diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17226 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.9973915878708834,
+  "eval_steps": 500,
+  "global_step": 2452,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016302575806977503,
+      "grad_norm": 0.7529258131980896,
+      "learning_rate": 0.0,
+      "loss": 0.4678,
+      "step": 1
+    },
+    {
+      "epoch": 0.0032605151613955006,
+      "grad_norm": 0.2298814058303833,
+      "learning_rate": 2.1533827903669654e-05,
+      "loss": 0.417,
+      "step": 2
+    },
+    {
+      "epoch": 0.00489077274209325,
+      "grad_norm": 0.20245032012462616,
+      "learning_rate": 3.413030972429927e-05,
+      "loss": 0.3776,
+      "step": 3
+    },
+    {
+      "epoch": 0.006521030322791001,
+      "grad_norm": 0.1889941245317459,
+      "learning_rate": 4.306765580733931e-05,
+      "loss": 0.3834,
+      "step": 4
+    },
+    {
+      "epoch": 0.008151287903488751,
+      "grad_norm": 0.16519580781459808,
+      "learning_rate": 5e-05,
+      "loss": 0.4055,
+      "step": 5
+    },
+    {
+      "epoch": 0.0097815454841865,
+      "grad_norm": 0.14732758700847626,
+      "learning_rate": 5.5664137627968925e-05,
+      "loss": 0.3698,
+      "step": 6
+    },
+    {
+      "epoch": 0.011411803064884252,
+      "grad_norm": 0.1339130848646164,
+      "learning_rate": 6.0453097756108376e-05,
+      "loss": 0.348,
+      "step": 7
+    },
+    {
+      "epoch": 0.013042060645582002,
+      "grad_norm": 0.1276892125606537,
+      "learning_rate": 6.460148371100896e-05,
+      "loss": 0.3469,
+      "step": 8
+    },
+    {
+      "epoch": 0.014672318226279752,
+      "grad_norm": 0.13217830657958984,
+      "learning_rate": 6.826061944859854e-05,
+      "loss": 0.351,
+      "step": 9
+    },
+    {
+      "epoch": 0.016302575806977502,
+      "grad_norm": 0.13900792598724365,
+      "learning_rate": 7.153382790366967e-05,
+      "loss": 0.33,
+      "step": 10
+    },
+    {
+      "epoch": 0.01793283338767525,
+      "grad_norm": 0.1303739994764328,
+      "learning_rate": 7.449480512024892e-05,
+      "loss": 0.3493,
+      "step": 11
+    },
+    {
+      "epoch": 0.019563090968373,
+      "grad_norm": 0.14218388497829437,
+      "learning_rate": 7.719796553163858e-05,
+      "loss": 0.3731,
+      "step": 12
+    },
+    {
+      "epoch": 0.021193348549070755,
+      "grad_norm": 0.11100596189498901,
+      "learning_rate": 7.968463205835412e-05,
+      "loss": 0.3385,
+      "step": 13
+    },
+    {
+      "epoch": 0.022823606129768505,
+      "grad_norm": 0.11745980381965637,
+      "learning_rate": 8.198692565977803e-05,
+      "loss": 0.3337,
+      "step": 14
+    },
+    {
+      "epoch": 0.024453863710466255,
+      "grad_norm": 0.11711365729570389,
+      "learning_rate": 8.413030972429928e-05,
+      "loss": 0.3117,
+      "step": 15
+    },
+    {
+      "epoch": 0.026084121291164004,
+      "grad_norm": 0.15356451272964478,
+      "learning_rate": 8.613531161467861e-05,
+      "loss": 0.3327,
+      "step": 16
+    },
+    {
+      "epoch": 0.027714378871861754,
+      "grad_norm": 0.14109951257705688,
+      "learning_rate": 8.80187213861294e-05,
+      "loss": 0.2942,
+      "step": 17
+    },
+    {
+      "epoch": 0.029344636452559504,
+      "grad_norm": 0.25882235169410706,
+      "learning_rate": 8.979444735226819e-05,
+      "loss": 0.316,
+      "step": 18
+    },
+    {
+      "epoch": 0.030974894033257254,
+      "grad_norm": 0.157696932554245,
+      "learning_rate": 9.147414002175752e-05,
+      "loss": 0.3284,
+      "step": 19
+    },
+    {
+      "epoch": 0.032605151613955004,
+      "grad_norm": 0.13975587487220764,
+      "learning_rate": 9.306765580733931e-05,
+      "loss": 0.3245,
+      "step": 20
+    },
+    {
+      "epoch": 0.034235409194652754,
+      "grad_norm": 0.8783023953437805,
+      "learning_rate": 9.458340748040766e-05,
+      "loss": 0.3181,
+      "step": 21
+    },
+    {
+      "epoch": 0.0358656667753505,
+      "grad_norm": 0.11985931545495987,
+      "learning_rate": 9.602863302391859e-05,
+      "loss": 0.3098,
+      "step": 22
+    },
+    {
+      "epoch": 0.03749592435604825,
+      "grad_norm": 0.14401742815971375,
+      "learning_rate": 9.740960467331899e-05,
+      "loss": 0.3132,
+      "step": 23
+    },
+    {
+      "epoch": 0.039126181936746,
+      "grad_norm": 0.17440500855445862,
+      "learning_rate": 9.873179343530825e-05,
+      "loss": 0.3074,
+      "step": 24
+    },
+    {
+      "epoch": 0.04075643951744375,
+      "grad_norm": 0.12916715443134308,
+      "learning_rate": 0.0001,
+      "loss": 0.2955,
+      "step": 25
+    },
+    {
+      "epoch": 0.04238669709814151,
+      "grad_norm": 0.5290112495422363,
+      "learning_rate": 0.0001,
+      "loss": 0.3143,
+      "step": 26
+    },
+    {
+      "epoch": 0.04401695467883926,
+      "grad_norm": 0.1266779601573944,
+      "learning_rate": 0.0001,
+      "loss": 0.2958,
+      "step": 27
+    },
+    {
+      "epoch": 0.04564721225953701,
+      "grad_norm": 0.13447235524654388,
+      "learning_rate": 0.0001,
+      "loss": 0.3211,
+      "step": 28
+    },
+    {
+      "epoch": 0.04727746984023476,
+      "grad_norm": 0.12066753953695297,
+      "learning_rate": 0.0001,
+      "loss": 0.2894,
+      "step": 29
+    },
+    {
+      "epoch": 0.04890772742093251,
+      "grad_norm": 0.11640455573797226,
+      "learning_rate": 0.0001,
+      "loss": 0.2891,
+      "step": 30
+    },
+    {
+      "epoch": 0.05053798500163026,
+      "grad_norm": 0.1315324753522873,
+      "learning_rate": 0.0001,
+      "loss": 0.292,
+      "step": 31
+    },
+    {
+      "epoch": 0.05216824258232801,
+      "grad_norm": 0.09173787385225296,
+      "learning_rate": 0.0001,
+      "loss": 0.2786,
+      "step": 32
+    },
+    {
+      "epoch": 0.05379850016302576,
+      "grad_norm": 0.14253424108028412,
+      "learning_rate": 0.0001,
+      "loss": 0.2924,
+      "step": 33
+    },
+    {
+      "epoch": 0.05542875774372351,
+      "grad_norm": 0.1317639797925949,
+      "learning_rate": 0.0001,
+      "loss": 0.29,
+      "step": 34
+    },
+    {
+      "epoch": 0.05705901532442126,
+      "grad_norm": 0.11155499517917633,
+      "learning_rate": 0.0001,
+      "loss": 0.2575,
+      "step": 35
+    },
+    {
+      "epoch": 0.05868927290511901,
+      "grad_norm": 0.10907690972089767,
+      "learning_rate": 0.0001,
+      "loss": 0.2745,
+      "step": 36
+    },
+    {
+      "epoch": 0.06031953048581676,
+      "grad_norm": 0.11807228624820709,
+      "learning_rate": 0.0001,
+      "loss": 0.2713,
+      "step": 37
+    },
+    {
+      "epoch": 0.06194978806651451,
+      "grad_norm": 0.1130286455154419,
+      "learning_rate": 0.0001,
+      "loss": 0.2869,
+      "step": 38
+    },
+    {
+      "epoch": 0.06358004564721226,
+      "grad_norm": 0.1267092525959015,
+      "learning_rate": 0.0001,
+      "loss": 0.287,
+      "step": 39
+    },
+    {
+      "epoch": 0.06521030322791001,
+      "grad_norm": 0.13711895048618317,
+      "learning_rate": 0.0001,
+      "loss": 0.2662,
+      "step": 40
+    },
+    {
+      "epoch": 0.06684056080860776,
+      "grad_norm": 0.1149134635925293,
+      "learning_rate": 0.0001,
+      "loss": 0.2743,
+      "step": 41
+    },
+    {
+      "epoch": 0.06847081838930551,
+      "grad_norm": 0.1379522979259491,
+      "learning_rate": 0.0001,
+      "loss": 0.2915,
+      "step": 42
+    },
+    {
+      "epoch": 0.07010107597000326,
+      "grad_norm": 0.126117542386055,
+      "learning_rate": 0.0001,
+      "loss": 0.2527,
+      "step": 43
+    },
+    {
+      "epoch": 0.071731333550701,
+      "grad_norm": 0.13249348104000092,
+      "learning_rate": 0.0001,
+      "loss": 0.2951,
+      "step": 44
+    },
+    {
+      "epoch": 0.07336159113139876,
+      "grad_norm": 0.13240788877010345,
+      "learning_rate": 0.0001,
+      "loss": 0.2828,
+      "step": 45
+    },
+    {
+      "epoch": 0.0749918487120965,
+      "grad_norm": 0.1182844415307045,
+      "learning_rate": 0.0001,
+      "loss": 0.2564,
+      "step": 46
+    },
+    {
+      "epoch": 0.07662210629279426,
+      "grad_norm": 0.13393035531044006,
+      "learning_rate": 0.0001,
+      "loss": 0.2742,
+      "step": 47
+    },
+    {
+      "epoch": 0.078252363873492,
+      "grad_norm": 0.11272592842578888,
+      "learning_rate": 0.0001,
+      "loss": 0.279,
+      "step": 48
+    },
+    {
+      "epoch": 0.07988262145418976,
+      "grad_norm": 0.12920530140399933,
+      "learning_rate": 0.0001,
+      "loss": 0.2733,
+      "step": 49
+    },
+    {
+      "epoch": 0.0815128790348875,
+      "grad_norm": 0.13335202634334564,
+      "learning_rate": 0.0001,
+      "loss": 0.2647,
+      "step": 50
+    },
+    {
+      "epoch": 0.08314313661558526,
+      "grad_norm": 0.13447648286819458,
+      "learning_rate": 0.0001,
+      "loss": 0.2674,
+      "step": 51
+    },
+    {
+      "epoch": 0.08477339419628302,
+      "grad_norm": 0.13474304974079132,
+      "learning_rate": 0.0001,
+      "loss": 0.274,
+      "step": 52
+    },
+    {
+      "epoch": 0.08640365177698077,
+      "grad_norm": 0.15840347111225128,
+      "learning_rate": 0.0001,
+      "loss": 0.2871,
+      "step": 53
+    },
+    {
+      "epoch": 0.08803390935767852,
+      "grad_norm": 0.1323927342891693,
+      "learning_rate": 0.0001,
+      "loss": 0.2797,
+      "step": 54
+    },
+    {
+      "epoch": 0.08966416693837627,
+      "grad_norm": 0.15678195655345917,
+      "learning_rate": 0.0001,
+      "loss": 0.2742,
+      "step": 55
+    },
+    {
+      "epoch": 0.09129442451907402,
+      "grad_norm": 0.1214122325181961,
+      "learning_rate": 0.0001,
+      "loss": 0.258,
+      "step": 56
+    },
+    {
+      "epoch": 0.09292468209977177,
+      "grad_norm": 0.15963681042194366,
+      "learning_rate": 0.0001,
+      "loss": 0.2767,
+      "step": 57
+    },
+    {
+      "epoch": 0.09455493968046952,
+      "grad_norm": 0.14005902409553528,
+      "learning_rate": 0.0001,
+      "loss": 0.2786,
+      "step": 58
+    },
+    {
+      "epoch": 0.09618519726116727,
+      "grad_norm": 0.14053602516651154,
+      "learning_rate": 0.0001,
+      "loss": 0.2673,
+      "step": 59
+    },
+    {
+      "epoch": 0.09781545484186502,
+      "grad_norm": 0.1037927120923996,
+      "learning_rate": 0.0001,
+      "loss": 0.2532,
+      "step": 60
+    },
+    {
+      "epoch": 0.09944571242256277,
+      "grad_norm": 0.16954319179058075,
+      "learning_rate": 0.0001,
+      "loss": 0.257,
+      "step": 61
+    },
+    {
+      "epoch": 0.10107597000326052,
+      "grad_norm": 0.13848423957824707,
+      "learning_rate": 0.0001,
+      "loss": 0.2633,
+      "step": 62
+    },
+    {
+      "epoch": 0.10270622758395827,
+      "grad_norm": 0.1805400550365448,
+      "learning_rate": 0.0001,
+      "loss": 0.263,
+      "step": 63
+    },
+    {
+      "epoch": 0.10433648516465602,
+      "grad_norm": 0.17179027199745178,
+      "learning_rate": 0.0001,
+      "loss": 0.2681,
+      "step": 64
+    },
+    {
+      "epoch": 0.10596674274535377,
+      "grad_norm": 0.1394517719745636,
+      "learning_rate": 0.0001,
+      "loss": 0.27,
+      "step": 65
+    },
+    {
+      "epoch": 0.10759700032605152,
+      "grad_norm": 0.1399444192647934,
+      "learning_rate": 0.0001,
+      "loss": 0.262,
+      "step": 66
+    },
+    {
+      "epoch": 0.10922725790674927,
+      "grad_norm": 0.1266821324825287,
+      "learning_rate": 0.0001,
+      "loss": 0.2599,
+      "step": 67
+    },
+    {
+      "epoch": 0.11085751548744702,
+      "grad_norm": 0.13088282942771912,
+      "learning_rate": 0.0001,
+      "loss": 0.2636,
+      "step": 68
+    },
+    {
+      "epoch": 0.11248777306814477,
+      "grad_norm": 0.1161791980266571,
+      "learning_rate": 0.0001,
+      "loss": 0.2739,
+      "step": 69
+    },
+    {
+      "epoch": 0.11411803064884252,
+      "grad_norm": 0.1133728101849556,
+      "learning_rate": 0.0001,
+      "loss": 0.2369,
+      "step": 70
+    },
+    {
+      "epoch": 0.11574828822954027,
+      "grad_norm": 0.140262633562088,
+      "learning_rate": 0.0001,
+      "loss": 0.2752,
+      "step": 71
+    },
+    {
+      "epoch": 0.11737854581023802,
+      "grad_norm": 0.12292502820491791,
+      "learning_rate": 0.0001,
+      "loss": 0.2541,
+      "step": 72
+    },
+    {
+      "epoch": 0.11900880339093577,
+      "grad_norm": 0.14118324220180511,
+      "learning_rate": 0.0001,
+      "loss": 0.2611,
+      "step": 73
+    },
+    {
+      "epoch": 0.12063906097163352,
+      "grad_norm": 0.12177952378988266,
+      "learning_rate": 0.0001,
+      "loss": 0.2605,
+      "step": 74
+    },
+    {
+      "epoch": 0.12226931855233127,
+      "grad_norm": 0.13002616167068481,
+      "learning_rate": 0.0001,
+      "loss": 0.2609,
+      "step": 75
+    },
+    {
+      "epoch": 0.12389957613302902,
+      "grad_norm": 0.11555729806423187,
+      "learning_rate": 0.0001,
+      "loss": 0.2404,
+      "step": 76
+    },
+    {
+      "epoch": 0.12552983371372678,
+      "grad_norm": 0.1303972601890564,
+      "learning_rate": 0.0001,
+      "loss": 0.251,
+      "step": 77
+    },
+    {
+      "epoch": 0.12716009129442452,
+      "grad_norm": 0.1312289834022522,
+      "learning_rate": 0.0001,
+      "loss": 0.2635,
+      "step": 78
+    },
+    {
+      "epoch": 0.12879034887512228,
+      "grad_norm": 0.12450554966926575,
+      "learning_rate": 0.0001,
+      "loss": 0.2545,
+      "step": 79
+    },
+    {
+      "epoch": 0.13042060645582002,
+      "grad_norm": 0.1366516351699829,
+      "learning_rate": 0.0001,
+      "loss": 0.2566,
+      "step": 80
+    },
+    {
+      "epoch": 0.13205086403651778,
+      "grad_norm": 0.13325075805187225,
+      "learning_rate": 0.0001,
+      "loss": 0.2557,
+      "step": 81
+    },
+    {
+      "epoch": 0.13368112161721551,
+      "grad_norm": 0.15881724655628204,
+      "learning_rate": 0.0001,
+      "loss": 0.2467,
+      "step": 82
+    },
+    {
+      "epoch": 0.13531137919791328,
+      "grad_norm": 0.12714703381061554,
+      "learning_rate": 0.0001,
+      "loss": 0.259,
+      "step": 83
+    },
+    {
+      "epoch": 0.13694163677861101,
+      "grad_norm": 0.13727723062038422,
+      "learning_rate": 0.0001,
+      "loss": 0.2737,
+      "step": 84
+    },
+    {
+      "epoch": 0.13857189435930878,
+      "grad_norm": 0.12891346216201782,
+      "learning_rate": 0.0001,
+      "loss": 0.2747,
+      "step": 85
+    },
+    {
+      "epoch": 0.14020215194000651,
+      "grad_norm": 0.10840713977813721,
+      "learning_rate": 0.0001,
+      "loss": 0.2481,
+      "step": 86
+    },
+    {
+      "epoch": 0.14183240952070428,
+      "grad_norm": 0.12393542379140854,
+      "learning_rate": 0.0001,
+      "loss": 0.2547,
+      "step": 87
+    },
+    {
+      "epoch": 0.143462667101402,
+      "grad_norm": 0.16842670738697052,
+      "learning_rate": 0.0001,
+      "loss": 0.28,
+      "step": 88
+    },
+    {
+      "epoch": 0.14509292468209978,
+      "grad_norm": 0.13346168398857117,
+      "learning_rate": 0.0001,
+      "loss": 0.2359,
+      "step": 89
+    },
+    {
+      "epoch": 0.1467231822627975,
+      "grad_norm": 0.14012813568115234,
+      "learning_rate": 0.0001,
+      "loss": 0.242,
+      "step": 90
+    },
+    {
+      "epoch": 0.14835343984349528,
+      "grad_norm": 0.1721285581588745,
+      "learning_rate": 0.0001,
+      "loss": 0.2644,
+      "step": 91
+    },
+    {
+      "epoch": 0.149983697424193,
+      "grad_norm": 0.14937177300453186,
+      "learning_rate": 0.0001,
+      "loss": 0.2541,
+      "step": 92
+    },
+    {
+      "epoch": 0.15161395500489078,
+      "grad_norm": 0.13868822157382965,
+      "learning_rate": 0.0001,
+      "loss": 0.2609,
+      "step": 93
+    },
+    {
+      "epoch": 0.1532442125855885,
+      "grad_norm": 0.14626921713352203,
+      "learning_rate": 0.0001,
+      "loss": 0.2442,
+      "step": 94
+    },
+    {
+      "epoch": 0.15487447016628628,
+      "grad_norm": 0.13605758547782898,
+      "learning_rate": 0.0001,
+      "loss": 0.2524,
+      "step": 95
+    },
+    {
+      "epoch": 0.156504727746984,
+      "grad_norm": 0.1354619562625885,
+      "learning_rate": 0.0001,
+      "loss": 0.2504,
+      "step": 96
+    },
+    {
+      "epoch": 0.15813498532768178,
+      "grad_norm": 0.1597517877817154,
+      "learning_rate": 0.0001,
+      "loss": 0.2477,
+      "step": 97
+    },
+    {
+      "epoch": 0.1597652429083795,
+      "grad_norm": 0.13010838627815247,
+      "learning_rate": 0.0001,
+      "loss": 0.271,
+      "step": 98
+    },
+    {
+      "epoch": 0.16139550048907728,
+      "grad_norm": 0.13291318714618683,
+      "learning_rate": 0.0001,
+      "loss": 0.2447,
+      "step": 99
+    },
+    {
+      "epoch": 0.163025758069775,
+      "grad_norm": 0.16925360262393951,
+      "learning_rate": 0.0001,
+      "loss": 0.2617,
+      "step": 100
+    },
+    {
+      "epoch": 0.16465601565047278,
+      "grad_norm": 0.13861025869846344,
+      "learning_rate": 0.0001,
+      "loss": 0.2503,
+      "step": 101
+    },
+    {
+      "epoch": 0.1662862732311705,
+      "grad_norm": 0.16513262689113617,
+      "learning_rate": 0.0001,
+      "loss": 0.2513,
+      "step": 102
+    },
+    {
+      "epoch": 0.16791653081186828,
+      "grad_norm": 0.14727436006069183,
+      "learning_rate": 0.0001,
+      "loss": 0.2769,
+      "step": 103
+    },
+    {
+      "epoch": 0.16954678839256604,
+      "grad_norm": 0.14519990980625153,
+      "learning_rate": 0.0001,
+      "loss": 0.2506,
+      "step": 104
+    },
+    {
+      "epoch": 0.17117704597326378,
+      "grad_norm": 0.14038777351379395,
+      "learning_rate": 0.0001,
+      "loss": 0.2619,
+      "step": 105
+    },
+    {
+      "epoch": 0.17280730355396154,
+      "grad_norm": 0.13004878163337708,
+      "learning_rate": 0.0001,
+      "loss": 0.2466,
+      "step": 106
+    },
+    {
+      "epoch": 0.17443756113465927,
+      "grad_norm": 0.11581247299909592,
+      "learning_rate": 0.0001,
+      "loss": 0.2365,
+      "step": 107
+    },
+    {
+      "epoch": 0.17606781871535704,
+      "grad_norm": 0.14593157172203064,
+      "learning_rate": 0.0001,
+      "loss": 0.2446,
+      "step": 108
+    },
+    {
+      "epoch": 0.17769807629605477,
+      "grad_norm": 0.15348902344703674,
+      "learning_rate": 0.0001,
+      "loss": 0.2613,
+      "step": 109
+    },
+    {
+      "epoch": 0.17932833387675254,
+      "grad_norm": 0.11216390132904053,
+      "learning_rate": 0.0001,
+      "loss": 0.2328,
+      "step": 110
+    },
+    {
+      "epoch": 0.18095859145745027,
+      "grad_norm": 0.17392517626285553,
+      "learning_rate": 0.0001,
+      "loss": 0.2449,
+      "step": 111
+    },
+    {
+      "epoch": 0.18258884903814804,
+      "grad_norm": 0.13363705575466156,
+      "learning_rate": 0.0001,
+      "loss": 0.2577,
+      "step": 112
+    },
+    {
+      "epoch": 0.18421910661884577,
+      "grad_norm": 0.12435399740934372,
+      "learning_rate": 0.0001,
+      "loss": 0.2478,
+      "step": 113
+    },
+    {
+      "epoch": 0.18584936419954354,
+      "grad_norm": 0.15440337359905243,
+      "learning_rate": 0.0001,
+      "loss": 0.2655,
+      "step": 114
+    },
+    {
+      "epoch": 0.18747962178024127,
+      "grad_norm": 0.12815740704536438,
+      "learning_rate": 0.0001,
+      "loss": 0.2642,
+      "step": 115
+    },
+    {
+      "epoch": 0.18910987936093904,
+      "grad_norm": 0.11212265491485596,
+      "learning_rate": 0.0001,
+      "loss": 0.2496,
+      "step": 116
+    },
+    {
+      "epoch": 0.19074013694163677,
+      "grad_norm": 0.13762663304805756,
+      "learning_rate": 0.0001,
+      "loss": 0.2382,
+      "step": 117
+    },
+    {
+      "epoch": 0.19237039452233454,
+      "grad_norm": 0.11899245530366898,
+      "learning_rate": 0.0001,
+      "loss": 0.2473,
+      "step": 118
+    },
+    {
+      "epoch": 0.19400065210303227,
+      "grad_norm": 0.12855815887451172,
+      "learning_rate": 0.0001,
+      "loss": 0.2478,
+      "step": 119
+    },
+    {
+      "epoch": 0.19563090968373004,
+      "grad_norm": 0.116991326212883,
+      "learning_rate": 0.0001,
+      "loss": 0.268,
+      "step": 120
+    },
+    {
+      "epoch": 0.19726116726442777,
+      "grad_norm": 0.11379344016313553,
+      "learning_rate": 0.0001,
+      "loss": 0.2391,
+      "step": 121
+    },
+    {
+      "epoch": 0.19889142484512554,
+      "grad_norm": 0.12145145982503891,
+      "learning_rate": 0.0001,
+      "loss": 0.2433,
+      "step": 122
+    },
+    {
+      "epoch": 0.20052168242582327,
+      "grad_norm": 0.12102972716093063,
+      "learning_rate": 0.0001,
+      "loss": 0.2384,
+      "step": 123
+    },
+    {
+      "epoch": 0.20215194000652104,
+      "grad_norm": 0.13323700428009033,
+      "learning_rate": 0.0001,
+      "loss": 0.2393,
+      "step": 124
+    },
+    {
+      "epoch": 0.20378219758721877,
+      "grad_norm": 0.13560782372951508,
+      "learning_rate": 0.0001,
+      "loss": 0.2434,
+      "step": 125
+    },
+    {
+      "epoch": 0.20541245516791654,
+      "grad_norm": 0.14215120673179626,
+      "learning_rate": 0.0001,
+      "loss": 0.2478,
+      "step": 126
+    },
+    {
+      "epoch": 0.20704271274861427,
+      "grad_norm": 0.11859458684921265,
+      "learning_rate": 0.0001,
+      "loss": 0.2274,
+      "step": 127
+    },
+    {
+      "epoch": 0.20867297032931204,
+      "grad_norm": 0.12831608951091766,
+      "learning_rate": 0.0001,
+      "loss": 0.2625,
+      "step": 128
+    },
+    {
+      "epoch": 0.21030322791000977,
+      "grad_norm": 0.15069334208965302,
+      "learning_rate": 0.0001,
+      "loss": 0.2542,
+      "step": 129
+    },
+    {
+      "epoch": 0.21193348549070753,
+      "grad_norm": 0.13063862919807434,
+      "learning_rate": 0.0001,
+      "loss": 0.2428,
+      "step": 130
+    },
+    {
+      "epoch": 0.21356374307140527,
+      "grad_norm": 0.12685725092887878,
+      "learning_rate": 0.0001,
+      "loss": 0.2312,
+      "step": 131
+    },
+    {
+      "epoch": 0.21519400065210303,
+      "grad_norm": 0.13156022131443024,
+      "learning_rate": 0.0001,
+      "loss": 0.2548,
+      "step": 132
+    },
+    {
+      "epoch": 0.21682425823280077,
+      "grad_norm": 0.12375540286302567,
+      "learning_rate": 0.0001,
+      "loss": 0.241,
+      "step": 133
+    },
+    {
+      "epoch": 0.21845451581349853,
+      "grad_norm": 0.12112291157245636,
+      "learning_rate": 0.0001,
+      "loss": 0.2483,
+      "step": 134
+    },
+    {
+      "epoch": 0.22008477339419627,
+      "grad_norm": 0.13035452365875244,
+      "learning_rate": 0.0001,
+      "loss": 0.2418,
+      "step": 135
+    },
+    {
+      "epoch": 0.22171503097489403,
+      "grad_norm": 0.14445112645626068,
+      "learning_rate": 0.0001,
+      "loss": 0.2613,
+      "step": 136
+    },
+    {
+      "epoch": 0.22334528855559177,
+      "grad_norm": 0.11515045166015625,
+      "learning_rate": 0.0001,
+      "loss": 0.2423,
+      "step": 137
+    },
+    {
+      "epoch": 0.22497554613628953,
+      "grad_norm": 0.1339021772146225,
+      "learning_rate": 0.0001,
+      "loss": 0.2462,
+      "step": 138
+    },
+    {
+      "epoch": 0.2266058037169873,
+      "grad_norm": 0.14942044019699097,
+      "learning_rate": 0.0001,
+      "loss": 0.2289,
+      "step": 139
+    },
+    {
+      "epoch": 0.22823606129768503,
+      "grad_norm": 0.15505343675613403,
+      "learning_rate": 0.0001,
+      "loss": 0.2507,
+      "step": 140
+    },
+    {
+      "epoch": 0.2298663188783828,
+      "grad_norm": 0.15206627547740936,
+      "learning_rate": 0.0001,
+      "loss": 0.2466,
+      "step": 141
+    },
+    {
+      "epoch": 0.23149657645908053,
+      "grad_norm": 0.21628950536251068,
+      "learning_rate": 0.0001,
+      "loss": 0.2506,
+      "step": 142
+    },
+    {
+      "epoch": 0.2331268340397783,
+      "grad_norm": 0.146853506565094,
+      "learning_rate": 0.0001,
+      "loss": 0.2378,
+      "step": 143
+    },
+    {
+      "epoch": 0.23475709162047603,
+      "grad_norm": 0.13807877898216248,
+      "learning_rate": 0.0001,
+      "loss": 0.2304,
+      "step": 144
+    },
+    {
+      "epoch": 0.2363873492011738,
+      "grad_norm": 0.1247410997748375,
+      "learning_rate": 0.0001,
+      "loss": 0.2154,
+      "step": 145
+    },
+    {
+      "epoch": 0.23801760678187153,
+      "grad_norm": 0.11460768431425095,
+      "learning_rate": 0.0001,
+      "loss": 0.2539,
+      "step": 146
+    },
+    {
+      "epoch": 0.2396478643625693,
+      "grad_norm": 0.13893291354179382,
+      "learning_rate": 0.0001,
+      "loss": 0.2265,
+      "step": 147
+    },
+    {
+      "epoch": 0.24127812194326703,
+      "grad_norm": 0.13536906242370605,
+      "learning_rate": 0.0001,
+      "loss": 0.2267,
+      "step": 148
+    },
+    {
+      "epoch": 0.2429083795239648,
+      "grad_norm": 0.14529214799404144,
+      "learning_rate": 0.0001,
+      "loss": 0.2559,
+      "step": 149
+    },
+    {
+      "epoch": 0.24453863710466253,
+      "grad_norm": 0.13840830326080322,
+      "learning_rate": 0.0001,
+      "loss": 0.2353,
+      "step": 150
+    },
+    {
+      "epoch": 0.2461688946853603,
+      "grad_norm": 0.15911462903022766,
+      "learning_rate": 0.0001,
+      "loss": 0.2527,
+      "step": 151
+    },
+    {
+      "epoch": 0.24779915226605803,
+      "grad_norm": 0.11530350893735886,
+      "learning_rate": 0.0001,
+      "loss": 0.2568,
+      "step": 152
+    },
+    {
+      "epoch": 0.2494294098467558,
+      "grad_norm": 0.14135918021202087,
+      "learning_rate": 0.0001,
+      "loss": 0.2266,
+      "step": 153
+    },
+    {
+      "epoch": 0.25105966742745356,
+      "grad_norm": 0.18591678142547607,
+      "learning_rate": 0.0001,
+      "loss": 0.2593,
+      "step": 154
+    },
+    {
+      "epoch": 0.2526899250081513,
+      "grad_norm": 0.13683144748210907,
+      "learning_rate": 0.0001,
+      "loss": 0.2469,
+      "step": 155
+    },
+    {
+      "epoch": 0.25432018258884903,
+      "grad_norm": 0.17030082643032074,
+      "learning_rate": 0.0001,
+      "loss": 0.2413,
+      "step": 156
+    },
+    {
+      "epoch": 0.25595044016954677,
+      "grad_norm": 0.1563073843717575,
+      "learning_rate": 0.0001,
+      "loss": 0.2435,
+      "step": 157
+    },
+    {
+      "epoch": 0.25758069775024456,
+      "grad_norm": 0.129858136177063,
+      "learning_rate": 0.0001,
+      "loss": 0.2515,
+      "step": 158
+    },
+    {
+      "epoch": 0.2592109553309423,
+      "grad_norm": 0.13964857161045074,
+      "learning_rate": 0.0001,
+      "loss": 0.2271,
+      "step": 159
+    },
+    {
+      "epoch": 0.26084121291164003,
+      "grad_norm": 0.12167726457118988,
+      "learning_rate": 0.0001,
+      "loss": 0.2092,
+      "step": 160
+    },
+    {
+      "epoch": 0.26247147049233777,
+      "grad_norm": 0.12050545960664749,
+      "learning_rate": 0.0001,
+      "loss": 0.2321,
+      "step": 161
+    },
+    {
+      "epoch": 0.26410172807303556,
+      "grad_norm": 0.14696434140205383,
+      "learning_rate": 0.0001,
+      "loss": 0.2545,
+      "step": 162
+    },
+    {
+      "epoch": 0.2657319856537333,
+      "grad_norm": 0.10441072285175323,
+      "learning_rate": 0.0001,
+      "loss": 0.2347,
+      "step": 163
+    },
+    {
+      "epoch": 0.26736224323443103,
+      "grad_norm": 0.11823071539402008,
+      "learning_rate": 0.0001,
+      "loss": 0.2242,
+      "step": 164
+    },
+    {
+      "epoch": 0.26899250081512877,
+      "grad_norm": 0.1426367461681366,
+      "learning_rate": 0.0001,
+      "loss": 0.2614,
+      "step": 165
+    },
+    {
+      "epoch": 0.27062275839582656,
+      "grad_norm": 0.11903452128171921,
+      "learning_rate": 0.0001,
+      "loss": 0.2469,
+      "step": 166
+    },
+    {
+      "epoch": 0.2722530159765243,
+      "grad_norm": 0.12051168829202652,
+      "learning_rate": 0.0001,
+      "loss": 0.246,
+      "step": 167
+    },
+    {
+      "epoch": 0.27388327355722203,
+      "grad_norm": 0.1146332174539566,
+      "learning_rate": 0.0001,
+      "loss": 0.2485,
+      "step": 168
+    },
+    {
+      "epoch": 0.27551353113791976,
+      "grad_norm": 0.1366330087184906,
+      "learning_rate": 0.0001,
+      "loss": 0.2364,
+      "step": 169
+    },
+    {
+      "epoch": 0.27714378871861756,
+      "grad_norm": 0.19316819310188293,
+      "learning_rate": 0.0001,
+      "loss": 0.2373,
+      "step": 170
+    },
+    {
+      "epoch": 0.2787740462993153,
+      "grad_norm": 0.13915376365184784,
+      "learning_rate": 0.0001,
+      "loss": 0.2354,
+      "step": 171
+    },
+    {
+      "epoch": 0.28040430388001303,
+      "grad_norm": 0.16246309876441956,
+      "learning_rate": 0.0001,
+      "loss": 0.2452,
+      "step": 172
+    },
+    {
+      "epoch": 0.2820345614607108,
+      "grad_norm": 0.11810585111379623,
+      "learning_rate": 0.0001,
+      "loss": 0.2312,
+      "step": 173
+    },
+    {
+      "epoch": 0.28366481904140856,
+      "grad_norm": 0.15839870274066925,
+      "learning_rate": 0.0001,
+      "loss": 0.2464,
+      "step": 174
+    },
+    {
+      "epoch": 0.2852950766221063,
+      "grad_norm": 0.1045420840382576,
+      "learning_rate": 0.0001,
+      "loss": 0.2257,
+      "step": 175
+    },
+    {
+      "epoch": 0.286925334202804,
+      "grad_norm": 0.14196939766407013,
+      "learning_rate": 0.0001,
+      "loss": 0.2523,
+      "step": 176
+    },
+    {
+      "epoch": 0.2885555917835018,
+      "grad_norm": 0.14301368594169617,
+      "learning_rate": 0.0001,
+      "loss": 0.2387,
+      "step": 177
+    },
+    {
+      "epoch": 0.29018584936419956,
+      "grad_norm": 0.13265396654605865,
+      "learning_rate": 0.0001,
+      "loss": 0.2428,
+      "step": 178
+    },
+    {
+      "epoch": 0.2918161069448973,
+      "grad_norm": 0.126626119017601,
+      "learning_rate": 0.0001,
+      "loss": 0.2494,
+      "step": 179
+    },
+    {
+      "epoch": 0.293446364525595,
+      "grad_norm": 0.13440294563770294,
+      "learning_rate": 0.0001,
+      "loss": 0.2457,
+      "step": 180
+    },
+    {
+      "epoch": 0.2950766221062928,
+      "grad_norm": 0.1397274285554886,
+      "learning_rate": 0.0001,
+      "loss": 0.2274,
+      "step": 181
+    },
+    {
+      "epoch": 0.29670687968699055,
+      "grad_norm": 0.09933532774448395,
+      "learning_rate": 0.0001,
+      "loss": 0.225,
+      "step": 182
+    },
+    {
+      "epoch": 0.2983371372676883,
+      "grad_norm": 0.16503044962882996,
+      "learning_rate": 0.0001,
+      "loss": 0.2293,
+      "step": 183
+    },
+    {
+      "epoch": 0.299967394848386,
+      "grad_norm": 0.12690463662147522,
+      "learning_rate": 0.0001,
+      "loss": 0.2447,
+      "step": 184
+    },
+    {
+      "epoch": 0.3015976524290838,
+      "grad_norm": 0.15822118520736694,
+      "learning_rate": 0.0001,
+      "loss": 0.2492,
+      "step": 185
+    },
+    {
+      "epoch": 0.30322791000978155,
+      "grad_norm": 0.10334572941064835,
+      "learning_rate": 0.0001,
+      "loss": 0.2334,
+      "step": 186
+    },
+    {
+      "epoch": 0.3048581675904793,
+      "grad_norm": 0.12315747141838074,
+      "learning_rate": 0.0001,
+      "loss": 0.2335,
+      "step": 187
+    },
+    {
+      "epoch": 0.306488425171177,
+      "grad_norm": 0.13603512942790985,
+      "learning_rate": 0.0001,
+      "loss": 0.228,
+      "step": 188
+    },
+    {
+      "epoch": 0.3081186827518748,
+      "grad_norm": 0.1011793464422226,
+      "learning_rate": 0.0001,
+      "loss": 0.2224,
+      "step": 189
+    },
+    {
+      "epoch": 0.30974894033257255,
+      "grad_norm": 0.10281293839216232,
+      "learning_rate": 0.0001,
+      "loss": 0.2277,
+      "step": 190
+    },
+    {
+      "epoch": 0.3113791979132703,
+      "grad_norm": 0.1417594999074936,
+      "learning_rate": 0.0001,
+      "loss": 0.2516,
+      "step": 191
+    },
+    {
+      "epoch": 0.313009455493968,
+      "grad_norm": 0.11304276436567307,
+      "learning_rate": 0.0001,
+      "loss": 0.2323,
+      "step": 192
+    },
+    {
+      "epoch": 0.3146397130746658,
+      "grad_norm": 0.15146027505397797,
+      "learning_rate": 0.0001,
+      "loss": 0.2446,
+      "step": 193
+    },
+    {
+      "epoch": 0.31626997065536355,
+      "grad_norm": 0.11567169427871704,
+      "learning_rate": 0.0001,
+      "loss": 0.2018,
+      "step": 194
+    },
+    {
+      "epoch": 0.3179002282360613,
+      "grad_norm": 0.12554962933063507,
+      "learning_rate": 0.0001,
+      "loss": 0.2317,
+      "step": 195
+    },
+    {
+      "epoch": 0.319530485816759,
+      "grad_norm": 0.15784506499767303,
+      "learning_rate": 0.0001,
+      "loss": 0.2372,
+      "step": 196
+    },
+    {
+      "epoch": 0.3211607433974568,
+      "grad_norm": 0.13723263144493103,
+      "learning_rate": 0.0001,
+      "loss": 0.2395,
+      "step": 197
+    },
+    {
+      "epoch": 0.32279100097815455,
+      "grad_norm": 0.16114649176597595,
+      "learning_rate": 0.0001,
+      "loss": 0.244,
+      "step": 198
+    },
+    {
+      "epoch": 0.3244212585588523,
+      "grad_norm": 0.11904025822877884,
+      "learning_rate": 0.0001,
+      "loss": 0.2366,
+      "step": 199
+    },
+    {
+      "epoch": 0.32605151613955,
+      "grad_norm": 0.1232055053114891,
+      "learning_rate": 0.0001,
+      "loss": 0.2266,
+      "step": 200
+    },
+    {
+      "epoch": 0.3276817737202478,
+      "grad_norm": 0.1447722166776657,
+      "learning_rate": 0.0001,
+      "loss": 0.2451,
+      "step": 201
+    },
+    {
+      "epoch": 0.32931203130094555,
+      "grad_norm": 0.10923551023006439,
+      "learning_rate": 0.0001,
+      "loss": 0.2242,
+      "step": 202
+    },
+    {
+      "epoch": 0.3309422888816433,
+      "grad_norm": 0.13391102850437164,
+      "learning_rate": 0.0001,
+      "loss": 0.2342,
+      "step": 203
+    },
+    {
+      "epoch": 0.332572546462341,
+      "grad_norm": 0.14067484438419342,
+      "learning_rate": 0.0001,
+      "loss": 0.2334,
+      "step": 204
+    },
+    {
+      "epoch": 0.3342028040430388,
+      "grad_norm": 0.13190071284770966,
+      "learning_rate": 0.0001,
+      "loss": 0.2398,
+      "step": 205
+    },
+    {
+      "epoch": 0.33583306162373655,
+      "grad_norm": 0.12358011305332184,
+      "learning_rate": 0.0001,
+      "loss": 0.2218,
+      "step": 206
+    },
+    {
+      "epoch": 0.3374633192044343,
+      "grad_norm": 0.11492815613746643,
+      "learning_rate": 0.0001,
+      "loss": 0.2284,
+      "step": 207
+    },
+    {
+      "epoch": 0.3390935767851321,
+      "grad_norm": 0.11512556672096252,
+      "learning_rate": 0.0001,
+      "loss": 0.2265,
+      "step": 208
+    },
+    {
+      "epoch": 0.3407238343658298,
+      "grad_norm": 0.11229848861694336,
+      "learning_rate": 0.0001,
+      "loss": 0.2253,
+      "step": 209
+    },
+    {
+      "epoch": 0.34235409194652755,
+      "grad_norm": 0.11124417930841446,
+      "learning_rate": 0.0001,
+      "loss": 0.2449,
+      "step": 210
+    },
+    {
+      "epoch": 0.3439843495272253,
+      "grad_norm": 0.11999543011188507,
+      "learning_rate": 0.0001,
+      "loss": 0.2255,
+      "step": 211
+    },
+    {
+      "epoch": 0.3456146071079231,
+      "grad_norm": 0.10648955404758453,
+      "learning_rate": 0.0001,
+      "loss": 0.2104,
+      "step": 212
+    },
+    {
+      "epoch": 0.3472448646886208,
+      "grad_norm": 0.12528495490550995,
+      "learning_rate": 0.0001,
+      "loss": 0.2459,
+      "step": 213
+    },
+    {
+      "epoch": 0.34887512226931855,
+      "grad_norm": 0.12371502816677094,
+      "learning_rate": 0.0001,
+      "loss": 0.2483,
+      "step": 214
+    },
+    {
+      "epoch": 0.3505053798500163,
+      "grad_norm": 0.15203551948070526,
+      "learning_rate": 0.0001,
+      "loss": 0.23,
+      "step": 215
+    },
+    {
+      "epoch": 0.3521356374307141,
+      "grad_norm": 0.12010787427425385,
+      "learning_rate": 0.0001,
+      "loss": 0.2221,
+      "step": 216
+    },
+    {
+      "epoch": 0.3537658950114118,
+      "grad_norm": 0.1346224844455719,
+      "learning_rate": 0.0001,
+      "loss": 0.2268,
+      "step": 217
+    },
+    {
+      "epoch": 0.35539615259210955,
+      "grad_norm": 0.15142722427845,
+      "learning_rate": 0.0001,
+      "loss": 0.2338,
+      "step": 218
+    },
+    {
+      "epoch": 0.3570264101728073,
+      "grad_norm": 0.129653200507164,
+      "learning_rate": 0.0001,
+      "loss": 0.2325,
+      "step": 219
+    },
+    {
+      "epoch": 0.3586566677535051,
+      "grad_norm": 0.13703054189682007,
+      "learning_rate": 0.0001,
+      "loss": 0.2278,
+      "step": 220
+    },
+    {
+      "epoch": 0.3602869253342028,
+      "grad_norm": 0.1323288083076477,
+      "learning_rate": 0.0001,
+      "loss": 0.2331,
+      "step": 221
+    },
+    {
+      "epoch": 0.36191718291490055,
+      "grad_norm": 0.14150388538837433,
+      "learning_rate": 0.0001,
+      "loss": 0.2408,
+      "step": 222
+    },
+    {
+      "epoch": 0.3635474404955983,
+      "grad_norm": 0.16232647001743317,
+      "learning_rate": 0.0001,
+      "loss": 0.2228,
+      "step": 223
+    },
+    {
+      "epoch": 0.3651776980762961,
+      "grad_norm": 0.13750429451465607,
+      "learning_rate": 0.0001,
+      "loss": 0.2578,
+      "step": 224
+    },
+    {
+      "epoch": 0.3668079556569938,
+      "grad_norm": 0.15641643106937408,
+      "learning_rate": 0.0001,
+      "loss": 0.2202,
+      "step": 225
+    },
+    {
+      "epoch": 0.36843821323769155,
+      "grad_norm": 0.14466796815395355,
+      "learning_rate": 0.0001,
+      "loss": 0.2148,
+      "step": 226
+    },
+    {
+      "epoch": 0.3700684708183893,
+      "grad_norm": 0.11243683099746704,
+      "learning_rate": 0.0001,
+      "loss": 0.2353,
+      "step": 227
+    },
+    {
+      "epoch": 0.3716987283990871,
+      "grad_norm": 0.12383094429969788,
+      "learning_rate": 0.0001,
+      "loss": 0.2254,
+      "step": 228
+    },
+    {
+      "epoch": 0.3733289859797848,
+      "grad_norm": 0.11095986515283585,
+      "learning_rate": 0.0001,
+      "loss": 0.225,
+      "step": 229
+    },
+    {
+      "epoch": 0.37495924356048255,
+      "grad_norm": 0.12428440898656845,
+      "learning_rate": 0.0001,
+      "loss": 0.233,
+      "step": 230
+    },
+    {
+      "epoch": 0.3765895011411803,
+      "grad_norm": 0.14329205453395844,
+      "learning_rate": 0.0001,
+      "loss": 0.2307,
+      "step": 231
+    },
+    {
+      "epoch": 0.3782197587218781,
+      "grad_norm": 0.14026683568954468,
+      "learning_rate": 0.0001,
+      "loss": 0.2139,
+      "step": 232
+    },
+    {
+      "epoch": 0.3798500163025758,
+      "grad_norm": 0.11123739928007126,
+      "learning_rate": 0.0001,
+      "loss": 0.2191,
+      "step": 233
+    },
+    {
+      "epoch": 0.38148027388327355,
+      "grad_norm": 0.1463911086320877,
+      "learning_rate": 0.0001,
+      "loss": 0.2388,
+      "step": 234
+    },
+    {
+      "epoch": 0.3831105314639713,
+      "grad_norm": 0.13069649040699005,
+      "learning_rate": 0.0001,
+      "loss": 0.2185,
+      "step": 235
+    },
+    {
+      "epoch": 0.3847407890446691,
+      "grad_norm": 0.14931800961494446,
+      "learning_rate": 0.0001,
+      "loss": 0.2416,
+      "step": 236
+    },
+    {
+      "epoch": 0.3863710466253668,
+      "grad_norm": 0.16097313165664673,
+      "learning_rate": 0.0001,
+      "loss": 0.2347,
+      "step": 237
+    },
+    {
+      "epoch": 0.38800130420606455,
+      "grad_norm": 0.15226350724697113,
+      "learning_rate": 0.0001,
+      "loss": 0.2258,
+      "step": 238
+    },
+    {
+      "epoch": 0.3896315617867623,
+      "grad_norm": 0.13384918868541718,
+      "learning_rate": 0.0001,
+      "loss": 0.2265,
+      "step": 239
+    },
+    {
+      "epoch": 0.3912618193674601,
+      "grad_norm": 0.11753430217504501,
+      "learning_rate": 0.0001,
+      "loss": 0.2106,
+      "step": 240
+    },
+    {
+      "epoch": 0.3928920769481578,
+      "grad_norm": 0.1108139306306839,
+      "learning_rate": 0.0001,
+      "loss": 0.2293,
+      "step": 241
+    },
+    {
+      "epoch": 0.39452233452885554,
+      "grad_norm": 0.10754958540201187,
+      "learning_rate": 0.0001,
+      "loss": 0.2251,
+      "step": 242
+    },
+    {
+      "epoch": 0.39615259210955334,
+      "grad_norm": 0.11824717372655869,
+      "learning_rate": 0.0001,
+      "loss": 0.2507,
+      "step": 243
+    },
+    {
+      "epoch": 0.3977828496902511,
+      "grad_norm": 0.14322006702423096,
+      "learning_rate": 0.0001,
+      "loss": 0.2348,
+      "step": 244
+    },
+    {
+      "epoch": 0.3994131072709488,
+      "grad_norm": 0.10835988074541092,
+      "learning_rate": 0.0001,
+      "loss": 0.2141,
+      "step": 245
+    },
+    {
+      "epoch": 0.40104336485164654,
+      "grad_norm": 0.11391481757164001,
+      "learning_rate": 0.0001,
+      "loss": 0.2139,
+      "step": 246
+    },
+    {
+      "epoch": 0.40267362243234434,
+      "grad_norm": 0.11263515055179596,
+      "learning_rate": 0.0001,
+      "loss": 0.2354,
+      "step": 247
+    },
+    {
+      "epoch": 0.40430388001304207,
+      "grad_norm": 0.12588317692279816,
+      "learning_rate": 0.0001,
+      "loss": 0.2354,
+      "step": 248
+    },
+    {
+      "epoch": 0.4059341375937398,
+      "grad_norm": 0.12664109468460083,
+      "learning_rate": 0.0001,
+      "loss": 0.2371,
+      "step": 249
+    },
+    {
+      "epoch": 0.40756439517443754,
+      "grad_norm": 0.12330986559391022,
+      "learning_rate": 0.0001,
+      "loss": 0.2338,
+      "step": 250
+    },
+    {
+      "epoch": 0.40919465275513534,
+      "grad_norm": 0.11349525302648544,
+      "learning_rate": 0.0001,
+      "loss": 0.2315,
+      "step": 251
+    },
+    {
+      "epoch": 0.41082491033583307,
+      "grad_norm": 0.11793797463178635,
+      "learning_rate": 0.0001,
+      "loss": 0.2309,
+      "step": 252
+    },
+    {
+      "epoch": 0.4124551679165308,
+      "grad_norm": 0.14245276153087616,
+      "learning_rate": 0.0001,
+      "loss": 0.2469,
+      "step": 253
+    },
+    {
+      "epoch": 0.41408542549722854,
+      "grad_norm": 0.11353014409542084,
+      "learning_rate": 0.0001,
+      "loss": 0.2238,
+      "step": 254
+    },
+    {
+      "epoch": 0.41571568307792633,
+      "grad_norm": 0.13224811851978302,
+      "learning_rate": 0.0001,
+      "loss": 0.2419,
+      "step": 255
+    },
+    {
+      "epoch": 0.41734594065862407,
+      "grad_norm": 0.10972931236028671,
+      "learning_rate": 0.0001,
+      "loss": 0.2139,
+      "step": 256
+    },
+    {
+      "epoch": 0.4189761982393218,
+      "grad_norm": 0.17038947343826294,
+      "learning_rate": 0.0001,
+      "loss": 0.2233,
+      "step": 257
+    },
+    {
+      "epoch": 0.42060645582001954,
+      "grad_norm": 0.1390901654958725,
+      "learning_rate": 0.0001,
+      "loss": 0.2347,
+      "step": 258
+    },
+    {
+      "epoch": 0.42223671340071733,
+      "grad_norm": 0.10498136281967163,
+      "learning_rate": 0.0001,
+      "loss": 0.2185,
+      "step": 259
+    },
+    {
+      "epoch": 0.42386697098141507,
+      "grad_norm": 0.11392710357904434,
+      "learning_rate": 0.0001,
+      "loss": 0.2317,
+      "step": 260
+    },
+    {
+      "epoch": 0.4254972285621128,
+      "grad_norm": 0.12423884123563766,
+      "learning_rate": 0.0001,
+      "loss": 0.238,
+      "step": 261
+    },
+    {
+      "epoch": 0.42712748614281054,
+      "grad_norm": 0.12265395373106003,
+      "learning_rate": 0.0001,
+      "loss": 0.2151,
+      "step": 262
+    },
+    {
+      "epoch": 0.42875774372350833,
+      "grad_norm": 0.1066151112318039,
+      "learning_rate": 0.0001,
+      "loss": 0.2188,
+      "step": 263
+    },
+    {
+      "epoch": 0.43038800130420607,
+      "grad_norm": 0.10593073070049286,
+      "learning_rate": 0.0001,
+      "loss": 0.2179,
+      "step": 264
+    },
+    {
+      "epoch": 0.4320182588849038,
+      "grad_norm": 0.12529350817203522,
+      "learning_rate": 0.0001,
+      "loss": 0.2366,
+      "step": 265
+    },
+    {
+      "epoch": 0.43364851646560154,
+      "grad_norm": 0.13577479124069214,
+      "learning_rate": 0.0001,
+      "loss": 0.2304,
+      "step": 266
+    },
+    {
+      "epoch": 0.43527877404629933,
+      "grad_norm": 0.15030768513679504,
+      "learning_rate": 0.0001,
+      "loss": 0.2421,
+      "step": 267
+    },
+    {
+      "epoch": 0.43690903162699707,
+      "grad_norm": 0.12566924095153809,
+      "learning_rate": 0.0001,
+      "loss": 0.2366,
+      "step": 268
+    },
+    {
+      "epoch": 0.4385392892076948,
+      "grad_norm": 0.1277414709329605,
+      "learning_rate": 0.0001,
+      "loss": 0.2413,
+      "step": 269
+    },
+    {
+      "epoch": 0.44016954678839254,
+      "grad_norm": 0.12140754610300064,
+      "learning_rate": 0.0001,
+      "loss": 0.2072,
+      "step": 270
+    },
+    {
+      "epoch": 0.44179980436909033,
+      "grad_norm": 0.16660185158252716,
+      "learning_rate": 0.0001,
+      "loss": 0.2513,
+      "step": 271
+    },
+    {
+      "epoch": 0.44343006194978807,
+      "grad_norm": 0.11619671434164047,
+      "learning_rate": 0.0001,
+      "loss": 0.2158,
+      "step": 272
+    },
+    {
+      "epoch": 0.4450603195304858,
+      "grad_norm": 0.16923661530017853,
+      "learning_rate": 0.0001,
+      "loss": 0.2479,
+      "step": 273
+    },
+    {
+      "epoch": 0.44669057711118354,
+      "grad_norm": 0.14053957164287567,
+      "learning_rate": 0.0001,
+      "loss": 0.2383,
+      "step": 274
+    },
+    {
+      "epoch": 0.44832083469188133,
+      "grad_norm": 0.1332424134016037,
+      "learning_rate": 0.0001,
+      "loss": 0.2202,
+      "step": 275
+    },
+    {
+      "epoch": 0.44995109227257907,
+      "grad_norm": 0.10877622663974762,
+      "learning_rate": 0.0001,
+      "loss": 0.234,
+      "step": 276
+    },
+    {
+      "epoch": 0.4515813498532768,
+      "grad_norm": 0.11536753922700882,
+      "learning_rate": 0.0001,
+      "loss": 0.2199,
+      "step": 277
+    },
+    {
+      "epoch": 0.4532116074339746,
+      "grad_norm": 0.12691466510295868,
+      "learning_rate": 0.0001,
+      "loss": 0.2391,
+      "step": 278
+    },
+    {
+      "epoch": 0.45484186501467233,
+      "grad_norm": 0.12029401957988739,
+      "learning_rate": 0.0001,
+      "loss": 0.2258,
+      "step": 279
+    },
+    {
+      "epoch": 0.45647212259537007,
+      "grad_norm": 0.14181189239025116,
+      "learning_rate": 0.0001,
+      "loss": 0.2202,
+      "step": 280
+    },
+    {
+      "epoch": 0.4581023801760678,
+      "grad_norm": 0.1265992522239685,
+      "learning_rate": 0.0001,
+      "loss": 0.2297,
+      "step": 281
+    },
+    {
+      "epoch": 0.4597326377567656,
+      "grad_norm": 0.1276220828294754,
+      "learning_rate": 0.0001,
+      "loss": 0.2282,
+      "step": 282
+    },
+    {
+      "epoch": 0.46136289533746333,
+      "grad_norm": 0.15334482491016388,
+      "learning_rate": 0.0001,
+      "loss": 0.2213,
+      "step": 283
+    },
+    {
+      "epoch": 0.46299315291816107,
+      "grad_norm": 0.14251714944839478,
+      "learning_rate": 0.0001,
+      "loss": 0.2244,
+      "step": 284
+    },
+    {
+      "epoch": 0.4646234104988588,
+      "grad_norm": 0.1373746544122696,
+      "learning_rate": 0.0001,
+      "loss": 0.2258,
+      "step": 285
+    },
+    {
+      "epoch": 0.4662536680795566,
+      "grad_norm": 0.14110830426216125,
+      "learning_rate": 0.0001,
+      "loss": 0.2192,
+      "step": 286
+    },
+    {
+      "epoch": 0.46788392566025433,
+      "grad_norm": 0.12478266656398773,
+      "learning_rate": 0.0001,
+      "loss": 0.2256,
+      "step": 287
+    },
+    {
+      "epoch": 0.46951418324095207,
+      "grad_norm": 0.13609923422336578,
+      "learning_rate": 0.0001,
+      "loss": 0.2273,
+      "step": 288
+    },
+    {
+      "epoch": 0.4711444408216498,
+      "grad_norm": 0.15859074890613556,
+      "learning_rate": 0.0001,
+      "loss": 0.2314,
+      "step": 289
+    },
+    {
+      "epoch": 0.4727746984023476,
+      "grad_norm": 0.1267993301153183,
+      "learning_rate": 0.0001,
+      "loss": 0.2431,
+      "step": 290
+    },
+    {
+      "epoch": 0.47440495598304533,
+      "grad_norm": 0.1309819370508194,
+      "learning_rate": 0.0001,
+      "loss": 0.2036,
+      "step": 291
+    },
+    {
+      "epoch": 0.47603521356374306,
+      "grad_norm": 0.14452795684337616,
+      "learning_rate": 0.0001,
+      "loss": 0.2113,
+      "step": 292
+    },
+    {
+      "epoch": 0.4776654711444408,
+      "grad_norm": 0.10817385464906693,
+      "learning_rate": 0.0001,
+      "loss": 0.2198,
+      "step": 293
+    },
+    {
+      "epoch": 0.4792957287251386,
+      "grad_norm": 0.10400962829589844,
+      "learning_rate": 0.0001,
+      "loss": 0.2107,
+      "step": 294
+    },
+    {
+      "epoch": 0.48092598630583633,
+      "grad_norm": 0.14228041470050812,
+      "learning_rate": 0.0001,
+      "loss": 0.2355,
+      "step": 295
+    },
+    {
+      "epoch": 0.48255624388653406,
+      "grad_norm": 0.144633486866951,
+      "learning_rate": 0.0001,
+      "loss": 0.237,
+      "step": 296
+    },
+    {
+      "epoch": 0.4841865014672318,
+      "grad_norm": 0.14238397777080536,
+      "learning_rate": 0.0001,
+      "loss": 0.2299,
+      "step": 297
+    },
+    {
+      "epoch": 0.4858167590479296,
+      "grad_norm": 0.12710346281528473,
+      "learning_rate": 0.0001,
+      "loss": 0.2267,
+      "step": 298
+    },
+    {
+      "epoch": 0.4874470166286273,
+      "grad_norm": 0.12529858946800232,
+      "learning_rate": 0.0001,
+      "loss": 0.2301,
+      "step": 299
+    },
+    {
+      "epoch": 0.48907727420932506,
+      "grad_norm": 0.12333487719297409,
+      "learning_rate": 0.0001,
+      "loss": 0.2457,
+      "step": 300
+    },
+    {
+      "epoch": 0.4907075317900228,
+      "grad_norm": 0.10910045355558395,
+      "learning_rate": 0.0001,
+      "loss": 0.2179,
+      "step": 301
+    },
+    {
+      "epoch": 0.4923377893707206,
+      "grad_norm": 0.13991764187812805,
+      "learning_rate": 0.0001,
+      "loss": 0.2361,
+      "step": 302
+    },
+    {
+      "epoch": 0.4939680469514183,
+      "grad_norm": 0.1286964863538742,
+      "learning_rate": 0.0001,
+      "loss": 0.2172,
+      "step": 303
+    },
+    {
+      "epoch": 0.49559830453211606,
+      "grad_norm": 0.13012994825839996,
+      "learning_rate": 0.0001,
+      "loss": 0.199,
+      "step": 304
+    },
+    {
+      "epoch": 0.4972285621128138,
+      "grad_norm": 0.12035807967185974,
+      "learning_rate": 0.0001,
+      "loss": 0.232,
+      "step": 305
+    },
+    {
+      "epoch": 0.4988588196935116,
+      "grad_norm": 0.1246248111128807,
+      "learning_rate": 0.0001,
+      "loss": 0.2271,
+      "step": 306
+    },
+    {
+      "epoch": 0.5004890772742093,
+      "grad_norm": 0.13514195382595062,
+      "learning_rate": 0.0001,
+      "loss": 0.2089,
+      "step": 307
+    },
+    {
+      "epoch": 0.5021193348549071,
+      "grad_norm": 0.1276530623435974,
+      "learning_rate": 0.0001,
+      "loss": 0.2282,
+      "step": 308
+    },
+    {
+      "epoch": 0.5037495924356048,
+      "grad_norm": 0.12651173770427704,
+      "learning_rate": 0.0001,
+      "loss": 0.2235,
+      "step": 309
+    },
+    {
+      "epoch": 0.5053798500163026,
+      "grad_norm": 0.12165319919586182,
+      "learning_rate": 0.0001,
+      "loss": 0.2283,
+      "step": 310
+    },
+    {
+      "epoch": 0.5070101075970004,
+      "grad_norm": 0.10687411576509476,
+      "learning_rate": 0.0001,
+      "loss": 0.2285,
+      "step": 311
+    },
+    {
+      "epoch": 0.5086403651776981,
+      "grad_norm": 0.10170154273509979,
+      "learning_rate": 0.0001,
+      "loss": 0.2165,
+      "step": 312
+    },
+    {
+      "epoch": 0.5102706227583959,
+      "grad_norm": 0.130455881357193,
+      "learning_rate": 0.0001,
+      "loss": 0.2236,
+      "step": 313
+    },
+    {
+      "epoch": 0.5119008803390935,
+      "grad_norm": 0.12910182774066925,
+      "learning_rate": 0.0001,
+      "loss": 0.2339,
+      "step": 314
+    },
+    {
+      "epoch": 0.5135311379197913,
+      "grad_norm": 0.12536931037902832,
+      "learning_rate": 0.0001,
+      "loss": 0.225,
+      "step": 315
+    },
+    {
+      "epoch": 0.5151613955004891,
+      "grad_norm": 0.11117815226316452,
+      "learning_rate": 0.0001,
+      "loss": 0.2167,
+      "step": 316
+    },
+    {
+      "epoch": 0.5167916530811868,
+      "grad_norm": 0.10664685070514679,
+      "learning_rate": 0.0001,
+      "loss": 0.2132,
+      "step": 317
+    },
+    {
+      "epoch": 0.5184219106618846,
+      "grad_norm": 0.11488650739192963,
+      "learning_rate": 0.0001,
+      "loss": 0.2032,
+      "step": 318
+    },
+    {
+      "epoch": 0.5200521682425824,
+      "grad_norm": 0.1376049518585205,
+      "learning_rate": 0.0001,
+      "loss": 0.2198,
+      "step": 319
+    },
+    {
+      "epoch": 0.5216824258232801,
+      "grad_norm": 0.11661184579133987,
+      "learning_rate": 0.0001,
+      "loss": 0.2225,
+      "step": 320
+    },
+    {
+      "epoch": 0.5233126834039779,
+      "grad_norm": 0.15519613027572632,
+      "learning_rate": 0.0001,
+      "loss": 0.2195,
+      "step": 321
+    },
+    {
+      "epoch": 0.5249429409846755,
+      "grad_norm": 0.13077083230018616,
+      "learning_rate": 0.0001,
+      "loss": 0.2413,
+      "step": 322
+    },
+    {
+      "epoch": 0.5265731985653733,
+      "grad_norm": 0.13975407183170319,
+      "learning_rate": 0.0001,
+      "loss": 0.2307,
+      "step": 323
+    },
+    {
+      "epoch": 0.5282034561460711,
+      "grad_norm": 0.12319333851337433,
+      "learning_rate": 0.0001,
+      "loss": 0.2122,
+      "step": 324
+    },
+    {
+      "epoch": 0.5298337137267688,
+      "grad_norm": 0.12614178657531738,
+      "learning_rate": 0.0001,
+      "loss": 0.2126,
+      "step": 325
+    },
+    {
+      "epoch": 0.5314639713074666,
+      "grad_norm": 0.12005390971899033,
+      "learning_rate": 0.0001,
+      "loss": 0.2261,
+      "step": 326
+    },
+    {
+      "epoch": 0.5330942288881644,
+      "grad_norm": 0.14331591129302979,
+      "learning_rate": 0.0001,
+      "loss": 0.2204,
+      "step": 327
+    },
+    {
+      "epoch": 0.5347244864688621,
+      "grad_norm": 0.1153935119509697,
+      "learning_rate": 0.0001,
+      "loss": 0.2221,
+      "step": 328
+    },
+    {
+      "epoch": 0.5363547440495599,
+      "grad_norm": 0.14517177641391754,
+      "learning_rate": 0.0001,
+      "loss": 0.2324,
+      "step": 329
+    },
+    {
+      "epoch": 0.5379850016302575,
+      "grad_norm": 0.1347828060388565,
+      "learning_rate": 0.0001,
+      "loss": 0.2157,
+      "step": 330
+    },
+    {
+      "epoch": 0.5396152592109553,
+      "grad_norm": 0.12938876450061798,
+      "learning_rate": 0.0001,
+      "loss": 0.221,
+      "step": 331
+    },
+    {
+      "epoch": 0.5412455167916531,
+      "grad_norm": 0.12898674607276917,
+      "learning_rate": 0.0001,
+      "loss": 0.2243,
+      "step": 332
+    },
+    {
+      "epoch": 0.5428757743723508,
+      "grad_norm": 0.1460563838481903,
+      "learning_rate": 0.0001,
+      "loss": 0.2091,
+      "step": 333
+    },
+    {
+      "epoch": 0.5445060319530486,
+      "grad_norm": 0.12404835224151611,
+      "learning_rate": 0.0001,
+      "loss": 0.2299,
+      "step": 334
+    },
+    {
+      "epoch": 0.5461362895337464,
+      "grad_norm": 0.10986137390136719,
+      "learning_rate": 0.0001,
+      "loss": 0.2119,
+      "step": 335
+    },
+    {
+      "epoch": 0.5477665471144441,
+      "grad_norm": 0.10479142516851425,
+      "learning_rate": 0.0001,
+      "loss": 0.2132,
+      "step": 336
+    },
+    {
+      "epoch": 0.5493968046951418,
+      "grad_norm": 0.14422518014907837,
+      "learning_rate": 0.0001,
+      "loss": 0.2379,
+      "step": 337
+    },
+    {
+      "epoch": 0.5510270622758395,
+      "grad_norm": 0.13756605982780457,
+      "learning_rate": 0.0001,
+      "loss": 0.2205,
+      "step": 338
+    },
+    {
+      "epoch": 0.5526573198565373,
+      "grad_norm": 0.15542298555374146,
+      "learning_rate": 0.0001,
+      "loss": 0.227,
+      "step": 339
+    },
+    {
+      "epoch": 0.5542875774372351,
+      "grad_norm": 0.11967030167579651,
+      "learning_rate": 0.0001,
+      "loss": 0.2336,
+      "step": 340
+    },
+    {
+      "epoch": 0.5559178350179328,
+      "grad_norm": 0.11930400133132935,
+      "learning_rate": 0.0001,
+      "loss": 0.2273,
+      "step": 341
+    },
+    {
+      "epoch": 0.5575480925986306,
+      "grad_norm": 0.13011619448661804,
+      "learning_rate": 0.0001,
+      "loss": 0.2386,
+      "step": 342
+    },
+    {
+      "epoch": 0.5591783501793284,
+      "grad_norm": 0.1742897927761078,
+      "learning_rate": 0.0001,
+      "loss": 0.2248,
+      "step": 343
+    },
+    {
+      "epoch": 0.5608086077600261,
+      "grad_norm": 0.09916642308235168,
+      "learning_rate": 0.0001,
+      "loss": 0.2169,
+      "step": 344
+    },
+    {
+      "epoch": 0.5624388653407238,
+      "grad_norm": 0.11124306917190552,
+      "learning_rate": 0.0001,
+      "loss": 0.2173,
+      "step": 345
+    },
+    {
+      "epoch": 0.5640691229214216,
+      "grad_norm": 0.13426317274570465,
+      "learning_rate": 0.0001,
+      "loss": 0.2235,
+      "step": 346
+    },
+    {
+      "epoch": 0.5656993805021193,
+      "grad_norm": 0.12277089059352875,
+      "learning_rate": 0.0001,
+      "loss": 0.2234,
+      "step": 347
+    },
+    {
+      "epoch": 0.5673296380828171,
+      "grad_norm": 0.1190582811832428,
+      "learning_rate": 0.0001,
+      "loss": 0.2173,
+      "step": 348
+    },
+    {
+      "epoch": 0.5689598956635148,
+      "grad_norm": 0.10108431428670883,
+      "learning_rate": 0.0001,
+      "loss": 0.2133,
+      "step": 349
+    },
+    {
+      "epoch": 0.5705901532442126,
+      "grad_norm": 0.10684805363416672,
+      "learning_rate": 0.0001,
+      "loss": 0.2311,
+      "step": 350
+    },
+    {
+      "epoch": 0.5722204108249104,
+      "grad_norm": 0.10276418924331665,
+      "learning_rate": 0.0001,
+      "loss": 0.2137,
+      "step": 351
+    },
+    {
+      "epoch": 0.573850668405608,
+      "grad_norm": 0.10777677595615387,
+      "learning_rate": 0.0001,
+      "loss": 0.2083,
+      "step": 352
+    },
+    {
+      "epoch": 0.5754809259863058,
+      "grad_norm": 0.12655171751976013,
+      "learning_rate": 0.0001,
+      "loss": 0.2269,
+      "step": 353
+    },
+    {
+      "epoch": 0.5771111835670036,
+      "grad_norm": 0.11766703426837921,
+      "learning_rate": 0.0001,
+      "loss": 0.2185,
+      "step": 354
+    },
+    {
+      "epoch": 0.5787414411477013,
+      "grad_norm": 0.10404328256845474,
+      "learning_rate": 0.0001,
+      "loss": 0.2027,
+      "step": 355
+    },
+    {
+      "epoch": 0.5803716987283991,
+      "grad_norm": 0.10919814556837082,
+      "learning_rate": 0.0001,
+      "loss": 0.2025,
+      "step": 356
+    },
+    {
+      "epoch": 0.5820019563090968,
+      "grad_norm": 0.12742376327514648,
+      "learning_rate": 0.0001,
+      "loss": 0.2129,
+      "step": 357
+    },
+    {
+      "epoch": 0.5836322138897946,
+      "grad_norm": 0.133440762758255,
+      "learning_rate": 0.0001,
+      "loss": 0.2142,
+      "step": 358
+    },
+    {
+      "epoch": 0.5852624714704924,
+      "grad_norm": 0.11150208115577698,
+      "learning_rate": 0.0001,
+      "loss": 0.216,
+      "step": 359
+    },
+    {
+      "epoch": 0.58689272905119,
+      "grad_norm": 0.12375590205192566,
+      "learning_rate": 0.0001,
+      "loss": 0.2148,
+      "step": 360
+    },
+    {
+      "epoch": 0.5885229866318878,
+      "grad_norm": 0.13547080755233765,
+      "learning_rate": 0.0001,
+      "loss": 0.2105,
+      "step": 361
+    },
+    {
+      "epoch": 0.5901532442125856,
+      "grad_norm": 0.1400783509016037,
+      "learning_rate": 0.0001,
+      "loss": 0.2314,
+      "step": 362
+    },
+    {
+      "epoch": 0.5917835017932833,
+      "grad_norm": 0.11818625032901764,
+      "learning_rate": 0.0001,
+      "loss": 0.2098,
+      "step": 363
+    },
+    {
+      "epoch": 0.5934137593739811,
+      "grad_norm": 0.11656329035758972,
+      "learning_rate": 0.0001,
+      "loss": 0.2124,
+      "step": 364
+    },
+    {
+      "epoch": 0.5950440169546788,
+      "grad_norm": 0.17707152664661407,
+      "learning_rate": 0.0001,
+      "loss": 0.2004,
+      "step": 365
+    },
+    {
+      "epoch": 0.5966742745353766,
+      "grad_norm": 0.15075908601284027,
+      "learning_rate": 0.0001,
+      "loss": 0.2043,
+      "step": 366
+    },
+    {
+      "epoch": 0.5983045321160744,
+      "grad_norm": 0.14985023438930511,
+      "learning_rate": 0.0001,
+      "loss": 0.2274,
+      "step": 367
+    },
+    {
+      "epoch": 0.599934789696772,
+      "grad_norm": 0.13255703449249268,
+      "learning_rate": 0.0001,
+      "loss": 0.2174,
+      "step": 368
+    },
+    {
+      "epoch": 0.6015650472774698,
+      "grad_norm": 0.11377043277025223,
+      "learning_rate": 0.0001,
+      "loss": 0.2285,
+      "step": 369
+    },
+    {
+      "epoch": 0.6031953048581676,
+      "grad_norm": 0.10583726316690445,
+      "learning_rate": 0.0001,
+      "loss": 0.2111,
+      "step": 370
+    },
+    {
+      "epoch": 0.6048255624388653,
+      "grad_norm": 0.10989916324615479,
+      "learning_rate": 0.0001,
+      "loss": 0.2277,
+      "step": 371
+    },
+    {
+      "epoch": 0.6064558200195631,
+      "grad_norm": 0.13273194432258606,
+      "learning_rate": 0.0001,
+      "loss": 0.2269,
+      "step": 372
+    },
+    {
+      "epoch": 0.6080860776002608,
+      "grad_norm": 0.12810608744621277,
+      "learning_rate": 0.0001,
+      "loss": 0.2369,
+      "step": 373
+    },
+    {
+      "epoch": 0.6097163351809586,
+      "grad_norm": 0.11391692608594894,
+      "learning_rate": 0.0001,
+      "loss": 0.2178,
+      "step": 374
+    },
+    {
+      "epoch": 0.6113465927616564,
+      "grad_norm": 0.18068620562553406,
+      "learning_rate": 0.0001,
+      "loss": 0.2257,
+      "step": 375
+    },
+    {
+      "epoch": 0.612976850342354,
+      "grad_norm": 0.11202447861433029,
+      "learning_rate": 0.0001,
+      "loss": 0.1979,
+      "step": 376
+    },
+    {
+      "epoch": 0.6146071079230518,
+      "grad_norm": 0.13879020512104034,
+      "learning_rate": 0.0001,
+      "loss": 0.2235,
+      "step": 377
+    },
+    {
+      "epoch": 0.6162373655037496,
+      "grad_norm": 0.11466659605503082,
+      "learning_rate": 0.0001,
+      "loss": 0.2146,
+      "step": 378
+    },
+    {
+      "epoch": 0.6178676230844473,
+      "grad_norm": 0.13301534950733185,
+      "learning_rate": 0.0001,
+      "loss": 0.2374,
+      "step": 379
+    },
+    {
+      "epoch": 0.6194978806651451,
+      "grad_norm": 0.10728432238101959,
+      "learning_rate": 0.0001,
+      "loss": 0.2002,
+      "step": 380
+    },
+    {
+      "epoch": 0.6211281382458429,
+      "grad_norm": 0.11212770640850067,
+      "learning_rate": 0.0001,
+      "loss": 0.2232,
+      "step": 381
+    },
+    {
+      "epoch": 0.6227583958265406,
+      "grad_norm": 0.12755908071994781,
+      "learning_rate": 0.0001,
+      "loss": 0.2284,
+      "step": 382
+    },
+    {
+      "epoch": 0.6243886534072384,
+      "grad_norm": 0.11284497380256653,
+      "learning_rate": 0.0001,
+      "loss": 0.2235,
+      "step": 383
+    },
+    {
+      "epoch": 0.626018910987936,
+      "grad_norm": 0.13458143174648285,
+      "learning_rate": 0.0001,
+      "loss": 0.2183,
+      "step": 384
+    },
+    {
+      "epoch": 0.6276491685686338,
+      "grad_norm": 0.13338471949100494,
+      "learning_rate": 0.0001,
+      "loss": 0.2222,
+      "step": 385
+    },
+    {
+      "epoch": 0.6292794261493316,
+      "grad_norm": 0.11197499930858612,
+      "learning_rate": 0.0001,
+      "loss": 0.2184,
+      "step": 386
+    },
+    {
+      "epoch": 0.6309096837300293,
+      "grad_norm": 0.11694587767124176,
+      "learning_rate": 0.0001,
+      "loss": 0.2231,
+      "step": 387
+    },
+    {
+      "epoch": 0.6325399413107271,
+      "grad_norm": 0.12493859976530075,
+      "learning_rate": 0.0001,
+      "loss": 0.2138,
+      "step": 388
+    },
+    {
+      "epoch": 0.6341701988914249,
+      "grad_norm": 0.12846441566944122,
+      "learning_rate": 0.0001,
+      "loss": 0.2244,
+      "step": 389
+    },
+    {
+      "epoch": 0.6358004564721226,
+      "grad_norm": 0.11621380597352982,
+      "learning_rate": 0.0001,
+      "loss": 0.2232,
+      "step": 390
+    },
+    {
+      "epoch": 0.6374307140528204,
+      "grad_norm": 0.12881864607334137,
+      "learning_rate": 0.0001,
+      "loss": 0.2076,
+      "step": 391
+    },
+    {
+      "epoch": 0.639060971633518,
+      "grad_norm": 0.14580029249191284,
+      "learning_rate": 0.0001,
+      "loss": 0.222,
+      "step": 392
+    },
+    {
+      "epoch": 0.6406912292142158,
+      "grad_norm": 0.11677811294794083,
+      "learning_rate": 0.0001,
+      "loss": 0.2239,
+      "step": 393
+    },
+    {
+      "epoch": 0.6423214867949136,
+      "grad_norm": 0.13098081946372986,
+      "learning_rate": 0.0001,
+      "loss": 0.2217,
+      "step": 394
+    },
+    {
+      "epoch": 0.6439517443756113,
+      "grad_norm": 0.13351522386074066,
+      "learning_rate": 0.0001,
+      "loss": 0.2129,
+      "step": 395
+    },
+    {
+      "epoch": 0.6455820019563091,
+      "grad_norm": 0.10511767864227295,
+      "learning_rate": 0.0001,
+      "loss": 0.1971,
+      "step": 396
+    },
+    {
+      "epoch": 0.6472122595370069,
+      "grad_norm": 0.14180706441402435,
+      "learning_rate": 0.0001,
+      "loss": 0.2195,
+      "step": 397
+    },
+    {
+      "epoch": 0.6488425171177046,
+      "grad_norm": 0.1285708099603653,
+      "learning_rate": 0.0001,
+      "loss": 0.2371,
+      "step": 398
+    },
+    {
+      "epoch": 0.6504727746984024,
+      "grad_norm": 0.11256030201911926,
+      "learning_rate": 0.0001,
+      "loss": 0.2181,
+      "step": 399
+    },
+    {
+      "epoch": 0.6521030322791,
+      "grad_norm": 0.13401979207992554,
+      "learning_rate": 0.0001,
+      "loss": 0.2226,
+      "step": 400
+    },
+    {
+      "epoch": 0.6537332898597978,
+      "grad_norm": 0.1150553822517395,
+      "learning_rate": 0.0001,
+      "loss": 0.2307,
+      "step": 401
+    },
+    {
+      "epoch": 0.6553635474404956,
+      "grad_norm": 0.12015953660011292,
+      "learning_rate": 0.0001,
+      "loss": 0.2107,
+      "step": 402
+    },
+    {
+      "epoch": 0.6569938050211933,
+      "grad_norm": 0.10118319094181061,
+      "learning_rate": 0.0001,
+      "loss": 0.2045,
+      "step": 403
+    },
+    {
+      "epoch": 0.6586240626018911,
+      "grad_norm": 0.12295553833246231,
+      "learning_rate": 0.0001,
+      "loss": 0.2206,
+      "step": 404
+    },
+    {
+      "epoch": 0.6602543201825889,
+      "grad_norm": 0.11807361245155334,
+      "learning_rate": 0.0001,
+      "loss": 0.2197,
+      "step": 405
+    },
+    {
+      "epoch": 0.6618845777632866,
+      "grad_norm": 0.12740248441696167,
+      "learning_rate": 0.0001,
+      "loss": 0.2172,
+      "step": 406
+    },
+    {
+      "epoch": 0.6635148353439844,
+      "grad_norm": 0.12042457610368729,
+      "learning_rate": 0.0001,
+      "loss": 0.2225,
+      "step": 407
+    },
+    {
+      "epoch": 0.665145092924682,
+      "grad_norm": 0.17038275301456451,
+      "learning_rate": 0.0001,
+      "loss": 0.233,
+      "step": 408
+    },
+    {
+      "epoch": 0.6667753505053798,
+      "grad_norm": 0.11151735484600067,
+      "learning_rate": 0.0001,
+      "loss": 0.2013,
+      "step": 409
+    },
+    {
+      "epoch": 0.6684056080860776,
+      "grad_norm": 0.12992088496685028,
+      "learning_rate": 0.0001,
+      "loss": 0.2128,
+      "step": 410
+    },
+    {
+      "epoch": 0.6700358656667753,
+      "grad_norm": 0.10836778581142426,
+      "learning_rate": 0.0001,
+      "loss": 0.2187,
+      "step": 411
+    },
+    {
+      "epoch": 0.6716661232474731,
+      "grad_norm": 0.128456249833107,
+      "learning_rate": 0.0001,
+      "loss": 0.2078,
+      "step": 412
+    },
+    {
+      "epoch": 0.6732963808281709,
+      "grad_norm": 0.12796828150749207,
+      "learning_rate": 0.0001,
+      "loss": 0.2222,
+      "step": 413
+    },
+    {
+      "epoch": 0.6749266384088686,
+      "grad_norm": 0.11841002106666565,
+      "learning_rate": 0.0001,
+      "loss": 0.2127,
+      "step": 414
+    },
+    {
+      "epoch": 0.6765568959895664,
+      "grad_norm": 0.11400352418422699,
+      "learning_rate": 0.0001,
+      "loss": 0.2004,
+      "step": 415
+    },
+    {
+      "epoch": 0.6781871535702642,
+      "grad_norm": 0.1325376331806183,
+      "learning_rate": 0.0001,
+      "loss": 0.218,
+      "step": 416
+    },
+    {
+      "epoch": 0.6798174111509618,
+      "grad_norm": 0.128141388297081,
+      "learning_rate": 0.0001,
+      "loss": 0.2095,
+      "step": 417
+    },
+    {
+      "epoch": 0.6814476687316596,
+      "grad_norm": 0.1291760355234146,
+      "learning_rate": 0.0001,
+      "loss": 0.2231,
+      "step": 418
+    },
+    {
+      "epoch": 0.6830779263123573,
+      "grad_norm": 0.12524424493312836,
+      "learning_rate": 0.0001,
+      "loss": 0.2143,
+      "step": 419
+    },
+    {
+      "epoch": 0.6847081838930551,
+      "grad_norm": 0.128330260515213,
+      "learning_rate": 0.0001,
+      "loss": 0.199,
+      "step": 420
+    },
+    {
+      "epoch": 0.6863384414737529,
+      "grad_norm": 0.12733139097690582,
+      "learning_rate": 0.0001,
+      "loss": 0.2145,
+      "step": 421
+    },
+    {
+      "epoch": 0.6879686990544506,
+      "grad_norm": 0.10170670598745346,
+      "learning_rate": 0.0001,
+      "loss": 0.2053,
+      "step": 422
+    },
+    {
+      "epoch": 0.6895989566351484,
+      "grad_norm": 0.11686375737190247,
+      "learning_rate": 0.0001,
+      "loss": 0.2193,
+      "step": 423
+    },
+    {
+      "epoch": 0.6912292142158462,
+      "grad_norm": 0.1116994172334671,
+      "learning_rate": 0.0001,
+      "loss": 0.22,
+      "step": 424
+    },
+    {
+      "epoch": 0.6928594717965438,
+      "grad_norm": 0.10491323471069336,
+      "learning_rate": 0.0001,
+      "loss": 0.2127,
+      "step": 425
+    },
+    {
+      "epoch": 0.6944897293772416,
+      "grad_norm": 0.10770369321107864,
+      "learning_rate": 0.0001,
+      "loss": 0.2046,
+      "step": 426
+    },
+    {
+      "epoch": 0.6961199869579393,
+      "grad_norm": 0.12084191292524338,
+      "learning_rate": 0.0001,
+      "loss": 0.2174,
+      "step": 427
+    },
+    {
+      "epoch": 0.6977502445386371,
+      "grad_norm": 0.12580984830856323,
+      "learning_rate": 0.0001,
+      "loss": 0.2173,
+      "step": 428
+    },
+    {
+      "epoch": 0.6993805021193349,
+      "grad_norm": 0.1020190566778183,
+      "learning_rate": 0.0001,
+      "loss": 0.213,
+      "step": 429
+    },
+    {
+      "epoch": 0.7010107597000326,
+      "grad_norm": 0.10437814146280289,
+      "learning_rate": 0.0001,
+      "loss": 0.2187,
+      "step": 430
+    },
+    {
+      "epoch": 0.7026410172807304,
+      "grad_norm": 0.13456971943378448,
+      "learning_rate": 0.0001,
+      "loss": 0.2153,
+      "step": 431
+    },
+    {
+      "epoch": 0.7042712748614282,
+      "grad_norm": 0.12021128088235855,
+      "learning_rate": 0.0001,
+      "loss": 0.2146,
+      "step": 432
+    },
+    {
+      "epoch": 0.7059015324421258,
+      "grad_norm": 0.15099307894706726,
+      "learning_rate": 0.0001,
+      "loss": 0.2154,
+      "step": 433
+    },
+    {
+      "epoch": 0.7075317900228236,
+      "grad_norm": 0.11958514899015427,
+      "learning_rate": 0.0001,
+      "loss": 0.216,
+      "step": 434
+    },
+    {
+      "epoch": 0.7091620476035213,
+      "grad_norm": 0.11153008043766022,
+      "learning_rate": 0.0001,
+      "loss": 0.2034,
+      "step": 435
+    },
+    {
+      "epoch": 0.7107923051842191,
+      "grad_norm": 0.14314942061901093,
+      "learning_rate": 0.0001,
+      "loss": 0.2005,
+      "step": 436
+    },
+    {
+      "epoch": 0.7124225627649169,
+      "grad_norm": 0.10860710591077805,
+      "learning_rate": 0.0001,
+      "loss": 0.21,
+      "step": 437
+    },
+    {
+      "epoch": 0.7140528203456146,
+      "grad_norm": 0.1300295740365982,
+      "learning_rate": 0.0001,
+      "loss": 0.2148,
+      "step": 438
+    },
+    {
+      "epoch": 0.7156830779263124,
+      "grad_norm": 0.11765281856060028,
+      "learning_rate": 0.0001,
+      "loss": 0.2144,
+      "step": 439
+    },
+    {
+      "epoch": 0.7173133355070102,
+      "grad_norm": 0.1349002867937088,
+      "learning_rate": 0.0001,
+      "loss": 0.2079,
+      "step": 440
+    },
+    {
+      "epoch": 0.7189435930877078,
+      "grad_norm": 0.15499438345432281,
+      "learning_rate": 0.0001,
+      "loss": 0.2223,
+      "step": 441
+    },
+    {
+      "epoch": 0.7205738506684056,
+      "grad_norm": 0.15518735349178314,
+      "learning_rate": 0.0001,
+      "loss": 0.2317,
+      "step": 442
+    },
+    {
+      "epoch": 0.7222041082491033,
+      "grad_norm": 0.16259780526161194,
+      "learning_rate": 0.0001,
+      "loss": 0.2194,
+      "step": 443
+    },
+    {
+      "epoch": 0.7238343658298011,
+      "grad_norm": 0.13394394516944885,
+      "learning_rate": 0.0001,
+      "loss": 0.2175,
+      "step": 444
+    },
+    {
+      "epoch": 0.7254646234104989,
+      "grad_norm": 0.1212131530046463,
+      "learning_rate": 0.0001,
+      "loss": 0.2172,
+      "step": 445
+    },
+    {
+      "epoch": 0.7270948809911966,
+      "grad_norm": 0.1333845555782318,
+      "learning_rate": 0.0001,
+      "loss": 0.2125,
+      "step": 446
+    },
+    {
+      "epoch": 0.7287251385718944,
+      "grad_norm": 0.1470658779144287,
+      "learning_rate": 0.0001,
+      "loss": 0.2263,
+      "step": 447
+    },
+    {
+      "epoch": 0.7303553961525922,
+      "grad_norm": 0.11653272062540054,
+      "learning_rate": 0.0001,
+      "loss": 0.2203,
+      "step": 448
+    },
+    {
+      "epoch": 0.7319856537332898,
+      "grad_norm": 0.11255240440368652,
+      "learning_rate": 0.0001,
+      "loss": 0.2063,
+      "step": 449
+    },
+    {
+      "epoch": 0.7336159113139876,
+      "grad_norm": 0.1200360506772995,
+      "learning_rate": 0.0001,
+      "loss": 0.217,
+      "step": 450
+    },
+    {
+      "epoch": 0.7352461688946854,
+      "grad_norm": 0.12475258111953735,
+      "learning_rate": 0.0001,
+      "loss": 0.2232,
+      "step": 451
+    },
+    {
+      "epoch": 0.7368764264753831,
+      "grad_norm": 0.11277184635400772,
+      "learning_rate": 0.0001,
+      "loss": 0.2314,
+      "step": 452
+    },
+    {
+      "epoch": 0.7385066840560809,
+      "grad_norm": 0.11270298063755035,
+      "learning_rate": 0.0001,
+      "loss": 0.2112,
+      "step": 453
+    },
+    {
+      "epoch": 0.7401369416367786,
+      "grad_norm": 0.12822334468364716,
+      "learning_rate": 0.0001,
+      "loss": 0.2233,
+      "step": 454
+    },
+    {
+      "epoch": 0.7417671992174764,
+      "grad_norm": 0.11773821711540222,
+      "learning_rate": 0.0001,
+      "loss": 0.2213,
+      "step": 455
+    },
+    {
+      "epoch": 0.7433974567981741,
+      "grad_norm": 0.12635211646556854,
+      "learning_rate": 0.0001,
+      "loss": 0.2001,
+      "step": 456
+    },
+    {
+      "epoch": 0.7450277143788718,
+      "grad_norm": 0.15887145698070526,
+      "learning_rate": 0.0001,
+      "loss": 0.2276,
+      "step": 457
+    },
+    {
+      "epoch": 0.7466579719595696,
+      "grad_norm": 0.12271532416343689,
+      "learning_rate": 0.0001,
+      "loss": 0.219,
+      "step": 458
+    },
+    {
+      "epoch": 0.7482882295402674,
+      "grad_norm": 0.14966937899589539,
+      "learning_rate": 0.0001,
+      "loss": 0.2119,
+      "step": 459
+    },
+    {
+      "epoch": 0.7499184871209651,
+      "grad_norm": 0.15153633058071136,
+      "learning_rate": 0.0001,
+      "loss": 0.2316,
+      "step": 460
+    },
+    {
+      "epoch": 0.7515487447016629,
+      "grad_norm": 0.14428603649139404,
+      "learning_rate": 0.0001,
+      "loss": 0.2197,
+      "step": 461
+    },
+    {
+      "epoch": 0.7531790022823606,
+      "grad_norm": 0.12772101163864136,
+      "learning_rate": 0.0001,
+      "loss": 0.2159,
+      "step": 462
+    },
+    {
+      "epoch": 0.7548092598630584,
+      "grad_norm": 0.12549546360969543,
+      "learning_rate": 0.0001,
+      "loss": 0.2176,
+      "step": 463
+    },
+    {
+      "epoch": 0.7564395174437561,
+      "grad_norm": 0.13936737179756165,
+      "learning_rate": 0.0001,
+      "loss": 0.2187,
+      "step": 464
+    },
+    {
+      "epoch": 0.7580697750244538,
+      "grad_norm": 0.11733359098434448,
+      "learning_rate": 0.0001,
+      "loss": 0.2013,
+      "step": 465
+    },
+    {
+      "epoch": 0.7597000326051516,
+      "grad_norm": 0.14219063520431519,
+      "learning_rate": 0.0001,
+      "loss": 0.2263,
+      "step": 466
+    },
+    {
+      "epoch": 0.7613302901858494,
+      "grad_norm": 0.11690539866685867,
+      "learning_rate": 0.0001,
+      "loss": 0.2108,
+      "step": 467
+    },
+    {
+      "epoch": 0.7629605477665471,
+      "grad_norm": 0.13221612572669983,
+      "learning_rate": 0.0001,
+      "loss": 0.2051,
+      "step": 468
+    },
+    {
+      "epoch": 0.7645908053472449,
+      "grad_norm": 0.14684675633907318,
+      "learning_rate": 0.0001,
+      "loss": 0.2168,
+      "step": 469
+    },
+    {
+      "epoch": 0.7662210629279426,
+      "grad_norm": 0.1189989224076271,
+      "learning_rate": 0.0001,
+      "loss": 0.2204,
+      "step": 470
+    },
+    {
+      "epoch": 0.7678513205086404,
+      "grad_norm": 0.12807010114192963,
+      "learning_rate": 0.0001,
+      "loss": 0.2285,
+      "step": 471
+    },
+    {
+      "epoch": 0.7694815780893381,
+      "grad_norm": 0.13275974988937378,
+      "learning_rate": 0.0001,
+      "loss": 0.219,
+      "step": 472
+    },
+    {
+      "epoch": 0.7711118356700358,
+      "grad_norm": 0.1249246746301651,
+      "learning_rate": 0.0001,
+      "loss": 0.2218,
+      "step": 473
+    },
+    {
+      "epoch": 0.7727420932507336,
+      "grad_norm": 0.13339447975158691,
+      "learning_rate": 0.0001,
+      "loss": 0.2068,
+      "step": 474
+    },
+    {
+      "epoch": 0.7743723508314314,
+      "grad_norm": 0.12550900876522064,
+      "learning_rate": 0.0001,
+      "loss": 0.2103,
+      "step": 475
+    },
+    {
+      "epoch": 0.7760026084121291,
+      "grad_norm": 0.11490708589553833,
+      "learning_rate": 0.0001,
+      "loss": 0.2033,
+      "step": 476
+    },
+    {
+      "epoch": 0.7776328659928269,
+      "grad_norm": 0.12597325444221497,
+      "learning_rate": 0.0001,
+      "loss": 0.2135,
+      "step": 477
+    },
+    {
+      "epoch": 0.7792631235735246,
+      "grad_norm": 0.15443576872348785,
+      "learning_rate": 0.0001,
+      "loss": 0.2133,
+      "step": 478
+    },
+    {
+      "epoch": 0.7808933811542224,
+      "grad_norm": 0.19461177289485931,
+      "learning_rate": 0.0001,
+      "loss": 0.1955,
+      "step": 479
+    },
+    {
+      "epoch": 0.7825236387349201,
+      "grad_norm": 0.1330460160970688,
+      "learning_rate": 0.0001,
+      "loss": 0.1979,
+      "step": 480
+    },
+    {
+      "epoch": 0.7841538963156178,
+      "grad_norm": 0.11665869504213333,
+      "learning_rate": 0.0001,
+      "loss": 0.2149,
+      "step": 481
+    },
+    {
+      "epoch": 0.7857841538963156,
+      "grad_norm": 0.09942878782749176,
+      "learning_rate": 0.0001,
+      "loss": 0.2009,
+      "step": 482
+    },
+    {
+      "epoch": 0.7874144114770134,
+      "grad_norm": 0.11635838449001312,
+      "learning_rate": 0.0001,
+      "loss": 0.2046,
+      "step": 483
+    },
+    {
+      "epoch": 0.7890446690577111,
+      "grad_norm": 0.11658889055252075,
+      "learning_rate": 0.0001,
+      "loss": 0.2154,
+      "step": 484
+    },
+    {
+      "epoch": 0.7906749266384089,
+      "grad_norm": 0.122380830347538,
+      "learning_rate": 0.0001,
+      "loss": 0.2003,
+      "step": 485
+    },
+    {
+      "epoch": 0.7923051842191067,
+      "grad_norm": 0.11796704679727554,
+      "learning_rate": 0.0001,
+      "loss": 0.1966,
+      "step": 486
+    },
+    {
+      "epoch": 0.7939354417998044,
+      "grad_norm": 0.10663303732872009,
+      "learning_rate": 0.0001,
+      "loss": 0.2166,
+      "step": 487
+    },
+    {
+      "epoch": 0.7955656993805021,
+      "grad_norm": 0.11099519580602646,
+      "learning_rate": 0.0001,
+      "loss": 0.209,
+      "step": 488
+    },
+    {
+      "epoch": 0.7971959569611998,
+      "grad_norm": 0.14217236638069153,
+      "learning_rate": 0.0001,
+      "loss": 0.1955,
+      "step": 489
+    },
+    {
+      "epoch": 0.7988262145418976,
+      "grad_norm": 0.10185975581407547,
+      "learning_rate": 0.0001,
+      "loss": 0.194,
+      "step": 490
+    },
+    {
+      "epoch": 0.8004564721225954,
+      "grad_norm": 0.11912322789430618,
+      "learning_rate": 0.0001,
+      "loss": 0.1987,
+      "step": 491
+    },
+    {
+      "epoch": 0.8020867297032931,
+      "grad_norm": 0.11036073416471481,
+      "learning_rate": 0.0001,
+      "loss": 0.199,
+      "step": 492
+    },
+    {
+      "epoch": 0.8037169872839909,
+      "grad_norm": 0.13446538150310516,
+      "learning_rate": 0.0001,
+      "loss": 0.214,
+      "step": 493
+    },
+    {
+      "epoch": 0.8053472448646887,
+      "grad_norm": 0.12443588674068451,
+      "learning_rate": 0.0001,
+      "loss": 0.2067,
+      "step": 494
+    },
+    {
+      "epoch": 0.8069775024453864,
+      "grad_norm": 0.12884607911109924,
+      "learning_rate": 0.0001,
+      "loss": 0.2093,
+      "step": 495
+    },
+    {
+      "epoch": 0.8086077600260841,
+      "grad_norm": 0.1282474249601364,
+      "learning_rate": 0.0001,
+      "loss": 0.1896,
+      "step": 496
+    },
+    {
+      "epoch": 0.8102380176067818,
+      "grad_norm": 0.15667201578617096,
+      "learning_rate": 0.0001,
+      "loss": 0.2275,
+      "step": 497
+    },
+    {
+      "epoch": 0.8118682751874796,
+      "grad_norm": 0.12083553522825241,
+      "learning_rate": 0.0001,
+      "loss": 0.2061,
+      "step": 498
+    },
+    {
+      "epoch": 0.8134985327681774,
+      "grad_norm": 0.1412494033575058,
+      "learning_rate": 0.0001,
+      "loss": 0.2288,
+      "step": 499
+    },
+    {
+      "epoch": 0.8151287903488751,
+      "grad_norm": 0.11393098533153534,
+      "learning_rate": 0.0001,
+      "loss": 0.1828,
+      "step": 500
+    },
+    {
+      "epoch": 0.8151287903488751,
+      "eval_loss": 0.21177859604358673,
+      "eval_runtime": 2800.3476,
+      "eval_samples_per_second": 0.674,
+      "eval_steps_per_second": 0.169,
+      "step": 500
+    },
+    {
+      "epoch": 0.8167590479295729,
+      "grad_norm": 0.14163638651371002,
+      "learning_rate": 0.0001,
+      "loss": 0.2251,
+      "step": 501
+    },
+    {
+      "epoch": 0.8183893055102707,
+      "grad_norm": 0.14434993267059326,
+      "learning_rate": 0.0001,
+      "loss": 0.2147,
+      "step": 502
+    },
+    {
+      "epoch": 0.8200195630909684,
+      "grad_norm": 0.11927150934934616,
+      "learning_rate": 0.0001,
+      "loss": 0.1887,
+      "step": 503
+    },
+    {
+      "epoch": 0.8216498206716661,
+      "grad_norm": 0.12366633117198944,
+      "learning_rate": 0.0001,
+      "loss": 0.1949,
+      "step": 504
+    },
+    {
+      "epoch": 0.8232800782523638,
+      "grad_norm": 0.1395760029554367,
+      "learning_rate": 0.0001,
+      "loss": 0.1998,
+      "step": 505
+    },
+    {
+      "epoch": 0.8249103358330616,
+      "grad_norm": 0.1172211617231369,
+      "learning_rate": 0.0001,
+      "loss": 0.196,
+      "step": 506
+    },
+    {
+      "epoch": 0.8265405934137594,
+      "grad_norm": 0.14511209726333618,
+      "learning_rate": 0.0001,
+      "loss": 0.2301,
+      "step": 507
+    },
+    {
+      "epoch": 0.8281708509944571,
+      "grad_norm": 0.17221054434776306,
+      "learning_rate": 0.0001,
+      "loss": 0.2243,
+      "step": 508
+    },
+    {
+      "epoch": 0.8298011085751549,
+      "grad_norm": 0.1269037127494812,
+      "learning_rate": 0.0001,
+      "loss": 0.2086,
+      "step": 509
+    },
+    {
+      "epoch": 0.8314313661558527,
+      "grad_norm": 0.13336747884750366,
+      "learning_rate": 0.0001,
+      "loss": 0.2085,
+      "step": 510
+    },
+    {
+      "epoch": 0.8330616237365503,
+      "grad_norm": 0.13246330618858337,
+      "learning_rate": 0.0001,
+      "loss": 0.2264,
+      "step": 511
+    },
+    {
+      "epoch": 0.8346918813172481,
+      "grad_norm": 0.11336122453212738,
+      "learning_rate": 0.0001,
+      "loss": 0.204,
+      "step": 512
+    },
+    {
+      "epoch": 0.8363221388979458,
+      "grad_norm": 0.12755842506885529,
+      "learning_rate": 0.0001,
+      "loss": 0.2088,
+      "step": 513
+    },
+    {
+      "epoch": 0.8379523964786436,
+      "grad_norm": 0.11611583828926086,
+      "learning_rate": 0.0001,
+      "loss": 0.2037,
+      "step": 514
+    },
+    {
+      "epoch": 0.8395826540593414,
+      "grad_norm": 0.09402882307767868,
+      "learning_rate": 0.0001,
+      "loss": 0.1939,
+      "step": 515
+    },
+    {
+      "epoch": 0.8412129116400391,
+      "grad_norm": 0.12100663781166077,
+      "learning_rate": 0.0001,
+      "loss": 0.2104,
+      "step": 516
+    },
+    {
+      "epoch": 0.8428431692207369,
+      "grad_norm": 0.14289285242557526,
+      "learning_rate": 0.0001,
+      "loss": 0.2216,
+      "step": 517
+    },
+    {
+      "epoch": 0.8444734268014347,
+      "grad_norm": 0.10300295054912567,
+      "learning_rate": 0.0001,
+      "loss": 0.1796,
+      "step": 518
+    },
+    {
+      "epoch": 0.8461036843821323,
+      "grad_norm": 0.12143992632627487,
+      "learning_rate": 0.0001,
+      "loss": 0.2177,
+      "step": 519
+    },
+    {
+      "epoch": 0.8477339419628301,
+      "grad_norm": 0.12010245770215988,
+      "learning_rate": 0.0001,
+      "loss": 0.2016,
+      "step": 520
+    },
+    {
+      "epoch": 0.8493641995435279,
+      "grad_norm": 0.1474478542804718,
+      "learning_rate": 0.0001,
+      "loss": 0.2062,
+      "step": 521
+    },
+    {
+      "epoch": 0.8509944571242256,
+      "grad_norm": 0.12784548103809357,
+      "learning_rate": 0.0001,
+      "loss": 0.2317,
+      "step": 522
+    },
+    {
+      "epoch": 0.8526247147049234,
+      "grad_norm": 0.11985955387353897,
+      "learning_rate": 0.0001,
+      "loss": 0.2196,
+      "step": 523
+    },
+    {
+      "epoch": 0.8542549722856211,
+      "grad_norm": 0.11471915990114212,
+      "learning_rate": 0.0001,
+      "loss": 0.2082,
+      "step": 524
+    },
+    {
+      "epoch": 0.8558852298663189,
+      "grad_norm": 0.135053813457489,
+      "learning_rate": 0.0001,
+      "loss": 0.215,
+      "step": 525
+    },
+    {
+      "epoch": 0.8575154874470167,
+      "grad_norm": 0.10875121504068375,
+      "learning_rate": 0.0001,
+      "loss": 0.2143,
+      "step": 526
+    },
+    {
+      "epoch": 0.8591457450277143,
+      "grad_norm": 0.11824619024991989,
+      "learning_rate": 0.0001,
+      "loss": 0.2276,
+      "step": 527
+    },
+    {
+      "epoch": 0.8607760026084121,
+      "grad_norm": 0.10967472940683365,
+      "learning_rate": 0.0001,
+      "loss": 0.1933,
+      "step": 528
+    },
+    {
+      "epoch": 0.8624062601891099,
+      "grad_norm": 0.09989364445209503,
+      "learning_rate": 0.0001,
+      "loss": 0.2196,
+      "step": 529
+    },
+    {
+      "epoch": 0.8640365177698076,
+      "grad_norm": 0.10137049108743668,
+      "learning_rate": 0.0001,
+      "loss": 0.1993,
+      "step": 530
+    },
+    {
+      "epoch": 0.8656667753505054,
+      "grad_norm": 0.13142383098602295,
+      "learning_rate": 0.0001,
+      "loss": 0.2221,
+      "step": 531
+    },
+    {
+      "epoch": 0.8672970329312031,
+      "grad_norm": 0.1282247006893158,
+      "learning_rate": 0.0001,
+      "loss": 0.2167,
+      "step": 532
+    },
+    {
+      "epoch": 0.8689272905119009,
+      "grad_norm": 0.10532195121049881,
+      "learning_rate": 0.0001,
+      "loss": 0.2143,
+      "step": 533
+    },
+    {
+      "epoch": 0.8705575480925987,
+      "grad_norm": 0.10713239759206772,
+      "learning_rate": 0.0001,
+      "loss": 0.2045,
+      "step": 534
+    },
+    {
+      "epoch": 0.8721878056732963,
+      "grad_norm": 0.12712952494621277,
+      "learning_rate": 0.0001,
+      "loss": 0.2264,
+      "step": 535
+    },
+    {
+      "epoch": 0.8738180632539941,
+      "grad_norm": 0.12850171327590942,
+      "learning_rate": 0.0001,
+      "loss": 0.2284,
+      "step": 536
+    },
+    {
+      "epoch": 0.8754483208346919,
+      "grad_norm": 0.12360873818397522,
+      "learning_rate": 0.0001,
+      "loss": 0.197,
+      "step": 537
+    },
+    {
+      "epoch": 0.8770785784153896,
+      "grad_norm": 0.12052212655544281,
+      "learning_rate": 0.0001,
+      "loss": 0.1906,
+      "step": 538
+    },
+    {
+      "epoch": 0.8787088359960874,
+      "grad_norm": 0.14144307374954224,
+      "learning_rate": 0.0001,
+      "loss": 0.2131,
+      "step": 539
+    },
+    {
+      "epoch": 0.8803390935767851,
+      "grad_norm": 0.11279332637786865,
+      "learning_rate": 0.0001,
+      "loss": 0.2145,
+      "step": 540
+    },
+    {
+      "epoch": 0.8819693511574829,
+      "grad_norm": 0.161105215549469,
+      "learning_rate": 0.0001,
+      "loss": 0.2212,
+      "step": 541
+    },
+    {
+      "epoch": 0.8835996087381807,
+      "grad_norm": 0.11746472120285034,
+      "learning_rate": 0.0001,
+      "loss": 0.213,
+      "step": 542
+    },
+    {
+      "epoch": 0.8852298663188783,
+      "grad_norm": 0.1244741752743721,
+      "learning_rate": 0.0001,
+      "loss": 0.2065,
+      "step": 543
+    },
+    {
+      "epoch": 0.8868601238995761,
+      "grad_norm": 0.1004570797085762,
+      "learning_rate": 0.0001,
+      "loss": 0.2086,
+      "step": 544
+    },
+    {
+      "epoch": 0.8884903814802739,
+      "grad_norm": 0.10560411214828491,
+      "learning_rate": 0.0001,
+      "loss": 0.204,
+      "step": 545
+    },
+    {
+      "epoch": 0.8901206390609716,
+      "grad_norm": 0.11466600000858307,
+      "learning_rate": 0.0001,
+      "loss": 0.2012,
+      "step": 546
+    },
+    {
+      "epoch": 0.8917508966416694,
+      "grad_norm": 0.11946078389883041,
+      "learning_rate": 0.0001,
+      "loss": 0.21,
+      "step": 547
+    },
+    {
+      "epoch": 0.8933811542223671,
+      "grad_norm": 0.13734155893325806,
+      "learning_rate": 0.0001,
+      "loss": 0.2166,
+      "step": 548
+    },
+    {
+      "epoch": 0.8950114118030649,
+      "grad_norm": 0.11657248437404633,
+      "learning_rate": 0.0001,
+      "loss": 0.2106,
+      "step": 549
+    },
+    {
+      "epoch": 0.8966416693837627,
+      "grad_norm": 0.1141151636838913,
+      "learning_rate": 0.0001,
+      "loss": 0.1827,
+      "step": 550
+    },
+    {
+      "epoch": 0.8982719269644603,
+      "grad_norm": 0.1335282176733017,
+      "learning_rate": 0.0001,
+      "loss": 0.2069,
+      "step": 551
+    },
+    {
+      "epoch": 0.8999021845451581,
+      "grad_norm": 0.12621258199214935,
+      "learning_rate": 0.0001,
+      "loss": 0.1985,
+      "step": 552
+    },
+    {
+      "epoch": 0.9015324421258559,
+      "grad_norm": 0.13729962706565857,
+      "learning_rate": 0.0001,
+      "loss": 0.2068,
+      "step": 553
+    },
+    {
+      "epoch": 0.9031626997065536,
+      "grad_norm": 0.11198980361223221,
+      "learning_rate": 0.0001,
+      "loss": 0.2078,
+      "step": 554
+    },
+    {
+      "epoch": 0.9047929572872514,
+      "grad_norm": 0.12503336369991302,
+      "learning_rate": 0.0001,
+      "loss": 0.2077,
+      "step": 555
+    },
+    {
+      "epoch": 0.9064232148679492,
+      "grad_norm": 0.14169982075691223,
+      "learning_rate": 0.0001,
+      "loss": 0.2052,
+      "step": 556
+    },
+    {
+      "epoch": 0.9080534724486469,
+      "grad_norm": 0.10698267072439194,
+      "learning_rate": 0.0001,
+      "loss": 0.1958,
+      "step": 557
+    },
+    {
+      "epoch": 0.9096837300293447,
+      "grad_norm": 0.1422925740480423,
+      "learning_rate": 0.0001,
+      "loss": 0.1995,
+      "step": 558
+    },
+    {
+      "epoch": 0.9113139876100423,
+      "grad_norm": 0.11455567926168442,
+      "learning_rate": 0.0001,
+      "loss": 0.1998,
+      "step": 559
+    },
+    {
+      "epoch": 0.9129442451907401,
+      "grad_norm": 0.1599031537771225,
+      "learning_rate": 0.0001,
+      "loss": 0.2113,
+      "step": 560
+    },
+    {
+      "epoch": 0.9145745027714379,
+      "grad_norm": 0.13150307536125183,
+      "learning_rate": 0.0001,
+      "loss": 0.217,
+      "step": 561
+    },
+    {
+      "epoch": 0.9162047603521356,
+      "grad_norm": 0.10883089154958725,
+      "learning_rate": 0.0001,
+      "loss": 0.1914,
+      "step": 562
+    },
+    {
+      "epoch": 0.9178350179328334,
+      "grad_norm": 0.12056294828653336,
+      "learning_rate": 0.0001,
+      "loss": 0.2008,
+      "step": 563
+    },
+    {
+      "epoch": 0.9194652755135312,
+      "grad_norm": 0.13235759735107422,
+      "learning_rate": 0.0001,
+      "loss": 0.2019,
+      "step": 564
+    },
+    {
+      "epoch": 0.9210955330942289,
+      "grad_norm": 0.11871761828660965,
+      "learning_rate": 0.0001,
+      "loss": 0.2075,
+      "step": 565
+    },
+    {
+      "epoch": 0.9227257906749267,
+      "grad_norm": 0.13773424923419952,
+      "learning_rate": 0.0001,
+      "loss": 0.2157,
+      "step": 566
+    },
+    {
+      "epoch": 0.9243560482556243,
+      "grad_norm": 0.12058678269386292,
+      "learning_rate": 0.0001,
+      "loss": 0.1872,
+      "step": 567
+    },
+    {
+      "epoch": 0.9259863058363221,
+      "grad_norm": 0.12389074265956879,
+      "learning_rate": 0.0001,
+      "loss": 0.2092,
+      "step": 568
+    },
+    {
+      "epoch": 0.9276165634170199,
+      "grad_norm": 0.13207301497459412,
+      "learning_rate": 0.0001,
+      "loss": 0.2111,
+      "step": 569
+    },
+    {
+      "epoch": 0.9292468209977176,
+      "grad_norm": 0.12088489532470703,
+      "learning_rate": 0.0001,
+      "loss": 0.1886,
+      "step": 570
+    },
+    {
+      "epoch": 0.9308770785784154,
+      "grad_norm": 0.1485392153263092,
+      "learning_rate": 0.0001,
+      "loss": 0.2148,
+      "step": 571
+    },
+    {
+      "epoch": 0.9325073361591132,
+      "grad_norm": 0.15602104365825653,
+      "learning_rate": 0.0001,
+      "loss": 0.2115,
+      "step": 572
+    },
+    {
+      "epoch": 0.9341375937398109,
+      "grad_norm": 0.1451883614063263,
+      "learning_rate": 0.0001,
+      "loss": 0.2147,
+      "step": 573
+    },
+    {
+      "epoch": 0.9357678513205087,
+      "grad_norm": 0.15947164595127106,
+      "learning_rate": 0.0001,
+      "loss": 0.2026,
+      "step": 574
+    },
+    {
+      "epoch": 0.9373981089012063,
+      "grad_norm": 0.15256251394748688,
+      "learning_rate": 0.0001,
+      "loss": 0.209,
+      "step": 575
+    },
+    {
+      "epoch": 0.9390283664819041,
+      "grad_norm": 0.1322648674249649,
+      "learning_rate": 0.0001,
+      "loss": 0.198,
+      "step": 576
+    },
+    {
+      "epoch": 0.9406586240626019,
+      "grad_norm": 0.11925152689218521,
+      "learning_rate": 0.0001,
+      "loss": 0.2055,
+      "step": 577
+    },
+    {
+      "epoch": 0.9422888816432996,
+      "grad_norm": 0.12666535377502441,
+      "learning_rate": 0.0001,
+      "loss": 0.2063,
+      "step": 578
+    },
+    {
+      "epoch": 0.9439191392239974,
+      "grad_norm": 0.10876161605119705,
+      "learning_rate": 0.0001,
+      "loss": 0.2113,
+      "step": 579
+    },
+    {
+      "epoch": 0.9455493968046952,
+      "grad_norm": 0.12513647973537445,
+      "learning_rate": 0.0001,
+      "loss": 0.2093,
+      "step": 580
+    },
+    {
+      "epoch": 0.9471796543853929,
+      "grad_norm": 0.13541190326213837,
+      "learning_rate": 0.0001,
+      "loss": 0.2048,
+      "step": 581
+    },
+    {
+      "epoch": 0.9488099119660907,
+      "grad_norm": 0.12051568925380707,
+      "learning_rate": 0.0001,
+      "loss": 0.2205,
+      "step": 582
+    },
+    {
+      "epoch": 0.9504401695467883,
+      "grad_norm": 0.11748453974723816,
+      "learning_rate": 0.0001,
+      "loss": 0.211,
+      "step": 583
+    },
+    {
+      "epoch": 0.9520704271274861,
+      "grad_norm": 0.11520763486623764,
+      "learning_rate": 0.0001,
+      "loss": 0.2033,
+      "step": 584
+    },
+    {
+      "epoch": 0.9537006847081839,
+      "grad_norm": 0.11155470460653305,
+      "learning_rate": 0.0001,
+      "loss": 0.2004,
+      "step": 585
+    },
+    {
+      "epoch": 0.9553309422888816,
+      "grad_norm": 0.12134213000535965,
+      "learning_rate": 0.0001,
+      "loss": 0.2137,
+      "step": 586
+    },
+    {
+      "epoch": 0.9569611998695794,
+      "grad_norm": 0.1297822743654251,
+      "learning_rate": 0.0001,
+      "loss": 0.2052,
+      "step": 587
+    },
+    {
+      "epoch": 0.9585914574502772,
+      "grad_norm": 0.12923404574394226,
+      "learning_rate": 0.0001,
+      "loss": 0.209,
+      "step": 588
+    },
+    {
+      "epoch": 0.9602217150309749,
+      "grad_norm": 0.20898625254631042,
+      "learning_rate": 0.0001,
+      "loss": 0.2047,
+      "step": 589
+    },
+    {
+      "epoch": 0.9618519726116727,
+      "grad_norm": 0.12227951735258102,
+      "learning_rate": 0.0001,
+      "loss": 0.2107,
+      "step": 590
+    },
+    {
+      "epoch": 0.9634822301923704,
+      "grad_norm": 0.12081098556518555,
+      "learning_rate": 0.0001,
+      "loss": 0.2093,
+      "step": 591
+    },
+    {
+      "epoch": 0.9651124877730681,
+      "grad_norm": 0.13986457884311676,
+      "learning_rate": 0.0001,
+      "loss": 0.2161,
+      "step": 592
+    },
+    {
+      "epoch": 0.9667427453537659,
+      "grad_norm": 0.12605050206184387,
+      "learning_rate": 0.0001,
+      "loss": 0.2153,
+      "step": 593
+    },
+    {
+      "epoch": 0.9683730029344636,
+      "grad_norm": 0.11643191426992416,
+      "learning_rate": 0.0001,
+      "loss": 0.2062,
+      "step": 594
+    },
+    {
+      "epoch": 0.9700032605151614,
+      "grad_norm": 0.12521113455295563,
+      "learning_rate": 0.0001,
+      "loss": 0.2121,
+      "step": 595
+    },
+    {
+      "epoch": 0.9716335180958592,
+      "grad_norm": 0.11723621189594269,
+      "learning_rate": 0.0001,
+      "loss": 0.2053,
+      "step": 596
+    },
+    {
+      "epoch": 0.9732637756765569,
+      "grad_norm": 0.12405609339475632,
+      "learning_rate": 0.0001,
+      "loss": 0.2104,
+      "step": 597
+    },
+    {
+      "epoch": 0.9748940332572547,
+      "grad_norm": 0.10823733359575272,
+      "learning_rate": 0.0001,
+      "loss": 0.2174,
+      "step": 598
+    },
+    {
+      "epoch": 0.9765242908379524,
+      "grad_norm": 0.11131487786769867,
+      "learning_rate": 0.0001,
+      "loss": 0.213,
+      "step": 599
+    },
+    {
+      "epoch": 0.9781545484186501,
+      "grad_norm": 0.13913866877555847,
+      "learning_rate": 0.0001,
+      "loss": 0.2086,
+      "step": 600
+    },
+    {
+      "epoch": 0.9797848059993479,
+      "grad_norm": 0.11523352563381195,
+      "learning_rate": 0.0001,
+      "loss": 0.2103,
+      "step": 601
+    },
+    {
+      "epoch": 0.9814150635800456,
+      "grad_norm": 0.1258571743965149,
+      "learning_rate": 0.0001,
+      "loss": 0.2082,
+      "step": 602
+    },
+    {
+      "epoch": 0.9830453211607434,
+      "grad_norm": 0.14156940579414368,
+      "learning_rate": 0.0001,
+      "loss": 0.2005,
+      "step": 603
+    },
+    {
+      "epoch": 0.9846755787414412,
+      "grad_norm": 0.1144745722413063,
+      "learning_rate": 0.0001,
+      "loss": 0.218,
+      "step": 604
+    },
+    {
+      "epoch": 0.9863058363221389,
+      "grad_norm": 0.12323161959648132,
+      "learning_rate": 0.0001,
+      "loss": 0.2007,
+      "step": 605
+    },
+    {
+      "epoch": 0.9879360939028367,
+      "grad_norm": 0.1369849145412445,
+      "learning_rate": 0.0001,
+      "loss": 0.2187,
+      "step": 606
+    },
+    {
+      "epoch": 0.9895663514835344,
+      "grad_norm": 0.13043704628944397,
+      "learning_rate": 0.0001,
+      "loss": 0.2111,
+      "step": 607
+    },
+    {
+      "epoch": 0.9911966090642321,
+      "grad_norm": 0.11909907311201096,
+      "learning_rate": 0.0001,
+      "loss": 0.2054,
+      "step": 608
+    },
+    {
+      "epoch": 0.9928268666449299,
+      "grad_norm": 0.12159440666437149,
+      "learning_rate": 0.0001,
+      "loss": 0.2168,
+      "step": 609
+    },
+    {
+      "epoch": 0.9944571242256276,
+      "grad_norm": 0.11479892581701279,
+      "learning_rate": 0.0001,
+      "loss": 0.2116,
+      "step": 610
+    },
+    {
+      "epoch": 0.9960873818063254,
+      "grad_norm": 0.12101448327302933,
+      "learning_rate": 0.0001,
+      "loss": 0.2054,
+      "step": 611
+    },
+    {
+      "epoch": 0.9977176393870232,
+      "grad_norm": 0.1083630621433258,
+      "learning_rate": 0.0001,
+      "loss": 0.1981,
+      "step": 612
+    },
+    {
+      "epoch": 0.9993478969677209,
+      "grad_norm": 0.09980759769678116,
+      "learning_rate": 0.0001,
+      "loss": 0.2045,
+      "step": 613
+    },
+    {
+      "epoch": 1.0009781545484187,
+      "grad_norm": 0.1311112940311432,
+      "learning_rate": 0.0001,
+      "loss": 0.2099,
+      "step": 614
+    },
+    {
+      "epoch": 1.0026084121291163,
+      "grad_norm": 0.12701699137687683,
+      "learning_rate": 0.0001,
+      "loss": 0.1989,
+      "step": 615
+    },
+    {
+      "epoch": 1.0042386697098142,
+      "grad_norm": 0.10910508036613464,
+      "learning_rate": 0.0001,
+      "loss": 0.2063,
+      "step": 616
+    },
+    {
+      "epoch": 1.005868927290512,
+      "grad_norm": 0.11964251101016998,
+      "learning_rate": 0.0001,
+      "loss": 0.2046,
+      "step": 617
+    },
+    {
+      "epoch": 1.0074991848712096,
+      "grad_norm": 0.11345476657152176,
+      "learning_rate": 0.0001,
+      "loss": 0.1933,
+      "step": 618
+    },
+    {
+      "epoch": 1.0091294424519075,
+      "grad_norm": 0.11260963976383209,
+      "learning_rate": 0.0001,
+      "loss": 0.2019,
+      "step": 619
+    },
+    {
+      "epoch": 1.0107597000326052,
+      "grad_norm": 0.13378599286079407,
+      "learning_rate": 0.0001,
+      "loss": 0.2074,
+      "step": 620
+    },
+    {
+      "epoch": 1.0123899576133029,
+      "grad_norm": 0.10661476850509644,
+      "learning_rate": 0.0001,
+      "loss": 0.1972,
+      "step": 621
+    },
+    {
+      "epoch": 1.0140202151940008,
+      "grad_norm": 0.12620778381824493,
+      "learning_rate": 0.0001,
+      "loss": 0.2111,
+      "step": 622
+    },
+    {
+      "epoch": 1.0156504727746984,
+      "grad_norm": 0.12133750319480896,
+      "learning_rate": 0.0001,
+      "loss": 0.2022,
+      "step": 623
+    },
+    {
+      "epoch": 1.0172807303553961,
+      "grad_norm": 0.10500383377075195,
+      "learning_rate": 0.0001,
+      "loss": 0.1778,
+      "step": 624
+    },
+    {
+      "epoch": 1.0189109879360938,
+      "grad_norm": 0.12143319845199585,
+      "learning_rate": 0.0001,
+      "loss": 0.211,
+      "step": 625
+    },
+    {
+      "epoch": 1.0205412455167917,
+      "grad_norm": 0.11858654022216797,
+      "learning_rate": 0.0001,
+      "loss": 0.1871,
+      "step": 626
+    },
+    {
+      "epoch": 1.0221715030974894,
+      "grad_norm": 0.12005385756492615,
+      "learning_rate": 0.0001,
+      "loss": 0.2077,
+      "step": 627
+    },
+    {
+      "epoch": 1.023801760678187,
+      "grad_norm": 0.13842667639255524,
+      "learning_rate": 0.0001,
+      "loss": 0.2263,
+      "step": 628
+    },
+    {
+      "epoch": 1.025432018258885,
+      "grad_norm": 0.14517144858837128,
+      "learning_rate": 0.0001,
+      "loss": 0.2117,
+      "step": 629
+    },
+    {
+      "epoch": 1.0270622758395827,
+      "grad_norm": 0.12258317321538925,
+      "learning_rate": 0.0001,
+      "loss": 0.1921,
+      "step": 630
+    },
+    {
+      "epoch": 1.0286925334202803,
+      "grad_norm": 0.10833742469549179,
+      "learning_rate": 0.0001,
+      "loss": 0.1868,
+      "step": 631
+    },
+    {
+      "epoch": 1.0303227910009782,
+      "grad_norm": 0.15066489577293396,
+      "learning_rate": 0.0001,
+      "loss": 0.206,
+      "step": 632
+    },
+    {
+      "epoch": 1.031953048581676,
+      "grad_norm": 0.11644082516431808,
+      "learning_rate": 0.0001,
+      "loss": 0.2128,
+      "step": 633
+    },
+    {
+      "epoch": 1.0335833061623736,
+      "grad_norm": 0.1423221379518509,
+      "learning_rate": 0.0001,
+      "loss": 0.1998,
+      "step": 634
+    },
+    {
+      "epoch": 1.0352135637430715,
+      "grad_norm": 0.12416679412126541,
+      "learning_rate": 0.0001,
+      "loss": 0.1915,
+      "step": 635
+    },
+    {
+      "epoch": 1.0368438213237692,
+      "grad_norm": 0.11678506433963776,
+      "learning_rate": 0.0001,
+      "loss": 0.189,
+      "step": 636
+    },
+    {
+      "epoch": 1.0384740789044669,
+      "grad_norm": 0.12458233535289764,
+      "learning_rate": 0.0001,
+      "loss": 0.1911,
+      "step": 637
+    },
+    {
+      "epoch": 1.0401043364851648,
+      "grad_norm": 0.12417783588171005,
+      "learning_rate": 0.0001,
+      "loss": 0.1773,
+      "step": 638
+    },
+    {
+      "epoch": 1.0417345940658624,
+      "grad_norm": 0.11989506334066391,
+      "learning_rate": 0.0001,
+      "loss": 0.1932,
+      "step": 639
+    },
+    {
+      "epoch": 1.0433648516465601,
+      "grad_norm": 0.1083277091383934,
+      "learning_rate": 0.0001,
+      "loss": 0.1952,
+      "step": 640
+    },
+    {
+      "epoch": 1.0449951092272578,
+      "grad_norm": 0.17112882435321808,
+      "learning_rate": 0.0001,
+      "loss": 0.2021,
+      "step": 641
+    },
+    {
+      "epoch": 1.0466253668079557,
+      "grad_norm": 0.14799998700618744,
+      "learning_rate": 0.0001,
+      "loss": 0.2038,
+      "step": 642
+    },
+    {
+      "epoch": 1.0482556243886534,
+      "grad_norm": 0.14156191051006317,
+      "learning_rate": 0.0001,
+      "loss": 0.1908,
+      "step": 643
+    },
+    {
+      "epoch": 1.049885881969351,
+      "grad_norm": 0.1254645138978958,
+      "learning_rate": 0.0001,
+      "loss": 0.2045,
+      "step": 644
+    },
+    {
+      "epoch": 1.051516139550049,
+      "grad_norm": 0.11081767827272415,
+      "learning_rate": 0.0001,
+      "loss": 0.195,
+      "step": 645
+    },
+    {
+      "epoch": 1.0531463971307466,
+      "grad_norm": 0.11810845881700516,
+      "learning_rate": 0.0001,
+      "loss": 0.1939,
+      "step": 646
+    },
+    {
+      "epoch": 1.0547766547114443,
+      "grad_norm": 0.12629976868629456,
+      "learning_rate": 0.0001,
+      "loss": 0.2117,
+      "step": 647
+    },
+    {
+      "epoch": 1.0564069122921422,
+      "grad_norm": 0.11882951855659485,
+      "learning_rate": 0.0001,
+      "loss": 0.192,
+      "step": 648
+    },
+    {
+      "epoch": 1.05803716987284,
+      "grad_norm": 0.11729996651411057,
+      "learning_rate": 0.0001,
+      "loss": 0.1916,
+      "step": 649
+    },
+    {
+      "epoch": 1.0596674274535376,
+      "grad_norm": 0.11437559127807617,
+      "learning_rate": 0.0001,
+      "loss": 0.1885,
+      "step": 650
+    },
+    {
+      "epoch": 1.0612976850342355,
+      "grad_norm": 0.13408319652080536,
+      "learning_rate": 0.0001,
+      "loss": 0.2138,
+      "step": 651
+    },
+    {
+      "epoch": 1.0629279426149332,
+      "grad_norm": 0.14032766222953796,
+      "learning_rate": 0.0001,
+      "loss": 0.205,
+      "step": 652
+    },
+    {
+      "epoch": 1.0645582001956309,
+      "grad_norm": 0.1438027173280716,
+      "learning_rate": 0.0001,
+      "loss": 0.1934,
+      "step": 653
+    },
+    {
+      "epoch": 1.0661884577763288,
+      "grad_norm": 0.11743790656328201,
+      "learning_rate": 0.0001,
+      "loss": 0.1944,
+      "step": 654
+    },
+    {
+      "epoch": 1.0678187153570264,
+      "grad_norm": 0.1314900517463684,
+      "learning_rate": 0.0001,
+      "loss": 0.2108,
+      "step": 655
+    },
+    {
+      "epoch": 1.0694489729377241,
+      "grad_norm": 0.11884745210409164,
+      "learning_rate": 0.0001,
+      "loss": 0.1858,
+      "step": 656
+    },
+    {
+      "epoch": 1.0710792305184218,
+      "grad_norm": 0.10163510590791702,
+      "learning_rate": 0.0001,
+      "loss": 0.1984,
+      "step": 657
+    },
+    {
+      "epoch": 1.0727094880991197,
+      "grad_norm": 0.11893923580646515,
+      "learning_rate": 0.0001,
+      "loss": 0.1999,
+      "step": 658
+    },
+    {
+      "epoch": 1.0743397456798174,
+      "grad_norm": 0.12390507012605667,
+      "learning_rate": 0.0001,
+      "loss": 0.2142,
+      "step": 659
+    },
+    {
+      "epoch": 1.075970003260515,
+      "grad_norm": 0.11425681412220001,
+      "learning_rate": 0.0001,
+      "loss": 0.1883,
+      "step": 660
+    },
+    {
+      "epoch": 1.077600260841213,
+      "grad_norm": 0.11586383730173111,
+      "learning_rate": 0.0001,
+      "loss": 0.1894,
+      "step": 661
+    },
+    {
+      "epoch": 1.0792305184219106,
+      "grad_norm": 0.1234535500407219,
+      "learning_rate": 0.0001,
+      "loss": 0.2,
+      "step": 662
+    },
+    {
+      "epoch": 1.0808607760026083,
+      "grad_norm": 0.107200987637043,
+      "learning_rate": 0.0001,
+      "loss": 0.1875,
+      "step": 663
+    },
+    {
+      "epoch": 1.0824910335833062,
+      "grad_norm": 0.12569493055343628,
+      "learning_rate": 0.0001,
+      "loss": 0.1936,
+      "step": 664
+    },
+    {
+      "epoch": 1.084121291164004,
+      "grad_norm": 0.12856988608837128,
+      "learning_rate": 0.0001,
+      "loss": 0.1905,
+      "step": 665
+    },
+    {
+      "epoch": 1.0857515487447016,
+      "grad_norm": 0.16755834221839905,
+      "learning_rate": 0.0001,
+      "loss": 0.2023,
+      "step": 666
+    },
+    {
+      "epoch": 1.0873818063253995,
+      "grad_norm": 0.13660454750061035,
+      "learning_rate": 0.0001,
+      "loss": 0.2032,
+      "step": 667
+    },
+    {
+      "epoch": 1.0890120639060972,
+      "grad_norm": 0.10664553940296173,
+      "learning_rate": 0.0001,
+      "loss": 0.1725,
+      "step": 668
+    },
+    {
+      "epoch": 1.0906423214867949,
+      "grad_norm": 0.1499480903148651,
+      "learning_rate": 0.0001,
+      "loss": 0.2057,
+      "step": 669
+    },
+    {
+      "epoch": 1.0922725790674928,
+      "grad_norm": 0.13117225468158722,
+      "learning_rate": 0.0001,
+      "loss": 0.2064,
+      "step": 670
+    },
+    {
+      "epoch": 1.0939028366481904,
+      "grad_norm": 0.12112154811620712,
+      "learning_rate": 0.0001,
+      "loss": 0.1952,
+      "step": 671
+    },
+    {
+      "epoch": 1.0955330942288881,
+      "grad_norm": 0.12471124529838562,
+      "learning_rate": 0.0001,
+      "loss": 0.2005,
+      "step": 672
+    },
+    {
+      "epoch": 1.097163351809586,
+      "grad_norm": 0.14131243526935577,
+      "learning_rate": 0.0001,
+      "loss": 0.2099,
+      "step": 673
+    },
+    {
+      "epoch": 1.0987936093902837,
+      "grad_norm": 0.1315171718597412,
+      "learning_rate": 0.0001,
+      "loss": 0.1955,
+      "step": 674
+    },
+    {
+      "epoch": 1.1004238669709814,
+      "grad_norm": 0.12525959312915802,
+      "learning_rate": 0.0001,
+      "loss": 0.1937,
+      "step": 675
+    },
+    {
+      "epoch": 1.102054124551679,
+      "grad_norm": 0.12638898193836212,
+      "learning_rate": 0.0001,
+      "loss": 0.2045,
+      "step": 676
+    },
+    {
+      "epoch": 1.103684382132377,
+      "grad_norm": 0.12964749336242676,
+      "learning_rate": 0.0001,
+      "loss": 0.1799,
+      "step": 677
+    },
+    {
+      "epoch": 1.1053146397130746,
+      "grad_norm": 0.14780597388744354,
+      "learning_rate": 0.0001,
+      "loss": 0.1987,
+      "step": 678
+    },
+    {
+      "epoch": 1.1069448972937723,
+      "grad_norm": 0.14882171154022217,
+      "learning_rate": 0.0001,
+      "loss": 0.2038,
+      "step": 679
+    },
+    {
+      "epoch": 1.1085751548744702,
+      "grad_norm": 0.131666362285614,
+      "learning_rate": 0.0001,
+      "loss": 0.1915,
+      "step": 680
+    },
+    {
+      "epoch": 1.110205412455168,
+      "grad_norm": 0.12266353517770767,
+      "learning_rate": 0.0001,
+      "loss": 0.19,
+      "step": 681
+    },
+    {
+      "epoch": 1.1118356700358656,
+      "grad_norm": 0.13120290637016296,
+      "learning_rate": 0.0001,
+      "loss": 0.2028,
+      "step": 682
+    },
+    {
+      "epoch": 1.1134659276165635,
+      "grad_norm": 0.12845005095005035,
+      "learning_rate": 0.0001,
+      "loss": 0.202,
+      "step": 683
+    },
+    {
+      "epoch": 1.1150961851972612,
+      "grad_norm": 0.12827853858470917,
+      "learning_rate": 0.0001,
+      "loss": 0.2101,
+      "step": 684
+    },
+    {
+      "epoch": 1.1167264427779588,
+      "grad_norm": 0.10449423640966415,
+      "learning_rate": 0.0001,
+      "loss": 0.1814,
+      "step": 685
+    },
+    {
+      "epoch": 1.1183567003586568,
+      "grad_norm": 0.12488257884979248,
+      "learning_rate": 0.0001,
+      "loss": 0.2008,
+      "step": 686
+    },
+    {
+      "epoch": 1.1199869579393544,
+      "grad_norm": 0.1401628851890564,
+      "learning_rate": 0.0001,
+      "loss": 0.2046,
+      "step": 687
+    },
+    {
+      "epoch": 1.1216172155200521,
+      "grad_norm": 0.11075824499130249,
+      "learning_rate": 0.0001,
+      "loss": 0.1908,
+      "step": 688
+    },
+    {
+      "epoch": 1.12324747310075,
+      "grad_norm": 0.14995628595352173,
+      "learning_rate": 0.0001,
+      "loss": 0.1993,
+      "step": 689
+    },
+    {
+      "epoch": 1.1248777306814477,
+      "grad_norm": 0.13028737902641296,
+      "learning_rate": 0.0001,
+      "loss": 0.2035,
+      "step": 690
+    },
+    {
+      "epoch": 1.1265079882621454,
+      "grad_norm": 0.12029106914997101,
+      "learning_rate": 0.0001,
+      "loss": 0.166,
+      "step": 691
+    },
+    {
+      "epoch": 1.1281382458428433,
+      "grad_norm": 0.11109007149934769,
+      "learning_rate": 0.0001,
+      "loss": 0.1921,
+      "step": 692
+    },
+    {
+      "epoch": 1.129768503423541,
+      "grad_norm": 0.112598717212677,
+      "learning_rate": 0.0001,
+      "loss": 0.1849,
+      "step": 693
+    },
+    {
+      "epoch": 1.1313987610042386,
+      "grad_norm": 0.11764469742774963,
+      "learning_rate": 0.0001,
+      "loss": 0.1979,
+      "step": 694
+    },
+    {
+      "epoch": 1.1330290185849363,
+      "grad_norm": 0.13900335133075714,
+      "learning_rate": 0.0001,
+      "loss": 0.1949,
+      "step": 695
+    },
+    {
+      "epoch": 1.1346592761656342,
+      "grad_norm": 0.11250978708267212,
+      "learning_rate": 0.0001,
+      "loss": 0.193,
+      "step": 696
+    },
+    {
+      "epoch": 1.136289533746332,
+      "grad_norm": 0.13568443059921265,
+      "learning_rate": 0.0001,
+      "loss": 0.1858,
+      "step": 697
+    },
+    {
+      "epoch": 1.1379197913270296,
+      "grad_norm": 0.15527978539466858,
+      "learning_rate": 0.0001,
+      "loss": 0.1957,
+      "step": 698
+    },
+    {
+      "epoch": 1.1395500489077275,
+      "grad_norm": 0.12831154465675354,
+      "learning_rate": 0.0001,
+      "loss": 0.19,
+      "step": 699
+    },
+    {
+      "epoch": 1.1411803064884252,
+      "grad_norm": 0.1345253884792328,
+      "learning_rate": 0.0001,
+      "loss": 0.1916,
+      "step": 700
+    },
+    {
+      "epoch": 1.1428105640691228,
+      "grad_norm": 0.13202865421772003,
+      "learning_rate": 0.0001,
+      "loss": 0.187,
+      "step": 701
+    },
+    {
+      "epoch": 1.1444408216498207,
+      "grad_norm": 0.12138167768716812,
+      "learning_rate": 0.0001,
+      "loss": 0.1923,
+      "step": 702
+    },
+    {
+      "epoch": 1.1460710792305184,
+      "grad_norm": 0.13008633255958557,
+      "learning_rate": 0.0001,
+      "loss": 0.2001,
+      "step": 703
+    },
+    {
+      "epoch": 1.147701336811216,
+      "grad_norm": 0.13472393155097961,
+      "learning_rate": 0.0001,
+      "loss": 0.1965,
+      "step": 704
+    },
+    {
+      "epoch": 1.149331594391914,
+      "grad_norm": 0.13745389878749847,
+      "learning_rate": 0.0001,
+      "loss": 0.1929,
+      "step": 705
+    },
+    {
+      "epoch": 1.1509618519726117,
+      "grad_norm": 0.13292807340621948,
+      "learning_rate": 0.0001,
+      "loss": 0.2109,
+      "step": 706
+    },
+    {
+      "epoch": 1.1525921095533094,
+      "grad_norm": 0.1446523666381836,
+      "learning_rate": 0.0001,
+      "loss": 0.1948,
+      "step": 707
+    },
+    {
+      "epoch": 1.154222367134007,
+      "grad_norm": 0.1305355578660965,
+      "learning_rate": 0.0001,
+      "loss": 0.187,
+      "step": 708
+    },
+    {
+      "epoch": 1.155852624714705,
+      "grad_norm": 0.11248882114887238,
+      "learning_rate": 0.0001,
+      "loss": 0.1878,
+      "step": 709
+    },
+    {
+      "epoch": 1.1574828822954026,
+      "grad_norm": 0.11229637265205383,
+      "learning_rate": 0.0001,
+      "loss": 0.1864,
+      "step": 710
+    },
+    {
+      "epoch": 1.1591131398761005,
+      "grad_norm": 0.13594146072864532,
+      "learning_rate": 0.0001,
+      "loss": 0.2021,
+      "step": 711
+    },
+    {
+      "epoch": 1.1607433974567982,
+      "grad_norm": 0.10707145929336548,
+      "learning_rate": 0.0001,
+      "loss": 0.1841,
+      "step": 712
+    },
+    {
+      "epoch": 1.162373655037496,
+      "grad_norm": 0.13550283014774323,
+      "learning_rate": 0.0001,
+      "loss": 0.1909,
+      "step": 713
+    },
+    {
+      "epoch": 1.1640039126181936,
+      "grad_norm": 0.16137608885765076,
+      "learning_rate": 0.0001,
+      "loss": 0.1904,
+      "step": 714
+    },
+    {
+      "epoch": 1.1656341701988915,
+      "grad_norm": 0.1217959076166153,
+      "learning_rate": 0.0001,
+      "loss": 0.1652,
+      "step": 715
+    },
+    {
+      "epoch": 1.1672644277795892,
+      "grad_norm": 0.12333795428276062,
+      "learning_rate": 0.0001,
+      "loss": 0.1774,
+      "step": 716
+    },
+    {
+      "epoch": 1.1688946853602868,
+      "grad_norm": 0.13329114019870758,
+      "learning_rate": 0.0001,
+      "loss": 0.1945,
+      "step": 717
+    },
+    {
+      "epoch": 1.1705249429409847,
+      "grad_norm": 0.13846194744110107,
+      "learning_rate": 0.0001,
+      "loss": 0.2064,
+      "step": 718
+    },
+    {
+      "epoch": 1.1721552005216824,
+      "grad_norm": 0.12300273030996323,
+      "learning_rate": 0.0001,
+      "loss": 0.1854,
+      "step": 719
+    },
+    {
+      "epoch": 1.17378545810238,
+      "grad_norm": 0.13478830456733704,
+      "learning_rate": 0.0001,
+      "loss": 0.192,
+      "step": 720
+    },
+    {
+      "epoch": 1.175415715683078,
+      "grad_norm": 0.12126126885414124,
+      "learning_rate": 0.0001,
+      "loss": 0.1976,
+      "step": 721
+    },
+    {
+      "epoch": 1.1770459732637757,
+      "grad_norm": 0.12510600686073303,
+      "learning_rate": 0.0001,
+      "loss": 0.2058,
+      "step": 722
+    },
+    {
+      "epoch": 1.1786762308444734,
+      "grad_norm": 0.1227242723107338,
+      "learning_rate": 0.0001,
+      "loss": 0.167,
+      "step": 723
+    },
+    {
+      "epoch": 1.1803064884251713,
+      "grad_norm": 0.15479400753974915,
+      "learning_rate": 0.0001,
+      "loss": 0.2009,
+      "step": 724
+    },
+    {
+      "epoch": 1.181936746005869,
+      "grad_norm": 0.1210135892033577,
+      "learning_rate": 0.0001,
+      "loss": 0.2009,
+      "step": 725
+    },
+    {
+      "epoch": 1.1835670035865666,
+      "grad_norm": 0.15612360835075378,
+      "learning_rate": 0.0001,
+      "loss": 0.191,
+      "step": 726
+    },
+    {
+      "epoch": 1.1851972611672643,
+      "grad_norm": 0.20443665981292725,
+      "learning_rate": 0.0001,
+      "loss": 0.1861,
+      "step": 727
+    },
+    {
+      "epoch": 1.1868275187479622,
+      "grad_norm": 0.13557809591293335,
+      "learning_rate": 0.0001,
+      "loss": 0.2098,
+      "step": 728
+    },
+    {
+      "epoch": 1.18845777632866,
+      "grad_norm": 0.14073556661605835,
+      "learning_rate": 0.0001,
+      "loss": 0.2046,
+      "step": 729
+    },
+    {
+      "epoch": 1.1900880339093578,
+      "grad_norm": 0.11344654113054276,
+      "learning_rate": 0.0001,
+      "loss": 0.1878,
+      "step": 730
+    },
+    {
+      "epoch": 1.1917182914900555,
+      "grad_norm": 0.12960317730903625,
+      "learning_rate": 0.0001,
+      "loss": 0.2068,
+      "step": 731
+    },
+    {
+      "epoch": 1.1933485490707532,
+      "grad_norm": 0.11689286679029465,
+      "learning_rate": 0.0001,
+      "loss": 0.1806,
+      "step": 732
+    },
+    {
+      "epoch": 1.1949788066514508,
+      "grad_norm": 0.12490563094615936,
+      "learning_rate": 0.0001,
+      "loss": 0.2013,
+      "step": 733
+    },
+    {
+      "epoch": 1.1966090642321487,
+      "grad_norm": 0.13222745060920715,
+      "learning_rate": 0.0001,
+      "loss": 0.1923,
+      "step": 734
+    },
+    {
+      "epoch": 1.1982393218128464,
+      "grad_norm": 0.14103910326957703,
+      "learning_rate": 0.0001,
+      "loss": 0.1973,
+      "step": 735
+    },
+    {
+      "epoch": 1.199869579393544,
+      "grad_norm": 0.1409367471933365,
+      "learning_rate": 0.0001,
+      "loss": 0.1971,
+      "step": 736
+    },
+    {
+      "epoch": 1.201499836974242,
+      "grad_norm": 0.13670028746128082,
+      "learning_rate": 0.0001,
+      "loss": 0.19,
+      "step": 737
+    },
+    {
+      "epoch": 1.2031300945549397,
+      "grad_norm": 0.11378346383571625,
+      "learning_rate": 0.0001,
+      "loss": 0.1846,
+      "step": 738
+    },
+    {
+      "epoch": 1.2047603521356374,
+      "grad_norm": 0.12771202623844147,
+      "learning_rate": 0.0001,
+      "loss": 0.1952,
+      "step": 739
+    },
+    {
+      "epoch": 1.2063906097163353,
+      "grad_norm": 0.1480347216129303,
+      "learning_rate": 0.0001,
+      "loss": 0.2065,
+      "step": 740
+    },
+    {
+      "epoch": 1.208020867297033,
+      "grad_norm": 0.14294308423995972,
+      "learning_rate": 0.0001,
+      "loss": 0.199,
+      "step": 741
+    },
+    {
+      "epoch": 1.2096511248777306,
+      "grad_norm": 0.11501043289899826,
+      "learning_rate": 0.0001,
+      "loss": 0.1888,
+      "step": 742
+    },
+    {
+      "epoch": 1.2112813824584285,
+      "grad_norm": 0.12315449863672256,
+      "learning_rate": 0.0001,
+      "loss": 0.1999,
+      "step": 743
+    },
+    {
+      "epoch": 1.2129116400391262,
+      "grad_norm": 0.11507046967744827,
+      "learning_rate": 0.0001,
+      "loss": 0.2075,
+      "step": 744
+    },
+    {
+      "epoch": 1.214541897619824,
+      "grad_norm": 0.11354836076498032,
+      "learning_rate": 0.0001,
+      "loss": 0.1861,
+      "step": 745
+    },
+    {
+      "epoch": 1.2161721552005216,
+      "grad_norm": 0.13437440991401672,
+      "learning_rate": 0.0001,
+      "loss": 0.1985,
+      "step": 746
+    },
+    {
+      "epoch": 1.2178024127812195,
+      "grad_norm": 0.11698954552412033,
+      "learning_rate": 0.0001,
+      "loss": 0.1959,
+      "step": 747
+    },
+    {
+      "epoch": 1.2194326703619172,
+      "grad_norm": 0.12441843748092651,
+      "learning_rate": 0.0001,
+      "loss": 0.1955,
+      "step": 748
+    },
+    {
+      "epoch": 1.2210629279426148,
+      "grad_norm": 1.1128816604614258,
+      "learning_rate": 0.0001,
+      "loss": 0.2114,
+      "step": 749
+    },
+    {
+      "epoch": 1.2226931855233127,
+      "grad_norm": 0.1364130824804306,
+      "learning_rate": 0.0001,
+      "loss": 0.1847,
+      "step": 750
+    },
+    {
+      "epoch": 1.2243234431040104,
+      "grad_norm": 0.14699682593345642,
+      "learning_rate": 0.0001,
+      "loss": 0.1955,
+      "step": 751
+    },
+    {
+      "epoch": 1.225953700684708,
+      "grad_norm": 0.16500285267829895,
+      "learning_rate": 0.0001,
+      "loss": 0.2026,
+      "step": 752
+    },
+    {
+      "epoch": 1.227583958265406,
+      "grad_norm": 0.12731170654296875,
+      "learning_rate": 0.0001,
+      "loss": 0.1935,
+      "step": 753
+    },
+    {
+      "epoch": 1.2292142158461037,
+      "grad_norm": 0.12876448035240173,
+      "learning_rate": 0.0001,
+      "loss": 0.2066,
+      "step": 754
+    },
+    {
+      "epoch": 1.2308444734268014,
+      "grad_norm": 0.14302317798137665,
+      "learning_rate": 0.0001,
+      "loss": 0.1946,
+      "step": 755
+    },
+    {
+      "epoch": 1.2324747310074993,
+      "grad_norm": 0.14912651479244232,
+      "learning_rate": 0.0001,
+      "loss": 0.1836,
+      "step": 756
+    },
+    {
+      "epoch": 1.234104988588197,
+      "grad_norm": 0.1160784587264061,
+      "learning_rate": 0.0001,
+      "loss": 0.1853,
+      "step": 757
+    },
+    {
+      "epoch": 1.2357352461688946,
+      "grad_norm": 0.11871147900819778,
+      "learning_rate": 0.0001,
+      "loss": 0.1845,
+      "step": 758
+    },
+    {
+      "epoch": 1.2373655037495925,
+      "grad_norm": 0.14770323038101196,
+      "learning_rate": 0.0001,
+      "loss": 0.2033,
+      "step": 759
+    },
+    {
+      "epoch": 1.2389957613302902,
+      "grad_norm": 0.2331717312335968,
+      "learning_rate": 0.0001,
+      "loss": 0.188,
+      "step": 760
+    },
+    {
+      "epoch": 1.240626018910988,
+      "grad_norm": 0.12382549792528152,
+      "learning_rate": 0.0001,
+      "loss": 0.1991,
+      "step": 761
+    },
+    {
+      "epoch": 1.2422562764916858,
+      "grad_norm": 0.16363592445850372,
+      "learning_rate": 0.0001,
+      "loss": 0.1922,
+      "step": 762
+    },
+    {
+      "epoch": 1.2438865340723835,
+      "grad_norm": 0.10923486948013306,
+      "learning_rate": 0.0001,
+      "loss": 0.1951,
+      "step": 763
+    },
+    {
+      "epoch": 1.2455167916530812,
+      "grad_norm": 0.15710005164146423,
+      "learning_rate": 0.0001,
+      "loss": 0.2105,
+      "step": 764
+    },
+    {
+      "epoch": 1.2471470492337788,
+      "grad_norm": 0.1298326998949051,
+      "learning_rate": 0.0001,
+      "loss": 0.1957,
+      "step": 765
+    },
+    {
+      "epoch": 1.2487773068144767,
+      "grad_norm": 0.12656745314598083,
+      "learning_rate": 0.0001,
+      "loss": 0.1908,
+      "step": 766
+    },
+    {
+      "epoch": 1.2504075643951744,
+      "grad_norm": 0.23991864919662476,
+      "learning_rate": 0.0001,
+      "loss": 0.2039,
+      "step": 767
+    },
+    {
+      "epoch": 1.2520378219758723,
+      "grad_norm": 0.12446881830692291,
+      "learning_rate": 0.0001,
+      "loss": 0.2051,
+      "step": 768
+    },
+    {
+      "epoch": 1.25366807955657,
+      "grad_norm": 0.13822287321090698,
+      "learning_rate": 0.0001,
+      "loss": 0.2118,
+      "step": 769
+    },
+    {
+      "epoch": 1.2552983371372677,
+      "grad_norm": 0.1322907656431198,
+      "learning_rate": 0.0001,
+      "loss": 0.2043,
+      "step": 770
+    },
+    {
+      "epoch": 1.2569285947179654,
+      "grad_norm": 0.13346537947654724,
+      "learning_rate": 0.0001,
+      "loss": 0.1907,
+      "step": 771
+    },
+    {
+      "epoch": 1.2585588522986633,
+      "grad_norm": 0.133008673787117,
+      "learning_rate": 0.0001,
+      "loss": 0.1957,
+      "step": 772
+    },
+    {
+      "epoch": 1.260189109879361,
+      "grad_norm": 0.13915961980819702,
+      "learning_rate": 0.0001,
+      "loss": 0.1996,
+      "step": 773
+    },
+    {
+      "epoch": 1.2618193674600586,
+      "grad_norm": 0.12022515386343002,
+      "learning_rate": 0.0001,
+      "loss": 0.2048,
+      "step": 774
+    },
+    {
+      "epoch": 1.2634496250407565,
+      "grad_norm": 0.12311067432165146,
+      "learning_rate": 0.0001,
+      "loss": 0.2076,
+      "step": 775
+    },
+    {
+      "epoch": 1.2650798826214542,
+      "grad_norm": 0.11143454164266586,
+      "learning_rate": 0.0001,
+      "loss": 0.192,
+      "step": 776
+    },
+    {
+      "epoch": 1.266710140202152,
+      "grad_norm": 0.11484785377979279,
+      "learning_rate": 0.0001,
+      "loss": 0.1797,
+      "step": 777
+    },
+    {
+      "epoch": 1.2683403977828496,
+      "grad_norm": 0.12231987714767456,
+      "learning_rate": 0.0001,
+      "loss": 0.2006,
+      "step": 778
+    },
+    {
+      "epoch": 1.2699706553635475,
+      "grad_norm": 0.138115793466568,
+      "learning_rate": 0.0001,
+      "loss": 0.1953,
+      "step": 779
+    },
+    {
+      "epoch": 1.2716009129442452,
+      "grad_norm": 0.1262485831975937,
+      "learning_rate": 0.0001,
+      "loss": 0.2104,
+      "step": 780
+    },
+    {
+      "epoch": 1.273231170524943,
+      "grad_norm": 0.10364367067813873,
+      "learning_rate": 0.0001,
+      "loss": 0.1913,
+      "step": 781
+    },
+    {
+      "epoch": 1.2748614281056407,
+      "grad_norm": 0.11666619032621384,
+      "learning_rate": 0.0001,
+      "loss": 0.2092,
+      "step": 782
+    },
+    {
+      "epoch": 1.2764916856863384,
+      "grad_norm": 0.10511959344148636,
+      "learning_rate": 0.0001,
+      "loss": 0.1821,
+      "step": 783
+    },
+    {
+      "epoch": 1.278121943267036,
+      "grad_norm": 0.12640027701854706,
+      "learning_rate": 0.0001,
+      "loss": 0.1969,
+      "step": 784
+    },
+    {
+      "epoch": 1.279752200847734,
+      "grad_norm": 0.13358190655708313,
+      "learning_rate": 0.0001,
+      "loss": 0.1874,
+      "step": 785
+    },
+    {
+      "epoch": 1.2813824584284317,
+      "grad_norm": 0.13797491788864136,
+      "learning_rate": 0.0001,
+      "loss": 0.2039,
+      "step": 786
+    },
+    {
+      "epoch": 1.2830127160091294,
+      "grad_norm": 0.16730040311813354,
+      "learning_rate": 0.0001,
+      "loss": 0.1812,
+      "step": 787
+    },
+    {
+      "epoch": 1.2846429735898273,
+      "grad_norm": 0.13052508234977722,
+      "learning_rate": 0.0001,
+      "loss": 0.2029,
+      "step": 788
+    },
+    {
+      "epoch": 1.286273231170525,
+      "grad_norm": 0.11100683361291885,
+      "learning_rate": 0.0001,
+      "loss": 0.1756,
+      "step": 789
+    },
+    {
+      "epoch": 1.2879034887512226,
+      "grad_norm": 0.12207093834877014,
+      "learning_rate": 0.0001,
+      "loss": 0.1951,
+      "step": 790
+    },
+    {
+      "epoch": 1.2895337463319203,
+      "grad_norm": 0.10500862449407578,
+      "learning_rate": 0.0001,
+      "loss": 0.187,
+      "step": 791
+    },
+    {
+      "epoch": 1.2911640039126182,
+      "grad_norm": 0.12395953387022018,
+      "learning_rate": 0.0001,
+      "loss": 0.1789,
+      "step": 792
+    },
+    {
+      "epoch": 1.2927942614933159,
+      "grad_norm": 0.12728352844715118,
+      "learning_rate": 0.0001,
+      "loss": 0.2055,
+      "step": 793
+    },
+    {
+      "epoch": 1.2944245190740138,
+      "grad_norm": 0.12210983037948608,
+      "learning_rate": 0.0001,
+      "loss": 0.1848,
+      "step": 794
+    },
+    {
+      "epoch": 1.2960547766547115,
+      "grad_norm": 0.1309320330619812,
+      "learning_rate": 0.0001,
+      "loss": 0.2019,
+      "step": 795
+    },
+    {
+      "epoch": 1.2976850342354092,
+      "grad_norm": 0.14810331165790558,
+      "learning_rate": 0.0001,
+      "loss": 0.1983,
+      "step": 796
+    },
+    {
+      "epoch": 1.2993152918161068,
+      "grad_norm": 0.11592376232147217,
+      "learning_rate": 0.0001,
+      "loss": 0.1839,
+      "step": 797
+    },
+    {
+      "epoch": 1.3009455493968047,
+      "grad_norm": 0.12396306544542313,
+      "learning_rate": 0.0001,
+      "loss": 0.1964,
+      "step": 798
+    },
+    {
+      "epoch": 1.3025758069775024,
+      "grad_norm": 0.13260740041732788,
+      "learning_rate": 0.0001,
+      "loss": 0.2012,
+      "step": 799
+    },
+    {
+      "epoch": 1.3042060645582003,
+      "grad_norm": 0.13183878362178802,
+      "learning_rate": 0.0001,
+      "loss": 0.2066,
+      "step": 800
+    },
+    {
+      "epoch": 1.305836322138898,
+      "grad_norm": 0.12593010067939758,
+      "learning_rate": 0.0001,
+      "loss": 0.1916,
+      "step": 801
+    },
+    {
+      "epoch": 1.3074665797195957,
+      "grad_norm": 0.11641526967287064,
+      "learning_rate": 0.0001,
+      "loss": 0.1847,
+      "step": 802
+    },
+    {
+      "epoch": 1.3090968373002934,
+      "grad_norm": 0.1082523912191391,
+      "learning_rate": 0.0001,
+      "loss": 0.1816,
+      "step": 803
+    },
+    {
+      "epoch": 1.3107270948809913,
+      "grad_norm": 0.11721451580524445,
+      "learning_rate": 0.0001,
+      "loss": 0.1913,
+      "step": 804
+    },
+    {
+      "epoch": 1.312357352461689,
+      "grad_norm": 0.14624540507793427,
+      "learning_rate": 0.0001,
+      "loss": 0.2057,
+      "step": 805
+    },
+    {
+      "epoch": 1.3139876100423866,
+      "grad_norm": 0.11817949265241623,
+      "learning_rate": 0.0001,
+      "loss": 0.1876,
+      "step": 806
+    },
+    {
+      "epoch": 1.3156178676230845,
+      "grad_norm": 0.12821592390537262,
+      "learning_rate": 0.0001,
+      "loss": 0.2099,
+      "step": 807
+    },
+    {
+      "epoch": 1.3172481252037822,
+      "grad_norm": 0.13968628644943237,
+      "learning_rate": 0.0001,
+      "loss": 0.2101,
+      "step": 808
+    },
+    {
+      "epoch": 1.3188783827844799,
+      "grad_norm": 0.1454535573720932,
+      "learning_rate": 0.0001,
+      "loss": 0.2062,
+      "step": 809
+    },
+    {
+      "epoch": 1.3205086403651776,
+      "grad_norm": 0.1232244223356247,
+      "learning_rate": 0.0001,
+      "loss": 0.1775,
+      "step": 810
+    },
+    {
+      "epoch": 1.3221388979458755,
+      "grad_norm": 0.12356823682785034,
+      "learning_rate": 0.0001,
+      "loss": 0.1945,
+      "step": 811
+    },
+    {
+      "epoch": 1.3237691555265731,
+      "grad_norm": 0.1321604698896408,
+      "learning_rate": 0.0001,
+      "loss": 0.1956,
+      "step": 812
+    },
+    {
+      "epoch": 1.325399413107271,
+      "grad_norm": 0.13686218857765198,
+      "learning_rate": 0.0001,
+      "loss": 0.1997,
+      "step": 813
+    },
+    {
+      "epoch": 1.3270296706879687,
+      "grad_norm": 0.1265021115541458,
+      "learning_rate": 0.0001,
+      "loss": 0.2003,
+      "step": 814
+    },
+    {
+      "epoch": 1.3286599282686664,
+      "grad_norm": 0.13822129368782043,
+      "learning_rate": 0.0001,
+      "loss": 0.2037,
+      "step": 815
+    },
+    {
+      "epoch": 1.330290185849364,
+      "grad_norm": 0.1474665403366089,
+      "learning_rate": 0.0001,
+      "loss": 0.2014,
+      "step": 816
+    },
+    {
+      "epoch": 1.331920443430062,
+      "grad_norm": 0.1312236338853836,
+      "learning_rate": 0.0001,
+      "loss": 0.2047,
+      "step": 817
+    },
+    {
+      "epoch": 1.3335507010107597,
+      "grad_norm": 0.10650108009576797,
+      "learning_rate": 0.0001,
+      "loss": 0.1774,
+      "step": 818
+    },
+    {
+      "epoch": 1.3351809585914576,
+      "grad_norm": 0.12446287274360657,
+      "learning_rate": 0.0001,
+      "loss": 0.1993,
+      "step": 819
+    },
+    {
+      "epoch": 1.3368112161721553,
+      "grad_norm": 0.1500684767961502,
+      "learning_rate": 0.0001,
+      "loss": 0.2065,
+      "step": 820
+    },
+    {
+      "epoch": 1.338441473752853,
+      "grad_norm": 0.11526080965995789,
+      "learning_rate": 0.0001,
+      "loss": 0.1929,
+      "step": 821
+    },
+    {
+      "epoch": 1.3400717313335506,
+      "grad_norm": 0.14538566768169403,
+      "learning_rate": 0.0001,
+      "loss": 0.1951,
+      "step": 822
+    },
+    {
+      "epoch": 1.3417019889142485,
+      "grad_norm": 0.11737110465765,
+      "learning_rate": 0.0001,
+      "loss": 0.2022,
+      "step": 823
+    },
+    {
+      "epoch": 1.3433322464949462,
+      "grad_norm": 0.1339058130979538,
+      "learning_rate": 0.0001,
+      "loss": 0.2056,
+      "step": 824
+    },
+    {
+      "epoch": 1.3449625040756439,
+      "grad_norm": 0.1357828974723816,
+      "learning_rate": 0.0001,
+      "loss": 0.1997,
+      "step": 825
+    },
+    {
+      "epoch": 1.3465927616563418,
+      "grad_norm": 0.13894008100032806,
+      "learning_rate": 0.0001,
+      "loss": 0.1976,
+      "step": 826
+    },
+    {
+      "epoch": 1.3482230192370395,
+      "grad_norm": 0.11640580743551254,
+      "learning_rate": 0.0001,
+      "loss": 0.1931,
+      "step": 827
+    },
+    {
+      "epoch": 1.3498532768177371,
+      "grad_norm": 0.11283834278583527,
+      "learning_rate": 0.0001,
+      "loss": 0.1968,
+      "step": 828
+    },
+    {
+      "epoch": 1.3514835343984348,
+      "grad_norm": 0.13692787289619446,
+      "learning_rate": 0.0001,
+      "loss": 0.2087,
+      "step": 829
+    },
+    {
+      "epoch": 1.3531137919791327,
+      "grad_norm": 0.0954209491610527,
+      "learning_rate": 0.0001,
+      "loss": 0.1828,
+      "step": 830
+    },
+    {
+      "epoch": 1.3547440495598304,
+      "grad_norm": 0.1314362734556198,
+      "learning_rate": 0.0001,
+      "loss": 0.2074,
+      "step": 831
+    },
+    {
+      "epoch": 1.3563743071405283,
+      "grad_norm": 0.12032657861709595,
+      "learning_rate": 0.0001,
+      "loss": 0.1797,
+      "step": 832
+    },
+    {
+      "epoch": 1.358004564721226,
+      "grad_norm": 0.12045875936746597,
+      "learning_rate": 0.0001,
+      "loss": 0.1948,
+      "step": 833
+    },
+    {
+      "epoch": 1.3596348223019237,
+      "grad_norm": 0.12139002233743668,
+      "learning_rate": 0.0001,
+      "loss": 0.2074,
+      "step": 834
+    },
+    {
+      "epoch": 1.3612650798826214,
+      "grad_norm": 0.125333771109581,
+      "learning_rate": 0.0001,
+      "loss": 0.189,
+      "step": 835
+    },
+    {
+      "epoch": 1.3628953374633193,
+      "grad_norm": 0.12336897104978561,
+      "learning_rate": 0.0001,
+      "loss": 0.2015,
+      "step": 836
+    },
+    {
+      "epoch": 1.364525595044017,
+      "grad_norm": 0.1447877436876297,
+      "learning_rate": 0.0001,
+      "loss": 0.1979,
+      "step": 837
+    },
+    {
+      "epoch": 1.3661558526247148,
+      "grad_norm": 0.12751318514347076,
+      "learning_rate": 0.0001,
+      "loss": 0.1924,
+      "step": 838
+    },
+    {
+      "epoch": 1.3677861102054125,
+      "grad_norm": 0.14476466178894043,
+      "learning_rate": 0.0001,
+      "loss": 0.2007,
+      "step": 839
+    },
+    {
+      "epoch": 1.3694163677861102,
+      "grad_norm": 0.12900257110595703,
+      "learning_rate": 0.0001,
+      "loss": 0.1685,
+      "step": 840
+    },
+    {
+      "epoch": 1.3710466253668079,
+      "grad_norm": 0.14788837730884552,
+      "learning_rate": 0.0001,
+      "loss": 0.188,
+      "step": 841
+    },
+    {
+      "epoch": 1.3726768829475058,
+      "grad_norm": 0.1143098995089531,
+      "learning_rate": 0.0001,
+      "loss": 0.1925,
+      "step": 842
+    },
+    {
+      "epoch": 1.3743071405282035,
+      "grad_norm": 0.16243141889572144,
+      "learning_rate": 0.0001,
+      "loss": 0.1904,
+      "step": 843
+    },
+    {
+      "epoch": 1.3759373981089011,
+      "grad_norm": 0.1285100132226944,
+      "learning_rate": 0.0001,
+      "loss": 0.2101,
+      "step": 844
+    },
+    {
+      "epoch": 1.377567655689599,
+      "grad_norm": 0.12278752028942108,
+      "learning_rate": 0.0001,
+      "loss": 0.1676,
+      "step": 845
+    },
+    {
+      "epoch": 1.3791979132702967,
+      "grad_norm": 0.11762263625860214,
+      "learning_rate": 0.0001,
+      "loss": 0.1857,
+      "step": 846
+    },
+    {
+      "epoch": 1.3808281708509944,
+      "grad_norm": 0.14315438270568848,
+      "learning_rate": 0.0001,
+      "loss": 0.1981,
+      "step": 847
+    },
+    {
+      "epoch": 1.382458428431692,
+      "grad_norm": 0.13514509797096252,
+      "learning_rate": 0.0001,
+      "loss": 0.1874,
+      "step": 848
+    },
+    {
+      "epoch": 1.38408868601239,
+      "grad_norm": 0.14663557708263397,
+      "learning_rate": 0.0001,
+      "loss": 0.1922,
+      "step": 849
+    },
+    {
+      "epoch": 1.3857189435930877,
+      "grad_norm": 0.11271989345550537,
+      "learning_rate": 0.0001,
+      "loss": 0.1741,
+      "step": 850
+    },
+    {
+      "epoch": 1.3873492011737856,
+      "grad_norm": 0.1267681121826172,
+      "learning_rate": 0.0001,
+      "loss": 0.2004,
+      "step": 851
+    },
+    {
+      "epoch": 1.3889794587544833,
+      "grad_norm": 0.13862234354019165,
+      "learning_rate": 0.0001,
+      "loss": 0.1833,
+      "step": 852
+    },
+    {
+      "epoch": 1.390609716335181,
+      "grad_norm": 0.14076252281665802,
+      "learning_rate": 0.0001,
+      "loss": 0.1946,
+      "step": 853
+    },
+    {
+      "epoch": 1.3922399739158786,
+      "grad_norm": 0.11336036026477814,
+      "learning_rate": 0.0001,
+      "loss": 0.1826,
+      "step": 854
+    },
+    {
+      "epoch": 1.3938702314965765,
+      "grad_norm": 0.12610171735286713,
+      "learning_rate": 0.0001,
+      "loss": 0.2061,
+      "step": 855
+    },
+    {
+      "epoch": 1.3955004890772742,
+      "grad_norm": 0.12771809101104736,
+      "learning_rate": 0.0001,
+      "loss": 0.1826,
+      "step": 856
+    },
+    {
+      "epoch": 1.3971307466579719,
+      "grad_norm": 0.13943758606910706,
+      "learning_rate": 0.0001,
+      "loss": 0.2154,
+      "step": 857
+    },
+    {
+      "epoch": 1.3987610042386698,
+      "grad_norm": 0.1181144267320633,
+      "learning_rate": 0.0001,
+      "loss": 0.2047,
+      "step": 858
+    },
+    {
+      "epoch": 1.4003912618193675,
+      "grad_norm": 0.12739057838916779,
+      "learning_rate": 0.0001,
+      "loss": 0.1924,
+      "step": 859
+    },
+    {
+      "epoch": 1.4020215194000651,
+      "grad_norm": 0.14955663681030273,
+      "learning_rate": 0.0001,
+      "loss": 0.204,
+      "step": 860
+    },
+    {
+      "epoch": 1.4036517769807628,
+      "grad_norm": 0.12984280288219452,
+      "learning_rate": 0.0001,
+      "loss": 0.1833,
+      "step": 861
+    },
+    {
+      "epoch": 1.4052820345614607,
+      "grad_norm": 0.1367812603712082,
+      "learning_rate": 0.0001,
+      "loss": 0.2041,
+      "step": 862
+    },
+    {
+      "epoch": 1.4069122921421584,
+      "grad_norm": 0.11790581792593002,
+      "learning_rate": 0.0001,
+      "loss": 0.1927,
+      "step": 863
+    },
+    {
+      "epoch": 1.4085425497228563,
+      "grad_norm": 0.13845130801200867,
+      "learning_rate": 0.0001,
+      "loss": 0.198,
+      "step": 864
+    },
+    {
+      "epoch": 1.410172807303554,
+      "grad_norm": 0.13803738355636597,
+      "learning_rate": 0.0001,
+      "loss": 0.1942,
+      "step": 865
+    },
+    {
+      "epoch": 1.4118030648842517,
+      "grad_norm": 0.11733663082122803,
+      "learning_rate": 0.0001,
+      "loss": 0.1811,
+      "step": 866
+    },
+    {
+      "epoch": 1.4134333224649493,
+      "grad_norm": 0.14048326015472412,
+      "learning_rate": 0.0001,
+      "loss": 0.2063,
+      "step": 867
+    },
+    {
+      "epoch": 1.4150635800456473,
+      "grad_norm": 0.11841481178998947,
+      "learning_rate": 0.0001,
+      "loss": 0.1783,
+      "step": 868
+    },
+    {
+      "epoch": 1.416693837626345,
+      "grad_norm": 0.13812898099422455,
+      "learning_rate": 0.0001,
+      "loss": 0.1945,
+      "step": 869
+    },
+    {
+      "epoch": 1.4183240952070428,
+      "grad_norm": 0.12648576498031616,
+      "learning_rate": 0.0001,
+      "loss": 0.2121,
+      "step": 870
+    },
+    {
+      "epoch": 1.4199543527877405,
+      "grad_norm": 0.12122874706983566,
+      "learning_rate": 0.0001,
+      "loss": 0.198,
+      "step": 871
+    },
+    {
+      "epoch": 1.4215846103684382,
+      "grad_norm": 0.12100232392549515,
+      "learning_rate": 0.0001,
+      "loss": 0.1896,
+      "step": 872
+    },
+    {
+      "epoch": 1.4232148679491359,
+      "grad_norm": 0.11330271512269974,
+      "learning_rate": 0.0001,
+      "loss": 0.1883,
+      "step": 873
+    },
+    {
+      "epoch": 1.4248451255298338,
+      "grad_norm": 0.12889273464679718,
+      "learning_rate": 0.0001,
+      "loss": 0.1913,
+      "step": 874
+    },
+    {
+      "epoch": 1.4264753831105315,
+      "grad_norm": 0.12529848515987396,
+      "learning_rate": 0.0001,
+      "loss": 0.1947,
+      "step": 875
+    },
+    {
+      "epoch": 1.4281056406912291,
+      "grad_norm": 0.11464565992355347,
+      "learning_rate": 0.0001,
+      "loss": 0.1928,
+      "step": 876
+    },
+    {
+      "epoch": 1.429735898271927,
+      "grad_norm": 0.09968235343694687,
+      "learning_rate": 0.0001,
+      "loss": 0.18,
+      "step": 877
+    },
+    {
+      "epoch": 1.4313661558526247,
+      "grad_norm": 0.12522117793560028,
+      "learning_rate": 0.0001,
+      "loss": 0.2025,
+      "step": 878
+    },
+    {
+      "epoch": 1.4329964134333224,
+      "grad_norm": 0.13101626932621002,
+      "learning_rate": 0.0001,
+      "loss": 0.1883,
+      "step": 879
+    },
+    {
+      "epoch": 1.43462667101402,
+      "grad_norm": 0.18192481994628906,
+      "learning_rate": 0.0001,
+      "loss": 0.1906,
+      "step": 880
+    },
+    {
+      "epoch": 1.436256928594718,
+      "grad_norm": 0.1182754710316658,
+      "learning_rate": 0.0001,
+      "loss": 0.1737,
+      "step": 881
+    },
+    {
+      "epoch": 1.4378871861754157,
+      "grad_norm": 0.1254081428050995,
+      "learning_rate": 0.0001,
+      "loss": 0.1919,
+      "step": 882
+    },
+    {
+      "epoch": 1.4395174437561136,
+      "grad_norm": 0.12255612015724182,
+      "learning_rate": 0.0001,
+      "loss": 0.1998,
+      "step": 883
+    },
+    {
+      "epoch": 1.4411477013368112,
+      "grad_norm": 0.12978112697601318,
+      "learning_rate": 0.0001,
+      "loss": 0.189,
+      "step": 884
+    },
+    {
+      "epoch": 1.442777958917509,
+      "grad_norm": 0.1313999891281128,
+      "learning_rate": 0.0001,
+      "loss": 0.1847,
+      "step": 885
+    },
+    {
+      "epoch": 1.4444082164982066,
+      "grad_norm": 0.1327790915966034,
+      "learning_rate": 0.0001,
+      "loss": 0.2005,
+      "step": 886
+    },
+    {
+      "epoch": 1.4460384740789045,
+      "grad_norm": 0.13885673880577087,
+      "learning_rate": 0.0001,
+      "loss": 0.1899,
+      "step": 887
+    },
+    {
+      "epoch": 1.4476687316596022,
+      "grad_norm": 0.13544884324073792,
+      "learning_rate": 0.0001,
+      "loss": 0.1944,
+      "step": 888
+    },
+    {
+      "epoch": 1.4492989892403,
+      "grad_norm": 0.12309068441390991,
+      "learning_rate": 0.0001,
+      "loss": 0.2011,
+      "step": 889
+    },
+    {
+      "epoch": 1.4509292468209978,
+      "grad_norm": 0.11212994158267975,
+      "learning_rate": 0.0001,
+      "loss": 0.1709,
+      "step": 890
+    },
+    {
+      "epoch": 1.4525595044016955,
+      "grad_norm": 0.1351044625043869,
+      "learning_rate": 0.0001,
+      "loss": 0.1765,
+      "step": 891
+    },
+    {
+      "epoch": 1.4541897619823931,
+      "grad_norm": 0.1273559033870697,
+      "learning_rate": 0.0001,
+      "loss": 0.1856,
+      "step": 892
+    },
+    {
+      "epoch": 1.455820019563091,
+      "grad_norm": 0.12091701477766037,
+      "learning_rate": 0.0001,
+      "loss": 0.2017,
+      "step": 893
+    },
+    {
+      "epoch": 1.4574502771437887,
+      "grad_norm": 0.12357049435377121,
+      "learning_rate": 0.0001,
+      "loss": 0.1997,
+      "step": 894
+    },
+    {
+      "epoch": 1.4590805347244864,
+      "grad_norm": 0.13915151357650757,
+      "learning_rate": 0.0001,
+      "loss": 0.2016,
+      "step": 895
+    },
+    {
+      "epoch": 1.4607107923051843,
+      "grad_norm": 0.15922588109970093,
+      "learning_rate": 0.0001,
+      "loss": 0.1946,
+      "step": 896
+    },
+    {
+      "epoch": 1.462341049885882,
+      "grad_norm": 0.1327250450849533,
+      "learning_rate": 0.0001,
+      "loss": 0.2009,
+      "step": 897
+    },
+    {
+      "epoch": 1.4639713074665797,
+      "grad_norm": 0.1281212717294693,
+      "learning_rate": 0.0001,
+      "loss": 0.1885,
+      "step": 898
+    },
+    {
+      "epoch": 1.4656015650472773,
+      "grad_norm": 0.12630882859230042,
+      "learning_rate": 0.0001,
+      "loss": 0.1758,
+      "step": 899
+    },
+    {
+      "epoch": 1.4672318226279752,
+      "grad_norm": 0.17888978123664856,
+      "learning_rate": 0.0001,
+      "loss": 0.1969,
+      "step": 900
+    },
+    {
+      "epoch": 1.468862080208673,
+      "grad_norm": 0.11581540107727051,
+      "learning_rate": 0.0001,
+      "loss": 0.1811,
+      "step": 901
+    },
+    {
+      "epoch": 1.4704923377893708,
+      "grad_norm": 0.13671360909938812,
+      "learning_rate": 0.0001,
+      "loss": 0.1921,
+      "step": 902
+    },
+    {
+      "epoch": 1.4721225953700685,
+      "grad_norm": 0.12927068769931793,
+      "learning_rate": 0.0001,
+      "loss": 0.1907,
+      "step": 903
+    },
+    {
+      "epoch": 1.4737528529507662,
+      "grad_norm": 0.1307104527950287,
+      "learning_rate": 0.0001,
+      "loss": 0.1774,
+      "step": 904
+    },
+    {
+      "epoch": 1.4753831105314639,
+      "grad_norm": 0.12712356448173523,
+      "learning_rate": 0.0001,
+      "loss": 0.1945,
+      "step": 905
+    },
+    {
+      "epoch": 1.4770133681121618,
+      "grad_norm": 0.11870065331459045,
+      "learning_rate": 0.0001,
+      "loss": 0.1764,
+      "step": 906
+    },
+    {
+      "epoch": 1.4786436256928595,
+      "grad_norm": 0.13073444366455078,
+      "learning_rate": 0.0001,
+      "loss": 0.1941,
+      "step": 907
+    },
+    {
+      "epoch": 1.4802738832735574,
+      "grad_norm": 0.12752722203731537,
+      "learning_rate": 0.0001,
+      "loss": 0.1781,
+      "step": 908
+    },
+    {
+      "epoch": 1.481904140854255,
+      "grad_norm": 0.18502075970172882,
+      "learning_rate": 0.0001,
+      "loss": 0.1963,
+      "step": 909
+    },
+    {
+      "epoch": 1.4835343984349527,
+      "grad_norm": 0.14644372463226318,
+      "learning_rate": 0.0001,
+      "loss": 0.2031,
+      "step": 910
+    },
+    {
+      "epoch": 1.4851646560156504,
+      "grad_norm": 0.1375977247953415,
+      "learning_rate": 0.0001,
+      "loss": 0.1872,
+      "step": 911
+    },
+    {
+      "epoch": 1.4867949135963483,
+      "grad_norm": 0.1250106245279312,
+      "learning_rate": 0.0001,
+      "loss": 0.1939,
+      "step": 912
+    },
+    {
+      "epoch": 1.488425171177046,
+      "grad_norm": 0.1255251169204712,
+      "learning_rate": 0.0001,
+      "loss": 0.1918,
+      "step": 913
+    },
+    {
+      "epoch": 1.4900554287577437,
+      "grad_norm": 0.11189655214548111,
+      "learning_rate": 0.0001,
+      "loss": 0.1897,
+      "step": 914
+    },
+    {
+      "epoch": 1.4916856863384416,
+      "grad_norm": 0.12730088829994202,
+      "learning_rate": 0.0001,
+      "loss": 0.1828,
+      "step": 915
+    },
+    {
+      "epoch": 1.4933159439191392,
+      "grad_norm": 0.10540137439966202,
+      "learning_rate": 0.0001,
+      "loss": 0.191,
+      "step": 916
+    },
+    {
+      "epoch": 1.494946201499837,
+      "grad_norm": 0.10388045758008957,
+      "learning_rate": 0.0001,
+      "loss": 0.1904,
+      "step": 917
+    },
+    {
+      "epoch": 1.4965764590805346,
+      "grad_norm": 0.1330886036157608,
+      "learning_rate": 0.0001,
+      "loss": 0.1946,
+      "step": 918
+    },
+    {
+      "epoch": 1.4982067166612325,
+      "grad_norm": 0.12678848206996918,
+      "learning_rate": 0.0001,
+      "loss": 0.2083,
+      "step": 919
+    },
+    {
+      "epoch": 1.4998369742419302,
+      "grad_norm": 0.12516173720359802,
+      "learning_rate": 0.0001,
+      "loss": 0.202,
+      "step": 920
+    },
+    {
+      "epoch": 1.501467231822628,
+      "grad_norm": 0.1177758052945137,
+      "learning_rate": 0.0001,
+      "loss": 0.1943,
+      "step": 921
+    },
+    {
+      "epoch": 1.5030974894033258,
+      "grad_norm": 0.10760951787233353,
+      "learning_rate": 0.0001,
+      "loss": 0.1738,
+      "step": 922
+    },
+    {
+      "epoch": 1.5047277469840235,
+      "grad_norm": 0.1439739167690277,
+      "learning_rate": 0.0001,
+      "loss": 0.2107,
+      "step": 923
+    },
+    {
+      "epoch": 1.5063580045647211,
+      "grad_norm": 0.127821683883667,
+      "learning_rate": 0.0001,
+      "loss": 0.1788,
+      "step": 924
+    },
+    {
+      "epoch": 1.5079882621454188,
+      "grad_norm": 0.11863593012094498,
+      "learning_rate": 0.0001,
+      "loss": 0.1928,
+      "step": 925
+    },
+    {
+      "epoch": 1.5096185197261167,
+      "grad_norm": 0.1257062703371048,
+      "learning_rate": 0.0001,
+      "loss": 0.1814,
+      "step": 926
+    },
+    {
+      "epoch": 1.5112487773068146,
+      "grad_norm": 0.14376260340213776,
+      "learning_rate": 0.0001,
+      "loss": 0.1876,
+      "step": 927
+    },
+    {
+      "epoch": 1.5128790348875123,
+      "grad_norm": 0.12747377157211304,
+      "learning_rate": 0.0001,
+      "loss": 0.2062,
+      "step": 928
+    },
+    {
+      "epoch": 1.51450929246821,
+      "grad_norm": 0.10602065175771713,
+      "learning_rate": 0.0001,
+      "loss": 0.1834,
+      "step": 929
+    },
+    {
+      "epoch": 1.5161395500489077,
+      "grad_norm": 0.11980943381786346,
+      "learning_rate": 0.0001,
+      "loss": 0.1932,
+      "step": 930
+    },
+    {
+      "epoch": 1.5177698076296053,
+      "grad_norm": 0.12690620124340057,
+      "learning_rate": 0.0001,
+      "loss": 0.2067,
+      "step": 931
+    },
+    {
+      "epoch": 1.5194000652103032,
+      "grad_norm": 0.10429069399833679,
+      "learning_rate": 0.0001,
+      "loss": 0.1845,
+      "step": 932
+    },
+    {
+      "epoch": 1.5210303227910011,
+      "grad_norm": 0.11686432361602783,
+      "learning_rate": 0.0001,
+      "loss": 0.2038,
+      "step": 933
+    },
+    {
+      "epoch": 1.5226605803716988,
+      "grad_norm": 0.13629446923732758,
+      "learning_rate": 0.0001,
+      "loss": 0.2071,
+      "step": 934
+    },
+    {
+      "epoch": 1.5242908379523965,
+      "grad_norm": 0.11884860694408417,
+      "learning_rate": 0.0001,
+      "loss": 0.1866,
+      "step": 935
+    },
+    {
+      "epoch": 1.5259210955330942,
+      "grad_norm": 0.11080071330070496,
+      "learning_rate": 0.0001,
+      "loss": 0.1925,
+      "step": 936
+    },
+    {
+      "epoch": 1.5275513531137919,
+      "grad_norm": 0.13977265357971191,
+      "learning_rate": 0.0001,
+      "loss": 0.2035,
+      "step": 937
+    },
+    {
+      "epoch": 1.5291816106944898,
+      "grad_norm": 0.1306087076663971,
+      "learning_rate": 0.0001,
+      "loss": 0.1906,
+      "step": 938
+    },
+    {
+      "epoch": 1.5308118682751874,
+      "grad_norm": 0.1292889267206192,
+      "learning_rate": 0.0001,
+      "loss": 0.2064,
+      "step": 939
+    },
+    {
+      "epoch": 1.5324421258558854,
+      "grad_norm": 0.1151227205991745,
+      "learning_rate": 0.0001,
+      "loss": 0.1889,
+      "step": 940
+    },
+    {
+      "epoch": 1.534072383436583,
+      "grad_norm": 0.23857301473617554,
+      "learning_rate": 0.0001,
+      "loss": 0.1936,
+      "step": 941
+    },
+    {
+      "epoch": 1.5357026410172807,
+      "grad_norm": 0.12873469293117523,
+      "learning_rate": 0.0001,
+      "loss": 0.1864,
+      "step": 942
+    },
+    {
+      "epoch": 1.5373328985979784,
+      "grad_norm": 0.1140642762184143,
+      "learning_rate": 0.0001,
+      "loss": 0.1844,
+      "step": 943
+    },
+    {
+      "epoch": 1.538963156178676,
+      "grad_norm": 0.12932021915912628,
+      "learning_rate": 0.0001,
+      "loss": 0.1717,
+      "step": 944
+    },
+    {
+      "epoch": 1.540593413759374,
+      "grad_norm": 0.11678753048181534,
+      "learning_rate": 0.0001,
+      "loss": 0.1814,
+      "step": 945
+    },
+    {
+      "epoch": 1.5422236713400719,
+      "grad_norm": 0.11084496229887009,
+      "learning_rate": 0.0001,
+      "loss": 0.1635,
+      "step": 946
+    },
+    {
+      "epoch": 1.5438539289207696,
+      "grad_norm": 0.11934110522270203,
+      "learning_rate": 0.0001,
+      "loss": 0.1897,
+      "step": 947
+    },
+    {
+      "epoch": 1.5454841865014672,
+      "grad_norm": 0.13100844621658325,
+      "learning_rate": 0.0001,
+      "loss": 0.1852,
+      "step": 948
+    },
+    {
+      "epoch": 1.547114444082165,
+      "grad_norm": 0.12914130091667175,
+      "learning_rate": 0.0001,
+      "loss": 0.1799,
+      "step": 949
+    },
+    {
+      "epoch": 1.5487447016628626,
+      "grad_norm": 0.1111481636762619,
+      "learning_rate": 0.0001,
+      "loss": 0.1854,
+      "step": 950
+    },
+    {
+      "epoch": 1.5503749592435605,
+      "grad_norm": 0.1460474282503128,
+      "learning_rate": 0.0001,
+      "loss": 0.1864,
+      "step": 951
+    },
+    {
+      "epoch": 1.5520052168242582,
+      "grad_norm": 0.13828031718730927,
+      "learning_rate": 0.0001,
+      "loss": 0.1893,
+      "step": 952
+    },
+    {
+      "epoch": 1.553635474404956,
+      "grad_norm": 0.1216074749827385,
+      "learning_rate": 0.0001,
+      "loss": 0.1866,
+      "step": 953
+    },
+    {
+      "epoch": 1.5552657319856538,
+      "grad_norm": 0.11165913194417953,
+      "learning_rate": 0.0001,
+      "loss": 0.1858,
+      "step": 954
+    },
+    {
+      "epoch": 1.5568959895663514,
+      "grad_norm": 0.12925301492214203,
+      "learning_rate": 0.0001,
+      "loss": 0.198,
+      "step": 955
+    },
+    {
+      "epoch": 1.5585262471470491,
+      "grad_norm": 0.10411135852336884,
+      "learning_rate": 0.0001,
+      "loss": 0.1836,
+      "step": 956
+    },
+    {
+      "epoch": 1.560156504727747,
+      "grad_norm": 0.1324637234210968,
+      "learning_rate": 0.0001,
+      "loss": 0.1752,
+      "step": 957
+    },
+    {
+      "epoch": 1.5617867623084447,
+      "grad_norm": 0.1328878253698349,
+      "learning_rate": 0.0001,
+      "loss": 0.1861,
+      "step": 958
+    },
+    {
+      "epoch": 1.5634170198891426,
+      "grad_norm": 0.1299583464860916,
+      "learning_rate": 0.0001,
+      "loss": 0.192,
+      "step": 959
+    },
+    {
+      "epoch": 1.5650472774698403,
+      "grad_norm": 0.14747850596904755,
+      "learning_rate": 0.0001,
+      "loss": 0.1826,
+      "step": 960
+    },
+    {
+      "epoch": 1.566677535050538,
+      "grad_norm": 0.1322248876094818,
+      "learning_rate": 0.0001,
+      "loss": 0.1924,
+      "step": 961
+    },
+    {
+      "epoch": 1.5683077926312357,
+      "grad_norm": 0.14771921932697296,
+      "learning_rate": 0.0001,
+      "loss": 0.1908,
+      "step": 962
+    },
+    {
+      "epoch": 1.5699380502119333,
+      "grad_norm": 0.12810666859149933,
+      "learning_rate": 0.0001,
+      "loss": 0.1933,
+      "step": 963
+    },
+    {
+      "epoch": 1.5715683077926312,
+      "grad_norm": 0.12468911707401276,
+      "learning_rate": 0.0001,
+      "loss": 0.1969,
+      "step": 964
+    },
+    {
+      "epoch": 1.5731985653733291,
+      "grad_norm": 0.13675253093242645,
+      "learning_rate": 0.0001,
+      "loss": 0.1838,
+      "step": 965
+    },
+    {
+      "epoch": 1.5748288229540268,
+      "grad_norm": 0.14610256254673004,
+      "learning_rate": 0.0001,
+      "loss": 0.1871,
+      "step": 966
+    },
+    {
+      "epoch": 1.5764590805347245,
+      "grad_norm": 0.16016973555088043,
+      "learning_rate": 0.0001,
+      "loss": 0.2068,
+      "step": 967
+    },
+    {
+      "epoch": 1.5780893381154222,
+      "grad_norm": 0.11924766004085541,
+      "learning_rate": 0.0001,
+      "loss": 0.1963,
+      "step": 968
+    },
+    {
+      "epoch": 1.5797195956961199,
+      "grad_norm": 0.14039623737335205,
+      "learning_rate": 0.0001,
+      "loss": 0.193,
+      "step": 969
+    },
+    {
+      "epoch": 1.5813498532768178,
+      "grad_norm": 0.13962414860725403,
+      "learning_rate": 0.0001,
+      "loss": 0.2006,
+      "step": 970
+    },
+    {
+      "epoch": 1.5829801108575154,
+      "grad_norm": 0.1010158583521843,
+      "learning_rate": 0.0001,
+      "loss": 0.1793,
+      "step": 971
+    },
+    {
+      "epoch": 1.5846103684382133,
+      "grad_norm": 0.13020698726177216,
+      "learning_rate": 0.0001,
+      "loss": 0.1936,
+      "step": 972
+    },
+    {
+      "epoch": 1.586240626018911,
+      "grad_norm": 0.13899926841259003,
+      "learning_rate": 0.0001,
+      "loss": 0.1855,
+      "step": 973
+    },
+    {
+      "epoch": 1.5878708835996087,
+      "grad_norm": 0.13506311178207397,
+      "learning_rate": 0.0001,
+      "loss": 0.2038,
+      "step": 974
+    },
+    {
+      "epoch": 1.5895011411803064,
+      "grad_norm": 0.10924082249403,
+      "learning_rate": 0.0001,
+      "loss": 0.189,
+      "step": 975
+    },
+    {
+      "epoch": 1.5911313987610043,
+      "grad_norm": 0.14923515915870667,
+      "learning_rate": 0.0001,
+      "loss": 0.1987,
+      "step": 976
+    },
+    {
+      "epoch": 1.592761656341702,
+      "grad_norm": 0.11555776745080948,
+      "learning_rate": 0.0001,
+      "loss": 0.1819,
+      "step": 977
+    },
+    {
+      "epoch": 1.5943919139223999,
+      "grad_norm": 0.11343539506196976,
+      "learning_rate": 0.0001,
+      "loss": 0.1926,
+      "step": 978
+    },
+    {
+      "epoch": 1.5960221715030976,
+      "grad_norm": 0.12200161069631577,
+      "learning_rate": 0.0001,
+      "loss": 0.1883,
+      "step": 979
+    },
+    {
+      "epoch": 1.5976524290837952,
+      "grad_norm": 0.14577162265777588,
+      "learning_rate": 0.0001,
+      "loss": 0.1957,
+      "step": 980
+    },
+    {
+      "epoch": 1.599282686664493,
+      "grad_norm": 0.12125492095947266,
+      "learning_rate": 0.0001,
+      "loss": 0.1873,
+      "step": 981
+    },
+    {
+      "epoch": 1.6009129442451906,
+      "grad_norm": 0.15497727692127228,
+      "learning_rate": 0.0001,
+      "loss": 0.2019,
+      "step": 982
+    },
+    {
+      "epoch": 1.6025432018258885,
+      "grad_norm": 0.13305650651454926,
+      "learning_rate": 0.0001,
+      "loss": 0.1951,
+      "step": 983
+    },
+    {
+      "epoch": 1.6041734594065864,
+      "grad_norm": 0.13118699193000793,
+      "learning_rate": 0.0001,
+      "loss": 0.1704,
+      "step": 984
+    },
+    {
+      "epoch": 1.605803716987284,
+      "grad_norm": 0.12898042798042297,
+      "learning_rate": 0.0001,
+      "loss": 0.1937,
+      "step": 985
+    },
+    {
+      "epoch": 1.6074339745679818,
+      "grad_norm": 0.13094180822372437,
+      "learning_rate": 0.0001,
+      "loss": 0.1938,
+      "step": 986
+    },
+    {
+      "epoch": 1.6090642321486794,
+      "grad_norm": 0.13454465568065643,
+      "learning_rate": 0.0001,
+      "loss": 0.1833,
+      "step": 987
+    },
+    {
+      "epoch": 1.6106944897293771,
+      "grad_norm": 0.10670379549264908,
+      "learning_rate": 0.0001,
+      "loss": 0.1951,
+      "step": 988
+    },
+    {
+      "epoch": 1.612324747310075,
+      "grad_norm": 0.1214357241988182,
+      "learning_rate": 0.0001,
+      "loss": 0.1837,
+      "step": 989
+    },
+    {
+      "epoch": 1.6139550048907727,
+      "grad_norm": 0.12821504473686218,
+      "learning_rate": 0.0001,
+      "loss": 0.1937,
+      "step": 990
+    },
+    {
+      "epoch": 1.6155852624714706,
+      "grad_norm": 0.1238052025437355,
+      "learning_rate": 0.0001,
+      "loss": 0.2057,
+      "step": 991
+    },
+    {
+      "epoch": 1.6172155200521683,
+      "grad_norm": 0.15875551104545593,
+      "learning_rate": 0.0001,
+      "loss": 0.2037,
+      "step": 992
+    },
+    {
+      "epoch": 1.618845777632866,
+      "grad_norm": 0.12168974429368973,
+      "learning_rate": 0.0001,
+      "loss": 0.1965,
+      "step": 993
+    },
+    {
+      "epoch": 1.6204760352135636,
+      "grad_norm": 0.20948785543441772,
+      "learning_rate": 0.0001,
+      "loss": 0.1885,
+      "step": 994
+    },
+    {
+      "epoch": 1.6221062927942613,
+      "grad_norm": 0.12937723100185394,
+      "learning_rate": 0.0001,
+      "loss": 0.1893,
+      "step": 995
+    },
+    {
+      "epoch": 1.6237365503749592,
+      "grad_norm": 0.11235075443983078,
+      "learning_rate": 0.0001,
+      "loss": 0.185,
+      "step": 996
+    },
+    {
+      "epoch": 1.6253668079556571,
+      "grad_norm": 0.1234828382730484,
+      "learning_rate": 0.0001,
+      "loss": 0.1938,
+      "step": 997
+    },
+    {
+      "epoch": 1.6269970655363548,
+      "grad_norm": 0.11725595593452454,
+      "learning_rate": 0.0001,
+      "loss": 0.1844,
+      "step": 998
+    },
+    {
+      "epoch": 1.6286273231170525,
+      "grad_norm": 0.1212792843580246,
+      "learning_rate": 0.0001,
+      "loss": 0.2011,
+      "step": 999
+    },
+    {
+      "epoch": 1.6302575806977502,
+      "grad_norm": 0.12212159484624863,
+      "learning_rate": 0.0001,
+      "loss": 0.1965,
+      "step": 1000
+    },
+    {
+      "epoch": 1.6302575806977502,
+      "eval_loss": 0.20038433372974396,
+      "eval_runtime": 2056.6687,
+      "eval_samples_per_second": 0.918,
+      "eval_steps_per_second": 0.229,
+      "step": 1000
+    },
+    {
+      "epoch": 1.6318878382784479,
+      "grad_norm": 0.102206751704216,
+      "learning_rate": 0.0001,
+      "loss": 0.1835,
+      "step": 1001
+    },
+    {
+      "epoch": 1.6335180958591458,
+      "grad_norm": 0.12285874038934708,
+      "learning_rate": 0.0001,
+      "loss": 0.1868,
+      "step": 1002
+    },
+    {
+      "epoch": 1.6351483534398437,
+      "grad_norm": 0.12729693949222565,
+      "learning_rate": 0.0001,
+      "loss": 0.185,
+      "step": 1003
+    },
+    {
+      "epoch": 1.6367786110205413,
+      "grad_norm": 0.12659911811351776,
+      "learning_rate": 0.0001,
+      "loss": 0.1847,
+      "step": 1004
+    },
+    {
+      "epoch": 1.638408868601239,
+      "grad_norm": 0.13969117403030396,
+      "learning_rate": 0.0001,
+      "loss": 0.1903,
+      "step": 1005
+    },
+    {
+      "epoch": 1.6400391261819367,
+      "grad_norm": 0.14476704597473145,
+      "learning_rate": 0.0001,
+      "loss": 0.1901,
+      "step": 1006
+    },
+    {
+      "epoch": 1.6416693837626344,
+      "grad_norm": 0.15515267848968506,
+      "learning_rate": 0.0001,
+      "loss": 0.206,
+      "step": 1007
+    },
+    {
+      "epoch": 1.6432996413433323,
+      "grad_norm": 0.140212744474411,
+      "learning_rate": 0.0001,
+      "loss": 0.1761,
+      "step": 1008
+    },
+    {
+      "epoch": 1.64492989892403,
+      "grad_norm": 0.149306982755661,
+      "learning_rate": 0.0001,
+      "loss": 0.1917,
+      "step": 1009
+    },
+    {
+      "epoch": 1.6465601565047279,
+      "grad_norm": 0.12449630349874496,
+      "learning_rate": 0.0001,
+      "loss": 0.1913,
+      "step": 1010
+    },
+    {
+      "epoch": 1.6481904140854255,
+      "grad_norm": 0.11689987778663635,
+      "learning_rate": 0.0001,
+      "loss": 0.1709,
+      "step": 1011
+    },
+    {
+      "epoch": 1.6498206716661232,
+      "grad_norm": 0.14068616926670074,
+      "learning_rate": 0.0001,
+      "loss": 0.1958,
+      "step": 1012
+    },
+    {
+      "epoch": 1.651450929246821,
+      "grad_norm": 0.11247096955776215,
+      "learning_rate": 0.0001,
+      "loss": 0.206,
+      "step": 1013
+    },
+    {
+      "epoch": 1.6530811868275186,
+      "grad_norm": 0.1336074322462082,
+      "learning_rate": 0.0001,
+      "loss": 0.1828,
+      "step": 1014
+    },
+    {
+      "epoch": 1.6547114444082165,
+      "grad_norm": 0.1070534735918045,
+      "learning_rate": 0.0001,
+      "loss": 0.188,
+      "step": 1015
+    },
+    {
+      "epoch": 1.6563417019889144,
+      "grad_norm": 0.1153770238161087,
+      "learning_rate": 0.0001,
+      "loss": 0.1818,
+      "step": 1016
+    },
+    {
+      "epoch": 1.657971959569612,
+      "grad_norm": 0.12715262174606323,
+      "learning_rate": 0.0001,
+      "loss": 0.1945,
+      "step": 1017
+    },
+    {
+      "epoch": 1.6596022171503098,
+      "grad_norm": 0.13281576335430145,
+      "learning_rate": 0.0001,
+      "loss": 0.1821,
+      "step": 1018
+    },
+    {
+      "epoch": 1.6612324747310074,
+      "grad_norm": 0.12460318207740784,
+      "learning_rate": 0.0001,
+      "loss": 0.1999,
+      "step": 1019
+    },
+    {
+      "epoch": 1.6628627323117051,
+      "grad_norm": 0.12163477391004562,
+      "learning_rate": 0.0001,
+      "loss": 0.1883,
+      "step": 1020
+    },
+    {
+      "epoch": 1.664492989892403,
+      "grad_norm": 0.11481994390487671,
+      "learning_rate": 0.0001,
+      "loss": 0.1898,
+      "step": 1021
+    },
+    {
+      "epoch": 1.6661232474731007,
+      "grad_norm": 0.11116690933704376,
+      "learning_rate": 0.0001,
+      "loss": 0.1947,
+      "step": 1022
+    },
+    {
+      "epoch": 1.6677535050537986,
+      "grad_norm": 0.12174486368894577,
+      "learning_rate": 0.0001,
+      "loss": 0.1876,
+      "step": 1023
+    },
+    {
+      "epoch": 1.6693837626344963,
+      "grad_norm": 0.13974875211715698,
+      "learning_rate": 0.0001,
+      "loss": 0.1769,
+      "step": 1024
+    },
+    {
+      "epoch": 1.671014020215194,
+      "grad_norm": 0.10470610111951828,
+      "learning_rate": 0.0001,
+      "loss": 0.1814,
+      "step": 1025
+    },
+    {
+      "epoch": 1.6726442777958916,
+      "grad_norm": 0.11099997162818909,
+      "learning_rate": 0.0001,
+      "loss": 0.1944,
+      "step": 1026
+    },
+    {
+      "epoch": 1.6742745353765895,
+      "grad_norm": 0.11604826152324677,
+      "learning_rate": 0.0001,
+      "loss": 0.1937,
+      "step": 1027
+    },
+    {
+      "epoch": 1.6759047929572872,
+      "grad_norm": 0.10957235097885132,
+      "learning_rate": 0.0001,
+      "loss": 0.1756,
+      "step": 1028
+    },
+    {
+      "epoch": 1.6775350505379851,
+      "grad_norm": 0.13742849230766296,
+      "learning_rate": 0.0001,
+      "loss": 0.182,
+      "step": 1029
+    },
+    {
+      "epoch": 1.6791653081186828,
+      "grad_norm": 0.12097762525081635,
+      "learning_rate": 0.0001,
+      "loss": 0.1996,
+      "step": 1030
+    },
+    {
+      "epoch": 1.6807955656993805,
+      "grad_norm": 0.12486281245946884,
+      "learning_rate": 0.0001,
+      "loss": 0.2007,
+      "step": 1031
+    },
+    {
+      "epoch": 1.6824258232800782,
+      "grad_norm": 0.112356036901474,
+      "learning_rate": 0.0001,
+      "loss": 0.1873,
+      "step": 1032
+    },
+    {
+      "epoch": 1.6840560808607759,
+      "grad_norm": 0.11297633498907089,
+      "learning_rate": 0.0001,
+      "loss": 0.1872,
+      "step": 1033
+    },
+    {
+      "epoch": 1.6856863384414738,
+      "grad_norm": 0.13877838850021362,
+      "learning_rate": 0.0001,
+      "loss": 0.2053,
+      "step": 1034
+    },
+    {
+      "epoch": 1.6873165960221717,
+      "grad_norm": 0.144087016582489,
+      "learning_rate": 0.0001,
+      "loss": 0.1827,
+      "step": 1035
+    },
+    {
+      "epoch": 1.6889468536028693,
+      "grad_norm": 0.1187191903591156,
+      "learning_rate": 0.0001,
+      "loss": 0.1854,
+      "step": 1036
+    },
+    {
+      "epoch": 1.690577111183567,
+      "grad_norm": 0.15161648392677307,
+      "learning_rate": 0.0001,
+      "loss": 0.1974,
+      "step": 1037
+    },
+    {
+      "epoch": 1.6922073687642647,
+      "grad_norm": 0.1534930318593979,
+      "learning_rate": 0.0001,
+      "loss": 0.1882,
+      "step": 1038
+    },
+    {
+      "epoch": 1.6938376263449624,
+      "grad_norm": 0.09909740835428238,
+      "learning_rate": 0.0001,
+      "loss": 0.1792,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6954678839256603,
+      "grad_norm": 0.13875959813594818,
+      "learning_rate": 0.0001,
+      "loss": 0.1884,
+      "step": 1040
+    },
+    {
+      "epoch": 1.697098141506358,
+      "grad_norm": 0.1244412511587143,
+      "learning_rate": 0.0001,
+      "loss": 0.1943,
+      "step": 1041
+    },
+    {
+      "epoch": 1.6987283990870559,
+      "grad_norm": 0.12313511967658997,
+      "learning_rate": 0.0001,
+      "loss": 0.1918,
+      "step": 1042
+    },
+    {
+      "epoch": 1.7003586566677535,
+      "grad_norm": 0.12056350708007812,
+      "learning_rate": 0.0001,
+      "loss": 0.1751,
+      "step": 1043
+    },
+    {
+      "epoch": 1.7019889142484512,
+      "grad_norm": 0.14450062811374664,
+      "learning_rate": 0.0001,
+      "loss": 0.1998,
+      "step": 1044
+    },
+    {
+      "epoch": 1.703619171829149,
+      "grad_norm": 0.13067997992038727,
+      "learning_rate": 0.0001,
+      "loss": 0.1956,
+      "step": 1045
+    },
+    {
+      "epoch": 1.7052494294098468,
+      "grad_norm": 0.1026124432682991,
+      "learning_rate": 0.0001,
+      "loss": 0.1906,
+      "step": 1046
+    },
+    {
+      "epoch": 1.7068796869905445,
+      "grad_norm": 0.12087104469537735,
+      "learning_rate": 0.0001,
+      "loss": 0.1803,
+      "step": 1047
+    },
+    {
+      "epoch": 1.7085099445712424,
+      "grad_norm": 0.12124813348054886,
+      "learning_rate": 0.0001,
+      "loss": 0.1974,
+      "step": 1048
+    },
+    {
+      "epoch": 1.71014020215194,
+      "grad_norm": 0.12620921432971954,
+      "learning_rate": 0.0001,
+      "loss": 0.1952,
+      "step": 1049
+    },
+    {
+      "epoch": 1.7117704597326378,
+      "grad_norm": 0.12863439321517944,
+      "learning_rate": 0.0001,
+      "loss": 0.1943,
+      "step": 1050
+    },
+    {
+      "epoch": 1.7134007173133354,
+      "grad_norm": 0.11415056884288788,
+      "learning_rate": 0.0001,
+      "loss": 0.1928,
+      "step": 1051
+    },
+    {
+      "epoch": 1.7150309748940331,
+      "grad_norm": 0.11588121205568314,
+      "learning_rate": 0.0001,
+      "loss": 0.1833,
+      "step": 1052
+    },
+    {
+      "epoch": 1.716661232474731,
+      "grad_norm": 0.11711575090885162,
+      "learning_rate": 0.0001,
+      "loss": 0.1854,
+      "step": 1053
+    },
+    {
+      "epoch": 1.718291490055429,
+      "grad_norm": 0.11321627348661423,
+      "learning_rate": 0.0001,
+      "loss": 0.1908,
+      "step": 1054
+    },
+    {
+      "epoch": 1.7199217476361266,
+      "grad_norm": 0.11224471032619476,
+      "learning_rate": 0.0001,
+      "loss": 0.1812,
+      "step": 1055
+    },
+    {
+      "epoch": 1.7215520052168243,
+      "grad_norm": 0.09902875125408173,
+      "learning_rate": 0.0001,
+      "loss": 0.185,
+      "step": 1056
+    },
+    {
+      "epoch": 1.723182262797522,
+      "grad_norm": 0.12637241184711456,
+      "learning_rate": 0.0001,
+      "loss": 0.2065,
+      "step": 1057
+    },
+    {
+      "epoch": 1.7248125203782196,
+      "grad_norm": 0.10967186093330383,
+      "learning_rate": 0.0001,
+      "loss": 0.1789,
+      "step": 1058
+    },
+    {
+      "epoch": 1.7264427779589175,
+      "grad_norm": 0.11306209117174149,
+      "learning_rate": 0.0001,
+      "loss": 0.1848,
+      "step": 1059
+    },
+    {
+      "epoch": 1.7280730355396152,
+      "grad_norm": 0.11642675846815109,
+      "learning_rate": 0.0001,
+      "loss": 0.19,
+      "step": 1060
+    },
+    {
+      "epoch": 1.7297032931203131,
+      "grad_norm": 0.14287948608398438,
+      "learning_rate": 0.0001,
+      "loss": 0.198,
+      "step": 1061
+    },
+    {
+      "epoch": 1.7313335507010108,
+      "grad_norm": 0.14489899575710297,
+      "learning_rate": 0.0001,
+      "loss": 0.1898,
+      "step": 1062
+    },
+    {
+      "epoch": 1.7329638082817085,
+      "grad_norm": 0.12616468966007233,
+      "learning_rate": 0.0001,
+      "loss": 0.181,
+      "step": 1063
+    },
+    {
+      "epoch": 1.7345940658624062,
+      "grad_norm": 0.1419239342212677,
+      "learning_rate": 0.0001,
+      "loss": 0.1981,
+      "step": 1064
+    },
+    {
+      "epoch": 1.7362243234431038,
+      "grad_norm": 0.12401966750621796,
+      "learning_rate": 0.0001,
+      "loss": 0.1822,
+      "step": 1065
+    },
+    {
+      "epoch": 1.7378545810238017,
+      "grad_norm": 0.15133000910282135,
+      "learning_rate": 0.0001,
+      "loss": 0.1974,
+      "step": 1066
+    },
+    {
+      "epoch": 1.7394848386044997,
+      "grad_norm": 0.1293191760778427,
+      "learning_rate": 0.0001,
+      "loss": 0.1904,
+      "step": 1067
+    },
+    {
+      "epoch": 1.7411150961851973,
+      "grad_norm": 0.12822459638118744,
+      "learning_rate": 0.0001,
+      "loss": 0.1893,
+      "step": 1068
+    },
+    {
+      "epoch": 1.742745353765895,
+      "grad_norm": 0.10721483081579208,
+      "learning_rate": 0.0001,
+      "loss": 0.1752,
+      "step": 1069
+    },
+    {
+      "epoch": 1.7443756113465927,
+      "grad_norm": 0.13893002271652222,
+      "learning_rate": 0.0001,
+      "loss": 0.1981,
+      "step": 1070
+    },
+    {
+      "epoch": 1.7460058689272904,
+      "grad_norm": 0.1082753837108612,
+      "learning_rate": 0.0001,
+      "loss": 0.1925,
+      "step": 1071
+    },
+    {
+      "epoch": 1.7476361265079883,
+      "grad_norm": 0.12488586455583572,
+      "learning_rate": 0.0001,
+      "loss": 0.1759,
+      "step": 1072
+    },
+    {
+      "epoch": 1.7492663840886862,
+      "grad_norm": 0.10307318717241287,
+      "learning_rate": 0.0001,
+      "loss": 0.1685,
+      "step": 1073
+    },
+    {
+      "epoch": 1.7508966416693839,
+      "grad_norm": 0.12732993066310883,
+      "learning_rate": 0.0001,
+      "loss": 0.1843,
+      "step": 1074
+    },
+    {
+      "epoch": 1.7525268992500815,
+      "grad_norm": 0.12339639663696289,
+      "learning_rate": 0.0001,
+      "loss": 0.185,
+      "step": 1075
+    },
+    {
+      "epoch": 1.7541571568307792,
+      "grad_norm": 0.10467349737882614,
+      "learning_rate": 0.0001,
+      "loss": 0.1787,
+      "step": 1076
+    },
+    {
+      "epoch": 1.755787414411477,
+      "grad_norm": 0.1297389268875122,
+      "learning_rate": 0.0001,
+      "loss": 0.1865,
+      "step": 1077
+    },
+    {
+      "epoch": 1.7574176719921748,
+      "grad_norm": 0.13278993964195251,
+      "learning_rate": 0.0001,
+      "loss": 0.1872,
+      "step": 1078
+    },
+    {
+      "epoch": 1.7590479295728725,
+      "grad_norm": 0.125654399394989,
+      "learning_rate": 0.0001,
+      "loss": 0.1731,
+      "step": 1079
+    },
+    {
+      "epoch": 1.7606781871535704,
+      "grad_norm": 0.12131417542695999,
+      "learning_rate": 0.0001,
+      "loss": 0.1717,
+      "step": 1080
+    },
+    {
+      "epoch": 1.762308444734268,
+      "grad_norm": 0.12412076443433762,
+      "learning_rate": 0.0001,
+      "loss": 0.1819,
+      "step": 1081
+    },
+    {
+      "epoch": 1.7639387023149657,
+      "grad_norm": 0.12504039704799652,
+      "learning_rate": 0.0001,
+      "loss": 0.1714,
+      "step": 1082
+    },
+    {
+      "epoch": 1.7655689598956634,
+      "grad_norm": 0.12615200877189636,
+      "learning_rate": 0.0001,
+      "loss": 0.1982,
+      "step": 1083
+    },
+    {
+      "epoch": 1.767199217476361,
+      "grad_norm": 0.13149379193782806,
+      "learning_rate": 0.0001,
+      "loss": 0.1983,
+      "step": 1084
+    },
+    {
+      "epoch": 1.768829475057059,
+      "grad_norm": 0.13638049364089966,
+      "learning_rate": 0.0001,
+      "loss": 0.1938,
+      "step": 1085
+    },
+    {
+      "epoch": 1.770459732637757,
+      "grad_norm": 0.10553500801324844,
+      "learning_rate": 0.0001,
+      "loss": 0.1867,
+      "step": 1086
+    },
+    {
+      "epoch": 1.7720899902184546,
+      "grad_norm": 0.13788992166519165,
+      "learning_rate": 0.0001,
+      "loss": 0.1825,
+      "step": 1087
+    },
+    {
+      "epoch": 1.7737202477991523,
+      "grad_norm": 0.11936747282743454,
+      "learning_rate": 0.0001,
+      "loss": 0.1885,
+      "step": 1088
+    },
+    {
+      "epoch": 1.77535050537985,
+      "grad_norm": 0.133977010846138,
+      "learning_rate": 0.0001,
+      "loss": 0.1954,
+      "step": 1089
+    },
+    {
+      "epoch": 1.7769807629605476,
+      "grad_norm": 0.12340915203094482,
+      "learning_rate": 0.0001,
+      "loss": 0.1825,
+      "step": 1090
+    },
+    {
+      "epoch": 1.7786110205412455,
+      "grad_norm": 0.11585668474435806,
+      "learning_rate": 0.0001,
+      "loss": 0.1854,
+      "step": 1091
+    },
+    {
+      "epoch": 1.7802412781219432,
+      "grad_norm": 0.1759333461523056,
+      "learning_rate": 0.0001,
+      "loss": 0.1947,
+      "step": 1092
+    },
+    {
+      "epoch": 1.7818715357026411,
+      "grad_norm": 0.12591791152954102,
+      "learning_rate": 0.0001,
+      "loss": 0.1897,
+      "step": 1093
+    },
+    {
+      "epoch": 1.7835017932833388,
+      "grad_norm": 0.11980108171701431,
+      "learning_rate": 0.0001,
+      "loss": 0.1962,
+      "step": 1094
+    },
+    {
+      "epoch": 1.7851320508640365,
+      "grad_norm": 0.12715961039066315,
+      "learning_rate": 0.0001,
+      "loss": 0.1961,
+      "step": 1095
+    },
+    {
+      "epoch": 1.7867623084447342,
+      "grad_norm": 0.1274608075618744,
+      "learning_rate": 0.0001,
+      "loss": 0.1855,
+      "step": 1096
+    },
+    {
+      "epoch": 1.788392566025432,
+      "grad_norm": 0.15268570184707642,
+      "learning_rate": 0.0001,
+      "loss": 0.1918,
+      "step": 1097
+    },
+    {
+      "epoch": 1.7900228236061297,
+      "grad_norm": 0.1284448206424713,
+      "learning_rate": 0.0001,
+      "loss": 0.1912,
+      "step": 1098
+    },
+    {
+      "epoch": 1.7916530811868276,
+      "grad_norm": 0.1219286322593689,
+      "learning_rate": 0.0001,
+      "loss": 0.1792,
+      "step": 1099
+    },
+    {
+      "epoch": 1.7932833387675253,
+      "grad_norm": 0.11822406947612762,
+      "learning_rate": 0.0001,
+      "loss": 0.1838,
+      "step": 1100
+    },
+    {
+      "epoch": 1.794913596348223,
+      "grad_norm": 0.12364920228719711,
+      "learning_rate": 0.0001,
+      "loss": 0.1897,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7965438539289207,
+      "grad_norm": 0.11299461871385574,
+      "learning_rate": 0.0001,
+      "loss": 0.1955,
+      "step": 1102
+    },
+    {
+      "epoch": 1.7981741115096184,
+      "grad_norm": 0.13961385190486908,
+      "learning_rate": 0.0001,
+      "loss": 0.1899,
+      "step": 1103
+    },
+    {
+      "epoch": 1.7998043690903163,
+      "grad_norm": 0.1028355211019516,
+      "learning_rate": 0.0001,
+      "loss": 0.1776,
+      "step": 1104
+    },
+    {
+      "epoch": 1.8014346266710142,
+      "grad_norm": 0.107155442237854,
+      "learning_rate": 0.0001,
+      "loss": 0.1731,
+      "step": 1105
+    },
+    {
+      "epoch": 1.8030648842517119,
+      "grad_norm": 0.11978352069854736,
+      "learning_rate": 0.0001,
+      "loss": 0.1929,
+      "step": 1106
+    },
+    {
+      "epoch": 1.8046951418324095,
+      "grad_norm": 0.12182936072349548,
+      "learning_rate": 0.0001,
+      "loss": 0.1874,
+      "step": 1107
+    },
+    {
+      "epoch": 1.8063253994131072,
+      "grad_norm": 0.1270841509103775,
+      "learning_rate": 0.0001,
+      "loss": 0.1832,
+      "step": 1108
+    },
+    {
+      "epoch": 1.807955656993805,
+      "grad_norm": 0.10812801122665405,
+      "learning_rate": 0.0001,
+      "loss": 0.1704,
+      "step": 1109
+    },
+    {
+      "epoch": 1.8095859145745028,
+      "grad_norm": 0.11686091870069504,
+      "learning_rate": 0.0001,
+      "loss": 0.1765,
+      "step": 1110
+    },
+    {
+      "epoch": 1.8112161721552005,
+      "grad_norm": 0.1108395904302597,
+      "learning_rate": 0.0001,
+      "loss": 0.1832,
+      "step": 1111
+    },
+    {
+      "epoch": 1.8128464297358984,
+      "grad_norm": 0.13820597529411316,
+      "learning_rate": 0.0001,
+      "loss": 0.186,
+      "step": 1112
+    },
+    {
+      "epoch": 1.814476687316596,
+      "grad_norm": 0.1270742267370224,
+      "learning_rate": 0.0001,
+      "loss": 0.191,
+      "step": 1113
+    },
+    {
+      "epoch": 1.8161069448972937,
+      "grad_norm": 0.160540372133255,
+      "learning_rate": 0.0001,
+      "loss": 0.1821,
+      "step": 1114
+    },
+    {
+      "epoch": 1.8177372024779914,
+      "grad_norm": 0.11482474952936172,
+      "learning_rate": 0.0001,
+      "loss": 0.1914,
+      "step": 1115
+    },
+    {
+      "epoch": 1.8193674600586893,
+      "grad_norm": 0.13658450543880463,
+      "learning_rate": 0.0001,
+      "loss": 0.173,
+      "step": 1116
+    },
+    {
+      "epoch": 1.820997717639387,
+      "grad_norm": 0.10865119844675064,
+      "learning_rate": 0.0001,
+      "loss": 0.1821,
+      "step": 1117
+    },
+    {
+      "epoch": 1.822627975220085,
+      "grad_norm": 0.10327646881341934,
+      "learning_rate": 0.0001,
+      "loss": 0.1744,
+      "step": 1118
+    },
+    {
+      "epoch": 1.8242582328007826,
+      "grad_norm": 0.10378043353557587,
+      "learning_rate": 0.0001,
+      "loss": 0.1717,
+      "step": 1119
+    },
+    {
+      "epoch": 1.8258884903814803,
+      "grad_norm": 0.11684548109769821,
+      "learning_rate": 0.0001,
+      "loss": 0.1893,
+      "step": 1120
+    },
+    {
+      "epoch": 1.827518747962178,
+      "grad_norm": 0.1185649037361145,
+      "learning_rate": 0.0001,
+      "loss": 0.1865,
+      "step": 1121
+    },
+    {
+      "epoch": 1.8291490055428756,
+      "grad_norm": 0.11936715245246887,
+      "learning_rate": 0.0001,
+      "loss": 0.1844,
+      "step": 1122
+    },
+    {
+      "epoch": 1.8307792631235735,
+      "grad_norm": 0.10836116224527359,
+      "learning_rate": 0.0001,
+      "loss": 0.1826,
+      "step": 1123
+    },
+    {
+      "epoch": 1.8324095207042714,
+      "grad_norm": 0.10425475239753723,
+      "learning_rate": 0.0001,
+      "loss": 0.1873,
+      "step": 1124
+    },
+    {
+      "epoch": 1.8340397782849691,
+      "grad_norm": 0.12443797290325165,
+      "learning_rate": 0.0001,
+      "loss": 0.1775,
+      "step": 1125
+    },
+    {
+      "epoch": 1.8356700358656668,
+      "grad_norm": 0.1314772516489029,
+      "learning_rate": 0.0001,
+      "loss": 0.1825,
+      "step": 1126
+    },
+    {
+      "epoch": 1.8373002934463645,
+      "grad_norm": 0.1321101188659668,
+      "learning_rate": 0.0001,
+      "loss": 0.1775,
+      "step": 1127
+    },
+    {
+      "epoch": 1.8389305510270622,
+      "grad_norm": 0.11217895895242691,
+      "learning_rate": 0.0001,
+      "loss": 0.1814,
+      "step": 1128
+    },
+    {
+      "epoch": 1.84056080860776,
+      "grad_norm": 0.11161702871322632,
+      "learning_rate": 0.0001,
+      "loss": 0.1856,
+      "step": 1129
+    },
+    {
+      "epoch": 1.8421910661884577,
+      "grad_norm": 0.11947924643754959,
+      "learning_rate": 0.0001,
+      "loss": 0.1818,
+      "step": 1130
+    },
+    {
+      "epoch": 1.8438213237691556,
+      "grad_norm": 0.15216386318206787,
+      "learning_rate": 0.0001,
+      "loss": 0.18,
+      "step": 1131
+    },
+    {
+      "epoch": 1.8454515813498533,
+      "grad_norm": 0.12306888401508331,
+      "learning_rate": 0.0001,
+      "loss": 0.1918,
+      "step": 1132
+    },
+    {
+      "epoch": 1.847081838930551,
+      "grad_norm": 0.1321091651916504,
+      "learning_rate": 0.0001,
+      "loss": 0.1945,
+      "step": 1133
+    },
+    {
+      "epoch": 1.8487120965112487,
+      "grad_norm": 0.12340245395898819,
+      "learning_rate": 0.0001,
+      "loss": 0.1853,
+      "step": 1134
+    },
+    {
+      "epoch": 1.8503423540919464,
+      "grad_norm": 0.1099599152803421,
+      "learning_rate": 0.0001,
+      "loss": 0.1979,
+      "step": 1135
+    },
+    {
+      "epoch": 1.8519726116726443,
+      "grad_norm": 0.13142207264900208,
+      "learning_rate": 0.0001,
+      "loss": 0.1811,
+      "step": 1136
+    },
+    {
+      "epoch": 1.8536028692533422,
+      "grad_norm": 0.11548743396997452,
+      "learning_rate": 0.0001,
+      "loss": 0.1958,
+      "step": 1137
+    },
+    {
+      "epoch": 1.8552331268340398,
+      "grad_norm": 0.1179068312048912,
+      "learning_rate": 0.0001,
+      "loss": 0.1836,
+      "step": 1138
+    },
+    {
+      "epoch": 1.8568633844147375,
+      "grad_norm": 0.1258961409330368,
+      "learning_rate": 0.0001,
+      "loss": 0.1883,
+      "step": 1139
+    },
+    {
+      "epoch": 1.8584936419954352,
+      "grad_norm": 0.11142059415578842,
+      "learning_rate": 0.0001,
+      "loss": 0.1921,
+      "step": 1140
+    },
+    {
+      "epoch": 1.8601238995761329,
+      "grad_norm": 0.11219684034585953,
+      "learning_rate": 0.0001,
+      "loss": 0.1818,
+      "step": 1141
+    },
+    {
+      "epoch": 1.8617541571568308,
+      "grad_norm": 0.11732007563114166,
+      "learning_rate": 0.0001,
+      "loss": 0.185,
+      "step": 1142
+    },
+    {
+      "epoch": 1.8633844147375287,
+      "grad_norm": 0.10912680625915527,
+      "learning_rate": 0.0001,
+      "loss": 0.1823,
+      "step": 1143
+    },
+    {
+      "epoch": 1.8650146723182264,
+      "grad_norm": 0.13096576929092407,
+      "learning_rate": 0.0001,
+      "loss": 0.1958,
+      "step": 1144
+    },
+    {
+      "epoch": 1.866644929898924,
+      "grad_norm": 0.1521504819393158,
+      "learning_rate": 0.0001,
+      "loss": 0.1884,
+      "step": 1145
+    },
+    {
+      "epoch": 1.8682751874796217,
+      "grad_norm": 0.12232112884521484,
+      "learning_rate": 0.0001,
+      "loss": 0.2035,
+      "step": 1146
+    },
+    {
+      "epoch": 1.8699054450603194,
+      "grad_norm": 0.10666659474372864,
+      "learning_rate": 0.0001,
+      "loss": 0.1678,
+      "step": 1147
+    },
+    {
+      "epoch": 1.8715357026410173,
+      "grad_norm": 0.12789608538150787,
+      "learning_rate": 0.0001,
+      "loss": 0.1705,
+      "step": 1148
+    },
+    {
+      "epoch": 1.873165960221715,
+      "grad_norm": 0.1496013104915619,
+      "learning_rate": 0.0001,
+      "loss": 0.2019,
+      "step": 1149
+    },
+    {
+      "epoch": 1.874796217802413,
+      "grad_norm": 0.1169729083776474,
+      "learning_rate": 0.0001,
+      "loss": 0.1706,
+      "step": 1150
+    },
+    {
+      "epoch": 1.8764264753831106,
+      "grad_norm": 0.14244718849658966,
+      "learning_rate": 0.0001,
+      "loss": 0.1868,
+      "step": 1151
+    },
+    {
+      "epoch": 1.8780567329638083,
+      "grad_norm": 0.12071932852268219,
+      "learning_rate": 0.0001,
+      "loss": 0.186,
+      "step": 1152
+    },
+    {
+      "epoch": 1.879686990544506,
+      "grad_norm": 0.12490394711494446,
+      "learning_rate": 0.0001,
+      "loss": 0.1912,
+      "step": 1153
+    },
+    {
+      "epoch": 1.8813172481252036,
+      "grad_norm": 0.13060052692890167,
+      "learning_rate": 0.0001,
+      "loss": 0.2214,
+      "step": 1154
+    },
+    {
+      "epoch": 1.8829475057059015,
+      "grad_norm": 0.12722712755203247,
+      "learning_rate": 0.0001,
+      "loss": 0.1813,
+      "step": 1155
+    },
+    {
+      "epoch": 1.8845777632865994,
+      "grad_norm": 0.12535281479358673,
+      "learning_rate": 0.0001,
+      "loss": 0.1961,
+      "step": 1156
+    },
+    {
+      "epoch": 1.886208020867297,
+      "grad_norm": 0.11369742453098297,
+      "learning_rate": 0.0001,
+      "loss": 0.192,
+      "step": 1157
+    },
+    {
+      "epoch": 1.8878382784479948,
+      "grad_norm": 0.1047392338514328,
+      "learning_rate": 0.0001,
+      "loss": 0.1765,
+      "step": 1158
+    },
+    {
+      "epoch": 1.8894685360286925,
+      "grad_norm": 0.13275443017482758,
+      "learning_rate": 0.0001,
+      "loss": 0.1865,
+      "step": 1159
+    },
+    {
+      "epoch": 1.8910987936093901,
+      "grad_norm": 0.11752204596996307,
+      "learning_rate": 0.0001,
+      "loss": 0.1886,
+      "step": 1160
+    },
+    {
+      "epoch": 1.892729051190088,
+      "grad_norm": 0.11460895091295242,
+      "learning_rate": 0.0001,
+      "loss": 0.1898,
+      "step": 1161
+    },
+    {
+      "epoch": 1.8943593087707857,
+      "grad_norm": 0.12889966368675232,
+      "learning_rate": 0.0001,
+      "loss": 0.1826,
+      "step": 1162
+    },
+    {
+      "epoch": 1.8959895663514836,
+      "grad_norm": 0.11824796348810196,
+      "learning_rate": 0.0001,
+      "loss": 0.1775,
+      "step": 1163
+    },
+    {
+      "epoch": 1.8976198239321813,
+      "grad_norm": 0.10617636889219284,
+      "learning_rate": 0.0001,
+      "loss": 0.1718,
+      "step": 1164
+    },
+    {
+      "epoch": 1.899250081512879,
+      "grad_norm": 0.12249031662940979,
+      "learning_rate": 0.0001,
+      "loss": 0.1732,
+      "step": 1165
+    },
+    {
+      "epoch": 1.9008803390935767,
+      "grad_norm": 0.1171279028058052,
+      "learning_rate": 0.0001,
+      "loss": 0.1837,
+      "step": 1166
+    },
+    {
+      "epoch": 1.9025105966742746,
+      "grad_norm": 0.12532579898834229,
+      "learning_rate": 0.0001,
+      "loss": 0.1846,
+      "step": 1167
+    },
+    {
+      "epoch": 1.9041408542549723,
+      "grad_norm": 0.13807226717472076,
+      "learning_rate": 0.0001,
+      "loss": 0.18,
+      "step": 1168
+    },
+    {
+      "epoch": 1.9057711118356702,
+      "grad_norm": 0.1306944191455841,
+      "learning_rate": 0.0001,
+      "loss": 0.1849,
+      "step": 1169
+    },
+    {
+      "epoch": 1.9074013694163678,
+      "grad_norm": 0.3306502401828766,
+      "learning_rate": 0.0001,
+      "loss": 0.2122,
+      "step": 1170
+    },
+    {
+      "epoch": 1.9090316269970655,
+      "grad_norm": 0.1267193704843521,
+      "learning_rate": 0.0001,
+      "loss": 0.1893,
+      "step": 1171
+    },
+    {
+      "epoch": 1.9106618845777632,
+      "grad_norm": 0.1224604994058609,
+      "learning_rate": 0.0001,
+      "loss": 0.1989,
+      "step": 1172
+    },
+    {
+      "epoch": 1.9122921421584609,
+      "grad_norm": 0.11894525587558746,
+      "learning_rate": 0.0001,
+      "loss": 0.1731,
+      "step": 1173
+    },
+    {
+      "epoch": 1.9139223997391588,
+      "grad_norm": 0.11913865804672241,
+      "learning_rate": 0.0001,
+      "loss": 0.1845,
+      "step": 1174
+    },
+    {
+      "epoch": 1.9155526573198567,
+      "grad_norm": 0.11289380490779877,
+      "learning_rate": 0.0001,
+      "loss": 0.182,
+      "step": 1175
+    },
+    {
+      "epoch": 1.9171829149005544,
+      "grad_norm": 0.13549114763736725,
+      "learning_rate": 0.0001,
+      "loss": 0.206,
+      "step": 1176
+    },
+    {
+      "epoch": 1.918813172481252,
+      "grad_norm": 0.1213872879743576,
+      "learning_rate": 0.0001,
+      "loss": 0.1779,
+      "step": 1177
+    },
+    {
+      "epoch": 1.9204434300619497,
+      "grad_norm": 0.13372060656547546,
+      "learning_rate": 0.0001,
+      "loss": 0.1761,
+      "step": 1178
+    },
+    {
+      "epoch": 1.9220736876426474,
+      "grad_norm": 0.09771383553743362,
+      "learning_rate": 0.0001,
+      "loss": 0.1687,
+      "step": 1179
+    },
+    {
+      "epoch": 1.9237039452233453,
+      "grad_norm": 0.12244701385498047,
+      "learning_rate": 0.0001,
+      "loss": 0.1827,
+      "step": 1180
+    },
+    {
+      "epoch": 1.925334202804043,
+      "grad_norm": 0.1301904022693634,
+      "learning_rate": 0.0001,
+      "loss": 0.1814,
+      "step": 1181
+    },
+    {
+      "epoch": 1.926964460384741,
+      "grad_norm": 0.14816220104694366,
+      "learning_rate": 0.0001,
+      "loss": 0.2002,
+      "step": 1182
+    },
+    {
+      "epoch": 1.9285947179654386,
+      "grad_norm": 0.11727407574653625,
+      "learning_rate": 0.0001,
+      "loss": 0.1835,
+      "step": 1183
+    },
+    {
+      "epoch": 1.9302249755461363,
+      "grad_norm": 0.13211950659751892,
+      "learning_rate": 0.0001,
+      "loss": 0.1859,
+      "step": 1184
+    },
+    {
+      "epoch": 1.931855233126834,
+      "grad_norm": 0.12375782430171967,
+      "learning_rate": 0.0001,
+      "loss": 0.1837,
+      "step": 1185
+    },
+    {
+      "epoch": 1.9334854907075318,
+      "grad_norm": 0.14515690505504608,
+      "learning_rate": 0.0001,
+      "loss": 0.1757,
+      "step": 1186
+    },
+    {
+      "epoch": 1.9351157482882295,
+      "grad_norm": 0.11579223722219467,
+      "learning_rate": 0.0001,
+      "loss": 0.1767,
+      "step": 1187
+    },
+    {
+      "epoch": 1.9367460058689274,
+      "grad_norm": 0.11580117791891098,
+      "learning_rate": 0.0001,
+      "loss": 0.1745,
+      "step": 1188
+    },
+    {
+      "epoch": 1.938376263449625,
+      "grad_norm": 0.13860304653644562,
+      "learning_rate": 0.0001,
+      "loss": 0.1899,
+      "step": 1189
+    },
+    {
+      "epoch": 1.9400065210303228,
+      "grad_norm": 0.10386091470718384,
+      "learning_rate": 0.0001,
+      "loss": 0.1799,
+      "step": 1190
+    },
+    {
+      "epoch": 1.9416367786110205,
+      "grad_norm": 0.12089010328054428,
+      "learning_rate": 0.0001,
+      "loss": 0.1886,
+      "step": 1191
+    },
+    {
+      "epoch": 1.9432670361917181,
+      "grad_norm": 0.13745586574077606,
+      "learning_rate": 0.0001,
+      "loss": 0.1865,
+      "step": 1192
+    },
+    {
+      "epoch": 1.944897293772416,
+      "grad_norm": 0.12096656113862991,
+      "learning_rate": 0.0001,
+      "loss": 0.1929,
+      "step": 1193
+    },
+    {
+      "epoch": 1.946527551353114,
+      "grad_norm": 0.24482771754264832,
+      "learning_rate": 0.0001,
+      "loss": 0.1911,
+      "step": 1194
+    },
+    {
+      "epoch": 1.9481578089338116,
+      "grad_norm": 0.13980430364608765,
+      "learning_rate": 0.0001,
+      "loss": 0.1951,
+      "step": 1195
+    },
+    {
+      "epoch": 1.9497880665145093,
+      "grad_norm": 0.1281464695930481,
+      "learning_rate": 0.0001,
+      "loss": 0.1797,
+      "step": 1196
+    },
+    {
+      "epoch": 1.951418324095207,
+      "grad_norm": 0.11813049763441086,
+      "learning_rate": 0.0001,
+      "loss": 0.1836,
+      "step": 1197
+    },
+    {
+      "epoch": 1.9530485816759047,
+      "grad_norm": 0.13674259185791016,
+      "learning_rate": 0.0001,
+      "loss": 0.1921,
+      "step": 1198
+    },
+    {
+      "epoch": 1.9546788392566026,
+      "grad_norm": 0.12205852568149567,
+      "learning_rate": 0.0001,
+      "loss": 0.1726,
+      "step": 1199
+    },
+    {
+      "epoch": 1.9563090968373003,
+      "grad_norm": 0.12858718633651733,
+      "learning_rate": 0.0001,
+      "loss": 0.1814,
+      "step": 1200
+    },
+    {
+      "epoch": 1.9579393544179982,
+      "grad_norm": 0.1414395272731781,
+      "learning_rate": 0.0001,
+      "loss": 0.1791,
+      "step": 1201
+    },
+    {
+      "epoch": 1.9595696119986958,
+      "grad_norm": 0.10766829550266266,
+      "learning_rate": 0.0001,
+      "loss": 0.1905,
+      "step": 1202
+    },
+    {
+      "epoch": 1.9611998695793935,
+      "grad_norm": 0.12443807721138,
+      "learning_rate": 0.0001,
+      "loss": 0.1847,
+      "step": 1203
+    },
+    {
+      "epoch": 1.9628301271600912,
+      "grad_norm": 0.1413409411907196,
+      "learning_rate": 0.0001,
+      "loss": 0.1943,
+      "step": 1204
+    },
+    {
+      "epoch": 1.9644603847407889,
+      "grad_norm": 0.12135512381792068,
+      "learning_rate": 0.0001,
+      "loss": 0.1957,
+      "step": 1205
+    },
+    {
+      "epoch": 1.9660906423214868,
+      "grad_norm": 0.15406377613544464,
+      "learning_rate": 0.0001,
+      "loss": 0.1799,
+      "step": 1206
+    },
+    {
+      "epoch": 1.9677208999021847,
+      "grad_norm": 0.10838975757360458,
+      "learning_rate": 0.0001,
+      "loss": 0.1849,
+      "step": 1207
+    },
+    {
+      "epoch": 1.9693511574828824,
+      "grad_norm": 0.12363716959953308,
+      "learning_rate": 0.0001,
+      "loss": 0.2111,
+      "step": 1208
+    },
+    {
+      "epoch": 1.97098141506358,
+      "grad_norm": 0.10429894924163818,
+      "learning_rate": 0.0001,
+      "loss": 0.1778,
+      "step": 1209
+    },
+    {
+      "epoch": 1.9726116726442777,
+      "grad_norm": 0.11979079246520996,
+      "learning_rate": 0.0001,
+      "loss": 0.197,
+      "step": 1210
+    },
+    {
+      "epoch": 1.9742419302249754,
+      "grad_norm": 0.10798018425703049,
+      "learning_rate": 0.0001,
+      "loss": 0.1697,
+      "step": 1211
+    },
+    {
+      "epoch": 1.9758721878056733,
+      "grad_norm": 0.11077526211738586,
+      "learning_rate": 0.0001,
+      "loss": 0.1881,
+      "step": 1212
+    },
+    {
+      "epoch": 1.9775024453863712,
+      "grad_norm": 0.13390393555164337,
+      "learning_rate": 0.0001,
+      "loss": 0.1909,
+      "step": 1213
+    },
+    {
+      "epoch": 1.979132702967069,
+      "grad_norm": 0.11331623047590256,
+      "learning_rate": 0.0001,
+      "loss": 0.1869,
+      "step": 1214
+    },
+    {
+      "epoch": 1.9807629605477666,
+      "grad_norm": 0.12054945528507233,
+      "learning_rate": 0.0001,
+      "loss": 0.2016,
+      "step": 1215
+    },
+    {
+      "epoch": 1.9823932181284643,
+      "grad_norm": 0.11420212686061859,
+      "learning_rate": 0.0001,
+      "loss": 0.1714,
+      "step": 1216
+    },
+    {
+      "epoch": 1.984023475709162,
+      "grad_norm": 0.1471630334854126,
+      "learning_rate": 0.0001,
+      "loss": 0.1991,
+      "step": 1217
+    },
+    {
+      "epoch": 1.9856537332898598,
+      "grad_norm": 0.11184452474117279,
+      "learning_rate": 0.0001,
+      "loss": 0.1728,
+      "step": 1218
+    },
+    {
+      "epoch": 1.9872839908705575,
+      "grad_norm": 0.12244154512882233,
+      "learning_rate": 0.0001,
+      "loss": 0.1725,
+      "step": 1219
+    },
+    {
+      "epoch": 1.9889142484512554,
+      "grad_norm": 0.11920984834432602,
+      "learning_rate": 0.0001,
+      "loss": 0.1746,
+      "step": 1220
+    },
+    {
+      "epoch": 1.990544506031953,
+      "grad_norm": 0.14389824867248535,
+      "learning_rate": 0.0001,
+      "loss": 0.192,
+      "step": 1221
+    },
+    {
+      "epoch": 1.9921747636126508,
+      "grad_norm": 0.13001768290996552,
+      "learning_rate": 0.0001,
+      "loss": 0.1988,
+      "step": 1222
+    },
+    {
+      "epoch": 1.9938050211933485,
+      "grad_norm": 0.2044341266155243,
+      "learning_rate": 0.0001,
+      "loss": 0.1783,
+      "step": 1223
+    },
+    {
+      "epoch": 1.9954352787740461,
+      "grad_norm": 0.11807113885879517,
+      "learning_rate": 0.0001,
+      "loss": 0.1748,
+      "step": 1224
+    },
+    {
+      "epoch": 1.997065536354744,
+      "grad_norm": 0.12389995157718658,
+      "learning_rate": 0.0001,
+      "loss": 0.1968,
+      "step": 1225
+    },
+    {
+      "epoch": 1.998695793935442,
+      "grad_norm": 0.11298985034227371,
+      "learning_rate": 0.0001,
+      "loss": 0.1781,
+      "step": 1226
+    },
+    {
+      "epoch": 2.0003260515161396,
+      "grad_norm": 0.15368470549583435,
+      "learning_rate": 0.0001,
+      "loss": 0.1889,
+      "step": 1227
+    },
+    {
+      "epoch": 2.0019563090968373,
+      "grad_norm": 0.13915176689624786,
+      "learning_rate": 0.0001,
+      "loss": 0.1791,
+      "step": 1228
+    },
+    {
+      "epoch": 2.003586566677535,
+      "grad_norm": 0.13117247819900513,
+      "learning_rate": 0.0001,
+      "loss": 0.1753,
+      "step": 1229
+    },
+    {
+      "epoch": 2.0052168242582327,
+      "grad_norm": 0.1464276909828186,
+      "learning_rate": 0.0001,
+      "loss": 0.1773,
+      "step": 1230
+    },
+    {
+      "epoch": 2.0068470818389303,
+      "grad_norm": 0.1339617669582367,
+      "learning_rate": 0.0001,
+      "loss": 0.1857,
+      "step": 1231
+    },
+    {
+      "epoch": 2.0084773394196285,
+      "grad_norm": 0.12631379067897797,
+      "learning_rate": 0.0001,
+      "loss": 0.1711,
+      "step": 1232
+    },
+    {
+      "epoch": 2.010107597000326,
+      "grad_norm": 0.1561838537454605,
+      "learning_rate": 0.0001,
+      "loss": 0.173,
+      "step": 1233
+    },
+    {
+      "epoch": 2.011737854581024,
+      "grad_norm": 0.13479989767074585,
+      "learning_rate": 0.0001,
+      "loss": 0.1661,
+      "step": 1234
+    },
+    {
+      "epoch": 2.0133681121617215,
+      "grad_norm": 0.14471793174743652,
+      "learning_rate": 0.0001,
+      "loss": 0.1712,
+      "step": 1235
+    },
+    {
+      "epoch": 2.014998369742419,
+      "grad_norm": 0.13273389637470245,
+      "learning_rate": 0.0001,
+      "loss": 0.1689,
+      "step": 1236
+    },
+    {
+      "epoch": 2.016628627323117,
+      "grad_norm": 0.15367534756660461,
+      "learning_rate": 0.0001,
+      "loss": 0.1917,
+      "step": 1237
+    },
+    {
+      "epoch": 2.018258884903815,
+      "grad_norm": 0.14998719096183777,
+      "learning_rate": 0.0001,
+      "loss": 0.1683,
+      "step": 1238
+    },
+    {
+      "epoch": 2.0198891424845127,
+      "grad_norm": 0.12296893447637558,
+      "learning_rate": 0.0001,
+      "loss": 0.1604,
+      "step": 1239
+    },
+    {
+      "epoch": 2.0215194000652104,
+      "grad_norm": 0.1817280501127243,
+      "learning_rate": 0.0001,
+      "loss": 0.1783,
+      "step": 1240
+    },
+    {
+      "epoch": 2.023149657645908,
+      "grad_norm": 0.13128723204135895,
+      "learning_rate": 0.0001,
+      "loss": 0.1664,
+      "step": 1241
+    },
+    {
+      "epoch": 2.0247799152266057,
+      "grad_norm": 0.10108043253421783,
+      "learning_rate": 0.0001,
+      "loss": 0.1612,
+      "step": 1242
+    },
+    {
+      "epoch": 2.0264101728073034,
+      "grad_norm": 0.12584680318832397,
+      "learning_rate": 0.0001,
+      "loss": 0.1769,
+      "step": 1243
+    },
+    {
+      "epoch": 2.0280404303880015,
+      "grad_norm": 0.11849281191825867,
+      "learning_rate": 0.0001,
+      "loss": 0.1684,
+      "step": 1244
+    },
+    {
+      "epoch": 2.029670687968699,
+      "grad_norm": 0.15767361223697662,
+      "learning_rate": 0.0001,
+      "loss": 0.172,
+      "step": 1245
+    },
+    {
+      "epoch": 2.031300945549397,
+      "grad_norm": 0.13649272918701172,
+      "learning_rate": 0.0001,
+      "loss": 0.1777,
+      "step": 1246
+    },
+    {
+      "epoch": 2.0329312031300946,
+      "grad_norm": 0.12619361281394958,
+      "learning_rate": 0.0001,
+      "loss": 0.1843,
+      "step": 1247
+    },
+    {
+      "epoch": 2.0345614607107922,
+      "grad_norm": 0.12181363254785538,
+      "learning_rate": 0.0001,
+      "loss": 0.1651,
+      "step": 1248
+    },
+    {
+      "epoch": 2.03619171829149,
+      "grad_norm": 0.14900757372379303,
+      "learning_rate": 0.0001,
+      "loss": 0.1777,
+      "step": 1249
+    },
+    {
+      "epoch": 2.0378219758721876,
+      "grad_norm": 0.13565577566623688,
+      "learning_rate": 0.0001,
+      "loss": 0.1797,
+      "step": 1250
+    },
+    {
+      "epoch": 2.0394522334528857,
+      "grad_norm": 0.11237785965204239,
+      "learning_rate": 0.0001,
+      "loss": 0.1577,
+      "step": 1251
+    },
+    {
+      "epoch": 2.0410824910335834,
+      "grad_norm": 0.14608171582221985,
+      "learning_rate": 0.0001,
+      "loss": 0.1792,
+      "step": 1252
+    },
+    {
+      "epoch": 2.042712748614281,
+      "grad_norm": 0.14351628720760345,
+      "learning_rate": 0.0001,
+      "loss": 0.1848,
+      "step": 1253
+    },
+    {
+      "epoch": 2.0443430061949788,
+      "grad_norm": 0.13642996549606323,
+      "learning_rate": 0.0001,
+      "loss": 0.1919,
+      "step": 1254
+    },
+    {
+      "epoch": 2.0459732637756765,
+      "grad_norm": 0.13869710266590118,
+      "learning_rate": 0.0001,
+      "loss": 0.1739,
+      "step": 1255
+    },
+    {
+      "epoch": 2.047603521356374,
+      "grad_norm": 0.12874935567378998,
+      "learning_rate": 0.0001,
+      "loss": 0.1785,
+      "step": 1256
+    },
+    {
+      "epoch": 2.0492337789370723,
+      "grad_norm": 0.12782157957553864,
+      "learning_rate": 0.0001,
+      "loss": 0.1802,
+      "step": 1257
+    },
+    {
+      "epoch": 2.05086403651777,
+      "grad_norm": 0.12892591953277588,
+      "learning_rate": 0.0001,
+      "loss": 0.1704,
+      "step": 1258
+    },
+    {
+      "epoch": 2.0524942940984676,
+      "grad_norm": 0.1089789867401123,
+      "learning_rate": 0.0001,
+      "loss": 0.1578,
+      "step": 1259
+    },
+    {
+      "epoch": 2.0541245516791653,
+      "grad_norm": 0.1354093849658966,
+      "learning_rate": 0.0001,
+      "loss": 0.1675,
+      "step": 1260
+    },
+    {
+      "epoch": 2.055754809259863,
+      "grad_norm": 0.13709183037281036,
+      "learning_rate": 0.0001,
+      "loss": 0.1735,
+      "step": 1261
+    },
+    {
+      "epoch": 2.0573850668405607,
+      "grad_norm": 0.14396348595619202,
+      "learning_rate": 0.0001,
+      "loss": 0.1815,
+      "step": 1262
+    },
+    {
+      "epoch": 2.0590153244212583,
+      "grad_norm": 0.1237788200378418,
+      "learning_rate": 0.0001,
+      "loss": 0.1565,
+      "step": 1263
+    },
+    {
+      "epoch": 2.0606455820019565,
+      "grad_norm": 0.13664290308952332,
+      "learning_rate": 0.0001,
+      "loss": 0.1786,
+      "step": 1264
+    },
+    {
+      "epoch": 2.062275839582654,
+      "grad_norm": 0.16328440606594086,
+      "learning_rate": 0.0001,
+      "loss": 0.1598,
+      "step": 1265
+    },
+    {
+      "epoch": 2.063906097163352,
+      "grad_norm": 0.1353476196527481,
+      "learning_rate": 0.0001,
+      "loss": 0.1659,
+      "step": 1266
+    },
+    {
+      "epoch": 2.0655363547440495,
+      "grad_norm": 0.16309651732444763,
+      "learning_rate": 0.0001,
+      "loss": 0.1999,
+      "step": 1267
+    },
+    {
+      "epoch": 2.067166612324747,
+      "grad_norm": 0.13841314613819122,
+      "learning_rate": 0.0001,
+      "loss": 0.1974,
+      "step": 1268
+    },
+    {
+      "epoch": 2.068796869905445,
+      "grad_norm": 0.12356504052877426,
+      "learning_rate": 0.0001,
+      "loss": 0.1695,
+      "step": 1269
+    },
+    {
+      "epoch": 2.070427127486143,
+      "grad_norm": 0.13916341960430145,
+      "learning_rate": 0.0001,
+      "loss": 0.165,
+      "step": 1270
+    },
+    {
+      "epoch": 2.0720573850668407,
+      "grad_norm": 0.13832920789718628,
+      "learning_rate": 0.0001,
+      "loss": 0.1619,
+      "step": 1271
+    },
+    {
+      "epoch": 2.0736876426475384,
+      "grad_norm": 0.14106802642345428,
+      "learning_rate": 0.0001,
+      "loss": 0.1749,
+      "step": 1272
+    },
+    {
+      "epoch": 2.075317900228236,
+      "grad_norm": 0.12422151118516922,
+      "learning_rate": 0.0001,
+      "loss": 0.1718,
+      "step": 1273
+    },
+    {
+      "epoch": 2.0769481578089337,
+      "grad_norm": 0.14482441544532776,
+      "learning_rate": 0.0001,
+      "loss": 0.1807,
+      "step": 1274
+    },
+    {
+      "epoch": 2.0785784153896314,
+      "grad_norm": 0.11608091741800308,
+      "learning_rate": 0.0001,
+      "loss": 0.1759,
+      "step": 1275
+    },
+    {
+      "epoch": 2.0802086729703295,
+      "grad_norm": 0.12500539422035217,
+      "learning_rate": 0.0001,
+      "loss": 0.1542,
+      "step": 1276
+    },
+    {
+      "epoch": 2.081838930551027,
+      "grad_norm": 0.13771778345108032,
+      "learning_rate": 0.0001,
+      "loss": 0.1899,
+      "step": 1277
+    },
+    {
+      "epoch": 2.083469188131725,
+      "grad_norm": 0.12829305231571198,
+      "learning_rate": 0.0001,
+      "loss": 0.1617,
+      "step": 1278
+    },
+    {
+      "epoch": 2.0850994457124226,
+      "grad_norm": 0.12664735317230225,
+      "learning_rate": 0.0001,
+      "loss": 0.1756,
+      "step": 1279
+    },
+    {
+      "epoch": 2.0867297032931202,
+      "grad_norm": 0.14798226952552795,
+      "learning_rate": 0.0001,
+      "loss": 0.175,
+      "step": 1280
+    },
+    {
+      "epoch": 2.088359960873818,
+      "grad_norm": 0.13553886115550995,
+      "learning_rate": 0.0001,
+      "loss": 0.1745,
+      "step": 1281
+    },
+    {
+      "epoch": 2.0899902184545156,
+      "grad_norm": 0.14772772789001465,
+      "learning_rate": 0.0001,
+      "loss": 0.1727,
+      "step": 1282
+    },
+    {
+      "epoch": 2.0916204760352137,
+      "grad_norm": 0.14489415287971497,
+      "learning_rate": 0.0001,
+      "loss": 0.1595,
+      "step": 1283
+    },
+    {
+      "epoch": 2.0932507336159114,
+      "grad_norm": 0.14262838661670685,
+      "learning_rate": 0.0001,
+      "loss": 0.1735,
+      "step": 1284
+    },
+    {
+      "epoch": 2.094880991196609,
+      "grad_norm": 0.13312998414039612,
+      "learning_rate": 0.0001,
+      "loss": 0.172,
+      "step": 1285
+    },
+    {
+      "epoch": 2.0965112487773068,
+      "grad_norm": 0.1361507624387741,
+      "learning_rate": 0.0001,
+      "loss": 0.1679,
+      "step": 1286
+    },
+    {
+      "epoch": 2.0981415063580044,
+      "grad_norm": 0.13537771999835968,
+      "learning_rate": 0.0001,
+      "loss": 0.1656,
+      "step": 1287
+    },
+    {
+      "epoch": 2.099771763938702,
+      "grad_norm": 0.16186851263046265,
+      "learning_rate": 0.0001,
+      "loss": 0.1786,
+      "step": 1288
+    },
+    {
+      "epoch": 2.1014020215194003,
+      "grad_norm": 0.1386481523513794,
+      "learning_rate": 0.0001,
+      "loss": 0.1745,
+      "step": 1289
+    },
+    {
+      "epoch": 2.103032279100098,
+      "grad_norm": 0.12809383869171143,
+      "learning_rate": 0.0001,
+      "loss": 0.1676,
+      "step": 1290
+    },
+    {
+      "epoch": 2.1046625366807956,
+      "grad_norm": 0.12493950873613358,
+      "learning_rate": 0.0001,
+      "loss": 0.1634,
+      "step": 1291
+    },
+    {
+      "epoch": 2.1062927942614933,
+      "grad_norm": 0.12215922027826309,
+      "learning_rate": 0.0001,
+      "loss": 0.1633,
+      "step": 1292
+    },
+    {
+      "epoch": 2.107923051842191,
+      "grad_norm": 0.14522279798984528,
+      "learning_rate": 0.0001,
+      "loss": 0.1785,
+      "step": 1293
+    },
+    {
+      "epoch": 2.1095533094228887,
+      "grad_norm": 0.16697941720485687,
+      "learning_rate": 0.0001,
+      "loss": 0.1714,
+      "step": 1294
+    },
+    {
+      "epoch": 2.111183567003587,
+      "grad_norm": 0.12275785952806473,
+      "learning_rate": 0.0001,
+      "loss": 0.174,
+      "step": 1295
+    },
+    {
+      "epoch": 2.1128138245842845,
+      "grad_norm": 0.12233259528875351,
+      "learning_rate": 0.0001,
+      "loss": 0.1639,
+      "step": 1296
+    },
+    {
+      "epoch": 2.114444082164982,
+      "grad_norm": 0.145602285861969,
+      "learning_rate": 0.0001,
+      "loss": 0.1725,
+      "step": 1297
+    },
+    {
+      "epoch": 2.11607433974568,
+      "grad_norm": 0.18430952727794647,
+      "learning_rate": 0.0001,
+      "loss": 0.1617,
+      "step": 1298
+    },
+    {
+      "epoch": 2.1177045973263775,
+      "grad_norm": 0.12870638072490692,
+      "learning_rate": 0.0001,
+      "loss": 0.1741,
+      "step": 1299
+    },
+    {
+      "epoch": 2.119334854907075,
+      "grad_norm": 0.1242339164018631,
+      "learning_rate": 0.0001,
+      "loss": 0.1614,
+      "step": 1300
+    },
+    {
+      "epoch": 2.120965112487773,
+      "grad_norm": 0.15219736099243164,
+      "learning_rate": 0.0001,
+      "loss": 0.1759,
+      "step": 1301
+    },
+    {
+      "epoch": 2.122595370068471,
+      "grad_norm": 0.14123129844665527,
+      "learning_rate": 0.0001,
+      "loss": 0.1637,
+      "step": 1302
+    },
+    {
+      "epoch": 2.1242256276491687,
+      "grad_norm": 0.11224009841680527,
+      "learning_rate": 0.0001,
+      "loss": 0.1586,
+      "step": 1303
+    },
+    {
+      "epoch": 2.1258558852298663,
+      "grad_norm": 0.1388261467218399,
+      "learning_rate": 0.0001,
+      "loss": 0.1607,
+      "step": 1304
+    },
+    {
+      "epoch": 2.127486142810564,
+      "grad_norm": 0.13148367404937744,
+      "learning_rate": 0.0001,
+      "loss": 0.1894,
+      "step": 1305
+    },
+    {
+      "epoch": 2.1291164003912617,
+      "grad_norm": 0.1266559213399887,
+      "learning_rate": 0.0001,
+      "loss": 0.1582,
+      "step": 1306
+    },
+    {
+      "epoch": 2.1307466579719594,
+      "grad_norm": 0.15286579728126526,
+      "learning_rate": 0.0001,
+      "loss": 0.1852,
+      "step": 1307
+    },
+    {
+      "epoch": 2.1323769155526575,
+      "grad_norm": 0.1486169546842575,
+      "learning_rate": 0.0001,
+      "loss": 0.1673,
+      "step": 1308
+    },
+    {
+      "epoch": 2.134007173133355,
+      "grad_norm": 0.1324825882911682,
+      "learning_rate": 0.0001,
+      "loss": 0.1745,
+      "step": 1309
+    },
+    {
+      "epoch": 2.135637430714053,
+      "grad_norm": 0.12974712252616882,
+      "learning_rate": 0.0001,
+      "loss": 0.1786,
+      "step": 1310
+    },
+    {
+      "epoch": 2.1372676882947506,
+      "grad_norm": 0.17382071912288666,
+      "learning_rate": 0.0001,
+      "loss": 0.1659,
+      "step": 1311
+    },
+    {
+      "epoch": 2.1388979458754482,
+      "grad_norm": 0.14870017766952515,
+      "learning_rate": 0.0001,
+      "loss": 0.1827,
+      "step": 1312
+    },
+    {
+      "epoch": 2.140528203456146,
+      "grad_norm": 0.1424129158258438,
+      "learning_rate": 0.0001,
+      "loss": 0.1813,
+      "step": 1313
+    },
+    {
+      "epoch": 2.1421584610368436,
+      "grad_norm": 0.15458005666732788,
+      "learning_rate": 0.0001,
+      "loss": 0.1705,
+      "step": 1314
+    },
+    {
+      "epoch": 2.1437887186175417,
+      "grad_norm": 0.1291740983724594,
+      "learning_rate": 0.0001,
+      "loss": 0.1563,
+      "step": 1315
+    },
+    {
+      "epoch": 2.1454189761982394,
+      "grad_norm": 0.14298751950263977,
+      "learning_rate": 0.0001,
+      "loss": 0.176,
+      "step": 1316
+    },
+    {
+      "epoch": 2.147049233778937,
+      "grad_norm": 0.1237107664346695,
+      "learning_rate": 0.0001,
+      "loss": 0.1636,
+      "step": 1317
+    },
+    {
+      "epoch": 2.1486794913596348,
+      "grad_norm": 0.11446285992860794,
+      "learning_rate": 0.0001,
+      "loss": 0.1658,
+      "step": 1318
+    },
+    {
+      "epoch": 2.1503097489403324,
+      "grad_norm": 0.11239829659461975,
+      "learning_rate": 0.0001,
+      "loss": 0.1606,
+      "step": 1319
+    },
+    {
+      "epoch": 2.15194000652103,
+      "grad_norm": 0.12736567854881287,
+      "learning_rate": 0.0001,
+      "loss": 0.1778,
+      "step": 1320
+    },
+    {
+      "epoch": 2.1535702641017282,
+      "grad_norm": 0.13921810686588287,
+      "learning_rate": 0.0001,
+      "loss": 0.1882,
+      "step": 1321
+    },
+    {
+      "epoch": 2.155200521682426,
+      "grad_norm": 0.12388879060745239,
+      "learning_rate": 0.0001,
+      "loss": 0.1824,
+      "step": 1322
+    },
+    {
+      "epoch": 2.1568307792631236,
+      "grad_norm": 0.1522189825773239,
+      "learning_rate": 0.0001,
+      "loss": 0.1556,
+      "step": 1323
+    },
+    {
+      "epoch": 2.1584610368438213,
+      "grad_norm": 0.13830043375492096,
+      "learning_rate": 0.0001,
+      "loss": 0.1706,
+      "step": 1324
+    },
+    {
+      "epoch": 2.160091294424519,
+      "grad_norm": 0.14144088327884674,
+      "learning_rate": 0.0001,
+      "loss": 0.1699,
+      "step": 1325
+    },
+    {
+      "epoch": 2.1617215520052167,
+      "grad_norm": 0.14037571847438812,
+      "learning_rate": 0.0001,
+      "loss": 0.1673,
+      "step": 1326
+    },
+    {
+      "epoch": 2.1633518095859148,
+      "grad_norm": 0.16178935766220093,
+      "learning_rate": 0.0001,
+      "loss": 0.1881,
+      "step": 1327
+    },
+    {
+      "epoch": 2.1649820671666125,
+      "grad_norm": 0.12526622414588928,
+      "learning_rate": 0.0001,
+      "loss": 0.1675,
+      "step": 1328
+    },
+    {
+      "epoch": 2.16661232474731,
+      "grad_norm": 0.1448407918214798,
+      "learning_rate": 0.0001,
+      "loss": 0.1731,
+      "step": 1329
+    },
+    {
+      "epoch": 2.168242582328008,
+      "grad_norm": 0.15621818602085114,
+      "learning_rate": 0.0001,
+      "loss": 0.1764,
+      "step": 1330
+    },
+    {
+      "epoch": 2.1698728399087055,
+      "grad_norm": 0.14016954600811005,
+      "learning_rate": 0.0001,
+      "loss": 0.1752,
+      "step": 1331
+    },
+    {
+      "epoch": 2.171503097489403,
+      "grad_norm": 0.15519127249717712,
+      "learning_rate": 0.0001,
+      "loss": 0.1701,
+      "step": 1332
+    },
+    {
+      "epoch": 2.1731333550701013,
+      "grad_norm": 0.12022172659635544,
+      "learning_rate": 0.0001,
+      "loss": 0.1639,
+      "step": 1333
+    },
+    {
+      "epoch": 2.174763612650799,
+      "grad_norm": 0.1351272314786911,
+      "learning_rate": 0.0001,
+      "loss": 0.1653,
+      "step": 1334
+    },
+    {
+      "epoch": 2.1763938702314967,
+      "grad_norm": 0.13195279240608215,
+      "learning_rate": 0.0001,
+      "loss": 0.1735,
+      "step": 1335
+    },
+    {
+      "epoch": 2.1780241278121943,
+      "grad_norm": 0.14435507357120514,
+      "learning_rate": 0.0001,
+      "loss": 0.1801,
+      "step": 1336
+    },
+    {
+      "epoch": 2.179654385392892,
+      "grad_norm": 0.13603608310222626,
+      "learning_rate": 0.0001,
+      "loss": 0.1643,
+      "step": 1337
+    },
+    {
+      "epoch": 2.1812846429735897,
+      "grad_norm": 0.14588388800621033,
+      "learning_rate": 0.0001,
+      "loss": 0.1711,
+      "step": 1338
+    },
+    {
+      "epoch": 2.1829149005542874,
+      "grad_norm": 0.13602250814437866,
+      "learning_rate": 0.0001,
+      "loss": 0.1854,
+      "step": 1339
+    },
+    {
+      "epoch": 2.1845451581349855,
+      "grad_norm": 0.13646137714385986,
+      "learning_rate": 0.0001,
+      "loss": 0.1881,
+      "step": 1340
+    },
+    {
+      "epoch": 2.186175415715683,
+      "grad_norm": 0.13280268013477325,
+      "learning_rate": 0.0001,
+      "loss": 0.1708,
+      "step": 1341
+    },
+    {
+      "epoch": 2.187805673296381,
+      "grad_norm": 0.13370732963085175,
+      "learning_rate": 0.0001,
+      "loss": 0.1667,
+      "step": 1342
+    },
+    {
+      "epoch": 2.1894359308770786,
+      "grad_norm": 0.14326944947242737,
+      "learning_rate": 0.0001,
+      "loss": 0.1686,
+      "step": 1343
+    },
+    {
+      "epoch": 2.1910661884577762,
+      "grad_norm": 0.12916752696037292,
+      "learning_rate": 0.0001,
+      "loss": 0.1779,
+      "step": 1344
+    },
+    {
+      "epoch": 2.192696446038474,
+      "grad_norm": 0.14404352009296417,
+      "learning_rate": 0.0001,
+      "loss": 0.171,
+      "step": 1345
+    },
+    {
+      "epoch": 2.194326703619172,
+      "grad_norm": 0.12311873584985733,
+      "learning_rate": 0.0001,
+      "loss": 0.1735,
+      "step": 1346
+    },
+    {
+      "epoch": 2.1959569611998697,
+      "grad_norm": 0.15599961578845978,
+      "learning_rate": 0.0001,
+      "loss": 0.1748,
+      "step": 1347
+    },
+    {
+      "epoch": 2.1975872187805674,
+      "grad_norm": 0.12893660366535187,
+      "learning_rate": 0.0001,
+      "loss": 0.1829,
+      "step": 1348
+    },
+    {
+      "epoch": 2.199217476361265,
+      "grad_norm": 0.12859125435352325,
+      "learning_rate": 0.0001,
+      "loss": 0.1797,
+      "step": 1349
+    },
+    {
+      "epoch": 2.2008477339419628,
+      "grad_norm": 0.1159624233841896,
+      "learning_rate": 0.0001,
+      "loss": 0.1658,
+      "step": 1350
+    },
+    {
+      "epoch": 2.2024779915226604,
+      "grad_norm": 0.13532379269599915,
+      "learning_rate": 0.0001,
+      "loss": 0.1973,
+      "step": 1351
+    },
+    {
+      "epoch": 2.204108249103358,
+      "grad_norm": 0.142998605966568,
+      "learning_rate": 0.0001,
+      "loss": 0.1763,
+      "step": 1352
+    },
+    {
+      "epoch": 2.2057385066840562,
+      "grad_norm": 0.15357817709445953,
+      "learning_rate": 0.0001,
+      "loss": 0.1733,
+      "step": 1353
+    },
+    {
+      "epoch": 2.207368764264754,
+      "grad_norm": 0.133555606007576,
+      "learning_rate": 0.0001,
+      "loss": 0.177,
+      "step": 1354
+    },
+    {
+      "epoch": 2.2089990218454516,
+      "grad_norm": 0.12514221668243408,
+      "learning_rate": 0.0001,
+      "loss": 0.1706,
+      "step": 1355
+    },
+    {
+      "epoch": 2.2106292794261493,
+      "grad_norm": 0.12143199145793915,
+      "learning_rate": 0.0001,
+      "loss": 0.1728,
+      "step": 1356
+    },
+    {
+      "epoch": 2.212259537006847,
+      "grad_norm": 0.12493051588535309,
+      "learning_rate": 0.0001,
+      "loss": 0.1735,
+      "step": 1357
+    },
+    {
+      "epoch": 2.2138897945875446,
+      "grad_norm": 0.12735562026500702,
+      "learning_rate": 0.0001,
+      "loss": 0.1692,
+      "step": 1358
+    },
+    {
+      "epoch": 2.2155200521682428,
+      "grad_norm": 0.15920554101467133,
+      "learning_rate": 0.0001,
+      "loss": 0.1946,
+      "step": 1359
+    },
+    {
+      "epoch": 2.2171503097489405,
+      "grad_norm": 0.12582500278949738,
+      "learning_rate": 0.0001,
+      "loss": 0.1585,
+      "step": 1360
+    },
+    {
+      "epoch": 2.218780567329638,
+      "grad_norm": 0.16344155371189117,
+      "learning_rate": 0.0001,
+      "loss": 0.1666,
+      "step": 1361
+    },
+    {
+      "epoch": 2.220410824910336,
+      "grad_norm": 0.12184584140777588,
+      "learning_rate": 0.0001,
+      "loss": 0.1804,
+      "step": 1362
+    },
+    {
+      "epoch": 2.2220410824910335,
+      "grad_norm": 0.16029442846775055,
+      "learning_rate": 0.0001,
+      "loss": 0.1911,
+      "step": 1363
+    },
+    {
+      "epoch": 2.223671340071731,
+      "grad_norm": 0.12164244800806046,
+      "learning_rate": 0.0001,
+      "loss": 0.1907,
+      "step": 1364
+    },
+    {
+      "epoch": 2.225301597652429,
+      "grad_norm": 0.15586431324481964,
+      "learning_rate": 0.0001,
+      "loss": 0.1762,
+      "step": 1365
+    },
+    {
+      "epoch": 2.226931855233127,
+      "grad_norm": 0.1489764153957367,
+      "learning_rate": 0.0001,
+      "loss": 0.1857,
+      "step": 1366
+    },
+    {
+      "epoch": 2.2285621128138247,
+      "grad_norm": 0.153330996632576,
+      "learning_rate": 0.0001,
+      "loss": 0.1756,
+      "step": 1367
+    },
+    {
+      "epoch": 2.2301923703945223,
+      "grad_norm": 0.11781416088342667,
+      "learning_rate": 0.0001,
+      "loss": 0.1556,
+      "step": 1368
+    },
+    {
+      "epoch": 2.23182262797522,
+      "grad_norm": 0.1183483824133873,
+      "learning_rate": 0.0001,
+      "loss": 0.1564,
+      "step": 1369
+    },
+    {
+      "epoch": 2.2334528855559177,
+      "grad_norm": 0.12061773240566254,
+      "learning_rate": 0.0001,
+      "loss": 0.1717,
+      "step": 1370
+    },
+    {
+      "epoch": 2.2350831431366154,
+      "grad_norm": 0.151438370347023,
+      "learning_rate": 0.0001,
+      "loss": 0.1794,
+      "step": 1371
+    },
+    {
+      "epoch": 2.2367134007173135,
+      "grad_norm": 0.12716467678546906,
+      "learning_rate": 0.0001,
+      "loss": 0.1778,
+      "step": 1372
+    },
+    {
+      "epoch": 2.238343658298011,
+      "grad_norm": 0.13353648781776428,
+      "learning_rate": 0.0001,
+      "loss": 0.181,
+      "step": 1373
+    },
+    {
+      "epoch": 2.239973915878709,
+      "grad_norm": 0.12627844512462616,
+      "learning_rate": 0.0001,
+      "loss": 0.1714,
+      "step": 1374
+    },
+    {
+      "epoch": 2.2416041734594065,
+      "grad_norm": 0.14614202082157135,
+      "learning_rate": 0.0001,
+      "loss": 0.1751,
+      "step": 1375
+    },
+    {
+      "epoch": 2.2432344310401042,
+      "grad_norm": 0.12322308868169785,
+      "learning_rate": 0.0001,
+      "loss": 0.1688,
+      "step": 1376
+    },
+    {
+      "epoch": 2.244864688620802,
+      "grad_norm": 0.1395500749349594,
+      "learning_rate": 0.0001,
+      "loss": 0.1777,
+      "step": 1377
+    },
+    {
+      "epoch": 2.2464949462015,
+      "grad_norm": 0.12420187145471573,
+      "learning_rate": 0.0001,
+      "loss": 0.1617,
+      "step": 1378
+    },
+    {
+      "epoch": 2.2481252037821977,
+      "grad_norm": 0.1447030007839203,
+      "learning_rate": 0.0001,
+      "loss": 0.1794,
+      "step": 1379
+    },
+    {
+      "epoch": 2.2497554613628954,
+      "grad_norm": 0.1566443145275116,
+      "learning_rate": 0.0001,
+      "loss": 0.1863,
+      "step": 1380
+    },
+    {
+      "epoch": 2.251385718943593,
+      "grad_norm": 0.1401456892490387,
+      "learning_rate": 0.0001,
+      "loss": 0.1822,
+      "step": 1381
+    },
+    {
+      "epoch": 2.2530159765242908,
+      "grad_norm": 0.13101978600025177,
+      "learning_rate": 0.0001,
+      "loss": 0.1672,
+      "step": 1382
+    },
+    {
+      "epoch": 2.2546462341049884,
+      "grad_norm": 0.12039732187986374,
+      "learning_rate": 0.0001,
+      "loss": 0.1679,
+      "step": 1383
+    },
+    {
+      "epoch": 2.2562764916856866,
+      "grad_norm": 0.12393520027399063,
+      "learning_rate": 0.0001,
+      "loss": 0.1728,
+      "step": 1384
+    },
+    {
+      "epoch": 2.2579067492663842,
+      "grad_norm": 0.16083435714244843,
+      "learning_rate": 0.0001,
+      "loss": 0.1709,
+      "step": 1385
+    },
+    {
+      "epoch": 2.259537006847082,
+      "grad_norm": 0.13336148858070374,
+      "learning_rate": 0.0001,
+      "loss": 0.1675,
+      "step": 1386
+    },
+    {
+      "epoch": 2.2611672644277796,
+      "grad_norm": 0.14768260717391968,
+      "learning_rate": 0.0001,
+      "loss": 0.1747,
+      "step": 1387
+    },
+    {
+      "epoch": 2.2627975220084773,
+      "grad_norm": 0.13280175626277924,
+      "learning_rate": 0.0001,
+      "loss": 0.1616,
+      "step": 1388
+    },
+    {
+      "epoch": 2.264427779589175,
+      "grad_norm": 0.12707430124282837,
+      "learning_rate": 0.0001,
+      "loss": 0.1687,
+      "step": 1389
+    },
+    {
+      "epoch": 2.2660580371698726,
+      "grad_norm": 0.14237570762634277,
+      "learning_rate": 0.0001,
+      "loss": 0.1598,
+      "step": 1390
+    },
+    {
+      "epoch": 2.2676882947505708,
+      "grad_norm": 0.1464899331331253,
+      "learning_rate": 0.0001,
+      "loss": 0.1795,
+      "step": 1391
+    },
+    {
+      "epoch": 2.2693185523312684,
+      "grad_norm": 0.15607094764709473,
+      "learning_rate": 0.0001,
+      "loss": 0.1775,
+      "step": 1392
+    },
+    {
+      "epoch": 2.270948809911966,
+      "grad_norm": 0.1322450190782547,
+      "learning_rate": 0.0001,
+      "loss": 0.1729,
+      "step": 1393
+    },
+    {
+      "epoch": 2.272579067492664,
+      "grad_norm": 0.17985309660434723,
+      "learning_rate": 0.0001,
+      "loss": 0.1649,
+      "step": 1394
+    },
+    {
+      "epoch": 2.2742093250733615,
+      "grad_norm": 0.1275913417339325,
+      "learning_rate": 0.0001,
+      "loss": 0.184,
+      "step": 1395
+    },
+    {
+      "epoch": 2.275839582654059,
+      "grad_norm": 0.1359567791223526,
+      "learning_rate": 0.0001,
+      "loss": 0.1807,
+      "step": 1396
+    },
+    {
+      "epoch": 2.2774698402347573,
+      "grad_norm": 0.13110549747943878,
+      "learning_rate": 0.0001,
+      "loss": 0.1736,
+      "step": 1397
+    },
+    {
+      "epoch": 2.279100097815455,
+      "grad_norm": 0.13743306696414948,
+      "learning_rate": 0.0001,
+      "loss": 0.1657,
+      "step": 1398
+    },
+    {
+      "epoch": 2.2807303553961527,
+      "grad_norm": 0.15267515182495117,
+      "learning_rate": 0.0001,
+      "loss": 0.1674,
+      "step": 1399
+    },
+    {
+      "epoch": 2.2823606129768503,
+      "grad_norm": 0.12119441479444504,
+      "learning_rate": 0.0001,
+      "loss": 0.1627,
+      "step": 1400
+    },
+    {
+      "epoch": 2.283990870557548,
+      "grad_norm": 0.1421733945608139,
+      "learning_rate": 0.0001,
+      "loss": 0.161,
+      "step": 1401
+    },
+    {
+      "epoch": 2.2856211281382457,
+      "grad_norm": 0.1504882127046585,
+      "learning_rate": 0.0001,
+      "loss": 0.1869,
+      "step": 1402
+    },
+    {
+      "epoch": 2.2872513857189434,
+      "grad_norm": 0.15724492073059082,
+      "learning_rate": 0.0001,
+      "loss": 0.1826,
+      "step": 1403
+    },
+    {
+      "epoch": 2.2888816432996415,
+      "grad_norm": 0.1295897364616394,
+      "learning_rate": 0.0001,
+      "loss": 0.1697,
+      "step": 1404
+    },
+    {
+      "epoch": 2.290511900880339,
+      "grad_norm": 0.13183419406414032,
+      "learning_rate": 0.0001,
+      "loss": 0.1675,
+      "step": 1405
+    },
+    {
+      "epoch": 2.292142158461037,
+      "grad_norm": 0.14537960290908813,
+      "learning_rate": 0.0001,
+      "loss": 0.1912,
+      "step": 1406
+    },
+    {
+      "epoch": 2.2937724160417345,
+      "grad_norm": 0.1461685299873352,
+      "learning_rate": 0.0001,
+      "loss": 0.1815,
+      "step": 1407
+    },
+    {
+      "epoch": 2.295402673622432,
+      "grad_norm": 0.13184034824371338,
+      "learning_rate": 0.0001,
+      "loss": 0.1764,
+      "step": 1408
+    },
+    {
+      "epoch": 2.29703293120313,
+      "grad_norm": 0.15598301589488983,
+      "learning_rate": 0.0001,
+      "loss": 0.1688,
+      "step": 1409
+    },
+    {
+      "epoch": 2.298663188783828,
+      "grad_norm": 0.17276427149772644,
+      "learning_rate": 0.0001,
+      "loss": 0.1794,
+      "step": 1410
+    },
+    {
+      "epoch": 2.3002934463645257,
+      "grad_norm": 0.12554508447647095,
+      "learning_rate": 0.0001,
+      "loss": 0.1788,
+      "step": 1411
+    },
+    {
+      "epoch": 2.3019237039452234,
+      "grad_norm": 0.12981827557086945,
+      "learning_rate": 0.0001,
+      "loss": 0.1658,
+      "step": 1412
+    },
+    {
+      "epoch": 2.303553961525921,
+      "grad_norm": 0.10513079911470413,
+      "learning_rate": 0.0001,
+      "loss": 0.1587,
+      "step": 1413
+    },
+    {
+      "epoch": 2.3051842191066187,
+      "grad_norm": 0.12070135772228241,
+      "learning_rate": 0.0001,
+      "loss": 0.1652,
+      "step": 1414
+    },
+    {
+      "epoch": 2.3068144766873164,
+      "grad_norm": 0.11308653652667999,
+      "learning_rate": 0.0001,
+      "loss": 0.1609,
+      "step": 1415
+    },
+    {
+      "epoch": 2.308444734268014,
+      "grad_norm": 0.1290631741285324,
+      "learning_rate": 0.0001,
+      "loss": 0.179,
+      "step": 1416
+    },
+    {
+      "epoch": 2.3100749918487122,
+      "grad_norm": 0.15130403637886047,
+      "learning_rate": 0.0001,
+      "loss": 0.192,
+      "step": 1417
+    },
+    {
+      "epoch": 2.31170524942941,
+      "grad_norm": 0.12521226704120636,
+      "learning_rate": 0.0001,
+      "loss": 0.1743,
+      "step": 1418
+    },
+    {
+      "epoch": 2.3133355070101076,
+      "grad_norm": 0.15238836407661438,
+      "learning_rate": 0.0001,
+      "loss": 0.1806,
+      "step": 1419
+    },
+    {
+      "epoch": 2.3149657645908053,
+      "grad_norm": 0.15979474782943726,
+      "learning_rate": 0.0001,
+      "loss": 0.1987,
+      "step": 1420
+    },
+    {
+      "epoch": 2.316596022171503,
+      "grad_norm": 0.1352485716342926,
+      "learning_rate": 0.0001,
+      "loss": 0.1868,
+      "step": 1421
+    },
+    {
+      "epoch": 2.318226279752201,
+      "grad_norm": 0.12670114636421204,
+      "learning_rate": 0.0001,
+      "loss": 0.1807,
+      "step": 1422
+    },
+    {
+      "epoch": 2.3198565373328988,
+      "grad_norm": 0.13426585495471954,
+      "learning_rate": 0.0001,
+      "loss": 0.1739,
+      "step": 1423
+    },
+    {
+      "epoch": 2.3214867949135964,
+      "grad_norm": 0.1362699568271637,
+      "learning_rate": 0.0001,
+      "loss": 0.1839,
+      "step": 1424
+    },
+    {
+      "epoch": 2.323117052494294,
+      "grad_norm": 0.12018754333257675,
+      "learning_rate": 0.0001,
+      "loss": 0.1607,
+      "step": 1425
+    },
+    {
+      "epoch": 2.324747310074992,
+      "grad_norm": 0.13709640502929688,
+      "learning_rate": 0.0001,
+      "loss": 0.1784,
+      "step": 1426
+    },
+    {
+      "epoch": 2.3263775676556895,
+      "grad_norm": 0.12608668208122253,
+      "learning_rate": 0.0001,
+      "loss": 0.1727,
+      "step": 1427
+    },
+    {
+      "epoch": 2.328007825236387,
+      "grad_norm": 0.13371500372886658,
+      "learning_rate": 0.0001,
+      "loss": 0.1669,
+      "step": 1428
+    },
+    {
+      "epoch": 2.3296380828170853,
+      "grad_norm": 0.13639819622039795,
+      "learning_rate": 0.0001,
+      "loss": 0.1755,
+      "step": 1429
+    },
+    {
+      "epoch": 2.331268340397783,
+      "grad_norm": 0.13079668581485748,
+      "learning_rate": 0.0001,
+      "loss": 0.1843,
+      "step": 1430
+    },
+    {
+      "epoch": 2.3328985979784806,
+      "grad_norm": 0.11506593972444534,
+      "learning_rate": 0.0001,
+      "loss": 0.1596,
+      "step": 1431
+    },
+    {
+      "epoch": 2.3345288555591783,
+      "grad_norm": 0.15789979696273804,
+      "learning_rate": 0.0001,
+      "loss": 0.1814,
+      "step": 1432
+    },
+    {
+      "epoch": 2.336159113139876,
+      "grad_norm": 0.1199864000082016,
+      "learning_rate": 0.0001,
+      "loss": 0.1635,
+      "step": 1433
+    },
+    {
+      "epoch": 2.3377893707205737,
+      "grad_norm": 0.14094781875610352,
+      "learning_rate": 0.0001,
+      "loss": 0.184,
+      "step": 1434
+    },
+    {
+      "epoch": 2.339419628301272,
+      "grad_norm": 0.13015292584896088,
+      "learning_rate": 0.0001,
+      "loss": 0.1643,
+      "step": 1435
+    },
+    {
+      "epoch": 2.3410498858819695,
+      "grad_norm": 0.13332228362560272,
+      "learning_rate": 0.0001,
+      "loss": 0.1596,
+      "step": 1436
+    },
+    {
+      "epoch": 2.342680143462667,
+      "grad_norm": 0.11775743216276169,
+      "learning_rate": 0.0001,
+      "loss": 0.1783,
+      "step": 1437
+    },
+    {
+      "epoch": 2.344310401043365,
+      "grad_norm": 0.15774385631084442,
+      "learning_rate": 0.0001,
+      "loss": 0.188,
+      "step": 1438
+    },
+    {
+      "epoch": 2.3459406586240625,
+      "grad_norm": 0.12837542593479156,
+      "learning_rate": 0.0001,
+      "loss": 0.1666,
+      "step": 1439
+    },
+    {
+      "epoch": 2.34757091620476,
+      "grad_norm": 0.14266343414783478,
+      "learning_rate": 0.0001,
+      "loss": 0.173,
+      "step": 1440
+    },
+    {
+      "epoch": 2.349201173785458,
+      "grad_norm": 0.14660033583641052,
+      "learning_rate": 0.0001,
+      "loss": 0.1724,
+      "step": 1441
+    },
+    {
+      "epoch": 2.350831431366156,
+      "grad_norm": 0.12945188581943512,
+      "learning_rate": 0.0001,
+      "loss": 0.1567,
+      "step": 1442
+    },
+    {
+      "epoch": 2.3524616889468537,
+      "grad_norm": 0.13205265998840332,
+      "learning_rate": 0.0001,
+      "loss": 0.1634,
+      "step": 1443
+    },
+    {
+      "epoch": 2.3540919465275514,
+      "grad_norm": 0.11146970093250275,
+      "learning_rate": 0.0001,
+      "loss": 0.1698,
+      "step": 1444
+    },
+    {
+      "epoch": 2.355722204108249,
+      "grad_norm": 0.1483961045742035,
+      "learning_rate": 0.0001,
+      "loss": 0.1705,
+      "step": 1445
+    },
+    {
+      "epoch": 2.3573524616889467,
+      "grad_norm": 0.1443917602300644,
+      "learning_rate": 0.0001,
+      "loss": 0.1881,
+      "step": 1446
+    },
+    {
+      "epoch": 2.3589827192696444,
+      "grad_norm": 0.11366801708936691,
+      "learning_rate": 0.0001,
+      "loss": 0.1673,
+      "step": 1447
+    },
+    {
+      "epoch": 2.3606129768503425,
+      "grad_norm": 0.12853001058101654,
+      "learning_rate": 0.0001,
+      "loss": 0.1682,
+      "step": 1448
+    },
+    {
+      "epoch": 2.3622432344310402,
+      "grad_norm": 0.11675182729959488,
+      "learning_rate": 0.0001,
+      "loss": 0.1607,
+      "step": 1449
+    },
+    {
+      "epoch": 2.363873492011738,
+      "grad_norm": 0.13047528266906738,
+      "learning_rate": 0.0001,
+      "loss": 0.1585,
+      "step": 1450
+    },
+    {
+      "epoch": 2.3655037495924356,
+      "grad_norm": 0.13428814709186554,
+      "learning_rate": 0.0001,
+      "loss": 0.1742,
+      "step": 1451
+    },
+    {
+      "epoch": 2.3671340071731333,
+      "grad_norm": 0.14489330351352692,
+      "learning_rate": 0.0001,
+      "loss": 0.1624,
+      "step": 1452
+    },
+    {
+      "epoch": 2.368764264753831,
+      "grad_norm": 0.126490980386734,
+      "learning_rate": 0.0001,
+      "loss": 0.1825,
+      "step": 1453
+    },
+    {
+      "epoch": 2.3703945223345286,
+      "grad_norm": 0.13337522745132446,
+      "learning_rate": 0.0001,
+      "loss": 0.1653,
+      "step": 1454
+    },
+    {
+      "epoch": 2.3720247799152268,
+      "grad_norm": 0.13690900802612305,
+      "learning_rate": 0.0001,
+      "loss": 0.1849,
+      "step": 1455
+    },
+    {
+      "epoch": 2.3736550374959244,
+      "grad_norm": 0.12736913561820984,
+      "learning_rate": 0.0001,
+      "loss": 0.175,
+      "step": 1456
+    },
+    {
+      "epoch": 2.375285295076622,
+      "grad_norm": 0.14143235981464386,
+      "learning_rate": 0.0001,
+      "loss": 0.181,
+      "step": 1457
+    },
+    {
+      "epoch": 2.37691555265732,
+      "grad_norm": 0.12882289290428162,
+      "learning_rate": 0.0001,
+      "loss": 0.1672,
+      "step": 1458
+    },
+    {
+      "epoch": 2.3785458102380175,
+      "grad_norm": 0.12184431403875351,
+      "learning_rate": 0.0001,
+      "loss": 0.1691,
+      "step": 1459
+    },
+    {
+      "epoch": 2.3801760678187156,
+      "grad_norm": 0.12170469015836716,
+      "learning_rate": 0.0001,
+      "loss": 0.1721,
+      "step": 1460
+    },
+    {
+      "epoch": 2.3818063253994133,
+      "grad_norm": 0.13725103437900543,
+      "learning_rate": 0.0001,
+      "loss": 0.1846,
+      "step": 1461
+    },
+    {
+      "epoch": 2.383436582980111,
+      "grad_norm": 0.137665793299675,
+      "learning_rate": 0.0001,
+      "loss": 0.1675,
+      "step": 1462
+    },
+    {
+      "epoch": 2.3850668405608086,
+      "grad_norm": 0.11704001575708389,
+      "learning_rate": 0.0001,
+      "loss": 0.1695,
+      "step": 1463
+    },
+    {
+      "epoch": 2.3866970981415063,
+      "grad_norm": 0.11113730818033218,
+      "learning_rate": 0.0001,
+      "loss": 0.1621,
+      "step": 1464
+    },
+    {
+      "epoch": 2.388327355722204,
+      "grad_norm": 0.1406833976507187,
+      "learning_rate": 0.0001,
+      "loss": 0.179,
+      "step": 1465
+    },
+    {
+      "epoch": 2.3899576133029017,
+      "grad_norm": 0.12707440555095673,
+      "learning_rate": 0.0001,
+      "loss": 0.1697,
+      "step": 1466
+    },
+    {
+      "epoch": 2.3915878708835994,
+      "grad_norm": 0.14010672271251678,
+      "learning_rate": 0.0001,
+      "loss": 0.1871,
+      "step": 1467
+    },
+    {
+      "epoch": 2.3932181284642975,
+      "grad_norm": 0.12332647293806076,
+      "learning_rate": 0.0001,
+      "loss": 0.1621,
+      "step": 1468
+    },
+    {
+      "epoch": 2.394848386044995,
+      "grad_norm": 0.1202240139245987,
+      "learning_rate": 0.0001,
+      "loss": 0.1718,
+      "step": 1469
+    },
+    {
+      "epoch": 2.396478643625693,
+      "grad_norm": 0.11759907752275467,
+      "learning_rate": 0.0001,
+      "loss": 0.1584,
+      "step": 1470
+    },
+    {
+      "epoch": 2.3981089012063905,
+      "grad_norm": 0.12136384099721909,
+      "learning_rate": 0.0001,
+      "loss": 0.1681,
+      "step": 1471
+    },
+    {
+      "epoch": 2.399739158787088,
+      "grad_norm": 0.1522362232208252,
+      "learning_rate": 0.0001,
+      "loss": 0.1706,
+      "step": 1472
+    },
+    {
+      "epoch": 2.4013694163677863,
+      "grad_norm": 0.13118572533130646,
+      "learning_rate": 0.0001,
+      "loss": 0.1694,
+      "step": 1473
+    },
+    {
+      "epoch": 2.402999673948484,
+      "grad_norm": 0.11596754938364029,
+      "learning_rate": 0.0001,
+      "loss": 0.1693,
+      "step": 1474
+    },
+    {
+      "epoch": 2.4046299315291817,
+      "grad_norm": 0.13307926058769226,
+      "learning_rate": 0.0001,
+      "loss": 0.1749,
+      "step": 1475
+    },
+    {
+      "epoch": 2.4062601891098794,
+      "grad_norm": 0.13675716519355774,
+      "learning_rate": 0.0001,
+      "loss": 0.1717,
+      "step": 1476
+    },
+    {
+      "epoch": 2.407890446690577,
+      "grad_norm": 0.17052991688251495,
+      "learning_rate": 0.0001,
+      "loss": 0.1788,
+      "step": 1477
+    },
+    {
+      "epoch": 2.4095207042712747,
+      "grad_norm": 0.13308259844779968,
+      "learning_rate": 0.0001,
+      "loss": 0.165,
+      "step": 1478
+    },
+    {
+      "epoch": 2.4111509618519724,
+      "grad_norm": 0.13663795590400696,
+      "learning_rate": 0.0001,
+      "loss": 0.177,
+      "step": 1479
+    },
+    {
+      "epoch": 2.4127812194326705,
+      "grad_norm": 0.1282254457473755,
+      "learning_rate": 0.0001,
+      "loss": 0.1809,
+      "step": 1480
+    },
+    {
+      "epoch": 2.4144114770133682,
+      "grad_norm": 0.14260698854923248,
+      "learning_rate": 0.0001,
+      "loss": 0.1635,
+      "step": 1481
+    },
+    {
+      "epoch": 2.416041734594066,
+      "grad_norm": 0.12570828199386597,
+      "learning_rate": 0.0001,
+      "loss": 0.1719,
+      "step": 1482
+    },
+    {
+      "epoch": 2.4176719921747636,
+      "grad_norm": 0.13560189306735992,
+      "learning_rate": 0.0001,
+      "loss": 0.1699,
+      "step": 1483
+    },
+    {
+      "epoch": 2.4193022497554613,
+      "grad_norm": 0.15842407941818237,
+      "learning_rate": 0.0001,
+      "loss": 0.1759,
+      "step": 1484
+    },
+    {
+      "epoch": 2.420932507336159,
+      "grad_norm": 0.13507573306560516,
+      "learning_rate": 0.0001,
+      "loss": 0.1751,
+      "step": 1485
+    },
+    {
+      "epoch": 2.422562764916857,
+      "grad_norm": 0.13392452895641327,
+      "learning_rate": 0.0001,
+      "loss": 0.1855,
+      "step": 1486
+    },
+    {
+      "epoch": 2.4241930224975548,
+      "grad_norm": 0.14085431396961212,
+      "learning_rate": 0.0001,
+      "loss": 0.1565,
+      "step": 1487
+    },
+    {
+      "epoch": 2.4258232800782524,
+      "grad_norm": 0.11966408044099808,
+      "learning_rate": 0.0001,
+      "loss": 0.1696,
+      "step": 1488
+    },
+    {
+      "epoch": 2.42745353765895,
+      "grad_norm": 0.14620652794837952,
+      "learning_rate": 0.0001,
+      "loss": 0.179,
+      "step": 1489
+    },
+    {
+      "epoch": 2.429083795239648,
+      "grad_norm": 0.15509895980358124,
+      "learning_rate": 0.0001,
+      "loss": 0.1641,
+      "step": 1490
+    },
+    {
+      "epoch": 2.4307140528203455,
+      "grad_norm": 0.1289438009262085,
+      "learning_rate": 0.0001,
+      "loss": 0.1754,
+      "step": 1491
+    },
+    {
+      "epoch": 2.432344310401043,
+      "grad_norm": 0.1589927077293396,
+      "learning_rate": 0.0001,
+      "loss": 0.1778,
+      "step": 1492
+    },
+    {
+      "epoch": 2.4339745679817413,
+      "grad_norm": 0.13043618202209473,
+      "learning_rate": 0.0001,
+      "loss": 0.1866,
+      "step": 1493
+    },
+    {
+      "epoch": 2.435604825562439,
+      "grad_norm": 0.12837867438793182,
+      "learning_rate": 0.0001,
+      "loss": 0.1644,
+      "step": 1494
+    },
+    {
+      "epoch": 2.4372350831431366,
+      "grad_norm": 0.13602042198181152,
+      "learning_rate": 0.0001,
+      "loss": 0.18,
+      "step": 1495
+    },
+    {
+      "epoch": 2.4388653407238343,
+      "grad_norm": 0.12518282234668732,
+      "learning_rate": 0.0001,
+      "loss": 0.1792,
+      "step": 1496
+    },
+    {
+      "epoch": 2.440495598304532,
+      "grad_norm": 0.1458752602338791,
+      "learning_rate": 0.0001,
+      "loss": 0.174,
+      "step": 1497
+    },
+    {
+      "epoch": 2.4421258558852297,
+      "grad_norm": 0.16408641636371613,
+      "learning_rate": 0.0001,
+      "loss": 0.1638,
+      "step": 1498
+    },
+    {
+      "epoch": 2.443756113465928,
+      "grad_norm": 0.1250162422657013,
+      "learning_rate": 0.0001,
+      "loss": 0.1799,
+      "step": 1499
+    },
+    {
+      "epoch": 2.4453863710466255,
+      "grad_norm": 0.14285162091255188,
+      "learning_rate": 0.0001,
+      "loss": 0.183,
+      "step": 1500
+    },
+    {
+      "epoch": 2.4453863710466255,
+      "eval_loss": 0.19733810424804688,
+      "eval_runtime": 2885.0678,
+      "eval_samples_per_second": 0.654,
+      "eval_steps_per_second": 0.164,
+      "step": 1500
+    },
+    {
+      "epoch": 2.447016628627323,
+      "grad_norm": 0.1123531311750412,
+      "learning_rate": 0.0001,
+      "loss": 0.1653,
+      "step": 1501
+    },
+    {
+      "epoch": 2.448646886208021,
+      "grad_norm": 0.1342422366142273,
+      "learning_rate": 0.0001,
+      "loss": 0.1652,
+      "step": 1502
+    },
+    {
+      "epoch": 2.4502771437887185,
+      "grad_norm": 0.13005055487155914,
+      "learning_rate": 0.0001,
+      "loss": 0.1513,
+      "step": 1503
+    },
+    {
+      "epoch": 2.451907401369416,
+      "grad_norm": 0.1291249841451645,
+      "learning_rate": 0.0001,
+      "loss": 0.1658,
+      "step": 1504
+    },
+    {
+      "epoch": 2.453537658950114,
+      "grad_norm": 0.15009792149066925,
+      "learning_rate": 0.0001,
+      "loss": 0.1582,
+      "step": 1505
+    },
+    {
+      "epoch": 2.455167916530812,
+      "grad_norm": 0.13734357059001923,
+      "learning_rate": 0.0001,
+      "loss": 0.1729,
+      "step": 1506
+    },
+    {
+      "epoch": 2.4567981741115097,
+      "grad_norm": 0.14626893401145935,
+      "learning_rate": 0.0001,
+      "loss": 0.1753,
+      "step": 1507
+    },
+    {
+      "epoch": 2.4584284316922074,
+      "grad_norm": 0.13441750407218933,
+      "learning_rate": 0.0001,
+      "loss": 0.168,
+      "step": 1508
+    },
+    {
+      "epoch": 2.460058689272905,
+      "grad_norm": 0.17883312702178955,
+      "learning_rate": 0.0001,
+      "loss": 0.1724,
+      "step": 1509
+    },
+    {
+      "epoch": 2.4616889468536027,
+      "grad_norm": 0.13286323845386505,
+      "learning_rate": 0.0001,
+      "loss": 0.1707,
+      "step": 1510
+    },
+    {
+      "epoch": 2.463319204434301,
+      "grad_norm": 0.11390995234251022,
+      "learning_rate": 0.0001,
+      "loss": 0.152,
+      "step": 1511
+    },
+    {
+      "epoch": 2.4649494620149985,
+      "grad_norm": 0.13794218003749847,
+      "learning_rate": 0.0001,
+      "loss": 0.1693,
+      "step": 1512
+    },
+    {
+      "epoch": 2.466579719595696,
+      "grad_norm": 0.13029181957244873,
+      "learning_rate": 0.0001,
+      "loss": 0.1885,
+      "step": 1513
+    },
+    {
+      "epoch": 2.468209977176394,
+      "grad_norm": 0.1590529978275299,
+      "learning_rate": 0.0001,
+      "loss": 0.1656,
+      "step": 1514
+    },
+    {
+      "epoch": 2.4698402347570916,
+      "grad_norm": 0.129343181848526,
+      "learning_rate": 0.0001,
+      "loss": 0.1645,
+      "step": 1515
+    },
+    {
+      "epoch": 2.4714704923377893,
+      "grad_norm": 0.13485710322856903,
+      "learning_rate": 0.0001,
+      "loss": 0.178,
+      "step": 1516
+    },
+    {
+      "epoch": 2.473100749918487,
+      "grad_norm": 0.13252151012420654,
+      "learning_rate": 0.0001,
+      "loss": 0.1807,
+      "step": 1517
+    },
+    {
+      "epoch": 2.474731007499185,
+      "grad_norm": 0.14138251543045044,
+      "learning_rate": 0.0001,
+      "loss": 0.1807,
+      "step": 1518
+    },
+    {
+      "epoch": 2.4763612650798827,
+      "grad_norm": 0.12066026031970978,
+      "learning_rate": 0.0001,
+      "loss": 0.1579,
+      "step": 1519
+    },
+    {
+      "epoch": 2.4779915226605804,
+      "grad_norm": 0.128167986869812,
+      "learning_rate": 0.0001,
+      "loss": 0.1765,
+      "step": 1520
+    },
+    {
+      "epoch": 2.479621780241278,
+      "grad_norm": 0.11725836247205734,
+      "learning_rate": 0.0001,
+      "loss": 0.1584,
+      "step": 1521
+    },
+    {
+      "epoch": 2.481252037821976,
+      "grad_norm": 0.12235096096992493,
+      "learning_rate": 0.0001,
+      "loss": 0.1752,
+      "step": 1522
+    },
+    {
+      "epoch": 2.4828822954026735,
+      "grad_norm": 0.13664376735687256,
+      "learning_rate": 0.0001,
+      "loss": 0.1722,
+      "step": 1523
+    },
+    {
+      "epoch": 2.4845125529833716,
+      "grad_norm": 0.12066584080457687,
+      "learning_rate": 0.0001,
+      "loss": 0.1805,
+      "step": 1524
+    },
+    {
+      "epoch": 2.4861428105640693,
+      "grad_norm": 0.15553556382656097,
+      "learning_rate": 0.0001,
+      "loss": 0.1763,
+      "step": 1525
+    },
+    {
+      "epoch": 2.487773068144767,
+      "grad_norm": 0.11850383877754211,
+      "learning_rate": 0.0001,
+      "loss": 0.1758,
+      "step": 1526
+    },
+    {
+      "epoch": 2.4894033257254646,
+      "grad_norm": 0.1156458929181099,
+      "learning_rate": 0.0001,
+      "loss": 0.1723,
+      "step": 1527
+    },
+    {
+      "epoch": 2.4910335833061623,
+      "grad_norm": 0.14494523406028748,
+      "learning_rate": 0.0001,
+      "loss": 0.1721,
+      "step": 1528
+    },
+    {
+      "epoch": 2.49266384088686,
+      "grad_norm": 0.19197629392147064,
+      "learning_rate": 0.0001,
+      "loss": 0.1844,
+      "step": 1529
+    },
+    {
+      "epoch": 2.4942940984675577,
+      "grad_norm": 0.13929398357868195,
+      "learning_rate": 0.0001,
+      "loss": 0.1605,
+      "step": 1530
+    },
+    {
+      "epoch": 2.495924356048256,
+      "grad_norm": 0.14377978444099426,
+      "learning_rate": 0.0001,
+      "loss": 0.1757,
+      "step": 1531
+    },
+    {
+      "epoch": 2.4975546136289535,
+      "grad_norm": 0.1260727494955063,
+      "learning_rate": 0.0001,
+      "loss": 0.1584,
+      "step": 1532
+    },
+    {
+      "epoch": 2.499184871209651,
+      "grad_norm": 0.13934855163097382,
+      "learning_rate": 0.0001,
+      "loss": 0.1906,
+      "step": 1533
+    },
+    {
+      "epoch": 2.500815128790349,
+      "grad_norm": 0.1342552751302719,
+      "learning_rate": 0.0001,
+      "loss": 0.1642,
+      "step": 1534
+    },
+    {
+      "epoch": 2.5024453863710465,
+      "grad_norm": 0.14849533140659332,
+      "learning_rate": 0.0001,
+      "loss": 0.1774,
+      "step": 1535
+    },
+    {
+      "epoch": 2.5040756439517446,
+      "grad_norm": 0.12595170736312866,
+      "learning_rate": 0.0001,
+      "loss": 0.1757,
+      "step": 1536
+    },
+    {
+      "epoch": 2.5057059015324423,
+      "grad_norm": 0.15015050768852234,
+      "learning_rate": 0.0001,
+      "loss": 0.1598,
+      "step": 1537
+    },
+    {
+      "epoch": 2.50733615911314,
+      "grad_norm": 0.1296335756778717,
+      "learning_rate": 0.0001,
+      "loss": 0.1725,
+      "step": 1538
+    },
+    {
+      "epoch": 2.5089664166938377,
+      "grad_norm": 0.12165997922420502,
+      "learning_rate": 0.0001,
+      "loss": 0.1647,
+      "step": 1539
+    },
+    {
+      "epoch": 2.5105966742745354,
+      "grad_norm": 0.14188778400421143,
+      "learning_rate": 0.0001,
+      "loss": 0.1675,
+      "step": 1540
+    },
+    {
+      "epoch": 2.512226931855233,
+      "grad_norm": 0.12897147238254547,
+      "learning_rate": 0.0001,
+      "loss": 0.1661,
+      "step": 1541
+    },
+    {
+      "epoch": 2.5138571894359307,
+      "grad_norm": 0.10982251167297363,
+      "learning_rate": 0.0001,
+      "loss": 0.1734,
+      "step": 1542
+    },
+    {
+      "epoch": 2.5154874470166284,
+      "grad_norm": 0.11350741237401962,
+      "learning_rate": 0.0001,
+      "loss": 0.1587,
+      "step": 1543
+    },
+    {
+      "epoch": 2.5171177045973265,
+      "grad_norm": 0.13737499713897705,
+      "learning_rate": 0.0001,
+      "loss": 0.1791,
+      "step": 1544
+    },
+    {
+      "epoch": 2.518747962178024,
+      "grad_norm": 0.17301884293556213,
+      "learning_rate": 0.0001,
+      "loss": 0.17,
+      "step": 1545
+    },
+    {
+      "epoch": 2.520378219758722,
+      "grad_norm": 0.12698844075202942,
+      "learning_rate": 0.0001,
+      "loss": 0.1803,
+      "step": 1546
+    },
+    {
+      "epoch": 2.5220084773394196,
+      "grad_norm": 0.1362551897764206,
+      "learning_rate": 0.0001,
+      "loss": 0.1666,
+      "step": 1547
+    },
+    {
+      "epoch": 2.5236387349201173,
+      "grad_norm": 0.13497772812843323,
+      "learning_rate": 0.0001,
+      "loss": 0.1757,
+      "step": 1548
+    },
+    {
+      "epoch": 2.5252689925008154,
+      "grad_norm": 0.13246221840381622,
+      "learning_rate": 0.0001,
+      "loss": 0.1626,
+      "step": 1549
+    },
+    {
+      "epoch": 2.526899250081513,
+      "grad_norm": 0.11976587027311325,
+      "learning_rate": 0.0001,
+      "loss": 0.1704,
+      "step": 1550
+    },
+    {
+      "epoch": 2.5285295076622107,
+      "grad_norm": 0.17471329867839813,
+      "learning_rate": 0.0001,
+      "loss": 0.1743,
+      "step": 1551
+    },
+    {
+      "epoch": 2.5301597652429084,
+      "grad_norm": 0.1331670731306076,
+      "learning_rate": 0.0001,
+      "loss": 0.1565,
+      "step": 1552
+    },
+    {
+      "epoch": 2.531790022823606,
+      "grad_norm": 0.12313182651996613,
+      "learning_rate": 0.0001,
+      "loss": 0.1508,
+      "step": 1553
+    },
+    {
+      "epoch": 2.533420280404304,
+      "grad_norm": 0.14300191402435303,
+      "learning_rate": 0.0001,
+      "loss": 0.1825,
+      "step": 1554
+    },
+    {
+      "epoch": 2.5350505379850015,
+      "grad_norm": 0.14361177384853363,
+      "learning_rate": 0.0001,
+      "loss": 0.1751,
+      "step": 1555
+    },
+    {
+      "epoch": 2.536680795565699,
+      "grad_norm": 0.1338524967432022,
+      "learning_rate": 0.0001,
+      "loss": 0.1714,
+      "step": 1556
+    },
+    {
+      "epoch": 2.5383110531463973,
+      "grad_norm": 0.12020072340965271,
+      "learning_rate": 0.0001,
+      "loss": 0.1745,
+      "step": 1557
+    },
+    {
+      "epoch": 2.539941310727095,
+      "grad_norm": 0.1332116425037384,
+      "learning_rate": 0.0001,
+      "loss": 0.1644,
+      "step": 1558
+    },
+    {
+      "epoch": 2.5415715683077926,
+      "grad_norm": 0.1357092410326004,
+      "learning_rate": 0.0001,
+      "loss": 0.1667,
+      "step": 1559
+    },
+    {
+      "epoch": 2.5432018258884903,
+      "grad_norm": 0.12332719564437866,
+      "learning_rate": 0.0001,
+      "loss": 0.178,
+      "step": 1560
+    },
+    {
+      "epoch": 2.544832083469188,
+      "grad_norm": 0.11925667524337769,
+      "learning_rate": 0.0001,
+      "loss": 0.1736,
+      "step": 1561
+    },
+    {
+      "epoch": 2.546462341049886,
+      "grad_norm": 0.14202888309955597,
+      "learning_rate": 0.0001,
+      "loss": 0.1795,
+      "step": 1562
+    },
+    {
+      "epoch": 2.548092598630584,
+      "grad_norm": 0.18294650316238403,
+      "learning_rate": 0.0001,
+      "loss": 0.1859,
+      "step": 1563
+    },
+    {
+      "epoch": 2.5497228562112815,
+      "grad_norm": 0.14184100925922394,
+      "learning_rate": 0.0001,
+      "loss": 0.1628,
+      "step": 1564
+    },
+    {
+      "epoch": 2.551353113791979,
+      "grad_norm": 0.1371891051530838,
+      "learning_rate": 0.0001,
+      "loss": 0.18,
+      "step": 1565
+    },
+    {
+      "epoch": 2.552983371372677,
+      "grad_norm": 0.20214851200580597,
+      "learning_rate": 0.0001,
+      "loss": 0.1813,
+      "step": 1566
+    },
+    {
+      "epoch": 2.5546136289533745,
+      "grad_norm": 0.12881435453891754,
+      "learning_rate": 0.0001,
+      "loss": 0.1657,
+      "step": 1567
+    },
+    {
+      "epoch": 2.556243886534072,
+      "grad_norm": 0.1161297932267189,
+      "learning_rate": 0.0001,
+      "loss": 0.1641,
+      "step": 1568
+    },
+    {
+      "epoch": 2.55787414411477,
+      "grad_norm": 0.13545578718185425,
+      "learning_rate": 0.0001,
+      "loss": 0.17,
+      "step": 1569
+    },
+    {
+      "epoch": 2.559504401695468,
+      "grad_norm": 0.13674761354923248,
+      "learning_rate": 0.0001,
+      "loss": 0.175,
+      "step": 1570
+    },
+    {
+      "epoch": 2.5611346592761657,
+      "grad_norm": 0.12067017704248428,
+      "learning_rate": 0.0001,
+      "loss": 0.1545,
+      "step": 1571
+    },
+    {
+      "epoch": 2.5627649168568634,
+      "grad_norm": 0.13512328267097473,
+      "learning_rate": 0.0001,
+      "loss": 0.1706,
+      "step": 1572
+    },
+    {
+      "epoch": 2.564395174437561,
+      "grad_norm": 0.14814037084579468,
+      "learning_rate": 0.0001,
+      "loss": 0.1824,
+      "step": 1573
+    },
+    {
+      "epoch": 2.5660254320182587,
+      "grad_norm": 0.13570277392864227,
+      "learning_rate": 0.0001,
+      "loss": 0.18,
+      "step": 1574
+    },
+    {
+      "epoch": 2.567655689598957,
+      "grad_norm": 0.14428043365478516,
+      "learning_rate": 0.0001,
+      "loss": 0.176,
+      "step": 1575
+    },
+    {
+      "epoch": 2.5692859471796545,
+      "grad_norm": 0.11816335469484329,
+      "learning_rate": 0.0001,
+      "loss": 0.1658,
+      "step": 1576
+    },
+    {
+      "epoch": 2.570916204760352,
+      "grad_norm": 0.12405506521463394,
+      "learning_rate": 0.0001,
+      "loss": 0.1704,
+      "step": 1577
+    },
+    {
+      "epoch": 2.57254646234105,
+      "grad_norm": 0.12209967523813248,
+      "learning_rate": 0.0001,
+      "loss": 0.1683,
+      "step": 1578
+    },
+    {
+      "epoch": 2.5741767199217476,
+      "grad_norm": 0.12708422541618347,
+      "learning_rate": 0.0001,
+      "loss": 0.1846,
+      "step": 1579
+    },
+    {
+      "epoch": 2.5758069775024452,
+      "grad_norm": 0.13518854975700378,
+      "learning_rate": 0.0001,
+      "loss": 0.1692,
+      "step": 1580
+    },
+    {
+      "epoch": 2.577437235083143,
+      "grad_norm": 0.1394702047109604,
+      "learning_rate": 0.0001,
+      "loss": 0.1793,
+      "step": 1581
+    },
+    {
+      "epoch": 2.5790674926638406,
+      "grad_norm": 0.1345827430486679,
+      "learning_rate": 0.0001,
+      "loss": 0.1578,
+      "step": 1582
+    },
+    {
+      "epoch": 2.5806977502445387,
+      "grad_norm": 0.1303935945034027,
+      "learning_rate": 0.0001,
+      "loss": 0.1754,
+      "step": 1583
+    },
+    {
+      "epoch": 2.5823280078252364,
+      "grad_norm": 0.1360052078962326,
+      "learning_rate": 0.0001,
+      "loss": 0.1825,
+      "step": 1584
+    },
+    {
+      "epoch": 2.583958265405934,
+      "grad_norm": 0.1259981095790863,
+      "learning_rate": 0.0001,
+      "loss": 0.1651,
+      "step": 1585
+    },
+    {
+      "epoch": 2.5855885229866318,
+      "grad_norm": 0.13449037075042725,
+      "learning_rate": 0.0001,
+      "loss": 0.1901,
+      "step": 1586
+    },
+    {
+      "epoch": 2.58721878056733,
+      "grad_norm": 0.14254209399223328,
+      "learning_rate": 0.0001,
+      "loss": 0.187,
+      "step": 1587
+    },
+    {
+      "epoch": 2.5888490381480276,
+      "grad_norm": 0.11680305004119873,
+      "learning_rate": 0.0001,
+      "loss": 0.1669,
+      "step": 1588
+    },
+    {
+      "epoch": 2.5904792957287253,
+      "grad_norm": 0.12670142948627472,
+      "learning_rate": 0.0001,
+      "loss": 0.1659,
+      "step": 1589
+    },
+    {
+      "epoch": 2.592109553309423,
+      "grad_norm": 0.11500969529151917,
+      "learning_rate": 0.0001,
+      "loss": 0.1523,
+      "step": 1590
+    },
+    {
+      "epoch": 2.5937398108901206,
+      "grad_norm": 0.15863120555877686,
+      "learning_rate": 0.0001,
+      "loss": 0.1731,
+      "step": 1591
+    },
+    {
+      "epoch": 2.5953700684708183,
+      "grad_norm": 0.12249530106782913,
+      "learning_rate": 0.0001,
+      "loss": 0.1722,
+      "step": 1592
+    },
+    {
+      "epoch": 2.597000326051516,
+      "grad_norm": 0.1352062076330185,
+      "learning_rate": 0.0001,
+      "loss": 0.1787,
+      "step": 1593
+    },
+    {
+      "epoch": 2.5986305836322137,
+      "grad_norm": 0.14207401871681213,
+      "learning_rate": 0.0001,
+      "loss": 0.1954,
+      "step": 1594
+    },
+    {
+      "epoch": 2.600260841212912,
+      "grad_norm": 0.12589991092681885,
+      "learning_rate": 0.0001,
+      "loss": 0.161,
+      "step": 1595
+    },
+    {
+      "epoch": 2.6018910987936095,
+      "grad_norm": 0.15374112129211426,
+      "learning_rate": 0.0001,
+      "loss": 0.1631,
+      "step": 1596
+    },
+    {
+      "epoch": 2.603521356374307,
+      "grad_norm": 0.17806057631969452,
+      "learning_rate": 0.0001,
+      "loss": 0.1749,
+      "step": 1597
+    },
+    {
+      "epoch": 2.605151613955005,
+      "grad_norm": 0.13044816255569458,
+      "learning_rate": 0.0001,
+      "loss": 0.1602,
+      "step": 1598
+    },
+    {
+      "epoch": 2.6067818715357025,
+      "grad_norm": 0.13261814415454865,
+      "learning_rate": 0.0001,
+      "loss": 0.175,
+      "step": 1599
+    },
+    {
+      "epoch": 2.6084121291164006,
+      "grad_norm": 0.16431686282157898,
+      "learning_rate": 0.0001,
+      "loss": 0.1711,
+      "step": 1600
+    },
+    {
+      "epoch": 2.6100423866970983,
+      "grad_norm": 0.1591222733259201,
+      "learning_rate": 0.0001,
+      "loss": 0.1782,
+      "step": 1601
+    },
+    {
+      "epoch": 2.611672644277796,
+      "grad_norm": 0.15045471489429474,
+      "learning_rate": 0.0001,
+      "loss": 0.1576,
+      "step": 1602
+    },
+    {
+      "epoch": 2.6133029018584937,
+      "grad_norm": 0.14002764225006104,
+      "learning_rate": 0.0001,
+      "loss": 0.1731,
+      "step": 1603
+    },
+    {
+      "epoch": 2.6149331594391914,
+      "grad_norm": 0.13808737695217133,
+      "learning_rate": 0.0001,
+      "loss": 0.1754,
+      "step": 1604
+    },
+    {
+      "epoch": 2.616563417019889,
+      "grad_norm": 0.14910684525966644,
+      "learning_rate": 0.0001,
+      "loss": 0.1834,
+      "step": 1605
+    },
+    {
+      "epoch": 2.6181936746005867,
+      "grad_norm": 0.1458890289068222,
+      "learning_rate": 0.0001,
+      "loss": 0.173,
+      "step": 1606
+    },
+    {
+      "epoch": 2.6198239321812844,
+      "grad_norm": 0.13817395269870758,
+      "learning_rate": 0.0001,
+      "loss": 0.1745,
+      "step": 1607
+    },
+    {
+      "epoch": 2.6214541897619825,
+      "grad_norm": 0.22925616800785065,
+      "learning_rate": 0.0001,
+      "loss": 0.1693,
+      "step": 1608
+    },
+    {
+      "epoch": 2.62308444734268,
+      "grad_norm": 0.13849849998950958,
+      "learning_rate": 0.0001,
+      "loss": 0.1734,
+      "step": 1609
+    },
+    {
+      "epoch": 2.624714704923378,
+      "grad_norm": 0.12482542544603348,
+      "learning_rate": 0.0001,
+      "loss": 0.1864,
+      "step": 1610
+    },
+    {
+      "epoch": 2.6263449625040756,
+      "grad_norm": 0.12124403566122055,
+      "learning_rate": 0.0001,
+      "loss": 0.1803,
+      "step": 1611
+    },
+    {
+      "epoch": 2.6279752200847732,
+      "grad_norm": 0.12327743321657181,
+      "learning_rate": 0.0001,
+      "loss": 0.1581,
+      "step": 1612
+    },
+    {
+      "epoch": 2.6296054776654714,
+      "grad_norm": 0.17278893291950226,
+      "learning_rate": 0.0001,
+      "loss": 0.1637,
+      "step": 1613
+    },
+    {
+      "epoch": 2.631235735246169,
+      "grad_norm": 0.14170318841934204,
+      "learning_rate": 0.0001,
+      "loss": 0.1793,
+      "step": 1614
+    },
+    {
+      "epoch": 2.6328659928268667,
+      "grad_norm": 0.1279682219028473,
+      "learning_rate": 0.0001,
+      "loss": 0.1657,
+      "step": 1615
+    },
+    {
+      "epoch": 2.6344962504075644,
+      "grad_norm": 0.1350080519914627,
+      "learning_rate": 0.0001,
+      "loss": 0.1854,
+      "step": 1616
+    },
+    {
+      "epoch": 2.636126507988262,
+      "grad_norm": 0.1315418928861618,
+      "learning_rate": 0.0001,
+      "loss": 0.1732,
+      "step": 1617
+    },
+    {
+      "epoch": 2.6377567655689598,
+      "grad_norm": 0.13384385406970978,
+      "learning_rate": 0.0001,
+      "loss": 0.1727,
+      "step": 1618
+    },
+    {
+      "epoch": 2.6393870231496575,
+      "grad_norm": 0.4788396954536438,
+      "learning_rate": 0.0001,
+      "loss": 0.1705,
+      "step": 1619
+    },
+    {
+      "epoch": 2.641017280730355,
+      "grad_norm": 0.15386568009853363,
+      "learning_rate": 0.0001,
+      "loss": 0.1653,
+      "step": 1620
+    },
+    {
+      "epoch": 2.6426475383110533,
+      "grad_norm": 0.12126284837722778,
+      "learning_rate": 0.0001,
+      "loss": 0.1645,
+      "step": 1621
+    },
+    {
+      "epoch": 2.644277795891751,
+      "grad_norm": 0.11877623945474625,
+      "learning_rate": 0.0001,
+      "loss": 0.1656,
+      "step": 1622
+    },
+    {
+      "epoch": 2.6459080534724486,
+      "grad_norm": 0.13270951807498932,
+      "learning_rate": 0.0001,
+      "loss": 0.1861,
+      "step": 1623
+    },
+    {
+      "epoch": 2.6475383110531463,
+      "grad_norm": 0.13268467783927917,
+      "learning_rate": 0.0001,
+      "loss": 0.1803,
+      "step": 1624
+    },
+    {
+      "epoch": 2.6491685686338444,
+      "grad_norm": 0.11460261046886444,
+      "learning_rate": 0.0001,
+      "loss": 0.1849,
+      "step": 1625
+    },
+    {
+      "epoch": 2.650798826214542,
+      "grad_norm": 0.1381942331790924,
+      "learning_rate": 0.0001,
+      "loss": 0.1726,
+      "step": 1626
+    },
+    {
+      "epoch": 2.65242908379524,
+      "grad_norm": 0.11194758862257004,
+      "learning_rate": 0.0001,
+      "loss": 0.174,
+      "step": 1627
+    },
+    {
+      "epoch": 2.6540593413759375,
+      "grad_norm": 0.13582143187522888,
+      "learning_rate": 0.0001,
+      "loss": 0.188,
+      "step": 1628
+    },
+    {
+      "epoch": 2.655689598956635,
+      "grad_norm": 0.13221515715122223,
+      "learning_rate": 0.0001,
+      "loss": 0.1784,
+      "step": 1629
+    },
+    {
+      "epoch": 2.657319856537333,
+      "grad_norm": 0.13104896247386932,
+      "learning_rate": 0.0001,
+      "loss": 0.1644,
+      "step": 1630
+    },
+    {
+      "epoch": 2.6589501141180305,
+      "grad_norm": 0.15535131096839905,
+      "learning_rate": 0.0001,
+      "loss": 0.1723,
+      "step": 1631
+    },
+    {
+      "epoch": 2.660580371698728,
+      "grad_norm": 0.1257065385580063,
+      "learning_rate": 0.0001,
+      "loss": 0.1437,
+      "step": 1632
+    },
+    {
+      "epoch": 2.6622106292794263,
+      "grad_norm": 0.12850813567638397,
+      "learning_rate": 0.0001,
+      "loss": 0.1735,
+      "step": 1633
+    },
+    {
+      "epoch": 2.663840886860124,
+      "grad_norm": 0.12646383047103882,
+      "learning_rate": 0.0001,
+      "loss": 0.1771,
+      "step": 1634
+    },
+    {
+      "epoch": 2.6654711444408217,
+      "grad_norm": 0.2852911651134491,
+      "learning_rate": 0.0001,
+      "loss": 0.1851,
+      "step": 1635
+    },
+    {
+      "epoch": 2.6671014020215194,
+      "grad_norm": 0.11714810878038406,
+      "learning_rate": 0.0001,
+      "loss": 0.1549,
+      "step": 1636
+    },
+    {
+      "epoch": 2.668731659602217,
+      "grad_norm": 0.14019465446472168,
+      "learning_rate": 0.0001,
+      "loss": 0.185,
+      "step": 1637
+    },
+    {
+      "epoch": 2.670361917182915,
+      "grad_norm": 0.12677006423473358,
+      "learning_rate": 0.0001,
+      "loss": 0.1657,
+      "step": 1638
+    },
+    {
+      "epoch": 2.671992174763613,
+      "grad_norm": 0.12617303431034088,
+      "learning_rate": 0.0001,
+      "loss": 0.1729,
+      "step": 1639
+    },
+    {
+      "epoch": 2.6736224323443105,
+      "grad_norm": 0.1309848427772522,
+      "learning_rate": 0.0001,
+      "loss": 0.1661,
+      "step": 1640
+    },
+    {
+      "epoch": 2.675252689925008,
+      "grad_norm": 0.1385383903980255,
+      "learning_rate": 0.0001,
+      "loss": 0.1794,
+      "step": 1641
+    },
+    {
+      "epoch": 2.676882947505706,
+      "grad_norm": 0.1279730498790741,
+      "learning_rate": 0.0001,
+      "loss": 0.1661,
+      "step": 1642
+    },
+    {
+      "epoch": 2.6785132050864036,
+      "grad_norm": 0.12799672782421112,
+      "learning_rate": 0.0001,
+      "loss": 0.1628,
+      "step": 1643
+    },
+    {
+      "epoch": 2.6801434626671012,
+      "grad_norm": 0.16915291547775269,
+      "learning_rate": 0.0001,
+      "loss": 0.1707,
+      "step": 1644
+    },
+    {
+      "epoch": 2.681773720247799,
+      "grad_norm": 0.12088014930486679,
+      "learning_rate": 0.0001,
+      "loss": 0.1671,
+      "step": 1645
+    },
+    {
+      "epoch": 2.683403977828497,
+      "grad_norm": 0.13347816467285156,
+      "learning_rate": 0.0001,
+      "loss": 0.1715,
+      "step": 1646
+    },
+    {
+      "epoch": 2.6850342354091947,
+      "grad_norm": 0.12426477670669556,
+      "learning_rate": 0.0001,
+      "loss": 0.1917,
+      "step": 1647
+    },
+    {
+      "epoch": 2.6866644929898924,
+      "grad_norm": 0.16994018852710724,
+      "learning_rate": 0.0001,
+      "loss": 0.1929,
+      "step": 1648
+    },
+    {
+      "epoch": 2.68829475057059,
+      "grad_norm": 0.1453407108783722,
+      "learning_rate": 0.0001,
+      "loss": 0.1711,
+      "step": 1649
+    },
+    {
+      "epoch": 2.6899250081512878,
+      "grad_norm": 0.15362900495529175,
+      "learning_rate": 0.0001,
+      "loss": 0.1829,
+      "step": 1650
+    },
+    {
+      "epoch": 2.691555265731986,
+      "grad_norm": 0.15922309458255768,
+      "learning_rate": 0.0001,
+      "loss": 0.1655,
+      "step": 1651
+    },
+    {
+      "epoch": 2.6931855233126836,
+      "grad_norm": 0.153437077999115,
+      "learning_rate": 0.0001,
+      "loss": 0.1776,
+      "step": 1652
+    },
+    {
+      "epoch": 2.6948157808933813,
+      "grad_norm": 0.15213799476623535,
+      "learning_rate": 0.0001,
+      "loss": 0.1835,
+      "step": 1653
+    },
+    {
+      "epoch": 2.696446038474079,
+      "grad_norm": 0.10505233705043793,
+      "learning_rate": 0.0001,
+      "loss": 0.1501,
+      "step": 1654
+    },
+    {
+      "epoch": 2.6980762960547766,
+      "grad_norm": 0.13109369575977325,
+      "learning_rate": 0.0001,
+      "loss": 0.1601,
+      "step": 1655
+    },
+    {
+      "epoch": 2.6997065536354743,
+      "grad_norm": 0.1327008157968521,
+      "learning_rate": 0.0001,
+      "loss": 0.1781,
+      "step": 1656
+    },
+    {
+      "epoch": 2.701336811216172,
+      "grad_norm": 0.12708407640457153,
+      "learning_rate": 0.0001,
+      "loss": 0.1599,
+      "step": 1657
+    },
+    {
+      "epoch": 2.7029670687968697,
+      "grad_norm": 0.12450224906206131,
+      "learning_rate": 0.0001,
+      "loss": 0.1711,
+      "step": 1658
+    },
+    {
+      "epoch": 2.704597326377568,
+      "grad_norm": 0.13624098896980286,
+      "learning_rate": 0.0001,
+      "loss": 0.1636,
+      "step": 1659
+    },
+    {
+      "epoch": 2.7062275839582655,
+      "grad_norm": 0.1352638155221939,
+      "learning_rate": 0.0001,
+      "loss": 0.1732,
+      "step": 1660
+    },
+    {
+      "epoch": 2.707857841538963,
+      "grad_norm": 0.1240679994225502,
+      "learning_rate": 0.0001,
+      "loss": 0.1715,
+      "step": 1661
+    },
+    {
+      "epoch": 2.709488099119661,
+      "grad_norm": 0.131019726395607,
+      "learning_rate": 0.0001,
+      "loss": 0.1743,
+      "step": 1662
+    },
+    {
+      "epoch": 2.7111183567003585,
+      "grad_norm": 0.11371717602014542,
+      "learning_rate": 0.0001,
+      "loss": 0.1653,
+      "step": 1663
+    },
+    {
+      "epoch": 2.7127486142810566,
+      "grad_norm": 0.1314827799797058,
+      "learning_rate": 0.0001,
+      "loss": 0.1788,
+      "step": 1664
+    },
+    {
+      "epoch": 2.7143788718617543,
+      "grad_norm": 0.11662380397319794,
+      "learning_rate": 0.0001,
+      "loss": 0.1596,
+      "step": 1665
+    },
+    {
+      "epoch": 2.716009129442452,
+      "grad_norm": 0.18013271689414978,
+      "learning_rate": 0.0001,
+      "loss": 0.1539,
+      "step": 1666
+    },
+    {
+      "epoch": 2.7176393870231497,
+      "grad_norm": 0.1361352503299713,
+      "learning_rate": 0.0001,
+      "loss": 0.1683,
+      "step": 1667
+    },
+    {
+      "epoch": 2.7192696446038473,
+      "grad_norm": 0.12468544393777847,
+      "learning_rate": 0.0001,
+      "loss": 0.1651,
+      "step": 1668
+    },
+    {
+      "epoch": 2.720899902184545,
+      "grad_norm": 0.14304213225841522,
+      "learning_rate": 0.0001,
+      "loss": 0.1766,
+      "step": 1669
+    },
+    {
+      "epoch": 2.7225301597652427,
+      "grad_norm": 0.13735635578632355,
+      "learning_rate": 0.0001,
+      "loss": 0.1828,
+      "step": 1670
+    },
+    {
+      "epoch": 2.7241604173459404,
+      "grad_norm": 0.1301272064447403,
+      "learning_rate": 0.0001,
+      "loss": 0.1694,
+      "step": 1671
+    },
+    {
+      "epoch": 2.7257906749266385,
+      "grad_norm": 0.13478688895702362,
+      "learning_rate": 0.0001,
+      "loss": 0.1809,
+      "step": 1672
+    },
+    {
+      "epoch": 2.727420932507336,
+      "grad_norm": 0.1462695151567459,
+      "learning_rate": 0.0001,
+      "loss": 0.1605,
+      "step": 1673
+    },
+    {
+      "epoch": 2.729051190088034,
+      "grad_norm": 0.13804949820041656,
+      "learning_rate": 0.0001,
+      "loss": 0.1672,
+      "step": 1674
+    },
+    {
+      "epoch": 2.7306814476687316,
+      "grad_norm": 0.12185340374708176,
+      "learning_rate": 0.0001,
+      "loss": 0.1665,
+      "step": 1675
+    },
+    {
+      "epoch": 2.7323117052494297,
+      "grad_norm": 0.13339075446128845,
+      "learning_rate": 0.0001,
+      "loss": 0.1788,
+      "step": 1676
+    },
+    {
+      "epoch": 2.7339419628301274,
+      "grad_norm": 0.12658098340034485,
+      "learning_rate": 0.0001,
+      "loss": 0.1549,
+      "step": 1677
+    },
+    {
+      "epoch": 2.735572220410825,
+      "grad_norm": 0.12869013845920563,
+      "learning_rate": 0.0001,
+      "loss": 0.1788,
+      "step": 1678
+    },
+    {
+      "epoch": 2.7372024779915227,
+      "grad_norm": 0.12980249524116516,
+      "learning_rate": 0.0001,
+      "loss": 0.1783,
+      "step": 1679
+    },
+    {
+      "epoch": 2.7388327355722204,
+      "grad_norm": 0.14288075268268585,
+      "learning_rate": 0.0001,
+      "loss": 0.1708,
+      "step": 1680
+    },
+    {
+      "epoch": 2.740462993152918,
+      "grad_norm": 0.15478843450546265,
+      "learning_rate": 0.0001,
+      "loss": 0.1803,
+      "step": 1681
+    },
+    {
+      "epoch": 2.7420932507336158,
+      "grad_norm": 0.13585714995861053,
+      "learning_rate": 0.0001,
+      "loss": 0.185,
+      "step": 1682
+    },
+    {
+      "epoch": 2.7437235083143134,
+      "grad_norm": 0.1448957920074463,
+      "learning_rate": 0.0001,
+      "loss": 0.1716,
+      "step": 1683
+    },
+    {
+      "epoch": 2.7453537658950116,
+      "grad_norm": 0.14890708029270172,
+      "learning_rate": 0.0001,
+      "loss": 0.181,
+      "step": 1684
+    },
+    {
+      "epoch": 2.7469840234757092,
+      "grad_norm": 0.13350339233875275,
+      "learning_rate": 0.0001,
+      "loss": 0.1757,
+      "step": 1685
+    },
+    {
+      "epoch": 2.748614281056407,
+      "grad_norm": 0.1049543023109436,
+      "learning_rate": 0.0001,
+      "loss": 0.1668,
+      "step": 1686
+    },
+    {
+      "epoch": 2.7502445386371046,
+      "grad_norm": 3.6135470867156982,
+      "learning_rate": 0.0001,
+      "loss": 0.2121,
+      "step": 1687
+    },
+    {
+      "epoch": 2.7518747962178023,
+      "grad_norm": 0.1557306945323944,
+      "learning_rate": 0.0001,
+      "loss": 0.1742,
+      "step": 1688
+    },
+    {
+      "epoch": 2.7535050537985004,
+      "grad_norm": 0.1619507372379303,
+      "learning_rate": 0.0001,
+      "loss": 0.1586,
+      "step": 1689
+    },
+    {
+      "epoch": 2.755135311379198,
+      "grad_norm": 0.12189088016748428,
+      "learning_rate": 0.0001,
+      "loss": 0.1789,
+      "step": 1690
+    },
+    {
+      "epoch": 2.7567655689598958,
+      "grad_norm": 0.15941926836967468,
+      "learning_rate": 0.0001,
+      "loss": 0.1727,
+      "step": 1691
+    },
+    {
+      "epoch": 2.7583958265405935,
+      "grad_norm": 0.11842834204435349,
+      "learning_rate": 0.0001,
+      "loss": 0.1592,
+      "step": 1692
+    },
+    {
+      "epoch": 2.760026084121291,
+      "grad_norm": 0.1222245842218399,
+      "learning_rate": 0.0001,
+      "loss": 0.1593,
+      "step": 1693
+    },
+    {
+      "epoch": 2.761656341701989,
+      "grad_norm": 0.1369628608226776,
+      "learning_rate": 0.0001,
+      "loss": 0.155,
+      "step": 1694
+    },
+    {
+      "epoch": 2.7632865992826865,
+      "grad_norm": 0.14212529361248016,
+      "learning_rate": 0.0001,
+      "loss": 0.1681,
+      "step": 1695
+    },
+    {
+      "epoch": 2.764916856863384,
+      "grad_norm": 0.14328838884830475,
+      "learning_rate": 0.0001,
+      "loss": 0.1844,
+      "step": 1696
+    },
+    {
+      "epoch": 2.7665471144440823,
+      "grad_norm": 0.14077410101890564,
+      "learning_rate": 0.0001,
+      "loss": 0.1856,
+      "step": 1697
+    },
+    {
+      "epoch": 2.76817737202478,
+      "grad_norm": 0.13928896188735962,
+      "learning_rate": 0.0001,
+      "loss": 0.1718,
+      "step": 1698
+    },
+    {
+      "epoch": 2.7698076296054777,
+      "grad_norm": 0.13314402103424072,
+      "learning_rate": 0.0001,
+      "loss": 0.1705,
+      "step": 1699
+    },
+    {
+      "epoch": 2.7714378871861753,
+      "grad_norm": 0.1279967725276947,
+      "learning_rate": 0.0001,
+      "loss": 0.1758,
+      "step": 1700
+    },
+    {
+      "epoch": 2.773068144766873,
+      "grad_norm": 0.11493542045354843,
+      "learning_rate": 0.0001,
+      "loss": 0.1539,
+      "step": 1701
+    },
+    {
+      "epoch": 2.774698402347571,
+      "grad_norm": 0.12252707779407501,
+      "learning_rate": 0.0001,
+      "loss": 0.1752,
+      "step": 1702
+    },
+    {
+      "epoch": 2.776328659928269,
+      "grad_norm": 0.19633330404758453,
+      "learning_rate": 0.0001,
+      "loss": 0.1582,
+      "step": 1703
+    },
+    {
+      "epoch": 2.7779589175089665,
+      "grad_norm": 0.1423460990190506,
+      "learning_rate": 0.0001,
+      "loss": 0.1668,
+      "step": 1704
+    },
+    {
+      "epoch": 2.779589175089664,
+      "grad_norm": 0.13523077964782715,
+      "learning_rate": 0.0001,
+      "loss": 0.1843,
+      "step": 1705
+    },
+    {
+      "epoch": 2.781219432670362,
+      "grad_norm": 0.13212746381759644,
+      "learning_rate": 0.0001,
+      "loss": 0.1783,
+      "step": 1706
+    },
+    {
+      "epoch": 2.7828496902510595,
+      "grad_norm": 0.1263880431652069,
+      "learning_rate": 0.0001,
+      "loss": 0.1579,
+      "step": 1707
+    },
+    {
+      "epoch": 2.7844799478317572,
+      "grad_norm": 0.12073154002428055,
+      "learning_rate": 0.0001,
+      "loss": 0.1569,
+      "step": 1708
+    },
+    {
+      "epoch": 2.786110205412455,
+      "grad_norm": 0.14019356667995453,
+      "learning_rate": 0.0001,
+      "loss": 0.1809,
+      "step": 1709
+    },
+    {
+      "epoch": 2.787740462993153,
+      "grad_norm": 0.11406267434358597,
+      "learning_rate": 0.0001,
+      "loss": 0.1814,
+      "step": 1710
+    },
+    {
+      "epoch": 2.7893707205738507,
+      "grad_norm": 0.11895663291215897,
+      "learning_rate": 0.0001,
+      "loss": 0.1783,
+      "step": 1711
+    },
+    {
+      "epoch": 2.7910009781545484,
+      "grad_norm": 0.14709694683551788,
+      "learning_rate": 0.0001,
+      "loss": 0.1686,
+      "step": 1712
+    },
+    {
+      "epoch": 2.792631235735246,
+      "grad_norm": 0.13388898968696594,
+      "learning_rate": 0.0001,
+      "loss": 0.1654,
+      "step": 1713
+    },
+    {
+      "epoch": 2.7942614933159438,
+      "grad_norm": 0.1226249486207962,
+      "learning_rate": 0.0001,
+      "loss": 0.1739,
+      "step": 1714
+    },
+    {
+      "epoch": 2.795891750896642,
+      "grad_norm": 0.12996350228786469,
+      "learning_rate": 0.0001,
+      "loss": 0.1699,
+      "step": 1715
+    },
+    {
+      "epoch": 2.7975220084773396,
+      "grad_norm": 0.13227185606956482,
+      "learning_rate": 0.0001,
+      "loss": 0.1774,
+      "step": 1716
+    },
+    {
+      "epoch": 2.7991522660580372,
+      "grad_norm": 0.12332453578710556,
+      "learning_rate": 0.0001,
+      "loss": 0.1602,
+      "step": 1717
+    },
+    {
+      "epoch": 2.800782523638735,
+      "grad_norm": 0.1187395229935646,
+      "learning_rate": 0.0001,
+      "loss": 0.1578,
+      "step": 1718
+    },
+    {
+      "epoch": 2.8024127812194326,
+      "grad_norm": 0.13751424849033356,
+      "learning_rate": 0.0001,
+      "loss": 0.1734,
+      "step": 1719
+    },
+    {
+      "epoch": 2.8040430388001303,
+      "grad_norm": 0.13629640638828278,
+      "learning_rate": 0.0001,
+      "loss": 0.173,
+      "step": 1720
+    },
+    {
+      "epoch": 2.805673296380828,
+      "grad_norm": 0.11772475391626358,
+      "learning_rate": 0.0001,
+      "loss": 0.1677,
+      "step": 1721
+    },
+    {
+      "epoch": 2.8073035539615256,
+      "grad_norm": 0.13224674761295319,
+      "learning_rate": 0.0001,
+      "loss": 0.1757,
+      "step": 1722
+    },
+    {
+      "epoch": 2.8089338115422238,
+      "grad_norm": 0.1200365424156189,
+      "learning_rate": 0.0001,
+      "loss": 0.1792,
+      "step": 1723
+    },
+    {
+      "epoch": 2.8105640691229214,
+      "grad_norm": 0.14300167560577393,
+      "learning_rate": 0.0001,
+      "loss": 0.1748,
+      "step": 1724
+    },
+    {
+      "epoch": 2.812194326703619,
+      "grad_norm": 0.10958956182003021,
+      "learning_rate": 0.0001,
+      "loss": 0.1633,
+      "step": 1725
+    },
+    {
+      "epoch": 2.813824584284317,
+      "grad_norm": 0.13673417270183563,
+      "learning_rate": 0.0001,
+      "loss": 0.1618,
+      "step": 1726
+    },
+    {
+      "epoch": 2.815454841865015,
+      "grad_norm": 0.1306380033493042,
+      "learning_rate": 0.0001,
+      "loss": 0.183,
+      "step": 1727
+    },
+    {
+      "epoch": 2.8170850994457126,
+      "grad_norm": 0.12033358216285706,
+      "learning_rate": 0.0001,
+      "loss": 0.1616,
+      "step": 1728
+    },
+    {
+      "epoch": 2.8187153570264103,
+      "grad_norm": 0.14117878675460815,
+      "learning_rate": 0.0001,
+      "loss": 0.1715,
+      "step": 1729
+    },
+    {
+      "epoch": 2.820345614607108,
+      "grad_norm": 0.13309049606323242,
+      "learning_rate": 0.0001,
+      "loss": 0.1773,
+      "step": 1730
+    },
+    {
+      "epoch": 2.8219758721878057,
+      "grad_norm": 0.12040352821350098,
+      "learning_rate": 0.0001,
+      "loss": 0.1656,
+      "step": 1731
+    },
+    {
+      "epoch": 2.8236061297685033,
+      "grad_norm": 0.12827961146831512,
+      "learning_rate": 0.0001,
+      "loss": 0.1661,
+      "step": 1732
+    },
+    {
+      "epoch": 2.825236387349201,
+      "grad_norm": 0.1280909925699234,
+      "learning_rate": 0.0001,
+      "loss": 0.1806,
+      "step": 1733
+    },
+    {
+      "epoch": 2.8268666449298987,
+      "grad_norm": 0.12497354298830032,
+      "learning_rate": 0.0001,
+      "loss": 0.1788,
+      "step": 1734
+    },
+    {
+      "epoch": 2.828496902510597,
+      "grad_norm": 0.1294706016778946,
+      "learning_rate": 0.0001,
+      "loss": 0.1743,
+      "step": 1735
+    },
+    {
+      "epoch": 2.8301271600912945,
+      "grad_norm": 0.14673174917697906,
+      "learning_rate": 0.0001,
+      "loss": 0.168,
+      "step": 1736
+    },
+    {
+      "epoch": 2.831757417671992,
+      "grad_norm": 0.14964553713798523,
+      "learning_rate": 0.0001,
+      "loss": 0.1751,
+      "step": 1737
+    },
+    {
+      "epoch": 2.83338767525269,
+      "grad_norm": 0.17080165445804596,
+      "learning_rate": 0.0001,
+      "loss": 0.1651,
+      "step": 1738
+    },
+    {
+      "epoch": 2.8350179328333875,
+      "grad_norm": 0.12748171389102936,
+      "learning_rate": 0.0001,
+      "loss": 0.1565,
+      "step": 1739
+    },
+    {
+      "epoch": 2.8366481904140857,
+      "grad_norm": 0.1372041255235672,
+      "learning_rate": 0.0001,
+      "loss": 0.1816,
+      "step": 1740
+    },
+    {
+      "epoch": 2.8382784479947833,
+      "grad_norm": 0.13018721342086792,
+      "learning_rate": 0.0001,
+      "loss": 0.1758,
+      "step": 1741
+    },
+    {
+      "epoch": 2.839908705575481,
+      "grad_norm": 0.13896982371807098,
+      "learning_rate": 0.0001,
+      "loss": 0.1713,
+      "step": 1742
+    },
+    {
+      "epoch": 2.8415389631561787,
+      "grad_norm": 0.13519124686717987,
+      "learning_rate": 0.0001,
+      "loss": 0.1826,
+      "step": 1743
+    },
+    {
+      "epoch": 2.8431692207368764,
+      "grad_norm": 0.12529662251472473,
+      "learning_rate": 0.0001,
+      "loss": 0.1657,
+      "step": 1744
+    },
+    {
+      "epoch": 2.844799478317574,
+      "grad_norm": 0.1385057270526886,
+      "learning_rate": 0.0001,
+      "loss": 0.1742,
+      "step": 1745
+    },
+    {
+      "epoch": 2.8464297358982718,
+      "grad_norm": 0.13610967993736267,
+      "learning_rate": 0.0001,
+      "loss": 0.1729,
+      "step": 1746
+    },
+    {
+      "epoch": 2.8480599934789694,
+      "grad_norm": 0.1311800330877304,
+      "learning_rate": 0.0001,
+      "loss": 0.1829,
+      "step": 1747
+    },
+    {
+      "epoch": 2.8496902510596676,
+      "grad_norm": 0.13023614883422852,
+      "learning_rate": 0.0001,
+      "loss": 0.1679,
+      "step": 1748
+    },
+    {
+      "epoch": 2.8513205086403652,
+      "grad_norm": 0.12245821207761765,
+      "learning_rate": 0.0001,
+      "loss": 0.1699,
+      "step": 1749
+    },
+    {
+      "epoch": 2.852950766221063,
+      "grad_norm": 0.1297629326581955,
+      "learning_rate": 0.0001,
+      "loss": 0.172,
+      "step": 1750
+    },
+    {
+      "epoch": 2.8545810238017606,
+      "grad_norm": 0.12192010134458542,
+      "learning_rate": 0.0001,
+      "loss": 0.1727,
+      "step": 1751
+    },
+    {
+      "epoch": 2.8562112813824583,
+      "grad_norm": 0.12483784556388855,
+      "learning_rate": 0.0001,
+      "loss": 0.1658,
+      "step": 1752
+    },
+    {
+      "epoch": 2.8578415389631564,
+      "grad_norm": 0.1716589629650116,
+      "learning_rate": 0.0001,
+      "loss": 0.1795,
+      "step": 1753
+    },
+    {
+      "epoch": 2.859471796543854,
+      "grad_norm": 0.1266416311264038,
+      "learning_rate": 0.0001,
+      "loss": 0.1572,
+      "step": 1754
+    },
+    {
+      "epoch": 2.8611020541245518,
+      "grad_norm": 0.15566131472587585,
+      "learning_rate": 0.0001,
+      "loss": 0.1812,
+      "step": 1755
+    },
+    {
+      "epoch": 2.8627323117052494,
+      "grad_norm": 0.1435212343931198,
+      "learning_rate": 0.0001,
+      "loss": 0.1682,
+      "step": 1756
+    },
+    {
+      "epoch": 2.864362569285947,
+      "grad_norm": 0.13166531920433044,
+      "learning_rate": 0.0001,
+      "loss": 0.17,
+      "step": 1757
+    },
+    {
+      "epoch": 2.865992826866645,
+      "grad_norm": 0.1274840533733368,
+      "learning_rate": 0.0001,
+      "loss": 0.1661,
+      "step": 1758
+    },
+    {
+      "epoch": 2.8676230844473425,
+      "grad_norm": 0.1092241182923317,
+      "learning_rate": 0.0001,
+      "loss": 0.1655,
+      "step": 1759
+    },
+    {
+      "epoch": 2.86925334202804,
+      "grad_norm": 0.13460928201675415,
+      "learning_rate": 0.0001,
+      "loss": 0.1627,
+      "step": 1760
+    },
+    {
+      "epoch": 2.8708835996087383,
+      "grad_norm": 0.12913478910923004,
+      "learning_rate": 0.0001,
+      "loss": 0.1618,
+      "step": 1761
+    },
+    {
+      "epoch": 2.872513857189436,
+      "grad_norm": 0.11661716550588608,
+      "learning_rate": 0.0001,
+      "loss": 0.169,
+      "step": 1762
+    },
+    {
+      "epoch": 2.8741441147701337,
+      "grad_norm": 0.11179400235414505,
+      "learning_rate": 0.0001,
+      "loss": 0.1725,
+      "step": 1763
+    },
+    {
+      "epoch": 2.8757743723508313,
+      "grad_norm": 0.1252732276916504,
+      "learning_rate": 0.0001,
+      "loss": 0.1827,
+      "step": 1764
+    },
+    {
+      "epoch": 2.8774046299315295,
+      "grad_norm": 0.3473842144012451,
+      "learning_rate": 0.0001,
+      "loss": 0.1693,
+      "step": 1765
+    },
+    {
+      "epoch": 2.879034887512227,
+      "grad_norm": 0.1262052208185196,
+      "learning_rate": 0.0001,
+      "loss": 0.1728,
+      "step": 1766
+    },
+    {
+      "epoch": 2.880665145092925,
+      "grad_norm": 0.1326300948858261,
+      "learning_rate": 0.0001,
+      "loss": 0.1744,
+      "step": 1767
+    },
+    {
+      "epoch": 2.8822954026736225,
+      "grad_norm": 0.1303003877401352,
+      "learning_rate": 0.0001,
+      "loss": 0.1714,
+      "step": 1768
+    },
+    {
+      "epoch": 2.88392566025432,
+      "grad_norm": 0.12167539447546005,
+      "learning_rate": 0.0001,
+      "loss": 0.1677,
+      "step": 1769
+    },
+    {
+      "epoch": 2.885555917835018,
+      "grad_norm": 0.13617132604122162,
+      "learning_rate": 0.0001,
+      "loss": 0.1646,
+      "step": 1770
+    },
+    {
+      "epoch": 2.8871861754157155,
+      "grad_norm": 0.13883374631404877,
+      "learning_rate": 0.0001,
+      "loss": 0.173,
+      "step": 1771
+    },
+    {
+      "epoch": 2.888816432996413,
+      "grad_norm": 0.12531040608882904,
+      "learning_rate": 0.0001,
+      "loss": 0.1678,
+      "step": 1772
+    },
+    {
+      "epoch": 2.8904466905771113,
+      "grad_norm": 0.13892248272895813,
+      "learning_rate": 0.0001,
+      "loss": 0.1677,
+      "step": 1773
+    },
+    {
+      "epoch": 2.892076948157809,
+      "grad_norm": 0.14446908235549927,
+      "learning_rate": 0.0001,
+      "loss": 0.1666,
+      "step": 1774
+    },
+    {
+      "epoch": 2.8937072057385067,
+      "grad_norm": 0.14126956462860107,
+      "learning_rate": 0.0001,
+      "loss": 0.1626,
+      "step": 1775
+    },
+    {
+      "epoch": 2.8953374633192044,
+      "grad_norm": 0.1542399823665619,
+      "learning_rate": 0.0001,
+      "loss": 0.1741,
+      "step": 1776
+    },
+    {
+      "epoch": 2.896967720899902,
+      "grad_norm": 0.14131003618240356,
+      "learning_rate": 0.0001,
+      "loss": 0.174,
+      "step": 1777
+    },
+    {
+      "epoch": 2.8985979784806,
+      "grad_norm": 0.15592537820339203,
+      "learning_rate": 0.0001,
+      "loss": 0.1847,
+      "step": 1778
+    },
+    {
+      "epoch": 2.900228236061298,
+      "grad_norm": 0.14722147583961487,
+      "learning_rate": 0.0001,
+      "loss": 0.1804,
+      "step": 1779
+    },
+    {
+      "epoch": 2.9018584936419956,
+      "grad_norm": 0.13757076859474182,
+      "learning_rate": 0.0001,
+      "loss": 0.1856,
+      "step": 1780
+    },
+    {
+      "epoch": 2.9034887512226932,
+      "grad_norm": 0.2536503076553345,
+      "learning_rate": 0.0001,
+      "loss": 0.1513,
+      "step": 1781
+    },
+    {
+      "epoch": 2.905119008803391,
+      "grad_norm": 0.11684080958366394,
+      "learning_rate": 0.0001,
+      "loss": 0.1723,
+      "step": 1782
+    },
+    {
+      "epoch": 2.9067492663840886,
+      "grad_norm": 0.1301044374704361,
+      "learning_rate": 0.0001,
+      "loss": 0.1756,
+      "step": 1783
+    },
+    {
+      "epoch": 2.9083795239647863,
+      "grad_norm": 0.11158990114927292,
+      "learning_rate": 0.0001,
+      "loss": 0.1627,
+      "step": 1784
+    },
+    {
+      "epoch": 2.910009781545484,
+      "grad_norm": 0.14631851017475128,
+      "learning_rate": 0.0001,
+      "loss": 0.1763,
+      "step": 1785
+    },
+    {
+      "epoch": 2.911640039126182,
+      "grad_norm": 0.12340549379587173,
+      "learning_rate": 0.0001,
+      "loss": 0.1826,
+      "step": 1786
+    },
+    {
+      "epoch": 2.9132702967068798,
+      "grad_norm": 0.15119358897209167,
+      "learning_rate": 0.0001,
+      "loss": 0.1517,
+      "step": 1787
+    },
+    {
+      "epoch": 2.9149005542875774,
+      "grad_norm": 0.10862945020198822,
+      "learning_rate": 0.0001,
+      "loss": 0.1636,
+      "step": 1788
+    },
+    {
+      "epoch": 2.916530811868275,
+      "grad_norm": 0.12316179275512695,
+      "learning_rate": 0.0001,
+      "loss": 0.1726,
+      "step": 1789
+    },
+    {
+      "epoch": 2.918161069448973,
+      "grad_norm": 0.14677046239376068,
+      "learning_rate": 0.0001,
+      "loss": 0.1705,
+      "step": 1790
+    },
+    {
+      "epoch": 2.919791327029671,
+      "grad_norm": 0.1339065581560135,
+      "learning_rate": 0.0001,
+      "loss": 0.1771,
+      "step": 1791
+    },
+    {
+      "epoch": 2.9214215846103686,
+      "grad_norm": 0.1291988492012024,
+      "learning_rate": 0.0001,
+      "loss": 0.1741,
+      "step": 1792
+    },
+    {
+      "epoch": 2.9230518421910663,
+      "grad_norm": 0.13565215468406677,
+      "learning_rate": 0.0001,
+      "loss": 0.1692,
+      "step": 1793
+    },
+    {
+      "epoch": 2.924682099771764,
+      "grad_norm": 0.13081243634223938,
+      "learning_rate": 0.0001,
+      "loss": 0.1622,
+      "step": 1794
+    },
+    {
+      "epoch": 2.9263123573524616,
+      "grad_norm": 0.12821637094020844,
+      "learning_rate": 0.0001,
+      "loss": 0.1592,
+      "step": 1795
+    },
+    {
+      "epoch": 2.9279426149331593,
+      "grad_norm": 0.17648915946483612,
+      "learning_rate": 0.0001,
+      "loss": 0.1875,
+      "step": 1796
+    },
+    {
+      "epoch": 2.929572872513857,
+      "grad_norm": 0.11781003326177597,
+      "learning_rate": 0.0001,
+      "loss": 0.1591,
+      "step": 1797
+    },
+    {
+      "epoch": 2.9312031300945547,
+      "grad_norm": 0.13282893598079681,
+      "learning_rate": 0.0001,
+      "loss": 0.1734,
+      "step": 1798
+    },
+    {
+      "epoch": 2.932833387675253,
+      "grad_norm": 0.14549997448921204,
+      "learning_rate": 0.0001,
+      "loss": 0.1792,
+      "step": 1799
+    },
+    {
+      "epoch": 2.9344636452559505,
+      "grad_norm": 0.12919341027736664,
+      "learning_rate": 0.0001,
+      "loss": 0.1682,
+      "step": 1800
+    },
+    {
+      "epoch": 2.936093902836648,
+      "grad_norm": 0.12719695270061493,
+      "learning_rate": 0.0001,
+      "loss": 0.1824,
+      "step": 1801
+    },
+    {
+      "epoch": 2.937724160417346,
+      "grad_norm": 0.14096450805664062,
+      "learning_rate": 0.0001,
+      "loss": 0.1609,
+      "step": 1802
+    },
+    {
+      "epoch": 2.9393544179980435,
+      "grad_norm": 0.11929788440465927,
+      "learning_rate": 0.0001,
+      "loss": 0.1602,
+      "step": 1803
+    },
+    {
+      "epoch": 2.9409846755787417,
+      "grad_norm": 0.16668562591075897,
+      "learning_rate": 0.0001,
+      "loss": 0.1767,
+      "step": 1804
+    },
+    {
+      "epoch": 2.9426149331594393,
+      "grad_norm": 0.14096197485923767,
+      "learning_rate": 0.0001,
+      "loss": 0.1696,
+      "step": 1805
+    },
+    {
+      "epoch": 2.944245190740137,
+      "grad_norm": 0.1540384292602539,
+      "learning_rate": 0.0001,
+      "loss": 0.1849,
+      "step": 1806
+    },
+    {
+      "epoch": 2.9458754483208347,
+      "grad_norm": 0.11548353731632233,
+      "learning_rate": 0.0001,
+      "loss": 0.1676,
+      "step": 1807
+    },
+    {
+      "epoch": 2.9475057059015324,
+      "grad_norm": 0.12311840057373047,
+      "learning_rate": 0.0001,
+      "loss": 0.1664,
+      "step": 1808
+    },
+    {
+      "epoch": 2.94913596348223,
+      "grad_norm": 0.1358504742383957,
+      "learning_rate": 0.0001,
+      "loss": 0.1752,
+      "step": 1809
+    },
+    {
+      "epoch": 2.9507662210629277,
+      "grad_norm": 0.12379997223615646,
+      "learning_rate": 0.0001,
+      "loss": 0.1656,
+      "step": 1810
+    },
+    {
+      "epoch": 2.9523964786436254,
+      "grad_norm": 0.13933683931827545,
+      "learning_rate": 0.0001,
+      "loss": 0.1797,
+      "step": 1811
+    },
+    {
+      "epoch": 2.9540267362243235,
+      "grad_norm": 0.1396905779838562,
+      "learning_rate": 0.0001,
+      "loss": 0.1761,
+      "step": 1812
+    },
+    {
+      "epoch": 2.9556569938050212,
+      "grad_norm": 0.12376896291971207,
+      "learning_rate": 0.0001,
+      "loss": 0.1608,
+      "step": 1813
+    },
+    {
+      "epoch": 2.957287251385719,
+      "grad_norm": 0.12583288550376892,
+      "learning_rate": 0.0001,
+      "loss": 0.1684,
+      "step": 1814
+    },
+    {
+      "epoch": 2.9589175089664166,
+      "grad_norm": 0.13755297660827637,
+      "learning_rate": 0.0001,
+      "loss": 0.144,
+      "step": 1815
+    },
+    {
+      "epoch": 2.9605477665471147,
+      "grad_norm": 0.24326035380363464,
+      "learning_rate": 0.0001,
+      "loss": 0.1885,
+      "step": 1816
+    },
+    {
+      "epoch": 2.9621780241278124,
+      "grad_norm": 0.14443641901016235,
+      "learning_rate": 0.0001,
+      "loss": 0.1885,
+      "step": 1817
+    },
+    {
+      "epoch": 2.96380828170851,
+      "grad_norm": 0.12700864672660828,
+      "learning_rate": 0.0001,
+      "loss": 0.1641,
+      "step": 1818
+    },
+    {
+      "epoch": 2.9654385392892078,
+      "grad_norm": 0.148148775100708,
+      "learning_rate": 0.0001,
+      "loss": 0.1792,
+      "step": 1819
+    },
+    {
+      "epoch": 2.9670687968699054,
+      "grad_norm": 0.14224189519882202,
+      "learning_rate": 0.0001,
+      "loss": 0.1692,
+      "step": 1820
+    },
+    {
+      "epoch": 2.968699054450603,
+      "grad_norm": 0.15919244289398193,
+      "learning_rate": 0.0001,
+      "loss": 0.174,
+      "step": 1821
+    },
+    {
+      "epoch": 2.970329312031301,
+      "grad_norm": 0.2009061574935913,
+      "learning_rate": 0.0001,
+      "loss": 0.1737,
+      "step": 1822
+    },
+    {
+      "epoch": 2.9719595696119985,
+      "grad_norm": 0.12979143857955933,
+      "learning_rate": 0.0001,
+      "loss": 0.1668,
+      "step": 1823
+    },
+    {
+      "epoch": 2.9735898271926966,
+      "grad_norm": 0.12879891693592072,
+      "learning_rate": 0.0001,
+      "loss": 0.1608,
+      "step": 1824
+    },
+    {
+      "epoch": 2.9752200847733943,
+      "grad_norm": 0.11508017033338547,
+      "learning_rate": 0.0001,
+      "loss": 0.1705,
+      "step": 1825
+    },
+    {
+      "epoch": 2.976850342354092,
+      "grad_norm": 0.12390004843473434,
+      "learning_rate": 0.0001,
+      "loss": 0.1697,
+      "step": 1826
+    },
+    {
+      "epoch": 2.9784805999347896,
+      "grad_norm": 0.13983500003814697,
+      "learning_rate": 0.0001,
+      "loss": 0.1744,
+      "step": 1827
+    },
+    {
+      "epoch": 2.9801108575154873,
+      "grad_norm": 0.12721027433872223,
+      "learning_rate": 0.0001,
+      "loss": 0.1799,
+      "step": 1828
+    },
+    {
+      "epoch": 2.9817411150961854,
+      "grad_norm": 0.12273556739091873,
+      "learning_rate": 0.0001,
+      "loss": 0.1664,
+      "step": 1829
+    },
+    {
+      "epoch": 2.983371372676883,
+      "grad_norm": 0.12730666995048523,
+      "learning_rate": 0.0001,
+      "loss": 0.1716,
+      "step": 1830
+    },
+    {
+      "epoch": 2.985001630257581,
+      "grad_norm": 0.12331203371286392,
+      "learning_rate": 0.0001,
+      "loss": 0.1641,
+      "step": 1831
+    },
+    {
+      "epoch": 2.9866318878382785,
+      "grad_norm": 0.1258230060338974,
+      "learning_rate": 0.0001,
+      "loss": 0.1778,
+      "step": 1832
+    },
+    {
+      "epoch": 2.988262145418976,
+      "grad_norm": 0.14592202007770538,
+      "learning_rate": 0.0001,
+      "loss": 0.1888,
+      "step": 1833
+    },
+    {
+      "epoch": 2.989892402999674,
+      "grad_norm": 0.12880602478981018,
+      "learning_rate": 0.0001,
+      "loss": 0.1731,
+      "step": 1834
+    },
+    {
+      "epoch": 2.9915226605803715,
+      "grad_norm": 0.14615651965141296,
+      "learning_rate": 0.0001,
+      "loss": 0.1853,
+      "step": 1835
+    },
+    {
+      "epoch": 2.993152918161069,
+      "grad_norm": 0.12081672251224518,
+      "learning_rate": 0.0001,
+      "loss": 0.1837,
+      "step": 1836
+    },
+    {
+      "epoch": 2.9947831757417673,
+      "grad_norm": 0.11534418165683746,
+      "learning_rate": 0.0001,
+      "loss": 0.1773,
+      "step": 1837
+    },
+    {
+      "epoch": 2.996413433322465,
+      "grad_norm": 0.12042343616485596,
+      "learning_rate": 0.0001,
+      "loss": 0.1594,
+      "step": 1838
+    },
+    {
+      "epoch": 2.9980436909031627,
+      "grad_norm": 0.13987913727760315,
+      "learning_rate": 0.0001,
+      "loss": 0.1785,
+      "step": 1839
+    },
+    {
+      "epoch": 2.9996739484838604,
+      "grad_norm": 0.12604045867919922,
+      "learning_rate": 0.0001,
+      "loss": 0.1708,
+      "step": 1840
+    },
+    {
+      "epoch": 3.001304206064558,
+      "grad_norm": 0.13217991590499878,
+      "learning_rate": 0.0001,
+      "loss": 0.1621,
+      "step": 1841
+    },
+    {
+      "epoch": 3.0029344636452557,
+      "grad_norm": 0.1318402737379074,
+      "learning_rate": 0.0001,
+      "loss": 0.1565,
+      "step": 1842
+    },
+    {
+      "epoch": 3.004564721225954,
+      "grad_norm": 0.12177060544490814,
+      "learning_rate": 0.0001,
+      "loss": 0.162,
+      "step": 1843
+    },
+    {
+      "epoch": 3.0061949788066515,
+      "grad_norm": 0.11984521895647049,
+      "learning_rate": 0.0001,
+      "loss": 0.1441,
+      "step": 1844
+    },
+    {
+      "epoch": 3.007825236387349,
+      "grad_norm": 0.13043449819087982,
+      "learning_rate": 0.0001,
+      "loss": 0.1419,
+      "step": 1845
+    },
+    {
+      "epoch": 3.009455493968047,
+      "grad_norm": 0.12092608213424683,
+      "learning_rate": 0.0001,
+      "loss": 0.1503,
+      "step": 1846
+    },
+    {
+      "epoch": 3.0110857515487446,
+      "grad_norm": 0.15196233987808228,
+      "learning_rate": 0.0001,
+      "loss": 0.161,
+      "step": 1847
+    },
+    {
+      "epoch": 3.0127160091294423,
+      "grad_norm": 0.14589709043502808,
+      "learning_rate": 0.0001,
+      "loss": 0.1564,
+      "step": 1848
+    },
+    {
+      "epoch": 3.0143462667101404,
+      "grad_norm": 0.1587577909231186,
+      "learning_rate": 0.0001,
+      "loss": 0.1673,
+      "step": 1849
+    },
+    {
+      "epoch": 3.015976524290838,
+      "grad_norm": 0.14549551904201508,
+      "learning_rate": 0.0001,
+      "loss": 0.1573,
+      "step": 1850
+    },
+    {
+      "epoch": 3.0176067818715357,
+      "grad_norm": 0.1457890272140503,
+      "learning_rate": 0.0001,
+      "loss": 0.1552,
+      "step": 1851
+    },
+    {
+      "epoch": 3.0192370394522334,
+      "grad_norm": 0.1330796182155609,
+      "learning_rate": 0.0001,
+      "loss": 0.1522,
+      "step": 1852
+    },
+    {
+      "epoch": 3.020867297032931,
+      "grad_norm": 0.16735856235027313,
+      "learning_rate": 0.0001,
+      "loss": 0.1618,
+      "step": 1853
+    },
+    {
+      "epoch": 3.022497554613629,
+      "grad_norm": 0.13368208706378937,
+      "learning_rate": 0.0001,
+      "loss": 0.1522,
+      "step": 1854
+    },
+    {
+      "epoch": 3.024127812194327,
+      "grad_norm": 0.14208734035491943,
+      "learning_rate": 0.0001,
+      "loss": 0.1537,
+      "step": 1855
+    },
+    {
+      "epoch": 3.0257580697750246,
+      "grad_norm": 0.17149995267391205,
+      "learning_rate": 0.0001,
+      "loss": 0.1688,
+      "step": 1856
+    },
+    {
+      "epoch": 3.0273883273557223,
+      "grad_norm": 0.16402693092823029,
+      "learning_rate": 0.0001,
+      "loss": 0.153,
+      "step": 1857
+    },
+    {
+      "epoch": 3.02901858493642,
+      "grad_norm": 0.14791880548000336,
+      "learning_rate": 0.0001,
+      "loss": 0.1415,
+      "step": 1858
+    },
+    {
+      "epoch": 3.0306488425171176,
+      "grad_norm": 0.13278983533382416,
+      "learning_rate": 0.0001,
+      "loss": 0.1624,
+      "step": 1859
+    },
+    {
+      "epoch": 3.0322791000978153,
+      "grad_norm": 0.1486644297838211,
+      "learning_rate": 0.0001,
+      "loss": 0.1489,
+      "step": 1860
+    },
+    {
+      "epoch": 3.033909357678513,
+      "grad_norm": 0.15381312370300293,
+      "learning_rate": 0.0001,
+      "loss": 0.1546,
+      "step": 1861
+    },
+    {
+      "epoch": 3.035539615259211,
+      "grad_norm": 0.1434403508901596,
+      "learning_rate": 0.0001,
+      "loss": 0.1595,
+      "step": 1862
+    },
+    {
+      "epoch": 3.037169872839909,
+      "grad_norm": 0.14210708439350128,
+      "learning_rate": 0.0001,
+      "loss": 0.1533,
+      "step": 1863
+    },
+    {
+      "epoch": 3.0388001304206065,
+      "grad_norm": 0.1316472291946411,
+      "learning_rate": 0.0001,
+      "loss": 0.1406,
+      "step": 1864
+    },
+    {
+      "epoch": 3.040430388001304,
+      "grad_norm": 0.17375947535037994,
+      "learning_rate": 0.0001,
+      "loss": 0.1727,
+      "step": 1865
+    },
+    {
+      "epoch": 3.042060645582002,
+      "grad_norm": 0.13698440790176392,
+      "learning_rate": 0.0001,
+      "loss": 0.1458,
+      "step": 1866
+    },
+    {
+      "epoch": 3.0436909031626995,
+      "grad_norm": 0.14209020137786865,
+      "learning_rate": 0.0001,
+      "loss": 0.1467,
+      "step": 1867
+    },
+    {
+      "epoch": 3.0453211607433976,
+      "grad_norm": 0.14422006905078888,
+      "learning_rate": 0.0001,
+      "loss": 0.1502,
+      "step": 1868
+    },
+    {
+      "epoch": 3.0469514183240953,
+      "grad_norm": 0.13612805306911469,
+      "learning_rate": 0.0001,
+      "loss": 0.1519,
+      "step": 1869
+    },
+    {
+      "epoch": 3.048581675904793,
+      "grad_norm": 0.16070745885372162,
+      "learning_rate": 0.0001,
+      "loss": 0.1606,
+      "step": 1870
+    },
+    {
+      "epoch": 3.0502119334854907,
+      "grad_norm": 0.1366911679506302,
+      "learning_rate": 0.0001,
+      "loss": 0.1532,
+      "step": 1871
+    },
+    {
+      "epoch": 3.0518421910661884,
+      "grad_norm": 0.1536417007446289,
+      "learning_rate": 0.0001,
+      "loss": 0.1624,
+      "step": 1872
+    },
+    {
+      "epoch": 3.053472448646886,
+      "grad_norm": 0.14232809841632843,
+      "learning_rate": 0.0001,
+      "loss": 0.161,
+      "step": 1873
+    },
+    {
+      "epoch": 3.055102706227584,
+      "grad_norm": 0.13991610705852509,
+      "learning_rate": 0.0001,
+      "loss": 0.1451,
+      "step": 1874
+    },
+    {
+      "epoch": 3.056732963808282,
+      "grad_norm": 0.13531944155693054,
+      "learning_rate": 0.0001,
+      "loss": 0.1551,
+      "step": 1875
+    },
+    {
+      "epoch": 3.0583632213889795,
+      "grad_norm": 0.1320919394493103,
+      "learning_rate": 0.0001,
+      "loss": 0.1637,
+      "step": 1876
+    },
+    {
+      "epoch": 3.059993478969677,
+      "grad_norm": 0.1661989539861679,
+      "learning_rate": 0.0001,
+      "loss": 0.168,
+      "step": 1877
+    },
+    {
+      "epoch": 3.061623736550375,
+      "grad_norm": 0.14109432697296143,
+      "learning_rate": 0.0001,
+      "loss": 0.1594,
+      "step": 1878
+    },
+    {
+      "epoch": 3.0632539941310726,
+      "grad_norm": 0.12099778652191162,
+      "learning_rate": 0.0001,
+      "loss": 0.1361,
+      "step": 1879
+    },
+    {
+      "epoch": 3.0648842517117703,
+      "grad_norm": 0.15160542726516724,
+      "learning_rate": 0.0001,
+      "loss": 0.1442,
+      "step": 1880
+    },
+    {
+      "epoch": 3.0665145092924684,
+      "grad_norm": 0.15511579811573029,
+      "learning_rate": 0.0001,
+      "loss": 0.1517,
+      "step": 1881
+    },
+    {
+      "epoch": 3.068144766873166,
+      "grad_norm": 0.15297721326351166,
+      "learning_rate": 0.0001,
+      "loss": 0.1484,
+      "step": 1882
+    },
+    {
+      "epoch": 3.0697750244538637,
+      "grad_norm": 0.15919755399227142,
+      "learning_rate": 0.0001,
+      "loss": 0.1538,
+      "step": 1883
+    },
+    {
+      "epoch": 3.0714052820345614,
+      "grad_norm": 0.16629868745803833,
+      "learning_rate": 0.0001,
+      "loss": 0.1727,
+      "step": 1884
+    },
+    {
+      "epoch": 3.073035539615259,
+      "grad_norm": 0.15279151499271393,
+      "learning_rate": 0.0001,
+      "loss": 0.167,
+      "step": 1885
+    },
+    {
+      "epoch": 3.074665797195957,
+      "grad_norm": 0.13714101910591125,
+      "learning_rate": 0.0001,
+      "loss": 0.1564,
+      "step": 1886
+    },
+    {
+      "epoch": 3.076296054776655,
+      "grad_norm": 0.13693886995315552,
+      "learning_rate": 0.0001,
+      "loss": 0.1528,
+      "step": 1887
+    },
+    {
+      "epoch": 3.0779263123573526,
+      "grad_norm": 0.1455615609884262,
+      "learning_rate": 0.0001,
+      "loss": 0.1606,
+      "step": 1888
+    },
+    {
+      "epoch": 3.0795565699380503,
+      "grad_norm": 0.12297336757183075,
+      "learning_rate": 0.0001,
+      "loss": 0.1513,
+      "step": 1889
+    },
+    {
+      "epoch": 3.081186827518748,
+      "grad_norm": 0.13729313015937805,
+      "learning_rate": 0.0001,
+      "loss": 0.1594,
+      "step": 1890
+    },
+    {
+      "epoch": 3.0828170850994456,
+      "grad_norm": 0.15476392209529877,
+      "learning_rate": 0.0001,
+      "loss": 0.1646,
+      "step": 1891
+    },
+    {
+      "epoch": 3.0844473426801433,
+      "grad_norm": 0.13134579360485077,
+      "learning_rate": 0.0001,
+      "loss": 0.145,
+      "step": 1892
+    },
+    {
+      "epoch": 3.0860776002608414,
+      "grad_norm": 0.12954765558242798,
+      "learning_rate": 0.0001,
+      "loss": 0.1463,
+      "step": 1893
+    },
+    {
+      "epoch": 3.087707857841539,
+      "grad_norm": 0.13961279392242432,
+      "learning_rate": 0.0001,
+      "loss": 0.1502,
+      "step": 1894
+    },
+    {
+      "epoch": 3.089338115422237,
+      "grad_norm": 0.1431739330291748,
+      "learning_rate": 0.0001,
+      "loss": 0.1542,
+      "step": 1895
+    },
+    {
+      "epoch": 3.0909683730029345,
+      "grad_norm": 0.13088122010231018,
+      "learning_rate": 0.0001,
+      "loss": 0.1418,
+      "step": 1896
+    },
+    {
+      "epoch": 3.092598630583632,
+      "grad_norm": 0.14733296632766724,
+      "learning_rate": 0.0001,
+      "loss": 0.1506,
+      "step": 1897
+    },
+    {
+      "epoch": 3.09422888816433,
+      "grad_norm": 0.14778561890125275,
+      "learning_rate": 0.0001,
+      "loss": 0.1492,
+      "step": 1898
+    },
+    {
+      "epoch": 3.0958591457450275,
+      "grad_norm": 0.17354664206504822,
+      "learning_rate": 0.0001,
+      "loss": 0.1725,
+      "step": 1899
+    },
+    {
+      "epoch": 3.0974894033257256,
+      "grad_norm": 0.17014777660369873,
+      "learning_rate": 0.0001,
+      "loss": 0.1653,
+      "step": 1900
+    },
+    {
+      "epoch": 3.0991196609064233,
+      "grad_norm": 0.1588321030139923,
+      "learning_rate": 0.0001,
+      "loss": 0.1618,
+      "step": 1901
+    },
+    {
+      "epoch": 3.100749918487121,
+      "grad_norm": 0.1591597944498062,
+      "learning_rate": 0.0001,
+      "loss": 0.1574,
+      "step": 1902
+    },
+    {
+      "epoch": 3.1023801760678187,
+      "grad_norm": 0.1335812509059906,
+      "learning_rate": 0.0001,
+      "loss": 0.1508,
+      "step": 1903
+    },
+    {
+      "epoch": 3.1040104336485164,
+      "grad_norm": 0.13163426518440247,
+      "learning_rate": 0.0001,
+      "loss": 0.134,
+      "step": 1904
+    },
+    {
+      "epoch": 3.105640691229214,
+      "grad_norm": 0.17262452840805054,
+      "learning_rate": 0.0001,
+      "loss": 0.1662,
+      "step": 1905
+    },
+    {
+      "epoch": 3.107270948809912,
+      "grad_norm": 0.1346106082201004,
+      "learning_rate": 0.0001,
+      "loss": 0.1519,
+      "step": 1906
+    },
+    {
+      "epoch": 3.10890120639061,
+      "grad_norm": 0.15132804214954376,
+      "learning_rate": 0.0001,
+      "loss": 0.1567,
+      "step": 1907
+    },
+    {
+      "epoch": 3.1105314639713075,
+      "grad_norm": 0.14656689763069153,
+      "learning_rate": 0.0001,
+      "loss": 0.174,
+      "step": 1908
+    },
+    {
+      "epoch": 3.112161721552005,
+      "grad_norm": 0.12101754546165466,
+      "learning_rate": 0.0001,
+      "loss": 0.1313,
+      "step": 1909
+    },
+    {
+      "epoch": 3.113791979132703,
+      "grad_norm": 0.13873204588890076,
+      "learning_rate": 0.0001,
+      "loss": 0.1685,
+      "step": 1910
+    },
+    {
+      "epoch": 3.1154222367134006,
+      "grad_norm": 0.14690999686717987,
+      "learning_rate": 0.0001,
+      "loss": 0.1522,
+      "step": 1911
+    },
+    {
+      "epoch": 3.1170524942940983,
+      "grad_norm": 0.14875850081443787,
+      "learning_rate": 0.0001,
+      "loss": 0.1669,
+      "step": 1912
+    },
+    {
+      "epoch": 3.1186827518747964,
+      "grad_norm": 0.12188220769166946,
+      "learning_rate": 0.0001,
+      "loss": 0.1423,
+      "step": 1913
+    },
+    {
+      "epoch": 3.120313009455494,
+      "grad_norm": 0.14481259882450104,
+      "learning_rate": 0.0001,
+      "loss": 0.1567,
+      "step": 1914
+    },
+    {
+      "epoch": 3.1219432670361917,
+      "grad_norm": 0.1382434368133545,
+      "learning_rate": 0.0001,
+      "loss": 0.1495,
+      "step": 1915
+    },
+    {
+      "epoch": 3.1235735246168894,
+      "grad_norm": 0.14566384255886078,
+      "learning_rate": 0.0001,
+      "loss": 0.1476,
+      "step": 1916
+    },
+    {
+      "epoch": 3.125203782197587,
+      "grad_norm": 0.14900517463684082,
+      "learning_rate": 0.0001,
+      "loss": 0.1642,
+      "step": 1917
+    },
+    {
+      "epoch": 3.126834039778285,
+      "grad_norm": 0.1237114742398262,
+      "learning_rate": 0.0001,
+      "loss": 0.1387,
+      "step": 1918
+    },
+    {
+      "epoch": 3.128464297358983,
+      "grad_norm": 0.13117025792598724,
+      "learning_rate": 0.0001,
+      "loss": 0.1515,
+      "step": 1919
+    },
+    {
+      "epoch": 3.1300945549396806,
+      "grad_norm": 0.13853514194488525,
+      "learning_rate": 0.0001,
+      "loss": 0.1496,
+      "step": 1920
+    },
+    {
+      "epoch": 3.1317248125203783,
+      "grad_norm": 0.17825017869472504,
+      "learning_rate": 0.0001,
+      "loss": 0.1578,
+      "step": 1921
+    },
+    {
+      "epoch": 3.133355070101076,
+      "grad_norm": 0.1505771279335022,
+      "learning_rate": 0.0001,
+      "loss": 0.1535,
+      "step": 1922
+    },
+    {
+      "epoch": 3.1349853276817736,
+      "grad_norm": 0.1433846801519394,
+      "learning_rate": 0.0001,
+      "loss": 0.1572,
+      "step": 1923
+    },
+    {
+      "epoch": 3.1366155852624713,
+      "grad_norm": 0.1596456617116928,
+      "learning_rate": 0.0001,
+      "loss": 0.1604,
+      "step": 1924
+    },
+    {
+      "epoch": 3.138245842843169,
+      "grad_norm": 0.14370723068714142,
+      "learning_rate": 0.0001,
+      "loss": 0.1608,
+      "step": 1925
+    },
+    {
+      "epoch": 3.139876100423867,
+      "grad_norm": 0.16248880326747894,
+      "learning_rate": 0.0001,
+      "loss": 0.171,
+      "step": 1926
+    },
+    {
+      "epoch": 3.141506358004565,
+      "grad_norm": 0.1393871307373047,
+      "learning_rate": 0.0001,
+      "loss": 0.1605,
+      "step": 1927
+    },
+    {
+      "epoch": 3.1431366155852625,
+      "grad_norm": 0.12134724855422974,
+      "learning_rate": 0.0001,
+      "loss": 0.1594,
+      "step": 1928
+    },
+    {
+      "epoch": 3.14476687316596,
+      "grad_norm": 0.16671213507652283,
+      "learning_rate": 0.0001,
+      "loss": 0.1675,
+      "step": 1929
+    },
+    {
+      "epoch": 3.146397130746658,
+      "grad_norm": 0.7389642596244812,
+      "learning_rate": 0.0001,
+      "loss": 0.1587,
+      "step": 1930
+    },
+    {
+      "epoch": 3.148027388327356,
+      "grad_norm": 0.1277475357055664,
+      "learning_rate": 0.0001,
+      "loss": 0.1467,
+      "step": 1931
+    },
+    {
+      "epoch": 3.1496576459080536,
+      "grad_norm": 0.14921234548091888,
+      "learning_rate": 0.0001,
+      "loss": 0.1521,
+      "step": 1932
+    },
+    {
+      "epoch": 3.1512879034887513,
+      "grad_norm": 0.14741483330726624,
+      "learning_rate": 0.0001,
+      "loss": 0.158,
+      "step": 1933
+    },
+    {
+      "epoch": 3.152918161069449,
+      "grad_norm": 0.14746753871440887,
+      "learning_rate": 0.0001,
+      "loss": 0.1479,
+      "step": 1934
+    },
+    {
+      "epoch": 3.1545484186501467,
+      "grad_norm": 0.13620176911354065,
+      "learning_rate": 0.0001,
+      "loss": 0.1499,
+      "step": 1935
+    },
+    {
+      "epoch": 3.1561786762308444,
+      "grad_norm": 1.9854735136032104,
+      "learning_rate": 0.0001,
+      "loss": 0.1762,
+      "step": 1936
+    },
+    {
+      "epoch": 3.157808933811542,
+      "grad_norm": 0.1485620141029358,
+      "learning_rate": 0.0001,
+      "loss": 0.1523,
+      "step": 1937
+    },
+    {
+      "epoch": 3.15943919139224,
+      "grad_norm": 0.1689678281545639,
+      "learning_rate": 0.0001,
+      "loss": 0.1563,
+      "step": 1938
+    },
+    {
+      "epoch": 3.161069448972938,
+      "grad_norm": 0.14907783269882202,
+      "learning_rate": 0.0001,
+      "loss": 0.1649,
+      "step": 1939
+    },
+    {
+      "epoch": 3.1626997065536355,
+      "grad_norm": 0.15304021537303925,
+      "learning_rate": 0.0001,
+      "loss": 0.1689,
+      "step": 1940
+    },
+    {
+      "epoch": 3.164329964134333,
+      "grad_norm": 0.16715513169765472,
+      "learning_rate": 0.0001,
+      "loss": 0.1601,
+      "step": 1941
+    },
+    {
+      "epoch": 3.165960221715031,
+      "grad_norm": 0.15324808657169342,
+      "learning_rate": 0.0001,
+      "loss": 0.1612,
+      "step": 1942
+    },
+    {
+      "epoch": 3.1675904792957286,
+      "grad_norm": 0.14634795486927032,
+      "learning_rate": 0.0001,
+      "loss": 0.1568,
+      "step": 1943
+    },
+    {
+      "epoch": 3.1692207368764267,
+      "grad_norm": 0.15744540095329285,
+      "learning_rate": 0.0001,
+      "loss": 0.1552,
+      "step": 1944
+    },
+    {
+      "epoch": 3.1708509944571244,
+      "grad_norm": 0.15544916689395905,
+      "learning_rate": 0.0001,
+      "loss": 0.16,
+      "step": 1945
+    },
+    {
+      "epoch": 3.172481252037822,
+      "grad_norm": 0.14175128936767578,
+      "learning_rate": 0.0001,
+      "loss": 0.1639,
+      "step": 1946
+    },
+    {
+      "epoch": 3.1741115096185197,
+      "grad_norm": 0.14051491022109985,
+      "learning_rate": 0.0001,
+      "loss": 0.1379,
+      "step": 1947
+    },
+    {
+      "epoch": 3.1757417671992174,
+      "grad_norm": 0.14642100036144257,
+      "learning_rate": 0.0001,
+      "loss": 0.1535,
+      "step": 1948
+    },
+    {
+      "epoch": 3.177372024779915,
+      "grad_norm": 0.1421380490064621,
+      "learning_rate": 0.0001,
+      "loss": 0.1599,
+      "step": 1949
+    },
+    {
+      "epoch": 3.1790022823606128,
+      "grad_norm": 0.15691342949867249,
+      "learning_rate": 0.0001,
+      "loss": 0.1589,
+      "step": 1950
+    },
+    {
+      "epoch": 3.180632539941311,
+      "grad_norm": 0.14704185724258423,
+      "learning_rate": 0.0001,
+      "loss": 0.1508,
+      "step": 1951
+    },
+    {
+      "epoch": 3.1822627975220086,
+      "grad_norm": 0.15139469504356384,
+      "learning_rate": 0.0001,
+      "loss": 0.151,
+      "step": 1952
+    },
+    {
+      "epoch": 3.1838930551027063,
+      "grad_norm": 0.15881513059139252,
+      "learning_rate": 0.0001,
+      "loss": 0.1604,
+      "step": 1953
+    },
+    {
+      "epoch": 3.185523312683404,
+      "grad_norm": 0.16441747546195984,
+      "learning_rate": 0.0001,
+      "loss": 0.1677,
+      "step": 1954
+    },
+    {
+      "epoch": 3.1871535702641016,
+      "grad_norm": 0.17274102568626404,
+      "learning_rate": 0.0001,
+      "loss": 0.1674,
+      "step": 1955
+    },
+    {
+      "epoch": 3.1887838278447993,
+      "grad_norm": 0.13821010291576385,
+      "learning_rate": 0.0001,
+      "loss": 0.1478,
+      "step": 1956
+    },
+    {
+      "epoch": 3.1904140854254974,
+      "grad_norm": 0.14471560716629028,
+      "learning_rate": 0.0001,
+      "loss": 0.1524,
+      "step": 1957
+    },
+    {
+      "epoch": 3.192044343006195,
+      "grad_norm": 0.14315104484558105,
+      "learning_rate": 0.0001,
+      "loss": 0.1528,
+      "step": 1958
+    },
+    {
+      "epoch": 3.193674600586893,
+      "grad_norm": 0.15579240024089813,
+      "learning_rate": 0.0001,
+      "loss": 0.1575,
+      "step": 1959
+    },
+    {
+      "epoch": 3.1953048581675905,
+      "grad_norm": 0.16074621677398682,
+      "learning_rate": 0.0001,
+      "loss": 0.1498,
+      "step": 1960
+    },
+    {
+      "epoch": 3.196935115748288,
+      "grad_norm": 0.14559932053089142,
+      "learning_rate": 0.0001,
+      "loss": 0.1525,
+      "step": 1961
+    },
+    {
+      "epoch": 3.198565373328986,
+      "grad_norm": 0.146773099899292,
+      "learning_rate": 0.0001,
+      "loss": 0.1568,
+      "step": 1962
+    },
+    {
+      "epoch": 3.2001956309096835,
+      "grad_norm": 0.1405738741159439,
+      "learning_rate": 0.0001,
+      "loss": 0.1553,
+      "step": 1963
+    },
+    {
+      "epoch": 3.2018258884903816,
+      "grad_norm": 0.1669120341539383,
+      "learning_rate": 0.0001,
+      "loss": 0.1652,
+      "step": 1964
+    },
+    {
+      "epoch": 3.2034561460710793,
+      "grad_norm": 0.14730647206306458,
+      "learning_rate": 0.0001,
+      "loss": 0.1532,
+      "step": 1965
+    },
+    {
+      "epoch": 3.205086403651777,
+      "grad_norm": 0.15380671620368958,
+      "learning_rate": 0.0001,
+      "loss": 0.1585,
+      "step": 1966
+    },
+    {
+      "epoch": 3.2067166612324747,
+      "grad_norm": 0.1475609540939331,
+      "learning_rate": 0.0001,
+      "loss": 0.1533,
+      "step": 1967
+    },
+    {
+      "epoch": 3.2083469188131724,
+      "grad_norm": 0.15261127054691315,
+      "learning_rate": 0.0001,
+      "loss": 0.1528,
+      "step": 1968
+    },
+    {
+      "epoch": 3.20997717639387,
+      "grad_norm": 0.13426612317562103,
+      "learning_rate": 0.0001,
+      "loss": 0.1459,
+      "step": 1969
+    },
+    {
+      "epoch": 3.211607433974568,
+      "grad_norm": 0.16383317112922668,
+      "learning_rate": 0.0001,
+      "loss": 0.1457,
+      "step": 1970
+    },
+    {
+      "epoch": 3.213237691555266,
+      "grad_norm": 0.12917453050613403,
+      "learning_rate": 0.0001,
+      "loss": 0.1447,
+      "step": 1971
+    },
+    {
+      "epoch": 3.2148679491359635,
+      "grad_norm": 0.14621089398860931,
+      "learning_rate": 0.0001,
+      "loss": 0.1425,
+      "step": 1972
+    },
+    {
+      "epoch": 3.216498206716661,
+      "grad_norm": 0.14736516773700714,
+      "learning_rate": 0.0001,
+      "loss": 0.1542,
+      "step": 1973
+    },
+    {
+      "epoch": 3.218128464297359,
+      "grad_norm": 0.18349121510982513,
+      "learning_rate": 0.0001,
+      "loss": 0.1581,
+      "step": 1974
+    },
+    {
+      "epoch": 3.2197587218780566,
+      "grad_norm": 0.13845403492450714,
+      "learning_rate": 0.0001,
+      "loss": 0.1531,
+      "step": 1975
+    },
+    {
+      "epoch": 3.2213889794587547,
+      "grad_norm": 0.17356093227863312,
+      "learning_rate": 0.0001,
+      "loss": 0.1499,
+      "step": 1976
+    },
+    {
+      "epoch": 3.2230192370394524,
+      "grad_norm": 0.13243253529071808,
+      "learning_rate": 0.0001,
+      "loss": 0.1551,
+      "step": 1977
+    },
+    {
+      "epoch": 3.22464949462015,
+      "grad_norm": 0.16005612909793854,
+      "learning_rate": 0.0001,
+      "loss": 0.1485,
+      "step": 1978
+    },
+    {
+      "epoch": 3.2262797522008477,
+      "grad_norm": 0.150725856423378,
+      "learning_rate": 0.0001,
+      "loss": 0.1598,
+      "step": 1979
+    },
+    {
+      "epoch": 3.2279100097815454,
+      "grad_norm": 0.1280834674835205,
+      "learning_rate": 0.0001,
+      "loss": 0.1508,
+      "step": 1980
+    },
+    {
+      "epoch": 3.229540267362243,
+      "grad_norm": 0.1567755490541458,
+      "learning_rate": 0.0001,
+      "loss": 0.1561,
+      "step": 1981
+    },
+    {
+      "epoch": 3.231170524942941,
+      "grad_norm": 0.13241910934448242,
+      "learning_rate": 0.0001,
+      "loss": 0.1483,
+      "step": 1982
+    },
+    {
+      "epoch": 3.232800782523639,
+      "grad_norm": 0.1435912549495697,
+      "learning_rate": 0.0001,
+      "loss": 0.1466,
+      "step": 1983
+    },
+    {
+      "epoch": 3.2344310401043366,
+      "grad_norm": 0.1577858179807663,
+      "learning_rate": 0.0001,
+      "loss": 0.1617,
+      "step": 1984
+    },
+    {
+      "epoch": 3.2360612976850343,
+      "grad_norm": 0.14068260788917542,
+      "learning_rate": 0.0001,
+      "loss": 0.1511,
+      "step": 1985
+    },
+    {
+      "epoch": 3.237691555265732,
+      "grad_norm": 0.1652795672416687,
+      "learning_rate": 0.0001,
+      "loss": 0.1744,
+      "step": 1986
+    },
+    {
+      "epoch": 3.2393218128464296,
+      "grad_norm": 0.14117799699306488,
+      "learning_rate": 0.0001,
+      "loss": 0.1525,
+      "step": 1987
+    },
+    {
+      "epoch": 3.2409520704271273,
+      "grad_norm": 0.18951740860939026,
+      "learning_rate": 0.0001,
+      "loss": 0.1491,
+      "step": 1988
+    },
+    {
+      "epoch": 3.2425823280078254,
+      "grad_norm": 0.17044013738632202,
+      "learning_rate": 0.0001,
+      "loss": 0.1416,
+      "step": 1989
+    },
+    {
+      "epoch": 3.244212585588523,
+      "grad_norm": 0.16022269427776337,
+      "learning_rate": 0.0001,
+      "loss": 0.1707,
+      "step": 1990
+    },
+    {
+      "epoch": 3.245842843169221,
+      "grad_norm": 0.17762519419193268,
+      "learning_rate": 0.0001,
+      "loss": 0.1641,
+      "step": 1991
+    },
+    {
+      "epoch": 3.2474731007499185,
+      "grad_norm": 0.15027785301208496,
+      "learning_rate": 0.0001,
+      "loss": 0.1665,
+      "step": 1992
+    },
+    {
+      "epoch": 3.249103358330616,
+      "grad_norm": 0.12501652538776398,
+      "learning_rate": 0.0001,
+      "loss": 0.1388,
+      "step": 1993
+    },
+    {
+      "epoch": 3.250733615911314,
+      "grad_norm": 0.14507539570331573,
+      "learning_rate": 0.0001,
+      "loss": 0.1667,
+      "step": 1994
+    },
+    {
+      "epoch": 3.252363873492012,
+      "grad_norm": 0.12410466372966766,
+      "learning_rate": 0.0001,
+      "loss": 0.152,
+      "step": 1995
+    },
+    {
+      "epoch": 3.2539941310727096,
+      "grad_norm": 0.1520231068134308,
+      "learning_rate": 0.0001,
+      "loss": 0.165,
+      "step": 1996
+    },
+    {
+      "epoch": 3.2556243886534073,
+      "grad_norm": 0.15757714211940765,
+      "learning_rate": 0.0001,
+      "loss": 0.1722,
+      "step": 1997
+    },
+    {
+      "epoch": 3.257254646234105,
+      "grad_norm": 0.19692480564117432,
+      "learning_rate": 0.0001,
+      "loss": 0.1586,
+      "step": 1998
+    },
+    {
+      "epoch": 3.2588849038148027,
+      "grad_norm": 0.1453189253807068,
+      "learning_rate": 0.0001,
+      "loss": 0.1524,
+      "step": 1999
+    },
+    {
+      "epoch": 3.2605151613955003,
+      "grad_norm": 0.14371079206466675,
+      "learning_rate": 0.0001,
+      "loss": 0.1596,
+      "step": 2000
+    },
+    {
+      "epoch": 3.2605151613955003,
+      "eval_loss": 0.19923283159732819,
+      "eval_runtime": 2814.5584,
+      "eval_samples_per_second": 0.67,
+      "eval_steps_per_second": 0.168,
+      "step": 2000
+    },
+    {
+      "epoch": 3.262145418976198,
+      "grad_norm": 0.15101583302021027,
+      "learning_rate": 0.0001,
+      "loss": 0.1628,
+      "step": 2001
+    },
+    {
+      "epoch": 3.263775676556896,
+      "grad_norm": 0.14068691432476044,
+      "learning_rate": 0.0001,
+      "loss": 0.1411,
+      "step": 2002
+    },
+    {
+      "epoch": 3.265405934137594,
+      "grad_norm": 0.1584228277206421,
+      "learning_rate": 0.0001,
+      "loss": 0.1623,
+      "step": 2003
+    },
+    {
+      "epoch": 3.2670361917182915,
+      "grad_norm": 0.14234709739685059,
+      "learning_rate": 0.0001,
+      "loss": 0.1547,
+      "step": 2004
+    },
+    {
+      "epoch": 3.268666449298989,
+      "grad_norm": 0.13787250220775604,
+      "learning_rate": 0.0001,
+      "loss": 0.1578,
+      "step": 2005
+    },
+    {
+      "epoch": 3.270296706879687,
+      "grad_norm": 0.13636542856693268,
+      "learning_rate": 0.0001,
+      "loss": 0.1501,
+      "step": 2006
+    },
+    {
+      "epoch": 3.2719269644603846,
+      "grad_norm": 0.1498252898454666,
+      "learning_rate": 0.0001,
+      "loss": 0.1628,
+      "step": 2007
+    },
+    {
+      "epoch": 3.2735572220410827,
+      "grad_norm": 0.15761011838912964,
+      "learning_rate": 0.0001,
+      "loss": 0.1533,
+      "step": 2008
+    },
+    {
+      "epoch": 3.2751874796217804,
+      "grad_norm": 0.13434697687625885,
+      "learning_rate": 0.0001,
+      "loss": 0.1681,
+      "step": 2009
+    },
+    {
+      "epoch": 3.276817737202478,
+      "grad_norm": 0.1439303308725357,
+      "learning_rate": 0.0001,
+      "loss": 0.1577,
+      "step": 2010
+    },
+    {
+      "epoch": 3.2784479947831757,
+      "grad_norm": 0.1279498040676117,
+      "learning_rate": 0.0001,
+      "loss": 0.15,
+      "step": 2011
+    },
+    {
+      "epoch": 3.2800782523638734,
+      "grad_norm": 0.17215445637702942,
+      "learning_rate": 0.0001,
+      "loss": 0.1632,
+      "step": 2012
+    },
+    {
+      "epoch": 3.281708509944571,
+      "grad_norm": 0.17374494671821594,
+      "learning_rate": 0.0001,
+      "loss": 0.1626,
+      "step": 2013
+    },
+    {
+      "epoch": 3.2833387675252688,
+      "grad_norm": 0.14224795997142792,
+      "learning_rate": 0.0001,
+      "loss": 0.1463,
+      "step": 2014
+    },
+    {
+      "epoch": 3.284969025105967,
+      "grad_norm": 0.13347376883029938,
+      "learning_rate": 0.0001,
+      "loss": 0.1594,
+      "step": 2015
+    },
+    {
+      "epoch": 3.2865992826866646,
+      "grad_norm": 0.12881265580654144,
+      "learning_rate": 0.0001,
+      "loss": 0.1569,
+      "step": 2016
+    },
+    {
+      "epoch": 3.2882295402673622,
+      "grad_norm": 0.14080087840557098,
+      "learning_rate": 0.0001,
+      "loss": 0.1522,
+      "step": 2017
+    },
+    {
+      "epoch": 3.28985979784806,
+      "grad_norm": 0.16095654666423798,
+      "learning_rate": 0.0001,
+      "loss": 0.1608,
+      "step": 2018
+    },
+    {
+      "epoch": 3.2914900554287576,
+      "grad_norm": 0.11454547941684723,
+      "learning_rate": 0.0001,
+      "loss": 0.1356,
+      "step": 2019
+    },
+    {
+      "epoch": 3.2931203130094557,
+      "grad_norm": 0.13725461065769196,
+      "learning_rate": 0.0001,
+      "loss": 0.1502,
+      "step": 2020
+    },
+    {
+      "epoch": 3.2947505705901534,
+      "grad_norm": 0.22002211213111877,
+      "learning_rate": 0.0001,
+      "loss": 0.1513,
+      "step": 2021
+    },
+    {
+      "epoch": 3.296380828170851,
+      "grad_norm": 0.15117506682872772,
+      "learning_rate": 0.0001,
+      "loss": 0.1673,
+      "step": 2022
+    },
+    {
+      "epoch": 3.2980110857515488,
+      "grad_norm": 0.14475668966770172,
+      "learning_rate": 0.0001,
+      "loss": 0.157,
+      "step": 2023
+    },
+    {
+      "epoch": 3.2996413433322465,
+      "grad_norm": 0.13530389964580536,
+      "learning_rate": 0.0001,
+      "loss": 0.1447,
+      "step": 2024
+    },
+    {
+      "epoch": 3.301271600912944,
+      "grad_norm": 0.14106786251068115,
+      "learning_rate": 0.0001,
+      "loss": 0.1565,
+      "step": 2025
+    },
+    {
+      "epoch": 3.302901858493642,
+      "grad_norm": 0.16226093471050262,
+      "learning_rate": 0.0001,
+      "loss": 0.1623,
+      "step": 2026
+    },
+    {
+      "epoch": 3.3045321160743395,
+      "grad_norm": 0.1750541478395462,
+      "learning_rate": 0.0001,
+      "loss": 0.1672,
+      "step": 2027
+    },
+    {
+      "epoch": 3.3061623736550376,
+      "grad_norm": 0.15967412292957306,
+      "learning_rate": 0.0001,
+      "loss": 0.1487,
+      "step": 2028
+    },
+    {
+      "epoch": 3.3077926312357353,
+      "grad_norm": 0.18519139289855957,
+      "learning_rate": 0.0001,
+      "loss": 0.1666,
+      "step": 2029
+    },
+    {
+      "epoch": 3.309422888816433,
+      "grad_norm": 0.14711232483386993,
+      "learning_rate": 0.0001,
+      "loss": 0.1617,
+      "step": 2030
+    },
+    {
+      "epoch": 3.3110531463971307,
+      "grad_norm": 0.14961479604244232,
+      "learning_rate": 0.0001,
+      "loss": 0.149,
+      "step": 2031
+    },
+    {
+      "epoch": 3.3126834039778283,
+      "grad_norm": 0.16532696783542633,
+      "learning_rate": 0.0001,
+      "loss": 0.157,
+      "step": 2032
+    },
+    {
+      "epoch": 3.3143136615585265,
+      "grad_norm": 0.1479751318693161,
+      "learning_rate": 0.0001,
+      "loss": 0.164,
+      "step": 2033
+    },
+    {
+      "epoch": 3.315943919139224,
+      "grad_norm": 0.16249758005142212,
+      "learning_rate": 0.0001,
+      "loss": 0.1607,
+      "step": 2034
+    },
+    {
+      "epoch": 3.317574176719922,
+      "grad_norm": 0.15464435517787933,
+      "learning_rate": 0.0001,
+      "loss": 0.1514,
+      "step": 2035
+    },
+    {
+      "epoch": 3.3192044343006195,
+      "grad_norm": 0.12793324887752533,
+      "learning_rate": 0.0001,
+      "loss": 0.1558,
+      "step": 2036
+    },
+    {
+      "epoch": 3.320834691881317,
+      "grad_norm": 0.1387512981891632,
+      "learning_rate": 0.0001,
+      "loss": 0.1568,
+      "step": 2037
+    },
+    {
+      "epoch": 3.322464949462015,
+      "grad_norm": 0.14236465096473694,
+      "learning_rate": 0.0001,
+      "loss": 0.1571,
+      "step": 2038
+    },
+    {
+      "epoch": 3.3240952070427126,
+      "grad_norm": 0.12923693656921387,
+      "learning_rate": 0.0001,
+      "loss": 0.1368,
+      "step": 2039
+    },
+    {
+      "epoch": 3.3257254646234107,
+      "grad_norm": 0.14453841745853424,
+      "learning_rate": 0.0001,
+      "loss": 0.1476,
+      "step": 2040
+    },
+    {
+      "epoch": 3.3273557222041084,
+      "grad_norm": 0.13822996616363525,
+      "learning_rate": 0.0001,
+      "loss": 0.1458,
+      "step": 2041
+    },
+    {
+      "epoch": 3.328985979784806,
+      "grad_norm": 0.1528143286705017,
+      "learning_rate": 0.0001,
+      "loss": 0.1694,
+      "step": 2042
+    },
+    {
+      "epoch": 3.3306162373655037,
+      "grad_norm": 0.14497984945774078,
+      "learning_rate": 0.0001,
+      "loss": 0.1446,
+      "step": 2043
+    },
+    {
+      "epoch": 3.3322464949462014,
+      "grad_norm": 0.16992256045341492,
+      "learning_rate": 0.0001,
+      "loss": 0.1561,
+      "step": 2044
+    },
+    {
+      "epoch": 3.333876752526899,
+      "grad_norm": 0.16492198407649994,
+      "learning_rate": 0.0001,
+      "loss": 0.1653,
+      "step": 2045
+    },
+    {
+      "epoch": 3.335507010107597,
+      "grad_norm": 0.14697159826755524,
+      "learning_rate": 0.0001,
+      "loss": 0.145,
+      "step": 2046
+    },
+    {
+      "epoch": 3.337137267688295,
+      "grad_norm": 0.1527179777622223,
+      "learning_rate": 0.0001,
+      "loss": 0.154,
+      "step": 2047
+    },
+    {
+      "epoch": 3.3387675252689926,
+      "grad_norm": 0.14120477437973022,
+      "learning_rate": 0.0001,
+      "loss": 0.1583,
+      "step": 2048
+    },
+    {
+      "epoch": 3.3403977828496902,
+      "grad_norm": 0.13726529479026794,
+      "learning_rate": 0.0001,
+      "loss": 0.1491,
+      "step": 2049
+    },
+    {
+      "epoch": 3.342028040430388,
+      "grad_norm": 0.13249808549880981,
+      "learning_rate": 0.0001,
+      "loss": 0.1615,
+      "step": 2050
+    },
+    {
+      "epoch": 3.3436582980110856,
+      "grad_norm": 0.14583012461662292,
+      "learning_rate": 0.0001,
+      "loss": 0.1515,
+      "step": 2051
+    },
+    {
+      "epoch": 3.3452885555917833,
+      "grad_norm": 0.12669849395751953,
+      "learning_rate": 0.0001,
+      "loss": 0.136,
+      "step": 2052
+    },
+    {
+      "epoch": 3.3469188131724814,
+      "grad_norm": 0.15459409356117249,
+      "learning_rate": 0.0001,
+      "loss": 0.1408,
+      "step": 2053
+    },
+    {
+      "epoch": 3.348549070753179,
+      "grad_norm": 0.15946894884109497,
+      "learning_rate": 0.0001,
+      "loss": 0.1532,
+      "step": 2054
+    },
+    {
+      "epoch": 3.3501793283338768,
+      "grad_norm": 0.15201634168624878,
+      "learning_rate": 0.0001,
+      "loss": 0.1593,
+      "step": 2055
+    },
+    {
+      "epoch": 3.3518095859145745,
+      "grad_norm": 0.15227296948432922,
+      "learning_rate": 0.0001,
+      "loss": 0.1444,
+      "step": 2056
+    },
+    {
+      "epoch": 3.353439843495272,
+      "grad_norm": 0.15703243017196655,
+      "learning_rate": 0.0001,
+      "loss": 0.1547,
+      "step": 2057
+    },
+    {
+      "epoch": 3.35507010107597,
+      "grad_norm": 0.14738823473453522,
+      "learning_rate": 0.0001,
+      "loss": 0.1587,
+      "step": 2058
+    },
+    {
+      "epoch": 3.356700358656668,
+      "grad_norm": 0.1499371975660324,
+      "learning_rate": 0.0001,
+      "loss": 0.1602,
+      "step": 2059
+    },
+    {
+      "epoch": 3.3583306162373656,
+      "grad_norm": 0.15711835026741028,
+      "learning_rate": 0.0001,
+      "loss": 0.1648,
+      "step": 2060
+    },
+    {
+      "epoch": 3.3599608738180633,
+      "grad_norm": 0.13238079845905304,
+      "learning_rate": 0.0001,
+      "loss": 0.1369,
+      "step": 2061
+    },
+    {
+      "epoch": 3.361591131398761,
+      "grad_norm": 0.13998901844024658,
+      "learning_rate": 0.0001,
+      "loss": 0.1578,
+      "step": 2062
+    },
+    {
+      "epoch": 3.3632213889794587,
+      "grad_norm": 0.13403183221817017,
+      "learning_rate": 0.0001,
+      "loss": 0.1491,
+      "step": 2063
+    },
+    {
+      "epoch": 3.3648516465601563,
+      "grad_norm": 0.12701204419136047,
+      "learning_rate": 0.0001,
+      "loss": 0.1428,
+      "step": 2064
+    },
+    {
+      "epoch": 3.366481904140854,
+      "grad_norm": 0.13955272734165192,
+      "learning_rate": 0.0001,
+      "loss": 0.1618,
+      "step": 2065
+    },
+    {
+      "epoch": 3.368112161721552,
+      "grad_norm": 0.1396985650062561,
+      "learning_rate": 0.0001,
+      "loss": 0.1558,
+      "step": 2066
+    },
+    {
+      "epoch": 3.36974241930225,
+      "grad_norm": 0.1466887891292572,
+      "learning_rate": 0.0001,
+      "loss": 0.1565,
+      "step": 2067
+    },
+    {
+      "epoch": 3.3713726768829475,
+      "grad_norm": 0.148866206407547,
+      "learning_rate": 0.0001,
+      "loss": 0.1525,
+      "step": 2068
+    },
+    {
+      "epoch": 3.373002934463645,
+      "grad_norm": 0.16650554537773132,
+      "learning_rate": 0.0001,
+      "loss": 0.1728,
+      "step": 2069
+    },
+    {
+      "epoch": 3.374633192044343,
+      "grad_norm": 0.14089016616344452,
+      "learning_rate": 0.0001,
+      "loss": 0.1527,
+      "step": 2070
+    },
+    {
+      "epoch": 3.376263449625041,
+      "grad_norm": 0.15425576269626617,
+      "learning_rate": 0.0001,
+      "loss": 0.1612,
+      "step": 2071
+    },
+    {
+      "epoch": 3.3778937072057387,
+      "grad_norm": 0.2497289627790451,
+      "learning_rate": 0.0001,
+      "loss": 0.1661,
+      "step": 2072
+    },
+    {
+      "epoch": 3.3795239647864364,
+      "grad_norm": 0.15727926790714264,
+      "learning_rate": 0.0001,
+      "loss": 0.1641,
+      "step": 2073
+    },
+    {
+      "epoch": 3.381154222367134,
+      "grad_norm": 0.13486291468143463,
+      "learning_rate": 0.0001,
+      "loss": 0.1435,
+      "step": 2074
+    },
+    {
+      "epoch": 3.3827844799478317,
+      "grad_norm": 0.13273045420646667,
+      "learning_rate": 0.0001,
+      "loss": 0.1599,
+      "step": 2075
+    },
+    {
+      "epoch": 3.3844147375285294,
+      "grad_norm": 0.14401297271251678,
+      "learning_rate": 0.0001,
+      "loss": 0.162,
+      "step": 2076
+    },
+    {
+      "epoch": 3.386044995109227,
+      "grad_norm": 0.1639464646577835,
+      "learning_rate": 0.0001,
+      "loss": 0.1749,
+      "step": 2077
+    },
+    {
+      "epoch": 3.387675252689925,
+      "grad_norm": 0.14354901015758514,
+      "learning_rate": 0.0001,
+      "loss": 0.1571,
+      "step": 2078
+    },
+    {
+      "epoch": 3.389305510270623,
+      "grad_norm": 0.1772812008857727,
+      "learning_rate": 0.0001,
+      "loss": 0.164,
+      "step": 2079
+    },
+    {
+      "epoch": 3.3909357678513206,
+      "grad_norm": 0.1357707381248474,
+      "learning_rate": 0.0001,
+      "loss": 0.1572,
+      "step": 2080
+    },
+    {
+      "epoch": 3.3925660254320182,
+      "grad_norm": 0.15610982477664948,
+      "learning_rate": 0.0001,
+      "loss": 0.1718,
+      "step": 2081
+    },
+    {
+      "epoch": 3.394196283012716,
+      "grad_norm": 0.1462925523519516,
+      "learning_rate": 0.0001,
+      "loss": 0.1458,
+      "step": 2082
+    },
+    {
+      "epoch": 3.3958265405934136,
+      "grad_norm": 0.1445331573486328,
+      "learning_rate": 0.0001,
+      "loss": 0.1614,
+      "step": 2083
+    },
+    {
+      "epoch": 3.3974567981741117,
+      "grad_norm": 0.13011384010314941,
+      "learning_rate": 0.0001,
+      "loss": 0.1629,
+      "step": 2084
+    },
+    {
+      "epoch": 3.3990870557548094,
+      "grad_norm": 0.13652506470680237,
+      "learning_rate": 0.0001,
+      "loss": 0.145,
+      "step": 2085
+    },
+    {
+      "epoch": 3.400717313335507,
+      "grad_norm": 0.15374945104122162,
+      "learning_rate": 0.0001,
+      "loss": 0.1607,
+      "step": 2086
+    },
+    {
+      "epoch": 3.4023475709162048,
+      "grad_norm": 0.17759419977664948,
+      "learning_rate": 0.0001,
+      "loss": 0.1831,
+      "step": 2087
+    },
+    {
+      "epoch": 3.4039778284969024,
+      "grad_norm": 0.13576067984104156,
+      "learning_rate": 0.0001,
+      "loss": 0.1609,
+      "step": 2088
+    },
+    {
+      "epoch": 3.4056080860776,
+      "grad_norm": 0.14758118987083435,
+      "learning_rate": 0.0001,
+      "loss": 0.1648,
+      "step": 2089
+    },
+    {
+      "epoch": 3.407238343658298,
+      "grad_norm": 0.14185144007205963,
+      "learning_rate": 0.0001,
+      "loss": 0.1553,
+      "step": 2090
+    },
+    {
+      "epoch": 3.408868601238996,
+      "grad_norm": 0.15567590296268463,
+      "learning_rate": 0.0001,
+      "loss": 0.162,
+      "step": 2091
+    },
+    {
+      "epoch": 3.4104988588196936,
+      "grad_norm": 0.14208033680915833,
+      "learning_rate": 0.0001,
+      "loss": 0.1648,
+      "step": 2092
+    },
+    {
+      "epoch": 3.4121291164003913,
+      "grad_norm": 0.150962695479393,
+      "learning_rate": 0.0001,
+      "loss": 0.1659,
+      "step": 2093
+    },
+    {
+      "epoch": 3.413759373981089,
+      "grad_norm": 0.18408554792404175,
+      "learning_rate": 0.0001,
+      "loss": 0.1564,
+      "step": 2094
+    },
+    {
+      "epoch": 3.4153896315617867,
+      "grad_norm": 0.1329030841588974,
+      "learning_rate": 0.0001,
+      "loss": 0.1434,
+      "step": 2095
+    },
+    {
+      "epoch": 3.4170198891424843,
+      "grad_norm": 0.19760195910930634,
+      "learning_rate": 0.0001,
+      "loss": 0.1616,
+      "step": 2096
+    },
+    {
+      "epoch": 3.4186501467231825,
+      "grad_norm": 0.13045673072338104,
+      "learning_rate": 0.0001,
+      "loss": 0.1482,
+      "step": 2097
+    },
+    {
+      "epoch": 3.42028040430388,
+      "grad_norm": 0.13792401552200317,
+      "learning_rate": 0.0001,
+      "loss": 0.1491,
+      "step": 2098
+    },
+    {
+      "epoch": 3.421910661884578,
+      "grad_norm": 0.15530133247375488,
+      "learning_rate": 0.0001,
+      "loss": 0.1569,
+      "step": 2099
+    },
+    {
+      "epoch": 3.4235409194652755,
+      "grad_norm": 0.16320976614952087,
+      "learning_rate": 0.0001,
+      "loss": 0.1504,
+      "step": 2100
+    },
+    {
+      "epoch": 3.425171177045973,
+      "grad_norm": 0.16057944297790527,
+      "learning_rate": 0.0001,
+      "loss": 0.1575,
+      "step": 2101
+    },
+    {
+      "epoch": 3.426801434626671,
+      "grad_norm": 0.16377048194408417,
+      "learning_rate": 0.0001,
+      "loss": 0.1587,
+      "step": 2102
+    },
+    {
+      "epoch": 3.4284316922073685,
+      "grad_norm": 0.14860613644123077,
+      "learning_rate": 0.0001,
+      "loss": 0.1666,
+      "step": 2103
+    },
+    {
+      "epoch": 3.4300619497880667,
+      "grad_norm": 0.1717861294746399,
+      "learning_rate": 0.0001,
+      "loss": 0.1727,
+      "step": 2104
+    },
+    {
+      "epoch": 3.4316922073687643,
+      "grad_norm": 0.18878856301307678,
+      "learning_rate": 0.0001,
+      "loss": 0.1744,
+      "step": 2105
+    },
+    {
+      "epoch": 3.433322464949462,
+      "grad_norm": 0.22134870290756226,
+      "learning_rate": 0.0001,
+      "loss": 0.1502,
+      "step": 2106
+    },
+    {
+      "epoch": 3.4349527225301597,
+      "grad_norm": 0.12796226143836975,
+      "learning_rate": 0.0001,
+      "loss": 0.1472,
+      "step": 2107
+    },
+    {
+      "epoch": 3.4365829801108574,
+      "grad_norm": 0.17766544222831726,
+      "learning_rate": 0.0001,
+      "loss": 0.1631,
+      "step": 2108
+    },
+    {
+      "epoch": 3.4382132376915555,
+      "grad_norm": 0.16302259266376495,
+      "learning_rate": 0.0001,
+      "loss": 0.154,
+      "step": 2109
+    },
+    {
+      "epoch": 3.439843495272253,
+      "grad_norm": 0.12639376521110535,
+      "learning_rate": 0.0001,
+      "loss": 0.1595,
+      "step": 2110
+    },
+    {
+      "epoch": 3.441473752852951,
+      "grad_norm": 0.13385580480098724,
+      "learning_rate": 0.0001,
+      "loss": 0.148,
+      "step": 2111
+    },
+    {
+      "epoch": 3.4431040104336486,
+      "grad_norm": 0.16364267468452454,
+      "learning_rate": 0.0001,
+      "loss": 0.1521,
+      "step": 2112
+    },
+    {
+      "epoch": 3.4447342680143462,
+      "grad_norm": 0.1307305246591568,
+      "learning_rate": 0.0001,
+      "loss": 0.1623,
+      "step": 2113
+    },
+    {
+      "epoch": 3.446364525595044,
+      "grad_norm": 0.1305004358291626,
+      "learning_rate": 0.0001,
+      "loss": 0.1428,
+      "step": 2114
+    },
+    {
+      "epoch": 3.4479947831757416,
+      "grad_norm": 0.1550890952348709,
+      "learning_rate": 0.0001,
+      "loss": 0.1521,
+      "step": 2115
+    },
+    {
+      "epoch": 3.4496250407564393,
+      "grad_norm": 0.14810912311077118,
+      "learning_rate": 0.0001,
+      "loss": 0.1635,
+      "step": 2116
+    },
+    {
+      "epoch": 3.4512552983371374,
+      "grad_norm": 0.15610523521900177,
+      "learning_rate": 0.0001,
+      "loss": 0.1619,
+      "step": 2117
+    },
+    {
+      "epoch": 3.452885555917835,
+      "grad_norm": 0.1532841920852661,
+      "learning_rate": 0.0001,
+      "loss": 0.1609,
+      "step": 2118
+    },
+    {
+      "epoch": 3.4545158134985328,
+      "grad_norm": 0.14741961658000946,
+      "learning_rate": 0.0001,
+      "loss": 0.167,
+      "step": 2119
+    },
+    {
+      "epoch": 3.4561460710792304,
+      "grad_norm": 0.13078215718269348,
+      "learning_rate": 0.0001,
+      "loss": 0.1428,
+      "step": 2120
+    },
+    {
+      "epoch": 3.457776328659928,
+      "grad_norm": 0.18994919955730438,
+      "learning_rate": 0.0001,
+      "loss": 0.1548,
+      "step": 2121
+    },
+    {
+      "epoch": 3.4594065862406262,
+      "grad_norm": 0.12827709317207336,
+      "learning_rate": 0.0001,
+      "loss": 0.1586,
+      "step": 2122
+    },
+    {
+      "epoch": 3.461036843821324,
+      "grad_norm": 0.15168218314647675,
+      "learning_rate": 0.0001,
+      "loss": 0.1641,
+      "step": 2123
+    },
+    {
+      "epoch": 3.4626671014020216,
+      "grad_norm": 0.13048698008060455,
+      "learning_rate": 0.0001,
+      "loss": 0.1566,
+      "step": 2124
+    },
+    {
+      "epoch": 3.4642973589827193,
+      "grad_norm": 0.13012981414794922,
+      "learning_rate": 0.0001,
+      "loss": 0.1523,
+      "step": 2125
+    },
+    {
+      "epoch": 3.465927616563417,
+      "grad_norm": 0.14694590866565704,
+      "learning_rate": 0.0001,
+      "loss": 0.1473,
+      "step": 2126
+    },
+    {
+      "epoch": 3.4675578741441146,
+      "grad_norm": 0.13174067437648773,
+      "learning_rate": 0.0001,
+      "loss": 0.1531,
+      "step": 2127
+    },
+    {
+      "epoch": 3.4691881317248123,
+      "grad_norm": 0.18024425208568573,
+      "learning_rate": 0.0001,
+      "loss": 0.1624,
+      "step": 2128
+    },
+    {
+      "epoch": 3.4708183893055105,
+      "grad_norm": 0.14632080495357513,
+      "learning_rate": 0.0001,
+      "loss": 0.1709,
+      "step": 2129
+    },
+    {
+      "epoch": 3.472448646886208,
+      "grad_norm": 0.13840989768505096,
+      "learning_rate": 0.0001,
+      "loss": 0.1576,
+      "step": 2130
+    },
+    {
+      "epoch": 3.474078904466906,
+      "grad_norm": 0.1256123036146164,
+      "learning_rate": 0.0001,
+      "loss": 0.1486,
+      "step": 2131
+    },
+    {
+      "epoch": 3.4757091620476035,
+      "grad_norm": 0.19333600997924805,
+      "learning_rate": 0.0001,
+      "loss": 0.1526,
+      "step": 2132
+    },
+    {
+      "epoch": 3.477339419628301,
+      "grad_norm": 0.1740863025188446,
+      "learning_rate": 0.0001,
+      "loss": 0.162,
+      "step": 2133
+    },
+    {
+      "epoch": 3.478969677208999,
+      "grad_norm": 0.17544031143188477,
+      "learning_rate": 0.0001,
+      "loss": 0.1743,
+      "step": 2134
+    },
+    {
+      "epoch": 3.480599934789697,
+      "grad_norm": 0.13726355135440826,
+      "learning_rate": 0.0001,
+      "loss": 0.146,
+      "step": 2135
+    },
+    {
+      "epoch": 3.4822301923703947,
+      "grad_norm": 0.13387274742126465,
+      "learning_rate": 0.0001,
+      "loss": 0.1444,
+      "step": 2136
+    },
+    {
+      "epoch": 3.4838604499510923,
+      "grad_norm": 0.1480950564146042,
+      "learning_rate": 0.0001,
+      "loss": 0.1656,
+      "step": 2137
+    },
+    {
+      "epoch": 3.48549070753179,
+      "grad_norm": 0.14710885286331177,
+      "learning_rate": 0.0001,
+      "loss": 0.1501,
+      "step": 2138
+    },
+    {
+      "epoch": 3.4871209651124877,
+      "grad_norm": 0.14011640846729279,
+      "learning_rate": 0.0001,
+      "loss": 0.1571,
+      "step": 2139
+    },
+    {
+      "epoch": 3.4887512226931854,
+      "grad_norm": 0.15596067905426025,
+      "learning_rate": 0.0001,
+      "loss": 0.1686,
+      "step": 2140
+    },
+    {
+      "epoch": 3.490381480273883,
+      "grad_norm": 0.14129222929477692,
+      "learning_rate": 0.0001,
+      "loss": 0.1597,
+      "step": 2141
+    },
+    {
+      "epoch": 3.492011737854581,
+      "grad_norm": 0.1319301277399063,
+      "learning_rate": 0.0001,
+      "loss": 0.1619,
+      "step": 2142
+    },
+    {
+      "epoch": 3.493641995435279,
+      "grad_norm": 0.16085407137870789,
+      "learning_rate": 0.0001,
+      "loss": 0.1542,
+      "step": 2143
+    },
+    {
+      "epoch": 3.4952722530159765,
+      "grad_norm": 0.1328543722629547,
+      "learning_rate": 0.0001,
+      "loss": 0.1501,
+      "step": 2144
+    },
+    {
+      "epoch": 3.4969025105966742,
+      "grad_norm": 0.14981688559055328,
+      "learning_rate": 0.0001,
+      "loss": 0.1601,
+      "step": 2145
+    },
+    {
+      "epoch": 3.498532768177372,
+      "grad_norm": 0.14664553105831146,
+      "learning_rate": 0.0001,
+      "loss": 0.1583,
+      "step": 2146
+    },
+    {
+      "epoch": 3.50016302575807,
+      "grad_norm": 0.13246874511241913,
+      "learning_rate": 0.0001,
+      "loss": 0.1491,
+      "step": 2147
+    },
+    {
+      "epoch": 3.5017932833387677,
+      "grad_norm": 0.15695396065711975,
+      "learning_rate": 0.0001,
+      "loss": 0.1793,
+      "step": 2148
+    },
+    {
+      "epoch": 3.5034235409194654,
+      "grad_norm": 0.12087570875883102,
+      "learning_rate": 0.0001,
+      "loss": 0.1396,
+      "step": 2149
+    },
+    {
+      "epoch": 3.505053798500163,
+      "grad_norm": 0.13022390007972717,
+      "learning_rate": 0.0001,
+      "loss": 0.1577,
+      "step": 2150
+    },
+    {
+      "epoch": 3.5066840560808608,
+      "grad_norm": 0.15572933852672577,
+      "learning_rate": 0.0001,
+      "loss": 0.1532,
+      "step": 2151
+    },
+    {
+      "epoch": 3.5083143136615584,
+      "grad_norm": 0.17626436054706573,
+      "learning_rate": 0.0001,
+      "loss": 0.1604,
+      "step": 2152
+    },
+    {
+      "epoch": 3.509944571242256,
+      "grad_norm": 0.14553359150886536,
+      "learning_rate": 0.0001,
+      "loss": 0.163,
+      "step": 2153
+    },
+    {
+      "epoch": 3.511574828822954,
+      "grad_norm": 0.17382773756980896,
+      "learning_rate": 0.0001,
+      "loss": 0.1641,
+      "step": 2154
+    },
+    {
+      "epoch": 3.513205086403652,
+      "grad_norm": 0.12575078010559082,
+      "learning_rate": 0.0001,
+      "loss": 0.1493,
+      "step": 2155
+    },
+    {
+      "epoch": 3.5148353439843496,
+      "grad_norm": 0.1463012993335724,
+      "learning_rate": 0.0001,
+      "loss": 0.159,
+      "step": 2156
+    },
+    {
+      "epoch": 3.5164656015650473,
+      "grad_norm": 0.16834314167499542,
+      "learning_rate": 0.0001,
+      "loss": 0.1659,
+      "step": 2157
+    },
+    {
+      "epoch": 3.518095859145745,
+      "grad_norm": 0.15499548614025116,
+      "learning_rate": 0.0001,
+      "loss": 0.1584,
+      "step": 2158
+    },
+    {
+      "epoch": 3.5197261167264426,
+      "grad_norm": 0.13314367830753326,
+      "learning_rate": 0.0001,
+      "loss": 0.1515,
+      "step": 2159
+    },
+    {
+      "epoch": 3.5213563743071408,
+      "grad_norm": 0.1413407176733017,
+      "learning_rate": 0.0001,
+      "loss": 0.1614,
+      "step": 2160
+    },
+    {
+      "epoch": 3.5229866318878384,
+      "grad_norm": 0.14445054531097412,
+      "learning_rate": 0.0001,
+      "loss": 0.1595,
+      "step": 2161
+    },
+    {
+      "epoch": 3.524616889468536,
+      "grad_norm": 0.14501655101776123,
+      "learning_rate": 0.0001,
+      "loss": 0.1607,
+      "step": 2162
+    },
+    {
+      "epoch": 3.526247147049234,
+      "grad_norm": 0.14677855372428894,
+      "learning_rate": 0.0001,
+      "loss": 0.1588,
+      "step": 2163
+    },
+    {
+      "epoch": 3.5278774046299315,
+      "grad_norm": 0.13915497064590454,
+      "learning_rate": 0.0001,
+      "loss": 0.164,
+      "step": 2164
+    },
+    {
+      "epoch": 3.529507662210629,
+      "grad_norm": 0.12632952630519867,
+      "learning_rate": 0.0001,
+      "loss": 0.1442,
+      "step": 2165
+    },
+    {
+      "epoch": 3.531137919791327,
+      "grad_norm": 0.13267867267131805,
+      "learning_rate": 0.0001,
+      "loss": 0.159,
+      "step": 2166
+    },
+    {
+      "epoch": 3.5327681773720245,
+      "grad_norm": 0.12538422644138336,
+      "learning_rate": 0.0001,
+      "loss": 0.1409,
+      "step": 2167
+    },
+    {
+      "epoch": 3.5343984349527227,
+      "grad_norm": 0.13974052667617798,
+      "learning_rate": 0.0001,
+      "loss": 0.1461,
+      "step": 2168
+    },
+    {
+      "epoch": 3.5360286925334203,
+      "grad_norm": 0.1575179398059845,
+      "learning_rate": 0.0001,
+      "loss": 0.1596,
+      "step": 2169
+    },
+    {
+      "epoch": 3.537658950114118,
+      "grad_norm": 0.20956085622310638,
+      "learning_rate": 0.0001,
+      "loss": 0.1583,
+      "step": 2170
+    },
+    {
+      "epoch": 3.5392892076948157,
+      "grad_norm": 0.13932162523269653,
+      "learning_rate": 0.0001,
+      "loss": 0.1493,
+      "step": 2171
+    },
+    {
+      "epoch": 3.5409194652755134,
+      "grad_norm": 0.16261732578277588,
+      "learning_rate": 0.0001,
+      "loss": 0.1603,
+      "step": 2172
+    },
+    {
+      "epoch": 3.5425497228562115,
+      "grad_norm": 0.15178941190242767,
+      "learning_rate": 0.0001,
+      "loss": 0.1601,
+      "step": 2173
+    },
+    {
+      "epoch": 3.544179980436909,
+      "grad_norm": 0.17579567432403564,
+      "learning_rate": 0.0001,
+      "loss": 0.1531,
+      "step": 2174
+    },
+    {
+      "epoch": 3.545810238017607,
+      "grad_norm": 0.1470760852098465,
+      "learning_rate": 0.0001,
+      "loss": 0.1498,
+      "step": 2175
+    },
+    {
+      "epoch": 3.5474404955983045,
+      "grad_norm": 0.1457168012857437,
+      "learning_rate": 0.0001,
+      "loss": 0.1667,
+      "step": 2176
+    },
+    {
+      "epoch": 3.5490707531790022,
+      "grad_norm": 0.1363757997751236,
+      "learning_rate": 0.0001,
+      "loss": 0.1538,
+      "step": 2177
+    },
+    {
+      "epoch": 3.5507010107597,
+      "grad_norm": 0.15826475620269775,
+      "learning_rate": 0.0001,
+      "loss": 0.1643,
+      "step": 2178
+    },
+    {
+      "epoch": 3.5523312683403976,
+      "grad_norm": 0.13432374596595764,
+      "learning_rate": 0.0001,
+      "loss": 0.1487,
+      "step": 2179
+    },
+    {
+      "epoch": 3.5539615259210953,
+      "grad_norm": 0.1316751092672348,
+      "learning_rate": 0.0001,
+      "loss": 0.1515,
+      "step": 2180
+    },
+    {
+      "epoch": 3.5555917835017934,
+      "grad_norm": 0.14341945946216583,
+      "learning_rate": 0.0001,
+      "loss": 0.1507,
+      "step": 2181
+    },
+    {
+      "epoch": 3.557222041082491,
+      "grad_norm": 0.15055926144123077,
+      "learning_rate": 0.0001,
+      "loss": 0.164,
+      "step": 2182
+    },
+    {
+      "epoch": 3.5588522986631888,
+      "grad_norm": 0.18501447141170502,
+      "learning_rate": 0.0001,
+      "loss": 0.1526,
+      "step": 2183
+    },
+    {
+      "epoch": 3.5604825562438864,
+      "grad_norm": 0.13240905106067657,
+      "learning_rate": 0.0001,
+      "loss": 0.1487,
+      "step": 2184
+    },
+    {
+      "epoch": 3.5621128138245846,
+      "grad_norm": 0.15811198949813843,
+      "learning_rate": 0.0001,
+      "loss": 0.1659,
+      "step": 2185
+    },
+    {
+      "epoch": 3.5637430714052822,
+      "grad_norm": 0.17472024261951447,
+      "learning_rate": 0.0001,
+      "loss": 0.1605,
+      "step": 2186
+    },
+    {
+      "epoch": 3.56537332898598,
+      "grad_norm": 0.1416904181241989,
+      "learning_rate": 0.0001,
+      "loss": 0.1563,
+      "step": 2187
+    },
+    {
+      "epoch": 3.5670035865666776,
+      "grad_norm": 0.18020367622375488,
+      "learning_rate": 0.0001,
+      "loss": 0.1733,
+      "step": 2188
+    },
+    {
+      "epoch": 3.5686338441473753,
+      "grad_norm": 0.1482088267803192,
+      "learning_rate": 0.0001,
+      "loss": 0.1619,
+      "step": 2189
+    },
+    {
+      "epoch": 3.570264101728073,
+      "grad_norm": 0.2286827713251114,
+      "learning_rate": 0.0001,
+      "loss": 0.1745,
+      "step": 2190
+    },
+    {
+      "epoch": 3.5718943593087706,
+      "grad_norm": 0.15454155206680298,
+      "learning_rate": 0.0001,
+      "loss": 0.1571,
+      "step": 2191
+    },
+    {
+      "epoch": 3.5735246168894683,
+      "grad_norm": 0.15183959901332855,
+      "learning_rate": 0.0001,
+      "loss": 0.1604,
+      "step": 2192
+    },
+    {
+      "epoch": 3.5751548744701664,
+      "grad_norm": 0.1492152214050293,
+      "learning_rate": 0.0001,
+      "loss": 0.1735,
+      "step": 2193
+    },
+    {
+      "epoch": 3.576785132050864,
+      "grad_norm": 0.15264061093330383,
+      "learning_rate": 0.0001,
+      "loss": 0.1541,
+      "step": 2194
+    },
+    {
+      "epoch": 3.578415389631562,
+      "grad_norm": 0.13549228012561798,
+      "learning_rate": 0.0001,
+      "loss": 0.1521,
+      "step": 2195
+    },
+    {
+      "epoch": 3.5800456472122595,
+      "grad_norm": 0.1467984914779663,
+      "learning_rate": 0.0001,
+      "loss": 0.1607,
+      "step": 2196
+    },
+    {
+      "epoch": 3.581675904792957,
+      "grad_norm": 0.13193894922733307,
+      "learning_rate": 0.0001,
+      "loss": 0.158,
+      "step": 2197
+    },
+    {
+      "epoch": 3.5833061623736553,
+      "grad_norm": 0.1380847692489624,
+      "learning_rate": 0.0001,
+      "loss": 0.1567,
+      "step": 2198
+    },
+    {
+      "epoch": 3.584936419954353,
+      "grad_norm": 0.12376043945550919,
+      "learning_rate": 0.0001,
+      "loss": 0.1642,
+      "step": 2199
+    },
+    {
+      "epoch": 3.5865666775350507,
+      "grad_norm": 0.1520168036222458,
+      "learning_rate": 0.0001,
+      "loss": 0.1668,
+      "step": 2200
+    },
+    {
+      "epoch": 3.5881969351157483,
+      "grad_norm": 0.15130695700645447,
+      "learning_rate": 0.0001,
+      "loss": 0.1607,
+      "step": 2201
+    },
+    {
+      "epoch": 3.589827192696446,
+      "grad_norm": 0.21414104104042053,
+      "learning_rate": 0.0001,
+      "loss": 0.1748,
+      "step": 2202
+    },
+    {
+      "epoch": 3.5914574502771437,
+      "grad_norm": 0.13339370489120483,
+      "learning_rate": 0.0001,
+      "loss": 0.1569,
+      "step": 2203
+    },
+    {
+      "epoch": 3.5930877078578414,
+      "grad_norm": 0.14717626571655273,
+      "learning_rate": 0.0001,
+      "loss": 0.1485,
+      "step": 2204
+    },
+    {
+      "epoch": 3.594717965438539,
+      "grad_norm": 0.14454171061515808,
+      "learning_rate": 0.0001,
+      "loss": 0.1561,
+      "step": 2205
+    },
+    {
+      "epoch": 3.596348223019237,
+      "grad_norm": 0.1337466984987259,
+      "learning_rate": 0.0001,
+      "loss": 0.1548,
+      "step": 2206
+    },
+    {
+      "epoch": 3.597978480599935,
+      "grad_norm": 0.13074032962322235,
+      "learning_rate": 0.0001,
+      "loss": 0.1602,
+      "step": 2207
+    },
+    {
+      "epoch": 3.5996087381806325,
+      "grad_norm": 0.1627005934715271,
+      "learning_rate": 0.0001,
+      "loss": 0.1582,
+      "step": 2208
+    },
+    {
+      "epoch": 3.60123899576133,
+      "grad_norm": 0.16543006896972656,
+      "learning_rate": 0.0001,
+      "loss": 0.1699,
+      "step": 2209
+    },
+    {
+      "epoch": 3.602869253342028,
+      "grad_norm": 0.15218278765678406,
+      "learning_rate": 0.0001,
+      "loss": 0.164,
+      "step": 2210
+    },
+    {
+      "epoch": 3.604499510922726,
+      "grad_norm": 0.15008680522441864,
+      "learning_rate": 0.0001,
+      "loss": 0.1567,
+      "step": 2211
+    },
+    {
+      "epoch": 3.6061297685034237,
+      "grad_norm": 0.1402311772108078,
+      "learning_rate": 0.0001,
+      "loss": 0.149,
+      "step": 2212
+    },
+    {
+      "epoch": 3.6077600260841214,
+      "grad_norm": 0.12671758234500885,
+      "learning_rate": 0.0001,
+      "loss": 0.1452,
+      "step": 2213
+    },
+    {
+      "epoch": 3.609390283664819,
+      "grad_norm": 0.13454151153564453,
+      "learning_rate": 0.0001,
+      "loss": 0.1565,
+      "step": 2214
+    },
+    {
+      "epoch": 3.6110205412455167,
+      "grad_norm": 0.13921399414539337,
+      "learning_rate": 0.0001,
+      "loss": 0.1686,
+      "step": 2215
+    },
+    {
+      "epoch": 3.6126507988262144,
+      "grad_norm": 0.15764585137367249,
+      "learning_rate": 0.0001,
+      "loss": 0.1696,
+      "step": 2216
+    },
+    {
+      "epoch": 3.614281056406912,
+      "grad_norm": 0.13245287537574768,
+      "learning_rate": 0.0001,
+      "loss": 0.1497,
+      "step": 2217
+    },
+    {
+      "epoch": 3.61591131398761,
+      "grad_norm": 0.15295889973640442,
+      "learning_rate": 0.0001,
+      "loss": 0.1548,
+      "step": 2218
+    },
+    {
+      "epoch": 3.617541571568308,
+      "grad_norm": 0.1387680172920227,
+      "learning_rate": 0.0001,
+      "loss": 0.1566,
+      "step": 2219
+    },
+    {
+      "epoch": 3.6191718291490056,
+      "grad_norm": 0.16183353960514069,
+      "learning_rate": 0.0001,
+      "loss": 0.1713,
+      "step": 2220
+    },
+    {
+      "epoch": 3.6208020867297033,
+      "grad_norm": 0.15998132526874542,
+      "learning_rate": 0.0001,
+      "loss": 0.1716,
+      "step": 2221
+    },
+    {
+      "epoch": 3.622432344310401,
+      "grad_norm": 0.14948931336402893,
+      "learning_rate": 0.0001,
+      "loss": 0.1558,
+      "step": 2222
+    },
+    {
+      "epoch": 3.6240626018910986,
+      "grad_norm": 0.13298356533050537,
+      "learning_rate": 0.0001,
+      "loss": 0.1497,
+      "step": 2223
+    },
+    {
+      "epoch": 3.6256928594717968,
+      "grad_norm": 0.1536460816860199,
+      "learning_rate": 0.0001,
+      "loss": 0.1637,
+      "step": 2224
+    },
+    {
+      "epoch": 3.6273231170524944,
+      "grad_norm": 0.13881844282150269,
+      "learning_rate": 0.0001,
+      "loss": 0.1637,
+      "step": 2225
+    },
+    {
+      "epoch": 3.628953374633192,
+      "grad_norm": 0.14646394550800323,
+      "learning_rate": 0.0001,
+      "loss": 0.1608,
+      "step": 2226
+    },
+    {
+      "epoch": 3.63058363221389,
+      "grad_norm": 0.16166602075099945,
+      "learning_rate": 0.0001,
+      "loss": 0.1587,
+      "step": 2227
+    },
+    {
+      "epoch": 3.6322138897945875,
+      "grad_norm": 0.1409720927476883,
+      "learning_rate": 0.0001,
+      "loss": 0.1553,
+      "step": 2228
+    },
+    {
+      "epoch": 3.633844147375285,
+      "grad_norm": 0.14227084815502167,
+      "learning_rate": 0.0001,
+      "loss": 0.164,
+      "step": 2229
+    },
+    {
+      "epoch": 3.635474404955983,
+      "grad_norm": 0.13686221837997437,
+      "learning_rate": 0.0001,
+      "loss": 0.1513,
+      "step": 2230
+    },
+    {
+      "epoch": 3.6371046625366805,
+      "grad_norm": 0.1413184404373169,
+      "learning_rate": 0.0001,
+      "loss": 0.1614,
+      "step": 2231
+    },
+    {
+      "epoch": 3.6387349201173786,
+      "grad_norm": 0.15510554611682892,
+      "learning_rate": 0.0001,
+      "loss": 0.1584,
+      "step": 2232
+    },
+    {
+      "epoch": 3.6403651776980763,
+      "grad_norm": 0.17040590941905975,
+      "learning_rate": 0.0001,
+      "loss": 0.1584,
+      "step": 2233
+    },
+    {
+      "epoch": 3.641995435278774,
+      "grad_norm": 0.14824065566062927,
+      "learning_rate": 0.0001,
+      "loss": 0.1626,
+      "step": 2234
+    },
+    {
+      "epoch": 3.6436256928594717,
+      "grad_norm": 0.13972316682338715,
+      "learning_rate": 0.0001,
+      "loss": 0.1615,
+      "step": 2235
+    },
+    {
+      "epoch": 3.64525595044017,
+      "grad_norm": 0.13162623345851898,
+      "learning_rate": 0.0001,
+      "loss": 0.1767,
+      "step": 2236
+    },
+    {
+      "epoch": 3.6468862080208675,
+      "grad_norm": 0.1307874321937561,
+      "learning_rate": 0.0001,
+      "loss": 0.1464,
+      "step": 2237
+    },
+    {
+      "epoch": 3.648516465601565,
+      "grad_norm": 0.13576972484588623,
+      "learning_rate": 0.0001,
+      "loss": 0.1629,
+      "step": 2238
+    },
+    {
+      "epoch": 3.650146723182263,
+      "grad_norm": 0.17965367436408997,
+      "learning_rate": 0.0001,
+      "loss": 0.1658,
+      "step": 2239
+    },
+    {
+      "epoch": 3.6517769807629605,
+      "grad_norm": 0.1429440826177597,
+      "learning_rate": 0.0001,
+      "loss": 0.1532,
+      "step": 2240
+    },
+    {
+      "epoch": 3.653407238343658,
+      "grad_norm": 0.1440856158733368,
+      "learning_rate": 0.0001,
+      "loss": 0.1483,
+      "step": 2241
+    },
+    {
+      "epoch": 3.655037495924356,
+      "grad_norm": 0.15915343165397644,
+      "learning_rate": 0.0001,
+      "loss": 0.1551,
+      "step": 2242
+    },
+    {
+      "epoch": 3.6566677535050536,
+      "grad_norm": 0.16080951690673828,
+      "learning_rate": 0.0001,
+      "loss": 0.1573,
+      "step": 2243
+    },
+    {
+      "epoch": 3.6582980110857517,
+      "grad_norm": 0.166608989238739,
+      "learning_rate": 0.0001,
+      "loss": 0.1501,
+      "step": 2244
+    },
+    {
+      "epoch": 3.6599282686664494,
+      "grad_norm": 0.16346345841884613,
+      "learning_rate": 0.0001,
+      "loss": 0.1637,
+      "step": 2245
+    },
+    {
+      "epoch": 3.661558526247147,
+      "grad_norm": 0.1675470620393753,
+      "learning_rate": 0.0001,
+      "loss": 0.1635,
+      "step": 2246
+    },
+    {
+      "epoch": 3.6631887838278447,
+      "grad_norm": 0.15890179574489594,
+      "learning_rate": 0.0001,
+      "loss": 0.1633,
+      "step": 2247
+    },
+    {
+      "epoch": 3.6648190414085424,
+      "grad_norm": 0.1385503113269806,
+      "learning_rate": 0.0001,
+      "loss": 0.1517,
+      "step": 2248
+    },
+    {
+      "epoch": 3.6664492989892405,
+      "grad_norm": 0.14586155116558075,
+      "learning_rate": 0.0001,
+      "loss": 0.1647,
+      "step": 2249
+    },
+    {
+      "epoch": 3.6680795565699382,
+      "grad_norm": 0.15654848515987396,
+      "learning_rate": 0.0001,
+      "loss": 0.1782,
+      "step": 2250
+    },
+    {
+      "epoch": 3.669709814150636,
+      "grad_norm": 0.14317664504051208,
+      "learning_rate": 0.0001,
+      "loss": 0.1527,
+      "step": 2251
+    },
+    {
+      "epoch": 3.6713400717313336,
+      "grad_norm": 0.1403859406709671,
+      "learning_rate": 0.0001,
+      "loss": 0.1611,
+      "step": 2252
+    },
+    {
+      "epoch": 3.6729703293120313,
+      "grad_norm": 0.16431018710136414,
+      "learning_rate": 0.0001,
+      "loss": 0.1633,
+      "step": 2253
+    },
+    {
+      "epoch": 3.674600586892729,
+      "grad_norm": 0.14151932299137115,
+      "learning_rate": 0.0001,
+      "loss": 0.1465,
+      "step": 2254
+    },
+    {
+      "epoch": 3.6762308444734266,
+      "grad_norm": 0.1487053781747818,
+      "learning_rate": 0.0001,
+      "loss": 0.1595,
+      "step": 2255
+    },
+    {
+      "epoch": 3.6778611020541243,
+      "grad_norm": 0.15442033112049103,
+      "learning_rate": 0.0001,
+      "loss": 0.1564,
+      "step": 2256
+    },
+    {
+      "epoch": 3.6794913596348224,
+      "grad_norm": 0.13494384288787842,
+      "learning_rate": 0.0001,
+      "loss": 0.1544,
+      "step": 2257
+    },
+    {
+      "epoch": 3.68112161721552,
+      "grad_norm": 0.1306290328502655,
+      "learning_rate": 0.0001,
+      "loss": 0.1472,
+      "step": 2258
+    },
+    {
+      "epoch": 3.682751874796218,
+      "grad_norm": 0.15780360996723175,
+      "learning_rate": 0.0001,
+      "loss": 0.1579,
+      "step": 2259
+    },
+    {
+      "epoch": 3.6843821323769155,
+      "grad_norm": 0.14508086442947388,
+      "learning_rate": 0.0001,
+      "loss": 0.1634,
+      "step": 2260
+    },
+    {
+      "epoch": 3.686012389957613,
+      "grad_norm": 0.21266472339630127,
+      "learning_rate": 0.0001,
+      "loss": 0.1639,
+      "step": 2261
+    },
+    {
+      "epoch": 3.6876426475383113,
+      "grad_norm": 0.14544187486171722,
+      "learning_rate": 0.0001,
+      "loss": 0.1665,
+      "step": 2262
+    },
+    {
+      "epoch": 3.689272905119009,
+      "grad_norm": 0.1321001499891281,
+      "learning_rate": 0.0001,
+      "loss": 0.1571,
+      "step": 2263
+    },
+    {
+      "epoch": 3.6909031626997066,
+      "grad_norm": 0.15363967418670654,
+      "learning_rate": 0.0001,
+      "loss": 0.1452,
+      "step": 2264
+    },
+    {
+      "epoch": 3.6925334202804043,
+      "grad_norm": 0.149633526802063,
+      "learning_rate": 0.0001,
+      "loss": 0.1603,
+      "step": 2265
+    },
+    {
+      "epoch": 3.694163677861102,
+      "grad_norm": 0.15210863947868347,
+      "learning_rate": 0.0001,
+      "loss": 0.1528,
+      "step": 2266
+    },
+    {
+      "epoch": 3.6957939354417997,
+      "grad_norm": 0.1460522711277008,
+      "learning_rate": 0.0001,
+      "loss": 0.1634,
+      "step": 2267
+    },
+    {
+      "epoch": 3.6974241930224974,
+      "grad_norm": 0.15007948875427246,
+      "learning_rate": 0.0001,
+      "loss": 0.1468,
+      "step": 2268
+    },
+    {
+      "epoch": 3.699054450603195,
+      "grad_norm": 0.15326817333698273,
+      "learning_rate": 0.0001,
+      "loss": 0.1586,
+      "step": 2269
+    },
+    {
+      "epoch": 3.700684708183893,
+      "grad_norm": 0.14278486371040344,
+      "learning_rate": 0.0001,
+      "loss": 0.1694,
+      "step": 2270
+    },
+    {
+      "epoch": 3.702314965764591,
+      "grad_norm": 0.14537227153778076,
+      "learning_rate": 0.0001,
+      "loss": 0.1559,
+      "step": 2271
+    },
+    {
+      "epoch": 3.7039452233452885,
+      "grad_norm": 0.1516200751066208,
+      "learning_rate": 0.0001,
+      "loss": 0.1577,
+      "step": 2272
+    },
+    {
+      "epoch": 3.705575480925986,
+      "grad_norm": 0.14554426074028015,
+      "learning_rate": 0.0001,
+      "loss": 0.1645,
+      "step": 2273
+    },
+    {
+      "epoch": 3.707205738506684,
+      "grad_norm": 0.1325369030237198,
+      "learning_rate": 0.0001,
+      "loss": 0.1614,
+      "step": 2274
+    },
+    {
+      "epoch": 3.708835996087382,
+      "grad_norm": 0.1257917732000351,
+      "learning_rate": 0.0001,
+      "loss": 0.1445,
+      "step": 2275
+    },
+    {
+      "epoch": 3.7104662536680797,
+      "grad_norm": 0.1275888830423355,
+      "learning_rate": 0.0001,
+      "loss": 0.16,
+      "step": 2276
+    },
+    {
+      "epoch": 3.7120965112487774,
+      "grad_norm": 0.14174294471740723,
+      "learning_rate": 0.0001,
+      "loss": 0.1551,
+      "step": 2277
+    },
+    {
+      "epoch": 3.713726768829475,
+      "grad_norm": 0.16611164808273315,
+      "learning_rate": 0.0001,
+      "loss": 0.1635,
+      "step": 2278
+    },
+    {
+      "epoch": 3.7153570264101727,
+      "grad_norm": 0.1329105794429779,
+      "learning_rate": 0.0001,
+      "loss": 0.153,
+      "step": 2279
+    },
+    {
+      "epoch": 3.7169872839908704,
+      "grad_norm": 0.13968057930469513,
+      "learning_rate": 0.0001,
+      "loss": 0.158,
+      "step": 2280
+    },
+    {
+      "epoch": 3.718617541571568,
+      "grad_norm": 0.1441374272108078,
+      "learning_rate": 0.0001,
+      "loss": 0.1452,
+      "step": 2281
+    },
+    {
+      "epoch": 3.720247799152266,
+      "grad_norm": 0.14315244555473328,
+      "learning_rate": 0.0001,
+      "loss": 0.1561,
+      "step": 2282
+    },
+    {
+      "epoch": 3.721878056732964,
+      "grad_norm": 0.12441976368427277,
+      "learning_rate": 0.0001,
+      "loss": 0.1544,
+      "step": 2283
+    },
+    {
+      "epoch": 3.7235083143136616,
+      "grad_norm": 0.14892461895942688,
+      "learning_rate": 0.0001,
+      "loss": 0.1622,
+      "step": 2284
+    },
+    {
+      "epoch": 3.7251385718943593,
+      "grad_norm": 0.11621859669685364,
+      "learning_rate": 0.0001,
+      "loss": 0.1441,
+      "step": 2285
+    },
+    {
+      "epoch": 3.726768829475057,
+      "grad_norm": 0.1611883044242859,
+      "learning_rate": 0.0001,
+      "loss": 0.1531,
+      "step": 2286
+    },
+    {
+      "epoch": 3.728399087055755,
+      "grad_norm": 0.15095670521259308,
+      "learning_rate": 0.0001,
+      "loss": 0.1671,
+      "step": 2287
+    },
+    {
+      "epoch": 3.7300293446364527,
+      "grad_norm": 0.1477982997894287,
+      "learning_rate": 0.0001,
+      "loss": 0.1529,
+      "step": 2288
+    },
+    {
+      "epoch": 3.7316596022171504,
+      "grad_norm": 0.17801976203918457,
+      "learning_rate": 0.0001,
+      "loss": 0.1698,
+      "step": 2289
+    },
+    {
+      "epoch": 3.733289859797848,
+      "grad_norm": 0.12449807673692703,
+      "learning_rate": 0.0001,
+      "loss": 0.1411,
+      "step": 2290
+    },
+    {
+      "epoch": 3.734920117378546,
+      "grad_norm": 0.14423514902591705,
+      "learning_rate": 0.0001,
+      "loss": 0.1507,
+      "step": 2291
+    },
+    {
+      "epoch": 3.7365503749592435,
+      "grad_norm": 0.15219520032405853,
+      "learning_rate": 0.0001,
+      "loss": 0.1746,
+      "step": 2292
+    },
+    {
+      "epoch": 3.738180632539941,
+      "grad_norm": 0.13563816249370575,
+      "learning_rate": 0.0001,
+      "loss": 0.1446,
+      "step": 2293
+    },
+    {
+      "epoch": 3.739810890120639,
+      "grad_norm": 0.13731209933757782,
+      "learning_rate": 0.0001,
+      "loss": 0.1467,
+      "step": 2294
+    },
+    {
+      "epoch": 3.741441147701337,
+      "grad_norm": 0.14237910509109497,
+      "learning_rate": 0.0001,
+      "loss": 0.1635,
+      "step": 2295
+    },
+    {
+      "epoch": 3.7430714052820346,
+      "grad_norm": 0.13505598902702332,
+      "learning_rate": 0.0001,
+      "loss": 0.148,
+      "step": 2296
+    },
+    {
+      "epoch": 3.7447016628627323,
+      "grad_norm": 0.17628848552703857,
+      "learning_rate": 0.0001,
+      "loss": 0.1539,
+      "step": 2297
+    },
+    {
+      "epoch": 3.74633192044343,
+      "grad_norm": 0.14530231058597565,
+      "learning_rate": 0.0001,
+      "loss": 0.1572,
+      "step": 2298
+    },
+    {
+      "epoch": 3.7479621780241277,
+      "grad_norm": 0.14856889843940735,
+      "learning_rate": 0.0001,
+      "loss": 0.1623,
+      "step": 2299
+    },
+    {
+      "epoch": 3.749592435604826,
+      "grad_norm": 0.170349583029747,
+      "learning_rate": 0.0001,
+      "loss": 0.1598,
+      "step": 2300
+    },
+    {
+      "epoch": 3.7512226931855235,
+      "grad_norm": 0.14849968254566193,
+      "learning_rate": 0.0001,
+      "loss": 0.1751,
+      "step": 2301
+    },
+    {
+      "epoch": 3.752852950766221,
+      "grad_norm": 0.1442185491323471,
+      "learning_rate": 0.0001,
+      "loss": 0.1454,
+      "step": 2302
+    },
+    {
+      "epoch": 3.754483208346919,
+      "grad_norm": 0.14942088723182678,
+      "learning_rate": 0.0001,
+      "loss": 0.151,
+      "step": 2303
+    },
+    {
+      "epoch": 3.7561134659276165,
+      "grad_norm": 0.15279610455036163,
+      "learning_rate": 0.0001,
+      "loss": 0.1574,
+      "step": 2304
+    },
+    {
+      "epoch": 3.757743723508314,
+      "grad_norm": 0.1706325262784958,
+      "learning_rate": 0.0001,
+      "loss": 0.1537,
+      "step": 2305
+    },
+    {
+      "epoch": 3.759373981089012,
+      "grad_norm": 0.15288381278514862,
+      "learning_rate": 0.0001,
+      "loss": 0.1618,
+      "step": 2306
+    },
+    {
+      "epoch": 3.7610042386697096,
+      "grad_norm": 0.1183413565158844,
+      "learning_rate": 0.0001,
+      "loss": 0.1478,
+      "step": 2307
+    },
+    {
+      "epoch": 3.7626344962504077,
+      "grad_norm": 0.13402055203914642,
+      "learning_rate": 0.0001,
+      "loss": 0.1464,
+      "step": 2308
+    },
+    {
+      "epoch": 3.7642647538311054,
+      "grad_norm": 0.14990192651748657,
+      "learning_rate": 0.0001,
+      "loss": 0.1568,
+      "step": 2309
+    },
+    {
+      "epoch": 3.765895011411803,
+      "grad_norm": 0.15893998742103577,
+      "learning_rate": 0.0001,
+      "loss": 0.1541,
+      "step": 2310
+    },
+    {
+      "epoch": 3.7675252689925007,
+      "grad_norm": 0.15546587109565735,
+      "learning_rate": 0.0001,
+      "loss": 0.1541,
+      "step": 2311
+    },
+    {
+      "epoch": 3.7691555265731984,
+      "grad_norm": 0.16592715680599213,
+      "learning_rate": 0.0001,
+      "loss": 0.1632,
+      "step": 2312
+    },
+    {
+      "epoch": 3.7707857841538965,
+      "grad_norm": 0.15784431993961334,
+      "learning_rate": 0.0001,
+      "loss": 0.1627,
+      "step": 2313
+    },
+    {
+      "epoch": 3.772416041734594,
+      "grad_norm": 0.16422522068023682,
+      "learning_rate": 0.0001,
+      "loss": 0.1589,
+      "step": 2314
+    },
+    {
+      "epoch": 3.774046299315292,
+      "grad_norm": 0.14868198335170746,
+      "learning_rate": 0.0001,
+      "loss": 0.168,
+      "step": 2315
+    },
+    {
+      "epoch": 3.7756765568959896,
+      "grad_norm": 0.1566770076751709,
+      "learning_rate": 0.0001,
+      "loss": 0.1657,
+      "step": 2316
+    },
+    {
+      "epoch": 3.7773068144766873,
+      "grad_norm": 0.16978543996810913,
+      "learning_rate": 0.0001,
+      "loss": 0.1442,
+      "step": 2317
+    },
+    {
+      "epoch": 3.778937072057385,
+      "grad_norm": 0.15802276134490967,
+      "learning_rate": 0.0001,
+      "loss": 0.1564,
+      "step": 2318
+    },
+    {
+      "epoch": 3.7805673296380826,
+      "grad_norm": 0.15518346428871155,
+      "learning_rate": 0.0001,
+      "loss": 0.156,
+      "step": 2319
+    },
+    {
+      "epoch": 3.7821975872187803,
+      "grad_norm": 0.1578715294599533,
+      "learning_rate": 0.0001,
+      "loss": 0.1687,
+      "step": 2320
+    },
+    {
+      "epoch": 3.7838278447994784,
+      "grad_norm": 0.14465771615505219,
+      "learning_rate": 0.0001,
+      "loss": 0.1474,
+      "step": 2321
+    },
+    {
+      "epoch": 3.785458102380176,
+      "grad_norm": 0.1761205941438675,
+      "learning_rate": 0.0001,
+      "loss": 0.1566,
+      "step": 2322
+    },
+    {
+      "epoch": 3.787088359960874,
+      "grad_norm": 0.1283109188079834,
+      "learning_rate": 0.0001,
+      "loss": 0.1473,
+      "step": 2323
+    },
+    {
+      "epoch": 3.7887186175415715,
+      "grad_norm": 0.14071565866470337,
+      "learning_rate": 0.0001,
+      "loss": 0.151,
+      "step": 2324
+    },
+    {
+      "epoch": 3.7903488751222696,
+      "grad_norm": 0.17781133949756622,
+      "learning_rate": 0.0001,
+      "loss": 0.1635,
+      "step": 2325
+    },
+    {
+      "epoch": 3.7919791327029673,
+      "grad_norm": 0.1420576274394989,
+      "learning_rate": 0.0001,
+      "loss": 0.1668,
+      "step": 2326
+    },
+    {
+      "epoch": 3.793609390283665,
+      "grad_norm": 0.1406291425228119,
+      "learning_rate": 0.0001,
+      "loss": 0.1454,
+      "step": 2327
+    },
+    {
+      "epoch": 3.7952396478643626,
+      "grad_norm": 0.14093580842018127,
+      "learning_rate": 0.0001,
+      "loss": 0.1495,
+      "step": 2328
+    },
+    {
+      "epoch": 3.7968699054450603,
+      "grad_norm": 0.12997625768184662,
+      "learning_rate": 0.0001,
+      "loss": 0.1512,
+      "step": 2329
+    },
+    {
+      "epoch": 3.798500163025758,
+      "grad_norm": 0.1375696063041687,
+      "learning_rate": 0.0001,
+      "loss": 0.1515,
+      "step": 2330
+    },
+    {
+      "epoch": 3.8001304206064557,
+      "grad_norm": 0.15136592090129852,
+      "learning_rate": 0.0001,
+      "loss": 0.1583,
+      "step": 2331
+    },
+    {
+      "epoch": 3.8017606781871534,
+      "grad_norm": 0.1388722062110901,
+      "learning_rate": 0.0001,
+      "loss": 0.1516,
+      "step": 2332
+    },
+    {
+      "epoch": 3.8033909357678515,
+      "grad_norm": 0.15683171153068542,
+      "learning_rate": 0.0001,
+      "loss": 0.1726,
+      "step": 2333
+    },
+    {
+      "epoch": 3.805021193348549,
+      "grad_norm": 0.13569128513336182,
+      "learning_rate": 0.0001,
+      "loss": 0.1585,
+      "step": 2334
+    },
+    {
+      "epoch": 3.806651450929247,
+      "grad_norm": 0.14862379431724548,
+      "learning_rate": 0.0001,
+      "loss": 0.1553,
+      "step": 2335
+    },
+    {
+      "epoch": 3.8082817085099445,
+      "grad_norm": 0.14512869715690613,
+      "learning_rate": 0.0001,
+      "loss": 0.1635,
+      "step": 2336
+    },
+    {
+      "epoch": 3.809911966090642,
+      "grad_norm": 0.13408806920051575,
+      "learning_rate": 0.0001,
+      "loss": 0.1732,
+      "step": 2337
+    },
+    {
+      "epoch": 3.8115422236713403,
+      "grad_norm": 0.1384328156709671,
+      "learning_rate": 0.0001,
+      "loss": 0.1631,
+      "step": 2338
+    },
+    {
+      "epoch": 3.813172481252038,
+      "grad_norm": 0.1299862563610077,
+      "learning_rate": 0.0001,
+      "loss": 0.16,
+      "step": 2339
+    },
+    {
+      "epoch": 3.8148027388327357,
+      "grad_norm": 0.13506759703159332,
+      "learning_rate": 0.0001,
+      "loss": 0.1719,
+      "step": 2340
+    },
+    {
+      "epoch": 3.8164329964134334,
+      "grad_norm": 0.1495598405599594,
+      "learning_rate": 0.0001,
+      "loss": 0.1501,
+      "step": 2341
+    },
+    {
+      "epoch": 3.818063253994131,
+      "grad_norm": 0.1325817108154297,
+      "learning_rate": 0.0001,
+      "loss": 0.1525,
+      "step": 2342
+    },
+    {
+      "epoch": 3.8196935115748287,
+      "grad_norm": 0.13198593258857727,
+      "learning_rate": 0.0001,
+      "loss": 0.1731,
+      "step": 2343
+    },
+    {
+      "epoch": 3.8213237691555264,
+      "grad_norm": 0.1505766659975052,
+      "learning_rate": 0.0001,
+      "loss": 0.1608,
+      "step": 2344
+    },
+    {
+      "epoch": 3.822954026736224,
+      "grad_norm": 0.13334894180297852,
+      "learning_rate": 0.0001,
+      "loss": 0.1497,
+      "step": 2345
+    },
+    {
+      "epoch": 3.824584284316922,
+      "grad_norm": 0.15418575704097748,
+      "learning_rate": 0.0001,
+      "loss": 0.1588,
+      "step": 2346
+    },
+    {
+      "epoch": 3.82621454189762,
+      "grad_norm": 0.1457456350326538,
+      "learning_rate": 0.0001,
+      "loss": 0.1482,
+      "step": 2347
+    },
+    {
+      "epoch": 3.8278447994783176,
+      "grad_norm": 0.20289213955402374,
+      "learning_rate": 0.0001,
+      "loss": 0.1491,
+      "step": 2348
+    },
+    {
+      "epoch": 3.8294750570590153,
+      "grad_norm": 0.15711665153503418,
+      "learning_rate": 0.0001,
+      "loss": 0.16,
+      "step": 2349
+    },
+    {
+      "epoch": 3.831105314639713,
+      "grad_norm": 0.17610590159893036,
+      "learning_rate": 0.0001,
+      "loss": 0.1741,
+      "step": 2350
+    },
+    {
+      "epoch": 3.832735572220411,
+      "grad_norm": 0.15204782783985138,
+      "learning_rate": 0.0001,
+      "loss": 0.1443,
+      "step": 2351
+    },
+    {
+      "epoch": 3.8343658298011087,
+      "grad_norm": 0.18176928162574768,
+      "learning_rate": 0.0001,
+      "loss": 0.1685,
+      "step": 2352
+    },
+    {
+      "epoch": 3.8359960873818064,
+      "grad_norm": 0.16083329916000366,
+      "learning_rate": 0.0001,
+      "loss": 0.1564,
+      "step": 2353
+    },
+    {
+      "epoch": 3.837626344962504,
+      "grad_norm": 0.1513577103614807,
+      "learning_rate": 0.0001,
+      "loss": 0.1556,
+      "step": 2354
+    },
+    {
+      "epoch": 3.839256602543202,
+      "grad_norm": 0.15226048231124878,
+      "learning_rate": 0.0001,
+      "loss": 0.1565,
+      "step": 2355
+    },
+    {
+      "epoch": 3.8408868601238995,
+      "grad_norm": 0.16025280952453613,
+      "learning_rate": 0.0001,
+      "loss": 0.1512,
+      "step": 2356
+    },
+    {
+      "epoch": 3.842517117704597,
+      "grad_norm": 0.16616910696029663,
+      "learning_rate": 0.0001,
+      "loss": 0.1675,
+      "step": 2357
+    },
+    {
+      "epoch": 3.844147375285295,
+      "grad_norm": 0.14858278632164001,
+      "learning_rate": 0.0001,
+      "loss": 0.1332,
+      "step": 2358
+    },
+    {
+      "epoch": 3.845777632865993,
+      "grad_norm": 0.15278182923793793,
+      "learning_rate": 0.0001,
+      "loss": 0.154,
+      "step": 2359
+    },
+    {
+      "epoch": 3.8474078904466906,
+      "grad_norm": 0.15966159105300903,
+      "learning_rate": 0.0001,
+      "loss": 0.1541,
+      "step": 2360
+    },
+    {
+      "epoch": 3.8490381480273883,
+      "grad_norm": 0.12534458935260773,
+      "learning_rate": 0.0001,
+      "loss": 0.1547,
+      "step": 2361
+    },
+    {
+      "epoch": 3.850668405608086,
+      "grad_norm": 0.14975754916667938,
+      "learning_rate": 0.0001,
+      "loss": 0.179,
+      "step": 2362
+    },
+    {
+      "epoch": 3.8522986631887837,
+      "grad_norm": 0.130054771900177,
+      "learning_rate": 0.0001,
+      "loss": 0.1596,
+      "step": 2363
+    },
+    {
+      "epoch": 3.853928920769482,
+      "grad_norm": 0.1336638629436493,
+      "learning_rate": 0.0001,
+      "loss": 0.1566,
+      "step": 2364
+    },
+    {
+      "epoch": 3.8555591783501795,
+      "grad_norm": 0.14911198616027832,
+      "learning_rate": 0.0001,
+      "loss": 0.1617,
+      "step": 2365
+    },
+    {
+      "epoch": 3.857189435930877,
+      "grad_norm": 0.1424766629934311,
+      "learning_rate": 0.0001,
+      "loss": 0.164,
+      "step": 2366
+    },
+    {
+      "epoch": 3.858819693511575,
+      "grad_norm": 0.13621950149536133,
+      "learning_rate": 0.0001,
+      "loss": 0.1584,
+      "step": 2367
+    },
+    {
+      "epoch": 3.8604499510922725,
+      "grad_norm": 0.13983507454395294,
+      "learning_rate": 0.0001,
+      "loss": 0.1441,
+      "step": 2368
+    },
+    {
+      "epoch": 3.86208020867297,
+      "grad_norm": 0.14439599215984344,
+      "learning_rate": 0.0001,
+      "loss": 0.1568,
+      "step": 2369
+    },
+    {
+      "epoch": 3.863710466253668,
+      "grad_norm": 0.14129005372524261,
+      "learning_rate": 0.0001,
+      "loss": 0.1675,
+      "step": 2370
+    },
+    {
+      "epoch": 3.8653407238343656,
+      "grad_norm": 0.14672988653182983,
+      "learning_rate": 0.0001,
+      "loss": 0.1706,
+      "step": 2371
+    },
+    {
+      "epoch": 3.8669709814150637,
+      "grad_norm": 0.14792685210704803,
+      "learning_rate": 0.0001,
+      "loss": 0.1692,
+      "step": 2372
+    },
+    {
+      "epoch": 3.8686012389957614,
+      "grad_norm": 0.1503150761127472,
+      "learning_rate": 0.0001,
+      "loss": 0.1619,
+      "step": 2373
+    },
+    {
+      "epoch": 3.870231496576459,
+      "grad_norm": 0.15507511794567108,
+      "learning_rate": 0.0001,
+      "loss": 0.1681,
+      "step": 2374
+    },
+    {
+      "epoch": 3.8718617541571567,
+      "grad_norm": 0.1679387390613556,
+      "learning_rate": 0.0001,
+      "loss": 0.1451,
+      "step": 2375
+    },
+    {
+      "epoch": 3.873492011737855,
+      "grad_norm": 0.15510492026805878,
+      "learning_rate": 0.0001,
+      "loss": 0.1719,
+      "step": 2376
+    },
+    {
+      "epoch": 3.8751222693185525,
+      "grad_norm": 0.13814286887645721,
+      "learning_rate": 0.0001,
+      "loss": 0.1529,
+      "step": 2377
+    },
+    {
+      "epoch": 3.87675252689925,
+      "grad_norm": 0.12971526384353638,
+      "learning_rate": 0.0001,
+      "loss": 0.1413,
+      "step": 2378
+    },
+    {
+      "epoch": 3.878382784479948,
+      "grad_norm": 0.13721339404582977,
+      "learning_rate": 0.0001,
+      "loss": 0.1527,
+      "step": 2379
+    },
+    {
+      "epoch": 3.8800130420606456,
+      "grad_norm": 0.1391274333000183,
+      "learning_rate": 0.0001,
+      "loss": 0.1544,
+      "step": 2380
+    },
+    {
+      "epoch": 3.8816432996413432,
+      "grad_norm": 0.12956668436527252,
+      "learning_rate": 0.0001,
+      "loss": 0.1391,
+      "step": 2381
+    },
+    {
+      "epoch": 3.883273557222041,
+      "grad_norm": 0.12296263873577118,
+      "learning_rate": 0.0001,
+      "loss": 0.1668,
+      "step": 2382
+    },
+    {
+      "epoch": 3.8849038148027386,
+      "grad_norm": 0.14773519337177277,
+      "learning_rate": 0.0001,
+      "loss": 0.1704,
+      "step": 2383
+    },
+    {
+      "epoch": 3.8865340723834367,
+      "grad_norm": 0.14175285398960114,
+      "learning_rate": 0.0001,
+      "loss": 0.1571,
+      "step": 2384
+    },
+    {
+      "epoch": 3.8881643299641344,
+      "grad_norm": 0.1267097443342209,
+      "learning_rate": 0.0001,
+      "loss": 0.1431,
+      "step": 2385
+    },
+    {
+      "epoch": 3.889794587544832,
+      "grad_norm": 0.29268741607666016,
+      "learning_rate": 0.0001,
+      "loss": 0.155,
+      "step": 2386
+    },
+    {
+      "epoch": 3.8914248451255298,
+      "grad_norm": 0.13753896951675415,
+      "learning_rate": 0.0001,
+      "loss": 0.1561,
+      "step": 2387
+    },
+    {
+      "epoch": 3.8930551027062275,
+      "grad_norm": 0.1538880318403244,
+      "learning_rate": 0.0001,
+      "loss": 0.1545,
+      "step": 2388
+    },
+    {
+      "epoch": 3.8946853602869256,
+      "grad_norm": 0.15996767580509186,
+      "learning_rate": 0.0001,
+      "loss": 0.1551,
+      "step": 2389
+    },
+    {
+      "epoch": 3.8963156178676233,
+      "grad_norm": 0.1439359188079834,
+      "learning_rate": 0.0001,
+      "loss": 0.1504,
+      "step": 2390
+    },
+    {
+      "epoch": 3.897945875448321,
+      "grad_norm": 0.23938068747520447,
+      "learning_rate": 0.0001,
+      "loss": 0.1591,
+      "step": 2391
+    },
+    {
+      "epoch": 3.8995761330290186,
+      "grad_norm": 0.1339603215456009,
+      "learning_rate": 0.0001,
+      "loss": 0.1541,
+      "step": 2392
+    },
+    {
+      "epoch": 3.9012063906097163,
+      "grad_norm": 0.146190345287323,
+      "learning_rate": 0.0001,
+      "loss": 0.1639,
+      "step": 2393
+    },
+    {
+      "epoch": 3.902836648190414,
+      "grad_norm": 0.14381268620491028,
+      "learning_rate": 0.0001,
+      "loss": 0.1553,
+      "step": 2394
+    },
+    {
+      "epoch": 3.9044669057711117,
+      "grad_norm": 0.1367824673652649,
+      "learning_rate": 0.0001,
+      "loss": 0.1551,
+      "step": 2395
+    },
+    {
+      "epoch": 3.9060971633518093,
+      "grad_norm": 0.15872205793857574,
+      "learning_rate": 0.0001,
+      "loss": 0.1591,
+      "step": 2396
+    },
+    {
+      "epoch": 3.9077274209325075,
+      "grad_norm": 0.12379555404186249,
+      "learning_rate": 0.0001,
+      "loss": 0.1388,
+      "step": 2397
+    },
+    {
+      "epoch": 3.909357678513205,
+      "grad_norm": 0.14201846718788147,
+      "learning_rate": 0.0001,
+      "loss": 0.1549,
+      "step": 2398
+    },
+    {
+      "epoch": 3.910987936093903,
+      "grad_norm": 0.1701783388853073,
+      "learning_rate": 0.0001,
+      "loss": 0.1743,
+      "step": 2399
+    },
+    {
+      "epoch": 3.9126181936746005,
+      "grad_norm": 0.13417696952819824,
+      "learning_rate": 0.0001,
+      "loss": 0.1526,
+      "step": 2400
+    },
+    {
+      "epoch": 3.914248451255298,
+      "grad_norm": 0.1509416103363037,
+      "learning_rate": 0.0001,
+      "loss": 0.1607,
+      "step": 2401
+    },
+    {
+      "epoch": 3.9158787088359963,
+      "grad_norm": 0.1474318504333496,
+      "learning_rate": 0.0001,
+      "loss": 0.1509,
+      "step": 2402
+    },
+    {
+      "epoch": 3.917508966416694,
+      "grad_norm": 0.18277962505817413,
+      "learning_rate": 0.0001,
+      "loss": 0.1673,
+      "step": 2403
+    },
+    {
+      "epoch": 3.9191392239973917,
+      "grad_norm": 0.14905720949172974,
+      "learning_rate": 0.0001,
+      "loss": 0.1513,
+      "step": 2404
+    },
+    {
+      "epoch": 3.9207694815780894,
+      "grad_norm": 0.13981392979621887,
+      "learning_rate": 0.0001,
+      "loss": 0.1615,
+      "step": 2405
+    },
+    {
+      "epoch": 3.922399739158787,
+      "grad_norm": 0.13442060351371765,
+      "learning_rate": 0.0001,
+      "loss": 0.1514,
+      "step": 2406
+    },
+    {
+      "epoch": 3.9240299967394847,
+      "grad_norm": 0.13714581727981567,
+      "learning_rate": 0.0001,
+      "loss": 0.1563,
+      "step": 2407
+    },
+    {
+      "epoch": 3.9256602543201824,
+      "grad_norm": 0.15870222449302673,
+      "learning_rate": 0.0001,
+      "loss": 0.166,
+      "step": 2408
+    },
+    {
+      "epoch": 3.92729051190088,
+      "grad_norm": 0.14761915802955627,
+      "learning_rate": 0.0001,
+      "loss": 0.1587,
+      "step": 2409
+    },
+    {
+      "epoch": 3.928920769481578,
+      "grad_norm": 0.13917234539985657,
+      "learning_rate": 0.0001,
+      "loss": 0.1609,
+      "step": 2410
+    },
+    {
+      "epoch": 3.930551027062276,
+      "grad_norm": 0.13690310716629028,
+      "learning_rate": 0.0001,
+      "loss": 0.1478,
+      "step": 2411
+    },
+    {
+      "epoch": 3.9321812846429736,
+      "grad_norm": 0.1417568325996399,
+      "learning_rate": 0.0001,
+      "loss": 0.1648,
+      "step": 2412
+    },
+    {
+      "epoch": 3.9338115422236712,
+      "grad_norm": 0.13829518854618073,
+      "learning_rate": 0.0001,
+      "loss": 0.1589,
+      "step": 2413
+    },
+    {
+      "epoch": 3.9354417998043694,
+      "grad_norm": 0.11972015351057053,
+      "learning_rate": 0.0001,
+      "loss": 0.125,
+      "step": 2414
+    },
+    {
+      "epoch": 3.937072057385067,
+      "grad_norm": 0.13998998701572418,
+      "learning_rate": 0.0001,
+      "loss": 0.1721,
+      "step": 2415
+    },
+    {
+      "epoch": 3.9387023149657647,
+      "grad_norm": 0.13850805163383484,
+      "learning_rate": 0.0001,
+      "loss": 0.1451,
+      "step": 2416
+    },
+    {
+      "epoch": 3.9403325725464624,
+      "grad_norm": 0.13838806748390198,
+      "learning_rate": 0.0001,
+      "loss": 0.1509,
+      "step": 2417
+    },
+    {
+      "epoch": 3.94196283012716,
+      "grad_norm": 0.13861410319805145,
+      "learning_rate": 0.0001,
+      "loss": 0.1512,
+      "step": 2418
+    },
+    {
+      "epoch": 3.9435930877078578,
+      "grad_norm": 0.14857284724712372,
+      "learning_rate": 0.0001,
+      "loss": 0.145,
+      "step": 2419
+    },
+    {
+      "epoch": 3.9452233452885554,
+      "grad_norm": 0.13676993548870087,
+      "learning_rate": 0.0001,
+      "loss": 0.1464,
+      "step": 2420
+    },
+    {
+      "epoch": 3.946853602869253,
+      "grad_norm": 0.1672215759754181,
+      "learning_rate": 0.0001,
+      "loss": 0.1506,
+      "step": 2421
+    },
+    {
+      "epoch": 3.9484838604499513,
+      "grad_norm": 0.1676868051290512,
+      "learning_rate": 0.0001,
+      "loss": 0.1714,
+      "step": 2422
+    },
+    {
+      "epoch": 3.950114118030649,
+      "grad_norm": 0.15942254662513733,
+      "learning_rate": 0.0001,
+      "loss": 0.1543,
+      "step": 2423
+    },
+    {
+      "epoch": 3.9517443756113466,
+      "grad_norm": 0.1640825718641281,
+      "learning_rate": 0.0001,
+      "loss": 0.1555,
+      "step": 2424
+    },
+    {
+      "epoch": 3.9533746331920443,
+      "grad_norm": 0.13515083491802216,
+      "learning_rate": 0.0001,
+      "loss": 0.1591,
+      "step": 2425
+    },
+    {
+      "epoch": 3.955004890772742,
+      "grad_norm": 0.13103432953357697,
+      "learning_rate": 0.0001,
+      "loss": 0.1465,
+      "step": 2426
+    },
+    {
+      "epoch": 3.95663514835344,
+      "grad_norm": 0.13040922582149506,
+      "learning_rate": 0.0001,
+      "loss": 0.1504,
+      "step": 2427
+    },
+    {
+      "epoch": 3.958265405934138,
+      "grad_norm": 0.14233194291591644,
+      "learning_rate": 0.0001,
+      "loss": 0.1623,
+      "step": 2428
+    },
+    {
+      "epoch": 3.9598956635148355,
+      "grad_norm": 0.14986193180084229,
+      "learning_rate": 0.0001,
+      "loss": 0.154,
+      "step": 2429
+    },
+    {
+      "epoch": 3.961525921095533,
+      "grad_norm": 0.14771242439746857,
+      "learning_rate": 0.0001,
+      "loss": 0.1595,
+      "step": 2430
+    },
+    {
+      "epoch": 3.963156178676231,
+      "grad_norm": 0.12477671355009079,
+      "learning_rate": 0.0001,
+      "loss": 0.1552,
+      "step": 2431
+    },
+    {
+      "epoch": 3.9647864362569285,
+      "grad_norm": 0.12812167406082153,
+      "learning_rate": 0.0001,
+      "loss": 0.1576,
+      "step": 2432
+    },
+    {
+      "epoch": 3.966416693837626,
+      "grad_norm": 0.12061511725187302,
+      "learning_rate": 0.0001,
+      "loss": 0.1474,
+      "step": 2433
+    },
+    {
+      "epoch": 3.968046951418324,
+      "grad_norm": 0.15570174157619476,
+      "learning_rate": 0.0001,
+      "loss": 0.1528,
+      "step": 2434
+    },
+    {
+      "epoch": 3.969677208999022,
+      "grad_norm": 0.12087506800889969,
+      "learning_rate": 0.0001,
+      "loss": 0.1376,
+      "step": 2435
+    },
+    {
+      "epoch": 3.9713074665797197,
+      "grad_norm": 0.16076244413852692,
+      "learning_rate": 0.0001,
+      "loss": 0.159,
+      "step": 2436
+    },
+    {
+      "epoch": 3.9729377241604173,
+      "grad_norm": 0.17178180813789368,
+      "learning_rate": 0.0001,
+      "loss": 0.1613,
+      "step": 2437
+    },
+    {
+      "epoch": 3.974567981741115,
+      "grad_norm": 0.15928059816360474,
+      "learning_rate": 0.0001,
+      "loss": 0.1714,
+      "step": 2438
+    },
+    {
+      "epoch": 3.9761982393218127,
+      "grad_norm": 0.14910367131233215,
+      "learning_rate": 0.0001,
+      "loss": 0.1522,
+      "step": 2439
+    },
+    {
+      "epoch": 3.977828496902511,
+      "grad_norm": 0.158916637301445,
+      "learning_rate": 0.0001,
+      "loss": 0.1715,
+      "step": 2440
+    },
+    {
+      "epoch": 3.9794587544832085,
+      "grad_norm": 0.13345277309417725,
+      "learning_rate": 0.0001,
+      "loss": 0.1463,
+      "step": 2441
+    },
+    {
+      "epoch": 3.981089012063906,
+      "grad_norm": 0.1387554258108139,
+      "learning_rate": 0.0001,
+      "loss": 0.1705,
+      "step": 2442
+    },
+    {
+      "epoch": 3.982719269644604,
+      "grad_norm": 0.1317049264907837,
+      "learning_rate": 0.0001,
+      "loss": 0.1444,
+      "step": 2443
+    },
+    {
+      "epoch": 3.9843495272253016,
+      "grad_norm": 0.1379990130662918,
+      "learning_rate": 0.0001,
+      "loss": 0.1636,
+      "step": 2444
+    },
+    {
+      "epoch": 3.9859797848059992,
+      "grad_norm": 0.14041666686534882,
+      "learning_rate": 0.0001,
+      "loss": 0.1543,
+      "step": 2445
+    },
+    {
+      "epoch": 3.987610042386697,
+      "grad_norm": 0.1239699274301529,
+      "learning_rate": 0.0001,
+      "loss": 0.1477,
+      "step": 2446
+    },
+    {
+      "epoch": 3.9892402999673946,
+      "grad_norm": 0.16149182617664337,
+      "learning_rate": 0.0001,
+      "loss": 0.1662,
+      "step": 2447
+    },
+    {
+      "epoch": 3.9908705575480927,
+      "grad_norm": 0.13410444557666779,
+      "learning_rate": 0.0001,
+      "loss": 0.1662,
+      "step": 2448
+    },
+    {
+      "epoch": 3.9925008151287904,
+      "grad_norm": 0.13568072021007538,
+      "learning_rate": 0.0001,
+      "loss": 0.1534,
+      "step": 2449
+    },
+    {
+      "epoch": 3.994131072709488,
+      "grad_norm": 0.15375611186027527,
+      "learning_rate": 0.0001,
+      "loss": 0.1758,
+      "step": 2450
+    },
+    {
+      "epoch": 3.9957613302901858,
+      "grad_norm": 0.13414491713047028,
+      "learning_rate": 0.0001,
+      "loss": 0.1674,
+      "step": 2451
+    },
+    {
+      "epoch": 3.9973915878708834,
+      "grad_norm": 0.15805280208587646,
+      "learning_rate": 0.0001,
+      "loss": 0.1665,
+      "step": 2452
+    },
+    {
+      "epoch": 3.9973915878708834,
+      "step": 2452,
+      "total_flos": 7.122222383569568e+18,
+      "train_loss": 0.188311614755535,
+      "train_runtime": 156391.4769,
+      "train_samples_per_second": 0.471,
+      "train_steps_per_second": 0.016
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2452,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "total_flos": 7.122222383569568e+18,
+  "train_batch_size": 6,
+  "trial_name": null,
+  "trial_params": null
+}