diff --git "a/qwen3-vl-4b-agentnet-bsz384-step3922/trainer_state.json" "b/qwen3-vl-4b-agentnet-bsz384-step3922/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/qwen3-vl-4b-agentnet-bsz384-step3922/trainer_state.json"
@@ -0,0 +1,27488 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 10000.0,
+  "global_step": 3922,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00025497195308516065,
+      "grad_norm": 14.721250534057617,
+      "learning_rate": 1.0152284263959391e-07,
+      "loss": 1.1753778457641602,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005099439061703213,
+      "grad_norm": 14.536355018615723,
+      "learning_rate": 2.0304568527918783e-07,
+      "loss": 1.176823377609253,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007649158592554819,
+      "grad_norm": 14.321669578552246,
+      "learning_rate": 3.0456852791878176e-07,
+      "loss": 1.1615561246871948,
+      "step": 3
+    },
+    {
+      "epoch": 0.0010198878123406426,
+      "grad_norm": 14.656594276428223,
+      "learning_rate": 4.0609137055837566e-07,
+      "loss": 1.1877045631408691,
+      "step": 4
+    },
+    {
+      "epoch": 0.0012748597654258032,
+      "grad_norm": 14.882043838500977,
+      "learning_rate": 5.076142131979696e-07,
+      "loss": 1.1730620861053467,
+      "step": 5
+    },
+    {
+      "epoch": 0.0015298317185109638,
+      "grad_norm": 14.535433769226074,
+      "learning_rate": 6.091370558375635e-07,
+      "loss": 1.1784110069274902,
+      "step": 6
+    },
+    {
+      "epoch": 0.0017848036715961244,
+      "grad_norm": 14.390504837036133,
+      "learning_rate": 7.106598984771574e-07,
+      "loss": 1.169837474822998,
+      "step": 7
+    },
+    {
+      "epoch": 0.002039775624681285,
+      "grad_norm": 14.107206344604492,
+      "learning_rate": 8.121827411167513e-07,
+      "loss": 1.1590073108673096,
+      "step": 8
+    },
+    {
+      "epoch": 0.0022947475777664456,
+      "grad_norm": 14.10217571258545,
+      "learning_rate": 9.137055837563452e-07,
+      "loss": 1.1481646299362183,
+      "step": 9
+    },
+    {
+      "epoch": 0.0025497195308516064,
+      "grad_norm": 13.8338623046875,
+      "learning_rate": 1.0152284263959392e-06,
+      "loss": 1.1406148672103882,
+      "step": 10
+    },
+    {
+      "epoch": 0.0028046914839367667,
+      "grad_norm": 12.136486053466797,
+      "learning_rate": 1.116751269035533e-06,
+      "loss": 1.0856387615203857,
+      "step": 11
+    },
+    {
+      "epoch": 0.0030596634370219276,
+      "grad_norm": 11.842436790466309,
+      "learning_rate": 1.218274111675127e-06,
+      "loss": 1.0627188682556152,
+      "step": 12
+    },
+    {
+      "epoch": 0.0033146353901070884,
+      "grad_norm": 11.395700454711914,
+      "learning_rate": 1.319796954314721e-06,
+      "loss": 1.0452030897140503,
+      "step": 13
+    },
+    {
+      "epoch": 0.0035696073431922487,
+      "grad_norm": 7.274648666381836,
+      "learning_rate": 1.4213197969543148e-06,
+      "loss": 0.9214948415756226,
+      "step": 14
+    },
+    {
+      "epoch": 0.0038245792962774095,
+      "grad_norm": 6.850907325744629,
+      "learning_rate": 1.5228426395939088e-06,
+      "loss": 0.8989903330802917,
+      "step": 15
+    },
+    {
+      "epoch": 0.00407955124936257,
+      "grad_norm": 6.552920818328857,
+      "learning_rate": 1.6243654822335026e-06,
+      "loss": 0.8858917951583862,
+      "step": 16
+    },
+    {
+      "epoch": 0.004334523202447731,
+      "grad_norm": 5.687392711639404,
+      "learning_rate": 1.7258883248730964e-06,
+      "loss": 0.8578486442565918,
+      "step": 17
+    },
+    {
+      "epoch": 0.004589495155532891,
+      "grad_norm": 5.373569965362549,
+      "learning_rate": 1.8274111675126904e-06,
+      "loss": 0.8581478595733643,
+      "step": 18
+    },
+    {
+      "epoch": 0.004844467108618052,
+      "grad_norm": 2.3742711544036865,
+      "learning_rate": 1.9289340101522844e-06,
+      "loss": 0.7883254289627075,
+      "step": 19
+    },
+    {
+      "epoch": 0.005099439061703213,
+      "grad_norm": 1.9187192916870117,
+      "learning_rate": 2.0304568527918785e-06,
+      "loss": 0.7528367042541504,
+      "step": 20
+    },
+    {
+      "epoch": 0.0053544110147883735,
+      "grad_norm": 1.869828701019287,
+      "learning_rate": 2.1319796954314725e-06,
+      "loss": 0.7535017728805542,
+      "step": 21
+    },
+    {
+      "epoch": 0.0056093829678735335,
+      "grad_norm": 1.7534024715423584,
+      "learning_rate": 2.233502538071066e-06,
+      "loss": 0.7501562833786011,
+      "step": 22
+    },
+    {
+      "epoch": 0.005864354920958694,
+      "grad_norm": 1.589906096458435,
+      "learning_rate": 2.33502538071066e-06,
+      "loss": 0.7379674911499023,
+      "step": 23
+    },
+    {
+      "epoch": 0.006119326874043855,
+      "grad_norm": 1.4798340797424316,
+      "learning_rate": 2.436548223350254e-06,
+      "loss": 0.7441307902336121,
+      "step": 24
+    },
+    {
+      "epoch": 0.006374298827129016,
+      "grad_norm": 1.3907454013824463,
+      "learning_rate": 2.5380710659898476e-06,
+      "loss": 0.7364725470542908,
+      "step": 25
+    },
+    {
+      "epoch": 0.006629270780214177,
+      "grad_norm": 1.217887043952942,
+      "learning_rate": 2.639593908629442e-06,
+      "loss": 0.7361563444137573,
+      "step": 26
+    },
+    {
+      "epoch": 0.006884242733299337,
+      "grad_norm": 1.6351842880249023,
+      "learning_rate": 2.7411167512690357e-06,
+      "loss": 0.7294541597366333,
+      "step": 27
+    },
+    {
+      "epoch": 0.0071392146863844975,
+      "grad_norm": 1.6783027648925781,
+      "learning_rate": 2.8426395939086297e-06,
+      "loss": 0.7308826446533203,
+      "step": 28
+    },
+    {
+      "epoch": 0.007394186639469658,
+      "grad_norm": 1.4227818250656128,
+      "learning_rate": 2.9441624365482237e-06,
+      "loss": 0.7099924087524414,
+      "step": 29
+    },
+    {
+      "epoch": 0.007649158592554819,
+      "grad_norm": 1.152917742729187,
+      "learning_rate": 3.0456852791878177e-06,
+      "loss": 0.705260157585144,
+      "step": 30
+    },
+    {
+      "epoch": 0.00790413054563998,
+      "grad_norm": 0.9976029992103577,
+      "learning_rate": 3.1472081218274113e-06,
+      "loss": 0.6943796873092651,
+      "step": 31
+    },
+    {
+      "epoch": 0.00815910249872514,
+      "grad_norm": 0.9264460206031799,
+      "learning_rate": 3.2487309644670053e-06,
+      "loss": 0.6871784925460815,
+      "step": 32
+    },
+    {
+      "epoch": 0.008414074451810302,
+      "grad_norm": 0.8817992806434631,
+      "learning_rate": 3.3502538071065993e-06,
+      "loss": 0.7034109830856323,
+      "step": 33
+    },
+    {
+      "epoch": 0.008669046404895462,
+      "grad_norm": 0.8824509382247925,
+      "learning_rate": 3.451776649746193e-06,
+      "loss": 0.6865764856338501,
+      "step": 34
+    },
+    {
+      "epoch": 0.008924018357980621,
+      "grad_norm": 0.8971969485282898,
+      "learning_rate": 3.5532994923857873e-06,
+      "loss": 0.6814626455307007,
+      "step": 35
+    },
+    {
+      "epoch": 0.009178990311065782,
+      "grad_norm": 0.8440399169921875,
+      "learning_rate": 3.654822335025381e-06,
+      "loss": 0.6798086166381836,
+      "step": 36
+    },
+    {
+      "epoch": 0.009433962264150943,
+      "grad_norm": 0.7719917297363281,
+      "learning_rate": 3.756345177664975e-06,
+      "loss": 0.670407235622406,
+      "step": 37
+    },
+    {
+      "epoch": 0.009688934217236104,
+      "grad_norm": 0.7223917841911316,
+      "learning_rate": 3.857868020304569e-06,
+      "loss": 0.6843222379684448,
+      "step": 38
+    },
+    {
+      "epoch": 0.009943906170321265,
+      "grad_norm": 0.8157918453216553,
+      "learning_rate": 3.959390862944163e-06,
+      "loss": 0.6840213537216187,
+      "step": 39
+    },
+    {
+      "epoch": 0.010198878123406425,
+      "grad_norm": 0.8798688650131226,
+      "learning_rate": 4.060913705583757e-06,
+      "loss": 0.6696931719779968,
+      "step": 40
+    },
+    {
+      "epoch": 0.010453850076491586,
+      "grad_norm": 0.9106945991516113,
+      "learning_rate": 4.162436548223351e-06,
+      "loss": 0.6799577474594116,
+      "step": 41
+    },
+    {
+      "epoch": 0.010708822029576747,
+      "grad_norm": 0.7262415289878845,
+      "learning_rate": 4.263959390862945e-06,
+      "loss": 0.6715244650840759,
+      "step": 42
+    },
+    {
+      "epoch": 0.010963793982661908,
+      "grad_norm": 0.7152193784713745,
+      "learning_rate": 4.365482233502538e-06,
+      "loss": 0.6724027395248413,
+      "step": 43
+    },
+    {
+      "epoch": 0.011218765935747067,
+      "grad_norm": 0.7244555950164795,
+      "learning_rate": 4.467005076142132e-06,
+      "loss": 0.6676362752914429,
+      "step": 44
+    },
+    {
+      "epoch": 0.011473737888832228,
+      "grad_norm": 0.7677158117294312,
+      "learning_rate": 4.568527918781726e-06,
+      "loss": 0.6664789319038391,
+      "step": 45
+    },
+    {
+      "epoch": 0.011728709841917389,
+      "grad_norm": 0.7524646520614624,
+      "learning_rate": 4.67005076142132e-06,
+      "loss": 0.6637834310531616,
+      "step": 46
+    },
+    {
+      "epoch": 0.01198368179500255,
+      "grad_norm": 0.7234571576118469,
+      "learning_rate": 4.771573604060914e-06,
+      "loss": 0.6616507768630981,
+      "step": 47
+    },
+    {
+      "epoch": 0.01223865374808771,
+      "grad_norm": 0.637627363204956,
+      "learning_rate": 4.873096446700508e-06,
+      "loss": 0.6517107486724854,
+      "step": 48
+    },
+    {
+      "epoch": 0.012493625701172871,
+      "grad_norm": 0.6411532759666443,
+      "learning_rate": 4.974619289340102e-06,
+      "loss": 0.6510401368141174,
+      "step": 49
+    },
+    {
+      "epoch": 0.012748597654258032,
+      "grad_norm": 0.6801655292510986,
+      "learning_rate": 5.076142131979695e-06,
+      "loss": 0.6529415845870972,
+      "step": 50
+    },
+    {
+      "epoch": 0.013003569607343193,
+      "grad_norm": 0.6227138042449951,
+      "learning_rate": 5.17766497461929e-06,
+      "loss": 0.6479157209396362,
+      "step": 51
+    },
+    {
+      "epoch": 0.013258541560428353,
+      "grad_norm": 0.6181749105453491,
+      "learning_rate": 5.279187817258884e-06,
+      "loss": 0.655814528465271,
+      "step": 52
+    },
+    {
+      "epoch": 0.013513513513513514,
+      "grad_norm": 0.6531626582145691,
+      "learning_rate": 5.380710659898477e-06,
+      "loss": 0.648429811000824,
+      "step": 53
+    },
+    {
+      "epoch": 0.013768485466598673,
+      "grad_norm": 0.6198768019676208,
+      "learning_rate": 5.482233502538071e-06,
+      "loss": 0.6356717348098755,
+      "step": 54
+    },
+    {
+      "epoch": 0.014023457419683834,
+      "grad_norm": 0.5874183177947998,
+      "learning_rate": 5.583756345177665e-06,
+      "loss": 0.6472568511962891,
+      "step": 55
+    },
+    {
+      "epoch": 0.014278429372768995,
+      "grad_norm": 0.5928446054458618,
+      "learning_rate": 5.685279187817259e-06,
+      "loss": 0.6416783332824707,
+      "step": 56
+    },
+    {
+      "epoch": 0.014533401325854156,
+      "grad_norm": 0.6353596448898315,
+      "learning_rate": 5.7868020304568525e-06,
+      "loss": 0.6454662084579468,
+      "step": 57
+    },
+    {
+      "epoch": 0.014788373278939317,
+      "grad_norm": 0.6318172216415405,
+      "learning_rate": 5.888324873096447e-06,
+      "loss": 0.6328722238540649,
+      "step": 58
+    },
+    {
+      "epoch": 0.015043345232024477,
+      "grad_norm": 0.6352128982543945,
+      "learning_rate": 5.989847715736041e-06,
+      "loss": 0.6377559900283813,
+      "step": 59
+    },
+    {
+      "epoch": 0.015298317185109638,
+      "grad_norm": 0.5992292761802673,
+      "learning_rate": 6.091370558375635e-06,
+      "loss": 0.6464930772781372,
+      "step": 60
+    },
+    {
+      "epoch": 0.015553289138194799,
+      "grad_norm": 0.6220185160636902,
+      "learning_rate": 6.1928934010152285e-06,
+      "loss": 0.6422460079193115,
+      "step": 61
+    },
+    {
+      "epoch": 0.01580826109127996,
+      "grad_norm": 0.5953255891799927,
+      "learning_rate": 6.2944162436548225e-06,
+      "loss": 0.6442855596542358,
+      "step": 62
+    },
+    {
+      "epoch": 0.01606323304436512,
+      "grad_norm": 0.6460153460502625,
+      "learning_rate": 6.395939086294417e-06,
+      "loss": 0.6416395902633667,
+      "step": 63
+    },
+    {
+      "epoch": 0.01631820499745028,
+      "grad_norm": 0.6249935626983643,
+      "learning_rate": 6.4974619289340105e-06,
+      "loss": 0.6346004605293274,
+      "step": 64
+    },
+    {
+      "epoch": 0.01657317695053544,
+      "grad_norm": 0.7142345309257507,
+      "learning_rate": 6.5989847715736045e-06,
+      "loss": 0.6336624622344971,
+      "step": 65
+    },
+    {
+      "epoch": 0.016828148903620603,
+      "grad_norm": 0.6218382716178894,
+      "learning_rate": 6.7005076142131985e-06,
+      "loss": 0.6198103427886963,
+      "step": 66
+    },
+    {
+      "epoch": 0.017083120856705762,
+      "grad_norm": 0.6140885353088379,
+      "learning_rate": 6.8020304568527926e-06,
+      "loss": 0.6337085962295532,
+      "step": 67
+    },
+    {
+      "epoch": 0.017338092809790925,
+      "grad_norm": 0.6162137985229492,
+      "learning_rate": 6.903553299492386e-06,
+      "loss": 0.6449379920959473,
+      "step": 68
+    },
+    {
+      "epoch": 0.017593064762876084,
+      "grad_norm": 0.6531879901885986,
+      "learning_rate": 7.0050761421319806e-06,
+      "loss": 0.6236452460289001,
+      "step": 69
+    },
+    {
+      "epoch": 0.017848036715961243,
+      "grad_norm": 0.6065159440040588,
+      "learning_rate": 7.106598984771575e-06,
+      "loss": 0.6327449679374695,
+      "step": 70
+    },
+    {
+      "epoch": 0.018103008669046405,
+      "grad_norm": 0.6237702965736389,
+      "learning_rate": 7.208121827411169e-06,
+      "loss": 0.620902955532074,
+      "step": 71
+    },
+    {
+      "epoch": 0.018357980622131564,
+      "grad_norm": 0.6284878849983215,
+      "learning_rate": 7.309644670050762e-06,
+      "loss": 0.6234478950500488,
+      "step": 72
+    },
+    {
+      "epoch": 0.018612952575216727,
+      "grad_norm": 0.6469556093215942,
+      "learning_rate": 7.411167512690356e-06,
+      "loss": 0.6349846124649048,
+      "step": 73
+    },
+    {
+      "epoch": 0.018867924528301886,
+      "grad_norm": 0.6482848525047302,
+      "learning_rate": 7.51269035532995e-06,
+      "loss": 0.633674144744873,
+      "step": 74
+    },
+    {
+      "epoch": 0.01912289648138705,
+      "grad_norm": 0.6971201300621033,
+      "learning_rate": 7.614213197969543e-06,
+      "loss": 0.6282392740249634,
+      "step": 75
+    },
+    {
+      "epoch": 0.019377868434472208,
+      "grad_norm": 0.6171407699584961,
+      "learning_rate": 7.715736040609138e-06,
+      "loss": 0.614886462688446,
+      "step": 76
+    },
+    {
+      "epoch": 0.01963284038755737,
+      "grad_norm": 0.6748046278953552,
+      "learning_rate": 7.817258883248731e-06,
+      "loss": 0.6317158937454224,
+      "step": 77
+    },
+    {
+      "epoch": 0.01988781234064253,
+      "grad_norm": 0.6388630270957947,
+      "learning_rate": 7.918781725888326e-06,
+      "loss": 0.6277610063552856,
+      "step": 78
+    },
+    {
+      "epoch": 0.02014278429372769,
+      "grad_norm": 0.6394343972206116,
+      "learning_rate": 8.020304568527919e-06,
+      "loss": 0.6236151456832886,
+      "step": 79
+    },
+    {
+      "epoch": 0.02039775624681285,
+      "grad_norm": 0.6329667568206787,
+      "learning_rate": 8.121827411167514e-06,
+      "loss": 0.6225125789642334,
+      "step": 80
+    },
+    {
+      "epoch": 0.02065272819989801,
+      "grad_norm": 0.6687252521514893,
+      "learning_rate": 8.223350253807107e-06,
+      "loss": 0.6116753816604614,
+      "step": 81
+    },
+    {
+      "epoch": 0.020907700152983173,
+      "grad_norm": 0.646135151386261,
+      "learning_rate": 8.324873096446702e-06,
+      "loss": 0.6251896023750305,
+      "step": 82
+    },
+    {
+      "epoch": 0.02116267210606833,
+      "grad_norm": 0.6251548528671265,
+      "learning_rate": 8.426395939086295e-06,
+      "loss": 0.6129327416419983,
+      "step": 83
+    },
+    {
+      "epoch": 0.021417644059153494,
+      "grad_norm": 0.6150680780410767,
+      "learning_rate": 8.52791878172589e-06,
+      "loss": 0.6178444623947144,
+      "step": 84
+    },
+    {
+      "epoch": 0.021672616012238653,
+      "grad_norm": 0.645462691783905,
+      "learning_rate": 8.629441624365483e-06,
+      "loss": 0.6154327392578125,
+      "step": 85
+    },
+    {
+      "epoch": 0.021927587965323816,
+      "grad_norm": 0.6240130662918091,
+      "learning_rate": 8.730964467005076e-06,
+      "loss": 0.6073774695396423,
+      "step": 86
+    },
+    {
+      "epoch": 0.022182559918408975,
+      "grad_norm": 0.6194576621055603,
+      "learning_rate": 8.832487309644671e-06,
+      "loss": 0.6132628321647644,
+      "step": 87
+    },
+    {
+      "epoch": 0.022437531871494134,
+      "grad_norm": 0.6547966599464417,
+      "learning_rate": 8.934010152284264e-06,
+      "loss": 0.6120939254760742,
+      "step": 88
+    },
+    {
+      "epoch": 0.022692503824579296,
+      "grad_norm": 0.6388934850692749,
+      "learning_rate": 9.035532994923859e-06,
+      "loss": 0.6320643424987793,
+      "step": 89
+    },
+    {
+      "epoch": 0.022947475777664456,
+      "grad_norm": 0.6568025350570679,
+      "learning_rate": 9.137055837563452e-06,
+      "loss": 0.6370781064033508,
+      "step": 90
+    },
+    {
+      "epoch": 0.023202447730749618,
+      "grad_norm": 0.6334405541419983,
+      "learning_rate": 9.238578680203047e-06,
+      "loss": 0.6066848039627075,
+      "step": 91
+    },
+    {
+      "epoch": 0.023457419683834777,
+      "grad_norm": 0.6498663425445557,
+      "learning_rate": 9.34010152284264e-06,
+      "loss": 0.6138375401496887,
+      "step": 92
+    },
+    {
+      "epoch": 0.02371239163691994,
+      "grad_norm": 0.634401261806488,
+      "learning_rate": 9.441624365482235e-06,
+      "loss": 0.611154317855835,
+      "step": 93
+    },
+    {
+      "epoch": 0.0239673635900051,
+      "grad_norm": 0.6656029224395752,
+      "learning_rate": 9.543147208121828e-06,
+      "loss": 0.6110818386077881,
+      "step": 94
+    },
+    {
+      "epoch": 0.02422233554309026,
+      "grad_norm": 0.652365505695343,
+      "learning_rate": 9.644670050761421e-06,
+      "loss": 0.6210362911224365,
+      "step": 95
+    },
+    {
+      "epoch": 0.02447730749617542,
+      "grad_norm": 0.6741998791694641,
+      "learning_rate": 9.746192893401016e-06,
+      "loss": 0.6129822134971619,
+      "step": 96
+    },
+    {
+      "epoch": 0.024732279449260583,
+      "grad_norm": 0.6117521524429321,
+      "learning_rate": 9.84771573604061e-06,
+      "loss": 0.6152582168579102,
+      "step": 97
+    },
+    {
+      "epoch": 0.024987251402345742,
+      "grad_norm": 0.6950223445892334,
+      "learning_rate": 9.949238578680204e-06,
+      "loss": 0.6205112934112549,
+      "step": 98
+    },
+    {
+      "epoch": 0.0252422233554309,
+      "grad_norm": 0.6633530259132385,
+      "learning_rate": 1.0050761421319797e-05,
+      "loss": 0.6119738221168518,
+      "step": 99
+    },
+    {
+      "epoch": 0.025497195308516064,
+      "grad_norm": 0.668982744216919,
+      "learning_rate": 1.015228426395939e-05,
+      "loss": 0.6128796935081482,
+      "step": 100
+    },
+    {
+      "epoch": 0.025752167261601223,
+      "grad_norm": 0.754180371761322,
+      "learning_rate": 1.0253807106598985e-05,
+      "loss": 0.6129236221313477,
+      "step": 101
+    },
+    {
+      "epoch": 0.026007139214686385,
+      "grad_norm": 0.6716321706771851,
+      "learning_rate": 1.035532994923858e-05,
+      "loss": 0.6249738931655884,
+      "step": 102
+    },
+    {
+      "epoch": 0.026262111167771544,
+      "grad_norm": 0.6696272492408752,
+      "learning_rate": 1.0456852791878173e-05,
+      "loss": 0.6208000779151917,
+      "step": 103
+    },
+    {
+      "epoch": 0.026517083120856707,
+      "grad_norm": 0.7047821283340454,
+      "learning_rate": 1.0558375634517768e-05,
+      "loss": 0.6148910522460938,
+      "step": 104
+    },
+    {
+      "epoch": 0.026772055073941866,
+      "grad_norm": 0.715707004070282,
+      "learning_rate": 1.0659898477157361e-05,
+      "loss": 0.6176806688308716,
+      "step": 105
+    },
+    {
+      "epoch": 0.02702702702702703,
+      "grad_norm": 0.674106776714325,
+      "learning_rate": 1.0761421319796955e-05,
+      "loss": 0.6050000786781311,
+      "step": 106
+    },
+    {
+      "epoch": 0.027281998980112188,
+      "grad_norm": 0.7002929449081421,
+      "learning_rate": 1.0862944162436548e-05,
+      "loss": 0.5905170440673828,
+      "step": 107
+    },
+    {
+      "epoch": 0.027536970933197347,
+      "grad_norm": 0.7115480899810791,
+      "learning_rate": 1.0964467005076143e-05,
+      "loss": 0.6017693281173706,
+      "step": 108
+    },
+    {
+      "epoch": 0.02779194288628251,
+      "grad_norm": 0.6301277279853821,
+      "learning_rate": 1.1065989847715737e-05,
+      "loss": 0.5972022414207458,
+      "step": 109
+    },
+    {
+      "epoch": 0.02804691483936767,
+      "grad_norm": 0.6617989540100098,
+      "learning_rate": 1.116751269035533e-05,
+      "loss": 0.6111524105072021,
+      "step": 110
+    },
+    {
+      "epoch": 0.02830188679245283,
+      "grad_norm": 0.6753233075141907,
+      "learning_rate": 1.1269035532994925e-05,
+      "loss": 0.6101921796798706,
+      "step": 111
+    },
+    {
+      "epoch": 0.02855685874553799,
+      "grad_norm": 0.6855331063270569,
+      "learning_rate": 1.1370558375634519e-05,
+      "loss": 0.6058908700942993,
+      "step": 112
+    },
+    {
+      "epoch": 0.028811830698623152,
+      "grad_norm": 0.6627272367477417,
+      "learning_rate": 1.1472081218274113e-05,
+      "loss": 0.61765456199646,
+      "step": 113
+    },
+    {
+      "epoch": 0.02906680265170831,
+      "grad_norm": 0.7235305309295654,
+      "learning_rate": 1.1573604060913705e-05,
+      "loss": 0.6114434599876404,
+      "step": 114
+    },
+    {
+      "epoch": 0.029321774604793474,
+      "grad_norm": 0.6944801211357117,
+      "learning_rate": 1.16751269035533e-05,
+      "loss": 0.5895316004753113,
+      "step": 115
+    },
+    {
+      "epoch": 0.029576746557878633,
+      "grad_norm": 0.6143687963485718,
+      "learning_rate": 1.1776649746192895e-05,
+      "loss": 0.6181043982505798,
+      "step": 116
+    },
+    {
+      "epoch": 0.029831718510963796,
+      "grad_norm": 0.702218770980835,
+      "learning_rate": 1.1878172588832488e-05,
+      "loss": 0.6035462021827698,
+      "step": 117
+    },
+    {
+      "epoch": 0.030086690464048955,
+      "grad_norm": 0.649609386920929,
+      "learning_rate": 1.1979695431472083e-05,
+      "loss": 0.6008034348487854,
+      "step": 118
+    },
+    {
+      "epoch": 0.030341662417134114,
+      "grad_norm": 0.7478456497192383,
+      "learning_rate": 1.2081218274111678e-05,
+      "loss": 0.6118278503417969,
+      "step": 119
+    },
+    {
+      "epoch": 0.030596634370219276,
+      "grad_norm": 0.6522828936576843,
+      "learning_rate": 1.218274111675127e-05,
+      "loss": 0.6109780669212341,
+      "step": 120
+    },
+    {
+      "epoch": 0.030851606323304435,
+      "grad_norm": 0.6617293953895569,
+      "learning_rate": 1.2284263959390864e-05,
+      "loss": 0.6059006452560425,
+      "step": 121
+    },
+    {
+      "epoch": 0.031106578276389598,
+      "grad_norm": 0.7125791311264038,
+      "learning_rate": 1.2385786802030457e-05,
+      "loss": 0.5911107063293457,
+      "step": 122
+    },
+    {
+      "epoch": 0.03136155022947476,
+      "grad_norm": 0.7545065879821777,
+      "learning_rate": 1.2487309644670052e-05,
+      "loss": 0.6011978983879089,
+      "step": 123
+    },
+    {
+      "epoch": 0.03161652218255992,
+      "grad_norm": 0.7025015950202942,
+      "learning_rate": 1.2588832487309645e-05,
+      "loss": 0.6012612581253052,
+      "step": 124
+    },
+    {
+      "epoch": 0.03187149413564508,
+      "grad_norm": 0.7487797737121582,
+      "learning_rate": 1.269035532994924e-05,
+      "loss": 0.5970184803009033,
+      "step": 125
+    },
+    {
+      "epoch": 0.03212646608873024,
+      "grad_norm": 0.6467530727386475,
+      "learning_rate": 1.2791878172588835e-05,
+      "loss": 0.5995411276817322,
+      "step": 126
+    },
+    {
+      "epoch": 0.0323814380418154,
+      "grad_norm": 0.6963929533958435,
+      "learning_rate": 1.2893401015228428e-05,
+      "loss": 0.6060649752616882,
+      "step": 127
+    },
+    {
+      "epoch": 0.03263640999490056,
+      "grad_norm": 0.7099882364273071,
+      "learning_rate": 1.2994923857868021e-05,
+      "loss": 0.6047376394271851,
+      "step": 128
+    },
+    {
+      "epoch": 0.03289138194798572,
+      "grad_norm": 0.6539604067802429,
+      "learning_rate": 1.3096446700507614e-05,
+      "loss": 0.586577832698822,
+      "step": 129
+    },
+    {
+      "epoch": 0.03314635390107088,
+      "grad_norm": 0.6586759686470032,
+      "learning_rate": 1.3197969543147209e-05,
+      "loss": 0.5893585681915283,
+      "step": 130
+    },
+    {
+      "epoch": 0.03340132585415604,
+      "grad_norm": 0.6708022356033325,
+      "learning_rate": 1.3299492385786802e-05,
+      "loss": 0.5953554511070251,
+      "step": 131
+    },
+    {
+      "epoch": 0.033656297807241206,
+      "grad_norm": 0.669730007648468,
+      "learning_rate": 1.3401015228426397e-05,
+      "loss": 0.6061270236968994,
+      "step": 132
+    },
+    {
+      "epoch": 0.033911269760326365,
+      "grad_norm": 0.7589663863182068,
+      "learning_rate": 1.3502538071065992e-05,
+      "loss": 0.6139975786209106,
+      "step": 133
+    },
+    {
+      "epoch": 0.034166241713411524,
+      "grad_norm": 0.7068342566490173,
+      "learning_rate": 1.3604060913705585e-05,
+      "loss": 0.5982658267021179,
+      "step": 134
+    },
+    {
+      "epoch": 0.03442121366649668,
+      "grad_norm": 0.7072777152061462,
+      "learning_rate": 1.3705583756345178e-05,
+      "loss": 0.5927161574363708,
+      "step": 135
+    },
+    {
+      "epoch": 0.03467618561958185,
+      "grad_norm": 0.8180981874465942,
+      "learning_rate": 1.3807106598984771e-05,
+      "loss": 0.6063701510429382,
+      "step": 136
+    },
+    {
+      "epoch": 0.03493115757266701,
+      "grad_norm": 0.601114809513092,
+      "learning_rate": 1.3908629441624366e-05,
+      "loss": 0.574968159198761,
+      "step": 137
+    },
+    {
+      "epoch": 0.03518612952575217,
+      "grad_norm": 0.7802360653877258,
+      "learning_rate": 1.4010152284263961e-05,
+      "loss": 0.5933843851089478,
+      "step": 138
+    },
+    {
+      "epoch": 0.03544110147883733,
+      "grad_norm": 0.7044690847396851,
+      "learning_rate": 1.4111675126903554e-05,
+      "loss": 0.5892487168312073,
+      "step": 139
+    },
+    {
+      "epoch": 0.035696073431922486,
+      "grad_norm": 0.7081224918365479,
+      "learning_rate": 1.421319796954315e-05,
+      "loss": 0.602599024772644,
+      "step": 140
+    },
+    {
+      "epoch": 0.03595104538500765,
+      "grad_norm": 0.6481499075889587,
+      "learning_rate": 1.4314720812182742e-05,
+      "loss": 0.6026634573936462,
+      "step": 141
+    },
+    {
+      "epoch": 0.03620601733809281,
+      "grad_norm": 0.7358949184417725,
+      "learning_rate": 1.4416243654822337e-05,
+      "loss": 0.609551191329956,
+      "step": 142
+    },
+    {
+      "epoch": 0.03646098929117797,
+      "grad_norm": 0.6168767809867859,
+      "learning_rate": 1.4517766497461929e-05,
+      "loss": 0.591394305229187,
+      "step": 143
+    },
+    {
+      "epoch": 0.03671596124426313,
+      "grad_norm": 0.6998642086982727,
+      "learning_rate": 1.4619289340101523e-05,
+      "loss": 0.606634795665741,
+      "step": 144
+    },
+    {
+      "epoch": 0.036970933197348295,
+      "grad_norm": 0.6324466466903687,
+      "learning_rate": 1.4720812182741118e-05,
+      "loss": 0.6037495136260986,
+      "step": 145
+    },
+    {
+      "epoch": 0.037225905150433454,
+      "grad_norm": 0.7579569220542908,
+      "learning_rate": 1.4822335025380712e-05,
+      "loss": 0.5949714779853821,
+      "step": 146
+    },
+    {
+      "epoch": 0.03748087710351861,
+      "grad_norm": 0.708786129951477,
+      "learning_rate": 1.4923857868020306e-05,
+      "loss": 0.5930920839309692,
+      "step": 147
+    },
+    {
+      "epoch": 0.03773584905660377,
+      "grad_norm": 0.6507843732833862,
+      "learning_rate": 1.50253807106599e-05,
+      "loss": 0.6192787885665894,
+      "step": 148
+    },
+    {
+      "epoch": 0.03799082100968893,
+      "grad_norm": 0.6853079795837402,
+      "learning_rate": 1.5126903553299494e-05,
+      "loss": 0.5902525782585144,
+      "step": 149
+    },
+    {
+      "epoch": 0.0382457929627741,
+      "grad_norm": 0.7084800601005554,
+      "learning_rate": 1.5228426395939086e-05,
+      "loss": 0.5898470878601074,
+      "step": 150
+    },
+    {
+      "epoch": 0.038500764915859256,
+      "grad_norm": 0.6946215033531189,
+      "learning_rate": 1.5329949238578682e-05,
+      "loss": 0.6009114384651184,
+      "step": 151
+    },
+    {
+      "epoch": 0.038755736868944415,
+      "grad_norm": 0.6725452542304993,
+      "learning_rate": 1.5431472081218276e-05,
+      "loss": 0.5899134874343872,
+      "step": 152
+    },
+    {
+      "epoch": 0.039010708822029574,
+      "grad_norm": 0.7391459345817566,
+      "learning_rate": 1.553299492385787e-05,
+      "loss": 0.6052982211112976,
+      "step": 153
+    },
+    {
+      "epoch": 0.03926568077511474,
+      "grad_norm": 0.653334379196167,
+      "learning_rate": 1.5634517766497462e-05,
+      "loss": 0.5881596803665161,
+      "step": 154
+    },
+    {
+      "epoch": 0.0395206527281999,
+      "grad_norm": 0.70075923204422,
+      "learning_rate": 1.573604060913706e-05,
+      "loss": 0.5922362208366394,
+      "step": 155
+    },
+    {
+      "epoch": 0.03977562468128506,
+      "grad_norm": 0.6499072313308716,
+      "learning_rate": 1.583756345177665e-05,
+      "loss": 0.6015427708625793,
+      "step": 156
+    },
+    {
+      "epoch": 0.04003059663437022,
+      "grad_norm": 0.6751073002815247,
+      "learning_rate": 1.5939086294416245e-05,
+      "loss": 0.5883609652519226,
+      "step": 157
+    },
+    {
+      "epoch": 0.04028556858745538,
+      "grad_norm": 0.6795080900192261,
+      "learning_rate": 1.6040609137055838e-05,
+      "loss": 0.6162182688713074,
+      "step": 158
+    },
+    {
+      "epoch": 0.04054054054054054,
+      "grad_norm": 0.7047825455665588,
+      "learning_rate": 1.614213197969543e-05,
+      "loss": 0.5969679355621338,
+      "step": 159
+    },
+    {
+      "epoch": 0.0407955124936257,
+      "grad_norm": 0.6710919737815857,
+      "learning_rate": 1.6243654822335028e-05,
+      "loss": 0.5879135131835938,
+      "step": 160
+    },
+    {
+      "epoch": 0.04105048444671086,
+      "grad_norm": 0.6726529002189636,
+      "learning_rate": 1.634517766497462e-05,
+      "loss": 0.5928366184234619,
+      "step": 161
+    },
+    {
+      "epoch": 0.04130545639979602,
+      "grad_norm": 0.6386929154396057,
+      "learning_rate": 1.6446700507614214e-05,
+      "loss": 0.6050669550895691,
+      "step": 162
+    },
+    {
+      "epoch": 0.041560428352881186,
+      "grad_norm": 0.7748963832855225,
+      "learning_rate": 1.654822335025381e-05,
+      "loss": 0.5833348035812378,
+      "step": 163
+    },
+    {
+      "epoch": 0.041815400305966345,
+      "grad_norm": 0.7063996195793152,
+      "learning_rate": 1.6649746192893404e-05,
+      "loss": 0.588036298751831,
+      "step": 164
+    },
+    {
+      "epoch": 0.042070372259051504,
+      "grad_norm": 0.663123369216919,
+      "learning_rate": 1.6751269035532997e-05,
+      "loss": 0.5778357982635498,
+      "step": 165
+    },
+    {
+      "epoch": 0.04232534421213666,
+      "grad_norm": 0.715749204158783,
+      "learning_rate": 1.685279187817259e-05,
+      "loss": 0.5950402617454529,
+      "step": 166
+    },
+    {
+      "epoch": 0.04258031616522182,
+      "grad_norm": 0.657593309879303,
+      "learning_rate": 1.6954314720812183e-05,
+      "loss": 0.5839961767196655,
+      "step": 167
+    },
+    {
+      "epoch": 0.04283528811830699,
+      "grad_norm": 0.7044532299041748,
+      "learning_rate": 1.705583756345178e-05,
+      "loss": 0.5974175930023193,
+      "step": 168
+    },
+    {
+      "epoch": 0.04309026007139215,
+      "grad_norm": 0.678438663482666,
+      "learning_rate": 1.7157360406091373e-05,
+      "loss": 0.5922316312789917,
+      "step": 169
+    },
+    {
+      "epoch": 0.043345232024477306,
+      "grad_norm": 0.7534765601158142,
+      "learning_rate": 1.7258883248730966e-05,
+      "loss": 0.5797853469848633,
+      "step": 170
+    },
+    {
+      "epoch": 0.043600203977562466,
+      "grad_norm": 0.6486764550209045,
+      "learning_rate": 1.736040609137056e-05,
+      "loss": 0.5776516795158386,
+      "step": 171
+    },
+    {
+      "epoch": 0.04385517593064763,
+      "grad_norm": 0.6965651512145996,
+      "learning_rate": 1.7461928934010152e-05,
+      "loss": 0.5867475271224976,
+      "step": 172
+    },
+    {
+      "epoch": 0.04411014788373279,
+      "grad_norm": 0.7084111571311951,
+      "learning_rate": 1.7563451776649745e-05,
+      "loss": 0.5911486744880676,
+      "step": 173
+    },
+    {
+      "epoch": 0.04436511983681795,
+      "grad_norm": 0.6508429646492004,
+      "learning_rate": 1.7664974619289342e-05,
+      "loss": 0.5861883163452148,
+      "step": 174
+    },
+    {
+      "epoch": 0.04462009178990311,
+      "grad_norm": 0.6957427859306335,
+      "learning_rate": 1.7766497461928935e-05,
+      "loss": 0.5958301424980164,
+      "step": 175
+    },
+    {
+      "epoch": 0.04487506374298827,
+      "grad_norm": 0.6256112456321716,
+      "learning_rate": 1.786802030456853e-05,
+      "loss": 0.5757142901420593,
+      "step": 176
+    },
+    {
+      "epoch": 0.045130035696073434,
+      "grad_norm": 0.6769145727157593,
+      "learning_rate": 1.7969543147208125e-05,
+      "loss": 0.6127809286117554,
+      "step": 177
+    },
+    {
+      "epoch": 0.04538500764915859,
+      "grad_norm": 0.6851696372032166,
+      "learning_rate": 1.8071065989847718e-05,
+      "loss": 0.5945026874542236,
+      "step": 178
+    },
+    {
+      "epoch": 0.04563997960224375,
+      "grad_norm": 0.6712672114372253,
+      "learning_rate": 1.817258883248731e-05,
+      "loss": 0.5810881853103638,
+      "step": 179
+    },
+    {
+      "epoch": 0.04589495155532891,
+      "grad_norm": 0.6257203817367554,
+      "learning_rate": 1.8274111675126904e-05,
+      "loss": 0.5861809253692627,
+      "step": 180
+    },
+    {
+      "epoch": 0.04614992350841408,
+      "grad_norm": 0.7018057107925415,
+      "learning_rate": 1.8375634517766498e-05,
+      "loss": 0.5822906494140625,
+      "step": 181
+    },
+    {
+      "epoch": 0.046404895461499236,
+      "grad_norm": 0.6754285097122192,
+      "learning_rate": 1.8477157360406094e-05,
+      "loss": 0.5803866386413574,
+      "step": 182
+    },
+    {
+      "epoch": 0.046659867414584395,
+      "grad_norm": 0.6747144460678101,
+      "learning_rate": 1.8578680203045687e-05,
+      "loss": 0.5928809642791748,
+      "step": 183
+    },
+    {
+      "epoch": 0.046914839367669554,
+      "grad_norm": 0.6424155235290527,
+      "learning_rate": 1.868020304568528e-05,
+      "loss": 0.6039294004440308,
+      "step": 184
+    },
+    {
+      "epoch": 0.04716981132075472,
+      "grad_norm": 0.6336119771003723,
+      "learning_rate": 1.8781725888324877e-05,
+      "loss": 0.5872094631195068,
+      "step": 185
+    },
+    {
+      "epoch": 0.04742478327383988,
+      "grad_norm": 0.8231200575828552,
+      "learning_rate": 1.888324873096447e-05,
+      "loss": 0.5870229005813599,
+      "step": 186
+    },
+    {
+      "epoch": 0.04767975522692504,
+      "grad_norm": 0.6766306161880493,
+      "learning_rate": 1.8984771573604063e-05,
+      "loss": 0.596156120300293,
+      "step": 187
+    },
+    {
+      "epoch": 0.0479347271800102,
+      "grad_norm": 0.7034800052642822,
+      "learning_rate": 1.9086294416243656e-05,
+      "loss": 0.5882968306541443,
+      "step": 188
+    },
+    {
+      "epoch": 0.04818969913309536,
+      "grad_norm": 0.6791765093803406,
+      "learning_rate": 1.918781725888325e-05,
+      "loss": 0.5723391771316528,
+      "step": 189
+    },
+    {
+      "epoch": 0.04844467108618052,
+      "grad_norm": 0.7587200999259949,
+      "learning_rate": 1.9289340101522843e-05,
+      "loss": 0.5900762677192688,
+      "step": 190
+    },
+    {
+      "epoch": 0.04869964303926568,
+      "grad_norm": 0.7111191749572754,
+      "learning_rate": 1.939086294416244e-05,
+      "loss": 0.5732899308204651,
+      "step": 191
+    },
+    {
+      "epoch": 0.04895461499235084,
+      "grad_norm": 0.6762776374816895,
+      "learning_rate": 1.9492385786802032e-05,
+      "loss": 0.5881006717681885,
+      "step": 192
+    },
+    {
+      "epoch": 0.049209586945436,
+      "grad_norm": 0.6622965335845947,
+      "learning_rate": 1.9593908629441626e-05,
+      "loss": 0.5785661935806274,
+      "step": 193
+    },
+    {
+      "epoch": 0.049464558898521166,
+      "grad_norm": 0.6711365580558777,
+      "learning_rate": 1.969543147208122e-05,
+      "loss": 0.5604966282844543,
+      "step": 194
+    },
+    {
+      "epoch": 0.049719530851606325,
+      "grad_norm": 0.6800675392150879,
+      "learning_rate": 1.9796954314720812e-05,
+      "loss": 0.5834771394729614,
+      "step": 195
+    },
+    {
+      "epoch": 0.049974502804691484,
+      "grad_norm": 0.7173908352851868,
+      "learning_rate": 1.989847715736041e-05,
+      "loss": 0.5812022686004639,
+      "step": 196
+    },
+    {
+      "epoch": 0.05022947475777664,
+      "grad_norm": 0.7289883494377136,
+      "learning_rate": 2e-05,
+      "loss": 0.5865119099617004,
+      "step": 197
+    },
+    {
+      "epoch": 0.0504844467108618,
+      "grad_norm": 0.6825205087661743,
+      "learning_rate": 1.9999996443546215e-05,
+      "loss": 0.583699643611908,
+      "step": 198
+    },
+    {
+      "epoch": 0.05073941866394697,
+      "grad_norm": 0.7007641196250916,
+      "learning_rate": 1.9999985774187383e-05,
+      "loss": 0.5956920981407166,
+      "step": 199
+    },
+    {
+      "epoch": 0.05099439061703213,
+      "grad_norm": 0.6625441312789917,
+      "learning_rate": 1.9999967991931094e-05,
+      "loss": 0.5949492454528809,
+      "step": 200
+    },
+    {
+      "epoch": 0.051249362570117286,
+      "grad_norm": 0.6780283451080322,
+      "learning_rate": 1.999994309679e-05,
+      "loss": 0.5827687978744507,
+      "step": 201
+    },
+    {
+      "epoch": 0.051504334523202445,
+      "grad_norm": 0.6601573824882507,
+      "learning_rate": 1.9999911088781804e-05,
+      "loss": 0.5779587626457214,
+      "step": 202
+    },
+    {
+      "epoch": 0.05175930647628761,
+      "grad_norm": 0.6573741436004639,
+      "learning_rate": 1.9999871967929278e-05,
+      "loss": 0.599594235420227,
+      "step": 203
+    },
+    {
+      "epoch": 0.05201427842937277,
+      "grad_norm": 0.639542281627655,
+      "learning_rate": 1.9999825734260247e-05,
+      "loss": 0.5684338808059692,
+      "step": 204
+    },
+    {
+      "epoch": 0.05226925038245793,
+      "grad_norm": 0.6658681035041809,
+      "learning_rate": 1.9999772387807592e-05,
+      "loss": 0.5895668268203735,
+      "step": 205
+    },
+    {
+      "epoch": 0.05252422233554309,
+      "grad_norm": 0.6262486577033997,
+      "learning_rate": 1.9999711928609265e-05,
+      "loss": 0.5960357189178467,
+      "step": 206
+    },
+    {
+      "epoch": 0.05277919428862825,
+      "grad_norm": 0.6094916462898254,
+      "learning_rate": 1.9999644356708262e-05,
+      "loss": 0.5768816471099854,
+      "step": 207
+    },
+    {
+      "epoch": 0.053034166241713414,
+      "grad_norm": 0.6605136394500732,
+      "learning_rate": 1.999956967215265e-05,
+      "loss": 0.5881094932556152,
+      "step": 208
+    },
+    {
+      "epoch": 0.05328913819479857,
+      "grad_norm": 0.7232625484466553,
+      "learning_rate": 1.9999487874995556e-05,
+      "loss": 0.5898884534835815,
+      "step": 209
+    },
+    {
+      "epoch": 0.05354411014788373,
+      "grad_norm": 0.5892757773399353,
+      "learning_rate": 1.9999398965295157e-05,
+      "loss": 0.5826902985572815,
+      "step": 210
+    },
+    {
+      "epoch": 0.05379908210096889,
+      "grad_norm": 0.6755984425544739,
+      "learning_rate": 1.9999302943114692e-05,
+      "loss": 0.5842537879943848,
+      "step": 211
+    },
+    {
+      "epoch": 0.05405405405405406,
+      "grad_norm": 0.6419636011123657,
+      "learning_rate": 1.999919980852246e-05,
+      "loss": 0.5654566287994385,
+      "step": 212
+    },
+    {
+      "epoch": 0.054309026007139216,
+      "grad_norm": 0.7050777673721313,
+      "learning_rate": 1.9999089561591826e-05,
+      "loss": 0.5780625343322754,
+      "step": 213
+    },
+    {
+      "epoch": 0.054563997960224375,
+      "grad_norm": 0.6014391183853149,
+      "learning_rate": 1.9998972202401206e-05,
+      "loss": 0.5929491519927979,
+      "step": 214
+    },
+    {
+      "epoch": 0.054818969913309534,
+      "grad_norm": 0.662752628326416,
+      "learning_rate": 1.9998847731034073e-05,
+      "loss": 0.5813001990318298,
+      "step": 215
+    },
+    {
+      "epoch": 0.05507394186639469,
+      "grad_norm": 0.6744083166122437,
+      "learning_rate": 1.9998716147578963e-05,
+      "loss": 0.5822510123252869,
+      "step": 216
+    },
+    {
+      "epoch": 0.05532891381947986,
+      "grad_norm": 0.6255607008934021,
+      "learning_rate": 1.999857745212947e-05,
+      "loss": 0.5750855803489685,
+      "step": 217
+    },
+    {
+      "epoch": 0.05558388577256502,
+      "grad_norm": 0.7068489789962769,
+      "learning_rate": 1.999843164478425e-05,
+      "loss": 0.5958958864212036,
+      "step": 218
+    },
+    {
+      "epoch": 0.05583885772565018,
+      "grad_norm": 0.6492578983306885,
+      "learning_rate": 1.9998278725647015e-05,
+      "loss": 0.5895745754241943,
+      "step": 219
+    },
+    {
+      "epoch": 0.05609382967873534,
+      "grad_norm": 0.6811429262161255,
+      "learning_rate": 1.9998118694826524e-05,
+      "loss": 0.5909900665283203,
+      "step": 220
+    },
+    {
+      "epoch": 0.0563488016318205,
+      "grad_norm": 0.6466650366783142,
+      "learning_rate": 1.999795155243662e-05,
+      "loss": 0.5822628140449524,
+      "step": 221
+    },
+    {
+      "epoch": 0.05660377358490566,
+      "grad_norm": 0.6460858583450317,
+      "learning_rate": 1.999777729859618e-05,
+      "loss": 0.5704171061515808,
+      "step": 222
+    },
+    {
+      "epoch": 0.05685874553799082,
+      "grad_norm": 0.7094568014144897,
+      "learning_rate": 1.999759593342916e-05,
+      "loss": 0.5730971693992615,
+      "step": 223
+    },
+    {
+      "epoch": 0.05711371749107598,
+      "grad_norm": 0.6798518896102905,
+      "learning_rate": 1.9997407457064547e-05,
+      "loss": 0.5954874157905579,
+      "step": 224
+    },
+    {
+      "epoch": 0.05736868944416114,
+      "grad_norm": 0.6695542335510254,
+      "learning_rate": 1.9997211869636417e-05,
+      "loss": 0.5870078802108765,
+      "step": 225
+    },
+    {
+      "epoch": 0.057623661397246305,
+      "grad_norm": 0.6144403219223022,
+      "learning_rate": 1.9997009171283882e-05,
+      "loss": 0.5849398374557495,
+      "step": 226
+    },
+    {
+      "epoch": 0.057878633350331464,
+      "grad_norm": 0.7087647318840027,
+      "learning_rate": 1.9996799362151124e-05,
+      "loss": 0.5610650777816772,
+      "step": 227
+    },
+    {
+      "epoch": 0.05813360530341662,
+      "grad_norm": 0.6279194951057434,
+      "learning_rate": 1.999658244238737e-05,
+      "loss": 0.5676144361495972,
+      "step": 228
+    },
+    {
+      "epoch": 0.05838857725650178,
+      "grad_norm": 0.7145519256591797,
+      "learning_rate": 1.9996358412146922e-05,
+      "loss": 0.5789085626602173,
+      "step": 229
+    },
+    {
+      "epoch": 0.05864354920958695,
+      "grad_norm": 0.6787282824516296,
+      "learning_rate": 1.999612727158913e-05,
+      "loss": 0.5748767852783203,
+      "step": 230
+    },
+    {
+      "epoch": 0.05889852116267211,
+      "grad_norm": 0.6754794120788574,
+      "learning_rate": 1.99958890208784e-05,
+      "loss": 0.5880950093269348,
+      "step": 231
+    },
+    {
+      "epoch": 0.059153493115757266,
+      "grad_norm": 3.1220149993896484,
+      "learning_rate": 1.999564366018419e-05,
+      "loss": 0.5785700082778931,
+      "step": 232
+    },
+    {
+      "epoch": 0.059408465068842425,
+      "grad_norm": 0.8344593644142151,
+      "learning_rate": 1.999539118968104e-05,
+      "loss": 0.5717012882232666,
+      "step": 233
+    },
+    {
+      "epoch": 0.05966343702192759,
+      "grad_norm": 0.6129970550537109,
+      "learning_rate": 1.9995131609548513e-05,
+      "loss": 0.5926172733306885,
+      "step": 234
+    },
+    {
+      "epoch": 0.05991840897501275,
+      "grad_norm": 0.7066675424575806,
+      "learning_rate": 1.9994864919971255e-05,
+      "loss": 0.5734254717826843,
+      "step": 235
+    },
+    {
+      "epoch": 0.06017338092809791,
+      "grad_norm": 0.7014307975769043,
+      "learning_rate": 1.9994591121138954e-05,
+      "loss": 0.5930376052856445,
+      "step": 236
+    },
+    {
+      "epoch": 0.06042835288118307,
+      "grad_norm": 0.6754729747772217,
+      "learning_rate": 1.999431021324637e-05,
+      "loss": 0.5725037455558777,
+      "step": 237
+    },
+    {
+      "epoch": 0.06068332483426823,
+      "grad_norm": 0.714168131351471,
+      "learning_rate": 1.99940221964933e-05,
+      "loss": 0.5748128890991211,
+      "step": 238
+    },
+    {
+      "epoch": 0.060938296787353394,
+      "grad_norm": 0.6713688969612122,
+      "learning_rate": 1.9993727071084613e-05,
+      "loss": 0.5799189805984497,
+      "step": 239
+    },
+    {
+      "epoch": 0.06119326874043855,
+      "grad_norm": 0.6011117696762085,
+      "learning_rate": 1.999342483723023e-05,
+      "loss": 0.5691922903060913,
+      "step": 240
+    },
+    {
+      "epoch": 0.06144824069352371,
+      "grad_norm": 0.6895564794540405,
+      "learning_rate": 1.9993115495145125e-05,
+      "loss": 0.5746656060218811,
+      "step": 241
+    },
+    {
+      "epoch": 0.06170321264660887,
+      "grad_norm": 0.6951008439064026,
+      "learning_rate": 1.999279904504933e-05,
+      "loss": 0.574233889579773,
+      "step": 242
+    },
+    {
+      "epoch": 0.06195818459969404,
+      "grad_norm": 0.6126100420951843,
+      "learning_rate": 1.999247548716793e-05,
+      "loss": 0.565665602684021,
+      "step": 243
+    },
+    {
+      "epoch": 0.062213156552779196,
+      "grad_norm": 0.6327571868896484,
+      "learning_rate": 1.9992144821731075e-05,
+      "loss": 0.5754146575927734,
+      "step": 244
+    },
+    {
+      "epoch": 0.062468128505864355,
+      "grad_norm": 0.6757508516311646,
+      "learning_rate": 1.999180704897396e-05,
+      "loss": 0.5776671171188354,
+      "step": 245
+    },
+    {
+      "epoch": 0.06272310045894952,
+      "grad_norm": 0.6368160843849182,
+      "learning_rate": 1.9991462169136843e-05,
+      "loss": 0.5814864635467529,
+      "step": 246
+    },
+    {
+      "epoch": 0.06297807241203468,
+      "grad_norm": 0.6689087748527527,
+      "learning_rate": 1.9991110182465032e-05,
+      "loss": 0.5699710845947266,
+      "step": 247
+    },
+    {
+      "epoch": 0.06323304436511984,
+      "grad_norm": 0.6050186157226562,
+      "learning_rate": 1.9990751089208894e-05,
+      "loss": 0.5749836564064026,
+      "step": 248
+    },
+    {
+      "epoch": 0.063488016318205,
+      "grad_norm": 0.6237219572067261,
+      "learning_rate": 1.999038488962384e-05,
+      "loss": 0.5625735521316528,
+      "step": 249
+    },
+    {
+      "epoch": 0.06374298827129016,
+      "grad_norm": 0.6202013492584229,
+      "learning_rate": 1.9990011583970355e-05,
+      "loss": 0.5654894113540649,
+      "step": 250
+    },
+    {
+      "epoch": 0.06399796022437532,
+      "grad_norm": 0.5948776602745056,
+      "learning_rate": 1.998963117251396e-05,
+      "loss": 0.5873595476150513,
+      "step": 251
+    },
+    {
+      "epoch": 0.06425293217746048,
+      "grad_norm": 0.6015293002128601,
+      "learning_rate": 1.9989243655525248e-05,
+      "loss": 0.5803345441818237,
+      "step": 252
+    },
+    {
+      "epoch": 0.06450790413054563,
+      "grad_norm": 0.5996026992797852,
+      "learning_rate": 1.9988849033279848e-05,
+      "loss": 0.5710455775260925,
+      "step": 253
+    },
+    {
+      "epoch": 0.0647628760836308,
+      "grad_norm": 0.5705514550209045,
+      "learning_rate": 1.998844730605845e-05,
+      "loss": 0.5731304883956909,
+      "step": 254
+    },
+    {
+      "epoch": 0.06501784803671597,
+      "grad_norm": 0.5985062718391418,
+      "learning_rate": 1.9988038474146804e-05,
+      "loss": 0.5627564191818237,
+      "step": 255
+    },
+    {
+      "epoch": 0.06527281998980113,
+      "grad_norm": 0.6054355502128601,
+      "learning_rate": 1.9987622537835708e-05,
+      "loss": 0.5643206834793091,
+      "step": 256
+    },
+    {
+      "epoch": 0.06552779194288628,
+      "grad_norm": 0.643243670463562,
+      "learning_rate": 1.998719949742101e-05,
+      "loss": 0.5702025890350342,
+      "step": 257
+    },
+    {
+      "epoch": 0.06578276389597144,
+      "grad_norm": 0.6212320327758789,
+      "learning_rate": 1.998676935320362e-05,
+      "loss": 0.5596565008163452,
+      "step": 258
+    },
+    {
+      "epoch": 0.0660377358490566,
+      "grad_norm": 0.6149182915687561,
+      "learning_rate": 1.998633210548949e-05,
+      "loss": 0.5833432078361511,
+      "step": 259
+    },
+    {
+      "epoch": 0.06629270780214176,
+      "grad_norm": 0.5876778960227966,
+      "learning_rate": 1.9985887754589634e-05,
+      "loss": 0.563776969909668,
+      "step": 260
+    },
+    {
+      "epoch": 0.06654767975522692,
+      "grad_norm": 0.6171606183052063,
+      "learning_rate": 1.9985436300820108e-05,
+      "loss": 0.5687385201454163,
+      "step": 261
+    },
+    {
+      "epoch": 0.06680265170831208,
+      "grad_norm": 0.6119545698165894,
+      "learning_rate": 1.9984977744502038e-05,
+      "loss": 0.5770053267478943,
+      "step": 262
+    },
+    {
+      "epoch": 0.06705762366139725,
+      "grad_norm": 0.625106155872345,
+      "learning_rate": 1.9984512085961584e-05,
+      "loss": 0.5761595964431763,
+      "step": 263
+    },
+    {
+      "epoch": 0.06731259561448241,
+      "grad_norm": 0.627996563911438,
+      "learning_rate": 1.9984039325529965e-05,
+      "loss": 0.5653480291366577,
+      "step": 264
+    },
+    {
+      "epoch": 0.06756756756756757,
+      "grad_norm": 0.6073549389839172,
+      "learning_rate": 1.9983559463543453e-05,
+      "loss": 0.5723285675048828,
+      "step": 265
+    },
+    {
+      "epoch": 0.06782253952065273,
+      "grad_norm": 0.6070031523704529,
+      "learning_rate": 1.9983072500343367e-05,
+      "loss": 0.5722025632858276,
+      "step": 266
+    },
+    {
+      "epoch": 0.06807751147373789,
+      "grad_norm": 0.6099826097488403,
+      "learning_rate": 1.9982578436276084e-05,
+      "loss": 0.5669994354248047,
+      "step": 267
+    },
+    {
+      "epoch": 0.06833248342682305,
+      "grad_norm": 0.565504252910614,
+      "learning_rate": 1.998207727169302e-05,
+      "loss": 0.5737874507904053,
+      "step": 268
+    },
+    {
+      "epoch": 0.06858745537990821,
+      "grad_norm": 0.6359668970108032,
+      "learning_rate": 1.9981569006950655e-05,
+      "loss": 0.5619522929191589,
+      "step": 269
+    },
+    {
+      "epoch": 0.06884242733299337,
+      "grad_norm": 0.5864189863204956,
+      "learning_rate": 1.998105364241051e-05,
+      "loss": 0.5720523595809937,
+      "step": 270
+    },
+    {
+      "epoch": 0.06909739928607853,
+      "grad_norm": 0.6055141091346741,
+      "learning_rate": 1.9980531178439157e-05,
+      "loss": 0.5729068517684937,
+      "step": 271
+    },
+    {
+      "epoch": 0.0693523712391637,
+      "grad_norm": 0.5796358585357666,
+      "learning_rate": 1.9980001615408228e-05,
+      "loss": 0.5904380083084106,
+      "step": 272
+    },
+    {
+      "epoch": 0.06960734319224886,
+      "grad_norm": 0.5816705822944641,
+      "learning_rate": 1.9979464953694388e-05,
+      "loss": 0.561646580696106,
+      "step": 273
+    },
+    {
+      "epoch": 0.06986231514533402,
+      "grad_norm": 0.5701183676719666,
+      "learning_rate": 1.9978921193679362e-05,
+      "loss": 0.5620859861373901,
+      "step": 274
+    },
+    {
+      "epoch": 0.07011728709841918,
+      "grad_norm": 0.640988826751709,
+      "learning_rate": 1.997837033574992e-05,
+      "loss": 0.5681776404380798,
+      "step": 275
+    },
+    {
+      "epoch": 0.07037225905150434,
+      "grad_norm": 0.5843631625175476,
+      "learning_rate": 1.9977812380297888e-05,
+      "loss": 0.5757836103439331,
+      "step": 276
+    },
+    {
+      "epoch": 0.0706272310045895,
+      "grad_norm": 0.6079508066177368,
+      "learning_rate": 1.9977247327720128e-05,
+      "loss": 0.5664551258087158,
+      "step": 277
+    },
+    {
+      "epoch": 0.07088220295767465,
+      "grad_norm": 0.6057885885238647,
+      "learning_rate": 1.9976675178418562e-05,
+      "loss": 0.5635764598846436,
+      "step": 278
+    },
+    {
+      "epoch": 0.07113717491075981,
+      "grad_norm": 0.5704780220985413,
+      "learning_rate": 1.997609593280015e-05,
+      "loss": 0.5763840675354004,
+      "step": 279
+    },
+    {
+      "epoch": 0.07139214686384497,
+      "grad_norm": 0.6236803531646729,
+      "learning_rate": 1.9975509591276902e-05,
+      "loss": 0.5770968794822693,
+      "step": 280
+    },
+    {
+      "epoch": 0.07164711881693014,
+      "grad_norm": 0.5445568561553955,
+      "learning_rate": 1.9974916154265887e-05,
+      "loss": 0.5587365627288818,
+      "step": 281
+    },
+    {
+      "epoch": 0.0719020907700153,
+      "grad_norm": 0.5997340679168701,
+      "learning_rate": 1.99743156221892e-05,
+      "loss": 0.5820911526679993,
+      "step": 282
+    },
+    {
+      "epoch": 0.07215706272310046,
+      "grad_norm": 0.6064010858535767,
+      "learning_rate": 1.9973707995474e-05,
+      "loss": 0.5616083145141602,
+      "step": 283
+    },
+    {
+      "epoch": 0.07241203467618562,
+      "grad_norm": 0.5523048043251038,
+      "learning_rate": 1.997309327455249e-05,
+      "loss": 0.5786893367767334,
+      "step": 284
+    },
+    {
+      "epoch": 0.07266700662927078,
+      "grad_norm": 0.544638454914093,
+      "learning_rate": 1.9972471459861908e-05,
+      "loss": 0.5746819972991943,
+      "step": 285
+    },
+    {
+      "epoch": 0.07292197858235594,
+      "grad_norm": 0.5817301273345947,
+      "learning_rate": 1.997184255184455e-05,
+      "loss": 0.5628771185874939,
+      "step": 286
+    },
+    {
+      "epoch": 0.0731769505354411,
+      "grad_norm": 0.5777922868728638,
+      "learning_rate": 1.9971206550947748e-05,
+      "loss": 0.5575648546218872,
+      "step": 287
+    },
+    {
+      "epoch": 0.07343192248852626,
+      "grad_norm": 0.591174840927124,
+      "learning_rate": 1.9970563457623885e-05,
+      "loss": 0.5619316101074219,
+      "step": 288
+    },
+    {
+      "epoch": 0.07368689444161142,
+      "grad_norm": 0.6272574067115784,
+      "learning_rate": 1.996991327233039e-05,
+      "loss": 0.554466962814331,
+      "step": 289
+    },
+    {
+      "epoch": 0.07394186639469659,
+      "grad_norm": 0.6420910954475403,
+      "learning_rate": 1.996925599552973e-05,
+      "loss": 0.5560216903686523,
+      "step": 290
+    },
+    {
+      "epoch": 0.07419683834778175,
+      "grad_norm": 0.6134066581726074,
+      "learning_rate": 1.9968591627689428e-05,
+      "loss": 0.5638004541397095,
+      "step": 291
+    },
+    {
+      "epoch": 0.07445181030086691,
+      "grad_norm": 0.5705217123031616,
+      "learning_rate": 1.996792016928203e-05,
+      "loss": 0.5734155178070068,
+      "step": 292
+    },
+    {
+      "epoch": 0.07470678225395207,
+      "grad_norm": 0.6209928393363953,
+      "learning_rate": 1.9967241620785148e-05,
+      "loss": 0.5502181053161621,
+      "step": 293
+    },
+    {
+      "epoch": 0.07496175420703723,
+      "grad_norm": 0.5950906276702881,
+      "learning_rate": 1.9966555982681418e-05,
+      "loss": 0.5590144991874695,
+      "step": 294
+    },
+    {
+      "epoch": 0.07521672616012239,
+      "grad_norm": 0.5660822987556458,
+      "learning_rate": 1.996586325545854e-05,
+      "loss": 0.5804231762886047,
+      "step": 295
+    },
+    {
+      "epoch": 0.07547169811320754,
+      "grad_norm": 0.6057717204093933,
+      "learning_rate": 1.996516343960924e-05,
+      "loss": 0.5541191101074219,
+      "step": 296
+    },
+    {
+      "epoch": 0.0757266700662927,
+      "grad_norm": 0.6694291234016418,
+      "learning_rate": 1.9964456535631287e-05,
+      "loss": 0.5697247385978699,
+      "step": 297
+    },
+    {
+      "epoch": 0.07598164201937786,
+      "grad_norm": 0.6713988184928894,
+      "learning_rate": 1.9963742544027495e-05,
+      "loss": 0.5601903200149536,
+      "step": 298
+    },
+    {
+      "epoch": 0.07623661397246304,
+      "grad_norm": 0.5943032503128052,
+      "learning_rate": 1.9963021465305724e-05,
+      "loss": 0.5566859841346741,
+      "step": 299
+    },
+    {
+      "epoch": 0.0764915859255482,
+      "grad_norm": 0.6334525942802429,
+      "learning_rate": 1.996229329997887e-05,
+      "loss": 0.5652881860733032,
+      "step": 300
+    },
+    {
+      "epoch": 0.07674655787863335,
+      "grad_norm": 0.6462168097496033,
+      "learning_rate": 1.996155804856487e-05,
+      "loss": 0.568797767162323,
+      "step": 301
+    },
+    {
+      "epoch": 0.07700152983171851,
+      "grad_norm": 0.5707945227622986,
+      "learning_rate": 1.9960815711586695e-05,
+      "loss": 0.5688931345939636,
+      "step": 302
+    },
+    {
+      "epoch": 0.07725650178480367,
+      "grad_norm": 0.5915647745132446,
+      "learning_rate": 1.996006628957237e-05,
+      "loss": 0.5620356798171997,
+      "step": 303
+    },
+    {
+      "epoch": 0.07751147373788883,
+      "grad_norm": 0.6422096490859985,
+      "learning_rate": 1.995930978305495e-05,
+      "loss": 0.5672441720962524,
+      "step": 304
+    },
+    {
+      "epoch": 0.07776644569097399,
+      "grad_norm": 0.5536637306213379,
+      "learning_rate": 1.9958546192572535e-05,
+      "loss": 0.567467987537384,
+      "step": 305
+    },
+    {
+      "epoch": 0.07802141764405915,
+      "grad_norm": 0.5747365355491638,
+      "learning_rate": 1.995777551866825e-05,
+      "loss": 0.5674901008605957,
+      "step": 306
+    },
+    {
+      "epoch": 0.07827638959714431,
+      "grad_norm": 0.5945401191711426,
+      "learning_rate": 1.9956997761890277e-05,
+      "loss": 0.5611063838005066,
+      "step": 307
+    },
+    {
+      "epoch": 0.07853136155022948,
+      "grad_norm": 0.5844597816467285,
+      "learning_rate": 1.9956212922791825e-05,
+      "loss": 0.5652987360954285,
+      "step": 308
+    },
+    {
+      "epoch": 0.07878633350331464,
+      "grad_norm": 0.5997607111930847,
+      "learning_rate": 1.995542100193114e-05,
+      "loss": 0.5385614633560181,
+      "step": 309
+    },
+    {
+      "epoch": 0.0790413054563998,
+      "grad_norm": 0.5956833362579346,
+      "learning_rate": 1.9954621999871512e-05,
+      "loss": 0.5502104759216309,
+      "step": 310
+    },
+    {
+      "epoch": 0.07929627740948496,
+      "grad_norm": 0.6196988821029663,
+      "learning_rate": 1.9953815917181258e-05,
+      "loss": 0.554365336894989,
+      "step": 311
+    },
+    {
+      "epoch": 0.07955124936257012,
+      "grad_norm": 0.5872536897659302,
+      "learning_rate": 1.9953002754433743e-05,
+      "loss": 0.5669418573379517,
+      "step": 312
+    },
+    {
+      "epoch": 0.07980622131565528,
+      "grad_norm": 0.5234706401824951,
+      "learning_rate": 1.995218251220736e-05,
+      "loss": 0.5619447231292725,
+      "step": 313
+    },
+    {
+      "epoch": 0.08006119326874044,
+      "grad_norm": 0.6450950503349304,
+      "learning_rate": 1.995135519108554e-05,
+      "loss": 0.5421331524848938,
+      "step": 314
+    },
+    {
+      "epoch": 0.0803161652218256,
+      "grad_norm": 0.5748299956321716,
+      "learning_rate": 1.995052079165675e-05,
+      "loss": 0.5626130700111389,
+      "step": 315
+    },
+    {
+      "epoch": 0.08057113717491075,
+      "grad_norm": 0.5997022390365601,
+      "learning_rate": 1.994967931451449e-05,
+      "loss": 0.5518707036972046,
+      "step": 316
+    },
+    {
+      "epoch": 0.08082610912799593,
+      "grad_norm": 0.6092442274093628,
+      "learning_rate": 1.994883076025729e-05,
+      "loss": 0.5542733073234558,
+      "step": 317
+    },
+    {
+      "epoch": 0.08108108108108109,
+      "grad_norm": 0.5561699271202087,
+      "learning_rate": 1.9947975129488726e-05,
+      "loss": 0.5553290247917175,
+      "step": 318
+    },
+    {
+      "epoch": 0.08133605303416624,
+      "grad_norm": 0.5849249362945557,
+      "learning_rate": 1.9947112422817397e-05,
+      "loss": 0.562719464302063,
+      "step": 319
+    },
+    {
+      "epoch": 0.0815910249872514,
+      "grad_norm": 0.5908464193344116,
+      "learning_rate": 1.994624264085694e-05,
+      "loss": 0.5615310668945312,
+      "step": 320
+    },
+    {
+      "epoch": 0.08184599694033656,
+      "grad_norm": 0.5735774040222168,
+      "learning_rate": 1.9945365784226018e-05,
+      "loss": 0.5658733248710632,
+      "step": 321
+    },
+    {
+      "epoch": 0.08210096889342172,
+      "grad_norm": 0.5955777168273926,
+      "learning_rate": 1.9944481853548335e-05,
+      "loss": 0.5692462921142578,
+      "step": 322
+    },
+    {
+      "epoch": 0.08235594084650688,
+      "grad_norm": 0.5768380761146545,
+      "learning_rate": 1.994359084945262e-05,
+      "loss": 0.5534395575523376,
+      "step": 323
+    },
+    {
+      "epoch": 0.08261091279959204,
+      "grad_norm": 0.5331956744194031,
+      "learning_rate": 1.9942692772572645e-05,
+      "loss": 0.561090886592865,
+      "step": 324
+    },
+    {
+      "epoch": 0.0828658847526772,
+      "grad_norm": 0.564062237739563,
+      "learning_rate": 1.9941787623547194e-05,
+      "loss": 0.5530338287353516,
+      "step": 325
+    },
+    {
+      "epoch": 0.08312085670576237,
+      "grad_norm": 0.5807541012763977,
+      "learning_rate": 1.9940875403020095e-05,
+      "loss": 0.5750647783279419,
+      "step": 326
+    },
+    {
+      "epoch": 0.08337582865884753,
+      "grad_norm": 0.5979731678962708,
+      "learning_rate": 1.9939956111640198e-05,
+      "loss": 0.5668591260910034,
+      "step": 327
+    },
+    {
+      "epoch": 0.08363080061193269,
+      "grad_norm": 0.5643814206123352,
+      "learning_rate": 1.993902975006139e-05,
+      "loss": 0.5625386834144592,
+      "step": 328
+    },
+    {
+      "epoch": 0.08388577256501785,
+      "grad_norm": 0.7797948122024536,
+      "learning_rate": 1.9938096318942583e-05,
+      "loss": 0.5680009722709656,
+      "step": 329
+    },
+    {
+      "epoch": 0.08414074451810301,
+      "grad_norm": 0.6033815741539001,
+      "learning_rate": 1.993715581894772e-05,
+      "loss": 0.551729679107666,
+      "step": 330
+    },
+    {
+      "epoch": 0.08439571647118817,
+      "grad_norm": 0.7249321937561035,
+      "learning_rate": 1.9936208250745767e-05,
+      "loss": 0.5621654987335205,
+      "step": 331
+    },
+    {
+      "epoch": 0.08465068842427333,
+      "grad_norm": 0.5542919635772705,
+      "learning_rate": 1.9935253615010718e-05,
+      "loss": 0.5623407363891602,
+      "step": 332
+    },
+    {
+      "epoch": 0.08490566037735849,
+      "grad_norm": 0.5352950692176819,
+      "learning_rate": 1.9934291912421602e-05,
+      "loss": 0.5668007135391235,
+      "step": 333
+    },
+    {
+      "epoch": 0.08516063233044364,
+      "grad_norm": 0.58268803358078,
+      "learning_rate": 1.993332314366247e-05,
+      "loss": 0.5632613301277161,
+      "step": 334
+    },
+    {
+      "epoch": 0.08541560428352882,
+      "grad_norm": 0.5689071416854858,
+      "learning_rate": 1.9932347309422393e-05,
+      "loss": 0.5548432469367981,
+      "step": 335
+    },
+    {
+      "epoch": 0.08567057623661398,
+      "grad_norm": 0.5977305173873901,
+      "learning_rate": 1.9931364410395474e-05,
+      "loss": 0.5616758465766907,
+      "step": 336
+    },
+    {
+      "epoch": 0.08592554818969914,
+      "grad_norm": 0.5791174173355103,
+      "learning_rate": 1.9930374447280844e-05,
+      "loss": 0.5640117526054382,
+      "step": 337
+    },
+    {
+      "epoch": 0.0861805201427843,
+      "grad_norm": 0.5817056894302368,
+      "learning_rate": 1.992937742078265e-05,
+      "loss": 0.5501354932785034,
+      "step": 338
+    },
+    {
+      "epoch": 0.08643549209586945,
+      "grad_norm": 0.555757462978363,
+      "learning_rate": 1.9928373331610068e-05,
+      "loss": 0.5771246552467346,
+      "step": 339
+    },
+    {
+      "epoch": 0.08669046404895461,
+      "grad_norm": 0.5622315406799316,
+      "learning_rate": 1.99273621804773e-05,
+      "loss": 0.5560314655303955,
+      "step": 340
+    },
+    {
+      "epoch": 0.08694543600203977,
+      "grad_norm": 0.5567899942398071,
+      "learning_rate": 1.992634396810357e-05,
+      "loss": 0.5638811588287354,
+      "step": 341
+    },
+    {
+      "epoch": 0.08720040795512493,
+      "grad_norm": 0.9306987524032593,
+      "learning_rate": 1.992531869521312e-05,
+      "loss": 0.5746092796325684,
+      "step": 342
+    },
+    {
+      "epoch": 0.08745537990821009,
+      "grad_norm": 0.6356567740440369,
+      "learning_rate": 1.9924286362535214e-05,
+      "loss": 0.5542607307434082,
+      "step": 343
+    },
+    {
+      "epoch": 0.08771035186129526,
+      "grad_norm": 0.5900321006774902,
+      "learning_rate": 1.9923246970804148e-05,
+      "loss": 0.5701124668121338,
+      "step": 344
+    },
+    {
+      "epoch": 0.08796532381438042,
+      "grad_norm": 0.5773775577545166,
+      "learning_rate": 1.9922200520759224e-05,
+      "loss": 0.567479133605957,
+      "step": 345
+    },
+    {
+      "epoch": 0.08822029576746558,
+      "grad_norm": 0.6096591949462891,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 0.5689170956611633,
+      "step": 346
+    },
+    {
+      "epoch": 0.08847526772055074,
+      "grad_norm": 0.5603031516075134,
+      "learning_rate": 1.9920086448710162e-05,
+      "loss": 0.5633706450462341,
+      "step": 347
+    },
+    {
+      "epoch": 0.0887302396736359,
+      "grad_norm": 0.5907047986984253,
+      "learning_rate": 1.9919018828209737e-05,
+      "loss": 0.5522236227989197,
+      "step": 348
+    },
+    {
+      "epoch": 0.08898521162672106,
+      "grad_norm": 0.5694087743759155,
+      "learning_rate": 1.99179441524029e-05,
+      "loss": 0.5575593113899231,
+      "step": 349
+    },
+    {
+      "epoch": 0.08924018357980622,
+      "grad_norm": 0.5818880796432495,
+      "learning_rate": 1.9916862422054053e-05,
+      "loss": 0.5685945749282837,
+      "step": 350
+    },
+    {
+      "epoch": 0.08949515553289138,
+      "grad_norm": 0.5598475933074951,
+      "learning_rate": 1.991577363793262e-05,
+      "loss": 0.5636742115020752,
+      "step": 351
+    },
+    {
+      "epoch": 0.08975012748597654,
+      "grad_norm": 0.5648708939552307,
+      "learning_rate": 1.991467780081305e-05,
+      "loss": 0.549278974533081,
+      "step": 352
+    },
+    {
+      "epoch": 0.09000509943906171,
+      "grad_norm": 0.5766149163246155,
+      "learning_rate": 1.9913574911474795e-05,
+      "loss": 0.5545766353607178,
+      "step": 353
+    },
+    {
+      "epoch": 0.09026007139214687,
+      "grad_norm": 0.5259166955947876,
+      "learning_rate": 1.9912464970702334e-05,
+      "loss": 0.5719385147094727,
+      "step": 354
+    },
+    {
+      "epoch": 0.09051504334523203,
+      "grad_norm": 0.5788517594337463,
+      "learning_rate": 1.9911347979285157e-05,
+      "loss": 0.5588270425796509,
+      "step": 355
+    },
+    {
+      "epoch": 0.09077001529831719,
+      "grad_norm": 0.5590032935142517,
+      "learning_rate": 1.9910223938017768e-05,
+      "loss": 0.5672649145126343,
+      "step": 356
+    },
+    {
+      "epoch": 0.09102498725140235,
+      "grad_norm": 0.5529116988182068,
+      "learning_rate": 1.9909092847699684e-05,
+      "loss": 0.5570537447929382,
+      "step": 357
+    },
+    {
+      "epoch": 0.0912799592044875,
+      "grad_norm": 0.5393619537353516,
+      "learning_rate": 1.9907954709135445e-05,
+      "loss": 0.5653344988822937,
+      "step": 358
+    },
+    {
+      "epoch": 0.09153493115757266,
+      "grad_norm": 0.5495200753211975,
+      "learning_rate": 1.9906809523134597e-05,
+      "loss": 0.5471888780593872,
+      "step": 359
+    },
+    {
+      "epoch": 0.09178990311065782,
+      "grad_norm": 0.5408692955970764,
+      "learning_rate": 1.9905657290511698e-05,
+      "loss": 0.5618144273757935,
+      "step": 360
+    },
+    {
+      "epoch": 0.09204487506374298,
+      "grad_norm": 0.5768983364105225,
+      "learning_rate": 1.9904498012086327e-05,
+      "loss": 0.556114673614502,
+      "step": 361
+    },
+    {
+      "epoch": 0.09229984701682815,
+      "grad_norm": 0.5828980207443237,
+      "learning_rate": 1.990333168868306e-05,
+      "loss": 0.5702318549156189,
+      "step": 362
+    },
+    {
+      "epoch": 0.09255481896991331,
+      "grad_norm": 0.5826273560523987,
+      "learning_rate": 1.9902158321131494e-05,
+      "loss": 0.5587303638458252,
+      "step": 363
+    },
+    {
+      "epoch": 0.09280979092299847,
+      "grad_norm": 0.5333880186080933,
+      "learning_rate": 1.9900977910266235e-05,
+      "loss": 0.5579195618629456,
+      "step": 364
+    },
+    {
+      "epoch": 0.09306476287608363,
+      "grad_norm": 0.5813481211662292,
+      "learning_rate": 1.98997904569269e-05,
+      "loss": 0.5583409070968628,
+      "step": 365
+    },
+    {
+      "epoch": 0.09331973482916879,
+      "grad_norm": 0.5596633553504944,
+      "learning_rate": 1.9898595961958112e-05,
+      "loss": 0.5528003573417664,
+      "step": 366
+    },
+    {
+      "epoch": 0.09357470678225395,
+      "grad_norm": 0.536656379699707,
+      "learning_rate": 1.9897394426209507e-05,
+      "loss": 0.5665175914764404,
+      "step": 367
+    },
+    {
+      "epoch": 0.09382967873533911,
+      "grad_norm": 0.561928391456604,
+      "learning_rate": 1.989618585053572e-05,
+      "loss": 0.5620167851448059,
+      "step": 368
+    },
+    {
+      "epoch": 0.09408465068842427,
+      "grad_norm": 0.6040554046630859,
+      "learning_rate": 1.9894970235796408e-05,
+      "loss": 0.5658797025680542,
+      "step": 369
+    },
+    {
+      "epoch": 0.09433962264150944,
+      "grad_norm": 0.5419930815696716,
+      "learning_rate": 1.9893747582856216e-05,
+      "loss": 0.5657435655593872,
+      "step": 370
+    },
+    {
+      "epoch": 0.0945945945945946,
+      "grad_norm": 0.5847061276435852,
+      "learning_rate": 1.989251789258482e-05,
+      "loss": 0.5543770790100098,
+      "step": 371
+    },
+    {
+      "epoch": 0.09484956654767976,
+      "grad_norm": 0.5925385355949402,
+      "learning_rate": 1.9891281165856876e-05,
+      "loss": 0.5622945427894592,
+      "step": 372
+    },
+    {
+      "epoch": 0.09510453850076492,
+      "grad_norm": 0.5780953168869019,
+      "learning_rate": 1.9890037403552055e-05,
+      "loss": 0.5627380609512329,
+      "step": 373
+    },
+    {
+      "epoch": 0.09535951045385008,
+      "grad_norm": 0.5803493857383728,
+      "learning_rate": 1.9888786606555045e-05,
+      "loss": 0.564932107925415,
+      "step": 374
+    },
+    {
+      "epoch": 0.09561448240693524,
+      "grad_norm": 0.5735729336738586,
+      "learning_rate": 1.9887528775755517e-05,
+      "loss": 0.5499950647354126,
+      "step": 375
+    },
+    {
+      "epoch": 0.0958694543600204,
+      "grad_norm": 0.5783923268318176,
+      "learning_rate": 1.9886263912048156e-05,
+      "loss": 0.5513893961906433,
+      "step": 376
+    },
+    {
+      "epoch": 0.09612442631310555,
+      "grad_norm": 0.544529378414154,
+      "learning_rate": 1.988499201633265e-05,
+      "loss": 0.5571684837341309,
+      "step": 377
+    },
+    {
+      "epoch": 0.09637939826619071,
+      "grad_norm": 0.547946035861969,
+      "learning_rate": 1.9883713089513686e-05,
+      "loss": 0.5583531856536865,
+      "step": 378
+    },
+    {
+      "epoch": 0.09663437021927589,
+      "grad_norm": 0.6041200160980225,
+      "learning_rate": 1.988242713250095e-05,
+      "loss": 0.5552366971969604,
+      "step": 379
+    },
+    {
+      "epoch": 0.09688934217236105,
+      "grad_norm": 0.5608474016189575,
+      "learning_rate": 1.9881134146209137e-05,
+      "loss": 0.5538177490234375,
+      "step": 380
+    },
+    {
+      "epoch": 0.0971443141254462,
+      "grad_norm": 0.5594263076782227,
+      "learning_rate": 1.987983413155793e-05,
+      "loss": 0.5568601489067078,
+      "step": 381
+    },
+    {
+      "epoch": 0.09739928607853136,
+      "grad_norm": 0.5792958736419678,
+      "learning_rate": 1.987852708947202e-05,
+      "loss": 0.541644811630249,
+      "step": 382
+    },
+    {
+      "epoch": 0.09765425803161652,
+      "grad_norm": 0.6069263219833374,
+      "learning_rate": 1.9877213020881096e-05,
+      "loss": 0.5578107833862305,
+      "step": 383
+    },
+    {
+      "epoch": 0.09790922998470168,
+      "grad_norm": 0.5543441772460938,
+      "learning_rate": 1.9875891926719838e-05,
+      "loss": 0.564591109752655,
+      "step": 384
+    },
+    {
+      "epoch": 0.09816420193778684,
+      "grad_norm": 0.5760945677757263,
+      "learning_rate": 1.9874563807927936e-05,
+      "loss": 0.5550767779350281,
+      "step": 385
+    },
+    {
+      "epoch": 0.098419173890872,
+      "grad_norm": 0.5751408934593201,
+      "learning_rate": 1.987322866545006e-05,
+      "loss": 0.5435432195663452,
+      "step": 386
+    },
+    {
+      "epoch": 0.09867414584395716,
+      "grad_norm": 0.5727622509002686,
+      "learning_rate": 1.987188650023589e-05,
+      "loss": 0.5572792291641235,
+      "step": 387
+    },
+    {
+      "epoch": 0.09892911779704233,
+      "grad_norm": 0.5712685585021973,
+      "learning_rate": 1.9870537313240096e-05,
+      "loss": 0.5545316934585571,
+      "step": 388
+    },
+    {
+      "epoch": 0.09918408975012749,
+      "grad_norm": 0.5742475986480713,
+      "learning_rate": 1.9869181105422335e-05,
+      "loss": 0.5356405973434448,
+      "step": 389
+    },
+    {
+      "epoch": 0.09943906170321265,
+      "grad_norm": 0.5529100298881531,
+      "learning_rate": 1.9867817877747276e-05,
+      "loss": 0.5502123236656189,
+      "step": 390
+    },
+    {
+      "epoch": 0.09969403365629781,
+      "grad_norm": 0.5420888662338257,
+      "learning_rate": 1.9866447631184556e-05,
+      "loss": 0.548579752445221,
+      "step": 391
+    },
+    {
+      "epoch": 0.09994900560938297,
+      "grad_norm": 0.6033974885940552,
+      "learning_rate": 1.9865070366708835e-05,
+      "loss": 0.551939845085144,
+      "step": 392
+    },
+    {
+      "epoch": 0.10020397756246813,
+      "grad_norm": 0.5301747918128967,
+      "learning_rate": 1.986368608529974e-05,
+      "loss": 0.5500297546386719,
+      "step": 393
+    },
+    {
+      "epoch": 0.10045894951555329,
+      "grad_norm": 0.6244176626205444,
+      "learning_rate": 1.9862294787941897e-05,
+      "loss": 0.5556297302246094,
+      "step": 394
+    },
+    {
+      "epoch": 0.10071392146863845,
+      "grad_norm": 0.5527774691581726,
+      "learning_rate": 1.9860896475624925e-05,
+      "loss": 0.5543330311775208,
+      "step": 395
+    },
+    {
+      "epoch": 0.1009688934217236,
+      "grad_norm": 0.5964891314506531,
+      "learning_rate": 1.985949114934343e-05,
+      "loss": 0.5608077049255371,
+      "step": 396
+    },
+    {
+      "epoch": 0.10122386537480878,
+      "grad_norm": 0.5965831875801086,
+      "learning_rate": 1.9858078810097004e-05,
+      "loss": 0.5453798770904541,
+      "step": 397
+    },
+    {
+      "epoch": 0.10147883732789394,
+      "grad_norm": 0.5628226399421692,
+      "learning_rate": 1.9856659458890237e-05,
+      "loss": 0.5567731857299805,
+      "step": 398
+    },
+    {
+      "epoch": 0.1017338092809791,
+      "grad_norm": 0.5935264825820923,
+      "learning_rate": 1.98552330967327e-05,
+      "loss": 0.5467360615730286,
+      "step": 399
+    },
+    {
+      "epoch": 0.10198878123406425,
+      "grad_norm": 0.5754925012588501,
+      "learning_rate": 1.9853799724638943e-05,
+      "loss": 0.5451920032501221,
+      "step": 400
+    },
+    {
+      "epoch": 0.10224375318714941,
+      "grad_norm": 0.5660573244094849,
+      "learning_rate": 1.985235934362852e-05,
+      "loss": 0.5535790324211121,
+      "step": 401
+    },
+    {
+      "epoch": 0.10249872514023457,
+      "grad_norm": 0.6029878854751587,
+      "learning_rate": 1.985091195472596e-05,
+      "loss": 0.5511131286621094,
+      "step": 402
+    },
+    {
+      "epoch": 0.10275369709331973,
+      "grad_norm": 0.5555292963981628,
+      "learning_rate": 1.984945755896077e-05,
+      "loss": 0.5386227369308472,
+      "step": 403
+    },
+    {
+      "epoch": 0.10300866904640489,
+      "grad_norm": 0.6450228095054626,
+      "learning_rate": 1.9847996157367455e-05,
+      "loss": 0.5452139973640442,
+      "step": 404
+    },
+    {
+      "epoch": 0.10326364099949005,
+      "grad_norm": 0.5396519303321838,
+      "learning_rate": 1.984652775098549e-05,
+      "loss": 0.5553629398345947,
+      "step": 405
+    },
+    {
+      "epoch": 0.10351861295257522,
+      "grad_norm": 1.392758846282959,
+      "learning_rate": 1.984505234085935e-05,
+      "loss": 0.5525428652763367,
+      "step": 406
+    },
+    {
+      "epoch": 0.10377358490566038,
+      "grad_norm": 0.593498706817627,
+      "learning_rate": 1.9843569928038468e-05,
+      "loss": 0.5546976327896118,
+      "step": 407
+    },
+    {
+      "epoch": 0.10402855685874554,
+      "grad_norm": 0.5546035170555115,
+      "learning_rate": 1.984208051357728e-05,
+      "loss": 0.5391968488693237,
+      "step": 408
+    },
+    {
+      "epoch": 0.1042835288118307,
+      "grad_norm": 0.5852994918823242,
+      "learning_rate": 1.9840584098535184e-05,
+      "loss": 0.5535978078842163,
+      "step": 409
+    },
+    {
+      "epoch": 0.10453850076491586,
+      "grad_norm": 0.5861825346946716,
+      "learning_rate": 1.9839080683976574e-05,
+      "loss": 0.5517117977142334,
+      "step": 410
+    },
+    {
+      "epoch": 0.10479347271800102,
+      "grad_norm": 0.5767414569854736,
+      "learning_rate": 1.983757027097081e-05,
+      "loss": 0.5635597109794617,
+      "step": 411
+    },
+    {
+      "epoch": 0.10504844467108618,
+      "grad_norm": 0.5809765458106995,
+      "learning_rate": 1.9836052860592237e-05,
+      "loss": 0.5357990264892578,
+      "step": 412
+    },
+    {
+      "epoch": 0.10530341662417134,
+      "grad_norm": 0.601608157157898,
+      "learning_rate": 1.9834528453920177e-05,
+      "loss": 0.549247145652771,
+      "step": 413
+    },
+    {
+      "epoch": 0.1055583885772565,
+      "grad_norm": 0.6048389673233032,
+      "learning_rate": 1.9832997052038922e-05,
+      "loss": 0.5468675494194031,
+      "step": 414
+    },
+    {
+      "epoch": 0.10581336053034167,
+      "grad_norm": 0.5551306009292603,
+      "learning_rate": 1.9831458656037748e-05,
+      "loss": 0.5442308187484741,
+      "step": 415
+    },
+    {
+      "epoch": 0.10606833248342683,
+      "grad_norm": 0.5358591675758362,
+      "learning_rate": 1.9829913267010898e-05,
+      "loss": 0.5525547862052917,
+      "step": 416
+    },
+    {
+      "epoch": 0.10632330443651199,
+      "grad_norm": 0.5656468272209167,
+      "learning_rate": 1.9828360886057597e-05,
+      "loss": 0.5506001710891724,
+      "step": 417
+    },
+    {
+      "epoch": 0.10657827638959715,
+      "grad_norm": 0.5704612135887146,
+      "learning_rate": 1.982680151428203e-05,
+      "loss": 0.5432718992233276,
+      "step": 418
+    },
+    {
+      "epoch": 0.1068332483426823,
+      "grad_norm": 0.5470494031906128,
+      "learning_rate": 1.982523515279338e-05,
+      "loss": 0.5554221868515015,
+      "step": 419
+    },
+    {
+      "epoch": 0.10708822029576746,
+      "grad_norm": 0.6086367964744568,
+      "learning_rate": 1.9823661802705773e-05,
+      "loss": 0.5490662455558777,
+      "step": 420
+    },
+    {
+      "epoch": 0.10734319224885262,
+      "grad_norm": 0.5634582042694092,
+      "learning_rate": 1.982208146513832e-05,
+      "loss": 0.5513579845428467,
+      "step": 421
+    },
+    {
+      "epoch": 0.10759816420193778,
+      "grad_norm": 0.5439919233322144,
+      "learning_rate": 1.98204941412151e-05,
+      "loss": 0.5450636744499207,
+      "step": 422
+    },
+    {
+      "epoch": 0.10785313615502294,
+      "grad_norm": 0.5421820878982544,
+      "learning_rate": 1.9818899832065172e-05,
+      "loss": 0.5400946736335754,
+      "step": 423
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 0.5724503397941589,
+      "learning_rate": 1.981729853882254e-05,
+      "loss": 0.5543707609176636,
+      "step": 424
+    },
+    {
+      "epoch": 0.10836308006119327,
+      "grad_norm": 0.5568701028823853,
+      "learning_rate": 1.9815690262626194e-05,
+      "loss": 0.5440797805786133,
+      "step": 425
+    },
+    {
+      "epoch": 0.10861805201427843,
+      "grad_norm": 0.5401669144630432,
+      "learning_rate": 1.981407500462009e-05,
+      "loss": 0.5601478815078735,
+      "step": 426
+    },
+    {
+      "epoch": 0.10887302396736359,
+      "grad_norm": 0.5365921854972839,
+      "learning_rate": 1.9812452765953137e-05,
+      "loss": 0.5376150012016296,
+      "step": 427
+    },
+    {
+      "epoch": 0.10912799592044875,
+      "grad_norm": 0.5326974391937256,
+      "learning_rate": 1.9810823547779226e-05,
+      "loss": 0.5423977375030518,
+      "step": 428
+    },
+    {
+      "epoch": 0.10938296787353391,
+      "grad_norm": 0.6022288203239441,
+      "learning_rate": 1.9809187351257205e-05,
+      "loss": 0.5444180965423584,
+      "step": 429
+    },
+    {
+      "epoch": 0.10963793982661907,
+      "grad_norm": 0.5397586226463318,
+      "learning_rate": 1.9807544177550882e-05,
+      "loss": 0.548419713973999,
+      "step": 430
+    },
+    {
+      "epoch": 0.10989291177970423,
+      "grad_norm": 0.5493386387825012,
+      "learning_rate": 1.980589402782903e-05,
+      "loss": 0.5459563732147217,
+      "step": 431
+    },
+    {
+      "epoch": 0.11014788373278939,
+      "grad_norm": 0.5432290434837341,
+      "learning_rate": 1.980423690326539e-05,
+      "loss": 0.5504255294799805,
+      "step": 432
+    },
+    {
+      "epoch": 0.11040285568587456,
+      "grad_norm": 0.5319405198097229,
+      "learning_rate": 1.9802572805038655e-05,
+      "loss": 0.5571215152740479,
+      "step": 433
+    },
+    {
+      "epoch": 0.11065782763895972,
+      "grad_norm": 0.5671195387840271,
+      "learning_rate": 1.9800901734332488e-05,
+      "loss": 0.5390557050704956,
+      "step": 434
+    },
+    {
+      "epoch": 0.11091279959204488,
+      "grad_norm": 0.5227031707763672,
+      "learning_rate": 1.97992236923355e-05,
+      "loss": 0.5487812757492065,
+      "step": 435
+    },
+    {
+      "epoch": 0.11116777154513004,
+      "grad_norm": 0.6271248459815979,
+      "learning_rate": 1.979753868024127e-05,
+      "loss": 0.5339798331260681,
+      "step": 436
+    },
+    {
+      "epoch": 0.1114227434982152,
+      "grad_norm": 0.5489307045936584,
+      "learning_rate": 1.979584669924833e-05,
+      "loss": 0.5359088182449341,
+      "step": 437
+    },
+    {
+      "epoch": 0.11167771545130036,
+      "grad_norm": 0.5491774082183838,
+      "learning_rate": 1.9794147750560176e-05,
+      "loss": 0.5513111352920532,
+      "step": 438
+    },
+    {
+      "epoch": 0.11193268740438551,
+      "grad_norm": 0.5912401676177979,
+      "learning_rate": 1.9792441835385247e-05,
+      "loss": 0.5447284579277039,
+      "step": 439
+    },
+    {
+      "epoch": 0.11218765935747067,
+      "grad_norm": 0.5678877830505371,
+      "learning_rate": 1.979072895493695e-05,
+      "loss": 0.5309037566184998,
+      "step": 440
+    },
+    {
+      "epoch": 0.11244263131055583,
+      "grad_norm": 0.5953914523124695,
+      "learning_rate": 1.9789009110433636e-05,
+      "loss": 0.5362557172775269,
+      "step": 441
+    },
+    {
+      "epoch": 0.112697603263641,
+      "grad_norm": 0.5275906920433044,
+      "learning_rate": 1.9787282303098618e-05,
+      "loss": 0.5531718730926514,
+      "step": 442
+    },
+    {
+      "epoch": 0.11295257521672616,
+      "grad_norm": 0.5663650631904602,
+      "learning_rate": 1.978554853416016e-05,
+      "loss": 0.5560038089752197,
+      "step": 443
+    },
+    {
+      "epoch": 0.11320754716981132,
+      "grad_norm": 0.5504028797149658,
+      "learning_rate": 1.978380780485147e-05,
+      "loss": 0.5488899946212769,
+      "step": 444
+    },
+    {
+      "epoch": 0.11346251912289648,
+      "grad_norm": 0.5290588736534119,
+      "learning_rate": 1.9782060116410718e-05,
+      "loss": 0.5435810089111328,
+      "step": 445
+    },
+    {
+      "epoch": 0.11371749107598164,
+      "grad_norm": 0.5408740639686584,
+      "learning_rate": 1.9780305470081017e-05,
+      "loss": 0.5513544082641602,
+      "step": 446
+    },
+    {
+      "epoch": 0.1139724630290668,
+      "grad_norm": 0.5024471879005432,
+      "learning_rate": 1.9778543867110428e-05,
+      "loss": 0.5466337203979492,
+      "step": 447
+    },
+    {
+      "epoch": 0.11422743498215196,
+      "grad_norm": 0.539569616317749,
+      "learning_rate": 1.9776775308751964e-05,
+      "loss": 0.5356583595275879,
+      "step": 448
+    },
+    {
+      "epoch": 0.11448240693523712,
+      "grad_norm": 0.525956392288208,
+      "learning_rate": 1.9774999796263588e-05,
+      "loss": 0.5408869385719299,
+      "step": 449
+    },
+    {
+      "epoch": 0.11473737888832228,
+      "grad_norm": 0.5690717697143555,
+      "learning_rate": 1.9773217330908205e-05,
+      "loss": 0.5401495695114136,
+      "step": 450
+    },
+    {
+      "epoch": 0.11499235084140745,
+      "grad_norm": 0.5895713567733765,
+      "learning_rate": 1.9771427913953662e-05,
+      "loss": 0.5400240421295166,
+      "step": 451
+    },
+    {
+      "epoch": 0.11524732279449261,
+      "grad_norm": 0.6181904077529907,
+      "learning_rate": 1.9769631546672756e-05,
+      "loss": 0.5605360269546509,
+      "step": 452
+    },
+    {
+      "epoch": 0.11550229474757777,
+      "grad_norm": 0.5505849123001099,
+      "learning_rate": 1.9767828230343228e-05,
+      "loss": 0.5433098077774048,
+      "step": 453
+    },
+    {
+      "epoch": 0.11575726670066293,
+      "grad_norm": 0.6274793744087219,
+      "learning_rate": 1.9766017966247763e-05,
+      "loss": 0.5562180876731873,
+      "step": 454
+    },
+    {
+      "epoch": 0.11601223865374809,
+      "grad_norm": 0.5555331110954285,
+      "learning_rate": 1.9764200755673976e-05,
+      "loss": 0.5439101457595825,
+      "step": 455
+    },
+    {
+      "epoch": 0.11626721060683325,
+      "grad_norm": 0.5220648050308228,
+      "learning_rate": 1.976237659991444e-05,
+      "loss": 0.5579456090927124,
+      "step": 456
+    },
+    {
+      "epoch": 0.1165221825599184,
+      "grad_norm": 0.5761458277702332,
+      "learning_rate": 1.9760545500266656e-05,
+      "loss": 0.5493845343589783,
+      "step": 457
+    },
+    {
+      "epoch": 0.11677715451300356,
+      "grad_norm": 0.5373028516769409,
+      "learning_rate": 1.9758707458033073e-05,
+      "loss": 0.5442485809326172,
+      "step": 458
+    },
+    {
+      "epoch": 0.11703212646608872,
+      "grad_norm": 0.4915338456630707,
+      "learning_rate": 1.975686247452107e-05,
+      "loss": 0.5558599233627319,
+      "step": 459
+    },
+    {
+      "epoch": 0.1172870984191739,
+      "grad_norm": 0.5607727766036987,
+      "learning_rate": 1.975501055104297e-05,
+      "loss": 0.5386903882026672,
+      "step": 460
+    },
+    {
+      "epoch": 0.11754207037225906,
+      "grad_norm": 0.5363933444023132,
+      "learning_rate": 1.9753151688916025e-05,
+      "loss": 0.5452768802642822,
+      "step": 461
+    },
+    {
+      "epoch": 0.11779704232534421,
+      "grad_norm": 0.5088923573493958,
+      "learning_rate": 1.9751285889462424e-05,
+      "loss": 0.5445141196250916,
+      "step": 462
+    },
+    {
+      "epoch": 0.11805201427842937,
+      "grad_norm": 0.5186890959739685,
+      "learning_rate": 1.97494131540093e-05,
+      "loss": 0.5501380562782288,
+      "step": 463
+    },
+    {
+      "epoch": 0.11830698623151453,
+      "grad_norm": 0.5275077223777771,
+      "learning_rate": 1.9747533483888707e-05,
+      "loss": 0.5424678325653076,
+      "step": 464
+    },
+    {
+      "epoch": 0.11856195818459969,
+      "grad_norm": 0.5187524557113647,
+      "learning_rate": 1.9745646880437642e-05,
+      "loss": 0.5477076768875122,
+      "step": 465
+    },
+    {
+      "epoch": 0.11881693013768485,
+      "grad_norm": 0.46981313824653625,
+      "learning_rate": 1.9743753344998024e-05,
+      "loss": 0.5471561551094055,
+      "step": 466
+    },
+    {
+      "epoch": 0.11907190209077001,
+      "grad_norm": 0.62347012758255,
+      "learning_rate": 1.974185287891671e-05,
+      "loss": 0.5401801466941833,
+      "step": 467
+    },
+    {
+      "epoch": 0.11932687404385518,
+      "grad_norm": 0.5400267243385315,
+      "learning_rate": 1.973994548354548e-05,
+      "loss": 0.5439719557762146,
+      "step": 468
+    },
+    {
+      "epoch": 0.11958184599694034,
+      "grad_norm": 0.5315633416175842,
+      "learning_rate": 1.9738031160241054e-05,
+      "loss": 0.5322132110595703,
+      "step": 469
+    },
+    {
+      "epoch": 0.1198368179500255,
+      "grad_norm": 0.5405152440071106,
+      "learning_rate": 1.9736109910365066e-05,
+      "loss": 0.5460124015808105,
+      "step": 470
+    },
+    {
+      "epoch": 0.12009178990311066,
+      "grad_norm": 0.5464207530021667,
+      "learning_rate": 1.9734181735284083e-05,
+      "loss": 0.547633171081543,
+      "step": 471
+    },
+    {
+      "epoch": 0.12034676185619582,
+      "grad_norm": 0.49561744928359985,
+      "learning_rate": 1.9732246636369605e-05,
+      "loss": 0.5439903736114502,
+      "step": 472
+    },
+    {
+      "epoch": 0.12060173380928098,
+      "grad_norm": 0.526726245880127,
+      "learning_rate": 1.9730304614998043e-05,
+      "loss": 0.5577775239944458,
+      "step": 473
+    },
+    {
+      "epoch": 0.12085670576236614,
+      "grad_norm": 0.5296305418014526,
+      "learning_rate": 1.9728355672550743e-05,
+      "loss": 0.5440423488616943,
+      "step": 474
+    },
+    {
+      "epoch": 0.1211116777154513,
+      "grad_norm": 0.5025987029075623,
+      "learning_rate": 1.9726399810413967e-05,
+      "loss": 0.5496869683265686,
+      "step": 475
+    },
+    {
+      "epoch": 0.12136664966853646,
+      "grad_norm": 0.5133289694786072,
+      "learning_rate": 1.9724437029978903e-05,
+      "loss": 0.5514233112335205,
+      "step": 476
+    },
+    {
+      "epoch": 0.12162162162162163,
+      "grad_norm": 0.5568787455558777,
+      "learning_rate": 1.9722467332641656e-05,
+      "loss": 0.5393109321594238,
+      "step": 477
+    },
+    {
+      "epoch": 0.12187659357470679,
+      "grad_norm": 0.5228671431541443,
+      "learning_rate": 1.972049071980326e-05,
+      "loss": 0.525666356086731,
+      "step": 478
+    },
+    {
+      "epoch": 0.12213156552779195,
+      "grad_norm": 0.5343871712684631,
+      "learning_rate": 1.9718507192869656e-05,
+      "loss": 0.5391042828559875,
+      "step": 479
+    },
+    {
+      "epoch": 0.1223865374808771,
+      "grad_norm": 0.5969329476356506,
+      "learning_rate": 1.971651675325171e-05,
+      "loss": 0.561370313167572,
+      "step": 480
+    },
+    {
+      "epoch": 0.12264150943396226,
+      "grad_norm": 0.5148950219154358,
+      "learning_rate": 1.97145194023652e-05,
+      "loss": 0.5295721888542175,
+      "step": 481
+    },
+    {
+      "epoch": 0.12289648138704742,
+      "grad_norm": 0.5520386099815369,
+      "learning_rate": 1.971251514163083e-05,
+      "loss": 0.5467032790184021,
+      "step": 482
+    },
+    {
+      "epoch": 0.12315145334013258,
+      "grad_norm": 0.5354210138320923,
+      "learning_rate": 1.9710503972474205e-05,
+      "loss": 0.5428868532180786,
+      "step": 483
+    },
+    {
+      "epoch": 0.12340642529321774,
+      "grad_norm": 0.5027364492416382,
+      "learning_rate": 1.970848589632586e-05,
+      "loss": 0.5456262826919556,
+      "step": 484
+    },
+    {
+      "epoch": 0.1236613972463029,
+      "grad_norm": 0.5828258395195007,
+      "learning_rate": 1.9706460914621225e-05,
+      "loss": 0.5422083735466003,
+      "step": 485
+    },
+    {
+      "epoch": 0.12391636919938807,
+      "grad_norm": 0.5776222348213196,
+      "learning_rate": 1.970442902880065e-05,
+      "loss": 0.5387598276138306,
+      "step": 486
+    },
+    {
+      "epoch": 0.12417134115247323,
+      "grad_norm": 0.5405538082122803,
+      "learning_rate": 1.9702390240309402e-05,
+      "loss": 0.5427982807159424,
+      "step": 487
+    },
+    {
+      "epoch": 0.12442631310555839,
+      "grad_norm": 0.5617440938949585,
+      "learning_rate": 1.9700344550597656e-05,
+      "loss": 0.5464503765106201,
+      "step": 488
+    },
+    {
+      "epoch": 0.12468128505864355,
+      "grad_norm": 0.5526654124259949,
+      "learning_rate": 1.9698291961120485e-05,
+      "loss": 0.5374647378921509,
+      "step": 489
+    },
+    {
+      "epoch": 0.12493625701172871,
+      "grad_norm": 0.5415617227554321,
+      "learning_rate": 1.9696232473337878e-05,
+      "loss": 0.5434882044792175,
+      "step": 490
+    },
+    {
+      "epoch": 0.12519122896481388,
+      "grad_norm": 0.555722713470459,
+      "learning_rate": 1.9694166088714733e-05,
+      "loss": 0.5455443859100342,
+      "step": 491
+    },
+    {
+      "epoch": 0.12544620091789904,
+      "grad_norm": 0.5384701490402222,
+      "learning_rate": 1.9692092808720846e-05,
+      "loss": 0.5436517000198364,
+      "step": 492
+    },
+    {
+      "epoch": 0.1257011728709842,
+      "grad_norm": 0.5429685711860657,
+      "learning_rate": 1.9690012634830925e-05,
+      "loss": 0.5500330924987793,
+      "step": 493
+    },
+    {
+      "epoch": 0.12595614482406936,
+      "grad_norm": 0.547223687171936,
+      "learning_rate": 1.9687925568524577e-05,
+      "loss": 0.5479034781455994,
+      "step": 494
+    },
+    {
+      "epoch": 0.12621111677715452,
+      "grad_norm": 0.5522466897964478,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 0.550029993057251,
+      "step": 495
+    },
+    {
+      "epoch": 0.12646608873023968,
+      "grad_norm": 0.5701102614402771,
+      "learning_rate": 1.9683730764605547e-05,
+      "loss": 0.5450952053070068,
+      "step": 496
+    },
+    {
+      "epoch": 0.12672106068332484,
+      "grad_norm": 0.5533367395401001,
+      "learning_rate": 1.968162302997659e-05,
+      "loss": 0.5455372333526611,
+      "step": 497
+    },
+    {
+      "epoch": 0.12697603263641,
+      "grad_norm": 0.5364789962768555,
+      "learning_rate": 1.9679508408898656e-05,
+      "loss": 0.5422064065933228,
+      "step": 498
+    },
+    {
+      "epoch": 0.12723100458949516,
+      "grad_norm": 0.5047568678855896,
+      "learning_rate": 1.967738690287585e-05,
+      "loss": 0.5391058921813965,
+      "step": 499
+    },
+    {
+      "epoch": 0.12748597654258031,
+      "grad_norm": 0.5302197933197021,
+      "learning_rate": 1.967525851341719e-05,
+      "loss": 0.5551446676254272,
+      "step": 500
+    },
+    {
+      "epoch": 0.12774094849566547,
+      "grad_norm": 0.4994960129261017,
+      "learning_rate": 1.967312324203657e-05,
+      "loss": 0.5429335832595825,
+      "step": 501
+    },
+    {
+      "epoch": 0.12799592044875063,
+      "grad_norm": 0.5566155314445496,
+      "learning_rate": 1.9670981090252794e-05,
+      "loss": 0.5487710237503052,
+      "step": 502
+    },
+    {
+      "epoch": 0.1282508924018358,
+      "grad_norm": 0.5283164381980896,
+      "learning_rate": 1.9668832059589552e-05,
+      "loss": 0.5358626842498779,
+      "step": 503
+    },
+    {
+      "epoch": 0.12850586435492095,
+      "grad_norm": 0.5359975099563599,
+      "learning_rate": 1.966667615157543e-05,
+      "loss": 0.5301843881607056,
+      "step": 504
+    },
+    {
+      "epoch": 0.1287608363080061,
+      "grad_norm": 0.48307645320892334,
+      "learning_rate": 1.9664513367743908e-05,
+      "loss": 0.5526771545410156,
+      "step": 505
+    },
+    {
+      "epoch": 0.12901580826109127,
+      "grad_norm": 0.5109233260154724,
+      "learning_rate": 1.9662343709633352e-05,
+      "loss": 0.5420863628387451,
+      "step": 506
+    },
+    {
+      "epoch": 0.12927078021417643,
+      "grad_norm": 0.4935479164123535,
+      "learning_rate": 1.966016717878702e-05,
+      "loss": 0.5470730066299438,
+      "step": 507
+    },
+    {
+      "epoch": 0.1295257521672616,
+      "grad_norm": 0.4998207986354828,
+      "learning_rate": 1.965798377675306e-05,
+      "loss": 0.5346579551696777,
+      "step": 508
+    },
+    {
+      "epoch": 0.12978072412034677,
+      "grad_norm": 0.5249969959259033,
+      "learning_rate": 1.9655793505084503e-05,
+      "loss": 0.5412853956222534,
+      "step": 509
+    },
+    {
+      "epoch": 0.13003569607343193,
+      "grad_norm": 0.5389821529388428,
+      "learning_rate": 1.965359636533927e-05,
+      "loss": 0.5390466451644897,
+      "step": 510
+    },
+    {
+      "epoch": 0.1302906680265171,
+      "grad_norm": 0.539438009262085,
+      "learning_rate": 1.9651392359080164e-05,
+      "loss": 0.5388988852500916,
+      "step": 511
+    },
+    {
+      "epoch": 0.13054563997960225,
+      "grad_norm": 0.483715295791626,
+      "learning_rate": 1.964918148787488e-05,
+      "loss": 0.5375275611877441,
+      "step": 512
+    },
+    {
+      "epoch": 0.1308006119326874,
+      "grad_norm": 0.5151199102401733,
+      "learning_rate": 1.9646963753295987e-05,
+      "loss": 0.5366615653038025,
+      "step": 513
+    },
+    {
+      "epoch": 0.13105558388577257,
+      "grad_norm": 0.5603130459785461,
+      "learning_rate": 1.964473915692094e-05,
+      "loss": 0.5489786863327026,
+      "step": 514
+    },
+    {
+      "epoch": 0.13131055583885773,
+      "grad_norm": 0.5356683731079102,
+      "learning_rate": 1.9642507700332068e-05,
+      "loss": 0.5374578833580017,
+      "step": 515
+    },
+    {
+      "epoch": 0.1315655277919429,
+      "grad_norm": 0.4916025996208191,
+      "learning_rate": 1.964026938511659e-05,
+      "loss": 0.5560331344604492,
+      "step": 516
+    },
+    {
+      "epoch": 0.13182049974502805,
+      "grad_norm": 0.5539834499359131,
+      "learning_rate": 1.9638024212866606e-05,
+      "loss": 0.5416618585586548,
+      "step": 517
+    },
+    {
+      "epoch": 0.1320754716981132,
+      "grad_norm": 0.5391684174537659,
+      "learning_rate": 1.9635772185179074e-05,
+      "loss": 0.5419684052467346,
+      "step": 518
+    },
+    {
+      "epoch": 0.13233044365119837,
+      "grad_norm": 0.5286181569099426,
+      "learning_rate": 1.963351330365585e-05,
+      "loss": 0.5357035398483276,
+      "step": 519
+    },
+    {
+      "epoch": 0.13258541560428352,
+      "grad_norm": 0.5388613343238831,
+      "learning_rate": 1.963124756990365e-05,
+      "loss": 0.5408596396446228,
+      "step": 520
+    },
+    {
+      "epoch": 0.13284038755736868,
+      "grad_norm": 0.5261210203170776,
+      "learning_rate": 1.962897498553407e-05,
+      "loss": 0.5369002819061279,
+      "step": 521
+    },
+    {
+      "epoch": 0.13309535951045384,
+      "grad_norm": 0.5403190851211548,
+      "learning_rate": 1.962669555216358e-05,
+      "loss": 0.5399401783943176,
+      "step": 522
+    },
+    {
+      "epoch": 0.133350331463539,
+      "grad_norm": 0.5058906078338623,
+      "learning_rate": 1.962440927141352e-05,
+      "loss": 0.5308097004890442,
+      "step": 523
+    },
+    {
+      "epoch": 0.13360530341662416,
+      "grad_norm": 0.5357410907745361,
+      "learning_rate": 1.9622116144910095e-05,
+      "loss": 0.5340980887413025,
+      "step": 524
+    },
+    {
+      "epoch": 0.13386027536970932,
+      "grad_norm": 0.5216020345687866,
+      "learning_rate": 1.9619816174284393e-05,
+      "loss": 0.5346090793609619,
+      "step": 525
+    },
+    {
+      "epoch": 0.1341152473227945,
+      "grad_norm": 0.5459076762199402,
+      "learning_rate": 1.9617509361172356e-05,
+      "loss": 0.53227698802948,
+      "step": 526
+    },
+    {
+      "epoch": 0.13437021927587967,
+      "grad_norm": 0.5702166557312012,
+      "learning_rate": 1.9615195707214805e-05,
+      "loss": 0.5388787984848022,
+      "step": 527
+    },
+    {
+      "epoch": 0.13462519122896482,
+      "grad_norm": 0.49151238799095154,
+      "learning_rate": 1.9612875214057414e-05,
+      "loss": 0.5381410121917725,
+      "step": 528
+    },
+    {
+      "epoch": 0.13488016318204998,
+      "grad_norm": 0.5661936402320862,
+      "learning_rate": 1.961054788335073e-05,
+      "loss": 0.55195152759552,
+      "step": 529
+    },
+    {
+      "epoch": 0.13513513513513514,
+      "grad_norm": 0.5514591932296753,
+      "learning_rate": 1.9608213716750164e-05,
+      "loss": 0.5301691293716431,
+      "step": 530
+    },
+    {
+      "epoch": 0.1353901070882203,
+      "grad_norm": 0.5307884216308594,
+      "learning_rate": 1.9605872715915986e-05,
+      "loss": 0.5460130572319031,
+      "step": 531
+    },
+    {
+      "epoch": 0.13564507904130546,
+      "grad_norm": 0.5439223647117615,
+      "learning_rate": 1.9603524882513326e-05,
+      "loss": 0.5416542887687683,
+      "step": 532
+    },
+    {
+      "epoch": 0.13590005099439062,
+      "grad_norm": 0.5474928021430969,
+      "learning_rate": 1.9601170218212183e-05,
+      "loss": 0.5408943891525269,
+      "step": 533
+    },
+    {
+      "epoch": 0.13615502294747578,
+      "grad_norm": 0.5338258743286133,
+      "learning_rate": 1.9598808724687398e-05,
+      "loss": 0.5379087924957275,
+      "step": 534
+    },
+    {
+      "epoch": 0.13640999490056094,
+      "grad_norm": 0.5032938122749329,
+      "learning_rate": 1.9596440403618688e-05,
+      "loss": 0.5256127715110779,
+      "step": 535
+    },
+    {
+      "epoch": 0.1366649668536461,
+      "grad_norm": 0.5296608805656433,
+      "learning_rate": 1.9594065256690614e-05,
+      "loss": 0.5323258638381958,
+      "step": 536
+    },
+    {
+      "epoch": 0.13691993880673126,
+      "grad_norm": 0.549721896648407,
+      "learning_rate": 1.9591683285592595e-05,
+      "loss": 0.5364140868186951,
+      "step": 537
+    },
+    {
+      "epoch": 0.13717491075981642,
+      "grad_norm": 0.5119733214378357,
+      "learning_rate": 1.9589294492018907e-05,
+      "loss": 0.5263822078704834,
+      "step": 538
+    },
+    {
+      "epoch": 0.13742988271290157,
+      "grad_norm": 0.5301274657249451,
+      "learning_rate": 1.9586898877668677e-05,
+      "loss": 0.5366970896720886,
+      "step": 539
+    },
+    {
+      "epoch": 0.13768485466598673,
+      "grad_norm": 0.5832563638687134,
+      "learning_rate": 1.9584496444245885e-05,
+      "loss": 0.5429117679595947,
+      "step": 540
+    },
+    {
+      "epoch": 0.1379398266190719,
+      "grad_norm": 0.5166734457015991,
+      "learning_rate": 1.9582087193459355e-05,
+      "loss": 0.5296197533607483,
+      "step": 541
+    },
+    {
+      "epoch": 0.13819479857215705,
+      "grad_norm": 0.5472302436828613,
+      "learning_rate": 1.957967112702277e-05,
+      "loss": 0.5465075969696045,
+      "step": 542
+    },
+    {
+      "epoch": 0.1384497705252422,
+      "grad_norm": 0.5155672430992126,
+      "learning_rate": 1.9577248246654653e-05,
+      "loss": 0.5279982089996338,
+      "step": 543
+    },
+    {
+      "epoch": 0.1387047424783274,
+      "grad_norm": 0.5144238471984863,
+      "learning_rate": 1.9574818554078373e-05,
+      "loss": 0.5338395833969116,
+      "step": 544
+    },
+    {
+      "epoch": 0.13895971443141256,
+      "grad_norm": 0.5261866450309753,
+      "learning_rate": 1.9572382051022154e-05,
+      "loss": 0.5367433428764343,
+      "step": 545
+    },
+    {
+      "epoch": 0.13921468638449772,
+      "grad_norm": 0.5071731209754944,
+      "learning_rate": 1.9569938739219054e-05,
+      "loss": 0.5295122265815735,
+      "step": 546
+    },
+    {
+      "epoch": 0.13946965833758287,
+      "grad_norm": 0.5251842737197876,
+      "learning_rate": 1.9567488620406984e-05,
+      "loss": 0.5354647636413574,
+      "step": 547
+    },
+    {
+      "epoch": 0.13972463029066803,
+      "grad_norm": 0.5434141159057617,
+      "learning_rate": 1.956503169632868e-05,
+      "loss": 0.5488410592079163,
+      "step": 548
+    },
+    {
+      "epoch": 0.1399796022437532,
+      "grad_norm": 0.5540900826454163,
+      "learning_rate": 1.9562567968731742e-05,
+      "loss": 0.5251233577728271,
+      "step": 549
+    },
+    {
+      "epoch": 0.14023457419683835,
+      "grad_norm": 0.4901406466960907,
+      "learning_rate": 1.9560097439368585e-05,
+      "loss": 0.5400822758674622,
+      "step": 550
+    },
+    {
+      "epoch": 0.1404895461499235,
+      "grad_norm": 0.5357276797294617,
+      "learning_rate": 1.9557620109996483e-05,
+      "loss": 0.528943657875061,
+      "step": 551
+    },
+    {
+      "epoch": 0.14074451810300867,
+      "grad_norm": 0.5773533582687378,
+      "learning_rate": 1.955513598237753e-05,
+      "loss": 0.5278090238571167,
+      "step": 552
+    },
+    {
+      "epoch": 0.14099949005609383,
+      "grad_norm": 0.4954776167869568,
+      "learning_rate": 1.9552645058278668e-05,
+      "loss": 0.5397365093231201,
+      "step": 553
+    },
+    {
+      "epoch": 0.141254462009179,
+      "grad_norm": 0.5578538775444031,
+      "learning_rate": 1.955014733947167e-05,
+      "loss": 0.5337741374969482,
+      "step": 554
+    },
+    {
+      "epoch": 0.14150943396226415,
+      "grad_norm": 0.5055522918701172,
+      "learning_rate": 1.9547642827733136e-05,
+      "loss": 0.5458861589431763,
+      "step": 555
+    },
+    {
+      "epoch": 0.1417644059153493,
+      "grad_norm": 0.4946596324443817,
+      "learning_rate": 1.9545131524844506e-05,
+      "loss": 0.5277013778686523,
+      "step": 556
+    },
+    {
+      "epoch": 0.14201937786843447,
+      "grad_norm": 0.5347071290016174,
+      "learning_rate": 1.954261343259204e-05,
+      "loss": 0.5519477128982544,
+      "step": 557
+    },
+    {
+      "epoch": 0.14227434982151962,
+      "grad_norm": 0.5242138504981995,
+      "learning_rate": 1.9540088552766837e-05,
+      "loss": 0.5345370173454285,
+      "step": 558
+    },
+    {
+      "epoch": 0.14252932177460478,
+      "grad_norm": 0.5394492149353027,
+      "learning_rate": 1.9537556887164823e-05,
+      "loss": 0.5402896404266357,
+      "step": 559
+    },
+    {
+      "epoch": 0.14278429372768994,
+      "grad_norm": 0.5234394073486328,
+      "learning_rate": 1.953501843758675e-05,
+      "loss": 0.5392587184906006,
+      "step": 560
+    },
+    {
+      "epoch": 0.1430392656807751,
+      "grad_norm": 0.511301577091217,
+      "learning_rate": 1.9532473205838184e-05,
+      "loss": 0.543667197227478,
+      "step": 561
+    },
+    {
+      "epoch": 0.1432942376338603,
+      "grad_norm": 0.5079246759414673,
+      "learning_rate": 1.9529921193729535e-05,
+      "loss": 0.5325132608413696,
+      "step": 562
+    },
+    {
+      "epoch": 0.14354920958694545,
+      "grad_norm": 0.5344441533088684,
+      "learning_rate": 1.9527362403076022e-05,
+      "loss": 0.5426052808761597,
+      "step": 563
+    },
+    {
+      "epoch": 0.1438041815400306,
+      "grad_norm": 0.5662505626678467,
+      "learning_rate": 1.9524796835697687e-05,
+      "loss": 0.5501530766487122,
+      "step": 564
+    },
+    {
+      "epoch": 0.14405915349311577,
+      "grad_norm": 0.479183554649353,
+      "learning_rate": 1.9522224493419396e-05,
+      "loss": 0.5392322540283203,
+      "step": 565
+    },
+    {
+      "epoch": 0.14431412544620092,
+      "grad_norm": 0.5352718234062195,
+      "learning_rate": 1.9519645378070833e-05,
+      "loss": 0.5259209871292114,
+      "step": 566
+    },
+    {
+      "epoch": 0.14456909739928608,
+      "grad_norm": 0.49504008889198303,
+      "learning_rate": 1.95170594914865e-05,
+      "loss": 0.5283001065254211,
+      "step": 567
+    },
+    {
+      "epoch": 0.14482406935237124,
+      "grad_norm": 0.5184426307678223,
+      "learning_rate": 1.9514466835505714e-05,
+      "loss": 0.5301646590232849,
+      "step": 568
+    },
+    {
+      "epoch": 0.1450790413054564,
+      "grad_norm": 0.550871729850769,
+      "learning_rate": 1.9511867411972606e-05,
+      "loss": 0.5460808873176575,
+      "step": 569
+    },
+    {
+      "epoch": 0.14533401325854156,
+      "grad_norm": 0.49909305572509766,
+      "learning_rate": 1.950926122273612e-05,
+      "loss": 0.5394014120101929,
+      "step": 570
+    },
+    {
+      "epoch": 0.14558898521162672,
+      "grad_norm": 0.5513647794723511,
+      "learning_rate": 1.9506648269650017e-05,
+      "loss": 0.5427008867263794,
+      "step": 571
+    },
+    {
+      "epoch": 0.14584395716471188,
+      "grad_norm": 0.48180994391441345,
+      "learning_rate": 1.9504028554572865e-05,
+      "loss": 0.5257720351219177,
+      "step": 572
+    },
+    {
+      "epoch": 0.14609892911779704,
+      "grad_norm": 0.6737922430038452,
+      "learning_rate": 1.9501402079368042e-05,
+      "loss": 0.5394344329833984,
+      "step": 573
+    },
+    {
+      "epoch": 0.1463539010708822,
+      "grad_norm": 0.48611074686050415,
+      "learning_rate": 1.949876884590374e-05,
+      "loss": 0.5424251556396484,
+      "step": 574
+    },
+    {
+      "epoch": 0.14660887302396736,
+      "grad_norm": 0.5731044411659241,
+      "learning_rate": 1.949612885605295e-05,
+      "loss": 0.5467298030853271,
+      "step": 575
+    },
+    {
+      "epoch": 0.14686384497705252,
+      "grad_norm": 0.4976442754268646,
+      "learning_rate": 1.9493482111693474e-05,
+      "loss": 0.5349081754684448,
+      "step": 576
+    },
+    {
+      "epoch": 0.14711881693013767,
+      "grad_norm": 0.49011242389678955,
+      "learning_rate": 1.9490828614707917e-05,
+      "loss": 0.5175978541374207,
+      "step": 577
+    },
+    {
+      "epoch": 0.14737378888322283,
+      "grad_norm": 0.5349713563919067,
+      "learning_rate": 1.9488168366983685e-05,
+      "loss": 0.5363706350326538,
+      "step": 578
+    },
+    {
+      "epoch": 0.147628760836308,
+      "grad_norm": 0.5079620480537415,
+      "learning_rate": 1.9485501370412988e-05,
+      "loss": 0.5362746119499207,
+      "step": 579
+    },
+    {
+      "epoch": 0.14788373278939318,
+      "grad_norm": 0.5510601997375488,
+      "learning_rate": 1.9482827626892838e-05,
+      "loss": 0.5300568342208862,
+      "step": 580
+    },
+    {
+      "epoch": 0.14813870474247834,
+      "grad_norm": 0.48822200298309326,
+      "learning_rate": 1.948014713832504e-05,
+      "loss": 0.5431665778160095,
+      "step": 581
+    },
+    {
+      "epoch": 0.1483936766955635,
+      "grad_norm": 0.5097545385360718,
+      "learning_rate": 1.947745990661621e-05,
+      "loss": 0.5365676879882812,
+      "step": 582
+    },
+    {
+      "epoch": 0.14864864864864866,
+      "grad_norm": 0.524516761302948,
+      "learning_rate": 1.9474765933677736e-05,
+      "loss": 0.5253903865814209,
+      "step": 583
+    },
+    {
+      "epoch": 0.14890362060173382,
+      "grad_norm": 0.5344324707984924,
+      "learning_rate": 1.9472065221425827e-05,
+      "loss": 0.5223650932312012,
+      "step": 584
+    },
+    {
+      "epoch": 0.14915859255481898,
+      "grad_norm": 0.5210974812507629,
+      "learning_rate": 1.9469357771781476e-05,
+      "loss": 0.5488340854644775,
+      "step": 585
+    },
+    {
+      "epoch": 0.14941356450790413,
+      "grad_norm": 0.5254884958267212,
+      "learning_rate": 1.946664358667046e-05,
+      "loss": 0.5375260710716248,
+      "step": 586
+    },
+    {
+      "epoch": 0.1496685364609893,
+      "grad_norm": 0.5536079406738281,
+      "learning_rate": 1.9463922668023358e-05,
+      "loss": 0.5502488017082214,
+      "step": 587
+    },
+    {
+      "epoch": 0.14992350841407445,
+      "grad_norm": 0.5063897967338562,
+      "learning_rate": 1.9461195017775534e-05,
+      "loss": 0.5332539081573486,
+      "step": 588
+    },
+    {
+      "epoch": 0.1501784803671596,
+      "grad_norm": 0.5346553921699524,
+      "learning_rate": 1.9458460637867137e-05,
+      "loss": 0.5380452871322632,
+      "step": 589
+    },
+    {
+      "epoch": 0.15043345232024477,
+      "grad_norm": 0.573293149471283,
+      "learning_rate": 1.945571953024311e-05,
+      "loss": 0.5183770656585693,
+      "step": 590
+    },
+    {
+      "epoch": 0.15068842427332993,
+      "grad_norm": 0.5627990365028381,
+      "learning_rate": 1.9452971696853175e-05,
+      "loss": 0.5426141619682312,
+      "step": 591
+    },
+    {
+      "epoch": 0.1509433962264151,
+      "grad_norm": 0.49688220024108887,
+      "learning_rate": 1.9450217139651845e-05,
+      "loss": 0.5325013399124146,
+      "step": 592
+    },
+    {
+      "epoch": 0.15119836817950025,
+      "grad_norm": 0.5401874780654907,
+      "learning_rate": 1.9447455860598407e-05,
+      "loss": 0.5306686758995056,
+      "step": 593
+    },
+    {
+      "epoch": 0.1514533401325854,
+      "grad_norm": 0.49251788854599,
+      "learning_rate": 1.9444687861656932e-05,
+      "loss": 0.5364015102386475,
+      "step": 594
+    },
+    {
+      "epoch": 0.15170831208567057,
+      "grad_norm": 0.4938039183616638,
+      "learning_rate": 1.9441913144796274e-05,
+      "loss": 0.5307292342185974,
+      "step": 595
+    },
+    {
+      "epoch": 0.15196328403875572,
+      "grad_norm": 0.5310829281806946,
+      "learning_rate": 1.9439131711990064e-05,
+      "loss": 0.524756669998169,
+      "step": 596
+    },
+    {
+      "epoch": 0.15221825599184088,
+      "grad_norm": 0.5013731718063354,
+      "learning_rate": 1.943634356521671e-05,
+      "loss": 0.5259018540382385,
+      "step": 597
+    },
+    {
+      "epoch": 0.15247322794492607,
+      "grad_norm": 0.49433621764183044,
+      "learning_rate": 1.9433548706459394e-05,
+      "loss": 0.540558934211731,
+      "step": 598
+    },
+    {
+      "epoch": 0.15272819989801123,
+      "grad_norm": 0.47782808542251587,
+      "learning_rate": 1.9430747137706075e-05,
+      "loss": 0.5381788015365601,
+      "step": 599
+    },
+    {
+      "epoch": 0.1529831718510964,
+      "grad_norm": 0.4966195821762085,
+      "learning_rate": 1.942793886094948e-05,
+      "loss": 0.5175549983978271,
+      "step": 600
+    },
+    {
+      "epoch": 0.15323814380418155,
+      "grad_norm": 0.5325585007667542,
+      "learning_rate": 1.942512387818711e-05,
+      "loss": 0.5350815653800964,
+      "step": 601
+    },
+    {
+      "epoch": 0.1534931157572667,
+      "grad_norm": 0.4902491867542267,
+      "learning_rate": 1.942230219142124e-05,
+      "loss": 0.5346603393554688,
+      "step": 602
+    },
+    {
+      "epoch": 0.15374808771035187,
+      "grad_norm": 0.5321222543716431,
+      "learning_rate": 1.941947380265891e-05,
+      "loss": 0.5547575950622559,
+      "step": 603
+    },
+    {
+      "epoch": 0.15400305966343703,
+      "grad_norm": 0.4908081591129303,
+      "learning_rate": 1.9416638713911918e-05,
+      "loss": 0.5460398197174072,
+      "step": 604
+    },
+    {
+      "epoch": 0.15425803161652218,
+      "grad_norm": 0.5266767144203186,
+      "learning_rate": 1.9413796927196843e-05,
+      "loss": 0.5413973331451416,
+      "step": 605
+    },
+    {
+      "epoch": 0.15451300356960734,
+      "grad_norm": 0.5171524882316589,
+      "learning_rate": 1.9410948444535026e-05,
+      "loss": 0.5335320234298706,
+      "step": 606
+    },
+    {
+      "epoch": 0.1547679755226925,
+      "grad_norm": 0.5076289176940918,
+      "learning_rate": 1.940809326795256e-05,
+      "loss": 0.5222631692886353,
+      "step": 607
+    },
+    {
+      "epoch": 0.15502294747577766,
+      "grad_norm": 0.5298922061920166,
+      "learning_rate": 1.9405231399480308e-05,
+      "loss": 0.5390610694885254,
+      "step": 608
+    },
+    {
+      "epoch": 0.15527791942886282,
+      "grad_norm": 0.4918667674064636,
+      "learning_rate": 1.9402362841153887e-05,
+      "loss": 0.5259802937507629,
+      "step": 609
+    },
+    {
+      "epoch": 0.15553289138194798,
+      "grad_norm": 0.49876874685287476,
+      "learning_rate": 1.9399487595013683e-05,
+      "loss": 0.5446267127990723,
+      "step": 610
+    },
+    {
+      "epoch": 0.15578786333503314,
+      "grad_norm": 0.5575383901596069,
+      "learning_rate": 1.9396605663104825e-05,
+      "loss": 0.5347251296043396,
+      "step": 611
+    },
+    {
+      "epoch": 0.1560428352881183,
+      "grad_norm": 0.47660186886787415,
+      "learning_rate": 1.9393717047477213e-05,
+      "loss": 0.5166308879852295,
+      "step": 612
+    },
+    {
+      "epoch": 0.15629780724120346,
+      "grad_norm": 0.5122316479682922,
+      "learning_rate": 1.9390821750185482e-05,
+      "loss": 0.5235748291015625,
+      "step": 613
+    },
+    {
+      "epoch": 0.15655277919428862,
+      "grad_norm": 0.5255897045135498,
+      "learning_rate": 1.9387919773289036e-05,
+      "loss": 0.531277060508728,
+      "step": 614
+    },
+    {
+      "epoch": 0.15680775114737378,
+      "grad_norm": 0.5205806493759155,
+      "learning_rate": 1.9385011118852023e-05,
+      "loss": 0.5403878688812256,
+      "step": 615
+    },
+    {
+      "epoch": 0.15706272310045896,
+      "grad_norm": 0.52583247423172,
+      "learning_rate": 1.9382095788943345e-05,
+      "loss": 0.5407297015190125,
+      "step": 616
+    },
+    {
+      "epoch": 0.15731769505354412,
+      "grad_norm": 0.49851149320602417,
+      "learning_rate": 1.9379173785636647e-05,
+      "loss": 0.5462212562561035,
+      "step": 617
+    },
+    {
+      "epoch": 0.15757266700662928,
+      "grad_norm": 0.5226295590400696,
+      "learning_rate": 1.9376245111010324e-05,
+      "loss": 0.5349191427230835,
+      "step": 618
+    },
+    {
+      "epoch": 0.15782763895971444,
+      "grad_norm": 0.4985732436180115,
+      "learning_rate": 1.937330976714751e-05,
+      "loss": 0.5181032419204712,
+      "step": 619
+    },
+    {
+      "epoch": 0.1580826109127996,
+      "grad_norm": 0.5089516043663025,
+      "learning_rate": 1.9370367756136093e-05,
+      "loss": 0.5246825218200684,
+      "step": 620
+    },
+    {
+      "epoch": 0.15833758286588476,
+      "grad_norm": 0.5574972629547119,
+      "learning_rate": 1.93674190800687e-05,
+      "loss": 0.5245094299316406,
+      "step": 621
+    },
+    {
+      "epoch": 0.15859255481896992,
+      "grad_norm": 0.5276218056678772,
+      "learning_rate": 1.9364463741042694e-05,
+      "loss": 0.542697012424469,
+      "step": 622
+    },
+    {
+      "epoch": 0.15884752677205508,
+      "grad_norm": 0.5665618777275085,
+      "learning_rate": 1.936150174116018e-05,
+      "loss": 0.5375609397888184,
+      "step": 623
+    },
+    {
+      "epoch": 0.15910249872514023,
+      "grad_norm": 0.5423240661621094,
+      "learning_rate": 1.9358533082528e-05,
+      "loss": 0.5459458827972412,
+      "step": 624
+    },
+    {
+      "epoch": 0.1593574706782254,
+      "grad_norm": 0.5253497958183289,
+      "learning_rate": 1.9355557767257738e-05,
+      "loss": 0.5167371034622192,
+      "step": 625
+    },
+    {
+      "epoch": 0.15961244263131055,
+      "grad_norm": 0.5296688675880432,
+      "learning_rate": 1.9352575797465704e-05,
+      "loss": 0.5493143200874329,
+      "step": 626
+    },
+    {
+      "epoch": 0.1598674145843957,
+      "grad_norm": 0.7724140882492065,
+      "learning_rate": 1.934958717527295e-05,
+      "loss": 0.5376444458961487,
+      "step": 627
+    },
+    {
+      "epoch": 0.16012238653748087,
+      "grad_norm": 0.5141907334327698,
+      "learning_rate": 1.9346591902805246e-05,
+      "loss": 0.5336338877677917,
+      "step": 628
+    },
+    {
+      "epoch": 0.16037735849056603,
+      "grad_norm": 0.49341434240341187,
+      "learning_rate": 1.9343589982193113e-05,
+      "loss": 0.529479444026947,
+      "step": 629
+    },
+    {
+      "epoch": 0.1606323304436512,
+      "grad_norm": 0.5087915062904358,
+      "learning_rate": 1.9340581415571783e-05,
+      "loss": 0.5335008502006531,
+      "step": 630
+    },
+    {
+      "epoch": 0.16088730239673635,
+      "grad_norm": 0.5580664873123169,
+      "learning_rate": 1.9337566205081225e-05,
+      "loss": 0.530624508857727,
+      "step": 631
+    },
+    {
+      "epoch": 0.1611422743498215,
+      "grad_norm": 0.49169811606407166,
+      "learning_rate": 1.9334544352866126e-05,
+      "loss": 0.5459813475608826,
+      "step": 632
+    },
+    {
+      "epoch": 0.1613972463029067,
+      "grad_norm": 0.49668213725090027,
+      "learning_rate": 1.9331515861075907e-05,
+      "loss": 0.5162781476974487,
+      "step": 633
+    },
+    {
+      "epoch": 0.16165221825599185,
+      "grad_norm": 0.5077573657035828,
+      "learning_rate": 1.9328480731864698e-05,
+      "loss": 0.5343323349952698,
+      "step": 634
+    },
+    {
+      "epoch": 0.161907190209077,
+      "grad_norm": 0.5330377817153931,
+      "learning_rate": 1.932543896739137e-05,
+      "loss": 0.5407090187072754,
+      "step": 635
+    },
+    {
+      "epoch": 0.16216216216216217,
+      "grad_norm": 0.5178587436676025,
+      "learning_rate": 1.932239056981949e-05,
+      "loss": 0.5372127890586853,
+      "step": 636
+    },
+    {
+      "epoch": 0.16241713411524733,
+      "grad_norm": 0.4959794878959656,
+      "learning_rate": 1.9319335541317363e-05,
+      "loss": 0.5220591425895691,
+      "step": 637
+    },
+    {
+      "epoch": 0.1626721060683325,
+      "grad_norm": 0.4854819178581238,
+      "learning_rate": 1.9316273884058e-05,
+      "loss": 0.544111430644989,
+      "step": 638
+    },
+    {
+      "epoch": 0.16292707802141765,
+      "grad_norm": 0.49497857689857483,
+      "learning_rate": 1.931320560021913e-05,
+      "loss": 0.5263684988021851,
+      "step": 639
+    },
+    {
+      "epoch": 0.1631820499745028,
+      "grad_norm": 0.5102571249008179,
+      "learning_rate": 1.9310130691983192e-05,
+      "loss": 0.5366497039794922,
+      "step": 640
+    },
+    {
+      "epoch": 0.16343702192758797,
+      "grad_norm": 0.5193189978599548,
+      "learning_rate": 1.9307049161537346e-05,
+      "loss": 0.5210189819335938,
+      "step": 641
+    },
+    {
+      "epoch": 0.16369199388067313,
+      "grad_norm": 0.5115967988967896,
+      "learning_rate": 1.9303961011073448e-05,
+      "loss": 0.532012403011322,
+      "step": 642
+    },
+    {
+      "epoch": 0.16394696583375828,
+      "grad_norm": 0.5676260590553284,
+      "learning_rate": 1.930086624278808e-05,
+      "loss": 0.523317813873291,
+      "step": 643
+    },
+    {
+      "epoch": 0.16420193778684344,
+      "grad_norm": 0.5026125907897949,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 0.539628267288208,
+      "step": 644
+    },
+    {
+      "epoch": 0.1644569097399286,
+      "grad_norm": 0.520626962184906,
+      "learning_rate": 1.929465686156274e-05,
+      "loss": 0.5272216200828552,
+      "step": 645
+    },
+    {
+      "epoch": 0.16471188169301376,
+      "grad_norm": 0.4822489619255066,
+      "learning_rate": 1.929154225303945e-05,
+      "loss": 0.5315180420875549,
+      "step": 646
+    },
+    {
+      "epoch": 0.16496685364609892,
+      "grad_norm": 0.5027172565460205,
+      "learning_rate": 1.928842103552803e-05,
+      "loss": 0.5313969254493713,
+      "step": 647
+    },
+    {
+      "epoch": 0.16522182559918408,
+      "grad_norm": 0.5247225761413574,
+      "learning_rate": 1.9285293211248578e-05,
+      "loss": 0.5298991203308105,
+      "step": 648
+    },
+    {
+      "epoch": 0.16547679755226924,
+      "grad_norm": 0.5090324878692627,
+      "learning_rate": 1.9282158782425884e-05,
+      "loss": 0.5201632976531982,
+      "step": 649
+    },
+    {
+      "epoch": 0.1657317695053544,
+      "grad_norm": 0.5293415784835815,
+      "learning_rate": 1.9279017751289438e-05,
+      "loss": 0.5382990837097168,
+      "step": 650
+    },
+    {
+      "epoch": 0.16598674145843959,
+      "grad_norm": 0.5205861330032349,
+      "learning_rate": 1.9275870120073426e-05,
+      "loss": 0.5317997932434082,
+      "step": 651
+    },
+    {
+      "epoch": 0.16624171341152474,
+      "grad_norm": 0.5162908434867859,
+      "learning_rate": 1.9272715891016736e-05,
+      "loss": 0.5323415994644165,
+      "step": 652
+    },
+    {
+      "epoch": 0.1664966853646099,
+      "grad_norm": 0.5015779733657837,
+      "learning_rate": 1.926955506636293e-05,
+      "loss": 0.5262338519096375,
+      "step": 653
+    },
+    {
+      "epoch": 0.16675165731769506,
+      "grad_norm": 0.49459603428840637,
+      "learning_rate": 1.9266387648360285e-05,
+      "loss": 0.5312216281890869,
+      "step": 654
+    },
+    {
+      "epoch": 0.16700662927078022,
+      "grad_norm": 0.5384743213653564,
+      "learning_rate": 1.926321363926175e-05,
+      "loss": 0.5280896425247192,
+      "step": 655
+    },
+    {
+      "epoch": 0.16726160122386538,
+      "grad_norm": 0.49762842059135437,
+      "learning_rate": 1.9260033041324964e-05,
+      "loss": 0.5348333120346069,
+      "step": 656
+    },
+    {
+      "epoch": 0.16751657317695054,
+      "grad_norm": 0.5026784539222717,
+      "learning_rate": 1.9256845856812266e-05,
+      "loss": 0.5293280482292175,
+      "step": 657
+    },
+    {
+      "epoch": 0.1677715451300357,
+      "grad_norm": 0.5365166068077087,
+      "learning_rate": 1.9253652087990667e-05,
+      "loss": 0.5373440980911255,
+      "step": 658
+    },
+    {
+      "epoch": 0.16802651708312086,
+      "grad_norm": 0.5215746164321899,
+      "learning_rate": 1.9250451737131866e-05,
+      "loss": 0.536878228187561,
+      "step": 659
+    },
+    {
+      "epoch": 0.16828148903620602,
+      "grad_norm": 0.4788154363632202,
+      "learning_rate": 1.9247244806512244e-05,
+      "loss": 0.5261363983154297,
+      "step": 660
+    },
+    {
+      "epoch": 0.16853646098929118,
+      "grad_norm": 0.4960463345050812,
+      "learning_rate": 1.9244031298412857e-05,
+      "loss": 0.5358573198318481,
+      "step": 661
+    },
+    {
+      "epoch": 0.16879143294237633,
+      "grad_norm": 0.5835654735565186,
+      "learning_rate": 1.9240811215119447e-05,
+      "loss": 0.5243600606918335,
+      "step": 662
+    },
+    {
+      "epoch": 0.1690464048954615,
+      "grad_norm": 0.4894512891769409,
+      "learning_rate": 1.9237584558922425e-05,
+      "loss": 0.5246574878692627,
+      "step": 663
+    },
+    {
+      "epoch": 0.16930137684854665,
+      "grad_norm": 0.519819974899292,
+      "learning_rate": 1.9234351332116888e-05,
+      "loss": 0.52531898021698,
+      "step": 664
+    },
+    {
+      "epoch": 0.1695563488016318,
+      "grad_norm": 0.46499520540237427,
+      "learning_rate": 1.92311115370026e-05,
+      "loss": 0.5293796062469482,
+      "step": 665
+    },
+    {
+      "epoch": 0.16981132075471697,
+      "grad_norm": 0.484613835811615,
+      "learning_rate": 1.9227865175883993e-05,
+      "loss": 0.5152720212936401,
+      "step": 666
+    },
+    {
+      "epoch": 0.17006629270780213,
+      "grad_norm": 0.49043798446655273,
+      "learning_rate": 1.9224612251070176e-05,
+      "loss": 0.5403691530227661,
+      "step": 667
+    },
+    {
+      "epoch": 0.1703212646608873,
+      "grad_norm": 0.5324785113334656,
+      "learning_rate": 1.9221352764874923e-05,
+      "loss": 0.5379540920257568,
+      "step": 668
+    },
+    {
+      "epoch": 0.17057623661397248,
+      "grad_norm": 0.4996960163116455,
+      "learning_rate": 1.9218086719616678e-05,
+      "loss": 0.5312554240226746,
+      "step": 669
+    },
+    {
+      "epoch": 0.17083120856705764,
+      "grad_norm": 0.46016421914100647,
+      "learning_rate": 1.9214814117618547e-05,
+      "loss": 0.5201725363731384,
+      "step": 670
+    },
+    {
+      "epoch": 0.1710861805201428,
+      "grad_norm": 0.499889999628067,
+      "learning_rate": 1.9211534961208303e-05,
+      "loss": 0.5231727957725525,
+      "step": 671
+    },
+    {
+      "epoch": 0.17134115247322795,
+      "grad_norm": 0.6887993216514587,
+      "learning_rate": 1.920824925271838e-05,
+      "loss": 0.5276427268981934,
+      "step": 672
+    },
+    {
+      "epoch": 0.1715961244263131,
+      "grad_norm": 0.5240216851234436,
+      "learning_rate": 1.9204956994485872e-05,
+      "loss": 0.5339590311050415,
+      "step": 673
+    },
+    {
+      "epoch": 0.17185109637939827,
+      "grad_norm": 0.5527732372283936,
+      "learning_rate": 1.920165818885253e-05,
+      "loss": 0.5293232798576355,
+      "step": 674
+    },
+    {
+      "epoch": 0.17210606833248343,
+      "grad_norm": 0.5374642610549927,
+      "learning_rate": 1.9198352838164767e-05,
+      "loss": 0.539370059967041,
+      "step": 675
+    },
+    {
+      "epoch": 0.1723610402855686,
+      "grad_norm": 0.658922553062439,
+      "learning_rate": 1.9195040944773645e-05,
+      "loss": 0.5308823585510254,
+      "step": 676
+    },
+    {
+      "epoch": 0.17261601223865375,
+      "grad_norm": 0.5424158573150635,
+      "learning_rate": 1.9191722511034884e-05,
+      "loss": 0.5404328107833862,
+      "step": 677
+    },
+    {
+      "epoch": 0.1728709841917389,
+      "grad_norm": 0.5352630019187927,
+      "learning_rate": 1.918839753930886e-05,
+      "loss": 0.5331385135650635,
+      "step": 678
+    },
+    {
+      "epoch": 0.17312595614482407,
+      "grad_norm": 0.5205698013305664,
+      "learning_rate": 1.9185066031960587e-05,
+      "loss": 0.5314624309539795,
+      "step": 679
+    },
+    {
+      "epoch": 0.17338092809790923,
+      "grad_norm": 0.5151124596595764,
+      "learning_rate": 1.918172799135974e-05,
+      "loss": 0.5420811176300049,
+      "step": 680
+    },
+    {
+      "epoch": 0.17363590005099439,
+      "grad_norm": 0.522510290145874,
+      "learning_rate": 1.9178383419880635e-05,
+      "loss": 0.5371588468551636,
+      "step": 681
+    },
+    {
+      "epoch": 0.17389087200407954,
+      "grad_norm": 0.5063507556915283,
+      "learning_rate": 1.9175032319902236e-05,
+      "loss": 0.5372850298881531,
+      "step": 682
+    },
+    {
+      "epoch": 0.1741458439571647,
+      "grad_norm": 0.5166487693786621,
+      "learning_rate": 1.917167469380815e-05,
+      "loss": 0.5451241731643677,
+      "step": 683
+    },
+    {
+      "epoch": 0.17440081591024986,
+      "grad_norm": 0.5515632033348083,
+      "learning_rate": 1.916831054398662e-05,
+      "loss": 0.5475760698318481,
+      "step": 684
+    },
+    {
+      "epoch": 0.17465578786333502,
+      "grad_norm": 0.5165284872055054,
+      "learning_rate": 1.9164939872830542e-05,
+      "loss": 0.5366044044494629,
+      "step": 685
+    },
+    {
+      "epoch": 0.17491075981642018,
+      "grad_norm": 0.5325884222984314,
+      "learning_rate": 1.916156268273744e-05,
+      "loss": 0.5388437509536743,
+      "step": 686
+    },
+    {
+      "epoch": 0.17516573176950537,
+      "grad_norm": 0.5325320959091187,
+      "learning_rate": 1.9158178976109475e-05,
+      "loss": 0.5334592461585999,
+      "step": 687
+    },
+    {
+      "epoch": 0.17542070372259053,
+      "grad_norm": 0.5373082160949707,
+      "learning_rate": 1.9154788755353454e-05,
+      "loss": 0.5266925692558289,
+      "step": 688
+    },
+    {
+      "epoch": 0.17567567567567569,
+      "grad_norm": 0.5505048036575317,
+      "learning_rate": 1.9151392022880802e-05,
+      "loss": 0.5321172475814819,
+      "step": 689
+    },
+    {
+      "epoch": 0.17593064762876084,
+      "grad_norm": 0.5130444169044495,
+      "learning_rate": 1.9147988781107588e-05,
+      "loss": 0.5240427851676941,
+      "step": 690
+    },
+    {
+      "epoch": 0.176185619581846,
+      "grad_norm": 0.5584560036659241,
+      "learning_rate": 1.9144579032454503e-05,
+      "loss": 0.5241004228591919,
+      "step": 691
+    },
+    {
+      "epoch": 0.17644059153493116,
+      "grad_norm": 0.46925339102745056,
+      "learning_rate": 1.9141162779346875e-05,
+      "loss": 0.5358484983444214,
+      "step": 692
+    },
+    {
+      "epoch": 0.17669556348801632,
+      "grad_norm": 0.5069420337677002,
+      "learning_rate": 1.9137740024214647e-05,
+      "loss": 0.518207311630249,
+      "step": 693
+    },
+    {
+      "epoch": 0.17695053544110148,
+      "grad_norm": 0.5575243234634399,
+      "learning_rate": 1.91343107694924e-05,
+      "loss": 0.5154577493667603,
+      "step": 694
+    },
+    {
+      "epoch": 0.17720550739418664,
+      "grad_norm": 0.510741651058197,
+      "learning_rate": 1.913087501761932e-05,
+      "loss": 0.5157490968704224,
+      "step": 695
+    },
+    {
+      "epoch": 0.1774604793472718,
+      "grad_norm": 0.5032539367675781,
+      "learning_rate": 1.912743277103924e-05,
+      "loss": 0.5181795954704285,
+      "step": 696
+    },
+    {
+      "epoch": 0.17771545130035696,
+      "grad_norm": 0.49168163537979126,
+      "learning_rate": 1.9123984032200586e-05,
+      "loss": 0.5217337608337402,
+      "step": 697
+    },
+    {
+      "epoch": 0.17797042325344212,
+      "grad_norm": 0.5148937702178955,
+      "learning_rate": 1.912052880355642e-05,
+      "loss": 0.5191341638565063,
+      "step": 698
+    },
+    {
+      "epoch": 0.17822539520652728,
+      "grad_norm": 0.5000565052032471,
+      "learning_rate": 1.9117067087564413e-05,
+      "loss": 0.5333778262138367,
+      "step": 699
+    },
+    {
+      "epoch": 0.17848036715961244,
+      "grad_norm": 0.5110312700271606,
+      "learning_rate": 1.9113598886686856e-05,
+      "loss": 0.5120112895965576,
+      "step": 700
+    },
+    {
+      "epoch": 0.1787353391126976,
+      "grad_norm": 0.5078974962234497,
+      "learning_rate": 1.911012420339064e-05,
+      "loss": 0.5289940237998962,
+      "step": 701
+    },
+    {
+      "epoch": 0.17899031106578275,
+      "grad_norm": 0.5486764311790466,
+      "learning_rate": 1.910664304014728e-05,
+      "loss": 0.5259402990341187,
+      "step": 702
+    },
+    {
+      "epoch": 0.1792452830188679,
+      "grad_norm": 0.5564773678779602,
+      "learning_rate": 1.910315539943289e-05,
+      "loss": 0.5347332954406738,
+      "step": 703
+    },
+    {
+      "epoch": 0.17950025497195307,
+      "grad_norm": 0.6226455569267273,
+      "learning_rate": 1.9099661283728207e-05,
+      "loss": 0.5321149826049805,
+      "step": 704
+    },
+    {
+      "epoch": 0.17975522692503826,
+      "grad_norm": 0.5020618438720703,
+      "learning_rate": 1.9096160695518552e-05,
+      "loss": 0.5225280523300171,
+      "step": 705
+    },
+    {
+      "epoch": 0.18001019887812342,
+      "grad_norm": 0.5472486019134521,
+      "learning_rate": 1.9092653637293873e-05,
+      "loss": 0.5333778858184814,
+      "step": 706
+    },
+    {
+      "epoch": 0.18026517083120858,
+      "grad_norm": 0.5323438048362732,
+      "learning_rate": 1.9089140111548695e-05,
+      "loss": 0.5389878153800964,
+      "step": 707
+    },
+    {
+      "epoch": 0.18052014278429374,
+      "grad_norm": 0.5173847079277039,
+      "learning_rate": 1.9085620120782165e-05,
+      "loss": 0.5251544713973999,
+      "step": 708
+    },
+    {
+      "epoch": 0.1807751147373789,
+      "grad_norm": 0.53608238697052,
+      "learning_rate": 1.9082093667498018e-05,
+      "loss": 0.5316119194030762,
+      "step": 709
+    },
+    {
+      "epoch": 0.18103008669046405,
+      "grad_norm": 0.5072662234306335,
+      "learning_rate": 1.907856075420459e-05,
+      "loss": 0.5360819101333618,
+      "step": 710
+    },
+    {
+      "epoch": 0.1812850586435492,
+      "grad_norm": 0.5388319492340088,
+      "learning_rate": 1.9075021383414804e-05,
+      "loss": 0.5342385172843933,
+      "step": 711
+    },
+    {
+      "epoch": 0.18154003059663437,
+      "grad_norm": 1.0042855739593506,
+      "learning_rate": 1.907147555764618e-05,
+      "loss": 0.5308269262313843,
+      "step": 712
+    },
+    {
+      "epoch": 0.18179500254971953,
+      "grad_norm": 0.5531407594680786,
+      "learning_rate": 1.9067923279420842e-05,
+      "loss": 0.5368529558181763,
+      "step": 713
+    },
+    {
+      "epoch": 0.1820499745028047,
+      "grad_norm": 0.5203389525413513,
+      "learning_rate": 1.9064364551265483e-05,
+      "loss": 0.5198056101799011,
+      "step": 714
+    },
+    {
+      "epoch": 0.18230494645588985,
+      "grad_norm": 0.5286828279495239,
+      "learning_rate": 1.9060799375711397e-05,
+      "loss": 0.5205039381980896,
+      "step": 715
+    },
+    {
+      "epoch": 0.182559918408975,
+      "grad_norm": 0.5192756056785583,
+      "learning_rate": 1.905722775529446e-05,
+      "loss": 0.5289533138275146,
+      "step": 716
+    },
+    {
+      "epoch": 0.18281489036206017,
+      "grad_norm": 0.5478755235671997,
+      "learning_rate": 1.9053649692555135e-05,
+      "loss": 0.5210616588592529,
+      "step": 717
+    },
+    {
+      "epoch": 0.18306986231514533,
+      "grad_norm": 0.7221400141716003,
+      "learning_rate": 1.905006519003846e-05,
+      "loss": 0.5364406108856201,
+      "step": 718
+    },
+    {
+      "epoch": 0.18332483426823049,
+      "grad_norm": 0.5132589340209961,
+      "learning_rate": 1.9046474250294062e-05,
+      "loss": 0.5295343399047852,
+      "step": 719
+    },
+    {
+      "epoch": 0.18357980622131564,
+      "grad_norm": 0.513512909412384,
+      "learning_rate": 1.9042876875876145e-05,
+      "loss": 0.5361637473106384,
+      "step": 720
+    },
+    {
+      "epoch": 0.1838347781744008,
+      "grad_norm": 0.5017802715301514,
+      "learning_rate": 1.9039273069343482e-05,
+      "loss": 0.514184296131134,
+      "step": 721
+    },
+    {
+      "epoch": 0.18408975012748596,
+      "grad_norm": 0.5017067193984985,
+      "learning_rate": 1.9035662833259433e-05,
+      "loss": 0.5232157707214355,
+      "step": 722
+    },
+    {
+      "epoch": 0.18434472208057115,
+      "grad_norm": 0.5659171938896179,
+      "learning_rate": 1.903204617019192e-05,
+      "loss": 0.5253515839576721,
+      "step": 723
+    },
+    {
+      "epoch": 0.1845996940336563,
+      "grad_norm": 0.5655650496482849,
+      "learning_rate": 1.902842308271345e-05,
+      "loss": 0.532926082611084,
+      "step": 724
+    },
+    {
+      "epoch": 0.18485466598674147,
+      "grad_norm": 0.6155320405960083,
+      "learning_rate": 1.9024793573401087e-05,
+      "loss": 0.5291966795921326,
+      "step": 725
+    },
+    {
+      "epoch": 0.18510963793982663,
+      "grad_norm": 0.5772908926010132,
+      "learning_rate": 1.9021157644836465e-05,
+      "loss": 0.5200927257537842,
+      "step": 726
+    },
+    {
+      "epoch": 0.18536460989291179,
+      "grad_norm": 0.4874703884124756,
+      "learning_rate": 1.901751529960579e-05,
+      "loss": 0.5172476768493652,
+      "step": 727
+    },
+    {
+      "epoch": 0.18561958184599694,
+      "grad_norm": 0.5260182619094849,
+      "learning_rate": 1.9013866540299825e-05,
+      "loss": 0.5145378112792969,
+      "step": 728
+    },
+    {
+      "epoch": 0.1858745537990821,
+      "grad_norm": 0.48625022172927856,
+      "learning_rate": 1.9010211369513905e-05,
+      "loss": 0.5356984734535217,
+      "step": 729
+    },
+    {
+      "epoch": 0.18612952575216726,
+      "grad_norm": 0.5057938098907471,
+      "learning_rate": 1.9006549789847912e-05,
+      "loss": 0.5326985120773315,
+      "step": 730
+    },
+    {
+      "epoch": 0.18638449770525242,
+      "grad_norm": 0.5073822140693665,
+      "learning_rate": 1.90028818039063e-05,
+      "loss": 0.5180009007453918,
+      "step": 731
+    },
+    {
+      "epoch": 0.18663946965833758,
+      "grad_norm": 0.47638195753097534,
+      "learning_rate": 1.899920741429807e-05,
+      "loss": 0.5337942838668823,
+      "step": 732
+    },
+    {
+      "epoch": 0.18689444161142274,
+      "grad_norm": 0.4857496917247772,
+      "learning_rate": 1.899552662363678e-05,
+      "loss": 0.5289947986602783,
+      "step": 733
+    },
+    {
+      "epoch": 0.1871494135645079,
+      "grad_norm": 0.483283668756485,
+      "learning_rate": 1.8991839434540543e-05,
+      "loss": 0.5122606754302979,
+      "step": 734
+    },
+    {
+      "epoch": 0.18740438551759306,
+      "grad_norm": 0.5082046985626221,
+      "learning_rate": 1.898814584963203e-05,
+      "loss": 0.5219171047210693,
+      "step": 735
+    },
+    {
+      "epoch": 0.18765935747067822,
+      "grad_norm": 0.7273216247558594,
+      "learning_rate": 1.8984445871538444e-05,
+      "loss": 0.5157103538513184,
+      "step": 736
+    },
+    {
+      "epoch": 0.18791432942376338,
+      "grad_norm": 0.49670591950416565,
+      "learning_rate": 1.8980739502891548e-05,
+      "loss": 0.5254105925559998,
+      "step": 737
+    },
+    {
+      "epoch": 0.18816930137684854,
+      "grad_norm": 0.47980067133903503,
+      "learning_rate": 1.897702674632765e-05,
+      "loss": 0.5299422740936279,
+      "step": 738
+    },
+    {
+      "epoch": 0.1884242733299337,
+      "grad_norm": 0.49665823578834534,
+      "learning_rate": 1.89733076044876e-05,
+      "loss": 0.5355211496353149,
+      "step": 739
+    },
+    {
+      "epoch": 0.18867924528301888,
+      "grad_norm": 0.4740928113460541,
+      "learning_rate": 1.8969582080016783e-05,
+      "loss": 0.5235376358032227,
+      "step": 740
+    },
+    {
+      "epoch": 0.18893421723610404,
+      "grad_norm": 0.49630486965179443,
+      "learning_rate": 1.8965850175565136e-05,
+      "loss": 0.5099401473999023,
+      "step": 741
+    },
+    {
+      "epoch": 0.1891891891891892,
+      "grad_norm": 0.47335001826286316,
+      "learning_rate": 1.896211189378713e-05,
+      "loss": 0.5353489518165588,
+      "step": 742
+    },
+    {
+      "epoch": 0.18944416114227436,
+      "grad_norm": 0.5141160488128662,
+      "learning_rate": 1.8958367237341763e-05,
+      "loss": 0.5258992314338684,
+      "step": 743
+    },
+    {
+      "epoch": 0.18969913309535952,
+      "grad_norm": 0.4717693030834198,
+      "learning_rate": 1.895461620889258e-05,
+      "loss": 0.5221823453903198,
+      "step": 744
+    },
+    {
+      "epoch": 0.18995410504844468,
+      "grad_norm": 0.49536171555519104,
+      "learning_rate": 1.8950858811107654e-05,
+      "loss": 0.5208817720413208,
+      "step": 745
+    },
+    {
+      "epoch": 0.19020907700152984,
+      "grad_norm": 0.482179194688797,
+      "learning_rate": 1.894709504665958e-05,
+      "loss": 0.5294069051742554,
+      "step": 746
+    },
+    {
+      "epoch": 0.190464048954615,
+      "grad_norm": 0.49157753586769104,
+      "learning_rate": 1.8943324918225495e-05,
+      "loss": 0.5172535181045532,
+      "step": 747
+    },
+    {
+      "epoch": 0.19071902090770015,
+      "grad_norm": 0.47853702306747437,
+      "learning_rate": 1.8939548428487055e-05,
+      "loss": 0.5317491292953491,
+      "step": 748
+    },
+    {
+      "epoch": 0.1909739928607853,
+      "grad_norm": 0.5791231989860535,
+      "learning_rate": 1.8935765580130443e-05,
+      "loss": 0.5277394652366638,
+      "step": 749
+    },
+    {
+      "epoch": 0.19122896481387047,
+      "grad_norm": 0.5110471248626709,
+      "learning_rate": 1.8931976375846363e-05,
+      "loss": 0.5369372367858887,
+      "step": 750
+    },
+    {
+      "epoch": 0.19148393676695563,
+      "grad_norm": 0.5030074119567871,
+      "learning_rate": 1.892818081833004e-05,
+      "loss": 0.5264465808868408,
+      "step": 751
+    },
+    {
+      "epoch": 0.1917389087200408,
+      "grad_norm": 0.554189145565033,
+      "learning_rate": 1.892437891028122e-05,
+      "loss": 0.5304807424545288,
+      "step": 752
+    },
+    {
+      "epoch": 0.19199388067312595,
+      "grad_norm": 0.518322229385376,
+      "learning_rate": 1.8920570654404168e-05,
+      "loss": 0.510017454624176,
+      "step": 753
+    },
+    {
+      "epoch": 0.1922488526262111,
+      "grad_norm": 0.47035935521125793,
+      "learning_rate": 1.8916756053407658e-05,
+      "loss": 0.518936276435852,
+      "step": 754
+    },
+    {
+      "epoch": 0.19250382457929627,
+      "grad_norm": 0.5004586577415466,
+      "learning_rate": 1.8912935110004977e-05,
+      "loss": 0.5358216762542725,
+      "step": 755
+    },
+    {
+      "epoch": 0.19275879653238143,
+      "grad_norm": 0.5754180550575256,
+      "learning_rate": 1.8909107826913933e-05,
+      "loss": 0.521419107913971,
+      "step": 756
+    },
+    {
+      "epoch": 0.19301376848546659,
+      "grad_norm": 0.4933522939682007,
+      "learning_rate": 1.890527420685684e-05,
+      "loss": 0.5265296697616577,
+      "step": 757
+    },
+    {
+      "epoch": 0.19326874043855177,
+      "grad_norm": 0.48939067125320435,
+      "learning_rate": 1.8901434252560503e-05,
+      "loss": 0.5180717706680298,
+      "step": 758
+    },
+    {
+      "epoch": 0.19352371239163693,
+      "grad_norm": 0.5148706436157227,
+      "learning_rate": 1.8897587966756258e-05,
+      "loss": 0.5236243009567261,
+      "step": 759
+    },
+    {
+      "epoch": 0.1937786843447221,
+      "grad_norm": 0.5004563331604004,
+      "learning_rate": 1.889373535217993e-05,
+      "loss": 0.509172797203064,
+      "step": 760
+    },
+    {
+      "epoch": 0.19403365629780725,
+      "grad_norm": 0.46056416630744934,
+      "learning_rate": 1.8889876411571847e-05,
+      "loss": 0.5105539560317993,
+      "step": 761
+    },
+    {
+      "epoch": 0.1942886282508924,
+      "grad_norm": 0.4904102385044098,
+      "learning_rate": 1.8886011147676835e-05,
+      "loss": 0.5207321047782898,
+      "step": 762
+    },
+    {
+      "epoch": 0.19454360020397757,
+      "grad_norm": 0.48944613337516785,
+      "learning_rate": 1.8882139563244225e-05,
+      "loss": 0.5243609547615051,
+      "step": 763
+    },
+    {
+      "epoch": 0.19479857215706273,
+      "grad_norm": 0.4759974777698517,
+      "learning_rate": 1.8878261661027838e-05,
+      "loss": 0.5158101916313171,
+      "step": 764
+    },
+    {
+      "epoch": 0.1950535441101479,
+      "grad_norm": 0.525944709777832,
+      "learning_rate": 1.8874377443785986e-05,
+      "loss": 0.5378414392471313,
+      "step": 765
+    },
+    {
+      "epoch": 0.19530851606323305,
+      "grad_norm": 0.5273529887199402,
+      "learning_rate": 1.887048691428148e-05,
+      "loss": 0.5196689963340759,
+      "step": 766
+    },
+    {
+      "epoch": 0.1955634880163182,
+      "grad_norm": 0.4938674569129944,
+      "learning_rate": 1.8866590075281624e-05,
+      "loss": 0.5200862288475037,
+      "step": 767
+    },
+    {
+      "epoch": 0.19581845996940336,
+      "grad_norm": 0.5003823041915894,
+      "learning_rate": 1.8862686929558195e-05,
+      "loss": 0.5321277976036072,
+      "step": 768
+    },
+    {
+      "epoch": 0.19607343192248852,
+      "grad_norm": 0.4937255084514618,
+      "learning_rate": 1.8858777479887465e-05,
+      "loss": 0.5314494967460632,
+      "step": 769
+    },
+    {
+      "epoch": 0.19632840387557368,
+      "grad_norm": 0.5254429578781128,
+      "learning_rate": 1.8854861729050194e-05,
+      "loss": 0.5185518860816956,
+      "step": 770
+    },
+    {
+      "epoch": 0.19658337582865884,
+      "grad_norm": 0.4962460994720459,
+      "learning_rate": 1.8850939679831614e-05,
+      "loss": 0.5256330966949463,
+      "step": 771
+    },
+    {
+      "epoch": 0.196838347781744,
+      "grad_norm": 0.49898064136505127,
+      "learning_rate": 1.8847011335021447e-05,
+      "loss": 0.5047429800033569,
+      "step": 772
+    },
+    {
+      "epoch": 0.19709331973482916,
+      "grad_norm": 0.4960572123527527,
+      "learning_rate": 1.8843076697413888e-05,
+      "loss": 0.5207228660583496,
+      "step": 773
+    },
+    {
+      "epoch": 0.19734829168791432,
+      "grad_norm": 0.5059972405433655,
+      "learning_rate": 1.8839135769807605e-05,
+      "loss": 0.5412514209747314,
+      "step": 774
+    },
+    {
+      "epoch": 0.19760326364099948,
+      "grad_norm": 0.4811801016330719,
+      "learning_rate": 1.8835188555005744e-05,
+      "loss": 0.52277010679245,
+      "step": 775
+    },
+    {
+      "epoch": 0.19785823559408466,
+      "grad_norm": 0.5361527800559998,
+      "learning_rate": 1.8831235055815927e-05,
+      "loss": 0.5182678699493408,
+      "step": 776
+    },
+    {
+      "epoch": 0.19811320754716982,
+      "grad_norm": 0.48322221636772156,
+      "learning_rate": 1.8827275275050233e-05,
+      "loss": 0.513138473033905,
+      "step": 777
+    },
+    {
+      "epoch": 0.19836817950025498,
+      "grad_norm": 0.4969165027141571,
+      "learning_rate": 1.8823309215525227e-05,
+      "loss": 0.5136932134628296,
+      "step": 778
+    },
+    {
+      "epoch": 0.19862315145334014,
+      "grad_norm": 0.4899584949016571,
+      "learning_rate": 1.8819336880061926e-05,
+      "loss": 0.5196957588195801,
+      "step": 779
+    },
+    {
+      "epoch": 0.1988781234064253,
+      "grad_norm": 0.49217143654823303,
+      "learning_rate": 1.8815358271485814e-05,
+      "loss": 0.5142471790313721,
+      "step": 780
+    },
+    {
+      "epoch": 0.19913309535951046,
+      "grad_norm": 0.4637305736541748,
+      "learning_rate": 1.881137339262684e-05,
+      "loss": 0.5263211131095886,
+      "step": 781
+    },
+    {
+      "epoch": 0.19938806731259562,
+      "grad_norm": 0.5127052664756775,
+      "learning_rate": 1.8807382246319413e-05,
+      "loss": 0.5331568717956543,
+      "step": 782
+    },
+    {
+      "epoch": 0.19964303926568078,
+      "grad_norm": 0.4882904589176178,
+      "learning_rate": 1.8803384835402395e-05,
+      "loss": 0.5182445049285889,
+      "step": 783
+    },
+    {
+      "epoch": 0.19989801121876594,
+      "grad_norm": 0.511396586894989,
+      "learning_rate": 1.879938116271911e-05,
+      "loss": 0.529838502407074,
+      "step": 784
+    },
+    {
+      "epoch": 0.2001529831718511,
+      "grad_norm": 0.476744681596756,
+      "learning_rate": 1.8795371231117334e-05,
+      "loss": 0.5218862295150757,
+      "step": 785
+    },
+    {
+      "epoch": 0.20040795512493625,
+      "grad_norm": 0.46153387427330017,
+      "learning_rate": 1.879135504344929e-05,
+      "loss": 0.5229381322860718,
+      "step": 786
+    },
+    {
+      "epoch": 0.2006629270780214,
+      "grad_norm": 0.5047249794006348,
+      "learning_rate": 1.8787332602571663e-05,
+      "loss": 0.5238070487976074,
+      "step": 787
+    },
+    {
+      "epoch": 0.20091789903110657,
+      "grad_norm": 0.48882344365119934,
+      "learning_rate": 1.878330391134557e-05,
+      "loss": 0.5123788714408875,
+      "step": 788
+    },
+    {
+      "epoch": 0.20117287098419173,
+      "grad_norm": 0.5000123977661133,
+      "learning_rate": 1.8779268972636587e-05,
+      "loss": 0.5192744731903076,
+      "step": 789
+    },
+    {
+      "epoch": 0.2014278429372769,
+      "grad_norm": 0.4846855700016022,
+      "learning_rate": 1.8775227789314723e-05,
+      "loss": 0.5199167728424072,
+      "step": 790
+    },
+    {
+      "epoch": 0.20168281489036205,
+      "grad_norm": 0.4933943450450897,
+      "learning_rate": 1.8771180364254443e-05,
+      "loss": 0.513075590133667,
+      "step": 791
+    },
+    {
+      "epoch": 0.2019377868434472,
+      "grad_norm": 0.4882349371910095,
+      "learning_rate": 1.8767126700334633e-05,
+      "loss": 0.5088496804237366,
+      "step": 792
+    },
+    {
+      "epoch": 0.20219275879653237,
+      "grad_norm": 0.5605803728103638,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 0.5345818996429443,
+      "step": 793
+    },
+    {
+      "epoch": 0.20244773074961755,
+      "grad_norm": 0.6046520471572876,
+      "learning_rate": 1.875900066745422e-05,
+      "loss": 0.522911787033081,
+      "step": 794
+    },
+    {
+      "epoch": 0.20270270270270271,
+      "grad_norm": 0.49711018800735474,
+      "learning_rate": 1.875492830427358e-05,
+      "loss": 0.5213525295257568,
+      "step": 795
+    },
+    {
+      "epoch": 0.20295767465578787,
+      "grad_norm": 0.4825911521911621,
+      "learning_rate": 1.8750849713793354e-05,
+      "loss": 0.5204763412475586,
+      "step": 796
+    },
+    {
+      "epoch": 0.20321264660887303,
+      "grad_norm": 0.4894961416721344,
+      "learning_rate": 1.874676489891461e-05,
+      "loss": 0.5249648690223694,
+      "step": 797
+    },
+    {
+      "epoch": 0.2034676185619582,
+      "grad_norm": 0.5199370980262756,
+      "learning_rate": 1.8742673862542834e-05,
+      "loss": 0.5049653053283691,
+      "step": 798
+    },
+    {
+      "epoch": 0.20372259051504335,
+      "grad_norm": 0.5881942510604858,
+      "learning_rate": 1.8738576607587946e-05,
+      "loss": 0.5105040073394775,
+      "step": 799
+    },
+    {
+      "epoch": 0.2039775624681285,
+      "grad_norm": 0.5093399882316589,
+      "learning_rate": 1.8734473136964283e-05,
+      "loss": 0.5335594415664673,
+      "step": 800
+    },
+    {
+      "epoch": 0.20423253442121367,
+      "grad_norm": 0.5274905562400818,
+      "learning_rate": 1.8730363453590607e-05,
+      "loss": 0.539413571357727,
+      "step": 801
+    },
+    {
+      "epoch": 0.20448750637429883,
+      "grad_norm": 0.4588984251022339,
+      "learning_rate": 1.87262475603901e-05,
+      "loss": 0.5119473934173584,
+      "step": 802
+    },
+    {
+      "epoch": 0.204742478327384,
+      "grad_norm": 0.5083139538764954,
+      "learning_rate": 1.872212546029035e-05,
+      "loss": 0.523451566696167,
+      "step": 803
+    },
+    {
+      "epoch": 0.20499745028046915,
+      "grad_norm": 0.5172069072723389,
+      "learning_rate": 1.8717997156223383e-05,
+      "loss": 0.49723517894744873,
+      "step": 804
+    },
+    {
+      "epoch": 0.2052524222335543,
+      "grad_norm": 0.5106121301651001,
+      "learning_rate": 1.8713862651125607e-05,
+      "loss": 0.5193416476249695,
+      "step": 805
+    },
+    {
+      "epoch": 0.20550739418663946,
+      "grad_norm": 0.5006569027900696,
+      "learning_rate": 1.870972194793787e-05,
+      "loss": 0.5207656621932983,
+      "step": 806
+    },
+    {
+      "epoch": 0.20576236613972462,
+      "grad_norm": 0.4751335680484772,
+      "learning_rate": 1.8705575049605415e-05,
+      "loss": 0.5251712799072266,
+      "step": 807
+    },
+    {
+      "epoch": 0.20601733809280978,
+      "grad_norm": 0.4934970438480377,
+      "learning_rate": 1.8701421959077884e-05,
+      "loss": 0.5144690871238708,
+      "step": 808
+    },
+    {
+      "epoch": 0.20627231004589494,
+      "grad_norm": 0.5313510298728943,
+      "learning_rate": 1.8697262679309335e-05,
+      "loss": 0.5420048236846924,
+      "step": 809
+    },
+    {
+      "epoch": 0.2065272819989801,
+      "grad_norm": 0.4750503599643707,
+      "learning_rate": 1.8693097213258235e-05,
+      "loss": 0.5261687636375427,
+      "step": 810
+    },
+    {
+      "epoch": 0.20678225395206526,
+      "grad_norm": 0.4985625743865967,
+      "learning_rate": 1.868892556388743e-05,
+      "loss": 0.5317679643630981,
+      "step": 811
+    },
+    {
+      "epoch": 0.20703722590515045,
+      "grad_norm": 0.5090624094009399,
+      "learning_rate": 1.8684747734164177e-05,
+      "loss": 0.5190541744232178,
+      "step": 812
+    },
+    {
+      "epoch": 0.2072921978582356,
+      "grad_norm": 0.48770126700401306,
+      "learning_rate": 1.8680563727060133e-05,
+      "loss": 0.5069824457168579,
+      "step": 813
+    },
+    {
+      "epoch": 0.20754716981132076,
+      "grad_norm": 0.49879923462867737,
+      "learning_rate": 1.867637354555134e-05,
+      "loss": 0.5261608362197876,
+      "step": 814
+    },
+    {
+      "epoch": 0.20780214176440592,
+      "grad_norm": 0.5008533000946045,
+      "learning_rate": 1.8672177192618234e-05,
+      "loss": 0.5157227516174316,
+      "step": 815
+    },
+    {
+      "epoch": 0.20805711371749108,
+      "grad_norm": 0.5074318051338196,
+      "learning_rate": 1.8667974671245645e-05,
+      "loss": 0.530713677406311,
+      "step": 816
+    },
+    {
+      "epoch": 0.20831208567057624,
+      "grad_norm": 0.48070028424263,
+      "learning_rate": 1.866376598442279e-05,
+      "loss": 0.5272051692008972,
+      "step": 817
+    },
+    {
+      "epoch": 0.2085670576236614,
+      "grad_norm": 0.5248457789421082,
+      "learning_rate": 1.8659551135143262e-05,
+      "loss": 0.5113994479179382,
+      "step": 818
+    },
+    {
+      "epoch": 0.20882202957674656,
+      "grad_norm": 0.5021365284919739,
+      "learning_rate": 1.865533012640505e-05,
+      "loss": 0.5125217437744141,
+      "step": 819
+    },
+    {
+      "epoch": 0.20907700152983172,
+      "grad_norm": 0.4999274015426636,
+      "learning_rate": 1.8651102961210516e-05,
+      "loss": 0.4997989535331726,
+      "step": 820
+    },
+    {
+      "epoch": 0.20933197348291688,
+      "grad_norm": 0.5285374522209167,
+      "learning_rate": 1.86468696425664e-05,
+      "loss": 0.5188378691673279,
+      "step": 821
+    },
+    {
+      "epoch": 0.20958694543600204,
+      "grad_norm": 0.49890968203544617,
+      "learning_rate": 1.8642630173483832e-05,
+      "loss": 0.5256074666976929,
+      "step": 822
+    },
+    {
+      "epoch": 0.2098419173890872,
+      "grad_norm": 0.49642476439476013,
+      "learning_rate": 1.8638384556978302e-05,
+      "loss": 0.5300315618515015,
+      "step": 823
+    },
+    {
+      "epoch": 0.21009688934217235,
+      "grad_norm": 0.52839595079422,
+      "learning_rate": 1.8634132796069674e-05,
+      "loss": 0.5289017558097839,
+      "step": 824
+    },
+    {
+      "epoch": 0.21035186129525751,
+      "grad_norm": 0.4960280656814575,
+      "learning_rate": 1.8629874893782195e-05,
+      "loss": 0.5117241144180298,
+      "step": 825
+    },
+    {
+      "epoch": 0.21060683324834267,
+      "grad_norm": 0.47426652908325195,
+      "learning_rate": 1.8625610853144463e-05,
+      "loss": 0.5090838670730591,
+      "step": 826
+    },
+    {
+      "epoch": 0.21086180520142783,
+      "grad_norm": 0.5177625417709351,
+      "learning_rate": 1.862134067718945e-05,
+      "loss": 0.5097435116767883,
+      "step": 827
+    },
+    {
+      "epoch": 0.211116777154513,
+      "grad_norm": 0.4960584044456482,
+      "learning_rate": 1.86170643689545e-05,
+      "loss": 0.5163490772247314,
+      "step": 828
+    },
+    {
+      "epoch": 0.21137174910759818,
+      "grad_norm": 0.5274242758750916,
+      "learning_rate": 1.861278193148131e-05,
+      "loss": 0.51758873462677,
+      "step": 829
+    },
+    {
+      "epoch": 0.21162672106068334,
+      "grad_norm": 0.533641517162323,
+      "learning_rate": 1.8608493367815933e-05,
+      "loss": 0.5184922218322754,
+      "step": 830
+    },
+    {
+      "epoch": 0.2118816930137685,
+      "grad_norm": 0.48885729908943176,
+      "learning_rate": 1.8604198681008793e-05,
+      "loss": 0.5227895975112915,
+      "step": 831
+    },
+    {
+      "epoch": 0.21213666496685366,
+      "grad_norm": 0.5027369856834412,
+      "learning_rate": 1.859989787411465e-05,
+      "loss": 0.5039633512496948,
+      "step": 832
+    },
+    {
+      "epoch": 0.21239163691993881,
+      "grad_norm": 0.49498140811920166,
+      "learning_rate": 1.859559095019264e-05,
+      "loss": 0.5215380191802979,
+      "step": 833
+    },
+    {
+      "epoch": 0.21264660887302397,
+      "grad_norm": 0.5174261331558228,
+      "learning_rate": 1.859127791230623e-05,
+      "loss": 0.5233101844787598,
+      "step": 834
+    },
+    {
+      "epoch": 0.21290158082610913,
+      "grad_norm": 0.49725356698036194,
+      "learning_rate": 1.8586958763523246e-05,
+      "loss": 0.5073238015174866,
+      "step": 835
+    },
+    {
+      "epoch": 0.2131565527791943,
+      "grad_norm": 0.48951879143714905,
+      "learning_rate": 1.858263350691586e-05,
+      "loss": 0.5236285924911499,
+      "step": 836
+    },
+    {
+      "epoch": 0.21341152473227945,
+      "grad_norm": 0.5008646845817566,
+      "learning_rate": 1.8578302145560586e-05,
+      "loss": 0.51987624168396,
+      "step": 837
+    },
+    {
+      "epoch": 0.2136664966853646,
+      "grad_norm": 0.5107278823852539,
+      "learning_rate": 1.857396468253828e-05,
+      "loss": 0.5350760221481323,
+      "step": 838
+    },
+    {
+      "epoch": 0.21392146863844977,
+      "grad_norm": 0.491146445274353,
+      "learning_rate": 1.856962112093414e-05,
+      "loss": 0.5397478342056274,
+      "step": 839
+    },
+    {
+      "epoch": 0.21417644059153493,
+      "grad_norm": 0.4717349708080292,
+      "learning_rate": 1.8565271463837706e-05,
+      "loss": 0.5235544443130493,
+      "step": 840
+    },
+    {
+      "epoch": 0.2144314125446201,
+      "grad_norm": 0.48198896646499634,
+      "learning_rate": 1.8560915714342842e-05,
+      "loss": 0.5057838559150696,
+      "step": 841
+    },
+    {
+      "epoch": 0.21468638449770525,
+      "grad_norm": 0.48079413175582886,
+      "learning_rate": 1.8556553875547755e-05,
+      "loss": 0.5073414444923401,
+      "step": 842
+    },
+    {
+      "epoch": 0.2149413564507904,
+      "grad_norm": 0.4871809780597687,
+      "learning_rate": 1.855218595055498e-05,
+      "loss": 0.5244035124778748,
+      "step": 843
+    },
+    {
+      "epoch": 0.21519632840387556,
+      "grad_norm": 0.47666579484939575,
+      "learning_rate": 1.8547811942471384e-05,
+      "loss": 0.515140950679779,
+      "step": 844
+    },
+    {
+      "epoch": 0.21545130035696072,
+      "grad_norm": 0.48149263858795166,
+      "learning_rate": 1.854343185440816e-05,
+      "loss": 0.5190153121948242,
+      "step": 845
+    },
+    {
+      "epoch": 0.21570627231004588,
+      "grad_norm": 0.5033882856369019,
+      "learning_rate": 1.8539045689480817e-05,
+      "loss": 0.5240796804428101,
+      "step": 846
+    },
+    {
+      "epoch": 0.21596124426313107,
+      "grad_norm": 0.47844886779785156,
+      "learning_rate": 1.85346534508092e-05,
+      "loss": 0.5184129476547241,
+      "step": 847
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 0.48677003383636475,
+      "learning_rate": 1.8530255141517465e-05,
+      "loss": 0.5040110349655151,
+      "step": 848
+    },
+    {
+      "epoch": 0.2164711881693014,
+      "grad_norm": 0.5304367542266846,
+      "learning_rate": 1.852585076473409e-05,
+      "loss": 0.504118800163269,
+      "step": 849
+    },
+    {
+      "epoch": 0.21672616012238655,
+      "grad_norm": 0.4905112087726593,
+      "learning_rate": 1.8521440323591866e-05,
+      "loss": 0.5151620507240295,
+      "step": 850
+    },
+    {
+      "epoch": 0.2169811320754717,
+      "grad_norm": 0.549328088760376,
+      "learning_rate": 1.85170238212279e-05,
+      "loss": 0.5425221920013428,
+      "step": 851
+    },
+    {
+      "epoch": 0.21723610402855686,
+      "grad_norm": 0.6872383952140808,
+      "learning_rate": 1.851260126078361e-05,
+      "loss": 0.5106749534606934,
+      "step": 852
+    },
+    {
+      "epoch": 0.21749107598164202,
+      "grad_norm": 0.4876777231693268,
+      "learning_rate": 1.850817264540472e-05,
+      "loss": 0.5124921202659607,
+      "step": 853
+    },
+    {
+      "epoch": 0.21774604793472718,
+      "grad_norm": 0.488199383020401,
+      "learning_rate": 1.8503737978241263e-05,
+      "loss": 0.5182828903198242,
+      "step": 854
+    },
+    {
+      "epoch": 0.21800101988781234,
+      "grad_norm": 0.4671249985694885,
+      "learning_rate": 1.849929726244758e-05,
+      "loss": 0.495769202709198,
+      "step": 855
+    },
+    {
+      "epoch": 0.2182559918408975,
+      "grad_norm": 0.5009353756904602,
+      "learning_rate": 1.849485050118231e-05,
+      "loss": 0.5188135504722595,
+      "step": 856
+    },
+    {
+      "epoch": 0.21851096379398266,
+      "grad_norm": 0.4932180643081665,
+      "learning_rate": 1.8490397697608394e-05,
+      "loss": 0.5201120972633362,
+      "step": 857
+    },
+    {
+      "epoch": 0.21876593574706782,
+      "grad_norm": 0.4824739992618561,
+      "learning_rate": 1.8485938854893067e-05,
+      "loss": 0.5199398398399353,
+      "step": 858
+    },
+    {
+      "epoch": 0.21902090770015298,
+      "grad_norm": 0.4766797721385956,
+      "learning_rate": 1.8481473976207866e-05,
+      "loss": 0.5209866762161255,
+      "step": 859
+    },
+    {
+      "epoch": 0.21927587965323814,
+      "grad_norm": 0.517713725566864,
+      "learning_rate": 1.8477003064728616e-05,
+      "loss": 0.5150636434555054,
+      "step": 860
+    },
+    {
+      "epoch": 0.2195308516063233,
+      "grad_norm": 0.4794352948665619,
+      "learning_rate": 1.8472526123635434e-05,
+      "loss": 0.5228415727615356,
+      "step": 861
+    },
+    {
+      "epoch": 0.21978582355940846,
+      "grad_norm": 0.5226784944534302,
+      "learning_rate": 1.846804315611273e-05,
+      "loss": 0.5156341791152954,
+      "step": 862
+    },
+    {
+      "epoch": 0.22004079551249361,
+      "grad_norm": 0.49431315064430237,
+      "learning_rate": 1.8463554165349195e-05,
+      "loss": 0.5137113928794861,
+      "step": 863
+    },
+    {
+      "epoch": 0.22029576746557877,
+      "grad_norm": 0.5030642151832581,
+      "learning_rate": 1.8459059154537806e-05,
+      "loss": 0.5218013525009155,
+      "step": 864
+    },
+    {
+      "epoch": 0.22055073941866396,
+      "grad_norm": 0.5052620768547058,
+      "learning_rate": 1.8454558126875825e-05,
+      "loss": 0.5037291646003723,
+      "step": 865
+    },
+    {
+      "epoch": 0.22080571137174912,
+      "grad_norm": 0.465748131275177,
+      "learning_rate": 1.845005108556479e-05,
+      "loss": 0.5203166007995605,
+      "step": 866
+    },
+    {
+      "epoch": 0.22106068332483428,
+      "grad_norm": 0.5093213319778442,
+      "learning_rate": 1.8445538033810515e-05,
+      "loss": 0.5137592554092407,
+      "step": 867
+    },
+    {
+      "epoch": 0.22131565527791944,
+      "grad_norm": 0.501781165599823,
+      "learning_rate": 1.8441018974823098e-05,
+      "loss": 0.5238536596298218,
+      "step": 868
+    },
+    {
+      "epoch": 0.2215706272310046,
+      "grad_norm": 0.48056527972221375,
+      "learning_rate": 1.8436493911816903e-05,
+      "loss": 0.5207821130752563,
+      "step": 869
+    },
+    {
+      "epoch": 0.22182559918408976,
+      "grad_norm": 0.4888818562030792,
+      "learning_rate": 1.8431962848010557e-05,
+      "loss": 0.5104517936706543,
+      "step": 870
+    },
+    {
+      "epoch": 0.22208057113717491,
+      "grad_norm": 0.5200516581535339,
+      "learning_rate": 1.8427425786626976e-05,
+      "loss": 0.5080585479736328,
+      "step": 871
+    },
+    {
+      "epoch": 0.22233554309026007,
+      "grad_norm": 0.5427224636077881,
+      "learning_rate": 1.8422882730893323e-05,
+      "loss": 0.4997643828392029,
+      "step": 872
+    },
+    {
+      "epoch": 0.22259051504334523,
+      "grad_norm": 0.4931401014328003,
+      "learning_rate": 1.8418333684041033e-05,
+      "loss": 0.5171279311180115,
+      "step": 873
+    },
+    {
+      "epoch": 0.2228454869964304,
+      "grad_norm": 0.5080557465553284,
+      "learning_rate": 1.8413778649305797e-05,
+      "loss": 0.5060961246490479,
+      "step": 874
+    },
+    {
+      "epoch": 0.22310045894951555,
+      "grad_norm": 0.5039787888526917,
+      "learning_rate": 1.8409217629927577e-05,
+      "loss": 0.5168291926383972,
+      "step": 875
+    },
+    {
+      "epoch": 0.2233554309026007,
+      "grad_norm": 0.4833467900753021,
+      "learning_rate": 1.8404650629150578e-05,
+      "loss": 0.5225592851638794,
+      "step": 876
+    },
+    {
+      "epoch": 0.22361040285568587,
+      "grad_norm": 0.5088043212890625,
+      "learning_rate": 1.8400077650223264e-05,
+      "loss": 0.5346219539642334,
+      "step": 877
+    },
+    {
+      "epoch": 0.22386537480877103,
+      "grad_norm": 0.5302433371543884,
+      "learning_rate": 1.839549869639836e-05,
+      "loss": 0.5219231843948364,
+      "step": 878
+    },
+    {
+      "epoch": 0.2241203467618562,
+      "grad_norm": 0.561053454875946,
+      "learning_rate": 1.8390913770932827e-05,
+      "loss": 0.5408127307891846,
+      "step": 879
+    },
+    {
+      "epoch": 0.22437531871494135,
+      "grad_norm": 0.47593629360198975,
+      "learning_rate": 1.8386322877087883e-05,
+      "loss": 0.5122065544128418,
+      "step": 880
+    },
+    {
+      "epoch": 0.2246302906680265,
+      "grad_norm": 0.5070580840110779,
+      "learning_rate": 1.838172601812899e-05,
+      "loss": 0.5167946815490723,
+      "step": 881
+    },
+    {
+      "epoch": 0.22488526262111166,
+      "grad_norm": 0.49819034337997437,
+      "learning_rate": 1.8377123197325843e-05,
+      "loss": 0.5193842649459839,
+      "step": 882
+    },
+    {
+      "epoch": 0.22514023457419685,
+      "grad_norm": 0.6639063358306885,
+      "learning_rate": 1.8372514417952398e-05,
+      "loss": 0.5219736099243164,
+      "step": 883
+    },
+    {
+      "epoch": 0.225395206527282,
+      "grad_norm": 0.5022265315055847,
+      "learning_rate": 1.8367899683286826e-05,
+      "loss": 0.5150482654571533,
+      "step": 884
+    },
+    {
+      "epoch": 0.22565017848036717,
+      "grad_norm": 0.5262812376022339,
+      "learning_rate": 1.8363278996611553e-05,
+      "loss": 0.5040405988693237,
+      "step": 885
+    },
+    {
+      "epoch": 0.22590515043345233,
+      "grad_norm": 0.48911675810813904,
+      "learning_rate": 1.8358652361213226e-05,
+      "loss": 0.5164714455604553,
+      "step": 886
+    },
+    {
+      "epoch": 0.2261601223865375,
+      "grad_norm": 0.48677948117256165,
+      "learning_rate": 1.8354019780382736e-05,
+      "loss": 0.514830470085144,
+      "step": 887
+    },
+    {
+      "epoch": 0.22641509433962265,
+      "grad_norm": 0.49844464659690857,
+      "learning_rate": 1.8349381257415185e-05,
+      "loss": 0.5102154016494751,
+      "step": 888
+    },
+    {
+      "epoch": 0.2266700662927078,
+      "grad_norm": 0.5603659152984619,
+      "learning_rate": 1.834473679560991e-05,
+      "loss": 0.5293639302253723,
+      "step": 889
+    },
+    {
+      "epoch": 0.22692503824579296,
+      "grad_norm": 0.538626492023468,
+      "learning_rate": 1.8340086398270486e-05,
+      "loss": 0.502727746963501,
+      "step": 890
+    },
+    {
+      "epoch": 0.22718001019887812,
+      "grad_norm": 0.4659905433654785,
+      "learning_rate": 1.8335430068704688e-05,
+      "loss": 0.5082594156265259,
+      "step": 891
+    },
+    {
+      "epoch": 0.22743498215196328,
+      "grad_norm": 0.4687235653400421,
+      "learning_rate": 1.8330767810224525e-05,
+      "loss": 0.502903938293457,
+      "step": 892
+    },
+    {
+      "epoch": 0.22768995410504844,
+      "grad_norm": 0.49669066071510315,
+      "learning_rate": 1.8326099626146214e-05,
+      "loss": 0.5199105143547058,
+      "step": 893
+    },
+    {
+      "epoch": 0.2279449260581336,
+      "grad_norm": 0.5097357034683228,
+      "learning_rate": 1.8321425519790193e-05,
+      "loss": 0.502683162689209,
+      "step": 894
+    },
+    {
+      "epoch": 0.22819989801121876,
+      "grad_norm": 0.5846214890480042,
+      "learning_rate": 1.8316745494481114e-05,
+      "loss": 0.5099455714225769,
+      "step": 895
+    },
+    {
+      "epoch": 0.22845486996430392,
+      "grad_norm": 0.4794066548347473,
+      "learning_rate": 1.8312059553547833e-05,
+      "loss": 0.5095470547676086,
+      "step": 896
+    },
+    {
+      "epoch": 0.22870984191738908,
+      "grad_norm": 0.5125945806503296,
+      "learning_rate": 1.8307367700323412e-05,
+      "loss": 0.5077476501464844,
+      "step": 897
+    },
+    {
+      "epoch": 0.22896481387047424,
+      "grad_norm": 0.46401146054267883,
+      "learning_rate": 1.830266993814513e-05,
+      "loss": 0.521671712398529,
+      "step": 898
+    },
+    {
+      "epoch": 0.2292197858235594,
+      "grad_norm": 0.46971526741981506,
+      "learning_rate": 1.829796627035446e-05,
+      "loss": 0.5065335035324097,
+      "step": 899
+    },
+    {
+      "epoch": 0.22947475777664456,
+      "grad_norm": 0.5173994898796082,
+      "learning_rate": 1.8293256700297072e-05,
+      "loss": 0.5077659487724304,
+      "step": 900
+    },
+    {
+      "epoch": 0.22972972972972974,
+      "grad_norm": 0.4828406572341919,
+      "learning_rate": 1.8288541231322853e-05,
+      "loss": 0.506639838218689,
+      "step": 901
+    },
+    {
+      "epoch": 0.2299847016828149,
+      "grad_norm": 0.4878862202167511,
+      "learning_rate": 1.8283819866785855e-05,
+      "loss": 0.5134996771812439,
+      "step": 902
+    },
+    {
+      "epoch": 0.23023967363590006,
+      "grad_norm": 0.5095584392547607,
+      "learning_rate": 1.827909261004435e-05,
+      "loss": 0.5173805356025696,
+      "step": 903
+    },
+    {
+      "epoch": 0.23049464558898522,
+      "grad_norm": 0.48897939920425415,
+      "learning_rate": 1.8274359464460796e-05,
+      "loss": 0.524322509765625,
+      "step": 904
+    },
+    {
+      "epoch": 0.23074961754207038,
+      "grad_norm": 0.47126173973083496,
+      "learning_rate": 1.826962043340183e-05,
+      "loss": 0.5017592310905457,
+      "step": 905
+    },
+    {
+      "epoch": 0.23100458949515554,
+      "grad_norm": 0.49605822563171387,
+      "learning_rate": 1.826487552023828e-05,
+      "loss": 0.5261249542236328,
+      "step": 906
+    },
+    {
+      "epoch": 0.2312595614482407,
+      "grad_norm": 0.5319313406944275,
+      "learning_rate": 1.8260124728345163e-05,
+      "loss": 0.5224013328552246,
+      "step": 907
+    },
+    {
+      "epoch": 0.23151453340132586,
+      "grad_norm": 0.47262606024742126,
+      "learning_rate": 1.825536806110167e-05,
+      "loss": 0.511128306388855,
+      "step": 908
+    },
+    {
+      "epoch": 0.23176950535441102,
+      "grad_norm": 0.4793703556060791,
+      "learning_rate": 1.825060552189118e-05,
+      "loss": 0.5372747778892517,
+      "step": 909
+    },
+    {
+      "epoch": 0.23202447730749617,
+      "grad_norm": 0.5034754872322083,
+      "learning_rate": 1.8245837114101235e-05,
+      "loss": 0.5220484733581543,
+      "step": 910
+    },
+    {
+      "epoch": 0.23227944926058133,
+      "grad_norm": 0.49909234046936035,
+      "learning_rate": 1.8241062841123564e-05,
+      "loss": 0.5091926455497742,
+      "step": 911
+    },
+    {
+      "epoch": 0.2325344212136665,
+      "grad_norm": 0.5578997731208801,
+      "learning_rate": 1.8236282706354064e-05,
+      "loss": 0.5100309252738953,
+      "step": 912
+    },
+    {
+      "epoch": 0.23278939316675165,
+      "grad_norm": 0.5156607031822205,
+      "learning_rate": 1.82314967131928e-05,
+      "loss": 0.5231087803840637,
+      "step": 913
+    },
+    {
+      "epoch": 0.2330443651198368,
+      "grad_norm": 0.478680819272995,
+      "learning_rate": 1.8226704865044e-05,
+      "loss": 0.508722722530365,
+      "step": 914
+    },
+    {
+      "epoch": 0.23329933707292197,
+      "grad_norm": 0.5336467027664185,
+      "learning_rate": 1.822190716531607e-05,
+      "loss": 0.5129567980766296,
+      "step": 915
+    },
+    {
+      "epoch": 0.23355430902600713,
+      "grad_norm": 0.471403032541275,
+      "learning_rate": 1.821710361742156e-05,
+      "loss": 0.5059640407562256,
+      "step": 916
+    },
+    {
+      "epoch": 0.2338092809790923,
+      "grad_norm": 0.48177075386047363,
+      "learning_rate": 1.82122942247772e-05,
+      "loss": 0.5262762308120728,
+      "step": 917
+    },
+    {
+      "epoch": 0.23406425293217745,
+      "grad_norm": 0.5124832391738892,
+      "learning_rate": 1.820747899080386e-05,
+      "loss": 0.5098938345909119,
+      "step": 918
+    },
+    {
+      "epoch": 0.23431922488526263,
+      "grad_norm": 0.4705912470817566,
+      "learning_rate": 1.8202657918926566e-05,
+      "loss": 0.49921292066574097,
+      "step": 919
+    },
+    {
+      "epoch": 0.2345741968383478,
+      "grad_norm": 0.5206021070480347,
+      "learning_rate": 1.819783101257451e-05,
+      "loss": 0.525272011756897,
+      "step": 920
+    },
+    {
+      "epoch": 0.23482916879143295,
+      "grad_norm": 0.5261361598968506,
+      "learning_rate": 1.8192998275181026e-05,
+      "loss": 0.5254808664321899,
+      "step": 921
+    },
+    {
+      "epoch": 0.2350841407445181,
+      "grad_norm": 0.5328448414802551,
+      "learning_rate": 1.8188159710183595e-05,
+      "loss": 0.5316905379295349,
+      "step": 922
+    },
+    {
+      "epoch": 0.23533911269760327,
+      "grad_norm": 0.4848942756652832,
+      "learning_rate": 1.8183315321023837e-05,
+      "loss": 0.5214201211929321,
+      "step": 923
+    },
+    {
+      "epoch": 0.23559408465068843,
+      "grad_norm": 0.49192479252815247,
+      "learning_rate": 1.817846511114753e-05,
+      "loss": 0.5270960330963135,
+      "step": 924
+    },
+    {
+      "epoch": 0.2358490566037736,
+      "grad_norm": 0.49427223205566406,
+      "learning_rate": 1.8173609084004577e-05,
+      "loss": 0.5243868827819824,
+      "step": 925
+    },
+    {
+      "epoch": 0.23610402855685875,
+      "grad_norm": 0.47979867458343506,
+      "learning_rate": 1.8168747243049026e-05,
+      "loss": 0.5152573585510254,
+      "step": 926
+    },
+    {
+      "epoch": 0.2363590005099439,
+      "grad_norm": 0.5014990568161011,
+      "learning_rate": 1.8163879591739067e-05,
+      "loss": 0.5148708820343018,
+      "step": 927
+    },
+    {
+      "epoch": 0.23661397246302907,
+      "grad_norm": 0.5115143656730652,
+      "learning_rate": 1.8159006133537004e-05,
+      "loss": 0.5039348602294922,
+      "step": 928
+    },
+    {
+      "epoch": 0.23686894441611422,
+      "grad_norm": 0.48674800992012024,
+      "learning_rate": 1.8154126871909296e-05,
+      "loss": 0.5124222636222839,
+      "step": 929
+    },
+    {
+      "epoch": 0.23712391636919938,
+      "grad_norm": 0.5349897146224976,
+      "learning_rate": 1.81492418103265e-05,
+      "loss": 0.5278897881507874,
+      "step": 930
+    },
+    {
+      "epoch": 0.23737888832228454,
+      "grad_norm": 0.5227882862091064,
+      "learning_rate": 1.814435095226333e-05,
+      "loss": 0.5197198390960693,
+      "step": 931
+    },
+    {
+      "epoch": 0.2376338602753697,
+      "grad_norm": 0.5830060839653015,
+      "learning_rate": 1.81394543011986e-05,
+      "loss": 0.5197795629501343,
+      "step": 932
+    },
+    {
+      "epoch": 0.23788883222845486,
+      "grad_norm": 0.4716755747795105,
+      "learning_rate": 1.8134551860615256e-05,
+      "loss": 0.5050005316734314,
+      "step": 933
+    },
+    {
+      "epoch": 0.23814380418154002,
+      "grad_norm": 0.4908246397972107,
+      "learning_rate": 1.8129643634000357e-05,
+      "loss": 0.5052043795585632,
+      "step": 934
+    },
+    {
+      "epoch": 0.23839877613462518,
+      "grad_norm": 0.49863147735595703,
+      "learning_rate": 1.8124729624845077e-05,
+      "loss": 0.5203200578689575,
+      "step": 935
+    },
+    {
+      "epoch": 0.23865374808771037,
+      "grad_norm": 0.49130505323410034,
+      "learning_rate": 1.8119809836644715e-05,
+      "loss": 0.511245846748352,
+      "step": 936
+    },
+    {
+      "epoch": 0.23890872004079552,
+      "grad_norm": 0.4879443347454071,
+      "learning_rate": 1.8114884272898662e-05,
+      "loss": 0.5171563029289246,
+      "step": 937
+    },
+    {
+      "epoch": 0.23916369199388068,
+      "grad_norm": 0.4863815903663635,
+      "learning_rate": 1.8109952937110425e-05,
+      "loss": 0.5202454328536987,
+      "step": 938
+    },
+    {
+      "epoch": 0.23941866394696584,
+      "grad_norm": 0.5152843594551086,
+      "learning_rate": 1.8105015832787623e-05,
+      "loss": 0.5006053447723389,
+      "step": 939
+    },
+    {
+      "epoch": 0.239673635900051,
+      "grad_norm": 0.533755898475647,
+      "learning_rate": 1.810007296344197e-05,
+      "loss": 0.5041182637214661,
+      "step": 940
+    },
+    {
+      "epoch": 0.23992860785313616,
+      "grad_norm": 0.4880644679069519,
+      "learning_rate": 1.8095124332589284e-05,
+      "loss": 0.5128837823867798,
+      "step": 941
+    },
+    {
+      "epoch": 0.24018357980622132,
+      "grad_norm": 0.5024210810661316,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.5149589776992798,
+      "step": 942
+    },
+    {
+      "epoch": 0.24043855175930648,
+      "grad_norm": 0.4898494780063629,
+      "learning_rate": 1.8085209800446564e-05,
+      "loss": 0.519150972366333,
+      "step": 943
+    },
+    {
+      "epoch": 0.24069352371239164,
+      "grad_norm": 0.5067214369773865,
+      "learning_rate": 1.808024390620865e-05,
+      "loss": 0.4951241910457611,
+      "step": 944
+    },
+    {
+      "epoch": 0.2409484956654768,
+      "grad_norm": 0.5119669437408447,
+      "learning_rate": 1.8075272264567925e-05,
+      "loss": 0.5104391574859619,
+      "step": 945
+    },
+    {
+      "epoch": 0.24120346761856196,
+      "grad_norm": 0.47356361150741577,
+      "learning_rate": 1.8070294879060678e-05,
+      "loss": 0.5130054354667664,
+      "step": 946
+    },
+    {
+      "epoch": 0.24145843957164712,
+      "grad_norm": 0.48387038707733154,
+      "learning_rate": 1.8065311753227272e-05,
+      "loss": 0.5082079768180847,
+      "step": 947
+    },
+    {
+      "epoch": 0.24171341152473227,
+      "grad_norm": 0.4830719530582428,
+      "learning_rate": 1.806032289061216e-05,
+      "loss": 0.5125942826271057,
+      "step": 948
+    },
+    {
+      "epoch": 0.24196838347781743,
+      "grad_norm": 0.4798039495944977,
+      "learning_rate": 1.8055328294763874e-05,
+      "loss": 0.5218983888626099,
+      "step": 949
+    },
+    {
+      "epoch": 0.2422233554309026,
+      "grad_norm": 0.47178006172180176,
+      "learning_rate": 1.8050327969235024e-05,
+      "loss": 0.5177539587020874,
+      "step": 950
+    },
+    {
+      "epoch": 0.24247832738398775,
+      "grad_norm": 0.5267336964607239,
+      "learning_rate": 1.8045321917582293e-05,
+      "loss": 0.5238089561462402,
+      "step": 951
+    },
+    {
+      "epoch": 0.2427332993370729,
+      "grad_norm": 0.4602765738964081,
+      "learning_rate": 1.8040310143366447e-05,
+      "loss": 0.4993072748184204,
+      "step": 952
+    },
+    {
+      "epoch": 0.24298827129015807,
+      "grad_norm": 0.4600813686847687,
+      "learning_rate": 1.8035292650152303e-05,
+      "loss": 0.5135704874992371,
+      "step": 953
+    },
+    {
+      "epoch": 0.24324324324324326,
+      "grad_norm": 0.475060373544693,
+      "learning_rate": 1.803026944150877e-05,
+      "loss": 0.5125201940536499,
+      "step": 954
+    },
+    {
+      "epoch": 0.24349821519632842,
+      "grad_norm": 0.48474636673927307,
+      "learning_rate": 1.8025240521008798e-05,
+      "loss": 0.4982396364212036,
+      "step": 955
+    },
+    {
+      "epoch": 0.24375318714941357,
+      "grad_norm": 0.5275590419769287,
+      "learning_rate": 1.802020589222942e-05,
+      "loss": 0.5201438069343567,
+      "step": 956
+    },
+    {
+      "epoch": 0.24400815910249873,
+      "grad_norm": 0.48619845509529114,
+      "learning_rate": 1.801516555875172e-05,
+      "loss": 0.521652102470398,
+      "step": 957
+    },
+    {
+      "epoch": 0.2442631310555839,
+      "grad_norm": 0.4878775477409363,
+      "learning_rate": 1.8010119524160834e-05,
+      "loss": 0.5095280408859253,
+      "step": 958
+    },
+    {
+      "epoch": 0.24451810300866905,
+      "grad_norm": 0.49116265773773193,
+      "learning_rate": 1.8005067792045965e-05,
+      "loss": 0.4987580180168152,
+      "step": 959
+    },
+    {
+      "epoch": 0.2447730749617542,
+      "grad_norm": 0.4573895335197449,
+      "learning_rate": 1.8000010366000366e-05,
+      "loss": 0.5119349956512451,
+      "step": 960
+    },
+    {
+      "epoch": 0.24502804691483937,
+      "grad_norm": 0.5173366069793701,
+      "learning_rate": 1.7994947249621333e-05,
+      "loss": 0.5049844980239868,
+      "step": 961
+    },
+    {
+      "epoch": 0.24528301886792453,
+      "grad_norm": 0.48248374462127686,
+      "learning_rate": 1.7989878446510215e-05,
+      "loss": 0.505333423614502,
+      "step": 962
+    },
+    {
+      "epoch": 0.2455379908210097,
+      "grad_norm": 0.46847787499427795,
+      "learning_rate": 1.7984803960272405e-05,
+      "loss": 0.4996873736381531,
+      "step": 963
+    },
+    {
+      "epoch": 0.24579296277409485,
+      "grad_norm": 0.4857884645462036,
+      "learning_rate": 1.7979723794517338e-05,
+      "loss": 0.5094532370567322,
+      "step": 964
+    },
+    {
+      "epoch": 0.24604793472718,
+      "grad_norm": 0.47273755073547363,
+      "learning_rate": 1.7974637952858493e-05,
+      "loss": 0.5089906454086304,
+      "step": 965
+    },
+    {
+      "epoch": 0.24630290668026517,
+      "grad_norm": 0.48037955164909363,
+      "learning_rate": 1.796954643891337e-05,
+      "loss": 0.5090777277946472,
+      "step": 966
+    },
+    {
+      "epoch": 0.24655787863335032,
+      "grad_norm": 0.4806813597679138,
+      "learning_rate": 1.796444925630353e-05,
+      "loss": 0.5150277614593506,
+      "step": 967
+    },
+    {
+      "epoch": 0.24681285058643548,
+      "grad_norm": 0.47713831067085266,
+      "learning_rate": 1.7959346408654547e-05,
+      "loss": 0.5063602924346924,
+      "step": 968
+    },
+    {
+      "epoch": 0.24706782253952064,
+      "grad_norm": 0.463236927986145,
+      "learning_rate": 1.7954237899596027e-05,
+      "loss": 0.5151035189628601,
+      "step": 969
+    },
+    {
+      "epoch": 0.2473227944926058,
+      "grad_norm": 0.496243417263031,
+      "learning_rate": 1.7949123732761603e-05,
+      "loss": 0.5035381317138672,
+      "step": 970
+    },
+    {
+      "epoch": 0.24757776644569096,
+      "grad_norm": 0.5023394227027893,
+      "learning_rate": 1.7944003911788943e-05,
+      "loss": 0.5111215710639954,
+      "step": 971
+    },
+    {
+      "epoch": 0.24783273839877615,
+      "grad_norm": 0.48129037022590637,
+      "learning_rate": 1.7938878440319722e-05,
+      "loss": 0.5170261263847351,
+      "step": 972
+    },
+    {
+      "epoch": 0.2480877103518613,
+      "grad_norm": 0.493413507938385,
+      "learning_rate": 1.7933747321999642e-05,
+      "loss": 0.5111275911331177,
+      "step": 973
+    },
+    {
+      "epoch": 0.24834268230494647,
+      "grad_norm": 0.4867372214794159,
+      "learning_rate": 1.792861056047842e-05,
+      "loss": 0.5078480243682861,
+      "step": 974
+    },
+    {
+      "epoch": 0.24859765425803163,
+      "grad_norm": 0.509955883026123,
+      "learning_rate": 1.7923468159409786e-05,
+      "loss": 0.5027446150779724,
+      "step": 975
+    },
+    {
+      "epoch": 0.24885262621111678,
+      "grad_norm": 0.4628150463104248,
+      "learning_rate": 1.7918320122451486e-05,
+      "loss": 0.5124740600585938,
+      "step": 976
+    },
+    {
+      "epoch": 0.24910759816420194,
+      "grad_norm": 0.45736798644065857,
+      "learning_rate": 1.7913166453265264e-05,
+      "loss": 0.509383499622345,
+      "step": 977
+    },
+    {
+      "epoch": 0.2493625701172871,
+      "grad_norm": 0.5170514583587646,
+      "learning_rate": 1.7908007155516887e-05,
+      "loss": 0.5118707418441772,
+      "step": 978
+    },
+    {
+      "epoch": 0.24961754207037226,
+      "grad_norm": 0.4959959089756012,
+      "learning_rate": 1.7902842232876105e-05,
+      "loss": 0.5089007616043091,
+      "step": 979
+    },
+    {
+      "epoch": 0.24987251402345742,
+      "grad_norm": 0.4515939950942993,
+      "learning_rate": 1.789767168901669e-05,
+      "loss": 0.5101932883262634,
+      "step": 980
+    },
+    {
+      "epoch": 0.2501274859765426,
+      "grad_norm": 0.46624132990837097,
+      "learning_rate": 1.7892495527616397e-05,
+      "loss": 0.5061900615692139,
+      "step": 981
+    },
+    {
+      "epoch": 0.25038245792962777,
+      "grad_norm": 0.44490790367126465,
+      "learning_rate": 1.788731375235698e-05,
+      "loss": 0.5091356039047241,
+      "step": 982
+    },
+    {
+      "epoch": 0.2506374298827129,
+      "grad_norm": 0.4683645963668823,
+      "learning_rate": 1.788212636692419e-05,
+      "loss": 0.5200745463371277,
+      "step": 983
+    },
+    {
+      "epoch": 0.2508924018357981,
+      "grad_norm": 0.4631229639053345,
+      "learning_rate": 1.787693337500777e-05,
+      "loss": 0.5115489363670349,
+      "step": 984
+    },
+    {
+      "epoch": 0.25114737378888324,
+      "grad_norm": 0.47554415464401245,
+      "learning_rate": 1.787173478030144e-05,
+      "loss": 0.5071451663970947,
+      "step": 985
+    },
+    {
+      "epoch": 0.2514023457419684,
+      "grad_norm": 0.45346564054489136,
+      "learning_rate": 1.7866530586502914e-05,
+      "loss": 0.5114556550979614,
+      "step": 986
+    },
+    {
+      "epoch": 0.25165731769505356,
+      "grad_norm": 0.4637988209724426,
+      "learning_rate": 1.786132079731389e-05,
+      "loss": 0.5111753344535828,
+      "step": 987
+    },
+    {
+      "epoch": 0.2519122896481387,
+      "grad_norm": 0.4647560119628906,
+      "learning_rate": 1.7856105416440046e-05,
+      "loss": 0.5089646577835083,
+      "step": 988
+    },
+    {
+      "epoch": 0.2521672616012239,
+      "grad_norm": 0.47878068685531616,
+      "learning_rate": 1.7850884447591025e-05,
+      "loss": 0.5066558122634888,
+      "step": 989
+    },
+    {
+      "epoch": 0.25242223355430904,
+      "grad_norm": 0.4984032213687897,
+      "learning_rate": 1.784565789448046e-05,
+      "loss": 0.5191236734390259,
+      "step": 990
+    },
+    {
+      "epoch": 0.2526772055073942,
+      "grad_norm": 0.4744832515716553,
+      "learning_rate": 1.784042576082595e-05,
+      "loss": 0.48514270782470703,
+      "step": 991
+    },
+    {
+      "epoch": 0.25293217746047936,
+      "grad_norm": 0.4698421061038971,
+      "learning_rate": 1.7835188050349064e-05,
+      "loss": 0.5130078196525574,
+      "step": 992
+    },
+    {
+      "epoch": 0.2531871494135645,
+      "grad_norm": 0.4446653127670288,
+      "learning_rate": 1.7829944766775333e-05,
+      "loss": 0.5099542140960693,
+      "step": 993
+    },
+    {
+      "epoch": 0.2534421213666497,
+      "grad_norm": 0.48649677634239197,
+      "learning_rate": 1.7824695913834257e-05,
+      "loss": 0.5056322813034058,
+      "step": 994
+    },
+    {
+      "epoch": 0.25369709331973483,
+      "grad_norm": 0.45543864369392395,
+      "learning_rate": 1.7819441495259297e-05,
+      "loss": 0.5074928998947144,
+      "step": 995
+    },
+    {
+      "epoch": 0.25395206527282,
+      "grad_norm": 0.4768550992012024,
+      "learning_rate": 1.7814181514787873e-05,
+      "loss": 0.5242055058479309,
+      "step": 996
+    },
+    {
+      "epoch": 0.25420703722590515,
+      "grad_norm": 0.48187726736068726,
+      "learning_rate": 1.7808915976161364e-05,
+      "loss": 0.5093708634376526,
+      "step": 997
+    },
+    {
+      "epoch": 0.2544620091789903,
+      "grad_norm": 0.4623137414455414,
+      "learning_rate": 1.7803644883125093e-05,
+      "loss": 0.49990314245224,
+      "step": 998
+    },
+    {
+      "epoch": 0.25471698113207547,
+      "grad_norm": 0.437720388174057,
+      "learning_rate": 1.779836823942834e-05,
+      "loss": 0.50358647108078,
+      "step": 999
+    },
+    {
+      "epoch": 0.25497195308516063,
+      "grad_norm": 0.4898304045200348,
+      "learning_rate": 1.779308604882434e-05,
+      "loss": 0.5073832869529724,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2552269250382458,
+      "grad_norm": 0.4843921363353729,
+      "learning_rate": 1.7787798315070263e-05,
+      "loss": 0.494332492351532,
+      "step": 1001
+    },
+    {
+      "epoch": 0.25548189699133095,
+      "grad_norm": 0.48126739263534546,
+      "learning_rate": 1.7782505041927218e-05,
+      "loss": 0.5016703009605408,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2557368689444161,
+      "grad_norm": 0.4517021179199219,
+      "learning_rate": 1.777720623316027e-05,
+      "loss": 0.49795588850975037,
+      "step": 1003
+    },
+    {
+      "epoch": 0.25599184089750127,
+      "grad_norm": 0.49286332726478577,
+      "learning_rate": 1.777190189253841e-05,
+      "loss": 0.4952079951763153,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2562468128505864,
+      "grad_norm": 0.4811077117919922,
+      "learning_rate": 1.7766592023834567e-05,
+      "loss": 0.5128995776176453,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2565017848036716,
+      "grad_norm": 0.4894014000892639,
+      "learning_rate": 1.77612766308256e-05,
+      "loss": 0.5113130807876587,
+      "step": 1006
+    },
+    {
+      "epoch": 0.25675675675675674,
+      "grad_norm": 0.47017577290534973,
+      "learning_rate": 1.77559557172923e-05,
+      "loss": 0.5148601531982422,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2570117287098419,
+      "grad_norm": 0.4653587341308594,
+      "learning_rate": 1.775062928701938e-05,
+      "loss": 0.49841389060020447,
+      "step": 1008
+    },
+    {
+      "epoch": 0.25726670066292706,
+      "grad_norm": 0.5072566270828247,
+      "learning_rate": 1.7745297343795487e-05,
+      "loss": 0.5057612657546997,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2575216726160122,
+      "grad_norm": 0.48639681935310364,
+      "learning_rate": 1.7739959891413175e-05,
+      "loss": 0.5132037997245789,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2577766445690974,
+      "grad_norm": 0.4910087287425995,
+      "learning_rate": 1.7734616933668936e-05,
+      "loss": 0.5040243864059448,
+      "step": 1011
+    },
+    {
+      "epoch": 0.25803161652218254,
+      "grad_norm": 0.4876771569252014,
+      "learning_rate": 1.7729268474363156e-05,
+      "loss": 0.5224146842956543,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2582865884752677,
+      "grad_norm": 0.4741784930229187,
+      "learning_rate": 1.772391451730015e-05,
+      "loss": 0.49925094842910767,
+      "step": 1013
+    },
+    {
+      "epoch": 0.25854156042835286,
+      "grad_norm": 0.48314356803894043,
+      "learning_rate": 1.7718555066288138e-05,
+      "loss": 0.4966428279876709,
+      "step": 1014
+    },
+    {
+      "epoch": 0.258796532381438,
+      "grad_norm": 0.48094186186790466,
+      "learning_rate": 1.7713190125139243e-05,
+      "loss": 0.5235569477081299,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2590515043345232,
+      "grad_norm": 0.47423312067985535,
+      "learning_rate": 1.7707819697669504e-05,
+      "loss": 0.5283186435699463,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2593064762876084,
+      "grad_norm": 0.49974769353866577,
+      "learning_rate": 1.7702443787698855e-05,
+      "loss": 0.5045661926269531,
+      "step": 1017
+    },
+    {
+      "epoch": 0.25956144824069355,
+      "grad_norm": 0.4792061448097229,
+      "learning_rate": 1.7697062399051126e-05,
+      "loss": 0.5071189999580383,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2598164201937787,
+      "grad_norm": 0.47046658396720886,
+      "learning_rate": 1.7691675535554057e-05,
+      "loss": 0.5016986727714539,
+      "step": 1019
+    },
+    {
+      "epoch": 0.26007139214686387,
+      "grad_norm": 0.540090024471283,
+      "learning_rate": 1.768628320103927e-05,
+      "loss": 0.5116336345672607,
+      "step": 1020
+    },
+    {
+      "epoch": 0.260326364099949,
+      "grad_norm": 0.47801148891448975,
+      "learning_rate": 1.7680885399342283e-05,
+      "loss": 0.5074282884597778,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2605813360530342,
+      "grad_norm": 0.4674408435821533,
+      "learning_rate": 1.7675482134302503e-05,
+      "loss": 0.5008707046508789,
+      "step": 1022
+    },
+    {
+      "epoch": 0.26083630800611934,
+      "grad_norm": 0.4926578104496002,
+      "learning_rate": 1.7670073409763218e-05,
+      "loss": 0.5130398273468018,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2610912799592045,
+      "grad_norm": 0.49176183342933655,
+      "learning_rate": 1.7664659229571612e-05,
+      "loss": 0.509769082069397,
+      "step": 1024
+    },
+    {
+      "epoch": 0.26134625191228966,
+      "grad_norm": 0.49057114124298096,
+      "learning_rate": 1.7659239597578738e-05,
+      "loss": 0.49287480115890503,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2616012238653748,
+      "grad_norm": 0.46402209997177124,
+      "learning_rate": 1.7653814517639525e-05,
+      "loss": 0.516800045967102,
+      "step": 1026
+    },
+    {
+      "epoch": 0.26185619581846,
+      "grad_norm": 0.5012035965919495,
+      "learning_rate": 1.764838399361279e-05,
+      "loss": 0.5082567930221558,
+      "step": 1027
+    },
+    {
+      "epoch": 0.26211116777154514,
+      "grad_norm": 0.47564834356307983,
+      "learning_rate": 1.764294802936121e-05,
+      "loss": 0.5159832835197449,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2623661397246303,
+      "grad_norm": 0.4816213548183441,
+      "learning_rate": 1.7637506628751335e-05,
+      "loss": 0.5017706155776978,
+      "step": 1029
+    },
+    {
+      "epoch": 0.26262111167771546,
+      "grad_norm": 0.4737376868724823,
+      "learning_rate": 1.763205979565359e-05,
+      "loss": 0.501492977142334,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2628760836308006,
+      "grad_norm": 0.49498605728149414,
+      "learning_rate": 1.762660753394225e-05,
+      "loss": 0.5140003561973572,
+      "step": 1031
+    },
+    {
+      "epoch": 0.2631310555838858,
+      "grad_norm": 0.46646806597709656,
+      "learning_rate": 1.7621149847495458e-05,
+      "loss": 0.5102156400680542,
+      "step": 1032
+    },
+    {
+      "epoch": 0.26338602753697093,
+      "grad_norm": 0.4754180908203125,
+      "learning_rate": 1.761568674019522e-05,
+      "loss": 0.49969398975372314,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2636409994900561,
+      "grad_norm": 0.46612799167633057,
+      "learning_rate": 1.7610218215927393e-05,
+      "loss": 0.5041083097457886,
+      "step": 1034
+    },
+    {
+      "epoch": 0.26389597144314125,
+      "grad_norm": 0.46223193407058716,
+      "learning_rate": 1.760474427858169e-05,
+      "loss": 0.488689661026001,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2641509433962264,
+      "grad_norm": 0.4570687711238861,
+      "learning_rate": 1.7599264932051664e-05,
+      "loss": 0.4999356269836426,
+      "step": 1036
+    },
+    {
+      "epoch": 0.26440591534931157,
+      "grad_norm": 0.4681054949760437,
+      "learning_rate": 1.759378018023473e-05,
+      "loss": 0.504220187664032,
+      "step": 1037
+    },
+    {
+      "epoch": 0.26466088730239673,
+      "grad_norm": 0.48575031757354736,
+      "learning_rate": 1.7588290027032137e-05,
+      "loss": 0.5085967779159546,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2649158592554819,
+      "grad_norm": 0.4551243185997009,
+      "learning_rate": 1.758279447634899e-05,
+      "loss": 0.5101237297058105,
+      "step": 1039
+    },
+    {
+      "epoch": 0.26517083120856705,
+      "grad_norm": 0.455987811088562,
+      "learning_rate": 1.757729353209421e-05,
+      "loss": 0.5142706632614136,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2654258031616522,
+      "grad_norm": 0.47160667181015015,
+      "learning_rate": 1.7571787198180578e-05,
+      "loss": 0.5077152252197266,
+      "step": 1041
+    },
+    {
+      "epoch": 0.26568077511473737,
+      "grad_norm": 0.45185431838035583,
+      "learning_rate": 1.7566275478524694e-05,
+      "loss": 0.5094373226165771,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2659357470678225,
+      "grad_norm": 0.4466405212879181,
+      "learning_rate": 1.7560758377046994e-05,
+      "loss": 0.4995154142379761,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2661907190209077,
+      "grad_norm": 0.4542228877544403,
+      "learning_rate": 1.755523589767174e-05,
+      "loss": 0.5100736618041992,
+      "step": 1044
+    },
+    {
+      "epoch": 0.26644569097399284,
+      "grad_norm": 0.4614347517490387,
+      "learning_rate": 1.7549708044327024e-05,
+      "loss": 0.5141627788543701,
+      "step": 1045
+    },
+    {
+      "epoch": 0.266700662927078,
+      "grad_norm": 0.44910943508148193,
+      "learning_rate": 1.7544174820944752e-05,
+      "loss": 0.5136409997940063,
+      "step": 1046
+    },
+    {
+      "epoch": 0.26695563488016316,
+      "grad_norm": 0.46518927812576294,
+      "learning_rate": 1.753863623146066e-05,
+      "loss": 0.4939352869987488,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2672106068332483,
+      "grad_norm": 0.4545368552207947,
+      "learning_rate": 1.753309227981429e-05,
+      "loss": 0.5089648962020874,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2674655787863335,
+      "grad_norm": 0.4595126509666443,
+      "learning_rate": 1.7527542969949008e-05,
+      "loss": 0.5027369260787964,
+      "step": 1049
+    },
+    {
+      "epoch": 0.26772055073941864,
+      "grad_norm": 0.45682191848754883,
+      "learning_rate": 1.7521988305811986e-05,
+      "loss": 0.5061858892440796,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2679755226925038,
+      "grad_norm": 0.5684783458709717,
+      "learning_rate": 1.7516428291354205e-05,
+      "loss": 0.4948621690273285,
+      "step": 1051
+    },
+    {
+      "epoch": 0.268230494645589,
+      "grad_norm": 0.4799656271934509,
+      "learning_rate": 1.751086293053045e-05,
+      "loss": 0.5029177665710449,
+      "step": 1052
+    },
+    {
+      "epoch": 0.26848546659867417,
+      "grad_norm": 0.48182642459869385,
+      "learning_rate": 1.7505292227299314e-05,
+      "loss": 0.5024133920669556,
+      "step": 1053
+    },
+    {
+      "epoch": 0.26874043855175933,
+      "grad_norm": 0.4441515803337097,
+      "learning_rate": 1.7499716185623183e-05,
+      "loss": 0.5072953104972839,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2689954105048445,
+      "grad_norm": 0.46608737111091614,
+      "learning_rate": 1.7494134809468248e-05,
+      "loss": 0.5079753398895264,
+      "step": 1055
+    },
+    {
+      "epoch": 0.26925038245792965,
+      "grad_norm": 0.47361519932746887,
+      "learning_rate": 1.7488548102804482e-05,
+      "loss": 0.5090345144271851,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2695053544110148,
+      "grad_norm": 0.5031419992446899,
+      "learning_rate": 1.748295606960567e-05,
+      "loss": 0.4993833899497986,
+      "step": 1057
+    },
+    {
+      "epoch": 0.26976032636409997,
+      "grad_norm": 0.46434053778648376,
+      "learning_rate": 1.7477358713849368e-05,
+      "loss": 0.5092716813087463,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2700152983171851,
+      "grad_norm": 0.4608798027038574,
+      "learning_rate": 1.7471756039516917e-05,
+      "loss": 0.5093937516212463,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2702702702702703,
+      "grad_norm": 0.47563883662223816,
+      "learning_rate": 1.7466148050593458e-05,
+      "loss": 0.5046992897987366,
+      "step": 1060
+    },
+    {
+      "epoch": 0.27052524222335544,
+      "grad_norm": 0.5088484287261963,
+      "learning_rate": 1.74605347510679e-05,
+      "loss": 0.509278416633606,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2707802141764406,
+      "grad_norm": 0.4759383797645569,
+      "learning_rate": 1.7454916144932923e-05,
+      "loss": 0.500342845916748,
+      "step": 1062
+    },
+    {
+      "epoch": 0.27103518612952576,
+      "grad_norm": 0.4903341233730316,
+      "learning_rate": 1.7449292236184994e-05,
+      "loss": 0.5136889219284058,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2712901580826109,
+      "grad_norm": 0.4696923494338989,
+      "learning_rate": 1.744366302882435e-05,
+      "loss": 0.5082424283027649,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2715451300356961,
+      "grad_norm": 0.46577882766723633,
+      "learning_rate": 1.7438028526854993e-05,
+      "loss": 0.5139687657356262,
+      "step": 1065
+    },
+    {
+      "epoch": 0.27180010198878124,
+      "grad_norm": 0.49244529008865356,
+      "learning_rate": 1.743238873428469e-05,
+      "loss": 0.5190452933311462,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2720550739418664,
+      "grad_norm": 0.4551467001438141,
+      "learning_rate": 1.7426743655124973e-05,
+      "loss": 0.5102889537811279,
+      "step": 1067
+    },
+    {
+      "epoch": 0.27231004589495156,
+      "grad_norm": 0.46808892488479614,
+      "learning_rate": 1.742109329339114e-05,
+      "loss": 0.49744510650634766,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2725650178480367,
+      "grad_norm": 0.4782489836215973,
+      "learning_rate": 1.7415437653102237e-05,
+      "loss": 0.502158522605896,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2728199898011219,
+      "grad_norm": 0.5074446201324463,
+      "learning_rate": 1.7409776738281067e-05,
+      "loss": 0.5154489278793335,
+      "step": 1070
+    },
+    {
+      "epoch": 0.27307496175420704,
+      "grad_norm": 0.4884415566921234,
+      "learning_rate": 1.740411055295419e-05,
+      "loss": 0.5210711359977722,
+      "step": 1071
+    },
+    {
+      "epoch": 0.2733299337072922,
+      "grad_norm": 0.4918287694454193,
+      "learning_rate": 1.7398439101151908e-05,
+      "loss": 0.49791979789733887,
+      "step": 1072
+    },
+    {
+      "epoch": 0.27358490566037735,
+      "grad_norm": 0.5024614334106445,
+      "learning_rate": 1.7392762386908275e-05,
+      "loss": 0.5004453063011169,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2738398776134625,
+      "grad_norm": 0.4789541959762573,
+      "learning_rate": 1.7387080414261084e-05,
+      "loss": 0.49946314096450806,
+      "step": 1074
+    },
+    {
+      "epoch": 0.27409484956654767,
+      "grad_norm": 0.5202962160110474,
+      "learning_rate": 1.738139318725187e-05,
+      "loss": 0.5039060711860657,
+      "step": 1075
+    },
+    {
+      "epoch": 0.27434982151963283,
+      "grad_norm": 0.4789374768733978,
+      "learning_rate": 1.7375700709925903e-05,
+      "loss": 0.5128728151321411,
+      "step": 1076
+    },
+    {
+      "epoch": 0.274604793472718,
+      "grad_norm": 0.46692997217178345,
+      "learning_rate": 1.7370002986332195e-05,
+      "loss": 0.5212869644165039,
+      "step": 1077
+    },
+    {
+      "epoch": 0.27485976542580315,
+      "grad_norm": 0.45039069652557373,
+      "learning_rate": 1.7364300020523475e-05,
+      "loss": 0.5023295879364014,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2751147373788883,
+      "grad_norm": 0.44396907091140747,
+      "learning_rate": 1.7358591816556223e-05,
+      "loss": 0.5003005266189575,
+      "step": 1079
+    },
+    {
+      "epoch": 0.27536970933197347,
+      "grad_norm": 0.5632413625717163,
+      "learning_rate": 1.7352878378490622e-05,
+      "loss": 0.4984963834285736,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2756246812850586,
+      "grad_norm": 0.4821857810020447,
+      "learning_rate": 1.7347159710390585e-05,
+      "loss": 0.5095998048782349,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2758796532381438,
+      "grad_norm": 0.509071409702301,
+      "learning_rate": 1.7341435816323756e-05,
+      "loss": 0.5028913617134094,
+      "step": 1082
+    },
+    {
+      "epoch": 0.27613462519122894,
+      "grad_norm": 0.4965466558933258,
+      "learning_rate": 1.7335706700361488e-05,
+      "loss": 0.5158098936080933,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2763895971443141,
+      "grad_norm": 0.4649892747402191,
+      "learning_rate": 1.732997236657884e-05,
+      "loss": 0.5070985555648804,
+      "step": 1084
+    },
+    {
+      "epoch": 0.27664456909739926,
+      "grad_norm": 0.46375176310539246,
+      "learning_rate": 1.73242328190546e-05,
+      "loss": 0.5019338130950928,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2768995410504844,
+      "grad_norm": 0.4885120689868927,
+      "learning_rate": 1.7318488061871245e-05,
+      "loss": 0.5054546594619751,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2771545130035696,
+      "grad_norm": 0.47661104798316956,
+      "learning_rate": 1.7312738099114973e-05,
+      "loss": 0.49401187896728516,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2774094849566548,
+      "grad_norm": 0.4707930088043213,
+      "learning_rate": 1.7306982934875684e-05,
+      "loss": 0.5095722079277039,
+      "step": 1088
+    },
+    {
+      "epoch": 0.27766445690973995,
+      "grad_norm": 0.45717406272888184,
+      "learning_rate": 1.730122257324697e-05,
+      "loss": 0.5120978951454163,
+      "step": 1089
+    },
+    {
+      "epoch": 0.2779194288628251,
+      "grad_norm": 0.46347159147262573,
+      "learning_rate": 1.7295457018326117e-05,
+      "loss": 0.4990503191947937,
+      "step": 1090
+    },
+    {
+      "epoch": 0.27817440081591027,
+      "grad_norm": 0.4729735255241394,
+      "learning_rate": 1.7289686274214116e-05,
+      "loss": 0.5116077661514282,
+      "step": 1091
+    },
+    {
+      "epoch": 0.27842937276899543,
+      "grad_norm": 0.4576144516468048,
+      "learning_rate": 1.7283910345015645e-05,
+      "loss": 0.5087347030639648,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2786843447220806,
+      "grad_norm": 0.4772886335849762,
+      "learning_rate": 1.727812923483907e-05,
+      "loss": 0.5092043280601501,
+      "step": 1093
+    },
+    {
+      "epoch": 0.27893931667516575,
+      "grad_norm": 0.4713665246963501,
+      "learning_rate": 1.7272342947796433e-05,
+      "loss": 0.5062289834022522,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2791942886282509,
+      "grad_norm": 0.4943859875202179,
+      "learning_rate": 1.7266551488003478e-05,
+      "loss": 0.5019166469573975,
+      "step": 1095
+    },
+    {
+      "epoch": 0.27944926058133607,
+      "grad_norm": 0.49818283319473267,
+      "learning_rate": 1.726075485957961e-05,
+      "loss": 0.5120074152946472,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2797042325344212,
+      "grad_norm": 0.4867881238460541,
+      "learning_rate": 1.7254953066647915e-05,
+      "loss": 0.4957153797149658,
+      "step": 1097
+    },
+    {
+      "epoch": 0.2799592044875064,
+      "grad_norm": 0.49564328789711,
+      "learning_rate": 1.724914611333516e-05,
+      "loss": 0.4979622960090637,
+      "step": 1098
+    },
+    {
+      "epoch": 0.28021417644059154,
+      "grad_norm": 0.47495225071907043,
+      "learning_rate": 1.7243334003771778e-05,
+      "loss": 0.4982151985168457,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2804691483936767,
+      "grad_norm": 0.45300769805908203,
+      "learning_rate": 1.7237516742091863e-05,
+      "loss": 0.5014975070953369,
+      "step": 1100
+    },
+    {
+      "epoch": 0.28072412034676186,
+      "grad_norm": 0.44557687640190125,
+      "learning_rate": 1.7231694332433184e-05,
+      "loss": 0.4973270297050476,
+      "step": 1101
+    },
+    {
+      "epoch": 0.280979092299847,
+      "grad_norm": 0.4794977009296417,
+      "learning_rate": 1.7225866778937168e-05,
+      "loss": 0.5073505640029907,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2812340642529322,
+      "grad_norm": 0.4475419819355011,
+      "learning_rate": 1.722003408574889e-05,
+      "loss": 0.5049219131469727,
+      "step": 1103
+    },
+    {
+      "epoch": 0.28148903620601734,
+      "grad_norm": 0.4797416925430298,
+      "learning_rate": 1.7214196257017106e-05,
+      "loss": 0.5003278851509094,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2817440081591025,
+      "grad_norm": 0.47968360781669617,
+      "learning_rate": 1.72083532968942e-05,
+      "loss": 0.5042279958724976,
+      "step": 1105
+    },
+    {
+      "epoch": 0.28199898011218766,
+      "grad_norm": 0.4770395755767822,
+      "learning_rate": 1.720250520953622e-05,
+      "loss": 0.5177443623542786,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2822539520652728,
+      "grad_norm": 0.4603266417980194,
+      "learning_rate": 1.719665199910285e-05,
+      "loss": 0.5057562589645386,
+      "step": 1107
+    },
+    {
+      "epoch": 0.282508924018358,
+      "grad_norm": 0.46381697058677673,
+      "learning_rate": 1.7190793669757427e-05,
+      "loss": 0.5043715238571167,
+      "step": 1108
+    },
+    {
+      "epoch": 0.28276389597144314,
+      "grad_norm": 0.4711867868900299,
+      "learning_rate": 1.7184930225666932e-05,
+      "loss": 0.5236672163009644,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2830188679245283,
+      "grad_norm": 0.4568409025669098,
+      "learning_rate": 1.7179061671001974e-05,
+      "loss": 0.5128495097160339,
+      "step": 1110
+    },
+    {
+      "epoch": 0.28327383987761345,
+      "grad_norm": 0.45788177847862244,
+      "learning_rate": 1.71731880099368e-05,
+      "loss": 0.5006764531135559,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2835288118306986,
+      "grad_norm": 0.44360288977622986,
+      "learning_rate": 1.7167309246649298e-05,
+      "loss": 0.49972233176231384,
+      "step": 1112
+    },
+    {
+      "epoch": 0.28378378378378377,
+      "grad_norm": 0.4582159221172333,
+      "learning_rate": 1.716142538532097e-05,
+      "loss": 0.5080081224441528,
+      "step": 1113
+    },
+    {
+      "epoch": 0.28403875573686893,
+      "grad_norm": 0.45554617047309875,
+      "learning_rate": 1.7155536430136958e-05,
+      "loss": 0.5192968845367432,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2842937276899541,
+      "grad_norm": 0.46286189556121826,
+      "learning_rate": 1.7149642385286015e-05,
+      "loss": 0.5007770657539368,
+      "step": 1115
+    },
+    {
+      "epoch": 0.28454869964303925,
+      "grad_norm": 0.45503437519073486,
+      "learning_rate": 1.7143743254960526e-05,
+      "loss": 0.49312448501586914,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2848036715961244,
+      "grad_norm": 0.4485381841659546,
+      "learning_rate": 1.7137839043356486e-05,
+      "loss": 0.5041566491127014,
+      "step": 1117
+    },
+    {
+      "epoch": 0.28505864354920957,
+      "grad_norm": 0.5287951231002808,
+      "learning_rate": 1.7131929754673505e-05,
+      "loss": 0.4854404926300049,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2853136155022947,
+      "grad_norm": 0.48092010617256165,
+      "learning_rate": 1.712601539311481e-05,
+      "loss": 0.5001811981201172,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2855685874553799,
+      "grad_norm": 0.4755527377128601,
+      "learning_rate": 1.7120095962887226e-05,
+      "loss": 0.5057350397109985,
+      "step": 1120
+    },
+    {
+      "epoch": 0.28582355940846504,
+      "grad_norm": 0.5042078495025635,
+      "learning_rate": 1.7114171468201192e-05,
+      "loss": 0.5058793425559998,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2860785313615502,
+      "grad_norm": 0.46610692143440247,
+      "learning_rate": 1.710824191327075e-05,
+      "loss": 0.5130771398544312,
+      "step": 1122
+    },
+    {
+      "epoch": 0.28633350331463536,
+      "grad_norm": 0.4715431332588196,
+      "learning_rate": 1.710230730231353e-05,
+      "loss": 0.5075538158416748,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2865884752677206,
+      "grad_norm": 0.4927000403404236,
+      "learning_rate": 1.7096367639550767e-05,
+      "loss": 0.5113017559051514,
+      "step": 1124
+    },
+    {
+      "epoch": 0.28684344722080574,
+      "grad_norm": 0.4718981683254242,
+      "learning_rate": 1.7090422929207295e-05,
+      "loss": 0.49854955077171326,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2870984191738909,
+      "grad_norm": 0.49637195467948914,
+      "learning_rate": 1.7084473175511527e-05,
+      "loss": 0.4942783713340759,
+      "step": 1126
+    },
+    {
+      "epoch": 0.28735339112697605,
+      "grad_norm": 0.48915430903434753,
+      "learning_rate": 1.7078518382695467e-05,
+      "loss": 0.49409493803977966,
+      "step": 1127
+    },
+    {
+      "epoch": 0.2876083630800612,
+      "grad_norm": 0.5150572061538696,
+      "learning_rate": 1.7072558554994705e-05,
+      "loss": 0.5223121047019958,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2878633350331464,
+      "grad_norm": 0.4622846841812134,
+      "learning_rate": 1.7066593696648408e-05,
+      "loss": 0.5007497668266296,
+      "step": 1129
+    },
+    {
+      "epoch": 0.28811830698623153,
+      "grad_norm": 0.48057541251182556,
+      "learning_rate": 1.7060623811899332e-05,
+      "loss": 0.49329742789268494,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2883732789393167,
+      "grad_norm": 0.47378435730934143,
+      "learning_rate": 1.7054648904993795e-05,
+      "loss": 0.5099858045578003,
+      "step": 1131
+    },
+    {
+      "epoch": 0.28862825089240185,
+      "grad_norm": 0.4834018647670746,
+      "learning_rate": 1.7048668980181696e-05,
+      "loss": 0.5121385455131531,
+      "step": 1132
+    },
+    {
+      "epoch": 0.288883222845487,
+      "grad_norm": 0.47347232699394226,
+      "learning_rate": 1.70426840417165e-05,
+      "loss": 0.5076555609703064,
+      "step": 1133
+    },
+    {
+      "epoch": 0.28913819479857217,
+      "grad_norm": 0.48978927731513977,
+      "learning_rate": 1.7036694093855238e-05,
+      "loss": 0.5149083137512207,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2893931667516573,
+      "grad_norm": 0.4732295572757721,
+      "learning_rate": 1.7030699140858503e-05,
+      "loss": 0.49698835611343384,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2896481387047425,
+      "grad_norm": 0.4847556948661804,
+      "learning_rate": 1.702469918699045e-05,
+      "loss": 0.5097410678863525,
+      "step": 1136
+    },
+    {
+      "epoch": 0.28990311065782765,
+      "grad_norm": 0.45591381192207336,
+      "learning_rate": 1.7018694236518792e-05,
+      "loss": 0.5003314018249512,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2901580826109128,
+      "grad_norm": 0.49350666999816895,
+      "learning_rate": 1.7012684293714797e-05,
+      "loss": 0.5001873970031738,
+      "step": 1138
+    },
+    {
+      "epoch": 0.29041305456399796,
+      "grad_norm": 0.467124879360199,
+      "learning_rate": 1.7006669362853275e-05,
+      "loss": 0.48504096269607544,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2906680265170831,
+      "grad_norm": 0.45942822098731995,
+      "learning_rate": 1.7000649448212595e-05,
+      "loss": 0.49284881353378296,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2909229984701683,
+      "grad_norm": 0.4662016034126282,
+      "learning_rate": 1.6994624554074667e-05,
+      "loss": 0.49966710805892944,
+      "step": 1141
+    },
+    {
+      "epoch": 0.29117797042325344,
+      "grad_norm": 0.46195200085639954,
+      "learning_rate": 1.6988594684724947e-05,
+      "loss": 0.49967098236083984,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2914329423763386,
+      "grad_norm": 0.4631154537200928,
+      "learning_rate": 1.6982559844452414e-05,
+      "loss": 0.4934503436088562,
+      "step": 1143
+    },
+    {
+      "epoch": 0.29168791432942376,
+      "grad_norm": 0.46881651878356934,
+      "learning_rate": 1.69765200375496e-05,
+      "loss": 0.4958922863006592,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2919428862825089,
+      "grad_norm": 0.47307780385017395,
+      "learning_rate": 1.6970475268312568e-05,
+      "loss": 0.49355167150497437,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2921978582355941,
+      "grad_norm": 0.48002174496650696,
+      "learning_rate": 1.6964425541040904e-05,
+      "loss": 0.4948916435241699,
+      "step": 1146
+    },
+    {
+      "epoch": 0.29245283018867924,
+      "grad_norm": 0.45839452743530273,
+      "learning_rate": 1.695837086003772e-05,
+      "loss": 0.5025448203086853,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2927078021417644,
+      "grad_norm": 0.4786364436149597,
+      "learning_rate": 1.6952311229609656e-05,
+      "loss": 0.49452894926071167,
+      "step": 1148
+    },
+    {
+      "epoch": 0.29296277409484955,
+      "grad_norm": 0.5036335587501526,
+      "learning_rate": 1.694624665406687e-05,
+      "loss": 0.4942960739135742,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2932177460479347,
+      "grad_norm": 0.47374340891838074,
+      "learning_rate": 1.6940177137723046e-05,
+      "loss": 0.48950761556625366,
+      "step": 1150
+    },
+    {
+      "epoch": 0.29347271800101987,
+      "grad_norm": 0.4503117799758911,
+      "learning_rate": 1.693410268489536e-05,
+      "loss": 0.4947274327278137,
+      "step": 1151
+    },
+    {
+      "epoch": 0.29372768995410503,
+      "grad_norm": 0.49066653847694397,
+      "learning_rate": 1.692802329990453e-05,
+      "loss": 0.5052638053894043,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2939826619071902,
+      "grad_norm": 0.5150389671325684,
+      "learning_rate": 1.6921938987074756e-05,
+      "loss": 0.5081547498703003,
+      "step": 1153
+    },
+    {
+      "epoch": 0.29423763386027535,
+      "grad_norm": 0.4666706621646881,
+      "learning_rate": 1.691584975073376e-05,
+      "loss": 0.5027748346328735,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2944926058133605,
+      "grad_norm": 0.5015774369239807,
+      "learning_rate": 1.6909755595212755e-05,
+      "loss": 0.5036760568618774,
+      "step": 1155
+    },
+    {
+      "epoch": 0.29474757776644567,
+      "grad_norm": 0.472493439912796,
+      "learning_rate": 1.6903656524846458e-05,
+      "loss": 0.5006581544876099,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2950025497195308,
+      "grad_norm": 0.5009905099868774,
+      "learning_rate": 1.6897552543973084e-05,
+      "loss": 0.5129563212394714,
+      "step": 1157
+    },
+    {
+      "epoch": 0.295257521672616,
+      "grad_norm": 0.46842071413993835,
+      "learning_rate": 1.6891443656934336e-05,
+      "loss": 0.4934506416320801,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2955124936257012,
+      "grad_norm": 0.46922436356544495,
+      "learning_rate": 1.688532986807541e-05,
+      "loss": 0.5039920806884766,
+      "step": 1159
+    },
+    {
+      "epoch": 0.29576746557878636,
+      "grad_norm": 0.4793666899204254,
+      "learning_rate": 1.6879211181744986e-05,
+      "loss": 0.4972963333129883,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2960224375318715,
+      "grad_norm": 0.47085341811180115,
+      "learning_rate": 1.6873087602295232e-05,
+      "loss": 0.5129951238632202,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2962774094849567,
+      "grad_norm": 0.45663273334503174,
+      "learning_rate": 1.686695913408179e-05,
+      "loss": 0.4943583607673645,
+      "step": 1162
+    },
+    {
+      "epoch": 0.29653238143804184,
+      "grad_norm": 0.5101933479309082,
+      "learning_rate": 1.6860825781463783e-05,
+      "loss": 0.49064767360687256,
+      "step": 1163
+    },
+    {
+      "epoch": 0.296787353391127,
+      "grad_norm": 0.5012008547782898,
+      "learning_rate": 1.685468754880381e-05,
+      "loss": 0.49900519847869873,
+      "step": 1164
+    },
+    {
+      "epoch": 0.29704232534421215,
+      "grad_norm": 0.4591823220252991,
+      "learning_rate": 1.684854444046794e-05,
+      "loss": 0.49930524826049805,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2972972972972973,
+      "grad_norm": 0.48761099576950073,
+      "learning_rate": 1.6842396460825706e-05,
+      "loss": 0.4923129975795746,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2975522692503825,
+      "grad_norm": 0.45887067914009094,
+      "learning_rate": 1.6836243614250113e-05,
+      "loss": 0.5167244672775269,
+      "step": 1167
+    },
+    {
+      "epoch": 0.29780724120346763,
+      "grad_norm": 0.48813778162002563,
+      "learning_rate": 1.683008590511762e-05,
+      "loss": 0.4933728575706482,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2980622131565528,
+      "grad_norm": 0.4750381410121918,
+      "learning_rate": 1.682392333780815e-05,
+      "loss": 0.4995809495449066,
+      "step": 1169
+    },
+    {
+      "epoch": 0.29831718510963795,
+      "grad_norm": 0.47015345096588135,
+      "learning_rate": 1.6817755916705084e-05,
+      "loss": 0.49063003063201904,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2985721570627231,
+      "grad_norm": 0.47086119651794434,
+      "learning_rate": 1.6811583646195242e-05,
+      "loss": 0.5013208985328674,
+      "step": 1171
+    },
+    {
+      "epoch": 0.29882712901580827,
+      "grad_norm": 0.48746731877326965,
+      "learning_rate": 1.680540653066891e-05,
+      "loss": 0.5028801560401917,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2990821009688934,
+      "grad_norm": 0.4941359758377075,
+      "learning_rate": 1.6799224574519818e-05,
+      "loss": 0.5044407844543457,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2993370729219786,
+      "grad_norm": 0.47060495615005493,
+      "learning_rate": 1.6793037782145122e-05,
+      "loss": 0.5153166055679321,
+      "step": 1174
+    },
+    {
+      "epoch": 0.29959204487506375,
+      "grad_norm": 0.4782000482082367,
+      "learning_rate": 1.6786846157945437e-05,
+      "loss": 0.49174898862838745,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2998470168281489,
+      "grad_norm": 0.47206997871398926,
+      "learning_rate": 1.678064970632481e-05,
+      "loss": 0.49420687556266785,
+      "step": 1176
+    },
+    {
+      "epoch": 0.30010198878123406,
+      "grad_norm": 0.5003753304481506,
+      "learning_rate": 1.677444843169072e-05,
+      "loss": 0.5016045570373535,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3003569607343192,
+      "grad_norm": 0.4960240423679352,
+      "learning_rate": 1.676824233845407e-05,
+      "loss": 0.4992179572582245,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3006119326874044,
+      "grad_norm": 0.4680156409740448,
+      "learning_rate": 1.6762031431029203e-05,
+      "loss": 0.5058457255363464,
+      "step": 1179
+    },
+    {
+      "epoch": 0.30086690464048954,
+      "grad_norm": 0.5267618894577026,
+      "learning_rate": 1.6755815713833876e-05,
+      "loss": 0.4940145015716553,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3011218765935747,
+      "grad_norm": 0.4637565612792969,
+      "learning_rate": 1.674959519128927e-05,
+      "loss": 0.5075904130935669,
+      "step": 1181
+    },
+    {
+      "epoch": 0.30137684854665986,
+      "grad_norm": 0.6613916754722595,
+      "learning_rate": 1.674336986781999e-05,
+      "loss": 0.5021583437919617,
+      "step": 1182
+    },
+    {
+      "epoch": 0.301631820499745,
+      "grad_norm": 0.5022047758102417,
+      "learning_rate": 1.6737139747854055e-05,
+      "loss": 0.49105164408683777,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3018867924528302,
+      "grad_norm": 0.5665522813796997,
+      "learning_rate": 1.673090483582288e-05,
+      "loss": 0.49535173177719116,
+      "step": 1184
+    },
+    {
+      "epoch": 0.30214176440591534,
+      "grad_norm": 0.4763643145561218,
+      "learning_rate": 1.6724665136161305e-05,
+      "loss": 0.496861070394516,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3023967363590005,
+      "grad_norm": 0.49327611923217773,
+      "learning_rate": 1.6718420653307578e-05,
+      "loss": 0.5091848373413086,
+      "step": 1186
+    },
+    {
+      "epoch": 0.30265170831208565,
+      "grad_norm": 0.44212591648101807,
+      "learning_rate": 1.6712171391703332e-05,
+      "loss": 0.49904412031173706,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3029066802651708,
+      "grad_norm": 0.48958224058151245,
+      "learning_rate": 1.670591735579361e-05,
+      "loss": 0.4978852868080139,
+      "step": 1188
+    },
+    {
+      "epoch": 0.303161652218256,
+      "grad_norm": 0.4796984791755676,
+      "learning_rate": 1.6699658550026855e-05,
+      "loss": 0.49702033400535583,
+      "step": 1189
+    },
+    {
+      "epoch": 0.30341662417134113,
+      "grad_norm": 0.6042547225952148,
+      "learning_rate": 1.6693394978854898e-05,
+      "loss": 0.5002682209014893,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3036715961244263,
+      "grad_norm": 0.4669792652130127,
+      "learning_rate": 1.6687126646732955e-05,
+      "loss": 0.4984990060329437,
+      "step": 1191
+    },
+    {
+      "epoch": 0.30392656807751145,
+      "grad_norm": 0.451721727848053,
+      "learning_rate": 1.668085355811963e-05,
+      "loss": 0.501367449760437,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3041815400305966,
+      "grad_norm": 0.44385984539985657,
+      "learning_rate": 1.6674575717476923e-05,
+      "loss": 0.4959375262260437,
+      "step": 1193
+    },
+    {
+      "epoch": 0.30443651198368177,
+      "grad_norm": 0.47235265374183655,
+      "learning_rate": 1.66682931292702e-05,
+      "loss": 0.49185383319854736,
+      "step": 1194
+    },
+    {
+      "epoch": 0.304691483936767,
+      "grad_norm": 0.46303853392601013,
+      "learning_rate": 1.66620057979682e-05,
+      "loss": 0.5002027153968811,
+      "step": 1195
+    },
+    {
+      "epoch": 0.30494645588985214,
+      "grad_norm": 0.4476638734340668,
+      "learning_rate": 1.6655713728043056e-05,
+      "loss": 0.49557623267173767,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3052014278429373,
+      "grad_norm": 0.47302544116973877,
+      "learning_rate": 1.6649416923970248e-05,
+      "loss": 0.4934016466140747,
+      "step": 1197
+    },
+    {
+      "epoch": 0.30545639979602246,
+      "grad_norm": 0.4599985182285309,
+      "learning_rate": 1.6643115390228642e-05,
+      "loss": 0.5042760372161865,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3057113717491076,
+      "grad_norm": 0.47824129462242126,
+      "learning_rate": 1.663680913130046e-05,
+      "loss": 0.4973282516002655,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3059663437021928,
+      "grad_norm": 0.4603481590747833,
+      "learning_rate": 1.6630498151671282e-05,
+      "loss": 0.485217422246933,
+      "step": 1200
+    },
+    {
+      "epoch": 0.30622131565527794,
+      "grad_norm": 0.48893457651138306,
+      "learning_rate": 1.662418245583005e-05,
+      "loss": 0.4935639798641205,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3064762876083631,
+      "grad_norm": 0.4877832233905792,
+      "learning_rate": 1.6617862048269065e-05,
+      "loss": 0.5135349631309509,
+      "step": 1202
+    },
+    {
+      "epoch": 0.30673125956144826,
+      "grad_norm": 0.45736855268478394,
+      "learning_rate": 1.6611536933483966e-05,
+      "loss": 0.48187631368637085,
+      "step": 1203
+    },
+    {
+      "epoch": 0.3069862315145334,
+      "grad_norm": 0.48175039887428284,
+      "learning_rate": 1.660520711597376e-05,
+      "loss": 0.5004374384880066,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3072412034676186,
+      "grad_norm": 0.45422685146331787,
+      "learning_rate": 1.659887260024078e-05,
+      "loss": 0.4838033616542816,
+      "step": 1205
+    },
+    {
+      "epoch": 0.30749617542070373,
+      "grad_norm": 0.49426570534706116,
+      "learning_rate": 1.659253339079071e-05,
+      "loss": 0.5089007616043091,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3077511473737889,
+      "grad_norm": 0.48816773295402527,
+      "learning_rate": 1.6586189492132567e-05,
+      "loss": 0.4888179898262024,
+      "step": 1207
+    },
+    {
+      "epoch": 0.30800611932687405,
+      "grad_norm": 0.4737013876438141,
+      "learning_rate": 1.6579840908778714e-05,
+      "loss": 0.5024397373199463,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3082610912799592,
+      "grad_norm": 0.4843386709690094,
+      "learning_rate": 1.6573487645244838e-05,
+      "loss": 0.5023159980773926,
+      "step": 1209
+    },
+    {
+      "epoch": 0.30851606323304437,
+      "grad_norm": 0.4812295436859131,
+      "learning_rate": 1.6567129706049953e-05,
+      "loss": 0.5020912885665894,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3087710351861295,
+      "grad_norm": 0.5185731053352356,
+      "learning_rate": 1.656076709571641e-05,
+      "loss": 0.4940309226512909,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3090260071392147,
+      "grad_norm": 0.5270474553108215,
+      "learning_rate": 1.655439981876987e-05,
+      "loss": 0.5050083994865417,
+      "step": 1212
+    },
+    {
+      "epoch": 0.30928097909229985,
+      "grad_norm": 0.4855048358440399,
+      "learning_rate": 1.6548027879739314e-05,
+      "loss": 0.5098885297775269,
+      "step": 1213
+    },
+    {
+      "epoch": 0.309535951045385,
+      "grad_norm": 0.4971138536930084,
+      "learning_rate": 1.654165128315705e-05,
+      "loss": 0.5043885707855225,
+      "step": 1214
+    },
+    {
+      "epoch": 0.30979092299847016,
+      "grad_norm": 0.47204726934432983,
+      "learning_rate": 1.6535270033558692e-05,
+      "loss": 0.5070247054100037,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3100458949515553,
+      "grad_norm": 0.8165231347084045,
+      "learning_rate": 1.652888413548316e-05,
+      "loss": 0.49824613332748413,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3103008669046405,
+      "grad_norm": 0.4515669047832489,
+      "learning_rate": 1.6522493593472683e-05,
+      "loss": 0.4811893403530121,
+      "step": 1217
+    },
+    {
+      "epoch": 0.31055583885772564,
+      "grad_norm": 0.4702485501766205,
+      "learning_rate": 1.65160984120728e-05,
+      "loss": 0.4961361289024353,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3108108108108108,
+      "grad_norm": 0.4543362855911255,
+      "learning_rate": 1.6509698595832343e-05,
+      "loss": 0.49158987402915955,
+      "step": 1219
+    },
+    {
+      "epoch": 0.31106578276389596,
+      "grad_norm": 0.4930214583873749,
+      "learning_rate": 1.650329414930344e-05,
+      "loss": 0.5052799582481384,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3113207547169811,
+      "grad_norm": 0.45656049251556396,
+      "learning_rate": 1.649688507704152e-05,
+      "loss": 0.49146366119384766,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3115757266700663,
+      "grad_norm": 0.4643915891647339,
+      "learning_rate": 1.649047138360529e-05,
+      "loss": 0.4719155728816986,
+      "step": 1222
+    },
+    {
+      "epoch": 0.31183069862315144,
+      "grad_norm": 0.4875204265117645,
+      "learning_rate": 1.648405307355675e-05,
+      "loss": 0.4880620241165161,
+      "step": 1223
+    },
+    {
+      "epoch": 0.3120856705762366,
+      "grad_norm": 0.8442656993865967,
+      "learning_rate": 1.6477630151461192e-05,
+      "loss": 0.5070099830627441,
+      "step": 1224
+    },
+    {
+      "epoch": 0.31234064252932175,
+      "grad_norm": 0.4763401448726654,
+      "learning_rate": 1.6471202621887173e-05,
+      "loss": 0.505492091178894,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3125956144824069,
+      "grad_norm": 0.44397473335266113,
+      "learning_rate": 1.6464770489406543e-05,
+      "loss": 0.5049811601638794,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3128505864354921,
+      "grad_norm": 0.470048725605011,
+      "learning_rate": 1.6458333758594415e-05,
+      "loss": 0.4918753504753113,
+      "step": 1227
+    },
+    {
+      "epoch": 0.31310555838857723,
+      "grad_norm": 0.4405840039253235,
+      "learning_rate": 1.6451892434029175e-05,
+      "loss": 0.48752328753471375,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3133605303416624,
+      "grad_norm": 0.4632117450237274,
+      "learning_rate": 1.6445446520292477e-05,
+      "loss": 0.4995397925376892,
+      "step": 1229
+    },
+    {
+      "epoch": 0.31361550229474755,
+      "grad_norm": 0.44940271973609924,
+      "learning_rate": 1.6438996021969245e-05,
+      "loss": 0.49737581610679626,
+      "step": 1230
+    },
+    {
+      "epoch": 0.31387047424783276,
+      "grad_norm": 0.45170217752456665,
+      "learning_rate": 1.6432540943647658e-05,
+      "loss": 0.4811992347240448,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3141254462009179,
+      "grad_norm": 0.4569378197193146,
+      "learning_rate": 1.6426081289919144e-05,
+      "loss": 0.4972044825553894,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3143804181540031,
+      "grad_norm": 0.468311607837677,
+      "learning_rate": 1.6419617065378405e-05,
+      "loss": 0.5013965368270874,
+      "step": 1233
+    },
+    {
+      "epoch": 0.31463539010708824,
+      "grad_norm": 0.48914453387260437,
+      "learning_rate": 1.6413148274623384e-05,
+      "loss": 0.5007435083389282,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3148903620601734,
+      "grad_norm": 0.5679212808609009,
+      "learning_rate": 1.6406674922255262e-05,
+      "loss": 0.49458086490631104,
+      "step": 1235
+    },
+    {
+      "epoch": 0.31514533401325856,
+      "grad_norm": 0.4546929597854614,
+      "learning_rate": 1.640019701287849e-05,
+      "loss": 0.4977070689201355,
+      "step": 1236
+    },
+    {
+      "epoch": 0.3154003059663437,
+      "grad_norm": 0.4733176827430725,
+      "learning_rate": 1.6393714551100734e-05,
+      "loss": 0.5150038599967957,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3156552779194289,
+      "grad_norm": 0.4902767837047577,
+      "learning_rate": 1.6387227541532912e-05,
+      "loss": 0.5028160810470581,
+      "step": 1238
+    },
+    {
+      "epoch": 0.31591024987251404,
+      "grad_norm": 0.4904122054576874,
+      "learning_rate": 1.6380735988789172e-05,
+      "loss": 0.48970431089401245,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3161652218255992,
+      "grad_norm": 0.4793023467063904,
+      "learning_rate": 1.63742398974869e-05,
+      "loss": 0.4952084422111511,
+      "step": 1240
+    },
+    {
+      "epoch": 0.31642019377868436,
+      "grad_norm": 0.5026772618293762,
+      "learning_rate": 1.6367739272246702e-05,
+      "loss": 0.48636317253112793,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3166751657317695,
+      "grad_norm": 0.48422929644584656,
+      "learning_rate": 1.6361234117692414e-05,
+      "loss": 0.4868415296077728,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3169301376848547,
+      "grad_norm": 0.4590810239315033,
+      "learning_rate": 1.6354724438451094e-05,
+      "loss": 0.49675533175468445,
+      "step": 1243
+    },
+    {
+      "epoch": 0.31718510963793983,
+      "grad_norm": 0.46764785051345825,
+      "learning_rate": 1.634821023915301e-05,
+      "loss": 0.49604731798171997,
+      "step": 1244
+    },
+    {
+      "epoch": 0.317440081591025,
+      "grad_norm": 0.47263312339782715,
+      "learning_rate": 1.6341691524431662e-05,
+      "loss": 0.5027939081192017,
+      "step": 1245
+    },
+    {
+      "epoch": 0.31769505354411015,
+      "grad_norm": 0.45748889446258545,
+      "learning_rate": 1.6335168298923743e-05,
+      "loss": 0.49228137731552124,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3179500254971953,
+      "grad_norm": 0.473275363445282,
+      "learning_rate": 1.632864056726917e-05,
+      "loss": 0.4914236068725586,
+      "step": 1247
+    },
+    {
+      "epoch": 0.31820499745028047,
+      "grad_norm": 0.45597827434539795,
+      "learning_rate": 1.632210833411105e-05,
+      "loss": 0.4925191104412079,
+      "step": 1248
+    },
+    {
+      "epoch": 0.31845996940336563,
+      "grad_norm": 0.4696197211742401,
+      "learning_rate": 1.6315571604095706e-05,
+      "loss": 0.4934743046760559,
+      "step": 1249
+    },
+    {
+      "epoch": 0.3187149413564508,
+      "grad_norm": 0.4852502644062042,
+      "learning_rate": 1.6309030381872656e-05,
+      "loss": 0.496171772480011,
+      "step": 1250
+    },
+    {
+      "epoch": 0.31896991330953595,
+      "grad_norm": 0.4474034905433655,
+      "learning_rate": 1.6302484672094604e-05,
+      "loss": 0.4935624301433563,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3192248852626211,
+      "grad_norm": 0.47168582677841187,
+      "learning_rate": 1.6295934479417454e-05,
+      "loss": 0.4883151054382324,
+      "step": 1252
+    },
+    {
+      "epoch": 0.31947985721570626,
+      "grad_norm": 0.4466879367828369,
+      "learning_rate": 1.6289379808500303e-05,
+      "loss": 0.4885556697845459,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3197348291687914,
+      "grad_norm": 0.47400885820388794,
+      "learning_rate": 1.628282066400542e-05,
+      "loss": 0.49073100090026855,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3199898011218766,
+      "grad_norm": 0.4575072228908539,
+      "learning_rate": 1.6276257050598273e-05,
+      "loss": 0.4954169988632202,
+      "step": 1255
+    },
+    {
+      "epoch": 0.32024477307496174,
+      "grad_norm": 0.4483925700187683,
+      "learning_rate": 1.626968897294749e-05,
+      "loss": 0.4948553442955017,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3204997450280469,
+      "grad_norm": 0.4554179906845093,
+      "learning_rate": 1.626311643572489e-05,
+      "loss": 0.4990859925746918,
+      "step": 1257
+    },
+    {
+      "epoch": 0.32075471698113206,
+      "grad_norm": 0.5914171934127808,
+      "learning_rate": 1.625653944360546e-05,
+      "loss": 0.49615174531936646,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3210096889342172,
+      "grad_norm": 0.4487502872943878,
+      "learning_rate": 1.6249958001267345e-05,
+      "loss": 0.503517746925354,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3212646608873024,
+      "grad_norm": 0.4658033549785614,
+      "learning_rate": 1.624337211339187e-05,
+      "loss": 0.49343857169151306,
+      "step": 1260
+    },
+    {
+      "epoch": 0.32151963284038754,
+      "grad_norm": 0.4669674038887024,
+      "learning_rate": 1.623678178466352e-05,
+      "loss": 0.5054748058319092,
+      "step": 1261
+    },
+    {
+      "epoch": 0.3217746047934727,
+      "grad_norm": 0.4506138265132904,
+      "learning_rate": 1.623018701976993e-05,
+      "loss": 0.4952157437801361,
+      "step": 1262
+    },
+    {
+      "epoch": 0.32202957674655786,
+      "grad_norm": 0.47126540541648865,
+      "learning_rate": 1.622358782340189e-05,
+      "loss": 0.5030245780944824,
+      "step": 1263
+    },
+    {
+      "epoch": 0.322284548699643,
+      "grad_norm": 0.44352665543556213,
+      "learning_rate": 1.6216984200253355e-05,
+      "loss": 0.48594093322753906,
+      "step": 1264
+    },
+    {
+      "epoch": 0.3225395206527282,
+      "grad_norm": 0.4284340441226959,
+      "learning_rate": 1.621037615502142e-05,
+      "loss": 0.49887198209762573,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3227944926058134,
+      "grad_norm": 0.44898107647895813,
+      "learning_rate": 1.6203763692406328e-05,
+      "loss": 0.48706620931625366,
+      "step": 1266
+    },
+    {
+      "epoch": 0.32304946455889855,
+      "grad_norm": 0.440380334854126,
+      "learning_rate": 1.619714681711146e-05,
+      "loss": 0.49467596411705017,
+      "step": 1267
+    },
+    {
+      "epoch": 0.3233044365119837,
+      "grad_norm": 0.46733948588371277,
+      "learning_rate": 1.6190525533843337e-05,
+      "loss": 0.5020816922187805,
+      "step": 1268
+    },
+    {
+      "epoch": 0.32355940846506887,
+      "grad_norm": 0.6229173541069031,
+      "learning_rate": 1.618389984731162e-05,
+      "loss": 0.49869582056999207,
+      "step": 1269
+    },
+    {
+      "epoch": 0.323814380418154,
+      "grad_norm": 0.48032182455062866,
+      "learning_rate": 1.6177269762229096e-05,
+      "loss": 0.4977878928184509,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3240693523712392,
+      "grad_norm": 0.46267253160476685,
+      "learning_rate": 1.6170635283311684e-05,
+      "loss": 0.5060316324234009,
+      "step": 1271
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 0.4646252989768982,
+      "learning_rate": 1.6163996415278423e-05,
+      "loss": 0.4943225085735321,
+      "step": 1272
+    },
+    {
+      "epoch": 0.3245792962774095,
+      "grad_norm": 0.44283971190452576,
+      "learning_rate": 1.6157353162851486e-05,
+      "loss": 0.47732409834861755,
+      "step": 1273
+    },
+    {
+      "epoch": 0.32483426823049466,
+      "grad_norm": 0.48292553424835205,
+      "learning_rate": 1.6150705530756155e-05,
+      "loss": 0.5045241117477417,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3250892401835798,
+      "grad_norm": 0.48017552495002747,
+      "learning_rate": 1.6144053523720832e-05,
+      "loss": 0.4933950901031494,
+      "step": 1275
+    },
+    {
+      "epoch": 0.325344212136665,
+      "grad_norm": 0.45280221104621887,
+      "learning_rate": 1.613739714647702e-05,
+      "loss": 0.49421560764312744,
+      "step": 1276
+    },
+    {
+      "epoch": 0.32559918408975014,
+      "grad_norm": 0.48675018548965454,
+      "learning_rate": 1.6130736403759342e-05,
+      "loss": 0.5033859014511108,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3258541560428353,
+      "grad_norm": 0.4708247780799866,
+      "learning_rate": 1.6124071300305527e-05,
+      "loss": 0.4810166358947754,
+      "step": 1278
+    },
+    {
+      "epoch": 0.32610912799592046,
+      "grad_norm": 0.4413648545742035,
+      "learning_rate": 1.61174018408564e-05,
+      "loss": 0.49215754866600037,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3263640999490056,
+      "grad_norm": 0.49567604064941406,
+      "learning_rate": 1.6110728030155875e-05,
+      "loss": 0.500187337398529,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3266190719020908,
+      "grad_norm": 0.47270676493644714,
+      "learning_rate": 1.6104049872950988e-05,
+      "loss": 0.5070374011993408,
+      "step": 1281
+    },
+    {
+      "epoch": 0.32687404385517593,
+      "grad_norm": 0.4364311099052429,
+      "learning_rate": 1.6097367373991842e-05,
+      "loss": 0.5007890462875366,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3271290158082611,
+      "grad_norm": 0.48195508122444153,
+      "learning_rate": 1.609068053803164e-05,
+      "loss": 0.48835358023643494,
+      "step": 1283
+    },
+    {
+      "epoch": 0.32738398776134625,
+      "grad_norm": 0.4663650691509247,
+      "learning_rate": 1.6083989369826664e-05,
+      "loss": 0.4924512803554535,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3276389597144314,
+      "grad_norm": 0.4929714500904083,
+      "learning_rate": 1.607729387413628e-05,
+      "loss": 0.48393774032592773,
+      "step": 1285
+    },
+    {
+      "epoch": 0.32789393166751657,
+      "grad_norm": 0.5165911912918091,
+      "learning_rate": 1.607059405572293e-05,
+      "loss": 0.4980706572532654,
+      "step": 1286
+    },
+    {
+      "epoch": 0.32814890362060173,
+      "grad_norm": 0.44888025522232056,
+      "learning_rate": 1.606388991935214e-05,
+      "loss": 0.5012993812561035,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3284038755736869,
+      "grad_norm": 0.4578009843826294,
+      "learning_rate": 1.6057181469792497e-05,
+      "loss": 0.49674296379089355,
+      "step": 1288
+    },
+    {
+      "epoch": 0.32865884752677205,
+      "grad_norm": 0.4720461368560791,
+      "learning_rate": 1.6050468711815658e-05,
+      "loss": 0.4820350110530853,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3289138194798572,
+      "grad_norm": 0.4323115050792694,
+      "learning_rate": 1.6043751650196344e-05,
+      "loss": 0.48184940218925476,
+      "step": 1290
+    },
+    {
+      "epoch": 0.32916879143294236,
+      "grad_norm": 0.5161405205726624,
+      "learning_rate": 1.6037030289712344e-05,
+      "loss": 0.5068054795265198,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3294237633860275,
+      "grad_norm": 0.44478678703308105,
+      "learning_rate": 1.6030304635144493e-05,
+      "loss": 0.5015352368354797,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3296787353391127,
+      "grad_norm": 0.46428152918815613,
+      "learning_rate": 1.6023574691276693e-05,
+      "loss": 0.49676769971847534,
+      "step": 1293
+    },
+    {
+      "epoch": 0.32993370729219784,
+      "grad_norm": 0.45837974548339844,
+      "learning_rate": 1.6016840462895887e-05,
+      "loss": 0.5077446103096008,
+      "step": 1294
+    },
+    {
+      "epoch": 0.330188679245283,
+      "grad_norm": 0.48795589804649353,
+      "learning_rate": 1.6010101954792073e-05,
+      "loss": 0.5017638206481934,
+      "step": 1295
+    },
+    {
+      "epoch": 0.33044365119836816,
+      "grad_norm": 0.4359217882156372,
+      "learning_rate": 1.6003359171758286e-05,
+      "loss": 0.5008919835090637,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3306986231514533,
+      "grad_norm": 0.4801797568798065,
+      "learning_rate": 1.5996612118590604e-05,
+      "loss": 0.5036278963088989,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3309535951045385,
+      "grad_norm": 0.45398226380348206,
+      "learning_rate": 1.598986080008815e-05,
+      "loss": 0.48134273290634155,
+      "step": 1298
+    },
+    {
+      "epoch": 0.33120856705762364,
+      "grad_norm": 0.469053715467453,
+      "learning_rate": 1.598310522105307e-05,
+      "loss": 0.4934031367301941,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3314635390107088,
+      "grad_norm": 0.44086888432502747,
+      "learning_rate": 1.5976345386290546e-05,
+      "loss": 0.499302476644516,
+      "step": 1300
+    },
+    {
+      "epoch": 0.33171851096379396,
+      "grad_norm": 0.47684571146965027,
+      "learning_rate": 1.5969581300608783e-05,
+      "loss": 0.4984762668609619,
+      "step": 1301
+    },
+    {
+      "epoch": 0.33197348291687917,
+      "grad_norm": 0.4727766215801239,
+      "learning_rate": 1.5962812968819018e-05,
+      "loss": 0.498884379863739,
+      "step": 1302
+    },
+    {
+      "epoch": 0.33222845486996433,
+      "grad_norm": 0.4573783874511719,
+      "learning_rate": 1.5956040395735498e-05,
+      "loss": 0.49887561798095703,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3324834268230495,
+      "grad_norm": 0.4735158681869507,
+      "learning_rate": 1.5949263586175492e-05,
+      "loss": 0.49824100732803345,
+      "step": 1304
+    },
+    {
+      "epoch": 0.33273839877613465,
+      "grad_norm": 0.45007798075675964,
+      "learning_rate": 1.5942482544959286e-05,
+      "loss": 0.48702430725097656,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3329933707292198,
+      "grad_norm": 0.44589129090309143,
+      "learning_rate": 1.593569727691017e-05,
+      "loss": 0.49279114603996277,
+      "step": 1306
+    },
+    {
+      "epoch": 0.33324834268230497,
+      "grad_norm": 0.4649209976196289,
+      "learning_rate": 1.592890778685444e-05,
+      "loss": 0.47921788692474365,
+      "step": 1307
+    },
+    {
+      "epoch": 0.3335033146353901,
+      "grad_norm": 0.4656672775745392,
+      "learning_rate": 1.59221140796214e-05,
+      "loss": 0.5048980712890625,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3337582865884753,
+      "grad_norm": 0.434708833694458,
+      "learning_rate": 1.5915316160043355e-05,
+      "loss": 0.504131555557251,
+      "step": 1309
+    },
+    {
+      "epoch": 0.33401325854156044,
+      "grad_norm": 0.443644255399704,
+      "learning_rate": 1.5908514032955592e-05,
+      "loss": 0.4883410930633545,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3342682304946456,
+      "grad_norm": 0.47313472628593445,
+      "learning_rate": 1.5901707703196412e-05,
+      "loss": 0.49209117889404297,
+      "step": 1311
+    },
+    {
+      "epoch": 0.33452320244773076,
+      "grad_norm": 0.4823039174079895,
+      "learning_rate": 1.589489717560709e-05,
+      "loss": 0.4971883296966553,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3347781744008159,
+      "grad_norm": 0.4580431282520294,
+      "learning_rate": 1.5888082455031885e-05,
+      "loss": 0.49591752886772156,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3350331463539011,
+      "grad_norm": 0.5018422603607178,
+      "learning_rate": 1.5881263546318052e-05,
+      "loss": 0.4978068768978119,
+      "step": 1314
+    },
+    {
+      "epoch": 0.33528811830698624,
+      "grad_norm": 0.5090048313140869,
+      "learning_rate": 1.5874440454315823e-05,
+      "loss": 0.4929397702217102,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3355430902600714,
+      "grad_norm": 0.4691426157951355,
+      "learning_rate": 1.5867613183878387e-05,
+      "loss": 0.49518275260925293,
+      "step": 1316
+    },
+    {
+      "epoch": 0.33579806221315656,
+      "grad_norm": 0.48400259017944336,
+      "learning_rate": 1.5860781739861928e-05,
+      "loss": 0.48870205879211426,
+      "step": 1317
+    },
+    {
+      "epoch": 0.3360530341662417,
+      "grad_norm": 0.48520103096961975,
+      "learning_rate": 1.5853946127125585e-05,
+      "loss": 0.4775869846343994,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3363080061193269,
+      "grad_norm": 0.4813341796398163,
+      "learning_rate": 1.584710635053147e-05,
+      "loss": 0.47866344451904297,
+      "step": 1319
+    },
+    {
+      "epoch": 0.33656297807241203,
+      "grad_norm": 0.46262988448143005,
+      "learning_rate": 1.5840262414944647e-05,
+      "loss": 0.48800715804100037,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3368179500254972,
+      "grad_norm": 0.4679766595363617,
+      "learning_rate": 1.583341432523315e-05,
+      "loss": 0.48198017477989197,
+      "step": 1321
+    },
+    {
+      "epoch": 0.33707292197858235,
+      "grad_norm": 0.458714097738266,
+      "learning_rate": 1.5826562086267956e-05,
+      "loss": 0.4898045063018799,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3373278939316675,
+      "grad_norm": 0.4584973156452179,
+      "learning_rate": 1.5819705702923004e-05,
+      "loss": 0.4798343777656555,
+      "step": 1323
+    },
+    {
+      "epoch": 0.33758286588475267,
+      "grad_norm": 0.4500448405742645,
+      "learning_rate": 1.5812845180075175e-05,
+      "loss": 0.49616456031799316,
+      "step": 1324
+    },
+    {
+      "epoch": 0.33783783783783783,
+      "grad_norm": 0.43624138832092285,
+      "learning_rate": 1.5805980522604295e-05,
+      "loss": 0.49544939398765564,
+      "step": 1325
+    },
+    {
+      "epoch": 0.338092809790923,
+      "grad_norm": 0.44867441058158875,
+      "learning_rate": 1.579911173539313e-05,
+      "loss": 0.49614858627319336,
+      "step": 1326
+    },
+    {
+      "epoch": 0.33834778174400815,
+      "grad_norm": 0.45365074276924133,
+      "learning_rate": 1.5792238823327387e-05,
+      "loss": 0.48979371786117554,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3386027536970933,
+      "grad_norm": 0.4778674840927124,
+      "learning_rate": 1.5785361791295706e-05,
+      "loss": 0.4943079948425293,
+      "step": 1328
+    },
+    {
+      "epoch": 0.33885772565017847,
+      "grad_norm": 0.45407992601394653,
+      "learning_rate": 1.5778480644189654e-05,
+      "loss": 0.501994252204895,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3391126976032636,
+      "grad_norm": 0.47102952003479004,
+      "learning_rate": 1.5771595386903728e-05,
+      "loss": 0.4933057725429535,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3393676695563488,
+      "grad_norm": 0.5028448700904846,
+      "learning_rate": 1.5764706024335343e-05,
+      "loss": 0.48624253273010254,
+      "step": 1331
+    },
+    {
+      "epoch": 0.33962264150943394,
+      "grad_norm": 0.4622834324836731,
+      "learning_rate": 1.5757812561384848e-05,
+      "loss": 0.5023224353790283,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3398776134625191,
+      "grad_norm": 0.4559676945209503,
+      "learning_rate": 1.5750915002955497e-05,
+      "loss": 0.4845350682735443,
+      "step": 1333
+    },
+    {
+      "epoch": 0.34013258541560426,
+      "grad_norm": 0.5004395246505737,
+      "learning_rate": 1.5744013353953455e-05,
+      "loss": 0.4978603720664978,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3403875573686894,
+      "grad_norm": 0.4740391969680786,
+      "learning_rate": 1.5737107619287802e-05,
+      "loss": 0.4802100658416748,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3406425293217746,
+      "grad_norm": 0.4601091146469116,
+      "learning_rate": 1.573019780387053e-05,
+      "loss": 0.49661630392074585,
+      "step": 1336
+    },
+    {
+      "epoch": 0.34089750127485974,
+      "grad_norm": 0.43999335169792175,
+      "learning_rate": 1.5723283912616515e-05,
+      "loss": 0.4958333373069763,
+      "step": 1337
+    },
+    {
+      "epoch": 0.34115247322794495,
+      "grad_norm": 0.4198979139328003,
+      "learning_rate": 1.5716365950443556e-05,
+      "loss": 0.49501991271972656,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3414074451810301,
+      "grad_norm": 0.45959314703941345,
+      "learning_rate": 1.5709443922272326e-05,
+      "loss": 0.4792822599411011,
+      "step": 1339
+    },
+    {
+      "epoch": 0.34166241713411527,
+      "grad_norm": 0.47105738520622253,
+      "learning_rate": 1.5702517833026408e-05,
+      "loss": 0.4770154654979706,
+      "step": 1340
+    },
+    {
+      "epoch": 0.34191738908720043,
+      "grad_norm": 0.4640108048915863,
+      "learning_rate": 1.569558768763226e-05,
+      "loss": 0.4815002381801605,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3421723610402856,
+      "grad_norm": 0.4672930836677551,
+      "learning_rate": 1.568865349101923e-05,
+      "loss": 0.4833976626396179,
+      "step": 1342
+    },
+    {
+      "epoch": 0.34242733299337075,
+      "grad_norm": 0.467336505651474,
+      "learning_rate": 1.5681715248119552e-05,
+      "loss": 0.4880855977535248,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3426823049464559,
+      "grad_norm": 0.45184525847435,
+      "learning_rate": 1.5674772963868333e-05,
+      "loss": 0.5052072405815125,
+      "step": 1344
+    },
+    {
+      "epoch": 0.34293727689954107,
+      "grad_norm": 0.4374590516090393,
+      "learning_rate": 1.566782664320355e-05,
+      "loss": 0.4983629584312439,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3431922488526262,
+      "grad_norm": 0.4904771149158478,
+      "learning_rate": 1.5660876291066066e-05,
+      "loss": 0.49454396963119507,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3434472208057114,
+      "grad_norm": 0.43185144662857056,
+      "learning_rate": 1.565392191239959e-05,
+      "loss": 0.4821247458457947,
+      "step": 1347
+    },
+    {
+      "epoch": 0.34370219275879654,
+      "grad_norm": 0.4500340223312378,
+      "learning_rate": 1.564696351215072e-05,
+      "loss": 0.5049868226051331,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3439571647118817,
+      "grad_norm": 0.44509389996528625,
+      "learning_rate": 1.5640001095268894e-05,
+      "loss": 0.4985738694667816,
+      "step": 1349
+    },
+    {
+      "epoch": 0.34421213666496686,
+      "grad_norm": 0.5017538666725159,
+      "learning_rate": 1.563303466670642e-05,
+      "loss": 0.47815823554992676,
+      "step": 1350
+    },
+    {
+      "epoch": 0.344467108618052,
+      "grad_norm": 0.4683600664138794,
+      "learning_rate": 1.562606423141845e-05,
+      "loss": 0.5037387609481812,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3447220805711372,
+      "grad_norm": 0.46206045150756836,
+      "learning_rate": 1.561908979436299e-05,
+      "loss": 0.49229001998901367,
+      "step": 1352
+    },
+    {
+      "epoch": 0.34497705252422234,
+      "grad_norm": 0.4904536008834839,
+      "learning_rate": 1.561211136050089e-05,
+      "loss": 0.4933377206325531,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3452320244773075,
+      "grad_norm": 0.47214558720588684,
+      "learning_rate": 1.5605128934795858e-05,
+      "loss": 0.4849925935268402,
+      "step": 1354
+    },
+    {
+      "epoch": 0.34548699643039266,
+      "grad_norm": 0.4356664717197418,
+      "learning_rate": 1.5598142522214417e-05,
+      "loss": 0.48362982273101807,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3457419683834778,
+      "grad_norm": 0.5014368295669556,
+      "learning_rate": 1.559115212772594e-05,
+      "loss": 0.4885830283164978,
+      "step": 1356
+    },
+    {
+      "epoch": 0.345996940336563,
+      "grad_norm": 0.4560520052909851,
+      "learning_rate": 1.5584157756302635e-05,
+      "loss": 0.4855990409851074,
+      "step": 1357
+    },
+    {
+      "epoch": 0.34625191228964813,
+      "grad_norm": 0.44068634510040283,
+      "learning_rate": 1.5577159412919524e-05,
+      "loss": 0.4836045205593109,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3465068842427333,
+      "grad_norm": 0.4414604604244232,
+      "learning_rate": 1.5570157102554475e-05,
+      "loss": 0.5005707144737244,
+      "step": 1359
+    },
+    {
+      "epoch": 0.34676185619581845,
+      "grad_norm": 0.47479185461997986,
+      "learning_rate": 1.556315083018816e-05,
+      "loss": 0.4892706573009491,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3470168281489036,
+      "grad_norm": 0.4664304256439209,
+      "learning_rate": 1.5556140600804076e-05,
+      "loss": 0.48691055178642273,
+      "step": 1361
+    },
+    {
+      "epoch": 0.34727180010198877,
+      "grad_norm": 0.4774048626422882,
+      "learning_rate": 1.5549126419388538e-05,
+      "loss": 0.49023565649986267,
+      "step": 1362
+    },
+    {
+      "epoch": 0.34752677205507393,
+      "grad_norm": 0.4945896565914154,
+      "learning_rate": 1.554210829093066e-05,
+      "loss": 0.49875587224960327,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3477817440081591,
+      "grad_norm": 0.49446311593055725,
+      "learning_rate": 1.5535086220422386e-05,
+      "loss": 0.5095691680908203,
+      "step": 1364
+    },
+    {
+      "epoch": 0.34803671596124425,
+      "grad_norm": 0.5272205471992493,
+      "learning_rate": 1.5528060212858436e-05,
+      "loss": 0.49271202087402344,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3482916879143294,
+      "grad_norm": 0.45778605341911316,
+      "learning_rate": 1.5521030273236354e-05,
+      "loss": 0.4978727698326111,
+      "step": 1366
+    },
+    {
+      "epoch": 0.34854665986741457,
+      "grad_norm": 0.4778561592102051,
+      "learning_rate": 1.5513996406556465e-05,
+      "loss": 0.4920888841152191,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3488016318204997,
+      "grad_norm": 0.47162318229675293,
+      "learning_rate": 1.55069586178219e-05,
+      "loss": 0.4907873570919037,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3490566037735849,
+      "grad_norm": 0.4643825590610504,
+      "learning_rate": 1.5499916912038563e-05,
+      "loss": 0.4940566420555115,
+      "step": 1369
+    },
+    {
+      "epoch": 0.34931157572667004,
+      "grad_norm": 0.48091721534729004,
+      "learning_rate": 1.5492871294215162e-05,
+      "loss": 0.48381727933883667,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3495665476797552,
+      "grad_norm": 0.4480636715888977,
+      "learning_rate": 1.5485821769363178e-05,
+      "loss": 0.4959699511528015,
+      "step": 1371
+    },
+    {
+      "epoch": 0.34982151963284036,
+      "grad_norm": 0.44783326983451843,
+      "learning_rate": 1.5478768342496872e-05,
+      "loss": 0.48227232694625854,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3500764915859256,
+      "grad_norm": 0.45729002356529236,
+      "learning_rate": 1.547171101863328e-05,
+      "loss": 0.4862489700317383,
+      "step": 1373
+    },
+    {
+      "epoch": 0.35033146353901073,
+      "grad_norm": 0.4617900848388672,
+      "learning_rate": 1.5464649802792215e-05,
+      "loss": 0.4825901985168457,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3505864354920959,
+      "grad_norm": 0.45521557331085205,
+      "learning_rate": 1.5457584699996253e-05,
+      "loss": 0.47123441100120544,
+      "step": 1375
+    },
+    {
+      "epoch": 0.35084140744518105,
+      "grad_norm": 0.49398496747016907,
+      "learning_rate": 1.5450515715270733e-05,
+      "loss": 0.4958046078681946,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3510963793982662,
+      "grad_norm": 0.506957471370697,
+      "learning_rate": 1.5443442853643765e-05,
+      "loss": 0.5013394355773926,
+      "step": 1377
+    },
+    {
+      "epoch": 0.35135135135135137,
+      "grad_norm": 0.467107892036438,
+      "learning_rate": 1.5436366120146206e-05,
+      "loss": 0.48585793375968933,
+      "step": 1378
+    },
+    {
+      "epoch": 0.35160632330443653,
+      "grad_norm": 0.4466833174228668,
+      "learning_rate": 1.5429285519811672e-05,
+      "loss": 0.5096202492713928,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3518612952575217,
+      "grad_norm": 0.5030296444892883,
+      "learning_rate": 1.5422201057676525e-05,
+      "loss": 0.4857207238674164,
+      "step": 1380
+    },
+    {
+      "epoch": 0.35211626721060685,
+      "grad_norm": 0.45916059613227844,
+      "learning_rate": 1.5415112738779887e-05,
+      "loss": 0.4840683937072754,
+      "step": 1381
+    },
+    {
+      "epoch": 0.352371239163692,
+      "grad_norm": 0.4854084849357605,
+      "learning_rate": 1.5408020568163602e-05,
+      "loss": 0.48948368430137634,
+      "step": 1382
+    },
+    {
+      "epoch": 0.35262621111677717,
+      "grad_norm": 0.49546507000923157,
+      "learning_rate": 1.540092455087227e-05,
+      "loss": 0.4880248010158539,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3528811830698623,
+      "grad_norm": 0.47287800908088684,
+      "learning_rate": 1.539382469195323e-05,
+      "loss": 0.5022507905960083,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3531361550229475,
+      "grad_norm": 0.4680861830711365,
+      "learning_rate": 1.5386720996456533e-05,
+      "loss": 0.4974364936351776,
+      "step": 1385
+    },
+    {
+      "epoch": 0.35339112697603264,
+      "grad_norm": 0.4794875383377075,
+      "learning_rate": 1.5379613469434983e-05,
+      "loss": 0.50507652759552,
+      "step": 1386
+    },
+    {
+      "epoch": 0.3536460989291178,
+      "grad_norm": 0.463410347700119,
+      "learning_rate": 1.537250211594409e-05,
+      "loss": 0.4969019889831543,
+      "step": 1387
+    },
+    {
+      "epoch": 0.35390107088220296,
+      "grad_norm": 0.4773065447807312,
+      "learning_rate": 1.53653869410421e-05,
+      "loss": 0.4828874468803406,
+      "step": 1388
+    },
+    {
+      "epoch": 0.3541560428352881,
+      "grad_norm": 0.5189251899719238,
+      "learning_rate": 1.5358267949789968e-05,
+      "loss": 0.4929976463317871,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3544110147883733,
+      "grad_norm": 0.49028658866882324,
+      "learning_rate": 1.5351145147251364e-05,
+      "loss": 0.49443262815475464,
+      "step": 1390
+    },
+    {
+      "epoch": 0.35466598674145844,
+      "grad_norm": 0.47114813327789307,
+      "learning_rate": 1.5344018538492682e-05,
+      "loss": 0.494369238615036,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3549209586945436,
+      "grad_norm": 0.4885484278202057,
+      "learning_rate": 1.5336888128583002e-05,
+      "loss": 0.48555788397789,
+      "step": 1392
+    },
+    {
+      "epoch": 0.35517593064762876,
+      "grad_norm": 0.47980996966362,
+      "learning_rate": 1.5329753922594122e-05,
+      "loss": 0.4955976903438568,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3554309026007139,
+      "grad_norm": 0.4988878667354584,
+      "learning_rate": 1.532261592560054e-05,
+      "loss": 0.49415457248687744,
+      "step": 1394
+    },
+    {
+      "epoch": 0.3556858745537991,
+      "grad_norm": 0.49632182717323303,
+      "learning_rate": 1.5315474142679442e-05,
+      "loss": 0.4775877892971039,
+      "step": 1395
+    },
+    {
+      "epoch": 0.35594084650688423,
+      "grad_norm": 0.48797720670700073,
+      "learning_rate": 1.530832857891072e-05,
+      "loss": 0.4879951477050781,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3561958184599694,
+      "grad_norm": 0.48648008704185486,
+      "learning_rate": 1.5301179239376936e-05,
+      "loss": 0.48652490973472595,
+      "step": 1397
+    },
+    {
+      "epoch": 0.35645079041305455,
+      "grad_norm": 0.46025142073631287,
+      "learning_rate": 1.529402612916336e-05,
+      "loss": 0.5025362968444824,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3567057623661397,
+      "grad_norm": 0.4712684750556946,
+      "learning_rate": 1.5286869253357925e-05,
+      "loss": 0.48671525716781616,
+      "step": 1399
+    },
+    {
+      "epoch": 0.35696073431922487,
+      "grad_norm": 0.4850005507469177,
+      "learning_rate": 1.5279708617051258e-05,
+      "loss": 0.4923925995826721,
+      "step": 1400
+    },
+    {
+      "epoch": 0.35721570627231003,
+      "grad_norm": 0.455744206905365,
+      "learning_rate": 1.5272544225336646e-05,
+      "loss": 0.4913860261440277,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3574706782253952,
+      "grad_norm": 0.4736967086791992,
+      "learning_rate": 1.5265376083310063e-05,
+      "loss": 0.4790980815887451,
+      "step": 1402
+    },
+    {
+      "epoch": 0.35772565017848035,
+      "grad_norm": 0.5233493447303772,
+      "learning_rate": 1.5258204196070136e-05,
+      "loss": 0.4792144298553467,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3579806221315655,
+      "grad_norm": 0.45144742727279663,
+      "learning_rate": 1.5251028568718163e-05,
+      "loss": 0.4849224090576172,
+      "step": 1404
+    },
+    {
+      "epoch": 0.35823559408465067,
+      "grad_norm": 0.45398980379104614,
+      "learning_rate": 1.5243849206358099e-05,
+      "loss": 0.4965559244155884,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3584905660377358,
+      "grad_norm": 0.49643367528915405,
+      "learning_rate": 1.5236666114096562e-05,
+      "loss": 0.4921931028366089,
+      "step": 1406
+    },
+    {
+      "epoch": 0.358745537990821,
+      "grad_norm": 0.4721378982067108,
+      "learning_rate": 1.5229479297042825e-05,
+      "loss": 0.49166640639305115,
+      "step": 1407
+    },
+    {
+      "epoch": 0.35900050994390614,
+      "grad_norm": 0.4492199718952179,
+      "learning_rate": 1.5222288760308792e-05,
+      "loss": 0.4912944734096527,
+      "step": 1408
+    },
+    {
+      "epoch": 0.35925548189699136,
+      "grad_norm": 0.4733271598815918,
+      "learning_rate": 1.521509450900903e-05,
+      "loss": 0.48487481474876404,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3595104538500765,
+      "grad_norm": 0.45325836539268494,
+      "learning_rate": 1.520789654826075e-05,
+      "loss": 0.4862060546875,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3597654258031617,
+      "grad_norm": 0.43667829036712646,
+      "learning_rate": 1.5200694883183787e-05,
+      "loss": 0.49290984869003296,
+      "step": 1411
+    },
+    {
+      "epoch": 0.36002039775624683,
+      "grad_norm": 0.4515577256679535,
+      "learning_rate": 1.519348951890062e-05,
+      "loss": 0.4928954541683197,
+      "step": 1412
+    },
+    {
+      "epoch": 0.360275369709332,
+      "grad_norm": 0.4574653208255768,
+      "learning_rate": 1.5186280460536358e-05,
+      "loss": 0.49473699927330017,
+      "step": 1413
+    },
+    {
+      "epoch": 0.36053034166241715,
+      "grad_norm": 0.4594244062900543,
+      "learning_rate": 1.5179067713218743e-05,
+      "loss": 0.4945780038833618,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3607853136155023,
+      "grad_norm": 0.46252748370170593,
+      "learning_rate": 1.517185128207813e-05,
+      "loss": 0.47525572776794434,
+      "step": 1415
+    },
+    {
+      "epoch": 0.36104028556858747,
+      "grad_norm": 0.4991248846054077,
+      "learning_rate": 1.5164631172247501e-05,
+      "loss": 0.48477134108543396,
+      "step": 1416
+    },
+    {
+      "epoch": 0.36129525752167263,
+      "grad_norm": 0.45575839281082153,
+      "learning_rate": 1.5157407388862452e-05,
+      "loss": 0.4752116799354553,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3615502294747578,
+      "grad_norm": 0.48317432403564453,
+      "learning_rate": 1.5150179937061196e-05,
+      "loss": 0.48912566900253296,
+      "step": 1418
+    },
+    {
+      "epoch": 0.36180520142784295,
+      "grad_norm": 0.7431215643882751,
+      "learning_rate": 1.5142948821984553e-05,
+      "loss": 0.4992019534111023,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3620601733809281,
+      "grad_norm": 0.45986902713775635,
+      "learning_rate": 1.513571404877595e-05,
+      "loss": 0.498704195022583,
+      "step": 1420
+    },
+    {
+      "epoch": 0.36231514533401327,
+      "grad_norm": 0.457722544670105,
+      "learning_rate": 1.5128475622581404e-05,
+      "loss": 0.48822641372680664,
+      "step": 1421
+    },
+    {
+      "epoch": 0.3625701172870984,
+      "grad_norm": 0.4633914828300476,
+      "learning_rate": 1.512123354854955e-05,
+      "loss": 0.4874158799648285,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3628250892401836,
+      "grad_norm": 0.48305845260620117,
+      "learning_rate": 1.5113987831831609e-05,
+      "loss": 0.49463486671447754,
+      "step": 1423
+    },
+    {
+      "epoch": 0.36308006119326874,
+      "grad_norm": 0.44547805190086365,
+      "learning_rate": 1.5106738477581388e-05,
+      "loss": 0.4956837296485901,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3633350331463539,
+      "grad_norm": 0.4877088963985443,
+      "learning_rate": 1.5099485490955288e-05,
+      "loss": 0.47981151938438416,
+      "step": 1425
+    },
+    {
+      "epoch": 0.36359000509943906,
+      "grad_norm": 0.44127556681632996,
+      "learning_rate": 1.509222887711229e-05,
+      "loss": 0.4929065704345703,
+      "step": 1426
+    },
+    {
+      "epoch": 0.3638449770525242,
+      "grad_norm": 0.47561880946159363,
+      "learning_rate": 1.5084968641213958e-05,
+      "loss": 0.48719948530197144,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3640999490056094,
+      "grad_norm": 0.4477303922176361,
+      "learning_rate": 1.507770478842443e-05,
+      "loss": 0.4831518828868866,
+      "step": 1428
+    },
+    {
+      "epoch": 0.36435492095869454,
+      "grad_norm": 0.45889630913734436,
+      "learning_rate": 1.5070437323910414e-05,
+      "loss": 0.4897419810295105,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3646098929117797,
+      "grad_norm": 0.46235835552215576,
+      "learning_rate": 1.5063166252841197e-05,
+      "loss": 0.48954832553863525,
+      "step": 1430
+    },
+    {
+      "epoch": 0.36486486486486486,
+      "grad_norm": 0.4915541410446167,
+      "learning_rate": 1.5055891580388617e-05,
+      "loss": 0.49446234107017517,
+      "step": 1431
+    },
+    {
+      "epoch": 0.36511983681795,
+      "grad_norm": 0.44475996494293213,
+      "learning_rate": 1.5048613311727089e-05,
+      "loss": 0.489834725856781,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3653748087710352,
+      "grad_norm": 0.5123434066772461,
+      "learning_rate": 1.5041331452033572e-05,
+      "loss": 0.48975151777267456,
+      "step": 1433
+    },
+    {
+      "epoch": 0.36562978072412033,
+      "grad_norm": 0.45602652430534363,
+      "learning_rate": 1.5034046006487591e-05,
+      "loss": 0.49672383069992065,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3658847526772055,
+      "grad_norm": 0.5002124905586243,
+      "learning_rate": 1.502675698027121e-05,
+      "loss": 0.48555001616477966,
+      "step": 1435
+    },
+    {
+      "epoch": 0.36613972463029065,
+      "grad_norm": 0.46187663078308105,
+      "learning_rate": 1.5019464378569054e-05,
+      "loss": 0.4871925115585327,
+      "step": 1436
+    },
+    {
+      "epoch": 0.3663946965833758,
+      "grad_norm": 0.4582308232784271,
+      "learning_rate": 1.501216820656827e-05,
+      "loss": 0.4774112105369568,
+      "step": 1437
+    },
+    {
+      "epoch": 0.36664966853646097,
+      "grad_norm": 0.44312283396720886,
+      "learning_rate": 1.5004868469458572e-05,
+      "loss": 0.4840548038482666,
+      "step": 1438
+    },
+    {
+      "epoch": 0.36690464048954613,
+      "grad_norm": 0.4665316939353943,
+      "learning_rate": 1.4997565172432189e-05,
+      "loss": 0.5036348700523376,
+      "step": 1439
+    },
+    {
+      "epoch": 0.3671596124426313,
+      "grad_norm": 0.4862591326236725,
+      "learning_rate": 1.4990258320683888e-05,
+      "loss": 0.48736369609832764,
+      "step": 1440
+    },
+    {
+      "epoch": 0.36741458439571645,
+      "grad_norm": 0.4475901424884796,
+      "learning_rate": 1.4982947919410966e-05,
+      "loss": 0.4928306043148041,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3676695563488016,
+      "grad_norm": 0.4491722285747528,
+      "learning_rate": 1.4975633973813244e-05,
+      "loss": 0.47156596183776855,
+      "step": 1442
+    },
+    {
+      "epoch": 0.36792452830188677,
+      "grad_norm": 0.4628240168094635,
+      "learning_rate": 1.4968316489093067e-05,
+      "loss": 0.5090243220329285,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3681795002549719,
+      "grad_norm": 0.4485695958137512,
+      "learning_rate": 1.4960995470455288e-05,
+      "loss": 0.5036436319351196,
+      "step": 1444
+    },
+    {
+      "epoch": 0.36843447220805714,
+      "grad_norm": 0.439031720161438,
+      "learning_rate": 1.4953670923107284e-05,
+      "loss": 0.4707303047180176,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3686894441611423,
+      "grad_norm": 0.4461120069026947,
+      "learning_rate": 1.4946342852258937e-05,
+      "loss": 0.4881828725337982,
+      "step": 1446
+    },
+    {
+      "epoch": 0.36894441611422746,
+      "grad_norm": 0.5070061683654785,
+      "learning_rate": 1.4939011263122635e-05,
+      "loss": 0.49580106139183044,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3691993880673126,
+      "grad_norm": 0.4657633900642395,
+      "learning_rate": 1.4931676160913276e-05,
+      "loss": 0.48893028497695923,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3694543600203978,
+      "grad_norm": 0.44490692019462585,
+      "learning_rate": 1.4924337550848238e-05,
+      "loss": 0.4781566262245178,
+      "step": 1449
+    },
+    {
+      "epoch": 0.36970933197348294,
+      "grad_norm": 0.4616687595844269,
+      "learning_rate": 1.4916995438147417e-05,
+      "loss": 0.4804666042327881,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3699643039265681,
+      "grad_norm": 0.49199023842811584,
+      "learning_rate": 1.4909649828033188e-05,
+      "loss": 0.49582383036613464,
+      "step": 1451
+    },
+    {
+      "epoch": 0.37021927587965325,
+      "grad_norm": 0.4535580277442932,
+      "learning_rate": 1.4902300725730414e-05,
+      "loss": 0.4697763919830322,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3704742478327384,
+      "grad_norm": 0.5966619253158569,
+      "learning_rate": 1.489494813646644e-05,
+      "loss": 0.48284289240837097,
+      "step": 1453
+    },
+    {
+      "epoch": 0.37072921978582357,
+      "grad_norm": 0.4896046817302704,
+      "learning_rate": 1.48875920654711e-05,
+      "loss": 0.4833661913871765,
+      "step": 1454
+    },
+    {
+      "epoch": 0.37098419173890873,
+      "grad_norm": 0.4610314965248108,
+      "learning_rate": 1.48802325179767e-05,
+      "loss": 0.48917660117149353,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3712391636919939,
+      "grad_norm": 0.44801920652389526,
+      "learning_rate": 1.4872869499218015e-05,
+      "loss": 0.48730477690696716,
+      "step": 1456
+    },
+    {
+      "epoch": 0.37149413564507905,
+      "grad_norm": 0.47437381744384766,
+      "learning_rate": 1.4865503014432291e-05,
+      "loss": 0.4959720969200134,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3717491075981642,
+      "grad_norm": 0.4891728162765503,
+      "learning_rate": 1.4858133068859245e-05,
+      "loss": 0.4845280349254608,
+      "step": 1458
+    },
+    {
+      "epoch": 0.37200407955124937,
+      "grad_norm": 0.46612614393234253,
+      "learning_rate": 1.4850759667741046e-05,
+      "loss": 0.4805614948272705,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3722590515043345,
+      "grad_norm": 0.4583685100078583,
+      "learning_rate": 1.4843382816322331e-05,
+      "loss": 0.47898387908935547,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3725140234574197,
+      "grad_norm": 0.4775019586086273,
+      "learning_rate": 1.4836002519850183e-05,
+      "loss": 0.47897419333457947,
+      "step": 1461
+    },
+    {
+      "epoch": 0.37276899541050484,
+      "grad_norm": 0.46514609456062317,
+      "learning_rate": 1.482861878357414e-05,
+      "loss": 0.4846462607383728,
+      "step": 1462
+    },
+    {
+      "epoch": 0.37302396736359,
+      "grad_norm": 0.4577699303627014,
+      "learning_rate": 1.4821231612746185e-05,
+      "loss": 0.4773539900779724,
+      "step": 1463
+    },
+    {
+      "epoch": 0.37327893931667516,
+      "grad_norm": 0.47564896941185,
+      "learning_rate": 1.4813841012620746e-05,
+      "loss": 0.48534271121025085,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3735339112697603,
+      "grad_norm": 0.4468478858470917,
+      "learning_rate": 1.4806446988454683e-05,
+      "loss": 0.4761880040168762,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3737888832228455,
+      "grad_norm": 0.4458772838115692,
+      "learning_rate": 1.4799049545507305e-05,
+      "loss": 0.47753477096557617,
+      "step": 1466
+    },
+    {
+      "epoch": 0.37404385517593064,
+      "grad_norm": 0.4572882354259491,
+      "learning_rate": 1.479164868904034e-05,
+      "loss": 0.4821872115135193,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3742988271290158,
+      "grad_norm": 0.4762170910835266,
+      "learning_rate": 1.478424442431795e-05,
+      "loss": 0.47601696848869324,
+      "step": 1468
+    },
+    {
+      "epoch": 0.37455379908210096,
+      "grad_norm": 0.4635324776172638,
+      "learning_rate": 1.477683675660672e-05,
+      "loss": 0.4912150502204895,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3748087710351861,
+      "grad_norm": 0.43993592262268066,
+      "learning_rate": 1.4769425691175652e-05,
+      "loss": 0.47345566749572754,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3750637429882713,
+      "grad_norm": 0.4620898962020874,
+      "learning_rate": 1.4762011233296174e-05,
+      "loss": 0.4752131402492523,
+      "step": 1471
+    },
+    {
+      "epoch": 0.37531871494135643,
+      "grad_norm": 0.46631693840026855,
+      "learning_rate": 1.4754593388242117e-05,
+      "loss": 0.49605101346969604,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3755736868944416,
+      "grad_norm": 0.43644899129867554,
+      "learning_rate": 1.4747172161289732e-05,
+      "loss": 0.4809049367904663,
+      "step": 1473
+    },
+    {
+      "epoch": 0.37582865884752675,
+      "grad_norm": 0.4631332457065582,
+      "learning_rate": 1.473974755771766e-05,
+      "loss": 0.4931192398071289,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3760836308006119,
+      "grad_norm": 0.4428750276565552,
+      "learning_rate": 1.473231958280696e-05,
+      "loss": 0.48310884833335876,
+      "step": 1475
+    },
+    {
+      "epoch": 0.37633860275369707,
+      "grad_norm": 0.4648493230342865,
+      "learning_rate": 1.472488824184108e-05,
+      "loss": 0.47687384486198425,
+      "step": 1476
+    },
+    {
+      "epoch": 0.37659357470678223,
+      "grad_norm": 0.47747692465782166,
+      "learning_rate": 1.4717453540105862e-05,
+      "loss": 0.49597570300102234,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3768485466598674,
+      "grad_norm": 0.45823588967323303,
+      "learning_rate": 1.4710015482889544e-05,
+      "loss": 0.4796806573867798,
+      "step": 1478
+    },
+    {
+      "epoch": 0.37710351861295255,
+      "grad_norm": 0.4600730538368225,
+      "learning_rate": 1.4702574075482746e-05,
+      "loss": 0.48539218306541443,
+      "step": 1479
+    },
+    {
+      "epoch": 0.37735849056603776,
+      "grad_norm": 0.4480803906917572,
+      "learning_rate": 1.469512932317847e-05,
+      "loss": 0.488616406917572,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3776134625191229,
+      "grad_norm": 0.48798975348472595,
+      "learning_rate": 1.4687681231272104e-05,
+      "loss": 0.4985082149505615,
+      "step": 1481
+    },
+    {
+      "epoch": 0.3778684344722081,
+      "grad_norm": 0.45104506611824036,
+      "learning_rate": 1.4680229805061399e-05,
+      "loss": 0.4792388379573822,
+      "step": 1482
+    },
+    {
+      "epoch": 0.37812340642529324,
+      "grad_norm": 0.44524678587913513,
+      "learning_rate": 1.4672775049846498e-05,
+      "loss": 0.47451266646385193,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3783783783783784,
+      "grad_norm": 0.46526607871055603,
+      "learning_rate": 1.4665316970929888e-05,
+      "loss": 0.48592525720596313,
+      "step": 1484
+    },
+    {
+      "epoch": 0.37863335033146356,
+      "grad_norm": 0.4819578230381012,
+      "learning_rate": 1.4657855573616437e-05,
+      "loss": 0.4838479459285736,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3788883222845487,
+      "grad_norm": 0.44666457176208496,
+      "learning_rate": 1.4650390863213366e-05,
+      "loss": 0.49645447731018066,
+      "step": 1486
+    },
+    {
+      "epoch": 0.3791432942376339,
+      "grad_norm": 0.45488491654396057,
+      "learning_rate": 1.4642922845030258e-05,
+      "loss": 0.4942684769630432,
+      "step": 1487
+    },
+    {
+      "epoch": 0.37939826619071904,
+      "grad_norm": 0.47846585512161255,
+      "learning_rate": 1.4635451524379043e-05,
+      "loss": 0.48699718713760376,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3796532381438042,
+      "grad_norm": 0.46078893542289734,
+      "learning_rate": 1.4627976906573999e-05,
+      "loss": 0.48280471563339233,
+      "step": 1489
+    },
+    {
+      "epoch": 0.37990821009688935,
+      "grad_norm": 0.4552493989467621,
+      "learning_rate": 1.462049899693176e-05,
+      "loss": 0.48949164152145386,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3801631820499745,
+      "grad_norm": 0.4367724061012268,
+      "learning_rate": 1.4613017800771286e-05,
+      "loss": 0.49780935049057007,
+      "step": 1491
+    },
+    {
+      "epoch": 0.38041815400305967,
+      "grad_norm": 0.4415394961833954,
+      "learning_rate": 1.4605533323413887e-05,
+      "loss": 0.5012547969818115,
+      "step": 1492
+    },
+    {
+      "epoch": 0.38067312595614483,
+      "grad_norm": 0.45368072390556335,
+      "learning_rate": 1.4598045570183203e-05,
+      "loss": 0.47883474826812744,
+      "step": 1493
+    },
+    {
+      "epoch": 0.38092809790923,
+      "grad_norm": 0.4611485004425049,
+      "learning_rate": 1.4590554546405203e-05,
+      "loss": 0.4865252673625946,
+      "step": 1494
+    },
+    {
+      "epoch": 0.38118306986231515,
+      "grad_norm": 0.47810161113739014,
+      "learning_rate": 1.4583060257408183e-05,
+      "loss": 0.47490066289901733,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3814380418154003,
+      "grad_norm": 0.45693275332450867,
+      "learning_rate": 1.4575562708522761e-05,
+      "loss": 0.4761950671672821,
+      "step": 1496
+    },
+    {
+      "epoch": 0.38169301376848547,
+      "grad_norm": 0.4426383972167969,
+      "learning_rate": 1.4568061905081874e-05,
+      "loss": 0.4906729459762573,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3819479857215706,
+      "grad_norm": 0.45328155159950256,
+      "learning_rate": 1.4560557852420773e-05,
+      "loss": 0.4902084767818451,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3822029576746558,
+      "grad_norm": 0.4416186511516571,
+      "learning_rate": 1.4553050555877026e-05,
+      "loss": 0.5001434087753296,
+      "step": 1499
+    },
+    {
+      "epoch": 0.38245792962774094,
+      "grad_norm": 0.44836804270744324,
+      "learning_rate": 1.45455400207905e-05,
+      "loss": 0.49423956871032715,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3827129015808261,
+      "grad_norm": 0.4632645547389984,
+      "learning_rate": 1.4538026252503367e-05,
+      "loss": 0.47067809104919434,
+      "step": 1501
+    },
+    {
+      "epoch": 0.38296787353391126,
+      "grad_norm": 0.4599841833114624,
+      "learning_rate": 1.4530509256360104e-05,
+      "loss": 0.4771389365196228,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3832228454869964,
+      "grad_norm": 0.4621497094631195,
+      "learning_rate": 1.4522989037707482e-05,
+      "loss": 0.49022993445396423,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3834778174400816,
+      "grad_norm": 0.4501090347766876,
+      "learning_rate": 1.4515465601894561e-05,
+      "loss": 0.5076841115951538,
+      "step": 1504
+    },
+    {
+      "epoch": 0.38373278939316674,
+      "grad_norm": 0.4534502923488617,
+      "learning_rate": 1.4507938954272693e-05,
+      "loss": 0.4904021918773651,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3839877613462519,
+      "grad_norm": 0.46439942717552185,
+      "learning_rate": 1.450040910019551e-05,
+      "loss": 0.4915975332260132,
+      "step": 1506
+    },
+    {
+      "epoch": 0.38424273329933706,
+      "grad_norm": 0.47011348605155945,
+      "learning_rate": 1.449287604501893e-05,
+      "loss": 0.4765165448188782,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3844977052524222,
+      "grad_norm": 0.45521828532218933,
+      "learning_rate": 1.4485339794101144e-05,
+      "loss": 0.49013710021972656,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3847526772055074,
+      "grad_norm": 0.4601080119609833,
+      "learning_rate": 1.4477800352802623e-05,
+      "loss": 0.47809481620788574,
+      "step": 1509
+    },
+    {
+      "epoch": 0.38500764915859254,
+      "grad_norm": 0.462210088968277,
+      "learning_rate": 1.447025772648609e-05,
+      "loss": 0.48346203565597534,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3852626211116777,
+      "grad_norm": 0.4716843366622925,
+      "learning_rate": 1.4462711920516557e-05,
+      "loss": 0.4802989363670349,
+      "step": 1511
+    },
+    {
+      "epoch": 0.38551759306476285,
+      "grad_norm": 0.6230674386024475,
+      "learning_rate": 1.4455162940261284e-05,
+      "loss": 0.47767776250839233,
+      "step": 1512
+    },
+    {
+      "epoch": 0.385772565017848,
+      "grad_norm": 0.44257497787475586,
+      "learning_rate": 1.4447610791089785e-05,
+      "loss": 0.47903773188591003,
+      "step": 1513
+    },
+    {
+      "epoch": 0.38602753697093317,
+      "grad_norm": 0.4705200493335724,
+      "learning_rate": 1.4440055478373839e-05,
+      "loss": 0.4785939157009125,
+      "step": 1514
+    },
+    {
+      "epoch": 0.38628250892401833,
+      "grad_norm": 0.4540731906890869,
+      "learning_rate": 1.443249700748747e-05,
+      "loss": 0.47946688532829285,
+      "step": 1515
+    },
+    {
+      "epoch": 0.38653748087710355,
+      "grad_norm": 0.4631092846393585,
+      "learning_rate": 1.4424935383806946e-05,
+      "loss": 0.48664236068725586,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3867924528301887,
+      "grad_norm": 0.4636557102203369,
+      "learning_rate": 1.441737061271078e-05,
+      "loss": 0.4810870289802551,
+      "step": 1517
+    },
+    {
+      "epoch": 0.38704742478327386,
+      "grad_norm": 0.4769573509693146,
+      "learning_rate": 1.4409802699579725e-05,
+      "loss": 0.4827399253845215,
+      "step": 1518
+    },
+    {
+      "epoch": 0.387302396736359,
+      "grad_norm": 0.46991199254989624,
+      "learning_rate": 1.4402231649796769e-05,
+      "loss": 0.5055210590362549,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3875573686894442,
+      "grad_norm": 0.47453632950782776,
+      "learning_rate": 1.4394657468747129e-05,
+      "loss": 0.48708269000053406,
+      "step": 1520
+    },
+    {
+      "epoch": 0.38781234064252934,
+      "grad_norm": 0.44457998871803284,
+      "learning_rate": 1.438708016181825e-05,
+      "loss": 0.4817838668823242,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3880673125956145,
+      "grad_norm": 0.45942017436027527,
+      "learning_rate": 1.4379499734399797e-05,
+      "loss": 0.4798167943954468,
+      "step": 1522
+    },
+    {
+      "epoch": 0.38832228454869966,
+      "grad_norm": 0.46707138419151306,
+      "learning_rate": 1.4371916191883663e-05,
+      "loss": 0.46871715784072876,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3885772565017848,
+      "grad_norm": 0.4826716184616089,
+      "learning_rate": 1.4364329539663949e-05,
+      "loss": 0.4892393946647644,
+      "step": 1524
+    },
+    {
+      "epoch": 0.38883222845487,
+      "grad_norm": 0.4482455849647522,
+      "learning_rate": 1.4356739783136972e-05,
+      "loss": 0.46966874599456787,
+      "step": 1525
+    },
+    {
+      "epoch": 0.38908720040795514,
+      "grad_norm": 0.4489781856536865,
+      "learning_rate": 1.4349146927701252e-05,
+      "loss": 0.47867918014526367,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3893421723610403,
+      "grad_norm": 0.47932168841362,
+      "learning_rate": 1.4341550978757521e-05,
+      "loss": 0.4873637557029724,
+      "step": 1527
+    },
+    {
+      "epoch": 0.38959714431412545,
+      "grad_norm": 0.44425198435783386,
+      "learning_rate": 1.4333951941708706e-05,
+      "loss": 0.4994814395904541,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3898521162672106,
+      "grad_norm": 0.45439931750297546,
+      "learning_rate": 1.4326349821959933e-05,
+      "loss": 0.47517848014831543,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3901070882202958,
+      "grad_norm": 0.4534931182861328,
+      "learning_rate": 1.4318744624918514e-05,
+      "loss": 0.48021823167800903,
+      "step": 1530
+    },
+    {
+      "epoch": 0.39036206017338093,
+      "grad_norm": 0.583962082862854,
+      "learning_rate": 1.4311136355993966e-05,
+      "loss": 0.49174126982688904,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3906170321264661,
+      "grad_norm": 0.46215638518333435,
+      "learning_rate": 1.430352502059797e-05,
+      "loss": 0.4798983037471771,
+      "step": 1532
+    },
+    {
+      "epoch": 0.39087200407955125,
+      "grad_norm": 0.5120609998703003,
+      "learning_rate": 1.4295910624144402e-05,
+      "loss": 0.4900760054588318,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3911269760326364,
+      "grad_norm": 0.45972445607185364,
+      "learning_rate": 1.4288293172049313e-05,
+      "loss": 0.47815102338790894,
+      "step": 1534
+    },
+    {
+      "epoch": 0.39138194798572157,
+      "grad_norm": 0.4667312800884247,
+      "learning_rate": 1.4280672669730927e-05,
+      "loss": 0.496964693069458,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3916369199388067,
+      "grad_norm": 0.47719672322273254,
+      "learning_rate": 1.4273049122609633e-05,
+      "loss": 0.48423081636428833,
+      "step": 1536
+    },
+    {
+      "epoch": 0.3918918918918919,
+      "grad_norm": 0.48460742831230164,
+      "learning_rate": 1.4265422536107994e-05,
+      "loss": 0.48138755559921265,
+      "step": 1537
+    },
+    {
+      "epoch": 0.39214686384497704,
+      "grad_norm": 0.5079189538955688,
+      "learning_rate": 1.4257792915650728e-05,
+      "loss": 0.4881475567817688,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3924018357980622,
+      "grad_norm": 0.4590676724910736,
+      "learning_rate": 1.4250160266664712e-05,
+      "loss": 0.49422332644462585,
+      "step": 1539
+    },
+    {
+      "epoch": 0.39265680775114736,
+      "grad_norm": 0.4414307177066803,
+      "learning_rate": 1.4242524594578984e-05,
+      "loss": 0.48674750328063965,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3929117797042325,
+      "grad_norm": 0.4570107161998749,
+      "learning_rate": 1.4234885904824725e-05,
+      "loss": 0.48515844345092773,
+      "step": 1541
+    },
+    {
+      "epoch": 0.3931667516573177,
+      "grad_norm": 0.47521156072616577,
+      "learning_rate": 1.4227244202835258e-05,
+      "loss": 0.4922315776348114,
+      "step": 1542
+    },
+    {
+      "epoch": 0.39342172361040284,
+      "grad_norm": 0.4279789328575134,
+      "learning_rate": 1.4219599494046062e-05,
+      "loss": 0.47620928287506104,
+      "step": 1543
+    },
+    {
+      "epoch": 0.393676695563488,
+      "grad_norm": 0.47957128286361694,
+      "learning_rate": 1.4211951783894749e-05,
+      "loss": 0.47388386726379395,
+      "step": 1544
+    },
+    {
+      "epoch": 0.39393166751657316,
+      "grad_norm": 0.45006492733955383,
+      "learning_rate": 1.4204301077821059e-05,
+      "loss": 0.46949005126953125,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3941866394696583,
+      "grad_norm": 0.5271633863449097,
+      "learning_rate": 1.419664738126687e-05,
+      "loss": 0.49612054228782654,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3944416114227435,
+      "grad_norm": 0.45255032181739807,
+      "learning_rate": 1.4188990699676186e-05,
+      "loss": 0.481834352016449,
+      "step": 1547
+    },
+    {
+      "epoch": 0.39469658337582864,
+      "grad_norm": 0.46046289801597595,
+      "learning_rate": 1.4181331038495136e-05,
+      "loss": 0.4820711016654968,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3949515553289138,
+      "grad_norm": 0.4877573549747467,
+      "learning_rate": 1.4173668403171962e-05,
+      "loss": 0.48516762256622314,
+      "step": 1549
+    },
+    {
+      "epoch": 0.39520652728199895,
+      "grad_norm": 0.49049946665763855,
+      "learning_rate": 1.4166002799157032e-05,
+      "loss": 0.48174232244491577,
+      "step": 1550
+    },
+    {
+      "epoch": 0.39546149923508417,
+      "grad_norm": 0.46354466676712036,
+      "learning_rate": 1.4158334231902812e-05,
+      "loss": 0.47682464122772217,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3957164711881693,
+      "grad_norm": 0.49040186405181885,
+      "learning_rate": 1.415066270686389e-05,
+      "loss": 0.48382410407066345,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3959714431412545,
+      "grad_norm": 0.43498021364212036,
+      "learning_rate": 1.4142988229496945e-05,
+      "loss": 0.49451208114624023,
+      "step": 1553
+    },
+    {
+      "epoch": 0.39622641509433965,
+      "grad_norm": 0.5171186923980713,
+      "learning_rate": 1.4135310805260766e-05,
+      "loss": 0.48215898871421814,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3964813870474248,
+      "grad_norm": 0.4854249954223633,
+      "learning_rate": 1.4127630439616231e-05,
+      "loss": 0.48182177543640137,
+      "step": 1555
+    },
+    {
+      "epoch": 0.39673635900050996,
+      "grad_norm": 0.4441761374473572,
+      "learning_rate": 1.4119947138026313e-05,
+      "loss": 0.47478580474853516,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3969913309535951,
+      "grad_norm": 0.4775533974170685,
+      "learning_rate": 1.4112260905956079e-05,
+      "loss": 0.47552570700645447,
+      "step": 1557
+    },
+    {
+      "epoch": 0.3972463029066803,
+      "grad_norm": 0.5103408694267273,
+      "learning_rate": 1.4104571748872667e-05,
+      "loss": 0.502683162689209,
+      "step": 1558
+    },
+    {
+      "epoch": 0.39750127485976544,
+      "grad_norm": 0.4726421535015106,
+      "learning_rate": 1.4096879672245305e-05,
+      "loss": 0.47500142455101013,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3977562468128506,
+      "grad_norm": 0.4710962176322937,
+      "learning_rate": 1.4089184681545303e-05,
+      "loss": 0.49768611788749695,
+      "step": 1560
+    },
+    {
+      "epoch": 0.39801121876593576,
+      "grad_norm": 0.509074330329895,
+      "learning_rate": 1.4081486782246027e-05,
+      "loss": 0.4758099913597107,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3982661907190209,
+      "grad_norm": 0.4465167224407196,
+      "learning_rate": 1.4073785979822929e-05,
+      "loss": 0.49819037318229675,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3985211626721061,
+      "grad_norm": 0.4520632326602936,
+      "learning_rate": 1.4066082279753515e-05,
+      "loss": 0.4983295798301697,
+      "step": 1563
+    },
+    {
+      "epoch": 0.39877613462519124,
+      "grad_norm": 0.46213945746421814,
+      "learning_rate": 1.4058375687517355e-05,
+      "loss": 0.46227526664733887,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3990311065782764,
+      "grad_norm": 0.4812057614326477,
+      "learning_rate": 1.405066620859608e-05,
+      "loss": 0.48829787969589233,
+      "step": 1565
+    },
+    {
+      "epoch": 0.39928607853136155,
+      "grad_norm": 0.7864221930503845,
+      "learning_rate": 1.4042953848473368e-05,
+      "loss": 0.4807780683040619,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3995410504844467,
+      "grad_norm": 0.46031373739242554,
+      "learning_rate": 1.4035238612634951e-05,
+      "loss": 0.49220967292785645,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3997960224375319,
+      "grad_norm": 0.478533536195755,
+      "learning_rate": 1.4027520506568605e-05,
+      "loss": 0.4768020808696747,
+      "step": 1568
+    },
+    {
+      "epoch": 0.40005099439061703,
+      "grad_norm": 0.45310717821121216,
+      "learning_rate": 1.401979953576415e-05,
+      "loss": 0.4713541865348816,
+      "step": 1569
+    },
+    {
+      "epoch": 0.4003059663437022,
+      "grad_norm": 0.4644032120704651,
+      "learning_rate": 1.4012075705713434e-05,
+      "loss": 0.4951431155204773,
+      "step": 1570
+    },
+    {
+      "epoch": 0.40056093829678735,
+      "grad_norm": 0.4290669858455658,
+      "learning_rate": 1.4004349021910353e-05,
+      "loss": 0.47089534997940063,
+      "step": 1571
+    },
+    {
+      "epoch": 0.4008159102498725,
+      "grad_norm": 0.46978506445884705,
+      "learning_rate": 1.3996619489850822e-05,
+      "loss": 0.4897878170013428,
+      "step": 1572
+    },
+    {
+      "epoch": 0.40107088220295767,
+      "grad_norm": 0.4528856873512268,
+      "learning_rate": 1.398888711503279e-05,
+      "loss": 0.48548150062561035,
+      "step": 1573
+    },
+    {
+      "epoch": 0.4013258541560428,
+      "grad_norm": 0.4511377811431885,
+      "learning_rate": 1.3981151902956217e-05,
+      "loss": 0.4705016016960144,
+      "step": 1574
+    },
+    {
+      "epoch": 0.401580826109128,
+      "grad_norm": 0.43902209401130676,
+      "learning_rate": 1.3973413859123092e-05,
+      "loss": 0.48379310965538025,
+      "step": 1575
+    },
+    {
+      "epoch": 0.40183579806221315,
+      "grad_norm": 0.45791682600975037,
+      "learning_rate": 1.3965672989037417e-05,
+      "loss": 0.48469385504722595,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4020907700152983,
+      "grad_norm": 0.4580422341823578,
+      "learning_rate": 1.3957929298205195e-05,
+      "loss": 0.48687800765037537,
+      "step": 1577
+    },
+    {
+      "epoch": 0.40234574196838346,
+      "grad_norm": 0.47267118096351624,
+      "learning_rate": 1.3950182792134446e-05,
+      "loss": 0.4877247214317322,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4026007139214686,
+      "grad_norm": 0.4548337757587433,
+      "learning_rate": 1.3942433476335184e-05,
+      "loss": 0.48541364073753357,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4028556858745538,
+      "grad_norm": 0.43207213282585144,
+      "learning_rate": 1.393468135631943e-05,
+      "loss": 0.4798010587692261,
+      "step": 1580
+    },
+    {
+      "epoch": 0.40311065782763894,
+      "grad_norm": 0.4571256637573242,
+      "learning_rate": 1.3926926437601197e-05,
+      "loss": 0.4921513497829437,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4033656297807241,
+      "grad_norm": 0.48809128999710083,
+      "learning_rate": 1.3919168725696478e-05,
+      "loss": 0.48802515864372253,
+      "step": 1582
+    },
+    {
+      "epoch": 0.40362060173380926,
+      "grad_norm": 0.45787227153778076,
+      "learning_rate": 1.391140822612327e-05,
+      "loss": 0.49700865149497986,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4038755736868944,
+      "grad_norm": 0.44923320412635803,
+      "learning_rate": 1.3903644944401545e-05,
+      "loss": 0.4944605529308319,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4041305456399796,
+      "grad_norm": 0.45642563700675964,
+      "learning_rate": 1.3895878886053246e-05,
+      "loss": 0.4735734462738037,
+      "step": 1585
+    },
+    {
+      "epoch": 0.40438551759306474,
+      "grad_norm": 0.4567015767097473,
+      "learning_rate": 1.3888110056602308e-05,
+      "loss": 0.4728502035140991,
+      "step": 1586
+    },
+    {
+      "epoch": 0.40464048954614995,
+      "grad_norm": 0.4450572729110718,
+      "learning_rate": 1.388033846157462e-05,
+      "loss": 0.49534106254577637,
+      "step": 1587
+    },
+    {
+      "epoch": 0.4048954614992351,
+      "grad_norm": 0.44211024045944214,
+      "learning_rate": 1.3872564106498047e-05,
+      "loss": 0.47716912627220154,
+      "step": 1588
+    },
+    {
+      "epoch": 0.40515043345232027,
+      "grad_norm": 0.4511987268924713,
+      "learning_rate": 1.3864786996902423e-05,
+      "loss": 0.4985756278038025,
+      "step": 1589
+    },
+    {
+      "epoch": 0.40540540540540543,
+      "grad_norm": 0.4538654088973999,
+      "learning_rate": 1.3857007138319527e-05,
+      "loss": 0.4770917296409607,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4056603773584906,
+      "grad_norm": 0.43509507179260254,
+      "learning_rate": 1.3849224536283097e-05,
+      "loss": 0.48660558462142944,
+      "step": 1591
+    },
+    {
+      "epoch": 0.40591534931157575,
+      "grad_norm": 0.4530210494995117,
+      "learning_rate": 1.3841439196328836e-05,
+      "loss": 0.49161213636398315,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4061703212646609,
+      "grad_norm": 0.47141772508621216,
+      "learning_rate": 1.3833651123994379e-05,
+      "loss": 0.4738878905773163,
+      "step": 1593
+    },
+    {
+      "epoch": 0.40642529321774606,
+      "grad_norm": 0.46208295226097107,
+      "learning_rate": 1.3825860324819308e-05,
+      "loss": 0.48128652572631836,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4066802651708312,
+      "grad_norm": 0.44921064376831055,
+      "learning_rate": 1.3818066804345151e-05,
+      "loss": 0.49665015935897827,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4069352371239164,
+      "grad_norm": 0.47046971321105957,
+      "learning_rate": 1.3810270568115366e-05,
+      "loss": 0.46492284536361694,
+      "step": 1596
+    },
+    {
+      "epoch": 0.40719020907700154,
+      "grad_norm": 0.4615946412086487,
+      "learning_rate": 1.3802471621675337e-05,
+      "loss": 0.48955702781677246,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4074451810300867,
+      "grad_norm": 0.46036478877067566,
+      "learning_rate": 1.3794669970572395e-05,
+      "loss": 0.48940813541412354,
+      "step": 1598
+    },
+    {
+      "epoch": 0.40770015298317186,
+      "grad_norm": 0.43279993534088135,
+      "learning_rate": 1.378686562035577e-05,
+      "loss": 0.480970561504364,
+      "step": 1599
+    },
+    {
+      "epoch": 0.407955124936257,
+      "grad_norm": 0.45972615480422974,
+      "learning_rate": 1.377905857657663e-05,
+      "loss": 0.4687748849391937,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4082100968893422,
+      "grad_norm": 0.4757559895515442,
+      "learning_rate": 1.3771248844788055e-05,
+      "loss": 0.4859820306301117,
+      "step": 1601
+    },
+    {
+      "epoch": 0.40846506884242734,
+      "grad_norm": 0.4991218149662018,
+      "learning_rate": 1.3763436430545034e-05,
+      "loss": 0.47516581416130066,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4087200407955125,
+      "grad_norm": 0.45847487449645996,
+      "learning_rate": 1.3755621339404459e-05,
+      "loss": 0.4729556441307068,
+      "step": 1603
+    },
+    {
+      "epoch": 0.40897501274859765,
+      "grad_norm": 0.4737587571144104,
+      "learning_rate": 1.374780357692514e-05,
+      "loss": 0.48022347688674927,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4092299847016828,
+      "grad_norm": 0.4732692539691925,
+      "learning_rate": 1.3739983148667776e-05,
+      "loss": 0.4782615900039673,
+      "step": 1605
+    },
+    {
+      "epoch": 0.409484956654768,
+      "grad_norm": 0.4965715706348419,
+      "learning_rate": 1.3732160060194968e-05,
+      "loss": 0.49145591259002686,
+      "step": 1606
+    },
+    {
+      "epoch": 0.40973992860785313,
+      "grad_norm": 0.4683915376663208,
+      "learning_rate": 1.3724334317071198e-05,
+      "loss": 0.4661458730697632,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4099949005609383,
+      "grad_norm": 0.47402238845825195,
+      "learning_rate": 1.3716505924862854e-05,
+      "loss": 0.4833154082298279,
+      "step": 1608
+    },
+    {
+      "epoch": 0.41024987251402345,
+      "grad_norm": 0.44059866666793823,
+      "learning_rate": 1.3708674889138196e-05,
+      "loss": 0.4715484082698822,
+      "step": 1609
+    },
+    {
+      "epoch": 0.4105048444671086,
+      "grad_norm": 0.4649759531021118,
+      "learning_rate": 1.3700841215467367e-05,
+      "loss": 0.4846433103084564,
+      "step": 1610
+    },
+    {
+      "epoch": 0.41075981642019377,
+      "grad_norm": 0.45463913679122925,
+      "learning_rate": 1.3693004909422386e-05,
+      "loss": 0.4861733615398407,
+      "step": 1611
+    },
+    {
+      "epoch": 0.4110147883732789,
+      "grad_norm": 0.49009451270103455,
+      "learning_rate": 1.3685165976577146e-05,
+      "loss": 0.47389039397239685,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4112697603263641,
+      "grad_norm": 0.4492635130882263,
+      "learning_rate": 1.3677324422507408e-05,
+      "loss": 0.47067707777023315,
+      "step": 1613
+    },
+    {
+      "epoch": 0.41152473227944925,
+      "grad_norm": 0.47212710976600647,
+      "learning_rate": 1.3669480252790797e-05,
+      "loss": 0.47150108218193054,
+      "step": 1614
+    },
+    {
+      "epoch": 0.4117797042325344,
+      "grad_norm": 0.4461907148361206,
+      "learning_rate": 1.3661633473006793e-05,
+      "loss": 0.48159798979759216,
+      "step": 1615
+    },
+    {
+      "epoch": 0.41203467618561956,
+      "grad_norm": 0.4642240107059479,
+      "learning_rate": 1.3653784088736746e-05,
+      "loss": 0.48190754652023315,
+      "step": 1616
+    },
+    {
+      "epoch": 0.4122896481387047,
+      "grad_norm": 0.4726830720901489,
+      "learning_rate": 1.3645932105563846e-05,
+      "loss": 0.49468010663986206,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4125446200917899,
+      "grad_norm": 0.4275886118412018,
+      "learning_rate": 1.363807752907314e-05,
+      "loss": 0.4874088168144226,
+      "step": 1618
+    },
+    {
+      "epoch": 0.41279959204487504,
+      "grad_norm": 0.6437248587608337,
+      "learning_rate": 1.363022036485151e-05,
+      "loss": 0.4754694402217865,
+      "step": 1619
+    },
+    {
+      "epoch": 0.4130545639979602,
+      "grad_norm": 0.46456652879714966,
+      "learning_rate": 1.3622360618487687e-05,
+      "loss": 0.48658570647239685,
+      "step": 1620
+    },
+    {
+      "epoch": 0.41330953595104536,
+      "grad_norm": 0.44917410612106323,
+      "learning_rate": 1.3614498295572238e-05,
+      "loss": 0.4856434762477875,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4135645079041305,
+      "grad_norm": 0.4353286921977997,
+      "learning_rate": 1.3606633401697557e-05,
+      "loss": 0.49146074056625366,
+      "step": 1622
+    },
+    {
+      "epoch": 0.41381947985721573,
+      "grad_norm": 0.4655471742153168,
+      "learning_rate": 1.3598765942457872e-05,
+      "loss": 0.47687792778015137,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4140744518103009,
+      "grad_norm": 0.4673038125038147,
+      "learning_rate": 1.3590895923449234e-05,
+      "loss": 0.45503830909729004,
+      "step": 1624
+    },
+    {
+      "epoch": 0.41432942376338605,
+      "grad_norm": 0.4729268252849579,
+      "learning_rate": 1.3583023350269513e-05,
+      "loss": 0.47268760204315186,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4145843957164712,
+      "grad_norm": 0.49650296568870544,
+      "learning_rate": 1.3575148228518403e-05,
+      "loss": 0.46334347128868103,
+      "step": 1626
+    },
+    {
+      "epoch": 0.41483936766955637,
+      "grad_norm": 0.4445992112159729,
+      "learning_rate": 1.3567270563797399e-05,
+      "loss": 0.47191429138183594,
+      "step": 1627
+    },
+    {
+      "epoch": 0.41509433962264153,
+      "grad_norm": 0.4386600852012634,
+      "learning_rate": 1.3559390361709815e-05,
+      "loss": 0.49940285086631775,
+      "step": 1628
+    },
+    {
+      "epoch": 0.4153493115757267,
+      "grad_norm": 0.4637932777404785,
+      "learning_rate": 1.3551507627860763e-05,
+      "loss": 0.48494452238082886,
+      "step": 1629
+    },
+    {
+      "epoch": 0.41560428352881185,
+      "grad_norm": 0.46620863676071167,
+      "learning_rate": 1.3543622367857163e-05,
+      "loss": 0.4677116274833679,
+      "step": 1630
+    },
+    {
+      "epoch": 0.415859255481897,
+      "grad_norm": 0.4731232523918152,
+      "learning_rate": 1.353573458730772e-05,
+      "loss": 0.4884800910949707,
+      "step": 1631
+    },
+    {
+      "epoch": 0.41611422743498216,
+      "grad_norm": 0.5170514583587646,
+      "learning_rate": 1.3527844291822949e-05,
+      "loss": 0.48462021350860596,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4163691993880673,
+      "grad_norm": 0.5338066816329956,
+      "learning_rate": 1.3519951487015139e-05,
+      "loss": 0.4877411723136902,
+      "step": 1633
+    },
+    {
+      "epoch": 0.4166241713411525,
+      "grad_norm": 0.6070787310600281,
+      "learning_rate": 1.351205617849837e-05,
+      "loss": 0.49505704641342163,
+      "step": 1634
+    },
+    {
+      "epoch": 0.41687914329423764,
+      "grad_norm": 0.44800975918769836,
+      "learning_rate": 1.3504158371888498e-05,
+      "loss": 0.47991299629211426,
+      "step": 1635
+    },
+    {
+      "epoch": 0.4171341152473228,
+      "grad_norm": 0.43597695231437683,
+      "learning_rate": 1.3496258072803166e-05,
+      "loss": 0.486041396856308,
+      "step": 1636
+    },
+    {
+      "epoch": 0.41738908720040796,
+      "grad_norm": 0.4952779710292816,
+      "learning_rate": 1.3488355286861783e-05,
+      "loss": 0.48333197832107544,
+      "step": 1637
+    },
+    {
+      "epoch": 0.4176440591534931,
+      "grad_norm": 0.4361210763454437,
+      "learning_rate": 1.3480450019685524e-05,
+      "loss": 0.46700620651245117,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4178990311065783,
+      "grad_norm": 0.4386427700519562,
+      "learning_rate": 1.3472542276897334e-05,
+      "loss": 0.4709675908088684,
+      "step": 1639
+    },
+    {
+      "epoch": 0.41815400305966344,
+      "grad_norm": 0.47871455550193787,
+      "learning_rate": 1.3464632064121918e-05,
+      "loss": 0.48721468448638916,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4184089750127486,
+      "grad_norm": 0.4435887634754181,
+      "learning_rate": 1.3456719386985736e-05,
+      "loss": 0.47918903827667236,
+      "step": 1641
+    },
+    {
+      "epoch": 0.41866394696583376,
+      "grad_norm": 0.4284169673919678,
+      "learning_rate": 1.3448804251117005e-05,
+      "loss": 0.47185349464416504,
+      "step": 1642
+    },
+    {
+      "epoch": 0.4189189189189189,
+      "grad_norm": 0.4468022882938385,
+      "learning_rate": 1.3440886662145684e-05,
+      "loss": 0.4968520998954773,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4191738908720041,
+      "grad_norm": 0.4575154483318329,
+      "learning_rate": 1.3432966625703486e-05,
+      "loss": 0.48007091879844666,
+      "step": 1644
+    },
+    {
+      "epoch": 0.41942886282508923,
+      "grad_norm": 0.43185529112815857,
+      "learning_rate": 1.3425044147423853e-05,
+      "loss": 0.47791963815689087,
+      "step": 1645
+    },
+    {
+      "epoch": 0.4196838347781744,
+      "grad_norm": 0.4488309919834137,
+      "learning_rate": 1.3417119232941978e-05,
+      "loss": 0.4668620228767395,
+      "step": 1646
+    },
+    {
+      "epoch": 0.41993880673125955,
+      "grad_norm": 0.4566347301006317,
+      "learning_rate": 1.340919188789477e-05,
+      "loss": 0.49318307638168335,
+      "step": 1647
+    },
+    {
+      "epoch": 0.4201937786843447,
+      "grad_norm": 0.4377533793449402,
+      "learning_rate": 1.3401262117920885e-05,
+      "loss": 0.49428534507751465,
+      "step": 1648
+    },
+    {
+      "epoch": 0.42044875063742987,
+      "grad_norm": 0.4332873225212097,
+      "learning_rate": 1.3393329928660694e-05,
+      "loss": 0.48560136556625366,
+      "step": 1649
+    },
+    {
+      "epoch": 0.42070372259051503,
+      "grad_norm": 0.44632354378700256,
+      "learning_rate": 1.3385395325756285e-05,
+      "loss": 0.4737345576286316,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4209586945436002,
+      "grad_norm": 0.44238245487213135,
+      "learning_rate": 1.3377458314851471e-05,
+      "loss": 0.47939157485961914,
+      "step": 1651
+    },
+    {
+      "epoch": 0.42121366649668535,
+      "grad_norm": 0.4579528272151947,
+      "learning_rate": 1.3369518901591771e-05,
+      "loss": 0.46565157175064087,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4214686384497705,
+      "grad_norm": 0.4802386164665222,
+      "learning_rate": 1.3361577091624428e-05,
+      "loss": 0.4831387996673584,
+      "step": 1653
+    },
+    {
+      "epoch": 0.42172361040285566,
+      "grad_norm": 0.46119028329849243,
+      "learning_rate": 1.3353632890598362e-05,
+      "loss": 0.472695529460907,
+      "step": 1654
+    },
+    {
+      "epoch": 0.4219785823559408,
+      "grad_norm": 0.45306500792503357,
+      "learning_rate": 1.3345686304164219e-05,
+      "loss": 0.4780885577201843,
+      "step": 1655
+    },
+    {
+      "epoch": 0.422233554309026,
+      "grad_norm": 0.4374094009399414,
+      "learning_rate": 1.333773733797433e-05,
+      "loss": 0.4786366820335388,
+      "step": 1656
+    },
+    {
+      "epoch": 0.42248852626211114,
+      "grad_norm": 0.5166929364204407,
+      "learning_rate": 1.3329785997682722e-05,
+      "loss": 0.48381268978118896,
+      "step": 1657
+    },
+    {
+      "epoch": 0.42274349821519636,
+      "grad_norm": 0.4499305784702301,
+      "learning_rate": 1.3321832288945111e-05,
+      "loss": 0.49744629859924316,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4229984701682815,
+      "grad_norm": 0.43170177936553955,
+      "learning_rate": 1.3313876217418893e-05,
+      "loss": 0.4850466251373291,
+      "step": 1659
+    },
+    {
+      "epoch": 0.4232534421213667,
+      "grad_norm": 0.46282199025154114,
+      "learning_rate": 1.3305917788763153e-05,
+      "loss": 0.47375819087028503,
+      "step": 1660
+    },
+    {
+      "epoch": 0.42350841407445183,
+      "grad_norm": 0.487903356552124,
+      "learning_rate": 1.3297957008638643e-05,
+      "loss": 0.46697551012039185,
+      "step": 1661
+    },
+    {
+      "epoch": 0.423763386027537,
+      "grad_norm": 0.4308168590068817,
+      "learning_rate": 1.3289993882707797e-05,
+      "loss": 0.4807865619659424,
+      "step": 1662
+    },
+    {
+      "epoch": 0.42401835798062215,
+      "grad_norm": 0.44673821330070496,
+      "learning_rate": 1.3282028416634706e-05,
+      "loss": 0.48854202032089233,
+      "step": 1663
+    },
+    {
+      "epoch": 0.4242733299337073,
+      "grad_norm": 0.44982901215553284,
+      "learning_rate": 1.3274060616085138e-05,
+      "loss": 0.4782748818397522,
+      "step": 1664
+    },
+    {
+      "epoch": 0.42452830188679247,
+      "grad_norm": 0.47188979387283325,
+      "learning_rate": 1.3266090486726517e-05,
+      "loss": 0.47001177072525024,
+      "step": 1665
+    },
+    {
+      "epoch": 0.42478327383987763,
+      "grad_norm": 0.4699896574020386,
+      "learning_rate": 1.3258118034227919e-05,
+      "loss": 0.46810004115104675,
+      "step": 1666
+    },
+    {
+      "epoch": 0.4250382457929628,
+      "grad_norm": 0.4989445209503174,
+      "learning_rate": 1.3250143264260073e-05,
+      "loss": 0.47544538974761963,
+      "step": 1667
+    },
+    {
+      "epoch": 0.42529321774604795,
+      "grad_norm": 0.45112210512161255,
+      "learning_rate": 1.3242166182495365e-05,
+      "loss": 0.4739255905151367,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4255481896991331,
+      "grad_norm": 0.44104111194610596,
+      "learning_rate": 1.3234186794607818e-05,
+      "loss": 0.4756285548210144,
+      "step": 1669
+    },
+    {
+      "epoch": 0.42580316165221826,
+      "grad_norm": 0.47478458285331726,
+      "learning_rate": 1.3226205106273096e-05,
+      "loss": 0.47291168570518494,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4260581336053034,
+      "grad_norm": 0.468546062707901,
+      "learning_rate": 1.32182211231685e-05,
+      "loss": 0.4908111095428467,
+      "step": 1671
+    },
+    {
+      "epoch": 0.4263131055583886,
+      "grad_norm": 0.4369455873966217,
+      "learning_rate": 1.3210234850972966e-05,
+      "loss": 0.484855979681015,
+      "step": 1672
+    },
+    {
+      "epoch": 0.42656807751147374,
+      "grad_norm": 0.4560834467411041,
+      "learning_rate": 1.320224629536705e-05,
+      "loss": 0.476165771484375,
+      "step": 1673
+    },
+    {
+      "epoch": 0.4268230494645589,
+      "grad_norm": 0.4688289165496826,
+      "learning_rate": 1.319425546203294e-05,
+      "loss": 0.4720986783504486,
+      "step": 1674
+    },
+    {
+      "epoch": 0.42707802141764406,
+      "grad_norm": 0.47327515482902527,
+      "learning_rate": 1.3186262356654446e-05,
+      "loss": 0.4842362403869629,
+      "step": 1675
+    },
+    {
+      "epoch": 0.4273329933707292,
+      "grad_norm": 0.45015227794647217,
+      "learning_rate": 1.3178266984916988e-05,
+      "loss": 0.46754440665245056,
+      "step": 1676
+    },
+    {
+      "epoch": 0.4275879653238144,
+      "grad_norm": 0.43983298540115356,
+      "learning_rate": 1.3170269352507596e-05,
+      "loss": 0.4953022003173828,
+      "step": 1677
+    },
+    {
+      "epoch": 0.42784293727689954,
+      "grad_norm": 0.44356074929237366,
+      "learning_rate": 1.3162269465114918e-05,
+      "loss": 0.4818527102470398,
+      "step": 1678
+    },
+    {
+      "epoch": 0.4280979092299847,
+      "grad_norm": 0.5327989459037781,
+      "learning_rate": 1.3154267328429195e-05,
+      "loss": 0.48241209983825684,
+      "step": 1679
+    },
+    {
+      "epoch": 0.42835288118306986,
+      "grad_norm": 0.46997660398483276,
+      "learning_rate": 1.3146262948142275e-05,
+      "loss": 0.4853352904319763,
+      "step": 1680
+    },
+    {
+      "epoch": 0.428607853136155,
+      "grad_norm": 0.4460082948207855,
+      "learning_rate": 1.31382563299476e-05,
+      "loss": 0.48609769344329834,
+      "step": 1681
+    },
+    {
+      "epoch": 0.4288628250892402,
+      "grad_norm": 0.4682973027229309,
+      "learning_rate": 1.3130247479540204e-05,
+      "loss": 0.47891318798065186,
+      "step": 1682
+    },
+    {
+      "epoch": 0.42911779704232533,
+      "grad_norm": 0.47408202290534973,
+      "learning_rate": 1.3122236402616704e-05,
+      "loss": 0.47502049803733826,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4293727689954105,
+      "grad_norm": 0.4689554274082184,
+      "learning_rate": 1.3114223104875308e-05,
+      "loss": 0.47387951612472534,
+      "step": 1684
+    },
+    {
+      "epoch": 0.42962774094849565,
+      "grad_norm": 0.49953290820121765,
+      "learning_rate": 1.3106207592015805e-05,
+      "loss": 0.47314685583114624,
+      "step": 1685
+    },
+    {
+      "epoch": 0.4298827129015808,
+      "grad_norm": 0.4629705846309662,
+      "learning_rate": 1.309818986973955e-05,
+      "loss": 0.48018312454223633,
+      "step": 1686
+    },
+    {
+      "epoch": 0.43013768485466597,
+      "grad_norm": 0.4409901201725006,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.47948044538497925,
+      "step": 1687
+    },
+    {
+      "epoch": 0.43039265680775113,
+      "grad_norm": 0.4578230082988739,
+      "learning_rate": 1.3082147819750082e-05,
+      "loss": 0.48771414160728455,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4306476287608363,
+      "grad_norm": 0.4722670018672943,
+      "learning_rate": 1.307412350344743e-05,
+      "loss": 0.47841060161590576,
+      "step": 1689
+    },
+    {
+      "epoch": 0.43090260071392145,
+      "grad_norm": 0.47902294993400574,
+      "learning_rate": 1.3066097000549144e-05,
+      "loss": 0.4815452992916107,
+      "step": 1690
+    },
+    {
+      "epoch": 0.4311575726670066,
+      "grad_norm": 0.45889562368392944,
+      "learning_rate": 1.3058068316764398e-05,
+      "loss": 0.4722667634487152,
+      "step": 1691
+    },
+    {
+      "epoch": 0.43141254462009176,
+      "grad_norm": 0.43147915601730347,
+      "learning_rate": 1.3050037457803923e-05,
+      "loss": 0.4807421565055847,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4316675165731769,
+      "grad_norm": 0.46262240409851074,
+      "learning_rate": 1.3042004429379996e-05,
+      "loss": 0.47799575328826904,
+      "step": 1693
+    },
+    {
+      "epoch": 0.43192248852626214,
+      "grad_norm": 0.4444118142127991,
+      "learning_rate": 1.3033969237206435e-05,
+      "loss": 0.48812946677207947,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4321774604793473,
+      "grad_norm": 0.45554131269454956,
+      "learning_rate": 1.3025931886998596e-05,
+      "loss": 0.4786272943019867,
+      "step": 1695
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 0.4367639422416687,
+      "learning_rate": 1.301789238447337e-05,
+      "loss": 0.4648439288139343,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4326874043855176,
+      "grad_norm": 0.4478955864906311,
+      "learning_rate": 1.300985073534919e-05,
+      "loss": 0.4997819662094116,
+      "step": 1697
+    },
+    {
+      "epoch": 0.4329423763386028,
+      "grad_norm": 0.44023770093917847,
+      "learning_rate": 1.3001806945345996e-05,
+      "loss": 0.47148871421813965,
+      "step": 1698
+    },
+    {
+      "epoch": 0.43319734829168793,
+      "grad_norm": 0.43281856179237366,
+      "learning_rate": 1.2993761020185267e-05,
+      "loss": 0.4947243630886078,
+      "step": 1699
+    },
+    {
+      "epoch": 0.4334523202447731,
+      "grad_norm": 0.4664633274078369,
+      "learning_rate": 1.2985712965589994e-05,
+      "loss": 0.4891164302825928,
+      "step": 1700
+    },
+    {
+      "epoch": 0.43370729219785825,
+      "grad_norm": 0.44780081510543823,
+      "learning_rate": 1.2977662787284689e-05,
+      "loss": 0.4701750874519348,
+      "step": 1701
+    },
+    {
+      "epoch": 0.4339622641509434,
+      "grad_norm": 0.45525118708610535,
+      "learning_rate": 1.296961049099536e-05,
+      "loss": 0.47872576117515564,
+      "step": 1702
+    },
+    {
+      "epoch": 0.43421723610402857,
+      "grad_norm": 0.4881942570209503,
+      "learning_rate": 1.2961556082449536e-05,
+      "loss": 0.45850667357444763,
+      "step": 1703
+    },
+    {
+      "epoch": 0.43447220805711373,
+      "grad_norm": 0.4705426096916199,
+      "learning_rate": 1.2953499567376244e-05,
+      "loss": 0.4757862091064453,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4347271800101989,
+      "grad_norm": 0.4563402533531189,
+      "learning_rate": 1.2945440951506007e-05,
+      "loss": 0.483933687210083,
+      "step": 1705
+    },
+    {
+      "epoch": 0.43498215196328405,
+      "grad_norm": 0.4657118320465088,
+      "learning_rate": 1.2937380240570845e-05,
+      "loss": 0.46637582778930664,
+      "step": 1706
+    },
+    {
+      "epoch": 0.4352371239163692,
+      "grad_norm": 0.48174551129341125,
+      "learning_rate": 1.292931744030427e-05,
+      "loss": 0.48372596502304077,
+      "step": 1707
+    },
+    {
+      "epoch": 0.43549209586945437,
+      "grad_norm": 0.46783116459846497,
+      "learning_rate": 1.2921252556441269e-05,
+      "loss": 0.47157931327819824,
+      "step": 1708
+    },
+    {
+      "epoch": 0.4357470678225395,
+      "grad_norm": 0.453797310590744,
+      "learning_rate": 1.2913185594718325e-05,
+      "loss": 0.4759908616542816,
+      "step": 1709
+    },
+    {
+      "epoch": 0.4360020397756247,
+      "grad_norm": 0.4898701012134552,
+      "learning_rate": 1.2905116560873396e-05,
+      "loss": 0.49279874563217163,
+      "step": 1710
+    },
+    {
+      "epoch": 0.43625701172870984,
+      "grad_norm": 0.44953569769859314,
+      "learning_rate": 1.2897045460645905e-05,
+      "loss": 0.47002968192100525,
+      "step": 1711
+    },
+    {
+      "epoch": 0.436511983681795,
+      "grad_norm": 0.453167587518692,
+      "learning_rate": 1.2888972299776754e-05,
+      "loss": 0.4644286334514618,
+      "step": 1712
+    },
+    {
+      "epoch": 0.43676695563488016,
+      "grad_norm": 0.43823450803756714,
+      "learning_rate": 1.288089708400831e-05,
+      "loss": 0.47578439116477966,
+      "step": 1713
+    },
+    {
+      "epoch": 0.4370219275879653,
+      "grad_norm": 0.4516087770462036,
+      "learning_rate": 1.2872819819084397e-05,
+      "loss": 0.46475034952163696,
+      "step": 1714
+    },
+    {
+      "epoch": 0.4372768995410505,
+      "grad_norm": 0.4563446044921875,
+      "learning_rate": 1.2864740510750298e-05,
+      "loss": 0.46562060713768005,
+      "step": 1715
+    },
+    {
+      "epoch": 0.43753187149413564,
+      "grad_norm": 0.4217168688774109,
+      "learning_rate": 1.285665916475275e-05,
+      "loss": 0.4905102252960205,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4377868434472208,
+      "grad_norm": 0.4617106318473816,
+      "learning_rate": 1.2848575786839943e-05,
+      "loss": 0.4758516550064087,
+      "step": 1717
+    },
+    {
+      "epoch": 0.43804181540030596,
+      "grad_norm": 0.4274753928184509,
+      "learning_rate": 1.2840490382761507e-05,
+      "loss": 0.47411856055259705,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4382967873533911,
+      "grad_norm": 0.4601130783557892,
+      "learning_rate": 1.2832402958268512e-05,
+      "loss": 0.4765852093696594,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4385517593064763,
+      "grad_norm": 0.4270618259906769,
+      "learning_rate": 1.2824313519113473e-05,
+      "loss": 0.5004135370254517,
+      "step": 1720
+    },
+    {
+      "epoch": 0.43880673125956143,
+      "grad_norm": 0.4411335289478302,
+      "learning_rate": 1.2816222071050333e-05,
+      "loss": 0.49290788173675537,
+      "step": 1721
+    },
+    {
+      "epoch": 0.4390617032126466,
+      "grad_norm": 0.4671216607093811,
+      "learning_rate": 1.280812861983446e-05,
+      "loss": 0.46440428495407104,
+      "step": 1722
+    },
+    {
+      "epoch": 0.43931667516573175,
+      "grad_norm": 0.4586823582649231,
+      "learning_rate": 1.2800033171222662e-05,
+      "loss": 0.469355046749115,
+      "step": 1723
+    },
+    {
+      "epoch": 0.4395716471188169,
+      "grad_norm": 0.47849181294441223,
+      "learning_rate": 1.279193573097314e-05,
+      "loss": 0.47547468543052673,
+      "step": 1724
+    },
+    {
+      "epoch": 0.43982661907190207,
+      "grad_norm": 0.446884423494339,
+      "learning_rate": 1.2783836304845541e-05,
+      "loss": 0.48563331365585327,
+      "step": 1725
+    },
+    {
+      "epoch": 0.44008159102498723,
+      "grad_norm": 0.4576139748096466,
+      "learning_rate": 1.277573489860091e-05,
+      "loss": 0.4876425266265869,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4403365629780724,
+      "grad_norm": 0.4311068654060364,
+      "learning_rate": 1.2767631518001697e-05,
+      "loss": 0.4762778878211975,
+      "step": 1727
+    },
+    {
+      "epoch": 0.44059153493115755,
+      "grad_norm": 0.44358745217323303,
+      "learning_rate": 1.2759526168811767e-05,
+      "loss": 0.48140454292297363,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4408465068842427,
+      "grad_norm": 0.45184627175331116,
+      "learning_rate": 1.2751418856796378e-05,
+      "loss": 0.46511802077293396,
+      "step": 1729
+    },
+    {
+      "epoch": 0.4411014788373279,
+      "grad_norm": 0.44608306884765625,
+      "learning_rate": 1.2743309587722186e-05,
+      "loss": 0.47363704442977905,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4413564507904131,
+      "grad_norm": 0.44456541538238525,
+      "learning_rate": 1.2735198367357238e-05,
+      "loss": 0.4707756042480469,
+      "step": 1731
+    },
+    {
+      "epoch": 0.44161142274349824,
+      "grad_norm": 0.4834231436252594,
+      "learning_rate": 1.2727085201470972e-05,
+      "loss": 0.47929319739341736,
+      "step": 1732
+    },
+    {
+      "epoch": 0.4418663946965834,
+      "grad_norm": 0.49860382080078125,
+      "learning_rate": 1.2718970095834208e-05,
+      "loss": 0.47534316778182983,
+      "step": 1733
+    },
+    {
+      "epoch": 0.44212136664966856,
+      "grad_norm": 1.011425256729126,
+      "learning_rate": 1.2710853056219147e-05,
+      "loss": 0.49217113852500916,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4423763386027537,
+      "grad_norm": 0.4697829782962799,
+      "learning_rate": 1.2702734088399363e-05,
+      "loss": 0.4657892882823944,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4426313105558389,
+      "grad_norm": 0.48535552620887756,
+      "learning_rate": 1.2694613198149796e-05,
+      "loss": 0.47705650329589844,
+      "step": 1736
+    },
+    {
+      "epoch": 0.44288628250892403,
+      "grad_norm": 0.4486839473247528,
+      "learning_rate": 1.2686490391246774e-05,
+      "loss": 0.46023058891296387,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4431412544620092,
+      "grad_norm": 0.4410373270511627,
+      "learning_rate": 1.2678365673467963e-05,
+      "loss": 0.4703472852706909,
+      "step": 1738
+    },
+    {
+      "epoch": 0.44339622641509435,
+      "grad_norm": 0.46294504404067993,
+      "learning_rate": 1.2670239050592403e-05,
+      "loss": 0.47119733691215515,
+      "step": 1739
+    },
+    {
+      "epoch": 0.4436511983681795,
+      "grad_norm": 0.46631738543510437,
+      "learning_rate": 1.2662110528400487e-05,
+      "loss": 0.47250574827194214,
+      "step": 1740
+    },
+    {
+      "epoch": 0.44390617032126467,
+      "grad_norm": 0.47729727625846863,
+      "learning_rate": 1.2653980112673955e-05,
+      "loss": 0.4722181260585785,
+      "step": 1741
+    },
+    {
+      "epoch": 0.44416114227434983,
+      "grad_norm": 0.42720356583595276,
+      "learning_rate": 1.2645847809195901e-05,
+      "loss": 0.47384095191955566,
+      "step": 1742
+    },
+    {
+      "epoch": 0.444416114227435,
+      "grad_norm": 0.4484679102897644,
+      "learning_rate": 1.2637713623750756e-05,
+      "loss": 0.4848884046077728,
+      "step": 1743
+    },
+    {
+      "epoch": 0.44467108618052015,
+      "grad_norm": 0.456752210855484,
+      "learning_rate": 1.2629577562124287e-05,
+      "loss": 0.4707961976528168,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4449260581336053,
+      "grad_norm": 0.42030447721481323,
+      "learning_rate": 1.2621439630103602e-05,
+      "loss": 0.47010743618011475,
+      "step": 1745
+    },
+    {
+      "epoch": 0.44518103008669047,
+      "grad_norm": 0.4455772042274475,
+      "learning_rate": 1.261329983347714e-05,
+      "loss": 0.46104955673217773,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4454360020397756,
+      "grad_norm": 0.44260451197624207,
+      "learning_rate": 1.2605158178034656e-05,
+      "loss": 0.4911135137081146,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4456909739928608,
+      "grad_norm": 0.4729379713535309,
+      "learning_rate": 1.259701466956724e-05,
+      "loss": 0.47768276929855347,
+      "step": 1748
+    },
+    {
+      "epoch": 0.44594594594594594,
+      "grad_norm": 0.4576628506183624,
+      "learning_rate": 1.2588869313867292e-05,
+      "loss": 0.47084134817123413,
+      "step": 1749
+    },
+    {
+      "epoch": 0.4462009178990311,
+      "grad_norm": 0.42538702487945557,
+      "learning_rate": 1.258072211672853e-05,
+      "loss": 0.47333893179893494,
+      "step": 1750
+    },
+    {
+      "epoch": 0.44645588985211626,
+      "grad_norm": 0.4477687180042267,
+      "learning_rate": 1.2572573083945978e-05,
+      "loss": 0.4792226254940033,
+      "step": 1751
+    },
+    {
+      "epoch": 0.4467108618052014,
+      "grad_norm": 0.4507501423358917,
+      "learning_rate": 1.2564422221315968e-05,
+      "loss": 0.4824841618537903,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4469658337582866,
+      "grad_norm": 0.44207245111465454,
+      "learning_rate": 1.2556269534636135e-05,
+      "loss": 0.475760281085968,
+      "step": 1753
+    },
+    {
+      "epoch": 0.44722080571137174,
+      "grad_norm": 0.4917367994785309,
+      "learning_rate": 1.254811502970541e-05,
+      "loss": 0.46215686202049255,
+      "step": 1754
+    },
+    {
+      "epoch": 0.4474757776644569,
+      "grad_norm": 0.444948673248291,
+      "learning_rate": 1.2539958712324014e-05,
+      "loss": 0.4736212491989136,
+      "step": 1755
+    },
+    {
+      "epoch": 0.44773074961754206,
+      "grad_norm": 0.45400217175483704,
+      "learning_rate": 1.2531800588293461e-05,
+      "loss": 0.4696999490261078,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4479857215706272,
+      "grad_norm": 0.4441147744655609,
+      "learning_rate": 1.252364066341655e-05,
+      "loss": 0.47615838050842285,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4482406935237124,
+      "grad_norm": 0.44922590255737305,
+      "learning_rate": 1.2515478943497361e-05,
+      "loss": 0.4806191325187683,
+      "step": 1758
+    },
+    {
+      "epoch": 0.44849566547679753,
+      "grad_norm": 0.44325754046440125,
+      "learning_rate": 1.2507315434341249e-05,
+      "loss": 0.46858978271484375,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4487506374298827,
+      "grad_norm": 0.45428788661956787,
+      "learning_rate": 1.2499150141754842e-05,
+      "loss": 0.4710807800292969,
+      "step": 1760
+    },
+    {
+      "epoch": 0.44900560938296785,
+      "grad_norm": 0.4337548017501831,
+      "learning_rate": 1.2490983071546037e-05,
+      "loss": 0.4801885485649109,
+      "step": 1761
+    },
+    {
+      "epoch": 0.449260581336053,
+      "grad_norm": 0.4419419765472412,
+      "learning_rate": 1.2482814229523996e-05,
+      "loss": 0.4899044930934906,
+      "step": 1762
+    },
+    {
+      "epoch": 0.44951555328913817,
+      "grad_norm": 0.45944783091545105,
+      "learning_rate": 1.2474643621499143e-05,
+      "loss": 0.4820578098297119,
+      "step": 1763
+    },
+    {
+      "epoch": 0.44977052524222333,
+      "grad_norm": 0.4322916269302368,
+      "learning_rate": 1.2466471253283152e-05,
+      "loss": 0.48402005434036255,
+      "step": 1764
+    },
+    {
+      "epoch": 0.45002549719530854,
+      "grad_norm": 0.45586666464805603,
+      "learning_rate": 1.2458297130688956e-05,
+      "loss": 0.4691661298274994,
+      "step": 1765
+    },
+    {
+      "epoch": 0.4502804691483937,
+      "grad_norm": 0.45265716314315796,
+      "learning_rate": 1.2450121259530734e-05,
+      "loss": 0.46303677558898926,
+      "step": 1766
+    },
+    {
+      "epoch": 0.45053544110147886,
+      "grad_norm": 0.466658353805542,
+      "learning_rate": 1.2441943645623904e-05,
+      "loss": 0.4779145121574402,
+      "step": 1767
+    },
+    {
+      "epoch": 0.450790413054564,
+      "grad_norm": 0.4724508821964264,
+      "learning_rate": 1.2433764294785129e-05,
+      "loss": 0.47466349601745605,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4510453850076492,
+      "grad_norm": 0.45049089193344116,
+      "learning_rate": 1.2425583212832302e-05,
+      "loss": 0.4861510992050171,
+      "step": 1769
+    },
+    {
+      "epoch": 0.45130035696073434,
+      "grad_norm": 0.4370516836643219,
+      "learning_rate": 1.2417400405584558e-05,
+      "loss": 0.4628305435180664,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4515553289138195,
+      "grad_norm": 0.45360061526298523,
+      "learning_rate": 1.2409215878862249e-05,
+      "loss": 0.4932473599910736,
+      "step": 1771
+    },
+    {
+      "epoch": 0.45181030086690466,
+      "grad_norm": 0.45101219415664673,
+      "learning_rate": 1.2401029638486952e-05,
+      "loss": 0.47766444087028503,
+      "step": 1772
+    },
+    {
+      "epoch": 0.4520652728199898,
+      "grad_norm": 0.44970381259918213,
+      "learning_rate": 1.239284169028146e-05,
+      "loss": 0.46386319398880005,
+      "step": 1773
+    },
+    {
+      "epoch": 0.452320244773075,
+      "grad_norm": 0.44679099321365356,
+      "learning_rate": 1.2384652040069797e-05,
+      "loss": 0.4587230086326599,
+      "step": 1774
+    },
+    {
+      "epoch": 0.45257521672616013,
+      "grad_norm": 0.44836875796318054,
+      "learning_rate": 1.2376460693677175e-05,
+      "loss": 0.4732503294944763,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4528301886792453,
+      "grad_norm": 0.5044194459915161,
+      "learning_rate": 1.2368267656930027e-05,
+      "loss": 0.4807708263397217,
+      "step": 1776
+    },
+    {
+      "epoch": 0.45308516063233045,
+      "grad_norm": 0.47821566462516785,
+      "learning_rate": 1.2360072935655982e-05,
+      "loss": 0.4741743505001068,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4533401325854156,
+      "grad_norm": 0.45061835646629333,
+      "learning_rate": 1.235187653568387e-05,
+      "loss": 0.46865314245224,
+      "step": 1778
+    },
+    {
+      "epoch": 0.45359510453850077,
+      "grad_norm": 0.44544747471809387,
+      "learning_rate": 1.2343678462843719e-05,
+      "loss": 0.4797539710998535,
+      "step": 1779
+    },
+    {
+      "epoch": 0.45385007649158593,
+      "grad_norm": 0.4373033344745636,
+      "learning_rate": 1.2335478722966738e-05,
+      "loss": 0.4692363440990448,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4541050484446711,
+      "grad_norm": 0.42586174607276917,
+      "learning_rate": 1.2327277321885326e-05,
+      "loss": 0.4837741255760193,
+      "step": 1781
+    },
+    {
+      "epoch": 0.45436002039775625,
+      "grad_norm": 0.4503430724143982,
+      "learning_rate": 1.2319074265433064e-05,
+      "loss": 0.473422110080719,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4546149923508414,
+      "grad_norm": 0.43340277671813965,
+      "learning_rate": 1.2310869559444714e-05,
+      "loss": 0.4896630048751831,
+      "step": 1783
+    },
+    {
+      "epoch": 0.45486996430392657,
+      "grad_norm": 0.5077560544013977,
+      "learning_rate": 1.2302663209756205e-05,
+      "loss": 0.4597128629684448,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4551249362570117,
+      "grad_norm": 0.4537369906902313,
+      "learning_rate": 1.2294455222204632e-05,
+      "loss": 0.46173548698425293,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4553799082100969,
+      "grad_norm": 0.46171754598617554,
+      "learning_rate": 1.228624560262827e-05,
+      "loss": 0.4772665500640869,
+      "step": 1786
+    },
+    {
+      "epoch": 0.45563488016318204,
+      "grad_norm": 0.4706333577632904,
+      "learning_rate": 1.2278034356866544e-05,
+      "loss": 0.4807857275009155,
+      "step": 1787
+    },
+    {
+      "epoch": 0.4558898521162672,
+      "grad_norm": 0.4414377212524414,
+      "learning_rate": 1.2269821490760032e-05,
+      "loss": 0.4800296425819397,
+      "step": 1788
+    },
+    {
+      "epoch": 0.45614482406935236,
+      "grad_norm": 0.43886110186576843,
+      "learning_rate": 1.226160701015047e-05,
+      "loss": 0.46951231360435486,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4563997960224375,
+      "grad_norm": 0.438742458820343,
+      "learning_rate": 1.2253390920880752e-05,
+      "loss": 0.4638747572898865,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4566547679755227,
+      "grad_norm": 0.4479273855686188,
+      "learning_rate": 1.2245173228794895e-05,
+      "loss": 0.4727884531021118,
+      "step": 1791
+    },
+    {
+      "epoch": 0.45690973992860784,
+      "grad_norm": 0.4348614811897278,
+      "learning_rate": 1.223695393973807e-05,
+      "loss": 0.4840061664581299,
+      "step": 1792
+    },
+    {
+      "epoch": 0.457164711881693,
+      "grad_norm": 0.463783860206604,
+      "learning_rate": 1.2228733059556586e-05,
+      "loss": 0.4745769500732422,
+      "step": 1793
+    },
+    {
+      "epoch": 0.45741968383477816,
+      "grad_norm": 0.46853530406951904,
+      "learning_rate": 1.2220510594097872e-05,
+      "loss": 0.49177151918411255,
+      "step": 1794
+    },
+    {
+      "epoch": 0.4576746557878633,
+      "grad_norm": 0.4315093457698822,
+      "learning_rate": 1.22122865492105e-05,
+      "loss": 0.47219523787498474,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4579296277409485,
+      "grad_norm": 0.44561806321144104,
+      "learning_rate": 1.220406093074415e-05,
+      "loss": 0.4777514934539795,
+      "step": 1796
+    },
+    {
+      "epoch": 0.45818459969403363,
+      "grad_norm": 0.4777575135231018,
+      "learning_rate": 1.219583374454963e-05,
+      "loss": 0.47510725259780884,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4584395716471188,
+      "grad_norm": 0.4492422044277191,
+      "learning_rate": 1.2187604996478866e-05,
+      "loss": 0.47745585441589355,
+      "step": 1798
+    },
+    {
+      "epoch": 0.45869454360020395,
+      "grad_norm": 0.4303736090660095,
+      "learning_rate": 1.2179374692384885e-05,
+      "loss": 0.462444543838501,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4589495155532891,
+      "grad_norm": 0.458484411239624,
+      "learning_rate": 1.217114283812183e-05,
+      "loss": 0.49000486731529236,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4592044875063743,
+      "grad_norm": 0.45598477125167847,
+      "learning_rate": 1.2162909439544937e-05,
+      "loss": 0.47185465693473816,
+      "step": 1801
+    },
+    {
+      "epoch": 0.4594594594594595,
+      "grad_norm": 0.4487041234970093,
+      "learning_rate": 1.2154674502510556e-05,
+      "loss": 0.4776914715766907,
+      "step": 1802
+    },
+    {
+      "epoch": 0.45971443141254464,
+      "grad_norm": 0.4446077048778534,
+      "learning_rate": 1.2146438032876113e-05,
+      "loss": 0.47020599246025085,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4599694033656298,
+      "grad_norm": 0.4974459111690521,
+      "learning_rate": 1.2138200036500135e-05,
+      "loss": 0.48040771484375,
+      "step": 1804
+    },
+    {
+      "epoch": 0.46022437531871496,
+      "grad_norm": 0.47240355610847473,
+      "learning_rate": 1.2129960519242235e-05,
+      "loss": 0.47204288840293884,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4604793472718001,
+      "grad_norm": 0.46213290095329285,
+      "learning_rate": 1.2121719486963105e-05,
+      "loss": 0.4674822688102722,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4607343192248853,
+      "grad_norm": 0.4440208971500397,
+      "learning_rate": 1.2113476945524513e-05,
+      "loss": 0.4881192147731781,
+      "step": 1807
+    },
+    {
+      "epoch": 0.46098929117797044,
+      "grad_norm": 0.4705323278903961,
+      "learning_rate": 1.2105232900789303e-05,
+      "loss": 0.4879416823387146,
+      "step": 1808
+    },
+    {
+      "epoch": 0.4612442631310556,
+      "grad_norm": 0.4392581582069397,
+      "learning_rate": 1.209698735862139e-05,
+      "loss": 0.48007023334503174,
+      "step": 1809
+    },
+    {
+      "epoch": 0.46149923508414076,
+      "grad_norm": 0.44329649209976196,
+      "learning_rate": 1.2088740324885748e-05,
+      "loss": 0.463228315114975,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4617542070372259,
+      "grad_norm": 0.43376657366752625,
+      "learning_rate": 1.208049180544842e-05,
+      "loss": 0.4700953960418701,
+      "step": 1811
+    },
+    {
+      "epoch": 0.4620091789903111,
+      "grad_norm": 0.4467110335826874,
+      "learning_rate": 1.2072241806176501e-05,
+      "loss": 0.4704214036464691,
+      "step": 1812
+    },
+    {
+      "epoch": 0.46226415094339623,
+      "grad_norm": 0.42056575417518616,
+      "learning_rate": 1.2063990332938134e-05,
+      "loss": 0.4868830442428589,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4625191228964814,
+      "grad_norm": 0.4433046579360962,
+      "learning_rate": 1.2055737391602522e-05,
+      "loss": 0.46415600180625916,
+      "step": 1814
+    },
+    {
+      "epoch": 0.46277409484956655,
+      "grad_norm": 0.4381503164768219,
+      "learning_rate": 1.2047482988039905e-05,
+      "loss": 0.4766816794872284,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4630290668026517,
+      "grad_norm": 0.45899951457977295,
+      "learning_rate": 1.2039227128121564e-05,
+      "loss": 0.4798325300216675,
+      "step": 1816
+    },
+    {
+      "epoch": 0.46328403875573687,
+      "grad_norm": 0.4464300870895386,
+      "learning_rate": 1.203096981771981e-05,
+      "loss": 0.47894152998924255,
+      "step": 1817
+    },
+    {
+      "epoch": 0.46353901070882203,
+      "grad_norm": 0.4569408893585205,
+      "learning_rate": 1.2022711062707997e-05,
+      "loss": 0.4581347107887268,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4637939826619072,
+      "grad_norm": 0.45413821935653687,
+      "learning_rate": 1.2014450868960504e-05,
+      "loss": 0.45865410566329956,
+      "step": 1819
+    },
+    {
+      "epoch": 0.46404895461499235,
+      "grad_norm": 0.4557012915611267,
+      "learning_rate": 1.2006189242352723e-05,
+      "loss": 0.48170050978660583,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4643039265680775,
+      "grad_norm": 0.4703729748725891,
+      "learning_rate": 1.1997926188761077e-05,
+      "loss": 0.4796580672264099,
+      "step": 1821
+    },
+    {
+      "epoch": 0.46455889852116267,
+      "grad_norm": 0.44692087173461914,
+      "learning_rate": 1.1989661714063e-05,
+      "loss": 0.47521767020225525,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4648138704742478,
+      "grad_norm": 0.43485578894615173,
+      "learning_rate": 1.1981395824136932e-05,
+      "loss": 0.47263312339782715,
+      "step": 1823
+    },
+    {
+      "epoch": 0.465068842427333,
+      "grad_norm": 0.44065967202186584,
+      "learning_rate": 1.197312852486233e-05,
+      "loss": 0.4700472354888916,
+      "step": 1824
+    },
+    {
+      "epoch": 0.46532381438041814,
+      "grad_norm": 0.44195377826690674,
+      "learning_rate": 1.1964859822119646e-05,
+      "loss": 0.46952304244041443,
+      "step": 1825
+    },
+    {
+      "epoch": 0.4655787863335033,
+      "grad_norm": 0.46734338998794556,
+      "learning_rate": 1.1956589721790329e-05,
+      "loss": 0.47229647636413574,
+      "step": 1826
+    },
+    {
+      "epoch": 0.46583375828658846,
+      "grad_norm": 0.4728843569755554,
+      "learning_rate": 1.1948318229756828e-05,
+      "loss": 0.47129344940185547,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4660887302396736,
+      "grad_norm": 0.48862841725349426,
+      "learning_rate": 1.1940045351902578e-05,
+      "loss": 0.4701687693595886,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4663437021927588,
+      "grad_norm": 0.46959683299064636,
+      "learning_rate": 1.1931771094111997e-05,
+      "loss": 0.4784017503261566,
+      "step": 1829
+    },
+    {
+      "epoch": 0.46659867414584394,
+      "grad_norm": 0.48078179359436035,
+      "learning_rate": 1.1923495462270494e-05,
+      "loss": 0.4602683186531067,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4668536460989291,
+      "grad_norm": 0.43931084871292114,
+      "learning_rate": 1.1915218462264446e-05,
+      "loss": 0.4679564833641052,
+      "step": 1831
+    },
+    {
+      "epoch": 0.46710861805201426,
+      "grad_norm": 0.48635151982307434,
+      "learning_rate": 1.190694009998121e-05,
+      "loss": 0.4717889726161957,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4673635900050994,
+      "grad_norm": 0.5232540369033813,
+      "learning_rate": 1.1898660381309104e-05,
+      "loss": 0.4683421552181244,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4676185619581846,
+      "grad_norm": 0.44738560914993286,
+      "learning_rate": 1.1890379312137414e-05,
+      "loss": 0.4686152935028076,
+      "step": 1834
+    },
+    {
+      "epoch": 0.46787353391126973,
+      "grad_norm": 0.4688865840435028,
+      "learning_rate": 1.1882096898356396e-05,
+      "loss": 0.4768770635128021,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4681285058643549,
+      "grad_norm": 0.43922460079193115,
+      "learning_rate": 1.187381314585725e-05,
+      "loss": 0.4662562310695648,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4683834778174401,
+      "grad_norm": 0.8413701057434082,
+      "learning_rate": 1.1865528060532127e-05,
+      "loss": 0.4704360365867615,
+      "step": 1837
+    },
+    {
+      "epoch": 0.46863844977052527,
+      "grad_norm": 0.46233436465263367,
+      "learning_rate": 1.185724164827414e-05,
+      "loss": 0.4667252004146576,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4688934217236104,
+      "grad_norm": 0.43266645073890686,
+      "learning_rate": 1.1848953914977335e-05,
+      "loss": 0.47639912366867065,
+      "step": 1839
+    },
+    {
+      "epoch": 0.4691483936766956,
+      "grad_norm": 0.4587555527687073,
+      "learning_rate": 1.1840664866536697e-05,
+      "loss": 0.47096312046051025,
+      "step": 1840
+    },
+    {
+      "epoch": 0.46940336562978074,
+      "grad_norm": 0.4797708988189697,
+      "learning_rate": 1.1832374508848154e-05,
+      "loss": 0.4757949709892273,
+      "step": 1841
+    },
+    {
+      "epoch": 0.4696583375828659,
+      "grad_norm": 0.4354809820652008,
+      "learning_rate": 1.1824082847808558e-05,
+      "loss": 0.4758743643760681,
+      "step": 1842
+    },
+    {
+      "epoch": 0.46991330953595106,
+      "grad_norm": 0.46408167481422424,
+      "learning_rate": 1.1815789889315691e-05,
+      "loss": 0.4625156819820404,
+      "step": 1843
+    },
+    {
+      "epoch": 0.4701682814890362,
+      "grad_norm": 0.4551498591899872,
+      "learning_rate": 1.180749563926826e-05,
+      "loss": 0.46774551272392273,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4704232534421214,
+      "grad_norm": 0.45411378145217896,
+      "learning_rate": 1.1799200103565887e-05,
+      "loss": 0.46884870529174805,
+      "step": 1845
+    },
+    {
+      "epoch": 0.47067822539520654,
+      "grad_norm": 0.4664252996444702,
+      "learning_rate": 1.1790903288109106e-05,
+      "loss": 0.4782741963863373,
+      "step": 1846
+    },
+    {
+      "epoch": 0.4709331973482917,
+      "grad_norm": 0.4781484305858612,
+      "learning_rate": 1.1782605198799371e-05,
+      "loss": 0.4723745286464691,
+      "step": 1847
+    },
+    {
+      "epoch": 0.47118816930137686,
+      "grad_norm": 0.4496593475341797,
+      "learning_rate": 1.1774305841539036e-05,
+      "loss": 0.46488043665885925,
+      "step": 1848
+    },
+    {
+      "epoch": 0.471443141254462,
+      "grad_norm": 0.45651766657829285,
+      "learning_rate": 1.1766005222231351e-05,
+      "loss": 0.47134900093078613,
+      "step": 1849
+    },
+    {
+      "epoch": 0.4716981132075472,
+      "grad_norm": 0.47809088230133057,
+      "learning_rate": 1.1757703346780473e-05,
+      "loss": 0.48021984100341797,
+      "step": 1850
+    },
+    {
+      "epoch": 0.47195308516063234,
+      "grad_norm": 0.46262407302856445,
+      "learning_rate": 1.1749400221091455e-05,
+      "loss": 0.4492693841457367,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4722080571137175,
+      "grad_norm": 0.4340875744819641,
+      "learning_rate": 1.1741095851070228e-05,
+      "loss": 0.4752400517463684,
+      "step": 1852
+    },
+    {
+      "epoch": 0.47246302906680265,
+      "grad_norm": 0.42423349618911743,
+      "learning_rate": 1.1732790242623612e-05,
+      "loss": 0.4710129201412201,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4727180010198878,
+      "grad_norm": 0.48380085825920105,
+      "learning_rate": 1.1724483401659313e-05,
+      "loss": 0.47220295667648315,
+      "step": 1854
+    },
+    {
+      "epoch": 0.47297297297297297,
+      "grad_norm": 0.458454966545105,
+      "learning_rate": 1.1716175334085908e-05,
+      "loss": 0.4703748822212219,
+      "step": 1855
+    },
+    {
+      "epoch": 0.47322794492605813,
+      "grad_norm": 0.4433039426803589,
+      "learning_rate": 1.170786604581285e-05,
+      "loss": 0.4689106345176697,
+      "step": 1856
+    },
+    {
+      "epoch": 0.4734829168791433,
+      "grad_norm": 0.43104684352874756,
+      "learning_rate": 1.169955554275046e-05,
+      "loss": 0.4718170166015625,
+      "step": 1857
+    },
+    {
+      "epoch": 0.47373788883222845,
+      "grad_norm": 0.5022832155227661,
+      "learning_rate": 1.169124383080992e-05,
+      "loss": 0.4683248996734619,
+      "step": 1858
+    },
+    {
+      "epoch": 0.4739928607853136,
+      "grad_norm": 0.48713451623916626,
+      "learning_rate": 1.1682930915903275e-05,
+      "loss": 0.4653375744819641,
+      "step": 1859
+    },
+    {
+      "epoch": 0.47424783273839877,
+      "grad_norm": 0.43980905413627625,
+      "learning_rate": 1.1674616803943425e-05,
+      "loss": 0.4635535180568695,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4745028046914839,
+      "grad_norm": 0.44994592666625977,
+      "learning_rate": 1.1666301500844117e-05,
+      "loss": 0.4867127537727356,
+      "step": 1861
+    },
+    {
+      "epoch": 0.4747577766445691,
+      "grad_norm": 0.45325347781181335,
+      "learning_rate": 1.1657985012519951e-05,
+      "loss": 0.4687597453594208,
+      "step": 1862
+    },
+    {
+      "epoch": 0.47501274859765424,
+      "grad_norm": 0.45354774594306946,
+      "learning_rate": 1.1649667344886376e-05,
+      "loss": 0.46883076429367065,
+      "step": 1863
+    },
+    {
+      "epoch": 0.4752677205507394,
+      "grad_norm": 0.4507564902305603,
+      "learning_rate": 1.1641348503859664e-05,
+      "loss": 0.4657594859600067,
+      "step": 1864
+    },
+    {
+      "epoch": 0.47552269250382456,
+      "grad_norm": 0.4488622844219208,
+      "learning_rate": 1.1633028495356928e-05,
+      "loss": 0.481199711561203,
+      "step": 1865
+    },
+    {
+      "epoch": 0.4757776644569097,
+      "grad_norm": 0.47723981738090515,
+      "learning_rate": 1.1624707325296117e-05,
+      "loss": 0.47260582447052,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4760326364099949,
+      "grad_norm": 0.43852660059928894,
+      "learning_rate": 1.1616384999596006e-05,
+      "loss": 0.4724685847759247,
+      "step": 1867
+    },
+    {
+      "epoch": 0.47628760836308004,
+      "grad_norm": 0.45562633872032166,
+      "learning_rate": 1.1608061524176182e-05,
+      "loss": 0.4831947982311249,
+      "step": 1868
+    },
+    {
+      "epoch": 0.4765425803161652,
+      "grad_norm": 0.4537625312805176,
+      "learning_rate": 1.1599736904957057e-05,
+      "loss": 0.4736064374446869,
+      "step": 1869
+    },
+    {
+      "epoch": 0.47679755226925036,
+      "grad_norm": 0.44323453307151794,
+      "learning_rate": 1.1591411147859859e-05,
+      "loss": 0.4788297414779663,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4770525242223355,
+      "grad_norm": 0.4866369962692261,
+      "learning_rate": 1.1583084258806621e-05,
+      "loss": 0.46678242087364197,
+      "step": 1871
+    },
+    {
+      "epoch": 0.47730749617542073,
+      "grad_norm": 0.45373299717903137,
+      "learning_rate": 1.157475624372018e-05,
+      "loss": 0.471998393535614,
+      "step": 1872
+    },
+    {
+      "epoch": 0.4775624681285059,
+      "grad_norm": 0.4975714087486267,
+      "learning_rate": 1.1566427108524179e-05,
+      "loss": 0.46503281593322754,
+      "step": 1873
+    },
+    {
+      "epoch": 0.47781744008159105,
+      "grad_norm": 0.440778911113739,
+      "learning_rate": 1.1558096859143051e-05,
+      "loss": 0.4698757827281952,
+      "step": 1874
+    },
+    {
+      "epoch": 0.4780724120346762,
+      "grad_norm": 0.44512689113616943,
+      "learning_rate": 1.1549765501502028e-05,
+      "loss": 0.4689352214336395,
+      "step": 1875
+    },
+    {
+      "epoch": 0.47832738398776137,
+      "grad_norm": 0.4634833335876465,
+      "learning_rate": 1.154143304152713e-05,
+      "loss": 0.468364417552948,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4785823559408465,
+      "grad_norm": 0.46329474449157715,
+      "learning_rate": 1.1533099485145156e-05,
+      "loss": 0.468635618686676,
+      "step": 1877
+    },
+    {
+      "epoch": 0.4788373278939317,
+      "grad_norm": 0.4551597535610199,
+      "learning_rate": 1.1524764838283686e-05,
+      "loss": 0.46521398425102234,
+      "step": 1878
+    },
+    {
+      "epoch": 0.47909229984701684,
+      "grad_norm": 0.4328210651874542,
+      "learning_rate": 1.1516429106871078e-05,
+      "loss": 0.47337955236434937,
+      "step": 1879
+    },
+    {
+      "epoch": 0.479347271800102,
+      "grad_norm": 0.4416564106941223,
+      "learning_rate": 1.1508092296836468e-05,
+      "loss": 0.4794473350048065,
+      "step": 1880
+    },
+    {
+      "epoch": 0.47960224375318716,
+      "grad_norm": 0.4334598779678345,
+      "learning_rate": 1.1499754414109743e-05,
+      "loss": 0.47388195991516113,
+      "step": 1881
+    },
+    {
+      "epoch": 0.4798572157062723,
+      "grad_norm": 0.4369765818119049,
+      "learning_rate": 1.1491415464621562e-05,
+      "loss": 0.48022717237472534,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4801121876593575,
+      "grad_norm": 0.4837839901447296,
+      "learning_rate": 1.148307545430335e-05,
+      "loss": 0.46111083030700684,
+      "step": 1883
+    },
+    {
+      "epoch": 0.48036715961244264,
+      "grad_norm": 0.43741574883461,
+      "learning_rate": 1.1474734389087271e-05,
+      "loss": 0.47851133346557617,
+      "step": 1884
+    },
+    {
+      "epoch": 0.4806221315655278,
+      "grad_norm": 0.48897045850753784,
+      "learning_rate": 1.1466392274906253e-05,
+      "loss": 0.47374314069747925,
+      "step": 1885
+    },
+    {
+      "epoch": 0.48087710351861296,
+      "grad_norm": 0.46187931299209595,
+      "learning_rate": 1.1458049117693963e-05,
+      "loss": 0.4768770933151245,
+      "step": 1886
+    },
+    {
+      "epoch": 0.4811320754716981,
+      "grad_norm": 0.4495793879032135,
+      "learning_rate": 1.1449704923384813e-05,
+      "loss": 0.4840967655181885,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4813870474247833,
+      "grad_norm": 0.4757740795612335,
+      "learning_rate": 1.1441359697913947e-05,
+      "loss": 0.470970094203949,
+      "step": 1888
+    },
+    {
+      "epoch": 0.48164201937786844,
+      "grad_norm": 0.4387821853160858,
+      "learning_rate": 1.1433013447217254e-05,
+      "loss": 0.47704702615737915,
+      "step": 1889
+    },
+    {
+      "epoch": 0.4818969913309536,
+      "grad_norm": 0.6305357217788696,
+      "learning_rate": 1.1424666177231335e-05,
+      "loss": 0.473327100276947,
+      "step": 1890
+    },
+    {
+      "epoch": 0.48215196328403875,
+      "grad_norm": 0.4685002267360687,
+      "learning_rate": 1.1416317893893537e-05,
+      "loss": 0.466896653175354,
+      "step": 1891
+    },
+    {
+      "epoch": 0.4824069352371239,
+      "grad_norm": 0.46780869364738464,
+      "learning_rate": 1.140796860314191e-05,
+      "loss": 0.480934739112854,
+      "step": 1892
+    },
+    {
+      "epoch": 0.48266190719020907,
+      "grad_norm": 0.4916781187057495,
+      "learning_rate": 1.1399618310915228e-05,
+      "loss": 0.4645150601863861,
+      "step": 1893
+    },
+    {
+      "epoch": 0.48291687914329423,
+      "grad_norm": 0.4845971465110779,
+      "learning_rate": 1.1391267023152977e-05,
+      "loss": 0.46976611018180847,
+      "step": 1894
+    },
+    {
+      "epoch": 0.4831718510963794,
+      "grad_norm": 0.4984472990036011,
+      "learning_rate": 1.1382914745795353e-05,
+      "loss": 0.4808133542537689,
+      "step": 1895
+    },
+    {
+      "epoch": 0.48342682304946455,
+      "grad_norm": 0.4666406512260437,
+      "learning_rate": 1.137456148478325e-05,
+      "loss": 0.4635807275772095,
+      "step": 1896
+    },
+    {
+      "epoch": 0.4836817950025497,
+      "grad_norm": 0.44348442554473877,
+      "learning_rate": 1.1366207246058269e-05,
+      "loss": 0.47462043166160583,
+      "step": 1897
+    },
+    {
+      "epoch": 0.48393676695563487,
+      "grad_norm": 0.44801104068756104,
+      "learning_rate": 1.1357852035562703e-05,
+      "loss": 0.4728727340698242,
+      "step": 1898
+    },
+    {
+      "epoch": 0.48419173890872,
+      "grad_norm": 0.5481351017951965,
+      "learning_rate": 1.134949585923953e-05,
+      "loss": 0.44794484972953796,
+      "step": 1899
+    },
+    {
+      "epoch": 0.4844467108618052,
+      "grad_norm": 0.45803800225257874,
+      "learning_rate": 1.1341138723032426e-05,
+      "loss": 0.47609731554985046,
+      "step": 1900
+    },
+    {
+      "epoch": 0.48470168281489034,
+      "grad_norm": 0.4401257038116455,
+      "learning_rate": 1.1332780632885745e-05,
+      "loss": 0.46779781579971313,
+      "step": 1901
+    },
+    {
+      "epoch": 0.4849566547679755,
+      "grad_norm": 0.4817562699317932,
+      "learning_rate": 1.1324421594744516e-05,
+      "loss": 0.46378979086875916,
+      "step": 1902
+    },
+    {
+      "epoch": 0.48521162672106066,
+      "grad_norm": 0.4679580628871918,
+      "learning_rate": 1.1316061614554452e-05,
+      "loss": 0.48464304208755493,
+      "step": 1903
+    },
+    {
+      "epoch": 0.4854665986741458,
+      "grad_norm": 0.42400920391082764,
+      "learning_rate": 1.130770069826192e-05,
+      "loss": 0.4615113139152527,
+      "step": 1904
+    },
+    {
+      "epoch": 0.485721570627231,
+      "grad_norm": 0.4422542452812195,
+      "learning_rate": 1.1299338851813975e-05,
+      "loss": 0.47650349140167236,
+      "step": 1905
+    },
+    {
+      "epoch": 0.48597654258031614,
+      "grad_norm": 0.44117411971092224,
+      "learning_rate": 1.1290976081158306e-05,
+      "loss": 0.4791657328605652,
+      "step": 1906
+    },
+    {
+      "epoch": 0.4862315145334013,
+      "grad_norm": 0.5534968972206116,
+      "learning_rate": 1.1282612392243286e-05,
+      "loss": 0.46617746353149414,
+      "step": 1907
+    },
+    {
+      "epoch": 0.4864864864864865,
+      "grad_norm": 0.44438889622688293,
+      "learning_rate": 1.127424779101793e-05,
+      "loss": 0.47839054465293884,
+      "step": 1908
+    },
+    {
+      "epoch": 0.4867414584395717,
+      "grad_norm": 0.43262767791748047,
+      "learning_rate": 1.1265882283431892e-05,
+      "loss": 0.4703233242034912,
+      "step": 1909
+    },
+    {
+      "epoch": 0.48699643039265683,
+      "grad_norm": 0.4824342131614685,
+      "learning_rate": 1.1257515875435488e-05,
+      "loss": 0.4696802794933319,
+      "step": 1910
+    },
+    {
+      "epoch": 0.487251402345742,
+      "grad_norm": 0.43984854221343994,
+      "learning_rate": 1.1249148572979664e-05,
+      "loss": 0.4753193259239197,
+      "step": 1911
+    },
+    {
+      "epoch": 0.48750637429882715,
+      "grad_norm": 0.561900794506073,
+      "learning_rate": 1.1240780382016006e-05,
+      "loss": 0.4705323576927185,
+      "step": 1912
+    },
+    {
+      "epoch": 0.4877613462519123,
+      "grad_norm": 0.459989070892334,
+      "learning_rate": 1.1232411308496729e-05,
+      "loss": 0.4775056540966034,
+      "step": 1913
+    },
+    {
+      "epoch": 0.48801631820499747,
+      "grad_norm": 0.4523221254348755,
+      "learning_rate": 1.122404135837468e-05,
+      "loss": 0.4824284017086029,
+      "step": 1914
+    },
+    {
+      "epoch": 0.4882712901580826,
+      "grad_norm": 0.4518112540245056,
+      "learning_rate": 1.1215670537603324e-05,
+      "loss": 0.4524316191673279,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4885262621111678,
+      "grad_norm": 0.5053389668464661,
+      "learning_rate": 1.1207298852136753e-05,
+      "loss": 0.4648011326789856,
+      "step": 1916
+    },
+    {
+      "epoch": 0.48878123406425295,
+      "grad_norm": 0.43176308274269104,
+      "learning_rate": 1.1198926307929665e-05,
+      "loss": 0.45798033475875854,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4890362060173381,
+      "grad_norm": 0.46930280327796936,
+      "learning_rate": 1.1190552910937378e-05,
+      "loss": 0.47919926047325134,
+      "step": 1918
+    },
+    {
+      "epoch": 0.48929117797042326,
+      "grad_norm": 0.4317914843559265,
+      "learning_rate": 1.1182178667115806e-05,
+      "loss": 0.4498417377471924,
+      "step": 1919
+    },
+    {
+      "epoch": 0.4895461499235084,
+      "grad_norm": 0.4929078221321106,
+      "learning_rate": 1.1173803582421477e-05,
+      "loss": 0.46412748098373413,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4898011218765936,
+      "grad_norm": 0.4420813024044037,
+      "learning_rate": 1.116542766281151e-05,
+      "loss": 0.4748162627220154,
+      "step": 1921
+    },
+    {
+      "epoch": 0.49005609382967874,
+      "grad_norm": 0.4476282596588135,
+      "learning_rate": 1.1157050914243614e-05,
+      "loss": 0.4865252375602722,
+      "step": 1922
+    },
+    {
+      "epoch": 0.4903110657827639,
+      "grad_norm": 0.46605515480041504,
+      "learning_rate": 1.11486733426761e-05,
+      "loss": 0.46148931980133057,
+      "step": 1923
+    },
+    {
+      "epoch": 0.49056603773584906,
+      "grad_norm": 0.4584254026412964,
+      "learning_rate": 1.1140294954067853e-05,
+      "loss": 0.48570913076400757,
+      "step": 1924
+    },
+    {
+      "epoch": 0.4908210096889342,
+      "grad_norm": 0.4745509624481201,
+      "learning_rate": 1.1131915754378344e-05,
+      "loss": 0.4840744733810425,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4910759816420194,
+      "grad_norm": 0.4451920986175537,
+      "learning_rate": 1.1123535749567622e-05,
+      "loss": 0.46334999799728394,
+      "step": 1926
+    },
+    {
+      "epoch": 0.49133095359510454,
+      "grad_norm": 0.4377116560935974,
+      "learning_rate": 1.1115154945596306e-05,
+      "loss": 0.46594923734664917,
+      "step": 1927
+    },
+    {
+      "epoch": 0.4915859255481897,
+      "grad_norm": 0.4420704245567322,
+      "learning_rate": 1.1106773348425581e-05,
+      "loss": 0.4736064672470093,
+      "step": 1928
+    },
+    {
+      "epoch": 0.49184089750127485,
+      "grad_norm": 0.4611606001853943,
+      "learning_rate": 1.1098390964017204e-05,
+      "loss": 0.4691614806652069,
+      "step": 1929
+    },
+    {
+      "epoch": 0.49209586945436,
+      "grad_norm": 0.4173363745212555,
+      "learning_rate": 1.1090007798333486e-05,
+      "loss": 0.4697965979576111,
+      "step": 1930
+    },
+    {
+      "epoch": 0.49235084140744517,
+      "grad_norm": 0.45429426431655884,
+      "learning_rate": 1.1081623857337297e-05,
+      "loss": 0.4782986342906952,
+      "step": 1931
+    },
+    {
+      "epoch": 0.49260581336053033,
+      "grad_norm": 0.44332918524742126,
+      "learning_rate": 1.1073239146992054e-05,
+      "loss": 0.4492153525352478,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4928607853136155,
+      "grad_norm": 0.44081881642341614,
+      "learning_rate": 1.1064853673261727e-05,
+      "loss": 0.4710718095302582,
+      "step": 1933
+    },
+    {
+      "epoch": 0.49311575726670065,
+      "grad_norm": 0.45330432057380676,
+      "learning_rate": 1.105646744211082e-05,
+      "loss": 0.4720541834831238,
+      "step": 1934
+    },
+    {
+      "epoch": 0.4933707292197858,
+      "grad_norm": 0.45238229632377625,
+      "learning_rate": 1.1048080459504388e-05,
+      "loss": 0.46549171209335327,
+      "step": 1935
+    },
+    {
+      "epoch": 0.49362570117287097,
+      "grad_norm": 0.4571751654148102,
+      "learning_rate": 1.1039692731408014e-05,
+      "loss": 0.46449217200279236,
+      "step": 1936
+    },
+    {
+      "epoch": 0.4938806731259561,
+      "grad_norm": 0.44922831654548645,
+      "learning_rate": 1.1031304263787811e-05,
+      "loss": 0.47091251611709595,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4941356450790413,
+      "grad_norm": 0.516984224319458,
+      "learning_rate": 1.1022915062610416e-05,
+      "loss": 0.45369237661361694,
+      "step": 1938
+    },
+    {
+      "epoch": 0.49439061703212644,
+      "grad_norm": 0.4570682942867279,
+      "learning_rate": 1.1014525133842986e-05,
+      "loss": 0.47135066986083984,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4946455889852116,
+      "grad_norm": 0.4430196285247803,
+      "learning_rate": 1.100613448345321e-05,
+      "loss": 0.4640349745750427,
+      "step": 1940
+    },
+    {
+      "epoch": 0.49490056093829676,
+      "grad_norm": 0.45458775758743286,
+      "learning_rate": 1.0997743117409277e-05,
+      "loss": 0.4690319001674652,
+      "step": 1941
+    },
+    {
+      "epoch": 0.4951555328913819,
+      "grad_norm": 0.47576791048049927,
+      "learning_rate": 1.098935104167988e-05,
+      "loss": 0.47333425283432007,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4954105048444671,
+      "grad_norm": 0.46054837107658386,
+      "learning_rate": 1.0980958262234235e-05,
+      "loss": 0.4702942371368408,
+      "step": 1943
+    },
+    {
+      "epoch": 0.4956654767975523,
+      "grad_norm": 0.4609912931919098,
+      "learning_rate": 1.0972564785042045e-05,
+      "loss": 0.47456806898117065,
+      "step": 1944
+    },
+    {
+      "epoch": 0.49592044875063745,
+      "grad_norm": 0.4505545198917389,
+      "learning_rate": 1.0964170616073512e-05,
+      "loss": 0.46712201833724976,
+      "step": 1945
+    },
+    {
+      "epoch": 0.4961754207037226,
+      "grad_norm": 0.46952521800994873,
+      "learning_rate": 1.0955775761299333e-05,
+      "loss": 0.4621439576148987,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4964303926568078,
+      "grad_norm": 0.4679666757583618,
+      "learning_rate": 1.0947380226690686e-05,
+      "loss": 0.4812907576560974,
+      "step": 1947
+    },
+    {
+      "epoch": 0.49668536460989293,
+      "grad_norm": 0.43158140778541565,
+      "learning_rate": 1.093898401821924e-05,
+      "loss": 0.462116003036499,
+      "step": 1948
+    },
+    {
+      "epoch": 0.4969403365629781,
+      "grad_norm": 0.4336107075214386,
+      "learning_rate": 1.093058714185714e-05,
+      "loss": 0.4580172300338745,
+      "step": 1949
+    },
+    {
+      "epoch": 0.49719530851606325,
+      "grad_norm": 0.5117933750152588,
+      "learning_rate": 1.0922189603577011e-05,
+      "loss": 0.46445757150650024,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4974502804691484,
+      "grad_norm": 0.4405916929244995,
+      "learning_rate": 1.0913791409351935e-05,
+      "loss": 0.4670753479003906,
+      "step": 1951
+    },
+    {
+      "epoch": 0.49770525242223357,
+      "grad_norm": 0.4495008587837219,
+      "learning_rate": 1.0905392565155477e-05,
+      "loss": 0.4655703902244568,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4979602243753187,
+      "grad_norm": 0.45106127858161926,
+      "learning_rate": 1.089699307696166e-05,
+      "loss": 0.4770054519176483,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4982151963284039,
+      "grad_norm": 0.44882631301879883,
+      "learning_rate": 1.0888592950744955e-05,
+      "loss": 0.46113085746765137,
+      "step": 1954
+    },
+    {
+      "epoch": 0.49847016828148905,
+      "grad_norm": 0.46270790696144104,
+      "learning_rate": 1.0880192192480291e-05,
+      "loss": 0.4824230670928955,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4987251402345742,
+      "grad_norm": 0.4500824213027954,
+      "learning_rate": 1.0871790808143063e-05,
+      "loss": 0.46038350462913513,
+      "step": 1956
+    },
+    {
+      "epoch": 0.49898011218765936,
+      "grad_norm": 0.600124180316925,
+      "learning_rate": 1.086338880370909e-05,
+      "loss": 0.4829053282737732,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4992350841407445,
+      "grad_norm": 0.435794472694397,
+      "learning_rate": 1.0854986185154638e-05,
+      "loss": 0.45988592505455017,
+      "step": 1958
+    },
+    {
+      "epoch": 0.4994900560938297,
+      "grad_norm": 0.4310831129550934,
+      "learning_rate": 1.0846582958456418e-05,
+      "loss": 0.45987948775291443,
+      "step": 1959
+    },
+    {
+      "epoch": 0.49974502804691484,
+      "grad_norm": 0.43353962898254395,
+      "learning_rate": 1.0838179129591562e-05,
+      "loss": 0.4516587257385254,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4458468556404114,
+      "learning_rate": 1.082977470453764e-05,
+      "loss": 0.469723641872406,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5002549719530852,
+      "grad_norm": 0.4295446574687958,
+      "learning_rate": 1.0821369689272639e-05,
+      "loss": 0.4734836518764496,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5005099439061703,
+      "grad_norm": 0.42044463753700256,
+      "learning_rate": 1.0812964089774968e-05,
+      "loss": 0.4678879678249359,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5007649158592555,
+      "grad_norm": 0.48239123821258545,
+      "learning_rate": 1.0804557912023455e-05,
+      "loss": 0.4636920690536499,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5010198878123406,
+      "grad_norm": 0.47312313318252563,
+      "learning_rate": 1.0796151161997334e-05,
+      "loss": 0.4822813868522644,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5012748597654259,
+      "grad_norm": 0.5328629016876221,
+      "learning_rate": 1.0787743845676256e-05,
+      "loss": 0.45847058296203613,
+      "step": 1966
+    },
+    {
+      "epoch": 0.501529831718511,
+      "grad_norm": 0.45701542496681213,
+      "learning_rate": 1.0779335969040252e-05,
+      "loss": 0.471851110458374,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5017848036715962,
+      "grad_norm": 0.45358672738075256,
+      "learning_rate": 1.0770927538069781e-05,
+      "loss": 0.4805513322353363,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5020397756246813,
+      "grad_norm": 0.4443429410457611,
+      "learning_rate": 1.076251855874568e-05,
+      "loss": 0.4669516682624817,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5022947475777665,
+      "grad_norm": 0.49044740200042725,
+      "learning_rate": 1.0754109037049174e-05,
+      "loss": 0.46486878395080566,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5025497195308516,
+      "grad_norm": 0.44463393092155457,
+      "learning_rate": 1.0745698978961875e-05,
+      "loss": 0.45962488651275635,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5028046914839368,
+      "grad_norm": 0.47817161679267883,
+      "learning_rate": 1.0737288390465792e-05,
+      "loss": 0.47821468114852905,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5030596634370219,
+      "grad_norm": 0.4611988663673401,
+      "learning_rate": 1.0728877277543288e-05,
+      "loss": 0.45836731791496277,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5033146353901071,
+      "grad_norm": 0.46209344267845154,
+      "learning_rate": 1.0720465646177111e-05,
+      "loss": 0.47259795665740967,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5035696073431922,
+      "grad_norm": 0.4538228213787079,
+      "learning_rate": 1.0712053502350384e-05,
+      "loss": 0.47403550148010254,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5038245792962774,
+      "grad_norm": 0.443917453289032,
+      "learning_rate": 1.0703640852046577e-05,
+      "loss": 0.4648395776748657,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5040795512493625,
+      "grad_norm": 0.45121023058891296,
+      "learning_rate": 1.0695227701249537e-05,
+      "loss": 0.4645318388938904,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5043345232024478,
+      "grad_norm": 0.47773703932762146,
+      "learning_rate": 1.0686814055943459e-05,
+      "loss": 0.46701762080192566,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5045894951555329,
+      "grad_norm": 0.46860113739967346,
+      "learning_rate": 1.0678399922112892e-05,
+      "loss": 0.47811728715896606,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5048444671086181,
+      "grad_norm": 0.4680320620536804,
+      "learning_rate": 1.066998530574273e-05,
+      "loss": 0.44749268889427185,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5050994390617032,
+      "grad_norm": 0.45026010274887085,
+      "learning_rate": 1.0661570212818211e-05,
+      "loss": 0.4677862524986267,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5053544110147884,
+      "grad_norm": 0.4540463387966156,
+      "learning_rate": 1.0653154649324918e-05,
+      "loss": 0.47464463114738464,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5056093829678735,
+      "grad_norm": 0.4496227502822876,
+      "learning_rate": 1.0644738621248753e-05,
+      "loss": 0.46616873145103455,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5058643549209587,
+      "grad_norm": 0.4450114667415619,
+      "learning_rate": 1.0636322134575972e-05,
+      "loss": 0.4677341878414154,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5061193268740438,
+      "grad_norm": 0.4330081343650818,
+      "learning_rate": 1.0627905195293135e-05,
+      "loss": 0.47850990295410156,
+      "step": 1985
+    },
+    {
+      "epoch": 0.506374298827129,
+      "grad_norm": 0.4307883083820343,
+      "learning_rate": 1.0619487809387137e-05,
+      "loss": 0.47758373618125916,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5066292707802141,
+      "grad_norm": 0.4381122887134552,
+      "learning_rate": 1.0611069982845183e-05,
+      "loss": 0.4692190885543823,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5068842427332994,
+      "grad_norm": 0.43644505739212036,
+      "learning_rate": 1.0602651721654804e-05,
+      "loss": 0.46923840045928955,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5071392146863845,
+      "grad_norm": 0.513249397277832,
+      "learning_rate": 1.0594233031803822e-05,
+      "loss": 0.4702669382095337,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5073941866394697,
+      "grad_norm": 0.4524279832839966,
+      "learning_rate": 1.0585813919280377e-05,
+      "loss": 0.4500643014907837,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5076491585925548,
+      "grad_norm": 0.4405861496925354,
+      "learning_rate": 1.0577394390072904e-05,
+      "loss": 0.47343379259109497,
+      "step": 1991
+    },
+    {
+      "epoch": 0.50790413054564,
+      "grad_norm": 0.44109389185905457,
+      "learning_rate": 1.056897445017014e-05,
+      "loss": 0.46540567278862,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5081591024987251,
+      "grad_norm": 0.46960288286209106,
+      "learning_rate": 1.0560554105561108e-05,
+      "loss": 0.4597005248069763,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5084140744518103,
+      "grad_norm": 0.46017128229141235,
+      "learning_rate": 1.055213336223512e-05,
+      "loss": 0.4604566693305969,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5086690464048954,
+      "grad_norm": 0.43463587760925293,
+      "learning_rate": 1.0543712226181776e-05,
+      "loss": 0.4535146653652191,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5089240183579806,
+      "grad_norm": 0.46322280168533325,
+      "learning_rate": 1.053529070339095e-05,
+      "loss": 0.47358083724975586,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5091789903110658,
+      "grad_norm": 0.4796248972415924,
+      "learning_rate": 1.0526868799852797e-05,
+      "loss": 0.47287294268608093,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5094339622641509,
+      "grad_norm": 0.5290987491607666,
+      "learning_rate": 1.0518446521557735e-05,
+      "loss": 0.46784111857414246,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5096889342172362,
+      "grad_norm": 0.4367986023426056,
+      "learning_rate": 1.0510023874496451e-05,
+      "loss": 0.4657739996910095,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5099439061703213,
+      "grad_norm": 0.446732759475708,
+      "learning_rate": 1.0501600864659902e-05,
+      "loss": 0.4691154956817627,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5101988781234065,
+      "grad_norm": 0.4529222846031189,
+      "learning_rate": 1.0493177498039295e-05,
+      "loss": 0.46201086044311523,
+      "step": 2001
+    },
+    {
+      "epoch": 0.5104538500764916,
+      "grad_norm": 0.44185587763786316,
+      "learning_rate": 1.0484753780626089e-05,
+      "loss": 0.47200000286102295,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5107088220295768,
+      "grad_norm": 0.4439525306224823,
+      "learning_rate": 1.0476329718412001e-05,
+      "loss": 0.47707054018974304,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5109637939826619,
+      "grad_norm": 0.4340081512928009,
+      "learning_rate": 1.0467905317388987e-05,
+      "loss": 0.45767197012901306,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5112187659357471,
+      "grad_norm": 0.4600531756877899,
+      "learning_rate": 1.0459480583549245e-05,
+      "loss": 0.4482811391353607,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5114737378888322,
+      "grad_norm": 0.4530130624771118,
+      "learning_rate": 1.0451055522885208e-05,
+      "loss": 0.4777918756008148,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5117287098419174,
+      "grad_norm": 0.4777997136116028,
+      "learning_rate": 1.044263014138955e-05,
+      "loss": 0.4598935544490814,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5119836817950025,
+      "grad_norm": 0.42285043001174927,
+      "learning_rate": 1.0434204445055161e-05,
+      "loss": 0.4733307361602783,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5122386537480877,
+      "grad_norm": 0.4836715757846832,
+      "learning_rate": 1.0425778439875164e-05,
+      "loss": 0.47103893756866455,
+      "step": 2009
+    },
+    {
+      "epoch": 0.5124936257011729,
+      "grad_norm": 0.45524367690086365,
+      "learning_rate": 1.04173521318429e-05,
+      "loss": 0.4625380039215088,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5127485976542581,
+      "grad_norm": 0.44874754548072815,
+      "learning_rate": 1.0408925526951919e-05,
+      "loss": 0.46822822093963623,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5130035696073432,
+      "grad_norm": 0.532166063785553,
+      "learning_rate": 1.0400498631195992e-05,
+      "loss": 0.46029651165008545,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5132585415604284,
+      "grad_norm": 0.4773326814174652,
+      "learning_rate": 1.039207145056909e-05,
+      "loss": 0.45712560415267944,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5135135135135135,
+      "grad_norm": 0.4635961353778839,
+      "learning_rate": 1.038364399106539e-05,
+      "loss": 0.4717424809932709,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5137684854665987,
+      "grad_norm": 0.4427274763584137,
+      "learning_rate": 1.0375216258679267e-05,
+      "loss": 0.4571472108364105,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5140234574196838,
+      "grad_norm": 0.4772450923919678,
+      "learning_rate": 1.0366788259405281e-05,
+      "loss": 0.46459245681762695,
+      "step": 2016
+    },
+    {
+      "epoch": 0.514278429372769,
+      "grad_norm": 0.44661223888397217,
+      "learning_rate": 1.0358359999238202e-05,
+      "loss": 0.46645474433898926,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5145334013258541,
+      "grad_norm": 0.4430338442325592,
+      "learning_rate": 1.0349931484172964e-05,
+      "loss": 0.4655427634716034,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5147883732789393,
+      "grad_norm": 0.4474931061267853,
+      "learning_rate": 1.0341502720204696e-05,
+      "loss": 0.4667506217956543,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5150433452320244,
+      "grad_norm": 0.46259480714797974,
+      "learning_rate": 1.03330737133287e-05,
+      "loss": 0.4495689570903778,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5152983171851097,
+      "grad_norm": 0.44517841935157776,
+      "learning_rate": 1.032464446954045e-05,
+      "loss": 0.47217345237731934,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5155532891381948,
+      "grad_norm": 0.4635341763496399,
+      "learning_rate": 1.031621499483559e-05,
+      "loss": 0.466999351978302,
+      "step": 2022
+    },
+    {
+      "epoch": 0.51580826109128,
+      "grad_norm": 0.4462071657180786,
+      "learning_rate": 1.0307785295209927e-05,
+      "loss": 0.47211140394210815,
+      "step": 2023
+    },
+    {
+      "epoch": 0.5160632330443651,
+      "grad_norm": 0.42271414399147034,
+      "learning_rate": 1.0299355376659426e-05,
+      "loss": 0.47100910544395447,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5163182049974503,
+      "grad_norm": 0.4277551472187042,
+      "learning_rate": 1.0290925245180214e-05,
+      "loss": 0.46340689063072205,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5165731769505354,
+      "grad_norm": 0.4386981725692749,
+      "learning_rate": 1.0282494906768567e-05,
+      "loss": 0.46859121322631836,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5168281489036206,
+      "grad_norm": 0.4549739956855774,
+      "learning_rate": 1.0274064367420897e-05,
+      "loss": 0.4723912477493286,
+      "step": 2027
+    },
+    {
+      "epoch": 0.5170831208567057,
+      "grad_norm": 0.46073293685913086,
+      "learning_rate": 1.026563363313378e-05,
+      "loss": 0.46750569343566895,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5173380928097909,
+      "grad_norm": 0.4419374465942383,
+      "learning_rate": 1.0257202709903915e-05,
+      "loss": 0.46308594942092896,
+      "step": 2029
+    },
+    {
+      "epoch": 0.517593064762876,
+      "grad_norm": 0.45616763830184937,
+      "learning_rate": 1.024877160372814e-05,
+      "loss": 0.4581572413444519,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5178480367159612,
+      "grad_norm": 0.5098860263824463,
+      "learning_rate": 1.0240340320603421e-05,
+      "loss": 0.4721102714538574,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5181030086690463,
+      "grad_norm": 0.42570966482162476,
+      "learning_rate": 1.0231908866526851e-05,
+      "loss": 0.46494436264038086,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5183579806221316,
+      "grad_norm": 0.44891291856765747,
+      "learning_rate": 1.0223477247495655e-05,
+      "loss": 0.4593876600265503,
+      "step": 2033
+    },
+    {
+      "epoch": 0.5186129525752168,
+      "grad_norm": 0.4411010444164276,
+      "learning_rate": 1.0215045469507152e-05,
+      "loss": 0.46814531087875366,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5188679245283019,
+      "grad_norm": 0.4394744634628296,
+      "learning_rate": 1.0206613538558797e-05,
+      "loss": 0.4597451090812683,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5191228964813871,
+      "grad_norm": 0.45414721965789795,
+      "learning_rate": 1.019818146064814e-05,
+      "loss": 0.46010684967041016,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5193778684344722,
+      "grad_norm": 0.4221673905849457,
+      "learning_rate": 1.0189749241772843e-05,
+      "loss": 0.4623414874076843,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5196328403875574,
+      "grad_norm": 0.45032015442848206,
+      "learning_rate": 1.0181316887930663e-05,
+      "loss": 0.45287615060806274,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5198878123406425,
+      "grad_norm": 0.4528999626636505,
+      "learning_rate": 1.0172884405119457e-05,
+      "loss": 0.4615684449672699,
+      "step": 2039
+    },
+    {
+      "epoch": 0.5201427842937277,
+      "grad_norm": 0.45465999841690063,
+      "learning_rate": 1.016445179933717e-05,
+      "loss": 0.46677809953689575,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5203977562468128,
+      "grad_norm": 0.45125019550323486,
+      "learning_rate": 1.015601907658184e-05,
+      "loss": 0.45892518758773804,
+      "step": 2041
+    },
+    {
+      "epoch": 0.520652728199898,
+      "grad_norm": 0.4755760133266449,
+      "learning_rate": 1.0147586242851586e-05,
+      "loss": 0.45814356207847595,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5209077001529832,
+      "grad_norm": 0.44352635741233826,
+      "learning_rate": 1.0139153304144597e-05,
+      "loss": 0.46379196643829346,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5211626721060684,
+      "grad_norm": 0.43896499276161194,
+      "learning_rate": 1.013072026645915e-05,
+      "loss": 0.46776318550109863,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5214176440591535,
+      "grad_norm": 0.47973087430000305,
+      "learning_rate": 1.0122287135793582e-05,
+      "loss": 0.46471667289733887,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5216726160122387,
+      "grad_norm": 0.4453974962234497,
+      "learning_rate": 1.011385391814631e-05,
+      "loss": 0.45457661151885986,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5219275879653238,
+      "grad_norm": 0.45497700572013855,
+      "learning_rate": 1.0105420619515798e-05,
+      "loss": 0.4676205813884735,
+      "step": 2047
+    },
+    {
+      "epoch": 0.522182559918409,
+      "grad_norm": 0.4510899782180786,
+      "learning_rate": 1.0096987245900569e-05,
+      "loss": 0.47454676032066345,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5224375318714941,
+      "grad_norm": 0.4445478916168213,
+      "learning_rate": 1.0088553803299215e-05,
+      "loss": 0.4578350782394409,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5226925038245793,
+      "grad_norm": 0.4726911783218384,
+      "learning_rate": 1.0080120297710355e-05,
+      "loss": 0.45180433988571167,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5229474757776644,
+      "grad_norm": 0.46544188261032104,
+      "learning_rate": 1.0071686735132669e-05,
+      "loss": 0.4593530297279358,
+      "step": 2051
+    },
+    {
+      "epoch": 0.5232024477307496,
+      "grad_norm": 0.44481343030929565,
+      "learning_rate": 1.0063253121564868e-05,
+      "loss": 0.45732077956199646,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5234574196838347,
+      "grad_norm": 0.4890146255493164,
+      "learning_rate": 1.005481946300571e-05,
+      "loss": 0.46841156482696533,
+      "step": 2053
+    },
+    {
+      "epoch": 0.52371239163692,
+      "grad_norm": 0.46075499057769775,
+      "learning_rate": 1.0046385765453972e-05,
+      "loss": 0.4659883975982666,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5239673635900051,
+      "grad_norm": 0.4498838484287262,
+      "learning_rate": 1.0037952034908472e-05,
+      "loss": 0.46704480051994324,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5242223355430903,
+      "grad_norm": 0.4813586473464966,
+      "learning_rate": 1.0029518277368034e-05,
+      "loss": 0.45885729789733887,
+      "step": 2056
+    },
+    {
+      "epoch": 0.5244773074961754,
+      "grad_norm": 0.45792996883392334,
+      "learning_rate": 1.0021084498831522e-05,
+      "loss": 0.4637394845485687,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5247322794492606,
+      "grad_norm": 0.4410227835178375,
+      "learning_rate": 1.0012650705297802e-05,
+      "loss": 0.4597325325012207,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5249872514023457,
+      "grad_norm": 0.4623326063156128,
+      "learning_rate": 1.000421690276575e-05,
+      "loss": 0.47189098596572876,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5252422233554309,
+      "grad_norm": 0.4475923776626587,
+      "learning_rate": 9.995783097234255e-06,
+      "loss": 0.46808943152427673,
+      "step": 2060
+    },
+    {
+      "epoch": 0.525497195308516,
+      "grad_norm": 0.45596396923065186,
+      "learning_rate": 9.987349294702203e-06,
+      "loss": 0.4588278532028198,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5257521672616012,
+      "grad_norm": 0.46377086639404297,
+      "learning_rate": 9.978915501168481e-06,
+      "loss": 0.468145489692688,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5260071392146863,
+      "grad_norm": 0.450214147567749,
+      "learning_rate": 9.970481722631968e-06,
+      "loss": 0.4631403386592865,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5262621111677716,
+      "grad_norm": 0.45720988512039185,
+      "learning_rate": 9.962047965091531e-06,
+      "loss": 0.4511702060699463,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5265170831208567,
+      "grad_norm": 0.4529277980327606,
+      "learning_rate": 9.95361423454603e-06,
+      "loss": 0.4667152762413025,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5267720550739419,
+      "grad_norm": 0.4591916799545288,
+      "learning_rate": 9.945180536994295e-06,
+      "loss": 0.46990442276000977,
+      "step": 2066
+    },
+    {
+      "epoch": 0.527027027027027,
+      "grad_norm": 0.43610069155693054,
+      "learning_rate": 9.936746878435137e-06,
+      "loss": 0.44906729459762573,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5272819989801122,
+      "grad_norm": 0.44433870911598206,
+      "learning_rate": 9.928313264867335e-06,
+      "loss": 0.45793232321739197,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5275369709331974,
+      "grad_norm": 0.4354260563850403,
+      "learning_rate": 9.919879702289649e-06,
+      "loss": 0.4704725444316864,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5277919428862825,
+      "grad_norm": 0.4708549678325653,
+      "learning_rate": 9.911446196700788e-06,
+      "loss": 0.4823994040489197,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5280469148393677,
+      "grad_norm": 0.44031861424446106,
+      "learning_rate": 9.903012754099431e-06,
+      "loss": 0.4768982529640198,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5283018867924528,
+      "grad_norm": 0.4868448078632355,
+      "learning_rate": 9.894579380484206e-06,
+      "loss": 0.460830420255661,
+      "step": 2072
+    },
+    {
+      "epoch": 0.528556858745538,
+      "grad_norm": 0.4490775167942047,
+      "learning_rate": 9.886146081853694e-06,
+      "loss": 0.4631894826889038,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5288118306986231,
+      "grad_norm": 0.4440162181854248,
+      "learning_rate": 9.87771286420642e-06,
+      "loss": 0.4668932855129242,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5290668026517084,
+      "grad_norm": 0.4418855607509613,
+      "learning_rate": 9.869279733540856e-06,
+      "loss": 0.46449732780456543,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5293217746047935,
+      "grad_norm": 0.4457247853279114,
+      "learning_rate": 9.860846695855406e-06,
+      "loss": 0.4614427089691162,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5295767465578787,
+      "grad_norm": 0.4354749619960785,
+      "learning_rate": 9.852413757148418e-06,
+      "loss": 0.4676775336265564,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5298317185109638,
+      "grad_norm": 0.42488738894462585,
+      "learning_rate": 9.84398092341816e-06,
+      "loss": 0.4580889940261841,
+      "step": 2078
+    },
+    {
+      "epoch": 0.530086690464049,
+      "grad_norm": 0.45775771141052246,
+      "learning_rate": 9.835548200662831e-06,
+      "loss": 0.46174877882003784,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5303416624171341,
+      "grad_norm": 0.4479270875453949,
+      "learning_rate": 9.827115594880545e-06,
+      "loss": 0.47181570529937744,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5305966343702193,
+      "grad_norm": 0.4417240023612976,
+      "learning_rate": 9.81868311206934e-06,
+      "loss": 0.4601489305496216,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5308516063233044,
+      "grad_norm": 0.43877771496772766,
+      "learning_rate": 9.81025075822716e-06,
+      "loss": 0.4636847972869873,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5311065782763896,
+      "grad_norm": 0.4579344391822815,
+      "learning_rate": 9.801818539351862e-06,
+      "loss": 0.4594728350639343,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5313615502294747,
+      "grad_norm": 0.49850475788116455,
+      "learning_rate": 9.793386461441205e-06,
+      "loss": 0.46444177627563477,
+      "step": 2084
+    },
+    {
+      "epoch": 0.53161652218256,
+      "grad_norm": 0.4226575493812561,
+      "learning_rate": 9.784954530492846e-06,
+      "loss": 0.45225656032562256,
+      "step": 2085
+    },
+    {
+      "epoch": 0.531871494135645,
+      "grad_norm": 0.4468752145767212,
+      "learning_rate": 9.77652275250435e-06,
+      "loss": 0.46698713302612305,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5321264660887303,
+      "grad_norm": 0.46996641159057617,
+      "learning_rate": 9.76809113347315e-06,
+      "loss": 0.4715825915336609,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5323814380418154,
+      "grad_norm": 0.4626185894012451,
+      "learning_rate": 9.75965967939658e-06,
+      "loss": 0.47345566749572754,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5326364099949006,
+      "grad_norm": 0.46349525451660156,
+      "learning_rate": 9.751228396271863e-06,
+      "loss": 0.4519904851913452,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5328913819479857,
+      "grad_norm": 0.4393472671508789,
+      "learning_rate": 9.742797290096085e-06,
+      "loss": 0.47511130571365356,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5331463539010709,
+      "grad_norm": 0.45189473032951355,
+      "learning_rate": 9.73436636686622e-06,
+      "loss": 0.4526618421077728,
+      "step": 2091
+    },
+    {
+      "epoch": 0.533401325854156,
+      "grad_norm": 0.44831642508506775,
+      "learning_rate": 9.725935632579105e-06,
+      "loss": 0.47018808126449585,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5336562978072412,
+      "grad_norm": 0.4449017345905304,
+      "learning_rate": 9.71750509323144e-06,
+      "loss": 0.463678777217865,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5339112697603263,
+      "grad_norm": 0.4990766942501068,
+      "learning_rate": 9.709074754819788e-06,
+      "loss": 0.4615117013454437,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5341662417134115,
+      "grad_norm": 0.4472992420196533,
+      "learning_rate": 9.700644623340576e-06,
+      "loss": 0.4790828227996826,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5344212136664966,
+      "grad_norm": 0.4707857072353363,
+      "learning_rate": 9.692214704790078e-06,
+      "loss": 0.4669959545135498,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5346761856195819,
+      "grad_norm": 0.47750231623649597,
+      "learning_rate": 9.683785005164412e-06,
+      "loss": 0.4698825478553772,
+      "step": 2097
+    },
+    {
+      "epoch": 0.534931157572667,
+      "grad_norm": 0.44692638516426086,
+      "learning_rate": 9.675355530459554e-06,
+      "loss": 0.4576447904109955,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5351861295257522,
+      "grad_norm": 0.4795735776424408,
+      "learning_rate": 9.666926286671306e-06,
+      "loss": 0.45990800857543945,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5354411014788373,
+      "grad_norm": 0.452642023563385,
+      "learning_rate": 9.658497279795307e-06,
+      "loss": 0.45492836833000183,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5356960734319225,
+      "grad_norm": 0.492758572101593,
+      "learning_rate": 9.650068515827039e-06,
+      "loss": 0.4594270884990692,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5359510453850076,
+      "grad_norm": 0.44014111161231995,
+      "learning_rate": 9.641640000761803e-06,
+      "loss": 0.4722544550895691,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5362060173380928,
+      "grad_norm": 0.4531368911266327,
+      "learning_rate": 9.63321174059472e-06,
+      "loss": 0.4664442539215088,
+      "step": 2103
+    },
+    {
+      "epoch": 0.536460989291178,
+      "grad_norm": 0.4387378692626953,
+      "learning_rate": 9.624783741320738e-06,
+      "loss": 0.48242801427841187,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5367159612442631,
+      "grad_norm": 0.46323493123054504,
+      "learning_rate": 9.616356008934615e-06,
+      "loss": 0.4569288194179535,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5369709331973483,
+      "grad_norm": 0.4467400014400482,
+      "learning_rate": 9.607928549430913e-06,
+      "loss": 0.443368136882782,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5372259051504334,
+      "grad_norm": 0.46812474727630615,
+      "learning_rate": 9.599501368804011e-06,
+      "loss": 0.47366127371788025,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5374808771035187,
+      "grad_norm": 0.45175814628601074,
+      "learning_rate": 9.591074473048083e-06,
+      "loss": 0.4567645192146301,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5377358490566038,
+      "grad_norm": 0.45011746883392334,
+      "learning_rate": 9.582647868157103e-06,
+      "loss": 0.4638264775276184,
+      "step": 2109
+    },
+    {
+      "epoch": 0.537990821009689,
+      "grad_norm": 0.46512457728385925,
+      "learning_rate": 9.574221560124837e-06,
+      "loss": 0.4629848897457123,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5382457929627741,
+      "grad_norm": 0.5198569297790527,
+      "learning_rate": 9.565795554944844e-06,
+      "loss": 0.46182310581207275,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5385007649158593,
+      "grad_norm": 0.41323816776275635,
+      "learning_rate": 9.557369858610454e-06,
+      "loss": 0.4664433002471924,
+      "step": 2112
+    },
+    {
+      "epoch": 0.5387557368689444,
+      "grad_norm": 0.4467845559120178,
+      "learning_rate": 9.548944477114794e-06,
+      "loss": 0.46828243136405945,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5390107088220296,
+      "grad_norm": 0.4535411298274994,
+      "learning_rate": 9.54051941645076e-06,
+      "loss": 0.47404253482818604,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5392656807751147,
+      "grad_norm": 0.4470507800579071,
+      "learning_rate": 9.532094682611018e-06,
+      "loss": 0.4727866053581238,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5395206527281999,
+      "grad_norm": 0.4469457268714905,
+      "learning_rate": 9.523670281588002e-06,
+      "loss": 0.45999619364738464,
+      "step": 2116
+    },
+    {
+      "epoch": 0.539775624681285,
+      "grad_norm": 0.4599631428718567,
+      "learning_rate": 9.515246219373911e-06,
+      "loss": 0.465770959854126,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5400305966343703,
+      "grad_norm": 0.4520121216773987,
+      "learning_rate": 9.50682250196071e-06,
+      "loss": 0.46152377128601074,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5402855685874554,
+      "grad_norm": 0.4410518705844879,
+      "learning_rate": 9.498399135340101e-06,
+      "loss": 0.47138556838035583,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 0.46009910106658936,
+      "learning_rate": 9.489976125503552e-06,
+      "loss": 0.4417486786842346,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5407955124936257,
+      "grad_norm": 0.4274013638496399,
+      "learning_rate": 9.481553478442268e-06,
+      "loss": 0.4546934962272644,
+      "step": 2121
+    },
+    {
+      "epoch": 0.5410504844467109,
+      "grad_norm": 0.46814054250717163,
+      "learning_rate": 9.473131200147205e-06,
+      "loss": 0.4571819305419922,
+      "step": 2122
+    },
+    {
+      "epoch": 0.541305456399796,
+      "grad_norm": 0.45113396644592285,
+      "learning_rate": 9.46470929660905e-06,
+      "loss": 0.4620974659919739,
+      "step": 2123
+    },
+    {
+      "epoch": 0.5415604283528812,
+      "grad_norm": 0.4376858174800873,
+      "learning_rate": 9.456287773818227e-06,
+      "loss": 0.45991358160972595,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5418154003059663,
+      "grad_norm": 0.44803953170776367,
+      "learning_rate": 9.447866637764882e-06,
+      "loss": 0.45416638255119324,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5420703722590515,
+      "grad_norm": 0.4605104327201843,
+      "learning_rate": 9.439445894438896e-06,
+      "loss": 0.4649544656276703,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5423253442121366,
+      "grad_norm": 0.46057993173599243,
+      "learning_rate": 9.431025549829864e-06,
+      "loss": 0.454551637172699,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5425803161652218,
+      "grad_norm": 0.5471037030220032,
+      "learning_rate": 9.422605609927098e-06,
+      "loss": 0.4616963267326355,
+      "step": 2128
+    },
+    {
+      "epoch": 0.542835288118307,
+      "grad_norm": 0.4633654057979584,
+      "learning_rate": 9.414186080719625e-06,
+      "loss": 0.4599413573741913,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5430902600713922,
+      "grad_norm": 0.4843316972255707,
+      "learning_rate": 9.405766968196183e-06,
+      "loss": 0.46958720684051514,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5433452320244773,
+      "grad_norm": 0.4385245442390442,
+      "learning_rate": 9.397348278345201e-06,
+      "loss": 0.4616304337978363,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5436002039775625,
+      "grad_norm": 0.43418222665786743,
+      "learning_rate": 9.388930017154819e-06,
+      "loss": 0.4705386757850647,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5438551759306476,
+      "grad_norm": 0.5231401920318604,
+      "learning_rate": 9.380512190612865e-06,
+      "loss": 0.46048712730407715,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5441101478837328,
+      "grad_norm": 0.44810429215431213,
+      "learning_rate": 9.372094804706867e-06,
+      "loss": 0.45276516675949097,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5443651198368179,
+      "grad_norm": 0.4766870439052582,
+      "learning_rate": 9.36367786542403e-06,
+      "loss": 0.45601996779441833,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5446200917899031,
+      "grad_norm": 0.4559542238712311,
+      "learning_rate": 9.355261378751247e-06,
+      "loss": 0.4878440797328949,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5448750637429882,
+      "grad_norm": 0.43964841961860657,
+      "learning_rate": 9.346845350675087e-06,
+      "loss": 0.4721733033657074,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5451300356960734,
+      "grad_norm": 0.4341309368610382,
+      "learning_rate": 9.338429787181792e-06,
+      "loss": 0.46167072653770447,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5453850076491585,
+      "grad_norm": 0.4392673075199127,
+      "learning_rate": 9.330014694257274e-06,
+      "loss": 0.45932063460350037,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5456399796022438,
+      "grad_norm": 0.43223118782043457,
+      "learning_rate": 9.321600077887111e-06,
+      "loss": 0.4435594081878662,
+      "step": 2140
+    },
+    {
+      "epoch": 0.545894951555329,
+      "grad_norm": 0.4600526988506317,
+      "learning_rate": 9.313185944056541e-06,
+      "loss": 0.4624207019805908,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5461499235084141,
+      "grad_norm": 0.4657096266746521,
+      "learning_rate": 9.304772298750463e-06,
+      "loss": 0.46557554602622986,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5464048954614993,
+      "grad_norm": 0.43160563707351685,
+      "learning_rate": 9.296359147953428e-06,
+      "loss": 0.4483899474143982,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5466598674145844,
+      "grad_norm": 0.43755489587783813,
+      "learning_rate": 9.287946497649621e-06,
+      "loss": 0.46107468008995056,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5469148393676696,
+      "grad_norm": 0.45328959822654724,
+      "learning_rate": 9.27953435382289e-06,
+      "loss": 0.4532812535762787,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5471698113207547,
+      "grad_norm": 0.4669122099876404,
+      "learning_rate": 9.271122722456715e-06,
+      "loss": 0.4514726996421814,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5474247832738399,
+      "grad_norm": 0.42995432019233704,
+      "learning_rate": 9.262711609534211e-06,
+      "loss": 0.4527595639228821,
+      "step": 2147
+    },
+    {
+      "epoch": 0.547679755226925,
+      "grad_norm": 0.46898776292800903,
+      "learning_rate": 9.254301021038125e-06,
+      "loss": 0.4473420977592468,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5479347271800102,
+      "grad_norm": 0.46627387404441833,
+      "learning_rate": 9.245890962950832e-06,
+      "loss": 0.45596832036972046,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5481896991330953,
+      "grad_norm": 0.4662887752056122,
+      "learning_rate": 9.237481441254325e-06,
+      "loss": 0.4651641249656677,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5484446710861806,
+      "grad_norm": 0.44494640827178955,
+      "learning_rate": 9.22907246193022e-06,
+      "loss": 0.45719462633132935,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5486996430392657,
+      "grad_norm": 0.4548644721508026,
+      "learning_rate": 9.22066403095975e-06,
+      "loss": 0.4613267779350281,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5489546149923509,
+      "grad_norm": 0.4366980493068695,
+      "learning_rate": 9.21225615432375e-06,
+      "loss": 0.4484979808330536,
+      "step": 2153
+    },
+    {
+      "epoch": 0.549209586945436,
+      "grad_norm": 0.4651254117488861,
+      "learning_rate": 9.203848838002666e-06,
+      "loss": 0.4629446864128113,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5494645588985212,
+      "grad_norm": 0.4780799150466919,
+      "learning_rate": 9.195442087976547e-06,
+      "loss": 0.4587783217430115,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5497195308516063,
+      "grad_norm": 0.4534330666065216,
+      "learning_rate": 9.187035910225035e-06,
+      "loss": 0.47006481885910034,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5499745028046915,
+      "grad_norm": 0.4560837149620056,
+      "learning_rate": 9.178630310727366e-06,
+      "loss": 0.4496500492095947,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5502294747577766,
+      "grad_norm": 0.4531233310699463,
+      "learning_rate": 9.170225295462364e-06,
+      "loss": 0.4603425860404968,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5504844467108618,
+      "grad_norm": 0.4529958963394165,
+      "learning_rate": 9.16182087040844e-06,
+      "loss": 0.47253990173339844,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5507394186639469,
+      "grad_norm": 0.4396999478340149,
+      "learning_rate": 9.153417041543586e-06,
+      "loss": 0.46572861075401306,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5509943906170321,
+      "grad_norm": 0.45875558257102966,
+      "learning_rate": 9.145013814845363e-06,
+      "loss": 0.453915536403656,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5512493625701173,
+      "grad_norm": 0.45509132742881775,
+      "learning_rate": 9.136611196290915e-06,
+      "loss": 0.4658550024032593,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5515043345232025,
+      "grad_norm": 0.6631203889846802,
+      "learning_rate": 9.12820919185694e-06,
+      "loss": 0.4767513871192932,
+      "step": 2163
+    },
+    {
+      "epoch": 0.5517593064762876,
+      "grad_norm": 0.4426300823688507,
+      "learning_rate": 9.11980780751971e-06,
+      "loss": 0.4608106017112732,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5520142784293728,
+      "grad_norm": 0.517145037651062,
+      "learning_rate": 9.11140704925505e-06,
+      "loss": 0.4579850137233734,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5522692503824579,
+      "grad_norm": 0.4606268107891083,
+      "learning_rate": 9.103006923038344e-06,
+      "loss": 0.4567311704158783,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5525242223355431,
+      "grad_norm": 0.46356263756752014,
+      "learning_rate": 9.094607434844523e-06,
+      "loss": 0.4575386047363281,
+      "step": 2167
+    },
+    {
+      "epoch": 0.5527791942886282,
+      "grad_norm": 0.4374256134033203,
+      "learning_rate": 9.086208590648066e-06,
+      "loss": 0.4539230763912201,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5530341662417134,
+      "grad_norm": 0.439361572265625,
+      "learning_rate": 9.077810396422995e-06,
+      "loss": 0.47697627544403076,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5532891381947985,
+      "grad_norm": 0.4530436396598816,
+      "learning_rate": 9.069412858142863e-06,
+      "loss": 0.4649924039840698,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5535441101478837,
+      "grad_norm": 0.4412326514720917,
+      "learning_rate": 9.061015981780764e-06,
+      "loss": 0.46756356954574585,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5537990821009688,
+      "grad_norm": 0.440147340297699,
+      "learning_rate": 9.052619773309318e-06,
+      "loss": 0.44795387983322144,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5540540540540541,
+      "grad_norm": 0.44182321429252625,
+      "learning_rate": 9.044224238700672e-06,
+      "loss": 0.4552813470363617,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5543090260071392,
+      "grad_norm": 0.544137179851532,
+      "learning_rate": 9.03582938392649e-06,
+      "loss": 0.4627988338470459,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5545639979602244,
+      "grad_norm": 0.4592606723308563,
+      "learning_rate": 9.02743521495796e-06,
+      "loss": 0.4631474018096924,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5548189699133096,
+      "grad_norm": 0.4471321403980255,
+      "learning_rate": 9.019041737765768e-06,
+      "loss": 0.4676770567893982,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5550739418663947,
+      "grad_norm": 0.48931652307510376,
+      "learning_rate": 9.010648958320121e-06,
+      "loss": 0.4544360041618347,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5553289138194799,
+      "grad_norm": 0.4572511315345764,
+      "learning_rate": 9.002256882590728e-06,
+      "loss": 0.46220335364341736,
+      "step": 2178
+    },
+    {
+      "epoch": 0.555583885772565,
+      "grad_norm": 0.43627431988716125,
+      "learning_rate": 8.993865516546791e-06,
+      "loss": 0.45467087626457214,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5558388577256502,
+      "grad_norm": 0.4847698211669922,
+      "learning_rate": 8.985474866157014e-06,
+      "loss": 0.4593415856361389,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5560938296787353,
+      "grad_norm": 0.5630824565887451,
+      "learning_rate": 8.977084937389591e-06,
+      "loss": 0.4610545337200165,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5563488016318205,
+      "grad_norm": 0.44593602418899536,
+      "learning_rate": 8.968695736212194e-06,
+      "loss": 0.46437281370162964,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5566037735849056,
+      "grad_norm": 0.45797938108444214,
+      "learning_rate": 8.960307268591988e-06,
+      "loss": 0.4567159116268158,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5568587455379909,
+      "grad_norm": 0.4415434002876282,
+      "learning_rate": 8.951919540495613e-06,
+      "loss": 0.4509863555431366,
+      "step": 2184
+    },
+    {
+      "epoch": 0.557113717491076,
+      "grad_norm": 0.4689483344554901,
+      "learning_rate": 8.943532557889182e-06,
+      "loss": 0.4577532708644867,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5573686894441612,
+      "grad_norm": 0.44582420587539673,
+      "learning_rate": 8.935146326738275e-06,
+      "loss": 0.44930100440979004,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5576236613972463,
+      "grad_norm": 0.4341600835323334,
+      "learning_rate": 8.926760853007946e-06,
+      "loss": 0.459422767162323,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5578786333503315,
+      "grad_norm": 0.4593459367752075,
+      "learning_rate": 8.918376142662709e-06,
+      "loss": 0.47731637954711914,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5581336053034166,
+      "grad_norm": 0.49128293991088867,
+      "learning_rate": 8.909992201666516e-06,
+      "loss": 0.44464847445487976,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5583885772565018,
+      "grad_norm": 0.4623543918132782,
+      "learning_rate": 8.901609035982799e-06,
+      "loss": 0.4652901589870453,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5586435492095869,
+      "grad_norm": 0.4317094683647156,
+      "learning_rate": 8.893226651574422e-06,
+      "loss": 0.46675562858581543,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5588985211626721,
+      "grad_norm": 0.43534040451049805,
+      "learning_rate": 8.8848450544037e-06,
+      "loss": 0.46179577708244324,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5591534931157572,
+      "grad_norm": 0.4367939233779907,
+      "learning_rate": 8.876464250432381e-06,
+      "loss": 0.4616091251373291,
+      "step": 2193
+    },
+    {
+      "epoch": 0.5594084650688425,
+      "grad_norm": 0.4712008535861969,
+      "learning_rate": 8.86808424562166e-06,
+      "loss": 0.45911508798599243,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5596634370219276,
+      "grad_norm": 0.4427134096622467,
+      "learning_rate": 8.859705045932152e-06,
+      "loss": 0.4717683792114258,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5599184089750128,
+      "grad_norm": 0.43177637457847595,
+      "learning_rate": 8.851326657323904e-06,
+      "loss": 0.4647178053855896,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5601733809280979,
+      "grad_norm": 0.5126813650131226,
+      "learning_rate": 8.842949085756389e-06,
+      "loss": 0.4650019407272339,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5604283528811831,
+      "grad_norm": 0.4845443069934845,
+      "learning_rate": 8.834572337188494e-06,
+      "loss": 0.4632229804992676,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5606833248342682,
+      "grad_norm": 0.4437451958656311,
+      "learning_rate": 8.826196417578523e-06,
+      "loss": 0.46996474266052246,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5609382967873534,
+      "grad_norm": 0.4660290777683258,
+      "learning_rate": 8.817821332884194e-06,
+      "loss": 0.4538179337978363,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5611932687404385,
+      "grad_norm": 0.48204517364501953,
+      "learning_rate": 8.809447089062625e-06,
+      "loss": 0.4647831320762634,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5614482406935237,
+      "grad_norm": 0.4758976399898529,
+      "learning_rate": 8.801073692070337e-06,
+      "loss": 0.45339640974998474,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5617032126466088,
+      "grad_norm": 0.45573097467422485,
+      "learning_rate": 8.792701147863249e-06,
+      "loss": 0.46297213435173035,
+      "step": 2203
+    },
+    {
+      "epoch": 0.561958184599694,
+      "grad_norm": 0.4524531662464142,
+      "learning_rate": 8.784329462396678e-06,
+      "loss": 0.45312342047691345,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5622131565527791,
+      "grad_norm": 0.47178715467453003,
+      "learning_rate": 8.775958641625323e-06,
+      "loss": 0.4536016881465912,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5624681285058644,
+      "grad_norm": 0.4592845141887665,
+      "learning_rate": 8.767588691503271e-06,
+      "loss": 0.46670663356781006,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5627231004589495,
+      "grad_norm": 0.453191339969635,
+      "learning_rate": 8.759219617984e-06,
+      "loss": 0.4639744162559509,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5629780724120347,
+      "grad_norm": 0.4972878694534302,
+      "learning_rate": 8.750851427020341e-06,
+      "loss": 0.46717357635498047,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5632330443651198,
+      "grad_norm": 0.4795783758163452,
+      "learning_rate": 8.742484124564514e-06,
+      "loss": 0.45974206924438477,
+      "step": 2209
+    },
+    {
+      "epoch": 0.563488016318205,
+      "grad_norm": 0.43076789379119873,
+      "learning_rate": 8.73411771656811e-06,
+      "loss": 0.446453332901001,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5637429882712902,
+      "grad_norm": 0.4527166783809662,
+      "learning_rate": 8.725752208982074e-06,
+      "loss": 0.47003960609436035,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5639979602243753,
+      "grad_norm": 0.4806315004825592,
+      "learning_rate": 8.717387607756714e-06,
+      "loss": 0.46837812662124634,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5642529321774605,
+      "grad_norm": 0.44723406434059143,
+      "learning_rate": 8.709023918841695e-06,
+      "loss": 0.4666522443294525,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5645079041305456,
+      "grad_norm": 0.46346336603164673,
+      "learning_rate": 8.70066114818603e-06,
+      "loss": 0.44260507822036743,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5647628760836308,
+      "grad_norm": 0.46325960755348206,
+      "learning_rate": 8.692299301738081e-06,
+      "loss": 0.46868035197257996,
+      "step": 2215
+    },
+    {
+      "epoch": 0.565017848036716,
+      "grad_norm": 0.4721716046333313,
+      "learning_rate": 8.683938385445552e-06,
+      "loss": 0.4612208604812622,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5652728199898012,
+      "grad_norm": 0.4544331133365631,
+      "learning_rate": 8.675578405255485e-06,
+      "loss": 0.47177284955978394,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5655277919428863,
+      "grad_norm": 0.4448011815547943,
+      "learning_rate": 8.667219367114257e-06,
+      "loss": 0.45070117712020874,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5657827638959715,
+      "grad_norm": 0.43457093834877014,
+      "learning_rate": 8.658861276967574e-06,
+      "loss": 0.4596124291419983,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5660377358490566,
+      "grad_norm": 0.5355086326599121,
+      "learning_rate": 8.650504140760474e-06,
+      "loss": 0.4691835343837738,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5662927078021418,
+      "grad_norm": 0.4600619971752167,
+      "learning_rate": 8.642147964437303e-06,
+      "loss": 0.45449429750442505,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5665476797552269,
+      "grad_norm": 0.48802223801612854,
+      "learning_rate": 8.633792753941733e-06,
+      "loss": 0.46611183881759644,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5668026517083121,
+      "grad_norm": 0.4627108573913574,
+      "learning_rate": 8.625438515216751e-06,
+      "loss": 0.47132524847984314,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5670576236613972,
+      "grad_norm": 0.4547104835510254,
+      "learning_rate": 8.617085254204648e-06,
+      "loss": 0.4444102644920349,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5673125956144824,
+      "grad_norm": 0.4633663594722748,
+      "learning_rate": 8.608732976847025e-06,
+      "loss": 0.4674640893936157,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5675675675675675,
+      "grad_norm": 0.44347018003463745,
+      "learning_rate": 8.600381689084775e-06,
+      "loss": 0.46025699377059937,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5678225395206528,
+      "grad_norm": 0.4436300992965698,
+      "learning_rate": 8.592031396858094e-06,
+      "loss": 0.45944809913635254,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5680775114737379,
+      "grad_norm": 0.45730459690093994,
+      "learning_rate": 8.583682106106467e-06,
+      "loss": 0.4482569098472595,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5683324834268231,
+      "grad_norm": 0.4409254491329193,
+      "learning_rate": 8.575333822768666e-06,
+      "loss": 0.47202134132385254,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5685874553799082,
+      "grad_norm": 0.46967801451683044,
+      "learning_rate": 8.566986552782747e-06,
+      "loss": 0.4568798542022705,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5688424273329934,
+      "grad_norm": 0.4574753940105438,
+      "learning_rate": 8.558640302086053e-06,
+      "loss": 0.45431026816368103,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5690973992860785,
+      "grad_norm": 0.45594164729118347,
+      "learning_rate": 8.550295076615189e-06,
+      "loss": 0.4549804627895355,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5693523712391637,
+      "grad_norm": 0.4559899866580963,
+      "learning_rate": 8.541950882306039e-06,
+      "loss": 0.4558350443840027,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5696073431922488,
+      "grad_norm": 0.46322691440582275,
+      "learning_rate": 8.533607725093749e-06,
+      "loss": 0.45330411195755005,
+      "step": 2234
+    },
+    {
+      "epoch": 0.569862315145334,
+      "grad_norm": 0.519966185092926,
+      "learning_rate": 8.52526561091273e-06,
+      "loss": 0.47055038809776306,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5701172870984191,
+      "grad_norm": 0.4469245970249176,
+      "learning_rate": 8.516924545696653e-06,
+      "loss": 0.4623914361000061,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5703722590515043,
+      "grad_norm": 0.48718273639678955,
+      "learning_rate": 8.50858453537844e-06,
+      "loss": 0.4567593038082123,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5706272310045895,
+      "grad_norm": 0.4596962034702301,
+      "learning_rate": 8.50024558589026e-06,
+      "loss": 0.4584554433822632,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5708822029576747,
+      "grad_norm": 0.4492977261543274,
+      "learning_rate": 8.491907703163537e-06,
+      "loss": 0.45537814497947693,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5711371749107598,
+      "grad_norm": 0.45445308089256287,
+      "learning_rate": 8.483570893128923e-06,
+      "loss": 0.4571847915649414,
+      "step": 2240
+    },
+    {
+      "epoch": 0.571392146863845,
+      "grad_norm": 0.4287809133529663,
+      "learning_rate": 8.475235161716317e-06,
+      "loss": 0.4541972577571869,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5716471188169301,
+      "grad_norm": 0.5193936824798584,
+      "learning_rate": 8.466900514854847e-06,
+      "loss": 0.44935959577560425,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5719020907700153,
+      "grad_norm": 0.4742075204849243,
+      "learning_rate": 8.458566958472872e-06,
+      "loss": 0.45262664556503296,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5721570627231004,
+      "grad_norm": 0.42204076051712036,
+      "learning_rate": 8.450234498497972e-06,
+      "loss": 0.46552449464797974,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5724120346761856,
+      "grad_norm": 0.44148775935173035,
+      "learning_rate": 8.441903140856954e-06,
+      "loss": 0.4528549611568451,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5726670066292707,
+      "grad_norm": 0.4461648166179657,
+      "learning_rate": 8.433572891475826e-06,
+      "loss": 0.45218607783317566,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5729219785823559,
+      "grad_norm": 0.4426140785217285,
+      "learning_rate": 8.425243756279824e-06,
+      "loss": 0.46453770995140076,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5731769505354412,
+      "grad_norm": 0.43527477979660034,
+      "learning_rate": 8.416915741193384e-06,
+      "loss": 0.46211183071136475,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5734319224885263,
+      "grad_norm": 0.4312806725502014,
+      "learning_rate": 8.408588852140144e-06,
+      "loss": 0.47130507230758667,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5736868944416115,
+      "grad_norm": 0.43725165724754333,
+      "learning_rate": 8.400263095042945e-06,
+      "loss": 0.4598652720451355,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5739418663946966,
+      "grad_norm": 0.45928505063056946,
+      "learning_rate": 8.39193847582382e-06,
+      "loss": 0.444231241941452,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5741968383477818,
+      "grad_norm": 0.4414902627468109,
+      "learning_rate": 8.383615000403999e-06,
+      "loss": 0.4602000117301941,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5744518103008669,
+      "grad_norm": 0.4411165118217468,
+      "learning_rate": 8.375292674703885e-06,
+      "loss": 0.4520033299922943,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5747067822539521,
+      "grad_norm": 0.4370625913143158,
+      "learning_rate": 8.366971504643074e-06,
+      "loss": 0.45016974210739136,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5749617542070372,
+      "grad_norm": 0.4558812379837036,
+      "learning_rate": 8.35865149614034e-06,
+      "loss": 0.4486446976661682,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5752167261601224,
+      "grad_norm": 0.4235498607158661,
+      "learning_rate": 8.350332655113626e-06,
+      "loss": 0.45743364095687866,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5754716981132075,
+      "grad_norm": 0.4559319019317627,
+      "learning_rate": 8.342014987480047e-06,
+      "loss": 0.45270293951034546,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5757266700662927,
+      "grad_norm": 0.44990503787994385,
+      "learning_rate": 8.333698499155886e-06,
+      "loss": 0.45790648460388184,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5759816420193778,
+      "grad_norm": 0.4234278202056885,
+      "learning_rate": 8.32538319605658e-06,
+      "loss": 0.4859004616737366,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5762366139724631,
+      "grad_norm": 0.44142118096351624,
+      "learning_rate": 8.31706908409673e-06,
+      "loss": 0.46571290493011475,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5764915859255482,
+      "grad_norm": 0.4531867504119873,
+      "learning_rate": 8.308756169190083e-06,
+      "loss": 0.45635876059532166,
+      "step": 2261
+    },
+    {
+      "epoch": 0.5767465578786334,
+      "grad_norm": 0.4511576294898987,
+      "learning_rate": 8.300444457249544e-06,
+      "loss": 0.4581971764564514,
+      "step": 2262
+    },
+    {
+      "epoch": 0.5770015298317185,
+      "grad_norm": 0.43757081031799316,
+      "learning_rate": 8.292133954187148e-06,
+      "loss": 0.46797725558280945,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5772565017848037,
+      "grad_norm": 0.47690480947494507,
+      "learning_rate": 8.283824665914097e-06,
+      "loss": 0.4525216519832611,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5775114737378888,
+      "grad_norm": 0.44798848032951355,
+      "learning_rate": 8.275516598340692e-06,
+      "loss": 0.4591936767101288,
+      "step": 2265
+    },
+    {
+      "epoch": 0.577766445690974,
+      "grad_norm": 0.45524078607559204,
+      "learning_rate": 8.267209757376391e-06,
+      "loss": 0.46707552671432495,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5780214176440591,
+      "grad_norm": 0.9517451524734497,
+      "learning_rate": 8.258904148929775e-06,
+      "loss": 0.4510538578033447,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5782763895971443,
+      "grad_norm": 0.4590145945549011,
+      "learning_rate": 8.250599778908546e-06,
+      "loss": 0.46022161841392517,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5785313615502294,
+      "grad_norm": 0.49070027470588684,
+      "learning_rate": 8.242296653219527e-06,
+      "loss": 0.45525944232940674,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5787863335033147,
+      "grad_norm": 0.44155219197273254,
+      "learning_rate": 8.23399477776865e-06,
+      "loss": 0.45879992842674255,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5790413054563998,
+      "grad_norm": 0.4478846788406372,
+      "learning_rate": 8.225694158460969e-06,
+      "loss": 0.4742863178253174,
+      "step": 2271
+    },
+    {
+      "epoch": 0.579296277409485,
+      "grad_norm": 0.4330812394618988,
+      "learning_rate": 8.217394801200632e-06,
+      "loss": 0.4589769244194031,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5795512493625701,
+      "grad_norm": 0.44530272483825684,
+      "learning_rate": 8.209096711890897e-06,
+      "loss": 0.46048784255981445,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5798062213156553,
+      "grad_norm": 0.44956183433532715,
+      "learning_rate": 8.200799896434116e-06,
+      "loss": 0.47708839178085327,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5800611932687404,
+      "grad_norm": 0.43646925687789917,
+      "learning_rate": 8.192504360731741e-06,
+      "loss": 0.4574781358242035,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5803161652218256,
+      "grad_norm": 0.4399811625480652,
+      "learning_rate": 8.18421011068431e-06,
+      "loss": 0.4625152349472046,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5805711371749107,
+      "grad_norm": 0.4326459765434265,
+      "learning_rate": 8.175917152191447e-06,
+      "loss": 0.4627325236797333,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5808261091279959,
+      "grad_norm": 0.4582973122596741,
+      "learning_rate": 8.167625491151849e-06,
+      "loss": 0.45126891136169434,
+      "step": 2278
+    },
+    {
+      "epoch": 0.581081081081081,
+      "grad_norm": 0.4950394928455353,
+      "learning_rate": 8.159335133463306e-06,
+      "loss": 0.45725926756858826,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5813360530341662,
+      "grad_norm": 0.43344494700431824,
+      "learning_rate": 8.151046085022668e-06,
+      "loss": 0.45229387283325195,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5815910249872513,
+      "grad_norm": 0.4514404833316803,
+      "learning_rate": 8.142758351725862e-06,
+      "loss": 0.4698212146759033,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5818459969403366,
+      "grad_norm": 0.46291786432266235,
+      "learning_rate": 8.134471939467874e-06,
+      "loss": 0.4666634500026703,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5821009688934218,
+      "grad_norm": 0.4638640582561493,
+      "learning_rate": 8.126186854142752e-06,
+      "loss": 0.4487994611263275,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5823559408465069,
+      "grad_norm": 0.44678041338920593,
+      "learning_rate": 8.117903101643609e-06,
+      "loss": 0.4675861597061157,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5826109127995921,
+      "grad_norm": 0.45098763704299927,
+      "learning_rate": 8.109620687862587e-06,
+      "loss": 0.46308618783950806,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5828658847526772,
+      "grad_norm": 0.43600428104400635,
+      "learning_rate": 8.1013396186909e-06,
+      "loss": 0.4676024317741394,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5831208567057624,
+      "grad_norm": 0.5147649645805359,
+      "learning_rate": 8.093059900018793e-06,
+      "loss": 0.4530147314071655,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5833758286588475,
+      "grad_norm": 0.4465515613555908,
+      "learning_rate": 8.084781537735554e-06,
+      "loss": 0.43942782282829285,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5836308006119327,
+      "grad_norm": 0.478004515171051,
+      "learning_rate": 8.076504537729506e-06,
+      "loss": 0.45939934253692627,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5838857725650178,
+      "grad_norm": 0.4839225113391876,
+      "learning_rate": 8.068228905888005e-06,
+      "loss": 0.4545905292034149,
+      "step": 2290
+    },
+    {
+      "epoch": 0.584140744518103,
+      "grad_norm": 0.45244279503822327,
+      "learning_rate": 8.059954648097427e-06,
+      "loss": 0.4498525857925415,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5843957164711882,
+      "grad_norm": 0.46127891540527344,
+      "learning_rate": 8.051681770243176e-06,
+      "loss": 0.4581502676010132,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5846506884242734,
+      "grad_norm": 0.467270165681839,
+      "learning_rate": 8.043410278209673e-06,
+      "loss": 0.470907062292099,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5849056603773585,
+      "grad_norm": 0.442678302526474,
+      "learning_rate": 8.035140177880358e-06,
+      "loss": 0.4593364894390106,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5851606323304437,
+      "grad_norm": 0.4396481513977051,
+      "learning_rate": 8.02687147513767e-06,
+      "loss": 0.45607924461364746,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5854156042835288,
+      "grad_norm": 0.4411541819572449,
+      "learning_rate": 8.018604175863072e-06,
+      "loss": 0.45380541682243347,
+      "step": 2296
+    },
+    {
+      "epoch": 0.585670576236614,
+      "grad_norm": 0.4444219470024109,
+      "learning_rate": 8.010338285937006e-06,
+      "loss": 0.4425385892391205,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5859255481896991,
+      "grad_norm": 0.47013190388679504,
+      "learning_rate": 8.002073811238926e-06,
+      "loss": 0.46726760268211365,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5861805201427843,
+      "grad_norm": 0.42536959052085876,
+      "learning_rate": 7.99381075764728e-06,
+      "loss": 0.4520610272884369,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5864354920958694,
+      "grad_norm": 0.45570358633995056,
+      "learning_rate": 7.9855491310395e-06,
+      "loss": 0.4546683728694916,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5866904640489546,
+      "grad_norm": 0.44183358550071716,
+      "learning_rate": 7.977288937292003e-06,
+      "loss": 0.4454396665096283,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5869454360020397,
+      "grad_norm": 0.46111831068992615,
+      "learning_rate": 7.969030182280192e-06,
+      "loss": 0.4599505364894867,
+      "step": 2302
+    },
+    {
+      "epoch": 0.587200407955125,
+      "grad_norm": 0.4493691027164459,
+      "learning_rate": 7.960772871878441e-06,
+      "loss": 0.444584459066391,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5874553799082101,
+      "grad_norm": 0.4369162917137146,
+      "learning_rate": 7.952517011960099e-06,
+      "loss": 0.4446493089199066,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5877103518612953,
+      "grad_norm": 0.4527932405471802,
+      "learning_rate": 7.94426260839748e-06,
+      "loss": 0.4618287980556488,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5879653238143804,
+      "grad_norm": 0.47581008076667786,
+      "learning_rate": 7.936009667061868e-06,
+      "loss": 0.4562625288963318,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5882202957674656,
+      "grad_norm": 0.44282785058021545,
+      "learning_rate": 7.927758193823502e-06,
+      "loss": 0.46942514181137085,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5884752677205507,
+      "grad_norm": 0.43578222393989563,
+      "learning_rate": 7.91950819455158e-06,
+      "loss": 0.44949501752853394,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5887302396736359,
+      "grad_norm": 0.4501938819885254,
+      "learning_rate": 7.911259675114257e-06,
+      "loss": 0.4654654264450073,
+      "step": 2309
+    },
+    {
+      "epoch": 0.588985211626721,
+      "grad_norm": 0.4405452609062195,
+      "learning_rate": 7.903012641378613e-06,
+      "loss": 0.4545036256313324,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5892401835798062,
+      "grad_norm": 0.5059425234794617,
+      "learning_rate": 7.8947670992107e-06,
+      "loss": 0.46662160754203796,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5894951555328913,
+      "grad_norm": 0.4511828124523163,
+      "learning_rate": 7.88652305447549e-06,
+      "loss": 0.4749602675437927,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5897501274859765,
+      "grad_norm": 0.4501926600933075,
+      "learning_rate": 7.878280513036898e-06,
+      "loss": 0.44702601432800293,
+      "step": 2313
+    },
+    {
+      "epoch": 0.5900050994390617,
+      "grad_norm": 0.4389598071575165,
+      "learning_rate": 7.870039480757767e-06,
+      "loss": 0.4614280164241791,
+      "step": 2314
+    },
+    {
+      "epoch": 0.5902600713921469,
+      "grad_norm": 0.44306790828704834,
+      "learning_rate": 7.861799963499866e-06,
+      "loss": 0.45800089836120605,
+      "step": 2315
+    },
+    {
+      "epoch": 0.590515043345232,
+      "grad_norm": 0.4772992432117462,
+      "learning_rate": 7.853561967123892e-06,
+      "loss": 0.47122466564178467,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5907700152983172,
+      "grad_norm": 0.46096673607826233,
+      "learning_rate": 7.84532549748945e-06,
+      "loss": 0.45394206047058105,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5910249872514024,
+      "grad_norm": 0.4446107745170593,
+      "learning_rate": 7.837090560455065e-06,
+      "loss": 0.4603210389614105,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5912799592044875,
+      "grad_norm": 0.5345133543014526,
+      "learning_rate": 7.828857161878172e-06,
+      "loss": 0.4523944854736328,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5915349311575727,
+      "grad_norm": 0.4459843635559082,
+      "learning_rate": 7.820625307615117e-06,
+      "loss": 0.4598017632961273,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5917899031106578,
+      "grad_norm": 0.4400486350059509,
+      "learning_rate": 7.812395003521136e-06,
+      "loss": 0.4436286389827728,
+      "step": 2321
+    },
+    {
+      "epoch": 0.592044875063743,
+      "grad_norm": 0.47732704877853394,
+      "learning_rate": 7.804166255450372e-06,
+      "loss": 0.44466397166252136,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5922998470168281,
+      "grad_norm": 0.4303358793258667,
+      "learning_rate": 7.795939069255853e-06,
+      "loss": 0.46469664573669434,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5925548189699134,
+      "grad_norm": 0.4471558928489685,
+      "learning_rate": 7.787713450789505e-06,
+      "loss": 0.46260079741477966,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5928097909229985,
+      "grad_norm": 0.43670353293418884,
+      "learning_rate": 7.77948940590213e-06,
+      "loss": 0.44456517696380615,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5930647628760837,
+      "grad_norm": 0.4920576512813568,
+      "learning_rate": 7.771266940443419e-06,
+      "loss": 0.45686396956443787,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5933197348291688,
+      "grad_norm": 0.5350610017776489,
+      "learning_rate": 7.763046060261931e-06,
+      "loss": 0.46714290976524353,
+      "step": 2327
+    },
+    {
+      "epoch": 0.593574706782254,
+      "grad_norm": 0.428666889667511,
+      "learning_rate": 7.754826771205111e-06,
+      "loss": 0.454071044921875,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5938296787353391,
+      "grad_norm": 0.617946982383728,
+      "learning_rate": 7.746609079119255e-06,
+      "loss": 0.4507453143596649,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5940846506884243,
+      "grad_norm": 0.46272704005241394,
+      "learning_rate": 7.738392989849531e-06,
+      "loss": 0.4464974105358124,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5943396226415094,
+      "grad_norm": 0.4412076771259308,
+      "learning_rate": 7.73017850923997e-06,
+      "loss": 0.4614940285682678,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5945945945945946,
+      "grad_norm": 0.4551214575767517,
+      "learning_rate": 7.721965643133458e-06,
+      "loss": 0.45627376437187195,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5948495665476797,
+      "grad_norm": 0.45512428879737854,
+      "learning_rate": 7.713754397371729e-06,
+      "loss": 0.463383287191391,
+      "step": 2333
+    },
+    {
+      "epoch": 0.595104538500765,
+      "grad_norm": 0.47260189056396484,
+      "learning_rate": 7.705544777795368e-06,
+      "loss": 0.45705240964889526,
+      "step": 2334
+    },
+    {
+      "epoch": 0.59535951045385,
+      "grad_norm": 0.4583861231803894,
+      "learning_rate": 7.6973367902438e-06,
+      "loss": 0.47041448950767517,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5956144824069353,
+      "grad_norm": 0.42344579100608826,
+      "learning_rate": 7.68913044055529e-06,
+      "loss": 0.4356708526611328,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5958694543600204,
+      "grad_norm": 0.47228240966796875,
+      "learning_rate": 7.680925734566938e-06,
+      "loss": 0.46566927433013916,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5961244263131056,
+      "grad_norm": 0.44690245389938354,
+      "learning_rate": 7.672722678114676e-06,
+      "loss": 0.46312588453292847,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5963793982661907,
+      "grad_norm": 0.45382794737815857,
+      "learning_rate": 7.664521277033264e-06,
+      "loss": 0.47106119990348816,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5966343702192759,
+      "grad_norm": 0.47584131360054016,
+      "learning_rate": 7.656321537156281e-06,
+      "loss": 0.4431549906730652,
+      "step": 2340
+    },
+    {
+      "epoch": 0.596889342172361,
+      "grad_norm": 0.4524940252304077,
+      "learning_rate": 7.648123464316133e-06,
+      "loss": 0.4706418216228485,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5971443141254462,
+      "grad_norm": 0.4546431303024292,
+      "learning_rate": 7.639927064344023e-06,
+      "loss": 0.4647432565689087,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5973992860785313,
+      "grad_norm": 0.47987425327301025,
+      "learning_rate": 7.631732343069976e-06,
+      "loss": 0.4546689987182617,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5976542580316165,
+      "grad_norm": 0.45354241132736206,
+      "learning_rate": 7.623539306322827e-06,
+      "loss": 0.4678219258785248,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5979092299847016,
+      "grad_norm": 0.4847342073917389,
+      "learning_rate": 7.615347959930206e-06,
+      "loss": 0.4393325746059418,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5981642019377869,
+      "grad_norm": 0.4791020452976227,
+      "learning_rate": 7.607158309718539e-06,
+      "loss": 0.4542630612850189,
+      "step": 2346
+    },
+    {
+      "epoch": 0.598419173890872,
+      "grad_norm": 0.458964079618454,
+      "learning_rate": 7.598970361513052e-06,
+      "loss": 0.4545365273952484,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5986741458439572,
+      "grad_norm": 0.4871685206890106,
+      "learning_rate": 7.590784121137755e-06,
+      "loss": 0.4562395513057709,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5989291177970423,
+      "grad_norm": 0.46031686663627625,
+      "learning_rate": 7.582599594415445e-06,
+      "loss": 0.44557708501815796,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5991840897501275,
+      "grad_norm": 0.4532771706581116,
+      "learning_rate": 7.5744167871677e-06,
+      "loss": 0.4576115608215332,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5994390617032126,
+      "grad_norm": 0.43669387698173523,
+      "learning_rate": 7.5662357052148735e-06,
+      "loss": 0.46957695484161377,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5996940336562978,
+      "grad_norm": 0.4384315311908722,
+      "learning_rate": 7.558056354376097e-06,
+      "loss": 0.4530205726623535,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5999490056093829,
+      "grad_norm": 0.49745088815689087,
+      "learning_rate": 7.549878740469267e-06,
+      "loss": 0.45755571126937866,
+      "step": 2353
+    },
+    {
+      "epoch": 0.6002039775624681,
+      "grad_norm": 0.4505481719970703,
+      "learning_rate": 7.541702869311047e-06,
+      "loss": 0.4640122354030609,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6004589495155533,
+      "grad_norm": 0.5628259778022766,
+      "learning_rate": 7.53352874671685e-06,
+      "loss": 0.4595435857772827,
+      "step": 2355
+    },
+    {
+      "epoch": 0.6007139214686384,
+      "grad_norm": 0.47558727860450745,
+      "learning_rate": 7.52535637850086e-06,
+      "loss": 0.46309083700180054,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6009688934217237,
+      "grad_norm": 0.44385313987731934,
+      "learning_rate": 7.517185770476005e-06,
+      "loss": 0.4619719386100769,
+      "step": 2357
+    },
+    {
+      "epoch": 0.6012238653748088,
+      "grad_norm": 0.4565849304199219,
+      "learning_rate": 7.509016928453966e-06,
+      "loss": 0.46770864725112915,
+      "step": 2358
+    },
+    {
+      "epoch": 0.601478837327894,
+      "grad_norm": 0.4452267587184906,
+      "learning_rate": 7.5008498582451615e-06,
+      "loss": 0.4678848385810852,
+      "step": 2359
+    },
+    {
+      "epoch": 0.6017338092809791,
+      "grad_norm": 0.4667193293571472,
+      "learning_rate": 7.492684565658756e-06,
+      "loss": 0.4526401460170746,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6019887812340643,
+      "grad_norm": 0.46906158328056335,
+      "learning_rate": 7.484521056502644e-06,
+      "loss": 0.4547154903411865,
+      "step": 2361
+    },
+    {
+      "epoch": 0.6022437531871494,
+      "grad_norm": 0.4575459361076355,
+      "learning_rate": 7.476359336583454e-06,
+      "loss": 0.455719530582428,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6024987251402346,
+      "grad_norm": 0.45700007677078247,
+      "learning_rate": 7.46819941170654e-06,
+      "loss": 0.4421968460083008,
+      "step": 2363
+    },
+    {
+      "epoch": 0.6027536970933197,
+      "grad_norm": 0.4708902835845947,
+      "learning_rate": 7.4600412876759874e-06,
+      "loss": 0.45173487067222595,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6030086690464049,
+      "grad_norm": 0.4684321880340576,
+      "learning_rate": 7.451884970294591e-06,
+      "loss": 0.4693717360496521,
+      "step": 2365
+    },
+    {
+      "epoch": 0.60326364099949,
+      "grad_norm": 0.4727350175380707,
+      "learning_rate": 7.443730465363865e-06,
+      "loss": 0.4616132378578186,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6035186129525753,
+      "grad_norm": 0.43844127655029297,
+      "learning_rate": 7.435577778684033e-06,
+      "loss": 0.4775257110595703,
+      "step": 2367
+    },
+    {
+      "epoch": 0.6037735849056604,
+      "grad_norm": 0.4586241543292999,
+      "learning_rate": 7.427426916054025e-06,
+      "loss": 0.4586012363433838,
+      "step": 2368
+    },
+    {
+      "epoch": 0.6040285568587456,
+      "grad_norm": 0.4408634305000305,
+      "learning_rate": 7.419277883271473e-06,
+      "loss": 0.4525206983089447,
+      "step": 2369
+    },
+    {
+      "epoch": 0.6042835288118307,
+      "grad_norm": 0.4376678168773651,
+      "learning_rate": 7.41113068613271e-06,
+      "loss": 0.4567801356315613,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6045385007649159,
+      "grad_norm": 0.4777446389198303,
+      "learning_rate": 7.4029853304327636e-06,
+      "loss": 0.4473497271537781,
+      "step": 2371
+    },
+    {
+      "epoch": 0.604793472718001,
+      "grad_norm": 0.7311729192733765,
+      "learning_rate": 7.394841821965345e-06,
+      "loss": 0.4430292546749115,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6050484446710862,
+      "grad_norm": 0.4651448726654053,
+      "learning_rate": 7.3867001665228666e-06,
+      "loss": 0.45642799139022827,
+      "step": 2373
+    },
+    {
+      "epoch": 0.6053034166241713,
+      "grad_norm": 0.49951180815696716,
+      "learning_rate": 7.378560369896403e-06,
+      "loss": 0.4494733214378357,
+      "step": 2374
+    },
+    {
+      "epoch": 0.6055583885772565,
+      "grad_norm": 0.431007444858551,
+      "learning_rate": 7.370422437875716e-06,
+      "loss": 0.4578118324279785,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6058133605303416,
+      "grad_norm": 0.44983765482902527,
+      "learning_rate": 7.362286376249248e-06,
+      "loss": 0.4628801941871643,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6060683324834268,
+      "grad_norm": 0.44235482811927795,
+      "learning_rate": 7.3541521908041005e-06,
+      "loss": 0.453955739736557,
+      "step": 2377
+    },
+    {
+      "epoch": 0.606323304436512,
+      "grad_norm": 0.4479529857635498,
+      "learning_rate": 7.346019887326046e-06,
+      "loss": 0.4570160508155823,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6065782763895972,
+      "grad_norm": 0.44753244519233704,
+      "learning_rate": 7.337889471599517e-06,
+      "loss": 0.4476889371871948,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6068332483426823,
+      "grad_norm": 0.4472813904285431,
+      "learning_rate": 7.3297609494076e-06,
+      "loss": 0.4456092119216919,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6070882202957675,
+      "grad_norm": 0.43575942516326904,
+      "learning_rate": 7.3216343265320414e-06,
+      "loss": 0.45226651430130005,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6073431922488526,
+      "grad_norm": 0.4231815040111542,
+      "learning_rate": 7.313509608753231e-06,
+      "loss": 0.45405086874961853,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6075981642019378,
+      "grad_norm": 0.4461728632450104,
+      "learning_rate": 7.305386801850205e-06,
+      "loss": 0.45775744318962097,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6078531361550229,
+      "grad_norm": 0.4281957149505615,
+      "learning_rate": 7.29726591160064e-06,
+      "loss": 0.458732932806015,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6081081081081081,
+      "grad_norm": 0.44112592935562134,
+      "learning_rate": 7.289146943780853e-06,
+      "loss": 0.4626714587211609,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6083630800611932,
+      "grad_norm": 0.4187681972980499,
+      "learning_rate": 7.281029904165795e-06,
+      "loss": 0.45753440260887146,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6086180520142784,
+      "grad_norm": 0.43930917978286743,
+      "learning_rate": 7.27291479852903e-06,
+      "loss": 0.4530356824398041,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6088730239673635,
+      "grad_norm": 0.48235180974006653,
+      "learning_rate": 7.264801632642764e-06,
+      "loss": 0.4509512782096863,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6091279959204488,
+      "grad_norm": 0.4398561418056488,
+      "learning_rate": 7.256690412277818e-06,
+      "loss": 0.4341862201690674,
+      "step": 2389
+    },
+    {
+      "epoch": 0.609382967873534,
+      "grad_norm": 0.429409384727478,
+      "learning_rate": 7.248581143203626e-06,
+      "loss": 0.440748393535614,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6096379398266191,
+      "grad_norm": 0.4427203834056854,
+      "learning_rate": 7.240473831188237e-06,
+      "loss": 0.4571680426597595,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6098929117797043,
+      "grad_norm": 0.46011874079704285,
+      "learning_rate": 7.232368481998309e-06,
+      "loss": 0.4616484045982361,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6101478837327894,
+      "grad_norm": 0.43624424934387207,
+      "learning_rate": 7.2242651013990965e-06,
+      "loss": 0.4401955008506775,
+      "step": 2393
+    },
+    {
+      "epoch": 0.6104028556858746,
+      "grad_norm": 0.44239377975463867,
+      "learning_rate": 7.216163695154463e-06,
+      "loss": 0.4554689824581146,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6106578276389597,
+      "grad_norm": 0.44673725962638855,
+      "learning_rate": 7.208064269026863e-06,
+      "loss": 0.43517106771469116,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6109127995920449,
+      "grad_norm": 0.4484138488769531,
+      "learning_rate": 7.1999668287773425e-06,
+      "loss": 0.4602460563182831,
+      "step": 2396
+    },
+    {
+      "epoch": 0.61116777154513,
+      "grad_norm": 0.44098860025405884,
+      "learning_rate": 7.191871380165538e-06,
+      "loss": 0.45212626457214355,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6114227434982152,
+      "grad_norm": 0.45807570219039917,
+      "learning_rate": 7.183777928949668e-06,
+      "loss": 0.44789016246795654,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6116777154513003,
+      "grad_norm": 0.45253509283065796,
+      "learning_rate": 7.175686480886529e-06,
+      "loss": 0.45914798974990845,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6119326874043856,
+      "grad_norm": 0.4360634982585907,
+      "learning_rate": 7.167597041731491e-06,
+      "loss": 0.4590366780757904,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6121876593574707,
+      "grad_norm": 0.436668336391449,
+      "learning_rate": 7.1595096172384985e-06,
+      "loss": 0.4478031098842621,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6124426313105559,
+      "grad_norm": 0.4420381188392639,
+      "learning_rate": 7.151424213160061e-06,
+      "loss": 0.4541397988796234,
+      "step": 2402
+    },
+    {
+      "epoch": 0.612697603263641,
+      "grad_norm": 0.42137983441352844,
+      "learning_rate": 7.143340835247252e-06,
+      "loss": 0.44909077882766724,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6129525752167262,
+      "grad_norm": 0.44776588678359985,
+      "learning_rate": 7.135259489249706e-06,
+      "loss": 0.46198710799217224,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6132075471698113,
+      "grad_norm": 0.4347139894962311,
+      "learning_rate": 7.127180180915609e-06,
+      "loss": 0.4604983329772949,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6134625191228965,
+      "grad_norm": 0.46223214268684387,
+      "learning_rate": 7.119102915991693e-06,
+      "loss": 0.44768351316452026,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6137174910759816,
+      "grad_norm": 0.4660816192626953,
+      "learning_rate": 7.111027700223249e-06,
+      "loss": 0.4439643621444702,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6139724630290668,
+      "grad_norm": 0.4395486116409302,
+      "learning_rate": 7.102954539354097e-06,
+      "loss": 0.4600735306739807,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6142274349821519,
+      "grad_norm": 0.4583393335342407,
+      "learning_rate": 7.094883439126607e-06,
+      "loss": 0.44173431396484375,
+      "step": 2409
+    },
+    {
+      "epoch": 0.6144824069352371,
+      "grad_norm": 0.43538230657577515,
+      "learning_rate": 7.086814405281677e-06,
+      "loss": 0.4389159083366394,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6147373788883223,
+      "grad_norm": 0.44855111837387085,
+      "learning_rate": 7.078747443558736e-06,
+      "loss": 0.4594293534755707,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6149923508414075,
+      "grad_norm": 0.4412626624107361,
+      "learning_rate": 7.070682559695737e-06,
+      "loss": 0.4607057571411133,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6152473227944926,
+      "grad_norm": 0.4448585510253906,
+      "learning_rate": 7.062619759429158e-06,
+      "loss": 0.4644583463668823,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6155022947475778,
+      "grad_norm": 0.4369284510612488,
+      "learning_rate": 7.054559048493996e-06,
+      "loss": 0.450400173664093,
+      "step": 2414
+    },
+    {
+      "epoch": 0.6157572667006629,
+      "grad_norm": 0.4468177556991577,
+      "learning_rate": 7.046500432623759e-06,
+      "loss": 0.4551388919353485,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6160122386537481,
+      "grad_norm": 0.4235501289367676,
+      "learning_rate": 7.038443917550464e-06,
+      "loss": 0.45600685477256775,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6162672106068332,
+      "grad_norm": 0.4289674460887909,
+      "learning_rate": 7.0303895090046405e-06,
+      "loss": 0.459359735250473,
+      "step": 2417
+    },
+    {
+      "epoch": 0.6165221825599184,
+      "grad_norm": 0.4907114505767822,
+      "learning_rate": 7.022337212715318e-06,
+      "loss": 0.4645995795726776,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6167771545130035,
+      "grad_norm": 0.4563983380794525,
+      "learning_rate": 7.014287034410009e-06,
+      "loss": 0.4514213502407074,
+      "step": 2419
+    },
+    {
+      "epoch": 0.6170321264660887,
+      "grad_norm": 0.44452720880508423,
+      "learning_rate": 7.006238979814735e-06,
+      "loss": 0.47157543897628784,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6172870984191738,
+      "grad_norm": 0.4489119052886963,
+      "learning_rate": 6.998193054654007e-06,
+      "loss": 0.4580475687980652,
+      "step": 2421
+    },
+    {
+      "epoch": 0.617542070372259,
+      "grad_norm": 0.4419918358325958,
+      "learning_rate": 6.990149264650814e-06,
+      "loss": 0.43857941031455994,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6177970423253442,
+      "grad_norm": 0.42903104424476624,
+      "learning_rate": 6.982107615526631e-06,
+      "loss": 0.46096014976501465,
+      "step": 2423
+    },
+    {
+      "epoch": 0.6180520142784294,
+      "grad_norm": 0.4421437978744507,
+      "learning_rate": 6.97406811300141e-06,
+      "loss": 0.44633787870407104,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6183069862315146,
+      "grad_norm": 0.44374901056289673,
+      "learning_rate": 6.966030762793569e-06,
+      "loss": 0.46367567777633667,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6185619581845997,
+      "grad_norm": 0.4307388365268707,
+      "learning_rate": 6.9579955706200086e-06,
+      "loss": 0.4521704912185669,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6188169301376849,
+      "grad_norm": 0.4368211627006531,
+      "learning_rate": 6.949962542196081e-06,
+      "loss": 0.4489135146141052,
+      "step": 2427
+    },
+    {
+      "epoch": 0.61907190209077,
+      "grad_norm": 0.4382425844669342,
+      "learning_rate": 6.941931683235603e-06,
+      "loss": 0.45526647567749023,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6193268740438552,
+      "grad_norm": 0.448438823223114,
+      "learning_rate": 6.933902999450859e-06,
+      "loss": 0.45010483264923096,
+      "step": 2429
+    },
+    {
+      "epoch": 0.6195818459969403,
+      "grad_norm": 0.43156129121780396,
+      "learning_rate": 6.925876496552572e-06,
+      "loss": 0.4513978660106659,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6198368179500255,
+      "grad_norm": 0.4507652223110199,
+      "learning_rate": 6.917852180249922e-06,
+      "loss": 0.458511084318161,
+      "step": 2431
+    },
+    {
+      "epoch": 0.6200917899031106,
+      "grad_norm": 0.43145832419395447,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.45626115798950195,
+      "step": 2432
+    },
+    {
+      "epoch": 0.6203467618561959,
+      "grad_norm": 0.4541116952896118,
+      "learning_rate": 6.901810130260452e-06,
+      "loss": 0.45203861594200134,
+      "step": 2433
+    },
+    {
+      "epoch": 0.620601733809281,
+      "grad_norm": 0.4439984858036041,
+      "learning_rate": 6.8937924079841964e-06,
+      "loss": 0.44320011138916016,
+      "step": 2434
+    },
+    {
+      "epoch": 0.6208567057623662,
+      "grad_norm": 0.4731787443161011,
+      "learning_rate": 6.885776895124692e-06,
+      "loss": 0.4608490467071533,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6211116777154513,
+      "grad_norm": 0.4167715907096863,
+      "learning_rate": 6.877763597383298e-06,
+      "loss": 0.46320483088493347,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6213666496685365,
+      "grad_norm": 0.45172250270843506,
+      "learning_rate": 6.869752520459803e-06,
+      "loss": 0.4560050368309021,
+      "step": 2437
+    },
+    {
+      "epoch": 0.6216216216216216,
+      "grad_norm": 0.4521925747394562,
+      "learning_rate": 6.861743670052404e-06,
+      "loss": 0.4454156756401062,
+      "step": 2438
+    },
+    {
+      "epoch": 0.6218765935747068,
+      "grad_norm": 0.44286608695983887,
+      "learning_rate": 6.853737051857729e-06,
+      "loss": 0.4496470093727112,
+      "step": 2439
+    },
+    {
+      "epoch": 0.6221315655277919,
+      "grad_norm": 0.4602040946483612,
+      "learning_rate": 6.845732671570806e-06,
+      "loss": 0.4612855911254883,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6223865374808771,
+      "grad_norm": 0.45351946353912354,
+      "learning_rate": 6.837730534885084e-06,
+      "loss": 0.45776110887527466,
+      "step": 2441
+    },
+    {
+      "epoch": 0.6226415094339622,
+      "grad_norm": 0.4474753439426422,
+      "learning_rate": 6.829730647492404e-06,
+      "loss": 0.45600712299346924,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6228964813870475,
+      "grad_norm": 0.4500572681427002,
+      "learning_rate": 6.821733015083016e-06,
+      "loss": 0.4475374221801758,
+      "step": 2443
+    },
+    {
+      "epoch": 0.6231514533401326,
+      "grad_norm": 0.43536829948425293,
+      "learning_rate": 6.813737643345556e-06,
+      "loss": 0.45218324661254883,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6234064252932178,
+      "grad_norm": 0.4450026750564575,
+      "learning_rate": 6.805744537967061e-06,
+      "loss": 0.4560825824737549,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6236613972463029,
+      "grad_norm": 0.5039249658584595,
+      "learning_rate": 6.797753704632953e-06,
+      "loss": 0.4593825340270996,
+      "step": 2446
+    },
+    {
+      "epoch": 0.6239163691993881,
+      "grad_norm": 0.4471895098686218,
+      "learning_rate": 6.789765149027039e-06,
+      "loss": 0.4518852233886719,
+      "step": 2447
+    },
+    {
+      "epoch": 0.6241713411524732,
+      "grad_norm": 0.44482189416885376,
+      "learning_rate": 6.781778876831502e-06,
+      "loss": 0.4477294087409973,
+      "step": 2448
+    },
+    {
+      "epoch": 0.6244263131055584,
+      "grad_norm": 0.43780866265296936,
+      "learning_rate": 6.773794893726904e-06,
+      "loss": 0.46982261538505554,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6246812850586435,
+      "grad_norm": 0.46241095662117004,
+      "learning_rate": 6.7658132053921864e-06,
+      "loss": 0.4516867399215698,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6249362570117287,
+      "grad_norm": 0.45196449756622314,
+      "learning_rate": 6.7578338175046394e-06,
+      "loss": 0.45734354853630066,
+      "step": 2451
+    },
+    {
+      "epoch": 0.6251912289648138,
+      "grad_norm": 0.44736403226852417,
+      "learning_rate": 6.749856735739929e-06,
+      "loss": 0.45609256625175476,
+      "step": 2452
+    },
+    {
+      "epoch": 0.625446200917899,
+      "grad_norm": 0.4608820378780365,
+      "learning_rate": 6.741881965772086e-06,
+      "loss": 0.45859766006469727,
+      "step": 2453
+    },
+    {
+      "epoch": 0.6257011728709841,
+      "grad_norm": 0.4258270561695099,
+      "learning_rate": 6.733909513273487e-06,
+      "loss": 0.44442683458328247,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6259561448240694,
+      "grad_norm": 0.4504850208759308,
+      "learning_rate": 6.725939383914864e-06,
+      "loss": 0.45771652460098267,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6262111167771545,
+      "grad_norm": 0.44868308305740356,
+      "learning_rate": 6.717971583365299e-06,
+      "loss": 0.4601963460445404,
+      "step": 2456
+    },
+    {
+      "epoch": 0.6264660887302397,
+      "grad_norm": 0.46132561564445496,
+      "learning_rate": 6.71000611729221e-06,
+      "loss": 0.45538172125816345,
+      "step": 2457
+    },
+    {
+      "epoch": 0.6267210606833248,
+      "grad_norm": 0.4325001835823059,
+      "learning_rate": 6.702042991361361e-06,
+      "loss": 0.45820948481559753,
+      "step": 2458
+    },
+    {
+      "epoch": 0.62697603263641,
+      "grad_norm": 0.4585916996002197,
+      "learning_rate": 6.694082211236852e-06,
+      "loss": 0.4583401679992676,
+      "step": 2459
+    },
+    {
+      "epoch": 0.6272310045894951,
+      "grad_norm": 0.4335605502128601,
+      "learning_rate": 6.686123782581109e-06,
+      "loss": 0.44409066438674927,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6274859765425803,
+      "grad_norm": 0.44804811477661133,
+      "learning_rate": 6.6781677110548904e-06,
+      "loss": 0.45734381675720215,
+      "step": 2461
+    },
+    {
+      "epoch": 0.6277409484956655,
+      "grad_norm": 0.4423656761646271,
+      "learning_rate": 6.670214002317279e-06,
+      "loss": 0.44959601759910583,
+      "step": 2462
+    },
+    {
+      "epoch": 0.6279959204487506,
+      "grad_norm": 0.45429491996765137,
+      "learning_rate": 6.6622626620256744e-06,
+      "loss": 0.4591550827026367,
+      "step": 2463
+    },
+    {
+      "epoch": 0.6282508924018358,
+      "grad_norm": 0.44495442509651184,
+      "learning_rate": 6.654313695835784e-06,
+      "loss": 0.4539817273616791,
+      "step": 2464
+    },
+    {
+      "epoch": 0.628505864354921,
+      "grad_norm": 0.43526822328567505,
+      "learning_rate": 6.6463671094016414e-06,
+      "loss": 0.46003812551498413,
+      "step": 2465
+    },
+    {
+      "epoch": 0.6287608363080062,
+      "grad_norm": 0.45048588514328003,
+      "learning_rate": 6.638422908375577e-06,
+      "loss": 0.4498358368873596,
+      "step": 2466
+    },
+    {
+      "epoch": 0.6290158082610913,
+      "grad_norm": 0.4301253855228424,
+      "learning_rate": 6.630481098408228e-06,
+      "loss": 0.46457192301750183,
+      "step": 2467
+    },
+    {
+      "epoch": 0.6292707802141765,
+      "grad_norm": 0.4593249261379242,
+      "learning_rate": 6.6225416851485315e-06,
+      "loss": 0.45766276121139526,
+      "step": 2468
+    },
+    {
+      "epoch": 0.6295257521672616,
+      "grad_norm": 0.47466710209846497,
+      "learning_rate": 6.6146046742437184e-06,
+      "loss": 0.4411028325557709,
+      "step": 2469
+    },
+    {
+      "epoch": 0.6297807241203468,
+      "grad_norm": 0.46918758749961853,
+      "learning_rate": 6.6066700713393106e-06,
+      "loss": 0.449954628944397,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6300356960734319,
+      "grad_norm": 0.45072171092033386,
+      "learning_rate": 6.598737882079118e-06,
+      "loss": 0.4382023811340332,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6302906680265171,
+      "grad_norm": 0.43589165806770325,
+      "learning_rate": 6.590808112105232e-06,
+      "loss": 0.4397895336151123,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6305456399796022,
+      "grad_norm": 0.44526374340057373,
+      "learning_rate": 6.5828807670580255e-06,
+      "loss": 0.4655776023864746,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6308006119326874,
+      "grad_norm": 0.432576984167099,
+      "learning_rate": 6.5749558525761485e-06,
+      "loss": 0.4525606334209442,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6310555838857725,
+      "grad_norm": 0.44795534014701843,
+      "learning_rate": 6.567033374296521e-06,
+      "loss": 0.4457744061946869,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6313105558388578,
+      "grad_norm": 0.46786946058273315,
+      "learning_rate": 6.5591133378543174e-06,
+      "loss": 0.4649132490158081,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6315655277919429,
+      "grad_norm": 0.43163734674453735,
+      "learning_rate": 6.551195748882997e-06,
+      "loss": 0.4570130705833435,
+      "step": 2477
+    },
+    {
+      "epoch": 0.6318204997450281,
+      "grad_norm": 0.47148701548576355,
+      "learning_rate": 6.5432806130142644e-06,
+      "loss": 0.45167526602745056,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6320754716981132,
+      "grad_norm": 0.44035637378692627,
+      "learning_rate": 6.535367935878084e-06,
+      "loss": 0.4530208706855774,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6323304436511984,
+      "grad_norm": 0.4428682029247284,
+      "learning_rate": 6.527457723102668e-06,
+      "loss": 0.44609448313713074,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6325854156042835,
+      "grad_norm": 0.4493977427482605,
+      "learning_rate": 6.519549980314477e-06,
+      "loss": 0.4558190703392029,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6328403875573687,
+      "grad_norm": 0.4617939889431,
+      "learning_rate": 6.511644713138221e-06,
+      "loss": 0.45700210332870483,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6330953595104538,
+      "grad_norm": 0.42630910873413086,
+      "learning_rate": 6.503741927196836e-06,
+      "loss": 0.4472409188747406,
+      "step": 2483
+    },
+    {
+      "epoch": 0.633350331463539,
+      "grad_norm": 0.4318679869174957,
+      "learning_rate": 6.495841628111505e-06,
+      "loss": 0.4415430426597595,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6336053034166241,
+      "grad_norm": 0.45561352372169495,
+      "learning_rate": 6.487943821501632e-06,
+      "loss": 0.46478819847106934,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6338602753697093,
+      "grad_norm": 0.45272183418273926,
+      "learning_rate": 6.480048512984863e-06,
+      "loss": 0.45466798543930054,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6341152473227945,
+      "grad_norm": 0.4237247407436371,
+      "learning_rate": 6.472155708177052e-06,
+      "loss": 0.4517762362957001,
+      "step": 2487
+    },
+    {
+      "epoch": 0.6343702192758797,
+      "grad_norm": 0.4491550922393799,
+      "learning_rate": 6.464265412692281e-06,
+      "loss": 0.4577600955963135,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6346251912289648,
+      "grad_norm": 0.4343549311161041,
+      "learning_rate": 6.456377632142842e-06,
+      "loss": 0.460031121969223,
+      "step": 2489
+    },
+    {
+      "epoch": 0.63488016318205,
+      "grad_norm": 0.43676286935806274,
+      "learning_rate": 6.44849237213924e-06,
+      "loss": 0.4529508352279663,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6351351351351351,
+      "grad_norm": 0.4376670718193054,
+      "learning_rate": 6.4406096382901895e-06,
+      "loss": 0.4492360055446625,
+      "step": 2491
+    },
+    {
+      "epoch": 0.6353901070882203,
+      "grad_norm": 0.44926509261131287,
+      "learning_rate": 6.432729436202605e-06,
+      "loss": 0.45888346433639526,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6356450790413054,
+      "grad_norm": 0.4386760890483856,
+      "learning_rate": 6.4248517714816e-06,
+      "loss": 0.44898495078086853,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6359000509943906,
+      "grad_norm": 0.4315982758998871,
+      "learning_rate": 6.4169766497304906e-06,
+      "loss": 0.4513634443283081,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6361550229474757,
+      "grad_norm": 0.4623962938785553,
+      "learning_rate": 6.4091040765507715e-06,
+      "loss": 0.4590742588043213,
+      "step": 2495
+    },
+    {
+      "epoch": 0.6364099949005609,
+      "grad_norm": 0.4697008728981018,
+      "learning_rate": 6.4012340575421316e-06,
+      "loss": 0.4546396732330322,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6366649668536462,
+      "grad_norm": 0.4347667098045349,
+      "learning_rate": 6.3933665983024465e-06,
+      "loss": 0.45000773668289185,
+      "step": 2497
+    },
+    {
+      "epoch": 0.6369199388067313,
+      "grad_norm": 0.4588465392589569,
+      "learning_rate": 6.3855017044277655e-06,
+      "loss": 0.45052796602249146,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6371749107598165,
+      "grad_norm": 0.43070754408836365,
+      "learning_rate": 6.377639381512315e-06,
+      "loss": 0.45954522490501404,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6374298827129016,
+      "grad_norm": 0.43200400471687317,
+      "learning_rate": 6.369779635148492e-06,
+      "loss": 0.45716235041618347,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6376848546659868,
+      "grad_norm": 0.44875460863113403,
+      "learning_rate": 6.3619224709268645e-06,
+      "loss": 0.45695194602012634,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6379398266190719,
+      "grad_norm": 0.4918175935745239,
+      "learning_rate": 6.3540678944361556e-06,
+      "loss": 0.4504631459712982,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6381947985721571,
+      "grad_norm": 0.44776660203933716,
+      "learning_rate": 6.346215911263257e-06,
+      "loss": 0.4516891837120056,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6384497705252422,
+      "grad_norm": 0.4455452561378479,
+      "learning_rate": 6.33836652699321e-06,
+      "loss": 0.4505302906036377,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6387047424783274,
+      "grad_norm": 0.4366051256656647,
+      "learning_rate": 6.3305197472092065e-06,
+      "loss": 0.4543272852897644,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6389597144314125,
+      "grad_norm": 0.4307226240634918,
+      "learning_rate": 6.322675577492594e-06,
+      "loss": 0.4597504138946533,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6392146863844977,
+      "grad_norm": 0.45603105425834656,
+      "learning_rate": 6.314834023422859e-06,
+      "loss": 0.44111618399620056,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6394696583375828,
+      "grad_norm": 0.42073628306388855,
+      "learning_rate": 6.306995090577617e-06,
+      "loss": 0.4541438817977905,
+      "step": 2508
+    },
+    {
+      "epoch": 0.6397246302906681,
+      "grad_norm": 0.45000380277633667,
+      "learning_rate": 6.299158784532635e-06,
+      "loss": 0.45004022121429443,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6399796022437532,
+      "grad_norm": 0.42614609003067017,
+      "learning_rate": 6.291325110861805e-06,
+      "loss": 0.4429459571838379,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6402345741968384,
+      "grad_norm": 0.4538845717906952,
+      "learning_rate": 6.283494075137147e-06,
+      "loss": 0.46047067642211914,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6404895461499235,
+      "grad_norm": 0.42154061794281006,
+      "learning_rate": 6.2756656829288035e-06,
+      "loss": 0.45054006576538086,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6407445181030087,
+      "grad_norm": 0.44144386053085327,
+      "learning_rate": 6.267839939805036e-06,
+      "loss": 0.45737284421920776,
+      "step": 2513
+    },
+    {
+      "epoch": 0.6409994900560938,
+      "grad_norm": 0.4459270238876343,
+      "learning_rate": 6.260016851332227e-06,
+      "loss": 0.44330281019210815,
+      "step": 2514
+    },
+    {
+      "epoch": 0.641254462009179,
+      "grad_norm": 0.43176722526550293,
+      "learning_rate": 6.252196423074864e-06,
+      "loss": 0.4518471360206604,
+      "step": 2515
+    },
+    {
+      "epoch": 0.6415094339622641,
+      "grad_norm": 0.4171527922153473,
+      "learning_rate": 6.244378660595544e-06,
+      "loss": 0.44375455379486084,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6417644059153493,
+      "grad_norm": 0.4783954322338104,
+      "learning_rate": 6.236563569454969e-06,
+      "loss": 0.45610353350639343,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6420193778684344,
+      "grad_norm": 0.4581439197063446,
+      "learning_rate": 6.228751155211946e-06,
+      "loss": 0.45231878757476807,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6422743498215197,
+      "grad_norm": 0.4603917598724365,
+      "learning_rate": 6.22094142342337e-06,
+      "loss": 0.4513261914253235,
+      "step": 2519
+    },
+    {
+      "epoch": 0.6425293217746048,
+      "grad_norm": 0.4171881675720215,
+      "learning_rate": 6.213134379644235e-06,
+      "loss": 0.45569711923599243,
+      "step": 2520
+    },
+    {
+      "epoch": 0.64278429372769,
+      "grad_norm": 0.438123881816864,
+      "learning_rate": 6.2053300294276094e-06,
+      "loss": 0.44259113073349,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6430392656807751,
+      "grad_norm": 0.4275990128517151,
+      "learning_rate": 6.197528378324664e-06,
+      "loss": 0.4489068388938904,
+      "step": 2522
+    },
+    {
+      "epoch": 0.6432942376338603,
+      "grad_norm": 0.44643691182136536,
+      "learning_rate": 6.1897294318846394e-06,
+      "loss": 0.45738035440444946,
+      "step": 2523
+    },
+    {
+      "epoch": 0.6435492095869454,
+      "grad_norm": 0.4457843601703644,
+      "learning_rate": 6.1819331956548525e-06,
+      "loss": 0.4524329602718353,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6438041815400306,
+      "grad_norm": 0.44476771354675293,
+      "learning_rate": 6.174139675180695e-06,
+      "loss": 0.43885207176208496,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6440591534931157,
+      "grad_norm": 0.4469187557697296,
+      "learning_rate": 6.1663488760056256e-06,
+      "loss": 0.4523865580558777,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6443141254462009,
+      "grad_norm": 0.4615035951137543,
+      "learning_rate": 6.158560803671168e-06,
+      "loss": 0.4440152049064636,
+      "step": 2527
+    },
+    {
+      "epoch": 0.644569097399286,
+      "grad_norm": 0.43872007727622986,
+      "learning_rate": 6.150775463716905e-06,
+      "loss": 0.4510747194290161,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6448240693523712,
+      "grad_norm": 0.45650118589401245,
+      "learning_rate": 6.142992861680477e-06,
+      "loss": 0.4303828477859497,
+      "step": 2529
+    },
+    {
+      "epoch": 0.6450790413054563,
+      "grad_norm": 0.4566303789615631,
+      "learning_rate": 6.13521300309758e-06,
+      "loss": 0.4519597291946411,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6453340132585416,
+      "grad_norm": 0.47050338983535767,
+      "learning_rate": 6.127435893501951e-06,
+      "loss": 0.4555937647819519,
+      "step": 2531
+    },
+    {
+      "epoch": 0.6455889852116268,
+      "grad_norm": 0.4484924077987671,
+      "learning_rate": 6.119661538425381e-06,
+      "loss": 0.44870179891586304,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6458439571647119,
+      "grad_norm": 0.45548173785209656,
+      "learning_rate": 6.111889943397695e-06,
+      "loss": 0.441594660282135,
+      "step": 2533
+    },
+    {
+      "epoch": 0.6460989291177971,
+      "grad_norm": 0.44344255328178406,
+      "learning_rate": 6.1041211139467545e-06,
+      "loss": 0.45279091596603394,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6463539010708822,
+      "grad_norm": 0.4390721619129181,
+      "learning_rate": 6.096355055598458e-06,
+      "loss": 0.45674705505371094,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6466088730239674,
+      "grad_norm": 0.44609135389328003,
+      "learning_rate": 6.088591773876731e-06,
+      "loss": 0.4513790011405945,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6468638449770525,
+      "grad_norm": 0.44705885648727417,
+      "learning_rate": 6.080831274303523e-06,
+      "loss": 0.43511801958084106,
+      "step": 2537
+    },
+    {
+      "epoch": 0.6471188169301377,
+      "grad_norm": 0.4256707727909088,
+      "learning_rate": 6.073073562398806e-06,
+      "loss": 0.4494359493255615,
+      "step": 2538
+    },
+    {
+      "epoch": 0.6473737888832228,
+      "grad_norm": 0.4406540095806122,
+      "learning_rate": 6.065318643680573e-06,
+      "loss": 0.44055265188217163,
+      "step": 2539
+    },
+    {
+      "epoch": 0.647628760836308,
+      "grad_norm": 0.47909441590309143,
+      "learning_rate": 6.057566523664819e-06,
+      "loss": 0.4648555517196655,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6478837327893932,
+      "grad_norm": 0.44395220279693604,
+      "learning_rate": 6.049817207865557e-06,
+      "loss": 0.44198960065841675,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6481387047424784,
+      "grad_norm": 0.4380699694156647,
+      "learning_rate": 6.042070701794806e-06,
+      "loss": 0.44920453429222107,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6483936766955635,
+      "grad_norm": 0.4582255780696869,
+      "learning_rate": 6.0343270109625855e-06,
+      "loss": 0.4538794457912445,
+      "step": 2543
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 0.43302080035209656,
+      "learning_rate": 6.026586140876908e-06,
+      "loss": 0.4537618160247803,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6489036206017338,
+      "grad_norm": 0.4683970510959625,
+      "learning_rate": 6.018848097043783e-06,
+      "loss": 0.45142412185668945,
+      "step": 2545
+    },
+    {
+      "epoch": 0.649158592554819,
+      "grad_norm": 0.4430343806743622,
+      "learning_rate": 6.011112884967214e-06,
+      "loss": 0.45917999744415283,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6494135645079041,
+      "grad_norm": 0.4492309093475342,
+      "learning_rate": 6.003380510149179e-06,
+      "loss": 0.4289212226867676,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6496685364609893,
+      "grad_norm": 0.4585253596305847,
+      "learning_rate": 5.995650978089649e-06,
+      "loss": 0.45338308811187744,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6499235084140744,
+      "grad_norm": 0.4318457841873169,
+      "learning_rate": 5.9879242942865675e-06,
+      "loss": 0.4537937045097351,
+      "step": 2549
+    },
+    {
+      "epoch": 0.6501784803671596,
+      "grad_norm": 0.438854455947876,
+      "learning_rate": 5.980200464235852e-06,
+      "loss": 0.44037315249443054,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6504334523202447,
+      "grad_norm": 0.44499143958091736,
+      "learning_rate": 5.972479493431395e-06,
+      "loss": 0.4457972049713135,
+      "step": 2551
+    },
+    {
+      "epoch": 0.65068842427333,
+      "grad_norm": 0.43149423599243164,
+      "learning_rate": 5.964761387365052e-06,
+      "loss": 0.46129220724105835,
+      "step": 2552
+    },
+    {
+      "epoch": 0.6509433962264151,
+      "grad_norm": 0.4269418716430664,
+      "learning_rate": 5.957046151526634e-06,
+      "loss": 0.44635066390037537,
+      "step": 2553
+    },
+    {
+      "epoch": 0.6511983681795003,
+      "grad_norm": 0.44937610626220703,
+      "learning_rate": 5.949333791403923e-06,
+      "loss": 0.4526378810405731,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6514533401325854,
+      "grad_norm": 0.4750503897666931,
+      "learning_rate": 5.941624312482648e-06,
+      "loss": 0.4481755495071411,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6517083120856706,
+      "grad_norm": 0.43597346544265747,
+      "learning_rate": 5.933917720246488e-06,
+      "loss": 0.451925128698349,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6519632840387557,
+      "grad_norm": 0.4327908456325531,
+      "learning_rate": 5.9262140201770745e-06,
+      "loss": 0.44594454765319824,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6522182559918409,
+      "grad_norm": 0.4297172725200653,
+      "learning_rate": 5.918513217753976e-06,
+      "loss": 0.4475826621055603,
+      "step": 2558
+    },
+    {
+      "epoch": 0.652473227944926,
+      "grad_norm": 0.4717065095901489,
+      "learning_rate": 5.910815318454703e-06,
+      "loss": 0.45425352454185486,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6527281998980112,
+      "grad_norm": 0.4428306221961975,
+      "learning_rate": 5.903120327754697e-06,
+      "loss": 0.4513399600982666,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6529831718510963,
+      "grad_norm": 0.43051621317863464,
+      "learning_rate": 5.8954282511273374e-06,
+      "loss": 0.4462912678718567,
+      "step": 2561
+    },
+    {
+      "epoch": 0.6532381438041815,
+      "grad_norm": 0.446906179189682,
+      "learning_rate": 5.887739094043923e-06,
+      "loss": 0.436113566160202,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6534931157572667,
+      "grad_norm": 0.46555471420288086,
+      "learning_rate": 5.880052861973686e-06,
+      "loss": 0.4455112814903259,
+      "step": 2563
+    },
+    {
+      "epoch": 0.6537480877103519,
+      "grad_norm": 0.4588684141635895,
+      "learning_rate": 5.872369560383771e-06,
+      "loss": 0.4556528925895691,
+      "step": 2564
+    },
+    {
+      "epoch": 0.654003059663437,
+      "grad_norm": 0.44467881321907043,
+      "learning_rate": 5.864689194739237e-06,
+      "loss": 0.44747036695480347,
+      "step": 2565
+    },
+    {
+      "epoch": 0.6542580316165222,
+      "grad_norm": 0.43433132767677307,
+      "learning_rate": 5.857011770503056e-06,
+      "loss": 0.437305212020874,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6545130035696073,
+      "grad_norm": 0.44600459933280945,
+      "learning_rate": 5.849337293136113e-06,
+      "loss": 0.44404035806655884,
+      "step": 2567
+    },
+    {
+      "epoch": 0.6547679755226925,
+      "grad_norm": 0.43104448914527893,
+      "learning_rate": 5.841665768097189e-06,
+      "loss": 0.44813108444213867,
+      "step": 2568
+    },
+    {
+      "epoch": 0.6550229474757777,
+      "grad_norm": 0.46145471930503845,
+      "learning_rate": 5.83399720084297e-06,
+      "loss": 0.448213666677475,
+      "step": 2569
+    },
+    {
+      "epoch": 0.6552779194288628,
+      "grad_norm": 0.47107648849487305,
+      "learning_rate": 5.826331596828037e-06,
+      "loss": 0.45392078161239624,
+      "step": 2570
+    },
+    {
+      "epoch": 0.655532891381948,
+      "grad_norm": 0.45138102769851685,
+      "learning_rate": 5.818668961504867e-06,
+      "loss": 0.4528809189796448,
+      "step": 2571
+    },
+    {
+      "epoch": 0.6557878633350331,
+      "grad_norm": 0.43152180314064026,
+      "learning_rate": 5.8110093003238175e-06,
+      "loss": 0.45032018423080444,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6560428352881184,
+      "grad_norm": 0.44748079776763916,
+      "learning_rate": 5.803352618733133e-06,
+      "loss": 0.44270965456962585,
+      "step": 2573
+    },
+    {
+      "epoch": 0.6562978072412035,
+      "grad_norm": 0.44162699580192566,
+      "learning_rate": 5.795698922178947e-06,
+      "loss": 0.45946773886680603,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6565527791942887,
+      "grad_norm": 0.5382289290428162,
+      "learning_rate": 5.788048216105254e-06,
+      "loss": 0.44528162479400635,
+      "step": 2575
+    },
+    {
+      "epoch": 0.6568077511473738,
+      "grad_norm": 0.4421813189983368,
+      "learning_rate": 5.7804005059539355e-06,
+      "loss": 0.4661310613155365,
+      "step": 2576
+    },
+    {
+      "epoch": 0.657062723100459,
+      "grad_norm": 0.4564555585384369,
+      "learning_rate": 5.772755797164743e-06,
+      "loss": 0.45135313272476196,
+      "step": 2577
+    },
+    {
+      "epoch": 0.6573176950535441,
+      "grad_norm": 0.45606404542922974,
+      "learning_rate": 5.765114095175281e-06,
+      "loss": 0.44829630851745605,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6575726670066293,
+      "grad_norm": 0.4391755759716034,
+      "learning_rate": 5.757475405421017e-06,
+      "loss": 0.44676074385643005,
+      "step": 2579
+    },
+    {
+      "epoch": 0.6578276389597144,
+      "grad_norm": 0.44971051812171936,
+      "learning_rate": 5.74983973333529e-06,
+      "loss": 0.4497562646865845,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6580826109127996,
+      "grad_norm": 0.4419224262237549,
+      "learning_rate": 5.742207084349274e-06,
+      "loss": 0.4493602514266968,
+      "step": 2581
+    },
+    {
+      "epoch": 0.6583375828658847,
+      "grad_norm": 0.42870786786079407,
+      "learning_rate": 5.734577463892008e-06,
+      "loss": 0.4335828721523285,
+      "step": 2582
+    },
+    {
+      "epoch": 0.6585925548189699,
+      "grad_norm": 0.5611152052879333,
+      "learning_rate": 5.726950877390367e-06,
+      "loss": 0.4474702775478363,
+      "step": 2583
+    },
+    {
+      "epoch": 0.658847526772055,
+      "grad_norm": 0.4444577693939209,
+      "learning_rate": 5.7193273302690755e-06,
+      "loss": 0.453504741191864,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6591024987251403,
+      "grad_norm": 0.44351455569267273,
+      "learning_rate": 5.711706827950691e-06,
+      "loss": 0.4382694661617279,
+      "step": 2585
+    },
+    {
+      "epoch": 0.6593574706782254,
+      "grad_norm": 0.44918960332870483,
+      "learning_rate": 5.7040893758556e-06,
+      "loss": 0.4414241313934326,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6596124426313106,
+      "grad_norm": 0.42611655592918396,
+      "learning_rate": 5.696474979402036e-06,
+      "loss": 0.4577499330043793,
+      "step": 2587
+    },
+    {
+      "epoch": 0.6598674145843957,
+      "grad_norm": 0.44104957580566406,
+      "learning_rate": 5.688863644006037e-06,
+      "loss": 0.4392445981502533,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6601223865374809,
+      "grad_norm": 0.4344611167907715,
+      "learning_rate": 5.681255375081485e-06,
+      "loss": 0.44575804471969604,
+      "step": 2589
+    },
+    {
+      "epoch": 0.660377358490566,
+      "grad_norm": 0.4342581033706665,
+      "learning_rate": 5.673650178040073e-06,
+      "loss": 0.4402153491973877,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6606323304436512,
+      "grad_norm": 0.4354460537433624,
+      "learning_rate": 5.666048058291298e-06,
+      "loss": 0.47594496607780457,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6608873023967363,
+      "grad_norm": 0.44662290811538696,
+      "learning_rate": 5.658449021242481e-06,
+      "loss": 0.4612034559249878,
+      "step": 2592
+    },
+    {
+      "epoch": 0.6611422743498215,
+      "grad_norm": 0.45198825001716614,
+      "learning_rate": 5.650853072298752e-06,
+      "loss": 0.44805312156677246,
+      "step": 2593
+    },
+    {
+      "epoch": 0.6613972463029066,
+      "grad_norm": 0.44156575202941895,
+      "learning_rate": 5.643260216863031e-06,
+      "loss": 0.4416990280151367,
+      "step": 2594
+    },
+    {
+      "epoch": 0.6616522182559919,
+      "grad_norm": 0.43809452652931213,
+      "learning_rate": 5.6356704603360545e-06,
+      "loss": 0.45295262336730957,
+      "step": 2595
+    },
+    {
+      "epoch": 0.661907190209077,
+      "grad_norm": 0.4427354037761688,
+      "learning_rate": 5.628083808116338e-06,
+      "loss": 0.44600507616996765,
+      "step": 2596
+    },
+    {
+      "epoch": 0.6621621621621622,
+      "grad_norm": 0.43249139189720154,
+      "learning_rate": 5.620500265600206e-06,
+      "loss": 0.4660620093345642,
+      "step": 2597
+    },
+    {
+      "epoch": 0.6624171341152473,
+      "grad_norm": 0.43712878227233887,
+      "learning_rate": 5.612919838181757e-06,
+      "loss": 0.43594229221343994,
+      "step": 2598
+    },
+    {
+      "epoch": 0.6626721060683325,
+      "grad_norm": 0.4529092609882355,
+      "learning_rate": 5.605342531252873e-06,
+      "loss": 0.4467432498931885,
+      "step": 2599
+    },
+    {
+      "epoch": 0.6629270780214176,
+      "grad_norm": 0.44129279255867004,
+      "learning_rate": 5.597768350203231e-06,
+      "loss": 0.451646625995636,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6631820499745028,
+      "grad_norm": 0.4357939660549164,
+      "learning_rate": 5.590197300420277e-06,
+      "loss": 0.44854840636253357,
+      "step": 2601
+    },
+    {
+      "epoch": 0.6634370219275879,
+      "grad_norm": 0.550003170967102,
+      "learning_rate": 5.58262938728922e-06,
+      "loss": 0.45516157150268555,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6636919938806731,
+      "grad_norm": 0.4335564076900482,
+      "learning_rate": 5.575064616193058e-06,
+      "loss": 0.45366984605789185,
+      "step": 2603
+    },
+    {
+      "epoch": 0.6639469658337583,
+      "grad_norm": 0.44977179169654846,
+      "learning_rate": 5.567502992512536e-06,
+      "loss": 0.4410482347011566,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6642019377868434,
+      "grad_norm": 0.43469348549842834,
+      "learning_rate": 5.559944521626164e-06,
+      "loss": 0.4456717371940613,
+      "step": 2605
+    },
+    {
+      "epoch": 0.6644569097399287,
+      "grad_norm": 0.45355284214019775,
+      "learning_rate": 5.552389208910219e-06,
+      "loss": 0.45244699716567993,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6647118816930138,
+      "grad_norm": 0.4623728096485138,
+      "learning_rate": 5.544837059738719e-06,
+      "loss": 0.45508691668510437,
+      "step": 2607
+    },
+    {
+      "epoch": 0.664966853646099,
+      "grad_norm": 0.44880279898643494,
+      "learning_rate": 5.537288079483445e-06,
+      "loss": 0.4548211991786957,
+      "step": 2608
+    },
+    {
+      "epoch": 0.6652218255991841,
+      "grad_norm": 0.4464181959629059,
+      "learning_rate": 5.529742273513914e-06,
+      "loss": 0.451995313167572,
+      "step": 2609
+    },
+    {
+      "epoch": 0.6654767975522693,
+      "grad_norm": 0.437755823135376,
+      "learning_rate": 5.522199647197384e-06,
+      "loss": 0.4451538920402527,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6657317695053544,
+      "grad_norm": 0.44787105917930603,
+      "learning_rate": 5.514660205898861e-06,
+      "loss": 0.44024309515953064,
+      "step": 2611
+    },
+    {
+      "epoch": 0.6659867414584396,
+      "grad_norm": 0.43366125226020813,
+      "learning_rate": 5.507123954981073e-06,
+      "loss": 0.45478901267051697,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6662417134115247,
+      "grad_norm": 0.43287813663482666,
+      "learning_rate": 5.4995908998044915e-06,
+      "loss": 0.44715046882629395,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6664966853646099,
+      "grad_norm": 0.4263473153114319,
+      "learning_rate": 5.49206104572731e-06,
+      "loss": 0.4595363736152649,
+      "step": 2614
+    },
+    {
+      "epoch": 0.666751657317695,
+      "grad_norm": 0.4816651940345764,
+      "learning_rate": 5.484534398105439e-06,
+      "loss": 0.45576655864715576,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6670066292707802,
+      "grad_norm": 0.4245269298553467,
+      "learning_rate": 5.47701096229252e-06,
+      "loss": 0.4497915506362915,
+      "step": 2616
+    },
+    {
+      "epoch": 0.6672616012238654,
+      "grad_norm": 0.43695083260536194,
+      "learning_rate": 5.4694907436399e-06,
+      "loss": 0.4417057931423187,
+      "step": 2617
+    },
+    {
+      "epoch": 0.6675165731769506,
+      "grad_norm": 0.5978035926818848,
+      "learning_rate": 5.461973747496636e-06,
+      "loss": 0.4519868791103363,
+      "step": 2618
+    },
+    {
+      "epoch": 0.6677715451300357,
+      "grad_norm": 0.4561963379383087,
+      "learning_rate": 5.454459979209506e-06,
+      "loss": 0.4437265396118164,
+      "step": 2619
+    },
+    {
+      "epoch": 0.6680265170831209,
+      "grad_norm": 0.43772682547569275,
+      "learning_rate": 5.446949444122976e-06,
+      "loss": 0.44676685333251953,
+      "step": 2620
+    },
+    {
+      "epoch": 0.668281489036206,
+      "grad_norm": 0.44906044006347656,
+      "learning_rate": 5.439442147579226e-06,
+      "loss": 0.4478064477443695,
+      "step": 2621
+    },
+    {
+      "epoch": 0.6685364609892912,
+      "grad_norm": 0.4388953745365143,
+      "learning_rate": 5.431938094918132e-06,
+      "loss": 0.4440498352050781,
+      "step": 2622
+    },
+    {
+      "epoch": 0.6687914329423763,
+      "grad_norm": 0.4350358247756958,
+      "learning_rate": 5.424437291477243e-06,
+      "loss": 0.4516732692718506,
+      "step": 2623
+    },
+    {
+      "epoch": 0.6690464048954615,
+      "grad_norm": 0.4323616623878479,
+      "learning_rate": 5.416939742591818e-06,
+      "loss": 0.45428466796875,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6693013768485466,
+      "grad_norm": 0.4260307550430298,
+      "learning_rate": 5.409445453594799e-06,
+      "loss": 0.4449426531791687,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6695563488016318,
+      "grad_norm": 0.4495390057563782,
+      "learning_rate": 5.401954429816797e-06,
+      "loss": 0.44637101888656616,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6698113207547169,
+      "grad_norm": 0.4546131193637848,
+      "learning_rate": 5.394466676586114e-06,
+      "loss": 0.4595838487148285,
+      "step": 2627
+    },
+    {
+      "epoch": 0.6700662927078022,
+      "grad_norm": 0.43755903840065,
+      "learning_rate": 5.386982199228714e-06,
+      "loss": 0.4451454281806946,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6703212646608873,
+      "grad_norm": 0.4387609660625458,
+      "learning_rate": 5.379501003068244e-06,
+      "loss": 0.44180548191070557,
+      "step": 2629
+    },
+    {
+      "epoch": 0.6705762366139725,
+      "grad_norm": 0.42819881439208984,
+      "learning_rate": 5.372023093426004e-06,
+      "loss": 0.4571313261985779,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6708312085670576,
+      "grad_norm": 0.4296024441719055,
+      "learning_rate": 5.3645484756209605e-06,
+      "loss": 0.4568363130092621,
+      "step": 2631
+    },
+    {
+      "epoch": 0.6710861805201428,
+      "grad_norm": 0.44118306040763855,
+      "learning_rate": 5.357077154969742e-06,
+      "loss": 0.4438619017601013,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6713411524732279,
+      "grad_norm": 0.4651823937892914,
+      "learning_rate": 5.349609136786636e-06,
+      "loss": 0.4524320960044861,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6715961244263131,
+      "grad_norm": 0.45834317803382874,
+      "learning_rate": 5.342144426383563e-06,
+      "loss": 0.4385998249053955,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6718510963793982,
+      "grad_norm": 0.43824106454849243,
+      "learning_rate": 5.334683029070118e-06,
+      "loss": 0.43711209297180176,
+      "step": 2635
+    },
+    {
+      "epoch": 0.6721060683324834,
+      "grad_norm": 0.4382132887840271,
+      "learning_rate": 5.327224950153507e-06,
+      "loss": 0.4439789056777954,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6723610402855685,
+      "grad_norm": 0.4576619565486908,
+      "learning_rate": 5.3197701949386e-06,
+      "loss": 0.45007187128067017,
+      "step": 2637
+    },
+    {
+      "epoch": 0.6726160122386537,
+      "grad_norm": 0.43934521079063416,
+      "learning_rate": 5.312318768727902e-06,
+      "loss": 0.4436546266078949,
+      "step": 2638
+    },
+    {
+      "epoch": 0.672870984191739,
+      "grad_norm": 0.4601533114910126,
+      "learning_rate": 5.304870676821532e-06,
+      "loss": 0.4548959732055664,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6731259561448241,
+      "grad_norm": 0.42206525802612305,
+      "learning_rate": 5.297425924517259e-06,
+      "loss": 0.45760273933410645,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6733809280979093,
+      "grad_norm": 0.4502648115158081,
+      "learning_rate": 5.2899845171104605e-06,
+      "loss": 0.4489644467830658,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6736359000509944,
+      "grad_norm": 0.756621241569519,
+      "learning_rate": 5.28254645989414e-06,
+      "loss": 0.46550774574279785,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6738908720040796,
+      "grad_norm": 0.45248404145240784,
+      "learning_rate": 5.275111758158925e-06,
+      "loss": 0.4535596966743469,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6741458439571647,
+      "grad_norm": 0.474893718957901,
+      "learning_rate": 5.267680417193042e-06,
+      "loss": 0.4447336196899414,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6744008159102499,
+      "grad_norm": 0.43572816252708435,
+      "learning_rate": 5.26025244228234e-06,
+      "loss": 0.450375497341156,
+      "step": 2645
+    },
+    {
+      "epoch": 0.674655787863335,
+      "grad_norm": 0.4469342827796936,
+      "learning_rate": 5.252827838710271e-06,
+      "loss": 0.44614309072494507,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6749107598164202,
+      "grad_norm": 0.4533426761627197,
+      "learning_rate": 5.245406611757882e-06,
+      "loss": 0.44949886202812195,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6751657317695053,
+      "grad_norm": 0.45130079984664917,
+      "learning_rate": 5.237988766703827e-06,
+      "loss": 0.44672998785972595,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6754207037225906,
+      "grad_norm": 0.4586695432662964,
+      "learning_rate": 5.230574308824352e-06,
+      "loss": 0.4617176055908203,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6756756756756757,
+      "grad_norm": 0.44763344526290894,
+      "learning_rate": 5.223163243393283e-06,
+      "loss": 0.4565013647079468,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6759306476287609,
+      "grad_norm": 0.4288623631000519,
+      "learning_rate": 5.215755575682054e-06,
+      "loss": 0.45748433470726013,
+      "step": 2651
+    },
+    {
+      "epoch": 0.676185619581846,
+      "grad_norm": 0.4510361850261688,
+      "learning_rate": 5.208351310959662e-06,
+      "loss": 0.4522618055343628,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6764405915349312,
+      "grad_norm": 0.4306660592556,
+      "learning_rate": 5.200950454492694e-06,
+      "loss": 0.44474905729293823,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6766955634880163,
+      "grad_norm": 0.4786277413368225,
+      "learning_rate": 5.19355301154532e-06,
+      "loss": 0.45132479071617126,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6769505354411015,
+      "grad_norm": 0.48917731642723083,
+      "learning_rate": 5.186158987379257e-06,
+      "loss": 0.4453713893890381,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6772055073941866,
+      "grad_norm": 0.4612053334712982,
+      "learning_rate": 5.178768387253815e-06,
+      "loss": 0.4520571231842041,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6774604793472718,
+      "grad_norm": 0.47949251532554626,
+      "learning_rate": 5.1713812164258625e-06,
+      "loss": 0.4475216269493103,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6777154513003569,
+      "grad_norm": 0.43021127581596375,
+      "learning_rate": 5.163997480149818e-06,
+      "loss": 0.45925724506378174,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6779704232534421,
+      "grad_norm": 0.4294150769710541,
+      "learning_rate": 5.156617183677672e-06,
+      "loss": 0.4427727162837982,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6782253952065272,
+      "grad_norm": 0.4174054265022278,
+      "learning_rate": 5.149240332258953e-06,
+      "loss": 0.4476696848869324,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6784803671596125,
+      "grad_norm": 0.44336920976638794,
+      "learning_rate": 5.1418669311407585e-06,
+      "loss": 0.4482848644256592,
+      "step": 2661
+    },
+    {
+      "epoch": 0.6787353391126976,
+      "grad_norm": 0.4349704086780548,
+      "learning_rate": 5.134496985567714e-06,
+      "loss": 0.4412155747413635,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6789903110657828,
+      "grad_norm": 0.4378722012042999,
+      "learning_rate": 5.127130500781988e-06,
+      "loss": 0.45444124937057495,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6792452830188679,
+      "grad_norm": 0.4330241084098816,
+      "learning_rate": 5.1197674820233045e-06,
+      "loss": 0.4420257806777954,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6795002549719531,
+      "grad_norm": 0.43697431683540344,
+      "learning_rate": 5.1124079345289e-06,
+      "loss": 0.44826623797416687,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6797552269250382,
+      "grad_norm": 0.44842299818992615,
+      "learning_rate": 5.105051863533559e-06,
+      "loss": 0.4589853286743164,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6800101988781234,
+      "grad_norm": 0.42814716696739197,
+      "learning_rate": 5.097699274269593e-06,
+      "loss": 0.45576342940330505,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6802651708312085,
+      "grad_norm": 0.4533100724220276,
+      "learning_rate": 5.0903501719668155e-06,
+      "loss": 0.4435397982597351,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6805201427842937,
+      "grad_norm": 0.4440280497074127,
+      "learning_rate": 5.083004561852583e-06,
+      "loss": 0.4435123801231384,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6807751147373788,
+      "grad_norm": 0.4627237021923065,
+      "learning_rate": 5.075662449151764e-06,
+      "loss": 0.44633591175079346,
+      "step": 2670
+    },
+    {
+      "epoch": 0.681030086690464,
+      "grad_norm": 0.4171501398086548,
+      "learning_rate": 5.068323839086728e-06,
+      "loss": 0.4337880313396454,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6812850586435492,
+      "grad_norm": 0.45280689001083374,
+      "learning_rate": 5.060988736877366e-06,
+      "loss": 0.45015066862106323,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6815400305966344,
+      "grad_norm": 0.4357544481754303,
+      "learning_rate": 5.053657147741067e-06,
+      "loss": 0.45653852820396423,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6817950025497195,
+      "grad_norm": 0.46855801343917847,
+      "learning_rate": 5.046329076892719e-06,
+      "loss": 0.44177255034446716,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6820499745028047,
+      "grad_norm": 0.42504367232322693,
+      "learning_rate": 5.039004529544718e-06,
+      "loss": 0.4501117169857025,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6823049464558899,
+      "grad_norm": 0.4342193007469177,
+      "learning_rate": 5.031683510906937e-06,
+      "loss": 0.44261711835861206,
+      "step": 2676
+    },
+    {
+      "epoch": 0.682559918408975,
+      "grad_norm": 0.44799095392227173,
+      "learning_rate": 5.0243660261867554e-06,
+      "loss": 0.4422341585159302,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6828148903620602,
+      "grad_norm": 0.4549873173236847,
+      "learning_rate": 5.0170520805890365e-06,
+      "loss": 0.45025262236595154,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6830698623151453,
+      "grad_norm": 0.4405505657196045,
+      "learning_rate": 5.009741679316113e-06,
+      "loss": 0.458317369222641,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6833248342682305,
+      "grad_norm": 0.4512025713920593,
+      "learning_rate": 5.0024348275678144e-06,
+      "loss": 0.45345792174339294,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6835798062213156,
+      "grad_norm": 0.457786500453949,
+      "learning_rate": 4.9951315305414316e-06,
+      "loss": 0.4395756721496582,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6838347781744009,
+      "grad_norm": 0.47746995091438293,
+      "learning_rate": 4.987831793431731e-06,
+      "loss": 0.45292338728904724,
+      "step": 2682
+    },
+    {
+      "epoch": 0.684089750127486,
+      "grad_norm": 0.49980446696281433,
+      "learning_rate": 4.980535621430953e-06,
+      "loss": 0.4654287099838257,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6843447220805712,
+      "grad_norm": 0.47383415699005127,
+      "learning_rate": 4.973243019728792e-06,
+      "loss": 0.4488328695297241,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6845996940336563,
+      "grad_norm": 0.44947731494903564,
+      "learning_rate": 4.965953993512413e-06,
+      "loss": 0.4427359998226166,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6848546659867415,
+      "grad_norm": 0.44544997811317444,
+      "learning_rate": 4.958668547966432e-06,
+      "loss": 0.4360436797142029,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6851096379398266,
+      "grad_norm": 0.4315359592437744,
+      "learning_rate": 4.951386688272915e-06,
+      "loss": 0.43482017517089844,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6853646098929118,
+      "grad_norm": 0.4480088949203491,
+      "learning_rate": 4.944108419611387e-06,
+      "loss": 0.43367239832878113,
+      "step": 2688
+    },
+    {
+      "epoch": 0.6856195818459969,
+      "grad_norm": 0.4426881670951843,
+      "learning_rate": 4.936833747158806e-06,
+      "loss": 0.441577285528183,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6858745537990821,
+      "grad_norm": 0.46219876408576965,
+      "learning_rate": 4.929562676089586e-06,
+      "loss": 0.4452260136604309,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6861295257521672,
+      "grad_norm": 0.43867194652557373,
+      "learning_rate": 4.922295211575574e-06,
+      "loss": 0.45053476095199585,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6863844977052524,
+      "grad_norm": 0.4794599711894989,
+      "learning_rate": 4.915031358786043e-06,
+      "loss": 0.442020982503891,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6866394696583376,
+      "grad_norm": 0.47966474294662476,
+      "learning_rate": 4.9077711228877125e-06,
+      "loss": 0.4580918252468109,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6868944416114228,
+      "grad_norm": 0.43187135457992554,
+      "learning_rate": 4.900514509044716e-06,
+      "loss": 0.42957156896591187,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6871494135645079,
+      "grad_norm": 0.4613824784755707,
+      "learning_rate": 4.893261522418614e-06,
+      "loss": 0.45658349990844727,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6874043855175931,
+      "grad_norm": 0.4548393487930298,
+      "learning_rate": 4.886012168168395e-06,
+      "loss": 0.45649951696395874,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6876593574706782,
+      "grad_norm": 0.4860595166683197,
+      "learning_rate": 4.878766451450451e-06,
+      "loss": 0.44177111983299255,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6879143294237634,
+      "grad_norm": 0.45349612832069397,
+      "learning_rate": 4.871524377418596e-06,
+      "loss": 0.449371337890625,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6881693013768485,
+      "grad_norm": 0.4503442049026489,
+      "learning_rate": 4.864285951224059e-06,
+      "loss": 0.45505473017692566,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6884242733299337,
+      "grad_norm": 0.4361976683139801,
+      "learning_rate": 4.857051178015451e-06,
+      "loss": 0.4406062066555023,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6886792452830188,
+      "grad_norm": 0.4606969952583313,
+      "learning_rate": 4.849820062938805e-06,
+      "loss": 0.4530186653137207,
+      "step": 2701
+    },
+    {
+      "epoch": 0.688934217236104,
+      "grad_norm": 0.43665894865989685,
+      "learning_rate": 4.842592611137551e-06,
+      "loss": 0.4582091271877289,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6891891891891891,
+      "grad_norm": 0.4496651589870453,
+      "learning_rate": 4.8353688277525015e-06,
+      "loss": 0.4425380229949951,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6894441611422744,
+      "grad_norm": 0.4403512477874756,
+      "learning_rate": 4.828148717921874e-06,
+      "loss": 0.4357679784297943,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6896991330953595,
+      "grad_norm": 0.4550356864929199,
+      "learning_rate": 4.820932286781263e-06,
+      "loss": 0.4607630670070648,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6899541050484447,
+      "grad_norm": 0.43422931432724,
+      "learning_rate": 4.813719539463644e-06,
+      "loss": 0.4445881247520447,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6902090770015298,
+      "grad_norm": 0.4598325490951538,
+      "learning_rate": 4.806510481099386e-06,
+      "loss": 0.43153107166290283,
+      "step": 2707
+    },
+    {
+      "epoch": 0.690464048954615,
+      "grad_norm": 0.48792895674705505,
+      "learning_rate": 4.799305116816217e-06,
+      "loss": 0.45328664779663086,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6907190209077001,
+      "grad_norm": 0.4260575473308563,
+      "learning_rate": 4.7921034517392516e-06,
+      "loss": 0.4630817174911499,
+      "step": 2709
+    },
+    {
+      "epoch": 0.6909739928607853,
+      "grad_norm": 0.426405131816864,
+      "learning_rate": 4.784905490990971e-06,
+      "loss": 0.45140355825424194,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6912289648138705,
+      "grad_norm": 0.44859254360198975,
+      "learning_rate": 4.777711239691208e-06,
+      "loss": 0.45345211029052734,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6914839367669556,
+      "grad_norm": 0.4419932961463928,
+      "learning_rate": 4.770520702957182e-06,
+      "loss": 0.44984006881713867,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6917389087200408,
+      "grad_norm": 0.483076274394989,
+      "learning_rate": 4.763333885903438e-06,
+      "loss": 0.4438518285751343,
+      "step": 2713
+    },
+    {
+      "epoch": 0.691993880673126,
+      "grad_norm": 0.5435576438903809,
+      "learning_rate": 4.756150793641902e-06,
+      "loss": 0.4360465407371521,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6922488526262112,
+      "grad_norm": 0.44988253712654114,
+      "learning_rate": 4.748971431281843e-06,
+      "loss": 0.43878430128097534,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6925038245792963,
+      "grad_norm": 0.47290316224098206,
+      "learning_rate": 4.741795803929867e-06,
+      "loss": 0.4487782418727875,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6927587965323815,
+      "grad_norm": 0.44757747650146484,
+      "learning_rate": 4.734623916689941e-06,
+      "loss": 0.4293305277824402,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6930137684854666,
+      "grad_norm": 0.44219687581062317,
+      "learning_rate": 4.727455774663357e-06,
+      "loss": 0.44206833839416504,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6932687404385518,
+      "grad_norm": 0.45080289244651794,
+      "learning_rate": 4.720291382948745e-06,
+      "loss": 0.44882553815841675,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6935237123916369,
+      "grad_norm": 0.4788825213909149,
+      "learning_rate": 4.713130746642078e-06,
+      "loss": 0.44349539279937744,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6937786843447221,
+      "grad_norm": 0.4739687442779541,
+      "learning_rate": 4.705973870836643e-06,
+      "loss": 0.43993064761161804,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6940336562978072,
+      "grad_norm": 0.4558063745498657,
+      "learning_rate": 4.698820760623064e-06,
+      "loss": 0.4564177393913269,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6942886282508924,
+      "grad_norm": 0.4424072206020355,
+      "learning_rate": 4.691671421089285e-06,
+      "loss": 0.44053369760513306,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6945436002039775,
+      "grad_norm": 0.4580995738506317,
+      "learning_rate": 4.684525857320561e-06,
+      "loss": 0.431607723236084,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6947985721570628,
+      "grad_norm": 0.43411484360694885,
+      "learning_rate": 4.677384074399462e-06,
+      "loss": 0.4485299289226532,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6950535441101479,
+      "grad_norm": 0.4523801803588867,
+      "learning_rate": 4.670246077405881e-06,
+      "loss": 0.4523230493068695,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6953085160632331,
+      "grad_norm": 0.44680526852607727,
+      "learning_rate": 4.663111871417e-06,
+      "loss": 0.43542763590812683,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6955634880163182,
+      "grad_norm": 0.44599083065986633,
+      "learning_rate": 4.6559814615073214e-06,
+      "loss": 0.4428485333919525,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6958184599694034,
+      "grad_norm": 0.4294494688510895,
+      "learning_rate": 4.648854852748635e-06,
+      "loss": 0.44599223136901855,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6960734319224885,
+      "grad_norm": 0.44038671255111694,
+      "learning_rate": 4.641732050210032e-06,
+      "loss": 0.44695937633514404,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6963284038755737,
+      "grad_norm": 0.4496418535709381,
+      "learning_rate": 4.6346130589579055e-06,
+      "loss": 0.4504983127117157,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6965833758286588,
+      "grad_norm": 0.9035135507583618,
+      "learning_rate": 4.627497884055912e-06,
+      "loss": 0.4419633746147156,
+      "step": 2732
+    },
+    {
+      "epoch": 0.696838347781744,
+      "grad_norm": 0.4334407448768616,
+      "learning_rate": 4.620386530565018e-06,
+      "loss": 0.44229233264923096,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6970933197348291,
+      "grad_norm": 0.4289253354072571,
+      "learning_rate": 4.613279003543468e-06,
+      "loss": 0.44758471846580505,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6973482916879143,
+      "grad_norm": 0.4214285612106323,
+      "learning_rate": 4.606175308046772e-06,
+      "loss": 0.43780988454818726,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6976032636409994,
+      "grad_norm": 0.45487329363822937,
+      "learning_rate": 4.59907544912773e-06,
+      "loss": 0.4521713852882385,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6978582355940847,
+      "grad_norm": 0.4270506203174591,
+      "learning_rate": 4.591979431836402e-06,
+      "loss": 0.43326398730278015,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6981132075471698,
+      "grad_norm": 0.4325590133666992,
+      "learning_rate": 4.584887261220117e-06,
+      "loss": 0.42342886328697205,
+      "step": 2738
+    },
+    {
+      "epoch": 0.698368179500255,
+      "grad_norm": 0.8522976636886597,
+      "learning_rate": 4.577798942323477e-06,
+      "loss": 0.43506431579589844,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6986231514533401,
+      "grad_norm": 0.491022527217865,
+      "learning_rate": 4.570714480188331e-06,
+      "loss": 0.4496845602989197,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6988781234064253,
+      "grad_norm": 0.4635283052921295,
+      "learning_rate": 4.563633879853797e-06,
+      "loss": 0.45886242389678955,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6991330953595104,
+      "grad_norm": 0.4629541039466858,
+      "learning_rate": 4.5565571463562365e-06,
+      "loss": 0.462287038564682,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6993880673125956,
+      "grad_norm": 0.44179099798202515,
+      "learning_rate": 4.549484284729265e-06,
+      "loss": 0.44741857051849365,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6996430392656807,
+      "grad_norm": 0.4561825394630432,
+      "learning_rate": 4.542415300003753e-06,
+      "loss": 0.44358640909194946,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6998980112187659,
+      "grad_norm": 0.49426209926605225,
+      "learning_rate": 4.5353501972077885e-06,
+      "loss": 0.4456411600112915,
+      "step": 2745
+    },
+    {
+      "epoch": 0.7001529831718512,
+      "grad_norm": 0.4641091227531433,
+      "learning_rate": 4.528288981366722e-06,
+      "loss": 0.442534863948822,
+      "step": 2746
+    },
+    {
+      "epoch": 0.7004079551249363,
+      "grad_norm": 0.433246374130249,
+      "learning_rate": 4.5212316575031325e-06,
+      "loss": 0.45268380641937256,
+      "step": 2747
+    },
+    {
+      "epoch": 0.7006629270780215,
+      "grad_norm": 0.44399333000183105,
+      "learning_rate": 4.514178230636824e-06,
+      "loss": 0.43584078550338745,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7009178990311066,
+      "grad_norm": 0.4575839936733246,
+      "learning_rate": 4.507128705784841e-06,
+      "loss": 0.4431343078613281,
+      "step": 2749
+    },
+    {
+      "epoch": 0.7011728709841918,
+      "grad_norm": 0.4551209509372711,
+      "learning_rate": 4.500083087961442e-06,
+      "loss": 0.4474720358848572,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7014278429372769,
+      "grad_norm": 0.48329710960388184,
+      "learning_rate": 4.493041382178105e-06,
+      "loss": 0.4326696991920471,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7016828148903621,
+      "grad_norm": 0.44136470556259155,
+      "learning_rate": 4.4860035934435375e-06,
+      "loss": 0.45715269446372986,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7019377868434472,
+      "grad_norm": 0.425485223531723,
+      "learning_rate": 4.478969726763648e-06,
+      "loss": 0.43804869055747986,
+      "step": 2753
+    },
+    {
+      "epoch": 0.7021927587965324,
+      "grad_norm": 0.4444881081581116,
+      "learning_rate": 4.471939787141561e-06,
+      "loss": 0.45472773909568787,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7024477307496175,
+      "grad_norm": 0.43286269903182983,
+      "learning_rate": 4.464913779577617e-06,
+      "loss": 0.4386185109615326,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7027027027027027,
+      "grad_norm": 0.45265665650367737,
+      "learning_rate": 4.457891709069341e-06,
+      "loss": 0.4427916705608368,
+      "step": 2756
+    },
+    {
+      "epoch": 0.7029576746557878,
+      "grad_norm": 0.6262911558151245,
+      "learning_rate": 4.450873580611466e-06,
+      "loss": 0.43673306703567505,
+      "step": 2757
+    },
+    {
+      "epoch": 0.7032126466088731,
+      "grad_norm": 0.4524223804473877,
+      "learning_rate": 4.443859399195928e-06,
+      "loss": 0.4502025842666626,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7034676185619582,
+      "grad_norm": 0.4436202645301819,
+      "learning_rate": 4.4368491698118425e-06,
+      "loss": 0.4552689492702484,
+      "step": 2759
+    },
+    {
+      "epoch": 0.7037225905150434,
+      "grad_norm": 0.4441589415073395,
+      "learning_rate": 4.429842897445529e-06,
+      "loss": 0.4459702968597412,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7039775624681285,
+      "grad_norm": 0.442243754863739,
+      "learning_rate": 4.422840587080476e-06,
+      "loss": 0.45736703276634216,
+      "step": 2761
+    },
+    {
+      "epoch": 0.7042325344212137,
+      "grad_norm": 0.5277527570724487,
+      "learning_rate": 4.415842243697369e-06,
+      "loss": 0.43360424041748047,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7044875063742988,
+      "grad_norm": 0.47451654076576233,
+      "learning_rate": 4.4088478722740635e-06,
+      "loss": 0.4568926692008972,
+      "step": 2763
+    },
+    {
+      "epoch": 0.704742478327384,
+      "grad_norm": 0.48908933997154236,
+      "learning_rate": 4.401857477785586e-06,
+      "loss": 0.4532874524593353,
+      "step": 2764
+    },
+    {
+      "epoch": 0.7049974502804691,
+      "grad_norm": 0.4372377097606659,
+      "learning_rate": 4.394871065204146e-06,
+      "loss": 0.4365415573120117,
+      "step": 2765
+    },
+    {
+      "epoch": 0.7052524222335543,
+      "grad_norm": 0.43009495735168457,
+      "learning_rate": 4.387888639499109e-06,
+      "loss": 0.4571770131587982,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7055073941866394,
+      "grad_norm": 0.4533731937408447,
+      "learning_rate": 4.380910205637012e-06,
+      "loss": 0.44067034125328064,
+      "step": 2767
+    },
+    {
+      "epoch": 0.7057623661397247,
+      "grad_norm": 0.444358229637146,
+      "learning_rate": 4.373935768581554e-06,
+      "loss": 0.4552963376045227,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7060173380928098,
+      "grad_norm": 0.44427451491355896,
+      "learning_rate": 4.366965333293586e-06,
+      "loss": 0.43634259700775146,
+      "step": 2769
+    },
+    {
+      "epoch": 0.706272310045895,
+      "grad_norm": 0.45108288526535034,
+      "learning_rate": 4.359998904731107e-06,
+      "loss": 0.4482075572013855,
+      "step": 2770
+    },
+    {
+      "epoch": 0.7065272819989801,
+      "grad_norm": 0.42642563581466675,
+      "learning_rate": 4.353036487849285e-06,
+      "loss": 0.44058769941329956,
+      "step": 2771
+    },
+    {
+      "epoch": 0.7067822539520653,
+      "grad_norm": 0.4614061415195465,
+      "learning_rate": 4.346078087600411e-06,
+      "loss": 0.43684858083724976,
+      "step": 2772
+    },
+    {
+      "epoch": 0.7070372259051504,
+      "grad_norm": 0.4506364166736603,
+      "learning_rate": 4.339123708933941e-06,
+      "loss": 0.44278430938720703,
+      "step": 2773
+    },
+    {
+      "epoch": 0.7072921978582356,
+      "grad_norm": 0.44530582427978516,
+      "learning_rate": 4.332173356796452e-06,
+      "loss": 0.43861445784568787,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7075471698113207,
+      "grad_norm": 0.4369005560874939,
+      "learning_rate": 4.3252270361316695e-06,
+      "loss": 0.4379958510398865,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7078021417644059,
+      "grad_norm": 0.43202000856399536,
+      "learning_rate": 4.3182847518804536e-06,
+      "loss": 0.4420497417449951,
+      "step": 2776
+    },
+    {
+      "epoch": 0.708057113717491,
+      "grad_norm": 0.43807902932167053,
+      "learning_rate": 4.311346508980773e-06,
+      "loss": 0.4428611993789673,
+      "step": 2777
+    },
+    {
+      "epoch": 0.7083120856705762,
+      "grad_norm": 0.42809081077575684,
+      "learning_rate": 4.304412312367743e-06,
+      "loss": 0.46127551794052124,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7085670576236613,
+      "grad_norm": 0.45571109652519226,
+      "learning_rate": 4.297482166973596e-06,
+      "loss": 0.45550617575645447,
+      "step": 2779
+    },
+    {
+      "epoch": 0.7088220295767466,
+      "grad_norm": 0.433366060256958,
+      "learning_rate": 4.290556077727675e-06,
+      "loss": 0.4468022286891937,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7090770015298317,
+      "grad_norm": 0.4316498041152954,
+      "learning_rate": 4.283634049556449e-06,
+      "loss": 0.4427945017814636,
+      "step": 2781
+    },
+    {
+      "epoch": 0.7093319734829169,
+      "grad_norm": 0.45922383666038513,
+      "learning_rate": 4.2767160873834904e-06,
+      "loss": 0.4336870610713959,
+      "step": 2782
+    },
+    {
+      "epoch": 0.7095869454360021,
+      "grad_norm": 0.44714367389678955,
+      "learning_rate": 4.269802196129477e-06,
+      "loss": 0.44349512457847595,
+      "step": 2783
+    },
+    {
+      "epoch": 0.7098419173890872,
+      "grad_norm": 0.44446757435798645,
+      "learning_rate": 4.2628923807122024e-06,
+      "loss": 0.44376105070114136,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7100968893421724,
+      "grad_norm": 0.4495147168636322,
+      "learning_rate": 4.255986646046549e-06,
+      "loss": 0.4431377649307251,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7103518612952575,
+      "grad_norm": 0.4551457166671753,
+      "learning_rate": 4.249084997044505e-06,
+      "loss": 0.44691866636276245,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7106068332483427,
+      "grad_norm": 0.44687506556510925,
+      "learning_rate": 4.242187438615153e-06,
+      "loss": 0.4500548541545868,
+      "step": 2787
+    },
+    {
+      "epoch": 0.7108618052014278,
+      "grad_norm": 0.4545448124408722,
+      "learning_rate": 4.235293975664658e-06,
+      "loss": 0.43516406416893005,
+      "step": 2788
+    },
+    {
+      "epoch": 0.711116777154513,
+      "grad_norm": 0.4348348379135132,
+      "learning_rate": 4.228404613096275e-06,
+      "loss": 0.4404086768627167,
+      "step": 2789
+    },
+    {
+      "epoch": 0.7113717491075982,
+      "grad_norm": 0.4600769877433777,
+      "learning_rate": 4.22151935581035e-06,
+      "loss": 0.44938015937805176,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7116267210606834,
+      "grad_norm": 0.45333588123321533,
+      "learning_rate": 4.214638208704295e-06,
+      "loss": 0.4313281774520874,
+      "step": 2791
+    },
+    {
+      "epoch": 0.7118816930137685,
+      "grad_norm": 0.4451024532318115,
+      "learning_rate": 4.207761176672615e-06,
+      "loss": 0.44445693492889404,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7121366649668537,
+      "grad_norm": 0.4356543719768524,
+      "learning_rate": 4.200888264606869e-06,
+      "loss": 0.4430459141731262,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7123916369199388,
+      "grad_norm": 0.43236348032951355,
+      "learning_rate": 4.194019477395708e-06,
+      "loss": 0.4271642565727234,
+      "step": 2794
+    },
+    {
+      "epoch": 0.712646608873024,
+      "grad_norm": 0.4406087398529053,
+      "learning_rate": 4.187154819924829e-06,
+      "loss": 0.4488407075405121,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7129015808261091,
+      "grad_norm": 0.47709518671035767,
+      "learning_rate": 4.180294297076997e-06,
+      "loss": 0.44116121530532837,
+      "step": 2796
+    },
+    {
+      "epoch": 0.7131565527791943,
+      "grad_norm": 0.4376133978366852,
+      "learning_rate": 4.173437913732048e-06,
+      "loss": 0.4433579444885254,
+      "step": 2797
+    },
+    {
+      "epoch": 0.7134115247322794,
+      "grad_norm": 0.4503537118434906,
+      "learning_rate": 4.166585674766853e-06,
+      "loss": 0.4418359696865082,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7136664966853646,
+      "grad_norm": 0.44435012340545654,
+      "learning_rate": 4.159737585055353e-06,
+      "loss": 0.453326940536499,
+      "step": 2799
+    },
+    {
+      "epoch": 0.7139214686384497,
+      "grad_norm": 0.4249648153781891,
+      "learning_rate": 4.152893649468533e-06,
+      "loss": 0.4535840153694153,
+      "step": 2800
+    },
+    {
+      "epoch": 0.714176440591535,
+      "grad_norm": 0.4280673861503601,
+      "learning_rate": 4.146053872874417e-06,
+      "loss": 0.4415363073348999,
+      "step": 2801
+    },
+    {
+      "epoch": 0.7144314125446201,
+      "grad_norm": 0.5632311105728149,
+      "learning_rate": 4.139218260138074e-06,
+      "loss": 0.43556785583496094,
+      "step": 2802
+    },
+    {
+      "epoch": 0.7146863844977053,
+      "grad_norm": 0.435977041721344,
+      "learning_rate": 4.132386816121616e-06,
+      "loss": 0.4506576955318451,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7149413564507904,
+      "grad_norm": 0.44648057222366333,
+      "learning_rate": 4.12555954568418e-06,
+      "loss": 0.43422961235046387,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7151963284038756,
+      "grad_norm": 0.4379127025604248,
+      "learning_rate": 4.118736453681949e-06,
+      "loss": 0.44232869148254395,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7154513003569607,
+      "grad_norm": 0.43217480182647705,
+      "learning_rate": 4.111917544968116e-06,
+      "loss": 0.44452977180480957,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7157062723100459,
+      "grad_norm": 0.4436221420764923,
+      "learning_rate": 4.105102824392912e-06,
+      "loss": 0.4404059052467346,
+      "step": 2807
+    },
+    {
+      "epoch": 0.715961244263131,
+      "grad_norm": 0.449494868516922,
+      "learning_rate": 4.098292296803593e-06,
+      "loss": 0.44439762830734253,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7162162162162162,
+      "grad_norm": 0.43304330110549927,
+      "learning_rate": 4.091485967044409e-06,
+      "loss": 0.4660702347755432,
+      "step": 2809
+    },
+    {
+      "epoch": 0.7164711881693013,
+      "grad_norm": 0.44333916902542114,
+      "learning_rate": 4.084683839956647e-06,
+      "loss": 0.4513348937034607,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7167261601223865,
+      "grad_norm": 0.44252026081085205,
+      "learning_rate": 4.0778859203786e-06,
+      "loss": 0.44663819670677185,
+      "step": 2811
+    },
+    {
+      "epoch": 0.7169811320754716,
+      "grad_norm": 0.46256279945373535,
+      "learning_rate": 4.0710922131455595e-06,
+      "loss": 0.4497140645980835,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7172361040285569,
+      "grad_norm": 0.5449967980384827,
+      "learning_rate": 4.064302723089832e-06,
+      "loss": 0.43796730041503906,
+      "step": 2813
+    },
+    {
+      "epoch": 0.717491075981642,
+      "grad_norm": 0.43101754784584045,
+      "learning_rate": 4.057517455040717e-06,
+      "loss": 0.4442185163497925,
+      "step": 2814
+    },
+    {
+      "epoch": 0.7177460479347272,
+      "grad_norm": 0.43088218569755554,
+      "learning_rate": 4.050736413824509e-06,
+      "loss": 0.45426806807518005,
+      "step": 2815
+    },
+    {
+      "epoch": 0.7180010198878123,
+      "grad_norm": 0.4459136724472046,
+      "learning_rate": 4.0439596042645065e-06,
+      "loss": 0.45037540793418884,
+      "step": 2816
+    },
+    {
+      "epoch": 0.7182559918408975,
+      "grad_norm": 0.4530165195465088,
+      "learning_rate": 4.037187031180986e-06,
+      "loss": 0.45242929458618164,
+      "step": 2817
+    },
+    {
+      "epoch": 0.7185109637939827,
+      "grad_norm": 0.443611204624176,
+      "learning_rate": 4.0304186993912205e-06,
+      "loss": 0.4525124430656433,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7187659357470678,
+      "grad_norm": 0.4550294876098633,
+      "learning_rate": 4.0236546137094576e-06,
+      "loss": 0.4362117052078247,
+      "step": 2819
+    },
+    {
+      "epoch": 0.719020907700153,
+      "grad_norm": 0.4554041922092438,
+      "learning_rate": 4.016894778946934e-06,
+      "loss": 0.44021332263946533,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7192758796532381,
+      "grad_norm": 0.49561822414398193,
+      "learning_rate": 4.010139199911854e-06,
+      "loss": 0.4398931860923767,
+      "step": 2821
+    },
+    {
+      "epoch": 0.7195308516063234,
+      "grad_norm": 0.4929938018321991,
+      "learning_rate": 4.003387881409397e-06,
+      "loss": 0.4550277590751648,
+      "step": 2822
+    },
+    {
+      "epoch": 0.7197858235594085,
+      "grad_norm": 0.4283452033996582,
+      "learning_rate": 3.9966408282417155e-06,
+      "loss": 0.44857415556907654,
+      "step": 2823
+    },
+    {
+      "epoch": 0.7200407955124937,
+      "grad_norm": 0.473547101020813,
+      "learning_rate": 3.989898045207931e-06,
+      "loss": 0.4334336221218109,
+      "step": 2824
+    },
+    {
+      "epoch": 0.7202957674655788,
+      "grad_norm": 0.4281342923641205,
+      "learning_rate": 3.983159537104113e-06,
+      "loss": 0.44296973943710327,
+      "step": 2825
+    },
+    {
+      "epoch": 0.720550739418664,
+      "grad_norm": 0.48224103450775146,
+      "learning_rate": 3.97642530872331e-06,
+      "loss": 0.44094353914260864,
+      "step": 2826
+    },
+    {
+      "epoch": 0.7208057113717491,
+      "grad_norm": 0.43880414962768555,
+      "learning_rate": 3.969695364855512e-06,
+      "loss": 0.44613009691238403,
+      "step": 2827
+    },
+    {
+      "epoch": 0.7210606833248343,
+      "grad_norm": 0.44216209650039673,
+      "learning_rate": 3.96296971028766e-06,
+      "loss": 0.4331624209880829,
+      "step": 2828
+    },
+    {
+      "epoch": 0.7213156552779194,
+      "grad_norm": 0.44664260745048523,
+      "learning_rate": 3.95624834980366e-06,
+      "loss": 0.4392186105251312,
+      "step": 2829
+    },
+    {
+      "epoch": 0.7215706272310046,
+      "grad_norm": 0.4344140589237213,
+      "learning_rate": 3.949531288184345e-06,
+      "loss": 0.4357095956802368,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7218255991840897,
+      "grad_norm": 0.4699135422706604,
+      "learning_rate": 3.942818530207504e-06,
+      "loss": 0.43042606115341187,
+      "step": 2831
+    },
+    {
+      "epoch": 0.7220805711371749,
+      "grad_norm": 0.4375179708003998,
+      "learning_rate": 3.93611008064786e-06,
+      "loss": 0.44665050506591797,
+      "step": 2832
+    },
+    {
+      "epoch": 0.72233554309026,
+      "grad_norm": 0.4416513442993164,
+      "learning_rate": 3.929405944277072e-06,
+      "loss": 0.4503394067287445,
+      "step": 2833
+    },
+    {
+      "epoch": 0.7225905150433453,
+      "grad_norm": 0.44849449396133423,
+      "learning_rate": 3.9227061258637236e-06,
+      "loss": 0.4251297116279602,
+      "step": 2834
+    },
+    {
+      "epoch": 0.7228454869964304,
+      "grad_norm": 0.45465877652168274,
+      "learning_rate": 3.916010630173341e-06,
+      "loss": 0.44149231910705566,
+      "step": 2835
+    },
+    {
+      "epoch": 0.7231004589495156,
+      "grad_norm": 0.45338740944862366,
+      "learning_rate": 3.909319461968363e-06,
+      "loss": 0.4441741406917572,
+      "step": 2836
+    },
+    {
+      "epoch": 0.7233554309026007,
+      "grad_norm": 0.4316529631614685,
+      "learning_rate": 3.90263262600816e-06,
+      "loss": 0.44860753417015076,
+      "step": 2837
+    },
+    {
+      "epoch": 0.7236104028556859,
+      "grad_norm": 0.44786977767944336,
+      "learning_rate": 3.895950127049013e-06,
+      "loss": 0.4485016465187073,
+      "step": 2838
+    },
+    {
+      "epoch": 0.723865374808771,
+      "grad_norm": 0.42609506845474243,
+      "learning_rate": 3.889271969844123e-06,
+      "loss": 0.4528493285179138,
+      "step": 2839
+    },
+    {
+      "epoch": 0.7241203467618562,
+      "grad_norm": 0.44411757588386536,
+      "learning_rate": 3.882598159143609e-06,
+      "loss": 0.4639114737510681,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7243753187149413,
+      "grad_norm": 0.45524635910987854,
+      "learning_rate": 3.875928699694477e-06,
+      "loss": 0.44869527220726013,
+      "step": 2841
+    },
+    {
+      "epoch": 0.7246302906680265,
+      "grad_norm": 0.4178290367126465,
+      "learning_rate": 3.869263596240661e-06,
+      "loss": 0.46156296133995056,
+      "step": 2842
+    },
+    {
+      "epoch": 0.7248852626211116,
+      "grad_norm": 0.43635308742523193,
+      "learning_rate": 3.862602853522983e-06,
+      "loss": 0.4542333781719208,
+      "step": 2843
+    },
+    {
+      "epoch": 0.7251402345741969,
+      "grad_norm": 0.42462316155433655,
+      "learning_rate": 3.85594647627917e-06,
+      "loss": 0.44222697615623474,
+      "step": 2844
+    },
+    {
+      "epoch": 0.725395206527282,
+      "grad_norm": 0.4641839265823364,
+      "learning_rate": 3.8492944692438446e-06,
+      "loss": 0.4510359466075897,
+      "step": 2845
+    },
+    {
+      "epoch": 0.7256501784803672,
+      "grad_norm": 0.44939640164375305,
+      "learning_rate": 3.842646837148515e-06,
+      "loss": 0.4434684216976166,
+      "step": 2846
+    },
+    {
+      "epoch": 0.7259051504334523,
+      "grad_norm": 0.44022607803344727,
+      "learning_rate": 3.836003584721577e-06,
+      "loss": 0.43942567706108093,
+      "step": 2847
+    },
+    {
+      "epoch": 0.7261601223865375,
+      "grad_norm": 0.45273932814598083,
+      "learning_rate": 3.8293647166883215e-06,
+      "loss": 0.44515150785446167,
+      "step": 2848
+    },
+    {
+      "epoch": 0.7264150943396226,
+      "grad_norm": 0.47738298773765564,
+      "learning_rate": 3.822730237770908e-06,
+      "loss": 0.4392485022544861,
+      "step": 2849
+    },
+    {
+      "epoch": 0.7266700662927078,
+      "grad_norm": 0.430848628282547,
+      "learning_rate": 3.816100152688385e-06,
+      "loss": 0.44411778450012207,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7269250382457929,
+      "grad_norm": 0.439805805683136,
+      "learning_rate": 3.809474466156664e-06,
+      "loss": 0.4501129984855652,
+      "step": 2851
+    },
+    {
+      "epoch": 0.7271800101988781,
+      "grad_norm": 0.46396151185035706,
+      "learning_rate": 3.802853182888543e-06,
+      "loss": 0.4317866265773773,
+      "step": 2852
+    },
+    {
+      "epoch": 0.7274349821519633,
+      "grad_norm": 0.5632092952728271,
+      "learning_rate": 3.796236307593676e-06,
+      "loss": 0.44473159313201904,
+      "step": 2853
+    },
+    {
+      "epoch": 0.7276899541050484,
+      "grad_norm": 0.5041026473045349,
+      "learning_rate": 3.7896238449785803e-06,
+      "loss": 0.44062238931655884,
+      "step": 2854
+    },
+    {
+      "epoch": 0.7279449260581337,
+      "grad_norm": 0.4421965777873993,
+      "learning_rate": 3.783015799746644e-06,
+      "loss": 0.46264347434043884,
+      "step": 2855
+    },
+    {
+      "epoch": 0.7281998980112188,
+      "grad_norm": 0.4480365514755249,
+      "learning_rate": 3.7764121765981133e-06,
+      "loss": 0.44427475333213806,
+      "step": 2856
+    },
+    {
+      "epoch": 0.728454869964304,
+      "grad_norm": 0.44947126507759094,
+      "learning_rate": 3.7698129802300744e-06,
+      "loss": 0.451511025428772,
+      "step": 2857
+    },
+    {
+      "epoch": 0.7287098419173891,
+      "grad_norm": 0.5991028547286987,
+      "learning_rate": 3.7632182153364838e-06,
+      "loss": 0.4486866891384125,
+      "step": 2858
+    },
+    {
+      "epoch": 0.7289648138704743,
+      "grad_norm": 0.4347856342792511,
+      "learning_rate": 3.756627886608133e-06,
+      "loss": 0.4447828233242035,
+      "step": 2859
+    },
+    {
+      "epoch": 0.7292197858235594,
+      "grad_norm": 0.42992183566093445,
+      "learning_rate": 3.7500419987326574e-06,
+      "loss": 0.44058841466903687,
+      "step": 2860
+    },
+    {
+      "epoch": 0.7294747577766446,
+      "grad_norm": 0.4386395215988159,
+      "learning_rate": 3.743460556394547e-06,
+      "loss": 0.44062167406082153,
+      "step": 2861
+    },
+    {
+      "epoch": 0.7297297297297297,
+      "grad_norm": 0.4650445878505707,
+      "learning_rate": 3.7368835642751123e-06,
+      "loss": 0.4424327313899994,
+      "step": 2862
+    },
+    {
+      "epoch": 0.7299847016828149,
+      "grad_norm": 0.4388737976551056,
+      "learning_rate": 3.730311027052511e-06,
+      "loss": 0.43483245372772217,
+      "step": 2863
+    },
+    {
+      "epoch": 0.7302396736359,
+      "grad_norm": 0.42335045337677,
+      "learning_rate": 3.7237429494017306e-06,
+      "loss": 0.4343976378440857,
+      "step": 2864
+    },
+    {
+      "epoch": 0.7304946455889852,
+      "grad_norm": 0.4400503933429718,
+      "learning_rate": 3.717179335994583e-06,
+      "loss": 0.4404789209365845,
+      "step": 2865
+    },
+    {
+      "epoch": 0.7307496175420704,
+      "grad_norm": 0.4329128563404083,
+      "learning_rate": 3.710620191499701e-06,
+      "loss": 0.4489637315273285,
+      "step": 2866
+    },
+    {
+      "epoch": 0.7310045894951556,
+      "grad_norm": 0.43637967109680176,
+      "learning_rate": 3.704065520582549e-06,
+      "loss": 0.44528207182884216,
+      "step": 2867
+    },
+    {
+      "epoch": 0.7312595614482407,
+      "grad_norm": 0.42939335107803345,
+      "learning_rate": 3.697515327905399e-06,
+      "loss": 0.434497207403183,
+      "step": 2868
+    },
+    {
+      "epoch": 0.7315145334013259,
+      "grad_norm": 0.5018973350524902,
+      "learning_rate": 3.690969618127348e-06,
+      "loss": 0.45246458053588867,
+      "step": 2869
+    },
+    {
+      "epoch": 0.731769505354411,
+      "grad_norm": 0.43627703189849854,
+      "learning_rate": 3.6844283959042924e-06,
+      "loss": 0.42965325713157654,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7320244773074962,
+      "grad_norm": 0.46314767003059387,
+      "learning_rate": 3.6778916658889506e-06,
+      "loss": 0.43974757194519043,
+      "step": 2871
+    },
+    {
+      "epoch": 0.7322794492605813,
+      "grad_norm": 0.47196269035339355,
+      "learning_rate": 3.6713594327308343e-06,
+      "loss": 0.4353893995285034,
+      "step": 2872
+    },
+    {
+      "epoch": 0.7325344212136665,
+      "grad_norm": 0.45461487770080566,
+      "learning_rate": 3.664831701076258e-06,
+      "loss": 0.44562897086143494,
+      "step": 2873
+    },
+    {
+      "epoch": 0.7327893931667516,
+      "grad_norm": 0.46071672439575195,
+      "learning_rate": 3.658308475568342e-06,
+      "loss": 0.43509751558303833,
+      "step": 2874
+    },
+    {
+      "epoch": 0.7330443651198368,
+      "grad_norm": 0.44231125712394714,
+      "learning_rate": 3.65178976084699e-06,
+      "loss": 0.442873477935791,
+      "step": 2875
+    },
+    {
+      "epoch": 0.7332993370729219,
+      "grad_norm": 0.4249587953090668,
+      "learning_rate": 3.6452755615489077e-06,
+      "loss": 0.43926429748535156,
+      "step": 2876
+    },
+    {
+      "epoch": 0.7335543090260072,
+      "grad_norm": 0.46666327118873596,
+      "learning_rate": 3.638765882307589e-06,
+      "loss": 0.4330987334251404,
+      "step": 2877
+    },
+    {
+      "epoch": 0.7338092809790923,
+      "grad_norm": 0.45984357595443726,
+      "learning_rate": 3.6322607277533027e-06,
+      "loss": 0.4495600461959839,
+      "step": 2878
+    },
+    {
+      "epoch": 0.7340642529321775,
+      "grad_norm": 0.4424566328525543,
+      "learning_rate": 3.625760102513103e-06,
+      "loss": 0.4446689486503601,
+      "step": 2879
+    },
+    {
+      "epoch": 0.7343192248852626,
+      "grad_norm": 0.44157835841178894,
+      "learning_rate": 3.6192640112108322e-06,
+      "loss": 0.44923990964889526,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7345741968383478,
+      "grad_norm": 0.4445481300354004,
+      "learning_rate": 3.612772458467092e-06,
+      "loss": 0.44120463728904724,
+      "step": 2881
+    },
+    {
+      "epoch": 0.7348291687914329,
+      "grad_norm": 0.5572802424430847,
+      "learning_rate": 3.6062854488992714e-06,
+      "loss": 0.43533897399902344,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7350841407445181,
+      "grad_norm": 0.45228099822998047,
+      "learning_rate": 3.599802987121512e-06,
+      "loss": 0.4434191584587097,
+      "step": 2883
+    },
+    {
+      "epoch": 0.7353391126976032,
+      "grad_norm": 0.46297982335090637,
+      "learning_rate": 3.593325077744739e-06,
+      "loss": 0.4504498839378357,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7355940846506884,
+      "grad_norm": 0.44489234685897827,
+      "learning_rate": 3.5868517253766235e-06,
+      "loss": 0.43913325667381287,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7358490566037735,
+      "grad_norm": 0.4312242567539215,
+      "learning_rate": 3.580382934621598e-06,
+      "loss": 0.4557768702507019,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7361040285568587,
+      "grad_norm": 0.44070443511009216,
+      "learning_rate": 3.573918710080857e-06,
+      "loss": 0.4463081359863281,
+      "step": 2887
+    },
+    {
+      "epoch": 0.7363590005099439,
+      "grad_norm": 0.4172808527946472,
+      "learning_rate": 3.567459056352347e-06,
+      "loss": 0.44776830077171326,
+      "step": 2888
+    },
+    {
+      "epoch": 0.7366139724630291,
+      "grad_norm": 0.43707603216171265,
+      "learning_rate": 3.5610039780307546e-06,
+      "loss": 0.4435421824455261,
+      "step": 2889
+    },
+    {
+      "epoch": 0.7368689444161143,
+      "grad_norm": 0.4548320770263672,
+      "learning_rate": 3.5545534797075234e-06,
+      "loss": 0.4290686845779419,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7371239163691994,
+      "grad_norm": 0.42244952917099,
+      "learning_rate": 3.5481075659708287e-06,
+      "loss": 0.4392220973968506,
+      "step": 2891
+    },
+    {
+      "epoch": 0.7373788883222846,
+      "grad_norm": 0.4414602518081665,
+      "learning_rate": 3.541666241405588e-06,
+      "loss": 0.45275434851646423,
+      "step": 2892
+    },
+    {
+      "epoch": 0.7376338602753697,
+      "grad_norm": 0.45304572582244873,
+      "learning_rate": 3.5352295105934607e-06,
+      "loss": 0.4434977173805237,
+      "step": 2893
+    },
+    {
+      "epoch": 0.7378888322284549,
+      "grad_norm": 0.45812326669692993,
+      "learning_rate": 3.528797378112828e-06,
+      "loss": 0.4410284459590912,
+      "step": 2894
+    },
+    {
+      "epoch": 0.73814380418154,
+      "grad_norm": 0.44536668062210083,
+      "learning_rate": 3.5223698485388136e-06,
+      "loss": 0.4378358721733093,
+      "step": 2895
+    },
+    {
+      "epoch": 0.7383987761346252,
+      "grad_norm": 0.43989458680152893,
+      "learning_rate": 3.515946926443252e-06,
+      "loss": 0.4467636048793793,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7386537480877103,
+      "grad_norm": 0.4682613015174866,
+      "learning_rate": 3.509528616394716e-06,
+      "loss": 0.4479398727416992,
+      "step": 2897
+    },
+    {
+      "epoch": 0.7389087200407956,
+      "grad_norm": 0.4957796633243561,
+      "learning_rate": 3.503114922958486e-06,
+      "loss": 0.4401901364326477,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7391636919938807,
+      "grad_norm": 0.4830598831176758,
+      "learning_rate": 3.496705850696561e-06,
+      "loss": 0.4409579634666443,
+      "step": 2899
+    },
+    {
+      "epoch": 0.7394186639469659,
+      "grad_norm": 0.43040627241134644,
+      "learning_rate": 3.4903014041676565e-06,
+      "loss": 0.43947097659111023,
+      "step": 2900
+    },
+    {
+      "epoch": 0.739673635900051,
+      "grad_norm": 0.4375455975532532,
+      "learning_rate": 3.483901587927201e-06,
+      "loss": 0.4315758943557739,
+      "step": 2901
+    },
+    {
+      "epoch": 0.7399286078531362,
+      "grad_norm": 0.44875892996788025,
+      "learning_rate": 3.477506406527317e-06,
+      "loss": 0.44477686285972595,
+      "step": 2902
+    },
+    {
+      "epoch": 0.7401835798062213,
+      "grad_norm": 0.5409514307975769,
+      "learning_rate": 3.4711158645168443e-06,
+      "loss": 0.4459382891654968,
+      "step": 2903
+    },
+    {
+      "epoch": 0.7404385517593065,
+      "grad_norm": 0.44331035017967224,
+      "learning_rate": 3.4647299664413136e-06,
+      "loss": 0.4305277466773987,
+      "step": 2904
+    },
+    {
+      "epoch": 0.7406935237123916,
+      "grad_norm": 0.46023833751678467,
+      "learning_rate": 3.458348716842952e-06,
+      "loss": 0.42848941683769226,
+      "step": 2905
+    },
+    {
+      "epoch": 0.7409484956654768,
+      "grad_norm": 0.453451544046402,
+      "learning_rate": 3.4519721202606895e-06,
+      "loss": 0.4502413272857666,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7412034676185619,
+      "grad_norm": 0.43192124366760254,
+      "learning_rate": 3.4456001812301344e-06,
+      "loss": 0.4459025263786316,
+      "step": 2907
+    },
+    {
+      "epoch": 0.7414584395716471,
+      "grad_norm": 0.4435686767101288,
+      "learning_rate": 3.4392329042835904e-06,
+      "loss": 0.44413721561431885,
+      "step": 2908
+    },
+    {
+      "epoch": 0.7417134115247322,
+      "grad_norm": 0.44830721616744995,
+      "learning_rate": 3.432870293950047e-06,
+      "loss": 0.45666563510894775,
+      "step": 2909
+    },
+    {
+      "epoch": 0.7419683834778175,
+      "grad_norm": 0.42248955368995667,
+      "learning_rate": 3.426512354755166e-06,
+      "loss": 0.43796348571777344,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7422233554309026,
+      "grad_norm": 0.437529057264328,
+      "learning_rate": 3.420159091221288e-06,
+      "loss": 0.4518410563468933,
+      "step": 2911
+    },
+    {
+      "epoch": 0.7424783273839878,
+      "grad_norm": 0.43285441398620605,
+      "learning_rate": 3.4138105078674366e-06,
+      "loss": 0.43692296743392944,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7427332993370729,
+      "grad_norm": 0.45706868171691895,
+      "learning_rate": 3.4074666092092945e-06,
+      "loss": 0.4546869397163391,
+      "step": 2913
+    },
+    {
+      "epoch": 0.7429882712901581,
+      "grad_norm": 0.44124263525009155,
+      "learning_rate": 3.401127399759225e-06,
+      "loss": 0.4413590431213379,
+      "step": 2914
+    },
+    {
+      "epoch": 0.7432432432432432,
+      "grad_norm": 0.4383951723575592,
+      "learning_rate": 3.3947928840262425e-06,
+      "loss": 0.4420372247695923,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7434982151963284,
+      "grad_norm": 0.44535309076309204,
+      "learning_rate": 3.3884630665160344e-06,
+      "loss": 0.4349723756313324,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7437531871494135,
+      "grad_norm": 0.41102197766304016,
+      "learning_rate": 3.3821379517309406e-06,
+      "loss": 0.44966214895248413,
+      "step": 2917
+    },
+    {
+      "epoch": 0.7440081591024987,
+      "grad_norm": 0.44411683082580566,
+      "learning_rate": 3.3758175441699525e-06,
+      "loss": 0.4433893859386444,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7442631310555838,
+      "grad_norm": 0.470836877822876,
+      "learning_rate": 3.3695018483287212e-06,
+      "loss": 0.44160282611846924,
+      "step": 2919
+    },
+    {
+      "epoch": 0.744518103008669,
+      "grad_norm": 0.5225307941436768,
+      "learning_rate": 3.363190868699544e-06,
+      "loss": 0.44049859046936035,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7447730749617542,
+      "grad_norm": 0.4494461119174957,
+      "learning_rate": 3.3568846097713594e-06,
+      "loss": 0.4394533038139343,
+      "step": 2921
+    },
+    {
+      "epoch": 0.7450280469148394,
+      "grad_norm": 0.4861459732055664,
+      "learning_rate": 3.3505830760297543e-06,
+      "loss": 0.4236561059951782,
+      "step": 2922
+    },
+    {
+      "epoch": 0.7452830188679245,
+      "grad_norm": 0.4226544201374054,
+      "learning_rate": 3.34428627195695e-06,
+      "loss": 0.4505658745765686,
+      "step": 2923
+    },
+    {
+      "epoch": 0.7455379908210097,
+      "grad_norm": 0.445040762424469,
+      "learning_rate": 3.3379942020318014e-06,
+      "loss": 0.46367499232292175,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7457929627740949,
+      "grad_norm": 0.43944481015205383,
+      "learning_rate": 3.331706870729806e-06,
+      "loss": 0.4512088894844055,
+      "step": 2925
+    },
+    {
+      "epoch": 0.74604793472718,
+      "grad_norm": 0.47440463304519653,
+      "learning_rate": 3.3254242825230777e-06,
+      "loss": 0.4355928301811218,
+      "step": 2926
+    },
+    {
+      "epoch": 0.7463029066802652,
+      "grad_norm": 0.4428374767303467,
+      "learning_rate": 3.319146441880371e-06,
+      "loss": 0.44209808111190796,
+      "step": 2927
+    },
+    {
+      "epoch": 0.7465578786333503,
+      "grad_norm": 0.4566478729248047,
+      "learning_rate": 3.3128733532670478e-06,
+      "loss": 0.4385034441947937,
+      "step": 2928
+    },
+    {
+      "epoch": 0.7468128505864355,
+      "grad_norm": 0.4758487045764923,
+      "learning_rate": 3.306605021145106e-06,
+      "loss": 0.4396563470363617,
+      "step": 2929
+    },
+    {
+      "epoch": 0.7470678225395206,
+      "grad_norm": 0.440157413482666,
+      "learning_rate": 3.300341449973148e-06,
+      "loss": 0.4350048899650574,
+      "step": 2930
+    },
+    {
+      "epoch": 0.7473227944926059,
+      "grad_norm": 0.4477387070655823,
+      "learning_rate": 3.2940826442063924e-06,
+      "loss": 0.43772047758102417,
+      "step": 2931
+    },
+    {
+      "epoch": 0.747577766445691,
+      "grad_norm": 0.42891183495521545,
+      "learning_rate": 3.2878286082966704e-06,
+      "loss": 0.44212839007377625,
+      "step": 2932
+    },
+    {
+      "epoch": 0.7478327383987762,
+      "grad_norm": 0.4327998459339142,
+      "learning_rate": 3.281579346692426e-06,
+      "loss": 0.4553752541542053,
+      "step": 2933
+    },
+    {
+      "epoch": 0.7480877103518613,
+      "grad_norm": 0.44166067242622375,
+      "learning_rate": 3.275334863838694e-06,
+      "loss": 0.41830703616142273,
+      "step": 2934
+    },
+    {
+      "epoch": 0.7483426823049465,
+      "grad_norm": 0.45415228605270386,
+      "learning_rate": 3.269095164177123e-06,
+      "loss": 0.4383961856365204,
+      "step": 2935
+    },
+    {
+      "epoch": 0.7485976542580316,
+      "grad_norm": 0.4422258138656616,
+      "learning_rate": 3.2628602521459496e-06,
+      "loss": 0.4465511739253998,
+      "step": 2936
+    },
+    {
+      "epoch": 0.7488526262111168,
+      "grad_norm": 0.4468015730381012,
+      "learning_rate": 3.256630132180009e-06,
+      "loss": 0.4185349941253662,
+      "step": 2937
+    },
+    {
+      "epoch": 0.7491075981642019,
+      "grad_norm": 0.4371696710586548,
+      "learning_rate": 3.250404808710731e-06,
+      "loss": 0.4337192475795746,
+      "step": 2938
+    },
+    {
+      "epoch": 0.7493625701172871,
+      "grad_norm": 0.4528083801269531,
+      "learning_rate": 3.2441842861661266e-06,
+      "loss": 0.44116291403770447,
+      "step": 2939
+    },
+    {
+      "epoch": 0.7496175420703722,
+      "grad_norm": 0.42228686809539795,
+      "learning_rate": 3.2379685689707985e-06,
+      "loss": 0.45381030440330505,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7498725140234574,
+      "grad_norm": 0.4403478503227234,
+      "learning_rate": 3.2317576615459324e-06,
+      "loss": 0.43453022837638855,
+      "step": 2941
+    },
+    {
+      "epoch": 0.7501274859765426,
+      "grad_norm": 0.42008721828460693,
+      "learning_rate": 3.225551568309284e-06,
+      "loss": 0.4479690492153168,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7503824579296278,
+      "grad_norm": 0.435314804315567,
+      "learning_rate": 3.2193502936751895e-06,
+      "loss": 0.43803542852401733,
+      "step": 2943
+    },
+    {
+      "epoch": 0.7506374298827129,
+      "grad_norm": 0.43313083052635193,
+      "learning_rate": 3.213153842054564e-06,
+      "loss": 0.4430111050605774,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7508924018357981,
+      "grad_norm": 0.4652822017669678,
+      "learning_rate": 3.20696221785488e-06,
+      "loss": 0.4293684959411621,
+      "step": 2945
+    },
+    {
+      "epoch": 0.7511473737888832,
+      "grad_norm": 0.46115192770957947,
+      "learning_rate": 3.2007754254801872e-06,
+      "loss": 0.43593597412109375,
+      "step": 2946
+    },
+    {
+      "epoch": 0.7514023457419684,
+      "grad_norm": 0.46106842160224915,
+      "learning_rate": 3.1945934693310897e-06,
+      "loss": 0.4410976767539978,
+      "step": 2947
+    },
+    {
+      "epoch": 0.7516573176950535,
+      "grad_norm": 0.45949119329452515,
+      "learning_rate": 3.1884163538047607e-06,
+      "loss": 0.44500434398651123,
+      "step": 2948
+    },
+    {
+      "epoch": 0.7519122896481387,
+      "grad_norm": 0.4491775631904602,
+      "learning_rate": 3.182244083294923e-06,
+      "loss": 0.44120240211486816,
+      "step": 2949
+    },
+    {
+      "epoch": 0.7521672616012238,
+      "grad_norm": 0.42800047993659973,
+      "learning_rate": 3.1760766621918515e-06,
+      "loss": 0.43075132369995117,
+      "step": 2950
+    },
+    {
+      "epoch": 0.752422233554309,
+      "grad_norm": 0.45833832025527954,
+      "learning_rate": 3.1699140948823835e-06,
+      "loss": 0.430724561214447,
+      "step": 2951
+    },
+    {
+      "epoch": 0.7526772055073941,
+      "grad_norm": 0.45019084215164185,
+      "learning_rate": 3.1637563857498886e-06,
+      "loss": 0.43860697746276855,
+      "step": 2952
+    },
+    {
+      "epoch": 0.7529321774604794,
+      "grad_norm": 0.45236337184906006,
+      "learning_rate": 3.157603539174293e-06,
+      "loss": 0.44204622507095337,
+      "step": 2953
+    },
+    {
+      "epoch": 0.7531871494135645,
+      "grad_norm": 0.43661150336265564,
+      "learning_rate": 3.1514555595320604e-06,
+      "loss": 0.4306318461894989,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7534421213666497,
+      "grad_norm": 0.43726611137390137,
+      "learning_rate": 3.145312451196192e-06,
+      "loss": 0.42486798763275146,
+      "step": 2955
+    },
+    {
+      "epoch": 0.7536970933197348,
+      "grad_norm": 0.4403837025165558,
+      "learning_rate": 3.1391742185362174e-06,
+      "loss": 0.446133553981781,
+      "step": 2956
+    },
+    {
+      "epoch": 0.75395206527282,
+      "grad_norm": 0.47683706879615784,
+      "learning_rate": 3.133040865918213e-06,
+      "loss": 0.4425346255302429,
+      "step": 2957
+    },
+    {
+      "epoch": 0.7542070372259051,
+      "grad_norm": 0.44009360671043396,
+      "learning_rate": 3.1269123977047687e-06,
+      "loss": 0.4544362425804138,
+      "step": 2958
+    },
+    {
+      "epoch": 0.7544620091789903,
+      "grad_norm": 0.43296313285827637,
+      "learning_rate": 3.120788818255015e-06,
+      "loss": 0.4442460536956787,
+      "step": 2959
+    },
+    {
+      "epoch": 0.7547169811320755,
+      "grad_norm": 0.42778074741363525,
+      "learning_rate": 3.1146701319245896e-06,
+      "loss": 0.4324518144130707,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7549719530851606,
+      "grad_norm": 0.44995859265327454,
+      "learning_rate": 3.1085563430656653e-06,
+      "loss": 0.4532390832901001,
+      "step": 2961
+    },
+    {
+      "epoch": 0.7552269250382458,
+      "grad_norm": 0.4304356873035431,
+      "learning_rate": 3.102447456026919e-06,
+      "loss": 0.4366806745529175,
+      "step": 2962
+    },
+    {
+      "epoch": 0.755481896991331,
+      "grad_norm": 0.466528981924057,
+      "learning_rate": 3.0963434751535427e-06,
+      "loss": 0.45421624183654785,
+      "step": 2963
+    },
+    {
+      "epoch": 0.7557368689444162,
+      "grad_norm": 0.576439380645752,
+      "learning_rate": 3.090244404787245e-06,
+      "loss": 0.43155166506767273,
+      "step": 2964
+    },
+    {
+      "epoch": 0.7559918408975013,
+      "grad_norm": 0.4535970091819763,
+      "learning_rate": 3.084150249266241e-06,
+      "loss": 0.43590113520622253,
+      "step": 2965
+    },
+    {
+      "epoch": 0.7562468128505865,
+      "grad_norm": 0.45630592107772827,
+      "learning_rate": 3.078061012925242e-06,
+      "loss": 0.43900856375694275,
+      "step": 2966
+    },
+    {
+      "epoch": 0.7565017848036716,
+      "grad_norm": 0.42787259817123413,
+      "learning_rate": 3.0719767000954714e-06,
+      "loss": 0.44750213623046875,
+      "step": 2967
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 0.4446224570274353,
+      "learning_rate": 3.065897315104641e-06,
+      "loss": 0.43496066331863403,
+      "step": 2968
+    },
+    {
+      "epoch": 0.7570117287098419,
+      "grad_norm": 0.4456896185874939,
+      "learning_rate": 3.059822862276958e-06,
+      "loss": 0.42623668909072876,
+      "step": 2969
+    },
+    {
+      "epoch": 0.7572667006629271,
+      "grad_norm": 0.4380398392677307,
+      "learning_rate": 3.0537533459331312e-06,
+      "loss": 0.42342907190322876,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7575216726160122,
+      "grad_norm": 0.45309311151504517,
+      "learning_rate": 3.0476887703903456e-06,
+      "loss": 0.44981467723846436,
+      "step": 2971
+    },
+    {
+      "epoch": 0.7577766445690974,
+      "grad_norm": 0.49701908230781555,
+      "learning_rate": 3.0416291399622834e-06,
+      "loss": 0.4507873058319092,
+      "step": 2972
+    },
+    {
+      "epoch": 0.7580316165221825,
+      "grad_norm": 0.41651859879493713,
+      "learning_rate": 3.0355744589590976e-06,
+      "loss": 0.42973122000694275,
+      "step": 2973
+    },
+    {
+      "epoch": 0.7582865884752678,
+      "grad_norm": 0.41733089089393616,
+      "learning_rate": 3.029524731687432e-06,
+      "loss": 0.4291536211967468,
+      "step": 2974
+    },
+    {
+      "epoch": 0.7585415604283529,
+      "grad_norm": 0.42825251817703247,
+      "learning_rate": 3.0234799624504008e-06,
+      "loss": 0.43868499994277954,
+      "step": 2975
+    },
+    {
+      "epoch": 0.7587965323814381,
+      "grad_norm": 0.45960578322410583,
+      "learning_rate": 3.0174401555475884e-06,
+      "loss": 0.4422037601470947,
+      "step": 2976
+    },
+    {
+      "epoch": 0.7590515043345232,
+      "grad_norm": 0.45203688740730286,
+      "learning_rate": 3.0114053152750557e-06,
+      "loss": 0.4476014971733093,
+      "step": 2977
+    },
+    {
+      "epoch": 0.7593064762876084,
+      "grad_norm": 0.4353548586368561,
+      "learning_rate": 3.0053754459253338e-06,
+      "loss": 0.4427911937236786,
+      "step": 2978
+    },
+    {
+      "epoch": 0.7595614482406935,
+      "grad_norm": 0.4374661147594452,
+      "learning_rate": 2.9993505517874043e-06,
+      "loss": 0.4469697177410126,
+      "step": 2979
+    },
+    {
+      "epoch": 0.7598164201937787,
+      "grad_norm": 0.42508962750434875,
+      "learning_rate": 2.9933306371467276e-06,
+      "loss": 0.44563043117523193,
+      "step": 2980
+    },
+    {
+      "epoch": 0.7600713921468638,
+      "grad_norm": 0.4326213300228119,
+      "learning_rate": 2.987315706285209e-06,
+      "loss": 0.4384555220603943,
+      "step": 2981
+    },
+    {
+      "epoch": 0.760326364099949,
+      "grad_norm": 0.4297401010990143,
+      "learning_rate": 2.9813057634812104e-06,
+      "loss": 0.4502851366996765,
+      "step": 2982
+    },
+    {
+      "epoch": 0.7605813360530341,
+      "grad_norm": 0.4279881715774536,
+      "learning_rate": 2.975300813009554e-06,
+      "loss": 0.4402778148651123,
+      "step": 2983
+    },
+    {
+      "epoch": 0.7608363080061193,
+      "grad_norm": 0.43452244997024536,
+      "learning_rate": 2.9693008591414998e-06,
+      "loss": 0.43243128061294556,
+      "step": 2984
+    },
+    {
+      "epoch": 0.7610912799592044,
+      "grad_norm": 0.4346815049648285,
+      "learning_rate": 2.963305906144763e-06,
+      "loss": 0.4579737186431885,
+      "step": 2985
+    },
+    {
+      "epoch": 0.7613462519122897,
+      "grad_norm": 0.44520387053489685,
+      "learning_rate": 2.957315958283504e-06,
+      "loss": 0.4395790994167328,
+      "step": 2986
+    },
+    {
+      "epoch": 0.7616012238653748,
+      "grad_norm": 0.42731431126594543,
+      "learning_rate": 2.9513310198183067e-06,
+      "loss": 0.446267694234848,
+      "step": 2987
+    },
+    {
+      "epoch": 0.76185619581846,
+      "grad_norm": 0.449008584022522,
+      "learning_rate": 2.9453510950062057e-06,
+      "loss": 0.45317041873931885,
+      "step": 2988
+    },
+    {
+      "epoch": 0.7621111677715451,
+      "grad_norm": 0.4196610748767853,
+      "learning_rate": 2.939376188100671e-06,
+      "loss": 0.4390903413295746,
+      "step": 2989
+    },
+    {
+      "epoch": 0.7623661397246303,
+      "grad_norm": 0.4226979613304138,
+      "learning_rate": 2.9334063033515926e-06,
+      "loss": 0.4473586976528168,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7626211116777154,
+      "grad_norm": 0.44532281160354614,
+      "learning_rate": 2.9274414450053003e-06,
+      "loss": 0.42633652687072754,
+      "step": 2991
+    },
+    {
+      "epoch": 0.7628760836308006,
+      "grad_norm": 0.4505890905857086,
+      "learning_rate": 2.921481617304536e-06,
+      "loss": 0.4464772641658783,
+      "step": 2992
+    },
+    {
+      "epoch": 0.7631310555838857,
+      "grad_norm": 0.46233832836151123,
+      "learning_rate": 2.9155268244884773e-06,
+      "loss": 0.43518686294555664,
+      "step": 2993
+    },
+    {
+      "epoch": 0.7633860275369709,
+      "grad_norm": 0.4125634729862213,
+      "learning_rate": 2.9095770707927086e-06,
+      "loss": 0.4407936930656433,
+      "step": 2994
+    },
+    {
+      "epoch": 0.7636409994900561,
+      "grad_norm": 0.4293194115161896,
+      "learning_rate": 2.9036323604492333e-06,
+      "loss": 0.43980568647384644,
+      "step": 2995
+    },
+    {
+      "epoch": 0.7638959714431413,
+      "grad_norm": 0.434973806142807,
+      "learning_rate": 2.897692697686472e-06,
+      "loss": 0.426477313041687,
+      "step": 2996
+    },
+    {
+      "epoch": 0.7641509433962265,
+      "grad_norm": 0.4353492259979248,
+      "learning_rate": 2.891758086729253e-06,
+      "loss": 0.43505096435546875,
+      "step": 2997
+    },
+    {
+      "epoch": 0.7644059153493116,
+      "grad_norm": 0.4570540189743042,
+      "learning_rate": 2.8858285317988065e-06,
+      "loss": 0.4340648055076599,
+      "step": 2998
+    },
+    {
+      "epoch": 0.7646608873023968,
+      "grad_norm": 0.43070581555366516,
+      "learning_rate": 2.8799040371127772e-06,
+      "loss": 0.4464573264122009,
+      "step": 2999
+    },
+    {
+      "epoch": 0.7649158592554819,
+      "grad_norm": 0.4650447964668274,
+      "learning_rate": 2.873984606885193e-06,
+      "loss": 0.4371274411678314,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7651708312085671,
+      "grad_norm": 0.4581238031387329,
+      "learning_rate": 2.8680702453264953e-06,
+      "loss": 0.44181880354881287,
+      "step": 3001
+    },
+    {
+      "epoch": 0.7654258031616522,
+      "grad_norm": 0.42868974804878235,
+      "learning_rate": 2.8621609566435173e-06,
+      "loss": 0.4534147381782532,
+      "step": 3002
+    },
+    {
+      "epoch": 0.7656807751147374,
+      "grad_norm": 0.4373212456703186,
+      "learning_rate": 2.856256745039476e-06,
+      "loss": 0.459540456533432,
+      "step": 3003
+    },
+    {
+      "epoch": 0.7659357470678225,
+      "grad_norm": 0.432050496339798,
+      "learning_rate": 2.8503576147139887e-06,
+      "loss": 0.4405026435852051,
+      "step": 3004
+    },
+    {
+      "epoch": 0.7661907190209077,
+      "grad_norm": 0.46295300126075745,
+      "learning_rate": 2.8444635698630464e-06,
+      "loss": 0.43494483828544617,
+      "step": 3005
+    },
+    {
+      "epoch": 0.7664456909739928,
+      "grad_norm": 0.43868863582611084,
+      "learning_rate": 2.838574614679034e-06,
+      "loss": 0.4434785544872284,
+      "step": 3006
+    },
+    {
+      "epoch": 0.7667006629270781,
+      "grad_norm": 0.451294481754303,
+      "learning_rate": 2.8326907533507074e-06,
+      "loss": 0.4428268373012543,
+      "step": 3007
+    },
+    {
+      "epoch": 0.7669556348801632,
+      "grad_norm": 0.43986567854881287,
+      "learning_rate": 2.826811990063201e-06,
+      "loss": 0.44829750061035156,
+      "step": 3008
+    },
+    {
+      "epoch": 0.7672106068332484,
+      "grad_norm": 0.45006152987480164,
+      "learning_rate": 2.8209383289980272e-06,
+      "loss": 0.4289799630641937,
+      "step": 3009
+    },
+    {
+      "epoch": 0.7674655787863335,
+      "grad_norm": 0.45366767048835754,
+      "learning_rate": 2.8150697743330713e-06,
+      "loss": 0.42882341146469116,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7677205507394187,
+      "grad_norm": 0.43628907203674316,
+      "learning_rate": 2.8092063302425733e-06,
+      "loss": 0.43587619066238403,
+      "step": 3011
+    },
+    {
+      "epoch": 0.7679755226925038,
+      "grad_norm": 0.4418754577636719,
+      "learning_rate": 2.803348000897155e-06,
+      "loss": 0.44426703453063965,
+      "step": 3012
+    },
+    {
+      "epoch": 0.768230494645589,
+      "grad_norm": 0.44897258281707764,
+      "learning_rate": 2.7974947904637872e-06,
+      "loss": 0.43549928069114685,
+      "step": 3013
+    },
+    {
+      "epoch": 0.7684854665986741,
+      "grad_norm": 0.4471312463283539,
+      "learning_rate": 2.7916467031058027e-06,
+      "loss": 0.4412408769130707,
+      "step": 3014
+    },
+    {
+      "epoch": 0.7687404385517593,
+      "grad_norm": 0.43710899353027344,
+      "learning_rate": 2.7858037429828978e-06,
+      "loss": 0.4489285349845886,
+      "step": 3015
+    },
+    {
+      "epoch": 0.7689954105048444,
+      "grad_norm": 0.4264802932739258,
+      "learning_rate": 2.779965914251109e-06,
+      "loss": 0.43275749683380127,
+      "step": 3016
+    },
+    {
+      "epoch": 0.7692503824579296,
+      "grad_norm": 0.42677900195121765,
+      "learning_rate": 2.7741332210628346e-06,
+      "loss": 0.44439825415611267,
+      "step": 3017
+    },
+    {
+      "epoch": 0.7695053544110148,
+      "grad_norm": 0.4425060749053955,
+      "learning_rate": 2.7683056675668206e-06,
+      "loss": 0.439004123210907,
+      "step": 3018
+    },
+    {
+      "epoch": 0.7697603263641,
+      "grad_norm": 0.43179410696029663,
+      "learning_rate": 2.7624832579081396e-06,
+      "loss": 0.45458805561065674,
+      "step": 3019
+    },
+    {
+      "epoch": 0.7700152983171851,
+      "grad_norm": 0.4506986737251282,
+      "learning_rate": 2.756665996228224e-06,
+      "loss": 0.4397001266479492,
+      "step": 3020
+    },
+    {
+      "epoch": 0.7702702702702703,
+      "grad_norm": 0.5024133324623108,
+      "learning_rate": 2.7508538866648417e-06,
+      "loss": 0.413383811712265,
+      "step": 3021
+    },
+    {
+      "epoch": 0.7705252422233554,
+      "grad_norm": 0.4732454717159271,
+      "learning_rate": 2.7450469333520856e-06,
+      "loss": 0.44492125511169434,
+      "step": 3022
+    },
+    {
+      "epoch": 0.7707802141764406,
+      "grad_norm": 0.4141547679901123,
+      "learning_rate": 2.7392451404203945e-06,
+      "loss": 0.4385744035243988,
+      "step": 3023
+    },
+    {
+      "epoch": 0.7710351861295257,
+      "grad_norm": 0.4609329402446747,
+      "learning_rate": 2.7334485119965235e-06,
+      "loss": 0.4328097999095917,
+      "step": 3024
+    },
+    {
+      "epoch": 0.7712901580826109,
+      "grad_norm": 0.45164668560028076,
+      "learning_rate": 2.7276570522035672e-06,
+      "loss": 0.42820605635643005,
+      "step": 3025
+    },
+    {
+      "epoch": 0.771545130035696,
+      "grad_norm": 0.4409070611000061,
+      "learning_rate": 2.7218707651609357e-06,
+      "loss": 0.4500933289527893,
+      "step": 3026
+    },
+    {
+      "epoch": 0.7718001019887812,
+      "grad_norm": 0.4577195942401886,
+      "learning_rate": 2.716089654984356e-06,
+      "loss": 0.4405101239681244,
+      "step": 3027
+    },
+    {
+      "epoch": 0.7720550739418663,
+      "grad_norm": 0.4214123487472534,
+      "learning_rate": 2.7103137257858867e-06,
+      "loss": 0.4437384009361267,
+      "step": 3028
+    },
+    {
+      "epoch": 0.7723100458949516,
+      "grad_norm": 0.43069449067115784,
+      "learning_rate": 2.7045429816738855e-06,
+      "loss": 0.45080098509788513,
+      "step": 3029
+    },
+    {
+      "epoch": 0.7725650178480367,
+      "grad_norm": 0.4492934048175812,
+      "learning_rate": 2.698777426753033e-06,
+      "loss": 0.4323241710662842,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7728199898011219,
+      "grad_norm": 0.4298577308654785,
+      "learning_rate": 2.6930170651243206e-06,
+      "loss": 0.43868204951286316,
+      "step": 3031
+    },
+    {
+      "epoch": 0.7730749617542071,
+      "grad_norm": 0.4327729046344757,
+      "learning_rate": 2.6872619008850274e-06,
+      "loss": 0.44194912910461426,
+      "step": 3032
+    },
+    {
+      "epoch": 0.7733299337072922,
+      "grad_norm": 0.44278308749198914,
+      "learning_rate": 2.681511938128757e-06,
+      "loss": 0.42782241106033325,
+      "step": 3033
+    },
+    {
+      "epoch": 0.7735849056603774,
+      "grad_norm": 0.4500921070575714,
+      "learning_rate": 2.6757671809454055e-06,
+      "loss": 0.42661377787590027,
+      "step": 3034
+    },
+    {
+      "epoch": 0.7738398776134625,
+      "grad_norm": 0.43823516368865967,
+      "learning_rate": 2.6700276334211605e-06,
+      "loss": 0.4408743381500244,
+      "step": 3035
+    },
+    {
+      "epoch": 0.7740948495665477,
+      "grad_norm": 0.4439128041267395,
+      "learning_rate": 2.6642932996385163e-06,
+      "loss": 0.43246230483055115,
+      "step": 3036
+    },
+    {
+      "epoch": 0.7743498215196328,
+      "grad_norm": 0.4504877030849457,
+      "learning_rate": 2.658564183676243e-06,
+      "loss": 0.4362192153930664,
+      "step": 3037
+    },
+    {
+      "epoch": 0.774604793472718,
+      "grad_norm": 0.4362858831882477,
+      "learning_rate": 2.6528402896094154e-06,
+      "loss": 0.43194517493247986,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7748597654258031,
+      "grad_norm": 0.43476131558418274,
+      "learning_rate": 2.647121621509383e-06,
+      "loss": 0.443441241979599,
+      "step": 3039
+    },
+    {
+      "epoch": 0.7751147373788884,
+      "grad_norm": 0.440054327249527,
+      "learning_rate": 2.64140818344378e-06,
+      "loss": 0.44788044691085815,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7753697093319735,
+      "grad_norm": 0.4809795320034027,
+      "learning_rate": 2.6356999794765226e-06,
+      "loss": 0.45182570815086365,
+      "step": 3041
+    },
+    {
+      "epoch": 0.7756246812850587,
+      "grad_norm": 0.447130024433136,
+      "learning_rate": 2.629997013667808e-06,
+      "loss": 0.44849371910095215,
+      "step": 3042
+    },
+    {
+      "epoch": 0.7758796532381438,
+      "grad_norm": 0.4554246962070465,
+      "learning_rate": 2.624299290074097e-06,
+      "loss": 0.44359302520751953,
+      "step": 3043
+    },
+    {
+      "epoch": 0.776134625191229,
+      "grad_norm": 0.447831392288208,
+      "learning_rate": 2.6186068127481333e-06,
+      "loss": 0.43672287464141846,
+      "step": 3044
+    },
+    {
+      "epoch": 0.7763895971443141,
+      "grad_norm": 0.4739379286766052,
+      "learning_rate": 2.61291958573892e-06,
+      "loss": 0.4372883439064026,
+      "step": 3045
+    },
+    {
+      "epoch": 0.7766445690973993,
+      "grad_norm": 0.45469939708709717,
+      "learning_rate": 2.6072376130917275e-06,
+      "loss": 0.4256613254547119,
+      "step": 3046
+    },
+    {
+      "epoch": 0.7768995410504844,
+      "grad_norm": 0.43982696533203125,
+      "learning_rate": 2.6015608988480956e-06,
+      "loss": 0.44357165694236755,
+      "step": 3047
+    },
+    {
+      "epoch": 0.7771545130035696,
+      "grad_norm": 0.44001829624176025,
+      "learning_rate": 2.595889447045813e-06,
+      "loss": 0.4202490448951721,
+      "step": 3048
+    },
+    {
+      "epoch": 0.7774094849566547,
+      "grad_norm": 0.4493784010410309,
+      "learning_rate": 2.590223261718937e-06,
+      "loss": 0.4398352801799774,
+      "step": 3049
+    },
+    {
+      "epoch": 0.77766445690974,
+      "grad_norm": 0.44004541635513306,
+      "learning_rate": 2.5845623468977687e-06,
+      "loss": 0.43794482946395874,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7779194288628251,
+      "grad_norm": 0.4580850303173065,
+      "learning_rate": 2.5789067066088633e-06,
+      "loss": 0.4361294209957123,
+      "step": 3051
+    },
+    {
+      "epoch": 0.7781744008159103,
+      "grad_norm": 0.45136088132858276,
+      "learning_rate": 2.5732563448750304e-06,
+      "loss": 0.43613120913505554,
+      "step": 3052
+    },
+    {
+      "epoch": 0.7784293727689954,
+      "grad_norm": 0.46002984046936035,
+      "learning_rate": 2.567611265715313e-06,
+      "loss": 0.43921753764152527,
+      "step": 3053
+    },
+    {
+      "epoch": 0.7786843447220806,
+      "grad_norm": 0.46570706367492676,
+      "learning_rate": 2.5619714731450086e-06,
+      "loss": 0.4551700949668884,
+      "step": 3054
+    },
+    {
+      "epoch": 0.7789393166751657,
+      "grad_norm": 0.4437997341156006,
+      "learning_rate": 2.5563369711756526e-06,
+      "loss": 0.45722657442092896,
+      "step": 3055
+    },
+    {
+      "epoch": 0.7791942886282509,
+      "grad_norm": 0.4493846893310547,
+      "learning_rate": 2.550707763815007e-06,
+      "loss": 0.42499202489852905,
+      "step": 3056
+    },
+    {
+      "epoch": 0.779449260581336,
+      "grad_norm": 0.4641744792461395,
+      "learning_rate": 2.5450838550670808e-06,
+      "loss": 0.4430501461029053,
+      "step": 3057
+    },
+    {
+      "epoch": 0.7797042325344212,
+      "grad_norm": 0.4606056213378906,
+      "learning_rate": 2.5394652489321057e-06,
+      "loss": 0.43809938430786133,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7799592044875063,
+      "grad_norm": 0.44892892241477966,
+      "learning_rate": 2.533851949406543e-06,
+      "loss": 0.4352834224700928,
+      "step": 3059
+    },
+    {
+      "epoch": 0.7802141764405915,
+      "grad_norm": 0.43062809109687805,
+      "learning_rate": 2.5282439604830845e-06,
+      "loss": 0.42994245886802673,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7804691483936766,
+      "grad_norm": 0.448716938495636,
+      "learning_rate": 2.5226412861506353e-06,
+      "loss": 0.43783435225486755,
+      "step": 3061
+    },
+    {
+      "epoch": 0.7807241203467619,
+      "grad_norm": 0.4420609176158905,
+      "learning_rate": 2.5170439303943295e-06,
+      "loss": 0.44785523414611816,
+      "step": 3062
+    },
+    {
+      "epoch": 0.780979092299847,
+      "grad_norm": 0.4563617408275604,
+      "learning_rate": 2.5114518971955205e-06,
+      "loss": 0.4503597617149353,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7812340642529322,
+      "grad_norm": 0.4480021595954895,
+      "learning_rate": 2.5058651905317577e-06,
+      "loss": 0.4416431188583374,
+      "step": 3064
+    },
+    {
+      "epoch": 0.7814890362060173,
+      "grad_norm": 0.46790507435798645,
+      "learning_rate": 2.5002838143768195e-06,
+      "loss": 0.44542765617370605,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7817440081591025,
+      "grad_norm": 0.44628477096557617,
+      "learning_rate": 2.494707772700691e-06,
+      "loss": 0.4343700408935547,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7819989801121877,
+      "grad_norm": 0.4471970796585083,
+      "learning_rate": 2.4891370694695516e-06,
+      "loss": 0.4321914315223694,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7822539520652728,
+      "grad_norm": 0.46172577142715454,
+      "learning_rate": 2.4835717086457988e-06,
+      "loss": 0.4497830271720886,
+      "step": 3068
+    },
+    {
+      "epoch": 0.782508924018358,
+      "grad_norm": 0.489749938249588,
+      "learning_rate": 2.478011694188015e-06,
+      "loss": 0.45339810848236084,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7827638959714431,
+      "grad_norm": 0.4978252649307251,
+      "learning_rate": 2.472457030050994e-06,
+      "loss": 0.4448505640029907,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7830188679245284,
+      "grad_norm": 0.4339428246021271,
+      "learning_rate": 2.4669077201857138e-06,
+      "loss": 0.435431569814682,
+      "step": 3071
+    },
+    {
+      "epoch": 0.7832738398776135,
+      "grad_norm": 0.4410724937915802,
+      "learning_rate": 2.4613637685393433e-06,
+      "loss": 0.45155227184295654,
+      "step": 3072
+    },
+    {
+      "epoch": 0.7835288118306987,
+      "grad_norm": 0.458036333322525,
+      "learning_rate": 2.455825179055249e-06,
+      "loss": 0.44565317034721375,
+      "step": 3073
+    },
+    {
+      "epoch": 0.7837837837837838,
+      "grad_norm": 0.42787349224090576,
+      "learning_rate": 2.4502919556729798e-06,
+      "loss": 0.43534278869628906,
+      "step": 3074
+    },
+    {
+      "epoch": 0.784038755736869,
+      "grad_norm": 0.45477649569511414,
+      "learning_rate": 2.4447641023282607e-06,
+      "loss": 0.44423800706863403,
+      "step": 3075
+    },
+    {
+      "epoch": 0.7842937276899541,
+      "grad_norm": 0.43623751401901245,
+      "learning_rate": 2.4392416229530093e-06,
+      "loss": 0.4423667788505554,
+      "step": 3076
+    },
+    {
+      "epoch": 0.7845486996430393,
+      "grad_norm": 0.4349161684513092,
+      "learning_rate": 2.4337245214753104e-06,
+      "loss": 0.44194480776786804,
+      "step": 3077
+    },
+    {
+      "epoch": 0.7848036715961244,
+      "grad_norm": 0.47349271178245544,
+      "learning_rate": 2.428212801819424e-06,
+      "loss": 0.432020366191864,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7850586435492096,
+      "grad_norm": 0.45997923612594604,
+      "learning_rate": 2.422706467905792e-06,
+      "loss": 0.42831337451934814,
+      "step": 3079
+    },
+    {
+      "epoch": 0.7853136155022947,
+      "grad_norm": 0.4358552396297455,
+      "learning_rate": 2.4172055236510127e-06,
+      "loss": 0.44012266397476196,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7855685874553799,
+      "grad_norm": 0.44077014923095703,
+      "learning_rate": 2.4117099729678638e-06,
+      "loss": 0.4489743113517761,
+      "step": 3081
+    },
+    {
+      "epoch": 0.785823559408465,
+      "grad_norm": 0.447301983833313,
+      "learning_rate": 2.406219819765275e-06,
+      "loss": 0.42973965406417847,
+      "step": 3082
+    },
+    {
+      "epoch": 0.7860785313615503,
+      "grad_norm": 0.4363212287425995,
+      "learning_rate": 2.4007350679483398e-06,
+      "loss": 0.4414937496185303,
+      "step": 3083
+    },
+    {
+      "epoch": 0.7863335033146354,
+      "grad_norm": 0.4348331391811371,
+      "learning_rate": 2.395255721418317e-06,
+      "loss": 0.4441762864589691,
+      "step": 3084
+    },
+    {
+      "epoch": 0.7865884752677206,
+      "grad_norm": 0.4310491383075714,
+      "learning_rate": 2.389781784072609e-06,
+      "loss": 0.4399312436580658,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7868434472208057,
+      "grad_norm": 0.46151554584503174,
+      "learning_rate": 2.38431325980478e-06,
+      "loss": 0.42329296469688416,
+      "step": 3086
+    },
+    {
+      "epoch": 0.7870984191738909,
+      "grad_norm": 0.4397614002227783,
+      "learning_rate": 2.378850152504544e-06,
+      "loss": 0.41856569051742554,
+      "step": 3087
+    },
+    {
+      "epoch": 0.787353391126976,
+      "grad_norm": 0.4383176565170288,
+      "learning_rate": 2.3733924660577524e-06,
+      "loss": 0.44408220052719116,
+      "step": 3088
+    },
+    {
+      "epoch": 0.7876083630800612,
+      "grad_norm": 0.45403745770454407,
+      "learning_rate": 2.367940204346414e-06,
+      "loss": 0.43578648567199707,
+      "step": 3089
+    },
+    {
+      "epoch": 0.7878633350331463,
+      "grad_norm": 0.43478861451148987,
+      "learning_rate": 2.3624933712486673e-06,
+      "loss": 0.43671154975891113,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7881183069862315,
+      "grad_norm": 0.4544563889503479,
+      "learning_rate": 2.3570519706387928e-06,
+      "loss": 0.43795761466026306,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7883732789393166,
+      "grad_norm": 0.4508332908153534,
+      "learning_rate": 2.351616006387214e-06,
+      "loss": 0.4364631175994873,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7886282508924018,
+      "grad_norm": 0.44363948702812195,
+      "learning_rate": 2.3461854823604766e-06,
+      "loss": 0.4498422145843506,
+      "step": 3093
+    },
+    {
+      "epoch": 0.788883222845487,
+      "grad_norm": 0.452188104391098,
+      "learning_rate": 2.3407604024212636e-06,
+      "loss": 0.44089120626449585,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7891381947985722,
+      "grad_norm": 0.45992568135261536,
+      "learning_rate": 2.3353407704283926e-06,
+      "loss": 0.43927067518234253,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7893931667516573,
+      "grad_norm": 0.44867175817489624,
+      "learning_rate": 2.329926590236784e-06,
+      "loss": 0.4402957558631897,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7896481387047425,
+      "grad_norm": 0.4226747751235962,
+      "learning_rate": 2.324517865697501e-06,
+      "loss": 0.4406220316886902,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7899031106578276,
+      "grad_norm": 0.45604899525642395,
+      "learning_rate": 2.3191146006577202e-06,
+      "loss": 0.44700348377227783,
+      "step": 3098
+    },
+    {
+      "epoch": 0.7901580826109128,
+      "grad_norm": 0.42548590898513794,
+      "learning_rate": 2.3137167989607324e-06,
+      "loss": 0.4498402774333954,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7904130545639979,
+      "grad_norm": 1.5527485609054565,
+      "learning_rate": 2.3083244644459457e-06,
+      "loss": 0.4142894744873047,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7906680265170831,
+      "grad_norm": 0.41954970359802246,
+      "learning_rate": 2.302937600948877e-06,
+      "loss": 0.4311685562133789,
+      "step": 3101
+    },
+    {
+      "epoch": 0.7909229984701683,
+      "grad_norm": 0.435149610042572,
+      "learning_rate": 2.2975562123011497e-06,
+      "loss": 0.45557522773742676,
+      "step": 3102
+    },
+    {
+      "epoch": 0.7911779704232534,
+      "grad_norm": 0.43851983547210693,
+      "learning_rate": 2.2921803023305e-06,
+      "loss": 0.43911582231521606,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7914329423763387,
+      "grad_norm": 0.43195638060569763,
+      "learning_rate": 2.2868098748607594e-06,
+      "loss": 0.44874662160873413,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7916879143294238,
+      "grad_norm": 0.42538851499557495,
+      "learning_rate": 2.281444933711867e-06,
+      "loss": 0.42983171343803406,
+      "step": 3105
+    },
+    {
+      "epoch": 0.791942886282509,
+      "grad_norm": 0.45717546343803406,
+      "learning_rate": 2.2760854826998523e-06,
+      "loss": 0.4311561584472656,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7921978582355941,
+      "grad_norm": 0.43942540884017944,
+      "learning_rate": 2.2707315256368434e-06,
+      "loss": 0.4412494897842407,
+      "step": 3107
+    },
+    {
+      "epoch": 0.7924528301886793,
+      "grad_norm": 0.4405248761177063,
+      "learning_rate": 2.2653830663310693e-06,
+      "loss": 0.4397597014904022,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7927078021417644,
+      "grad_norm": 0.4440111517906189,
+      "learning_rate": 2.2600401085868263e-06,
+      "loss": 0.4302397668361664,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7929627740948496,
+      "grad_norm": 0.4543129503726959,
+      "learning_rate": 2.254702656204516e-06,
+      "loss": 0.4415404796600342,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7932177460479347,
+      "grad_norm": 0.4537447988986969,
+      "learning_rate": 2.249370712980624e-06,
+      "loss": 0.4310953617095947,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7934727180010199,
+      "grad_norm": 0.4390498399734497,
+      "learning_rate": 2.244044282707705e-06,
+      "loss": 0.4310564696788788,
+      "step": 3112
+    },
+    {
+      "epoch": 0.793727689954105,
+      "grad_norm": 0.4333949387073517,
+      "learning_rate": 2.2387233691744047e-06,
+      "loss": 0.4509139657020569,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7939826619071902,
+      "grad_norm": 0.4292167127132416,
+      "learning_rate": 2.233407976165438e-06,
+      "loss": 0.43772897124290466,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7942376338602753,
+      "grad_norm": 0.4735633134841919,
+      "learning_rate": 2.2280981074615926e-06,
+      "loss": 0.445385217666626,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7944926058133606,
+      "grad_norm": 0.42386460304260254,
+      "learning_rate": 2.222793766839734e-06,
+      "loss": 0.4210456609725952,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7947475777664457,
+      "grad_norm": 0.4557611346244812,
+      "learning_rate": 2.2174949580727835e-06,
+      "loss": 0.4388166666030884,
+      "step": 3117
+    },
+    {
+      "epoch": 0.7950025497195309,
+      "grad_norm": 0.4518273174762726,
+      "learning_rate": 2.21220168492974e-06,
+      "loss": 0.43927669525146484,
+      "step": 3118
+    },
+    {
+      "epoch": 0.795257521672616,
+      "grad_norm": 0.4459928870201111,
+      "learning_rate": 2.206913951175661e-06,
+      "loss": 0.4367228150367737,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7955124936257012,
+      "grad_norm": 0.444635272026062,
+      "learning_rate": 2.2016317605716577e-06,
+      "loss": 0.4311097264289856,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7957674655787863,
+      "grad_norm": 0.4263254702091217,
+      "learning_rate": 2.1963551168749097e-06,
+      "loss": 0.44298744201660156,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7960224375318715,
+      "grad_norm": 0.4463154077529907,
+      "learning_rate": 2.19108402383864e-06,
+      "loss": 0.43678268790245056,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7962774094849566,
+      "grad_norm": 0.43629640340805054,
+      "learning_rate": 2.185818485212128e-06,
+      "loss": 0.4486580491065979,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7965323814380418,
+      "grad_norm": 0.44561508297920227,
+      "learning_rate": 2.1805585047407063e-06,
+      "loss": 0.45028600096702576,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7967873533911269,
+      "grad_norm": 0.4386787712574005,
+      "learning_rate": 2.175304086165747e-06,
+      "loss": 0.43141692876815796,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7970423253442122,
+      "grad_norm": 0.4359768033027649,
+      "learning_rate": 2.1700552332246695e-06,
+      "loss": 0.4307355284690857,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7972972972972973,
+      "grad_norm": 0.4674769341945648,
+      "learning_rate": 2.1648119496509423e-06,
+      "loss": 0.4411862790584564,
+      "step": 3127
+    },
+    {
+      "epoch": 0.7975522692503825,
+      "grad_norm": 0.436692476272583,
+      "learning_rate": 2.1595742391740516e-06,
+      "loss": 0.43730098009109497,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7978072412034676,
+      "grad_norm": 0.4336155354976654,
+      "learning_rate": 2.1543421055195425e-06,
+      "loss": 0.43345558643341064,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7980622131565528,
+      "grad_norm": 0.43793782591819763,
+      "learning_rate": 2.1491155524089767e-06,
+      "loss": 0.43637627363204956,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7983171851096379,
+      "grad_norm": 0.43755635619163513,
+      "learning_rate": 2.1438945835599556e-06,
+      "loss": 0.44143471121788025,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7985721570627231,
+      "grad_norm": 0.45615267753601074,
+      "learning_rate": 2.1386792026861103e-06,
+      "loss": 0.4418499171733856,
+      "step": 3132
+    },
+    {
+      "epoch": 0.7988271290158082,
+      "grad_norm": 0.45956259965896606,
+      "learning_rate": 2.1334694134970892e-06,
+      "loss": 0.43504881858825684,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7990821009688934,
+      "grad_norm": 0.46168217062950134,
+      "learning_rate": 2.1282652196985643e-06,
+      "loss": 0.426724374294281,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7993370729219785,
+      "grad_norm": 0.4162600338459015,
+      "learning_rate": 2.123066624992236e-06,
+      "loss": 0.4324457049369812,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7995920448750637,
+      "grad_norm": 0.43563950061798096,
+      "learning_rate": 2.117873633075812e-06,
+      "loss": 0.4513908326625824,
+      "step": 3136
+    },
+    {
+      "epoch": 0.7998470168281488,
+      "grad_norm": 0.4455576539039612,
+      "learning_rate": 2.1126862476430245e-06,
+      "loss": 0.43949684500694275,
+      "step": 3137
+    },
+    {
+      "epoch": 0.8001019887812341,
+      "grad_norm": 0.44459694623947144,
+      "learning_rate": 2.107504472383606e-06,
+      "loss": 0.4357977509498596,
+      "step": 3138
+    },
+    {
+      "epoch": 0.8003569607343193,
+      "grad_norm": 0.4605623781681061,
+      "learning_rate": 2.1023283109833106e-06,
+      "loss": 0.429220050573349,
+      "step": 3139
+    },
+    {
+      "epoch": 0.8006119326874044,
+      "grad_norm": 0.43185335397720337,
+      "learning_rate": 2.0971577671238987e-06,
+      "loss": 0.43885111808776855,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8008669046404896,
+      "grad_norm": 0.42108774185180664,
+      "learning_rate": 2.0919928444831174e-06,
+      "loss": 0.4211594760417938,
+      "step": 3141
+    },
+    {
+      "epoch": 0.8011218765935747,
+      "grad_norm": 0.47686824202537537,
+      "learning_rate": 2.0868335467347367e-06,
+      "loss": 0.4349004924297333,
+      "step": 3142
+    },
+    {
+      "epoch": 0.8013768485466599,
+      "grad_norm": 0.45625442266464233,
+      "learning_rate": 2.081679877548519e-06,
+      "loss": 0.43549275398254395,
+      "step": 3143
+    },
+    {
+      "epoch": 0.801631820499745,
+      "grad_norm": 0.43894386291503906,
+      "learning_rate": 2.0765318405902158e-06,
+      "loss": 0.4453749656677246,
+      "step": 3144
+    },
+    {
+      "epoch": 0.8018867924528302,
+      "grad_norm": 0.44987383484840393,
+      "learning_rate": 2.071389439521584e-06,
+      "loss": 0.4443401098251343,
+      "step": 3145
+    },
+    {
+      "epoch": 0.8021417644059153,
+      "grad_norm": 0.42041513323783875,
+      "learning_rate": 2.066252678000362e-06,
+      "loss": 0.4356636703014374,
+      "step": 3146
+    },
+    {
+      "epoch": 0.8023967363590006,
+      "grad_norm": 0.46287089586257935,
+      "learning_rate": 2.06112155968028e-06,
+      "loss": 0.44034093618392944,
+      "step": 3147
+    },
+    {
+      "epoch": 0.8026517083120857,
+      "grad_norm": 0.44690874218940735,
+      "learning_rate": 2.055996088211061e-06,
+      "loss": 0.4372246265411377,
+      "step": 3148
+    },
+    {
+      "epoch": 0.8029066802651709,
+      "grad_norm": 0.4576677978038788,
+      "learning_rate": 2.0508762672383975e-06,
+      "loss": 0.43028396368026733,
+      "step": 3149
+    },
+    {
+      "epoch": 0.803161652218256,
+      "grad_norm": 0.46135053038597107,
+      "learning_rate": 2.045762100403975e-06,
+      "loss": 0.4304881691932678,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8034166241713412,
+      "grad_norm": 0.43473130464553833,
+      "learning_rate": 2.040653591345455e-06,
+      "loss": 0.4307914674282074,
+      "step": 3151
+    },
+    {
+      "epoch": 0.8036715961244263,
+      "grad_norm": 0.42412862181663513,
+      "learning_rate": 2.0355507436964684e-06,
+      "loss": 0.4323742389678955,
+      "step": 3152
+    },
+    {
+      "epoch": 0.8039265680775115,
+      "grad_norm": 0.4608285129070282,
+      "learning_rate": 2.0304535610866285e-06,
+      "loss": 0.4228378236293793,
+      "step": 3153
+    },
+    {
+      "epoch": 0.8041815400305966,
+      "grad_norm": 0.44225743412971497,
+      "learning_rate": 2.0253620471415115e-06,
+      "loss": 0.45546650886535645,
+      "step": 3154
+    },
+    {
+      "epoch": 0.8044365119836818,
+      "grad_norm": 0.4197104573249817,
+      "learning_rate": 2.0202762054826617e-06,
+      "loss": 0.4291185736656189,
+      "step": 3155
+    },
+    {
+      "epoch": 0.8046914839367669,
+      "grad_norm": 0.4479086101055145,
+      "learning_rate": 2.015196039727597e-06,
+      "loss": 0.43910154700279236,
+      "step": 3156
+    },
+    {
+      "epoch": 0.8049464558898521,
+      "grad_norm": 0.4188553988933563,
+      "learning_rate": 2.0101215534897856e-06,
+      "loss": 0.4407055377960205,
+      "step": 3157
+    },
+    {
+      "epoch": 0.8052014278429372,
+      "grad_norm": 0.4461323022842407,
+      "learning_rate": 2.0050527503786686e-06,
+      "loss": 0.4346162676811218,
+      "step": 3158
+    },
+    {
+      "epoch": 0.8054563997960225,
+      "grad_norm": 0.4502837061882019,
+      "learning_rate": 1.9999896339996373e-06,
+      "loss": 0.4362154006958008,
+      "step": 3159
+    },
+    {
+      "epoch": 0.8057113717491076,
+      "grad_norm": 0.45032528042793274,
+      "learning_rate": 1.9949322079540354e-06,
+      "loss": 0.4282383322715759,
+      "step": 3160
+    },
+    {
+      "epoch": 0.8059663437021928,
+      "grad_norm": 0.4429026246070862,
+      "learning_rate": 1.9898804758391697e-06,
+      "loss": 0.4364677369594574,
+      "step": 3161
+    },
+    {
+      "epoch": 0.8062213156552779,
+      "grad_norm": 0.45870086550712585,
+      "learning_rate": 1.9848344412482856e-06,
+      "loss": 0.4519317150115967,
+      "step": 3162
+    },
+    {
+      "epoch": 0.8064762876083631,
+      "grad_norm": 0.4462481141090393,
+      "learning_rate": 1.979794107770582e-06,
+      "loss": 0.4192603528499603,
+      "step": 3163
+    },
+    {
+      "epoch": 0.8067312595614482,
+      "grad_norm": 0.4347079396247864,
+      "learning_rate": 1.9747594789912038e-06,
+      "loss": 0.45258253812789917,
+      "step": 3164
+    },
+    {
+      "epoch": 0.8069862315145334,
+      "grad_norm": 0.42841705679893494,
+      "learning_rate": 1.969730558491235e-06,
+      "loss": 0.4304022490978241,
+      "step": 3165
+    },
+    {
+      "epoch": 0.8072412034676185,
+      "grad_norm": 0.42129963636398315,
+      "learning_rate": 1.9647073498476977e-06,
+      "loss": 0.4637308120727539,
+      "step": 3166
+    },
+    {
+      "epoch": 0.8074961754207037,
+      "grad_norm": 0.4346625506877899,
+      "learning_rate": 1.9596898566335575e-06,
+      "loss": 0.43997952342033386,
+      "step": 3167
+    },
+    {
+      "epoch": 0.8077511473737888,
+      "grad_norm": 0.44551050662994385,
+      "learning_rate": 1.954678082417707e-06,
+      "loss": 0.4464991092681885,
+      "step": 3168
+    },
+    {
+      "epoch": 0.808006119326874,
+      "grad_norm": 0.44049108028411865,
+      "learning_rate": 1.9496720307649797e-06,
+      "loss": 0.4449331760406494,
+      "step": 3169
+    },
+    {
+      "epoch": 0.8082610912799592,
+      "grad_norm": 0.48131170868873596,
+      "learning_rate": 1.9446717052361286e-06,
+      "loss": 0.44099175930023193,
+      "step": 3170
+    },
+    {
+      "epoch": 0.8085160632330444,
+      "grad_norm": 0.44364720582962036,
+      "learning_rate": 1.939677109387841e-06,
+      "loss": 0.4477325677871704,
+      "step": 3171
+    },
+    {
+      "epoch": 0.8087710351861295,
+      "grad_norm": 0.422466903924942,
+      "learning_rate": 1.9346882467727323e-06,
+      "loss": 0.43594223260879517,
+      "step": 3172
+    },
+    {
+      "epoch": 0.8090260071392147,
+      "grad_norm": 0.4314025938510895,
+      "learning_rate": 1.929705120939326e-06,
+      "loss": 0.434553325176239,
+      "step": 3173
+    },
+    {
+      "epoch": 0.8092809790922999,
+      "grad_norm": 0.45429590344429016,
+      "learning_rate": 1.924727735432076e-06,
+      "loss": 0.44354602694511414,
+      "step": 3174
+    },
+    {
+      "epoch": 0.809535951045385,
+      "grad_norm": 0.4488116502761841,
+      "learning_rate": 1.9197560937913538e-06,
+      "loss": 0.4514884650707245,
+      "step": 3175
+    },
+    {
+      "epoch": 0.8097909229984702,
+      "grad_norm": 0.43512481451034546,
+      "learning_rate": 1.914790199553437e-06,
+      "loss": 0.43873876333236694,
+      "step": 3176
+    },
+    {
+      "epoch": 0.8100458949515553,
+      "grad_norm": 0.42961785197257996,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.4447658956050873,
+      "step": 3177
+    },
+    {
+      "epoch": 0.8103008669046405,
+      "grad_norm": 0.43365493416786194,
+      "learning_rate": 1.9048756674107228e-06,
+      "loss": 0.4348711669445038,
+      "step": 3178
+    },
+    {
+      "epoch": 0.8105558388577256,
+      "grad_norm": 0.4372931718826294,
+      "learning_rate": 1.8999270365580347e-06,
+      "loss": 0.453824520111084,
+      "step": 3179
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 0.42043957114219666,
+      "learning_rate": 1.8949841672123826e-06,
+      "loss": 0.4466125965118408,
+      "step": 3180
+    },
+    {
+      "epoch": 0.811065782763896,
+      "grad_norm": 0.4353785812854767,
+      "learning_rate": 1.8900470628895783e-06,
+      "loss": 0.44272512197494507,
+      "step": 3181
+    },
+    {
+      "epoch": 0.8113207547169812,
+      "grad_norm": 0.4317415654659271,
+      "learning_rate": 1.8851157271013443e-06,
+      "loss": 0.4332931935787201,
+      "step": 3182
+    },
+    {
+      "epoch": 0.8115757266700663,
+      "grad_norm": 0.4312629997730255,
+      "learning_rate": 1.8801901633552878e-06,
+      "loss": 0.43788594007492065,
+      "step": 3183
+    },
+    {
+      "epoch": 0.8118306986231515,
+      "grad_norm": 0.45187297463417053,
+      "learning_rate": 1.8752703751549207e-06,
+      "loss": 0.4394760727882385,
+      "step": 3184
+    },
+    {
+      "epoch": 0.8120856705762366,
+      "grad_norm": 0.44457483291625977,
+      "learning_rate": 1.8703563659996483e-06,
+      "loss": 0.42946380376815796,
+      "step": 3185
+    },
+    {
+      "epoch": 0.8123406425293218,
+      "grad_norm": 0.47339144349098206,
+      "learning_rate": 1.865448139384748e-06,
+      "loss": 0.4360730051994324,
+      "step": 3186
+    },
+    {
+      "epoch": 0.8125956144824069,
+      "grad_norm": 0.42321452498435974,
+      "learning_rate": 1.8605456988014014e-06,
+      "loss": 0.43954116106033325,
+      "step": 3187
+    },
+    {
+      "epoch": 0.8128505864354921,
+      "grad_norm": 0.4437929689884186,
+      "learning_rate": 1.8556490477366739e-06,
+      "loss": 0.42642879486083984,
+      "step": 3188
+    },
+    {
+      "epoch": 0.8131055583885772,
+      "grad_norm": 0.4351274371147156,
+      "learning_rate": 1.8507581896735004e-06,
+      "loss": 0.43298158049583435,
+      "step": 3189
+    },
+    {
+      "epoch": 0.8133605303416624,
+      "grad_norm": 0.4379459619522095,
+      "learning_rate": 1.8458731280907093e-06,
+      "loss": 0.4267469048500061,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8136155022947476,
+      "grad_norm": 0.43378978967666626,
+      "learning_rate": 1.8409938664629978e-06,
+      "loss": 0.4531802535057068,
+      "step": 3191
+    },
+    {
+      "epoch": 0.8138704742478328,
+      "grad_norm": 0.4308216869831085,
+      "learning_rate": 1.8361204082609353e-06,
+      "loss": 0.44718247652053833,
+      "step": 3192
+    },
+    {
+      "epoch": 0.8141254462009179,
+      "grad_norm": 0.42755618691444397,
+      "learning_rate": 1.8312527569509741e-06,
+      "loss": 0.44727540016174316,
+      "step": 3193
+    },
+    {
+      "epoch": 0.8143804181540031,
+      "grad_norm": 0.4255657196044922,
+      "learning_rate": 1.8263909159954253e-06,
+      "loss": 0.4319005012512207,
+      "step": 3194
+    },
+    {
+      "epoch": 0.8146353901070882,
+      "grad_norm": 0.46537476778030396,
+      "learning_rate": 1.8215348888524709e-06,
+      "loss": 0.42908668518066406,
+      "step": 3195
+    },
+    {
+      "epoch": 0.8148903620601734,
+      "grad_norm": 0.4323773980140686,
+      "learning_rate": 1.8166846789761638e-06,
+      "loss": 0.434833288192749,
+      "step": 3196
+    },
+    {
+      "epoch": 0.8151453340132585,
+      "grad_norm": 0.4433038532733917,
+      "learning_rate": 1.811840289816409e-06,
+      "loss": 0.4452885091304779,
+      "step": 3197
+    },
+    {
+      "epoch": 0.8154003059663437,
+      "grad_norm": 0.43234574794769287,
+      "learning_rate": 1.8070017248189743e-06,
+      "loss": 0.4334978461265564,
+      "step": 3198
+    },
+    {
+      "epoch": 0.8156552779194288,
+      "grad_norm": 0.4315129816532135,
+      "learning_rate": 1.802168987425491e-06,
+      "loss": 0.44864726066589355,
+      "step": 3199
+    },
+    {
+      "epoch": 0.815910249872514,
+      "grad_norm": 0.46172434091567993,
+      "learning_rate": 1.7973420810734354e-06,
+      "loss": 0.43277662992477417,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8161652218255991,
+      "grad_norm": 0.4356657564640045,
+      "learning_rate": 1.7925210091961464e-06,
+      "loss": 0.4305696189403534,
+      "step": 3201
+    },
+    {
+      "epoch": 0.8164201937786844,
+      "grad_norm": 0.4411664307117462,
+      "learning_rate": 1.7877057752228022e-06,
+      "loss": 0.42217326164245605,
+      "step": 3202
+    },
+    {
+      "epoch": 0.8166751657317695,
+      "grad_norm": 0.4304858148097992,
+      "learning_rate": 1.7828963825784385e-06,
+      "loss": 0.44613325595855713,
+      "step": 3203
+    },
+    {
+      "epoch": 0.8169301376848547,
+      "grad_norm": 0.4482412338256836,
+      "learning_rate": 1.7780928346839355e-06,
+      "loss": 0.445767879486084,
+      "step": 3204
+    },
+    {
+      "epoch": 0.8171851096379398,
+      "grad_norm": 0.4683200716972351,
+      "learning_rate": 1.7732951349560024e-06,
+      "loss": 0.4428790509700775,
+      "step": 3205
+    },
+    {
+      "epoch": 0.817440081591025,
+      "grad_norm": 0.4400186538696289,
+      "learning_rate": 1.768503286807206e-06,
+      "loss": 0.4351852536201477,
+      "step": 3206
+    },
+    {
+      "epoch": 0.8176950535441101,
+      "grad_norm": 0.4413929283618927,
+      "learning_rate": 1.763717293645939e-06,
+      "loss": 0.4290766417980194,
+      "step": 3207
+    },
+    {
+      "epoch": 0.8179500254971953,
+      "grad_norm": 0.44811487197875977,
+      "learning_rate": 1.758937158876437e-06,
+      "loss": 0.43229806423187256,
+      "step": 3208
+    },
+    {
+      "epoch": 0.8182049974502805,
+      "grad_norm": 0.44337981939315796,
+      "learning_rate": 1.754162885898768e-06,
+      "loss": 0.42842841148376465,
+      "step": 3209
+    },
+    {
+      "epoch": 0.8184599694033656,
+      "grad_norm": 0.4307434856891632,
+      "learning_rate": 1.7493944781088245e-06,
+      "loss": 0.44062185287475586,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8187149413564508,
+      "grad_norm": 0.45656681060791016,
+      "learning_rate": 1.7446319388983312e-06,
+      "loss": 0.4341960549354553,
+      "step": 3211
+    },
+    {
+      "epoch": 0.818969913309536,
+      "grad_norm": 0.4163688123226166,
+      "learning_rate": 1.7398752716548395e-06,
+      "loss": 0.4489062428474426,
+      "step": 3212
+    },
+    {
+      "epoch": 0.8192248852626212,
+      "grad_norm": 0.4449394941329956,
+      "learning_rate": 1.7351244797617218e-06,
+      "loss": 0.44037866592407227,
+      "step": 3213
+    },
+    {
+      "epoch": 0.8194798572157063,
+      "grad_norm": 0.4564632177352905,
+      "learning_rate": 1.7303795665981738e-06,
+      "loss": 0.4470926523208618,
+      "step": 3214
+    },
+    {
+      "epoch": 0.8197348291687915,
+      "grad_norm": 0.44864529371261597,
+      "learning_rate": 1.725640535539206e-06,
+      "loss": 0.4593150317668915,
+      "step": 3215
+    },
+    {
+      "epoch": 0.8199898011218766,
+      "grad_norm": 0.4310421943664551,
+      "learning_rate": 1.7209073899556504e-06,
+      "loss": 0.4326831102371216,
+      "step": 3216
+    },
+    {
+      "epoch": 0.8202447730749618,
+      "grad_norm": 0.4456369876861572,
+      "learning_rate": 1.7161801332141492e-06,
+      "loss": 0.43467840552330017,
+      "step": 3217
+    },
+    {
+      "epoch": 0.8204997450280469,
+      "grad_norm": 0.45582112669944763,
+      "learning_rate": 1.7114587686771533e-06,
+      "loss": 0.4258680045604706,
+      "step": 3218
+    },
+    {
+      "epoch": 0.8207547169811321,
+      "grad_norm": 0.4281396269798279,
+      "learning_rate": 1.7067432997029265e-06,
+      "loss": 0.4291533827781677,
+      "step": 3219
+    },
+    {
+      "epoch": 0.8210096889342172,
+      "grad_norm": 0.4517004191875458,
+      "learning_rate": 1.7020337296455435e-06,
+      "loss": 0.4399279057979584,
+      "step": 3220
+    },
+    {
+      "epoch": 0.8212646608873024,
+      "grad_norm": 0.4636811912059784,
+      "learning_rate": 1.6973300618548705e-06,
+      "loss": 0.44841358065605164,
+      "step": 3221
+    },
+    {
+      "epoch": 0.8215196328403875,
+      "grad_norm": 0.44340163469314575,
+      "learning_rate": 1.6926322996765899e-06,
+      "loss": 0.42988523840904236,
+      "step": 3222
+    },
+    {
+      "epoch": 0.8217746047934728,
+      "grad_norm": 0.43543392419815063,
+      "learning_rate": 1.687940446452172e-06,
+      "loss": 0.4322485327720642,
+      "step": 3223
+    },
+    {
+      "epoch": 0.8220295767465579,
+      "grad_norm": 0.4247783422470093,
+      "learning_rate": 1.6832545055188886e-06,
+      "loss": 0.4405077397823334,
+      "step": 3224
+    },
+    {
+      "epoch": 0.8222845486996431,
+      "grad_norm": 0.4364262819290161,
+      "learning_rate": 1.678574480209809e-06,
+      "loss": 0.4284090995788574,
+      "step": 3225
+    },
+    {
+      "epoch": 0.8225395206527282,
+      "grad_norm": 0.4634256660938263,
+      "learning_rate": 1.6739003738537874e-06,
+      "loss": 0.43978071212768555,
+      "step": 3226
+    },
+    {
+      "epoch": 0.8227944926058134,
+      "grad_norm": 0.44353532791137695,
+      "learning_rate": 1.669232189775476e-06,
+      "loss": 0.43755969405174255,
+      "step": 3227
+    },
+    {
+      "epoch": 0.8230494645588985,
+      "grad_norm": 0.43164727091789246,
+      "learning_rate": 1.6645699312953123e-06,
+      "loss": 0.44416743516921997,
+      "step": 3228
+    },
+    {
+      "epoch": 0.8233044365119837,
+      "grad_norm": 0.5508279204368591,
+      "learning_rate": 1.6599136017295159e-06,
+      "loss": 0.42011135816574097,
+      "step": 3229
+    },
+    {
+      "epoch": 0.8235594084650688,
+      "grad_norm": 0.4682815670967102,
+      "learning_rate": 1.6552632043900885e-06,
+      "loss": 0.45232391357421875,
+      "step": 3230
+    },
+    {
+      "epoch": 0.823814380418154,
+      "grad_norm": 0.442646324634552,
+      "learning_rate": 1.6506187425848196e-06,
+      "loss": 0.4356235861778259,
+      "step": 3231
+    },
+    {
+      "epoch": 0.8240693523712391,
+      "grad_norm": 0.45952385663986206,
+      "learning_rate": 1.645980219617267e-06,
+      "loss": 0.4367219805717468,
+      "step": 3232
+    },
+    {
+      "epoch": 0.8243243243243243,
+      "grad_norm": 0.44393765926361084,
+      "learning_rate": 1.6413476387867732e-06,
+      "loss": 0.44239887595176697,
+      "step": 3233
+    },
+    {
+      "epoch": 0.8245792962774094,
+      "grad_norm": 0.4498979151248932,
+      "learning_rate": 1.6367210033884474e-06,
+      "loss": 0.4374240040779114,
+      "step": 3234
+    },
+    {
+      "epoch": 0.8248342682304947,
+      "grad_norm": 0.4651869833469391,
+      "learning_rate": 1.6321003167131743e-06,
+      "loss": 0.4354780912399292,
+      "step": 3235
+    },
+    {
+      "epoch": 0.8250892401835798,
+      "grad_norm": 0.42792996764183044,
+      "learning_rate": 1.6274855820476065e-06,
+      "loss": 0.44007962942123413,
+      "step": 3236
+    },
+    {
+      "epoch": 0.825344212136665,
+      "grad_norm": 0.4541287422180176,
+      "learning_rate": 1.622876802674158e-06,
+      "loss": 0.43488869071006775,
+      "step": 3237
+    },
+    {
+      "epoch": 0.8255991840897501,
+      "grad_norm": 0.4381212294101715,
+      "learning_rate": 1.6182739818710158e-06,
+      "loss": 0.4432847499847412,
+      "step": 3238
+    },
+    {
+      "epoch": 0.8258541560428353,
+      "grad_norm": 0.47068390250205994,
+      "learning_rate": 1.6136771229121195e-06,
+      "loss": 0.43802642822265625,
+      "step": 3239
+    },
+    {
+      "epoch": 0.8261091279959204,
+      "grad_norm": 0.4551355242729187,
+      "learning_rate": 1.6090862290671738e-06,
+      "loss": 0.43210625648498535,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8263640999490056,
+      "grad_norm": 0.44001293182373047,
+      "learning_rate": 1.6045013036016422e-06,
+      "loss": 0.44448336958885193,
+      "step": 3241
+    },
+    {
+      "epoch": 0.8266190719020907,
+      "grad_norm": 0.43821027874946594,
+      "learning_rate": 1.5999223497767379e-06,
+      "loss": 0.4449557065963745,
+      "step": 3242
+    },
+    {
+      "epoch": 0.8268740438551759,
+      "grad_norm": 0.4449080526828766,
+      "learning_rate": 1.5953493708494262e-06,
+      "loss": 0.4334353804588318,
+      "step": 3243
+    },
+    {
+      "epoch": 0.827129015808261,
+      "grad_norm": 0.42738792300224304,
+      "learning_rate": 1.590782370072428e-06,
+      "loss": 0.43336939811706543,
+      "step": 3244
+    },
+    {
+      "epoch": 0.8273839877613463,
+      "grad_norm": 0.43144291639328003,
+      "learning_rate": 1.5862213506942037e-06,
+      "loss": 0.43666523694992065,
+      "step": 3245
+    },
+    {
+      "epoch": 0.8276389597144315,
+      "grad_norm": 0.4348338842391968,
+      "learning_rate": 1.5816663159589719e-06,
+      "loss": 0.4477895498275757,
+      "step": 3246
+    },
+    {
+      "epoch": 0.8278939316675166,
+      "grad_norm": 0.4303411543369293,
+      "learning_rate": 1.5771172691066793e-06,
+      "loss": 0.4419710338115692,
+      "step": 3247
+    },
+    {
+      "epoch": 0.8281489036206018,
+      "grad_norm": 0.46197861433029175,
+      "learning_rate": 1.572574213373027e-06,
+      "loss": 0.42960870265960693,
+      "step": 3248
+    },
+    {
+      "epoch": 0.8284038755736869,
+      "grad_norm": 0.4203360974788666,
+      "learning_rate": 1.5680371519894444e-06,
+      "loss": 0.443248450756073,
+      "step": 3249
+    },
+    {
+      "epoch": 0.8286588475267721,
+      "grad_norm": 0.47939595580101013,
+      "learning_rate": 1.563506088183102e-06,
+      "loss": 0.4249131381511688,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8289138194798572,
+      "grad_norm": 0.4603368639945984,
+      "learning_rate": 1.5589810251769034e-06,
+      "loss": 0.43305230140686035,
+      "step": 3251
+    },
+    {
+      "epoch": 0.8291687914329424,
+      "grad_norm": 0.43101152777671814,
+      "learning_rate": 1.5544619661894866e-06,
+      "loss": 0.42263245582580566,
+      "step": 3252
+    },
+    {
+      "epoch": 0.8294237633860275,
+      "grad_norm": 0.46168243885040283,
+      "learning_rate": 1.5499489144352131e-06,
+      "loss": 0.41834792494773865,
+      "step": 3253
+    },
+    {
+      "epoch": 0.8296787353391127,
+      "grad_norm": 0.47389018535614014,
+      "learning_rate": 1.5454418731241793e-06,
+      "loss": 0.45298704504966736,
+      "step": 3254
+    },
+    {
+      "epoch": 0.8299337072921978,
+      "grad_norm": 0.4682997167110443,
+      "learning_rate": 1.5409408454621987e-06,
+      "loss": 0.44076794385910034,
+      "step": 3255
+    },
+    {
+      "epoch": 0.8301886792452831,
+      "grad_norm": 0.45663902163505554,
+      "learning_rate": 1.5364458346508093e-06,
+      "loss": 0.44266772270202637,
+      "step": 3256
+    },
+    {
+      "epoch": 0.8304436511983682,
+      "grad_norm": 0.4399891793727875,
+      "learning_rate": 1.5319568438872746e-06,
+      "loss": 0.4307417869567871,
+      "step": 3257
+    },
+    {
+      "epoch": 0.8306986231514534,
+      "grad_norm": 0.4241107702255249,
+      "learning_rate": 1.527473876364568e-06,
+      "loss": 0.43898722529411316,
+      "step": 3258
+    },
+    {
+      "epoch": 0.8309535951045385,
+      "grad_norm": 0.5689083933830261,
+      "learning_rate": 1.522996935271388e-06,
+      "loss": 0.44163841009140015,
+      "step": 3259
+    },
+    {
+      "epoch": 0.8312085670576237,
+      "grad_norm": 0.4378184974193573,
+      "learning_rate": 1.5185260237921361e-06,
+      "loss": 0.43709036707878113,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8314635390107088,
+      "grad_norm": 0.43571507930755615,
+      "learning_rate": 1.5140611451069354e-06,
+      "loss": 0.44041547179222107,
+      "step": 3261
+    },
+    {
+      "epoch": 0.831718510963794,
+      "grad_norm": 0.44554468989372253,
+      "learning_rate": 1.5096023023916096e-06,
+      "loss": 0.435102641582489,
+      "step": 3262
+    },
+    {
+      "epoch": 0.8319734829168791,
+      "grad_norm": 0.46228763461112976,
+      "learning_rate": 1.5051494988176917e-06,
+      "loss": 0.43231040239334106,
+      "step": 3263
+    },
+    {
+      "epoch": 0.8322284548699643,
+      "grad_norm": 0.45202240347862244,
+      "learning_rate": 1.5007027375524209e-06,
+      "loss": 0.4369967579841614,
+      "step": 3264
+    },
+    {
+      "epoch": 0.8324834268230494,
+      "grad_norm": 0.45653805136680603,
+      "learning_rate": 1.4962620217587386e-06,
+      "loss": 0.4365310072898865,
+      "step": 3265
+    },
+    {
+      "epoch": 0.8327383987761346,
+      "grad_norm": 0.44089001417160034,
+      "learning_rate": 1.4918273545952833e-06,
+      "loss": 0.4460240304470062,
+      "step": 3266
+    },
+    {
+      "epoch": 0.8329933707292198,
+      "grad_norm": 0.43628019094467163,
+      "learning_rate": 1.4873987392163946e-06,
+      "loss": 0.4349481463432312,
+      "step": 3267
+    },
+    {
+      "epoch": 0.833248342682305,
+      "grad_norm": 0.4397881031036377,
+      "learning_rate": 1.4829761787721042e-06,
+      "loss": 0.438412070274353,
+      "step": 3268
+    },
+    {
+      "epoch": 0.8335033146353901,
+      "grad_norm": 0.4743090569972992,
+      "learning_rate": 1.4785596764081366e-06,
+      "loss": 0.4575331211090088,
+      "step": 3269
+    },
+    {
+      "epoch": 0.8337582865884753,
+      "grad_norm": 0.4318879246711731,
+      "learning_rate": 1.4741492352659137e-06,
+      "loss": 0.43636226654052734,
+      "step": 3270
+    },
+    {
+      "epoch": 0.8340132585415604,
+      "grad_norm": 0.4453240931034088,
+      "learning_rate": 1.4697448584825369e-06,
+      "loss": 0.4309747815132141,
+      "step": 3271
+    },
+    {
+      "epoch": 0.8342682304946456,
+      "grad_norm": 0.42908501625061035,
+      "learning_rate": 1.4653465491908003e-06,
+      "loss": 0.445771187543869,
+      "step": 3272
+    },
+    {
+      "epoch": 0.8345232024477307,
+      "grad_norm": 0.4467611610889435,
+      "learning_rate": 1.4609543105191837e-06,
+      "loss": 0.45422136783599854,
+      "step": 3273
+    },
+    {
+      "epoch": 0.8347781744008159,
+      "grad_norm": 0.4240168631076813,
+      "learning_rate": 1.4565681455918434e-06,
+      "loss": 0.442562460899353,
+      "step": 3274
+    },
+    {
+      "epoch": 0.835033146353901,
+      "grad_norm": 0.4435538649559021,
+      "learning_rate": 1.452188057528615e-06,
+      "loss": 0.42605406045913696,
+      "step": 3275
+    },
+    {
+      "epoch": 0.8352881183069862,
+      "grad_norm": 0.46490219235420227,
+      "learning_rate": 1.4478140494450211e-06,
+      "loss": 0.4366442561149597,
+      "step": 3276
+    },
+    {
+      "epoch": 0.8355430902600713,
+      "grad_norm": 0.43472519516944885,
+      "learning_rate": 1.4434461244522458e-06,
+      "loss": 0.44394561648368835,
+      "step": 3277
+    },
+    {
+      "epoch": 0.8357980622131566,
+      "grad_norm": 0.424283504486084,
+      "learning_rate": 1.4390842856571607e-06,
+      "loss": 0.4361807703971863,
+      "step": 3278
+    },
+    {
+      "epoch": 0.8360530341662417,
+      "grad_norm": 0.4511633813381195,
+      "learning_rate": 1.4347285361622966e-06,
+      "loss": 0.4422985911369324,
+      "step": 3279
+    },
+    {
+      "epoch": 0.8363080061193269,
+      "grad_norm": 0.45547357201576233,
+      "learning_rate": 1.4303788790658613e-06,
+      "loss": 0.4446526765823364,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8365629780724121,
+      "grad_norm": 0.44244518876075745,
+      "learning_rate": 1.4260353174617237e-06,
+      "loss": 0.4395790994167328,
+      "step": 3281
+    },
+    {
+      "epoch": 0.8368179500254972,
+      "grad_norm": 0.44289788603782654,
+      "learning_rate": 1.4216978544394177e-06,
+      "loss": 0.4399866759777069,
+      "step": 3282
+    },
+    {
+      "epoch": 0.8370729219785824,
+      "grad_norm": 0.4806705117225647,
+      "learning_rate": 1.4173664930841414e-06,
+      "loss": 0.43484628200531006,
+      "step": 3283
+    },
+    {
+      "epoch": 0.8373278939316675,
+      "grad_norm": 0.444036602973938,
+      "learning_rate": 1.4130412364767565e-06,
+      "loss": 0.43715572357177734,
+      "step": 3284
+    },
+    {
+      "epoch": 0.8375828658847527,
+      "grad_norm": 0.44062647223472595,
+      "learning_rate": 1.4087220876937714e-06,
+      "loss": 0.4454675316810608,
+      "step": 3285
+    },
+    {
+      "epoch": 0.8378378378378378,
+      "grad_norm": 0.44224995374679565,
+      "learning_rate": 1.4044090498073625e-06,
+      "loss": 0.44212135672569275,
+      "step": 3286
+    },
+    {
+      "epoch": 0.838092809790923,
+      "grad_norm": 0.41917848587036133,
+      "learning_rate": 1.400102125885351e-06,
+      "loss": 0.427121102809906,
+      "step": 3287
+    },
+    {
+      "epoch": 0.8383477817440081,
+      "grad_norm": 0.4165647625923157,
+      "learning_rate": 1.395801318991209e-06,
+      "loss": 0.4368363320827484,
+      "step": 3288
+    },
+    {
+      "epoch": 0.8386027536970934,
+      "grad_norm": 0.4660046100616455,
+      "learning_rate": 1.3915066321840675e-06,
+      "loss": 0.4286162853240967,
+      "step": 3289
+    },
+    {
+      "epoch": 0.8388577256501785,
+      "grad_norm": 0.4394930899143219,
+      "learning_rate": 1.387218068518691e-06,
+      "loss": 0.44216713309288025,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8391126976032637,
+      "grad_norm": 0.4506072998046875,
+      "learning_rate": 1.382935631045501e-06,
+      "loss": 0.422912061214447,
+      "step": 3291
+    },
+    {
+      "epoch": 0.8393676695563488,
+      "grad_norm": 0.4367363154888153,
+      "learning_rate": 1.3786593228105494e-06,
+      "loss": 0.4353519082069397,
+      "step": 3292
+    },
+    {
+      "epoch": 0.839622641509434,
+      "grad_norm": 0.441277414560318,
+      "learning_rate": 1.3743891468555415e-06,
+      "loss": 0.45142579078674316,
+      "step": 3293
+    },
+    {
+      "epoch": 0.8398776134625191,
+      "grad_norm": 0.4367586672306061,
+      "learning_rate": 1.3701251062178101e-06,
+      "loss": 0.4367350935935974,
+      "step": 3294
+    },
+    {
+      "epoch": 0.8401325854156043,
+      "grad_norm": 0.4555458426475525,
+      "learning_rate": 1.3658672039303266e-06,
+      "loss": 0.45928090810775757,
+      "step": 3295
+    },
+    {
+      "epoch": 0.8403875573686894,
+      "grad_norm": 0.43432533740997314,
+      "learning_rate": 1.3616154430216989e-06,
+      "loss": 0.44207262992858887,
+      "step": 3296
+    },
+    {
+      "epoch": 0.8406425293217746,
+      "grad_norm": 0.4285244941711426,
+      "learning_rate": 1.3573698265161683e-06,
+      "loss": 0.4286430776119232,
+      "step": 3297
+    },
+    {
+      "epoch": 0.8408975012748597,
+      "grad_norm": 0.43814343214035034,
+      "learning_rate": 1.3531303574335973e-06,
+      "loss": 0.4361557960510254,
+      "step": 3298
+    },
+    {
+      "epoch": 0.841152473227945,
+      "grad_norm": 0.43010541796684265,
+      "learning_rate": 1.3488970387894874e-06,
+      "loss": 0.4290144145488739,
+      "step": 3299
+    },
+    {
+      "epoch": 0.8414074451810301,
+      "grad_norm": 0.5369753837585449,
+      "learning_rate": 1.3446698735949537e-06,
+      "loss": 0.43052634596824646,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8416624171341153,
+      "grad_norm": 0.4521333873271942,
+      "learning_rate": 1.3404488648567405e-06,
+      "loss": 0.4416801631450653,
+      "step": 3301
+    },
+    {
+      "epoch": 0.8419173890872004,
+      "grad_norm": 0.420360803604126,
+      "learning_rate": 1.3362340155772147e-06,
+      "loss": 0.4436587691307068,
+      "step": 3302
+    },
+    {
+      "epoch": 0.8421723610402856,
+      "grad_norm": 0.4203634262084961,
+      "learning_rate": 1.332025328754356e-06,
+      "loss": 0.4210076332092285,
+      "step": 3303
+    },
+    {
+      "epoch": 0.8424273329933707,
+      "grad_norm": 0.448469340801239,
+      "learning_rate": 1.3278228073817667e-06,
+      "loss": 0.4314681887626648,
+      "step": 3304
+    },
+    {
+      "epoch": 0.8426823049464559,
+      "grad_norm": 0.4256279766559601,
+      "learning_rate": 1.3236264544486631e-06,
+      "loss": 0.4425191879272461,
+      "step": 3305
+    },
+    {
+      "epoch": 0.842937276899541,
+      "grad_norm": 0.4700623154640198,
+      "learning_rate": 1.3194362729398702e-06,
+      "loss": 0.44431278109550476,
+      "step": 3306
+    },
+    {
+      "epoch": 0.8431922488526262,
+      "grad_norm": 0.4359423518180847,
+      "learning_rate": 1.3152522658358247e-06,
+      "loss": 0.4293837249279022,
+      "step": 3307
+    },
+    {
+      "epoch": 0.8434472208057113,
+      "grad_norm": 0.4656182825565338,
+      "learning_rate": 1.311074436112575e-06,
+      "loss": 0.4318646788597107,
+      "step": 3308
+    },
+    {
+      "epoch": 0.8437021927587965,
+      "grad_norm": 0.46945562958717346,
+      "learning_rate": 1.3069027867417682e-06,
+      "loss": 0.4293580651283264,
+      "step": 3309
+    },
+    {
+      "epoch": 0.8439571647118816,
+      "grad_norm": 0.4682983458042145,
+      "learning_rate": 1.3027373206906646e-06,
+      "loss": 0.4242051839828491,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8442121366649669,
+      "grad_norm": 0.4534981846809387,
+      "learning_rate": 1.2985780409221183e-06,
+      "loss": 0.4334355592727661,
+      "step": 3311
+    },
+    {
+      "epoch": 0.844467108618052,
+      "grad_norm": 0.4599556028842926,
+      "learning_rate": 1.2944249503945894e-06,
+      "loss": 0.4311177134513855,
+      "step": 3312
+    },
+    {
+      "epoch": 0.8447220805711372,
+      "grad_norm": 0.41913193464279175,
+      "learning_rate": 1.2902780520621328e-06,
+      "loss": 0.4342377185821533,
+      "step": 3313
+    },
+    {
+      "epoch": 0.8449770525242223,
+      "grad_norm": 0.4389280378818512,
+      "learning_rate": 1.2861373488743934e-06,
+      "loss": 0.4300832152366638,
+      "step": 3314
+    },
+    {
+      "epoch": 0.8452320244773075,
+      "grad_norm": 0.45004579424858093,
+      "learning_rate": 1.282002843776623e-06,
+      "loss": 0.4427967071533203,
+      "step": 3315
+    },
+    {
+      "epoch": 0.8454869964303927,
+      "grad_norm": 0.42154547572135925,
+      "learning_rate": 1.2778745397096503e-06,
+      "loss": 0.42205655574798584,
+      "step": 3316
+    },
+    {
+      "epoch": 0.8457419683834778,
+      "grad_norm": 0.43388786911964417,
+      "learning_rate": 1.2737524396099033e-06,
+      "loss": 0.43303999304771423,
+      "step": 3317
+    },
+    {
+      "epoch": 0.845996940336563,
+      "grad_norm": 0.445159375667572,
+      "learning_rate": 1.269636546409394e-06,
+      "loss": 0.43945175409317017,
+      "step": 3318
+    },
+    {
+      "epoch": 0.8462519122896481,
+      "grad_norm": 0.43009161949157715,
+      "learning_rate": 1.2655268630357197e-06,
+      "loss": 0.43136143684387207,
+      "step": 3319
+    },
+    {
+      "epoch": 0.8465068842427333,
+      "grad_norm": 0.43622779846191406,
+      "learning_rate": 1.2614233924120555e-06,
+      "loss": 0.4358550012111664,
+      "step": 3320
+    },
+    {
+      "epoch": 0.8467618561958185,
+      "grad_norm": 0.44786137342453003,
+      "learning_rate": 1.2573261374571677e-06,
+      "loss": 0.4262104034423828,
+      "step": 3321
+    },
+    {
+      "epoch": 0.8470168281489037,
+      "grad_norm": 0.4478050172328949,
+      "learning_rate": 1.2532351010853916e-06,
+      "loss": 0.43247491121292114,
+      "step": 3322
+    },
+    {
+      "epoch": 0.8472718001019888,
+      "grad_norm": 0.46326589584350586,
+      "learning_rate": 1.2491502862066484e-06,
+      "loss": 0.42917340993881226,
+      "step": 3323
+    },
+    {
+      "epoch": 0.847526772055074,
+      "grad_norm": 0.4584805965423584,
+      "learning_rate": 1.2450716957264232e-06,
+      "loss": 0.43868932127952576,
+      "step": 3324
+    },
+    {
+      "epoch": 0.8477817440081591,
+      "grad_norm": 0.4304777979850769,
+      "learning_rate": 1.2409993325457848e-06,
+      "loss": 0.4443057179450989,
+      "step": 3325
+    },
+    {
+      "epoch": 0.8480367159612443,
+      "grad_norm": 0.45579901337623596,
+      "learning_rate": 1.2369331995613664e-06,
+      "loss": 0.449260413646698,
+      "step": 3326
+    },
+    {
+      "epoch": 0.8482916879143294,
+      "grad_norm": 0.43227526545524597,
+      "learning_rate": 1.232873299665367e-06,
+      "loss": 0.436143159866333,
+      "step": 3327
+    },
+    {
+      "epoch": 0.8485466598674146,
+      "grad_norm": 0.45995062589645386,
+      "learning_rate": 1.2288196357455584e-06,
+      "loss": 0.455481618642807,
+      "step": 3328
+    },
+    {
+      "epoch": 0.8488016318204997,
+      "grad_norm": 0.5433409810066223,
+      "learning_rate": 1.2247722106852777e-06,
+      "loss": 0.4205161929130554,
+      "step": 3329
+    },
+    {
+      "epoch": 0.8490566037735849,
+      "grad_norm": 0.44447267055511475,
+      "learning_rate": 1.2207310273634155e-06,
+      "loss": 0.44169408082962036,
+      "step": 3330
+    },
+    {
+      "epoch": 0.84931157572667,
+      "grad_norm": 0.4399317502975464,
+      "learning_rate": 1.2166960886544322e-06,
+      "loss": 0.4465487003326416,
+      "step": 3331
+    },
+    {
+      "epoch": 0.8495665476797553,
+      "grad_norm": 0.4159085750579834,
+      "learning_rate": 1.212667397428342e-06,
+      "loss": 0.4228609502315521,
+      "step": 3332
+    },
+    {
+      "epoch": 0.8498215196328404,
+      "grad_norm": 0.43474212288856506,
+      "learning_rate": 1.2086449565507107e-06,
+      "loss": 0.4253726899623871,
+      "step": 3333
+    },
+    {
+      "epoch": 0.8500764915859256,
+      "grad_norm": 0.44543296098709106,
+      "learning_rate": 1.2046287688826708e-06,
+      "loss": 0.4422019422054291,
+      "step": 3334
+    },
+    {
+      "epoch": 0.8503314635390107,
+      "grad_norm": 0.4402267336845398,
+      "learning_rate": 1.2006188372808925e-06,
+      "loss": 0.44364506006240845,
+      "step": 3335
+    },
+    {
+      "epoch": 0.8505864354920959,
+      "grad_norm": 0.44199803471565247,
+      "learning_rate": 1.1966151645976087e-06,
+      "loss": 0.42757415771484375,
+      "step": 3336
+    },
+    {
+      "epoch": 0.850841407445181,
+      "grad_norm": 0.43642616271972656,
+      "learning_rate": 1.1926177536805905e-06,
+      "loss": 0.43112000823020935,
+      "step": 3337
+    },
+    {
+      "epoch": 0.8510963793982662,
+      "grad_norm": 0.45532628893852234,
+      "learning_rate": 1.1886266073731633e-06,
+      "loss": 0.4462944269180298,
+      "step": 3338
+    },
+    {
+      "epoch": 0.8513513513513513,
+      "grad_norm": 0.4233347475528717,
+      "learning_rate": 1.184641728514191e-06,
+      "loss": 0.43310827016830444,
+      "step": 3339
+    },
+    {
+      "epoch": 0.8516063233044365,
+      "grad_norm": 0.4422180652618408,
+      "learning_rate": 1.1806631199380781e-06,
+      "loss": 0.4407688081264496,
+      "step": 3340
+    },
+    {
+      "epoch": 0.8518612952575216,
+      "grad_norm": 0.4382367730140686,
+      "learning_rate": 1.1766907844747744e-06,
+      "loss": 0.4392991065979004,
+      "step": 3341
+    },
+    {
+      "epoch": 0.8521162672106068,
+      "grad_norm": 0.44065603613853455,
+      "learning_rate": 1.1727247249497686e-06,
+      "loss": 0.4309106767177582,
+      "step": 3342
+    },
+    {
+      "epoch": 0.852371239163692,
+      "grad_norm": 0.43926700949668884,
+      "learning_rate": 1.1687649441840764e-06,
+      "loss": 0.436871737241745,
+      "step": 3343
+    },
+    {
+      "epoch": 0.8526262111167772,
+      "grad_norm": 0.4672529399394989,
+      "learning_rate": 1.1648114449942583e-06,
+      "loss": 0.4357472360134125,
+      "step": 3344
+    },
+    {
+      "epoch": 0.8528811830698623,
+      "grad_norm": 0.4307812452316284,
+      "learning_rate": 1.160864230192399e-06,
+      "loss": 0.4353850781917572,
+      "step": 3345
+    },
+    {
+      "epoch": 0.8531361550229475,
+      "grad_norm": 0.44444510340690613,
+      "learning_rate": 1.1569233025861147e-06,
+      "loss": 0.4323183000087738,
+      "step": 3346
+    },
+    {
+      "epoch": 0.8533911269760326,
+      "grad_norm": 0.42350655794143677,
+      "learning_rate": 1.152988664978556e-06,
+      "loss": 0.4409070909023285,
+      "step": 3347
+    },
+    {
+      "epoch": 0.8536460989291178,
+      "grad_norm": 0.44289278984069824,
+      "learning_rate": 1.149060320168387e-06,
+      "loss": 0.44024693965911865,
+      "step": 3348
+    },
+    {
+      "epoch": 0.8539010708822029,
+      "grad_norm": 0.45462486147880554,
+      "learning_rate": 1.145138270949807e-06,
+      "loss": 0.4491046369075775,
+      "step": 3349
+    },
+    {
+      "epoch": 0.8541560428352881,
+      "grad_norm": 0.42163506150245667,
+      "learning_rate": 1.1412225201125359e-06,
+      "loss": 0.4273730218410492,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8544110147883732,
+      "grad_norm": 0.4252319037914276,
+      "learning_rate": 1.1373130704418089e-06,
+      "loss": 0.4343377351760864,
+      "step": 3351
+    },
+    {
+      "epoch": 0.8546659867414584,
+      "grad_norm": 0.43090590834617615,
+      "learning_rate": 1.1334099247183782e-06,
+      "loss": 0.44295868277549744,
+      "step": 3352
+    },
+    {
+      "epoch": 0.8549209586945437,
+      "grad_norm": 0.43933114409446716,
+      "learning_rate": 1.12951308571852e-06,
+      "loss": 0.42852818965911865,
+      "step": 3353
+    },
+    {
+      "epoch": 0.8551759306476288,
+      "grad_norm": 0.8181011080741882,
+      "learning_rate": 1.1256225562140155e-06,
+      "loss": 0.4561828374862671,
+      "step": 3354
+    },
+    {
+      "epoch": 0.855430902600714,
+      "grad_norm": 0.4378681778907776,
+      "learning_rate": 1.121738338972167e-06,
+      "loss": 0.4349752366542816,
+      "step": 3355
+    },
+    {
+      "epoch": 0.8556858745537991,
+      "grad_norm": 0.42967164516448975,
+      "learning_rate": 1.1178604367557776e-06,
+      "loss": 0.437456339597702,
+      "step": 3356
+    },
+    {
+      "epoch": 0.8559408465068843,
+      "grad_norm": 0.4497753381729126,
+      "learning_rate": 1.113988852323168e-06,
+      "loss": 0.42805370688438416,
+      "step": 3357
+    },
+    {
+      "epoch": 0.8561958184599694,
+      "grad_norm": 0.5577053427696228,
+      "learning_rate": 1.110123588428159e-06,
+      "loss": 0.4310586452484131,
+      "step": 3358
+    },
+    {
+      "epoch": 0.8564507904130546,
+      "grad_norm": 0.44477880001068115,
+      "learning_rate": 1.1062646478200735e-06,
+      "loss": 0.43208011984825134,
+      "step": 3359
+    },
+    {
+      "epoch": 0.8567057623661397,
+      "grad_norm": 0.43552449345588684,
+      "learning_rate": 1.1024120332437428e-06,
+      "loss": 0.4297208786010742,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8569607343192249,
+      "grad_norm": 0.453593909740448,
+      "learning_rate": 1.0985657474394984e-06,
+      "loss": 0.4424916207790375,
+      "step": 3361
+    },
+    {
+      "epoch": 0.85721570627231,
+      "grad_norm": 0.4612596035003662,
+      "learning_rate": 1.0947257931431644e-06,
+      "loss": 0.44119787216186523,
+      "step": 3362
+    },
+    {
+      "epoch": 0.8574706782253952,
+      "grad_norm": 0.44879037141799927,
+      "learning_rate": 1.09089217308607e-06,
+      "loss": 0.4400937557220459,
+      "step": 3363
+    },
+    {
+      "epoch": 0.8577256501784803,
+      "grad_norm": 0.442455530166626,
+      "learning_rate": 1.0870648899950254e-06,
+      "loss": 0.4569939076900482,
+      "step": 3364
+    },
+    {
+      "epoch": 0.8579806221315656,
+      "grad_norm": 0.44891437888145447,
+      "learning_rate": 1.083243946592346e-06,
+      "loss": 0.4352353513240814,
+      "step": 3365
+    },
+    {
+      "epoch": 0.8582355940846507,
+      "grad_norm": 0.4200226962566376,
+      "learning_rate": 1.079429345595836e-06,
+      "loss": 0.44492048025131226,
+      "step": 3366
+    },
+    {
+      "epoch": 0.8584905660377359,
+      "grad_norm": 0.41359463334083557,
+      "learning_rate": 1.0756210897187812e-06,
+      "loss": 0.4444793164730072,
+      "step": 3367
+    },
+    {
+      "epoch": 0.858745537990821,
+      "grad_norm": 0.45372694730758667,
+      "learning_rate": 1.0718191816699631e-06,
+      "loss": 0.4389488101005554,
+      "step": 3368
+    },
+    {
+      "epoch": 0.8590005099439062,
+      "grad_norm": 0.4190084636211395,
+      "learning_rate": 1.06802362415364e-06,
+      "loss": 0.45514458417892456,
+      "step": 3369
+    },
+    {
+      "epoch": 0.8592554818969913,
+      "grad_norm": 0.41038715839385986,
+      "learning_rate": 1.064234419869561e-06,
+      "loss": 0.43614017963409424,
+      "step": 3370
+    },
+    {
+      "epoch": 0.8595104538500765,
+      "grad_norm": 0.42078742384910583,
+      "learning_rate": 1.0604515715129481e-06,
+      "loss": 0.43874263763427734,
+      "step": 3371
+    },
+    {
+      "epoch": 0.8597654258031616,
+      "grad_norm": 0.42877811193466187,
+      "learning_rate": 1.0566750817745076e-06,
+      "loss": 0.43939733505249023,
+      "step": 3372
+    },
+    {
+      "epoch": 0.8600203977562468,
+      "grad_norm": 0.5327430367469788,
+      "learning_rate": 1.0529049533404223e-06,
+      "loss": 0.4264140725135803,
+      "step": 3373
+    },
+    {
+      "epoch": 0.8602753697093319,
+      "grad_norm": 0.45152685046195984,
+      "learning_rate": 1.0491411888923509e-06,
+      "loss": 0.430926650762558,
+      "step": 3374
+    },
+    {
+      "epoch": 0.8605303416624172,
+      "grad_norm": 0.43535324931144714,
+      "learning_rate": 1.0453837911074204e-06,
+      "loss": 0.43031561374664307,
+      "step": 3375
+    },
+    {
+      "epoch": 0.8607853136155023,
+      "grad_norm": 0.43099328875541687,
+      "learning_rate": 1.041632762658238e-06,
+      "loss": 0.44012102484703064,
+      "step": 3376
+    },
+    {
+      "epoch": 0.8610402855685875,
+      "grad_norm": 0.4511074125766754,
+      "learning_rate": 1.0378881062128732e-06,
+      "loss": 0.4425766170024872,
+      "step": 3377
+    },
+    {
+      "epoch": 0.8612952575216726,
+      "grad_norm": 0.4516361951828003,
+      "learning_rate": 1.0341498244348636e-06,
+      "loss": 0.43071991205215454,
+      "step": 3378
+    },
+    {
+      "epoch": 0.8615502294747578,
+      "grad_norm": 0.424569696187973,
+      "learning_rate": 1.0304179199832187e-06,
+      "loss": 0.44217395782470703,
+      "step": 3379
+    },
+    {
+      "epoch": 0.8618052014278429,
+      "grad_norm": 0.735274612903595,
+      "learning_rate": 1.0266923955124032e-06,
+      "loss": 0.4275852143764496,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8620601733809281,
+      "grad_norm": 0.4615504741668701,
+      "learning_rate": 1.0229732536723502e-06,
+      "loss": 0.4398762583732605,
+      "step": 3381
+    },
+    {
+      "epoch": 0.8623151453340132,
+      "grad_norm": 0.43356427550315857,
+      "learning_rate": 1.019260497108453e-06,
+      "loss": 0.4225383400917053,
+      "step": 3382
+    },
+    {
+      "epoch": 0.8625701172870984,
+      "grad_norm": 0.4367641806602478,
+      "learning_rate": 1.01555412846156e-06,
+      "loss": 0.4380742013454437,
+      "step": 3383
+    },
+    {
+      "epoch": 0.8628250892401835,
+      "grad_norm": 0.44011420011520386,
+      "learning_rate": 1.0118541503679724e-06,
+      "loss": 0.4346046447753906,
+      "step": 3384
+    },
+    {
+      "epoch": 0.8630800611932687,
+      "grad_norm": 0.4376404881477356,
+      "learning_rate": 1.0081605654594573e-06,
+      "loss": 0.42278382182121277,
+      "step": 3385
+    },
+    {
+      "epoch": 0.8633350331463538,
+      "grad_norm": 0.46831297874450684,
+      "learning_rate": 1.004473376363222e-06,
+      "loss": 0.43210935592651367,
+      "step": 3386
+    },
+    {
+      "epoch": 0.8635900050994391,
+      "grad_norm": 0.42311087250709534,
+      "learning_rate": 1.000792585701934e-06,
+      "loss": 0.4354386329650879,
+      "step": 3387
+    },
+    {
+      "epoch": 0.8638449770525243,
+      "grad_norm": 0.4499448835849762,
+      "learning_rate": 9.971181960937027e-07,
+      "loss": 0.4326968193054199,
+      "step": 3388
+    },
+    {
+      "epoch": 0.8640999490056094,
+      "grad_norm": 0.45473289489746094,
+      "learning_rate": 9.934502101520904e-07,
+      "loss": 0.4448380172252655,
+      "step": 3389
+    },
+    {
+      "epoch": 0.8643549209586946,
+      "grad_norm": 0.4439501464366913,
+      "learning_rate": 9.897886304860994e-07,
+      "loss": 0.41785746812820435,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8646098929117797,
+      "grad_norm": 0.4552282989025116,
+      "learning_rate": 9.86133459700177e-07,
+      "loss": 0.45230287313461304,
+      "step": 3391
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 0.43610405921936035,
+      "learning_rate": 9.824847003942152e-07,
+      "loss": 0.4380587041378021,
+      "step": 3392
+    },
+    {
+      "epoch": 0.86511983681795,
+      "grad_norm": 0.43049919605255127,
+      "learning_rate": 9.788423551635384e-07,
+      "loss": 0.4224151074886322,
+      "step": 3393
+    },
+    {
+      "epoch": 0.8653748087710352,
+      "grad_norm": 0.4411639869213104,
+      "learning_rate": 9.752064265989158e-07,
+      "loss": 0.43679267168045044,
+      "step": 3394
+    },
+    {
+      "epoch": 0.8656297807241203,
+      "grad_norm": 0.45369890332221985,
+      "learning_rate": 9.715769172865542e-07,
+      "loss": 0.44111520051956177,
+      "step": 3395
+    },
+    {
+      "epoch": 0.8658847526772055,
+      "grad_norm": 0.44774970412254333,
+      "learning_rate": 9.679538298080804e-07,
+      "loss": 0.41965052485466003,
+      "step": 3396
+    },
+    {
+      "epoch": 0.8661397246302907,
+      "grad_norm": 0.42721036076545715,
+      "learning_rate": 9.6433716674057e-07,
+      "loss": 0.4369600713253021,
+      "step": 3397
+    },
+    {
+      "epoch": 0.8663946965833759,
+      "grad_norm": 0.4301885664463043,
+      "learning_rate": 9.607269306565202e-07,
+      "loss": 0.450730562210083,
+      "step": 3398
+    },
+    {
+      "epoch": 0.866649668536461,
+      "grad_norm": 0.44836947321891785,
+      "learning_rate": 9.571231241238578e-07,
+      "loss": 0.4442753791809082,
+      "step": 3399
+    },
+    {
+      "epoch": 0.8669046404895462,
+      "grad_norm": 0.4398816227912903,
+      "learning_rate": 9.535257497059391e-07,
+      "loss": 0.437112420797348,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8671596124426313,
+      "grad_norm": 0.44665607810020447,
+      "learning_rate": 9.499348099615413e-07,
+      "loss": 0.4469532370567322,
+      "step": 3401
+    },
+    {
+      "epoch": 0.8674145843957165,
+      "grad_norm": 0.45355597138404846,
+      "learning_rate": 9.463503074448676e-07,
+      "loss": 0.42864227294921875,
+      "step": 3402
+    },
+    {
+      "epoch": 0.8676695563488016,
+      "grad_norm": 0.43615153431892395,
+      "learning_rate": 9.42772244705542e-07,
+      "loss": 0.4274861216545105,
+      "step": 3403
+    },
+    {
+      "epoch": 0.8679245283018868,
+      "grad_norm": 0.4431467056274414,
+      "learning_rate": 9.392006242886054e-07,
+      "loss": 0.4423532485961914,
+      "step": 3404
+    },
+    {
+      "epoch": 0.8681795002549719,
+      "grad_norm": 0.44587862491607666,
+      "learning_rate": 9.356354487345187e-07,
+      "loss": 0.4221605658531189,
+      "step": 3405
+    },
+    {
+      "epoch": 0.8684344722080571,
+      "grad_norm": 0.4480460584163666,
+      "learning_rate": 9.320767205791614e-07,
+      "loss": 0.4308694303035736,
+      "step": 3406
+    },
+    {
+      "epoch": 0.8686894441611422,
+      "grad_norm": 0.46006181836128235,
+      "learning_rate": 9.285244423538198e-07,
+      "loss": 0.43253546953201294,
+      "step": 3407
+    },
+    {
+      "epoch": 0.8689444161142275,
+      "grad_norm": 0.43100929260253906,
+      "learning_rate": 9.249786165852004e-07,
+      "loss": 0.43706291913986206,
+      "step": 3408
+    },
+    {
+      "epoch": 0.8691993880673126,
+      "grad_norm": 0.438680499792099,
+      "learning_rate": 9.214392457954158e-07,
+      "loss": 0.4269286096096039,
+      "step": 3409
+    },
+    {
+      "epoch": 0.8694543600203978,
+      "grad_norm": 0.43813207745552063,
+      "learning_rate": 9.179063325019832e-07,
+      "loss": 0.4430767893791199,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8697093319734829,
+      "grad_norm": 0.434760719537735,
+      "learning_rate": 9.143798792178371e-07,
+      "loss": 0.42191946506500244,
+      "step": 3411
+    },
+    {
+      "epoch": 0.8699643039265681,
+      "grad_norm": 0.44063812494277954,
+      "learning_rate": 9.108598884513053e-07,
+      "loss": 0.44339871406555176,
+      "step": 3412
+    },
+    {
+      "epoch": 0.8702192758796532,
+      "grad_norm": 0.4579174816608429,
+      "learning_rate": 9.073463627061285e-07,
+      "loss": 0.44739198684692383,
+      "step": 3413
+    },
+    {
+      "epoch": 0.8704742478327384,
+      "grad_norm": 0.45135340094566345,
+      "learning_rate": 9.038393044814475e-07,
+      "loss": 0.4417533278465271,
+      "step": 3414
+    },
+    {
+      "epoch": 0.8707292197858235,
+      "grad_norm": 0.44124355912208557,
+      "learning_rate": 9.003387162717959e-07,
+      "loss": 0.43357163667678833,
+      "step": 3415
+    },
+    {
+      "epoch": 0.8709841917389087,
+      "grad_norm": 0.44779467582702637,
+      "learning_rate": 8.968446005671117e-07,
+      "loss": 0.43666166067123413,
+      "step": 3416
+    },
+    {
+      "epoch": 0.8712391636919938,
+      "grad_norm": 0.44657012820243835,
+      "learning_rate": 8.933569598527247e-07,
+      "loss": 0.43212875723838806,
+      "step": 3417
+    },
+    {
+      "epoch": 0.871494135645079,
+      "grad_norm": 0.42799729108810425,
+      "learning_rate": 8.898757966093629e-07,
+      "loss": 0.4432377219200134,
+      "step": 3418
+    },
+    {
+      "epoch": 0.8717491075981642,
+      "grad_norm": 0.4438287317752838,
+      "learning_rate": 8.864011133131489e-07,
+      "loss": 0.43603086471557617,
+      "step": 3419
+    },
+    {
+      "epoch": 0.8720040795512494,
+      "grad_norm": 0.44265010952949524,
+      "learning_rate": 8.829329124355868e-07,
+      "loss": 0.43077313899993896,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8722590515043345,
+      "grad_norm": 0.4266294836997986,
+      "learning_rate": 8.794711964435821e-07,
+      "loss": 0.44449281692504883,
+      "step": 3421
+    },
+    {
+      "epoch": 0.8725140234574197,
+      "grad_norm": 0.43002915382385254,
+      "learning_rate": 8.760159677994174e-07,
+      "loss": 0.4223582148551941,
+      "step": 3422
+    },
+    {
+      "epoch": 0.8727689954105049,
+      "grad_norm": 0.4211961030960083,
+      "learning_rate": 8.725672289607634e-07,
+      "loss": 0.42744097113609314,
+      "step": 3423
+    },
+    {
+      "epoch": 0.87302396736359,
+      "grad_norm": 0.45127928256988525,
+      "learning_rate": 8.691249823806813e-07,
+      "loss": 0.43572863936424255,
+      "step": 3424
+    },
+    {
+      "epoch": 0.8732789393166752,
+      "grad_norm": 0.4222966730594635,
+      "learning_rate": 8.656892305076047e-07,
+      "loss": 0.44335150718688965,
+      "step": 3425
+    },
+    {
+      "epoch": 0.8735339112697603,
+      "grad_norm": 0.45571115612983704,
+      "learning_rate": 8.622599757853533e-07,
+      "loss": 0.4440256953239441,
+      "step": 3426
+    },
+    {
+      "epoch": 0.8737888832228455,
+      "grad_norm": 0.45147934556007385,
+      "learning_rate": 8.588372206531293e-07,
+      "loss": 0.44460529088974,
+      "step": 3427
+    },
+    {
+      "epoch": 0.8740438551759306,
+      "grad_norm": 0.4408029317855835,
+      "learning_rate": 8.554209675454984e-07,
+      "loss": 0.4386541247367859,
+      "step": 3428
+    },
+    {
+      "epoch": 0.8742988271290159,
+      "grad_norm": 0.4156317114830017,
+      "learning_rate": 8.520112188924135e-07,
+      "loss": 0.4272143840789795,
+      "step": 3429
+    },
+    {
+      "epoch": 0.874553799082101,
+      "grad_norm": 0.4348088204860687,
+      "learning_rate": 8.486079771192001e-07,
+      "loss": 0.4361407458782196,
+      "step": 3430
+    },
+    {
+      "epoch": 0.8748087710351862,
+      "grad_norm": 0.4363251328468323,
+      "learning_rate": 8.452112446465477e-07,
+      "loss": 0.4313136041164398,
+      "step": 3431
+    },
+    {
+      "epoch": 0.8750637429882713,
+      "grad_norm": 0.4510456621646881,
+      "learning_rate": 8.418210238905256e-07,
+      "loss": 0.44090861082077026,
+      "step": 3432
+    },
+    {
+      "epoch": 0.8753187149413565,
+      "grad_norm": 0.43219730257987976,
+      "learning_rate": 8.384373172625626e-07,
+      "loss": 0.43194374442100525,
+      "step": 3433
+    },
+    {
+      "epoch": 0.8755736868944416,
+      "grad_norm": 0.4394565522670746,
+      "learning_rate": 8.350601271694614e-07,
+      "loss": 0.42495352029800415,
+      "step": 3434
+    },
+    {
+      "epoch": 0.8758286588475268,
+      "grad_norm": 0.6174067854881287,
+      "learning_rate": 8.316894560133826e-07,
+      "loss": 0.4395310580730438,
+      "step": 3435
+    },
+    {
+      "epoch": 0.8760836308006119,
+      "grad_norm": 0.4343336820602417,
+      "learning_rate": 8.28325306191855e-07,
+      "loss": 0.4309213161468506,
+      "step": 3436
+    },
+    {
+      "epoch": 0.8763386027536971,
+      "grad_norm": 0.45802173018455505,
+      "learning_rate": 8.249676800977657e-07,
+      "loss": 0.44333964586257935,
+      "step": 3437
+    },
+    {
+      "epoch": 0.8765935747067822,
+      "grad_norm": 0.4481572210788727,
+      "learning_rate": 8.216165801193676e-07,
+      "loss": 0.43676790595054626,
+      "step": 3438
+    },
+    {
+      "epoch": 0.8768485466598674,
+      "grad_norm": 0.4250171482563019,
+      "learning_rate": 8.182720086402618e-07,
+      "loss": 0.4320659041404724,
+      "step": 3439
+    },
+    {
+      "epoch": 0.8771035186129525,
+      "grad_norm": 0.4334368407726288,
+      "learning_rate": 8.149339680394164e-07,
+      "loss": 0.4514850378036499,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8773584905660378,
+      "grad_norm": 0.41408801078796387,
+      "learning_rate": 8.116024606911444e-07,
+      "loss": 0.43668755888938904,
+      "step": 3441
+    },
+    {
+      "epoch": 0.8776134625191229,
+      "grad_norm": 0.429240345954895,
+      "learning_rate": 8.082774889651168e-07,
+      "loss": 0.43987786769866943,
+      "step": 3442
+    },
+    {
+      "epoch": 0.8778684344722081,
+      "grad_norm": 0.4321937561035156,
+      "learning_rate": 8.049590552263575e-07,
+      "loss": 0.43867743015289307,
+      "step": 3443
+    },
+    {
+      "epoch": 0.8781234064252932,
+      "grad_norm": 0.43646883964538574,
+      "learning_rate": 8.01647161835235e-07,
+      "loss": 0.44158801436424255,
+      "step": 3444
+    },
+    {
+      "epoch": 0.8783783783783784,
+      "grad_norm": 0.4338165521621704,
+      "learning_rate": 7.983418111474717e-07,
+      "loss": 0.43622469902038574,
+      "step": 3445
+    },
+    {
+      "epoch": 0.8786333503314635,
+      "grad_norm": 0.42995837330818176,
+      "learning_rate": 7.950430055141312e-07,
+      "loss": 0.4489513337612152,
+      "step": 3446
+    },
+    {
+      "epoch": 0.8788883222845487,
+      "grad_norm": 0.4226197302341461,
+      "learning_rate": 7.91750747281621e-07,
+      "loss": 0.43353432416915894,
+      "step": 3447
+    },
+    {
+      "epoch": 0.8791432942376338,
+      "grad_norm": 0.48948803544044495,
+      "learning_rate": 7.884650387916992e-07,
+      "loss": 0.4460390508174896,
+      "step": 3448
+    },
+    {
+      "epoch": 0.879398266190719,
+      "grad_norm": 0.42587754130363464,
+      "learning_rate": 7.85185882381454e-07,
+      "loss": 0.43023681640625,
+      "step": 3449
+    },
+    {
+      "epoch": 0.8796532381438041,
+      "grad_norm": 0.44122031331062317,
+      "learning_rate": 7.819132803833229e-07,
+      "loss": 0.44698429107666016,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8799082100968894,
+      "grad_norm": 0.43389633297920227,
+      "learning_rate": 7.786472351250785e-07,
+      "loss": 0.4534319341182709,
+      "step": 3451
+    },
+    {
+      "epoch": 0.8801631820499745,
+      "grad_norm": 0.44629064202308655,
+      "learning_rate": 7.753877489298245e-07,
+      "loss": 0.4405984878540039,
+      "step": 3452
+    },
+    {
+      "epoch": 0.8804181540030597,
+      "grad_norm": 0.44391247630119324,
+      "learning_rate": 7.721348241160087e-07,
+      "loss": 0.4256533086299896,
+      "step": 3453
+    },
+    {
+      "epoch": 0.8806731259561448,
+      "grad_norm": 0.44749441742897034,
+      "learning_rate": 7.688884629974013e-07,
+      "loss": 0.42853403091430664,
+      "step": 3454
+    },
+    {
+      "epoch": 0.88092809790923,
+      "grad_norm": 0.4337758719921112,
+      "learning_rate": 7.65648667883111e-07,
+      "loss": 0.43673738837242126,
+      "step": 3455
+    },
+    {
+      "epoch": 0.8811830698623151,
+      "grad_norm": 0.44894421100616455,
+      "learning_rate": 7.624154410775753e-07,
+      "loss": 0.4365650415420532,
+      "step": 3456
+    },
+    {
+      "epoch": 0.8814380418154003,
+      "grad_norm": 0.4481375813484192,
+      "learning_rate": 7.591887848805545e-07,
+      "loss": 0.4377281069755554,
+      "step": 3457
+    },
+    {
+      "epoch": 0.8816930137684854,
+      "grad_norm": 0.4205091893672943,
+      "learning_rate": 7.559687015871431e-07,
+      "loss": 0.4369480311870575,
+      "step": 3458
+    },
+    {
+      "epoch": 0.8819479857215706,
+      "grad_norm": 0.4384622573852539,
+      "learning_rate": 7.527551934877597e-07,
+      "loss": 0.436867892742157,
+      "step": 3459
+    },
+    {
+      "epoch": 0.8822029576746558,
+      "grad_norm": 0.42390331625938416,
+      "learning_rate": 7.49548262868135e-07,
+      "loss": 0.4484369158744812,
+      "step": 3460
+    },
+    {
+      "epoch": 0.882457929627741,
+      "grad_norm": 0.43532341718673706,
+      "learning_rate": 7.46347912009332e-07,
+      "loss": 0.44156980514526367,
+      "step": 3461
+    },
+    {
+      "epoch": 0.8827129015808262,
+      "grad_norm": 0.44449347257614136,
+      "learning_rate": 7.431541431877343e-07,
+      "loss": 0.4474422335624695,
+      "step": 3462
+    },
+    {
+      "epoch": 0.8829678735339113,
+      "grad_norm": 0.4316299259662628,
+      "learning_rate": 7.399669586750357e-07,
+      "loss": 0.4310445189476013,
+      "step": 3463
+    },
+    {
+      "epoch": 0.8832228454869965,
+      "grad_norm": 0.4550790786743164,
+      "learning_rate": 7.367863607382542e-07,
+      "loss": 0.4407844543457031,
+      "step": 3464
+    },
+    {
+      "epoch": 0.8834778174400816,
+      "grad_norm": 0.435812771320343,
+      "learning_rate": 7.336123516397176e-07,
+      "loss": 0.438144326210022,
+      "step": 3465
+    },
+    {
+      "epoch": 0.8837327893931668,
+      "grad_norm": 0.42210423946380615,
+      "learning_rate": 7.304449336370711e-07,
+      "loss": 0.4265088737010956,
+      "step": 3466
+    },
+    {
+      "epoch": 0.8839877613462519,
+      "grad_norm": 0.4351652264595032,
+      "learning_rate": 7.272841089832694e-07,
+      "loss": 0.4282117486000061,
+      "step": 3467
+    },
+    {
+      "epoch": 0.8842427332993371,
+      "grad_norm": 0.4479033350944519,
+      "learning_rate": 7.24129879926575e-07,
+      "loss": 0.4302578866481781,
+      "step": 3468
+    },
+    {
+      "epoch": 0.8844977052524222,
+      "grad_norm": 0.43092960119247437,
+      "learning_rate": 7.209822487105655e-07,
+      "loss": 0.43718647956848145,
+      "step": 3469
+    },
+    {
+      "epoch": 0.8847526772055074,
+      "grad_norm": 0.44142788648605347,
+      "learning_rate": 7.17841217574119e-07,
+      "loss": 0.43912026286125183,
+      "step": 3470
+    },
+    {
+      "epoch": 0.8850076491585925,
+      "grad_norm": 0.4471288323402405,
+      "learning_rate": 7.147067887514236e-07,
+      "loss": 0.43443840742111206,
+      "step": 3471
+    },
+    {
+      "epoch": 0.8852626211116777,
+      "grad_norm": 0.43807896971702576,
+      "learning_rate": 7.115789644719728e-07,
+      "loss": 0.43923306465148926,
+      "step": 3472
+    },
+    {
+      "epoch": 0.8855175930647629,
+      "grad_norm": 0.4477294385433197,
+      "learning_rate": 7.084577469605525e-07,
+      "loss": 0.4377750754356384,
+      "step": 3473
+    },
+    {
+      "epoch": 0.8857725650178481,
+      "grad_norm": 0.4518868327140808,
+      "learning_rate": 7.053431384372599e-07,
+      "loss": 0.43124985694885254,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8860275369709332,
+      "grad_norm": 0.44317349791526794,
+      "learning_rate": 7.022351411174866e-07,
+      "loss": 0.44053560495376587,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8862825089240184,
+      "grad_norm": 0.4312305748462677,
+      "learning_rate": 6.991337572119217e-07,
+      "loss": 0.44413360953330994,
+      "step": 3476
+    },
+    {
+      "epoch": 0.8865374808771035,
+      "grad_norm": 0.4528648555278778,
+      "learning_rate": 6.960389889265517e-07,
+      "loss": 0.4408388137817383,
+      "step": 3477
+    },
+    {
+      "epoch": 0.8867924528301887,
+      "grad_norm": 0.4547745883464813,
+      "learning_rate": 6.929508384626571e-07,
+      "loss": 0.4394215941429138,
+      "step": 3478
+    },
+    {
+      "epoch": 0.8870474247832738,
+      "grad_norm": 0.4406372010707855,
+      "learning_rate": 6.898693080168084e-07,
+      "loss": 0.43522459268569946,
+      "step": 3479
+    },
+    {
+      "epoch": 0.887302396736359,
+      "grad_norm": 0.4650934934616089,
+      "learning_rate": 6.867943997808735e-07,
+      "loss": 0.44631558656692505,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8875573686894441,
+      "grad_norm": 0.4555158019065857,
+      "learning_rate": 6.837261159420016e-07,
+      "loss": 0.43532800674438477,
+      "step": 3481
+    },
+    {
+      "epoch": 0.8878123406425293,
+      "grad_norm": 0.4111659824848175,
+      "learning_rate": 6.806644586826383e-07,
+      "loss": 0.4364461302757263,
+      "step": 3482
+    },
+    {
+      "epoch": 0.8880673125956144,
+      "grad_norm": 0.4358905553817749,
+      "learning_rate": 6.776094301805114e-07,
+      "loss": 0.4406110942363739,
+      "step": 3483
+    },
+    {
+      "epoch": 0.8883222845486997,
+      "grad_norm": 0.4338662326335907,
+      "learning_rate": 6.745610326086327e-07,
+      "loss": 0.41641271114349365,
+      "step": 3484
+    },
+    {
+      "epoch": 0.8885772565017848,
+      "grad_norm": 0.46959635615348816,
+      "learning_rate": 6.715192681353022e-07,
+      "loss": 0.4398638904094696,
+      "step": 3485
+    },
+    {
+      "epoch": 0.88883222845487,
+      "grad_norm": 0.4260825514793396,
+      "learning_rate": 6.684841389240971e-07,
+      "loss": 0.4335266351699829,
+      "step": 3486
+    },
+    {
+      "epoch": 0.8890872004079551,
+      "grad_norm": 0.4593506455421448,
+      "learning_rate": 6.654556471338747e-07,
+      "loss": 0.4404299259185791,
+      "step": 3487
+    },
+    {
+      "epoch": 0.8893421723610403,
+      "grad_norm": 0.44930601119995117,
+      "learning_rate": 6.624337949187776e-07,
+      "loss": 0.43138647079467773,
+      "step": 3488
+    },
+    {
+      "epoch": 0.8895971443141254,
+      "grad_norm": 0.4422794282436371,
+      "learning_rate": 6.594185844282175e-07,
+      "loss": 0.4225475788116455,
+      "step": 3489
+    },
+    {
+      "epoch": 0.8898521162672106,
+      "grad_norm": 0.42670273780822754,
+      "learning_rate": 6.564100178068867e-07,
+      "loss": 0.41914981603622437,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8901070882202957,
+      "grad_norm": 0.4556596875190735,
+      "learning_rate": 6.534080971947554e-07,
+      "loss": 0.4371855854988098,
+      "step": 3491
+    },
+    {
+      "epoch": 0.8903620601733809,
+      "grad_norm": 0.41994306445121765,
+      "learning_rate": 6.504128247270547e-07,
+      "loss": 0.4364043176174164,
+      "step": 3492
+    },
+    {
+      "epoch": 0.890617032126466,
+      "grad_norm": 0.467311829328537,
+      "learning_rate": 6.474242025342991e-07,
+      "loss": 0.4354938864707947,
+      "step": 3493
+    },
+    {
+      "epoch": 0.8908720040795512,
+      "grad_norm": 0.4528542160987854,
+      "learning_rate": 6.44442232742265e-07,
+      "loss": 0.42929911613464355,
+      "step": 3494
+    },
+    {
+      "epoch": 0.8911269760326365,
+      "grad_norm": 0.45788252353668213,
+      "learning_rate": 6.414669174720001e-07,
+      "loss": 0.4449249505996704,
+      "step": 3495
+    },
+    {
+      "epoch": 0.8913819479857216,
+      "grad_norm": 0.45263516902923584,
+      "learning_rate": 6.384982588398225e-07,
+      "loss": 0.4337441325187683,
+      "step": 3496
+    },
+    {
+      "epoch": 0.8916369199388068,
+      "grad_norm": 0.4540601968765259,
+      "learning_rate": 6.355362589573078e-07,
+      "loss": 0.4453548192977905,
+      "step": 3497
+    },
+    {
+      "epoch": 0.8918918918918919,
+      "grad_norm": 0.4698966443538666,
+      "learning_rate": 6.325809199313015e-07,
+      "loss": 0.4402586817741394,
+      "step": 3498
+    },
+    {
+      "epoch": 0.8921468638449771,
+      "grad_norm": 0.45818012952804565,
+      "learning_rate": 6.29632243863908e-07,
+      "loss": 0.41622447967529297,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8924018357980622,
+      "grad_norm": 0.4561059772968292,
+      "learning_rate": 6.266902328524915e-07,
+      "loss": 0.42476189136505127,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8926568077511474,
+      "grad_norm": 0.45123928785324097,
+      "learning_rate": 6.237548889896805e-07,
+      "loss": 0.43091583251953125,
+      "step": 3501
+    },
+    {
+      "epoch": 0.8929117797042325,
+      "grad_norm": 0.4374764859676361,
+      "learning_rate": 6.208262143633548e-07,
+      "loss": 0.427135169506073,
+      "step": 3502
+    },
+    {
+      "epoch": 0.8931667516573177,
+      "grad_norm": 0.43948739767074585,
+      "learning_rate": 6.179042110566558e-07,
+      "loss": 0.43492886424064636,
+      "step": 3503
+    },
+    {
+      "epoch": 0.8934217236104028,
+      "grad_norm": 0.43435484170913696,
+      "learning_rate": 6.149888811479787e-07,
+      "loss": 0.4235824942588806,
+      "step": 3504
+    },
+    {
+      "epoch": 0.893676695563488,
+      "grad_norm": 0.42728570103645325,
+      "learning_rate": 6.12080226710966e-07,
+      "loss": 0.4385955035686493,
+      "step": 3505
+    },
+    {
+      "epoch": 0.8939316675165732,
+      "grad_norm": 0.4453405737876892,
+      "learning_rate": 6.091782498145204e-07,
+      "loss": 0.44239622354507446,
+      "step": 3506
+    },
+    {
+      "epoch": 0.8941866394696584,
+      "grad_norm": 0.45244428515434265,
+      "learning_rate": 6.062829525227909e-07,
+      "loss": 0.4283861517906189,
+      "step": 3507
+    },
+    {
+      "epoch": 0.8944416114227435,
+      "grad_norm": 0.42372480034828186,
+      "learning_rate": 6.033943368951745e-07,
+      "loss": 0.42147597670555115,
+      "step": 3508
+    },
+    {
+      "epoch": 0.8946965833758287,
+      "grad_norm": 0.4271625876426697,
+      "learning_rate": 6.005124049863187e-07,
+      "loss": 0.4466112554073334,
+      "step": 3509
+    },
+    {
+      "epoch": 0.8949515553289138,
+      "grad_norm": 0.4325474798679352,
+      "learning_rate": 5.976371588461138e-07,
+      "loss": 0.4398912787437439,
+      "step": 3510
+    },
+    {
+      "epoch": 0.895206527281999,
+      "grad_norm": 0.4471154510974884,
+      "learning_rate": 5.94768600519694e-07,
+      "loss": 0.43469834327697754,
+      "step": 3511
+    },
+    {
+      "epoch": 0.8954614992350841,
+      "grad_norm": 0.4322305917739868,
+      "learning_rate": 5.919067320474425e-07,
+      "loss": 0.41751420497894287,
+      "step": 3512
+    },
+    {
+      "epoch": 0.8957164711881693,
+      "grad_norm": 0.44496405124664307,
+      "learning_rate": 5.890515554649756e-07,
+      "loss": 0.4205957055091858,
+      "step": 3513
+    },
+    {
+      "epoch": 0.8959714431412544,
+      "grad_norm": 0.41922953724861145,
+      "learning_rate": 5.862030728031554e-07,
+      "loss": 0.4464760422706604,
+      "step": 3514
+    },
+    {
+      "epoch": 0.8962264150943396,
+      "grad_norm": 0.4451551139354706,
+      "learning_rate": 5.833612860880844e-07,
+      "loss": 0.434225857257843,
+      "step": 3515
+    },
+    {
+      "epoch": 0.8964813870474247,
+      "grad_norm": 0.4299732744693756,
+      "learning_rate": 5.805261973410936e-07,
+      "loss": 0.4436805844306946,
+      "step": 3516
+    },
+    {
+      "epoch": 0.89673635900051,
+      "grad_norm": 0.442019522190094,
+      "learning_rate": 5.77697808578761e-07,
+      "loss": 0.4386199116706848,
+      "step": 3517
+    },
+    {
+      "epoch": 0.8969913309535951,
+      "grad_norm": 0.4500035345554352,
+      "learning_rate": 5.748761218128918e-07,
+      "loss": 0.4361865222454071,
+      "step": 3518
+    },
+    {
+      "epoch": 0.8972463029066803,
+      "grad_norm": 0.42633429169654846,
+      "learning_rate": 5.720611390505215e-07,
+      "loss": 0.4213171899318695,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8975012748597654,
+      "grad_norm": 0.4605890214443207,
+      "learning_rate": 5.692528622939275e-07,
+      "loss": 0.42544159293174744,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8977562468128506,
+      "grad_norm": 0.4514401853084564,
+      "learning_rate": 5.664512935406063e-07,
+      "loss": 0.4186820685863495,
+      "step": 3521
+    },
+    {
+      "epoch": 0.8980112187659357,
+      "grad_norm": 0.43343934416770935,
+      "learning_rate": 5.636564347832907e-07,
+      "loss": 0.41555485129356384,
+      "step": 3522
+    },
+    {
+      "epoch": 0.8982661907190209,
+      "grad_norm": 0.46844685077667236,
+      "learning_rate": 5.60868288009937e-07,
+      "loss": 0.4361535906791687,
+      "step": 3523
+    },
+    {
+      "epoch": 0.898521162672106,
+      "grad_norm": 0.4280957877635956,
+      "learning_rate": 5.580868552037277e-07,
+      "loss": 0.4293152987957001,
+      "step": 3524
+    },
+    {
+      "epoch": 0.8987761346251912,
+      "grad_norm": 0.4266355633735657,
+      "learning_rate": 5.553121383430715e-07,
+      "loss": 0.4427189230918884,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8990311065782763,
+      "grad_norm": 0.425010085105896,
+      "learning_rate": 5.525441394015973e-07,
+      "loss": 0.43918511271476746,
+      "step": 3526
+    },
+    {
+      "epoch": 0.8992860785313616,
+      "grad_norm": 0.42402413487434387,
+      "learning_rate": 5.49782860348157e-07,
+      "loss": 0.41261357069015503,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8995410504844467,
+      "grad_norm": 0.46322718262672424,
+      "learning_rate": 5.470283031468259e-07,
+      "loss": 0.4363265037536621,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8997960224375319,
+      "grad_norm": 0.4325588643550873,
+      "learning_rate": 5.442804697568915e-07,
+      "loss": 0.4324786067008972,
+      "step": 3529
+    },
+    {
+      "epoch": 0.9000509943906171,
+      "grad_norm": 0.4412180781364441,
+      "learning_rate": 5.415393621328658e-07,
+      "loss": 0.4369555413722992,
+      "step": 3530
+    },
+    {
+      "epoch": 0.9003059663437022,
+      "grad_norm": 0.4297512471675873,
+      "learning_rate": 5.388049822244701e-07,
+      "loss": 0.43392789363861084,
+      "step": 3531
+    },
+    {
+      "epoch": 0.9005609382967874,
+      "grad_norm": 0.4450361132621765,
+      "learning_rate": 5.360773319766443e-07,
+      "loss": 0.43461984395980835,
+      "step": 3532
+    },
+    {
+      "epoch": 0.9008159102498725,
+      "grad_norm": 0.44984447956085205,
+      "learning_rate": 5.333564133295421e-07,
+      "loss": 0.4311913847923279,
+      "step": 3533
+    },
+    {
+      "epoch": 0.9010708822029577,
+      "grad_norm": 0.4270000159740448,
+      "learning_rate": 5.306422282185264e-07,
+      "loss": 0.4367779493331909,
+      "step": 3534
+    },
+    {
+      "epoch": 0.9013258541560428,
+      "grad_norm": 0.4270723760128021,
+      "learning_rate": 5.279347785741718e-07,
+      "loss": 0.4244728982448578,
+      "step": 3535
+    },
+    {
+      "epoch": 0.901580826109128,
+      "grad_norm": 0.44941937923431396,
+      "learning_rate": 5.252340663222666e-07,
+      "loss": 0.4332941770553589,
+      "step": 3536
+    },
+    {
+      "epoch": 0.9018357980622131,
+      "grad_norm": 0.43194061517715454,
+      "learning_rate": 5.225400933837954e-07,
+      "loss": 0.4334859848022461,
+      "step": 3537
+    },
+    {
+      "epoch": 0.9020907700152984,
+      "grad_norm": 0.42677220702171326,
+      "learning_rate": 5.198528616749599e-07,
+      "loss": 0.4181649088859558,
+      "step": 3538
+    },
+    {
+      "epoch": 0.9023457419683835,
+      "grad_norm": 0.43355873227119446,
+      "learning_rate": 5.171723731071643e-07,
+      "loss": 0.44718942046165466,
+      "step": 3539
+    },
+    {
+      "epoch": 0.9026007139214687,
+      "grad_norm": 0.42261987924575806,
+      "learning_rate": 5.144986295870125e-07,
+      "loss": 0.4279809892177582,
+      "step": 3540
+    },
+    {
+      "epoch": 0.9028556858745538,
+      "grad_norm": 0.43762263655662537,
+      "learning_rate": 5.118316330163165e-07,
+      "loss": 0.43330079317092896,
+      "step": 3541
+    },
+    {
+      "epoch": 0.903110657827639,
+      "grad_norm": 0.4268043339252472,
+      "learning_rate": 5.091713852920854e-07,
+      "loss": 0.4417315423488617,
+      "step": 3542
+    },
+    {
+      "epoch": 0.9033656297807241,
+      "grad_norm": 0.45512381196022034,
+      "learning_rate": 5.065178883065258e-07,
+      "loss": 0.4269631803035736,
+      "step": 3543
+    },
+    {
+      "epoch": 0.9036206017338093,
+      "grad_norm": 0.42681342363357544,
+      "learning_rate": 5.038711439470511e-07,
+      "loss": 0.439994752407074,
+      "step": 3544
+    },
+    {
+      "epoch": 0.9038755736868944,
+      "grad_norm": 0.4336576759815216,
+      "learning_rate": 5.0123115409626e-07,
+      "loss": 0.43351709842681885,
+      "step": 3545
+    },
+    {
+      "epoch": 0.9041305456399796,
+      "grad_norm": 0.4286787211894989,
+      "learning_rate": 4.98597920631958e-07,
+      "loss": 0.43429574370384216,
+      "step": 3546
+    },
+    {
+      "epoch": 0.9043855175930647,
+      "grad_norm": 0.449375718832016,
+      "learning_rate": 4.95971445427137e-07,
+      "loss": 0.45396527647972107,
+      "step": 3547
+    },
+    {
+      "epoch": 0.90464048954615,
+      "grad_norm": 0.4293923079967499,
+      "learning_rate": 4.933517303499846e-07,
+      "loss": 0.43527752161026,
+      "step": 3548
+    },
+    {
+      "epoch": 0.904895461499235,
+      "grad_norm": 0.4441344738006592,
+      "learning_rate": 4.907387772638839e-07,
+      "loss": 0.44533535838127136,
+      "step": 3549
+    },
+    {
+      "epoch": 0.9051504334523203,
+      "grad_norm": 0.44053396582603455,
+      "learning_rate": 4.881325880273968e-07,
+      "loss": 0.44325876235961914,
+      "step": 3550
+    },
+    {
+      "epoch": 0.9054054054054054,
+      "grad_norm": 0.4394433796405792,
+      "learning_rate": 4.855331644942873e-07,
+      "loss": 0.4454990327358246,
+      "step": 3551
+    },
+    {
+      "epoch": 0.9056603773584906,
+      "grad_norm": 0.4348095953464508,
+      "learning_rate": 4.829405085134997e-07,
+      "loss": 0.4354587197303772,
+      "step": 3552
+    },
+    {
+      "epoch": 0.9059153493115757,
+      "grad_norm": 0.4534893333911896,
+      "learning_rate": 4.803546219291655e-07,
+      "loss": 0.4252866506576538,
+      "step": 3553
+    },
+    {
+      "epoch": 0.9061703212646609,
+      "grad_norm": 0.4208613634109497,
+      "learning_rate": 4.777755065806044e-07,
+      "loss": 0.4360646605491638,
+      "step": 3554
+    },
+    {
+      "epoch": 0.906425293217746,
+      "grad_norm": 0.4523314833641052,
+      "learning_rate": 4.752031643023158e-07,
+      "loss": 0.4250752925872803,
+      "step": 3555
+    },
+    {
+      "epoch": 0.9066802651708312,
+      "grad_norm": 0.4313192665576935,
+      "learning_rate": 4.726375969239816e-07,
+      "loss": 0.4444052577018738,
+      "step": 3556
+    },
+    {
+      "epoch": 0.9069352371239163,
+      "grad_norm": 0.4894581735134125,
+      "learning_rate": 4.700788062704686e-07,
+      "loss": 0.43131422996520996,
+      "step": 3557
+    },
+    {
+      "epoch": 0.9071902090770015,
+      "grad_norm": 0.45645684003829956,
+      "learning_rate": 4.675267941618178e-07,
+      "loss": 0.4375073313713074,
+      "step": 3558
+    },
+    {
+      "epoch": 0.9074451810300866,
+      "grad_norm": 0.43935051560401917,
+      "learning_rate": 4.6498156241325364e-07,
+      "loss": 0.43498870730400085,
+      "step": 3559
+    },
+    {
+      "epoch": 0.9077001529831719,
+      "grad_norm": 0.4343809485435486,
+      "learning_rate": 4.6244311283517805e-07,
+      "loss": 0.43834230303764343,
+      "step": 3560
+    },
+    {
+      "epoch": 0.907955124936257,
+      "grad_norm": 0.4208904802799225,
+      "learning_rate": 4.599114472331634e-07,
+      "loss": 0.4369846284389496,
+      "step": 3561
+    },
+    {
+      "epoch": 0.9082100968893422,
+      "grad_norm": 0.44323861598968506,
+      "learning_rate": 4.573865674079625e-07,
+      "loss": 0.4203859567642212,
+      "step": 3562
+    },
+    {
+      "epoch": 0.9084650688424273,
+      "grad_norm": 0.48349323868751526,
+      "learning_rate": 4.548684751554988e-07,
+      "loss": 0.4276666045188904,
+      "step": 3563
+    },
+    {
+      "epoch": 0.9087200407955125,
+      "grad_norm": 0.4553949236869812,
+      "learning_rate": 4.5235717226686516e-07,
+      "loss": 0.43263620138168335,
+      "step": 3564
+    },
+    {
+      "epoch": 0.9089750127485976,
+      "grad_norm": 0.437443345785141,
+      "learning_rate": 4.4985266052833043e-07,
+      "loss": 0.4380514621734619,
+      "step": 3565
+    },
+    {
+      "epoch": 0.9092299847016828,
+      "grad_norm": 0.41560494899749756,
+      "learning_rate": 4.4735494172133074e-07,
+      "loss": 0.4358428418636322,
+      "step": 3566
+    },
+    {
+      "epoch": 0.909484956654768,
+      "grad_norm": 0.44587910175323486,
+      "learning_rate": 4.4486401762246947e-07,
+      "loss": 0.44075244665145874,
+      "step": 3567
+    },
+    {
+      "epoch": 0.9097399286078531,
+      "grad_norm": 0.4215124547481537,
+      "learning_rate": 4.423798900035203e-07,
+      "loss": 0.42927083373069763,
+      "step": 3568
+    },
+    {
+      "epoch": 0.9099949005609383,
+      "grad_norm": 0.44220197200775146,
+      "learning_rate": 4.3990256063141667e-07,
+      "loss": 0.4306790232658386,
+      "step": 3569
+    },
+    {
+      "epoch": 0.9102498725140235,
+      "grad_norm": 0.44558820128440857,
+      "learning_rate": 4.374320312682612e-07,
+      "loss": 0.4551706910133362,
+      "step": 3570
+    },
+    {
+      "epoch": 0.9105048444671087,
+      "grad_norm": 0.4541255533695221,
+      "learning_rate": 4.349683036713215e-07,
+      "loss": 0.43386468291282654,
+      "step": 3571
+    },
+    {
+      "epoch": 0.9107598164201938,
+      "grad_norm": 0.42653003334999084,
+      "learning_rate": 4.3251137959302023e-07,
+      "loss": 0.4319538474082947,
+      "step": 3572
+    },
+    {
+      "epoch": 0.911014788373279,
+      "grad_norm": 0.436084508895874,
+      "learning_rate": 4.300612607809473e-07,
+      "loss": 0.4303576946258545,
+      "step": 3573
+    },
+    {
+      "epoch": 0.9112697603263641,
+      "grad_norm": 0.452581524848938,
+      "learning_rate": 4.276179489778487e-07,
+      "loss": 0.4242931306362152,
+      "step": 3574
+    },
+    {
+      "epoch": 0.9115247322794493,
+      "grad_norm": 0.42535877227783203,
+      "learning_rate": 4.251814459216286e-07,
+      "loss": 0.4394821524620056,
+      "step": 3575
+    },
+    {
+      "epoch": 0.9117797042325344,
+      "grad_norm": 0.45626556873321533,
+      "learning_rate": 4.227517533453518e-07,
+      "loss": 0.45337963104248047,
+      "step": 3576
+    },
+    {
+      "epoch": 0.9120346761856196,
+      "grad_norm": 0.43536752462387085,
+      "learning_rate": 4.203288729772326e-07,
+      "loss": 0.442407488822937,
+      "step": 3577
+    },
+    {
+      "epoch": 0.9122896481387047,
+      "grad_norm": 0.43327459692955017,
+      "learning_rate": 4.179128065406468e-07,
+      "loss": 0.43041032552719116,
+      "step": 3578
+    },
+    {
+      "epoch": 0.9125446200917899,
+      "grad_norm": 0.4425260126590729,
+      "learning_rate": 4.155035557541165e-07,
+      "loss": 0.4451976716518402,
+      "step": 3579
+    },
+    {
+      "epoch": 0.912799592044875,
+      "grad_norm": 0.4322320818901062,
+      "learning_rate": 4.1310112233132303e-07,
+      "loss": 0.4459317922592163,
+      "step": 3580
+    },
+    {
+      "epoch": 0.9130545639979603,
+      "grad_norm": 0.44417110085487366,
+      "learning_rate": 4.107055079810951e-07,
+      "loss": 0.42950934171676636,
+      "step": 3581
+    },
+    {
+      "epoch": 0.9133095359510454,
+      "grad_norm": 0.44251391291618347,
+      "learning_rate": 4.083167144074074e-07,
+      "loss": 0.43427062034606934,
+      "step": 3582
+    },
+    {
+      "epoch": 0.9135645079041306,
+      "grad_norm": 0.4517573118209839,
+      "learning_rate": 4.059347433093896e-07,
+      "loss": 0.43051815032958984,
+      "step": 3583
+    },
+    {
+      "epoch": 0.9138194798572157,
+      "grad_norm": 0.44148555397987366,
+      "learning_rate": 4.0355959638131545e-07,
+      "loss": 0.43934953212738037,
+      "step": 3584
+    },
+    {
+      "epoch": 0.9140744518103009,
+      "grad_norm": 0.4492388367652893,
+      "learning_rate": 4.0119127531260347e-07,
+      "loss": 0.4464166760444641,
+      "step": 3585
+    },
+    {
+      "epoch": 0.914329423763386,
+      "grad_norm": 0.4425787329673767,
+      "learning_rate": 3.988297817878217e-07,
+      "loss": 0.44234633445739746,
+      "step": 3586
+    },
+    {
+      "epoch": 0.9145843957164712,
+      "grad_norm": 0.43447956442832947,
+      "learning_rate": 3.9647511748667653e-07,
+      "loss": 0.444368451833725,
+      "step": 3587
+    },
+    {
+      "epoch": 0.9148393676695563,
+      "grad_norm": 0.44007408618927,
+      "learning_rate": 3.9412728408401714e-07,
+      "loss": 0.43814414739608765,
+      "step": 3588
+    },
+    {
+      "epoch": 0.9150943396226415,
+      "grad_norm": 0.42095082998275757,
+      "learning_rate": 3.917862832498387e-07,
+      "loss": 0.4246536195278168,
+      "step": 3589
+    },
+    {
+      "epoch": 0.9153493115757266,
+      "grad_norm": 0.42961618304252625,
+      "learning_rate": 3.8945211664927154e-07,
+      "loss": 0.42815613746643066,
+      "step": 3590
+    },
+    {
+      "epoch": 0.9156042835288118,
+      "grad_norm": 0.4285229444503784,
+      "learning_rate": 3.871247859425886e-07,
+      "loss": 0.43332967162132263,
+      "step": 3591
+    },
+    {
+      "epoch": 0.915859255481897,
+      "grad_norm": 0.42834195494651794,
+      "learning_rate": 3.84804292785198e-07,
+      "loss": 0.4423152208328247,
+      "step": 3592
+    },
+    {
+      "epoch": 0.9161142274349822,
+      "grad_norm": 0.5144716501235962,
+      "learning_rate": 3.82490638827645e-07,
+      "loss": 0.43932145833969116,
+      "step": 3593
+    },
+    {
+      "epoch": 0.9163691993880673,
+      "grad_norm": 0.41799429059028625,
+      "learning_rate": 3.8018382571560876e-07,
+      "loss": 0.4334896206855774,
+      "step": 3594
+    },
+    {
+      "epoch": 0.9166241713411525,
+      "grad_norm": 0.427685409784317,
+      "learning_rate": 3.7788385508990667e-07,
+      "loss": 0.4288986921310425,
+      "step": 3595
+    },
+    {
+      "epoch": 0.9168791432942376,
+      "grad_norm": 0.4473314881324768,
+      "learning_rate": 3.755907285864846e-07,
+      "loss": 0.44231194257736206,
+      "step": 3596
+    },
+    {
+      "epoch": 0.9171341152473228,
+      "grad_norm": 0.45096901059150696,
+      "learning_rate": 3.733044478364234e-07,
+      "loss": 0.43934544920921326,
+      "step": 3597
+    },
+    {
+      "epoch": 0.9173890872004079,
+      "grad_norm": 0.43266093730926514,
+      "learning_rate": 3.710250144659322e-07,
+      "loss": 0.43362313508987427,
+      "step": 3598
+    },
+    {
+      "epoch": 0.9176440591534931,
+      "grad_norm": 0.42727604508399963,
+      "learning_rate": 3.6875243009635406e-07,
+      "loss": 0.4410891532897949,
+      "step": 3599
+    },
+    {
+      "epoch": 0.9178990311065782,
+      "grad_norm": 0.4408717751502991,
+      "learning_rate": 3.664866963441538e-07,
+      "loss": 0.43474286794662476,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9181540030596634,
+      "grad_norm": 0.4211311936378479,
+      "learning_rate": 3.6422781482092664e-07,
+      "loss": 0.43140873312950134,
+      "step": 3601
+    },
+    {
+      "epoch": 0.9184089750127487,
+      "grad_norm": 0.4141659736633301,
+      "learning_rate": 3.619757871333973e-07,
+      "loss": 0.4358154535293579,
+      "step": 3602
+    },
+    {
+      "epoch": 0.9186639469658338,
+      "grad_norm": 0.4296232759952545,
+      "learning_rate": 3.59730614883409e-07,
+      "loss": 0.43090933561325073,
+      "step": 3603
+    },
+    {
+      "epoch": 0.918918918918919,
+      "grad_norm": 0.4304314851760864,
+      "learning_rate": 3.574922996679331e-07,
+      "loss": 0.44070133566856384,
+      "step": 3604
+    },
+    {
+      "epoch": 0.9191738908720041,
+      "grad_norm": 0.45988893508911133,
+      "learning_rate": 3.552608430790649e-07,
+      "loss": 0.4354715049266815,
+      "step": 3605
+    },
+    {
+      "epoch": 0.9194288628250893,
+      "grad_norm": 0.43448030948638916,
+      "learning_rate": 3.53036246704016e-07,
+      "loss": 0.43007493019104004,
+      "step": 3606
+    },
+    {
+      "epoch": 0.9196838347781744,
+      "grad_norm": 0.4371713399887085,
+      "learning_rate": 3.5081851212512176e-07,
+      "loss": 0.44341692328453064,
+      "step": 3607
+    },
+    {
+      "epoch": 0.9199388067312596,
+      "grad_norm": 0.44628259539604187,
+      "learning_rate": 3.486076409198369e-07,
+      "loss": 0.42946192622184753,
+      "step": 3608
+    },
+    {
+      "epoch": 0.9201937786843447,
+      "grad_norm": 0.47736960649490356,
+      "learning_rate": 3.464036346607336e-07,
+      "loss": 0.45520347356796265,
+      "step": 3609
+    },
+    {
+      "epoch": 0.9204487506374299,
+      "grad_norm": 0.4316670894622803,
+      "learning_rate": 3.442064949155011e-07,
+      "loss": 0.42847108840942383,
+      "step": 3610
+    },
+    {
+      "epoch": 0.920703722590515,
+      "grad_norm": 0.4310346841812134,
+      "learning_rate": 3.420162232469437e-07,
+      "loss": 0.43078848719596863,
+      "step": 3611
+    },
+    {
+      "epoch": 0.9209586945436002,
+      "grad_norm": 0.4333515167236328,
+      "learning_rate": 3.398328212129809e-07,
+      "loss": 0.42883485555648804,
+      "step": 3612
+    },
+    {
+      "epoch": 0.9212136664966853,
+      "grad_norm": 0.45875313878059387,
+      "learning_rate": 3.3765629036665135e-07,
+      "loss": 0.42375749349594116,
+      "step": 3613
+    },
+    {
+      "epoch": 0.9214686384497706,
+      "grad_norm": 0.43513643741607666,
+      "learning_rate": 3.354866322560946e-07,
+      "loss": 0.4320690631866455,
+      "step": 3614
+    },
+    {
+      "epoch": 0.9217236104028557,
+      "grad_norm": 0.43181368708610535,
+      "learning_rate": 3.3332384842457156e-07,
+      "loss": 0.43557125329971313,
+      "step": 3615
+    },
+    {
+      "epoch": 0.9219785823559409,
+      "grad_norm": 0.4194730818271637,
+      "learning_rate": 3.3116794041045174e-07,
+      "loss": 0.43037256598472595,
+      "step": 3616
+    },
+    {
+      "epoch": 0.922233554309026,
+      "grad_norm": 0.4568606913089752,
+      "learning_rate": 3.290189097472096e-07,
+      "loss": 0.43266376852989197,
+      "step": 3617
+    },
+    {
+      "epoch": 0.9224885262621112,
+      "grad_norm": 0.435941606760025,
+      "learning_rate": 3.2687675796343244e-07,
+      "loss": 0.43532881140708923,
+      "step": 3618
+    },
+    {
+      "epoch": 0.9227434982151963,
+      "grad_norm": 0.47094932198524475,
+      "learning_rate": 3.247414865828136e-07,
+      "loss": 0.43585801124572754,
+      "step": 3619
+    },
+    {
+      "epoch": 0.9229984701682815,
+      "grad_norm": 0.45321404933929443,
+      "learning_rate": 3.226130971241503e-07,
+      "loss": 0.4414428472518921,
+      "step": 3620
+    },
+    {
+      "epoch": 0.9232534421213666,
+      "grad_norm": 0.4174649119377136,
+      "learning_rate": 3.204915911013484e-07,
+      "loss": 0.44382810592651367,
+      "step": 3621
+    },
+    {
+      "epoch": 0.9235084140744518,
+      "grad_norm": 0.4502234160900116,
+      "learning_rate": 3.1837697002341293e-07,
+      "loss": 0.41893133521080017,
+      "step": 3622
+    },
+    {
+      "epoch": 0.9237633860275369,
+      "grad_norm": 0.4951938986778259,
+      "learning_rate": 3.162692353944563e-07,
+      "loss": 0.4423009157180786,
+      "step": 3623
+    },
+    {
+      "epoch": 0.9240183579806222,
+      "grad_norm": 0.471845805644989,
+      "learning_rate": 3.1416838871368925e-07,
+      "loss": 0.4448579251766205,
+      "step": 3624
+    },
+    {
+      "epoch": 0.9242733299337073,
+      "grad_norm": 0.4562813937664032,
+      "learning_rate": 3.120744314754254e-07,
+      "loss": 0.4233444333076477,
+      "step": 3625
+    },
+    {
+      "epoch": 0.9245283018867925,
+      "grad_norm": 0.4315374195575714,
+      "learning_rate": 3.099873651690788e-07,
+      "loss": 0.4357251524925232,
+      "step": 3626
+    },
+    {
+      "epoch": 0.9247832738398776,
+      "grad_norm": 0.4478853642940521,
+      "learning_rate": 3.0790719127915647e-07,
+      "loss": 0.45080792903900146,
+      "step": 3627
+    },
+    {
+      "epoch": 0.9250382457929628,
+      "grad_norm": 0.443620502948761,
+      "learning_rate": 3.058339112852693e-07,
+      "loss": 0.4329032897949219,
+      "step": 3628
+    },
+    {
+      "epoch": 0.9252932177460479,
+      "grad_norm": 0.42865297198295593,
+      "learning_rate": 3.0376752666212317e-07,
+      "loss": 0.4263918399810791,
+      "step": 3629
+    },
+    {
+      "epoch": 0.9255481896991331,
+      "grad_norm": 0.43004220724105835,
+      "learning_rate": 3.017080388795168e-07,
+      "loss": 0.4549320936203003,
+      "step": 3630
+    },
+    {
+      "epoch": 0.9258031616522182,
+      "grad_norm": 0.43752700090408325,
+      "learning_rate": 2.9965544940234624e-07,
+      "loss": 0.4366416931152344,
+      "step": 3631
+    },
+    {
+      "epoch": 0.9260581336053034,
+      "grad_norm": 0.432623028755188,
+      "learning_rate": 2.976097596905969e-07,
+      "loss": 0.4313819110393524,
+      "step": 3632
+    },
+    {
+      "epoch": 0.9263131055583885,
+      "grad_norm": 0.43730250000953674,
+      "learning_rate": 2.955709711993515e-07,
+      "loss": 0.43685126304626465,
+      "step": 3633
+    },
+    {
+      "epoch": 0.9265680775114737,
+      "grad_norm": 0.4221460521221161,
+      "learning_rate": 2.9353908537878004e-07,
+      "loss": 0.42311978340148926,
+      "step": 3634
+    },
+    {
+      "epoch": 0.9268230494645588,
+      "grad_norm": 0.43018046021461487,
+      "learning_rate": 2.915141036741442e-07,
+      "loss": 0.43955671787261963,
+      "step": 3635
+    },
+    {
+      "epoch": 0.9270780214176441,
+      "grad_norm": 0.4328952431678772,
+      "learning_rate": 2.894960275257941e-07,
+      "loss": 0.4308651387691498,
+      "step": 3636
+    },
+    {
+      "epoch": 0.9273329933707293,
+      "grad_norm": 0.44335830211639404,
+      "learning_rate": 2.874848583691714e-07,
+      "loss": 0.438312828540802,
+      "step": 3637
+    },
+    {
+      "epoch": 0.9275879653238144,
+      "grad_norm": 0.4323786497116089,
+      "learning_rate": 2.854805976348007e-07,
+      "loss": 0.43586835265159607,
+      "step": 3638
+    },
+    {
+      "epoch": 0.9278429372768996,
+      "grad_norm": 0.42834949493408203,
+      "learning_rate": 2.8348324674829174e-07,
+      "loss": 0.442486047744751,
+      "step": 3639
+    },
+    {
+      "epoch": 0.9280979092299847,
+      "grad_norm": 0.4232763350009918,
+      "learning_rate": 2.8149280713034575e-07,
+      "loss": 0.43259090185165405,
+      "step": 3640
+    },
+    {
+      "epoch": 0.9283528811830699,
+      "grad_norm": 0.4621821641921997,
+      "learning_rate": 2.795092801967414e-07,
+      "loss": 0.4362543225288391,
+      "step": 3641
+    },
+    {
+      "epoch": 0.928607853136155,
+      "grad_norm": 0.4357510805130005,
+      "learning_rate": 2.775326673583434e-07,
+      "loss": 0.45379358530044556,
+      "step": 3642
+    },
+    {
+      "epoch": 0.9288628250892402,
+      "grad_norm": 0.4349755048751831,
+      "learning_rate": 2.755629700210982e-07,
+      "loss": 0.4319487512111664,
+      "step": 3643
+    },
+    {
+      "epoch": 0.9291177970423253,
+      "grad_norm": 0.4547295868396759,
+      "learning_rate": 2.7360018958603405e-07,
+      "loss": 0.43946123123168945,
+      "step": 3644
+    },
+    {
+      "epoch": 0.9293727689954105,
+      "grad_norm": 0.44599592685699463,
+      "learning_rate": 2.716443274492597e-07,
+      "loss": 0.4380370080471039,
+      "step": 3645
+    },
+    {
+      "epoch": 0.9296277409484957,
+      "grad_norm": 0.44466736912727356,
+      "learning_rate": 2.696953850019579e-07,
+      "loss": 0.43402618169784546,
+      "step": 3646
+    },
+    {
+      "epoch": 0.9298827129015809,
+      "grad_norm": 0.4327200949192047,
+      "learning_rate": 2.677533636303964e-07,
+      "loss": 0.44082748889923096,
+      "step": 3647
+    },
+    {
+      "epoch": 0.930137684854666,
+      "grad_norm": 0.439943790435791,
+      "learning_rate": 2.658182647159169e-07,
+      "loss": 0.43326807022094727,
+      "step": 3648
+    },
+    {
+      "epoch": 0.9303926568077512,
+      "grad_norm": 0.42451268434524536,
+      "learning_rate": 2.6389008963493613e-07,
+      "loss": 0.43715566396713257,
+      "step": 3649
+    },
+    {
+      "epoch": 0.9306476287608363,
+      "grad_norm": 0.42902666330337524,
+      "learning_rate": 2.6196883975894817e-07,
+      "loss": 0.42805609107017517,
+      "step": 3650
+    },
+    {
+      "epoch": 0.9309026007139215,
+      "grad_norm": 0.44421282410621643,
+      "learning_rate": 2.60054516454521e-07,
+      "loss": 0.44107586145401,
+      "step": 3651
+    },
+    {
+      "epoch": 0.9311575726670066,
+      "grad_norm": 0.44369035959243774,
+      "learning_rate": 2.581471210832931e-07,
+      "loss": 0.43590283393859863,
+      "step": 3652
+    },
+    {
+      "epoch": 0.9314125446200918,
+      "grad_norm": 0.4411109983921051,
+      "learning_rate": 2.5624665500197933e-07,
+      "loss": 0.4290148615837097,
+      "step": 3653
+    },
+    {
+      "epoch": 0.9316675165731769,
+      "grad_norm": 0.566591739654541,
+      "learning_rate": 2.543531195623605e-07,
+      "loss": 0.4412996172904968,
+      "step": 3654
+    },
+    {
+      "epoch": 0.9319224885262621,
+      "grad_norm": 0.44682958722114563,
+      "learning_rate": 2.524665161112949e-07,
+      "loss": 0.44440752267837524,
+      "step": 3655
+    },
+    {
+      "epoch": 0.9321774604793472,
+      "grad_norm": 0.43499064445495605,
+      "learning_rate": 2.505868459907035e-07,
+      "loss": 0.43677276372909546,
+      "step": 3656
+    },
+    {
+      "epoch": 0.9324324324324325,
+      "grad_norm": 0.45111894607543945,
+      "learning_rate": 2.48714110537579e-07,
+      "loss": 0.43129217624664307,
+      "step": 3657
+    },
+    {
+      "epoch": 0.9326874043855176,
+      "grad_norm": 0.44929251074790955,
+      "learning_rate": 2.4684831108398033e-07,
+      "loss": 0.4395376443862915,
+      "step": 3658
+    },
+    {
+      "epoch": 0.9329423763386028,
+      "grad_norm": 0.4287400245666504,
+      "learning_rate": 2.449894489570337e-07,
+      "loss": 0.4335207939147949,
+      "step": 3659
+    },
+    {
+      "epoch": 0.9331973482916879,
+      "grad_norm": 0.42606982588768005,
+      "learning_rate": 2.4313752547893034e-07,
+      "loss": 0.4391172528266907,
+      "step": 3660
+    },
+    {
+      "epoch": 0.9334523202447731,
+      "grad_norm": 0.43672558665275574,
+      "learning_rate": 2.412925419669276e-07,
+      "loss": 0.4412034749984741,
+      "step": 3661
+    },
+    {
+      "epoch": 0.9337072921978582,
+      "grad_norm": 0.43024998903274536,
+      "learning_rate": 2.3945449973334366e-07,
+      "loss": 0.4417959451675415,
+      "step": 3662
+    },
+    {
+      "epoch": 0.9339622641509434,
+      "grad_norm": 0.45105281472206116,
+      "learning_rate": 2.3762340008556261e-07,
+      "loss": 0.44574111700057983,
+      "step": 3663
+    },
+    {
+      "epoch": 0.9342172361040285,
+      "grad_norm": 0.42282834649086,
+      "learning_rate": 2.3579924432602708e-07,
+      "loss": 0.435298889875412,
+      "step": 3664
+    },
+    {
+      "epoch": 0.9344722080571137,
+      "grad_norm": 0.4477471709251404,
+      "learning_rate": 2.339820337522425e-07,
+      "loss": 0.44897374510765076,
+      "step": 3665
+    },
+    {
+      "epoch": 0.9347271800101988,
+      "grad_norm": 0.42740267515182495,
+      "learning_rate": 2.321717696567738e-07,
+      "loss": 0.43435829877853394,
+      "step": 3666
+    },
+    {
+      "epoch": 0.934982151963284,
+      "grad_norm": 0.4464569389820099,
+      "learning_rate": 2.3036845332724545e-07,
+      "loss": 0.44783616065979004,
+      "step": 3667
+    },
+    {
+      "epoch": 0.9352371239163692,
+      "grad_norm": 0.4314689636230469,
+      "learning_rate": 2.2857208604634028e-07,
+      "loss": 0.43364688754081726,
+      "step": 3668
+    },
+    {
+      "epoch": 0.9354920958694544,
+      "grad_norm": 0.42942556738853455,
+      "learning_rate": 2.2678266909179737e-07,
+      "loss": 0.4412001967430115,
+      "step": 3669
+    },
+    {
+      "epoch": 0.9357470678225395,
+      "grad_norm": 0.42958909273147583,
+      "learning_rate": 2.2500020373641297e-07,
+      "loss": 0.4348294734954834,
+      "step": 3670
+    },
+    {
+      "epoch": 0.9360020397756247,
+      "grad_norm": 0.4455057978630066,
+      "learning_rate": 2.2322469124803624e-07,
+      "loss": 0.43147438764572144,
+      "step": 3671
+    },
+    {
+      "epoch": 0.9362570117287098,
+      "grad_norm": 0.45338964462280273,
+      "learning_rate": 2.214561328895748e-07,
+      "loss": 0.442391961812973,
+      "step": 3672
+    },
+    {
+      "epoch": 0.936511983681795,
+      "grad_norm": 0.473303884267807,
+      "learning_rate": 2.1969452991898678e-07,
+      "loss": 0.4215404689311981,
+      "step": 3673
+    },
+    {
+      "epoch": 0.9367669556348802,
+      "grad_norm": 0.46086111664772034,
+      "learning_rate": 2.1793988358928432e-07,
+      "loss": 0.4291722774505615,
+      "step": 3674
+    },
+    {
+      "epoch": 0.9370219275879653,
+      "grad_norm": 0.4351825416088104,
+      "learning_rate": 2.161921951485313e-07,
+      "loss": 0.4376457929611206,
+      "step": 3675
+    },
+    {
+      "epoch": 0.9372768995410505,
+      "grad_norm": 0.4450758397579193,
+      "learning_rate": 2.1445146583984334e-07,
+      "loss": 0.4214126765727997,
+      "step": 3676
+    },
+    {
+      "epoch": 0.9375318714941356,
+      "grad_norm": 0.44977065920829773,
+      "learning_rate": 2.1271769690138334e-07,
+      "loss": 0.42246899008750916,
+      "step": 3677
+    },
+    {
+      "epoch": 0.9377868434472209,
+      "grad_norm": 0.43122920393943787,
+      "learning_rate": 2.1099088956636594e-07,
+      "loss": 0.43919774889945984,
+      "step": 3678
+    },
+    {
+      "epoch": 0.938041815400306,
+      "grad_norm": 0.5114609599113464,
+      "learning_rate": 2.0927104506305308e-07,
+      "loss": 0.4412056803703308,
+      "step": 3679
+    },
+    {
+      "epoch": 0.9382967873533912,
+      "grad_norm": 0.4477209746837616,
+      "learning_rate": 2.0755816461475398e-07,
+      "loss": 0.4255548417568207,
+      "step": 3680
+    },
+    {
+      "epoch": 0.9385517593064763,
+      "grad_norm": 0.4294905960559845,
+      "learning_rate": 2.0585224943982518e-07,
+      "loss": 0.4375752806663513,
+      "step": 3681
+    },
+    {
+      "epoch": 0.9388067312595615,
+      "grad_norm": 0.43354830145835876,
+      "learning_rate": 2.041533007516694e-07,
+      "loss": 0.4328603446483612,
+      "step": 3682
+    },
+    {
+      "epoch": 0.9390617032126466,
+      "grad_norm": 0.4309961199760437,
+      "learning_rate": 2.0246131975873217e-07,
+      "loss": 0.4423938989639282,
+      "step": 3683
+    },
+    {
+      "epoch": 0.9393166751657318,
+      "grad_norm": 0.4355313181877136,
+      "learning_rate": 2.0077630766450195e-07,
+      "loss": 0.4296422004699707,
+      "step": 3684
+    },
+    {
+      "epoch": 0.9395716471188169,
+      "grad_norm": 0.4247247874736786,
+      "learning_rate": 1.9909826566751555e-07,
+      "loss": 0.44954705238342285,
+      "step": 3685
+    },
+    {
+      "epoch": 0.9398266190719021,
+      "grad_norm": 0.45959627628326416,
+      "learning_rate": 1.97427194961346e-07,
+      "loss": 0.4236677885055542,
+      "step": 3686
+    },
+    {
+      "epoch": 0.9400815910249872,
+      "grad_norm": 0.4352707862854004,
+      "learning_rate": 1.9576309673461358e-07,
+      "loss": 0.42161887884140015,
+      "step": 3687
+    },
+    {
+      "epoch": 0.9403365629780724,
+      "grad_norm": 0.4265056550502777,
+      "learning_rate": 1.941059721709715e-07,
+      "loss": 0.4375278353691101,
+      "step": 3688
+    },
+    {
+      "epoch": 0.9405915349311575,
+      "grad_norm": 0.43939751386642456,
+      "learning_rate": 1.9245582244912241e-07,
+      "loss": 0.4362907409667969,
+      "step": 3689
+    },
+    {
+      "epoch": 0.9408465068842428,
+      "grad_norm": 0.41473960876464844,
+      "learning_rate": 1.9081264874279858e-07,
+      "loss": 0.4331972301006317,
+      "step": 3690
+    },
+    {
+      "epoch": 0.9411014788373279,
+      "grad_norm": 0.43197643756866455,
+      "learning_rate": 1.891764522207762e-07,
+      "loss": 0.4336824417114258,
+      "step": 3691
+    },
+    {
+      "epoch": 0.9413564507904131,
+      "grad_norm": 0.4420890808105469,
+      "learning_rate": 1.8754723404686425e-07,
+      "loss": 0.4414839744567871,
+      "step": 3692
+    },
+    {
+      "epoch": 0.9416114227434982,
+      "grad_norm": 0.4495333135128021,
+      "learning_rate": 1.859249953799147e-07,
+      "loss": 0.42230579257011414,
+      "step": 3693
+    },
+    {
+      "epoch": 0.9418663946965834,
+      "grad_norm": 0.44855111837387085,
+      "learning_rate": 1.8430973737380785e-07,
+      "loss": 0.4318239390850067,
+      "step": 3694
+    },
+    {
+      "epoch": 0.9421213666496685,
+      "grad_norm": 0.4337852895259857,
+      "learning_rate": 1.8270146117746245e-07,
+      "loss": 0.43734127283096313,
+      "step": 3695
+    },
+    {
+      "epoch": 0.9423763386027537,
+      "grad_norm": 0.431687593460083,
+      "learning_rate": 1.8110016793483122e-07,
+      "loss": 0.44457387924194336,
+      "step": 3696
+    },
+    {
+      "epoch": 0.9426313105558388,
+      "grad_norm": 0.43883758783340454,
+      "learning_rate": 1.7950585878489856e-07,
+      "loss": 0.418931245803833,
+      "step": 3697
+    },
+    {
+      "epoch": 0.942886282508924,
+      "grad_norm": 0.46315500140190125,
+      "learning_rate": 1.7791853486168188e-07,
+      "loss": 0.4317883253097534,
+      "step": 3698
+    },
+    {
+      "epoch": 0.9431412544620091,
+      "grad_norm": 0.43983328342437744,
+      "learning_rate": 1.7633819729422907e-07,
+      "loss": 0.4401555061340332,
+      "step": 3699
+    },
+    {
+      "epoch": 0.9433962264150944,
+      "grad_norm": 0.42895379662513733,
+      "learning_rate": 1.7476484720662103e-07,
+      "loss": 0.4379884898662567,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9436511983681795,
+      "grad_norm": 0.43615207076072693,
+      "learning_rate": 1.7319848571796805e-07,
+      "loss": 0.4482054114341736,
+      "step": 3701
+    },
+    {
+      "epoch": 0.9439061703212647,
+      "grad_norm": 0.4342696964740753,
+      "learning_rate": 1.7163911394240675e-07,
+      "loss": 0.42798078060150146,
+      "step": 3702
+    },
+    {
+      "epoch": 0.9441611422743498,
+      "grad_norm": 0.4399515390396118,
+      "learning_rate": 1.700867329891043e-07,
+      "loss": 0.43327248096466064,
+      "step": 3703
+    },
+    {
+      "epoch": 0.944416114227435,
+      "grad_norm": 0.44510048627853394,
+      "learning_rate": 1.6854134396225407e-07,
+      "loss": 0.4332675337791443,
+      "step": 3704
+    },
+    {
+      "epoch": 0.9446710861805201,
+      "grad_norm": 0.4416811764240265,
+      "learning_rate": 1.6700294796107907e-07,
+      "loss": 0.44257673621177673,
+      "step": 3705
+    },
+    {
+      "epoch": 0.9449260581336053,
+      "grad_norm": 0.44168102741241455,
+      "learning_rate": 1.6547154607982508e-07,
+      "loss": 0.4348487854003906,
+      "step": 3706
+    },
+    {
+      "epoch": 0.9451810300866904,
+      "grad_norm": 0.43375131487846375,
+      "learning_rate": 1.6394713940776296e-07,
+      "loss": 0.4283615052700043,
+      "step": 3707
+    },
+    {
+      "epoch": 0.9454360020397756,
+      "grad_norm": 0.42909738421440125,
+      "learning_rate": 1.6242972902919095e-07,
+      "loss": 0.43982595205307007,
+      "step": 3708
+    },
+    {
+      "epoch": 0.9456909739928608,
+      "grad_norm": 0.45628172159194946,
+      "learning_rate": 1.6091931602342903e-07,
+      "loss": 0.42930155992507935,
+      "step": 3709
+    },
+    {
+      "epoch": 0.9459459459459459,
+      "grad_norm": 0.4570588171482086,
+      "learning_rate": 1.594159014648178e-07,
+      "loss": 0.4148539900779724,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9462009178990312,
+      "grad_norm": 0.4399675130844116,
+      "learning_rate": 1.5791948642272404e-07,
+      "loss": 0.4340864419937134,
+      "step": 3711
+    },
+    {
+      "epoch": 0.9464558898521163,
+      "grad_norm": 0.439888060092926,
+      "learning_rate": 1.5643007196153303e-07,
+      "loss": 0.43424567580223083,
+      "step": 3712
+    },
+    {
+      "epoch": 0.9467108618052015,
+      "grad_norm": 0.42277246713638306,
+      "learning_rate": 1.5494765914065292e-07,
+      "loss": 0.4288041591644287,
+      "step": 3713
+    },
+    {
+      "epoch": 0.9469658337582866,
+      "grad_norm": 0.42644914984703064,
+      "learning_rate": 1.5347224901450908e-07,
+      "loss": 0.4338779151439667,
+      "step": 3714
+    },
+    {
+      "epoch": 0.9472208057113718,
+      "grad_norm": 0.5191669464111328,
+      "learning_rate": 1.5200384263254763e-07,
+      "loss": 0.4319589138031006,
+      "step": 3715
+    },
+    {
+      "epoch": 0.9474757776644569,
+      "grad_norm": 0.443218469619751,
+      "learning_rate": 1.5054244103923087e-07,
+      "loss": 0.432138055562973,
+      "step": 3716
+    },
+    {
+      "epoch": 0.9477307496175421,
+      "grad_norm": 0.44982388615608215,
+      "learning_rate": 1.4908804527404287e-07,
+      "loss": 0.4303247928619385,
+      "step": 3717
+    },
+    {
+      "epoch": 0.9479857215706272,
+      "grad_norm": 0.454996258020401,
+      "learning_rate": 1.476406563714794e-07,
+      "loss": 0.4347943663597107,
+      "step": 3718
+    },
+    {
+      "epoch": 0.9482406935237124,
+      "grad_norm": 0.43886685371398926,
+      "learning_rate": 1.4620027536105697e-07,
+      "loss": 0.43539103865623474,
+      "step": 3719
+    },
+    {
+      "epoch": 0.9484956654767975,
+      "grad_norm": 0.4551427960395813,
+      "learning_rate": 1.447669032673027e-07,
+      "loss": 0.4332054853439331,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9487506374298827,
+      "grad_norm": 0.4315807819366455,
+      "learning_rate": 1.4334054110976437e-07,
+      "loss": 0.4265736937522888,
+      "step": 3721
+    },
+    {
+      "epoch": 0.9490056093829679,
+      "grad_norm": 0.43956583738327026,
+      "learning_rate": 1.419211899029971e-07,
+      "loss": 0.42740926146507263,
+      "step": 3722
+    },
+    {
+      "epoch": 0.9492605813360531,
+      "grad_norm": 0.4354575276374817,
+      "learning_rate": 1.4050885065657326e-07,
+      "loss": 0.4265577793121338,
+      "step": 3723
+    },
+    {
+      "epoch": 0.9495155532891382,
+      "grad_norm": 0.4236654043197632,
+      "learning_rate": 1.3910352437507711e-07,
+      "loss": 0.435396671295166,
+      "step": 3724
+    },
+    {
+      "epoch": 0.9497705252422234,
+      "grad_norm": 0.4544011056423187,
+      "learning_rate": 1.377052120581046e-07,
+      "loss": 0.4167858064174652,
+      "step": 3725
+    },
+    {
+      "epoch": 0.9500254971953085,
+      "grad_norm": 0.43282467126846313,
+      "learning_rate": 1.3631391470026124e-07,
+      "loss": 0.41828426718711853,
+      "step": 3726
+    },
+    {
+      "epoch": 0.9502804691483937,
+      "grad_norm": 0.4354334771633148,
+      "learning_rate": 1.3492963329116538e-07,
+      "loss": 0.4377305507659912,
+      "step": 3727
+    },
+    {
+      "epoch": 0.9505354411014788,
+      "grad_norm": 0.4522036910057068,
+      "learning_rate": 1.3355236881544276e-07,
+      "loss": 0.44088101387023926,
+      "step": 3728
+    },
+    {
+      "epoch": 0.950790413054564,
+      "grad_norm": 0.4242122173309326,
+      "learning_rate": 1.3218212225272753e-07,
+      "loss": 0.43219560384750366,
+      "step": 3729
+    },
+    {
+      "epoch": 0.9510453850076491,
+      "grad_norm": 0.43885210156440735,
+      "learning_rate": 1.3081889457766672e-07,
+      "loss": 0.43644246459007263,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9513003569607343,
+      "grad_norm": 0.42883220314979553,
+      "learning_rate": 1.2946268675990804e-07,
+      "loss": 0.43907150626182556,
+      "step": 3731
+    },
+    {
+      "epoch": 0.9515553289138194,
+      "grad_norm": 0.43720653653144836,
+      "learning_rate": 1.2811349976411204e-07,
+      "loss": 0.43129271268844604,
+      "step": 3732
+    },
+    {
+      "epoch": 0.9518103008669047,
+      "grad_norm": 0.4418365955352783,
+      "learning_rate": 1.2677133454994105e-07,
+      "loss": 0.4380974769592285,
+      "step": 3733
+    },
+    {
+      "epoch": 0.9520652728199898,
+      "grad_norm": 0.4314889907836914,
+      "learning_rate": 1.2543619207206704e-07,
+      "loss": 0.4347606599330902,
+      "step": 3734
+    },
+    {
+      "epoch": 0.952320244773075,
+      "grad_norm": 0.4545467793941498,
+      "learning_rate": 1.241080732801625e-07,
+      "loss": 0.43105897307395935,
+      "step": 3735
+    },
+    {
+      "epoch": 0.9525752167261601,
+      "grad_norm": 0.4504272937774658,
+      "learning_rate": 1.2278697911890625e-07,
+      "loss": 0.442145437002182,
+      "step": 3736
+    },
+    {
+      "epoch": 0.9528301886792453,
+      "grad_norm": 0.431334525346756,
+      "learning_rate": 1.2147291052798217e-07,
+      "loss": 0.4376698136329651,
+      "step": 3737
+    },
+    {
+      "epoch": 0.9530851606323304,
+      "grad_norm": 0.42173293232917786,
+      "learning_rate": 1.2016586844207368e-07,
+      "loss": 0.4280625283718109,
+      "step": 3738
+    },
+    {
+      "epoch": 0.9533401325854156,
+      "grad_norm": 0.4419265687465668,
+      "learning_rate": 1.1886585379086713e-07,
+      "loss": 0.4444487988948822,
+      "step": 3739
+    },
+    {
+      "epoch": 0.9535951045385007,
+      "grad_norm": 0.42979907989501953,
+      "learning_rate": 1.1757286749905172e-07,
+      "loss": 0.43843722343444824,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9538500764915859,
+      "grad_norm": 0.43650469183921814,
+      "learning_rate": 1.1628691048631735e-07,
+      "loss": 0.41674113273620605,
+      "step": 3741
+    },
+    {
+      "epoch": 0.954105048444671,
+      "grad_norm": 0.5075280666351318,
+      "learning_rate": 1.1500798366735234e-07,
+      "loss": 0.4313585162162781,
+      "step": 3742
+    },
+    {
+      "epoch": 0.9543600203977562,
+      "grad_norm": 0.45171964168548584,
+      "learning_rate": 1.1373608795184566e-07,
+      "loss": 0.43857893347740173,
+      "step": 3743
+    },
+    {
+      "epoch": 0.9546149923508415,
+      "grad_norm": 0.4379873275756836,
+      "learning_rate": 1.1247122424448586e-07,
+      "loss": 0.4406069219112396,
+      "step": 3744
+    },
+    {
+      "epoch": 0.9548699643039266,
+      "grad_norm": 0.4228276312351227,
+      "learning_rate": 1.1121339344495663e-07,
+      "loss": 0.43608206510543823,
+      "step": 3745
+    },
+    {
+      "epoch": 0.9551249362570118,
+      "grad_norm": 0.4296625554561615,
+      "learning_rate": 1.0996259644794449e-07,
+      "loss": 0.433504581451416,
+      "step": 3746
+    },
+    {
+      "epoch": 0.9553799082100969,
+      "grad_norm": 0.4238683879375458,
+      "learning_rate": 1.0871883414312778e-07,
+      "loss": 0.4223134517669678,
+      "step": 3747
+    },
+    {
+      "epoch": 0.9556348801631821,
+      "grad_norm": 0.5850834846496582,
+      "learning_rate": 1.0748210741518217e-07,
+      "loss": 0.43754324316978455,
+      "step": 3748
+    },
+    {
+      "epoch": 0.9558898521162672,
+      "grad_norm": 0.4604136049747467,
+      "learning_rate": 1.0625241714378287e-07,
+      "loss": 0.41852694749832153,
+      "step": 3749
+    },
+    {
+      "epoch": 0.9561448240693524,
+      "grad_norm": 0.42672300338745117,
+      "learning_rate": 1.0502976420359468e-07,
+      "loss": 0.4334542155265808,
+      "step": 3750
+    },
+    {
+      "epoch": 0.9563997960224375,
+      "grad_norm": 0.43828144669532776,
+      "learning_rate": 1.0381414946428081e-07,
+      "loss": 0.42628201842308044,
+      "step": 3751
+    },
+    {
+      "epoch": 0.9566547679755227,
+      "grad_norm": 0.4435332119464874,
+      "learning_rate": 1.0260557379049518e-07,
+      "loss": 0.42938369512557983,
+      "step": 3752
+    },
+    {
+      "epoch": 0.9569097399286078,
+      "grad_norm": 0.4323669373989105,
+      "learning_rate": 1.0140403804188903e-07,
+      "loss": 0.43224120140075684,
+      "step": 3753
+    },
+    {
+      "epoch": 0.957164711881693,
+      "grad_norm": 0.46962302923202515,
+      "learning_rate": 1.0020954307310205e-07,
+      "loss": 0.44025421142578125,
+      "step": 3754
+    },
+    {
+      "epoch": 0.9574196838347782,
+      "grad_norm": 0.43256840109825134,
+      "learning_rate": 9.902208973376682e-08,
+      "loss": 0.43730872869491577,
+      "step": 3755
+    },
+    {
+      "epoch": 0.9576746557878634,
+      "grad_norm": 0.4514023959636688,
+      "learning_rate": 9.784167886850881e-08,
+      "loss": 0.442075252532959,
+      "step": 3756
+    },
+    {
+      "epoch": 0.9579296277409485,
+      "grad_norm": 0.4285776615142822,
+      "learning_rate": 9.66683113169431e-08,
+      "loss": 0.43864166736602783,
+      "step": 3757
+    },
+    {
+      "epoch": 0.9581845996940337,
+      "grad_norm": 0.4442579746246338,
+      "learning_rate": 9.550198791367538e-08,
+      "loss": 0.440030574798584,
+      "step": 3758
+    },
+    {
+      "epoch": 0.9584395716471188,
+      "grad_norm": 0.43228745460510254,
+      "learning_rate": 9.434270948830093e-08,
+      "loss": 0.44247886538505554,
+      "step": 3759
+    },
+    {
+      "epoch": 0.958694543600204,
+      "grad_norm": 0.42962801456451416,
+      "learning_rate": 9.31904768654035e-08,
+      "loss": 0.4262453019618988,
+      "step": 3760
+    },
+    {
+      "epoch": 0.9589495155532891,
+      "grad_norm": 0.43366146087646484,
+      "learning_rate": 9.204529086455527e-08,
+      "loss": 0.433780312538147,
+      "step": 3761
+    },
+    {
+      "epoch": 0.9592044875063743,
+      "grad_norm": 0.4430442750453949,
+      "learning_rate": 9.090715230031688e-08,
+      "loss": 0.4226398169994354,
+      "step": 3762
+    },
+    {
+      "epoch": 0.9594594594594594,
+      "grad_norm": 0.4477764964103699,
+      "learning_rate": 8.977606198223521e-08,
+      "loss": 0.4350840747356415,
+      "step": 3763
+    },
+    {
+      "epoch": 0.9597144314125446,
+      "grad_norm": 0.4348631203174591,
+      "learning_rate": 8.865202071484558e-08,
+      "loss": 0.43202561140060425,
+      "step": 3764
+    },
+    {
+      "epoch": 0.9599694033656297,
+      "grad_norm": 0.44474688172340393,
+      "learning_rate": 8.753502929766622e-08,
+      "loss": 0.421417236328125,
+      "step": 3765
+    },
+    {
+      "epoch": 0.960224375318715,
+      "grad_norm": 0.43490901589393616,
+      "learning_rate": 8.642508852520492e-08,
+      "loss": 0.42618608474731445,
+      "step": 3766
+    },
+    {
+      "epoch": 0.9604793472718001,
+      "grad_norm": 0.4435284435749054,
+      "learning_rate": 8.532219918695128e-08,
+      "loss": 0.42826855182647705,
+      "step": 3767
+    },
+    {
+      "epoch": 0.9607343192248853,
+      "grad_norm": 0.4380919635295868,
+      "learning_rate": 8.422636206737888e-08,
+      "loss": 0.4380806088447571,
+      "step": 3768
+    },
+    {
+      "epoch": 0.9609892911779704,
+      "grad_norm": 0.4442375898361206,
+      "learning_rate": 8.313757794594867e-08,
+      "loss": 0.44663724303245544,
+      "step": 3769
+    },
+    {
+      "epoch": 0.9612442631310556,
+      "grad_norm": 0.42946797609329224,
+      "learning_rate": 8.205584759710228e-08,
+      "loss": 0.42948731780052185,
+      "step": 3770
+    },
+    {
+      "epoch": 0.9614992350841407,
+      "grad_norm": 0.45053452253341675,
+      "learning_rate": 8.09811717902631e-08,
+      "loss": 0.4317397475242615,
+      "step": 3771
+    },
+    {
+      "epoch": 0.9617542070372259,
+      "grad_norm": 0.4437471330165863,
+      "learning_rate": 7.99135512898408e-08,
+      "loss": 0.4413641393184662,
+      "step": 3772
+    },
+    {
+      "epoch": 0.962009178990311,
+      "grad_norm": 0.44477081298828125,
+      "learning_rate": 7.885298685522235e-08,
+      "loss": 0.44333338737487793,
+      "step": 3773
+    },
+    {
+      "epoch": 0.9622641509433962,
+      "grad_norm": 0.4432341456413269,
+      "learning_rate": 7.779947924077658e-08,
+      "loss": 0.4412536025047302,
+      "step": 3774
+    },
+    {
+      "epoch": 0.9625191228964813,
+      "grad_norm": 0.4365348517894745,
+      "learning_rate": 7.675302919585514e-08,
+      "loss": 0.43959423899650574,
+      "step": 3775
+    },
+    {
+      "epoch": 0.9627740948495666,
+      "grad_norm": 0.43227964639663696,
+      "learning_rate": 7.57136374647871e-08,
+      "loss": 0.4292386472225189,
+      "step": 3776
+    },
+    {
+      "epoch": 0.9630290668026517,
+      "grad_norm": 0.4467385411262512,
+      "learning_rate": 7.468130478688218e-08,
+      "loss": 0.44225990772247314,
+      "step": 3777
+    },
+    {
+      "epoch": 0.9632840387557369,
+      "grad_norm": 0.4632147252559662,
+      "learning_rate": 7.365603189643078e-08,
+      "loss": 0.4331645369529724,
+      "step": 3778
+    },
+    {
+      "epoch": 0.963539010708822,
+      "grad_norm": 0.4216609001159668,
+      "learning_rate": 7.263781952269954e-08,
+      "loss": 0.43343281745910645,
+      "step": 3779
+    },
+    {
+      "epoch": 0.9637939826619072,
+      "grad_norm": 0.45064857602119446,
+      "learning_rate": 7.162666838993249e-08,
+      "loss": 0.4255315065383911,
+      "step": 3780
+    },
+    {
+      "epoch": 0.9640489546149924,
+      "grad_norm": 0.4347366392612457,
+      "learning_rate": 7.062257921735205e-08,
+      "loss": 0.4380302429199219,
+      "step": 3781
+    },
+    {
+      "epoch": 0.9643039265680775,
+      "grad_norm": 0.4600411057472229,
+      "learning_rate": 6.962555271915806e-08,
+      "loss": 0.4289246201515198,
+      "step": 3782
+    },
+    {
+      "epoch": 0.9645588985211627,
+      "grad_norm": 0.44208458065986633,
+      "learning_rate": 6.863558960452654e-08,
+      "loss": 0.4249235987663269,
+      "step": 3783
+    },
+    {
+      "epoch": 0.9648138704742478,
+      "grad_norm": 0.44760599732398987,
+      "learning_rate": 6.76526905776087e-08,
+      "loss": 0.4318837821483612,
+      "step": 3784
+    },
+    {
+      "epoch": 0.965068842427333,
+      "grad_norm": 0.43746480345726013,
+      "learning_rate": 6.667685633753196e-08,
+      "loss": 0.4212871491909027,
+      "step": 3785
+    },
+    {
+      "epoch": 0.9653238143804181,
+      "grad_norm": 0.4536183178424835,
+      "learning_rate": 6.570808757839775e-08,
+      "loss": 0.42356353998184204,
+      "step": 3786
+    },
+    {
+      "epoch": 0.9655787863335034,
+      "grad_norm": 0.42770835757255554,
+      "learning_rate": 6.474638498928265e-08,
+      "loss": 0.4301995038986206,
+      "step": 3787
+    },
+    {
+      "epoch": 0.9658337582865885,
+      "grad_norm": 0.4357539713382721,
+      "learning_rate": 6.379174925423614e-08,
+      "loss": 0.4319273829460144,
+      "step": 3788
+    },
+    {
+      "epoch": 0.9660887302396737,
+      "grad_norm": 0.46971604228019714,
+      "learning_rate": 6.284418105228284e-08,
+      "loss": 0.4219222068786621,
+      "step": 3789
+    },
+    {
+      "epoch": 0.9663437021927588,
+      "grad_norm": 0.45036160945892334,
+      "learning_rate": 6.190368105741806e-08,
+      "loss": 0.41687309741973877,
+      "step": 3790
+    },
+    {
+      "epoch": 0.966598674145844,
+      "grad_norm": 0.4009671211242676,
+      "learning_rate": 6.097024993861112e-08,
+      "loss": 0.43720635771751404,
+      "step": 3791
+    },
+    {
+      "epoch": 0.9668536460989291,
+      "grad_norm": 0.42961275577545166,
+      "learning_rate": 6.004388835980424e-08,
+      "loss": 0.4499396085739136,
+      "step": 3792
+    },
+    {
+      "epoch": 0.9671086180520143,
+      "grad_norm": 0.45251771807670593,
+      "learning_rate": 5.9124596979908136e-08,
+      "loss": 0.45437192916870117,
+      "step": 3793
+    },
+    {
+      "epoch": 0.9673635900050994,
+      "grad_norm": 0.4467000365257263,
+      "learning_rate": 5.821237645280753e-08,
+      "loss": 0.43350905179977417,
+      "step": 3794
+    },
+    {
+      "epoch": 0.9676185619581846,
+      "grad_norm": 0.44402676820755005,
+      "learning_rate": 5.730722742735562e-08,
+      "loss": 0.42394256591796875,
+      "step": 3795
+    },
+    {
+      "epoch": 0.9678735339112697,
+      "grad_norm": 0.42971062660217285,
+      "learning_rate": 5.640915054737739e-08,
+      "loss": 0.43170222640037537,
+      "step": 3796
+    },
+    {
+      "epoch": 0.968128505864355,
+      "grad_norm": 0.46525314450263977,
+      "learning_rate": 5.55181464516652e-08,
+      "loss": 0.4343894422054291,
+      "step": 3797
+    },
+    {
+      "epoch": 0.96838347781744,
+      "grad_norm": 0.425347238779068,
+      "learning_rate": 5.46342157739832e-08,
+      "loss": 0.4369485378265381,
+      "step": 3798
+    },
+    {
+      "epoch": 0.9686384497705253,
+      "grad_norm": 0.447989284992218,
+      "learning_rate": 5.375735914306291e-08,
+      "loss": 0.4052256941795349,
+      "step": 3799
+    },
+    {
+      "epoch": 0.9688934217236104,
+      "grad_norm": 0.4680423438549042,
+      "learning_rate": 5.288757718260429e-08,
+      "loss": 0.4462357759475708,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9691483936766956,
+      "grad_norm": 0.4516068696975708,
+      "learning_rate": 5.2024870511274694e-08,
+      "loss": 0.43032628297805786,
+      "step": 3801
+    },
+    {
+      "epoch": 0.9694033656297807,
+      "grad_norm": 0.4310958683490753,
+      "learning_rate": 5.116923974270993e-08,
+      "loss": 0.41478651762008667,
+      "step": 3802
+    },
+    {
+      "epoch": 0.9696583375828659,
+      "grad_norm": 0.41717851161956787,
+      "learning_rate": 5.032068548551205e-08,
+      "loss": 0.42103320360183716,
+      "step": 3803
+    },
+    {
+      "epoch": 0.969913309535951,
+      "grad_norm": 0.44686272740364075,
+      "learning_rate": 4.947920834325048e-08,
+      "loss": 0.42334866523742676,
+      "step": 3804
+    },
+    {
+      "epoch": 0.9701682814890362,
+      "grad_norm": 0.43319904804229736,
+      "learning_rate": 4.8644808914459776e-08,
+      "loss": 0.44510579109191895,
+      "step": 3805
+    },
+    {
+      "epoch": 0.9704232534421213,
+      "grad_norm": 0.42840850353240967,
+      "learning_rate": 4.781748779263962e-08,
+      "loss": 0.42885157465934753,
+      "step": 3806
+    },
+    {
+      "epoch": 0.9706782253952065,
+      "grad_norm": 0.41729164123535156,
+      "learning_rate": 4.6997245566257066e-08,
+      "loss": 0.43627625703811646,
+      "step": 3807
+    },
+    {
+      "epoch": 0.9709331973482916,
+      "grad_norm": 0.5475233197212219,
+      "learning_rate": 4.618408281874209e-08,
+      "loss": 0.4385756254196167,
+      "step": 3808
+    },
+    {
+      "epoch": 0.9711881693013769,
+      "grad_norm": 0.4557948708534241,
+      "learning_rate": 4.537800012849092e-08,
+      "loss": 0.42454612255096436,
+      "step": 3809
+    },
+    {
+      "epoch": 0.971443141254462,
+      "grad_norm": 0.42507830262184143,
+      "learning_rate": 4.4578998068861566e-08,
+      "loss": 0.4411567151546478,
+      "step": 3810
+    },
+    {
+      "epoch": 0.9716981132075472,
+      "grad_norm": 0.4329739511013031,
+      "learning_rate": 4.378707720817721e-08,
+      "loss": 0.44114264845848083,
+      "step": 3811
+    },
+    {
+      "epoch": 0.9719530851606323,
+      "grad_norm": 0.4257124960422516,
+      "learning_rate": 4.3002238109723927e-08,
+      "loss": 0.43023261427879333,
+      "step": 3812
+    },
+    {
+      "epoch": 0.9722080571137175,
+      "grad_norm": 0.4460616111755371,
+      "learning_rate": 4.2224481331750723e-08,
+      "loss": 0.4382517337799072,
+      "step": 3813
+    },
+    {
+      "epoch": 0.9724630290668026,
+      "grad_norm": 0.4248466491699219,
+      "learning_rate": 4.1453807427467297e-08,
+      "loss": 0.4303026795387268,
+      "step": 3814
+    },
+    {
+      "epoch": 0.9727180010198878,
+      "grad_norm": 0.42271357774734497,
+      "learning_rate": 4.06902169450496e-08,
+      "loss": 0.42649969458580017,
+      "step": 3815
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 0.42175567150115967,
+      "learning_rate": 3.9933710427629837e-08,
+      "loss": 0.44406718015670776,
+      "step": 3816
+    },
+    {
+      "epoch": 0.9732279449260581,
+      "grad_norm": 0.439832866191864,
+      "learning_rate": 3.918428841330646e-08,
+      "loss": 0.4441678524017334,
+      "step": 3817
+    },
+    {
+      "epoch": 0.9734829168791433,
+      "grad_norm": 0.4424952566623688,
+      "learning_rate": 3.844195143513418e-08,
+      "loss": 0.4407542645931244,
+      "step": 3818
+    },
+    {
+      "epoch": 0.9737378888322284,
+      "grad_norm": 0.4503242075443268,
+      "learning_rate": 3.770670002113175e-08,
+      "loss": 0.44105467200279236,
+      "step": 3819
+    },
+    {
+      "epoch": 0.9739928607853137,
+      "grad_norm": 0.43343716859817505,
+      "learning_rate": 3.697853469427637e-08,
+      "loss": 0.4408354163169861,
+      "step": 3820
+    },
+    {
+      "epoch": 0.9742478327383988,
+      "grad_norm": 0.44179418683052063,
+      "learning_rate": 3.6257455972505964e-08,
+      "loss": 0.42275065183639526,
+      "step": 3821
+    },
+    {
+      "epoch": 0.974502804691484,
+      "grad_norm": 0.4276719391345978,
+      "learning_rate": 3.554346436871581e-08,
+      "loss": 0.44853538274765015,
+      "step": 3822
+    },
+    {
+      "epoch": 0.9747577766445691,
+      "grad_norm": 0.4315888285636902,
+      "learning_rate": 3.4836560390762995e-08,
+      "loss": 0.43066248297691345,
+      "step": 3823
+    },
+    {
+      "epoch": 0.9750127485976543,
+      "grad_norm": 0.4345954358577728,
+      "learning_rate": 3.4136744541459764e-08,
+      "loss": 0.4355141818523407,
+      "step": 3824
+    },
+    {
+      "epoch": 0.9752677205507394,
+      "grad_norm": 0.43663638830184937,
+      "learning_rate": 3.344401731858127e-08,
+      "loss": 0.42832696437835693,
+      "step": 3825
+    },
+    {
+      "epoch": 0.9755226925038246,
+      "grad_norm": 0.4864625632762909,
+      "learning_rate": 3.2758379214855584e-08,
+      "loss": 0.436530202627182,
+      "step": 3826
+    },
+    {
+      "epoch": 0.9757776644569097,
+      "grad_norm": 0.42900747060775757,
+      "learning_rate": 3.207983071797261e-08,
+      "loss": 0.44682687520980835,
+      "step": 3827
+    },
+    {
+      "epoch": 0.9760326364099949,
+      "grad_norm": 0.44420379400253296,
+      "learning_rate": 3.140837231057625e-08,
+      "loss": 0.41383880376815796,
+      "step": 3828
+    },
+    {
+      "epoch": 0.97628760836308,
+      "grad_norm": 0.44567447900772095,
+      "learning_rate": 3.074400447027004e-08,
+      "loss": 0.4374455213546753,
+      "step": 3829
+    },
+    {
+      "epoch": 0.9765425803161653,
+      "grad_norm": 0.4449840486049652,
+      "learning_rate": 3.008672766961151e-08,
+      "loss": 0.42075401544570923,
+      "step": 3830
+    },
+    {
+      "epoch": 0.9767975522692504,
+      "grad_norm": 0.44207072257995605,
+      "learning_rate": 2.9436542376116706e-08,
+      "loss": 0.42974042892456055,
+      "step": 3831
+    },
+    {
+      "epoch": 0.9770525242223356,
+      "grad_norm": 0.41998976469039917,
+      "learning_rate": 2.8793449052254564e-08,
+      "loss": 0.4292892813682556,
+      "step": 3832
+    },
+    {
+      "epoch": 0.9773074961754207,
+      "grad_norm": 0.45919719338417053,
+      "learning_rate": 2.8157448155452515e-08,
+      "loss": 0.43220004439353943,
+      "step": 3833
+    },
+    {
+      "epoch": 0.9775624681285059,
+      "grad_norm": 0.5310207605361938,
+      "learning_rate": 2.752854013809314e-08,
+      "loss": 0.4327763020992279,
+      "step": 3834
+    },
+    {
+      "epoch": 0.977817440081591,
+      "grad_norm": 0.4287780523300171,
+      "learning_rate": 2.6906725447510828e-08,
+      "loss": 0.4327196478843689,
+      "step": 3835
+    },
+    {
+      "epoch": 0.9780724120346762,
+      "grad_norm": 0.4250393509864807,
+      "learning_rate": 2.629200452599845e-08,
+      "loss": 0.4387291669845581,
+      "step": 3836
+    },
+    {
+      "epoch": 0.9783273839877613,
+      "grad_norm": 0.41067278385162354,
+      "learning_rate": 2.5684377810800687e-08,
+      "loss": 0.44172751903533936,
+      "step": 3837
+    },
+    {
+      "epoch": 0.9785823559408465,
+      "grad_norm": 0.46978065371513367,
+      "learning_rate": 2.5083845734116264e-08,
+      "loss": 0.4297478199005127,
+      "step": 3838
+    },
+    {
+      "epoch": 0.9788373278939316,
+      "grad_norm": 0.43518373370170593,
+      "learning_rate": 2.4490408723097936e-08,
+      "loss": 0.43526455760002136,
+      "step": 3839
+    },
+    {
+      "epoch": 0.9790922998470168,
+      "grad_norm": 0.4430595338344574,
+      "learning_rate": 2.3904067199853607e-08,
+      "loss": 0.44849714636802673,
+      "step": 3840
+    },
+    {
+      "epoch": 0.979347271800102,
+      "grad_norm": 0.5210933089256287,
+      "learning_rate": 2.332482158144078e-08,
+      "loss": 0.4486989974975586,
+      "step": 3841
+    },
+    {
+      "epoch": 0.9796022437531872,
+      "grad_norm": 0.45702749490737915,
+      "learning_rate": 2.2752672279873212e-08,
+      "loss": 0.4192723035812378,
+      "step": 3842
+    },
+    {
+      "epoch": 0.9798572157062723,
+      "grad_norm": 0.4321991205215454,
+      "learning_rate": 2.2187619702113138e-08,
+      "loss": 0.4389612078666687,
+      "step": 3843
+    },
+    {
+      "epoch": 0.9801121876593575,
+      "grad_norm": 0.435012549161911,
+      "learning_rate": 2.1629664250080172e-08,
+      "loss": 0.43103253841400146,
+      "step": 3844
+    },
+    {
+      "epoch": 0.9803671596124426,
+      "grad_norm": 0.4384916424751282,
+      "learning_rate": 2.1078806320640187e-08,
+      "loss": 0.42653101682662964,
+      "step": 3845
+    },
+    {
+      "epoch": 0.9806221315655278,
+      "grad_norm": 0.42758908867836,
+      "learning_rate": 2.0535046305614202e-08,
+      "loss": 0.42084115743637085,
+      "step": 3846
+    },
+    {
+      "epoch": 0.9808771035186129,
+      "grad_norm": 0.4485051929950714,
+      "learning_rate": 1.9998384591773945e-08,
+      "loss": 0.4244152009487152,
+      "step": 3847
+    },
+    {
+      "epoch": 0.9811320754716981,
+      "grad_norm": 0.43234965205192566,
+      "learning_rate": 1.9468821560841845e-08,
+      "loss": 0.43603652715682983,
+      "step": 3848
+    },
+    {
+      "epoch": 0.9813870474247832,
+      "grad_norm": 0.4192225933074951,
+      "learning_rate": 1.8946357589492147e-08,
+      "loss": 0.42890340089797974,
+      "step": 3849
+    },
+    {
+      "epoch": 0.9816420193778684,
+      "grad_norm": 0.4449460804462433,
+      "learning_rate": 1.843099304934759e-08,
+      "loss": 0.4340965151786804,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9818969913309537,
+      "grad_norm": 0.46211758255958557,
+      "learning_rate": 1.792272830698161e-08,
+      "loss": 0.43783053755760193,
+      "step": 3851
+    },
+    {
+      "epoch": 0.9821519632840388,
+      "grad_norm": 0.41678836941719055,
+      "learning_rate": 1.7421563723919453e-08,
+      "loss": 0.43343478441238403,
+      "step": 3852
+    },
+    {
+      "epoch": 0.982406935237124,
+      "grad_norm": 0.4770534336566925,
+      "learning_rate": 1.692749965663376e-08,
+      "loss": 0.42233383655548096,
+      "step": 3853
+    },
+    {
+      "epoch": 0.9826619071902091,
+      "grad_norm": 0.44247758388519287,
+      "learning_rate": 1.6440536456547863e-08,
+      "loss": 0.4290817975997925,
+      "step": 3854
+    },
+    {
+      "epoch": 0.9829168791432943,
+      "grad_norm": 0.4584869146347046,
+      "learning_rate": 1.5960674470035796e-08,
+      "loss": 0.43462952971458435,
+      "step": 3855
+    },
+    {
+      "epoch": 0.9831718510963794,
+      "grad_norm": 0.42376816272735596,
+      "learning_rate": 1.5487914038417873e-08,
+      "loss": 0.4478336274623871,
+      "step": 3856
+    },
+    {
+      "epoch": 0.9834268230494646,
+      "grad_norm": 0.5084519386291504,
+      "learning_rate": 1.502225549796288e-08,
+      "loss": 0.4331415593624115,
+      "step": 3857
+    },
+    {
+      "epoch": 0.9836817950025497,
+      "grad_norm": 0.4290451109409332,
+      "learning_rate": 1.4563699179892532e-08,
+      "loss": 0.4360179305076599,
+      "step": 3858
+    },
+    {
+      "epoch": 0.9839367669556349,
+      "grad_norm": 0.44008415937423706,
+      "learning_rate": 1.4112245410370373e-08,
+      "loss": 0.4326023459434509,
+      "step": 3859
+    },
+    {
+      "epoch": 0.98419173890872,
+      "grad_norm": 0.4248690605163574,
+      "learning_rate": 1.366789451051287e-08,
+      "loss": 0.4331281781196594,
+      "step": 3860
+    },
+    {
+      "epoch": 0.9844467108618052,
+      "grad_norm": 0.43502941727638245,
+      "learning_rate": 1.3230646796383862e-08,
+      "loss": 0.4257779121398926,
+      "step": 3861
+    },
+    {
+      "epoch": 0.9847016828148903,
+      "grad_norm": 0.4281660318374634,
+      "learning_rate": 1.2800502578991236e-08,
+      "loss": 0.43861016631126404,
+      "step": 3862
+    },
+    {
+      "epoch": 0.9849566547679756,
+      "grad_norm": 0.4268348813056946,
+      "learning_rate": 1.2377462164294695e-08,
+      "loss": 0.4324760138988495,
+      "step": 3863
+    },
+    {
+      "epoch": 0.9852116267210607,
+      "grad_norm": 0.42239123582839966,
+      "learning_rate": 1.1961525853196876e-08,
+      "loss": 0.4312841296195984,
+      "step": 3864
+    },
+    {
+      "epoch": 0.9854665986741459,
+      "grad_norm": 0.43889138102531433,
+      "learning_rate": 1.1552693941551118e-08,
+      "loss": 0.4337885081768036,
+      "step": 3865
+    },
+    {
+      "epoch": 0.985721570627231,
+      "grad_norm": 0.43608731031417847,
+      "learning_rate": 1.115096672015481e-08,
+      "loss": 0.4339744448661804,
+      "step": 3866
+    },
+    {
+      "epoch": 0.9859765425803162,
+      "grad_norm": 0.41787976026535034,
+      "learning_rate": 1.0756344474753822e-08,
+      "loss": 0.4270380735397339,
+      "step": 3867
+    },
+    {
+      "epoch": 0.9862315145334013,
+      "grad_norm": 0.4472630023956299,
+      "learning_rate": 1.036882748603918e-08,
+      "loss": 0.43142664432525635,
+      "step": 3868
+    },
+    {
+      "epoch": 0.9864864864864865,
+      "grad_norm": 0.42906099557876587,
+      "learning_rate": 9.988416029647063e-09,
+      "loss": 0.43397778272628784,
+      "step": 3869
+    },
+    {
+      "epoch": 0.9867414584395716,
+      "grad_norm": 0.4523125886917114,
+      "learning_rate": 9.61511037616103e-09,
+      "loss": 0.4238213896751404,
+      "step": 3870
+    },
+    {
+      "epoch": 0.9869964303926568,
+      "grad_norm": 0.4824058711528778,
+      "learning_rate": 9.248910791109789e-09,
+      "loss": 0.43059441447257996,
+      "step": 3871
+    },
+    {
+      "epoch": 0.9872514023457419,
+      "grad_norm": 0.4468144178390503,
+      "learning_rate": 8.889817534969425e-09,
+      "loss": 0.42374369502067566,
+      "step": 3872
+    },
+    {
+      "epoch": 0.9875063742988271,
+      "grad_norm": 0.4466455578804016,
+      "learning_rate": 8.537830863157848e-09,
+      "loss": 0.43019899725914,
+      "step": 3873
+    },
+    {
+      "epoch": 0.9877613462519123,
+      "grad_norm": 0.4248717427253723,
+      "learning_rate": 8.19295102603923e-09,
+      "loss": 0.4276416301727295,
+      "step": 3874
+    },
+    {
+      "epoch": 0.9880163182049975,
+      "grad_norm": 0.4355369508266449,
+      "learning_rate": 7.855178268926233e-09,
+      "loss": 0.44368404150009155,
+      "step": 3875
+    },
+    {
+      "epoch": 0.9882712901580826,
+      "grad_norm": 0.4394513964653015,
+      "learning_rate": 7.524512832071117e-09,
+      "loss": 0.42616432905197144,
+      "step": 3876
+    },
+    {
+      "epoch": 0.9885262621111678,
+      "grad_norm": 0.45191922783851624,
+      "learning_rate": 7.200954950673522e-09,
+      "loss": 0.4272156357765198,
+      "step": 3877
+    },
+    {
+      "epoch": 0.9887812340642529,
+      "grad_norm": 0.42931079864501953,
+      "learning_rate": 6.8845048548782375e-09,
+      "loss": 0.42737168073654175,
+      "step": 3878
+    },
+    {
+      "epoch": 0.9890362060173381,
+      "grad_norm": 0.46329110860824585,
+      "learning_rate": 6.575162769771881e-09,
+      "loss": 0.4369872510433197,
+      "step": 3879
+    },
+    {
+      "epoch": 0.9892911779704232,
+      "grad_norm": 0.42689400911331177,
+      "learning_rate": 6.272928915387333e-09,
+      "loss": 0.44038841128349304,
+      "step": 3880
+    },
+    {
+      "epoch": 0.9895461499235084,
+      "grad_norm": 0.4330800771713257,
+      "learning_rate": 5.977803506701518e-09,
+      "loss": 0.42824047803878784,
+      "step": 3881
+    },
+    {
+      "epoch": 0.9898011218765935,
+      "grad_norm": 0.4149211645126343,
+      "learning_rate": 5.689786753633186e-09,
+      "loss": 0.4450784921646118,
+      "step": 3882
+    },
+    {
+      "epoch": 0.9900560938296787,
+      "grad_norm": 0.4341184198856354,
+      "learning_rate": 5.40887886104624e-09,
+      "loss": 0.4351019859313965,
+      "step": 3883
+    },
+    {
+      "epoch": 0.9903110657827638,
+      "grad_norm": 0.46087369322776794,
+      "learning_rate": 5.135080028748629e-09,
+      "loss": 0.4396669566631317,
+      "step": 3884
+    },
+    {
+      "epoch": 0.9905660377358491,
+      "grad_norm": 0.4476439654827118,
+      "learning_rate": 4.868390451490123e-09,
+      "loss": 0.4363985061645508,
+      "step": 3885
+    },
+    {
+      "epoch": 0.9908210096889342,
+      "grad_norm": 0.44064781069755554,
+      "learning_rate": 4.608810318964541e-09,
+      "loss": 0.42739638686180115,
+      "step": 3886
+    },
+    {
+      "epoch": 0.9910759816420194,
+      "grad_norm": 0.4554571211338043,
+      "learning_rate": 4.35633981580974e-09,
+      "loss": 0.4383185803890228,
+      "step": 3887
+    },
+    {
+      "epoch": 0.9913309535951046,
+      "grad_norm": 0.4218502640724182,
+      "learning_rate": 4.110979121604297e-09,
+      "loss": 0.4426953196525574,
+      "step": 3888
+    },
+    {
+      "epoch": 0.9915859255481897,
+      "grad_norm": 0.4349331557750702,
+      "learning_rate": 3.87272841087194e-09,
+      "loss": 0.4466049075126648,
+      "step": 3889
+    },
+    {
+      "epoch": 0.9918408975012749,
+      "grad_norm": 0.42966899275779724,
+      "learning_rate": 3.641587853077111e-09,
+      "loss": 0.43008658289909363,
+      "step": 3890
+    },
+    {
+      "epoch": 0.99209586945436,
+      "grad_norm": 0.4374047815799713,
+      "learning_rate": 3.4175576126305177e-09,
+      "loss": 0.4254233241081238,
+      "step": 3891
+    },
+    {
+      "epoch": 0.9923508414074452,
+      "grad_norm": 0.4157470464706421,
+      "learning_rate": 3.20063784888025e-09,
+      "loss": 0.4410827159881592,
+      "step": 3892
+    },
+    {
+      "epoch": 0.9926058133605303,
+      "grad_norm": 0.42432066798210144,
+      "learning_rate": 2.9908287161195537e-09,
+      "loss": 0.4353354871273041,
+      "step": 3893
+    },
+    {
+      "epoch": 0.9928607853136155,
+      "grad_norm": 0.5254224538803101,
+      "learning_rate": 2.788130363584607e-09,
+      "loss": 0.42760753631591797,
+      "step": 3894
+    },
+    {
+      "epoch": 0.9931157572667006,
+      "grad_norm": 0.4437570869922638,
+      "learning_rate": 2.5925429354534126e-09,
+      "loss": 0.4352593421936035,
+      "step": 3895
+    },
+    {
+      "epoch": 0.9933707292197859,
+      "grad_norm": 0.45760712027549744,
+      "learning_rate": 2.404066570843577e-09,
+      "loss": 0.4349639415740967,
+      "step": 3896
+    },
+    {
+      "epoch": 0.993625701172871,
+      "grad_norm": 0.4340839684009552,
+      "learning_rate": 2.222701403818972e-09,
+      "loss": 0.4380471110343933,
+      "step": 3897
+    },
+    {
+      "epoch": 0.9938806731259562,
+      "grad_norm": 0.4430701434612274,
+      "learning_rate": 2.048447563380851e-09,
+      "loss": 0.43017488718032837,
+      "step": 3898
+    },
+    {
+      "epoch": 0.9941356450790413,
+      "grad_norm": 0.43190470337867737,
+      "learning_rate": 1.881305173475623e-09,
+      "loss": 0.41849285364151,
+      "step": 3899
+    },
+    {
+      "epoch": 0.9943906170321265,
+      "grad_norm": 0.4917444884777069,
+      "learning_rate": 1.7212743529892996e-09,
+      "loss": 0.4396299123764038,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9946455889852116,
+      "grad_norm": 0.4348951280117035,
+      "learning_rate": 1.5683552157508275e-09,
+      "loss": 0.4330969452857971,
+      "step": 3901
+    },
+    {
+      "epoch": 0.9949005609382968,
+      "grad_norm": 0.43384942412376404,
+      "learning_rate": 1.422547870530977e-09,
+      "loss": 0.4398225247859955,
+      "step": 3902
+    },
+    {
+      "epoch": 0.9951555328913819,
+      "grad_norm": 0.4424363374710083,
+      "learning_rate": 1.2838524210390113e-09,
+      "loss": 0.42855989933013916,
+      "step": 3903
+    },
+    {
+      "epoch": 0.9954105048444671,
+      "grad_norm": 0.4463779926300049,
+      "learning_rate": 1.1522689659293484e-09,
+      "loss": 0.4319448471069336,
+      "step": 3904
+    },
+    {
+      "epoch": 0.9956654767975522,
+      "grad_norm": 0.4403187036514282,
+      "learning_rate": 1.0277975987960098e-09,
+      "loss": 0.43797823786735535,
+      "step": 3905
+    },
+    {
+      "epoch": 0.9959204487506375,
+      "grad_norm": 0.43161749839782715,
+      "learning_rate": 9.104384081737305e-10,
+      "loss": 0.43103861808776855,
+      "step": 3906
+    },
+    {
+      "epoch": 0.9961754207037226,
+      "grad_norm": 0.44430285692214966,
+      "learning_rate": 8.001914775401798e-10,
+      "loss": 0.44083261489868164,
+      "step": 3907
+    },
+    {
+      "epoch": 0.9964303926568078,
+      "grad_norm": 0.4711357355117798,
+      "learning_rate": 6.970568853115201e-10,
+      "loss": 0.43897342681884766,
+      "step": 3908
+    },
+    {
+      "epoch": 0.9966853646098929,
+      "grad_norm": 0.4367968440055847,
+      "learning_rate": 6.010347048468479e-10,
+      "loss": 0.4301096200942993,
+      "step": 3909
+    },
+    {
+      "epoch": 0.9969403365629781,
+      "grad_norm": 0.4353834092617035,
+      "learning_rate": 5.121250044459736e-10,
+      "loss": 0.42462557554244995,
+      "step": 3910
+    },
+    {
+      "epoch": 0.9971953085160632,
+      "grad_norm": 0.4247950613498688,
+      "learning_rate": 4.3032784735053123e-10,
+      "loss": 0.42845088243484497,
+      "step": 3911
+    },
+    {
+      "epoch": 0.9974502804691484,
+      "grad_norm": 0.4372820258140564,
+      "learning_rate": 3.5564329174064826e-10,
+      "loss": 0.4321706295013428,
+      "step": 3912
+    },
+    {
+      "epoch": 0.9977052524222335,
+      "grad_norm": 0.46002548933029175,
+      "learning_rate": 2.8807139073827597e-10,
+      "loss": 0.43641120195388794,
+      "step": 3913
+    },
+    {
+      "epoch": 0.9979602243753187,
+      "grad_norm": 0.4415642321109772,
+      "learning_rate": 2.276121924094099e-10,
+      "loss": 0.4387682378292084,
+      "step": 3914
+    },
+    {
+      "epoch": 0.9982151963284038,
+      "grad_norm": 0.43121448159217834,
+      "learning_rate": 1.7426573975520834e-10,
+      "loss": 0.4379480481147766,
+      "step": 3915
+    },
+    {
+      "epoch": 0.998470168281489,
+      "grad_norm": 0.4421471059322357,
+      "learning_rate": 1.2803207072198398e-10,
+      "loss": 0.419895738363266,
+      "step": 3916
+    },
+    {
+      "epoch": 0.9987251402345741,
+      "grad_norm": 0.4430276155471802,
+      "learning_rate": 8.891121819565307e-11,
+      "loss": 0.43177294731140137,
+      "step": 3917
+    },
+    {
+      "epoch": 0.9989801121876594,
+      "grad_norm": 0.429384708404541,
+      "learning_rate": 5.690321000062504e-11,
+      "loss": 0.4321375787258148,
+      "step": 3918
+    },
+    {
+      "epoch": 0.9992350841407445,
+      "grad_norm": 0.41285955905914307,
+      "learning_rate": 3.200806890646391e-11,
+      "loss": 0.44204482436180115,
+      "step": 3919
+    },
+    {
+      "epoch": 0.9994900560938297,
+      "grad_norm": 0.42164909839630127,
+      "learning_rate": 1.4225812619006462e-11,
+      "loss": 0.4531802535057068,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9997450280469148,
+      "grad_norm": 0.4804461896419525,
+      "learning_rate": 3.556453787023628e-12,
+      "loss": 0.43864184617996216,
+      "step": 3921
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.46723073720932007,
+      "learning_rate": 0.0,
+      "loss": 0.42660099267959595,
+      "step": 3922
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3922,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.4774551385835726e+20,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}