diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9123047097730642,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004561523548865321,
+      "grad_norm": 10.147029876708984,
+      "learning_rate": 3.787878787878788e-08,
+      "loss": 0.4344,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009123047097730642,
+      "grad_norm": 9.425281524658203,
+      "learning_rate": 7.575757575757576e-08,
+      "loss": 0.4245,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013684570646595963,
+      "grad_norm": 9.732573509216309,
+      "learning_rate": 1.1363636363636364e-07,
+      "loss": 0.4299,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018246094195461283,
+      "grad_norm": 9.073328971862793,
+      "learning_rate": 1.5151515151515152e-07,
+      "loss": 0.4278,
+      "step": 4
+    },
+    {
+      "epoch": 0.0022807617744326604,
+      "grad_norm": 7.501804351806641,
+      "learning_rate": 1.893939393939394e-07,
+      "loss": 0.4319,
+      "step": 5
+    },
+    {
+      "epoch": 0.0027369141293191925,
+      "grad_norm": 7.199859142303467,
+      "learning_rate": 2.2727272727272729e-07,
+      "loss": 0.4347,
+      "step": 6
+    },
+    {
+      "epoch": 0.0031930664842057246,
+      "grad_norm": 4.714947700500488,
+      "learning_rate": 2.651515151515152e-07,
+      "loss": 0.4275,
+      "step": 7
+    },
+    {
+      "epoch": 0.0036492188390922567,
+      "grad_norm": 4.641926288604736,
+      "learning_rate": 3.0303030303030305e-07,
+      "loss": 0.4307,
+      "step": 8
+    },
+    {
+      "epoch": 0.004105371193978789,
+      "grad_norm": 4.131030082702637,
+      "learning_rate": 3.409090909090909e-07,
+      "loss": 0.4235,
+      "step": 9
+    },
+    {
+      "epoch": 0.004561523548865321,
+      "grad_norm": 4.803492069244385,
+      "learning_rate": 3.787878787878788e-07,
+      "loss": 0.4491,
+      "step": 10
+    },
+    {
+      "epoch": 0.005017675903751853,
+      "grad_norm": 4.609049320220947,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 0.441,
+      "step": 11
+    },
+    {
+      "epoch": 0.005473828258638385,
+      "grad_norm": 5.404257774353027,
+      "learning_rate": 4.5454545454545457e-07,
+      "loss": 0.4246,
+      "step": 12
+    },
+    {
+      "epoch": 0.005929980613524917,
+      "grad_norm": 5.120520114898682,
+      "learning_rate": 4.924242424242425e-07,
+      "loss": 0.4157,
+      "step": 13
+    },
+    {
+      "epoch": 0.006386132968411449,
+      "grad_norm": 5.029507637023926,
+      "learning_rate": 5.303030303030304e-07,
+      "loss": 0.4307,
+      "step": 14
+    },
+    {
+      "epoch": 0.006842285323297981,
+      "grad_norm": 5.628627300262451,
+      "learning_rate": 5.681818181818182e-07,
+      "loss": 0.42,
+      "step": 15
+    },
+    {
+      "epoch": 0.007298437678184513,
+      "grad_norm": 6.867446422576904,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 0.4163,
+      "step": 16
+    },
+    {
+      "epoch": 0.0077545900330710455,
+      "grad_norm": 6.876864433288574,
+      "learning_rate": 6.43939393939394e-07,
+      "loss": 0.4093,
+      "step": 17
+    },
+    {
+      "epoch": 0.008210742387957578,
+      "grad_norm": 7.119993209838867,
+      "learning_rate": 6.818181818181818e-07,
+      "loss": 0.3981,
+      "step": 18
+    },
+    {
+      "epoch": 0.00866689474284411,
+      "grad_norm": 6.600307464599609,
+      "learning_rate": 7.196969696969698e-07,
+      "loss": 0.4,
+      "step": 19
+    },
+    {
+      "epoch": 0.009123047097730642,
+      "grad_norm": 6.1094489097595215,
+      "learning_rate": 7.575757575757576e-07,
+      "loss": 0.384,
+      "step": 20
+    },
+    {
+      "epoch": 0.009579199452617175,
+      "grad_norm": 5.499886512756348,
+      "learning_rate": 7.954545454545455e-07,
+      "loss": 0.3947,
+      "step": 21
+    },
+    {
+      "epoch": 0.010035351807503706,
+      "grad_norm": 17.080503463745117,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.3733,
+      "step": 22
+    },
+    {
+      "epoch": 0.010491504162390239,
+      "grad_norm": 20.898975372314453,
+      "learning_rate": 8.712121212121213e-07,
+      "loss": 0.3793,
+      "step": 23
+    },
+    {
+      "epoch": 0.01094765651727677,
+      "grad_norm": 17.195316314697266,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.3461,
+      "step": 24
+    },
+    {
+      "epoch": 0.011403808872163303,
+      "grad_norm": 15.659423828125,
+      "learning_rate": 9.469696969696971e-07,
+      "loss": 0.3624,
+      "step": 25
+    },
+    {
+      "epoch": 0.011859961227049834,
+      "grad_norm": 5.359454154968262,
+      "learning_rate": 9.84848484848485e-07,
+      "loss": 0.3458,
+      "step": 26
+    },
+    {
+      "epoch": 0.012316113581936367,
+      "grad_norm": 4.0308380126953125,
+      "learning_rate": 1.0227272727272729e-06,
+      "loss": 0.3527,
+      "step": 27
+    },
+    {
+      "epoch": 0.012772265936822898,
+      "grad_norm": 9.697226524353027,
+      "learning_rate": 1.0606060606060608e-06,
+      "loss": 0.3379,
+      "step": 28
+    },
+    {
+      "epoch": 0.013228418291709431,
+      "grad_norm": 3.423438549041748,
+      "learning_rate": 1.0984848484848485e-06,
+      "loss": 0.3416,
+      "step": 29
+    },
+    {
+      "epoch": 0.013684570646595963,
+      "grad_norm": 4.606201648712158,
+      "learning_rate": 1.1363636363636364e-06,
+      "loss": 0.3267,
+      "step": 30
+    },
+    {
+      "epoch": 0.014140723001482496,
+      "grad_norm": 3.0502517223358154,
+      "learning_rate": 1.1742424242424245e-06,
+      "loss": 0.2893,
+      "step": 31
+    },
+    {
+      "epoch": 0.014596875356369027,
+      "grad_norm": 2.1415719985961914,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 0.2903,
+      "step": 32
+    },
+    {
+      "epoch": 0.01505302771125556,
+      "grad_norm": 11.651140213012695,
+      "learning_rate": 1.25e-06,
+      "loss": 0.2836,
+      "step": 33
+    },
+    {
+      "epoch": 0.015509180066142091,
+      "grad_norm": 5.237735748291016,
+      "learning_rate": 1.287878787878788e-06,
+      "loss": 0.2949,
+      "step": 34
+    },
+    {
+      "epoch": 0.015965332421028624,
+      "grad_norm": 9.150033950805664,
+      "learning_rate": 1.3257575757575757e-06,
+      "loss": 0.2716,
+      "step": 35
+    },
+    {
+      "epoch": 0.016421484775915155,
+      "grad_norm": 1.0813686847686768,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 0.2712,
+      "step": 36
+    },
+    {
+      "epoch": 0.016877637130801686,
+      "grad_norm": 2.1033756732940674,
+      "learning_rate": 1.4015151515151515e-06,
+      "loss": 0.2715,
+      "step": 37
+    },
+    {
+      "epoch": 0.01733378948568822,
+      "grad_norm": 1.5601638555526733,
+      "learning_rate": 1.4393939393939396e-06,
+      "loss": 0.2643,
+      "step": 38
+    },
+    {
+      "epoch": 0.017789941840574752,
+      "grad_norm": 0.8007458448410034,
+      "learning_rate": 1.4772727272727275e-06,
+      "loss": 0.2627,
+      "step": 39
+    },
+    {
+      "epoch": 0.018246094195461283,
+      "grad_norm": 2.066539764404297,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 0.2689,
+      "step": 40
+    },
+    {
+      "epoch": 0.018702246550347815,
+      "grad_norm": 2.0211870670318604,
+      "learning_rate": 1.5530303030303032e-06,
+      "loss": 0.2589,
+      "step": 41
+    },
+    {
+      "epoch": 0.01915839890523435,
+      "grad_norm": 0.9270282983779907,
+      "learning_rate": 1.590909090909091e-06,
+      "loss": 0.2548,
+      "step": 42
+    },
+    {
+      "epoch": 0.01961455126012088,
+      "grad_norm": 1.046525239944458,
+      "learning_rate": 1.628787878787879e-06,
+      "loss": 0.2383,
+      "step": 43
+    },
+    {
+      "epoch": 0.020070703615007412,
+      "grad_norm": 0.6939360499382019,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.2698,
+      "step": 44
+    },
+    {
+      "epoch": 0.020526855969893943,
+      "grad_norm": 1.193113088607788,
+      "learning_rate": 1.7045454545454546e-06,
+      "loss": 0.2534,
+      "step": 45
+    },
+    {
+      "epoch": 0.020983008324780478,
+      "grad_norm": 0.9029439091682434,
+      "learning_rate": 1.7424242424242427e-06,
+      "loss": 0.2238,
+      "step": 46
+    },
+    {
+      "epoch": 0.02143916067966701,
+      "grad_norm": 0.6781004667282104,
+      "learning_rate": 1.7803030303030306e-06,
+      "loss": 0.2384,
+      "step": 47
+    },
+    {
+      "epoch": 0.02189531303455354,
+      "grad_norm": 0.7137442827224731,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.2491,
+      "step": 48
+    },
+    {
+      "epoch": 0.02235146538944007,
+      "grad_norm": 0.6012441515922546,
+      "learning_rate": 1.8560606060606062e-06,
+      "loss": 0.2407,
+      "step": 49
+    },
+    {
+      "epoch": 0.022807617744326606,
+      "grad_norm": 1.0792092084884644,
+      "learning_rate": 1.8939393939393941e-06,
+      "loss": 0.2497,
+      "step": 50
+    },
+    {
+      "epoch": 0.023263770099213137,
+      "grad_norm": 0.6669754385948181,
+      "learning_rate": 1.931818181818182e-06,
+      "loss": 0.2381,
+      "step": 51
+    },
+    {
+      "epoch": 0.02371992245409967,
+      "grad_norm": 0.982768177986145,
+      "learning_rate": 1.96969696969697e-06,
+      "loss": 0.2385,
+      "step": 52
+    },
+    {
+      "epoch": 0.0241760748089862,
+      "grad_norm": 0.5674754977226257,
+      "learning_rate": 2.0075757575757576e-06,
+      "loss": 0.2301,
+      "step": 53
+    },
+    {
+      "epoch": 0.024632227163872734,
+      "grad_norm": 0.5736498832702637,
+      "learning_rate": 2.0454545454545457e-06,
+      "loss": 0.2237,
+      "step": 54
+    },
+    {
+      "epoch": 0.025088379518759266,
+      "grad_norm": 1.8275068998336792,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.2308,
+      "step": 55
+    },
+    {
+      "epoch": 0.025544531873645797,
+      "grad_norm": 0.7837151288986206,
+      "learning_rate": 2.1212121212121216e-06,
+      "loss": 0.2212,
+      "step": 56
+    },
+    {
+      "epoch": 0.026000684228532328,
+      "grad_norm": 0.8162587881088257,
+      "learning_rate": 2.1590909090909092e-06,
+      "loss": 0.2373,
+      "step": 57
+    },
+    {
+      "epoch": 0.026456836583418863,
+      "grad_norm": 0.8167909979820251,
+      "learning_rate": 2.196969696969697e-06,
+      "loss": 0.2362,
+      "step": 58
+    },
+    {
+      "epoch": 0.026912988938305394,
+      "grad_norm": 0.9385275840759277,
+      "learning_rate": 2.234848484848485e-06,
+      "loss": 0.221,
+      "step": 59
+    },
+    {
+      "epoch": 0.027369141293191925,
+      "grad_norm": 1.3416460752487183,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 0.2312,
+      "step": 60
+    },
+    {
+      "epoch": 0.02782529364807846,
+      "grad_norm": 0.8656715750694275,
+      "learning_rate": 2.310606060606061e-06,
+      "loss": 0.2306,
+      "step": 61
+    },
+    {
+      "epoch": 0.02828144600296499,
+      "grad_norm": 0.8939012885093689,
+      "learning_rate": 2.348484848484849e-06,
+      "loss": 0.2241,
+      "step": 62
+    },
+    {
+      "epoch": 0.028737598357851522,
+      "grad_norm": 1.5157166719436646,
+      "learning_rate": 2.3863636363636367e-06,
+      "loss": 0.2306,
+      "step": 63
+    },
+    {
+      "epoch": 0.029193750712738054,
+      "grad_norm": 2.806877374649048,
+      "learning_rate": 2.4242424242424244e-06,
+      "loss": 0.2312,
+      "step": 64
+    },
+    {
+      "epoch": 0.029649903067624588,
+      "grad_norm": 0.6272144317626953,
+      "learning_rate": 2.4621212121212125e-06,
+      "loss": 0.2271,
+      "step": 65
+    },
+    {
+      "epoch": 0.03010605542251112,
+      "grad_norm": 0.5834864377975464,
+      "learning_rate": 2.5e-06,
+      "loss": 0.2176,
+      "step": 66
+    },
+    {
+      "epoch": 0.03056220777739765,
+      "grad_norm": 0.5433618426322937,
+      "learning_rate": 2.537878787878788e-06,
+      "loss": 0.2218,
+      "step": 67
+    },
+    {
+      "epoch": 0.031018360132284182,
+      "grad_norm": 0.715444803237915,
+      "learning_rate": 2.575757575757576e-06,
+      "loss": 0.2349,
+      "step": 68
+    },
+    {
+      "epoch": 0.03147451248717072,
+      "grad_norm": 0.5452340841293335,
+      "learning_rate": 2.6136363636363637e-06,
+      "loss": 0.226,
+      "step": 69
+    },
+    {
+      "epoch": 0.03193066484205725,
+      "grad_norm": 0.5565879344940186,
+      "learning_rate": 2.6515151515151514e-06,
+      "loss": 0.2247,
+      "step": 70
+    },
+    {
+      "epoch": 0.03238681719694378,
+      "grad_norm": 0.5290928483009338,
+      "learning_rate": 2.6893939393939395e-06,
+      "loss": 0.2235,
+      "step": 71
+    },
+    {
+      "epoch": 0.03284296955183031,
+      "grad_norm": 0.5101991295814514,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.2257,
+      "step": 72
+    },
+    {
+      "epoch": 0.03329912190671684,
+      "grad_norm": 0.5286222100257874,
+      "learning_rate": 2.7651515151515153e-06,
+      "loss": 0.2292,
+      "step": 73
+    },
+    {
+      "epoch": 0.03375527426160337,
+      "grad_norm": 0.5347652435302734,
+      "learning_rate": 2.803030303030303e-06,
+      "loss": 0.2062,
+      "step": 74
+    },
+    {
+      "epoch": 0.03421142661648991,
+      "grad_norm": 0.5243805050849915,
+      "learning_rate": 2.8409090909090916e-06,
+      "loss": 0.2204,
+      "step": 75
+    },
+    {
+      "epoch": 0.03466757897137644,
+      "grad_norm": 0.49387186765670776,
+      "learning_rate": 2.8787878787878793e-06,
+      "loss": 0.212,
+      "step": 76
+    },
+    {
+      "epoch": 0.03512373132626297,
+      "grad_norm": 0.5565639138221741,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 0.2334,
+      "step": 77
+    },
+    {
+      "epoch": 0.035579883681149505,
+      "grad_norm": 0.5147691965103149,
+      "learning_rate": 2.954545454545455e-06,
+      "loss": 0.2194,
+      "step": 78
+    },
+    {
+      "epoch": 0.036036036036036036,
+      "grad_norm": 0.5143291354179382,
+      "learning_rate": 2.992424242424243e-06,
+      "loss": 0.2272,
+      "step": 79
+    },
+    {
+      "epoch": 0.03649218839092257,
+      "grad_norm": 0.5480614304542542,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.2288,
+      "step": 80
+    },
+    {
+      "epoch": 0.0369483407458091,
+      "grad_norm": 0.5317783951759338,
+      "learning_rate": 3.0681818181818186e-06,
+      "loss": 0.2105,
+      "step": 81
+    },
+    {
+      "epoch": 0.03740449310069563,
+      "grad_norm": 0.5546643137931824,
+      "learning_rate": 3.1060606060606063e-06,
+      "loss": 0.2276,
+      "step": 82
+    },
+    {
+      "epoch": 0.03786064545558217,
+      "grad_norm": 0.5155991911888123,
+      "learning_rate": 3.143939393939394e-06,
+      "loss": 0.2205,
+      "step": 83
+    },
+    {
+      "epoch": 0.0383167978104687,
+      "grad_norm": 0.5074533224105835,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 0.2128,
+      "step": 84
+    },
+    {
+      "epoch": 0.03877295016535523,
+      "grad_norm": 0.5260673761367798,
+      "learning_rate": 3.21969696969697e-06,
+      "loss": 0.2096,
+      "step": 85
+    },
+    {
+      "epoch": 0.03922910252024176,
+      "grad_norm": 0.4996931850910187,
+      "learning_rate": 3.257575757575758e-06,
+      "loss": 0.2203,
+      "step": 86
+    },
+    {
+      "epoch": 0.03968525487512829,
+      "grad_norm": 0.5263378024101257,
+      "learning_rate": 3.2954545454545456e-06,
+      "loss": 0.2233,
+      "step": 87
+    },
+    {
+      "epoch": 0.040141407230014824,
+      "grad_norm": 0.5121489763259888,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.2086,
+      "step": 88
+    },
+    {
+      "epoch": 0.040597559584901355,
+      "grad_norm": 0.5465315580368042,
+      "learning_rate": 3.3712121212121214e-06,
+      "loss": 0.2288,
+      "step": 89
+    },
+    {
+      "epoch": 0.041053711939787886,
+      "grad_norm": 0.5363483428955078,
+      "learning_rate": 3.409090909090909e-06,
+      "loss": 0.203,
+      "step": 90
+    },
+    {
+      "epoch": 0.041509864294674424,
+      "grad_norm": 0.515140950679779,
+      "learning_rate": 3.4469696969696977e-06,
+      "loss": 0.2214,
+      "step": 91
+    },
+    {
+      "epoch": 0.041966016649560955,
+      "grad_norm": 0.5285958647727966,
+      "learning_rate": 3.4848484848484854e-06,
+      "loss": 0.2235,
+      "step": 92
+    },
+    {
+      "epoch": 0.04242216900444749,
+      "grad_norm": 0.5260041356086731,
+      "learning_rate": 3.522727272727273e-06,
+      "loss": 0.2141,
+      "step": 93
+    },
+    {
+      "epoch": 0.04287832135933402,
+      "grad_norm": 0.4971928596496582,
+      "learning_rate": 3.560606060606061e-06,
+      "loss": 0.2008,
+      "step": 94
+    },
+    {
+      "epoch": 0.04333447371422055,
+      "grad_norm": 0.5243147015571594,
+      "learning_rate": 3.598484848484849e-06,
+      "loss": 0.2252,
+      "step": 95
+    },
+    {
+      "epoch": 0.04379062606910708,
+      "grad_norm": 0.5313335061073303,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.2209,
+      "step": 96
+    },
+    {
+      "epoch": 0.04424677842399361,
+      "grad_norm": 0.5319870114326477,
+      "learning_rate": 3.6742424242424247e-06,
+      "loss": 0.2185,
+      "step": 97
+    },
+    {
+      "epoch": 0.04470293077888014,
+      "grad_norm": 0.49033528566360474,
+      "learning_rate": 3.7121212121212124e-06,
+      "loss": 0.2069,
+      "step": 98
+    },
+    {
+      "epoch": 0.04515908313376668,
+      "grad_norm": 0.5386990308761597,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.2273,
+      "step": 99
+    },
+    {
+      "epoch": 0.04561523548865321,
+      "grad_norm": 0.5092862844467163,
+      "learning_rate": 3.7878787878787882e-06,
+      "loss": 0.2122,
+      "step": 100
+    },
+    {
+      "epoch": 0.04607138784353974,
+      "grad_norm": 0.5264452695846558,
+      "learning_rate": 3.825757575757576e-06,
+      "loss": 0.2118,
+      "step": 101
+    },
+    {
+      "epoch": 0.046527540198426275,
+      "grad_norm": 0.4846494793891907,
+      "learning_rate": 3.863636363636364e-06,
+      "loss": 0.2062,
+      "step": 102
+    },
+    {
+      "epoch": 0.046983692553312806,
+      "grad_norm": 0.5260981917381287,
+      "learning_rate": 3.901515151515151e-06,
+      "loss": 0.2055,
+      "step": 103
+    },
+    {
+      "epoch": 0.04743984490819934,
+      "grad_norm": 0.506851077079773,
+      "learning_rate": 3.93939393939394e-06,
+      "loss": 0.2061,
+      "step": 104
+    },
+    {
+      "epoch": 0.04789599726308587,
+      "grad_norm": 0.5081851482391357,
+      "learning_rate": 3.9772727272727275e-06,
+      "loss": 0.2069,
+      "step": 105
+    },
+    {
+      "epoch": 0.0483521496179724,
+      "grad_norm": 0.5418718457221985,
+      "learning_rate": 4.015151515151515e-06,
+      "loss": 0.2176,
+      "step": 106
+    },
+    {
+      "epoch": 0.04880830197285894,
+      "grad_norm": 0.5091400146484375,
+      "learning_rate": 4.053030303030303e-06,
+      "loss": 0.2022,
+      "step": 107
+    },
+    {
+      "epoch": 0.04926445432774547,
+      "grad_norm": 0.5231116414070129,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 0.2166,
+      "step": 108
+    },
+    {
+      "epoch": 0.049720606682632,
+      "grad_norm": 0.515403151512146,
+      "learning_rate": 4.128787878787879e-06,
+      "loss": 0.2116,
+      "step": 109
+    },
+    {
+      "epoch": 0.05017675903751853,
+      "grad_norm": 0.507393479347229,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.2018,
+      "step": 110
+    },
+    {
+      "epoch": 0.05063291139240506,
+      "grad_norm": 0.5226441025733948,
+      "learning_rate": 4.204545454545455e-06,
+      "loss": 0.2108,
+      "step": 111
+    },
+    {
+      "epoch": 0.051089063747291594,
+      "grad_norm": 0.5319175720214844,
+      "learning_rate": 4.242424242424243e-06,
+      "loss": 0.2221,
+      "step": 112
+    },
+    {
+      "epoch": 0.051545216102178125,
+      "grad_norm": 0.49702200293540955,
+      "learning_rate": 4.280303030303031e-06,
+      "loss": 0.2116,
+      "step": 113
+    },
+    {
+      "epoch": 0.052001368457064656,
+      "grad_norm": 0.5483461022377014,
+      "learning_rate": 4.3181818181818185e-06,
+      "loss": 0.2044,
+      "step": 114
+    },
+    {
+      "epoch": 0.052457520811951194,
+      "grad_norm": 0.5081018805503845,
+      "learning_rate": 4.356060606060606e-06,
+      "loss": 0.203,
+      "step": 115
+    },
+    {
+      "epoch": 0.052913673166837726,
+      "grad_norm": 0.49701428413391113,
+      "learning_rate": 4.393939393939394e-06,
+      "loss": 0.1908,
+      "step": 116
+    },
+    {
+      "epoch": 0.05336982552172426,
+      "grad_norm": 0.501356303691864,
+      "learning_rate": 4.4318181818181824e-06,
+      "loss": 0.2104,
+      "step": 117
+    },
+    {
+      "epoch": 0.05382597787661079,
+      "grad_norm": 0.49473562836647034,
+      "learning_rate": 4.46969696969697e-06,
+      "loss": 0.1978,
+      "step": 118
+    },
+    {
+      "epoch": 0.05428213023149732,
+      "grad_norm": 0.5121810436248779,
+      "learning_rate": 4.507575757575758e-06,
+      "loss": 0.2117,
+      "step": 119
+    },
+    {
+      "epoch": 0.05473828258638385,
+      "grad_norm": 0.4914250075817108,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.2024,
+      "step": 120
+    },
+    {
+      "epoch": 0.05519443494127038,
+      "grad_norm": 0.5256963968276978,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 0.2125,
+      "step": 121
+    },
+    {
+      "epoch": 0.05565058729615692,
+      "grad_norm": 0.4932485520839691,
+      "learning_rate": 4.621212121212122e-06,
+      "loss": 0.2012,
+      "step": 122
+    },
+    {
+      "epoch": 0.05610673965104345,
+      "grad_norm": 0.47854381799697876,
+      "learning_rate": 4.6590909090909095e-06,
+      "loss": 0.2221,
+      "step": 123
+    },
+    {
+      "epoch": 0.05656289200592998,
+      "grad_norm": 0.4833521246910095,
+      "learning_rate": 4.696969696969698e-06,
+      "loss": 0.2134,
+      "step": 124
+    },
+    {
+      "epoch": 0.05701904436081651,
+      "grad_norm": 0.43368273973464966,
+      "learning_rate": 4.734848484848486e-06,
+      "loss": 0.1985,
+      "step": 125
+    },
+    {
+      "epoch": 0.057475196715703045,
+      "grad_norm": 0.46237725019454956,
+      "learning_rate": 4.772727272727273e-06,
+      "loss": 0.1878,
+      "step": 126
+    },
+    {
+      "epoch": 0.057931349070589576,
+      "grad_norm": 0.4770749509334564,
+      "learning_rate": 4.810606060606061e-06,
+      "loss": 0.2137,
+      "step": 127
+    },
+    {
+      "epoch": 0.05838750142547611,
+      "grad_norm": 0.4436318874359131,
+      "learning_rate": 4.848484848484849e-06,
+      "loss": 0.2042,
+      "step": 128
+    },
+    {
+      "epoch": 0.05884365378036264,
+      "grad_norm": 0.41609877347946167,
+      "learning_rate": 4.8863636363636365e-06,
+      "loss": 0.1878,
+      "step": 129
+    },
+    {
+      "epoch": 0.059299806135249176,
+      "grad_norm": 0.42990630865097046,
+      "learning_rate": 4.924242424242425e-06,
+      "loss": 0.1965,
+      "step": 130
+    },
+    {
+      "epoch": 0.05975595849013571,
+      "grad_norm": 0.45672428607940674,
+      "learning_rate": 4.962121212121213e-06,
+      "loss": 0.2283,
+      "step": 131
+    },
+    {
+      "epoch": 0.06021211084502224,
+      "grad_norm": 0.3977130651473999,
+      "learning_rate": 5e-06,
+      "loss": 0.203,
+      "step": 132
+    },
+    {
+      "epoch": 0.06066826319990877,
+      "grad_norm": 0.40768885612487793,
+      "learning_rate": 4.9999993176248125e-06,
+      "loss": 0.188,
+      "step": 133
+    },
+    {
+      "epoch": 0.0611244155547953,
+      "grad_norm": 0.40733861923217773,
+      "learning_rate": 4.9999972704996216e-06,
+      "loss": 0.1977,
+      "step": 134
+    },
+    {
+      "epoch": 0.06158056790968183,
+      "grad_norm": 0.4187253415584564,
+      "learning_rate": 4.999993858625544e-06,
+      "loss": 0.1969,
+      "step": 135
+    },
+    {
+      "epoch": 0.062036720264568364,
+      "grad_norm": 0.401579886674881,
+      "learning_rate": 4.999989082004443e-06,
+      "loss": 0.1941,
+      "step": 136
+    },
+    {
+      "epoch": 0.062492872619454895,
+      "grad_norm": 0.3997587561607361,
+      "learning_rate": 4.999982940638927e-06,
+      "loss": 0.1969,
+      "step": 137
+    },
+    {
+      "epoch": 0.06294902497434143,
+      "grad_norm": 0.37221795320510864,
+      "learning_rate": 4.999975434532347e-06,
+      "loss": 0.1801,
+      "step": 138
+    },
+    {
+      "epoch": 0.06340517732922796,
+      "grad_norm": 0.41053351759910583,
+      "learning_rate": 4.9999665636888025e-06,
+      "loss": 0.2059,
+      "step": 139
+    },
+    {
+      "epoch": 0.0638613296841145,
+      "grad_norm": 0.369053453207016,
+      "learning_rate": 4.999956328113134e-06,
+      "loss": 0.1913,
+      "step": 140
+    },
+    {
+      "epoch": 0.06431748203900102,
+      "grad_norm": 0.40374141931533813,
+      "learning_rate": 4.99994472781093e-06,
+      "loss": 0.2117,
+      "step": 141
+    },
+    {
+      "epoch": 0.06477363439388756,
+      "grad_norm": 0.38011205196380615,
+      "learning_rate": 4.999931762788524e-06,
+      "loss": 0.1971,
+      "step": 142
+    },
+    {
+      "epoch": 0.0652297867487741,
+      "grad_norm": 0.3858908414840698,
+      "learning_rate": 4.999917433052992e-06,
+      "loss": 0.1997,
+      "step": 143
+    },
+    {
+      "epoch": 0.06568593910366062,
+      "grad_norm": 0.35961923003196716,
+      "learning_rate": 4.999901738612159e-06,
+      "loss": 0.1926,
+      "step": 144
+    },
+    {
+      "epoch": 0.06614209145854716,
+      "grad_norm": 0.35613885521888733,
+      "learning_rate": 4.999884679474589e-06,
+      "loss": 0.1942,
+      "step": 145
+    },
+    {
+      "epoch": 0.06659824381343368,
+      "grad_norm": 0.38008368015289307,
+      "learning_rate": 4.9998662556495975e-06,
+      "loss": 0.1954,
+      "step": 146
+    },
+    {
+      "epoch": 0.06705439616832022,
+      "grad_norm": 0.3533459007740021,
+      "learning_rate": 4.999846467147241e-06,
+      "loss": 0.2004,
+      "step": 147
+    },
+    {
+      "epoch": 0.06751054852320675,
+      "grad_norm": 0.38875409960746765,
+      "learning_rate": 4.999825313978322e-06,
+      "loss": 0.1928,
+      "step": 148
+    },
+    {
+      "epoch": 0.06796670087809328,
+      "grad_norm": 0.3910053074359894,
+      "learning_rate": 4.999802796154388e-06,
+      "loss": 0.2065,
+      "step": 149
+    },
+    {
+      "epoch": 0.06842285323297982,
+      "grad_norm": 0.35528501868247986,
+      "learning_rate": 4.999778913687733e-06,
+      "loss": 0.2041,
+      "step": 150
+    },
+    {
+      "epoch": 0.06887900558786635,
+      "grad_norm": 0.3690823018550873,
+      "learning_rate": 4.999753666591392e-06,
+      "loss": 0.2018,
+      "step": 151
+    },
+    {
+      "epoch": 0.06933515794275288,
+      "grad_norm": 0.3650220036506653,
+      "learning_rate": 4.999727054879149e-06,
+      "loss": 0.1915,
+      "step": 152
+    },
+    {
+      "epoch": 0.06979131029763941,
+      "grad_norm": 0.36004355549812317,
+      "learning_rate": 4.99969907856553e-06,
+      "loss": 0.2073,
+      "step": 153
+    },
+    {
+      "epoch": 0.07024746265252595,
+      "grad_norm": 0.3482271134853363,
+      "learning_rate": 4.999669737665809e-06,
+      "loss": 0.1956,
+      "step": 154
+    },
+    {
+      "epoch": 0.07070361500741247,
+      "grad_norm": 0.3585127592086792,
+      "learning_rate": 4.999639032196002e-06,
+      "loss": 0.2027,
+      "step": 155
+    },
+    {
+      "epoch": 0.07115976736229901,
+      "grad_norm": 0.3586574196815491,
+      "learning_rate": 4.999606962172872e-06,
+      "loss": 0.1943,
+      "step": 156
+    },
+    {
+      "epoch": 0.07161591971718553,
+      "grad_norm": 0.35172027349472046,
+      "learning_rate": 4.999573527613925e-06,
+      "loss": 0.187,
+      "step": 157
+    },
+    {
+      "epoch": 0.07207207207207207,
+      "grad_norm": 0.3584582805633545,
+      "learning_rate": 4.999538728537414e-06,
+      "loss": 0.1962,
+      "step": 158
+    },
+    {
+      "epoch": 0.07252822442695861,
+      "grad_norm": 0.35542625188827515,
+      "learning_rate": 4.999502564962335e-06,
+      "loss": 0.1963,
+      "step": 159
+    },
+    {
+      "epoch": 0.07298437678184513,
+      "grad_norm": 0.38614457845687866,
+      "learning_rate": 4.999465036908429e-06,
+      "loss": 0.2151,
+      "step": 160
+    },
+    {
+      "epoch": 0.07344052913673167,
+      "grad_norm": 0.3604886531829834,
+      "learning_rate": 4.999426144396184e-06,
+      "loss": 0.2078,
+      "step": 161
+    },
+    {
+      "epoch": 0.0738966814916182,
+      "grad_norm": 0.35215020179748535,
+      "learning_rate": 4.999385887446831e-06,
+      "loss": 0.1968,
+      "step": 162
+    },
+    {
+      "epoch": 0.07435283384650473,
+      "grad_norm": 0.354849636554718,
+      "learning_rate": 4.999344266082348e-06,
+      "loss": 0.202,
+      "step": 163
+    },
+    {
+      "epoch": 0.07480898620139126,
+      "grad_norm": 0.36668819189071655,
+      "learning_rate": 4.999301280325452e-06,
+      "loss": 0.1977,
+      "step": 164
+    },
+    {
+      "epoch": 0.0752651385562778,
+      "grad_norm": 0.36435064673423767,
+      "learning_rate": 4.999256930199612e-06,
+      "loss": 0.1972,
+      "step": 165
+    },
+    {
+      "epoch": 0.07572129091116434,
+      "grad_norm": 0.3671411871910095,
+      "learning_rate": 4.999211215729039e-06,
+      "loss": 0.193,
+      "step": 166
+    },
+    {
+      "epoch": 0.07617744326605086,
+      "grad_norm": 0.3575402498245239,
+      "learning_rate": 4.999164136938686e-06,
+      "loss": 0.1968,
+      "step": 167
+    },
+    {
+      "epoch": 0.0766335956209374,
+      "grad_norm": 0.36870336532592773,
+      "learning_rate": 4.999115693854255e-06,
+      "loss": 0.1918,
+      "step": 168
+    },
+    {
+      "epoch": 0.07708974797582392,
+      "grad_norm": 0.3807375729084015,
+      "learning_rate": 4.999065886502192e-06,
+      "loss": 0.204,
+      "step": 169
+    },
+    {
+      "epoch": 0.07754590033071046,
+      "grad_norm": 0.3614742159843445,
+      "learning_rate": 4.999014714909684e-06,
+      "loss": 0.2034,
+      "step": 170
+    },
+    {
+      "epoch": 0.07800205268559698,
+      "grad_norm": 0.3583196997642517,
+      "learning_rate": 4.99896217910467e-06,
+      "loss": 0.1903,
+      "step": 171
+    },
+    {
+      "epoch": 0.07845820504048352,
+      "grad_norm": 0.36009538173675537,
+      "learning_rate": 4.998908279115825e-06,
+      "loss": 0.2014,
+      "step": 172
+    },
+    {
+      "epoch": 0.07891435739537005,
+      "grad_norm": 0.38533395528793335,
+      "learning_rate": 4.998853014972575e-06,
+      "loss": 0.1991,
+      "step": 173
+    },
+    {
+      "epoch": 0.07937050975025658,
+      "grad_norm": 0.3520246148109436,
+      "learning_rate": 4.9987963867050886e-06,
+      "loss": 0.1924,
+      "step": 174
+    },
+    {
+      "epoch": 0.07982666210514312,
+      "grad_norm": 0.3675670027732849,
+      "learning_rate": 4.998738394344278e-06,
+      "loss": 0.2139,
+      "step": 175
+    },
+    {
+      "epoch": 0.08028281446002965,
+      "grad_norm": 0.3831228017807007,
+      "learning_rate": 4.998679037921803e-06,
+      "loss": 0.2088,
+      "step": 176
+    },
+    {
+      "epoch": 0.08073896681491619,
+      "grad_norm": 0.36646074056625366,
+      "learning_rate": 4.998618317470065e-06,
+      "loss": 0.2009,
+      "step": 177
+    },
+    {
+      "epoch": 0.08119511916980271,
+      "grad_norm": 0.36600837111473083,
+      "learning_rate": 4.998556233022212e-06,
+      "loss": 0.1935,
+      "step": 178
+    },
+    {
+      "epoch": 0.08165127152468925,
+      "grad_norm": 0.3981282711029053,
+      "learning_rate": 4.998492784612137e-06,
+      "loss": 0.2069,
+      "step": 179
+    },
+    {
+      "epoch": 0.08210742387957577,
+      "grad_norm": 0.3594295084476471,
+      "learning_rate": 4.998427972274473e-06,
+      "loss": 0.1938,
+      "step": 180
+    },
+    {
+      "epoch": 0.08256357623446231,
+      "grad_norm": 0.3490099608898163,
+      "learning_rate": 4.998361796044605e-06,
+      "loss": 0.1951,
+      "step": 181
+    },
+    {
+      "epoch": 0.08301972858934885,
+      "grad_norm": 0.34965503215789795,
+      "learning_rate": 4.998294255958656e-06,
+      "loss": 0.1943,
+      "step": 182
+    },
+    {
+      "epoch": 0.08347588094423537,
+      "grad_norm": 0.3506588041782379,
+      "learning_rate": 4.998225352053497e-06,
+      "loss": 0.1948,
+      "step": 183
+    },
+    {
+      "epoch": 0.08393203329912191,
+      "grad_norm": 0.3694084584712982,
+      "learning_rate": 4.998155084366744e-06,
+      "loss": 0.1947,
+      "step": 184
+    },
+    {
+      "epoch": 0.08438818565400844,
+      "grad_norm": 0.36333993077278137,
+      "learning_rate": 4.9980834529367525e-06,
+      "loss": 0.1951,
+      "step": 185
+    },
+    {
+      "epoch": 0.08484433800889497,
+      "grad_norm": 0.38360416889190674,
+      "learning_rate": 4.99801045780263e-06,
+      "loss": 0.2103,
+      "step": 186
+    },
+    {
+      "epoch": 0.0853004903637815,
+      "grad_norm": 0.32897692918777466,
+      "learning_rate": 4.997936099004222e-06,
+      "loss": 0.2006,
+      "step": 187
+    },
+    {
+      "epoch": 0.08575664271866804,
+      "grad_norm": 0.34873566031455994,
+      "learning_rate": 4.997860376582123e-06,
+      "loss": 0.1908,
+      "step": 188
+    },
+    {
+      "epoch": 0.08621279507355457,
+      "grad_norm": 0.37413525581359863,
+      "learning_rate": 4.997783290577669e-06,
+      "loss": 0.1984,
+      "step": 189
+    },
+    {
+      "epoch": 0.0866689474284411,
+      "grad_norm": 0.35462263226509094,
+      "learning_rate": 4.997704841032941e-06,
+      "loss": 0.1996,
+      "step": 190
+    },
+    {
+      "epoch": 0.08712509978332764,
+      "grad_norm": 0.3693375885486603,
+      "learning_rate": 4.997625027990763e-06,
+      "loss": 0.2102,
+      "step": 191
+    },
+    {
+      "epoch": 0.08758125213821416,
+      "grad_norm": 0.35890457034111023,
+      "learning_rate": 4.997543851494709e-06,
+      "loss": 0.198,
+      "step": 192
+    },
+    {
+      "epoch": 0.0880374044931007,
+      "grad_norm": 0.34527066349983215,
+      "learning_rate": 4.997461311589089e-06,
+      "loss": 0.1993,
+      "step": 193
+    },
+    {
+      "epoch": 0.08849355684798722,
+      "grad_norm": 0.3605559766292572,
+      "learning_rate": 4.997377408318965e-06,
+      "loss": 0.199,
+      "step": 194
+    },
+    {
+      "epoch": 0.08894970920287376,
+      "grad_norm": 0.3567296266555786,
+      "learning_rate": 4.997292141730137e-06,
+      "loss": 0.1994,
+      "step": 195
+    },
+    {
+      "epoch": 0.08940586155776029,
+      "grad_norm": 0.3458857834339142,
+      "learning_rate": 4.9972055118691545e-06,
+      "loss": 0.1827,
+      "step": 196
+    },
+    {
+      "epoch": 0.08986201391264682,
+      "grad_norm": 0.34776148200035095,
+      "learning_rate": 4.997117518783307e-06,
+      "loss": 0.2079,
+      "step": 197
+    },
+    {
+      "epoch": 0.09031816626753336,
+      "grad_norm": 0.37243497371673584,
+      "learning_rate": 4.9970281625206306e-06,
+      "loss": 0.2142,
+      "step": 198
+    },
+    {
+      "epoch": 0.09077431862241989,
+      "grad_norm": 0.37486183643341064,
+      "learning_rate": 4.996937443129904e-06,
+      "loss": 0.1966,
+      "step": 199
+    },
+    {
+      "epoch": 0.09123047097730642,
+      "grad_norm": 0.34183305501937866,
+      "learning_rate": 4.996845360660652e-06,
+      "loss": 0.1956,
+      "step": 200
+    },
+    {
+      "epoch": 0.09168662333219295,
+      "grad_norm": 0.33284252882003784,
+      "learning_rate": 4.996751915163143e-06,
+      "loss": 0.1886,
+      "step": 201
+    },
+    {
+      "epoch": 0.09214277568707949,
+      "grad_norm": 0.3514299690723419,
+      "learning_rate": 4.996657106688388e-06,
+      "loss": 0.2045,
+      "step": 202
+    },
+    {
+      "epoch": 0.09259892804196601,
+      "grad_norm": 0.34648382663726807,
+      "learning_rate": 4.996560935288143e-06,
+      "loss": 0.1897,
+      "step": 203
+    },
+    {
+      "epoch": 0.09305508039685255,
+      "grad_norm": 0.3549829423427582,
+      "learning_rate": 4.996463401014908e-06,
+      "loss": 0.2049,
+      "step": 204
+    },
+    {
+      "epoch": 0.09351123275173909,
+      "grad_norm": 0.3465183675289154,
+      "learning_rate": 4.9963645039219266e-06,
+      "loss": 0.1996,
+      "step": 205
+    },
+    {
+      "epoch": 0.09396738510662561,
+      "grad_norm": 0.34694546461105347,
+      "learning_rate": 4.996264244063187e-06,
+      "loss": 0.1999,
+      "step": 206
+    },
+    {
+      "epoch": 0.09442353746151215,
+      "grad_norm": 0.3358584940433502,
+      "learning_rate": 4.9961626214934225e-06,
+      "loss": 0.19,
+      "step": 207
+    },
+    {
+      "epoch": 0.09487968981639867,
+      "grad_norm": 0.3806498944759369,
+      "learning_rate": 4.9960596362681054e-06,
+      "loss": 0.193,
+      "step": 208
+    },
+    {
+      "epoch": 0.09533584217128521,
+      "grad_norm": 0.343035489320755,
+      "learning_rate": 4.9959552884434596e-06,
+      "loss": 0.1974,
+      "step": 209
+    },
+    {
+      "epoch": 0.09579199452617174,
+      "grad_norm": 0.3670075833797455,
+      "learning_rate": 4.995849578076446e-06,
+      "loss": 0.1947,
+      "step": 210
+    },
+    {
+      "epoch": 0.09624814688105827,
+      "grad_norm": 0.3301618993282318,
+      "learning_rate": 4.995742505224772e-06,
+      "loss": 0.1888,
+      "step": 211
+    },
+    {
+      "epoch": 0.0967042992359448,
+      "grad_norm": 0.3443339169025421,
+      "learning_rate": 4.9956340699468896e-06,
+      "loss": 0.1914,
+      "step": 212
+    },
+    {
+      "epoch": 0.09716045159083134,
+      "grad_norm": 0.3581489622592926,
+      "learning_rate": 4.9955242723019925e-06,
+      "loss": 0.1915,
+      "step": 213
+    },
+    {
+      "epoch": 0.09761660394571788,
+      "grad_norm": 0.37551987171173096,
+      "learning_rate": 4.995413112350021e-06,
+      "loss": 0.1974,
+      "step": 214
+    },
+    {
+      "epoch": 0.0980727563006044,
+      "grad_norm": 0.34931662678718567,
+      "learning_rate": 4.9953005901516546e-06,
+      "loss": 0.2001,
+      "step": 215
+    },
+    {
+      "epoch": 0.09852890865549094,
+      "grad_norm": 0.3806745707988739,
+      "learning_rate": 4.995186705768322e-06,
+      "loss": 0.2005,
+      "step": 216
+    },
+    {
+      "epoch": 0.09898506101037746,
+      "grad_norm": 0.33778902888298035,
+      "learning_rate": 4.995071459262191e-06,
+      "loss": 0.1875,
+      "step": 217
+    },
+    {
+      "epoch": 0.099441213365264,
+      "grad_norm": 0.35212385654449463,
+      "learning_rate": 4.994954850696176e-06,
+      "loss": 0.1931,
+      "step": 218
+    },
+    {
+      "epoch": 0.09989736572015052,
+      "grad_norm": 0.3523644506931305,
+      "learning_rate": 4.994836880133932e-06,
+      "loss": 0.1909,
+      "step": 219
+    },
+    {
+      "epoch": 0.10035351807503706,
+      "grad_norm": 0.3814811706542969,
+      "learning_rate": 4.9947175476398606e-06,
+      "loss": 0.2154,
+      "step": 220
+    },
+    {
+      "epoch": 0.1008096704299236,
+      "grad_norm": 0.371563583612442,
+      "learning_rate": 4.994596853279105e-06,
+      "loss": 0.1935,
+      "step": 221
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.34389588236808777,
+      "learning_rate": 4.994474797117551e-06,
+      "loss": 0.1923,
+      "step": 222
+    },
+    {
+      "epoch": 0.10172197513969666,
+      "grad_norm": 0.353318452835083,
+      "learning_rate": 4.994351379221831e-06,
+      "loss": 0.1907,
+      "step": 223
+    },
+    {
+      "epoch": 0.10217812749458319,
+      "grad_norm": 0.3768618106842041,
+      "learning_rate": 4.994226599659319e-06,
+      "loss": 0.2019,
+      "step": 224
+    },
+    {
+      "epoch": 0.10263427984946973,
+      "grad_norm": 0.36510515213012695,
+      "learning_rate": 4.99410045849813e-06,
+      "loss": 0.2002,
+      "step": 225
+    },
+    {
+      "epoch": 0.10309043220435625,
+      "grad_norm": 0.3604632019996643,
+      "learning_rate": 4.993972955807125e-06,
+      "loss": 0.2032,
+      "step": 226
+    },
+    {
+      "epoch": 0.10354658455924279,
+      "grad_norm": 0.3545345366001129,
+      "learning_rate": 4.99384409165591e-06,
+      "loss": 0.1957,
+      "step": 227
+    },
+    {
+      "epoch": 0.10400273691412931,
+      "grad_norm": 0.3762911260128021,
+      "learning_rate": 4.993713866114829e-06,
+      "loss": 0.1978,
+      "step": 228
+    },
+    {
+      "epoch": 0.10445888926901585,
+      "grad_norm": 0.3659232556819916,
+      "learning_rate": 4.993582279254974e-06,
+      "loss": 0.1986,
+      "step": 229
+    },
+    {
+      "epoch": 0.10491504162390239,
+      "grad_norm": 0.3906863033771515,
+      "learning_rate": 4.993449331148178e-06,
+      "loss": 0.2126,
+      "step": 230
+    },
+    {
+      "epoch": 0.10537119397878891,
+      "grad_norm": 0.3383077383041382,
+      "learning_rate": 4.993315021867018e-06,
+      "loss": 0.2035,
+      "step": 231
+    },
+    {
+      "epoch": 0.10582734633367545,
+      "grad_norm": 0.35205110907554626,
+      "learning_rate": 4.993179351484811e-06,
+      "loss": 0.1862,
+      "step": 232
+    },
+    {
+      "epoch": 0.10628349868856198,
+      "grad_norm": 0.3486587107181549,
+      "learning_rate": 4.993042320075622e-06,
+      "loss": 0.193,
+      "step": 233
+    },
+    {
+      "epoch": 0.10673965104344851,
+      "grad_norm": 0.365501344203949,
+      "learning_rate": 4.992903927714255e-06,
+      "loss": 0.1888,
+      "step": 234
+    },
+    {
+      "epoch": 0.10719580339833504,
+      "grad_norm": 0.34221065044403076,
+      "learning_rate": 4.99276417447626e-06,
+      "loss": 0.1955,
+      "step": 235
+    },
+    {
+      "epoch": 0.10765195575322158,
+      "grad_norm": 0.34505724906921387,
+      "learning_rate": 4.9926230604379275e-06,
+      "loss": 0.1878,
+      "step": 236
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 0.314775288105011,
+      "learning_rate": 4.99248058567629e-06,
+      "loss": 0.1718,
+      "step": 237
+    },
+    {
+      "epoch": 0.10856426046299464,
+      "grad_norm": 0.3540940284729004,
+      "learning_rate": 4.992336750269128e-06,
+      "loss": 0.1986,
+      "step": 238
+    },
+    {
+      "epoch": 0.10902041281788118,
+      "grad_norm": 0.3652269244194031,
+      "learning_rate": 4.992191554294958e-06,
+      "loss": 0.199,
+      "step": 239
+    },
+    {
+      "epoch": 0.1094765651727677,
+      "grad_norm": 0.3575532138347626,
+      "learning_rate": 4.992044997833044e-06,
+      "loss": 0.2039,
+      "step": 240
+    },
+    {
+      "epoch": 0.10993271752765424,
+      "grad_norm": 0.3454466462135315,
+      "learning_rate": 4.991897080963391e-06,
+      "loss": 0.196,
+      "step": 241
+    },
+    {
+      "epoch": 0.11038886988254076,
+      "grad_norm": 0.3385619819164276,
+      "learning_rate": 4.991747803766747e-06,
+      "loss": 0.1927,
+      "step": 242
+    },
+    {
+      "epoch": 0.1108450222374273,
+      "grad_norm": 0.3477875888347626,
+      "learning_rate": 4.9915971663246015e-06,
+      "loss": 0.1891,
+      "step": 243
+    },
+    {
+      "epoch": 0.11130117459231384,
+      "grad_norm": 0.35544857382774353,
+      "learning_rate": 4.991445168719189e-06,
+      "loss": 0.2023,
+      "step": 244
+    },
+    {
+      "epoch": 0.11175732694720036,
+      "grad_norm": 0.3432496190071106,
+      "learning_rate": 4.991291811033484e-06,
+      "loss": 0.2062,
+      "step": 245
+    },
+    {
+      "epoch": 0.1122134793020869,
+      "grad_norm": 0.3488253355026245,
+      "learning_rate": 4.9911370933512046e-06,
+      "loss": 0.1973,
+      "step": 246
+    },
+    {
+      "epoch": 0.11266963165697343,
+      "grad_norm": 0.3326709568500519,
+      "learning_rate": 4.990981015756812e-06,
+      "loss": 0.1907,
+      "step": 247
+    },
+    {
+      "epoch": 0.11312578401185996,
+      "grad_norm": 0.3441293239593506,
+      "learning_rate": 4.9908235783355075e-06,
+      "loss": 0.1925,
+      "step": 248
+    },
+    {
+      "epoch": 0.11358193636674649,
+      "grad_norm": 0.32878321409225464,
+      "learning_rate": 4.990664781173237e-06,
+      "loss": 0.1856,
+      "step": 249
+    },
+    {
+      "epoch": 0.11403808872163303,
+      "grad_norm": 0.3548767864704132,
+      "learning_rate": 4.990504624356689e-06,
+      "loss": 0.1931,
+      "step": 250
+    },
+    {
+      "epoch": 0.11449424107651955,
+      "grad_norm": 0.34791386127471924,
+      "learning_rate": 4.990343107973291e-06,
+      "loss": 0.1881,
+      "step": 251
+    },
+    {
+      "epoch": 0.11495039343140609,
+      "grad_norm": 0.3431399464607239,
+      "learning_rate": 4.990180232111217e-06,
+      "loss": 0.1953,
+      "step": 252
+    },
+    {
+      "epoch": 0.11540654578629263,
+      "grad_norm": 0.3286855220794678,
+      "learning_rate": 4.990015996859379e-06,
+      "loss": 0.1774,
+      "step": 253
+    },
+    {
+      "epoch": 0.11586269814117915,
+      "grad_norm": 0.3628874719142914,
+      "learning_rate": 4.989850402307434e-06,
+      "loss": 0.1856,
+      "step": 254
+    },
+    {
+      "epoch": 0.11631885049606569,
+      "grad_norm": 0.3715766668319702,
+      "learning_rate": 4.9896834485457805e-06,
+      "loss": 0.1919,
+      "step": 255
+    },
+    {
+      "epoch": 0.11677500285095221,
+      "grad_norm": 0.3670489490032196,
+      "learning_rate": 4.989515135665558e-06,
+      "loss": 0.2053,
+      "step": 256
+    },
+    {
+      "epoch": 0.11723115520583875,
+      "grad_norm": 0.3520582616329193,
+      "learning_rate": 4.989345463758648e-06,
+      "loss": 0.1937,
+      "step": 257
+    },
+    {
+      "epoch": 0.11768730756072528,
+      "grad_norm": 0.35874396562576294,
+      "learning_rate": 4.989174432917675e-06,
+      "loss": 0.1927,
+      "step": 258
+    },
+    {
+      "epoch": 0.11814345991561181,
+      "grad_norm": 0.3420073091983795,
+      "learning_rate": 4.989002043236005e-06,
+      "loss": 0.1891,
+      "step": 259
+    },
+    {
+      "epoch": 0.11859961227049835,
+      "grad_norm": 0.33540236949920654,
+      "learning_rate": 4.988828294807746e-06,
+      "loss": 0.1945,
+      "step": 260
+    },
+    {
+      "epoch": 0.11905576462538488,
+      "grad_norm": 0.3865305781364441,
+      "learning_rate": 4.988653187727745e-06,
+      "loss": 0.2027,
+      "step": 261
+    },
+    {
+      "epoch": 0.11951191698027142,
+      "grad_norm": 0.3342127501964569,
+      "learning_rate": 4.988476722091596e-06,
+      "loss": 0.19,
+      "step": 262
+    },
+    {
+      "epoch": 0.11996806933515794,
+      "grad_norm": 0.34244269132614136,
+      "learning_rate": 4.98829889799563e-06,
+      "loss": 0.1938,
+      "step": 263
+    },
+    {
+      "epoch": 0.12042422169004448,
+      "grad_norm": 0.3308809995651245,
+      "learning_rate": 4.988119715536922e-06,
+      "loss": 0.1848,
+      "step": 264
+    },
+    {
+      "epoch": 0.120880374044931,
+      "grad_norm": 0.33956554532051086,
+      "learning_rate": 4.9879391748132865e-06,
+      "loss": 0.1941,
+      "step": 265
+    },
+    {
+      "epoch": 0.12133652639981754,
+      "grad_norm": 0.3551222085952759,
+      "learning_rate": 4.987757275923282e-06,
+      "loss": 0.1887,
+      "step": 266
+    },
+    {
+      "epoch": 0.12179267875470406,
+      "grad_norm": 0.31548917293548584,
+      "learning_rate": 4.987574018966205e-06,
+      "loss": 0.1745,
+      "step": 267
+    },
+    {
+      "epoch": 0.1222488311095906,
+      "grad_norm": 0.32318633794784546,
+      "learning_rate": 4.987389404042098e-06,
+      "loss": 0.1873,
+      "step": 268
+    },
+    {
+      "epoch": 0.12270498346447714,
+      "grad_norm": 0.34276658296585083,
+      "learning_rate": 4.987203431251741e-06,
+      "loss": 0.1971,
+      "step": 269
+    },
+    {
+      "epoch": 0.12316113581936367,
+      "grad_norm": 0.35004177689552307,
+      "learning_rate": 4.987016100696658e-06,
+      "loss": 0.1841,
+      "step": 270
+    },
+    {
+      "epoch": 0.1236172881742502,
+      "grad_norm": 0.33097681403160095,
+      "learning_rate": 4.986827412479111e-06,
+      "loss": 0.1805,
+      "step": 271
+    },
+    {
+      "epoch": 0.12407344052913673,
+      "grad_norm": 0.3408569097518921,
+      "learning_rate": 4.986637366702105e-06,
+      "loss": 0.1943,
+      "step": 272
+    },
+    {
+      "epoch": 0.12452959288402327,
+      "grad_norm": 0.3659879267215729,
+      "learning_rate": 4.986445963469388e-06,
+      "loss": 0.2015,
+      "step": 273
+    },
+    {
+      "epoch": 0.12498574523890979,
+      "grad_norm": 0.3407228887081146,
+      "learning_rate": 4.986253202885444e-06,
+      "loss": 0.1954,
+      "step": 274
+    },
+    {
+      "epoch": 0.12544189759379631,
+      "grad_norm": 0.33964788913726807,
+      "learning_rate": 4.986059085055504e-06,
+      "loss": 0.1777,
+      "step": 275
+    },
+    {
+      "epoch": 0.12589804994868287,
+      "grad_norm": 0.3378674387931824,
+      "learning_rate": 4.985863610085534e-06,
+      "loss": 0.1822,
+      "step": 276
+    },
+    {
+      "epoch": 0.1263542023035694,
+      "grad_norm": 0.3634103238582611,
+      "learning_rate": 4.985666778082247e-06,
+      "loss": 0.1958,
+      "step": 277
+    },
+    {
+      "epoch": 0.12681035465845591,
+      "grad_norm": 0.35330814123153687,
+      "learning_rate": 4.985468589153092e-06,
+      "loss": 0.1871,
+      "step": 278
+    },
+    {
+      "epoch": 0.12726650701334247,
+      "grad_norm": 0.33935442566871643,
+      "learning_rate": 4.98526904340626e-06,
+      "loss": 0.1818,
+      "step": 279
+    },
+    {
+      "epoch": 0.127722659368229,
+      "grad_norm": 0.3432380259037018,
+      "learning_rate": 4.985068140950683e-06,
+      "loss": 0.1871,
+      "step": 280
+    },
+    {
+      "epoch": 0.12817881172311552,
+      "grad_norm": 0.3669479489326477,
+      "learning_rate": 4.984865881896035e-06,
+      "loss": 0.1925,
+      "step": 281
+    },
+    {
+      "epoch": 0.12863496407800204,
+      "grad_norm": 0.3525663912296295,
+      "learning_rate": 4.984662266352728e-06,
+      "loss": 0.2012,
+      "step": 282
+    },
+    {
+      "epoch": 0.1290911164328886,
+      "grad_norm": 0.34177345037460327,
+      "learning_rate": 4.984457294431917e-06,
+      "loss": 0.1952,
+      "step": 283
+    },
+    {
+      "epoch": 0.12954726878777512,
+      "grad_norm": 0.35861608386039734,
+      "learning_rate": 4.984250966245495e-06,
+      "loss": 0.203,
+      "step": 284
+    },
+    {
+      "epoch": 0.13000342114266164,
+      "grad_norm": 0.3607324957847595,
+      "learning_rate": 4.984043281906096e-06,
+      "loss": 0.1992,
+      "step": 285
+    },
+    {
+      "epoch": 0.1304595734975482,
+      "grad_norm": 0.3345677852630615,
+      "learning_rate": 4.983834241527097e-06,
+      "loss": 0.1893,
+      "step": 286
+    },
+    {
+      "epoch": 0.13091572585243472,
+      "grad_norm": 0.3196655511856079,
+      "learning_rate": 4.983623845222611e-06,
+      "loss": 0.1833,
+      "step": 287
+    },
+    {
+      "epoch": 0.13137187820732124,
+      "grad_norm": 0.33671531081199646,
+      "learning_rate": 4.983412093107496e-06,
+      "loss": 0.1837,
+      "step": 288
+    },
+    {
+      "epoch": 0.13182803056220777,
+      "grad_norm": 0.34064242243766785,
+      "learning_rate": 4.983198985297345e-06,
+      "loss": 0.1807,
+      "step": 289
+    },
+    {
+      "epoch": 0.13228418291709432,
+      "grad_norm": 0.3591769337654114,
+      "learning_rate": 4.982984521908495e-06,
+      "loss": 0.1984,
+      "step": 290
+    },
+    {
+      "epoch": 0.13274033527198084,
+      "grad_norm": 0.3471721410751343,
+      "learning_rate": 4.9827687030580205e-06,
+      "loss": 0.1951,
+      "step": 291
+    },
+    {
+      "epoch": 0.13319648762686737,
+      "grad_norm": 0.35906657576560974,
+      "learning_rate": 4.982551528863738e-06,
+      "loss": 0.1934,
+      "step": 292
+    },
+    {
+      "epoch": 0.13365263998175392,
+      "grad_norm": 0.3696458637714386,
+      "learning_rate": 4.982332999444203e-06,
+      "loss": 0.1977,
+      "step": 293
+    },
+    {
+      "epoch": 0.13410879233664044,
+      "grad_norm": 0.35301026701927185,
+      "learning_rate": 4.98211311491871e-06,
+      "loss": 0.1823,
+      "step": 294
+    },
+    {
+      "epoch": 0.13456494469152697,
+      "grad_norm": 0.32495495676994324,
+      "learning_rate": 4.981891875407294e-06,
+      "loss": 0.1908,
+      "step": 295
+    },
+    {
+      "epoch": 0.1350210970464135,
+      "grad_norm": 0.3348126709461212,
+      "learning_rate": 4.981669281030731e-06,
+      "loss": 0.2015,
+      "step": 296
+    },
+    {
+      "epoch": 0.13547724940130004,
+      "grad_norm": 0.34072747826576233,
+      "learning_rate": 4.981445331910534e-06,
+      "loss": 0.1862,
+      "step": 297
+    },
+    {
+      "epoch": 0.13593340175618657,
+      "grad_norm": 0.3544979393482208,
+      "learning_rate": 4.981220028168957e-06,
+      "loss": 0.1951,
+      "step": 298
+    },
+    {
+      "epoch": 0.1363895541110731,
+      "grad_norm": 0.3534928262233734,
+      "learning_rate": 4.980993369928993e-06,
+      "loss": 0.208,
+      "step": 299
+    },
+    {
+      "epoch": 0.13684570646595964,
+      "grad_norm": 0.33685001730918884,
+      "learning_rate": 4.980765357314376e-06,
+      "loss": 0.1966,
+      "step": 300
+    },
+    {
+      "epoch": 0.13730185882084617,
+      "grad_norm": 0.33794280886650085,
+      "learning_rate": 4.980535990449577e-06,
+      "loss": 0.1945,
+      "step": 301
+    },
+    {
+      "epoch": 0.1377580111757327,
+      "grad_norm": 0.3310077488422394,
+      "learning_rate": 4.9803052694598075e-06,
+      "loss": 0.1865,
+      "step": 302
+    },
+    {
+      "epoch": 0.13821416353061922,
+      "grad_norm": 0.3224427402019501,
+      "learning_rate": 4.980073194471019e-06,
+      "loss": 0.1783,
+      "step": 303
+    },
+    {
+      "epoch": 0.13867031588550577,
+      "grad_norm": 0.3626343607902527,
+      "learning_rate": 4.9798397656099005e-06,
+      "loss": 0.1854,
+      "step": 304
+    },
+    {
+      "epoch": 0.1391264682403923,
+      "grad_norm": 0.330517053604126,
+      "learning_rate": 4.979604983003882e-06,
+      "loss": 0.1915,
+      "step": 305
+    },
+    {
+      "epoch": 0.13958262059527882,
+      "grad_norm": 0.3335297107696533,
+      "learning_rate": 4.9793688467811295e-06,
+      "loss": 0.1903,
+      "step": 306
+    },
+    {
+      "epoch": 0.14003877295016537,
+      "grad_norm": 0.3258552551269531,
+      "learning_rate": 4.979131357070551e-06,
+      "loss": 0.1804,
+      "step": 307
+    },
+    {
+      "epoch": 0.1404949253050519,
+      "grad_norm": 0.33972159028053284,
+      "learning_rate": 4.978892514001792e-06,
+      "loss": 0.1819,
+      "step": 308
+    },
+    {
+      "epoch": 0.14095107765993842,
+      "grad_norm": 0.3701869249343872,
+      "learning_rate": 4.978652317705236e-06,
+      "loss": 0.2048,
+      "step": 309
+    },
+    {
+      "epoch": 0.14140723001482494,
+      "grad_norm": 0.36860236525535583,
+      "learning_rate": 4.978410768312008e-06,
+      "loss": 0.197,
+      "step": 310
+    },
+    {
+      "epoch": 0.1418633823697115,
+      "grad_norm": 0.3520394563674927,
+      "learning_rate": 4.978167865953969e-06,
+      "loss": 0.1996,
+      "step": 311
+    },
+    {
+      "epoch": 0.14231953472459802,
+      "grad_norm": 0.3518396019935608,
+      "learning_rate": 4.977923610763719e-06,
+      "loss": 0.2086,
+      "step": 312
+    },
+    {
+      "epoch": 0.14277568707948454,
+      "grad_norm": 0.32982707023620605,
+      "learning_rate": 4.977678002874597e-06,
+      "loss": 0.1859,
+      "step": 313
+    },
+    {
+      "epoch": 0.14323183943437107,
+      "grad_norm": 0.32479897141456604,
+      "learning_rate": 4.977431042420681e-06,
+      "loss": 0.185,
+      "step": 314
+    },
+    {
+      "epoch": 0.14368799178925762,
+      "grad_norm": 0.35447508096694946,
+      "learning_rate": 4.977182729536787e-06,
+      "loss": 0.1926,
+      "step": 315
+    },
+    {
+      "epoch": 0.14414414414414414,
+      "grad_norm": 0.3521266281604767,
+      "learning_rate": 4.976933064358467e-06,
+      "loss": 0.1821,
+      "step": 316
+    },
+    {
+      "epoch": 0.14460029649903067,
+      "grad_norm": 0.35105085372924805,
+      "learning_rate": 4.976682047022016e-06,
+      "loss": 0.1931,
+      "step": 317
+    },
+    {
+      "epoch": 0.14505644885391722,
+      "grad_norm": 0.36129453778266907,
+      "learning_rate": 4.976429677664463e-06,
+      "loss": 0.1894,
+      "step": 318
+    },
+    {
+      "epoch": 0.14551260120880374,
+      "grad_norm": 0.330276757478714,
+      "learning_rate": 4.976175956423576e-06,
+      "loss": 0.1858,
+      "step": 319
+    },
+    {
+      "epoch": 0.14596875356369027,
+      "grad_norm": 0.3706572651863098,
+      "learning_rate": 4.975920883437862e-06,
+      "loss": 0.2042,
+      "step": 320
+    },
+    {
+      "epoch": 0.1464249059185768,
+      "grad_norm": 0.34393006563186646,
+      "learning_rate": 4.975664458846566e-06,
+      "loss": 0.1849,
+      "step": 321
+    },
+    {
+      "epoch": 0.14688105827346334,
+      "grad_norm": 0.36969077587127686,
+      "learning_rate": 4.9754066827896695e-06,
+      "loss": 0.1997,
+      "step": 322
+    },
+    {
+      "epoch": 0.14733721062834987,
+      "grad_norm": 0.34701862931251526,
+      "learning_rate": 4.975147555407893e-06,
+      "loss": 0.1907,
+      "step": 323
+    },
+    {
+      "epoch": 0.1477933629832364,
+      "grad_norm": 0.3762175440788269,
+      "learning_rate": 4.974887076842694e-06,
+      "loss": 0.2018,
+      "step": 324
+    },
+    {
+      "epoch": 0.14824951533812294,
+      "grad_norm": 0.3407662808895111,
+      "learning_rate": 4.974625247236267e-06,
+      "loss": 0.1997,
+      "step": 325
+    },
+    {
+      "epoch": 0.14870566769300947,
+      "grad_norm": 0.3483238220214844,
+      "learning_rate": 4.974362066731545e-06,
+      "loss": 0.1881,
+      "step": 326
+    },
+    {
+      "epoch": 0.149161820047896,
+      "grad_norm": 0.3363865315914154,
+      "learning_rate": 4.9740975354722e-06,
+      "loss": 0.2006,
+      "step": 327
+    },
+    {
+      "epoch": 0.14961797240278252,
+      "grad_norm": 0.3370267450809479,
+      "learning_rate": 4.973831653602637e-06,
+      "loss": 0.1859,
+      "step": 328
+    },
+    {
+      "epoch": 0.15007412475766907,
+      "grad_norm": 0.3611241281032562,
+      "learning_rate": 4.9735644212680035e-06,
+      "loss": 0.2005,
+      "step": 329
+    },
+    {
+      "epoch": 0.1505302771125556,
+      "grad_norm": 0.3441029489040375,
+      "learning_rate": 4.973295838614179e-06,
+      "loss": 0.189,
+      "step": 330
+    },
+    {
+      "epoch": 0.15098642946744212,
+      "grad_norm": 0.3048926591873169,
+      "learning_rate": 4.973025905787786e-06,
+      "loss": 0.1789,
+      "step": 331
+    },
+    {
+      "epoch": 0.15144258182232867,
+      "grad_norm": 0.3375416100025177,
+      "learning_rate": 4.972754622936178e-06,
+      "loss": 0.2078,
+      "step": 332
+    },
+    {
+      "epoch": 0.1518987341772152,
+      "grad_norm": 0.34900805354118347,
+      "learning_rate": 4.97248199020745e-06,
+      "loss": 0.1985,
+      "step": 333
+    },
+    {
+      "epoch": 0.15235488653210172,
+      "grad_norm": 0.33640843629837036,
+      "learning_rate": 4.972208007750432e-06,
+      "loss": 0.1902,
+      "step": 334
+    },
+    {
+      "epoch": 0.15281103888698824,
+      "grad_norm": 0.35257476568222046,
+      "learning_rate": 4.97193267571469e-06,
+      "loss": 0.1911,
+      "step": 335
+    },
+    {
+      "epoch": 0.1532671912418748,
+      "grad_norm": 0.34630295634269714,
+      "learning_rate": 4.971655994250529e-06,
+      "loss": 0.2064,
+      "step": 336
+    },
+    {
+      "epoch": 0.15372334359676132,
+      "grad_norm": 0.3275966942310333,
+      "learning_rate": 4.97137796350899e-06,
+      "loss": 0.1898,
+      "step": 337
+    },
+    {
+      "epoch": 0.15417949595164784,
+      "grad_norm": 0.32383328676223755,
+      "learning_rate": 4.971098583641849e-06,
+      "loss": 0.1795,
+      "step": 338
+    },
+    {
+      "epoch": 0.1546356483065344,
+      "grad_norm": 0.31406620144844055,
+      "learning_rate": 4.970817854801618e-06,
+      "loss": 0.1864,
+      "step": 339
+    },
+    {
+      "epoch": 0.15509180066142092,
+      "grad_norm": 0.3566049337387085,
+      "learning_rate": 4.97053577714155e-06,
+      "loss": 0.1955,
+      "step": 340
+    },
+    {
+      "epoch": 0.15554795301630744,
+      "grad_norm": 0.3355313241481781,
+      "learning_rate": 4.97025235081563e-06,
+      "loss": 0.1869,
+      "step": 341
+    },
+    {
+      "epoch": 0.15600410537119397,
+      "grad_norm": 0.35014480352401733,
+      "learning_rate": 4.969967575978579e-06,
+      "loss": 0.1919,
+      "step": 342
+    },
+    {
+      "epoch": 0.15646025772608052,
+      "grad_norm": 0.32399502396583557,
+      "learning_rate": 4.969681452785857e-06,
+      "loss": 0.1803,
+      "step": 343
+    },
+    {
+      "epoch": 0.15691641008096704,
+      "grad_norm": 0.32772696018218994,
+      "learning_rate": 4.96939398139366e-06,
+      "loss": 0.2029,
+      "step": 344
+    },
+    {
+      "epoch": 0.15737256243585357,
+      "grad_norm": 0.3334849178791046,
+      "learning_rate": 4.969105161958916e-06,
+      "loss": 0.1916,
+      "step": 345
+    },
+    {
+      "epoch": 0.1578287147907401,
+      "grad_norm": 0.3391447365283966,
+      "learning_rate": 4.968814994639293e-06,
+      "loss": 0.1975,
+      "step": 346
+    },
+    {
+      "epoch": 0.15828486714562665,
+      "grad_norm": 0.35992148518562317,
+      "learning_rate": 4.968523479593193e-06,
+      "loss": 0.2009,
+      "step": 347
+    },
+    {
+      "epoch": 0.15874101950051317,
+      "grad_norm": 0.33622291684150696,
+      "learning_rate": 4.968230616979755e-06,
+      "loss": 0.1883,
+      "step": 348
+    },
+    {
+      "epoch": 0.1591971718553997,
+      "grad_norm": 0.34112611413002014,
+      "learning_rate": 4.967936406958852e-06,
+      "loss": 0.1956,
+      "step": 349
+    },
+    {
+      "epoch": 0.15965332421028625,
+      "grad_norm": 0.35878798365592957,
+      "learning_rate": 4.967640849691093e-06,
+      "loss": 0.1921,
+      "step": 350
+    },
+    {
+      "epoch": 0.16010947656517277,
+      "grad_norm": 0.35000935196876526,
+      "learning_rate": 4.967343945337823e-06,
+      "loss": 0.1999,
+      "step": 351
+    },
+    {
+      "epoch": 0.1605656289200593,
+      "grad_norm": 0.3344356119632721,
+      "learning_rate": 4.967045694061122e-06,
+      "loss": 0.1817,
+      "step": 352
+    },
+    {
+      "epoch": 0.16102178127494582,
+      "grad_norm": 0.3495728671550751,
+      "learning_rate": 4.9667460960238064e-06,
+      "loss": 0.1855,
+      "step": 353
+    },
+    {
+      "epoch": 0.16147793362983237,
+      "grad_norm": 0.3327745199203491,
+      "learning_rate": 4.966445151389426e-06,
+      "loss": 0.1945,
+      "step": 354
+    },
+    {
+      "epoch": 0.1619340859847189,
+      "grad_norm": 0.3585168421268463,
+      "learning_rate": 4.966142860322266e-06,
+      "loss": 0.2013,
+      "step": 355
+    },
+    {
+      "epoch": 0.16239023833960542,
+      "grad_norm": 0.32381704449653625,
+      "learning_rate": 4.965839222987348e-06,
+      "loss": 0.1855,
+      "step": 356
+    },
+    {
+      "epoch": 0.16284639069449197,
+      "grad_norm": 0.3340785801410675,
+      "learning_rate": 4.965534239550428e-06,
+      "loss": 0.1921,
+      "step": 357
+    },
+    {
+      "epoch": 0.1633025430493785,
+      "grad_norm": 0.33418938517570496,
+      "learning_rate": 4.965227910177996e-06,
+      "loss": 0.1792,
+      "step": 358
+    },
+    {
+      "epoch": 0.16375869540426502,
+      "grad_norm": 0.33342260122299194,
+      "learning_rate": 4.964920235037276e-06,
+      "loss": 0.1823,
+      "step": 359
+    },
+    {
+      "epoch": 0.16421484775915154,
+      "grad_norm": 0.3288987874984741,
+      "learning_rate": 4.9646112142962295e-06,
+      "loss": 0.1854,
+      "step": 360
+    },
+    {
+      "epoch": 0.1646710001140381,
+      "grad_norm": 0.34447282552719116,
+      "learning_rate": 4.9643008481235505e-06,
+      "loss": 0.1971,
+      "step": 361
+    },
+    {
+      "epoch": 0.16512715246892462,
+      "grad_norm": 0.33075904846191406,
+      "learning_rate": 4.963989136688668e-06,
+      "loss": 0.1826,
+      "step": 362
+    },
+    {
+      "epoch": 0.16558330482381114,
+      "grad_norm": 0.34057560563087463,
+      "learning_rate": 4.963676080161745e-06,
+      "loss": 0.1915,
+      "step": 363
+    },
+    {
+      "epoch": 0.1660394571786977,
+      "grad_norm": 0.33922502398490906,
+      "learning_rate": 4.96336167871368e-06,
+      "loss": 0.1921,
+      "step": 364
+    },
+    {
+      "epoch": 0.16649560953358422,
+      "grad_norm": 0.3598951995372772,
+      "learning_rate": 4.963045932516104e-06,
+      "loss": 0.1974,
+      "step": 365
+    },
+    {
+      "epoch": 0.16695176188847075,
+      "grad_norm": 0.34719255566596985,
+      "learning_rate": 4.962728841741383e-06,
+      "loss": 0.1923,
+      "step": 366
+    },
+    {
+      "epoch": 0.16740791424335727,
+      "grad_norm": 0.3541870713233948,
+      "learning_rate": 4.962410406562617e-06,
+      "loss": 0.1984,
+      "step": 367
+    },
+    {
+      "epoch": 0.16786406659824382,
+      "grad_norm": 0.33247461915016174,
+      "learning_rate": 4.96209062715364e-06,
+      "loss": 0.1851,
+      "step": 368
+    },
+    {
+      "epoch": 0.16832021895313035,
+      "grad_norm": 0.32637760043144226,
+      "learning_rate": 4.961769503689019e-06,
+      "loss": 0.1897,
+      "step": 369
+    },
+    {
+      "epoch": 0.16877637130801687,
+      "grad_norm": 0.34264907240867615,
+      "learning_rate": 4.961447036344056e-06,
+      "loss": 0.187,
+      "step": 370
+    },
+    {
+      "epoch": 0.16923252366290342,
+      "grad_norm": 0.33460769057273865,
+      "learning_rate": 4.961123225294786e-06,
+      "loss": 0.1968,
+      "step": 371
+    },
+    {
+      "epoch": 0.16968867601778995,
+      "grad_norm": 0.33467501401901245,
+      "learning_rate": 4.960798070717977e-06,
+      "loss": 0.2027,
+      "step": 372
+    },
+    {
+      "epoch": 0.17014482837267647,
+      "grad_norm": 0.33618053793907166,
+      "learning_rate": 4.960471572791131e-06,
+      "loss": 0.1848,
+      "step": 373
+    },
+    {
+      "epoch": 0.170600980727563,
+      "grad_norm": 0.3525373637676239,
+      "learning_rate": 4.960143731692484e-06,
+      "loss": 0.196,
+      "step": 374
+    },
+    {
+      "epoch": 0.17105713308244955,
+      "grad_norm": 0.3740014135837555,
+      "learning_rate": 4.959814547601004e-06,
+      "loss": 0.204,
+      "step": 375
+    },
+    {
+      "epoch": 0.17151328543733607,
+      "grad_norm": 0.3468645513057709,
+      "learning_rate": 4.959484020696392e-06,
+      "loss": 0.1899,
+      "step": 376
+    },
+    {
+      "epoch": 0.1719694377922226,
+      "grad_norm": 0.31559595465660095,
+      "learning_rate": 4.9591521511590834e-06,
+      "loss": 0.1837,
+      "step": 377
+    },
+    {
+      "epoch": 0.17242559014710915,
+      "grad_norm": 0.3296249508857727,
+      "learning_rate": 4.958818939170246e-06,
+      "loss": 0.1841,
+      "step": 378
+    },
+    {
+      "epoch": 0.17288174250199567,
+      "grad_norm": 0.3379790186882019,
+      "learning_rate": 4.9584843849117794e-06,
+      "loss": 0.2054,
+      "step": 379
+    },
+    {
+      "epoch": 0.1733378948568822,
+      "grad_norm": 0.33702710270881653,
+      "learning_rate": 4.9581484885663175e-06,
+      "loss": 0.202,
+      "step": 380
+    },
+    {
+      "epoch": 0.17379404721176872,
+      "grad_norm": 0.3467918634414673,
+      "learning_rate": 4.957811250317227e-06,
+      "loss": 0.1915,
+      "step": 381
+    },
+    {
+      "epoch": 0.17425019956665527,
+      "grad_norm": 0.33218732476234436,
+      "learning_rate": 4.957472670348605e-06,
+      "loss": 0.1977,
+      "step": 382
+    },
+    {
+      "epoch": 0.1747063519215418,
+      "grad_norm": 0.3500065505504608,
+      "learning_rate": 4.957132748845282e-06,
+      "loss": 0.1926,
+      "step": 383
+    },
+    {
+      "epoch": 0.17516250427642832,
+      "grad_norm": 0.33966535329818726,
+      "learning_rate": 4.956791485992823e-06,
+      "loss": 0.2017,
+      "step": 384
+    },
+    {
+      "epoch": 0.17561865663131485,
+      "grad_norm": 0.34644681215286255,
+      "learning_rate": 4.956448881977521e-06,
+      "loss": 0.1953,
+      "step": 385
+    },
+    {
+      "epoch": 0.1760748089862014,
+      "grad_norm": 0.31539812684059143,
+      "learning_rate": 4.9561049369864054e-06,
+      "loss": 0.1824,
+      "step": 386
+    },
+    {
+      "epoch": 0.17653096134108792,
+      "grad_norm": 0.349380761384964,
+      "learning_rate": 4.955759651207236e-06,
+      "loss": 0.1995,
+      "step": 387
+    },
+    {
+      "epoch": 0.17698711369597445,
+      "grad_norm": 0.3420228660106659,
+      "learning_rate": 4.955413024828504e-06,
+      "loss": 0.1955,
+      "step": 388
+    },
+    {
+      "epoch": 0.177443266050861,
+      "grad_norm": 0.3577854633331299,
+      "learning_rate": 4.955065058039432e-06,
+      "loss": 0.1998,
+      "step": 389
+    },
+    {
+      "epoch": 0.17789941840574752,
+      "grad_norm": 0.3404674530029297,
+      "learning_rate": 4.954715751029976e-06,
+      "loss": 0.1959,
+      "step": 390
+    },
+    {
+      "epoch": 0.17835557076063405,
+      "grad_norm": 0.32474109530448914,
+      "learning_rate": 4.9543651039908215e-06,
+      "loss": 0.1758,
+      "step": 391
+    },
+    {
+      "epoch": 0.17881172311552057,
+      "grad_norm": 0.3270639479160309,
+      "learning_rate": 4.9540131171133884e-06,
+      "loss": 0.1836,
+      "step": 392
+    },
+    {
+      "epoch": 0.17926787547040712,
+      "grad_norm": 0.31391867995262146,
+      "learning_rate": 4.953659790589825e-06,
+      "loss": 0.1712,
+      "step": 393
+    },
+    {
+      "epoch": 0.17972402782529365,
+      "grad_norm": 0.3314268887042999,
+      "learning_rate": 4.953305124613014e-06,
+      "loss": 0.1868,
+      "step": 394
+    },
+    {
+      "epoch": 0.18018018018018017,
+      "grad_norm": 0.32445767521858215,
+      "learning_rate": 4.952949119376567e-06,
+      "loss": 0.1802,
+      "step": 395
+    },
+    {
+      "epoch": 0.18063633253506672,
+      "grad_norm": 0.33939313888549805,
+      "learning_rate": 4.952591775074825e-06,
+      "loss": 0.1832,
+      "step": 396
+    },
+    {
+      "epoch": 0.18109248488995325,
+      "grad_norm": 0.3279170095920563,
+      "learning_rate": 4.9522330919028665e-06,
+      "loss": 0.1879,
+      "step": 397
+    },
+    {
+      "epoch": 0.18154863724483977,
+      "grad_norm": 0.3530663251876831,
+      "learning_rate": 4.951873070056492e-06,
+      "loss": 0.2027,
+      "step": 398
+    },
+    {
+      "epoch": 0.1820047895997263,
+      "grad_norm": 0.3501916527748108,
+      "learning_rate": 4.951511709732242e-06,
+      "loss": 0.199,
+      "step": 399
+    },
+    {
+      "epoch": 0.18246094195461285,
+      "grad_norm": 0.33651891350746155,
+      "learning_rate": 4.951149011127379e-06,
+      "loss": 0.2036,
+      "step": 400
+    },
+    {
+      "epoch": 0.18291709430949937,
+      "grad_norm": 0.33945992588996887,
+      "learning_rate": 4.950784974439904e-06,
+      "loss": 0.1868,
+      "step": 401
+    },
+    {
+      "epoch": 0.1833732466643859,
+      "grad_norm": 0.33853235840797424,
+      "learning_rate": 4.950419599868542e-06,
+      "loss": 0.1892,
+      "step": 402
+    },
+    {
+      "epoch": 0.18382939901927245,
+      "grad_norm": 0.337922066450119,
+      "learning_rate": 4.950052887612753e-06,
+      "loss": 0.1745,
+      "step": 403
+    },
+    {
+      "epoch": 0.18428555137415897,
+      "grad_norm": 0.3489057719707489,
+      "learning_rate": 4.949684837872723e-06,
+      "loss": 0.1969,
+      "step": 404
+    },
+    {
+      "epoch": 0.1847417037290455,
+      "grad_norm": 0.34555283188819885,
+      "learning_rate": 4.949315450849373e-06,
+      "loss": 0.2052,
+      "step": 405
+    },
+    {
+      "epoch": 0.18519785608393202,
+      "grad_norm": 0.34546807408332825,
+      "learning_rate": 4.948944726744349e-06,
+      "loss": 0.2008,
+      "step": 406
+    },
+    {
+      "epoch": 0.18565400843881857,
+      "grad_norm": 0.3540160059928894,
+      "learning_rate": 4.948572665760031e-06,
+      "loss": 0.1891,
+      "step": 407
+    },
+    {
+      "epoch": 0.1861101607937051,
+      "grad_norm": 0.32495954632759094,
+      "learning_rate": 4.948199268099525e-06,
+      "loss": 0.1917,
+      "step": 408
+    },
+    {
+      "epoch": 0.18656631314859162,
+      "grad_norm": 0.3689875304698944,
+      "learning_rate": 4.947824533966672e-06,
+      "loss": 0.202,
+      "step": 409
+    },
+    {
+      "epoch": 0.18702246550347817,
+      "grad_norm": 0.33589452505111694,
+      "learning_rate": 4.947448463566038e-06,
+      "loss": 0.1847,
+      "step": 410
+    },
+    {
+      "epoch": 0.1874786178583647,
+      "grad_norm": 0.3329775929450989,
+      "learning_rate": 4.947071057102919e-06,
+      "loss": 0.1905,
+      "step": 411
+    },
+    {
+      "epoch": 0.18793477021325122,
+      "grad_norm": 0.3307413160800934,
+      "learning_rate": 4.946692314783342e-06,
+      "loss": 0.1858,
+      "step": 412
+    },
+    {
+      "epoch": 0.18839092256813775,
+      "grad_norm": 0.3514516055583954,
+      "learning_rate": 4.946312236814063e-06,
+      "loss": 0.1992,
+      "step": 413
+    },
+    {
+      "epoch": 0.1888470749230243,
+      "grad_norm": 0.3269518315792084,
+      "learning_rate": 4.945930823402566e-06,
+      "loss": 0.1891,
+      "step": 414
+    },
+    {
+      "epoch": 0.18930322727791082,
+      "grad_norm": 0.33550408482551575,
+      "learning_rate": 4.9455480747570645e-06,
+      "loss": 0.1946,
+      "step": 415
+    },
+    {
+      "epoch": 0.18975937963279735,
+      "grad_norm": 0.33317622542381287,
+      "learning_rate": 4.9451639910865016e-06,
+      "loss": 0.1875,
+      "step": 416
+    },
+    {
+      "epoch": 0.1902155319876839,
+      "grad_norm": 0.3332635462284088,
+      "learning_rate": 4.9447785726005475e-06,
+      "loss": 0.1872,
+      "step": 417
+    },
+    {
+      "epoch": 0.19067168434257042,
+      "grad_norm": 0.3485168218612671,
+      "learning_rate": 4.944391819509604e-06,
+      "loss": 0.1989,
+      "step": 418
+    },
+    {
+      "epoch": 0.19112783669745695,
+      "grad_norm": 0.34602800011634827,
+      "learning_rate": 4.944003732024798e-06,
+      "loss": 0.1857,
+      "step": 419
+    },
+    {
+      "epoch": 0.19158398905234347,
+      "grad_norm": 0.3513849377632141,
+      "learning_rate": 4.943614310357987e-06,
+      "loss": 0.1857,
+      "step": 420
+    },
+    {
+      "epoch": 0.19204014140723003,
+      "grad_norm": 0.37508660554885864,
+      "learning_rate": 4.943223554721758e-06,
+      "loss": 0.1871,
+      "step": 421
+    },
+    {
+      "epoch": 0.19249629376211655,
+      "grad_norm": 2.0501863956451416,
+      "learning_rate": 4.942831465329421e-06,
+      "loss": 0.1896,
+      "step": 422
+    },
+    {
+      "epoch": 0.19295244611700307,
+      "grad_norm": 0.3143482208251953,
+      "learning_rate": 4.9424380423950215e-06,
+      "loss": 0.174,
+      "step": 423
+    },
+    {
+      "epoch": 0.1934085984718896,
+      "grad_norm": 0.33220532536506653,
+      "learning_rate": 4.942043286133326e-06,
+      "loss": 0.1983,
+      "step": 424
+    },
+    {
+      "epoch": 0.19386475082677615,
+      "grad_norm": 0.33381083607673645,
+      "learning_rate": 4.941647196759834e-06,
+      "loss": 0.1927,
+      "step": 425
+    },
+    {
+      "epoch": 0.19432090318166267,
+      "grad_norm": 0.34169742465019226,
+      "learning_rate": 4.941249774490771e-06,
+      "loss": 0.1975,
+      "step": 426
+    },
+    {
+      "epoch": 0.1947770555365492,
+      "grad_norm": 0.3176627457141876,
+      "learning_rate": 4.940851019543088e-06,
+      "loss": 0.1888,
+      "step": 427
+    },
+    {
+      "epoch": 0.19523320789143575,
+      "grad_norm": 0.3162442147731781,
+      "learning_rate": 4.940450932134467e-06,
+      "loss": 0.1939,
+      "step": 428
+    },
+    {
+      "epoch": 0.19568936024632227,
+      "grad_norm": 0.33328139781951904,
+      "learning_rate": 4.940049512483315e-06,
+      "loss": 0.192,
+      "step": 429
+    },
+    {
+      "epoch": 0.1961455126012088,
+      "grad_norm": 0.33137211203575134,
+      "learning_rate": 4.939646760808767e-06,
+      "loss": 0.1829,
+      "step": 430
+    },
+    {
+      "epoch": 0.19660166495609532,
+      "grad_norm": 0.33801645040512085,
+      "learning_rate": 4.939242677330686e-06,
+      "loss": 0.1969,
+      "step": 431
+    },
+    {
+      "epoch": 0.19705781731098188,
+      "grad_norm": 0.3268057405948639,
+      "learning_rate": 4.9388372622696605e-06,
+      "loss": 0.1824,
+      "step": 432
+    },
+    {
+      "epoch": 0.1975139696658684,
+      "grad_norm": 0.33720341324806213,
+      "learning_rate": 4.938430515847006e-06,
+      "loss": 0.1916,
+      "step": 433
+    },
+    {
+      "epoch": 0.19797012202075492,
+      "grad_norm": 0.3099447786808014,
+      "learning_rate": 4.938022438284766e-06,
+      "loss": 0.169,
+      "step": 434
+    },
+    {
+      "epoch": 0.19842627437564148,
+      "grad_norm": 0.3325330317020416,
+      "learning_rate": 4.937613029805713e-06,
+      "loss": 0.1839,
+      "step": 435
+    },
+    {
+      "epoch": 0.198882426730528,
+      "grad_norm": 0.34183311462402344,
+      "learning_rate": 4.937202290633337e-06,
+      "loss": 0.2086,
+      "step": 436
+    },
+    {
+      "epoch": 0.19933857908541452,
+      "grad_norm": 0.3339877426624298,
+      "learning_rate": 4.936790220991866e-06,
+      "loss": 0.1934,
+      "step": 437
+    },
+    {
+      "epoch": 0.19979473144030105,
+      "grad_norm": 0.32080769538879395,
+      "learning_rate": 4.936376821106247e-06,
+      "loss": 0.1849,
+      "step": 438
+    },
+    {
+      "epoch": 0.2002508837951876,
+      "grad_norm": 0.322485089302063,
+      "learning_rate": 4.935962091202154e-06,
+      "loss": 0.1869,
+      "step": 439
+    },
+    {
+      "epoch": 0.20070703615007413,
+      "grad_norm": 0.34053468704223633,
+      "learning_rate": 4.935546031505991e-06,
+      "loss": 0.2019,
+      "step": 440
+    },
+    {
+      "epoch": 0.20116318850496065,
+      "grad_norm": 0.33340534567832947,
+      "learning_rate": 4.935128642244882e-06,
+      "loss": 0.1862,
+      "step": 441
+    },
+    {
+      "epoch": 0.2016193408598472,
+      "grad_norm": 0.3638502359390259,
+      "learning_rate": 4.934709923646681e-06,
+      "loss": 0.2015,
+      "step": 442
+    },
+    {
+      "epoch": 0.20207549321473373,
+      "grad_norm": 0.3303164541721344,
+      "learning_rate": 4.934289875939966e-06,
+      "loss": 0.1817,
+      "step": 443
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.35156112909317017,
+      "learning_rate": 4.933868499354043e-06,
+      "loss": 0.1976,
+      "step": 444
+    },
+    {
+      "epoch": 0.20298779792450677,
+      "grad_norm": 0.33329400420188904,
+      "learning_rate": 4.9334457941189395e-06,
+      "loss": 0.1784,
+      "step": 445
+    },
+    {
+      "epoch": 0.20344395027939333,
+      "grad_norm": 0.3373907208442688,
+      "learning_rate": 4.933021760465411e-06,
+      "loss": 0.1895,
+      "step": 446
+    },
+    {
+      "epoch": 0.20390010263427985,
+      "grad_norm": 0.3146997392177582,
+      "learning_rate": 4.932596398624938e-06,
+      "loss": 0.1773,
+      "step": 447
+    },
+    {
+      "epoch": 0.20435625498916637,
+      "grad_norm": 0.322661429643631,
+      "learning_rate": 4.932169708829725e-06,
+      "loss": 0.1805,
+      "step": 448
+    },
+    {
+      "epoch": 0.20481240734405293,
+      "grad_norm": 0.3433276116847992,
+      "learning_rate": 4.931741691312703e-06,
+      "loss": 0.1829,
+      "step": 449
+    },
+    {
+      "epoch": 0.20526855969893945,
+      "grad_norm": 0.3343183696269989,
+      "learning_rate": 4.931312346307525e-06,
+      "loss": 0.1854,
+      "step": 450
+    },
+    {
+      "epoch": 0.20572471205382598,
+      "grad_norm": 0.34893640875816345,
+      "learning_rate": 4.930881674048571e-06,
+      "loss": 0.1893,
+      "step": 451
+    },
+    {
+      "epoch": 0.2061808644087125,
+      "grad_norm": 0.3287903368473053,
+      "learning_rate": 4.930449674770947e-06,
+      "loss": 0.1871,
+      "step": 452
+    },
+    {
+      "epoch": 0.20663701676359905,
+      "grad_norm": 0.34509751200675964,
+      "learning_rate": 4.93001634871048e-06,
+      "loss": 0.1848,
+      "step": 453
+    },
+    {
+      "epoch": 0.20709316911848558,
+      "grad_norm": 0.35198959708213806,
+      "learning_rate": 4.9295816961037224e-06,
+      "loss": 0.1897,
+      "step": 454
+    },
+    {
+      "epoch": 0.2075493214733721,
+      "grad_norm": 0.354473352432251,
+      "learning_rate": 4.929145717187951e-06,
+      "loss": 0.1996,
+      "step": 455
+    },
+    {
+      "epoch": 0.20800547382825862,
+      "grad_norm": 0.3427160382270813,
+      "learning_rate": 4.928708412201169e-06,
+      "loss": 0.1773,
+      "step": 456
+    },
+    {
+      "epoch": 0.20846162618314518,
+      "grad_norm": 0.36603277921676636,
+      "learning_rate": 4.928269781382098e-06,
+      "loss": 0.1916,
+      "step": 457
+    },
+    {
+      "epoch": 0.2089177785380317,
+      "grad_norm": 0.3512258231639862,
+      "learning_rate": 4.927829824970188e-06,
+      "loss": 0.1871,
+      "step": 458
+    },
+    {
+      "epoch": 0.20937393089291823,
+      "grad_norm": 0.34780237078666687,
+      "learning_rate": 4.927388543205613e-06,
+      "loss": 0.1991,
+      "step": 459
+    },
+    {
+      "epoch": 0.20983008324780478,
+      "grad_norm": 0.41480913758277893,
+      "learning_rate": 4.926945936329266e-06,
+      "loss": 0.192,
+      "step": 460
+    },
+    {
+      "epoch": 0.2102862356026913,
+      "grad_norm": 0.36109524965286255,
+      "learning_rate": 4.926502004582768e-06,
+      "loss": 0.1794,
+      "step": 461
+    },
+    {
+      "epoch": 0.21074238795757783,
+      "grad_norm": 0.3389061689376831,
+      "learning_rate": 4.926056748208461e-06,
+      "loss": 0.2007,
+      "step": 462
+    },
+    {
+      "epoch": 0.21119854031246435,
+      "grad_norm": 0.3711278438568115,
+      "learning_rate": 4.92561016744941e-06,
+      "loss": 0.1963,
+      "step": 463
+    },
+    {
+      "epoch": 0.2116546926673509,
+      "grad_norm": 0.3651914894580841,
+      "learning_rate": 4.925162262549405e-06,
+      "loss": 0.1928,
+      "step": 464
+    },
+    {
+      "epoch": 0.21211084502223743,
+      "grad_norm": 0.3350219428539276,
+      "learning_rate": 4.9247130337529545e-06,
+      "loss": 0.1737,
+      "step": 465
+    },
+    {
+      "epoch": 0.21256699737712395,
+      "grad_norm": 0.3586723208427429,
+      "learning_rate": 4.924262481305295e-06,
+      "loss": 0.1893,
+      "step": 466
+    },
+    {
+      "epoch": 0.2130231497320105,
+      "grad_norm": 0.33117642998695374,
+      "learning_rate": 4.923810605452383e-06,
+      "loss": 0.1915,
+      "step": 467
+    },
+    {
+      "epoch": 0.21347930208689703,
+      "grad_norm": 0.35808396339416504,
+      "learning_rate": 4.923357406440896e-06,
+      "loss": 0.1904,
+      "step": 468
+    },
+    {
+      "epoch": 0.21393545444178355,
+      "grad_norm": 0.3280753195285797,
+      "learning_rate": 4.9229028845182366e-06,
+      "loss": 0.1841,
+      "step": 469
+    },
+    {
+      "epoch": 0.21439160679667008,
+      "grad_norm": 0.3263069987297058,
+      "learning_rate": 4.92244703993253e-06,
+      "loss": 0.18,
+      "step": 470
+    },
+    {
+      "epoch": 0.21484775915155663,
+      "grad_norm": 0.3604918122291565,
+      "learning_rate": 4.921989872932617e-06,
+      "loss": 0.1835,
+      "step": 471
+    },
+    {
+      "epoch": 0.21530391150644315,
+      "grad_norm": 0.33646416664123535,
+      "learning_rate": 4.921531383768071e-06,
+      "loss": 0.1926,
+      "step": 472
+    },
+    {
+      "epoch": 0.21576006386132968,
+      "grad_norm": 0.314749538898468,
+      "learning_rate": 4.9210715726891765e-06,
+      "loss": 0.1713,
+      "step": 473
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 0.37420058250427246,
+      "learning_rate": 4.9206104399469475e-06,
+      "loss": 0.1954,
+      "step": 474
+    },
+    {
+      "epoch": 0.21667236857110275,
+      "grad_norm": 0.34236428141593933,
+      "learning_rate": 4.9201479857931155e-06,
+      "loss": 0.1935,
+      "step": 475
+    },
+    {
+      "epoch": 0.21712852092598928,
+      "grad_norm": 0.3258551359176636,
+      "learning_rate": 4.919684210480134e-06,
+      "loss": 0.1964,
+      "step": 476
+    },
+    {
+      "epoch": 0.2175846732808758,
+      "grad_norm": 0.3280400335788727,
+      "learning_rate": 4.9192191142611775e-06,
+      "loss": 0.1867,
+      "step": 477
+    },
+    {
+      "epoch": 0.21804082563576235,
+      "grad_norm": 0.32536911964416504,
+      "learning_rate": 4.9187526973901434e-06,
+      "loss": 0.1845,
+      "step": 478
+    },
+    {
+      "epoch": 0.21849697799064888,
+      "grad_norm": 0.3250200152397156,
+      "learning_rate": 4.918284960121648e-06,
+      "loss": 0.1814,
+      "step": 479
+    },
+    {
+      "epoch": 0.2189531303455354,
+      "grad_norm": 0.31664541363716125,
+      "learning_rate": 4.917815902711029e-06,
+      "loss": 0.1789,
+      "step": 480
+    },
+    {
+      "epoch": 0.21940928270042195,
+      "grad_norm": 0.352123498916626,
+      "learning_rate": 4.917345525414345e-06,
+      "loss": 0.1977,
+      "step": 481
+    },
+    {
+      "epoch": 0.21986543505530848,
+      "grad_norm": 0.33699682354927063,
+      "learning_rate": 4.916873828488375e-06,
+      "loss": 0.1881,
+      "step": 482
+    },
+    {
+      "epoch": 0.220321587410195,
+      "grad_norm": 0.319257527589798,
+      "learning_rate": 4.916400812190619e-06,
+      "loss": 0.1824,
+      "step": 483
+    },
+    {
+      "epoch": 0.22077773976508153,
+      "grad_norm": 0.3320259749889374,
+      "learning_rate": 4.915926476779297e-06,
+      "loss": 0.1904,
+      "step": 484
+    },
+    {
+      "epoch": 0.22123389211996808,
+      "grad_norm": 0.35232606530189514,
+      "learning_rate": 4.915450822513348e-06,
+      "loss": 0.1797,
+      "step": 485
+    },
+    {
+      "epoch": 0.2216900444748546,
+      "grad_norm": 0.33443185687065125,
+      "learning_rate": 4.91497384965243e-06,
+      "loss": 0.1812,
+      "step": 486
+    },
+    {
+      "epoch": 0.22214619682974113,
+      "grad_norm": 0.3441826105117798,
+      "learning_rate": 4.914495558456926e-06,
+      "loss": 0.1931,
+      "step": 487
+    },
+    {
+      "epoch": 0.22260234918462768,
+      "grad_norm": 0.32324960827827454,
+      "learning_rate": 4.914015949187934e-06,
+      "loss": 0.1766,
+      "step": 488
+    },
+    {
+      "epoch": 0.2230585015395142,
+      "grad_norm": 0.34965988993644714,
+      "learning_rate": 4.913535022107271e-06,
+      "loss": 0.2023,
+      "step": 489
+    },
+    {
+      "epoch": 0.22351465389440073,
+      "grad_norm": 0.3410987854003906,
+      "learning_rate": 4.913052777477477e-06,
+      "loss": 0.1942,
+      "step": 490
+    },
+    {
+      "epoch": 0.22397080624928725,
+      "grad_norm": 0.34191426634788513,
+      "learning_rate": 4.912569215561809e-06,
+      "loss": 0.1833,
+      "step": 491
+    },
+    {
+      "epoch": 0.2244269586041738,
+      "grad_norm": 0.3284359872341156,
+      "learning_rate": 4.912084336624243e-06,
+      "loss": 0.2023,
+      "step": 492
+    },
+    {
+      "epoch": 0.22488311095906033,
+      "grad_norm": 0.3127383291721344,
+      "learning_rate": 4.911598140929476e-06,
+      "loss": 0.1845,
+      "step": 493
+    },
+    {
+      "epoch": 0.22533926331394685,
+      "grad_norm": 0.33405444025993347,
+      "learning_rate": 4.91111062874292e-06,
+      "loss": 0.1742,
+      "step": 494
+    },
+    {
+      "epoch": 0.22579541566883338,
+      "grad_norm": 0.3368760645389557,
+      "learning_rate": 4.910621800330711e-06,
+      "loss": 0.1949,
+      "step": 495
+    },
+    {
+      "epoch": 0.22625156802371993,
+      "grad_norm": 0.33372989296913147,
+      "learning_rate": 4.910131655959697e-06,
+      "loss": 0.1843,
+      "step": 496
+    },
+    {
+      "epoch": 0.22670772037860645,
+      "grad_norm": 0.35408976674079895,
+      "learning_rate": 4.909640195897451e-06,
+      "loss": 0.1995,
+      "step": 497
+    },
+    {
+      "epoch": 0.22716387273349298,
+      "grad_norm": 0.2991258502006531,
+      "learning_rate": 4.909147420412258e-06,
+      "loss": 0.1714,
+      "step": 498
+    },
+    {
+      "epoch": 0.22762002508837953,
+      "grad_norm": 0.34220778942108154,
+      "learning_rate": 4.908653329773127e-06,
+      "loss": 0.1981,
+      "step": 499
+    },
+    {
+      "epoch": 0.22807617744326605,
+      "grad_norm": 0.3388160765171051,
+      "learning_rate": 4.908157924249781e-06,
+      "loss": 0.1947,
+      "step": 500
+    },
+    {
+      "epoch": 0.22853232979815258,
+      "grad_norm": 0.3370591104030609,
+      "learning_rate": 4.907661204112662e-06,
+      "loss": 0.1881,
+      "step": 501
+    },
+    {
+      "epoch": 0.2289884821530391,
+      "grad_norm": 0.3480423390865326,
+      "learning_rate": 4.90716316963293e-06,
+      "loss": 0.184,
+      "step": 502
+    },
+    {
+      "epoch": 0.22944463450792565,
+      "grad_norm": 0.3363558053970337,
+      "learning_rate": 4.906663821082461e-06,
+      "loss": 0.1884,
+      "step": 503
+    },
+    {
+      "epoch": 0.22990078686281218,
+      "grad_norm": 0.33011311292648315,
+      "learning_rate": 4.906163158733851e-06,
+      "loss": 0.1873,
+      "step": 504
+    },
+    {
+      "epoch": 0.2303569392176987,
+      "grad_norm": 0.35187768936157227,
+      "learning_rate": 4.9056611828604105e-06,
+      "loss": 0.1988,
+      "step": 505
+    },
+    {
+      "epoch": 0.23081309157258525,
+      "grad_norm": 0.3446900546550751,
+      "learning_rate": 4.905157893736169e-06,
+      "loss": 0.2026,
+      "step": 506
+    },
+    {
+      "epoch": 0.23126924392747178,
+      "grad_norm": 0.34599074721336365,
+      "learning_rate": 4.90465329163587e-06,
+      "loss": 0.1974,
+      "step": 507
+    },
+    {
+      "epoch": 0.2317253962823583,
+      "grad_norm": 0.32214030623435974,
+      "learning_rate": 4.904147376834979e-06,
+      "loss": 0.1795,
+      "step": 508
+    },
+    {
+      "epoch": 0.23218154863724483,
+      "grad_norm": 0.30877596139907837,
+      "learning_rate": 4.903640149609674e-06,
+      "loss": 0.1854,
+      "step": 509
+    },
+    {
+      "epoch": 0.23263770099213138,
+      "grad_norm": 0.3001999258995056,
+      "learning_rate": 4.903131610236849e-06,
+      "loss": 0.1861,
+      "step": 510
+    },
+    {
+      "epoch": 0.2330938533470179,
+      "grad_norm": 0.3578189015388489,
+      "learning_rate": 4.902621758994116e-06,
+      "loss": 0.2034,
+      "step": 511
+    },
+    {
+      "epoch": 0.23355000570190443,
+      "grad_norm": 0.32655709981918335,
+      "learning_rate": 4.9021105961598046e-06,
+      "loss": 0.1787,
+      "step": 512
+    },
+    {
+      "epoch": 0.23400615805679098,
+      "grad_norm": 0.33656859397888184,
+      "learning_rate": 4.9015981220129575e-06,
+      "loss": 0.188,
+      "step": 513
+    },
+    {
+      "epoch": 0.2344623104116775,
+      "grad_norm": 0.3175767660140991,
+      "learning_rate": 4.901084336833334e-06,
+      "loss": 0.1855,
+      "step": 514
+    },
+    {
+      "epoch": 0.23491846276656403,
+      "grad_norm": 0.3438035249710083,
+      "learning_rate": 4.90056924090141e-06,
+      "loss": 0.2051,
+      "step": 515
+    },
+    {
+      "epoch": 0.23537461512145055,
+      "grad_norm": 0.3083075284957886,
+      "learning_rate": 4.900052834498377e-06,
+      "loss": 0.1859,
+      "step": 516
+    },
+    {
+      "epoch": 0.2358307674763371,
+      "grad_norm": 0.31989985704421997,
+      "learning_rate": 4.89953511790614e-06,
+      "loss": 0.1836,
+      "step": 517
+    },
+    {
+      "epoch": 0.23628691983122363,
+      "grad_norm": 0.338565468788147,
+      "learning_rate": 4.899016091407321e-06,
+      "loss": 0.1877,
+      "step": 518
+    },
+    {
+      "epoch": 0.23674307218611015,
+      "grad_norm": 0.3327895402908325,
+      "learning_rate": 4.898495755285258e-06,
+      "loss": 0.1885,
+      "step": 519
+    },
+    {
+      "epoch": 0.2371992245409967,
+      "grad_norm": 0.3383982181549072,
+      "learning_rate": 4.897974109824002e-06,
+      "loss": 0.1966,
+      "step": 520
+    },
+    {
+      "epoch": 0.23765537689588323,
+      "grad_norm": 0.3337146043777466,
+      "learning_rate": 4.897451155308318e-06,
+      "loss": 0.1895,
+      "step": 521
+    },
+    {
+      "epoch": 0.23811152925076975,
+      "grad_norm": 0.33300086855888367,
+      "learning_rate": 4.8969268920236875e-06,
+      "loss": 0.1903,
+      "step": 522
+    },
+    {
+      "epoch": 0.23856768160565628,
+      "grad_norm": 0.3378000557422638,
+      "learning_rate": 4.896401320256307e-06,
+      "loss": 0.1979,
+      "step": 523
+    },
+    {
+      "epoch": 0.23902383396054283,
+      "grad_norm": 0.3299938142299652,
+      "learning_rate": 4.895874440293085e-06,
+      "loss": 0.1863,
+      "step": 524
+    },
+    {
+      "epoch": 0.23947998631542936,
+      "grad_norm": 0.3388802707195282,
+      "learning_rate": 4.895346252421646e-06,
+      "loss": 0.194,
+      "step": 525
+    },
+    {
+      "epoch": 0.23993613867031588,
+      "grad_norm": 0.32321327924728394,
+      "learning_rate": 4.894816756930327e-06,
+      "loss": 0.1847,
+      "step": 526
+    },
+    {
+      "epoch": 0.24039229102520243,
+      "grad_norm": 0.3241207003593445,
+      "learning_rate": 4.894285954108181e-06,
+      "loss": 0.1991,
+      "step": 527
+    },
+    {
+      "epoch": 0.24084844338008896,
+      "grad_norm": 0.31251421570777893,
+      "learning_rate": 4.8937538442449724e-06,
+      "loss": 0.1818,
+      "step": 528
+    },
+    {
+      "epoch": 0.24130459573497548,
+      "grad_norm": 0.35212618112564087,
+      "learning_rate": 4.893220427631181e-06,
+      "loss": 0.2053,
+      "step": 529
+    },
+    {
+      "epoch": 0.241760748089862,
+      "grad_norm": 0.3470520079135895,
+      "learning_rate": 4.892685704557998e-06,
+      "loss": 0.1778,
+      "step": 530
+    },
+    {
+      "epoch": 0.24221690044474856,
+      "grad_norm": 0.34073078632354736,
+      "learning_rate": 4.892149675317328e-06,
+      "loss": 0.1789,
+      "step": 531
+    },
+    {
+      "epoch": 0.24267305279963508,
+      "grad_norm": 0.327871173620224,
+      "learning_rate": 4.891612340201791e-06,
+      "loss": 0.1903,
+      "step": 532
+    },
+    {
+      "epoch": 0.2431292051545216,
+      "grad_norm": 0.36025846004486084,
+      "learning_rate": 4.891073699504719e-06,
+      "loss": 0.2023,
+      "step": 533
+    },
+    {
+      "epoch": 0.24358535750940813,
+      "grad_norm": 0.34813451766967773,
+      "learning_rate": 4.8905337535201536e-06,
+      "loss": 0.2092,
+      "step": 534
+    },
+    {
+      "epoch": 0.24404150986429468,
+      "grad_norm": 0.3350469768047333,
+      "learning_rate": 4.889992502542853e-06,
+      "loss": 0.1767,
+      "step": 535
+    },
+    {
+      "epoch": 0.2444976622191812,
+      "grad_norm": 0.34527790546417236,
+      "learning_rate": 4.8894499468682865e-06,
+      "loss": 0.1929,
+      "step": 536
+    },
+    {
+      "epoch": 0.24495381457406773,
+      "grad_norm": 0.32161593437194824,
+      "learning_rate": 4.888906086792634e-06,
+      "loss": 0.1802,
+      "step": 537
+    },
+    {
+      "epoch": 0.24540996692895428,
+      "grad_norm": 0.3323829472064972,
+      "learning_rate": 4.88836092261279e-06,
+      "loss": 0.1858,
+      "step": 538
+    },
+    {
+      "epoch": 0.2458661192838408,
+      "grad_norm": 0.32991620898246765,
+      "learning_rate": 4.88781445462636e-06,
+      "loss": 0.1906,
+      "step": 539
+    },
+    {
+      "epoch": 0.24632227163872733,
+      "grad_norm": 0.3399929702281952,
+      "learning_rate": 4.887266683131659e-06,
+      "loss": 0.2009,
+      "step": 540
+    },
+    {
+      "epoch": 0.24677842399361385,
+      "grad_norm": 0.31057384610176086,
+      "learning_rate": 4.8867176084277166e-06,
+      "loss": 0.1801,
+      "step": 541
+    },
+    {
+      "epoch": 0.2472345763485004,
+      "grad_norm": 0.355825811624527,
+      "learning_rate": 4.886167230814273e-06,
+      "loss": 0.1841,
+      "step": 542
+    },
+    {
+      "epoch": 0.24769072870338693,
+      "grad_norm": 0.34464454650878906,
+      "learning_rate": 4.88561555059178e-06,
+      "loss": 0.1863,
+      "step": 543
+    },
+    {
+      "epoch": 0.24814688105827346,
+      "grad_norm": 0.31960487365722656,
+      "learning_rate": 4.885062568061399e-06,
+      "loss": 0.18,
+      "step": 544
+    },
+    {
+      "epoch": 0.24860303341316,
+      "grad_norm": 0.3348695635795593,
+      "learning_rate": 4.884508283525002e-06,
+      "loss": 0.1896,
+      "step": 545
+    },
+    {
+      "epoch": 0.24905918576804653,
+      "grad_norm": 0.34271687269210815,
+      "learning_rate": 4.883952697285176e-06,
+      "loss": 0.1973,
+      "step": 546
+    },
+    {
+      "epoch": 0.24951533812293306,
+      "grad_norm": 0.32807254791259766,
+      "learning_rate": 4.883395809645214e-06,
+      "loss": 0.1782,
+      "step": 547
+    },
+    {
+      "epoch": 0.24997149047781958,
+      "grad_norm": 0.3190830945968628,
+      "learning_rate": 4.882837620909121e-06,
+      "loss": 0.1843,
+      "step": 548
+    },
+    {
+      "epoch": 0.25042764283270613,
+      "grad_norm": 0.3390159010887146,
+      "learning_rate": 4.882278131381612e-06,
+      "loss": 0.1967,
+      "step": 549
+    },
+    {
+      "epoch": 0.25088379518759263,
+      "grad_norm": 0.32320883870124817,
+      "learning_rate": 4.881717341368113e-06,
+      "loss": 0.1821,
+      "step": 550
+    },
+    {
+      "epoch": 0.2513399475424792,
+      "grad_norm": 0.32232770323753357,
+      "learning_rate": 4.88115525117476e-06,
+      "loss": 0.1823,
+      "step": 551
+    },
+    {
+      "epoch": 0.25179609989736573,
+      "grad_norm": 0.3226570785045624,
+      "learning_rate": 4.880591861108397e-06,
+      "loss": 0.1894,
+      "step": 552
+    },
+    {
+      "epoch": 0.25225225225225223,
+      "grad_norm": 0.3332536220550537,
+      "learning_rate": 4.880027171476579e-06,
+      "loss": 0.1916,
+      "step": 553
+    },
+    {
+      "epoch": 0.2527084046071388,
+      "grad_norm": 0.3378143310546875,
+      "learning_rate": 4.8794611825875694e-06,
+      "loss": 0.1943,
+      "step": 554
+    },
+    {
+      "epoch": 0.25316455696202533,
+      "grad_norm": 0.3257540762424469,
+      "learning_rate": 4.878893894750344e-06,
+      "loss": 0.1956,
+      "step": 555
+    },
+    {
+      "epoch": 0.25362070931691183,
+      "grad_norm": 0.34001460671424866,
+      "learning_rate": 4.878325308274583e-06,
+      "loss": 0.1889,
+      "step": 556
+    },
+    {
+      "epoch": 0.2540768616717984,
+      "grad_norm": 0.3512193262577057,
+      "learning_rate": 4.877755423470678e-06,
+      "loss": 0.1992,
+      "step": 557
+    },
+    {
+      "epoch": 0.25453301402668493,
+      "grad_norm": 0.34466034173965454,
+      "learning_rate": 4.87718424064973e-06,
+      "loss": 0.1964,
+      "step": 558
+    },
+    {
+      "epoch": 0.25498916638157143,
+      "grad_norm": 0.32443082332611084,
+      "learning_rate": 4.8766117601235486e-06,
+      "loss": 0.1795,
+      "step": 559
+    },
+    {
+      "epoch": 0.255445318736458,
+      "grad_norm": 0.3286967873573303,
+      "learning_rate": 4.876037982204649e-06,
+      "loss": 0.1888,
+      "step": 560
+    },
+    {
+      "epoch": 0.25590147109134453,
+      "grad_norm": 0.33227288722991943,
+      "learning_rate": 4.875462907206258e-06,
+      "loss": 0.1887,
+      "step": 561
+    },
+    {
+      "epoch": 0.25635762344623103,
+      "grad_norm": 0.33176088333129883,
+      "learning_rate": 4.874886535442308e-06,
+      "loss": 0.1761,
+      "step": 562
+    },
+    {
+      "epoch": 0.2568137758011176,
+      "grad_norm": 0.3200024366378784,
+      "learning_rate": 4.874308867227442e-06,
+      "loss": 0.1803,
+      "step": 563
+    },
+    {
+      "epoch": 0.2572699281560041,
+      "grad_norm": 0.33369627594947815,
+      "learning_rate": 4.873729902877009e-06,
+      "loss": 0.1798,
+      "step": 564
+    },
+    {
+      "epoch": 0.25772608051089063,
+      "grad_norm": 0.32088449597358704,
+      "learning_rate": 4.873149642707064e-06,
+      "loss": 0.2007,
+      "step": 565
+    },
+    {
+      "epoch": 0.2581822328657772,
+      "grad_norm": 0.31753721833229065,
+      "learning_rate": 4.872568087034373e-06,
+      "loss": 0.1872,
+      "step": 566
+    },
+    {
+      "epoch": 0.2586383852206637,
+      "grad_norm": 0.33731013536453247,
+      "learning_rate": 4.871985236176406e-06,
+      "loss": 0.1885,
+      "step": 567
+    },
+    {
+      "epoch": 0.25909453757555023,
+      "grad_norm": 0.3191109001636505,
+      "learning_rate": 4.871401090451342e-06,
+      "loss": 0.1784,
+      "step": 568
+    },
+    {
+      "epoch": 0.2595506899304368,
+      "grad_norm": 0.32786479592323303,
+      "learning_rate": 4.870815650178066e-06,
+      "loss": 0.1852,
+      "step": 569
+    },
+    {
+      "epoch": 0.2600068422853233,
+      "grad_norm": 0.34695082902908325,
+      "learning_rate": 4.870228915676171e-06,
+      "loss": 0.1811,
+      "step": 570
+    },
+    {
+      "epoch": 0.26046299464020983,
+      "grad_norm": 0.33102044463157654,
+      "learning_rate": 4.869640887265953e-06,
+      "loss": 0.1892,
+      "step": 571
+    },
+    {
+      "epoch": 0.2609191469950964,
+      "grad_norm": 0.32680588960647583,
+      "learning_rate": 4.869051565268419e-06,
+      "loss": 0.1861,
+      "step": 572
+    },
+    {
+      "epoch": 0.2613752993499829,
+      "grad_norm": 0.3280143141746521,
+      "learning_rate": 4.86846095000528e-06,
+      "loss": 0.1967,
+      "step": 573
+    },
+    {
+      "epoch": 0.26183145170486943,
+      "grad_norm": 0.338615745306015,
+      "learning_rate": 4.867869041798951e-06,
+      "loss": 0.1809,
+      "step": 574
+    },
+    {
+      "epoch": 0.262287604059756,
+      "grad_norm": 0.3240947723388672,
+      "learning_rate": 4.867275840972558e-06,
+      "loss": 0.1917,
+      "step": 575
+    },
+    {
+      "epoch": 0.2627437564146425,
+      "grad_norm": 0.3325682580471039,
+      "learning_rate": 4.866681347849925e-06,
+      "loss": 0.1968,
+      "step": 576
+    },
+    {
+      "epoch": 0.26319990876952903,
+      "grad_norm": 0.31293338537216187,
+      "learning_rate": 4.86608556275559e-06,
+      "loss": 0.1731,
+      "step": 577
+    },
+    {
+      "epoch": 0.26365606112441553,
+      "grad_norm": 0.32401880621910095,
+      "learning_rate": 4.86548848601479e-06,
+      "loss": 0.1778,
+      "step": 578
+    },
+    {
+      "epoch": 0.2641122134793021,
+      "grad_norm": 0.3436855971813202,
+      "learning_rate": 4.86489011795347e-06,
+      "loss": 0.187,
+      "step": 579
+    },
+    {
+      "epoch": 0.26456836583418863,
+      "grad_norm": 0.3250855803489685,
+      "learning_rate": 4.8642904588982785e-06,
+      "loss": 0.1862,
+      "step": 580
+    },
+    {
+      "epoch": 0.26502451818907513,
+      "grad_norm": 0.32397133111953735,
+      "learning_rate": 4.8636895091765704e-06,
+      "loss": 0.1846,
+      "step": 581
+    },
+    {
+      "epoch": 0.2654806705439617,
+      "grad_norm": 0.3123103976249695,
+      "learning_rate": 4.863087269116404e-06,
+      "loss": 0.1817,
+      "step": 582
+    },
+    {
+      "epoch": 0.26593682289884824,
+      "grad_norm": 0.34392285346984863,
+      "learning_rate": 4.862483739046542e-06,
+      "loss": 0.1961,
+      "step": 583
+    },
+    {
+      "epoch": 0.26639297525373473,
+      "grad_norm": 0.32618820667266846,
+      "learning_rate": 4.861878919296451e-06,
+      "loss": 0.1823,
+      "step": 584
+    },
+    {
+      "epoch": 0.2668491276086213,
+      "grad_norm": 0.3391662538051605,
+      "learning_rate": 4.861272810196304e-06,
+      "loss": 0.1932,
+      "step": 585
+    },
+    {
+      "epoch": 0.26730527996350784,
+      "grad_norm": 0.32017019391059875,
+      "learning_rate": 4.8606654120769735e-06,
+      "loss": 0.1779,
+      "step": 586
+    },
+    {
+      "epoch": 0.26776143231839433,
+      "grad_norm": 0.3086513876914978,
+      "learning_rate": 4.860056725270041e-06,
+      "loss": 0.1772,
+      "step": 587
+    },
+    {
+      "epoch": 0.2682175846732809,
+      "grad_norm": 0.3114010691642761,
+      "learning_rate": 4.859446750107786e-06,
+      "loss": 0.1776,
+      "step": 588
+    },
+    {
+      "epoch": 0.2686737370281674,
+      "grad_norm": 0.3291904628276825,
+      "learning_rate": 4.858835486923196e-06,
+      "loss": 0.1899,
+      "step": 589
+    },
+    {
+      "epoch": 0.26912988938305393,
+      "grad_norm": 0.32585060596466064,
+      "learning_rate": 4.8582229360499585e-06,
+      "loss": 0.1866,
+      "step": 590
+    },
+    {
+      "epoch": 0.2695860417379405,
+      "grad_norm": 0.32723599672317505,
+      "learning_rate": 4.857609097822467e-06,
+      "loss": 0.1924,
+      "step": 591
+    },
+    {
+      "epoch": 0.270042194092827,
+      "grad_norm": 0.32842934131622314,
+      "learning_rate": 4.856993972575813e-06,
+      "loss": 0.1867,
+      "step": 592
+    },
+    {
+      "epoch": 0.27049834644771353,
+      "grad_norm": 0.328803688287735,
+      "learning_rate": 4.856377560645795e-06,
+      "loss": 0.1833,
+      "step": 593
+    },
+    {
+      "epoch": 0.2709544988026001,
+      "grad_norm": 0.34655970335006714,
+      "learning_rate": 4.855759862368913e-06,
+      "loss": 0.1977,
+      "step": 594
+    },
+    {
+      "epoch": 0.2714106511574866,
+      "grad_norm": 0.30820634961128235,
+      "learning_rate": 4.855140878082368e-06,
+      "loss": 0.2003,
+      "step": 595
+    },
+    {
+      "epoch": 0.27186680351237313,
+      "grad_norm": 0.3195788562297821,
+      "learning_rate": 4.854520608124063e-06,
+      "loss": 0.1857,
+      "step": 596
+    },
+    {
+      "epoch": 0.2723229558672597,
+      "grad_norm": 0.3271220624446869,
+      "learning_rate": 4.853899052832604e-06,
+      "loss": 0.1866,
+      "step": 597
+    },
+    {
+      "epoch": 0.2727791082221462,
+      "grad_norm": 0.3150160610675812,
+      "learning_rate": 4.853276212547299e-06,
+      "loss": 0.1844,
+      "step": 598
+    },
+    {
+      "epoch": 0.27323526057703273,
+      "grad_norm": 0.3389405608177185,
+      "learning_rate": 4.852652087608155e-06,
+      "loss": 0.1921,
+      "step": 599
+    },
+    {
+      "epoch": 0.2736914129319193,
+      "grad_norm": 0.33866310119628906,
+      "learning_rate": 4.8520266783558825e-06,
+      "loss": 0.193,
+      "step": 600
+    },
+    {
+      "epoch": 0.2741475652868058,
+      "grad_norm": 0.31101903319358826,
+      "learning_rate": 4.8513999851318924e-06,
+      "loss": 0.1722,
+      "step": 601
+    },
+    {
+      "epoch": 0.27460371764169234,
+      "grad_norm": 0.3181251287460327,
+      "learning_rate": 4.850772008278297e-06,
+      "loss": 0.1884,
+      "step": 602
+    },
+    {
+      "epoch": 0.27505986999657883,
+      "grad_norm": 0.3256985545158386,
+      "learning_rate": 4.85014274813791e-06,
+      "loss": 0.1885,
+      "step": 603
+    },
+    {
+      "epoch": 0.2755160223514654,
+      "grad_norm": 0.32810190320014954,
+      "learning_rate": 4.849512205054242e-06,
+      "loss": 0.178,
+      "step": 604
+    },
+    {
+      "epoch": 0.27597217470635194,
+      "grad_norm": 0.32210129499435425,
+      "learning_rate": 4.848880379371509e-06,
+      "loss": 0.1837,
+      "step": 605
+    },
+    {
+      "epoch": 0.27642832706123843,
+      "grad_norm": 0.3388887345790863,
+      "learning_rate": 4.848247271434623e-06,
+      "loss": 0.1839,
+      "step": 606
+    },
+    {
+      "epoch": 0.276884479416125,
+      "grad_norm": 0.33636143803596497,
+      "learning_rate": 4.847612881589199e-06,
+      "loss": 0.1952,
+      "step": 607
+    },
+    {
+      "epoch": 0.27734063177101154,
+      "grad_norm": 0.3170711398124695,
+      "learning_rate": 4.846977210181549e-06,
+      "loss": 0.1761,
+      "step": 608
+    },
+    {
+      "epoch": 0.27779678412589803,
+      "grad_norm": 0.3371245265007019,
+      "learning_rate": 4.846340257558688e-06,
+      "loss": 0.1942,
+      "step": 609
+    },
+    {
+      "epoch": 0.2782529364807846,
+      "grad_norm": 0.31507205963134766,
+      "learning_rate": 4.845702024068327e-06,
+      "loss": 0.1749,
+      "step": 610
+    },
+    {
+      "epoch": 0.27870908883567114,
+      "grad_norm": 0.3270321190357208,
+      "learning_rate": 4.845062510058878e-06,
+      "loss": 0.1895,
+      "step": 611
+    },
+    {
+      "epoch": 0.27916524119055763,
+      "grad_norm": 0.34832537174224854,
+      "learning_rate": 4.844421715879453e-06,
+      "loss": 0.1887,
+      "step": 612
+    },
+    {
+      "epoch": 0.2796213935454442,
+      "grad_norm": 0.33865275979042053,
+      "learning_rate": 4.843779641879861e-06,
+      "loss": 0.1826,
+      "step": 613
+    },
+    {
+      "epoch": 0.28007754590033074,
+      "grad_norm": 0.3251104950904846,
+      "learning_rate": 4.843136288410609e-06,
+      "loss": 0.1846,
+      "step": 614
+    },
+    {
+      "epoch": 0.28053369825521723,
+      "grad_norm": 0.33893606066703796,
+      "learning_rate": 4.842491655822906e-06,
+      "loss": 0.188,
+      "step": 615
+    },
+    {
+      "epoch": 0.2809898506101038,
+      "grad_norm": 0.3232625424861908,
+      "learning_rate": 4.841845744468655e-06,
+      "loss": 0.1843,
+      "step": 616
+    },
+    {
+      "epoch": 0.2814460029649903,
+      "grad_norm": 0.3356733024120331,
+      "learning_rate": 4.841198554700461e-06,
+      "loss": 0.1854,
+      "step": 617
+    },
+    {
+      "epoch": 0.28190215531987683,
+      "grad_norm": 0.3200189471244812,
+      "learning_rate": 4.840550086871623e-06,
+      "loss": 0.1879,
+      "step": 618
+    },
+    {
+      "epoch": 0.2823583076747634,
+      "grad_norm": 0.3267514109611511,
+      "learning_rate": 4.83990034133614e-06,
+      "loss": 0.1899,
+      "step": 619
+    },
+    {
+      "epoch": 0.2828144600296499,
+      "grad_norm": 0.3436368405818939,
+      "learning_rate": 4.83924931844871e-06,
+      "loss": 0.1876,
+      "step": 620
+    },
+    {
+      "epoch": 0.28327061238453644,
+      "grad_norm": 0.3443828225135803,
+      "learning_rate": 4.838597018564725e-06,
+      "loss": 0.1969,
+      "step": 621
+    },
+    {
+      "epoch": 0.283726764739423,
+      "grad_norm": 0.3394968807697296,
+      "learning_rate": 4.837943442040276e-06,
+      "loss": 0.1808,
+      "step": 622
+    },
+    {
+      "epoch": 0.2841829170943095,
+      "grad_norm": 0.3129294216632843,
+      "learning_rate": 4.83728858923215e-06,
+      "loss": 0.189,
+      "step": 623
+    },
+    {
+      "epoch": 0.28463906944919604,
+      "grad_norm": 0.3206091523170471,
+      "learning_rate": 4.836632460497832e-06,
+      "loss": 0.1875,
+      "step": 624
+    },
+    {
+      "epoch": 0.2850952218040826,
+      "grad_norm": 0.3501910865306854,
+      "learning_rate": 4.835975056195503e-06,
+      "loss": 0.1886,
+      "step": 625
+    },
+    {
+      "epoch": 0.2855513741589691,
+      "grad_norm": 0.39122045040130615,
+      "learning_rate": 4.835316376684039e-06,
+      "loss": 0.1953,
+      "step": 626
+    },
+    {
+      "epoch": 0.28600752651385564,
+      "grad_norm": 0.34854045510292053,
+      "learning_rate": 4.8346564223230145e-06,
+      "loss": 0.1863,
+      "step": 627
+    },
+    {
+      "epoch": 0.28646367886874213,
+      "grad_norm": 0.33810970187187195,
+      "learning_rate": 4.833995193472697e-06,
+      "loss": 0.1903,
+      "step": 628
+    },
+    {
+      "epoch": 0.2869198312236287,
+      "grad_norm": 0.30962079763412476,
+      "learning_rate": 4.833332690494053e-06,
+      "loss": 0.1804,
+      "step": 629
+    },
+    {
+      "epoch": 0.28737598357851524,
+      "grad_norm": 0.3120759129524231,
+      "learning_rate": 4.832668913748741e-06,
+      "loss": 0.1763,
+      "step": 630
+    },
+    {
+      "epoch": 0.28783213593340173,
+      "grad_norm": 0.3218189477920532,
+      "learning_rate": 4.832003863599121e-06,
+      "loss": 0.178,
+      "step": 631
+    },
+    {
+      "epoch": 0.2882882882882883,
+      "grad_norm": 0.3556252717971802,
+      "learning_rate": 4.831337540408239e-06,
+      "loss": 0.1947,
+      "step": 632
+    },
+    {
+      "epoch": 0.28874444064317484,
+      "grad_norm": 0.339598685503006,
+      "learning_rate": 4.830669944539844e-06,
+      "loss": 0.1929,
+      "step": 633
+    },
+    {
+      "epoch": 0.28920059299806133,
+      "grad_norm": 0.32749828696250916,
+      "learning_rate": 4.830001076358376e-06,
+      "loss": 0.1877,
+      "step": 634
+    },
+    {
+      "epoch": 0.2896567453529479,
+      "grad_norm": 0.33081379532814026,
+      "learning_rate": 4.829330936228969e-06,
+      "loss": 0.1878,
+      "step": 635
+    },
+    {
+      "epoch": 0.29011289770783444,
+      "grad_norm": 0.31877443194389343,
+      "learning_rate": 4.828659524517455e-06,
+      "loss": 0.1924,
+      "step": 636
+    },
+    {
+      "epoch": 0.29056905006272093,
+      "grad_norm": 0.3343644142150879,
+      "learning_rate": 4.827986841590355e-06,
+      "loss": 0.1859,
+      "step": 637
+    },
+    {
+      "epoch": 0.2910252024176075,
+      "grad_norm": 0.3220553994178772,
+      "learning_rate": 4.827312887814889e-06,
+      "loss": 0.1962,
+      "step": 638
+    },
+    {
+      "epoch": 0.29148135477249404,
+      "grad_norm": 0.31882330775260925,
+      "learning_rate": 4.8266376635589685e-06,
+      "loss": 0.1769,
+      "step": 639
+    },
+    {
+      "epoch": 0.29193750712738054,
+      "grad_norm": 0.33859291672706604,
+      "learning_rate": 4.825961169191196e-06,
+      "loss": 0.1928,
+      "step": 640
+    },
+    {
+      "epoch": 0.2923936594822671,
+      "grad_norm": 0.3168545663356781,
+      "learning_rate": 4.825283405080873e-06,
+      "loss": 0.1844,
+      "step": 641
+    },
+    {
+      "epoch": 0.2928498118371536,
+      "grad_norm": 0.307955801486969,
+      "learning_rate": 4.8246043715979885e-06,
+      "loss": 0.1788,
+      "step": 642
+    },
+    {
+      "epoch": 0.29330596419204014,
+      "grad_norm": 0.3036643862724304,
+      "learning_rate": 4.8239240691132285e-06,
+      "loss": 0.1751,
+      "step": 643
+    },
+    {
+      "epoch": 0.2937621165469267,
+      "grad_norm": 0.3087560534477234,
+      "learning_rate": 4.8232424979979684e-06,
+      "loss": 0.1807,
+      "step": 644
+    },
+    {
+      "epoch": 0.2942182689018132,
+      "grad_norm": 0.32730385661125183,
+      "learning_rate": 4.82255965862428e-06,
+      "loss": 0.1959,
+      "step": 645
+    },
+    {
+      "epoch": 0.29467442125669974,
+      "grad_norm": 0.32819244265556335,
+      "learning_rate": 4.821875551364927e-06,
+      "loss": 0.1758,
+      "step": 646
+    },
+    {
+      "epoch": 0.2951305736115863,
+      "grad_norm": 0.3385339081287384,
+      "learning_rate": 4.821190176593359e-06,
+      "loss": 0.1914,
+      "step": 647
+    },
+    {
+      "epoch": 0.2955867259664728,
+      "grad_norm": 0.3380511999130249,
+      "learning_rate": 4.820503534683725e-06,
+      "loss": 0.1934,
+      "step": 648
+    },
+    {
+      "epoch": 0.29604287832135934,
+      "grad_norm": 0.3233324885368347,
+      "learning_rate": 4.819815626010862e-06,
+      "loss": 0.1668,
+      "step": 649
+    },
+    {
+      "epoch": 0.2964990306762459,
+      "grad_norm": 0.3080935776233673,
+      "learning_rate": 4.819126450950301e-06,
+      "loss": 0.1859,
+      "step": 650
+    },
+    {
+      "epoch": 0.2969551830311324,
+      "grad_norm": 0.31115442514419556,
+      "learning_rate": 4.818436009878261e-06,
+      "loss": 0.1837,
+      "step": 651
+    },
+    {
+      "epoch": 0.29741133538601894,
+      "grad_norm": 0.33515509963035583,
+      "learning_rate": 4.8177443031716545e-06,
+      "loss": 0.1861,
+      "step": 652
+    },
+    {
+      "epoch": 0.2978674877409055,
+      "grad_norm": 0.3230332136154175,
+      "learning_rate": 4.817051331208084e-06,
+      "loss": 0.1844,
+      "step": 653
+    },
+    {
+      "epoch": 0.298323640095792,
+      "grad_norm": 0.3206697106361389,
+      "learning_rate": 4.816357094365845e-06,
+      "loss": 0.1869,
+      "step": 654
+    },
+    {
+      "epoch": 0.29877979245067854,
+      "grad_norm": 0.34294918179512024,
+      "learning_rate": 4.81566159302392e-06,
+      "loss": 0.1883,
+      "step": 655
+    },
+    {
+      "epoch": 0.29923594480556504,
+      "grad_norm": 0.33375123143196106,
+      "learning_rate": 4.814964827561981e-06,
+      "loss": 0.1998,
+      "step": 656
+    },
+    {
+      "epoch": 0.2996920971604516,
+      "grad_norm": 0.36231178045272827,
+      "learning_rate": 4.8142667983603954e-06,
+      "loss": 0.1882,
+      "step": 657
+    },
+    {
+      "epoch": 0.30014824951533814,
+      "grad_norm": 0.3381553292274475,
+      "learning_rate": 4.8135675058002176e-06,
+      "loss": 0.1957,
+      "step": 658
+    },
+    {
+      "epoch": 0.30060440187022464,
+      "grad_norm": 0.3309899568557739,
+      "learning_rate": 4.81286695026319e-06,
+      "loss": 0.1932,
+      "step": 659
+    },
+    {
+      "epoch": 0.3010605542251112,
+      "grad_norm": 0.3273334503173828,
+      "learning_rate": 4.812165132131746e-06,
+      "loss": 0.1836,
+      "step": 660
+    },
+    {
+      "epoch": 0.30151670657999774,
+      "grad_norm": 0.32859018445014954,
+      "learning_rate": 4.811462051789009e-06,
+      "loss": 0.1828,
+      "step": 661
+    },
+    {
+      "epoch": 0.30197285893488424,
+      "grad_norm": 0.33868131041526794,
+      "learning_rate": 4.81075770961879e-06,
+      "loss": 0.1923,
+      "step": 662
+    },
+    {
+      "epoch": 0.3024290112897708,
+      "grad_norm": 0.3335169851779938,
+      "learning_rate": 4.810052106005591e-06,
+      "loss": 0.1951,
+      "step": 663
+    },
+    {
+      "epoch": 0.30288516364465734,
+      "grad_norm": 0.3228609561920166,
+      "learning_rate": 4.809345241334598e-06,
+      "loss": 0.1808,
+      "step": 664
+    },
+    {
+      "epoch": 0.30334131599954384,
+      "grad_norm": 0.34415724873542786,
+      "learning_rate": 4.808637115991692e-06,
+      "loss": 0.1806,
+      "step": 665
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.30662181973457336,
+      "learning_rate": 4.807927730363438e-06,
+      "loss": 0.1848,
+      "step": 666
+    },
+    {
+      "epoch": 0.3042536207093169,
+      "grad_norm": 0.30515527725219727,
+      "learning_rate": 4.807217084837087e-06,
+      "loss": 0.1731,
+      "step": 667
+    },
+    {
+      "epoch": 0.30470977306420344,
+      "grad_norm": 0.3008563220500946,
+      "learning_rate": 4.806505179800583e-06,
+      "loss": 0.1803,
+      "step": 668
+    },
+    {
+      "epoch": 0.30516592541909,
+      "grad_norm": 0.3207714259624481,
+      "learning_rate": 4.805792015642555e-06,
+      "loss": 0.1738,
+      "step": 669
+    },
+    {
+      "epoch": 0.3056220777739765,
+      "grad_norm": 0.3091341257095337,
+      "learning_rate": 4.805077592752319e-06,
+      "loss": 0.1821,
+      "step": 670
+    },
+    {
+      "epoch": 0.30607823012886304,
+      "grad_norm": 0.3387342691421509,
+      "learning_rate": 4.804361911519879e-06,
+      "loss": 0.1791,
+      "step": 671
+    },
+    {
+      "epoch": 0.3065343824837496,
+      "grad_norm": 0.3401768207550049,
+      "learning_rate": 4.803644972335925e-06,
+      "loss": 0.1836,
+      "step": 672
+    },
+    {
+      "epoch": 0.3069905348386361,
+      "grad_norm": 0.32615286111831665,
+      "learning_rate": 4.8029267755918325e-06,
+      "loss": 0.1821,
+      "step": 673
+    },
+    {
+      "epoch": 0.30744668719352264,
+      "grad_norm": 0.35637909173965454,
+      "learning_rate": 4.802207321679669e-06,
+      "loss": 0.1927,
+      "step": 674
+    },
+    {
+      "epoch": 0.3079028395484092,
+      "grad_norm": 0.3130182921886444,
+      "learning_rate": 4.8014866109921805e-06,
+      "loss": 0.1807,
+      "step": 675
+    },
+    {
+      "epoch": 0.3083589919032957,
+      "grad_norm": 0.3272121846675873,
+      "learning_rate": 4.800764643922806e-06,
+      "loss": 0.1868,
+      "step": 676
+    },
+    {
+      "epoch": 0.30881514425818224,
+      "grad_norm": 0.33786797523498535,
+      "learning_rate": 4.8000414208656665e-06,
+      "loss": 0.1715,
+      "step": 677
+    },
+    {
+      "epoch": 0.3092712966130688,
+      "grad_norm": 0.33341503143310547,
+      "learning_rate": 4.79931694221557e-06,
+      "loss": 0.1877,
+      "step": 678
+    },
+    {
+      "epoch": 0.3097274489679553,
+      "grad_norm": 0.3236059546470642,
+      "learning_rate": 4.798591208368008e-06,
+      "loss": 0.1961,
+      "step": 679
+    },
+    {
+      "epoch": 0.31018360132284184,
+      "grad_norm": 0.3282153010368347,
+      "learning_rate": 4.797864219719161e-06,
+      "loss": 0.1884,
+      "step": 680
+    },
+    {
+      "epoch": 0.31063975367772834,
+      "grad_norm": 0.3386630415916443,
+      "learning_rate": 4.797135976665889e-06,
+      "loss": 0.1823,
+      "step": 681
+    },
+    {
+      "epoch": 0.3110959060326149,
+      "grad_norm": 0.32271265983581543,
+      "learning_rate": 4.796406479605743e-06,
+      "loss": 0.1826,
+      "step": 682
+    },
+    {
+      "epoch": 0.31155205838750144,
+      "grad_norm": 0.32866325974464417,
+      "learning_rate": 4.7956757289369545e-06,
+      "loss": 0.182,
+      "step": 683
+    },
+    {
+      "epoch": 0.31200821074238794,
+      "grad_norm": 0.3348342180252075,
+      "learning_rate": 4.794943725058441e-06,
+      "loss": 0.1898,
+      "step": 684
+    },
+    {
+      "epoch": 0.3124643630972745,
+      "grad_norm": 0.31914207339286804,
+      "learning_rate": 4.794210468369802e-06,
+      "loss": 0.1836,
+      "step": 685
+    },
+    {
+      "epoch": 0.31292051545216104,
+      "grad_norm": 0.3274868130683899,
+      "learning_rate": 4.793475959271324e-06,
+      "loss": 0.1877,
+      "step": 686
+    },
+    {
+      "epoch": 0.31337666780704754,
+      "grad_norm": 0.34306320548057556,
+      "learning_rate": 4.792740198163974e-06,
+      "loss": 0.1929,
+      "step": 687
+    },
+    {
+      "epoch": 0.3138328201619341,
+      "grad_norm": 0.3342038094997406,
+      "learning_rate": 4.792003185449406e-06,
+      "loss": 0.1921,
+      "step": 688
+    },
+    {
+      "epoch": 0.31428897251682064,
+      "grad_norm": 0.33734455704689026,
+      "learning_rate": 4.791264921529954e-06,
+      "loss": 0.1793,
+      "step": 689
+    },
+    {
+      "epoch": 0.31474512487170714,
+      "grad_norm": 0.33868956565856934,
+      "learning_rate": 4.790525406808637e-06,
+      "loss": 0.1832,
+      "step": 690
+    },
+    {
+      "epoch": 0.3152012772265937,
+      "grad_norm": 0.3032340407371521,
+      "learning_rate": 4.789784641689156e-06,
+      "loss": 0.1745,
+      "step": 691
+    },
+    {
+      "epoch": 0.3156574295814802,
+      "grad_norm": 0.3176478445529938,
+      "learning_rate": 4.789042626575895e-06,
+      "loss": 0.1872,
+      "step": 692
+    },
+    {
+      "epoch": 0.31611358193636674,
+      "grad_norm": 0.32524633407592773,
+      "learning_rate": 4.7882993618739195e-06,
+      "loss": 0.1807,
+      "step": 693
+    },
+    {
+      "epoch": 0.3165697342912533,
+      "grad_norm": 0.33026036620140076,
+      "learning_rate": 4.787554847988979e-06,
+      "loss": 0.1959,
+      "step": 694
+    },
+    {
+      "epoch": 0.3170258866461398,
+      "grad_norm": 0.32658708095550537,
+      "learning_rate": 4.786809085327502e-06,
+      "loss": 0.1913,
+      "step": 695
+    },
+    {
+      "epoch": 0.31748203900102634,
+      "grad_norm": 0.3467511832714081,
+      "learning_rate": 4.786062074296602e-06,
+      "loss": 0.1978,
+      "step": 696
+    },
+    {
+      "epoch": 0.3179381913559129,
+      "grad_norm": 0.342766135931015,
+      "learning_rate": 4.7853138153040725e-06,
+      "loss": 0.1885,
+      "step": 697
+    },
+    {
+      "epoch": 0.3183943437107994,
+      "grad_norm": 0.3280925154685974,
+      "learning_rate": 4.784564308758386e-06,
+      "loss": 0.1798,
+      "step": 698
+    },
+    {
+      "epoch": 0.31885049606568594,
+      "grad_norm": 0.3347899317741394,
+      "learning_rate": 4.783813555068701e-06,
+      "loss": 0.185,
+      "step": 699
+    },
+    {
+      "epoch": 0.3193066484205725,
+      "grad_norm": 0.321150004863739,
+      "learning_rate": 4.783061554644853e-06,
+      "loss": 0.1749,
+      "step": 700
+    },
+    {
+      "epoch": 0.319762800775459,
+      "grad_norm": 0.33380016684532166,
+      "learning_rate": 4.782308307897359e-06,
+      "loss": 0.1967,
+      "step": 701
+    },
+    {
+      "epoch": 0.32021895313034554,
+      "grad_norm": 0.31356051564216614,
+      "learning_rate": 4.7815538152374165e-06,
+      "loss": 0.1851,
+      "step": 702
+    },
+    {
+      "epoch": 0.3206751054852321,
+      "grad_norm": 0.31227415800094604,
+      "learning_rate": 4.780798077076903e-06,
+      "loss": 0.1826,
+      "step": 703
+    },
+    {
+      "epoch": 0.3211312578401186,
+      "grad_norm": 0.3198064863681793,
+      "learning_rate": 4.780041093828376e-06,
+      "loss": 0.1862,
+      "step": 704
+    },
+    {
+      "epoch": 0.32158741019500514,
+      "grad_norm": 0.3399809002876282,
+      "learning_rate": 4.7792828659050735e-06,
+      "loss": 0.1936,
+      "step": 705
+    },
+    {
+      "epoch": 0.32204356254989164,
+      "grad_norm": 0.3192659914493561,
+      "learning_rate": 4.778523393720911e-06,
+      "loss": 0.1803,
+      "step": 706
+    },
+    {
+      "epoch": 0.3224997149047782,
+      "grad_norm": 0.3209479749202728,
+      "learning_rate": 4.777762677690486e-06,
+      "loss": 0.1834,
+      "step": 707
+    },
+    {
+      "epoch": 0.32295586725966474,
+      "grad_norm": 0.31552308797836304,
+      "learning_rate": 4.777000718229072e-06,
+      "loss": 0.1771,
+      "step": 708
+    },
+    {
+      "epoch": 0.32341201961455124,
+      "grad_norm": 0.30813658237457275,
+      "learning_rate": 4.776237515752625e-06,
+      "loss": 0.1795,
+      "step": 709
+    },
+    {
+      "epoch": 0.3238681719694378,
+      "grad_norm": 0.3131260573863983,
+      "learning_rate": 4.775473070677775e-06,
+      "loss": 0.1835,
+      "step": 710
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 0.3254958391189575,
+      "learning_rate": 4.774707383421833e-06,
+      "loss": 0.178,
+      "step": 711
+    },
+    {
+      "epoch": 0.32478047667921084,
+      "grad_norm": 0.3313031494617462,
+      "learning_rate": 4.773940454402789e-06,
+      "loss": 0.1886,
+      "step": 712
+    },
+    {
+      "epoch": 0.3252366290340974,
+      "grad_norm": 0.3319737911224365,
+      "learning_rate": 4.773172284039309e-06,
+      "loss": 0.1836,
+      "step": 713
+    },
+    {
+      "epoch": 0.32569278138898394,
+      "grad_norm": 0.3118657171726227,
+      "learning_rate": 4.772402872750738e-06,
+      "loss": 0.1839,
+      "step": 714
+    },
+    {
+      "epoch": 0.32614893374387044,
+      "grad_norm": 0.33755141496658325,
+      "learning_rate": 4.771632220957096e-06,
+      "loss": 0.1835,
+      "step": 715
+    },
+    {
+      "epoch": 0.326605086098757,
+      "grad_norm": 0.31846287846565247,
+      "learning_rate": 4.770860329079083e-06,
+      "loss": 0.1759,
+      "step": 716
+    },
+    {
+      "epoch": 0.32706123845364354,
+      "grad_norm": 0.32218137383461,
+      "learning_rate": 4.7700871975380755e-06,
+      "loss": 0.1812,
+      "step": 717
+    },
+    {
+      "epoch": 0.32751739080853004,
+      "grad_norm": 0.3278481364250183,
+      "learning_rate": 4.769312826756126e-06,
+      "loss": 0.1865,
+      "step": 718
+    },
+    {
+      "epoch": 0.3279735431634166,
+      "grad_norm": 0.31637537479400635,
+      "learning_rate": 4.768537217155962e-06,
+      "loss": 0.1818,
+      "step": 719
+    },
+    {
+      "epoch": 0.3284296955183031,
+      "grad_norm": 0.30573880672454834,
+      "learning_rate": 4.7677603691609905e-06,
+      "loss": 0.1698,
+      "step": 720
+    },
+    {
+      "epoch": 0.32888584787318964,
+      "grad_norm": 0.31987571716308594,
+      "learning_rate": 4.766982283195292e-06,
+      "loss": 0.1812,
+      "step": 721
+    },
+    {
+      "epoch": 0.3293420002280762,
+      "grad_norm": 0.32963794469833374,
+      "learning_rate": 4.7662029596836255e-06,
+      "loss": 0.1779,
+      "step": 722
+    },
+    {
+      "epoch": 0.3297981525829627,
+      "grad_norm": 0.3098617196083069,
+      "learning_rate": 4.7654223990514214e-06,
+      "loss": 0.185,
+      "step": 723
+    },
+    {
+      "epoch": 0.33025430493784924,
+      "grad_norm": 0.3117121756076813,
+      "learning_rate": 4.7646406017247895e-06,
+      "loss": 0.1774,
+      "step": 724
+    },
+    {
+      "epoch": 0.3307104572927358,
+      "grad_norm": 0.3244931101799011,
+      "learning_rate": 4.763857568130512e-06,
+      "loss": 0.1881,
+      "step": 725
+    },
+    {
+      "epoch": 0.3311666096476223,
+      "grad_norm": 0.31307151913642883,
+      "learning_rate": 4.763073298696049e-06,
+      "loss": 0.1935,
+      "step": 726
+    },
+    {
+      "epoch": 0.33162276200250884,
+      "grad_norm": 0.3236372768878937,
+      "learning_rate": 4.76228779384953e-06,
+      "loss": 0.1805,
+      "step": 727
+    },
+    {
+      "epoch": 0.3320789143573954,
+      "grad_norm": 0.34224697947502136,
+      "learning_rate": 4.761501054019766e-06,
+      "loss": 0.1865,
+      "step": 728
+    },
+    {
+      "epoch": 0.3325350667122819,
+      "grad_norm": 0.3122299313545227,
+      "learning_rate": 4.760713079636236e-06,
+      "loss": 0.1813,
+      "step": 729
+    },
+    {
+      "epoch": 0.33299121906716844,
+      "grad_norm": 0.3301961421966553,
+      "learning_rate": 4.759923871129096e-06,
+      "loss": 0.1882,
+      "step": 730
+    },
+    {
+      "epoch": 0.33344737142205494,
+      "grad_norm": 0.3279363811016083,
+      "learning_rate": 4.759133428929174e-06,
+      "loss": 0.1875,
+      "step": 731
+    },
+    {
+      "epoch": 0.3339035237769415,
+      "grad_norm": 0.3137544095516205,
+      "learning_rate": 4.758341753467975e-06,
+      "loss": 0.179,
+      "step": 732
+    },
+    {
+      "epoch": 0.33435967613182804,
+      "grad_norm": 0.3000304400920868,
+      "learning_rate": 4.757548845177672e-06,
+      "loss": 0.1703,
+      "step": 733
+    },
+    {
+      "epoch": 0.33481582848671454,
+      "grad_norm": 0.3185427784919739,
+      "learning_rate": 4.756754704491115e-06,
+      "loss": 0.1831,
+      "step": 734
+    },
+    {
+      "epoch": 0.3352719808416011,
+      "grad_norm": 0.3135800063610077,
+      "learning_rate": 4.755959331841826e-06,
+      "loss": 0.1818,
+      "step": 735
+    },
+    {
+      "epoch": 0.33572813319648764,
+      "grad_norm": 0.3081668019294739,
+      "learning_rate": 4.755162727663998e-06,
+      "loss": 0.1765,
+      "step": 736
+    },
+    {
+      "epoch": 0.33618428555137414,
+      "grad_norm": 0.3195345103740692,
+      "learning_rate": 4.754364892392499e-06,
+      "loss": 0.1854,
+      "step": 737
+    },
+    {
+      "epoch": 0.3366404379062607,
+      "grad_norm": 0.34218689799308777,
+      "learning_rate": 4.7535658264628646e-06,
+      "loss": 0.1942,
+      "step": 738
+    },
+    {
+      "epoch": 0.33709659026114724,
+      "grad_norm": 0.36137616634368896,
+      "learning_rate": 4.752765530311308e-06,
+      "loss": 0.1942,
+      "step": 739
+    },
+    {
+      "epoch": 0.33755274261603374,
+      "grad_norm": 0.30246302485466003,
+      "learning_rate": 4.751964004374709e-06,
+      "loss": 0.1731,
+      "step": 740
+    },
+    {
+      "epoch": 0.3380088949709203,
+      "grad_norm": 0.31947270035743713,
+      "learning_rate": 4.751161249090621e-06,
+      "loss": 0.1837,
+      "step": 741
+    },
+    {
+      "epoch": 0.33846504732580684,
+      "grad_norm": 0.32863304018974304,
+      "learning_rate": 4.75035726489727e-06,
+      "loss": 0.1858,
+      "step": 742
+    },
+    {
+      "epoch": 0.33892119968069334,
+      "grad_norm": 0.31922587752342224,
+      "learning_rate": 4.749552052233549e-06,
+      "loss": 0.184,
+      "step": 743
+    },
+    {
+      "epoch": 0.3393773520355799,
+      "grad_norm": 0.314726859331131,
+      "learning_rate": 4.748745611539024e-06,
+      "loss": 0.1791,
+      "step": 744
+    },
+    {
+      "epoch": 0.3398335043904664,
+      "grad_norm": 0.33290842175483704,
+      "learning_rate": 4.747937943253932e-06,
+      "loss": 0.1825,
+      "step": 745
+    },
+    {
+      "epoch": 0.34028965674535294,
+      "grad_norm": 0.34637632966041565,
+      "learning_rate": 4.747129047819179e-06,
+      "loss": 0.1923,
+      "step": 746
+    },
+    {
+      "epoch": 0.3407458091002395,
+      "grad_norm": 0.34507614374160767,
+      "learning_rate": 4.746318925676341e-06,
+      "loss": 0.193,
+      "step": 747
+    },
+    {
+      "epoch": 0.341201961455126,
+      "grad_norm": 0.3102602958679199,
+      "learning_rate": 4.745507577267663e-06,
+      "loss": 0.1847,
+      "step": 748
+    },
+    {
+      "epoch": 0.34165811381001254,
+      "grad_norm": 0.34404727816581726,
+      "learning_rate": 4.744695003036061e-06,
+      "loss": 0.192,
+      "step": 749
+    },
+    {
+      "epoch": 0.3421142661648991,
+      "grad_norm": 0.33767130970954895,
+      "learning_rate": 4.74388120342512e-06,
+      "loss": 0.1913,
+      "step": 750
+    },
+    {
+      "epoch": 0.3425704185197856,
+      "grad_norm": 0.29585522413253784,
+      "learning_rate": 4.743066178879092e-06,
+      "loss": 0.1765,
+      "step": 751
+    },
+    {
+      "epoch": 0.34302657087467214,
+      "grad_norm": 0.31045567989349365,
+      "learning_rate": 4.7422499298429e-06,
+      "loss": 0.1825,
+      "step": 752
+    },
+    {
+      "epoch": 0.3434827232295587,
+      "grad_norm": 0.3103623390197754,
+      "learning_rate": 4.741432456762134e-06,
+      "loss": 0.1874,
+      "step": 753
+    },
+    {
+      "epoch": 0.3439388755844452,
+      "grad_norm": 0.34627896547317505,
+      "learning_rate": 4.740613760083052e-06,
+      "loss": 0.2076,
+      "step": 754
+    },
+    {
+      "epoch": 0.34439502793933174,
+      "grad_norm": 0.321254700422287,
+      "learning_rate": 4.739793840252583e-06,
+      "loss": 0.1895,
+      "step": 755
+    },
+    {
+      "epoch": 0.3448511802942183,
+      "grad_norm": 0.31450793147087097,
+      "learning_rate": 4.738972697718319e-06,
+      "loss": 0.1767,
+      "step": 756
+    },
+    {
+      "epoch": 0.3453073326491048,
+      "grad_norm": 0.3260285556316376,
+      "learning_rate": 4.738150332928523e-06,
+      "loss": 0.1769,
+      "step": 757
+    },
+    {
+      "epoch": 0.34576348500399134,
+      "grad_norm": 0.30420246720314026,
+      "learning_rate": 4.737326746332124e-06,
+      "loss": 0.1844,
+      "step": 758
+    },
+    {
+      "epoch": 0.34621963735887784,
+      "grad_norm": 0.3235284984111786,
+      "learning_rate": 4.736501938378717e-06,
+      "loss": 0.1854,
+      "step": 759
+    },
+    {
+      "epoch": 0.3466757897137644,
+      "grad_norm": 0.3209013342857361,
+      "learning_rate": 4.735675909518565e-06,
+      "loss": 0.1873,
+      "step": 760
+    },
+    {
+      "epoch": 0.34713194206865094,
+      "grad_norm": 0.3339729309082031,
+      "learning_rate": 4.734848660202599e-06,
+      "loss": 0.1893,
+      "step": 761
+    },
+    {
+      "epoch": 0.34758809442353744,
+      "grad_norm": 0.329868346452713,
+      "learning_rate": 4.734020190882413e-06,
+      "loss": 0.1829,
+      "step": 762
+    },
+    {
+      "epoch": 0.348044246778424,
+      "grad_norm": 0.30221396684646606,
+      "learning_rate": 4.733190502010269e-06,
+      "loss": 0.1714,
+      "step": 763
+    },
+    {
+      "epoch": 0.34850039913331055,
+      "grad_norm": 0.32563963532447815,
+      "learning_rate": 4.732359594039094e-06,
+      "loss": 0.1881,
+      "step": 764
+    },
+    {
+      "epoch": 0.34895655148819704,
+      "grad_norm": 0.3266116678714752,
+      "learning_rate": 4.731527467422481e-06,
+      "loss": 0.1862,
+      "step": 765
+    },
+    {
+      "epoch": 0.3494127038430836,
+      "grad_norm": 0.30555808544158936,
+      "learning_rate": 4.730694122614688e-06,
+      "loss": 0.1818,
+      "step": 766
+    },
+    {
+      "epoch": 0.34986885619797015,
+      "grad_norm": 0.32462236285209656,
+      "learning_rate": 4.729859560070638e-06,
+      "loss": 0.1934,
+      "step": 767
+    },
+    {
+      "epoch": 0.35032500855285664,
+      "grad_norm": 0.3051515817642212,
+      "learning_rate": 4.729023780245919e-06,
+      "loss": 0.1772,
+      "step": 768
+    },
+    {
+      "epoch": 0.3507811609077432,
+      "grad_norm": 0.31362780928611755,
+      "learning_rate": 4.728186783596783e-06,
+      "loss": 0.1755,
+      "step": 769
+    },
+    {
+      "epoch": 0.3512373132626297,
+      "grad_norm": 0.3330015540122986,
+      "learning_rate": 4.727348570580147e-06,
+      "loss": 0.1838,
+      "step": 770
+    },
+    {
+      "epoch": 0.35169346561751624,
+      "grad_norm": 0.3189384639263153,
+      "learning_rate": 4.726509141653591e-06,
+      "loss": 0.1789,
+      "step": 771
+    },
+    {
+      "epoch": 0.3521496179724028,
+      "grad_norm": 0.3253416121006012,
+      "learning_rate": 4.725668497275361e-06,
+      "loss": 0.1829,
+      "step": 772
+    },
+    {
+      "epoch": 0.3526057703272893,
+      "grad_norm": 0.3212781250476837,
+      "learning_rate": 4.724826637904362e-06,
+      "loss": 0.1927,
+      "step": 773
+    },
+    {
+      "epoch": 0.35306192268217584,
+      "grad_norm": 0.32843947410583496,
+      "learning_rate": 4.723983564000168e-06,
+      "loss": 0.1816,
+      "step": 774
+    },
+    {
+      "epoch": 0.3535180750370624,
+      "grad_norm": 0.3207141160964966,
+      "learning_rate": 4.723139276023012e-06,
+      "loss": 0.1872,
+      "step": 775
+    },
+    {
+      "epoch": 0.3539742273919489,
+      "grad_norm": 0.3114449381828308,
+      "learning_rate": 4.72229377443379e-06,
+      "loss": 0.1755,
+      "step": 776
+    },
+    {
+      "epoch": 0.35443037974683544,
+      "grad_norm": 0.32432764768600464,
+      "learning_rate": 4.721447059694063e-06,
+      "loss": 0.1843,
+      "step": 777
+    },
+    {
+      "epoch": 0.354886532101722,
+      "grad_norm": 0.3253295123577118,
+      "learning_rate": 4.720599132266052e-06,
+      "loss": 0.1785,
+      "step": 778
+    },
+    {
+      "epoch": 0.3553426844566085,
+      "grad_norm": 0.3052341639995575,
+      "learning_rate": 4.719749992612641e-06,
+      "loss": 0.1783,
+      "step": 779
+    },
+    {
+      "epoch": 0.35579883681149505,
+      "grad_norm": 0.33046671748161316,
+      "learning_rate": 4.718899641197375e-06,
+      "loss": 0.1891,
+      "step": 780
+    },
+    {
+      "epoch": 0.3562549891663816,
+      "grad_norm": 0.3296401798725128,
+      "learning_rate": 4.718048078484461e-06,
+      "loss": 0.1858,
+      "step": 781
+    },
+    {
+      "epoch": 0.3567111415212681,
+      "grad_norm": 0.32206159830093384,
+      "learning_rate": 4.717195304938768e-06,
+      "loss": 0.1803,
+      "step": 782
+    },
+    {
+      "epoch": 0.35716729387615465,
+      "grad_norm": 0.31181758642196655,
+      "learning_rate": 4.716341321025825e-06,
+      "loss": 0.1791,
+      "step": 783
+    },
+    {
+      "epoch": 0.35762344623104114,
+      "grad_norm": 0.320717453956604,
+      "learning_rate": 4.71548612721182e-06,
+      "loss": 0.1811,
+      "step": 784
+    },
+    {
+      "epoch": 0.3580795985859277,
+      "grad_norm": 0.32423365116119385,
+      "learning_rate": 4.714629723963606e-06,
+      "loss": 0.1925,
+      "step": 785
+    },
+    {
+      "epoch": 0.35853575094081425,
+      "grad_norm": 0.3308560848236084,
+      "learning_rate": 4.713772111748693e-06,
+      "loss": 0.187,
+      "step": 786
+    },
+    {
+      "epoch": 0.35899190329570074,
+      "grad_norm": 0.3121544420719147,
+      "learning_rate": 4.712913291035251e-06,
+      "loss": 0.1756,
+      "step": 787
+    },
+    {
+      "epoch": 0.3594480556505873,
+      "grad_norm": 0.3259446620941162,
+      "learning_rate": 4.712053262292111e-06,
+      "loss": 0.1829,
+      "step": 788
+    },
+    {
+      "epoch": 0.35990420800547385,
+      "grad_norm": 0.34747666120529175,
+      "learning_rate": 4.711192025988762e-06,
+      "loss": 0.1977,
+      "step": 789
+    },
+    {
+      "epoch": 0.36036036036036034,
+      "grad_norm": 0.32702142000198364,
+      "learning_rate": 4.710329582595354e-06,
+      "loss": 0.1818,
+      "step": 790
+    },
+    {
+      "epoch": 0.3608165127152469,
+      "grad_norm": 0.3245146870613098,
+      "learning_rate": 4.709465932582694e-06,
+      "loss": 0.1868,
+      "step": 791
+    },
+    {
+      "epoch": 0.36127266507013345,
+      "grad_norm": 0.3361288011074066,
+      "learning_rate": 4.70860107642225e-06,
+      "loss": 0.1828,
+      "step": 792
+    },
+    {
+      "epoch": 0.36172881742501994,
+      "grad_norm": 0.31716376543045044,
+      "learning_rate": 4.707735014586146e-06,
+      "loss": 0.1784,
+      "step": 793
+    },
+    {
+      "epoch": 0.3621849697799065,
+      "grad_norm": 0.32183414697647095,
+      "learning_rate": 4.7068677475471644e-06,
+      "loss": 0.1822,
+      "step": 794
+    },
+    {
+      "epoch": 0.36264112213479305,
+      "grad_norm": 0.31124773621559143,
+      "learning_rate": 4.705999275778749e-06,
+      "loss": 0.1801,
+      "step": 795
+    },
+    {
+      "epoch": 0.36309727448967954,
+      "grad_norm": 0.3171737492084503,
+      "learning_rate": 4.7051295997549964e-06,
+      "loss": 0.1843,
+      "step": 796
+    },
+    {
+      "epoch": 0.3635534268445661,
+      "grad_norm": 0.3090074062347412,
+      "learning_rate": 4.704258719950664e-06,
+      "loss": 0.1715,
+      "step": 797
+    },
+    {
+      "epoch": 0.3640095791994526,
+      "grad_norm": 0.33406710624694824,
+      "learning_rate": 4.7033866368411655e-06,
+      "loss": 0.1921,
+      "step": 798
+    },
+    {
+      "epoch": 0.36446573155433915,
+      "grad_norm": 0.3011876046657562,
+      "learning_rate": 4.70251335090257e-06,
+      "loss": 0.1738,
+      "step": 799
+    },
+    {
+      "epoch": 0.3649218839092257,
+      "grad_norm": 0.32348188757896423,
+      "learning_rate": 4.701638862611605e-06,
+      "loss": 0.1833,
+      "step": 800
+    },
+    {
+      "epoch": 0.3653780362641122,
+      "grad_norm": 0.32232600450515747,
+      "learning_rate": 4.7007631724456534e-06,
+      "loss": 0.1839,
+      "step": 801
+    },
+    {
+      "epoch": 0.36583418861899875,
+      "grad_norm": 0.3285371661186218,
+      "learning_rate": 4.6998862808827564e-06,
+      "loss": 0.1907,
+      "step": 802
+    },
+    {
+      "epoch": 0.3662903409738853,
+      "grad_norm": 0.3231484293937683,
+      "learning_rate": 4.699008188401607e-06,
+      "loss": 0.1797,
+      "step": 803
+    },
+    {
+      "epoch": 0.3667464933287718,
+      "grad_norm": 0.3512055277824402,
+      "learning_rate": 4.698128895481557e-06,
+      "loss": 0.1856,
+      "step": 804
+    },
+    {
+      "epoch": 0.36720264568365835,
+      "grad_norm": 0.33501172065734863,
+      "learning_rate": 4.697248402602613e-06,
+      "loss": 0.1909,
+      "step": 805
+    },
+    {
+      "epoch": 0.3676587980385449,
+      "grad_norm": 0.3313238024711609,
+      "learning_rate": 4.696366710245435e-06,
+      "loss": 0.1858,
+      "step": 806
+    },
+    {
+      "epoch": 0.3681149503934314,
+      "grad_norm": 0.3317754566669464,
+      "learning_rate": 4.69548381889134e-06,
+      "loss": 0.1768,
+      "step": 807
+    },
+    {
+      "epoch": 0.36857110274831795,
+      "grad_norm": 0.31768131256103516,
+      "learning_rate": 4.694599729022297e-06,
+      "loss": 0.1869,
+      "step": 808
+    },
+    {
+      "epoch": 0.36902725510320444,
+      "grad_norm": 0.35013508796691895,
+      "learning_rate": 4.693714441120933e-06,
+      "loss": 0.1929,
+      "step": 809
+    },
+    {
+      "epoch": 0.369483407458091,
+      "grad_norm": 0.33270907402038574,
+      "learning_rate": 4.692827955670526e-06,
+      "loss": 0.1899,
+      "step": 810
+    },
+    {
+      "epoch": 0.36993955981297755,
+      "grad_norm": 0.33312204480171204,
+      "learning_rate": 4.691940273155008e-06,
+      "loss": 0.1839,
+      "step": 811
+    },
+    {
+      "epoch": 0.37039571216786404,
+      "grad_norm": 0.33113792538642883,
+      "learning_rate": 4.691051394058965e-06,
+      "loss": 0.1803,
+      "step": 812
+    },
+    {
+      "epoch": 0.3708518645227506,
+      "grad_norm": 0.3166777789592743,
+      "learning_rate": 4.690161318867636e-06,
+      "loss": 0.1746,
+      "step": 813
+    },
+    {
+      "epoch": 0.37130801687763715,
+      "grad_norm": 0.3250664472579956,
+      "learning_rate": 4.6892700480669155e-06,
+      "loss": 0.1796,
+      "step": 814
+    },
+    {
+      "epoch": 0.37176416923252364,
+      "grad_norm": 0.326790988445282,
+      "learning_rate": 4.688377582143345e-06,
+      "loss": 0.1851,
+      "step": 815
+    },
+    {
+      "epoch": 0.3722203215874102,
+      "grad_norm": 0.33060479164123535,
+      "learning_rate": 4.687483921584124e-06,
+      "loss": 0.191,
+      "step": 816
+    },
+    {
+      "epoch": 0.37267647394229675,
+      "grad_norm": 0.31363940238952637,
+      "learning_rate": 4.686589066877101e-06,
+      "loss": 0.1916,
+      "step": 817
+    },
+    {
+      "epoch": 0.37313262629718325,
+      "grad_norm": 0.3291283845901489,
+      "learning_rate": 4.6856930185107776e-06,
+      "loss": 0.1872,
+      "step": 818
+    },
+    {
+      "epoch": 0.3735887786520698,
+      "grad_norm": 0.3110540807247162,
+      "learning_rate": 4.684795776974307e-06,
+      "loss": 0.1806,
+      "step": 819
+    },
+    {
+      "epoch": 0.37404493100695635,
+      "grad_norm": 0.33516925573349,
+      "learning_rate": 4.683897342757493e-06,
+      "loss": 0.1849,
+      "step": 820
+    },
+    {
+      "epoch": 0.37450108336184285,
+      "grad_norm": 0.3327985405921936,
+      "learning_rate": 4.6829977163507916e-06,
+      "loss": 0.1858,
+      "step": 821
+    },
+    {
+      "epoch": 0.3749572357167294,
+      "grad_norm": 0.3493301570415497,
+      "learning_rate": 4.682096898245308e-06,
+      "loss": 0.1986,
+      "step": 822
+    },
+    {
+      "epoch": 0.3754133880716159,
+      "grad_norm": 0.33128485083580017,
+      "learning_rate": 4.6811948889327995e-06,
+      "loss": 0.1833,
+      "step": 823
+    },
+    {
+      "epoch": 0.37586954042650245,
+      "grad_norm": 0.30384600162506104,
+      "learning_rate": 4.680291688905674e-06,
+      "loss": 0.1854,
+      "step": 824
+    },
+    {
+      "epoch": 0.376325692781389,
+      "grad_norm": 0.31636878848075867,
+      "learning_rate": 4.6793872986569875e-06,
+      "loss": 0.1865,
+      "step": 825
+    },
+    {
+      "epoch": 0.3767818451362755,
+      "grad_norm": 0.32121264934539795,
+      "learning_rate": 4.678481718680446e-06,
+      "loss": 0.1819,
+      "step": 826
+    },
+    {
+      "epoch": 0.37723799749116205,
+      "grad_norm": 0.3307656943798065,
+      "learning_rate": 4.677574949470408e-06,
+      "loss": 0.1873,
+      "step": 827
+    },
+    {
+      "epoch": 0.3776941498460486,
+      "grad_norm": 0.32052236795425415,
+      "learning_rate": 4.676666991521876e-06,
+      "loss": 0.1831,
+      "step": 828
+    },
+    {
+      "epoch": 0.3781503022009351,
+      "grad_norm": 0.31357574462890625,
+      "learning_rate": 4.675757845330508e-06,
+      "loss": 0.1754,
+      "step": 829
+    },
+    {
+      "epoch": 0.37860645455582165,
+      "grad_norm": 0.31481024622917175,
+      "learning_rate": 4.674847511392604e-06,
+      "loss": 0.1826,
+      "step": 830
+    },
+    {
+      "epoch": 0.3790626069107082,
+      "grad_norm": 0.31135737895965576,
+      "learning_rate": 4.673935990205116e-06,
+      "loss": 0.1789,
+      "step": 831
+    },
+    {
+      "epoch": 0.3795187592655947,
+      "grad_norm": 0.3337706923484802,
+      "learning_rate": 4.673023282265645e-06,
+      "loss": 0.188,
+      "step": 832
+    },
+    {
+      "epoch": 0.37997491162048125,
+      "grad_norm": 0.32273125648498535,
+      "learning_rate": 4.672109388072438e-06,
+      "loss": 0.1923,
+      "step": 833
+    },
+    {
+      "epoch": 0.3804310639753678,
+      "grad_norm": 0.3034324645996094,
+      "learning_rate": 4.671194308124388e-06,
+      "loss": 0.1676,
+      "step": 834
+    },
+    {
+      "epoch": 0.3808872163302543,
+      "grad_norm": 0.3207845389842987,
+      "learning_rate": 4.67027804292104e-06,
+      "loss": 0.1791,
+      "step": 835
+    },
+    {
+      "epoch": 0.38134336868514085,
+      "grad_norm": 0.3326779901981354,
+      "learning_rate": 4.669360592962581e-06,
+      "loss": 0.1854,
+      "step": 836
+    },
+    {
+      "epoch": 0.38179952104002735,
+      "grad_norm": 0.32716864347457886,
+      "learning_rate": 4.668441958749848e-06,
+      "loss": 0.1843,
+      "step": 837
+    },
+    {
+      "epoch": 0.3822556733949139,
+      "grad_norm": 0.31938284635543823,
+      "learning_rate": 4.667522140784324e-06,
+      "loss": 0.1922,
+      "step": 838
+    },
+    {
+      "epoch": 0.38271182574980045,
+      "grad_norm": 0.32528191804885864,
+      "learning_rate": 4.6666011395681375e-06,
+      "loss": 0.1832,
+      "step": 839
+    },
+    {
+      "epoch": 0.38316797810468695,
+      "grad_norm": 0.3120166063308716,
+      "learning_rate": 4.665678955604064e-06,
+      "loss": 0.1845,
+      "step": 840
+    },
+    {
+      "epoch": 0.3836241304595735,
+      "grad_norm": 0.3062160313129425,
+      "learning_rate": 4.664755589395522e-06,
+      "loss": 0.1878,
+      "step": 841
+    },
+    {
+      "epoch": 0.38408028281446005,
+      "grad_norm": 0.31849101185798645,
+      "learning_rate": 4.663831041446578e-06,
+      "loss": 0.1826,
+      "step": 842
+    },
+    {
+      "epoch": 0.38453643516934655,
+      "grad_norm": 0.3389931917190552,
+      "learning_rate": 4.662905312261944e-06,
+      "loss": 0.186,
+      "step": 843
+    },
+    {
+      "epoch": 0.3849925875242331,
+      "grad_norm": 0.3107641339302063,
+      "learning_rate": 4.661978402346974e-06,
+      "loss": 0.1938,
+      "step": 844
+    },
+    {
+      "epoch": 0.38544873987911965,
+      "grad_norm": 0.32172077894210815,
+      "learning_rate": 4.66105031220767e-06,
+      "loss": 0.1842,
+      "step": 845
+    },
+    {
+      "epoch": 0.38590489223400615,
+      "grad_norm": 0.30493271350860596,
+      "learning_rate": 4.6601210423506746e-06,
+      "loss": 0.1774,
+      "step": 846
+    },
+    {
+      "epoch": 0.3863610445888927,
+      "grad_norm": 0.32963982224464417,
+      "learning_rate": 4.659190593283278e-06,
+      "loss": 0.1816,
+      "step": 847
+    },
+    {
+      "epoch": 0.3868171969437792,
+      "grad_norm": 0.3351307511329651,
+      "learning_rate": 4.658258965513412e-06,
+      "loss": 0.178,
+      "step": 848
+    },
+    {
+      "epoch": 0.38727334929866575,
+      "grad_norm": 0.31453830003738403,
+      "learning_rate": 4.657326159549652e-06,
+      "loss": 0.1858,
+      "step": 849
+    },
+    {
+      "epoch": 0.3877295016535523,
+      "grad_norm": 0.33353087306022644,
+      "learning_rate": 4.656392175901217e-06,
+      "loss": 0.1852,
+      "step": 850
+    },
+    {
+      "epoch": 0.3881856540084388,
+      "grad_norm": 0.2970513701438904,
+      "learning_rate": 4.655457015077969e-06,
+      "loss": 0.1881,
+      "step": 851
+    },
+    {
+      "epoch": 0.38864180636332535,
+      "grad_norm": 0.32646217942237854,
+      "learning_rate": 4.654520677590412e-06,
+      "loss": 0.1914,
+      "step": 852
+    },
+    {
+      "epoch": 0.3890979587182119,
+      "grad_norm": 0.32203829288482666,
+      "learning_rate": 4.653583163949694e-06,
+      "loss": 0.1853,
+      "step": 853
+    },
+    {
+      "epoch": 0.3895541110730984,
+      "grad_norm": 0.3246288597583771,
+      "learning_rate": 4.652644474667603e-06,
+      "loss": 0.1854,
+      "step": 854
+    },
+    {
+      "epoch": 0.39001026342798495,
+      "grad_norm": 0.3387046754360199,
+      "learning_rate": 4.651704610256568e-06,
+      "loss": 0.1896,
+      "step": 855
+    },
+    {
+      "epoch": 0.3904664157828715,
+      "grad_norm": 0.3063892722129822,
+      "learning_rate": 4.650763571229664e-06,
+      "loss": 0.183,
+      "step": 856
+    },
+    {
+      "epoch": 0.390922568137758,
+      "grad_norm": 0.3367040753364563,
+      "learning_rate": 4.649821358100603e-06,
+      "loss": 0.1825,
+      "step": 857
+    },
+    {
+      "epoch": 0.39137872049264455,
+      "grad_norm": 0.3044869601726532,
+      "learning_rate": 4.648877971383739e-06,
+      "loss": 0.1781,
+      "step": 858
+    },
+    {
+      "epoch": 0.3918348728475311,
+      "grad_norm": 0.3244490623474121,
+      "learning_rate": 4.647933411594068e-06,
+      "loss": 0.1844,
+      "step": 859
+    },
+    {
+      "epoch": 0.3922910252024176,
+      "grad_norm": 0.3292739987373352,
+      "learning_rate": 4.646987679247223e-06,
+      "loss": 0.1882,
+      "step": 860
+    },
+    {
+      "epoch": 0.39274717755730415,
+      "grad_norm": 0.3370338976383209,
+      "learning_rate": 4.646040774859483e-06,
+      "loss": 0.1988,
+      "step": 861
+    },
+    {
+      "epoch": 0.39320332991219065,
+      "grad_norm": 0.3305016756057739,
+      "learning_rate": 4.645092698947761e-06,
+      "loss": 0.2008,
+      "step": 862
+    },
+    {
+      "epoch": 0.3936594822670772,
+      "grad_norm": 0.3123026192188263,
+      "learning_rate": 4.644143452029611e-06,
+      "loss": 0.1765,
+      "step": 863
+    },
+    {
+      "epoch": 0.39411563462196375,
+      "grad_norm": 0.3094756603240967,
+      "learning_rate": 4.643193034623229e-06,
+      "loss": 0.1775,
+      "step": 864
+    },
+    {
+      "epoch": 0.39457178697685025,
+      "grad_norm": 0.30703040957450867,
+      "learning_rate": 4.642241447247447e-06,
+      "loss": 0.1801,
+      "step": 865
+    },
+    {
+      "epoch": 0.3950279393317368,
+      "grad_norm": 0.31493717432022095,
+      "learning_rate": 4.641288690421737e-06,
+      "loss": 0.1753,
+      "step": 866
+    },
+    {
+      "epoch": 0.39548409168662335,
+      "grad_norm": 0.3090752959251404,
+      "learning_rate": 4.640334764666208e-06,
+      "loss": 0.1834,
+      "step": 867
+    },
+    {
+      "epoch": 0.39594024404150985,
+      "grad_norm": 0.3200157582759857,
+      "learning_rate": 4.6393796705016105e-06,
+      "loss": 0.1864,
+      "step": 868
+    },
+    {
+      "epoch": 0.3963963963963964,
+      "grad_norm": 0.3082413375377655,
+      "learning_rate": 4.638423408449329e-06,
+      "loss": 0.1874,
+      "step": 869
+    },
+    {
+      "epoch": 0.39685254875128295,
+      "grad_norm": 0.3035981059074402,
+      "learning_rate": 4.6374659790313855e-06,
+      "loss": 0.1746,
+      "step": 870
+    },
+    {
+      "epoch": 0.39730870110616945,
+      "grad_norm": 0.32199594378471375,
+      "learning_rate": 4.6365073827704445e-06,
+      "loss": 0.1769,
+      "step": 871
+    },
+    {
+      "epoch": 0.397764853461056,
+      "grad_norm": 0.3274296522140503,
+      "learning_rate": 4.635547620189802e-06,
+      "loss": 0.1902,
+      "step": 872
+    },
+    {
+      "epoch": 0.39822100581594255,
+      "grad_norm": 0.30711469054222107,
+      "learning_rate": 4.634586691813391e-06,
+      "loss": 0.1776,
+      "step": 873
+    },
+    {
+      "epoch": 0.39867715817082905,
+      "grad_norm": 0.32410046458244324,
+      "learning_rate": 4.633624598165785e-06,
+      "loss": 0.1845,
+      "step": 874
+    },
+    {
+      "epoch": 0.3991333105257156,
+      "grad_norm": 0.3116839826107025,
+      "learning_rate": 4.63266133977219e-06,
+      "loss": 0.1778,
+      "step": 875
+    },
+    {
+      "epoch": 0.3995894628806021,
+      "grad_norm": 0.3269287943840027,
+      "learning_rate": 4.631696917158449e-06,
+      "loss": 0.1865,
+      "step": 876
+    },
+    {
+      "epoch": 0.40004561523548865,
+      "grad_norm": 0.32391420006752014,
+      "learning_rate": 4.630731330851041e-06,
+      "loss": 0.1852,
+      "step": 877
+    },
+    {
+      "epoch": 0.4005017675903752,
+      "grad_norm": 0.3267417848110199,
+      "learning_rate": 4.629764581377077e-06,
+      "loss": 0.1889,
+      "step": 878
+    },
+    {
+      "epoch": 0.4009579199452617,
+      "grad_norm": 0.33576029539108276,
+      "learning_rate": 4.628796669264311e-06,
+      "loss": 0.186,
+      "step": 879
+    },
+    {
+      "epoch": 0.40141407230014825,
+      "grad_norm": 0.31001895666122437,
+      "learning_rate": 4.62782759504112e-06,
+      "loss": 0.1909,
+      "step": 880
+    },
+    {
+      "epoch": 0.4018702246550348,
+      "grad_norm": 0.301137775182724,
+      "learning_rate": 4.626857359236527e-06,
+      "loss": 0.1707,
+      "step": 881
+    },
+    {
+      "epoch": 0.4023263770099213,
+      "grad_norm": 0.33584073185920715,
+      "learning_rate": 4.625885962380181e-06,
+      "loss": 0.1962,
+      "step": 882
+    },
+    {
+      "epoch": 0.40278252936480785,
+      "grad_norm": 0.31457439064979553,
+      "learning_rate": 4.624913405002368e-06,
+      "loss": 0.185,
+      "step": 883
+    },
+    {
+      "epoch": 0.4032386817196944,
+      "grad_norm": 0.30383747816085815,
+      "learning_rate": 4.623939687634009e-06,
+      "loss": 0.1747,
+      "step": 884
+    },
+    {
+      "epoch": 0.4036948340745809,
+      "grad_norm": 0.3180914521217346,
+      "learning_rate": 4.622964810806654e-06,
+      "loss": 0.189,
+      "step": 885
+    },
+    {
+      "epoch": 0.40415098642946745,
+      "grad_norm": 0.3123425841331482,
+      "learning_rate": 4.62198877505249e-06,
+      "loss": 0.1748,
+      "step": 886
+    },
+    {
+      "epoch": 0.40460713878435395,
+      "grad_norm": 0.3254973888397217,
+      "learning_rate": 4.621011580904334e-06,
+      "loss": 0.1831,
+      "step": 887
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.3253493905067444,
+      "learning_rate": 4.620033228895639e-06,
+      "loss": 0.1978,
+      "step": 888
+    },
+    {
+      "epoch": 0.40551944349412705,
+      "grad_norm": 0.33728930354118347,
+      "learning_rate": 4.619053719560483e-06,
+      "loss": 0.1763,
+      "step": 889
+    },
+    {
+      "epoch": 0.40597559584901355,
+      "grad_norm": 0.3005589246749878,
+      "learning_rate": 4.618073053433585e-06,
+      "loss": 0.1657,
+      "step": 890
+    },
+    {
+      "epoch": 0.4064317482039001,
+      "grad_norm": 0.3256130814552307,
+      "learning_rate": 4.617091231050287e-06,
+      "loss": 0.1778,
+      "step": 891
+    },
+    {
+      "epoch": 0.40688790055878665,
+      "grad_norm": 0.31482994556427,
+      "learning_rate": 4.616108252946568e-06,
+      "loss": 0.1852,
+      "step": 892
+    },
+    {
+      "epoch": 0.40734405291367315,
+      "grad_norm": 0.3129308521747589,
+      "learning_rate": 4.6151241196590365e-06,
+      "loss": 0.1795,
+      "step": 893
+    },
+    {
+      "epoch": 0.4078002052685597,
+      "grad_norm": 0.3356710970401764,
+      "learning_rate": 4.61413883172493e-06,
+      "loss": 0.187,
+      "step": 894
+    },
+    {
+      "epoch": 0.40825635762344625,
+      "grad_norm": 0.3260447680950165,
+      "learning_rate": 4.613152389682117e-06,
+      "loss": 0.1839,
+      "step": 895
+    },
+    {
+      "epoch": 0.40871250997833275,
+      "grad_norm": 0.30863022804260254,
+      "learning_rate": 4.612164794069096e-06,
+      "loss": 0.1848,
+      "step": 896
+    },
+    {
+      "epoch": 0.4091686623332193,
+      "grad_norm": 0.31158310174942017,
+      "learning_rate": 4.6111760454249965e-06,
+      "loss": 0.181,
+      "step": 897
+    },
+    {
+      "epoch": 0.40962481468810585,
+      "grad_norm": 0.361092746257782,
+      "learning_rate": 4.610186144289577e-06,
+      "loss": 0.192,
+      "step": 898
+    },
+    {
+      "epoch": 0.41008096704299235,
+      "grad_norm": 0.3518179953098297,
+      "learning_rate": 4.6091950912032235e-06,
+      "loss": 0.1842,
+      "step": 899
+    },
+    {
+      "epoch": 0.4105371193978789,
+      "grad_norm": 0.3163907527923584,
+      "learning_rate": 4.608202886706953e-06,
+      "loss": 0.1826,
+      "step": 900
+    },
+    {
+      "epoch": 0.4109932717527654,
+      "grad_norm": 0.30330589413642883,
+      "learning_rate": 4.607209531342409e-06,
+      "loss": 0.1778,
+      "step": 901
+    },
+    {
+      "epoch": 0.41144942410765195,
+      "grad_norm": 0.3233715295791626,
+      "learning_rate": 4.606215025651866e-06,
+      "loss": 0.1834,
+      "step": 902
+    },
+    {
+      "epoch": 0.4119055764625385,
+      "grad_norm": 0.30021730065345764,
+      "learning_rate": 4.6052193701782235e-06,
+      "loss": 0.1683,
+      "step": 903
+    },
+    {
+      "epoch": 0.412361728817425,
+      "grad_norm": 0.3365362286567688,
+      "learning_rate": 4.6042225654650096e-06,
+      "loss": 0.1822,
+      "step": 904
+    },
+    {
+      "epoch": 0.41281788117231155,
+      "grad_norm": 0.3295752704143524,
+      "learning_rate": 4.603224612056382e-06,
+      "loss": 0.1786,
+      "step": 905
+    },
+    {
+      "epoch": 0.4132740335271981,
+      "grad_norm": 0.3206825256347656,
+      "learning_rate": 4.602225510497121e-06,
+      "loss": 0.1828,
+      "step": 906
+    },
+    {
+      "epoch": 0.4137301858820846,
+      "grad_norm": 0.35625067353248596,
+      "learning_rate": 4.6012252613326385e-06,
+      "loss": 0.1985,
+      "step": 907
+    },
+    {
+      "epoch": 0.41418633823697115,
+      "grad_norm": 0.3219689726829529,
+      "learning_rate": 4.60022386510897e-06,
+      "loss": 0.1857,
+      "step": 908
+    },
+    {
+      "epoch": 0.4146424905918577,
+      "grad_norm": 0.3333626091480255,
+      "learning_rate": 4.5992213223727775e-06,
+      "loss": 0.1815,
+      "step": 909
+    },
+    {
+      "epoch": 0.4150986429467442,
+      "grad_norm": 0.3252718150615692,
+      "learning_rate": 4.598217633671349e-06,
+      "loss": 0.1925,
+      "step": 910
+    },
+    {
+      "epoch": 0.41555479530163075,
+      "grad_norm": 0.3060569167137146,
+      "learning_rate": 4.597212799552599e-06,
+      "loss": 0.1756,
+      "step": 911
+    },
+    {
+      "epoch": 0.41601094765651725,
+      "grad_norm": 0.3107914924621582,
+      "learning_rate": 4.5962068205650674e-06,
+      "loss": 0.1835,
+      "step": 912
+    },
+    {
+      "epoch": 0.4164671000114038,
+      "grad_norm": 0.31634923815727234,
+      "learning_rate": 4.595199697257917e-06,
+      "loss": 0.1899,
+      "step": 913
+    },
+    {
+      "epoch": 0.41692325236629035,
+      "grad_norm": 0.32350093126296997,
+      "learning_rate": 4.594191430180936e-06,
+      "loss": 0.1757,
+      "step": 914
+    },
+    {
+      "epoch": 0.41737940472117685,
+      "grad_norm": 0.33588549494743347,
+      "learning_rate": 4.593182019884539e-06,
+      "loss": 0.1749,
+      "step": 915
+    },
+    {
+      "epoch": 0.4178355570760634,
+      "grad_norm": 0.34761321544647217,
+      "learning_rate": 4.592171466919762e-06,
+      "loss": 0.1726,
+      "step": 916
+    },
+    {
+      "epoch": 0.41829170943094995,
+      "grad_norm": 0.3587896227836609,
+      "learning_rate": 4.591159771838267e-06,
+      "loss": 0.1953,
+      "step": 917
+    },
+    {
+      "epoch": 0.41874786178583645,
+      "grad_norm": 0.3153538107872009,
+      "learning_rate": 4.5901469351923375e-06,
+      "loss": 0.1699,
+      "step": 918
+    },
+    {
+      "epoch": 0.419204014140723,
+      "grad_norm": 0.3278065025806427,
+      "learning_rate": 4.589132957534883e-06,
+      "loss": 0.1843,
+      "step": 919
+    },
+    {
+      "epoch": 0.41966016649560955,
+      "grad_norm": 0.30511313676834106,
+      "learning_rate": 4.588117839419432e-06,
+      "loss": 0.1799,
+      "step": 920
+    },
+    {
+      "epoch": 0.42011631885049605,
+      "grad_norm": 0.3209461271762848,
+      "learning_rate": 4.587101581400139e-06,
+      "loss": 0.1797,
+      "step": 921
+    },
+    {
+      "epoch": 0.4205724712053826,
+      "grad_norm": 0.31860652565956116,
+      "learning_rate": 4.586084184031778e-06,
+      "loss": 0.1864,
+      "step": 922
+    },
+    {
+      "epoch": 0.42102862356026916,
+      "grad_norm": 0.35664108395576477,
+      "learning_rate": 4.585065647869748e-06,
+      "loss": 0.19,
+      "step": 923
+    },
+    {
+      "epoch": 0.42148477591515565,
+      "grad_norm": 0.31350767612457275,
+      "learning_rate": 4.584045973470067e-06,
+      "loss": 0.1752,
+      "step": 924
+    },
+    {
+      "epoch": 0.4219409282700422,
+      "grad_norm": 0.3262849748134613,
+      "learning_rate": 4.583025161389376e-06,
+      "loss": 0.1949,
+      "step": 925
+    },
+    {
+      "epoch": 0.4223970806249287,
+      "grad_norm": 0.3226277232170105,
+      "learning_rate": 4.582003212184937e-06,
+      "loss": 0.1847,
+      "step": 926
+    },
+    {
+      "epoch": 0.42285323297981525,
+      "grad_norm": 0.3453037142753601,
+      "learning_rate": 4.580980126414629e-06,
+      "loss": 0.1889,
+      "step": 927
+    },
+    {
+      "epoch": 0.4233093853347018,
+      "grad_norm": 0.3371812403202057,
+      "learning_rate": 4.579955904636959e-06,
+      "loss": 0.1937,
+      "step": 928
+    },
+    {
+      "epoch": 0.4237655376895883,
+      "grad_norm": 0.33281436562538147,
+      "learning_rate": 4.578930547411048e-06,
+      "loss": 0.1833,
+      "step": 929
+    },
+    {
+      "epoch": 0.42422169004447485,
+      "grad_norm": 0.32040923833847046,
+      "learning_rate": 4.577904055296638e-06,
+      "loss": 0.1761,
+      "step": 930
+    },
+    {
+      "epoch": 0.4246778423993614,
+      "grad_norm": 0.31436535716056824,
+      "learning_rate": 4.576876428854091e-06,
+      "loss": 0.1816,
+      "step": 931
+    },
+    {
+      "epoch": 0.4251339947542479,
+      "grad_norm": 0.3223901093006134,
+      "learning_rate": 4.5758476686443905e-06,
+      "loss": 0.1705,
+      "step": 932
+    },
+    {
+      "epoch": 0.42559014710913445,
+      "grad_norm": 0.3209950923919678,
+      "learning_rate": 4.5748177752291356e-06,
+      "loss": 0.1768,
+      "step": 933
+    },
+    {
+      "epoch": 0.426046299464021,
+      "grad_norm": 0.3268280327320099,
+      "learning_rate": 4.573786749170544e-06,
+      "loss": 0.1814,
+      "step": 934
+    },
+    {
+      "epoch": 0.4265024518189075,
+      "grad_norm": 0.3238708972930908,
+      "learning_rate": 4.572754591031455e-06,
+      "loss": 0.1869,
+      "step": 935
+    },
+    {
+      "epoch": 0.42695860417379405,
+      "grad_norm": 0.33111047744750977,
+      "learning_rate": 4.571721301375323e-06,
+      "loss": 0.1848,
+      "step": 936
+    },
+    {
+      "epoch": 0.4274147565286806,
+      "grad_norm": 0.31891363859176636,
+      "learning_rate": 4.570686880766222e-06,
+      "loss": 0.1832,
+      "step": 937
+    },
+    {
+      "epoch": 0.4278709088835671,
+      "grad_norm": 0.29862523078918457,
+      "learning_rate": 4.56965132976884e-06,
+      "loss": 0.1723,
+      "step": 938
+    },
+    {
+      "epoch": 0.42832706123845365,
+      "grad_norm": 0.3275577425956726,
+      "learning_rate": 4.568614648948486e-06,
+      "loss": 0.1849,
+      "step": 939
+    },
+    {
+      "epoch": 0.42878321359334015,
+      "grad_norm": 0.32358846068382263,
+      "learning_rate": 4.5675768388710855e-06,
+      "loss": 0.1785,
+      "step": 940
+    },
+    {
+      "epoch": 0.4292393659482267,
+      "grad_norm": 0.31089192628860474,
+      "learning_rate": 4.566537900103176e-06,
+      "loss": 0.1715,
+      "step": 941
+    },
+    {
+      "epoch": 0.42969551830311326,
+      "grad_norm": 0.31759703159332275,
+      "learning_rate": 4.565497833211917e-06,
+      "loss": 0.1831,
+      "step": 942
+    },
+    {
+      "epoch": 0.43015167065799975,
+      "grad_norm": 0.33461329340934753,
+      "learning_rate": 4.5644566387650805e-06,
+      "loss": 0.1987,
+      "step": 943
+    },
+    {
+      "epoch": 0.4306078230128863,
+      "grad_norm": 0.3221607506275177,
+      "learning_rate": 4.563414317331053e-06,
+      "loss": 0.1874,
+      "step": 944
+    },
+    {
+      "epoch": 0.43106397536777286,
+      "grad_norm": 0.332340806722641,
+      "learning_rate": 4.562370869478841e-06,
+      "loss": 0.1876,
+      "step": 945
+    },
+    {
+      "epoch": 0.43152012772265935,
+      "grad_norm": 0.2983841001987457,
+      "learning_rate": 4.56132629577806e-06,
+      "loss": 0.1761,
+      "step": 946
+    },
+    {
+      "epoch": 0.4319762800775459,
+      "grad_norm": 0.30900871753692627,
+      "learning_rate": 4.560280596798946e-06,
+      "loss": 0.1837,
+      "step": 947
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 0.3051842749118805,
+      "learning_rate": 4.559233773112343e-06,
+      "loss": 0.185,
+      "step": 948
+    },
+    {
+      "epoch": 0.43288858478731895,
+      "grad_norm": 0.31908267736434937,
+      "learning_rate": 4.558185825289714e-06,
+      "loss": 0.1776,
+      "step": 949
+    },
+    {
+      "epoch": 0.4333447371422055,
+      "grad_norm": 0.33089491724967957,
+      "learning_rate": 4.557136753903133e-06,
+      "loss": 0.1802,
+      "step": 950
+    },
+    {
+      "epoch": 0.433800889497092,
+      "grad_norm": 0.33471742272377014,
+      "learning_rate": 4.556086559525289e-06,
+      "loss": 0.1789,
+      "step": 951
+    },
+    {
+      "epoch": 0.43425704185197855,
+      "grad_norm": 0.32713890075683594,
+      "learning_rate": 4.5550352427294836e-06,
+      "loss": 0.1797,
+      "step": 952
+    },
+    {
+      "epoch": 0.4347131942068651,
+      "grad_norm": 0.3173023760318756,
+      "learning_rate": 4.55398280408963e-06,
+      "loss": 0.1802,
+      "step": 953
+    },
+    {
+      "epoch": 0.4351693465617516,
+      "grad_norm": 0.3266894817352295,
+      "learning_rate": 4.552929244180254e-06,
+      "loss": 0.1956,
+      "step": 954
+    },
+    {
+      "epoch": 0.43562549891663815,
+      "grad_norm": 0.3276515007019043,
+      "learning_rate": 4.551874563576496e-06,
+      "loss": 0.1794,
+      "step": 955
+    },
+    {
+      "epoch": 0.4360816512715247,
+      "grad_norm": 0.31627368927001953,
+      "learning_rate": 4.550818762854105e-06,
+      "loss": 0.1789,
+      "step": 956
+    },
+    {
+      "epoch": 0.4365378036264112,
+      "grad_norm": 0.30590522289276123,
+      "learning_rate": 4.549761842589442e-06,
+      "loss": 0.1773,
+      "step": 957
+    },
+    {
+      "epoch": 0.43699395598129775,
+      "grad_norm": 0.3182070553302765,
+      "learning_rate": 4.548703803359482e-06,
+      "loss": 0.1874,
+      "step": 958
+    },
+    {
+      "epoch": 0.4374501083361843,
+      "grad_norm": 0.3101583421230316,
+      "learning_rate": 4.547644645741807e-06,
+      "loss": 0.1775,
+      "step": 959
+    },
+    {
+      "epoch": 0.4379062606910708,
+      "grad_norm": 0.29207471013069153,
+      "learning_rate": 4.546584370314613e-06,
+      "loss": 0.1732,
+      "step": 960
+    },
+    {
+      "epoch": 0.43836241304595736,
+      "grad_norm": 0.3010220229625702,
+      "learning_rate": 4.545522977656702e-06,
+      "loss": 0.1748,
+      "step": 961
+    },
+    {
+      "epoch": 0.4388185654008439,
+      "grad_norm": 0.32280778884887695,
+      "learning_rate": 4.5444604683474915e-06,
+      "loss": 0.181,
+      "step": 962
+    },
+    {
+      "epoch": 0.4392747177557304,
+      "grad_norm": 0.33040815591812134,
+      "learning_rate": 4.543396842967003e-06,
+      "loss": 0.1812,
+      "step": 963
+    },
+    {
+      "epoch": 0.43973087011061696,
+      "grad_norm": 0.3100931644439697,
+      "learning_rate": 4.542332102095871e-06,
+      "loss": 0.179,
+      "step": 964
+    },
+    {
+      "epoch": 0.44018702246550345,
+      "grad_norm": 0.3178858757019043,
+      "learning_rate": 4.541266246315337e-06,
+      "loss": 0.1774,
+      "step": 965
+    },
+    {
+      "epoch": 0.44064317482039,
+      "grad_norm": 0.30414026975631714,
+      "learning_rate": 4.5401992762072525e-06,
+      "loss": 0.1747,
+      "step": 966
+    },
+    {
+      "epoch": 0.44109932717527656,
+      "grad_norm": 0.3274037539958954,
+      "learning_rate": 4.539131192354076e-06,
+      "loss": 0.1829,
+      "step": 967
+    },
+    {
+      "epoch": 0.44155547953016305,
+      "grad_norm": 0.3198661804199219,
+      "learning_rate": 4.538061995338875e-06,
+      "loss": 0.177,
+      "step": 968
+    },
+    {
+      "epoch": 0.4420116318850496,
+      "grad_norm": 0.3221239745616913,
+      "learning_rate": 4.536991685745325e-06,
+      "loss": 0.1852,
+      "step": 969
+    },
+    {
+      "epoch": 0.44246778423993616,
+      "grad_norm": 0.3158005177974701,
+      "learning_rate": 4.535920264157706e-06,
+      "loss": 0.1851,
+      "step": 970
+    },
+    {
+      "epoch": 0.44292393659482265,
+      "grad_norm": 0.3480580151081085,
+      "learning_rate": 4.534847731160911e-06,
+      "loss": 0.199,
+      "step": 971
+    },
+    {
+      "epoch": 0.4433800889497092,
+      "grad_norm": 0.31699442863464355,
+      "learning_rate": 4.533774087340431e-06,
+      "loss": 0.1745,
+      "step": 972
+    },
+    {
+      "epoch": 0.44383624130459576,
+      "grad_norm": 0.31012094020843506,
+      "learning_rate": 4.5326993332823735e-06,
+      "loss": 0.1757,
+      "step": 973
+    },
+    {
+      "epoch": 0.44429239365948225,
+      "grad_norm": 0.31004664301872253,
+      "learning_rate": 4.5316234695734424e-06,
+      "loss": 0.1825,
+      "step": 974
+    },
+    {
+      "epoch": 0.4447485460143688,
+      "grad_norm": 0.3240938186645508,
+      "learning_rate": 4.5305464968009545e-06,
+      "loss": 0.1828,
+      "step": 975
+    },
+    {
+      "epoch": 0.44520469836925536,
+      "grad_norm": 0.31526750326156616,
+      "learning_rate": 4.529468415552829e-06,
+      "loss": 0.1875,
+      "step": 976
+    },
+    {
+      "epoch": 0.44566085072414185,
+      "grad_norm": 0.31451746821403503,
+      "learning_rate": 4.528389226417589e-06,
+      "loss": 0.1842,
+      "step": 977
+    },
+    {
+      "epoch": 0.4461170030790284,
+      "grad_norm": 0.30880266427993774,
+      "learning_rate": 4.527308929984366e-06,
+      "loss": 0.1826,
+      "step": 978
+    },
+    {
+      "epoch": 0.4465731554339149,
+      "grad_norm": 0.3010977506637573,
+      "learning_rate": 4.5262275268428936e-06,
+      "loss": 0.1801,
+      "step": 979
+    },
+    {
+      "epoch": 0.44702930778880146,
+      "grad_norm": 0.32594212889671326,
+      "learning_rate": 4.52514501758351e-06,
+      "loss": 0.1824,
+      "step": 980
+    },
+    {
+      "epoch": 0.447485460143688,
+      "grad_norm": 0.3239097595214844,
+      "learning_rate": 4.524061402797155e-06,
+      "loss": 0.1846,
+      "step": 981
+    },
+    {
+      "epoch": 0.4479416124985745,
+      "grad_norm": 0.33570918440818787,
+      "learning_rate": 4.522976683075377e-06,
+      "loss": 0.1751,
+      "step": 982
+    },
+    {
+      "epoch": 0.44839776485346106,
+      "grad_norm": 0.32964858412742615,
+      "learning_rate": 4.521890859010324e-06,
+      "loss": 0.198,
+      "step": 983
+    },
+    {
+      "epoch": 0.4488539172083476,
+      "grad_norm": 0.30950650572776794,
+      "learning_rate": 4.520803931194747e-06,
+      "loss": 0.1845,
+      "step": 984
+    },
+    {
+      "epoch": 0.4493100695632341,
+      "grad_norm": 0.34167081117630005,
+      "learning_rate": 4.5197159002219995e-06,
+      "loss": 0.205,
+      "step": 985
+    },
+    {
+      "epoch": 0.44976622191812066,
+      "grad_norm": 0.3003822863101959,
+      "learning_rate": 4.518626766686039e-06,
+      "loss": 0.1751,
+      "step": 986
+    },
+    {
+      "epoch": 0.4502223742730072,
+      "grad_norm": 0.30479103326797485,
+      "learning_rate": 4.517536531181423e-06,
+      "loss": 0.1744,
+      "step": 987
+    },
+    {
+      "epoch": 0.4506785266278937,
+      "grad_norm": 0.3145301938056946,
+      "learning_rate": 4.5164451943033105e-06,
+      "loss": 0.1802,
+      "step": 988
+    },
+    {
+      "epoch": 0.45113467898278026,
+      "grad_norm": 0.33077433705329895,
+      "learning_rate": 4.5153527566474645e-06,
+      "loss": 0.197,
+      "step": 989
+    },
+    {
+      "epoch": 0.45159083133766675,
+      "grad_norm": 0.3036116063594818,
+      "learning_rate": 4.514259218810244e-06,
+      "loss": 0.1749,
+      "step": 990
+    },
+    {
+      "epoch": 0.4520469836925533,
+      "grad_norm": 0.3283868730068207,
+      "learning_rate": 4.513164581388614e-06,
+      "loss": 0.1954,
+      "step": 991
+    },
+    {
+      "epoch": 0.45250313604743986,
+      "grad_norm": 0.3224160373210907,
+      "learning_rate": 4.512068844980136e-06,
+      "loss": 0.1794,
+      "step": 992
+    },
+    {
+      "epoch": 0.45295928840232635,
+      "grad_norm": 0.31259632110595703,
+      "learning_rate": 4.510972010182973e-06,
+      "loss": 0.1787,
+      "step": 993
+    },
+    {
+      "epoch": 0.4534154407572129,
+      "grad_norm": 0.340762197971344,
+      "learning_rate": 4.509874077595888e-06,
+      "loss": 0.184,
+      "step": 994
+    },
+    {
+      "epoch": 0.45387159311209946,
+      "grad_norm": 0.3239292800426483,
+      "learning_rate": 4.508775047818241e-06,
+      "loss": 0.1852,
+      "step": 995
+    },
+    {
+      "epoch": 0.45432774546698595,
+      "grad_norm": 0.3123815953731537,
+      "learning_rate": 4.507674921449994e-06,
+      "loss": 0.1776,
+      "step": 996
+    },
+    {
+      "epoch": 0.4547838978218725,
+      "grad_norm": 0.324153870344162,
+      "learning_rate": 4.506573699091705e-06,
+      "loss": 0.1784,
+      "step": 997
+    },
+    {
+      "epoch": 0.45524005017675906,
+      "grad_norm": 0.33521366119384766,
+      "learning_rate": 4.505471381344532e-06,
+      "loss": 0.1876,
+      "step": 998
+    },
+    {
+      "epoch": 0.45569620253164556,
+      "grad_norm": 0.30435261130332947,
+      "learning_rate": 4.5043679688102305e-06,
+      "loss": 0.1751,
+      "step": 999
+    },
+    {
+      "epoch": 0.4561523548865321,
+      "grad_norm": 0.3096970021724701,
+      "learning_rate": 4.503263462091153e-06,
+      "loss": 0.1772,
+      "step": 1000
+    },
+    {
+      "epoch": 0.45660850724141866,
+      "grad_norm": 0.31173086166381836,
+      "learning_rate": 4.502157861790251e-06,
+      "loss": 0.1831,
+      "step": 1001
+    },
+    {
+      "epoch": 0.45706465959630516,
+      "grad_norm": 0.30706673860549927,
+      "learning_rate": 4.501051168511071e-06,
+      "loss": 0.1855,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4575208119511917,
+      "grad_norm": 0.3058623671531677,
+      "learning_rate": 4.499943382857757e-06,
+      "loss": 0.174,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4579769643060782,
+      "grad_norm": 0.3191463053226471,
+      "learning_rate": 4.49883450543505e-06,
+      "loss": 0.1905,
+      "step": 1004
+    },
+    {
+      "epoch": 0.45843311666096476,
+      "grad_norm": 0.32602742314338684,
+      "learning_rate": 4.497724536848285e-06,
+      "loss": 0.1855,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4588892690158513,
+      "grad_norm": 0.31472569704055786,
+      "learning_rate": 4.496613477703396e-06,
+      "loss": 0.1739,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4593454213707378,
+      "grad_norm": 0.3072871267795563,
+      "learning_rate": 4.4955013286069086e-06,
+      "loss": 0.172,
+      "step": 1007
+    },
+    {
+      "epoch": 0.45980157372562436,
+      "grad_norm": 0.3149460554122925,
+      "learning_rate": 4.494388090165947e-06,
+      "loss": 0.1904,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4602577260805109,
+      "grad_norm": 0.30591487884521484,
+      "learning_rate": 4.493273762988226e-06,
+      "loss": 0.179,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4607138784353974,
+      "grad_norm": 0.3252745270729065,
+      "learning_rate": 4.492158347682058e-06,
+      "loss": 0.1827,
+      "step": 1010
+    },
+    {
+      "epoch": 0.46117003079028396,
+      "grad_norm": 0.3028480112552643,
+      "learning_rate": 4.491041844856348e-06,
+      "loss": 0.1785,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4616261831451705,
+      "grad_norm": 0.3118377923965454,
+      "learning_rate": 4.489924255120597e-06,
+      "loss": 0.1915,
+      "step": 1012
+    },
+    {
+      "epoch": 0.462082335500057,
+      "grad_norm": 0.3277367353439331,
+      "learning_rate": 4.488805579084895e-06,
+      "loss": 0.1793,
+      "step": 1013
+    },
+    {
+      "epoch": 0.46253848785494356,
+      "grad_norm": 0.31564077734947205,
+      "learning_rate": 4.487685817359929e-06,
+      "loss": 0.186,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4629946402098301,
+      "grad_norm": 0.3242451548576355,
+      "learning_rate": 4.486564970556975e-06,
+      "loss": 0.1876,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4634507925647166,
+      "grad_norm": 0.33848950266838074,
+      "learning_rate": 4.485443039287907e-06,
+      "loss": 0.1965,
+      "step": 1016
+    },
+    {
+      "epoch": 0.46390694491960316,
+      "grad_norm": 0.3352382481098175,
+      "learning_rate": 4.4843200241651835e-06,
+      "loss": 0.1823,
+      "step": 1017
+    },
+    {
+      "epoch": 0.46436309727448966,
+      "grad_norm": 0.3247874081134796,
+      "learning_rate": 4.483195925801861e-06,
+      "loss": 0.1926,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4648192496293762,
+      "grad_norm": 0.3780592978000641,
+      "learning_rate": 4.482070744811584e-06,
+      "loss": 0.1845,
+      "step": 1019
+    },
+    {
+      "epoch": 0.46527540198426276,
+      "grad_norm": 0.3001246154308319,
+      "learning_rate": 4.48094448180859e-06,
+      "loss": 0.1726,
+      "step": 1020
+    },
+    {
+      "epoch": 0.46573155433914926,
+      "grad_norm": 0.30914705991744995,
+      "learning_rate": 4.479817137407706e-06,
+      "loss": 0.1832,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4661877066940358,
+      "grad_norm": 0.31331297755241394,
+      "learning_rate": 4.478688712224348e-06,
+      "loss": 0.1868,
+      "step": 1022
+    },
+    {
+      "epoch": 0.46664385904892236,
+      "grad_norm": 0.3110509514808655,
+      "learning_rate": 4.477559206874525e-06,
+      "loss": 0.1847,
+      "step": 1023
+    },
+    {
+      "epoch": 0.46710001140380886,
+      "grad_norm": 0.3159843385219574,
+      "learning_rate": 4.476428621974833e-06,
+      "loss": 0.1823,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4675561637586954,
+      "grad_norm": 0.338840126991272,
+      "learning_rate": 4.47529695814246e-06,
+      "loss": 0.1764,
+      "step": 1025
+    },
+    {
+      "epoch": 0.46801231611358196,
+      "grad_norm": 0.3200490176677704,
+      "learning_rate": 4.47416421599518e-06,
+      "loss": 0.1811,
+      "step": 1026
+    },
+    {
+      "epoch": 0.46846846846846846,
+      "grad_norm": 0.3208402693271637,
+      "learning_rate": 4.473030396151358e-06,
+      "loss": 0.1763,
+      "step": 1027
+    },
+    {
+      "epoch": 0.468924620823355,
+      "grad_norm": 0.3356301784515381,
+      "learning_rate": 4.471895499229946e-06,
+      "loss": 0.1878,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4693807731782415,
+      "grad_norm": 0.3069671392440796,
+      "learning_rate": 4.470759525850484e-06,
+      "loss": 0.1819,
+      "step": 1029
+    },
+    {
+      "epoch": 0.46983692553312806,
+      "grad_norm": 0.3169107139110565,
+      "learning_rate": 4.4696224766331005e-06,
+      "loss": 0.1795,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4702930778880146,
+      "grad_norm": 0.32077765464782715,
+      "learning_rate": 4.468484352198511e-06,
+      "loss": 0.186,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4707492302429011,
+      "grad_norm": 0.323901891708374,
+      "learning_rate": 4.467345153168018e-06,
+      "loss": 0.1905,
+      "step": 1032
+    },
+    {
+      "epoch": 0.47120538259778766,
+      "grad_norm": 0.3261025547981262,
+      "learning_rate": 4.466204880163509e-06,
+      "loss": 0.1826,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4716615349526742,
+      "grad_norm": 0.3143298327922821,
+      "learning_rate": 4.465063533807462e-06,
+      "loss": 0.1875,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4721176873075607,
+      "grad_norm": 0.3154652714729309,
+      "learning_rate": 4.463921114722935e-06,
+      "loss": 0.1835,
+      "step": 1035
+    },
+    {
+      "epoch": 0.47257383966244726,
+      "grad_norm": 0.31739693880081177,
+      "learning_rate": 4.462777623533577e-06,
+      "loss": 0.1882,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4730299920173338,
+      "grad_norm": 0.308622270822525,
+      "learning_rate": 4.4616330608636185e-06,
+      "loss": 0.1863,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4734861443722203,
+      "grad_norm": 0.3299105763435364,
+      "learning_rate": 4.460487427337877e-06,
+      "loss": 0.1836,
+      "step": 1038
+    },
+    {
+      "epoch": 0.47394229672710686,
+      "grad_norm": 0.32627055048942566,
+      "learning_rate": 4.459340723581755e-06,
+      "loss": 0.1777,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4743984490819934,
+      "grad_norm": 0.3012368381023407,
+      "learning_rate": 4.458192950221237e-06,
+      "loss": 0.1724,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4748546014368799,
+      "grad_norm": 0.3031756579875946,
+      "learning_rate": 4.457044107882893e-06,
+      "loss": 0.1816,
+      "step": 1041
+    },
+    {
+      "epoch": 0.47531075379176646,
+      "grad_norm": 0.3183117210865021,
+      "learning_rate": 4.4558941971938765e-06,
+      "loss": 0.176,
+      "step": 1042
+    },
+    {
+      "epoch": 0.47576690614665296,
+      "grad_norm": 0.31405413150787354,
+      "learning_rate": 4.454743218781924e-06,
+      "loss": 0.1775,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4762230585015395,
+      "grad_norm": 0.3318958580493927,
+      "learning_rate": 4.4535911732753535e-06,
+      "loss": 0.1892,
+      "step": 1044
+    },
+    {
+      "epoch": 0.47667921085642606,
+      "grad_norm": 0.30204522609710693,
+      "learning_rate": 4.452438061303068e-06,
+      "loss": 0.1775,
+      "step": 1045
+    },
+    {
+      "epoch": 0.47713536321131256,
+      "grad_norm": 0.32054299116134644,
+      "learning_rate": 4.451283883494551e-06,
+      "loss": 0.1772,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4775915155661991,
+      "grad_norm": 0.32408997416496277,
+      "learning_rate": 4.450128640479869e-06,
+      "loss": 0.183,
+      "step": 1047
+    },
+    {
+      "epoch": 0.47804766792108566,
+      "grad_norm": 0.30600759387016296,
+      "learning_rate": 4.448972332889669e-06,
+      "loss": 0.1828,
+      "step": 1048
+    },
+    {
+      "epoch": 0.47850382027597216,
+      "grad_norm": 0.32001373171806335,
+      "learning_rate": 4.447814961355179e-06,
+      "loss": 0.1885,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4789599726308587,
+      "grad_norm": 0.3291076719760895,
+      "learning_rate": 4.44665652650821e-06,
+      "loss": 0.1704,
+      "step": 1050
+    },
+    {
+      "epoch": 0.47941612498574526,
+      "grad_norm": 0.32346177101135254,
+      "learning_rate": 4.445497028981149e-06,
+      "loss": 0.1775,
+      "step": 1051
+    },
+    {
+      "epoch": 0.47987227734063176,
+      "grad_norm": 0.3201315402984619,
+      "learning_rate": 4.444336469406968e-06,
+      "loss": 0.1925,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4803284296955183,
+      "grad_norm": 0.3116174638271332,
+      "learning_rate": 4.443174848419216e-06,
+      "loss": 0.1836,
+      "step": 1053
+    },
+    {
+      "epoch": 0.48078458205040486,
+      "grad_norm": 0.3288903534412384,
+      "learning_rate": 4.442012166652022e-06,
+      "loss": 0.1769,
+      "step": 1054
+    },
+    {
+      "epoch": 0.48124073440529136,
+      "grad_norm": 0.3382035791873932,
+      "learning_rate": 4.440848424740094e-06,
+      "loss": 0.183,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4816968867601779,
+      "grad_norm": 0.3126995265483856,
+      "learning_rate": 4.4396836233187195e-06,
+      "loss": 0.1709,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4821530391150644,
+      "grad_norm": 0.32766982913017273,
+      "learning_rate": 4.438517763023763e-06,
+      "loss": 0.1755,
+      "step": 1057
+    },
+    {
+      "epoch": 0.48260919146995096,
+      "grad_norm": 0.32621482014656067,
+      "learning_rate": 4.437350844491668e-06,
+      "loss": 0.1824,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4830653438248375,
+      "grad_norm": 0.3366355001926422,
+      "learning_rate": 4.436182868359456e-06,
+      "loss": 0.1875,
+      "step": 1059
+    },
+    {
+      "epoch": 0.483521496179724,
+      "grad_norm": 0.3044385612010956,
+      "learning_rate": 4.435013835264725e-06,
+      "loss": 0.1648,
+      "step": 1060
+    },
+    {
+      "epoch": 0.48397764853461056,
+      "grad_norm": 0.31879478693008423,
+      "learning_rate": 4.43384374584565e-06,
+      "loss": 0.1787,
+      "step": 1061
+    },
+    {
+      "epoch": 0.4844338008894971,
+      "grad_norm": 0.33908283710479736,
+      "learning_rate": 4.432672600740984e-06,
+      "loss": 0.1968,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4848899532443836,
+      "grad_norm": 0.3292134702205658,
+      "learning_rate": 4.431500400590054e-06,
+      "loss": 0.1829,
+      "step": 1063
+    },
+    {
+      "epoch": 0.48534610559927016,
+      "grad_norm": 0.3177955150604248,
+      "learning_rate": 4.4303271460327655e-06,
+      "loss": 0.1801,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4858022579541567,
+      "grad_norm": 0.29104816913604736,
+      "learning_rate": 4.429152837709597e-06,
+      "loss": 0.1787,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4862584103090432,
+      "grad_norm": 0.30899256467819214,
+      "learning_rate": 4.427977476261606e-06,
+      "loss": 0.1821,
+      "step": 1066
+    },
+    {
+      "epoch": 0.48671456266392976,
+      "grad_norm": 0.31402498483657837,
+      "learning_rate": 4.4268010623304195e-06,
+      "loss": 0.1767,
+      "step": 1067
+    },
+    {
+      "epoch": 0.48717071501881626,
+      "grad_norm": 0.30521360039711,
+      "learning_rate": 4.425623596558243e-06,
+      "loss": 0.1829,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4876268673737028,
+      "grad_norm": 0.3126015365123749,
+      "learning_rate": 4.424445079587855e-06,
+      "loss": 0.1825,
+      "step": 1069
+    },
+    {
+      "epoch": 0.48808301972858936,
+      "grad_norm": 0.31246834993362427,
+      "learning_rate": 4.423265512062611e-06,
+      "loss": 0.1765,
+      "step": 1070
+    },
+    {
+      "epoch": 0.48853917208347586,
+      "grad_norm": 0.32182776927948,
+      "learning_rate": 4.422084894626434e-06,
+      "loss": 0.1877,
+      "step": 1071
+    },
+    {
+      "epoch": 0.4889953244383624,
+      "grad_norm": 0.3359273374080658,
+      "learning_rate": 4.420903227923823e-06,
+      "loss": 0.1939,
+      "step": 1072
+    },
+    {
+      "epoch": 0.48945147679324896,
+      "grad_norm": 0.2938372790813446,
+      "learning_rate": 4.419720512599852e-06,
+      "loss": 0.1675,
+      "step": 1073
+    },
+    {
+      "epoch": 0.48990762914813546,
+      "grad_norm": 0.30865082144737244,
+      "learning_rate": 4.418536749300164e-06,
+      "loss": 0.1791,
+      "step": 1074
+    },
+    {
+      "epoch": 0.490363781503022,
+      "grad_norm": 0.3115358352661133,
+      "learning_rate": 4.4173519386709765e-06,
+      "loss": 0.1746,
+      "step": 1075
+    },
+    {
+      "epoch": 0.49081993385790856,
+      "grad_norm": 0.33995020389556885,
+      "learning_rate": 4.416166081359077e-06,
+      "loss": 0.1876,
+      "step": 1076
+    },
+    {
+      "epoch": 0.49127608621279506,
+      "grad_norm": 0.310163676738739,
+      "learning_rate": 4.414979178011825e-06,
+      "loss": 0.1787,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4917322385676816,
+      "grad_norm": 0.3275308609008789,
+      "learning_rate": 4.413791229277152e-06,
+      "loss": 0.1787,
+      "step": 1078
+    },
+    {
+      "epoch": 0.49218839092256816,
+      "grad_norm": 0.3173719048500061,
+      "learning_rate": 4.41260223580356e-06,
+      "loss": 0.1805,
+      "step": 1079
+    },
+    {
+      "epoch": 0.49264454327745466,
+      "grad_norm": 0.35063788294792175,
+      "learning_rate": 4.411412198240119e-06,
+      "loss": 0.1904,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4931006956323412,
+      "grad_norm": 0.32667285203933716,
+      "learning_rate": 4.410221117236473e-06,
+      "loss": 0.195,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4935568479872277,
+      "grad_norm": 0.32641226053237915,
+      "learning_rate": 4.409028993442831e-06,
+      "loss": 0.1821,
+      "step": 1082
+    },
+    {
+      "epoch": 0.49401300034211426,
+      "grad_norm": 0.31927937269210815,
+      "learning_rate": 4.407835827509974e-06,
+      "loss": 0.1778,
+      "step": 1083
+    },
+    {
+      "epoch": 0.4944691526970008,
+      "grad_norm": 0.32031726837158203,
+      "learning_rate": 4.406641620089252e-06,
+      "loss": 0.1896,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4949253050518873,
+      "grad_norm": 0.3153742253780365,
+      "learning_rate": 4.405446371832583e-06,
+      "loss": 0.1924,
+      "step": 1085
+    },
+    {
+      "epoch": 0.49538145740677386,
+      "grad_norm": 0.311893492937088,
+      "learning_rate": 4.404250083392454e-06,
+      "loss": 0.1728,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4958376097616604,
+      "grad_norm": 0.31848254799842834,
+      "learning_rate": 4.403052755421917e-06,
+      "loss": 0.1865,
+      "step": 1087
+    },
+    {
+      "epoch": 0.4962937621165469,
+      "grad_norm": 0.31412747502326965,
+      "learning_rate": 4.401854388574595e-06,
+      "loss": 0.1792,
+      "step": 1088
+    },
+    {
+      "epoch": 0.49674991447143346,
+      "grad_norm": 0.31871849298477173,
+      "learning_rate": 4.4006549835046766e-06,
+      "loss": 0.1864,
+      "step": 1089
+    },
+    {
+      "epoch": 0.49720606682632,
+      "grad_norm": 0.30212581157684326,
+      "learning_rate": 4.399454540866917e-06,
+      "loss": 0.1806,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4976622191812065,
+      "grad_norm": 0.3130529820919037,
+      "learning_rate": 4.398253061316638e-06,
+      "loss": 0.1768,
+      "step": 1091
+    },
+    {
+      "epoch": 0.49811837153609306,
+      "grad_norm": 0.3267703652381897,
+      "learning_rate": 4.397050545509726e-06,
+      "loss": 0.1841,
+      "step": 1092
+    },
+    {
+      "epoch": 0.49857452389097956,
+      "grad_norm": 0.31406691670417786,
+      "learning_rate": 4.395846994102637e-06,
+      "loss": 0.1813,
+      "step": 1093
+    },
+    {
+      "epoch": 0.4990306762458661,
+      "grad_norm": 0.31510791182518005,
+      "learning_rate": 4.394642407752389e-06,
+      "loss": 0.1755,
+      "step": 1094
+    },
+    {
+      "epoch": 0.49948682860075266,
+      "grad_norm": 0.30812543630599976,
+      "learning_rate": 4.393436787116565e-06,
+      "loss": 0.1807,
+      "step": 1095
+    },
+    {
+      "epoch": 0.49994298095563916,
+      "grad_norm": 0.3103227913379669,
+      "learning_rate": 4.392230132853316e-06,
+      "loss": 0.1881,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5003991333105258,
+      "grad_norm": 0.3123283386230469,
+      "learning_rate": 4.391022445621351e-06,
+      "loss": 0.1781,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5008552856654123,
+      "grad_norm": 0.3193586766719818,
+      "learning_rate": 4.38981372607995e-06,
+      "loss": 0.1827,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5013114380202988,
+      "grad_norm": 0.32207897305488586,
+      "learning_rate": 4.388603974888951e-06,
+      "loss": 0.1885,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5017675903751853,
+      "grad_norm": 0.3117227554321289,
+      "learning_rate": 4.387393192708758e-06,
+      "loss": 0.1812,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5022237427300719,
+      "grad_norm": 0.33583545684814453,
+      "learning_rate": 4.386181380200337e-06,
+      "loss": 0.1975,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5026798950849584,
+      "grad_norm": 0.3330482542514801,
+      "learning_rate": 4.384968538025217e-06,
+      "loss": 0.1902,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5031360474398449,
+      "grad_norm": 0.3201743960380554,
+      "learning_rate": 4.38375466684549e-06,
+      "loss": 0.1767,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5035921997947315,
+      "grad_norm": 0.3165605962276459,
+      "learning_rate": 4.382539767323805e-06,
+      "loss": 0.1776,
+      "step": 1104
+    },
+    {
+      "epoch": 0.504048352149618,
+      "grad_norm": 0.319797158241272,
+      "learning_rate": 4.3813238401233805e-06,
+      "loss": 0.1822,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5045045045045045,
+      "grad_norm": 0.31622326374053955,
+      "learning_rate": 4.380106885907987e-06,
+      "loss": 0.178,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5049606568593911,
+      "grad_norm": 0.34379029273986816,
+      "learning_rate": 4.378888905341963e-06,
+      "loss": 0.1775,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5054168092142776,
+      "grad_norm": 0.3279973566532135,
+      "learning_rate": 4.377669899090202e-06,
+      "loss": 0.1918,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5058729615691641,
+      "grad_norm": 0.30299675464630127,
+      "learning_rate": 4.376449867818161e-06,
+      "loss": 0.1795,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.31859180331230164,
+      "learning_rate": 4.375228812191855e-06,
+      "loss": 0.1744,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5067852662789372,
+      "grad_norm": 0.29021868109703064,
+      "learning_rate": 4.374006732877858e-06,
+      "loss": 0.1729,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5072414186338237,
+      "grad_norm": 0.32044151425361633,
+      "learning_rate": 4.372783630543305e-06,
+      "loss": 0.1806,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5076975709887103,
+      "grad_norm": 0.302049458026886,
+      "learning_rate": 4.371559505855886e-06,
+      "loss": 0.1734,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5081537233435968,
+      "grad_norm": 0.3195923864841461,
+      "learning_rate": 4.370334359483851e-06,
+      "loss": 0.1724,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5086098756984833,
+      "grad_norm": 0.3208305537700653,
+      "learning_rate": 4.369108192096007e-06,
+      "loss": 0.186,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5090660280533699,
+      "grad_norm": 0.29478055238723755,
+      "learning_rate": 4.3678810043617215e-06,
+      "loss": 0.174,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5095221804082564,
+      "grad_norm": 0.317287802696228,
+      "learning_rate": 4.366652796950914e-06,
+      "loss": 0.1926,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5099783327631429,
+      "grad_norm": 0.31819984316825867,
+      "learning_rate": 4.365423570534064e-06,
+      "loss": 0.182,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5104344851180295,
+      "grad_norm": 0.32738587260246277,
+      "learning_rate": 4.364193325782207e-06,
+      "loss": 0.1847,
+      "step": 1119
+    },
+    {
+      "epoch": 0.510890637472916,
+      "grad_norm": 0.29867494106292725,
+      "learning_rate": 4.362962063366933e-06,
+      "loss": 0.1688,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5113467898278025,
+      "grad_norm": 0.32660600543022156,
+      "learning_rate": 4.361729783960388e-06,
+      "loss": 0.1796,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5118029421826891,
+      "grad_norm": 0.3067159354686737,
+      "learning_rate": 4.360496488235274e-06,
+      "loss": 0.1763,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5122590945375756,
+      "grad_norm": 0.31153634190559387,
+      "learning_rate": 4.359262176864848e-06,
+      "loss": 0.1791,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5127152468924621,
+      "grad_norm": 0.30642974376678467,
+      "learning_rate": 4.358026850522919e-06,
+      "loss": 0.1709,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5131713992473487,
+      "grad_norm": 0.29793867468833923,
+      "learning_rate": 4.356790509883854e-06,
+      "loss": 0.1758,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5136275516022352,
+      "grad_norm": 0.29833337664604187,
+      "learning_rate": 4.35555315562257e-06,
+      "loss": 0.166,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5140837039571217,
+      "grad_norm": 0.31274035573005676,
+      "learning_rate": 4.3543147884145394e-06,
+      "loss": 0.1721,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5145398563120082,
+      "grad_norm": 0.3146253526210785,
+      "learning_rate": 4.353075408935787e-06,
+      "loss": 0.177,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5149960086668948,
+      "grad_norm": 0.3214576244354248,
+      "learning_rate": 4.35183501786289e-06,
+      "loss": 0.1851,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5154521610217813,
+      "grad_norm": 0.3149864673614502,
+      "learning_rate": 4.350593615872978e-06,
+      "loss": 0.1785,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5159083133766678,
+      "grad_norm": 0.30646955966949463,
+      "learning_rate": 4.3493512036437335e-06,
+      "loss": 0.1769,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5163644657315544,
+      "grad_norm": 0.3173981010913849,
+      "learning_rate": 4.348107781853389e-06,
+      "loss": 0.1802,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5168206180864409,
+      "grad_norm": 0.3099330961704254,
+      "learning_rate": 4.346863351180728e-06,
+      "loss": 0.1851,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5172767704413274,
+      "grad_norm": 0.30647557973861694,
+      "learning_rate": 4.345617912305086e-06,
+      "loss": 0.1749,
+      "step": 1134
+    },
+    {
+      "epoch": 0.517732922796214,
+      "grad_norm": 0.31544044613838196,
+      "learning_rate": 4.3443714659063486e-06,
+      "loss": 0.1834,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5181890751511005,
+      "grad_norm": 0.3377651274204254,
+      "learning_rate": 4.34312401266495e-06,
+      "loss": 0.1848,
+      "step": 1136
+    },
+    {
+      "epoch": 0.518645227505987,
+      "grad_norm": 0.30423247814178467,
+      "learning_rate": 4.341875553261876e-06,
+      "loss": 0.1819,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5191013798608736,
+      "grad_norm": 0.3267976641654968,
+      "learning_rate": 4.34062608837866e-06,
+      "loss": 0.1791,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5195575322157601,
+      "grad_norm": 0.3072841465473175,
+      "learning_rate": 4.3393756186973866e-06,
+      "loss": 0.1757,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5200136845706466,
+      "grad_norm": 0.32858094573020935,
+      "learning_rate": 4.338124144900685e-06,
+      "loss": 0.1885,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5204698369255332,
+      "grad_norm": 0.31543034315109253,
+      "learning_rate": 4.336871667671738e-06,
+      "loss": 0.1666,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5209259892804197,
+      "grad_norm": 0.31898143887519836,
+      "learning_rate": 4.33561818769427e-06,
+      "loss": 0.1836,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5213821416353062,
+      "grad_norm": 0.299163818359375,
+      "learning_rate": 4.334363705652559e-06,
+      "loss": 0.1793,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5218382939901928,
+      "grad_norm": 0.3210349678993225,
+      "learning_rate": 4.333108222231423e-06,
+      "loss": 0.17,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5222944463450793,
+      "grad_norm": 0.28622928261756897,
+      "learning_rate": 4.331851738116235e-06,
+      "loss": 0.1711,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5227505986999658,
+      "grad_norm": 0.30328530073165894,
+      "learning_rate": 4.3305942539929055e-06,
+      "loss": 0.1844,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5232067510548524,
+      "grad_norm": 0.3457143008708954,
+      "learning_rate": 4.3293357705479e-06,
+      "loss": 0.1874,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5236629034097389,
+      "grad_norm": 0.29161518812179565,
+      "learning_rate": 4.32807628846822e-06,
+      "loss": 0.1749,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5241190557646254,
+      "grad_norm": 0.30757150053977966,
+      "learning_rate": 4.326815808441421e-06,
+      "loss": 0.1695,
+      "step": 1149
+    },
+    {
+      "epoch": 0.524575208119512,
+      "grad_norm": 0.32009485363960266,
+      "learning_rate": 4.325554331155597e-06,
+      "loss": 0.1823,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5250313604743985,
+      "grad_norm": 0.30738401412963867,
+      "learning_rate": 4.324291857299389e-06,
+      "loss": 0.1716,
+      "step": 1151
+    },
+    {
+      "epoch": 0.525487512829285,
+      "grad_norm": 0.31683769822120667,
+      "learning_rate": 4.3230283875619815e-06,
+      "loss": 0.1752,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5259436651841715,
+      "grad_norm": 0.30964645743370056,
+      "learning_rate": 4.321763922633103e-06,
+      "loss": 0.1774,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5263998175390581,
+      "grad_norm": 0.3028723895549774,
+      "learning_rate": 4.320498463203026e-06,
+      "loss": 0.1776,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5268559698939446,
+      "grad_norm": 0.30387482047080994,
+      "learning_rate": 4.319232009962563e-06,
+      "loss": 0.1767,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5273121222488311,
+      "grad_norm": 0.2991229295730591,
+      "learning_rate": 4.317964563603073e-06,
+      "loss": 0.1695,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5277682746037177,
+      "grad_norm": 0.3026543855667114,
+      "learning_rate": 4.316696124816454e-06,
+      "loss": 0.1731,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5282244269586042,
+      "grad_norm": 0.30617398023605347,
+      "learning_rate": 4.315426694295148e-06,
+      "loss": 0.177,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5286805793134907,
+      "grad_norm": 0.3052368760108948,
+      "learning_rate": 4.314156272732136e-06,
+      "loss": 0.1844,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5291367316683773,
+      "grad_norm": 0.31815728545188904,
+      "learning_rate": 4.312884860820942e-06,
+      "loss": 0.1793,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5295928840232638,
+      "grad_norm": 0.3251146078109741,
+      "learning_rate": 4.31161245925563e-06,
+      "loss": 0.1889,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5300490363781503,
+      "grad_norm": 0.33524826169013977,
+      "learning_rate": 4.310339068730805e-06,
+      "loss": 0.1886,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5305051887330369,
+      "grad_norm": 0.3078482449054718,
+      "learning_rate": 4.3090646899416085e-06,
+      "loss": 0.1831,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5309613410879234,
+      "grad_norm": 0.31183791160583496,
+      "learning_rate": 4.307789323583727e-06,
+      "loss": 0.1841,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5314174934428099,
+      "grad_norm": 0.3018867075443268,
+      "learning_rate": 4.306512970353382e-06,
+      "loss": 0.1718,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5318736457976965,
+      "grad_norm": 0.31354305148124695,
+      "learning_rate": 4.305235630947333e-06,
+      "loss": 0.1852,
+      "step": 1166
+    },
+    {
+      "epoch": 0.532329798152583,
+      "grad_norm": 0.31621262431144714,
+      "learning_rate": 4.3039573060628834e-06,
+      "loss": 0.1782,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5327859505074695,
+      "grad_norm": 0.3158392906188965,
+      "learning_rate": 4.302677996397868e-06,
+      "loss": 0.177,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5332421028623561,
+      "grad_norm": 0.32750409841537476,
+      "learning_rate": 4.301397702650665e-06,
+      "loss": 0.1735,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5336982552172426,
+      "grad_norm": 0.3193983733654022,
+      "learning_rate": 4.3001164255201836e-06,
+      "loss": 0.1832,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5341544075721291,
+      "grad_norm": 0.35041525959968567,
+      "learning_rate": 4.298834165705876e-06,
+      "loss": 0.1957,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5346105599270157,
+      "grad_norm": 0.32777348160743713,
+      "learning_rate": 4.297550923907726e-06,
+      "loss": 0.1869,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5350667122819022,
+      "grad_norm": 0.3178562819957733,
+      "learning_rate": 4.296266700826257e-06,
+      "loss": 0.1884,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5355228646367887,
+      "grad_norm": 0.31194210052490234,
+      "learning_rate": 4.294981497162525e-06,
+      "loss": 0.1767,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5359790169916753,
+      "grad_norm": 0.31972649693489075,
+      "learning_rate": 4.293695313618125e-06,
+      "loss": 0.1783,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5364351693465618,
+      "grad_norm": 0.31819331645965576,
+      "learning_rate": 4.2924081508951824e-06,
+      "loss": 0.1787,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5368913217014483,
+      "grad_norm": 0.3497715890407562,
+      "learning_rate": 4.291120009696361e-06,
+      "loss": 0.1894,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5373474740563348,
+      "grad_norm": 0.31592386960983276,
+      "learning_rate": 4.289830890724858e-06,
+      "loss": 0.1878,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5378036264112214,
+      "grad_norm": 0.30878397822380066,
+      "learning_rate": 4.288540794684402e-06,
+      "loss": 0.1728,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5382597787661079,
+      "grad_norm": 0.3047281503677368,
+      "learning_rate": 4.287249722279257e-06,
+      "loss": 0.1821,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5387159311209944,
+      "grad_norm": 0.29890167713165283,
+      "learning_rate": 4.28595767421422e-06,
+      "loss": 0.1776,
+      "step": 1181
+    },
+    {
+      "epoch": 0.539172083475881,
+      "grad_norm": 0.31495875120162964,
+      "learning_rate": 4.284664651194621e-06,
+      "loss": 0.1835,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5396282358307675,
+      "grad_norm": 0.32684817910194397,
+      "learning_rate": 4.283370653926319e-06,
+      "loss": 0.186,
+      "step": 1183
+    },
+    {
+      "epoch": 0.540084388185654,
+      "grad_norm": 0.3131972551345825,
+      "learning_rate": 4.28207568311571e-06,
+      "loss": 0.1701,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 0.3264712691307068,
+      "learning_rate": 4.280779739469717e-06,
+      "loss": 0.186,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5409966928954271,
+      "grad_norm": 0.3139888346195221,
+      "learning_rate": 4.279482823695797e-06,
+      "loss": 0.1762,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5414528452503136,
+      "grad_norm": 0.3101795017719269,
+      "learning_rate": 4.278184936501936e-06,
+      "loss": 0.1774,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5419089976052002,
+      "grad_norm": 0.3101790249347687,
+      "learning_rate": 4.27688607859665e-06,
+      "loss": 0.1734,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5423651499600867,
+      "grad_norm": 0.3040241599082947,
+      "learning_rate": 4.275586250688987e-06,
+      "loss": 0.1784,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5428213023149732,
+      "grad_norm": 0.3099677562713623,
+      "learning_rate": 4.274285453488522e-06,
+      "loss": 0.1802,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5432774546698598,
+      "grad_norm": 0.32185667753219604,
+      "learning_rate": 4.272983687705362e-06,
+      "loss": 0.1799,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5437336070247463,
+      "grad_norm": 0.31333786249160767,
+      "learning_rate": 4.27168095405014e-06,
+      "loss": 0.1912,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5441897593796328,
+      "grad_norm": 0.3191187381744385,
+      "learning_rate": 4.270377253234019e-06,
+      "loss": 0.1834,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5446459117345194,
+      "grad_norm": 0.3310491740703583,
+      "learning_rate": 4.26907258596869e-06,
+      "loss": 0.1878,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5451020640894059,
+      "grad_norm": 0.30478915572166443,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.1675,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5455582164442924,
+      "grad_norm": 0.3024504780769348,
+      "learning_rate": 4.266460354939803e-06,
+      "loss": 0.1824,
+      "step": 1196
+    },
+    {
+      "epoch": 0.546014368799179,
+      "grad_norm": 0.3188779950141907,
+      "learning_rate": 4.265152792602264e-06,
+      "loss": 0.1769,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5464705211540655,
+      "grad_norm": 0.3264593482017517,
+      "learning_rate": 4.263844266667549e-06,
+      "loss": 0.1857,
+      "step": 1198
+    },
+    {
+      "epoch": 0.546926673508952,
+      "grad_norm": 0.30892127752304077,
+      "learning_rate": 4.2625347778499845e-06,
+      "loss": 0.1904,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5473828258638386,
+      "grad_norm": 0.3225180208683014,
+      "learning_rate": 4.26122432686442e-06,
+      "loss": 0.1801,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5478389782187251,
+      "grad_norm": 0.32953789830207825,
+      "learning_rate": 4.259912914426231e-06,
+      "loss": 0.1802,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5482951305736116,
+      "grad_norm": 0.34198135137557983,
+      "learning_rate": 4.258600541251317e-06,
+      "loss": 0.1833,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5487512829284981,
+      "grad_norm": 0.6198412775993347,
+      "learning_rate": 4.2572872080561024e-06,
+      "loss": 0.1772,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5492074352833847,
+      "grad_norm": 0.33181968331336975,
+      "learning_rate": 4.255972915557537e-06,
+      "loss": 0.1869,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5496635876382712,
+      "grad_norm": 0.3115808665752411,
+      "learning_rate": 4.254657664473093e-06,
+      "loss": 0.1766,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5501197399931577,
+      "grad_norm": 0.3069736361503601,
+      "learning_rate": 4.253341455520767e-06,
+      "loss": 0.1814,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5505758923480443,
+      "grad_norm": 0.30829694867134094,
+      "learning_rate": 4.252024289419076e-06,
+      "loss": 0.1706,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5510320447029308,
+      "grad_norm": 0.29807761311531067,
+      "learning_rate": 4.250706166887061e-06,
+      "loss": 0.1665,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5514881970578173,
+      "grad_norm": 0.30204683542251587,
+      "learning_rate": 4.249387088644287e-06,
+      "loss": 0.1642,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5519443494127039,
+      "grad_norm": 0.33678415417671204,
+      "learning_rate": 4.2480670554108386e-06,
+      "loss": 0.1919,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5524005017675904,
+      "grad_norm": 0.3355265259742737,
+      "learning_rate": 4.246746067907322e-06,
+      "loss": 0.2016,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5528566541224769,
+      "grad_norm": 0.31889843940734863,
+      "learning_rate": 4.245424126854864e-06,
+      "loss": 0.1922,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5533128064773635,
+      "grad_norm": 0.2929706871509552,
+      "learning_rate": 4.244101232975113e-06,
+      "loss": 0.1676,
+      "step": 1213
+    },
+    {
+      "epoch": 0.55376895883225,
+      "grad_norm": 0.33658766746520996,
+      "learning_rate": 4.242777386990236e-06,
+      "loss": 0.1824,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5542251111871365,
+      "grad_norm": 0.3305429220199585,
+      "learning_rate": 4.241452589622921e-06,
+      "loss": 0.1809,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5546812635420231,
+      "grad_norm": 0.3344593644142151,
+      "learning_rate": 4.240126841596377e-06,
+      "loss": 0.1884,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5551374158969096,
+      "grad_norm": 0.33831673860549927,
+      "learning_rate": 4.238800143634327e-06,
+      "loss": 0.178,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5555935682517961,
+      "grad_norm": 0.3098374307155609,
+      "learning_rate": 4.237472496461018e-06,
+      "loss": 0.1748,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5560497206066827,
+      "grad_norm": 0.32199814915657043,
+      "learning_rate": 4.236143900801211e-06,
+      "loss": 0.1819,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5565058729615692,
+      "grad_norm": 0.34061112999916077,
+      "learning_rate": 4.234814357380189e-06,
+      "loss": 0.1847,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5569620253164557,
+      "grad_norm": 0.3061227798461914,
+      "learning_rate": 4.233483866923749e-06,
+      "loss": 0.183,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5574181776713423,
+      "grad_norm": 0.30358588695526123,
+      "learning_rate": 4.232152430158203e-06,
+      "loss": 0.1765,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5578743300262288,
+      "grad_norm": 0.31991446018218994,
+      "learning_rate": 4.230820047810387e-06,
+      "loss": 0.1886,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5583304823811153,
+      "grad_norm": 0.30070960521698,
+      "learning_rate": 4.229486720607645e-06,
+      "loss": 0.1682,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5587866347360019,
+      "grad_norm": 0.30840572714805603,
+      "learning_rate": 4.228152449277842e-06,
+      "loss": 0.1823,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5592427870908884,
+      "grad_norm": 0.3182162940502167,
+      "learning_rate": 4.226817234549359e-06,
+      "loss": 0.1833,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5596989394457749,
+      "grad_norm": 0.30193209648132324,
+      "learning_rate": 4.2254810771510864e-06,
+      "loss": 0.1744,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5601550918006615,
+      "grad_norm": 0.33933258056640625,
+      "learning_rate": 4.224143977812435e-06,
+      "loss": 0.176,
+      "step": 1228
+    },
+    {
+      "epoch": 0.560611244155548,
+      "grad_norm": 0.31610801815986633,
+      "learning_rate": 4.222805937263327e-06,
+      "loss": 0.1721,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5610673965104345,
+      "grad_norm": 0.312422513961792,
+      "learning_rate": 4.221466956234198e-06,
+      "loss": 0.1727,
+      "step": 1230
+    },
+    {
+      "epoch": 0.561523548865321,
+      "grad_norm": 0.32197174429893494,
+      "learning_rate": 4.220127035456e-06,
+      "loss": 0.1768,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5619797012202076,
+      "grad_norm": 0.315396249294281,
+      "learning_rate": 4.218786175660194e-06,
+      "loss": 0.1802,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5624358535750941,
+      "grad_norm": 0.3244089186191559,
+      "learning_rate": 4.2174443775787564e-06,
+      "loss": 0.1735,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5628920059299806,
+      "grad_norm": 0.3139295279979706,
+      "learning_rate": 4.216101641944174e-06,
+      "loss": 0.1789,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5633481582848672,
+      "grad_norm": 0.34641653299331665,
+      "learning_rate": 4.214757969489448e-06,
+      "loss": 0.1816,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5638043106397537,
+      "grad_norm": 0.3126952648162842,
+      "learning_rate": 4.213413360948089e-06,
+      "loss": 0.1816,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5642604629946402,
+      "grad_norm": 0.3047172427177429,
+      "learning_rate": 4.212067817054118e-06,
+      "loss": 0.1816,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5647166153495268,
+      "grad_norm": 0.31668224930763245,
+      "learning_rate": 4.210721338542068e-06,
+      "loss": 0.1717,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5651727677044133,
+      "grad_norm": 0.3106701672077179,
+      "learning_rate": 4.209373926146983e-06,
+      "loss": 0.1748,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5656289200592998,
+      "grad_norm": 0.3245757520198822,
+      "learning_rate": 4.208025580604413e-06,
+      "loss": 0.1837,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5660850724141864,
+      "grad_norm": 0.3108798563480377,
+      "learning_rate": 4.206676302650424e-06,
+      "loss": 0.1812,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5665412247690729,
+      "grad_norm": 0.3091650903224945,
+      "learning_rate": 4.205326093021583e-06,
+      "loss": 0.1806,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5669973771239594,
+      "grad_norm": 0.3153209090232849,
+      "learning_rate": 4.203974952454972e-06,
+      "loss": 0.1775,
+      "step": 1243
+    },
+    {
+      "epoch": 0.567453529478846,
+      "grad_norm": 0.32562071084976196,
+      "learning_rate": 4.202622881688178e-06,
+      "loss": 0.1801,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5679096818337325,
+      "grad_norm": 0.3106227219104767,
+      "learning_rate": 4.201269881459298e-06,
+      "loss": 0.1668,
+      "step": 1245
+    },
+    {
+      "epoch": 0.568365834188619,
+      "grad_norm": 0.33634430170059204,
+      "learning_rate": 4.199915952506934e-06,
+      "loss": 0.1861,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5688219865435056,
+      "grad_norm": 0.31277039647102356,
+      "learning_rate": 4.198561095570194e-06,
+      "loss": 0.1871,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5692781388983921,
+      "grad_norm": 0.33361855149269104,
+      "learning_rate": 4.197205311388698e-06,
+      "loss": 0.1986,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5697342912532786,
+      "grad_norm": 0.32146215438842773,
+      "learning_rate": 4.195848600702565e-06,
+      "loss": 0.1848,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5701904436081652,
+      "grad_norm": 0.32315775752067566,
+      "learning_rate": 4.194490964252429e-06,
+      "loss": 0.1881,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5706465959630517,
+      "grad_norm": 0.2935425639152527,
+      "learning_rate": 4.1931324027794195e-06,
+      "loss": 0.173,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5711027483179382,
+      "grad_norm": 0.3100287616252899,
+      "learning_rate": 4.1917729170251765e-06,
+      "loss": 0.1708,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5715589006728248,
+      "grad_norm": 0.3283054828643799,
+      "learning_rate": 4.190412507731844e-06,
+      "loss": 0.1922,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5720150530277113,
+      "grad_norm": 0.3061600923538208,
+      "learning_rate": 4.1890511756420695e-06,
+      "loss": 0.1719,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5724712053825978,
+      "grad_norm": 0.3020835518836975,
+      "learning_rate": 4.187688921499004e-06,
+      "loss": 0.1751,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5729273577374843,
+      "grad_norm": 0.34978342056274414,
+      "learning_rate": 4.186325746046302e-06,
+      "loss": 0.1899,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5733835100923709,
+      "grad_norm": 0.31503593921661377,
+      "learning_rate": 4.184961650028123e-06,
+      "loss": 0.1767,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5738396624472574,
+      "grad_norm": 0.31431183218955994,
+      "learning_rate": 4.183596634189125e-06,
+      "loss": 0.1798,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5742958148021439,
+      "grad_norm": 0.31539273262023926,
+      "learning_rate": 4.182230699274471e-06,
+      "loss": 0.1793,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5747519671570305,
+      "grad_norm": 0.3235699236392975,
+      "learning_rate": 4.180863846029825e-06,
+      "loss": 0.1898,
+      "step": 1260
+    },
+    {
+      "epoch": 0.575208119511917,
+      "grad_norm": 0.3153180480003357,
+      "learning_rate": 4.179496075201354e-06,
+      "loss": 0.173,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5756642718668035,
+      "grad_norm": 0.3054327666759491,
+      "learning_rate": 4.178127387535722e-06,
+      "loss": 0.1737,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5761204242216901,
+      "grad_norm": 0.3344239592552185,
+      "learning_rate": 4.176757783780097e-06,
+      "loss": 0.1829,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5765765765765766,
+      "grad_norm": 0.31499776244163513,
+      "learning_rate": 4.175387264682146e-06,
+      "loss": 0.178,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5770327289314631,
+      "grad_norm": 0.32077229022979736,
+      "learning_rate": 4.174015830990036e-06,
+      "loss": 0.1824,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5774888812863497,
+      "grad_norm": 0.30602914094924927,
+      "learning_rate": 4.172643483452431e-06,
+      "loss": 0.1789,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5779450336412362,
+      "grad_norm": 0.30904388427734375,
+      "learning_rate": 4.171270222818497e-06,
+      "loss": 0.1726,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5784011859961227,
+      "grad_norm": 0.3176083266735077,
+      "learning_rate": 4.169896049837899e-06,
+      "loss": 0.1814,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5788573383510093,
+      "grad_norm": 0.3289932310581207,
+      "learning_rate": 4.1685209652607935e-06,
+      "loss": 0.191,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5793134907058958,
+      "grad_norm": 0.33303940296173096,
+      "learning_rate": 4.167144969837843e-06,
+      "loss": 0.1718,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5797696430607823,
+      "grad_norm": 0.30857717990875244,
+      "learning_rate": 4.165768064320203e-06,
+      "loss": 0.1752,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5802257954156689,
+      "grad_norm": 0.31593722105026245,
+      "learning_rate": 4.164390249459526e-06,
+      "loss": 0.1808,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5806819477705554,
+      "grad_norm": 0.3228105306625366,
+      "learning_rate": 4.163011526007963e-06,
+      "loss": 0.1848,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5811381001254419,
+      "grad_norm": 0.3089195191860199,
+      "learning_rate": 4.161631894718157e-06,
+      "loss": 0.163,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5815942524803285,
+      "grad_norm": 0.3149510324001312,
+      "learning_rate": 4.160251356343249e-06,
+      "loss": 0.1875,
+      "step": 1275
+    },
+    {
+      "epoch": 0.582050404835215,
+      "grad_norm": 0.3298307955265045,
+      "learning_rate": 4.158869911636876e-06,
+      "loss": 0.1888,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5825065571901015,
+      "grad_norm": 0.2937168478965759,
+      "learning_rate": 4.157487561353169e-06,
+      "loss": 0.1761,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5829627095449881,
+      "grad_norm": 0.31423869729042053,
+      "learning_rate": 4.156104306246754e-06,
+      "loss": 0.181,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5834188618998746,
+      "grad_norm": 0.31659549474716187,
+      "learning_rate": 4.154720147072748e-06,
+      "loss": 0.1725,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5838750142547611,
+      "grad_norm": 0.3194673955440521,
+      "learning_rate": 4.153335084586766e-06,
+      "loss": 0.1719,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5843311666096476,
+      "grad_norm": 0.33193960785865784,
+      "learning_rate": 4.1519491195449115e-06,
+      "loss": 0.1808,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5847873189645342,
+      "grad_norm": 0.31641775369644165,
+      "learning_rate": 4.150562252703786e-06,
+      "loss": 0.171,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5852434713194207,
+      "grad_norm": 0.3171290159225464,
+      "learning_rate": 4.149174484820477e-06,
+      "loss": 0.1779,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5856996236743072,
+      "grad_norm": 0.2875344157218933,
+      "learning_rate": 4.147785816652569e-06,
+      "loss": 0.1704,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5861557760291938,
+      "grad_norm": 0.32229799032211304,
+      "learning_rate": 4.146396248958136e-06,
+      "loss": 0.1727,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5866119283840803,
+      "grad_norm": 0.32215970754623413,
+      "learning_rate": 4.145005782495743e-06,
+      "loss": 0.1805,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5870680807389668,
+      "grad_norm": 0.32583972811698914,
+      "learning_rate": 4.143614418024446e-06,
+      "loss": 0.1856,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5875242330938534,
+      "grad_norm": 0.31126630306243896,
+      "learning_rate": 4.142222156303792e-06,
+      "loss": 0.1708,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5879803854487399,
+      "grad_norm": 0.3259385824203491,
+      "learning_rate": 4.1408289980938154e-06,
+      "loss": 0.1734,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5884365378036264,
+      "grad_norm": 0.28777754306793213,
+      "learning_rate": 4.139434944155042e-06,
+      "loss": 0.1649,
+      "step": 1290
+    },
+    {
+      "epoch": 0.588892690158513,
+      "grad_norm": 0.3149665892124176,
+      "learning_rate": 4.138039995248486e-06,
+      "loss": 0.1736,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5893488425133995,
+      "grad_norm": 0.3060600459575653,
+      "learning_rate": 4.13664415213565e-06,
+      "loss": 0.1682,
+      "step": 1292
+    },
+    {
+      "epoch": 0.589804994868286,
+      "grad_norm": 0.3319993019104004,
+      "learning_rate": 4.1352474155785264e-06,
+      "loss": 0.1805,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5902611472231726,
+      "grad_norm": 0.31701231002807617,
+      "learning_rate": 4.133849786339591e-06,
+      "loss": 0.1875,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5907172995780591,
+      "grad_norm": 0.29577770829200745,
+      "learning_rate": 4.132451265181813e-06,
+      "loss": 0.1767,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5911734519329456,
+      "grad_norm": 0.3056115508079529,
+      "learning_rate": 4.131051852868643e-06,
+      "loss": 0.1836,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5916296042878322,
+      "grad_norm": 0.309233158826828,
+      "learning_rate": 4.129651550164021e-06,
+      "loss": 0.1825,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5920857566427187,
+      "grad_norm": 0.31639593839645386,
+      "learning_rate": 4.128250357832374e-06,
+      "loss": 0.1835,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5925419089976052,
+      "grad_norm": 0.3226074278354645,
+      "learning_rate": 4.1268482766386106e-06,
+      "loss": 0.1836,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5929980613524918,
+      "grad_norm": 0.3120798170566559,
+      "learning_rate": 4.125445307348129e-06,
+      "loss": 0.1829,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5934542137073783,
+      "grad_norm": 0.3283137381076813,
+      "learning_rate": 4.124041450726809e-06,
+      "loss": 0.1786,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5939103660622648,
+      "grad_norm": 0.30323657393455505,
+      "learning_rate": 4.122636707541018e-06,
+      "loss": 0.1783,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5943665184171514,
+      "grad_norm": 0.3175077438354492,
+      "learning_rate": 4.121231078557603e-06,
+      "loss": 0.1709,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5948226707720379,
+      "grad_norm": 0.32461968064308167,
+      "learning_rate": 4.119824564543901e-06,
+      "loss": 0.1905,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5952788231269244,
+      "grad_norm": 0.31229013204574585,
+      "learning_rate": 4.1184171662677234e-06,
+      "loss": 0.1728,
+      "step": 1305
+    },
+    {
+      "epoch": 0.595734975481811,
+      "grad_norm": 0.3311435580253601,
+      "learning_rate": 4.117008884497371e-06,
+      "loss": 0.1837,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5961911278366975,
+      "grad_norm": 0.32371848821640015,
+      "learning_rate": 4.115599720001627e-06,
+      "loss": 0.1768,
+      "step": 1307
+    },
+    {
+      "epoch": 0.596647280191584,
+      "grad_norm": 0.33422669768333435,
+      "learning_rate": 4.114189673549752e-06,
+      "loss": 0.1897,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5971034325464705,
+      "grad_norm": 0.30875715613365173,
+      "learning_rate": 4.112778745911492e-06,
+      "loss": 0.1865,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5975595849013571,
+      "grad_norm": 0.3149239420890808,
+      "learning_rate": 4.111366937857072e-06,
+      "loss": 0.182,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5980157372562436,
+      "grad_norm": 0.3077428638935089,
+      "learning_rate": 4.109954250157198e-06,
+      "loss": 0.1717,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5984718896111301,
+      "grad_norm": 0.31923070549964905,
+      "learning_rate": 4.108540683583057e-06,
+      "loss": 0.1671,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5989280419660167,
+      "grad_norm": 0.3275984227657318,
+      "learning_rate": 4.107126238906315e-06,
+      "loss": 0.1897,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5993841943209032,
+      "grad_norm": 0.3374669849872589,
+      "learning_rate": 4.1057109168991176e-06,
+      "loss": 0.1967,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5998403466757897,
+      "grad_norm": 0.31628432869911194,
+      "learning_rate": 4.104294718334089e-06,
+      "loss": 0.1834,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6002964990306763,
+      "grad_norm": 0.31935304403305054,
+      "learning_rate": 4.102877643984332e-06,
+      "loss": 0.1736,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6007526513855628,
+      "grad_norm": 0.3521789014339447,
+      "learning_rate": 4.101459694623428e-06,
+      "loss": 0.1756,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6012088037404493,
+      "grad_norm": 0.30787909030914307,
+      "learning_rate": 4.100040871025437e-06,
+      "loss": 0.1737,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6016649560953359,
+      "grad_norm": 0.3424101173877716,
+      "learning_rate": 4.098621173964894e-06,
+      "loss": 0.182,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6021211084502224,
+      "grad_norm": 0.31995871663093567,
+      "learning_rate": 4.097200604216811e-06,
+      "loss": 0.1859,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6025772608051089,
+      "grad_norm": 0.29601237177848816,
+      "learning_rate": 4.095779162556678e-06,
+      "loss": 0.1754,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6030334131599955,
+      "grad_norm": 0.3001455068588257,
+      "learning_rate": 4.09435684976046e-06,
+      "loss": 0.1699,
+      "step": 1322
+    },
+    {
+      "epoch": 0.603489565514882,
+      "grad_norm": 0.31265977025032043,
+      "learning_rate": 4.092933666604599e-06,
+      "loss": 0.1773,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6039457178697685,
+      "grad_norm": 0.310334175825119,
+      "learning_rate": 4.09150961386601e-06,
+      "loss": 0.1805,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6044018702246551,
+      "grad_norm": 0.3342970013618469,
+      "learning_rate": 4.090084692322082e-06,
+      "loss": 0.1734,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6048580225795416,
+      "grad_norm": 0.326528400182724,
+      "learning_rate": 4.088658902750682e-06,
+      "loss": 0.1802,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6053141749344281,
+      "grad_norm": 0.2973044514656067,
+      "learning_rate": 4.0872322459301485e-06,
+      "loss": 0.1676,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6057703272893147,
+      "grad_norm": 0.31312698125839233,
+      "learning_rate": 4.085804722639293e-06,
+      "loss": 0.1724,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6062264796442012,
+      "grad_norm": 0.31389057636260986,
+      "learning_rate": 4.084376333657401e-06,
+      "loss": 0.1825,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6066826319990877,
+      "grad_norm": 0.3122806251049042,
+      "learning_rate": 4.08294707976423e-06,
+      "loss": 0.1696,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6071387843539743,
+      "grad_norm": 0.3433966636657715,
+      "learning_rate": 4.081516961740009e-06,
+      "loss": 0.1867,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.3198709189891815,
+      "learning_rate": 4.0800859803654436e-06,
+      "loss": 0.1799,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6080510890637473,
+      "grad_norm": 0.33092859387397766,
+      "learning_rate": 4.078654136421702e-06,
+      "loss": 0.18,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6085072414186338,
+      "grad_norm": 0.31809109449386597,
+      "learning_rate": 4.077221430690431e-06,
+      "loss": 0.1831,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6089633937735204,
+      "grad_norm": 0.3202691972255707,
+      "learning_rate": 4.075787863953743e-06,
+      "loss": 0.1758,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6094195461284069,
+      "grad_norm": 0.3313249349594116,
+      "learning_rate": 4.074353436994223e-06,
+      "loss": 0.1858,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6098756984832934,
+      "grad_norm": 0.3066006302833557,
+      "learning_rate": 4.072918150594926e-06,
+      "loss": 0.1794,
+      "step": 1337
+    },
+    {
+      "epoch": 0.61033185083818,
+      "grad_norm": 0.30945560336112976,
+      "learning_rate": 4.071482005539373e-06,
+      "loss": 0.1747,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6107880031930665,
+      "grad_norm": 0.296085387468338,
+      "learning_rate": 4.070045002611557e-06,
+      "loss": 0.1761,
+      "step": 1339
+    },
+    {
+      "epoch": 0.611244155547953,
+      "grad_norm": 0.3362123966217041,
+      "learning_rate": 4.068607142595939e-06,
+      "loss": 0.179,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6117003079028396,
+      "grad_norm": 0.3089531362056732,
+      "learning_rate": 4.067168426277445e-06,
+      "loss": 0.1721,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6121564602577261,
+      "grad_norm": 0.32868391275405884,
+      "learning_rate": 4.065728854441472e-06,
+      "loss": 0.1892,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6126126126126126,
+      "grad_norm": 0.33618587255477905,
+      "learning_rate": 4.064288427873881e-06,
+      "loss": 0.1848,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6130687649674992,
+      "grad_norm": 0.3443443477153778,
+      "learning_rate": 4.062847147361003e-06,
+      "loss": 0.1861,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6135249173223857,
+      "grad_norm": 0.3269481956958771,
+      "learning_rate": 4.061405013689632e-06,
+      "loss": 0.176,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6139810696772722,
+      "grad_norm": 0.30411383509635925,
+      "learning_rate": 4.0599620276470295e-06,
+      "loss": 0.1621,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6144372220321588,
+      "grad_norm": 0.31699591875076294,
+      "learning_rate": 4.058518190020921e-06,
+      "loss": 0.1748,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6148933743870453,
+      "grad_norm": 0.32867613434791565,
+      "learning_rate": 4.0570735015994986e-06,
+      "loss": 0.1774,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6153495267419318,
+      "grad_norm": 0.3087550401687622,
+      "learning_rate": 4.055627963171418e-06,
+      "loss": 0.1812,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6158056790968184,
+      "grad_norm": 0.3262806534767151,
+      "learning_rate": 4.054181575525797e-06,
+      "loss": 0.1779,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6162618314517049,
+      "grad_norm": 0.29999014735221863,
+      "learning_rate": 4.052734339452221e-06,
+      "loss": 0.1785,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6167179838065914,
+      "grad_norm": 0.30150437355041504,
+      "learning_rate": 4.0512862557407365e-06,
+      "loss": 0.1801,
+      "step": 1352
+    },
+    {
+      "epoch": 0.617174136161478,
+      "grad_norm": 0.3183686435222626,
+      "learning_rate": 4.049837325181851e-06,
+      "loss": 0.1816,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6176302885163645,
+      "grad_norm": 0.3078478276729584,
+      "learning_rate": 4.048387548566536e-06,
+      "loss": 0.1684,
+      "step": 1354
+    },
+    {
+      "epoch": 0.618086440871251,
+      "grad_norm": 0.3235067129135132,
+      "learning_rate": 4.046936926686226e-06,
+      "loss": 0.1785,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6185425932261376,
+      "grad_norm": 0.33682987093925476,
+      "learning_rate": 4.045485460332815e-06,
+      "loss": 0.1737,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6189987455810241,
+      "grad_norm": 0.3047386705875397,
+      "learning_rate": 4.044033150298659e-06,
+      "loss": 0.1806,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6194548979359106,
+      "grad_norm": 0.30245861411094666,
+      "learning_rate": 4.042579997376574e-06,
+      "loss": 0.1747,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6199110502907971,
+      "grad_norm": 0.30369704961776733,
+      "learning_rate": 4.041126002359837e-06,
+      "loss": 0.1781,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6203672026456837,
+      "grad_norm": 0.326681911945343,
+      "learning_rate": 4.0396711660421825e-06,
+      "loss": 0.1898,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6208233550005702,
+      "grad_norm": 0.3057461977005005,
+      "learning_rate": 4.038215489217808e-06,
+      "loss": 0.1681,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6212795073554567,
+      "grad_norm": 0.3232496976852417,
+      "learning_rate": 4.036758972681366e-06,
+      "loss": 0.1797,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6217356597103433,
+      "grad_norm": 0.3215213716030121,
+      "learning_rate": 4.03530161722797e-06,
+      "loss": 0.1793,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6221918120652298,
+      "grad_norm": 0.30891114473342896,
+      "learning_rate": 4.03384342365319e-06,
+      "loss": 0.1675,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6226479644201163,
+      "grad_norm": 0.3129180669784546,
+      "learning_rate": 4.0323843927530545e-06,
+      "loss": 0.1803,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6231041167750029,
+      "grad_norm": 0.310806006193161,
+      "learning_rate": 4.030924525324049e-06,
+      "loss": 0.1713,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6235602691298894,
+      "grad_norm": 0.30714869499206543,
+      "learning_rate": 4.029463822163114e-06,
+      "loss": 0.1697,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6240164214847759,
+      "grad_norm": 0.32029086351394653,
+      "learning_rate": 4.02800228406765e-06,
+      "loss": 0.1894,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6244725738396625,
+      "grad_norm": 0.3005598783493042,
+      "learning_rate": 4.026539911835508e-06,
+      "loss": 0.172,
+      "step": 1369
+    },
+    {
+      "epoch": 0.624928726194549,
+      "grad_norm": 0.3286812901496887,
+      "learning_rate": 4.025076706264999e-06,
+      "loss": 0.1884,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6253848785494355,
+      "grad_norm": 0.31396982073783875,
+      "learning_rate": 4.023612668154887e-06,
+      "loss": 0.1676,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6258410309043221,
+      "grad_norm": 0.32762932777404785,
+      "learning_rate": 4.02214779830439e-06,
+      "loss": 0.1912,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6262971832592086,
+      "grad_norm": 0.3083411455154419,
+      "learning_rate": 4.020682097513181e-06,
+      "loss": 0.1779,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6267533356140951,
+      "grad_norm": 0.29053211212158203,
+      "learning_rate": 4.019215566581387e-06,
+      "loss": 0.1672,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6272094879689817,
+      "grad_norm": 0.30577850341796875,
+      "learning_rate": 4.0177482063095866e-06,
+      "loss": 0.174,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6276656403238682,
+      "grad_norm": 0.3015972077846527,
+      "learning_rate": 4.016280017498812e-06,
+      "loss": 0.1684,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6281217926787547,
+      "grad_norm": 0.3299334943294525,
+      "learning_rate": 4.014811000950547e-06,
+      "loss": 0.1887,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6285779450336413,
+      "grad_norm": 0.31303149461746216,
+      "learning_rate": 4.013341157466729e-06,
+      "loss": 0.1807,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6290340973885278,
+      "grad_norm": 0.32456493377685547,
+      "learning_rate": 4.011870487849747e-06,
+      "loss": 0.1887,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6294902497434143,
+      "grad_norm": 0.3507488965988159,
+      "learning_rate": 4.010398992902437e-06,
+      "loss": 0.1778,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6299464020983009,
+      "grad_norm": 0.301596999168396,
+      "learning_rate": 4.00892667342809e-06,
+      "loss": 0.1755,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6304025544531874,
+      "grad_norm": 0.30903780460357666,
+      "learning_rate": 4.0074535302304454e-06,
+      "loss": 0.1816,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6308587068080739,
+      "grad_norm": 0.3234713673591614,
+      "learning_rate": 4.0059795641136916e-06,
+      "loss": 0.1799,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6313148591629604,
+      "grad_norm": 0.30154043436050415,
+      "learning_rate": 4.004504775882467e-06,
+      "loss": 0.1802,
+      "step": 1384
+    },
+    {
+      "epoch": 0.631771011517847,
+      "grad_norm": 0.2991478145122528,
+      "learning_rate": 4.00302916634186e-06,
+      "loss": 0.177,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6322271638727335,
+      "grad_norm": 0.31340357661247253,
+      "learning_rate": 4.001552736297405e-06,
+      "loss": 0.1705,
+      "step": 1386
+    },
+    {
+      "epoch": 0.63268331622762,
+      "grad_norm": 0.30802538990974426,
+      "learning_rate": 4.000075486555085e-06,
+      "loss": 0.1748,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6331394685825066,
+      "grad_norm": 0.31243154406547546,
+      "learning_rate": 3.998597417921331e-06,
+      "loss": 0.1867,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6335956209373931,
+      "grad_norm": 0.31176456809043884,
+      "learning_rate": 3.997118531203022e-06,
+      "loss": 0.1713,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6340517732922796,
+      "grad_norm": 0.3167435824871063,
+      "learning_rate": 3.9956388272074806e-06,
+      "loss": 0.1753,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6345079256471662,
+      "grad_norm": 0.32061755657196045,
+      "learning_rate": 3.99415830674248e-06,
+      "loss": 0.1852,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6349640780020527,
+      "grad_norm": 0.3040485084056854,
+      "learning_rate": 3.992676970616233e-06,
+      "loss": 0.1694,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6354202303569392,
+      "grad_norm": 0.3159121870994568,
+      "learning_rate": 3.991194819637406e-06,
+      "loss": 0.1785,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6358763827118258,
+      "grad_norm": 0.3032073974609375,
+      "learning_rate": 3.989711854615101e-06,
+      "loss": 0.1751,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6363325350667123,
+      "grad_norm": 0.3233620822429657,
+      "learning_rate": 3.988228076358871e-06,
+      "loss": 0.1808,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6367886874215988,
+      "grad_norm": 0.2979919910430908,
+      "learning_rate": 3.98674348567871e-06,
+      "loss": 0.171,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6372448397764854,
+      "grad_norm": 0.3137773871421814,
+      "learning_rate": 3.985258083385057e-06,
+      "loss": 0.1722,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6377009921313719,
+      "grad_norm": 0.31898343563079834,
+      "learning_rate": 3.983771870288793e-06,
+      "loss": 0.1765,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6381571444862584,
+      "grad_norm": 0.31646081805229187,
+      "learning_rate": 3.982284847201242e-06,
+      "loss": 0.1862,
+      "step": 1399
+    },
+    {
+      "epoch": 0.638613296841145,
+      "grad_norm": 0.33168384432792664,
+      "learning_rate": 3.980797014934169e-06,
+      "loss": 0.1861,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6390694491960315,
+      "grad_norm": 0.31025266647338867,
+      "learning_rate": 3.979308374299785e-06,
+      "loss": 0.1792,
+      "step": 1401
+    },
+    {
+      "epoch": 0.639525601550918,
+      "grad_norm": 0.31457510590553284,
+      "learning_rate": 3.977818926110735e-06,
+      "loss": 0.1863,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6399817539058046,
+      "grad_norm": 0.31398001313209534,
+      "learning_rate": 3.976328671180112e-06,
+      "loss": 0.1778,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6404379062606911,
+      "grad_norm": 0.302520215511322,
+      "learning_rate": 3.974837610321445e-06,
+      "loss": 0.1766,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6408940586155776,
+      "grad_norm": 0.3468776047229767,
+      "learning_rate": 3.973345744348705e-06,
+      "loss": 0.1796,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6413502109704642,
+      "grad_norm": 0.3197506070137024,
+      "learning_rate": 3.9718530740763014e-06,
+      "loss": 0.1848,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6418063633253507,
+      "grad_norm": 0.3236994743347168,
+      "learning_rate": 3.970359600319083e-06,
+      "loss": 0.1928,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6422625156802372,
+      "grad_norm": 0.29972779750823975,
+      "learning_rate": 3.968865323892339e-06,
+      "loss": 0.1896,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6427186680351238,
+      "grad_norm": 0.3099987804889679,
+      "learning_rate": 3.967370245611792e-06,
+      "loss": 0.1752,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6431748203900103,
+      "grad_norm": 0.3267914056777954,
+      "learning_rate": 3.965874366293609e-06,
+      "loss": 0.1763,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6436309727448968,
+      "grad_norm": 0.3156987428665161,
+      "learning_rate": 3.964377686754388e-06,
+      "loss": 0.1789,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6440871250997833,
+      "grad_norm": 0.3007858991622925,
+      "learning_rate": 3.962880207811168e-06,
+      "loss": 0.1738,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6445432774546699,
+      "grad_norm": 0.3187737464904785,
+      "learning_rate": 3.961381930281423e-06,
+      "loss": 0.1946,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6449994298095564,
+      "grad_norm": 0.31307703256607056,
+      "learning_rate": 3.959882854983061e-06,
+      "loss": 0.184,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6454555821644429,
+      "grad_norm": 0.3020404279232025,
+      "learning_rate": 3.958382982734431e-06,
+      "loss": 0.1692,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6459117345193295,
+      "grad_norm": 0.31720787286758423,
+      "learning_rate": 3.95688231435431e-06,
+      "loss": 0.1804,
+      "step": 1416
+    },
+    {
+      "epoch": 0.646367886874216,
+      "grad_norm": 0.2947949767112732,
+      "learning_rate": 3.955380850661915e-06,
+      "loss": 0.167,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6468240392291025,
+      "grad_norm": 0.3007355034351349,
+      "learning_rate": 3.953878592476895e-06,
+      "loss": 0.1739,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6472801915839891,
+      "grad_norm": 0.33176013827323914,
+      "learning_rate": 3.952375540619332e-06,
+      "loss": 0.1874,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6477363439388756,
+      "grad_norm": 0.3278287947177887,
+      "learning_rate": 3.950871695909744e-06,
+      "loss": 0.179,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6481924962937621,
+      "grad_norm": 0.3073302209377289,
+      "learning_rate": 3.949367059169079e-06,
+      "loss": 0.1729,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 0.3037928342819214,
+      "learning_rate": 3.947861631218718e-06,
+      "loss": 0.1747,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6491048010035352,
+      "grad_norm": 0.3492172956466675,
+      "learning_rate": 3.9463554128804735e-06,
+      "loss": 0.1948,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6495609533584217,
+      "grad_norm": 0.34276771545410156,
+      "learning_rate": 3.944848404976593e-06,
+      "loss": 0.1895,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6500171057133083,
+      "grad_norm": 0.3183286488056183,
+      "learning_rate": 3.9433406083297515e-06,
+      "loss": 0.1811,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6504732580681948,
+      "grad_norm": 0.3316059112548828,
+      "learning_rate": 3.941832023763055e-06,
+      "loss": 0.1847,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6509294104230813,
+      "grad_norm": 0.33066871762275696,
+      "learning_rate": 3.940322652100039e-06,
+      "loss": 0.1817,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6513855627779679,
+      "grad_norm": 0.35097697377204895,
+      "learning_rate": 3.93881249416467e-06,
+      "loss": 0.1792,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6518417151328544,
+      "grad_norm": 0.3205112814903259,
+      "learning_rate": 3.937301550781345e-06,
+      "loss": 0.182,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6522978674877409,
+      "grad_norm": 0.3211330473423004,
+      "learning_rate": 3.935789822774887e-06,
+      "loss": 0.1813,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6527540198426275,
+      "grad_norm": 0.31473127007484436,
+      "learning_rate": 3.9342773109705505e-06,
+      "loss": 0.1675,
+      "step": 1431
+    },
+    {
+      "epoch": 0.653210172197514,
+      "grad_norm": 0.3301873207092285,
+      "learning_rate": 3.932764016194013e-06,
+      "loss": 0.1807,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6536663245524005,
+      "grad_norm": 0.29816126823425293,
+      "learning_rate": 3.931249939271382e-06,
+      "loss": 0.168,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6541224769072871,
+      "grad_norm": 0.31088292598724365,
+      "learning_rate": 3.929735081029196e-06,
+      "loss": 0.1777,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6545786292621736,
+      "grad_norm": 0.310939759016037,
+      "learning_rate": 3.928219442294414e-06,
+      "loss": 0.1723,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6550347816170601,
+      "grad_norm": 0.3389177918434143,
+      "learning_rate": 3.926703023894424e-06,
+      "loss": 0.1819,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6554909339719466,
+      "grad_norm": 0.32863786816596985,
+      "learning_rate": 3.925185826657038e-06,
+      "loss": 0.1804,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6559470863268332,
+      "grad_norm": 0.312286913394928,
+      "learning_rate": 3.923667851410494e-06,
+      "loss": 0.1797,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6564032386817197,
+      "grad_norm": 0.31244441866874695,
+      "learning_rate": 3.922149098983458e-06,
+      "loss": 0.1759,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6568593910366062,
+      "grad_norm": 0.3101857006549835,
+      "learning_rate": 3.920629570205014e-06,
+      "loss": 0.1827,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6573155433914928,
+      "grad_norm": 0.31726837158203125,
+      "learning_rate": 3.919109265904674e-06,
+      "loss": 0.1757,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6577716957463793,
+      "grad_norm": 0.3110054135322571,
+      "learning_rate": 3.917588186912372e-06,
+      "loss": 0.1776,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6582278481012658,
+      "grad_norm": 0.31904950737953186,
+      "learning_rate": 3.916066334058466e-06,
+      "loss": 0.1812,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6586840004561524,
+      "grad_norm": 0.3162565529346466,
+      "learning_rate": 3.914543708173735e-06,
+      "loss": 0.1702,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6591401528110389,
+      "grad_norm": 0.3111371397972107,
+      "learning_rate": 3.913020310089381e-06,
+      "loss": 0.1798,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6595963051659254,
+      "grad_norm": 0.30505436658859253,
+      "learning_rate": 3.911496140637029e-06,
+      "loss": 0.1715,
+      "step": 1446
+    },
+    {
+      "epoch": 0.660052457520812,
+      "grad_norm": 0.3237900733947754,
+      "learning_rate": 3.909971200648719e-06,
+      "loss": 0.1845,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6605086098756985,
+      "grad_norm": 0.32273730635643005,
+      "learning_rate": 3.90844549095692e-06,
+      "loss": 0.1814,
+      "step": 1448
+    },
+    {
+      "epoch": 0.660964762230585,
+      "grad_norm": 0.298110693693161,
+      "learning_rate": 3.9069190123945155e-06,
+      "loss": 0.1631,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6614209145854716,
+      "grad_norm": 0.2936743497848511,
+      "learning_rate": 3.90539176579481e-06,
+      "loss": 0.174,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6618770669403581,
+      "grad_norm": 0.31390246748924255,
+      "learning_rate": 3.903863751991529e-06,
+      "loss": 0.1813,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6623332192952446,
+      "grad_norm": 0.3006286323070526,
+      "learning_rate": 3.9023349718188155e-06,
+      "loss": 0.1692,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6627893716501312,
+      "grad_norm": 0.3290579617023468,
+      "learning_rate": 3.900805426111228e-06,
+      "loss": 0.1838,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6632455240050177,
+      "grad_norm": 0.31978702545166016,
+      "learning_rate": 3.8992751157037486e-06,
+      "loss": 0.1812,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6637016763599042,
+      "grad_norm": 0.3119882345199585,
+      "learning_rate": 3.897744041431773e-06,
+      "loss": 0.182,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6641578287147908,
+      "grad_norm": 0.3049165904521942,
+      "learning_rate": 3.8962122041311155e-06,
+      "loss": 0.1831,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6646139810696773,
+      "grad_norm": 0.2950694262981415,
+      "learning_rate": 3.894679604638006e-06,
+      "loss": 0.168,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6650701334245638,
+      "grad_norm": 0.3176410496234894,
+      "learning_rate": 3.89314624378909e-06,
+      "loss": 0.1803,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6655262857794504,
+      "grad_norm": 0.3513249158859253,
+      "learning_rate": 3.89161212242143e-06,
+      "loss": 0.1933,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6659824381343369,
+      "grad_norm": 0.31532546877861023,
+      "learning_rate": 3.890077241372503e-06,
+      "loss": 0.1693,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6664385904892234,
+      "grad_norm": 0.3120201528072357,
+      "learning_rate": 3.888541601480202e-06,
+      "loss": 0.1754,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6668947428441099,
+      "grad_norm": 0.312911719083786,
+      "learning_rate": 3.887005203582831e-06,
+      "loss": 0.1765,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6673508951989965,
+      "grad_norm": 0.32318025827407837,
+      "learning_rate": 3.885468048519111e-06,
+      "loss": 0.1832,
+      "step": 1463
+    },
+    {
+      "epoch": 0.667807047553883,
+      "grad_norm": 0.30750149488449097,
+      "learning_rate": 3.883930137128175e-06,
+      "loss": 0.1755,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6682631999087695,
+      "grad_norm": 0.30958881974220276,
+      "learning_rate": 3.882391470249569e-06,
+      "loss": 0.1815,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6687193522636561,
+      "grad_norm": 0.3040544092655182,
+      "learning_rate": 3.880852048723252e-06,
+      "loss": 0.1609,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6691755046185426,
+      "grad_norm": 0.2950521409511566,
+      "learning_rate": 3.879311873389594e-06,
+      "loss": 0.1665,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6696316569734291,
+      "grad_norm": 0.3143308758735657,
+      "learning_rate": 3.877770945089377e-06,
+      "loss": 0.1802,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6700878093283157,
+      "grad_norm": 0.3149574100971222,
+      "learning_rate": 3.876229264663792e-06,
+      "loss": 0.1849,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6705439616832022,
+      "grad_norm": 0.32009440660476685,
+      "learning_rate": 3.874686832954446e-06,
+      "loss": 0.1859,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6710001140380887,
+      "grad_norm": 0.32351619005203247,
+      "learning_rate": 3.8731436508033514e-06,
+      "loss": 0.1783,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6714562663929753,
+      "grad_norm": 0.3076383173465729,
+      "learning_rate": 3.871599719052931e-06,
+      "loss": 0.1863,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6719124187478618,
+      "grad_norm": 0.3264526128768921,
+      "learning_rate": 3.870055038546018e-06,
+      "loss": 0.1772,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6723685711027483,
+      "grad_norm": 0.32554197311401367,
+      "learning_rate": 3.868509610125853e-06,
+      "loss": 0.1847,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6728247234576349,
+      "grad_norm": 0.31761273741722107,
+      "learning_rate": 3.866963434636086e-06,
+      "loss": 0.1763,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6732808758125214,
+      "grad_norm": 0.30616438388824463,
+      "learning_rate": 3.865416512920776e-06,
+      "loss": 0.1752,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6737370281674079,
+      "grad_norm": 0.31411558389663696,
+      "learning_rate": 3.863868845824384e-06,
+      "loss": 0.1818,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6741931805222945,
+      "grad_norm": 0.311724454164505,
+      "learning_rate": 3.862320434191785e-06,
+      "loss": 0.1781,
+      "step": 1478
+    },
+    {
+      "epoch": 0.674649332877181,
+      "grad_norm": 0.3308613896369934,
+      "learning_rate": 3.860771278868256e-06,
+      "loss": 0.1824,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6751054852320675,
+      "grad_norm": 0.33403635025024414,
+      "learning_rate": 3.859221380699482e-06,
+      "loss": 0.1705,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6755616375869541,
+      "grad_norm": 0.31827399134635925,
+      "learning_rate": 3.857670740531551e-06,
+      "loss": 0.1776,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6760177899418406,
+      "grad_norm": 0.3525606393814087,
+      "learning_rate": 3.856119359210959e-06,
+      "loss": 0.1911,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6764739422967271,
+      "grad_norm": 0.32447677850723267,
+      "learning_rate": 3.854567237584605e-06,
+      "loss": 0.1758,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6769300946516137,
+      "grad_norm": 0.32655781507492065,
+      "learning_rate": 3.853014376499792e-06,
+      "loss": 0.1766,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6773862470065002,
+      "grad_norm": 0.30575570464134216,
+      "learning_rate": 3.851460776804228e-06,
+      "loss": 0.1719,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6778423993613867,
+      "grad_norm": 0.30875733494758606,
+      "learning_rate": 3.849906439346022e-06,
+      "loss": 0.1777,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6782985517162733,
+      "grad_norm": 0.28560709953308105,
+      "learning_rate": 3.848351364973688e-06,
+      "loss": 0.1715,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6787547040711598,
+      "grad_norm": 0.330027312040329,
+      "learning_rate": 3.846795554536141e-06,
+      "loss": 0.1795,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6792108564260463,
+      "grad_norm": 0.3124745190143585,
+      "learning_rate": 3.845239008882698e-06,
+      "loss": 0.1815,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6796670087809328,
+      "grad_norm": 0.31482622027397156,
+      "learning_rate": 3.843681728863079e-06,
+      "loss": 0.173,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6801231611358194,
+      "grad_norm": 0.3253953456878662,
+      "learning_rate": 3.842123715327402e-06,
+      "loss": 0.185,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6805793134907059,
+      "grad_norm": 0.30116111040115356,
+      "learning_rate": 3.840564969126186e-06,
+      "loss": 0.1767,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6810354658455924,
+      "grad_norm": 0.3304510712623596,
+      "learning_rate": 3.839005491110352e-06,
+      "loss": 0.173,
+      "step": 1493
+    },
+    {
+      "epoch": 0.681491618200479,
+      "grad_norm": 0.3365388810634613,
+      "learning_rate": 3.837445282131219e-06,
+      "loss": 0.1907,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6819477705553655,
+      "grad_norm": 0.30363014340400696,
+      "learning_rate": 3.835884343040506e-06,
+      "loss": 0.1764,
+      "step": 1495
+    },
+    {
+      "epoch": 0.682403922910252,
+      "grad_norm": 0.31243661046028137,
+      "learning_rate": 3.834322674690329e-06,
+      "loss": 0.1711,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6828600752651386,
+      "grad_norm": 0.30954939126968384,
+      "learning_rate": 3.832760277933203e-06,
+      "loss": 0.174,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6833162276200251,
+      "grad_norm": 0.31466034054756165,
+      "learning_rate": 3.831197153622041e-06,
+      "loss": 0.1756,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6837723799749116,
+      "grad_norm": 0.3177936375141144,
+      "learning_rate": 3.829633302610153e-06,
+      "loss": 0.179,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6842285323297982,
+      "grad_norm": 0.33621156215667725,
+      "learning_rate": 3.828068725751245e-06,
+      "loss": 0.174,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6846846846846847,
+      "grad_norm": 0.33815908432006836,
+      "learning_rate": 3.826503423899419e-06,
+      "loss": 0.1795,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6851408370395712,
+      "grad_norm": 0.30445554852485657,
+      "learning_rate": 3.824937397909175e-06,
+      "loss": 0.165,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6855969893944578,
+      "grad_norm": 0.32556524872779846,
+      "learning_rate": 3.823370648635407e-06,
+      "loss": 0.1843,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6860531417493443,
+      "grad_norm": 0.2974243760108948,
+      "learning_rate": 3.8218031769334024e-06,
+      "loss": 0.1772,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6865092941042308,
+      "grad_norm": 0.31662964820861816,
+      "learning_rate": 3.820234983658845e-06,
+      "loss": 0.1769,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6869654464591174,
+      "grad_norm": 0.3074699938297272,
+      "learning_rate": 3.818666069667811e-06,
+      "loss": 0.1775,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6874215988140039,
+      "grad_norm": 0.31894782185554504,
+      "learning_rate": 3.817096435816772e-06,
+      "loss": 0.1747,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6878777511688904,
+      "grad_norm": 0.30165767669677734,
+      "learning_rate": 3.81552608296259e-06,
+      "loss": 0.177,
+      "step": 1508
+    },
+    {
+      "epoch": 0.688333903523777,
+      "grad_norm": 0.30644863843917847,
+      "learning_rate": 3.813955011962522e-06,
+      "loss": 0.1805,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6887900558786635,
+      "grad_norm": 0.30376261472702026,
+      "learning_rate": 3.812383223674215e-06,
+      "loss": 0.1762,
+      "step": 1510
+    },
+    {
+      "epoch": 0.68924620823355,
+      "grad_norm": 0.31697842478752136,
+      "learning_rate": 3.8108107189557087e-06,
+      "loss": 0.1785,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6897023605884366,
+      "grad_norm": 0.32420986890792847,
+      "learning_rate": 3.809237498665434e-06,
+      "loss": 0.1729,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6901585129433231,
+      "grad_norm": 0.32394835352897644,
+      "learning_rate": 3.807663563662212e-06,
+      "loss": 0.1802,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6906146652982096,
+      "grad_norm": 0.30360114574432373,
+      "learning_rate": 3.806088914805255e-06,
+      "loss": 0.1677,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6910708176530961,
+      "grad_norm": 0.3022841811180115,
+      "learning_rate": 3.8045135529541612e-06,
+      "loss": 0.1731,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6915269700079827,
+      "grad_norm": 0.31654092669487,
+      "learning_rate": 3.8029374789689234e-06,
+      "loss": 0.1865,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6919831223628692,
+      "grad_norm": 0.31795841455459595,
+      "learning_rate": 3.80136069370992e-06,
+      "loss": 0.1824,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6924392747177557,
+      "grad_norm": 0.32236313819885254,
+      "learning_rate": 3.7997831980379186e-06,
+      "loss": 0.1812,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6928954270726423,
+      "grad_norm": 0.302062451839447,
+      "learning_rate": 3.7982049928140734e-06,
+      "loss": 0.1748,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6933515794275288,
+      "grad_norm": 0.32038962841033936,
+      "learning_rate": 3.7966260788999278e-06,
+      "loss": 0.1765,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6938077317824153,
+      "grad_norm": 0.30634284019470215,
+      "learning_rate": 3.7950464571574104e-06,
+      "loss": 0.1758,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6942638841373019,
+      "grad_norm": 0.31122341752052307,
+      "learning_rate": 3.793466128448838e-06,
+      "loss": 0.1744,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6947200364921884,
+      "grad_norm": 0.30366992950439453,
+      "learning_rate": 3.7918850936369104e-06,
+      "loss": 0.1862,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6951761888470749,
+      "grad_norm": 0.31658536195755005,
+      "learning_rate": 3.7903033535847167e-06,
+      "loss": 0.1849,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6956323412019615,
+      "grad_norm": 0.2998535931110382,
+      "learning_rate": 3.788720909155728e-06,
+      "loss": 0.1742,
+      "step": 1525
+    },
+    {
+      "epoch": 0.696088493556848,
+      "grad_norm": 0.3026093542575836,
+      "learning_rate": 3.7871377612138015e-06,
+      "loss": 0.1711,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6965446459117345,
+      "grad_norm": 0.3110896050930023,
+      "learning_rate": 3.7855539106231775e-06,
+      "loss": 0.1817,
+      "step": 1527
+    },
+    {
+      "epoch": 0.6970007982666211,
+      "grad_norm": 0.304331511259079,
+      "learning_rate": 3.7839693582484806e-06,
+      "loss": 0.1788,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6974569506215076,
+      "grad_norm": 0.3040176033973694,
+      "learning_rate": 3.7823841049547183e-06,
+      "loss": 0.1726,
+      "step": 1529
+    },
+    {
+      "epoch": 0.6979131029763941,
+      "grad_norm": 0.3107008934020996,
+      "learning_rate": 3.780798151607279e-06,
+      "loss": 0.1769,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6983692553312807,
+      "grad_norm": 0.3180543780326843,
+      "learning_rate": 3.7792114990719365e-06,
+      "loss": 0.1876,
+      "step": 1531
+    },
+    {
+      "epoch": 0.6988254076861672,
+      "grad_norm": 0.29411572217941284,
+      "learning_rate": 3.7776241482148452e-06,
+      "loss": 0.1634,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6992815600410537,
+      "grad_norm": 0.31372714042663574,
+      "learning_rate": 3.776036099902539e-06,
+      "loss": 0.1805,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6997377123959403,
+      "grad_norm": 0.30445849895477295,
+      "learning_rate": 3.7744473550019335e-06,
+      "loss": 0.1698,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7001938647508268,
+      "grad_norm": 0.30601853132247925,
+      "learning_rate": 3.7728579143803258e-06,
+      "loss": 0.1768,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7006500171057133,
+      "grad_norm": 0.3063245713710785,
+      "learning_rate": 3.771267778905391e-06,
+      "loss": 0.1816,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7011061694605999,
+      "grad_norm": 0.29846036434173584,
+      "learning_rate": 3.7696769494451846e-06,
+      "loss": 0.1711,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7015623218154864,
+      "grad_norm": 0.28382304310798645,
+      "learning_rate": 3.76808542686814e-06,
+      "loss": 0.1617,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7020184741703729,
+      "grad_norm": 0.3130197823047638,
+      "learning_rate": 3.766493212043071e-06,
+      "loss": 0.1776,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7024746265252594,
+      "grad_norm": 0.325830340385437,
+      "learning_rate": 3.7649003058391664e-06,
+      "loss": 0.182,
+      "step": 1540
+    },
+    {
+      "epoch": 0.702930778880146,
+      "grad_norm": 0.32922762632369995,
+      "learning_rate": 3.7633067091259954e-06,
+      "loss": 0.1826,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7033869312350325,
+      "grad_norm": 0.33391472697257996,
+      "learning_rate": 3.7617124227735015e-06,
+      "loss": 0.1815,
+      "step": 1542
+    },
+    {
+      "epoch": 0.703843083589919,
+      "grad_norm": 0.3020962178707123,
+      "learning_rate": 3.7601174476520062e-06,
+      "loss": 0.1786,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7042992359448056,
+      "grad_norm": 0.2875640094280243,
+      "learning_rate": 3.7585217846322075e-06,
+      "loss": 0.163,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7047553882996921,
+      "grad_norm": 0.31332284212112427,
+      "learning_rate": 3.756925434585177e-06,
+      "loss": 0.1802,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7052115406545786,
+      "grad_norm": 0.31729385256767273,
+      "learning_rate": 3.755328398382362e-06,
+      "loss": 0.1749,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7056676930094652,
+      "grad_norm": 0.33252960443496704,
+      "learning_rate": 3.753730676895586e-06,
+      "loss": 0.1821,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7061238453643517,
+      "grad_norm": 0.3068774938583374,
+      "learning_rate": 3.7521322709970454e-06,
+      "loss": 0.174,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7065799977192382,
+      "grad_norm": 0.32396289706230164,
+      "learning_rate": 3.7505331815593093e-06,
+      "loss": 0.1796,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7070361500741248,
+      "grad_norm": 0.32775676250457764,
+      "learning_rate": 3.748933409455322e-06,
+      "loss": 0.1895,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7074923024290113,
+      "grad_norm": 0.33597442507743835,
+      "learning_rate": 3.7473329555583983e-06,
+      "loss": 0.1846,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7079484547838978,
+      "grad_norm": 0.2888537347316742,
+      "learning_rate": 3.745731820742227e-06,
+      "loss": 0.161,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7084046071387844,
+      "grad_norm": 0.3230069875717163,
+      "learning_rate": 3.744130005880866e-06,
+      "loss": 0.1831,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.31857040524482727,
+      "learning_rate": 3.742527511848749e-06,
+      "loss": 0.1896,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7093169118485574,
+      "grad_norm": 0.29441314935684204,
+      "learning_rate": 3.740924339520676e-06,
+      "loss": 0.1643,
+      "step": 1555
+    },
+    {
+      "epoch": 0.709773064203444,
+      "grad_norm": 0.3000827729701996,
+      "learning_rate": 3.7393204897718194e-06,
+      "loss": 0.1797,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7102292165583305,
+      "grad_norm": 0.3140794336795807,
+      "learning_rate": 3.7377159634777217e-06,
+      "loss": 0.1782,
+      "step": 1557
+    },
+    {
+      "epoch": 0.710685368913217,
+      "grad_norm": 0.3227851688861847,
+      "learning_rate": 3.7361107615142926e-06,
+      "loss": 0.18,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7111415212681036,
+      "grad_norm": 0.2972438931465149,
+      "learning_rate": 3.7345048847578136e-06,
+      "loss": 0.1743,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7115976736229901,
+      "grad_norm": 0.3304663598537445,
+      "learning_rate": 3.7328983340849324e-06,
+      "loss": 0.1773,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7120538259778766,
+      "grad_norm": 0.31646305322647095,
+      "learning_rate": 3.7312911103726648e-06,
+      "loss": 0.1859,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7125099783327632,
+      "grad_norm": 0.310647189617157,
+      "learning_rate": 3.7296832144983945e-06,
+      "loss": 0.1789,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7129661306876497,
+      "grad_norm": 0.31869450211524963,
+      "learning_rate": 3.7280746473398736e-06,
+      "loss": 0.1853,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7134222830425362,
+      "grad_norm": 0.30761584639549255,
+      "learning_rate": 3.7264654097752173e-06,
+      "loss": 0.1831,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7138784353974227,
+      "grad_norm": 0.3035110831260681,
+      "learning_rate": 3.7248555026829096e-06,
+      "loss": 0.1741,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7143345877523093,
+      "grad_norm": 0.3166104853153229,
+      "learning_rate": 3.7232449269417985e-06,
+      "loss": 0.1847,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7147907401071958,
+      "grad_norm": 0.31113430857658386,
+      "learning_rate": 3.7216336834310982e-06,
+      "loss": 0.181,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7152468924620823,
+      "grad_norm": 0.319774329662323,
+      "learning_rate": 3.7200217730303865e-06,
+      "loss": 0.1772,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7157030448169689,
+      "grad_norm": 0.3270570933818817,
+      "learning_rate": 3.7184091966196056e-06,
+      "loss": 0.1838,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7161591971718554,
+      "grad_norm": 0.3104270100593567,
+      "learning_rate": 3.716795955079061e-06,
+      "loss": 0.1688,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7166153495267419,
+      "grad_norm": 0.30215322971343994,
+      "learning_rate": 3.7151820492894214e-06,
+      "loss": 0.1719,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7170715018816285,
+      "grad_norm": 0.30320432782173157,
+      "learning_rate": 3.713567480131718e-06,
+      "loss": 0.1776,
+      "step": 1572
+    },
+    {
+      "epoch": 0.717527654236515,
+      "grad_norm": 0.31449124217033386,
+      "learning_rate": 3.7119522484873453e-06,
+      "loss": 0.1821,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7179838065914015,
+      "grad_norm": 0.31486573815345764,
+      "learning_rate": 3.710336355238058e-06,
+      "loss": 0.1785,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7184399589462881,
+      "grad_norm": 0.296237587928772,
+      "learning_rate": 3.708719801265972e-06,
+      "loss": 0.1718,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7188961113011746,
+      "grad_norm": 0.332865834236145,
+      "learning_rate": 3.7071025874535643e-06,
+      "loss": 0.1876,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7193522636560611,
+      "grad_norm": 0.31560611724853516,
+      "learning_rate": 3.7054847146836735e-06,
+      "loss": 0.1816,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7198084160109477,
+      "grad_norm": 0.3118779957294464,
+      "learning_rate": 3.7038661838394953e-06,
+      "loss": 0.1816,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7202645683658342,
+      "grad_norm": 0.3085733950138092,
+      "learning_rate": 3.702246995804586e-06,
+      "loss": 0.1712,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7207207207207207,
+      "grad_norm": 0.3030265271663666,
+      "learning_rate": 3.7006271514628617e-06,
+      "loss": 0.1732,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7211768730756073,
+      "grad_norm": 0.3013935387134552,
+      "learning_rate": 3.699006651698594e-06,
+      "loss": 0.1784,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7216330254304938,
+      "grad_norm": 0.31913870573043823,
+      "learning_rate": 3.697385497396416e-06,
+      "loss": 0.1752,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7220891777853803,
+      "grad_norm": 0.3300533592700958,
+      "learning_rate": 3.6957636894413136e-06,
+      "loss": 0.1936,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7225453301402669,
+      "grad_norm": 0.2990119159221649,
+      "learning_rate": 3.694141228718634e-06,
+      "loss": 0.1738,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7230014824951534,
+      "grad_norm": 0.3139911890029907,
+      "learning_rate": 3.6925181161140783e-06,
+      "loss": 0.1744,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7234576348500399,
+      "grad_norm": 0.3176199793815613,
+      "learning_rate": 3.690894352513703e-06,
+      "loss": 0.1876,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7239137872049265,
+      "grad_norm": 0.3010476231575012,
+      "learning_rate": 3.6892699388039223e-06,
+      "loss": 0.1668,
+      "step": 1587
+    },
+    {
+      "epoch": 0.724369939559813,
+      "grad_norm": 0.304487407207489,
+      "learning_rate": 3.6876448758715028e-06,
+      "loss": 0.1717,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7248260919146995,
+      "grad_norm": 0.3229767084121704,
+      "learning_rate": 3.6860191646035677e-06,
+      "loss": 0.1612,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7252822442695861,
+      "grad_norm": 0.3187156319618225,
+      "learning_rate": 3.684392805887591e-06,
+      "loss": 0.1781,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7257383966244726,
+      "grad_norm": 0.32359451055526733,
+      "learning_rate": 3.6827658006114046e-06,
+      "loss": 0.1827,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7261945489793591,
+      "grad_norm": 0.31425759196281433,
+      "learning_rate": 3.681138149663189e-06,
+      "loss": 0.1764,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7266507013342456,
+      "grad_norm": 0.3116792142391205,
+      "learning_rate": 3.6795098539314806e-06,
+      "loss": 0.1726,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7271068536891322,
+      "grad_norm": 0.32859688997268677,
+      "learning_rate": 3.677880914305165e-06,
+      "loss": 0.1802,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7275630060440187,
+      "grad_norm": 0.31765374541282654,
+      "learning_rate": 3.6762513316734814e-06,
+      "loss": 0.1773,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7280191583989052,
+      "grad_norm": 0.3266363739967346,
+      "learning_rate": 3.6746211069260197e-06,
+      "loss": 0.179,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7284753107537918,
+      "grad_norm": 0.32359579205513,
+      "learning_rate": 3.672990240952718e-06,
+      "loss": 0.1777,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7289314631086783,
+      "grad_norm": 0.3184155821800232,
+      "learning_rate": 3.671358734643867e-06,
+      "loss": 0.165,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7293876154635648,
+      "grad_norm": 0.3010857105255127,
+      "learning_rate": 3.669726588890108e-06,
+      "loss": 0.1701,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7298437678184514,
+      "grad_norm": 0.29946383833885193,
+      "learning_rate": 3.6680938045824284e-06,
+      "loss": 0.163,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7302999201733379,
+      "grad_norm": 0.3143792748451233,
+      "learning_rate": 3.6664603826121634e-06,
+      "loss": 0.1712,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7307560725282244,
+      "grad_norm": 0.31581059098243713,
+      "learning_rate": 3.6648263238710014e-06,
+      "loss": 0.1813,
+      "step": 1602
+    },
+    {
+      "epoch": 0.731212224883111,
+      "grad_norm": 0.3098791241645813,
+      "learning_rate": 3.663191629250974e-06,
+      "loss": 0.1735,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7316683772379975,
+      "grad_norm": 0.32205960154533386,
+      "learning_rate": 3.661556299644462e-06,
+      "loss": 0.1826,
+      "step": 1604
+    },
+    {
+      "epoch": 0.732124529592884,
+      "grad_norm": 0.3265833258628845,
+      "learning_rate": 3.65992033594419e-06,
+      "loss": 0.1734,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7325806819477706,
+      "grad_norm": 0.3077433705329895,
+      "learning_rate": 3.658283739043233e-06,
+      "loss": 0.1782,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7330368343026571,
+      "grad_norm": 0.31840401887893677,
+      "learning_rate": 3.656646509835008e-06,
+      "loss": 0.1672,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7334929866575436,
+      "grad_norm": 0.30904659628868103,
+      "learning_rate": 3.6550086492132804e-06,
+      "loss": 0.1683,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7339491390124302,
+      "grad_norm": 0.307577908039093,
+      "learning_rate": 3.6533701580721566e-06,
+      "loss": 0.1756,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7344052913673167,
+      "grad_norm": 0.3144417405128479,
+      "learning_rate": 3.6517310373060905e-06,
+      "loss": 0.1798,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7348614437222032,
+      "grad_norm": 0.33439385890960693,
+      "learning_rate": 3.650091287809878e-06,
+      "loss": 0.1749,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7353175960770898,
+      "grad_norm": 0.29521438479423523,
+      "learning_rate": 3.6484509104786582e-06,
+      "loss": 0.1649,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7357737484319763,
+      "grad_norm": 0.32933667302131653,
+      "learning_rate": 3.6468099062079137e-06,
+      "loss": 0.1817,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7362299007868628,
+      "grad_norm": 0.3304448425769806,
+      "learning_rate": 3.645168275893469e-06,
+      "loss": 0.1737,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7366860531417494,
+      "grad_norm": 0.31729376316070557,
+      "learning_rate": 3.6435260204314916e-06,
+      "loss": 0.1751,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7371422054966359,
+      "grad_norm": 0.3009093105792999,
+      "learning_rate": 3.6418831407184856e-06,
+      "loss": 0.1724,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7375983578515224,
+      "grad_norm": 0.32070010900497437,
+      "learning_rate": 3.6402396376513027e-06,
+      "loss": 0.1758,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7380545102064089,
+      "grad_norm": 0.3074767291545868,
+      "learning_rate": 3.63859551212713e-06,
+      "loss": 0.1695,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7385106625612955,
+      "grad_norm": 0.3176269233226776,
+      "learning_rate": 3.636950765043496e-06,
+      "loss": 0.1768,
+      "step": 1619
+    },
+    {
+      "epoch": 0.738966814916182,
+      "grad_norm": 0.31392332911491394,
+      "learning_rate": 3.6353053972982676e-06,
+      "loss": 0.1771,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7394229672710685,
+      "grad_norm": 0.3235676884651184,
+      "learning_rate": 3.6336594097896533e-06,
+      "loss": 0.1906,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7398791196259551,
+      "grad_norm": 0.31795981526374817,
+      "learning_rate": 3.632012803416196e-06,
+      "loss": 0.1729,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7403352719808416,
+      "grad_norm": 0.3236065208911896,
+      "learning_rate": 3.6303655790767795e-06,
+      "loss": 0.1861,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7407914243357281,
+      "grad_norm": 0.3049633502960205,
+      "learning_rate": 3.628717737670623e-06,
+      "loss": 0.1782,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7412475766906147,
+      "grad_norm": 0.3074170649051666,
+      "learning_rate": 3.627069280097284e-06,
+      "loss": 0.1739,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7417037290455012,
+      "grad_norm": 0.30917996168136597,
+      "learning_rate": 3.625420207256656e-06,
+      "loss": 0.1774,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7421598814003877,
+      "grad_norm": 0.32159876823425293,
+      "learning_rate": 3.6237705200489663e-06,
+      "loss": 0.1822,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7426160337552743,
+      "grad_norm": 0.3114009499549866,
+      "learning_rate": 3.6221202193747818e-06,
+      "loss": 0.182,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7430721861101608,
+      "grad_norm": 0.33464324474334717,
+      "learning_rate": 3.6204693061350003e-06,
+      "loss": 0.1793,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7435283384650473,
+      "grad_norm": 0.30630210041999817,
+      "learning_rate": 3.6188177812308567e-06,
+      "loss": 0.1709,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7439844908199339,
+      "grad_norm": 0.31895655393600464,
+      "learning_rate": 3.617165645563918e-06,
+      "loss": 0.1837,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7444406431748204,
+      "grad_norm": 0.3423558473587036,
+      "learning_rate": 3.6155129000360846e-06,
+      "loss": 0.1747,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7448967955297069,
+      "grad_norm": 0.3048911988735199,
+      "learning_rate": 3.613859545549592e-06,
+      "loss": 0.1735,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7453529478845935,
+      "grad_norm": 0.3302198052406311,
+      "learning_rate": 3.612205583007007e-06,
+      "loss": 0.1775,
+      "step": 1634
+    },
+    {
+      "epoch": 0.74580910023948,
+      "grad_norm": 0.31379425525665283,
+      "learning_rate": 3.610551013311225e-06,
+      "loss": 0.1834,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7462652525943665,
+      "grad_norm": 0.32724007964134216,
+      "learning_rate": 3.6088958373654794e-06,
+      "loss": 0.1866,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7467214049492531,
+      "grad_norm": 0.319243848323822,
+      "learning_rate": 3.607240056073329e-06,
+      "loss": 0.1718,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7471775573041396,
+      "grad_norm": 0.3103865385055542,
+      "learning_rate": 3.605583670338665e-06,
+      "loss": 0.1688,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7476337096590261,
+      "grad_norm": 0.33787649869918823,
+      "learning_rate": 3.603926681065709e-06,
+      "loss": 0.1826,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7480898620139127,
+      "grad_norm": 0.3046160638332367,
+      "learning_rate": 3.602269089159013e-06,
+      "loss": 0.1714,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7485460143687992,
+      "grad_norm": 0.29203805327415466,
+      "learning_rate": 3.6006108955234546e-06,
+      "loss": 0.1685,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7490021667236857,
+      "grad_norm": 0.3286929130554199,
+      "learning_rate": 3.5989521010642423e-06,
+      "loss": 0.1806,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7494583190785722,
+      "grad_norm": 0.3124166429042816,
+      "learning_rate": 3.5972927066869135e-06,
+      "loss": 0.1726,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7499144714334588,
+      "grad_norm": 0.32953932881355286,
+      "learning_rate": 3.5956327132973313e-06,
+      "loss": 0.1749,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7503706237883453,
+      "grad_norm": 0.3174798786640167,
+      "learning_rate": 3.5939721218016864e-06,
+      "loss": 0.1758,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7508267761432318,
+      "grad_norm": 0.3085630238056183,
+      "learning_rate": 3.5923109331064954e-06,
+      "loss": 0.1753,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7512829284981184,
+      "grad_norm": 0.31211572885513306,
+      "learning_rate": 3.590649148118602e-06,
+      "loss": 0.1685,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7517390808530049,
+      "grad_norm": 0.3159165680408478,
+      "learning_rate": 3.588986767745174e-06,
+      "loss": 0.1652,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7521952332078914,
+      "grad_norm": 0.3205738067626953,
+      "learning_rate": 3.587323792893706e-06,
+      "loss": 0.1736,
+      "step": 1649
+    },
+    {
+      "epoch": 0.752651385562778,
+      "grad_norm": 0.29008302092552185,
+      "learning_rate": 3.585660224472016e-06,
+      "loss": 0.1604,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7531075379176645,
+      "grad_norm": 0.3097541630268097,
+      "learning_rate": 3.5839960633882466e-06,
+      "loss": 0.1781,
+      "step": 1651
+    },
+    {
+      "epoch": 0.753563690272551,
+      "grad_norm": 0.3127402067184448,
+      "learning_rate": 3.5823313105508626e-06,
+      "loss": 0.1852,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7540198426274376,
+      "grad_norm": 0.3087466359138489,
+      "learning_rate": 3.5806659668686533e-06,
+      "loss": 0.1751,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7544759949823241,
+      "grad_norm": 0.2982191741466522,
+      "learning_rate": 3.57900003325073e-06,
+      "loss": 0.1729,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7549321473372106,
+      "grad_norm": 0.33493128418922424,
+      "learning_rate": 3.5773335106065275e-06,
+      "loss": 0.1841,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7553882996920972,
+      "grad_norm": 0.306412935256958,
+      "learning_rate": 3.575666399845799e-06,
+      "loss": 0.1767,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7558444520469837,
+      "grad_norm": 0.298348605632782,
+      "learning_rate": 3.5739987018786208e-06,
+      "loss": 0.174,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7563006044018702,
+      "grad_norm": 0.312343955039978,
+      "learning_rate": 3.57233041761539e-06,
+      "loss": 0.1789,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 0.3003070652484894,
+      "learning_rate": 3.570661547966823e-06,
+      "loss": 0.1704,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7572129091116433,
+      "grad_norm": 0.321424663066864,
+      "learning_rate": 3.568992093843956e-06,
+      "loss": 0.1827,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7576690614665298,
+      "grad_norm": 0.32341936230659485,
+      "learning_rate": 3.567322056158144e-06,
+      "loss": 0.1838,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7581252138214164,
+      "grad_norm": 0.3097274899482727,
+      "learning_rate": 3.5656514358210614e-06,
+      "loss": 0.1759,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7585813661763029,
+      "grad_norm": 0.3126348555088043,
+      "learning_rate": 3.5639802337446994e-06,
+      "loss": 0.1724,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7590375185311894,
+      "grad_norm": 0.3032686412334442,
+      "learning_rate": 3.5623084508413685e-06,
+      "loss": 0.1661,
+      "step": 1664
+    },
+    {
+      "epoch": 0.759493670886076,
+      "grad_norm": 0.3123730421066284,
+      "learning_rate": 3.5606360880236935e-06,
+      "loss": 0.1719,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7599498232409625,
+      "grad_norm": 0.3254353404045105,
+      "learning_rate": 3.558963146204619e-06,
+      "loss": 0.1776,
+      "step": 1666
+    },
+    {
+      "epoch": 0.760405975595849,
+      "grad_norm": 0.3022678792476654,
+      "learning_rate": 3.5572896262974044e-06,
+      "loss": 0.1655,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7608621279507356,
+      "grad_norm": 0.30680909752845764,
+      "learning_rate": 3.555615529215623e-06,
+      "loss": 0.1684,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7613182803056221,
+      "grad_norm": 0.30180925130844116,
+      "learning_rate": 3.5539408558731657e-06,
+      "loss": 0.1728,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7617744326605086,
+      "grad_norm": 0.3078424036502838,
+      "learning_rate": 3.552265607184237e-06,
+      "loss": 0.1774,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7622305850153951,
+      "grad_norm": 0.3140369653701782,
+      "learning_rate": 3.5505897840633552e-06,
+      "loss": 0.1699,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7626867373702817,
+      "grad_norm": 0.3076333999633789,
+      "learning_rate": 3.5489133874253516e-06,
+      "loss": 0.1686,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7631428897251682,
+      "grad_norm": 0.31459537148475647,
+      "learning_rate": 3.5472364181853726e-06,
+      "loss": 0.1804,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7635990420800547,
+      "grad_norm": 0.33527055382728577,
+      "learning_rate": 3.545558877258875e-06,
+      "loss": 0.1763,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7640551944349413,
+      "grad_norm": 0.30008918046951294,
+      "learning_rate": 3.54388076556163e-06,
+      "loss": 0.1655,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7645113467898278,
+      "grad_norm": 0.31347891688346863,
+      "learning_rate": 3.5422020840097173e-06,
+      "loss": 0.1626,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7649674991447143,
+      "grad_norm": 0.3020315170288086,
+      "learning_rate": 3.5405228335195306e-06,
+      "loss": 0.1676,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7654236514996009,
+      "grad_norm": 0.31036731600761414,
+      "learning_rate": 3.5388430150077726e-06,
+      "loss": 0.1785,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7658798038544874,
+      "grad_norm": 0.3237437605857849,
+      "learning_rate": 3.5371626293914563e-06,
+      "loss": 0.1743,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7663359562093739,
+      "grad_norm": 0.2993324398994446,
+      "learning_rate": 3.535481677587904e-06,
+      "loss": 0.1705,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7667921085642605,
+      "grad_norm": 0.3123346269130707,
+      "learning_rate": 3.5338001605147496e-06,
+      "loss": 0.1863,
+      "step": 1681
+    },
+    {
+      "epoch": 0.767248260919147,
+      "grad_norm": 0.30745282769203186,
+      "learning_rate": 3.5321180790899317e-06,
+      "loss": 0.1707,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7677044132740335,
+      "grad_norm": 0.29562121629714966,
+      "learning_rate": 3.530435434231699e-06,
+      "loss": 0.1722,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7681605656289201,
+      "grad_norm": 0.28618761897087097,
+      "learning_rate": 3.5287522268586074e-06,
+      "loss": 0.1628,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7686167179838066,
+      "grad_norm": 0.31206732988357544,
+      "learning_rate": 3.527068457889521e-06,
+      "loss": 0.1764,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7690728703386931,
+      "grad_norm": 0.2976711094379425,
+      "learning_rate": 3.525384128243609e-06,
+      "loss": 0.1766,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7695290226935797,
+      "grad_norm": 0.30767711997032166,
+      "learning_rate": 3.5236992388403467e-06,
+      "loss": 0.1701,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7699851750484662,
+      "grad_norm": 0.31731992959976196,
+      "learning_rate": 3.5220137905995165e-06,
+      "loss": 0.1813,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7704413274033527,
+      "grad_norm": 0.31212568283081055,
+      "learning_rate": 3.520327784441204e-06,
+      "loss": 0.1715,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7708974797582393,
+      "grad_norm": 0.3051251471042633,
+      "learning_rate": 3.518641221285801e-06,
+      "loss": 0.1752,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7713536321131258,
+      "grad_norm": 0.3114382326602936,
+      "learning_rate": 3.5169541020540023e-06,
+      "loss": 0.1778,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7718097844680123,
+      "grad_norm": 0.2959032654762268,
+      "learning_rate": 3.515266427666806e-06,
+      "loss": 0.1733,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7722659368228989,
+      "grad_norm": 0.3071809709072113,
+      "learning_rate": 3.5135781990455144e-06,
+      "loss": 0.1761,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7727220891777854,
+      "grad_norm": 0.3115699291229248,
+      "learning_rate": 3.511889417111731e-06,
+      "loss": 0.1755,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7731782415326719,
+      "grad_norm": 0.32594627141952515,
+      "learning_rate": 3.510200082787363e-06,
+      "loss": 0.1835,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7736343938875584,
+      "grad_norm": 0.3007581830024719,
+      "learning_rate": 3.508510196994618e-06,
+      "loss": 0.181,
+      "step": 1696
+    },
+    {
+      "epoch": 0.774090546242445,
+      "grad_norm": 0.3074135482311249,
+      "learning_rate": 3.506819760656005e-06,
+      "loss": 0.1795,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7745466985973315,
+      "grad_norm": 0.3041733503341675,
+      "learning_rate": 3.505128774694333e-06,
+      "loss": 0.172,
+      "step": 1698
+    },
+    {
+      "epoch": 0.775002850952218,
+      "grad_norm": 0.33501139283180237,
+      "learning_rate": 3.503437240032711e-06,
+      "loss": 0.1833,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7754590033071046,
+      "grad_norm": 0.32189062237739563,
+      "learning_rate": 3.50174515759455e-06,
+      "loss": 0.1777,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7759151556619911,
+      "grad_norm": 0.3350752592086792,
+      "learning_rate": 3.500052528303556e-06,
+      "loss": 0.1822,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7763713080168776,
+      "grad_norm": 0.2975068986415863,
+      "learning_rate": 3.4983593530837373e-06,
+      "loss": 0.1699,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7768274603717642,
+      "grad_norm": 0.31347838044166565,
+      "learning_rate": 3.496665632859397e-06,
+      "loss": 0.1732,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7772836127266507,
+      "grad_norm": 0.32078996300697327,
+      "learning_rate": 3.4949713685551377e-06,
+      "loss": 0.1656,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7777397650815372,
+      "grad_norm": 0.33548134565353394,
+      "learning_rate": 3.4932765610958592e-06,
+      "loss": 0.1848,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7781959174364238,
+      "grad_norm": 0.2939399778842926,
+      "learning_rate": 3.491581211406756e-06,
+      "loss": 0.1671,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7786520697913103,
+      "grad_norm": 0.3254663944244385,
+      "learning_rate": 3.489885320413321e-06,
+      "loss": 0.1848,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7791082221461968,
+      "grad_norm": 0.3030187785625458,
+      "learning_rate": 3.488188889041341e-06,
+      "loss": 0.176,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7795643745010834,
+      "grad_norm": 0.30583539605140686,
+      "learning_rate": 3.4864919182168973e-06,
+      "loss": 0.1708,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7800205268559699,
+      "grad_norm": 0.32346537709236145,
+      "learning_rate": 3.4847944088663666e-06,
+      "loss": 0.1743,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7804766792108564,
+      "grad_norm": 0.328371524810791,
+      "learning_rate": 3.483096361916421e-06,
+      "loss": 0.1896,
+      "step": 1711
+    },
+    {
+      "epoch": 0.780932831565743,
+      "grad_norm": 0.2983171045780182,
+      "learning_rate": 3.4813977782940234e-06,
+      "loss": 0.1647,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7813889839206295,
+      "grad_norm": 0.3053433895111084,
+      "learning_rate": 3.4796986589264316e-06,
+      "loss": 0.1755,
+      "step": 1713
+    },
+    {
+      "epoch": 0.781845136275516,
+      "grad_norm": 0.30685046315193176,
+      "learning_rate": 3.4779990047411926e-06,
+      "loss": 0.1773,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7823012886304026,
+      "grad_norm": 0.3175199329853058,
+      "learning_rate": 3.4762988166661516e-06,
+      "loss": 0.1803,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7827574409852891,
+      "grad_norm": 0.3137868344783783,
+      "learning_rate": 3.4745980956294396e-06,
+      "loss": 0.1699,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7832135933401756,
+      "grad_norm": 0.3052430748939514,
+      "learning_rate": 3.4728968425594805e-06,
+      "loss": 0.1668,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7836697456950622,
+      "grad_norm": 0.3157985508441925,
+      "learning_rate": 3.4711950583849883e-06,
+      "loss": 0.1778,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7841258980499487,
+      "grad_norm": 0.30576908588409424,
+      "learning_rate": 3.469492744034968e-06,
+      "loss": 0.1781,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7845820504048352,
+      "grad_norm": 0.3303896188735962,
+      "learning_rate": 3.4677899004387134e-06,
+      "loss": 0.1773,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7850382027597217,
+      "grad_norm": 0.306974858045578,
+      "learning_rate": 3.4660865285258063e-06,
+      "loss": 0.1736,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7854943551146083,
+      "grad_norm": 0.31403952836990356,
+      "learning_rate": 3.4643826292261178e-06,
+      "loss": 0.1745,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7859505074694948,
+      "grad_norm": 0.31387513875961304,
+      "learning_rate": 3.4626782034698074e-06,
+      "loss": 0.168,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7864066598243813,
+      "grad_norm": 0.31111714243888855,
+      "learning_rate": 3.460973252187321e-06,
+      "loss": 0.1743,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7868628121792679,
+      "grad_norm": 0.3043362498283386,
+      "learning_rate": 3.4592677763093907e-06,
+      "loss": 0.1734,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7873189645341544,
+      "grad_norm": 0.3131301999092102,
+      "learning_rate": 3.457561776767038e-06,
+      "loss": 0.1769,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7877751168890409,
+      "grad_norm": 0.312832772731781,
+      "learning_rate": 3.4558552544915674e-06,
+      "loss": 0.1759,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7882312692439275,
+      "grad_norm": 0.3067186176776886,
+      "learning_rate": 3.4541482104145695e-06,
+      "loss": 0.1645,
+      "step": 1728
+    },
+    {
+      "epoch": 0.788687421598814,
+      "grad_norm": 0.31117093563079834,
+      "learning_rate": 3.4524406454679206e-06,
+      "loss": 0.1714,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7891435739537005,
+      "grad_norm": 0.31211569905281067,
+      "learning_rate": 3.4507325605837787e-06,
+      "loss": 0.1848,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7895997263085871,
+      "grad_norm": 0.3020468056201935,
+      "learning_rate": 3.4490239566945903e-06,
+      "loss": 0.1773,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7900558786634736,
+      "grad_norm": 0.3162781596183777,
+      "learning_rate": 3.447314834733081e-06,
+      "loss": 0.1727,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7905120310183601,
+      "grad_norm": 0.31081804633140564,
+      "learning_rate": 3.4456051956322605e-06,
+      "loss": 0.1745,
+      "step": 1733
+    },
+    {
+      "epoch": 0.7909681833732467,
+      "grad_norm": 0.3408621549606323,
+      "learning_rate": 3.4438950403254224e-06,
+      "loss": 0.1877,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7914243357281332,
+      "grad_norm": 0.3137703835964203,
+      "learning_rate": 3.442184369746139e-06,
+      "loss": 0.1693,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7918804880830197,
+      "grad_norm": 0.32482755184173584,
+      "learning_rate": 3.440473184828266e-06,
+      "loss": 0.1698,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7923366404379063,
+      "grad_norm": 0.31918951869010925,
+      "learning_rate": 3.438761486505941e-06,
+      "loss": 0.1754,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7927927927927928,
+      "grad_norm": 0.3231935501098633,
+      "learning_rate": 3.43704927571358e-06,
+      "loss": 0.179,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7932489451476793,
+      "grad_norm": 0.30907079577445984,
+      "learning_rate": 3.435336553385877e-06,
+      "loss": 0.1755,
+      "step": 1739
+    },
+    {
+      "epoch": 0.7937050975025659,
+      "grad_norm": 0.32392293214797974,
+      "learning_rate": 3.433623320457809e-06,
+      "loss": 0.1864,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7941612498574524,
+      "grad_norm": 0.32874321937561035,
+      "learning_rate": 3.4319095778646305e-06,
+      "loss": 0.1894,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7946174022123389,
+      "grad_norm": 0.3175441324710846,
+      "learning_rate": 3.4301953265418736e-06,
+      "loss": 0.1691,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7950735545672255,
+      "grad_norm": 0.3188350796699524,
+      "learning_rate": 3.428480567425348e-06,
+      "loss": 0.1763,
+      "step": 1743
+    },
+    {
+      "epoch": 0.795529706922112,
+      "grad_norm": 0.32800957560539246,
+      "learning_rate": 3.4267653014511405e-06,
+      "loss": 0.1854,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7959858592769985,
+      "grad_norm": 0.3219795525074005,
+      "learning_rate": 3.4250495295556157e-06,
+      "loss": 0.1789,
+      "step": 1745
+    },
+    {
+      "epoch": 0.7964420116318851,
+      "grad_norm": 0.30824023485183716,
+      "learning_rate": 3.4233332526754137e-06,
+      "loss": 0.1781,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7968981639867716,
+      "grad_norm": 0.29734349250793457,
+      "learning_rate": 3.421616471747451e-06,
+      "loss": 0.1653,
+      "step": 1747
+    },
+    {
+      "epoch": 0.7973543163416581,
+      "grad_norm": 0.31022900342941284,
+      "learning_rate": 3.419899187708917e-06,
+      "loss": 0.1687,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7978104686965446,
+      "grad_norm": 0.3076399266719818,
+      "learning_rate": 3.418181401497278e-06,
+      "loss": 0.1751,
+      "step": 1749
+    },
+    {
+      "epoch": 0.7982666210514312,
+      "grad_norm": 0.32529518008232117,
+      "learning_rate": 3.416463114050274e-06,
+      "loss": 0.1707,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7987227734063177,
+      "grad_norm": 0.30859947204589844,
+      "learning_rate": 3.4147443263059176e-06,
+      "loss": 0.1797,
+      "step": 1751
+    },
+    {
+      "epoch": 0.7991789257612042,
+      "grad_norm": 0.3178499937057495,
+      "learning_rate": 3.4130250392024973e-06,
+      "loss": 0.174,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7996350781160908,
+      "grad_norm": 0.30828773975372314,
+      "learning_rate": 3.4113052536785705e-06,
+      "loss": 0.1736,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8000912304709773,
+      "grad_norm": 0.32288694381713867,
+      "learning_rate": 3.4095849706729684e-06,
+      "loss": 0.1872,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8005473828258638,
+      "grad_norm": 0.2964501976966858,
+      "learning_rate": 3.407864191124793e-06,
+      "loss": 0.1687,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8010035351807504,
+      "grad_norm": 0.31898024678230286,
+      "learning_rate": 3.4061429159734207e-06,
+      "loss": 0.1831,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8014596875356369,
+      "grad_norm": 0.3297187387943268,
+      "learning_rate": 3.404421146158494e-06,
+      "loss": 0.1829,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8019158398905234,
+      "grad_norm": 0.3069886565208435,
+      "learning_rate": 3.4026988826199273e-06,
+      "loss": 0.1753,
+      "step": 1758
+    },
+    {
+      "epoch": 0.80237199224541,
+      "grad_norm": 0.32255181670188904,
+      "learning_rate": 3.4009761262979046e-06,
+      "loss": 0.1821,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8028281446002965,
+      "grad_norm": 0.3090084195137024,
+      "learning_rate": 3.3992528781328793e-06,
+      "loss": 0.1766,
+      "step": 1760
+    },
+    {
+      "epoch": 0.803284296955183,
+      "grad_norm": 0.32624325156211853,
+      "learning_rate": 3.397529139065573e-06,
+      "loss": 0.1691,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8037404493100696,
+      "grad_norm": 0.30940142273902893,
+      "learning_rate": 3.395804910036975e-06,
+      "loss": 0.1727,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8041966016649561,
+      "grad_norm": 0.31408458948135376,
+      "learning_rate": 3.394080191988341e-06,
+      "loss": 0.1791,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8046527540198426,
+      "grad_norm": 0.307859867811203,
+      "learning_rate": 3.3923549858611958e-06,
+      "loss": 0.1772,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8051089063747292,
+      "grad_norm": 0.3106807768344879,
+      "learning_rate": 3.39062929259733e-06,
+      "loss": 0.1842,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8055650587296157,
+      "grad_norm": 0.30824077129364014,
+      "learning_rate": 3.3889031131387995e-06,
+      "loss": 0.1776,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8060212110845022,
+      "grad_norm": 0.33875778317451477,
+      "learning_rate": 3.3871764484279258e-06,
+      "loss": 0.1924,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8064773634393888,
+      "grad_norm": 0.3071323037147522,
+      "learning_rate": 3.385449299407296e-06,
+      "loss": 0.1821,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8069335157942753,
+      "grad_norm": 0.317123144865036,
+      "learning_rate": 3.38372166701976e-06,
+      "loss": 0.1775,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8073896681491618,
+      "grad_norm": 0.3307800889015198,
+      "learning_rate": 3.3819935522084322e-06,
+      "loss": 0.1655,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8078458205040484,
+      "grad_norm": 0.31006714701652527,
+      "learning_rate": 3.3802649559166926e-06,
+      "loss": 0.1766,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8083019728589349,
+      "grad_norm": 0.32217007875442505,
+      "learning_rate": 3.378535879088182e-06,
+      "loss": 0.1839,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8087581252138214,
+      "grad_norm": 0.305215060710907,
+      "learning_rate": 3.376806322666802e-06,
+      "loss": 0.1668,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8092142775687079,
+      "grad_norm": 0.30105552077293396,
+      "learning_rate": 3.375076287596718e-06,
+      "loss": 0.1705,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8096704299235945,
+      "grad_norm": 0.30478084087371826,
+      "learning_rate": 3.3733457748223582e-06,
+      "loss": 0.1794,
+      "step": 1775
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.30508729815483093,
+      "learning_rate": 3.3716147852884073e-06,
+      "loss": 0.1673,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8105827346333675,
+      "grad_norm": 0.31005942821502686,
+      "learning_rate": 3.369883319939815e-06,
+      "loss": 0.1712,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8110388869882541,
+      "grad_norm": 0.30497846007347107,
+      "learning_rate": 3.3681513797217874e-06,
+      "loss": 0.1656,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8114950393431406,
+      "grad_norm": 0.33867478370666504,
+      "learning_rate": 3.3664189655797912e-06,
+      "loss": 0.1768,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8119511916980271,
+      "grad_norm": 0.29837220907211304,
+      "learning_rate": 3.3646860784595512e-06,
+      "loss": 0.1711,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8124073440529137,
+      "grad_norm": 0.3795611560344696,
+      "learning_rate": 3.362952719307051e-06,
+      "loss": 0.1823,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8128634964078002,
+      "grad_norm": 0.31635069847106934,
+      "learning_rate": 3.3612188890685317e-06,
+      "loss": 0.1781,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8133196487626867,
+      "grad_norm": 0.30627480149269104,
+      "learning_rate": 3.3594845886904913e-06,
+      "loss": 0.1708,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8137758011175733,
+      "grad_norm": 0.3109096884727478,
+      "learning_rate": 3.357749819119685e-06,
+      "loss": 0.173,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8142319534724598,
+      "grad_norm": 0.3274926543235779,
+      "learning_rate": 3.3560145813031226e-06,
+      "loss": 0.1907,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8146881058273463,
+      "grad_norm": 0.2913208603858948,
+      "learning_rate": 3.3542788761880716e-06,
+      "loss": 0.1665,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8151442581822329,
+      "grad_norm": 0.3218519985675812,
+      "learning_rate": 3.352542704722055e-06,
+      "loss": 0.1824,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8156004105371194,
+      "grad_norm": 0.32326894998550415,
+      "learning_rate": 3.3508060678528464e-06,
+      "loss": 0.1825,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8160565628920059,
+      "grad_norm": 0.31490081548690796,
+      "learning_rate": 3.349068966528478e-06,
+      "loss": 0.1888,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8165127152468925,
+      "grad_norm": 0.3084615170955658,
+      "learning_rate": 3.347331401697233e-06,
+      "loss": 0.1702,
+      "step": 1790
+    },
+    {
+      "epoch": 0.816968867601779,
+      "grad_norm": 0.31111517548561096,
+      "learning_rate": 3.345593374307648e-06,
+      "loss": 0.1741,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8174250199566655,
+      "grad_norm": 0.31124627590179443,
+      "learning_rate": 3.3438548853085135e-06,
+      "loss": 0.1655,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8178811723115521,
+      "grad_norm": 0.29866474866867065,
+      "learning_rate": 3.3421159356488696e-06,
+      "loss": 0.1663,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8183373246664386,
+      "grad_norm": 0.3189578652381897,
+      "learning_rate": 3.34037652627801e-06,
+      "loss": 0.176,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8187934770213251,
+      "grad_norm": 0.29917868971824646,
+      "learning_rate": 3.3386366581454786e-06,
+      "loss": 0.1717,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8192496293762117,
+      "grad_norm": 0.3366638123989105,
+      "learning_rate": 3.3368963322010695e-06,
+      "loss": 0.175,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8197057817310982,
+      "grad_norm": 0.289279580116272,
+      "learning_rate": 3.335155549394826e-06,
+      "loss": 0.164,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8201619340859847,
+      "grad_norm": 0.31346961855888367,
+      "learning_rate": 3.3334143106770433e-06,
+      "loss": 0.1817,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8206180864408712,
+      "grad_norm": 0.309415340423584,
+      "learning_rate": 3.3316726169982635e-06,
+      "loss": 0.182,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8210742387957578,
+      "grad_norm": 0.31800204515457153,
+      "learning_rate": 3.329930469309276e-06,
+      "loss": 0.1675,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8215303911506443,
+      "grad_norm": 0.31752607226371765,
+      "learning_rate": 3.32818786856112e-06,
+      "loss": 0.1821,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8219865435055308,
+      "grad_norm": 0.30387353897094727,
+      "learning_rate": 3.3264448157050834e-06,
+      "loss": 0.1706,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8224426958604174,
+      "grad_norm": 0.3141171932220459,
+      "learning_rate": 3.3247013116926975e-06,
+      "loss": 0.1607,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8228988482153039,
+      "grad_norm": 0.3053392767906189,
+      "learning_rate": 3.322957357475741e-06,
+      "loss": 0.1719,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8233550005701904,
+      "grad_norm": 0.3272973895072937,
+      "learning_rate": 3.32121295400624e-06,
+      "loss": 0.1807,
+      "step": 1805
+    },
+    {
+      "epoch": 0.823811152925077,
+      "grad_norm": 0.33620065450668335,
+      "learning_rate": 3.3194681022364626e-06,
+      "loss": 0.1671,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8242673052799635,
+      "grad_norm": 0.3148181736469269,
+      "learning_rate": 3.3177228031189262e-06,
+      "loss": 0.1681,
+      "step": 1807
+    },
+    {
+      "epoch": 0.82472345763485,
+      "grad_norm": 0.32004332542419434,
+      "learning_rate": 3.315977057606388e-06,
+      "loss": 0.1829,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8251796099897366,
+      "grad_norm": 0.3128789961338043,
+      "learning_rate": 3.314230866651852e-06,
+      "loss": 0.186,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8256357623446231,
+      "grad_norm": 0.3105550706386566,
+      "learning_rate": 3.312484231208563e-06,
+      "loss": 0.1722,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8260919146995096,
+      "grad_norm": 0.3062818646430969,
+      "learning_rate": 3.31073715223001e-06,
+      "loss": 0.1709,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8265480670543962,
+      "grad_norm": 0.3230028450489044,
+      "learning_rate": 3.3089896306699233e-06,
+      "loss": 0.1711,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8270042194092827,
+      "grad_norm": 0.3506735563278198,
+      "learning_rate": 3.3072416674822768e-06,
+      "loss": 0.1835,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8274603717641692,
+      "grad_norm": 0.3038380444049835,
+      "learning_rate": 3.3054932636212815e-06,
+      "loss": 0.1709,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8279165241190558,
+      "grad_norm": 0.31784698367118835,
+      "learning_rate": 3.303744420041393e-06,
+      "loss": 0.184,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8283726764739423,
+      "grad_norm": 0.31457674503326416,
+      "learning_rate": 3.301995137697304e-06,
+      "loss": 0.1705,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8288288288288288,
+      "grad_norm": 0.3141183853149414,
+      "learning_rate": 3.30024541754395e-06,
+      "loss": 0.1721,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8292849811837154,
+      "grad_norm": 0.32975050806999207,
+      "learning_rate": 3.298495260536502e-06,
+      "loss": 0.1711,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8297411335386019,
+      "grad_norm": 0.3369347155094147,
+      "learning_rate": 3.29674466763037e-06,
+      "loss": 0.1812,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8301972858934884,
+      "grad_norm": 0.3278859555721283,
+      "learning_rate": 3.2949936397812055e-06,
+      "loss": 0.1791,
+      "step": 1820
+    },
+    {
+      "epoch": 0.830653438248375,
+      "grad_norm": 0.2986181974411011,
+      "learning_rate": 3.2932421779448933e-06,
+      "loss": 0.1717,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8311095906032615,
+      "grad_norm": 0.3137131333351135,
+      "learning_rate": 3.2914902830775558e-06,
+      "loss": 0.1801,
+      "step": 1822
+    },
+    {
+      "epoch": 0.831565742958148,
+      "grad_norm": 0.34094566106796265,
+      "learning_rate": 3.289737956135554e-06,
+      "loss": 0.1919,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8320218953130345,
+      "grad_norm": 0.3042123019695282,
+      "learning_rate": 3.287985198075484e-06,
+      "loss": 0.1736,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8324780476679211,
+      "grad_norm": 0.2937975525856018,
+      "learning_rate": 3.2862320098541755e-06,
+      "loss": 0.1619,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8329342000228076,
+      "grad_norm": 0.31467536091804504,
+      "learning_rate": 3.284478392428695e-06,
+      "loss": 0.1748,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8333903523776941,
+      "grad_norm": 0.3342815935611725,
+      "learning_rate": 3.2827243467563406e-06,
+      "loss": 0.1837,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8338465047325807,
+      "grad_norm": 0.305915892124176,
+      "learning_rate": 3.2809698737946494e-06,
+      "loss": 0.1801,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8343026570874672,
+      "grad_norm": 0.30489709973335266,
+      "learning_rate": 3.279214974501386e-06,
+      "loss": 0.1759,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8347588094423537,
+      "grad_norm": 0.331374853849411,
+      "learning_rate": 3.277459649834551e-06,
+      "loss": 0.1728,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8352149617972403,
+      "grad_norm": 0.3291572034358978,
+      "learning_rate": 3.275703900752376e-06,
+      "loss": 0.1782,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8356711141521268,
+      "grad_norm": 0.32215580344200134,
+      "learning_rate": 3.2739477282133253e-06,
+      "loss": 0.169,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8361272665070133,
+      "grad_norm": 0.3016672432422638,
+      "learning_rate": 3.2721911331760936e-06,
+      "loss": 0.1735,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8365834188618999,
+      "grad_norm": 0.304810494184494,
+      "learning_rate": 3.270434116599605e-06,
+      "loss": 0.1786,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8370395712167864,
+      "grad_norm": 0.3194803297519684,
+      "learning_rate": 3.2686766794430176e-06,
+      "loss": 0.1661,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8374957235716729,
+      "grad_norm": 0.3065241277217865,
+      "learning_rate": 3.266918822665715e-06,
+      "loss": 0.1641,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8379518759265595,
+      "grad_norm": 0.3140125274658203,
+      "learning_rate": 3.2651605472273113e-06,
+      "loss": 0.1801,
+      "step": 1837
+    },
+    {
+      "epoch": 0.838408028281446,
+      "grad_norm": 0.3100372850894928,
+      "learning_rate": 3.26340185408765e-06,
+      "loss": 0.1764,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8388641806363325,
+      "grad_norm": 0.33244743943214417,
+      "learning_rate": 3.2616427442068017e-06,
+      "loss": 0.1708,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8393203329912191,
+      "grad_norm": 0.3000289499759674,
+      "learning_rate": 3.259883218545065e-06,
+      "loss": 0.1651,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8397764853461056,
+      "grad_norm": 0.3246243894100189,
+      "learning_rate": 3.258123278062965e-06,
+      "loss": 0.1676,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8402326377009921,
+      "grad_norm": 0.32207009196281433,
+      "learning_rate": 3.2563629237212534e-06,
+      "loss": 0.1836,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8406887900558787,
+      "grad_norm": 0.32155367732048035,
+      "learning_rate": 3.2546021564809084e-06,
+      "loss": 0.1762,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8411449424107652,
+      "grad_norm": 0.3279334604740143,
+      "learning_rate": 3.2528409773031322e-06,
+      "loss": 0.174,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8416010947656517,
+      "grad_norm": 0.3021336793899536,
+      "learning_rate": 3.2510793871493535e-06,
+      "loss": 0.1796,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8420572471205383,
+      "grad_norm": 0.32889553904533386,
+      "learning_rate": 3.2493173869812243e-06,
+      "loss": 0.1829,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8425133994754248,
+      "grad_norm": 0.30873119831085205,
+      "learning_rate": 3.2475549777606213e-06,
+      "loss": 0.1647,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8429695518303113,
+      "grad_norm": 0.30581632256507874,
+      "learning_rate": 3.2457921604496435e-06,
+      "loss": 0.1717,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8434257041851979,
+      "grad_norm": 0.31490737199783325,
+      "learning_rate": 3.2440289360106126e-06,
+      "loss": 0.1838,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8438818565400844,
+      "grad_norm": 0.30432435870170593,
+      "learning_rate": 3.2422653054060745e-06,
+      "loss": 0.1763,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8443380088949709,
+      "grad_norm": 0.29845231771469116,
+      "learning_rate": 3.2405012695987943e-06,
+      "loss": 0.1704,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8447941612498574,
+      "grad_norm": 0.31056126952171326,
+      "learning_rate": 3.2387368295517586e-06,
+      "loss": 0.1675,
+      "step": 1852
+    },
+    {
+      "epoch": 0.845250313604744,
+      "grad_norm": 0.30088019371032715,
+      "learning_rate": 3.2369719862281775e-06,
+      "loss": 0.1694,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8457064659596305,
+      "grad_norm": 0.3124479353427887,
+      "learning_rate": 3.2352067405914783e-06,
+      "loss": 0.1787,
+      "step": 1854
+    },
+    {
+      "epoch": 0.846162618314517,
+      "grad_norm": 0.30317366123199463,
+      "learning_rate": 3.233441093605309e-06,
+      "loss": 0.1846,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8466187706694036,
+      "grad_norm": 0.3233534097671509,
+      "learning_rate": 3.231675046233536e-06,
+      "loss": 0.1802,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8470749230242901,
+      "grad_norm": 0.32793286442756653,
+      "learning_rate": 3.229908599440245e-06,
+      "loss": 0.1842,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8475310753791766,
+      "grad_norm": 0.3085339069366455,
+      "learning_rate": 3.228141754189741e-06,
+      "loss": 0.1775,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8479872277340632,
+      "grad_norm": 0.31056535243988037,
+      "learning_rate": 3.2263745114465428e-06,
+      "loss": 0.1782,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8484433800889497,
+      "grad_norm": 0.31079089641571045,
+      "learning_rate": 3.22460687217539e-06,
+      "loss": 0.1799,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8488995324438362,
+      "grad_norm": 0.32323333621025085,
+      "learning_rate": 3.2228388373412366e-06,
+      "loss": 0.1746,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8493556847987228,
+      "grad_norm": 0.30707842111587524,
+      "learning_rate": 3.221070407909253e-06,
+      "loss": 0.1767,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8498118371536093,
+      "grad_norm": 0.2940948009490967,
+      "learning_rate": 3.219301584844826e-06,
+      "loss": 0.1714,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8502679895084958,
+      "grad_norm": 0.295808345079422,
+      "learning_rate": 3.217532369113555e-06,
+      "loss": 0.1591,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8507241418633824,
+      "grad_norm": 0.2904025614261627,
+      "learning_rate": 3.215762761681256e-06,
+      "loss": 0.1668,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8511802942182689,
+      "grad_norm": 0.2962084412574768,
+      "learning_rate": 3.2139927635139586e-06,
+      "loss": 0.1638,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8516364465731554,
+      "grad_norm": 0.31741929054260254,
+      "learning_rate": 3.2122223755779025e-06,
+      "loss": 0.17,
+      "step": 1867
+    },
+    {
+      "epoch": 0.852092598928042,
+      "grad_norm": 0.3114364445209503,
+      "learning_rate": 3.2104515988395456e-06,
+      "loss": 0.1765,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8525487512829285,
+      "grad_norm": 0.3309611976146698,
+      "learning_rate": 3.2086804342655544e-06,
+      "loss": 0.1748,
+      "step": 1869
+    },
+    {
+      "epoch": 0.853004903637815,
+      "grad_norm": 0.30942338705062866,
+      "learning_rate": 3.206908882822807e-06,
+      "loss": 0.176,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8534610559927016,
+      "grad_norm": 0.3057447671890259,
+      "learning_rate": 3.2051369454783937e-06,
+      "loss": 0.168,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8539172083475881,
+      "grad_norm": 0.31072261929512024,
+      "learning_rate": 3.2033646231996167e-06,
+      "loss": 0.1829,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8543733607024746,
+      "grad_norm": 0.302334725856781,
+      "learning_rate": 3.2015919169539856e-06,
+      "loss": 0.1816,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8548295130573612,
+      "grad_norm": 0.30069172382354736,
+      "learning_rate": 3.1998188277092224e-06,
+      "loss": 0.17,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8552856654122477,
+      "grad_norm": 0.2981019914150238,
+      "learning_rate": 3.1980453564332547e-06,
+      "loss": 0.1717,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8557418177671342,
+      "grad_norm": 0.29103586077690125,
+      "learning_rate": 3.196271504094223e-06,
+      "loss": 0.1666,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8561979701220207,
+      "grad_norm": 0.34074217081069946,
+      "learning_rate": 3.1944972716604723e-06,
+      "loss": 0.1797,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8566541224769073,
+      "grad_norm": 0.3058969974517822,
+      "learning_rate": 3.1927226601005555e-06,
+      "loss": 0.1775,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8571102748317938,
+      "grad_norm": 0.3224133551120758,
+      "learning_rate": 3.1909476703832355e-06,
+      "loss": 0.1816,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8575664271866803,
+      "grad_norm": 0.3107955753803253,
+      "learning_rate": 3.189172303477478e-06,
+      "loss": 0.1641,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8580225795415669,
+      "grad_norm": 0.3106577694416046,
+      "learning_rate": 3.187396560352457e-06,
+      "loss": 0.1764,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8584787318964534,
+      "grad_norm": 0.30549386143684387,
+      "learning_rate": 3.1856204419775493e-06,
+      "loss": 0.171,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8589348842513399,
+      "grad_norm": 0.329221248626709,
+      "learning_rate": 3.1838439493223392e-06,
+      "loss": 0.1862,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8593910366062265,
+      "grad_norm": 0.3206157088279724,
+      "learning_rate": 3.182067083356616e-06,
+      "loss": 0.1776,
+      "step": 1884
+    },
+    {
+      "epoch": 0.859847188961113,
+      "grad_norm": 0.31115755438804626,
+      "learning_rate": 3.180289845050368e-06,
+      "loss": 0.1791,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8603033413159995,
+      "grad_norm": 0.3184621334075928,
+      "learning_rate": 3.178512235373791e-06,
+      "loss": 0.1786,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8607594936708861,
+      "grad_norm": 0.3181290328502655,
+      "learning_rate": 3.1767342552972843e-06,
+      "loss": 0.1708,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8612156460257726,
+      "grad_norm": 0.3009862005710602,
+      "learning_rate": 3.174955905791444e-06,
+      "loss": 0.1722,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8616717983806591,
+      "grad_norm": 0.329120010137558,
+      "learning_rate": 3.1731771878270746e-06,
+      "loss": 0.1747,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8621279507355457,
+      "grad_norm": 0.32426172494888306,
+      "learning_rate": 3.1713981023751767e-06,
+      "loss": 0.1734,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8625841030904322,
+      "grad_norm": 0.2980930805206299,
+      "learning_rate": 3.169618650406954e-06,
+      "loss": 0.1784,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8630402554453187,
+      "grad_norm": 0.3111628592014313,
+      "learning_rate": 3.1678388328938093e-06,
+      "loss": 0.1772,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8634964078002053,
+      "grad_norm": 0.3209223747253418,
+      "learning_rate": 3.166058650807345e-06,
+      "loss": 0.1743,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8639525601550918,
+      "grad_norm": 0.30618196725845337,
+      "learning_rate": 3.1642781051193626e-06,
+      "loss": 0.1716,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8644087125099783,
+      "grad_norm": 0.32629266381263733,
+      "learning_rate": 3.1624971968018634e-06,
+      "loss": 0.1701,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 0.3280145525932312,
+      "learning_rate": 3.1607159268270447e-06,
+      "loss": 0.1944,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8653210172197514,
+      "grad_norm": 0.32206812500953674,
+      "learning_rate": 3.1589342961673024e-06,
+      "loss": 0.1758,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8657771695746379,
+      "grad_norm": 0.31884250044822693,
+      "learning_rate": 3.157152305795228e-06,
+      "loss": 0.1789,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8662333219295245,
+      "grad_norm": 0.3230218291282654,
+      "learning_rate": 3.155369956683612e-06,
+      "loss": 0.1722,
+      "step": 1899
+    },
+    {
+      "epoch": 0.866689474284411,
+      "grad_norm": 0.31271427869796753,
+      "learning_rate": 3.153587249805438e-06,
+      "loss": 0.1746,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8671456266392975,
+      "grad_norm": 0.31306135654449463,
+      "learning_rate": 3.1518041861338856e-06,
+      "loss": 0.1691,
+      "step": 1901
+    },
+    {
+      "epoch": 0.867601778994184,
+      "grad_norm": 0.31754305958747864,
+      "learning_rate": 3.1500207666423306e-06,
+      "loss": 0.1667,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8680579313490706,
+      "grad_norm": 0.30968138575553894,
+      "learning_rate": 3.1482369923043403e-06,
+      "loss": 0.1768,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8685140837039571,
+      "grad_norm": 0.30352306365966797,
+      "learning_rate": 3.1464528640936797e-06,
+      "loss": 0.1708,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8689702360588436,
+      "grad_norm": 0.3019421398639679,
+      "learning_rate": 3.1446683829843027e-06,
+      "loss": 0.1711,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8694263884137302,
+      "grad_norm": 0.32632267475128174,
+      "learning_rate": 3.1428835499503586e-06,
+      "loss": 0.1815,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8698825407686167,
+      "grad_norm": 0.2981249690055847,
+      "learning_rate": 3.1410983659661882e-06,
+      "loss": 0.1823,
+      "step": 1907
+    },
+    {
+      "epoch": 0.8703386931235032,
+      "grad_norm": 0.3024674952030182,
+      "learning_rate": 3.139312832006323e-06,
+      "loss": 0.1768,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8707948454783898,
+      "grad_norm": 0.31647157669067383,
+      "learning_rate": 3.1375269490454864e-06,
+      "loss": 0.1763,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8712509978332763,
+      "grad_norm": 0.31125423312187195,
+      "learning_rate": 3.135740718058593e-06,
+      "loss": 0.1728,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8717071501881628,
+      "grad_norm": 0.29286032915115356,
+      "learning_rate": 3.1339541400207456e-06,
+      "loss": 0.1609,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8721633025430494,
+      "grad_norm": 0.3155742883682251,
+      "learning_rate": 3.132167215907238e-06,
+      "loss": 0.1717,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8726194548979359,
+      "grad_norm": 0.33474037051200867,
+      "learning_rate": 3.1303799466935523e-06,
+      "loss": 0.1812,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8730756072528224,
+      "grad_norm": 0.32887232303619385,
+      "learning_rate": 3.128592333355359e-06,
+      "loss": 0.1704,
+      "step": 1914
+    },
+    {
+      "epoch": 0.873531759607709,
+      "grad_norm": 0.33778512477874756,
+      "learning_rate": 3.1268043768685163e-06,
+      "loss": 0.1779,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8739879119625955,
+      "grad_norm": 0.3021858036518097,
+      "learning_rate": 3.12501607820907e-06,
+      "loss": 0.1695,
+      "step": 1916
+    },
+    {
+      "epoch": 0.874444064317482,
+      "grad_norm": 0.31008246541023254,
+      "learning_rate": 3.1232274383532528e-06,
+      "loss": 0.1745,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8749002166723686,
+      "grad_norm": 0.3025192618370056,
+      "learning_rate": 3.121438458277483e-06,
+      "loss": 0.1712,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8753563690272551,
+      "grad_norm": 0.3145863115787506,
+      "learning_rate": 3.1196491389583656e-06,
+      "loss": 0.1837,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8758125213821416,
+      "grad_norm": 0.31107470393180847,
+      "learning_rate": 3.11785948137269e-06,
+      "loss": 0.1775,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8762686737370282,
+      "grad_norm": 0.30264562368392944,
+      "learning_rate": 3.1160694864974304e-06,
+      "loss": 0.1654,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8767248260919147,
+      "grad_norm": 0.31219393014907837,
+      "learning_rate": 3.114279155309746e-06,
+      "loss": 0.1752,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8771809784468012,
+      "grad_norm": 0.30373796820640564,
+      "learning_rate": 3.112488488786978e-06,
+      "loss": 0.1697,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8776371308016878,
+      "grad_norm": 0.310825377702713,
+      "learning_rate": 3.1106974879066514e-06,
+      "loss": 0.1795,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8780932831565743,
+      "grad_norm": 0.30181989073753357,
+      "learning_rate": 3.108906153646475e-06,
+      "loss": 0.166,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8785494355114608,
+      "grad_norm": 0.3220551311969757,
+      "learning_rate": 3.107114486984338e-06,
+      "loss": 0.1773,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8790055878663474,
+      "grad_norm": 0.299573689699173,
+      "learning_rate": 3.10532248889831e-06,
+      "loss": 0.1647,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8794617402212339,
+      "grad_norm": 0.3220743238925934,
+      "learning_rate": 3.1035301603666456e-06,
+      "loss": 0.1778,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8799178925761204,
+      "grad_norm": 0.2995295226573944,
+      "learning_rate": 3.1017375023677755e-06,
+      "loss": 0.1705,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8803740449310069,
+      "grad_norm": 0.3305610418319702,
+      "learning_rate": 3.099944515880312e-06,
+      "loss": 0.177,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8808301972858935,
+      "grad_norm": 0.3195752203464508,
+      "learning_rate": 3.0981512018830473e-06,
+      "loss": 0.1727,
+      "step": 1931
+    },
+    {
+      "epoch": 0.88128634964078,
+      "grad_norm": 0.32316330075263977,
+      "learning_rate": 3.0963575613549523e-06,
+      "loss": 0.1718,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8817425019956665,
+      "grad_norm": 0.30436742305755615,
+      "learning_rate": 3.094563595275174e-06,
+      "loss": 0.1606,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8821986543505531,
+      "grad_norm": 0.3236144185066223,
+      "learning_rate": 3.09276930462304e-06,
+      "loss": 0.1761,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8826548067054396,
+      "grad_norm": 0.304534375667572,
+      "learning_rate": 3.090974690378053e-06,
+      "loss": 0.1706,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8831109590603261,
+      "grad_norm": 0.3272643983364105,
+      "learning_rate": 3.089179753519894e-06,
+      "loss": 0.187,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8835671114152127,
+      "grad_norm": 0.326433926820755,
+      "learning_rate": 3.0873844950284193e-06,
+      "loss": 0.182,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8840232637700992,
+      "grad_norm": 0.3116316795349121,
+      "learning_rate": 3.0855889158836598e-06,
+      "loss": 0.1785,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8844794161249857,
+      "grad_norm": 0.3136686384677887,
+      "learning_rate": 3.083793017065823e-06,
+      "loss": 0.1731,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8849355684798723,
+      "grad_norm": 0.3122534155845642,
+      "learning_rate": 3.0819967995552913e-06,
+      "loss": 0.1698,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8853917208347588,
+      "grad_norm": 0.3312046229839325,
+      "learning_rate": 3.080200264332619e-06,
+      "loss": 0.1809,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8858478731896453,
+      "grad_norm": 0.31901177763938904,
+      "learning_rate": 3.0784034123785345e-06,
+      "loss": 0.1664,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8863040255445319,
+      "grad_norm": 0.2952611744403839,
+      "learning_rate": 3.076606244673941e-06,
+      "loss": 0.1715,
+      "step": 1943
+    },
+    {
+      "epoch": 0.8867601778994184,
+      "grad_norm": 0.3115205764770508,
+      "learning_rate": 3.074808762199911e-06,
+      "loss": 0.1837,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8872163302543049,
+      "grad_norm": 0.3021535277366638,
+      "learning_rate": 3.0730109659376916e-06,
+      "loss": 0.1707,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8876724826091915,
+      "grad_norm": 0.3121006190776825,
+      "learning_rate": 3.0712128568687e-06,
+      "loss": 0.1674,
+      "step": 1946
+    },
+    {
+      "epoch": 0.888128634964078,
+      "grad_norm": 0.31935596466064453,
+      "learning_rate": 3.069414435974524e-06,
+      "loss": 0.1715,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8885847873189645,
+      "grad_norm": 0.33563101291656494,
+      "learning_rate": 3.0676157042369213e-06,
+      "loss": 0.1849,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8890409396738511,
+      "grad_norm": 0.3145502805709839,
+      "learning_rate": 3.0658166626378205e-06,
+      "loss": 0.1802,
+      "step": 1949
+    },
+    {
+      "epoch": 0.8894970920287376,
+      "grad_norm": 0.35879671573638916,
+      "learning_rate": 3.0640173121593188e-06,
+      "loss": 0.1967,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8899532443836241,
+      "grad_norm": 0.3013881742954254,
+      "learning_rate": 3.0622176537836813e-06,
+      "loss": 0.1727,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8904093967385107,
+      "grad_norm": 0.30106544494628906,
+      "learning_rate": 3.0604176884933422e-06,
+      "loss": 0.1812,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8908655490933972,
+      "grad_norm": 0.3225567936897278,
+      "learning_rate": 3.058617417270902e-06,
+      "loss": 0.1754,
+      "step": 1953
+    },
+    {
+      "epoch": 0.8913217014482837,
+      "grad_norm": 0.29570358991622925,
+      "learning_rate": 3.0568168410991305e-06,
+      "loss": 0.1733,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8917778538031702,
+      "grad_norm": 0.3089274764060974,
+      "learning_rate": 3.0550159609609613e-06,
+      "loss": 0.1818,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8922340061580568,
+      "grad_norm": 0.2963835895061493,
+      "learning_rate": 3.053214777839496e-06,
+      "loss": 0.1686,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8926901585129433,
+      "grad_norm": 0.29800137877464294,
+      "learning_rate": 3.0514132927180002e-06,
+      "loss": 0.1708,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8931463108678298,
+      "grad_norm": 0.31042057275772095,
+      "learning_rate": 3.0496115065799046e-06,
+      "loss": 0.1757,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8936024632227164,
+      "grad_norm": 0.29591378569602966,
+      "learning_rate": 3.047809420408806e-06,
+      "loss": 0.1696,
+      "step": 1959
+    },
+    {
+      "epoch": 0.8940586155776029,
+      "grad_norm": 0.32200944423675537,
+      "learning_rate": 3.0460070351884614e-06,
+      "loss": 0.1736,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8945147679324894,
+      "grad_norm": 0.3139008581638336,
+      "learning_rate": 3.0442043519027938e-06,
+      "loss": 0.1742,
+      "step": 1961
+    },
+    {
+      "epoch": 0.894970920287376,
+      "grad_norm": 0.33181479573249817,
+      "learning_rate": 3.0424013715358897e-06,
+      "loss": 0.1665,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8954270726422625,
+      "grad_norm": 0.30871087312698364,
+      "learning_rate": 3.040598095071995e-06,
+      "loss": 0.1754,
+      "step": 1963
+    },
+    {
+      "epoch": 0.895883224997149,
+      "grad_norm": 0.31437596678733826,
+      "learning_rate": 3.0387945234955187e-06,
+      "loss": 0.1711,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8963393773520356,
+      "grad_norm": 0.31460368633270264,
+      "learning_rate": 3.0369906577910307e-06,
+      "loss": 0.1647,
+      "step": 1965
+    },
+    {
+      "epoch": 0.8967955297069221,
+      "grad_norm": 0.31222400069236755,
+      "learning_rate": 3.0351864989432624e-06,
+      "loss": 0.1698,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8972516820618086,
+      "grad_norm": 0.3121219277381897,
+      "learning_rate": 3.033382047937104e-06,
+      "loss": 0.1811,
+      "step": 1967
+    },
+    {
+      "epoch": 0.8977078344166952,
+      "grad_norm": 0.30796417593955994,
+      "learning_rate": 3.031577305757605e-06,
+      "loss": 0.1656,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8981639867715817,
+      "grad_norm": 0.33585023880004883,
+      "learning_rate": 3.0297722733899755e-06,
+      "loss": 0.1708,
+      "step": 1969
+    },
+    {
+      "epoch": 0.8986201391264682,
+      "grad_norm": 0.322227418422699,
+      "learning_rate": 3.0279669518195807e-06,
+      "loss": 0.1777,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8990762914813548,
+      "grad_norm": 0.32161715626716614,
+      "learning_rate": 3.026161342031949e-06,
+      "loss": 0.1736,
+      "step": 1971
+    },
+    {
+      "epoch": 0.8995324438362413,
+      "grad_norm": 0.3084839880466461,
+      "learning_rate": 3.024355445012761e-06,
+      "loss": 0.1695,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8999885961911278,
+      "grad_norm": 0.32000112533569336,
+      "learning_rate": 3.0225492617478574e-06,
+      "loss": 0.1743,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9004447485460144,
+      "grad_norm": 0.3198784589767456,
+      "learning_rate": 3.0207427932232333e-06,
+      "loss": 0.1846,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9009009009009009,
+      "grad_norm": 0.33122551441192627,
+      "learning_rate": 3.018936040425039e-06,
+      "loss": 0.1694,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9013570532557874,
+      "grad_norm": 0.30618634819984436,
+      "learning_rate": 3.0171290043395823e-06,
+      "loss": 0.1819,
+      "step": 1976
+    },
+    {
+      "epoch": 0.901813205610674,
+      "grad_norm": 0.34121984243392944,
+      "learning_rate": 3.0153216859533254e-06,
+      "loss": 0.1785,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9022693579655605,
+      "grad_norm": 0.30839937925338745,
+      "learning_rate": 3.013514086252882e-06,
+      "loss": 0.1822,
+      "step": 1978
+    },
+    {
+      "epoch": 0.902725510320447,
+      "grad_norm": 0.313555508852005,
+      "learning_rate": 3.0117062062250213e-06,
+      "loss": 0.1733,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9031816626753335,
+      "grad_norm": 0.3175600469112396,
+      "learning_rate": 3.0098980468566663e-06,
+      "loss": 0.1774,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9036378150302201,
+      "grad_norm": 0.3060867488384247,
+      "learning_rate": 3.008089609134891e-06,
+      "loss": 0.1698,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9040939673851066,
+      "grad_norm": 0.3148481845855713,
+      "learning_rate": 3.0062808940469212e-06,
+      "loss": 0.1781,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9045501197399931,
+      "grad_norm": 0.32034122943878174,
+      "learning_rate": 3.004471902580135e-06,
+      "loss": 0.1894,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9050062720948797,
+      "grad_norm": 0.3200012743473053,
+      "learning_rate": 3.0026626357220623e-06,
+      "loss": 0.177,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9054624244497662,
+      "grad_norm": 0.3045053482055664,
+      "learning_rate": 3.0008530944603804e-06,
+      "loss": 0.1634,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9059185768046527,
+      "grad_norm": 0.3209184408187866,
+      "learning_rate": 2.9990432797829193e-06,
+      "loss": 0.1769,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9063747291595393,
+      "grad_norm": 0.3025554120540619,
+      "learning_rate": 2.997233192677656e-06,
+      "loss": 0.1696,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9068308815144258,
+      "grad_norm": 0.3227944076061249,
+      "learning_rate": 2.9954228341327192e-06,
+      "loss": 0.1826,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9072870338693123,
+      "grad_norm": 0.3202097713947296,
+      "learning_rate": 2.9936122051363818e-06,
+      "loss": 0.1784,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9077431862241989,
+      "grad_norm": 0.30975785851478577,
+      "learning_rate": 2.991801306677068e-06,
+      "loss": 0.1654,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9081993385790854,
+      "grad_norm": 0.3029954433441162,
+      "learning_rate": 2.989990139743346e-06,
+      "loss": 0.1684,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9086554909339719,
+      "grad_norm": 0.3254508972167969,
+      "learning_rate": 2.988178705323934e-06,
+      "loss": 0.1842,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9091116432888585,
+      "grad_norm": 0.2830461263656616,
+      "learning_rate": 2.9863670044076935e-06,
+      "loss": 0.1635,
+      "step": 1993
+    },
+    {
+      "epoch": 0.909567795643745,
+      "grad_norm": 0.3307308256626129,
+      "learning_rate": 2.9845550379836314e-06,
+      "loss": 0.1743,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9100239479986315,
+      "grad_norm": 0.31417912244796753,
+      "learning_rate": 2.9827428070409013e-06,
+      "loss": 0.1837,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9104801003535181,
+      "grad_norm": 0.3405453860759735,
+      "learning_rate": 2.9809303125688004e-06,
+      "loss": 0.1746,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9109362527084046,
+      "grad_norm": 0.31599441170692444,
+      "learning_rate": 2.9791175555567702e-06,
+      "loss": 0.1778,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.32422518730163574,
+      "learning_rate": 2.9773045369943936e-06,
+      "loss": 0.1756,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9118485574181777,
+      "grad_norm": 0.3074676990509033,
+      "learning_rate": 2.975491257871399e-06,
+      "loss": 0.1761,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9123047097730642,
+      "grad_norm": 0.2959085702896118,
+      "learning_rate": 2.9736777191776543e-06,
+      "loss": 0.1749,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4384,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.521897993857663e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}