diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,30730 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9997719238225566,
+  "eval_steps": 500,
+  "global_step": 4384,
+  "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
+    },
+    {
+      "epoch": 0.9127608621279507,
+      "grad_norm": 0.315239816904068,
+      "learning_rate": 2.9718639219031724e-06,
+      "loss": 0.1744,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9132170144828373,
+      "grad_norm": 0.3079422414302826,
+      "learning_rate": 2.970049867038104e-06,
+      "loss": 0.1676,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9136731668377238,
+      "grad_norm": 0.31476151943206787,
+      "learning_rate": 2.968235555572743e-06,
+      "loss": 0.1752,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9141293191926103,
+      "grad_norm": 0.29994356632232666,
+      "learning_rate": 2.966420988497522e-06,
+      "loss": 0.1736,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9145854715474968,
+      "grad_norm": 0.31020399928092957,
+      "learning_rate": 2.9646061668030123e-06,
+      "loss": 0.1766,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9150416239023834,
+      "grad_norm": 0.310195654630661,
+      "learning_rate": 2.962791091479927e-06,
+      "loss": 0.1774,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9154977762572699,
+      "grad_norm": 0.3052014410495758,
+      "learning_rate": 2.9609757635191155e-06,
+      "loss": 0.1699,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9159539286121564,
+      "grad_norm": 0.30509743094444275,
+      "learning_rate": 2.959160183911565e-06,
+      "loss": 0.179,
+      "step": 2008
+    },
+    {
+      "epoch": 0.916410080967043,
+      "grad_norm": 0.2921556532382965,
+      "learning_rate": 2.957344353648401e-06,
+      "loss": 0.1674,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9168662333219295,
+      "grad_norm": 0.3185284435749054,
+      "learning_rate": 2.955528273720886e-06,
+      "loss": 0.1708,
+      "step": 2010
+    },
+    {
+      "epoch": 0.917322385676816,
+      "grad_norm": 0.3152352273464203,
+      "learning_rate": 2.953711945120417e-06,
+      "loss": 0.1696,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9177785380317026,
+      "grad_norm": 0.3126958906650543,
+      "learning_rate": 2.9518953688385298e-06,
+      "loss": 0.1752,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9182346903865891,
+      "grad_norm": 0.3017788529396057,
+      "learning_rate": 2.950078545866892e-06,
+      "loss": 0.1733,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9186908427414756,
+      "grad_norm": 0.30587199330329895,
+      "learning_rate": 2.948261477197309e-06,
+      "loss": 0.167,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9191469950963622,
+      "grad_norm": 0.3012850284576416,
+      "learning_rate": 2.946444163821718e-06,
+      "loss": 0.1696,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9196031474512487,
+      "grad_norm": 0.32750070095062256,
+      "learning_rate": 2.9446266067321904e-06,
+      "loss": 0.1801,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9200592998061352,
+      "grad_norm": 0.30671969056129456,
+      "learning_rate": 2.9428088069209315e-06,
+      "loss": 0.1742,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9205154521610218,
+      "grad_norm": 0.3050041198730469,
+      "learning_rate": 2.9409907653802794e-06,
+      "loss": 0.1699,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9209716045159083,
+      "grad_norm": 0.31370341777801514,
+      "learning_rate": 2.9391724831027015e-06,
+      "loss": 0.175,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9214277568707948,
+      "grad_norm": 0.3043113052845001,
+      "learning_rate": 2.9373539610807983e-06,
+      "loss": 0.1765,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9218839092256814,
+      "grad_norm": 0.33342990279197693,
+      "learning_rate": 2.935535200307302e-06,
+      "loss": 0.1733,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9223400615805679,
+      "grad_norm": 0.29968947172164917,
+      "learning_rate": 2.933716201775075e-06,
+      "loss": 0.1731,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9227962139354544,
+      "grad_norm": 0.31365442276000977,
+      "learning_rate": 2.931896966477109e-06,
+      "loss": 0.1813,
+      "step": 2023
+    },
+    {
+      "epoch": 0.923252366290341,
+      "grad_norm": 0.3206329047679901,
+      "learning_rate": 2.930077495406523e-06,
+      "loss": 0.1815,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9237085186452275,
+      "grad_norm": 0.30596745014190674,
+      "learning_rate": 2.9282577895565676e-06,
+      "loss": 0.1709,
+      "step": 2025
+    },
+    {
+      "epoch": 0.924164671000114,
+      "grad_norm": 0.30679023265838623,
+      "learning_rate": 2.9264378499206207e-06,
+      "loss": 0.174,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9246208233550006,
+      "grad_norm": 0.30678436160087585,
+      "learning_rate": 2.9246176774921875e-06,
+      "loss": 0.169,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9250769757098871,
+      "grad_norm": 0.3298848867416382,
+      "learning_rate": 2.9227972732649e-06,
+      "loss": 0.1791,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9255331280647736,
+      "grad_norm": 0.3009489178657532,
+      "learning_rate": 2.9209766382325176e-06,
+      "loss": 0.1665,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9259892804196602,
+      "grad_norm": 0.31149813532829285,
+      "learning_rate": 2.9191557733889257e-06,
+      "loss": 0.1751,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9264454327745467,
+      "grad_norm": 0.2840937077999115,
+      "learning_rate": 2.9173346797281334e-06,
+      "loss": 0.1613,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9269015851294332,
+      "grad_norm": 0.28505340218544006,
+      "learning_rate": 2.915513358244276e-06,
+      "loss": 0.1713,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9273577374843197,
+      "grad_norm": 0.32732346653938293,
+      "learning_rate": 2.913691809931615e-06,
+      "loss": 0.1894,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9278138898392063,
+      "grad_norm": 0.33758285641670227,
+      "learning_rate": 2.911870035784533e-06,
+      "loss": 0.1868,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9282700421940928,
+      "grad_norm": 0.31489428877830505,
+      "learning_rate": 2.9100480367975364e-06,
+      "loss": 0.1764,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9287261945489793,
+      "grad_norm": 0.3046037256717682,
+      "learning_rate": 2.9082258139652536e-06,
+      "loss": 0.1682,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9291823469038659,
+      "grad_norm": 0.3068292737007141,
+      "learning_rate": 2.9064033682824387e-06,
+      "loss": 0.1597,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9296384992587524,
+      "grad_norm": 0.31091052293777466,
+      "learning_rate": 2.9045807007439634e-06,
+      "loss": 0.172,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9300946516136389,
+      "grad_norm": 0.3191123902797699,
+      "learning_rate": 2.902757812344823e-06,
+      "loss": 0.1832,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9305508039685255,
+      "grad_norm": 0.3178643584251404,
+      "learning_rate": 2.900934704080133e-06,
+      "loss": 0.1602,
+      "step": 2040
+    },
+    {
+      "epoch": 0.931006956323412,
+      "grad_norm": 0.3080301880836487,
+      "learning_rate": 2.8991113769451262e-06,
+      "loss": 0.1774,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9314631086782985,
+      "grad_norm": 0.3150477409362793,
+      "learning_rate": 2.89728783193516e-06,
+      "loss": 0.1726,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9319192610331851,
+      "grad_norm": 0.32018113136291504,
+      "learning_rate": 2.895464070045706e-06,
+      "loss": 0.1781,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9323754133880716,
+      "grad_norm": 0.31867480278015137,
+      "learning_rate": 2.893640092272357e-06,
+      "loss": 0.1682,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9328315657429581,
+      "grad_norm": 0.31242507696151733,
+      "learning_rate": 2.8918158996108227e-06,
+      "loss": 0.1803,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9332877180978447,
+      "grad_norm": 0.28953930735588074,
+      "learning_rate": 2.889991493056929e-06,
+      "loss": 0.1553,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9337438704527312,
+      "grad_norm": 0.313681036233902,
+      "learning_rate": 2.8881668736066206e-06,
+      "loss": 0.1803,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9342000228076177,
+      "grad_norm": 0.3214857578277588,
+      "learning_rate": 2.8863420422559577e-06,
+      "loss": 0.1834,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9346561751625043,
+      "grad_norm": 0.29927003383636475,
+      "learning_rate": 2.8845170000011167e-06,
+      "loss": 0.1763,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9351123275173908,
+      "grad_norm": 0.31733405590057373,
+      "learning_rate": 2.8826917478383875e-06,
+      "loss": 0.168,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9355684798722773,
+      "grad_norm": 0.3365289568901062,
+      "learning_rate": 2.8808662867641746e-06,
+      "loss": 0.1878,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9360246322271639,
+      "grad_norm": 0.3141789734363556,
+      "learning_rate": 2.8790406177749985e-06,
+      "loss": 0.1717,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9364807845820504,
+      "grad_norm": 0.30329155921936035,
+      "learning_rate": 2.877214741867493e-06,
+      "loss": 0.1744,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9369369369369369,
+      "grad_norm": 0.3056482672691345,
+      "learning_rate": 2.8753886600384028e-06,
+      "loss": 0.1739,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9373930892918235,
+      "grad_norm": 0.30756711959838867,
+      "learning_rate": 2.873562373284587e-06,
+      "loss": 0.1696,
+      "step": 2055
+    },
+    {
+      "epoch": 0.93784924164671,
+      "grad_norm": 0.30760154128074646,
+      "learning_rate": 2.8717358826030158e-06,
+      "loss": 0.1733,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9383053940015965,
+      "grad_norm": 0.30770784616470337,
+      "learning_rate": 2.8699091889907695e-06,
+      "loss": 0.171,
+      "step": 2057
+    },
+    {
+      "epoch": 0.938761546356483,
+      "grad_norm": 0.30789944529533386,
+      "learning_rate": 2.868082293445041e-06,
+      "loss": 0.1757,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9392176987113696,
+      "grad_norm": 0.3200453817844391,
+      "learning_rate": 2.8662551969631337e-06,
+      "loss": 0.168,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9396738510662561,
+      "grad_norm": 0.31140875816345215,
+      "learning_rate": 2.86442790054246e-06,
+      "loss": 0.1654,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9401300034211426,
+      "grad_norm": 0.3055671751499176,
+      "learning_rate": 2.86260040518054e-06,
+      "loss": 0.1708,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9405861557760292,
+      "grad_norm": 0.312163382768631,
+      "learning_rate": 2.8607727118750032e-06,
+      "loss": 0.1703,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9410423081309157,
+      "grad_norm": 0.3036530613899231,
+      "learning_rate": 2.85894482162359e-06,
+      "loss": 0.1691,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9414984604858022,
+      "grad_norm": 0.30906611680984497,
+      "learning_rate": 2.8571167354241445e-06,
+      "loss": 0.1714,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9419546128406888,
+      "grad_norm": 0.3080875277519226,
+      "learning_rate": 2.85528845427462e-06,
+      "loss": 0.1648,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9424107651955753,
+      "grad_norm": 0.30116966366767883,
+      "learning_rate": 2.853459979173074e-06,
+      "loss": 0.1744,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9428669175504618,
+      "grad_norm": 0.30044636130332947,
+      "learning_rate": 2.851631311117672e-06,
+      "loss": 0.1575,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9433230699053484,
+      "grad_norm": 0.32075974345207214,
+      "learning_rate": 2.849802451106685e-06,
+      "loss": 0.1753,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9437792222602349,
+      "grad_norm": 0.3079969584941864,
+      "learning_rate": 2.8479734001384875e-06,
+      "loss": 0.174,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9442353746151214,
+      "grad_norm": 0.2999923527240753,
+      "learning_rate": 2.846144159211558e-06,
+      "loss": 0.1714,
+      "step": 2070
+    },
+    {
+      "epoch": 0.944691526970008,
+      "grad_norm": 0.305400013923645,
+      "learning_rate": 2.8443147293244805e-06,
+      "loss": 0.1738,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9451476793248945,
+      "grad_norm": 0.3257659375667572,
+      "learning_rate": 2.84248511147594e-06,
+      "loss": 0.1757,
+      "step": 2072
+    },
+    {
+      "epoch": 0.945603831679781,
+      "grad_norm": 0.2996188998222351,
+      "learning_rate": 2.8406553066647257e-06,
+      "loss": 0.169,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9460599840346676,
+      "grad_norm": 0.3152143359184265,
+      "learning_rate": 2.8388253158897283e-06,
+      "loss": 0.1793,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9465161363895541,
+      "grad_norm": 0.30808013677597046,
+      "learning_rate": 2.8369951401499408e-06,
+      "loss": 0.1721,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9469722887444406,
+      "grad_norm": 0.3054342269897461,
+      "learning_rate": 2.835164780444455e-06,
+      "loss": 0.172,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9474284410993272,
+      "grad_norm": 0.32644006609916687,
+      "learning_rate": 2.8333342377724644e-06,
+      "loss": 0.1812,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9478845934542137,
+      "grad_norm": 0.3102589249610901,
+      "learning_rate": 2.831503513133264e-06,
+      "loss": 0.1647,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9483407458091002,
+      "grad_norm": 0.3049914240837097,
+      "learning_rate": 2.8296726075262464e-06,
+      "loss": 0.1671,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9487968981639868,
+      "grad_norm": 0.3153834342956543,
+      "learning_rate": 2.8278415219509025e-06,
+      "loss": 0.171,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9492530505188733,
+      "grad_norm": 0.3235340118408203,
+      "learning_rate": 2.8260102574068223e-06,
+      "loss": 0.173,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9497092028737598,
+      "grad_norm": 0.30418190360069275,
+      "learning_rate": 2.8241788148936935e-06,
+      "loss": 0.1705,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9501653552286463,
+      "grad_norm": 0.32771942019462585,
+      "learning_rate": 2.822347195411301e-06,
+      "loss": 0.1814,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9506215075835329,
+      "grad_norm": 0.31460103392601013,
+      "learning_rate": 2.8205153999595253e-06,
+      "loss": 0.17,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9510776599384194,
+      "grad_norm": 0.3007590174674988,
+      "learning_rate": 2.818683429538346e-06,
+      "loss": 0.1679,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9515338122933059,
+      "grad_norm": 0.3325331211090088,
+      "learning_rate": 2.8168512851478334e-06,
+      "loss": 0.1779,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9519899646481925,
+      "grad_norm": 0.3317030966281891,
+      "learning_rate": 2.8150189677881574e-06,
+      "loss": 0.185,
+      "step": 2087
+    },
+    {
+      "epoch": 0.952446117003079,
+      "grad_norm": 0.30043888092041016,
+      "learning_rate": 2.8131864784595788e-06,
+      "loss": 0.1702,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9529022693579655,
+      "grad_norm": 0.3076836168766022,
+      "learning_rate": 2.811353818162455e-06,
+      "loss": 0.1681,
+      "step": 2089
+    },
+    {
+      "epoch": 0.9533584217128521,
+      "grad_norm": 0.2904086112976074,
+      "learning_rate": 2.8095209878972353e-06,
+      "loss": 0.1671,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9538145740677386,
+      "grad_norm": 0.31719329953193665,
+      "learning_rate": 2.8076879886644616e-06,
+      "loss": 0.1704,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9542707264226251,
+      "grad_norm": 0.29341357946395874,
+      "learning_rate": 2.8058548214647674e-06,
+      "loss": 0.168,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9547268787775117,
+      "grad_norm": 0.2954353392124176,
+      "learning_rate": 2.8040214872988814e-06,
+      "loss": 0.1626,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9551830311323982,
+      "grad_norm": 0.31150898337364197,
+      "learning_rate": 2.80218798716762e-06,
+      "loss": 0.1713,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9556391834872847,
+      "grad_norm": 0.3197208046913147,
+      "learning_rate": 2.800354322071891e-06,
+      "loss": 0.179,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9560953358421713,
+      "grad_norm": 0.32801467180252075,
+      "learning_rate": 2.798520493012691e-06,
+      "loss": 0.1739,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9565514881970578,
+      "grad_norm": 0.3069617748260498,
+      "learning_rate": 2.7966865009911103e-06,
+      "loss": 0.1734,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9570076405519443,
+      "grad_norm": 0.30003073811531067,
+      "learning_rate": 2.794852347008324e-06,
+      "loss": 0.1789,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9574637929068309,
+      "grad_norm": 0.3202616572380066,
+      "learning_rate": 2.793018032065597e-06,
+      "loss": 0.1689,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9579199452617174,
+      "grad_norm": 0.29878199100494385,
+      "learning_rate": 2.7911835571642816e-06,
+      "loss": 0.1624,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9583760976166039,
+      "grad_norm": 0.29897353053092957,
+      "learning_rate": 2.789348923305819e-06,
+      "loss": 0.1713,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9588322499714905,
+      "grad_norm": 0.3061663508415222,
+      "learning_rate": 2.7875141314917366e-06,
+      "loss": 0.1805,
+      "step": 2102
+    },
+    {
+      "epoch": 0.959288402326377,
+      "grad_norm": 0.3218860328197479,
+      "learning_rate": 2.7856791827236447e-06,
+      "loss": 0.1845,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9597445546812635,
+      "grad_norm": 0.3210839629173279,
+      "learning_rate": 2.783844078003245e-06,
+      "loss": 0.1814,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9602007070361501,
+      "grad_norm": 0.3209848701953888,
+      "learning_rate": 2.782008818332321e-06,
+      "loss": 0.1801,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9606568593910366,
+      "grad_norm": 0.3130209147930145,
+      "learning_rate": 2.7801734047127403e-06,
+      "loss": 0.173,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9611130117459231,
+      "grad_norm": 0.3018120527267456,
+      "learning_rate": 2.7783378381464554e-06,
+      "loss": 0.1577,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9615691641008097,
+      "grad_norm": 0.33262181282043457,
+      "learning_rate": 2.7765021196355023e-06,
+      "loss": 0.1824,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9620253164556962,
+      "grad_norm": 0.31890246272087097,
+      "learning_rate": 2.774666250182002e-06,
+      "loss": 0.1775,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9624814688105827,
+      "grad_norm": 0.294516384601593,
+      "learning_rate": 2.7728302307881537e-06,
+      "loss": 0.1714,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9629376211654692,
+      "grad_norm": 0.30404695868492126,
+      "learning_rate": 2.770994062456241e-06,
+      "loss": 0.1769,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9633937735203558,
+      "grad_norm": 0.3051758408546448,
+      "learning_rate": 2.76915774618863e-06,
+      "loss": 0.1785,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9638499258752423,
+      "grad_norm": 0.2989727854728699,
+      "learning_rate": 2.7673212829877633e-06,
+      "loss": 0.1706,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9643060782301288,
+      "grad_norm": 0.330750972032547,
+      "learning_rate": 2.7654846738561693e-06,
+      "loss": 0.1945,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9647622305850154,
+      "grad_norm": 0.3040163815021515,
+      "learning_rate": 2.763647919796451e-06,
+      "loss": 0.1768,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9652183829399019,
+      "grad_norm": 0.31833416223526,
+      "learning_rate": 2.761811021811295e-06,
+      "loss": 0.1794,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9656745352947884,
+      "grad_norm": 0.3233703672885895,
+      "learning_rate": 2.7599739809034613e-06,
+      "loss": 0.1811,
+      "step": 2117
+    },
+    {
+      "epoch": 0.966130687649675,
+      "grad_norm": 0.30766183137893677,
+      "learning_rate": 2.7581367980757923e-06,
+      "loss": 0.1679,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9665868400045615,
+      "grad_norm": 0.31805744767189026,
+      "learning_rate": 2.756299474331208e-06,
+      "loss": 0.1822,
+      "step": 2119
+    },
+    {
+      "epoch": 0.967042992359448,
+      "grad_norm": 0.31153884530067444,
+      "learning_rate": 2.754462010672701e-06,
+      "loss": 0.1708,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9674991447143346,
+      "grad_norm": 0.29028862714767456,
+      "learning_rate": 2.752624408103345e-06,
+      "loss": 0.1745,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9679552970692211,
+      "grad_norm": 0.3012767434120178,
+      "learning_rate": 2.750786667626285e-06,
+      "loss": 0.1769,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9684114494241076,
+      "grad_norm": 0.3117278814315796,
+      "learning_rate": 2.7489487902447466e-06,
+      "loss": 0.1731,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9688676017789942,
+      "grad_norm": 0.29669642448425293,
+      "learning_rate": 2.7471107769620258e-06,
+      "loss": 0.1656,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9693237541338807,
+      "grad_norm": 0.30316677689552307,
+      "learning_rate": 2.745272628781495e-06,
+      "loss": 0.1659,
+      "step": 2125
+    },
+    {
+      "epoch": 0.9697799064887672,
+      "grad_norm": 0.3249475061893463,
+      "learning_rate": 2.7434343467065993e-06,
+      "loss": 0.1859,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9702360588436538,
+      "grad_norm": 0.32630324363708496,
+      "learning_rate": 2.741595931740857e-06,
+      "loss": 0.178,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9706922111985403,
+      "grad_norm": 0.29812943935394287,
+      "learning_rate": 2.739757384887859e-06,
+      "loss": 0.1626,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9711483635534268,
+      "grad_norm": 0.3231302499771118,
+      "learning_rate": 2.7379187071512694e-06,
+      "loss": 0.1818,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9716045159083134,
+      "grad_norm": 0.3188766539096832,
+      "learning_rate": 2.7360798995348204e-06,
+      "loss": 0.1822,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9720606682631999,
+      "grad_norm": 0.30705463886260986,
+      "learning_rate": 2.7342409630423195e-06,
+      "loss": 0.1719,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9725168206180864,
+      "grad_norm": 0.3012809157371521,
+      "learning_rate": 2.732401898677642e-06,
+      "loss": 0.1685,
+      "step": 2132
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 0.2905045747756958,
+      "learning_rate": 2.7305627074447318e-06,
+      "loss": 0.1708,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9734291253278595,
+      "grad_norm": 0.30509981513023376,
+      "learning_rate": 2.7287233903476064e-06,
+      "loss": 0.1753,
+      "step": 2134
+    },
+    {
+      "epoch": 0.973885277682746,
+      "grad_norm": 0.30524182319641113,
+      "learning_rate": 2.7268839483903474e-06,
+      "loss": 0.1742,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9743414300376325,
+      "grad_norm": 0.30904024839401245,
+      "learning_rate": 2.725044382577107e-06,
+      "loss": 0.1726,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9747975823925191,
+      "grad_norm": 0.2919926345348358,
+      "learning_rate": 2.7232046939121037e-06,
+      "loss": 0.1631,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9752537347474056,
+      "grad_norm": 0.2963598966598511,
+      "learning_rate": 2.7213648833996243e-06,
+      "loss": 0.1665,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9757098871022921,
+      "grad_norm": 0.3199349045753479,
+      "learning_rate": 2.7195249520440227e-06,
+      "loss": 0.1722,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9761660394571787,
+      "grad_norm": 0.3061026334762573,
+      "learning_rate": 2.7176849008497165e-06,
+      "loss": 0.1653,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9766221918120652,
+      "grad_norm": 0.3068127930164337,
+      "learning_rate": 2.7158447308211893e-06,
+      "loss": 0.1718,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9770783441669517,
+      "grad_norm": 0.3143504559993744,
+      "learning_rate": 2.7140044429629918e-06,
+      "loss": 0.1665,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9775344965218383,
+      "grad_norm": 0.3093785047531128,
+      "learning_rate": 2.712164038279737e-06,
+      "loss": 0.1767,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9779906488767248,
+      "grad_norm": 0.30348923802375793,
+      "learning_rate": 2.7103235177761018e-06,
+      "loss": 0.1754,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9784468012316113,
+      "grad_norm": 0.31281015276908875,
+      "learning_rate": 2.7084828824568256e-06,
+      "loss": 0.1749,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9789029535864979,
+      "grad_norm": 0.30854693055152893,
+      "learning_rate": 2.7066421333267136e-06,
+      "loss": 0.1716,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9793591059413844,
+      "grad_norm": 0.3090038001537323,
+      "learning_rate": 2.7048012713906296e-06,
+      "loss": 0.1782,
+      "step": 2147
+    },
+    {
+      "epoch": 0.9798152582962709,
+      "grad_norm": 0.30777889490127563,
+      "learning_rate": 2.702960297653501e-06,
+      "loss": 0.1798,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9802714106511575,
+      "grad_norm": 0.30617496371269226,
+      "learning_rate": 2.7011192131203147e-06,
+      "loss": 0.1702,
+      "step": 2149
+    },
+    {
+      "epoch": 0.980727563006044,
+      "grad_norm": 0.3002450168132782,
+      "learning_rate": 2.6992780187961197e-06,
+      "loss": 0.1652,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9811837153609305,
+      "grad_norm": 0.2985190153121948,
+      "learning_rate": 2.697436715686025e-06,
+      "loss": 0.1695,
+      "step": 2151
+    },
+    {
+      "epoch": 0.9816398677158171,
+      "grad_norm": 0.3162119388580322,
+      "learning_rate": 2.695595304795197e-06,
+      "loss": 0.1756,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9820960200707036,
+      "grad_norm": 0.30369311571121216,
+      "learning_rate": 2.693753787128863e-06,
+      "loss": 0.1713,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9825521724255901,
+      "grad_norm": 0.3029561936855316,
+      "learning_rate": 2.6919121636923073e-06,
+      "loss": 0.169,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9830083247804767,
+      "grad_norm": 0.2929121255874634,
+      "learning_rate": 2.690070435490873e-06,
+      "loss": 0.1635,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9834644771353632,
+      "grad_norm": 0.30831754207611084,
+      "learning_rate": 2.688228603529959e-06,
+      "loss": 0.1647,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9839206294902497,
+      "grad_norm": 0.3395315706729889,
+      "learning_rate": 2.6863866688150225e-06,
+      "loss": 0.1822,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9843767818451363,
+      "grad_norm": 0.31319454312324524,
+      "learning_rate": 2.684544632351575e-06,
+      "loss": 0.1711,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9848329342000228,
+      "grad_norm": 0.3161505460739136,
+      "learning_rate": 2.682702495145185e-06,
+      "loss": 0.171,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9852890865549093,
+      "grad_norm": 0.2971371114253998,
+      "learning_rate": 2.680860258201475e-06,
+      "loss": 0.1649,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9857452389097958,
+      "grad_norm": 0.29178425669670105,
+      "learning_rate": 2.6790179225261236e-06,
+      "loss": 0.171,
+      "step": 2161
+    },
+    {
+      "epoch": 0.9862013912646824,
+      "grad_norm": 0.3111404478549957,
+      "learning_rate": 2.6771754891248614e-06,
+      "loss": 0.1601,
+      "step": 2162
+    },
+    {
+      "epoch": 0.9866575436195689,
+      "grad_norm": 0.32499390840530396,
+      "learning_rate": 2.6753329590034723e-06,
+      "loss": 0.1806,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9871136959744554,
+      "grad_norm": 0.3276522159576416,
+      "learning_rate": 2.6734903331677946e-06,
+      "loss": 0.178,
+      "step": 2164
+    },
+    {
+      "epoch": 0.987569848329342,
+      "grad_norm": 0.3157401382923126,
+      "learning_rate": 2.6716476126237185e-06,
+      "loss": 0.176,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9880260006842285,
+      "grad_norm": 0.31126946210861206,
+      "learning_rate": 2.6698047983771847e-06,
+      "loss": 0.1737,
+      "step": 2166
+    },
+    {
+      "epoch": 0.988482153039115,
+      "grad_norm": 0.2875387668609619,
+      "learning_rate": 2.6679618914341858e-06,
+      "loss": 0.1686,
+      "step": 2167
+    },
+    {
+      "epoch": 0.9889383053940016,
+      "grad_norm": 0.2988435626029968,
+      "learning_rate": 2.666118892800765e-06,
+      "loss": 0.1637,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9893944577488881,
+      "grad_norm": 0.303847998380661,
+      "learning_rate": 2.664275803483016e-06,
+      "loss": 0.1685,
+      "step": 2169
+    },
+    {
+      "epoch": 0.9898506101037746,
+      "grad_norm": 0.32962435483932495,
+      "learning_rate": 2.662432624487081e-06,
+      "loss": 0.1803,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9903067624586612,
+      "grad_norm": 0.318949431180954,
+      "learning_rate": 2.6605893568191515e-06,
+      "loss": 0.1692,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9907629148135477,
+      "grad_norm": 0.30072420835494995,
+      "learning_rate": 2.658746001485469e-06,
+      "loss": 0.1751,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9912190671684342,
+      "grad_norm": 0.3179664611816406,
+      "learning_rate": 2.6569025594923197e-06,
+      "loss": 0.16,
+      "step": 2173
+    },
+    {
+      "epoch": 0.9916752195233208,
+      "grad_norm": 0.31549254059791565,
+      "learning_rate": 2.6550590318460397e-06,
+      "loss": 0.1817,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9921313718782073,
+      "grad_norm": 0.3236657679080963,
+      "learning_rate": 2.65321541955301e-06,
+      "loss": 0.1761,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9925875242330938,
+      "grad_norm": 0.32296234369277954,
+      "learning_rate": 2.651371723619661e-06,
+      "loss": 0.1767,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9930436765879804,
+      "grad_norm": 0.3087942600250244,
+      "learning_rate": 2.649527945052465e-06,
+      "loss": 0.1753,
+      "step": 2177
+    },
+    {
+      "epoch": 0.9934998289428669,
+      "grad_norm": 0.3095649778842926,
+      "learning_rate": 2.647684084857941e-06,
+      "loss": 0.1714,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9939559812977534,
+      "grad_norm": 0.2999379634857178,
+      "learning_rate": 2.645840144042653e-06,
+      "loss": 0.177,
+      "step": 2179
+    },
+    {
+      "epoch": 0.99441213365264,
+      "grad_norm": 0.33610638976097107,
+      "learning_rate": 2.6439961236132083e-06,
+      "loss": 0.1761,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9948682860075265,
+      "grad_norm": 0.32506120204925537,
+      "learning_rate": 2.642152024576259e-06,
+      "loss": 0.1835,
+      "step": 2181
+    },
+    {
+      "epoch": 0.995324438362413,
+      "grad_norm": 0.32096078991889954,
+      "learning_rate": 2.640307847938497e-06,
+      "loss": 0.1736,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9957805907172996,
+      "grad_norm": 0.31698352098464966,
+      "learning_rate": 2.63846359470666e-06,
+      "loss": 0.1737,
+      "step": 2183
+    },
+    {
+      "epoch": 0.9962367430721861,
+      "grad_norm": 0.3094034790992737,
+      "learning_rate": 2.6366192658875256e-06,
+      "loss": 0.1739,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9966928954270726,
+      "grad_norm": 0.2955319285392761,
+      "learning_rate": 2.6347748624879145e-06,
+      "loss": 0.1683,
+      "step": 2185
+    },
+    {
+      "epoch": 0.9971490477819591,
+      "grad_norm": 0.3264576494693756,
+      "learning_rate": 2.632930385514684e-06,
+      "loss": 0.1847,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9976052001368457,
+      "grad_norm": 0.3200644850730896,
+      "learning_rate": 2.6310858359747376e-06,
+      "loss": 0.1718,
+      "step": 2187
+    },
+    {
+      "epoch": 0.9980613524917322,
+      "grad_norm": 0.33452606201171875,
+      "learning_rate": 2.629241214875013e-06,
+      "loss": 0.1731,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9985175048466187,
+      "grad_norm": 0.30290496349334717,
+      "learning_rate": 2.6273965232224895e-06,
+      "loss": 0.167,
+      "step": 2189
+    },
+    {
+      "epoch": 0.9989736572015053,
+      "grad_norm": 0.31510254740715027,
+      "learning_rate": 2.625551762024185e-06,
+      "loss": 0.1806,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9994298095563918,
+      "grad_norm": 0.3053393065929413,
+      "learning_rate": 2.623706932287155e-06,
+      "loss": 0.1714,
+      "step": 2191
+    },
+    {
+      "epoch": 0.9998859619112783,
+      "grad_norm": 0.3165036141872406,
+      "learning_rate": 2.621862035018492e-06,
+      "loss": 0.1781,
+      "step": 2192
+    },
+    {
+      "epoch": 1.0003421142661648,
+      "grad_norm": 0.2965346872806549,
+      "learning_rate": 2.6200170712253253e-06,
+      "loss": 0.1531,
+      "step": 2193
+    },
+    {
+      "epoch": 1.0007982666210515,
+      "grad_norm": 0.3012768626213074,
+      "learning_rate": 2.618172041914822e-06,
+      "loss": 0.1628,
+      "step": 2194
+    },
+    {
+      "epoch": 1.001254418975938,
+      "grad_norm": 0.2806198298931122,
+      "learning_rate": 2.6163269480941826e-06,
+      "loss": 0.1446,
+      "step": 2195
+    },
+    {
+      "epoch": 1.0017105713308245,
+      "grad_norm": 0.30723991990089417,
+      "learning_rate": 2.6144817907706453e-06,
+      "loss": 0.1509,
+      "step": 2196
+    },
+    {
+      "epoch": 1.002166723685711,
+      "grad_norm": 0.3128562569618225,
+      "learning_rate": 2.612636570951481e-06,
+      "loss": 0.1643,
+      "step": 2197
+    },
+    {
+      "epoch": 1.0026228760405975,
+      "grad_norm": 0.27928367257118225,
+      "learning_rate": 2.610791289643996e-06,
+      "loss": 0.1422,
+      "step": 2198
+    },
+    {
+      "epoch": 1.003079028395484,
+      "grad_norm": 0.30517908930778503,
+      "learning_rate": 2.608945947855529e-06,
+      "loss": 0.1546,
+      "step": 2199
+    },
+    {
+      "epoch": 1.0035351807503705,
+      "grad_norm": 0.2774403989315033,
+      "learning_rate": 2.607100546593453e-06,
+      "loss": 0.1503,
+      "step": 2200
+    },
+    {
+      "epoch": 1.0039913331052572,
+      "grad_norm": 0.2793969511985779,
+      "learning_rate": 2.6052550868651715e-06,
+      "loss": 0.1486,
+      "step": 2201
+    },
+    {
+      "epoch": 1.0044474854601437,
+      "grad_norm": 0.3040004074573517,
+      "learning_rate": 2.603409569678123e-06,
+      "loss": 0.1596,
+      "step": 2202
+    },
+    {
+      "epoch": 1.0049036378150302,
+      "grad_norm": 0.2952609360218048,
+      "learning_rate": 2.6015639960397754e-06,
+      "loss": 0.153,
+      "step": 2203
+    },
+    {
+      "epoch": 1.0053597901699167,
+      "grad_norm": 0.30683255195617676,
+      "learning_rate": 2.5997183669576264e-06,
+      "loss": 0.1542,
+      "step": 2204
+    },
+    {
+      "epoch": 1.0058159425248032,
+      "grad_norm": 0.30917850136756897,
+      "learning_rate": 2.5978726834392055e-06,
+      "loss": 0.1562,
+      "step": 2205
+    },
+    {
+      "epoch": 1.0062720948796897,
+      "grad_norm": 0.31984400749206543,
+      "learning_rate": 2.5960269464920727e-06,
+      "loss": 0.1643,
+      "step": 2206
+    },
+    {
+      "epoch": 1.0067282472345764,
+      "grad_norm": 0.3237025737762451,
+      "learning_rate": 2.5941811571238157e-06,
+      "loss": 0.1613,
+      "step": 2207
+    },
+    {
+      "epoch": 1.007184399589463,
+      "grad_norm": 0.29265496134757996,
+      "learning_rate": 2.59233531634205e-06,
+      "loss": 0.1495,
+      "step": 2208
+    },
+    {
+      "epoch": 1.0076405519443494,
+      "grad_norm": 0.31332945823669434,
+      "learning_rate": 2.5904894251544214e-06,
+      "loss": 0.1566,
+      "step": 2209
+    },
+    {
+      "epoch": 1.008096704299236,
+      "grad_norm": 0.314137727022171,
+      "learning_rate": 2.588643484568602e-06,
+      "loss": 0.1607,
+      "step": 2210
+    },
+    {
+      "epoch": 1.0085528566541224,
+      "grad_norm": 0.3159463405609131,
+      "learning_rate": 2.5867974955922915e-06,
+      "loss": 0.1562,
+      "step": 2211
+    },
+    {
+      "epoch": 1.009009009009009,
+      "grad_norm": 0.3082673251628876,
+      "learning_rate": 2.584951459233215e-06,
+      "loss": 0.1548,
+      "step": 2212
+    },
+    {
+      "epoch": 1.0094651613638956,
+      "grad_norm": 0.29175305366516113,
+      "learning_rate": 2.583105376499124e-06,
+      "loss": 0.1581,
+      "step": 2213
+    },
+    {
+      "epoch": 1.0099213137187821,
+      "grad_norm": 0.28802669048309326,
+      "learning_rate": 2.5812592483977956e-06,
+      "loss": 0.1526,
+      "step": 2214
+    },
+    {
+      "epoch": 1.0103774660736686,
+      "grad_norm": 0.3163564205169678,
+      "learning_rate": 2.5794130759370305e-06,
+      "loss": 0.16,
+      "step": 2215
+    },
+    {
+      "epoch": 1.0108336184285551,
+      "grad_norm": 0.3020622134208679,
+      "learning_rate": 2.5775668601246555e-06,
+      "loss": 0.1616,
+      "step": 2216
+    },
+    {
+      "epoch": 1.0112897707834416,
+      "grad_norm": 0.2995922565460205,
+      "learning_rate": 2.575720601968521e-06,
+      "loss": 0.1463,
+      "step": 2217
+    },
+    {
+      "epoch": 1.0117459231383281,
+      "grad_norm": 0.3059779107570648,
+      "learning_rate": 2.5738743024764977e-06,
+      "loss": 0.1605,
+      "step": 2218
+    },
+    {
+      "epoch": 1.0122020754932148,
+      "grad_norm": 0.3093193769454956,
+      "learning_rate": 2.5720279626564816e-06,
+      "loss": 0.156,
+      "step": 2219
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.33121731877326965,
+      "learning_rate": 2.5701815835163896e-06,
+      "loss": 0.1568,
+      "step": 2220
+    },
+    {
+      "epoch": 1.0131143802029878,
+      "grad_norm": 0.2849758565425873,
+      "learning_rate": 2.5683351660641614e-06,
+      "loss": 0.1477,
+      "step": 2221
+    },
+    {
+      "epoch": 1.0135705325578743,
+      "grad_norm": 0.29688236117362976,
+      "learning_rate": 2.566488711307755e-06,
+      "loss": 0.1543,
+      "step": 2222
+    },
+    {
+      "epoch": 1.0140266849127608,
+      "grad_norm": 0.30517637729644775,
+      "learning_rate": 2.564642220255151e-06,
+      "loss": 0.1576,
+      "step": 2223
+    },
+    {
+      "epoch": 1.0144828372676473,
+      "grad_norm": 0.2935236394405365,
+      "learning_rate": 2.5627956939143507e-06,
+      "loss": 0.1518,
+      "step": 2224
+    },
+    {
+      "epoch": 1.014938989622534,
+      "grad_norm": 0.2980383634567261,
+      "learning_rate": 2.56094913329337e-06,
+      "loss": 0.1551,
+      "step": 2225
+    },
+    {
+      "epoch": 1.0153951419774205,
+      "grad_norm": 0.3179318904876709,
+      "learning_rate": 2.5591025394002487e-06,
+      "loss": 0.1603,
+      "step": 2226
+    },
+    {
+      "epoch": 1.015851294332307,
+      "grad_norm": 0.3024425804615021,
+      "learning_rate": 2.5572559132430422e-06,
+      "loss": 0.1529,
+      "step": 2227
+    },
+    {
+      "epoch": 1.0163074466871935,
+      "grad_norm": 0.30249929428100586,
+      "learning_rate": 2.555409255829825e-06,
+      "loss": 0.1493,
+      "step": 2228
+    },
+    {
+      "epoch": 1.01676359904208,
+      "grad_norm": 0.33018338680267334,
+      "learning_rate": 2.553562568168686e-06,
+      "loss": 0.1553,
+      "step": 2229
+    },
+    {
+      "epoch": 1.0172197513969665,
+      "grad_norm": 0.30230361223220825,
+      "learning_rate": 2.5517158512677328e-06,
+      "loss": 0.1486,
+      "step": 2230
+    },
+    {
+      "epoch": 1.017675903751853,
+      "grad_norm": 0.2952478229999542,
+      "learning_rate": 2.5498691061350885e-06,
+      "loss": 0.1525,
+      "step": 2231
+    },
+    {
+      "epoch": 1.0181320561067397,
+      "grad_norm": 0.31507235765457153,
+      "learning_rate": 2.548022333778892e-06,
+      "loss": 0.1533,
+      "step": 2232
+    },
+    {
+      "epoch": 1.0185882084616262,
+      "grad_norm": 0.3138676583766937,
+      "learning_rate": 2.5461755352072962e-06,
+      "loss": 0.1548,
+      "step": 2233
+    },
+    {
+      "epoch": 1.0190443608165127,
+      "grad_norm": 0.3054126501083374,
+      "learning_rate": 2.5443287114284687e-06,
+      "loss": 0.1619,
+      "step": 2234
+    },
+    {
+      "epoch": 1.0195005131713992,
+      "grad_norm": 0.30826279520988464,
+      "learning_rate": 2.5424818634505916e-06,
+      "loss": 0.1535,
+      "step": 2235
+    },
+    {
+      "epoch": 1.0199566655262857,
+      "grad_norm": 0.3322764039039612,
+      "learning_rate": 2.540634992281858e-06,
+      "loss": 0.1623,
+      "step": 2236
+    },
+    {
+      "epoch": 1.0204128178811722,
+      "grad_norm": 0.33079448342323303,
+      "learning_rate": 2.5387880989304762e-06,
+      "loss": 0.159,
+      "step": 2237
+    },
+    {
+      "epoch": 1.020868970236059,
+      "grad_norm": 0.3026164174079895,
+      "learning_rate": 2.5369411844046654e-06,
+      "loss": 0.1571,
+      "step": 2238
+    },
+    {
+      "epoch": 1.0213251225909454,
+      "grad_norm": 0.29058417677879333,
+      "learning_rate": 2.535094249712657e-06,
+      "loss": 0.1492,
+      "step": 2239
+    },
+    {
+      "epoch": 1.021781274945832,
+      "grad_norm": 0.2958424985408783,
+      "learning_rate": 2.5332472958626923e-06,
+      "loss": 0.1466,
+      "step": 2240
+    },
+    {
+      "epoch": 1.0222374273007184,
+      "grad_norm": 0.32235005497932434,
+      "learning_rate": 2.531400323863023e-06,
+      "loss": 0.1637,
+      "step": 2241
+    },
+    {
+      "epoch": 1.022693579655605,
+      "grad_norm": 0.33696475625038147,
+      "learning_rate": 2.5295533347219124e-06,
+      "loss": 0.1649,
+      "step": 2242
+    },
+    {
+      "epoch": 1.0231497320104914,
+      "grad_norm": 0.3055405914783478,
+      "learning_rate": 2.5277063294476323e-06,
+      "loss": 0.1551,
+      "step": 2243
+    },
+    {
+      "epoch": 1.0236058843653781,
+      "grad_norm": 0.3082205057144165,
+      "learning_rate": 2.525859309048463e-06,
+      "loss": 0.1659,
+      "step": 2244
+    },
+    {
+      "epoch": 1.0240620367202646,
+      "grad_norm": 0.29391857981681824,
+      "learning_rate": 2.5240122745326924e-06,
+      "loss": 0.1539,
+      "step": 2245
+    },
+    {
+      "epoch": 1.0245181890751511,
+      "grad_norm": 0.312833309173584,
+      "learning_rate": 2.5221652269086178e-06,
+      "loss": 0.159,
+      "step": 2246
+    },
+    {
+      "epoch": 1.0249743414300376,
+      "grad_norm": 0.31205928325653076,
+      "learning_rate": 2.5203181671845424e-06,
+      "loss": 0.1629,
+      "step": 2247
+    },
+    {
+      "epoch": 1.0254304937849241,
+      "grad_norm": 0.3087555170059204,
+      "learning_rate": 2.518471096368777e-06,
+      "loss": 0.1503,
+      "step": 2248
+    },
+    {
+      "epoch": 1.0258866461398106,
+      "grad_norm": 0.31644195318222046,
+      "learning_rate": 2.5166240154696363e-06,
+      "loss": 0.1638,
+      "step": 2249
+    },
+    {
+      "epoch": 1.0263427984946971,
+      "grad_norm": 0.30362093448638916,
+      "learning_rate": 2.514776925495444e-06,
+      "loss": 0.1551,
+      "step": 2250
+    },
+    {
+      "epoch": 1.0267989508495838,
+      "grad_norm": 0.30277901887893677,
+      "learning_rate": 2.5129298274545255e-06,
+      "loss": 0.1611,
+      "step": 2251
+    },
+    {
+      "epoch": 1.0272551032044703,
+      "grad_norm": 0.30254557728767395,
+      "learning_rate": 2.511082722355212e-06,
+      "loss": 0.1555,
+      "step": 2252
+    },
+    {
+      "epoch": 1.0277112555593568,
+      "grad_norm": 0.3123329281806946,
+      "learning_rate": 2.5092356112058393e-06,
+      "loss": 0.1612,
+      "step": 2253
+    },
+    {
+      "epoch": 1.0281674079142433,
+      "grad_norm": 0.29911330342292786,
+      "learning_rate": 2.507388495014744e-06,
+      "loss": 0.1555,
+      "step": 2254
+    },
+    {
+      "epoch": 1.0286235602691298,
+      "grad_norm": 0.31661123037338257,
+      "learning_rate": 2.5055413747902694e-06,
+      "loss": 0.1562,
+      "step": 2255
+    },
+    {
+      "epoch": 1.0290797126240163,
+      "grad_norm": 0.3083839416503906,
+      "learning_rate": 2.503694251540757e-06,
+      "loss": 0.1552,
+      "step": 2256
+    },
+    {
+      "epoch": 1.029535864978903,
+      "grad_norm": 0.31602880358695984,
+      "learning_rate": 2.5018471262745513e-06,
+      "loss": 0.1615,
+      "step": 2257
+    },
+    {
+      "epoch": 1.0299920173337895,
+      "grad_norm": 0.32992851734161377,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1652,
+      "step": 2258
+    },
+    {
+      "epoch": 1.030448169688676,
+      "grad_norm": 0.30462247133255005,
+      "learning_rate": 2.498152873725449e-06,
+      "loss": 0.1544,
+      "step": 2259
+    },
+    {
+      "epoch": 1.0309043220435625,
+      "grad_norm": 0.30221256613731384,
+      "learning_rate": 2.496305748459244e-06,
+      "loss": 0.1525,
+      "step": 2260
+    },
+    {
+      "epoch": 1.031360474398449,
+      "grad_norm": 0.29837363958358765,
+      "learning_rate": 2.4944586252097314e-06,
+      "loss": 0.1556,
+      "step": 2261
+    },
+    {
+      "epoch": 1.0318166267533355,
+      "grad_norm": 0.30723658204078674,
+      "learning_rate": 2.4926115049852567e-06,
+      "loss": 0.1544,
+      "step": 2262
+    },
+    {
+      "epoch": 1.0322727791082222,
+      "grad_norm": 0.3049622178077698,
+      "learning_rate": 2.490764388794162e-06,
+      "loss": 0.1512,
+      "step": 2263
+    },
+    {
+      "epoch": 1.0327289314631087,
+      "grad_norm": 0.3173573911190033,
+      "learning_rate": 2.4889172776447885e-06,
+      "loss": 0.157,
+      "step": 2264
+    },
+    {
+      "epoch": 1.0331850838179952,
+      "grad_norm": 0.30256882309913635,
+      "learning_rate": 2.487070172545475e-06,
+      "loss": 0.1538,
+      "step": 2265
+    },
+    {
+      "epoch": 1.0336412361728817,
+      "grad_norm": 0.28806358575820923,
+      "learning_rate": 2.4852230745045568e-06,
+      "loss": 0.1518,
+      "step": 2266
+    },
+    {
+      "epoch": 1.0340973885277682,
+      "grad_norm": 0.3219054043292999,
+      "learning_rate": 2.483375984530364e-06,
+      "loss": 0.1515,
+      "step": 2267
+    },
+    {
+      "epoch": 1.0345535408826547,
+      "grad_norm": 0.29139336943626404,
+      "learning_rate": 2.4815289036312236e-06,
+      "loss": 0.1552,
+      "step": 2268
+    },
+    {
+      "epoch": 1.0350096932375414,
+      "grad_norm": 0.28353777527809143,
+      "learning_rate": 2.479681832815458e-06,
+      "loss": 0.1429,
+      "step": 2269
+    },
+    {
+      "epoch": 1.035465845592428,
+      "grad_norm": 0.315647155046463,
+      "learning_rate": 2.477834773091383e-06,
+      "loss": 0.1569,
+      "step": 2270
+    },
+    {
+      "epoch": 1.0359219979473144,
+      "grad_norm": 0.30197957158088684,
+      "learning_rate": 2.475987725467308e-06,
+      "loss": 0.1597,
+      "step": 2271
+    },
+    {
+      "epoch": 1.036378150302201,
+      "grad_norm": 0.3065713346004486,
+      "learning_rate": 2.474140690951538e-06,
+      "loss": 0.1491,
+      "step": 2272
+    },
+    {
+      "epoch": 1.0368343026570874,
+      "grad_norm": 0.3195562958717346,
+      "learning_rate": 2.4722936705523685e-06,
+      "loss": 0.155,
+      "step": 2273
+    },
+    {
+      "epoch": 1.037290455011974,
+      "grad_norm": 0.3105924129486084,
+      "learning_rate": 2.470446665278088e-06,
+      "loss": 0.16,
+      "step": 2274
+    },
+    {
+      "epoch": 1.0377466073668606,
+      "grad_norm": 0.30231398344039917,
+      "learning_rate": 2.4685996761369776e-06,
+      "loss": 0.153,
+      "step": 2275
+    },
+    {
+      "epoch": 1.0382027597217471,
+      "grad_norm": 0.29108160734176636,
+      "learning_rate": 2.4667527041373085e-06,
+      "loss": 0.1512,
+      "step": 2276
+    },
+    {
+      "epoch": 1.0386589120766336,
+      "grad_norm": 0.30156388878822327,
+      "learning_rate": 2.4649057502873434e-06,
+      "loss": 0.1459,
+      "step": 2277
+    },
+    {
+      "epoch": 1.0391150644315201,
+      "grad_norm": 0.30411794781684875,
+      "learning_rate": 2.463058815595335e-06,
+      "loss": 0.1541,
+      "step": 2278
+    },
+    {
+      "epoch": 1.0395712167864066,
+      "grad_norm": 0.3148190677165985,
+      "learning_rate": 2.461211901069524e-06,
+      "loss": 0.1526,
+      "step": 2279
+    },
+    {
+      "epoch": 1.0400273691412931,
+      "grad_norm": 0.3218180239200592,
+      "learning_rate": 2.459365007718143e-06,
+      "loss": 0.1583,
+      "step": 2280
+    },
+    {
+      "epoch": 1.0404835214961796,
+      "grad_norm": 0.3015739619731903,
+      "learning_rate": 2.4575181365494096e-06,
+      "loss": 0.1606,
+      "step": 2281
+    },
+    {
+      "epoch": 1.0409396738510663,
+      "grad_norm": 0.2966444492340088,
+      "learning_rate": 2.4556712885715318e-06,
+      "loss": 0.1529,
+      "step": 2282
+    },
+    {
+      "epoch": 1.0413958262059528,
+      "grad_norm": 0.3134633004665375,
+      "learning_rate": 2.4538244647927046e-06,
+      "loss": 0.1522,
+      "step": 2283
+    },
+    {
+      "epoch": 1.0418519785608393,
+      "grad_norm": 0.3058009147644043,
+      "learning_rate": 2.4519776662211083e-06,
+      "loss": 0.1535,
+      "step": 2284
+    },
+    {
+      "epoch": 1.0423081309157258,
+      "grad_norm": 0.31028133630752563,
+      "learning_rate": 2.4501308938649115e-06,
+      "loss": 0.1571,
+      "step": 2285
+    },
+    {
+      "epoch": 1.0427642832706123,
+      "grad_norm": 0.2927097976207733,
+      "learning_rate": 2.4482841487322676e-06,
+      "loss": 0.1409,
+      "step": 2286
+    },
+    {
+      "epoch": 1.0432204356254988,
+      "grad_norm": 0.29738518595695496,
+      "learning_rate": 2.4464374318313145e-06,
+      "loss": 0.1507,
+      "step": 2287
+    },
+    {
+      "epoch": 1.0436765879803855,
+      "grad_norm": 0.3097405731678009,
+      "learning_rate": 2.444590744170176e-06,
+      "loss": 0.1497,
+      "step": 2288
+    },
+    {
+      "epoch": 1.044132740335272,
+      "grad_norm": 0.3048311471939087,
+      "learning_rate": 2.442744086756958e-06,
+      "loss": 0.1526,
+      "step": 2289
+    },
+    {
+      "epoch": 1.0445888926901585,
+      "grad_norm": 0.2945125102996826,
+      "learning_rate": 2.4408974605997517e-06,
+      "loss": 0.1502,
+      "step": 2290
+    },
+    {
+      "epoch": 1.045045045045045,
+      "grad_norm": 0.3074931800365448,
+      "learning_rate": 2.439050866706631e-06,
+      "loss": 0.1492,
+      "step": 2291
+    },
+    {
+      "epoch": 1.0455011973999315,
+      "grad_norm": 0.29824402928352356,
+      "learning_rate": 2.437204306085651e-06,
+      "loss": 0.1493,
+      "step": 2292
+    },
+    {
+      "epoch": 1.045957349754818,
+      "grad_norm": 0.293407678604126,
+      "learning_rate": 2.4353577797448493e-06,
+      "loss": 0.1574,
+      "step": 2293
+    },
+    {
+      "epoch": 1.0464135021097047,
+      "grad_norm": 0.3277481496334076,
+      "learning_rate": 2.4335112886922452e-06,
+      "loss": 0.1631,
+      "step": 2294
+    },
+    {
+      "epoch": 1.0468696544645912,
+      "grad_norm": 0.3238832950592041,
+      "learning_rate": 2.431664833935839e-06,
+      "loss": 0.158,
+      "step": 2295
+    },
+    {
+      "epoch": 1.0473258068194777,
+      "grad_norm": 0.3290277421474457,
+      "learning_rate": 2.4298184164836104e-06,
+      "loss": 0.1592,
+      "step": 2296
+    },
+    {
+      "epoch": 1.0477819591743642,
+      "grad_norm": 0.30506718158721924,
+      "learning_rate": 2.427972037343519e-06,
+      "loss": 0.1566,
+      "step": 2297
+    },
+    {
+      "epoch": 1.0482381115292507,
+      "grad_norm": 0.30278289318084717,
+      "learning_rate": 2.4261256975235027e-06,
+      "loss": 0.1489,
+      "step": 2298
+    },
+    {
+      "epoch": 1.0486942638841372,
+      "grad_norm": 0.3021693229675293,
+      "learning_rate": 2.42427939803148e-06,
+      "loss": 0.1492,
+      "step": 2299
+    },
+    {
+      "epoch": 1.0491504162390237,
+      "grad_norm": 0.32313692569732666,
+      "learning_rate": 2.422433139875345e-06,
+      "loss": 0.1594,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0496065685939104,
+      "grad_norm": 0.29050105810165405,
+      "learning_rate": 2.42058692406297e-06,
+      "loss": 0.1471,
+      "step": 2301
+    },
+    {
+      "epoch": 1.050062720948797,
+      "grad_norm": 0.31139975786209106,
+      "learning_rate": 2.4187407516022056e-06,
+      "loss": 0.1645,
+      "step": 2302
+    },
+    {
+      "epoch": 1.0505188733036834,
+      "grad_norm": 0.29966482520103455,
+      "learning_rate": 2.4168946235008774e-06,
+      "loss": 0.1514,
+      "step": 2303
+    },
+    {
+      "epoch": 1.05097502565857,
+      "grad_norm": 0.30819904804229736,
+      "learning_rate": 2.4150485407667855e-06,
+      "loss": 0.1565,
+      "step": 2304
+    },
+    {
+      "epoch": 1.0514311780134564,
+      "grad_norm": 0.3194921612739563,
+      "learning_rate": 2.413202504407709e-06,
+      "loss": 0.1627,
+      "step": 2305
+    },
+    {
+      "epoch": 1.051887330368343,
+      "grad_norm": 0.3258031904697418,
+      "learning_rate": 2.411356515431398e-06,
+      "loss": 0.1672,
+      "step": 2306
+    },
+    {
+      "epoch": 1.0523434827232296,
+      "grad_norm": 0.3231761157512665,
+      "learning_rate": 2.409510574845579e-06,
+      "loss": 0.1612,
+      "step": 2307
+    },
+    {
+      "epoch": 1.0527996350781161,
+      "grad_norm": 0.30210381746292114,
+      "learning_rate": 2.4076646836579508e-06,
+      "loss": 0.1561,
+      "step": 2308
+    },
+    {
+      "epoch": 1.0532557874330026,
+      "grad_norm": 0.31846266984939575,
+      "learning_rate": 2.405818842876185e-06,
+      "loss": 0.1583,
+      "step": 2309
+    },
+    {
+      "epoch": 1.0537119397878891,
+      "grad_norm": 0.30924952030181885,
+      "learning_rate": 2.4039730535079277e-06,
+      "loss": 0.1585,
+      "step": 2310
+    },
+    {
+      "epoch": 1.0541680921427756,
+      "grad_norm": 0.32560425996780396,
+      "learning_rate": 2.4021273165607954e-06,
+      "loss": 0.1549,
+      "step": 2311
+    },
+    {
+      "epoch": 1.0546242444976621,
+      "grad_norm": 0.3123047649860382,
+      "learning_rate": 2.4002816330423744e-06,
+      "loss": 0.1631,
+      "step": 2312
+    },
+    {
+      "epoch": 1.0550803968525488,
+      "grad_norm": 0.3088406026363373,
+      "learning_rate": 2.398436003960226e-06,
+      "loss": 0.1723,
+      "step": 2313
+    },
+    {
+      "epoch": 1.0555365492074353,
+      "grad_norm": 0.299550861120224,
+      "learning_rate": 2.396590430321878e-06,
+      "loss": 0.1602,
+      "step": 2314
+    },
+    {
+      "epoch": 1.0559927015623218,
+      "grad_norm": 0.2949889302253723,
+      "learning_rate": 2.3947449131348285e-06,
+      "loss": 0.1554,
+      "step": 2315
+    },
+    {
+      "epoch": 1.0564488539172083,
+      "grad_norm": 0.318534255027771,
+      "learning_rate": 2.392899453406548e-06,
+      "loss": 0.1577,
+      "step": 2316
+    },
+    {
+      "epoch": 1.0569050062720948,
+      "grad_norm": 0.3184177875518799,
+      "learning_rate": 2.3910540521444713e-06,
+      "loss": 0.1643,
+      "step": 2317
+    },
+    {
+      "epoch": 1.0573611586269813,
+      "grad_norm": 0.3195169270038605,
+      "learning_rate": 2.3892087103560045e-06,
+      "loss": 0.158,
+      "step": 2318
+    },
+    {
+      "epoch": 1.057817310981868,
+      "grad_norm": 0.298846572637558,
+      "learning_rate": 2.3873634290485198e-06,
+      "loss": 0.147,
+      "step": 2319
+    },
+    {
+      "epoch": 1.0582734633367545,
+      "grad_norm": 0.30916425585746765,
+      "learning_rate": 2.385518209229355e-06,
+      "loss": 0.1552,
+      "step": 2320
+    },
+    {
+      "epoch": 1.058729615691641,
+      "grad_norm": 0.31216779351234436,
+      "learning_rate": 2.3836730519058178e-06,
+      "loss": 0.1514,
+      "step": 2321
+    },
+    {
+      "epoch": 1.0591857680465275,
+      "grad_norm": 0.31287074089050293,
+      "learning_rate": 2.381827958085179e-06,
+      "loss": 0.1472,
+      "step": 2322
+    },
+    {
+      "epoch": 1.059641920401414,
+      "grad_norm": 0.30948731303215027,
+      "learning_rate": 2.379982928774675e-06,
+      "loss": 0.158,
+      "step": 2323
+    },
+    {
+      "epoch": 1.0600980727563005,
+      "grad_norm": 0.2959662973880768,
+      "learning_rate": 2.3781379649815094e-06,
+      "loss": 0.1497,
+      "step": 2324
+    },
+    {
+      "epoch": 1.0605542251111872,
+      "grad_norm": 0.30378252267837524,
+      "learning_rate": 2.376293067712845e-06,
+      "loss": 0.1574,
+      "step": 2325
+    },
+    {
+      "epoch": 1.0610103774660737,
+      "grad_norm": 0.30417200922966003,
+      "learning_rate": 2.374448237975815e-06,
+      "loss": 0.1564,
+      "step": 2326
+    },
+    {
+      "epoch": 1.0614665298209602,
+      "grad_norm": 0.31169456243515015,
+      "learning_rate": 2.372603476777511e-06,
+      "loss": 0.1498,
+      "step": 2327
+    },
+    {
+      "epoch": 1.0619226821758467,
+      "grad_norm": 0.31076836585998535,
+      "learning_rate": 2.3707587851249875e-06,
+      "loss": 0.1472,
+      "step": 2328
+    },
+    {
+      "epoch": 1.0623788345307332,
+      "grad_norm": 0.3299168646335602,
+      "learning_rate": 2.368914164025263e-06,
+      "loss": 0.1574,
+      "step": 2329
+    },
+    {
+      "epoch": 1.0628349868856197,
+      "grad_norm": 0.3123321831226349,
+      "learning_rate": 2.3670696144853163e-06,
+      "loss": 0.1533,
+      "step": 2330
+    },
+    {
+      "epoch": 1.0632911392405062,
+      "grad_norm": 0.30647292733192444,
+      "learning_rate": 2.3652251375120867e-06,
+      "loss": 0.1549,
+      "step": 2331
+    },
+    {
+      "epoch": 1.063747291595393,
+      "grad_norm": 0.29056668281555176,
+      "learning_rate": 2.3633807341124753e-06,
+      "loss": 0.1438,
+      "step": 2332
+    },
+    {
+      "epoch": 1.0642034439502794,
+      "grad_norm": 0.295025110244751,
+      "learning_rate": 2.361536405293341e-06,
+      "loss": 0.1535,
+      "step": 2333
+    },
+    {
+      "epoch": 1.064659596305166,
+      "grad_norm": 0.30024731159210205,
+      "learning_rate": 2.3596921520615047e-06,
+      "loss": 0.1496,
+      "step": 2334
+    },
+    {
+      "epoch": 1.0651157486600524,
+      "grad_norm": 0.3065096437931061,
+      "learning_rate": 2.357847975423742e-06,
+      "loss": 0.1503,
+      "step": 2335
+    },
+    {
+      "epoch": 1.065571901014939,
+      "grad_norm": 0.3045707046985626,
+      "learning_rate": 2.3560038763867913e-06,
+      "loss": 0.1602,
+      "step": 2336
+    },
+    {
+      "epoch": 1.0660280533698254,
+      "grad_norm": 0.3058183193206787,
+      "learning_rate": 2.354159855957347e-06,
+      "loss": 0.1572,
+      "step": 2337
+    },
+    {
+      "epoch": 1.0664842057247121,
+      "grad_norm": 0.29959750175476074,
+      "learning_rate": 2.35231591514206e-06,
+      "loss": 0.1538,
+      "step": 2338
+    },
+    {
+      "epoch": 1.0669403580795986,
+      "grad_norm": 0.3200899362564087,
+      "learning_rate": 2.3504720549475354e-06,
+      "loss": 0.1566,
+      "step": 2339
+    },
+    {
+      "epoch": 1.0673965104344851,
+      "grad_norm": 0.3018283247947693,
+      "learning_rate": 2.3486282763803397e-06,
+      "loss": 0.153,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0678526627893716,
+      "grad_norm": 0.29666638374328613,
+      "learning_rate": 2.3467845804469904e-06,
+      "loss": 0.1456,
+      "step": 2341
+    },
+    {
+      "epoch": 1.0683088151442581,
+      "grad_norm": 0.3126644194126129,
+      "learning_rate": 2.3449409681539615e-06,
+      "loss": 0.1604,
+      "step": 2342
+    },
+    {
+      "epoch": 1.0687649674991446,
+      "grad_norm": 0.3108762502670288,
+      "learning_rate": 2.3430974405076816e-06,
+      "loss": 0.1542,
+      "step": 2343
+    },
+    {
+      "epoch": 1.0692211198540313,
+      "grad_norm": 0.31192293763160706,
+      "learning_rate": 2.3412539985145324e-06,
+      "loss": 0.1602,
+      "step": 2344
+    },
+    {
+      "epoch": 1.0696772722089178,
+      "grad_norm": 0.3227790892124176,
+      "learning_rate": 2.3394106431808485e-06,
+      "loss": 0.1526,
+      "step": 2345
+    },
+    {
+      "epoch": 1.0701334245638043,
+      "grad_norm": 0.3191806972026825,
+      "learning_rate": 2.3375673755129197e-06,
+      "loss": 0.1533,
+      "step": 2346
+    },
+    {
+      "epoch": 1.0705895769186908,
+      "grad_norm": 0.2930872142314911,
+      "learning_rate": 2.3357241965169845e-06,
+      "loss": 0.1516,
+      "step": 2347
+    },
+    {
+      "epoch": 1.0710457292735773,
+      "grad_norm": 0.31123727560043335,
+      "learning_rate": 2.3338811071992353e-06,
+      "loss": 0.1515,
+      "step": 2348
+    },
+    {
+      "epoch": 1.0715018816284638,
+      "grad_norm": 0.305807501077652,
+      "learning_rate": 2.332038108565815e-06,
+      "loss": 0.149,
+      "step": 2349
+    },
+    {
+      "epoch": 1.0719580339833503,
+      "grad_norm": 0.32649123668670654,
+      "learning_rate": 2.3301952016228157e-06,
+      "loss": 0.1565,
+      "step": 2350
+    },
+    {
+      "epoch": 1.072414186338237,
+      "grad_norm": 0.31828823685646057,
+      "learning_rate": 2.3283523873762823e-06,
+      "loss": 0.1531,
+      "step": 2351
+    },
+    {
+      "epoch": 1.0728703386931235,
+      "grad_norm": 0.3053196668624878,
+      "learning_rate": 2.3265096668322063e-06,
+      "loss": 0.1626,
+      "step": 2352
+    },
+    {
+      "epoch": 1.07332649104801,
+      "grad_norm": 0.31288042664527893,
+      "learning_rate": 2.3246670409965286e-06,
+      "loss": 0.1522,
+      "step": 2353
+    },
+    {
+      "epoch": 1.0737826434028965,
+      "grad_norm": 0.30660977959632874,
+      "learning_rate": 2.3228245108751402e-06,
+      "loss": 0.1534,
+      "step": 2354
+    },
+    {
+      "epoch": 1.074238795757783,
+      "grad_norm": 0.31357109546661377,
+      "learning_rate": 2.3209820774738777e-06,
+      "loss": 0.1503,
+      "step": 2355
+    },
+    {
+      "epoch": 1.0746949481126697,
+      "grad_norm": 0.29743650555610657,
+      "learning_rate": 2.319139741798525e-06,
+      "loss": 0.1571,
+      "step": 2356
+    },
+    {
+      "epoch": 1.0751511004675562,
+      "grad_norm": 0.3009781241416931,
+      "learning_rate": 2.3172975048548156e-06,
+      "loss": 0.1497,
+      "step": 2357
+    },
+    {
+      "epoch": 1.0756072528224427,
+      "grad_norm": 0.30445727705955505,
+      "learning_rate": 2.315455367648426e-06,
+      "loss": 0.1575,
+      "step": 2358
+    },
+    {
+      "epoch": 1.0760634051773292,
+      "grad_norm": 0.31645891070365906,
+      "learning_rate": 2.3136133311849784e-06,
+      "loss": 0.1636,
+      "step": 2359
+    },
+    {
+      "epoch": 1.0765195575322157,
+      "grad_norm": 0.3001002371311188,
+      "learning_rate": 2.3117713964700415e-06,
+      "loss": 0.1475,
+      "step": 2360
+    },
+    {
+      "epoch": 1.0769757098871022,
+      "grad_norm": 0.30241382122039795,
+      "learning_rate": 2.3099295645091276e-06,
+      "loss": 0.1509,
+      "step": 2361
+    },
+    {
+      "epoch": 1.0774318622419887,
+      "grad_norm": 0.31122294068336487,
+      "learning_rate": 2.308087836307693e-06,
+      "loss": 0.1525,
+      "step": 2362
+    },
+    {
+      "epoch": 1.0778880145968754,
+      "grad_norm": 0.30662044882774353,
+      "learning_rate": 2.3062462128711378e-06,
+      "loss": 0.1569,
+      "step": 2363
+    },
+    {
+      "epoch": 1.078344166951762,
+      "grad_norm": 0.301777184009552,
+      "learning_rate": 2.304404695204804e-06,
+      "loss": 0.1538,
+      "step": 2364
+    },
+    {
+      "epoch": 1.0788003193066484,
+      "grad_norm": 0.3187249004840851,
+      "learning_rate": 2.302563284313976e-06,
+      "loss": 0.1562,
+      "step": 2365
+    },
+    {
+      "epoch": 1.079256471661535,
+      "grad_norm": 0.30848801136016846,
+      "learning_rate": 2.30072198120388e-06,
+      "loss": 0.1507,
+      "step": 2366
+    },
+    {
+      "epoch": 1.0797126240164214,
+      "grad_norm": 0.34001514315605164,
+      "learning_rate": 2.2988807868796857e-06,
+      "loss": 0.165,
+      "step": 2367
+    },
+    {
+      "epoch": 1.080168776371308,
+      "grad_norm": 0.31782227754592896,
+      "learning_rate": 2.2970397023465e-06,
+      "loss": 0.1568,
+      "step": 2368
+    },
+    {
+      "epoch": 1.0806249287261946,
+      "grad_norm": 0.29252395033836365,
+      "learning_rate": 2.295198728609371e-06,
+      "loss": 0.1529,
+      "step": 2369
+    },
+    {
+      "epoch": 1.0810810810810811,
+      "grad_norm": 0.3122186064720154,
+      "learning_rate": 2.293357866673287e-06,
+      "loss": 0.1505,
+      "step": 2370
+    },
+    {
+      "epoch": 1.0815372334359676,
+      "grad_norm": 0.3349948823451996,
+      "learning_rate": 2.291517117543175e-06,
+      "loss": 0.1447,
+      "step": 2371
+    },
+    {
+      "epoch": 1.0819933857908541,
+      "grad_norm": 0.30371904373168945,
+      "learning_rate": 2.289676482223899e-06,
+      "loss": 0.1537,
+      "step": 2372
+    },
+    {
+      "epoch": 1.0824495381457406,
+      "grad_norm": 0.3028498888015747,
+      "learning_rate": 2.2878359617202637e-06,
+      "loss": 0.1543,
+      "step": 2373
+    },
+    {
+      "epoch": 1.0829056905006271,
+      "grad_norm": 0.29874497652053833,
+      "learning_rate": 2.285995557037009e-06,
+      "loss": 0.1557,
+      "step": 2374
+    },
+    {
+      "epoch": 1.0833618428555138,
+      "grad_norm": 0.3105461597442627,
+      "learning_rate": 2.284155269178812e-06,
+      "loss": 0.1484,
+      "step": 2375
+    },
+    {
+      "epoch": 1.0838179952104003,
+      "grad_norm": 0.3080073595046997,
+      "learning_rate": 2.2823150991502844e-06,
+      "loss": 0.158,
+      "step": 2376
+    },
+    {
+      "epoch": 1.0842741475652868,
+      "grad_norm": 0.30038782954216003,
+      "learning_rate": 2.2804750479559777e-06,
+      "loss": 0.151,
+      "step": 2377
+    },
+    {
+      "epoch": 1.0847302999201733,
+      "grad_norm": 0.3226597011089325,
+      "learning_rate": 2.278635116600376e-06,
+      "loss": 0.1611,
+      "step": 2378
+    },
+    {
+      "epoch": 1.0851864522750598,
+      "grad_norm": 0.29122862219810486,
+      "learning_rate": 2.276795306087897e-06,
+      "loss": 0.1469,
+      "step": 2379
+    },
+    {
+      "epoch": 1.0856426046299463,
+      "grad_norm": 0.30449211597442627,
+      "learning_rate": 2.274955617422894e-06,
+      "loss": 0.1527,
+      "step": 2380
+    },
+    {
+      "epoch": 1.0860987569848328,
+      "grad_norm": 0.29464495182037354,
+      "learning_rate": 2.2731160516096534e-06,
+      "loss": 0.1527,
+      "step": 2381
+    },
+    {
+      "epoch": 1.0865549093397195,
+      "grad_norm": 0.3376303017139435,
+      "learning_rate": 2.271276609652395e-06,
+      "loss": 0.1524,
+      "step": 2382
+    },
+    {
+      "epoch": 1.087011061694606,
+      "grad_norm": 0.31834450364112854,
+      "learning_rate": 2.2694372925552686e-06,
+      "loss": 0.1595,
+      "step": 2383
+    },
+    {
+      "epoch": 1.0874672140494925,
+      "grad_norm": 0.307799756526947,
+      "learning_rate": 2.267598101322359e-06,
+      "loss": 0.1583,
+      "step": 2384
+    },
+    {
+      "epoch": 1.087923366404379,
+      "grad_norm": 0.30637145042419434,
+      "learning_rate": 2.2657590369576813e-06,
+      "loss": 0.1497,
+      "step": 2385
+    },
+    {
+      "epoch": 1.0883795187592655,
+      "grad_norm": 0.3328933119773865,
+      "learning_rate": 2.2639201004651796e-06,
+      "loss": 0.1549,
+      "step": 2386
+    },
+    {
+      "epoch": 1.088835671114152,
+      "grad_norm": 0.29612571001052856,
+      "learning_rate": 2.262081292848732e-06,
+      "loss": 0.1435,
+      "step": 2387
+    },
+    {
+      "epoch": 1.0892918234690387,
+      "grad_norm": 0.311734139919281,
+      "learning_rate": 2.2602426151121413e-06,
+      "loss": 0.1649,
+      "step": 2388
+    },
+    {
+      "epoch": 1.0897479758239252,
+      "grad_norm": 0.31515946984291077,
+      "learning_rate": 2.2584040682591436e-06,
+      "loss": 0.172,
+      "step": 2389
+    },
+    {
+      "epoch": 1.0902041281788117,
+      "grad_norm": 0.3302708864212036,
+      "learning_rate": 2.2565656532934016e-06,
+      "loss": 0.1663,
+      "step": 2390
+    },
+    {
+      "epoch": 1.0906602805336982,
+      "grad_norm": 0.3079644441604614,
+      "learning_rate": 2.2547273712185053e-06,
+      "loss": 0.1575,
+      "step": 2391
+    },
+    {
+      "epoch": 1.0911164328885847,
+      "grad_norm": 0.31906506419181824,
+      "learning_rate": 2.2528892230379746e-06,
+      "loss": 0.1532,
+      "step": 2392
+    },
+    {
+      "epoch": 1.0915725852434712,
+      "grad_norm": 0.3005848228931427,
+      "learning_rate": 2.2510512097552547e-06,
+      "loss": 0.1534,
+      "step": 2393
+    },
+    {
+      "epoch": 1.092028737598358,
+      "grad_norm": 0.2951228618621826,
+      "learning_rate": 2.2492133323737157e-06,
+      "loss": 0.1525,
+      "step": 2394
+    },
+    {
+      "epoch": 1.0924848899532444,
+      "grad_norm": 0.3060276508331299,
+      "learning_rate": 2.2473755918966567e-06,
+      "loss": 0.1512,
+      "step": 2395
+    },
+    {
+      "epoch": 1.092941042308131,
+      "grad_norm": 0.31095948815345764,
+      "learning_rate": 2.245537989327299e-06,
+      "loss": 0.1529,
+      "step": 2396
+    },
+    {
+      "epoch": 1.0933971946630174,
+      "grad_norm": 0.29899078607559204,
+      "learning_rate": 2.2437005256687926e-06,
+      "loss": 0.1506,
+      "step": 2397
+    },
+    {
+      "epoch": 1.093853347017904,
+      "grad_norm": 0.29391351342201233,
+      "learning_rate": 2.2418632019242077e-06,
+      "loss": 0.1489,
+      "step": 2398
+    },
+    {
+      "epoch": 1.0943094993727904,
+      "grad_norm": 0.31103986501693726,
+      "learning_rate": 2.240026019096539e-06,
+      "loss": 0.1549,
+      "step": 2399
+    },
+    {
+      "epoch": 1.0947656517276771,
+      "grad_norm": 0.31178927421569824,
+      "learning_rate": 2.2381889781887063e-06,
+      "loss": 0.1563,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0952218040825636,
+      "grad_norm": 0.3093305826187134,
+      "learning_rate": 2.2363520802035498e-06,
+      "loss": 0.1545,
+      "step": 2401
+    },
+    {
+      "epoch": 1.0956779564374501,
+      "grad_norm": 0.33527037501335144,
+      "learning_rate": 2.2345153261438316e-06,
+      "loss": 0.1711,
+      "step": 2402
+    },
+    {
+      "epoch": 1.0961341087923366,
+      "grad_norm": 0.3177525997161865,
+      "learning_rate": 2.232678717012237e-06,
+      "loss": 0.1564,
+      "step": 2403
+    },
+    {
+      "epoch": 1.0965902611472231,
+      "grad_norm": 0.2990678548812866,
+      "learning_rate": 2.2308422538113718e-06,
+      "loss": 0.158,
+      "step": 2404
+    },
+    {
+      "epoch": 1.0970464135021096,
+      "grad_norm": 0.3025558590888977,
+      "learning_rate": 2.22900593754376e-06,
+      "loss": 0.1568,
+      "step": 2405
+    },
+    {
+      "epoch": 1.0975025658569963,
+      "grad_norm": 0.31103450059890747,
+      "learning_rate": 2.2271697692118476e-06,
+      "loss": 0.1577,
+      "step": 2406
+    },
+    {
+      "epoch": 1.0979587182118828,
+      "grad_norm": 0.3023872971534729,
+      "learning_rate": 2.2253337498179984e-06,
+      "loss": 0.1525,
+      "step": 2407
+    },
+    {
+      "epoch": 1.0984148705667693,
+      "grad_norm": 0.3037280738353729,
+      "learning_rate": 2.2234978803644972e-06,
+      "loss": 0.1551,
+      "step": 2408
+    },
+    {
+      "epoch": 1.0988710229216558,
+      "grad_norm": 0.298149049282074,
+      "learning_rate": 2.2216621618535455e-06,
+      "loss": 0.1504,
+      "step": 2409
+    },
+    {
+      "epoch": 1.0993271752765423,
+      "grad_norm": 0.2978970408439636,
+      "learning_rate": 2.2198265952872606e-06,
+      "loss": 0.1516,
+      "step": 2410
+    },
+    {
+      "epoch": 1.0997833276314288,
+      "grad_norm": 0.33059462904930115,
+      "learning_rate": 2.21799118166768e-06,
+      "loss": 0.1625,
+      "step": 2411
+    },
+    {
+      "epoch": 1.1002394799863153,
+      "grad_norm": 0.30958178639411926,
+      "learning_rate": 2.2161559219967555e-06,
+      "loss": 0.1585,
+      "step": 2412
+    },
+    {
+      "epoch": 1.100695632341202,
+      "grad_norm": 0.3242631256580353,
+      "learning_rate": 2.214320817276356e-06,
+      "loss": 0.1596,
+      "step": 2413
+    },
+    {
+      "epoch": 1.1011517846960885,
+      "grad_norm": 0.3253602087497711,
+      "learning_rate": 2.212485868508265e-06,
+      "loss": 0.1518,
+      "step": 2414
+    },
+    {
+      "epoch": 1.101607937050975,
+      "grad_norm": 0.31701716780662537,
+      "learning_rate": 2.210651076694182e-06,
+      "loss": 0.1537,
+      "step": 2415
+    },
+    {
+      "epoch": 1.1020640894058615,
+      "grad_norm": 0.30287981033325195,
+      "learning_rate": 2.2088164428357196e-06,
+      "loss": 0.1541,
+      "step": 2416
+    },
+    {
+      "epoch": 1.102520241760748,
+      "grad_norm": 0.3010953366756439,
+      "learning_rate": 2.206981967934404e-06,
+      "loss": 0.1533,
+      "step": 2417
+    },
+    {
+      "epoch": 1.1029763941156345,
+      "grad_norm": 0.30580252408981323,
+      "learning_rate": 2.2051476529916765e-06,
+      "loss": 0.1501,
+      "step": 2418
+    },
+    {
+      "epoch": 1.1034325464705212,
+      "grad_norm": 0.2954094409942627,
+      "learning_rate": 2.20331349900889e-06,
+      "loss": 0.1439,
+      "step": 2419
+    },
+    {
+      "epoch": 1.1038886988254077,
+      "grad_norm": 0.31143760681152344,
+      "learning_rate": 2.2014795069873092e-06,
+      "loss": 0.1525,
+      "step": 2420
+    },
+    {
+      "epoch": 1.1043448511802942,
+      "grad_norm": 0.29542872309684753,
+      "learning_rate": 2.19964567792811e-06,
+      "loss": 0.1521,
+      "step": 2421
+    },
+    {
+      "epoch": 1.1048010035351807,
+      "grad_norm": 0.31463319063186646,
+      "learning_rate": 2.197812012832381e-06,
+      "loss": 0.158,
+      "step": 2422
+    },
+    {
+      "epoch": 1.1052571558900672,
+      "grad_norm": 0.2935430407524109,
+      "learning_rate": 2.1959785127011195e-06,
+      "loss": 0.1551,
+      "step": 2423
+    },
+    {
+      "epoch": 1.1057133082449537,
+      "grad_norm": 0.32637569308280945,
+      "learning_rate": 2.1941451785352334e-06,
+      "loss": 0.159,
+      "step": 2424
+    },
+    {
+      "epoch": 1.1061694605998404,
+      "grad_norm": 0.3115178048610687,
+      "learning_rate": 2.1923120113355396e-06,
+      "loss": 0.1589,
+      "step": 2425
+    },
+    {
+      "epoch": 1.106625612954727,
+      "grad_norm": 0.3223641812801361,
+      "learning_rate": 2.190479012102766e-06,
+      "loss": 0.1611,
+      "step": 2426
+    },
+    {
+      "epoch": 1.1070817653096134,
+      "grad_norm": 0.3285447359085083,
+      "learning_rate": 2.1886461818375456e-06,
+      "loss": 0.1586,
+      "step": 2427
+    },
+    {
+      "epoch": 1.1075379176645,
+      "grad_norm": 0.29261723160743713,
+      "learning_rate": 2.1868135215404216e-06,
+      "loss": 0.1443,
+      "step": 2428
+    },
+    {
+      "epoch": 1.1079940700193864,
+      "grad_norm": 0.3278425931930542,
+      "learning_rate": 2.184981032211843e-06,
+      "loss": 0.1618,
+      "step": 2429
+    },
+    {
+      "epoch": 1.108450222374273,
+      "grad_norm": 0.3188801407814026,
+      "learning_rate": 2.183148714852167e-06,
+      "loss": 0.1581,
+      "step": 2430
+    },
+    {
+      "epoch": 1.1089063747291594,
+      "grad_norm": 0.33092331886291504,
+      "learning_rate": 2.181316570461655e-06,
+      "loss": 0.1596,
+      "step": 2431
+    },
+    {
+      "epoch": 1.1093625270840461,
+      "grad_norm": 0.2973078787326813,
+      "learning_rate": 2.1794846000404756e-06,
+      "loss": 0.158,
+      "step": 2432
+    },
+    {
+      "epoch": 1.1098186794389326,
+      "grad_norm": 0.31557056307792664,
+      "learning_rate": 2.1776528045887e-06,
+      "loss": 0.16,
+      "step": 2433
+    },
+    {
+      "epoch": 1.1102748317938191,
+      "grad_norm": 0.28650033473968506,
+      "learning_rate": 2.175821185106308e-06,
+      "loss": 0.1429,
+      "step": 2434
+    },
+    {
+      "epoch": 1.1107309841487056,
+      "grad_norm": 0.2963087558746338,
+      "learning_rate": 2.173989742593179e-06,
+      "loss": 0.144,
+      "step": 2435
+    },
+    {
+      "epoch": 1.1111871365035921,
+      "grad_norm": 0.3010818064212799,
+      "learning_rate": 2.1721584780490988e-06,
+      "loss": 0.1495,
+      "step": 2436
+    },
+    {
+      "epoch": 1.1116432888584786,
+      "grad_norm": 0.3129344880580902,
+      "learning_rate": 2.170327392473754e-06,
+      "loss": 0.155,
+      "step": 2437
+    },
+    {
+      "epoch": 1.1120994412133653,
+      "grad_norm": 0.32375624775886536,
+      "learning_rate": 2.168496486866736e-06,
+      "loss": 0.1618,
+      "step": 2438
+    },
+    {
+      "epoch": 1.1125555935682518,
+      "grad_norm": 0.32489460706710815,
+      "learning_rate": 2.166665762227536e-06,
+      "loss": 0.1566,
+      "step": 2439
+    },
+    {
+      "epoch": 1.1130117459231383,
+      "grad_norm": 0.3014304041862488,
+      "learning_rate": 2.1648352195555456e-06,
+      "loss": 0.1602,
+      "step": 2440
+    },
+    {
+      "epoch": 1.1134678982780248,
+      "grad_norm": 0.3152945935726166,
+      "learning_rate": 2.16300485985006e-06,
+      "loss": 0.1569,
+      "step": 2441
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.3113345205783844,
+      "learning_rate": 2.161174684110272e-06,
+      "loss": 0.1502,
+      "step": 2442
+    },
+    {
+      "epoch": 1.1143802029877978,
+      "grad_norm": 0.3059302866458893,
+      "learning_rate": 2.159344693335275e-06,
+      "loss": 0.1523,
+      "step": 2443
+    },
+    {
+      "epoch": 1.1148363553426845,
+      "grad_norm": 0.33054935932159424,
+      "learning_rate": 2.157514888524061e-06,
+      "loss": 0.1504,
+      "step": 2444
+    },
+    {
+      "epoch": 1.115292507697571,
+      "grad_norm": 0.29914331436157227,
+      "learning_rate": 2.1556852706755203e-06,
+      "loss": 0.1558,
+      "step": 2445
+    },
+    {
+      "epoch": 1.1157486600524575,
+      "grad_norm": 0.28529635071754456,
+      "learning_rate": 2.153855840788443e-06,
+      "loss": 0.1488,
+      "step": 2446
+    },
+    {
+      "epoch": 1.116204812407344,
+      "grad_norm": 0.3082621395587921,
+      "learning_rate": 2.1520265998615137e-06,
+      "loss": 0.157,
+      "step": 2447
+    },
+    {
+      "epoch": 1.1166609647622305,
+      "grad_norm": 0.30169448256492615,
+      "learning_rate": 2.150197548893315e-06,
+      "loss": 0.1525,
+      "step": 2448
+    },
+    {
+      "epoch": 1.117117117117117,
+      "grad_norm": 0.3060326874256134,
+      "learning_rate": 2.148368688882328e-06,
+      "loss": 0.1559,
+      "step": 2449
+    },
+    {
+      "epoch": 1.1175732694720037,
+      "grad_norm": 0.29175877571105957,
+      "learning_rate": 2.146540020826927e-06,
+      "loss": 0.1467,
+      "step": 2450
+    },
+    {
+      "epoch": 1.1180294218268902,
+      "grad_norm": 0.31199726462364197,
+      "learning_rate": 2.144711545725381e-06,
+      "loss": 0.1572,
+      "step": 2451
+    },
+    {
+      "epoch": 1.1184855741817767,
+      "grad_norm": 0.3203292787075043,
+      "learning_rate": 2.142883264575856e-06,
+      "loss": 0.1596,
+      "step": 2452
+    },
+    {
+      "epoch": 1.1189417265366632,
+      "grad_norm": 0.3079957962036133,
+      "learning_rate": 2.1410551783764104e-06,
+      "loss": 0.1572,
+      "step": 2453
+    },
+    {
+      "epoch": 1.1193978788915497,
+      "grad_norm": 0.32759830355644226,
+      "learning_rate": 2.1392272881249976e-06,
+      "loss": 0.1579,
+      "step": 2454
+    },
+    {
+      "epoch": 1.1198540312464362,
+      "grad_norm": 0.31564974784851074,
+      "learning_rate": 2.1373995948194613e-06,
+      "loss": 0.1578,
+      "step": 2455
+    },
+    {
+      "epoch": 1.120310183601323,
+      "grad_norm": 0.32277682423591614,
+      "learning_rate": 2.1355720994575414e-06,
+      "loss": 0.159,
+      "step": 2456
+    },
+    {
+      "epoch": 1.1207663359562094,
+      "grad_norm": 0.3213791847229004,
+      "learning_rate": 2.1337448030368675e-06,
+      "loss": 0.1576,
+      "step": 2457
+    },
+    {
+      "epoch": 1.121222488311096,
+      "grad_norm": 0.30738502740859985,
+      "learning_rate": 2.1319177065549594e-06,
+      "loss": 0.1553,
+      "step": 2458
+    },
+    {
+      "epoch": 1.1216786406659824,
+      "grad_norm": 0.3020482063293457,
+      "learning_rate": 2.130090811009231e-06,
+      "loss": 0.1501,
+      "step": 2459
+    },
+    {
+      "epoch": 1.122134793020869,
+      "grad_norm": 0.2827683985233307,
+      "learning_rate": 2.1282641173969855e-06,
+      "loss": 0.1447,
+      "step": 2460
+    },
+    {
+      "epoch": 1.1225909453757554,
+      "grad_norm": 0.294086217880249,
+      "learning_rate": 2.1264376267154138e-06,
+      "loss": 0.1534,
+      "step": 2461
+    },
+    {
+      "epoch": 1.123047097730642,
+      "grad_norm": 0.3120155334472656,
+      "learning_rate": 2.1246113399615977e-06,
+      "loss": 0.1498,
+      "step": 2462
+    },
+    {
+      "epoch": 1.1235032500855286,
+      "grad_norm": 0.32669350504875183,
+      "learning_rate": 2.1227852581325076e-06,
+      "loss": 0.1624,
+      "step": 2463
+    },
+    {
+      "epoch": 1.1239594024404151,
+      "grad_norm": 0.3262674808502197,
+      "learning_rate": 2.120959382225002e-06,
+      "loss": 0.1586,
+      "step": 2464
+    },
+    {
+      "epoch": 1.1244155547953016,
+      "grad_norm": 0.306515634059906,
+      "learning_rate": 2.1191337132358262e-06,
+      "loss": 0.1498,
+      "step": 2465
+    },
+    {
+      "epoch": 1.1248717071501881,
+      "grad_norm": 0.31069692969322205,
+      "learning_rate": 2.117308252161614e-06,
+      "loss": 0.1489,
+      "step": 2466
+    },
+    {
+      "epoch": 1.1253278595050746,
+      "grad_norm": 0.3287632465362549,
+      "learning_rate": 2.1154829999988845e-06,
+      "loss": 0.1558,
+      "step": 2467
+    },
+    {
+      "epoch": 1.1257840118599611,
+      "grad_norm": 0.2938937842845917,
+      "learning_rate": 2.113657957744042e-06,
+      "loss": 0.1435,
+      "step": 2468
+    },
+    {
+      "epoch": 1.1262401642148478,
+      "grad_norm": 0.3082640469074249,
+      "learning_rate": 2.11183312639338e-06,
+      "loss": 0.1505,
+      "step": 2469
+    },
+    {
+      "epoch": 1.1266963165697343,
+      "grad_norm": 0.31762850284576416,
+      "learning_rate": 2.1100085069430715e-06,
+      "loss": 0.159,
+      "step": 2470
+    },
+    {
+      "epoch": 1.1271524689246208,
+      "grad_norm": 0.3084609806537628,
+      "learning_rate": 2.1081841003891785e-06,
+      "loss": 0.1523,
+      "step": 2471
+    },
+    {
+      "epoch": 1.1276086212795073,
+      "grad_norm": 0.3295886814594269,
+      "learning_rate": 2.106359907727644e-06,
+      "loss": 0.1607,
+      "step": 2472
+    },
+    {
+      "epoch": 1.1280647736343938,
+      "grad_norm": 0.3112865090370178,
+      "learning_rate": 2.104535929954295e-06,
+      "loss": 0.1623,
+      "step": 2473
+    },
+    {
+      "epoch": 1.1285209259892803,
+      "grad_norm": 0.30768048763275146,
+      "learning_rate": 2.102712168064841e-06,
+      "loss": 0.1545,
+      "step": 2474
+    },
+    {
+      "epoch": 1.128977078344167,
+      "grad_norm": 0.2967469394207001,
+      "learning_rate": 2.1008886230548746e-06,
+      "loss": 0.1573,
+      "step": 2475
+    },
+    {
+      "epoch": 1.1294332306990535,
+      "grad_norm": 0.3131488859653473,
+      "learning_rate": 2.099065295919869e-06,
+      "loss": 0.1481,
+      "step": 2476
+    },
+    {
+      "epoch": 1.12988938305394,
+      "grad_norm": 0.3084881901741028,
+      "learning_rate": 2.0972421876551777e-06,
+      "loss": 0.1539,
+      "step": 2477
+    },
+    {
+      "epoch": 1.1303455354088265,
+      "grad_norm": 0.31541505455970764,
+      "learning_rate": 2.095419299256036e-06,
+      "loss": 0.1579,
+      "step": 2478
+    },
+    {
+      "epoch": 1.130801687763713,
+      "grad_norm": 0.31113600730895996,
+      "learning_rate": 2.0935966317175618e-06,
+      "loss": 0.1594,
+      "step": 2479
+    },
+    {
+      "epoch": 1.1312578401185995,
+      "grad_norm": 0.298733115196228,
+      "learning_rate": 2.0917741860347463e-06,
+      "loss": 0.1524,
+      "step": 2480
+    },
+    {
+      "epoch": 1.131713992473486,
+      "grad_norm": 0.3065612018108368,
+      "learning_rate": 2.0899519632024644e-06,
+      "loss": 0.1577,
+      "step": 2481
+    },
+    {
+      "epoch": 1.1321701448283727,
+      "grad_norm": 0.3024389147758484,
+      "learning_rate": 2.0881299642154676e-06,
+      "loss": 0.1513,
+      "step": 2482
+    },
+    {
+      "epoch": 1.1326262971832592,
+      "grad_norm": 0.2894464135169983,
+      "learning_rate": 2.0863081900683853e-06,
+      "loss": 0.1459,
+      "step": 2483
+    },
+    {
+      "epoch": 1.1330824495381457,
+      "grad_norm": 0.3041086494922638,
+      "learning_rate": 2.0844866417557243e-06,
+      "loss": 0.1516,
+      "step": 2484
+    },
+    {
+      "epoch": 1.1335386018930322,
+      "grad_norm": 0.3138040006160736,
+      "learning_rate": 2.082665320271867e-06,
+      "loss": 0.1529,
+      "step": 2485
+    },
+    {
+      "epoch": 1.1339947542479187,
+      "grad_norm": 0.3205288350582123,
+      "learning_rate": 2.0808442266110756e-06,
+      "loss": 0.1606,
+      "step": 2486
+    },
+    {
+      "epoch": 1.1344509066028055,
+      "grad_norm": 0.32108408212661743,
+      "learning_rate": 2.0790233617674833e-06,
+      "loss": 0.1571,
+      "step": 2487
+    },
+    {
+      "epoch": 1.134907058957692,
+      "grad_norm": 0.305132657289505,
+      "learning_rate": 2.0772027267351e-06,
+      "loss": 0.1605,
+      "step": 2488
+    },
+    {
+      "epoch": 1.1353632113125784,
+      "grad_norm": 0.3122641146183014,
+      "learning_rate": 2.0753823225078124e-06,
+      "loss": 0.1528,
+      "step": 2489
+    },
+    {
+      "epoch": 1.135819363667465,
+      "grad_norm": 0.3220573663711548,
+      "learning_rate": 2.0735621500793798e-06,
+      "loss": 0.1695,
+      "step": 2490
+    },
+    {
+      "epoch": 1.1362755160223514,
+      "grad_norm": 0.31758105754852295,
+      "learning_rate": 2.0717422104434328e-06,
+      "loss": 0.1494,
+      "step": 2491
+    },
+    {
+      "epoch": 1.136731668377238,
+      "grad_norm": 0.33168843388557434,
+      "learning_rate": 2.069922504593478e-06,
+      "loss": 0.1613,
+      "step": 2492
+    },
+    {
+      "epoch": 1.1371878207321244,
+      "grad_norm": 0.3141818046569824,
+      "learning_rate": 2.0681030335228917e-06,
+      "loss": 0.1532,
+      "step": 2493
+    },
+    {
+      "epoch": 1.1376439730870112,
+      "grad_norm": 0.30216023325920105,
+      "learning_rate": 2.0662837982249253e-06,
+      "loss": 0.1452,
+      "step": 2494
+    },
+    {
+      "epoch": 1.1381001254418976,
+      "grad_norm": 0.29277002811431885,
+      "learning_rate": 2.0644647996926982e-06,
+      "loss": 0.148,
+      "step": 2495
+    },
+    {
+      "epoch": 1.1385562777967841,
+      "grad_norm": 0.3166434168815613,
+      "learning_rate": 2.062646038919202e-06,
+      "loss": 0.1542,
+      "step": 2496
+    },
+    {
+      "epoch": 1.1390124301516706,
+      "grad_norm": 0.3139345943927765,
+      "learning_rate": 2.0608275168973e-06,
+      "loss": 0.1542,
+      "step": 2497
+    },
+    {
+      "epoch": 1.1394685825065571,
+      "grad_norm": 0.30859139561653137,
+      "learning_rate": 2.0590092346197223e-06,
+      "loss": 0.1483,
+      "step": 2498
+    },
+    {
+      "epoch": 1.1399247348614436,
+      "grad_norm": 0.2961941361427307,
+      "learning_rate": 2.0571911930790685e-06,
+      "loss": 0.1501,
+      "step": 2499
+    },
+    {
+      "epoch": 1.1403808872163301,
+      "grad_norm": 0.3011581301689148,
+      "learning_rate": 2.0553733932678096e-06,
+      "loss": 0.1545,
+      "step": 2500
+    },
+    {
+      "epoch": 1.1408370395712168,
+      "grad_norm": 0.2961842119693756,
+      "learning_rate": 2.0535558361782827e-06,
+      "loss": 0.1437,
+      "step": 2501
+    },
+    {
+      "epoch": 1.1412931919261033,
+      "grad_norm": 0.3130788505077362,
+      "learning_rate": 2.0517385228026916e-06,
+      "loss": 0.1528,
+      "step": 2502
+    },
+    {
+      "epoch": 1.1417493442809898,
+      "grad_norm": 0.308366596698761,
+      "learning_rate": 2.0499214541331084e-06,
+      "loss": 0.1548,
+      "step": 2503
+    },
+    {
+      "epoch": 1.1422054966358763,
+      "grad_norm": 0.32291486859321594,
+      "learning_rate": 2.048104631161471e-06,
+      "loss": 0.1525,
+      "step": 2504
+    },
+    {
+      "epoch": 1.1426616489907628,
+      "grad_norm": 0.3222767114639282,
+      "learning_rate": 2.046288054879584e-06,
+      "loss": 0.1573,
+      "step": 2505
+    },
+    {
+      "epoch": 1.1431178013456496,
+      "grad_norm": 0.3051936626434326,
+      "learning_rate": 2.0444717262791154e-06,
+      "loss": 0.1571,
+      "step": 2506
+    },
+    {
+      "epoch": 1.143573953700536,
+      "grad_norm": 0.29791855812072754,
+      "learning_rate": 2.0426556463516e-06,
+      "loss": 0.1508,
+      "step": 2507
+    },
+    {
+      "epoch": 1.1440301060554225,
+      "grad_norm": 0.31174027919769287,
+      "learning_rate": 2.040839816088436e-06,
+      "loss": 0.1478,
+      "step": 2508
+    },
+    {
+      "epoch": 1.144486258410309,
+      "grad_norm": 0.31370431184768677,
+      "learning_rate": 2.039024236480885e-06,
+      "loss": 0.161,
+      "step": 2509
+    },
+    {
+      "epoch": 1.1449424107651955,
+      "grad_norm": 0.3067293167114258,
+      "learning_rate": 2.0372089085200734e-06,
+      "loss": 0.1502,
+      "step": 2510
+    },
+    {
+      "epoch": 1.145398563120082,
+      "grad_norm": 0.30066871643066406,
+      "learning_rate": 2.035393833196988e-06,
+      "loss": 0.1448,
+      "step": 2511
+    },
+    {
+      "epoch": 1.1458547154749685,
+      "grad_norm": 0.29360148310661316,
+      "learning_rate": 2.0335790115024787e-06,
+      "loss": 0.1545,
+      "step": 2512
+    },
+    {
+      "epoch": 1.1463108678298553,
+      "grad_norm": 0.3109211325645447,
+      "learning_rate": 2.0317644444272573e-06,
+      "loss": 0.1621,
+      "step": 2513
+    },
+    {
+      "epoch": 1.1467670201847417,
+      "grad_norm": 0.32034653425216675,
+      "learning_rate": 2.0299501329618965e-06,
+      "loss": 0.1614,
+      "step": 2514
+    },
+    {
+      "epoch": 1.1472231725396282,
+      "grad_norm": 0.3384312093257904,
+      "learning_rate": 2.028136078096828e-06,
+      "loss": 0.1592,
+      "step": 2515
+    },
+    {
+      "epoch": 1.1476793248945147,
+      "grad_norm": 0.29340147972106934,
+      "learning_rate": 2.026322280822346e-06,
+      "loss": 0.1422,
+      "step": 2516
+    },
+    {
+      "epoch": 1.1481354772494012,
+      "grad_norm": 0.3106522858142853,
+      "learning_rate": 2.0245087421286025e-06,
+      "loss": 0.1589,
+      "step": 2517
+    },
+    {
+      "epoch": 1.148591629604288,
+      "grad_norm": 0.3177458941936493,
+      "learning_rate": 2.0226954630056072e-06,
+      "loss": 0.1512,
+      "step": 2518
+    },
+    {
+      "epoch": 1.1490477819591745,
+      "grad_norm": 0.3203088343143463,
+      "learning_rate": 2.02088244444323e-06,
+      "loss": 0.1586,
+      "step": 2519
+    },
+    {
+      "epoch": 1.149503934314061,
+      "grad_norm": 0.3190899193286896,
+      "learning_rate": 2.019069687431199e-06,
+      "loss": 0.1633,
+      "step": 2520
+    },
+    {
+      "epoch": 1.1499600866689474,
+      "grad_norm": 0.3084123432636261,
+      "learning_rate": 2.017257192959099e-06,
+      "loss": 0.151,
+      "step": 2521
+    },
+    {
+      "epoch": 1.150416239023834,
+      "grad_norm": 0.3053361177444458,
+      "learning_rate": 2.0154449620163694e-06,
+      "loss": 0.1576,
+      "step": 2522
+    },
+    {
+      "epoch": 1.1508723913787204,
+      "grad_norm": 0.32078465819358826,
+      "learning_rate": 2.0136329955923073e-06,
+      "loss": 0.1539,
+      "step": 2523
+    },
+    {
+      "epoch": 1.151328543733607,
+      "grad_norm": 0.3125305473804474,
+      "learning_rate": 2.0118212946760664e-06,
+      "loss": 0.1482,
+      "step": 2524
+    },
+    {
+      "epoch": 1.1517846960884937,
+      "grad_norm": 0.30333492159843445,
+      "learning_rate": 2.0100098602566544e-06,
+      "loss": 0.1539,
+      "step": 2525
+    },
+    {
+      "epoch": 1.1522408484433802,
+      "grad_norm": 0.31321239471435547,
+      "learning_rate": 2.008198693322933e-06,
+      "loss": 0.1539,
+      "step": 2526
+    },
+    {
+      "epoch": 1.1526970007982666,
+      "grad_norm": 0.3032292425632477,
+      "learning_rate": 2.006387794863619e-06,
+      "loss": 0.1546,
+      "step": 2527
+    },
+    {
+      "epoch": 1.1531531531531531,
+      "grad_norm": 0.3145170509815216,
+      "learning_rate": 2.004577165867282e-06,
+      "loss": 0.152,
+      "step": 2528
+    },
+    {
+      "epoch": 1.1536093055080396,
+      "grad_norm": 0.3092341721057892,
+      "learning_rate": 2.002766807322344e-06,
+      "loss": 0.1551,
+      "step": 2529
+    },
+    {
+      "epoch": 1.1540654578629261,
+      "grad_norm": 0.32281920313835144,
+      "learning_rate": 2.000956720217081e-06,
+      "loss": 0.1561,
+      "step": 2530
+    },
+    {
+      "epoch": 1.1545216102178126,
+      "grad_norm": 0.300767183303833,
+      "learning_rate": 1.99914690553962e-06,
+      "loss": 0.1523,
+      "step": 2531
+    },
+    {
+      "epoch": 1.1549777625726994,
+      "grad_norm": 0.310223251581192,
+      "learning_rate": 1.9973373642779385e-06,
+      "loss": 0.1513,
+      "step": 2532
+    },
+    {
+      "epoch": 1.1554339149275858,
+      "grad_norm": 0.3141593933105469,
+      "learning_rate": 1.9955280974198653e-06,
+      "loss": 0.1543,
+      "step": 2533
+    },
+    {
+      "epoch": 1.1558900672824723,
+      "grad_norm": 0.3046564757823944,
+      "learning_rate": 1.9937191059530796e-06,
+      "loss": 0.1577,
+      "step": 2534
+    },
+    {
+      "epoch": 1.1563462196373588,
+      "grad_norm": 0.314810574054718,
+      "learning_rate": 1.99191039086511e-06,
+      "loss": 0.1671,
+      "step": 2535
+    },
+    {
+      "epoch": 1.1568023719922453,
+      "grad_norm": 0.3175409138202667,
+      "learning_rate": 1.9901019531433345e-06,
+      "loss": 0.1527,
+      "step": 2536
+    },
+    {
+      "epoch": 1.157258524347132,
+      "grad_norm": 0.32860544323921204,
+      "learning_rate": 1.988293793774979e-06,
+      "loss": 0.1585,
+      "step": 2537
+    },
+    {
+      "epoch": 1.1577146767020186,
+      "grad_norm": 0.3256966471672058,
+      "learning_rate": 1.9864859137471195e-06,
+      "loss": 0.1541,
+      "step": 2538
+    },
+    {
+      "epoch": 1.158170829056905,
+      "grad_norm": 0.3185851275920868,
+      "learning_rate": 1.984678314046676e-06,
+      "loss": 0.1523,
+      "step": 2539
+    },
+    {
+      "epoch": 1.1586269814117915,
+      "grad_norm": 0.31023722887039185,
+      "learning_rate": 1.9828709956604176e-06,
+      "loss": 0.1516,
+      "step": 2540
+    },
+    {
+      "epoch": 1.159083133766678,
+      "grad_norm": 0.31944340467453003,
+      "learning_rate": 1.981063959574962e-06,
+      "loss": 0.1547,
+      "step": 2541
+    },
+    {
+      "epoch": 1.1595392861215645,
+      "grad_norm": 0.3001713156700134,
+      "learning_rate": 1.9792572067767676e-06,
+      "loss": 0.1556,
+      "step": 2542
+    },
+    {
+      "epoch": 1.159995438476451,
+      "grad_norm": 0.31969380378723145,
+      "learning_rate": 1.9774507382521434e-06,
+      "loss": 0.1565,
+      "step": 2543
+    },
+    {
+      "epoch": 1.1604515908313378,
+      "grad_norm": 0.3322685658931732,
+      "learning_rate": 1.9756445549872396e-06,
+      "loss": 0.1549,
+      "step": 2544
+    },
+    {
+      "epoch": 1.1609077431862243,
+      "grad_norm": 0.3241439759731293,
+      "learning_rate": 1.9738386579680515e-06,
+      "loss": 0.1594,
+      "step": 2545
+    },
+    {
+      "epoch": 1.1613638955411107,
+      "grad_norm": 0.309921532869339,
+      "learning_rate": 1.9720330481804197e-06,
+      "loss": 0.1529,
+      "step": 2546
+    },
+    {
+      "epoch": 1.1618200478959972,
+      "grad_norm": 0.3175309896469116,
+      "learning_rate": 1.970227726610026e-06,
+      "loss": 0.1537,
+      "step": 2547
+    },
+    {
+      "epoch": 1.1622762002508837,
+      "grad_norm": 0.30349457263946533,
+      "learning_rate": 1.968422694242396e-06,
+      "loss": 0.1605,
+      "step": 2548
+    },
+    {
+      "epoch": 1.1627323526057702,
+      "grad_norm": 0.30369940400123596,
+      "learning_rate": 1.966617952062897e-06,
+      "loss": 0.1531,
+      "step": 2549
+    },
+    {
+      "epoch": 1.163188504960657,
+      "grad_norm": 0.32541221380233765,
+      "learning_rate": 1.9648135010567376e-06,
+      "loss": 0.1668,
+      "step": 2550
+    },
+    {
+      "epoch": 1.1636446573155435,
+      "grad_norm": 0.31525540351867676,
+      "learning_rate": 1.9630093422089697e-06,
+      "loss": 0.1557,
+      "step": 2551
+    },
+    {
+      "epoch": 1.16410080967043,
+      "grad_norm": 0.2928027808666229,
+      "learning_rate": 1.961205476504482e-06,
+      "loss": 0.1527,
+      "step": 2552
+    },
+    {
+      "epoch": 1.1645569620253164,
+      "grad_norm": 0.313533216714859,
+      "learning_rate": 1.959401904928006e-06,
+      "loss": 0.1521,
+      "step": 2553
+    },
+    {
+      "epoch": 1.165013114380203,
+      "grad_norm": 0.3197345435619354,
+      "learning_rate": 1.957598628464111e-06,
+      "loss": 0.1644,
+      "step": 2554
+    },
+    {
+      "epoch": 1.1654692667350894,
+      "grad_norm": 0.3331969082355499,
+      "learning_rate": 1.9557956480972066e-06,
+      "loss": 0.1601,
+      "step": 2555
+    },
+    {
+      "epoch": 1.1659254190899762,
+      "grad_norm": 0.3225902020931244,
+      "learning_rate": 1.9539929648115395e-06,
+      "loss": 0.1496,
+      "step": 2556
+    },
+    {
+      "epoch": 1.1663815714448627,
+      "grad_norm": 0.3071746230125427,
+      "learning_rate": 1.9521905795911955e-06,
+      "loss": 0.1532,
+      "step": 2557
+    },
+    {
+      "epoch": 1.1668377237997491,
+      "grad_norm": 0.3109276294708252,
+      "learning_rate": 1.9503884934200962e-06,
+      "loss": 0.1509,
+      "step": 2558
+    },
+    {
+      "epoch": 1.1672938761546356,
+      "grad_norm": 0.31990760564804077,
+      "learning_rate": 1.9485867072820006e-06,
+      "loss": 0.1618,
+      "step": 2559
+    },
+    {
+      "epoch": 1.1677500285095221,
+      "grad_norm": 0.31767144799232483,
+      "learning_rate": 1.9467852221605044e-06,
+      "loss": 0.1559,
+      "step": 2560
+    },
+    {
+      "epoch": 1.1682061808644086,
+      "grad_norm": 0.3203451931476593,
+      "learning_rate": 1.9449840390390383e-06,
+      "loss": 0.1608,
+      "step": 2561
+    },
+    {
+      "epoch": 1.1686623332192951,
+      "grad_norm": 0.3119843006134033,
+      "learning_rate": 1.94318315890087e-06,
+      "loss": 0.1582,
+      "step": 2562
+    },
+    {
+      "epoch": 1.1691184855741819,
+      "grad_norm": 0.28726524114608765,
+      "learning_rate": 1.9413825827290983e-06,
+      "loss": 0.1474,
+      "step": 2563
+    },
+    {
+      "epoch": 1.1695746379290684,
+      "grad_norm": 0.29348814487457275,
+      "learning_rate": 1.939582311506658e-06,
+      "loss": 0.1577,
+      "step": 2564
+    },
+    {
+      "epoch": 1.1700307902839548,
+      "grad_norm": 0.31846269965171814,
+      "learning_rate": 1.9377823462163195e-06,
+      "loss": 0.1555,
+      "step": 2565
+    },
+    {
+      "epoch": 1.1704869426388413,
+      "grad_norm": 0.3218359649181366,
+      "learning_rate": 1.9359826878406825e-06,
+      "loss": 0.1605,
+      "step": 2566
+    },
+    {
+      "epoch": 1.1709430949937278,
+      "grad_norm": 0.30212482810020447,
+      "learning_rate": 1.93418333736218e-06,
+      "loss": 0.15,
+      "step": 2567
+    },
+    {
+      "epoch": 1.1713992473486146,
+      "grad_norm": 0.32369521260261536,
+      "learning_rate": 1.9323842957630795e-06,
+      "loss": 0.1632,
+      "step": 2568
+    },
+    {
+      "epoch": 1.171855399703501,
+      "grad_norm": 0.3052249252796173,
+      "learning_rate": 1.9305855640254774e-06,
+      "loss": 0.1535,
+      "step": 2569
+    },
+    {
+      "epoch": 1.1723115520583876,
+      "grad_norm": 0.3070298433303833,
+      "learning_rate": 1.9287871431313004e-06,
+      "loss": 0.1557,
+      "step": 2570
+    },
+    {
+      "epoch": 1.172767704413274,
+      "grad_norm": 0.3102405369281769,
+      "learning_rate": 1.9269890340623088e-06,
+      "loss": 0.155,
+      "step": 2571
+    },
+    {
+      "epoch": 1.1732238567681605,
+      "grad_norm": 0.29954037070274353,
+      "learning_rate": 1.925191237800089e-06,
+      "loss": 0.1477,
+      "step": 2572
+    },
+    {
+      "epoch": 1.173680009123047,
+      "grad_norm": 0.33765316009521484,
+      "learning_rate": 1.92339375532606e-06,
+      "loss": 0.1608,
+      "step": 2573
+    },
+    {
+      "epoch": 1.1741361614779335,
+      "grad_norm": 0.3113259971141815,
+      "learning_rate": 1.9215965876214663e-06,
+      "loss": 0.1553,
+      "step": 2574
+    },
+    {
+      "epoch": 1.1745923138328203,
+      "grad_norm": 0.3262672424316406,
+      "learning_rate": 1.9197997356673817e-06,
+      "loss": 0.1557,
+      "step": 2575
+    },
+    {
+      "epoch": 1.1750484661877068,
+      "grad_norm": 0.32911545038223267,
+      "learning_rate": 1.9180032004447095e-06,
+      "loss": 0.1521,
+      "step": 2576
+    },
+    {
+      "epoch": 1.1755046185425932,
+      "grad_norm": 0.29119279980659485,
+      "learning_rate": 1.9162069829341775e-06,
+      "loss": 0.1481,
+      "step": 2577
+    },
+    {
+      "epoch": 1.1759607708974797,
+      "grad_norm": 0.29823169112205505,
+      "learning_rate": 1.9144110841163406e-06,
+      "loss": 0.1483,
+      "step": 2578
+    },
+    {
+      "epoch": 1.1764169232523662,
+      "grad_norm": 0.3086245656013489,
+      "learning_rate": 1.9126155049715816e-06,
+      "loss": 0.1528,
+      "step": 2579
+    },
+    {
+      "epoch": 1.1768730756072527,
+      "grad_norm": 0.3069640100002289,
+      "learning_rate": 1.910820246480106e-06,
+      "loss": 0.1544,
+      "step": 2580
+    },
+    {
+      "epoch": 1.1773292279621392,
+      "grad_norm": 0.3237389028072357,
+      "learning_rate": 1.909025309621947e-06,
+      "loss": 0.1525,
+      "step": 2581
+    },
+    {
+      "epoch": 1.177785380317026,
+      "grad_norm": 0.31238242983818054,
+      "learning_rate": 1.907230695376961e-06,
+      "loss": 0.1567,
+      "step": 2582
+    },
+    {
+      "epoch": 1.1782415326719125,
+      "grad_norm": 0.32844918966293335,
+      "learning_rate": 1.9054364047248263e-06,
+      "loss": 0.1624,
+      "step": 2583
+    },
+    {
+      "epoch": 1.178697685026799,
+      "grad_norm": 0.3161335289478302,
+      "learning_rate": 1.9036424386450487e-06,
+      "loss": 0.1608,
+      "step": 2584
+    },
+    {
+      "epoch": 1.1791538373816854,
+      "grad_norm": 0.3095495104789734,
+      "learning_rate": 1.9018487981169532e-06,
+      "loss": 0.1552,
+      "step": 2585
+    },
+    {
+      "epoch": 1.179609989736572,
+      "grad_norm": 0.32562121748924255,
+      "learning_rate": 1.9000554841196884e-06,
+      "loss": 0.168,
+      "step": 2586
+    },
+    {
+      "epoch": 1.1800661420914587,
+      "grad_norm": 0.3023471534252167,
+      "learning_rate": 1.8982624976322253e-06,
+      "loss": 0.1558,
+      "step": 2587
+    },
+    {
+      "epoch": 1.1805222944463452,
+      "grad_norm": 0.3121137320995331,
+      "learning_rate": 1.8964698396333555e-06,
+      "loss": 0.151,
+      "step": 2588
+    },
+    {
+      "epoch": 1.1809784468012317,
+      "grad_norm": 0.3004269301891327,
+      "learning_rate": 1.8946775111016904e-06,
+      "loss": 0.1477,
+      "step": 2589
+    },
+    {
+      "epoch": 1.1814345991561181,
+      "grad_norm": 0.3145170211791992,
+      "learning_rate": 1.8928855130156635e-06,
+      "loss": 0.1583,
+      "step": 2590
+    },
+    {
+      "epoch": 1.1818907515110046,
+      "grad_norm": 0.3035140931606293,
+      "learning_rate": 1.8910938463535252e-06,
+      "loss": 0.1502,
+      "step": 2591
+    },
+    {
+      "epoch": 1.1823469038658911,
+      "grad_norm": 0.31445202231407166,
+      "learning_rate": 1.8893025120933488e-06,
+      "loss": 0.1569,
+      "step": 2592
+    },
+    {
+      "epoch": 1.1828030562207776,
+      "grad_norm": 0.31155338883399963,
+      "learning_rate": 1.8875115112130231e-06,
+      "loss": 0.1619,
+      "step": 2593
+    },
+    {
+      "epoch": 1.1832592085756644,
+      "grad_norm": 0.2908079922199249,
+      "learning_rate": 1.8857208446902548e-06,
+      "loss": 0.1471,
+      "step": 2594
+    },
+    {
+      "epoch": 1.1837153609305509,
+      "grad_norm": 0.312664657831192,
+      "learning_rate": 1.8839305135025704e-06,
+      "loss": 0.1561,
+      "step": 2595
+    },
+    {
+      "epoch": 1.1841715132854373,
+      "grad_norm": 0.3114558756351471,
+      "learning_rate": 1.882140518627311e-06,
+      "loss": 0.1542,
+      "step": 2596
+    },
+    {
+      "epoch": 1.1846276656403238,
+      "grad_norm": 0.309690922498703,
+      "learning_rate": 1.8803508610416352e-06,
+      "loss": 0.1518,
+      "step": 2597
+    },
+    {
+      "epoch": 1.1850838179952103,
+      "grad_norm": 0.3130999803543091,
+      "learning_rate": 1.8785615417225179e-06,
+      "loss": 0.1601,
+      "step": 2598
+    },
+    {
+      "epoch": 1.1855399703500968,
+      "grad_norm": 0.28748223185539246,
+      "learning_rate": 1.8767725616467487e-06,
+      "loss": 0.1479,
+      "step": 2599
+    },
+    {
+      "epoch": 1.1859961227049836,
+      "grad_norm": 0.3250840902328491,
+      "learning_rate": 1.874983921790931e-06,
+      "loss": 0.1537,
+      "step": 2600
+    },
+    {
+      "epoch": 1.18645227505987,
+      "grad_norm": 0.306423544883728,
+      "learning_rate": 1.8731956231314843e-06,
+      "loss": 0.1581,
+      "step": 2601
+    },
+    {
+      "epoch": 1.1869084274147566,
+      "grad_norm": 0.3731411099433899,
+      "learning_rate": 1.8714076666446413e-06,
+      "loss": 0.1561,
+      "step": 2602
+    },
+    {
+      "epoch": 1.187364579769643,
+      "grad_norm": 0.29688945412635803,
+      "learning_rate": 1.869620053306448e-06,
+      "loss": 0.148,
+      "step": 2603
+    },
+    {
+      "epoch": 1.1878207321245295,
+      "grad_norm": 0.33435988426208496,
+      "learning_rate": 1.8678327840927627e-06,
+      "loss": 0.1651,
+      "step": 2604
+    },
+    {
+      "epoch": 1.188276884479416,
+      "grad_norm": 0.32039669156074524,
+      "learning_rate": 1.8660458599792546e-06,
+      "loss": 0.1593,
+      "step": 2605
+    },
+    {
+      "epoch": 1.1887330368343028,
+      "grad_norm": 0.30875256657600403,
+      "learning_rate": 1.864259281941408e-06,
+      "loss": 0.1443,
+      "step": 2606
+    },
+    {
+      "epoch": 1.1891891891891893,
+      "grad_norm": 0.2998284697532654,
+      "learning_rate": 1.8624730509545147e-06,
+      "loss": 0.1507,
+      "step": 2607
+    },
+    {
+      "epoch": 1.1896453415440758,
+      "grad_norm": 0.2920983135700226,
+      "learning_rate": 1.860687167993678e-06,
+      "loss": 0.1442,
+      "step": 2608
+    },
+    {
+      "epoch": 1.1901014938989622,
+      "grad_norm": 0.30350929498672485,
+      "learning_rate": 1.858901634033813e-06,
+      "loss": 0.1478,
+      "step": 2609
+    },
+    {
+      "epoch": 1.1905576462538487,
+      "grad_norm": 0.30006861686706543,
+      "learning_rate": 1.8571164500496428e-06,
+      "loss": 0.1482,
+      "step": 2610
+    },
+    {
+      "epoch": 1.1910137986087352,
+      "grad_norm": 0.31983694434165955,
+      "learning_rate": 1.8553316170156977e-06,
+      "loss": 0.1697,
+      "step": 2611
+    },
+    {
+      "epoch": 1.1914699509636217,
+      "grad_norm": 0.30616286396980286,
+      "learning_rate": 1.8535471359063211e-06,
+      "loss": 0.1609,
+      "step": 2612
+    },
+    {
+      "epoch": 1.1919261033185085,
+      "grad_norm": 0.32357361912727356,
+      "learning_rate": 1.8517630076956594e-06,
+      "loss": 0.158,
+      "step": 2613
+    },
+    {
+      "epoch": 1.192382255673395,
+      "grad_norm": 0.3030652403831482,
+      "learning_rate": 1.8499792333576698e-06,
+      "loss": 0.1505,
+      "step": 2614
+    },
+    {
+      "epoch": 1.1928384080282814,
+      "grad_norm": 0.30310359597206116,
+      "learning_rate": 1.848195813866115e-06,
+      "loss": 0.1614,
+      "step": 2615
+    },
+    {
+      "epoch": 1.193294560383168,
+      "grad_norm": 0.2982158362865448,
+      "learning_rate": 1.8464127501945625e-06,
+      "loss": 0.1514,
+      "step": 2616
+    },
+    {
+      "epoch": 1.1937507127380544,
+      "grad_norm": 0.30620935559272766,
+      "learning_rate": 1.8446300433163883e-06,
+      "loss": 0.1434,
+      "step": 2617
+    },
+    {
+      "epoch": 1.1942068650929412,
+      "grad_norm": 0.3338318169116974,
+      "learning_rate": 1.8428476942047723e-06,
+      "loss": 0.1574,
+      "step": 2618
+    },
+    {
+      "epoch": 1.1946630174478277,
+      "grad_norm": 0.30019357800483704,
+      "learning_rate": 1.8410657038326984e-06,
+      "loss": 0.1519,
+      "step": 2619
+    },
+    {
+      "epoch": 1.1951191698027142,
+      "grad_norm": 0.3047332465648651,
+      "learning_rate": 1.8392840731729561e-06,
+      "loss": 0.155,
+      "step": 2620
+    },
+    {
+      "epoch": 1.1955753221576007,
+      "grad_norm": 0.3109712302684784,
+      "learning_rate": 1.8375028031981368e-06,
+      "loss": 0.1634,
+      "step": 2621
+    },
+    {
+      "epoch": 1.1960314745124871,
+      "grad_norm": 0.321063756942749,
+      "learning_rate": 1.8357218948806378e-06,
+      "loss": 0.1478,
+      "step": 2622
+    },
+    {
+      "epoch": 1.1964876268673736,
+      "grad_norm": 0.3099934756755829,
+      "learning_rate": 1.833941349192656e-06,
+      "loss": 0.1631,
+      "step": 2623
+    },
+    {
+      "epoch": 1.1969437792222601,
+      "grad_norm": 0.3017142415046692,
+      "learning_rate": 1.8321611671061915e-06,
+      "loss": 0.1485,
+      "step": 2624
+    },
+    {
+      "epoch": 1.1973999315771469,
+      "grad_norm": 0.3092878758907318,
+      "learning_rate": 1.8303813495930466e-06,
+      "loss": 0.1558,
+      "step": 2625
+    },
+    {
+      "epoch": 1.1978560839320334,
+      "grad_norm": 0.2847161889076233,
+      "learning_rate": 1.8286018976248244e-06,
+      "loss": 0.1404,
+      "step": 2626
+    },
+    {
+      "epoch": 1.1983122362869199,
+      "grad_norm": 0.30538350343704224,
+      "learning_rate": 1.826822812172926e-06,
+      "loss": 0.1511,
+      "step": 2627
+    },
+    {
+      "epoch": 1.1987683886418063,
+      "grad_norm": 0.31767594814300537,
+      "learning_rate": 1.8250440942085562e-06,
+      "loss": 0.162,
+      "step": 2628
+    },
+    {
+      "epoch": 1.1992245409966928,
+      "grad_norm": 0.31899115443229675,
+      "learning_rate": 1.8232657447027172e-06,
+      "loss": 0.1511,
+      "step": 2629
+    },
+    {
+      "epoch": 1.1996806933515793,
+      "grad_norm": 0.30557581782341003,
+      "learning_rate": 1.821487764626209e-06,
+      "loss": 0.1607,
+      "step": 2630
+    },
+    {
+      "epoch": 1.2001368457064658,
+      "grad_norm": 0.3107191026210785,
+      "learning_rate": 1.8197101549496333e-06,
+      "loss": 0.1543,
+      "step": 2631
+    },
+    {
+      "epoch": 1.2005929980613526,
+      "grad_norm": 0.2884511649608612,
+      "learning_rate": 1.8179329166433846e-06,
+      "loss": 0.1535,
+      "step": 2632
+    },
+    {
+      "epoch": 1.201049150416239,
+      "grad_norm": 0.2948748767375946,
+      "learning_rate": 1.8161560506776603e-06,
+      "loss": 0.1452,
+      "step": 2633
+    },
+    {
+      "epoch": 1.2015053027711255,
+      "grad_norm": 0.29231536388397217,
+      "learning_rate": 1.8143795580224514e-06,
+      "loss": 0.1458,
+      "step": 2634
+    },
+    {
+      "epoch": 1.201961455126012,
+      "grad_norm": 0.3279147744178772,
+      "learning_rate": 1.8126034396475439e-06,
+      "loss": 0.1644,
+      "step": 2635
+    },
+    {
+      "epoch": 1.2024176074808985,
+      "grad_norm": 0.329790860414505,
+      "learning_rate": 1.8108276965225225e-06,
+      "loss": 0.1637,
+      "step": 2636
+    },
+    {
+      "epoch": 1.2028737598357853,
+      "grad_norm": 0.3062582314014435,
+      "learning_rate": 1.8090523296167655e-06,
+      "loss": 0.1503,
+      "step": 2637
+    },
+    {
+      "epoch": 1.2033299121906718,
+      "grad_norm": 0.28904667496681213,
+      "learning_rate": 1.8072773398994447e-06,
+      "loss": 0.1457,
+      "step": 2638
+    },
+    {
+      "epoch": 1.2037860645455583,
+      "grad_norm": 0.32898733019828796,
+      "learning_rate": 1.8055027283395294e-06,
+      "loss": 0.1561,
+      "step": 2639
+    },
+    {
+      "epoch": 1.2042422169004448,
+      "grad_norm": 0.3165503144264221,
+      "learning_rate": 1.8037284959057782e-06,
+      "loss": 0.1553,
+      "step": 2640
+    },
+    {
+      "epoch": 1.2046983692553312,
+      "grad_norm": 0.31806641817092896,
+      "learning_rate": 1.8019546435667461e-06,
+      "loss": 0.1565,
+      "step": 2641
+    },
+    {
+      "epoch": 1.2051545216102177,
+      "grad_norm": 0.30143144726753235,
+      "learning_rate": 1.8001811722907786e-06,
+      "loss": 0.1465,
+      "step": 2642
+    },
+    {
+      "epoch": 1.2056106739651042,
+      "grad_norm": 0.33881014585494995,
+      "learning_rate": 1.7984080830460144e-06,
+      "loss": 0.1652,
+      "step": 2643
+    },
+    {
+      "epoch": 1.206066826319991,
+      "grad_norm": 0.30798473954200745,
+      "learning_rate": 1.7966353768003838e-06,
+      "loss": 0.148,
+      "step": 2644
+    },
+    {
+      "epoch": 1.2065229786748775,
+      "grad_norm": 0.3176652193069458,
+      "learning_rate": 1.7948630545216067e-06,
+      "loss": 0.1581,
+      "step": 2645
+    },
+    {
+      "epoch": 1.206979131029764,
+      "grad_norm": 0.3237309455871582,
+      "learning_rate": 1.7930911171771937e-06,
+      "loss": 0.1611,
+      "step": 2646
+    },
+    {
+      "epoch": 1.2074352833846504,
+      "grad_norm": 0.32610419392585754,
+      "learning_rate": 1.7913195657344468e-06,
+      "loss": 0.1596,
+      "step": 2647
+    },
+    {
+      "epoch": 1.207891435739537,
+      "grad_norm": 0.30673009157180786,
+      "learning_rate": 1.7895484011604553e-06,
+      "loss": 0.1567,
+      "step": 2648
+    },
+    {
+      "epoch": 1.2083475880944234,
+      "grad_norm": 0.3054012954235077,
+      "learning_rate": 1.787777624422098e-06,
+      "loss": 0.1565,
+      "step": 2649
+    },
+    {
+      "epoch": 1.2088037404493102,
+      "grad_norm": 0.3162326216697693,
+      "learning_rate": 1.786007236486043e-06,
+      "loss": 0.1565,
+      "step": 2650
+    },
+    {
+      "epoch": 1.2092598928041967,
+      "grad_norm": 0.31434953212738037,
+      "learning_rate": 1.784237238318745e-06,
+      "loss": 0.158,
+      "step": 2651
+    },
+    {
+      "epoch": 1.2097160451590832,
+      "grad_norm": 0.3150542080402374,
+      "learning_rate": 1.7824676308864452e-06,
+      "loss": 0.1548,
+      "step": 2652
+    },
+    {
+      "epoch": 1.2101721975139696,
+      "grad_norm": 0.2970605790615082,
+      "learning_rate": 1.7806984151551748e-06,
+      "loss": 0.1506,
+      "step": 2653
+    },
+    {
+      "epoch": 1.2106283498688561,
+      "grad_norm": 0.32505783438682556,
+      "learning_rate": 1.778929592090747e-06,
+      "loss": 0.1617,
+      "step": 2654
+    },
+    {
+      "epoch": 1.2110845022237426,
+      "grad_norm": 0.3411644995212555,
+      "learning_rate": 1.777161162658764e-06,
+      "loss": 0.1651,
+      "step": 2655
+    },
+    {
+      "epoch": 1.2115406545786294,
+      "grad_norm": 0.3094402849674225,
+      "learning_rate": 1.7753931278246109e-06,
+      "loss": 0.1496,
+      "step": 2656
+    },
+    {
+      "epoch": 1.2119968069335159,
+      "grad_norm": 0.3243854343891144,
+      "learning_rate": 1.7736254885534577e-06,
+      "loss": 0.1618,
+      "step": 2657
+    },
+    {
+      "epoch": 1.2124529592884024,
+      "grad_norm": 0.30191361904144287,
+      "learning_rate": 1.77185824581026e-06,
+      "loss": 0.1542,
+      "step": 2658
+    },
+    {
+      "epoch": 1.2129091116432889,
+      "grad_norm": 0.3013656437397003,
+      "learning_rate": 1.7700914005597557e-06,
+      "loss": 0.1578,
+      "step": 2659
+    },
+    {
+      "epoch": 1.2133652639981753,
+      "grad_norm": 0.3291957676410675,
+      "learning_rate": 1.768324953766465e-06,
+      "loss": 0.1638,
+      "step": 2660
+    },
+    {
+      "epoch": 1.2138214163530618,
+      "grad_norm": 0.2955394983291626,
+      "learning_rate": 1.7665589063946923e-06,
+      "loss": 0.1479,
+      "step": 2661
+    },
+    {
+      "epoch": 1.2142775687079483,
+      "grad_norm": 0.3093152642250061,
+      "learning_rate": 1.764793259408522e-06,
+      "loss": 0.154,
+      "step": 2662
+    },
+    {
+      "epoch": 1.214733721062835,
+      "grad_norm": 0.29417502880096436,
+      "learning_rate": 1.7630280137718227e-06,
+      "loss": 0.146,
+      "step": 2663
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.2877040505409241,
+      "learning_rate": 1.761263170448242e-06,
+      "loss": 0.1459,
+      "step": 2664
+    },
+    {
+      "epoch": 1.215646025772608,
+      "grad_norm": 0.31246253848075867,
+      "learning_rate": 1.7594987304012068e-06,
+      "loss": 0.1589,
+      "step": 2665
+    },
+    {
+      "epoch": 1.2161021781274945,
+      "grad_norm": 0.3212037682533264,
+      "learning_rate": 1.7577346945939266e-06,
+      "loss": 0.1592,
+      "step": 2666
+    },
+    {
+      "epoch": 1.216558330482381,
+      "grad_norm": 0.3107142448425293,
+      "learning_rate": 1.7559710639893882e-06,
+      "loss": 0.1548,
+      "step": 2667
+    },
+    {
+      "epoch": 1.2170144828372678,
+      "grad_norm": 0.33747154474258423,
+      "learning_rate": 1.7542078395503574e-06,
+      "loss": 0.1593,
+      "step": 2668
+    },
+    {
+      "epoch": 1.2174706351921543,
+      "grad_norm": 0.3010542690753937,
+      "learning_rate": 1.7524450222393796e-06,
+      "loss": 0.1568,
+      "step": 2669
+    },
+    {
+      "epoch": 1.2179267875470408,
+      "grad_norm": 0.30928072333335876,
+      "learning_rate": 1.7506826130187766e-06,
+      "loss": 0.1469,
+      "step": 2670
+    },
+    {
+      "epoch": 1.2183829399019273,
+      "grad_norm": 0.309754878282547,
+      "learning_rate": 1.748920612850647e-06,
+      "loss": 0.1514,
+      "step": 2671
+    },
+    {
+      "epoch": 1.2188390922568137,
+      "grad_norm": 0.310050904750824,
+      "learning_rate": 1.7471590226968682e-06,
+      "loss": 0.1555,
+      "step": 2672
+    },
+    {
+      "epoch": 1.2192952446117002,
+      "grad_norm": 0.307734876871109,
+      "learning_rate": 1.7453978435190922e-06,
+      "loss": 0.152,
+      "step": 2673
+    },
+    {
+      "epoch": 1.2197513969665867,
+      "grad_norm": 0.30353859066963196,
+      "learning_rate": 1.7436370762787468e-06,
+      "loss": 0.1451,
+      "step": 2674
+    },
+    {
+      "epoch": 1.2202075493214735,
+      "grad_norm": 0.3155065178871155,
+      "learning_rate": 1.7418767219370359e-06,
+      "loss": 0.1499,
+      "step": 2675
+    },
+    {
+      "epoch": 1.22066370167636,
+      "grad_norm": 0.3251703083515167,
+      "learning_rate": 1.7401167814549353e-06,
+      "loss": 0.1522,
+      "step": 2676
+    },
+    {
+      "epoch": 1.2211198540312465,
+      "grad_norm": 0.30963969230651855,
+      "learning_rate": 1.7383572557931987e-06,
+      "loss": 0.1557,
+      "step": 2677
+    },
+    {
+      "epoch": 1.221576006386133,
+      "grad_norm": 0.31001678109169006,
+      "learning_rate": 1.736598145912351e-06,
+      "loss": 0.1528,
+      "step": 2678
+    },
+    {
+      "epoch": 1.2220321587410194,
+      "grad_norm": 0.3227824866771698,
+      "learning_rate": 1.7348394527726894e-06,
+      "loss": 0.1577,
+      "step": 2679
+    },
+    {
+      "epoch": 1.222488311095906,
+      "grad_norm": 0.32634344696998596,
+      "learning_rate": 1.7330811773342864e-06,
+      "loss": 0.1631,
+      "step": 2680
+    },
+    {
+      "epoch": 1.2229444634507924,
+      "grad_norm": 0.3060514032840729,
+      "learning_rate": 1.7313233205569838e-06,
+      "loss": 0.1545,
+      "step": 2681
+    },
+    {
+      "epoch": 1.2234006158056792,
+      "grad_norm": 0.3172318637371063,
+      "learning_rate": 1.7295658834003958e-06,
+      "loss": 0.1535,
+      "step": 2682
+    },
+    {
+      "epoch": 1.2238567681605657,
+      "grad_norm": 0.3039669096469879,
+      "learning_rate": 1.7278088668239074e-06,
+      "loss": 0.1501,
+      "step": 2683
+    },
+    {
+      "epoch": 1.2243129205154522,
+      "grad_norm": 0.31188148260116577,
+      "learning_rate": 1.7260522717866751e-06,
+      "loss": 0.1587,
+      "step": 2684
+    },
+    {
+      "epoch": 1.2247690728703386,
+      "grad_norm": 0.3121003806591034,
+      "learning_rate": 1.7242960992476242e-06,
+      "loss": 0.1526,
+      "step": 2685
+    },
+    {
+      "epoch": 1.2252252252252251,
+      "grad_norm": 0.31302228569984436,
+      "learning_rate": 1.7225403501654497e-06,
+      "loss": 0.1585,
+      "step": 2686
+    },
+    {
+      "epoch": 1.2256813775801119,
+      "grad_norm": 0.307232141494751,
+      "learning_rate": 1.720785025498614e-06,
+      "loss": 0.1496,
+      "step": 2687
+    },
+    {
+      "epoch": 1.2261375299349984,
+      "grad_norm": 0.31790757179260254,
+      "learning_rate": 1.719030126205351e-06,
+      "loss": 0.1559,
+      "step": 2688
+    },
+    {
+      "epoch": 1.2265936822898849,
+      "grad_norm": 0.32046571373939514,
+      "learning_rate": 1.7172756532436596e-06,
+      "loss": 0.159,
+      "step": 2689
+    },
+    {
+      "epoch": 1.2270498346447714,
+      "grad_norm": 0.31658434867858887,
+      "learning_rate": 1.715521607571306e-06,
+      "loss": 0.1534,
+      "step": 2690
+    },
+    {
+      "epoch": 1.2275059869996578,
+      "grad_norm": 0.2986975908279419,
+      "learning_rate": 1.7137679901458253e-06,
+      "loss": 0.1461,
+      "step": 2691
+    },
+    {
+      "epoch": 1.2279621393545443,
+      "grad_norm": 0.3127783238887787,
+      "learning_rate": 1.7120148019245173e-06,
+      "loss": 0.1549,
+      "step": 2692
+    },
+    {
+      "epoch": 1.2284182917094308,
+      "grad_norm": 0.3254636824131012,
+      "learning_rate": 1.7102620438644463e-06,
+      "loss": 0.1557,
+      "step": 2693
+    },
+    {
+      "epoch": 1.2288744440643176,
+      "grad_norm": 0.29877039790153503,
+      "learning_rate": 1.7085097169224453e-06,
+      "loss": 0.148,
+      "step": 2694
+    },
+    {
+      "epoch": 1.229330596419204,
+      "grad_norm": 0.33304280042648315,
+      "learning_rate": 1.706757822055108e-06,
+      "loss": 0.1524,
+      "step": 2695
+    },
+    {
+      "epoch": 1.2297867487740906,
+      "grad_norm": 0.3092648983001709,
+      "learning_rate": 1.7050063602187956e-06,
+      "loss": 0.1583,
+      "step": 2696
+    },
+    {
+      "epoch": 1.230242901128977,
+      "grad_norm": 0.32401609420776367,
+      "learning_rate": 1.7032553323696308e-06,
+      "loss": 0.157,
+      "step": 2697
+    },
+    {
+      "epoch": 1.2306990534838635,
+      "grad_norm": 0.3087688088417053,
+      "learning_rate": 1.7015047394634993e-06,
+      "loss": 0.1462,
+      "step": 2698
+    },
+    {
+      "epoch": 1.2311552058387503,
+      "grad_norm": 0.3159826099872589,
+      "learning_rate": 1.699754582456051e-06,
+      "loss": 0.1542,
+      "step": 2699
+    },
+    {
+      "epoch": 1.2316113581936368,
+      "grad_norm": 0.318142294883728,
+      "learning_rate": 1.6980048623026967e-06,
+      "loss": 0.1482,
+      "step": 2700
+    },
+    {
+      "epoch": 1.2320675105485233,
+      "grad_norm": 0.30811622738838196,
+      "learning_rate": 1.6962555799586078e-06,
+      "loss": 0.1558,
+      "step": 2701
+    },
+    {
+      "epoch": 1.2325236629034098,
+      "grad_norm": 0.3205932080745697,
+      "learning_rate": 1.6945067363787194e-06,
+      "loss": 0.1541,
+      "step": 2702
+    },
+    {
+      "epoch": 1.2329798152582963,
+      "grad_norm": 0.31248947978019714,
+      "learning_rate": 1.6927583325177239e-06,
+      "loss": 0.1548,
+      "step": 2703
+    },
+    {
+      "epoch": 1.2334359676131827,
+      "grad_norm": 0.32467710971832275,
+      "learning_rate": 1.6910103693300767e-06,
+      "loss": 0.159,
+      "step": 2704
+    },
+    {
+      "epoch": 1.2338921199680692,
+      "grad_norm": 0.3079068064689636,
+      "learning_rate": 1.6892628477699907e-06,
+      "loss": 0.1451,
+      "step": 2705
+    },
+    {
+      "epoch": 1.234348272322956,
+      "grad_norm": 0.3093346655368805,
+      "learning_rate": 1.6875157687914375e-06,
+      "loss": 0.1437,
+      "step": 2706
+    },
+    {
+      "epoch": 1.2348044246778425,
+      "grad_norm": 0.29650968313217163,
+      "learning_rate": 1.6857691333481487e-06,
+      "loss": 0.1504,
+      "step": 2707
+    },
+    {
+      "epoch": 1.235260577032729,
+      "grad_norm": 0.31869250535964966,
+      "learning_rate": 1.6840229423936127e-06,
+      "loss": 0.1502,
+      "step": 2708
+    },
+    {
+      "epoch": 1.2357167293876155,
+      "grad_norm": 0.3020484149456024,
+      "learning_rate": 1.6822771968810744e-06,
+      "loss": 0.1486,
+      "step": 2709
+    },
+    {
+      "epoch": 1.236172881742502,
+      "grad_norm": 0.31009912490844727,
+      "learning_rate": 1.6805318977635375e-06,
+      "loss": 0.1589,
+      "step": 2710
+    },
+    {
+      "epoch": 1.2366290340973884,
+      "grad_norm": 0.32181861996650696,
+      "learning_rate": 1.6787870459937615e-06,
+      "loss": 0.1566,
+      "step": 2711
+    },
+    {
+      "epoch": 1.237085186452275,
+      "grad_norm": 0.31387802958488464,
+      "learning_rate": 1.6770426425242603e-06,
+      "loss": 0.1574,
+      "step": 2712
+    },
+    {
+      "epoch": 1.2375413388071617,
+      "grad_norm": 0.287138432264328,
+      "learning_rate": 1.6752986883073031e-06,
+      "loss": 0.1431,
+      "step": 2713
+    },
+    {
+      "epoch": 1.2379974911620482,
+      "grad_norm": 0.30575278401374817,
+      "learning_rate": 1.6735551842949166e-06,
+      "loss": 0.1544,
+      "step": 2714
+    },
+    {
+      "epoch": 1.2384536435169347,
+      "grad_norm": 0.317719429731369,
+      "learning_rate": 1.6718121314388796e-06,
+      "loss": 0.1586,
+      "step": 2715
+    },
+    {
+      "epoch": 1.2389097958718212,
+      "grad_norm": 0.3255164623260498,
+      "learning_rate": 1.670069530690725e-06,
+      "loss": 0.1553,
+      "step": 2716
+    },
+    {
+      "epoch": 1.2393659482267076,
+      "grad_norm": 0.3311634361743927,
+      "learning_rate": 1.6683273830017378e-06,
+      "loss": 0.1581,
+      "step": 2717
+    },
+    {
+      "epoch": 1.2398221005815944,
+      "grad_norm": 0.3169220983982086,
+      "learning_rate": 1.6665856893229573e-06,
+      "loss": 0.1535,
+      "step": 2718
+    },
+    {
+      "epoch": 1.2402782529364809,
+      "grad_norm": 0.3144884407520294,
+      "learning_rate": 1.6648444506051748e-06,
+      "loss": 0.1513,
+      "step": 2719
+    },
+    {
+      "epoch": 1.2407344052913674,
+      "grad_norm": 0.30270740389823914,
+      "learning_rate": 1.6631036677989315e-06,
+      "loss": 0.1533,
+      "step": 2720
+    },
+    {
+      "epoch": 1.2411905576462539,
+      "grad_norm": 0.29269200563430786,
+      "learning_rate": 1.6613633418545222e-06,
+      "loss": 0.1512,
+      "step": 2721
+    },
+    {
+      "epoch": 1.2416467100011404,
+      "grad_norm": 0.3148145377635956,
+      "learning_rate": 1.659623473721991e-06,
+      "loss": 0.1529,
+      "step": 2722
+    },
+    {
+      "epoch": 1.2421028623560268,
+      "grad_norm": 0.3186376094818115,
+      "learning_rate": 1.6578840643511318e-06,
+      "loss": 0.1626,
+      "step": 2723
+    },
+    {
+      "epoch": 1.2425590147109133,
+      "grad_norm": 0.3129862844944,
+      "learning_rate": 1.6561451146914873e-06,
+      "loss": 0.154,
+      "step": 2724
+    },
+    {
+      "epoch": 1.2430151670658,
+      "grad_norm": 0.3209805488586426,
+      "learning_rate": 1.6544066256923522e-06,
+      "loss": 0.1565,
+      "step": 2725
+    },
+    {
+      "epoch": 1.2434713194206866,
+      "grad_norm": 0.299464613199234,
+      "learning_rate": 1.6526685983027676e-06,
+      "loss": 0.1485,
+      "step": 2726
+    },
+    {
+      "epoch": 1.243927471775573,
+      "grad_norm": 0.3183022141456604,
+      "learning_rate": 1.6509310334715229e-06,
+      "loss": 0.1603,
+      "step": 2727
+    },
+    {
+      "epoch": 1.2443836241304596,
+      "grad_norm": 0.3087853789329529,
+      "learning_rate": 1.649193932147154e-06,
+      "loss": 0.1557,
+      "step": 2728
+    },
+    {
+      "epoch": 1.244839776485346,
+      "grad_norm": 0.31835296750068665,
+      "learning_rate": 1.6474572952779461e-06,
+      "loss": 0.1621,
+      "step": 2729
+    },
+    {
+      "epoch": 1.2452959288402325,
+      "grad_norm": 0.33173424005508423,
+      "learning_rate": 1.6457211238119286e-06,
+      "loss": 0.1541,
+      "step": 2730
+    },
+    {
+      "epoch": 1.2457520811951193,
+      "grad_norm": 0.30072087049484253,
+      "learning_rate": 1.6439854186968785e-06,
+      "loss": 0.1583,
+      "step": 2731
+    },
+    {
+      "epoch": 1.2462082335500058,
+      "grad_norm": 0.3210989236831665,
+      "learning_rate": 1.6422501808803165e-06,
+      "loss": 0.1553,
+      "step": 2732
+    },
+    {
+      "epoch": 1.2466643859048923,
+      "grad_norm": 0.30897119641304016,
+      "learning_rate": 1.6405154113095102e-06,
+      "loss": 0.1512,
+      "step": 2733
+    },
+    {
+      "epoch": 1.2471205382597788,
+      "grad_norm": 0.30960527062416077,
+      "learning_rate": 1.6387811109314688e-06,
+      "loss": 0.1574,
+      "step": 2734
+    },
+    {
+      "epoch": 1.2475766906146653,
+      "grad_norm": 0.30513590574264526,
+      "learning_rate": 1.6370472806929493e-06,
+      "loss": 0.1493,
+      "step": 2735
+    },
+    {
+      "epoch": 1.2480328429695517,
+      "grad_norm": 0.30977416038513184,
+      "learning_rate": 1.635313921540449e-06,
+      "loss": 0.1556,
+      "step": 2736
+    },
+    {
+      "epoch": 1.2484889953244385,
+      "grad_norm": 0.30246755480766296,
+      "learning_rate": 1.6335810344202094e-06,
+      "loss": 0.1524,
+      "step": 2737
+    },
+    {
+      "epoch": 1.248945147679325,
+      "grad_norm": 0.30922985076904297,
+      "learning_rate": 1.631848620278213e-06,
+      "loss": 0.1517,
+      "step": 2738
+    },
+    {
+      "epoch": 1.2494013000342115,
+      "grad_norm": 0.3096439242362976,
+      "learning_rate": 1.6301166800601853e-06,
+      "loss": 0.1534,
+      "step": 2739
+    },
+    {
+      "epoch": 1.249857452389098,
+      "grad_norm": 0.30746814608573914,
+      "learning_rate": 1.6283852147115931e-06,
+      "loss": 0.1578,
+      "step": 2740
+    },
+    {
+      "epoch": 1.2503136047439845,
+      "grad_norm": 0.3158532977104187,
+      "learning_rate": 1.6266542251776432e-06,
+      "loss": 0.1561,
+      "step": 2741
+    },
+    {
+      "epoch": 1.250769757098871,
+      "grad_norm": 0.3057871460914612,
+      "learning_rate": 1.624923712403283e-06,
+      "loss": 0.1523,
+      "step": 2742
+    },
+    {
+      "epoch": 1.2512259094537574,
+      "grad_norm": 0.30569472908973694,
+      "learning_rate": 1.6231936773331997e-06,
+      "loss": 0.1567,
+      "step": 2743
+    },
+    {
+      "epoch": 1.2516820618086442,
+      "grad_norm": 0.3193475604057312,
+      "learning_rate": 1.6214641209118186e-06,
+      "loss": 0.1589,
+      "step": 2744
+    },
+    {
+      "epoch": 1.2521382141635307,
+      "grad_norm": 0.30626213550567627,
+      "learning_rate": 1.6197350440833076e-06,
+      "loss": 0.1523,
+      "step": 2745
+    },
+    {
+      "epoch": 1.2525943665184172,
+      "grad_norm": 0.3074076473712921,
+      "learning_rate": 1.618006447791568e-06,
+      "loss": 0.1477,
+      "step": 2746
+    },
+    {
+      "epoch": 1.2530505188733037,
+      "grad_norm": 0.3103839159011841,
+      "learning_rate": 1.6162783329802408e-06,
+      "loss": 0.1548,
+      "step": 2747
+    },
+    {
+      "epoch": 1.2535066712281902,
+      "grad_norm": 0.31095272302627563,
+      "learning_rate": 1.6145507005927052e-06,
+      "loss": 0.1567,
+      "step": 2748
+    },
+    {
+      "epoch": 1.2539628235830769,
+      "grad_norm": 0.3184444010257721,
+      "learning_rate": 1.612823551572075e-06,
+      "loss": 0.1597,
+      "step": 2749
+    },
+    {
+      "epoch": 1.2544189759379631,
+      "grad_norm": 0.30295413732528687,
+      "learning_rate": 1.611096886861201e-06,
+      "loss": 0.1532,
+      "step": 2750
+    },
+    {
+      "epoch": 1.2548751282928499,
+      "grad_norm": 0.2971593737602234,
+      "learning_rate": 1.6093707074026707e-06,
+      "loss": 0.1442,
+      "step": 2751
+    },
+    {
+      "epoch": 1.2553312806477364,
+      "grad_norm": 0.30472785234451294,
+      "learning_rate": 1.607645014138805e-06,
+      "loss": 0.1569,
+      "step": 2752
+    },
+    {
+      "epoch": 1.2557874330026229,
+      "grad_norm": 0.3229816257953644,
+      "learning_rate": 1.6059198080116605e-06,
+      "loss": 0.1481,
+      "step": 2753
+    },
+    {
+      "epoch": 1.2562435853575094,
+      "grad_norm": 0.3252611756324768,
+      "learning_rate": 1.604195089963026e-06,
+      "loss": 0.1594,
+      "step": 2754
+    },
+    {
+      "epoch": 1.2566997377123958,
+      "grad_norm": 0.30860254168510437,
+      "learning_rate": 1.602470860934427e-06,
+      "loss": 0.1547,
+      "step": 2755
+    },
+    {
+      "epoch": 1.2571558900672826,
+      "grad_norm": 0.322961688041687,
+      "learning_rate": 1.6007471218671209e-06,
+      "loss": 0.1513,
+      "step": 2756
+    },
+    {
+      "epoch": 1.257612042422169,
+      "grad_norm": 0.3013272285461426,
+      "learning_rate": 1.5990238737020958e-06,
+      "loss": 0.1497,
+      "step": 2757
+    },
+    {
+      "epoch": 1.2580681947770556,
+      "grad_norm": 0.3016090989112854,
+      "learning_rate": 1.597301117380073e-06,
+      "loss": 0.1542,
+      "step": 2758
+    },
+    {
+      "epoch": 1.258524347131942,
+      "grad_norm": 0.3073787987232208,
+      "learning_rate": 1.5955788538415068e-06,
+      "loss": 0.1499,
+      "step": 2759
+    },
+    {
+      "epoch": 1.2589804994868286,
+      "grad_norm": 0.30865272879600525,
+      "learning_rate": 1.59385708402658e-06,
+      "loss": 0.1547,
+      "step": 2760
+    },
+    {
+      "epoch": 1.2594366518417153,
+      "grad_norm": 0.3037521243095398,
+      "learning_rate": 1.5921358088752076e-06,
+      "loss": 0.1471,
+      "step": 2761
+    },
+    {
+      "epoch": 1.2598928041966015,
+      "grad_norm": 0.32083818316459656,
+      "learning_rate": 1.5904150293270327e-06,
+      "loss": 0.1528,
+      "step": 2762
+    },
+    {
+      "epoch": 1.2603489565514883,
+      "grad_norm": 0.3303329646587372,
+      "learning_rate": 1.588694746321431e-06,
+      "loss": 0.1595,
+      "step": 2763
+    },
+    {
+      "epoch": 1.2608051089063748,
+      "grad_norm": 0.31374749541282654,
+      "learning_rate": 1.586974960797503e-06,
+      "loss": 0.1454,
+      "step": 2764
+    },
+    {
+      "epoch": 1.2612612612612613,
+      "grad_norm": 0.2799595594406128,
+      "learning_rate": 1.5852556736940822e-06,
+      "loss": 0.1432,
+      "step": 2765
+    },
+    {
+      "epoch": 1.2617174136161478,
+      "grad_norm": 0.311143159866333,
+      "learning_rate": 1.5835368859497264e-06,
+      "loss": 0.1554,
+      "step": 2766
+    },
+    {
+      "epoch": 1.2621735659710343,
+      "grad_norm": 0.31572243571281433,
+      "learning_rate": 1.5818185985027223e-06,
+      "loss": 0.1537,
+      "step": 2767
+    },
+    {
+      "epoch": 1.262629718325921,
+      "grad_norm": 0.31115269660949707,
+      "learning_rate": 1.5801008122910838e-06,
+      "loss": 0.1454,
+      "step": 2768
+    },
+    {
+      "epoch": 1.2630858706808075,
+      "grad_norm": 0.3117983937263489,
+      "learning_rate": 1.5783835282525498e-06,
+      "loss": 0.1534,
+      "step": 2769
+    },
+    {
+      "epoch": 1.263542023035694,
+      "grad_norm": 0.3179197311401367,
+      "learning_rate": 1.5766667473245867e-06,
+      "loss": 0.1592,
+      "step": 2770
+    },
+    {
+      "epoch": 1.2639981753905805,
+      "grad_norm": 0.3219977617263794,
+      "learning_rate": 1.5749504704443849e-06,
+      "loss": 0.1569,
+      "step": 2771
+    },
+    {
+      "epoch": 1.264454327745467,
+      "grad_norm": 0.30277568101882935,
+      "learning_rate": 1.5732346985488605e-06,
+      "loss": 0.1472,
+      "step": 2772
+    },
+    {
+      "epoch": 1.2649104801003535,
+      "grad_norm": 0.310068815946579,
+      "learning_rate": 1.571519432574653e-06,
+      "loss": 0.1543,
+      "step": 2773
+    },
+    {
+      "epoch": 1.26536663245524,
+      "grad_norm": 0.3069816827774048,
+      "learning_rate": 1.5698046734581274e-06,
+      "loss": 0.1585,
+      "step": 2774
+    },
+    {
+      "epoch": 1.2658227848101267,
+      "grad_norm": 0.32326608896255493,
+      "learning_rate": 1.5680904221353693e-06,
+      "loss": 0.1535,
+      "step": 2775
+    },
+    {
+      "epoch": 1.2662789371650132,
+      "grad_norm": 0.31554704904556274,
+      "learning_rate": 1.5663766795421912e-06,
+      "loss": 0.1545,
+      "step": 2776
+    },
+    {
+      "epoch": 1.2667350895198997,
+      "grad_norm": 0.2983073592185974,
+      "learning_rate": 1.5646634466141234e-06,
+      "loss": 0.1518,
+      "step": 2777
+    },
+    {
+      "epoch": 1.2671912418747862,
+      "grad_norm": 0.3003422021865845,
+      "learning_rate": 1.562950724286421e-06,
+      "loss": 0.15,
+      "step": 2778
+    },
+    {
+      "epoch": 1.2676473942296727,
+      "grad_norm": 0.3148253262042999,
+      "learning_rate": 1.5612385134940593e-06,
+      "loss": 0.1544,
+      "step": 2779
+    },
+    {
+      "epoch": 1.2681035465845594,
+      "grad_norm": 0.304943323135376,
+      "learning_rate": 1.5595268151717347e-06,
+      "loss": 0.1512,
+      "step": 2780
+    },
+    {
+      "epoch": 1.2685596989394456,
+      "grad_norm": 0.30289942026138306,
+      "learning_rate": 1.5578156302538622e-06,
+      "loss": 0.1475,
+      "step": 2781
+    },
+    {
+      "epoch": 1.2690158512943324,
+      "grad_norm": 0.29996365308761597,
+      "learning_rate": 1.5561049596745793e-06,
+      "loss": 0.149,
+      "step": 2782
+    },
+    {
+      "epoch": 1.2694720036492189,
+      "grad_norm": 0.31475630402565,
+      "learning_rate": 1.5543948043677407e-06,
+      "loss": 0.1471,
+      "step": 2783
+    },
+    {
+      "epoch": 1.2699281560041054,
+      "grad_norm": 0.3142929971218109,
+      "learning_rate": 1.55268516526692e-06,
+      "loss": 0.1531,
+      "step": 2784
+    },
+    {
+      "epoch": 1.2703843083589919,
+      "grad_norm": 0.3168845474720001,
+      "learning_rate": 1.5509760433054099e-06,
+      "loss": 0.1585,
+      "step": 2785
+    },
+    {
+      "epoch": 1.2708404607138784,
+      "grad_norm": 0.3155894875526428,
+      "learning_rate": 1.5492674394162213e-06,
+      "loss": 0.1571,
+      "step": 2786
+    },
+    {
+      "epoch": 1.271296613068765,
+      "grad_norm": 0.3478847146034241,
+      "learning_rate": 1.5475593545320804e-06,
+      "loss": 0.1555,
+      "step": 2787
+    },
+    {
+      "epoch": 1.2717527654236516,
+      "grad_norm": 0.3011278212070465,
+      "learning_rate": 1.5458517895854309e-06,
+      "loss": 0.1468,
+      "step": 2788
+    },
+    {
+      "epoch": 1.272208917778538,
+      "grad_norm": 0.3064025342464447,
+      "learning_rate": 1.544144745508433e-06,
+      "loss": 0.1558,
+      "step": 2789
+    },
+    {
+      "epoch": 1.2726650701334246,
+      "grad_norm": 0.30734509229660034,
+      "learning_rate": 1.5424382232329628e-06,
+      "loss": 0.1585,
+      "step": 2790
+    },
+    {
+      "epoch": 1.273121222488311,
+      "grad_norm": 0.3009508550167084,
+      "learning_rate": 1.54073222369061e-06,
+      "loss": 0.1501,
+      "step": 2791
+    },
+    {
+      "epoch": 1.2735773748431976,
+      "grad_norm": 0.3109910190105438,
+      "learning_rate": 1.5390267478126802e-06,
+      "loss": 0.1587,
+      "step": 2792
+    },
+    {
+      "epoch": 1.274033527198084,
+      "grad_norm": 0.30920159816741943,
+      "learning_rate": 1.5373217965301937e-06,
+      "loss": 0.1475,
+      "step": 2793
+    },
+    {
+      "epoch": 1.2744896795529708,
+      "grad_norm": 0.3178183436393738,
+      "learning_rate": 1.5356173707738833e-06,
+      "loss": 0.1548,
+      "step": 2794
+    },
+    {
+      "epoch": 1.2749458319078573,
+      "grad_norm": 0.3134722411632538,
+      "learning_rate": 1.5339134714741943e-06,
+      "loss": 0.1579,
+      "step": 2795
+    },
+    {
+      "epoch": 1.2754019842627438,
+      "grad_norm": 0.3062033951282501,
+      "learning_rate": 1.5322100995612868e-06,
+      "loss": 0.1595,
+      "step": 2796
+    },
+    {
+      "epoch": 1.2758581366176303,
+      "grad_norm": 0.3179316520690918,
+      "learning_rate": 1.530507255965032e-06,
+      "loss": 0.1618,
+      "step": 2797
+    },
+    {
+      "epoch": 1.2763142889725168,
+      "grad_norm": 0.30741018056869507,
+      "learning_rate": 1.5288049416150119e-06,
+      "loss": 0.151,
+      "step": 2798
+    },
+    {
+      "epoch": 1.2767704413274035,
+      "grad_norm": 0.35391443967819214,
+      "learning_rate": 1.5271031574405199e-06,
+      "loss": 0.1525,
+      "step": 2799
+    },
+    {
+      "epoch": 1.27722659368229,
+      "grad_norm": 0.31200337409973145,
+      "learning_rate": 1.525401904370561e-06,
+      "loss": 0.1503,
+      "step": 2800
+    },
+    {
+      "epoch": 1.2776827460371765,
+      "grad_norm": 0.29565033316612244,
+      "learning_rate": 1.5237011833338488e-06,
+      "loss": 0.1466,
+      "step": 2801
+    },
+    {
+      "epoch": 1.278138898392063,
+      "grad_norm": 0.32427778840065,
+      "learning_rate": 1.5220009952588076e-06,
+      "loss": 0.159,
+      "step": 2802
+    },
+    {
+      "epoch": 1.2785950507469495,
+      "grad_norm": 0.29709169268608093,
+      "learning_rate": 1.52030134107357e-06,
+      "loss": 0.1473,
+      "step": 2803
+    },
+    {
+      "epoch": 1.279051203101836,
+      "grad_norm": 0.2913004159927368,
+      "learning_rate": 1.5186022217059777e-06,
+      "loss": 0.1469,
+      "step": 2804
+    },
+    {
+      "epoch": 1.2795073554567225,
+      "grad_norm": 0.3017559349536896,
+      "learning_rate": 1.5169036380835792e-06,
+      "loss": 0.1548,
+      "step": 2805
+    },
+    {
+      "epoch": 1.2799635078116092,
+      "grad_norm": 0.3405783474445343,
+      "learning_rate": 1.5152055911336334e-06,
+      "loss": 0.1567,
+      "step": 2806
+    },
+    {
+      "epoch": 1.2804196601664957,
+      "grad_norm": 0.31293705105781555,
+      "learning_rate": 1.5135080817831033e-06,
+      "loss": 0.1536,
+      "step": 2807
+    },
+    {
+      "epoch": 1.2808758125213822,
+      "grad_norm": 0.3153549134731293,
+      "learning_rate": 1.5118111109586598e-06,
+      "loss": 0.1569,
+      "step": 2808
+    },
+    {
+      "epoch": 1.2813319648762687,
+      "grad_norm": 0.297101229429245,
+      "learning_rate": 1.5101146795866795e-06,
+      "loss": 0.1455,
+      "step": 2809
+    },
+    {
+      "epoch": 1.2817881172311552,
+      "grad_norm": 0.31153643131256104,
+      "learning_rate": 1.5084187885932447e-06,
+      "loss": 0.1612,
+      "step": 2810
+    },
+    {
+      "epoch": 1.2822442695860419,
+      "grad_norm": 0.32876113057136536,
+      "learning_rate": 1.5067234389041416e-06,
+      "loss": 0.1643,
+      "step": 2811
+    },
+    {
+      "epoch": 1.2827004219409281,
+      "grad_norm": 0.3016121983528137,
+      "learning_rate": 1.5050286314448632e-06,
+      "loss": 0.1516,
+      "step": 2812
+    },
+    {
+      "epoch": 1.2831565742958149,
+      "grad_norm": 0.29664552211761475,
+      "learning_rate": 1.5033343671406042e-06,
+      "loss": 0.1445,
+      "step": 2813
+    },
+    {
+      "epoch": 1.2836127266507014,
+      "grad_norm": 0.3113342821598053,
+      "learning_rate": 1.5016406469162637e-06,
+      "loss": 0.1479,
+      "step": 2814
+    },
+    {
+      "epoch": 1.2840688790055879,
+      "grad_norm": 0.3214724361896515,
+      "learning_rate": 1.4999474716964449e-06,
+      "loss": 0.1597,
+      "step": 2815
+    },
+    {
+      "epoch": 1.2845250313604744,
+      "grad_norm": 0.31613484025001526,
+      "learning_rate": 1.4982548424054505e-06,
+      "loss": 0.1599,
+      "step": 2816
+    },
+    {
+      "epoch": 1.2849811837153609,
+      "grad_norm": 0.3102830946445465,
+      "learning_rate": 1.496562759967289e-06,
+      "loss": 0.1515,
+      "step": 2817
+    },
+    {
+      "epoch": 1.2854373360702476,
+      "grad_norm": 0.3270382285118103,
+      "learning_rate": 1.4948712253056677e-06,
+      "loss": 0.1468,
+      "step": 2818
+    },
+    {
+      "epoch": 1.285893488425134,
+      "grad_norm": 0.31916970014572144,
+      "learning_rate": 1.493180239343996e-06,
+      "loss": 0.1579,
+      "step": 2819
+    },
+    {
+      "epoch": 1.2863496407800206,
+      "grad_norm": 0.30484479665756226,
+      "learning_rate": 1.4914898030053827e-06,
+      "loss": 0.1509,
+      "step": 2820
+    },
+    {
+      "epoch": 1.286805793134907,
+      "grad_norm": 0.30585789680480957,
+      "learning_rate": 1.4897999172126378e-06,
+      "loss": 0.1443,
+      "step": 2821
+    },
+    {
+      "epoch": 1.2872619454897936,
+      "grad_norm": 0.3167206346988678,
+      "learning_rate": 1.4881105828882697e-06,
+      "loss": 0.1572,
+      "step": 2822
+    },
+    {
+      "epoch": 1.28771809784468,
+      "grad_norm": 0.31721216440200806,
+      "learning_rate": 1.4864218009544867e-06,
+      "loss": 0.1566,
+      "step": 2823
+    },
+    {
+      "epoch": 1.2881742501995666,
+      "grad_norm": 0.323087602853775,
+      "learning_rate": 1.4847335723331952e-06,
+      "loss": 0.1611,
+      "step": 2824
+    },
+    {
+      "epoch": 1.2886304025544533,
+      "grad_norm": 0.3117333948612213,
+      "learning_rate": 1.4830458979459986e-06,
+      "loss": 0.1557,
+      "step": 2825
+    },
+    {
+      "epoch": 1.2890865549093398,
+      "grad_norm": 0.31498292088508606,
+      "learning_rate": 1.4813587787141986e-06,
+      "loss": 0.1518,
+      "step": 2826
+    },
+    {
+      "epoch": 1.2895427072642263,
+      "grad_norm": 0.3316204249858856,
+      "learning_rate": 1.4796722155587956e-06,
+      "loss": 0.1657,
+      "step": 2827
+    },
+    {
+      "epoch": 1.2899988596191128,
+      "grad_norm": 0.31495368480682373,
+      "learning_rate": 1.4779862094004837e-06,
+      "loss": 0.1589,
+      "step": 2828
+    },
+    {
+      "epoch": 1.2904550119739993,
+      "grad_norm": 0.3010186553001404,
+      "learning_rate": 1.4763007611596535e-06,
+      "loss": 0.1491,
+      "step": 2829
+    },
+    {
+      "epoch": 1.290911164328886,
+      "grad_norm": 0.32025548815727234,
+      "learning_rate": 1.4746158717563916e-06,
+      "loss": 0.1544,
+      "step": 2830
+    },
+    {
+      "epoch": 1.2913673166837722,
+      "grad_norm": 0.3204079270362854,
+      "learning_rate": 1.4729315421104795e-06,
+      "loss": 0.1525,
+      "step": 2831
+    },
+    {
+      "epoch": 1.291823469038659,
+      "grad_norm": 0.3094485104084015,
+      "learning_rate": 1.4712477731413935e-06,
+      "loss": 0.1592,
+      "step": 2832
+    },
+    {
+      "epoch": 1.2922796213935455,
+      "grad_norm": 0.3160625398159027,
+      "learning_rate": 1.4695645657683023e-06,
+      "loss": 0.1553,
+      "step": 2833
+    },
+    {
+      "epoch": 1.292735773748432,
+      "grad_norm": 0.31368139386177063,
+      "learning_rate": 1.4678819209100698e-06,
+      "loss": 0.1564,
+      "step": 2834
+    },
+    {
+      "epoch": 1.2931919261033185,
+      "grad_norm": 0.32229942083358765,
+      "learning_rate": 1.4661998394852516e-06,
+      "loss": 0.1477,
+      "step": 2835
+    },
+    {
+      "epoch": 1.293648078458205,
+      "grad_norm": 0.3030005097389221,
+      "learning_rate": 1.4645183224120956e-06,
+      "loss": 0.1507,
+      "step": 2836
+    },
+    {
+      "epoch": 1.2941042308130917,
+      "grad_norm": 0.3204514682292938,
+      "learning_rate": 1.462837370608544e-06,
+      "loss": 0.1558,
+      "step": 2837
+    },
+    {
+      "epoch": 1.2945603831679782,
+      "grad_norm": 0.2939911186695099,
+      "learning_rate": 1.461156984992228e-06,
+      "loss": 0.1492,
+      "step": 2838
+    },
+    {
+      "epoch": 1.2950165355228647,
+      "grad_norm": 0.3315194845199585,
+      "learning_rate": 1.4594771664804696e-06,
+      "loss": 0.1661,
+      "step": 2839
+    },
+    {
+      "epoch": 1.2954726878777512,
+      "grad_norm": 0.33271101117134094,
+      "learning_rate": 1.457797915990283e-06,
+      "loss": 0.1627,
+      "step": 2840
+    },
+    {
+      "epoch": 1.2959288402326377,
+      "grad_norm": 0.34080472588539124,
+      "learning_rate": 1.4561192344383707e-06,
+      "loss": 0.1612,
+      "step": 2841
+    },
+    {
+      "epoch": 1.2963849925875242,
+      "grad_norm": 0.30395233631134033,
+      "learning_rate": 1.4544411227411253e-06,
+      "loss": 0.152,
+      "step": 2842
+    },
+    {
+      "epoch": 1.2968411449424107,
+      "grad_norm": 0.3104346692562103,
+      "learning_rate": 1.4527635818146284e-06,
+      "loss": 0.1591,
+      "step": 2843
+    },
+    {
+      "epoch": 1.2972972972972974,
+      "grad_norm": 0.3082316517829895,
+      "learning_rate": 1.4510866125746497e-06,
+      "loss": 0.1523,
+      "step": 2844
+    },
+    {
+      "epoch": 1.2977534496521839,
+      "grad_norm": 0.2906036972999573,
+      "learning_rate": 1.4494102159366458e-06,
+      "loss": 0.1531,
+      "step": 2845
+    },
+    {
+      "epoch": 1.2982096020070704,
+      "grad_norm": 0.31794828176498413,
+      "learning_rate": 1.4477343928157638e-06,
+      "loss": 0.1531,
+      "step": 2846
+    },
+    {
+      "epoch": 1.2986657543619569,
+      "grad_norm": 0.31362074613571167,
+      "learning_rate": 1.4460591441268341e-06,
+      "loss": 0.1489,
+      "step": 2847
+    },
+    {
+      "epoch": 1.2991219067168434,
+      "grad_norm": 0.313071608543396,
+      "learning_rate": 1.4443844707843767e-06,
+      "loss": 0.1511,
+      "step": 2848
+    },
+    {
+      "epoch": 1.29957805907173,
+      "grad_norm": 0.3071008026599884,
+      "learning_rate": 1.442710373702596e-06,
+      "loss": 0.149,
+      "step": 2849
+    },
+    {
+      "epoch": 1.3000342114266166,
+      "grad_norm": 0.3161664307117462,
+      "learning_rate": 1.441036853795381e-06,
+      "loss": 0.1586,
+      "step": 2850
+    },
+    {
+      "epoch": 1.300490363781503,
+      "grad_norm": 0.31835851073265076,
+      "learning_rate": 1.439363911976307e-06,
+      "loss": 0.1538,
+      "step": 2851
+    },
+    {
+      "epoch": 1.3009465161363896,
+      "grad_norm": 0.27921515703201294,
+      "learning_rate": 1.437691549158633e-06,
+      "loss": 0.1436,
+      "step": 2852
+    },
+    {
+      "epoch": 1.301402668491276,
+      "grad_norm": 0.3164553940296173,
+      "learning_rate": 1.4360197662553004e-06,
+      "loss": 0.1557,
+      "step": 2853
+    },
+    {
+      "epoch": 1.3018588208461626,
+      "grad_norm": 0.315032958984375,
+      "learning_rate": 1.434348564178939e-06,
+      "loss": 0.1512,
+      "step": 2854
+    },
+    {
+      "epoch": 1.302314973201049,
+      "grad_norm": 0.31959787011146545,
+      "learning_rate": 1.4326779438418565e-06,
+      "loss": 0.1582,
+      "step": 2855
+    },
+    {
+      "epoch": 1.3027711255559358,
+      "grad_norm": 0.3174087703227997,
+      "learning_rate": 1.4310079061560438e-06,
+      "loss": 0.1573,
+      "step": 2856
+    },
+    {
+      "epoch": 1.3032272779108223,
+      "grad_norm": 0.3158583641052246,
+      "learning_rate": 1.4293384520331772e-06,
+      "loss": 0.1506,
+      "step": 2857
+    },
+    {
+      "epoch": 1.3036834302657088,
+      "grad_norm": 0.30257847905158997,
+      "learning_rate": 1.4276695823846104e-06,
+      "loss": 0.1567,
+      "step": 2858
+    },
+    {
+      "epoch": 1.3041395826205953,
+      "grad_norm": 0.30727842450141907,
+      "learning_rate": 1.4260012981213796e-06,
+      "loss": 0.1509,
+      "step": 2859
+    },
+    {
+      "epoch": 1.3045957349754818,
+      "grad_norm": 0.3066655397415161,
+      "learning_rate": 1.424333600154202e-06,
+      "loss": 0.1495,
+      "step": 2860
+    },
+    {
+      "epoch": 1.3050518873303685,
+      "grad_norm": 0.3215961456298828,
+      "learning_rate": 1.4226664893934727e-06,
+      "loss": 0.1482,
+      "step": 2861
+    },
+    {
+      "epoch": 1.3055080396852548,
+      "grad_norm": 0.29976993799209595,
+      "learning_rate": 1.4209999667492696e-06,
+      "loss": 0.1548,
+      "step": 2862
+    },
+    {
+      "epoch": 1.3059641920401415,
+      "grad_norm": 0.31503909826278687,
+      "learning_rate": 1.419334033131347e-06,
+      "loss": 0.1534,
+      "step": 2863
+    },
+    {
+      "epoch": 1.306420344395028,
+      "grad_norm": 0.3186908960342407,
+      "learning_rate": 1.4176686894491386e-06,
+      "loss": 0.1571,
+      "step": 2864
+    },
+    {
+      "epoch": 1.3068764967499145,
+      "grad_norm": 0.3151373863220215,
+      "learning_rate": 1.416003936611755e-06,
+      "loss": 0.1454,
+      "step": 2865
+    },
+    {
+      "epoch": 1.307332649104801,
+      "grad_norm": 0.32813790440559387,
+      "learning_rate": 1.4143397755279852e-06,
+      "loss": 0.1536,
+      "step": 2866
+    },
+    {
+      "epoch": 1.3077888014596875,
+      "grad_norm": 0.3181378245353699,
+      "learning_rate": 1.412676207106295e-06,
+      "loss": 0.1549,
+      "step": 2867
+    },
+    {
+      "epoch": 1.3082449538145742,
+      "grad_norm": 0.30067533254623413,
+      "learning_rate": 1.411013232254827e-06,
+      "loss": 0.1443,
+      "step": 2868
+    },
+    {
+      "epoch": 1.3087011061694607,
+      "grad_norm": 0.33433085680007935,
+      "learning_rate": 1.4093508518813987e-06,
+      "loss": 0.1624,
+      "step": 2869
+    },
+    {
+      "epoch": 1.3091572585243472,
+      "grad_norm": 0.3428129553794861,
+      "learning_rate": 1.407689066893505e-06,
+      "loss": 0.153,
+      "step": 2870
+    },
+    {
+      "epoch": 1.3096134108792337,
+      "grad_norm": 0.2925196588039398,
+      "learning_rate": 1.4060278781983144e-06,
+      "loss": 0.1434,
+      "step": 2871
+    },
+    {
+      "epoch": 1.3100695632341202,
+      "grad_norm": 0.3108038604259491,
+      "learning_rate": 1.4043672867026693e-06,
+      "loss": 0.158,
+      "step": 2872
+    },
+    {
+      "epoch": 1.3105257155890067,
+      "grad_norm": 0.3009796142578125,
+      "learning_rate": 1.402707293313087e-06,
+      "loss": 0.1485,
+      "step": 2873
+    },
+    {
+      "epoch": 1.3109818679438932,
+      "grad_norm": 0.3226380944252014,
+      "learning_rate": 1.4010478989357585e-06,
+      "loss": 0.1566,
+      "step": 2874
+    },
+    {
+      "epoch": 1.3114380202987799,
+      "grad_norm": 0.31393906474113464,
+      "learning_rate": 1.399389104476547e-06,
+      "loss": 0.152,
+      "step": 2875
+    },
+    {
+      "epoch": 1.3118941726536664,
+      "grad_norm": 0.3250656723976135,
+      "learning_rate": 1.3977309108409882e-06,
+      "loss": 0.1579,
+      "step": 2876
+    },
+    {
+      "epoch": 1.3123503250085529,
+      "grad_norm": 0.31130245327949524,
+      "learning_rate": 1.3960733189342907e-06,
+      "loss": 0.1504,
+      "step": 2877
+    },
+    {
+      "epoch": 1.3128064773634394,
+      "grad_norm": 0.31418856978416443,
+      "learning_rate": 1.3944163296613354e-06,
+      "loss": 0.1511,
+      "step": 2878
+    },
+    {
+      "epoch": 1.3132626297183259,
+      "grad_norm": 0.300039678812027,
+      "learning_rate": 1.3927599439266717e-06,
+      "loss": 0.1512,
+      "step": 2879
+    },
+    {
+      "epoch": 1.3137187820732126,
+      "grad_norm": 0.3436013162136078,
+      "learning_rate": 1.3911041626345215e-06,
+      "loss": 0.1583,
+      "step": 2880
+    },
+    {
+      "epoch": 1.3141749344280989,
+      "grad_norm": 0.329882949590683,
+      "learning_rate": 1.3894489866887757e-06,
+      "loss": 0.1608,
+      "step": 2881
+    },
+    {
+      "epoch": 1.3146310867829856,
+      "grad_norm": 0.30026715993881226,
+      "learning_rate": 1.3877944169929947e-06,
+      "loss": 0.1593,
+      "step": 2882
+    },
+    {
+      "epoch": 1.315087239137872,
+      "grad_norm": 0.29953205585479736,
+      "learning_rate": 1.3861404544504078e-06,
+      "loss": 0.1478,
+      "step": 2883
+    },
+    {
+      "epoch": 1.3155433914927586,
+      "grad_norm": 0.2963915467262268,
+      "learning_rate": 1.3844870999639154e-06,
+      "loss": 0.1416,
+      "step": 2884
+    },
+    {
+      "epoch": 1.315999543847645,
+      "grad_norm": 0.3181801736354828,
+      "learning_rate": 1.3828343544360829e-06,
+      "loss": 0.1644,
+      "step": 2885
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.32313159108161926,
+      "learning_rate": 1.3811822187691443e-06,
+      "loss": 0.1559,
+      "step": 2886
+    },
+    {
+      "epoch": 1.3169118485574183,
+      "grad_norm": 0.3114675283432007,
+      "learning_rate": 1.3795306938649993e-06,
+      "loss": 0.158,
+      "step": 2887
+    },
+    {
+      "epoch": 1.3173680009123048,
+      "grad_norm": 0.33845824003219604,
+      "learning_rate": 1.3778797806252186e-06,
+      "loss": 0.1638,
+      "step": 2888
+    },
+    {
+      "epoch": 1.3178241532671913,
+      "grad_norm": 0.3146835267543793,
+      "learning_rate": 1.376229479951034e-06,
+      "loss": 0.1541,
+      "step": 2889
+    },
+    {
+      "epoch": 1.3182803056220778,
+      "grad_norm": 0.3196883797645569,
+      "learning_rate": 1.3745797927433454e-06,
+      "loss": 0.1538,
+      "step": 2890
+    },
+    {
+      "epoch": 1.3187364579769643,
+      "grad_norm": 0.30379146337509155,
+      "learning_rate": 1.3729307199027162e-06,
+      "loss": 0.1581,
+      "step": 2891
+    },
+    {
+      "epoch": 1.3191926103318508,
+      "grad_norm": 0.2985891103744507,
+      "learning_rate": 1.3712822623293776e-06,
+      "loss": 0.1513,
+      "step": 2892
+    },
+    {
+      "epoch": 1.3196487626867373,
+      "grad_norm": 0.31963807344436646,
+      "learning_rate": 1.3696344209232216e-06,
+      "loss": 0.1467,
+      "step": 2893
+    },
+    {
+      "epoch": 1.320104915041624,
+      "grad_norm": 0.3292235732078552,
+      "learning_rate": 1.3679871965838048e-06,
+      "loss": 0.1526,
+      "step": 2894
+    },
+    {
+      "epoch": 1.3205610673965105,
+      "grad_norm": 0.33198538422584534,
+      "learning_rate": 1.366340590210348e-06,
+      "loss": 0.164,
+      "step": 2895
+    },
+    {
+      "epoch": 1.321017219751397,
+      "grad_norm": 0.3235149681568146,
+      "learning_rate": 1.3646946027017333e-06,
+      "loss": 0.1696,
+      "step": 2896
+    },
+    {
+      "epoch": 1.3214733721062835,
+      "grad_norm": 0.30945852398872375,
+      "learning_rate": 1.363049234956505e-06,
+      "loss": 0.1548,
+      "step": 2897
+    },
+    {
+      "epoch": 1.32192952446117,
+      "grad_norm": 0.31654229760169983,
+      "learning_rate": 1.3614044878728711e-06,
+      "loss": 0.1514,
+      "step": 2898
+    },
+    {
+      "epoch": 1.3223856768160567,
+      "grad_norm": 0.3196168839931488,
+      "learning_rate": 1.359760362348697e-06,
+      "loss": 0.1543,
+      "step": 2899
+    },
+    {
+      "epoch": 1.3228418291709432,
+      "grad_norm": 0.3089727759361267,
+      "learning_rate": 1.3581168592815142e-06,
+      "loss": 0.1532,
+      "step": 2900
+    },
+    {
+      "epoch": 1.3232979815258297,
+      "grad_norm": 0.31021323800086975,
+      "learning_rate": 1.3564739795685095e-06,
+      "loss": 0.1568,
+      "step": 2901
+    },
+    {
+      "epoch": 1.3237541338807162,
+      "grad_norm": 0.3066624402999878,
+      "learning_rate": 1.3548317241065311e-06,
+      "loss": 0.1525,
+      "step": 2902
+    },
+    {
+      "epoch": 1.3242102862356027,
+      "grad_norm": 0.3322006165981293,
+      "learning_rate": 1.353190093792087e-06,
+      "loss": 0.1481,
+      "step": 2903
+    },
+    {
+      "epoch": 1.3246664385904892,
+      "grad_norm": 0.302090048789978,
+      "learning_rate": 1.3515490895213428e-06,
+      "loss": 0.1479,
+      "step": 2904
+    },
+    {
+      "epoch": 1.3251225909453757,
+      "grad_norm": 0.32400602102279663,
+      "learning_rate": 1.3499087121901234e-06,
+      "loss": 0.1632,
+      "step": 2905
+    },
+    {
+      "epoch": 1.3255787433002624,
+      "grad_norm": 0.31271302700042725,
+      "learning_rate": 1.3482689626939099e-06,
+      "loss": 0.158,
+      "step": 2906
+    },
+    {
+      "epoch": 1.3260348956551489,
+      "grad_norm": 0.3004334270954132,
+      "learning_rate": 1.3466298419278442e-06,
+      "loss": 0.1493,
+      "step": 2907
+    },
+    {
+      "epoch": 1.3264910480100354,
+      "grad_norm": 0.31443971395492554,
+      "learning_rate": 1.3449913507867202e-06,
+      "loss": 0.1501,
+      "step": 2908
+    },
+    {
+      "epoch": 1.3269472003649219,
+      "grad_norm": 0.30077528953552246,
+      "learning_rate": 1.343353490164992e-06,
+      "loss": 0.1516,
+      "step": 2909
+    },
+    {
+      "epoch": 1.3274033527198084,
+      "grad_norm": 0.31191545724868774,
+      "learning_rate": 1.3417162609567677e-06,
+      "loss": 0.1566,
+      "step": 2910
+    },
+    {
+      "epoch": 1.327859505074695,
+      "grad_norm": 0.3076206147670746,
+      "learning_rate": 1.3400796640558106e-06,
+      "loss": 0.1479,
+      "step": 2911
+    },
+    {
+      "epoch": 1.3283156574295814,
+      "grad_norm": 0.330691397190094,
+      "learning_rate": 1.3384437003555394e-06,
+      "loss": 0.1665,
+      "step": 2912
+    },
+    {
+      "epoch": 1.328771809784468,
+      "grad_norm": 0.30503392219543457,
+      "learning_rate": 1.3368083707490259e-06,
+      "loss": 0.157,
+      "step": 2913
+    },
+    {
+      "epoch": 1.3292279621393546,
+      "grad_norm": 0.3229046165943146,
+      "learning_rate": 1.3351736761289984e-06,
+      "loss": 0.1529,
+      "step": 2914
+    },
+    {
+      "epoch": 1.329684114494241,
+      "grad_norm": 0.2919079661369324,
+      "learning_rate": 1.3335396173878367e-06,
+      "loss": 0.1464,
+      "step": 2915
+    },
+    {
+      "epoch": 1.3301402668491276,
+      "grad_norm": 0.3344881236553192,
+      "learning_rate": 1.3319061954175733e-06,
+      "loss": 0.1648,
+      "step": 2916
+    },
+    {
+      "epoch": 1.330596419204014,
+      "grad_norm": 0.3266892731189728,
+      "learning_rate": 1.330273411109893e-06,
+      "loss": 0.1572,
+      "step": 2917
+    },
+    {
+      "epoch": 1.3310525715589008,
+      "grad_norm": 0.3084025979042053,
+      "learning_rate": 1.3286412653561326e-06,
+      "loss": 0.153,
+      "step": 2918
+    },
+    {
+      "epoch": 1.3315087239137873,
+      "grad_norm": 0.3008584678173065,
+      "learning_rate": 1.3270097590472828e-06,
+      "loss": 0.1429,
+      "step": 2919
+    },
+    {
+      "epoch": 1.3319648762686738,
+      "grad_norm": 0.292179137468338,
+      "learning_rate": 1.3253788930739816e-06,
+      "loss": 0.145,
+      "step": 2920
+    },
+    {
+      "epoch": 1.3324210286235603,
+      "grad_norm": 0.2984185218811035,
+      "learning_rate": 1.3237486683265186e-06,
+      "loss": 0.1498,
+      "step": 2921
+    },
+    {
+      "epoch": 1.3328771809784468,
+      "grad_norm": 0.3196554481983185,
+      "learning_rate": 1.3221190856948352e-06,
+      "loss": 0.1532,
+      "step": 2922
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.3272205591201782,
+      "learning_rate": 1.3204901460685198e-06,
+      "loss": 0.1595,
+      "step": 2923
+    },
+    {
+      "epoch": 1.3337894856882198,
+      "grad_norm": 0.2987115681171417,
+      "learning_rate": 1.3188618503368117e-06,
+      "loss": 0.1504,
+      "step": 2924
+    },
+    {
+      "epoch": 1.3342456380431065,
+      "grad_norm": 0.3112976551055908,
+      "learning_rate": 1.3172341993885968e-06,
+      "loss": 0.156,
+      "step": 2925
+    },
+    {
+      "epoch": 1.334701790397993,
+      "grad_norm": 0.3252262473106384,
+      "learning_rate": 1.31560719411241e-06,
+      "loss": 0.1611,
+      "step": 2926
+    },
+    {
+      "epoch": 1.3351579427528795,
+      "grad_norm": 0.31838521361351013,
+      "learning_rate": 1.3139808353964342e-06,
+      "loss": 0.1468,
+      "step": 2927
+    },
+    {
+      "epoch": 1.335614095107766,
+      "grad_norm": 0.30408743023872375,
+      "learning_rate": 1.3123551241284976e-06,
+      "loss": 0.1443,
+      "step": 2928
+    },
+    {
+      "epoch": 1.3360702474626525,
+      "grad_norm": 0.31184643507003784,
+      "learning_rate": 1.3107300611960776e-06,
+      "loss": 0.1501,
+      "step": 2929
+    },
+    {
+      "epoch": 1.3365263998175392,
+      "grad_norm": 0.3136065900325775,
+      "learning_rate": 1.3091056474862968e-06,
+      "loss": 0.1569,
+      "step": 2930
+    },
+    {
+      "epoch": 1.3369825521724255,
+      "grad_norm": 0.302523136138916,
+      "learning_rate": 1.3074818838859221e-06,
+      "loss": 0.1521,
+      "step": 2931
+    },
+    {
+      "epoch": 1.3374387045273122,
+      "grad_norm": 0.31496572494506836,
+      "learning_rate": 1.3058587712813665e-06,
+      "loss": 0.1598,
+      "step": 2932
+    },
+    {
+      "epoch": 1.3378948568821987,
+      "grad_norm": 0.33140820264816284,
+      "learning_rate": 1.304236310558687e-06,
+      "loss": 0.1551,
+      "step": 2933
+    },
+    {
+      "epoch": 1.3383510092370852,
+      "grad_norm": 0.31709909439086914,
+      "learning_rate": 1.3026145026035854e-06,
+      "loss": 0.1622,
+      "step": 2934
+    },
+    {
+      "epoch": 1.3388071615919717,
+      "grad_norm": 0.3094816207885742,
+      "learning_rate": 1.300993348301407e-06,
+      "loss": 0.1581,
+      "step": 2935
+    },
+    {
+      "epoch": 1.3392633139468582,
+      "grad_norm": 0.310981810092926,
+      "learning_rate": 1.299372848537139e-06,
+      "loss": 0.1492,
+      "step": 2936
+    },
+    {
+      "epoch": 1.3397194663017449,
+      "grad_norm": 0.3038754463195801,
+      "learning_rate": 1.2977530041954146e-06,
+      "loss": 0.153,
+      "step": 2937
+    },
+    {
+      "epoch": 1.3401756186566314,
+      "grad_norm": 0.309890478849411,
+      "learning_rate": 1.2961338161605047e-06,
+      "loss": 0.1547,
+      "step": 2938
+    },
+    {
+      "epoch": 1.3406317710115179,
+      "grad_norm": 0.3166119456291199,
+      "learning_rate": 1.2945152853163267e-06,
+      "loss": 0.1472,
+      "step": 2939
+    },
+    {
+      "epoch": 1.3410879233664044,
+      "grad_norm": 0.30829304456710815,
+      "learning_rate": 1.292897412546436e-06,
+      "loss": 0.1605,
+      "step": 2940
+    },
+    {
+      "epoch": 1.3415440757212909,
+      "grad_norm": 0.31166979670524597,
+      "learning_rate": 1.291280198734029e-06,
+      "loss": 0.1519,
+      "step": 2941
+    },
+    {
+      "epoch": 1.3420002280761776,
+      "grad_norm": 0.3307502269744873,
+      "learning_rate": 1.2896636447619431e-06,
+      "loss": 0.1553,
+      "step": 2942
+    },
+    {
+      "epoch": 1.3424563804310639,
+      "grad_norm": 0.32278013229370117,
+      "learning_rate": 1.288047751512655e-06,
+      "loss": 0.1608,
+      "step": 2943
+    },
+    {
+      "epoch": 1.3429125327859506,
+      "grad_norm": 0.31143462657928467,
+      "learning_rate": 1.286432519868282e-06,
+      "loss": 0.148,
+      "step": 2944
+    },
+    {
+      "epoch": 1.343368685140837,
+      "grad_norm": 0.3010978400707245,
+      "learning_rate": 1.2848179507105796e-06,
+      "loss": 0.1431,
+      "step": 2945
+    },
+    {
+      "epoch": 1.3438248374957236,
+      "grad_norm": 0.3140721917152405,
+      "learning_rate": 1.2832040449209399e-06,
+      "loss": 0.1543,
+      "step": 2946
+    },
+    {
+      "epoch": 1.34428098985061,
+      "grad_norm": 0.3114619255065918,
+      "learning_rate": 1.2815908033803952e-06,
+      "loss": 0.155,
+      "step": 2947
+    },
+    {
+      "epoch": 1.3447371422054966,
+      "grad_norm": 0.3143191933631897,
+      "learning_rate": 1.2799782269696137e-06,
+      "loss": 0.1502,
+      "step": 2948
+    },
+    {
+      "epoch": 1.3451932945603833,
+      "grad_norm": 0.2854684591293335,
+      "learning_rate": 1.278366316568902e-06,
+      "loss": 0.1414,
+      "step": 2949
+    },
+    {
+      "epoch": 1.3456494469152698,
+      "grad_norm": 0.31013891100883484,
+      "learning_rate": 1.2767550730582017e-06,
+      "loss": 0.1592,
+      "step": 2950
+    },
+    {
+      "epoch": 1.3461055992701563,
+      "grad_norm": 0.32352420687675476,
+      "learning_rate": 1.2751444973170904e-06,
+      "loss": 0.1532,
+      "step": 2951
+    },
+    {
+      "epoch": 1.3465617516250428,
+      "grad_norm": 0.3085663616657257,
+      "learning_rate": 1.2735345902247831e-06,
+      "loss": 0.1564,
+      "step": 2952
+    },
+    {
+      "epoch": 1.3470179039799293,
+      "grad_norm": 0.3159683644771576,
+      "learning_rate": 1.2719253526601272e-06,
+      "loss": 0.1471,
+      "step": 2953
+    },
+    {
+      "epoch": 1.3474740563348158,
+      "grad_norm": 0.30028852820396423,
+      "learning_rate": 1.2703167855016057e-06,
+      "loss": 0.1526,
+      "step": 2954
+    },
+    {
+      "epoch": 1.3479302086897023,
+      "grad_norm": 0.33059003949165344,
+      "learning_rate": 1.2687088896273365e-06,
+      "loss": 0.1585,
+      "step": 2955
+    },
+    {
+      "epoch": 1.348386361044589,
+      "grad_norm": 0.3213483691215515,
+      "learning_rate": 1.2671016659150693e-06,
+      "loss": 0.1621,
+      "step": 2956
+    },
+    {
+      "epoch": 1.3488425133994755,
+      "grad_norm": 0.30662602186203003,
+      "learning_rate": 1.2654951152421879e-06,
+      "loss": 0.1485,
+      "step": 2957
+    },
+    {
+      "epoch": 1.349298665754362,
+      "grad_norm": 0.3204701840877533,
+      "learning_rate": 1.2638892384857078e-06,
+      "loss": 0.1629,
+      "step": 2958
+    },
+    {
+      "epoch": 1.3497548181092485,
+      "grad_norm": 0.3220400810241699,
+      "learning_rate": 1.2622840365222783e-06,
+      "loss": 0.1515,
+      "step": 2959
+    },
+    {
+      "epoch": 1.350210970464135,
+      "grad_norm": 0.33348965644836426,
+      "learning_rate": 1.2606795102281806e-06,
+      "loss": 0.1519,
+      "step": 2960
+    },
+    {
+      "epoch": 1.3506671228190217,
+      "grad_norm": 0.3342219889163971,
+      "learning_rate": 1.2590756604793242e-06,
+      "loss": 0.158,
+      "step": 2961
+    },
+    {
+      "epoch": 1.351123275173908,
+      "grad_norm": 0.31400221586227417,
+      "learning_rate": 1.2574724881512517e-06,
+      "loss": 0.1507,
+      "step": 2962
+    },
+    {
+      "epoch": 1.3515794275287947,
+      "grad_norm": 0.30305933952331543,
+      "learning_rate": 1.2558699941191347e-06,
+      "loss": 0.1506,
+      "step": 2963
+    },
+    {
+      "epoch": 1.3520355798836812,
+      "grad_norm": 0.3086269795894623,
+      "learning_rate": 1.2542681792577749e-06,
+      "loss": 0.1527,
+      "step": 2964
+    },
+    {
+      "epoch": 1.3524917322385677,
+      "grad_norm": 0.30883389711380005,
+      "learning_rate": 1.2526670444416034e-06,
+      "loss": 0.1545,
+      "step": 2965
+    },
+    {
+      "epoch": 1.3529478845934542,
+      "grad_norm": 0.31932634115219116,
+      "learning_rate": 1.2510665905446787e-06,
+      "loss": 0.1506,
+      "step": 2966
+    },
+    {
+      "epoch": 1.3534040369483407,
+      "grad_norm": 0.30006587505340576,
+      "learning_rate": 1.249466818440691e-06,
+      "loss": 0.1468,
+      "step": 2967
+    },
+    {
+      "epoch": 1.3538601893032274,
+      "grad_norm": 0.2946045696735382,
+      "learning_rate": 1.2478677290029554e-06,
+      "loss": 0.1507,
+      "step": 2968
+    },
+    {
+      "epoch": 1.3543163416581139,
+      "grad_norm": 0.3050330877304077,
+      "learning_rate": 1.2462693231044138e-06,
+      "loss": 0.1566,
+      "step": 2969
+    },
+    {
+      "epoch": 1.3547724940130004,
+      "grad_norm": 0.30813461542129517,
+      "learning_rate": 1.244671601617638e-06,
+      "loss": 0.1532,
+      "step": 2970
+    },
+    {
+      "epoch": 1.3552286463678869,
+      "grad_norm": 0.32525256276130676,
+      "learning_rate": 1.2430745654148238e-06,
+      "loss": 0.1554,
+      "step": 2971
+    },
+    {
+      "epoch": 1.3556847987227734,
+      "grad_norm": 0.3163076639175415,
+      "learning_rate": 1.2414782153677935e-06,
+      "loss": 0.1555,
+      "step": 2972
+    },
+    {
+      "epoch": 1.3561409510776599,
+      "grad_norm": 0.31215327978134155,
+      "learning_rate": 1.2398825523479935e-06,
+      "loss": 0.1552,
+      "step": 2973
+    },
+    {
+      "epoch": 1.3565971034325464,
+      "grad_norm": 0.3032691776752472,
+      "learning_rate": 1.2382875772264985e-06,
+      "loss": 0.1515,
+      "step": 2974
+    },
+    {
+      "epoch": 1.357053255787433,
+      "grad_norm": 0.314069539308548,
+      "learning_rate": 1.236693290874005e-06,
+      "loss": 0.1583,
+      "step": 2975
+    },
+    {
+      "epoch": 1.3575094081423196,
+      "grad_norm": 0.31185275316238403,
+      "learning_rate": 1.235099694160834e-06,
+      "loss": 0.1513,
+      "step": 2976
+    },
+    {
+      "epoch": 1.357965560497206,
+      "grad_norm": 0.3366699814796448,
+      "learning_rate": 1.2335067879569298e-06,
+      "loss": 0.1607,
+      "step": 2977
+    },
+    {
+      "epoch": 1.3584217128520926,
+      "grad_norm": 0.3172808885574341,
+      "learning_rate": 1.231914573131861e-06,
+      "loss": 0.1567,
+      "step": 2978
+    },
+    {
+      "epoch": 1.358877865206979,
+      "grad_norm": 0.32084766030311584,
+      "learning_rate": 1.2303230505548162e-06,
+      "loss": 0.1562,
+      "step": 2979
+    },
+    {
+      "epoch": 1.3593340175618658,
+      "grad_norm": 0.307922899723053,
+      "learning_rate": 1.2287322210946098e-06,
+      "loss": 0.1538,
+      "step": 2980
+    },
+    {
+      "epoch": 1.3597901699167523,
+      "grad_norm": 0.30017930269241333,
+      "learning_rate": 1.2271420856196744e-06,
+      "loss": 0.1488,
+      "step": 2981
+    },
+    {
+      "epoch": 1.3602463222716388,
+      "grad_norm": 0.306976318359375,
+      "learning_rate": 1.2255526449980667e-06,
+      "loss": 0.1547,
+      "step": 2982
+    },
+    {
+      "epoch": 1.3607024746265253,
+      "grad_norm": 0.2977949380874634,
+      "learning_rate": 1.2239639000974615e-06,
+      "loss": 0.1413,
+      "step": 2983
+    },
+    {
+      "epoch": 1.3611586269814118,
+      "grad_norm": 0.30933767557144165,
+      "learning_rate": 1.2223758517851556e-06,
+      "loss": 0.1485,
+      "step": 2984
+    },
+    {
+      "epoch": 1.3616147793362983,
+      "grad_norm": 0.30757781863212585,
+      "learning_rate": 1.2207885009280637e-06,
+      "loss": 0.1525,
+      "step": 2985
+    },
+    {
+      "epoch": 1.3620709316911848,
+      "grad_norm": 0.29795169830322266,
+      "learning_rate": 1.219201848392722e-06,
+      "loss": 0.1468,
+      "step": 2986
+    },
+    {
+      "epoch": 1.3625270840460715,
+      "grad_norm": 0.3072497248649597,
+      "learning_rate": 1.2176158950452838e-06,
+      "loss": 0.1477,
+      "step": 2987
+    },
+    {
+      "epoch": 1.362983236400958,
+      "grad_norm": 0.31369003653526306,
+      "learning_rate": 1.21603064175152e-06,
+      "loss": 0.1522,
+      "step": 2988
+    },
+    {
+      "epoch": 1.3634393887558445,
+      "grad_norm": 0.302862286567688,
+      "learning_rate": 1.2144460893768223e-06,
+      "loss": 0.1484,
+      "step": 2989
+    },
+    {
+      "epoch": 1.363895541110731,
+      "grad_norm": 0.30438125133514404,
+      "learning_rate": 1.2128622387861985e-06,
+      "loss": 0.1456,
+      "step": 2990
+    },
+    {
+      "epoch": 1.3643516934656175,
+      "grad_norm": 0.2957283854484558,
+      "learning_rate": 1.211279090844272e-06,
+      "loss": 0.148,
+      "step": 2991
+    },
+    {
+      "epoch": 1.3648078458205042,
+      "grad_norm": 0.30728259682655334,
+      "learning_rate": 1.2096966464152837e-06,
+      "loss": 0.148,
+      "step": 2992
+    },
+    {
+      "epoch": 1.3652639981753905,
+      "grad_norm": 0.30239006876945496,
+      "learning_rate": 1.2081149063630902e-06,
+      "loss": 0.1534,
+      "step": 2993
+    },
+    {
+      "epoch": 1.3657201505302772,
+      "grad_norm": 0.3303487300872803,
+      "learning_rate": 1.2065338715511634e-06,
+      "loss": 0.1675,
+      "step": 2994
+    },
+    {
+      "epoch": 1.3661763028851637,
+      "grad_norm": 0.31265461444854736,
+      "learning_rate": 1.2049535428425906e-06,
+      "loss": 0.1624,
+      "step": 2995
+    },
+    {
+      "epoch": 1.3666324552400502,
+      "grad_norm": 0.29741230607032776,
+      "learning_rate": 1.2033739211000729e-06,
+      "loss": 0.1443,
+      "step": 2996
+    },
+    {
+      "epoch": 1.3670886075949367,
+      "grad_norm": 0.31509754061698914,
+      "learning_rate": 1.2017950071859272e-06,
+      "loss": 0.156,
+      "step": 2997
+    },
+    {
+      "epoch": 1.3675447599498232,
+      "grad_norm": 0.30511605739593506,
+      "learning_rate": 1.2002168019620825e-06,
+      "loss": 0.1519,
+      "step": 2998
+    },
+    {
+      "epoch": 1.3680009123047099,
+      "grad_norm": 0.3008752763271332,
+      "learning_rate": 1.198639306290081e-06,
+      "loss": 0.1566,
+      "step": 2999
+    },
+    {
+      "epoch": 1.3684570646595964,
+      "grad_norm": 0.3070181608200073,
+      "learning_rate": 1.1970625210310768e-06,
+      "loss": 0.148,
+      "step": 3000
+    },
+    {
+      "epoch": 1.3689132170144829,
+      "grad_norm": 0.3170846104621887,
+      "learning_rate": 1.1954864470458394e-06,
+      "loss": 0.1572,
+      "step": 3001
+    },
+    {
+      "epoch": 1.3693693693693694,
+      "grad_norm": 0.3217135965824127,
+      "learning_rate": 1.1939110851947466e-06,
+      "loss": 0.1534,
+      "step": 3002
+    },
+    {
+      "epoch": 1.3698255217242559,
+      "grad_norm": 0.2972395718097687,
+      "learning_rate": 1.1923364363377887e-06,
+      "loss": 0.1468,
+      "step": 3003
+    },
+    {
+      "epoch": 1.3702816740791424,
+      "grad_norm": 0.30074235796928406,
+      "learning_rate": 1.190762501334566e-06,
+      "loss": 0.1563,
+      "step": 3004
+    },
+    {
+      "epoch": 1.3707378264340289,
+      "grad_norm": 0.3159099817276001,
+      "learning_rate": 1.1891892810442917e-06,
+      "loss": 0.1515,
+      "step": 3005
+    },
+    {
+      "epoch": 1.3711939787889156,
+      "grad_norm": 0.3177895247936249,
+      "learning_rate": 1.187616776325786e-06,
+      "loss": 0.1539,
+      "step": 3006
+    },
+    {
+      "epoch": 1.371650131143802,
+      "grad_norm": 0.3224925994873047,
+      "learning_rate": 1.1860449880374794e-06,
+      "loss": 0.1602,
+      "step": 3007
+    },
+    {
+      "epoch": 1.3721062834986886,
+      "grad_norm": 0.3165651857852936,
+      "learning_rate": 1.184473917037411e-06,
+      "loss": 0.1546,
+      "step": 3008
+    },
+    {
+      "epoch": 1.372562435853575,
+      "grad_norm": 0.32890525460243225,
+      "learning_rate": 1.18290356418323e-06,
+      "loss": 0.1515,
+      "step": 3009
+    },
+    {
+      "epoch": 1.3730185882084616,
+      "grad_norm": 0.3059317171573639,
+      "learning_rate": 1.1813339303321895e-06,
+      "loss": 0.1513,
+      "step": 3010
+    },
+    {
+      "epoch": 1.3734747405633483,
+      "grad_norm": 0.3198723793029785,
+      "learning_rate": 1.179765016341155e-06,
+      "loss": 0.1632,
+      "step": 3011
+    },
+    {
+      "epoch": 1.3739308929182346,
+      "grad_norm": 0.29671710729599,
+      "learning_rate": 1.1781968230665978e-06,
+      "loss": 0.1505,
+      "step": 3012
+    },
+    {
+      "epoch": 1.3743870452731213,
+      "grad_norm": 0.3080565631389618,
+      "learning_rate": 1.1766293513645932e-06,
+      "loss": 0.1485,
+      "step": 3013
+    },
+    {
+      "epoch": 1.3748431976280078,
+      "grad_norm": 0.29941487312316895,
+      "learning_rate": 1.175062602090825e-06,
+      "loss": 0.1548,
+      "step": 3014
+    },
+    {
+      "epoch": 1.3752993499828943,
+      "grad_norm": 0.30851706862449646,
+      "learning_rate": 1.1734965761005815e-06,
+      "loss": 0.1485,
+      "step": 3015
+    },
+    {
+      "epoch": 1.3757555023377808,
+      "grad_norm": 0.29139479994773865,
+      "learning_rate": 1.1719312742487565e-06,
+      "loss": 0.1445,
+      "step": 3016
+    },
+    {
+      "epoch": 1.3762116546926673,
+      "grad_norm": 0.31408241391181946,
+      "learning_rate": 1.1703666973898487e-06,
+      "loss": 0.146,
+      "step": 3017
+    },
+    {
+      "epoch": 1.376667807047554,
+      "grad_norm": 0.3236776888370514,
+      "learning_rate": 1.1688028463779597e-06,
+      "loss": 0.1498,
+      "step": 3018
+    },
+    {
+      "epoch": 1.3771239594024405,
+      "grad_norm": 0.2994031012058258,
+      "learning_rate": 1.167239722066798e-06,
+      "loss": 0.1481,
+      "step": 3019
+    },
+    {
+      "epoch": 1.377580111757327,
+      "grad_norm": 0.3160935342311859,
+      "learning_rate": 1.1656773253096716e-06,
+      "loss": 0.1578,
+      "step": 3020
+    },
+    {
+      "epoch": 1.3780362641122135,
+      "grad_norm": 0.30692413449287415,
+      "learning_rate": 1.1641156569594944e-06,
+      "loss": 0.1469,
+      "step": 3021
+    },
+    {
+      "epoch": 1.3784924164671,
+      "grad_norm": 0.3264402747154236,
+      "learning_rate": 1.1625547178687812e-06,
+      "loss": 0.1511,
+      "step": 3022
+    },
+    {
+      "epoch": 1.3789485688219865,
+      "grad_norm": 0.3243708610534668,
+      "learning_rate": 1.1609945088896488e-06,
+      "loss": 0.163,
+      "step": 3023
+    },
+    {
+      "epoch": 1.379404721176873,
+      "grad_norm": 0.3036910891532898,
+      "learning_rate": 1.1594350308738154e-06,
+      "loss": 0.1529,
+      "step": 3024
+    },
+    {
+      "epoch": 1.3798608735317597,
+      "grad_norm": 0.32628411054611206,
+      "learning_rate": 1.1578762846726e-06,
+      "loss": 0.1573,
+      "step": 3025
+    },
+    {
+      "epoch": 1.3803170258866462,
+      "grad_norm": 0.30349403619766235,
+      "learning_rate": 1.1563182711369218e-06,
+      "loss": 0.1485,
+      "step": 3026
+    },
+    {
+      "epoch": 1.3807731782415327,
+      "grad_norm": 0.3151264786720276,
+      "learning_rate": 1.154760991117302e-06,
+      "loss": 0.1434,
+      "step": 3027
+    },
+    {
+      "epoch": 1.3812293305964192,
+      "grad_norm": 0.3015769124031067,
+      "learning_rate": 1.1532044454638598e-06,
+      "loss": 0.1518,
+      "step": 3028
+    },
+    {
+      "epoch": 1.3816854829513057,
+      "grad_norm": 0.31136202812194824,
+      "learning_rate": 1.1516486350263131e-06,
+      "loss": 0.1461,
+      "step": 3029
+    },
+    {
+      "epoch": 1.3821416353061924,
+      "grad_norm": 0.3052590489387512,
+      "learning_rate": 1.1500935606539784e-06,
+      "loss": 0.1543,
+      "step": 3030
+    },
+    {
+      "epoch": 1.3825977876610789,
+      "grad_norm": 0.31879687309265137,
+      "learning_rate": 1.1485392231957728e-06,
+      "loss": 0.1596,
+      "step": 3031
+    },
+    {
+      "epoch": 1.3830539400159654,
+      "grad_norm": 0.3095165193080902,
+      "learning_rate": 1.1469856235002088e-06,
+      "loss": 0.1544,
+      "step": 3032
+    },
+    {
+      "epoch": 1.3835100923708519,
+      "grad_norm": 0.33869850635528564,
+      "learning_rate": 1.145432762415396e-06,
+      "loss": 0.1657,
+      "step": 3033
+    },
+    {
+      "epoch": 1.3839662447257384,
+      "grad_norm": 0.32992473244667053,
+      "learning_rate": 1.1438806407890413e-06,
+      "loss": 0.1555,
+      "step": 3034
+    },
+    {
+      "epoch": 1.3844223970806249,
+      "grad_norm": 0.3257879614830017,
+      "learning_rate": 1.1423292594684492e-06,
+      "loss": 0.1562,
+      "step": 3035
+    },
+    {
+      "epoch": 1.3848785494355114,
+      "grad_norm": 0.31231051683425903,
+      "learning_rate": 1.140778619300519e-06,
+      "loss": 0.1565,
+      "step": 3036
+    },
+    {
+      "epoch": 1.385334701790398,
+      "grad_norm": 0.3484162986278534,
+      "learning_rate": 1.1392287211317447e-06,
+      "loss": 0.163,
+      "step": 3037
+    },
+    {
+      "epoch": 1.3857908541452846,
+      "grad_norm": 0.3225407302379608,
+      "learning_rate": 1.137679565808216e-06,
+      "loss": 0.1501,
+      "step": 3038
+    },
+    {
+      "epoch": 1.386247006500171,
+      "grad_norm": 0.31159690022468567,
+      "learning_rate": 1.1361311541756174e-06,
+      "loss": 0.1546,
+      "step": 3039
+    },
+    {
+      "epoch": 1.3867031588550576,
+      "grad_norm": 0.31157949566841125,
+      "learning_rate": 1.1345834870792256e-06,
+      "loss": 0.1622,
+      "step": 3040
+    },
+    {
+      "epoch": 1.387159311209944,
+      "grad_norm": 0.32223495841026306,
+      "learning_rate": 1.1330365653639137e-06,
+      "loss": 0.1535,
+      "step": 3041
+    },
+    {
+      "epoch": 1.3876154635648308,
+      "grad_norm": 0.6560049057006836,
+      "learning_rate": 1.1314903898741472e-06,
+      "loss": 0.1492,
+      "step": 3042
+    },
+    {
+      "epoch": 1.388071615919717,
+      "grad_norm": 0.3122204840183258,
+      "learning_rate": 1.1299449614539827e-06,
+      "loss": 0.1482,
+      "step": 3043
+    },
+    {
+      "epoch": 1.3885277682746038,
+      "grad_norm": 0.30853062868118286,
+      "learning_rate": 1.1284002809470695e-06,
+      "loss": 0.1526,
+      "step": 3044
+    },
+    {
+      "epoch": 1.3889839206294903,
+      "grad_norm": 0.3152260482311249,
+      "learning_rate": 1.1268563491966494e-06,
+      "loss": 0.1491,
+      "step": 3045
+    },
+    {
+      "epoch": 1.3894400729843768,
+      "grad_norm": 0.3197248876094818,
+      "learning_rate": 1.1253131670455548e-06,
+      "loss": 0.1528,
+      "step": 3046
+    },
+    {
+      "epoch": 1.3898962253392633,
+      "grad_norm": 0.3285209536552429,
+      "learning_rate": 1.1237707353362087e-06,
+      "loss": 0.1582,
+      "step": 3047
+    },
+    {
+      "epoch": 1.3903523776941498,
+      "grad_norm": 0.30746176838874817,
+      "learning_rate": 1.1222290549106245e-06,
+      "loss": 0.152,
+      "step": 3048
+    },
+    {
+      "epoch": 1.3908085300490365,
+      "grad_norm": 0.3091007471084595,
+      "learning_rate": 1.1206881266104071e-06,
+      "loss": 0.1463,
+      "step": 3049
+    },
+    {
+      "epoch": 1.391264682403923,
+      "grad_norm": 0.31541907787323,
+      "learning_rate": 1.119147951276749e-06,
+      "loss": 0.1565,
+      "step": 3050
+    },
+    {
+      "epoch": 1.3917208347588095,
+      "grad_norm": 0.30186277627944946,
+      "learning_rate": 1.117608529750431e-06,
+      "loss": 0.1422,
+      "step": 3051
+    },
+    {
+      "epoch": 1.392176987113696,
+      "grad_norm": 0.30865976214408875,
+      "learning_rate": 1.1160698628718253e-06,
+      "loss": 0.1526,
+      "step": 3052
+    },
+    {
+      "epoch": 1.3926331394685825,
+      "grad_norm": 0.29759830236434937,
+      "learning_rate": 1.1145319514808895e-06,
+      "loss": 0.1473,
+      "step": 3053
+    },
+    {
+      "epoch": 1.393089291823469,
+      "grad_norm": 0.3094671368598938,
+      "learning_rate": 1.1129947964171698e-06,
+      "loss": 0.1568,
+      "step": 3054
+    },
+    {
+      "epoch": 1.3935454441783555,
+      "grad_norm": 0.30431169271469116,
+      "learning_rate": 1.111458398519799e-06,
+      "loss": 0.1479,
+      "step": 3055
+    },
+    {
+      "epoch": 1.3940015965332422,
+      "grad_norm": 0.3121318221092224,
+      "learning_rate": 1.1099227586274965e-06,
+      "loss": 0.1551,
+      "step": 3056
+    },
+    {
+      "epoch": 1.3944577488881287,
+      "grad_norm": 0.29787304997444153,
+      "learning_rate": 1.1083878775785703e-06,
+      "loss": 0.1494,
+      "step": 3057
+    },
+    {
+      "epoch": 1.3949139012430152,
+      "grad_norm": 0.3125587999820709,
+      "learning_rate": 1.1068537562109109e-06,
+      "loss": 0.1527,
+      "step": 3058
+    },
+    {
+      "epoch": 1.3953700535979017,
+      "grad_norm": 0.319680780172348,
+      "learning_rate": 1.105320395361995e-06,
+      "loss": 0.1575,
+      "step": 3059
+    },
+    {
+      "epoch": 1.3958262059527882,
+      "grad_norm": 0.3143640458583832,
+      "learning_rate": 1.1037877958688856e-06,
+      "loss": 0.1463,
+      "step": 3060
+    },
+    {
+      "epoch": 1.3962823583076749,
+      "grad_norm": 0.30363672971725464,
+      "learning_rate": 1.1022559585682269e-06,
+      "loss": 0.1506,
+      "step": 3061
+    },
+    {
+      "epoch": 1.3967385106625612,
+      "grad_norm": 0.3088383972644806,
+      "learning_rate": 1.1007248842962516e-06,
+      "loss": 0.1568,
+      "step": 3062
+    },
+    {
+      "epoch": 1.3971946630174479,
+      "grad_norm": 0.39603710174560547,
+      "learning_rate": 1.0991945738887727e-06,
+      "loss": 0.1611,
+      "step": 3063
+    },
+    {
+      "epoch": 1.3976508153723344,
+      "grad_norm": 0.3133648633956909,
+      "learning_rate": 1.0976650281811855e-06,
+      "loss": 0.1547,
+      "step": 3064
+    },
+    {
+      "epoch": 1.3981069677272209,
+      "grad_norm": 0.31006720662117004,
+      "learning_rate": 1.0961362480084711e-06,
+      "loss": 0.1482,
+      "step": 3065
+    },
+    {
+      "epoch": 1.3985631200821074,
+      "grad_norm": 0.31392380595207214,
+      "learning_rate": 1.09460823420519e-06,
+      "loss": 0.1517,
+      "step": 3066
+    },
+    {
+      "epoch": 1.3990192724369939,
+      "grad_norm": 0.33342450857162476,
+      "learning_rate": 1.0930809876054853e-06,
+      "loss": 0.1498,
+      "step": 3067
+    },
+    {
+      "epoch": 1.3994754247918806,
+      "grad_norm": 0.3142637014389038,
+      "learning_rate": 1.091554509043081e-06,
+      "loss": 0.1499,
+      "step": 3068
+    },
+    {
+      "epoch": 1.399931577146767,
+      "grad_norm": 0.31120747327804565,
+      "learning_rate": 1.0900287993512818e-06,
+      "loss": 0.1524,
+      "step": 3069
+    },
+    {
+      "epoch": 1.4003877295016536,
+      "grad_norm": 0.32112669944763184,
+      "learning_rate": 1.0885038593629724e-06,
+      "loss": 0.1543,
+      "step": 3070
+    },
+    {
+      "epoch": 1.40084388185654,
+      "grad_norm": 0.30108073353767395,
+      "learning_rate": 1.086979689910618e-06,
+      "loss": 0.1411,
+      "step": 3071
+    },
+    {
+      "epoch": 1.4013000342114266,
+      "grad_norm": 0.2958739399909973,
+      "learning_rate": 1.0854562918262645e-06,
+      "loss": 0.1465,
+      "step": 3072
+    },
+    {
+      "epoch": 1.401756186566313,
+      "grad_norm": 0.3173351585865021,
+      "learning_rate": 1.0839336659415342e-06,
+      "loss": 0.1522,
+      "step": 3073
+    },
+    {
+      "epoch": 1.4022123389211996,
+      "grad_norm": 0.3175341784954071,
+      "learning_rate": 1.0824118130876283e-06,
+      "loss": 0.1532,
+      "step": 3074
+    },
+    {
+      "epoch": 1.4026684912760863,
+      "grad_norm": 0.313238263130188,
+      "learning_rate": 1.080890734095327e-06,
+      "loss": 0.1443,
+      "step": 3075
+    },
+    {
+      "epoch": 1.4031246436309728,
+      "grad_norm": 0.32441985607147217,
+      "learning_rate": 1.0793704297949872e-06,
+      "loss": 0.15,
+      "step": 3076
+    },
+    {
+      "epoch": 1.4035807959858593,
+      "grad_norm": 0.31845900416374207,
+      "learning_rate": 1.0778509010165434e-06,
+      "loss": 0.1534,
+      "step": 3077
+    },
+    {
+      "epoch": 1.4040369483407458,
+      "grad_norm": 0.3017372190952301,
+      "learning_rate": 1.0763321485895056e-06,
+      "loss": 0.1481,
+      "step": 3078
+    },
+    {
+      "epoch": 1.4044931006956323,
+      "grad_norm": 0.3080170154571533,
+      "learning_rate": 1.074814173342963e-06,
+      "loss": 0.1547,
+      "step": 3079
+    },
+    {
+      "epoch": 1.404949253050519,
+      "grad_norm": 0.30554628372192383,
+      "learning_rate": 1.0732969761055774e-06,
+      "loss": 0.1535,
+      "step": 3080
+    },
+    {
+      "epoch": 1.4054054054054055,
+      "grad_norm": 0.3063387870788574,
+      "learning_rate": 1.0717805577055862e-06,
+      "loss": 0.1549,
+      "step": 3081
+    },
+    {
+      "epoch": 1.405861557760292,
+      "grad_norm": 0.3212406039237976,
+      "learning_rate": 1.070264918970804e-06,
+      "loss": 0.1567,
+      "step": 3082
+    },
+    {
+      "epoch": 1.4063177101151785,
+      "grad_norm": 0.3066968023777008,
+      "learning_rate": 1.0687500607286177e-06,
+      "loss": 0.1493,
+      "step": 3083
+    },
+    {
+      "epoch": 1.406773862470065,
+      "grad_norm": 0.3106542229652405,
+      "learning_rate": 1.0672359838059884e-06,
+      "loss": 0.1547,
+      "step": 3084
+    },
+    {
+      "epoch": 1.4072300148249515,
+      "grad_norm": 0.30837446451187134,
+      "learning_rate": 1.0657226890294512e-06,
+      "loss": 0.1514,
+      "step": 3085
+    },
+    {
+      "epoch": 1.407686167179838,
+      "grad_norm": 0.32347390055656433,
+      "learning_rate": 1.0642101772251126e-06,
+      "loss": 0.1643,
+      "step": 3086
+    },
+    {
+      "epoch": 1.4081423195347247,
+      "grad_norm": 0.3266824781894684,
+      "learning_rate": 1.0626984492186552e-06,
+      "loss": 0.1544,
+      "step": 3087
+    },
+    {
+      "epoch": 1.4085984718896112,
+      "grad_norm": 0.31205129623413086,
+      "learning_rate": 1.0611875058353301e-06,
+      "loss": 0.1493,
+      "step": 3088
+    },
+    {
+      "epoch": 1.4090546242444977,
+      "grad_norm": 0.3109511435031891,
+      "learning_rate": 1.0596773478999622e-06,
+      "loss": 0.1511,
+      "step": 3089
+    },
+    {
+      "epoch": 1.4095107765993842,
+      "grad_norm": 0.30402329564094543,
+      "learning_rate": 1.0581679762369468e-06,
+      "loss": 0.1503,
+      "step": 3090
+    },
+    {
+      "epoch": 1.4099669289542707,
+      "grad_norm": 0.31207895278930664,
+      "learning_rate": 1.0566593916702497e-06,
+      "loss": 0.1537,
+      "step": 3091
+    },
+    {
+      "epoch": 1.4104230813091574,
+      "grad_norm": 0.298210471868515,
+      "learning_rate": 1.055151595023407e-06,
+      "loss": 0.1457,
+      "step": 3092
+    },
+    {
+      "epoch": 1.4108792336640437,
+      "grad_norm": 0.3115929365158081,
+      "learning_rate": 1.0536445871195267e-06,
+      "loss": 0.1532,
+      "step": 3093
+    },
+    {
+      "epoch": 1.4113353860189304,
+      "grad_norm": 0.3212464451789856,
+      "learning_rate": 1.0521383687812828e-06,
+      "loss": 0.1551,
+      "step": 3094
+    },
+    {
+      "epoch": 1.4117915383738169,
+      "grad_norm": 0.31525203585624695,
+      "learning_rate": 1.0506329408309217e-06,
+      "loss": 0.1466,
+      "step": 3095
+    },
+    {
+      "epoch": 1.4122476907287034,
+      "grad_norm": 0.3074638843536377,
+      "learning_rate": 1.0491283040902567e-06,
+      "loss": 0.1513,
+      "step": 3096
+    },
+    {
+      "epoch": 1.4127038430835899,
+      "grad_norm": 0.3096350133419037,
+      "learning_rate": 1.0476244593806683e-06,
+      "loss": 0.153,
+      "step": 3097
+    },
+    {
+      "epoch": 1.4131599954384764,
+      "grad_norm": 0.3063823878765106,
+      "learning_rate": 1.046121407523106e-06,
+      "loss": 0.1571,
+      "step": 3098
+    },
+    {
+      "epoch": 1.413616147793363,
+      "grad_norm": 0.3134966790676117,
+      "learning_rate": 1.044619149338086e-06,
+      "loss": 0.1614,
+      "step": 3099
+    },
+    {
+      "epoch": 1.4140723001482496,
+      "grad_norm": 0.29941999912261963,
+      "learning_rate": 1.0431176856456904e-06,
+      "loss": 0.1475,
+      "step": 3100
+    },
+    {
+      "epoch": 1.414528452503136,
+      "grad_norm": 0.345453143119812,
+      "learning_rate": 1.04161701726557e-06,
+      "loss": 0.1587,
+      "step": 3101
+    },
+    {
+      "epoch": 1.4149846048580226,
+      "grad_norm": 0.31012415885925293,
+      "learning_rate": 1.0401171450169382e-06,
+      "loss": 0.1501,
+      "step": 3102
+    },
+    {
+      "epoch": 1.415440757212909,
+      "grad_norm": 0.3076293468475342,
+      "learning_rate": 1.0386180697185773e-06,
+      "loss": 0.1521,
+      "step": 3103
+    },
+    {
+      "epoch": 1.4158969095677956,
+      "grad_norm": 0.30451568961143494,
+      "learning_rate": 1.0371197921888324e-06,
+      "loss": 0.1488,
+      "step": 3104
+    },
+    {
+      "epoch": 1.416353061922682,
+      "grad_norm": 0.30719640851020813,
+      "learning_rate": 1.0356223132456123e-06,
+      "loss": 0.1472,
+      "step": 3105
+    },
+    {
+      "epoch": 1.4168092142775688,
+      "grad_norm": 0.2984181344509125,
+      "learning_rate": 1.0341256337063917e-06,
+      "loss": 0.1439,
+      "step": 3106
+    },
+    {
+      "epoch": 1.4172653666324553,
+      "grad_norm": 0.3150872588157654,
+      "learning_rate": 1.0326297543882083e-06,
+      "loss": 0.1547,
+      "step": 3107
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.3073064088821411,
+      "learning_rate": 1.0311346761076619e-06,
+      "loss": 0.1574,
+      "step": 3108
+    },
+    {
+      "epoch": 1.4181776713422283,
+      "grad_norm": 0.3216606676578522,
+      "learning_rate": 1.029640399680917e-06,
+      "loss": 0.1571,
+      "step": 3109
+    },
+    {
+      "epoch": 1.4186338236971148,
+      "grad_norm": 0.3034321069717407,
+      "learning_rate": 1.0281469259236992e-06,
+      "loss": 0.1486,
+      "step": 3110
+    },
+    {
+      "epoch": 1.4190899760520015,
+      "grad_norm": 0.30615031719207764,
+      "learning_rate": 1.0266542556512962e-06,
+      "loss": 0.1449,
+      "step": 3111
+    },
+    {
+      "epoch": 1.4195461284068878,
+      "grad_norm": 0.3222392201423645,
+      "learning_rate": 1.0251623896785553e-06,
+      "loss": 0.1607,
+      "step": 3112
+    },
+    {
+      "epoch": 1.4200022807617745,
+      "grad_norm": 0.32082414627075195,
+      "learning_rate": 1.0236713288198887e-06,
+      "loss": 0.1585,
+      "step": 3113
+    },
+    {
+      "epoch": 1.420458433116661,
+      "grad_norm": 0.31318649649620056,
+      "learning_rate": 1.0221810738892658e-06,
+      "loss": 0.1545,
+      "step": 3114
+    },
+    {
+      "epoch": 1.4209145854715475,
+      "grad_norm": 0.32253462076187134,
+      "learning_rate": 1.0206916257002168e-06,
+      "loss": 0.1595,
+      "step": 3115
+    },
+    {
+      "epoch": 1.421370737826434,
+      "grad_norm": 0.3235491216182709,
+      "learning_rate": 1.0192029850658306e-06,
+      "loss": 0.1566,
+      "step": 3116
+    },
+    {
+      "epoch": 1.4218268901813205,
+      "grad_norm": 0.32143735885620117,
+      "learning_rate": 1.0177151527987586e-06,
+      "loss": 0.1607,
+      "step": 3117
+    },
+    {
+      "epoch": 1.4222830425362072,
+      "grad_norm": 0.3123950958251953,
+      "learning_rate": 1.0162281297112076e-06,
+      "loss": 0.1531,
+      "step": 3118
+    },
+    {
+      "epoch": 1.4227391948910937,
+      "grad_norm": 0.3127868175506592,
+      "learning_rate": 1.0147419166149436e-06,
+      "loss": 0.1488,
+      "step": 3119
+    },
+    {
+      "epoch": 1.4231953472459802,
+      "grad_norm": 0.3116130232810974,
+      "learning_rate": 1.0132565143212907e-06,
+      "loss": 0.1579,
+      "step": 3120
+    },
+    {
+      "epoch": 1.4236514996008667,
+      "grad_norm": 0.3248913586139679,
+      "learning_rate": 1.0117719236411301e-06,
+      "loss": 0.1477,
+      "step": 3121
+    },
+    {
+      "epoch": 1.4241076519557532,
+      "grad_norm": 0.3187323212623596,
+      "learning_rate": 1.0102881453848996e-06,
+      "loss": 0.161,
+      "step": 3122
+    },
+    {
+      "epoch": 1.42456380431064,
+      "grad_norm": 0.31156131625175476,
+      "learning_rate": 1.008805180362595e-06,
+      "loss": 0.1633,
+      "step": 3123
+    },
+    {
+      "epoch": 1.4250199566655262,
+      "grad_norm": 0.30793336033821106,
+      "learning_rate": 1.007323029383766e-06,
+      "loss": 0.1501,
+      "step": 3124
+    },
+    {
+      "epoch": 1.4254761090204129,
+      "grad_norm": 0.2975230813026428,
+      "learning_rate": 1.0058416932575204e-06,
+      "loss": 0.1469,
+      "step": 3125
+    },
+    {
+      "epoch": 1.4259322613752994,
+      "grad_norm": 0.32109805941581726,
+      "learning_rate": 1.0043611727925196e-06,
+      "loss": 0.1572,
+      "step": 3126
+    },
+    {
+      "epoch": 1.4263884137301859,
+      "grad_norm": 0.30318859219551086,
+      "learning_rate": 1.002881468796979e-06,
+      "loss": 0.1507,
+      "step": 3127
+    },
+    {
+      "epoch": 1.4268445660850724,
+      "grad_norm": 0.33701032400131226,
+      "learning_rate": 1.00140258207867e-06,
+      "loss": 0.1595,
+      "step": 3128
+    },
+    {
+      "epoch": 1.4273007184399589,
+      "grad_norm": 0.2992219924926758,
+      "learning_rate": 9.999245134449164e-07,
+      "loss": 0.1461,
+      "step": 3129
+    },
+    {
+      "epoch": 1.4277568707948456,
+      "grad_norm": 0.30897483229637146,
+      "learning_rate": 9.984472637025957e-07,
+      "loss": 0.1526,
+      "step": 3130
+    },
+    {
+      "epoch": 1.428213023149732,
+      "grad_norm": 0.3063428997993469,
+      "learning_rate": 9.969708336581405e-07,
+      "loss": 0.1565,
+      "step": 3131
+    },
+    {
+      "epoch": 1.4286691755046186,
+      "grad_norm": 0.2905179560184479,
+      "learning_rate": 9.954952241175325e-07,
+      "loss": 0.1433,
+      "step": 3132
+    },
+    {
+      "epoch": 1.429125327859505,
+      "grad_norm": 0.30064401030540466,
+      "learning_rate": 9.940204358863084e-07,
+      "loss": 0.1484,
+      "step": 3133
+    },
+    {
+      "epoch": 1.4295814802143916,
+      "grad_norm": 0.2968345880508423,
+      "learning_rate": 9.92546469769555e-07,
+      "loss": 0.1526,
+      "step": 3134
+    },
+    {
+      "epoch": 1.430037632569278,
+      "grad_norm": 0.32591161131858826,
+      "learning_rate": 9.910733265719104e-07,
+      "loss": 0.159,
+      "step": 3135
+    },
+    {
+      "epoch": 1.4304937849241646,
+      "grad_norm": 0.3167390823364258,
+      "learning_rate": 9.896010070975638e-07,
+      "loss": 0.1499,
+      "step": 3136
+    },
+    {
+      "epoch": 1.4309499372790513,
+      "grad_norm": 0.3145119249820709,
+      "learning_rate": 9.881295121502544e-07,
+      "loss": 0.1557,
+      "step": 3137
+    },
+    {
+      "epoch": 1.4314060896339378,
+      "grad_norm": 0.31079769134521484,
+      "learning_rate": 9.866588425332708e-07,
+      "loss": 0.1513,
+      "step": 3138
+    },
+    {
+      "epoch": 1.4318622419888243,
+      "grad_norm": 0.30512410402297974,
+      "learning_rate": 9.851889990494535e-07,
+      "loss": 0.1494,
+      "step": 3139
+    },
+    {
+      "epoch": 1.4323183943437108,
+      "grad_norm": 0.3090335726737976,
+      "learning_rate": 9.837199825011895e-07,
+      "loss": 0.1513,
+      "step": 3140
+    },
+    {
+      "epoch": 1.4327745466985973,
+      "grad_norm": 0.3088358938694,
+      "learning_rate": 9.82251793690415e-07,
+      "loss": 0.1535,
+      "step": 3141
+    },
+    {
+      "epoch": 1.433230699053484,
+      "grad_norm": 0.3097587525844574,
+      "learning_rate": 9.807844334186142e-07,
+      "loss": 0.1561,
+      "step": 3142
+    },
+    {
+      "epoch": 1.4336868514083703,
+      "grad_norm": 0.33403947949409485,
+      "learning_rate": 9.79317902486819e-07,
+      "loss": 0.1645,
+      "step": 3143
+    },
+    {
+      "epoch": 1.434143003763257,
+      "grad_norm": 0.303890198469162,
+      "learning_rate": 9.778522016956105e-07,
+      "loss": 0.1457,
+      "step": 3144
+    },
+    {
+      "epoch": 1.4345991561181435,
+      "grad_norm": 0.3050207793712616,
+      "learning_rate": 9.76387331845114e-07,
+      "loss": 0.154,
+      "step": 3145
+    },
+    {
+      "epoch": 1.43505530847303,
+      "grad_norm": 0.3216145634651184,
+      "learning_rate": 9.74923293735001e-07,
+      "loss": 0.1597,
+      "step": 3146
+    },
+    {
+      "epoch": 1.4355114608279165,
+      "grad_norm": 0.30170512199401855,
+      "learning_rate": 9.734600881644922e-07,
+      "loss": 0.1476,
+      "step": 3147
+    },
+    {
+      "epoch": 1.435967613182803,
+      "grad_norm": 0.3042486011981964,
+      "learning_rate": 9.71997715932351e-07,
+      "loss": 0.1553,
+      "step": 3148
+    },
+    {
+      "epoch": 1.4364237655376897,
+      "grad_norm": 0.3273428678512573,
+      "learning_rate": 9.705361778368862e-07,
+      "loss": 0.1565,
+      "step": 3149
+    },
+    {
+      "epoch": 1.4368799178925762,
+      "grad_norm": 0.34081923961639404,
+      "learning_rate": 9.690754746759523e-07,
+      "loss": 0.1617,
+      "step": 3150
+    },
+    {
+      "epoch": 1.4373360702474627,
+      "grad_norm": 0.3350405693054199,
+      "learning_rate": 9.676156072469468e-07,
+      "loss": 0.1582,
+      "step": 3151
+    },
+    {
+      "epoch": 1.4377922226023492,
+      "grad_norm": 0.2830207645893097,
+      "learning_rate": 9.661565763468106e-07,
+      "loss": 0.1436,
+      "step": 3152
+    },
+    {
+      "epoch": 1.4382483749572357,
+      "grad_norm": 0.3127972483634949,
+      "learning_rate": 9.64698382772031e-07,
+      "loss": 0.1571,
+      "step": 3153
+    },
+    {
+      "epoch": 1.4387045273121222,
+      "grad_norm": 0.30612418055534363,
+      "learning_rate": 9.632410273186341e-07,
+      "loss": 0.1491,
+      "step": 3154
+    },
+    {
+      "epoch": 1.4391606796670087,
+      "grad_norm": 0.2978619933128357,
+      "learning_rate": 9.617845107821923e-07,
+      "loss": 0.1508,
+      "step": 3155
+    },
+    {
+      "epoch": 1.4396168320218954,
+      "grad_norm": 0.3209926187992096,
+      "learning_rate": 9.603288339578177e-07,
+      "loss": 0.1515,
+      "step": 3156
+    },
+    {
+      "epoch": 1.4400729843767819,
+      "grad_norm": 0.3027195632457733,
+      "learning_rate": 9.58873997640164e-07,
+      "loss": 0.1495,
+      "step": 3157
+    },
+    {
+      "epoch": 1.4405291367316684,
+      "grad_norm": 0.29430443048477173,
+      "learning_rate": 9.574200026234267e-07,
+      "loss": 0.1409,
+      "step": 3158
+    },
+    {
+      "epoch": 1.4409852890865549,
+      "grad_norm": 0.2968393564224243,
+      "learning_rate": 9.55966849701342e-07,
+      "loss": 0.1502,
+      "step": 3159
+    },
+    {
+      "epoch": 1.4414414414414414,
+      "grad_norm": 0.30550527572631836,
+      "learning_rate": 9.54514539667185e-07,
+      "loss": 0.1517,
+      "step": 3160
+    },
+    {
+      "epoch": 1.441897593796328,
+      "grad_norm": 0.29717135429382324,
+      "learning_rate": 9.530630733137744e-07,
+      "loss": 0.1489,
+      "step": 3161
+    },
+    {
+      "epoch": 1.4423537461512146,
+      "grad_norm": 0.32080647349357605,
+      "learning_rate": 9.516124514334645e-07,
+      "loss": 0.1574,
+      "step": 3162
+    },
+    {
+      "epoch": 1.442809898506101,
+      "grad_norm": 0.30421873927116394,
+      "learning_rate": 9.501626748181495e-07,
+      "loss": 0.1526,
+      "step": 3163
+    },
+    {
+      "epoch": 1.4432660508609876,
+      "grad_norm": 0.3065115213394165,
+      "learning_rate": 9.487137442592642e-07,
+      "loss": 0.1498,
+      "step": 3164
+    },
+    {
+      "epoch": 1.443722203215874,
+      "grad_norm": 0.30278193950653076,
+      "learning_rate": 9.472656605477789e-07,
+      "loss": 0.1483,
+      "step": 3165
+    },
+    {
+      "epoch": 1.4441783555707606,
+      "grad_norm": 0.3078298568725586,
+      "learning_rate": 9.458184244742033e-07,
+      "loss": 0.1497,
+      "step": 3166
+    },
+    {
+      "epoch": 1.444634507925647,
+      "grad_norm": 0.3090832829475403,
+      "learning_rate": 9.443720368285833e-07,
+      "loss": 0.1495,
+      "step": 3167
+    },
+    {
+      "epoch": 1.4450906602805338,
+      "grad_norm": 0.3104894161224365,
+      "learning_rate": 9.429264984005015e-07,
+      "loss": 0.1536,
+      "step": 3168
+    },
+    {
+      "epoch": 1.4455468126354203,
+      "grad_norm": 0.32480907440185547,
+      "learning_rate": 9.414818099790793e-07,
+      "loss": 0.1505,
+      "step": 3169
+    },
+    {
+      "epoch": 1.4460029649903068,
+      "grad_norm": 0.31823861598968506,
+      "learning_rate": 9.400379723529712e-07,
+      "loss": 0.1653,
+      "step": 3170
+    },
+    {
+      "epoch": 1.4464591173451933,
+      "grad_norm": 0.31258171796798706,
+      "learning_rate": 9.385949863103686e-07,
+      "loss": 0.1494,
+      "step": 3171
+    },
+    {
+      "epoch": 1.4469152697000798,
+      "grad_norm": 0.3286343216896057,
+      "learning_rate": 9.371528526389978e-07,
+      "loss": 0.1527,
+      "step": 3172
+    },
+    {
+      "epoch": 1.4473714220549665,
+      "grad_norm": 0.2966218888759613,
+      "learning_rate": 9.357115721261189e-07,
+      "loss": 0.1468,
+      "step": 3173
+    },
+    {
+      "epoch": 1.4478275744098528,
+      "grad_norm": 0.329210489988327,
+      "learning_rate": 9.342711455585288e-07,
+      "loss": 0.1622,
+      "step": 3174
+    },
+    {
+      "epoch": 1.4482837267647395,
+      "grad_norm": 0.30394673347473145,
+      "learning_rate": 9.328315737225557e-07,
+      "loss": 0.1512,
+      "step": 3175
+    },
+    {
+      "epoch": 1.448739879119626,
+      "grad_norm": 0.3039047122001648,
+      "learning_rate": 9.313928574040615e-07,
+      "loss": 0.1559,
+      "step": 3176
+    },
+    {
+      "epoch": 1.4491960314745125,
+      "grad_norm": 0.317135751247406,
+      "learning_rate": 9.299549973884431e-07,
+      "loss": 0.1613,
+      "step": 3177
+    },
+    {
+      "epoch": 1.449652183829399,
+      "grad_norm": 0.31129783391952515,
+      "learning_rate": 9.285179944606277e-07,
+      "loss": 0.1499,
+      "step": 3178
+    },
+    {
+      "epoch": 1.4501083361842855,
+      "grad_norm": 0.30169281363487244,
+      "learning_rate": 9.270818494050751e-07,
+      "loss": 0.1456,
+      "step": 3179
+    },
+    {
+      "epoch": 1.4505644885391722,
+      "grad_norm": 0.30908894538879395,
+      "learning_rate": 9.256465630057778e-07,
+      "loss": 0.1542,
+      "step": 3180
+    },
+    {
+      "epoch": 1.4510206408940587,
+      "grad_norm": 0.2985111176967621,
+      "learning_rate": 9.242121360462583e-07,
+      "loss": 0.1408,
+      "step": 3181
+    },
+    {
+      "epoch": 1.4514767932489452,
+      "grad_norm": 0.31691789627075195,
+      "learning_rate": 9.227785693095701e-07,
+      "loss": 0.1492,
+      "step": 3182
+    },
+    {
+      "epoch": 1.4519329456038317,
+      "grad_norm": 0.3289337754249573,
+      "learning_rate": 9.213458635782985e-07,
+      "loss": 0.1631,
+      "step": 3183
+    },
+    {
+      "epoch": 1.4523890979587182,
+      "grad_norm": 0.30601766705513,
+      "learning_rate": 9.199140196345569e-07,
+      "loss": 0.1556,
+      "step": 3184
+    },
+    {
+      "epoch": 1.4528452503136047,
+      "grad_norm": 0.3339848518371582,
+      "learning_rate": 9.184830382599902e-07,
+      "loss": 0.1605,
+      "step": 3185
+    },
+    {
+      "epoch": 1.4533014026684912,
+      "grad_norm": 0.3070511519908905,
+      "learning_rate": 9.170529202357705e-07,
+      "loss": 0.1539,
+      "step": 3186
+    },
+    {
+      "epoch": 1.453757555023378,
+      "grad_norm": 0.3366701304912567,
+      "learning_rate": 9.156236663425999e-07,
+      "loss": 0.1611,
+      "step": 3187
+    },
+    {
+      "epoch": 1.4542137073782644,
+      "grad_norm": 0.3169521391391754,
+      "learning_rate": 9.141952773607082e-07,
+      "loss": 0.1493,
+      "step": 3188
+    },
+    {
+      "epoch": 1.4546698597331509,
+      "grad_norm": 0.30830278992652893,
+      "learning_rate": 9.127677540698529e-07,
+      "loss": 0.1555,
+      "step": 3189
+    },
+    {
+      "epoch": 1.4551260120880374,
+      "grad_norm": 0.3099665939807892,
+      "learning_rate": 9.113410972493184e-07,
+      "loss": 0.1541,
+      "step": 3190
+    },
+    {
+      "epoch": 1.4555821644429239,
+      "grad_norm": 0.29520052671432495,
+      "learning_rate": 9.099153076779188e-07,
+      "loss": 0.1488,
+      "step": 3191
+    },
+    {
+      "epoch": 1.4560383167978106,
+      "grad_norm": 0.3152410686016083,
+      "learning_rate": 9.084903861339916e-07,
+      "loss": 0.1531,
+      "step": 3192
+    },
+    {
+      "epoch": 1.4564944691526969,
+      "grad_norm": 0.3060758411884308,
+      "learning_rate": 9.070663333954022e-07,
+      "loss": 0.1514,
+      "step": 3193
+    },
+    {
+      "epoch": 1.4569506215075836,
+      "grad_norm": 0.32142001390457153,
+      "learning_rate": 9.0564315023954e-07,
+      "loss": 0.1538,
+      "step": 3194
+    },
+    {
+      "epoch": 1.45740677386247,
+      "grad_norm": 0.30876317620277405,
+      "learning_rate": 9.042208374433226e-07,
+      "loss": 0.1611,
+      "step": 3195
+    },
+    {
+      "epoch": 1.4578629262173566,
+      "grad_norm": 0.30111321806907654,
+      "learning_rate": 9.027993957831898e-07,
+      "loss": 0.1512,
+      "step": 3196
+    },
+    {
+      "epoch": 1.458319078572243,
+      "grad_norm": 0.3058169484138489,
+      "learning_rate": 9.013788260351072e-07,
+      "loss": 0.1468,
+      "step": 3197
+    },
+    {
+      "epoch": 1.4587752309271296,
+      "grad_norm": 0.3133831322193146,
+      "learning_rate": 8.999591289745629e-07,
+      "loss": 0.1515,
+      "step": 3198
+    },
+    {
+      "epoch": 1.4592313832820163,
+      "grad_norm": 0.30585533380508423,
+      "learning_rate": 8.985403053765717e-07,
+      "loss": 0.1504,
+      "step": 3199
+    },
+    {
+      "epoch": 1.4596875356369028,
+      "grad_norm": 0.3147274851799011,
+      "learning_rate": 8.971223560156686e-07,
+      "loss": 0.1497,
+      "step": 3200
+    },
+    {
+      "epoch": 1.4601436879917893,
+      "grad_norm": 0.2890297770500183,
+      "learning_rate": 8.957052816659123e-07,
+      "loss": 0.1495,
+      "step": 3201
+    },
+    {
+      "epoch": 1.4605998403466758,
+      "grad_norm": 0.29727575182914734,
+      "learning_rate": 8.942890831008838e-07,
+      "loss": 0.145,
+      "step": 3202
+    },
+    {
+      "epoch": 1.4610559927015623,
+      "grad_norm": 0.3083558678627014,
+      "learning_rate": 8.928737610936864e-07,
+      "loss": 0.1491,
+      "step": 3203
+    },
+    {
+      "epoch": 1.4615121450564488,
+      "grad_norm": 0.3140217661857605,
+      "learning_rate": 8.914593164169438e-07,
+      "loss": 0.1549,
+      "step": 3204
+    },
+    {
+      "epoch": 1.4619682974113353,
+      "grad_norm": 0.3274988532066345,
+      "learning_rate": 8.900457498428027e-07,
+      "loss": 0.148,
+      "step": 3205
+    },
+    {
+      "epoch": 1.462424449766222,
+      "grad_norm": 0.3139221966266632,
+      "learning_rate": 8.886330621429285e-07,
+      "loss": 0.1515,
+      "step": 3206
+    },
+    {
+      "epoch": 1.4628806021211085,
+      "grad_norm": 0.3175349235534668,
+      "learning_rate": 8.872212540885083e-07,
+      "loss": 0.1523,
+      "step": 3207
+    },
+    {
+      "epoch": 1.463336754475995,
+      "grad_norm": 0.30219534039497375,
+      "learning_rate": 8.858103264502482e-07,
+      "loss": 0.1513,
+      "step": 3208
+    },
+    {
+      "epoch": 1.4637929068308815,
+      "grad_norm": 0.3193797767162323,
+      "learning_rate": 8.844002799983736e-07,
+      "loss": 0.1568,
+      "step": 3209
+    },
+    {
+      "epoch": 1.464249059185768,
+      "grad_norm": 0.3105214834213257,
+      "learning_rate": 8.829911155026292e-07,
+      "loss": 0.1526,
+      "step": 3210
+    },
+    {
+      "epoch": 1.4647052115406547,
+      "grad_norm": 0.31633415818214417,
+      "learning_rate": 8.815828337322779e-07,
+      "loss": 0.1539,
+      "step": 3211
+    },
+    {
+      "epoch": 1.4651613638955412,
+      "grad_norm": 0.3036090135574341,
+      "learning_rate": 8.801754354561001e-07,
+      "loss": 0.1493,
+      "step": 3212
+    },
+    {
+      "epoch": 1.4656175162504277,
+      "grad_norm": 0.2986423671245575,
+      "learning_rate": 8.787689214423967e-07,
+      "loss": 0.1518,
+      "step": 3213
+    },
+    {
+      "epoch": 1.4660736686053142,
+      "grad_norm": 0.2999612092971802,
+      "learning_rate": 8.773632924589817e-07,
+      "loss": 0.1525,
+      "step": 3214
+    },
+    {
+      "epoch": 1.4665298209602007,
+      "grad_norm": 0.3044091761112213,
+      "learning_rate": 8.759585492731903e-07,
+      "loss": 0.1492,
+      "step": 3215
+    },
+    {
+      "epoch": 1.4669859733150872,
+      "grad_norm": 0.3045421540737152,
+      "learning_rate": 8.745546926518714e-07,
+      "loss": 0.1503,
+      "step": 3216
+    },
+    {
+      "epoch": 1.4674421256699737,
+      "grad_norm": 0.33883824944496155,
+      "learning_rate": 8.731517233613898e-07,
+      "loss": 0.158,
+      "step": 3217
+    },
+    {
+      "epoch": 1.4678982780248604,
+      "grad_norm": 0.2929799258708954,
+      "learning_rate": 8.717496421676272e-07,
+      "loss": 0.1499,
+      "step": 3218
+    },
+    {
+      "epoch": 1.4683544303797469,
+      "grad_norm": 0.3102925419807434,
+      "learning_rate": 8.703484498359796e-07,
+      "loss": 0.1568,
+      "step": 3219
+    },
+    {
+      "epoch": 1.4688105827346334,
+      "grad_norm": 0.30890464782714844,
+      "learning_rate": 8.689481471313574e-07,
+      "loss": 0.156,
+      "step": 3220
+    },
+    {
+      "epoch": 1.4692667350895199,
+      "grad_norm": 0.31544947624206543,
+      "learning_rate": 8.67548734818188e-07,
+      "loss": 0.1518,
+      "step": 3221
+    },
+    {
+      "epoch": 1.4697228874444064,
+      "grad_norm": 0.30898991227149963,
+      "learning_rate": 8.661502136604097e-07,
+      "loss": 0.1543,
+      "step": 3222
+    },
+    {
+      "epoch": 1.470179039799293,
+      "grad_norm": 0.31116682291030884,
+      "learning_rate": 8.647525844214752e-07,
+      "loss": 0.1481,
+      "step": 3223
+    },
+    {
+      "epoch": 1.4706351921541794,
+      "grad_norm": 0.30223411321640015,
+      "learning_rate": 8.633558478643503e-07,
+      "loss": 0.1465,
+      "step": 3224
+    },
+    {
+      "epoch": 1.471091344509066,
+      "grad_norm": 0.309527188539505,
+      "learning_rate": 8.619600047515148e-07,
+      "loss": 0.1524,
+      "step": 3225
+    },
+    {
+      "epoch": 1.4715474968639526,
+      "grad_norm": 0.29924291372299194,
+      "learning_rate": 8.605650558449588e-07,
+      "loss": 0.144,
+      "step": 3226
+    },
+    {
+      "epoch": 1.472003649218839,
+      "grad_norm": 0.32415664196014404,
+      "learning_rate": 8.591710019061855e-07,
+      "loss": 0.1575,
+      "step": 3227
+    },
+    {
+      "epoch": 1.4724598015737256,
+      "grad_norm": 0.32692041993141174,
+      "learning_rate": 8.577778436962081e-07,
+      "loss": 0.1597,
+      "step": 3228
+    },
+    {
+      "epoch": 1.472915953928612,
+      "grad_norm": 0.31633323431015015,
+      "learning_rate": 8.563855819755538e-07,
+      "loss": 0.1546,
+      "step": 3229
+    },
+    {
+      "epoch": 1.4733721062834988,
+      "grad_norm": 0.311994343996048,
+      "learning_rate": 8.549942175042572e-07,
+      "loss": 0.1472,
+      "step": 3230
+    },
+    {
+      "epoch": 1.4738282586383853,
+      "grad_norm": 0.31836631894111633,
+      "learning_rate": 8.536037510418646e-07,
+      "loss": 0.1488,
+      "step": 3231
+    },
+    {
+      "epoch": 1.4742844109932718,
+      "grad_norm": 0.3100402355194092,
+      "learning_rate": 8.522141833474318e-07,
+      "loss": 0.1524,
+      "step": 3232
+    },
+    {
+      "epoch": 1.4747405633481583,
+      "grad_norm": 0.2894825339317322,
+      "learning_rate": 8.508255151795245e-07,
+      "loss": 0.1417,
+      "step": 3233
+    },
+    {
+      "epoch": 1.4751967157030448,
+      "grad_norm": 0.3087346851825714,
+      "learning_rate": 8.494377472962159e-07,
+      "loss": 0.1556,
+      "step": 3234
+    },
+    {
+      "epoch": 1.4756528680579313,
+      "grad_norm": 0.30577149987220764,
+      "learning_rate": 8.480508804550888e-07,
+      "loss": 0.1514,
+      "step": 3235
+    },
+    {
+      "epoch": 1.4761090204128178,
+      "grad_norm": 0.32549619674682617,
+      "learning_rate": 8.466649154132342e-07,
+      "loss": 0.1502,
+      "step": 3236
+    },
+    {
+      "epoch": 1.4765651727677045,
+      "grad_norm": 0.3030546307563782,
+      "learning_rate": 8.452798529272519e-07,
+      "loss": 0.1484,
+      "step": 3237
+    },
+    {
+      "epoch": 1.477021325122591,
+      "grad_norm": 0.30905619263648987,
+      "learning_rate": 8.438956937532466e-07,
+      "loss": 0.154,
+      "step": 3238
+    },
+    {
+      "epoch": 1.4774774774774775,
+      "grad_norm": 0.2988545894622803,
+      "learning_rate": 8.425124386468311e-07,
+      "loss": 0.1525,
+      "step": 3239
+    },
+    {
+      "epoch": 1.477933629832364,
+      "grad_norm": 0.30539819598197937,
+      "learning_rate": 8.411300883631246e-07,
+      "loss": 0.1518,
+      "step": 3240
+    },
+    {
+      "epoch": 1.4783897821872505,
+      "grad_norm": 0.2963528335094452,
+      "learning_rate": 8.397486436567523e-07,
+      "loss": 0.148,
+      "step": 3241
+    },
+    {
+      "epoch": 1.4788459345421372,
+      "grad_norm": 0.29646122455596924,
+      "learning_rate": 8.383681052818443e-07,
+      "loss": 0.1457,
+      "step": 3242
+    },
+    {
+      "epoch": 1.4793020868970235,
+      "grad_norm": 0.3156827390193939,
+      "learning_rate": 8.369884739920381e-07,
+      "loss": 0.1492,
+      "step": 3243
+    },
+    {
+      "epoch": 1.4797582392519102,
+      "grad_norm": 0.337334007024765,
+      "learning_rate": 8.356097505404742e-07,
+      "loss": 0.1599,
+      "step": 3244
+    },
+    {
+      "epoch": 1.4802143916067967,
+      "grad_norm": 0.3082313537597656,
+      "learning_rate": 8.342319356797968e-07,
+      "loss": 0.1582,
+      "step": 3245
+    },
+    {
+      "epoch": 1.4806705439616832,
+      "grad_norm": 0.3213750422000885,
+      "learning_rate": 8.328550301621571e-07,
+      "loss": 0.1589,
+      "step": 3246
+    },
+    {
+      "epoch": 1.4811266963165697,
+      "grad_norm": 0.3032473027706146,
+      "learning_rate": 8.314790347392071e-07,
+      "loss": 0.1567,
+      "step": 3247
+    },
+    {
+      "epoch": 1.4815828486714562,
+      "grad_norm": 0.32334375381469727,
+      "learning_rate": 8.301039501621028e-07,
+      "loss": 0.1515,
+      "step": 3248
+    },
+    {
+      "epoch": 1.482039001026343,
+      "grad_norm": 0.30944037437438965,
+      "learning_rate": 8.287297771815034e-07,
+      "loss": 0.1481,
+      "step": 3249
+    },
+    {
+      "epoch": 1.4824951533812294,
+      "grad_norm": 0.3286755084991455,
+      "learning_rate": 8.273565165475689e-07,
+      "loss": 0.1622,
+      "step": 3250
+    },
+    {
+      "epoch": 1.4829513057361159,
+      "grad_norm": 0.3079632520675659,
+      "learning_rate": 8.259841690099646e-07,
+      "loss": 0.1483,
+      "step": 3251
+    },
+    {
+      "epoch": 1.4834074580910024,
+      "grad_norm": 0.29280510544776917,
+      "learning_rate": 8.24612735317854e-07,
+      "loss": 0.1407,
+      "step": 3252
+    },
+    {
+      "epoch": 1.4838636104458889,
+      "grad_norm": 0.31254562735557556,
+      "learning_rate": 8.23242216219903e-07,
+      "loss": 0.1575,
+      "step": 3253
+    },
+    {
+      "epoch": 1.4843197628007754,
+      "grad_norm": 0.30869993567466736,
+      "learning_rate": 8.218726124642784e-07,
+      "loss": 0.1475,
+      "step": 3254
+    },
+    {
+      "epoch": 1.4847759151556619,
+      "grad_norm": 0.3260568678379059,
+      "learning_rate": 8.205039247986463e-07,
+      "loss": 0.1496,
+      "step": 3255
+    },
+    {
+      "epoch": 1.4852320675105486,
+      "grad_norm": 0.3133602440357208,
+      "learning_rate": 8.191361539701748e-07,
+      "loss": 0.1584,
+      "step": 3256
+    },
+    {
+      "epoch": 1.485688219865435,
+      "grad_norm": 0.31250137090682983,
+      "learning_rate": 8.177693007255297e-07,
+      "loss": 0.154,
+      "step": 3257
+    },
+    {
+      "epoch": 1.4861443722203216,
+      "grad_norm": 0.3081834614276886,
+      "learning_rate": 8.164033658108753e-07,
+      "loss": 0.1516,
+      "step": 3258
+    },
+    {
+      "epoch": 1.486600524575208,
+      "grad_norm": 0.3257152736186981,
+      "learning_rate": 8.150383499718776e-07,
+      "loss": 0.1629,
+      "step": 3259
+    },
+    {
+      "epoch": 1.4870566769300946,
+      "grad_norm": 0.3028790056705475,
+      "learning_rate": 8.136742539536979e-07,
+      "loss": 0.1457,
+      "step": 3260
+    },
+    {
+      "epoch": 1.4875128292849813,
+      "grad_norm": 0.29866430163383484,
+      "learning_rate": 8.123110785009969e-07,
+      "loss": 0.1533,
+      "step": 3261
+    },
+    {
+      "epoch": 1.4879689816398678,
+      "grad_norm": 0.3184605836868286,
+      "learning_rate": 8.109488243579317e-07,
+      "loss": 0.1612,
+      "step": 3262
+    },
+    {
+      "epoch": 1.4884251339947543,
+      "grad_norm": 0.30884847044944763,
+      "learning_rate": 8.095874922681571e-07,
+      "loss": 0.154,
+      "step": 3263
+    },
+    {
+      "epoch": 1.4888812863496408,
+      "grad_norm": 0.3255379796028137,
+      "learning_rate": 8.082270829748245e-07,
+      "loss": 0.1603,
+      "step": 3264
+    },
+    {
+      "epoch": 1.4893374387045273,
+      "grad_norm": 0.3141554296016693,
+      "learning_rate": 8.068675972205814e-07,
+      "loss": 0.152,
+      "step": 3265
+    },
+    {
+      "epoch": 1.4897935910594138,
+      "grad_norm": 0.3216108977794647,
+      "learning_rate": 8.055090357475712e-07,
+      "loss": 0.1559,
+      "step": 3266
+    },
+    {
+      "epoch": 1.4902497434143003,
+      "grad_norm": 0.31121939420700073,
+      "learning_rate": 8.04151399297434e-07,
+      "loss": 0.1551,
+      "step": 3267
+    },
+    {
+      "epoch": 1.490705895769187,
+      "grad_norm": 0.2986626625061035,
+      "learning_rate": 8.02794688611303e-07,
+      "loss": 0.1484,
+      "step": 3268
+    },
+    {
+      "epoch": 1.4911620481240735,
+      "grad_norm": 0.3181208074092865,
+      "learning_rate": 8.014389044298068e-07,
+      "loss": 0.1615,
+      "step": 3269
+    },
+    {
+      "epoch": 1.49161820047896,
+      "grad_norm": 0.30053120851516724,
+      "learning_rate": 8.00084047493068e-07,
+      "loss": 0.1575,
+      "step": 3270
+    },
+    {
+      "epoch": 1.4920743528338465,
+      "grad_norm": 0.30814942717552185,
+      "learning_rate": 7.987301185407031e-07,
+      "loss": 0.1507,
+      "step": 3271
+    },
+    {
+      "epoch": 1.492530505188733,
+      "grad_norm": 0.30331018567085266,
+      "learning_rate": 7.973771183118222e-07,
+      "loss": 0.1477,
+      "step": 3272
+    },
+    {
+      "epoch": 1.4929866575436197,
+      "grad_norm": 0.31159040331840515,
+      "learning_rate": 7.960250475450282e-07,
+      "loss": 0.1543,
+      "step": 3273
+    },
+    {
+      "epoch": 1.493442809898506,
+      "grad_norm": 0.3249748647212982,
+      "learning_rate": 7.946739069784174e-07,
+      "loss": 0.1595,
+      "step": 3274
+    },
+    {
+      "epoch": 1.4938989622533927,
+      "grad_norm": 0.3200061321258545,
+      "learning_rate": 7.933236973495769e-07,
+      "loss": 0.1496,
+      "step": 3275
+    },
+    {
+      "epoch": 1.4943551146082792,
+      "grad_norm": 0.3016737103462219,
+      "learning_rate": 7.919744193955864e-07,
+      "loss": 0.156,
+      "step": 3276
+    },
+    {
+      "epoch": 1.4948112669631657,
+      "grad_norm": 0.3198837637901306,
+      "learning_rate": 7.906260738530178e-07,
+      "loss": 0.1536,
+      "step": 3277
+    },
+    {
+      "epoch": 1.4952674193180522,
+      "grad_norm": 0.3238000273704529,
+      "learning_rate": 7.892786614579323e-07,
+      "loss": 0.1647,
+      "step": 3278
+    },
+    {
+      "epoch": 1.4957235716729387,
+      "grad_norm": 0.31604671478271484,
+      "learning_rate": 7.87932182945883e-07,
+      "loss": 0.1476,
+      "step": 3279
+    },
+    {
+      "epoch": 1.4961797240278254,
+      "grad_norm": 0.32189393043518066,
+      "learning_rate": 7.865866390519114e-07,
+      "loss": 0.1571,
+      "step": 3280
+    },
+    {
+      "epoch": 1.496635876382712,
+      "grad_norm": 0.30032822489738464,
+      "learning_rate": 7.852420305105521e-07,
+      "loss": 0.1574,
+      "step": 3281
+    },
+    {
+      "epoch": 1.4970920287375984,
+      "grad_norm": 0.3214437961578369,
+      "learning_rate": 7.838983580558262e-07,
+      "loss": 0.1553,
+      "step": 3282
+    },
+    {
+      "epoch": 1.4975481810924849,
+      "grad_norm": 0.3088015019893646,
+      "learning_rate": 7.825556224212446e-07,
+      "loss": 0.1505,
+      "step": 3283
+    },
+    {
+      "epoch": 1.4980043334473714,
+      "grad_norm": 0.3141407072544098,
+      "learning_rate": 7.812138243398071e-07,
+      "loss": 0.1558,
+      "step": 3284
+    },
+    {
+      "epoch": 1.4984604858022579,
+      "grad_norm": 0.31576451659202576,
+      "learning_rate": 7.798729645440014e-07,
+      "loss": 0.1503,
+      "step": 3285
+    },
+    {
+      "epoch": 1.4989166381571444,
+      "grad_norm": 0.3214374780654907,
+      "learning_rate": 7.78533043765802e-07,
+      "loss": 0.1619,
+      "step": 3286
+    },
+    {
+      "epoch": 1.499372790512031,
+      "grad_norm": 0.30440840125083923,
+      "learning_rate": 7.771940627366739e-07,
+      "loss": 0.1537,
+      "step": 3287
+    },
+    {
+      "epoch": 1.4998289428669176,
+      "grad_norm": 0.28711816668510437,
+      "learning_rate": 7.75856022187565e-07,
+      "loss": 0.145,
+      "step": 3288
+    },
+    {
+      "epoch": 1.500285095221804,
+      "grad_norm": 0.30153515934944153,
+      "learning_rate": 7.745189228489137e-07,
+      "loss": 0.1496,
+      "step": 3289
+    },
+    {
+      "epoch": 1.5007412475766906,
+      "grad_norm": 0.31528177857398987,
+      "learning_rate": 7.731827654506419e-07,
+      "loss": 0.1498,
+      "step": 3290
+    },
+    {
+      "epoch": 1.501197399931577,
+      "grad_norm": 0.2987953722476959,
+      "learning_rate": 7.718475507221579e-07,
+      "loss": 0.1512,
+      "step": 3291
+    },
+    {
+      "epoch": 1.5016535522864638,
+      "grad_norm": 0.3184186816215515,
+      "learning_rate": 7.705132793923559e-07,
+      "loss": 0.1579,
+      "step": 3292
+    },
+    {
+      "epoch": 1.50210970464135,
+      "grad_norm": 0.3025065064430237,
+      "learning_rate": 7.691799521896149e-07,
+      "loss": 0.1494,
+      "step": 3293
+    },
+    {
+      "epoch": 1.5025658569962368,
+      "grad_norm": 0.3177792727947235,
+      "learning_rate": 7.678475698417981e-07,
+      "loss": 0.16,
+      "step": 3294
+    },
+    {
+      "epoch": 1.5030220093511233,
+      "grad_norm": 0.308728963136673,
+      "learning_rate": 7.665161330762524e-07,
+      "loss": 0.1555,
+      "step": 3295
+    },
+    {
+      "epoch": 1.5034781617060098,
+      "grad_norm": 0.3133706748485565,
+      "learning_rate": 7.651856426198106e-07,
+      "loss": 0.1517,
+      "step": 3296
+    },
+    {
+      "epoch": 1.5039343140608965,
+      "grad_norm": 0.3139544725418091,
+      "learning_rate": 7.638560991987881e-07,
+      "loss": 0.1505,
+      "step": 3297
+    },
+    {
+      "epoch": 1.5043904664157828,
+      "grad_norm": 0.3034341037273407,
+      "learning_rate": 7.62527503538982e-07,
+      "loss": 0.1534,
+      "step": 3298
+    },
+    {
+      "epoch": 1.5048466187706695,
+      "grad_norm": 0.3203388750553131,
+      "learning_rate": 7.61199856365673e-07,
+      "loss": 0.1577,
+      "step": 3299
+    },
+    {
+      "epoch": 1.5053027711255558,
+      "grad_norm": 0.286771297454834,
+      "learning_rate": 7.59873158403624e-07,
+      "loss": 0.1493,
+      "step": 3300
+    },
+    {
+      "epoch": 1.5057589234804425,
+      "grad_norm": 0.3069736659526825,
+      "learning_rate": 7.585474103770793e-07,
+      "loss": 0.1492,
+      "step": 3301
+    },
+    {
+      "epoch": 1.506215075835329,
+      "grad_norm": 0.3017934262752533,
+      "learning_rate": 7.572226130097651e-07,
+      "loss": 0.1433,
+      "step": 3302
+    },
+    {
+      "epoch": 1.5066712281902155,
+      "grad_norm": 0.305084764957428,
+      "learning_rate": 7.55898767024888e-07,
+      "loss": 0.1468,
+      "step": 3303
+    },
+    {
+      "epoch": 1.5071273805451022,
+      "grad_norm": 0.3239816725254059,
+      "learning_rate": 7.54575873145137e-07,
+      "loss": 0.1539,
+      "step": 3304
+    },
+    {
+      "epoch": 1.5075835328999885,
+      "grad_norm": 0.30694708228111267,
+      "learning_rate": 7.532539320926791e-07,
+      "loss": 0.1471,
+      "step": 3305
+    },
+    {
+      "epoch": 1.5080396852548752,
+      "grad_norm": 0.3359319865703583,
+      "learning_rate": 7.519329445891613e-07,
+      "loss": 0.1606,
+      "step": 3306
+    },
+    {
+      "epoch": 1.5084958376097617,
+      "grad_norm": 0.3332662582397461,
+      "learning_rate": 7.506129113557132e-07,
+      "loss": 0.1634,
+      "step": 3307
+    },
+    {
+      "epoch": 1.5089519899646482,
+      "grad_norm": 0.31325846910476685,
+      "learning_rate": 7.492938331129393e-07,
+      "loss": 0.1583,
+      "step": 3308
+    },
+    {
+      "epoch": 1.5094081423195347,
+      "grad_norm": 0.3116482198238373,
+      "learning_rate": 7.479757105809254e-07,
+      "loss": 0.1498,
+      "step": 3309
+    },
+    {
+      "epoch": 1.5098642946744212,
+      "grad_norm": 0.29472580552101135,
+      "learning_rate": 7.46658544479234e-07,
+      "loss": 0.1523,
+      "step": 3310
+    },
+    {
+      "epoch": 1.510320447029308,
+      "grad_norm": 0.32367345690727234,
+      "learning_rate": 7.453423355269074e-07,
+      "loss": 0.1696,
+      "step": 3311
+    },
+    {
+      "epoch": 1.5107765993841942,
+      "grad_norm": 0.3050374388694763,
+      "learning_rate": 7.440270844424635e-07,
+      "loss": 0.1522,
+      "step": 3312
+    },
+    {
+      "epoch": 1.511232751739081,
+      "grad_norm": 0.32662537693977356,
+      "learning_rate": 7.427127919438984e-07,
+      "loss": 0.1598,
+      "step": 3313
+    },
+    {
+      "epoch": 1.5116889040939674,
+      "grad_norm": 0.29112929105758667,
+      "learning_rate": 7.413994587486844e-07,
+      "loss": 0.1442,
+      "step": 3314
+    },
+    {
+      "epoch": 1.5121450564488539,
+      "grad_norm": 0.3221563696861267,
+      "learning_rate": 7.400870855737701e-07,
+      "loss": 0.1511,
+      "step": 3315
+    },
+    {
+      "epoch": 1.5126012088037406,
+      "grad_norm": 0.28454089164733887,
+      "learning_rate": 7.3877567313558e-07,
+      "loss": 0.147,
+      "step": 3316
+    },
+    {
+      "epoch": 1.5130573611586269,
+      "grad_norm": 0.31670263409614563,
+      "learning_rate": 7.374652221500153e-07,
+      "loss": 0.1596,
+      "step": 3317
+    },
+    {
+      "epoch": 1.5135135135135136,
+      "grad_norm": 0.31325897574424744,
+      "learning_rate": 7.361557333324501e-07,
+      "loss": 0.1581,
+      "step": 3318
+    },
+    {
+      "epoch": 1.5139696658684,
+      "grad_norm": 0.32172390818595886,
+      "learning_rate": 7.348472073977362e-07,
+      "loss": 0.1628,
+      "step": 3319
+    },
+    {
+      "epoch": 1.5144258182232866,
+      "grad_norm": 0.29374465346336365,
+      "learning_rate": 7.335396450601972e-07,
+      "loss": 0.1438,
+      "step": 3320
+    },
+    {
+      "epoch": 1.514881970578173,
+      "grad_norm": 0.3137332797050476,
+      "learning_rate": 7.322330470336314e-07,
+      "loss": 0.1586,
+      "step": 3321
+    },
+    {
+      "epoch": 1.5153381229330596,
+      "grad_norm": 0.31671056151390076,
+      "learning_rate": 7.309274140313111e-07,
+      "loss": 0.155,
+      "step": 3322
+    },
+    {
+      "epoch": 1.5157942752879463,
+      "grad_norm": 0.32312607765197754,
+      "learning_rate": 7.296227467659814e-07,
+      "loss": 0.1627,
+      "step": 3323
+    },
+    {
+      "epoch": 1.5162504276428326,
+      "grad_norm": 0.31486013531684875,
+      "learning_rate": 7.283190459498607e-07,
+      "loss": 0.1496,
+      "step": 3324
+    },
+    {
+      "epoch": 1.5167065799977193,
+      "grad_norm": 0.3037280738353729,
+      "learning_rate": 7.270163122946383e-07,
+      "loss": 0.1566,
+      "step": 3325
+    },
+    {
+      "epoch": 1.5171627323526058,
+      "grad_norm": 0.3155437111854553,
+      "learning_rate": 7.257145465114781e-07,
+      "loss": 0.1625,
+      "step": 3326
+    },
+    {
+      "epoch": 1.5176188847074923,
+      "grad_norm": 0.3133535385131836,
+      "learning_rate": 7.244137493110135e-07,
+      "loss": 0.158,
+      "step": 3327
+    },
+    {
+      "epoch": 1.5180750370623788,
+      "grad_norm": 0.3095282316207886,
+      "learning_rate": 7.231139214033505e-07,
+      "loss": 0.1523,
+      "step": 3328
+    },
+    {
+      "epoch": 1.5185311894172653,
+      "grad_norm": 0.29443180561065674,
+      "learning_rate": 7.218150634980652e-07,
+      "loss": 0.1497,
+      "step": 3329
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.30814483761787415,
+      "learning_rate": 7.20517176304204e-07,
+      "loss": 0.1517,
+      "step": 3330
+    },
+    {
+      "epoch": 1.5194434941270383,
+      "grad_norm": 0.3101062476634979,
+      "learning_rate": 7.19220260530284e-07,
+      "loss": 0.1502,
+      "step": 3331
+    },
+    {
+      "epoch": 1.519899646481925,
+      "grad_norm": 0.3157236576080322,
+      "learning_rate": 7.179243168842911e-07,
+      "loss": 0.1538,
+      "step": 3332
+    },
+    {
+      "epoch": 1.5203557988368115,
+      "grad_norm": 0.30044591426849365,
+      "learning_rate": 7.166293460736812e-07,
+      "loss": 0.1574,
+      "step": 3333
+    },
+    {
+      "epoch": 1.520811951191698,
+      "grad_norm": 0.3099837005138397,
+      "learning_rate": 7.153353488053802e-07,
+      "loss": 0.1553,
+      "step": 3334
+    },
+    {
+      "epoch": 1.5212681035465847,
+      "grad_norm": 0.32365456223487854,
+      "learning_rate": 7.140423257857807e-07,
+      "loss": 0.1546,
+      "step": 3335
+    },
+    {
+      "epoch": 1.521724255901471,
+      "grad_norm": 0.317545622587204,
+      "learning_rate": 7.127502777207437e-07,
+      "loss": 0.1517,
+      "step": 3336
+    },
+    {
+      "epoch": 1.5221804082563577,
+      "grad_norm": 0.31620028614997864,
+      "learning_rate": 7.114592053155984e-07,
+      "loss": 0.1573,
+      "step": 3337
+    },
+    {
+      "epoch": 1.5226365606112442,
+      "grad_norm": 0.3101683557033539,
+      "learning_rate": 7.101691092751425e-07,
+      "loss": 0.1564,
+      "step": 3338
+    },
+    {
+      "epoch": 1.5230927129661307,
+      "grad_norm": 0.3162214159965515,
+      "learning_rate": 7.088799903036389e-07,
+      "loss": 0.1511,
+      "step": 3339
+    },
+    {
+      "epoch": 1.5235488653210172,
+      "grad_norm": 0.2912890315055847,
+      "learning_rate": 7.075918491048172e-07,
+      "loss": 0.1479,
+      "step": 3340
+    },
+    {
+      "epoch": 1.5240050176759037,
+      "grad_norm": 0.3113529682159424,
+      "learning_rate": 7.063046863818757e-07,
+      "loss": 0.1524,
+      "step": 3341
+    },
+    {
+      "epoch": 1.5244611700307904,
+      "grad_norm": 0.3238811492919922,
+      "learning_rate": 7.05018502837475e-07,
+      "loss": 0.1541,
+      "step": 3342
+    },
+    {
+      "epoch": 1.5249173223856767,
+      "grad_norm": 0.30948635935783386,
+      "learning_rate": 7.037332991737438e-07,
+      "loss": 0.1544,
+      "step": 3343
+    },
+    {
+      "epoch": 1.5253734747405634,
+      "grad_norm": 0.309459924697876,
+      "learning_rate": 7.024490760922748e-07,
+      "loss": 0.1581,
+      "step": 3344
+    },
+    {
+      "epoch": 1.52582962709545,
+      "grad_norm": 0.31893423199653625,
+      "learning_rate": 7.011658342941252e-07,
+      "loss": 0.1598,
+      "step": 3345
+    },
+    {
+      "epoch": 1.5262857794503364,
+      "grad_norm": 0.3032120168209076,
+      "learning_rate": 6.998835744798175e-07,
+      "loss": 0.1488,
+      "step": 3346
+    },
+    {
+      "epoch": 1.526741931805223,
+      "grad_norm": 0.3243098855018616,
+      "learning_rate": 6.986022973493361e-07,
+      "loss": 0.1617,
+      "step": 3347
+    },
+    {
+      "epoch": 1.5271980841601094,
+      "grad_norm": 0.31308120489120483,
+      "learning_rate": 6.973220036021313e-07,
+      "loss": 0.152,
+      "step": 3348
+    },
+    {
+      "epoch": 1.527654236514996,
+      "grad_norm": 0.3219483196735382,
+      "learning_rate": 6.960426939371171e-07,
+      "loss": 0.1564,
+      "step": 3349
+    },
+    {
+      "epoch": 1.5281103888698824,
+      "grad_norm": 0.3056938946247101,
+      "learning_rate": 6.947643690526671e-07,
+      "loss": 0.1515,
+      "step": 3350
+    },
+    {
+      "epoch": 1.528566541224769,
+      "grad_norm": 0.3281216323375702,
+      "learning_rate": 6.934870296466195e-07,
+      "loss": 0.1548,
+      "step": 3351
+    },
+    {
+      "epoch": 1.5290226935796556,
+      "grad_norm": 0.3130066394805908,
+      "learning_rate": 6.922106764162742e-07,
+      "loss": 0.1494,
+      "step": 3352
+    },
+    {
+      "epoch": 1.529478845934542,
+      "grad_norm": 0.2870764434337616,
+      "learning_rate": 6.909353100583923e-07,
+      "loss": 0.1504,
+      "step": 3353
+    },
+    {
+      "epoch": 1.5299349982894288,
+      "grad_norm": 0.2951715588569641,
+      "learning_rate": 6.896609312691968e-07,
+      "loss": 0.146,
+      "step": 3354
+    },
+    {
+      "epoch": 1.530391150644315,
+      "grad_norm": 0.31744328141212463,
+      "learning_rate": 6.883875407443702e-07,
+      "loss": 0.152,
+      "step": 3355
+    },
+    {
+      "epoch": 1.5308473029992018,
+      "grad_norm": 0.3135569095611572,
+      "learning_rate": 6.871151391790584e-07,
+      "loss": 0.1545,
+      "step": 3356
+    },
+    {
+      "epoch": 1.5313034553540883,
+      "grad_norm": 0.3057423233985901,
+      "learning_rate": 6.858437272678639e-07,
+      "loss": 0.1489,
+      "step": 3357
+    },
+    {
+      "epoch": 1.5317596077089748,
+      "grad_norm": 0.317620187997818,
+      "learning_rate": 6.845733057048523e-07,
+      "loss": 0.154,
+      "step": 3358
+    },
+    {
+      "epoch": 1.5322157600638613,
+      "grad_norm": 0.3051481544971466,
+      "learning_rate": 6.833038751835461e-07,
+      "loss": 0.1465,
+      "step": 3359
+    },
+    {
+      "epoch": 1.5326719124187478,
+      "grad_norm": 0.3214899003505707,
+      "learning_rate": 6.820354363969276e-07,
+      "loss": 0.1529,
+      "step": 3360
+    },
+    {
+      "epoch": 1.5331280647736345,
+      "grad_norm": 0.30182966589927673,
+      "learning_rate": 6.807679900374378e-07,
+      "loss": 0.1468,
+      "step": 3361
+    },
+    {
+      "epoch": 1.5335842171285208,
+      "grad_norm": 0.29656708240509033,
+      "learning_rate": 6.795015367969757e-07,
+      "loss": 0.1506,
+      "step": 3362
+    },
+    {
+      "epoch": 1.5340403694834075,
+      "grad_norm": 0.29666009545326233,
+      "learning_rate": 6.782360773668975e-07,
+      "loss": 0.1463,
+      "step": 3363
+    },
+    {
+      "epoch": 1.534496521838294,
+      "grad_norm": 0.32493188977241516,
+      "learning_rate": 6.769716124380193e-07,
+      "loss": 0.1608,
+      "step": 3364
+    },
+    {
+      "epoch": 1.5349526741931805,
+      "grad_norm": 0.31728872656822205,
+      "learning_rate": 6.757081427006123e-07,
+      "loss": 0.1511,
+      "step": 3365
+    },
+    {
+      "epoch": 1.5354088265480672,
+      "grad_norm": 0.3071468770503998,
+      "learning_rate": 6.744456688444043e-07,
+      "loss": 0.1531,
+      "step": 3366
+    },
+    {
+      "epoch": 1.5358649789029535,
+      "grad_norm": 0.321716845035553,
+      "learning_rate": 6.731841915585796e-07,
+      "loss": 0.1542,
+      "step": 3367
+    },
+    {
+      "epoch": 1.5363211312578402,
+      "grad_norm": 0.31244996190071106,
+      "learning_rate": 6.7192371153178e-07,
+      "loss": 0.1574,
+      "step": 3368
+    },
+    {
+      "epoch": 1.5367772836127267,
+      "grad_norm": 0.31048959493637085,
+      "learning_rate": 6.706642294521012e-07,
+      "loss": 0.1465,
+      "step": 3369
+    },
+    {
+      "epoch": 1.5372334359676132,
+      "grad_norm": 0.3205074965953827,
+      "learning_rate": 6.694057460070938e-07,
+      "loss": 0.1588,
+      "step": 3370
+    },
+    {
+      "epoch": 1.5376895883224997,
+      "grad_norm": 0.31585344672203064,
+      "learning_rate": 6.68148261883766e-07,
+      "loss": 0.1566,
+      "step": 3371
+    },
+    {
+      "epoch": 1.5381457406773862,
+      "grad_norm": 0.32582157850265503,
+      "learning_rate": 6.668917777685771e-07,
+      "loss": 0.1595,
+      "step": 3372
+    },
+    {
+      "epoch": 1.538601893032273,
+      "grad_norm": 0.32495325803756714,
+      "learning_rate": 6.656362943474423e-07,
+      "loss": 0.1594,
+      "step": 3373
+    },
+    {
+      "epoch": 1.5390580453871592,
+      "grad_norm": 0.31340718269348145,
+      "learning_rate": 6.643818123057305e-07,
+      "loss": 0.1584,
+      "step": 3374
+    },
+    {
+      "epoch": 1.539514197742046,
+      "grad_norm": 0.3246331214904785,
+      "learning_rate": 6.631283323282631e-07,
+      "loss": 0.1478,
+      "step": 3375
+    },
+    {
+      "epoch": 1.5399703500969324,
+      "grad_norm": 0.3190922141075134,
+      "learning_rate": 6.618758550993157e-07,
+      "loss": 0.1584,
+      "step": 3376
+    },
+    {
+      "epoch": 1.540426502451819,
+      "grad_norm": 0.44338318705558777,
+      "learning_rate": 6.606243813026142e-07,
+      "loss": 0.1511,
+      "step": 3377
+    },
+    {
+      "epoch": 1.5408826548067054,
+      "grad_norm": 0.322183758020401,
+      "learning_rate": 6.593739116213396e-07,
+      "loss": 0.1678,
+      "step": 3378
+    },
+    {
+      "epoch": 1.5413388071615919,
+      "grad_norm": 0.3071935474872589,
+      "learning_rate": 6.581244467381243e-07,
+      "loss": 0.159,
+      "step": 3379
+    },
+    {
+      "epoch": 1.5417949595164786,
+      "grad_norm": 0.30973145365715027,
+      "learning_rate": 6.568759873350505e-07,
+      "loss": 0.1513,
+      "step": 3380
+    },
+    {
+      "epoch": 1.5422511118713649,
+      "grad_norm": 0.30505236983299255,
+      "learning_rate": 6.556285340936525e-07,
+      "loss": 0.1541,
+      "step": 3381
+    },
+    {
+      "epoch": 1.5427072642262516,
+      "grad_norm": 0.30744948983192444,
+      "learning_rate": 6.543820876949147e-07,
+      "loss": 0.1519,
+      "step": 3382
+    },
+    {
+      "epoch": 1.543163416581138,
+      "grad_norm": 0.31609633564949036,
+      "learning_rate": 6.531366488192728e-07,
+      "loss": 0.1525,
+      "step": 3383
+    },
+    {
+      "epoch": 1.5436195689360246,
+      "grad_norm": 0.3089766502380371,
+      "learning_rate": 6.518922181466123e-07,
+      "loss": 0.1524,
+      "step": 3384
+    },
+    {
+      "epoch": 1.5440757212909113,
+      "grad_norm": 0.29330557584762573,
+      "learning_rate": 6.506487963562669e-07,
+      "loss": 0.1403,
+      "step": 3385
+    },
+    {
+      "epoch": 1.5445318736457976,
+      "grad_norm": 0.3207091987133026,
+      "learning_rate": 6.494063841270224e-07,
+      "loss": 0.1593,
+      "step": 3386
+    },
+    {
+      "epoch": 1.5449880260006843,
+      "grad_norm": 0.3045348525047302,
+      "learning_rate": 6.481649821371108e-07,
+      "loss": 0.1477,
+      "step": 3387
+    },
+    {
+      "epoch": 1.5454441783555708,
+      "grad_norm": 0.30639615654945374,
+      "learning_rate": 6.469245910642135e-07,
+      "loss": 0.1569,
+      "step": 3388
+    },
+    {
+      "epoch": 1.5459003307104573,
+      "grad_norm": 0.31870657205581665,
+      "learning_rate": 6.456852115854612e-07,
+      "loss": 0.1516,
+      "step": 3389
+    },
+    {
+      "epoch": 1.5463564830653438,
+      "grad_norm": 0.31109416484832764,
+      "learning_rate": 6.444468443774307e-07,
+      "loss": 0.1493,
+      "step": 3390
+    },
+    {
+      "epoch": 1.5468126354202303,
+      "grad_norm": 0.3142516613006592,
+      "learning_rate": 6.432094901161468e-07,
+      "loss": 0.1535,
+      "step": 3391
+    },
+    {
+      "epoch": 1.547268787775117,
+      "grad_norm": 0.31115826964378357,
+      "learning_rate": 6.419731494770814e-07,
+      "loss": 0.1519,
+      "step": 3392
+    },
+    {
+      "epoch": 1.5477249401300033,
+      "grad_norm": 0.32070258259773254,
+      "learning_rate": 6.407378231351527e-07,
+      "loss": 0.1569,
+      "step": 3393
+    },
+    {
+      "epoch": 1.54818109248489,
+      "grad_norm": 0.29688215255737305,
+      "learning_rate": 6.395035117647263e-07,
+      "loss": 0.1446,
+      "step": 3394
+    },
+    {
+      "epoch": 1.5486372448397765,
+      "grad_norm": 0.3047696352005005,
+      "learning_rate": 6.382702160396126e-07,
+      "loss": 0.1491,
+      "step": 3395
+    },
+    {
+      "epoch": 1.549093397194663,
+      "grad_norm": 0.31923699378967285,
+      "learning_rate": 6.370379366330682e-07,
+      "loss": 0.1532,
+      "step": 3396
+    },
+    {
+      "epoch": 1.5495495495495497,
+      "grad_norm": 0.3031322956085205,
+      "learning_rate": 6.358066742177941e-07,
+      "loss": 0.1487,
+      "step": 3397
+    },
+    {
+      "epoch": 1.550005701904436,
+      "grad_norm": 0.2950242757797241,
+      "learning_rate": 6.34576429465936e-07,
+      "loss": 0.1471,
+      "step": 3398
+    },
+    {
+      "epoch": 1.5504618542593227,
+      "grad_norm": 0.3169761598110199,
+      "learning_rate": 6.333472030490865e-07,
+      "loss": 0.1608,
+      "step": 3399
+    },
+    {
+      "epoch": 1.5509180066142092,
+      "grad_norm": 0.3052075207233429,
+      "learning_rate": 6.321189956382795e-07,
+      "loss": 0.1505,
+      "step": 3400
+    },
+    {
+      "epoch": 1.5513741589690957,
+      "grad_norm": 0.32011815905570984,
+      "learning_rate": 6.308918079039927e-07,
+      "loss": 0.1497,
+      "step": 3401
+    },
+    {
+      "epoch": 1.5518303113239822,
+      "grad_norm": 0.296032577753067,
+      "learning_rate": 6.2966564051615e-07,
+      "loss": 0.1498,
+      "step": 3402
+    },
+    {
+      "epoch": 1.5522864636788687,
+      "grad_norm": 0.31267207860946655,
+      "learning_rate": 6.284404941441153e-07,
+      "loss": 0.1557,
+      "step": 3403
+    },
+    {
+      "epoch": 1.5527426160337554,
+      "grad_norm": 0.31805986166000366,
+      "learning_rate": 6.27216369456696e-07,
+      "loss": 0.1549,
+      "step": 3404
+    },
+    {
+      "epoch": 1.5531987683886417,
+      "grad_norm": 0.30880048871040344,
+      "learning_rate": 6.259932671221425e-07,
+      "loss": 0.1464,
+      "step": 3405
+    },
+    {
+      "epoch": 1.5536549207435284,
+      "grad_norm": 0.2941443622112274,
+      "learning_rate": 6.247711878081464e-07,
+      "loss": 0.1444,
+      "step": 3406
+    },
+    {
+      "epoch": 1.554111073098415,
+      "grad_norm": 0.31114882230758667,
+      "learning_rate": 6.235501321818399e-07,
+      "loss": 0.1499,
+      "step": 3407
+    },
+    {
+      "epoch": 1.5545672254533014,
+      "grad_norm": 0.3184615969657898,
+      "learning_rate": 6.223301009097982e-07,
+      "loss": 0.1563,
+      "step": 3408
+    },
+    {
+      "epoch": 1.555023377808188,
+      "grad_norm": 0.31234195828437805,
+      "learning_rate": 6.211110946580379e-07,
+      "loss": 0.1566,
+      "step": 3409
+    },
+    {
+      "epoch": 1.5554795301630744,
+      "grad_norm": 0.2887580096721649,
+      "learning_rate": 6.198931140920131e-07,
+      "loss": 0.136,
+      "step": 3410
+    },
+    {
+      "epoch": 1.555935682517961,
+      "grad_norm": 0.3166784942150116,
+      "learning_rate": 6.186761598766203e-07,
+      "loss": 0.1532,
+      "step": 3411
+    },
+    {
+      "epoch": 1.5563918348728474,
+      "grad_norm": 0.3218485414981842,
+      "learning_rate": 6.174602326761947e-07,
+      "loss": 0.155,
+      "step": 3412
+    },
+    {
+      "epoch": 1.556847987227734,
+      "grad_norm": 0.30574384331703186,
+      "learning_rate": 6.162453331545112e-07,
+      "loss": 0.1528,
+      "step": 3413
+    },
+    {
+      "epoch": 1.5573041395826206,
+      "grad_norm": 0.30976760387420654,
+      "learning_rate": 6.150314619747832e-07,
+      "loss": 0.1551,
+      "step": 3414
+    },
+    {
+      "epoch": 1.557760291937507,
+      "grad_norm": 0.30279844999313354,
+      "learning_rate": 6.138186197996632e-07,
+      "loss": 0.1605,
+      "step": 3415
+    },
+    {
+      "epoch": 1.5582164442923938,
+      "grad_norm": 0.3292537331581116,
+      "learning_rate": 6.126068072912431e-07,
+      "loss": 0.1615,
+      "step": 3416
+    },
+    {
+      "epoch": 1.55867259664728,
+      "grad_norm": 0.3096506893634796,
+      "learning_rate": 6.113960251110504e-07,
+      "loss": 0.1547,
+      "step": 3417
+    },
+    {
+      "epoch": 1.5591287490021668,
+      "grad_norm": 0.3094627857208252,
+      "learning_rate": 6.101862739200515e-07,
+      "loss": 0.1518,
+      "step": 3418
+    },
+    {
+      "epoch": 1.5595849013570533,
+      "grad_norm": 0.2902866303920746,
+      "learning_rate": 6.089775543786495e-07,
+      "loss": 0.1432,
+      "step": 3419
+    },
+    {
+      "epoch": 1.5600410537119398,
+      "grad_norm": 0.3151184916496277,
+      "learning_rate": 6.077698671466852e-07,
+      "loss": 0.1551,
+      "step": 3420
+    },
+    {
+      "epoch": 1.5604972060668263,
+      "grad_norm": 0.32338881492614746,
+      "learning_rate": 6.06563212883435e-07,
+      "loss": 0.1559,
+      "step": 3421
+    },
+    {
+      "epoch": 1.5609533584217128,
+      "grad_norm": 0.32234862446784973,
+      "learning_rate": 6.053575922476118e-07,
+      "loss": 0.1569,
+      "step": 3422
+    },
+    {
+      "epoch": 1.5614095107765995,
+      "grad_norm": 0.29854637384414673,
+      "learning_rate": 6.041530058973633e-07,
+      "loss": 0.1505,
+      "step": 3423
+    },
+    {
+      "epoch": 1.5618656631314858,
+      "grad_norm": 0.29216888546943665,
+      "learning_rate": 6.029494544902742e-07,
+      "loss": 0.1515,
+      "step": 3424
+    },
+    {
+      "epoch": 1.5623218154863725,
+      "grad_norm": 0.3148649036884308,
+      "learning_rate": 6.017469386833635e-07,
+      "loss": 0.1553,
+      "step": 3425
+    },
+    {
+      "epoch": 1.562777967841259,
+      "grad_norm": 0.3224228322505951,
+      "learning_rate": 6.005454591330842e-07,
+      "loss": 0.1505,
+      "step": 3426
+    },
+    {
+      "epoch": 1.5632341201961455,
+      "grad_norm": 0.3333069384098053,
+      "learning_rate": 5.993450164953246e-07,
+      "loss": 0.1603,
+      "step": 3427
+    },
+    {
+      "epoch": 1.563690272551032,
+      "grad_norm": 0.31756389141082764,
+      "learning_rate": 5.981456114254061e-07,
+      "loss": 0.1547,
+      "step": 3428
+    },
+    {
+      "epoch": 1.5641464249059185,
+      "grad_norm": 0.3326030373573303,
+      "learning_rate": 5.969472445780836e-07,
+      "loss": 0.16,
+      "step": 3429
+    },
+    {
+      "epoch": 1.5646025772608052,
+      "grad_norm": 0.3064666986465454,
+      "learning_rate": 5.95749916607547e-07,
+      "loss": 0.1413,
+      "step": 3430
+    },
+    {
+      "epoch": 1.5650587296156915,
+      "grad_norm": 0.3054736256599426,
+      "learning_rate": 5.945536281674169e-07,
+      "loss": 0.1531,
+      "step": 3431
+    },
+    {
+      "epoch": 1.5655148819705782,
+      "grad_norm": 0.32204848527908325,
+      "learning_rate": 5.933583799107482e-07,
+      "loss": 0.1561,
+      "step": 3432
+    },
+    {
+      "epoch": 1.5659710343254647,
+      "grad_norm": 0.3190995752811432,
+      "learning_rate": 5.921641724900265e-07,
+      "loss": 0.1546,
+      "step": 3433
+    },
+    {
+      "epoch": 1.5664271866803512,
+      "grad_norm": 0.3066360056400299,
+      "learning_rate": 5.9097100655717e-07,
+      "loss": 0.1515,
+      "step": 3434
+    },
+    {
+      "epoch": 1.566883339035238,
+      "grad_norm": 0.31656771898269653,
+      "learning_rate": 5.897788827635281e-07,
+      "loss": 0.164,
+      "step": 3435
+    },
+    {
+      "epoch": 1.5673394913901242,
+      "grad_norm": 0.30090463161468506,
+      "learning_rate": 5.885878017598815e-07,
+      "loss": 0.151,
+      "step": 3436
+    },
+    {
+      "epoch": 1.567795643745011,
+      "grad_norm": 0.31196528673171997,
+      "learning_rate": 5.873977641964404e-07,
+      "loss": 0.1504,
+      "step": 3437
+    },
+    {
+      "epoch": 1.5682517960998974,
+      "grad_norm": 0.30652493238449097,
+      "learning_rate": 5.862087707228481e-07,
+      "loss": 0.1493,
+      "step": 3438
+    },
+    {
+      "epoch": 1.568707948454784,
+      "grad_norm": 0.29871073365211487,
+      "learning_rate": 5.850208219881747e-07,
+      "loss": 0.1461,
+      "step": 3439
+    },
+    {
+      "epoch": 1.5691641008096704,
+      "grad_norm": 0.3195111155509949,
+      "learning_rate": 5.838339186409236e-07,
+      "loss": 0.1505,
+      "step": 3440
+    },
+    {
+      "epoch": 1.5696202531645569,
+      "grad_norm": 0.30468082427978516,
+      "learning_rate": 5.826480613290244e-07,
+      "loss": 0.155,
+      "step": 3441
+    },
+    {
+      "epoch": 1.5700764055194436,
+      "grad_norm": 0.3015684187412262,
+      "learning_rate": 5.814632506998369e-07,
+      "loss": 0.1472,
+      "step": 3442
+    },
+    {
+      "epoch": 1.5705325578743299,
+      "grad_norm": 0.2937803566455841,
+      "learning_rate": 5.802794874001492e-07,
+      "loss": 0.151,
+      "step": 3443
+    },
+    {
+      "epoch": 1.5709887102292166,
+      "grad_norm": 0.30181747674942017,
+      "learning_rate": 5.790967720761778e-07,
+      "loss": 0.1503,
+      "step": 3444
+    },
+    {
+      "epoch": 1.571444862584103,
+      "grad_norm": 0.3078451156616211,
+      "learning_rate": 5.779151053735668e-07,
+      "loss": 0.1517,
+      "step": 3445
+    },
+    {
+      "epoch": 1.5719010149389896,
+      "grad_norm": 0.30520886182785034,
+      "learning_rate": 5.767344879373895e-07,
+      "loss": 0.1501,
+      "step": 3446
+    },
+    {
+      "epoch": 1.5723571672938763,
+      "grad_norm": 0.3197169303894043,
+      "learning_rate": 5.755549204121444e-07,
+      "loss": 0.1582,
+      "step": 3447
+    },
+    {
+      "epoch": 1.5728133196487626,
+      "grad_norm": 0.29458707571029663,
+      "learning_rate": 5.743764034417579e-07,
+      "loss": 0.1481,
+      "step": 3448
+    },
+    {
+      "epoch": 1.5732694720036493,
+      "grad_norm": 0.305230051279068,
+      "learning_rate": 5.731989376695813e-07,
+      "loss": 0.1521,
+      "step": 3449
+    },
+    {
+      "epoch": 1.5737256243585358,
+      "grad_norm": 0.30674469470977783,
+      "learning_rate": 5.72022523738395e-07,
+      "loss": 0.1527,
+      "step": 3450
+    },
+    {
+      "epoch": 1.5741817767134223,
+      "grad_norm": 0.32161375880241394,
+      "learning_rate": 5.70847162290403e-07,
+      "loss": 0.152,
+      "step": 3451
+    },
+    {
+      "epoch": 1.5746379290683088,
+      "grad_norm": 0.30655717849731445,
+      "learning_rate": 5.696728539672353e-07,
+      "loss": 0.1557,
+      "step": 3452
+    },
+    {
+      "epoch": 1.5750940814231953,
+      "grad_norm": 0.2979765236377716,
+      "learning_rate": 5.68499599409946e-07,
+      "loss": 0.1526,
+      "step": 3453
+    },
+    {
+      "epoch": 1.575550233778082,
+      "grad_norm": 0.3173069655895233,
+      "learning_rate": 5.673273992590165e-07,
+      "loss": 0.1557,
+      "step": 3454
+    },
+    {
+      "epoch": 1.5760063861329683,
+      "grad_norm": 0.29276156425476074,
+      "learning_rate": 5.661562541543503e-07,
+      "loss": 0.1431,
+      "step": 3455
+    },
+    {
+      "epoch": 1.576462538487855,
+      "grad_norm": 0.3056100904941559,
+      "learning_rate": 5.649861647352758e-07,
+      "loss": 0.1504,
+      "step": 3456
+    },
+    {
+      "epoch": 1.5769186908427415,
+      "grad_norm": 0.3291006088256836,
+      "learning_rate": 5.638171316405449e-07,
+      "loss": 0.1545,
+      "step": 3457
+    },
+    {
+      "epoch": 1.577374843197628,
+      "grad_norm": 0.3001651167869568,
+      "learning_rate": 5.626491555083333e-07,
+      "loss": 0.1501,
+      "step": 3458
+    },
+    {
+      "epoch": 1.5778309955525145,
+      "grad_norm": 0.2978948652744293,
+      "learning_rate": 5.61482236976238e-07,
+      "loss": 0.1447,
+      "step": 3459
+    },
+    {
+      "epoch": 1.578287147907401,
+      "grad_norm": 0.3034391403198242,
+      "learning_rate": 5.603163766812817e-07,
+      "loss": 0.146,
+      "step": 3460
+    },
+    {
+      "epoch": 1.5787433002622877,
+      "grad_norm": 0.3025749921798706,
+      "learning_rate": 5.591515752599064e-07,
+      "loss": 0.1547,
+      "step": 3461
+    },
+    {
+      "epoch": 1.579199452617174,
+      "grad_norm": 0.3021927773952484,
+      "learning_rate": 5.579878333479788e-07,
+      "loss": 0.1514,
+      "step": 3462
+    },
+    {
+      "epoch": 1.5796556049720607,
+      "grad_norm": 0.29340580105781555,
+      "learning_rate": 5.568251515807848e-07,
+      "loss": 0.1419,
+      "step": 3463
+    },
+    {
+      "epoch": 1.5801117573269472,
+      "grad_norm": 0.3142057955265045,
+      "learning_rate": 5.556635305930327e-07,
+      "loss": 0.1582,
+      "step": 3464
+    },
+    {
+      "epoch": 1.5805679096818337,
+      "grad_norm": 0.31142109632492065,
+      "learning_rate": 5.545029710188518e-07,
+      "loss": 0.158,
+      "step": 3465
+    },
+    {
+      "epoch": 1.5810240620367204,
+      "grad_norm": 0.32449841499328613,
+      "learning_rate": 5.533434734917914e-07,
+      "loss": 0.157,
+      "step": 3466
+    },
+    {
+      "epoch": 1.5814802143916067,
+      "grad_norm": 0.3110409677028656,
+      "learning_rate": 5.52185038644821e-07,
+      "loss": 0.1503,
+      "step": 3467
+    },
+    {
+      "epoch": 1.5819363667464934,
+      "grad_norm": 0.317279577255249,
+      "learning_rate": 5.510276671103315e-07,
+      "loss": 0.1565,
+      "step": 3468
+    },
+    {
+      "epoch": 1.58239251910138,
+      "grad_norm": 0.32877445220947266,
+      "learning_rate": 5.498713595201316e-07,
+      "loss": 0.1644,
+      "step": 3469
+    },
+    {
+      "epoch": 1.5828486714562664,
+      "grad_norm": 0.3234955370426178,
+      "learning_rate": 5.487161165054489e-07,
+      "loss": 0.1572,
+      "step": 3470
+    },
+    {
+      "epoch": 1.583304823811153,
+      "grad_norm": 0.29937177896499634,
+      "learning_rate": 5.475619386969325e-07,
+      "loss": 0.149,
+      "step": 3471
+    },
+    {
+      "epoch": 1.5837609761660394,
+      "grad_norm": 0.34193578362464905,
+      "learning_rate": 5.464088267246473e-07,
+      "loss": 0.1537,
+      "step": 3472
+    },
+    {
+      "epoch": 1.584217128520926,
+      "grad_norm": 0.3189118504524231,
+      "learning_rate": 5.452567812180773e-07,
+      "loss": 0.1571,
+      "step": 3473
+    },
+    {
+      "epoch": 1.5846732808758124,
+      "grad_norm": 0.2953558564186096,
+      "learning_rate": 5.441058028061244e-07,
+      "loss": 0.1465,
+      "step": 3474
+    },
+    {
+      "epoch": 1.585129433230699,
+      "grad_norm": 0.2988276481628418,
+      "learning_rate": 5.429558921171072e-07,
+      "loss": 0.1466,
+      "step": 3475
+    },
+    {
+      "epoch": 1.5855855855855856,
+      "grad_norm": 0.2927725911140442,
+      "learning_rate": 5.418070497787634e-07,
+      "loss": 0.1495,
+      "step": 3476
+    },
+    {
+      "epoch": 1.586041737940472,
+      "grad_norm": 0.316119909286499,
+      "learning_rate": 5.406592764182455e-07,
+      "loss": 0.1566,
+      "step": 3477
+    },
+    {
+      "epoch": 1.5864978902953588,
+      "grad_norm": 0.31945639848709106,
+      "learning_rate": 5.395125726621231e-07,
+      "loss": 0.1527,
+      "step": 3478
+    },
+    {
+      "epoch": 1.586954042650245,
+      "grad_norm": 0.31205880641937256,
+      "learning_rate": 5.383669391363824e-07,
+      "loss": 0.1585,
+      "step": 3479
+    },
+    {
+      "epoch": 1.5874101950051318,
+      "grad_norm": 0.3154524862766266,
+      "learning_rate": 5.372223764664236e-07,
+      "loss": 0.1533,
+      "step": 3480
+    },
+    {
+      "epoch": 1.587866347360018,
+      "grad_norm": 0.3169793486595154,
+      "learning_rate": 5.360788852770654e-07,
+      "loss": 0.1572,
+      "step": 3481
+    },
+    {
+      "epoch": 1.5883224997149048,
+      "grad_norm": 0.3148733377456665,
+      "learning_rate": 5.34936466192539e-07,
+      "loss": 0.1629,
+      "step": 3482
+    },
+    {
+      "epoch": 1.5887786520697913,
+      "grad_norm": 0.32352304458618164,
+      "learning_rate": 5.337951198364907e-07,
+      "loss": 0.1642,
+      "step": 3483
+    },
+    {
+      "epoch": 1.5892348044246778,
+      "grad_norm": 0.3167645335197449,
+      "learning_rate": 5.326548468319825e-07,
+      "loss": 0.1508,
+      "step": 3484
+    },
+    {
+      "epoch": 1.5896909567795645,
+      "grad_norm": 0.31507450342178345,
+      "learning_rate": 5.315156478014896e-07,
+      "loss": 0.1537,
+      "step": 3485
+    },
+    {
+      "epoch": 1.5901471091344508,
+      "grad_norm": 0.30381253361701965,
+      "learning_rate": 5.303775233669004e-07,
+      "loss": 0.1493,
+      "step": 3486
+    },
+    {
+      "epoch": 1.5906032614893375,
+      "grad_norm": 0.3034569025039673,
+      "learning_rate": 5.292404741495172e-07,
+      "loss": 0.1494,
+      "step": 3487
+    },
+    {
+      "epoch": 1.591059413844224,
+      "grad_norm": 0.3198554813861847,
+      "learning_rate": 5.281045007700555e-07,
+      "loss": 0.1514,
+      "step": 3488
+    },
+    {
+      "epoch": 1.5915155661991105,
+      "grad_norm": 0.31144264340400696,
+      "learning_rate": 5.26969603848643e-07,
+      "loss": 0.1497,
+      "step": 3489
+    },
+    {
+      "epoch": 1.591971718553997,
+      "grad_norm": 0.29551345109939575,
+      "learning_rate": 5.258357840048209e-07,
+      "loss": 0.1492,
+      "step": 3490
+    },
+    {
+      "epoch": 1.5924278709088835,
+      "grad_norm": 0.32346290349960327,
+      "learning_rate": 5.247030418575402e-07,
+      "loss": 0.1639,
+      "step": 3491
+    },
+    {
+      "epoch": 1.5928840232637702,
+      "grad_norm": 0.30338621139526367,
+      "learning_rate": 5.235713780251669e-07,
+      "loss": 0.1503,
+      "step": 3492
+    },
+    {
+      "epoch": 1.5933401756186565,
+      "grad_norm": 0.29151561856269836,
+      "learning_rate": 5.224407931254755e-07,
+      "loss": 0.1463,
+      "step": 3493
+    },
+    {
+      "epoch": 1.5937963279735432,
+      "grad_norm": 0.2989885210990906,
+      "learning_rate": 5.213112877756527e-07,
+      "loss": 0.147,
+      "step": 3494
+    },
+    {
+      "epoch": 1.5942524803284297,
+      "grad_norm": 0.31166550517082214,
+      "learning_rate": 5.20182862592295e-07,
+      "loss": 0.1471,
+      "step": 3495
+    },
+    {
+      "epoch": 1.5947086326833162,
+      "grad_norm": 0.3077526390552521,
+      "learning_rate": 5.190555181914106e-07,
+      "loss": 0.1481,
+      "step": 3496
+    },
+    {
+      "epoch": 1.595164785038203,
+      "grad_norm": 0.31219279766082764,
+      "learning_rate": 5.17929255188416e-07,
+      "loss": 0.1512,
+      "step": 3497
+    },
+    {
+      "epoch": 1.5956209373930892,
+      "grad_norm": 0.3102685809135437,
+      "learning_rate": 5.168040741981395e-07,
+      "loss": 0.1484,
+      "step": 3498
+    },
+    {
+      "epoch": 1.596077089747976,
+      "grad_norm": 0.32463139295578003,
+      "learning_rate": 5.156799758348171e-07,
+      "loss": 0.1629,
+      "step": 3499
+    },
+    {
+      "epoch": 1.5965332421028624,
+      "grad_norm": 0.3114151358604431,
+      "learning_rate": 5.14556960712094e-07,
+      "loss": 0.1496,
+      "step": 3500
+    },
+    {
+      "epoch": 1.596989394457749,
+      "grad_norm": 0.30154433846473694,
+      "learning_rate": 5.134350294430249e-07,
+      "loss": 0.1549,
+      "step": 3501
+    },
+    {
+      "epoch": 1.5974455468126354,
+      "grad_norm": 0.3116212785243988,
+      "learning_rate": 5.123141826400718e-07,
+      "loss": 0.1559,
+      "step": 3502
+    },
+    {
+      "epoch": 1.597901699167522,
+      "grad_norm": 0.31571850180625916,
+      "learning_rate": 5.111944209151057e-07,
+      "loss": 0.1492,
+      "step": 3503
+    },
+    {
+      "epoch": 1.5983578515224086,
+      "grad_norm": 0.30454325675964355,
+      "learning_rate": 5.100757448794039e-07,
+      "loss": 0.1501,
+      "step": 3504
+    },
+    {
+      "epoch": 1.5988140038772949,
+      "grad_norm": 0.29900699853897095,
+      "learning_rate": 5.089581551436518e-07,
+      "loss": 0.1497,
+      "step": 3505
+    },
+    {
+      "epoch": 1.5992701562321816,
+      "grad_norm": 0.321414053440094,
+      "learning_rate": 5.078416523179427e-07,
+      "loss": 0.1508,
+      "step": 3506
+    },
+    {
+      "epoch": 1.599726308587068,
+      "grad_norm": 0.3069469928741455,
+      "learning_rate": 5.067262370117751e-07,
+      "loss": 0.1534,
+      "step": 3507
+    },
+    {
+      "epoch": 1.6001824609419546,
+      "grad_norm": 0.31895050406455994,
+      "learning_rate": 5.056119098340545e-07,
+      "loss": 0.1567,
+      "step": 3508
+    },
+    {
+      "epoch": 1.600638613296841,
+      "grad_norm": 0.2984812259674072,
+      "learning_rate": 5.044986713930922e-07,
+      "loss": 0.1436,
+      "step": 3509
+    },
+    {
+      "epoch": 1.6010947656517276,
+      "grad_norm": 0.29863056540489197,
+      "learning_rate": 5.033865222966053e-07,
+      "loss": 0.1457,
+      "step": 3510
+    },
+    {
+      "epoch": 1.6015509180066143,
+      "grad_norm": 0.30738797783851624,
+      "learning_rate": 5.022754631517154e-07,
+      "loss": 0.1516,
+      "step": 3511
+    },
+    {
+      "epoch": 1.6020070703615006,
+      "grad_norm": 0.3156004846096039,
+      "learning_rate": 5.01165494564951e-07,
+      "loss": 0.1545,
+      "step": 3512
+    },
+    {
+      "epoch": 1.6024632227163873,
+      "grad_norm": 0.307125449180603,
+      "learning_rate": 5.000566171422433e-07,
+      "loss": 0.1516,
+      "step": 3513
+    },
+    {
+      "epoch": 1.6029193750712738,
+      "grad_norm": 0.3126176595687866,
+      "learning_rate": 4.989488314889295e-07,
+      "loss": 0.1403,
+      "step": 3514
+    },
+    {
+      "epoch": 1.6033755274261603,
+      "grad_norm": 0.3013862371444702,
+      "learning_rate": 4.978421382097495e-07,
+      "loss": 0.1528,
+      "step": 3515
+    },
+    {
+      "epoch": 1.603831679781047,
+      "grad_norm": 0.30904343724250793,
+      "learning_rate": 4.967365379088473e-07,
+      "loss": 0.1473,
+      "step": 3516
+    },
+    {
+      "epoch": 1.6042878321359333,
+      "grad_norm": 0.3083403706550598,
+      "learning_rate": 4.956320311897705e-07,
+      "loss": 0.1407,
+      "step": 3517
+    },
+    {
+      "epoch": 1.60474398449082,
+      "grad_norm": 0.30868539214134216,
+      "learning_rate": 4.945286186554691e-07,
+      "loss": 0.148,
+      "step": 3518
+    },
+    {
+      "epoch": 1.6052001368457065,
+      "grad_norm": 0.32043275237083435,
+      "learning_rate": 4.934263009082957e-07,
+      "loss": 0.1525,
+      "step": 3519
+    },
+    {
+      "epoch": 1.605656289200593,
+      "grad_norm": 0.31782957911491394,
+      "learning_rate": 4.923250785500069e-07,
+      "loss": 0.1609,
+      "step": 3520
+    },
+    {
+      "epoch": 1.6061124415554795,
+      "grad_norm": 0.3057280480861664,
+      "learning_rate": 4.912249521817589e-07,
+      "loss": 0.1533,
+      "step": 3521
+    },
+    {
+      "epoch": 1.606568593910366,
+      "grad_norm": 0.3016074001789093,
+      "learning_rate": 4.901259224041124e-07,
+      "loss": 0.1448,
+      "step": 3522
+    },
+    {
+      "epoch": 1.6070247462652527,
+      "grad_norm": 0.31116631627082825,
+      "learning_rate": 4.890279898170269e-07,
+      "loss": 0.157,
+      "step": 3523
+    },
+    {
+      "epoch": 1.607480898620139,
+      "grad_norm": 0.3060493469238281,
+      "learning_rate": 4.879311550198643e-07,
+      "loss": 0.1555,
+      "step": 3524
+    },
+    {
+      "epoch": 1.6079370509750257,
+      "grad_norm": 0.29398325085639954,
+      "learning_rate": 4.868354186113866e-07,
+      "loss": 0.1501,
+      "step": 3525
+    },
+    {
+      "epoch": 1.6083932033299122,
+      "grad_norm": 0.29186704754829407,
+      "learning_rate": 4.857407811897566e-07,
+      "loss": 0.1475,
+      "step": 3526
+    },
+    {
+      "epoch": 1.6088493556847987,
+      "grad_norm": 0.3177958130836487,
+      "learning_rate": 4.846472433525365e-07,
+      "loss": 0.1581,
+      "step": 3527
+    },
+    {
+      "epoch": 1.6093055080396854,
+      "grad_norm": 0.2719574570655823,
+      "learning_rate": 4.835548056966899e-07,
+      "loss": 0.1394,
+      "step": 3528
+    },
+    {
+      "epoch": 1.6097616603945717,
+      "grad_norm": 0.3154228627681732,
+      "learning_rate": 4.824634688185778e-07,
+      "loss": 0.1548,
+      "step": 3529
+    },
+    {
+      "epoch": 1.6102178127494584,
+      "grad_norm": 0.31578510999679565,
+      "learning_rate": 4.813732333139617e-07,
+      "loss": 0.1556,
+      "step": 3530
+    },
+    {
+      "epoch": 1.6106739651043447,
+      "grad_norm": 0.31592386960983276,
+      "learning_rate": 4.802840997780003e-07,
+      "loss": 0.1465,
+      "step": 3531
+    },
+    {
+      "epoch": 1.6111301174592314,
+      "grad_norm": 0.3141348659992218,
+      "learning_rate": 4.791960688052533e-07,
+      "loss": 0.159,
+      "step": 3532
+    },
+    {
+      "epoch": 1.611586269814118,
+      "grad_norm": 0.3136829733848572,
+      "learning_rate": 4.781091409896763e-07,
+      "loss": 0.1538,
+      "step": 3533
+    },
+    {
+      "epoch": 1.6120424221690044,
+      "grad_norm": 0.2995051145553589,
+      "learning_rate": 4.77023316924623e-07,
+      "loss": 0.1448,
+      "step": 3534
+    },
+    {
+      "epoch": 1.6124985745238911,
+      "grad_norm": 0.3133722245693207,
+      "learning_rate": 4.7593859720284465e-07,
+      "loss": 0.154,
+      "step": 3535
+    },
+    {
+      "epoch": 1.6129547268787774,
+      "grad_norm": 0.3033691346645355,
+      "learning_rate": 4.748549824164908e-07,
+      "loss": 0.1591,
+      "step": 3536
+    },
+    {
+      "epoch": 1.613410879233664,
+      "grad_norm": 0.3108067810535431,
+      "learning_rate": 4.7377247315710694e-07,
+      "loss": 0.1494,
+      "step": 3537
+    },
+    {
+      "epoch": 1.6138670315885506,
+      "grad_norm": 0.31021302938461304,
+      "learning_rate": 4.7269107001563453e-07,
+      "loss": 0.1627,
+      "step": 3538
+    },
+    {
+      "epoch": 1.614323183943437,
+      "grad_norm": 0.3097059726715088,
+      "learning_rate": 4.716107735824119e-07,
+      "loss": 0.1528,
+      "step": 3539
+    },
+    {
+      "epoch": 1.6147793362983236,
+      "grad_norm": 0.31308677792549133,
+      "learning_rate": 4.705315844471728e-07,
+      "loss": 0.1528,
+      "step": 3540
+    },
+    {
+      "epoch": 1.61523548865321,
+      "grad_norm": 0.3023461401462555,
+      "learning_rate": 4.694535031990463e-07,
+      "loss": 0.1431,
+      "step": 3541
+    },
+    {
+      "epoch": 1.6156916410080968,
+      "grad_norm": 0.2913013696670532,
+      "learning_rate": 4.6837653042655836e-07,
+      "loss": 0.143,
+      "step": 3542
+    },
+    {
+      "epoch": 1.616147793362983,
+      "grad_norm": 0.3190494179725647,
+      "learning_rate": 4.673006667176275e-07,
+      "loss": 0.1556,
+      "step": 3543
+    },
+    {
+      "epoch": 1.6166039457178698,
+      "grad_norm": 0.3215368688106537,
+      "learning_rate": 4.6622591265956863e-07,
+      "loss": 0.1617,
+      "step": 3544
+    },
+    {
+      "epoch": 1.6170600980727563,
+      "grad_norm": 0.3251458406448364,
+      "learning_rate": 4.6515226883909e-07,
+      "loss": 0.1729,
+      "step": 3545
+    },
+    {
+      "epoch": 1.6175162504276428,
+      "grad_norm": 0.30709657073020935,
+      "learning_rate": 4.640797358422938e-07,
+      "loss": 0.1505,
+      "step": 3546
+    },
+    {
+      "epoch": 1.6179724027825295,
+      "grad_norm": 0.32719293236732483,
+      "learning_rate": 4.6300831425467595e-07,
+      "loss": 0.1552,
+      "step": 3547
+    },
+    {
+      "epoch": 1.6184285551374158,
+      "grad_norm": 0.3129293620586395,
+      "learning_rate": 4.6193800466112576e-07,
+      "loss": 0.1528,
+      "step": 3548
+    },
+    {
+      "epoch": 1.6188847074923025,
+      "grad_norm": 0.3009616732597351,
+      "learning_rate": 4.6086880764592444e-07,
+      "loss": 0.1517,
+      "step": 3549
+    },
+    {
+      "epoch": 1.619340859847189,
+      "grad_norm": 0.31559130549430847,
+      "learning_rate": 4.5980072379274846e-07,
+      "loss": 0.1539,
+      "step": 3550
+    },
+    {
+      "epoch": 1.6197970122020755,
+      "grad_norm": 0.3186354339122772,
+      "learning_rate": 4.587337536846634e-07,
+      "loss": 0.1611,
+      "step": 3551
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.30500730872154236,
+      "learning_rate": 4.5766789790412963e-07,
+      "loss": 0.1521,
+      "step": 3552
+    },
+    {
+      "epoch": 1.6207093169118485,
+      "grad_norm": 0.31652188301086426,
+      "learning_rate": 4.566031570329976e-07,
+      "loss": 0.1611,
+      "step": 3553
+    },
+    {
+      "epoch": 1.6211654692667352,
+      "grad_norm": 0.3092023432254791,
+      "learning_rate": 4.5553953165250935e-07,
+      "loss": 0.1551,
+      "step": 3554
+    },
+    {
+      "epoch": 1.6216216216216215,
+      "grad_norm": 0.3010091185569763,
+      "learning_rate": 4.5447702234329846e-07,
+      "loss": 0.1491,
+      "step": 3555
+    },
+    {
+      "epoch": 1.6220777739765082,
+      "grad_norm": 0.3242601752281189,
+      "learning_rate": 4.534156296853884e-07,
+      "loss": 0.1517,
+      "step": 3556
+    },
+    {
+      "epoch": 1.6225339263313947,
+      "grad_norm": 0.3044527769088745,
+      "learning_rate": 4.523553542581932e-07,
+      "loss": 0.1467,
+      "step": 3557
+    },
+    {
+      "epoch": 1.6229900786862812,
+      "grad_norm": 0.3253837823867798,
+      "learning_rate": 4.512961966405185e-07,
+      "loss": 0.1552,
+      "step": 3558
+    },
+    {
+      "epoch": 1.6234462310411677,
+      "grad_norm": 0.31709450483322144,
+      "learning_rate": 4.5023815741055827e-07,
+      "loss": 0.154,
+      "step": 3559
+    },
+    {
+      "epoch": 1.6239023833960542,
+      "grad_norm": 0.3060818016529083,
+      "learning_rate": 4.491812371458962e-07,
+      "loss": 0.1588,
+      "step": 3560
+    },
+    {
+      "epoch": 1.624358535750941,
+      "grad_norm": 0.3102993965148926,
+      "learning_rate": 4.481254364235049e-07,
+      "loss": 0.1493,
+      "step": 3561
+    },
+    {
+      "epoch": 1.6248146881058272,
+      "grad_norm": 0.3216089904308319,
+      "learning_rate": 4.4707075581974597e-07,
+      "loss": 0.1467,
+      "step": 3562
+    },
+    {
+      "epoch": 1.625270840460714,
+      "grad_norm": 0.31532058119773865,
+      "learning_rate": 4.460171959103707e-07,
+      "loss": 0.1536,
+      "step": 3563
+    },
+    {
+      "epoch": 1.6257269928156004,
+      "grad_norm": 0.3276218771934509,
+      "learning_rate": 4.449647572705171e-07,
+      "loss": 0.1567,
+      "step": 3564
+    },
+    {
+      "epoch": 1.626183145170487,
+      "grad_norm": 0.2949657142162323,
+      "learning_rate": 4.439134404747109e-07,
+      "loss": 0.1437,
+      "step": 3565
+    },
+    {
+      "epoch": 1.6266392975253736,
+      "grad_norm": 0.28823360800743103,
+      "learning_rate": 4.428632460968671e-07,
+      "loss": 0.1437,
+      "step": 3566
+    },
+    {
+      "epoch": 1.62709544988026,
+      "grad_norm": 0.31108757853507996,
+      "learning_rate": 4.4181417471028664e-07,
+      "loss": 0.1578,
+      "step": 3567
+    },
+    {
+      "epoch": 1.6275516022351466,
+      "grad_norm": 0.31237173080444336,
+      "learning_rate": 4.407662268876578e-07,
+      "loss": 0.1498,
+      "step": 3568
+    },
+    {
+      "epoch": 1.628007754590033,
+      "grad_norm": 0.30817535519599915,
+      "learning_rate": 4.3971940320105516e-07,
+      "loss": 0.1507,
+      "step": 3569
+    },
+    {
+      "epoch": 1.6284639069449196,
+      "grad_norm": 0.29511311650276184,
+      "learning_rate": 4.3867370422194025e-07,
+      "loss": 0.1481,
+      "step": 3570
+    },
+    {
+      "epoch": 1.628920059299806,
+      "grad_norm": 0.30186212062835693,
+      "learning_rate": 4.3762913052116026e-07,
+      "loss": 0.1512,
+      "step": 3571
+    },
+    {
+      "epoch": 1.6293762116546926,
+      "grad_norm": 0.2990473806858063,
+      "learning_rate": 4.3658568266894716e-07,
+      "loss": 0.1471,
+      "step": 3572
+    },
+    {
+      "epoch": 1.6298323640095793,
+      "grad_norm": 0.3132086396217346,
+      "learning_rate": 4.3554336123492026e-07,
+      "loss": 0.151,
+      "step": 3573
+    },
+    {
+      "epoch": 1.6302885163644656,
+      "grad_norm": 0.3084031939506531,
+      "learning_rate": 4.345021667880836e-07,
+      "loss": 0.1497,
+      "step": 3574
+    },
+    {
+      "epoch": 1.6307446687193523,
+      "grad_norm": 0.2985324263572693,
+      "learning_rate": 4.3346209989682446e-07,
+      "loss": 0.1461,
+      "step": 3575
+    },
+    {
+      "epoch": 1.6312008210742388,
+      "grad_norm": 0.2958592474460602,
+      "learning_rate": 4.324231611289159e-07,
+      "loss": 0.1506,
+      "step": 3576
+    },
+    {
+      "epoch": 1.6316569734291253,
+      "grad_norm": 0.31981906294822693,
+      "learning_rate": 4.313853510515145e-07,
+      "loss": 0.1577,
+      "step": 3577
+    },
+    {
+      "epoch": 1.632113125784012,
+      "grad_norm": 0.3309996426105499,
+      "learning_rate": 4.30348670231161e-07,
+      "loss": 0.151,
+      "step": 3578
+    },
+    {
+      "epoch": 1.6325692781388983,
+      "grad_norm": 0.3010771870613098,
+      "learning_rate": 4.2931311923377904e-07,
+      "loss": 0.1482,
+      "step": 3579
+    },
+    {
+      "epoch": 1.633025430493785,
+      "grad_norm": 0.29624560475349426,
+      "learning_rate": 4.282786986246773e-07,
+      "loss": 0.1466,
+      "step": 3580
+    },
+    {
+      "epoch": 1.6334815828486715,
+      "grad_norm": 0.31187066435813904,
+      "learning_rate": 4.272454089685454e-07,
+      "loss": 0.1524,
+      "step": 3581
+    },
+    {
+      "epoch": 1.633937735203558,
+      "grad_norm": 0.3079606592655182,
+      "learning_rate": 4.2621325082945584e-07,
+      "loss": 0.1386,
+      "step": 3582
+    },
+    {
+      "epoch": 1.6343938875584445,
+      "grad_norm": 0.28866150975227356,
+      "learning_rate": 4.2518222477086507e-07,
+      "loss": 0.1447,
+      "step": 3583
+    },
+    {
+      "epoch": 1.634850039913331,
+      "grad_norm": 0.31563887000083923,
+      "learning_rate": 4.2415233135560977e-07,
+      "loss": 0.1543,
+      "step": 3584
+    },
+    {
+      "epoch": 1.6353061922682177,
+      "grad_norm": 0.2855626344680786,
+      "learning_rate": 4.231235711459089e-07,
+      "loss": 0.1457,
+      "step": 3585
+    },
+    {
+      "epoch": 1.635762344623104,
+      "grad_norm": 0.3184981346130371,
+      "learning_rate": 4.220959447033629e-07,
+      "loss": 0.151,
+      "step": 3586
+    },
+    {
+      "epoch": 1.6362184969779907,
+      "grad_norm": 0.3216342628002167,
+      "learning_rate": 4.210694525889525e-07,
+      "loss": 0.1486,
+      "step": 3587
+    },
+    {
+      "epoch": 1.6366746493328772,
+      "grad_norm": 0.334500789642334,
+      "learning_rate": 4.200440953630411e-07,
+      "loss": 0.1656,
+      "step": 3588
+    },
+    {
+      "epoch": 1.6371308016877637,
+      "grad_norm": 0.3204385042190552,
+      "learning_rate": 4.1901987358537094e-07,
+      "loss": 0.1555,
+      "step": 3589
+    },
+    {
+      "epoch": 1.6375869540426502,
+      "grad_norm": 0.30264127254486084,
+      "learning_rate": 4.1799678781506455e-07,
+      "loss": 0.1529,
+      "step": 3590
+    },
+    {
+      "epoch": 1.6380431063975367,
+      "grad_norm": 0.3188522458076477,
+      "learning_rate": 4.1697483861062455e-07,
+      "loss": 0.1543,
+      "step": 3591
+    },
+    {
+      "epoch": 1.6384992587524234,
+      "grad_norm": 0.317746639251709,
+      "learning_rate": 4.1595402652993304e-07,
+      "loss": 0.1567,
+      "step": 3592
+    },
+    {
+      "epoch": 1.6389554111073097,
+      "grad_norm": 0.30626434087753296,
+      "learning_rate": 4.1493435213025237e-07,
+      "loss": 0.1481,
+      "step": 3593
+    },
+    {
+      "epoch": 1.6394115634621964,
+      "grad_norm": 0.3014124631881714,
+      "learning_rate": 4.1391581596822245e-07,
+      "loss": 0.1453,
+      "step": 3594
+    },
+    {
+      "epoch": 1.639867715817083,
+      "grad_norm": 0.3092781901359558,
+      "learning_rate": 4.1289841859986135e-07,
+      "loss": 0.1511,
+      "step": 3595
+    },
+    {
+      "epoch": 1.6403238681719694,
+      "grad_norm": 0.3005068302154541,
+      "learning_rate": 4.118821605805684e-07,
+      "loss": 0.1488,
+      "step": 3596
+    },
+    {
+      "epoch": 1.6407800205268561,
+      "grad_norm": 0.336172878742218,
+      "learning_rate": 4.108670424651176e-07,
+      "loss": 0.1592,
+      "step": 3597
+    },
+    {
+      "epoch": 1.6412361728817424,
+      "grad_norm": 0.3191240727901459,
+      "learning_rate": 4.098530648076626e-07,
+      "loss": 0.1551,
+      "step": 3598
+    },
+    {
+      "epoch": 1.6416923252366291,
+      "grad_norm": 0.31726711988449097,
+      "learning_rate": 4.088402281617337e-07,
+      "loss": 0.1606,
+      "step": 3599
+    },
+    {
+      "epoch": 1.6421484775915156,
+      "grad_norm": 0.30808186531066895,
+      "learning_rate": 4.0782853308023895e-07,
+      "loss": 0.1605,
+      "step": 3600
+    },
+    {
+      "epoch": 1.642604629946402,
+      "grad_norm": 0.31458622217178345,
+      "learning_rate": 4.068179801154623e-07,
+      "loss": 0.1511,
+      "step": 3601
+    },
+    {
+      "epoch": 1.6430607823012886,
+      "grad_norm": 0.3228464722633362,
+      "learning_rate": 4.0580856981906477e-07,
+      "loss": 0.1451,
+      "step": 3602
+    },
+    {
+      "epoch": 1.643516934656175,
+      "grad_norm": 0.3140561282634735,
+      "learning_rate": 4.0480030274208356e-07,
+      "loss": 0.1545,
+      "step": 3603
+    },
+    {
+      "epoch": 1.6439730870110618,
+      "grad_norm": 0.3280344605445862,
+      "learning_rate": 4.0379317943493287e-07,
+      "loss": 0.1578,
+      "step": 3604
+    },
+    {
+      "epoch": 1.644429239365948,
+      "grad_norm": 0.3193259537220001,
+      "learning_rate": 4.0278720044740066e-07,
+      "loss": 0.1431,
+      "step": 3605
+    },
+    {
+      "epoch": 1.6448853917208348,
+      "grad_norm": 0.32660019397735596,
+      "learning_rate": 4.0178236632865105e-07,
+      "loss": 0.158,
+      "step": 3606
+    },
+    {
+      "epoch": 1.6453415440757213,
+      "grad_norm": 0.32375189661979675,
+      "learning_rate": 4.007786776272232e-07,
+      "loss": 0.1561,
+      "step": 3607
+    },
+    {
+      "epoch": 1.6457976964306078,
+      "grad_norm": 0.3131552040576935,
+      "learning_rate": 3.997761348910309e-07,
+      "loss": 0.1525,
+      "step": 3608
+    },
+    {
+      "epoch": 1.6462538487854943,
+      "grad_norm": 0.3274765908718109,
+      "learning_rate": 3.9877473866736185e-07,
+      "loss": 0.1561,
+      "step": 3609
+    },
+    {
+      "epoch": 1.6467100011403808,
+      "grad_norm": 0.32120758295059204,
+      "learning_rate": 3.977744895028793e-07,
+      "loss": 0.1589,
+      "step": 3610
+    },
+    {
+      "epoch": 1.6471661534952675,
+      "grad_norm": 0.3003138601779938,
+      "learning_rate": 3.96775387943619e-07,
+      "loss": 0.1542,
+      "step": 3611
+    },
+    {
+      "epoch": 1.6476223058501538,
+      "grad_norm": 0.31413623690605164,
+      "learning_rate": 3.95777434534991e-07,
+      "loss": 0.1504,
+      "step": 3612
+    },
+    {
+      "epoch": 1.6480784582050405,
+      "grad_norm": 0.3205919861793518,
+      "learning_rate": 3.9478062982177686e-07,
+      "loss": 0.1553,
+      "step": 3613
+    },
+    {
+      "epoch": 1.648534610559927,
+      "grad_norm": 0.30864235758781433,
+      "learning_rate": 3.937849743481345e-07,
+      "loss": 0.1614,
+      "step": 3614
+    },
+    {
+      "epoch": 1.6489907629148135,
+      "grad_norm": 0.310150146484375,
+      "learning_rate": 3.927904686575909e-07,
+      "loss": 0.1528,
+      "step": 3615
+    },
+    {
+      "epoch": 1.6494469152697002,
+      "grad_norm": 0.304085373878479,
+      "learning_rate": 3.917971132930476e-07,
+      "loss": 0.1523,
+      "step": 3616
+    },
+    {
+      "epoch": 1.6499030676245865,
+      "grad_norm": 0.31853482127189636,
+      "learning_rate": 3.9080490879677636e-07,
+      "loss": 0.1569,
+      "step": 3617
+    },
+    {
+      "epoch": 1.6503592199794732,
+      "grad_norm": 0.3135950267314911,
+      "learning_rate": 3.8981385571042334e-07,
+      "loss": 0.1524,
+      "step": 3618
+    },
+    {
+      "epoch": 1.6508153723343597,
+      "grad_norm": 0.31441137194633484,
+      "learning_rate": 3.888239545750039e-07,
+      "loss": 0.1567,
+      "step": 3619
+    },
+    {
+      "epoch": 1.6512715246892462,
+      "grad_norm": 0.2919193506240845,
+      "learning_rate": 3.8783520593090486e-07,
+      "loss": 0.1507,
+      "step": 3620
+    },
+    {
+      "epoch": 1.6517276770441327,
+      "grad_norm": 0.29587820172309875,
+      "learning_rate": 3.868476103178845e-07,
+      "loss": 0.1463,
+      "step": 3621
+    },
+    {
+      "epoch": 1.6521838293990192,
+      "grad_norm": 0.3096805810928345,
+      "learning_rate": 3.858611682750715e-07,
+      "loss": 0.1479,
+      "step": 3622
+    },
+    {
+      "epoch": 1.652639981753906,
+      "grad_norm": 0.31654176115989685,
+      "learning_rate": 3.848758803409638e-07,
+      "loss": 0.1558,
+      "step": 3623
+    },
+    {
+      "epoch": 1.6530961341087922,
+      "grad_norm": 0.31190750002861023,
+      "learning_rate": 3.83891747053432e-07,
+      "loss": 0.1509,
+      "step": 3624
+    },
+    {
+      "epoch": 1.653552286463679,
+      "grad_norm": 0.30706021189689636,
+      "learning_rate": 3.8290876894971294e-07,
+      "loss": 0.1416,
+      "step": 3625
+    },
+    {
+      "epoch": 1.6540084388185654,
+      "grad_norm": 0.3092644512653351,
+      "learning_rate": 3.8192694656641577e-07,
+      "loss": 0.1554,
+      "step": 3626
+    },
+    {
+      "epoch": 1.654464591173452,
+      "grad_norm": 0.3133431375026703,
+      "learning_rate": 3.8094628043951696e-07,
+      "loss": 0.1545,
+      "step": 3627
+    },
+    {
+      "epoch": 1.6549207435283386,
+      "grad_norm": 0.3105962574481964,
+      "learning_rate": 3.7996677110436226e-07,
+      "loss": 0.1546,
+      "step": 3628
+    },
+    {
+      "epoch": 1.655376895883225,
+      "grad_norm": 0.3142092227935791,
+      "learning_rate": 3.789884190956661e-07,
+      "loss": 0.1612,
+      "step": 3629
+    },
+    {
+      "epoch": 1.6558330482381116,
+      "grad_norm": 0.2971070110797882,
+      "learning_rate": 3.78011224947511e-07,
+      "loss": 0.1458,
+      "step": 3630
+    },
+    {
+      "epoch": 1.6562892005929981,
+      "grad_norm": 0.3151058554649353,
+      "learning_rate": 3.770351891933469e-07,
+      "loss": 0.1492,
+      "step": 3631
+    },
+    {
+      "epoch": 1.6567453529478846,
+      "grad_norm": 0.31935760378837585,
+      "learning_rate": 3.76060312365992e-07,
+      "loss": 0.162,
+      "step": 3632
+    },
+    {
+      "epoch": 1.657201505302771,
+      "grad_norm": 0.3099088966846466,
+      "learning_rate": 3.750865949976318e-07,
+      "loss": 0.1538,
+      "step": 3633
+    },
+    {
+      "epoch": 1.6576576576576576,
+      "grad_norm": 0.30606716871261597,
+      "learning_rate": 3.7411403761981936e-07,
+      "loss": 0.1625,
+      "step": 3634
+    },
+    {
+      "epoch": 1.6581138100125443,
+      "grad_norm": 0.3116915225982666,
+      "learning_rate": 3.7314264076347364e-07,
+      "loss": 0.1496,
+      "step": 3635
+    },
+    {
+      "epoch": 1.6585699623674306,
+      "grad_norm": 0.3003934323787689,
+      "learning_rate": 3.721724049588801e-07,
+      "loss": 0.1516,
+      "step": 3636
+    },
+    {
+      "epoch": 1.6590261147223173,
+      "grad_norm": 0.30263230204582214,
+      "learning_rate": 3.7120333073569046e-07,
+      "loss": 0.1533,
+      "step": 3637
+    },
+    {
+      "epoch": 1.6594822670772038,
+      "grad_norm": 0.32135114073753357,
+      "learning_rate": 3.702354186229229e-07,
+      "loss": 0.1538,
+      "step": 3638
+    },
+    {
+      "epoch": 1.6599384194320903,
+      "grad_norm": 0.3027350902557373,
+      "learning_rate": 3.692686691489597e-07,
+      "loss": 0.1464,
+      "step": 3639
+    },
+    {
+      "epoch": 1.6603945717869768,
+      "grad_norm": 0.3049584925174713,
+      "learning_rate": 3.6830308284155126e-07,
+      "loss": 0.1512,
+      "step": 3640
+    },
+    {
+      "epoch": 1.6608507241418633,
+      "grad_norm": 0.3140557110309601,
+      "learning_rate": 3.673386602278103e-07,
+      "loss": 0.1518,
+      "step": 3641
+    },
+    {
+      "epoch": 1.66130687649675,
+      "grad_norm": 0.29846030473709106,
+      "learning_rate": 3.6637540183421556e-07,
+      "loss": 0.1503,
+      "step": 3642
+    },
+    {
+      "epoch": 1.6617630288516363,
+      "grad_norm": 0.3167378902435303,
+      "learning_rate": 3.65413308186609e-07,
+      "loss": 0.1575,
+      "step": 3643
+    },
+    {
+      "epoch": 1.662219181206523,
+      "grad_norm": 0.30666813254356384,
+      "learning_rate": 3.64452379810199e-07,
+      "loss": 0.152,
+      "step": 3644
+    },
+    {
+      "epoch": 1.6626753335614095,
+      "grad_norm": 0.3021090626716614,
+      "learning_rate": 3.634926172295561e-07,
+      "loss": 0.1453,
+      "step": 3645
+    },
+    {
+      "epoch": 1.663131485916296,
+      "grad_norm": 0.3104269206523895,
+      "learning_rate": 3.6253402096861455e-07,
+      "loss": 0.157,
+      "step": 3646
+    },
+    {
+      "epoch": 1.6635876382711827,
+      "grad_norm": 0.3087924122810364,
+      "learning_rate": 3.6157659155067196e-07,
+      "loss": 0.1551,
+      "step": 3647
+    },
+    {
+      "epoch": 1.664043790626069,
+      "grad_norm": 0.3005847930908203,
+      "learning_rate": 3.606203294983901e-07,
+      "loss": 0.1501,
+      "step": 3648
+    },
+    {
+      "epoch": 1.6644999429809557,
+      "grad_norm": 0.29601916670799255,
+      "learning_rate": 3.5966523533379204e-07,
+      "loss": 0.1531,
+      "step": 3649
+    },
+    {
+      "epoch": 1.6649560953358422,
+      "grad_norm": 0.3251452147960663,
+      "learning_rate": 3.5871130957826404e-07,
+      "loss": 0.1561,
+      "step": 3650
+    },
+    {
+      "epoch": 1.6654122476907287,
+      "grad_norm": 0.30789023637771606,
+      "learning_rate": 3.577585527525537e-07,
+      "loss": 0.1572,
+      "step": 3651
+    },
+    {
+      "epoch": 1.6658684000456152,
+      "grad_norm": 0.30209922790527344,
+      "learning_rate": 3.5680696537677193e-07,
+      "loss": 0.1531,
+      "step": 3652
+    },
+    {
+      "epoch": 1.6663245524005017,
+      "grad_norm": 0.3040454387664795,
+      "learning_rate": 3.558565479703899e-07,
+      "loss": 0.1459,
+      "step": 3653
+    },
+    {
+      "epoch": 1.6667807047553884,
+      "grad_norm": 0.31154385209083557,
+      "learning_rate": 3.5490730105223996e-07,
+      "loss": 0.1493,
+      "step": 3654
+    },
+    {
+      "epoch": 1.6672368571102747,
+      "grad_norm": 0.30782178044319153,
+      "learning_rate": 3.539592251405172e-07,
+      "loss": 0.1522,
+      "step": 3655
+    },
+    {
+      "epoch": 1.6676930094651614,
+      "grad_norm": 0.3217276334762573,
+      "learning_rate": 3.5301232075277635e-07,
+      "loss": 0.1558,
+      "step": 3656
+    },
+    {
+      "epoch": 1.668149161820048,
+      "grad_norm": 0.30550268292427063,
+      "learning_rate": 3.5206658840593266e-07,
+      "loss": 0.1551,
+      "step": 3657
+    },
+    {
+      "epoch": 1.6686053141749344,
+      "grad_norm": 0.32060930132865906,
+      "learning_rate": 3.5112202861626145e-07,
+      "loss": 0.1587,
+      "step": 3658
+    },
+    {
+      "epoch": 1.6690614665298211,
+      "grad_norm": 0.3114902377128601,
+      "learning_rate": 3.501786418993977e-07,
+      "loss": 0.1566,
+      "step": 3659
+    },
+    {
+      "epoch": 1.6695176188847074,
+      "grad_norm": 0.31009548902511597,
+      "learning_rate": 3.4923642877033674e-07,
+      "loss": 0.1489,
+      "step": 3660
+    },
+    {
+      "epoch": 1.6699737712395941,
+      "grad_norm": 0.3054540157318115,
+      "learning_rate": 3.482953897434327e-07,
+      "loss": 0.1517,
+      "step": 3661
+    },
+    {
+      "epoch": 1.6704299235944804,
+      "grad_norm": 0.3006349802017212,
+      "learning_rate": 3.473555253323982e-07,
+      "loss": 0.1524,
+      "step": 3662
+    },
+    {
+      "epoch": 1.6708860759493671,
+      "grad_norm": 0.31384778022766113,
+      "learning_rate": 3.46416836050307e-07,
+      "loss": 0.1501,
+      "step": 3663
+    },
+    {
+      "epoch": 1.6713422283042536,
+      "grad_norm": 0.3021233081817627,
+      "learning_rate": 3.454793224095879e-07,
+      "loss": 0.149,
+      "step": 3664
+    },
+    {
+      "epoch": 1.67179838065914,
+      "grad_norm": 0.2960616648197174,
+      "learning_rate": 3.445429849220314e-07,
+      "loss": 0.1489,
+      "step": 3665
+    },
+    {
+      "epoch": 1.6722545330140268,
+      "grad_norm": 0.30893638730049133,
+      "learning_rate": 3.436078240987836e-07,
+      "loss": 0.1503,
+      "step": 3666
+    },
+    {
+      "epoch": 1.672710685368913,
+      "grad_norm": 0.3147350251674652,
+      "learning_rate": 3.426738404503488e-07,
+      "loss": 0.1591,
+      "step": 3667
+    },
+    {
+      "epoch": 1.6731668377237998,
+      "grad_norm": 0.3257388770580292,
+      "learning_rate": 3.4174103448658875e-07,
+      "loss": 0.1569,
+      "step": 3668
+    },
+    {
+      "epoch": 1.6736229900786863,
+      "grad_norm": 0.3123188614845276,
+      "learning_rate": 3.408094067167228e-07,
+      "loss": 0.1592,
+      "step": 3669
+    },
+    {
+      "epoch": 1.6740791424335728,
+      "grad_norm": 0.3212963044643402,
+      "learning_rate": 3.3987895764932553e-07,
+      "loss": 0.1595,
+      "step": 3670
+    },
+    {
+      "epoch": 1.6745352947884593,
+      "grad_norm": 0.31580206751823425,
+      "learning_rate": 3.389496877923307e-07,
+      "loss": 0.157,
+      "step": 3671
+    },
+    {
+      "epoch": 1.6749914471433458,
+      "grad_norm": 0.3075730800628662,
+      "learning_rate": 3.3802159765302627e-07,
+      "loss": 0.1474,
+      "step": 3672
+    },
+    {
+      "epoch": 1.6754475994982325,
+      "grad_norm": 0.3059377670288086,
+      "learning_rate": 3.3709468773805683e-07,
+      "loss": 0.1504,
+      "step": 3673
+    },
+    {
+      "epoch": 1.6759037518531188,
+      "grad_norm": 0.3095957934856415,
+      "learning_rate": 3.3616895855342195e-07,
+      "loss": 0.1593,
+      "step": 3674
+    },
+    {
+      "epoch": 1.6763599042080055,
+      "grad_norm": 0.3157307505607605,
+      "learning_rate": 3.352444106044786e-07,
+      "loss": 0.1486,
+      "step": 3675
+    },
+    {
+      "epoch": 1.676816056562892,
+      "grad_norm": 0.2993835210800171,
+      "learning_rate": 3.343210443959369e-07,
+      "loss": 0.1511,
+      "step": 3676
+    },
+    {
+      "epoch": 1.6772722089177785,
+      "grad_norm": 0.3100443482398987,
+      "learning_rate": 3.333988604318622e-07,
+      "loss": 0.1515,
+      "step": 3677
+    },
+    {
+      "epoch": 1.6777283612726652,
+      "grad_norm": 0.3142582178115845,
+      "learning_rate": 3.3247785921567607e-07,
+      "loss": 0.1524,
+      "step": 3678
+    },
+    {
+      "epoch": 1.6781845136275515,
+      "grad_norm": 0.31276392936706543,
+      "learning_rate": 3.3155804125015236e-07,
+      "loss": 0.1545,
+      "step": 3679
+    },
+    {
+      "epoch": 1.6786406659824382,
+      "grad_norm": 0.31469160318374634,
+      "learning_rate": 3.3063940703742e-07,
+      "loss": 0.1498,
+      "step": 3680
+    },
+    {
+      "epoch": 1.6790968183373247,
+      "grad_norm": 0.31484293937683105,
+      "learning_rate": 3.2972195707896144e-07,
+      "loss": 0.1608,
+      "step": 3681
+    },
+    {
+      "epoch": 1.6795529706922112,
+      "grad_norm": 0.31062158942222595,
+      "learning_rate": 3.288056918756127e-07,
+      "loss": 0.1496,
+      "step": 3682
+    },
+    {
+      "epoch": 1.6800091230470977,
+      "grad_norm": 0.30368563532829285,
+      "learning_rate": 3.278906119275635e-07,
+      "loss": 0.1524,
+      "step": 3683
+    },
+    {
+      "epoch": 1.6804652754019842,
+      "grad_norm": 0.30971863865852356,
+      "learning_rate": 3.269767177343555e-07,
+      "loss": 0.1518,
+      "step": 3684
+    },
+    {
+      "epoch": 1.680921427756871,
+      "grad_norm": 0.3160220980644226,
+      "learning_rate": 3.260640097948836e-07,
+      "loss": 0.1584,
+      "step": 3685
+    },
+    {
+      "epoch": 1.6813775801117572,
+      "grad_norm": 0.32227766513824463,
+      "learning_rate": 3.2515248860739647e-07,
+      "loss": 0.155,
+      "step": 3686
+    },
+    {
+      "epoch": 1.681833732466644,
+      "grad_norm": 0.3217791020870209,
+      "learning_rate": 3.2424215466949265e-07,
+      "loss": 0.1544,
+      "step": 3687
+    },
+    {
+      "epoch": 1.6822898848215304,
+      "grad_norm": 0.30283334851264954,
+      "learning_rate": 3.2333300847812387e-07,
+      "loss": 0.1567,
+      "step": 3688
+    },
+    {
+      "epoch": 1.682746037176417,
+      "grad_norm": 0.32226166129112244,
+      "learning_rate": 3.224250505295931e-07,
+      "loss": 0.161,
+      "step": 3689
+    },
+    {
+      "epoch": 1.6832021895313034,
+      "grad_norm": 0.3330966532230377,
+      "learning_rate": 3.2151828131955445e-07,
+      "loss": 0.1665,
+      "step": 3690
+    },
+    {
+      "epoch": 1.68365834188619,
+      "grad_norm": 0.3182884156703949,
+      "learning_rate": 3.206127013430138e-07,
+      "loss": 0.1554,
+      "step": 3691
+    },
+    {
+      "epoch": 1.6841144942410766,
+      "grad_norm": 0.31381475925445557,
+      "learning_rate": 3.1970831109432666e-07,
+      "loss": 0.1566,
+      "step": 3692
+    },
+    {
+      "epoch": 1.684570646595963,
+      "grad_norm": 0.29699620604515076,
+      "learning_rate": 3.188051110672008e-07,
+      "loss": 0.1462,
+      "step": 3693
+    },
+    {
+      "epoch": 1.6850267989508496,
+      "grad_norm": 0.31075018644332886,
+      "learning_rate": 3.179031017546927e-07,
+      "loss": 0.147,
+      "step": 3694
+    },
+    {
+      "epoch": 1.685482951305736,
+      "grad_norm": 0.31716203689575195,
+      "learning_rate": 3.17002283649209e-07,
+      "loss": 0.1571,
+      "step": 3695
+    },
+    {
+      "epoch": 1.6859391036606226,
+      "grad_norm": 0.31677770614624023,
+      "learning_rate": 3.161026572425074e-07,
+      "loss": 0.1471,
+      "step": 3696
+    },
+    {
+      "epoch": 1.6863952560155093,
+      "grad_norm": 0.317151814699173,
+      "learning_rate": 3.1520422302569383e-07,
+      "loss": 0.1499,
+      "step": 3697
+    },
+    {
+      "epoch": 1.6868514083703956,
+      "grad_norm": 0.31506291031837463,
+      "learning_rate": 3.14306981489223e-07,
+      "loss": 0.1532,
+      "step": 3698
+    },
+    {
+      "epoch": 1.6873075607252823,
+      "grad_norm": 0.31491878628730774,
+      "learning_rate": 3.1341093312289987e-07,
+      "loss": 0.1568,
+      "step": 3699
+    },
+    {
+      "epoch": 1.6877637130801688,
+      "grad_norm": 0.30678194761276245,
+      "learning_rate": 3.125160784158765e-07,
+      "loss": 0.1567,
+      "step": 3700
+    },
+    {
+      "epoch": 1.6882198654350553,
+      "grad_norm": 0.2992657721042633,
+      "learning_rate": 3.116224178566557e-07,
+      "loss": 0.1473,
+      "step": 3701
+    },
+    {
+      "epoch": 1.6886760177899418,
+      "grad_norm": 0.30563104152679443,
+      "learning_rate": 3.107299519330856e-07,
+      "loss": 0.1459,
+      "step": 3702
+    },
+    {
+      "epoch": 1.6891321701448283,
+      "grad_norm": 0.30882734060287476,
+      "learning_rate": 3.098386811323642e-07,
+      "loss": 0.1531,
+      "step": 3703
+    },
+    {
+      "epoch": 1.689588322499715,
+      "grad_norm": 0.31071731448173523,
+      "learning_rate": 3.0894860594103597e-07,
+      "loss": 0.1542,
+      "step": 3704
+    },
+    {
+      "epoch": 1.6900444748546013,
+      "grad_norm": 0.3254215121269226,
+      "learning_rate": 3.080597268449928e-07,
+      "loss": 0.1606,
+      "step": 3705
+    },
+    {
+      "epoch": 1.690500627209488,
+      "grad_norm": 0.31661999225616455,
+      "learning_rate": 3.071720443294748e-07,
+      "loss": 0.1559,
+      "step": 3706
+    },
+    {
+      "epoch": 1.6909567795643745,
+      "grad_norm": 0.31921330094337463,
+      "learning_rate": 3.0628555887906694e-07,
+      "loss": 0.1548,
+      "step": 3707
+    },
+    {
+      "epoch": 1.691412931919261,
+      "grad_norm": 0.32790619134902954,
+      "learning_rate": 3.054002709777029e-07,
+      "loss": 0.1551,
+      "step": 3708
+    },
+    {
+      "epoch": 1.6918690842741477,
+      "grad_norm": 0.30928003787994385,
+      "learning_rate": 3.045161811086608e-07,
+      "loss": 0.1538,
+      "step": 3709
+    },
+    {
+      "epoch": 1.692325236629034,
+      "grad_norm": 0.2973518669605255,
+      "learning_rate": 3.0363328975456574e-07,
+      "loss": 0.1584,
+      "step": 3710
+    },
+    {
+      "epoch": 1.6927813889839207,
+      "grad_norm": 0.32293006777763367,
+      "learning_rate": 3.0275159739738795e-07,
+      "loss": 0.1568,
+      "step": 3711
+    },
+    {
+      "epoch": 1.693237541338807,
+      "grad_norm": 0.3103497326374054,
+      "learning_rate": 3.018711045184436e-07,
+      "loss": 0.1503,
+      "step": 3712
+    },
+    {
+      "epoch": 1.6936936936936937,
+      "grad_norm": 0.31037601828575134,
+      "learning_rate": 3.0099181159839395e-07,
+      "loss": 0.1552,
+      "step": 3713
+    },
+    {
+      "epoch": 1.6941498460485802,
+      "grad_norm": 0.3057253956794739,
+      "learning_rate": 3.001137191172443e-07,
+      "loss": 0.1558,
+      "step": 3714
+    },
+    {
+      "epoch": 1.6946059984034667,
+      "grad_norm": 0.31681665778160095,
+      "learning_rate": 2.992368275543461e-07,
+      "loss": 0.1529,
+      "step": 3715
+    },
+    {
+      "epoch": 1.6950621507583534,
+      "grad_norm": 0.3103741407394409,
+      "learning_rate": 2.9836113738839544e-07,
+      "loss": 0.1575,
+      "step": 3716
+    },
+    {
+      "epoch": 1.6955183031132397,
+      "grad_norm": 0.29841843247413635,
+      "learning_rate": 2.9748664909743067e-07,
+      "loss": 0.1445,
+      "step": 3717
+    },
+    {
+      "epoch": 1.6959744554681264,
+      "grad_norm": 0.3250938057899475,
+      "learning_rate": 2.966133631588353e-07,
+      "loss": 0.1591,
+      "step": 3718
+    },
+    {
+      "epoch": 1.696430607823013,
+      "grad_norm": 0.3179364502429962,
+      "learning_rate": 2.9574128004933645e-07,
+      "loss": 0.1498,
+      "step": 3719
+    },
+    {
+      "epoch": 1.6968867601778994,
+      "grad_norm": 0.31617099046707153,
+      "learning_rate": 2.9487040024500424e-07,
+      "loss": 0.1544,
+      "step": 3720
+    },
+    {
+      "epoch": 1.697342912532786,
+      "grad_norm": 0.3070639967918396,
+      "learning_rate": 2.9400072422125225e-07,
+      "loss": 0.1485,
+      "step": 3721
+    },
+    {
+      "epoch": 1.6977990648876724,
+      "grad_norm": 0.3113795816898346,
+      "learning_rate": 2.931322524528357e-07,
+      "loss": 0.151,
+      "step": 3722
+    },
+    {
+      "epoch": 1.6982552172425591,
+      "grad_norm": 0.31374844908714294,
+      "learning_rate": 2.9226498541385514e-07,
+      "loss": 0.1457,
+      "step": 3723
+    },
+    {
+      "epoch": 1.6987113695974454,
+      "grad_norm": 0.3126968741416931,
+      "learning_rate": 2.9139892357775115e-07,
+      "loss": 0.1592,
+      "step": 3724
+    },
+    {
+      "epoch": 1.6991675219523321,
+      "grad_norm": 0.32987886667251587,
+      "learning_rate": 2.90534067417306e-07,
+      "loss": 0.1595,
+      "step": 3725
+    },
+    {
+      "epoch": 1.6996236743072186,
+      "grad_norm": 0.3147214949131012,
+      "learning_rate": 2.8967041740464633e-07,
+      "loss": 0.159,
+      "step": 3726
+    },
+    {
+      "epoch": 1.700079826662105,
+      "grad_norm": 0.3148607015609741,
+      "learning_rate": 2.8880797401123845e-07,
+      "loss": 0.1488,
+      "step": 3727
+    },
+    {
+      "epoch": 1.7005359790169918,
+      "grad_norm": 0.31231263279914856,
+      "learning_rate": 2.879467377078896e-07,
+      "loss": 0.152,
+      "step": 3728
+    },
+    {
+      "epoch": 1.700992131371878,
+      "grad_norm": 0.3058549463748932,
+      "learning_rate": 2.8708670896474944e-07,
+      "loss": 0.1505,
+      "step": 3729
+    },
+    {
+      "epoch": 1.7014482837267648,
+      "grad_norm": 0.31522607803344727,
+      "learning_rate": 2.8622788825130727e-07,
+      "loss": 0.1612,
+      "step": 3730
+    },
+    {
+      "epoch": 1.7019044360816513,
+      "grad_norm": 0.31844085454940796,
+      "learning_rate": 2.853702760363941e-07,
+      "loss": 0.1573,
+      "step": 3731
+    },
+    {
+      "epoch": 1.7023605884365378,
+      "grad_norm": 0.31204599142074585,
+      "learning_rate": 2.845138727881802e-07,
+      "loss": 0.1564,
+      "step": 3732
+    },
+    {
+      "epoch": 1.7028167407914243,
+      "grad_norm": 0.327084481716156,
+      "learning_rate": 2.836586789741763e-07,
+      "loss": 0.1601,
+      "step": 3733
+    },
+    {
+      "epoch": 1.7032728931463108,
+      "grad_norm": 0.3034561276435852,
+      "learning_rate": 2.8280469506123286e-07,
+      "loss": 0.1453,
+      "step": 3734
+    },
+    {
+      "epoch": 1.7037290455011975,
+      "grad_norm": 0.31749987602233887,
+      "learning_rate": 2.819519215155392e-07,
+      "loss": 0.1449,
+      "step": 3735
+    },
+    {
+      "epoch": 1.7041851978560838,
+      "grad_norm": 0.31425538659095764,
+      "learning_rate": 2.8110035880262563e-07,
+      "loss": 0.15,
+      "step": 3736
+    },
+    {
+      "epoch": 1.7046413502109705,
+      "grad_norm": 0.301835834980011,
+      "learning_rate": 2.802500073873593e-07,
+      "loss": 0.152,
+      "step": 3737
+    },
+    {
+      "epoch": 1.705097502565857,
+      "grad_norm": 0.3118385970592499,
+      "learning_rate": 2.794008677339483e-07,
+      "loss": 0.1561,
+      "step": 3738
+    },
+    {
+      "epoch": 1.7055536549207435,
+      "grad_norm": 0.3004243075847626,
+      "learning_rate": 2.785529403059373e-07,
+      "loss": 0.1494,
+      "step": 3739
+    },
+    {
+      "epoch": 1.70600980727563,
+      "grad_norm": 0.28671497106552124,
+      "learning_rate": 2.7770622556621054e-07,
+      "loss": 0.1459,
+      "step": 3740
+    },
+    {
+      "epoch": 1.7064659596305165,
+      "grad_norm": 0.315546452999115,
+      "learning_rate": 2.768607239769891e-07,
+      "loss": 0.1501,
+      "step": 3741
+    },
+    {
+      "epoch": 1.7069221119854032,
+      "grad_norm": 0.30942997336387634,
+      "learning_rate": 2.7601643599983297e-07,
+      "loss": 0.151,
+      "step": 3742
+    },
+    {
+      "epoch": 1.7073782643402895,
+      "grad_norm": 0.3129417598247528,
+      "learning_rate": 2.7517336209563845e-07,
+      "loss": 0.1591,
+      "step": 3743
+    },
+    {
+      "epoch": 1.7078344166951762,
+      "grad_norm": 0.3114008605480194,
+      "learning_rate": 2.7433150272463985e-07,
+      "loss": 0.158,
+      "step": 3744
+    },
+    {
+      "epoch": 1.7082905690500627,
+      "grad_norm": 0.299777626991272,
+      "learning_rate": 2.734908583464091e-07,
+      "loss": 0.1476,
+      "step": 3745
+    },
+    {
+      "epoch": 1.7087467214049492,
+      "grad_norm": 0.3086201250553131,
+      "learning_rate": 2.7265142941985315e-07,
+      "loss": 0.1494,
+      "step": 3746
+    },
+    {
+      "epoch": 1.709202873759836,
+      "grad_norm": 0.31288114190101624,
+      "learning_rate": 2.7181321640321724e-07,
+      "loss": 0.1557,
+      "step": 3747
+    },
+    {
+      "epoch": 1.7096590261147222,
+      "grad_norm": 0.2973141372203827,
+      "learning_rate": 2.7097621975408136e-07,
+      "loss": 0.1498,
+      "step": 3748
+    },
+    {
+      "epoch": 1.710115178469609,
+      "grad_norm": 0.2966330647468567,
+      "learning_rate": 2.7014043992936223e-07,
+      "loss": 0.1477,
+      "step": 3749
+    },
+    {
+      "epoch": 1.7105713308244954,
+      "grad_norm": 0.3240519165992737,
+      "learning_rate": 2.693058773853124e-07,
+      "loss": 0.1582,
+      "step": 3750
+    },
+    {
+      "epoch": 1.711027483179382,
+      "grad_norm": 0.30358991026878357,
+      "learning_rate": 2.6847253257751976e-07,
+      "loss": 0.1561,
+      "step": 3751
+    },
+    {
+      "epoch": 1.7114836355342684,
+      "grad_norm": 0.3080276846885681,
+      "learning_rate": 2.676404059609061e-07,
+      "loss": 0.1559,
+      "step": 3752
+    },
+    {
+      "epoch": 1.711939787889155,
+      "grad_norm": 0.30543097853660583,
+      "learning_rate": 2.668094979897315e-07,
+      "loss": 0.1506,
+      "step": 3753
+    },
+    {
+      "epoch": 1.7123959402440416,
+      "grad_norm": 0.30791059136390686,
+      "learning_rate": 2.659798091175875e-07,
+      "loss": 0.1546,
+      "step": 3754
+    },
+    {
+      "epoch": 1.712852092598928,
+      "grad_norm": 0.304038405418396,
+      "learning_rate": 2.651513397974015e-07,
+      "loss": 0.1586,
+      "step": 3755
+    },
+    {
+      "epoch": 1.7133082449538146,
+      "grad_norm": 0.3026449382305145,
+      "learning_rate": 2.6432409048143483e-07,
+      "loss": 0.1424,
+      "step": 3756
+    },
+    {
+      "epoch": 1.7137643973087011,
+      "grad_norm": 0.30461442470550537,
+      "learning_rate": 2.6349806162128374e-07,
+      "loss": 0.1564,
+      "step": 3757
+    },
+    {
+      "epoch": 1.7142205496635876,
+      "grad_norm": 0.3181486427783966,
+      "learning_rate": 2.626732536678772e-07,
+      "loss": 0.1506,
+      "step": 3758
+    },
+    {
+      "epoch": 1.7146767020184743,
+      "grad_norm": 0.31676286458969116,
+      "learning_rate": 2.618496670714779e-07,
+      "loss": 0.1494,
+      "step": 3759
+    },
+    {
+      "epoch": 1.7151328543733606,
+      "grad_norm": 0.2967580258846283,
+      "learning_rate": 2.610273022816812e-07,
+      "loss": 0.1472,
+      "step": 3760
+    },
+    {
+      "epoch": 1.7155890067282473,
+      "grad_norm": 0.3084806203842163,
+      "learning_rate": 2.602061597474176e-07,
+      "loss": 0.1534,
+      "step": 3761
+    },
+    {
+      "epoch": 1.7160451590831338,
+      "grad_norm": 0.32001441717147827,
+      "learning_rate": 2.593862399169478e-07,
+      "loss": 0.1586,
+      "step": 3762
+    },
+    {
+      "epoch": 1.7165013114380203,
+      "grad_norm": 0.3095225393772125,
+      "learning_rate": 2.5856754323786676e-07,
+      "loss": 0.1528,
+      "step": 3763
+    },
+    {
+      "epoch": 1.7169574637929068,
+      "grad_norm": 0.30565565824508667,
+      "learning_rate": 2.577500701571009e-07,
+      "loss": 0.1599,
+      "step": 3764
+    },
+    {
+      "epoch": 1.7174136161477933,
+      "grad_norm": 0.30566373467445374,
+      "learning_rate": 2.569338211209091e-07,
+      "loss": 0.1502,
+      "step": 3765
+    },
+    {
+      "epoch": 1.71786976850268,
+      "grad_norm": 0.3035769462585449,
+      "learning_rate": 2.5611879657488096e-07,
+      "loss": 0.1508,
+      "step": 3766
+    },
+    {
+      "epoch": 1.7183259208575663,
+      "grad_norm": 0.3105255961418152,
+      "learning_rate": 2.55304996963939e-07,
+      "loss": 0.1501,
+      "step": 3767
+    },
+    {
+      "epoch": 1.718782073212453,
+      "grad_norm": 0.3071368634700775,
+      "learning_rate": 2.5449242273233753e-07,
+      "loss": 0.1526,
+      "step": 3768
+    },
+    {
+      "epoch": 1.7192382255673395,
+      "grad_norm": 0.30545637011528015,
+      "learning_rate": 2.5368107432365977e-07,
+      "loss": 0.1534,
+      "step": 3769
+    },
+    {
+      "epoch": 1.719694377922226,
+      "grad_norm": 0.3120470345020294,
+      "learning_rate": 2.528709521808215e-07,
+      "loss": 0.1526,
+      "step": 3770
+    },
+    {
+      "epoch": 1.7201505302771125,
+      "grad_norm": 0.32080304622650146,
+      "learning_rate": 2.5206205674606825e-07,
+      "loss": 0.1534,
+      "step": 3771
+    },
+    {
+      "epoch": 1.720606682631999,
+      "grad_norm": 0.2998294532299042,
+      "learning_rate": 2.5125438846097636e-07,
+      "loss": 0.1481,
+      "step": 3772
+    },
+    {
+      "epoch": 1.7210628349868857,
+      "grad_norm": 0.2931610941886902,
+      "learning_rate": 2.5044794776645204e-07,
+      "loss": 0.1458,
+      "step": 3773
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.3261629641056061,
+      "learning_rate": 2.4964273510273056e-07,
+      "loss": 0.1614,
+      "step": 3774
+    },
+    {
+      "epoch": 1.7219751396966587,
+      "grad_norm": 0.32197389006614685,
+      "learning_rate": 2.48838750909379e-07,
+      "loss": 0.1567,
+      "step": 3775
+    },
+    {
+      "epoch": 1.7224312920515452,
+      "grad_norm": 0.31405317783355713,
+      "learning_rate": 2.480359956252912e-07,
+      "loss": 0.1552,
+      "step": 3776
+    },
+    {
+      "epoch": 1.7228874444064317,
+      "grad_norm": 0.3031854033470154,
+      "learning_rate": 2.4723446968869246e-07,
+      "loss": 0.1431,
+      "step": 3777
+    },
+    {
+      "epoch": 1.7233435967613184,
+      "grad_norm": 0.2973707318305969,
+      "learning_rate": 2.4643417353713563e-07,
+      "loss": 0.1418,
+      "step": 3778
+    },
+    {
+      "epoch": 1.7237997491162047,
+      "grad_norm": 0.296577125787735,
+      "learning_rate": 2.45635107607502e-07,
+      "loss": 0.1426,
+      "step": 3779
+    },
+    {
+      "epoch": 1.7242559014710914,
+      "grad_norm": 0.3165368437767029,
+      "learning_rate": 2.4483727233600237e-07,
+      "loss": 0.1565,
+      "step": 3780
+    },
+    {
+      "epoch": 1.724712053825978,
+      "grad_norm": 0.3143184781074524,
+      "learning_rate": 2.440406681581747e-07,
+      "loss": 0.1489,
+      "step": 3781
+    },
+    {
+      "epoch": 1.7251682061808644,
+      "grad_norm": 0.32117438316345215,
+      "learning_rate": 2.4324529550888516e-07,
+      "loss": 0.1535,
+      "step": 3782
+    },
+    {
+      "epoch": 1.725624358535751,
+      "grad_norm": 0.3097103536128998,
+      "learning_rate": 2.4245115482232876e-07,
+      "loss": 0.1505,
+      "step": 3783
+    },
+    {
+      "epoch": 1.7260805108906374,
+      "grad_norm": 0.3214317858219147,
+      "learning_rate": 2.4165824653202605e-07,
+      "loss": 0.1631,
+      "step": 3784
+    },
+    {
+      "epoch": 1.7265366632455241,
+      "grad_norm": 0.3127461373806,
+      "learning_rate": 2.408665710708263e-07,
+      "loss": 0.1531,
+      "step": 3785
+    },
+    {
+      "epoch": 1.7269928156004104,
+      "grad_norm": 0.3036994934082031,
+      "learning_rate": 2.4007612887090525e-07,
+      "loss": 0.1483,
+      "step": 3786
+    },
+    {
+      "epoch": 1.7274489679552971,
+      "grad_norm": 0.30662062764167786,
+      "learning_rate": 2.392869203637646e-07,
+      "loss": 0.1477,
+      "step": 3787
+    },
+    {
+      "epoch": 1.7279051203101836,
+      "grad_norm": 0.2999681532382965,
+      "learning_rate": 2.384989459802345e-07,
+      "loss": 0.1428,
+      "step": 3788
+    },
+    {
+      "epoch": 1.7283612726650701,
+      "grad_norm": 0.3167629837989807,
+      "learning_rate": 2.3771220615046996e-07,
+      "loss": 0.1574,
+      "step": 3789
+    },
+    {
+      "epoch": 1.7288174250199566,
+      "grad_norm": 0.30609169602394104,
+      "learning_rate": 2.3692670130395167e-07,
+      "loss": 0.1493,
+      "step": 3790
+    },
+    {
+      "epoch": 1.729273577374843,
+      "grad_norm": 0.3148931562900543,
+      "learning_rate": 2.3614243186948798e-07,
+      "loss": 0.1568,
+      "step": 3791
+    },
+    {
+      "epoch": 1.7297297297297298,
+      "grad_norm": 0.3236512243747711,
+      "learning_rate": 2.3535939827521114e-07,
+      "loss": 0.1552,
+      "step": 3792
+    },
+    {
+      "epoch": 1.730185882084616,
+      "grad_norm": 0.312885582447052,
+      "learning_rate": 2.3457760094857923e-07,
+      "loss": 0.1556,
+      "step": 3793
+    },
+    {
+      "epoch": 1.7306420344395028,
+      "grad_norm": 0.3259197473526001,
+      "learning_rate": 2.3379704031637557e-07,
+      "loss": 0.1531,
+      "step": 3794
+    },
+    {
+      "epoch": 1.7310981867943893,
+      "grad_norm": 0.30803072452545166,
+      "learning_rate": 2.3301771680470846e-07,
+      "loss": 0.1494,
+      "step": 3795
+    },
+    {
+      "epoch": 1.7315543391492758,
+      "grad_norm": 0.3061206340789795,
+      "learning_rate": 2.3223963083901014e-07,
+      "loss": 0.1585,
+      "step": 3796
+    },
+    {
+      "epoch": 1.7320104915041625,
+      "grad_norm": 0.29331690073013306,
+      "learning_rate": 2.3146278284403857e-07,
+      "loss": 0.1402,
+      "step": 3797
+    },
+    {
+      "epoch": 1.7324666438590488,
+      "grad_norm": 0.31748324632644653,
+      "learning_rate": 2.3068717324387479e-07,
+      "loss": 0.1549,
+      "step": 3798
+    },
+    {
+      "epoch": 1.7329227962139355,
+      "grad_norm": 0.30344825983047485,
+      "learning_rate": 2.299128024619246e-07,
+      "loss": 0.1508,
+      "step": 3799
+    },
+    {
+      "epoch": 1.733378948568822,
+      "grad_norm": 0.31567999720573425,
+      "learning_rate": 2.2913967092091706e-07,
+      "loss": 0.1629,
+      "step": 3800
+    },
+    {
+      "epoch": 1.7338351009237085,
+      "grad_norm": 0.30371516942977905,
+      "learning_rate": 2.2836777904290454e-07,
+      "loss": 0.1446,
+      "step": 3801
+    },
+    {
+      "epoch": 1.734291253278595,
+      "grad_norm": 0.31138890981674194,
+      "learning_rate": 2.2759712724926304e-07,
+      "loss": 0.1541,
+      "step": 3802
+    },
+    {
+      "epoch": 1.7347474056334815,
+      "grad_norm": 0.2812519073486328,
+      "learning_rate": 2.2682771596069158e-07,
+      "loss": 0.1371,
+      "step": 3803
+    },
+    {
+      "epoch": 1.7352035579883682,
+      "grad_norm": 0.2973928153514862,
+      "learning_rate": 2.2605954559721116e-07,
+      "loss": 0.1446,
+      "step": 3804
+    },
+    {
+      "epoch": 1.7356597103432545,
+      "grad_norm": 0.3021758198738098,
+      "learning_rate": 2.2529261657816715e-07,
+      "loss": 0.147,
+      "step": 3805
+    },
+    {
+      "epoch": 1.7361158626981412,
+      "grad_norm": 0.29365262389183044,
+      "learning_rate": 2.2452692932222575e-07,
+      "loss": 0.1468,
+      "step": 3806
+    },
+    {
+      "epoch": 1.7365720150530277,
+      "grad_norm": 0.31687840819358826,
+      "learning_rate": 2.237624842473754e-07,
+      "loss": 0.1549,
+      "step": 3807
+    },
+    {
+      "epoch": 1.7370281674079142,
+      "grad_norm": 0.3148006200790405,
+      "learning_rate": 2.2299928177092784e-07,
+      "loss": 0.158,
+      "step": 3808
+    },
+    {
+      "epoch": 1.737484319762801,
+      "grad_norm": 0.3191524147987366,
+      "learning_rate": 2.222373223095145e-07,
+      "loss": 0.1552,
+      "step": 3809
+    },
+    {
+      "epoch": 1.7379404721176872,
+      "grad_norm": 0.31188297271728516,
+      "learning_rate": 2.2147660627908935e-07,
+      "loss": 0.1507,
+      "step": 3810
+    },
+    {
+      "epoch": 1.738396624472574,
+      "grad_norm": 0.3140028715133667,
+      "learning_rate": 2.2071713409492768e-07,
+      "loss": 0.1501,
+      "step": 3811
+    },
+    {
+      "epoch": 1.7388527768274604,
+      "grad_norm": 0.3273194432258606,
+      "learning_rate": 2.199589061716245e-07,
+      "loss": 0.1646,
+      "step": 3812
+    },
+    {
+      "epoch": 1.739308929182347,
+      "grad_norm": 0.3324027955532074,
+      "learning_rate": 2.1920192292309795e-07,
+      "loss": 0.1606,
+      "step": 3813
+    },
+    {
+      "epoch": 1.7397650815372334,
+      "grad_norm": 0.30967575311660767,
+      "learning_rate": 2.1844618476258435e-07,
+      "loss": 0.1527,
+      "step": 3814
+    },
+    {
+      "epoch": 1.74022123389212,
+      "grad_norm": 0.3013947904109955,
+      "learning_rate": 2.1769169210264174e-07,
+      "loss": 0.149,
+      "step": 3815
+    },
+    {
+      "epoch": 1.7406773862470066,
+      "grad_norm": 0.31766411662101746,
+      "learning_rate": 2.1693844535514763e-07,
+      "loss": 0.158,
+      "step": 3816
+    },
+    {
+      "epoch": 1.741133538601893,
+      "grad_norm": 0.3132251501083374,
+      "learning_rate": 2.1618644493129893e-07,
+      "loss": 0.1547,
+      "step": 3817
+    },
+    {
+      "epoch": 1.7415896909567796,
+      "grad_norm": 0.31327876448631287,
+      "learning_rate": 2.1543569124161389e-07,
+      "loss": 0.1529,
+      "step": 3818
+    },
+    {
+      "epoch": 1.7420458433116661,
+      "grad_norm": 0.31884223222732544,
+      "learning_rate": 2.1468618469592862e-07,
+      "loss": 0.1626,
+      "step": 3819
+    },
+    {
+      "epoch": 1.7425019956665526,
+      "grad_norm": 0.3187384605407715,
+      "learning_rate": 2.13937925703398e-07,
+      "loss": 0.1595,
+      "step": 3820
+    },
+    {
+      "epoch": 1.7429581480214391,
+      "grad_norm": 0.31518635153770447,
+      "learning_rate": 2.1319091467249807e-07,
+      "loss": 0.1601,
+      "step": 3821
+    },
+    {
+      "epoch": 1.7434143003763256,
+      "grad_norm": 0.30604448914527893,
+      "learning_rate": 2.1244515201102185e-07,
+      "loss": 0.1489,
+      "step": 3822
+    },
+    {
+      "epoch": 1.7438704527312123,
+      "grad_norm": 0.3182459771633148,
+      "learning_rate": 2.1170063812608084e-07,
+      "loss": 0.1573,
+      "step": 3823
+    },
+    {
+      "epoch": 1.7443266050860986,
+      "grad_norm": 0.30910247564315796,
+      "learning_rate": 2.1095737342410577e-07,
+      "loss": 0.1502,
+      "step": 3824
+    },
+    {
+      "epoch": 1.7447827574409853,
+      "grad_norm": 0.30253705382347107,
+      "learning_rate": 2.102153583108446e-07,
+      "loss": 0.1479,
+      "step": 3825
+    },
+    {
+      "epoch": 1.7452389097958718,
+      "grad_norm": 0.3141220510005951,
+      "learning_rate": 2.0947459319136332e-07,
+      "loss": 0.1575,
+      "step": 3826
+    },
+    {
+      "epoch": 1.7456950621507583,
+      "grad_norm": 0.31064289808273315,
+      "learning_rate": 2.087350784700462e-07,
+      "loss": 0.161,
+      "step": 3827
+    },
+    {
+      "epoch": 1.746151214505645,
+      "grad_norm": 0.2949979901313782,
+      "learning_rate": 2.0799681455059423e-07,
+      "loss": 0.1512,
+      "step": 3828
+    },
+    {
+      "epoch": 1.7466073668605313,
+      "grad_norm": 0.31937408447265625,
+      "learning_rate": 2.0725980183602584e-07,
+      "loss": 0.1622,
+      "step": 3829
+    },
+    {
+      "epoch": 1.747063519215418,
+      "grad_norm": 0.3182561695575714,
+      "learning_rate": 2.065240407286767e-07,
+      "loss": 0.1575,
+      "step": 3830
+    },
+    {
+      "epoch": 1.7475196715703045,
+      "grad_norm": 0.3306896388530731,
+      "learning_rate": 2.0578953163019832e-07,
+      "loss": 0.1517,
+      "step": 3831
+    },
+    {
+      "epoch": 1.747975823925191,
+      "grad_norm": 0.3067154288291931,
+      "learning_rate": 2.0505627494155994e-07,
+      "loss": 0.1532,
+      "step": 3832
+    },
+    {
+      "epoch": 1.7484319762800775,
+      "grad_norm": 0.3205502927303314,
+      "learning_rate": 2.043242710630458e-07,
+      "loss": 0.1591,
+      "step": 3833
+    },
+    {
+      "epoch": 1.748888128634964,
+      "grad_norm": 0.33209317922592163,
+      "learning_rate": 2.0359352039425712e-07,
+      "loss": 0.1591,
+      "step": 3834
+    },
+    {
+      "epoch": 1.7493442809898507,
+      "grad_norm": 0.3140029013156891,
+      "learning_rate": 2.0286402333411115e-07,
+      "loss": 0.1551,
+      "step": 3835
+    },
+    {
+      "epoch": 1.749800433344737,
+      "grad_norm": 0.309185653924942,
+      "learning_rate": 2.0213578028084018e-07,
+      "loss": 0.1464,
+      "step": 3836
+    },
+    {
+      "epoch": 1.7502565856996237,
+      "grad_norm": 0.31191909313201904,
+      "learning_rate": 2.0140879163199263e-07,
+      "loss": 0.1531,
+      "step": 3837
+    },
+    {
+      "epoch": 1.7507127380545102,
+      "grad_norm": 0.31868529319763184,
+      "learning_rate": 2.006830577844307e-07,
+      "loss": 0.1515,
+      "step": 3838
+    },
+    {
+      "epoch": 1.7511688904093967,
+      "grad_norm": 0.30582472681999207,
+      "learning_rate": 1.9995857913433363e-07,
+      "loss": 0.1503,
+      "step": 3839
+    },
+    {
+      "epoch": 1.7516250427642834,
+      "grad_norm": 0.32486236095428467,
+      "learning_rate": 1.9923535607719423e-07,
+      "loss": 0.1561,
+      "step": 3840
+    },
+    {
+      "epoch": 1.7520811951191697,
+      "grad_norm": 0.325562059879303,
+      "learning_rate": 1.9851338900781974e-07,
+      "loss": 0.1555,
+      "step": 3841
+    },
+    {
+      "epoch": 1.7525373474740564,
+      "grad_norm": 0.3121775984764099,
+      "learning_rate": 1.9779267832033182e-07,
+      "loss": 0.1566,
+      "step": 3842
+    },
+    {
+      "epoch": 1.7529934998289427,
+      "grad_norm": 0.2984280288219452,
+      "learning_rate": 1.9707322440816744e-07,
+      "loss": 0.1474,
+      "step": 3843
+    },
+    {
+      "epoch": 1.7534496521838294,
+      "grad_norm": 0.31493279337882996,
+      "learning_rate": 1.9635502766407632e-07,
+      "loss": 0.1527,
+      "step": 3844
+    },
+    {
+      "epoch": 1.753905804538716,
+      "grad_norm": 0.31919893622398376,
+      "learning_rate": 1.9563808848012178e-07,
+      "loss": 0.1637,
+      "step": 3845
+    },
+    {
+      "epoch": 1.7543619568936024,
+      "grad_norm": 0.3184933662414551,
+      "learning_rate": 1.949224072476813e-07,
+      "loss": 0.1542,
+      "step": 3846
+    },
+    {
+      "epoch": 1.7548181092484891,
+      "grad_norm": 0.3189142942428589,
+      "learning_rate": 1.9420798435744513e-07,
+      "loss": 0.1495,
+      "step": 3847
+    },
+    {
+      "epoch": 1.7552742616033754,
+      "grad_norm": 0.3065921366214752,
+      "learning_rate": 1.934948201994169e-07,
+      "loss": 0.1491,
+      "step": 3848
+    },
+    {
+      "epoch": 1.7557304139582621,
+      "grad_norm": 0.3141615688800812,
+      "learning_rate": 1.9278291516291347e-07,
+      "loss": 0.1583,
+      "step": 3849
+    },
+    {
+      "epoch": 1.7561865663131486,
+      "grad_norm": 0.3099326193332672,
+      "learning_rate": 1.9207226963656289e-07,
+      "loss": 0.1496,
+      "step": 3850
+    },
+    {
+      "epoch": 1.7566427186680351,
+      "grad_norm": 0.3254956901073456,
+      "learning_rate": 1.9136288400830817e-07,
+      "loss": 0.1535,
+      "step": 3851
+    },
+    {
+      "epoch": 1.7570988710229216,
+      "grad_norm": 0.2959491014480591,
+      "learning_rate": 1.9065475866540172e-07,
+      "loss": 0.1462,
+      "step": 3852
+    },
+    {
+      "epoch": 1.7575550233778081,
+      "grad_norm": 0.2889825403690338,
+      "learning_rate": 1.899478939944102e-07,
+      "loss": 0.1458,
+      "step": 3853
+    },
+    {
+      "epoch": 1.7580111757326948,
+      "grad_norm": 0.31652894616127014,
+      "learning_rate": 1.892422903812105e-07,
+      "loss": 0.1539,
+      "step": 3854
+    },
+    {
+      "epoch": 1.758467328087581,
+      "grad_norm": 0.3272593915462494,
+      "learning_rate": 1.8853794821099198e-07,
+      "loss": 0.1593,
+      "step": 3855
+    },
+    {
+      "epoch": 1.7589234804424678,
+      "grad_norm": 0.316224604845047,
+      "learning_rate": 1.8783486786825455e-07,
+      "loss": 0.1519,
+      "step": 3856
+    },
+    {
+      "epoch": 1.7593796327973543,
+      "grad_norm": 0.3165028393268585,
+      "learning_rate": 1.8713304973681096e-07,
+      "loss": 0.1539,
+      "step": 3857
+    },
+    {
+      "epoch": 1.7598357851522408,
+      "grad_norm": 0.29622161388397217,
+      "learning_rate": 1.8643249419978276e-07,
+      "loss": 0.146,
+      "step": 3858
+    },
+    {
+      "epoch": 1.7602919375071275,
+      "grad_norm": 0.3198483884334564,
+      "learning_rate": 1.857332016396046e-07,
+      "loss": 0.155,
+      "step": 3859
+    },
+    {
+      "epoch": 1.7607480898620138,
+      "grad_norm": 0.3207602798938751,
+      "learning_rate": 1.8503517243801922e-07,
+      "loss": 0.1598,
+      "step": 3860
+    },
+    {
+      "epoch": 1.7612042422169005,
+      "grad_norm": 0.3029036521911621,
+      "learning_rate": 1.8433840697608147e-07,
+      "loss": 0.1568,
+      "step": 3861
+    },
+    {
+      "epoch": 1.761660394571787,
+      "grad_norm": 0.30976957082748413,
+      "learning_rate": 1.836429056341557e-07,
+      "loss": 0.1481,
+      "step": 3862
+    },
+    {
+      "epoch": 1.7621165469266735,
+      "grad_norm": 0.3119716942310333,
+      "learning_rate": 1.8294866879191574e-07,
+      "loss": 0.1471,
+      "step": 3863
+    },
+    {
+      "epoch": 1.76257269928156,
+      "grad_norm": 0.32527920603752136,
+      "learning_rate": 1.8225569682834565e-07,
+      "loss": 0.1552,
+      "step": 3864
+    },
+    {
+      "epoch": 1.7630288516364465,
+      "grad_norm": 0.3159204125404358,
+      "learning_rate": 1.8156399012173974e-07,
+      "loss": 0.1504,
+      "step": 3865
+    },
+    {
+      "epoch": 1.7634850039913332,
+      "grad_norm": 0.30727481842041016,
+      "learning_rate": 1.8087354904969988e-07,
+      "loss": 0.1508,
+      "step": 3866
+    },
+    {
+      "epoch": 1.7639411563462195,
+      "grad_norm": 0.3094164729118347,
+      "learning_rate": 1.8018437398913847e-07,
+      "loss": 0.1484,
+      "step": 3867
+    },
+    {
+      "epoch": 1.7643973087011062,
+      "grad_norm": 0.2928580045700073,
+      "learning_rate": 1.794964653162759e-07,
+      "loss": 0.1469,
+      "step": 3868
+    },
+    {
+      "epoch": 1.7648534610559927,
+      "grad_norm": 0.3132634460926056,
+      "learning_rate": 1.7880982340664182e-07,
+      "loss": 0.1521,
+      "step": 3869
+    },
+    {
+      "epoch": 1.7653096134108792,
+      "grad_norm": 0.30082637071609497,
+      "learning_rate": 1.7812444863507444e-07,
+      "loss": 0.1495,
+      "step": 3870
+    },
+    {
+      "epoch": 1.7657657657657657,
+      "grad_norm": 0.3163737654685974,
+      "learning_rate": 1.7744034137571975e-07,
+      "loss": 0.1452,
+      "step": 3871
+    },
+    {
+      "epoch": 1.7662219181206522,
+      "grad_norm": 0.31443390250205994,
+      "learning_rate": 1.7675750200203152e-07,
+      "loss": 0.1566,
+      "step": 3872
+    },
+    {
+      "epoch": 1.766678070475539,
+      "grad_norm": 0.2988242208957672,
+      "learning_rate": 1.7607593088677238e-07,
+      "loss": 0.1532,
+      "step": 3873
+    },
+    {
+      "epoch": 1.7671342228304252,
+      "grad_norm": 0.3275682032108307,
+      "learning_rate": 1.7539562840201223e-07,
+      "loss": 0.1526,
+      "step": 3874
+    },
+    {
+      "epoch": 1.767590375185312,
+      "grad_norm": 0.2954939007759094,
+      "learning_rate": 1.7471659491912818e-07,
+      "loss": 0.1445,
+      "step": 3875
+    },
+    {
+      "epoch": 1.7680465275401984,
+      "grad_norm": 0.2918040454387665,
+      "learning_rate": 1.7403883080880424e-07,
+      "loss": 0.1479,
+      "step": 3876
+    },
+    {
+      "epoch": 1.768502679895085,
+      "grad_norm": 0.30294790863990784,
+      "learning_rate": 1.7336233644103279e-07,
+      "loss": 0.1497,
+      "step": 3877
+    },
+    {
+      "epoch": 1.7689588322499716,
+      "grad_norm": 0.3147088289260864,
+      "learning_rate": 1.7268711218511092e-07,
+      "loss": 0.1464,
+      "step": 3878
+    },
+    {
+      "epoch": 1.769414984604858,
+      "grad_norm": 0.3061390221118927,
+      "learning_rate": 1.720131584096449e-07,
+      "loss": 0.1533,
+      "step": 3879
+    },
+    {
+      "epoch": 1.7698711369597446,
+      "grad_norm": 0.30812811851501465,
+      "learning_rate": 1.713404754825454e-07,
+      "loss": 0.149,
+      "step": 3880
+    },
+    {
+      "epoch": 1.7703272893146311,
+      "grad_norm": 0.31023749709129333,
+      "learning_rate": 1.7066906377103093e-07,
+      "loss": 0.1544,
+      "step": 3881
+    },
+    {
+      "epoch": 1.7707834416695176,
+      "grad_norm": 0.3018827736377716,
+      "learning_rate": 1.6999892364162497e-07,
+      "loss": 0.1426,
+      "step": 3882
+    },
+    {
+      "epoch": 1.7712395940244041,
+      "grad_norm": 0.3102011978626251,
+      "learning_rate": 1.693300554601568e-07,
+      "loss": 0.1591,
+      "step": 3883
+    },
+    {
+      "epoch": 1.7716957463792906,
+      "grad_norm": 0.31897613406181335,
+      "learning_rate": 1.6866245959176157e-07,
+      "loss": 0.1498,
+      "step": 3884
+    },
+    {
+      "epoch": 1.7721518987341773,
+      "grad_norm": 0.30343523621559143,
+      "learning_rate": 1.679961364008803e-07,
+      "loss": 0.1401,
+      "step": 3885
+    },
+    {
+      "epoch": 1.7726080510890636,
+      "grad_norm": 0.3148173391819,
+      "learning_rate": 1.6733108625125866e-07,
+      "loss": 0.1635,
+      "step": 3886
+    },
+    {
+      "epoch": 1.7730642034439503,
+      "grad_norm": 0.29460960626602173,
+      "learning_rate": 1.666673095059479e-07,
+      "loss": 0.1453,
+      "step": 3887
+    },
+    {
+      "epoch": 1.7735203557988368,
+      "grad_norm": 0.3192636966705322,
+      "learning_rate": 1.66004806527304e-07,
+      "loss": 0.1508,
+      "step": 3888
+    },
+    {
+      "epoch": 1.7739765081537233,
+      "grad_norm": 0.3023424446582794,
+      "learning_rate": 1.6534357767698656e-07,
+      "loss": 0.1494,
+      "step": 3889
+    },
+    {
+      "epoch": 1.77443266050861,
+      "grad_norm": 0.32384225726127625,
+      "learning_rate": 1.6468362331596156e-07,
+      "loss": 0.1567,
+      "step": 3890
+    },
+    {
+      "epoch": 1.7748888128634963,
+      "grad_norm": 0.3079982399940491,
+      "learning_rate": 1.6402494380449773e-07,
+      "loss": 0.1514,
+      "step": 3891
+    },
+    {
+      "epoch": 1.775344965218383,
+      "grad_norm": 0.3033638000488281,
+      "learning_rate": 1.6336753950216833e-07,
+      "loss": 0.1515,
+      "step": 3892
+    },
+    {
+      "epoch": 1.7758011175732695,
+      "grad_norm": 0.3139593005180359,
+      "learning_rate": 1.6271141076785068e-07,
+      "loss": 0.1579,
+      "step": 3893
+    },
+    {
+      "epoch": 1.776257269928156,
+      "grad_norm": 0.32384148240089417,
+      "learning_rate": 1.6205655795972469e-07,
+      "loss": 0.157,
+      "step": 3894
+    },
+    {
+      "epoch": 1.7767134222830425,
+      "grad_norm": 0.30533429980278015,
+      "learning_rate": 1.614029814352755e-07,
+      "loss": 0.1444,
+      "step": 3895
+    },
+    {
+      "epoch": 1.777169574637929,
+      "grad_norm": 0.30882418155670166,
+      "learning_rate": 1.6075068155129075e-07,
+      "loss": 0.1505,
+      "step": 3896
+    },
+    {
+      "epoch": 1.7776257269928157,
+      "grad_norm": 0.31655317544937134,
+      "learning_rate": 1.6009965866386034e-07,
+      "loss": 0.1504,
+      "step": 3897
+    },
+    {
+      "epoch": 1.778081879347702,
+      "grad_norm": 0.2847752571105957,
+      "learning_rate": 1.5944991312837832e-07,
+      "loss": 0.1444,
+      "step": 3898
+    },
+    {
+      "epoch": 1.7785380317025887,
+      "grad_norm": 0.3142144978046417,
+      "learning_rate": 1.5880144529954038e-07,
+      "loss": 0.1556,
+      "step": 3899
+    },
+    {
+      "epoch": 1.7789941840574752,
+      "grad_norm": 0.3130475580692291,
+      "learning_rate": 1.5815425553134562e-07,
+      "loss": 0.159,
+      "step": 3900
+    },
+    {
+      "epoch": 1.7794503364123617,
+      "grad_norm": 0.31044310331344604,
+      "learning_rate": 1.57508344177095e-07,
+      "loss": 0.1566,
+      "step": 3901
+    },
+    {
+      "epoch": 1.7799064887672482,
+      "grad_norm": 0.3058020770549774,
+      "learning_rate": 1.5686371158939118e-07,
+      "loss": 0.1477,
+      "step": 3902
+    },
+    {
+      "epoch": 1.7803626411221347,
+      "grad_norm": 0.3123171329498291,
+      "learning_rate": 1.5622035812013964e-07,
+      "loss": 0.153,
+      "step": 3903
+    },
+    {
+      "epoch": 1.7808187934770214,
+      "grad_norm": 0.319073885679245,
+      "learning_rate": 1.555782841205475e-07,
+      "loss": 0.1545,
+      "step": 3904
+    },
+    {
+      "epoch": 1.7812749458319077,
+      "grad_norm": 0.31320637464523315,
+      "learning_rate": 1.5493748994112211e-07,
+      "loss": 0.1539,
+      "step": 3905
+    },
+    {
+      "epoch": 1.7817310981867944,
+      "grad_norm": 0.3183026611804962,
+      "learning_rate": 1.5429797593167373e-07,
+      "loss": 0.1498,
+      "step": 3906
+    },
+    {
+      "epoch": 1.782187250541681,
+      "grad_norm": 0.3146790862083435,
+      "learning_rate": 1.5365974244131304e-07,
+      "loss": 0.1501,
+      "step": 3907
+    },
+    {
+      "epoch": 1.7826434028965674,
+      "grad_norm": 0.31185182929039,
+      "learning_rate": 1.5302278981845136e-07,
+      "loss": 0.1493,
+      "step": 3908
+    },
+    {
+      "epoch": 1.7830995552514541,
+      "grad_norm": 0.319352388381958,
+      "learning_rate": 1.5238711841080222e-07,
+      "loss": 0.1541,
+      "step": 3909
+    },
+    {
+      "epoch": 1.7835557076063404,
+      "grad_norm": 0.3251616060733795,
+      "learning_rate": 1.517527285653775e-07,
+      "loss": 0.1598,
+      "step": 3910
+    },
+    {
+      "epoch": 1.7840118599612271,
+      "grad_norm": 0.30998292565345764,
+      "learning_rate": 1.511196206284918e-07,
+      "loss": 0.158,
+      "step": 3911
+    },
+    {
+      "epoch": 1.7844680123161136,
+      "grad_norm": 0.3228926956653595,
+      "learning_rate": 1.5048779494575838e-07,
+      "loss": 0.162,
+      "step": 3912
+    },
+    {
+      "epoch": 1.7849241646710001,
+      "grad_norm": 0.31751784682273865,
+      "learning_rate": 1.4985725186209084e-07,
+      "loss": 0.1517,
+      "step": 3913
+    },
+    {
+      "epoch": 1.7853803170258866,
+      "grad_norm": 0.3163653612136841,
+      "learning_rate": 1.4922799172170295e-07,
+      "loss": 0.1555,
+      "step": 3914
+    },
+    {
+      "epoch": 1.7858364693807731,
+      "grad_norm": 0.305044949054718,
+      "learning_rate": 1.48600014868108e-07,
+      "loss": 0.149,
+      "step": 3915
+    },
+    {
+      "epoch": 1.7862926217356598,
+      "grad_norm": 0.32684648036956787,
+      "learning_rate": 1.4797332164411816e-07,
+      "loss": 0.1508,
+      "step": 3916
+    },
+    {
+      "epoch": 1.786748774090546,
+      "grad_norm": 0.31179702281951904,
+      "learning_rate": 1.4734791239184586e-07,
+      "loss": 0.1523,
+      "step": 3917
+    },
+    {
+      "epoch": 1.7872049264454328,
+      "grad_norm": 0.3156852126121521,
+      "learning_rate": 1.4672378745270193e-07,
+      "loss": 0.1535,
+      "step": 3918
+    },
+    {
+      "epoch": 1.7876610788003193,
+      "grad_norm": 0.3063938617706299,
+      "learning_rate": 1.4610094716739604e-07,
+      "loss": 0.1475,
+      "step": 3919
+    },
+    {
+      "epoch": 1.7881172311552058,
+      "grad_norm": 0.31440865993499756,
+      "learning_rate": 1.454793918759373e-07,
+      "loss": 0.1536,
+      "step": 3920
+    },
+    {
+      "epoch": 1.7885733835100923,
+      "grad_norm": 0.29051798582077026,
+      "learning_rate": 1.4485912191763262e-07,
+      "loss": 0.1396,
+      "step": 3921
+    },
+    {
+      "epoch": 1.7890295358649788,
+      "grad_norm": 0.30701833963394165,
+      "learning_rate": 1.4424013763108758e-07,
+      "loss": 0.1543,
+      "step": 3922
+    },
+    {
+      "epoch": 1.7894856882198655,
+      "grad_norm": 0.30064529180526733,
+      "learning_rate": 1.4362243935420545e-07,
+      "loss": 0.1428,
+      "step": 3923
+    },
+    {
+      "epoch": 1.7899418405747518,
+      "grad_norm": 0.3048557639122009,
+      "learning_rate": 1.430060274241876e-07,
+      "loss": 0.1477,
+      "step": 3924
+    },
+    {
+      "epoch": 1.7903979929296385,
+      "grad_norm": 0.3085685074329376,
+      "learning_rate": 1.423909021775341e-07,
+      "loss": 0.1475,
+      "step": 3925
+    },
+    {
+      "epoch": 1.790854145284525,
+      "grad_norm": 0.3075980544090271,
+      "learning_rate": 1.4177706395004182e-07,
+      "loss": 0.1558,
+      "step": 3926
+    },
+    {
+      "epoch": 1.7913102976394115,
+      "grad_norm": 0.3340415060520172,
+      "learning_rate": 1.4116451307680467e-07,
+      "loss": 0.1631,
+      "step": 3927
+    },
+    {
+      "epoch": 1.7917664499942982,
+      "grad_norm": 0.29679277539253235,
+      "learning_rate": 1.4055324989221464e-07,
+      "loss": 0.1488,
+      "step": 3928
+    },
+    {
+      "epoch": 1.7922226023491845,
+      "grad_norm": 0.3233134150505066,
+      "learning_rate": 1.3994327472996033e-07,
+      "loss": 0.1559,
+      "step": 3929
+    },
+    {
+      "epoch": 1.7926787547040712,
+      "grad_norm": 0.3093467354774475,
+      "learning_rate": 1.393345879230268e-07,
+      "loss": 0.1467,
+      "step": 3930
+    },
+    {
+      "epoch": 1.7931349070589577,
+      "grad_norm": 0.31063348054885864,
+      "learning_rate": 1.387271898036971e-07,
+      "loss": 0.1551,
+      "step": 3931
+    },
+    {
+      "epoch": 1.7935910594138442,
+      "grad_norm": 0.3240969479084015,
+      "learning_rate": 1.3812108070354908e-07,
+      "loss": 0.1521,
+      "step": 3932
+    },
+    {
+      "epoch": 1.7940472117687307,
+      "grad_norm": 0.30702143907546997,
+      "learning_rate": 1.375162609534586e-07,
+      "loss": 0.1502,
+      "step": 3933
+    },
+    {
+      "epoch": 1.7945033641236172,
+      "grad_norm": 0.3138985335826874,
+      "learning_rate": 1.3691273088359658e-07,
+      "loss": 0.1534,
+      "step": 3934
+    },
+    {
+      "epoch": 1.794959516478504,
+      "grad_norm": 0.3021513521671295,
+      "learning_rate": 1.3631049082343002e-07,
+      "loss": 0.1464,
+      "step": 3935
+    },
+    {
+      "epoch": 1.7954156688333902,
+      "grad_norm": 0.31062158942222595,
+      "learning_rate": 1.3570954110172203e-07,
+      "loss": 0.1516,
+      "step": 3936
+    },
+    {
+      "epoch": 1.795871821188277,
+      "grad_norm": 0.2992668151855469,
+      "learning_rate": 1.3510988204653103e-07,
+      "loss": 0.1459,
+      "step": 3937
+    },
+    {
+      "epoch": 1.7963279735431634,
+      "grad_norm": 0.31658652424812317,
+      "learning_rate": 1.3451151398521072e-07,
+      "loss": 0.1624,
+      "step": 3938
+    },
+    {
+      "epoch": 1.79678412589805,
+      "grad_norm": 0.30990517139434814,
+      "learning_rate": 1.3391443724441066e-07,
+      "loss": 0.1496,
+      "step": 3939
+    },
+    {
+      "epoch": 1.7972402782529366,
+      "grad_norm": 0.32465288043022156,
+      "learning_rate": 1.3331865215007482e-07,
+      "loss": 0.1602,
+      "step": 3940
+    },
+    {
+      "epoch": 1.797696430607823,
+      "grad_norm": 0.3142143189907074,
+      "learning_rate": 1.3272415902744306e-07,
+      "loss": 0.1503,
+      "step": 3941
+    },
+    {
+      "epoch": 1.7981525829627096,
+      "grad_norm": 0.29844924807548523,
+      "learning_rate": 1.321309582010491e-07,
+      "loss": 0.1485,
+      "step": 3942
+    },
+    {
+      "epoch": 1.7986087353175961,
+      "grad_norm": 0.32214438915252686,
+      "learning_rate": 1.3153904999472083e-07,
+      "loss": 0.1545,
+      "step": 3943
+    },
+    {
+      "epoch": 1.7990648876724826,
+      "grad_norm": 0.3080918490886688,
+      "learning_rate": 1.309484347315812e-07,
+      "loss": 0.1488,
+      "step": 3944
+    },
+    {
+      "epoch": 1.7995210400273691,
+      "grad_norm": 0.30766546726226807,
+      "learning_rate": 1.3035911273404756e-07,
+      "loss": 0.1483,
+      "step": 3945
+    },
+    {
+      "epoch": 1.7999771923822556,
+      "grad_norm": 0.3130015432834625,
+      "learning_rate": 1.2977108432383006e-07,
+      "loss": 0.1537,
+      "step": 3946
+    },
+    {
+      "epoch": 1.8004333447371423,
+      "grad_norm": 0.3149689733982086,
+      "learning_rate": 1.2918434982193446e-07,
+      "loss": 0.157,
+      "step": 3947
+    },
+    {
+      "epoch": 1.8008894970920286,
+      "grad_norm": 0.30314093828201294,
+      "learning_rate": 1.2859890954865867e-07,
+      "loss": 0.1535,
+      "step": 3948
+    },
+    {
+      "epoch": 1.8013456494469153,
+      "grad_norm": 0.3214733600616455,
+      "learning_rate": 1.2801476382359484e-07,
+      "loss": 0.1507,
+      "step": 3949
+    },
+    {
+      "epoch": 1.8018018018018018,
+      "grad_norm": 0.3278011679649353,
+      "learning_rate": 1.2743191296562762e-07,
+      "loss": 0.16,
+      "step": 3950
+    },
+    {
+      "epoch": 1.8022579541566883,
+      "grad_norm": 0.329528272151947,
+      "learning_rate": 1.2685035729293636e-07,
+      "loss": 0.1704,
+      "step": 3951
+    },
+    {
+      "epoch": 1.8027141065115748,
+      "grad_norm": 0.3273981213569641,
+      "learning_rate": 1.262700971229916e-07,
+      "loss": 0.1561,
+      "step": 3952
+    },
+    {
+      "epoch": 1.8031702588664613,
+      "grad_norm": 0.3025049865245819,
+      "learning_rate": 1.2569113277255796e-07,
+      "loss": 0.1486,
+      "step": 3953
+    },
+    {
+      "epoch": 1.803626411221348,
+      "grad_norm": 0.31925293803215027,
+      "learning_rate": 1.251134645576918e-07,
+      "loss": 0.1543,
+      "step": 3954
+    },
+    {
+      "epoch": 1.8040825635762343,
+      "grad_norm": 0.29942646622657776,
+      "learning_rate": 1.2453709279374255e-07,
+      "loss": 0.1492,
+      "step": 3955
+    },
+    {
+      "epoch": 1.804538715931121,
+      "grad_norm": 0.3033880889415741,
+      "learning_rate": 1.2396201779535154e-07,
+      "loss": 0.1514,
+      "step": 3956
+    },
+    {
+      "epoch": 1.8049948682860075,
+      "grad_norm": 0.3091328740119934,
+      "learning_rate": 1.2338823987645216e-07,
+      "loss": 0.1488,
+      "step": 3957
+    },
+    {
+      "epoch": 1.805451020640894,
+      "grad_norm": 0.3139559328556061,
+      "learning_rate": 1.2281575935027018e-07,
+      "loss": 0.1571,
+      "step": 3958
+    },
+    {
+      "epoch": 1.8059071729957807,
+      "grad_norm": 0.3051230013370514,
+      "learning_rate": 1.2224457652932258e-07,
+      "loss": 0.1519,
+      "step": 3959
+    },
+    {
+      "epoch": 1.806363325350667,
+      "grad_norm": 0.303396999835968,
+      "learning_rate": 1.2167469172541773e-07,
+      "loss": 0.1525,
+      "step": 3960
+    },
+    {
+      "epoch": 1.8068194777055537,
+      "grad_norm": 0.305642306804657,
+      "learning_rate": 1.2110610524965678e-07,
+      "loss": 0.1465,
+      "step": 3961
+    },
+    {
+      "epoch": 1.8072756300604402,
+      "grad_norm": 0.3080022931098938,
+      "learning_rate": 1.2053881741243028e-07,
+      "loss": 0.1581,
+      "step": 3962
+    },
+    {
+      "epoch": 1.8077317824153267,
+      "grad_norm": 0.2865687608718872,
+      "learning_rate": 1.199728285234214e-07,
+      "loss": 0.14,
+      "step": 3963
+    },
+    {
+      "epoch": 1.8081879347702132,
+      "grad_norm": 0.3068815767765045,
+      "learning_rate": 1.194081388916035e-07,
+      "loss": 0.1572,
+      "step": 3964
+    },
+    {
+      "epoch": 1.8086440871250997,
+      "grad_norm": 0.30765724182128906,
+      "learning_rate": 1.1884474882524033e-07,
+      "loss": 0.1544,
+      "step": 3965
+    },
+    {
+      "epoch": 1.8091002394799864,
+      "grad_norm": 0.31275638937950134,
+      "learning_rate": 1.1828265863188693e-07,
+      "loss": 0.1513,
+      "step": 3966
+    },
+    {
+      "epoch": 1.8095563918348727,
+      "grad_norm": 0.31821131706237793,
+      "learning_rate": 1.1772186861838842e-07,
+      "loss": 0.1516,
+      "step": 3967
+    },
+    {
+      "epoch": 1.8100125441897594,
+      "grad_norm": 0.3181442618370056,
+      "learning_rate": 1.1716237909087991e-07,
+      "loss": 0.1544,
+      "step": 3968
+    },
+    {
+      "epoch": 1.810468696544646,
+      "grad_norm": 0.3058021664619446,
+      "learning_rate": 1.1660419035478654e-07,
+      "loss": 0.1563,
+      "step": 3969
+    },
+    {
+      "epoch": 1.8109248488995324,
+      "grad_norm": 0.3056985139846802,
+      "learning_rate": 1.1604730271482429e-07,
+      "loss": 0.152,
+      "step": 3970
+    },
+    {
+      "epoch": 1.811381001254419,
+      "grad_norm": 0.3018074333667755,
+      "learning_rate": 1.1549171647499773e-07,
+      "loss": 0.151,
+      "step": 3971
+    },
+    {
+      "epoch": 1.8118371536093054,
+      "grad_norm": 0.31270530819892883,
+      "learning_rate": 1.1493743193860207e-07,
+      "loss": 0.1506,
+      "step": 3972
+    },
+    {
+      "epoch": 1.8122933059641921,
+      "grad_norm": 0.2971462607383728,
+      "learning_rate": 1.143844494082208e-07,
+      "loss": 0.1482,
+      "step": 3973
+    },
+    {
+      "epoch": 1.8127494583190784,
+      "grad_norm": 0.3096173405647278,
+      "learning_rate": 1.1383276918572744e-07,
+      "loss": 0.1494,
+      "step": 3974
+    },
+    {
+      "epoch": 1.8132056106739651,
+      "grad_norm": 0.2918759882450104,
+      "learning_rate": 1.1328239157228415e-07,
+      "loss": 0.1394,
+      "step": 3975
+    },
+    {
+      "epoch": 1.8136617630288516,
+      "grad_norm": 0.30815228819847107,
+      "learning_rate": 1.12733316868342e-07,
+      "loss": 0.1559,
+      "step": 3976
+    },
+    {
+      "epoch": 1.8141179153837381,
+      "grad_norm": 0.29987481236457825,
+      "learning_rate": 1.1218554537364118e-07,
+      "loss": 0.1544,
+      "step": 3977
+    },
+    {
+      "epoch": 1.8145740677386248,
+      "grad_norm": 0.3271811604499817,
+      "learning_rate": 1.116390773872103e-07,
+      "loss": 0.1582,
+      "step": 3978
+    },
+    {
+      "epoch": 1.8150302200935111,
+      "grad_norm": 0.2820909917354584,
+      "learning_rate": 1.1109391320736628e-07,
+      "loss": 0.1462,
+      "step": 3979
+    },
+    {
+      "epoch": 1.8154863724483978,
+      "grad_norm": 0.31981268525123596,
+      "learning_rate": 1.1055005313171413e-07,
+      "loss": 0.1558,
+      "step": 3980
+    },
+    {
+      "epoch": 1.8159425248032843,
+      "grad_norm": 0.3149668276309967,
+      "learning_rate": 1.100074974571469e-07,
+      "loss": 0.1482,
+      "step": 3981
+    },
+    {
+      "epoch": 1.8163986771581708,
+      "grad_norm": 0.312797874212265,
+      "learning_rate": 1.0946624647984655e-07,
+      "loss": 0.154,
+      "step": 3982
+    },
+    {
+      "epoch": 1.8168548295130573,
+      "grad_norm": 0.3090377748012543,
+      "learning_rate": 1.0892630049528175e-07,
+      "loss": 0.1513,
+      "step": 3983
+    },
+    {
+      "epoch": 1.8173109818679438,
+      "grad_norm": 0.3144042193889618,
+      "learning_rate": 1.0838765979820892e-07,
+      "loss": 0.1557,
+      "step": 3984
+    },
+    {
+      "epoch": 1.8177671342228305,
+      "grad_norm": 0.3171330690383911,
+      "learning_rate": 1.0785032468267232e-07,
+      "loss": 0.1515,
+      "step": 3985
+    },
+    {
+      "epoch": 1.8182232865777168,
+      "grad_norm": 0.32198697328567505,
+      "learning_rate": 1.0731429544200312e-07,
+      "loss": 0.1575,
+      "step": 3986
+    },
+    {
+      "epoch": 1.8186794389326035,
+      "grad_norm": 0.315370112657547,
+      "learning_rate": 1.0677957236881975e-07,
+      "loss": 0.1523,
+      "step": 3987
+    },
+    {
+      "epoch": 1.81913559128749,
+      "grad_norm": 0.3180811107158661,
+      "learning_rate": 1.0624615575502789e-07,
+      "loss": 0.1491,
+      "step": 3988
+    },
+    {
+      "epoch": 1.8195917436423765,
+      "grad_norm": 0.308650940656662,
+      "learning_rate": 1.0571404589181933e-07,
+      "loss": 0.1546,
+      "step": 3989
+    },
+    {
+      "epoch": 1.8200478959972632,
+      "grad_norm": 0.3229517638683319,
+      "learning_rate": 1.051832430696731e-07,
+      "loss": 0.1492,
+      "step": 3990
+    },
+    {
+      "epoch": 1.8205040483521495,
+      "grad_norm": 0.32051587104797363,
+      "learning_rate": 1.0465374757835461e-07,
+      "loss": 0.1597,
+      "step": 3991
+    },
+    {
+      "epoch": 1.8209602007070362,
+      "grad_norm": 0.3195159137248993,
+      "learning_rate": 1.0412555970691519e-07,
+      "loss": 0.1596,
+      "step": 3992
+    },
+    {
+      "epoch": 1.8214163530619227,
+      "grad_norm": 0.30502697825431824,
+      "learning_rate": 1.0359867974369331e-07,
+      "loss": 0.1515,
+      "step": 3993
+    },
+    {
+      "epoch": 1.8218725054168092,
+      "grad_norm": 0.30719655752182007,
+      "learning_rate": 1.0307310797631253e-07,
+      "loss": 0.1436,
+      "step": 3994
+    },
+    {
+      "epoch": 1.8223286577716957,
+      "grad_norm": 0.30686649680137634,
+      "learning_rate": 1.0254884469168253e-07,
+      "loss": 0.1497,
+      "step": 3995
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.3082399070262909,
+      "learning_rate": 1.0202589017599878e-07,
+      "loss": 0.1482,
+      "step": 3996
+    },
+    {
+      "epoch": 1.823240962481469,
+      "grad_norm": 0.3087017238140106,
+      "learning_rate": 1.0150424471474213e-07,
+      "loss": 0.1574,
+      "step": 3997
+    },
+    {
+      "epoch": 1.8236971148363552,
+      "grad_norm": 0.307973176240921,
+      "learning_rate": 1.0098390859267893e-07,
+      "loss": 0.1532,
+      "step": 3998
+    },
+    {
+      "epoch": 1.824153267191242,
+      "grad_norm": 0.3320818841457367,
+      "learning_rate": 1.0046488209386029e-07,
+      "loss": 0.1635,
+      "step": 3999
+    },
+    {
+      "epoch": 1.8246094195461284,
+      "grad_norm": 0.2974562346935272,
+      "learning_rate": 9.994716550162376e-08,
+      "loss": 0.1526,
+      "step": 4000
+    },
+    {
+      "epoch": 1.825065571901015,
+      "grad_norm": 0.3090340793132782,
+      "learning_rate": 9.943075909859017e-08,
+      "loss": 0.1526,
+      "step": 4001
+    },
+    {
+      "epoch": 1.8255217242559014,
+      "grad_norm": 0.29951414465904236,
+      "learning_rate": 9.891566316666628e-08,
+      "loss": 0.1463,
+      "step": 4002
+    },
+    {
+      "epoch": 1.825977876610788,
+      "grad_norm": 0.30130425095558167,
+      "learning_rate": 9.840187798704326e-08,
+      "loss": 0.1487,
+      "step": 4003
+    },
+    {
+      "epoch": 1.8264340289656746,
+      "grad_norm": 0.30784785747528076,
+      "learning_rate": 9.788940384019591e-08,
+      "loss": 0.1562,
+      "step": 4004
+    },
+    {
+      "epoch": 1.826890181320561,
+      "grad_norm": 0.30998003482818604,
+      "learning_rate": 9.737824100588406e-08,
+      "loss": 0.149,
+      "step": 4005
+    },
+    {
+      "epoch": 1.8273463336754476,
+      "grad_norm": 0.3123064935207367,
+      "learning_rate": 9.686838976315171e-08,
+      "loss": 0.1517,
+      "step": 4006
+    },
+    {
+      "epoch": 1.8278024860303341,
+      "grad_norm": 0.29321834444999695,
+      "learning_rate": 9.635985039032681e-08,
+      "loss": 0.1479,
+      "step": 4007
+    },
+    {
+      "epoch": 1.8282586383852206,
+      "grad_norm": 0.30024078488349915,
+      "learning_rate": 9.585262316502114e-08,
+      "loss": 0.1456,
+      "step": 4008
+    },
+    {
+      "epoch": 1.8287147907401073,
+      "grad_norm": 0.3033483028411865,
+      "learning_rate": 9.534670836412991e-08,
+      "loss": 0.1448,
+      "step": 4009
+    },
+    {
+      "epoch": 1.8291709430949936,
+      "grad_norm": 0.29524654150009155,
+      "learning_rate": 9.484210626383217e-08,
+      "loss": 0.1531,
+      "step": 4010
+    },
+    {
+      "epoch": 1.8296270954498803,
+      "grad_norm": 0.31876012682914734,
+      "learning_rate": 9.433881713959008e-08,
+      "loss": 0.1573,
+      "step": 4011
+    },
+    {
+      "epoch": 1.8300832478047668,
+      "grad_norm": 0.3208746612071991,
+      "learning_rate": 9.383684126614945e-08,
+      "loss": 0.156,
+      "step": 4012
+    },
+    {
+      "epoch": 1.8305394001596533,
+      "grad_norm": 0.31692856550216675,
+      "learning_rate": 9.333617891753938e-08,
+      "loss": 0.1556,
+      "step": 4013
+    },
+    {
+      "epoch": 1.8309955525145398,
+      "grad_norm": 0.3363637328147888,
+      "learning_rate": 9.283683036707041e-08,
+      "loss": 0.1559,
+      "step": 4014
+    },
+    {
+      "epoch": 1.8314517048694263,
+      "grad_norm": 0.321727454662323,
+      "learning_rate": 9.23387958873384e-08,
+      "loss": 0.154,
+      "step": 4015
+    },
+    {
+      "epoch": 1.831907857224313,
+      "grad_norm": 0.3119792938232422,
+      "learning_rate": 9.184207575021947e-08,
+      "loss": 0.1521,
+      "step": 4016
+    },
+    {
+      "epoch": 1.8323640095791993,
+      "grad_norm": 0.309603750705719,
+      "learning_rate": 9.134667022687338e-08,
+      "loss": 0.1527,
+      "step": 4017
+    },
+    {
+      "epoch": 1.832820161934086,
+      "grad_norm": 0.2901313304901123,
+      "learning_rate": 9.08525795877424e-08,
+      "loss": 0.1415,
+      "step": 4018
+    },
+    {
+      "epoch": 1.8332763142889725,
+      "grad_norm": 0.31954070925712585,
+      "learning_rate": 9.035980410255052e-08,
+      "loss": 0.1519,
+      "step": 4019
+    },
+    {
+      "epoch": 1.833732466643859,
+      "grad_norm": 0.33050987124443054,
+      "learning_rate": 8.986834404030364e-08,
+      "loss": 0.1663,
+      "step": 4020
+    },
+    {
+      "epoch": 1.8341886189987457,
+      "grad_norm": 0.2971230745315552,
+      "learning_rate": 8.937819966929023e-08,
+      "loss": 0.1447,
+      "step": 4021
+    },
+    {
+      "epoch": 1.834644771353632,
+      "grad_norm": 0.3163544237613678,
+      "learning_rate": 8.888937125707986e-08,
+      "loss": 0.1604,
+      "step": 4022
+    },
+    {
+      "epoch": 1.8351009237085187,
+      "grad_norm": 0.30251064896583557,
+      "learning_rate": 8.840185907052434e-08,
+      "loss": 0.1487,
+      "step": 4023
+    },
+    {
+      "epoch": 1.835557076063405,
+      "grad_norm": 0.3192661702632904,
+      "learning_rate": 8.79156633757569e-08,
+      "loss": 0.149,
+      "step": 4024
+    },
+    {
+      "epoch": 1.8360132284182917,
+      "grad_norm": 0.31080418825149536,
+      "learning_rate": 8.743078443819137e-08,
+      "loss": 0.1571,
+      "step": 4025
+    },
+    {
+      "epoch": 1.8364693807731782,
+      "grad_norm": 0.3174874782562256,
+      "learning_rate": 8.694722252252347e-08,
+      "loss": 0.1523,
+      "step": 4026
+    },
+    {
+      "epoch": 1.8369255331280647,
+      "grad_norm": 0.3181985914707184,
+      "learning_rate": 8.646497789272956e-08,
+      "loss": 0.1572,
+      "step": 4027
+    },
+    {
+      "epoch": 1.8373816854829514,
+      "grad_norm": 0.30056819319725037,
+      "learning_rate": 8.598405081206712e-08,
+      "loss": 0.1491,
+      "step": 4028
+    },
+    {
+      "epoch": 1.8378378378378377,
+      "grad_norm": 0.3146818280220032,
+      "learning_rate": 8.550444154307419e-08,
+      "loss": 0.1562,
+      "step": 4029
+    },
+    {
+      "epoch": 1.8382939901927244,
+      "grad_norm": 0.2932460606098175,
+      "learning_rate": 8.502615034756967e-08,
+      "loss": 0.1473,
+      "step": 4030
+    },
+    {
+      "epoch": 1.838750142547611,
+      "grad_norm": 0.2993987202644348,
+      "learning_rate": 8.454917748665303e-08,
+      "loss": 0.1532,
+      "step": 4031
+    },
+    {
+      "epoch": 1.8392062949024974,
+      "grad_norm": 0.31465497612953186,
+      "learning_rate": 8.407352322070323e-08,
+      "loss": 0.1539,
+      "step": 4032
+    },
+    {
+      "epoch": 1.839662447257384,
+      "grad_norm": 0.3242221474647522,
+      "learning_rate": 8.359918780938086e-08,
+      "loss": 0.1569,
+      "step": 4033
+    },
+    {
+      "epoch": 1.8401185996122704,
+      "grad_norm": 0.2993384003639221,
+      "learning_rate": 8.312617151162494e-08,
+      "loss": 0.1448,
+      "step": 4034
+    },
+    {
+      "epoch": 1.8405747519671571,
+      "grad_norm": 0.29778656363487244,
+      "learning_rate": 8.265447458565557e-08,
+      "loss": 0.1507,
+      "step": 4035
+    },
+    {
+      "epoch": 1.8410309043220434,
+      "grad_norm": 0.3276660144329071,
+      "learning_rate": 8.218409728897148e-08,
+      "loss": 0.1638,
+      "step": 4036
+    },
+    {
+      "epoch": 1.8414870566769301,
+      "grad_norm": 0.33212798833847046,
+      "learning_rate": 8.171503987835283e-08,
+      "loss": 0.153,
+      "step": 4037
+    },
+    {
+      "epoch": 1.8419432090318166,
+      "grad_norm": 0.3048456609249115,
+      "learning_rate": 8.12473026098573e-08,
+      "loss": 0.1508,
+      "step": 4038
+    },
+    {
+      "epoch": 1.8423993613867031,
+      "grad_norm": 0.30965879559516907,
+      "learning_rate": 8.078088573882287e-08,
+      "loss": 0.1469,
+      "step": 4039
+    },
+    {
+      "epoch": 1.8428555137415898,
+      "grad_norm": 0.30721354484558105,
+      "learning_rate": 8.031578951986697e-08,
+      "loss": 0.1529,
+      "step": 4040
+    },
+    {
+      "epoch": 1.8433116660964761,
+      "grad_norm": 0.304039865732193,
+      "learning_rate": 7.985201420688538e-08,
+      "loss": 0.1462,
+      "step": 4041
+    },
+    {
+      "epoch": 1.8437678184513628,
+      "grad_norm": 0.3046776354312897,
+      "learning_rate": 7.938956005305254e-08,
+      "loss": 0.1563,
+      "step": 4042
+    },
+    {
+      "epoch": 1.8442239708062493,
+      "grad_norm": 0.31509044766426086,
+      "learning_rate": 7.892842731082346e-08,
+      "loss": 0.1567,
+      "step": 4043
+    },
+    {
+      "epoch": 1.8446801231611358,
+      "grad_norm": 0.3241669535636902,
+      "learning_rate": 7.846861623192953e-08,
+      "loss": 0.1551,
+      "step": 4044
+    },
+    {
+      "epoch": 1.8451362755160223,
+      "grad_norm": 0.29685989022254944,
+      "learning_rate": 7.801012706738248e-08,
+      "loss": 0.1442,
+      "step": 4045
+    },
+    {
+      "epoch": 1.8455924278709088,
+      "grad_norm": 0.3015742301940918,
+      "learning_rate": 7.755296006747099e-08,
+      "loss": 0.1476,
+      "step": 4046
+    },
+    {
+      "epoch": 1.8460485802257955,
+      "grad_norm": 0.3101988434791565,
+      "learning_rate": 7.709711548176318e-08,
+      "loss": 0.1534,
+      "step": 4047
+    },
+    {
+      "epoch": 1.8465047325806818,
+      "grad_norm": 0.2922309935092926,
+      "learning_rate": 7.66425935591042e-08,
+      "loss": 0.1466,
+      "step": 4048
+    },
+    {
+      "epoch": 1.8469608849355685,
+      "grad_norm": 0.3130817711353302,
+      "learning_rate": 7.618939454761775e-08,
+      "loss": 0.1504,
+      "step": 4049
+    },
+    {
+      "epoch": 1.847417037290455,
+      "grad_norm": 0.2975600063800812,
+      "learning_rate": 7.573751869470541e-08,
+      "loss": 0.1469,
+      "step": 4050
+    },
+    {
+      "epoch": 1.8478731896453415,
+      "grad_norm": 0.3103300929069519,
+      "learning_rate": 7.528696624704568e-08,
+      "loss": 0.156,
+      "step": 4051
+    },
+    {
+      "epoch": 1.848329342000228,
+      "grad_norm": 0.32366883754730225,
+      "learning_rate": 7.483773745059569e-08,
+      "loss": 0.1598,
+      "step": 4052
+    },
+    {
+      "epoch": 1.8487854943551145,
+      "grad_norm": 0.31122487783432007,
+      "learning_rate": 7.438983255058984e-08,
+      "loss": 0.1485,
+      "step": 4053
+    },
+    {
+      "epoch": 1.8492416467100012,
+      "grad_norm": 0.3201131224632263,
+      "learning_rate": 7.394325179153893e-08,
+      "loss": 0.1601,
+      "step": 4054
+    },
+    {
+      "epoch": 1.8496977990648875,
+      "grad_norm": 0.3057609796524048,
+      "learning_rate": 7.349799541723184e-08,
+      "loss": 0.1487,
+      "step": 4055
+    },
+    {
+      "epoch": 1.8501539514197742,
+      "grad_norm": 0.2988307476043701,
+      "learning_rate": 7.305406367073387e-08,
+      "loss": 0.1453,
+      "step": 4056
+    },
+    {
+      "epoch": 1.8506101037746607,
+      "grad_norm": 0.30251824855804443,
+      "learning_rate": 7.261145679438752e-08,
+      "loss": 0.1448,
+      "step": 4057
+    },
+    {
+      "epoch": 1.8510662561295472,
+      "grad_norm": 0.31161943078041077,
+      "learning_rate": 7.217017502981206e-08,
+      "loss": 0.1603,
+      "step": 4058
+    },
+    {
+      "epoch": 1.851522408484434,
+      "grad_norm": 0.3044317960739136,
+      "learning_rate": 7.173021861790252e-08,
+      "loss": 0.1491,
+      "step": 4059
+    },
+    {
+      "epoch": 1.8519785608393202,
+      "grad_norm": 0.3090611696243286,
+      "learning_rate": 7.129158779883211e-08,
+      "loss": 0.1455,
+      "step": 4060
+    },
+    {
+      "epoch": 1.852434713194207,
+      "grad_norm": 0.33301857113838196,
+      "learning_rate": 7.085428281204925e-08,
+      "loss": 0.1646,
+      "step": 4061
+    },
+    {
+      "epoch": 1.8528908655490934,
+      "grad_norm": 0.3140386939048767,
+      "learning_rate": 7.041830389627857e-08,
+      "loss": 0.1507,
+      "step": 4062
+    },
+    {
+      "epoch": 1.85334701790398,
+      "grad_norm": 0.3001684546470642,
+      "learning_rate": 6.998365128952078e-08,
+      "loss": 0.1498,
+      "step": 4063
+    },
+    {
+      "epoch": 1.8538031702588664,
+      "grad_norm": 0.3103906810283661,
+      "learning_rate": 6.955032522905331e-08,
+      "loss": 0.1554,
+      "step": 4064
+    },
+    {
+      "epoch": 1.854259322613753,
+      "grad_norm": 0.2982504069805145,
+      "learning_rate": 6.911832595142915e-08,
+      "loss": 0.1456,
+      "step": 4065
+    },
+    {
+      "epoch": 1.8547154749686396,
+      "grad_norm": 0.309812068939209,
+      "learning_rate": 6.868765369247605e-08,
+      "loss": 0.1443,
+      "step": 4066
+    },
+    {
+      "epoch": 1.855171627323526,
+      "grad_norm": 0.3089972734451294,
+      "learning_rate": 6.825830868729816e-08,
+      "loss": 0.1521,
+      "step": 4067
+    },
+    {
+      "epoch": 1.8556277796784126,
+      "grad_norm": 0.3262918293476105,
+      "learning_rate": 6.78302911702755e-08,
+      "loss": 0.1571,
+      "step": 4068
+    },
+    {
+      "epoch": 1.8560839320332991,
+      "grad_norm": 0.2950558066368103,
+      "learning_rate": 6.740360137506258e-08,
+      "loss": 0.1464,
+      "step": 4069
+    },
+    {
+      "epoch": 1.8565400843881856,
+      "grad_norm": 0.30253860354423523,
+      "learning_rate": 6.697823953458943e-08,
+      "loss": 0.1505,
+      "step": 4070
+    },
+    {
+      "epoch": 1.8569962367430723,
+      "grad_norm": 0.29020312428474426,
+      "learning_rate": 6.655420588106115e-08,
+      "loss": 0.1444,
+      "step": 4071
+    },
+    {
+      "epoch": 1.8574523890979586,
+      "grad_norm": 0.322103887796402,
+      "learning_rate": 6.613150064595786e-08,
+      "loss": 0.1553,
+      "step": 4072
+    },
+    {
+      "epoch": 1.8579085414528453,
+      "grad_norm": 0.3157060146331787,
+      "learning_rate": 6.571012406003413e-08,
+      "loss": 0.156,
+      "step": 4073
+    },
+    {
+      "epoch": 1.8583646938077318,
+      "grad_norm": 0.3270284831523895,
+      "learning_rate": 6.529007635331958e-08,
+      "loss": 0.1558,
+      "step": 4074
+    },
+    {
+      "epoch": 1.8588208461626183,
+      "grad_norm": 0.3166934847831726,
+      "learning_rate": 6.487135775511854e-08,
+      "loss": 0.1527,
+      "step": 4075
+    },
+    {
+      "epoch": 1.8592769985175048,
+      "grad_norm": 0.3026927709579468,
+      "learning_rate": 6.445396849400987e-08,
+      "loss": 0.1438,
+      "step": 4076
+    },
+    {
+      "epoch": 1.8597331508723913,
+      "grad_norm": 0.31397131085395813,
+      "learning_rate": 6.403790879784572e-08,
+      "loss": 0.1515,
+      "step": 4077
+    },
+    {
+      "epoch": 1.860189303227278,
+      "grad_norm": 0.3106239438056946,
+      "learning_rate": 6.362317889375357e-08,
+      "loss": 0.1545,
+      "step": 4078
+    },
+    {
+      "epoch": 1.8606454555821643,
+      "grad_norm": 0.29761916399002075,
+      "learning_rate": 6.32097790081343e-08,
+      "loss": 0.1424,
+      "step": 4079
+    },
+    {
+      "epoch": 1.861101607937051,
+      "grad_norm": 0.2998462915420532,
+      "learning_rate": 6.27977093666629e-08,
+      "loss": 0.1453,
+      "step": 4080
+    },
+    {
+      "epoch": 1.8615577602919375,
+      "grad_norm": 0.3046252727508545,
+      "learning_rate": 6.238697019428863e-08,
+      "loss": 0.1476,
+      "step": 4081
+    },
+    {
+      "epoch": 1.862013912646824,
+      "grad_norm": 0.3134392499923706,
+      "learning_rate": 6.197756171523378e-08,
+      "loss": 0.1495,
+      "step": 4082
+    },
+    {
+      "epoch": 1.8624700650017105,
+      "grad_norm": 0.30088722705841064,
+      "learning_rate": 6.15694841529943e-08,
+      "loss": 0.1473,
+      "step": 4083
+    },
+    {
+      "epoch": 1.862926217356597,
+      "grad_norm": 0.32777875661849976,
+      "learning_rate": 6.11627377303406e-08,
+      "loss": 0.1593,
+      "step": 4084
+    },
+    {
+      "epoch": 1.8633823697114837,
+      "grad_norm": 0.2943339943885803,
+      "learning_rate": 6.075732266931478e-08,
+      "loss": 0.1485,
+      "step": 4085
+    },
+    {
+      "epoch": 1.86383852206637,
+      "grad_norm": 0.30925142765045166,
+      "learning_rate": 6.035323919123343e-08,
+      "loss": 0.1526,
+      "step": 4086
+    },
+    {
+      "epoch": 1.8642946744212567,
+      "grad_norm": 0.30068930983543396,
+      "learning_rate": 5.995048751668563e-08,
+      "loss": 0.1442,
+      "step": 4087
+    },
+    {
+      "epoch": 1.8647508267761432,
+      "grad_norm": 0.29883620142936707,
+      "learning_rate": 5.954906786553361e-08,
+      "loss": 0.1489,
+      "step": 4088
+    },
+    {
+      "epoch": 1.8652069791310297,
+      "grad_norm": 0.3294494152069092,
+      "learning_rate": 5.9148980456912305e-08,
+      "loss": 0.1556,
+      "step": 4089
+    },
+    {
+      "epoch": 1.8656631314859164,
+      "grad_norm": 0.3055429458618164,
+      "learning_rate": 5.875022550922954e-08,
+      "loss": 0.1443,
+      "step": 4090
+    },
+    {
+      "epoch": 1.8661192838408027,
+      "grad_norm": 0.3181114196777344,
+      "learning_rate": 5.83528032401659e-08,
+      "loss": 0.1528,
+      "step": 4091
+    },
+    {
+      "epoch": 1.8665754361956894,
+      "grad_norm": 0.3127560019493103,
+      "learning_rate": 5.795671386667423e-08,
+      "loss": 0.147,
+      "step": 4092
+    },
+    {
+      "epoch": 1.867031588550576,
+      "grad_norm": 0.30426082015037537,
+      "learning_rate": 5.756195760497907e-08,
+      "loss": 0.1509,
+      "step": 4093
+    },
+    {
+      "epoch": 1.8674877409054624,
+      "grad_norm": 0.30323028564453125,
+      "learning_rate": 5.716853467057915e-08,
+      "loss": 0.1585,
+      "step": 4094
+    },
+    {
+      "epoch": 1.867943893260349,
+      "grad_norm": 0.312921941280365,
+      "learning_rate": 5.677644527824322e-08,
+      "loss": 0.1534,
+      "step": 4095
+    },
+    {
+      "epoch": 1.8684000456152354,
+      "grad_norm": 0.31728360056877136,
+      "learning_rate": 5.638568964201313e-08,
+      "loss": 0.1582,
+      "step": 4096
+    },
+    {
+      "epoch": 1.8688561979701221,
+      "grad_norm": 0.3177885413169861,
+      "learning_rate": 5.599626797520241e-08,
+      "loss": 0.149,
+      "step": 4097
+    },
+    {
+      "epoch": 1.8693123503250084,
+      "grad_norm": 0.30900123715400696,
+      "learning_rate": 5.560818049039657e-08,
+      "loss": 0.1538,
+      "step": 4098
+    },
+    {
+      "epoch": 1.8697685026798951,
+      "grad_norm": 0.3214104175567627,
+      "learning_rate": 5.52214273994528e-08,
+      "loss": 0.1688,
+      "step": 4099
+    },
+    {
+      "epoch": 1.8702246550347816,
+      "grad_norm": 0.31845441460609436,
+      "learning_rate": 5.4836008913499164e-08,
+      "loss": 0.1598,
+      "step": 4100
+    },
+    {
+      "epoch": 1.8706808073896681,
+      "grad_norm": 0.3014170229434967,
+      "learning_rate": 5.445192524293597e-08,
+      "loss": 0.1457,
+      "step": 4101
+    },
+    {
+      "epoch": 1.8711369597445546,
+      "grad_norm": 0.3255158066749573,
+      "learning_rate": 5.406917659743466e-08,
+      "loss": 0.1565,
+      "step": 4102
+    },
+    {
+      "epoch": 1.8715931120994411,
+      "grad_norm": 0.31474485993385315,
+      "learning_rate": 5.368776318593727e-08,
+      "loss": 0.1608,
+      "step": 4103
+    },
+    {
+      "epoch": 1.8720492644543278,
+      "grad_norm": 0.30476266145706177,
+      "learning_rate": 5.33076852166578e-08,
+      "loss": 0.1467,
+      "step": 4104
+    },
+    {
+      "epoch": 1.8725054168092141,
+      "grad_norm": 0.30424967408180237,
+      "learning_rate": 5.292894289708111e-08,
+      "loss": 0.1561,
+      "step": 4105
+    },
+    {
+      "epoch": 1.8729615691641008,
+      "grad_norm": 0.30015653371810913,
+      "learning_rate": 5.255153643396266e-08,
+      "loss": 0.1441,
+      "step": 4106
+    },
+    {
+      "epoch": 1.8734177215189873,
+      "grad_norm": 0.31790515780448914,
+      "learning_rate": 5.2175466033328184e-08,
+      "loss": 0.1497,
+      "step": 4107
+    },
+    {
+      "epoch": 1.8738738738738738,
+      "grad_norm": 0.31600046157836914,
+      "learning_rate": 5.1800731900474855e-08,
+      "loss": 0.156,
+      "step": 4108
+    },
+    {
+      "epoch": 1.8743300262287605,
+      "grad_norm": 0.3092789053916931,
+      "learning_rate": 5.142733423997043e-08,
+      "loss": 0.1496,
+      "step": 4109
+    },
+    {
+      "epoch": 1.8747861785836468,
+      "grad_norm": 0.31084370613098145,
+      "learning_rate": 5.1055273255652127e-08,
+      "loss": 0.1536,
+      "step": 4110
+    },
+    {
+      "epoch": 1.8752423309385335,
+      "grad_norm": 0.32189297676086426,
+      "learning_rate": 5.068454915062832e-08,
+      "loss": 0.1526,
+      "step": 4111
+    },
+    {
+      "epoch": 1.87569848329342,
+      "grad_norm": 0.3167175352573395,
+      "learning_rate": 5.0315162127277384e-08,
+      "loss": 0.151,
+      "step": 4112
+    },
+    {
+      "epoch": 1.8761546356483065,
+      "grad_norm": 0.29758134484291077,
+      "learning_rate": 4.994711238724803e-08,
+      "loss": 0.1482,
+      "step": 4113
+    },
+    {
+      "epoch": 1.876610788003193,
+      "grad_norm": 0.3226402997970581,
+      "learning_rate": 4.958040013145815e-08,
+      "loss": 0.1553,
+      "step": 4114
+    },
+    {
+      "epoch": 1.8770669403580795,
+      "grad_norm": 0.32380685210227966,
+      "learning_rate": 4.92150255600965e-08,
+      "loss": 0.159,
+      "step": 4115
+    },
+    {
+      "epoch": 1.8775230927129662,
+      "grad_norm": 0.3136303424835205,
+      "learning_rate": 4.885098887262074e-08,
+      "loss": 0.154,
+      "step": 4116
+    },
+    {
+      "epoch": 1.8779792450678525,
+      "grad_norm": 0.3161921501159668,
+      "learning_rate": 4.8488290267758864e-08,
+      "loss": 0.1491,
+      "step": 4117
+    },
+    {
+      "epoch": 1.8784353974227392,
+      "grad_norm": 0.28935566544532776,
+      "learning_rate": 4.812692994350804e-08,
+      "loss": 0.1457,
+      "step": 4118
+    },
+    {
+      "epoch": 1.8788915497776257,
+      "grad_norm": 0.32881560921669006,
+      "learning_rate": 4.776690809713463e-08,
+      "loss": 0.1586,
+      "step": 4119
+    },
+    {
+      "epoch": 1.8793477021325122,
+      "grad_norm": 0.31737688183784485,
+      "learning_rate": 4.7408224925174776e-08,
+      "loss": 0.1512,
+      "step": 4120
+    },
+    {
+      "epoch": 1.879803854487399,
+      "grad_norm": 0.2896677553653717,
+      "learning_rate": 4.705088062343405e-08,
+      "loss": 0.1439,
+      "step": 4121
+    },
+    {
+      "epoch": 1.8802600068422852,
+      "grad_norm": 0.29725193977355957,
+      "learning_rate": 4.669487538698614e-08,
+      "loss": 0.1434,
+      "step": 4122
+    },
+    {
+      "epoch": 1.880716159197172,
+      "grad_norm": 0.31548237800598145,
+      "learning_rate": 4.634020941017475e-08,
+      "loss": 0.1529,
+      "step": 4123
+    },
+    {
+      "epoch": 1.8811723115520584,
+      "grad_norm": 0.3144434094429016,
+      "learning_rate": 4.598688288661196e-08,
+      "loss": 0.153,
+      "step": 4124
+    },
+    {
+      "epoch": 1.881628463906945,
+      "grad_norm": 0.30898335576057434,
+      "learning_rate": 4.563489600917875e-08,
+      "loss": 0.1518,
+      "step": 4125
+    },
+    {
+      "epoch": 1.8820846162618314,
+      "grad_norm": 0.3203905522823334,
+      "learning_rate": 4.528424897002503e-08,
+      "loss": 0.151,
+      "step": 4126
+    },
+    {
+      "epoch": 1.882540768616718,
+      "grad_norm": 0.3028983175754547,
+      "learning_rate": 4.4934941960568525e-08,
+      "loss": 0.1499,
+      "step": 4127
+    },
+    {
+      "epoch": 1.8829969209716046,
+      "grad_norm": 0.3071635663509369,
+      "learning_rate": 4.4586975171496686e-08,
+      "loss": 0.1498,
+      "step": 4128
+    },
+    {
+      "epoch": 1.883453073326491,
+      "grad_norm": 0.3218410611152649,
+      "learning_rate": 4.424034879276423e-08,
+      "loss": 0.1524,
+      "step": 4129
+    },
+    {
+      "epoch": 1.8839092256813776,
+      "grad_norm": 0.31365498900413513,
+      "learning_rate": 4.389506301359453e-08,
+      "loss": 0.1432,
+      "step": 4130
+    },
+    {
+      "epoch": 1.8843653780362641,
+      "grad_norm": 0.31670045852661133,
+      "learning_rate": 4.3551118022479575e-08,
+      "loss": 0.1591,
+      "step": 4131
+    },
+    {
+      "epoch": 1.8848215303911506,
+      "grad_norm": 0.30101335048675537,
+      "learning_rate": 4.320851400717835e-08,
+      "loss": 0.1439,
+      "step": 4132
+    },
+    {
+      "epoch": 1.8852776827460371,
+      "grad_norm": 0.31160426139831543,
+      "learning_rate": 4.2867251154718456e-08,
+      "loss": 0.1558,
+      "step": 4133
+    },
+    {
+      "epoch": 1.8857338351009236,
+      "grad_norm": 0.30103379487991333,
+      "learning_rate": 4.252732965139589e-08,
+      "loss": 0.1531,
+      "step": 4134
+    },
+    {
+      "epoch": 1.8861899874558103,
+      "grad_norm": 0.30498963594436646,
+      "learning_rate": 4.218874968277387e-08,
+      "loss": 0.1538,
+      "step": 4135
+    },
+    {
+      "epoch": 1.8866461398106966,
+      "grad_norm": 0.3057640790939331,
+      "learning_rate": 4.18515114336826e-08,
+      "loss": 0.1462,
+      "step": 4136
+    },
+    {
+      "epoch": 1.8871022921655833,
+      "grad_norm": 0.3127767741680145,
+      "learning_rate": 4.151561508822122e-08,
+      "loss": 0.1452,
+      "step": 4137
+    },
+    {
+      "epoch": 1.8875584445204698,
+      "grad_norm": 0.31729856133461,
+      "learning_rate": 4.118106082975498e-08,
+      "loss": 0.153,
+      "step": 4138
+    },
+    {
+      "epoch": 1.8880145968753563,
+      "grad_norm": 0.31838279962539673,
+      "learning_rate": 4.084784884091725e-08,
+      "loss": 0.1599,
+      "step": 4139
+    },
+    {
+      "epoch": 1.888470749230243,
+      "grad_norm": 0.30022817850112915,
+      "learning_rate": 4.05159793036089e-08,
+      "loss": 0.1454,
+      "step": 4140
+    },
+    {
+      "epoch": 1.8889269015851293,
+      "grad_norm": 0.31650739908218384,
+      "learning_rate": 4.018545239899696e-08,
+      "loss": 0.1573,
+      "step": 4141
+    },
+    {
+      "epoch": 1.889383053940016,
+      "grad_norm": 0.3004850745201111,
+      "learning_rate": 3.9856268307516545e-08,
+      "loss": 0.1446,
+      "step": 4142
+    },
+    {
+      "epoch": 1.8898392062949025,
+      "grad_norm": 0.32467272877693176,
+      "learning_rate": 3.952842720886918e-08,
+      "loss": 0.1515,
+      "step": 4143
+    },
+    {
+      "epoch": 1.890295358649789,
+      "grad_norm": 0.31169238686561584,
+      "learning_rate": 3.9201929282023375e-08,
+      "loss": 0.1546,
+      "step": 4144
+    },
+    {
+      "epoch": 1.8907515110046755,
+      "grad_norm": 0.32125574350357056,
+      "learning_rate": 3.887677470521434e-08,
+      "loss": 0.1603,
+      "step": 4145
+    },
+    {
+      "epoch": 1.891207663359562,
+      "grad_norm": 0.3054753243923187,
+      "learning_rate": 3.855296365594424e-08,
+      "loss": 0.1548,
+      "step": 4146
+    },
+    {
+      "epoch": 1.8916638157144487,
+      "grad_norm": 0.31254300475120544,
+      "learning_rate": 3.8230496310981415e-08,
+      "loss": 0.1464,
+      "step": 4147
+    },
+    {
+      "epoch": 1.892119968069335,
+      "grad_norm": 0.3083164691925049,
+      "learning_rate": 3.7909372846360593e-08,
+      "loss": 0.1483,
+      "step": 4148
+    },
+    {
+      "epoch": 1.8925761204242217,
+      "grad_norm": 0.32612234354019165,
+      "learning_rate": 3.758959343738322e-08,
+      "loss": 0.157,
+      "step": 4149
+    },
+    {
+      "epoch": 1.8930322727791082,
+      "grad_norm": 0.3106308579444885,
+      "learning_rate": 3.727115825861744e-08,
+      "loss": 0.1602,
+      "step": 4150
+    },
+    {
+      "epoch": 1.8934884251339947,
+      "grad_norm": 0.3090287148952484,
+      "learning_rate": 3.69540674838964e-08,
+      "loss": 0.1533,
+      "step": 4151
+    },
+    {
+      "epoch": 1.8939445774888812,
+      "grad_norm": 0.3133898675441742,
+      "learning_rate": 3.663832128632028e-08,
+      "loss": 0.1485,
+      "step": 4152
+    },
+    {
+      "epoch": 1.8944007298437677,
+      "grad_norm": 0.3172975480556488,
+      "learning_rate": 3.6323919838255064e-08,
+      "loss": 0.1444,
+      "step": 4153
+    },
+    {
+      "epoch": 1.8948568821986544,
+      "grad_norm": 0.2908395826816559,
+      "learning_rate": 3.6010863311332634e-08,
+      "loss": 0.1464,
+      "step": 4154
+    },
+    {
+      "epoch": 1.8953130345535407,
+      "grad_norm": 0.3151981234550476,
+      "learning_rate": 3.56991518764499e-08,
+      "loss": 0.1626,
+      "step": 4155
+    },
+    {
+      "epoch": 1.8957691869084274,
+      "grad_norm": 0.31290143728256226,
+      "learning_rate": 3.53887857037713e-08,
+      "loss": 0.158,
+      "step": 4156
+    },
+    {
+      "epoch": 1.896225339263314,
+      "grad_norm": 0.28666338324546814,
+      "learning_rate": 3.5079764962724636e-08,
+      "loss": 0.1421,
+      "step": 4157
+    },
+    {
+      "epoch": 1.8966814916182004,
+      "grad_norm": 0.30781060457229614,
+      "learning_rate": 3.477208982200525e-08,
+      "loss": 0.1499,
+      "step": 4158
+    },
+    {
+      "epoch": 1.8971376439730872,
+      "grad_norm": 0.29811081290245056,
+      "learning_rate": 3.446576044957267e-08,
+      "loss": 0.1518,
+      "step": 4159
+    },
+    {
+      "epoch": 1.8975937963279734,
+      "grad_norm": 0.30607905983924866,
+      "learning_rate": 3.416077701265203e-08,
+      "loss": 0.1496,
+      "step": 4160
+    },
+    {
+      "epoch": 1.8980499486828601,
+      "grad_norm": 0.3152986466884613,
+      "learning_rate": 3.38571396777343e-08,
+      "loss": 0.1523,
+      "step": 4161
+    },
+    {
+      "epoch": 1.8985061010377466,
+      "grad_norm": 0.31666451692581177,
+      "learning_rate": 3.355484861057468e-08,
+      "loss": 0.1501,
+      "step": 4162
+    },
+    {
+      "epoch": 1.8989622533926331,
+      "grad_norm": 0.30806514620780945,
+      "learning_rate": 3.325390397619394e-08,
+      "loss": 0.1459,
+      "step": 4163
+    },
+    {
+      "epoch": 1.8994184057475196,
+      "grad_norm": 0.3030230700969696,
+      "learning_rate": 3.295430593887788e-08,
+      "loss": 0.1518,
+      "step": 4164
+    },
+    {
+      "epoch": 1.8998745581024061,
+      "grad_norm": 0.30199679732322693,
+      "learning_rate": 3.265605466217736e-08,
+      "loss": 0.1539,
+      "step": 4165
+    },
+    {
+      "epoch": 1.9003307104572928,
+      "grad_norm": 0.319665789604187,
+      "learning_rate": 3.2359150308907404e-08,
+      "loss": 0.1552,
+      "step": 4166
+    },
+    {
+      "epoch": 1.9007868628121791,
+      "grad_norm": 0.29318103194236755,
+      "learning_rate": 3.206359304114864e-08,
+      "loss": 0.1465,
+      "step": 4167
+    },
+    {
+      "epoch": 1.9012430151670658,
+      "grad_norm": 0.28794869780540466,
+      "learning_rate": 3.176938302024535e-08,
+      "loss": 0.1403,
+      "step": 4168
+    },
+    {
+      "epoch": 1.9016991675219523,
+      "grad_norm": 0.3013702929019928,
+      "learning_rate": 3.14765204068071e-08,
+      "loss": 0.15,
+      "step": 4169
+    },
+    {
+      "epoch": 1.9021553198768388,
+      "grad_norm": 0.30245229601860046,
+      "learning_rate": 3.11850053607074e-08,
+      "loss": 0.1568,
+      "step": 4170
+    },
+    {
+      "epoch": 1.9026114722317256,
+      "grad_norm": 0.3028806746006012,
+      "learning_rate": 3.089483804108451e-08,
+      "loss": 0.1512,
+      "step": 4171
+    },
+    {
+      "epoch": 1.9030676245866118,
+      "grad_norm": 0.30867111682891846,
+      "learning_rate": 3.060601860634088e-08,
+      "loss": 0.1465,
+      "step": 4172
+    },
+    {
+      "epoch": 1.9035237769414985,
+      "grad_norm": 0.30924585461616516,
+      "learning_rate": 3.03185472141429e-08,
+      "loss": 0.1519,
+      "step": 4173
+    },
+    {
+      "epoch": 1.903979929296385,
+      "grad_norm": 0.3047725558280945,
+      "learning_rate": 3.003242402142143e-08,
+      "loss": 0.1464,
+      "step": 4174
+    },
+    {
+      "epoch": 1.9044360816512715,
+      "grad_norm": 0.2993828356266022,
+      "learning_rate": 2.9747649184370952e-08,
+      "loss": 0.1521,
+      "step": 4175
+    },
+    {
+      "epoch": 1.904892234006158,
+      "grad_norm": 0.3048347234725952,
+      "learning_rate": 2.9464222858450465e-08,
+      "loss": 0.1486,
+      "step": 4176
+    },
+    {
+      "epoch": 1.9053483863610445,
+      "grad_norm": 0.3246069550514221,
+      "learning_rate": 2.9182145198382028e-08,
+      "loss": 0.153,
+      "step": 4177
+    },
+    {
+      "epoch": 1.9058045387159313,
+      "grad_norm": 0.31859466433525085,
+      "learning_rate": 2.890141635815219e-08,
+      "loss": 0.1532,
+      "step": 4178
+    },
+    {
+      "epoch": 1.9062606910708175,
+      "grad_norm": 0.31309759616851807,
+      "learning_rate": 2.8622036491010307e-08,
+      "loss": 0.1462,
+      "step": 4179
+    },
+    {
+      "epoch": 1.9067168434257042,
+      "grad_norm": 0.3028302490711212,
+      "learning_rate": 2.834400574947077e-08,
+      "loss": 0.1466,
+      "step": 4180
+    },
+    {
+      "epoch": 1.9071729957805907,
+      "grad_norm": 0.28399717807769775,
+      "learning_rate": 2.806732428530995e-08,
+      "loss": 0.1414,
+      "step": 4181
+    },
+    {
+      "epoch": 1.9076291481354772,
+      "grad_norm": 0.305713951587677,
+      "learning_rate": 2.7791992249568688e-08,
+      "loss": 0.1573,
+      "step": 4182
+    },
+    {
+      "epoch": 1.9080853004903637,
+      "grad_norm": 0.29851067066192627,
+      "learning_rate": 2.751800979255037e-08,
+      "loss": 0.1448,
+      "step": 4183
+    },
+    {
+      "epoch": 1.9085414528452502,
+      "grad_norm": 0.31877046823501587,
+      "learning_rate": 2.72453770638223e-08,
+      "loss": 0.1553,
+      "step": 4184
+    },
+    {
+      "epoch": 1.908997605200137,
+      "grad_norm": 0.31680044531822205,
+      "learning_rate": 2.6974094212214586e-08,
+      "loss": 0.157,
+      "step": 4185
+    },
+    {
+      "epoch": 1.9094537575550232,
+      "grad_norm": 0.3161558508872986,
+      "learning_rate": 2.6704161385820993e-08,
+      "loss": 0.1575,
+      "step": 4186
+    },
+    {
+      "epoch": 1.90990990990991,
+      "grad_norm": 0.32829272747039795,
+      "learning_rate": 2.6435578731997258e-08,
+      "loss": 0.1618,
+      "step": 4187
+    },
+    {
+      "epoch": 1.9103660622647964,
+      "grad_norm": 0.326959490776062,
+      "learning_rate": 2.616834639736332e-08,
+      "loss": 0.16,
+      "step": 4188
+    },
+    {
+      "epoch": 1.910822214619683,
+      "grad_norm": 0.30527809262275696,
+      "learning_rate": 2.5902464527801097e-08,
+      "loss": 0.1491,
+      "step": 4189
+    },
+    {
+      "epoch": 1.9112783669745697,
+      "grad_norm": 0.3215686082839966,
+      "learning_rate": 2.5637933268455316e-08,
+      "loss": 0.1579,
+      "step": 4190
+    },
+    {
+      "epoch": 1.911734519329456,
+      "grad_norm": 0.306034654378891,
+      "learning_rate": 2.5374752763733802e-08,
+      "loss": 0.1514,
+      "step": 4191
+    },
+    {
+      "epoch": 1.9121906716843426,
+      "grad_norm": 0.3113567531108856,
+      "learning_rate": 2.5112923157306902e-08,
+      "loss": 0.1518,
+      "step": 4192
+    },
+    {
+      "epoch": 1.9126468240392291,
+      "grad_norm": 0.30300551652908325,
+      "learning_rate": 2.4852444592107227e-08,
+      "loss": 0.1485,
+      "step": 4193
+    },
+    {
+      "epoch": 1.9131029763941156,
+      "grad_norm": 0.3017410635948181,
+      "learning_rate": 2.459331721033048e-08,
+      "loss": 0.1529,
+      "step": 4194
+    },
+    {
+      "epoch": 1.9135591287490021,
+      "grad_norm": 0.29747387766838074,
+      "learning_rate": 2.4335541153434062e-08,
+      "loss": 0.1488,
+      "step": 4195
+    },
+    {
+      "epoch": 1.9140152811038886,
+      "grad_norm": 0.31298211216926575,
+      "learning_rate": 2.4079116562137906e-08,
+      "loss": 0.1552,
+      "step": 4196
+    },
+    {
+      "epoch": 1.9144714334587754,
+      "grad_norm": 0.3254447877407074,
+      "learning_rate": 2.3824043576424204e-08,
+      "loss": 0.1655,
+      "step": 4197
+    },
+    {
+      "epoch": 1.9149275858136616,
+      "grad_norm": 0.3182838261127472,
+      "learning_rate": 2.3570322335537688e-08,
+      "loss": 0.1545,
+      "step": 4198
+    },
+    {
+      "epoch": 1.9153837381685483,
+      "grad_norm": 0.30687761306762695,
+      "learning_rate": 2.3317952977984504e-08,
+      "loss": 0.1514,
+      "step": 4199
+    },
+    {
+      "epoch": 1.9158398905234348,
+      "grad_norm": 0.3025464713573456,
+      "learning_rate": 2.3066935641533063e-08,
+      "loss": 0.1573,
+      "step": 4200
+    },
+    {
+      "epoch": 1.9162960428783213,
+      "grad_norm": 0.3035159707069397,
+      "learning_rate": 2.2817270463213747e-08,
+      "loss": 0.1448,
+      "step": 4201
+    },
+    {
+      "epoch": 1.916752195233208,
+      "grad_norm": 0.30867886543273926,
+      "learning_rate": 2.2568957579319206e-08,
+      "loss": 0.1485,
+      "step": 4202
+    },
+    {
+      "epoch": 1.9172083475880943,
+      "grad_norm": 0.2988411784172058,
+      "learning_rate": 2.2321997125403227e-08,
+      "loss": 0.1483,
+      "step": 4203
+    },
+    {
+      "epoch": 1.917664499942981,
+      "grad_norm": 0.3118591010570526,
+      "learning_rate": 2.207638923628158e-08,
+      "loss": 0.1518,
+      "step": 4204
+    },
+    {
+      "epoch": 1.9181206522978673,
+      "grad_norm": 0.31566333770751953,
+      "learning_rate": 2.183213404603174e-08,
+      "loss": 0.1523,
+      "step": 4205
+    },
+    {
+      "epoch": 1.918576804652754,
+      "grad_norm": 0.3073529601097107,
+      "learning_rate": 2.1589231687992317e-08,
+      "loss": 0.1508,
+      "step": 4206
+    },
+    {
+      "epoch": 1.9190329570076405,
+      "grad_norm": 0.29325228929519653,
+      "learning_rate": 2.1347682294764195e-08,
+      "loss": 0.1424,
+      "step": 4207
+    },
+    {
+      "epoch": 1.919489109362527,
+      "grad_norm": 0.3260296583175659,
+      "learning_rate": 2.110748599820883e-08,
+      "loss": 0.1602,
+      "step": 4208
+    },
+    {
+      "epoch": 1.9199452617174138,
+      "grad_norm": 0.2806605100631714,
+      "learning_rate": 2.0868642929449678e-08,
+      "loss": 0.1351,
+      "step": 4209
+    },
+    {
+      "epoch": 1.9204014140723,
+      "grad_norm": 0.30611488223075867,
+      "learning_rate": 2.0631153218871047e-08,
+      "loss": 0.1478,
+      "step": 4210
+    },
+    {
+      "epoch": 1.9208575664271867,
+      "grad_norm": 0.32045602798461914,
+      "learning_rate": 2.0395016996118677e-08,
+      "loss": 0.1503,
+      "step": 4211
+    },
+    {
+      "epoch": 1.9213137187820732,
+      "grad_norm": 0.32939472794532776,
+      "learning_rate": 2.0160234390099454e-08,
+      "loss": 0.145,
+      "step": 4212
+    },
+    {
+      "epoch": 1.9217698711369597,
+      "grad_norm": 0.31411582231521606,
+      "learning_rate": 1.992680552898113e-08,
+      "loss": 0.1556,
+      "step": 4213
+    },
+    {
+      "epoch": 1.9222260234918462,
+      "grad_norm": 0.30685755610466003,
+      "learning_rate": 1.9694730540192607e-08,
+      "loss": 0.1509,
+      "step": 4214
+    },
+    {
+      "epoch": 1.9226821758467327,
+      "grad_norm": 0.29877859354019165,
+      "learning_rate": 1.9464009550423656e-08,
+      "loss": 0.1471,
+      "step": 4215
+    },
+    {
+      "epoch": 1.9231383282016195,
+      "grad_norm": 0.2963372766971588,
+      "learning_rate": 1.9234642685624915e-08,
+      "loss": 0.1456,
+      "step": 4216
+    },
+    {
+      "epoch": 1.9235944805565057,
+      "grad_norm": 0.3263612687587738,
+      "learning_rate": 1.9006630071007338e-08,
+      "loss": 0.1594,
+      "step": 4217
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.2938482463359833,
+      "learning_rate": 1.877997183104385e-08,
+      "loss": 0.1438,
+      "step": 4218
+    },
+    {
+      "epoch": 1.924506785266279,
+      "grad_norm": 0.31226950883865356,
+      "learning_rate": 1.8554668089466875e-08,
+      "loss": 0.1476,
+      "step": 4219
+    },
+    {
+      "epoch": 1.9249629376211654,
+      "grad_norm": 0.31848567724227905,
+      "learning_rate": 1.833071896926969e-08,
+      "loss": 0.1507,
+      "step": 4220
+    },
+    {
+      "epoch": 1.9254190899760522,
+      "grad_norm": 0.3016994893550873,
+      "learning_rate": 1.810812459270589e-08,
+      "loss": 0.1446,
+      "step": 4221
+    },
+    {
+      "epoch": 1.9258752423309384,
+      "grad_norm": 0.3061560392379761,
+      "learning_rate": 1.7886885081290496e-08,
+      "loss": 0.1515,
+      "step": 4222
+    },
+    {
+      "epoch": 1.9263313946858251,
+      "grad_norm": 0.3124854266643524,
+      "learning_rate": 1.7667000555797174e-08,
+      "loss": 0.1526,
+      "step": 4223
+    },
+    {
+      "epoch": 1.9267875470407116,
+      "grad_norm": 0.3131293058395386,
+      "learning_rate": 1.744847113626186e-08,
+      "loss": 0.1557,
+      "step": 4224
+    },
+    {
+      "epoch": 1.9272436993955981,
+      "grad_norm": 0.31350550055503845,
+      "learning_rate": 1.723129694197967e-08,
+      "loss": 0.1546,
+      "step": 4225
+    },
+    {
+      "epoch": 1.9276998517504846,
+      "grad_norm": 0.31424471735954285,
+      "learning_rate": 1.701547809150522e-08,
+      "loss": 0.156,
+      "step": 4226
+    },
+    {
+      "epoch": 1.9281560041053711,
+      "grad_norm": 0.3137062191963196,
+      "learning_rate": 1.6801014702655107e-08,
+      "loss": 0.1544,
+      "step": 4227
+    },
+    {
+      "epoch": 1.9286121564602579,
+      "grad_norm": 0.3185259997844696,
+      "learning_rate": 1.658790689250428e-08,
+      "loss": 0.159,
+      "step": 4228
+    },
+    {
+      "epoch": 1.9290683088151441,
+      "grad_norm": 0.3141513168811798,
+      "learning_rate": 1.6376154777388575e-08,
+      "loss": 0.1532,
+      "step": 4229
+    },
+    {
+      "epoch": 1.9295244611700308,
+      "grad_norm": 0.31117168068885803,
+      "learning_rate": 1.6165758472903305e-08,
+      "loss": 0.1497,
+      "step": 4230
+    },
+    {
+      "epoch": 1.9299806135249173,
+      "grad_norm": 0.30958041548728943,
+      "learning_rate": 1.595671809390409e-08,
+      "loss": 0.1481,
+      "step": 4231
+    },
+    {
+      "epoch": 1.9304367658798038,
+      "grad_norm": 0.3209620714187622,
+      "learning_rate": 1.574903375450576e-08,
+      "loss": 0.1516,
+      "step": 4232
+    },
+    {
+      "epoch": 1.9308929182346903,
+      "grad_norm": 0.3040211796760559,
+      "learning_rate": 1.5542705568083737e-08,
+      "loss": 0.1469,
+      "step": 4233
+    },
+    {
+      "epoch": 1.9313490705895768,
+      "grad_norm": 0.31526580452919006,
+      "learning_rate": 1.533773364727209e-08,
+      "loss": 0.1519,
+      "step": 4234
+    },
+    {
+      "epoch": 1.9318052229444636,
+      "grad_norm": 0.32222312688827515,
+      "learning_rate": 1.513411810396548e-08,
+      "loss": 0.1513,
+      "step": 4235
+    },
+    {
+      "epoch": 1.9322613752993498,
+      "grad_norm": 0.3070926070213318,
+      "learning_rate": 1.493185904931721e-08,
+      "loss": 0.1515,
+      "step": 4236
+    },
+    {
+      "epoch": 1.9327175276542365,
+      "grad_norm": 0.3051070272922516,
+      "learning_rate": 1.473095659374063e-08,
+      "loss": 0.1523,
+      "step": 4237
+    },
+    {
+      "epoch": 1.933173680009123,
+      "grad_norm": 0.29667696356773376,
+      "learning_rate": 1.4531410846908845e-08,
+      "loss": 0.1472,
+      "step": 4238
+    },
+    {
+      "epoch": 1.9336298323640095,
+      "grad_norm": 0.3049919009208679,
+      "learning_rate": 1.4333221917753325e-08,
+      "loss": 0.1499,
+      "step": 4239
+    },
+    {
+      "epoch": 1.9340859847188963,
+      "grad_norm": 0.301386296749115,
+      "learning_rate": 1.4136389914466142e-08,
+      "loss": 0.1522,
+      "step": 4240
+    },
+    {
+      "epoch": 1.9345421370737825,
+      "grad_norm": 0.29736459255218506,
+      "learning_rate": 1.3940914944497175e-08,
+      "loss": 0.1437,
+      "step": 4241
+    },
+    {
+      "epoch": 1.9349982894286692,
+      "grad_norm": 0.3111778795719147,
+      "learning_rate": 1.374679711455662e-08,
+      "loss": 0.1518,
+      "step": 4242
+    },
+    {
+      "epoch": 1.9354544417835557,
+      "grad_norm": 0.298543244600296,
+      "learning_rate": 1.3554036530613324e-08,
+      "loss": 0.1421,
+      "step": 4243
+    },
+    {
+      "epoch": 1.9359105941384422,
+      "grad_norm": 0.3185845613479614,
+      "learning_rate": 1.3362633297895334e-08,
+      "loss": 0.1547,
+      "step": 4244
+    },
+    {
+      "epoch": 1.9363667464933287,
+      "grad_norm": 0.33471280336380005,
+      "learning_rate": 1.3172587520889623e-08,
+      "loss": 0.1607,
+      "step": 4245
+    },
+    {
+      "epoch": 1.9368228988482152,
+      "grad_norm": 0.3081263601779938,
+      "learning_rate": 1.298389930334265e-08,
+      "loss": 0.1559,
+      "step": 4246
+    },
+    {
+      "epoch": 1.937279051203102,
+      "grad_norm": 0.2950994372367859,
+      "learning_rate": 1.2796568748258687e-08,
+      "loss": 0.1529,
+      "step": 4247
+    },
+    {
+      "epoch": 1.9377352035579882,
+      "grad_norm": 0.32963621616363525,
+      "learning_rate": 1.2610595957902039e-08,
+      "loss": 0.1566,
+      "step": 4248
+    },
+    {
+      "epoch": 1.938191355912875,
+      "grad_norm": 0.2945248484611511,
+      "learning_rate": 1.242598103379511e-08,
+      "loss": 0.1478,
+      "step": 4249
+    },
+    {
+      "epoch": 1.9386475082677614,
+      "grad_norm": 0.3133743107318878,
+      "learning_rate": 1.2242724076718948e-08,
+      "loss": 0.1527,
+      "step": 4250
+    },
+    {
+      "epoch": 1.939103660622648,
+      "grad_norm": 0.3225638270378113,
+      "learning_rate": 1.2060825186713809e-08,
+      "loss": 0.1551,
+      "step": 4251
+    },
+    {
+      "epoch": 1.9395598129775347,
+      "grad_norm": 0.3212038576602936,
+      "learning_rate": 1.1880284463078596e-08,
+      "loss": 0.157,
+      "step": 4252
+    },
+    {
+      "epoch": 1.940015965332421,
+      "grad_norm": 0.2965296804904938,
+      "learning_rate": 1.1701102004369746e-08,
+      "loss": 0.1556,
+      "step": 4253
+    },
+    {
+      "epoch": 1.9404721176873077,
+      "grad_norm": 0.3178795874118805,
+      "learning_rate": 1.152327790840374e-08,
+      "loss": 0.1489,
+      "step": 4254
+    },
+    {
+      "epoch": 1.9409282700421941,
+      "grad_norm": 0.28629595041275024,
+      "learning_rate": 1.1346812272254592e-08,
+      "loss": 0.1484,
+      "step": 4255
+    },
+    {
+      "epoch": 1.9413844223970806,
+      "grad_norm": 0.3096116781234741,
+      "learning_rate": 1.1171705192254689e-08,
+      "loss": 0.1487,
+      "step": 4256
+    },
+    {
+      "epoch": 1.9418405747519671,
+      "grad_norm": 0.3186762034893036,
+      "learning_rate": 1.0997956763995066e-08,
+      "loss": 0.1545,
+      "step": 4257
+    },
+    {
+      "epoch": 1.9422967271068536,
+      "grad_norm": 0.2991839051246643,
+      "learning_rate": 1.0825567082325405e-08,
+      "loss": 0.1506,
+      "step": 4258
+    },
+    {
+      "epoch": 1.9427528794617404,
+      "grad_norm": 0.3070531487464905,
+      "learning_rate": 1.0654536241352653e-08,
+      "loss": 0.1475,
+      "step": 4259
+    },
+    {
+      "epoch": 1.9432090318166266,
+      "grad_norm": 0.30476951599121094,
+      "learning_rate": 1.0484864334442956e-08,
+      "loss": 0.146,
+      "step": 4260
+    },
+    {
+      "epoch": 1.9436651841715133,
+      "grad_norm": 0.3032197654247284,
+      "learning_rate": 1.0316551454220281e-08,
+      "loss": 0.1462,
+      "step": 4261
+    },
+    {
+      "epoch": 1.9441213365263998,
+      "grad_norm": 0.32814499735832214,
+      "learning_rate": 1.0149597692566405e-08,
+      "loss": 0.1572,
+      "step": 4262
+    },
+    {
+      "epoch": 1.9445774888812863,
+      "grad_norm": 0.30672693252563477,
+      "learning_rate": 9.98400314062148e-09,
+      "loss": 0.1472,
+      "step": 4263
+    },
+    {
+      "epoch": 1.9450336412361728,
+      "grad_norm": 0.30709654092788696,
+      "learning_rate": 9.819767888783749e-09,
+      "loss": 0.1558,
+      "step": 4264
+    },
+    {
+      "epoch": 1.9454897935910593,
+      "grad_norm": 0.3125017583370209,
+      "learning_rate": 9.656892026709275e-09,
+      "loss": 0.161,
+      "step": 4265
+    },
+    {
+      "epoch": 1.945945945945946,
+      "grad_norm": 0.30347585678100586,
+      "learning_rate": 9.495375643311656e-09,
+      "loss": 0.158,
+      "step": 4266
+    },
+    {
+      "epoch": 1.9464020983008323,
+      "grad_norm": 0.3210424482822418,
+      "learning_rate": 9.335218826763137e-09,
+      "loss": 0.1644,
+      "step": 4267
+    },
+    {
+      "epoch": 1.946858250655719,
+      "grad_norm": 0.31803062558174133,
+      "learning_rate": 9.176421664492952e-09,
+      "loss": 0.154,
+      "step": 4268
+    },
+    {
+      "epoch": 1.9473144030106055,
+      "grad_norm": 0.32156792283058167,
+      "learning_rate": 9.018984243188977e-09,
+      "loss": 0.1494,
+      "step": 4269
+    },
+    {
+      "epoch": 1.947770555365492,
+      "grad_norm": 0.3266351819038391,
+      "learning_rate": 8.8629066487958e-09,
+      "loss": 0.1527,
+      "step": 4270
+    },
+    {
+      "epoch": 1.9482267077203788,
+      "grad_norm": 0.29932209849357605,
+      "learning_rate": 8.70818896651665e-09,
+      "loss": 0.1542,
+      "step": 4271
+    },
+    {
+      "epoch": 1.948682860075265,
+      "grad_norm": 0.30628523230552673,
+      "learning_rate": 8.554831280811748e-09,
+      "loss": 0.1575,
+      "step": 4272
+    },
+    {
+      "epoch": 1.9491390124301518,
+      "grad_norm": 0.2996629774570465,
+      "learning_rate": 8.402833675398847e-09,
+      "loss": 0.1462,
+      "step": 4273
+    },
+    {
+      "epoch": 1.9495951647850382,
+      "grad_norm": 0.31015118956565857,
+      "learning_rate": 8.252196233254073e-09,
+      "loss": 0.1486,
+      "step": 4274
+    },
+    {
+      "epoch": 1.9500513171399247,
+      "grad_norm": 0.31504955887794495,
+      "learning_rate": 8.102919036609702e-09,
+      "loss": 0.1598,
+      "step": 4275
+    },
+    {
+      "epoch": 1.9505074694948112,
+      "grad_norm": 0.3145448863506317,
+      "learning_rate": 7.95500216695666e-09,
+      "loss": 0.1599,
+      "step": 4276
+    },
+    {
+      "epoch": 1.9509636218496977,
+      "grad_norm": 0.3210938274860382,
+      "learning_rate": 7.808445705042578e-09,
+      "loss": 0.1593,
+      "step": 4277
+    },
+    {
+      "epoch": 1.9514197742045845,
+      "grad_norm": 0.2968498766422272,
+      "learning_rate": 7.663249730872624e-09,
+      "loss": 0.1436,
+      "step": 4278
+    },
+    {
+      "epoch": 1.9518759265594707,
+      "grad_norm": 0.3029457926750183,
+      "learning_rate": 7.519414323709783e-09,
+      "loss": 0.1512,
+      "step": 4279
+    },
+    {
+      "epoch": 1.9523320789143574,
+      "grad_norm": 0.331482857465744,
+      "learning_rate": 7.3769395620731934e-09,
+      "loss": 0.1568,
+      "step": 4280
+    },
+    {
+      "epoch": 1.952788231269244,
+      "grad_norm": 0.32181012630462646,
+      "learning_rate": 7.235825523740081e-09,
+      "loss": 0.1617,
+      "step": 4281
+    },
+    {
+      "epoch": 1.9532443836241304,
+      "grad_norm": 0.30754491686820984,
+      "learning_rate": 7.09607228574466e-09,
+      "loss": 0.1418,
+      "step": 4282
+    },
+    {
+      "epoch": 1.953700535979017,
+      "grad_norm": 0.2961413562297821,
+      "learning_rate": 6.957679924378124e-09,
+      "loss": 0.1482,
+      "step": 4283
+    },
+    {
+      "epoch": 1.9541566883339034,
+      "grad_norm": 0.3218229115009308,
+      "learning_rate": 6.82064851518921e-09,
+      "loss": 0.158,
+      "step": 4284
+    },
+    {
+      "epoch": 1.9546128406887902,
+      "grad_norm": 0.30200693011283875,
+      "learning_rate": 6.684978132983078e-09,
+      "loss": 0.1524,
+      "step": 4285
+    },
+    {
+      "epoch": 1.9550689930436764,
+      "grad_norm": 0.30520495772361755,
+      "learning_rate": 6.55066885182215e-09,
+      "loss": 0.1525,
+      "step": 4286
+    },
+    {
+      "epoch": 1.9555251453985631,
+      "grad_norm": 0.3005165457725525,
+      "learning_rate": 6.417720745026113e-09,
+      "loss": 0.1491,
+      "step": 4287
+    },
+    {
+      "epoch": 1.9559812977534496,
+      "grad_norm": 0.30482473969459534,
+      "learning_rate": 6.286133885171075e-09,
+      "loss": 0.1537,
+      "step": 4288
+    },
+    {
+      "epoch": 1.9564374501083361,
+      "grad_norm": 0.3127326965332031,
+      "learning_rate": 6.155908344090688e-09,
+      "loss": 0.1518,
+      "step": 4289
+    },
+    {
+      "epoch": 1.9568936024632229,
+      "grad_norm": 0.29569685459136963,
+      "learning_rate": 6.027044192875031e-09,
+      "loss": 0.1478,
+      "step": 4290
+    },
+    {
+      "epoch": 1.9573497548181091,
+      "grad_norm": 0.30033546686172485,
+      "learning_rate": 5.899541501870887e-09,
+      "loss": 0.1424,
+      "step": 4291
+    },
+    {
+      "epoch": 1.9578059071729959,
+      "grad_norm": 0.3073458671569824,
+      "learning_rate": 5.773400340682023e-09,
+      "loss": 0.1527,
+      "step": 4292
+    },
+    {
+      "epoch": 1.9582620595278823,
+      "grad_norm": 0.33385777473449707,
+      "learning_rate": 5.648620778168912e-09,
+      "loss": 0.1599,
+      "step": 4293
+    },
+    {
+      "epoch": 1.9587182118827688,
+      "grad_norm": 0.31177210807800293,
+      "learning_rate": 5.52520288244901e-09,
+      "loss": 0.1589,
+      "step": 4294
+    },
+    {
+      "epoch": 1.9591743642376553,
+      "grad_norm": 0.3075171411037445,
+      "learning_rate": 5.403146720895647e-09,
+      "loss": 0.1499,
+      "step": 4295
+    },
+    {
+      "epoch": 1.9596305165925418,
+      "grad_norm": 0.3147117495536804,
+      "learning_rate": 5.282452360139689e-09,
+      "loss": 0.1639,
+      "step": 4296
+    },
+    {
+      "epoch": 1.9600866689474286,
+      "grad_norm": 0.3060784935951233,
+      "learning_rate": 5.163119866068156e-09,
+      "loss": 0.1481,
+      "step": 4297
+    },
+    {
+      "epoch": 1.9605428213023148,
+      "grad_norm": 0.32398009300231934,
+      "learning_rate": 5.045149303824492e-09,
+      "loss": 0.1591,
+      "step": 4298
+    },
+    {
+      "epoch": 1.9609989736572015,
+      "grad_norm": 0.31675153970718384,
+      "learning_rate": 4.9285407378088515e-09,
+      "loss": 0.1473,
+      "step": 4299
+    },
+    {
+      "epoch": 1.961455126012088,
+      "grad_norm": 0.3197675347328186,
+      "learning_rate": 4.813294231678367e-09,
+      "loss": 0.1648,
+      "step": 4300
+    },
+    {
+      "epoch": 1.9619112783669745,
+      "grad_norm": 0.31326648592948914,
+      "learning_rate": 4.699409848345493e-09,
+      "loss": 0.1513,
+      "step": 4301
+    },
+    {
+      "epoch": 1.9623674307218613,
+      "grad_norm": 0.3160121738910675,
+      "learning_rate": 4.586887649980221e-09,
+      "loss": 0.1568,
+      "step": 4302
+    },
+    {
+      "epoch": 1.9628235830767475,
+      "grad_norm": 0.31106242537498474,
+      "learning_rate": 4.475727698007859e-09,
+      "loss": 0.1505,
+      "step": 4303
+    },
+    {
+      "epoch": 1.9632797354316343,
+      "grad_norm": 0.29899683594703674,
+      "learning_rate": 4.3659300531112555e-09,
+      "loss": 0.15,
+      "step": 4304
+    },
+    {
+      "epoch": 1.9637358877865208,
+      "grad_norm": 0.32101061940193176,
+      "learning_rate": 4.257494775228576e-09,
+      "loss": 0.1528,
+      "step": 4305
+    },
+    {
+      "epoch": 1.9641920401414072,
+      "grad_norm": 0.3137584626674652,
+      "learning_rate": 4.150421923554693e-09,
+      "loss": 0.1553,
+      "step": 4306
+    },
+    {
+      "epoch": 1.9646481924962937,
+      "grad_norm": 0.3097729980945587,
+      "learning_rate": 4.044711556540903e-09,
+      "loss": 0.1479,
+      "step": 4307
+    },
+    {
+      "epoch": 1.9651043448511802,
+      "grad_norm": 0.308292955160141,
+      "learning_rate": 3.9403637318943814e-09,
+      "loss": 0.1518,
+      "step": 4308
+    },
+    {
+      "epoch": 1.965560497206067,
+      "grad_norm": 0.3185825049877167,
+      "learning_rate": 3.837378506578449e-09,
+      "loss": 0.1505,
+      "step": 4309
+    },
+    {
+      "epoch": 1.9660166495609532,
+      "grad_norm": 0.3013981878757477,
+      "learning_rate": 3.735755936813135e-09,
+      "loss": 0.1509,
+      "step": 4310
+    },
+    {
+      "epoch": 1.96647280191584,
+      "grad_norm": 0.31009748578071594,
+      "learning_rate": 3.6354960780740634e-09,
+      "loss": 0.1446,
+      "step": 4311
+    },
+    {
+      "epoch": 1.9669289542707264,
+      "grad_norm": 0.2991821765899658,
+      "learning_rate": 3.536598985092732e-09,
+      "loss": 0.1491,
+      "step": 4312
+    },
+    {
+      "epoch": 1.967385106625613,
+      "grad_norm": 0.30359578132629395,
+      "learning_rate": 3.43906471185762e-09,
+      "loss": 0.1481,
+      "step": 4313
+    },
+    {
+      "epoch": 1.9678412589804994,
+      "grad_norm": 0.327327698469162,
+      "learning_rate": 3.3428933116125273e-09,
+      "loss": 0.1523,
+      "step": 4314
+    },
+    {
+      "epoch": 1.968297411335386,
+      "grad_norm": 0.3296803832054138,
+      "learning_rate": 3.248084836857124e-09,
+      "loss": 0.1593,
+      "step": 4315
+    },
+    {
+      "epoch": 1.9687535636902727,
+      "grad_norm": 0.3021911084651947,
+      "learning_rate": 3.154639339347787e-09,
+      "loss": 0.1507,
+      "step": 4316
+    },
+    {
+      "epoch": 1.969209716045159,
+      "grad_norm": 0.31282705068588257,
+      "learning_rate": 3.0625568700962116e-09,
+      "loss": 0.1536,
+      "step": 4317
+    },
+    {
+      "epoch": 1.9696658684000456,
+      "grad_norm": 0.3006461560726166,
+      "learning_rate": 2.9718374793702433e-09,
+      "loss": 0.1537,
+      "step": 4318
+    },
+    {
+      "epoch": 1.9701220207549321,
+      "grad_norm": 0.32612496614456177,
+      "learning_rate": 2.8824812166938776e-09,
+      "loss": 0.1563,
+      "step": 4319
+    },
+    {
+      "epoch": 1.9705781731098186,
+      "grad_norm": 0.3165525496006012,
+      "learning_rate": 2.7944881308464288e-09,
+      "loss": 0.1482,
+      "step": 4320
+    },
+    {
+      "epoch": 1.9710343254647054,
+      "grad_norm": 0.32714709639549255,
+      "learning_rate": 2.7078582698633615e-09,
+      "loss": 0.1564,
+      "step": 4321
+    },
+    {
+      "epoch": 1.9714904778195916,
+      "grad_norm": 0.3190106153488159,
+      "learning_rate": 2.6225916810357354e-09,
+      "loss": 0.1531,
+      "step": 4322
+    },
+    {
+      "epoch": 1.9719466301744784,
+      "grad_norm": 0.3216022849082947,
+      "learning_rate": 2.5386884109110387e-09,
+      "loss": 0.1475,
+      "step": 4323
+    },
+    {
+      "epoch": 1.9724027825293649,
+      "grad_norm": 0.30688899755477905,
+      "learning_rate": 2.4561485052920774e-09,
+      "loss": 0.1528,
+      "step": 4324
+    },
+    {
+      "epoch": 1.9728589348842513,
+      "grad_norm": 0.30783724784851074,
+      "learning_rate": 2.3749720092369753e-09,
+      "loss": 0.1521,
+      "step": 4325
+    },
+    {
+      "epoch": 1.9733150872391378,
+      "grad_norm": 0.31521403789520264,
+      "learning_rate": 2.2951589670600073e-09,
+      "loss": 0.1498,
+      "step": 4326
+    },
+    {
+      "epoch": 1.9737712395940243,
+      "grad_norm": 0.31150615215301514,
+      "learning_rate": 2.2167094223315978e-09,
+      "loss": 0.15,
+      "step": 4327
+    },
+    {
+      "epoch": 1.974227391948911,
+      "grad_norm": 0.3133440911769867,
+      "learning_rate": 2.1396234178769347e-09,
+      "loss": 0.1585,
+      "step": 4328
+    },
+    {
+      "epoch": 1.9746835443037973,
+      "grad_norm": 0.3090335428714752,
+      "learning_rate": 2.0639009957776344e-09,
+      "loss": 0.1481,
+      "step": 4329
+    },
+    {
+      "epoch": 1.975139696658684,
+      "grad_norm": 0.3068694472312927,
+      "learning_rate": 1.9895421973700756e-09,
+      "loss": 0.1532,
+      "step": 4330
+    },
+    {
+      "epoch": 1.9755958490135705,
+      "grad_norm": 0.31113746762275696,
+      "learning_rate": 1.9165470632473424e-09,
+      "loss": 0.1508,
+      "step": 4331
+    },
+    {
+      "epoch": 1.976052001368457,
+      "grad_norm": 0.3010239601135254,
+      "learning_rate": 1.844915633257005e-09,
+      "loss": 0.1487,
+      "step": 4332
+    },
+    {
+      "epoch": 1.9765081537233435,
+      "grad_norm": 0.305885374546051,
+      "learning_rate": 1.774647946503061e-09,
+      "loss": 0.1504,
+      "step": 4333
+    },
+    {
+      "epoch": 1.97696430607823,
+      "grad_norm": 0.3153265118598938,
+      "learning_rate": 1.7057440413442706e-09,
+      "loss": 0.1596,
+      "step": 4334
+    },
+    {
+      "epoch": 1.9774204584331168,
+      "grad_norm": 0.3111816644668579,
+      "learning_rate": 1.6382039553955453e-09,
+      "loss": 0.1519,
+      "step": 4335
+    },
+    {
+      "epoch": 1.977876610788003,
+      "grad_norm": 0.3076009154319763,
+      "learning_rate": 1.5720277255268368e-09,
+      "loss": 0.1486,
+      "step": 4336
+    },
+    {
+      "epoch": 1.9783327631428897,
+      "grad_norm": 0.32736456394195557,
+      "learning_rate": 1.5072153878639695e-09,
+      "loss": 0.1606,
+      "step": 4337
+    },
+    {
+      "epoch": 1.9787889154977762,
+      "grad_norm": 0.3252892792224884,
+      "learning_rate": 1.4437669777878083e-09,
+      "loss": 0.1605,
+      "step": 4338
+    },
+    {
+      "epoch": 1.9792450678526627,
+      "grad_norm": 0.29461824893951416,
+      "learning_rate": 1.3816825299348136e-09,
+      "loss": 0.1465,
+      "step": 4339
+    },
+    {
+      "epoch": 1.9797012202075495,
+      "grad_norm": 0.31674814224243164,
+      "learning_rate": 1.3209620781973188e-09,
+      "loss": 0.1525,
+      "step": 4340
+    },
+    {
+      "epoch": 1.9801573725624357,
+      "grad_norm": 0.3087109327316284,
+      "learning_rate": 1.2616056557221423e-09,
+      "loss": 0.1542,
+      "step": 4341
+    },
+    {
+      "epoch": 1.9806135249173225,
+      "grad_norm": 0.3171952962875366,
+      "learning_rate": 1.2036132949122535e-09,
+      "loss": 0.1585,
+      "step": 4342
+    },
+    {
+      "epoch": 1.981069677272209,
+      "grad_norm": 0.30287599563598633,
+      "learning_rate": 1.1469850274256621e-09,
+      "loss": 0.1483,
+      "step": 4343
+    },
+    {
+      "epoch": 1.9815258296270954,
+      "grad_norm": 0.3056921064853668,
+      "learning_rate": 1.0917208841756954e-09,
+      "loss": 0.1518,
+      "step": 4344
+    },
+    {
+      "epoch": 1.981981981981982,
+      "grad_norm": 0.3208737075328827,
+      "learning_rate": 1.037820895330721e-09,
+      "loss": 0.1559,
+      "step": 4345
+    },
+    {
+      "epoch": 1.9824381343368684,
+      "grad_norm": 0.3139932155609131,
+      "learning_rate": 9.852850903155353e-10,
+      "loss": 0.1563,
+      "step": 4346
+    },
+    {
+      "epoch": 1.9828942866917552,
+      "grad_norm": 0.32776904106140137,
+      "learning_rate": 9.341134978085865e-10,
+      "loss": 0.1564,
+      "step": 4347
+    },
+    {
+      "epoch": 1.9833504390466414,
+      "grad_norm": 0.31289342045783997,
+      "learning_rate": 8.843061457450286e-10,
+      "loss": 0.1554,
+      "step": 4348
+    },
+    {
+      "epoch": 1.9838065914015282,
+      "grad_norm": 0.3432624638080597,
+      "learning_rate": 8.358630613145014e-10,
+      "loss": 0.1655,
+      "step": 4349
+    },
+    {
+      "epoch": 1.9842627437564146,
+      "grad_norm": 0.30396798253059387,
+      "learning_rate": 7.887842709622395e-10,
+      "loss": 0.1423,
+      "step": 4350
+    },
+    {
+      "epoch": 1.9847188961113011,
+      "grad_norm": 0.3127341866493225,
+      "learning_rate": 7.430698003882409e-10,
+      "loss": 0.1525,
+      "step": 4351
+    },
+    {
+      "epoch": 1.9851750484661879,
+      "grad_norm": 0.2993481755256653,
+      "learning_rate": 6.98719674548376e-10,
+      "loss": 0.1383,
+      "step": 4352
+    },
+    {
+      "epoch": 1.9856312008210741,
+      "grad_norm": 0.2947846055030823,
+      "learning_rate": 6.557339176530009e-10,
+      "loss": 0.1449,
+      "step": 4353
+    },
+    {
+      "epoch": 1.9860873531759609,
+      "grad_norm": 0.2962585389614105,
+      "learning_rate": 6.141125531686221e-10,
+      "loss": 0.1488,
+      "step": 4354
+    },
+    {
+      "epoch": 1.9865435055308474,
+      "grad_norm": 0.30033570528030396,
+      "learning_rate": 5.738556038159537e-10,
+      "loss": 0.1483,
+      "step": 4355
+    },
+    {
+      "epoch": 1.9869996578857338,
+      "grad_norm": 0.31666630506515503,
+      "learning_rate": 5.349630915710279e-10,
+      "loss": 0.154,
+      "step": 4356
+    },
+    {
+      "epoch": 1.9874558102406203,
+      "grad_norm": 0.32846659421920776,
+      "learning_rate": 4.974350376660275e-10,
+      "loss": 0.159,
+      "step": 4357
+    },
+    {
+      "epoch": 1.9879119625955068,
+      "grad_norm": 0.31118103861808777,
+      "learning_rate": 4.612714625867876e-10,
+      "loss": 0.152,
+      "step": 4358
+    },
+    {
+      "epoch": 1.9883681149503936,
+      "grad_norm": 0.3125511705875397,
+      "learning_rate": 4.2647238607529396e-10,
+      "loss": 0.1448,
+      "step": 4359
+    },
+    {
+      "epoch": 1.9888242673052798,
+      "grad_norm": 0.31213322281837463,
+      "learning_rate": 3.930378271282953e-10,
+      "loss": 0.1559,
+      "step": 4360
+    },
+    {
+      "epoch": 1.9892804196601666,
+      "grad_norm": 0.306888222694397,
+      "learning_rate": 3.609678039981357e-10,
+      "loss": 0.1505,
+      "step": 4361
+    },
+    {
+      "epoch": 1.989736572015053,
+      "grad_norm": 0.30274081230163574,
+      "learning_rate": 3.3026233419108934e-10,
+      "loss": 0.1474,
+      "step": 4362
+    },
+    {
+      "epoch": 1.9901927243699395,
+      "grad_norm": 0.32113727927207947,
+      "learning_rate": 3.0092143447013613e-10,
+      "loss": 0.1597,
+      "step": 4363
+    },
+    {
+      "epoch": 1.990648876724826,
+      "grad_norm": 0.3094944655895233,
+      "learning_rate": 2.7294512085163093e-10,
+      "loss": 0.1462,
+      "step": 4364
+    },
+    {
+      "epoch": 1.9911050290797125,
+      "grad_norm": 0.3215519189834595,
+      "learning_rate": 2.463334086086344e-10,
+      "loss": 0.157,
+      "step": 4365
+    },
+    {
+      "epoch": 1.9915611814345993,
+      "grad_norm": 0.3137107789516449,
+      "learning_rate": 2.210863122678597e-10,
+      "loss": 0.1556,
+      "step": 4366
+    },
+    {
+      "epoch": 1.9920173337894855,
+      "grad_norm": 0.3107103109359741,
+      "learning_rate": 1.97203845611893e-10,
+      "loss": 0.1531,
+      "step": 4367
+    },
+    {
+      "epoch": 1.9924734861443723,
+      "grad_norm": 0.32824116945266724,
+      "learning_rate": 1.746860216783608e-10,
+      "loss": 0.1536,
+      "step": 4368
+    },
+    {
+      "epoch": 1.9929296384992587,
+      "grad_norm": 0.30260515213012695,
+      "learning_rate": 1.5353285275965246e-10,
+      "loss": 0.1587,
+      "step": 4369
+    },
+    {
+      "epoch": 1.9933857908541452,
+      "grad_norm": 0.3209952414035797,
+      "learning_rate": 1.337443504029201e-10,
+      "loss": 0.1603,
+      "step": 4370
+    },
+    {
+      "epoch": 1.993841943209032,
+      "grad_norm": 0.3102266490459442,
+      "learning_rate": 1.1532052541146644e-10,
+      "loss": 0.1562,
+      "step": 4371
+    },
+    {
+      "epoch": 1.9942980955639182,
+      "grad_norm": 0.3023976981639862,
+      "learning_rate": 9.82613878422467e-11,
+      "loss": 0.1496,
+      "step": 4372
+    },
+    {
+      "epoch": 1.994754247918805,
+      "grad_norm": 0.31487998366355896,
+      "learning_rate": 8.256694700781164e-11,
+      "loss": 0.1577,
+      "step": 4373
+    },
+    {
+      "epoch": 1.9952104002736915,
+      "grad_norm": 0.3100518584251404,
+      "learning_rate": 6.823721147630746e-11,
+      "loss": 0.1486,
+      "step": 4374
+    },
+    {
+      "epoch": 1.995666552628578,
+      "grad_norm": 0.29897409677505493,
+      "learning_rate": 5.527218906981047e-11,
+      "loss": 0.1423,
+      "step": 4375
+    },
+    {
+      "epoch": 1.9961227049834644,
+      "grad_norm": 0.29801782965660095,
+      "learning_rate": 4.3671886866269995e-11,
+      "loss": 0.1488,
+      "step": 4376
+    },
+    {
+      "epoch": 1.996578857338351,
+      "grad_norm": 0.299030065536499,
+      "learning_rate": 3.343631119812063e-11,
+      "loss": 0.1519,
+      "step": 4377
+    },
+    {
+      "epoch": 1.9970350096932377,
+      "grad_norm": 0.3054922819137573,
+      "learning_rate": 2.456546765311485e-11,
+      "loss": 0.1453,
+      "step": 4378
+    },
+    {
+      "epoch": 1.997491162048124,
+      "grad_norm": 0.29340651631355286,
+      "learning_rate": 1.7059361073490377e-11,
+      "loss": 0.148,
+      "step": 4379
+    },
+    {
+      "epoch": 1.9979473144030107,
+      "grad_norm": 0.3030911982059479,
+      "learning_rate": 1.0917995557080397e-11,
+      "loss": 0.1464,
+      "step": 4380
+    },
+    {
+      "epoch": 1.9984034667578972,
+      "grad_norm": 0.2997358739376068,
+      "learning_rate": 6.141374456480886e-12,
+      "loss": 0.1517,
+      "step": 4381
+    },
+    {
+      "epoch": 1.9988596191127836,
+      "grad_norm": 0.3059263825416565,
+      "learning_rate": 2.729500379328176e-12,
+      "loss": 0.1491,
+      "step": 4382
+    },
+    {
+      "epoch": 1.9993157714676704,
+      "grad_norm": 0.3209269344806671,
+      "learning_rate": 6.823751880213892e-13,
+      "loss": 0.161,
+      "step": 4383
+    },
+    {
+      "epoch": 1.9997719238225566,
+      "grad_norm": 0.3166249990463257,
+      "learning_rate": 0.0,
+      "loss": 0.1613,
+      "step": 4384
+    },
+    {
+      "epoch": 1.9997719238225566,
+      "step": 4384,
+      "total_flos": 1.8679640179110576e+19,
+      "train_loss": 0.08417626275565393,
+      "train_runtime": 10353.7265,
+      "train_samples_per_second": 54.203,
+      "train_steps_per_second": 0.423
+    }
+  ],
+  "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": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.8679640179110576e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}