diff --git "a/checkpoint-1000/trainer_state.json" "b/checkpoint-1000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1000/trainer_state.json"
@@ -0,0 +1,7033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.5005860748612956,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0025005860748612955,
+      "grad_norm": 0.7594180107116699,
+      "learning_rate": 4e-05,
+      "loss": 2.0425,
+      "step": 1
+    },
+    {
+      "epoch": 0.005001172149722591,
+      "grad_norm": 0.7701483964920044,
+      "learning_rate": 8e-05,
+      "loss": 2.0851,
+      "step": 2
+    },
+    {
+      "epoch": 0.007501758224583887,
+      "grad_norm": 0.7642767429351807,
+      "learning_rate": 0.00012,
+      "loss": 2.0201,
+      "step": 3
+    },
+    {
+      "epoch": 0.010002344299445182,
+      "grad_norm": 0.8546866774559021,
+      "learning_rate": 0.00016,
+      "loss": 1.9645,
+      "step": 4
+    },
+    {
+      "epoch": 0.012502930374306479,
+      "grad_norm": 0.8759480714797974,
+      "learning_rate": 0.0002,
+      "loss": 1.8278,
+      "step": 5
+    },
+    {
+      "epoch": 0.015003516449167774,
+      "grad_norm": 0.7554371953010559,
+      "learning_rate": 0.0001998322147651007,
+      "loss": 1.5279,
+      "step": 6
+    },
+    {
+      "epoch": 0.01750410252402907,
+      "grad_norm": 0.9531473517417908,
+      "learning_rate": 0.00019966442953020134,
+      "loss": 1.3597,
+      "step": 7
+    },
+    {
+      "epoch": 0.020004688598890364,
+      "grad_norm": 0.8676754236221313,
+      "learning_rate": 0.00019949664429530202,
+      "loss": 1.1228,
+      "step": 8
+    },
+    {
+      "epoch": 0.02250527467375166,
+      "grad_norm": 0.7342363595962524,
+      "learning_rate": 0.00019932885906040267,
+      "loss": 0.9392,
+      "step": 9
+    },
+    {
+      "epoch": 0.025005860748612958,
+      "grad_norm": 0.5210846066474915,
+      "learning_rate": 0.00019916107382550338,
+      "loss": 0.8306,
+      "step": 10
+    },
+    {
+      "epoch": 0.02750644682347425,
+      "grad_norm": 0.33346614241600037,
+      "learning_rate": 0.00019899328859060403,
+      "loss": 0.7801,
+      "step": 11
+    },
+    {
+      "epoch": 0.030007032898335548,
+      "grad_norm": 0.3581950068473816,
+      "learning_rate": 0.00019882550335570471,
+      "loss": 0.7862,
+      "step": 12
+    },
+    {
+      "epoch": 0.032507618973196845,
+      "grad_norm": 0.49463245272636414,
+      "learning_rate": 0.0001986577181208054,
+      "loss": 0.7424,
+      "step": 13
+    },
+    {
+      "epoch": 0.03500820504805814,
+      "grad_norm": 0.5055575370788574,
+      "learning_rate": 0.00019848993288590605,
+      "loss": 0.7309,
+      "step": 14
+    },
+    {
+      "epoch": 0.03750879112291943,
+      "grad_norm": 0.4792090058326721,
+      "learning_rate": 0.00019832214765100673,
+      "loss": 0.7158,
+      "step": 15
+    },
+    {
+      "epoch": 0.04000937719778073,
+      "grad_norm": 0.6231745481491089,
+      "learning_rate": 0.00019815436241610738,
+      "loss": 0.7298,
+      "step": 16
+    },
+    {
+      "epoch": 0.042509963272642025,
+      "grad_norm": 0.587597668170929,
+      "learning_rate": 0.00019798657718120806,
+      "loss": 0.6779,
+      "step": 17
+    },
+    {
+      "epoch": 0.04501054934750332,
+      "grad_norm": 0.6789360642433167,
+      "learning_rate": 0.0001978187919463087,
+      "loss": 0.6621,
+      "step": 18
+    },
+    {
+      "epoch": 0.04751113542236462,
+      "grad_norm": 0.7238547801971436,
+      "learning_rate": 0.00019765100671140942,
+      "loss": 0.6643,
+      "step": 19
+    },
+    {
+      "epoch": 0.050011721497225915,
+      "grad_norm": 0.44374287128448486,
+      "learning_rate": 0.00019748322147651007,
+      "loss": 0.6295,
+      "step": 20
+    },
+    {
+      "epoch": 0.052512307572087205,
+      "grad_norm": 0.16430321335792542,
+      "learning_rate": 0.00019731543624161075,
+      "loss": 0.65,
+      "step": 21
+    },
+    {
+      "epoch": 0.0550128936469485,
+      "grad_norm": 0.1478523313999176,
+      "learning_rate": 0.00019714765100671143,
+      "loss": 0.6412,
+      "step": 22
+    },
+    {
+      "epoch": 0.0575134797218098,
+      "grad_norm": 0.15282399952411652,
+      "learning_rate": 0.00019697986577181208,
+      "loss": 0.624,
+      "step": 23
+    },
+    {
+      "epoch": 0.060014065796671096,
+      "grad_norm": 0.15637804567813873,
+      "learning_rate": 0.00019681208053691276,
+      "loss": 0.633,
+      "step": 24
+    },
+    {
+      "epoch": 0.06251465187153239,
+      "grad_norm": 0.1476782113313675,
+      "learning_rate": 0.00019664429530201342,
+      "loss": 0.6406,
+      "step": 25
+    },
+    {
+      "epoch": 0.06501523794639369,
+      "grad_norm": 0.14628452062606812,
+      "learning_rate": 0.0001964765100671141,
+      "loss": 0.6197,
+      "step": 26
+    },
+    {
+      "epoch": 0.06751582402125499,
+      "grad_norm": 0.14777548611164093,
+      "learning_rate": 0.00019630872483221478,
+      "loss": 0.6153,
+      "step": 27
+    },
+    {
+      "epoch": 0.07001641009611628,
+      "grad_norm": 0.13805972039699554,
+      "learning_rate": 0.00019614093959731546,
+      "loss": 0.6014,
+      "step": 28
+    },
+    {
+      "epoch": 0.07251699617097758,
+      "grad_norm": 0.1502184271812439,
+      "learning_rate": 0.00019597315436241613,
+      "loss": 0.5974,
+      "step": 29
+    },
+    {
+      "epoch": 0.07501758224583886,
+      "grad_norm": 0.13858534395694733,
+      "learning_rate": 0.0001958053691275168,
+      "loss": 0.6144,
+      "step": 30
+    },
+    {
+      "epoch": 0.07751816832070016,
+      "grad_norm": 0.14372514188289642,
+      "learning_rate": 0.00019563758389261747,
+      "loss": 0.5941,
+      "step": 31
+    },
+    {
+      "epoch": 0.08001875439556146,
+      "grad_norm": 0.13981372117996216,
+      "learning_rate": 0.00019546979865771812,
+      "loss": 0.5945,
+      "step": 32
+    },
+    {
+      "epoch": 0.08251934047042275,
+      "grad_norm": 0.14265769720077515,
+      "learning_rate": 0.0001953020134228188,
+      "loss": 0.5848,
+      "step": 33
+    },
+    {
+      "epoch": 0.08501992654528405,
+      "grad_norm": 0.12909187376499176,
+      "learning_rate": 0.00019513422818791945,
+      "loss": 0.5802,
+      "step": 34
+    },
+    {
+      "epoch": 0.08752051262014535,
+      "grad_norm": 0.14645205438137054,
+      "learning_rate": 0.00019496644295302013,
+      "loss": 0.5689,
+      "step": 35
+    },
+    {
+      "epoch": 0.09002109869500664,
+      "grad_norm": 0.1427399069070816,
+      "learning_rate": 0.0001947986577181208,
+      "loss": 0.5726,
+      "step": 36
+    },
+    {
+      "epoch": 0.09252168476986794,
+      "grad_norm": 0.14320509135723114,
+      "learning_rate": 0.0001946308724832215,
+      "loss": 0.5544,
+      "step": 37
+    },
+    {
+      "epoch": 0.09502227084472924,
+      "grad_norm": 0.13662615418434143,
+      "learning_rate": 0.00019446308724832217,
+      "loss": 0.5793,
+      "step": 38
+    },
+    {
+      "epoch": 0.09752285691959053,
+      "grad_norm": 0.13153070211410522,
+      "learning_rate": 0.00019429530201342282,
+      "loss": 0.5846,
+      "step": 39
+    },
+    {
+      "epoch": 0.10002344299445183,
+      "grad_norm": 0.16413502395153046,
+      "learning_rate": 0.0001941275167785235,
+      "loss": 0.5574,
+      "step": 40
+    },
+    {
+      "epoch": 0.10252402906931311,
+      "grad_norm": 0.1483771800994873,
+      "learning_rate": 0.00019395973154362416,
+      "loss": 0.5716,
+      "step": 41
+    },
+    {
+      "epoch": 0.10502461514417441,
+      "grad_norm": 0.14123424887657166,
+      "learning_rate": 0.00019379194630872484,
+      "loss": 0.5619,
+      "step": 42
+    },
+    {
+      "epoch": 0.10752520121903571,
+      "grad_norm": 0.15064753592014313,
+      "learning_rate": 0.00019362416107382552,
+      "loss": 0.5769,
+      "step": 43
+    },
+    {
+      "epoch": 0.110025787293897,
+      "grad_norm": 0.15008266270160675,
+      "learning_rate": 0.00019345637583892617,
+      "loss": 0.555,
+      "step": 44
+    },
+    {
+      "epoch": 0.1125263733687583,
+      "grad_norm": 0.14735136926174164,
+      "learning_rate": 0.00019328859060402688,
+      "loss": 0.5331,
+      "step": 45
+    },
+    {
+      "epoch": 0.1150269594436196,
+      "grad_norm": 0.13642771542072296,
+      "learning_rate": 0.00019312080536912753,
+      "loss": 0.5397,
+      "step": 46
+    },
+    {
+      "epoch": 0.1175275455184809,
+      "grad_norm": 0.1735912710428238,
+      "learning_rate": 0.0001929530201342282,
+      "loss": 0.5451,
+      "step": 47
+    },
+    {
+      "epoch": 0.12002813159334219,
+      "grad_norm": 0.1445857733488083,
+      "learning_rate": 0.00019278523489932886,
+      "loss": 0.5739,
+      "step": 48
+    },
+    {
+      "epoch": 0.12252871766820349,
+      "grad_norm": 0.1417866051197052,
+      "learning_rate": 0.00019261744966442954,
+      "loss": 0.5435,
+      "step": 49
+    },
+    {
+      "epoch": 0.12502930374306478,
+      "grad_norm": 0.17940974235534668,
+      "learning_rate": 0.00019244966442953022,
+      "loss": 0.5602,
+      "step": 50
+    },
+    {
+      "epoch": 0.12752988981792607,
+      "grad_norm": 0.14657525718212128,
+      "learning_rate": 0.00019228187919463087,
+      "loss": 0.5227,
+      "step": 51
+    },
+    {
+      "epoch": 0.13003047589278738,
+      "grad_norm": 0.1798638254404068,
+      "learning_rate": 0.00019211409395973155,
+      "loss": 0.5395,
+      "step": 52
+    },
+    {
+      "epoch": 0.13253106196764866,
+      "grad_norm": 0.145979106426239,
+      "learning_rate": 0.0001919463087248322,
+      "loss": 0.5485,
+      "step": 53
+    },
+    {
+      "epoch": 0.13503164804250997,
+      "grad_norm": 0.14515143632888794,
+      "learning_rate": 0.0001917785234899329,
+      "loss": 0.5341,
+      "step": 54
+    },
+    {
+      "epoch": 0.13753223411737125,
+      "grad_norm": 0.14358939230442047,
+      "learning_rate": 0.00019161073825503357,
+      "loss": 0.5211,
+      "step": 55
+    },
+    {
+      "epoch": 0.14003282019223257,
+      "grad_norm": 0.14795799553394318,
+      "learning_rate": 0.00019144295302013425,
+      "loss": 0.5294,
+      "step": 56
+    },
+    {
+      "epoch": 0.14253340626709385,
+      "grad_norm": 0.14450453221797943,
+      "learning_rate": 0.0001912751677852349,
+      "loss": 0.5273,
+      "step": 57
+    },
+    {
+      "epoch": 0.14503399234195516,
+      "grad_norm": 0.15771551430225372,
+      "learning_rate": 0.00019110738255033558,
+      "loss": 0.5412,
+      "step": 58
+    },
+    {
+      "epoch": 0.14753457841681644,
+      "grad_norm": 0.18084295094013214,
+      "learning_rate": 0.00019093959731543626,
+      "loss": 0.5385,
+      "step": 59
+    },
+    {
+      "epoch": 0.15003516449167772,
+      "grad_norm": 0.1659678816795349,
+      "learning_rate": 0.0001907718120805369,
+      "loss": 0.5401,
+      "step": 60
+    },
+    {
+      "epoch": 0.15253575056653904,
+      "grad_norm": 0.15592657029628754,
+      "learning_rate": 0.0001906040268456376,
+      "loss": 0.5358,
+      "step": 61
+    },
+    {
+      "epoch": 0.15503633664140032,
+      "grad_norm": 0.19171522557735443,
+      "learning_rate": 0.00019043624161073827,
+      "loss": 0.5589,
+      "step": 62
+    },
+    {
+      "epoch": 0.15753692271626163,
+      "grad_norm": 0.17024098336696625,
+      "learning_rate": 0.00019026845637583895,
+      "loss": 0.5399,
+      "step": 63
+    },
+    {
+      "epoch": 0.1600375087911229,
+      "grad_norm": 0.16276618838310242,
+      "learning_rate": 0.0001901006711409396,
+      "loss": 0.5271,
+      "step": 64
+    },
+    {
+      "epoch": 0.16253809486598422,
+      "grad_norm": 0.1863684058189392,
+      "learning_rate": 0.00018993288590604028,
+      "loss": 0.5356,
+      "step": 65
+    },
+    {
+      "epoch": 0.1650386809408455,
+      "grad_norm": 0.15433144569396973,
+      "learning_rate": 0.00018976510067114096,
+      "loss": 0.5101,
+      "step": 66
+    },
+    {
+      "epoch": 0.16753926701570682,
+      "grad_norm": 0.1568458527326584,
+      "learning_rate": 0.00018959731543624161,
+      "loss": 0.5178,
+      "step": 67
+    },
+    {
+      "epoch": 0.1700398530905681,
+      "grad_norm": 0.19393286108970642,
+      "learning_rate": 0.0001894295302013423,
+      "loss": 0.5261,
+      "step": 68
+    },
+    {
+      "epoch": 0.1725404391654294,
+      "grad_norm": 0.15781769156455994,
+      "learning_rate": 0.00018926174496644295,
+      "loss": 0.5106,
+      "step": 69
+    },
+    {
+      "epoch": 0.1750410252402907,
+      "grad_norm": 0.1883280724287033,
+      "learning_rate": 0.00018909395973154363,
+      "loss": 0.5336,
+      "step": 70
+    },
+    {
+      "epoch": 0.17754161131515198,
+      "grad_norm": 0.184098020195961,
+      "learning_rate": 0.0001889261744966443,
+      "loss": 0.5079,
+      "step": 71
+    },
+    {
+      "epoch": 0.1800421973900133,
+      "grad_norm": 0.1611829400062561,
+      "learning_rate": 0.000188758389261745,
+      "loss": 0.5002,
+      "step": 72
+    },
+    {
+      "epoch": 0.18254278346487457,
+      "grad_norm": 0.21065734326839447,
+      "learning_rate": 0.00018859060402684564,
+      "loss": 0.5238,
+      "step": 73
+    },
+    {
+      "epoch": 0.18504336953973588,
+      "grad_norm": 0.16418707370758057,
+      "learning_rate": 0.00018842281879194632,
+      "loss": 0.5183,
+      "step": 74
+    },
+    {
+      "epoch": 0.18754395561459716,
+      "grad_norm": 0.17874297499656677,
+      "learning_rate": 0.000188255033557047,
+      "loss": 0.5206,
+      "step": 75
+    },
+    {
+      "epoch": 0.19004454168945847,
+      "grad_norm": 0.18666118383407593,
+      "learning_rate": 0.00018808724832214765,
+      "loss": 0.5143,
+      "step": 76
+    },
+    {
+      "epoch": 0.19254512776431976,
+      "grad_norm": 0.15301728248596191,
+      "learning_rate": 0.00018791946308724833,
+      "loss": 0.5038,
+      "step": 77
+    },
+    {
+      "epoch": 0.19504571383918107,
+      "grad_norm": 0.1661822497844696,
+      "learning_rate": 0.00018775167785234898,
+      "loss": 0.5182,
+      "step": 78
+    },
+    {
+      "epoch": 0.19754629991404235,
+      "grad_norm": 0.1866961419582367,
+      "learning_rate": 0.00018758389261744966,
+      "loss": 0.521,
+      "step": 79
+    },
+    {
+      "epoch": 0.20004688598890366,
+      "grad_norm": 0.178809255361557,
+      "learning_rate": 0.00018741610738255034,
+      "loss": 0.5088,
+      "step": 80
+    },
+    {
+      "epoch": 0.20254747206376494,
+      "grad_norm": 0.16507765650749207,
+      "learning_rate": 0.00018724832214765102,
+      "loss": 0.4998,
+      "step": 81
+    },
+    {
+      "epoch": 0.20504805813862623,
+      "grad_norm": 0.18188826739788055,
+      "learning_rate": 0.0001870805369127517,
+      "loss": 0.4959,
+      "step": 82
+    },
+    {
+      "epoch": 0.20754864421348754,
+      "grad_norm": 0.18728744983673096,
+      "learning_rate": 0.00018691275167785236,
+      "loss": 0.5192,
+      "step": 83
+    },
+    {
+      "epoch": 0.21004923028834882,
+      "grad_norm": 0.16816700994968414,
+      "learning_rate": 0.00018674496644295304,
+      "loss": 0.5021,
+      "step": 84
+    },
+    {
+      "epoch": 0.21254981636321013,
+      "grad_norm": 0.14855366945266724,
+      "learning_rate": 0.0001865771812080537,
+      "loss": 0.4962,
+      "step": 85
+    },
+    {
+      "epoch": 0.21505040243807141,
+      "grad_norm": 0.16501812636852264,
+      "learning_rate": 0.00018640939597315437,
+      "loss": 0.4927,
+      "step": 86
+    },
+    {
+      "epoch": 0.21755098851293272,
+      "grad_norm": 0.1743089109659195,
+      "learning_rate": 0.00018624161073825502,
+      "loss": 0.5142,
+      "step": 87
+    },
+    {
+      "epoch": 0.220051574587794,
+      "grad_norm": 0.1650586873292923,
+      "learning_rate": 0.0001860738255033557,
+      "loss": 0.4933,
+      "step": 88
+    },
+    {
+      "epoch": 0.22255216066265532,
+      "grad_norm": 0.16853803396224976,
+      "learning_rate": 0.00018590604026845638,
+      "loss": 0.5043,
+      "step": 89
+    },
+    {
+      "epoch": 0.2250527467375166,
+      "grad_norm": 0.16465862095355988,
+      "learning_rate": 0.00018573825503355706,
+      "loss": 0.5064,
+      "step": 90
+    },
+    {
+      "epoch": 0.2275533328123779,
+      "grad_norm": 0.17402450740337372,
+      "learning_rate": 0.00018557046979865774,
+      "loss": 0.5129,
+      "step": 91
+    },
+    {
+      "epoch": 0.2300539188872392,
+      "grad_norm": 0.16535209119319916,
+      "learning_rate": 0.0001854026845637584,
+      "loss": 0.4878,
+      "step": 92
+    },
+    {
+      "epoch": 0.2325545049621005,
+      "grad_norm": 0.19670793414115906,
+      "learning_rate": 0.00018523489932885907,
+      "loss": 0.4835,
+      "step": 93
+    },
+    {
+      "epoch": 0.2350550910369618,
+      "grad_norm": 0.15545278787612915,
+      "learning_rate": 0.00018506711409395973,
+      "loss": 0.4938,
+      "step": 94
+    },
+    {
+      "epoch": 0.23755567711182307,
+      "grad_norm": 0.17252099514007568,
+      "learning_rate": 0.0001848993288590604,
+      "loss": 0.5035,
+      "step": 95
+    },
+    {
+      "epoch": 0.24005626318668438,
+      "grad_norm": 0.20185311138629913,
+      "learning_rate": 0.00018473154362416109,
+      "loss": 0.5079,
+      "step": 96
+    },
+    {
+      "epoch": 0.24255684926154567,
+      "grad_norm": 0.1610802859067917,
+      "learning_rate": 0.00018456375838926174,
+      "loss": 0.4946,
+      "step": 97
+    },
+    {
+      "epoch": 0.24505743533640698,
+      "grad_norm": 0.2017122060060501,
+      "learning_rate": 0.00018439597315436244,
+      "loss": 0.496,
+      "step": 98
+    },
+    {
+      "epoch": 0.24755802141126826,
+      "grad_norm": 0.16128034889698029,
+      "learning_rate": 0.0001842281879194631,
+      "loss": 0.4813,
+      "step": 99
+    },
+    {
+      "epoch": 0.25005860748612957,
+      "grad_norm": 0.22258538007736206,
+      "learning_rate": 0.00018406040268456378,
+      "loss": 0.494,
+      "step": 100
+    },
+    {
+      "epoch": 0.2525591935609909,
+      "grad_norm": 0.18942825496196747,
+      "learning_rate": 0.00018389261744966443,
+      "loss": 0.489,
+      "step": 101
+    },
+    {
+      "epoch": 0.25505977963585214,
+      "grad_norm": 0.17586448788642883,
+      "learning_rate": 0.0001837248322147651,
+      "loss": 0.487,
+      "step": 102
+    },
+    {
+      "epoch": 0.25756036571071345,
+      "grad_norm": 0.2008223831653595,
+      "learning_rate": 0.0001835570469798658,
+      "loss": 0.494,
+      "step": 103
+    },
+    {
+      "epoch": 0.26006095178557476,
+      "grad_norm": 0.1799156367778778,
+      "learning_rate": 0.00018338926174496644,
+      "loss": 0.5082,
+      "step": 104
+    },
+    {
+      "epoch": 0.26256153786043607,
+      "grad_norm": 0.17686983942985535,
+      "learning_rate": 0.00018322147651006712,
+      "loss": 0.4727,
+      "step": 105
+    },
+    {
+      "epoch": 0.2650621239352973,
+      "grad_norm": 0.22485797107219696,
+      "learning_rate": 0.0001830536912751678,
+      "loss": 0.5017,
+      "step": 106
+    },
+    {
+      "epoch": 0.26756271001015863,
+      "grad_norm": 0.16205894947052002,
+      "learning_rate": 0.00018288590604026848,
+      "loss": 0.4744,
+      "step": 107
+    },
+    {
+      "epoch": 0.27006329608501994,
+      "grad_norm": 0.16790597140789032,
+      "learning_rate": 0.00018271812080536913,
+      "loss": 0.4773,
+      "step": 108
+    },
+    {
+      "epoch": 0.2725638821598812,
+      "grad_norm": 0.17916059494018555,
+      "learning_rate": 0.00018255033557046981,
+      "loss": 0.4779,
+      "step": 109
+    },
+    {
+      "epoch": 0.2750644682347425,
+      "grad_norm": 0.17400622367858887,
+      "learning_rate": 0.00018238255033557047,
+      "loss": 0.4977,
+      "step": 110
+    },
+    {
+      "epoch": 0.2775650543096038,
+      "grad_norm": 0.16923707723617554,
+      "learning_rate": 0.00018221476510067115,
+      "loss": 0.4933,
+      "step": 111
+    },
+    {
+      "epoch": 0.28006564038446513,
+      "grad_norm": 0.17827562987804413,
+      "learning_rate": 0.00018204697986577183,
+      "loss": 0.4888,
+      "step": 112
+    },
+    {
+      "epoch": 0.2825662264593264,
+      "grad_norm": 0.1732293665409088,
+      "learning_rate": 0.00018187919463087248,
+      "loss": 0.5098,
+      "step": 113
+    },
+    {
+      "epoch": 0.2850668125341877,
+      "grad_norm": 0.17349784076213837,
+      "learning_rate": 0.00018171140939597316,
+      "loss": 0.4924,
+      "step": 114
+    },
+    {
+      "epoch": 0.287567398609049,
+      "grad_norm": 0.20687106251716614,
+      "learning_rate": 0.00018154362416107384,
+      "loss": 0.5006,
+      "step": 115
+    },
+    {
+      "epoch": 0.2900679846839103,
+      "grad_norm": 0.15973134338855743,
+      "learning_rate": 0.00018137583892617452,
+      "loss": 0.4909,
+      "step": 116
+    },
+    {
+      "epoch": 0.2925685707587716,
+      "grad_norm": 0.21071845293045044,
+      "learning_rate": 0.00018120805369127517,
+      "loss": 0.4763,
+      "step": 117
+    },
+    {
+      "epoch": 0.2950691568336329,
+      "grad_norm": 0.19466304779052734,
+      "learning_rate": 0.00018104026845637585,
+      "loss": 0.4896,
+      "step": 118
+    },
+    {
+      "epoch": 0.2975697429084942,
+      "grad_norm": 0.19403570890426636,
+      "learning_rate": 0.00018087248322147653,
+      "loss": 0.4902,
+      "step": 119
+    },
+    {
+      "epoch": 0.30007032898335545,
+      "grad_norm": 0.219278946518898,
+      "learning_rate": 0.00018070469798657718,
+      "loss": 0.4883,
+      "step": 120
+    },
+    {
+      "epoch": 0.30257091505821676,
+      "grad_norm": 0.17110252380371094,
+      "learning_rate": 0.00018053691275167786,
+      "loss": 0.5022,
+      "step": 121
+    },
+    {
+      "epoch": 0.30507150113307807,
+      "grad_norm": 0.2356841117143631,
+      "learning_rate": 0.00018036912751677852,
+      "loss": 0.4794,
+      "step": 122
+    },
+    {
+      "epoch": 0.3075720872079394,
+      "grad_norm": 0.17025519907474518,
+      "learning_rate": 0.0001802013422818792,
+      "loss": 0.4828,
+      "step": 123
+    },
+    {
+      "epoch": 0.31007267328280064,
+      "grad_norm": 0.2149350345134735,
+      "learning_rate": 0.00018003355704697988,
+      "loss": 0.464,
+      "step": 124
+    },
+    {
+      "epoch": 0.31257325935766195,
+      "grad_norm": 0.17595884203910828,
+      "learning_rate": 0.00017986577181208056,
+      "loss": 0.4682,
+      "step": 125
+    },
+    {
+      "epoch": 0.31507384543252326,
+      "grad_norm": 0.17032761871814728,
+      "learning_rate": 0.0001796979865771812,
+      "loss": 0.4594,
+      "step": 126
+    },
+    {
+      "epoch": 0.31757443150738457,
+      "grad_norm": 0.15546788275241852,
+      "learning_rate": 0.0001795302013422819,
+      "loss": 0.4835,
+      "step": 127
+    },
+    {
+      "epoch": 0.3200750175822458,
+      "grad_norm": 0.18081508576869965,
+      "learning_rate": 0.00017936241610738257,
+      "loss": 0.477,
+      "step": 128
+    },
+    {
+      "epoch": 0.32257560365710714,
+      "grad_norm": 0.1631155163049698,
+      "learning_rate": 0.00017919463087248322,
+      "loss": 0.4768,
+      "step": 129
+    },
+    {
+      "epoch": 0.32507618973196845,
+      "grad_norm": 0.17103877663612366,
+      "learning_rate": 0.0001790268456375839,
+      "loss": 0.4725,
+      "step": 130
+    },
+    {
+      "epoch": 0.3275767758068297,
+      "grad_norm": 0.15857450664043427,
+      "learning_rate": 0.00017885906040268455,
+      "loss": 0.4777,
+      "step": 131
+    },
+    {
+      "epoch": 0.330077361881691,
+      "grad_norm": 0.1584555059671402,
+      "learning_rate": 0.00017869127516778523,
+      "loss": 0.4619,
+      "step": 132
+    },
+    {
+      "epoch": 0.3325779479565523,
+      "grad_norm": 0.18434159457683563,
+      "learning_rate": 0.0001785234899328859,
+      "loss": 0.4947,
+      "step": 133
+    },
+    {
+      "epoch": 0.33507853403141363,
+      "grad_norm": 0.17680639028549194,
+      "learning_rate": 0.0001783557046979866,
+      "loss": 0.4877,
+      "step": 134
+    },
+    {
+      "epoch": 0.3375791201062749,
+      "grad_norm": 0.17073361575603485,
+      "learning_rate": 0.00017818791946308727,
+      "loss": 0.4607,
+      "step": 135
+    },
+    {
+      "epoch": 0.3400797061811362,
+      "grad_norm": 0.15861110389232635,
+      "learning_rate": 0.00017802013422818792,
+      "loss": 0.4866,
+      "step": 136
+    },
+    {
+      "epoch": 0.3425802922559975,
+      "grad_norm": 0.16533519327640533,
+      "learning_rate": 0.0001778523489932886,
+      "loss": 0.4844,
+      "step": 137
+    },
+    {
+      "epoch": 0.3450808783308588,
+      "grad_norm": 0.1725560426712036,
+      "learning_rate": 0.00017768456375838926,
+      "loss": 0.4997,
+      "step": 138
+    },
+    {
+      "epoch": 0.3475814644057201,
+      "grad_norm": 0.16393449902534485,
+      "learning_rate": 0.00017751677852348994,
+      "loss": 0.4751,
+      "step": 139
+    },
+    {
+      "epoch": 0.3500820504805814,
+      "grad_norm": 0.16054843366146088,
+      "learning_rate": 0.00017734899328859062,
+      "loss": 0.4762,
+      "step": 140
+    },
+    {
+      "epoch": 0.3525826365554427,
+      "grad_norm": 0.17005555331707,
+      "learning_rate": 0.0001771812080536913,
+      "loss": 0.4669,
+      "step": 141
+    },
+    {
+      "epoch": 0.35508322263030395,
+      "grad_norm": 0.17783407866954803,
+      "learning_rate": 0.00017701342281879198,
+      "loss": 0.4824,
+      "step": 142
+    },
+    {
+      "epoch": 0.35758380870516526,
+      "grad_norm": 0.16758868098258972,
+      "learning_rate": 0.00017684563758389263,
+      "loss": 0.4742,
+      "step": 143
+    },
+    {
+      "epoch": 0.3600843947800266,
+      "grad_norm": 0.15669575333595276,
+      "learning_rate": 0.0001766778523489933,
+      "loss": 0.4594,
+      "step": 144
+    },
+    {
+      "epoch": 0.3625849808548879,
+      "grad_norm": 0.17958584427833557,
+      "learning_rate": 0.00017651006711409396,
+      "loss": 0.4786,
+      "step": 145
+    },
+    {
+      "epoch": 0.36508556692974914,
+      "grad_norm": 0.19689996540546417,
+      "learning_rate": 0.00017634228187919464,
+      "loss": 0.4858,
+      "step": 146
+    },
+    {
+      "epoch": 0.36758615300461045,
+      "grad_norm": 0.17314714193344116,
+      "learning_rate": 0.0001761744966442953,
+      "loss": 0.4773,
+      "step": 147
+    },
+    {
+      "epoch": 0.37008673907947176,
+      "grad_norm": 0.1754344254732132,
+      "learning_rate": 0.00017600671140939597,
+      "loss": 0.4848,
+      "step": 148
+    },
+    {
+      "epoch": 0.37258732515433307,
+      "grad_norm": 0.16518406569957733,
+      "learning_rate": 0.00017583892617449665,
+      "loss": 0.4593,
+      "step": 149
+    },
+    {
+      "epoch": 0.3750879112291943,
+      "grad_norm": 0.16496793925762177,
+      "learning_rate": 0.00017567114093959733,
+      "loss": 0.4755,
+      "step": 150
+    },
+    {
+      "epoch": 0.37758849730405564,
+      "grad_norm": 0.1820346862077713,
+      "learning_rate": 0.000175503355704698,
+      "loss": 0.4758,
+      "step": 151
+    },
+    {
+      "epoch": 0.38008908337891695,
+      "grad_norm": 0.1669500470161438,
+      "learning_rate": 0.00017533557046979867,
+      "loss": 0.4834,
+      "step": 152
+    },
+    {
+      "epoch": 0.3825896694537782,
+      "grad_norm": 0.18701374530792236,
+      "learning_rate": 0.00017516778523489935,
+      "loss": 0.4613,
+      "step": 153
+    },
+    {
+      "epoch": 0.3850902555286395,
+      "grad_norm": 0.1735297441482544,
+      "learning_rate": 0.000175,
+      "loss": 0.4742,
+      "step": 154
+    },
+    {
+      "epoch": 0.3875908416035008,
+      "grad_norm": 0.17582108080387115,
+      "learning_rate": 0.00017483221476510068,
+      "loss": 0.4712,
+      "step": 155
+    },
+    {
+      "epoch": 0.39009142767836213,
+      "grad_norm": 0.17562179267406464,
+      "learning_rate": 0.00017466442953020136,
+      "loss": 0.4543,
+      "step": 156
+    },
+    {
+      "epoch": 0.3925920137532234,
+      "grad_norm": 0.1820077896118164,
+      "learning_rate": 0.000174496644295302,
+      "loss": 0.4624,
+      "step": 157
+    },
+    {
+      "epoch": 0.3950925998280847,
+      "grad_norm": 0.19618383049964905,
+      "learning_rate": 0.0001743288590604027,
+      "loss": 0.48,
+      "step": 158
+    },
+    {
+      "epoch": 0.397593185902946,
+      "grad_norm": 0.20992861688137054,
+      "learning_rate": 0.00017416107382550337,
+      "loss": 0.4806,
+      "step": 159
+    },
+    {
+      "epoch": 0.4000937719778073,
+      "grad_norm": 0.17106316983699799,
+      "learning_rate": 0.00017399328859060405,
+      "loss": 0.4666,
+      "step": 160
+    },
+    {
+      "epoch": 0.4025943580526686,
+      "grad_norm": 0.20506031811237335,
+      "learning_rate": 0.0001738255033557047,
+      "loss": 0.465,
+      "step": 161
+    },
+    {
+      "epoch": 0.4050949441275299,
+      "grad_norm": 0.16451655328273773,
+      "learning_rate": 0.00017365771812080538,
+      "loss": 0.492,
+      "step": 162
+    },
+    {
+      "epoch": 0.4075955302023912,
+      "grad_norm": 0.17921902239322662,
+      "learning_rate": 0.00017348993288590604,
+      "loss": 0.4786,
+      "step": 163
+    },
+    {
+      "epoch": 0.41009611627725245,
+      "grad_norm": 0.19945573806762695,
+      "learning_rate": 0.00017332214765100671,
+      "loss": 0.4578,
+      "step": 164
+    },
+    {
+      "epoch": 0.41259670235211376,
+      "grad_norm": 0.1591523438692093,
+      "learning_rate": 0.0001731543624161074,
+      "loss": 0.4698,
+      "step": 165
+    },
+    {
+      "epoch": 0.4150972884269751,
+      "grad_norm": 0.2252337783575058,
+      "learning_rate": 0.00017298657718120805,
+      "loss": 0.4669,
+      "step": 166
+    },
+    {
+      "epoch": 0.4175978745018364,
+      "grad_norm": 0.17984913289546967,
+      "learning_rate": 0.00017281879194630873,
+      "loss": 0.4672,
+      "step": 167
+    },
+    {
+      "epoch": 0.42009846057669764,
+      "grad_norm": 0.1938534379005432,
+      "learning_rate": 0.0001726510067114094,
+      "loss": 0.4718,
+      "step": 168
+    },
+    {
+      "epoch": 0.42259904665155895,
+      "grad_norm": 0.18750695884227753,
+      "learning_rate": 0.0001724832214765101,
+      "loss": 0.4747,
+      "step": 169
+    },
+    {
+      "epoch": 0.42509963272642026,
+      "grad_norm": 0.17218442261219025,
+      "learning_rate": 0.00017231543624161074,
+      "loss": 0.4637,
+      "step": 170
+    },
+    {
+      "epoch": 0.4276002188012816,
+      "grad_norm": 0.1735689640045166,
+      "learning_rate": 0.00017214765100671142,
+      "loss": 0.4842,
+      "step": 171
+    },
+    {
+      "epoch": 0.43010080487614283,
+      "grad_norm": 0.18109886348247528,
+      "learning_rate": 0.0001719798657718121,
+      "loss": 0.4927,
+      "step": 172
+    },
+    {
+      "epoch": 0.43260139095100414,
+      "grad_norm": 0.18199589848518372,
+      "learning_rate": 0.00017181208053691275,
+      "loss": 0.4677,
+      "step": 173
+    },
+    {
+      "epoch": 0.43510197702586545,
+      "grad_norm": 0.1843508929014206,
+      "learning_rate": 0.00017164429530201343,
+      "loss": 0.4521,
+      "step": 174
+    },
+    {
+      "epoch": 0.43760256310072676,
+      "grad_norm": 0.1735106110572815,
+      "learning_rate": 0.00017147651006711408,
+      "loss": 0.4761,
+      "step": 175
+    },
+    {
+      "epoch": 0.440103149175588,
+      "grad_norm": 0.16991625726222992,
+      "learning_rate": 0.0001713087248322148,
+      "loss": 0.4507,
+      "step": 176
+    },
+    {
+      "epoch": 0.4426037352504493,
+      "grad_norm": 0.16398338973522186,
+      "learning_rate": 0.00017114093959731544,
+      "loss": 0.46,
+      "step": 177
+    },
+    {
+      "epoch": 0.44510432132531064,
+      "grad_norm": 0.18237490952014923,
+      "learning_rate": 0.00017097315436241612,
+      "loss": 0.4627,
+      "step": 178
+    },
+    {
+      "epoch": 0.4476049074001719,
+      "grad_norm": 0.174173042178154,
+      "learning_rate": 0.0001708053691275168,
+      "loss": 0.4514,
+      "step": 179
+    },
+    {
+      "epoch": 0.4501054934750332,
+      "grad_norm": 0.1653733253479004,
+      "learning_rate": 0.00017063758389261746,
+      "loss": 0.4551,
+      "step": 180
+    },
+    {
+      "epoch": 0.4526060795498945,
+      "grad_norm": 0.19772449135780334,
+      "learning_rate": 0.00017046979865771814,
+      "loss": 0.4463,
+      "step": 181
+    },
+    {
+      "epoch": 0.4551066656247558,
+      "grad_norm": 0.1714743971824646,
+      "learning_rate": 0.0001703020134228188,
+      "loss": 0.4673,
+      "step": 182
+    },
+    {
+      "epoch": 0.4576072516996171,
+      "grad_norm": 0.181261345744133,
+      "learning_rate": 0.00017013422818791947,
+      "loss": 0.4556,
+      "step": 183
+    },
+    {
+      "epoch": 0.4601078377744784,
+      "grad_norm": 0.15487723052501678,
+      "learning_rate": 0.00016996644295302012,
+      "loss": 0.4568,
+      "step": 184
+    },
+    {
+      "epoch": 0.4626084238493397,
+      "grad_norm": 0.16992811858654022,
+      "learning_rate": 0.00016979865771812083,
+      "loss": 0.4686,
+      "step": 185
+    },
+    {
+      "epoch": 0.465109009924201,
+      "grad_norm": 0.1934443712234497,
+      "learning_rate": 0.00016963087248322148,
+      "loss": 0.4695,
+      "step": 186
+    },
+    {
+      "epoch": 0.46760959599906227,
+      "grad_norm": 0.16562269628047943,
+      "learning_rate": 0.00016946308724832216,
+      "loss": 0.449,
+      "step": 187
+    },
+    {
+      "epoch": 0.4701101820739236,
+      "grad_norm": 0.19825531542301178,
+      "learning_rate": 0.00016929530201342284,
+      "loss": 0.4623,
+      "step": 188
+    },
+    {
+      "epoch": 0.4726107681487849,
+      "grad_norm": 0.1690351963043213,
+      "learning_rate": 0.0001691275167785235,
+      "loss": 0.4584,
+      "step": 189
+    },
+    {
+      "epoch": 0.47511135422364614,
+      "grad_norm": 0.1793663501739502,
+      "learning_rate": 0.00016895973154362417,
+      "loss": 0.4598,
+      "step": 190
+    },
+    {
+      "epoch": 0.47761194029850745,
+      "grad_norm": 0.1759204864501953,
+      "learning_rate": 0.00016879194630872483,
+      "loss": 0.4601,
+      "step": 191
+    },
+    {
+      "epoch": 0.48011252637336876,
+      "grad_norm": 0.16646680235862732,
+      "learning_rate": 0.0001686241610738255,
+      "loss": 0.4662,
+      "step": 192
+    },
+    {
+      "epoch": 0.4826131124482301,
+      "grad_norm": 0.1967586725950241,
+      "learning_rate": 0.00016845637583892619,
+      "loss": 0.4782,
+      "step": 193
+    },
+    {
+      "epoch": 0.48511369852309133,
+      "grad_norm": 0.16055084764957428,
+      "learning_rate": 0.00016828859060402686,
+      "loss": 0.4603,
+      "step": 194
+    },
+    {
+      "epoch": 0.48761428459795264,
+      "grad_norm": 0.2091376930475235,
+      "learning_rate": 0.00016812080536912754,
+      "loss": 0.4591,
+      "step": 195
+    },
+    {
+      "epoch": 0.49011487067281395,
+      "grad_norm": 0.17712931334972382,
+      "learning_rate": 0.0001679530201342282,
+      "loss": 0.4405,
+      "step": 196
+    },
+    {
+      "epoch": 0.49261545674767526,
+      "grad_norm": 0.16605544090270996,
+      "learning_rate": 0.00016778523489932888,
+      "loss": 0.4502,
+      "step": 197
+    },
+    {
+      "epoch": 0.4951160428225365,
+      "grad_norm": 0.18599025905132294,
+      "learning_rate": 0.00016761744966442953,
+      "loss": 0.449,
+      "step": 198
+    },
+    {
+      "epoch": 0.49761662889739783,
+      "grad_norm": 0.17631739377975464,
+      "learning_rate": 0.0001674496644295302,
+      "loss": 0.4646,
+      "step": 199
+    },
+    {
+      "epoch": 0.5001172149722591,
+      "grad_norm": 0.17810176312923431,
+      "learning_rate": 0.00016728187919463086,
+      "loss": 0.4538,
+      "step": 200
+    },
+    {
+      "epoch": 0.5026178010471204,
+      "grad_norm": 0.18166030943393707,
+      "learning_rate": 0.00016711409395973154,
+      "loss": 0.4596,
+      "step": 201
+    },
+    {
+      "epoch": 0.5051183871219818,
+      "grad_norm": 0.1877959668636322,
+      "learning_rate": 0.00016694630872483222,
+      "loss": 0.4421,
+      "step": 202
+    },
+    {
+      "epoch": 0.507618973196843,
+      "grad_norm": 0.16203252971172333,
+      "learning_rate": 0.0001667785234899329,
+      "loss": 0.4435,
+      "step": 203
+    },
+    {
+      "epoch": 0.5101195592717043,
+      "grad_norm": 0.22125209867954254,
+      "learning_rate": 0.00016661073825503358,
+      "loss": 0.4568,
+      "step": 204
+    },
+    {
+      "epoch": 0.5126201453465656,
+      "grad_norm": 0.16617968678474426,
+      "learning_rate": 0.00016644295302013423,
+      "loss": 0.4524,
+      "step": 205
+    },
+    {
+      "epoch": 0.5151207314214269,
+      "grad_norm": 0.1956106424331665,
+      "learning_rate": 0.00016627516778523491,
+      "loss": 0.4568,
+      "step": 206
+    },
+    {
+      "epoch": 0.5176213174962881,
+      "grad_norm": 0.17417706549167633,
+      "learning_rate": 0.00016610738255033557,
+      "loss": 0.4393,
+      "step": 207
+    },
+    {
+      "epoch": 0.5201219035711495,
+      "grad_norm": 0.1713646799325943,
+      "learning_rate": 0.00016593959731543625,
+      "loss": 0.443,
+      "step": 208
+    },
+    {
+      "epoch": 0.5226224896460108,
+      "grad_norm": 0.18685942888259888,
+      "learning_rate": 0.00016577181208053693,
+      "loss": 0.4451,
+      "step": 209
+    },
+    {
+      "epoch": 0.5251230757208721,
+      "grad_norm": 0.17455795407295227,
+      "learning_rate": 0.00016560402684563758,
+      "loss": 0.4538,
+      "step": 210
+    },
+    {
+      "epoch": 0.5276236617957334,
+      "grad_norm": 0.17413204908370972,
+      "learning_rate": 0.00016543624161073829,
+      "loss": 0.4453,
+      "step": 211
+    },
+    {
+      "epoch": 0.5301242478705946,
+      "grad_norm": 0.16172130405902863,
+      "learning_rate": 0.00016526845637583894,
+      "loss": 0.4584,
+      "step": 212
+    },
+    {
+      "epoch": 0.532624833945456,
+      "grad_norm": 0.16567657887935638,
+      "learning_rate": 0.00016510067114093962,
+      "loss": 0.4429,
+      "step": 213
+    },
+    {
+      "epoch": 0.5351254200203173,
+      "grad_norm": 0.17749889194965363,
+      "learning_rate": 0.00016493288590604027,
+      "loss": 0.4395,
+      "step": 214
+    },
+    {
+      "epoch": 0.5376260060951785,
+      "grad_norm": 0.16244365274906158,
+      "learning_rate": 0.00016476510067114095,
+      "loss": 0.455,
+      "step": 215
+    },
+    {
+      "epoch": 0.5401265921700399,
+      "grad_norm": 0.16314131021499634,
+      "learning_rate": 0.0001645973154362416,
+      "loss": 0.4383,
+      "step": 216
+    },
+    {
+      "epoch": 0.5426271782449011,
+      "grad_norm": 0.16629692912101746,
+      "learning_rate": 0.00016442953020134228,
+      "loss": 0.4391,
+      "step": 217
+    },
+    {
+      "epoch": 0.5451277643197624,
+      "grad_norm": 0.18032211065292358,
+      "learning_rate": 0.00016426174496644296,
+      "loss": 0.4502,
+      "step": 218
+    },
+    {
+      "epoch": 0.5476283503946238,
+      "grad_norm": 0.1673993170261383,
+      "learning_rate": 0.00016409395973154362,
+      "loss": 0.4518,
+      "step": 219
+    },
+    {
+      "epoch": 0.550128936469485,
+      "grad_norm": 0.17051811516284943,
+      "learning_rate": 0.00016392617449664432,
+      "loss": 0.4658,
+      "step": 220
+    },
+    {
+      "epoch": 0.5526295225443464,
+      "grad_norm": 0.17144182324409485,
+      "learning_rate": 0.00016375838926174498,
+      "loss": 0.455,
+      "step": 221
+    },
+    {
+      "epoch": 0.5551301086192076,
+      "grad_norm": 0.16543865203857422,
+      "learning_rate": 0.00016359060402684566,
+      "loss": 0.4524,
+      "step": 222
+    },
+    {
+      "epoch": 0.5576306946940689,
+      "grad_norm": 0.1819363236427307,
+      "learning_rate": 0.0001634228187919463,
+      "loss": 0.4374,
+      "step": 223
+    },
+    {
+      "epoch": 0.5601312807689303,
+      "grad_norm": 0.1822425127029419,
+      "learning_rate": 0.000163255033557047,
+      "loss": 0.4437,
+      "step": 224
+    },
+    {
+      "epoch": 0.5626318668437915,
+      "grad_norm": 0.17073342204093933,
+      "learning_rate": 0.00016308724832214767,
+      "loss": 0.4654,
+      "step": 225
+    },
+    {
+      "epoch": 0.5651324529186528,
+      "grad_norm": 0.1630779355764389,
+      "learning_rate": 0.00016291946308724832,
+      "loss": 0.4577,
+      "step": 226
+    },
+    {
+      "epoch": 0.5676330389935141,
+      "grad_norm": 0.17408612370491028,
+      "learning_rate": 0.000162751677852349,
+      "loss": 0.445,
+      "step": 227
+    },
+    {
+      "epoch": 0.5701336250683754,
+      "grad_norm": 0.2113485187292099,
+      "learning_rate": 0.00016258389261744965,
+      "loss": 0.4486,
+      "step": 228
+    },
+    {
+      "epoch": 0.5726342111432366,
+      "grad_norm": 0.16299131512641907,
+      "learning_rate": 0.00016241610738255036,
+      "loss": 0.4508,
+      "step": 229
+    },
+    {
+      "epoch": 0.575134797218098,
+      "grad_norm": 0.20489177107810974,
+      "learning_rate": 0.000162248322147651,
+      "loss": 0.4742,
+      "step": 230
+    },
+    {
+      "epoch": 0.5776353832929593,
+      "grad_norm": 0.1746734082698822,
+      "learning_rate": 0.0001620805369127517,
+      "loss": 0.4382,
+      "step": 231
+    },
+    {
+      "epoch": 0.5801359693678206,
+      "grad_norm": 0.1877690851688385,
+      "learning_rate": 0.00016191275167785237,
+      "loss": 0.4483,
+      "step": 232
+    },
+    {
+      "epoch": 0.5826365554426819,
+      "grad_norm": 0.19249488413333893,
+      "learning_rate": 0.00016174496644295302,
+      "loss": 0.4458,
+      "step": 233
+    },
+    {
+      "epoch": 0.5851371415175431,
+      "grad_norm": 0.1570655256509781,
+      "learning_rate": 0.0001615771812080537,
+      "loss": 0.4474,
+      "step": 234
+    },
+    {
+      "epoch": 0.5876377275924045,
+      "grad_norm": 0.188018798828125,
+      "learning_rate": 0.00016140939597315436,
+      "loss": 0.4345,
+      "step": 235
+    },
+    {
+      "epoch": 0.5901383136672658,
+      "grad_norm": 0.20541886985301971,
+      "learning_rate": 0.00016124161073825504,
+      "loss": 0.462,
+      "step": 236
+    },
+    {
+      "epoch": 0.592638899742127,
+      "grad_norm": 0.17590679228305817,
+      "learning_rate": 0.0001610738255033557,
+      "loss": 0.4613,
+      "step": 237
+    },
+    {
+      "epoch": 0.5951394858169884,
+      "grad_norm": 0.21810193359851837,
+      "learning_rate": 0.0001609060402684564,
+      "loss": 0.4319,
+      "step": 238
+    },
+    {
+      "epoch": 0.5976400718918496,
+      "grad_norm": 0.20115698873996735,
+      "learning_rate": 0.00016073825503355705,
+      "loss": 0.4347,
+      "step": 239
+    },
+    {
+      "epoch": 0.6001406579667109,
+      "grad_norm": 0.17849840223789215,
+      "learning_rate": 0.00016057046979865773,
+      "loss": 0.4478,
+      "step": 240
+    },
+    {
+      "epoch": 0.6026412440415723,
+      "grad_norm": 0.1893172413110733,
+      "learning_rate": 0.0001604026845637584,
+      "loss": 0.451,
+      "step": 241
+    },
+    {
+      "epoch": 0.6051418301164335,
+      "grad_norm": 0.19342820346355438,
+      "learning_rate": 0.00016023489932885906,
+      "loss": 0.4513,
+      "step": 242
+    },
+    {
+      "epoch": 0.6076424161912949,
+      "grad_norm": 0.19177012145519257,
+      "learning_rate": 0.00016006711409395974,
+      "loss": 0.4487,
+      "step": 243
+    },
+    {
+      "epoch": 0.6101430022661561,
+      "grad_norm": 0.2117731124162674,
+      "learning_rate": 0.0001598993288590604,
+      "loss": 0.4461,
+      "step": 244
+    },
+    {
+      "epoch": 0.6126435883410174,
+      "grad_norm": 0.16634789109230042,
+      "learning_rate": 0.00015973154362416107,
+      "loss": 0.4482,
+      "step": 245
+    },
+    {
+      "epoch": 0.6151441744158788,
+      "grad_norm": 0.19687195122241974,
+      "learning_rate": 0.00015956375838926175,
+      "loss": 0.4445,
+      "step": 246
+    },
+    {
+      "epoch": 0.61764476049074,
+      "grad_norm": 0.17551076412200928,
+      "learning_rate": 0.00015939597315436243,
+      "loss": 0.445,
+      "step": 247
+    },
+    {
+      "epoch": 0.6201453465656013,
+      "grad_norm": 0.16498766839504242,
+      "learning_rate": 0.0001592281879194631,
+      "loss": 0.4397,
+      "step": 248
+    },
+    {
+      "epoch": 0.6226459326404626,
+      "grad_norm": 0.17818611860275269,
+      "learning_rate": 0.00015906040268456377,
+      "loss": 0.4466,
+      "step": 249
+    },
+    {
+      "epoch": 0.6251465187153239,
+      "grad_norm": 0.205690398812294,
+      "learning_rate": 0.00015889261744966445,
+      "loss": 0.4448,
+      "step": 250
+    },
+    {
+      "epoch": 0.6276471047901852,
+      "grad_norm": 0.17474979162216187,
+      "learning_rate": 0.0001587248322147651,
+      "loss": 0.4293,
+      "step": 251
+    },
+    {
+      "epoch": 0.6301476908650465,
+      "grad_norm": 0.21222396194934845,
+      "learning_rate": 0.00015855704697986578,
+      "loss": 0.4368,
+      "step": 252
+    },
+    {
+      "epoch": 0.6326482769399078,
+      "grad_norm": 0.1702364981174469,
+      "learning_rate": 0.00015838926174496643,
+      "loss": 0.4283,
+      "step": 253
+    },
+    {
+      "epoch": 0.6351488630147691,
+      "grad_norm": 0.1743675321340561,
+      "learning_rate": 0.0001582214765100671,
+      "loss": 0.4619,
+      "step": 254
+    },
+    {
+      "epoch": 0.6376494490896304,
+      "grad_norm": 0.17541950941085815,
+      "learning_rate": 0.0001580536912751678,
+      "loss": 0.4344,
+      "step": 255
+    },
+    {
+      "epoch": 0.6401500351644916,
+      "grad_norm": 0.154206320643425,
+      "learning_rate": 0.00015788590604026847,
+      "loss": 0.4299,
+      "step": 256
+    },
+    {
+      "epoch": 0.642650621239353,
+      "grad_norm": 0.16429094970226288,
+      "learning_rate": 0.00015771812080536915,
+      "loss": 0.4367,
+      "step": 257
+    },
+    {
+      "epoch": 0.6451512073142143,
+      "grad_norm": 0.21027475595474243,
+      "learning_rate": 0.0001575503355704698,
+      "loss": 0.4668,
+      "step": 258
+    },
+    {
+      "epoch": 0.6476517933890755,
+      "grad_norm": 0.16886752843856812,
+      "learning_rate": 0.00015738255033557048,
+      "loss": 0.4521,
+      "step": 259
+    },
+    {
+      "epoch": 0.6501523794639369,
+      "grad_norm": 0.17572204768657684,
+      "learning_rate": 0.00015721476510067114,
+      "loss": 0.4235,
+      "step": 260
+    },
+    {
+      "epoch": 0.6526529655387981,
+      "grad_norm": 0.1727110594511032,
+      "learning_rate": 0.00015704697986577181,
+      "loss": 0.4611,
+      "step": 261
+    },
+    {
+      "epoch": 0.6551535516136594,
+      "grad_norm": 0.16705434024333954,
+      "learning_rate": 0.0001568791946308725,
+      "loss": 0.4488,
+      "step": 262
+    },
+    {
+      "epoch": 0.6576541376885208,
+      "grad_norm": 0.17407532036304474,
+      "learning_rate": 0.00015671140939597315,
+      "loss": 0.4524,
+      "step": 263
+    },
+    {
+      "epoch": 0.660154723763382,
+      "grad_norm": 0.17913798987865448,
+      "learning_rate": 0.00015654362416107385,
+      "loss": 0.4479,
+      "step": 264
+    },
+    {
+      "epoch": 0.6626553098382434,
+      "grad_norm": 0.1647716462612152,
+      "learning_rate": 0.0001563758389261745,
+      "loss": 0.4231,
+      "step": 265
+    },
+    {
+      "epoch": 0.6651558959131046,
+      "grad_norm": 0.20202037692070007,
+      "learning_rate": 0.0001562080536912752,
+      "loss": 0.4341,
+      "step": 266
+    },
+    {
+      "epoch": 0.6676564819879659,
+      "grad_norm": 0.16514341533184052,
+      "learning_rate": 0.00015604026845637584,
+      "loss": 0.4298,
+      "step": 267
+    },
+    {
+      "epoch": 0.6701570680628273,
+      "grad_norm": 0.18647202849388123,
+      "learning_rate": 0.00015587248322147652,
+      "loss": 0.4459,
+      "step": 268
+    },
+    {
+      "epoch": 0.6726576541376885,
+      "grad_norm": 0.1699058711528778,
+      "learning_rate": 0.0001557046979865772,
+      "loss": 0.45,
+      "step": 269
+    },
+    {
+      "epoch": 0.6751582402125498,
+      "grad_norm": 0.15985633432865143,
+      "learning_rate": 0.00015553691275167785,
+      "loss": 0.434,
+      "step": 270
+    },
+    {
+      "epoch": 0.6776588262874111,
+      "grad_norm": 0.18722815811634064,
+      "learning_rate": 0.00015536912751677853,
+      "loss": 0.4667,
+      "step": 271
+    },
+    {
+      "epoch": 0.6801594123622724,
+      "grad_norm": 0.19050948321819305,
+      "learning_rate": 0.00015520134228187918,
+      "loss": 0.4518,
+      "step": 272
+    },
+    {
+      "epoch": 0.6826599984371337,
+      "grad_norm": 0.16122502088546753,
+      "learning_rate": 0.0001550335570469799,
+      "loss": 0.4253,
+      "step": 273
+    },
+    {
+      "epoch": 0.685160584511995,
+      "grad_norm": 0.19916869699954987,
+      "learning_rate": 0.00015486577181208054,
+      "loss": 0.4563,
+      "step": 274
+    },
+    {
+      "epoch": 0.6876611705868563,
+      "grad_norm": 0.16123981773853302,
+      "learning_rate": 0.00015469798657718122,
+      "loss": 0.4341,
+      "step": 275
+    },
+    {
+      "epoch": 0.6901617566617176,
+      "grad_norm": 0.1763208657503128,
+      "learning_rate": 0.00015453020134228188,
+      "loss": 0.437,
+      "step": 276
+    },
+    {
+      "epoch": 0.6926623427365789,
+      "grad_norm": 0.17635302245616913,
+      "learning_rate": 0.00015436241610738256,
+      "loss": 0.4443,
+      "step": 277
+    },
+    {
+      "epoch": 0.6951629288114402,
+      "grad_norm": 0.17334021627902985,
+      "learning_rate": 0.00015419463087248324,
+      "loss": 0.4571,
+      "step": 278
+    },
+    {
+      "epoch": 0.6976635148863015,
+      "grad_norm": 0.1741616278886795,
+      "learning_rate": 0.0001540268456375839,
+      "loss": 0.4286,
+      "step": 279
+    },
+    {
+      "epoch": 0.7001641009611628,
+      "grad_norm": 0.1762605905532837,
+      "learning_rate": 0.00015385906040268457,
+      "loss": 0.4433,
+      "step": 280
+    },
+    {
+      "epoch": 0.702664687036024,
+      "grad_norm": 0.17451149225234985,
+      "learning_rate": 0.00015369127516778522,
+      "loss": 0.426,
+      "step": 281
+    },
+    {
+      "epoch": 0.7051652731108854,
+      "grad_norm": 0.16833680868148804,
+      "learning_rate": 0.00015352348993288593,
+      "loss": 0.44,
+      "step": 282
+    },
+    {
+      "epoch": 0.7076658591857466,
+      "grad_norm": 0.18105420470237732,
+      "learning_rate": 0.00015335570469798658,
+      "loss": 0.4415,
+      "step": 283
+    },
+    {
+      "epoch": 0.7101664452606079,
+      "grad_norm": 0.16717125475406647,
+      "learning_rate": 0.00015318791946308726,
+      "loss": 0.4353,
+      "step": 284
+    },
+    {
+      "epoch": 0.7126670313354693,
+      "grad_norm": 0.16678327322006226,
+      "learning_rate": 0.00015302013422818794,
+      "loss": 0.4171,
+      "step": 285
+    },
+    {
+      "epoch": 0.7151676174103305,
+      "grad_norm": 0.16035279631614685,
+      "learning_rate": 0.0001528523489932886,
+      "loss": 0.4553,
+      "step": 286
+    },
+    {
+      "epoch": 0.7176682034851919,
+      "grad_norm": 0.1686025708913803,
+      "learning_rate": 0.00015268456375838927,
+      "loss": 0.4301,
+      "step": 287
+    },
+    {
+      "epoch": 0.7201687895600531,
+      "grad_norm": 0.16823913156986237,
+      "learning_rate": 0.00015251677852348993,
+      "loss": 0.4426,
+      "step": 288
+    },
+    {
+      "epoch": 0.7226693756349144,
+      "grad_norm": 0.16710013151168823,
+      "learning_rate": 0.0001523489932885906,
+      "loss": 0.4258,
+      "step": 289
+    },
+    {
+      "epoch": 0.7251699617097758,
+      "grad_norm": 0.16891926527023315,
+      "learning_rate": 0.00015218120805369129,
+      "loss": 0.4316,
+      "step": 290
+    },
+    {
+      "epoch": 0.727670547784637,
+      "grad_norm": 0.18143802881240845,
+      "learning_rate": 0.00015201342281879196,
+      "loss": 0.4441,
+      "step": 291
+    },
+    {
+      "epoch": 0.7301711338594983,
+      "grad_norm": 0.1883898675441742,
+      "learning_rate": 0.00015184563758389262,
+      "loss": 0.4428,
+      "step": 292
+    },
+    {
+      "epoch": 0.7326717199343596,
+      "grad_norm": 0.19613900780677795,
+      "learning_rate": 0.0001516778523489933,
+      "loss": 0.443,
+      "step": 293
+    },
+    {
+      "epoch": 0.7351723060092209,
+      "grad_norm": 0.16892360150814056,
+      "learning_rate": 0.00015151006711409398,
+      "loss": 0.4373,
+      "step": 294
+    },
+    {
+      "epoch": 0.7376728920840822,
+      "grad_norm": 0.1972968578338623,
+      "learning_rate": 0.00015134228187919463,
+      "loss": 0.4349,
+      "step": 295
+    },
+    {
+      "epoch": 0.7401734781589435,
+      "grad_norm": 0.1730852872133255,
+      "learning_rate": 0.0001511744966442953,
+      "loss": 0.4382,
+      "step": 296
+    },
+    {
+      "epoch": 0.7426740642338048,
+      "grad_norm": 0.1790764033794403,
+      "learning_rate": 0.00015100671140939596,
+      "loss": 0.4348,
+      "step": 297
+    },
+    {
+      "epoch": 0.7451746503086661,
+      "grad_norm": 0.20283092558383942,
+      "learning_rate": 0.00015083892617449664,
+      "loss": 0.4385,
+      "step": 298
+    },
+    {
+      "epoch": 0.7476752363835274,
+      "grad_norm": 0.17277692258358002,
+      "learning_rate": 0.00015067114093959732,
+      "loss": 0.4465,
+      "step": 299
+    },
+    {
+      "epoch": 0.7501758224583887,
+      "grad_norm": 0.19094397127628326,
+      "learning_rate": 0.000150503355704698,
+      "loss": 0.427,
+      "step": 300
+    },
+    {
+      "epoch": 0.75267640853325,
+      "grad_norm": 0.17495505511760712,
+      "learning_rate": 0.00015033557046979868,
+      "loss": 0.4405,
+      "step": 301
+    },
+    {
+      "epoch": 0.7551769946081113,
+      "grad_norm": 0.1863364279270172,
+      "learning_rate": 0.00015016778523489933,
+      "loss": 0.4349,
+      "step": 302
+    },
+    {
+      "epoch": 0.7576775806829725,
+      "grad_norm": 0.185690775513649,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.4465,
+      "step": 303
+    },
+    {
+      "epoch": 0.7601781667578339,
+      "grad_norm": 0.17615684866905212,
+      "learning_rate": 0.00014983221476510067,
+      "loss": 0.4521,
+      "step": 304
+    },
+    {
+      "epoch": 0.7626787528326952,
+      "grad_norm": 0.16942819952964783,
+      "learning_rate": 0.00014966442953020135,
+      "loss": 0.4493,
+      "step": 305
+    },
+    {
+      "epoch": 0.7651793389075564,
+      "grad_norm": 0.17781828343868256,
+      "learning_rate": 0.00014949664429530203,
+      "loss": 0.4497,
+      "step": 306
+    },
+    {
+      "epoch": 0.7676799249824178,
+      "grad_norm": 0.1723792403936386,
+      "learning_rate": 0.00014932885906040268,
+      "loss": 0.4274,
+      "step": 307
+    },
+    {
+      "epoch": 0.770180511057279,
+      "grad_norm": 0.17277772724628448,
+      "learning_rate": 0.00014916107382550339,
+      "loss": 0.4325,
+      "step": 308
+    },
+    {
+      "epoch": 0.7726810971321404,
+      "grad_norm": 0.16546839475631714,
+      "learning_rate": 0.00014899328859060404,
+      "loss": 0.4449,
+      "step": 309
+    },
+    {
+      "epoch": 0.7751816832070016,
+      "grad_norm": 0.1712421327829361,
+      "learning_rate": 0.00014882550335570472,
+      "loss": 0.4398,
+      "step": 310
+    },
+    {
+      "epoch": 0.7776822692818629,
+      "grad_norm": 0.16765455901622772,
+      "learning_rate": 0.00014865771812080537,
+      "loss": 0.4114,
+      "step": 311
+    },
+    {
+      "epoch": 0.7801828553567243,
+      "grad_norm": 0.16849608719348907,
+      "learning_rate": 0.00014848993288590605,
+      "loss": 0.4351,
+      "step": 312
+    },
+    {
+      "epoch": 0.7826834414315855,
+      "grad_norm": 0.16892583668231964,
+      "learning_rate": 0.0001483221476510067,
+      "loss": 0.4381,
+      "step": 313
+    },
+    {
+      "epoch": 0.7851840275064468,
+      "grad_norm": 0.16866593062877655,
+      "learning_rate": 0.00014815436241610738,
+      "loss": 0.4413,
+      "step": 314
+    },
+    {
+      "epoch": 0.7876846135813081,
+      "grad_norm": 0.18011045455932617,
+      "learning_rate": 0.00014798657718120806,
+      "loss": 0.4219,
+      "step": 315
+    },
+    {
+      "epoch": 0.7901851996561694,
+      "grad_norm": 0.17827914655208588,
+      "learning_rate": 0.00014781879194630872,
+      "loss": 0.4484,
+      "step": 316
+    },
+    {
+      "epoch": 0.7926857857310307,
+      "grad_norm": 0.17511065304279327,
+      "learning_rate": 0.00014765100671140942,
+      "loss": 0.4246,
+      "step": 317
+    },
+    {
+      "epoch": 0.795186371805892,
+      "grad_norm": 0.17613396048545837,
+      "learning_rate": 0.00014748322147651008,
+      "loss": 0.432,
+      "step": 318
+    },
+    {
+      "epoch": 0.7976869578807533,
+      "grad_norm": 0.17050482332706451,
+      "learning_rate": 0.00014731543624161076,
+      "loss": 0.435,
+      "step": 319
+    },
+    {
+      "epoch": 0.8001875439556146,
+      "grad_norm": 0.16121627390384674,
+      "learning_rate": 0.0001471476510067114,
+      "loss": 0.4171,
+      "step": 320
+    },
+    {
+      "epoch": 0.8026881300304759,
+      "grad_norm": 0.172186940908432,
+      "learning_rate": 0.0001469798657718121,
+      "loss": 0.4235,
+      "step": 321
+    },
+    {
+      "epoch": 0.8051887161053372,
+      "grad_norm": 0.16914747655391693,
+      "learning_rate": 0.00014681208053691277,
+      "loss": 0.4348,
+      "step": 322
+    },
+    {
+      "epoch": 0.8076893021801985,
+      "grad_norm": 0.16869330406188965,
+      "learning_rate": 0.00014664429530201342,
+      "loss": 0.4436,
+      "step": 323
+    },
+    {
+      "epoch": 0.8101898882550598,
+      "grad_norm": 0.16513273119926453,
+      "learning_rate": 0.0001464765100671141,
+      "loss": 0.4189,
+      "step": 324
+    },
+    {
+      "epoch": 0.812690474329921,
+      "grad_norm": 0.17720884084701538,
+      "learning_rate": 0.00014630872483221478,
+      "loss": 0.4351,
+      "step": 325
+    },
+    {
+      "epoch": 0.8151910604047824,
+      "grad_norm": 0.17662091553211212,
+      "learning_rate": 0.00014614093959731546,
+      "loss": 0.4325,
+      "step": 326
+    },
+    {
+      "epoch": 0.8176916464796437,
+      "grad_norm": 0.1817537397146225,
+      "learning_rate": 0.0001459731543624161,
+      "loss": 0.4587,
+      "step": 327
+    },
+    {
+      "epoch": 0.8201922325545049,
+      "grad_norm": 0.18581923842430115,
+      "learning_rate": 0.0001458053691275168,
+      "loss": 0.4397,
+      "step": 328
+    },
+    {
+      "epoch": 0.8226928186293663,
+      "grad_norm": 0.17366306483745575,
+      "learning_rate": 0.00014563758389261744,
+      "loss": 0.4343,
+      "step": 329
+    },
+    {
+      "epoch": 0.8251934047042275,
+      "grad_norm": 0.19753865897655487,
+      "learning_rate": 0.00014546979865771812,
+      "loss": 0.4268,
+      "step": 330
+    },
+    {
+      "epoch": 0.8276939907790889,
+      "grad_norm": 0.1656249463558197,
+      "learning_rate": 0.0001453020134228188,
+      "loss": 0.4319,
+      "step": 331
+    },
+    {
+      "epoch": 0.8301945768539502,
+      "grad_norm": 0.1635424792766571,
+      "learning_rate": 0.00014513422818791946,
+      "loss": 0.4461,
+      "step": 332
+    },
+    {
+      "epoch": 0.8326951629288114,
+      "grad_norm": 0.19888846576213837,
+      "learning_rate": 0.00014496644295302014,
+      "loss": 0.4129,
+      "step": 333
+    },
+    {
+      "epoch": 0.8351957490036728,
+      "grad_norm": 0.16208425164222717,
+      "learning_rate": 0.00014479865771812082,
+      "loss": 0.4337,
+      "step": 334
+    },
+    {
+      "epoch": 0.837696335078534,
+      "grad_norm": 0.19178390502929688,
+      "learning_rate": 0.0001446308724832215,
+      "loss": 0.4344,
+      "step": 335
+    },
+    {
+      "epoch": 0.8401969211533953,
+      "grad_norm": 0.17292825877666473,
+      "learning_rate": 0.00014446308724832215,
+      "loss": 0.4218,
+      "step": 336
+    },
+    {
+      "epoch": 0.8426975072282566,
+      "grad_norm": 0.1647634506225586,
+      "learning_rate": 0.00014429530201342283,
+      "loss": 0.443,
+      "step": 337
+    },
+    {
+      "epoch": 0.8451980933031179,
+      "grad_norm": 0.18745389580726624,
+      "learning_rate": 0.0001441275167785235,
+      "loss": 0.4292,
+      "step": 338
+    },
+    {
+      "epoch": 0.8476986793779792,
+      "grad_norm": 0.20261400938034058,
+      "learning_rate": 0.00014395973154362416,
+      "loss": 0.4407,
+      "step": 339
+    },
+    {
+      "epoch": 0.8501992654528405,
+      "grad_norm": 0.16733746230602264,
+      "learning_rate": 0.00014379194630872484,
+      "loss": 0.432,
+      "step": 340
+    },
+    {
+      "epoch": 0.8526998515277018,
+      "grad_norm": 0.20043610036373138,
+      "learning_rate": 0.0001436241610738255,
+      "loss": 0.4403,
+      "step": 341
+    },
+    {
+      "epoch": 0.8552004376025631,
+      "grad_norm": 0.18660157918930054,
+      "learning_rate": 0.00014345637583892617,
+      "loss": 0.4177,
+      "step": 342
+    },
+    {
+      "epoch": 0.8577010236774244,
+      "grad_norm": 0.1695999950170517,
+      "learning_rate": 0.00014328859060402685,
+      "loss": 0.4173,
+      "step": 343
+    },
+    {
+      "epoch": 0.8602016097522857,
+      "grad_norm": 0.19332537055015564,
+      "learning_rate": 0.00014312080536912753,
+      "loss": 0.4196,
+      "step": 344
+    },
+    {
+      "epoch": 0.862702195827147,
+      "grad_norm": 0.17062701284885406,
+      "learning_rate": 0.00014295302013422819,
+      "loss": 0.4224,
+      "step": 345
+    },
+    {
+      "epoch": 0.8652027819020083,
+      "grad_norm": 0.16863779723644257,
+      "learning_rate": 0.00014278523489932887,
+      "loss": 0.432,
+      "step": 346
+    },
+    {
+      "epoch": 0.8677033679768695,
+      "grad_norm": 0.19555053114891052,
+      "learning_rate": 0.00014261744966442955,
+      "loss": 0.4361,
+      "step": 347
+    },
+    {
+      "epoch": 0.8702039540517309,
+      "grad_norm": 0.1860707402229309,
+      "learning_rate": 0.0001424496644295302,
+      "loss": 0.43,
+      "step": 348
+    },
+    {
+      "epoch": 0.8727045401265922,
+      "grad_norm": 0.16459205746650696,
+      "learning_rate": 0.00014228187919463088,
+      "loss": 0.4459,
+      "step": 349
+    },
+    {
+      "epoch": 0.8752051262014535,
+      "grad_norm": 0.21114307641983032,
+      "learning_rate": 0.00014211409395973153,
+      "loss": 0.4382,
+      "step": 350
+    },
+    {
+      "epoch": 0.8777057122763148,
+      "grad_norm": 0.1811704784631729,
+      "learning_rate": 0.0001419463087248322,
+      "loss": 0.4303,
+      "step": 351
+    },
+    {
+      "epoch": 0.880206298351176,
+      "grad_norm": 0.20086246728897095,
+      "learning_rate": 0.0001417785234899329,
+      "loss": 0.413,
+      "step": 352
+    },
+    {
+      "epoch": 0.8827068844260374,
+      "grad_norm": 0.17836225032806396,
+      "learning_rate": 0.00014161073825503357,
+      "loss": 0.4315,
+      "step": 353
+    },
+    {
+      "epoch": 0.8852074705008987,
+      "grad_norm": 0.1710200011730194,
+      "learning_rate": 0.00014144295302013425,
+      "loss": 0.4331,
+      "step": 354
+    },
+    {
+      "epoch": 0.8877080565757599,
+      "grad_norm": 0.17866557836532593,
+      "learning_rate": 0.0001412751677852349,
+      "loss": 0.4203,
+      "step": 355
+    },
+    {
+      "epoch": 0.8902086426506213,
+      "grad_norm": 0.19929209351539612,
+      "learning_rate": 0.00014110738255033558,
+      "loss": 0.4251,
+      "step": 356
+    },
+    {
+      "epoch": 0.8927092287254825,
+      "grad_norm": 0.1633501946926117,
+      "learning_rate": 0.00014093959731543624,
+      "loss": 0.4168,
+      "step": 357
+    },
+    {
+      "epoch": 0.8952098148003438,
+      "grad_norm": 0.19982637465000153,
+      "learning_rate": 0.00014077181208053691,
+      "loss": 0.4279,
+      "step": 358
+    },
+    {
+      "epoch": 0.8977104008752052,
+      "grad_norm": 0.2030080407857895,
+      "learning_rate": 0.0001406040268456376,
+      "loss": 0.4565,
+      "step": 359
+    },
+    {
+      "epoch": 0.9002109869500664,
+      "grad_norm": 0.17041054368019104,
+      "learning_rate": 0.00014043624161073827,
+      "loss": 0.447,
+      "step": 360
+    },
+    {
+      "epoch": 0.9027115730249278,
+      "grad_norm": 0.19890151917934418,
+      "learning_rate": 0.00014026845637583895,
+      "loss": 0.4212,
+      "step": 361
+    },
+    {
+      "epoch": 0.905212159099789,
+      "grad_norm": 0.21048621833324432,
+      "learning_rate": 0.0001401006711409396,
+      "loss": 0.4271,
+      "step": 362
+    },
+    {
+      "epoch": 0.9077127451746503,
+      "grad_norm": 0.16634558141231537,
+      "learning_rate": 0.0001399328859060403,
+      "loss": 0.4408,
+      "step": 363
+    },
+    {
+      "epoch": 0.9102133312495116,
+      "grad_norm": 0.2424069195985794,
+      "learning_rate": 0.00013976510067114094,
+      "loss": 0.4445,
+      "step": 364
+    },
+    {
+      "epoch": 0.9127139173243729,
+      "grad_norm": 0.16701576113700867,
+      "learning_rate": 0.00013959731543624162,
+      "loss": 0.4203,
+      "step": 365
+    },
+    {
+      "epoch": 0.9152145033992342,
+      "grad_norm": 0.17247918248176575,
+      "learning_rate": 0.00013942953020134227,
+      "loss": 0.4283,
+      "step": 366
+    },
+    {
+      "epoch": 0.9177150894740955,
+      "grad_norm": 0.19492937624454498,
+      "learning_rate": 0.00013926174496644295,
+      "loss": 0.4363,
+      "step": 367
+    },
+    {
+      "epoch": 0.9202156755489568,
+      "grad_norm": 0.16968198120594025,
+      "learning_rate": 0.00013909395973154363,
+      "loss": 0.4157,
+      "step": 368
+    },
+    {
+      "epoch": 0.922716261623818,
+      "grad_norm": 0.15889081358909607,
+      "learning_rate": 0.0001389261744966443,
+      "loss": 0.4557,
+      "step": 369
+    },
+    {
+      "epoch": 0.9252168476986794,
+      "grad_norm": 0.20287366211414337,
+      "learning_rate": 0.000138758389261745,
+      "loss": 0.4316,
+      "step": 370
+    },
+    {
+      "epoch": 0.9277174337735407,
+      "grad_norm": 0.18467019498348236,
+      "learning_rate": 0.00013859060402684564,
+      "loss": 0.4193,
+      "step": 371
+    },
+    {
+      "epoch": 0.930218019848402,
+      "grad_norm": 0.16602379083633423,
+      "learning_rate": 0.00013842281879194632,
+      "loss": 0.4291,
+      "step": 372
+    },
+    {
+      "epoch": 0.9327186059232633,
+      "grad_norm": 0.19791395962238312,
+      "learning_rate": 0.00013825503355704698,
+      "loss": 0.4242,
+      "step": 373
+    },
+    {
+      "epoch": 0.9352191919981245,
+      "grad_norm": 0.17725242674350739,
+      "learning_rate": 0.00013808724832214766,
+      "loss": 0.4268,
+      "step": 374
+    },
+    {
+      "epoch": 0.9377197780729859,
+      "grad_norm": 0.16493122279644012,
+      "learning_rate": 0.00013791946308724834,
+      "loss": 0.4289,
+      "step": 375
+    },
+    {
+      "epoch": 0.9402203641478472,
+      "grad_norm": 0.1670459806919098,
+      "learning_rate": 0.000137751677852349,
+      "loss": 0.4393,
+      "step": 376
+    },
+    {
+      "epoch": 0.9427209502227084,
+      "grad_norm": 0.19634924829006195,
+      "learning_rate": 0.00013758389261744967,
+      "loss": 0.4224,
+      "step": 377
+    },
+    {
+      "epoch": 0.9452215362975698,
+      "grad_norm": 0.1566496044397354,
+      "learning_rate": 0.00013741610738255035,
+      "loss": 0.4306,
+      "step": 378
+    },
+    {
+      "epoch": 0.947722122372431,
+      "grad_norm": 0.1697397530078888,
+      "learning_rate": 0.00013724832214765103,
+      "loss": 0.4159,
+      "step": 379
+    },
+    {
+      "epoch": 0.9502227084472923,
+      "grad_norm": 0.18244680762290955,
+      "learning_rate": 0.00013708053691275168,
+      "loss": 0.4325,
+      "step": 380
+    },
+    {
+      "epoch": 0.9527232945221537,
+      "grad_norm": 0.16451160609722137,
+      "learning_rate": 0.00013691275167785236,
+      "loss": 0.4224,
+      "step": 381
+    },
+    {
+      "epoch": 0.9552238805970149,
+      "grad_norm": 0.20068246126174927,
+      "learning_rate": 0.000136744966442953,
+      "loss": 0.4217,
+      "step": 382
+    },
+    {
+      "epoch": 0.9577244666718763,
+      "grad_norm": 0.19040897488594055,
+      "learning_rate": 0.0001365771812080537,
+      "loss": 0.4266,
+      "step": 383
+    },
+    {
+      "epoch": 0.9602250527467375,
+      "grad_norm": 0.16136051714420319,
+      "learning_rate": 0.00013640939597315437,
+      "loss": 0.4267,
+      "step": 384
+    },
+    {
+      "epoch": 0.9627256388215988,
+      "grad_norm": 0.1725684553384781,
+      "learning_rate": 0.00013624161073825503,
+      "loss": 0.4385,
+      "step": 385
+    },
+    {
+      "epoch": 0.9652262248964601,
+      "grad_norm": 0.2096080183982849,
+      "learning_rate": 0.0001360738255033557,
+      "loss": 0.4163,
+      "step": 386
+    },
+    {
+      "epoch": 0.9677268109713214,
+      "grad_norm": 0.16560566425323486,
+      "learning_rate": 0.00013590604026845639,
+      "loss": 0.4403,
+      "step": 387
+    },
+    {
+      "epoch": 0.9702273970461827,
+      "grad_norm": 0.18897269666194916,
+      "learning_rate": 0.00013573825503355706,
+      "loss": 0.415,
+      "step": 388
+    },
+    {
+      "epoch": 0.972727983121044,
+      "grad_norm": 0.18220849335193634,
+      "learning_rate": 0.00013557046979865772,
+      "loss": 0.4281,
+      "step": 389
+    },
+    {
+      "epoch": 0.9752285691959053,
+      "grad_norm": 0.16022995114326477,
+      "learning_rate": 0.0001354026845637584,
+      "loss": 0.4101,
+      "step": 390
+    },
+    {
+      "epoch": 0.9777291552707665,
+      "grad_norm": 0.17846792936325073,
+      "learning_rate": 0.00013523489932885908,
+      "loss": 0.4399,
+      "step": 391
+    },
+    {
+      "epoch": 0.9802297413456279,
+      "grad_norm": 0.17017945647239685,
+      "learning_rate": 0.00013506711409395973,
+      "loss": 0.4176,
+      "step": 392
+    },
+    {
+      "epoch": 0.9827303274204892,
+      "grad_norm": 0.1695704609155655,
+      "learning_rate": 0.0001348993288590604,
+      "loss": 0.4421,
+      "step": 393
+    },
+    {
+      "epoch": 0.9852309134953505,
+      "grad_norm": 0.1747402846813202,
+      "learning_rate": 0.00013473154362416106,
+      "loss": 0.4268,
+      "step": 394
+    },
+    {
+      "epoch": 0.9877314995702118,
+      "grad_norm": 0.1625954657793045,
+      "learning_rate": 0.00013456375838926174,
+      "loss": 0.4095,
+      "step": 395
+    },
+    {
+      "epoch": 0.990232085645073,
+      "grad_norm": 0.1669810563325882,
+      "learning_rate": 0.00013439597315436242,
+      "loss": 0.4258,
+      "step": 396
+    },
+    {
+      "epoch": 0.9927326717199344,
+      "grad_norm": 0.1578301340341568,
+      "learning_rate": 0.0001342281879194631,
+      "loss": 0.4155,
+      "step": 397
+    },
+    {
+      "epoch": 0.9952332577947957,
+      "grad_norm": 0.17688880860805511,
+      "learning_rate": 0.00013406040268456378,
+      "loss": 0.4338,
+      "step": 398
+    },
+    {
+      "epoch": 0.9977338438696569,
+      "grad_norm": 0.18112221360206604,
+      "learning_rate": 0.00013389261744966443,
+      "loss": 0.4057,
+      "step": 399
+    },
+    {
+      "epoch": 1.0002344299445183,
+      "grad_norm": 0.1682415008544922,
+      "learning_rate": 0.00013372483221476511,
+      "loss": 0.4413,
+      "step": 400
+    },
+    {
+      "epoch": 1.0027350160193795,
+      "grad_norm": 0.17804764211177826,
+      "learning_rate": 0.00013355704697986577,
+      "loss": 0.402,
+      "step": 401
+    },
+    {
+      "epoch": 1.0052356020942408,
+      "grad_norm": 0.17376191914081573,
+      "learning_rate": 0.00013338926174496645,
+      "loss": 0.4052,
+      "step": 402
+    },
+    {
+      "epoch": 1.007736188169102,
+      "grad_norm": 0.17517007887363434,
+      "learning_rate": 0.0001332214765100671,
+      "loss": 0.4019,
+      "step": 403
+    },
+    {
+      "epoch": 1.0102367742439635,
+      "grad_norm": 0.1827448606491089,
+      "learning_rate": 0.0001330536912751678,
+      "loss": 0.3953,
+      "step": 404
+    },
+    {
+      "epoch": 1.0127373603188248,
+      "grad_norm": 0.16910028457641602,
+      "learning_rate": 0.00013288590604026846,
+      "loss": 0.3991,
+      "step": 405
+    },
+    {
+      "epoch": 1.015237946393686,
+      "grad_norm": 0.18308071792125702,
+      "learning_rate": 0.00013271812080536914,
+      "loss": 0.3878,
+      "step": 406
+    },
+    {
+      "epoch": 1.0177385324685473,
+      "grad_norm": 0.20386648178100586,
+      "learning_rate": 0.00013255033557046982,
+      "loss": 0.402,
+      "step": 407
+    },
+    {
+      "epoch": 1.0202391185434085,
+      "grad_norm": 0.16715078055858612,
+      "learning_rate": 0.00013238255033557047,
+      "loss": 0.4025,
+      "step": 408
+    },
+    {
+      "epoch": 1.02273970461827,
+      "grad_norm": 0.18801377713680267,
+      "learning_rate": 0.00013221476510067115,
+      "loss": 0.4062,
+      "step": 409
+    },
+    {
+      "epoch": 1.0252402906931313,
+      "grad_norm": 0.1752709448337555,
+      "learning_rate": 0.0001320469798657718,
+      "loss": 0.4003,
+      "step": 410
+    },
+    {
+      "epoch": 1.0277408767679925,
+      "grad_norm": 0.16399900615215302,
+      "learning_rate": 0.00013187919463087248,
+      "loss": 0.396,
+      "step": 411
+    },
+    {
+      "epoch": 1.0302414628428538,
+      "grad_norm": 0.2031196504831314,
+      "learning_rate": 0.00013171140939597316,
+      "loss": 0.4086,
+      "step": 412
+    },
+    {
+      "epoch": 1.032742048917715,
+      "grad_norm": 0.17596857249736786,
+      "learning_rate": 0.00013154362416107384,
+      "loss": 0.394,
+      "step": 413
+    },
+    {
+      "epoch": 1.0352426349925763,
+      "grad_norm": 0.1620572954416275,
+      "learning_rate": 0.00013137583892617452,
+      "loss": 0.3983,
+      "step": 414
+    },
+    {
+      "epoch": 1.0377432210674378,
+      "grad_norm": 0.1825663447380066,
+      "learning_rate": 0.00013120805369127518,
+      "loss": 0.4103,
+      "step": 415
+    },
+    {
+      "epoch": 1.040243807142299,
+      "grad_norm": 0.19497060775756836,
+      "learning_rate": 0.00013104026845637586,
+      "loss": 0.3989,
+      "step": 416
+    },
+    {
+      "epoch": 1.0427443932171603,
+      "grad_norm": 0.1783512830734253,
+      "learning_rate": 0.0001308724832214765,
+      "loss": 0.3952,
+      "step": 417
+    },
+    {
+      "epoch": 1.0452449792920215,
+      "grad_norm": 0.17473211884498596,
+      "learning_rate": 0.0001307046979865772,
+      "loss": 0.3901,
+      "step": 418
+    },
+    {
+      "epoch": 1.0477455653668828,
+      "grad_norm": 0.1653796285390854,
+      "learning_rate": 0.00013053691275167784,
+      "loss": 0.38,
+      "step": 419
+    },
+    {
+      "epoch": 1.050246151441744,
+      "grad_norm": 0.17586909234523773,
+      "learning_rate": 0.00013036912751677852,
+      "loss": 0.4026,
+      "step": 420
+    },
+    {
+      "epoch": 1.0527467375166055,
+      "grad_norm": 0.16513925790786743,
+      "learning_rate": 0.0001302013422818792,
+      "loss": 0.3922,
+      "step": 421
+    },
+    {
+      "epoch": 1.0552473235914668,
+      "grad_norm": 0.20568743348121643,
+      "learning_rate": 0.00013003355704697988,
+      "loss": 0.3996,
+      "step": 422
+    },
+    {
+      "epoch": 1.057747909666328,
+      "grad_norm": 0.1725350171327591,
+      "learning_rate": 0.00012986577181208056,
+      "loss": 0.3912,
+      "step": 423
+    },
+    {
+      "epoch": 1.0602484957411893,
+      "grad_norm": 0.17537754774093628,
+      "learning_rate": 0.0001296979865771812,
+      "loss": 0.3893,
+      "step": 424
+    },
+    {
+      "epoch": 1.0627490818160505,
+      "grad_norm": 0.17607562243938446,
+      "learning_rate": 0.0001295302013422819,
+      "loss": 0.3894,
+      "step": 425
+    },
+    {
+      "epoch": 1.065249667890912,
+      "grad_norm": 0.16470113396644592,
+      "learning_rate": 0.00012936241610738254,
+      "loss": 0.391,
+      "step": 426
+    },
+    {
+      "epoch": 1.0677502539657733,
+      "grad_norm": 0.17240354418754578,
+      "learning_rate": 0.00012919463087248322,
+      "loss": 0.392,
+      "step": 427
+    },
+    {
+      "epoch": 1.0702508400406345,
+      "grad_norm": 0.17529991269111633,
+      "learning_rate": 0.0001290268456375839,
+      "loss": 0.3953,
+      "step": 428
+    },
+    {
+      "epoch": 1.0727514261154958,
+      "grad_norm": 0.17728881537914276,
+      "learning_rate": 0.00012885906040268456,
+      "loss": 0.3818,
+      "step": 429
+    },
+    {
+      "epoch": 1.075252012190357,
+      "grad_norm": 0.17223317921161652,
+      "learning_rate": 0.00012869127516778524,
+      "loss": 0.3825,
+      "step": 430
+    },
+    {
+      "epoch": 1.0777525982652185,
+      "grad_norm": 0.18325600028038025,
+      "learning_rate": 0.00012852348993288592,
+      "loss": 0.3927,
+      "step": 431
+    },
+    {
+      "epoch": 1.0802531843400798,
+      "grad_norm": 0.17309826612472534,
+      "learning_rate": 0.0001283557046979866,
+      "loss": 0.3942,
+      "step": 432
+    },
+    {
+      "epoch": 1.082753770414941,
+      "grad_norm": 0.1761365830898285,
+      "learning_rate": 0.00012818791946308725,
+      "loss": 0.3958,
+      "step": 433
+    },
+    {
+      "epoch": 1.0852543564898023,
+      "grad_norm": 0.18157604336738586,
+      "learning_rate": 0.00012802013422818793,
+      "loss": 0.3894,
+      "step": 434
+    },
+    {
+      "epoch": 1.0877549425646635,
+      "grad_norm": 0.17641478776931763,
+      "learning_rate": 0.0001278523489932886,
+      "loss": 0.3872,
+      "step": 435
+    },
+    {
+      "epoch": 1.0902555286395248,
+      "grad_norm": 0.1755024641752243,
+      "learning_rate": 0.00012768456375838926,
+      "loss": 0.4044,
+      "step": 436
+    },
+    {
+      "epoch": 1.0927561147143863,
+      "grad_norm": 0.17728440463542938,
+      "learning_rate": 0.00012751677852348994,
+      "loss": 0.4024,
+      "step": 437
+    },
+    {
+      "epoch": 1.0952567007892475,
+      "grad_norm": 0.18590691685676575,
+      "learning_rate": 0.0001273489932885906,
+      "loss": 0.4054,
+      "step": 438
+    },
+    {
+      "epoch": 1.0977572868641088,
+      "grad_norm": 0.16991455852985382,
+      "learning_rate": 0.0001271812080536913,
+      "loss": 0.3892,
+      "step": 439
+    },
+    {
+      "epoch": 1.10025787293897,
+      "grad_norm": 0.18331153690814972,
+      "learning_rate": 0.00012701342281879195,
+      "loss": 0.3946,
+      "step": 440
+    },
+    {
+      "epoch": 1.1027584590138313,
+      "grad_norm": 0.17147310078144073,
+      "learning_rate": 0.00012684563758389263,
+      "loss": 0.3841,
+      "step": 441
+    },
+    {
+      "epoch": 1.1052590450886925,
+      "grad_norm": 0.1769578754901886,
+      "learning_rate": 0.00012667785234899329,
+      "loss": 0.4014,
+      "step": 442
+    },
+    {
+      "epoch": 1.107759631163554,
+      "grad_norm": 0.1756420135498047,
+      "learning_rate": 0.00012651006711409397,
+      "loss": 0.4177,
+      "step": 443
+    },
+    {
+      "epoch": 1.1102602172384153,
+      "grad_norm": 0.1766512542963028,
+      "learning_rate": 0.00012634228187919465,
+      "loss": 0.3983,
+      "step": 444
+    },
+    {
+      "epoch": 1.1127608033132765,
+      "grad_norm": 0.16898678243160248,
+      "learning_rate": 0.0001261744966442953,
+      "loss": 0.3915,
+      "step": 445
+    },
+    {
+      "epoch": 1.1152613893881378,
+      "grad_norm": 0.17253310978412628,
+      "learning_rate": 0.00012600671140939598,
+      "loss": 0.3954,
+      "step": 446
+    },
+    {
+      "epoch": 1.117761975462999,
+      "grad_norm": 0.17454257607460022,
+      "learning_rate": 0.00012583892617449663,
+      "loss": 0.3984,
+      "step": 447
+    },
+    {
+      "epoch": 1.1202625615378605,
+      "grad_norm": 0.17149844765663147,
+      "learning_rate": 0.00012567114093959734,
+      "loss": 0.3912,
+      "step": 448
+    },
+    {
+      "epoch": 1.1227631476127218,
+      "grad_norm": 0.17389146983623505,
+      "learning_rate": 0.000125503355704698,
+      "loss": 0.4149,
+      "step": 449
+    },
+    {
+      "epoch": 1.125263733687583,
+      "grad_norm": 0.16737361252307892,
+      "learning_rate": 0.00012533557046979867,
+      "loss": 0.3889,
+      "step": 450
+    },
+    {
+      "epoch": 1.1277643197624443,
+      "grad_norm": 0.1793391853570938,
+      "learning_rate": 0.00012516778523489935,
+      "loss": 0.4098,
+      "step": 451
+    },
+    {
+      "epoch": 1.1302649058373055,
+      "grad_norm": 0.18140512704849243,
+      "learning_rate": 0.000125,
+      "loss": 0.4167,
+      "step": 452
+    },
+    {
+      "epoch": 1.132765491912167,
+      "grad_norm": 0.16896022856235504,
+      "learning_rate": 0.00012483221476510068,
+      "loss": 0.3959,
+      "step": 453
+    },
+    {
+      "epoch": 1.1352660779870283,
+      "grad_norm": 0.18008409440517426,
+      "learning_rate": 0.00012466442953020134,
+      "loss": 0.3892,
+      "step": 454
+    },
+    {
+      "epoch": 1.1377666640618895,
+      "grad_norm": 0.17115534842014313,
+      "learning_rate": 0.00012449664429530201,
+      "loss": 0.3937,
+      "step": 455
+    },
+    {
+      "epoch": 1.1402672501367508,
+      "grad_norm": 0.18967628479003906,
+      "learning_rate": 0.00012432885906040267,
+      "loss": 0.3992,
+      "step": 456
+    },
+    {
+      "epoch": 1.142767836211612,
+      "grad_norm": 0.18216951191425323,
+      "learning_rate": 0.00012416107382550337,
+      "loss": 0.4089,
+      "step": 457
+    },
+    {
+      "epoch": 1.1452684222864733,
+      "grad_norm": 0.19067801535129547,
+      "learning_rate": 0.00012399328859060403,
+      "loss": 0.3811,
+      "step": 458
+    },
+    {
+      "epoch": 1.1477690083613348,
+      "grad_norm": 0.17509539425373077,
+      "learning_rate": 0.0001238255033557047,
+      "loss": 0.3821,
+      "step": 459
+    },
+    {
+      "epoch": 1.150269594436196,
+      "grad_norm": 0.16954122483730316,
+      "learning_rate": 0.0001236577181208054,
+      "loss": 0.3856,
+      "step": 460
+    },
+    {
+      "epoch": 1.1527701805110573,
+      "grad_norm": 0.19434818625450134,
+      "learning_rate": 0.00012348993288590604,
+      "loss": 0.3823,
+      "step": 461
+    },
+    {
+      "epoch": 1.1552707665859185,
+      "grad_norm": 0.1899290233850479,
+      "learning_rate": 0.00012332214765100672,
+      "loss": 0.3957,
+      "step": 462
+    },
+    {
+      "epoch": 1.1577713526607798,
+      "grad_norm": 0.1762360781431198,
+      "learning_rate": 0.00012315436241610737,
+      "loss": 0.3802,
+      "step": 463
+    },
+    {
+      "epoch": 1.160271938735641,
+      "grad_norm": 0.19524188339710236,
+      "learning_rate": 0.00012298657718120805,
+      "loss": 0.4001,
+      "step": 464
+    },
+    {
+      "epoch": 1.1627725248105025,
+      "grad_norm": 0.17824076116085052,
+      "learning_rate": 0.00012281879194630873,
+      "loss": 0.4001,
+      "step": 465
+    },
+    {
+      "epoch": 1.1652731108853638,
+      "grad_norm": 0.18208450078964233,
+      "learning_rate": 0.0001226510067114094,
+      "loss": 0.3961,
+      "step": 466
+    },
+    {
+      "epoch": 1.167773696960225,
+      "grad_norm": 0.17536821961402893,
+      "learning_rate": 0.0001224832214765101,
+      "loss": 0.3976,
+      "step": 467
+    },
+    {
+      "epoch": 1.1702742830350863,
+      "grad_norm": 0.17048102617263794,
+      "learning_rate": 0.00012231543624161074,
+      "loss": 0.3874,
+      "step": 468
+    },
+    {
+      "epoch": 1.1727748691099475,
+      "grad_norm": 0.186661958694458,
+      "learning_rate": 0.00012214765100671142,
+      "loss": 0.4077,
+      "step": 469
+    },
+    {
+      "epoch": 1.175275455184809,
+      "grad_norm": 0.17875522375106812,
+      "learning_rate": 0.00012197986577181209,
+      "loss": 0.3976,
+      "step": 470
+    },
+    {
+      "epoch": 1.1777760412596703,
+      "grad_norm": 0.17223358154296875,
+      "learning_rate": 0.00012181208053691276,
+      "loss": 0.3941,
+      "step": 471
+    },
+    {
+      "epoch": 1.1802766273345315,
+      "grad_norm": 0.1784411072731018,
+      "learning_rate": 0.00012164429530201342,
+      "loss": 0.3918,
+      "step": 472
+    },
+    {
+      "epoch": 1.1827772134093928,
+      "grad_norm": 0.176781564950943,
+      "learning_rate": 0.00012147651006711409,
+      "loss": 0.3879,
+      "step": 473
+    },
+    {
+      "epoch": 1.185277799484254,
+      "grad_norm": 0.18470625579357147,
+      "learning_rate": 0.00012130872483221478,
+      "loss": 0.3806,
+      "step": 474
+    },
+    {
+      "epoch": 1.1877783855591155,
+      "grad_norm": 0.1847718358039856,
+      "learning_rate": 0.00012114093959731545,
+      "loss": 0.3999,
+      "step": 475
+    },
+    {
+      "epoch": 1.1902789716339768,
+      "grad_norm": 0.18402062356472015,
+      "learning_rate": 0.00012097315436241611,
+      "loss": 0.3905,
+      "step": 476
+    },
+    {
+      "epoch": 1.192779557708838,
+      "grad_norm": 0.1812908798456192,
+      "learning_rate": 0.0001208053691275168,
+      "loss": 0.3821,
+      "step": 477
+    },
+    {
+      "epoch": 1.1952801437836993,
+      "grad_norm": 0.18111836910247803,
+      "learning_rate": 0.00012063758389261746,
+      "loss": 0.3902,
+      "step": 478
+    },
+    {
+      "epoch": 1.1977807298585605,
+      "grad_norm": 0.17667655646800995,
+      "learning_rate": 0.00012046979865771813,
+      "loss": 0.3829,
+      "step": 479
+    },
+    {
+      "epoch": 1.2002813159334218,
+      "grad_norm": 0.17511925101280212,
+      "learning_rate": 0.00012030201342281879,
+      "loss": 0.3882,
+      "step": 480
+    },
+    {
+      "epoch": 1.2027819020082833,
+      "grad_norm": 0.1772802323102951,
+      "learning_rate": 0.00012013422818791946,
+      "loss": 0.3963,
+      "step": 481
+    },
+    {
+      "epoch": 1.2052824880831445,
+      "grad_norm": 0.17284731566905975,
+      "learning_rate": 0.00011996644295302013,
+      "loss": 0.3918,
+      "step": 482
+    },
+    {
+      "epoch": 1.2077830741580058,
+      "grad_norm": 0.18320755660533905,
+      "learning_rate": 0.00011979865771812082,
+      "loss": 0.4003,
+      "step": 483
+    },
+    {
+      "epoch": 1.210283660232867,
+      "grad_norm": 0.1786193549633026,
+      "learning_rate": 0.00011963087248322149,
+      "loss": 0.4007,
+      "step": 484
+    },
+    {
+      "epoch": 1.2127842463077283,
+      "grad_norm": 0.17019547522068024,
+      "learning_rate": 0.00011946308724832216,
+      "loss": 0.3973,
+      "step": 485
+    },
+    {
+      "epoch": 1.2152848323825896,
+      "grad_norm": 0.187973752617836,
+      "learning_rate": 0.00011929530201342283,
+      "loss": 0.4242,
+      "step": 486
+    },
+    {
+      "epoch": 1.217785418457451,
+      "grad_norm": 0.17459571361541748,
+      "learning_rate": 0.0001191275167785235,
+      "loss": 0.3957,
+      "step": 487
+    },
+    {
+      "epoch": 1.2202860045323123,
+      "grad_norm": 0.186808243393898,
+      "learning_rate": 0.00011895973154362416,
+      "loss": 0.3809,
+      "step": 488
+    },
+    {
+      "epoch": 1.2227865906071735,
+      "grad_norm": 0.1886894702911377,
+      "learning_rate": 0.00011879194630872483,
+      "loss": 0.3999,
+      "step": 489
+    },
+    {
+      "epoch": 1.2252871766820348,
+      "grad_norm": 0.17145392298698425,
+      "learning_rate": 0.0001186241610738255,
+      "loss": 0.3792,
+      "step": 490
+    },
+    {
+      "epoch": 1.227787762756896,
+      "grad_norm": 0.18122152984142303,
+      "learning_rate": 0.00011845637583892618,
+      "loss": 0.3954,
+      "step": 491
+    },
+    {
+      "epoch": 1.2302883488317575,
+      "grad_norm": 0.194286048412323,
+      "learning_rate": 0.00011828859060402686,
+      "loss": 0.3825,
+      "step": 492
+    },
+    {
+      "epoch": 1.2327889349066188,
+      "grad_norm": 0.17431862652301788,
+      "learning_rate": 0.00011812080536912754,
+      "loss": 0.3871,
+      "step": 493
+    },
+    {
+      "epoch": 1.23528952098148,
+      "grad_norm": 0.19097591936588287,
+      "learning_rate": 0.0001179530201342282,
+      "loss": 0.4018,
+      "step": 494
+    },
+    {
+      "epoch": 1.2377901070563413,
+      "grad_norm": 0.18145328760147095,
+      "learning_rate": 0.00011778523489932887,
+      "loss": 0.3874,
+      "step": 495
+    },
+    {
+      "epoch": 1.2402906931312025,
+      "grad_norm": 0.1778709590435028,
+      "learning_rate": 0.00011761744966442953,
+      "loss": 0.3955,
+      "step": 496
+    },
+    {
+      "epoch": 1.242791279206064,
+      "grad_norm": 0.20905902981758118,
+      "learning_rate": 0.0001174496644295302,
+      "loss": 0.398,
+      "step": 497
+    },
+    {
+      "epoch": 1.2452918652809253,
+      "grad_norm": 0.19288387894630432,
+      "learning_rate": 0.00011728187919463087,
+      "loss": 0.4031,
+      "step": 498
+    },
+    {
+      "epoch": 1.2477924513557865,
+      "grad_norm": 0.19638553261756897,
+      "learning_rate": 0.00011711409395973155,
+      "loss": 0.3967,
+      "step": 499
+    },
+    {
+      "epoch": 1.2502930374306478,
+      "grad_norm": 0.18342918157577515,
+      "learning_rate": 0.00011694630872483221,
+      "loss": 0.3844,
+      "step": 500
+    },
+    {
+      "epoch": 1.252793623505509,
+      "grad_norm": 0.17720724642276764,
+      "learning_rate": 0.0001167785234899329,
+      "loss": 0.3995,
+      "step": 501
+    },
+    {
+      "epoch": 1.2552942095803705,
+      "grad_norm": 0.1841309517621994,
+      "learning_rate": 0.00011661073825503357,
+      "loss": 0.3804,
+      "step": 502
+    },
+    {
+      "epoch": 1.2577947956552316,
+      "grad_norm": 0.19018958508968353,
+      "learning_rate": 0.00011644295302013424,
+      "loss": 0.368,
+      "step": 503
+    },
+    {
+      "epoch": 1.260295381730093,
+      "grad_norm": 0.17050069570541382,
+      "learning_rate": 0.0001162751677852349,
+      "loss": 0.3928,
+      "step": 504
+    },
+    {
+      "epoch": 1.2627959678049543,
+      "grad_norm": 0.19299225509166718,
+      "learning_rate": 0.00011610738255033557,
+      "loss": 0.3893,
+      "step": 505
+    },
+    {
+      "epoch": 1.2652965538798155,
+      "grad_norm": 0.18787892162799835,
+      "learning_rate": 0.00011593959731543624,
+      "loss": 0.3891,
+      "step": 506
+    },
+    {
+      "epoch": 1.2677971399546768,
+      "grad_norm": 0.1820453405380249,
+      "learning_rate": 0.00011577181208053692,
+      "loss": 0.397,
+      "step": 507
+    },
+    {
+      "epoch": 1.270297726029538,
+      "grad_norm": 0.18383005261421204,
+      "learning_rate": 0.00011560402684563758,
+      "loss": 0.3894,
+      "step": 508
+    },
+    {
+      "epoch": 1.2727983121043995,
+      "grad_norm": 0.20577560365200043,
+      "learning_rate": 0.00011543624161073828,
+      "loss": 0.4075,
+      "step": 509
+    },
+    {
+      "epoch": 1.2752988981792608,
+      "grad_norm": 0.18062449991703033,
+      "learning_rate": 0.00011526845637583894,
+      "loss": 0.4046,
+      "step": 510
+    },
+    {
+      "epoch": 1.277799484254122,
+      "grad_norm": 0.20840467512607574,
+      "learning_rate": 0.00011510067114093961,
+      "loss": 0.3876,
+      "step": 511
+    },
+    {
+      "epoch": 1.2803000703289833,
+      "grad_norm": 0.2160758078098297,
+      "learning_rate": 0.00011493288590604028,
+      "loss": 0.3972,
+      "step": 512
+    },
+    {
+      "epoch": 1.2828006564038446,
+      "grad_norm": 0.17154747247695923,
+      "learning_rate": 0.00011476510067114094,
+      "loss": 0.3875,
+      "step": 513
+    },
+    {
+      "epoch": 1.285301242478706,
+      "grad_norm": 0.18271404504776,
+      "learning_rate": 0.00011459731543624161,
+      "loss": 0.3903,
+      "step": 514
+    },
+    {
+      "epoch": 1.2878018285535673,
+      "grad_norm": 0.2114219218492508,
+      "learning_rate": 0.00011442953020134229,
+      "loss": 0.3846,
+      "step": 515
+    },
+    {
+      "epoch": 1.2903024146284285,
+      "grad_norm": 0.17632241547107697,
+      "learning_rate": 0.00011426174496644295,
+      "loss": 0.382,
+      "step": 516
+    },
+    {
+      "epoch": 1.2928030007032898,
+      "grad_norm": 0.18310467898845673,
+      "learning_rate": 0.00011409395973154362,
+      "loss": 0.3818,
+      "step": 517
+    },
+    {
+      "epoch": 1.295303586778151,
+      "grad_norm": 0.19026687741279602,
+      "learning_rate": 0.00011392617449664431,
+      "loss": 0.4004,
+      "step": 518
+    },
+    {
+      "epoch": 1.2978041728530125,
+      "grad_norm": 0.17775025963783264,
+      "learning_rate": 0.00011375838926174498,
+      "loss": 0.4029,
+      "step": 519
+    },
+    {
+      "epoch": 1.3003047589278738,
+      "grad_norm": 0.18015576899051666,
+      "learning_rate": 0.00011359060402684565,
+      "loss": 0.394,
+      "step": 520
+    },
+    {
+      "epoch": 1.302805345002735,
+      "grad_norm": 0.19737471640110016,
+      "learning_rate": 0.00011342281879194631,
+      "loss": 0.3799,
+      "step": 521
+    },
+    {
+      "epoch": 1.3053059310775963,
+      "grad_norm": 0.19162492454051971,
+      "learning_rate": 0.00011325503355704699,
+      "loss": 0.3965,
+      "step": 522
+    },
+    {
+      "epoch": 1.3078065171524575,
+      "grad_norm": 0.1779211312532425,
+      "learning_rate": 0.00011308724832214766,
+      "loss": 0.3967,
+      "step": 523
+    },
+    {
+      "epoch": 1.310307103227319,
+      "grad_norm": 0.2138555943965912,
+      "learning_rate": 0.00011291946308724832,
+      "loss": 0.3836,
+      "step": 524
+    },
+    {
+      "epoch": 1.3128076893021803,
+      "grad_norm": 0.2159447968006134,
+      "learning_rate": 0.00011275167785234899,
+      "loss": 0.3878,
+      "step": 525
+    },
+    {
+      "epoch": 1.3153082753770415,
+      "grad_norm": 0.17350620031356812,
+      "learning_rate": 0.00011258389261744966,
+      "loss": 0.3753,
+      "step": 526
+    },
+    {
+      "epoch": 1.3178088614519028,
+      "grad_norm": 0.2060236930847168,
+      "learning_rate": 0.00011241610738255035,
+      "loss": 0.3916,
+      "step": 527
+    },
+    {
+      "epoch": 1.320309447526764,
+      "grad_norm": 0.19645291566848755,
+      "learning_rate": 0.00011224832214765102,
+      "loss": 0.3984,
+      "step": 528
+    },
+    {
+      "epoch": 1.3228100336016253,
+      "grad_norm": 0.17098011076450348,
+      "learning_rate": 0.00011208053691275168,
+      "loss": 0.3915,
+      "step": 529
+    },
+    {
+      "epoch": 1.3253106196764866,
+      "grad_norm": 0.18419700860977173,
+      "learning_rate": 0.00011191275167785236,
+      "loss": 0.3867,
+      "step": 530
+    },
+    {
+      "epoch": 1.327811205751348,
+      "grad_norm": 0.19081442058086395,
+      "learning_rate": 0.00011174496644295303,
+      "loss": 0.3921,
+      "step": 531
+    },
+    {
+      "epoch": 1.3303117918262093,
+      "grad_norm": 0.17477817833423615,
+      "learning_rate": 0.0001115771812080537,
+      "loss": 0.4023,
+      "step": 532
+    },
+    {
+      "epoch": 1.3328123779010705,
+      "grad_norm": 0.18040700256824493,
+      "learning_rate": 0.00011140939597315436,
+      "loss": 0.3882,
+      "step": 533
+    },
+    {
+      "epoch": 1.3353129639759318,
+      "grad_norm": 0.18590973317623138,
+      "learning_rate": 0.00011124161073825503,
+      "loss": 0.3889,
+      "step": 534
+    },
+    {
+      "epoch": 1.337813550050793,
+      "grad_norm": 0.1844896674156189,
+      "learning_rate": 0.0001110738255033557,
+      "loss": 0.399,
+      "step": 535
+    },
+    {
+      "epoch": 1.3403141361256545,
+      "grad_norm": 0.1925611048936844,
+      "learning_rate": 0.00011090604026845639,
+      "loss": 0.4048,
+      "step": 536
+    },
+    {
+      "epoch": 1.3428147222005158,
+      "grad_norm": 0.17830778658390045,
+      "learning_rate": 0.00011073825503355705,
+      "loss": 0.4006,
+      "step": 537
+    },
+    {
+      "epoch": 1.345315308275377,
+      "grad_norm": 0.1769605129957199,
+      "learning_rate": 0.00011057046979865773,
+      "loss": 0.4116,
+      "step": 538
+    },
+    {
+      "epoch": 1.3478158943502383,
+      "grad_norm": 0.18456001579761505,
+      "learning_rate": 0.0001104026845637584,
+      "loss": 0.3915,
+      "step": 539
+    },
+    {
+      "epoch": 1.3503164804250996,
+      "grad_norm": 0.17792654037475586,
+      "learning_rate": 0.00011023489932885907,
+      "loss": 0.3901,
+      "step": 540
+    },
+    {
+      "epoch": 1.352817066499961,
+      "grad_norm": 0.1855223923921585,
+      "learning_rate": 0.00011006711409395973,
+      "loss": 0.3928,
+      "step": 541
+    },
+    {
+      "epoch": 1.3553176525748223,
+      "grad_norm": 0.17973613739013672,
+      "learning_rate": 0.0001098993288590604,
+      "loss": 0.392,
+      "step": 542
+    },
+    {
+      "epoch": 1.3578182386496835,
+      "grad_norm": 0.17758318781852722,
+      "learning_rate": 0.00010973154362416106,
+      "loss": 0.3817,
+      "step": 543
+    },
+    {
+      "epoch": 1.3603188247245448,
+      "grad_norm": 0.169822558760643,
+      "learning_rate": 0.00010956375838926174,
+      "loss": 0.3931,
+      "step": 544
+    },
+    {
+      "epoch": 1.362819410799406,
+      "grad_norm": 0.18156425654888153,
+      "learning_rate": 0.00010939597315436242,
+      "loss": 0.3819,
+      "step": 545
+    },
+    {
+      "epoch": 1.3653199968742675,
+      "grad_norm": 0.18494713306427002,
+      "learning_rate": 0.0001092281879194631,
+      "loss": 0.3711,
+      "step": 546
+    },
+    {
+      "epoch": 1.3678205829491288,
+      "grad_norm": 0.1997014880180359,
+      "learning_rate": 0.00010906040268456377,
+      "loss": 0.3942,
+      "step": 547
+    },
+    {
+      "epoch": 1.37032116902399,
+      "grad_norm": 0.1784486621618271,
+      "learning_rate": 0.00010889261744966444,
+      "loss": 0.3913,
+      "step": 548
+    },
+    {
+      "epoch": 1.3728217550988513,
+      "grad_norm": 0.17925742268562317,
+      "learning_rate": 0.0001087248322147651,
+      "loss": 0.3935,
+      "step": 549
+    },
+    {
+      "epoch": 1.3753223411737125,
+      "grad_norm": 0.18862564861774445,
+      "learning_rate": 0.00010855704697986577,
+      "loss": 0.3997,
+      "step": 550
+    },
+    {
+      "epoch": 1.3778229272485738,
+      "grad_norm": 0.1790710985660553,
+      "learning_rate": 0.00010838926174496644,
+      "loss": 0.3956,
+      "step": 551
+    },
+    {
+      "epoch": 1.380323513323435,
+      "grad_norm": 0.1938549429178238,
+      "learning_rate": 0.00010822147651006711,
+      "loss": 0.3994,
+      "step": 552
+    },
+    {
+      "epoch": 1.3828240993982965,
+      "grad_norm": 0.18064497411251068,
+      "learning_rate": 0.0001080536912751678,
+      "loss": 0.3812,
+      "step": 553
+    },
+    {
+      "epoch": 1.3853246854731578,
+      "grad_norm": 0.1703406572341919,
+      "learning_rate": 0.00010788590604026847,
+      "loss": 0.3908,
+      "step": 554
+    },
+    {
+      "epoch": 1.387825271548019,
+      "grad_norm": 0.18674202263355255,
+      "learning_rate": 0.00010771812080536914,
+      "loss": 0.3776,
+      "step": 555
+    },
+    {
+      "epoch": 1.3903258576228803,
+      "grad_norm": 0.17181988060474396,
+      "learning_rate": 0.00010755033557046981,
+      "loss": 0.3789,
+      "step": 556
+    },
+    {
+      "epoch": 1.3928264436977416,
+      "grad_norm": 0.16567173600196838,
+      "learning_rate": 0.00010738255033557047,
+      "loss": 0.392,
+      "step": 557
+    },
+    {
+      "epoch": 1.395327029772603,
+      "grad_norm": 0.18388572335243225,
+      "learning_rate": 0.00010721476510067114,
+      "loss": 0.3918,
+      "step": 558
+    },
+    {
+      "epoch": 1.3978276158474643,
+      "grad_norm": 0.17347316443920135,
+      "learning_rate": 0.00010704697986577182,
+      "loss": 0.3831,
+      "step": 559
+    },
+    {
+      "epoch": 1.4003282019223255,
+      "grad_norm": 0.18346688151359558,
+      "learning_rate": 0.00010687919463087249,
+      "loss": 0.4114,
+      "step": 560
+    },
+    {
+      "epoch": 1.4028287879971868,
+      "grad_norm": 0.1893952488899231,
+      "learning_rate": 0.00010671140939597315,
+      "loss": 0.4171,
+      "step": 561
+    },
+    {
+      "epoch": 1.405329374072048,
+      "grad_norm": 0.1657368689775467,
+      "learning_rate": 0.00010654362416107385,
+      "loss": 0.3831,
+      "step": 562
+    },
+    {
+      "epoch": 1.4078299601469095,
+      "grad_norm": 0.17420844733715057,
+      "learning_rate": 0.00010637583892617451,
+      "loss": 0.383,
+      "step": 563
+    },
+    {
+      "epoch": 1.4103305462217708,
+      "grad_norm": 0.1941898614168167,
+      "learning_rate": 0.00010620805369127518,
+      "loss": 0.3835,
+      "step": 564
+    },
+    {
+      "epoch": 1.412831132296632,
+      "grad_norm": 0.1878691017627716,
+      "learning_rate": 0.00010604026845637584,
+      "loss": 0.3931,
+      "step": 565
+    },
+    {
+      "epoch": 1.4153317183714933,
+      "grad_norm": 0.1785835325717926,
+      "learning_rate": 0.00010587248322147651,
+      "loss": 0.3935,
+      "step": 566
+    },
+    {
+      "epoch": 1.4178323044463546,
+      "grad_norm": 0.1761426031589508,
+      "learning_rate": 0.00010570469798657719,
+      "loss": 0.392,
+      "step": 567
+    },
+    {
+      "epoch": 1.420332890521216,
+      "grad_norm": 0.1820986419916153,
+      "learning_rate": 0.00010553691275167786,
+      "loss": 0.3951,
+      "step": 568
+    },
+    {
+      "epoch": 1.4228334765960773,
+      "grad_norm": 0.17546477913856506,
+      "learning_rate": 0.00010536912751677852,
+      "loss": 0.3801,
+      "step": 569
+    },
+    {
+      "epoch": 1.4253340626709385,
+      "grad_norm": 0.16867369413375854,
+      "learning_rate": 0.00010520134228187919,
+      "loss": 0.3942,
+      "step": 570
+    },
+    {
+      "epoch": 1.4278346487457998,
+      "grad_norm": 0.17508560419082642,
+      "learning_rate": 0.00010503355704697988,
+      "loss": 0.3967,
+      "step": 571
+    },
+    {
+      "epoch": 1.430335234820661,
+      "grad_norm": 0.17600426077842712,
+      "learning_rate": 0.00010486577181208055,
+      "loss": 0.3842,
+      "step": 572
+    },
+    {
+      "epoch": 1.4328358208955223,
+      "grad_norm": 0.17705383896827698,
+      "learning_rate": 0.00010469798657718121,
+      "loss": 0.3882,
+      "step": 573
+    },
+    {
+      "epoch": 1.4353364069703836,
+      "grad_norm": 0.16688145697116852,
+      "learning_rate": 0.00010453020134228188,
+      "loss": 0.3945,
+      "step": 574
+    },
+    {
+      "epoch": 1.437836993045245,
+      "grad_norm": 0.17111937701702118,
+      "learning_rate": 0.00010436241610738256,
+      "loss": 0.3928,
+      "step": 575
+    },
+    {
+      "epoch": 1.4403375791201063,
+      "grad_norm": 0.1712803840637207,
+      "learning_rate": 0.00010419463087248323,
+      "loss": 0.391,
+      "step": 576
+    },
+    {
+      "epoch": 1.4428381651949675,
+      "grad_norm": 0.17339582741260529,
+      "learning_rate": 0.00010402684563758389,
+      "loss": 0.3974,
+      "step": 577
+    },
+    {
+      "epoch": 1.4453387512698288,
+      "grad_norm": 0.1708223521709442,
+      "learning_rate": 0.00010385906040268456,
+      "loss": 0.3927,
+      "step": 578
+    },
+    {
+      "epoch": 1.44783933734469,
+      "grad_norm": 0.17007163166999817,
+      "learning_rate": 0.00010369127516778523,
+      "loss": 0.3784,
+      "step": 579
+    },
+    {
+      "epoch": 1.4503399234195515,
+      "grad_norm": 0.17226113379001617,
+      "learning_rate": 0.00010352348993288592,
+      "loss": 0.3937,
+      "step": 580
+    },
+    {
+      "epoch": 1.4528405094944128,
+      "grad_norm": 0.17065738141536713,
+      "learning_rate": 0.00010335570469798659,
+      "loss": 0.3941,
+      "step": 581
+    },
+    {
+      "epoch": 1.455341095569274,
+      "grad_norm": 0.17562048137187958,
+      "learning_rate": 0.00010318791946308725,
+      "loss": 0.3764,
+      "step": 582
+    },
+    {
+      "epoch": 1.4578416816441353,
+      "grad_norm": 0.17384320497512817,
+      "learning_rate": 0.00010302013422818793,
+      "loss": 0.3931,
+      "step": 583
+    },
+    {
+      "epoch": 1.4603422677189966,
+      "grad_norm": 0.1761142611503601,
+      "learning_rate": 0.0001028523489932886,
+      "loss": 0.3789,
+      "step": 584
+    },
+    {
+      "epoch": 1.462842853793858,
+      "grad_norm": 0.1703573316335678,
+      "learning_rate": 0.00010268456375838926,
+      "loss": 0.3825,
+      "step": 585
+    },
+    {
+      "epoch": 1.4653434398687193,
+      "grad_norm": 0.17177999019622803,
+      "learning_rate": 0.00010251677852348993,
+      "loss": 0.3888,
+      "step": 586
+    },
+    {
+      "epoch": 1.4678440259435805,
+      "grad_norm": 0.17300187051296234,
+      "learning_rate": 0.0001023489932885906,
+      "loss": 0.3943,
+      "step": 587
+    },
+    {
+      "epoch": 1.4703446120184418,
+      "grad_norm": 0.17999659478664398,
+      "learning_rate": 0.00010218120805369129,
+      "loss": 0.3973,
+      "step": 588
+    },
+    {
+      "epoch": 1.472845198093303,
+      "grad_norm": 0.18522197008132935,
+      "learning_rate": 0.00010201342281879196,
+      "loss": 0.3908,
+      "step": 589
+    },
+    {
+      "epoch": 1.4753457841681645,
+      "grad_norm": 0.174427330493927,
+      "learning_rate": 0.00010184563758389262,
+      "loss": 0.3932,
+      "step": 590
+    },
+    {
+      "epoch": 1.4778463702430258,
+      "grad_norm": 0.18314112722873688,
+      "learning_rate": 0.0001016778523489933,
+      "loss": 0.4066,
+      "step": 591
+    },
+    {
+      "epoch": 1.480346956317887,
+      "grad_norm": 0.17771361768245697,
+      "learning_rate": 0.00010151006711409397,
+      "loss": 0.3906,
+      "step": 592
+    },
+    {
+      "epoch": 1.4828475423927483,
+      "grad_norm": 0.16917257010936737,
+      "learning_rate": 0.00010134228187919463,
+      "loss": 0.371,
+      "step": 593
+    },
+    {
+      "epoch": 1.4853481284676096,
+      "grad_norm": 0.17607718706130981,
+      "learning_rate": 0.0001011744966442953,
+      "loss": 0.3841,
+      "step": 594
+    },
+    {
+      "epoch": 1.487848714542471,
+      "grad_norm": 0.17842422425746918,
+      "learning_rate": 0.00010100671140939597,
+      "loss": 0.3989,
+      "step": 595
+    },
+    {
+      "epoch": 1.490349300617332,
+      "grad_norm": 0.17947570979595184,
+      "learning_rate": 0.00010083892617449663,
+      "loss": 0.3986,
+      "step": 596
+    },
+    {
+      "epoch": 1.4928498866921935,
+      "grad_norm": 0.17755621671676636,
+      "learning_rate": 0.00010067114093959733,
+      "loss": 0.3823,
+      "step": 597
+    },
+    {
+      "epoch": 1.4953504727670548,
+      "grad_norm": 0.17342810332775116,
+      "learning_rate": 0.00010050335570469799,
+      "loss": 0.383,
+      "step": 598
+    },
+    {
+      "epoch": 1.497851058841916,
+      "grad_norm": 0.17453037202358246,
+      "learning_rate": 0.00010033557046979867,
+      "loss": 0.3837,
+      "step": 599
+    },
+    {
+      "epoch": 1.5003516449167775,
+      "grad_norm": 0.17941130697727203,
+      "learning_rate": 0.00010016778523489934,
+      "loss": 0.3878,
+      "step": 600
+    },
+    {
+      "epoch": 1.5028522309916386,
+      "grad_norm": 0.1898651123046875,
+      "learning_rate": 0.0001,
+      "loss": 0.3887,
+      "step": 601
+    },
+    {
+      "epoch": 1.5053528170665,
+      "grad_norm": 0.18540070950984955,
+      "learning_rate": 9.983221476510067e-05,
+      "loss": 0.3766,
+      "step": 602
+    },
+    {
+      "epoch": 1.5078534031413613,
+      "grad_norm": 0.1742309033870697,
+      "learning_rate": 9.966442953020134e-05,
+      "loss": 0.3911,
+      "step": 603
+    },
+    {
+      "epoch": 1.5103539892162225,
+      "grad_norm": 0.18229171633720398,
+      "learning_rate": 9.949664429530202e-05,
+      "loss": 0.4041,
+      "step": 604
+    },
+    {
+      "epoch": 1.5128545752910838,
+      "grad_norm": 0.1907065212726593,
+      "learning_rate": 9.93288590604027e-05,
+      "loss": 0.3736,
+      "step": 605
+    },
+    {
+      "epoch": 1.515355161365945,
+      "grad_norm": 0.17920121550559998,
+      "learning_rate": 9.916107382550336e-05,
+      "loss": 0.3915,
+      "step": 606
+    },
+    {
+      "epoch": 1.5178557474408065,
+      "grad_norm": 0.17517173290252686,
+      "learning_rate": 9.899328859060403e-05,
+      "loss": 0.3791,
+      "step": 607
+    },
+    {
+      "epoch": 1.5203563335156676,
+      "grad_norm": 0.1752045750617981,
+      "learning_rate": 9.882550335570471e-05,
+      "loss": 0.3972,
+      "step": 608
+    },
+    {
+      "epoch": 1.522856919590529,
+      "grad_norm": 0.1808864325284958,
+      "learning_rate": 9.865771812080538e-05,
+      "loss": 0.3861,
+      "step": 609
+    },
+    {
+      "epoch": 1.5253575056653903,
+      "grad_norm": 0.17795230448246002,
+      "learning_rate": 9.848993288590604e-05,
+      "loss": 0.3819,
+      "step": 610
+    },
+    {
+      "epoch": 1.5278580917402516,
+      "grad_norm": 0.18556566536426544,
+      "learning_rate": 9.832214765100671e-05,
+      "loss": 0.4073,
+      "step": 611
+    },
+    {
+      "epoch": 1.530358677815113,
+      "grad_norm": 0.17612846195697784,
+      "learning_rate": 9.815436241610739e-05,
+      "loss": 0.3824,
+      "step": 612
+    },
+    {
+      "epoch": 1.532859263889974,
+      "grad_norm": 0.17632323503494263,
+      "learning_rate": 9.798657718120807e-05,
+      "loss": 0.38,
+      "step": 613
+    },
+    {
+      "epoch": 1.5353598499648355,
+      "grad_norm": 0.17899446189403534,
+      "learning_rate": 9.781879194630873e-05,
+      "loss": 0.3792,
+      "step": 614
+    },
+    {
+      "epoch": 1.5378604360396968,
+      "grad_norm": 0.1789308488368988,
+      "learning_rate": 9.76510067114094e-05,
+      "loss": 0.3939,
+      "step": 615
+    },
+    {
+      "epoch": 1.540361022114558,
+      "grad_norm": 0.17783896625041962,
+      "learning_rate": 9.748322147651007e-05,
+      "loss": 0.3811,
+      "step": 616
+    },
+    {
+      "epoch": 1.5428616081894195,
+      "grad_norm": 0.18002574145793915,
+      "learning_rate": 9.731543624161075e-05,
+      "loss": 0.3912,
+      "step": 617
+    },
+    {
+      "epoch": 1.5453621942642806,
+      "grad_norm": 0.17521120607852936,
+      "learning_rate": 9.714765100671141e-05,
+      "loss": 0.3771,
+      "step": 618
+    },
+    {
+      "epoch": 1.547862780339142,
+      "grad_norm": 0.177418515086174,
+      "learning_rate": 9.697986577181208e-05,
+      "loss": 0.3823,
+      "step": 619
+    },
+    {
+      "epoch": 1.5503633664140033,
+      "grad_norm": 0.1844666749238968,
+      "learning_rate": 9.681208053691276e-05,
+      "loss": 0.3867,
+      "step": 620
+    },
+    {
+      "epoch": 1.5528639524888646,
+      "grad_norm": 0.17712990939617157,
+      "learning_rate": 9.664429530201344e-05,
+      "loss": 0.3769,
+      "step": 621
+    },
+    {
+      "epoch": 1.555364538563726,
+      "grad_norm": 0.1784822940826416,
+      "learning_rate": 9.64765100671141e-05,
+      "loss": 0.3812,
+      "step": 622
+    },
+    {
+      "epoch": 1.557865124638587,
+      "grad_norm": 0.18097448348999023,
+      "learning_rate": 9.630872483221477e-05,
+      "loss": 0.3969,
+      "step": 623
+    },
+    {
+      "epoch": 1.5603657107134485,
+      "grad_norm": 0.1777218133211136,
+      "learning_rate": 9.614093959731544e-05,
+      "loss": 0.3895,
+      "step": 624
+    },
+    {
+      "epoch": 1.5628662967883098,
+      "grad_norm": 0.17573560774326324,
+      "learning_rate": 9.59731543624161e-05,
+      "loss": 0.3807,
+      "step": 625
+    },
+    {
+      "epoch": 1.565366882863171,
+      "grad_norm": 0.1790502518415451,
+      "learning_rate": 9.580536912751678e-05,
+      "loss": 0.3744,
+      "step": 626
+    },
+    {
+      "epoch": 1.5678674689380323,
+      "grad_norm": 0.17353123426437378,
+      "learning_rate": 9.563758389261745e-05,
+      "loss": 0.3839,
+      "step": 627
+    },
+    {
+      "epoch": 1.5703680550128936,
+      "grad_norm": 0.1754903495311737,
+      "learning_rate": 9.546979865771813e-05,
+      "loss": 0.3796,
+      "step": 628
+    },
+    {
+      "epoch": 1.572868641087755,
+      "grad_norm": 0.174009770154953,
+      "learning_rate": 9.53020134228188e-05,
+      "loss": 0.374,
+      "step": 629
+    },
+    {
+      "epoch": 1.5753692271626163,
+      "grad_norm": 0.1829177439212799,
+      "learning_rate": 9.513422818791947e-05,
+      "loss": 0.387,
+      "step": 630
+    },
+    {
+      "epoch": 1.5778698132374775,
+      "grad_norm": 0.17329353094100952,
+      "learning_rate": 9.496644295302014e-05,
+      "loss": 0.3942,
+      "step": 631
+    },
+    {
+      "epoch": 1.5803703993123388,
+      "grad_norm": 0.16881446540355682,
+      "learning_rate": 9.479865771812081e-05,
+      "loss": 0.3831,
+      "step": 632
+    },
+    {
+      "epoch": 1.5828709853872,
+      "grad_norm": 0.18109095096588135,
+      "learning_rate": 9.463087248322147e-05,
+      "loss": 0.3834,
+      "step": 633
+    },
+    {
+      "epoch": 1.5853715714620615,
+      "grad_norm": 0.18489782512187958,
+      "learning_rate": 9.446308724832215e-05,
+      "loss": 0.3892,
+      "step": 634
+    },
+    {
+      "epoch": 1.5878721575369226,
+      "grad_norm": 0.19050204753875732,
+      "learning_rate": 9.429530201342282e-05,
+      "loss": 0.4003,
+      "step": 635
+    },
+    {
+      "epoch": 1.590372743611784,
+      "grad_norm": 0.19048739969730377,
+      "learning_rate": 9.41275167785235e-05,
+      "loss": 0.3871,
+      "step": 636
+    },
+    {
+      "epoch": 1.5928733296866453,
+      "grad_norm": 0.185329407453537,
+      "learning_rate": 9.395973154362417e-05,
+      "loss": 0.3946,
+      "step": 637
+    },
+    {
+      "epoch": 1.5953739157615066,
+      "grad_norm": 0.18026521801948547,
+      "learning_rate": 9.379194630872483e-05,
+      "loss": 0.3687,
+      "step": 638
+    },
+    {
+      "epoch": 1.597874501836368,
+      "grad_norm": 0.17552684247493744,
+      "learning_rate": 9.362416107382551e-05,
+      "loss": 0.3852,
+      "step": 639
+    },
+    {
+      "epoch": 1.600375087911229,
+      "grad_norm": 0.16905976831912994,
+      "learning_rate": 9.345637583892618e-05,
+      "loss": 0.3821,
+      "step": 640
+    },
+    {
+      "epoch": 1.6028756739860905,
+      "grad_norm": 0.17816942930221558,
+      "learning_rate": 9.328859060402684e-05,
+      "loss": 0.3885,
+      "step": 641
+    },
+    {
+      "epoch": 1.6053762600609518,
+      "grad_norm": 0.18062306940555573,
+      "learning_rate": 9.312080536912751e-05,
+      "loss": 0.3997,
+      "step": 642
+    },
+    {
+      "epoch": 1.607876846135813,
+      "grad_norm": 0.18119755387306213,
+      "learning_rate": 9.295302013422819e-05,
+      "loss": 0.3887,
+      "step": 643
+    },
+    {
+      "epoch": 1.6103774322106745,
+      "grad_norm": 0.17793086171150208,
+      "learning_rate": 9.278523489932887e-05,
+      "loss": 0.3814,
+      "step": 644
+    },
+    {
+      "epoch": 1.6128780182855356,
+      "grad_norm": 0.17421945929527283,
+      "learning_rate": 9.261744966442954e-05,
+      "loss": 0.3698,
+      "step": 645
+    },
+    {
+      "epoch": 1.615378604360397,
+      "grad_norm": 0.18807508051395416,
+      "learning_rate": 9.24496644295302e-05,
+      "loss": 0.3956,
+      "step": 646
+    },
+    {
+      "epoch": 1.6178791904352583,
+      "grad_norm": 0.1884203851222992,
+      "learning_rate": 9.228187919463087e-05,
+      "loss": 0.3884,
+      "step": 647
+    },
+    {
+      "epoch": 1.6203797765101196,
+      "grad_norm": 0.1798408031463623,
+      "learning_rate": 9.211409395973155e-05,
+      "loss": 0.386,
+      "step": 648
+    },
+    {
+      "epoch": 1.6228803625849808,
+      "grad_norm": 0.17403843998908997,
+      "learning_rate": 9.194630872483221e-05,
+      "loss": 0.3795,
+      "step": 649
+    },
+    {
+      "epoch": 1.625380948659842,
+      "grad_norm": 0.176111102104187,
+      "learning_rate": 9.17785234899329e-05,
+      "loss": 0.3891,
+      "step": 650
+    },
+    {
+      "epoch": 1.6278815347347035,
+      "grad_norm": 0.17879219353199005,
+      "learning_rate": 9.161073825503356e-05,
+      "loss": 0.3961,
+      "step": 651
+    },
+    {
+      "epoch": 1.6303821208095648,
+      "grad_norm": 0.17962028086185455,
+      "learning_rate": 9.144295302013424e-05,
+      "loss": 0.391,
+      "step": 652
+    },
+    {
+      "epoch": 1.632882706884426,
+      "grad_norm": 0.18730774521827698,
+      "learning_rate": 9.127516778523491e-05,
+      "loss": 0.3844,
+      "step": 653
+    },
+    {
+      "epoch": 1.6353832929592873,
+      "grad_norm": 0.17960268259048462,
+      "learning_rate": 9.110738255033557e-05,
+      "loss": 0.384,
+      "step": 654
+    },
+    {
+      "epoch": 1.6378838790341486,
+      "grad_norm": 0.17033277451992035,
+      "learning_rate": 9.093959731543624e-05,
+      "loss": 0.3882,
+      "step": 655
+    },
+    {
+      "epoch": 1.64038446510901,
+      "grad_norm": 0.17765159904956818,
+      "learning_rate": 9.077181208053692e-05,
+      "loss": 0.3753,
+      "step": 656
+    },
+    {
+      "epoch": 1.642885051183871,
+      "grad_norm": 0.18699118494987488,
+      "learning_rate": 9.060402684563759e-05,
+      "loss": 0.4095,
+      "step": 657
+    },
+    {
+      "epoch": 1.6453856372587325,
+      "grad_norm": 0.18108096718788147,
+      "learning_rate": 9.043624161073827e-05,
+      "loss": 0.3918,
+      "step": 658
+    },
+    {
+      "epoch": 1.6478862233335938,
+      "grad_norm": 0.18553516268730164,
+      "learning_rate": 9.026845637583893e-05,
+      "loss": 0.379,
+      "step": 659
+    },
+    {
+      "epoch": 1.650386809408455,
+      "grad_norm": 0.1940268576145172,
+      "learning_rate": 9.01006711409396e-05,
+      "loss": 0.3855,
+      "step": 660
+    },
+    {
+      "epoch": 1.6528873954833165,
+      "grad_norm": 0.17603854835033417,
+      "learning_rate": 8.993288590604028e-05,
+      "loss": 0.3804,
+      "step": 661
+    },
+    {
+      "epoch": 1.6553879815581776,
+      "grad_norm": 0.1748921126127243,
+      "learning_rate": 8.976510067114094e-05,
+      "loss": 0.3645,
+      "step": 662
+    },
+    {
+      "epoch": 1.657888567633039,
+      "grad_norm": 0.20754989981651306,
+      "learning_rate": 8.959731543624161e-05,
+      "loss": 0.396,
+      "step": 663
+    },
+    {
+      "epoch": 1.6603891537079003,
+      "grad_norm": 0.1931428760290146,
+      "learning_rate": 8.942953020134228e-05,
+      "loss": 0.3874,
+      "step": 664
+    },
+    {
+      "epoch": 1.6628897397827616,
+      "grad_norm": 0.18024632334709167,
+      "learning_rate": 8.926174496644296e-05,
+      "loss": 0.3803,
+      "step": 665
+    },
+    {
+      "epoch": 1.665390325857623,
+      "grad_norm": 0.18505463004112244,
+      "learning_rate": 8.909395973154364e-05,
+      "loss": 0.3682,
+      "step": 666
+    },
+    {
+      "epoch": 1.667890911932484,
+      "grad_norm": 0.20315155386924744,
+      "learning_rate": 8.89261744966443e-05,
+      "loss": 0.3814,
+      "step": 667
+    },
+    {
+      "epoch": 1.6703914980073455,
+      "grad_norm": 0.19038356840610504,
+      "learning_rate": 8.875838926174497e-05,
+      "loss": 0.3865,
+      "step": 668
+    },
+    {
+      "epoch": 1.6728920840822068,
+      "grad_norm": 0.20000489056110382,
+      "learning_rate": 8.859060402684565e-05,
+      "loss": 0.3921,
+      "step": 669
+    },
+    {
+      "epoch": 1.675392670157068,
+      "grad_norm": 0.19777032732963562,
+      "learning_rate": 8.842281879194631e-05,
+      "loss": 0.405,
+      "step": 670
+    },
+    {
+      "epoch": 1.6778932562319293,
+      "grad_norm": 0.1800341010093689,
+      "learning_rate": 8.825503355704698e-05,
+      "loss": 0.3773,
+      "step": 671
+    },
+    {
+      "epoch": 1.6803938423067906,
+      "grad_norm": 0.176991805434227,
+      "learning_rate": 8.808724832214765e-05,
+      "loss": 0.3717,
+      "step": 672
+    },
+    {
+      "epoch": 1.682894428381652,
+      "grad_norm": 0.1964898705482483,
+      "learning_rate": 8.791946308724833e-05,
+      "loss": 0.3824,
+      "step": 673
+    },
+    {
+      "epoch": 1.6853950144565133,
+      "grad_norm": 0.17583096027374268,
+      "learning_rate": 8.7751677852349e-05,
+      "loss": 0.3764,
+      "step": 674
+    },
+    {
+      "epoch": 1.6878956005313746,
+      "grad_norm": 0.1807302087545395,
+      "learning_rate": 8.758389261744967e-05,
+      "loss": 0.3803,
+      "step": 675
+    },
+    {
+      "epoch": 1.6903961866062358,
+      "grad_norm": 0.18745875358581543,
+      "learning_rate": 8.741610738255034e-05,
+      "loss": 0.3882,
+      "step": 676
+    },
+    {
+      "epoch": 1.692896772681097,
+      "grad_norm": 0.19276930391788483,
+      "learning_rate": 8.7248322147651e-05,
+      "loss": 0.3808,
+      "step": 677
+    },
+    {
+      "epoch": 1.6953973587559585,
+      "grad_norm": 0.18820412456989288,
+      "learning_rate": 8.708053691275169e-05,
+      "loss": 0.3948,
+      "step": 678
+    },
+    {
+      "epoch": 1.6978979448308196,
+      "grad_norm": 0.1914145052433014,
+      "learning_rate": 8.691275167785235e-05,
+      "loss": 0.386,
+      "step": 679
+    },
+    {
+      "epoch": 1.700398530905681,
+      "grad_norm": 0.2073824256658554,
+      "learning_rate": 8.674496644295302e-05,
+      "loss": 0.3929,
+      "step": 680
+    },
+    {
+      "epoch": 1.7028991169805423,
+      "grad_norm": 0.17458835244178772,
+      "learning_rate": 8.65771812080537e-05,
+      "loss": 0.3729,
+      "step": 681
+    },
+    {
+      "epoch": 1.7053997030554036,
+      "grad_norm": 0.17508700489997864,
+      "learning_rate": 8.640939597315436e-05,
+      "loss": 0.3744,
+      "step": 682
+    },
+    {
+      "epoch": 1.707900289130265,
+      "grad_norm": 0.18630559742450714,
+      "learning_rate": 8.624161073825504e-05,
+      "loss": 0.3913,
+      "step": 683
+    },
+    {
+      "epoch": 1.710400875205126,
+      "grad_norm": 0.1864834427833557,
+      "learning_rate": 8.607382550335571e-05,
+      "loss": 0.3903,
+      "step": 684
+    },
+    {
+      "epoch": 1.7129014612799875,
+      "grad_norm": 0.1861889809370041,
+      "learning_rate": 8.590604026845638e-05,
+      "loss": 0.3934,
+      "step": 685
+    },
+    {
+      "epoch": 1.7154020473548488,
+      "grad_norm": 0.1995447874069214,
+      "learning_rate": 8.573825503355704e-05,
+      "loss": 0.3955,
+      "step": 686
+    },
+    {
+      "epoch": 1.71790263342971,
+      "grad_norm": 0.18536217510700226,
+      "learning_rate": 8.557046979865772e-05,
+      "loss": 0.3803,
+      "step": 687
+    },
+    {
+      "epoch": 1.7204032195045715,
+      "grad_norm": 0.20880867540836334,
+      "learning_rate": 8.54026845637584e-05,
+      "loss": 0.3862,
+      "step": 688
+    },
+    {
+      "epoch": 1.7229038055794326,
+      "grad_norm": 0.17543533444404602,
+      "learning_rate": 8.523489932885907e-05,
+      "loss": 0.389,
+      "step": 689
+    },
+    {
+      "epoch": 1.725404391654294,
+      "grad_norm": 0.18419736623764038,
+      "learning_rate": 8.506711409395973e-05,
+      "loss": 0.3697,
+      "step": 690
+    },
+    {
+      "epoch": 1.7279049777291553,
+      "grad_norm": 0.17430716753005981,
+      "learning_rate": 8.489932885906041e-05,
+      "loss": 0.376,
+      "step": 691
+    },
+    {
+      "epoch": 1.7304055638040166,
+      "grad_norm": 0.17576076090335846,
+      "learning_rate": 8.473154362416108e-05,
+      "loss": 0.3717,
+      "step": 692
+    },
+    {
+      "epoch": 1.7329061498788778,
+      "grad_norm": 0.1788487732410431,
+      "learning_rate": 8.456375838926175e-05,
+      "loss": 0.391,
+      "step": 693
+    },
+    {
+      "epoch": 1.735406735953739,
+      "grad_norm": 0.17405882477760315,
+      "learning_rate": 8.439597315436241e-05,
+      "loss": 0.3824,
+      "step": 694
+    },
+    {
+      "epoch": 1.7379073220286005,
+      "grad_norm": 0.18588879704475403,
+      "learning_rate": 8.422818791946309e-05,
+      "loss": 0.4043,
+      "step": 695
+    },
+    {
+      "epoch": 1.7404079081034618,
+      "grad_norm": 0.17953068017959595,
+      "learning_rate": 8.406040268456377e-05,
+      "loss": 0.3816,
+      "step": 696
+    },
+    {
+      "epoch": 1.742908494178323,
+      "grad_norm": 0.17672885954380035,
+      "learning_rate": 8.389261744966444e-05,
+      "loss": 0.3921,
+      "step": 697
+    },
+    {
+      "epoch": 1.7454090802531843,
+      "grad_norm": 0.1765912026166916,
+      "learning_rate": 8.37248322147651e-05,
+      "loss": 0.3675,
+      "step": 698
+    },
+    {
+      "epoch": 1.7479096663280456,
+      "grad_norm": 0.16937559843063354,
+      "learning_rate": 8.355704697986577e-05,
+      "loss": 0.3658,
+      "step": 699
+    },
+    {
+      "epoch": 1.750410252402907,
+      "grad_norm": 0.18374691903591156,
+      "learning_rate": 8.338926174496645e-05,
+      "loss": 0.3936,
+      "step": 700
+    },
+    {
+      "epoch": 1.752910838477768,
+      "grad_norm": 0.1693335324525833,
+      "learning_rate": 8.322147651006712e-05,
+      "loss": 0.3726,
+      "step": 701
+    },
+    {
+      "epoch": 1.7554114245526296,
+      "grad_norm": 0.17510807514190674,
+      "learning_rate": 8.305369127516778e-05,
+      "loss": 0.3799,
+      "step": 702
+    },
+    {
+      "epoch": 1.7579120106274908,
+      "grad_norm": 0.19161054491996765,
+      "learning_rate": 8.288590604026846e-05,
+      "loss": 0.3982,
+      "step": 703
+    },
+    {
+      "epoch": 1.760412596702352,
+      "grad_norm": 0.18057194352149963,
+      "learning_rate": 8.271812080536914e-05,
+      "loss": 0.3786,
+      "step": 704
+    },
+    {
+      "epoch": 1.7629131827772135,
+      "grad_norm": 0.18895000219345093,
+      "learning_rate": 8.255033557046981e-05,
+      "loss": 0.3768,
+      "step": 705
+    },
+    {
+      "epoch": 1.7654137688520746,
+      "grad_norm": 0.18634304404258728,
+      "learning_rate": 8.238255033557048e-05,
+      "loss": 0.3907,
+      "step": 706
+    },
+    {
+      "epoch": 1.767914354926936,
+      "grad_norm": 0.1855781376361847,
+      "learning_rate": 8.221476510067114e-05,
+      "loss": 0.3957,
+      "step": 707
+    },
+    {
+      "epoch": 1.7704149410017973,
+      "grad_norm": 0.1801329404115677,
+      "learning_rate": 8.204697986577181e-05,
+      "loss": 0.3742,
+      "step": 708
+    },
+    {
+      "epoch": 1.7729155270766586,
+      "grad_norm": 0.1889009326696396,
+      "learning_rate": 8.187919463087249e-05,
+      "loss": 0.383,
+      "step": 709
+    },
+    {
+      "epoch": 1.77541611315152,
+      "grad_norm": 0.1734006106853485,
+      "learning_rate": 8.171140939597315e-05,
+      "loss": 0.3956,
+      "step": 710
+    },
+    {
+      "epoch": 1.777916699226381,
+      "grad_norm": 0.17539815604686737,
+      "learning_rate": 8.154362416107383e-05,
+      "loss": 0.3781,
+      "step": 711
+    },
+    {
+      "epoch": 1.7804172853012425,
+      "grad_norm": 0.2037649005651474,
+      "learning_rate": 8.13758389261745e-05,
+      "loss": 0.3868,
+      "step": 712
+    },
+    {
+      "epoch": 1.7829178713761038,
+      "grad_norm": 0.17677001655101776,
+      "learning_rate": 8.120805369127518e-05,
+      "loss": 0.3769,
+      "step": 713
+    },
+    {
+      "epoch": 1.785418457450965,
+      "grad_norm": 0.18312117457389832,
+      "learning_rate": 8.104026845637585e-05,
+      "loss": 0.3641,
+      "step": 714
+    },
+    {
+      "epoch": 1.7879190435258263,
+      "grad_norm": 0.19968779385089874,
+      "learning_rate": 8.087248322147651e-05,
+      "loss": 0.3888,
+      "step": 715
+    },
+    {
+      "epoch": 1.7904196296006876,
+      "grad_norm": 0.17776362597942352,
+      "learning_rate": 8.070469798657718e-05,
+      "loss": 0.3799,
+      "step": 716
+    },
+    {
+      "epoch": 1.792920215675549,
+      "grad_norm": 0.18887469172477722,
+      "learning_rate": 8.053691275167784e-05,
+      "loss": 0.3811,
+      "step": 717
+    },
+    {
+      "epoch": 1.7954208017504103,
+      "grad_norm": 0.18911400437355042,
+      "learning_rate": 8.036912751677852e-05,
+      "loss": 0.3659,
+      "step": 718
+    },
+    {
+      "epoch": 1.7979213878252716,
+      "grad_norm": 0.18411016464233398,
+      "learning_rate": 8.02013422818792e-05,
+      "loss": 0.3738,
+      "step": 719
+    },
+    {
+      "epoch": 1.8004219739001328,
+      "grad_norm": 0.17410506308078766,
+      "learning_rate": 8.003355704697987e-05,
+      "loss": 0.3697,
+      "step": 720
+    },
+    {
+      "epoch": 1.802922559974994,
+      "grad_norm": 0.1747969537973404,
+      "learning_rate": 7.986577181208054e-05,
+      "loss": 0.3742,
+      "step": 721
+    },
+    {
+      "epoch": 1.8054231460498555,
+      "grad_norm": 0.18556781113147736,
+      "learning_rate": 7.969798657718122e-05,
+      "loss": 0.3836,
+      "step": 722
+    },
+    {
+      "epoch": 1.8079237321247166,
+      "grad_norm": 0.18280568718910217,
+      "learning_rate": 7.953020134228188e-05,
+      "loss": 0.3635,
+      "step": 723
+    },
+    {
+      "epoch": 1.810424318199578,
+      "grad_norm": 0.17794889211654663,
+      "learning_rate": 7.936241610738255e-05,
+      "loss": 0.3863,
+      "step": 724
+    },
+    {
+      "epoch": 1.8129249042744393,
+      "grad_norm": 0.18276084959506989,
+      "learning_rate": 7.919463087248322e-05,
+      "loss": 0.394,
+      "step": 725
+    },
+    {
+      "epoch": 1.8154254903493006,
+      "grad_norm": 0.18190203607082367,
+      "learning_rate": 7.90268456375839e-05,
+      "loss": 0.3813,
+      "step": 726
+    },
+    {
+      "epoch": 1.817926076424162,
+      "grad_norm": 0.1755213439464569,
+      "learning_rate": 7.885906040268457e-05,
+      "loss": 0.3847,
+      "step": 727
+    },
+    {
+      "epoch": 1.820426662499023,
+      "grad_norm": 0.17683593928813934,
+      "learning_rate": 7.869127516778524e-05,
+      "loss": 0.389,
+      "step": 728
+    },
+    {
+      "epoch": 1.8229272485738846,
+      "grad_norm": 0.1868712455034256,
+      "learning_rate": 7.852348993288591e-05,
+      "loss": 0.3825,
+      "step": 729
+    },
+    {
+      "epoch": 1.8254278346487458,
+      "grad_norm": 0.18461084365844727,
+      "learning_rate": 7.835570469798657e-05,
+      "loss": 0.3904,
+      "step": 730
+    },
+    {
+      "epoch": 1.827928420723607,
+      "grad_norm": 0.17919501662254333,
+      "learning_rate": 7.818791946308725e-05,
+      "loss": 0.3738,
+      "step": 731
+    },
+    {
+      "epoch": 1.8304290067984685,
+      "grad_norm": 0.1868787705898285,
+      "learning_rate": 7.802013422818792e-05,
+      "loss": 0.3847,
+      "step": 732
+    },
+    {
+      "epoch": 1.8329295928733296,
+      "grad_norm": 0.1995868980884552,
+      "learning_rate": 7.78523489932886e-05,
+      "loss": 0.3927,
+      "step": 733
+    },
+    {
+      "epoch": 1.835430178948191,
+      "grad_norm": 0.17577460408210754,
+      "learning_rate": 7.768456375838927e-05,
+      "loss": 0.3756,
+      "step": 734
+    },
+    {
+      "epoch": 1.8379307650230523,
+      "grad_norm": 0.17471660673618317,
+      "learning_rate": 7.751677852348995e-05,
+      "loss": 0.3605,
+      "step": 735
+    },
+    {
+      "epoch": 1.8404313510979136,
+      "grad_norm": 0.1837376207113266,
+      "learning_rate": 7.734899328859061e-05,
+      "loss": 0.3839,
+      "step": 736
+    },
+    {
+      "epoch": 1.8429319371727748,
+      "grad_norm": 0.17759470641613007,
+      "learning_rate": 7.718120805369128e-05,
+      "loss": 0.3743,
+      "step": 737
+    },
+    {
+      "epoch": 1.845432523247636,
+      "grad_norm": 0.17445184290409088,
+      "learning_rate": 7.701342281879194e-05,
+      "loss": 0.3806,
+      "step": 738
+    },
+    {
+      "epoch": 1.8479331093224975,
+      "grad_norm": 0.18021027743816376,
+      "learning_rate": 7.684563758389261e-05,
+      "loss": 0.3812,
+      "step": 739
+    },
+    {
+      "epoch": 1.8504336953973588,
+      "grad_norm": 0.17419204115867615,
+      "learning_rate": 7.667785234899329e-05,
+      "loss": 0.377,
+      "step": 740
+    },
+    {
+      "epoch": 1.85293428147222,
+      "grad_norm": 0.18193626403808594,
+      "learning_rate": 7.651006711409397e-05,
+      "loss": 0.3909,
+      "step": 741
+    },
+    {
+      "epoch": 1.8554348675470813,
+      "grad_norm": 0.17466546595096588,
+      "learning_rate": 7.634228187919464e-05,
+      "loss": 0.3772,
+      "step": 742
+    },
+    {
+      "epoch": 1.8579354536219426,
+      "grad_norm": 0.17153415083885193,
+      "learning_rate": 7.61744966442953e-05,
+      "loss": 0.3717,
+      "step": 743
+    },
+    {
+      "epoch": 1.860436039696804,
+      "grad_norm": 0.17623353004455566,
+      "learning_rate": 7.600671140939598e-05,
+      "loss": 0.39,
+      "step": 744
+    },
+    {
+      "epoch": 1.862936625771665,
+      "grad_norm": 0.1748313009738922,
+      "learning_rate": 7.583892617449665e-05,
+      "loss": 0.3799,
+      "step": 745
+    },
+    {
+      "epoch": 1.8654372118465266,
+      "grad_norm": 0.1791505664587021,
+      "learning_rate": 7.567114093959731e-05,
+      "loss": 0.3718,
+      "step": 746
+    },
+    {
+      "epoch": 1.8679377979213878,
+      "grad_norm": 0.17601267993450165,
+      "learning_rate": 7.550335570469798e-05,
+      "loss": 0.3808,
+      "step": 747
+    },
+    {
+      "epoch": 1.870438383996249,
+      "grad_norm": 0.1794864535331726,
+      "learning_rate": 7.533557046979866e-05,
+      "loss": 0.3778,
+      "step": 748
+    },
+    {
+      "epoch": 1.8729389700711105,
+      "grad_norm": 0.18690125644207,
+      "learning_rate": 7.516778523489934e-05,
+      "loss": 0.395,
+      "step": 749
+    },
+    {
+      "epoch": 1.8754395561459716,
+      "grad_norm": 0.18212953209877014,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.3894,
+      "step": 750
+    },
+    {
+      "epoch": 1.877940142220833,
+      "grad_norm": 0.19056768715381622,
+      "learning_rate": 7.483221476510067e-05,
+      "loss": 0.3923,
+      "step": 751
+    },
+    {
+      "epoch": 1.8804407282956943,
+      "grad_norm": 0.1803092211484909,
+      "learning_rate": 7.466442953020134e-05,
+      "loss": 0.3574,
+      "step": 752
+    },
+    {
+      "epoch": 1.8829413143705556,
+      "grad_norm": 0.18736688792705536,
+      "learning_rate": 7.449664429530202e-05,
+      "loss": 0.3755,
+      "step": 753
+    },
+    {
+      "epoch": 1.885441900445417,
+      "grad_norm": 0.17693312466144562,
+      "learning_rate": 7.432885906040269e-05,
+      "loss": 0.3828,
+      "step": 754
+    },
+    {
+      "epoch": 1.887942486520278,
+      "grad_norm": 0.18736334145069122,
+      "learning_rate": 7.416107382550335e-05,
+      "loss": 0.3846,
+      "step": 755
+    },
+    {
+      "epoch": 1.8904430725951396,
+      "grad_norm": 0.1876332014799118,
+      "learning_rate": 7.399328859060403e-05,
+      "loss": 0.3869,
+      "step": 756
+    },
+    {
+      "epoch": 1.8929436586700008,
+      "grad_norm": 0.17936262488365173,
+      "learning_rate": 7.382550335570471e-05,
+      "loss": 0.3794,
+      "step": 757
+    },
+    {
+      "epoch": 1.895444244744862,
+      "grad_norm": 0.18268559873104095,
+      "learning_rate": 7.365771812080538e-05,
+      "loss": 0.391,
+      "step": 758
+    },
+    {
+      "epoch": 1.8979448308197235,
+      "grad_norm": 0.18068313598632812,
+      "learning_rate": 7.348993288590604e-05,
+      "loss": 0.3772,
+      "step": 759
+    },
+    {
+      "epoch": 1.9004454168945846,
+      "grad_norm": 0.18184617161750793,
+      "learning_rate": 7.332214765100671e-05,
+      "loss": 0.386,
+      "step": 760
+    },
+    {
+      "epoch": 1.902946002969446,
+      "grad_norm": 0.17513622343540192,
+      "learning_rate": 7.315436241610739e-05,
+      "loss": 0.3628,
+      "step": 761
+    },
+    {
+      "epoch": 1.9054465890443073,
+      "grad_norm": 0.18235018849372864,
+      "learning_rate": 7.298657718120806e-05,
+      "loss": 0.3881,
+      "step": 762
+    },
+    {
+      "epoch": 1.9079471751191686,
+      "grad_norm": 0.17305365204811096,
+      "learning_rate": 7.281879194630872e-05,
+      "loss": 0.3798,
+      "step": 763
+    },
+    {
+      "epoch": 1.9104477611940298,
+      "grad_norm": 0.1890551894903183,
+      "learning_rate": 7.26510067114094e-05,
+      "loss": 0.3911,
+      "step": 764
+    },
+    {
+      "epoch": 1.912948347268891,
+      "grad_norm": 0.17433114349842072,
+      "learning_rate": 7.248322147651007e-05,
+      "loss": 0.368,
+      "step": 765
+    },
+    {
+      "epoch": 1.9154489333437525,
+      "grad_norm": 0.172682985663414,
+      "learning_rate": 7.231543624161075e-05,
+      "loss": 0.3613,
+      "step": 766
+    },
+    {
+      "epoch": 1.9179495194186136,
+      "grad_norm": 0.1792813241481781,
+      "learning_rate": 7.214765100671141e-05,
+      "loss": 0.3663,
+      "step": 767
+    },
+    {
+      "epoch": 1.920450105493475,
+      "grad_norm": 0.17770087718963623,
+      "learning_rate": 7.197986577181208e-05,
+      "loss": 0.3846,
+      "step": 768
+    },
+    {
+      "epoch": 1.9229506915683363,
+      "grad_norm": 0.17440468072891235,
+      "learning_rate": 7.181208053691275e-05,
+      "loss": 0.3811,
+      "step": 769
+    },
+    {
+      "epoch": 1.9254512776431976,
+      "grad_norm": 0.1783992350101471,
+      "learning_rate": 7.164429530201343e-05,
+      "loss": 0.3702,
+      "step": 770
+    },
+    {
+      "epoch": 1.927951863718059,
+      "grad_norm": 0.18420962989330292,
+      "learning_rate": 7.147651006711409e-05,
+      "loss": 0.38,
+      "step": 771
+    },
+    {
+      "epoch": 1.93045244979292,
+      "grad_norm": 0.18442897498607635,
+      "learning_rate": 7.130872483221477e-05,
+      "loss": 0.3913,
+      "step": 772
+    },
+    {
+      "epoch": 1.9329530358677816,
+      "grad_norm": 0.19213555753231049,
+      "learning_rate": 7.114093959731544e-05,
+      "loss": 0.3849,
+      "step": 773
+    },
+    {
+      "epoch": 1.9354536219426428,
+      "grad_norm": 0.17588800191879272,
+      "learning_rate": 7.09731543624161e-05,
+      "loss": 0.3814,
+      "step": 774
+    },
+    {
+      "epoch": 1.937954208017504,
+      "grad_norm": 0.17929621040821075,
+      "learning_rate": 7.080536912751679e-05,
+      "loss": 0.3706,
+      "step": 775
+    },
+    {
+      "epoch": 1.9404547940923655,
+      "grad_norm": 0.17633728682994843,
+      "learning_rate": 7.063758389261745e-05,
+      "loss": 0.365,
+      "step": 776
+    },
+    {
+      "epoch": 1.9429553801672266,
+      "grad_norm": 0.1767294555902481,
+      "learning_rate": 7.046979865771812e-05,
+      "loss": 0.3677,
+      "step": 777
+    },
+    {
+      "epoch": 1.945455966242088,
+      "grad_norm": 0.17600083351135254,
+      "learning_rate": 7.03020134228188e-05,
+      "loss": 0.3566,
+      "step": 778
+    },
+    {
+      "epoch": 1.9479565523169493,
+      "grad_norm": 0.17859125137329102,
+      "learning_rate": 7.013422818791948e-05,
+      "loss": 0.3713,
+      "step": 779
+    },
+    {
+      "epoch": 1.9504571383918106,
+      "grad_norm": 0.183541402220726,
+      "learning_rate": 6.996644295302014e-05,
+      "loss": 0.3845,
+      "step": 780
+    },
+    {
+      "epoch": 1.952957724466672,
+      "grad_norm": 0.17881344258785248,
+      "learning_rate": 6.979865771812081e-05,
+      "loss": 0.3819,
+      "step": 781
+    },
+    {
+      "epoch": 1.955458310541533,
+      "grad_norm": 0.1792813092470169,
+      "learning_rate": 6.963087248322148e-05,
+      "loss": 0.373,
+      "step": 782
+    },
+    {
+      "epoch": 1.9579588966163946,
+      "grad_norm": 0.1860746592283249,
+      "learning_rate": 6.946308724832216e-05,
+      "loss": 0.3892,
+      "step": 783
+    },
+    {
+      "epoch": 1.9604594826912558,
+      "grad_norm": 0.17330695688724518,
+      "learning_rate": 6.929530201342282e-05,
+      "loss": 0.3675,
+      "step": 784
+    },
+    {
+      "epoch": 1.962960068766117,
+      "grad_norm": 0.1818770319223404,
+      "learning_rate": 6.912751677852349e-05,
+      "loss": 0.3744,
+      "step": 785
+    },
+    {
+      "epoch": 1.9654606548409783,
+      "grad_norm": 0.18105196952819824,
+      "learning_rate": 6.895973154362417e-05,
+      "loss": 0.3789,
+      "step": 786
+    },
+    {
+      "epoch": 1.9679612409158396,
+      "grad_norm": 0.17302206158638,
+      "learning_rate": 6.879194630872483e-05,
+      "loss": 0.3752,
+      "step": 787
+    },
+    {
+      "epoch": 1.970461826990701,
+      "grad_norm": 0.17827250063419342,
+      "learning_rate": 6.862416107382551e-05,
+      "loss": 0.3888,
+      "step": 788
+    },
+    {
+      "epoch": 1.972962413065562,
+      "grad_norm": 0.1822272539138794,
+      "learning_rate": 6.845637583892618e-05,
+      "loss": 0.3722,
+      "step": 789
+    },
+    {
+      "epoch": 1.9754629991404236,
+      "grad_norm": 0.1848425418138504,
+      "learning_rate": 6.828859060402685e-05,
+      "loss": 0.3601,
+      "step": 790
+    },
+    {
+      "epoch": 1.9779635852152848,
+      "grad_norm": 0.19786860048770905,
+      "learning_rate": 6.812080536912751e-05,
+      "loss": 0.3786,
+      "step": 791
+    },
+    {
+      "epoch": 1.980464171290146,
+      "grad_norm": 0.1781584918498993,
+      "learning_rate": 6.795302013422819e-05,
+      "loss": 0.3805,
+      "step": 792
+    },
+    {
+      "epoch": 1.9829647573650075,
+      "grad_norm": 0.1861296147108078,
+      "learning_rate": 6.778523489932886e-05,
+      "loss": 0.3676,
+      "step": 793
+    },
+    {
+      "epoch": 1.9854653434398686,
+      "grad_norm": 0.19738426804542542,
+      "learning_rate": 6.761744966442954e-05,
+      "loss": 0.3909,
+      "step": 794
+    },
+    {
+      "epoch": 1.98796592951473,
+      "grad_norm": 0.18954281508922577,
+      "learning_rate": 6.74496644295302e-05,
+      "loss": 0.3853,
+      "step": 795
+    },
+    {
+      "epoch": 1.9904665155895913,
+      "grad_norm": 0.18448856472969055,
+      "learning_rate": 6.728187919463087e-05,
+      "loss": 0.3921,
+      "step": 796
+    },
+    {
+      "epoch": 1.9929671016644526,
+      "grad_norm": 0.19441081583499908,
+      "learning_rate": 6.711409395973155e-05,
+      "loss": 0.3916,
+      "step": 797
+    },
+    {
+      "epoch": 1.995467687739314,
+      "grad_norm": 0.18970319628715515,
+      "learning_rate": 6.694630872483222e-05,
+      "loss": 0.3773,
+      "step": 798
+    },
+    {
+      "epoch": 1.997968273814175,
+      "grad_norm": 0.17508842051029205,
+      "learning_rate": 6.677852348993288e-05,
+      "loss": 0.3817,
+      "step": 799
+    },
+    {
+      "epoch": 2.0004688598890366,
+      "grad_norm": 0.18185412883758545,
+      "learning_rate": 6.661073825503355e-05,
+      "loss": 0.37,
+      "step": 800
+    },
+    {
+      "epoch": 2.0029694459638976,
+      "grad_norm": 0.17345231771469116,
+      "learning_rate": 6.644295302013423e-05,
+      "loss": 0.3386,
+      "step": 801
+    },
+    {
+      "epoch": 2.005470032038759,
+      "grad_norm": 0.16668309271335602,
+      "learning_rate": 6.627516778523491e-05,
+      "loss": 0.3422,
+      "step": 802
+    },
+    {
+      "epoch": 2.0079706181136205,
+      "grad_norm": 0.17595726251602173,
+      "learning_rate": 6.610738255033558e-05,
+      "loss": 0.3645,
+      "step": 803
+    },
+    {
+      "epoch": 2.0104712041884816,
+      "grad_norm": 0.17527499794960022,
+      "learning_rate": 6.593959731543624e-05,
+      "loss": 0.3381,
+      "step": 804
+    },
+    {
+      "epoch": 2.012971790263343,
+      "grad_norm": 0.17427921295166016,
+      "learning_rate": 6.577181208053692e-05,
+      "loss": 0.3543,
+      "step": 805
+    },
+    {
+      "epoch": 2.015472376338204,
+      "grad_norm": 0.18612544238567352,
+      "learning_rate": 6.560402684563759e-05,
+      "loss": 0.3503,
+      "step": 806
+    },
+    {
+      "epoch": 2.0179729624130656,
+      "grad_norm": 0.1776886284351349,
+      "learning_rate": 6.543624161073825e-05,
+      "loss": 0.3465,
+      "step": 807
+    },
+    {
+      "epoch": 2.020473548487927,
+      "grad_norm": 0.18326617777347565,
+      "learning_rate": 6.526845637583892e-05,
+      "loss": 0.3565,
+      "step": 808
+    },
+    {
+      "epoch": 2.022974134562788,
+      "grad_norm": 0.1882122904062271,
+      "learning_rate": 6.51006711409396e-05,
+      "loss": 0.3526,
+      "step": 809
+    },
+    {
+      "epoch": 2.0254747206376496,
+      "grad_norm": 0.1967739462852478,
+      "learning_rate": 6.493288590604028e-05,
+      "loss": 0.3351,
+      "step": 810
+    },
+    {
+      "epoch": 2.0279753067125106,
+      "grad_norm": 0.18970568478107452,
+      "learning_rate": 6.476510067114095e-05,
+      "loss": 0.3461,
+      "step": 811
+    },
+    {
+      "epoch": 2.030475892787372,
+      "grad_norm": 0.18940550088882446,
+      "learning_rate": 6.459731543624161e-05,
+      "loss": 0.3501,
+      "step": 812
+    },
+    {
+      "epoch": 2.0329764788622335,
+      "grad_norm": 0.18976517021656036,
+      "learning_rate": 6.442953020134228e-05,
+      "loss": 0.3336,
+      "step": 813
+    },
+    {
+      "epoch": 2.0354770649370946,
+      "grad_norm": 0.19836163520812988,
+      "learning_rate": 6.426174496644296e-05,
+      "loss": 0.3482,
+      "step": 814
+    },
+    {
+      "epoch": 2.037977651011956,
+      "grad_norm": 0.1996692419052124,
+      "learning_rate": 6.409395973154362e-05,
+      "loss": 0.3414,
+      "step": 815
+    },
+    {
+      "epoch": 2.040478237086817,
+      "grad_norm": 0.18734270334243774,
+      "learning_rate": 6.39261744966443e-05,
+      "loss": 0.3649,
+      "step": 816
+    },
+    {
+      "epoch": 2.0429788231616786,
+      "grad_norm": 0.18595857918262482,
+      "learning_rate": 6.375838926174497e-05,
+      "loss": 0.341,
+      "step": 817
+    },
+    {
+      "epoch": 2.04547940923654,
+      "grad_norm": 0.1862795203924179,
+      "learning_rate": 6.359060402684565e-05,
+      "loss": 0.343,
+      "step": 818
+    },
+    {
+      "epoch": 2.047979995311401,
+      "grad_norm": 0.18239374458789825,
+      "learning_rate": 6.342281879194632e-05,
+      "loss": 0.3411,
+      "step": 819
+    },
+    {
+      "epoch": 2.0504805813862625,
+      "grad_norm": 0.18192386627197266,
+      "learning_rate": 6.325503355704698e-05,
+      "loss": 0.3367,
+      "step": 820
+    },
+    {
+      "epoch": 2.0529811674611236,
+      "grad_norm": 0.20054876804351807,
+      "learning_rate": 6.308724832214765e-05,
+      "loss": 0.3465,
+      "step": 821
+    },
+    {
+      "epoch": 2.055481753535985,
+      "grad_norm": 0.18257856369018555,
+      "learning_rate": 6.291946308724832e-05,
+      "loss": 0.358,
+      "step": 822
+    },
+    {
+      "epoch": 2.057982339610846,
+      "grad_norm": 0.18463286757469177,
+      "learning_rate": 6.2751677852349e-05,
+      "loss": 0.3592,
+      "step": 823
+    },
+    {
+      "epoch": 2.0604829256857076,
+      "grad_norm": 0.1811976283788681,
+      "learning_rate": 6.258389261744967e-05,
+      "loss": 0.3449,
+      "step": 824
+    },
+    {
+      "epoch": 2.062983511760569,
+      "grad_norm": 0.19554132223129272,
+      "learning_rate": 6.241610738255034e-05,
+      "loss": 0.3465,
+      "step": 825
+    },
+    {
+      "epoch": 2.06548409783543,
+      "grad_norm": 0.1971265822649002,
+      "learning_rate": 6.224832214765101e-05,
+      "loss": 0.3481,
+      "step": 826
+    },
+    {
+      "epoch": 2.0679846839102916,
+      "grad_norm": 0.1899784356355667,
+      "learning_rate": 6.208053691275169e-05,
+      "loss": 0.3437,
+      "step": 827
+    },
+    {
+      "epoch": 2.0704852699851526,
+      "grad_norm": 0.1862245500087738,
+      "learning_rate": 6.191275167785235e-05,
+      "loss": 0.3393,
+      "step": 828
+    },
+    {
+      "epoch": 2.072985856060014,
+      "grad_norm": 0.18746799230575562,
+      "learning_rate": 6.174496644295302e-05,
+      "loss": 0.363,
+      "step": 829
+    },
+    {
+      "epoch": 2.0754864421348755,
+      "grad_norm": 0.19035547971725464,
+      "learning_rate": 6.157718120805369e-05,
+      "loss": 0.3616,
+      "step": 830
+    },
+    {
+      "epoch": 2.0779870282097366,
+      "grad_norm": 0.19137556850910187,
+      "learning_rate": 6.140939597315437e-05,
+      "loss": 0.3355,
+      "step": 831
+    },
+    {
+      "epoch": 2.080487614284598,
+      "grad_norm": 0.18697211146354675,
+      "learning_rate": 6.124161073825505e-05,
+      "loss": 0.3527,
+      "step": 832
+    },
+    {
+      "epoch": 2.082988200359459,
+      "grad_norm": 0.1894364058971405,
+      "learning_rate": 6.107382550335571e-05,
+      "loss": 0.3497,
+      "step": 833
+    },
+    {
+      "epoch": 2.0854887864343206,
+      "grad_norm": 0.1954149305820465,
+      "learning_rate": 6.090604026845638e-05,
+      "loss": 0.3375,
+      "step": 834
+    },
+    {
+      "epoch": 2.087989372509182,
+      "grad_norm": 0.19052596390247345,
+      "learning_rate": 6.0738255033557044e-05,
+      "loss": 0.3281,
+      "step": 835
+    },
+    {
+      "epoch": 2.090489958584043,
+      "grad_norm": 0.18177033960819244,
+      "learning_rate": 6.0570469798657724e-05,
+      "loss": 0.3369,
+      "step": 836
+    },
+    {
+      "epoch": 2.0929905446589046,
+      "grad_norm": 0.19367891550064087,
+      "learning_rate": 6.04026845637584e-05,
+      "loss": 0.3561,
+      "step": 837
+    },
+    {
+      "epoch": 2.0954911307337656,
+      "grad_norm": 0.19525274634361267,
+      "learning_rate": 6.023489932885906e-05,
+      "loss": 0.3481,
+      "step": 838
+    },
+    {
+      "epoch": 2.097991716808627,
+      "grad_norm": 0.18661166727542877,
+      "learning_rate": 6.006711409395973e-05,
+      "loss": 0.3549,
+      "step": 839
+    },
+    {
+      "epoch": 2.100492302883488,
+      "grad_norm": 0.1898164004087448,
+      "learning_rate": 5.989932885906041e-05,
+      "loss": 0.345,
+      "step": 840
+    },
+    {
+      "epoch": 2.1029928889583496,
+      "grad_norm": 0.1943848729133606,
+      "learning_rate": 5.973154362416108e-05,
+      "loss": 0.3574,
+      "step": 841
+    },
+    {
+      "epoch": 2.105493475033211,
+      "grad_norm": 0.1996321827173233,
+      "learning_rate": 5.956375838926175e-05,
+      "loss": 0.3608,
+      "step": 842
+    },
+    {
+      "epoch": 2.107994061108072,
+      "grad_norm": 0.18836036324501038,
+      "learning_rate": 5.9395973154362415e-05,
+      "loss": 0.3296,
+      "step": 843
+    },
+    {
+      "epoch": 2.1104946471829336,
+      "grad_norm": 0.2173769474029541,
+      "learning_rate": 5.922818791946309e-05,
+      "loss": 0.35,
+      "step": 844
+    },
+    {
+      "epoch": 2.1129952332577946,
+      "grad_norm": 0.20380008220672607,
+      "learning_rate": 5.906040268456377e-05,
+      "loss": 0.3466,
+      "step": 845
+    },
+    {
+      "epoch": 2.115495819332656,
+      "grad_norm": 0.20660364627838135,
+      "learning_rate": 5.8892617449664434e-05,
+      "loss": 0.3429,
+      "step": 846
+    },
+    {
+      "epoch": 2.1179964054075175,
+      "grad_norm": 0.1951860785484314,
+      "learning_rate": 5.87248322147651e-05,
+      "loss": 0.3394,
+      "step": 847
+    },
+    {
+      "epoch": 2.1204969914823786,
+      "grad_norm": 0.19565948843955994,
+      "learning_rate": 5.855704697986577e-05,
+      "loss": 0.3403,
+      "step": 848
+    },
+    {
+      "epoch": 2.12299757755724,
+      "grad_norm": 0.1973269134759903,
+      "learning_rate": 5.838926174496645e-05,
+      "loss": 0.3263,
+      "step": 849
+    },
+    {
+      "epoch": 2.125498163632101,
+      "grad_norm": 0.19696135818958282,
+      "learning_rate": 5.822147651006712e-05,
+      "loss": 0.3463,
+      "step": 850
+    },
+    {
+      "epoch": 2.1279987497069626,
+      "grad_norm": 0.1903723180294037,
+      "learning_rate": 5.8053691275167786e-05,
+      "loss": 0.3546,
+      "step": 851
+    },
+    {
+      "epoch": 2.130499335781824,
+      "grad_norm": 0.18617288768291473,
+      "learning_rate": 5.788590604026846e-05,
+      "loss": 0.3398,
+      "step": 852
+    },
+    {
+      "epoch": 2.132999921856685,
+      "grad_norm": 0.18958479166030884,
+      "learning_rate": 5.771812080536914e-05,
+      "loss": 0.3343,
+      "step": 853
+    },
+    {
+      "epoch": 2.1355005079315466,
+      "grad_norm": 0.19173558056354523,
+      "learning_rate": 5.7550335570469805e-05,
+      "loss": 0.3451,
+      "step": 854
+    },
+    {
+      "epoch": 2.1380010940064076,
+      "grad_norm": 0.19786284863948822,
+      "learning_rate": 5.738255033557047e-05,
+      "loss": 0.3583,
+      "step": 855
+    },
+    {
+      "epoch": 2.140501680081269,
+      "grad_norm": 0.1867261677980423,
+      "learning_rate": 5.7214765100671144e-05,
+      "loss": 0.3405,
+      "step": 856
+    },
+    {
+      "epoch": 2.1430022661561305,
+      "grad_norm": 0.20245207846164703,
+      "learning_rate": 5.704697986577181e-05,
+      "loss": 0.3444,
+      "step": 857
+    },
+    {
+      "epoch": 2.1455028522309916,
+      "grad_norm": 0.1897013932466507,
+      "learning_rate": 5.687919463087249e-05,
+      "loss": 0.333,
+      "step": 858
+    },
+    {
+      "epoch": 2.148003438305853,
+      "grad_norm": 0.19288712739944458,
+      "learning_rate": 5.6711409395973156e-05,
+      "loss": 0.3509,
+      "step": 859
+    },
+    {
+      "epoch": 2.150504024380714,
+      "grad_norm": 0.2178923487663269,
+      "learning_rate": 5.654362416107383e-05,
+      "loss": 0.3313,
+      "step": 860
+    },
+    {
+      "epoch": 2.1530046104555756,
+      "grad_norm": 0.19486026465892792,
+      "learning_rate": 5.6375838926174495e-05,
+      "loss": 0.3391,
+      "step": 861
+    },
+    {
+      "epoch": 2.155505196530437,
+      "grad_norm": 0.19711101055145264,
+      "learning_rate": 5.6208053691275175e-05,
+      "loss": 0.3543,
+      "step": 862
+    },
+    {
+      "epoch": 2.158005782605298,
+      "grad_norm": 0.19923795759677887,
+      "learning_rate": 5.604026845637584e-05,
+      "loss": 0.351,
+      "step": 863
+    },
+    {
+      "epoch": 2.1605063686801596,
+      "grad_norm": 0.19494760036468506,
+      "learning_rate": 5.5872483221476514e-05,
+      "loss": 0.3337,
+      "step": 864
+    },
+    {
+      "epoch": 2.1630069547550206,
+      "grad_norm": 0.19318631291389465,
+      "learning_rate": 5.570469798657718e-05,
+      "loss": 0.341,
+      "step": 865
+    },
+    {
+      "epoch": 2.165507540829882,
+      "grad_norm": 0.1909896433353424,
+      "learning_rate": 5.553691275167785e-05,
+      "loss": 0.3431,
+      "step": 866
+    },
+    {
+      "epoch": 2.168008126904743,
+      "grad_norm": 0.20463673770427704,
+      "learning_rate": 5.536912751677853e-05,
+      "loss": 0.3391,
+      "step": 867
+    },
+    {
+      "epoch": 2.1705087129796046,
+      "grad_norm": 0.19371578097343445,
+      "learning_rate": 5.52013422818792e-05,
+      "loss": 0.3627,
+      "step": 868
+    },
+    {
+      "epoch": 2.173009299054466,
+      "grad_norm": 0.19092465937137604,
+      "learning_rate": 5.5033557046979866e-05,
+      "loss": 0.3548,
+      "step": 869
+    },
+    {
+      "epoch": 2.175509885129327,
+      "grad_norm": 0.1930968463420868,
+      "learning_rate": 5.486577181208053e-05,
+      "loss": 0.3336,
+      "step": 870
+    },
+    {
+      "epoch": 2.1780104712041886,
+      "grad_norm": 0.19714932143688202,
+      "learning_rate": 5.469798657718121e-05,
+      "loss": 0.3436,
+      "step": 871
+    },
+    {
+      "epoch": 2.1805110572790496,
+      "grad_norm": 0.18730884790420532,
+      "learning_rate": 5.4530201342281885e-05,
+      "loss": 0.3542,
+      "step": 872
+    },
+    {
+      "epoch": 2.183011643353911,
+      "grad_norm": 0.1901921033859253,
+      "learning_rate": 5.436241610738255e-05,
+      "loss": 0.3354,
+      "step": 873
+    },
+    {
+      "epoch": 2.1855122294287725,
+      "grad_norm": 0.19920407235622406,
+      "learning_rate": 5.419463087248322e-05,
+      "loss": 0.3481,
+      "step": 874
+    },
+    {
+      "epoch": 2.1880128155036336,
+      "grad_norm": 0.1935439258813858,
+      "learning_rate": 5.40268456375839e-05,
+      "loss": 0.3267,
+      "step": 875
+    },
+    {
+      "epoch": 2.190513401578495,
+      "grad_norm": 0.19596831500530243,
+      "learning_rate": 5.385906040268457e-05,
+      "loss": 0.3539,
+      "step": 876
+    },
+    {
+      "epoch": 2.193013987653356,
+      "grad_norm": 0.19396671652793884,
+      "learning_rate": 5.3691275167785237e-05,
+      "loss": 0.3446,
+      "step": 877
+    },
+    {
+      "epoch": 2.1955145737282176,
+      "grad_norm": 0.2029576599597931,
+      "learning_rate": 5.352348993288591e-05,
+      "loss": 0.3409,
+      "step": 878
+    },
+    {
+      "epoch": 2.198015159803079,
+      "grad_norm": 0.18819305300712585,
+      "learning_rate": 5.3355704697986576e-05,
+      "loss": 0.3394,
+      "step": 879
+    },
+    {
+      "epoch": 2.20051574587794,
+      "grad_norm": 0.19793127477169037,
+      "learning_rate": 5.3187919463087256e-05,
+      "loss": 0.3532,
+      "step": 880
+    },
+    {
+      "epoch": 2.2030163319528016,
+      "grad_norm": 0.19650737941265106,
+      "learning_rate": 5.302013422818792e-05,
+      "loss": 0.349,
+      "step": 881
+    },
+    {
+      "epoch": 2.2055169180276626,
+      "grad_norm": 0.1919449269771576,
+      "learning_rate": 5.2852348993288595e-05,
+      "loss": 0.3523,
+      "step": 882
+    },
+    {
+      "epoch": 2.208017504102524,
+      "grad_norm": 0.1961842030286789,
+      "learning_rate": 5.268456375838926e-05,
+      "loss": 0.3308,
+      "step": 883
+    },
+    {
+      "epoch": 2.210518090177385,
+      "grad_norm": 0.19604089856147766,
+      "learning_rate": 5.251677852348994e-05,
+      "loss": 0.351,
+      "step": 884
+    },
+    {
+      "epoch": 2.2130186762522466,
+      "grad_norm": 0.19933487474918365,
+      "learning_rate": 5.234899328859061e-05,
+      "loss": 0.3542,
+      "step": 885
+    },
+    {
+      "epoch": 2.215519262327108,
+      "grad_norm": 0.20644378662109375,
+      "learning_rate": 5.218120805369128e-05,
+      "loss": 0.3422,
+      "step": 886
+    },
+    {
+      "epoch": 2.218019848401969,
+      "grad_norm": 0.19198018312454224,
+      "learning_rate": 5.2013422818791946e-05,
+      "loss": 0.3463,
+      "step": 887
+    },
+    {
+      "epoch": 2.2205204344768306,
+      "grad_norm": 0.19478443264961243,
+      "learning_rate": 5.184563758389261e-05,
+      "loss": 0.3477,
+      "step": 888
+    },
+    {
+      "epoch": 2.2230210205516916,
+      "grad_norm": 0.18789392709732056,
+      "learning_rate": 5.167785234899329e-05,
+      "loss": 0.3414,
+      "step": 889
+    },
+    {
+      "epoch": 2.225521606626553,
+      "grad_norm": 0.19023758172988892,
+      "learning_rate": 5.1510067114093966e-05,
+      "loss": 0.3453,
+      "step": 890
+    },
+    {
+      "epoch": 2.2280221927014145,
+      "grad_norm": 0.19009388983249664,
+      "learning_rate": 5.134228187919463e-05,
+      "loss": 0.3413,
+      "step": 891
+    },
+    {
+      "epoch": 2.2305227787762756,
+      "grad_norm": 0.18771737813949585,
+      "learning_rate": 5.11744966442953e-05,
+      "loss": 0.3342,
+      "step": 892
+    },
+    {
+      "epoch": 2.233023364851137,
+      "grad_norm": 0.19222122430801392,
+      "learning_rate": 5.100671140939598e-05,
+      "loss": 0.3529,
+      "step": 893
+    },
+    {
+      "epoch": 2.235523950925998,
+      "grad_norm": 0.19601918756961823,
+      "learning_rate": 5.083892617449665e-05,
+      "loss": 0.3489,
+      "step": 894
+    },
+    {
+      "epoch": 2.2380245370008596,
+      "grad_norm": 0.1902284473180771,
+      "learning_rate": 5.067114093959732e-05,
+      "loss": 0.362,
+      "step": 895
+    },
+    {
+      "epoch": 2.240525123075721,
+      "grad_norm": 0.18784965574741364,
+      "learning_rate": 5.050335570469798e-05,
+      "loss": 0.3373,
+      "step": 896
+    },
+    {
+      "epoch": 2.243025709150582,
+      "grad_norm": 0.19413816928863525,
+      "learning_rate": 5.033557046979866e-05,
+      "loss": 0.3481,
+      "step": 897
+    },
+    {
+      "epoch": 2.2455262952254436,
+      "grad_norm": 0.19530415534973145,
+      "learning_rate": 5.0167785234899336e-05,
+      "loss": 0.3602,
+      "step": 898
+    },
+    {
+      "epoch": 2.2480268813003046,
+      "grad_norm": 0.20132146775722504,
+      "learning_rate": 5e-05,
+      "loss": 0.335,
+      "step": 899
+    },
+    {
+      "epoch": 2.250527467375166,
+      "grad_norm": 0.19600576162338257,
+      "learning_rate": 4.983221476510067e-05,
+      "loss": 0.3382,
+      "step": 900
+    },
+    {
+      "epoch": 2.253028053450027,
+      "grad_norm": 0.19266889989376068,
+      "learning_rate": 4.966442953020135e-05,
+      "loss": 0.3372,
+      "step": 901
+    },
+    {
+      "epoch": 2.2555286395248886,
+      "grad_norm": 0.18912461400032043,
+      "learning_rate": 4.9496644295302015e-05,
+      "loss": 0.3464,
+      "step": 902
+    },
+    {
+      "epoch": 2.25802922559975,
+      "grad_norm": 0.19175934791564941,
+      "learning_rate": 4.932885906040269e-05,
+      "loss": 0.3452,
+      "step": 903
+    },
+    {
+      "epoch": 2.260529811674611,
+      "grad_norm": 0.19833925366401672,
+      "learning_rate": 4.9161073825503354e-05,
+      "loss": 0.3421,
+      "step": 904
+    },
+    {
+      "epoch": 2.2630303977494726,
+      "grad_norm": 0.1921466439962387,
+      "learning_rate": 4.8993288590604034e-05,
+      "loss": 0.34,
+      "step": 905
+    },
+    {
+      "epoch": 2.265530983824334,
+      "grad_norm": 0.200669065117836,
+      "learning_rate": 4.88255033557047e-05,
+      "loss": 0.3529,
+      "step": 906
+    },
+    {
+      "epoch": 2.268031569899195,
+      "grad_norm": 0.19716723263263702,
+      "learning_rate": 4.865771812080537e-05,
+      "loss": 0.342,
+      "step": 907
+    },
+    {
+      "epoch": 2.2705321559740566,
+      "grad_norm": 0.19816866517066956,
+      "learning_rate": 4.848993288590604e-05,
+      "loss": 0.3417,
+      "step": 908
+    },
+    {
+      "epoch": 2.2730327420489176,
+      "grad_norm": 0.1944151371717453,
+      "learning_rate": 4.832214765100672e-05,
+      "loss": 0.3445,
+      "step": 909
+    },
+    {
+      "epoch": 2.275533328123779,
+      "grad_norm": 0.19737473130226135,
+      "learning_rate": 4.8154362416107385e-05,
+      "loss": 0.3447,
+      "step": 910
+    },
+    {
+      "epoch": 2.27803391419864,
+      "grad_norm": 0.19146928191184998,
+      "learning_rate": 4.798657718120805e-05,
+      "loss": 0.3449,
+      "step": 911
+    },
+    {
+      "epoch": 2.2805345002735016,
+      "grad_norm": 0.1853460818529129,
+      "learning_rate": 4.7818791946308725e-05,
+      "loss": 0.3332,
+      "step": 912
+    },
+    {
+      "epoch": 2.283035086348363,
+      "grad_norm": 0.1998416632413864,
+      "learning_rate": 4.76510067114094e-05,
+      "loss": 0.3494,
+      "step": 913
+    },
+    {
+      "epoch": 2.285535672423224,
+      "grad_norm": 0.1972053200006485,
+      "learning_rate": 4.748322147651007e-05,
+      "loss": 0.3512,
+      "step": 914
+    },
+    {
+      "epoch": 2.2880362584980856,
+      "grad_norm": 0.18613013625144958,
+      "learning_rate": 4.731543624161074e-05,
+      "loss": 0.3401,
+      "step": 915
+    },
+    {
+      "epoch": 2.2905368445729466,
+      "grad_norm": 0.18879657983779907,
+      "learning_rate": 4.714765100671141e-05,
+      "loss": 0.341,
+      "step": 916
+    },
+    {
+      "epoch": 2.293037430647808,
+      "grad_norm": 0.19808343052864075,
+      "learning_rate": 4.697986577181208e-05,
+      "loss": 0.3312,
+      "step": 917
+    },
+    {
+      "epoch": 2.2955380167226695,
+      "grad_norm": 0.19908861815929413,
+      "learning_rate": 4.6812080536912756e-05,
+      "loss": 0.357,
+      "step": 918
+    },
+    {
+      "epoch": 2.2980386027975306,
+      "grad_norm": 0.19145451486110687,
+      "learning_rate": 4.664429530201342e-05,
+      "loss": 0.3318,
+      "step": 919
+    },
+    {
+      "epoch": 2.300539188872392,
+      "grad_norm": 0.19837158918380737,
+      "learning_rate": 4.6476510067114095e-05,
+      "loss": 0.3462,
+      "step": 920
+    },
+    {
+      "epoch": 2.303039774947253,
+      "grad_norm": 0.1906367689371109,
+      "learning_rate": 4.630872483221477e-05,
+      "loss": 0.3366,
+      "step": 921
+    },
+    {
+      "epoch": 2.3055403610221146,
+      "grad_norm": 0.20077019929885864,
+      "learning_rate": 4.6140939597315434e-05,
+      "loss": 0.3573,
+      "step": 922
+    },
+    {
+      "epoch": 2.308040947096976,
+      "grad_norm": 0.19852228462696075,
+      "learning_rate": 4.597315436241611e-05,
+      "loss": 0.3375,
+      "step": 923
+    },
+    {
+      "epoch": 2.310541533171837,
+      "grad_norm": 0.2120746225118637,
+      "learning_rate": 4.580536912751678e-05,
+      "loss": 0.341,
+      "step": 924
+    },
+    {
+      "epoch": 2.3130421192466986,
+      "grad_norm": 0.19465214014053345,
+      "learning_rate": 4.5637583892617453e-05,
+      "loss": 0.3519,
+      "step": 925
+    },
+    {
+      "epoch": 2.3155427053215596,
+      "grad_norm": 0.20303763449192047,
+      "learning_rate": 4.546979865771812e-05,
+      "loss": 0.3454,
+      "step": 926
+    },
+    {
+      "epoch": 2.318043291396421,
+      "grad_norm": 0.19843918085098267,
+      "learning_rate": 4.530201342281879e-05,
+      "loss": 0.3476,
+      "step": 927
+    },
+    {
+      "epoch": 2.320543877471282,
+      "grad_norm": 0.19577498733997345,
+      "learning_rate": 4.5134228187919466e-05,
+      "loss": 0.3438,
+      "step": 928
+    },
+    {
+      "epoch": 2.3230444635461436,
+      "grad_norm": 0.1818462759256363,
+      "learning_rate": 4.496644295302014e-05,
+      "loss": 0.3245,
+      "step": 929
+    },
+    {
+      "epoch": 2.325545049621005,
+      "grad_norm": 0.1994531899690628,
+      "learning_rate": 4.4798657718120805e-05,
+      "loss": 0.349,
+      "step": 930
+    },
+    {
+      "epoch": 2.328045635695866,
+      "grad_norm": 0.19218860566616058,
+      "learning_rate": 4.463087248322148e-05,
+      "loss": 0.3319,
+      "step": 931
+    },
+    {
+      "epoch": 2.3305462217707276,
+      "grad_norm": 0.18792758882045746,
+      "learning_rate": 4.446308724832215e-05,
+      "loss": 0.3438,
+      "step": 932
+    },
+    {
+      "epoch": 2.333046807845589,
+      "grad_norm": 0.18755611777305603,
+      "learning_rate": 4.4295302013422824e-05,
+      "loss": 0.3292,
+      "step": 933
+    },
+    {
+      "epoch": 2.33554739392045,
+      "grad_norm": 0.19642862677574158,
+      "learning_rate": 4.412751677852349e-05,
+      "loss": 0.3417,
+      "step": 934
+    },
+    {
+      "epoch": 2.3380479799953116,
+      "grad_norm": 0.193852499127388,
+      "learning_rate": 4.395973154362416e-05,
+      "loss": 0.3421,
+      "step": 935
+    },
+    {
+      "epoch": 2.3405485660701726,
+      "grad_norm": 0.19550983607769012,
+      "learning_rate": 4.3791946308724836e-05,
+      "loss": 0.3442,
+      "step": 936
+    },
+    {
+      "epoch": 2.343049152145034,
+      "grad_norm": 0.18619222939014435,
+      "learning_rate": 4.36241610738255e-05,
+      "loss": 0.3363,
+      "step": 937
+    },
+    {
+      "epoch": 2.345549738219895,
+      "grad_norm": 0.20551933348178864,
+      "learning_rate": 4.3456375838926176e-05,
+      "loss": 0.3573,
+      "step": 938
+    },
+    {
+      "epoch": 2.3480503242947566,
+      "grad_norm": 0.205500990152359,
+      "learning_rate": 4.328859060402685e-05,
+      "loss": 0.3425,
+      "step": 939
+    },
+    {
+      "epoch": 2.350550910369618,
+      "grad_norm": 0.194464772939682,
+      "learning_rate": 4.312080536912752e-05,
+      "loss": 0.3399,
+      "step": 940
+    },
+    {
+      "epoch": 2.353051496444479,
+      "grad_norm": 0.20024992525577545,
+      "learning_rate": 4.295302013422819e-05,
+      "loss": 0.3632,
+      "step": 941
+    },
+    {
+      "epoch": 2.3555520825193406,
+      "grad_norm": 0.20086492598056793,
+      "learning_rate": 4.278523489932886e-05,
+      "loss": 0.3385,
+      "step": 942
+    },
+    {
+      "epoch": 2.3580526685942016,
+      "grad_norm": 0.20552556216716766,
+      "learning_rate": 4.2617449664429534e-05,
+      "loss": 0.341,
+      "step": 943
+    },
+    {
+      "epoch": 2.360553254669063,
+      "grad_norm": 0.18927432596683502,
+      "learning_rate": 4.244966442953021e-05,
+      "loss": 0.3347,
+      "step": 944
+    },
+    {
+      "epoch": 2.3630538407439245,
+      "grad_norm": 0.19806334376335144,
+      "learning_rate": 4.228187919463087e-05,
+      "loss": 0.3409,
+      "step": 945
+    },
+    {
+      "epoch": 2.3655544268187856,
+      "grad_norm": 0.1992732286453247,
+      "learning_rate": 4.2114093959731546e-05,
+      "loss": 0.341,
+      "step": 946
+    },
+    {
+      "epoch": 2.368055012893647,
+      "grad_norm": 0.19874417781829834,
+      "learning_rate": 4.194630872483222e-05,
+      "loss": 0.3379,
+      "step": 947
+    },
+    {
+      "epoch": 2.370555598968508,
+      "grad_norm": 0.21121694147586823,
+      "learning_rate": 4.1778523489932886e-05,
+      "loss": 0.351,
+      "step": 948
+    },
+    {
+      "epoch": 2.3730561850433696,
+      "grad_norm": 0.18384341895580292,
+      "learning_rate": 4.161073825503356e-05,
+      "loss": 0.3239,
+      "step": 949
+    },
+    {
+      "epoch": 2.375556771118231,
+      "grad_norm": 0.1884583830833435,
+      "learning_rate": 4.144295302013423e-05,
+      "loss": 0.3383,
+      "step": 950
+    },
+    {
+      "epoch": 2.378057357193092,
+      "grad_norm": 0.20122064650058746,
+      "learning_rate": 4.1275167785234905e-05,
+      "loss": 0.3504,
+      "step": 951
+    },
+    {
+      "epoch": 2.3805579432679536,
+      "grad_norm": 0.19695253670215607,
+      "learning_rate": 4.110738255033557e-05,
+      "loss": 0.3413,
+      "step": 952
+    },
+    {
+      "epoch": 2.3830585293428146,
+      "grad_norm": 0.2073272317647934,
+      "learning_rate": 4.0939597315436244e-05,
+      "loss": 0.3434,
+      "step": 953
+    },
+    {
+      "epoch": 2.385559115417676,
+      "grad_norm": 0.19181008636951447,
+      "learning_rate": 4.077181208053692e-05,
+      "loss": 0.3557,
+      "step": 954
+    },
+    {
+      "epoch": 2.388059701492537,
+      "grad_norm": 0.20325197279453278,
+      "learning_rate": 4.060402684563759e-05,
+      "loss": 0.3281,
+      "step": 955
+    },
+    {
+      "epoch": 2.3905602875673986,
+      "grad_norm": 0.19629107415676117,
+      "learning_rate": 4.0436241610738256e-05,
+      "loss": 0.3401,
+      "step": 956
+    },
+    {
+      "epoch": 2.39306087364226,
+      "grad_norm": 0.199611634016037,
+      "learning_rate": 4.026845637583892e-05,
+      "loss": 0.3389,
+      "step": 957
+    },
+    {
+      "epoch": 2.395561459717121,
+      "grad_norm": 0.1981804221868515,
+      "learning_rate": 4.01006711409396e-05,
+      "loss": 0.3373,
+      "step": 958
+    },
+    {
+      "epoch": 2.3980620457919826,
+      "grad_norm": 0.20194140076637268,
+      "learning_rate": 3.993288590604027e-05,
+      "loss": 0.3392,
+      "step": 959
+    },
+    {
+      "epoch": 2.4005626318668436,
+      "grad_norm": 0.19979354739189148,
+      "learning_rate": 3.976510067114094e-05,
+      "loss": 0.3462,
+      "step": 960
+    },
+    {
+      "epoch": 2.403063217941705,
+      "grad_norm": 0.1924726366996765,
+      "learning_rate": 3.959731543624161e-05,
+      "loss": 0.3482,
+      "step": 961
+    },
+    {
+      "epoch": 2.4055638040165666,
+      "grad_norm": 0.19590695202350616,
+      "learning_rate": 3.942953020134229e-05,
+      "loss": 0.3577,
+      "step": 962
+    },
+    {
+      "epoch": 2.4080643900914276,
+      "grad_norm": 0.1921459585428238,
+      "learning_rate": 3.9261744966442954e-05,
+      "loss": 0.3389,
+      "step": 963
+    },
+    {
+      "epoch": 2.410564976166289,
+      "grad_norm": 0.19755838811397552,
+      "learning_rate": 3.909395973154363e-05,
+      "loss": 0.3523,
+      "step": 964
+    },
+    {
+      "epoch": 2.41306556224115,
+      "grad_norm": 0.19130590558052063,
+      "learning_rate": 3.89261744966443e-05,
+      "loss": 0.3355,
+      "step": 965
+    },
+    {
+      "epoch": 2.4155661483160116,
+      "grad_norm": 0.1915493607521057,
+      "learning_rate": 3.875838926174497e-05,
+      "loss": 0.3348,
+      "step": 966
+    },
+    {
+      "epoch": 2.418066734390873,
+      "grad_norm": 0.1937149167060852,
+      "learning_rate": 3.859060402684564e-05,
+      "loss": 0.3472,
+      "step": 967
+    },
+    {
+      "epoch": 2.420567320465734,
+      "grad_norm": 0.19370803236961365,
+      "learning_rate": 3.8422818791946305e-05,
+      "loss": 0.3503,
+      "step": 968
+    },
+    {
+      "epoch": 2.4230679065405956,
+      "grad_norm": 0.1892373114824295,
+      "learning_rate": 3.8255033557046985e-05,
+      "loss": 0.3357,
+      "step": 969
+    },
+    {
+      "epoch": 2.4255684926154566,
+      "grad_norm": 0.19388633966445923,
+      "learning_rate": 3.808724832214765e-05,
+      "loss": 0.3437,
+      "step": 970
+    },
+    {
+      "epoch": 2.428069078690318,
+      "grad_norm": 0.20671792328357697,
+      "learning_rate": 3.7919463087248324e-05,
+      "loss": 0.3645,
+      "step": 971
+    },
+    {
+      "epoch": 2.430569664765179,
+      "grad_norm": 0.19062629342079163,
+      "learning_rate": 3.775167785234899e-05,
+      "loss": 0.3419,
+      "step": 972
+    },
+    {
+      "epoch": 2.4330702508400406,
+      "grad_norm": 0.19693328440189362,
+      "learning_rate": 3.758389261744967e-05,
+      "loss": 0.3401,
+      "step": 973
+    },
+    {
+      "epoch": 2.435570836914902,
+      "grad_norm": 0.21329520642757416,
+      "learning_rate": 3.741610738255034e-05,
+      "loss": 0.3492,
+      "step": 974
+    },
+    {
+      "epoch": 2.438071422989763,
+      "grad_norm": 0.20143064856529236,
+      "learning_rate": 3.724832214765101e-05,
+      "loss": 0.3502,
+      "step": 975
+    },
+    {
+      "epoch": 2.4405720090646246,
+      "grad_norm": 0.19883887469768524,
+      "learning_rate": 3.7080536912751676e-05,
+      "loss": 0.3449,
+      "step": 976
+    },
+    {
+      "epoch": 2.443072595139486,
+      "grad_norm": 0.19424356520175934,
+      "learning_rate": 3.6912751677852356e-05,
+      "loss": 0.3529,
+      "step": 977
+    },
+    {
+      "epoch": 2.445573181214347,
+      "grad_norm": 0.19794312119483948,
+      "learning_rate": 3.674496644295302e-05,
+      "loss": 0.3501,
+      "step": 978
+    },
+    {
+      "epoch": 2.4480737672892086,
+      "grad_norm": 0.21012213826179504,
+      "learning_rate": 3.6577181208053695e-05,
+      "loss": 0.3361,
+      "step": 979
+    },
+    {
+      "epoch": 2.4505743533640696,
+      "grad_norm": 0.20088094472885132,
+      "learning_rate": 3.640939597315436e-05,
+      "loss": 0.3516,
+      "step": 980
+    },
+    {
+      "epoch": 2.453074939438931,
+      "grad_norm": 0.19473795592784882,
+      "learning_rate": 3.6241610738255034e-05,
+      "loss": 0.3343,
+      "step": 981
+    },
+    {
+      "epoch": 2.455575525513792,
+      "grad_norm": 0.20544959604740143,
+      "learning_rate": 3.607382550335571e-05,
+      "loss": 0.3368,
+      "step": 982
+    },
+    {
+      "epoch": 2.4580761115886536,
+      "grad_norm": 0.19593705236911774,
+      "learning_rate": 3.5906040268456373e-05,
+      "loss": 0.3327,
+      "step": 983
+    },
+    {
+      "epoch": 2.460576697663515,
+      "grad_norm": 0.1958618313074112,
+      "learning_rate": 3.5738255033557046e-05,
+      "loss": 0.3427,
+      "step": 984
+    },
+    {
+      "epoch": 2.463077283738376,
+      "grad_norm": 0.1891411989927292,
+      "learning_rate": 3.557046979865772e-05,
+      "loss": 0.3384,
+      "step": 985
+    },
+    {
+      "epoch": 2.4655778698132376,
+      "grad_norm": 0.20175737142562866,
+      "learning_rate": 3.540268456375839e-05,
+      "loss": 0.3446,
+      "step": 986
+    },
+    {
+      "epoch": 2.4680784558880986,
+      "grad_norm": 0.19774186611175537,
+      "learning_rate": 3.523489932885906e-05,
+      "loss": 0.3354,
+      "step": 987
+    },
+    {
+      "epoch": 2.47057904196296,
+      "grad_norm": 0.19549371302127838,
+      "learning_rate": 3.506711409395974e-05,
+      "loss": 0.3324,
+      "step": 988
+    },
+    {
+      "epoch": 2.4730796280378216,
+      "grad_norm": 0.20362330973148346,
+      "learning_rate": 3.4899328859060405e-05,
+      "loss": 0.355,
+      "step": 989
+    },
+    {
+      "epoch": 2.4755802141126826,
+      "grad_norm": 0.18917375802993774,
+      "learning_rate": 3.473154362416108e-05,
+      "loss": 0.3391,
+      "step": 990
+    },
+    {
+      "epoch": 2.478080800187544,
+      "grad_norm": 0.19393205642700195,
+      "learning_rate": 3.4563758389261744e-05,
+      "loss": 0.3429,
+      "step": 991
+    },
+    {
+      "epoch": 2.480581386262405,
+      "grad_norm": 0.20192348957061768,
+      "learning_rate": 3.439597315436242e-05,
+      "loss": 0.3403,
+      "step": 992
+    },
+    {
+      "epoch": 2.4830819723372666,
+      "grad_norm": 0.19449514150619507,
+      "learning_rate": 3.422818791946309e-05,
+      "loss": 0.3403,
+      "step": 993
+    },
+    {
+      "epoch": 2.485582558412128,
+      "grad_norm": 0.19650086760520935,
+      "learning_rate": 3.4060402684563756e-05,
+      "loss": 0.3478,
+      "step": 994
+    },
+    {
+      "epoch": 2.488083144486989,
+      "grad_norm": 0.1927471160888672,
+      "learning_rate": 3.389261744966443e-05,
+      "loss": 0.3344,
+      "step": 995
+    },
+    {
+      "epoch": 2.4905837305618506,
+      "grad_norm": 0.19746330380439758,
+      "learning_rate": 3.37248322147651e-05,
+      "loss": 0.3453,
+      "step": 996
+    },
+    {
+      "epoch": 2.4930843166367116,
+      "grad_norm": 0.19895830750465393,
+      "learning_rate": 3.3557046979865775e-05,
+      "loss": 0.353,
+      "step": 997
+    },
+    {
+      "epoch": 2.495584902711573,
+      "grad_norm": 0.19538374245166779,
+      "learning_rate": 3.338926174496644e-05,
+      "loss": 0.3429,
+      "step": 998
+    },
+    {
+      "epoch": 2.498085488786434,
+      "grad_norm": 0.20408596098423004,
+      "learning_rate": 3.3221476510067115e-05,
+      "loss": 0.3537,
+      "step": 999
+    },
+    {
+      "epoch": 2.5005860748612956,
+      "grad_norm": 0.18683022260665894,
+      "learning_rate": 3.305369127516779e-05,
+      "loss": 0.3372,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1197,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.0860708045154796e+18,
+  "train_batch_size": 5,
+  "trial_name": null,
+  "trial_params": null
+}