      "learning_rate": 9.988337414456355e-06,
+      "loss": 0.8769,
+      "step": 888
+    },
+    {
+      "epoch": 0.0489294952941824,
+      "grad_norm": 0.8554086089134216,
+      "learning_rate": 9.988307806752603e-06,
+      "loss": 0.892,
+      "step": 889
+    },
+    {
+      "epoch": 0.04898453409653806,
+      "grad_norm": 0.8418886661529541,
+      "learning_rate": 9.988278161558067e-06,
+      "loss": 0.7568,
+      "step": 890
+    },
+    {
+      "epoch": 0.04903957289889372,
+      "grad_norm": 1.4780360460281372,
+      "learning_rate": 9.988248478872967e-06,
+      "loss": 0.9126,
+      "step": 891
+    },
+    {
+      "epoch": 0.04909461170124938,
+      "grad_norm": 0.8236714005470276,
+      "learning_rate": 9.988218758697526e-06,
+      "loss": 0.7317,
+      "step": 892
+    },
+    {
+      "epoch": 0.04914965050360504,
+      "grad_norm": 0.8777141571044922,
+      "learning_rate": 9.988189001031968e-06,
+      "loss": 0.7989,
+      "step": 893
+    },
+    {
+      "epoch": 0.049204689305960704,
+      "grad_norm": 1.0235031843185425,
+      "learning_rate": 9.988159205876516e-06,
+      "loss": 0.8335,
+      "step": 894
+    },
+    {
+      "epoch": 0.049259728108316365,
+      "grad_norm": 0.9340357184410095,
+      "learning_rate": 9.988129373231395e-06,
+      "loss": 0.8129,
+      "step": 895
+    },
+    {
+      "epoch": 0.04931476691067203,
+      "grad_norm": 1.7686667442321777,
+      "learning_rate": 9.98809950309683e-06,
+      "loss": 0.9792,
+      "step": 896
+    },
+    {
+      "epoch": 0.04936980571302768,
+      "grad_norm": 0.9252369403839111,
+      "learning_rate": 9.988069595473044e-06,
+      "loss": 0.8671,
+      "step": 897
+    },
+    {
+      "epoch": 0.04942484451538334,
+      "grad_norm": 0.9989960789680481,
+      "learning_rate": 9.988039650360262e-06,
+      "loss": 0.9245,
+      "step": 898
+    },
+    {
+      "epoch": 0.049479883317739004,
+      "grad_norm": 1.062912106513977,
+      "learning_rate": 9.98800966775871e-06,
+      "loss": 0.9146,
+      "step": 899
+    },
+    {
+      "epoch": 0.049534922120094665,
+      "grad_norm": 0.8698169589042664,
+      "learning_rate": 9.98797964766861e-06,
+      "loss": 0.8606,
+      "step": 900
+    },
+    {
+      "epoch": 0.04958996092245033,
+      "grad_norm": 1.6754224300384521,
+      "learning_rate": 9.98794959009019e-06,
+      "loss": 0.9236,
+      "step": 901
+    },
+    {
+      "epoch": 0.04964499972480599,
+      "grad_norm": 1.084174394607544,
+      "learning_rate": 9.98791949502368e-06,
+      "loss": 0.9252,
+      "step": 902
+    },
+    {
+      "epoch": 0.04970003852716165,
+      "grad_norm": 0.9866724610328674,
+      "learning_rate": 9.987889362469301e-06,
+      "loss": 0.9096,
+      "step": 903
+    },
+    {
+      "epoch": 0.04975507732951731,
+      "grad_norm": 0.8814040422439575,
+      "learning_rate": 9.987859192427279e-06,
+      "loss": 0.8475,
+      "step": 904
+    },
+    {
+      "epoch": 0.04981011613187297,
+      "grad_norm": 0.8796457052230835,
+      "learning_rate": 9.987828984897843e-06,
+      "loss": 0.8478,
+      "step": 905
+    },
+    {
+      "epoch": 0.049865154934228634,
+      "grad_norm": 1.0541884899139404,
+      "learning_rate": 9.98779873988122e-06,
+      "loss": 0.9799,
+      "step": 906
+    },
+    {
+      "epoch": 0.049920193736584295,
+      "grad_norm": 0.91409832239151,
+      "learning_rate": 9.987768457377636e-06,
+      "loss": 0.8701,
+      "step": 907
+    },
+    {
+      "epoch": 0.04997523253893995,
+      "grad_norm": 1.0120370388031006,
+      "learning_rate": 9.98773813738732e-06,
+      "loss": 0.8417,
+      "step": 908
+    },
+    {
+      "epoch": 0.05003027134129561,
+      "grad_norm": 1.7744206190109253,
+      "learning_rate": 9.987707779910499e-06,
+      "loss": 0.9263,
+      "step": 909
+    },
+    {
+      "epoch": 0.05008531014365127,
+      "grad_norm": 0.9423969388008118,
+      "learning_rate": 9.987677384947402e-06,
+      "loss": 0.9667,
+      "step": 910
+    },
+    {
+      "epoch": 0.050140348946006934,
+      "grad_norm": 1.5940319299697876,
+      "learning_rate": 9.987646952498256e-06,
+      "loss": 0.9223,
+      "step": 911
+    },
+    {
+      "epoch": 0.050195387748362595,
+      "grad_norm": 0.941792368888855,
+      "learning_rate": 9.987616482563292e-06,
+      "loss": 0.895,
+      "step": 912
+    },
+    {
+      "epoch": 0.05025042655071826,
+      "grad_norm": 3.1945221424102783,
+      "learning_rate": 9.987585975142738e-06,
+      "loss": 0.837,
+      "step": 913
+    },
+    {
+      "epoch": 0.05030546535307392,
+      "grad_norm": 2.0819199085235596,
+      "learning_rate": 9.98755543023682e-06,
+      "loss": 0.918,
+      "step": 914
+    },
+    {
+      "epoch": 0.05036050415542958,
+      "grad_norm": 0.984282910823822,
+      "learning_rate": 9.987524847845773e-06,
+      "loss": 0.8589,
+      "step": 915
+    },
+    {
+      "epoch": 0.05041554295778524,
+      "grad_norm": 0.9021026492118835,
+      "learning_rate": 9.987494227969823e-06,
+      "loss": 0.9053,
+      "step": 916
+    },
+    {
+      "epoch": 0.0504705817601409,
+      "grad_norm": 2.6515462398529053,
+      "learning_rate": 9.9874635706092e-06,
+      "loss": 0.8874,
+      "step": 917
+    },
+    {
+      "epoch": 0.050525620562496563,
+      "grad_norm": 0.8893095254898071,
+      "learning_rate": 9.98743287576414e-06,
+      "loss": 0.8259,
+      "step": 918
+    },
+    {
+      "epoch": 0.05058065936485222,
+      "grad_norm": 0.9897775650024414,
+      "learning_rate": 9.987402143434868e-06,
+      "loss": 0.877,
+      "step": 919
+    },
+    {
+      "epoch": 0.05063569816720788,
+      "grad_norm": 0.9391944408416748,
+      "learning_rate": 9.987371373621614e-06,
+      "loss": 0.9363,
+      "step": 920
+    },
+    {
+      "epoch": 0.05069073696956354,
+      "grad_norm": 0.9585913419723511,
+      "learning_rate": 9.987340566324615e-06,
+      "loss": 0.8704,
+      "step": 921
+    },
+    {
+      "epoch": 0.0507457757719192,
+      "grad_norm": 0.9210980534553528,
+      "learning_rate": 9.987309721544098e-06,
+      "loss": 0.9321,
+      "step": 922
+    },
+    {
+      "epoch": 0.05080081457427486,
+      "grad_norm": 1.0713307857513428,
+      "learning_rate": 9.987278839280295e-06,
+      "loss": 0.9489,
+      "step": 923
+    },
+    {
+      "epoch": 0.050855853376630525,
+      "grad_norm": 1.0178636312484741,
+      "learning_rate": 9.98724791953344e-06,
+      "loss": 0.853,
+      "step": 924
+    },
+    {
+      "epoch": 0.050910892178986186,
+      "grad_norm": 0.9782636761665344,
+      "learning_rate": 9.987216962303766e-06,
+      "loss": 0.924,
+      "step": 925
+    },
+    {
+      "epoch": 0.05096593098134185,
+      "grad_norm": 0.9474522471427917,
+      "learning_rate": 9.987185967591503e-06,
+      "loss": 0.8619,
+      "step": 926
+    },
+    {
+      "epoch": 0.05102096978369751,
+      "grad_norm": 1.1875778436660767,
+      "learning_rate": 9.987154935396885e-06,
+      "loss": 1.012,
+      "step": 927
+    },
+    {
+      "epoch": 0.05107600858605317,
+      "grad_norm": 1.0585243701934814,
+      "learning_rate": 9.987123865720147e-06,
+      "loss": 0.9019,
+      "step": 928
+    },
+    {
+      "epoch": 0.051131047388408825,
+      "grad_norm": 0.9848800897598267,
+      "learning_rate": 9.98709275856152e-06,
+      "loss": 0.9665,
+      "step": 929
+    },
+    {
+      "epoch": 0.051186086190764486,
+      "grad_norm": 1.04201078414917,
+      "learning_rate": 9.987061613921238e-06,
+      "loss": 0.9269,
+      "step": 930
+    },
+    {
+      "epoch": 0.05124112499312015,
+      "grad_norm": 1.1748600006103516,
+      "learning_rate": 9.987030431799537e-06,
+      "loss": 0.8565,
+      "step": 931
+    },
+    {
+      "epoch": 0.05129616379547581,
+      "grad_norm": 1.879232406616211,
+      "learning_rate": 9.98699921219665e-06,
+      "loss": 0.8535,
+      "step": 932
+    },
+    {
+      "epoch": 0.05135120259783147,
+      "grad_norm": 0.9837847948074341,
+      "learning_rate": 9.986967955112812e-06,
+      "loss": 0.927,
+      "step": 933
+    },
+    {
+      "epoch": 0.05140624140018713,
+      "grad_norm": 0.8637211918830872,
+      "learning_rate": 9.986936660548257e-06,
+      "loss": 0.7903,
+      "step": 934
+    },
+    {
+      "epoch": 0.05146128020254279,
+      "grad_norm": 0.9078792929649353,
+      "learning_rate": 9.986905328503222e-06,
+      "loss": 0.9135,
+      "step": 935
+    },
+    {
+      "epoch": 0.051516319004898455,
+      "grad_norm": 0.9763005971908569,
+      "learning_rate": 9.98687395897794e-06,
+      "loss": 0.9006,
+      "step": 936
+    },
+    {
+      "epoch": 0.051571357807254116,
+      "grad_norm": 1.0174345970153809,
+      "learning_rate": 9.98684255197265e-06,
+      "loss": 0.9294,
+      "step": 937
+    },
+    {
+      "epoch": 0.05162639660960978,
+      "grad_norm": 0.8709769248962402,
+      "learning_rate": 9.986811107487584e-06,
+      "loss": 0.7986,
+      "step": 938
+    },
+    {
+      "epoch": 0.05168143541196544,
+      "grad_norm": 0.8717525601387024,
+      "learning_rate": 9.986779625522983e-06,
+      "loss": 0.8705,
+      "step": 939
+    },
+    {
+      "epoch": 0.05173647421432109,
+      "grad_norm": 0.9682945013046265,
+      "learning_rate": 9.98674810607908e-06,
+      "loss": 0.8127,
+      "step": 940
+    },
+    {
+      "epoch": 0.051791513016676755,
+      "grad_norm": 1.0248037576675415,
+      "learning_rate": 9.986716549156113e-06,
+      "loss": 0.9217,
+      "step": 941
+    },
+    {
+      "epoch": 0.051846551819032416,
+      "grad_norm": 0.9883397221565247,
+      "learning_rate": 9.98668495475432e-06,
+      "loss": 0.853,
+      "step": 942
+    },
+    {
+      "epoch": 0.05190159062138808,
+      "grad_norm": 0.9271108508110046,
+      "learning_rate": 9.986653322873937e-06,
+      "loss": 0.8807,
+      "step": 943
+    },
+    {
+      "epoch": 0.05195662942374374,
+      "grad_norm": 0.9027101397514343,
+      "learning_rate": 9.986621653515203e-06,
+      "loss": 0.88,
+      "step": 944
+    },
+    {
+      "epoch": 0.0520116682260994,
+      "grad_norm": 0.9807021617889404,
+      "learning_rate": 9.986589946678354e-06,
+      "loss": 0.8922,
+      "step": 945
+    },
+    {
+      "epoch": 0.05206670702845506,
+      "grad_norm": 0.8779157400131226,
+      "learning_rate": 9.98655820236363e-06,
+      "loss": 0.8988,
+      "step": 946
+    },
+    {
+      "epoch": 0.05212174583081072,
+      "grad_norm": 0.8182910680770874,
+      "learning_rate": 9.986526420571272e-06,
+      "loss": 0.7534,
+      "step": 947
+    },
+    {
+      "epoch": 0.052176784633166384,
+      "grad_norm": 0.9205981492996216,
+      "learning_rate": 9.986494601301513e-06,
+      "loss": 0.7516,
+      "step": 948
+    },
+    {
+      "epoch": 0.052231823435522046,
+      "grad_norm": 0.9829681515693665,
+      "learning_rate": 9.986462744554598e-06,
+      "loss": 0.9358,
+      "step": 949
+    },
+    {
+      "epoch": 0.05228686223787771,
+      "grad_norm": 0.8869890570640564,
+      "learning_rate": 9.986430850330762e-06,
+      "loss": 0.7933,
+      "step": 950
+    },
+    {
+      "epoch": 0.05234190104023336,
+      "grad_norm": 1.0226716995239258,
+      "learning_rate": 9.986398918630248e-06,
+      "loss": 0.9523,
+      "step": 951
+    },
+    {
+      "epoch": 0.05239693984258902,
+      "grad_norm": 0.9549778699874878,
+      "learning_rate": 9.986366949453293e-06,
+      "loss": 0.9368,
+      "step": 952
+    },
+    {
+      "epoch": 0.052451978644944684,
+      "grad_norm": 0.860454797744751,
+      "learning_rate": 9.98633494280014e-06,
+      "loss": 0.7618,
+      "step": 953
+    },
+    {
+      "epoch": 0.052507017447300346,
+      "grad_norm": 0.9623841643333435,
+      "learning_rate": 9.986302898671027e-06,
+      "loss": 0.8356,
+      "step": 954
+    },
+    {
+      "epoch": 0.05256205624965601,
+      "grad_norm": 0.9236606359481812,
+      "learning_rate": 9.986270817066196e-06,
+      "loss": 0.921,
+      "step": 955
+    },
+    {
+      "epoch": 0.05261709505201167,
+      "grad_norm": 1.0599812269210815,
+      "learning_rate": 9.98623869798589e-06,
+      "loss": 0.8082,
+      "step": 956
+    },
+    {
+      "epoch": 0.05267213385436733,
+      "grad_norm": 1.0321687459945679,
+      "learning_rate": 9.986206541430347e-06,
+      "loss": 0.9001,
+      "step": 957
+    },
+    {
+      "epoch": 0.05272717265672299,
+      "grad_norm": 0.8884543776512146,
+      "learning_rate": 9.986174347399813e-06,
+      "loss": 0.8317,
+      "step": 958
+    },
+    {
+      "epoch": 0.05278221145907865,
+      "grad_norm": 0.9592668414115906,
+      "learning_rate": 9.986142115894526e-06,
+      "loss": 0.9955,
+      "step": 959
+    },
+    {
+      "epoch": 0.052837250261434314,
+      "grad_norm": 0.9604032039642334,
+      "learning_rate": 9.986109846914729e-06,
+      "loss": 0.876,
+      "step": 960
+    },
+    {
+      "epoch": 0.052892289063789975,
+      "grad_norm": 0.9837536811828613,
+      "learning_rate": 9.986077540460664e-06,
+      "loss": 0.8247,
+      "step": 961
+    },
+    {
+      "epoch": 0.05294732786614563,
+      "grad_norm": 0.8570861220359802,
+      "learning_rate": 9.986045196532576e-06,
+      "loss": 0.879,
+      "step": 962
+    },
+    {
+      "epoch": 0.05300236666850129,
+      "grad_norm": 0.8441471457481384,
+      "learning_rate": 9.986012815130708e-06,
+      "loss": 0.8979,
+      "step": 963
+    },
+    {
+      "epoch": 0.05305740547085695,
+      "grad_norm": 0.8976197838783264,
+      "learning_rate": 9.985980396255302e-06,
+      "loss": 0.9382,
+      "step": 964
+    },
+    {
+      "epoch": 0.053112444273212614,
+      "grad_norm": 0.9685307741165161,
+      "learning_rate": 9.985947939906599e-06,
+      "loss": 0.8627,
+      "step": 965
+    },
+    {
+      "epoch": 0.053167483075568275,
+      "grad_norm": 0.8939018249511719,
+      "learning_rate": 9.98591544608485e-06,
+      "loss": 0.9221,
+      "step": 966
+    },
+    {
+      "epoch": 0.05322252187792394,
+      "grad_norm": 0.9218310713768005,
+      "learning_rate": 9.985882914790292e-06,
+      "loss": 0.8356,
+      "step": 967
+    },
+    {
+      "epoch": 0.0532775606802796,
+      "grad_norm": 0.9342261552810669,
+      "learning_rate": 9.985850346023174e-06,
+      "loss": 0.971,
+      "step": 968
+    },
+    {
+      "epoch": 0.05333259948263526,
+      "grad_norm": 1.0860705375671387,
+      "learning_rate": 9.985817739783741e-06,
+      "loss": 0.906,
+      "step": 969
+    },
+    {
+      "epoch": 0.05338763828499092,
+      "grad_norm": 0.8675006031990051,
+      "learning_rate": 9.985785096072234e-06,
+      "loss": 0.906,
+      "step": 970
+    },
+    {
+      "epoch": 0.05344267708734658,
+      "grad_norm": 0.8170626163482666,
+      "learning_rate": 9.985752414888903e-06,
+      "loss": 0.8109,
+      "step": 971
+    },
+    {
+      "epoch": 0.05349771588970224,
+      "grad_norm": 0.936434805393219,
+      "learning_rate": 9.98571969623399e-06,
+      "loss": 0.9219,
+      "step": 972
+    },
+    {
+      "epoch": 0.0535527546920579,
+      "grad_norm": 0.9316715002059937,
+      "learning_rate": 9.985686940107741e-06,
+      "loss": 0.8569,
+      "step": 973
+    },
+    {
+      "epoch": 0.05360779349441356,
+      "grad_norm": 1.183008074760437,
+      "learning_rate": 9.985654146510405e-06,
+      "loss": 0.837,
+      "step": 974
+    },
+    {
+      "epoch": 0.05366283229676922,
+      "grad_norm": 1.0788745880126953,
+      "learning_rate": 9.98562131544223e-06,
+      "loss": 0.8822,
+      "step": 975
+    },
+    {
+      "epoch": 0.05371787109912488,
+      "grad_norm": 0.9285461902618408,
+      "learning_rate": 9.985588446903455e-06,
+      "loss": 0.9279,
+      "step": 976
+    },
+    {
+      "epoch": 0.053772909901480544,
+      "grad_norm": 0.9389022588729858,
+      "learning_rate": 9.985555540894334e-06,
+      "loss": 0.9839,
+      "step": 977
+    },
+    {
+      "epoch": 0.053827948703836205,
+      "grad_norm": 0.8920616507530212,
+      "learning_rate": 9.985522597415112e-06,
+      "loss": 0.9205,
+      "step": 978
+    },
+    {
+      "epoch": 0.053882987506191866,
+      "grad_norm": 0.9755093455314636,
+      "learning_rate": 9.985489616466035e-06,
+      "loss": 0.8987,
+      "step": 979
+    },
+    {
+      "epoch": 0.05393802630854753,
+      "grad_norm": 0.96027010679245,
+      "learning_rate": 9.985456598047356e-06,
+      "loss": 0.8543,
+      "step": 980
+    },
+    {
+      "epoch": 0.05399306511090319,
+      "grad_norm": 1.0489718914031982,
+      "learning_rate": 9.985423542159317e-06,
+      "loss": 0.9179,
+      "step": 981
+    },
+    {
+      "epoch": 0.05404810391325885,
+      "grad_norm": 0.8665526509284973,
+      "learning_rate": 9.985390448802171e-06,
+      "loss": 0.9047,
+      "step": 982
+    },
+    {
+      "epoch": 0.054103142715614505,
+      "grad_norm": 0.8849464654922485,
+      "learning_rate": 9.985357317976163e-06,
+      "loss": 0.8892,
+      "step": 983
+    },
+    {
+      "epoch": 0.054158181517970166,
+      "grad_norm": 1.0083115100860596,
+      "learning_rate": 9.985324149681545e-06,
+      "loss": 0.7713,
+      "step": 984
+    },
+    {
+      "epoch": 0.05421322032032583,
+      "grad_norm": 0.8233863711357117,
+      "learning_rate": 9.985290943918565e-06,
+      "loss": 0.7967,
+      "step": 985
+    },
+    {
+      "epoch": 0.05426825912268149,
+      "grad_norm": 0.9615303874015808,
+      "learning_rate": 9.985257700687472e-06,
+      "loss": 0.8576,
+      "step": 986
+    },
+    {
+      "epoch": 0.05432329792503715,
+      "grad_norm": 0.8856416344642639,
+      "learning_rate": 9.985224419988517e-06,
+      "loss": 0.8614,
+      "step": 987
+    },
+    {
+      "epoch": 0.05437833672739281,
+      "grad_norm": 0.968325674533844,
+      "learning_rate": 9.98519110182195e-06,
+      "loss": 0.8247,
+      "step": 988
+    },
+    {
+      "epoch": 0.05443337552974847,
+      "grad_norm": 0.878402054309845,
+      "learning_rate": 9.985157746188021e-06,
+      "loss": 0.8661,
+      "step": 989
+    },
+    {
+      "epoch": 0.054488414332104135,
+      "grad_norm": 0.8376438021659851,
+      "learning_rate": 9.985124353086981e-06,
+      "loss": 0.8554,
+      "step": 990
+    },
+    {
+      "epoch": 0.054543453134459796,
+      "grad_norm": 1.0293036699295044,
+      "learning_rate": 9.98509092251908e-06,
+      "loss": 0.8049,
+      "step": 991
+    },
+    {
+      "epoch": 0.05459849193681546,
+      "grad_norm": 1.2345234155654907,
+      "learning_rate": 9.98505745448457e-06,
+      "loss": 1.0358,
+      "step": 992
+    },
+    {
+      "epoch": 0.05465353073917112,
+      "grad_norm": 0.9974482655525208,
+      "learning_rate": 9.985023948983703e-06,
+      "loss": 0.9329,
+      "step": 993
+    },
+    {
+      "epoch": 0.05470856954152677,
+      "grad_norm": 1.383955478668213,
+      "learning_rate": 9.984990406016732e-06,
+      "loss": 0.8688,
+      "step": 994
+    },
+    {
+      "epoch": 0.054763608343882435,
+      "grad_norm": 0.9369306564331055,
+      "learning_rate": 9.984956825583906e-06,
+      "loss": 0.8308,
+      "step": 995
+    },
+    {
+      "epoch": 0.054818647146238096,
+      "grad_norm": 0.8676120042800903,
+      "learning_rate": 9.984923207685478e-06,
+      "loss": 0.8283,
+      "step": 996
+    },
+    {
+      "epoch": 0.05487368594859376,
+      "grad_norm": 0.9218453168869019,
+      "learning_rate": 9.984889552321704e-06,
+      "loss": 0.7247,
+      "step": 997
+    },
+    {
+      "epoch": 0.05492872475094942,
+      "grad_norm": 0.8575478196144104,
+      "learning_rate": 9.984855859492833e-06,
+      "loss": 0.8462,
+      "step": 998
+    },
+    {
+      "epoch": 0.05498376355330508,
+      "grad_norm": 1.0042616128921509,
+      "learning_rate": 9.98482212919912e-06,
+      "loss": 0.9383,
+      "step": 999
+    },
+    {
+      "epoch": 0.05503880235566074,
+      "grad_norm": 0.8642181158065796,
+      "learning_rate": 9.984788361440817e-06,
+      "loss": 0.8805,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0550938411580164,
+      "grad_norm": 0.8413823843002319,
+      "learning_rate": 9.984754556218178e-06,
+      "loss": 0.8161,
+      "step": 1001
+    },
+    {
+      "epoch": 0.055148879960372064,
+      "grad_norm": 0.9473856091499329,
+      "learning_rate": 9.984720713531462e-06,
+      "loss": 0.8425,
+      "step": 1002
+    },
+    {
+      "epoch": 0.055203918762727726,
+      "grad_norm": 0.7854379415512085,
+      "learning_rate": 9.984686833380917e-06,
+      "loss": 0.7506,
+      "step": 1003
+    },
+    {
+      "epoch": 0.05525895756508339,
+      "grad_norm": 0.9481745958328247,
+      "learning_rate": 9.984652915766801e-06,
+      "loss": 0.954,
+      "step": 1004
+    },
+    {
+      "epoch": 0.05531399636743904,
+      "grad_norm": 0.767803966999054,
+      "learning_rate": 9.984618960689366e-06,
+      "loss": 0.8113,
+      "step": 1005
+    },
+    {
+      "epoch": 0.0553690351697947,
+      "grad_norm": 0.8957781195640564,
+      "learning_rate": 9.984584968148871e-06,
+      "loss": 0.9042,
+      "step": 1006
+    },
+    {
+      "epoch": 0.055424073972150364,
+      "grad_norm": 1.116646409034729,
+      "learning_rate": 9.98455093814557e-06,
+      "loss": 0.8648,
+      "step": 1007
+    },
+    {
+      "epoch": 0.055479112774506026,
+      "grad_norm": 0.9567018151283264,
+      "learning_rate": 9.98451687067972e-06,
+      "loss": 0.9446,
+      "step": 1008
+    },
+    {
+      "epoch": 0.05553415157686169,
+      "grad_norm": 0.8470665812492371,
+      "learning_rate": 9.98448276575157e-06,
+      "loss": 0.8186,
+      "step": 1009
+    },
+    {
+      "epoch": 0.05558919037921735,
+      "grad_norm": 0.9595193862915039,
+      "learning_rate": 9.984448623361387e-06,
+      "loss": 0.8406,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05564422918157301,
+      "grad_norm": 1.0579735040664673,
+      "learning_rate": 9.98441444350942e-06,
+      "loss": 0.9676,
+      "step": 1011
+    },
+    {
+      "epoch": 0.05569926798392867,
+      "grad_norm": 0.8693701028823853,
+      "learning_rate": 9.98438022619593e-06,
+      "loss": 0.9451,
+      "step": 1012
+    },
+    {
+      "epoch": 0.05575430678628433,
+      "grad_norm": 0.9251859784126282,
+      "learning_rate": 9.98434597142117e-06,
+      "loss": 0.7858,
+      "step": 1013
+    },
+    {
+      "epoch": 0.055809345588639994,
+      "grad_norm": 0.8584280014038086,
+      "learning_rate": 9.984311679185402e-06,
+      "loss": 0.8481,
+      "step": 1014
+    },
+    {
+      "epoch": 0.05586438439099565,
+      "grad_norm": 0.8903968334197998,
+      "learning_rate": 9.98427734948888e-06,
+      "loss": 0.7832,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05591942319335131,
+      "grad_norm": 0.905581533908844,
+      "learning_rate": 9.984242982331864e-06,
+      "loss": 0.8088,
+      "step": 1016
+    },
+    {
+      "epoch": 0.05597446199570697,
+      "grad_norm": 0.9866476655006409,
+      "learning_rate": 9.984208577714612e-06,
+      "loss": 0.8366,
+      "step": 1017
+    },
+    {
+      "epoch": 0.05602950079806263,
+      "grad_norm": 0.8843809962272644,
+      "learning_rate": 9.984174135637384e-06,
+      "loss": 0.8961,
+      "step": 1018
+    },
+    {
+      "epoch": 0.056084539600418294,
+      "grad_norm": 0.9071753621101379,
+      "learning_rate": 9.984139656100435e-06,
+      "loss": 0.8671,
+      "step": 1019
+    },
+    {
+      "epoch": 0.056139578402773956,
+      "grad_norm": 0.9894018173217773,
+      "learning_rate": 9.984105139104028e-06,
+      "loss": 0.9099,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05619461720512962,
+      "grad_norm": 0.8432741165161133,
+      "learning_rate": 9.98407058464842e-06,
+      "loss": 0.7817,
+      "step": 1021
+    },
+    {
+      "epoch": 0.05624965600748528,
+      "grad_norm": 0.9538390040397644,
+      "learning_rate": 9.984035992733873e-06,
+      "loss": 0.8689,
+      "step": 1022
+    },
+    {
+      "epoch": 0.05630469480984094,
+      "grad_norm": 0.9263421297073364,
+      "learning_rate": 9.984001363360645e-06,
+      "loss": 0.9066,
+      "step": 1023
+    },
+    {
+      "epoch": 0.0563597336121966,
+      "grad_norm": 0.8921047449111938,
+      "learning_rate": 9.983966696528996e-06,
+      "loss": 0.8304,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05641477241455226,
+      "grad_norm": 0.8379812240600586,
+      "learning_rate": 9.983931992239188e-06,
+      "loss": 0.866,
+      "step": 1025
+    },
+    {
+      "epoch": 0.05646981121690792,
+      "grad_norm": 0.9444219470024109,
+      "learning_rate": 9.983897250491481e-06,
+      "loss": 0.9456,
+      "step": 1026
+    },
+    {
+      "epoch": 0.05652485001926358,
+      "grad_norm": 1.0268759727478027,
+      "learning_rate": 9.983862471286137e-06,
+      "loss": 0.8277,
+      "step": 1027
+    },
+    {
+      "epoch": 0.05657988882161924,
+      "grad_norm": 1.3949217796325684,
+      "learning_rate": 9.983827654623418e-06,
+      "loss": 0.9721,
+      "step": 1028
+    },
+    {
+      "epoch": 0.0566349276239749,
+      "grad_norm": 0.8899377584457397,
+      "learning_rate": 9.983792800503582e-06,
+      "loss": 0.8794,
+      "step": 1029
+    },
+    {
+      "epoch": 0.05668996642633056,
+      "grad_norm": 0.989072322845459,
+      "learning_rate": 9.983757908926895e-06,
+      "loss": 0.8852,
+      "step": 1030
+    },
+    {
+      "epoch": 0.056745005228686224,
+      "grad_norm": 0.9797759056091309,
+      "learning_rate": 9.983722979893615e-06,
+      "loss": 1.0405,
+      "step": 1031
+    },
+    {
+      "epoch": 0.056800044031041885,
+      "grad_norm": 0.9044767618179321,
+      "learning_rate": 9.98368801340401e-06,
+      "loss": 0.7243,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05685508283339755,
+      "grad_norm": 1.116324782371521,
+      "learning_rate": 9.983653009458338e-06,
+      "loss": 0.9183,
+      "step": 1033
+    },
+    {
+      "epoch": 0.05691012163575321,
+      "grad_norm": 0.9373337030410767,
+      "learning_rate": 9.983617968056866e-06,
+      "loss": 0.9417,
+      "step": 1034
+    },
+    {
+      "epoch": 0.05696516043810887,
+      "grad_norm": 1.0587197542190552,
+      "learning_rate": 9.983582889199855e-06,
+      "loss": 0.896,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05702019924046453,
+      "grad_norm": 1.0080119371414185,
+      "learning_rate": 9.983547772887568e-06,
+      "loss": 0.924,
+      "step": 1036
+    },
+    {
+      "epoch": 0.057075238042820185,
+      "grad_norm": 0.847091019153595,
+      "learning_rate": 9.98351261912027e-06,
+      "loss": 0.7443,
+      "step": 1037
+    },
+    {
+      "epoch": 0.05713027684517585,
+      "grad_norm": 0.9876272082328796,
+      "learning_rate": 9.983477427898225e-06,
+      "loss": 0.9365,
+      "step": 1038
+    },
+    {
+      "epoch": 0.05718531564753151,
+      "grad_norm": 0.9188169240951538,
+      "learning_rate": 9.983442199221698e-06,
+      "loss": 0.9213,
+      "step": 1039
+    },
+    {
+      "epoch": 0.05724035444988717,
+      "grad_norm": 0.932399332523346,
+      "learning_rate": 9.983406933090954e-06,
+      "loss": 0.958,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05729539325224283,
+      "grad_norm": 0.9126465320587158,
+      "learning_rate": 9.983371629506258e-06,
+      "loss": 0.8913,
+      "step": 1041
+    },
+    {
+      "epoch": 0.05735043205459849,
+      "grad_norm": 0.80904620885849,
+      "learning_rate": 9.983336288467873e-06,
+      "loss": 0.7719,
+      "step": 1042
+    },
+    {
+      "epoch": 0.057405470856954154,
+      "grad_norm": 0.873833417892456,
+      "learning_rate": 9.983300909976067e-06,
+      "loss": 0.9201,
+      "step": 1043
+    },
+    {
+      "epoch": 0.057460509659309815,
+      "grad_norm": 0.8331829309463501,
+      "learning_rate": 9.983265494031107e-06,
+      "loss": 0.8605,
+      "step": 1044
+    },
+    {
+      "epoch": 0.057515548461665476,
+      "grad_norm": 0.8364768624305725,
+      "learning_rate": 9.983230040633255e-06,
+      "loss": 0.8627,
+      "step": 1045
+    },
+    {
+      "epoch": 0.05757058726402114,
+      "grad_norm": 0.9226736426353455,
+      "learning_rate": 9.98319454978278e-06,
+      "loss": 0.9759,
+      "step": 1046
+    },
+    {
+      "epoch": 0.05762562606637679,
+      "grad_norm": 0.8174427151679993,
+      "learning_rate": 9.98315902147995e-06,
+      "loss": 0.8066,
+      "step": 1047
+    },
+    {
+      "epoch": 0.057680664868732454,
+      "grad_norm": 0.9154924750328064,
+      "learning_rate": 9.98312345572503e-06,
+      "loss": 0.9112,
+      "step": 1048
+    },
+    {
+      "epoch": 0.057735703671088115,
+      "grad_norm": 0.8884655237197876,
+      "learning_rate": 9.983087852518289e-06,
+      "loss": 0.8699,
+      "step": 1049
+    },
+    {
+      "epoch": 0.057790742473443776,
+      "grad_norm": 0.8849230408668518,
+      "learning_rate": 9.983052211859992e-06,
+      "loss": 0.8999,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05784578127579944,
+      "grad_norm": 1.025843858718872,
+      "learning_rate": 9.98301653375041e-06,
+      "loss": 0.7764,
+      "step": 1051
+    },
+    {
+      "epoch": 0.0579008200781551,
+      "grad_norm": 0.900505006313324,
+      "learning_rate": 9.98298081818981e-06,
+      "loss": 0.9196,
+      "step": 1052
+    },
+    {
+      "epoch": 0.05795585888051076,
+      "grad_norm": 0.9506704211235046,
+      "learning_rate": 9.982945065178457e-06,
+      "loss": 0.8319,
+      "step": 1053
+    },
+    {
+      "epoch": 0.05801089768286642,
+      "grad_norm": 0.9439849853515625,
+      "learning_rate": 9.982909274716626e-06,
+      "loss": 0.8561,
+      "step": 1054
+    },
+    {
+      "epoch": 0.05806593648522208,
+      "grad_norm": 0.8761240243911743,
+      "learning_rate": 9.982873446804579e-06,
+      "loss": 0.9681,
+      "step": 1055
+    },
+    {
+      "epoch": 0.058120975287577745,
+      "grad_norm": 0.8756145238876343,
+      "learning_rate": 9.982837581442592e-06,
+      "loss": 0.8452,
+      "step": 1056
+    },
+    {
+      "epoch": 0.058176014089933406,
+      "grad_norm": 0.8732383847236633,
+      "learning_rate": 9.982801678630932e-06,
+      "loss": 0.9018,
+      "step": 1057
+    },
+    {
+      "epoch": 0.05823105289228906,
+      "grad_norm": 0.8338272571563721,
+      "learning_rate": 9.982765738369867e-06,
+      "loss": 0.9308,
+      "step": 1058
+    },
+    {
+      "epoch": 0.05828609169464472,
+      "grad_norm": 0.843163013458252,
+      "learning_rate": 9.982729760659669e-06,
+      "loss": 0.7802,
+      "step": 1059
+    },
+    {
+      "epoch": 0.05834113049700038,
+      "grad_norm": 1.2007580995559692,
+      "learning_rate": 9.982693745500606e-06,
+      "loss": 0.8406,
+      "step": 1060
+    },
+    {
+      "epoch": 0.058396169299356045,
+      "grad_norm": 0.8760073781013489,
+      "learning_rate": 9.982657692892954e-06,
+      "loss": 0.8528,
+      "step": 1061
+    },
+    {
+      "epoch": 0.058451208101711706,
+      "grad_norm": 0.925309419631958,
+      "learning_rate": 9.982621602836978e-06,
+      "loss": 0.9601,
+      "step": 1062
+    },
+    {
+      "epoch": 0.05850624690406737,
+      "grad_norm": 0.9277135133743286,
+      "learning_rate": 9.982585475332952e-06,
+      "loss": 0.8405,
+      "step": 1063
+    },
+    {
+      "epoch": 0.05856128570642303,
+      "grad_norm": 0.928044319152832,
+      "learning_rate": 9.98254931038115e-06,
+      "loss": 0.8259,
+      "step": 1064
+    },
+    {
+      "epoch": 0.05861632450877869,
+      "grad_norm": 0.8363838195800781,
+      "learning_rate": 9.982513107981837e-06,
+      "loss": 0.8655,
+      "step": 1065
+    },
+    {
+      "epoch": 0.05867136331113435,
+      "grad_norm": 0.9800984859466553,
+      "learning_rate": 9.982476868135292e-06,
+      "loss": 0.9285,
+      "step": 1066
+    },
+    {
+      "epoch": 0.05872640211349001,
+      "grad_norm": 0.8062636256217957,
+      "learning_rate": 9.982440590841785e-06,
+      "loss": 0.754,
+      "step": 1067
+    },
+    {
+      "epoch": 0.058781440915845674,
+      "grad_norm": 1.2010705471038818,
+      "learning_rate": 9.982404276101586e-06,
+      "loss": 0.9872,
+      "step": 1068
+    },
+    {
+      "epoch": 0.05883647971820133,
+      "grad_norm": 1.0036406517028809,
+      "learning_rate": 9.982367923914971e-06,
+      "loss": 0.8724,
+      "step": 1069
+    },
+    {
+      "epoch": 0.05889151852055699,
+      "grad_norm": 0.8768866658210754,
+      "learning_rate": 9.982331534282212e-06,
+      "loss": 0.838,
+      "step": 1070
+    },
+    {
+      "epoch": 0.05894655732291265,
+      "grad_norm": 0.7892739176750183,
+      "learning_rate": 9.982295107203584e-06,
+      "loss": 0.6974,
+      "step": 1071
+    },
+    {
+      "epoch": 0.05900159612526831,
+      "grad_norm": 0.863315999507904,
+      "learning_rate": 9.982258642679358e-06,
+      "loss": 0.9282,
+      "step": 1072
+    },
+    {
+      "epoch": 0.059056634927623974,
+      "grad_norm": 0.8645132780075073,
+      "learning_rate": 9.982222140709812e-06,
+      "loss": 0.8504,
+      "step": 1073
+    },
+    {
+      "epoch": 0.059111673729979636,
+      "grad_norm": 1.0003199577331543,
+      "learning_rate": 9.982185601295216e-06,
+      "loss": 1.0293,
+      "step": 1074
+    },
+    {
+      "epoch": 0.0591667125323353,
+      "grad_norm": 0.8391831517219543,
+      "learning_rate": 9.982149024435848e-06,
+      "loss": 0.8609,
+      "step": 1075
+    },
+    {
+      "epoch": 0.05922175133469096,
+      "grad_norm": 0.9940230846405029,
+      "learning_rate": 9.982112410131981e-06,
+      "loss": 0.9623,
+      "step": 1076
+    },
+    {
+      "epoch": 0.05927679013704662,
+      "grad_norm": 1.0670262575149536,
+      "learning_rate": 9.98207575838389e-06,
+      "loss": 0.9952,
+      "step": 1077
+    },
+    {
+      "epoch": 0.05933182893940228,
+      "grad_norm": 0.8506165742874146,
+      "learning_rate": 9.982039069191853e-06,
+      "loss": 0.8401,
+      "step": 1078
+    },
+    {
+      "epoch": 0.05938686774175794,
+      "grad_norm": 0.8956409096717834,
+      "learning_rate": 9.982002342556144e-06,
+      "loss": 0.8779,
+      "step": 1079
+    },
+    {
+      "epoch": 0.0594419065441136,
+      "grad_norm": 0.8955749273300171,
+      "learning_rate": 9.981965578477038e-06,
+      "loss": 0.8946,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05949694534646926,
+      "grad_norm": 0.9035234451293945,
+      "learning_rate": 9.981928776954811e-06,
+      "loss": 0.9352,
+      "step": 1081
+    },
+    {
+      "epoch": 0.05955198414882492,
+      "grad_norm": 0.8748759627342224,
+      "learning_rate": 9.981891937989743e-06,
+      "loss": 0.8803,
+      "step": 1082
+    },
+    {
+      "epoch": 0.05960702295118058,
+      "grad_norm": 0.9966281056404114,
+      "learning_rate": 9.981855061582108e-06,
+      "loss": 0.9304,
+      "step": 1083
+    },
+    {
+      "epoch": 0.05966206175353624,
+      "grad_norm": 0.8696668148040771,
+      "learning_rate": 9.981818147732183e-06,
+      "loss": 0.8706,
+      "step": 1084
+    },
+    {
+      "epoch": 0.059717100555891904,
+      "grad_norm": 0.9823188185691833,
+      "learning_rate": 9.981781196440249e-06,
+      "loss": 0.9431,
+      "step": 1085
+    },
+    {
+      "epoch": 0.059772139358247565,
+      "grad_norm": 0.8401583433151245,
+      "learning_rate": 9.981744207706577e-06,
+      "loss": 0.8369,
+      "step": 1086
+    },
+    {
+      "epoch": 0.05982717816060323,
+      "grad_norm": 0.8775757551193237,
+      "learning_rate": 9.981707181531452e-06,
+      "loss": 0.9516,
+      "step": 1087
+    },
+    {
+      "epoch": 0.05988221696295889,
+      "grad_norm": 0.9153465628623962,
+      "learning_rate": 9.981670117915148e-06,
+      "loss": 0.8997,
+      "step": 1088
+    },
+    {
+      "epoch": 0.05993725576531455,
+      "grad_norm": 0.9053078889846802,
+      "learning_rate": 9.981633016857946e-06,
+      "loss": 0.9452,
+      "step": 1089
+    },
+    {
+      "epoch": 0.059992294567670204,
+      "grad_norm": 0.9154480695724487,
+      "learning_rate": 9.981595878360123e-06,
+      "loss": 0.8293,
+      "step": 1090
+    },
+    {
+      "epoch": 0.060047333370025865,
+      "grad_norm": 0.85718834400177,
+      "learning_rate": 9.981558702421958e-06,
+      "loss": 0.876,
+      "step": 1091
+    },
+    {
+      "epoch": 0.06010237217238153,
+      "grad_norm": 0.9437130689620972,
+      "learning_rate": 9.981521489043734e-06,
+      "loss": 0.9731,
+      "step": 1092
+    },
+    {
+      "epoch": 0.06015741097473719,
+      "grad_norm": 0.9014891386032104,
+      "learning_rate": 9.981484238225725e-06,
+      "loss": 0.811,
+      "step": 1093
+    },
+    {
+      "epoch": 0.06021244977709285,
+      "grad_norm": 0.8942846655845642,
+      "learning_rate": 9.981446949968216e-06,
+      "loss": 0.808,
+      "step": 1094
+    },
+    {
+      "epoch": 0.06026748857944851,
+      "grad_norm": 0.855297863483429,
+      "learning_rate": 9.981409624271483e-06,
+      "loss": 0.8319,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06032252738180417,
+      "grad_norm": 0.9310913681983948,
+      "learning_rate": 9.981372261135811e-06,
+      "loss": 0.899,
+      "step": 1096
+    },
+    {
+      "epoch": 0.060377566184159834,
+      "grad_norm": 0.8472979664802551,
+      "learning_rate": 9.981334860561478e-06,
+      "loss": 0.8818,
+      "step": 1097
+    },
+    {
+      "epoch": 0.060432604986515495,
+      "grad_norm": 0.896617591381073,
+      "learning_rate": 9.981297422548764e-06,
+      "loss": 0.8991,
+      "step": 1098
+    },
+    {
+      "epoch": 0.06048764378887116,
+      "grad_norm": 0.8543037176132202,
+      "learning_rate": 9.981259947097954e-06,
+      "loss": 0.8595,
+      "step": 1099
+    },
+    {
+      "epoch": 0.06054268259122682,
+      "grad_norm": 0.8794904947280884,
+      "learning_rate": 9.981222434209327e-06,
+      "loss": 0.8561,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06059772139358247,
+      "grad_norm": 0.8882116675376892,
+      "learning_rate": 9.981184883883165e-06,
+      "loss": 0.8099,
+      "step": 1101
+    },
+    {
+      "epoch": 0.060652760195938134,
+      "grad_norm": 1.0068262815475464,
+      "learning_rate": 9.98114729611975e-06,
+      "loss": 0.8104,
+      "step": 1102
+    },
+    {
+      "epoch": 0.060707798998293795,
+      "grad_norm": 1.072316288948059,
+      "learning_rate": 9.981109670919366e-06,
+      "loss": 0.9877,
+      "step": 1103
+    },
+    {
+      "epoch": 0.06076283780064946,
+      "grad_norm": 0.9959045052528381,
+      "learning_rate": 9.981072008282298e-06,
+      "loss": 0.906,
+      "step": 1104
+    },
+    {
+      "epoch": 0.06081787660300512,
+      "grad_norm": 0.8712790608406067,
+      "learning_rate": 9.981034308208823e-06,
+      "loss": 0.8725,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06087291540536078,
+      "grad_norm": 0.9114679098129272,
+      "learning_rate": 9.980996570699228e-06,
+      "loss": 0.8385,
+      "step": 1106
+    },
+    {
+      "epoch": 0.06092795420771644,
+      "grad_norm": 1.0024466514587402,
+      "learning_rate": 9.980958795753796e-06,
+      "loss": 0.8661,
+      "step": 1107
+    },
+    {
+      "epoch": 0.0609829930100721,
+      "grad_norm": 0.9578461050987244,
+      "learning_rate": 9.98092098337281e-06,
+      "loss": 0.9358,
+      "step": 1108
+    },
+    {
+      "epoch": 0.061038031812427763,
+      "grad_norm": 0.8677787780761719,
+      "learning_rate": 9.980883133556557e-06,
+      "loss": 0.8146,
+      "step": 1109
+    },
+    {
+      "epoch": 0.061093070614783425,
+      "grad_norm": 0.9072276949882507,
+      "learning_rate": 9.98084524630532e-06,
+      "loss": 0.91,
+      "step": 1110
+    },
+    {
+      "epoch": 0.061148109417139086,
+      "grad_norm": 0.8827292919158936,
+      "learning_rate": 9.980807321619381e-06,
+      "loss": 0.8854,
+      "step": 1111
+    },
+    {
+      "epoch": 0.06120314821949474,
+      "grad_norm": 1.0012744665145874,
+      "learning_rate": 9.98076935949903e-06,
+      "loss": 0.8242,
+      "step": 1112
+    },
+    {
+      "epoch": 0.0612581870218504,
+      "grad_norm": 0.9152620434761047,
+      "learning_rate": 9.980731359944548e-06,
+      "loss": 0.8832,
+      "step": 1113
+    },
+    {
+      "epoch": 0.061313225824206063,
+      "grad_norm": 0.8986824750900269,
+      "learning_rate": 9.980693322956222e-06,
+      "loss": 0.7975,
+      "step": 1114
+    },
+    {
+      "epoch": 0.061368264626561725,
+      "grad_norm": 0.9373019933700562,
+      "learning_rate": 9.98065524853434e-06,
+      "loss": 0.9541,
+      "step": 1115
+    },
+    {
+      "epoch": 0.061423303428917386,
+      "grad_norm": 0.9875593781471252,
+      "learning_rate": 9.980617136679185e-06,
+      "loss": 1.0052,
+      "step": 1116
+    },
+    {
+      "epoch": 0.06147834223127305,
+      "grad_norm": 1.0664819478988647,
+      "learning_rate": 9.980578987391045e-06,
+      "loss": 0.8584,
+      "step": 1117
+    },
+    {
+      "epoch": 0.06153338103362871,
+      "grad_norm": 0.9149377942085266,
+      "learning_rate": 9.980540800670207e-06,
+      "loss": 0.8467,
+      "step": 1118
+    },
+    {
+      "epoch": 0.06158841983598437,
+      "grad_norm": 0.9303194284439087,
+      "learning_rate": 9.980502576516959e-06,
+      "loss": 0.8219,
+      "step": 1119
+    },
+    {
+      "epoch": 0.06164345863834003,
+      "grad_norm": 0.9059457778930664,
+      "learning_rate": 9.980464314931583e-06,
+      "loss": 0.8459,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06169849744069569,
+      "grad_norm": 0.9368849396705627,
+      "learning_rate": 9.980426015914375e-06,
+      "loss": 0.8933,
+      "step": 1121
+    },
+    {
+      "epoch": 0.061753536243051355,
+      "grad_norm": 0.8188626766204834,
+      "learning_rate": 9.980387679465615e-06,
+      "loss": 0.807,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06180857504540701,
+      "grad_norm": 1.027171015739441,
+      "learning_rate": 9.980349305585595e-06,
+      "loss": 0.8919,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06186361384776267,
+      "grad_norm": 0.831649899482727,
+      "learning_rate": 9.980310894274603e-06,
+      "loss": 0.8109,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06191865265011833,
+      "grad_norm": 1.0170252323150635,
+      "learning_rate": 9.980272445532928e-06,
+      "loss": 0.9537,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06197369145247399,
+      "grad_norm": 0.97837233543396,
+      "learning_rate": 9.980233959360858e-06,
+      "loss": 0.9104,
+      "step": 1126
+    },
+    {
+      "epoch": 0.062028730254829655,
+      "grad_norm": 0.9548324942588806,
+      "learning_rate": 9.980195435758681e-06,
+      "loss": 0.9473,
+      "step": 1127
+    },
+    {
+      "epoch": 0.062083769057185316,
+      "grad_norm": 0.8675842881202698,
+      "learning_rate": 9.980156874726692e-06,
+      "loss": 0.8313,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06213880785954098,
+      "grad_norm": 0.8948968052864075,
+      "learning_rate": 9.980118276265173e-06,
+      "loss": 0.8008,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06219384666189664,
+      "grad_norm": 0.8914239406585693,
+      "learning_rate": 9.98007964037442e-06,
+      "loss": 0.7642,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0622488854642523,
+      "grad_norm": 0.9499951004981995,
+      "learning_rate": 9.980040967054723e-06,
+      "loss": 0.8669,
+      "step": 1131
+    },
+    {
+      "epoch": 0.06230392426660796,
+      "grad_norm": 0.8959251642227173,
+      "learning_rate": 9.980002256306369e-06,
+      "loss": 0.9177,
+      "step": 1132
+    },
+    {
+      "epoch": 0.062358963068963616,
+      "grad_norm": 0.8634380102157593,
+      "learning_rate": 9.97996350812965e-06,
+      "loss": 0.8252,
+      "step": 1133
+    },
+    {
+      "epoch": 0.06241400187131928,
+      "grad_norm": 0.9380598068237305,
+      "learning_rate": 9.97992472252486e-06,
+      "loss": 0.9335,
+      "step": 1134
+    },
+    {
+      "epoch": 0.06246904067367494,
+      "grad_norm": 0.8373183608055115,
+      "learning_rate": 9.97988589949229e-06,
+      "loss": 0.848,
+      "step": 1135
+    },
+    {
+      "epoch": 0.0625240794760306,
+      "grad_norm": 0.9649023413658142,
+      "learning_rate": 9.97984703903223e-06,
+      "loss": 0.9648,
+      "step": 1136
+    },
+    {
+      "epoch": 0.06257911827838626,
+      "grad_norm": 0.9972373843193054,
+      "learning_rate": 9.979808141144972e-06,
+      "loss": 0.9104,
+      "step": 1137
+    },
+    {
+      "epoch": 0.06263415708074192,
+      "grad_norm": 0.8230985403060913,
+      "learning_rate": 9.97976920583081e-06,
+      "loss": 0.8393,
+      "step": 1138
+    },
+    {
+      "epoch": 0.06268919588309758,
+      "grad_norm": 0.9775324463844299,
+      "learning_rate": 9.979730233090034e-06,
+      "loss": 0.8385,
+      "step": 1139
+    },
+    {
+      "epoch": 0.06274423468545325,
+      "grad_norm": 0.8288110494613647,
+      "learning_rate": 9.97969122292294e-06,
+      "loss": 0.7308,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06279927348780891,
+      "grad_norm": 0.8980758786201477,
+      "learning_rate": 9.979652175329819e-06,
+      "loss": 0.863,
+      "step": 1141
+    },
+    {
+      "epoch": 0.06285431229016457,
+      "grad_norm": 7.43889045715332,
+      "learning_rate": 9.979613090310965e-06,
+      "loss": 0.9412,
+      "step": 1142
+    },
+    {
+      "epoch": 0.06290935109252023,
+      "grad_norm": 0.9758191704750061,
+      "learning_rate": 9.97957396786667e-06,
+      "loss": 0.8896,
+      "step": 1143
+    },
+    {
+      "epoch": 0.06296438989487589,
+      "grad_norm": 0.8211693167686462,
+      "learning_rate": 9.979534807997234e-06,
+      "loss": 0.7352,
+      "step": 1144
+    },
+    {
+      "epoch": 0.06301942869723155,
+      "grad_norm": 0.8643441796302795,
+      "learning_rate": 9.979495610702945e-06,
+      "loss": 0.8701,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06307446749958721,
+      "grad_norm": 1.0199437141418457,
+      "learning_rate": 9.9794563759841e-06,
+      "loss": 0.9025,
+      "step": 1146
+    },
+    {
+      "epoch": 0.06312950630194288,
+      "grad_norm": 0.8367893695831299,
+      "learning_rate": 9.979417103840994e-06,
+      "loss": 0.8491,
+      "step": 1147
+    },
+    {
+      "epoch": 0.06318454510429854,
+      "grad_norm": 0.9411819577217102,
+      "learning_rate": 9.979377794273923e-06,
+      "loss": 0.8501,
+      "step": 1148
+    },
+    {
+      "epoch": 0.06323958390665418,
+      "grad_norm": 1.1497365236282349,
+      "learning_rate": 9.97933844728318e-06,
+      "loss": 1.0227,
+      "step": 1149
+    },
+    {
+      "epoch": 0.06329462270900985,
+      "grad_norm": 0.9892984628677368,
+      "learning_rate": 9.979299062869064e-06,
+      "loss": 0.8942,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06334966151136551,
+      "grad_norm": 0.947952926158905,
+      "learning_rate": 9.979259641031867e-06,
+      "loss": 1.0149,
+      "step": 1151
+    },
+    {
+      "epoch": 0.06340470031372117,
+      "grad_norm": 0.9060251712799072,
+      "learning_rate": 9.979220181771889e-06,
+      "loss": 0.8607,
+      "step": 1152
+    },
+    {
+      "epoch": 0.06345973911607683,
+      "grad_norm": 0.8331984281539917,
+      "learning_rate": 9.979180685089424e-06,
+      "loss": 0.8777,
+      "step": 1153
+    },
+    {
+      "epoch": 0.06351477791843249,
+      "grad_norm": 0.9133188724517822,
+      "learning_rate": 9.97914115098477e-06,
+      "loss": 0.7409,
+      "step": 1154
+    },
+    {
+      "epoch": 0.06356981672078815,
+      "grad_norm": 0.9095513820648193,
+      "learning_rate": 9.979101579458224e-06,
+      "loss": 0.8938,
+      "step": 1155
+    },
+    {
+      "epoch": 0.06362485552314381,
+      "grad_norm": 0.9584553241729736,
+      "learning_rate": 9.979061970510082e-06,
+      "loss": 0.8765,
+      "step": 1156
+    },
+    {
+      "epoch": 0.06367989432549948,
+      "grad_norm": 0.8742124438285828,
+      "learning_rate": 9.979022324140644e-06,
+      "loss": 0.8564,
+      "step": 1157
+    },
+    {
+      "epoch": 0.06373493312785514,
+      "grad_norm": 0.8776904344558716,
+      "learning_rate": 9.978982640350208e-06,
+      "loss": 0.8713,
+      "step": 1158
+    },
+    {
+      "epoch": 0.0637899719302108,
+      "grad_norm": 0.8667464852333069,
+      "learning_rate": 9.97894291913907e-06,
+      "loss": 0.8705,
+      "step": 1159
+    },
+    {
+      "epoch": 0.06384501073256646,
+      "grad_norm": 0.9028087854385376,
+      "learning_rate": 9.978903160507531e-06,
+      "loss": 0.8297,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06390004953492212,
+      "grad_norm": 0.900812029838562,
+      "learning_rate": 9.978863364455887e-06,
+      "loss": 0.8456,
+      "step": 1161
+    },
+    {
+      "epoch": 0.06395508833727778,
+      "grad_norm": 0.9667207598686218,
+      "learning_rate": 9.97882353098444e-06,
+      "loss": 0.8081,
+      "step": 1162
+    },
+    {
+      "epoch": 0.06401012713963344,
+      "grad_norm": 0.8959711194038391,
+      "learning_rate": 9.978783660093488e-06,
+      "loss": 0.8455,
+      "step": 1163
+    },
+    {
+      "epoch": 0.0640651659419891,
+      "grad_norm": 0.8519117832183838,
+      "learning_rate": 9.97874375178333e-06,
+      "loss": 0.849,
+      "step": 1164
+    },
+    {
+      "epoch": 0.06412020474434477,
+      "grad_norm": 1.0532654523849487,
+      "learning_rate": 9.978703806054267e-06,
+      "loss": 0.7356,
+      "step": 1165
+    },
+    {
+      "epoch": 0.06417524354670043,
+      "grad_norm": 1.0313252210617065,
+      "learning_rate": 9.9786638229066e-06,
+      "loss": 1.024,
+      "step": 1166
+    },
+    {
+      "epoch": 0.06423028234905609,
+      "grad_norm": 1.0567537546157837,
+      "learning_rate": 9.978623802340627e-06,
+      "loss": 0.9423,
+      "step": 1167
+    },
+    {
+      "epoch": 0.06428532115141175,
+      "grad_norm": 0.8198097348213196,
+      "learning_rate": 9.97858374435665e-06,
+      "loss": 0.829,
+      "step": 1168
+    },
+    {
+      "epoch": 0.06434035995376741,
+      "grad_norm": 0.8718193173408508,
+      "learning_rate": 9.97854364895497e-06,
+      "loss": 0.7184,
+      "step": 1169
+    },
+    {
+      "epoch": 0.06439539875612307,
+      "grad_norm": 0.8037594556808472,
+      "learning_rate": 9.978503516135892e-06,
+      "loss": 0.7961,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06445043755847872,
+      "grad_norm": 0.9052229523658752,
+      "learning_rate": 9.978463345899709e-06,
+      "loss": 0.8016,
+      "step": 1171
+    },
+    {
+      "epoch": 0.06450547636083438,
+      "grad_norm": 1.0194638967514038,
+      "learning_rate": 9.978423138246731e-06,
+      "loss": 0.9045,
+      "step": 1172
+    },
+    {
+      "epoch": 0.06456051516319004,
+      "grad_norm": 0.953078031539917,
+      "learning_rate": 9.978382893177259e-06,
+      "loss": 0.9661,
+      "step": 1173
+    },
+    {
+      "epoch": 0.0646155539655457,
+      "grad_norm": 0.8777341842651367,
+      "learning_rate": 9.978342610691592e-06,
+      "loss": 0.8685,
+      "step": 1174
+    },
+    {
+      "epoch": 0.06467059276790137,
+      "grad_norm": 1.0178394317626953,
+      "learning_rate": 9.978302290790034e-06,
+      "loss": 0.9075,
+      "step": 1175
+    },
+    {
+      "epoch": 0.06472563157025703,
+      "grad_norm": 0.935694694519043,
+      "learning_rate": 9.978261933472889e-06,
+      "loss": 0.8438,
+      "step": 1176
+    },
+    {
+      "epoch": 0.06478067037261269,
+      "grad_norm": 1.0022411346435547,
+      "learning_rate": 9.97822153874046e-06,
+      "loss": 0.8701,
+      "step": 1177
+    },
+    {
+      "epoch": 0.06483570917496835,
+      "grad_norm": 1.0371203422546387,
+      "learning_rate": 9.97818110659305e-06,
+      "loss": 0.9111,
+      "step": 1178
+    },
+    {
+      "epoch": 0.06489074797732401,
+      "grad_norm": 0.7972478866577148,
+      "learning_rate": 9.978140637030963e-06,
+      "loss": 0.8602,
+      "step": 1179
+    },
+    {
+      "epoch": 0.06494578677967967,
+      "grad_norm": 0.8556679487228394,
+      "learning_rate": 9.978100130054505e-06,
+      "loss": 0.9149,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06500082558203533,
+      "grad_norm": 0.92474365234375,
+      "learning_rate": 9.978059585663979e-06,
+      "loss": 0.8608,
+      "step": 1181
+    },
+    {
+      "epoch": 0.065055864384391,
+      "grad_norm": 1.0170830488204956,
+      "learning_rate": 9.978019003859687e-06,
+      "loss": 0.9986,
+      "step": 1182
+    },
+    {
+      "epoch": 0.06511090318674666,
+      "grad_norm": 0.9405049681663513,
+      "learning_rate": 9.97797838464194e-06,
+      "loss": 0.9023,
+      "step": 1183
+    },
+    {
+      "epoch": 0.06516594198910232,
+      "grad_norm": 0.9351203441619873,
+      "learning_rate": 9.977937728011038e-06,
+      "loss": 0.8698,
+      "step": 1184
+    },
+    {
+      "epoch": 0.06522098079145798,
+      "grad_norm": 0.8620241284370422,
+      "learning_rate": 9.97789703396729e-06,
+      "loss": 0.9393,
+      "step": 1185
+    },
+    {
+      "epoch": 0.06527601959381364,
+      "grad_norm": 0.9440441131591797,
+      "learning_rate": 9.977856302511e-06,
+      "loss": 0.8249,
+      "step": 1186
+    },
+    {
+      "epoch": 0.0653310583961693,
+      "grad_norm": 0.8311079144477844,
+      "learning_rate": 9.977815533642474e-06,
+      "loss": 0.8614,
+      "step": 1187
+    },
+    {
+      "epoch": 0.06538609719852496,
+      "grad_norm": 0.8911672830581665,
+      "learning_rate": 9.977774727362018e-06,
+      "loss": 0.7909,
+      "step": 1188
+    },
+    {
+      "epoch": 0.06544113600088063,
+      "grad_norm": 0.9237088561058044,
+      "learning_rate": 9.97773388366994e-06,
+      "loss": 0.7116,
+      "step": 1189
+    },
+    {
+      "epoch": 0.06549617480323629,
+      "grad_norm": 1.1155747175216675,
+      "learning_rate": 9.977693002566549e-06,
+      "loss": 0.9248,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06555121360559195,
+      "grad_norm": 0.9386736750602722,
+      "learning_rate": 9.977652084052148e-06,
+      "loss": 0.8307,
+      "step": 1191
+    },
+    {
+      "epoch": 0.0656062524079476,
+      "grad_norm": 1.1666040420532227,
+      "learning_rate": 9.977611128127044e-06,
+      "loss": 0.9723,
+      "step": 1192
+    },
+    {
+      "epoch": 0.06566129121030326,
+      "grad_norm": 1.2366368770599365,
+      "learning_rate": 9.977570134791552e-06,
+      "loss": 0.8253,
+      "step": 1193
+    },
+    {
+      "epoch": 0.06571633001265892,
+      "grad_norm": 0.823443591594696,
+      "learning_rate": 9.977529104045971e-06,
+      "loss": 0.7472,
+      "step": 1194
+    },
+    {
+      "epoch": 0.06577136881501458,
+      "grad_norm": 0.9481683969497681,
+      "learning_rate": 9.977488035890617e-06,
+      "loss": 0.887,
+      "step": 1195
+    },
+    {
+      "epoch": 0.06582640761737024,
+      "grad_norm": 0.9120422005653381,
+      "learning_rate": 9.977446930325794e-06,
+      "loss": 0.867,
+      "step": 1196
+    },
+    {
+      "epoch": 0.0658814464197259,
+      "grad_norm": 0.8595587015151978,
+      "learning_rate": 9.977405787351811e-06,
+      "loss": 0.8532,
+      "step": 1197
+    },
+    {
+      "epoch": 0.06593648522208156,
+      "grad_norm": 0.8590419888496399,
+      "learning_rate": 9.97736460696898e-06,
+      "loss": 0.8998,
+      "step": 1198
+    },
+    {
+      "epoch": 0.06599152402443723,
+      "grad_norm": 0.9670939445495605,
+      "learning_rate": 9.977323389177609e-06,
+      "loss": 0.8964,
+      "step": 1199
+    },
+    {
+      "epoch": 0.06604656282679289,
+      "grad_norm": 0.8870261907577515,
+      "learning_rate": 9.977282133978006e-06,
+      "loss": 0.9542,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06610160162914855,
+      "grad_norm": 0.942294180393219,
+      "learning_rate": 9.977240841370484e-06,
+      "loss": 0.8681,
+      "step": 1201
+    },
+    {
+      "epoch": 0.06615664043150421,
+      "grad_norm": 0.9632517099380493,
+      "learning_rate": 9.977199511355353e-06,
+      "loss": 0.7327,
+      "step": 1202
+    },
+    {
+      "epoch": 0.06621167923385987,
+      "grad_norm": 4.8085479736328125,
+      "learning_rate": 9.97715814393292e-06,
+      "loss": 0.8528,
+      "step": 1203
+    },
+    {
+      "epoch": 0.06626671803621553,
+      "grad_norm": 0.9084093570709229,
+      "learning_rate": 9.977116739103503e-06,
+      "loss": 0.7836,
+      "step": 1204
+    },
+    {
+      "epoch": 0.0663217568385712,
+      "grad_norm": 0.8961902260780334,
+      "learning_rate": 9.977075296867406e-06,
+      "loss": 0.854,
+      "step": 1205
+    },
+    {
+      "epoch": 0.06637679564092686,
+      "grad_norm": 0.8727987408638,
+      "learning_rate": 9.977033817224945e-06,
+      "loss": 0.7931,
+      "step": 1206
+    },
+    {
+      "epoch": 0.06643183444328252,
+      "grad_norm": 0.8263267874717712,
+      "learning_rate": 9.976992300176428e-06,
+      "loss": 0.852,
+      "step": 1207
+    },
+    {
+      "epoch": 0.06648687324563818,
+      "grad_norm": 1.0499639511108398,
+      "learning_rate": 9.97695074572217e-06,
+      "loss": 1.0427,
+      "step": 1208
+    },
+    {
+      "epoch": 0.06654191204799384,
+      "grad_norm": 0.9337313771247864,
+      "learning_rate": 9.976909153862482e-06,
+      "loss": 0.8035,
+      "step": 1209
+    },
+    {
+      "epoch": 0.0665969508503495,
+      "grad_norm": 0.8795992732048035,
+      "learning_rate": 9.976867524597678e-06,
+      "loss": 0.9022,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06665198965270516,
+      "grad_norm": 0.9787294268608093,
+      "learning_rate": 9.976825857928069e-06,
+      "loss": 0.8259,
+      "step": 1211
+    },
+    {
+      "epoch": 0.06670702845506082,
+      "grad_norm": 0.8570082187652588,
+      "learning_rate": 9.976784153853969e-06,
+      "loss": 0.8567,
+      "step": 1212
+    },
+    {
+      "epoch": 0.06676206725741649,
+      "grad_norm": 1.0620380640029907,
+      "learning_rate": 9.976742412375694e-06,
+      "loss": 0.851,
+      "step": 1213
+    },
+    {
+      "epoch": 0.06681710605977213,
+      "grad_norm": 0.8545439839363098,
+      "learning_rate": 9.976700633493551e-06,
+      "loss": 0.8827,
+      "step": 1214
+    },
+    {
+      "epoch": 0.0668721448621278,
+      "grad_norm": 0.8543682098388672,
+      "learning_rate": 9.97665881720786e-06,
+      "loss": 0.8524,
+      "step": 1215
+    },
+    {
+      "epoch": 0.06692718366448346,
+      "grad_norm": 0.7748527526855469,
+      "learning_rate": 9.976616963518935e-06,
+      "loss": 0.7459,
+      "step": 1216
+    },
+    {
+      "epoch": 0.06698222246683912,
+      "grad_norm": 0.9876659512519836,
+      "learning_rate": 9.976575072427087e-06,
+      "loss": 0.8426,
+      "step": 1217
+    },
+    {
+      "epoch": 0.06703726126919478,
+      "grad_norm": 0.8763901591300964,
+      "learning_rate": 9.976533143932635e-06,
+      "loss": 0.8561,
+      "step": 1218
+    },
+    {
+      "epoch": 0.06709230007155044,
+      "grad_norm": 0.7816654443740845,
+      "learning_rate": 9.97649117803589e-06,
+      "loss": 0.8361,
+      "step": 1219
+    },
+    {
+      "epoch": 0.0671473388739061,
+      "grad_norm": 0.8659802675247192,
+      "learning_rate": 9.97644917473717e-06,
+      "loss": 0.897,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06720237767626176,
+      "grad_norm": 0.9180877208709717,
+      "learning_rate": 9.97640713403679e-06,
+      "loss": 0.9516,
+      "step": 1221
+    },
+    {
+      "epoch": 0.06725741647861742,
+      "grad_norm": 0.9624410271644592,
+      "learning_rate": 9.976365055935067e-06,
+      "loss": 0.9119,
+      "step": 1222
+    },
+    {
+      "epoch": 0.06731245528097309,
+      "grad_norm": 0.8291105031967163,
+      "learning_rate": 9.976322940432314e-06,
+      "loss": 0.788,
+      "step": 1223
+    },
+    {
+      "epoch": 0.06736749408332875,
+      "grad_norm": 0.9858983755111694,
+      "learning_rate": 9.976280787528854e-06,
+      "loss": 0.8794,
+      "step": 1224
+    },
+    {
+      "epoch": 0.06742253288568441,
+      "grad_norm": 0.8283948302268982,
+      "learning_rate": 9.976238597224996e-06,
+      "loss": 0.8571,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06747757168804007,
+      "grad_norm": 0.8585363626480103,
+      "learning_rate": 9.976196369521063e-06,
+      "loss": 0.9005,
+      "step": 1226
+    },
+    {
+      "epoch": 0.06753261049039573,
+      "grad_norm": 0.847882091999054,
+      "learning_rate": 9.976154104417369e-06,
+      "loss": 0.8058,
+      "step": 1227
+    },
+    {
+      "epoch": 0.06758764929275139,
+      "grad_norm": 0.9045611023902893,
+      "learning_rate": 9.976111801914232e-06,
+      "loss": 0.7864,
+      "step": 1228
+    },
+    {
+      "epoch": 0.06764268809510705,
+      "grad_norm": 0.805932879447937,
+      "learning_rate": 9.976069462011972e-06,
+      "loss": 0.8436,
+      "step": 1229
+    },
+    {
+      "epoch": 0.06769772689746271,
+      "grad_norm": 0.8809003233909607,
+      "learning_rate": 9.976027084710906e-06,
+      "loss": 0.7876,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06775276569981838,
+      "grad_norm": 0.8681740760803223,
+      "learning_rate": 9.975984670011352e-06,
+      "loss": 0.877,
+      "step": 1231
+    },
+    {
+      "epoch": 0.06780780450217404,
+      "grad_norm": 0.9909854531288147,
+      "learning_rate": 9.975942217913627e-06,
+      "loss": 0.8957,
+      "step": 1232
+    },
+    {
+      "epoch": 0.0678628433045297,
+      "grad_norm": 0.9213934540748596,
+      "learning_rate": 9.975899728418056e-06,
+      "loss": 0.8344,
+      "step": 1233
+    },
+    {
+      "epoch": 0.06791788210688536,
+      "grad_norm": 0.8289967179298401,
+      "learning_rate": 9.975857201524952e-06,
+      "loss": 0.876,
+      "step": 1234
+    },
+    {
+      "epoch": 0.06797292090924101,
+      "grad_norm": 0.891812264919281,
+      "learning_rate": 9.97581463723464e-06,
+      "loss": 0.8611,
+      "step": 1235
+    },
+    {
+      "epoch": 0.06802795971159667,
+      "grad_norm": 1.0301382541656494,
+      "learning_rate": 9.975772035547435e-06,
+      "loss": 0.8177,
+      "step": 1236
+    },
+    {
+      "epoch": 0.06808299851395233,
+      "grad_norm": 0.8380662798881531,
+      "learning_rate": 9.975729396463659e-06,
+      "loss": 0.8631,
+      "step": 1237
+    },
+    {
+      "epoch": 0.06813803731630799,
+      "grad_norm": 0.9226046204566956,
+      "learning_rate": 9.975686719983633e-06,
+      "loss": 0.8927,
+      "step": 1238
+    },
+    {
+      "epoch": 0.06819307611866365,
+      "grad_norm": 0.8917136192321777,
+      "learning_rate": 9.975644006107679e-06,
+      "loss": 0.9048,
+      "step": 1239
+    },
+    {
+      "epoch": 0.06824811492101931,
+      "grad_norm": 0.8559191226959229,
+      "learning_rate": 9.975601254836114e-06,
+      "loss": 0.8169,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06830315372337498,
+      "grad_norm": 0.9345341920852661,
+      "learning_rate": 9.975558466169263e-06,
+      "loss": 0.7929,
+      "step": 1241
+    },
+    {
+      "epoch": 0.06835819252573064,
+      "grad_norm": 0.9155850410461426,
+      "learning_rate": 9.975515640107447e-06,
+      "loss": 0.8825,
+      "step": 1242
+    },
+    {
+      "epoch": 0.0684132313280863,
+      "grad_norm": 0.899712860584259,
+      "learning_rate": 9.975472776650987e-06,
+      "loss": 0.825,
+      "step": 1243
+    },
+    {
+      "epoch": 0.06846827013044196,
+      "grad_norm": 0.8280880451202393,
+      "learning_rate": 9.975429875800206e-06,
+      "loss": 0.8539,
+      "step": 1244
+    },
+    {
+      "epoch": 0.06852330893279762,
+      "grad_norm": 0.9589636325836182,
+      "learning_rate": 9.975386937555426e-06,
+      "loss": 0.9465,
+      "step": 1245
+    },
+    {
+      "epoch": 0.06857834773515328,
+      "grad_norm": 1.1027253866195679,
+      "learning_rate": 9.97534396191697e-06,
+      "loss": 0.87,
+      "step": 1246
+    },
+    {
+      "epoch": 0.06863338653750894,
+      "grad_norm": 1.0510318279266357,
+      "learning_rate": 9.975300948885158e-06,
+      "loss": 0.8569,
+      "step": 1247
+    },
+    {
+      "epoch": 0.0686884253398646,
+      "grad_norm": 0.8897958397865295,
+      "learning_rate": 9.975257898460317e-06,
+      "loss": 0.8431,
+      "step": 1248
+    },
+    {
+      "epoch": 0.06874346414222027,
+      "grad_norm": 0.8827036619186401,
+      "learning_rate": 9.975214810642771e-06,
+      "loss": 0.922,
+      "step": 1249
+    },
+    {
+      "epoch": 0.06879850294457593,
+      "grad_norm": 0.8798324465751648,
+      "learning_rate": 9.97517168543284e-06,
+      "loss": 0.7837,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06885354174693159,
+      "grad_norm": 0.9053803086280823,
+      "learning_rate": 9.975128522830853e-06,
+      "loss": 0.82,
+      "step": 1251
+    },
+    {
+      "epoch": 0.06890858054928725,
+      "grad_norm": 0.8362607359886169,
+      "learning_rate": 9.975085322837129e-06,
+      "loss": 0.7684,
+      "step": 1252
+    },
+    {
+      "epoch": 0.06896361935164291,
+      "grad_norm": 0.8898602724075317,
+      "learning_rate": 9.975042085451997e-06,
+      "loss": 0.8205,
+      "step": 1253
+    },
+    {
+      "epoch": 0.06901865815399857,
+      "grad_norm": 0.9210274815559387,
+      "learning_rate": 9.97499881067578e-06,
+      "loss": 0.8364,
+      "step": 1254
+    },
+    {
+      "epoch": 0.06907369695635424,
+      "grad_norm": 1.0881952047348022,
+      "learning_rate": 9.974955498508804e-06,
+      "loss": 0.8234,
+      "step": 1255
+    },
+    {
+      "epoch": 0.0691287357587099,
+      "grad_norm": 0.8875024914741516,
+      "learning_rate": 9.974912148951394e-06,
+      "loss": 0.7974,
+      "step": 1256
+    },
+    {
+      "epoch": 0.06918377456106554,
+      "grad_norm": 0.9065666794776917,
+      "learning_rate": 9.974868762003876e-06,
+      "loss": 0.7721,
+      "step": 1257
+    },
+    {
+      "epoch": 0.0692388133634212,
+      "grad_norm": 0.8904553651809692,
+      "learning_rate": 9.974825337666576e-06,
+      "loss": 0.8551,
+      "step": 1258
+    },
+    {
+      "epoch": 0.06929385216577687,
+      "grad_norm": 0.8586102724075317,
+      "learning_rate": 9.974781875939821e-06,
+      "loss": 0.8666,
+      "step": 1259
+    },
+    {
+      "epoch": 0.06934889096813253,
+      "grad_norm": 0.9103402495384216,
+      "learning_rate": 9.974738376823935e-06,
+      "loss": 0.8361,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06940392977048819,
+      "grad_norm": 0.8657701015472412,
+      "learning_rate": 9.974694840319249e-06,
+      "loss": 0.8217,
+      "step": 1261
+    },
+    {
+      "epoch": 0.06945896857284385,
+      "grad_norm": 0.865703821182251,
+      "learning_rate": 9.974651266426088e-06,
+      "loss": 0.8751,
+      "step": 1262
+    },
+    {
+      "epoch": 0.06951400737519951,
+      "grad_norm": 0.8932577967643738,
+      "learning_rate": 9.974607655144779e-06,
+      "loss": 0.8709,
+      "step": 1263
+    },
+    {
+      "epoch": 0.06956904617755517,
+      "grad_norm": 0.8417405486106873,
+      "learning_rate": 9.97456400647565e-06,
+      "loss": 0.8104,
+      "step": 1264
+    },
+    {
+      "epoch": 0.06962408497991084,
+      "grad_norm": 0.8578035235404968,
+      "learning_rate": 9.974520320419032e-06,
+      "loss": 0.9173,
+      "step": 1265
+    },
+    {
+      "epoch": 0.0696791237822665,
+      "grad_norm": 0.957539439201355,
+      "learning_rate": 9.974476596975249e-06,
+      "loss": 0.8955,
+      "step": 1266
+    },
+    {
+      "epoch": 0.06973416258462216,
+      "grad_norm": 0.851222038269043,
+      "learning_rate": 9.974432836144632e-06,
+      "loss": 0.8696,
+      "step": 1267
+    },
+    {
+      "epoch": 0.06978920138697782,
+      "grad_norm": 0.8178789615631104,
+      "learning_rate": 9.974389037927508e-06,
+      "loss": 0.7921,
+      "step": 1268
+    },
+    {
+      "epoch": 0.06984424018933348,
+      "grad_norm": 0.954091489315033,
+      "learning_rate": 9.97434520232421e-06,
+      "loss": 0.9362,
+      "step": 1269
+    },
+    {
+      "epoch": 0.06989927899168914,
+      "grad_norm": 0.8525053858757019,
+      "learning_rate": 9.974301329335063e-06,
+      "loss": 0.7996,
+      "step": 1270
+    },
+    {
+      "epoch": 0.0699543177940448,
+      "grad_norm": 0.9340476393699646,
+      "learning_rate": 9.9742574189604e-06,
+      "loss": 0.9091,
+      "step": 1271
+    },
+    {
+      "epoch": 0.07000935659640047,
+      "grad_norm": 0.7946187257766724,
+      "learning_rate": 9.974213471200548e-06,
+      "loss": 0.874,
+      "step": 1272
+    },
+    {
+      "epoch": 0.07006439539875613,
+      "grad_norm": 0.8048381209373474,
+      "learning_rate": 9.97416948605584e-06,
+      "loss": 0.8557,
+      "step": 1273
+    },
+    {
+      "epoch": 0.07011943420111179,
+      "grad_norm": 0.9849064946174622,
+      "learning_rate": 9.974125463526607e-06,
+      "loss": 0.8154,
+      "step": 1274
+    },
+    {
+      "epoch": 0.07017447300346745,
+      "grad_norm": 0.9030239582061768,
+      "learning_rate": 9.974081403613178e-06,
+      "loss": 0.9411,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07022951180582311,
+      "grad_norm": 0.8869300484657288,
+      "learning_rate": 9.974037306315882e-06,
+      "loss": 0.8978,
+      "step": 1276
+    },
+    {
+      "epoch": 0.07028455060817877,
+      "grad_norm": 0.8558536767959595,
+      "learning_rate": 9.973993171635057e-06,
+      "loss": 0.8937,
+      "step": 1277
+    },
+    {
+      "epoch": 0.07033958941053442,
+      "grad_norm": 0.9005453586578369,
+      "learning_rate": 9.973948999571029e-06,
+      "loss": 0.9336,
+      "step": 1278
+    },
+    {
+      "epoch": 0.07039462821289008,
+      "grad_norm": 0.8489978909492493,
+      "learning_rate": 9.973904790124131e-06,
+      "loss": 0.8267,
+      "step": 1279
+    },
+    {
+      "epoch": 0.07044966701524574,
+      "grad_norm": 0.8295948505401611,
+      "learning_rate": 9.973860543294696e-06,
+      "loss": 0.8478,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0705047058176014,
+      "grad_norm": 0.8111379742622375,
+      "learning_rate": 9.973816259083058e-06,
+      "loss": 0.8333,
+      "step": 1281
+    },
+    {
+      "epoch": 0.07055974461995707,
+      "grad_norm": 0.9380189776420593,
+      "learning_rate": 9.973771937489547e-06,
+      "loss": 0.9718,
+      "step": 1282
+    },
+    {
+      "epoch": 0.07061478342231273,
+      "grad_norm": 1.251194953918457,
+      "learning_rate": 9.973727578514499e-06,
+      "loss": 0.9531,
+      "step": 1283
+    },
+    {
+      "epoch": 0.07066982222466839,
+      "grad_norm": 0.9897224307060242,
+      "learning_rate": 9.973683182158243e-06,
+      "loss": 0.7853,
+      "step": 1284
+    },
+    {
+      "epoch": 0.07072486102702405,
+      "grad_norm": 0.8409335017204285,
+      "learning_rate": 9.973638748421119e-06,
+      "loss": 0.7692,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07077989982937971,
+      "grad_norm": 0.9019681215286255,
+      "learning_rate": 9.973594277303456e-06,
+      "loss": 0.8135,
+      "step": 1286
+    },
+    {
+      "epoch": 0.07083493863173537,
+      "grad_norm": 0.9236096739768982,
+      "learning_rate": 9.973549768805588e-06,
+      "loss": 0.9304,
+      "step": 1287
+    },
+    {
+      "epoch": 0.07088997743409103,
+      "grad_norm": 0.9244743585586548,
+      "learning_rate": 9.973505222927854e-06,
+      "loss": 0.9056,
+      "step": 1288
+    },
+    {
+      "epoch": 0.0709450162364467,
+      "grad_norm": 1.3418753147125244,
+      "learning_rate": 9.973460639670585e-06,
+      "loss": 0.8419,
+      "step": 1289
+    },
+    {
+      "epoch": 0.07100005503880236,
+      "grad_norm": 0.8715767860412598,
+      "learning_rate": 9.973416019034117e-06,
+      "loss": 0.9704,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07105509384115802,
+      "grad_norm": 0.9609012007713318,
+      "learning_rate": 9.973371361018787e-06,
+      "loss": 0.8807,
+      "step": 1291
+    },
+    {
+      "epoch": 0.07111013264351368,
+      "grad_norm": 0.8085873126983643,
+      "learning_rate": 9.973326665624927e-06,
+      "loss": 0.7947,
+      "step": 1292
+    },
+    {
+      "epoch": 0.07116517144586934,
+      "grad_norm": 0.919280469417572,
+      "learning_rate": 9.973281932852877e-06,
+      "loss": 0.9743,
+      "step": 1293
+    },
+    {
+      "epoch": 0.071220210248225,
+      "grad_norm": 1.0651074647903442,
+      "learning_rate": 9.973237162702968e-06,
+      "loss": 0.7164,
+      "step": 1294
+    },
+    {
+      "epoch": 0.07127524905058066,
+      "grad_norm": 0.987251341342926,
+      "learning_rate": 9.973192355175542e-06,
+      "loss": 0.9286,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07133028785293632,
+      "grad_norm": 1.5507274866104126,
+      "learning_rate": 9.973147510270935e-06,
+      "loss": 0.9733,
+      "step": 1296
+    },
+    {
+      "epoch": 0.07138532665529199,
+      "grad_norm": 0.8439416885375977,
+      "learning_rate": 9.97310262798948e-06,
+      "loss": 0.7462,
+      "step": 1297
+    },
+    {
+      "epoch": 0.07144036545764765,
+      "grad_norm": 0.9604889750480652,
+      "learning_rate": 9.973057708331519e-06,
+      "loss": 1.0006,
+      "step": 1298
+    },
+    {
+      "epoch": 0.07149540426000331,
+      "grad_norm": 0.8568960428237915,
+      "learning_rate": 9.973012751297386e-06,
+      "loss": 0.878,
+      "step": 1299
+    },
+    {
+      "epoch": 0.07155044306235896,
+      "grad_norm": 0.8169522285461426,
+      "learning_rate": 9.972967756887419e-06,
+      "loss": 0.8241,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07160548186471462,
+      "grad_norm": 0.875738799571991,
+      "learning_rate": 9.97292272510196e-06,
+      "loss": 0.854,
+      "step": 1301
+    },
+    {
+      "epoch": 0.07166052066707028,
+      "grad_norm": 0.7877739071846008,
+      "learning_rate": 9.972877655941345e-06,
+      "loss": 0.779,
+      "step": 1302
+    },
+    {
+      "epoch": 0.07171555946942594,
+      "grad_norm": 0.8148574829101562,
+      "learning_rate": 9.972832549405912e-06,
+      "loss": 0.6965,
+      "step": 1303
+    },
+    {
+      "epoch": 0.0717705982717816,
+      "grad_norm": 0.936720609664917,
+      "learning_rate": 9.972787405495998e-06,
+      "loss": 0.798,
+      "step": 1304
+    },
+    {
+      "epoch": 0.07182563707413726,
+      "grad_norm": 0.8932886123657227,
+      "learning_rate": 9.972742224211949e-06,
+      "loss": 0.9196,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07188067587649292,
+      "grad_norm": 0.899246871471405,
+      "learning_rate": 9.972697005554099e-06,
+      "loss": 0.8081,
+      "step": 1306
+    },
+    {
+      "epoch": 0.07193571467884859,
+      "grad_norm": 0.8789899349212646,
+      "learning_rate": 9.972651749522788e-06,
+      "loss": 0.89,
+      "step": 1307
+    },
+    {
+      "epoch": 0.07199075348120425,
+      "grad_norm": 1.2412173748016357,
+      "learning_rate": 9.97260645611836e-06,
+      "loss": 0.9866,
+      "step": 1308
+    },
+    {
+      "epoch": 0.07204579228355991,
+      "grad_norm": 0.8655833005905151,
+      "learning_rate": 9.972561125341152e-06,
+      "loss": 0.8144,
+      "step": 1309
+    },
+    {
+      "epoch": 0.07210083108591557,
+      "grad_norm": 0.8705299496650696,
+      "learning_rate": 9.972515757191506e-06,
+      "loss": 0.8431,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07215586988827123,
+      "grad_norm": 0.8813188672065735,
+      "learning_rate": 9.972470351669761e-06,
+      "loss": 0.859,
+      "step": 1311
+    },
+    {
+      "epoch": 0.0722109086906269,
+      "grad_norm": 2.043627977371216,
+      "learning_rate": 9.972424908776262e-06,
+      "loss": 0.9886,
+      "step": 1312
+    },
+    {
+      "epoch": 0.07226594749298255,
+      "grad_norm": 0.9167500734329224,
+      "learning_rate": 9.972379428511348e-06,
+      "loss": 0.7203,
+      "step": 1313
+    },
+    {
+      "epoch": 0.07232098629533822,
+      "grad_norm": 1.3145136833190918,
+      "learning_rate": 9.972333910875358e-06,
+      "loss": 0.9325,
+      "step": 1314
+    },
+    {
+      "epoch": 0.07237602509769388,
+      "grad_norm": 0.834710419178009,
+      "learning_rate": 9.972288355868641e-06,
+      "loss": 0.9361,
+      "step": 1315
+    },
+    {
+      "epoch": 0.07243106390004954,
+      "grad_norm": 0.9039230942726135,
+      "learning_rate": 9.972242763491535e-06,
+      "loss": 0.8027,
+      "step": 1316
+    },
+    {
+      "epoch": 0.0724861027024052,
+      "grad_norm": 0.8911495208740234,
+      "learning_rate": 9.972197133744384e-06,
+      "loss": 0.951,
+      "step": 1317
+    },
+    {
+      "epoch": 0.07254114150476086,
+      "grad_norm": 1.0752439498901367,
+      "learning_rate": 9.972151466627529e-06,
+      "loss": 0.8421,
+      "step": 1318
+    },
+    {
+      "epoch": 0.07259618030711652,
+      "grad_norm": 0.926135778427124,
+      "learning_rate": 9.972105762141314e-06,
+      "loss": 0.8901,
+      "step": 1319
+    },
+    {
+      "epoch": 0.07265121910947218,
+      "grad_norm": 0.8166295289993286,
+      "learning_rate": 9.972060020286085e-06,
+      "loss": 0.7845,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07270625791182783,
+      "grad_norm": 1.0000934600830078,
+      "learning_rate": 9.972014241062182e-06,
+      "loss": 0.8383,
+      "step": 1321
+    },
+    {
+      "epoch": 0.0727612967141835,
+      "grad_norm": 1.2617899179458618,
+      "learning_rate": 9.971968424469951e-06,
+      "loss": 0.9826,
+      "step": 1322
+    },
+    {
+      "epoch": 0.07281633551653915,
+      "grad_norm": 0.8451040983200073,
+      "learning_rate": 9.971922570509738e-06,
+      "loss": 0.8262,
+      "step": 1323
+    },
+    {
+      "epoch": 0.07287137431889482,
+      "grad_norm": 0.8101939558982849,
+      "learning_rate": 9.971876679181884e-06,
+      "loss": 0.6904,
+      "step": 1324
+    },
+    {
+      "epoch": 0.07292641312125048,
+      "grad_norm": 0.8805514574050903,
+      "learning_rate": 9.971830750486736e-06,
+      "loss": 0.8491,
+      "step": 1325
+    },
+    {
+      "epoch": 0.07298145192360614,
+      "grad_norm": 0.8236901164054871,
+      "learning_rate": 9.97178478442464e-06,
+      "loss": 0.8462,
+      "step": 1326
+    },
+    {
+      "epoch": 0.0730364907259618,
+      "grad_norm": 0.9183042645454407,
+      "learning_rate": 9.971738780995938e-06,
+      "loss": 0.7577,
+      "step": 1327
+    },
+    {
+      "epoch": 0.07309152952831746,
+      "grad_norm": 0.8425934314727783,
+      "learning_rate": 9.971692740200982e-06,
+      "loss": 0.8462,
+      "step": 1328
+    },
+    {
+      "epoch": 0.07314656833067312,
+      "grad_norm": 0.9114993214607239,
+      "learning_rate": 9.971646662040112e-06,
+      "loss": 0.9132,
+      "step": 1329
+    },
+    {
+      "epoch": 0.07320160713302878,
+      "grad_norm": 0.8516649603843689,
+      "learning_rate": 9.971600546513675e-06,
+      "loss": 0.8819,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07325664593538445,
+      "grad_norm": 1.0859558582305908,
+      "learning_rate": 9.971554393622023e-06,
+      "loss": 0.9929,
+      "step": 1331
+    },
+    {
+      "epoch": 0.07331168473774011,
+      "grad_norm": 0.8906900882720947,
+      "learning_rate": 9.971508203365497e-06,
+      "loss": 0.9166,
+      "step": 1332
+    },
+    {
+      "epoch": 0.07336672354009577,
+      "grad_norm": 0.8931803703308105,
+      "learning_rate": 9.971461975744445e-06,
+      "loss": 0.864,
+      "step": 1333
+    },
+    {
+      "epoch": 0.07342176234245143,
+      "grad_norm": 0.8404982686042786,
+      "learning_rate": 9.971415710759216e-06,
+      "loss": 0.8609,
+      "step": 1334
+    },
+    {
+      "epoch": 0.07347680114480709,
+      "grad_norm": 0.8016490340232849,
+      "learning_rate": 9.971369408410157e-06,
+      "loss": 0.7694,
+      "step": 1335
+    },
+    {
+      "epoch": 0.07353183994716275,
+      "grad_norm": 0.7700600028038025,
+      "learning_rate": 9.971323068697618e-06,
+      "loss": 0.7875,
+      "step": 1336
+    },
+    {
+      "epoch": 0.07358687874951841,
+      "grad_norm": 0.8679799437522888,
+      "learning_rate": 9.971276691621946e-06,
+      "loss": 0.8409,
+      "step": 1337
+    },
+    {
+      "epoch": 0.07364191755187408,
+      "grad_norm": 0.8329173922538757,
+      "learning_rate": 9.971230277183486e-06,
+      "loss": 0.8707,
+      "step": 1338
+    },
+    {
+      "epoch": 0.07369695635422974,
+      "grad_norm": 0.8790140151977539,
+      "learning_rate": 9.97118382538259e-06,
+      "loss": 0.7631,
+      "step": 1339
+    },
+    {
+      "epoch": 0.0737519951565854,
+      "grad_norm": 1.1895341873168945,
+      "learning_rate": 9.97113733621961e-06,
+      "loss": 0.8555,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07380703395894106,
+      "grad_norm": 0.8531593680381775,
+      "learning_rate": 9.97109080969489e-06,
+      "loss": 0.7192,
+      "step": 1341
+    },
+    {
+      "epoch": 0.07386207276129672,
+      "grad_norm": 1.0388946533203125,
+      "learning_rate": 9.971044245808784e-06,
+      "loss": 0.8182,
+      "step": 1342
+    },
+    {
+      "epoch": 0.07391711156365237,
+      "grad_norm": 0.8858556747436523,
+      "learning_rate": 9.970997644561639e-06,
+      "loss": 0.7981,
+      "step": 1343
+    },
+    {
+      "epoch": 0.07397215036600803,
+      "grad_norm": 0.8710204362869263,
+      "learning_rate": 9.970951005953807e-06,
+      "loss": 0.7667,
+      "step": 1344
+    },
+    {
+      "epoch": 0.07402718916836369,
+      "grad_norm": 0.9788708090782166,
+      "learning_rate": 9.970904329985638e-06,
+      "loss": 0.9693,
+      "step": 1345
+    },
+    {
+      "epoch": 0.07408222797071935,
+      "grad_norm": 0.7805914878845215,
+      "learning_rate": 9.970857616657482e-06,
+      "loss": 0.6683,
+      "step": 1346
+    },
+    {
+      "epoch": 0.07413726677307501,
+      "grad_norm": 0.9977933168411255,
+      "learning_rate": 9.97081086596969e-06,
+      "loss": 0.8288,
+      "step": 1347
+    },
+    {
+      "epoch": 0.07419230557543068,
+      "grad_norm": 0.829115629196167,
+      "learning_rate": 9.970764077922617e-06,
+      "loss": 0.8361,
+      "step": 1348
+    },
+    {
+      "epoch": 0.07424734437778634,
+      "grad_norm": 1.226120114326477,
+      "learning_rate": 9.97071725251661e-06,
+      "loss": 1.0008,
+      "step": 1349
+    },
+    {
+      "epoch": 0.074302383180142,
+      "grad_norm": 0.8997750878334045,
+      "learning_rate": 9.970670389752021e-06,
+      "loss": 0.8048,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07435742198249766,
+      "grad_norm": 1.0885238647460938,
+      "learning_rate": 9.970623489629205e-06,
+      "loss": 0.9202,
+      "step": 1351
+    },
+    {
+      "epoch": 0.07441246078485332,
+      "grad_norm": 0.8736100792884827,
+      "learning_rate": 9.970576552148515e-06,
+      "loss": 0.8515,
+      "step": 1352
+    },
+    {
+      "epoch": 0.07446749958720898,
+      "grad_norm": 0.9211294651031494,
+      "learning_rate": 9.970529577310301e-06,
+      "loss": 0.9389,
+      "step": 1353
+    },
+    {
+      "epoch": 0.07452253838956464,
+      "grad_norm": 0.9334765672683716,
+      "learning_rate": 9.970482565114917e-06,
+      "loss": 0.8165,
+      "step": 1354
+    },
+    {
+      "epoch": 0.0745775771919203,
+      "grad_norm": 0.8307162523269653,
+      "learning_rate": 9.970435515562717e-06,
+      "loss": 0.7829,
+      "step": 1355
+    },
+    {
+      "epoch": 0.07463261599427597,
+      "grad_norm": 0.987634003162384,
+      "learning_rate": 9.970388428654055e-06,
+      "loss": 0.848,
+      "step": 1356
+    },
+    {
+      "epoch": 0.07468765479663163,
+      "grad_norm": 1.094752311706543,
+      "learning_rate": 9.970341304389281e-06,
+      "loss": 1.003,
+      "step": 1357
+    },
+    {
+      "epoch": 0.07474269359898729,
+      "grad_norm": 0.9865909814834595,
+      "learning_rate": 9.970294142768755e-06,
+      "loss": 0.9116,
+      "step": 1358
+    },
+    {
+      "epoch": 0.07479773240134295,
+      "grad_norm": 0.8404149413108826,
+      "learning_rate": 9.970246943792828e-06,
+      "loss": 0.8699,
+      "step": 1359
+    },
+    {
+      "epoch": 0.07485277120369861,
+      "grad_norm": 0.9602416753768921,
+      "learning_rate": 9.970199707461855e-06,
+      "loss": 0.8166,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07490781000605427,
+      "grad_norm": 0.9748693704605103,
+      "learning_rate": 9.970152433776193e-06,
+      "loss": 0.8767,
+      "step": 1361
+    },
+    {
+      "epoch": 0.07496284880840993,
+      "grad_norm": 0.8721657991409302,
+      "learning_rate": 9.970105122736194e-06,
+      "loss": 0.8825,
+      "step": 1362
+    },
+    {
+      "epoch": 0.0750178876107656,
+      "grad_norm": 0.8683610558509827,
+      "learning_rate": 9.970057774342215e-06,
+      "loss": 0.7873,
+      "step": 1363
+    },
+    {
+      "epoch": 0.07507292641312124,
+      "grad_norm": 0.856396496295929,
+      "learning_rate": 9.970010388594613e-06,
+      "loss": 0.8505,
+      "step": 1364
+    },
+    {
+      "epoch": 0.0751279652154769,
+      "grad_norm": 1.0709880590438843,
+      "learning_rate": 9.969962965493744e-06,
+      "loss": 0.9519,
+      "step": 1365
+    },
+    {
+      "epoch": 0.07518300401783257,
+      "grad_norm": 0.8839450478553772,
+      "learning_rate": 9.969915505039963e-06,
+      "loss": 0.8041,
+      "step": 1366
+    },
+    {
+      "epoch": 0.07523804282018823,
+      "grad_norm": 0.89545738697052,
+      "learning_rate": 9.969868007233627e-06,
+      "loss": 0.8713,
+      "step": 1367
+    },
+    {
+      "epoch": 0.07529308162254389,
+      "grad_norm": 0.9870849251747131,
+      "learning_rate": 9.969820472075094e-06,
+      "loss": 0.8655,
+      "step": 1368
+    },
+    {
+      "epoch": 0.07534812042489955,
+      "grad_norm": 1.3123797178268433,
+      "learning_rate": 9.96977289956472e-06,
+      "loss": 1.0425,
+      "step": 1369
+    },
+    {
+      "epoch": 0.07540315922725521,
+      "grad_norm": 0.8538400530815125,
+      "learning_rate": 9.969725289702865e-06,
+      "loss": 0.7052,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07545819802961087,
+      "grad_norm": 0.933397114276886,
+      "learning_rate": 9.969677642489884e-06,
+      "loss": 0.9819,
+      "step": 1371
+    },
+    {
+      "epoch": 0.07551323683196653,
+      "grad_norm": 0.8428112268447876,
+      "learning_rate": 9.969629957926134e-06,
+      "loss": 0.7313,
+      "step": 1372
+    },
+    {
+      "epoch": 0.0755682756343222,
+      "grad_norm": 0.9023239612579346,
+      "learning_rate": 9.96958223601198e-06,
+      "loss": 0.8297,
+      "step": 1373
+    },
+    {
+      "epoch": 0.07562331443667786,
+      "grad_norm": 0.8971324563026428,
+      "learning_rate": 9.969534476747771e-06,
+      "loss": 0.8832,
+      "step": 1374
+    },
+    {
+      "epoch": 0.07567835323903352,
+      "grad_norm": 0.8709388375282288,
+      "learning_rate": 9.969486680133874e-06,
+      "loss": 0.743,
+      "step": 1375
+    },
+    {
+      "epoch": 0.07573339204138918,
+      "grad_norm": 0.9094591736793518,
+      "learning_rate": 9.969438846170644e-06,
+      "loss": 0.8294,
+      "step": 1376
+    },
+    {
+      "epoch": 0.07578843084374484,
+      "grad_norm": 1.0753988027572632,
+      "learning_rate": 9.969390974858444e-06,
+      "loss": 0.7479,
+      "step": 1377
+    },
+    {
+      "epoch": 0.0758434696461005,
+      "grad_norm": 0.933775007724762,
+      "learning_rate": 9.96934306619763e-06,
+      "loss": 0.8235,
+      "step": 1378
+    },
+    {
+      "epoch": 0.07589850844845616,
+      "grad_norm": 0.8419735431671143,
+      "learning_rate": 9.969295120188565e-06,
+      "loss": 0.8103,
+      "step": 1379
+    },
+    {
+      "epoch": 0.07595354725081183,
+      "grad_norm": 0.8912790417671204,
+      "learning_rate": 9.969247136831606e-06,
+      "loss": 0.911,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07600858605316749,
+      "grad_norm": 0.8780983090400696,
+      "learning_rate": 9.969199116127118e-06,
+      "loss": 0.8619,
+      "step": 1381
+    },
+    {
+      "epoch": 0.07606362485552315,
+      "grad_norm": 0.8503809571266174,
+      "learning_rate": 9.969151058075459e-06,
+      "loss": 0.8093,
+      "step": 1382
+    },
+    {
+      "epoch": 0.07611866365787881,
+      "grad_norm": 0.8633087277412415,
+      "learning_rate": 9.96910296267699e-06,
+      "loss": 0.7524,
+      "step": 1383
+    },
+    {
+      "epoch": 0.07617370246023447,
+      "grad_norm": 1.1203595399856567,
+      "learning_rate": 9.969054829932074e-06,
+      "loss": 0.945,
+      "step": 1384
+    },
+    {
+      "epoch": 0.07622874126259013,
+      "grad_norm": 0.8766878843307495,
+      "learning_rate": 9.969006659841072e-06,
+      "loss": 0.7537,
+      "step": 1385
+    },
+    {
+      "epoch": 0.07628378006494578,
+      "grad_norm": 0.9795958399772644,
+      "learning_rate": 9.968958452404345e-06,
+      "loss": 0.7963,
+      "step": 1386
+    },
+    {
+      "epoch": 0.07633881886730144,
+      "grad_norm": 0.9117506146430969,
+      "learning_rate": 9.968910207622257e-06,
+      "loss": 0.9469,
+      "step": 1387
+    },
+    {
+      "epoch": 0.0763938576696571,
+      "grad_norm": 0.9731466770172119,
+      "learning_rate": 9.96886192549517e-06,
+      "loss": 0.9536,
+      "step": 1388
+    },
+    {
+      "epoch": 0.07644889647201276,
+      "grad_norm": 0.8923571109771729,
+      "learning_rate": 9.968813606023446e-06,
+      "loss": 0.8362,
+      "step": 1389
+    },
+    {
+      "epoch": 0.07650393527436843,
+      "grad_norm": 0.8819600343704224,
+      "learning_rate": 9.96876524920745e-06,
+      "loss": 0.6938,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07655897407672409,
+      "grad_norm": 0.9629887342453003,
+      "learning_rate": 9.968716855047545e-06,
+      "loss": 0.9104,
+      "step": 1391
+    },
+    {
+      "epoch": 0.07661401287907975,
+      "grad_norm": 0.992770254611969,
+      "learning_rate": 9.968668423544093e-06,
+      "loss": 0.944,
+      "step": 1392
+    },
+    {
+      "epoch": 0.07666905168143541,
+      "grad_norm": 0.8578491806983948,
+      "learning_rate": 9.96861995469746e-06,
+      "loss": 0.898,
+      "step": 1393
+    },
+    {
+      "epoch": 0.07672409048379107,
+      "grad_norm": 1.1169229745864868,
+      "learning_rate": 9.968571448508008e-06,
+      "loss": 0.8324,
+      "step": 1394
+    },
+    {
+      "epoch": 0.07677912928614673,
+      "grad_norm": 0.9600160121917725,
+      "learning_rate": 9.968522904976106e-06,
+      "loss": 0.9519,
+      "step": 1395
+    },
+    {
+      "epoch": 0.0768341680885024,
+      "grad_norm": 0.8271373510360718,
+      "learning_rate": 9.968474324102112e-06,
+      "loss": 0.8576,
+      "step": 1396
+    },
+    {
+      "epoch": 0.07688920689085806,
+      "grad_norm": 0.9437325596809387,
+      "learning_rate": 9.968425705886397e-06,
+      "loss": 0.9201,
+      "step": 1397
+    },
+    {
+      "epoch": 0.07694424569321372,
+      "grad_norm": 0.8679039478302002,
+      "learning_rate": 9.968377050329325e-06,
+      "loss": 0.8893,
+      "step": 1398
+    },
+    {
+      "epoch": 0.07699928449556938,
+      "grad_norm": 1.0178717374801636,
+      "learning_rate": 9.96832835743126e-06,
+      "loss": 0.9718,
+      "step": 1399
+    },
+    {
+      "epoch": 0.07705432329792504,
+      "grad_norm": 0.8354432582855225,
+      "learning_rate": 9.96827962719257e-06,
+      "loss": 0.83,
+      "step": 1400
+    },
+    {
+      "epoch": 0.0771093621002807,
+      "grad_norm": 1.2244631052017212,
+      "learning_rate": 9.968230859613619e-06,
+      "loss": 0.907,
+      "step": 1401
+    },
+    {
+      "epoch": 0.07716440090263636,
+      "grad_norm": 0.9099625945091248,
+      "learning_rate": 9.968182054694775e-06,
+      "loss": 0.809,
+      "step": 1402
+    },
+    {
+      "epoch": 0.07721943970499202,
+      "grad_norm": 0.8591424226760864,
+      "learning_rate": 9.968133212436404e-06,
+      "loss": 0.8869,
+      "step": 1403
+    },
+    {
+      "epoch": 0.07727447850734769,
+      "grad_norm": 1.068003535270691,
+      "learning_rate": 9.968084332838876e-06,
+      "loss": 0.8747,
+      "step": 1404
+    },
+    {
+      "epoch": 0.07732951730970335,
+      "grad_norm": 0.8503691554069519,
+      "learning_rate": 9.968035415902555e-06,
+      "loss": 0.7478,
+      "step": 1405
+    },
+    {
+      "epoch": 0.07738455611205901,
+      "grad_norm": 0.9209537506103516,
+      "learning_rate": 9.967986461627808e-06,
+      "loss": 0.9052,
+      "step": 1406
+    },
+    {
+      "epoch": 0.07743959491441466,
+      "grad_norm": 0.8447962999343872,
+      "learning_rate": 9.967937470015006e-06,
+      "loss": 0.7897,
+      "step": 1407
+    },
+    {
+      "epoch": 0.07749463371677032,
+      "grad_norm": 0.8731846809387207,
+      "learning_rate": 9.967888441064515e-06,
+      "loss": 0.837,
+      "step": 1408
+    },
+    {
+      "epoch": 0.07754967251912598,
+      "grad_norm": 0.9810444712638855,
+      "learning_rate": 9.967839374776705e-06,
+      "loss": 0.8236,
+      "step": 1409
+    },
+    {
+      "epoch": 0.07760471132148164,
+      "grad_norm": 0.8283190131187439,
+      "learning_rate": 9.967790271151944e-06,
+      "loss": 0.8443,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0776597501238373,
+      "grad_norm": 0.7999932765960693,
+      "learning_rate": 9.9677411301906e-06,
+      "loss": 0.7945,
+      "step": 1411
+    },
+    {
+      "epoch": 0.07771478892619296,
+      "grad_norm": 0.9435983300209045,
+      "learning_rate": 9.967691951893044e-06,
+      "loss": 0.9745,
+      "step": 1412
+    },
+    {
+      "epoch": 0.07776982772854862,
+      "grad_norm": 0.8885984420776367,
+      "learning_rate": 9.967642736259646e-06,
+      "loss": 0.9163,
+      "step": 1413
+    },
+    {
+      "epoch": 0.07782486653090429,
+      "grad_norm": 0.993928074836731,
+      "learning_rate": 9.967593483290776e-06,
+      "loss": 0.7797,
+      "step": 1414
+    },
+    {
+      "epoch": 0.07787990533325995,
+      "grad_norm": 1.058830976486206,
+      "learning_rate": 9.9675441929868e-06,
+      "loss": 0.8671,
+      "step": 1415
+    },
+    {
+      "epoch": 0.07793494413561561,
+      "grad_norm": 1.0469766855239868,
+      "learning_rate": 9.967494865348093e-06,
+      "loss": 0.8671,
+      "step": 1416
+    },
+    {
+      "epoch": 0.07798998293797127,
+      "grad_norm": 0.902729868888855,
+      "learning_rate": 9.967445500375025e-06,
+      "loss": 0.8748,
+      "step": 1417
+    },
+    {
+      "epoch": 0.07804502174032693,
+      "grad_norm": 0.90755295753479,
+      "learning_rate": 9.967396098067965e-06,
+      "loss": 0.8279,
+      "step": 1418
+    },
+    {
+      "epoch": 0.07810006054268259,
+      "grad_norm": 0.8822374939918518,
+      "learning_rate": 9.967346658427287e-06,
+      "loss": 0.9386,
+      "step": 1419
+    },
+    {
+      "epoch": 0.07815509934503825,
+      "grad_norm": 0.9201469421386719,
+      "learning_rate": 9.96729718145336e-06,
+      "loss": 0.8684,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07821013814739392,
+      "grad_norm": 0.9451109766960144,
+      "learning_rate": 9.967247667146558e-06,
+      "loss": 0.7854,
+      "step": 1421
+    },
+    {
+      "epoch": 0.07826517694974958,
+      "grad_norm": 0.9146197438240051,
+      "learning_rate": 9.96719811550725e-06,
+      "loss": 0.8496,
+      "step": 1422
+    },
+    {
+      "epoch": 0.07832021575210524,
+      "grad_norm": 0.9771224856376648,
+      "learning_rate": 9.967148526535813e-06,
+      "loss": 0.9657,
+      "step": 1423
+    },
+    {
+      "epoch": 0.0783752545544609,
+      "grad_norm": 0.8437683582305908,
+      "learning_rate": 9.967098900232616e-06,
+      "loss": 0.8336,
+      "step": 1424
+    },
+    {
+      "epoch": 0.07843029335681656,
+      "grad_norm": 0.8232185244560242,
+      "learning_rate": 9.967049236598034e-06,
+      "loss": 0.8878,
+      "step": 1425
+    },
+    {
+      "epoch": 0.07848533215917222,
+      "grad_norm": 1.0200369358062744,
+      "learning_rate": 9.96699953563244e-06,
+      "loss": 0.8135,
+      "step": 1426
+    },
+    {
+      "epoch": 0.07854037096152788,
+      "grad_norm": 0.8779187202453613,
+      "learning_rate": 9.966949797336208e-06,
+      "loss": 0.9124,
+      "step": 1427
+    },
+    {
+      "epoch": 0.07859540976388354,
+      "grad_norm": 0.9557466506958008,
+      "learning_rate": 9.966900021709708e-06,
+      "loss": 0.9118,
+      "step": 1428
+    },
+    {
+      "epoch": 0.07865044856623919,
+      "grad_norm": 0.8431050777435303,
+      "learning_rate": 9.966850208753317e-06,
+      "loss": 0.8361,
+      "step": 1429
+    },
+    {
+      "epoch": 0.07870548736859485,
+      "grad_norm": 0.9269648194313049,
+      "learning_rate": 9.966800358467412e-06,
+      "loss": 0.9194,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07876052617095052,
+      "grad_norm": 0.818681538105011,
+      "learning_rate": 9.966750470852363e-06,
+      "loss": 0.7483,
+      "step": 1431
+    },
+    {
+      "epoch": 0.07881556497330618,
+      "grad_norm": 0.8788284659385681,
+      "learning_rate": 9.966700545908547e-06,
+      "loss": 0.858,
+      "step": 1432
+    },
+    {
+      "epoch": 0.07887060377566184,
+      "grad_norm": 0.7734160423278809,
+      "learning_rate": 9.966650583636342e-06,
+      "loss": 0.694,
+      "step": 1433
+    },
+    {
+      "epoch": 0.0789256425780175,
+      "grad_norm": 0.8846608996391296,
+      "learning_rate": 9.966600584036117e-06,
+      "loss": 0.8144,
+      "step": 1434
+    },
+    {
+      "epoch": 0.07898068138037316,
+      "grad_norm": 0.9740058183670044,
+      "learning_rate": 9.966550547108254e-06,
+      "loss": 0.9314,
+      "step": 1435
+    },
+    {
+      "epoch": 0.07903572018272882,
+      "grad_norm": 0.8731759786605835,
+      "learning_rate": 9.966500472853124e-06,
+      "loss": 0.8475,
+      "step": 1436
+    },
+    {
+      "epoch": 0.07909075898508448,
+      "grad_norm": 0.8984843492507935,
+      "learning_rate": 9.966450361271109e-06,
+      "loss": 0.7803,
+      "step": 1437
+    },
+    {
+      "epoch": 0.07914579778744014,
+      "grad_norm": 0.8897966742515564,
+      "learning_rate": 9.96640021236258e-06,
+      "loss": 0.8879,
+      "step": 1438
+    },
+    {
+      "epoch": 0.0792008365897958,
+      "grad_norm": 0.80704265832901,
+      "learning_rate": 9.966350026127917e-06,
+      "loss": 0.7585,
+      "step": 1439
+    },
+    {
+      "epoch": 0.07925587539215147,
+      "grad_norm": 1.0807467699050903,
+      "learning_rate": 9.966299802567499e-06,
+      "loss": 1.078,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07931091419450713,
+      "grad_norm": 0.7994028925895691,
+      "learning_rate": 9.966249541681697e-06,
+      "loss": 0.8074,
+      "step": 1441
+    },
+    {
+      "epoch": 0.07936595299686279,
+      "grad_norm": 0.877592921257019,
+      "learning_rate": 9.966199243470895e-06,
+      "loss": 0.8084,
+      "step": 1442
+    },
+    {
+      "epoch": 0.07942099179921845,
+      "grad_norm": 0.7704572081565857,
+      "learning_rate": 9.966148907935469e-06,
+      "loss": 0.7206,
+      "step": 1443
+    },
+    {
+      "epoch": 0.07947603060157411,
+      "grad_norm": 0.8222140669822693,
+      "learning_rate": 9.966098535075797e-06,
+      "loss": 0.7768,
+      "step": 1444
+    },
+    {
+      "epoch": 0.07953106940392977,
+      "grad_norm": 1.389320731163025,
+      "learning_rate": 9.966048124892257e-06,
+      "loss": 1.0356,
+      "step": 1445
+    },
+    {
+      "epoch": 0.07958610820628544,
+      "grad_norm": 0.9082457423210144,
+      "learning_rate": 9.965997677385229e-06,
+      "loss": 0.7379,
+      "step": 1446
+    },
+    {
+      "epoch": 0.0796411470086411,
+      "grad_norm": 0.8029153943061829,
+      "learning_rate": 9.965947192555093e-06,
+      "loss": 0.7826,
+      "step": 1447
+    },
+    {
+      "epoch": 0.07969618581099676,
+      "grad_norm": 0.8752758502960205,
+      "learning_rate": 9.965896670402227e-06,
+      "loss": 0.8526,
+      "step": 1448
+    },
+    {
+      "epoch": 0.07975122461335242,
+      "grad_norm": 1.0665404796600342,
+      "learning_rate": 9.965846110927009e-06,
+      "loss": 0.858,
+      "step": 1449
+    },
+    {
+      "epoch": 0.07980626341570807,
+      "grad_norm": 0.9468502402305603,
+      "learning_rate": 9.96579551412982e-06,
+      "loss": 0.9658,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07986130221806373,
+      "grad_norm": 1.0239403247833252,
+      "learning_rate": 9.965744880011046e-06,
+      "loss": 0.7995,
+      "step": 1451
+    },
+    {
+      "epoch": 0.07991634102041939,
+      "grad_norm": 0.9808099865913391,
+      "learning_rate": 9.965694208571059e-06,
+      "loss": 1.0173,
+      "step": 1452
+    },
+    {
+      "epoch": 0.07997137982277505,
+      "grad_norm": 0.9338780641555786,
+      "learning_rate": 9.965643499810245e-06,
+      "loss": 0.7917,
+      "step": 1453
+    },
+    {
+      "epoch": 0.08002641862513071,
+      "grad_norm": 0.9294295310974121,
+      "learning_rate": 9.965592753728981e-06,
+      "loss": 0.88,
+      "step": 1454
+    },
+    {
+      "epoch": 0.08008145742748637,
+      "grad_norm": 1.0261508226394653,
+      "learning_rate": 9.965541970327654e-06,
+      "loss": 0.8825,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08013649622984204,
+      "grad_norm": 0.8964946269989014,
+      "learning_rate": 9.965491149606642e-06,
+      "loss": 0.81,
+      "step": 1456
+    },
+    {
+      "epoch": 0.0801915350321977,
+      "grad_norm": 0.9468267560005188,
+      "learning_rate": 9.965440291566329e-06,
+      "loss": 0.9453,
+      "step": 1457
+    },
+    {
+      "epoch": 0.08024657383455336,
+      "grad_norm": 0.8289040327072144,
+      "learning_rate": 9.965389396207092e-06,
+      "loss": 0.7373,
+      "step": 1458
+    },
+    {
+      "epoch": 0.08030161263690902,
+      "grad_norm": 0.8782384991645813,
+      "learning_rate": 9.965338463529322e-06,
+      "loss": 0.9199,
+      "step": 1459
+    },
+    {
+      "epoch": 0.08035665143926468,
+      "grad_norm": 0.8613787293434143,
+      "learning_rate": 9.965287493533395e-06,
+      "loss": 0.8719,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08041169024162034,
+      "grad_norm": 0.8474903106689453,
+      "learning_rate": 9.965236486219696e-06,
+      "loss": 0.8033,
+      "step": 1461
+    },
+    {
+      "epoch": 0.080466729043976,
+      "grad_norm": 1.1442681550979614,
+      "learning_rate": 9.965185441588609e-06,
+      "loss": 0.8996,
+      "step": 1462
+    },
+    {
+      "epoch": 0.08052176784633167,
+      "grad_norm": 1.564138412475586,
+      "learning_rate": 9.965134359640518e-06,
+      "loss": 0.7451,
+      "step": 1463
+    },
+    {
+      "epoch": 0.08057680664868733,
+      "grad_norm": 0.9211083054542542,
+      "learning_rate": 9.965083240375806e-06,
+      "loss": 0.8939,
+      "step": 1464
+    },
+    {
+      "epoch": 0.08063184545104299,
+      "grad_norm": 0.9503418207168579,
+      "learning_rate": 9.965032083794856e-06,
+      "loss": 0.8544,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08068688425339865,
+      "grad_norm": 0.9304021596908569,
+      "learning_rate": 9.964980889898055e-06,
+      "loss": 0.9192,
+      "step": 1466
+    },
+    {
+      "epoch": 0.08074192305575431,
+      "grad_norm": 0.8430425524711609,
+      "learning_rate": 9.964929658685787e-06,
+      "loss": 0.8586,
+      "step": 1467
+    },
+    {
+      "epoch": 0.08079696185810997,
+      "grad_norm": 0.8671759366989136,
+      "learning_rate": 9.964878390158437e-06,
+      "loss": 0.8807,
+      "step": 1468
+    },
+    {
+      "epoch": 0.08085200066046563,
+      "grad_norm": 0.9548830986022949,
+      "learning_rate": 9.964827084316389e-06,
+      "loss": 0.9033,
+      "step": 1469
+    },
+    {
+      "epoch": 0.0809070394628213,
+      "grad_norm": 0.8736767768859863,
+      "learning_rate": 9.964775741160029e-06,
+      "loss": 0.8509,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08096207826517696,
+      "grad_norm": 0.8827025890350342,
+      "learning_rate": 9.964724360689745e-06,
+      "loss": 0.897,
+      "step": 1471
+    },
+    {
+      "epoch": 0.0810171170675326,
+      "grad_norm": 1.02822744846344,
+      "learning_rate": 9.964672942905921e-06,
+      "loss": 1.0371,
+      "step": 1472
+    },
+    {
+      "epoch": 0.08107215586988827,
+      "grad_norm": 0.8619557619094849,
+      "learning_rate": 9.964621487808946e-06,
+      "loss": 0.7654,
+      "step": 1473
+    },
+    {
+      "epoch": 0.08112719467224393,
+      "grad_norm": 0.7855951189994812,
+      "learning_rate": 9.9645699953992e-06,
+      "loss": 0.7767,
+      "step": 1474
+    },
+    {
+      "epoch": 0.08118223347459959,
+      "grad_norm": 0.8139809370040894,
+      "learning_rate": 9.96451846567708e-06,
+      "loss": 0.7535,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08123727227695525,
+      "grad_norm": 0.8491657376289368,
+      "learning_rate": 9.964466898642966e-06,
+      "loss": 0.854,
+      "step": 1476
+    },
+    {
+      "epoch": 0.08129231107931091,
+      "grad_norm": 0.8968605399131775,
+      "learning_rate": 9.964415294297247e-06,
+      "loss": 0.8914,
+      "step": 1477
+    },
+    {
+      "epoch": 0.08134734988166657,
+      "grad_norm": 0.8692505359649658,
+      "learning_rate": 9.964363652640313e-06,
+      "loss": 0.9245,
+      "step": 1478
+    },
+    {
+      "epoch": 0.08140238868402223,
+      "grad_norm": 0.8916530013084412,
+      "learning_rate": 9.964311973672549e-06,
+      "loss": 0.7662,
+      "step": 1479
+    },
+    {
+      "epoch": 0.0814574274863779,
+      "grad_norm": 0.8239215612411499,
+      "learning_rate": 9.964260257394347e-06,
+      "loss": 0.9191,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08151246628873356,
+      "grad_norm": 0.8672100901603699,
+      "learning_rate": 9.964208503806092e-06,
+      "loss": 0.7656,
+      "step": 1481
+    },
+    {
+      "epoch": 0.08156750509108922,
+      "grad_norm": 0.9195712208747864,
+      "learning_rate": 9.964156712908177e-06,
+      "loss": 0.8656,
+      "step": 1482
+    },
+    {
+      "epoch": 0.08162254389344488,
+      "grad_norm": 0.8282535672187805,
+      "learning_rate": 9.964104884700986e-06,
+      "loss": 0.8264,
+      "step": 1483
+    },
+    {
+      "epoch": 0.08167758269580054,
+      "grad_norm": 0.8492032289505005,
+      "learning_rate": 9.964053019184913e-06,
+      "loss": 0.7816,
+      "step": 1484
+    },
+    {
+      "epoch": 0.0817326214981562,
+      "grad_norm": 0.8491117358207703,
+      "learning_rate": 9.964001116360347e-06,
+      "loss": 0.7885,
+      "step": 1485
+    },
+    {
+      "epoch": 0.08178766030051186,
+      "grad_norm": 0.9415153861045837,
+      "learning_rate": 9.963949176227677e-06,
+      "loss": 0.8165,
+      "step": 1486
+    },
+    {
+      "epoch": 0.08184269910286752,
+      "grad_norm": 0.8462526202201843,
+      "learning_rate": 9.963897198787294e-06,
+      "loss": 0.8498,
+      "step": 1487
+    },
+    {
+      "epoch": 0.08189773790522319,
+      "grad_norm": 0.8591959476470947,
+      "learning_rate": 9.963845184039586e-06,
+      "loss": 0.8906,
+      "step": 1488
+    },
+    {
+      "epoch": 0.08195277670757885,
+      "grad_norm": 0.840761661529541,
+      "learning_rate": 9.963793131984949e-06,
+      "loss": 0.7831,
+      "step": 1489
+    },
+    {
+      "epoch": 0.08200781550993451,
+      "grad_norm": 0.931404173374176,
+      "learning_rate": 9.96374104262377e-06,
+      "loss": 0.889,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08206285431229017,
+      "grad_norm": 0.9048783779144287,
+      "learning_rate": 9.963688915956443e-06,
+      "loss": 0.8321,
+      "step": 1491
+    },
+    {
+      "epoch": 0.08211789311464583,
+      "grad_norm": 0.9145931601524353,
+      "learning_rate": 9.96363675198336e-06,
+      "loss": 0.9918,
+      "step": 1492
+    },
+    {
+      "epoch": 0.08217293191700148,
+      "grad_norm": 0.9256643652915955,
+      "learning_rate": 9.963584550704908e-06,
+      "loss": 0.8731,
+      "step": 1493
+    },
+    {
+      "epoch": 0.08222797071935714,
+      "grad_norm": 1.0212007761001587,
+      "learning_rate": 9.963532312121486e-06,
+      "loss": 0.9077,
+      "step": 1494
+    },
+    {
+      "epoch": 0.0822830095217128,
+      "grad_norm": 0.9206242561340332,
+      "learning_rate": 9.963480036233483e-06,
+      "loss": 0.9076,
+      "step": 1495
+    },
+    {
+      "epoch": 0.08233804832406846,
+      "grad_norm": 0.8846865296363831,
+      "learning_rate": 9.963427723041294e-06,
+      "loss": 0.6826,
+      "step": 1496
+    },
+    {
+      "epoch": 0.08239308712642412,
+      "grad_norm": 0.8745351433753967,
+      "learning_rate": 9.963375372545309e-06,
+      "loss": 0.7935,
+      "step": 1497
+    },
+    {
+      "epoch": 0.08244812592877979,
+      "grad_norm": 0.9019666314125061,
+      "learning_rate": 9.963322984745924e-06,
+      "loss": 0.8435,
+      "step": 1498
+    },
+    {
+      "epoch": 0.08250316473113545,
+      "grad_norm": 0.8586859703063965,
+      "learning_rate": 9.963270559643531e-06,
+      "loss": 0.8118,
+      "step": 1499
+    },
+    {
+      "epoch": 0.08255820353349111,
+      "grad_norm": 0.9192817807197571,
+      "learning_rate": 9.963218097238528e-06,
+      "loss": 0.824,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08261324233584677,
+      "grad_norm": 0.8972243070602417,
+      "learning_rate": 9.963165597531304e-06,
+      "loss": 0.8404,
+      "step": 1501
+    },
+    {
+      "epoch": 0.08266828113820243,
+      "grad_norm": 0.8953961133956909,
+      "learning_rate": 9.963113060522256e-06,
+      "loss": 0.9031,
+      "step": 1502
+    },
+    {
+      "epoch": 0.0827233199405581,
+      "grad_norm": 0.9551270604133606,
+      "learning_rate": 9.963060486211779e-06,
+      "loss": 0.9177,
+      "step": 1503
+    },
+    {
+      "epoch": 0.08277835874291375,
+      "grad_norm": 0.8524616956710815,
+      "learning_rate": 9.963007874600268e-06,
+      "loss": 0.8582,
+      "step": 1504
+    },
+    {
+      "epoch": 0.08283339754526942,
+      "grad_norm": 0.8148764371871948,
+      "learning_rate": 9.962955225688118e-06,
+      "loss": 0.6859,
+      "step": 1505
+    },
+    {
+      "epoch": 0.08288843634762508,
+      "grad_norm": 0.9110590219497681,
+      "learning_rate": 9.962902539475728e-06,
+      "loss": 0.7189,
+      "step": 1506
+    },
+    {
+      "epoch": 0.08294347514998074,
+      "grad_norm": 0.8700116872787476,
+      "learning_rate": 9.962849815963487e-06,
+      "loss": 0.9462,
+      "step": 1507
+    },
+    {
+      "epoch": 0.0829985139523364,
+      "grad_norm": 0.877109706401825,
+      "learning_rate": 9.962797055151797e-06,
+      "loss": 0.8138,
+      "step": 1508
+    },
+    {
+      "epoch": 0.08305355275469206,
+      "grad_norm": 0.7818365097045898,
+      "learning_rate": 9.962744257041053e-06,
+      "loss": 0.8474,
+      "step": 1509
+    },
+    {
+      "epoch": 0.08310859155704772,
+      "grad_norm": 0.88360196352005,
+      "learning_rate": 9.96269142163165e-06,
+      "loss": 0.8724,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08316363035940338,
+      "grad_norm": 0.8982682228088379,
+      "learning_rate": 9.962638548923988e-06,
+      "loss": 0.9687,
+      "step": 1511
+    },
+    {
+      "epoch": 0.08321866916175905,
+      "grad_norm": 0.7362002730369568,
+      "learning_rate": 9.962585638918462e-06,
+      "loss": 0.7666,
+      "step": 1512
+    },
+    {
+      "epoch": 0.08327370796411471,
+      "grad_norm": 1.0993375778198242,
+      "learning_rate": 9.962532691615472e-06,
+      "loss": 0.8869,
+      "step": 1513
+    },
+    {
+      "epoch": 0.08332874676647037,
+      "grad_norm": 0.8684842586517334,
+      "learning_rate": 9.962479707015415e-06,
+      "loss": 0.872,
+      "step": 1514
+    },
+    {
+      "epoch": 0.08338378556882602,
+      "grad_norm": 1.0598478317260742,
+      "learning_rate": 9.962426685118689e-06,
+      "loss": 0.9102,
+      "step": 1515
+    },
+    {
+      "epoch": 0.08343882437118168,
+      "grad_norm": 0.8492125272750854,
+      "learning_rate": 9.96237362592569e-06,
+      "loss": 0.7554,
+      "step": 1516
+    },
+    {
+      "epoch": 0.08349386317353734,
+      "grad_norm": 0.8489052653312683,
+      "learning_rate": 9.962320529436821e-06,
+      "loss": 0.9139,
+      "step": 1517
+    },
+    {
+      "epoch": 0.083548901975893,
+      "grad_norm": 0.8650774359703064,
+      "learning_rate": 9.962267395652479e-06,
+      "loss": 0.8717,
+      "step": 1518
+    },
+    {
+      "epoch": 0.08360394077824866,
+      "grad_norm": 0.8393206596374512,
+      "learning_rate": 9.962214224573064e-06,
+      "loss": 0.8256,
+      "step": 1519
+    },
+    {
+      "epoch": 0.08365897958060432,
+      "grad_norm": 0.8304896354675293,
+      "learning_rate": 9.962161016198974e-06,
+      "loss": 0.8232,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08371401838295998,
+      "grad_norm": 0.8718386292457581,
+      "learning_rate": 9.962107770530612e-06,
+      "loss": 0.8206,
+      "step": 1521
+    },
+    {
+      "epoch": 0.08376905718531565,
+      "grad_norm": 0.9109341502189636,
+      "learning_rate": 9.962054487568373e-06,
+      "loss": 0.9576,
+      "step": 1522
+    },
+    {
+      "epoch": 0.08382409598767131,
+      "grad_norm": 0.9543303847312927,
+      "learning_rate": 9.962001167312663e-06,
+      "loss": 0.8816,
+      "step": 1523
+    },
+    {
+      "epoch": 0.08387913479002697,
+      "grad_norm": 0.9992844462394714,
+      "learning_rate": 9.961947809763881e-06,
+      "loss": 0.8682,
+      "step": 1524
+    },
+    {
+      "epoch": 0.08393417359238263,
+      "grad_norm": 0.8092770576477051,
+      "learning_rate": 9.961894414922425e-06,
+      "loss": 0.6352,
+      "step": 1525
+    },
+    {
+      "epoch": 0.08398921239473829,
+      "grad_norm": 0.9888653755187988,
+      "learning_rate": 9.961840982788703e-06,
+      "loss": 0.8721,
+      "step": 1526
+    },
+    {
+      "epoch": 0.08404425119709395,
+      "grad_norm": 1.0092703104019165,
+      "learning_rate": 9.961787513363108e-06,
+      "loss": 0.7776,
+      "step": 1527
+    },
+    {
+      "epoch": 0.08409928999944961,
+      "grad_norm": 0.8654646277427673,
+      "learning_rate": 9.961734006646049e-06,
+      "loss": 0.8835,
+      "step": 1528
+    },
+    {
+      "epoch": 0.08415432880180528,
+      "grad_norm": 0.7630153298377991,
+      "learning_rate": 9.961680462637924e-06,
+      "loss": 0.7501,
+      "step": 1529
+    },
+    {
+      "epoch": 0.08420936760416094,
+      "grad_norm": 1.1883158683776855,
+      "learning_rate": 9.961626881339138e-06,
+      "loss": 0.9476,
+      "step": 1530
+    },
+    {
+      "epoch": 0.0842644064065166,
+      "grad_norm": 0.8710927963256836,
+      "learning_rate": 9.96157326275009e-06,
+      "loss": 0.749,
+      "step": 1531
+    },
+    {
+      "epoch": 0.08431944520887226,
+      "grad_norm": 0.9500633478164673,
+      "learning_rate": 9.961519606871188e-06,
+      "loss": 0.8994,
+      "step": 1532
+    },
+    {
+      "epoch": 0.08437448401122792,
+      "grad_norm": 0.873257577419281,
+      "learning_rate": 9.961465913702833e-06,
+      "loss": 0.816,
+      "step": 1533
+    },
+    {
+      "epoch": 0.08442952281358358,
+      "grad_norm": 0.8007022142410278,
+      "learning_rate": 9.961412183245426e-06,
+      "loss": 0.787,
+      "step": 1534
+    },
+    {
+      "epoch": 0.08448456161593924,
+      "grad_norm": 0.8998435139656067,
+      "learning_rate": 9.961358415499374e-06,
+      "loss": 0.8741,
+      "step": 1535
+    },
+    {
+      "epoch": 0.08453960041829489,
+      "grad_norm": 0.9152502417564392,
+      "learning_rate": 9.961304610465081e-06,
+      "loss": 0.9749,
+      "step": 1536
+    },
+    {
+      "epoch": 0.08459463922065055,
+      "grad_norm": 0.8961958289146423,
+      "learning_rate": 9.961250768142949e-06,
+      "loss": 0.8683,
+      "step": 1537
+    },
+    {
+      "epoch": 0.08464967802300621,
+      "grad_norm": 0.8683995008468628,
+      "learning_rate": 9.961196888533387e-06,
+      "loss": 0.8347,
+      "step": 1538
+    },
+    {
+      "epoch": 0.08470471682536188,
+      "grad_norm": 0.835221529006958,
+      "learning_rate": 9.961142971636795e-06,
+      "loss": 0.8936,
+      "step": 1539
+    },
+    {
+      "epoch": 0.08475975562771754,
+      "grad_norm": 0.8666725158691406,
+      "learning_rate": 9.96108901745358e-06,
+      "loss": 0.7344,
+      "step": 1540
+    },
+    {
+      "epoch": 0.0848147944300732,
+      "grad_norm": 0.9509082436561584,
+      "learning_rate": 9.96103502598415e-06,
+      "loss": 0.8965,
+      "step": 1541
+    },
+    {
+      "epoch": 0.08486983323242886,
+      "grad_norm": 0.8134233951568604,
+      "learning_rate": 9.960980997228908e-06,
+      "loss": 0.797,
+      "step": 1542
+    },
+    {
+      "epoch": 0.08492487203478452,
+      "grad_norm": 1.0432242155075073,
+      "learning_rate": 9.96092693118826e-06,
+      "loss": 0.8754,
+      "step": 1543
+    },
+    {
+      "epoch": 0.08497991083714018,
+      "grad_norm": 0.9560218453407288,
+      "learning_rate": 9.960872827862613e-06,
+      "loss": 0.9238,
+      "step": 1544
+    },
+    {
+      "epoch": 0.08503494963949584,
+      "grad_norm": 0.8471649885177612,
+      "learning_rate": 9.960818687252374e-06,
+      "loss": 0.8622,
+      "step": 1545
+    },
+    {
+      "epoch": 0.0850899884418515,
+      "grad_norm": 1.2584747076034546,
+      "learning_rate": 9.960764509357951e-06,
+      "loss": 0.8007,
+      "step": 1546
+    },
+    {
+      "epoch": 0.08514502724420717,
+      "grad_norm": 0.8730618953704834,
+      "learning_rate": 9.960710294179748e-06,
+      "loss": 0.7412,
+      "step": 1547
+    },
+    {
+      "epoch": 0.08520006604656283,
+      "grad_norm": 0.8361592292785645,
+      "learning_rate": 9.960656041718176e-06,
+      "loss": 0.7018,
+      "step": 1548
+    },
+    {
+      "epoch": 0.08525510484891849,
+      "grad_norm": 0.8351722359657288,
+      "learning_rate": 9.96060175197364e-06,
+      "loss": 0.843,
+      "step": 1549
+    },
+    {
+      "epoch": 0.08531014365127415,
+      "grad_norm": 0.8665090203285217,
+      "learning_rate": 9.960547424946549e-06,
+      "loss": 0.8235,
+      "step": 1550
+    },
+    {
+      "epoch": 0.08536518245362981,
+      "grad_norm": 0.9254478812217712,
+      "learning_rate": 9.960493060637313e-06,
+      "loss": 0.8122,
+      "step": 1551
+    },
+    {
+      "epoch": 0.08542022125598547,
+      "grad_norm": 0.8712261319160461,
+      "learning_rate": 9.960438659046337e-06,
+      "loss": 0.823,
+      "step": 1552
+    },
+    {
+      "epoch": 0.08547526005834113,
+      "grad_norm": 0.9027207493782043,
+      "learning_rate": 9.960384220174033e-06,
+      "loss": 0.7964,
+      "step": 1553
+    },
+    {
+      "epoch": 0.0855302988606968,
+      "grad_norm": 0.854626476764679,
+      "learning_rate": 9.960329744020808e-06,
+      "loss": 0.755,
+      "step": 1554
+    },
+    {
+      "epoch": 0.08558533766305246,
+      "grad_norm": 0.9398048520088196,
+      "learning_rate": 9.960275230587073e-06,
+      "loss": 0.8607,
+      "step": 1555
+    },
+    {
+      "epoch": 0.08564037646540812,
+      "grad_norm": 1.008002758026123,
+      "learning_rate": 9.960220679873238e-06,
+      "loss": 0.9711,
+      "step": 1556
+    },
+    {
+      "epoch": 0.08569541526776378,
+      "grad_norm": 0.8999453783035278,
+      "learning_rate": 9.96016609187971e-06,
+      "loss": 0.8233,
+      "step": 1557
+    },
+    {
+      "epoch": 0.08575045407011943,
+      "grad_norm": 0.8912106156349182,
+      "learning_rate": 9.960111466606903e-06,
+      "loss": 0.8271,
+      "step": 1558
+    },
+    {
+      "epoch": 0.08580549287247509,
+      "grad_norm": 0.9269998073577881,
+      "learning_rate": 9.960056804055227e-06,
+      "loss": 0.7959,
+      "step": 1559
+    },
+    {
+      "epoch": 0.08586053167483075,
+      "grad_norm": 1.083815336227417,
+      "learning_rate": 9.96000210422509e-06,
+      "loss": 0.9436,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08591557047718641,
+      "grad_norm": 0.8906280398368835,
+      "learning_rate": 9.959947367116905e-06,
+      "loss": 0.9317,
+      "step": 1561
+    },
+    {
+      "epoch": 0.08597060927954207,
+      "grad_norm": 1.211696743965149,
+      "learning_rate": 9.959892592731084e-06,
+      "loss": 0.9076,
+      "step": 1562
+    },
+    {
+      "epoch": 0.08602564808189773,
+      "grad_norm": 0.9050534963607788,
+      "learning_rate": 9.959837781068038e-06,
+      "loss": 0.8728,
+      "step": 1563
+    },
+    {
+      "epoch": 0.0860806868842534,
+      "grad_norm": 0.9384796619415283,
+      "learning_rate": 9.959782932128178e-06,
+      "loss": 0.9277,
+      "step": 1564
+    },
+    {
+      "epoch": 0.08613572568660906,
+      "grad_norm": 0.795844316482544,
+      "learning_rate": 9.959728045911915e-06,
+      "loss": 0.7666,
+      "step": 1565
+    },
+    {
+      "epoch": 0.08619076448896472,
+      "grad_norm": 0.925956666469574,
+      "learning_rate": 9.959673122419668e-06,
+      "loss": 0.815,
+      "step": 1566
+    },
+    {
+      "epoch": 0.08624580329132038,
+      "grad_norm": 0.898047924041748,
+      "learning_rate": 9.959618161651843e-06,
+      "loss": 0.8131,
+      "step": 1567
+    },
+    {
+      "epoch": 0.08630084209367604,
+      "grad_norm": 0.8656220436096191,
+      "learning_rate": 9.959563163608856e-06,
+      "loss": 0.9336,
+      "step": 1568
+    },
+    {
+      "epoch": 0.0863558808960317,
+      "grad_norm": 0.9184645414352417,
+      "learning_rate": 9.95950812829112e-06,
+      "loss": 0.9557,
+      "step": 1569
+    },
+    {
+      "epoch": 0.08641091969838736,
+      "grad_norm": 0.8607667684555054,
+      "learning_rate": 9.959453055699048e-06,
+      "loss": 0.8272,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08646595850074303,
+      "grad_norm": 0.9561272263526917,
+      "learning_rate": 9.959397945833056e-06,
+      "loss": 0.8876,
+      "step": 1571
+    },
+    {
+      "epoch": 0.08652099730309869,
+      "grad_norm": 0.8562412261962891,
+      "learning_rate": 9.959342798693556e-06,
+      "loss": 0.8404,
+      "step": 1572
+    },
+    {
+      "epoch": 0.08657603610545435,
+      "grad_norm": 0.8924610614776611,
+      "learning_rate": 9.95928761428096e-06,
+      "loss": 0.8779,
+      "step": 1573
+    },
+    {
+      "epoch": 0.08663107490781001,
+      "grad_norm": 0.8343208432197571,
+      "learning_rate": 9.95923239259569e-06,
+      "loss": 0.8992,
+      "step": 1574
+    },
+    {
+      "epoch": 0.08668611371016567,
+      "grad_norm": 0.8835015296936035,
+      "learning_rate": 9.959177133638155e-06,
+      "loss": 1.0026,
+      "step": 1575
+    },
+    {
+      "epoch": 0.08674115251252133,
+      "grad_norm": 0.9540221095085144,
+      "learning_rate": 9.959121837408771e-06,
+      "loss": 0.8507,
+      "step": 1576
+    },
+    {
+      "epoch": 0.086796191314877,
+      "grad_norm": 1.087817668914795,
+      "learning_rate": 9.959066503907957e-06,
+      "loss": 0.8607,
+      "step": 1577
+    },
+    {
+      "epoch": 0.08685123011723266,
+      "grad_norm": 0.8072447180747986,
+      "learning_rate": 9.959011133136124e-06,
+      "loss": 0.882,
+      "step": 1578
+    },
+    {
+      "epoch": 0.0869062689195883,
+      "grad_norm": 0.7646876573562622,
+      "learning_rate": 9.958955725093694e-06,
+      "loss": 0.7653,
+      "step": 1579
+    },
+    {
+      "epoch": 0.08696130772194396,
+      "grad_norm": 0.8979537487030029,
+      "learning_rate": 9.958900279781078e-06,
+      "loss": 0.9033,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08701634652429963,
+      "grad_norm": 0.9445611834526062,
+      "learning_rate": 9.958844797198696e-06,
+      "loss": 0.9423,
+      "step": 1581
+    },
+    {
+      "epoch": 0.08707138532665529,
+      "grad_norm": 0.8836671113967896,
+      "learning_rate": 9.958789277346963e-06,
+      "loss": 0.839,
+      "step": 1582
+    },
+    {
+      "epoch": 0.08712642412901095,
+      "grad_norm": 1.0333542823791504,
+      "learning_rate": 9.958733720226296e-06,
+      "loss": 0.9211,
+      "step": 1583
+    },
+    {
+      "epoch": 0.08718146293136661,
+      "grad_norm": 0.8084085583686829,
+      "learning_rate": 9.958678125837117e-06,
+      "loss": 0.8387,
+      "step": 1584
+    },
+    {
+      "epoch": 0.08723650173372227,
+      "grad_norm": 0.7769419550895691,
+      "learning_rate": 9.958622494179838e-06,
+      "loss": 0.8307,
+      "step": 1585
+    },
+    {
+      "epoch": 0.08729154053607793,
+      "grad_norm": 0.8387578129768372,
+      "learning_rate": 9.95856682525488e-06,
+      "loss": 0.8001,
+      "step": 1586
+    },
+    {
+      "epoch": 0.0873465793384336,
+      "grad_norm": 0.8989812731742859,
+      "learning_rate": 9.95851111906266e-06,
+      "loss": 0.7752,
+      "step": 1587
+    },
+    {
+      "epoch": 0.08740161814078926,
+      "grad_norm": 0.8558734655380249,
+      "learning_rate": 9.958455375603602e-06,
+      "loss": 0.8149,
+      "step": 1588
+    },
+    {
+      "epoch": 0.08745665694314492,
+      "grad_norm": 0.8890896439552307,
+      "learning_rate": 9.958399594878117e-06,
+      "loss": 0.8232,
+      "step": 1589
+    },
+    {
+      "epoch": 0.08751169574550058,
+      "grad_norm": 0.875912070274353,
+      "learning_rate": 9.95834377688663e-06,
+      "loss": 0.7458,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08756673454785624,
+      "grad_norm": 0.808355987071991,
+      "learning_rate": 9.958287921629557e-06,
+      "loss": 0.8296,
+      "step": 1591
+    },
+    {
+      "epoch": 0.0876217733502119,
+      "grad_norm": 0.9637090563774109,
+      "learning_rate": 9.958232029107318e-06,
+      "loss": 0.8769,
+      "step": 1592
+    },
+    {
+      "epoch": 0.08767681215256756,
+      "grad_norm": 0.8980715870857239,
+      "learning_rate": 9.958176099320336e-06,
+      "loss": 0.7995,
+      "step": 1593
+    },
+    {
+      "epoch": 0.08773185095492322,
+      "grad_norm": 0.9369860291481018,
+      "learning_rate": 9.95812013226903e-06,
+      "loss": 0.8545,
+      "step": 1594
+    },
+    {
+      "epoch": 0.08778688975727889,
+      "grad_norm": 0.8589349389076233,
+      "learning_rate": 9.958064127953819e-06,
+      "loss": 0.8693,
+      "step": 1595
+    },
+    {
+      "epoch": 0.08784192855963455,
+      "grad_norm": 0.929207444190979,
+      "learning_rate": 9.958008086375126e-06,
+      "loss": 0.811,
+      "step": 1596
+    },
+    {
+      "epoch": 0.08789696736199021,
+      "grad_norm": 1.0825661420822144,
+      "learning_rate": 9.957952007533371e-06,
+      "loss": 1.0145,
+      "step": 1597
+    },
+    {
+      "epoch": 0.08795200616434587,
+      "grad_norm": 0.8818382024765015,
+      "learning_rate": 9.957895891428978e-06,
+      "loss": 0.7771,
+      "step": 1598
+    },
+    {
+      "epoch": 0.08800704496670153,
+      "grad_norm": 0.882780909538269,
+      "learning_rate": 9.957839738062363e-06,
+      "loss": 0.8857,
+      "step": 1599
+    },
+    {
+      "epoch": 0.08806208376905718,
+      "grad_norm": 0.9136924743652344,
+      "learning_rate": 9.957783547433955e-06,
+      "loss": 0.8873,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08811712257141284,
+      "grad_norm": 0.8896858096122742,
+      "learning_rate": 9.95772731954417e-06,
+      "loss": 0.8463,
+      "step": 1601
+    },
+    {
+      "epoch": 0.0881721613737685,
+      "grad_norm": 0.8671631813049316,
+      "learning_rate": 9.957671054393436e-06,
+      "loss": 0.8333,
+      "step": 1602
+    },
+    {
+      "epoch": 0.08822720017612416,
+      "grad_norm": 0.9442896246910095,
+      "learning_rate": 9.957614751982172e-06,
+      "loss": 0.9676,
+      "step": 1603
+    },
+    {
+      "epoch": 0.08828223897847982,
+      "grad_norm": 0.8249240517616272,
+      "learning_rate": 9.957558412310803e-06,
+      "loss": 0.7746,
+      "step": 1604
+    },
+    {
+      "epoch": 0.08833727778083549,
+      "grad_norm": 0.8125253319740295,
+      "learning_rate": 9.957502035379751e-06,
+      "loss": 0.7816,
+      "step": 1605
+    },
+    {
+      "epoch": 0.08839231658319115,
+      "grad_norm": 0.8467233777046204,
+      "learning_rate": 9.957445621189442e-06,
+      "loss": 0.7697,
+      "step": 1606
+    },
+    {
+      "epoch": 0.08844735538554681,
+      "grad_norm": 0.8322175145149231,
+      "learning_rate": 9.957389169740299e-06,
+      "loss": 0.7561,
+      "step": 1607
+    },
+    {
+      "epoch": 0.08850239418790247,
+      "grad_norm": 0.869163453578949,
+      "learning_rate": 9.957332681032746e-06,
+      "loss": 0.8984,
+      "step": 1608
+    },
+    {
+      "epoch": 0.08855743299025813,
+      "grad_norm": 0.8755944967269897,
+      "learning_rate": 9.957276155067206e-06,
+      "loss": 0.8016,
+      "step": 1609
+    },
+    {
+      "epoch": 0.08861247179261379,
+      "grad_norm": 0.8152669668197632,
+      "learning_rate": 9.957219591844108e-06,
+      "loss": 0.7763,
+      "step": 1610
+    },
+    {
+      "epoch": 0.08866751059496945,
+      "grad_norm": 0.979752779006958,
+      "learning_rate": 9.957162991363871e-06,
+      "loss": 0.7755,
+      "step": 1611
+    },
+    {
+      "epoch": 0.08872254939732512,
+      "grad_norm": 1.0481054782867432,
+      "learning_rate": 9.957106353626926e-06,
+      "loss": 0.9395,
+      "step": 1612
+    },
+    {
+      "epoch": 0.08877758819968078,
+      "grad_norm": 0.7773686647415161,
+      "learning_rate": 9.957049678633697e-06,
+      "loss": 0.7713,
+      "step": 1613
+    },
+    {
+      "epoch": 0.08883262700203644,
+      "grad_norm": 0.838979959487915,
+      "learning_rate": 9.956992966384609e-06,
+      "loss": 0.7909,
+      "step": 1614
+    },
+    {
+      "epoch": 0.0888876658043921,
+      "grad_norm": 0.9527049660682678,
+      "learning_rate": 9.956936216880089e-06,
+      "loss": 0.7944,
+      "step": 1615
+    },
+    {
+      "epoch": 0.08894270460674776,
+      "grad_norm": 0.7967305183410645,
+      "learning_rate": 9.956879430120561e-06,
+      "loss": 0.7703,
+      "step": 1616
+    },
+    {
+      "epoch": 0.08899774340910342,
+      "grad_norm": 0.9065802097320557,
+      "learning_rate": 9.956822606106456e-06,
+      "loss": 0.8188,
+      "step": 1617
+    },
+    {
+      "epoch": 0.08905278221145908,
+      "grad_norm": 0.7329322099685669,
+      "learning_rate": 9.956765744838199e-06,
+      "loss": 0.8043,
+      "step": 1618
+    },
+    {
+      "epoch": 0.08910782101381474,
+      "grad_norm": 0.864973247051239,
+      "learning_rate": 9.95670884631622e-06,
+      "loss": 0.8334,
+      "step": 1619
+    },
+    {
+      "epoch": 0.0891628598161704,
+      "grad_norm": 1.073559045791626,
+      "learning_rate": 9.95665191054094e-06,
+      "loss": 0.7755,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08921789861852607,
+      "grad_norm": 0.7347918748855591,
+      "learning_rate": 9.956594937512794e-06,
+      "loss": 0.7556,
+      "step": 1621
+    },
+    {
+      "epoch": 0.08927293742088172,
+      "grad_norm": 0.8756610751152039,
+      "learning_rate": 9.956537927232205e-06,
+      "loss": 0.8129,
+      "step": 1622
+    },
+    {
+      "epoch": 0.08932797622323738,
+      "grad_norm": 0.9132435917854309,
+      "learning_rate": 9.956480879699605e-06,
+      "loss": 0.8221,
+      "step": 1623
+    },
+    {
+      "epoch": 0.08938301502559304,
+      "grad_norm": 1.1978256702423096,
+      "learning_rate": 9.956423794915421e-06,
+      "loss": 0.8651,
+      "step": 1624
+    },
+    {
+      "epoch": 0.0894380538279487,
+      "grad_norm": 0.8493894934654236,
+      "learning_rate": 9.956366672880082e-06,
+      "loss": 0.7267,
+      "step": 1625
+    },
+    {
+      "epoch": 0.08949309263030436,
+      "grad_norm": 1.0971951484680176,
+      "learning_rate": 9.956309513594019e-06,
+      "loss": 0.7852,
+      "step": 1626
+    },
+    {
+      "epoch": 0.08954813143266002,
+      "grad_norm": 0.899974524974823,
+      "learning_rate": 9.95625231705766e-06,
+      "loss": 0.8868,
+      "step": 1627
+    },
+    {
+      "epoch": 0.08960317023501568,
+      "grad_norm": 0.8995566368103027,
+      "learning_rate": 9.956195083271436e-06,
+      "loss": 0.87,
+      "step": 1628
+    },
+    {
+      "epoch": 0.08965820903737134,
+      "grad_norm": 0.8924218416213989,
+      "learning_rate": 9.956137812235776e-06,
+      "loss": 0.7885,
+      "step": 1629
+    },
+    {
+      "epoch": 0.089713247839727,
+      "grad_norm": 0.9232820868492126,
+      "learning_rate": 9.956080503951108e-06,
+      "loss": 0.7923,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08976828664208267,
+      "grad_norm": 0.9298982620239258,
+      "learning_rate": 9.956023158417869e-06,
+      "loss": 0.8625,
+      "step": 1631
+    },
+    {
+      "epoch": 0.08982332544443833,
+      "grad_norm": 0.86515212059021,
+      "learning_rate": 9.955965775636488e-06,
+      "loss": 0.7683,
+      "step": 1632
+    },
+    {
+      "epoch": 0.08987836424679399,
+      "grad_norm": 0.8016952276229858,
+      "learning_rate": 9.955908355607392e-06,
+      "loss": 0.8122,
+      "step": 1633
+    },
+    {
+      "epoch": 0.08993340304914965,
+      "grad_norm": 0.842703640460968,
+      "learning_rate": 9.955850898331015e-06,
+      "loss": 0.8487,
+      "step": 1634
+    },
+    {
+      "epoch": 0.08998844185150531,
+      "grad_norm": 0.8239083886146545,
+      "learning_rate": 9.95579340380779e-06,
+      "loss": 0.8701,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09004348065386097,
+      "grad_norm": 0.8575418591499329,
+      "learning_rate": 9.955735872038149e-06,
+      "loss": 0.8263,
+      "step": 1636
+    },
+    {
+      "epoch": 0.09009851945621664,
+      "grad_norm": 0.8884586095809937,
+      "learning_rate": 9.955678303022522e-06,
+      "loss": 0.8112,
+      "step": 1637
+    },
+    {
+      "epoch": 0.0901535582585723,
+      "grad_norm": 0.9024681448936462,
+      "learning_rate": 9.955620696761345e-06,
+      "loss": 0.9174,
+      "step": 1638
+    },
+    {
+      "epoch": 0.09020859706092796,
+      "grad_norm": 0.8151944875717163,
+      "learning_rate": 9.955563053255049e-06,
+      "loss": 0.806,
+      "step": 1639
+    },
+    {
+      "epoch": 0.09026363586328362,
+      "grad_norm": 0.8292184472084045,
+      "learning_rate": 9.955505372504069e-06,
+      "loss": 0.8007,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09031867466563928,
+      "grad_norm": 0.9445936679840088,
+      "learning_rate": 9.955447654508835e-06,
+      "loss": 0.7089,
+      "step": 1641
+    },
+    {
+      "epoch": 0.09037371346799494,
+      "grad_norm": 0.781579315662384,
+      "learning_rate": 9.955389899269782e-06,
+      "loss": 0.8224,
+      "step": 1642
+    },
+    {
+      "epoch": 0.09042875227035059,
+      "grad_norm": 0.9028880596160889,
+      "learning_rate": 9.955332106787348e-06,
+      "loss": 0.7976,
+      "step": 1643
+    },
+    {
+      "epoch": 0.09048379107270625,
+      "grad_norm": 1.0336887836456299,
+      "learning_rate": 9.955274277061963e-06,
+      "loss": 0.9296,
+      "step": 1644
+    },
+    {
+      "epoch": 0.09053882987506191,
+      "grad_norm": 0.8894197940826416,
+      "learning_rate": 9.955216410094062e-06,
+      "loss": 0.815,
+      "step": 1645
+    },
+    {
+      "epoch": 0.09059386867741757,
+      "grad_norm": 0.8955528140068054,
+      "learning_rate": 9.955158505884083e-06,
+      "loss": 0.8707,
+      "step": 1646
+    },
+    {
+      "epoch": 0.09064890747977324,
+      "grad_norm": 0.8012683987617493,
+      "learning_rate": 9.955100564432458e-06,
+      "loss": 0.7467,
+      "step": 1647
+    },
+    {
+      "epoch": 0.0907039462821289,
+      "grad_norm": 0.917969286441803,
+      "learning_rate": 9.955042585739623e-06,
+      "loss": 0.8835,
+      "step": 1648
+    },
+    {
+      "epoch": 0.09075898508448456,
+      "grad_norm": 0.8066666722297668,
+      "learning_rate": 9.954984569806014e-06,
+      "loss": 0.8338,
+      "step": 1649
+    },
+    {
+      "epoch": 0.09081402388684022,
+      "grad_norm": 1.1324070692062378,
+      "learning_rate": 9.954926516632069e-06,
+      "loss": 0.8245,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09086906268919588,
+      "grad_norm": 0.8196014761924744,
+      "learning_rate": 9.954868426218222e-06,
+      "loss": 0.7897,
+      "step": 1651
+    },
+    {
+      "epoch": 0.09092410149155154,
+      "grad_norm": 0.8713478446006775,
+      "learning_rate": 9.95481029856491e-06,
+      "loss": 0.891,
+      "step": 1652
+    },
+    {
+      "epoch": 0.0909791402939072,
+      "grad_norm": 0.8489059805870056,
+      "learning_rate": 9.954752133672569e-06,
+      "loss": 0.7748,
+      "step": 1653
+    },
+    {
+      "epoch": 0.09103417909626287,
+      "grad_norm": 0.8914602994918823,
+      "learning_rate": 9.954693931541638e-06,
+      "loss": 0.8657,
+      "step": 1654
+    },
+    {
+      "epoch": 0.09108921789861853,
+      "grad_norm": 0.9031614661216736,
+      "learning_rate": 9.954635692172555e-06,
+      "loss": 0.7409,
+      "step": 1655
+    },
+    {
+      "epoch": 0.09114425670097419,
+      "grad_norm": 0.8680000305175781,
+      "learning_rate": 9.954577415565756e-06,
+      "loss": 0.8535,
+      "step": 1656
+    },
+    {
+      "epoch": 0.09119929550332985,
+      "grad_norm": 0.830596923828125,
+      "learning_rate": 9.954519101721679e-06,
+      "loss": 0.8601,
+      "step": 1657
+    },
+    {
+      "epoch": 0.09125433430568551,
+      "grad_norm": 0.9041332602500916,
+      "learning_rate": 9.954460750640762e-06,
+      "loss": 0.9104,
+      "step": 1658
+    },
+    {
+      "epoch": 0.09130937310804117,
+      "grad_norm": 0.7786296606063843,
+      "learning_rate": 9.954402362323445e-06,
+      "loss": 0.7671,
+      "step": 1659
+    },
+    {
+      "epoch": 0.09136441191039683,
+      "grad_norm": 1.0363564491271973,
+      "learning_rate": 9.954343936770165e-06,
+      "loss": 0.9339,
+      "step": 1660
+    },
+    {
+      "epoch": 0.0914194507127525,
+      "grad_norm": 0.8049986958503723,
+      "learning_rate": 9.954285473981363e-06,
+      "loss": 0.8125,
+      "step": 1661
+    },
+    {
+      "epoch": 0.09147448951510816,
+      "grad_norm": 0.7842011451721191,
+      "learning_rate": 9.954226973957477e-06,
+      "loss": 0.7153,
+      "step": 1662
+    },
+    {
+      "epoch": 0.09152952831746382,
+      "grad_norm": 0.8929729461669922,
+      "learning_rate": 9.954168436698948e-06,
+      "loss": 0.9563,
+      "step": 1663
+    },
+    {
+      "epoch": 0.09158456711981948,
+      "grad_norm": 0.8850226402282715,
+      "learning_rate": 9.954109862206216e-06,
+      "loss": 0.8257,
+      "step": 1664
+    },
+    {
+      "epoch": 0.09163960592217513,
+      "grad_norm": 0.8673348426818848,
+      "learning_rate": 9.954051250479719e-06,
+      "loss": 0.9489,
+      "step": 1665
+    },
+    {
+      "epoch": 0.09169464472453079,
+      "grad_norm": 0.8726119995117188,
+      "learning_rate": 9.9539926015199e-06,
+      "loss": 0.8222,
+      "step": 1666
+    },
+    {
+      "epoch": 0.09174968352688645,
+      "grad_norm": 0.7609312534332275,
+      "learning_rate": 9.953933915327196e-06,
+      "loss": 0.7749,
+      "step": 1667
+    },
+    {
+      "epoch": 0.09180472232924211,
+      "grad_norm": 0.857404887676239,
+      "learning_rate": 9.953875191902055e-06,
+      "loss": 0.8496,
+      "step": 1668
+    },
+    {
+      "epoch": 0.09185976113159777,
+      "grad_norm": 0.7835526466369629,
+      "learning_rate": 9.953816431244909e-06,
+      "loss": 0.7258,
+      "step": 1669
+    },
+    {
+      "epoch": 0.09191479993395343,
+      "grad_norm": 0.944984495639801,
+      "learning_rate": 9.95375763335621e-06,
+      "loss": 0.902,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0919698387363091,
+      "grad_norm": 0.9038936495780945,
+      "learning_rate": 9.953698798236391e-06,
+      "loss": 0.7559,
+      "step": 1671
+    },
+    {
+      "epoch": 0.09202487753866476,
+      "grad_norm": 0.8450848460197449,
+      "learning_rate": 9.953639925885898e-06,
+      "loss": 0.8338,
+      "step": 1672
+    },
+    {
+      "epoch": 0.09207991634102042,
+      "grad_norm": 0.827419102191925,
+      "learning_rate": 9.953581016305175e-06,
+      "loss": 0.8167,
+      "step": 1673
+    },
+    {
+      "epoch": 0.09213495514337608,
+      "grad_norm": 0.8517075777053833,
+      "learning_rate": 9.953522069494663e-06,
+      "loss": 0.8681,
+      "step": 1674
+    },
+    {
+      "epoch": 0.09218999394573174,
+      "grad_norm": 0.9504323601722717,
+      "learning_rate": 9.953463085454804e-06,
+      "loss": 0.8688,
+      "step": 1675
+    },
+    {
+      "epoch": 0.0922450327480874,
+      "grad_norm": 0.8905719518661499,
+      "learning_rate": 9.953404064186044e-06,
+      "loss": 0.8818,
+      "step": 1676
+    },
+    {
+      "epoch": 0.09230007155044306,
+      "grad_norm": 0.9223340153694153,
+      "learning_rate": 9.953345005688822e-06,
+      "loss": 0.8752,
+      "step": 1677
+    },
+    {
+      "epoch": 0.09235511035279872,
+      "grad_norm": 1.0500547885894775,
+      "learning_rate": 9.953285909963588e-06,
+      "loss": 0.7816,
+      "step": 1678
+    },
+    {
+      "epoch": 0.09241014915515439,
+      "grad_norm": 0.8407441973686218,
+      "learning_rate": 9.953226777010781e-06,
+      "loss": 0.745,
+      "step": 1679
+    },
+    {
+      "epoch": 0.09246518795751005,
+      "grad_norm": 0.7997288107872009,
+      "learning_rate": 9.953167606830847e-06,
+      "loss": 0.8171,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09252022675986571,
+      "grad_norm": 0.9752318859100342,
+      "learning_rate": 9.953108399424234e-06,
+      "loss": 0.8719,
+      "step": 1681
+    },
+    {
+      "epoch": 0.09257526556222137,
+      "grad_norm": 0.8524298667907715,
+      "learning_rate": 9.953049154791382e-06,
+      "loss": 0.8257,
+      "step": 1682
+    },
+    {
+      "epoch": 0.09263030436457703,
+      "grad_norm": 0.9460529088973999,
+      "learning_rate": 9.952989872932739e-06,
+      "loss": 0.7278,
+      "step": 1683
+    },
+    {
+      "epoch": 0.0926853431669327,
+      "grad_norm": 0.8959575891494751,
+      "learning_rate": 9.95293055384875e-06,
+      "loss": 0.903,
+      "step": 1684
+    },
+    {
+      "epoch": 0.09274038196928835,
+      "grad_norm": 0.8764386177062988,
+      "learning_rate": 9.95287119753986e-06,
+      "loss": 0.7958,
+      "step": 1685
+    },
+    {
+      "epoch": 0.092795420771644,
+      "grad_norm": 0.9611337184906006,
+      "learning_rate": 9.952811804006517e-06,
+      "loss": 0.8726,
+      "step": 1686
+    },
+    {
+      "epoch": 0.09285045957399966,
+      "grad_norm": 0.8155574202537537,
+      "learning_rate": 9.952752373249165e-06,
+      "loss": 0.7882,
+      "step": 1687
+    },
+    {
+      "epoch": 0.09290549837635532,
+      "grad_norm": 0.8789697289466858,
+      "learning_rate": 9.952692905268253e-06,
+      "loss": 0.8642,
+      "step": 1688
+    },
+    {
+      "epoch": 0.09296053717871099,
+      "grad_norm": 0.7910027503967285,
+      "learning_rate": 9.952633400064227e-06,
+      "loss": 0.7852,
+      "step": 1689
+    },
+    {
+      "epoch": 0.09301557598106665,
+      "grad_norm": 0.815819501876831,
+      "learning_rate": 9.952573857637533e-06,
+      "loss": 0.8606,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09307061478342231,
+      "grad_norm": 0.9840701818466187,
+      "learning_rate": 9.95251427798862e-06,
+      "loss": 0.9349,
+      "step": 1691
+    },
+    {
+      "epoch": 0.09312565358577797,
+      "grad_norm": 0.8715788722038269,
+      "learning_rate": 9.952454661117936e-06,
+      "loss": 0.813,
+      "step": 1692
+    },
+    {
+      "epoch": 0.09318069238813363,
+      "grad_norm": 0.8287779092788696,
+      "learning_rate": 9.952395007025926e-06,
+      "loss": 0.8346,
+      "step": 1693
+    },
+    {
+      "epoch": 0.0932357311904893,
+      "grad_norm": 0.9375059008598328,
+      "learning_rate": 9.952335315713044e-06,
+      "loss": 0.8868,
+      "step": 1694
+    },
+    {
+      "epoch": 0.09329076999284495,
+      "grad_norm": 0.9063667058944702,
+      "learning_rate": 9.952275587179734e-06,
+      "loss": 0.9562,
+      "step": 1695
+    },
+    {
+      "epoch": 0.09334580879520062,
+      "grad_norm": 0.816643476486206,
+      "learning_rate": 9.952215821426447e-06,
+      "loss": 0.7456,
+      "step": 1696
+    },
+    {
+      "epoch": 0.09340084759755628,
+      "grad_norm": 0.9004347324371338,
+      "learning_rate": 9.95215601845363e-06,
+      "loss": 0.8545,
+      "step": 1697
+    },
+    {
+      "epoch": 0.09345588639991194,
+      "grad_norm": 0.919195830821991,
+      "learning_rate": 9.952096178261736e-06,
+      "loss": 0.9347,
+      "step": 1698
+    },
+    {
+      "epoch": 0.0935109252022676,
+      "grad_norm": 0.8313261866569519,
+      "learning_rate": 9.952036300851211e-06,
+      "loss": 0.9169,
+      "step": 1699
+    },
+    {
+      "epoch": 0.09356596400462326,
+      "grad_norm": 0.8674910664558411,
+      "learning_rate": 9.951976386222507e-06,
+      "loss": 0.7621,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09362100280697892,
+      "grad_norm": 0.8931052684783936,
+      "learning_rate": 9.951916434376074e-06,
+      "loss": 0.8702,
+      "step": 1701
+    },
+    {
+      "epoch": 0.09367604160933458,
+      "grad_norm": 0.8748393058776855,
+      "learning_rate": 9.951856445312364e-06,
+      "loss": 0.7446,
+      "step": 1702
+    },
+    {
+      "epoch": 0.09373108041169025,
+      "grad_norm": 1.005459189414978,
+      "learning_rate": 9.951796419031825e-06,
+      "loss": 0.9843,
+      "step": 1703
+    },
+    {
+      "epoch": 0.09378611921404591,
+      "grad_norm": 1.0155184268951416,
+      "learning_rate": 9.95173635553491e-06,
+      "loss": 0.8868,
+      "step": 1704
+    },
+    {
+      "epoch": 0.09384115801640157,
+      "grad_norm": 2.1387271881103516,
+      "learning_rate": 9.951676254822072e-06,
+      "loss": 0.8691,
+      "step": 1705
+    },
+    {
+      "epoch": 0.09389619681875723,
+      "grad_norm": 0.9768403768539429,
+      "learning_rate": 9.951616116893757e-06,
+      "loss": 0.8409,
+      "step": 1706
+    },
+    {
+      "epoch": 0.09395123562111289,
+      "grad_norm": 0.7994607090950012,
+      "learning_rate": 9.951555941750424e-06,
+      "loss": 0.7836,
+      "step": 1707
+    },
+    {
+      "epoch": 0.09400627442346854,
+      "grad_norm": 0.8460201025009155,
+      "learning_rate": 9.95149572939252e-06,
+      "loss": 0.8216,
+      "step": 1708
+    },
+    {
+      "epoch": 0.0940613132258242,
+      "grad_norm": 0.8904135227203369,
+      "learning_rate": 9.951435479820499e-06,
+      "loss": 0.9053,
+      "step": 1709
+    },
+    {
+      "epoch": 0.09411635202817986,
+      "grad_norm": 0.9084494113922119,
+      "learning_rate": 9.951375193034815e-06,
+      "loss": 0.9308,
+      "step": 1710
+    },
+    {
+      "epoch": 0.09417139083053552,
+      "grad_norm": 1.0826482772827148,
+      "learning_rate": 9.951314869035921e-06,
+      "loss": 0.8468,
+      "step": 1711
+    },
+    {
+      "epoch": 0.09422642963289118,
+      "grad_norm": 0.8068915009498596,
+      "learning_rate": 9.95125450782427e-06,
+      "loss": 0.8253,
+      "step": 1712
+    },
+    {
+      "epoch": 0.09428146843524685,
+      "grad_norm": 0.8445400595664978,
+      "learning_rate": 9.951194109400316e-06,
+      "loss": 0.8386,
+      "step": 1713
+    },
+    {
+      "epoch": 0.09433650723760251,
+      "grad_norm": 0.8180645704269409,
+      "learning_rate": 9.951133673764513e-06,
+      "loss": 0.7907,
+      "step": 1714
+    },
+    {
+      "epoch": 0.09439154603995817,
+      "grad_norm": 0.8111036419868469,
+      "learning_rate": 9.951073200917311e-06,
+      "loss": 0.7918,
+      "step": 1715
+    },
+    {
+      "epoch": 0.09444658484231383,
+      "grad_norm": 0.862042248249054,
+      "learning_rate": 9.951012690859172e-06,
+      "loss": 0.783,
+      "step": 1716
+    },
+    {
+      "epoch": 0.09450162364466949,
+      "grad_norm": 0.8189615607261658,
+      "learning_rate": 9.950952143590544e-06,
+      "loss": 0.8192,
+      "step": 1717
+    },
+    {
+      "epoch": 0.09455666244702515,
+      "grad_norm": 0.9714062809944153,
+      "learning_rate": 9.950891559111887e-06,
+      "loss": 0.774,
+      "step": 1718
+    },
+    {
+      "epoch": 0.09461170124938081,
+      "grad_norm": 0.9691846370697021,
+      "learning_rate": 9.950830937423655e-06,
+      "loss": 0.8347,
+      "step": 1719
+    },
+    {
+      "epoch": 0.09466674005173648,
+      "grad_norm": 0.8488250970840454,
+      "learning_rate": 9.950770278526301e-06,
+      "loss": 0.8228,
+      "step": 1720
+    },
+    {
+      "epoch": 0.09472177885409214,
+      "grad_norm": 0.8638359904289246,
+      "learning_rate": 9.950709582420282e-06,
+      "loss": 0.8973,
+      "step": 1721
+    },
+    {
+      "epoch": 0.0947768176564478,
+      "grad_norm": 1.0148643255233765,
+      "learning_rate": 9.950648849106058e-06,
+      "loss": 0.9638,
+      "step": 1722
+    },
+    {
+      "epoch": 0.09483185645880346,
+      "grad_norm": 0.8870131969451904,
+      "learning_rate": 9.95058807858408e-06,
+      "loss": 0.8259,
+      "step": 1723
+    },
+    {
+      "epoch": 0.09488689526115912,
+      "grad_norm": 0.9134769439697266,
+      "learning_rate": 9.950527270854807e-06,
+      "loss": 0.865,
+      "step": 1724
+    },
+    {
+      "epoch": 0.09494193406351478,
+      "grad_norm": 0.7221654653549194,
+      "learning_rate": 9.950466425918697e-06,
+      "loss": 0.7593,
+      "step": 1725
+    },
+    {
+      "epoch": 0.09499697286587044,
+      "grad_norm": 0.9386674165725708,
+      "learning_rate": 9.950405543776207e-06,
+      "loss": 0.9508,
+      "step": 1726
+    },
+    {
+      "epoch": 0.0950520116682261,
+      "grad_norm": 0.7850627899169922,
+      "learning_rate": 9.950344624427795e-06,
+      "loss": 0.7999,
+      "step": 1727
+    },
+    {
+      "epoch": 0.09510705047058177,
+      "grad_norm": 0.921198308467865,
+      "learning_rate": 9.950283667873916e-06,
+      "loss": 0.8249,
+      "step": 1728
+    },
+    {
+      "epoch": 0.09516208927293741,
+      "grad_norm": 0.9503389000892639,
+      "learning_rate": 9.95022267411503e-06,
+      "loss": 0.901,
+      "step": 1729
+    },
+    {
+      "epoch": 0.09521712807529308,
+      "grad_norm": 0.7977343201637268,
+      "learning_rate": 9.950161643151597e-06,
+      "loss": 0.838,
+      "step": 1730
+    },
+    {
+      "epoch": 0.09527216687764874,
+      "grad_norm": 0.9056238532066345,
+      "learning_rate": 9.950100574984072e-06,
+      "loss": 0.9756,
+      "step": 1731
+    },
+    {
+      "epoch": 0.0953272056800044,
+      "grad_norm": 0.8092935681343079,
+      "learning_rate": 9.950039469612918e-06,
+      "loss": 0.8812,
+      "step": 1732
+    },
+    {
+      "epoch": 0.09538224448236006,
+      "grad_norm": 0.823693573474884,
+      "learning_rate": 9.949978327038592e-06,
+      "loss": 0.7914,
+      "step": 1733
+    },
+    {
+      "epoch": 0.09543728328471572,
+      "grad_norm": 0.9114876389503479,
+      "learning_rate": 9.949917147261554e-06,
+      "loss": 0.7944,
+      "step": 1734
+    },
+    {
+      "epoch": 0.09549232208707138,
+      "grad_norm": 1.0084123611450195,
+      "learning_rate": 9.949855930282262e-06,
+      "loss": 0.8544,
+      "step": 1735
+    },
+    {
+      "epoch": 0.09554736088942704,
+      "grad_norm": 0.842462956905365,
+      "learning_rate": 9.949794676101181e-06,
+      "loss": 0.7056,
+      "step": 1736
+    },
+    {
+      "epoch": 0.0956023996917827,
+      "grad_norm": 1.00497305393219,
+      "learning_rate": 9.949733384718766e-06,
+      "loss": 0.8372,
+      "step": 1737
+    },
+    {
+      "epoch": 0.09565743849413837,
+      "grad_norm": 1.0166410207748413,
+      "learning_rate": 9.94967205613548e-06,
+      "loss": 0.9316,
+      "step": 1738
+    },
+    {
+      "epoch": 0.09571247729649403,
+      "grad_norm": 0.8520192503929138,
+      "learning_rate": 9.949610690351784e-06,
+      "loss": 0.786,
+      "step": 1739
+    },
+    {
+      "epoch": 0.09576751609884969,
+      "grad_norm": 0.8003227114677429,
+      "learning_rate": 9.949549287368139e-06,
+      "loss": 0.8003,
+      "step": 1740
+    },
+    {
+      "epoch": 0.09582255490120535,
+      "grad_norm": 0.8657151460647583,
+      "learning_rate": 9.949487847185006e-06,
+      "loss": 0.8407,
+      "step": 1741
+    },
+    {
+      "epoch": 0.09587759370356101,
+      "grad_norm": 1.1119858026504517,
+      "learning_rate": 9.949426369802848e-06,
+      "loss": 0.8594,
+      "step": 1742
+    },
+    {
+      "epoch": 0.09593263250591667,
+      "grad_norm": 0.8968474864959717,
+      "learning_rate": 9.949364855222126e-06,
+      "loss": 0.8254,
+      "step": 1743
+    },
+    {
+      "epoch": 0.09598767130827233,
+      "grad_norm": 0.8740531206130981,
+      "learning_rate": 9.949303303443304e-06,
+      "loss": 0.8748,
+      "step": 1744
+    },
+    {
+      "epoch": 0.096042710110628,
+      "grad_norm": 0.8833459615707397,
+      "learning_rate": 9.94924171446684e-06,
+      "loss": 0.838,
+      "step": 1745
+    },
+    {
+      "epoch": 0.09609774891298366,
+      "grad_norm": 0.8783486485481262,
+      "learning_rate": 9.949180088293201e-06,
+      "loss": 0.7972,
+      "step": 1746
+    },
+    {
+      "epoch": 0.09615278771533932,
+      "grad_norm": 0.9197877049446106,
+      "learning_rate": 9.949118424922852e-06,
+      "loss": 0.8669,
+      "step": 1747
+    },
+    {
+      "epoch": 0.09620782651769498,
+      "grad_norm": 0.9771283864974976,
+      "learning_rate": 9.949056724356251e-06,
+      "loss": 0.8461,
+      "step": 1748
+    },
+    {
+      "epoch": 0.09626286532005064,
+      "grad_norm": 0.8325022459030151,
+      "learning_rate": 9.948994986593864e-06,
+      "loss": 0.8482,
+      "step": 1749
+    },
+    {
+      "epoch": 0.0963179041224063,
+      "grad_norm": 0.9732363224029541,
+      "learning_rate": 9.948933211636158e-06,
+      "loss": 0.8825,
+      "step": 1750
+    },
+    {
+      "epoch": 0.09637294292476195,
+      "grad_norm": 0.8229798078536987,
+      "learning_rate": 9.948871399483592e-06,
+      "loss": 0.8079,
+      "step": 1751
+    },
+    {
+      "epoch": 0.09642798172711761,
+      "grad_norm": 0.8861554265022278,
+      "learning_rate": 9.948809550136635e-06,
+      "loss": 0.8323,
+      "step": 1752
+    },
+    {
+      "epoch": 0.09648302052947327,
+      "grad_norm": 1.0618904829025269,
+      "learning_rate": 9.94874766359575e-06,
+      "loss": 0.8519,
+      "step": 1753
+    },
+    {
+      "epoch": 0.09653805933182893,
+      "grad_norm": 0.8494864702224731,
+      "learning_rate": 9.948685739861403e-06,
+      "loss": 0.961,
+      "step": 1754
+    },
+    {
+      "epoch": 0.0965930981341846,
+      "grad_norm": 0.8872213959693909,
+      "learning_rate": 9.948623778934058e-06,
+      "loss": 0.9367,
+      "step": 1755
+    },
+    {
+      "epoch": 0.09664813693654026,
+      "grad_norm": 0.8441230058670044,
+      "learning_rate": 9.948561780814181e-06,
+      "loss": 0.7654,
+      "step": 1756
+    },
+    {
+      "epoch": 0.09670317573889592,
+      "grad_norm": 0.8072223663330078,
+      "learning_rate": 9.948499745502239e-06,
+      "loss": 0.7894,
+      "step": 1757
+    },
+    {
+      "epoch": 0.09675821454125158,
+      "grad_norm": 0.8285261392593384,
+      "learning_rate": 9.948437672998696e-06,
+      "loss": 0.8351,
+      "step": 1758
+    },
+    {
+      "epoch": 0.09681325334360724,
+      "grad_norm": 0.9272124767303467,
+      "learning_rate": 9.94837556330402e-06,
+      "loss": 0.8708,
+      "step": 1759
+    },
+    {
+      "epoch": 0.0968682921459629,
+      "grad_norm": 0.8689375519752502,
+      "learning_rate": 9.94831341641868e-06,
+      "loss": 0.8478,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09692333094831856,
+      "grad_norm": 1.040784239768982,
+      "learning_rate": 9.94825123234314e-06,
+      "loss": 0.8915,
+      "step": 1761
+    },
+    {
+      "epoch": 0.09697836975067423,
+      "grad_norm": 0.7819718718528748,
+      "learning_rate": 9.948189011077867e-06,
+      "loss": 0.7728,
+      "step": 1762
+    },
+    {
+      "epoch": 0.09703340855302989,
+      "grad_norm": 0.7959379553794861,
+      "learning_rate": 9.948126752623331e-06,
+      "loss": 0.8248,
+      "step": 1763
+    },
+    {
+      "epoch": 0.09708844735538555,
+      "grad_norm": 0.8844753503799438,
+      "learning_rate": 9.94806445698e-06,
+      "loss": 0.7742,
+      "step": 1764
+    },
+    {
+      "epoch": 0.09714348615774121,
+      "grad_norm": 0.9168505668640137,
+      "learning_rate": 9.948002124148339e-06,
+      "loss": 0.9145,
+      "step": 1765
+    },
+    {
+      "epoch": 0.09719852496009687,
+      "grad_norm": 0.7199662923812866,
+      "learning_rate": 9.947939754128819e-06,
+      "loss": 0.6652,
+      "step": 1766
+    },
+    {
+      "epoch": 0.09725356376245253,
+      "grad_norm": 0.866470992565155,
+      "learning_rate": 9.947877346921909e-06,
+      "loss": 0.8293,
+      "step": 1767
+    },
+    {
+      "epoch": 0.0973086025648082,
+      "grad_norm": 0.9124754667282104,
+      "learning_rate": 9.947814902528078e-06,
+      "loss": 0.8599,
+      "step": 1768
+    },
+    {
+      "epoch": 0.09736364136716386,
+      "grad_norm": 0.9169870615005493,
+      "learning_rate": 9.947752420947792e-06,
+      "loss": 0.8382,
+      "step": 1769
+    },
+    {
+      "epoch": 0.09741868016951952,
+      "grad_norm": 1.0147640705108643,
+      "learning_rate": 9.947689902181526e-06,
+      "loss": 0.8425,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09747371897187518,
+      "grad_norm": 0.778575599193573,
+      "learning_rate": 9.947627346229745e-06,
+      "loss": 0.6979,
+      "step": 1771
+    },
+    {
+      "epoch": 0.09752875777423083,
+      "grad_norm": 0.815101146697998,
+      "learning_rate": 9.947564753092922e-06,
+      "loss": 0.8617,
+      "step": 1772
+    },
+    {
+      "epoch": 0.09758379657658649,
+      "grad_norm": 0.9556358456611633,
+      "learning_rate": 9.947502122771527e-06,
+      "loss": 0.9009,
+      "step": 1773
+    },
+    {
+      "epoch": 0.09763883537894215,
+      "grad_norm": 0.8603761196136475,
+      "learning_rate": 9.94743945526603e-06,
+      "loss": 0.9443,
+      "step": 1774
+    },
+    {
+      "epoch": 0.09769387418129781,
+      "grad_norm": 0.8621761798858643,
+      "learning_rate": 9.947376750576903e-06,
+      "loss": 0.7537,
+      "step": 1775
+    },
+    {
+      "epoch": 0.09774891298365347,
+      "grad_norm": 0.7399948835372925,
+      "learning_rate": 9.947314008704616e-06,
+      "loss": 0.7477,
+      "step": 1776
+    },
+    {
+      "epoch": 0.09780395178600913,
+      "grad_norm": 0.8855582475662231,
+      "learning_rate": 9.947251229649641e-06,
+      "loss": 0.8745,
+      "step": 1777
+    },
+    {
+      "epoch": 0.0978589905883648,
+      "grad_norm": 0.8718472719192505,
+      "learning_rate": 9.947188413412452e-06,
+      "loss": 0.9672,
+      "step": 1778
+    },
+    {
+      "epoch": 0.09791402939072046,
+      "grad_norm": 0.8598514795303345,
+      "learning_rate": 9.947125559993517e-06,
+      "loss": 0.8278,
+      "step": 1779
+    },
+    {
+      "epoch": 0.09796906819307612,
+      "grad_norm": 1.0373798608779907,
+      "learning_rate": 9.947062669393312e-06,
+      "loss": 0.8123,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09802410699543178,
+      "grad_norm": 1.0198705196380615,
+      "learning_rate": 9.946999741612306e-06,
+      "loss": 0.9039,
+      "step": 1781
+    },
+    {
+      "epoch": 0.09807914579778744,
+      "grad_norm": 0.8770025968551636,
+      "learning_rate": 9.946936776650977e-06,
+      "loss": 0.8326,
+      "step": 1782
+    },
+    {
+      "epoch": 0.0981341846001431,
+      "grad_norm": 0.7970215678215027,
+      "learning_rate": 9.946873774509794e-06,
+      "loss": 0.848,
+      "step": 1783
+    },
+    {
+      "epoch": 0.09818922340249876,
+      "grad_norm": 0.90342777967453,
+      "learning_rate": 9.946810735189231e-06,
+      "loss": 0.7993,
+      "step": 1784
+    },
+    {
+      "epoch": 0.09824426220485442,
+      "grad_norm": 1.2095681428909302,
+      "learning_rate": 9.946747658689763e-06,
+      "loss": 0.8544,
+      "step": 1785
+    },
+    {
+      "epoch": 0.09829930100721009,
+      "grad_norm": 0.8500953316688538,
+      "learning_rate": 9.946684545011866e-06,
+      "loss": 0.8398,
+      "step": 1786
+    },
+    {
+      "epoch": 0.09835433980956575,
+      "grad_norm": 0.8570724725723267,
+      "learning_rate": 9.946621394156011e-06,
+      "loss": 0.9255,
+      "step": 1787
+    },
+    {
+      "epoch": 0.09840937861192141,
+      "grad_norm": 0.8314846158027649,
+      "learning_rate": 9.946558206122672e-06,
+      "loss": 0.8398,
+      "step": 1788
+    },
+    {
+      "epoch": 0.09846441741427707,
+      "grad_norm": 0.8894716501235962,
+      "learning_rate": 9.946494980912326e-06,
+      "loss": 0.8612,
+      "step": 1789
+    },
+    {
+      "epoch": 0.09851945621663273,
+      "grad_norm": 0.9555756449699402,
+      "learning_rate": 9.94643171852545e-06,
+      "loss": 0.9551,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09857449501898839,
+      "grad_norm": 0.9556692838668823,
+      "learning_rate": 9.946368418962515e-06,
+      "loss": 0.8175,
+      "step": 1791
+    },
+    {
+      "epoch": 0.09862953382134405,
+      "grad_norm": 0.7288535833358765,
+      "learning_rate": 9.946305082224e-06,
+      "loss": 0.6162,
+      "step": 1792
+    },
+    {
+      "epoch": 0.09868457262369972,
+      "grad_norm": 0.95478355884552,
+      "learning_rate": 9.94624170831038e-06,
+      "loss": 0.9089,
+      "step": 1793
+    },
+    {
+      "epoch": 0.09873961142605536,
+      "grad_norm": 0.9080137610435486,
+      "learning_rate": 9.946178297222133e-06,
+      "loss": 0.9443,
+      "step": 1794
+    },
+    {
+      "epoch": 0.09879465022841102,
+      "grad_norm": 0.8060124516487122,
+      "learning_rate": 9.946114848959732e-06,
+      "loss": 0.7412,
+      "step": 1795
+    },
+    {
+      "epoch": 0.09884968903076669,
+      "grad_norm": 0.8487932085990906,
+      "learning_rate": 9.946051363523655e-06,
+      "loss": 0.7098,
+      "step": 1796
+    },
+    {
+      "epoch": 0.09890472783312235,
+      "grad_norm": 0.8982037901878357,
+      "learning_rate": 9.945987840914381e-06,
+      "loss": 0.8304,
+      "step": 1797
+    },
+    {
+      "epoch": 0.09895976663547801,
+      "grad_norm": 0.8124602437019348,
+      "learning_rate": 9.945924281132386e-06,
+      "loss": 0.8441,
+      "step": 1798
+    },
+    {
+      "epoch": 0.09901480543783367,
+      "grad_norm": 0.8081663250923157,
+      "learning_rate": 9.945860684178147e-06,
+      "loss": 0.732,
+      "step": 1799
+    },
+    {
+      "epoch": 0.09906984424018933,
+      "grad_norm": 0.7662907242774963,
+      "learning_rate": 9.945797050052147e-06,
+      "loss": 0.7538,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09912488304254499,
+      "grad_norm": 0.8418399095535278,
+      "learning_rate": 9.945733378754856e-06,
+      "loss": 0.8488,
+      "step": 1801
+    },
+    {
+      "epoch": 0.09917992184490065,
+      "grad_norm": 0.7298988699913025,
+      "learning_rate": 9.94566967028676e-06,
+      "loss": 0.7822,
+      "step": 1802
+    },
+    {
+      "epoch": 0.09923496064725632,
+      "grad_norm": 0.7788695693016052,
+      "learning_rate": 9.945605924648332e-06,
+      "loss": 0.8037,
+      "step": 1803
+    },
+    {
+      "epoch": 0.09928999944961198,
+      "grad_norm": 0.939297080039978,
+      "learning_rate": 9.945542141840054e-06,
+      "loss": 0.8654,
+      "step": 1804
+    },
+    {
+      "epoch": 0.09934503825196764,
+      "grad_norm": 0.9274358749389648,
+      "learning_rate": 9.945478321862406e-06,
+      "loss": 0.7712,
+      "step": 1805
+    },
+    {
+      "epoch": 0.0994000770543233,
+      "grad_norm": 0.816561222076416,
+      "learning_rate": 9.945414464715866e-06,
+      "loss": 0.7676,
+      "step": 1806
+    },
+    {
+      "epoch": 0.09945511585667896,
+      "grad_norm": 0.867915153503418,
+      "learning_rate": 9.945350570400916e-06,
+      "loss": 0.8343,
+      "step": 1807
+    },
+    {
+      "epoch": 0.09951015465903462,
+      "grad_norm": 0.8446162939071655,
+      "learning_rate": 9.945286638918034e-06,
+      "loss": 0.8128,
+      "step": 1808
+    },
+    {
+      "epoch": 0.09956519346139028,
+      "grad_norm": 0.8372986316680908,
+      "learning_rate": 9.945222670267703e-06,
+      "loss": 0.8611,
+      "step": 1809
+    },
+    {
+      "epoch": 0.09962023226374594,
+      "grad_norm": 0.787836492061615,
+      "learning_rate": 9.945158664450399e-06,
+      "loss": 0.7286,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0996752710661016,
+      "grad_norm": 0.9293436408042908,
+      "learning_rate": 9.945094621466609e-06,
+      "loss": 0.8699,
+      "step": 1811
+    },
+    {
+      "epoch": 0.09973030986845727,
+      "grad_norm": 0.8336932063102722,
+      "learning_rate": 9.94503054131681e-06,
+      "loss": 0.8222,
+      "step": 1812
+    },
+    {
+      "epoch": 0.09978534867081293,
+      "grad_norm": 0.8310953378677368,
+      "learning_rate": 9.944966424001486e-06,
+      "loss": 0.8131,
+      "step": 1813
+    },
+    {
+      "epoch": 0.09984038747316859,
+      "grad_norm": 0.7703443169593811,
+      "learning_rate": 9.944902269521117e-06,
+      "loss": 0.8135,
+      "step": 1814
+    },
+    {
+      "epoch": 0.09989542627552424,
+      "grad_norm": 0.750990092754364,
+      "learning_rate": 9.944838077876186e-06,
+      "loss": 0.8137,
+      "step": 1815
+    },
+    {
+      "epoch": 0.0999504650778799,
+      "grad_norm": 0.8502481579780579,
+      "learning_rate": 9.944773849067178e-06,
+      "loss": 0.8973,
+      "step": 1816
+    },
+    {
+      "epoch": 0.10000550388023556,
+      "grad_norm": 0.8299791812896729,
+      "learning_rate": 9.94470958309457e-06,
+      "loss": 0.8341,
+      "step": 1817
+    },
+    {
+      "epoch": 0.10006054268259122,
+      "grad_norm": 0.8519022464752197,
+      "learning_rate": 9.94464527995885e-06,
+      "loss": 0.8529,
+      "step": 1818
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 36338,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 909,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.365029428243988e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}