diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8412 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.993201531608971,
+  "eval_steps": 500,
+  "global_step": 1197,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0025005860748612955,
+      "grad_norm": 0.9269847273826599,
+      "learning_rate": 4e-05,
+      "loss": 2.5172,
+      "step": 1
+    },
+    {
+      "epoch": 0.005001172149722591,
+      "grad_norm": 0.8822197318077087,
+      "learning_rate": 8e-05,
+      "loss": 2.5129,
+      "step": 2
+    },
+    {
+      "epoch": 0.007501758224583887,
+      "grad_norm": 0.9105328917503357,
+      "learning_rate": 0.00012,
+      "loss": 2.4871,
+      "step": 3
+    },
+    {
+      "epoch": 0.010002344299445182,
+      "grad_norm": 1.040933609008789,
+      "learning_rate": 0.00016,
+      "loss": 2.4756,
+      "step": 4
+    },
+    {
+      "epoch": 0.012502930374306479,
+      "grad_norm": 0.9631335735321045,
+      "learning_rate": 0.0002,
+      "loss": 2.2208,
+      "step": 5
+    },
+    {
+      "epoch": 0.015003516449167774,
+      "grad_norm": 0.9249287247657776,
+      "learning_rate": 0.0001998322147651007,
+      "loss": 1.9561,
+      "step": 6
+    },
+    {
+      "epoch": 0.01750410252402907,
+      "grad_norm": 1.0489821434020996,
+      "learning_rate": 0.00019966442953020134,
+      "loss": 1.7816,
+      "step": 7
+    },
+    {
+      "epoch": 0.020004688598890364,
+      "grad_norm": 0.7718613147735596,
+      "learning_rate": 0.00019949664429530202,
+      "loss": 1.5392,
+      "step": 8
+    },
+    {
+      "epoch": 0.02250527467375166,
+      "grad_norm": 0.6948288679122925,
+      "learning_rate": 0.00019932885906040267,
+      "loss": 1.3313,
+      "step": 9
+    },
+    {
+      "epoch": 0.025005860748612958,
+      "grad_norm": 0.4082912504673004,
+      "learning_rate": 0.00019916107382550338,
+      "loss": 1.284,
+      "step": 10
+    },
+    {
+      "epoch": 0.02750644682347425,
+      "grad_norm": 0.39728468656539917,
+      "learning_rate": 0.00019899328859060403,
+      "loss": 1.263,
+      "step": 11
+    },
+    {
+      "epoch": 0.030007032898335548,
+      "grad_norm": 0.44735899567604065,
+      "learning_rate": 0.00019882550335570471,
+      "loss": 1.2095,
+      "step": 12
+    },
+    {
+      "epoch": 0.032507618973196845,
+      "grad_norm": 0.5090083479881287,
+      "learning_rate": 0.0001986577181208054,
+      "loss": 1.2178,
+      "step": 13
+    },
+    {
+      "epoch": 0.03500820504805814,
+      "grad_norm": 0.5362133979797363,
+      "learning_rate": 0.00019848993288590605,
+      "loss": 1.2615,
+      "step": 14
+    },
+    {
+      "epoch": 0.03750879112291943,
+      "grad_norm": 0.6663621664047241,
+      "learning_rate": 0.00019832214765100673,
+      "loss": 1.1609,
+      "step": 15
+    },
+    {
+      "epoch": 0.04000937719778073,
+      "grad_norm": 0.8338614106178284,
+      "learning_rate": 0.00019815436241610738,
+      "loss": 1.0952,
+      "step": 16
+    },
+    {
+      "epoch": 0.042509963272642025,
+      "grad_norm": 0.8141220211982727,
+      "learning_rate": 0.00019798657718120806,
+      "loss": 1.1661,
+      "step": 17
+    },
+    {
+      "epoch": 0.04501054934750332,
+      "grad_norm": 0.8922987580299377,
+      "learning_rate": 0.0001978187919463087,
+      "loss": 1.1088,
+      "step": 18
+    },
+    {
+      "epoch": 0.04751113542236462,
+      "grad_norm": 0.4849202334880829,
+      "learning_rate": 0.00019765100671140942,
+      "loss": 1.0611,
+      "step": 19
+    },
+    {
+      "epoch": 0.050011721497225915,
+      "grad_norm": 0.20992806553840637,
+      "learning_rate": 0.00019748322147651007,
+      "loss": 1.1372,
+      "step": 20
+    },
+    {
+      "epoch": 0.052512307572087205,
+      "grad_norm": 0.20208662748336792,
+      "learning_rate": 0.00019731543624161075,
+      "loss": 1.1255,
+      "step": 21
+    },
+    {
+      "epoch": 0.0550128936469485,
+      "grad_norm": 0.19485554099082947,
+      "learning_rate": 0.00019714765100671143,
+      "loss": 1.1028,
+      "step": 22
+    },
+    {
+      "epoch": 0.0575134797218098,
+      "grad_norm": 0.2088543176651001,
+      "learning_rate": 0.00019697986577181208,
+      "loss": 1.0978,
+      "step": 23
+    },
+    {
+      "epoch": 0.060014065796671096,
+      "grad_norm": 0.21725241839885712,
+      "learning_rate": 0.00019681208053691276,
+      "loss": 1.0679,
+      "step": 24
+    },
+    {
+      "epoch": 0.06251465187153239,
+      "grad_norm": 0.21685351431369781,
+      "learning_rate": 0.00019664429530201342,
+      "loss": 1.08,
+      "step": 25
+    },
+    {
+      "epoch": 0.06501523794639369,
+      "grad_norm": 0.22109799087047577,
+      "learning_rate": 0.0001964765100671141,
+      "loss": 1.0804,
+      "step": 26
+    },
+    {
+      "epoch": 0.06751582402125499,
+      "grad_norm": 0.1903853863477707,
+      "learning_rate": 0.00019630872483221478,
+      "loss": 1.0673,
+      "step": 27
+    },
+    {
+      "epoch": 0.07001641009611628,
+      "grad_norm": 0.2121160328388214,
+      "learning_rate": 0.00019614093959731546,
+      "loss": 1.029,
+      "step": 28
+    },
+    {
+      "epoch": 0.07251699617097758,
+      "grad_norm": 0.17948715388774872,
+      "learning_rate": 0.00019597315436241613,
+      "loss": 1.0788,
+      "step": 29
+    },
+    {
+      "epoch": 0.07501758224583886,
+      "grad_norm": 0.18863339722156525,
+      "learning_rate": 0.0001958053691275168,
+      "loss": 1.1238,
+      "step": 30
+    },
+    {
+      "epoch": 0.07751816832070016,
+      "grad_norm": 0.18628939986228943,
+      "learning_rate": 0.00019563758389261747,
+      "loss": 1.0934,
+      "step": 31
+    },
+    {
+      "epoch": 0.08001875439556146,
+      "grad_norm": 0.16961562633514404,
+      "learning_rate": 0.00019546979865771812,
+      "loss": 0.983,
+      "step": 32
+    },
+    {
+      "epoch": 0.08251934047042275,
+      "grad_norm": 0.19999153912067413,
+      "learning_rate": 0.0001953020134228188,
+      "loss": 1.0447,
+      "step": 33
+    },
+    {
+      "epoch": 0.08501992654528405,
+      "grad_norm": 0.18386781215667725,
+      "learning_rate": 0.00019513422818791945,
+      "loss": 1.1093,
+      "step": 34
+    },
+    {
+      "epoch": 0.08752051262014535,
+      "grad_norm": 0.19115187227725983,
+      "learning_rate": 0.00019496644295302013,
+      "loss": 1.0573,
+      "step": 35
+    },
+    {
+      "epoch": 0.09002109869500664,
+      "grad_norm": 0.17980585992336273,
+      "learning_rate": 0.0001947986577181208,
+      "loss": 1.0733,
+      "step": 36
+    },
+    {
+      "epoch": 0.09252168476986794,
+      "grad_norm": 0.17686498165130615,
+      "learning_rate": 0.0001946308724832215,
+      "loss": 1.1531,
+      "step": 37
+    },
+    {
+      "epoch": 0.09502227084472924,
+      "grad_norm": 0.19470718502998352,
+      "learning_rate": 0.00019446308724832217,
+      "loss": 1.134,
+      "step": 38
+    },
+    {
+      "epoch": 0.09752285691959053,
+      "grad_norm": 0.2018202841281891,
+      "learning_rate": 0.00019429530201342282,
+      "loss": 1.0134,
+      "step": 39
+    },
+    {
+      "epoch": 0.10002344299445183,
+      "grad_norm": 0.19689340889453888,
+      "learning_rate": 0.0001941275167785235,
+      "loss": 1.0327,
+      "step": 40
+    },
+    {
+      "epoch": 0.10252402906931311,
+      "grad_norm": 0.19733501970767975,
+      "learning_rate": 0.00019395973154362416,
+      "loss": 1.0542,
+      "step": 41
+    },
+    {
+      "epoch": 0.10502461514417441,
+      "grad_norm": 0.20530100166797638,
+      "learning_rate": 0.00019379194630872484,
+      "loss": 1.0185,
+      "step": 42
+    },
+    {
+      "epoch": 0.10752520121903571,
+      "grad_norm": 0.20656965672969818,
+      "learning_rate": 0.00019362416107382552,
+      "loss": 0.9785,
+      "step": 43
+    },
+    {
+      "epoch": 0.110025787293897,
+      "grad_norm": 0.1925908923149109,
+      "learning_rate": 0.00019345637583892617,
+      "loss": 1.0632,
+      "step": 44
+    },
+    {
+      "epoch": 0.1125263733687583,
+      "grad_norm": 0.1836056113243103,
+      "learning_rate": 0.00019328859060402688,
+      "loss": 1.0046,
+      "step": 45
+    },
+    {
+      "epoch": 0.1150269594436196,
+      "grad_norm": 0.1750098317861557,
+      "learning_rate": 0.00019312080536912753,
+      "loss": 1.0088,
+      "step": 46
+    },
+    {
+      "epoch": 0.1175275455184809,
+      "grad_norm": 0.17450976371765137,
+      "learning_rate": 0.0001929530201342282,
+      "loss": 1.0329,
+      "step": 47
+    },
+    {
+      "epoch": 0.12002813159334219,
+      "grad_norm": 0.17772209644317627,
+      "learning_rate": 0.00019278523489932886,
+      "loss": 1.0071,
+      "step": 48
+    },
+    {
+      "epoch": 0.12252871766820349,
+      "grad_norm": 0.19410440325737,
+      "learning_rate": 0.00019261744966442954,
+      "loss": 1.0533,
+      "step": 49
+    },
+    {
+      "epoch": 0.12502930374306478,
+      "grad_norm": 0.18247346580028534,
+      "learning_rate": 0.00019244966442953022,
+      "loss": 0.9578,
+      "step": 50
+    },
+    {
+      "epoch": 0.12752988981792607,
+      "grad_norm": 0.19110924005508423,
+      "learning_rate": 0.00019228187919463087,
+      "loss": 1.0692,
+      "step": 51
+    },
+    {
+      "epoch": 0.13003047589278738,
+      "grad_norm": 0.19710960984230042,
+      "learning_rate": 0.00019211409395973155,
+      "loss": 1.0528,
+      "step": 52
+    },
+    {
+      "epoch": 0.13253106196764866,
+      "grad_norm": 0.19629694521427155,
+      "learning_rate": 0.0001919463087248322,
+      "loss": 0.976,
+      "step": 53
+    },
+    {
+      "epoch": 0.13503164804250997,
+      "grad_norm": 0.19290994107723236,
+      "learning_rate": 0.0001917785234899329,
+      "loss": 1.0599,
+      "step": 54
+    },
+    {
+      "epoch": 0.13753223411737125,
+      "grad_norm": 0.18594202399253845,
+      "learning_rate": 0.00019161073825503357,
+      "loss": 1.0428,
+      "step": 55
+    },
+    {
+      "epoch": 0.14003282019223257,
+      "grad_norm": 0.19421391189098358,
+      "learning_rate": 0.00019144295302013425,
+      "loss": 1.0367,
+      "step": 56
+    },
+    {
+      "epoch": 0.14253340626709385,
+      "grad_norm": 0.19107426702976227,
+      "learning_rate": 0.0001912751677852349,
+      "loss": 1.0337,
+      "step": 57
+    },
+    {
+      "epoch": 0.14503399234195516,
+      "grad_norm": 0.19118137657642365,
+      "learning_rate": 0.00019110738255033558,
+      "loss": 1.0528,
+      "step": 58
+    },
+    {
+      "epoch": 0.14753457841681644,
+      "grad_norm": 0.19613830745220184,
+      "learning_rate": 0.00019093959731543626,
+      "loss": 1.0288,
+      "step": 59
+    },
+    {
+      "epoch": 0.15003516449167772,
+      "grad_norm": 0.19156797230243683,
+      "learning_rate": 0.0001907718120805369,
+      "loss": 1.0409,
+      "step": 60
+    },
+    {
+      "epoch": 0.15253575056653904,
+      "grad_norm": 0.1780826449394226,
+      "learning_rate": 0.0001906040268456376,
+      "loss": 0.9963,
+      "step": 61
+    },
+    {
+      "epoch": 0.15503633664140032,
+      "grad_norm": 0.20309200882911682,
+      "learning_rate": 0.00019043624161073827,
+      "loss": 0.9674,
+      "step": 62
+    },
+    {
+      "epoch": 0.15753692271626163,
+      "grad_norm": 0.20832936465740204,
+      "learning_rate": 0.00019026845637583895,
+      "loss": 1.0019,
+      "step": 63
+    },
+    {
+      "epoch": 0.1600375087911229,
+      "grad_norm": 0.18242955207824707,
+      "learning_rate": 0.0001901006711409396,
+      "loss": 0.9708,
+      "step": 64
+    },
+    {
+      "epoch": 0.16253809486598422,
+      "grad_norm": 0.1992724984884262,
+      "learning_rate": 0.00018993288590604028,
+      "loss": 1.0076,
+      "step": 65
+    },
+    {
+      "epoch": 0.1650386809408455,
+      "grad_norm": 0.20120841264724731,
+      "learning_rate": 0.00018976510067114096,
+      "loss": 0.9708,
+      "step": 66
+    },
+    {
+      "epoch": 0.16753926701570682,
+      "grad_norm": 0.21059167385101318,
+      "learning_rate": 0.00018959731543624161,
+      "loss": 1.0076,
+      "step": 67
+    },
+    {
+      "epoch": 0.1700398530905681,
+      "grad_norm": 0.1988890916109085,
+      "learning_rate": 0.0001894295302013423,
+      "loss": 1.0054,
+      "step": 68
+    },
+    {
+      "epoch": 0.1725404391654294,
+      "grad_norm": 0.1952708512544632,
+      "learning_rate": 0.00018926174496644295,
+      "loss": 1.0351,
+      "step": 69
+    },
+    {
+      "epoch": 0.1750410252402907,
+      "grad_norm": 0.23466959595680237,
+      "learning_rate": 0.00018909395973154363,
+      "loss": 0.9814,
+      "step": 70
+    },
+    {
+      "epoch": 0.17754161131515198,
+      "grad_norm": 0.20591197907924652,
+      "learning_rate": 0.0001889261744966443,
+      "loss": 0.9691,
+      "step": 71
+    },
+    {
+      "epoch": 0.1800421973900133,
+      "grad_norm": 0.2023579180240631,
+      "learning_rate": 0.000188758389261745,
+      "loss": 1.0003,
+      "step": 72
+    },
+    {
+      "epoch": 0.18254278346487457,
+      "grad_norm": 0.20543326437473297,
+      "learning_rate": 0.00018859060402684564,
+      "loss": 0.9683,
+      "step": 73
+    },
+    {
+      "epoch": 0.18504336953973588,
+      "grad_norm": 0.20226700603961945,
+      "learning_rate": 0.00018842281879194632,
+      "loss": 1.0036,
+      "step": 74
+    },
+    {
+      "epoch": 0.18754395561459716,
+      "grad_norm": 0.19225859642028809,
+      "learning_rate": 0.000188255033557047,
+      "loss": 0.9743,
+      "step": 75
+    },
+    {
+      "epoch": 0.19004454168945847,
+      "grad_norm": 0.21251380443572998,
+      "learning_rate": 0.00018808724832214765,
+      "loss": 1.0073,
+      "step": 76
+    },
+    {
+      "epoch": 0.19254512776431976,
+      "grad_norm": 0.2131969928741455,
+      "learning_rate": 0.00018791946308724833,
+      "loss": 0.9934,
+      "step": 77
+    },
+    {
+      "epoch": 0.19504571383918107,
+      "grad_norm": 0.19129155576229095,
+      "learning_rate": 0.00018775167785234898,
+      "loss": 1.0119,
+      "step": 78
+    },
+    {
+      "epoch": 0.19754629991404235,
+      "grad_norm": 0.20633932948112488,
+      "learning_rate": 0.00018758389261744966,
+      "loss": 1.0344,
+      "step": 79
+    },
+    {
+      "epoch": 0.20004688598890366,
+      "grad_norm": 0.21870285272598267,
+      "learning_rate": 0.00018741610738255034,
+      "loss": 1.0207,
+      "step": 80
+    },
+    {
+      "epoch": 0.20254747206376494,
+      "grad_norm": 0.2022286206483841,
+      "learning_rate": 0.00018724832214765102,
+      "loss": 1.026,
+      "step": 81
+    },
+    {
+      "epoch": 0.20504805813862623,
+      "grad_norm": 0.21065476536750793,
+      "learning_rate": 0.0001870805369127517,
+      "loss": 1.0396,
+      "step": 82
+    },
+    {
+      "epoch": 0.20754864421348754,
+      "grad_norm": 0.19675321877002716,
+      "learning_rate": 0.00018691275167785236,
+      "loss": 0.9884,
+      "step": 83
+    },
+    {
+      "epoch": 0.21004923028834882,
+      "grad_norm": 0.22069697082042694,
+      "learning_rate": 0.00018674496644295304,
+      "loss": 1.0053,
+      "step": 84
+    },
+    {
+      "epoch": 0.21254981636321013,
+      "grad_norm": 0.19039078056812286,
+      "learning_rate": 0.0001865771812080537,
+      "loss": 0.9954,
+      "step": 85
+    },
+    {
+      "epoch": 0.21505040243807141,
+      "grad_norm": 0.2229384034872055,
+      "learning_rate": 0.00018640939597315437,
+      "loss": 0.9337,
+      "step": 86
+    },
+    {
+      "epoch": 0.21755098851293272,
+      "grad_norm": 0.19707229733467102,
+      "learning_rate": 0.00018624161073825502,
+      "loss": 1.01,
+      "step": 87
+    },
+    {
+      "epoch": 0.220051574587794,
+      "grad_norm": 0.19986790418624878,
+      "learning_rate": 0.0001860738255033557,
+      "loss": 0.9801,
+      "step": 88
+    },
+    {
+      "epoch": 0.22255216066265532,
+      "grad_norm": 0.2047620266675949,
+      "learning_rate": 0.00018590604026845638,
+      "loss": 0.9507,
+      "step": 89
+    },
+    {
+      "epoch": 0.2250527467375166,
+      "grad_norm": 0.20746001601219177,
+      "learning_rate": 0.00018573825503355706,
+      "loss": 1.0103,
+      "step": 90
+    },
+    {
+      "epoch": 0.2275533328123779,
+      "grad_norm": 0.2248009592294693,
+      "learning_rate": 0.00018557046979865774,
+      "loss": 0.9869,
+      "step": 91
+    },
+    {
+      "epoch": 0.2300539188872392,
+      "grad_norm": 0.20264257490634918,
+      "learning_rate": 0.0001854026845637584,
+      "loss": 0.9913,
+      "step": 92
+    },
+    {
+      "epoch": 0.2325545049621005,
+      "grad_norm": 0.20468488335609436,
+      "learning_rate": 0.00018523489932885907,
+      "loss": 0.9607,
+      "step": 93
+    },
+    {
+      "epoch": 0.2350550910369618,
+      "grad_norm": 0.23831503093242645,
+      "learning_rate": 0.00018506711409395973,
+      "loss": 1.0247,
+      "step": 94
+    },
+    {
+      "epoch": 0.23755567711182307,
+      "grad_norm": 0.18826642632484436,
+      "learning_rate": 0.0001848993288590604,
+      "loss": 0.9482,
+      "step": 95
+    },
+    {
+      "epoch": 0.24005626318668438,
+      "grad_norm": 0.19682301580905914,
+      "learning_rate": 0.00018473154362416109,
+      "loss": 1.001,
+      "step": 96
+    },
+    {
+      "epoch": 0.24255684926154567,
+      "grad_norm": 0.2301325500011444,
+      "learning_rate": 0.00018456375838926174,
+      "loss": 1.0467,
+      "step": 97
+    },
+    {
+      "epoch": 0.24505743533640698,
+      "grad_norm": 0.20185357332229614,
+      "learning_rate": 0.00018439597315436244,
+      "loss": 0.9464,
+      "step": 98
+    },
+    {
+      "epoch": 0.24755802141126826,
+      "grad_norm": 0.2085759937763214,
+      "learning_rate": 0.0001842281879194631,
+      "loss": 1.0131,
+      "step": 99
+    },
+    {
+      "epoch": 0.25005860748612957,
+      "grad_norm": 0.20679374039173126,
+      "learning_rate": 0.00018406040268456378,
+      "loss": 1.0529,
+      "step": 100
+    },
+    {
+      "epoch": 0.2525591935609909,
+      "grad_norm": 0.21764545142650604,
+      "learning_rate": 0.00018389261744966443,
+      "loss": 1.017,
+      "step": 101
+    },
+    {
+      "epoch": 0.25505977963585214,
+      "grad_norm": 0.2044128179550171,
+      "learning_rate": 0.0001837248322147651,
+      "loss": 0.9725,
+      "step": 102
+    },
+    {
+      "epoch": 0.25756036571071345,
+      "grad_norm": 0.19335702061653137,
+      "learning_rate": 0.0001835570469798658,
+      "loss": 0.9906,
+      "step": 103
+    },
+    {
+      "epoch": 0.26006095178557476,
+      "grad_norm": 0.195275217294693,
+      "learning_rate": 0.00018338926174496644,
+      "loss": 0.967,
+      "step": 104
+    },
+    {
+      "epoch": 0.26256153786043607,
+      "grad_norm": 0.2024708092212677,
+      "learning_rate": 0.00018322147651006712,
+      "loss": 0.9881,
+      "step": 105
+    },
+    {
+      "epoch": 0.2650621239352973,
+      "grad_norm": 0.1868991106748581,
+      "learning_rate": 0.0001830536912751678,
+      "loss": 0.9886,
+      "step": 106
+    },
+    {
+      "epoch": 0.26756271001015863,
+      "grad_norm": 0.20454883575439453,
+      "learning_rate": 0.00018288590604026848,
+      "loss": 1.0121,
+      "step": 107
+    },
+    {
+      "epoch": 0.27006329608501994,
+      "grad_norm": 0.20686966180801392,
+      "learning_rate": 0.00018271812080536913,
+      "loss": 0.9569,
+      "step": 108
+    },
+    {
+      "epoch": 0.2725638821598812,
+      "grad_norm": 0.20435401797294617,
+      "learning_rate": 0.00018255033557046981,
+      "loss": 0.9742,
+      "step": 109
+    },
+    {
+      "epoch": 0.2750644682347425,
+      "grad_norm": 0.18989619612693787,
+      "learning_rate": 0.00018238255033557047,
+      "loss": 0.9616,
+      "step": 110
+    },
+    {
+      "epoch": 0.2775650543096038,
+      "grad_norm": 0.2045942097902298,
+      "learning_rate": 0.00018221476510067115,
+      "loss": 0.9302,
+      "step": 111
+    },
+    {
+      "epoch": 0.28006564038446513,
+      "grad_norm": 0.20995117723941803,
+      "learning_rate": 0.00018204697986577183,
+      "loss": 1.0072,
+      "step": 112
+    },
+    {
+      "epoch": 0.2825662264593264,
+      "grad_norm": 0.19431668519973755,
+      "learning_rate": 0.00018187919463087248,
+      "loss": 0.9932,
+      "step": 113
+    },
+    {
+      "epoch": 0.2850668125341877,
+      "grad_norm": 0.20368535816669464,
+      "learning_rate": 0.00018171140939597316,
+      "loss": 0.9882,
+      "step": 114
+    },
+    {
+      "epoch": 0.287567398609049,
+      "grad_norm": 0.1923864632844925,
+      "learning_rate": 0.00018154362416107384,
+      "loss": 0.9779,
+      "step": 115
+    },
+    {
+      "epoch": 0.2900679846839103,
+      "grad_norm": 0.19875271618366241,
+      "learning_rate": 0.00018137583892617452,
+      "loss": 0.9934,
+      "step": 116
+    },
+    {
+      "epoch": 0.2925685707587716,
+      "grad_norm": 0.21339964866638184,
+      "learning_rate": 0.00018120805369127517,
+      "loss": 0.9366,
+      "step": 117
+    },
+    {
+      "epoch": 0.2950691568336329,
+      "grad_norm": 0.20380328595638275,
+      "learning_rate": 0.00018104026845637585,
+      "loss": 0.9729,
+      "step": 118
+    },
+    {
+      "epoch": 0.2975697429084942,
+      "grad_norm": 0.20112350583076477,
+      "learning_rate": 0.00018087248322147653,
+      "loss": 0.9686,
+      "step": 119
+    },
+    {
+      "epoch": 0.30007032898335545,
+      "grad_norm": 0.2018086016178131,
+      "learning_rate": 0.00018070469798657718,
+      "loss": 1.0412,
+      "step": 120
+    },
+    {
+      "epoch": 0.30257091505821676,
+      "grad_norm": 0.20576363801956177,
+      "learning_rate": 0.00018053691275167786,
+      "loss": 0.9666,
+      "step": 121
+    },
+    {
+      "epoch": 0.30507150113307807,
+      "grad_norm": 0.22615446150302887,
+      "learning_rate": 0.00018036912751677852,
+      "loss": 0.9518,
+      "step": 122
+    },
+    {
+      "epoch": 0.3075720872079394,
+      "grad_norm": 0.19727525115013123,
+      "learning_rate": 0.0001802013422818792,
+      "loss": 1.0006,
+      "step": 123
+    },
+    {
+      "epoch": 0.31007267328280064,
+      "grad_norm": 0.20703637599945068,
+      "learning_rate": 0.00018003355704697988,
+      "loss": 0.994,
+      "step": 124
+    },
+    {
+      "epoch": 0.31257325935766195,
+      "grad_norm": 0.2170065939426422,
+      "learning_rate": 0.00017986577181208056,
+      "loss": 0.9449,
+      "step": 125
+    },
+    {
+      "epoch": 0.31507384543252326,
+      "grad_norm": 0.21106544137001038,
+      "learning_rate": 0.0001796979865771812,
+      "loss": 0.999,
+      "step": 126
+    },
+    {
+      "epoch": 0.31757443150738457,
+      "grad_norm": 0.20383739471435547,
+      "learning_rate": 0.0001795302013422819,
+      "loss": 0.9679,
+      "step": 127
+    },
+    {
+      "epoch": 0.3200750175822458,
+      "grad_norm": 0.20005188882350922,
+      "learning_rate": 0.00017936241610738257,
+      "loss": 0.9361,
+      "step": 128
+    },
+    {
+      "epoch": 0.32257560365710714,
+      "grad_norm": 0.21745333075523376,
+      "learning_rate": 0.00017919463087248322,
+      "loss": 0.9293,
+      "step": 129
+    },
+    {
+      "epoch": 0.32507618973196845,
+      "grad_norm": 0.2078651785850525,
+      "learning_rate": 0.0001790268456375839,
+      "loss": 0.9523,
+      "step": 130
+    },
+    {
+      "epoch": 0.3275767758068297,
+      "grad_norm": 0.18979056179523468,
+      "learning_rate": 0.00017885906040268455,
+      "loss": 0.9534,
+      "step": 131
+    },
+    {
+      "epoch": 0.330077361881691,
+      "grad_norm": 0.20072820782661438,
+      "learning_rate": 0.00017869127516778523,
+      "loss": 0.9605,
+      "step": 132
+    },
+    {
+      "epoch": 0.3325779479565523,
+      "grad_norm": 0.19807642698287964,
+      "learning_rate": 0.0001785234899328859,
+      "loss": 0.9626,
+      "step": 133
+    },
+    {
+      "epoch": 0.33507853403141363,
+      "grad_norm": 0.20472458004951477,
+      "learning_rate": 0.0001783557046979866,
+      "loss": 0.9659,
+      "step": 134
+    },
+    {
+      "epoch": 0.3375791201062749,
+      "grad_norm": 0.19827553629875183,
+      "learning_rate": 0.00017818791946308727,
+      "loss": 0.9452,
+      "step": 135
+    },
+    {
+      "epoch": 0.3400797061811362,
+      "grad_norm": 0.21104063093662262,
+      "learning_rate": 0.00017802013422818792,
+      "loss": 0.9595,
+      "step": 136
+    },
+    {
+      "epoch": 0.3425802922559975,
+      "grad_norm": 0.18921682238578796,
+      "learning_rate": 0.0001778523489932886,
+      "loss": 0.9619,
+      "step": 137
+    },
+    {
+      "epoch": 0.3450808783308588,
+      "grad_norm": 0.2175590544939041,
+      "learning_rate": 0.00017768456375838926,
+      "loss": 0.9695,
+      "step": 138
+    },
+    {
+      "epoch": 0.3475814644057201,
+      "grad_norm": 0.1978655904531479,
+      "learning_rate": 0.00017751677852348994,
+      "loss": 0.9493,
+      "step": 139
+    },
+    {
+      "epoch": 0.3500820504805814,
+      "grad_norm": 0.19963012635707855,
+      "learning_rate": 0.00017734899328859062,
+      "loss": 0.9141,
+      "step": 140
+    },
+    {
+      "epoch": 0.3525826365554427,
+      "grad_norm": 0.20079948008060455,
+      "learning_rate": 0.0001771812080536913,
+      "loss": 0.9649,
+      "step": 141
+    },
+    {
+      "epoch": 0.35508322263030395,
+      "grad_norm": 0.19476771354675293,
+      "learning_rate": 0.00017701342281879198,
+      "loss": 0.9056,
+      "step": 142
+    },
+    {
+      "epoch": 0.35758380870516526,
+      "grad_norm": 0.18434399366378784,
+      "learning_rate": 0.00017684563758389263,
+      "loss": 0.9166,
+      "step": 143
+    },
+    {
+      "epoch": 0.3600843947800266,
+      "grad_norm": 0.20533886551856995,
+      "learning_rate": 0.0001766778523489933,
+      "loss": 0.974,
+      "step": 144
+    },
+    {
+      "epoch": 0.3625849808548879,
+      "grad_norm": 0.2047143578529358,
+      "learning_rate": 0.00017651006711409396,
+      "loss": 0.9513,
+      "step": 145
+    },
+    {
+      "epoch": 0.36508556692974914,
+      "grad_norm": 0.19077660143375397,
+      "learning_rate": 0.00017634228187919464,
+      "loss": 0.9358,
+      "step": 146
+    },
+    {
+      "epoch": 0.36758615300461045,
+      "grad_norm": 0.21231941878795624,
+      "learning_rate": 0.0001761744966442953,
+      "loss": 0.9875,
+      "step": 147
+    },
+    {
+      "epoch": 0.37008673907947176,
+      "grad_norm": 0.20398589968681335,
+      "learning_rate": 0.00017600671140939597,
+      "loss": 0.9638,
+      "step": 148
+    },
+    {
+      "epoch": 0.37258732515433307,
+      "grad_norm": 0.1901940405368805,
+      "learning_rate": 0.00017583892617449665,
+      "loss": 0.954,
+      "step": 149
+    },
+    {
+      "epoch": 0.3750879112291943,
+      "grad_norm": 0.19586050510406494,
+      "learning_rate": 0.00017567114093959733,
+      "loss": 0.943,
+      "step": 150
+    },
+    {
+      "epoch": 0.37758849730405564,
+      "grad_norm": 0.20500780642032623,
+      "learning_rate": 0.000175503355704698,
+      "loss": 0.9939,
+      "step": 151
+    },
+    {
+      "epoch": 0.38008908337891695,
+      "grad_norm": 0.2010958045721054,
+      "learning_rate": 0.00017533557046979867,
+      "loss": 0.9963,
+      "step": 152
+    },
+    {
+      "epoch": 0.3825896694537782,
+      "grad_norm": 0.20818760991096497,
+      "learning_rate": 0.00017516778523489935,
+      "loss": 0.9368,
+      "step": 153
+    },
+    {
+      "epoch": 0.3850902555286395,
+      "grad_norm": 0.20369277894496918,
+      "learning_rate": 0.000175,
+      "loss": 0.9901,
+      "step": 154
+    },
+    {
+      "epoch": 0.3875908416035008,
+      "grad_norm": 0.20049050450325012,
+      "learning_rate": 0.00017483221476510068,
+      "loss": 0.9104,
+      "step": 155
+    },
+    {
+      "epoch": 0.39009142767836213,
+      "grad_norm": 0.20531973242759705,
+      "learning_rate": 0.00017466442953020136,
+      "loss": 0.9193,
+      "step": 156
+    },
+    {
+      "epoch": 0.3925920137532234,
+      "grad_norm": 0.19537755846977234,
+      "learning_rate": 0.000174496644295302,
+      "loss": 0.9368,
+      "step": 157
+    },
+    {
+      "epoch": 0.3950925998280847,
+      "grad_norm": 0.21875996887683868,
+      "learning_rate": 0.0001743288590604027,
+      "loss": 0.9799,
+      "step": 158
+    },
+    {
+      "epoch": 0.397593185902946,
+      "grad_norm": 0.21377162635326385,
+      "learning_rate": 0.00017416107382550337,
+      "loss": 1.0224,
+      "step": 159
+    },
+    {
+      "epoch": 0.4000937719778073,
+      "grad_norm": 0.20617148280143738,
+      "learning_rate": 0.00017399328859060405,
+      "loss": 0.9825,
+      "step": 160
+    },
+    {
+      "epoch": 0.4025943580526686,
+      "grad_norm": 0.19385607540607452,
+      "learning_rate": 0.0001738255033557047,
+      "loss": 0.9567,
+      "step": 161
+    },
+    {
+      "epoch": 0.4050949441275299,
+      "grad_norm": 0.20452259480953217,
+      "learning_rate": 0.00017365771812080538,
+      "loss": 0.8854,
+      "step": 162
+    },
+    {
+      "epoch": 0.4075955302023912,
+      "grad_norm": 0.20258085429668427,
+      "learning_rate": 0.00017348993288590604,
+      "loss": 1.0458,
+      "step": 163
+    },
+    {
+      "epoch": 0.41009611627725245,
+      "grad_norm": 0.21564531326293945,
+      "learning_rate": 0.00017332214765100671,
+      "loss": 0.9514,
+      "step": 164
+    },
+    {
+      "epoch": 0.41259670235211376,
+      "grad_norm": 0.21387049555778503,
+      "learning_rate": 0.0001731543624161074,
+      "loss": 1.0035,
+      "step": 165
+    },
+    {
+      "epoch": 0.4150972884269751,
+      "grad_norm": 0.20146381855010986,
+      "learning_rate": 0.00017298657718120805,
+      "loss": 0.9247,
+      "step": 166
+    },
+    {
+      "epoch": 0.4175978745018364,
+      "grad_norm": 0.20543105900287628,
+      "learning_rate": 0.00017281879194630873,
+      "loss": 0.9505,
+      "step": 167
+    },
+    {
+      "epoch": 0.42009846057669764,
+      "grad_norm": 0.21827350556850433,
+      "learning_rate": 0.0001726510067114094,
+      "loss": 0.9993,
+      "step": 168
+    },
+    {
+      "epoch": 0.42259904665155895,
+      "grad_norm": 0.22608278691768646,
+      "learning_rate": 0.0001724832214765101,
+      "loss": 0.9393,
+      "step": 169
+    },
+    {
+      "epoch": 0.42509963272642026,
+      "grad_norm": 0.21495342254638672,
+      "learning_rate": 0.00017231543624161074,
+      "loss": 1.0455,
+      "step": 170
+    },
+    {
+      "epoch": 0.4276002188012816,
+      "grad_norm": 0.21326063573360443,
+      "learning_rate": 0.00017214765100671142,
+      "loss": 1.017,
+      "step": 171
+    },
+    {
+      "epoch": 0.43010080487614283,
+      "grad_norm": 0.2060144692659378,
+      "learning_rate": 0.0001719798657718121,
+      "loss": 0.9671,
+      "step": 172
+    },
+    {
+      "epoch": 0.43260139095100414,
+      "grad_norm": 0.21247944235801697,
+      "learning_rate": 0.00017181208053691275,
+      "loss": 0.912,
+      "step": 173
+    },
+    {
+      "epoch": 0.43510197702586545,
+      "grad_norm": 0.20313112437725067,
+      "learning_rate": 0.00017164429530201343,
+      "loss": 1.021,
+      "step": 174
+    },
+    {
+      "epoch": 0.43760256310072676,
+      "grad_norm": 0.20316354930400848,
+      "learning_rate": 0.00017147651006711408,
+      "loss": 0.9542,
+      "step": 175
+    },
+    {
+      "epoch": 0.440103149175588,
+      "grad_norm": 0.19503550231456757,
+      "learning_rate": 0.0001713087248322148,
+      "loss": 0.9495,
+      "step": 176
+    },
+    {
+      "epoch": 0.4426037352504493,
+      "grad_norm": 0.19268807768821716,
+      "learning_rate": 0.00017114093959731544,
+      "loss": 0.9913,
+      "step": 177
+    },
+    {
+      "epoch": 0.44510432132531064,
+      "grad_norm": 0.19619421660900116,
+      "learning_rate": 0.00017097315436241612,
+      "loss": 0.9136,
+      "step": 178
+    },
+    {
+      "epoch": 0.4476049074001719,
+      "grad_norm": 0.21306179463863373,
+      "learning_rate": 0.0001708053691275168,
+      "loss": 0.9468,
+      "step": 179
+    },
+    {
+      "epoch": 0.4501054934750332,
+      "grad_norm": 0.1926979422569275,
+      "learning_rate": 0.00017063758389261746,
+      "loss": 0.9138,
+      "step": 180
+    },
+    {
+      "epoch": 0.4526060795498945,
+      "grad_norm": 0.20351314544677734,
+      "learning_rate": 0.00017046979865771814,
+      "loss": 0.9424,
+      "step": 181
+    },
+    {
+      "epoch": 0.4551066656247558,
+      "grad_norm": 0.20273157954216003,
+      "learning_rate": 0.0001703020134228188,
+      "loss": 0.9901,
+      "step": 182
+    },
+    {
+      "epoch": 0.4576072516996171,
+      "grad_norm": 0.19532570242881775,
+      "learning_rate": 0.00017013422818791947,
+      "loss": 0.9278,
+      "step": 183
+    },
+    {
+      "epoch": 0.4601078377744784,
+      "grad_norm": 0.2106495201587677,
+      "learning_rate": 0.00016996644295302012,
+      "loss": 0.9103,
+      "step": 184
+    },
+    {
+      "epoch": 0.4626084238493397,
+      "grad_norm": 0.19396382570266724,
+      "learning_rate": 0.00016979865771812083,
+      "loss": 0.9493,
+      "step": 185
+    },
+    {
+      "epoch": 0.465109009924201,
+      "grad_norm": 0.19334650039672852,
+      "learning_rate": 0.00016963087248322148,
+      "loss": 0.9813,
+      "step": 186
+    },
+    {
+      "epoch": 0.46760959599906227,
+      "grad_norm": 0.20727260410785675,
+      "learning_rate": 0.00016946308724832216,
+      "loss": 0.9721,
+      "step": 187
+    },
+    {
+      "epoch": 0.4701101820739236,
+      "grad_norm": 0.1958143711090088,
+      "learning_rate": 0.00016929530201342284,
+      "loss": 0.975,
+      "step": 188
+    },
+    {
+      "epoch": 0.4726107681487849,
+      "grad_norm": 0.2058304399251938,
+      "learning_rate": 0.0001691275167785235,
+      "loss": 0.9718,
+      "step": 189
+    },
+    {
+      "epoch": 0.47511135422364614,
+      "grad_norm": 0.19323155283927917,
+      "learning_rate": 0.00016895973154362417,
+      "loss": 0.933,
+      "step": 190
+    },
+    {
+      "epoch": 0.47761194029850745,
+      "grad_norm": 0.2087579071521759,
+      "learning_rate": 0.00016879194630872483,
+      "loss": 0.9643,
+      "step": 191
+    },
+    {
+      "epoch": 0.48011252637336876,
+      "grad_norm": 0.19288691878318787,
+      "learning_rate": 0.0001686241610738255,
+      "loss": 0.9452,
+      "step": 192
+    },
+    {
+      "epoch": 0.4826131124482301,
+      "grad_norm": 0.20016451179981232,
+      "learning_rate": 0.00016845637583892619,
+      "loss": 1.0201,
+      "step": 193
+    },
+    {
+      "epoch": 0.48511369852309133,
+      "grad_norm": 0.20064382255077362,
+      "learning_rate": 0.00016828859060402686,
+      "loss": 0.9329,
+      "step": 194
+    },
+    {
+      "epoch": 0.48761428459795264,
+      "grad_norm": 0.20682549476623535,
+      "learning_rate": 0.00016812080536912754,
+      "loss": 0.9067,
+      "step": 195
+    },
+    {
+      "epoch": 0.49011487067281395,
+      "grad_norm": 0.2095707803964615,
+      "learning_rate": 0.0001679530201342282,
+      "loss": 0.964,
+      "step": 196
+    },
+    {
+      "epoch": 0.49261545674767526,
+      "grad_norm": 0.19708895683288574,
+      "learning_rate": 0.00016778523489932888,
+      "loss": 0.9555,
+      "step": 197
+    },
+    {
+      "epoch": 0.4951160428225365,
+      "grad_norm": 0.21263785660266876,
+      "learning_rate": 0.00016761744966442953,
+      "loss": 0.9607,
+      "step": 198
+    },
+    {
+      "epoch": 0.49761662889739783,
+      "grad_norm": 0.2375573068857193,
+      "learning_rate": 0.0001674496644295302,
+      "loss": 0.93,
+      "step": 199
+    },
+    {
+      "epoch": 0.5001172149722591,
+      "grad_norm": 0.19389759004116058,
+      "learning_rate": 0.00016728187919463086,
+      "loss": 0.9339,
+      "step": 200
+    },
+    {
+      "epoch": 0.5026178010471204,
+      "grad_norm": 0.21175242960453033,
+      "learning_rate": 0.00016711409395973154,
+      "loss": 0.9171,
+      "step": 201
+    },
+    {
+      "epoch": 0.5051183871219818,
+      "grad_norm": 0.21076923608779907,
+      "learning_rate": 0.00016694630872483222,
+      "loss": 0.872,
+      "step": 202
+    },
+    {
+      "epoch": 0.507618973196843,
+      "grad_norm": 0.20474554598331451,
+      "learning_rate": 0.0001667785234899329,
+      "loss": 0.9452,
+      "step": 203
+    },
+    {
+      "epoch": 0.5101195592717043,
+      "grad_norm": 0.21046273410320282,
+      "learning_rate": 0.00016661073825503358,
+      "loss": 0.9573,
+      "step": 204
+    },
+    {
+      "epoch": 0.5126201453465656,
+      "grad_norm": 0.2026781141757965,
+      "learning_rate": 0.00016644295302013423,
+      "loss": 0.9899,
+      "step": 205
+    },
+    {
+      "epoch": 0.5151207314214269,
+      "grad_norm": 0.22526352107524872,
+      "learning_rate": 0.00016627516778523491,
+      "loss": 0.9826,
+      "step": 206
+    },
+    {
+      "epoch": 0.5176213174962881,
+      "grad_norm": 0.20340263843536377,
+      "learning_rate": 0.00016610738255033557,
+      "loss": 0.9356,
+      "step": 207
+    },
+    {
+      "epoch": 0.5201219035711495,
+      "grad_norm": 0.1983516812324524,
+      "learning_rate": 0.00016593959731543625,
+      "loss": 0.9663,
+      "step": 208
+    },
+    {
+      "epoch": 0.5226224896460108,
+      "grad_norm": 0.19852115213871002,
+      "learning_rate": 0.00016577181208053693,
+      "loss": 0.97,
+      "step": 209
+    },
+    {
+      "epoch": 0.5251230757208721,
+      "grad_norm": 0.20500719547271729,
+      "learning_rate": 0.00016560402684563758,
+      "loss": 0.9304,
+      "step": 210
+    },
+    {
+      "epoch": 0.5276236617957334,
+      "grad_norm": 0.2010101079940796,
+      "learning_rate": 0.00016543624161073829,
+      "loss": 0.9102,
+      "step": 211
+    },
+    {
+      "epoch": 0.5301242478705946,
+      "grad_norm": 0.21323886513710022,
+      "learning_rate": 0.00016526845637583894,
+      "loss": 0.9794,
+      "step": 212
+    },
+    {
+      "epoch": 0.532624833945456,
+      "grad_norm": 0.20126739144325256,
+      "learning_rate": 0.00016510067114093962,
+      "loss": 0.9326,
+      "step": 213
+    },
+    {
+      "epoch": 0.5351254200203173,
+      "grad_norm": 0.2181158810853958,
+      "learning_rate": 0.00016493288590604027,
+      "loss": 0.9925,
+      "step": 214
+    },
+    {
+      "epoch": 0.5376260060951785,
+      "grad_norm": 0.2002563178539276,
+      "learning_rate": 0.00016476510067114095,
+      "loss": 0.9283,
+      "step": 215
+    },
+    {
+      "epoch": 0.5401265921700399,
+      "grad_norm": 0.1901605725288391,
+      "learning_rate": 0.0001645973154362416,
+      "loss": 0.9481,
+      "step": 216
+    },
+    {
+      "epoch": 0.5426271782449011,
+      "grad_norm": 0.20211246609687805,
+      "learning_rate": 0.00016442953020134228,
+      "loss": 0.9343,
+      "step": 217
+    },
+    {
+      "epoch": 0.5451277643197624,
+      "grad_norm": 0.19748587906360626,
+      "learning_rate": 0.00016426174496644296,
+      "loss": 0.9967,
+      "step": 218
+    },
+    {
+      "epoch": 0.5476283503946238,
+      "grad_norm": 0.2044479250907898,
+      "learning_rate": 0.00016409395973154362,
+      "loss": 0.9436,
+      "step": 219
+    },
+    {
+      "epoch": 0.550128936469485,
+      "grad_norm": 0.1923314779996872,
+      "learning_rate": 0.00016392617449664432,
+      "loss": 0.9535,
+      "step": 220
+    },
+    {
+      "epoch": 0.5526295225443464,
+      "grad_norm": 0.1981159895658493,
+      "learning_rate": 0.00016375838926174498,
+      "loss": 0.9031,
+      "step": 221
+    },
+    {
+      "epoch": 0.5551301086192076,
+      "grad_norm": 0.19014570116996765,
+      "learning_rate": 0.00016359060402684566,
+      "loss": 0.9497,
+      "step": 222
+    },
+    {
+      "epoch": 0.5576306946940689,
+      "grad_norm": 0.19516299664974213,
+      "learning_rate": 0.0001634228187919463,
+      "loss": 0.9481,
+      "step": 223
+    },
+    {
+      "epoch": 0.5601312807689303,
+      "grad_norm": 0.2148059755563736,
+      "learning_rate": 0.000163255033557047,
+      "loss": 0.9571,
+      "step": 224
+    },
+    {
+      "epoch": 0.5626318668437915,
+      "grad_norm": 0.20576442778110504,
+      "learning_rate": 0.00016308724832214767,
+      "loss": 1.0018,
+      "step": 225
+    },
+    {
+      "epoch": 0.5651324529186528,
+      "grad_norm": 0.20371298491954803,
+      "learning_rate": 0.00016291946308724832,
+      "loss": 0.8954,
+      "step": 226
+    },
+    {
+      "epoch": 0.5676330389935141,
+      "grad_norm": 0.1865856647491455,
+      "learning_rate": 0.000162751677852349,
+      "loss": 0.9592,
+      "step": 227
+    },
+    {
+      "epoch": 0.5701336250683754,
+      "grad_norm": 0.20932011306285858,
+      "learning_rate": 0.00016258389261744965,
+      "loss": 0.9412,
+      "step": 228
+    },
+    {
+      "epoch": 0.5726342111432366,
+      "grad_norm": 0.2131514698266983,
+      "learning_rate": 0.00016241610738255036,
+      "loss": 0.9413,
+      "step": 229
+    },
+    {
+      "epoch": 0.575134797218098,
+      "grad_norm": 0.19242137670516968,
+      "learning_rate": 0.000162248322147651,
+      "loss": 0.9244,
+      "step": 230
+    },
+    {
+      "epoch": 0.5776353832929593,
+      "grad_norm": 0.20453810691833496,
+      "learning_rate": 0.0001620805369127517,
+      "loss": 0.9972,
+      "step": 231
+    },
+    {
+      "epoch": 0.5801359693678206,
+      "grad_norm": 0.19789357483386993,
+      "learning_rate": 0.00016191275167785237,
+      "loss": 0.9411,
+      "step": 232
+    },
+    {
+      "epoch": 0.5826365554426819,
+      "grad_norm": 0.1875084787607193,
+      "learning_rate": 0.00016174496644295302,
+      "loss": 0.8877,
+      "step": 233
+    },
+    {
+      "epoch": 0.5851371415175431,
+      "grad_norm": 0.21921932697296143,
+      "learning_rate": 0.0001615771812080537,
+      "loss": 0.8957,
+      "step": 234
+    },
+    {
+      "epoch": 0.5876377275924045,
+      "grad_norm": 0.19880889356136322,
+      "learning_rate": 0.00016140939597315436,
+      "loss": 1.0228,
+      "step": 235
+    },
+    {
+      "epoch": 0.5901383136672658,
+      "grad_norm": 0.19592322409152985,
+      "learning_rate": 0.00016124161073825504,
+      "loss": 0.9665,
+      "step": 236
+    },
+    {
+      "epoch": 0.592638899742127,
+      "grad_norm": 0.19552145898342133,
+      "learning_rate": 0.0001610738255033557,
+      "loss": 0.8964,
+      "step": 237
+    },
+    {
+      "epoch": 0.5951394858169884,
+      "grad_norm": 0.20831383764743805,
+      "learning_rate": 0.0001609060402684564,
+      "loss": 0.9962,
+      "step": 238
+    },
+    {
+      "epoch": 0.5976400718918496,
+      "grad_norm": 0.20157143473625183,
+      "learning_rate": 0.00016073825503355705,
+      "loss": 0.9736,
+      "step": 239
+    },
+    {
+      "epoch": 0.6001406579667109,
+      "grad_norm": 0.19122959673404694,
+      "learning_rate": 0.00016057046979865773,
+      "loss": 0.9004,
+      "step": 240
+    },
+    {
+      "epoch": 0.6026412440415723,
+      "grad_norm": 0.19804859161376953,
+      "learning_rate": 0.0001604026845637584,
+      "loss": 0.9438,
+      "step": 241
+    },
+    {
+      "epoch": 0.6051418301164335,
+      "grad_norm": 0.20546962320804596,
+      "learning_rate": 0.00016023489932885906,
+      "loss": 0.9389,
+      "step": 242
+    },
+    {
+      "epoch": 0.6076424161912949,
+      "grad_norm": 0.20515991747379303,
+      "learning_rate": 0.00016006711409395974,
+      "loss": 0.9325,
+      "step": 243
+    },
+    {
+      "epoch": 0.6101430022661561,
+      "grad_norm": 0.20005853474140167,
+      "learning_rate": 0.0001598993288590604,
+      "loss": 0.9827,
+      "step": 244
+    },
+    {
+      "epoch": 0.6126435883410174,
+      "grad_norm": 0.22935104370117188,
+      "learning_rate": 0.00015973154362416107,
+      "loss": 0.9733,
+      "step": 245
+    },
+    {
+      "epoch": 0.6151441744158788,
+      "grad_norm": 0.19898930191993713,
+      "learning_rate": 0.00015956375838926175,
+      "loss": 0.976,
+      "step": 246
+    },
+    {
+      "epoch": 0.61764476049074,
+      "grad_norm": 0.19529736042022705,
+      "learning_rate": 0.00015939597315436243,
+      "loss": 0.9177,
+      "step": 247
+    },
+    {
+      "epoch": 0.6201453465656013,
+      "grad_norm": 0.2278802990913391,
+      "learning_rate": 0.0001592281879194631,
+      "loss": 0.9902,
+      "step": 248
+    },
+    {
+      "epoch": 0.6226459326404626,
+      "grad_norm": 0.21218013763427734,
+      "learning_rate": 0.00015906040268456377,
+      "loss": 0.9617,
+      "step": 249
+    },
+    {
+      "epoch": 0.6251465187153239,
+      "grad_norm": 0.20722109079360962,
+      "learning_rate": 0.00015889261744966445,
+      "loss": 0.9364,
+      "step": 250
+    },
+    {
+      "epoch": 0.6276471047901852,
+      "grad_norm": 0.21361340582370758,
+      "learning_rate": 0.0001587248322147651,
+      "loss": 0.9022,
+      "step": 251
+    },
+    {
+      "epoch": 0.6301476908650465,
+      "grad_norm": 0.2029501050710678,
+      "learning_rate": 0.00015855704697986578,
+      "loss": 0.9071,
+      "step": 252
+    },
+    {
+      "epoch": 0.6326482769399078,
+      "grad_norm": 0.21769894659519196,
+      "learning_rate": 0.00015838926174496643,
+      "loss": 0.9579,
+      "step": 253
+    },
+    {
+      "epoch": 0.6351488630147691,
+      "grad_norm": 0.2267335206270218,
+      "learning_rate": 0.0001582214765100671,
+      "loss": 0.9522,
+      "step": 254
+    },
+    {
+      "epoch": 0.6376494490896304,
+      "grad_norm": 0.20004138350486755,
+      "learning_rate": 0.0001580536912751678,
+      "loss": 0.9873,
+      "step": 255
+    },
+    {
+      "epoch": 0.6401500351644916,
+      "grad_norm": 0.1902337670326233,
+      "learning_rate": 0.00015788590604026847,
+      "loss": 0.9001,
+      "step": 256
+    },
+    {
+      "epoch": 0.642650621239353,
+      "grad_norm": 0.2013217657804489,
+      "learning_rate": 0.00015771812080536915,
+      "loss": 0.9644,
+      "step": 257
+    },
+    {
+      "epoch": 0.6451512073142143,
+      "grad_norm": 0.1940159648656845,
+      "learning_rate": 0.0001575503355704698,
+      "loss": 0.9844,
+      "step": 258
+    },
+    {
+      "epoch": 0.6476517933890755,
+      "grad_norm": 0.20359578728675842,
+      "learning_rate": 0.00015738255033557048,
+      "loss": 0.9652,
+      "step": 259
+    },
+    {
+      "epoch": 0.6501523794639369,
+      "grad_norm": 0.18913763761520386,
+      "learning_rate": 0.00015721476510067114,
+      "loss": 0.9698,
+      "step": 260
+    },
+    {
+      "epoch": 0.6526529655387981,
+      "grad_norm": 0.19329224526882172,
+      "learning_rate": 0.00015704697986577181,
+      "loss": 0.9558,
+      "step": 261
+    },
+    {
+      "epoch": 0.6551535516136594,
+      "grad_norm": 0.20145884156227112,
+      "learning_rate": 0.0001568791946308725,
+      "loss": 1.0232,
+      "step": 262
+    },
+    {
+      "epoch": 0.6576541376885208,
+      "grad_norm": 0.18802523612976074,
+      "learning_rate": 0.00015671140939597315,
+      "loss": 0.9492,
+      "step": 263
+    },
+    {
+      "epoch": 0.660154723763382,
+      "grad_norm": 0.19531258940696716,
+      "learning_rate": 0.00015654362416107385,
+      "loss": 0.9791,
+      "step": 264
+    },
+    {
+      "epoch": 0.6626553098382434,
+      "grad_norm": 0.1917187124490738,
+      "learning_rate": 0.0001563758389261745,
+      "loss": 0.9741,
+      "step": 265
+    },
+    {
+      "epoch": 0.6651558959131046,
+      "grad_norm": 0.19565711915493011,
+      "learning_rate": 0.0001562080536912752,
+      "loss": 0.9876,
+      "step": 266
+    },
+    {
+      "epoch": 0.6676564819879659,
+      "grad_norm": 0.19605979323387146,
+      "learning_rate": 0.00015604026845637584,
+      "loss": 0.8845,
+      "step": 267
+    },
+    {
+      "epoch": 0.6701570680628273,
+      "grad_norm": 0.20349417626857758,
+      "learning_rate": 0.00015587248322147652,
+      "loss": 0.9595,
+      "step": 268
+    },
+    {
+      "epoch": 0.6726576541376885,
+      "grad_norm": 0.19674451649188995,
+      "learning_rate": 0.0001557046979865772,
+      "loss": 0.9113,
+      "step": 269
+    },
+    {
+      "epoch": 0.6751582402125498,
+      "grad_norm": 0.19608086347579956,
+      "learning_rate": 0.00015553691275167785,
+      "loss": 0.8849,
+      "step": 270
+    },
+    {
+      "epoch": 0.6776588262874111,
+      "grad_norm": 0.1995520442724228,
+      "learning_rate": 0.00015536912751677853,
+      "loss": 0.9341,
+      "step": 271
+    },
+    {
+      "epoch": 0.6801594123622724,
+      "grad_norm": 0.20707178115844727,
+      "learning_rate": 0.00015520134228187918,
+      "loss": 0.9262,
+      "step": 272
+    },
+    {
+      "epoch": 0.6826599984371337,
+      "grad_norm": 0.20216511189937592,
+      "learning_rate": 0.0001550335570469799,
+      "loss": 0.9162,
+      "step": 273
+    },
+    {
+      "epoch": 0.685160584511995,
+      "grad_norm": 0.19384877383708954,
+      "learning_rate": 0.00015486577181208054,
+      "loss": 0.9264,
+      "step": 274
+    },
+    {
+      "epoch": 0.6876611705868563,
+      "grad_norm": 0.21822841465473175,
+      "learning_rate": 0.00015469798657718122,
+      "loss": 0.9299,
+      "step": 275
+    },
+    {
+      "epoch": 0.6901617566617176,
+      "grad_norm": 0.195476695895195,
+      "learning_rate": 0.00015453020134228188,
+      "loss": 0.9246,
+      "step": 276
+    },
+    {
+      "epoch": 0.6926623427365789,
+      "grad_norm": 0.2062695026397705,
+      "learning_rate": 0.00015436241610738256,
+      "loss": 0.9143,
+      "step": 277
+    },
+    {
+      "epoch": 0.6951629288114402,
+      "grad_norm": 0.20120318233966827,
+      "learning_rate": 0.00015419463087248324,
+      "loss": 0.9244,
+      "step": 278
+    },
+    {
+      "epoch": 0.6976635148863015,
+      "grad_norm": 0.19195373356342316,
+      "learning_rate": 0.0001540268456375839,
+      "loss": 0.9629,
+      "step": 279
+    },
+    {
+      "epoch": 0.7001641009611628,
+      "grad_norm": 0.20471717417240143,
+      "learning_rate": 0.00015385906040268457,
+      "loss": 0.9761,
+      "step": 280
+    },
+    {
+      "epoch": 0.702664687036024,
+      "grad_norm": 0.192673459649086,
+      "learning_rate": 0.00015369127516778522,
+      "loss": 0.9769,
+      "step": 281
+    },
+    {
+      "epoch": 0.7051652731108854,
+      "grad_norm": 0.18674695491790771,
+      "learning_rate": 0.00015352348993288593,
+      "loss": 0.908,
+      "step": 282
+    },
+    {
+      "epoch": 0.7076658591857466,
+      "grad_norm": 0.2033309042453766,
+      "learning_rate": 0.00015335570469798658,
+      "loss": 0.9619,
+      "step": 283
+    },
+    {
+      "epoch": 0.7101664452606079,
+      "grad_norm": 0.20086367428302765,
+      "learning_rate": 0.00015318791946308726,
+      "loss": 0.9986,
+      "step": 284
+    },
+    {
+      "epoch": 0.7126670313354693,
+      "grad_norm": 0.197890967130661,
+      "learning_rate": 0.00015302013422818794,
+      "loss": 0.9179,
+      "step": 285
+    },
+    {
+      "epoch": 0.7151676174103305,
+      "grad_norm": 0.19057410955429077,
+      "learning_rate": 0.0001528523489932886,
+      "loss": 0.956,
+      "step": 286
+    },
+    {
+      "epoch": 0.7176682034851919,
+      "grad_norm": 0.1913488209247589,
+      "learning_rate": 0.00015268456375838927,
+      "loss": 0.8821,
+      "step": 287
+    },
+    {
+      "epoch": 0.7201687895600531,
+      "grad_norm": 0.19540366530418396,
+      "learning_rate": 0.00015251677852348993,
+      "loss": 0.8933,
+      "step": 288
+    },
+    {
+      "epoch": 0.7226693756349144,
+      "grad_norm": 0.20779870450496674,
+      "learning_rate": 0.0001523489932885906,
+      "loss": 0.9556,
+      "step": 289
+    },
+    {
+      "epoch": 0.7251699617097758,
+      "grad_norm": 0.19707651436328888,
+      "learning_rate": 0.00015218120805369129,
+      "loss": 0.9283,
+      "step": 290
+    },
+    {
+      "epoch": 0.727670547784637,
+      "grad_norm": 0.19629661738872528,
+      "learning_rate": 0.00015201342281879196,
+      "loss": 0.9423,
+      "step": 291
+    },
+    {
+      "epoch": 0.7301711338594983,
+      "grad_norm": 0.19412437081336975,
+      "learning_rate": 0.00015184563758389262,
+      "loss": 0.9046,
+      "step": 292
+    },
+    {
+      "epoch": 0.7326717199343596,
+      "grad_norm": 0.19544045627117157,
+      "learning_rate": 0.0001516778523489933,
+      "loss": 0.8976,
+      "step": 293
+    },
+    {
+      "epoch": 0.7351723060092209,
+      "grad_norm": 0.19209088385105133,
+      "learning_rate": 0.00015151006711409398,
+      "loss": 0.9031,
+      "step": 294
+    },
+    {
+      "epoch": 0.7376728920840822,
+      "grad_norm": 0.20053426921367645,
+      "learning_rate": 0.00015134228187919463,
+      "loss": 0.9324,
+      "step": 295
+    },
+    {
+      "epoch": 0.7401734781589435,
+      "grad_norm": 0.2002066671848297,
+      "learning_rate": 0.0001511744966442953,
+      "loss": 0.995,
+      "step": 296
+    },
+    {
+      "epoch": 0.7426740642338048,
+      "grad_norm": 0.1972004473209381,
+      "learning_rate": 0.00015100671140939596,
+      "loss": 0.9182,
+      "step": 297
+    },
+    {
+      "epoch": 0.7451746503086661,
+      "grad_norm": 0.19501295685768127,
+      "learning_rate": 0.00015083892617449664,
+      "loss": 0.9431,
+      "step": 298
+    },
+    {
+      "epoch": 0.7476752363835274,
+      "grad_norm": 0.20371028780937195,
+      "learning_rate": 0.00015067114093959732,
+      "loss": 0.9059,
+      "step": 299
+    },
+    {
+      "epoch": 0.7501758224583887,
+      "grad_norm": 0.23006804287433624,
+      "learning_rate": 0.000150503355704698,
+      "loss": 0.9876,
+      "step": 300
+    },
+    {
+      "epoch": 0.75267640853325,
+      "grad_norm": 0.18891564011573792,
+      "learning_rate": 0.00015033557046979868,
+      "loss": 0.9465,
+      "step": 301
+    },
+    {
+      "epoch": 0.7551769946081113,
+      "grad_norm": 0.1992032378911972,
+      "learning_rate": 0.00015016778523489933,
+      "loss": 0.939,
+      "step": 302
+    },
+    {
+      "epoch": 0.7576775806829725,
+      "grad_norm": 0.20672324299812317,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.8997,
+      "step": 303
+    },
+    {
+      "epoch": 0.7601781667578339,
+      "grad_norm": 0.20514875650405884,
+      "learning_rate": 0.00014983221476510067,
+      "loss": 0.9027,
+      "step": 304
+    },
+    {
+      "epoch": 0.7626787528326952,
+      "grad_norm": 0.19662059843540192,
+      "learning_rate": 0.00014966442953020135,
+      "loss": 0.9504,
+      "step": 305
+    },
+    {
+      "epoch": 0.7651793389075564,
+      "grad_norm": 0.19113412499427795,
+      "learning_rate": 0.00014949664429530203,
+      "loss": 0.8445,
+      "step": 306
+    },
+    {
+      "epoch": 0.7676799249824178,
+      "grad_norm": 0.20623773336410522,
+      "learning_rate": 0.00014932885906040268,
+      "loss": 0.9669,
+      "step": 307
+    },
+    {
+      "epoch": 0.770180511057279,
+      "grad_norm": 0.20261040329933167,
+      "learning_rate": 0.00014916107382550339,
+      "loss": 0.9047,
+      "step": 308
+    },
+    {
+      "epoch": 0.7726810971321404,
+      "grad_norm": 0.19355683028697968,
+      "learning_rate": 0.00014899328859060404,
+      "loss": 0.9422,
+      "step": 309
+    },
+    {
+      "epoch": 0.7751816832070016,
+      "grad_norm": 0.21379676461219788,
+      "learning_rate": 0.00014882550335570472,
+      "loss": 0.9782,
+      "step": 310
+    },
+    {
+      "epoch": 0.7776822692818629,
+      "grad_norm": 0.20317813754081726,
+      "learning_rate": 0.00014865771812080537,
+      "loss": 0.931,
+      "step": 311
+    },
+    {
+      "epoch": 0.7801828553567243,
+      "grad_norm": 0.20362474024295807,
+      "learning_rate": 0.00014848993288590605,
+      "loss": 0.8662,
+      "step": 312
+    },
+    {
+      "epoch": 0.7826834414315855,
+      "grad_norm": 0.20510531961917877,
+      "learning_rate": 0.0001483221476510067,
+      "loss": 1.0119,
+      "step": 313
+    },
+    {
+      "epoch": 0.7851840275064468,
+      "grad_norm": 0.19395427405834198,
+      "learning_rate": 0.00014815436241610738,
+      "loss": 0.9228,
+      "step": 314
+    },
+    {
+      "epoch": 0.7876846135813081,
+      "grad_norm": 0.1922769993543625,
+      "learning_rate": 0.00014798657718120806,
+      "loss": 0.9197,
+      "step": 315
+    },
+    {
+      "epoch": 0.7901851996561694,
+      "grad_norm": 0.2138831913471222,
+      "learning_rate": 0.00014781879194630872,
+      "loss": 0.9551,
+      "step": 316
+    },
+    {
+      "epoch": 0.7926857857310307,
+      "grad_norm": 0.1988006830215454,
+      "learning_rate": 0.00014765100671140942,
+      "loss": 0.9624,
+      "step": 317
+    },
+    {
+      "epoch": 0.795186371805892,
+      "grad_norm": 0.19090843200683594,
+      "learning_rate": 0.00014748322147651008,
+      "loss": 0.9382,
+      "step": 318
+    },
+    {
+      "epoch": 0.7976869578807533,
+      "grad_norm": 0.19148418307304382,
+      "learning_rate": 0.00014731543624161076,
+      "loss": 0.9222,
+      "step": 319
+    },
+    {
+      "epoch": 0.8001875439556146,
+      "grad_norm": 0.19201642274856567,
+      "learning_rate": 0.0001471476510067114,
+      "loss": 0.9356,
+      "step": 320
+    },
+    {
+      "epoch": 0.8026881300304759,
+      "grad_norm": 0.18535450100898743,
+      "learning_rate": 0.0001469798657718121,
+      "loss": 0.9196,
+      "step": 321
+    },
+    {
+      "epoch": 0.8051887161053372,
+      "grad_norm": 0.19076795876026154,
+      "learning_rate": 0.00014681208053691277,
+      "loss": 0.9523,
+      "step": 322
+    },
+    {
+      "epoch": 0.8076893021801985,
+      "grad_norm": 0.2030184417963028,
+      "learning_rate": 0.00014664429530201342,
+      "loss": 0.9695,
+      "step": 323
+    },
+    {
+      "epoch": 0.8101898882550598,
+      "grad_norm": 0.18903104960918427,
+      "learning_rate": 0.0001464765100671141,
+      "loss": 0.9315,
+      "step": 324
+    },
+    {
+      "epoch": 0.812690474329921,
+      "grad_norm": 0.18826857209205627,
+      "learning_rate": 0.00014630872483221478,
+      "loss": 0.9309,
+      "step": 325
+    },
+    {
+      "epoch": 0.8151910604047824,
+      "grad_norm": 0.1963105946779251,
+      "learning_rate": 0.00014614093959731546,
+      "loss": 0.8993,
+      "step": 326
+    },
+    {
+      "epoch": 0.8176916464796437,
+      "grad_norm": 0.20250087976455688,
+      "learning_rate": 0.0001459731543624161,
+      "loss": 0.9275,
+      "step": 327
+    },
+    {
+      "epoch": 0.8201922325545049,
+      "grad_norm": 0.20003145933151245,
+      "learning_rate": 0.0001458053691275168,
+      "loss": 0.9373,
+      "step": 328
+    },
+    {
+      "epoch": 0.8226928186293663,
+      "grad_norm": 0.18792639672756195,
+      "learning_rate": 0.00014563758389261744,
+      "loss": 0.861,
+      "step": 329
+    },
+    {
+      "epoch": 0.8251934047042275,
+      "grad_norm": 0.19464413821697235,
+      "learning_rate": 0.00014546979865771812,
+      "loss": 0.9148,
+      "step": 330
+    },
+    {
+      "epoch": 0.8276939907790889,
+      "grad_norm": 0.19977621734142303,
+      "learning_rate": 0.0001453020134228188,
+      "loss": 0.9171,
+      "step": 331
+    },
+    {
+      "epoch": 0.8301945768539502,
+      "grad_norm": 0.20777297019958496,
+      "learning_rate": 0.00014513422818791946,
+      "loss": 0.9772,
+      "step": 332
+    },
+    {
+      "epoch": 0.8326951629288114,
+      "grad_norm": 0.2004861682653427,
+      "learning_rate": 0.00014496644295302014,
+      "loss": 0.9501,
+      "step": 333
+    },
+    {
+      "epoch": 0.8351957490036728,
+      "grad_norm": 0.19017265737056732,
+      "learning_rate": 0.00014479865771812082,
+      "loss": 0.8614,
+      "step": 334
+    },
+    {
+      "epoch": 0.837696335078534,
+      "grad_norm": 0.20576266944408417,
+      "learning_rate": 0.0001446308724832215,
+      "loss": 0.9464,
+      "step": 335
+    },
+    {
+      "epoch": 0.8401969211533953,
+      "grad_norm": 0.2054765373468399,
+      "learning_rate": 0.00014446308724832215,
+      "loss": 0.9268,
+      "step": 336
+    },
+    {
+      "epoch": 0.8426975072282566,
+      "grad_norm": 0.19856566190719604,
+      "learning_rate": 0.00014429530201342283,
+      "loss": 0.9044,
+      "step": 337
+    },
+    {
+      "epoch": 0.8451980933031179,
+      "grad_norm": 0.20033764839172363,
+      "learning_rate": 0.0001441275167785235,
+      "loss": 0.9379,
+      "step": 338
+    },
+    {
+      "epoch": 0.8476986793779792,
+      "grad_norm": 0.19791364669799805,
+      "learning_rate": 0.00014395973154362416,
+      "loss": 0.8634,
+      "step": 339
+    },
+    {
+      "epoch": 0.8501992654528405,
+      "grad_norm": 0.19763191044330597,
+      "learning_rate": 0.00014379194630872484,
+      "loss": 0.9071,
+      "step": 340
+    },
+    {
+      "epoch": 0.8526998515277018,
+      "grad_norm": 0.2084035873413086,
+      "learning_rate": 0.0001436241610738255,
+      "loss": 0.9449,
+      "step": 341
+    },
+    {
+      "epoch": 0.8552004376025631,
+      "grad_norm": 0.20290718972682953,
+      "learning_rate": 0.00014345637583892617,
+      "loss": 0.9239,
+      "step": 342
+    },
+    {
+      "epoch": 0.8577010236774244,
+      "grad_norm": 0.20317819714546204,
+      "learning_rate": 0.00014328859060402685,
+      "loss": 0.9065,
+      "step": 343
+    },
+    {
+      "epoch": 0.8602016097522857,
+      "grad_norm": 0.19348326325416565,
+      "learning_rate": 0.00014312080536912753,
+      "loss": 0.8981,
+      "step": 344
+    },
+    {
+      "epoch": 0.862702195827147,
+      "grad_norm": 0.20921850204467773,
+      "learning_rate": 0.00014295302013422819,
+      "loss": 0.8473,
+      "step": 345
+    },
+    {
+      "epoch": 0.8652027819020083,
+      "grad_norm": 0.1959138959646225,
+      "learning_rate": 0.00014278523489932887,
+      "loss": 0.9086,
+      "step": 346
+    },
+    {
+      "epoch": 0.8677033679768695,
+      "grad_norm": 0.18429264426231384,
+      "learning_rate": 0.00014261744966442955,
+      "loss": 0.9492,
+      "step": 347
+    },
+    {
+      "epoch": 0.8702039540517309,
+      "grad_norm": 0.19408418238162994,
+      "learning_rate": 0.0001424496644295302,
+      "loss": 0.9161,
+      "step": 348
+    },
+    {
+      "epoch": 0.8727045401265922,
+      "grad_norm": 0.19348210096359253,
+      "learning_rate": 0.00014228187919463088,
+      "loss": 0.8937,
+      "step": 349
+    },
+    {
+      "epoch": 0.8752051262014535,
+      "grad_norm": 0.20142613351345062,
+      "learning_rate": 0.00014211409395973153,
+      "loss": 0.928,
+      "step": 350
+    },
+    {
+      "epoch": 0.8777057122763148,
+      "grad_norm": 0.19699464738368988,
+      "learning_rate": 0.0001419463087248322,
+      "loss": 0.9856,
+      "step": 351
+    },
+    {
+      "epoch": 0.880206298351176,
+      "grad_norm": 0.19020123779773712,
+      "learning_rate": 0.0001417785234899329,
+      "loss": 0.9221,
+      "step": 352
+    },
+    {
+      "epoch": 0.8827068844260374,
+      "grad_norm": 0.18610498309135437,
+      "learning_rate": 0.00014161073825503357,
+      "loss": 0.8846,
+      "step": 353
+    },
+    {
+      "epoch": 0.8852074705008987,
+      "grad_norm": 0.20566537976264954,
+      "learning_rate": 0.00014144295302013425,
+      "loss": 0.9766,
+      "step": 354
+    },
+    {
+      "epoch": 0.8877080565757599,
+      "grad_norm": 0.18680301308631897,
+      "learning_rate": 0.0001412751677852349,
+      "loss": 0.8741,
+      "step": 355
+    },
+    {
+      "epoch": 0.8902086426506213,
+      "grad_norm": 0.19031471014022827,
+      "learning_rate": 0.00014110738255033558,
+      "loss": 0.9063,
+      "step": 356
+    },
+    {
+      "epoch": 0.8927092287254825,
+      "grad_norm": 0.18606281280517578,
+      "learning_rate": 0.00014093959731543624,
+      "loss": 0.9157,
+      "step": 357
+    },
+    {
+      "epoch": 0.8952098148003438,
+      "grad_norm": 0.19429543614387512,
+      "learning_rate": 0.00014077181208053691,
+      "loss": 0.9557,
+      "step": 358
+    },
+    {
+      "epoch": 0.8977104008752052,
+      "grad_norm": 0.205494686961174,
+      "learning_rate": 0.0001406040268456376,
+      "loss": 0.9504,
+      "step": 359
+    },
+    {
+      "epoch": 0.9002109869500664,
+      "grad_norm": 0.21151868999004364,
+      "learning_rate": 0.00014043624161073827,
+      "loss": 0.9404,
+      "step": 360
+    },
+    {
+      "epoch": 0.9027115730249278,
+      "grad_norm": 0.19715413451194763,
+      "learning_rate": 0.00014026845637583895,
+      "loss": 0.9368,
+      "step": 361
+    },
+    {
+      "epoch": 0.905212159099789,
+      "grad_norm": 0.20190200209617615,
+      "learning_rate": 0.0001401006711409396,
+      "loss": 0.8945,
+      "step": 362
+    },
+    {
+      "epoch": 0.9077127451746503,
+      "grad_norm": 0.19868037104606628,
+      "learning_rate": 0.0001399328859060403,
+      "loss": 0.914,
+      "step": 363
+    },
+    {
+      "epoch": 0.9102133312495116,
+      "grad_norm": 0.1939004808664322,
+      "learning_rate": 0.00013976510067114094,
+      "loss": 0.9319,
+      "step": 364
+    },
+    {
+      "epoch": 0.9127139173243729,
+      "grad_norm": 0.21012766659259796,
+      "learning_rate": 0.00013959731543624162,
+      "loss": 0.9357,
+      "step": 365
+    },
+    {
+      "epoch": 0.9152145033992342,
+      "grad_norm": 0.19320148229599,
+      "learning_rate": 0.00013942953020134227,
+      "loss": 0.9171,
+      "step": 366
+    },
+    {
+      "epoch": 0.9177150894740955,
+      "grad_norm": 0.19462816417217255,
+      "learning_rate": 0.00013926174496644295,
+      "loss": 0.8754,
+      "step": 367
+    },
+    {
+      "epoch": 0.9202156755489568,
+      "grad_norm": 0.19171462953090668,
+      "learning_rate": 0.00013909395973154363,
+      "loss": 0.8649,
+      "step": 368
+    },
+    {
+      "epoch": 0.922716261623818,
+      "grad_norm": 0.19302335381507874,
+      "learning_rate": 0.0001389261744966443,
+      "loss": 0.8818,
+      "step": 369
+    },
+    {
+      "epoch": 0.9252168476986794,
+      "grad_norm": 0.19845272600650787,
+      "learning_rate": 0.000138758389261745,
+      "loss": 0.9664,
+      "step": 370
+    },
+    {
+      "epoch": 0.9277174337735407,
+      "grad_norm": 0.19285465776920319,
+      "learning_rate": 0.00013859060402684564,
+      "loss": 0.938,
+      "step": 371
+    },
+    {
+      "epoch": 0.930218019848402,
+      "grad_norm": 0.19765235483646393,
+      "learning_rate": 0.00013842281879194632,
+      "loss": 0.9196,
+      "step": 372
+    },
+    {
+      "epoch": 0.9327186059232633,
+      "grad_norm": 0.19708356261253357,
+      "learning_rate": 0.00013825503355704698,
+      "loss": 0.9941,
+      "step": 373
+    },
+    {
+      "epoch": 0.9352191919981245,
+      "grad_norm": 0.20135480165481567,
+      "learning_rate": 0.00013808724832214766,
+      "loss": 0.9202,
+      "step": 374
+    },
+    {
+      "epoch": 0.9377197780729859,
+      "grad_norm": 0.20005004107952118,
+      "learning_rate": 0.00013791946308724834,
+      "loss": 0.9331,
+      "step": 375
+    },
+    {
+      "epoch": 0.9402203641478472,
+      "grad_norm": 0.19034837186336517,
+      "learning_rate": 0.000137751677852349,
+      "loss": 0.8916,
+      "step": 376
+    },
+    {
+      "epoch": 0.9427209502227084,
+      "grad_norm": 0.19903801381587982,
+      "learning_rate": 0.00013758389261744967,
+      "loss": 0.8614,
+      "step": 377
+    },
+    {
+      "epoch": 0.9452215362975698,
+      "grad_norm": 0.19618506729602814,
+      "learning_rate": 0.00013741610738255035,
+      "loss": 0.9527,
+      "step": 378
+    },
+    {
+      "epoch": 0.947722122372431,
+      "grad_norm": 0.1987503319978714,
+      "learning_rate": 0.00013724832214765103,
+      "loss": 0.8987,
+      "step": 379
+    },
+    {
+      "epoch": 0.9502227084472923,
+      "grad_norm": 0.1936248540878296,
+      "learning_rate": 0.00013708053691275168,
+      "loss": 0.8943,
+      "step": 380
+    },
+    {
+      "epoch": 0.9527232945221537,
+      "grad_norm": 0.19770075380802155,
+      "learning_rate": 0.00013691275167785236,
+      "loss": 0.918,
+      "step": 381
+    },
+    {
+      "epoch": 0.9552238805970149,
+      "grad_norm": 0.19526584446430206,
+      "learning_rate": 0.000136744966442953,
+      "loss": 0.9709,
+      "step": 382
+    },
+    {
+      "epoch": 0.9577244666718763,
+      "grad_norm": 0.20261335372924805,
+      "learning_rate": 0.0001365771812080537,
+      "loss": 0.9189,
+      "step": 383
+    },
+    {
+      "epoch": 0.9602250527467375,
+      "grad_norm": 0.2129514068365097,
+      "learning_rate": 0.00013640939597315437,
+      "loss": 0.9054,
+      "step": 384
+    },
+    {
+      "epoch": 0.9627256388215988,
+      "grad_norm": 0.20009979605674744,
+      "learning_rate": 0.00013624161073825503,
+      "loss": 0.8555,
+      "step": 385
+    },
+    {
+      "epoch": 0.9652262248964601,
+      "grad_norm": 0.22270314395427704,
+      "learning_rate": 0.0001360738255033557,
+      "loss": 0.9823,
+      "step": 386
+    },
+    {
+      "epoch": 0.9677268109713214,
+      "grad_norm": 0.20095857977867126,
+      "learning_rate": 0.00013590604026845639,
+      "loss": 0.8974,
+      "step": 387
+    },
+    {
+      "epoch": 0.9702273970461827,
+      "grad_norm": 0.21703922748565674,
+      "learning_rate": 0.00013573825503355706,
+      "loss": 0.9612,
+      "step": 388
+    },
+    {
+      "epoch": 0.972727983121044,
+      "grad_norm": 0.20171691477298737,
+      "learning_rate": 0.00013557046979865772,
+      "loss": 0.9066,
+      "step": 389
+    },
+    {
+      "epoch": 0.9752285691959053,
+      "grad_norm": 0.21097329258918762,
+      "learning_rate": 0.0001354026845637584,
+      "loss": 0.9024,
+      "step": 390
+    },
+    {
+      "epoch": 0.9777291552707665,
+      "grad_norm": 0.1949266642332077,
+      "learning_rate": 0.00013523489932885908,
+      "loss": 0.8925,
+      "step": 391
+    },
+    {
+      "epoch": 0.9802297413456279,
+      "grad_norm": 0.20178557932376862,
+      "learning_rate": 0.00013506711409395973,
+      "loss": 0.9671,
+      "step": 392
+    },
+    {
+      "epoch": 0.9827303274204892,
+      "grad_norm": 0.1976921707391739,
+      "learning_rate": 0.0001348993288590604,
+      "loss": 0.8416,
+      "step": 393
+    },
+    {
+      "epoch": 0.9852309134953505,
+      "grad_norm": 0.20067280530929565,
+      "learning_rate": 0.00013473154362416106,
+      "loss": 0.9176,
+      "step": 394
+    },
+    {
+      "epoch": 0.9877314995702118,
+      "grad_norm": 0.20026645064353943,
+      "learning_rate": 0.00013456375838926174,
+      "loss": 0.9605,
+      "step": 395
+    },
+    {
+      "epoch": 0.990232085645073,
+      "grad_norm": 0.2071172446012497,
+      "learning_rate": 0.00013439597315436242,
+      "loss": 0.8994,
+      "step": 396
+    },
+    {
+      "epoch": 0.9927326717199344,
+      "grad_norm": 0.20478585362434387,
+      "learning_rate": 0.0001342281879194631,
+      "loss": 0.8575,
+      "step": 397
+    },
+    {
+      "epoch": 0.9952332577947957,
+      "grad_norm": 0.21187831461429596,
+      "learning_rate": 0.00013406040268456378,
+      "loss": 0.9329,
+      "step": 398
+    },
+    {
+      "epoch": 0.9977338438696569,
+      "grad_norm": 0.20360170304775238,
+      "learning_rate": 0.00013389261744966443,
+      "loss": 0.912,
+      "step": 399
+    },
+    {
+      "epoch": 1.0002344299445183,
+      "grad_norm": 0.20638534426689148,
+      "learning_rate": 0.00013372483221476511,
+      "loss": 0.9548,
+      "step": 400
+    },
+    {
+      "epoch": 1.0027350160193795,
+      "grad_norm": 0.20016121864318848,
+      "learning_rate": 0.00013355704697986577,
+      "loss": 0.8672,
+      "step": 401
+    },
+    {
+      "epoch": 1.0052356020942408,
+      "grad_norm": 0.19950714707374573,
+      "learning_rate": 0.00013338926174496645,
+      "loss": 0.9067,
+      "step": 402
+    },
+    {
+      "epoch": 1.007736188169102,
+      "grad_norm": 0.19529801607131958,
+      "learning_rate": 0.0001332214765100671,
+      "loss": 0.9288,
+      "step": 403
+    },
+    {
+      "epoch": 1.0102367742439635,
+      "grad_norm": 0.19638019800186157,
+      "learning_rate": 0.0001330536912751678,
+      "loss": 0.8655,
+      "step": 404
+    },
+    {
+      "epoch": 1.0127373603188248,
+      "grad_norm": 0.2076416313648224,
+      "learning_rate": 0.00013288590604026846,
+      "loss": 0.8996,
+      "step": 405
+    },
+    {
+      "epoch": 1.015237946393686,
+      "grad_norm": 0.19819454848766327,
+      "learning_rate": 0.00013271812080536914,
+      "loss": 0.8955,
+      "step": 406
+    },
+    {
+      "epoch": 1.0177385324685473,
+      "grad_norm": 0.2053442746400833,
+      "learning_rate": 0.00013255033557046982,
+      "loss": 0.8902,
+      "step": 407
+    },
+    {
+      "epoch": 1.0202391185434085,
+      "grad_norm": 0.211964949965477,
+      "learning_rate": 0.00013238255033557047,
+      "loss": 0.8836,
+      "step": 408
+    },
+    {
+      "epoch": 1.02273970461827,
+      "grad_norm": 0.21029819548130035,
+      "learning_rate": 0.00013221476510067115,
+      "loss": 0.8894,
+      "step": 409
+    },
+    {
+      "epoch": 1.0252402906931313,
+      "grad_norm": 0.20015518367290497,
+      "learning_rate": 0.0001320469798657718,
+      "loss": 0.9041,
+      "step": 410
+    },
+    {
+      "epoch": 1.0277408767679925,
+      "grad_norm": 0.2067364752292633,
+      "learning_rate": 0.00013187919463087248,
+      "loss": 0.8769,
+      "step": 411
+    },
+    {
+      "epoch": 1.0302414628428538,
+      "grad_norm": 0.21249079704284668,
+      "learning_rate": 0.00013171140939597316,
+      "loss": 0.8352,
+      "step": 412
+    },
+    {
+      "epoch": 1.032742048917715,
+      "grad_norm": 0.20129866898059845,
+      "learning_rate": 0.00013154362416107384,
+      "loss": 0.8499,
+      "step": 413
+    },
+    {
+      "epoch": 1.0352426349925763,
+      "grad_norm": 0.19970041513442993,
+      "learning_rate": 0.00013137583892617452,
+      "loss": 0.8604,
+      "step": 414
+    },
+    {
+      "epoch": 1.0377432210674378,
+      "grad_norm": 0.20939721167087555,
+      "learning_rate": 0.00013120805369127518,
+      "loss": 0.875,
+      "step": 415
+    },
+    {
+      "epoch": 1.040243807142299,
+      "grad_norm": 0.19969108700752258,
+      "learning_rate": 0.00013104026845637586,
+      "loss": 0.8776,
+      "step": 416
+    },
+    {
+      "epoch": 1.0427443932171603,
+      "grad_norm": 0.21169830858707428,
+      "learning_rate": 0.0001308724832214765,
+      "loss": 0.8733,
+      "step": 417
+    },
+    {
+      "epoch": 1.0452449792920215,
+      "grad_norm": 0.20807021856307983,
+      "learning_rate": 0.0001307046979865772,
+      "loss": 0.9011,
+      "step": 418
+    },
+    {
+      "epoch": 1.0477455653668828,
+      "grad_norm": 0.20478905737400055,
+      "learning_rate": 0.00013053691275167784,
+      "loss": 0.8469,
+      "step": 419
+    },
+    {
+      "epoch": 1.050246151441744,
+      "grad_norm": 0.20112550258636475,
+      "learning_rate": 0.00013036912751677852,
+      "loss": 0.8337,
+      "step": 420
+    },
+    {
+      "epoch": 1.0527467375166055,
+      "grad_norm": 0.20829133689403534,
+      "learning_rate": 0.0001302013422818792,
+      "loss": 0.8658,
+      "step": 421
+    },
+    {
+      "epoch": 1.0552473235914668,
+      "grad_norm": 0.2128250151872635,
+      "learning_rate": 0.00013003355704697988,
+      "loss": 0.9182,
+      "step": 422
+    },
+    {
+      "epoch": 1.057747909666328,
+      "grad_norm": 0.21892035007476807,
+      "learning_rate": 0.00012986577181208056,
+      "loss": 0.8941,
+      "step": 423
+    },
+    {
+      "epoch": 1.0602484957411893,
+      "grad_norm": 0.21170388162136078,
+      "learning_rate": 0.0001296979865771812,
+      "loss": 0.8941,
+      "step": 424
+    },
+    {
+      "epoch": 1.0627490818160505,
+      "grad_norm": 0.22825805842876434,
+      "learning_rate": 0.0001295302013422819,
+      "loss": 0.8802,
+      "step": 425
+    },
+    {
+      "epoch": 1.065249667890912,
+      "grad_norm": 0.22174450755119324,
+      "learning_rate": 0.00012936241610738254,
+      "loss": 0.8551,
+      "step": 426
+    },
+    {
+      "epoch": 1.0677502539657733,
+      "grad_norm": 0.21162670850753784,
+      "learning_rate": 0.00012919463087248322,
+      "loss": 0.8937,
+      "step": 427
+    },
+    {
+      "epoch": 1.0702508400406345,
+      "grad_norm": 0.21269601583480835,
+      "learning_rate": 0.0001290268456375839,
+      "loss": 0.8639,
+      "step": 428
+    },
+    {
+      "epoch": 1.0727514261154958,
+      "grad_norm": 0.21762597560882568,
+      "learning_rate": 0.00012885906040268456,
+      "loss": 0.844,
+      "step": 429
+    },
+    {
+      "epoch": 1.075252012190357,
+      "grad_norm": 0.22435954213142395,
+      "learning_rate": 0.00012869127516778524,
+      "loss": 0.9042,
+      "step": 430
+    },
+    {
+      "epoch": 1.0777525982652185,
+      "grad_norm": 0.21457557380199432,
+      "learning_rate": 0.00012852348993288592,
+      "loss": 0.8402,
+      "step": 431
+    },
+    {
+      "epoch": 1.0802531843400798,
+      "grad_norm": 0.21020552515983582,
+      "learning_rate": 0.0001283557046979866,
+      "loss": 0.9129,
+      "step": 432
+    },
+    {
+      "epoch": 1.082753770414941,
+      "grad_norm": 0.21104837954044342,
+      "learning_rate": 0.00012818791946308725,
+      "loss": 0.9079,
+      "step": 433
+    },
+    {
+      "epoch": 1.0852543564898023,
+      "grad_norm": 0.21720872819423676,
+      "learning_rate": 0.00012802013422818793,
+      "loss": 0.858,
+      "step": 434
+    },
+    {
+      "epoch": 1.0877549425646635,
+      "grad_norm": 0.21718966960906982,
+      "learning_rate": 0.0001278523489932886,
+      "loss": 0.9565,
+      "step": 435
+    },
+    {
+      "epoch": 1.0902555286395248,
+      "grad_norm": 0.21655264496803284,
+      "learning_rate": 0.00012768456375838926,
+      "loss": 0.8807,
+      "step": 436
+    },
+    {
+      "epoch": 1.0927561147143863,
+      "grad_norm": 0.22126564383506775,
+      "learning_rate": 0.00012751677852348994,
+      "loss": 0.8452,
+      "step": 437
+    },
+    {
+      "epoch": 1.0952567007892475,
+      "grad_norm": 0.20520944893360138,
+      "learning_rate": 0.0001273489932885906,
+      "loss": 0.8467,
+      "step": 438
+    },
+    {
+      "epoch": 1.0977572868641088,
+      "grad_norm": 0.21571533381938934,
+      "learning_rate": 0.0001271812080536913,
+      "loss": 0.8682,
+      "step": 439
+    },
+    {
+      "epoch": 1.10025787293897,
+      "grad_norm": 0.22515913844108582,
+      "learning_rate": 0.00012701342281879195,
+      "loss": 0.862,
+      "step": 440
+    },
+    {
+      "epoch": 1.1027584590138313,
+      "grad_norm": 0.2109398990869522,
+      "learning_rate": 0.00012684563758389263,
+      "loss": 0.8754,
+      "step": 441
+    },
+    {
+      "epoch": 1.1052590450886925,
+      "grad_norm": 0.21629829704761505,
+      "learning_rate": 0.00012667785234899329,
+      "loss": 0.8876,
+      "step": 442
+    },
+    {
+      "epoch": 1.107759631163554,
+      "grad_norm": 0.22433191537857056,
+      "learning_rate": 0.00012651006711409397,
+      "loss": 0.9026,
+      "step": 443
+    },
+    {
+      "epoch": 1.1102602172384153,
+      "grad_norm": 0.2074134796857834,
+      "learning_rate": 0.00012634228187919465,
+      "loss": 0.8615,
+      "step": 444
+    },
+    {
+      "epoch": 1.1127608033132765,
+      "grad_norm": 0.20749977231025696,
+      "learning_rate": 0.0001261744966442953,
+      "loss": 0.8419,
+      "step": 445
+    },
+    {
+      "epoch": 1.1152613893881378,
+      "grad_norm": 0.22443081438541412,
+      "learning_rate": 0.00012600671140939598,
+      "loss": 0.8668,
+      "step": 446
+    },
+    {
+      "epoch": 1.117761975462999,
+      "grad_norm": 0.21158166229724884,
+      "learning_rate": 0.00012583892617449663,
+      "loss": 0.8208,
+      "step": 447
+    },
+    {
+      "epoch": 1.1202625615378605,
+      "grad_norm": 0.2097133845090866,
+      "learning_rate": 0.00012567114093959734,
+      "loss": 0.9029,
+      "step": 448
+    },
+    {
+      "epoch": 1.1227631476127218,
+      "grad_norm": 0.22370265424251556,
+      "learning_rate": 0.000125503355704698,
+      "loss": 0.8474,
+      "step": 449
+    },
+    {
+      "epoch": 1.125263733687583,
+      "grad_norm": 0.214742511510849,
+      "learning_rate": 0.00012533557046979867,
+      "loss": 0.8601,
+      "step": 450
+    },
+    {
+      "epoch": 1.1277643197624443,
+      "grad_norm": 0.2197917401790619,
+      "learning_rate": 0.00012516778523489935,
+      "loss": 0.809,
+      "step": 451
+    },
+    {
+      "epoch": 1.1302649058373055,
+      "grad_norm": 0.22799834609031677,
+      "learning_rate": 0.000125,
+      "loss": 0.8841,
+      "step": 452
+    },
+    {
+      "epoch": 1.132765491912167,
+      "grad_norm": 0.22149550914764404,
+      "learning_rate": 0.00012483221476510068,
+      "loss": 0.9131,
+      "step": 453
+    },
+    {
+      "epoch": 1.1352660779870283,
+      "grad_norm": 0.21215297281742096,
+      "learning_rate": 0.00012466442953020134,
+      "loss": 0.8944,
+      "step": 454
+    },
+    {
+      "epoch": 1.1377666640618895,
+      "grad_norm": 0.22512802481651306,
+      "learning_rate": 0.00012449664429530201,
+      "loss": 0.8935,
+      "step": 455
+    },
+    {
+      "epoch": 1.1402672501367508,
+      "grad_norm": 0.21118547022342682,
+      "learning_rate": 0.00012432885906040267,
+      "loss": 0.9184,
+      "step": 456
+    },
+    {
+      "epoch": 1.142767836211612,
+      "grad_norm": 0.21238967776298523,
+      "learning_rate": 0.00012416107382550337,
+      "loss": 0.8627,
+      "step": 457
+    },
+    {
+      "epoch": 1.1452684222864733,
+      "grad_norm": 0.22024264931678772,
+      "learning_rate": 0.00012399328859060403,
+      "loss": 0.8885,
+      "step": 458
+    },
+    {
+      "epoch": 1.1477690083613348,
+      "grad_norm": 0.21847938001155853,
+      "learning_rate": 0.0001238255033557047,
+      "loss": 0.8765,
+      "step": 459
+    },
+    {
+      "epoch": 1.150269594436196,
+      "grad_norm": 0.21836036443710327,
+      "learning_rate": 0.0001236577181208054,
+      "loss": 0.8597,
+      "step": 460
+    },
+    {
+      "epoch": 1.1527701805110573,
+      "grad_norm": 0.2130286544561386,
+      "learning_rate": 0.00012348993288590604,
+      "loss": 0.8768,
+      "step": 461
+    },
+    {
+      "epoch": 1.1552707665859185,
+      "grad_norm": 0.21369999647140503,
+      "learning_rate": 0.00012332214765100672,
+      "loss": 0.8017,
+      "step": 462
+    },
+    {
+      "epoch": 1.1577713526607798,
+      "grad_norm": 0.2259240299463272,
+      "learning_rate": 0.00012315436241610737,
+      "loss": 0.8715,
+      "step": 463
+    },
+    {
+      "epoch": 1.160271938735641,
+      "grad_norm": 0.21691924333572388,
+      "learning_rate": 0.00012298657718120805,
+      "loss": 0.9279,
+      "step": 464
+    },
+    {
+      "epoch": 1.1627725248105025,
+      "grad_norm": 0.20966695249080658,
+      "learning_rate": 0.00012281879194630873,
+      "loss": 0.8204,
+      "step": 465
+    },
+    {
+      "epoch": 1.1652731108853638,
+      "grad_norm": 0.2113383263349533,
+      "learning_rate": 0.0001226510067114094,
+      "loss": 0.8876,
+      "step": 466
+    },
+    {
+      "epoch": 1.167773696960225,
+      "grad_norm": 0.2332780808210373,
+      "learning_rate": 0.0001224832214765101,
+      "loss": 0.8446,
+      "step": 467
+    },
+    {
+      "epoch": 1.1702742830350863,
+      "grad_norm": 0.21881066262722015,
+      "learning_rate": 0.00012231543624161074,
+      "loss": 0.8721,
+      "step": 468
+    },
+    {
+      "epoch": 1.1727748691099475,
+      "grad_norm": 0.21677622199058533,
+      "learning_rate": 0.00012214765100671142,
+      "loss": 0.8083,
+      "step": 469
+    },
+    {
+      "epoch": 1.175275455184809,
+      "grad_norm": 0.22742587327957153,
+      "learning_rate": 0.00012197986577181209,
+      "loss": 0.8901,
+      "step": 470
+    },
+    {
+      "epoch": 1.1777760412596703,
+      "grad_norm": 0.21044526994228363,
+      "learning_rate": 0.00012181208053691276,
+      "loss": 0.8411,
+      "step": 471
+    },
+    {
+      "epoch": 1.1802766273345315,
+      "grad_norm": 0.2221437394618988,
+      "learning_rate": 0.00012164429530201342,
+      "loss": 0.8908,
+      "step": 472
+    },
+    {
+      "epoch": 1.1827772134093928,
+      "grad_norm": 0.22299771010875702,
+      "learning_rate": 0.00012147651006711409,
+      "loss": 0.8717,
+      "step": 473
+    },
+    {
+      "epoch": 1.185277799484254,
+      "grad_norm": 0.2145218700170517,
+      "learning_rate": 0.00012130872483221478,
+      "loss": 0.8593,
+      "step": 474
+    },
+    {
+      "epoch": 1.1877783855591155,
+      "grad_norm": 0.2120436728000641,
+      "learning_rate": 0.00012114093959731545,
+      "loss": 0.8369,
+      "step": 475
+    },
+    {
+      "epoch": 1.1902789716339768,
+      "grad_norm": 0.24196121096611023,
+      "learning_rate": 0.00012097315436241611,
+      "loss": 0.8599,
+      "step": 476
+    },
+    {
+      "epoch": 1.192779557708838,
+      "grad_norm": 0.2327934354543686,
+      "learning_rate": 0.0001208053691275168,
+      "loss": 0.8856,
+      "step": 477
+    },
+    {
+      "epoch": 1.1952801437836993,
+      "grad_norm": 0.2271525263786316,
+      "learning_rate": 0.00012063758389261746,
+      "loss": 0.8695,
+      "step": 478
+    },
+    {
+      "epoch": 1.1977807298585605,
+      "grad_norm": 0.22260010242462158,
+      "learning_rate": 0.00012046979865771813,
+      "loss": 0.8731,
+      "step": 479
+    },
+    {
+      "epoch": 1.2002813159334218,
+      "grad_norm": 0.21468815207481384,
+      "learning_rate": 0.00012030201342281879,
+      "loss": 0.9019,
+      "step": 480
+    },
+    {
+      "epoch": 1.2027819020082833,
+      "grad_norm": 0.20896653831005096,
+      "learning_rate": 0.00012013422818791946,
+      "loss": 0.8784,
+      "step": 481
+    },
+    {
+      "epoch": 1.2052824880831445,
+      "grad_norm": 0.2142104059457779,
+      "learning_rate": 0.00011996644295302013,
+      "loss": 0.8668,
+      "step": 482
+    },
+    {
+      "epoch": 1.2077830741580058,
+      "grad_norm": 0.2258850336074829,
+      "learning_rate": 0.00011979865771812082,
+      "loss": 0.8904,
+      "step": 483
+    },
+    {
+      "epoch": 1.210283660232867,
+      "grad_norm": 0.21830275654792786,
+      "learning_rate": 0.00011963087248322149,
+      "loss": 0.8879,
+      "step": 484
+    },
+    {
+      "epoch": 1.2127842463077283,
+      "grad_norm": 0.2106282114982605,
+      "learning_rate": 0.00011946308724832216,
+      "loss": 0.8283,
+      "step": 485
+    },
+    {
+      "epoch": 1.2152848323825896,
+      "grad_norm": 0.21384255588054657,
+      "learning_rate": 0.00011929530201342283,
+      "loss": 0.8397,
+      "step": 486
+    },
+    {
+      "epoch": 1.217785418457451,
+      "grad_norm": 0.22718624770641327,
+      "learning_rate": 0.0001191275167785235,
+      "loss": 0.8968,
+      "step": 487
+    },
+    {
+      "epoch": 1.2202860045323123,
+      "grad_norm": 0.21989531815052032,
+      "learning_rate": 0.00011895973154362416,
+      "loss": 0.8884,
+      "step": 488
+    },
+    {
+      "epoch": 1.2227865906071735,
+      "grad_norm": 0.224409282207489,
+      "learning_rate": 0.00011879194630872483,
+      "loss": 0.9135,
+      "step": 489
+    },
+    {
+      "epoch": 1.2252871766820348,
+      "grad_norm": 0.21789176762104034,
+      "learning_rate": 0.0001186241610738255,
+      "loss": 0.8462,
+      "step": 490
+    },
+    {
+      "epoch": 1.227787762756896,
+      "grad_norm": 0.21743899583816528,
+      "learning_rate": 0.00011845637583892618,
+      "loss": 0.8244,
+      "step": 491
+    },
+    {
+      "epoch": 1.2302883488317575,
+      "grad_norm": 0.22736044228076935,
+      "learning_rate": 0.00011828859060402686,
+      "loss": 0.9055,
+      "step": 492
+    },
+    {
+      "epoch": 1.2327889349066188,
+      "grad_norm": 0.22548165917396545,
+      "learning_rate": 0.00011812080536912754,
+      "loss": 0.821,
+      "step": 493
+    },
+    {
+      "epoch": 1.23528952098148,
+      "grad_norm": 0.22150689363479614,
+      "learning_rate": 0.0001179530201342282,
+      "loss": 0.8681,
+      "step": 494
+    },
+    {
+      "epoch": 1.2377901070563413,
+      "grad_norm": 0.21708747744560242,
+      "learning_rate": 0.00011778523489932887,
+      "loss": 0.8926,
+      "step": 495
+    },
+    {
+      "epoch": 1.2402906931312025,
+      "grad_norm": 0.21993860602378845,
+      "learning_rate": 0.00011761744966442953,
+      "loss": 0.8876,
+      "step": 496
+    },
+    {
+      "epoch": 1.242791279206064,
+      "grad_norm": 0.21238721907138824,
+      "learning_rate": 0.0001174496644295302,
+      "loss": 0.8245,
+      "step": 497
+    },
+    {
+      "epoch": 1.2452918652809253,
+      "grad_norm": 0.22117340564727783,
+      "learning_rate": 0.00011728187919463087,
+      "loss": 0.9044,
+      "step": 498
+    },
+    {
+      "epoch": 1.2477924513557865,
+      "grad_norm": 0.21898695826530457,
+      "learning_rate": 0.00011711409395973155,
+      "loss": 0.9172,
+      "step": 499
+    },
+    {
+      "epoch": 1.2502930374306478,
+      "grad_norm": 0.22206924855709076,
+      "learning_rate": 0.00011694630872483221,
+      "loss": 0.8968,
+      "step": 500
+    },
+    {
+      "epoch": 1.252793623505509,
+      "grad_norm": 0.21640115976333618,
+      "learning_rate": 0.0001167785234899329,
+      "loss": 0.8456,
+      "step": 501
+    },
+    {
+      "epoch": 1.2552942095803705,
+      "grad_norm": 0.21233998239040375,
+      "learning_rate": 0.00011661073825503357,
+      "loss": 0.8522,
+      "step": 502
+    },
+    {
+      "epoch": 1.2577947956552316,
+      "grad_norm": 0.20975713431835175,
+      "learning_rate": 0.00011644295302013424,
+      "loss": 0.8824,
+      "step": 503
+    },
+    {
+      "epoch": 1.260295381730093,
+      "grad_norm": 0.21990106999874115,
+      "learning_rate": 0.0001162751677852349,
+      "loss": 0.8796,
+      "step": 504
+    },
+    {
+      "epoch": 1.2627959678049543,
+      "grad_norm": 0.21988140046596527,
+      "learning_rate": 0.00011610738255033557,
+      "loss": 0.8631,
+      "step": 505
+    },
+    {
+      "epoch": 1.2652965538798155,
+      "grad_norm": 0.20216408371925354,
+      "learning_rate": 0.00011593959731543624,
+      "loss": 0.8221,
+      "step": 506
+    },
+    {
+      "epoch": 1.2677971399546768,
+      "grad_norm": 0.20301879942417145,
+      "learning_rate": 0.00011577181208053692,
+      "loss": 0.8653,
+      "step": 507
+    },
+    {
+      "epoch": 1.270297726029538,
+      "grad_norm": 0.215713769197464,
+      "learning_rate": 0.00011560402684563758,
+      "loss": 0.8828,
+      "step": 508
+    },
+    {
+      "epoch": 1.2727983121043995,
+      "grad_norm": 0.22484052181243896,
+      "learning_rate": 0.00011543624161073828,
+      "loss": 0.8867,
+      "step": 509
+    },
+    {
+      "epoch": 1.2752988981792608,
+      "grad_norm": 0.2336317002773285,
+      "learning_rate": 0.00011526845637583894,
+      "loss": 0.8458,
+      "step": 510
+    },
+    {
+      "epoch": 1.277799484254122,
+      "grad_norm": 0.21190722286701202,
+      "learning_rate": 0.00011510067114093961,
+      "loss": 0.9453,
+      "step": 511
+    },
+    {
+      "epoch": 1.2803000703289833,
+      "grad_norm": 0.22565364837646484,
+      "learning_rate": 0.00011493288590604028,
+      "loss": 0.8232,
+      "step": 512
+    },
+    {
+      "epoch": 1.2828006564038446,
+      "grad_norm": 0.220944344997406,
+      "learning_rate": 0.00011476510067114094,
+      "loss": 0.9304,
+      "step": 513
+    },
+    {
+      "epoch": 1.285301242478706,
+      "grad_norm": 0.22200433909893036,
+      "learning_rate": 0.00011459731543624161,
+      "loss": 0.9019,
+      "step": 514
+    },
+    {
+      "epoch": 1.2878018285535673,
+      "grad_norm": 0.2165757268667221,
+      "learning_rate": 0.00011442953020134229,
+      "loss": 0.9023,
+      "step": 515
+    },
+    {
+      "epoch": 1.2903024146284285,
+      "grad_norm": 0.22672519087791443,
+      "learning_rate": 0.00011426174496644295,
+      "loss": 0.9161,
+      "step": 516
+    },
+    {
+      "epoch": 1.2928030007032898,
+      "grad_norm": 0.23044028878211975,
+      "learning_rate": 0.00011409395973154362,
+      "loss": 0.8841,
+      "step": 517
+    },
+    {
+      "epoch": 1.295303586778151,
+      "grad_norm": 0.20665666460990906,
+      "learning_rate": 0.00011392617449664431,
+      "loss": 0.8676,
+      "step": 518
+    },
+    {
+      "epoch": 1.2978041728530125,
+      "grad_norm": 0.21918492019176483,
+      "learning_rate": 0.00011375838926174498,
+      "loss": 0.9035,
+      "step": 519
+    },
+    {
+      "epoch": 1.3003047589278738,
+      "grad_norm": 0.2200019806623459,
+      "learning_rate": 0.00011359060402684565,
+      "loss": 0.8404,
+      "step": 520
+    },
+    {
+      "epoch": 1.302805345002735,
+      "grad_norm": 0.23530909419059753,
+      "learning_rate": 0.00011342281879194631,
+      "loss": 0.841,
+      "step": 521
+    },
+    {
+      "epoch": 1.3053059310775963,
+      "grad_norm": 0.2111617475748062,
+      "learning_rate": 0.00011325503355704699,
+      "loss": 0.8449,
+      "step": 522
+    },
+    {
+      "epoch": 1.3078065171524575,
+      "grad_norm": 0.24097535014152527,
+      "learning_rate": 0.00011308724832214766,
+      "loss": 0.9217,
+      "step": 523
+    },
+    {
+      "epoch": 1.310307103227319,
+      "grad_norm": 0.22256498038768768,
+      "learning_rate": 0.00011291946308724832,
+      "loss": 0.852,
+      "step": 524
+    },
+    {
+      "epoch": 1.3128076893021803,
+      "grad_norm": 0.22531381249427795,
+      "learning_rate": 0.00011275167785234899,
+      "loss": 0.8572,
+      "step": 525
+    },
+    {
+      "epoch": 1.3153082753770415,
+      "grad_norm": 0.22543148696422577,
+      "learning_rate": 0.00011258389261744966,
+      "loss": 0.8746,
+      "step": 526
+    },
+    {
+      "epoch": 1.3178088614519028,
+      "grad_norm": 0.2404658943414688,
+      "learning_rate": 0.00011241610738255035,
+      "loss": 0.8863,
+      "step": 527
+    },
+    {
+      "epoch": 1.320309447526764,
+      "grad_norm": 0.2268204241991043,
+      "learning_rate": 0.00011224832214765102,
+      "loss": 0.8459,
+      "step": 528
+    },
+    {
+      "epoch": 1.3228100336016253,
+      "grad_norm": 0.22067099809646606,
+      "learning_rate": 0.00011208053691275168,
+      "loss": 0.8461,
+      "step": 529
+    },
+    {
+      "epoch": 1.3253106196764866,
+      "grad_norm": 0.2447136491537094,
+      "learning_rate": 0.00011191275167785236,
+      "loss": 0.8311,
+      "step": 530
+    },
+    {
+      "epoch": 1.327811205751348,
+      "grad_norm": 0.22762243449687958,
+      "learning_rate": 0.00011174496644295303,
+      "loss": 0.8719,
+      "step": 531
+    },
+    {
+      "epoch": 1.3303117918262093,
+      "grad_norm": 0.2170940786600113,
+      "learning_rate": 0.0001115771812080537,
+      "loss": 0.8385,
+      "step": 532
+    },
+    {
+      "epoch": 1.3328123779010705,
+      "grad_norm": 0.22466519474983215,
+      "learning_rate": 0.00011140939597315436,
+      "loss": 0.892,
+      "step": 533
+    },
+    {
+      "epoch": 1.3353129639759318,
+      "grad_norm": 0.22515322268009186,
+      "learning_rate": 0.00011124161073825503,
+      "loss": 0.8454,
+      "step": 534
+    },
+    {
+      "epoch": 1.337813550050793,
+      "grad_norm": 0.22016112506389618,
+      "learning_rate": 0.0001110738255033557,
+      "loss": 0.8898,
+      "step": 535
+    },
+    {
+      "epoch": 1.3403141361256545,
+      "grad_norm": 0.2235654890537262,
+      "learning_rate": 0.00011090604026845639,
+      "loss": 0.8653,
+      "step": 536
+    },
+    {
+      "epoch": 1.3428147222005158,
+      "grad_norm": 0.2224671095609665,
+      "learning_rate": 0.00011073825503355705,
+      "loss": 0.835,
+      "step": 537
+    },
+    {
+      "epoch": 1.345315308275377,
+      "grad_norm": 0.21484190225601196,
+      "learning_rate": 0.00011057046979865773,
+      "loss": 0.9014,
+      "step": 538
+    },
+    {
+      "epoch": 1.3478158943502383,
+      "grad_norm": 0.2207605540752411,
+      "learning_rate": 0.0001104026845637584,
+      "loss": 0.8689,
+      "step": 539
+    },
+    {
+      "epoch": 1.3503164804250996,
+      "grad_norm": 0.21565458178520203,
+      "learning_rate": 0.00011023489932885907,
+      "loss": 0.833,
+      "step": 540
+    },
+    {
+      "epoch": 1.352817066499961,
+      "grad_norm": 0.21194227039813995,
+      "learning_rate": 0.00011006711409395973,
+      "loss": 0.791,
+      "step": 541
+    },
+    {
+      "epoch": 1.3553176525748223,
+      "grad_norm": 0.21446503698825836,
+      "learning_rate": 0.0001098993288590604,
+      "loss": 0.8244,
+      "step": 542
+    },
+    {
+      "epoch": 1.3578182386496835,
+      "grad_norm": 0.22496376931667328,
+      "learning_rate": 0.00010973154362416106,
+      "loss": 0.846,
+      "step": 543
+    },
+    {
+      "epoch": 1.3603188247245448,
+      "grad_norm": 0.22478918731212616,
+      "learning_rate": 0.00010956375838926174,
+      "loss": 0.8229,
+      "step": 544
+    },
+    {
+      "epoch": 1.362819410799406,
+      "grad_norm": 0.2262524515390396,
+      "learning_rate": 0.00010939597315436242,
+      "loss": 0.8492,
+      "step": 545
+    },
+    {
+      "epoch": 1.3653199968742675,
+      "grad_norm": 0.23414728045463562,
+      "learning_rate": 0.0001092281879194631,
+      "loss": 0.8673,
+      "step": 546
+    },
+    {
+      "epoch": 1.3678205829491288,
+      "grad_norm": 0.2247745245695114,
+      "learning_rate": 0.00010906040268456377,
+      "loss": 0.8434,
+      "step": 547
+    },
+    {
+      "epoch": 1.37032116902399,
+      "grad_norm": 0.22427400946617126,
+      "learning_rate": 0.00010889261744966444,
+      "loss": 0.8608,
+      "step": 548
+    },
+    {
+      "epoch": 1.3728217550988513,
+      "grad_norm": 0.2288215309381485,
+      "learning_rate": 0.0001087248322147651,
+      "loss": 0.8366,
+      "step": 549
+    },
+    {
+      "epoch": 1.3753223411737125,
+      "grad_norm": 0.22003835439682007,
+      "learning_rate": 0.00010855704697986577,
+      "loss": 0.8375,
+      "step": 550
+    },
+    {
+      "epoch": 1.3778229272485738,
+      "grad_norm": 0.2228715866804123,
+      "learning_rate": 0.00010838926174496644,
+      "loss": 0.832,
+      "step": 551
+    },
+    {
+      "epoch": 1.380323513323435,
+      "grad_norm": 0.21661970019340515,
+      "learning_rate": 0.00010822147651006711,
+      "loss": 0.8939,
+      "step": 552
+    },
+    {
+      "epoch": 1.3828240993982965,
+      "grad_norm": 0.24519488215446472,
+      "learning_rate": 0.0001080536912751678,
+      "loss": 0.8524,
+      "step": 553
+    },
+    {
+      "epoch": 1.3853246854731578,
+      "grad_norm": 0.2249690592288971,
+      "learning_rate": 0.00010788590604026847,
+      "loss": 0.8269,
+      "step": 554
+    },
+    {
+      "epoch": 1.387825271548019,
+      "grad_norm": 0.22413256764411926,
+      "learning_rate": 0.00010771812080536914,
+      "loss": 0.8482,
+      "step": 555
+    },
+    {
+      "epoch": 1.3903258576228803,
+      "grad_norm": 0.21265004575252533,
+      "learning_rate": 0.00010755033557046981,
+      "loss": 0.8818,
+      "step": 556
+    },
+    {
+      "epoch": 1.3928264436977416,
+      "grad_norm": 0.20840033888816833,
+      "learning_rate": 0.00010738255033557047,
+      "loss": 0.8407,
+      "step": 557
+    },
+    {
+      "epoch": 1.395327029772603,
+      "grad_norm": 0.226187065243721,
+      "learning_rate": 0.00010721476510067114,
+      "loss": 0.9044,
+      "step": 558
+    },
+    {
+      "epoch": 1.3978276158474643,
+      "grad_norm": 0.21541696786880493,
+      "learning_rate": 0.00010704697986577182,
+      "loss": 0.8897,
+      "step": 559
+    },
+    {
+      "epoch": 1.4003282019223255,
+      "grad_norm": 0.22921550273895264,
+      "learning_rate": 0.00010687919463087249,
+      "loss": 0.8933,
+      "step": 560
+    },
+    {
+      "epoch": 1.4028287879971868,
+      "grad_norm": 0.2249586284160614,
+      "learning_rate": 0.00010671140939597315,
+      "loss": 0.8736,
+      "step": 561
+    },
+    {
+      "epoch": 1.405329374072048,
+      "grad_norm": 0.21724307537078857,
+      "learning_rate": 0.00010654362416107385,
+      "loss": 0.8191,
+      "step": 562
+    },
+    {
+      "epoch": 1.4078299601469095,
+      "grad_norm": 0.2127966582775116,
+      "learning_rate": 0.00010637583892617451,
+      "loss": 0.8848,
+      "step": 563
+    },
+    {
+      "epoch": 1.4103305462217708,
+      "grad_norm": 0.21389761567115784,
+      "learning_rate": 0.00010620805369127518,
+      "loss": 0.8329,
+      "step": 564
+    },
+    {
+      "epoch": 1.412831132296632,
+      "grad_norm": 0.2194439023733139,
+      "learning_rate": 0.00010604026845637584,
+      "loss": 0.8426,
+      "step": 565
+    },
+    {
+      "epoch": 1.4153317183714933,
+      "grad_norm": 0.2189185917377472,
+      "learning_rate": 0.00010587248322147651,
+      "loss": 0.8279,
+      "step": 566
+    },
+    {
+      "epoch": 1.4178323044463546,
+      "grad_norm": 0.22137369215488434,
+      "learning_rate": 0.00010570469798657719,
+      "loss": 0.9193,
+      "step": 567
+    },
+    {
+      "epoch": 1.420332890521216,
+      "grad_norm": 0.22823460400104523,
+      "learning_rate": 0.00010553691275167786,
+      "loss": 0.8647,
+      "step": 568
+    },
+    {
+      "epoch": 1.4228334765960773,
+      "grad_norm": 0.20698755979537964,
+      "learning_rate": 0.00010536912751677852,
+      "loss": 0.8856,
+      "step": 569
+    },
+    {
+      "epoch": 1.4253340626709385,
+      "grad_norm": 0.22674210369586945,
+      "learning_rate": 0.00010520134228187919,
+      "loss": 0.8128,
+      "step": 570
+    },
+    {
+      "epoch": 1.4278346487457998,
+      "grad_norm": 0.2153196781873703,
+      "learning_rate": 0.00010503355704697988,
+      "loss": 0.812,
+      "step": 571
+    },
+    {
+      "epoch": 1.430335234820661,
+      "grad_norm": 0.21849411725997925,
+      "learning_rate": 0.00010486577181208055,
+      "loss": 0.8489,
+      "step": 572
+    },
+    {
+      "epoch": 1.4328358208955223,
+      "grad_norm": 0.23009636998176575,
+      "learning_rate": 0.00010469798657718121,
+      "loss": 0.8414,
+      "step": 573
+    },
+    {
+      "epoch": 1.4353364069703836,
+      "grad_norm": 0.22925853729248047,
+      "learning_rate": 0.00010453020134228188,
+      "loss": 0.8803,
+      "step": 574
+    },
+    {
+      "epoch": 1.437836993045245,
+      "grad_norm": 0.22622284293174744,
+      "learning_rate": 0.00010436241610738256,
+      "loss": 0.8608,
+      "step": 575
+    },
+    {
+      "epoch": 1.4403375791201063,
+      "grad_norm": 0.23347701132297516,
+      "learning_rate": 0.00010419463087248323,
+      "loss": 0.9013,
+      "step": 576
+    },
+    {
+      "epoch": 1.4428381651949675,
+      "grad_norm": 0.2254674881696701,
+      "learning_rate": 0.00010402684563758389,
+      "loss": 0.9116,
+      "step": 577
+    },
+    {
+      "epoch": 1.4453387512698288,
+      "grad_norm": 0.21847471594810486,
+      "learning_rate": 0.00010385906040268456,
+      "loss": 0.8496,
+      "step": 578
+    },
+    {
+      "epoch": 1.44783933734469,
+      "grad_norm": 0.21672604978084564,
+      "learning_rate": 0.00010369127516778523,
+      "loss": 0.8255,
+      "step": 579
+    },
+    {
+      "epoch": 1.4503399234195515,
+      "grad_norm": 0.22864758968353271,
+      "learning_rate": 0.00010352348993288592,
+      "loss": 0.8923,
+      "step": 580
+    },
+    {
+      "epoch": 1.4528405094944128,
+      "grad_norm": 0.22856175899505615,
+      "learning_rate": 0.00010335570469798659,
+      "loss": 0.8764,
+      "step": 581
+    },
+    {
+      "epoch": 1.455341095569274,
+      "grad_norm": 0.23273620009422302,
+      "learning_rate": 0.00010318791946308725,
+      "loss": 0.8663,
+      "step": 582
+    },
+    {
+      "epoch": 1.4578416816441353,
+      "grad_norm": 0.22231343388557434,
+      "learning_rate": 0.00010302013422818793,
+      "loss": 0.8071,
+      "step": 583
+    },
+    {
+      "epoch": 1.4603422677189966,
+      "grad_norm": 0.2304021120071411,
+      "learning_rate": 0.0001028523489932886,
+      "loss": 0.9086,
+      "step": 584
+    },
+    {
+      "epoch": 1.462842853793858,
+      "grad_norm": 0.22386790812015533,
+      "learning_rate": 0.00010268456375838926,
+      "loss": 0.819,
+      "step": 585
+    },
+    {
+      "epoch": 1.4653434398687193,
+      "grad_norm": 0.22458797693252563,
+      "learning_rate": 0.00010251677852348993,
+      "loss": 0.7978,
+      "step": 586
+    },
+    {
+      "epoch": 1.4678440259435805,
+      "grad_norm": 0.21856428682804108,
+      "learning_rate": 0.0001023489932885906,
+      "loss": 0.8349,
+      "step": 587
+    },
+    {
+      "epoch": 1.4703446120184418,
+      "grad_norm": 0.23313550651073456,
+      "learning_rate": 0.00010218120805369129,
+      "loss": 0.8727,
+      "step": 588
+    },
+    {
+      "epoch": 1.472845198093303,
+      "grad_norm": 0.23794901371002197,
+      "learning_rate": 0.00010201342281879196,
+      "loss": 0.8482,
+      "step": 589
+    },
+    {
+      "epoch": 1.4753457841681645,
+      "grad_norm": 0.2224135547876358,
+      "learning_rate": 0.00010184563758389262,
+      "loss": 0.875,
+      "step": 590
+    },
+    {
+      "epoch": 1.4778463702430258,
+      "grad_norm": 0.2217201292514801,
+      "learning_rate": 0.0001016778523489933,
+      "loss": 0.8375,
+      "step": 591
+    },
+    {
+      "epoch": 1.480346956317887,
+      "grad_norm": 0.2245808243751526,
+      "learning_rate": 0.00010151006711409397,
+      "loss": 0.8592,
+      "step": 592
+    },
+    {
+      "epoch": 1.4828475423927483,
+      "grad_norm": 0.22494655847549438,
+      "learning_rate": 0.00010134228187919463,
+      "loss": 0.8862,
+      "step": 593
+    },
+    {
+      "epoch": 1.4853481284676096,
+      "grad_norm": 0.21345628798007965,
+      "learning_rate": 0.0001011744966442953,
+      "loss": 0.8319,
+      "step": 594
+    },
+    {
+      "epoch": 1.487848714542471,
+      "grad_norm": 0.22361332178115845,
+      "learning_rate": 0.00010100671140939597,
+      "loss": 0.8489,
+      "step": 595
+    },
+    {
+      "epoch": 1.490349300617332,
+      "grad_norm": 0.2232082635164261,
+      "learning_rate": 0.00010083892617449663,
+      "loss": 0.8381,
+      "step": 596
+    },
+    {
+      "epoch": 1.4928498866921935,
+      "grad_norm": 0.22092638909816742,
+      "learning_rate": 0.00010067114093959733,
+      "loss": 0.8037,
+      "step": 597
+    },
+    {
+      "epoch": 1.4953504727670548,
+      "grad_norm": 0.22525610029697418,
+      "learning_rate": 0.00010050335570469799,
+      "loss": 0.8146,
+      "step": 598
+    },
+    {
+      "epoch": 1.497851058841916,
+      "grad_norm": 0.2206062525510788,
+      "learning_rate": 0.00010033557046979867,
+      "loss": 0.8784,
+      "step": 599
+    },
+    {
+      "epoch": 1.5003516449167775,
+      "grad_norm": 0.24318885803222656,
+      "learning_rate": 0.00010016778523489934,
+      "loss": 0.8551,
+      "step": 600
+    },
+    {
+      "epoch": 1.5028522309916386,
+      "grad_norm": 0.21619702875614166,
+      "learning_rate": 0.0001,
+      "loss": 0.8425,
+      "step": 601
+    },
+    {
+      "epoch": 1.5053528170665,
+      "grad_norm": 0.2159229964017868,
+      "learning_rate": 9.983221476510067e-05,
+      "loss": 0.8347,
+      "step": 602
+    },
+    {
+      "epoch": 1.5078534031413613,
+      "grad_norm": 0.22807268798351288,
+      "learning_rate": 9.966442953020134e-05,
+      "loss": 0.861,
+      "step": 603
+    },
+    {
+      "epoch": 1.5103539892162225,
+      "grad_norm": 0.22900323569774628,
+      "learning_rate": 9.949664429530202e-05,
+      "loss": 0.8868,
+      "step": 604
+    },
+    {
+      "epoch": 1.5128545752910838,
+      "grad_norm": 0.23081159591674805,
+      "learning_rate": 9.93288590604027e-05,
+      "loss": 0.8667,
+      "step": 605
+    },
+    {
+      "epoch": 1.515355161365945,
+      "grad_norm": 0.21598628163337708,
+      "learning_rate": 9.916107382550336e-05,
+      "loss": 0.8762,
+      "step": 606
+    },
+    {
+      "epoch": 1.5178557474408065,
+      "grad_norm": 0.2459835261106491,
+      "learning_rate": 9.899328859060403e-05,
+      "loss": 0.8683,
+      "step": 607
+    },
+    {
+      "epoch": 1.5203563335156676,
+      "grad_norm": 0.23040466010570526,
+      "learning_rate": 9.882550335570471e-05,
+      "loss": 0.803,
+      "step": 608
+    },
+    {
+      "epoch": 1.522856919590529,
+      "grad_norm": 0.21725349128246307,
+      "learning_rate": 9.865771812080538e-05,
+      "loss": 0.8495,
+      "step": 609
+    },
+    {
+      "epoch": 1.5253575056653903,
+      "grad_norm": 0.22498026490211487,
+      "learning_rate": 9.848993288590604e-05,
+      "loss": 0.8316,
+      "step": 610
+    },
+    {
+      "epoch": 1.5278580917402516,
+      "grad_norm": 0.22257371246814728,
+      "learning_rate": 9.832214765100671e-05,
+      "loss": 0.8404,
+      "step": 611
+    },
+    {
+      "epoch": 1.530358677815113,
+      "grad_norm": 0.2255915403366089,
+      "learning_rate": 9.815436241610739e-05,
+      "loss": 0.8486,
+      "step": 612
+    },
+    {
+      "epoch": 1.532859263889974,
+      "grad_norm": 0.22992673516273499,
+      "learning_rate": 9.798657718120807e-05,
+      "loss": 0.8409,
+      "step": 613
+    },
+    {
+      "epoch": 1.5353598499648355,
+      "grad_norm": 0.2263323813676834,
+      "learning_rate": 9.781879194630873e-05,
+      "loss": 0.7965,
+      "step": 614
+    },
+    {
+      "epoch": 1.5378604360396968,
+      "grad_norm": 0.23883429169654846,
+      "learning_rate": 9.76510067114094e-05,
+      "loss": 0.7889,
+      "step": 615
+    },
+    {
+      "epoch": 1.540361022114558,
+      "grad_norm": 0.23129902780056,
+      "learning_rate": 9.748322147651007e-05,
+      "loss": 0.8467,
+      "step": 616
+    },
+    {
+      "epoch": 1.5428616081894195,
+      "grad_norm": 0.2277020663022995,
+      "learning_rate": 9.731543624161075e-05,
+      "loss": 0.8758,
+      "step": 617
+    },
+    {
+      "epoch": 1.5453621942642806,
+      "grad_norm": 0.23400239646434784,
+      "learning_rate": 9.714765100671141e-05,
+      "loss": 0.8699,
+      "step": 618
+    },
+    {
+      "epoch": 1.547862780339142,
+      "grad_norm": 0.2309494912624359,
+      "learning_rate": 9.697986577181208e-05,
+      "loss": 0.8757,
+      "step": 619
+    },
+    {
+      "epoch": 1.5503633664140033,
+      "grad_norm": 0.22626151144504547,
+      "learning_rate": 9.681208053691276e-05,
+      "loss": 0.8471,
+      "step": 620
+    },
+    {
+      "epoch": 1.5528639524888646,
+      "grad_norm": 0.22830310463905334,
+      "learning_rate": 9.664429530201344e-05,
+      "loss": 0.8608,
+      "step": 621
+    },
+    {
+      "epoch": 1.555364538563726,
+      "grad_norm": 0.2295883148908615,
+      "learning_rate": 9.64765100671141e-05,
+      "loss": 0.8534,
+      "step": 622
+    },
+    {
+      "epoch": 1.557865124638587,
+      "grad_norm": 0.21810783445835114,
+      "learning_rate": 9.630872483221477e-05,
+      "loss": 0.8253,
+      "step": 623
+    },
+    {
+      "epoch": 1.5603657107134485,
+      "grad_norm": 0.21453121304512024,
+      "learning_rate": 9.614093959731544e-05,
+      "loss": 0.8384,
+      "step": 624
+    },
+    {
+      "epoch": 1.5628662967883098,
+      "grad_norm": 0.23363538086414337,
+      "learning_rate": 9.59731543624161e-05,
+      "loss": 0.8275,
+      "step": 625
+    },
+    {
+      "epoch": 1.565366882863171,
+      "grad_norm": 0.2258090227842331,
+      "learning_rate": 9.580536912751678e-05,
+      "loss": 0.8836,
+      "step": 626
+    },
+    {
+      "epoch": 1.5678674689380323,
+      "grad_norm": 0.23416352272033691,
+      "learning_rate": 9.563758389261745e-05,
+      "loss": 0.8198,
+      "step": 627
+    },
+    {
+      "epoch": 1.5703680550128936,
+      "grad_norm": 0.22489909827709198,
+      "learning_rate": 9.546979865771813e-05,
+      "loss": 0.862,
+      "step": 628
+    },
+    {
+      "epoch": 1.572868641087755,
+      "grad_norm": 0.2406560480594635,
+      "learning_rate": 9.53020134228188e-05,
+      "loss": 0.8648,
+      "step": 629
+    },
+    {
+      "epoch": 1.5753692271626163,
+      "grad_norm": 0.2455638349056244,
+      "learning_rate": 9.513422818791947e-05,
+      "loss": 0.8588,
+      "step": 630
+    },
+    {
+      "epoch": 1.5778698132374775,
+      "grad_norm": 0.21869002282619476,
+      "learning_rate": 9.496644295302014e-05,
+      "loss": 0.8354,
+      "step": 631
+    },
+    {
+      "epoch": 1.5803703993123388,
+      "grad_norm": 0.22325509786605835,
+      "learning_rate": 9.479865771812081e-05,
+      "loss": 0.8831,
+      "step": 632
+    },
+    {
+      "epoch": 1.5828709853872,
+      "grad_norm": 0.22697119414806366,
+      "learning_rate": 9.463087248322147e-05,
+      "loss": 0.8441,
+      "step": 633
+    },
+    {
+      "epoch": 1.5853715714620615,
+      "grad_norm": 0.2169933319091797,
+      "learning_rate": 9.446308724832215e-05,
+      "loss": 0.8339,
+      "step": 634
+    },
+    {
+      "epoch": 1.5878721575369226,
+      "grad_norm": 0.21638698875904083,
+      "learning_rate": 9.429530201342282e-05,
+      "loss": 0.7588,
+      "step": 635
+    },
+    {
+      "epoch": 1.590372743611784,
+      "grad_norm": 0.22790271043777466,
+      "learning_rate": 9.41275167785235e-05,
+      "loss": 0.8788,
+      "step": 636
+    },
+    {
+      "epoch": 1.5928733296866453,
+      "grad_norm": 0.22425347566604614,
+      "learning_rate": 9.395973154362417e-05,
+      "loss": 0.8306,
+      "step": 637
+    },
+    {
+      "epoch": 1.5953739157615066,
+      "grad_norm": 0.22301225364208221,
+      "learning_rate": 9.379194630872483e-05,
+      "loss": 0.8079,
+      "step": 638
+    },
+    {
+      "epoch": 1.597874501836368,
+      "grad_norm": 0.21887917816638947,
+      "learning_rate": 9.362416107382551e-05,
+      "loss": 0.868,
+      "step": 639
+    },
+    {
+      "epoch": 1.600375087911229,
+      "grad_norm": 0.2356102466583252,
+      "learning_rate": 9.345637583892618e-05,
+      "loss": 0.83,
+      "step": 640
+    },
+    {
+      "epoch": 1.6028756739860905,
+      "grad_norm": 0.2219531387090683,
+      "learning_rate": 9.328859060402684e-05,
+      "loss": 0.8631,
+      "step": 641
+    },
+    {
+      "epoch": 1.6053762600609518,
+      "grad_norm": 0.22262972593307495,
+      "learning_rate": 9.312080536912751e-05,
+      "loss": 0.8269,
+      "step": 642
+    },
+    {
+      "epoch": 1.607876846135813,
+      "grad_norm": 0.2268291860818863,
+      "learning_rate": 9.295302013422819e-05,
+      "loss": 0.8336,
+      "step": 643
+    },
+    {
+      "epoch": 1.6103774322106745,
+      "grad_norm": 0.2192980945110321,
+      "learning_rate": 9.278523489932887e-05,
+      "loss": 0.8978,
+      "step": 644
+    },
+    {
+      "epoch": 1.6128780182855356,
+      "grad_norm": 0.22530333697795868,
+      "learning_rate": 9.261744966442954e-05,
+      "loss": 0.8256,
+      "step": 645
+    },
+    {
+      "epoch": 1.615378604360397,
+      "grad_norm": 0.22764861583709717,
+      "learning_rate": 9.24496644295302e-05,
+      "loss": 0.8948,
+      "step": 646
+    },
+    {
+      "epoch": 1.6178791904352583,
+      "grad_norm": 0.2260412871837616,
+      "learning_rate": 9.228187919463087e-05,
+      "loss": 0.8654,
+      "step": 647
+    },
+    {
+      "epoch": 1.6203797765101196,
+      "grad_norm": 0.2225823849439621,
+      "learning_rate": 9.211409395973155e-05,
+      "loss": 0.8481,
+      "step": 648
+    },
+    {
+      "epoch": 1.6228803625849808,
+      "grad_norm": 0.21501144766807556,
+      "learning_rate": 9.194630872483221e-05,
+      "loss": 0.8549,
+      "step": 649
+    },
+    {
+      "epoch": 1.625380948659842,
+      "grad_norm": 0.22539402544498444,
+      "learning_rate": 9.17785234899329e-05,
+      "loss": 0.9034,
+      "step": 650
+    },
+    {
+      "epoch": 1.6278815347347035,
+      "grad_norm": 0.22741571068763733,
+      "learning_rate": 9.161073825503356e-05,
+      "loss": 0.8634,
+      "step": 651
+    },
+    {
+      "epoch": 1.6303821208095648,
+      "grad_norm": 0.224626824259758,
+      "learning_rate": 9.144295302013424e-05,
+      "loss": 0.8347,
+      "step": 652
+    },
+    {
+      "epoch": 1.632882706884426,
+      "grad_norm": 0.22715142369270325,
+      "learning_rate": 9.127516778523491e-05,
+      "loss": 0.8584,
+      "step": 653
+    },
+    {
+      "epoch": 1.6353832929592873,
+      "grad_norm": 0.2303231656551361,
+      "learning_rate": 9.110738255033557e-05,
+      "loss": 0.8659,
+      "step": 654
+    },
+    {
+      "epoch": 1.6378838790341486,
+      "grad_norm": 0.2309199720621109,
+      "learning_rate": 9.093959731543624e-05,
+      "loss": 0.8252,
+      "step": 655
+    },
+    {
+      "epoch": 1.64038446510901,
+      "grad_norm": 0.21832013130187988,
+      "learning_rate": 9.077181208053692e-05,
+      "loss": 0.8502,
+      "step": 656
+    },
+    {
+      "epoch": 1.642885051183871,
+      "grad_norm": 0.21867471933364868,
+      "learning_rate": 9.060402684563759e-05,
+      "loss": 0.8292,
+      "step": 657
+    },
+    {
+      "epoch": 1.6453856372587325,
+      "grad_norm": 0.21259494125843048,
+      "learning_rate": 9.043624161073827e-05,
+      "loss": 0.8247,
+      "step": 658
+    },
+    {
+      "epoch": 1.6478862233335938,
+      "grad_norm": 0.2309045046567917,
+      "learning_rate": 9.026845637583893e-05,
+      "loss": 0.8403,
+      "step": 659
+    },
+    {
+      "epoch": 1.650386809408455,
+      "grad_norm": 0.23072423040866852,
+      "learning_rate": 9.01006711409396e-05,
+      "loss": 0.8539,
+      "step": 660
+    },
+    {
+      "epoch": 1.6528873954833165,
+      "grad_norm": 0.2295873463153839,
+      "learning_rate": 8.993288590604028e-05,
+      "loss": 0.8867,
+      "step": 661
+    },
+    {
+      "epoch": 1.6553879815581776,
+      "grad_norm": 0.22668391466140747,
+      "learning_rate": 8.976510067114094e-05,
+      "loss": 0.8672,
+      "step": 662
+    },
+    {
+      "epoch": 1.657888567633039,
+      "grad_norm": 0.2313709855079651,
+      "learning_rate": 8.959731543624161e-05,
+      "loss": 0.8628,
+      "step": 663
+    },
+    {
+      "epoch": 1.6603891537079003,
+      "grad_norm": 0.2152678519487381,
+      "learning_rate": 8.942953020134228e-05,
+      "loss": 0.8556,
+      "step": 664
+    },
+    {
+      "epoch": 1.6628897397827616,
+      "grad_norm": 0.21895945072174072,
+      "learning_rate": 8.926174496644296e-05,
+      "loss": 0.8467,
+      "step": 665
+    },
+    {
+      "epoch": 1.665390325857623,
+      "grad_norm": 0.21720942854881287,
+      "learning_rate": 8.909395973154364e-05,
+      "loss": 0.8465,
+      "step": 666
+    },
+    {
+      "epoch": 1.667890911932484,
+      "grad_norm": 0.21633416414260864,
+      "learning_rate": 8.89261744966443e-05,
+      "loss": 0.8874,
+      "step": 667
+    },
+    {
+      "epoch": 1.6703914980073455,
+      "grad_norm": 0.23894019424915314,
+      "learning_rate": 8.875838926174497e-05,
+      "loss": 0.8756,
+      "step": 668
+    },
+    {
+      "epoch": 1.6728920840822068,
+      "grad_norm": 0.2241196483373642,
+      "learning_rate": 8.859060402684565e-05,
+      "loss": 0.8464,
+      "step": 669
+    },
+    {
+      "epoch": 1.675392670157068,
+      "grad_norm": 0.23065364360809326,
+      "learning_rate": 8.842281879194631e-05,
+      "loss": 0.7974,
+      "step": 670
+    },
+    {
+      "epoch": 1.6778932562319293,
+      "grad_norm": 0.2277778834104538,
+      "learning_rate": 8.825503355704698e-05,
+      "loss": 0.8772,
+      "step": 671
+    },
+    {
+      "epoch": 1.6803938423067906,
+      "grad_norm": 0.22781328856945038,
+      "learning_rate": 8.808724832214765e-05,
+      "loss": 0.8172,
+      "step": 672
+    },
+    {
+      "epoch": 1.682894428381652,
+      "grad_norm": 0.23119190335273743,
+      "learning_rate": 8.791946308724833e-05,
+      "loss": 0.8341,
+      "step": 673
+    },
+    {
+      "epoch": 1.6853950144565133,
+      "grad_norm": 0.22981172800064087,
+      "learning_rate": 8.7751677852349e-05,
+      "loss": 0.8244,
+      "step": 674
+    },
+    {
+      "epoch": 1.6878956005313746,
+      "grad_norm": 0.23412439227104187,
+      "learning_rate": 8.758389261744967e-05,
+      "loss": 0.9385,
+      "step": 675
+    },
+    {
+      "epoch": 1.6903961866062358,
+      "grad_norm": 0.22738024592399597,
+      "learning_rate": 8.741610738255034e-05,
+      "loss": 0.8494,
+      "step": 676
+    },
+    {
+      "epoch": 1.692896772681097,
+      "grad_norm": 0.2343178242444992,
+      "learning_rate": 8.7248322147651e-05,
+      "loss": 0.883,
+      "step": 677
+    },
+    {
+      "epoch": 1.6953973587559585,
+      "grad_norm": 0.21647284924983978,
+      "learning_rate": 8.708053691275169e-05,
+      "loss": 0.8039,
+      "step": 678
+    },
+    {
+      "epoch": 1.6978979448308196,
+      "grad_norm": 0.2313070297241211,
+      "learning_rate": 8.691275167785235e-05,
+      "loss": 0.8636,
+      "step": 679
+    },
+    {
+      "epoch": 1.700398530905681,
+      "grad_norm": 0.24407722055912018,
+      "learning_rate": 8.674496644295302e-05,
+      "loss": 0.929,
+      "step": 680
+    },
+    {
+      "epoch": 1.7028991169805423,
+      "grad_norm": 0.2223997265100479,
+      "learning_rate": 8.65771812080537e-05,
+      "loss": 0.8592,
+      "step": 681
+    },
+    {
+      "epoch": 1.7053997030554036,
+      "grad_norm": 0.2273021787405014,
+      "learning_rate": 8.640939597315436e-05,
+      "loss": 0.8467,
+      "step": 682
+    },
+    {
+      "epoch": 1.707900289130265,
+      "grad_norm": 0.23875978589057922,
+      "learning_rate": 8.624161073825504e-05,
+      "loss": 0.816,
+      "step": 683
+    },
+    {
+      "epoch": 1.710400875205126,
+      "grad_norm": 0.22687745094299316,
+      "learning_rate": 8.607382550335571e-05,
+      "loss": 0.8571,
+      "step": 684
+    },
+    {
+      "epoch": 1.7129014612799875,
+      "grad_norm": 0.22464463114738464,
+      "learning_rate": 8.590604026845638e-05,
+      "loss": 0.8187,
+      "step": 685
+    },
+    {
+      "epoch": 1.7154020473548488,
+      "grad_norm": 0.22605834901332855,
+      "learning_rate": 8.573825503355704e-05,
+      "loss": 0.8531,
+      "step": 686
+    },
+    {
+      "epoch": 1.71790263342971,
+      "grad_norm": 0.23256367444992065,
+      "learning_rate": 8.557046979865772e-05,
+      "loss": 0.8316,
+      "step": 687
+    },
+    {
+      "epoch": 1.7204032195045715,
+      "grad_norm": 0.2281472086906433,
+      "learning_rate": 8.54026845637584e-05,
+      "loss": 0.8047,
+      "step": 688
+    },
+    {
+      "epoch": 1.7229038055794326,
+      "grad_norm": 0.23774386942386627,
+      "learning_rate": 8.523489932885907e-05,
+      "loss": 0.9113,
+      "step": 689
+    },
+    {
+      "epoch": 1.725404391654294,
+      "grad_norm": 0.2390678972005844,
+      "learning_rate": 8.506711409395973e-05,
+      "loss": 0.8327,
+      "step": 690
+    },
+    {
+      "epoch": 1.7279049777291553,
+      "grad_norm": 0.22964608669281006,
+      "learning_rate": 8.489932885906041e-05,
+      "loss": 0.8622,
+      "step": 691
+    },
+    {
+      "epoch": 1.7304055638040166,
+      "grad_norm": 0.24142296612262726,
+      "learning_rate": 8.473154362416108e-05,
+      "loss": 0.8602,
+      "step": 692
+    },
+    {
+      "epoch": 1.7329061498788778,
+      "grad_norm": 0.23625676333904266,
+      "learning_rate": 8.456375838926175e-05,
+      "loss": 0.8654,
+      "step": 693
+    },
+    {
+      "epoch": 1.735406735953739,
+      "grad_norm": 0.2253704071044922,
+      "learning_rate": 8.439597315436241e-05,
+      "loss": 0.8768,
+      "step": 694
+    },
+    {
+      "epoch": 1.7379073220286005,
+      "grad_norm": 0.23764395713806152,
+      "learning_rate": 8.422818791946309e-05,
+      "loss": 0.8118,
+      "step": 695
+    },
+    {
+      "epoch": 1.7404079081034618,
+      "grad_norm": 0.24044619500637054,
+      "learning_rate": 8.406040268456377e-05,
+      "loss": 0.8566,
+      "step": 696
+    },
+    {
+      "epoch": 1.742908494178323,
+      "grad_norm": 0.22577708959579468,
+      "learning_rate": 8.389261744966444e-05,
+      "loss": 0.8766,
+      "step": 697
+    },
+    {
+      "epoch": 1.7454090802531843,
+      "grad_norm": 0.23613469302654266,
+      "learning_rate": 8.37248322147651e-05,
+      "loss": 0.8329,
+      "step": 698
+    },
+    {
+      "epoch": 1.7479096663280456,
+      "grad_norm": 0.23061516880989075,
+      "learning_rate": 8.355704697986577e-05,
+      "loss": 0.8599,
+      "step": 699
+    },
+    {
+      "epoch": 1.750410252402907,
+      "grad_norm": 0.22349974513053894,
+      "learning_rate": 8.338926174496645e-05,
+      "loss": 0.8065,
+      "step": 700
+    },
+    {
+      "epoch": 1.752910838477768,
+      "grad_norm": 0.21980604529380798,
+      "learning_rate": 8.322147651006712e-05,
+      "loss": 0.8679,
+      "step": 701
+    },
+    {
+      "epoch": 1.7554114245526296,
+      "grad_norm": 0.2166835069656372,
+      "learning_rate": 8.305369127516778e-05,
+      "loss": 0.9043,
+      "step": 702
+    },
+    {
+      "epoch": 1.7579120106274908,
+      "grad_norm": 0.23712614178657532,
+      "learning_rate": 8.288590604026846e-05,
+      "loss": 0.8614,
+      "step": 703
+    },
+    {
+      "epoch": 1.760412596702352,
+      "grad_norm": 0.23757979273796082,
+      "learning_rate": 8.271812080536914e-05,
+      "loss": 0.9244,
+      "step": 704
+    },
+    {
+      "epoch": 1.7629131827772135,
+      "grad_norm": 0.234315425157547,
+      "learning_rate": 8.255033557046981e-05,
+      "loss": 0.8441,
+      "step": 705
+    },
+    {
+      "epoch": 1.7654137688520746,
+      "grad_norm": 0.2295805960893631,
+      "learning_rate": 8.238255033557048e-05,
+      "loss": 0.8959,
+      "step": 706
+    },
+    {
+      "epoch": 1.767914354926936,
+      "grad_norm": 0.22519482672214508,
+      "learning_rate": 8.221476510067114e-05,
+      "loss": 0.8719,
+      "step": 707
+    },
+    {
+      "epoch": 1.7704149410017973,
+      "grad_norm": 0.23038741946220398,
+      "learning_rate": 8.204697986577181e-05,
+      "loss": 0.8181,
+      "step": 708
+    },
+    {
+      "epoch": 1.7729155270766586,
+      "grad_norm": 0.2364625334739685,
+      "learning_rate": 8.187919463087249e-05,
+      "loss": 0.8573,
+      "step": 709
+    },
+    {
+      "epoch": 1.77541611315152,
+      "grad_norm": 0.23509962856769562,
+      "learning_rate": 8.171140939597315e-05,
+      "loss": 0.9127,
+      "step": 710
+    },
+    {
+      "epoch": 1.777916699226381,
+      "grad_norm": 0.23874014616012573,
+      "learning_rate": 8.154362416107383e-05,
+      "loss": 0.8757,
+      "step": 711
+    },
+    {
+      "epoch": 1.7804172853012425,
+      "grad_norm": 0.2340226024389267,
+      "learning_rate": 8.13758389261745e-05,
+      "loss": 0.8317,
+      "step": 712
+    },
+    {
+      "epoch": 1.7829178713761038,
+      "grad_norm": 0.22894112765789032,
+      "learning_rate": 8.120805369127518e-05,
+      "loss": 0.8421,
+      "step": 713
+    },
+    {
+      "epoch": 1.785418457450965,
+      "grad_norm": 0.21780632436275482,
+      "learning_rate": 8.104026845637585e-05,
+      "loss": 0.919,
+      "step": 714
+    },
+    {
+      "epoch": 1.7879190435258263,
+      "grad_norm": 0.23393581807613373,
+      "learning_rate": 8.087248322147651e-05,
+      "loss": 0.8736,
+      "step": 715
+    },
+    {
+      "epoch": 1.7904196296006876,
+      "grad_norm": 0.23031313717365265,
+      "learning_rate": 8.070469798657718e-05,
+      "loss": 0.9054,
+      "step": 716
+    },
+    {
+      "epoch": 1.792920215675549,
+      "grad_norm": 0.23269620537757874,
+      "learning_rate": 8.053691275167784e-05,
+      "loss": 0.8225,
+      "step": 717
+    },
+    {
+      "epoch": 1.7954208017504103,
+      "grad_norm": 0.2282569706439972,
+      "learning_rate": 8.036912751677852e-05,
+      "loss": 0.8775,
+      "step": 718
+    },
+    {
+      "epoch": 1.7979213878252716,
+      "grad_norm": 0.2403908669948578,
+      "learning_rate": 8.02013422818792e-05,
+      "loss": 0.8492,
+      "step": 719
+    },
+    {
+      "epoch": 1.8004219739001328,
+      "grad_norm": 0.23468557000160217,
+      "learning_rate": 8.003355704697987e-05,
+      "loss": 0.8807,
+      "step": 720
+    },
+    {
+      "epoch": 1.802922559974994,
+      "grad_norm": 0.23158180713653564,
+      "learning_rate": 7.986577181208054e-05,
+      "loss": 0.8796,
+      "step": 721
+    },
+    {
+      "epoch": 1.8054231460498555,
+      "grad_norm": 0.2401338666677475,
+      "learning_rate": 7.969798657718122e-05,
+      "loss": 0.871,
+      "step": 722
+    },
+    {
+      "epoch": 1.8079237321247166,
+      "grad_norm": 0.2347831428050995,
+      "learning_rate": 7.953020134228188e-05,
+      "loss": 0.8536,
+      "step": 723
+    },
+    {
+      "epoch": 1.810424318199578,
+      "grad_norm": 0.2261980026960373,
+      "learning_rate": 7.936241610738255e-05,
+      "loss": 0.7623,
+      "step": 724
+    },
+    {
+      "epoch": 1.8129249042744393,
+      "grad_norm": 0.23553761839866638,
+      "learning_rate": 7.919463087248322e-05,
+      "loss": 0.8679,
+      "step": 725
+    },
+    {
+      "epoch": 1.8154254903493006,
+      "grad_norm": 0.2279273271560669,
+      "learning_rate": 7.90268456375839e-05,
+      "loss": 0.8387,
+      "step": 726
+    },
+    {
+      "epoch": 1.817926076424162,
+      "grad_norm": 0.23345427215099335,
+      "learning_rate": 7.885906040268457e-05,
+      "loss": 0.842,
+      "step": 727
+    },
+    {
+      "epoch": 1.820426662499023,
+      "grad_norm": 0.2320943921804428,
+      "learning_rate": 7.869127516778524e-05,
+      "loss": 0.8422,
+      "step": 728
+    },
+    {
+      "epoch": 1.8229272485738846,
+      "grad_norm": 0.23071599006652832,
+      "learning_rate": 7.852348993288591e-05,
+      "loss": 0.8294,
+      "step": 729
+    },
+    {
+      "epoch": 1.8254278346487458,
+      "grad_norm": 0.22273236513137817,
+      "learning_rate": 7.835570469798657e-05,
+      "loss": 0.8639,
+      "step": 730
+    },
+    {
+      "epoch": 1.827928420723607,
+      "grad_norm": 0.22813788056373596,
+      "learning_rate": 7.818791946308725e-05,
+      "loss": 0.8102,
+      "step": 731
+    },
+    {
+      "epoch": 1.8304290067984685,
+      "grad_norm": 0.21939919888973236,
+      "learning_rate": 7.802013422818792e-05,
+      "loss": 0.8169,
+      "step": 732
+    },
+    {
+      "epoch": 1.8329295928733296,
+      "grad_norm": 0.23192724585533142,
+      "learning_rate": 7.78523489932886e-05,
+      "loss": 0.8679,
+      "step": 733
+    },
+    {
+      "epoch": 1.835430178948191,
+      "grad_norm": 0.2325032353401184,
+      "learning_rate": 7.768456375838927e-05,
+      "loss": 0.8898,
+      "step": 734
+    },
+    {
+      "epoch": 1.8379307650230523,
+      "grad_norm": 0.23566211760044098,
+      "learning_rate": 7.751677852348995e-05,
+      "loss": 0.8269,
+      "step": 735
+    },
+    {
+      "epoch": 1.8404313510979136,
+      "grad_norm": 0.23325851559638977,
+      "learning_rate": 7.734899328859061e-05,
+      "loss": 0.8297,
+      "step": 736
+    },
+    {
+      "epoch": 1.8429319371727748,
+      "grad_norm": 0.242010697722435,
+      "learning_rate": 7.718120805369128e-05,
+      "loss": 0.8461,
+      "step": 737
+    },
+    {
+      "epoch": 1.845432523247636,
+      "grad_norm": 0.23698940873146057,
+      "learning_rate": 7.701342281879194e-05,
+      "loss": 0.8768,
+      "step": 738
+    },
+    {
+      "epoch": 1.8479331093224975,
+      "grad_norm": 0.2337309718132019,
+      "learning_rate": 7.684563758389261e-05,
+      "loss": 0.8345,
+      "step": 739
+    },
+    {
+      "epoch": 1.8504336953973588,
+      "grad_norm": 0.24542047083377838,
+      "learning_rate": 7.667785234899329e-05,
+      "loss": 0.8726,
+      "step": 740
+    },
+    {
+      "epoch": 1.85293428147222,
+      "grad_norm": 0.23877234756946564,
+      "learning_rate": 7.651006711409397e-05,
+      "loss": 0.8408,
+      "step": 741
+    },
+    {
+      "epoch": 1.8554348675470813,
+      "grad_norm": 0.23101742565631866,
+      "learning_rate": 7.634228187919464e-05,
+      "loss": 0.8295,
+      "step": 742
+    },
+    {
+      "epoch": 1.8579354536219426,
+      "grad_norm": 0.2338128685951233,
+      "learning_rate": 7.61744966442953e-05,
+      "loss": 0.858,
+      "step": 743
+    },
+    {
+      "epoch": 1.860436039696804,
+      "grad_norm": 0.22591547667980194,
+      "learning_rate": 7.600671140939598e-05,
+      "loss": 0.8102,
+      "step": 744
+    },
+    {
+      "epoch": 1.862936625771665,
+      "grad_norm": 0.24917657673358917,
+      "learning_rate": 7.583892617449665e-05,
+      "loss": 0.8945,
+      "step": 745
+    },
+    {
+      "epoch": 1.8654372118465266,
+      "grad_norm": 0.226085364818573,
+      "learning_rate": 7.567114093959731e-05,
+      "loss": 0.8706,
+      "step": 746
+    },
+    {
+      "epoch": 1.8679377979213878,
+      "grad_norm": 0.22864364087581635,
+      "learning_rate": 7.550335570469798e-05,
+      "loss": 0.8347,
+      "step": 747
+    },
+    {
+      "epoch": 1.870438383996249,
+      "grad_norm": 0.25771352648735046,
+      "learning_rate": 7.533557046979866e-05,
+      "loss": 0.9215,
+      "step": 748
+    },
+    {
+      "epoch": 1.8729389700711105,
+      "grad_norm": 0.2244727909564972,
+      "learning_rate": 7.516778523489934e-05,
+      "loss": 0.8664,
+      "step": 749
+    },
+    {
+      "epoch": 1.8754395561459716,
+      "grad_norm": 0.23526032269001007,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.8067,
+      "step": 750
+    },
+    {
+      "epoch": 1.877940142220833,
+      "grad_norm": 0.22905413806438446,
+      "learning_rate": 7.483221476510067e-05,
+      "loss": 0.8383,
+      "step": 751
+    },
+    {
+      "epoch": 1.8804407282956943,
+      "grad_norm": 0.23472002148628235,
+      "learning_rate": 7.466442953020134e-05,
+      "loss": 0.8951,
+      "step": 752
+    },
+    {
+      "epoch": 1.8829413143705556,
+      "grad_norm": 0.22684089839458466,
+      "learning_rate": 7.449664429530202e-05,
+      "loss": 0.8137,
+      "step": 753
+    },
+    {
+      "epoch": 1.885441900445417,
+      "grad_norm": 0.21714366972446442,
+      "learning_rate": 7.432885906040269e-05,
+      "loss": 0.8265,
+      "step": 754
+    },
+    {
+      "epoch": 1.887942486520278,
+      "grad_norm": 0.2347375750541687,
+      "learning_rate": 7.416107382550335e-05,
+      "loss": 0.8575,
+      "step": 755
+    },
+    {
+      "epoch": 1.8904430725951396,
+      "grad_norm": 0.22555333375930786,
+      "learning_rate": 7.399328859060403e-05,
+      "loss": 0.8205,
+      "step": 756
+    },
+    {
+      "epoch": 1.8929436586700008,
+      "grad_norm": 0.2268943339586258,
+      "learning_rate": 7.382550335570471e-05,
+      "loss": 0.8221,
+      "step": 757
+    },
+    {
+      "epoch": 1.895444244744862,
+      "grad_norm": 0.23352165520191193,
+      "learning_rate": 7.365771812080538e-05,
+      "loss": 0.8542,
+      "step": 758
+    },
+    {
+      "epoch": 1.8979448308197235,
+      "grad_norm": 0.24461059272289276,
+      "learning_rate": 7.348993288590604e-05,
+      "loss": 0.8635,
+      "step": 759
+    },
+    {
+      "epoch": 1.9004454168945846,
+      "grad_norm": 0.22984404861927032,
+      "learning_rate": 7.332214765100671e-05,
+      "loss": 0.8289,
+      "step": 760
+    },
+    {
+      "epoch": 1.902946002969446,
+      "grad_norm": 0.23233383893966675,
+      "learning_rate": 7.315436241610739e-05,
+      "loss": 0.8862,
+      "step": 761
+    },
+    {
+      "epoch": 1.9054465890443073,
+      "grad_norm": 0.22527334094047546,
+      "learning_rate": 7.298657718120806e-05,
+      "loss": 0.8558,
+      "step": 762
+    },
+    {
+      "epoch": 1.9079471751191686,
+      "grad_norm": 0.23167213797569275,
+      "learning_rate": 7.281879194630872e-05,
+      "loss": 0.8486,
+      "step": 763
+    },
+    {
+      "epoch": 1.9104477611940298,
+      "grad_norm": 0.23401731252670288,
+      "learning_rate": 7.26510067114094e-05,
+      "loss": 0.8078,
+      "step": 764
+    },
+    {
+      "epoch": 1.912948347268891,
+      "grad_norm": 0.23133142292499542,
+      "learning_rate": 7.248322147651007e-05,
+      "loss": 0.8902,
+      "step": 765
+    },
+    {
+      "epoch": 1.9154489333437525,
+      "grad_norm": 0.22788746654987335,
+      "learning_rate": 7.231543624161075e-05,
+      "loss": 0.8506,
+      "step": 766
+    },
+    {
+      "epoch": 1.9179495194186136,
+      "grad_norm": 0.23246771097183228,
+      "learning_rate": 7.214765100671141e-05,
+      "loss": 0.8304,
+      "step": 767
+    },
+    {
+      "epoch": 1.920450105493475,
+      "grad_norm": 0.22091439366340637,
+      "learning_rate": 7.197986577181208e-05,
+      "loss": 0.859,
+      "step": 768
+    },
+    {
+      "epoch": 1.9229506915683363,
+      "grad_norm": 0.22299793362617493,
+      "learning_rate": 7.181208053691275e-05,
+      "loss": 0.8522,
+      "step": 769
+    },
+    {
+      "epoch": 1.9254512776431976,
+      "grad_norm": 0.23084191977977753,
+      "learning_rate": 7.164429530201343e-05,
+      "loss": 0.8025,
+      "step": 770
+    },
+    {
+      "epoch": 1.927951863718059,
+      "grad_norm": 0.23488053679466248,
+      "learning_rate": 7.147651006711409e-05,
+      "loss": 0.8641,
+      "step": 771
+    },
+    {
+      "epoch": 1.93045244979292,
+      "grad_norm": 0.22758269309997559,
+      "learning_rate": 7.130872483221477e-05,
+      "loss": 0.8388,
+      "step": 772
+    },
+    {
+      "epoch": 1.9329530358677816,
+      "grad_norm": 0.2267666459083557,
+      "learning_rate": 7.114093959731544e-05,
+      "loss": 0.8579,
+      "step": 773
+    },
+    {
+      "epoch": 1.9354536219426428,
+      "grad_norm": 0.24212439358234406,
+      "learning_rate": 7.09731543624161e-05,
+      "loss": 0.8483,
+      "step": 774
+    },
+    {
+      "epoch": 1.937954208017504,
+      "grad_norm": 0.2339584231376648,
+      "learning_rate": 7.080536912751679e-05,
+      "loss": 0.9082,
+      "step": 775
+    },
+    {
+      "epoch": 1.9404547940923655,
+      "grad_norm": 0.22963057458400726,
+      "learning_rate": 7.063758389261745e-05,
+      "loss": 0.8596,
+      "step": 776
+    },
+    {
+      "epoch": 1.9429553801672266,
+      "grad_norm": 0.2314060777425766,
+      "learning_rate": 7.046979865771812e-05,
+      "loss": 0.833,
+      "step": 777
+    },
+    {
+      "epoch": 1.945455966242088,
+      "grad_norm": 0.2469668984413147,
+      "learning_rate": 7.03020134228188e-05,
+      "loss": 0.8124,
+      "step": 778
+    },
+    {
+      "epoch": 1.9479565523169493,
+      "grad_norm": 0.241798534989357,
+      "learning_rate": 7.013422818791948e-05,
+      "loss": 0.926,
+      "step": 779
+    },
+    {
+      "epoch": 1.9504571383918106,
+      "grad_norm": 0.223639577627182,
+      "learning_rate": 6.996644295302014e-05,
+      "loss": 0.8311,
+      "step": 780
+    },
+    {
+      "epoch": 1.952957724466672,
+      "grad_norm": 0.24033206701278687,
+      "learning_rate": 6.979865771812081e-05,
+      "loss": 0.8626,
+      "step": 781
+    },
+    {
+      "epoch": 1.955458310541533,
+      "grad_norm": 0.23264126479625702,
+      "learning_rate": 6.963087248322148e-05,
+      "loss": 0.8073,
+      "step": 782
+    },
+    {
+      "epoch": 1.9579588966163946,
+      "grad_norm": 0.23773786425590515,
+      "learning_rate": 6.946308724832216e-05,
+      "loss": 0.8845,
+      "step": 783
+    },
+    {
+      "epoch": 1.9604594826912558,
+      "grad_norm": 0.22041310369968414,
+      "learning_rate": 6.929530201342282e-05,
+      "loss": 0.8257,
+      "step": 784
+    },
+    {
+      "epoch": 1.962960068766117,
+      "grad_norm": 0.23237738013267517,
+      "learning_rate": 6.912751677852349e-05,
+      "loss": 0.8345,
+      "step": 785
+    },
+    {
+      "epoch": 1.9654606548409783,
+      "grad_norm": 0.25310811400413513,
+      "learning_rate": 6.895973154362417e-05,
+      "loss": 0.8662,
+      "step": 786
+    },
+    {
+      "epoch": 1.9679612409158396,
+      "grad_norm": 0.236192524433136,
+      "learning_rate": 6.879194630872483e-05,
+      "loss": 0.835,
+      "step": 787
+    },
+    {
+      "epoch": 1.970461826990701,
+      "grad_norm": 0.23081590235233307,
+      "learning_rate": 6.862416107382551e-05,
+      "loss": 0.7944,
+      "step": 788
+    },
+    {
+      "epoch": 1.972962413065562,
+      "grad_norm": 0.23320381343364716,
+      "learning_rate": 6.845637583892618e-05,
+      "loss": 0.8309,
+      "step": 789
+    },
+    {
+      "epoch": 1.9754629991404236,
+      "grad_norm": 0.23424074053764343,
+      "learning_rate": 6.828859060402685e-05,
+      "loss": 0.7982,
+      "step": 790
+    },
+    {
+      "epoch": 1.9779635852152848,
+      "grad_norm": 0.22726400196552277,
+      "learning_rate": 6.812080536912751e-05,
+      "loss": 0.8307,
+      "step": 791
+    },
+    {
+      "epoch": 1.980464171290146,
+      "grad_norm": 0.2355743944644928,
+      "learning_rate": 6.795302013422819e-05,
+      "loss": 0.8246,
+      "step": 792
+    },
+    {
+      "epoch": 1.9829647573650075,
+      "grad_norm": 0.24635888636112213,
+      "learning_rate": 6.778523489932886e-05,
+      "loss": 0.8803,
+      "step": 793
+    },
+    {
+      "epoch": 1.9854653434398686,
+      "grad_norm": 0.24838051199913025,
+      "learning_rate": 6.761744966442954e-05,
+      "loss": 0.8632,
+      "step": 794
+    },
+    {
+      "epoch": 1.98796592951473,
+      "grad_norm": 0.23996835947036743,
+      "learning_rate": 6.74496644295302e-05,
+      "loss": 0.8426,
+      "step": 795
+    },
+    {
+      "epoch": 1.9904665155895913,
+      "grad_norm": 0.25930771231651306,
+      "learning_rate": 6.728187919463087e-05,
+      "loss": 0.8341,
+      "step": 796
+    },
+    {
+      "epoch": 1.9929671016644526,
+      "grad_norm": 0.23352138698101044,
+      "learning_rate": 6.711409395973155e-05,
+      "loss": 0.8789,
+      "step": 797
+    },
+    {
+      "epoch": 1.995467687739314,
+      "grad_norm": 0.23742055892944336,
+      "learning_rate": 6.694630872483222e-05,
+      "loss": 0.8334,
+      "step": 798
+    },
+    {
+      "epoch": 1.997968273814175,
+      "grad_norm": 0.23012104630470276,
+      "learning_rate": 6.677852348993288e-05,
+      "loss": 0.8301,
+      "step": 799
+    },
+    {
+      "epoch": 2.0004688598890366,
+      "grad_norm": 0.23470798134803772,
+      "learning_rate": 6.661073825503355e-05,
+      "loss": 0.8068,
+      "step": 800
+    },
+    {
+      "epoch": 2.0029694459638976,
+      "grad_norm": 0.24607907235622406,
+      "learning_rate": 6.644295302013423e-05,
+      "loss": 0.7283,
+      "step": 801
+    },
+    {
+      "epoch": 2.005470032038759,
+      "grad_norm": 0.2318810075521469,
+      "learning_rate": 6.627516778523491e-05,
+      "loss": 0.7378,
+      "step": 802
+    },
+    {
+      "epoch": 2.0079706181136205,
+      "grad_norm": 0.2269017994403839,
+      "learning_rate": 6.610738255033558e-05,
+      "loss": 0.8176,
+      "step": 803
+    },
+    {
+      "epoch": 2.0104712041884816,
+      "grad_norm": 0.22210507094860077,
+      "learning_rate": 6.593959731543624e-05,
+      "loss": 0.7855,
+      "step": 804
+    },
+    {
+      "epoch": 2.012971790263343,
+      "grad_norm": 0.24078026413917542,
+      "learning_rate": 6.577181208053692e-05,
+      "loss": 0.7465,
+      "step": 805
+    },
+    {
+      "epoch": 2.015472376338204,
+      "grad_norm": 0.23802946507930756,
+      "learning_rate": 6.560402684563759e-05,
+      "loss": 0.7995,
+      "step": 806
+    },
+    {
+      "epoch": 2.0179729624130656,
+      "grad_norm": 0.2371736466884613,
+      "learning_rate": 6.543624161073825e-05,
+      "loss": 0.7822,
+      "step": 807
+    },
+    {
+      "epoch": 2.020473548487927,
+      "grad_norm": 0.23657207190990448,
+      "learning_rate": 6.526845637583892e-05,
+      "loss": 0.7848,
+      "step": 808
+    },
+    {
+      "epoch": 2.022974134562788,
+      "grad_norm": 0.23179924488067627,
+      "learning_rate": 6.51006711409396e-05,
+      "loss": 0.7733,
+      "step": 809
+    },
+    {
+      "epoch": 2.0254747206376496,
+      "grad_norm": 0.23265567421913147,
+      "learning_rate": 6.493288590604028e-05,
+      "loss": 0.804,
+      "step": 810
+    },
+    {
+      "epoch": 2.0279753067125106,
+      "grad_norm": 0.23541800677776337,
+      "learning_rate": 6.476510067114095e-05,
+      "loss": 0.7794,
+      "step": 811
+    },
+    {
+      "epoch": 2.030475892787372,
+      "grad_norm": 0.22652539610862732,
+      "learning_rate": 6.459731543624161e-05,
+      "loss": 0.8141,
+      "step": 812
+    },
+    {
+      "epoch": 2.0329764788622335,
+      "grad_norm": 0.2315577119588852,
+      "learning_rate": 6.442953020134228e-05,
+      "loss": 0.7734,
+      "step": 813
+    },
+    {
+      "epoch": 2.0354770649370946,
+      "grad_norm": 0.24747934937477112,
+      "learning_rate": 6.426174496644296e-05,
+      "loss": 0.8433,
+      "step": 814
+    },
+    {
+      "epoch": 2.037977651011956,
+      "grad_norm": 0.22663728892803192,
+      "learning_rate": 6.409395973154362e-05,
+      "loss": 0.7918,
+      "step": 815
+    },
+    {
+      "epoch": 2.040478237086817,
+      "grad_norm": 0.23609162867069244,
+      "learning_rate": 6.39261744966443e-05,
+      "loss": 0.799,
+      "step": 816
+    },
+    {
+      "epoch": 2.0429788231616786,
+      "grad_norm": 0.228263720870018,
+      "learning_rate": 6.375838926174497e-05,
+      "loss": 0.7819,
+      "step": 817
+    },
+    {
+      "epoch": 2.04547940923654,
+      "grad_norm": 0.22902126610279083,
+      "learning_rate": 6.359060402684565e-05,
+      "loss": 0.815,
+      "step": 818
+    },
+    {
+      "epoch": 2.047979995311401,
+      "grad_norm": 0.2405027449131012,
+      "learning_rate": 6.342281879194632e-05,
+      "loss": 0.8518,
+      "step": 819
+    },
+    {
+      "epoch": 2.0504805813862625,
+      "grad_norm": 0.23810040950775146,
+      "learning_rate": 6.325503355704698e-05,
+      "loss": 0.7877,
+      "step": 820
+    },
+    {
+      "epoch": 2.0529811674611236,
+      "grad_norm": 0.2406715303659439,
+      "learning_rate": 6.308724832214765e-05,
+      "loss": 0.8054,
+      "step": 821
+    },
+    {
+      "epoch": 2.055481753535985,
+      "grad_norm": 0.23643344640731812,
+      "learning_rate": 6.291946308724832e-05,
+      "loss": 0.7866,
+      "step": 822
+    },
+    {
+      "epoch": 2.057982339610846,
+      "grad_norm": 0.23765967786312103,
+      "learning_rate": 6.2751677852349e-05,
+      "loss": 0.7552,
+      "step": 823
+    },
+    {
+      "epoch": 2.0604829256857076,
+      "grad_norm": 0.2526334822177887,
+      "learning_rate": 6.258389261744967e-05,
+      "loss": 0.8114,
+      "step": 824
+    },
+    {
+      "epoch": 2.062983511760569,
+      "grad_norm": 0.24000713229179382,
+      "learning_rate": 6.241610738255034e-05,
+      "loss": 0.8218,
+      "step": 825
+    },
+    {
+      "epoch": 2.06548409783543,
+      "grad_norm": 0.25650936365127563,
+      "learning_rate": 6.224832214765101e-05,
+      "loss": 0.716,
+      "step": 826
+    },
+    {
+      "epoch": 2.0679846839102916,
+      "grad_norm": 0.2417072206735611,
+      "learning_rate": 6.208053691275169e-05,
+      "loss": 0.7678,
+      "step": 827
+    },
+    {
+      "epoch": 2.0704852699851526,
+      "grad_norm": 0.24069854617118835,
+      "learning_rate": 6.191275167785235e-05,
+      "loss": 0.7615,
+      "step": 828
+    },
+    {
+      "epoch": 2.072985856060014,
+      "grad_norm": 0.24259264767169952,
+      "learning_rate": 6.174496644295302e-05,
+      "loss": 0.7609,
+      "step": 829
+    },
+    {
+      "epoch": 2.0754864421348755,
+      "grad_norm": 0.23391026258468628,
+      "learning_rate": 6.157718120805369e-05,
+      "loss": 0.7833,
+      "step": 830
+    },
+    {
+      "epoch": 2.0779870282097366,
+      "grad_norm": 0.24214905500411987,
+      "learning_rate": 6.140939597315437e-05,
+      "loss": 0.8001,
+      "step": 831
+    },
+    {
+      "epoch": 2.080487614284598,
+      "grad_norm": 0.2601361572742462,
+      "learning_rate": 6.124161073825505e-05,
+      "loss": 0.857,
+      "step": 832
+    },
+    {
+      "epoch": 2.082988200359459,
+      "grad_norm": 0.2372775375843048,
+      "learning_rate": 6.107382550335571e-05,
+      "loss": 0.7888,
+      "step": 833
+    },
+    {
+      "epoch": 2.0854887864343206,
+      "grad_norm": 0.2427842766046524,
+      "learning_rate": 6.090604026845638e-05,
+      "loss": 0.8059,
+      "step": 834
+    },
+    {
+      "epoch": 2.087989372509182,
+      "grad_norm": 0.24764439463615417,
+      "learning_rate": 6.0738255033557044e-05,
+      "loss": 0.8016,
+      "step": 835
+    },
+    {
+      "epoch": 2.090489958584043,
+      "grad_norm": 0.2483157515525818,
+      "learning_rate": 6.0570469798657724e-05,
+      "loss": 0.8065,
+      "step": 836
+    },
+    {
+      "epoch": 2.0929905446589046,
+      "grad_norm": 0.2539243996143341,
+      "learning_rate": 6.04026845637584e-05,
+      "loss": 0.8485,
+      "step": 837
+    },
+    {
+      "epoch": 2.0954911307337656,
+      "grad_norm": 0.24617519974708557,
+      "learning_rate": 6.023489932885906e-05,
+      "loss": 0.8298,
+      "step": 838
+    },
+    {
+      "epoch": 2.097991716808627,
+      "grad_norm": 0.24019426107406616,
+      "learning_rate": 6.006711409395973e-05,
+      "loss": 0.7624,
+      "step": 839
+    },
+    {
+      "epoch": 2.100492302883488,
+      "grad_norm": 0.24285243451595306,
+      "learning_rate": 5.989932885906041e-05,
+      "loss": 0.7847,
+      "step": 840
+    },
+    {
+      "epoch": 2.1029928889583496,
+      "grad_norm": 0.2526892125606537,
+      "learning_rate": 5.973154362416108e-05,
+      "loss": 0.7552,
+      "step": 841
+    },
+    {
+      "epoch": 2.105493475033211,
+      "grad_norm": 0.2470400482416153,
+      "learning_rate": 5.956375838926175e-05,
+      "loss": 0.8403,
+      "step": 842
+    },
+    {
+      "epoch": 2.107994061108072,
+      "grad_norm": 0.2545977830886841,
+      "learning_rate": 5.9395973154362415e-05,
+      "loss": 0.8149,
+      "step": 843
+    },
+    {
+      "epoch": 2.1104946471829336,
+      "grad_norm": 0.2411421686410904,
+      "learning_rate": 5.922818791946309e-05,
+      "loss": 0.8499,
+      "step": 844
+    },
+    {
+      "epoch": 2.1129952332577946,
+      "grad_norm": 0.2427968829870224,
+      "learning_rate": 5.906040268456377e-05,
+      "loss": 0.7372,
+      "step": 845
+    },
+    {
+      "epoch": 2.115495819332656,
+      "grad_norm": 0.23140354454517365,
+      "learning_rate": 5.8892617449664434e-05,
+      "loss": 0.74,
+      "step": 846
+    },
+    {
+      "epoch": 2.1179964054075175,
+      "grad_norm": 0.24942906200885773,
+      "learning_rate": 5.87248322147651e-05,
+      "loss": 0.7843,
+      "step": 847
+    },
+    {
+      "epoch": 2.1204969914823786,
+      "grad_norm": 0.24845479428768158,
+      "learning_rate": 5.855704697986577e-05,
+      "loss": 0.8706,
+      "step": 848
+    },
+    {
+      "epoch": 2.12299757755724,
+      "grad_norm": 0.24324162304401398,
+      "learning_rate": 5.838926174496645e-05,
+      "loss": 0.7282,
+      "step": 849
+    },
+    {
+      "epoch": 2.125498163632101,
+      "grad_norm": 0.2480383813381195,
+      "learning_rate": 5.822147651006712e-05,
+      "loss": 0.7658,
+      "step": 850
+    },
+    {
+      "epoch": 2.1279987497069626,
+      "grad_norm": 0.24291519820690155,
+      "learning_rate": 5.8053691275167786e-05,
+      "loss": 0.7468,
+      "step": 851
+    },
+    {
+      "epoch": 2.130499335781824,
+      "grad_norm": 0.2464919537305832,
+      "learning_rate": 5.788590604026846e-05,
+      "loss": 0.8322,
+      "step": 852
+    },
+    {
+      "epoch": 2.132999921856685,
+      "grad_norm": 0.25378695130348206,
+      "learning_rate": 5.771812080536914e-05,
+      "loss": 0.7854,
+      "step": 853
+    },
+    {
+      "epoch": 2.1355005079315466,
+      "grad_norm": 0.2418270856142044,
+      "learning_rate": 5.7550335570469805e-05,
+      "loss": 0.82,
+      "step": 854
+    },
+    {
+      "epoch": 2.1380010940064076,
+      "grad_norm": 0.24473296105861664,
+      "learning_rate": 5.738255033557047e-05,
+      "loss": 0.7671,
+      "step": 855
+    },
+    {
+      "epoch": 2.140501680081269,
+      "grad_norm": 0.2462645024061203,
+      "learning_rate": 5.7214765100671144e-05,
+      "loss": 0.7632,
+      "step": 856
+    },
+    {
+      "epoch": 2.1430022661561305,
+      "grad_norm": 0.2552192211151123,
+      "learning_rate": 5.704697986577181e-05,
+      "loss": 0.7534,
+      "step": 857
+    },
+    {
+      "epoch": 2.1455028522309916,
+      "grad_norm": 0.2530480623245239,
+      "learning_rate": 5.687919463087249e-05,
+      "loss": 0.7887,
+      "step": 858
+    },
+    {
+      "epoch": 2.148003438305853,
+      "grad_norm": 0.24546147882938385,
+      "learning_rate": 5.6711409395973156e-05,
+      "loss": 0.7508,
+      "step": 859
+    },
+    {
+      "epoch": 2.150504024380714,
+      "grad_norm": 0.2479485124349594,
+      "learning_rate": 5.654362416107383e-05,
+      "loss": 0.7557,
+      "step": 860
+    },
+    {
+      "epoch": 2.1530046104555756,
+      "grad_norm": 0.25150132179260254,
+      "learning_rate": 5.6375838926174495e-05,
+      "loss": 0.7664,
+      "step": 861
+    },
+    {
+      "epoch": 2.155505196530437,
+      "grad_norm": 0.2424343079328537,
+      "learning_rate": 5.6208053691275175e-05,
+      "loss": 0.758,
+      "step": 862
+    },
+    {
+      "epoch": 2.158005782605298,
+      "grad_norm": 0.2511323094367981,
+      "learning_rate": 5.604026845637584e-05,
+      "loss": 0.7794,
+      "step": 863
+    },
+    {
+      "epoch": 2.1605063686801596,
+      "grad_norm": 0.2579136788845062,
+      "learning_rate": 5.5872483221476514e-05,
+      "loss": 0.7752,
+      "step": 864
+    },
+    {
+      "epoch": 2.1630069547550206,
+      "grad_norm": 0.24404679238796234,
+      "learning_rate": 5.570469798657718e-05,
+      "loss": 0.7855,
+      "step": 865
+    },
+    {
+      "epoch": 2.165507540829882,
+      "grad_norm": 0.2586469352245331,
+      "learning_rate": 5.553691275167785e-05,
+      "loss": 0.7756,
+      "step": 866
+    },
+    {
+      "epoch": 2.168008126904743,
+      "grad_norm": 0.25084635615348816,
+      "learning_rate": 5.536912751677853e-05,
+      "loss": 0.7624,
+      "step": 867
+    },
+    {
+      "epoch": 2.1705087129796046,
+      "grad_norm": 0.2515811026096344,
+      "learning_rate": 5.52013422818792e-05,
+      "loss": 0.8047,
+      "step": 868
+    },
+    {
+      "epoch": 2.173009299054466,
+      "grad_norm": 0.24664728343486786,
+      "learning_rate": 5.5033557046979866e-05,
+      "loss": 0.7461,
+      "step": 869
+    },
+    {
+      "epoch": 2.175509885129327,
+      "grad_norm": 0.2585003972053528,
+      "learning_rate": 5.486577181208053e-05,
+      "loss": 0.7517,
+      "step": 870
+    },
+    {
+      "epoch": 2.1780104712041886,
+      "grad_norm": 0.24943596124649048,
+      "learning_rate": 5.469798657718121e-05,
+      "loss": 0.8033,
+      "step": 871
+    },
+    {
+      "epoch": 2.1805110572790496,
+      "grad_norm": 0.24782632291316986,
+      "learning_rate": 5.4530201342281885e-05,
+      "loss": 0.763,
+      "step": 872
+    },
+    {
+      "epoch": 2.183011643353911,
+      "grad_norm": 0.26548439264297485,
+      "learning_rate": 5.436241610738255e-05,
+      "loss": 0.8328,
+      "step": 873
+    },
+    {
+      "epoch": 2.1855122294287725,
+      "grad_norm": 0.2599753141403198,
+      "learning_rate": 5.419463087248322e-05,
+      "loss": 0.7684,
+      "step": 874
+    },
+    {
+      "epoch": 2.1880128155036336,
+      "grad_norm": 0.2500552237033844,
+      "learning_rate": 5.40268456375839e-05,
+      "loss": 0.7289,
+      "step": 875
+    },
+    {
+      "epoch": 2.190513401578495,
+      "grad_norm": 0.25470981001853943,
+      "learning_rate": 5.385906040268457e-05,
+      "loss": 0.7501,
+      "step": 876
+    },
+    {
+      "epoch": 2.193013987653356,
+      "grad_norm": 0.24483001232147217,
+      "learning_rate": 5.3691275167785237e-05,
+      "loss": 0.7744,
+      "step": 877
+    },
+    {
+      "epoch": 2.1955145737282176,
+      "grad_norm": 0.24729371070861816,
+      "learning_rate": 5.352348993288591e-05,
+      "loss": 0.8136,
+      "step": 878
+    },
+    {
+      "epoch": 2.198015159803079,
+      "grad_norm": 0.25430426001548767,
+      "learning_rate": 5.3355704697986576e-05,
+      "loss": 0.8182,
+      "step": 879
+    },
+    {
+      "epoch": 2.20051574587794,
+      "grad_norm": 0.2573440670967102,
+      "learning_rate": 5.3187919463087256e-05,
+      "loss": 0.8044,
+      "step": 880
+    },
+    {
+      "epoch": 2.2030163319528016,
+      "grad_norm": 0.24993404746055603,
+      "learning_rate": 5.302013422818792e-05,
+      "loss": 0.7518,
+      "step": 881
+    },
+    {
+      "epoch": 2.2055169180276626,
+      "grad_norm": 0.26014912128448486,
+      "learning_rate": 5.2852348993288595e-05,
+      "loss": 0.7879,
+      "step": 882
+    },
+    {
+      "epoch": 2.208017504102524,
+      "grad_norm": 0.26472708582878113,
+      "learning_rate": 5.268456375838926e-05,
+      "loss": 0.7472,
+      "step": 883
+    },
+    {
+      "epoch": 2.210518090177385,
+      "grad_norm": 0.2627474367618561,
+      "learning_rate": 5.251677852348994e-05,
+      "loss": 0.7461,
+      "step": 884
+    },
+    {
+      "epoch": 2.2130186762522466,
+      "grad_norm": 0.23871855437755585,
+      "learning_rate": 5.234899328859061e-05,
+      "loss": 0.7501,
+      "step": 885
+    },
+    {
+      "epoch": 2.215519262327108,
+      "grad_norm": 0.2663973569869995,
+      "learning_rate": 5.218120805369128e-05,
+      "loss": 0.7197,
+      "step": 886
+    },
+    {
+      "epoch": 2.218019848401969,
+      "grad_norm": 0.2534399628639221,
+      "learning_rate": 5.2013422818791946e-05,
+      "loss": 0.8195,
+      "step": 887
+    },
+    {
+      "epoch": 2.2205204344768306,
+      "grad_norm": 0.2622665762901306,
+      "learning_rate": 5.184563758389261e-05,
+      "loss": 0.7615,
+      "step": 888
+    },
+    {
+      "epoch": 2.2230210205516916,
+      "grad_norm": 0.26300376653671265,
+      "learning_rate": 5.167785234899329e-05,
+      "loss": 0.8142,
+      "step": 889
+    },
+    {
+      "epoch": 2.225521606626553,
+      "grad_norm": 0.25385987758636475,
+      "learning_rate": 5.1510067114093966e-05,
+      "loss": 0.7643,
+      "step": 890
+    },
+    {
+      "epoch": 2.2280221927014145,
+      "grad_norm": 0.2716410458087921,
+      "learning_rate": 5.134228187919463e-05,
+      "loss": 0.8024,
+      "step": 891
+    },
+    {
+      "epoch": 2.2305227787762756,
+      "grad_norm": 0.2592583894729614,
+      "learning_rate": 5.11744966442953e-05,
+      "loss": 0.7872,
+      "step": 892
+    },
+    {
+      "epoch": 2.233023364851137,
+      "grad_norm": 0.24683411419391632,
+      "learning_rate": 5.100671140939598e-05,
+      "loss": 0.81,
+      "step": 893
+    },
+    {
+      "epoch": 2.235523950925998,
+      "grad_norm": 0.25091201066970825,
+      "learning_rate": 5.083892617449665e-05,
+      "loss": 0.766,
+      "step": 894
+    },
+    {
+      "epoch": 2.2380245370008596,
+      "grad_norm": 0.25476816296577454,
+      "learning_rate": 5.067114093959732e-05,
+      "loss": 0.7607,
+      "step": 895
+    },
+    {
+      "epoch": 2.240525123075721,
+      "grad_norm": 0.24960653483867645,
+      "learning_rate": 5.050335570469798e-05,
+      "loss": 0.8004,
+      "step": 896
+    },
+    {
+      "epoch": 2.243025709150582,
+      "grad_norm": 0.2472076714038849,
+      "learning_rate": 5.033557046979866e-05,
+      "loss": 0.7665,
+      "step": 897
+    },
+    {
+      "epoch": 2.2455262952254436,
+      "grad_norm": 0.2540365159511566,
+      "learning_rate": 5.0167785234899336e-05,
+      "loss": 0.8094,
+      "step": 898
+    },
+    {
+      "epoch": 2.2480268813003046,
+      "grad_norm": 0.2585788071155548,
+      "learning_rate": 5e-05,
+      "loss": 0.7663,
+      "step": 899
+    },
+    {
+      "epoch": 2.250527467375166,
+      "grad_norm": 0.2619767487049103,
+      "learning_rate": 4.983221476510067e-05,
+      "loss": 0.8323,
+      "step": 900
+    },
+    {
+      "epoch": 2.253028053450027,
+      "grad_norm": 0.26195207238197327,
+      "learning_rate": 4.966442953020135e-05,
+      "loss": 0.7951,
+      "step": 901
+    },
+    {
+      "epoch": 2.2555286395248886,
+      "grad_norm": 0.2557896673679352,
+      "learning_rate": 4.9496644295302015e-05,
+      "loss": 0.8498,
+      "step": 902
+    },
+    {
+      "epoch": 2.25802922559975,
+      "grad_norm": 0.24389132857322693,
+      "learning_rate": 4.932885906040269e-05,
+      "loss": 0.8198,
+      "step": 903
+    },
+    {
+      "epoch": 2.260529811674611,
+      "grad_norm": 0.25792619585990906,
+      "learning_rate": 4.9161073825503354e-05,
+      "loss": 0.8333,
+      "step": 904
+    },
+    {
+      "epoch": 2.2630303977494726,
+      "grad_norm": 0.26080524921417236,
+      "learning_rate": 4.8993288590604034e-05,
+      "loss": 0.8061,
+      "step": 905
+    },
+    {
+      "epoch": 2.265530983824334,
+      "grad_norm": 0.26113346219062805,
+      "learning_rate": 4.88255033557047e-05,
+      "loss": 0.7772,
+      "step": 906
+    },
+    {
+      "epoch": 2.268031569899195,
+      "grad_norm": 0.2677919268608093,
+      "learning_rate": 4.865771812080537e-05,
+      "loss": 0.7154,
+      "step": 907
+    },
+    {
+      "epoch": 2.2705321559740566,
+      "grad_norm": 0.2447947859764099,
+      "learning_rate": 4.848993288590604e-05,
+      "loss": 0.7082,
+      "step": 908
+    },
+    {
+      "epoch": 2.2730327420489176,
+      "grad_norm": 0.2506728768348694,
+      "learning_rate": 4.832214765100672e-05,
+      "loss": 0.7783,
+      "step": 909
+    },
+    {
+      "epoch": 2.275533328123779,
+      "grad_norm": 0.2573874294757843,
+      "learning_rate": 4.8154362416107385e-05,
+      "loss": 0.8102,
+      "step": 910
+    },
+    {
+      "epoch": 2.27803391419864,
+      "grad_norm": 0.26858681440353394,
+      "learning_rate": 4.798657718120805e-05,
+      "loss": 0.8285,
+      "step": 911
+    },
+    {
+      "epoch": 2.2805345002735016,
+      "grad_norm": 0.26032930612564087,
+      "learning_rate": 4.7818791946308725e-05,
+      "loss": 0.8049,
+      "step": 912
+    },
+    {
+      "epoch": 2.283035086348363,
+      "grad_norm": 0.2530040442943573,
+      "learning_rate": 4.76510067114094e-05,
+      "loss": 0.7438,
+      "step": 913
+    },
+    {
+      "epoch": 2.285535672423224,
+      "grad_norm": 0.25307929515838623,
+      "learning_rate": 4.748322147651007e-05,
+      "loss": 0.756,
+      "step": 914
+    },
+    {
+      "epoch": 2.2880362584980856,
+      "grad_norm": 0.2576766610145569,
+      "learning_rate": 4.731543624161074e-05,
+      "loss": 0.7778,
+      "step": 915
+    },
+    {
+      "epoch": 2.2905368445729466,
+      "grad_norm": 0.2508293390274048,
+      "learning_rate": 4.714765100671141e-05,
+      "loss": 0.7159,
+      "step": 916
+    },
+    {
+      "epoch": 2.293037430647808,
+      "grad_norm": 0.26381373405456543,
+      "learning_rate": 4.697986577181208e-05,
+      "loss": 0.787,
+      "step": 917
+    },
+    {
+      "epoch": 2.2955380167226695,
+      "grad_norm": 0.25328952074050903,
+      "learning_rate": 4.6812080536912756e-05,
+      "loss": 0.8,
+      "step": 918
+    },
+    {
+      "epoch": 2.2980386027975306,
+      "grad_norm": 0.25167882442474365,
+      "learning_rate": 4.664429530201342e-05,
+      "loss": 0.7657,
+      "step": 919
+    },
+    {
+      "epoch": 2.300539188872392,
+      "grad_norm": 0.26473554968833923,
+      "learning_rate": 4.6476510067114095e-05,
+      "loss": 0.7448,
+      "step": 920
+    },
+    {
+      "epoch": 2.303039774947253,
+      "grad_norm": 0.2561410963535309,
+      "learning_rate": 4.630872483221477e-05,
+      "loss": 0.7943,
+      "step": 921
+    },
+    {
+      "epoch": 2.3055403610221146,
+      "grad_norm": 0.27880439162254333,
+      "learning_rate": 4.6140939597315434e-05,
+      "loss": 0.7945,
+      "step": 922
+    },
+    {
+      "epoch": 2.308040947096976,
+      "grad_norm": 0.25818249583244324,
+      "learning_rate": 4.597315436241611e-05,
+      "loss": 0.7956,
+      "step": 923
+    },
+    {
+      "epoch": 2.310541533171837,
+      "grad_norm": 0.2587163746356964,
+      "learning_rate": 4.580536912751678e-05,
+      "loss": 0.7975,
+      "step": 924
+    },
+    {
+      "epoch": 2.3130421192466986,
+      "grad_norm": 0.25379982590675354,
+      "learning_rate": 4.5637583892617453e-05,
+      "loss": 0.8047,
+      "step": 925
+    },
+    {
+      "epoch": 2.3155427053215596,
+      "grad_norm": 0.2601042687892914,
+      "learning_rate": 4.546979865771812e-05,
+      "loss": 0.7841,
+      "step": 926
+    },
+    {
+      "epoch": 2.318043291396421,
+      "grad_norm": 0.25389307737350464,
+      "learning_rate": 4.530201342281879e-05,
+      "loss": 0.801,
+      "step": 927
+    },
+    {
+      "epoch": 2.320543877471282,
+      "grad_norm": 0.26475653052330017,
+      "learning_rate": 4.5134228187919466e-05,
+      "loss": 0.7941,
+      "step": 928
+    },
+    {
+      "epoch": 2.3230444635461436,
+      "grad_norm": 0.2604124844074249,
+      "learning_rate": 4.496644295302014e-05,
+      "loss": 0.7723,
+      "step": 929
+    },
+    {
+      "epoch": 2.325545049621005,
+      "grad_norm": 0.2608570158481598,
+      "learning_rate": 4.4798657718120805e-05,
+      "loss": 0.7831,
+      "step": 930
+    },
+    {
+      "epoch": 2.328045635695866,
+      "grad_norm": 0.24734510481357574,
+      "learning_rate": 4.463087248322148e-05,
+      "loss": 0.7536,
+      "step": 931
+    },
+    {
+      "epoch": 2.3305462217707276,
+      "grad_norm": 0.24852076172828674,
+      "learning_rate": 4.446308724832215e-05,
+      "loss": 0.8115,
+      "step": 932
+    },
+    {
+      "epoch": 2.333046807845589,
+      "grad_norm": 0.253268837928772,
+      "learning_rate": 4.4295302013422824e-05,
+      "loss": 0.7981,
+      "step": 933
+    },
+    {
+      "epoch": 2.33554739392045,
+      "grad_norm": 0.2487640231847763,
+      "learning_rate": 4.412751677852349e-05,
+      "loss": 0.7818,
+      "step": 934
+    },
+    {
+      "epoch": 2.3380479799953116,
+      "grad_norm": 0.2593766450881958,
+      "learning_rate": 4.395973154362416e-05,
+      "loss": 0.7527,
+      "step": 935
+    },
+    {
+      "epoch": 2.3405485660701726,
+      "grad_norm": 0.2596462666988373,
+      "learning_rate": 4.3791946308724836e-05,
+      "loss": 0.784,
+      "step": 936
+    },
+    {
+      "epoch": 2.343049152145034,
+      "grad_norm": 0.2679142653942108,
+      "learning_rate": 4.36241610738255e-05,
+      "loss": 0.8091,
+      "step": 937
+    },
+    {
+      "epoch": 2.345549738219895,
+      "grad_norm": 0.2611890137195587,
+      "learning_rate": 4.3456375838926176e-05,
+      "loss": 0.7759,
+      "step": 938
+    },
+    {
+      "epoch": 2.3480503242947566,
+      "grad_norm": 0.2712072432041168,
+      "learning_rate": 4.328859060402685e-05,
+      "loss": 0.772,
+      "step": 939
+    },
+    {
+      "epoch": 2.350550910369618,
+      "grad_norm": 0.2659551501274109,
+      "learning_rate": 4.312080536912752e-05,
+      "loss": 0.7155,
+      "step": 940
+    },
+    {
+      "epoch": 2.353051496444479,
+      "grad_norm": 0.26499849557876587,
+      "learning_rate": 4.295302013422819e-05,
+      "loss": 0.761,
+      "step": 941
+    },
+    {
+      "epoch": 2.3555520825193406,
+      "grad_norm": 0.26622283458709717,
+      "learning_rate": 4.278523489932886e-05,
+      "loss": 0.7402,
+      "step": 942
+    },
+    {
+      "epoch": 2.3580526685942016,
+      "grad_norm": 0.2556600570678711,
+      "learning_rate": 4.2617449664429534e-05,
+      "loss": 0.8002,
+      "step": 943
+    },
+    {
+      "epoch": 2.360553254669063,
+      "grad_norm": 0.25517985224723816,
+      "learning_rate": 4.244966442953021e-05,
+      "loss": 0.7729,
+      "step": 944
+    },
+    {
+      "epoch": 2.3630538407439245,
+      "grad_norm": 0.2658151388168335,
+      "learning_rate": 4.228187919463087e-05,
+      "loss": 0.7753,
+      "step": 945
+    },
+    {
+      "epoch": 2.3655544268187856,
+      "grad_norm": 0.25279146432876587,
+      "learning_rate": 4.2114093959731546e-05,
+      "loss": 0.7395,
+      "step": 946
+    },
+    {
+      "epoch": 2.368055012893647,
+      "grad_norm": 0.26842814683914185,
+      "learning_rate": 4.194630872483222e-05,
+      "loss": 0.7578,
+      "step": 947
+    },
+    {
+      "epoch": 2.370555598968508,
+      "grad_norm": 0.2501954138278961,
+      "learning_rate": 4.1778523489932886e-05,
+      "loss": 0.7525,
+      "step": 948
+    },
+    {
+      "epoch": 2.3730561850433696,
+      "grad_norm": 0.26409366726875305,
+      "learning_rate": 4.161073825503356e-05,
+      "loss": 0.7955,
+      "step": 949
+    },
+    {
+      "epoch": 2.375556771118231,
+      "grad_norm": 0.25643742084503174,
+      "learning_rate": 4.144295302013423e-05,
+      "loss": 0.8443,
+      "step": 950
+    },
+    {
+      "epoch": 2.378057357193092,
+      "grad_norm": 0.25786539912223816,
+      "learning_rate": 4.1275167785234905e-05,
+      "loss": 0.7609,
+      "step": 951
+    },
+    {
+      "epoch": 2.3805579432679536,
+      "grad_norm": 0.2559066414833069,
+      "learning_rate": 4.110738255033557e-05,
+      "loss": 0.7591,
+      "step": 952
+    },
+    {
+      "epoch": 2.3830585293428146,
+      "grad_norm": 0.24959532916545868,
+      "learning_rate": 4.0939597315436244e-05,
+      "loss": 0.731,
+      "step": 953
+    },
+    {
+      "epoch": 2.385559115417676,
+      "grad_norm": 0.24601855874061584,
+      "learning_rate": 4.077181208053692e-05,
+      "loss": 0.7295,
+      "step": 954
+    },
+    {
+      "epoch": 2.388059701492537,
+      "grad_norm": 0.2617453932762146,
+      "learning_rate": 4.060402684563759e-05,
+      "loss": 0.8341,
+      "step": 955
+    },
+    {
+      "epoch": 2.3905602875673986,
+      "grad_norm": 0.25553691387176514,
+      "learning_rate": 4.0436241610738256e-05,
+      "loss": 0.8194,
+      "step": 956
+    },
+    {
+      "epoch": 2.39306087364226,
+      "grad_norm": 0.24928820133209229,
+      "learning_rate": 4.026845637583892e-05,
+      "loss": 0.7825,
+      "step": 957
+    },
+    {
+      "epoch": 2.395561459717121,
+      "grad_norm": 0.2611832916736603,
+      "learning_rate": 4.01006711409396e-05,
+      "loss": 0.7878,
+      "step": 958
+    },
+    {
+      "epoch": 2.3980620457919826,
+      "grad_norm": 0.2605842649936676,
+      "learning_rate": 3.993288590604027e-05,
+      "loss": 0.831,
+      "step": 959
+    },
+    {
+      "epoch": 2.4005626318668436,
+      "grad_norm": 0.257687509059906,
+      "learning_rate": 3.976510067114094e-05,
+      "loss": 0.7808,
+      "step": 960
+    },
+    {
+      "epoch": 2.403063217941705,
+      "grad_norm": 0.2710731327533722,
+      "learning_rate": 3.959731543624161e-05,
+      "loss": 0.7869,
+      "step": 961
+    },
+    {
+      "epoch": 2.4055638040165666,
+      "grad_norm": 0.25934329628944397,
+      "learning_rate": 3.942953020134229e-05,
+      "loss": 0.7863,
+      "step": 962
+    },
+    {
+      "epoch": 2.4080643900914276,
+      "grad_norm": 0.26316648721694946,
+      "learning_rate": 3.9261744966442954e-05,
+      "loss": 0.7821,
+      "step": 963
+    },
+    {
+      "epoch": 2.410564976166289,
+      "grad_norm": 0.2509710192680359,
+      "learning_rate": 3.909395973154363e-05,
+      "loss": 0.7608,
+      "step": 964
+    },
+    {
+      "epoch": 2.41306556224115,
+      "grad_norm": 0.2608097195625305,
+      "learning_rate": 3.89261744966443e-05,
+      "loss": 0.7678,
+      "step": 965
+    },
+    {
+      "epoch": 2.4155661483160116,
+      "grad_norm": 0.26854583621025085,
+      "learning_rate": 3.875838926174497e-05,
+      "loss": 0.8275,
+      "step": 966
+    },
+    {
+      "epoch": 2.418066734390873,
+      "grad_norm": 0.2555355429649353,
+      "learning_rate": 3.859060402684564e-05,
+      "loss": 0.8006,
+      "step": 967
+    },
+    {
+      "epoch": 2.420567320465734,
+      "grad_norm": 0.2654256224632263,
+      "learning_rate": 3.8422818791946305e-05,
+      "loss": 0.7844,
+      "step": 968
+    },
+    {
+      "epoch": 2.4230679065405956,
+      "grad_norm": 0.2471611350774765,
+      "learning_rate": 3.8255033557046985e-05,
+      "loss": 0.7873,
+      "step": 969
+    },
+    {
+      "epoch": 2.4255684926154566,
+      "grad_norm": 0.26298415660858154,
+      "learning_rate": 3.808724832214765e-05,
+      "loss": 0.755,
+      "step": 970
+    },
+    {
+      "epoch": 2.428069078690318,
+      "grad_norm": 0.2553905248641968,
+      "learning_rate": 3.7919463087248324e-05,
+      "loss": 0.7782,
+      "step": 971
+    },
+    {
+      "epoch": 2.430569664765179,
+      "grad_norm": 0.2564135789871216,
+      "learning_rate": 3.775167785234899e-05,
+      "loss": 0.803,
+      "step": 972
+    },
+    {
+      "epoch": 2.4330702508400406,
+      "grad_norm": 0.26329025626182556,
+      "learning_rate": 3.758389261744967e-05,
+      "loss": 0.836,
+      "step": 973
+    },
+    {
+      "epoch": 2.435570836914902,
+      "grad_norm": 0.2499067187309265,
+      "learning_rate": 3.741610738255034e-05,
+      "loss": 0.7808,
+      "step": 974
+    },
+    {
+      "epoch": 2.438071422989763,
+      "grad_norm": 0.26111480593681335,
+      "learning_rate": 3.724832214765101e-05,
+      "loss": 0.7939,
+      "step": 975
+    },
+    {
+      "epoch": 2.4405720090646246,
+      "grad_norm": 0.26202917098999023,
+      "learning_rate": 3.7080536912751676e-05,
+      "loss": 0.7997,
+      "step": 976
+    },
+    {
+      "epoch": 2.443072595139486,
+      "grad_norm": 0.257347047328949,
+      "learning_rate": 3.6912751677852356e-05,
+      "loss": 0.8006,
+      "step": 977
+    },
+    {
+      "epoch": 2.445573181214347,
+      "grad_norm": 0.2559988796710968,
+      "learning_rate": 3.674496644295302e-05,
+      "loss": 0.7818,
+      "step": 978
+    },
+    {
+      "epoch": 2.4480737672892086,
+      "grad_norm": 0.25502046942710876,
+      "learning_rate": 3.6577181208053695e-05,
+      "loss": 0.7671,
+      "step": 979
+    },
+    {
+      "epoch": 2.4505743533640696,
+      "grad_norm": 0.26668861508369446,
+      "learning_rate": 3.640939597315436e-05,
+      "loss": 0.7798,
+      "step": 980
+    },
+    {
+      "epoch": 2.453074939438931,
+      "grad_norm": 0.25872403383255005,
+      "learning_rate": 3.6241610738255034e-05,
+      "loss": 0.7761,
+      "step": 981
+    },
+    {
+      "epoch": 2.455575525513792,
+      "grad_norm": 0.26160773634910583,
+      "learning_rate": 3.607382550335571e-05,
+      "loss": 0.7881,
+      "step": 982
+    },
+    {
+      "epoch": 2.4580761115886536,
+      "grad_norm": 0.2661009132862091,
+      "learning_rate": 3.5906040268456373e-05,
+      "loss": 0.8239,
+      "step": 983
+    },
+    {
+      "epoch": 2.460576697663515,
+      "grad_norm": 0.2684977650642395,
+      "learning_rate": 3.5738255033557046e-05,
+      "loss": 0.7761,
+      "step": 984
+    },
+    {
+      "epoch": 2.463077283738376,
+      "grad_norm": 0.25709253549575806,
+      "learning_rate": 3.557046979865772e-05,
+      "loss": 0.7922,
+      "step": 985
+    },
+    {
+      "epoch": 2.4655778698132376,
+      "grad_norm": 0.2595881521701813,
+      "learning_rate": 3.540268456375839e-05,
+      "loss": 0.8389,
+      "step": 986
+    },
+    {
+      "epoch": 2.4680784558880986,
+      "grad_norm": 0.27629536390304565,
+      "learning_rate": 3.523489932885906e-05,
+      "loss": 0.7846,
+      "step": 987
+    },
+    {
+      "epoch": 2.47057904196296,
+      "grad_norm": 0.2594124972820282,
+      "learning_rate": 3.506711409395974e-05,
+      "loss": 0.7962,
+      "step": 988
+    },
+    {
+      "epoch": 2.4730796280378216,
+      "grad_norm": 0.2742137908935547,
+      "learning_rate": 3.4899328859060405e-05,
+      "loss": 0.7744,
+      "step": 989
+    },
+    {
+      "epoch": 2.4755802141126826,
+      "grad_norm": 0.2660992443561554,
+      "learning_rate": 3.473154362416108e-05,
+      "loss": 0.7791,
+      "step": 990
+    },
+    {
+      "epoch": 2.478080800187544,
+      "grad_norm": 0.26402851939201355,
+      "learning_rate": 3.4563758389261744e-05,
+      "loss": 0.7854,
+      "step": 991
+    },
+    {
+      "epoch": 2.480581386262405,
+      "grad_norm": 0.2618241012096405,
+      "learning_rate": 3.439597315436242e-05,
+      "loss": 0.8587,
+      "step": 992
+    },
+    {
+      "epoch": 2.4830819723372666,
+      "grad_norm": 0.2527400255203247,
+      "learning_rate": 3.422818791946309e-05,
+      "loss": 0.778,
+      "step": 993
+    },
+    {
+      "epoch": 2.485582558412128,
+      "grad_norm": 0.266406387090683,
+      "learning_rate": 3.4060402684563756e-05,
+      "loss": 0.837,
+      "step": 994
+    },
+    {
+      "epoch": 2.488083144486989,
+      "grad_norm": 0.2607893645763397,
+      "learning_rate": 3.389261744966443e-05,
+      "loss": 0.7812,
+      "step": 995
+    },
+    {
+      "epoch": 2.4905837305618506,
+      "grad_norm": 0.26734456419944763,
+      "learning_rate": 3.37248322147651e-05,
+      "loss": 0.7932,
+      "step": 996
+    },
+    {
+      "epoch": 2.4930843166367116,
+      "grad_norm": 0.26186403632164,
+      "learning_rate": 3.3557046979865775e-05,
+      "loss": 0.7983,
+      "step": 997
+    },
+    {
+      "epoch": 2.495584902711573,
+      "grad_norm": 0.2572309076786041,
+      "learning_rate": 3.338926174496644e-05,
+      "loss": 0.754,
+      "step": 998
+    },
+    {
+      "epoch": 2.498085488786434,
+      "grad_norm": 0.2533490061759949,
+      "learning_rate": 3.3221476510067115e-05,
+      "loss": 0.7599,
+      "step": 999
+    },
+    {
+      "epoch": 2.5005860748612956,
+      "grad_norm": 0.2629091441631317,
+      "learning_rate": 3.305369127516779e-05,
+      "loss": 0.7471,
+      "step": 1000
+    },
+    {
+      "epoch": 2.503086660936157,
+      "grad_norm": 0.26545149087905884,
+      "learning_rate": 3.288590604026846e-05,
+      "loss": 0.7993,
+      "step": 1001
+    },
+    {
+      "epoch": 2.505587247011018,
+      "grad_norm": 0.27176633477211,
+      "learning_rate": 3.271812080536913e-05,
+      "loss": 0.7136,
+      "step": 1002
+    },
+    {
+      "epoch": 2.5080878330858796,
+      "grad_norm": 0.2601138949394226,
+      "learning_rate": 3.25503355704698e-05,
+      "loss": 0.7845,
+      "step": 1003
+    },
+    {
+      "epoch": 2.510588419160741,
+      "grad_norm": 0.2692638039588928,
+      "learning_rate": 3.238255033557047e-05,
+      "loss": 0.746,
+      "step": 1004
+    },
+    {
+      "epoch": 2.513089005235602,
+      "grad_norm": 0.26251697540283203,
+      "learning_rate": 3.221476510067114e-05,
+      "loss": 0.7552,
+      "step": 1005
+    },
+    {
+      "epoch": 2.515589591310463,
+      "grad_norm": 0.2608598470687866,
+      "learning_rate": 3.204697986577181e-05,
+      "loss": 0.7993,
+      "step": 1006
+    },
+    {
+      "epoch": 2.5180901773853246,
+      "grad_norm": 0.26370328664779663,
+      "learning_rate": 3.1879194630872485e-05,
+      "loss": 0.729,
+      "step": 1007
+    },
+    {
+      "epoch": 2.520590763460186,
+      "grad_norm": 0.2586556673049927,
+      "learning_rate": 3.171140939597316e-05,
+      "loss": 0.7759,
+      "step": 1008
+    },
+    {
+      "epoch": 2.523091349535047,
+      "grad_norm": 0.258453369140625,
+      "learning_rate": 3.1543624161073825e-05,
+      "loss": 0.7842,
+      "step": 1009
+    },
+    {
+      "epoch": 2.5255919356099086,
+      "grad_norm": 0.25726741552352905,
+      "learning_rate": 3.13758389261745e-05,
+      "loss": 0.7959,
+      "step": 1010
+    },
+    {
+      "epoch": 2.52809252168477,
+      "grad_norm": 0.2684222459793091,
+      "learning_rate": 3.120805369127517e-05,
+      "loss": 0.7602,
+      "step": 1011
+    },
+    {
+      "epoch": 2.530593107759631,
+      "grad_norm": 0.25897446274757385,
+      "learning_rate": 3.1040268456375844e-05,
+      "loss": 0.7748,
+      "step": 1012
+    },
+    {
+      "epoch": 2.5330936938344926,
+      "grad_norm": 0.26660779118537903,
+      "learning_rate": 3.087248322147651e-05,
+      "loss": 0.7613,
+      "step": 1013
+    },
+    {
+      "epoch": 2.5355942799093536,
+      "grad_norm": 0.25708746910095215,
+      "learning_rate": 3.070469798657718e-05,
+      "loss": 0.7915,
+      "step": 1014
+    },
+    {
+      "epoch": 2.538094865984215,
+      "grad_norm": 0.26296505331993103,
+      "learning_rate": 3.0536912751677856e-05,
+      "loss": 0.7485,
+      "step": 1015
+    },
+    {
+      "epoch": 2.540595452059076,
+      "grad_norm": 0.2721368670463562,
+      "learning_rate": 3.0369127516778522e-05,
+      "loss": 0.7031,
+      "step": 1016
+    },
+    {
+      "epoch": 2.5430960381339376,
+      "grad_norm": 0.2720634937286377,
+      "learning_rate": 3.02013422818792e-05,
+      "loss": 0.7754,
+      "step": 1017
+    },
+    {
+      "epoch": 2.545596624208799,
+      "grad_norm": 0.271311491727829,
+      "learning_rate": 3.0033557046979865e-05,
+      "loss": 0.7635,
+      "step": 1018
+    },
+    {
+      "epoch": 2.54809721028366,
+      "grad_norm": 0.26483824849128723,
+      "learning_rate": 2.986577181208054e-05,
+      "loss": 0.7518,
+      "step": 1019
+    },
+    {
+      "epoch": 2.5505977963585216,
+      "grad_norm": 0.2710844576358795,
+      "learning_rate": 2.9697986577181207e-05,
+      "loss": 0.8257,
+      "step": 1020
+    },
+    {
+      "epoch": 2.553098382433383,
+      "grad_norm": 0.26673269271850586,
+      "learning_rate": 2.9530201342281884e-05,
+      "loss": 0.7597,
+      "step": 1021
+    },
+    {
+      "epoch": 2.555598968508244,
+      "grad_norm": 0.26416197419166565,
+      "learning_rate": 2.936241610738255e-05,
+      "loss": 0.8071,
+      "step": 1022
+    },
+    {
+      "epoch": 2.5580995545831056,
+      "grad_norm": 0.2669306993484497,
+      "learning_rate": 2.9194630872483227e-05,
+      "loss": 0.7959,
+      "step": 1023
+    },
+    {
+      "epoch": 2.5606001406579666,
+      "grad_norm": 0.266413152217865,
+      "learning_rate": 2.9026845637583893e-05,
+      "loss": 0.7798,
+      "step": 1024
+    },
+    {
+      "epoch": 2.563100726732828,
+      "grad_norm": 0.2639547288417816,
+      "learning_rate": 2.885906040268457e-05,
+      "loss": 0.7373,
+      "step": 1025
+    },
+    {
+      "epoch": 2.565601312807689,
+      "grad_norm": 0.25265610218048096,
+      "learning_rate": 2.8691275167785235e-05,
+      "loss": 0.7428,
+      "step": 1026
+    },
+    {
+      "epoch": 2.5681018988825506,
+      "grad_norm": 0.2605838179588318,
+      "learning_rate": 2.8523489932885905e-05,
+      "loss": 0.7994,
+      "step": 1027
+    },
+    {
+      "epoch": 2.570602484957412,
+      "grad_norm": 0.2643495500087738,
+      "learning_rate": 2.8355704697986578e-05,
+      "loss": 0.8302,
+      "step": 1028
+    },
+    {
+      "epoch": 2.573103071032273,
+      "grad_norm": 0.2639097273349762,
+      "learning_rate": 2.8187919463087248e-05,
+      "loss": 0.7403,
+      "step": 1029
+    },
+    {
+      "epoch": 2.5756036571071346,
+      "grad_norm": 0.2704942226409912,
+      "learning_rate": 2.802013422818792e-05,
+      "loss": 0.742,
+      "step": 1030
+    },
+    {
+      "epoch": 2.578104243181996,
+      "grad_norm": 0.27417805790901184,
+      "learning_rate": 2.785234899328859e-05,
+      "loss": 0.7951,
+      "step": 1031
+    },
+    {
+      "epoch": 2.580604829256857,
+      "grad_norm": 0.2682810127735138,
+      "learning_rate": 2.7684563758389263e-05,
+      "loss": 0.7708,
+      "step": 1032
+    },
+    {
+      "epoch": 2.583105415331718,
+      "grad_norm": 0.25544607639312744,
+      "learning_rate": 2.7516778523489933e-05,
+      "loss": 0.7629,
+      "step": 1033
+    },
+    {
+      "epoch": 2.5856060014065796,
+      "grad_norm": 0.2566584646701813,
+      "learning_rate": 2.7348993288590606e-05,
+      "loss": 0.7672,
+      "step": 1034
+    },
+    {
+      "epoch": 2.588106587481441,
+      "grad_norm": 0.27147865295410156,
+      "learning_rate": 2.7181208053691276e-05,
+      "loss": 0.8237,
+      "step": 1035
+    },
+    {
+      "epoch": 2.590607173556302,
+      "grad_norm": 0.2876383662223816,
+      "learning_rate": 2.701342281879195e-05,
+      "loss": 0.7542,
+      "step": 1036
+    },
+    {
+      "epoch": 2.5931077596311636,
+      "grad_norm": 0.2700903117656708,
+      "learning_rate": 2.6845637583892618e-05,
+      "loss": 0.7472,
+      "step": 1037
+    },
+    {
+      "epoch": 2.595608345706025,
+      "grad_norm": 0.2761889696121216,
+      "learning_rate": 2.6677852348993288e-05,
+      "loss": 0.7608,
+      "step": 1038
+    },
+    {
+      "epoch": 2.598108931780886,
+      "grad_norm": 0.2627350986003876,
+      "learning_rate": 2.651006711409396e-05,
+      "loss": 0.7884,
+      "step": 1039
+    },
+    {
+      "epoch": 2.6006095178557476,
+      "grad_norm": 0.2600443661212921,
+      "learning_rate": 2.634228187919463e-05,
+      "loss": 0.7646,
+      "step": 1040
+    },
+    {
+      "epoch": 2.6031101039306086,
+      "grad_norm": 0.2687239646911621,
+      "learning_rate": 2.6174496644295304e-05,
+      "loss": 0.7837,
+      "step": 1041
+    },
+    {
+      "epoch": 2.60561069000547,
+      "grad_norm": 0.2561565935611725,
+      "learning_rate": 2.6006711409395973e-05,
+      "loss": 0.7751,
+      "step": 1042
+    },
+    {
+      "epoch": 2.608111276080331,
+      "grad_norm": 0.27367132902145386,
+      "learning_rate": 2.5838926174496646e-05,
+      "loss": 0.7827,
+      "step": 1043
+    },
+    {
+      "epoch": 2.6106118621551926,
+      "grad_norm": 0.2676515579223633,
+      "learning_rate": 2.5671140939597316e-05,
+      "loss": 0.7498,
+      "step": 1044
+    },
+    {
+      "epoch": 2.613112448230054,
+      "grad_norm": 0.2701331377029419,
+      "learning_rate": 2.550335570469799e-05,
+      "loss": 0.7557,
+      "step": 1045
+    },
+    {
+      "epoch": 2.615613034304915,
+      "grad_norm": 0.2638761103153229,
+      "learning_rate": 2.533557046979866e-05,
+      "loss": 0.756,
+      "step": 1046
+    },
+    {
+      "epoch": 2.6181136203797766,
+      "grad_norm": 0.26615577936172485,
+      "learning_rate": 2.516778523489933e-05,
+      "loss": 0.753,
+      "step": 1047
+    },
+    {
+      "epoch": 2.620614206454638,
+      "grad_norm": 0.2674086093902588,
+      "learning_rate": 2.5e-05,
+      "loss": 0.7683,
+      "step": 1048
+    },
+    {
+      "epoch": 2.623114792529499,
+      "grad_norm": 0.2707188129425049,
+      "learning_rate": 2.4832214765100674e-05,
+      "loss": 0.7737,
+      "step": 1049
+    },
+    {
+      "epoch": 2.6256153786043606,
+      "grad_norm": 0.26382195949554443,
+      "learning_rate": 2.4664429530201344e-05,
+      "loss": 0.773,
+      "step": 1050
+    },
+    {
+      "epoch": 2.6281159646792216,
+      "grad_norm": 0.27367544174194336,
+      "learning_rate": 2.4496644295302017e-05,
+      "loss": 0.7559,
+      "step": 1051
+    },
+    {
+      "epoch": 2.630616550754083,
+      "grad_norm": 0.27113792300224304,
+      "learning_rate": 2.4328859060402687e-05,
+      "loss": 0.8,
+      "step": 1052
+    },
+    {
+      "epoch": 2.633117136828944,
+      "grad_norm": 0.26364073157310486,
+      "learning_rate": 2.416107382550336e-05,
+      "loss": 0.7433,
+      "step": 1053
+    },
+    {
+      "epoch": 2.6356177229038056,
+      "grad_norm": 0.2878578007221222,
+      "learning_rate": 2.3993288590604026e-05,
+      "loss": 0.8011,
+      "step": 1054
+    },
+    {
+      "epoch": 2.638118308978667,
+      "grad_norm": 0.2688160240650177,
+      "learning_rate": 2.38255033557047e-05,
+      "loss": 0.7709,
+      "step": 1055
+    },
+    {
+      "epoch": 2.640618895053528,
+      "grad_norm": 0.2768557071685791,
+      "learning_rate": 2.365771812080537e-05,
+      "loss": 0.7644,
+      "step": 1056
+    },
+    {
+      "epoch": 2.6431194811283896,
+      "grad_norm": 0.25318166613578796,
+      "learning_rate": 2.348993288590604e-05,
+      "loss": 0.7462,
+      "step": 1057
+    },
+    {
+      "epoch": 2.6456200672032506,
+      "grad_norm": 0.26254960894584656,
+      "learning_rate": 2.332214765100671e-05,
+      "loss": 0.8197,
+      "step": 1058
+    },
+    {
+      "epoch": 2.648120653278112,
+      "grad_norm": 0.25807276368141174,
+      "learning_rate": 2.3154362416107384e-05,
+      "loss": 0.8078,
+      "step": 1059
+    },
+    {
+      "epoch": 2.650621239352973,
+      "grad_norm": 0.2553829252719879,
+      "learning_rate": 2.2986577181208054e-05,
+      "loss": 0.7824,
+      "step": 1060
+    },
+    {
+      "epoch": 2.6531218254278346,
+      "grad_norm": 0.25177881121635437,
+      "learning_rate": 2.2818791946308727e-05,
+      "loss": 0.816,
+      "step": 1061
+    },
+    {
+      "epoch": 2.655622411502696,
+      "grad_norm": 0.26017314195632935,
+      "learning_rate": 2.2651006711409396e-05,
+      "loss": 0.7933,
+      "step": 1062
+    },
+    {
+      "epoch": 2.658122997577557,
+      "grad_norm": 0.27956074476242065,
+      "learning_rate": 2.248322147651007e-05,
+      "loss": 0.7318,
+      "step": 1063
+    },
+    {
+      "epoch": 2.6606235836524186,
+      "grad_norm": 0.2631371021270752,
+      "learning_rate": 2.231543624161074e-05,
+      "loss": 0.8124,
+      "step": 1064
+    },
+    {
+      "epoch": 2.66312416972728,
+      "grad_norm": 0.2630529999732971,
+      "learning_rate": 2.2147651006711412e-05,
+      "loss": 0.7798,
+      "step": 1065
+    },
+    {
+      "epoch": 2.665624755802141,
+      "grad_norm": 0.2698495388031006,
+      "learning_rate": 2.197986577181208e-05,
+      "loss": 0.7402,
+      "step": 1066
+    },
+    {
+      "epoch": 2.6681253418770026,
+      "grad_norm": 0.2685375213623047,
+      "learning_rate": 2.181208053691275e-05,
+      "loss": 0.7634,
+      "step": 1067
+    },
+    {
+      "epoch": 2.6706259279518636,
+      "grad_norm": 0.27497631311416626,
+      "learning_rate": 2.1644295302013424e-05,
+      "loss": 0.8211,
+      "step": 1068
+    },
+    {
+      "epoch": 2.673126514026725,
+      "grad_norm": 0.2591504156589508,
+      "learning_rate": 2.1476510067114094e-05,
+      "loss": 0.7949,
+      "step": 1069
+    },
+    {
+      "epoch": 2.675627100101586,
+      "grad_norm": 0.26306867599487305,
+      "learning_rate": 2.1308724832214767e-05,
+      "loss": 0.7033,
+      "step": 1070
+    },
+    {
+      "epoch": 2.6781276861764476,
+      "grad_norm": 0.26264435052871704,
+      "learning_rate": 2.1140939597315437e-05,
+      "loss": 0.7679,
+      "step": 1071
+    },
+    {
+      "epoch": 2.680628272251309,
+      "grad_norm": 0.2685528099536896,
+      "learning_rate": 2.097315436241611e-05,
+      "loss": 0.7324,
+      "step": 1072
+    },
+    {
+      "epoch": 2.68312885832617,
+      "grad_norm": 0.25186851620674133,
+      "learning_rate": 2.080536912751678e-05,
+      "loss": 0.7889,
+      "step": 1073
+    },
+    {
+      "epoch": 2.6856294444010316,
+      "grad_norm": 0.26912397146224976,
+      "learning_rate": 2.0637583892617452e-05,
+      "loss": 0.7485,
+      "step": 1074
+    },
+    {
+      "epoch": 2.688130030475893,
+      "grad_norm": 0.280704140663147,
+      "learning_rate": 2.0469798657718122e-05,
+      "loss": 0.7907,
+      "step": 1075
+    },
+    {
+      "epoch": 2.690630616550754,
+      "grad_norm": 0.2644129991531372,
+      "learning_rate": 2.0302013422818795e-05,
+      "loss": 0.7597,
+      "step": 1076
+    },
+    {
+      "epoch": 2.693131202625615,
+      "grad_norm": 0.26235321164131165,
+      "learning_rate": 2.013422818791946e-05,
+      "loss": 0.8103,
+      "step": 1077
+    },
+    {
+      "epoch": 2.6956317887004766,
+      "grad_norm": 0.2663373649120331,
+      "learning_rate": 1.9966442953020134e-05,
+      "loss": 0.6938,
+      "step": 1078
+    },
+    {
+      "epoch": 2.698132374775338,
+      "grad_norm": 0.26949477195739746,
+      "learning_rate": 1.9798657718120804e-05,
+      "loss": 0.8105,
+      "step": 1079
+    },
+    {
+      "epoch": 2.700632960850199,
+      "grad_norm": 0.25903284549713135,
+      "learning_rate": 1.9630872483221477e-05,
+      "loss": 0.7911,
+      "step": 1080
+    },
+    {
+      "epoch": 2.7031335469250606,
+      "grad_norm": 0.25764670968055725,
+      "learning_rate": 1.946308724832215e-05,
+      "loss": 0.8019,
+      "step": 1081
+    },
+    {
+      "epoch": 2.705634132999922,
+      "grad_norm": 0.2695031762123108,
+      "learning_rate": 1.929530201342282e-05,
+      "loss": 0.7335,
+      "step": 1082
+    },
+    {
+      "epoch": 2.708134719074783,
+      "grad_norm": 0.27201348543167114,
+      "learning_rate": 1.9127516778523493e-05,
+      "loss": 0.811,
+      "step": 1083
+    },
+    {
+      "epoch": 2.7106353051496446,
+      "grad_norm": 0.25319722294807434,
+      "learning_rate": 1.8959731543624162e-05,
+      "loss": 0.7351,
+      "step": 1084
+    },
+    {
+      "epoch": 2.7131358912245056,
+      "grad_norm": 0.26669758558273315,
+      "learning_rate": 1.8791946308724835e-05,
+      "loss": 0.8104,
+      "step": 1085
+    },
+    {
+      "epoch": 2.715636477299367,
+      "grad_norm": 0.2601892054080963,
+      "learning_rate": 1.8624161073825505e-05,
+      "loss": 0.7458,
+      "step": 1086
+    },
+    {
+      "epoch": 2.718137063374228,
+      "grad_norm": 0.26129859685897827,
+      "learning_rate": 1.8456375838926178e-05,
+      "loss": 0.7658,
+      "step": 1087
+    },
+    {
+      "epoch": 2.7206376494490896,
+      "grad_norm": 0.2716621458530426,
+      "learning_rate": 1.8288590604026847e-05,
+      "loss": 0.7958,
+      "step": 1088
+    },
+    {
+      "epoch": 2.723138235523951,
+      "grad_norm": 0.26098430156707764,
+      "learning_rate": 1.8120805369127517e-05,
+      "loss": 0.727,
+      "step": 1089
+    },
+    {
+      "epoch": 2.725638821598812,
+      "grad_norm": 0.2624058723449707,
+      "learning_rate": 1.7953020134228187e-05,
+      "loss": 0.8095,
+      "step": 1090
+    },
+    {
+      "epoch": 2.7281394076736736,
+      "grad_norm": 0.26594632863998413,
+      "learning_rate": 1.778523489932886e-05,
+      "loss": 0.7466,
+      "step": 1091
+    },
+    {
+      "epoch": 2.730639993748535,
+      "grad_norm": 0.2688511908054352,
+      "learning_rate": 1.761744966442953e-05,
+      "loss": 0.8009,
+      "step": 1092
+    },
+    {
+      "epoch": 2.733140579823396,
+      "grad_norm": 0.27263709902763367,
+      "learning_rate": 1.7449664429530202e-05,
+      "loss": 0.7008,
+      "step": 1093
+    },
+    {
+      "epoch": 2.7356411658982576,
+      "grad_norm": 0.25561875104904175,
+      "learning_rate": 1.7281879194630872e-05,
+      "loss": 0.7201,
+      "step": 1094
+    },
+    {
+      "epoch": 2.7381417519731186,
+      "grad_norm": 0.2564723491668701,
+      "learning_rate": 1.7114093959731545e-05,
+      "loss": 0.7618,
+      "step": 1095
+    },
+    {
+      "epoch": 2.74064233804798,
+      "grad_norm": 0.2520231604576111,
+      "learning_rate": 1.6946308724832215e-05,
+      "loss": 0.7512,
+      "step": 1096
+    },
+    {
+      "epoch": 2.743142924122841,
+      "grad_norm": 0.2692018747329712,
+      "learning_rate": 1.6778523489932888e-05,
+      "loss": 0.7949,
+      "step": 1097
+    },
+    {
+      "epoch": 2.7456435101977026,
+      "grad_norm": 0.27487489581108093,
+      "learning_rate": 1.6610738255033557e-05,
+      "loss": 0.8054,
+      "step": 1098
+    },
+    {
+      "epoch": 2.748144096272564,
+      "grad_norm": 0.2683469355106354,
+      "learning_rate": 1.644295302013423e-05,
+      "loss": 0.7998,
+      "step": 1099
+    },
+    {
+      "epoch": 2.750644682347425,
+      "grad_norm": 0.25603777170181274,
+      "learning_rate": 1.62751677852349e-05,
+      "loss": 0.7705,
+      "step": 1100
+    },
+    {
+      "epoch": 2.7531452684222866,
+      "grad_norm": 0.2521064579486847,
+      "learning_rate": 1.610738255033557e-05,
+      "loss": 0.8382,
+      "step": 1101
+    },
+    {
+      "epoch": 2.7556458544971476,
+      "grad_norm": 0.2849092185497284,
+      "learning_rate": 1.5939597315436243e-05,
+      "loss": 0.7469,
+      "step": 1102
+    },
+    {
+      "epoch": 2.758146440572009,
+      "grad_norm": 0.2665373384952545,
+      "learning_rate": 1.5771812080536912e-05,
+      "loss": 0.7705,
+      "step": 1103
+    },
+    {
+      "epoch": 2.76064702664687,
+      "grad_norm": 0.25976264476776123,
+      "learning_rate": 1.5604026845637585e-05,
+      "loss": 0.7345,
+      "step": 1104
+    },
+    {
+      "epoch": 2.7631476127217316,
+      "grad_norm": 0.2643131613731384,
+      "learning_rate": 1.5436241610738255e-05,
+      "loss": 0.7876,
+      "step": 1105
+    },
+    {
+      "epoch": 2.765648198796593,
+      "grad_norm": 0.25333520770072937,
+      "learning_rate": 1.5268456375838928e-05,
+      "loss": 0.7383,
+      "step": 1106
+    },
+    {
+      "epoch": 2.768148784871454,
+      "grad_norm": 0.26729080080986023,
+      "learning_rate": 1.51006711409396e-05,
+      "loss": 0.7723,
+      "step": 1107
+    },
+    {
+      "epoch": 2.7706493709463156,
+      "grad_norm": 0.27518323063850403,
+      "learning_rate": 1.493288590604027e-05,
+      "loss": 0.8084,
+      "step": 1108
+    },
+    {
+      "epoch": 2.773149957021177,
+      "grad_norm": 0.2624513804912567,
+      "learning_rate": 1.4765100671140942e-05,
+      "loss": 0.7574,
+      "step": 1109
+    },
+    {
+      "epoch": 2.775650543096038,
+      "grad_norm": 0.2630423605442047,
+      "learning_rate": 1.4597315436241613e-05,
+      "loss": 0.7838,
+      "step": 1110
+    },
+    {
+      "epoch": 2.7781511291708996,
+      "grad_norm": 0.28019875288009644,
+      "learning_rate": 1.4429530201342285e-05,
+      "loss": 0.7646,
+      "step": 1111
+    },
+    {
+      "epoch": 2.7806517152457606,
+      "grad_norm": 0.253106027841568,
+      "learning_rate": 1.4261744966442953e-05,
+      "loss": 0.7455,
+      "step": 1112
+    },
+    {
+      "epoch": 2.783152301320622,
+      "grad_norm": 0.2707953453063965,
+      "learning_rate": 1.4093959731543624e-05,
+      "loss": 0.7182,
+      "step": 1113
+    },
+    {
+      "epoch": 2.785652887395483,
+      "grad_norm": 0.2584267258644104,
+      "learning_rate": 1.3926174496644295e-05,
+      "loss": 0.7682,
+      "step": 1114
+    },
+    {
+      "epoch": 2.7881534734703446,
+      "grad_norm": 0.2669549882411957,
+      "learning_rate": 1.3758389261744966e-05,
+      "loss": 0.7997,
+      "step": 1115
+    },
+    {
+      "epoch": 2.790654059545206,
+      "grad_norm": 0.2616058588027954,
+      "learning_rate": 1.3590604026845638e-05,
+      "loss": 0.7427,
+      "step": 1116
+    },
+    {
+      "epoch": 2.793154645620067,
+      "grad_norm": 0.2657564878463745,
+      "learning_rate": 1.3422818791946309e-05,
+      "loss": 0.7603,
+      "step": 1117
+    },
+    {
+      "epoch": 2.7956552316949286,
+      "grad_norm": 0.28465598821640015,
+      "learning_rate": 1.325503355704698e-05,
+      "loss": 0.7881,
+      "step": 1118
+    },
+    {
+      "epoch": 2.79815581776979,
+      "grad_norm": 0.2708990275859833,
+      "learning_rate": 1.3087248322147652e-05,
+      "loss": 0.7801,
+      "step": 1119
+    },
+    {
+      "epoch": 2.800656403844651,
+      "grad_norm": 0.2551485300064087,
+      "learning_rate": 1.2919463087248323e-05,
+      "loss": 0.7936,
+      "step": 1120
+    },
+    {
+      "epoch": 2.803156989919512,
+      "grad_norm": 0.26873987913131714,
+      "learning_rate": 1.2751677852348994e-05,
+      "loss": 0.7999,
+      "step": 1121
+    },
+    {
+      "epoch": 2.8056575759943736,
+      "grad_norm": 0.26690325140953064,
+      "learning_rate": 1.2583892617449666e-05,
+      "loss": 0.7492,
+      "step": 1122
+    },
+    {
+      "epoch": 2.808158162069235,
+      "grad_norm": 0.2745894193649292,
+      "learning_rate": 1.2416107382550337e-05,
+      "loss": 0.7771,
+      "step": 1123
+    },
+    {
+      "epoch": 2.810658748144096,
+      "grad_norm": 0.268500417470932,
+      "learning_rate": 1.2248322147651008e-05,
+      "loss": 0.7878,
+      "step": 1124
+    },
+    {
+      "epoch": 2.8131593342189576,
+      "grad_norm": 0.2567993402481079,
+      "learning_rate": 1.208053691275168e-05,
+      "loss": 0.7367,
+      "step": 1125
+    },
+    {
+      "epoch": 2.815659920293819,
+      "grad_norm": 0.270757257938385,
+      "learning_rate": 1.191275167785235e-05,
+      "loss": 0.784,
+      "step": 1126
+    },
+    {
+      "epoch": 2.81816050636868,
+      "grad_norm": 0.26289331912994385,
+      "learning_rate": 1.174496644295302e-05,
+      "loss": 0.7858,
+      "step": 1127
+    },
+    {
+      "epoch": 2.8206610924435416,
+      "grad_norm": 0.25360968708992004,
+      "learning_rate": 1.1577181208053692e-05,
+      "loss": 0.7789,
+      "step": 1128
+    },
+    {
+      "epoch": 2.8231616785184026,
+      "grad_norm": 0.2786359488964081,
+      "learning_rate": 1.1409395973154363e-05,
+      "loss": 0.7303,
+      "step": 1129
+    },
+    {
+      "epoch": 2.825662264593264,
+      "grad_norm": 0.2741607129573822,
+      "learning_rate": 1.1241610738255035e-05,
+      "loss": 0.7255,
+      "step": 1130
+    },
+    {
+      "epoch": 2.828162850668125,
+      "grad_norm": 0.26171642541885376,
+      "learning_rate": 1.1073825503355706e-05,
+      "loss": 0.749,
+      "step": 1131
+    },
+    {
+      "epoch": 2.8306634367429866,
+      "grad_norm": 0.2629507780075073,
+      "learning_rate": 1.0906040268456376e-05,
+      "loss": 0.7439,
+      "step": 1132
+    },
+    {
+      "epoch": 2.833164022817848,
+      "grad_norm": 0.2608429789543152,
+      "learning_rate": 1.0738255033557047e-05,
+      "loss": 0.7967,
+      "step": 1133
+    },
+    {
+      "epoch": 2.835664608892709,
+      "grad_norm": 0.2539840042591095,
+      "learning_rate": 1.0570469798657718e-05,
+      "loss": 0.7632,
+      "step": 1134
+    },
+    {
+      "epoch": 2.8381651949675706,
+      "grad_norm": 0.26609182357788086,
+      "learning_rate": 1.040268456375839e-05,
+      "loss": 0.7588,
+      "step": 1135
+    },
+    {
+      "epoch": 2.840665781042432,
+      "grad_norm": 0.2662641704082489,
+      "learning_rate": 1.0234899328859061e-05,
+      "loss": 0.767,
+      "step": 1136
+    },
+    {
+      "epoch": 2.843166367117293,
+      "grad_norm": 0.2716527581214905,
+      "learning_rate": 1.006711409395973e-05,
+      "loss": 0.7974,
+      "step": 1137
+    },
+    {
+      "epoch": 2.8456669531921546,
+      "grad_norm": 0.2681698799133301,
+      "learning_rate": 9.899328859060402e-06,
+      "loss": 0.7991,
+      "step": 1138
+    },
+    {
+      "epoch": 2.8481675392670156,
+      "grad_norm": 0.2646220624446869,
+      "learning_rate": 9.731543624161075e-06,
+      "loss": 0.8014,
+      "step": 1139
+    },
+    {
+      "epoch": 2.850668125341877,
+      "grad_norm": 0.26614639163017273,
+      "learning_rate": 9.563758389261746e-06,
+      "loss": 0.7139,
+      "step": 1140
+    },
+    {
+      "epoch": 2.853168711416738,
+      "grad_norm": 0.2810419797897339,
+      "learning_rate": 9.395973154362418e-06,
+      "loss": 0.7731,
+      "step": 1141
+    },
+    {
+      "epoch": 2.8556692974915996,
+      "grad_norm": 0.25968021154403687,
+      "learning_rate": 9.228187919463089e-06,
+      "loss": 0.8051,
+      "step": 1142
+    },
+    {
+      "epoch": 2.858169883566461,
+      "grad_norm": 0.2626001834869385,
+      "learning_rate": 9.060402684563759e-06,
+      "loss": 0.7884,
+      "step": 1143
+    },
+    {
+      "epoch": 2.860670469641322,
+      "grad_norm": 0.2615593373775482,
+      "learning_rate": 8.89261744966443e-06,
+      "loss": 0.7743,
+      "step": 1144
+    },
+    {
+      "epoch": 2.8631710557161836,
+      "grad_norm": 0.2630126178264618,
+      "learning_rate": 8.724832214765101e-06,
+      "loss": 0.7901,
+      "step": 1145
+    },
+    {
+      "epoch": 2.8656716417910446,
+      "grad_norm": 0.262552410364151,
+      "learning_rate": 8.557046979865773e-06,
+      "loss": 0.7725,
+      "step": 1146
+    },
+    {
+      "epoch": 2.868172227865906,
+      "grad_norm": 0.26387330889701843,
+      "learning_rate": 8.389261744966444e-06,
+      "loss": 0.7346,
+      "step": 1147
+    },
+    {
+      "epoch": 2.870672813940767,
+      "grad_norm": 0.26157933473587036,
+      "learning_rate": 8.221476510067115e-06,
+      "loss": 0.7882,
+      "step": 1148
+    },
+    {
+      "epoch": 2.8731734000156286,
+      "grad_norm": 0.26413771510124207,
+      "learning_rate": 8.053691275167785e-06,
+      "loss": 0.8176,
+      "step": 1149
+    },
+    {
+      "epoch": 2.87567398609049,
+      "grad_norm": 0.26703476905822754,
+      "learning_rate": 7.885906040268456e-06,
+      "loss": 0.7919,
+      "step": 1150
+    },
+    {
+      "epoch": 2.878174572165351,
+      "grad_norm": 0.2677638530731201,
+      "learning_rate": 7.718120805369127e-06,
+      "loss": 0.7637,
+      "step": 1151
+    },
+    {
+      "epoch": 2.8806751582402126,
+      "grad_norm": 0.28061383962631226,
+      "learning_rate": 7.5503355704698e-06,
+      "loss": 0.764,
+      "step": 1152
+    },
+    {
+      "epoch": 2.883175744315074,
+      "grad_norm": 0.30881813168525696,
+      "learning_rate": 7.382550335570471e-06,
+      "loss": 0.7423,
+      "step": 1153
+    },
+    {
+      "epoch": 2.885676330389935,
+      "grad_norm": 0.2670859396457672,
+      "learning_rate": 7.214765100671142e-06,
+      "loss": 0.8358,
+      "step": 1154
+    },
+    {
+      "epoch": 2.8881769164647966,
+      "grad_norm": 0.27003154158592224,
+      "learning_rate": 7.046979865771812e-06,
+      "loss": 0.7337,
+      "step": 1155
+    },
+    {
+      "epoch": 2.8906775025396576,
+      "grad_norm": 0.27815598249435425,
+      "learning_rate": 6.879194630872483e-06,
+      "loss": 0.8295,
+      "step": 1156
+    },
+    {
+      "epoch": 2.893178088614519,
+      "grad_norm": 0.267985075712204,
+      "learning_rate": 6.7114093959731546e-06,
+      "loss": 0.6885,
+      "step": 1157
+    },
+    {
+      "epoch": 2.89567867468938,
+      "grad_norm": 0.2884340286254883,
+      "learning_rate": 6.543624161073826e-06,
+      "loss": 0.7936,
+      "step": 1158
+    },
+    {
+      "epoch": 2.8981792607642416,
+      "grad_norm": 0.2722429931163788,
+      "learning_rate": 6.375838926174497e-06,
+      "loss": 0.7452,
+      "step": 1159
+    },
+    {
+      "epoch": 2.900679846839103,
+      "grad_norm": 0.2809288799762726,
+      "learning_rate": 6.2080536912751686e-06,
+      "loss": 0.803,
+      "step": 1160
+    },
+    {
+      "epoch": 2.903180432913964,
+      "grad_norm": 0.26661309599876404,
+      "learning_rate": 6.04026845637584e-06,
+      "loss": 0.7759,
+      "step": 1161
+    },
+    {
+      "epoch": 2.9056810189888256,
+      "grad_norm": 0.2651159465312958,
+      "learning_rate": 5.87248322147651e-06,
+      "loss": 0.7968,
+      "step": 1162
+    },
+    {
+      "epoch": 2.908181605063687,
+      "grad_norm": 0.2623818516731262,
+      "learning_rate": 5.704697986577182e-06,
+      "loss": 0.7441,
+      "step": 1163
+    },
+    {
+      "epoch": 2.910682191138548,
+      "grad_norm": 0.2637712359428406,
+      "learning_rate": 5.536912751677853e-06,
+      "loss": 0.7279,
+      "step": 1164
+    },
+    {
+      "epoch": 2.913182777213409,
+      "grad_norm": 0.2682763934135437,
+      "learning_rate": 5.3691275167785235e-06,
+      "loss": 0.7956,
+      "step": 1165
+    },
+    {
+      "epoch": 2.9156833632882706,
+      "grad_norm": 0.26668110489845276,
+      "learning_rate": 5.201342281879195e-06,
+      "loss": 0.7671,
+      "step": 1166
+    },
+    {
+      "epoch": 2.918183949363132,
+      "grad_norm": 0.2577240765094757,
+      "learning_rate": 5.033557046979865e-06,
+      "loss": 0.7932,
+      "step": 1167
+    },
+    {
+      "epoch": 2.920684535437993,
+      "grad_norm": 0.258863627910614,
+      "learning_rate": 4.8657718120805375e-06,
+      "loss": 0.8114,
+      "step": 1168
+    },
+    {
+      "epoch": 2.9231851215128546,
+      "grad_norm": 0.26777341961860657,
+      "learning_rate": 4.697986577181209e-06,
+      "loss": 0.7866,
+      "step": 1169
+    },
+    {
+      "epoch": 2.925685707587716,
+      "grad_norm": 0.2765660285949707,
+      "learning_rate": 4.530201342281879e-06,
+      "loss": 0.8136,
+      "step": 1170
+    },
+    {
+      "epoch": 2.928186293662577,
+      "grad_norm": 0.2771073281764984,
+      "learning_rate": 4.362416107382551e-06,
+      "loss": 0.7867,
+      "step": 1171
+    },
+    {
+      "epoch": 2.9306868797374386,
+      "grad_norm": 0.2580552399158478,
+      "learning_rate": 4.194630872483222e-06,
+      "loss": 0.7421,
+      "step": 1172
+    },
+    {
+      "epoch": 2.9331874658122996,
+      "grad_norm": 0.2619837522506714,
+      "learning_rate": 4.026845637583892e-06,
+      "loss": 0.7749,
+      "step": 1173
+    },
+    {
+      "epoch": 2.935688051887161,
+      "grad_norm": 0.263916939496994,
+      "learning_rate": 3.859060402684564e-06,
+      "loss": 0.7685,
+      "step": 1174
+    },
+    {
+      "epoch": 2.938188637962022,
+      "grad_norm": 0.2695319652557373,
+      "learning_rate": 3.6912751677852355e-06,
+      "loss": 0.7899,
+      "step": 1175
+    },
+    {
+      "epoch": 2.9406892240368836,
+      "grad_norm": 0.25818875432014465,
+      "learning_rate": 3.523489932885906e-06,
+      "loss": 0.7952,
+      "step": 1176
+    },
+    {
+      "epoch": 2.943189810111745,
+      "grad_norm": 0.273348331451416,
+      "learning_rate": 3.3557046979865773e-06,
+      "loss": 0.7718,
+      "step": 1177
+    },
+    {
+      "epoch": 2.945690396186606,
+      "grad_norm": 0.26397809386253357,
+      "learning_rate": 3.1879194630872486e-06,
+      "loss": 0.694,
+      "step": 1178
+    },
+    {
+      "epoch": 2.9481909822614676,
+      "grad_norm": 0.25927719473838806,
+      "learning_rate": 3.02013422818792e-06,
+      "loss": 0.7591,
+      "step": 1179
+    },
+    {
+      "epoch": 2.950691568336329,
+      "grad_norm": 0.26979297399520874,
+      "learning_rate": 2.852348993288591e-06,
+      "loss": 0.8195,
+      "step": 1180
+    },
+    {
+      "epoch": 2.95319215441119,
+      "grad_norm": 0.2654463052749634,
+      "learning_rate": 2.6845637583892617e-06,
+      "loss": 0.769,
+      "step": 1181
+    },
+    {
+      "epoch": 2.9556927404860516,
+      "grad_norm": 0.2589269280433655,
+      "learning_rate": 2.5167785234899326e-06,
+      "loss": 0.762,
+      "step": 1182
+    },
+    {
+      "epoch": 2.9581933265609126,
+      "grad_norm": 0.26951074600219727,
+      "learning_rate": 2.3489932885906044e-06,
+      "loss": 0.7893,
+      "step": 1183
+    },
+    {
+      "epoch": 2.960693912635774,
+      "grad_norm": 0.27295491099357605,
+      "learning_rate": 2.1812080536912753e-06,
+      "loss": 0.7942,
+      "step": 1184
+    },
+    {
+      "epoch": 2.963194498710635,
+      "grad_norm": 0.27274471521377563,
+      "learning_rate": 2.013422818791946e-06,
+      "loss": 0.7727,
+      "step": 1185
+    },
+    {
+      "epoch": 2.9656950847854966,
+      "grad_norm": 0.2667292654514313,
+      "learning_rate": 1.8456375838926177e-06,
+      "loss": 0.7749,
+      "step": 1186
+    },
+    {
+      "epoch": 2.968195670860358,
+      "grad_norm": 0.25694823265075684,
+      "learning_rate": 1.6778523489932886e-06,
+      "loss": 0.7478,
+      "step": 1187
+    },
+    {
+      "epoch": 2.970696256935219,
+      "grad_norm": 0.25563356280326843,
+      "learning_rate": 1.51006711409396e-06,
+      "loss": 0.7042,
+      "step": 1188
+    },
+    {
+      "epoch": 2.9731968430100806,
+      "grad_norm": 0.26784104108810425,
+      "learning_rate": 1.3422818791946309e-06,
+      "loss": 0.7381,
+      "step": 1189
+    },
+    {
+      "epoch": 2.975697429084942,
+      "grad_norm": 0.27679285407066345,
+      "learning_rate": 1.1744966442953022e-06,
+      "loss": 0.7915,
+      "step": 1190
+    },
+    {
+      "epoch": 2.978198015159803,
+      "grad_norm": 0.27582457661628723,
+      "learning_rate": 1.006711409395973e-06,
+      "loss": 0.8018,
+      "step": 1191
+    },
+    {
+      "epoch": 2.980698601234664,
+      "grad_norm": 0.26547786593437195,
+      "learning_rate": 8.389261744966443e-07,
+      "loss": 0.8207,
+      "step": 1192
+    },
+    {
+      "epoch": 2.9831991873095256,
+      "grad_norm": 0.26468461751937866,
+      "learning_rate": 6.711409395973154e-07,
+      "loss": 0.79,
+      "step": 1193
+    },
+    {
+      "epoch": 2.985699773384387,
+      "grad_norm": 0.2692301571369171,
+      "learning_rate": 5.033557046979866e-07,
+      "loss": 0.7913,
+      "step": 1194
+    },
+    {
+      "epoch": 2.988200359459248,
+      "grad_norm": 0.2573905289173126,
+      "learning_rate": 3.355704697986577e-07,
+      "loss": 0.7475,
+      "step": 1195
+    },
+    {
+      "epoch": 2.9907009455341096,
+      "grad_norm": 0.2607206106185913,
+      "learning_rate": 1.6778523489932886e-07,
+      "loss": 0.761,
+      "step": 1196
+    },
+    {
+      "epoch": 2.993201531608971,
+      "grad_norm": 0.2700795829296112,
+      "learning_rate": 0.0,
+      "loss": 0.7753,
+      "step": 1197
+    }
+  ],
+  "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": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.818762878799479e+18,
+  "train_batch_size": 5,
+  "trial_name": null,
+  "trial_params": null
+}