Files
TR_SFT/last-checkpoint/trainer_state.json
ModelHub XC d4c567ffae 初始化项目,由ModelHub XC社区提供模型
Model: Ba2han/TR_SFT
Source: Original Platform
2026-08-05 11:39:17 +08:00

16665 lines
416 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 948,
"global_step": 4740,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0004219743123137379,
"grad_norm": 0.3493342399597168,
"learning_rate": 2.8169014084507042e-06,
"loss": 2.3018248081207275,
"step": 2
},
{
"epoch": 0.0008439486246274758,
"grad_norm": 0.36698463559150696,
"learning_rate": 8.450704225352112e-06,
"loss": 2.3482322692871094,
"step": 4
},
{
"epoch": 0.0012659229369412137,
"grad_norm": 0.3716534376144409,
"learning_rate": 1.4084507042253523e-05,
"loss": 2.298100709915161,
"step": 6
},
{
"epoch": 0.0016878972492549516,
"grad_norm": 0.37269285321235657,
"learning_rate": 1.971830985915493e-05,
"loss": 2.3087313175201416,
"step": 8
},
{
"epoch": 0.0021098715615686897,
"grad_norm": 0.3556362986564636,
"learning_rate": 2.535211267605634e-05,
"loss": 2.353048801422119,
"step": 10
},
{
"epoch": 0.0025318458738824274,
"grad_norm": 0.368386447429657,
"learning_rate": 3.0985915492957744e-05,
"loss": 2.326158046722412,
"step": 12
},
{
"epoch": 0.0029538201861961655,
"grad_norm": 0.4067550301551819,
"learning_rate": 3.661971830985916e-05,
"loss": 2.333683729171753,
"step": 14
},
{
"epoch": 0.003375794498509903,
"grad_norm": 0.36954906582832336,
"learning_rate": 4.225352112676056e-05,
"loss": 2.290576934814453,
"step": 16
},
{
"epoch": 0.0037977688108236413,
"grad_norm": 0.35227838158607483,
"learning_rate": 4.788732394366197e-05,
"loss": 2.3628523349761963,
"step": 18
},
{
"epoch": 0.004219743123137379,
"grad_norm": 0.35705694556236267,
"learning_rate": 5.352112676056338e-05,
"loss": 2.377419948577881,
"step": 20
},
{
"epoch": 0.004641717435451117,
"grad_norm": 0.35026416182518005,
"learning_rate": 5.915492957746479e-05,
"loss": 2.349910020828247,
"step": 22
},
{
"epoch": 0.005063691747764855,
"grad_norm": 0.3596958518028259,
"learning_rate": 6.47887323943662e-05,
"loss": 2.3134686946868896,
"step": 24
},
{
"epoch": 0.005485666060078592,
"grad_norm": 0.34006455540657043,
"learning_rate": 7.042253521126761e-05,
"loss": 2.256657123565674,
"step": 26
},
{
"epoch": 0.005907640372392331,
"grad_norm": 0.36008140444755554,
"learning_rate": 7.605633802816902e-05,
"loss": 2.362276315689087,
"step": 28
},
{
"epoch": 0.006329614684706069,
"grad_norm": 0.3368754982948303,
"learning_rate": 8.169014084507043e-05,
"loss": 2.33296799659729,
"step": 30
},
{
"epoch": 0.006751588997019806,
"grad_norm": 0.3484903573989868,
"learning_rate": 8.732394366197182e-05,
"loss": 2.316190242767334,
"step": 32
},
{
"epoch": 0.007173563309333544,
"grad_norm": 0.3375855088233948,
"learning_rate": 9.295774647887325e-05,
"loss": 2.388054370880127,
"step": 34
},
{
"epoch": 0.0075955376216472826,
"grad_norm": 0.35737884044647217,
"learning_rate": 9.859154929577466e-05,
"loss": 2.3748927116394043,
"step": 36
},
{
"epoch": 0.00801751193396102,
"grad_norm": 0.3502481281757355,
"learning_rate": 0.00010422535211267605,
"loss": 2.3398969173431396,
"step": 38
},
{
"epoch": 0.008439486246274759,
"grad_norm": 0.33203786611557007,
"learning_rate": 0.00010985915492957747,
"loss": 2.4001569747924805,
"step": 40
},
{
"epoch": 0.008861460558588496,
"grad_norm": 0.30521464347839355,
"learning_rate": 0.00011549295774647888,
"loss": 2.338710308074951,
"step": 42
},
{
"epoch": 0.009283434870902234,
"grad_norm": 0.2896929383277893,
"learning_rate": 0.00012112676056338028,
"loss": 2.371890068054199,
"step": 44
},
{
"epoch": 0.009705409183215972,
"grad_norm": 0.31149810552597046,
"learning_rate": 0.0001267605633802817,
"loss": 2.3320541381835938,
"step": 46
},
{
"epoch": 0.01012738349552971,
"grad_norm": 0.3013763129711151,
"learning_rate": 0.0001323943661971831,
"loss": 2.2925612926483154,
"step": 48
},
{
"epoch": 0.010549357807843447,
"grad_norm": 0.2899003326892853,
"learning_rate": 0.00013802816901408452,
"loss": 2.3739919662475586,
"step": 50
},
{
"epoch": 0.010971332120157185,
"grad_norm": 0.28234249353408813,
"learning_rate": 0.00014366197183098594,
"loss": 2.3084330558776855,
"step": 52
},
{
"epoch": 0.011393306432470923,
"grad_norm": 0.2665596902370453,
"learning_rate": 0.00014929577464788734,
"loss": 2.32132625579834,
"step": 54
},
{
"epoch": 0.011815280744784662,
"grad_norm": 0.2801983058452606,
"learning_rate": 0.00015492957746478874,
"loss": 2.3287699222564697,
"step": 56
},
{
"epoch": 0.0122372550570984,
"grad_norm": 0.26391807198524475,
"learning_rate": 0.00016056338028169016,
"loss": 2.3672239780426025,
"step": 58
},
{
"epoch": 0.012659229369412137,
"grad_norm": 0.25507083535194397,
"learning_rate": 0.00016619718309859155,
"loss": 2.312551498413086,
"step": 60
},
{
"epoch": 0.013081203681725875,
"grad_norm": 0.24501122534275055,
"learning_rate": 0.00017183098591549295,
"loss": 2.3597629070281982,
"step": 62
},
{
"epoch": 0.013503177994039613,
"grad_norm": 0.24523726105690002,
"learning_rate": 0.00017746478873239437,
"loss": 2.3705875873565674,
"step": 64
},
{
"epoch": 0.01392515230635335,
"grad_norm": 0.23595115542411804,
"learning_rate": 0.0001830985915492958,
"loss": 2.3179006576538086,
"step": 66
},
{
"epoch": 0.014347126618667088,
"grad_norm": 0.2102200835943222,
"learning_rate": 0.0001887323943661972,
"loss": 2.2976298332214355,
"step": 68
},
{
"epoch": 0.014769100930980826,
"grad_norm": 0.208624005317688,
"learning_rate": 0.00019436619718309861,
"loss": 2.3517978191375732,
"step": 70
},
{
"epoch": 0.015191075243294565,
"grad_norm": 0.21405042707920074,
"learning_rate": 0.0002,
"loss": 2.2818446159362793,
"step": 72
},
{
"epoch": 0.015613049555608303,
"grad_norm": 0.2184291034936905,
"learning_rate": 0.0002056338028169014,
"loss": 2.229463815689087,
"step": 74
},
{
"epoch": 0.01603502386792204,
"grad_norm": 0.20547175407409668,
"learning_rate": 0.0002112676056338028,
"loss": 2.3209261894226074,
"step": 76
},
{
"epoch": 0.016456998180235778,
"grad_norm": 0.21404209733009338,
"learning_rate": 0.00021690140845070425,
"loss": 2.3340766429901123,
"step": 78
},
{
"epoch": 0.016878972492549518,
"grad_norm": 0.21448743343353271,
"learning_rate": 0.00022253521126760565,
"loss": 2.335631847381592,
"step": 80
},
{
"epoch": 0.017300946804863254,
"grad_norm": 0.22131432592868805,
"learning_rate": 0.00022816901408450707,
"loss": 2.3307600021362305,
"step": 82
},
{
"epoch": 0.017722921117176993,
"grad_norm": 0.20646852254867554,
"learning_rate": 0.00023380281690140847,
"loss": 2.3425393104553223,
"step": 84
},
{
"epoch": 0.01814489542949073,
"grad_norm": 0.20190732181072235,
"learning_rate": 0.00023943661971830986,
"loss": 2.351015329360962,
"step": 86
},
{
"epoch": 0.01856686974180447,
"grad_norm": 0.20206691324710846,
"learning_rate": 0.00024507042253521126,
"loss": 2.3061749935150146,
"step": 88
},
{
"epoch": 0.018988844054118204,
"grad_norm": 0.2107827067375183,
"learning_rate": 0.0002507042253521127,
"loss": 2.335526704788208,
"step": 90
},
{
"epoch": 0.019410818366431944,
"grad_norm": 0.2030932754278183,
"learning_rate": 0.0002563380281690141,
"loss": 2.3454275131225586,
"step": 92
},
{
"epoch": 0.019832792678745683,
"grad_norm": 0.20020480453968048,
"learning_rate": 0.0002619718309859155,
"loss": 2.3096773624420166,
"step": 94
},
{
"epoch": 0.02025476699105942,
"grad_norm": 0.18525218963623047,
"learning_rate": 0.0002676056338028169,
"loss": 2.265061855316162,
"step": 96
},
{
"epoch": 0.02067674130337316,
"grad_norm": 0.1751384735107422,
"learning_rate": 0.0002732394366197183,
"loss": 2.2887485027313232,
"step": 98
},
{
"epoch": 0.021098715615686894,
"grad_norm": 0.19603054225444794,
"learning_rate": 0.00027887323943661974,
"loss": 2.313230037689209,
"step": 100
},
{
"epoch": 0.021520689928000634,
"grad_norm": 0.17642512917518616,
"learning_rate": 0.00028450704225352114,
"loss": 2.2815263271331787,
"step": 102
},
{
"epoch": 0.02194266424031437,
"grad_norm": 0.18744505941867828,
"learning_rate": 0.0002901408450704226,
"loss": 2.2875919342041016,
"step": 104
},
{
"epoch": 0.02236463855262811,
"grad_norm": 0.18320868909358978,
"learning_rate": 0.000295774647887324,
"loss": 2.3087263107299805,
"step": 106
},
{
"epoch": 0.022786612864941845,
"grad_norm": 0.2057608962059021,
"learning_rate": 0.0003014084507042254,
"loss": 2.289973258972168,
"step": 108
},
{
"epoch": 0.023208587177255584,
"grad_norm": 0.16337190568447113,
"learning_rate": 0.00030704225352112677,
"loss": 2.270524024963379,
"step": 110
},
{
"epoch": 0.023630561489569324,
"grad_norm": 0.1729978770017624,
"learning_rate": 0.00031267605633802817,
"loss": 2.235231399536133,
"step": 112
},
{
"epoch": 0.02405253580188306,
"grad_norm": 0.17837706208229065,
"learning_rate": 0.0003183098591549296,
"loss": 2.1949877738952637,
"step": 114
},
{
"epoch": 0.0244745101141968,
"grad_norm": 0.1744994968175888,
"learning_rate": 0.000323943661971831,
"loss": 2.387223958969116,
"step": 116
},
{
"epoch": 0.024896484426510535,
"grad_norm": 0.16475647687911987,
"learning_rate": 0.0003295774647887324,
"loss": 2.296471118927002,
"step": 118
},
{
"epoch": 0.025318458738824275,
"grad_norm": 0.1803482472896576,
"learning_rate": 0.0003352112676056338,
"loss": 2.25184965133667,
"step": 120
},
{
"epoch": 0.02574043305113801,
"grad_norm": 0.16526687145233154,
"learning_rate": 0.0003408450704225352,
"loss": 2.266052484512329,
"step": 122
},
{
"epoch": 0.02616240736345175,
"grad_norm": 0.15544477105140686,
"learning_rate": 0.0003464788732394366,
"loss": 2.24651837348938,
"step": 124
},
{
"epoch": 0.02658438167576549,
"grad_norm": 0.16564379632472992,
"learning_rate": 0.00035211267605633805,
"loss": 2.26149320602417,
"step": 126
},
{
"epoch": 0.027006355988079225,
"grad_norm": 0.16480442881584167,
"learning_rate": 0.00035774647887323944,
"loss": 2.299612283706665,
"step": 128
},
{
"epoch": 0.027428330300392965,
"grad_norm": 0.16218475997447968,
"learning_rate": 0.0003633802816901409,
"loss": 2.3423268795013428,
"step": 130
},
{
"epoch": 0.0278503046127067,
"grad_norm": 0.16491222381591797,
"learning_rate": 0.0003690140845070423,
"loss": 2.2605960369110107,
"step": 132
},
{
"epoch": 0.02827227892502044,
"grad_norm": 0.16141007840633392,
"learning_rate": 0.0003746478873239437,
"loss": 2.2661232948303223,
"step": 134
},
{
"epoch": 0.028694253237334176,
"grad_norm": 0.1549309939146042,
"learning_rate": 0.0003802816901408451,
"loss": 2.344064950942993,
"step": 136
},
{
"epoch": 0.029116227549647915,
"grad_norm": 0.1570488065481186,
"learning_rate": 0.00038591549295774653,
"loss": 2.3017823696136475,
"step": 138
},
{
"epoch": 0.02953820186196165,
"grad_norm": 0.17115966975688934,
"learning_rate": 0.0003915492957746479,
"loss": 2.334764242172241,
"step": 140
},
{
"epoch": 0.02996017617427539,
"grad_norm": 0.1523764729499817,
"learning_rate": 0.0003971830985915493,
"loss": 2.2963616847991943,
"step": 142
},
{
"epoch": 0.03038215048658913,
"grad_norm": 0.1497291475534439,
"learning_rate": 0.0004,
"loss": 2.298975944519043,
"step": 144
},
{
"epoch": 0.030804124798902866,
"grad_norm": 0.14731788635253906,
"learning_rate": 0.0004,
"loss": 2.2866318225860596,
"step": 146
},
{
"epoch": 0.031226099111216606,
"grad_norm": 0.1455175280570984,
"learning_rate": 0.0004,
"loss": 2.3155295848846436,
"step": 148
},
{
"epoch": 0.03164807342353034,
"grad_norm": 0.13517186045646667,
"learning_rate": 0.0004,
"loss": 2.202117443084717,
"step": 150
},
{
"epoch": 0.03207004773584408,
"grad_norm": 0.14148102700710297,
"learning_rate": 0.0004,
"loss": 2.28737473487854,
"step": 152
},
{
"epoch": 0.03249202204815782,
"grad_norm": 0.1418687403202057,
"learning_rate": 0.0004,
"loss": 2.34505033493042,
"step": 154
},
{
"epoch": 0.032913996360471556,
"grad_norm": 0.14460255205631256,
"learning_rate": 0.0004,
"loss": 2.285484790802002,
"step": 156
},
{
"epoch": 0.03333597067278529,
"grad_norm": 0.14013704657554626,
"learning_rate": 0.0004,
"loss": 2.3012213706970215,
"step": 158
},
{
"epoch": 0.033757944985099035,
"grad_norm": 0.138973206281662,
"learning_rate": 0.0004,
"loss": 2.23698353767395,
"step": 160
},
{
"epoch": 0.03417991929741277,
"grad_norm": 0.13872702419757843,
"learning_rate": 0.0004,
"loss": 2.263179063796997,
"step": 162
},
{
"epoch": 0.03460189360972651,
"grad_norm": 0.1377047747373581,
"learning_rate": 0.0004,
"loss": 2.207634925842285,
"step": 164
},
{
"epoch": 0.03502386792204024,
"grad_norm": 0.13006150722503662,
"learning_rate": 0.0004,
"loss": 2.2157764434814453,
"step": 166
},
{
"epoch": 0.035445842234353986,
"grad_norm": 0.13557542860507965,
"learning_rate": 0.0004,
"loss": 2.209714889526367,
"step": 168
},
{
"epoch": 0.03586781654666772,
"grad_norm": 0.1439904272556305,
"learning_rate": 0.0004,
"loss": 2.2272696495056152,
"step": 170
},
{
"epoch": 0.03628979085898146,
"grad_norm": 0.13868501782417297,
"learning_rate": 0.0004,
"loss": 2.212550640106201,
"step": 172
},
{
"epoch": 0.0367117651712952,
"grad_norm": 0.13962973654270172,
"learning_rate": 0.0004,
"loss": 2.1771223545074463,
"step": 174
},
{
"epoch": 0.03713373948360894,
"grad_norm": 0.1326267123222351,
"learning_rate": 0.0004,
"loss": 2.295919895172119,
"step": 176
},
{
"epoch": 0.03755571379592267,
"grad_norm": 0.13839052617549896,
"learning_rate": 0.0004,
"loss": 2.216897487640381,
"step": 178
},
{
"epoch": 0.03797768810823641,
"grad_norm": 0.1442750245332718,
"learning_rate": 0.0004,
"loss": 2.239670515060425,
"step": 180
},
{
"epoch": 0.03839966242055015,
"grad_norm": 0.12919053435325623,
"learning_rate": 0.0004,
"loss": 2.283879041671753,
"step": 182
},
{
"epoch": 0.03882163673286389,
"grad_norm": 0.1274397224187851,
"learning_rate": 0.0004,
"loss": 2.2696444988250732,
"step": 184
},
{
"epoch": 0.03924361104517762,
"grad_norm": 0.12707428634166718,
"learning_rate": 0.0004,
"loss": 2.2171404361724854,
"step": 186
},
{
"epoch": 0.039665585357491366,
"grad_norm": 0.12736071646213531,
"learning_rate": 0.0004,
"loss": 2.2601091861724854,
"step": 188
},
{
"epoch": 0.0400875596698051,
"grad_norm": 0.1365179419517517,
"learning_rate": 0.0004,
"loss": 2.2088310718536377,
"step": 190
},
{
"epoch": 0.04050953398211884,
"grad_norm": 0.12932445108890533,
"learning_rate": 0.0004,
"loss": 2.239926815032959,
"step": 192
},
{
"epoch": 0.040931508294432574,
"grad_norm": 0.12791691720485687,
"learning_rate": 0.0004,
"loss": 2.289055824279785,
"step": 194
},
{
"epoch": 0.04135348260674632,
"grad_norm": 0.1365281045436859,
"learning_rate": 0.0004,
"loss": 2.2621095180511475,
"step": 196
},
{
"epoch": 0.04177545691906005,
"grad_norm": 0.1271848827600479,
"learning_rate": 0.0004,
"loss": 2.2374236583709717,
"step": 198
},
{
"epoch": 0.04219743123137379,
"grad_norm": 0.13125190138816833,
"learning_rate": 0.0004,
"loss": 2.245105743408203,
"step": 200
},
{
"epoch": 0.042619405543687525,
"grad_norm": 0.11939355731010437,
"learning_rate": 0.0004,
"loss": 2.2083916664123535,
"step": 202
},
{
"epoch": 0.04304137985600127,
"grad_norm": 0.11357792466878891,
"learning_rate": 0.0004,
"loss": 2.2386586666107178,
"step": 204
},
{
"epoch": 0.043463354168315003,
"grad_norm": 0.12360386550426483,
"learning_rate": 0.0004,
"loss": 2.2166037559509277,
"step": 206
},
{
"epoch": 0.04388532848062874,
"grad_norm": 0.11785627156496048,
"learning_rate": 0.0004,
"loss": 2.1768479347229004,
"step": 208
},
{
"epoch": 0.04430730279294248,
"grad_norm": 0.12361887842416763,
"learning_rate": 0.0004,
"loss": 2.2362818717956543,
"step": 210
},
{
"epoch": 0.04472927710525622,
"grad_norm": 0.11254332214593887,
"learning_rate": 0.0004,
"loss": 2.2256295680999756,
"step": 212
},
{
"epoch": 0.045151251417569954,
"grad_norm": 0.12254314124584198,
"learning_rate": 0.0004,
"loss": 2.21423077583313,
"step": 214
},
{
"epoch": 0.04557322572988369,
"grad_norm": 0.1274789422750473,
"learning_rate": 0.0004,
"loss": 2.2605161666870117,
"step": 216
},
{
"epoch": 0.04599520004219743,
"grad_norm": 0.11587799340486526,
"learning_rate": 0.0004,
"loss": 2.202139377593994,
"step": 218
},
{
"epoch": 0.04641717435451117,
"grad_norm": 0.11587532609701157,
"learning_rate": 0.0004,
"loss": 2.2536325454711914,
"step": 220
},
{
"epoch": 0.046839148666824905,
"grad_norm": 0.12374117225408554,
"learning_rate": 0.0004,
"loss": 2.273543357849121,
"step": 222
},
{
"epoch": 0.04726112297913865,
"grad_norm": 0.11590168625116348,
"learning_rate": 0.0004,
"loss": 2.1677937507629395,
"step": 224
},
{
"epoch": 0.047683097291452384,
"grad_norm": 0.11316622048616409,
"learning_rate": 0.0004,
"loss": 2.260772228240967,
"step": 226
},
{
"epoch": 0.04810507160376612,
"grad_norm": 0.11277369409799576,
"learning_rate": 0.0004,
"loss": 2.189258337020874,
"step": 228
},
{
"epoch": 0.048527045916079856,
"grad_norm": 0.1201343908905983,
"learning_rate": 0.0004,
"loss": 2.23919415473938,
"step": 230
},
{
"epoch": 0.0489490202283936,
"grad_norm": 0.11876040697097778,
"learning_rate": 0.0004,
"loss": 2.176072120666504,
"step": 232
},
{
"epoch": 0.049370994540707334,
"grad_norm": 0.11445920169353485,
"learning_rate": 0.0004,
"loss": 2.164640426635742,
"step": 234
},
{
"epoch": 0.04979296885302107,
"grad_norm": 0.12036921083927155,
"learning_rate": 0.0004,
"loss": 2.2662060260772705,
"step": 236
},
{
"epoch": 0.05021494316533481,
"grad_norm": 0.1221209317445755,
"learning_rate": 0.0004,
"loss": 2.2302322387695312,
"step": 238
},
{
"epoch": 0.05063691747764855,
"grad_norm": 0.11136005073785782,
"learning_rate": 0.0004,
"loss": 2.2404186725616455,
"step": 240
},
{
"epoch": 0.051058891789962285,
"grad_norm": 0.12716515362262726,
"learning_rate": 0.0004,
"loss": 2.2142770290374756,
"step": 242
},
{
"epoch": 0.05148086610227602,
"grad_norm": 0.11764153093099594,
"learning_rate": 0.0004,
"loss": 2.223127841949463,
"step": 244
},
{
"epoch": 0.051902840414589764,
"grad_norm": 0.11981026828289032,
"learning_rate": 0.0004,
"loss": 2.1965675354003906,
"step": 246
},
{
"epoch": 0.0523248147269035,
"grad_norm": 0.11381073296070099,
"learning_rate": 0.0004,
"loss": 2.1804094314575195,
"step": 248
},
{
"epoch": 0.052746789039217236,
"grad_norm": 0.11537916213274002,
"learning_rate": 0.0004,
"loss": 2.220707893371582,
"step": 250
},
{
"epoch": 0.05316876335153098,
"grad_norm": 0.12030758708715439,
"learning_rate": 0.0004,
"loss": 2.2037346363067627,
"step": 252
},
{
"epoch": 0.053590737663844715,
"grad_norm": 0.11342093348503113,
"learning_rate": 0.0004,
"loss": 2.203667640686035,
"step": 254
},
{
"epoch": 0.05401271197615845,
"grad_norm": 0.1088421493768692,
"learning_rate": 0.0004,
"loss": 2.1932451725006104,
"step": 256
},
{
"epoch": 0.05443468628847219,
"grad_norm": 0.12139564752578735,
"learning_rate": 0.0004,
"loss": 2.245731830596924,
"step": 258
},
{
"epoch": 0.05485666060078593,
"grad_norm": 0.12005486339330673,
"learning_rate": 0.0004,
"loss": 2.204179286956787,
"step": 260
},
{
"epoch": 0.055278634913099665,
"grad_norm": 0.12492313235998154,
"learning_rate": 0.0004,
"loss": 2.2297842502593994,
"step": 262
},
{
"epoch": 0.0557006092254134,
"grad_norm": 0.11256934702396393,
"learning_rate": 0.0004,
"loss": 2.217824935913086,
"step": 264
},
{
"epoch": 0.056122583537727144,
"grad_norm": 0.10663718730211258,
"learning_rate": 0.0004,
"loss": 2.2163045406341553,
"step": 266
},
{
"epoch": 0.05654455785004088,
"grad_norm": 0.11845961213111877,
"learning_rate": 0.0004,
"loss": 2.245414972305298,
"step": 268
},
{
"epoch": 0.056966532162354616,
"grad_norm": 0.11159045994281769,
"learning_rate": 0.0004,
"loss": 2.215888738632202,
"step": 270
},
{
"epoch": 0.05738850647466835,
"grad_norm": 0.11191422492265701,
"learning_rate": 0.0004,
"loss": 2.205087900161743,
"step": 272
},
{
"epoch": 0.057810480786982095,
"grad_norm": 0.11480437964200974,
"learning_rate": 0.0004,
"loss": 2.166315793991089,
"step": 274
},
{
"epoch": 0.05823245509929583,
"grad_norm": 0.10876529663801193,
"learning_rate": 0.0004,
"loss": 2.1795241832733154,
"step": 276
},
{
"epoch": 0.05865442941160957,
"grad_norm": 0.12178465723991394,
"learning_rate": 0.0004,
"loss": 2.2085864543914795,
"step": 278
},
{
"epoch": 0.0590764037239233,
"grad_norm": 0.11340827494859695,
"learning_rate": 0.0004,
"loss": 2.2166683673858643,
"step": 280
},
{
"epoch": 0.059498378036237046,
"grad_norm": 0.10359188914299011,
"learning_rate": 0.0004,
"loss": 2.166851043701172,
"step": 282
},
{
"epoch": 0.05992035234855078,
"grad_norm": 0.11261853575706482,
"learning_rate": 0.0004,
"loss": 2.167685031890869,
"step": 284
},
{
"epoch": 0.06034232666086452,
"grad_norm": 0.10509487241506577,
"learning_rate": 0.0004,
"loss": 2.151081085205078,
"step": 286
},
{
"epoch": 0.06076430097317826,
"grad_norm": 0.11053803563117981,
"learning_rate": 0.0004,
"loss": 2.250033140182495,
"step": 288
},
{
"epoch": 0.061186275285491996,
"grad_norm": 0.11651729047298431,
"learning_rate": 0.0004,
"loss": 2.1693906784057617,
"step": 290
},
{
"epoch": 0.06160824959780573,
"grad_norm": 0.10973582416772842,
"learning_rate": 0.0004,
"loss": 2.1263091564178467,
"step": 292
},
{
"epoch": 0.06203022391011947,
"grad_norm": 0.10662171244621277,
"learning_rate": 0.0004,
"loss": 2.1781439781188965,
"step": 294
},
{
"epoch": 0.06245219822243321,
"grad_norm": 0.11675313860177994,
"learning_rate": 0.0004,
"loss": 2.2503230571746826,
"step": 296
},
{
"epoch": 0.06287417253474695,
"grad_norm": 0.11857406795024872,
"learning_rate": 0.0004,
"loss": 2.258619546890259,
"step": 298
},
{
"epoch": 0.06329614684706068,
"grad_norm": 0.11008710414171219,
"learning_rate": 0.0004,
"loss": 2.1991045475006104,
"step": 300
},
{
"epoch": 0.06371812115937442,
"grad_norm": 0.11976458132266998,
"learning_rate": 0.0004,
"loss": 2.2022125720977783,
"step": 302
},
{
"epoch": 0.06414009547168815,
"grad_norm": 0.1103653758764267,
"learning_rate": 0.0004,
"loss": 2.2240095138549805,
"step": 304
},
{
"epoch": 0.0645620697840019,
"grad_norm": 0.10308089852333069,
"learning_rate": 0.0004,
"loss": 2.151815176010132,
"step": 306
},
{
"epoch": 0.06498404409631564,
"grad_norm": 0.11728787422180176,
"learning_rate": 0.0004,
"loss": 2.1404049396514893,
"step": 308
},
{
"epoch": 0.06540601840862938,
"grad_norm": 0.11225762218236923,
"learning_rate": 0.0004,
"loss": 2.2400128841400146,
"step": 310
},
{
"epoch": 0.06582799272094311,
"grad_norm": 0.1165354773402214,
"learning_rate": 0.0004,
"loss": 2.2270724773406982,
"step": 312
},
{
"epoch": 0.06624996703325685,
"grad_norm": 0.11236653476953506,
"learning_rate": 0.0004,
"loss": 2.1535444259643555,
"step": 314
},
{
"epoch": 0.06667194134557058,
"grad_norm": 0.10641568154096603,
"learning_rate": 0.0004,
"loss": 2.161425828933716,
"step": 316
},
{
"epoch": 0.06709391565788432,
"grad_norm": 0.10073976218700409,
"learning_rate": 0.0004,
"loss": 2.154911994934082,
"step": 318
},
{
"epoch": 0.06751588997019807,
"grad_norm": 0.11362458765506744,
"learning_rate": 0.0004,
"loss": 2.1593425273895264,
"step": 320
},
{
"epoch": 0.0679378642825118,
"grad_norm": 0.13053755462169647,
"learning_rate": 0.0004,
"loss": 2.231839895248413,
"step": 322
},
{
"epoch": 0.06835983859482554,
"grad_norm": 0.1055619940161705,
"learning_rate": 0.0004,
"loss": 2.230339527130127,
"step": 324
},
{
"epoch": 0.06878181290713928,
"grad_norm": 0.11623658239841461,
"learning_rate": 0.0004,
"loss": 2.111253499984741,
"step": 326
},
{
"epoch": 0.06920378721945301,
"grad_norm": 0.1088699996471405,
"learning_rate": 0.0004,
"loss": 2.1576030254364014,
"step": 328
},
{
"epoch": 0.06962576153176675,
"grad_norm": 0.11168421059846878,
"learning_rate": 0.0004,
"loss": 2.1996970176696777,
"step": 330
},
{
"epoch": 0.07004773584408049,
"grad_norm": 0.10531529784202576,
"learning_rate": 0.0004,
"loss": 2.2236742973327637,
"step": 332
},
{
"epoch": 0.07046971015639424,
"grad_norm": 0.10702145099639893,
"learning_rate": 0.0004,
"loss": 2.152845621109009,
"step": 334
},
{
"epoch": 0.07089168446870797,
"grad_norm": 0.10920841246843338,
"learning_rate": 0.0004,
"loss": 2.132333278656006,
"step": 336
},
{
"epoch": 0.07131365878102171,
"grad_norm": 0.10891810059547424,
"learning_rate": 0.0004,
"loss": 2.090947151184082,
"step": 338
},
{
"epoch": 0.07173563309333544,
"grad_norm": 0.11763957142829895,
"learning_rate": 0.0004,
"loss": 2.2283878326416016,
"step": 340
},
{
"epoch": 0.07215760740564918,
"grad_norm": 0.1033068373799324,
"learning_rate": 0.0004,
"loss": 2.1318740844726562,
"step": 342
},
{
"epoch": 0.07257958171796292,
"grad_norm": 0.11240024864673615,
"learning_rate": 0.0004,
"loss": 2.2340359687805176,
"step": 344
},
{
"epoch": 0.07300155603027665,
"grad_norm": 0.10983721911907196,
"learning_rate": 0.0004,
"loss": 2.1715822219848633,
"step": 346
},
{
"epoch": 0.0734235303425904,
"grad_norm": 0.11801565438508987,
"learning_rate": 0.0004,
"loss": 2.234556198120117,
"step": 348
},
{
"epoch": 0.07384550465490414,
"grad_norm": 0.10622187703847885,
"learning_rate": 0.0004,
"loss": 2.206968069076538,
"step": 350
},
{
"epoch": 0.07426747896721787,
"grad_norm": 0.11609214544296265,
"learning_rate": 0.0004,
"loss": 2.1820499897003174,
"step": 352
},
{
"epoch": 0.07468945327953161,
"grad_norm": 0.11098748445510864,
"learning_rate": 0.0004,
"loss": 2.1106762886047363,
"step": 354
},
{
"epoch": 0.07511142759184535,
"grad_norm": 0.11470896005630493,
"learning_rate": 0.0004,
"loss": 2.1696507930755615,
"step": 356
},
{
"epoch": 0.07553340190415908,
"grad_norm": 0.10947126895189285,
"learning_rate": 0.0004,
"loss": 2.1818830966949463,
"step": 358
},
{
"epoch": 0.07595537621647282,
"grad_norm": 0.11120952665805817,
"learning_rate": 0.0004,
"loss": 2.2424967288970947,
"step": 360
},
{
"epoch": 0.07637735052878657,
"grad_norm": 0.0992329865694046,
"learning_rate": 0.0004,
"loss": 2.165295124053955,
"step": 362
},
{
"epoch": 0.0767993248411003,
"grad_norm": 0.1052723303437233,
"learning_rate": 0.0004,
"loss": 2.13628888130188,
"step": 364
},
{
"epoch": 0.07722129915341404,
"grad_norm": 0.11137950420379639,
"learning_rate": 0.0004,
"loss": 2.161299467086792,
"step": 366
},
{
"epoch": 0.07764327346572777,
"grad_norm": 0.10961909592151642,
"learning_rate": 0.0004,
"loss": 2.067732810974121,
"step": 368
},
{
"epoch": 0.07806524777804151,
"grad_norm": 0.10626599192619324,
"learning_rate": 0.0004,
"loss": 2.221228837966919,
"step": 370
},
{
"epoch": 0.07848722209035525,
"grad_norm": 0.1092216894030571,
"learning_rate": 0.0004,
"loss": 2.237510919570923,
"step": 372
},
{
"epoch": 0.07890919640266898,
"grad_norm": 0.10934820771217346,
"learning_rate": 0.0004,
"loss": 2.1890249252319336,
"step": 374
},
{
"epoch": 0.07933117071498273,
"grad_norm": 0.10721974819898605,
"learning_rate": 0.0004,
"loss": 2.163759708404541,
"step": 376
},
{
"epoch": 0.07975314502729647,
"grad_norm": 0.10925968736410141,
"learning_rate": 0.0004,
"loss": 2.2155017852783203,
"step": 378
},
{
"epoch": 0.0801751193396102,
"grad_norm": 0.10603740811347961,
"learning_rate": 0.0004,
"loss": 2.177000045776367,
"step": 380
},
{
"epoch": 0.08059709365192394,
"grad_norm": 0.10820295661687851,
"learning_rate": 0.0004,
"loss": 2.1381659507751465,
"step": 382
},
{
"epoch": 0.08101906796423768,
"grad_norm": 0.11538650095462799,
"learning_rate": 0.0004,
"loss": 2.1240668296813965,
"step": 384
},
{
"epoch": 0.08144104227655141,
"grad_norm": 0.10810080915689468,
"learning_rate": 0.0004,
"loss": 2.144092559814453,
"step": 386
},
{
"epoch": 0.08186301658886515,
"grad_norm": 0.10264503210783005,
"learning_rate": 0.0004,
"loss": 2.1740212440490723,
"step": 388
},
{
"epoch": 0.08228499090117888,
"grad_norm": 0.10461390763521194,
"learning_rate": 0.0004,
"loss": 2.2094638347625732,
"step": 390
},
{
"epoch": 0.08270696521349263,
"grad_norm": 0.1066846251487732,
"learning_rate": 0.0004,
"loss": 2.1450464725494385,
"step": 392
},
{
"epoch": 0.08312893952580637,
"grad_norm": 0.11936092376708984,
"learning_rate": 0.0004,
"loss": 2.249525785446167,
"step": 394
},
{
"epoch": 0.0835509138381201,
"grad_norm": 0.10146570950746536,
"learning_rate": 0.0004,
"loss": 2.172985792160034,
"step": 396
},
{
"epoch": 0.08397288815043384,
"grad_norm": 0.10137574374675751,
"learning_rate": 0.0004,
"loss": 2.20550799369812,
"step": 398
},
{
"epoch": 0.08439486246274758,
"grad_norm": 0.10552877187728882,
"learning_rate": 0.0004,
"loss": 2.108227491378784,
"step": 400
},
{
"epoch": 0.08481683677506131,
"grad_norm": 0.11072912067174911,
"learning_rate": 0.0004,
"loss": 2.113502264022827,
"step": 402
},
{
"epoch": 0.08523881108737505,
"grad_norm": 0.110051229596138,
"learning_rate": 0.0004,
"loss": 2.11564040184021,
"step": 404
},
{
"epoch": 0.0856607853996888,
"grad_norm": 0.1032586321234703,
"learning_rate": 0.0004,
"loss": 2.1291587352752686,
"step": 406
},
{
"epoch": 0.08608275971200254,
"grad_norm": 0.12690642476081848,
"learning_rate": 0.0004,
"loss": 2.15268611907959,
"step": 408
},
{
"epoch": 0.08650473402431627,
"grad_norm": 0.10999573767185211,
"learning_rate": 0.0004,
"loss": 2.1778106689453125,
"step": 410
},
{
"epoch": 0.08692670833663001,
"grad_norm": 0.1014145016670227,
"learning_rate": 0.0004,
"loss": 2.1154181957244873,
"step": 412
},
{
"epoch": 0.08734868264894374,
"grad_norm": 0.09770940989255905,
"learning_rate": 0.0004,
"loss": 2.173128604888916,
"step": 414
},
{
"epoch": 0.08777065696125748,
"grad_norm": 0.10573363304138184,
"learning_rate": 0.0004,
"loss": 2.2228899002075195,
"step": 416
},
{
"epoch": 0.08819263127357121,
"grad_norm": 0.10503196716308594,
"learning_rate": 0.0004,
"loss": 2.1295382976531982,
"step": 418
},
{
"epoch": 0.08861460558588496,
"grad_norm": 0.10423147678375244,
"learning_rate": 0.0004,
"loss": 2.1498734951019287,
"step": 420
},
{
"epoch": 0.0890365798981987,
"grad_norm": 0.10368449240922928,
"learning_rate": 0.0004,
"loss": 2.165558099746704,
"step": 422
},
{
"epoch": 0.08945855421051244,
"grad_norm": 0.11377409845590591,
"learning_rate": 0.0004,
"loss": 2.1532649993896484,
"step": 424
},
{
"epoch": 0.08988052852282617,
"grad_norm": 0.10557077080011368,
"learning_rate": 0.0004,
"loss": 2.185331344604492,
"step": 426
},
{
"epoch": 0.09030250283513991,
"grad_norm": 0.11954786628484726,
"learning_rate": 0.0004,
"loss": 2.1350667476654053,
"step": 428
},
{
"epoch": 0.09072447714745364,
"grad_norm": 0.10571636259555817,
"learning_rate": 0.0004,
"loss": 2.111095905303955,
"step": 430
},
{
"epoch": 0.09114645145976738,
"grad_norm": 0.10475218296051025,
"learning_rate": 0.0004,
"loss": 2.1517415046691895,
"step": 432
},
{
"epoch": 0.09156842577208113,
"grad_norm": 0.10855066031217575,
"learning_rate": 0.0004,
"loss": 2.067002058029175,
"step": 434
},
{
"epoch": 0.09199040008439487,
"grad_norm": 0.10663752257823944,
"learning_rate": 0.0004,
"loss": 2.1381430625915527,
"step": 436
},
{
"epoch": 0.0924123743967086,
"grad_norm": 0.12324759364128113,
"learning_rate": 0.0004,
"loss": 2.198122501373291,
"step": 438
},
{
"epoch": 0.09283434870902234,
"grad_norm": 0.10396029055118561,
"learning_rate": 0.0004,
"loss": 2.172659397125244,
"step": 440
},
{
"epoch": 0.09325632302133607,
"grad_norm": 0.10649340599775314,
"learning_rate": 0.0004,
"loss": 2.0995426177978516,
"step": 442
},
{
"epoch": 0.09367829733364981,
"grad_norm": 0.1164933368563652,
"learning_rate": 0.0004,
"loss": 2.158050537109375,
"step": 444
},
{
"epoch": 0.09410027164596355,
"grad_norm": 0.10775702446699142,
"learning_rate": 0.0004,
"loss": 2.178995132446289,
"step": 446
},
{
"epoch": 0.0945222459582773,
"grad_norm": 0.09974786639213562,
"learning_rate": 0.0004,
"loss": 2.1841464042663574,
"step": 448
},
{
"epoch": 0.09494422027059103,
"grad_norm": 0.10044031590223312,
"learning_rate": 0.0004,
"loss": 2.1888539791107178,
"step": 450
},
{
"epoch": 0.09536619458290477,
"grad_norm": 0.1048150286078453,
"learning_rate": 0.0004,
"loss": 2.098114252090454,
"step": 452
},
{
"epoch": 0.0957881688952185,
"grad_norm": 0.11264552175998688,
"learning_rate": 0.0004,
"loss": 2.1471683979034424,
"step": 454
},
{
"epoch": 0.09621014320753224,
"grad_norm": 0.10416817665100098,
"learning_rate": 0.0004,
"loss": 2.185790777206421,
"step": 456
},
{
"epoch": 0.09663211751984598,
"grad_norm": 0.106419138610363,
"learning_rate": 0.0004,
"loss": 2.110515594482422,
"step": 458
},
{
"epoch": 0.09705409183215971,
"grad_norm": 0.10050857812166214,
"learning_rate": 0.0004,
"loss": 2.176166296005249,
"step": 460
},
{
"epoch": 0.09747606614447346,
"grad_norm": 0.11539862304925919,
"learning_rate": 0.0004,
"loss": 2.0740013122558594,
"step": 462
},
{
"epoch": 0.0978980404567872,
"grad_norm": 0.10971912741661072,
"learning_rate": 0.0004,
"loss": 2.2293474674224854,
"step": 464
},
{
"epoch": 0.09832001476910093,
"grad_norm": 0.09800975769758224,
"learning_rate": 0.0004,
"loss": 2.1657423973083496,
"step": 466
},
{
"epoch": 0.09874198908141467,
"grad_norm": 0.11508442461490631,
"learning_rate": 0.0004,
"loss": 2.183786392211914,
"step": 468
},
{
"epoch": 0.0991639633937284,
"grad_norm": 0.10498715937137604,
"learning_rate": 0.0004,
"loss": 2.1677026748657227,
"step": 470
},
{
"epoch": 0.09958593770604214,
"grad_norm": 0.11336793750524521,
"learning_rate": 0.0004,
"loss": 2.1494476795196533,
"step": 472
},
{
"epoch": 0.10000791201835588,
"grad_norm": 0.10106341540813446,
"learning_rate": 0.0004,
"loss": 2.117507219314575,
"step": 474
},
{
"epoch": 0.10042988633066963,
"grad_norm": 0.10885264724493027,
"learning_rate": 0.0004,
"loss": 2.1691131591796875,
"step": 476
},
{
"epoch": 0.10085186064298336,
"grad_norm": 0.1028209999203682,
"learning_rate": 0.0004,
"loss": 2.1025424003601074,
"step": 478
},
{
"epoch": 0.1012738349552971,
"grad_norm": 0.10343431681394577,
"learning_rate": 0.0004,
"loss": 2.135477066040039,
"step": 480
},
{
"epoch": 0.10169580926761083,
"grad_norm": 0.10494992882013321,
"learning_rate": 0.0004,
"loss": 2.0871782302856445,
"step": 482
},
{
"epoch": 0.10211778357992457,
"grad_norm": 0.10521426051855087,
"learning_rate": 0.0004,
"loss": 2.1012284755706787,
"step": 484
},
{
"epoch": 0.1025397578922383,
"grad_norm": 0.09971863776445389,
"learning_rate": 0.0004,
"loss": 2.113525152206421,
"step": 486
},
{
"epoch": 0.10296173220455204,
"grad_norm": 0.09890742599964142,
"learning_rate": 0.0004,
"loss": 2.13696551322937,
"step": 488
},
{
"epoch": 0.10338370651686579,
"grad_norm": 0.1144305095076561,
"learning_rate": 0.0004,
"loss": 2.169257640838623,
"step": 490
},
{
"epoch": 0.10380568082917953,
"grad_norm": 0.10876721888780594,
"learning_rate": 0.0004,
"loss": 2.1465811729431152,
"step": 492
},
{
"epoch": 0.10422765514149326,
"grad_norm": 0.11005362123250961,
"learning_rate": 0.0004,
"loss": 2.1479861736297607,
"step": 494
},
{
"epoch": 0.104649629453807,
"grad_norm": 0.12227214127779007,
"learning_rate": 0.0004,
"loss": 2.1559154987335205,
"step": 496
},
{
"epoch": 0.10507160376612074,
"grad_norm": 0.10492133349180222,
"learning_rate": 0.0004,
"loss": 2.1358585357666016,
"step": 498
},
{
"epoch": 0.10549357807843447,
"grad_norm": 0.12116564810276031,
"learning_rate": 0.0004,
"loss": 2.0748181343078613,
"step": 500
},
{
"epoch": 0.10591555239074821,
"grad_norm": 0.10974402725696564,
"learning_rate": 0.0004,
"loss": 2.1734626293182373,
"step": 502
},
{
"epoch": 0.10633752670306196,
"grad_norm": 0.10158853232860565,
"learning_rate": 0.0004,
"loss": 2.0756964683532715,
"step": 504
},
{
"epoch": 0.1067595010153757,
"grad_norm": 0.11271041631698608,
"learning_rate": 0.0004,
"loss": 2.1280670166015625,
"step": 506
},
{
"epoch": 0.10718147532768943,
"grad_norm": 0.10798516869544983,
"learning_rate": 0.0004,
"loss": 2.144733190536499,
"step": 508
},
{
"epoch": 0.10760344964000317,
"grad_norm": 0.10032257437705994,
"learning_rate": 0.0004,
"loss": 2.12662672996521,
"step": 510
},
{
"epoch": 0.1080254239523169,
"grad_norm": 0.10111194103956223,
"learning_rate": 0.0004,
"loss": 2.130863666534424,
"step": 512
},
{
"epoch": 0.10844739826463064,
"grad_norm": 0.12060201168060303,
"learning_rate": 0.0004,
"loss": 2.115837335586548,
"step": 514
},
{
"epoch": 0.10886937257694437,
"grad_norm": 0.10744975507259369,
"learning_rate": 0.0004,
"loss": 2.115091323852539,
"step": 516
},
{
"epoch": 0.10929134688925812,
"grad_norm": 0.11691953241825104,
"learning_rate": 0.0004,
"loss": 2.1441359519958496,
"step": 518
},
{
"epoch": 0.10971332120157186,
"grad_norm": 0.10967298597097397,
"learning_rate": 0.0004,
"loss": 2.1961259841918945,
"step": 520
},
{
"epoch": 0.1101352955138856,
"grad_norm": 0.11839812248945236,
"learning_rate": 0.0004,
"loss": 2.107771873474121,
"step": 522
},
{
"epoch": 0.11055726982619933,
"grad_norm": 0.10187339782714844,
"learning_rate": 0.0004,
"loss": 2.188927173614502,
"step": 524
},
{
"epoch": 0.11097924413851307,
"grad_norm": 0.09853896498680115,
"learning_rate": 0.0004,
"loss": 2.089794874191284,
"step": 526
},
{
"epoch": 0.1114012184508268,
"grad_norm": 0.10944125056266785,
"learning_rate": 0.0004,
"loss": 2.1138784885406494,
"step": 528
},
{
"epoch": 0.11182319276314054,
"grad_norm": 0.11422785371541977,
"learning_rate": 0.0004,
"loss": 2.141139030456543,
"step": 530
},
{
"epoch": 0.11224516707545429,
"grad_norm": 0.12035656720399857,
"learning_rate": 0.0004,
"loss": 2.063276529312134,
"step": 532
},
{
"epoch": 0.11266714138776802,
"grad_norm": 0.11225082725286484,
"learning_rate": 0.0004,
"loss": 2.1716654300689697,
"step": 534
},
{
"epoch": 0.11308911570008176,
"grad_norm": 0.10329756885766983,
"learning_rate": 0.0004,
"loss": 2.1288933753967285,
"step": 536
},
{
"epoch": 0.1135110900123955,
"grad_norm": 0.10448078066110611,
"learning_rate": 0.0004,
"loss": 2.1369645595550537,
"step": 538
},
{
"epoch": 0.11393306432470923,
"grad_norm": 0.11511237919330597,
"learning_rate": 0.0004,
"loss": 2.1774070262908936,
"step": 540
},
{
"epoch": 0.11435503863702297,
"grad_norm": 0.10099837183952332,
"learning_rate": 0.0004,
"loss": 2.089512825012207,
"step": 542
},
{
"epoch": 0.1147770129493367,
"grad_norm": 0.10087133198976517,
"learning_rate": 0.0004,
"loss": 2.1339664459228516,
"step": 544
},
{
"epoch": 0.11519898726165044,
"grad_norm": 0.10706653445959091,
"learning_rate": 0.0004,
"loss": 2.094083309173584,
"step": 546
},
{
"epoch": 0.11562096157396419,
"grad_norm": 0.10590770840644836,
"learning_rate": 0.0004,
"loss": 2.1993956565856934,
"step": 548
},
{
"epoch": 0.11604293588627793,
"grad_norm": 0.09958631545305252,
"learning_rate": 0.0004,
"loss": 2.1873281002044678,
"step": 550
},
{
"epoch": 0.11646491019859166,
"grad_norm": 0.10155294090509415,
"learning_rate": 0.0004,
"loss": 2.083519458770752,
"step": 552
},
{
"epoch": 0.1168868845109054,
"grad_norm": 0.09406159818172455,
"learning_rate": 0.0004,
"loss": 2.1541647911071777,
"step": 554
},
{
"epoch": 0.11730885882321913,
"grad_norm": 0.11386291682720184,
"learning_rate": 0.0004,
"loss": 2.097377061843872,
"step": 556
},
{
"epoch": 0.11773083313553287,
"grad_norm": 0.12189048528671265,
"learning_rate": 0.0004,
"loss": 2.106928586959839,
"step": 558
},
{
"epoch": 0.1181528074478466,
"grad_norm": 0.10407424718141556,
"learning_rate": 0.0004,
"loss": 2.1374409198760986,
"step": 560
},
{
"epoch": 0.11857478176016036,
"grad_norm": 0.10302145779132843,
"learning_rate": 0.0004,
"loss": 2.1138768196105957,
"step": 562
},
{
"epoch": 0.11899675607247409,
"grad_norm": 0.10530761629343033,
"learning_rate": 0.0004,
"loss": 2.123506546020508,
"step": 564
},
{
"epoch": 0.11941873038478783,
"grad_norm": 0.09889083355665207,
"learning_rate": 0.0004,
"loss": 2.1065826416015625,
"step": 566
},
{
"epoch": 0.11984070469710156,
"grad_norm": 0.12569454312324524,
"learning_rate": 0.0004,
"loss": 2.119546890258789,
"step": 568
},
{
"epoch": 0.1202626790094153,
"grad_norm": 0.12317321449518204,
"learning_rate": 0.0004,
"loss": 2.109205961227417,
"step": 570
},
{
"epoch": 0.12068465332172904,
"grad_norm": 0.10445666313171387,
"learning_rate": 0.0004,
"loss": 2.14752459526062,
"step": 572
},
{
"epoch": 0.12110662763404277,
"grad_norm": 0.11950821429491043,
"learning_rate": 0.0004,
"loss": 2.1294028759002686,
"step": 574
},
{
"epoch": 0.12152860194635652,
"grad_norm": 0.11578522622585297,
"learning_rate": 0.0004,
"loss": 2.149550437927246,
"step": 576
},
{
"epoch": 0.12195057625867026,
"grad_norm": 0.09948468208312988,
"learning_rate": 0.0004,
"loss": 2.1196377277374268,
"step": 578
},
{
"epoch": 0.12237255057098399,
"grad_norm": 0.11028246581554413,
"learning_rate": 0.0004,
"loss": 2.117483377456665,
"step": 580
},
{
"epoch": 0.12279452488329773,
"grad_norm": 0.11241725832223892,
"learning_rate": 0.0004,
"loss": 2.142362356185913,
"step": 582
},
{
"epoch": 0.12321649919561146,
"grad_norm": 0.11857046186923981,
"learning_rate": 0.0004,
"loss": 2.0809926986694336,
"step": 584
},
{
"epoch": 0.1236384735079252,
"grad_norm": 0.11449336260557175,
"learning_rate": 0.0004,
"loss": 2.129199266433716,
"step": 586
},
{
"epoch": 0.12406044782023894,
"grad_norm": 0.10256743431091309,
"learning_rate": 0.0004,
"loss": 2.0935845375061035,
"step": 588
},
{
"epoch": 0.12448242213255269,
"grad_norm": 0.09965559095144272,
"learning_rate": 0.0004,
"loss": 2.1434988975524902,
"step": 590
},
{
"epoch": 0.12490439644486642,
"grad_norm": 0.11713854223489761,
"learning_rate": 0.0004,
"loss": 2.1678171157836914,
"step": 592
},
{
"epoch": 0.12532637075718014,
"grad_norm": 0.10646077990531921,
"learning_rate": 0.0004,
"loss": 2.166114330291748,
"step": 594
},
{
"epoch": 0.1257483450694939,
"grad_norm": 0.1113065630197525,
"learning_rate": 0.0004,
"loss": 2.1068637371063232,
"step": 596
},
{
"epoch": 0.12617031938180764,
"grad_norm": 0.11009582132101059,
"learning_rate": 0.0004,
"loss": 2.090066432952881,
"step": 598
},
{
"epoch": 0.12659229369412137,
"grad_norm": 0.1059311181306839,
"learning_rate": 0.0004,
"loss": 2.101567268371582,
"step": 600
},
{
"epoch": 0.12701426800643512,
"grad_norm": 0.09663299471139908,
"learning_rate": 0.0004,
"loss": 2.0435891151428223,
"step": 602
},
{
"epoch": 0.12743624231874884,
"grad_norm": 0.1101335659623146,
"learning_rate": 0.0004,
"loss": 2.0809757709503174,
"step": 604
},
{
"epoch": 0.1278582166310626,
"grad_norm": 0.1137552261352539,
"learning_rate": 0.0004,
"loss": 2.1570522785186768,
"step": 606
},
{
"epoch": 0.1282801909433763,
"grad_norm": 0.09979414939880371,
"learning_rate": 0.0004,
"loss": 2.148672103881836,
"step": 608
},
{
"epoch": 0.12870216525569006,
"grad_norm": 0.11131460219621658,
"learning_rate": 0.0004,
"loss": 2.0637567043304443,
"step": 610
},
{
"epoch": 0.1291241395680038,
"grad_norm": 0.10011547803878784,
"learning_rate": 0.0004,
"loss": 2.113464832305908,
"step": 612
},
{
"epoch": 0.12954611388031753,
"grad_norm": 0.0972154513001442,
"learning_rate": 0.0004,
"loss": 2.12298321723938,
"step": 614
},
{
"epoch": 0.12996808819263128,
"grad_norm": 0.10826016217470169,
"learning_rate": 0.0004,
"loss": 2.125950574874878,
"step": 616
},
{
"epoch": 0.130390062504945,
"grad_norm": 0.10634735226631165,
"learning_rate": 0.0004,
"loss": 2.074981212615967,
"step": 618
},
{
"epoch": 0.13081203681725875,
"grad_norm": 0.0998908281326294,
"learning_rate": 0.0004,
"loss": 2.117025375366211,
"step": 620
},
{
"epoch": 0.13123401112957248,
"grad_norm": 0.09957288205623627,
"learning_rate": 0.0004,
"loss": 2.1138041019439697,
"step": 622
},
{
"epoch": 0.13165598544188623,
"grad_norm": 0.11162349581718445,
"learning_rate": 0.0004,
"loss": 2.083672285079956,
"step": 624
},
{
"epoch": 0.13207795975419998,
"grad_norm": 0.10052193701267242,
"learning_rate": 0.0004,
"loss": 2.096446990966797,
"step": 626
},
{
"epoch": 0.1324999340665137,
"grad_norm": 0.11923649907112122,
"learning_rate": 0.0004,
"loss": 2.1616833209991455,
"step": 628
},
{
"epoch": 0.13292190837882745,
"grad_norm": 0.10124363005161285,
"learning_rate": 0.0004,
"loss": 2.0721309185028076,
"step": 630
},
{
"epoch": 0.13334388269114117,
"grad_norm": 0.10707041621208191,
"learning_rate": 0.0004,
"loss": 2.01750111579895,
"step": 632
},
{
"epoch": 0.13376585700345492,
"grad_norm": 0.11144565045833588,
"learning_rate": 0.0004,
"loss": 2.02260422706604,
"step": 634
},
{
"epoch": 0.13418783131576864,
"grad_norm": 0.11089108139276505,
"learning_rate": 0.0004,
"loss": 2.1625897884368896,
"step": 636
},
{
"epoch": 0.1346098056280824,
"grad_norm": 0.10813088715076447,
"learning_rate": 0.0004,
"loss": 2.118436098098755,
"step": 638
},
{
"epoch": 0.13503177994039614,
"grad_norm": 0.10602413862943649,
"learning_rate": 0.0004,
"loss": 2.1581308841705322,
"step": 640
},
{
"epoch": 0.13545375425270986,
"grad_norm": 0.1079295352101326,
"learning_rate": 0.0004,
"loss": 2.090075731277466,
"step": 642
},
{
"epoch": 0.1358757285650236,
"grad_norm": 0.10124252736568451,
"learning_rate": 0.0004,
"loss": 2.0788414478302,
"step": 644
},
{
"epoch": 0.13629770287733733,
"grad_norm": 0.11989205330610275,
"learning_rate": 0.0004,
"loss": 2.0845541954040527,
"step": 646
},
{
"epoch": 0.13671967718965108,
"grad_norm": 0.09583327919244766,
"learning_rate": 0.0004,
"loss": 2.060208559036255,
"step": 648
},
{
"epoch": 0.1371416515019648,
"grad_norm": 0.11604829877614975,
"learning_rate": 0.0004,
"loss": 2.110811948776245,
"step": 650
},
{
"epoch": 0.13756362581427856,
"grad_norm": 0.12645196914672852,
"learning_rate": 0.0004,
"loss": 2.1087844371795654,
"step": 652
},
{
"epoch": 0.1379856001265923,
"grad_norm": 0.11102139204740524,
"learning_rate": 0.0004,
"loss": 2.1757702827453613,
"step": 654
},
{
"epoch": 0.13840757443890603,
"grad_norm": 0.11270277202129364,
"learning_rate": 0.0004,
"loss": 2.113804578781128,
"step": 656
},
{
"epoch": 0.13882954875121978,
"grad_norm": 0.10956049710512161,
"learning_rate": 0.0004,
"loss": 2.0794570446014404,
"step": 658
},
{
"epoch": 0.1392515230635335,
"grad_norm": 0.09589512646198273,
"learning_rate": 0.0004,
"loss": 2.1060094833374023,
"step": 660
},
{
"epoch": 0.13967349737584725,
"grad_norm": 0.12028387188911438,
"learning_rate": 0.0004,
"loss": 2.0634701251983643,
"step": 662
},
{
"epoch": 0.14009547168816097,
"grad_norm": 0.10321275144815445,
"learning_rate": 0.0004,
"loss": 2.1140222549438477,
"step": 664
},
{
"epoch": 0.14051744600047472,
"grad_norm": 0.13416993618011475,
"learning_rate": 0.0004,
"loss": 2.1182377338409424,
"step": 666
},
{
"epoch": 0.14093942031278847,
"grad_norm": 0.0967794805765152,
"learning_rate": 0.0004,
"loss": 2.185577392578125,
"step": 668
},
{
"epoch": 0.1413613946251022,
"grad_norm": 0.10812490433454514,
"learning_rate": 0.0004,
"loss": 2.093771457672119,
"step": 670
},
{
"epoch": 0.14178336893741594,
"grad_norm": 0.0973934531211853,
"learning_rate": 0.0004,
"loss": 2.1098852157592773,
"step": 672
},
{
"epoch": 0.14220534324972967,
"grad_norm": 0.09942277520895004,
"learning_rate": 0.0004,
"loss": 2.1654629707336426,
"step": 674
},
{
"epoch": 0.14262731756204342,
"grad_norm": 0.10842843353748322,
"learning_rate": 0.0004,
"loss": 2.089427947998047,
"step": 676
},
{
"epoch": 0.14304929187435714,
"grad_norm": 0.10706840455532074,
"learning_rate": 0.0004,
"loss": 2.144439697265625,
"step": 678
},
{
"epoch": 0.1434712661866709,
"grad_norm": 0.10574708878993988,
"learning_rate": 0.0004,
"loss": 2.1588997840881348,
"step": 680
},
{
"epoch": 0.14389324049898464,
"grad_norm": 0.11435360461473465,
"learning_rate": 0.0004,
"loss": 2.1252830028533936,
"step": 682
},
{
"epoch": 0.14431521481129836,
"grad_norm": 0.09806880354881287,
"learning_rate": 0.0004,
"loss": 2.1206793785095215,
"step": 684
},
{
"epoch": 0.1447371891236121,
"grad_norm": 0.10450348258018494,
"learning_rate": 0.0004,
"loss": 2.0693018436431885,
"step": 686
},
{
"epoch": 0.14515916343592583,
"grad_norm": 0.09918571263551712,
"learning_rate": 0.0004,
"loss": 2.066645860671997,
"step": 688
},
{
"epoch": 0.14558113774823958,
"grad_norm": 0.10964078456163406,
"learning_rate": 0.0004,
"loss": 2.1329102516174316,
"step": 690
},
{
"epoch": 0.1460031120605533,
"grad_norm": 0.1003369390964508,
"learning_rate": 0.0004,
"loss": 2.1030282974243164,
"step": 692
},
{
"epoch": 0.14642508637286705,
"grad_norm": 0.09732041507959366,
"learning_rate": 0.0004,
"loss": 2.105342388153076,
"step": 694
},
{
"epoch": 0.1468470606851808,
"grad_norm": 0.11758318543434143,
"learning_rate": 0.0004,
"loss": 2.0845322608947754,
"step": 696
},
{
"epoch": 0.14726903499749452,
"grad_norm": 0.1175965890288353,
"learning_rate": 0.0004,
"loss": 2.1936488151550293,
"step": 698
},
{
"epoch": 0.14769100930980827,
"grad_norm": 0.11464764922857285,
"learning_rate": 0.0004,
"loss": 2.1604058742523193,
"step": 700
},
{
"epoch": 0.148112983622122,
"grad_norm": 0.10833900421857834,
"learning_rate": 0.0004,
"loss": 2.03999662399292,
"step": 702
},
{
"epoch": 0.14853495793443575,
"grad_norm": 0.10867384821176529,
"learning_rate": 0.0004,
"loss": 2.105238676071167,
"step": 704
},
{
"epoch": 0.14895693224674947,
"grad_norm": 0.10172431170940399,
"learning_rate": 0.0004,
"loss": 2.0846638679504395,
"step": 706
},
{
"epoch": 0.14937890655906322,
"grad_norm": 0.10123492032289505,
"learning_rate": 0.0004,
"loss": 2.123725414276123,
"step": 708
},
{
"epoch": 0.14980088087137697,
"grad_norm": 0.11058921366930008,
"learning_rate": 0.0004,
"loss": 1.9794389009475708,
"step": 710
},
{
"epoch": 0.1502228551836907,
"grad_norm": 0.09867732971906662,
"learning_rate": 0.0004,
"loss": 2.041548013687134,
"step": 712
},
{
"epoch": 0.15064482949600444,
"grad_norm": 0.13508032262325287,
"learning_rate": 0.0004,
"loss": 2.0762624740600586,
"step": 714
},
{
"epoch": 0.15106680380831816,
"grad_norm": 0.1391245722770691,
"learning_rate": 0.0004,
"loss": 2.135963201522827,
"step": 716
},
{
"epoch": 0.1514887781206319,
"grad_norm": 0.11377518624067307,
"learning_rate": 0.0004,
"loss": 2.0835230350494385,
"step": 718
},
{
"epoch": 0.15191075243294563,
"grad_norm": 0.10214204341173172,
"learning_rate": 0.0004,
"loss": 2.0786423683166504,
"step": 720
},
{
"epoch": 0.15233272674525938,
"grad_norm": 0.12621009349822998,
"learning_rate": 0.0004,
"loss": 2.073246955871582,
"step": 722
},
{
"epoch": 0.15275470105757313,
"grad_norm": 0.10633029043674469,
"learning_rate": 0.0004,
"loss": 2.117952346801758,
"step": 724
},
{
"epoch": 0.15317667536988686,
"grad_norm": 0.0995170921087265,
"learning_rate": 0.0004,
"loss": 2.107276678085327,
"step": 726
},
{
"epoch": 0.1535986496822006,
"grad_norm": 0.10811508446931839,
"learning_rate": 0.0004,
"loss": 2.111994504928589,
"step": 728
},
{
"epoch": 0.15402062399451433,
"grad_norm": 0.09587360173463821,
"learning_rate": 0.0004,
"loss": 2.092371940612793,
"step": 730
},
{
"epoch": 0.15444259830682808,
"grad_norm": 0.10326235741376877,
"learning_rate": 0.0004,
"loss": 2.1675448417663574,
"step": 732
},
{
"epoch": 0.1548645726191418,
"grad_norm": 0.1008804440498352,
"learning_rate": 0.0004,
"loss": 2.087733745574951,
"step": 734
},
{
"epoch": 0.15528654693145555,
"grad_norm": 0.13531406223773956,
"learning_rate": 0.0004,
"loss": 2.027266502380371,
"step": 736
},
{
"epoch": 0.1557085212437693,
"grad_norm": 0.10144936293363571,
"learning_rate": 0.0004,
"loss": 2.0607857704162598,
"step": 738
},
{
"epoch": 0.15613049555608302,
"grad_norm": 0.11017454415559769,
"learning_rate": 0.0004,
"loss": 2.1015186309814453,
"step": 740
},
{
"epoch": 0.15655246986839677,
"grad_norm": 0.09735028445720673,
"learning_rate": 0.0004,
"loss": 2.1204636096954346,
"step": 742
},
{
"epoch": 0.1569744441807105,
"grad_norm": 0.10278435051441193,
"learning_rate": 0.0004,
"loss": 2.0729727745056152,
"step": 744
},
{
"epoch": 0.15739641849302424,
"grad_norm": 0.109861359000206,
"learning_rate": 0.0004,
"loss": 2.1217892169952393,
"step": 746
},
{
"epoch": 0.15781839280533796,
"grad_norm": 0.11959517747163773,
"learning_rate": 0.0004,
"loss": 2.0969040393829346,
"step": 748
},
{
"epoch": 0.15824036711765171,
"grad_norm": 0.11523259431123734,
"learning_rate": 0.0004,
"loss": 2.15737247467041,
"step": 750
},
{
"epoch": 0.15866234142996546,
"grad_norm": 0.10795868933200836,
"learning_rate": 0.0004,
"loss": 2.179013967514038,
"step": 752
},
{
"epoch": 0.1590843157422792,
"grad_norm": 0.10769423097372055,
"learning_rate": 0.0004,
"loss": 2.11958909034729,
"step": 754
},
{
"epoch": 0.15950629005459294,
"grad_norm": 0.10231717675924301,
"learning_rate": 0.0004,
"loss": 2.046900510787964,
"step": 756
},
{
"epoch": 0.15992826436690666,
"grad_norm": 0.10377570241689682,
"learning_rate": 0.0004,
"loss": 2.0804035663604736,
"step": 758
},
{
"epoch": 0.1603502386792204,
"grad_norm": 0.09826632589101791,
"learning_rate": 0.0004,
"loss": 2.121006965637207,
"step": 760
},
{
"epoch": 0.16077221299153413,
"grad_norm": 0.10372547805309296,
"learning_rate": 0.0004,
"loss": 2.0499508380889893,
"step": 762
},
{
"epoch": 0.16119418730384788,
"grad_norm": 0.11276647448539734,
"learning_rate": 0.0004,
"loss": 2.0513172149658203,
"step": 764
},
{
"epoch": 0.16161616161616163,
"grad_norm": 0.10348805785179138,
"learning_rate": 0.0004,
"loss": 2.0763437747955322,
"step": 766
},
{
"epoch": 0.16203813592847535,
"grad_norm": 0.10468009859323502,
"learning_rate": 0.0004,
"loss": 2.155933141708374,
"step": 768
},
{
"epoch": 0.1624601102407891,
"grad_norm": 0.1013028621673584,
"learning_rate": 0.0004,
"loss": 2.096817970275879,
"step": 770
},
{
"epoch": 0.16288208455310282,
"grad_norm": 0.0988638624548912,
"learning_rate": 0.0004,
"loss": 2.115628242492676,
"step": 772
},
{
"epoch": 0.16330405886541657,
"grad_norm": 0.10291626304388046,
"learning_rate": 0.0004,
"loss": 2.0738699436187744,
"step": 774
},
{
"epoch": 0.1637260331777303,
"grad_norm": 0.1355632245540619,
"learning_rate": 0.0004,
"loss": 2.110413074493408,
"step": 776
},
{
"epoch": 0.16414800749004405,
"grad_norm": 0.10514026880264282,
"learning_rate": 0.0004,
"loss": 2.044823408126831,
"step": 778
},
{
"epoch": 0.16456998180235777,
"grad_norm": 0.1026536375284195,
"learning_rate": 0.0004,
"loss": 2.0761866569519043,
"step": 780
},
{
"epoch": 0.16499195611467152,
"grad_norm": 0.10273230075836182,
"learning_rate": 0.0004,
"loss": 2.155762195587158,
"step": 782
},
{
"epoch": 0.16541393042698527,
"grad_norm": 0.1126902773976326,
"learning_rate": 0.0004,
"loss": 2.115461587905884,
"step": 784
},
{
"epoch": 0.165835904739299,
"grad_norm": 0.10173066705465317,
"learning_rate": 0.0004,
"loss": 2.0853469371795654,
"step": 786
},
{
"epoch": 0.16625787905161274,
"grad_norm": 0.10205572098493576,
"learning_rate": 0.0004,
"loss": 2.0874886512756348,
"step": 788
},
{
"epoch": 0.16667985336392646,
"grad_norm": 0.0917569249868393,
"learning_rate": 0.0004,
"loss": 2.132352590560913,
"step": 790
},
{
"epoch": 0.1671018276762402,
"grad_norm": 0.10074055939912796,
"learning_rate": 0.0004,
"loss": 2.084622621536255,
"step": 792
},
{
"epoch": 0.16752380198855393,
"grad_norm": 0.1239466592669487,
"learning_rate": 0.0004,
"loss": 2.1048314571380615,
"step": 794
},
{
"epoch": 0.16794577630086768,
"grad_norm": 0.09913384169340134,
"learning_rate": 0.0004,
"loss": 2.0674514770507812,
"step": 796
},
{
"epoch": 0.16836775061318143,
"grad_norm": 0.1087324395775795,
"learning_rate": 0.0004,
"loss": 2.117096424102783,
"step": 798
},
{
"epoch": 0.16878972492549515,
"grad_norm": 0.10408477485179901,
"learning_rate": 0.0004,
"loss": 2.050633668899536,
"step": 800
},
{
"epoch": 0.1692116992378089,
"grad_norm": 0.1000882238149643,
"learning_rate": 0.0004,
"loss": 2.1600115299224854,
"step": 802
},
{
"epoch": 0.16963367355012263,
"grad_norm": 0.10445057600736618,
"learning_rate": 0.0004,
"loss": 2.107898712158203,
"step": 804
},
{
"epoch": 0.17005564786243638,
"grad_norm": 0.09820706397294998,
"learning_rate": 0.0004,
"loss": 2.1222548484802246,
"step": 806
},
{
"epoch": 0.1704776221747501,
"grad_norm": 0.1043691635131836,
"learning_rate": 0.0004,
"loss": 2.114928722381592,
"step": 808
},
{
"epoch": 0.17089959648706385,
"grad_norm": 0.10397274792194366,
"learning_rate": 0.0004,
"loss": 2.0936038494110107,
"step": 810
},
{
"epoch": 0.1713215707993776,
"grad_norm": 0.10748063027858734,
"learning_rate": 0.0004,
"loss": 2.0815281867980957,
"step": 812
},
{
"epoch": 0.17174354511169132,
"grad_norm": 0.0982804223895073,
"learning_rate": 0.0004,
"loss": 2.0949110984802246,
"step": 814
},
{
"epoch": 0.17216551942400507,
"grad_norm": 0.11106297373771667,
"learning_rate": 0.0004,
"loss": 2.1133642196655273,
"step": 816
},
{
"epoch": 0.1725874937363188,
"grad_norm": 0.10691525787115097,
"learning_rate": 0.0004,
"loss": 2.0815305709838867,
"step": 818
},
{
"epoch": 0.17300946804863254,
"grad_norm": 0.10917850583791733,
"learning_rate": 0.0004,
"loss": 2.0350093841552734,
"step": 820
},
{
"epoch": 0.17343144236094626,
"grad_norm": 0.09457892924547195,
"learning_rate": 0.0004,
"loss": 2.1236209869384766,
"step": 822
},
{
"epoch": 0.17385341667326001,
"grad_norm": 0.10020066797733307,
"learning_rate": 0.0004,
"loss": 2.089968204498291,
"step": 824
},
{
"epoch": 0.17427539098557376,
"grad_norm": 0.10689288377761841,
"learning_rate": 0.0004,
"loss": 2.0820720195770264,
"step": 826
},
{
"epoch": 0.17469736529788749,
"grad_norm": 0.12644343078136444,
"learning_rate": 0.0004,
"loss": 2.084280490875244,
"step": 828
},
{
"epoch": 0.17511933961020124,
"grad_norm": 0.10654092580080032,
"learning_rate": 0.0004,
"loss": 2.0762181282043457,
"step": 830
},
{
"epoch": 0.17554131392251496,
"grad_norm": 0.1107269749045372,
"learning_rate": 0.0004,
"loss": 2.0836570262908936,
"step": 832
},
{
"epoch": 0.1759632882348287,
"grad_norm": 0.09906051307916641,
"learning_rate": 0.0004,
"loss": 2.081645965576172,
"step": 834
},
{
"epoch": 0.17638526254714243,
"grad_norm": 0.10999570041894913,
"learning_rate": 0.0004,
"loss": 2.0746512413024902,
"step": 836
},
{
"epoch": 0.17680723685945618,
"grad_norm": 0.09614890068769455,
"learning_rate": 0.0004,
"loss": 2.0921099185943604,
"step": 838
},
{
"epoch": 0.17722921117176993,
"grad_norm": 0.10425500571727753,
"learning_rate": 0.0004,
"loss": 2.1366209983825684,
"step": 840
},
{
"epoch": 0.17765118548408365,
"grad_norm": 0.1360170692205429,
"learning_rate": 0.0004,
"loss": 2.038316249847412,
"step": 842
},
{
"epoch": 0.1780731597963974,
"grad_norm": 0.1053030714392662,
"learning_rate": 0.0004,
"loss": 2.1270430088043213,
"step": 844
},
{
"epoch": 0.17849513410871112,
"grad_norm": 0.1127590462565422,
"learning_rate": 0.0004,
"loss": 2.0368661880493164,
"step": 846
},
{
"epoch": 0.17891710842102487,
"grad_norm": 0.10196477174758911,
"learning_rate": 0.0004,
"loss": 2.0663537979125977,
"step": 848
},
{
"epoch": 0.1793390827333386,
"grad_norm": 0.10385768115520477,
"learning_rate": 0.0004,
"loss": 2.1200766563415527,
"step": 850
},
{
"epoch": 0.17976105704565234,
"grad_norm": 0.10235826671123505,
"learning_rate": 0.0004,
"loss": 2.099470853805542,
"step": 852
},
{
"epoch": 0.1801830313579661,
"grad_norm": 0.10381414741277695,
"learning_rate": 0.0004,
"loss": 2.154128074645996,
"step": 854
},
{
"epoch": 0.18060500567027982,
"grad_norm": 0.09879153221845627,
"learning_rate": 0.0004,
"loss": 2.0719563961029053,
"step": 856
},
{
"epoch": 0.18102697998259357,
"grad_norm": 0.11098369210958481,
"learning_rate": 0.0004,
"loss": 2.063169002532959,
"step": 858
},
{
"epoch": 0.1814489542949073,
"grad_norm": 0.10949750989675522,
"learning_rate": 0.0004,
"loss": 2.1000149250030518,
"step": 860
},
{
"epoch": 0.18187092860722104,
"grad_norm": 0.10339707136154175,
"learning_rate": 0.0004,
"loss": 2.089691400527954,
"step": 862
},
{
"epoch": 0.18229290291953476,
"grad_norm": 0.11655983328819275,
"learning_rate": 0.0004,
"loss": 2.0740272998809814,
"step": 864
},
{
"epoch": 0.1827148772318485,
"grad_norm": 0.10416608303785324,
"learning_rate": 0.0004,
"loss": 2.024918794631958,
"step": 866
},
{
"epoch": 0.18313685154416226,
"grad_norm": 0.10420812666416168,
"learning_rate": 0.0004,
"loss": 2.0802533626556396,
"step": 868
},
{
"epoch": 0.18355882585647598,
"grad_norm": 0.1012798622250557,
"learning_rate": 0.0004,
"loss": 2.0436084270477295,
"step": 870
},
{
"epoch": 0.18398080016878973,
"grad_norm": 0.10507721453905106,
"learning_rate": 0.0004,
"loss": 2.0282247066497803,
"step": 872
},
{
"epoch": 0.18440277448110345,
"grad_norm": 0.10625720024108887,
"learning_rate": 0.0004,
"loss": 2.0734364986419678,
"step": 874
},
{
"epoch": 0.1848247487934172,
"grad_norm": 0.121348537504673,
"learning_rate": 0.0004,
"loss": 1.9815733432769775,
"step": 876
},
{
"epoch": 0.18524672310573093,
"grad_norm": 0.11806593835353851,
"learning_rate": 0.0004,
"loss": 2.03564453125,
"step": 878
},
{
"epoch": 0.18566869741804468,
"grad_norm": 0.10645047575235367,
"learning_rate": 0.0004,
"loss": 2.051527500152588,
"step": 880
},
{
"epoch": 0.18609067173035843,
"grad_norm": 0.09987389296293259,
"learning_rate": 0.0004,
"loss": 2.1459152698516846,
"step": 882
},
{
"epoch": 0.18651264604267215,
"grad_norm": 0.11512590199708939,
"learning_rate": 0.0004,
"loss": 2.042651414871216,
"step": 884
},
{
"epoch": 0.1869346203549859,
"grad_norm": 0.10150270909070969,
"learning_rate": 0.0004,
"loss": 2.062952995300293,
"step": 886
},
{
"epoch": 0.18735659466729962,
"grad_norm": 0.09909050911664963,
"learning_rate": 0.0004,
"loss": 2.032675266265869,
"step": 888
},
{
"epoch": 0.18777856897961337,
"grad_norm": 0.11986971646547318,
"learning_rate": 0.0004,
"loss": 2.066922426223755,
"step": 890
},
{
"epoch": 0.1882005432919271,
"grad_norm": 0.11826105415821075,
"learning_rate": 0.0004,
"loss": 2.068296194076538,
"step": 892
},
{
"epoch": 0.18862251760424084,
"grad_norm": 0.098719023168087,
"learning_rate": 0.0004,
"loss": 2.1250462532043457,
"step": 894
},
{
"epoch": 0.1890444919165546,
"grad_norm": 0.10234636813402176,
"learning_rate": 0.0004,
"loss": 2.03814697265625,
"step": 896
},
{
"epoch": 0.1894664662288683,
"grad_norm": 0.10218203812837601,
"learning_rate": 0.0004,
"loss": 2.0698814392089844,
"step": 898
},
{
"epoch": 0.18988844054118206,
"grad_norm": 0.09778252243995667,
"learning_rate": 0.0004,
"loss": 2.104174852371216,
"step": 900
},
{
"epoch": 0.19031041485349579,
"grad_norm": 0.10931072384119034,
"learning_rate": 0.0004,
"loss": 2.020155429840088,
"step": 902
},
{
"epoch": 0.19073238916580953,
"grad_norm": 0.09138067066669464,
"learning_rate": 0.0004,
"loss": 2.0176219940185547,
"step": 904
},
{
"epoch": 0.19115436347812326,
"grad_norm": 0.11851845681667328,
"learning_rate": 0.0004,
"loss": 2.053734302520752,
"step": 906
},
{
"epoch": 0.191576337790437,
"grad_norm": 0.11843250691890717,
"learning_rate": 0.0004,
"loss": 2.0874314308166504,
"step": 908
},
{
"epoch": 0.19199831210275076,
"grad_norm": 0.1034797951579094,
"learning_rate": 0.0004,
"loss": 2.0914268493652344,
"step": 910
},
{
"epoch": 0.19242028641506448,
"grad_norm": 0.0955357626080513,
"learning_rate": 0.0004,
"loss": 2.104736566543579,
"step": 912
},
{
"epoch": 0.19284226072737823,
"grad_norm": 0.10485327988862991,
"learning_rate": 0.0004,
"loss": 2.060455799102783,
"step": 914
},
{
"epoch": 0.19326423503969195,
"grad_norm": 0.11095664650201797,
"learning_rate": 0.0004,
"loss": 2.0719492435455322,
"step": 916
},
{
"epoch": 0.1936862093520057,
"grad_norm": 0.10153471678495407,
"learning_rate": 0.0004,
"loss": 2.09536075592041,
"step": 918
},
{
"epoch": 0.19410818366431942,
"grad_norm": 0.10129594802856445,
"learning_rate": 0.0004,
"loss": 2.083777904510498,
"step": 920
},
{
"epoch": 0.19453015797663317,
"grad_norm": 0.10285522043704987,
"learning_rate": 0.0004,
"loss": 2.056734085083008,
"step": 922
},
{
"epoch": 0.19495213228894692,
"grad_norm": 0.10165376961231232,
"learning_rate": 0.0004,
"loss": 2.1503188610076904,
"step": 924
},
{
"epoch": 0.19537410660126064,
"grad_norm": 0.10450077801942825,
"learning_rate": 0.0004,
"loss": 2.0430591106414795,
"step": 926
},
{
"epoch": 0.1957960809135744,
"grad_norm": 0.09593604505062103,
"learning_rate": 0.0004,
"loss": 2.1289496421813965,
"step": 928
},
{
"epoch": 0.19621805522588812,
"grad_norm": 0.10251755267381668,
"learning_rate": 0.0004,
"loss": 2.0544192790985107,
"step": 930
},
{
"epoch": 0.19664002953820187,
"grad_norm": 0.10145176947116852,
"learning_rate": 0.0004,
"loss": 2.095219850540161,
"step": 932
},
{
"epoch": 0.1970620038505156,
"grad_norm": 0.10890334099531174,
"learning_rate": 0.0004,
"loss": 2.0197975635528564,
"step": 934
},
{
"epoch": 0.19748397816282934,
"grad_norm": 0.10958156734704971,
"learning_rate": 0.0004,
"loss": 2.1150569915771484,
"step": 936
},
{
"epoch": 0.1979059524751431,
"grad_norm": 0.10154721140861511,
"learning_rate": 0.0004,
"loss": 2.0210187435150146,
"step": 938
},
{
"epoch": 0.1983279267874568,
"grad_norm": 0.10598000138998032,
"learning_rate": 0.0004,
"loss": 2.1473076343536377,
"step": 940
},
{
"epoch": 0.19874990109977056,
"grad_norm": 0.09814070910215378,
"learning_rate": 0.0004,
"loss": 2.0979013442993164,
"step": 942
},
{
"epoch": 0.19917187541208428,
"grad_norm": 0.09910205751657486,
"learning_rate": 0.0004,
"loss": 1.997559905052185,
"step": 944
},
{
"epoch": 0.19959384972439803,
"grad_norm": 0.11446119844913483,
"learning_rate": 0.0004,
"loss": 2.0604631900787354,
"step": 946
},
{
"epoch": 0.20001582403671175,
"grad_norm": 0.10531014949083328,
"learning_rate": 0.0004,
"loss": 2.0578184127807617,
"step": 948
},
{
"epoch": 0.20001582403671175,
"eval_loss": 2.0671815872192383,
"eval_runtime": 86.8816,
"eval_samples_per_second": 35.266,
"eval_steps_per_second": 4.408,
"step": 948
},
{
"epoch": 0.2004377983490255,
"grad_norm": 0.10325891524553299,
"learning_rate": 0.0004,
"loss": 2.1023595333099365,
"step": 950
},
{
"epoch": 0.20085977266133925,
"grad_norm": 0.10410068929195404,
"learning_rate": 0.0004,
"loss": 2.0740129947662354,
"step": 952
},
{
"epoch": 0.20128174697365298,
"grad_norm": 0.12220326066017151,
"learning_rate": 0.0004,
"loss": 2.0601963996887207,
"step": 954
},
{
"epoch": 0.20170372128596673,
"grad_norm": 0.10639762133359909,
"learning_rate": 0.0004,
"loss": 2.0554206371307373,
"step": 956
},
{
"epoch": 0.20212569559828045,
"grad_norm": 0.09522542357444763,
"learning_rate": 0.0004,
"loss": 2.0102133750915527,
"step": 958
},
{
"epoch": 0.2025476699105942,
"grad_norm": 0.09354773908853531,
"learning_rate": 0.0004,
"loss": 2.132495164871216,
"step": 960
},
{
"epoch": 0.20296964422290792,
"grad_norm": 0.11503658443689346,
"learning_rate": 0.0004,
"loss": 2.106981039047241,
"step": 962
},
{
"epoch": 0.20339161853522167,
"grad_norm": 0.09721265733242035,
"learning_rate": 0.0004,
"loss": 1.9978598356246948,
"step": 964
},
{
"epoch": 0.20381359284753542,
"grad_norm": 0.10172916203737259,
"learning_rate": 0.0004,
"loss": 2.101057291030884,
"step": 966
},
{
"epoch": 0.20423556715984914,
"grad_norm": 0.10961701720952988,
"learning_rate": 0.0004,
"loss": 2.0402886867523193,
"step": 968
},
{
"epoch": 0.2046575414721629,
"grad_norm": 0.09250684082508087,
"learning_rate": 0.0004,
"loss": 2.067052125930786,
"step": 970
},
{
"epoch": 0.2050795157844766,
"grad_norm": 0.09473984688520432,
"learning_rate": 0.0004,
"loss": 2.0596187114715576,
"step": 972
},
{
"epoch": 0.20550149009679036,
"grad_norm": 0.09915156662464142,
"learning_rate": 0.0004,
"loss": 2.0481491088867188,
"step": 974
},
{
"epoch": 0.20592346440910408,
"grad_norm": 0.09621451795101166,
"learning_rate": 0.0004,
"loss": 2.080754280090332,
"step": 976
},
{
"epoch": 0.20634543872141783,
"grad_norm": 0.095055490732193,
"learning_rate": 0.0004,
"loss": 2.0555837154388428,
"step": 978
},
{
"epoch": 0.20676741303373158,
"grad_norm": 0.09834734350442886,
"learning_rate": 0.0004,
"loss": 2.144092559814453,
"step": 980
},
{
"epoch": 0.2071893873460453,
"grad_norm": 0.12396267056465149,
"learning_rate": 0.0004,
"loss": 2.0947763919830322,
"step": 982
},
{
"epoch": 0.20761136165835906,
"grad_norm": 0.13575908541679382,
"learning_rate": 0.0004,
"loss": 2.0995829105377197,
"step": 984
},
{
"epoch": 0.20803333597067278,
"grad_norm": 0.09544304013252258,
"learning_rate": 0.0004,
"loss": 2.0954782962799072,
"step": 986
},
{
"epoch": 0.20845531028298653,
"grad_norm": 0.09818162769079208,
"learning_rate": 0.0004,
"loss": 2.140582323074341,
"step": 988
},
{
"epoch": 0.20887728459530025,
"grad_norm": 0.10510848462581635,
"learning_rate": 0.0004,
"loss": 2.055241584777832,
"step": 990
},
{
"epoch": 0.209299258907614,
"grad_norm": 0.09998326748609543,
"learning_rate": 0.0004,
"loss": 2.1074514389038086,
"step": 992
},
{
"epoch": 0.20972123321992775,
"grad_norm": 0.09891974180936813,
"learning_rate": 0.0004,
"loss": 2.055959463119507,
"step": 994
},
{
"epoch": 0.21014320753224147,
"grad_norm": 0.09861279278993607,
"learning_rate": 0.0004,
"loss": 2.0759928226470947,
"step": 996
},
{
"epoch": 0.21056518184455522,
"grad_norm": 0.10738018900156021,
"learning_rate": 0.0004,
"loss": 2.0847115516662598,
"step": 998
},
{
"epoch": 0.21098715615686894,
"grad_norm": 0.09597820043563843,
"learning_rate": 0.0004,
"loss": 2.091839075088501,
"step": 1000
},
{
"epoch": 0.2114091304691827,
"grad_norm": 0.10140739381313324,
"learning_rate": 0.0004,
"loss": 2.0645995140075684,
"step": 1002
},
{
"epoch": 0.21183110478149642,
"grad_norm": 0.0963931679725647,
"learning_rate": 0.0004,
"loss": 2.0418381690979004,
"step": 1004
},
{
"epoch": 0.21225307909381017,
"grad_norm": 0.10696844011545181,
"learning_rate": 0.0004,
"loss": 2.080503225326538,
"step": 1006
},
{
"epoch": 0.21267505340612392,
"grad_norm": 0.10360875725746155,
"learning_rate": 0.0004,
"loss": 2.0619759559631348,
"step": 1008
},
{
"epoch": 0.21309702771843764,
"grad_norm": 0.09701628983020782,
"learning_rate": 0.0004,
"loss": 2.075305700302124,
"step": 1010
},
{
"epoch": 0.2135190020307514,
"grad_norm": 0.13253048062324524,
"learning_rate": 0.0004,
"loss": 2.059206247329712,
"step": 1012
},
{
"epoch": 0.2139409763430651,
"grad_norm": 0.1022191196680069,
"learning_rate": 0.0004,
"loss": 2.103555679321289,
"step": 1014
},
{
"epoch": 0.21436295065537886,
"grad_norm": 0.12009415775537491,
"learning_rate": 0.0004,
"loss": 2.0632853507995605,
"step": 1016
},
{
"epoch": 0.21478492496769258,
"grad_norm": 0.10627920180559158,
"learning_rate": 0.0004,
"loss": 2.1009135246276855,
"step": 1018
},
{
"epoch": 0.21520689928000633,
"grad_norm": 0.10539852827787399,
"learning_rate": 0.0004,
"loss": 2.074646234512329,
"step": 1020
},
{
"epoch": 0.21562887359232008,
"grad_norm": 0.102335125207901,
"learning_rate": 0.0004,
"loss": 1.9752756357192993,
"step": 1022
},
{
"epoch": 0.2160508479046338,
"grad_norm": 0.10604073852300644,
"learning_rate": 0.0004,
"loss": 2.088203191757202,
"step": 1024
},
{
"epoch": 0.21647282221694755,
"grad_norm": 0.10646755993366241,
"learning_rate": 0.0004,
"loss": 2.1028835773468018,
"step": 1026
},
{
"epoch": 0.21689479652926127,
"grad_norm": 0.10872533172369003,
"learning_rate": 0.0004,
"loss": 2.0974297523498535,
"step": 1028
},
{
"epoch": 0.21731677084157502,
"grad_norm": 0.09953556954860687,
"learning_rate": 0.0004,
"loss": 2.0770680904388428,
"step": 1030
},
{
"epoch": 0.21773874515388875,
"grad_norm": 0.10345607250928879,
"learning_rate": 0.0004,
"loss": 2.0911922454833984,
"step": 1032
},
{
"epoch": 0.2181607194662025,
"grad_norm": 0.10142865031957626,
"learning_rate": 0.0004,
"loss": 2.0636839866638184,
"step": 1034
},
{
"epoch": 0.21858269377851625,
"grad_norm": 0.12310226261615753,
"learning_rate": 0.0004,
"loss": 2.0318033695220947,
"step": 1036
},
{
"epoch": 0.21900466809082997,
"grad_norm": 0.11234645545482635,
"learning_rate": 0.0004,
"loss": 2.037473440170288,
"step": 1038
},
{
"epoch": 0.21942664240314372,
"grad_norm": 0.10977761447429657,
"learning_rate": 0.0004,
"loss": 2.050473213195801,
"step": 1040
},
{
"epoch": 0.21984861671545744,
"grad_norm": 0.10015005618333817,
"learning_rate": 0.0004,
"loss": 2.0523502826690674,
"step": 1042
},
{
"epoch": 0.2202705910277712,
"grad_norm": 0.10433825105428696,
"learning_rate": 0.0004,
"loss": 2.0895588397979736,
"step": 1044
},
{
"epoch": 0.2206925653400849,
"grad_norm": 0.09808175265789032,
"learning_rate": 0.0004,
"loss": 2.0453665256500244,
"step": 1046
},
{
"epoch": 0.22111453965239866,
"grad_norm": 0.10073351114988327,
"learning_rate": 0.0004,
"loss": 2.0840749740600586,
"step": 1048
},
{
"epoch": 0.2215365139647124,
"grad_norm": 0.11345913261175156,
"learning_rate": 0.0004,
"loss": 2.0683348178863525,
"step": 1050
},
{
"epoch": 0.22195848827702613,
"grad_norm": 0.09906921535730362,
"learning_rate": 0.0004,
"loss": 2.1286203861236572,
"step": 1052
},
{
"epoch": 0.22238046258933988,
"grad_norm": 0.1055447906255722,
"learning_rate": 0.0004,
"loss": 2.057116746902466,
"step": 1054
},
{
"epoch": 0.2228024369016536,
"grad_norm": 0.11254426091909409,
"learning_rate": 0.0004,
"loss": 2.048659324645996,
"step": 1056
},
{
"epoch": 0.22322441121396736,
"grad_norm": 0.12048386037349701,
"learning_rate": 0.0004,
"loss": 2.0781071186065674,
"step": 1058
},
{
"epoch": 0.22364638552628108,
"grad_norm": 0.09799184650182724,
"learning_rate": 0.0004,
"loss": 2.08701491355896,
"step": 1060
},
{
"epoch": 0.22406835983859483,
"grad_norm": 0.11438445746898651,
"learning_rate": 0.0004,
"loss": 2.0588808059692383,
"step": 1062
},
{
"epoch": 0.22449033415090858,
"grad_norm": 0.10387099534273148,
"learning_rate": 0.0004,
"loss": 2.08544921875,
"step": 1064
},
{
"epoch": 0.2249123084632223,
"grad_norm": 0.10655742883682251,
"learning_rate": 0.0004,
"loss": 2.041012763977051,
"step": 1066
},
{
"epoch": 0.22533428277553605,
"grad_norm": 0.10069374740123749,
"learning_rate": 0.0004,
"loss": 1.982102394104004,
"step": 1068
},
{
"epoch": 0.22575625708784977,
"grad_norm": 0.1755983978509903,
"learning_rate": 0.0004,
"loss": 2.046351432800293,
"step": 1070
},
{
"epoch": 0.22617823140016352,
"grad_norm": 0.11466466635465622,
"learning_rate": 0.0004,
"loss": 2.0358128547668457,
"step": 1072
},
{
"epoch": 0.22660020571247724,
"grad_norm": 0.11308802664279938,
"learning_rate": 0.0004,
"loss": 2.0717899799346924,
"step": 1074
},
{
"epoch": 0.227022180024791,
"grad_norm": 0.10079872608184814,
"learning_rate": 0.0004,
"loss": 2.1229147911071777,
"step": 1076
},
{
"epoch": 0.22744415433710474,
"grad_norm": 0.1067342758178711,
"learning_rate": 0.0004,
"loss": 2.0142710208892822,
"step": 1078
},
{
"epoch": 0.22786612864941846,
"grad_norm": 0.11104470491409302,
"learning_rate": 0.0004,
"loss": 2.004711389541626,
"step": 1080
},
{
"epoch": 0.22828810296173221,
"grad_norm": 0.10653650760650635,
"learning_rate": 0.0004,
"loss": 2.0473668575286865,
"step": 1082
},
{
"epoch": 0.22871007727404594,
"grad_norm": 0.09132052212953568,
"learning_rate": 0.0004,
"loss": 2.0147132873535156,
"step": 1084
},
{
"epoch": 0.2291320515863597,
"grad_norm": 0.12246864289045334,
"learning_rate": 0.0004,
"loss": 1.9932918548583984,
"step": 1086
},
{
"epoch": 0.2295540258986734,
"grad_norm": 0.10114414989948273,
"learning_rate": 0.0004,
"loss": 2.0464165210723877,
"step": 1088
},
{
"epoch": 0.22997600021098716,
"grad_norm": 0.12049400806427002,
"learning_rate": 0.0004,
"loss": 2.0571815967559814,
"step": 1090
},
{
"epoch": 0.23039797452330088,
"grad_norm": 0.10335483402013779,
"learning_rate": 0.0004,
"loss": 2.0648066997528076,
"step": 1092
},
{
"epoch": 0.23081994883561463,
"grad_norm": 0.10721594840288162,
"learning_rate": 0.0004,
"loss": 2.0379750728607178,
"step": 1094
},
{
"epoch": 0.23124192314792838,
"grad_norm": 0.11306493729352951,
"learning_rate": 0.0004,
"loss": 2.0113027095794678,
"step": 1096
},
{
"epoch": 0.2316638974602421,
"grad_norm": 0.09969770163297653,
"learning_rate": 0.0004,
"loss": 2.0276451110839844,
"step": 1098
},
{
"epoch": 0.23208587177255585,
"grad_norm": 0.09972522407770157,
"learning_rate": 0.0004,
"loss": 2.1196937561035156,
"step": 1100
},
{
"epoch": 0.23250784608486957,
"grad_norm": 0.09788248687982559,
"learning_rate": 0.0004,
"loss": 2.09224796295166,
"step": 1102
},
{
"epoch": 0.23292982039718332,
"grad_norm": 0.1213034987449646,
"learning_rate": 0.0004,
"loss": 2.0813941955566406,
"step": 1104
},
{
"epoch": 0.23335179470949705,
"grad_norm": 0.1194828599691391,
"learning_rate": 0.0004,
"loss": 2.064236640930176,
"step": 1106
},
{
"epoch": 0.2337737690218108,
"grad_norm": 0.10941118001937866,
"learning_rate": 0.0004,
"loss": 2.0620577335357666,
"step": 1108
},
{
"epoch": 0.23419574333412455,
"grad_norm": 0.11109137535095215,
"learning_rate": 0.0004,
"loss": 2.092865228652954,
"step": 1110
},
{
"epoch": 0.23461771764643827,
"grad_norm": 0.10167246311903,
"learning_rate": 0.0004,
"loss": 2.066153049468994,
"step": 1112
},
{
"epoch": 0.23503969195875202,
"grad_norm": 0.09550223499536514,
"learning_rate": 0.0004,
"loss": 2.0813565254211426,
"step": 1114
},
{
"epoch": 0.23546166627106574,
"grad_norm": 0.10247048735618591,
"learning_rate": 0.0004,
"loss": 2.0835540294647217,
"step": 1116
},
{
"epoch": 0.2358836405833795,
"grad_norm": 0.10965420305728912,
"learning_rate": 0.0004,
"loss": 2.056603193283081,
"step": 1118
},
{
"epoch": 0.2363056148956932,
"grad_norm": 0.10589442402124405,
"learning_rate": 0.0004,
"loss": 2.077146291732788,
"step": 1120
},
{
"epoch": 0.23672758920800696,
"grad_norm": 0.11574966460466385,
"learning_rate": 0.0004,
"loss": 2.0568768978118896,
"step": 1122
},
{
"epoch": 0.2371495635203207,
"grad_norm": 0.13056907057762146,
"learning_rate": 0.0004,
"loss": 2.109354019165039,
"step": 1124
},
{
"epoch": 0.23757153783263443,
"grad_norm": 0.10184738039970398,
"learning_rate": 0.0004,
"loss": 2.1134679317474365,
"step": 1126
},
{
"epoch": 0.23799351214494818,
"grad_norm": 0.10987573117017746,
"learning_rate": 0.0004,
"loss": 2.0756723880767822,
"step": 1128
},
{
"epoch": 0.2384154864572619,
"grad_norm": 0.11375586688518524,
"learning_rate": 0.0004,
"loss": 2.1034908294677734,
"step": 1130
},
{
"epoch": 0.23883746076957565,
"grad_norm": 0.09235178679227829,
"learning_rate": 0.0004,
"loss": 2.0417909622192383,
"step": 1132
},
{
"epoch": 0.23925943508188938,
"grad_norm": 0.10113172978162766,
"learning_rate": 0.0004,
"loss": 2.1055049896240234,
"step": 1134
},
{
"epoch": 0.23968140939420313,
"grad_norm": 0.10936136543750763,
"learning_rate": 0.0004,
"loss": 2.0581185817718506,
"step": 1136
},
{
"epoch": 0.24010338370651688,
"grad_norm": 0.10075986385345459,
"learning_rate": 0.0004,
"loss": 2.0408172607421875,
"step": 1138
},
{
"epoch": 0.2405253580188306,
"grad_norm": 0.10835301131010056,
"learning_rate": 0.0004,
"loss": 2.0132951736450195,
"step": 1140
},
{
"epoch": 0.24094733233114435,
"grad_norm": 0.09728336334228516,
"learning_rate": 0.0004,
"loss": 2.0589864253997803,
"step": 1142
},
{
"epoch": 0.24136930664345807,
"grad_norm": 0.11029426753520966,
"learning_rate": 0.0004,
"loss": 2.0645458698272705,
"step": 1144
},
{
"epoch": 0.24179128095577182,
"grad_norm": 0.09583398699760437,
"learning_rate": 0.0004,
"loss": 2.0375349521636963,
"step": 1146
},
{
"epoch": 0.24221325526808554,
"grad_norm": 0.12009010463953018,
"learning_rate": 0.0004,
"loss": 2.092754602432251,
"step": 1148
},
{
"epoch": 0.2426352295803993,
"grad_norm": 0.12068220973014832,
"learning_rate": 0.0004,
"loss": 2.088794469833374,
"step": 1150
},
{
"epoch": 0.24305720389271304,
"grad_norm": 0.10811642557382584,
"learning_rate": 0.0004,
"loss": 2.0643973350524902,
"step": 1152
},
{
"epoch": 0.24347917820502676,
"grad_norm": 0.10594642907381058,
"learning_rate": 0.0004,
"loss": 2.042525053024292,
"step": 1154
},
{
"epoch": 0.2439011525173405,
"grad_norm": 0.10060539841651917,
"learning_rate": 0.0004,
"loss": 2.0604500770568848,
"step": 1156
},
{
"epoch": 0.24432312682965424,
"grad_norm": 0.11867481470108032,
"learning_rate": 0.0004,
"loss": 2.0957038402557373,
"step": 1158
},
{
"epoch": 0.24474510114196799,
"grad_norm": 0.10814383625984192,
"learning_rate": 0.0004,
"loss": 2.115431308746338,
"step": 1160
},
{
"epoch": 0.2451670754542817,
"grad_norm": 0.10752452909946442,
"learning_rate": 0.0004,
"loss": 2.0574448108673096,
"step": 1162
},
{
"epoch": 0.24558904976659546,
"grad_norm": 0.09640331566333771,
"learning_rate": 0.0004,
"loss": 2.0233981609344482,
"step": 1164
},
{
"epoch": 0.2460110240789092,
"grad_norm": 0.0982162356376648,
"learning_rate": 0.0004,
"loss": 2.0687575340270996,
"step": 1166
},
{
"epoch": 0.24643299839122293,
"grad_norm": 0.0957588478922844,
"learning_rate": 0.0004,
"loss": 2.021697998046875,
"step": 1168
},
{
"epoch": 0.24685497270353668,
"grad_norm": 0.11075134575366974,
"learning_rate": 0.0004,
"loss": 2.0079972743988037,
"step": 1170
},
{
"epoch": 0.2472769470158504,
"grad_norm": 0.0999542847275734,
"learning_rate": 0.0004,
"loss": 2.0607051849365234,
"step": 1172
},
{
"epoch": 0.24769892132816415,
"grad_norm": 0.10201392322778702,
"learning_rate": 0.0004,
"loss": 2.036114454269409,
"step": 1174
},
{
"epoch": 0.24812089564047787,
"grad_norm": 0.10176117718219757,
"learning_rate": 0.0004,
"loss": 2.009420156478882,
"step": 1176
},
{
"epoch": 0.24854286995279162,
"grad_norm": 0.10821950435638428,
"learning_rate": 0.0004,
"loss": 2.0584752559661865,
"step": 1178
},
{
"epoch": 0.24896484426510537,
"grad_norm": 0.11679771542549133,
"learning_rate": 0.0004,
"loss": 2.0315864086151123,
"step": 1180
},
{
"epoch": 0.2493868185774191,
"grad_norm": 0.09440344572067261,
"learning_rate": 0.0004,
"loss": 2.0383923053741455,
"step": 1182
},
{
"epoch": 0.24980879288973284,
"grad_norm": 0.10794760286808014,
"learning_rate": 0.0004,
"loss": 2.0885820388793945,
"step": 1184
},
{
"epoch": 0.25023076720204657,
"grad_norm": 0.11708028614521027,
"learning_rate": 0.0004,
"loss": 2.0421855449676514,
"step": 1186
},
{
"epoch": 0.2506527415143603,
"grad_norm": 0.10240083187818527,
"learning_rate": 0.0004,
"loss": 2.0341176986694336,
"step": 1188
},
{
"epoch": 0.25107471582667407,
"grad_norm": 0.10290568321943283,
"learning_rate": 0.0004,
"loss": 2.057615280151367,
"step": 1190
},
{
"epoch": 0.2514966901389878,
"grad_norm": 0.104375921189785,
"learning_rate": 0.0004,
"loss": 2.0583226680755615,
"step": 1192
},
{
"epoch": 0.2519186644513015,
"grad_norm": 0.10543190687894821,
"learning_rate": 0.0004,
"loss": 2.0246315002441406,
"step": 1194
},
{
"epoch": 0.2523406387636153,
"grad_norm": 0.11114701628684998,
"learning_rate": 0.0004,
"loss": 1.991916298866272,
"step": 1196
},
{
"epoch": 0.252762613075929,
"grad_norm": 0.10640053451061249,
"learning_rate": 0.0004,
"loss": 2.030595302581787,
"step": 1198
},
{
"epoch": 0.25318458738824273,
"grad_norm": 0.10556380450725555,
"learning_rate": 0.0004,
"loss": 2.082707405090332,
"step": 1200
},
{
"epoch": 0.25360656170055645,
"grad_norm": 0.1172170415520668,
"learning_rate": 0.0004,
"loss": 2.1146867275238037,
"step": 1202
},
{
"epoch": 0.25402853601287023,
"grad_norm": 0.1403823047876358,
"learning_rate": 0.0004,
"loss": 2.0604794025421143,
"step": 1204
},
{
"epoch": 0.25445051032518395,
"grad_norm": 0.10991206765174866,
"learning_rate": 0.0004,
"loss": 2.0372889041900635,
"step": 1206
},
{
"epoch": 0.2548724846374977,
"grad_norm": 0.10129861533641815,
"learning_rate": 0.0004,
"loss": 2.0699565410614014,
"step": 1208
},
{
"epoch": 0.25529445894981145,
"grad_norm": 0.09616662561893463,
"learning_rate": 0.0004,
"loss": 2.0831363201141357,
"step": 1210
},
{
"epoch": 0.2557164332621252,
"grad_norm": 0.10269881784915924,
"learning_rate": 0.0004,
"loss": 1.9777147769927979,
"step": 1212
},
{
"epoch": 0.2561384075744389,
"grad_norm": 0.09806791692972183,
"learning_rate": 0.0004,
"loss": 2.0336923599243164,
"step": 1214
},
{
"epoch": 0.2565603818867526,
"grad_norm": 0.10466054081916809,
"learning_rate": 0.0004,
"loss": 1.9841004610061646,
"step": 1216
},
{
"epoch": 0.2569823561990664,
"grad_norm": 0.10455029457807541,
"learning_rate": 0.0004,
"loss": 2.09592604637146,
"step": 1218
},
{
"epoch": 0.2574043305113801,
"grad_norm": 0.09201107174158096,
"learning_rate": 0.0004,
"loss": 2.10310697555542,
"step": 1220
},
{
"epoch": 0.25782630482369384,
"grad_norm": 0.09736589342355728,
"learning_rate": 0.0004,
"loss": 2.0038955211639404,
"step": 1222
},
{
"epoch": 0.2582482791360076,
"grad_norm": 0.09855983406305313,
"learning_rate": 0.0004,
"loss": 2.0687828063964844,
"step": 1224
},
{
"epoch": 0.25867025344832134,
"grad_norm": 0.09814353287220001,
"learning_rate": 0.0004,
"loss": 2.0119292736053467,
"step": 1226
},
{
"epoch": 0.25909222776063506,
"grad_norm": 0.11095716059207916,
"learning_rate": 0.0004,
"loss": 2.0202019214630127,
"step": 1228
},
{
"epoch": 0.2595142020729488,
"grad_norm": 0.11089792847633362,
"learning_rate": 0.0004,
"loss": 2.004791259765625,
"step": 1230
},
{
"epoch": 0.25993617638526256,
"grad_norm": 0.10363394021987915,
"learning_rate": 0.0004,
"loss": 2.0279629230499268,
"step": 1232
},
{
"epoch": 0.2603581506975763,
"grad_norm": 0.10285910964012146,
"learning_rate": 0.0004,
"loss": 2.0563113689422607,
"step": 1234
},
{
"epoch": 0.26078012500989,
"grad_norm": 0.12145205587148666,
"learning_rate": 0.0004,
"loss": 2.0327982902526855,
"step": 1236
},
{
"epoch": 0.2612020993222038,
"grad_norm": 0.0953412726521492,
"learning_rate": 0.0004,
"loss": 2.0196712017059326,
"step": 1238
},
{
"epoch": 0.2616240736345175,
"grad_norm": 0.10441157221794128,
"learning_rate": 0.0004,
"loss": 2.0695459842681885,
"step": 1240
},
{
"epoch": 0.26204604794683123,
"grad_norm": 0.12015742063522339,
"learning_rate": 0.0004,
"loss": 2.0084877014160156,
"step": 1242
},
{
"epoch": 0.26246802225914495,
"grad_norm": 0.14984363317489624,
"learning_rate": 0.0004,
"loss": 2.061483144760132,
"step": 1244
},
{
"epoch": 0.26288999657145873,
"grad_norm": 0.15144675970077515,
"learning_rate": 0.0004,
"loss": 2.0400681495666504,
"step": 1246
},
{
"epoch": 0.26331197088377245,
"grad_norm": 0.10156477987766266,
"learning_rate": 0.0004,
"loss": 2.0286359786987305,
"step": 1248
},
{
"epoch": 0.2637339451960862,
"grad_norm": 0.10266448557376862,
"learning_rate": 0.0004,
"loss": 2.041102170944214,
"step": 1250
},
{
"epoch": 0.26415591950839995,
"grad_norm": 0.10112863034009933,
"learning_rate": 0.0004,
"loss": 2.0677483081817627,
"step": 1252
},
{
"epoch": 0.26457789382071367,
"grad_norm": 0.09947121888399124,
"learning_rate": 0.0004,
"loss": 2.05094051361084,
"step": 1254
},
{
"epoch": 0.2649998681330274,
"grad_norm": 0.0929548367857933,
"learning_rate": 0.0004,
"loss": 2.05031156539917,
"step": 1256
},
{
"epoch": 0.2654218424453411,
"grad_norm": 0.10338550806045532,
"learning_rate": 0.0004,
"loss": 2.060044050216675,
"step": 1258
},
{
"epoch": 0.2658438167576549,
"grad_norm": 0.10694804042577744,
"learning_rate": 0.0004,
"loss": 2.053286075592041,
"step": 1260
},
{
"epoch": 0.2662657910699686,
"grad_norm": 0.10682427883148193,
"learning_rate": 0.0004,
"loss": 2.0644848346710205,
"step": 1262
},
{
"epoch": 0.26668776538228234,
"grad_norm": 0.09777329117059708,
"learning_rate": 0.0004,
"loss": 2.057276725769043,
"step": 1264
},
{
"epoch": 0.2671097396945961,
"grad_norm": 0.11592648178339005,
"learning_rate": 0.0004,
"loss": 2.0669894218444824,
"step": 1266
},
{
"epoch": 0.26753171400690984,
"grad_norm": 0.10248623788356781,
"learning_rate": 0.0004,
"loss": 1.994166612625122,
"step": 1268
},
{
"epoch": 0.26795368831922356,
"grad_norm": 0.09868704527616501,
"learning_rate": 0.0004,
"loss": 2.0670111179351807,
"step": 1270
},
{
"epoch": 0.2683756626315373,
"grad_norm": 0.11977045238018036,
"learning_rate": 0.0004,
"loss": 2.0150840282440186,
"step": 1272
},
{
"epoch": 0.26879763694385106,
"grad_norm": 0.09880789369344711,
"learning_rate": 0.0004,
"loss": 2.04841947555542,
"step": 1274
},
{
"epoch": 0.2692196112561648,
"grad_norm": 0.10139486193656921,
"learning_rate": 0.0004,
"loss": 2.0589547157287598,
"step": 1276
},
{
"epoch": 0.2696415855684785,
"grad_norm": 0.10906106233596802,
"learning_rate": 0.0004,
"loss": 2.010936975479126,
"step": 1278
},
{
"epoch": 0.2700635598807923,
"grad_norm": 0.1240103468298912,
"learning_rate": 0.0004,
"loss": 2.0552594661712646,
"step": 1280
},
{
"epoch": 0.270485534193106,
"grad_norm": 0.10055209696292877,
"learning_rate": 0.0004,
"loss": 2.0927560329437256,
"step": 1282
},
{
"epoch": 0.2709075085054197,
"grad_norm": 0.10815662890672684,
"learning_rate": 0.0004,
"loss": 2.0449116230010986,
"step": 1284
},
{
"epoch": 0.27132948281773345,
"grad_norm": 0.10944771766662598,
"learning_rate": 0.0004,
"loss": 2.0614092350006104,
"step": 1286
},
{
"epoch": 0.2717514571300472,
"grad_norm": 0.1065296083688736,
"learning_rate": 0.0004,
"loss": 2.0241000652313232,
"step": 1288
},
{
"epoch": 0.27217343144236095,
"grad_norm": 0.10042498260736465,
"learning_rate": 0.0004,
"loss": 2.108473777770996,
"step": 1290
},
{
"epoch": 0.27259540575467467,
"grad_norm": 0.09964487701654434,
"learning_rate": 0.0004,
"loss": 2.0980069637298584,
"step": 1292
},
{
"epoch": 0.27301738006698845,
"grad_norm": 0.1096806451678276,
"learning_rate": 0.0004,
"loss": 2.0962815284729004,
"step": 1294
},
{
"epoch": 0.27343935437930217,
"grad_norm": 0.10912879556417465,
"learning_rate": 0.0004,
"loss": 2.0583014488220215,
"step": 1296
},
{
"epoch": 0.2738613286916159,
"grad_norm": 0.10333561152219772,
"learning_rate": 0.0004,
"loss": 1.9904428720474243,
"step": 1298
},
{
"epoch": 0.2742833030039296,
"grad_norm": 0.10392451286315918,
"learning_rate": 0.0004,
"loss": 2.006603956222534,
"step": 1300
},
{
"epoch": 0.2747052773162434,
"grad_norm": 0.1168191209435463,
"learning_rate": 0.0004,
"loss": 2.0531041622161865,
"step": 1302
},
{
"epoch": 0.2751272516285571,
"grad_norm": 0.10119763016700745,
"learning_rate": 0.0004,
"loss": 2.1008636951446533,
"step": 1304
},
{
"epoch": 0.27554922594087083,
"grad_norm": 0.10331698507070541,
"learning_rate": 0.0004,
"loss": 2.029780626296997,
"step": 1306
},
{
"epoch": 0.2759712002531846,
"grad_norm": 0.10079697519540787,
"learning_rate": 0.0004,
"loss": 2.0278022289276123,
"step": 1308
},
{
"epoch": 0.27639317456549833,
"grad_norm": 0.5980570316314697,
"learning_rate": 0.0004,
"loss": 2.0641822814941406,
"step": 1310
},
{
"epoch": 0.27681514887781206,
"grad_norm": 0.11719910055398941,
"learning_rate": 0.0004,
"loss": 2.010359048843384,
"step": 1312
},
{
"epoch": 0.2772371231901258,
"grad_norm": 0.10348856449127197,
"learning_rate": 0.0004,
"loss": 2.030585527420044,
"step": 1314
},
{
"epoch": 0.27765909750243956,
"grad_norm": 0.10512405633926392,
"learning_rate": 0.0004,
"loss": 1.9943598508834839,
"step": 1316
},
{
"epoch": 0.2780810718147533,
"grad_norm": 0.11220675706863403,
"learning_rate": 0.0004,
"loss": 2.081284284591675,
"step": 1318
},
{
"epoch": 0.278503046127067,
"grad_norm": 0.10254770517349243,
"learning_rate": 0.0004,
"loss": 2.02280330657959,
"step": 1320
},
{
"epoch": 0.2789250204393808,
"grad_norm": 0.11460383981466293,
"learning_rate": 0.0004,
"loss": 2.0646212100982666,
"step": 1322
},
{
"epoch": 0.2793469947516945,
"grad_norm": 0.10143101215362549,
"learning_rate": 0.0004,
"loss": 2.0430362224578857,
"step": 1324
},
{
"epoch": 0.2797689690640082,
"grad_norm": 0.10769624263048172,
"learning_rate": 0.0004,
"loss": 2.038762331008911,
"step": 1326
},
{
"epoch": 0.28019094337632194,
"grad_norm": 0.11159703135490417,
"learning_rate": 0.0004,
"loss": 2.0418667793273926,
"step": 1328
},
{
"epoch": 0.2806129176886357,
"grad_norm": 0.11087030172348022,
"learning_rate": 0.0004,
"loss": 1.9600088596343994,
"step": 1330
},
{
"epoch": 0.28103489200094944,
"grad_norm": 0.10535233467817307,
"learning_rate": 0.0004,
"loss": 2.0000345706939697,
"step": 1332
},
{
"epoch": 0.28145686631326317,
"grad_norm": 0.09890623390674591,
"learning_rate": 0.0004,
"loss": 2.100395917892456,
"step": 1334
},
{
"epoch": 0.28187884062557694,
"grad_norm": 0.11283712089061737,
"learning_rate": 0.0004,
"loss": 2.0914347171783447,
"step": 1336
},
{
"epoch": 0.28230081493789067,
"grad_norm": 0.09981510788202286,
"learning_rate": 0.0004,
"loss": 2.036780834197998,
"step": 1338
},
{
"epoch": 0.2827227892502044,
"grad_norm": 0.1112729087471962,
"learning_rate": 0.0004,
"loss": 2.0381102561950684,
"step": 1340
},
{
"epoch": 0.2831447635625181,
"grad_norm": 0.1074172705411911,
"learning_rate": 0.0004,
"loss": 2.0957303047180176,
"step": 1342
},
{
"epoch": 0.2835667378748319,
"grad_norm": 0.09753048419952393,
"learning_rate": 0.0004,
"loss": 2.013277053833008,
"step": 1344
},
{
"epoch": 0.2839887121871456,
"grad_norm": 0.11461804062128067,
"learning_rate": 0.0004,
"loss": 2.107050657272339,
"step": 1346
},
{
"epoch": 0.28441068649945933,
"grad_norm": 0.15218521654605865,
"learning_rate": 0.0004,
"loss": 2.0824432373046875,
"step": 1348
},
{
"epoch": 0.2848326608117731,
"grad_norm": 0.09826374053955078,
"learning_rate": 0.0004,
"loss": 2.0368142127990723,
"step": 1350
},
{
"epoch": 0.28525463512408683,
"grad_norm": 0.1263633519411087,
"learning_rate": 0.0004,
"loss": 2.0246639251708984,
"step": 1352
},
{
"epoch": 0.28567660943640055,
"grad_norm": 0.10714877396821976,
"learning_rate": 0.0004,
"loss": 2.0293967723846436,
"step": 1354
},
{
"epoch": 0.2860985837487143,
"grad_norm": 0.10271991789340973,
"learning_rate": 0.0004,
"loss": 2.110764741897583,
"step": 1356
},
{
"epoch": 0.28652055806102805,
"grad_norm": 0.1031140685081482,
"learning_rate": 0.0004,
"loss": 2.045133590698242,
"step": 1358
},
{
"epoch": 0.2869425323733418,
"grad_norm": 0.11628556251525879,
"learning_rate": 0.0004,
"loss": 2.051297664642334,
"step": 1360
},
{
"epoch": 0.2873645066856555,
"grad_norm": 0.10816340893507004,
"learning_rate": 0.0004,
"loss": 2.0311293601989746,
"step": 1362
},
{
"epoch": 0.2877864809979693,
"grad_norm": 0.11754567921161652,
"learning_rate": 0.0004,
"loss": 1.993975043296814,
"step": 1364
},
{
"epoch": 0.288208455310283,
"grad_norm": 0.10457561165094376,
"learning_rate": 0.0004,
"loss": 2.0523953437805176,
"step": 1366
},
{
"epoch": 0.2886304296225967,
"grad_norm": 0.09719069302082062,
"learning_rate": 0.0004,
"loss": 2.0614097118377686,
"step": 1368
},
{
"epoch": 0.28905240393491044,
"grad_norm": 0.10489129275083542,
"learning_rate": 0.0004,
"loss": 2.0822484493255615,
"step": 1370
},
{
"epoch": 0.2894743782472242,
"grad_norm": 0.11243846267461777,
"learning_rate": 0.0004,
"loss": 2.036621332168579,
"step": 1372
},
{
"epoch": 0.28989635255953794,
"grad_norm": 0.10766646265983582,
"learning_rate": 0.0004,
"loss": 2.0342800617218018,
"step": 1374
},
{
"epoch": 0.29031832687185166,
"grad_norm": 0.10231495648622513,
"learning_rate": 0.0004,
"loss": 2.0997602939605713,
"step": 1376
},
{
"epoch": 0.29074030118416544,
"grad_norm": 0.09925436228513718,
"learning_rate": 0.0004,
"loss": 1.9890317916870117,
"step": 1378
},
{
"epoch": 0.29116227549647916,
"grad_norm": 0.12656085193157196,
"learning_rate": 0.0004,
"loss": 2.054431438446045,
"step": 1380
},
{
"epoch": 0.2915842498087929,
"grad_norm": 0.12750673294067383,
"learning_rate": 0.0004,
"loss": 2.102907657623291,
"step": 1382
},
{
"epoch": 0.2920062241211066,
"grad_norm": 0.11814823746681213,
"learning_rate": 0.0004,
"loss": 2.0192980766296387,
"step": 1384
},
{
"epoch": 0.2924281984334204,
"grad_norm": 0.10131193697452545,
"learning_rate": 0.0004,
"loss": 2.0269525051116943,
"step": 1386
},
{
"epoch": 0.2928501727457341,
"grad_norm": 0.10024850815534592,
"learning_rate": 0.0004,
"loss": 2.0104033946990967,
"step": 1388
},
{
"epoch": 0.2932721470580478,
"grad_norm": 0.11509843170642853,
"learning_rate": 0.0004,
"loss": 1.9751825332641602,
"step": 1390
},
{
"epoch": 0.2936941213703616,
"grad_norm": 0.10131978988647461,
"learning_rate": 0.0004,
"loss": 2.015760898590088,
"step": 1392
},
{
"epoch": 0.2941160956826753,
"grad_norm": 0.10398828983306885,
"learning_rate": 0.0004,
"loss": 2.0820837020874023,
"step": 1394
},
{
"epoch": 0.29453806999498905,
"grad_norm": 0.10714879631996155,
"learning_rate": 0.0004,
"loss": 1.9949901103973389,
"step": 1396
},
{
"epoch": 0.29496004430730277,
"grad_norm": 0.10393649339675903,
"learning_rate": 0.0004,
"loss": 2.0516409873962402,
"step": 1398
},
{
"epoch": 0.29538201861961655,
"grad_norm": 0.1124054566025734,
"learning_rate": 0.0004,
"loss": 1.9853947162628174,
"step": 1400
},
{
"epoch": 0.29580399293193027,
"grad_norm": 0.10428536683320999,
"learning_rate": 0.0004,
"loss": 2.0621979236602783,
"step": 1402
},
{
"epoch": 0.296225967244244,
"grad_norm": 0.10993221402168274,
"learning_rate": 0.0004,
"loss": 2.0149316787719727,
"step": 1404
},
{
"epoch": 0.29664794155655777,
"grad_norm": 0.10147418826818466,
"learning_rate": 0.0004,
"loss": 2.0695698261260986,
"step": 1406
},
{
"epoch": 0.2970699158688715,
"grad_norm": 0.11199607700109482,
"learning_rate": 0.0004,
"loss": 2.016078472137451,
"step": 1408
},
{
"epoch": 0.2974918901811852,
"grad_norm": 0.09972609579563141,
"learning_rate": 0.0004,
"loss": 2.047872543334961,
"step": 1410
},
{
"epoch": 0.29791386449349894,
"grad_norm": 0.10419731587171555,
"learning_rate": 0.0004,
"loss": 2.061795234680176,
"step": 1412
},
{
"epoch": 0.2983358388058127,
"grad_norm": 0.10370040684938431,
"learning_rate": 0.0004,
"loss": 1.9889763593673706,
"step": 1414
},
{
"epoch": 0.29875781311812644,
"grad_norm": 0.11165943741798401,
"learning_rate": 0.0004,
"loss": 1.9462987184524536,
"step": 1416
},
{
"epoch": 0.29917978743044016,
"grad_norm": 0.09943879395723343,
"learning_rate": 0.0004,
"loss": 2.010300636291504,
"step": 1418
},
{
"epoch": 0.29960176174275394,
"grad_norm": 0.11090060323476791,
"learning_rate": 0.0004,
"loss": 2.0240564346313477,
"step": 1420
},
{
"epoch": 0.30002373605506766,
"grad_norm": 0.11876624822616577,
"learning_rate": 0.0004,
"loss": 2.052326202392578,
"step": 1422
},
{
"epoch": 0.3004457103673814,
"grad_norm": 0.09885898232460022,
"learning_rate": 0.0004,
"loss": 2.066755533218384,
"step": 1424
},
{
"epoch": 0.3008676846796951,
"grad_norm": 0.1066424623131752,
"learning_rate": 0.0004,
"loss": 1.9553253650665283,
"step": 1426
},
{
"epoch": 0.3012896589920089,
"grad_norm": 0.10109975934028625,
"learning_rate": 0.0004,
"loss": 1.958069086074829,
"step": 1428
},
{
"epoch": 0.3017116333043226,
"grad_norm": 0.10391309857368469,
"learning_rate": 0.0004,
"loss": 2.0515308380126953,
"step": 1430
},
{
"epoch": 0.3021336076166363,
"grad_norm": 0.11120624095201492,
"learning_rate": 0.0004,
"loss": 2.0212414264678955,
"step": 1432
},
{
"epoch": 0.3025555819289501,
"grad_norm": 0.10360627621412277,
"learning_rate": 0.0004,
"loss": 2.0257387161254883,
"step": 1434
},
{
"epoch": 0.3029775562412638,
"grad_norm": 0.11205482482910156,
"learning_rate": 0.0004,
"loss": 2.055927276611328,
"step": 1436
},
{
"epoch": 0.30339953055357755,
"grad_norm": 0.10705963522195816,
"learning_rate": 0.0004,
"loss": 2.103196859359741,
"step": 1438
},
{
"epoch": 0.30382150486589127,
"grad_norm": 0.10121170431375504,
"learning_rate": 0.0004,
"loss": 1.98408842086792,
"step": 1440
},
{
"epoch": 0.30424347917820505,
"grad_norm": 0.10510347038507462,
"learning_rate": 0.0004,
"loss": 2.0021445751190186,
"step": 1442
},
{
"epoch": 0.30466545349051877,
"grad_norm": 0.09092416614294052,
"learning_rate": 0.0004,
"loss": 2.0650837421417236,
"step": 1444
},
{
"epoch": 0.3050874278028325,
"grad_norm": 0.12853112816810608,
"learning_rate": 0.0004,
"loss": 2.0214555263519287,
"step": 1446
},
{
"epoch": 0.30550940211514627,
"grad_norm": 0.12484022974967957,
"learning_rate": 0.0004,
"loss": 2.004979133605957,
"step": 1448
},
{
"epoch": 0.30593137642746,
"grad_norm": 0.10841439664363861,
"learning_rate": 0.0004,
"loss": 2.0770885944366455,
"step": 1450
},
{
"epoch": 0.3063533507397737,
"grad_norm": 0.10934076458215714,
"learning_rate": 0.0004,
"loss": 2.0190446376800537,
"step": 1452
},
{
"epoch": 0.30677532505208743,
"grad_norm": 0.10753870755434036,
"learning_rate": 0.0004,
"loss": 2.0732760429382324,
"step": 1454
},
{
"epoch": 0.3071972993644012,
"grad_norm": 0.10213645547628403,
"learning_rate": 0.0004,
"loss": 2.0003771781921387,
"step": 1456
},
{
"epoch": 0.30761927367671493,
"grad_norm": 0.10572090744972229,
"learning_rate": 0.0004,
"loss": 2.0285959243774414,
"step": 1458
},
{
"epoch": 0.30804124798902865,
"grad_norm": 0.10494815558195114,
"learning_rate": 0.0004,
"loss": 1.998738408088684,
"step": 1460
},
{
"epoch": 0.30846322230134243,
"grad_norm": 0.10957774519920349,
"learning_rate": 0.0004,
"loss": 2.0375754833221436,
"step": 1462
},
{
"epoch": 0.30888519661365615,
"grad_norm": 0.11592728644609451,
"learning_rate": 0.0004,
"loss": 2.016571283340454,
"step": 1464
},
{
"epoch": 0.3093071709259699,
"grad_norm": 0.1006048321723938,
"learning_rate": 0.0004,
"loss": 2.045219659805298,
"step": 1466
},
{
"epoch": 0.3097291452382836,
"grad_norm": 0.12066938728094101,
"learning_rate": 0.0004,
"loss": 2.0009267330169678,
"step": 1468
},
{
"epoch": 0.3101511195505974,
"grad_norm": 0.10149630904197693,
"learning_rate": 0.0004,
"loss": 2.0407376289367676,
"step": 1470
},
{
"epoch": 0.3105730938629111,
"grad_norm": 0.10064069926738739,
"learning_rate": 0.0004,
"loss": 2.0730724334716797,
"step": 1472
},
{
"epoch": 0.3109950681752248,
"grad_norm": 0.10382674634456635,
"learning_rate": 0.0004,
"loss": 2.1180226802825928,
"step": 1474
},
{
"epoch": 0.3114170424875386,
"grad_norm": 0.1059107705950737,
"learning_rate": 0.0004,
"loss": 2.004983901977539,
"step": 1476
},
{
"epoch": 0.3118390167998523,
"grad_norm": 0.09575723856687546,
"learning_rate": 0.0004,
"loss": 2.0194239616394043,
"step": 1478
},
{
"epoch": 0.31226099111216604,
"grad_norm": 0.1048755794763565,
"learning_rate": 0.0004,
"loss": 1.9876059293746948,
"step": 1480
},
{
"epoch": 0.31268296542447976,
"grad_norm": 0.11005643755197525,
"learning_rate": 0.0004,
"loss": 2.0378410816192627,
"step": 1482
},
{
"epoch": 0.31310493973679354,
"grad_norm": 0.09710226207971573,
"learning_rate": 0.0004,
"loss": 2.055088758468628,
"step": 1484
},
{
"epoch": 0.31352691404910726,
"grad_norm": 0.11550967395305634,
"learning_rate": 0.0004,
"loss": 2.045725107192993,
"step": 1486
},
{
"epoch": 0.313948888361421,
"grad_norm": 0.10504131019115448,
"learning_rate": 0.0004,
"loss": 2.017702102661133,
"step": 1488
},
{
"epoch": 0.31437086267373476,
"grad_norm": 0.10367326438426971,
"learning_rate": 0.0004,
"loss": 2.0419342517852783,
"step": 1490
},
{
"epoch": 0.3147928369860485,
"grad_norm": 0.11854102462530136,
"learning_rate": 0.0004,
"loss": 2.0199780464172363,
"step": 1492
},
{
"epoch": 0.3152148112983622,
"grad_norm": 0.09546536207199097,
"learning_rate": 0.0004,
"loss": 2.0593788623809814,
"step": 1494
},
{
"epoch": 0.31563678561067593,
"grad_norm": 0.09722495824098587,
"learning_rate": 0.0004,
"loss": 2.0179624557495117,
"step": 1496
},
{
"epoch": 0.3160587599229897,
"grad_norm": 0.1031755656003952,
"learning_rate": 0.0004,
"loss": 2.00142502784729,
"step": 1498
},
{
"epoch": 0.31648073423530343,
"grad_norm": 0.10381591320037842,
"learning_rate": 0.0004,
"loss": 2.0483436584472656,
"step": 1500
},
{
"epoch": 0.31690270854761715,
"grad_norm": 0.09587974846363068,
"learning_rate": 0.0004,
"loss": 2.0501937866210938,
"step": 1502
},
{
"epoch": 0.31732468285993093,
"grad_norm": 0.10678374022245407,
"learning_rate": 0.0004,
"loss": 2.0197954177856445,
"step": 1504
},
{
"epoch": 0.31774665717224465,
"grad_norm": 0.10604098439216614,
"learning_rate": 0.0004,
"loss": 1.9633119106292725,
"step": 1506
},
{
"epoch": 0.3181686314845584,
"grad_norm": 0.10567682236433029,
"learning_rate": 0.0004,
"loss": 1.9999604225158691,
"step": 1508
},
{
"epoch": 0.3185906057968721,
"grad_norm": 0.10761748254299164,
"learning_rate": 0.0004,
"loss": 2.028287887573242,
"step": 1510
},
{
"epoch": 0.3190125801091859,
"grad_norm": 0.1289740353822708,
"learning_rate": 0.0004,
"loss": 1.997977375984192,
"step": 1512
},
{
"epoch": 0.3194345544214996,
"grad_norm": 0.10368921607732773,
"learning_rate": 0.0004,
"loss": 2.0591623783111572,
"step": 1514
},
{
"epoch": 0.3198565287338133,
"grad_norm": 0.09553994983434677,
"learning_rate": 0.0004,
"loss": 2.0922186374664307,
"step": 1516
},
{
"epoch": 0.3202785030461271,
"grad_norm": 0.10956684499979019,
"learning_rate": 0.0004,
"loss": 2.0688552856445312,
"step": 1518
},
{
"epoch": 0.3207004773584408,
"grad_norm": 0.10936163365840912,
"learning_rate": 0.0004,
"loss": 2.0323116779327393,
"step": 1520
},
{
"epoch": 0.32112245167075454,
"grad_norm": 0.10848774760961533,
"learning_rate": 0.0004,
"loss": 1.9892557859420776,
"step": 1522
},
{
"epoch": 0.32154442598306826,
"grad_norm": 0.11885710060596466,
"learning_rate": 0.0004,
"loss": 2.02767276763916,
"step": 1524
},
{
"epoch": 0.32196640029538204,
"grad_norm": 0.11445765197277069,
"learning_rate": 0.0004,
"loss": 2.0190091133117676,
"step": 1526
},
{
"epoch": 0.32238837460769576,
"grad_norm": 0.10351983457803726,
"learning_rate": 0.0004,
"loss": 2.043550729751587,
"step": 1528
},
{
"epoch": 0.3228103489200095,
"grad_norm": 0.09981498122215271,
"learning_rate": 0.0004,
"loss": 1.982284426689148,
"step": 1530
},
{
"epoch": 0.32323232323232326,
"grad_norm": 0.10223311185836792,
"learning_rate": 0.0004,
"loss": 2.0192668437957764,
"step": 1532
},
{
"epoch": 0.323654297544637,
"grad_norm": 0.09700711816549301,
"learning_rate": 0.0004,
"loss": 2.0648601055145264,
"step": 1534
},
{
"epoch": 0.3240762718569507,
"grad_norm": 0.10281526297330856,
"learning_rate": 0.0004,
"loss": 2.018752098083496,
"step": 1536
},
{
"epoch": 0.3244982461692644,
"grad_norm": 0.10738297551870346,
"learning_rate": 0.0004,
"loss": 2.0688846111297607,
"step": 1538
},
{
"epoch": 0.3249202204815782,
"grad_norm": 0.10620314627885818,
"learning_rate": 0.0004,
"loss": 2.0286612510681152,
"step": 1540
},
{
"epoch": 0.3253421947938919,
"grad_norm": 0.12181965261697769,
"learning_rate": 0.0004,
"loss": 1.991784691810608,
"step": 1542
},
{
"epoch": 0.32576416910620565,
"grad_norm": 0.1114184558391571,
"learning_rate": 0.0004,
"loss": 1.9791942834854126,
"step": 1544
},
{
"epoch": 0.3261861434185194,
"grad_norm": 0.10929999500513077,
"learning_rate": 0.0004,
"loss": 2.0243911743164062,
"step": 1546
},
{
"epoch": 0.32660811773083315,
"grad_norm": 0.1039823517203331,
"learning_rate": 0.0004,
"loss": 2.0071513652801514,
"step": 1548
},
{
"epoch": 0.32703009204314687,
"grad_norm": 0.11235126107931137,
"learning_rate": 0.0004,
"loss": 2.022231101989746,
"step": 1550
},
{
"epoch": 0.3274520663554606,
"grad_norm": 0.1006590723991394,
"learning_rate": 0.0004,
"loss": 2.0311830043792725,
"step": 1552
},
{
"epoch": 0.32787404066777437,
"grad_norm": 0.09724811464548111,
"learning_rate": 0.0004,
"loss": 2.0610275268554688,
"step": 1554
},
{
"epoch": 0.3282960149800881,
"grad_norm": 0.10974770039319992,
"learning_rate": 0.0004,
"loss": 2.0103087425231934,
"step": 1556
},
{
"epoch": 0.3287179892924018,
"grad_norm": 0.11593059450387955,
"learning_rate": 0.0004,
"loss": 1.9464669227600098,
"step": 1558
},
{
"epoch": 0.32913996360471554,
"grad_norm": 0.09999720752239227,
"learning_rate": 0.0004,
"loss": 2.030454397201538,
"step": 1560
},
{
"epoch": 0.3295619379170293,
"grad_norm": 0.11147542297840118,
"learning_rate": 0.0004,
"loss": 2.081592559814453,
"step": 1562
},
{
"epoch": 0.32998391222934303,
"grad_norm": 0.10786650329828262,
"learning_rate": 0.0004,
"loss": 2.0256025791168213,
"step": 1564
},
{
"epoch": 0.33040588654165676,
"grad_norm": 0.1080414354801178,
"learning_rate": 0.0004,
"loss": 1.9759323596954346,
"step": 1566
},
{
"epoch": 0.33082786085397053,
"grad_norm": 0.10195384919643402,
"learning_rate": 0.0004,
"loss": 2.0485634803771973,
"step": 1568
},
{
"epoch": 0.33124983516628426,
"grad_norm": 0.09070245921611786,
"learning_rate": 0.0004,
"loss": 2.038480043411255,
"step": 1570
},
{
"epoch": 0.331671809478598,
"grad_norm": 0.10886934399604797,
"learning_rate": 0.0004,
"loss": 2.0486931800842285,
"step": 1572
},
{
"epoch": 0.3320937837909117,
"grad_norm": 0.09848352521657944,
"learning_rate": 0.0004,
"loss": 1.9118547439575195,
"step": 1574
},
{
"epoch": 0.3325157581032255,
"grad_norm": 0.09974077343940735,
"learning_rate": 0.0004,
"loss": 2.0117897987365723,
"step": 1576
},
{
"epoch": 0.3329377324155392,
"grad_norm": 0.1102852001786232,
"learning_rate": 0.0004,
"loss": 2.0351204872131348,
"step": 1578
},
{
"epoch": 0.3333597067278529,
"grad_norm": 0.1087120771408081,
"learning_rate": 0.0004,
"loss": 2.0185048580169678,
"step": 1580
},
{
"epoch": 0.3337816810401667,
"grad_norm": 0.09539604187011719,
"learning_rate": 0.0004,
"loss": 1.9460692405700684,
"step": 1582
},
{
"epoch": 0.3342036553524804,
"grad_norm": 0.10597709566354752,
"learning_rate": 0.0004,
"loss": 2.049802780151367,
"step": 1584
},
{
"epoch": 0.33462562966479414,
"grad_norm": 0.09302937984466553,
"learning_rate": 0.0004,
"loss": 2.0211756229400635,
"step": 1586
},
{
"epoch": 0.33504760397710787,
"grad_norm": 0.1053960993885994,
"learning_rate": 0.0004,
"loss": 2.0463926792144775,
"step": 1588
},
{
"epoch": 0.33546957828942164,
"grad_norm": 0.10724469274282455,
"learning_rate": 0.0004,
"loss": 1.9799323081970215,
"step": 1590
},
{
"epoch": 0.33589155260173537,
"grad_norm": 0.10626333206892014,
"learning_rate": 0.0004,
"loss": 2.003952741622925,
"step": 1592
},
{
"epoch": 0.3363135269140491,
"grad_norm": 0.11065094918012619,
"learning_rate": 0.0004,
"loss": 2.0335397720336914,
"step": 1594
},
{
"epoch": 0.33673550122636287,
"grad_norm": 0.11334291100502014,
"learning_rate": 0.0004,
"loss": 2.043454885482788,
"step": 1596
},
{
"epoch": 0.3371574755386766,
"grad_norm": 0.1212797686457634,
"learning_rate": 0.0004,
"loss": 2.011169195175171,
"step": 1598
},
{
"epoch": 0.3375794498509903,
"grad_norm": 0.10967233031988144,
"learning_rate": 0.0004,
"loss": 2.029719591140747,
"step": 1600
},
{
"epoch": 0.33800142416330403,
"grad_norm": 0.11400245875120163,
"learning_rate": 0.0004,
"loss": 2.0578675270080566,
"step": 1602
},
{
"epoch": 0.3384233984756178,
"grad_norm": 0.10003125667572021,
"learning_rate": 0.0004,
"loss": 2.052380084991455,
"step": 1604
},
{
"epoch": 0.33884537278793153,
"grad_norm": 0.10874568670988083,
"learning_rate": 0.0004,
"loss": 2.049670934677124,
"step": 1606
},
{
"epoch": 0.33926734710024525,
"grad_norm": 0.09906256198883057,
"learning_rate": 0.0004,
"loss": 2.0688095092773438,
"step": 1608
},
{
"epoch": 0.33968932141255903,
"grad_norm": 0.10611707717180252,
"learning_rate": 0.0004,
"loss": 1.9722919464111328,
"step": 1610
},
{
"epoch": 0.34011129572487275,
"grad_norm": 0.10133318603038788,
"learning_rate": 0.0004,
"loss": 2.0635390281677246,
"step": 1612
},
{
"epoch": 0.3405332700371865,
"grad_norm": 0.11087331175804138,
"learning_rate": 0.0004,
"loss": 2.0617902278900146,
"step": 1614
},
{
"epoch": 0.3409552443495002,
"grad_norm": 0.10502845793962479,
"learning_rate": 0.0004,
"loss": 2.0175294876098633,
"step": 1616
},
{
"epoch": 0.341377218661814,
"grad_norm": 0.10377192497253418,
"learning_rate": 0.0004,
"loss": 2.0534160137176514,
"step": 1618
},
{
"epoch": 0.3417991929741277,
"grad_norm": 0.11163768172264099,
"learning_rate": 0.0004,
"loss": 1.9704924821853638,
"step": 1620
},
{
"epoch": 0.3422211672864414,
"grad_norm": 0.10641206800937653,
"learning_rate": 0.0004,
"loss": 2.0665714740753174,
"step": 1622
},
{
"epoch": 0.3426431415987552,
"grad_norm": 0.11096753925085068,
"learning_rate": 0.0004,
"loss": 2.0521395206451416,
"step": 1624
},
{
"epoch": 0.3430651159110689,
"grad_norm": 0.10977613925933838,
"learning_rate": 0.0004,
"loss": 1.9546387195587158,
"step": 1626
},
{
"epoch": 0.34348709022338264,
"grad_norm": 0.09999173879623413,
"learning_rate": 0.0004,
"loss": 2.030972719192505,
"step": 1628
},
{
"epoch": 0.34390906453569636,
"grad_norm": 0.10646118223667145,
"learning_rate": 0.0004,
"loss": 1.9898555278778076,
"step": 1630
},
{
"epoch": 0.34433103884801014,
"grad_norm": 0.10504277795553207,
"learning_rate": 0.0004,
"loss": 2.0311975479125977,
"step": 1632
},
{
"epoch": 0.34475301316032386,
"grad_norm": 0.10043928772211075,
"learning_rate": 0.0004,
"loss": 2.016771078109741,
"step": 1634
},
{
"epoch": 0.3451749874726376,
"grad_norm": 0.10873425006866455,
"learning_rate": 0.0004,
"loss": 1.9871256351470947,
"step": 1636
},
{
"epoch": 0.34559696178495136,
"grad_norm": 0.12411544471979141,
"learning_rate": 0.0004,
"loss": 2.0406675338745117,
"step": 1638
},
{
"epoch": 0.3460189360972651,
"grad_norm": 0.10564137995243073,
"learning_rate": 0.0004,
"loss": 1.999247431755066,
"step": 1640
},
{
"epoch": 0.3464409104095788,
"grad_norm": 0.10970747470855713,
"learning_rate": 0.0004,
"loss": 1.9848129749298096,
"step": 1642
},
{
"epoch": 0.34686288472189253,
"grad_norm": 0.10622348636388779,
"learning_rate": 0.0004,
"loss": 2.0356855392456055,
"step": 1644
},
{
"epoch": 0.3472848590342063,
"grad_norm": 0.10049936175346375,
"learning_rate": 0.0004,
"loss": 1.9777731895446777,
"step": 1646
},
{
"epoch": 0.34770683334652003,
"grad_norm": 0.0955222025513649,
"learning_rate": 0.0004,
"loss": 2.0240771770477295,
"step": 1648
},
{
"epoch": 0.34812880765883375,
"grad_norm": 0.12071427702903748,
"learning_rate": 0.0004,
"loss": 2.031479835510254,
"step": 1650
},
{
"epoch": 0.3485507819711475,
"grad_norm": 0.10751160234212875,
"learning_rate": 0.0004,
"loss": 2.0816104412078857,
"step": 1652
},
{
"epoch": 0.34897275628346125,
"grad_norm": 0.12422254681587219,
"learning_rate": 0.0004,
"loss": 1.9731907844543457,
"step": 1654
},
{
"epoch": 0.34939473059577497,
"grad_norm": 0.10663653165102005,
"learning_rate": 0.0004,
"loss": 2.0478322505950928,
"step": 1656
},
{
"epoch": 0.3498167049080887,
"grad_norm": 0.11180845648050308,
"learning_rate": 0.0004,
"loss": 2.0372681617736816,
"step": 1658
},
{
"epoch": 0.35023867922040247,
"grad_norm": 0.10808839648962021,
"learning_rate": 0.0004,
"loss": 2.0175206661224365,
"step": 1660
},
{
"epoch": 0.3506606535327162,
"grad_norm": 0.10172480344772339,
"learning_rate": 0.0004,
"loss": 2.0679140090942383,
"step": 1662
},
{
"epoch": 0.3510826278450299,
"grad_norm": 0.11048221588134766,
"learning_rate": 0.0004,
"loss": 1.9486480951309204,
"step": 1664
},
{
"epoch": 0.3515046021573437,
"grad_norm": 0.1193045824766159,
"learning_rate": 0.0004,
"loss": 2.05427885055542,
"step": 1666
},
{
"epoch": 0.3519265764696574,
"grad_norm": 0.10146333277225494,
"learning_rate": 0.0004,
"loss": 2.0048863887786865,
"step": 1668
},
{
"epoch": 0.35234855078197114,
"grad_norm": 0.11769036203622818,
"learning_rate": 0.0004,
"loss": 2.0274953842163086,
"step": 1670
},
{
"epoch": 0.35277052509428486,
"grad_norm": 0.11162961274385452,
"learning_rate": 0.0004,
"loss": 2.0341732501983643,
"step": 1672
},
{
"epoch": 0.35319249940659864,
"grad_norm": 0.09634927660226822,
"learning_rate": 0.0004,
"loss": 2.031299591064453,
"step": 1674
},
{
"epoch": 0.35361447371891236,
"grad_norm": 0.10206437110900879,
"learning_rate": 0.0004,
"loss": 2.0100746154785156,
"step": 1676
},
{
"epoch": 0.3540364480312261,
"grad_norm": 0.10727409273386002,
"learning_rate": 0.0004,
"loss": 2.0030932426452637,
"step": 1678
},
{
"epoch": 0.35445842234353986,
"grad_norm": 0.09987301379442215,
"learning_rate": 0.0004,
"loss": 2.0274906158447266,
"step": 1680
},
{
"epoch": 0.3548803966558536,
"grad_norm": 0.0995340645313263,
"learning_rate": 0.0004,
"loss": 2.0236926078796387,
"step": 1682
},
{
"epoch": 0.3553023709681673,
"grad_norm": 0.09428931027650833,
"learning_rate": 0.0004,
"loss": 1.983730673789978,
"step": 1684
},
{
"epoch": 0.355724345280481,
"grad_norm": 0.10755246132612228,
"learning_rate": 0.0004,
"loss": 2.08170223236084,
"step": 1686
},
{
"epoch": 0.3561463195927948,
"grad_norm": 0.10375849902629852,
"learning_rate": 0.0004,
"loss": 2.0407795906066895,
"step": 1688
},
{
"epoch": 0.3565682939051085,
"grad_norm": 0.1375063955783844,
"learning_rate": 0.0004,
"loss": 2.002896547317505,
"step": 1690
},
{
"epoch": 0.35699026821742225,
"grad_norm": 0.10168766975402832,
"learning_rate": 0.0004,
"loss": 2.055514097213745,
"step": 1692
},
{
"epoch": 0.357412242529736,
"grad_norm": 0.11130502074956894,
"learning_rate": 0.0004,
"loss": 2.0014407634735107,
"step": 1694
},
{
"epoch": 0.35783421684204975,
"grad_norm": 0.10267114639282227,
"learning_rate": 0.0004,
"loss": 2.0570027828216553,
"step": 1696
},
{
"epoch": 0.35825619115436347,
"grad_norm": 0.12095794081687927,
"learning_rate": 0.0004,
"loss": 2.01104474067688,
"step": 1698
},
{
"epoch": 0.3586781654666772,
"grad_norm": 0.10102632641792297,
"learning_rate": 0.0004,
"loss": 2.000636100769043,
"step": 1700
},
{
"epoch": 0.35910013977899097,
"grad_norm": 0.11164496839046478,
"learning_rate": 0.0004,
"loss": 1.9788124561309814,
"step": 1702
},
{
"epoch": 0.3595221140913047,
"grad_norm": 0.10351933538913727,
"learning_rate": 0.0004,
"loss": 1.9406626224517822,
"step": 1704
},
{
"epoch": 0.3599440884036184,
"grad_norm": 0.12388698756694794,
"learning_rate": 0.0004,
"loss": 1.97909677028656,
"step": 1706
},
{
"epoch": 0.3603660627159322,
"grad_norm": 0.09685781598091125,
"learning_rate": 0.0004,
"loss": 2.0334558486938477,
"step": 1708
},
{
"epoch": 0.3607880370282459,
"grad_norm": 0.12370890378952026,
"learning_rate": 0.0004,
"loss": 2.0211026668548584,
"step": 1710
},
{
"epoch": 0.36121001134055963,
"grad_norm": 0.09980598092079163,
"learning_rate": 0.0004,
"loss": 2.015707015991211,
"step": 1712
},
{
"epoch": 0.36163198565287336,
"grad_norm": 0.1057153046131134,
"learning_rate": 0.0004,
"loss": 2.009878635406494,
"step": 1714
},
{
"epoch": 0.36205395996518713,
"grad_norm": 0.10937082022428513,
"learning_rate": 0.0004,
"loss": 1.9818382263183594,
"step": 1716
},
{
"epoch": 0.36247593427750086,
"grad_norm": 0.10777908563613892,
"learning_rate": 0.0004,
"loss": 2.020676851272583,
"step": 1718
},
{
"epoch": 0.3628979085898146,
"grad_norm": 0.11114877462387085,
"learning_rate": 0.0004,
"loss": 1.964246392250061,
"step": 1720
},
{
"epoch": 0.36331988290212835,
"grad_norm": 0.10094496607780457,
"learning_rate": 0.0004,
"loss": 1.9906814098358154,
"step": 1722
},
{
"epoch": 0.3637418572144421,
"grad_norm": 0.10544519871473312,
"learning_rate": 0.0004,
"loss": 2.0072786808013916,
"step": 1724
},
{
"epoch": 0.3641638315267558,
"grad_norm": 0.11200228333473206,
"learning_rate": 0.0004,
"loss": 2.031675100326538,
"step": 1726
},
{
"epoch": 0.3645858058390695,
"grad_norm": 0.10408878326416016,
"learning_rate": 0.0004,
"loss": 2.029160499572754,
"step": 1728
},
{
"epoch": 0.3650077801513833,
"grad_norm": 0.09775891155004501,
"learning_rate": 0.0004,
"loss": 1.9884238243103027,
"step": 1730
},
{
"epoch": 0.365429754463697,
"grad_norm": 0.10195142030715942,
"learning_rate": 0.0004,
"loss": 2.0471689701080322,
"step": 1732
},
{
"epoch": 0.36585172877601074,
"grad_norm": 0.10464382916688919,
"learning_rate": 0.0004,
"loss": 1.9236515760421753,
"step": 1734
},
{
"epoch": 0.3662737030883245,
"grad_norm": 0.11086594313383102,
"learning_rate": 0.0004,
"loss": 1.9768121242523193,
"step": 1736
},
{
"epoch": 0.36669567740063824,
"grad_norm": 0.10608164966106415,
"learning_rate": 0.0004,
"loss": 2.025625467300415,
"step": 1738
},
{
"epoch": 0.36711765171295196,
"grad_norm": 0.12192277610301971,
"learning_rate": 0.0004,
"loss": 2.003598690032959,
"step": 1740
},
{
"epoch": 0.3675396260252657,
"grad_norm": 0.09979259222745895,
"learning_rate": 0.0004,
"loss": 1.9627618789672852,
"step": 1742
},
{
"epoch": 0.36796160033757946,
"grad_norm": 0.10990424454212189,
"learning_rate": 0.0004,
"loss": 2.0553503036499023,
"step": 1744
},
{
"epoch": 0.3683835746498932,
"grad_norm": 0.0946066603064537,
"learning_rate": 0.0004,
"loss": 1.9856538772583008,
"step": 1746
},
{
"epoch": 0.3688055489622069,
"grad_norm": 0.10665788501501083,
"learning_rate": 0.0004,
"loss": 2.036868095397949,
"step": 1748
},
{
"epoch": 0.3692275232745207,
"grad_norm": 0.10382699221372604,
"learning_rate": 0.0004,
"loss": 2.0172998905181885,
"step": 1750
},
{
"epoch": 0.3696494975868344,
"grad_norm": 0.11515715718269348,
"learning_rate": 0.0004,
"loss": 2.0313518047332764,
"step": 1752
},
{
"epoch": 0.37007147189914813,
"grad_norm": 0.11131878942251205,
"learning_rate": 0.0004,
"loss": 1.9803560972213745,
"step": 1754
},
{
"epoch": 0.37049344621146185,
"grad_norm": 0.11094849556684494,
"learning_rate": 0.0004,
"loss": 2.0181262493133545,
"step": 1756
},
{
"epoch": 0.37091542052377563,
"grad_norm": 0.10003557801246643,
"learning_rate": 0.0004,
"loss": 2.0257697105407715,
"step": 1758
},
{
"epoch": 0.37133739483608935,
"grad_norm": 0.12847717106342316,
"learning_rate": 0.0004,
"loss": 2.0548484325408936,
"step": 1760
},
{
"epoch": 0.3717593691484031,
"grad_norm": 0.11092184484004974,
"learning_rate": 0.0004,
"loss": 2.0151097774505615,
"step": 1762
},
{
"epoch": 0.37218134346071685,
"grad_norm": 0.1026356890797615,
"learning_rate": 0.0004,
"loss": 2.0361034870147705,
"step": 1764
},
{
"epoch": 0.3726033177730306,
"grad_norm": 0.12397974729537964,
"learning_rate": 0.0004,
"loss": 2.0416955947875977,
"step": 1766
},
{
"epoch": 0.3730252920853443,
"grad_norm": 0.10785350948572159,
"learning_rate": 0.0004,
"loss": 2.0184898376464844,
"step": 1768
},
{
"epoch": 0.373447266397658,
"grad_norm": 0.09647557139396667,
"learning_rate": 0.0004,
"loss": 2.0348358154296875,
"step": 1770
},
{
"epoch": 0.3738692407099718,
"grad_norm": 0.10894135385751724,
"learning_rate": 0.0004,
"loss": 2.0238351821899414,
"step": 1772
},
{
"epoch": 0.3742912150222855,
"grad_norm": 0.11743897199630737,
"learning_rate": 0.0004,
"loss": 2.040525436401367,
"step": 1774
},
{
"epoch": 0.37471318933459924,
"grad_norm": 0.11659212410449982,
"learning_rate": 0.0004,
"loss": 2.018855333328247,
"step": 1776
},
{
"epoch": 0.375135163646913,
"grad_norm": 0.10117500275373459,
"learning_rate": 0.0004,
"loss": 2.040416717529297,
"step": 1778
},
{
"epoch": 0.37555713795922674,
"grad_norm": 0.10208738595247269,
"learning_rate": 0.0004,
"loss": 2.0061869621276855,
"step": 1780
},
{
"epoch": 0.37597911227154046,
"grad_norm": 0.09548701345920563,
"learning_rate": 0.0004,
"loss": 2.014235496520996,
"step": 1782
},
{
"epoch": 0.3764010865838542,
"grad_norm": 0.1199650913476944,
"learning_rate": 0.0004,
"loss": 1.977577567100525,
"step": 1784
},
{
"epoch": 0.37682306089616796,
"grad_norm": 0.12102092057466507,
"learning_rate": 0.0004,
"loss": 2.0604355335235596,
"step": 1786
},
{
"epoch": 0.3772450352084817,
"grad_norm": 0.11080902069807053,
"learning_rate": 0.0004,
"loss": 2.023509979248047,
"step": 1788
},
{
"epoch": 0.3776670095207954,
"grad_norm": 0.11269386112689972,
"learning_rate": 0.0004,
"loss": 2.011964797973633,
"step": 1790
},
{
"epoch": 0.3780889838331092,
"grad_norm": 0.10254208743572235,
"learning_rate": 0.0004,
"loss": 2.060206890106201,
"step": 1792
},
{
"epoch": 0.3785109581454229,
"grad_norm": 0.10592658072710037,
"learning_rate": 0.0004,
"loss": 2.0015995502471924,
"step": 1794
},
{
"epoch": 0.3789329324577366,
"grad_norm": 0.09882373362779617,
"learning_rate": 0.0004,
"loss": 1.9931437969207764,
"step": 1796
},
{
"epoch": 0.37935490677005035,
"grad_norm": 0.10435090214014053,
"learning_rate": 0.0004,
"loss": 2.011550188064575,
"step": 1798
},
{
"epoch": 0.3797768810823641,
"grad_norm": 0.11017206311225891,
"learning_rate": 0.0004,
"loss": 1.949293851852417,
"step": 1800
},
{
"epoch": 0.38019885539467785,
"grad_norm": 0.09750290960073471,
"learning_rate": 0.0004,
"loss": 1.9660183191299438,
"step": 1802
},
{
"epoch": 0.38062082970699157,
"grad_norm": 0.10377368330955505,
"learning_rate": 0.0004,
"loss": 2.033520221710205,
"step": 1804
},
{
"epoch": 0.38104280401930535,
"grad_norm": 0.10056043416261673,
"learning_rate": 0.0004,
"loss": 2.050577402114868,
"step": 1806
},
{
"epoch": 0.38146477833161907,
"grad_norm": 0.11384923756122589,
"learning_rate": 0.0004,
"loss": 1.9606878757476807,
"step": 1808
},
{
"epoch": 0.3818867526439328,
"grad_norm": 0.12534750998020172,
"learning_rate": 0.0004,
"loss": 1.9906620979309082,
"step": 1810
},
{
"epoch": 0.3823087269562465,
"grad_norm": 0.10672692209482193,
"learning_rate": 0.0004,
"loss": 1.9776018857955933,
"step": 1812
},
{
"epoch": 0.3827307012685603,
"grad_norm": 0.10126478224992752,
"learning_rate": 0.0004,
"loss": 2.0241119861602783,
"step": 1814
},
{
"epoch": 0.383152675580874,
"grad_norm": 0.10249096155166626,
"learning_rate": 0.0004,
"loss": 2.0062613487243652,
"step": 1816
},
{
"epoch": 0.38357464989318774,
"grad_norm": 0.10473034530878067,
"learning_rate": 0.0004,
"loss": 1.9895328283309937,
"step": 1818
},
{
"epoch": 0.3839966242055015,
"grad_norm": 0.1216689944267273,
"learning_rate": 0.0004,
"loss": 1.9631131887435913,
"step": 1820
},
{
"epoch": 0.38441859851781524,
"grad_norm": 0.1025574803352356,
"learning_rate": 0.0004,
"loss": 2.0995559692382812,
"step": 1822
},
{
"epoch": 0.38484057283012896,
"grad_norm": 0.09407646209001541,
"learning_rate": 0.0004,
"loss": 2.0030133724212646,
"step": 1824
},
{
"epoch": 0.3852625471424427,
"grad_norm": 0.10963959991931915,
"learning_rate": 0.0004,
"loss": 1.9982473850250244,
"step": 1826
},
{
"epoch": 0.38568452145475646,
"grad_norm": 0.09796998649835587,
"learning_rate": 0.0004,
"loss": 2.0309770107269287,
"step": 1828
},
{
"epoch": 0.3861064957670702,
"grad_norm": 0.1048675924539566,
"learning_rate": 0.0004,
"loss": 1.9741034507751465,
"step": 1830
},
{
"epoch": 0.3865284700793839,
"grad_norm": 0.10768698900938034,
"learning_rate": 0.0004,
"loss": 2.0877230167388916,
"step": 1832
},
{
"epoch": 0.3869504443916977,
"grad_norm": 0.10516445338726044,
"learning_rate": 0.0004,
"loss": 1.968559980392456,
"step": 1834
},
{
"epoch": 0.3873724187040114,
"grad_norm": 0.10021158307790756,
"learning_rate": 0.0004,
"loss": 2.0329670906066895,
"step": 1836
},
{
"epoch": 0.3877943930163251,
"grad_norm": 0.10462050884962082,
"learning_rate": 0.0004,
"loss": 1.9913111925125122,
"step": 1838
},
{
"epoch": 0.38821636732863885,
"grad_norm": 0.10117878019809723,
"learning_rate": 0.0004,
"loss": 2.00445818901062,
"step": 1840
},
{
"epoch": 0.3886383416409526,
"grad_norm": 0.10890329629182816,
"learning_rate": 0.0004,
"loss": 2.0784430503845215,
"step": 1842
},
{
"epoch": 0.38906031595326634,
"grad_norm": 0.10154514759778976,
"learning_rate": 0.0004,
"loss": 2.0695645809173584,
"step": 1844
},
{
"epoch": 0.38948229026558007,
"grad_norm": 0.11083352565765381,
"learning_rate": 0.0004,
"loss": 1.9813780784606934,
"step": 1846
},
{
"epoch": 0.38990426457789384,
"grad_norm": 0.1024395078420639,
"learning_rate": 0.0004,
"loss": 1.9848823547363281,
"step": 1848
},
{
"epoch": 0.39032623889020757,
"grad_norm": 0.12603364884853363,
"learning_rate": 0.0004,
"loss": 2.0116779804229736,
"step": 1850
},
{
"epoch": 0.3907482132025213,
"grad_norm": 0.09810984134674072,
"learning_rate": 0.0004,
"loss": 2.0123283863067627,
"step": 1852
},
{
"epoch": 0.391170187514835,
"grad_norm": 0.11026440560817719,
"learning_rate": 0.0004,
"loss": 2.0084388256073,
"step": 1854
},
{
"epoch": 0.3915921618271488,
"grad_norm": 0.1009652316570282,
"learning_rate": 0.0004,
"loss": 2.0251901149749756,
"step": 1856
},
{
"epoch": 0.3920141361394625,
"grad_norm": 0.11032655835151672,
"learning_rate": 0.0004,
"loss": 1.9531500339508057,
"step": 1858
},
{
"epoch": 0.39243611045177623,
"grad_norm": 0.10076446086168289,
"learning_rate": 0.0004,
"loss": 2.0322368144989014,
"step": 1860
},
{
"epoch": 0.39285808476409,
"grad_norm": 0.13186770677566528,
"learning_rate": 0.0004,
"loss": 2.0439555644989014,
"step": 1862
},
{
"epoch": 0.39328005907640373,
"grad_norm": 0.09583664685487747,
"learning_rate": 0.0004,
"loss": 2.0485589504241943,
"step": 1864
},
{
"epoch": 0.39370203338871745,
"grad_norm": 0.10402042418718338,
"learning_rate": 0.0004,
"loss": 2.0431933403015137,
"step": 1866
},
{
"epoch": 0.3941240077010312,
"grad_norm": 0.10985450446605682,
"learning_rate": 0.0004,
"loss": 1.923365592956543,
"step": 1868
},
{
"epoch": 0.39454598201334495,
"grad_norm": 0.11034008115530014,
"learning_rate": 0.0004,
"loss": 2.018366813659668,
"step": 1870
},
{
"epoch": 0.3949679563256587,
"grad_norm": 0.1043843924999237,
"learning_rate": 0.0004,
"loss": 2.004805088043213,
"step": 1872
},
{
"epoch": 0.3953899306379724,
"grad_norm": 0.10831619054079056,
"learning_rate": 0.0004,
"loss": 2.0457000732421875,
"step": 1874
},
{
"epoch": 0.3958119049502862,
"grad_norm": 0.12373778969049454,
"learning_rate": 0.0004,
"loss": 2.0204694271087646,
"step": 1876
},
{
"epoch": 0.3962338792625999,
"grad_norm": 0.12175879627466202,
"learning_rate": 0.0004,
"loss": 2.0352983474731445,
"step": 1878
},
{
"epoch": 0.3966558535749136,
"grad_norm": 0.10968101769685745,
"learning_rate": 0.0004,
"loss": 2.0558152198791504,
"step": 1880
},
{
"epoch": 0.39707782788722734,
"grad_norm": 0.10321266204118729,
"learning_rate": 0.0004,
"loss": 1.992050051689148,
"step": 1882
},
{
"epoch": 0.3974998021995411,
"grad_norm": 0.100496806204319,
"learning_rate": 0.0004,
"loss": 2.0391061305999756,
"step": 1884
},
{
"epoch": 0.39792177651185484,
"grad_norm": 0.10099262744188309,
"learning_rate": 0.0004,
"loss": 1.994731068611145,
"step": 1886
},
{
"epoch": 0.39834375082416856,
"grad_norm": 0.10313970595598221,
"learning_rate": 0.0004,
"loss": 1.9221858978271484,
"step": 1888
},
{
"epoch": 0.39876572513648234,
"grad_norm": 0.09847329556941986,
"learning_rate": 0.0004,
"loss": 2.039625406265259,
"step": 1890
},
{
"epoch": 0.39918769944879606,
"grad_norm": 0.10227619111537933,
"learning_rate": 0.0004,
"loss": 1.985874891281128,
"step": 1892
},
{
"epoch": 0.3996096737611098,
"grad_norm": 0.10636409372091293,
"learning_rate": 0.0004,
"loss": 1.9808831214904785,
"step": 1894
},
{
"epoch": 0.4000316480734235,
"grad_norm": 0.10621833056211472,
"learning_rate": 0.0004,
"loss": 1.9487495422363281,
"step": 1896
},
{
"epoch": 0.4000316480734235,
"eval_loss": 2.007476568222046,
"eval_runtime": 85.8704,
"eval_samples_per_second": 35.682,
"eval_steps_per_second": 4.46,
"step": 1896
},
{
"epoch": 0.4004536223857373,
"grad_norm": 0.11206910014152527,
"learning_rate": 0.0004,
"loss": 2.0210747718811035,
"step": 1898
},
{
"epoch": 0.400875596698051,
"grad_norm": 0.11541873961687088,
"learning_rate": 0.0004,
"loss": 2.012225389480591,
"step": 1900
},
{
"epoch": 0.40129757101036473,
"grad_norm": 0.1049957349896431,
"learning_rate": 0.0004,
"loss": 2.0555529594421387,
"step": 1902
},
{
"epoch": 0.4017195453226785,
"grad_norm": 0.1135619580745697,
"learning_rate": 0.0004,
"loss": 2.064544916152954,
"step": 1904
},
{
"epoch": 0.40214151963499223,
"grad_norm": 0.1060766726732254,
"learning_rate": 0.0004,
"loss": 2.0196597576141357,
"step": 1906
},
{
"epoch": 0.40256349394730595,
"grad_norm": 0.1219051256775856,
"learning_rate": 0.0004,
"loss": 1.9635637998580933,
"step": 1908
},
{
"epoch": 0.4029854682596197,
"grad_norm": 0.11062590777873993,
"learning_rate": 0.0004,
"loss": 2.0131888389587402,
"step": 1910
},
{
"epoch": 0.40340744257193345,
"grad_norm": 0.09908871352672577,
"learning_rate": 0.0004,
"loss": 1.983797550201416,
"step": 1912
},
{
"epoch": 0.40382941688424717,
"grad_norm": 0.09351597726345062,
"learning_rate": 0.0004,
"loss": 1.987016201019287,
"step": 1914
},
{
"epoch": 0.4042513911965609,
"grad_norm": 0.10428667813539505,
"learning_rate": 0.0004,
"loss": 2.0556464195251465,
"step": 1916
},
{
"epoch": 0.40467336550887467,
"grad_norm": 0.10044407844543457,
"learning_rate": 0.0004,
"loss": 2.0800204277038574,
"step": 1918
},
{
"epoch": 0.4050953398211884,
"grad_norm": 0.09969620406627655,
"learning_rate": 0.0004,
"loss": 2.0359761714935303,
"step": 1920
},
{
"epoch": 0.4055173141335021,
"grad_norm": 0.10544329136610031,
"learning_rate": 0.0004,
"loss": 2.056628465652466,
"step": 1922
},
{
"epoch": 0.40593928844581584,
"grad_norm": 0.11956953257322311,
"learning_rate": 0.0004,
"loss": 1.988150715827942,
"step": 1924
},
{
"epoch": 0.4063612627581296,
"grad_norm": 0.11934361606836319,
"learning_rate": 0.0004,
"loss": 2.032003402709961,
"step": 1926
},
{
"epoch": 0.40678323707044334,
"grad_norm": 0.09921639412641525,
"learning_rate": 0.0004,
"loss": 2.0804972648620605,
"step": 1928
},
{
"epoch": 0.40720521138275706,
"grad_norm": 0.11270210891962051,
"learning_rate": 0.0004,
"loss": 1.9156255722045898,
"step": 1930
},
{
"epoch": 0.40762718569507084,
"grad_norm": 0.10954868048429489,
"learning_rate": 0.0004,
"loss": 2.0205962657928467,
"step": 1932
},
{
"epoch": 0.40804916000738456,
"grad_norm": 0.10299162566661835,
"learning_rate": 0.0004,
"loss": 1.9980796575546265,
"step": 1934
},
{
"epoch": 0.4084711343196983,
"grad_norm": 0.09984418749809265,
"learning_rate": 0.0004,
"loss": 2.0021860599517822,
"step": 1936
},
{
"epoch": 0.408893108632012,
"grad_norm": 0.10401031374931335,
"learning_rate": 0.0004,
"loss": 1.965635895729065,
"step": 1938
},
{
"epoch": 0.4093150829443258,
"grad_norm": 0.10177171975374222,
"learning_rate": 0.0004,
"loss": 2.036318778991699,
"step": 1940
},
{
"epoch": 0.4097370572566395,
"grad_norm": 0.10262718051671982,
"learning_rate": 0.0004,
"loss": 1.9998247623443604,
"step": 1942
},
{
"epoch": 0.4101590315689532,
"grad_norm": 0.1086772233247757,
"learning_rate": 0.0004,
"loss": 1.991339087486267,
"step": 1944
},
{
"epoch": 0.410581005881267,
"grad_norm": 0.10860958695411682,
"learning_rate": 0.0004,
"loss": 2.01627516746521,
"step": 1946
},
{
"epoch": 0.4110029801935807,
"grad_norm": 0.10782618075609207,
"learning_rate": 0.0004,
"loss": 2.01762318611145,
"step": 1948
},
{
"epoch": 0.41142495450589445,
"grad_norm": 0.09959777444601059,
"learning_rate": 0.0004,
"loss": 1.9831655025482178,
"step": 1950
},
{
"epoch": 0.41184692881820817,
"grad_norm": 0.10234978049993515,
"learning_rate": 0.0004,
"loss": 1.9721789360046387,
"step": 1952
},
{
"epoch": 0.41226890313052195,
"grad_norm": 0.10592693835496902,
"learning_rate": 0.0004,
"loss": 1.9608489274978638,
"step": 1954
},
{
"epoch": 0.41269087744283567,
"grad_norm": 0.10494564473628998,
"learning_rate": 0.0004,
"loss": 1.9390430450439453,
"step": 1956
},
{
"epoch": 0.4131128517551494,
"grad_norm": 0.13374176621437073,
"learning_rate": 0.0004,
"loss": 1.9894862174987793,
"step": 1958
},
{
"epoch": 0.41353482606746317,
"grad_norm": 0.11623284220695496,
"learning_rate": 0.0004,
"loss": 1.9661037921905518,
"step": 1960
},
{
"epoch": 0.4139568003797769,
"grad_norm": 0.10733483731746674,
"learning_rate": 0.0004,
"loss": 2.040645122528076,
"step": 1962
},
{
"epoch": 0.4143787746920906,
"grad_norm": 0.11347152292728424,
"learning_rate": 0.0004,
"loss": 2.012077808380127,
"step": 1964
},
{
"epoch": 0.41480074900440433,
"grad_norm": 0.11153286695480347,
"learning_rate": 0.0004,
"loss": 1.9731961488723755,
"step": 1966
},
{
"epoch": 0.4152227233167181,
"grad_norm": 0.10009129345417023,
"learning_rate": 0.0004,
"loss": 2.009488582611084,
"step": 1968
},
{
"epoch": 0.41564469762903183,
"grad_norm": 0.10955440253019333,
"learning_rate": 0.0004,
"loss": 2.0032448768615723,
"step": 1970
},
{
"epoch": 0.41606667194134556,
"grad_norm": 0.1244802251458168,
"learning_rate": 0.0004,
"loss": 1.9455724954605103,
"step": 1972
},
{
"epoch": 0.41648864625365933,
"grad_norm": 0.09637445211410522,
"learning_rate": 0.0004,
"loss": 2.012906074523926,
"step": 1974
},
{
"epoch": 0.41691062056597306,
"grad_norm": 0.10025131702423096,
"learning_rate": 0.0004,
"loss": 2.019721508026123,
"step": 1976
},
{
"epoch": 0.4173325948782868,
"grad_norm": 0.11075221002101898,
"learning_rate": 0.0004,
"loss": 1.9751369953155518,
"step": 1978
},
{
"epoch": 0.4177545691906005,
"grad_norm": 0.10105768591165543,
"learning_rate": 0.0004,
"loss": 2.010686159133911,
"step": 1980
},
{
"epoch": 0.4181765435029143,
"grad_norm": 0.09841880947351456,
"learning_rate": 0.0004,
"loss": 1.9785358905792236,
"step": 1982
},
{
"epoch": 0.418598517815228,
"grad_norm": 0.11022451519966125,
"learning_rate": 0.0004,
"loss": 2.051682472229004,
"step": 1984
},
{
"epoch": 0.4190204921275417,
"grad_norm": 0.11154559254646301,
"learning_rate": 0.0004,
"loss": 2.019987106323242,
"step": 1986
},
{
"epoch": 0.4194424664398555,
"grad_norm": 0.13209353387355804,
"learning_rate": 0.0004,
"loss": 2.0060925483703613,
"step": 1988
},
{
"epoch": 0.4198644407521692,
"grad_norm": 0.11981359124183655,
"learning_rate": 0.0004,
"loss": 1.957808494567871,
"step": 1990
},
{
"epoch": 0.42028641506448294,
"grad_norm": 0.11082693934440613,
"learning_rate": 0.0004,
"loss": 2.043726682662964,
"step": 1992
},
{
"epoch": 0.42070838937679667,
"grad_norm": 0.10682928562164307,
"learning_rate": 0.0004,
"loss": 2.047206401824951,
"step": 1994
},
{
"epoch": 0.42113036368911044,
"grad_norm": 0.10359685868024826,
"learning_rate": 0.0004,
"loss": 1.9776005744934082,
"step": 1996
},
{
"epoch": 0.42155233800142417,
"grad_norm": 0.11906304955482483,
"learning_rate": 0.0004,
"loss": 2.004349946975708,
"step": 1998
},
{
"epoch": 0.4219743123137379,
"grad_norm": 0.10153121501207352,
"learning_rate": 0.0004,
"loss": 1.973453164100647,
"step": 2000
},
{
"epoch": 0.42239628662605166,
"grad_norm": 0.11691207438707352,
"learning_rate": 0.0004,
"loss": 2.0719709396362305,
"step": 2002
},
{
"epoch": 0.4228182609383654,
"grad_norm": 0.12921367585659027,
"learning_rate": 0.0004,
"loss": 1.9696711301803589,
"step": 2004
},
{
"epoch": 0.4232402352506791,
"grad_norm": 0.11897560954093933,
"learning_rate": 0.0004,
"loss": 1.9256787300109863,
"step": 2006
},
{
"epoch": 0.42366220956299283,
"grad_norm": 0.09961225837469101,
"learning_rate": 0.0004,
"loss": 2.013353109359741,
"step": 2008
},
{
"epoch": 0.4240841838753066,
"grad_norm": 0.09646215289831161,
"learning_rate": 0.0004,
"loss": 1.9872149229049683,
"step": 2010
},
{
"epoch": 0.42450615818762033,
"grad_norm": 0.09996464103460312,
"learning_rate": 0.0004,
"loss": 1.9781116247177124,
"step": 2012
},
{
"epoch": 0.42492813249993405,
"grad_norm": 0.10583832859992981,
"learning_rate": 0.0004,
"loss": 1.9643330574035645,
"step": 2014
},
{
"epoch": 0.42535010681224783,
"grad_norm": 0.11507900059223175,
"learning_rate": 0.0004,
"loss": 1.9224776029586792,
"step": 2016
},
{
"epoch": 0.42577208112456155,
"grad_norm": 0.10323559492826462,
"learning_rate": 0.0004,
"loss": 2.0397515296936035,
"step": 2018
},
{
"epoch": 0.4261940554368753,
"grad_norm": 0.09832274168729782,
"learning_rate": 0.0004,
"loss": 1.9954586029052734,
"step": 2020
},
{
"epoch": 0.426616029749189,
"grad_norm": 0.11246035993099213,
"learning_rate": 0.0004,
"loss": 2.0209174156188965,
"step": 2022
},
{
"epoch": 0.4270380040615028,
"grad_norm": 0.09806378930807114,
"learning_rate": 0.0004,
"loss": 2.0529582500457764,
"step": 2024
},
{
"epoch": 0.4274599783738165,
"grad_norm": 0.09668385982513428,
"learning_rate": 0.0004,
"loss": 2.0477523803710938,
"step": 2026
},
{
"epoch": 0.4278819526861302,
"grad_norm": 0.1092454195022583,
"learning_rate": 0.0004,
"loss": 1.9736766815185547,
"step": 2028
},
{
"epoch": 0.428303926998444,
"grad_norm": 0.11669601500034332,
"learning_rate": 0.0004,
"loss": 2.0277116298675537,
"step": 2030
},
{
"epoch": 0.4287259013107577,
"grad_norm": 0.10303463786840439,
"learning_rate": 0.0004,
"loss": 1.9472200870513916,
"step": 2032
},
{
"epoch": 0.42914787562307144,
"grad_norm": 0.10777775943279266,
"learning_rate": 0.0004,
"loss": 1.9354808330535889,
"step": 2034
},
{
"epoch": 0.42956984993538516,
"grad_norm": 0.09631859511137009,
"learning_rate": 0.0004,
"loss": 2.0478014945983887,
"step": 2036
},
{
"epoch": 0.42999182424769894,
"grad_norm": 0.10766435414552689,
"learning_rate": 0.0004,
"loss": 2.0453927516937256,
"step": 2038
},
{
"epoch": 0.43041379856001266,
"grad_norm": 0.11941702663898468,
"learning_rate": 0.0004,
"loss": 1.9345844984054565,
"step": 2040
},
{
"epoch": 0.4308357728723264,
"grad_norm": 0.11215683817863464,
"learning_rate": 0.0004,
"loss": 2.026921510696411,
"step": 2042
},
{
"epoch": 0.43125774718464016,
"grad_norm": 0.10473362356424332,
"learning_rate": 0.0004,
"loss": 1.9552847146987915,
"step": 2044
},
{
"epoch": 0.4316797214969539,
"grad_norm": 0.11646725237369537,
"learning_rate": 0.0004,
"loss": 1.979164719581604,
"step": 2046
},
{
"epoch": 0.4321016958092676,
"grad_norm": 0.10798052698373795,
"learning_rate": 0.0004,
"loss": 2.0296213626861572,
"step": 2048
},
{
"epoch": 0.4325236701215813,
"grad_norm": 0.11152573674917221,
"learning_rate": 0.0004,
"loss": 1.9848031997680664,
"step": 2050
},
{
"epoch": 0.4329456444338951,
"grad_norm": 0.1181911751627922,
"learning_rate": 0.0004,
"loss": 2.019930362701416,
"step": 2052
},
{
"epoch": 0.4333676187462088,
"grad_norm": 0.09586432576179504,
"learning_rate": 0.0004,
"loss": 2.0762386322021484,
"step": 2054
},
{
"epoch": 0.43378959305852255,
"grad_norm": 0.10266046226024628,
"learning_rate": 0.0004,
"loss": 1.9970450401306152,
"step": 2056
},
{
"epoch": 0.4342115673708363,
"grad_norm": 0.10519974678754807,
"learning_rate": 0.0004,
"loss": 2.005641460418701,
"step": 2058
},
{
"epoch": 0.43463354168315005,
"grad_norm": 0.09681975841522217,
"learning_rate": 0.0004,
"loss": 2.027987241744995,
"step": 2060
},
{
"epoch": 0.43505551599546377,
"grad_norm": 0.11653660237789154,
"learning_rate": 0.0004,
"loss": 2.016028642654419,
"step": 2062
},
{
"epoch": 0.4354774903077775,
"grad_norm": 0.13977235555648804,
"learning_rate": 0.0004,
"loss": 1.9971271753311157,
"step": 2064
},
{
"epoch": 0.43589946462009127,
"grad_norm": 0.10620417445898056,
"learning_rate": 0.0004,
"loss": 2.0887300968170166,
"step": 2066
},
{
"epoch": 0.436321438932405,
"grad_norm": 0.11155949532985687,
"learning_rate": 0.0004,
"loss": 2.011681318283081,
"step": 2068
},
{
"epoch": 0.4367434132447187,
"grad_norm": 0.10156253725290298,
"learning_rate": 0.0004,
"loss": 2.0017611980438232,
"step": 2070
},
{
"epoch": 0.4371653875570325,
"grad_norm": 0.10610520839691162,
"learning_rate": 0.0004,
"loss": 2.0229945182800293,
"step": 2072
},
{
"epoch": 0.4375873618693462,
"grad_norm": 0.11126831918954849,
"learning_rate": 0.0004,
"loss": 1.993042230606079,
"step": 2074
},
{
"epoch": 0.43800933618165994,
"grad_norm": 0.09987028688192368,
"learning_rate": 0.0004,
"loss": 1.9925321340560913,
"step": 2076
},
{
"epoch": 0.43843131049397366,
"grad_norm": 0.11103028804063797,
"learning_rate": 0.0004,
"loss": 1.982485055923462,
"step": 2078
},
{
"epoch": 0.43885328480628744,
"grad_norm": 0.11351363360881805,
"learning_rate": 0.0004,
"loss": 2.005112648010254,
"step": 2080
},
{
"epoch": 0.43927525911860116,
"grad_norm": 0.10984377562999725,
"learning_rate": 0.0004,
"loss": 2.0160129070281982,
"step": 2082
},
{
"epoch": 0.4396972334309149,
"grad_norm": 0.11146343499422073,
"learning_rate": 0.0004,
"loss": 1.990799903869629,
"step": 2084
},
{
"epoch": 0.44011920774322866,
"grad_norm": 0.11536877602338791,
"learning_rate": 0.0004,
"loss": 2.0534658432006836,
"step": 2086
},
{
"epoch": 0.4405411820555424,
"grad_norm": 0.10090973973274231,
"learning_rate": 0.0004,
"loss": 2.0760488510131836,
"step": 2088
},
{
"epoch": 0.4409631563678561,
"grad_norm": 0.11541594564914703,
"learning_rate": 0.0004,
"loss": 2.0133144855499268,
"step": 2090
},
{
"epoch": 0.4413851306801698,
"grad_norm": 0.16618090867996216,
"learning_rate": 0.0004,
"loss": 2.045398473739624,
"step": 2092
},
{
"epoch": 0.4418071049924836,
"grad_norm": 0.10003620386123657,
"learning_rate": 0.0004,
"loss": 2.004598379135132,
"step": 2094
},
{
"epoch": 0.4422290793047973,
"grad_norm": 0.11203385144472122,
"learning_rate": 0.0004,
"loss": 2.0071098804473877,
"step": 2096
},
{
"epoch": 0.44265105361711105,
"grad_norm": 0.10174703598022461,
"learning_rate": 0.0004,
"loss": 2.032463550567627,
"step": 2098
},
{
"epoch": 0.4430730279294248,
"grad_norm": 0.10276448726654053,
"learning_rate": 0.0004,
"loss": 2.021678924560547,
"step": 2100
},
{
"epoch": 0.44349500224173855,
"grad_norm": 0.09845304489135742,
"learning_rate": 0.0004,
"loss": 2.0061347484588623,
"step": 2102
},
{
"epoch": 0.44391697655405227,
"grad_norm": 0.11297474056482315,
"learning_rate": 0.0004,
"loss": 1.9412140846252441,
"step": 2104
},
{
"epoch": 0.444338950866366,
"grad_norm": 0.1126975417137146,
"learning_rate": 0.0004,
"loss": 2.0096042156219482,
"step": 2106
},
{
"epoch": 0.44476092517867977,
"grad_norm": 0.11344601958990097,
"learning_rate": 0.0004,
"loss": 1.946977972984314,
"step": 2108
},
{
"epoch": 0.4451828994909935,
"grad_norm": 0.09927421063184738,
"learning_rate": 0.0004,
"loss": 1.976645827293396,
"step": 2110
},
{
"epoch": 0.4456048738033072,
"grad_norm": 0.12145490199327469,
"learning_rate": 0.0004,
"loss": 2.0486249923706055,
"step": 2112
},
{
"epoch": 0.446026848115621,
"grad_norm": 0.10307963192462921,
"learning_rate": 0.0004,
"loss": 1.9108384847640991,
"step": 2114
},
{
"epoch": 0.4464488224279347,
"grad_norm": 0.10217355191707611,
"learning_rate": 0.0004,
"loss": 1.9963434934616089,
"step": 2116
},
{
"epoch": 0.44687079674024843,
"grad_norm": 0.113113172352314,
"learning_rate": 0.0004,
"loss": 2.0183160305023193,
"step": 2118
},
{
"epoch": 0.44729277105256215,
"grad_norm": 0.12282717227935791,
"learning_rate": 0.0004,
"loss": 2.0508859157562256,
"step": 2120
},
{
"epoch": 0.44771474536487593,
"grad_norm": 0.11350229382514954,
"learning_rate": 0.0004,
"loss": 2.0262668132781982,
"step": 2122
},
{
"epoch": 0.44813671967718965,
"grad_norm": 0.10847412794828415,
"learning_rate": 0.0004,
"loss": 1.9544306993484497,
"step": 2124
},
{
"epoch": 0.4485586939895034,
"grad_norm": 0.10568704456090927,
"learning_rate": 0.0004,
"loss": 2.0258617401123047,
"step": 2126
},
{
"epoch": 0.44898066830181715,
"grad_norm": 0.11809568852186203,
"learning_rate": 0.0004,
"loss": 2.0351240634918213,
"step": 2128
},
{
"epoch": 0.4494026426141309,
"grad_norm": 0.10306884348392487,
"learning_rate": 0.0004,
"loss": 2.0120677947998047,
"step": 2130
},
{
"epoch": 0.4498246169264446,
"grad_norm": 0.09341007471084595,
"learning_rate": 0.0004,
"loss": 2.0146405696868896,
"step": 2132
},
{
"epoch": 0.4502465912387583,
"grad_norm": 0.10219047963619232,
"learning_rate": 0.0004,
"loss": 2.0321600437164307,
"step": 2134
},
{
"epoch": 0.4506685655510721,
"grad_norm": 0.09620510041713715,
"learning_rate": 0.0004,
"loss": 2.0584912300109863,
"step": 2136
},
{
"epoch": 0.4510905398633858,
"grad_norm": 0.11491062492132187,
"learning_rate": 0.0004,
"loss": 2.043121814727783,
"step": 2138
},
{
"epoch": 0.45151251417569954,
"grad_norm": 0.10299600660800934,
"learning_rate": 0.0004,
"loss": 1.9601823091506958,
"step": 2140
},
{
"epoch": 0.4519344884880133,
"grad_norm": 0.1007743775844574,
"learning_rate": 0.0004,
"loss": 1.990552306175232,
"step": 2142
},
{
"epoch": 0.45235646280032704,
"grad_norm": 0.10193657875061035,
"learning_rate": 0.0004,
"loss": 2.0295917987823486,
"step": 2144
},
{
"epoch": 0.45277843711264076,
"grad_norm": 0.10482946038246155,
"learning_rate": 0.0004,
"loss": 2.053595781326294,
"step": 2146
},
{
"epoch": 0.4532004114249545,
"grad_norm": 0.10135386884212494,
"learning_rate": 0.0004,
"loss": 1.975598931312561,
"step": 2148
},
{
"epoch": 0.45362238573726826,
"grad_norm": 0.10648709535598755,
"learning_rate": 0.0004,
"loss": 1.94498610496521,
"step": 2150
},
{
"epoch": 0.454044360049582,
"grad_norm": 0.10134290903806686,
"learning_rate": 0.0004,
"loss": 1.995693564414978,
"step": 2152
},
{
"epoch": 0.4544663343618957,
"grad_norm": 0.10461129993200302,
"learning_rate": 0.0004,
"loss": 1.9785168170928955,
"step": 2154
},
{
"epoch": 0.4548883086742095,
"grad_norm": 0.10920045524835587,
"learning_rate": 0.0004,
"loss": 1.9855715036392212,
"step": 2156
},
{
"epoch": 0.4553102829865232,
"grad_norm": 0.09787940979003906,
"learning_rate": 0.0004,
"loss": 1.937028169631958,
"step": 2158
},
{
"epoch": 0.45573225729883693,
"grad_norm": 0.12362446635961533,
"learning_rate": 0.0004,
"loss": 2.0165865421295166,
"step": 2160
},
{
"epoch": 0.45615423161115065,
"grad_norm": 0.103085458278656,
"learning_rate": 0.0004,
"loss": 2.031452178955078,
"step": 2162
},
{
"epoch": 0.45657620592346443,
"grad_norm": 0.11231916397809982,
"learning_rate": 0.0004,
"loss": 2.0273871421813965,
"step": 2164
},
{
"epoch": 0.45699818023577815,
"grad_norm": 0.1059272512793541,
"learning_rate": 0.0004,
"loss": 1.9979679584503174,
"step": 2166
},
{
"epoch": 0.4574201545480919,
"grad_norm": 0.10523281991481781,
"learning_rate": 0.0004,
"loss": 1.89382004737854,
"step": 2168
},
{
"epoch": 0.45784212886040565,
"grad_norm": 0.10214272141456604,
"learning_rate": 0.0004,
"loss": 2.0123682022094727,
"step": 2170
},
{
"epoch": 0.4582641031727194,
"grad_norm": 0.09984613955020905,
"learning_rate": 0.0004,
"loss": 1.988513708114624,
"step": 2172
},
{
"epoch": 0.4586860774850331,
"grad_norm": 0.10118819028139114,
"learning_rate": 0.0004,
"loss": 1.9661070108413696,
"step": 2174
},
{
"epoch": 0.4591080517973468,
"grad_norm": 0.10467533767223358,
"learning_rate": 0.0004,
"loss": 1.9890488386154175,
"step": 2176
},
{
"epoch": 0.4595300261096606,
"grad_norm": 0.10961633920669556,
"learning_rate": 0.0004,
"loss": 1.9771366119384766,
"step": 2178
},
{
"epoch": 0.4599520004219743,
"grad_norm": 0.12882454693317413,
"learning_rate": 0.0004,
"loss": 1.984815239906311,
"step": 2180
},
{
"epoch": 0.46037397473428804,
"grad_norm": 0.10777653008699417,
"learning_rate": 0.0004,
"loss": 2.0275185108184814,
"step": 2182
},
{
"epoch": 0.46079594904660176,
"grad_norm": 0.11796388775110245,
"learning_rate": 0.0004,
"loss": 2.0496714115142822,
"step": 2184
},
{
"epoch": 0.46121792335891554,
"grad_norm": 0.10598327964544296,
"learning_rate": 0.0004,
"loss": 2.000676155090332,
"step": 2186
},
{
"epoch": 0.46163989767122926,
"grad_norm": 0.10859408974647522,
"learning_rate": 0.0004,
"loss": 2.060457229614258,
"step": 2188
},
{
"epoch": 0.462061871983543,
"grad_norm": 0.10652462393045425,
"learning_rate": 0.0004,
"loss": 1.986275553703308,
"step": 2190
},
{
"epoch": 0.46248384629585676,
"grad_norm": 0.10302762687206268,
"learning_rate": 0.0004,
"loss": 2.058629274368286,
"step": 2192
},
{
"epoch": 0.4629058206081705,
"grad_norm": 0.11035233736038208,
"learning_rate": 0.0004,
"loss": 2.061227798461914,
"step": 2194
},
{
"epoch": 0.4633277949204842,
"grad_norm": 0.10551932454109192,
"learning_rate": 0.0004,
"loss": 1.9984420537948608,
"step": 2196
},
{
"epoch": 0.4637497692327979,
"grad_norm": 0.10912052541971207,
"learning_rate": 0.0004,
"loss": 2.0092406272888184,
"step": 2198
},
{
"epoch": 0.4641717435451117,
"grad_norm": 0.10125407576560974,
"learning_rate": 0.0004,
"loss": 1.9692168235778809,
"step": 2200
},
{
"epoch": 0.4645937178574254,
"grad_norm": 0.10106629878282547,
"learning_rate": 0.0004,
"loss": 2.0404555797576904,
"step": 2202
},
{
"epoch": 0.46501569216973915,
"grad_norm": 0.10835615545511246,
"learning_rate": 0.0004,
"loss": 1.9469314813613892,
"step": 2204
},
{
"epoch": 0.4654376664820529,
"grad_norm": 0.12198732793331146,
"learning_rate": 0.0004,
"loss": 1.972214698791504,
"step": 2206
},
{
"epoch": 0.46585964079436665,
"grad_norm": 0.11745113134384155,
"learning_rate": 0.0004,
"loss": 1.925649642944336,
"step": 2208
},
{
"epoch": 0.46628161510668037,
"grad_norm": 0.10877996683120728,
"learning_rate": 0.0004,
"loss": 1.9826334714889526,
"step": 2210
},
{
"epoch": 0.4667035894189941,
"grad_norm": 0.10451096296310425,
"learning_rate": 0.0004,
"loss": 2.0056700706481934,
"step": 2212
},
{
"epoch": 0.46712556373130787,
"grad_norm": 0.11445613205432892,
"learning_rate": 0.0004,
"loss": 1.9893043041229248,
"step": 2214
},
{
"epoch": 0.4675475380436216,
"grad_norm": 0.0957336574792862,
"learning_rate": 0.0004,
"loss": 1.9916956424713135,
"step": 2216
},
{
"epoch": 0.4679695123559353,
"grad_norm": 0.11256985366344452,
"learning_rate": 0.0004,
"loss": 2.006793737411499,
"step": 2218
},
{
"epoch": 0.4683914866682491,
"grad_norm": 0.10433495044708252,
"learning_rate": 0.0004,
"loss": 2.026931047439575,
"step": 2220
},
{
"epoch": 0.4688134609805628,
"grad_norm": 0.11034650355577469,
"learning_rate": 0.0004,
"loss": 2.034743070602417,
"step": 2222
},
{
"epoch": 0.46923543529287653,
"grad_norm": 0.10520098358392715,
"learning_rate": 0.0004,
"loss": 1.9349656105041504,
"step": 2224
},
{
"epoch": 0.46965740960519026,
"grad_norm": 0.1050449013710022,
"learning_rate": 0.0004,
"loss": 2.034933567047119,
"step": 2226
},
{
"epoch": 0.47007938391750403,
"grad_norm": 0.108608178794384,
"learning_rate": 0.0004,
"loss": 1.9491335153579712,
"step": 2228
},
{
"epoch": 0.47050135822981776,
"grad_norm": 0.10998080670833588,
"learning_rate": 0.0004,
"loss": 1.966327428817749,
"step": 2230
},
{
"epoch": 0.4709233325421315,
"grad_norm": 0.11723295599222183,
"learning_rate": 0.0004,
"loss": 1.9978810548782349,
"step": 2232
},
{
"epoch": 0.47134530685444526,
"grad_norm": 0.11298651993274689,
"learning_rate": 0.0004,
"loss": 1.9546273946762085,
"step": 2234
},
{
"epoch": 0.471767281166759,
"grad_norm": 0.10582790523767471,
"learning_rate": 0.0004,
"loss": 1.9944738149642944,
"step": 2236
},
{
"epoch": 0.4721892554790727,
"grad_norm": 0.14811986684799194,
"learning_rate": 0.0004,
"loss": 2.006772994995117,
"step": 2238
},
{
"epoch": 0.4726112297913864,
"grad_norm": 0.10637582838535309,
"learning_rate": 0.0004,
"loss": 2.0300967693328857,
"step": 2240
},
{
"epoch": 0.4730332041037002,
"grad_norm": 0.09996522963047028,
"learning_rate": 0.0004,
"loss": 2.013122797012329,
"step": 2242
},
{
"epoch": 0.4734551784160139,
"grad_norm": 0.0978463739156723,
"learning_rate": 0.0004,
"loss": 2.048208475112915,
"step": 2244
},
{
"epoch": 0.47387715272832764,
"grad_norm": 0.09821776300668716,
"learning_rate": 0.0004,
"loss": 2.016819477081299,
"step": 2246
},
{
"epoch": 0.4742991270406414,
"grad_norm": 0.10800539702177048,
"learning_rate": 0.0004,
"loss": 1.99642014503479,
"step": 2248
},
{
"epoch": 0.47472110135295514,
"grad_norm": 0.0965987965464592,
"learning_rate": 0.0004,
"loss": 1.9972846508026123,
"step": 2250
},
{
"epoch": 0.47514307566526887,
"grad_norm": 0.09641961008310318,
"learning_rate": 0.0004,
"loss": 1.981523036956787,
"step": 2252
},
{
"epoch": 0.4755650499775826,
"grad_norm": 0.10403846204280853,
"learning_rate": 0.0004,
"loss": 1.9715113639831543,
"step": 2254
},
{
"epoch": 0.47598702428989637,
"grad_norm": 0.10933565348386765,
"learning_rate": 0.0004,
"loss": 1.9954628944396973,
"step": 2256
},
{
"epoch": 0.4764089986022101,
"grad_norm": 0.09355618059635162,
"learning_rate": 0.0004,
"loss": 2.017261266708374,
"step": 2258
},
{
"epoch": 0.4768309729145238,
"grad_norm": 0.10328681766986847,
"learning_rate": 0.0004,
"loss": 1.9323660135269165,
"step": 2260
},
{
"epoch": 0.4772529472268376,
"grad_norm": 0.09867209941148758,
"learning_rate": 0.0004,
"loss": 2.004274845123291,
"step": 2262
},
{
"epoch": 0.4776749215391513,
"grad_norm": 0.1116461530327797,
"learning_rate": 0.0004,
"loss": 1.9664078950881958,
"step": 2264
},
{
"epoch": 0.47809689585146503,
"grad_norm": 0.10858102887868881,
"learning_rate": 0.0004,
"loss": 1.9316598176956177,
"step": 2266
},
{
"epoch": 0.47851887016377875,
"grad_norm": 0.14670735597610474,
"learning_rate": 0.0004,
"loss": 1.9770874977111816,
"step": 2268
},
{
"epoch": 0.47894084447609253,
"grad_norm": 0.09770026057958603,
"learning_rate": 0.0004,
"loss": 2.062384843826294,
"step": 2270
},
{
"epoch": 0.47936281878840625,
"grad_norm": 0.14390859007835388,
"learning_rate": 0.0004,
"loss": 1.9850780963897705,
"step": 2272
},
{
"epoch": 0.47978479310072,
"grad_norm": 0.11551516503095627,
"learning_rate": 0.0004,
"loss": 2.036221981048584,
"step": 2274
},
{
"epoch": 0.48020676741303375,
"grad_norm": 0.1167362853884697,
"learning_rate": 0.0004,
"loss": 2.019850015640259,
"step": 2276
},
{
"epoch": 0.4806287417253475,
"grad_norm": 0.10289856791496277,
"learning_rate": 0.0004,
"loss": 2.013359785079956,
"step": 2278
},
{
"epoch": 0.4810507160376612,
"grad_norm": 0.10182593017816544,
"learning_rate": 0.0004,
"loss": 2.0449326038360596,
"step": 2280
},
{
"epoch": 0.4814726903499749,
"grad_norm": 0.10485699772834778,
"learning_rate": 0.0004,
"loss": 1.9769946336746216,
"step": 2282
},
{
"epoch": 0.4818946646622887,
"grad_norm": 0.1177060604095459,
"learning_rate": 0.0004,
"loss": 1.9607036113739014,
"step": 2284
},
{
"epoch": 0.4823166389746024,
"grad_norm": 0.12169861793518066,
"learning_rate": 0.0004,
"loss": 2.036513090133667,
"step": 2286
},
{
"epoch": 0.48273861328691614,
"grad_norm": 0.1090228334069252,
"learning_rate": 0.0004,
"loss": 1.994579553604126,
"step": 2288
},
{
"epoch": 0.4831605875992299,
"grad_norm": 0.11544869095087051,
"learning_rate": 0.0004,
"loss": 1.9541361331939697,
"step": 2290
},
{
"epoch": 0.48358256191154364,
"grad_norm": 0.1081470400094986,
"learning_rate": 0.0004,
"loss": 2.0582902431488037,
"step": 2292
},
{
"epoch": 0.48400453622385736,
"grad_norm": 0.10911247879266739,
"learning_rate": 0.0004,
"loss": 2.0111896991729736,
"step": 2294
},
{
"epoch": 0.4844265105361711,
"grad_norm": 0.1209435686469078,
"learning_rate": 0.0004,
"loss": 2.0389513969421387,
"step": 2296
},
{
"epoch": 0.48484848484848486,
"grad_norm": 0.10092119872570038,
"learning_rate": 0.0004,
"loss": 1.9691393375396729,
"step": 2298
},
{
"epoch": 0.4852704591607986,
"grad_norm": 0.11130089312791824,
"learning_rate": 0.0004,
"loss": 1.9911330938339233,
"step": 2300
},
{
"epoch": 0.4856924334731123,
"grad_norm": 0.10630028694868088,
"learning_rate": 0.0004,
"loss": 1.9568989276885986,
"step": 2302
},
{
"epoch": 0.4861144077854261,
"grad_norm": 0.11338246613740921,
"learning_rate": 0.0004,
"loss": 2.014009952545166,
"step": 2304
},
{
"epoch": 0.4865363820977398,
"grad_norm": 0.10607319325208664,
"learning_rate": 0.0004,
"loss": 2.0352091789245605,
"step": 2306
},
{
"epoch": 0.48695835641005353,
"grad_norm": 0.09900247305631638,
"learning_rate": 0.0004,
"loss": 1.9241132736206055,
"step": 2308
},
{
"epoch": 0.48738033072236725,
"grad_norm": 0.14401216804981232,
"learning_rate": 0.0004,
"loss": 1.9510775804519653,
"step": 2310
},
{
"epoch": 0.487802305034681,
"grad_norm": 0.11244620382785797,
"learning_rate": 0.0004,
"loss": 1.9602727890014648,
"step": 2312
},
{
"epoch": 0.48822427934699475,
"grad_norm": 0.11766840517520905,
"learning_rate": 0.0004,
"loss": 2.080589532852173,
"step": 2314
},
{
"epoch": 0.48864625365930847,
"grad_norm": 0.09981772303581238,
"learning_rate": 0.0004,
"loss": 1.9802720546722412,
"step": 2316
},
{
"epoch": 0.48906822797162225,
"grad_norm": 0.11033106595277786,
"learning_rate": 0.0004,
"loss": 2.0002493858337402,
"step": 2318
},
{
"epoch": 0.48949020228393597,
"grad_norm": 0.1256871521472931,
"learning_rate": 0.0004,
"loss": 2.022390365600586,
"step": 2320
},
{
"epoch": 0.4899121765962497,
"grad_norm": 0.10457857698202133,
"learning_rate": 0.0004,
"loss": 2.029090404510498,
"step": 2322
},
{
"epoch": 0.4903341509085634,
"grad_norm": 0.11304888874292374,
"learning_rate": 0.0004,
"loss": 2.0570192337036133,
"step": 2324
},
{
"epoch": 0.4907561252208772,
"grad_norm": 0.10582602769136429,
"learning_rate": 0.0004,
"loss": 1.9759361743927002,
"step": 2326
},
{
"epoch": 0.4911780995331909,
"grad_norm": 0.10448603332042694,
"learning_rate": 0.0004,
"loss": 2.0392684936523438,
"step": 2328
},
{
"epoch": 0.49160007384550464,
"grad_norm": 0.11255750060081482,
"learning_rate": 0.0004,
"loss": 1.9985418319702148,
"step": 2330
},
{
"epoch": 0.4920220481578184,
"grad_norm": 0.09878233075141907,
"learning_rate": 0.0004,
"loss": 1.9562793970108032,
"step": 2332
},
{
"epoch": 0.49244402247013214,
"grad_norm": 0.10621915757656097,
"learning_rate": 0.0004,
"loss": 1.9893743991851807,
"step": 2334
},
{
"epoch": 0.49286599678244586,
"grad_norm": 0.12052877247333527,
"learning_rate": 0.0004,
"loss": 2.04366135597229,
"step": 2336
},
{
"epoch": 0.4932879710947596,
"grad_norm": 0.10063917189836502,
"learning_rate": 0.0004,
"loss": 1.9235209226608276,
"step": 2338
},
{
"epoch": 0.49370994540707336,
"grad_norm": 0.14204511046409607,
"learning_rate": 0.0004,
"loss": 2.032701253890991,
"step": 2340
},
{
"epoch": 0.4941319197193871,
"grad_norm": 0.11547805368900299,
"learning_rate": 0.0004,
"loss": 2.006054639816284,
"step": 2342
},
{
"epoch": 0.4945538940317008,
"grad_norm": 0.11362501233816147,
"learning_rate": 0.0004,
"loss": 1.9638580083847046,
"step": 2344
},
{
"epoch": 0.4949758683440146,
"grad_norm": 0.10387091338634491,
"learning_rate": 0.0004,
"loss": 2.0145423412323,
"step": 2346
},
{
"epoch": 0.4953978426563283,
"grad_norm": 0.10092498362064362,
"learning_rate": 0.0004,
"loss": 1.9703021049499512,
"step": 2348
},
{
"epoch": 0.495819816968642,
"grad_norm": 0.11994252353906631,
"learning_rate": 0.0004,
"loss": 1.979360818862915,
"step": 2350
},
{
"epoch": 0.49624179128095575,
"grad_norm": 0.10781528800725937,
"learning_rate": 0.0004,
"loss": 1.98626708984375,
"step": 2352
},
{
"epoch": 0.4966637655932695,
"grad_norm": 0.10405923426151276,
"learning_rate": 0.0004,
"loss": 1.9539060592651367,
"step": 2354
},
{
"epoch": 0.49708573990558325,
"grad_norm": 0.10370170325040817,
"learning_rate": 0.0004,
"loss": 1.9100884199142456,
"step": 2356
},
{
"epoch": 0.49750771421789697,
"grad_norm": 0.10538933426141739,
"learning_rate": 0.0004,
"loss": 1.9766314029693604,
"step": 2358
},
{
"epoch": 0.49792968853021075,
"grad_norm": 0.10167869925498962,
"learning_rate": 0.0004,
"loss": 2.0158731937408447,
"step": 2360
},
{
"epoch": 0.49835166284252447,
"grad_norm": 0.0969918742775917,
"learning_rate": 0.0004,
"loss": 1.954291582107544,
"step": 2362
},
{
"epoch": 0.4987736371548382,
"grad_norm": 0.11679546535015106,
"learning_rate": 0.0004,
"loss": 1.9071389436721802,
"step": 2364
},
{
"epoch": 0.4991956114671519,
"grad_norm": 0.09594209492206573,
"learning_rate": 0.0004,
"loss": 2.0036873817443848,
"step": 2366
},
{
"epoch": 0.4996175857794657,
"grad_norm": 0.10378118604421616,
"learning_rate": 0.0004,
"loss": 2.0291147232055664,
"step": 2368
},
{
"epoch": 0.5000395600917794,
"grad_norm": 0.10891497135162354,
"learning_rate": 0.0004,
"loss": 1.9265341758728027,
"step": 2370
},
{
"epoch": 0.5004615344040931,
"grad_norm": 0.1099167913198471,
"learning_rate": 0.0004,
"loss": 2.0173258781433105,
"step": 2372
},
{
"epoch": 0.5008835087164069,
"grad_norm": 0.11504310369491577,
"learning_rate": 0.0004,
"loss": 1.986431360244751,
"step": 2374
},
{
"epoch": 0.5013054830287206,
"grad_norm": 0.10948949307203293,
"learning_rate": 0.0004,
"loss": 1.9303865432739258,
"step": 2376
},
{
"epoch": 0.5017274573410344,
"grad_norm": 0.1008806973695755,
"learning_rate": 0.0004,
"loss": 2.014810085296631,
"step": 2378
},
{
"epoch": 0.5021494316533481,
"grad_norm": 0.1239369586110115,
"learning_rate": 0.0004,
"loss": 1.993377923965454,
"step": 2380
},
{
"epoch": 0.5025714059656619,
"grad_norm": 0.10770989954471588,
"learning_rate": 0.0004,
"loss": 1.9748649597167969,
"step": 2382
},
{
"epoch": 0.5029933802779756,
"grad_norm": 0.10484594851732254,
"learning_rate": 0.0004,
"loss": 2.001687526702881,
"step": 2384
},
{
"epoch": 0.5034153545902893,
"grad_norm": 0.1216248944401741,
"learning_rate": 0.0004,
"loss": 1.9519157409667969,
"step": 2386
},
{
"epoch": 0.503837328902603,
"grad_norm": 0.11535459011793137,
"learning_rate": 0.0004,
"loss": 2.0479423999786377,
"step": 2388
},
{
"epoch": 0.5042593032149167,
"grad_norm": 0.10593268275260925,
"learning_rate": 0.0004,
"loss": 2.027888298034668,
"step": 2390
},
{
"epoch": 0.5046812775272306,
"grad_norm": 0.11867792159318924,
"learning_rate": 0.0004,
"loss": 1.9541786909103394,
"step": 2392
},
{
"epoch": 0.5051032518395443,
"grad_norm": 0.10116704553365707,
"learning_rate": 0.0004,
"loss": 1.9260892868041992,
"step": 2394
},
{
"epoch": 0.505525226151858,
"grad_norm": 0.10534977912902832,
"learning_rate": 0.0004,
"loss": 1.9572213888168335,
"step": 2396
},
{
"epoch": 0.5059472004641717,
"grad_norm": 0.11376344412565231,
"learning_rate": 0.0004,
"loss": 2.0021169185638428,
"step": 2398
},
{
"epoch": 0.5063691747764855,
"grad_norm": 0.11419486999511719,
"learning_rate": 0.0004,
"loss": 1.9917021989822388,
"step": 2400
},
{
"epoch": 0.5067911490887992,
"grad_norm": 0.10225561261177063,
"learning_rate": 0.0004,
"loss": 1.9665294885635376,
"step": 2402
},
{
"epoch": 0.5072131234011129,
"grad_norm": 0.10482429713010788,
"learning_rate": 0.0004,
"loss": 1.9898320436477661,
"step": 2404
},
{
"epoch": 0.5076350977134267,
"grad_norm": 0.09937436878681183,
"learning_rate": 0.0004,
"loss": 2.0081775188446045,
"step": 2406
},
{
"epoch": 0.5080570720257405,
"grad_norm": 0.10898911207914352,
"learning_rate": 0.0004,
"loss": 1.9538551568984985,
"step": 2408
},
{
"epoch": 0.5084790463380542,
"grad_norm": 0.10950437933206558,
"learning_rate": 0.0004,
"loss": 2.0497536659240723,
"step": 2410
},
{
"epoch": 0.5089010206503679,
"grad_norm": 0.12011754512786865,
"learning_rate": 0.0004,
"loss": 1.9786605834960938,
"step": 2412
},
{
"epoch": 0.5093229949626816,
"grad_norm": 0.10911696404218674,
"learning_rate": 0.0004,
"loss": 2.0652666091918945,
"step": 2414
},
{
"epoch": 0.5097449692749954,
"grad_norm": 0.10547386854887009,
"learning_rate": 0.0004,
"loss": 2.0417640209198,
"step": 2416
},
{
"epoch": 0.5101669435873091,
"grad_norm": 0.1104971393942833,
"learning_rate": 0.0004,
"loss": 1.964789628982544,
"step": 2418
},
{
"epoch": 0.5105889178996229,
"grad_norm": 0.10839339345693588,
"learning_rate": 0.0004,
"loss": 1.9659734964370728,
"step": 2420
},
{
"epoch": 0.5110108922119366,
"grad_norm": 0.11670142412185669,
"learning_rate": 0.0004,
"loss": 1.9137362241744995,
"step": 2422
},
{
"epoch": 0.5114328665242504,
"grad_norm": 0.12092326581478119,
"learning_rate": 0.0004,
"loss": 2.0218706130981445,
"step": 2424
},
{
"epoch": 0.5118548408365641,
"grad_norm": 0.10749033838510513,
"learning_rate": 0.0004,
"loss": 1.9647752046585083,
"step": 2426
},
{
"epoch": 0.5122768151488778,
"grad_norm": 0.11355699598789215,
"learning_rate": 0.0004,
"loss": 2.0163910388946533,
"step": 2428
},
{
"epoch": 0.5126987894611915,
"grad_norm": 0.12003671377897263,
"learning_rate": 0.0004,
"loss": 2.034925699234009,
"step": 2430
},
{
"epoch": 0.5131207637735052,
"grad_norm": 0.10423602163791656,
"learning_rate": 0.0004,
"loss": 1.945415735244751,
"step": 2432
},
{
"epoch": 0.5135427380858191,
"grad_norm": 0.11564397811889648,
"learning_rate": 0.0004,
"loss": 1.9838409423828125,
"step": 2434
},
{
"epoch": 0.5139647123981328,
"grad_norm": 0.10946027934551239,
"learning_rate": 0.0004,
"loss": 1.9665216207504272,
"step": 2436
},
{
"epoch": 0.5143866867104465,
"grad_norm": 0.09866265207529068,
"learning_rate": 0.0004,
"loss": 2.0311176776885986,
"step": 2438
},
{
"epoch": 0.5148086610227602,
"grad_norm": 0.11777540296316147,
"learning_rate": 0.0004,
"loss": 1.9856728315353394,
"step": 2440
},
{
"epoch": 0.515230635335074,
"grad_norm": 0.11609726399183273,
"learning_rate": 0.0004,
"loss": 1.9872740507125854,
"step": 2442
},
{
"epoch": 0.5156526096473877,
"grad_norm": 0.10349444299936295,
"learning_rate": 0.0004,
"loss": 1.9598498344421387,
"step": 2444
},
{
"epoch": 0.5160745839597014,
"grad_norm": 0.10568153858184814,
"learning_rate": 0.0004,
"loss": 1.9816974401474,
"step": 2446
},
{
"epoch": 0.5164965582720152,
"grad_norm": 0.11174038797616959,
"learning_rate": 0.0004,
"loss": 2.0950863361358643,
"step": 2448
},
{
"epoch": 0.516918532584329,
"grad_norm": 0.10832858085632324,
"learning_rate": 0.0004,
"loss": 1.977849006652832,
"step": 2450
},
{
"epoch": 0.5173405068966427,
"grad_norm": 0.10564068704843521,
"learning_rate": 0.0004,
"loss": 2.010291337966919,
"step": 2452
},
{
"epoch": 0.5177624812089564,
"grad_norm": 0.102881520986557,
"learning_rate": 0.0004,
"loss": 1.971081256866455,
"step": 2454
},
{
"epoch": 0.5181844555212701,
"grad_norm": 0.1098896786570549,
"learning_rate": 0.0004,
"loss": 1.9491922855377197,
"step": 2456
},
{
"epoch": 0.5186064298335838,
"grad_norm": 0.09652575105428696,
"learning_rate": 0.0004,
"loss": 2.027902364730835,
"step": 2458
},
{
"epoch": 0.5190284041458976,
"grad_norm": 0.1103406697511673,
"learning_rate": 0.0004,
"loss": 1.9653058052062988,
"step": 2460
},
{
"epoch": 0.5194503784582114,
"grad_norm": 0.09861093014478683,
"learning_rate": 0.0004,
"loss": 1.9974720478057861,
"step": 2462
},
{
"epoch": 0.5198723527705251,
"grad_norm": 0.1260729730129242,
"learning_rate": 0.0004,
"loss": 1.945346474647522,
"step": 2464
},
{
"epoch": 0.5202943270828388,
"grad_norm": 0.1017293930053711,
"learning_rate": 0.0004,
"loss": 1.9658123254776,
"step": 2466
},
{
"epoch": 0.5207163013951526,
"grad_norm": 0.11294218897819519,
"learning_rate": 0.0004,
"loss": 1.9527469873428345,
"step": 2468
},
{
"epoch": 0.5211382757074663,
"grad_norm": 0.11961628496646881,
"learning_rate": 0.0004,
"loss": 1.9800727367401123,
"step": 2470
},
{
"epoch": 0.52156025001978,
"grad_norm": 0.10792697221040726,
"learning_rate": 0.0004,
"loss": 1.9181640148162842,
"step": 2472
},
{
"epoch": 0.5219822243320937,
"grad_norm": 0.10481695830821991,
"learning_rate": 0.0004,
"loss": 1.9995081424713135,
"step": 2474
},
{
"epoch": 0.5224041986444076,
"grad_norm": 0.1029142513871193,
"learning_rate": 0.0004,
"loss": 1.9846267700195312,
"step": 2476
},
{
"epoch": 0.5228261729567213,
"grad_norm": 0.11301450431346893,
"learning_rate": 0.0004,
"loss": 1.9883666038513184,
"step": 2478
},
{
"epoch": 0.523248147269035,
"grad_norm": 0.10226210206747055,
"learning_rate": 0.0004,
"loss": 2.0130116939544678,
"step": 2480
},
{
"epoch": 0.5236701215813487,
"grad_norm": 0.11259128153324127,
"learning_rate": 0.0004,
"loss": 1.9783650636672974,
"step": 2482
},
{
"epoch": 0.5240920958936625,
"grad_norm": 0.11570736020803452,
"learning_rate": 0.0004,
"loss": 1.9685189723968506,
"step": 2484
},
{
"epoch": 0.5245140702059762,
"grad_norm": 0.0965210422873497,
"learning_rate": 0.0004,
"loss": 2.00887393951416,
"step": 2486
},
{
"epoch": 0.5249360445182899,
"grad_norm": 0.10556851327419281,
"learning_rate": 0.0004,
"loss": 2.0122082233428955,
"step": 2488
},
{
"epoch": 0.5253580188306037,
"grad_norm": 0.10208205878734589,
"learning_rate": 0.0004,
"loss": 1.9899144172668457,
"step": 2490
},
{
"epoch": 0.5257799931429175,
"grad_norm": 0.10466782003641129,
"learning_rate": 0.0004,
"loss": 2.028722047805786,
"step": 2492
},
{
"epoch": 0.5262019674552312,
"grad_norm": 0.11742915213108063,
"learning_rate": 0.0004,
"loss": 1.9135500192642212,
"step": 2494
},
{
"epoch": 0.5266239417675449,
"grad_norm": 0.12357254326343536,
"learning_rate": 0.0004,
"loss": 1.9647915363311768,
"step": 2496
},
{
"epoch": 0.5270459160798586,
"grad_norm": 0.12987452745437622,
"learning_rate": 0.0004,
"loss": 1.950881004333496,
"step": 2498
},
{
"epoch": 0.5274678903921723,
"grad_norm": 0.10171229392290115,
"learning_rate": 0.0004,
"loss": 1.9946374893188477,
"step": 2500
},
{
"epoch": 0.5278898647044861,
"grad_norm": 0.11121941357851028,
"learning_rate": 0.0004,
"loss": 1.9161549806594849,
"step": 2502
},
{
"epoch": 0.5283118390167999,
"grad_norm": 0.11911734938621521,
"learning_rate": 0.0004,
"loss": 2.05631685256958,
"step": 2504
},
{
"epoch": 0.5287338133291136,
"grad_norm": 0.114515520632267,
"learning_rate": 0.0004,
"loss": 1.9781222343444824,
"step": 2506
},
{
"epoch": 0.5291557876414273,
"grad_norm": 0.09910862147808075,
"learning_rate": 0.0004,
"loss": 1.9896719455718994,
"step": 2508
},
{
"epoch": 0.5295777619537411,
"grad_norm": 0.10130604356527328,
"learning_rate": 0.0004,
"loss": 1.961145043373108,
"step": 2510
},
{
"epoch": 0.5299997362660548,
"grad_norm": 0.11756748706102371,
"learning_rate": 0.0004,
"loss": 1.9919493198394775,
"step": 2512
},
{
"epoch": 0.5304217105783685,
"grad_norm": 0.10637060552835464,
"learning_rate": 0.00039982046678635545,
"loss": 1.946498155593872,
"step": 2514
},
{
"epoch": 0.5308436848906822,
"grad_norm": 0.10987117886543274,
"learning_rate": 0.0003994614003590664,
"loss": 2.048797369003296,
"step": 2516
},
{
"epoch": 0.5312656592029961,
"grad_norm": 0.13016334176063538,
"learning_rate": 0.0003991023339317774,
"loss": 1.9630935192108154,
"step": 2518
},
{
"epoch": 0.5316876335153098,
"grad_norm": 0.12178076058626175,
"learning_rate": 0.00039874326750448836,
"loss": 2.0223701000213623,
"step": 2520
},
{
"epoch": 0.5321096078276235,
"grad_norm": 0.12135455012321472,
"learning_rate": 0.00039838420107719927,
"loss": 2.002438545227051,
"step": 2522
},
{
"epoch": 0.5325315821399372,
"grad_norm": 0.10319202393293381,
"learning_rate": 0.00039802513464991024,
"loss": 2.0108494758605957,
"step": 2524
},
{
"epoch": 0.532953556452251,
"grad_norm": 0.09335236251354218,
"learning_rate": 0.0003976660682226212,
"loss": 1.9903230667114258,
"step": 2526
},
{
"epoch": 0.5333755307645647,
"grad_norm": 0.09771422296762466,
"learning_rate": 0.0003973070017953322,
"loss": 2.0218493938446045,
"step": 2528
},
{
"epoch": 0.5337975050768784,
"grad_norm": 0.10455936938524246,
"learning_rate": 0.0003969479353680431,
"loss": 1.9547443389892578,
"step": 2530
},
{
"epoch": 0.5342194793891922,
"grad_norm": 0.10861096531152725,
"learning_rate": 0.00039658886894075406,
"loss": 1.9429821968078613,
"step": 2532
},
{
"epoch": 0.534641453701506,
"grad_norm": 0.11580679565668106,
"learning_rate": 0.00039622980251346503,
"loss": 1.9221971035003662,
"step": 2534
},
{
"epoch": 0.5350634280138197,
"grad_norm": 0.11841796338558197,
"learning_rate": 0.000395870736086176,
"loss": 1.9951834678649902,
"step": 2536
},
{
"epoch": 0.5354854023261334,
"grad_norm": 0.105585016310215,
"learning_rate": 0.0003955116696588869,
"loss": 1.9619605541229248,
"step": 2538
},
{
"epoch": 0.5359073766384471,
"grad_norm": 0.11960457265377045,
"learning_rate": 0.0003951526032315979,
"loss": 2.0006375312805176,
"step": 2540
},
{
"epoch": 0.5363293509507608,
"grad_norm": 0.1099238321185112,
"learning_rate": 0.0003947935368043088,
"loss": 1.9983689785003662,
"step": 2542
},
{
"epoch": 0.5367513252630746,
"grad_norm": 0.11364013701677322,
"learning_rate": 0.00039443447037701976,
"loss": 1.952035903930664,
"step": 2544
},
{
"epoch": 0.5371732995753884,
"grad_norm": 0.10834506899118423,
"learning_rate": 0.0003940754039497307,
"loss": 2.0013222694396973,
"step": 2546
},
{
"epoch": 0.5375952738877021,
"grad_norm": 0.12964634597301483,
"learning_rate": 0.00039371633752244165,
"loss": 1.9360271692276,
"step": 2548
},
{
"epoch": 0.5380172482000158,
"grad_norm": 0.1075846254825592,
"learning_rate": 0.0003933572710951526,
"loss": 1.9721664190292358,
"step": 2550
},
{
"epoch": 0.5384392225123296,
"grad_norm": 0.1187378317117691,
"learning_rate": 0.0003929982046678636,
"loss": 1.9838147163391113,
"step": 2552
},
{
"epoch": 0.5388611968246433,
"grad_norm": 0.10463961213827133,
"learning_rate": 0.0003926391382405745,
"loss": 1.9461811780929565,
"step": 2554
},
{
"epoch": 0.539283171136957,
"grad_norm": 0.10743307322263718,
"learning_rate": 0.00039228007181328547,
"loss": 1.9757204055786133,
"step": 2556
},
{
"epoch": 0.5397051454492707,
"grad_norm": 0.10600344091653824,
"learning_rate": 0.00039192100538599644,
"loss": 1.976940631866455,
"step": 2558
},
{
"epoch": 0.5401271197615846,
"grad_norm": 0.10643648356199265,
"learning_rate": 0.0003915619389587074,
"loss": 1.9855600595474243,
"step": 2560
},
{
"epoch": 0.5405490940738983,
"grad_norm": 0.1129879355430603,
"learning_rate": 0.0003912028725314183,
"loss": 1.952157735824585,
"step": 2562
},
{
"epoch": 0.540971068386212,
"grad_norm": 0.129954531788826,
"learning_rate": 0.0003908438061041293,
"loss": 2.0546724796295166,
"step": 2564
},
{
"epoch": 0.5413930426985257,
"grad_norm": 0.10813000798225403,
"learning_rate": 0.00039048473967684026,
"loss": 1.9646235704421997,
"step": 2566
},
{
"epoch": 0.5418150170108395,
"grad_norm": 0.10587615519762039,
"learning_rate": 0.0003901256732495512,
"loss": 1.955272912979126,
"step": 2568
},
{
"epoch": 0.5422369913231532,
"grad_norm": 0.11749233305454254,
"learning_rate": 0.00038976660682226214,
"loss": 2.0047690868377686,
"step": 2570
},
{
"epoch": 0.5426589656354669,
"grad_norm": 0.1152234822511673,
"learning_rate": 0.00038940754039497305,
"loss": 2.021273374557495,
"step": 2572
},
{
"epoch": 0.5430809399477807,
"grad_norm": 0.10496938973665237,
"learning_rate": 0.000389048473967684,
"loss": 1.9686576128005981,
"step": 2574
},
{
"epoch": 0.5435029142600944,
"grad_norm": 0.10399749130010605,
"learning_rate": 0.000388689407540395,
"loss": 1.9918639659881592,
"step": 2576
},
{
"epoch": 0.5439248885724082,
"grad_norm": 0.09593001753091812,
"learning_rate": 0.00038833034111310596,
"loss": 1.9384801387786865,
"step": 2578
},
{
"epoch": 0.5443468628847219,
"grad_norm": 0.11743129044771194,
"learning_rate": 0.0003879712746858169,
"loss": 1.986926555633545,
"step": 2580
},
{
"epoch": 0.5447688371970356,
"grad_norm": 0.11106555163860321,
"learning_rate": 0.00038761220825852784,
"loss": 1.9247616529464722,
"step": 2582
},
{
"epoch": 0.5451908115093493,
"grad_norm": 0.10058481991291046,
"learning_rate": 0.0003872531418312388,
"loss": 1.9998693466186523,
"step": 2584
},
{
"epoch": 0.5456127858216631,
"grad_norm": 0.10874099284410477,
"learning_rate": 0.0003868940754039497,
"loss": 2.022087812423706,
"step": 2586
},
{
"epoch": 0.5460347601339769,
"grad_norm": 0.11434465646743774,
"learning_rate": 0.0003865350089766607,
"loss": 1.9178524017333984,
"step": 2588
},
{
"epoch": 0.5464567344462906,
"grad_norm": 0.10662268102169037,
"learning_rate": 0.00038617594254937166,
"loss": 2.0450174808502197,
"step": 2590
},
{
"epoch": 0.5468787087586043,
"grad_norm": 0.10540720075368881,
"learning_rate": 0.00038581687612208263,
"loss": 1.9318300485610962,
"step": 2592
},
{
"epoch": 0.5473006830709181,
"grad_norm": 0.10311566293239594,
"learning_rate": 0.00038545780969479355,
"loss": 1.9866130352020264,
"step": 2594
},
{
"epoch": 0.5477226573832318,
"grad_norm": 0.10430900007486343,
"learning_rate": 0.0003850987432675045,
"loss": 2.0241453647613525,
"step": 2596
},
{
"epoch": 0.5481446316955455,
"grad_norm": 0.10165537893772125,
"learning_rate": 0.0003847396768402155,
"loss": 2.010296583175659,
"step": 2598
},
{
"epoch": 0.5485666060078592,
"grad_norm": 0.10807012021541595,
"learning_rate": 0.0003843806104129264,
"loss": 2.0034232139587402,
"step": 2600
},
{
"epoch": 0.5489885803201731,
"grad_norm": 0.12723782658576965,
"learning_rate": 0.00038402154398563737,
"loss": 1.9255671501159668,
"step": 2602
},
{
"epoch": 0.5494105546324868,
"grad_norm": 0.09370072931051254,
"learning_rate": 0.0003836624775583483,
"loss": 1.9674731492996216,
"step": 2604
},
{
"epoch": 0.5498325289448005,
"grad_norm": 0.12219969928264618,
"learning_rate": 0.00038330341113105925,
"loss": 1.9914250373840332,
"step": 2606
},
{
"epoch": 0.5502545032571142,
"grad_norm": 0.09217988699674606,
"learning_rate": 0.0003829443447037702,
"loss": 1.9637184143066406,
"step": 2608
},
{
"epoch": 0.550676477569428,
"grad_norm": 0.10950630903244019,
"learning_rate": 0.0003825852782764812,
"loss": 1.9374934434890747,
"step": 2610
},
{
"epoch": 0.5510984518817417,
"grad_norm": 0.09815633296966553,
"learning_rate": 0.0003822262118491921,
"loss": 1.9912998676300049,
"step": 2612
},
{
"epoch": 0.5515204261940554,
"grad_norm": 0.09668152034282684,
"learning_rate": 0.00038186714542190307,
"loss": 1.995656132698059,
"step": 2614
},
{
"epoch": 0.5519424005063692,
"grad_norm": 0.11295585334300995,
"learning_rate": 0.00038150807899461404,
"loss": 1.9732999801635742,
"step": 2616
},
{
"epoch": 0.552364374818683,
"grad_norm": 0.11746437847614288,
"learning_rate": 0.000381149012567325,
"loss": 1.997145175933838,
"step": 2618
},
{
"epoch": 0.5527863491309967,
"grad_norm": 0.11727092415094376,
"learning_rate": 0.0003807899461400359,
"loss": 2.006805658340454,
"step": 2620
},
{
"epoch": 0.5532083234433104,
"grad_norm": 0.10990002006292343,
"learning_rate": 0.0003804308797127469,
"loss": 1.9893418550491333,
"step": 2622
},
{
"epoch": 0.5536302977556241,
"grad_norm": 0.11007990688085556,
"learning_rate": 0.00038007181328545786,
"loss": 1.9090021848678589,
"step": 2624
},
{
"epoch": 0.5540522720679378,
"grad_norm": 0.11116435378789902,
"learning_rate": 0.0003797127468581688,
"loss": 2.0294933319091797,
"step": 2626
},
{
"epoch": 0.5544742463802516,
"grad_norm": 0.11658182740211487,
"learning_rate": 0.0003793536804308797,
"loss": 1.9707906246185303,
"step": 2628
},
{
"epoch": 0.5548962206925654,
"grad_norm": 0.11011900752782822,
"learning_rate": 0.00037899461400359066,
"loss": 2.0327563285827637,
"step": 2630
},
{
"epoch": 0.5553181950048791,
"grad_norm": 0.10226181149482727,
"learning_rate": 0.0003786355475763016,
"loss": 2.015045166015625,
"step": 2632
},
{
"epoch": 0.5557401693171928,
"grad_norm": 0.11436307430267334,
"learning_rate": 0.0003782764811490126,
"loss": 1.9695391654968262,
"step": 2634
},
{
"epoch": 0.5561621436295066,
"grad_norm": 0.10111020505428314,
"learning_rate": 0.0003779174147217235,
"loss": 1.9782130718231201,
"step": 2636
},
{
"epoch": 0.5565841179418203,
"grad_norm": 0.11542998254299164,
"learning_rate": 0.0003775583482944345,
"loss": 2.021540880203247,
"step": 2638
},
{
"epoch": 0.557006092254134,
"grad_norm": 0.0996742844581604,
"learning_rate": 0.00037719928186714545,
"loss": 2.0101797580718994,
"step": 2640
},
{
"epoch": 0.5574280665664477,
"grad_norm": 0.1194087490439415,
"learning_rate": 0.0003768402154398564,
"loss": 2.053339958190918,
"step": 2642
},
{
"epoch": 0.5578500408787616,
"grad_norm": 0.12618958950042725,
"learning_rate": 0.00037648114901256733,
"loss": 1.9788166284561157,
"step": 2644
},
{
"epoch": 0.5582720151910753,
"grad_norm": 0.104707732796669,
"learning_rate": 0.0003761220825852783,
"loss": 1.9036158323287964,
"step": 2646
},
{
"epoch": 0.558693989503389,
"grad_norm": 0.10827819257974625,
"learning_rate": 0.00037576301615798927,
"loss": 1.9888815879821777,
"step": 2648
},
{
"epoch": 0.5591159638157027,
"grad_norm": 0.10772369056940079,
"learning_rate": 0.00037540394973070024,
"loss": 2.028395652770996,
"step": 2650
},
{
"epoch": 0.5595379381280164,
"grad_norm": 0.11533436924219131,
"learning_rate": 0.00037504488330341115,
"loss": 1.980991244316101,
"step": 2652
},
{
"epoch": 0.5599599124403302,
"grad_norm": 0.10628778487443924,
"learning_rate": 0.0003746858168761221,
"loss": 2.019418239593506,
"step": 2654
},
{
"epoch": 0.5603818867526439,
"grad_norm": 0.12098736315965652,
"learning_rate": 0.00037432675044883303,
"loss": 1.966699242591858,
"step": 2656
},
{
"epoch": 0.5608038610649577,
"grad_norm": 0.11796276271343231,
"learning_rate": 0.000373967684021544,
"loss": 1.9171817302703857,
"step": 2658
},
{
"epoch": 0.5612258353772714,
"grad_norm": 0.11030521243810654,
"learning_rate": 0.0003736086175942549,
"loss": 1.897729516029358,
"step": 2660
},
{
"epoch": 0.5616478096895852,
"grad_norm": 0.10991235077381134,
"learning_rate": 0.0003732495511669659,
"loss": 1.9078779220581055,
"step": 2662
},
{
"epoch": 0.5620697840018989,
"grad_norm": 0.1217423602938652,
"learning_rate": 0.00037289048473967685,
"loss": 2.041266918182373,
"step": 2664
},
{
"epoch": 0.5624917583142126,
"grad_norm": 0.11433570832014084,
"learning_rate": 0.0003725314183123878,
"loss": 1.9241865873336792,
"step": 2666
},
{
"epoch": 0.5629137326265263,
"grad_norm": 0.1101423129439354,
"learning_rate": 0.00037217235188509874,
"loss": 2.0498433113098145,
"step": 2668
},
{
"epoch": 0.56333570693884,
"grad_norm": 0.10628620535135269,
"learning_rate": 0.0003718132854578097,
"loss": 1.9455523490905762,
"step": 2670
},
{
"epoch": 0.5637576812511539,
"grad_norm": 0.10538875311613083,
"learning_rate": 0.0003714542190305207,
"loss": 2.0205154418945312,
"step": 2672
},
{
"epoch": 0.5641796555634676,
"grad_norm": 0.10258401930332184,
"learning_rate": 0.00037109515260323164,
"loss": 2.0379655361175537,
"step": 2674
},
{
"epoch": 0.5646016298757813,
"grad_norm": 0.10649662464857101,
"learning_rate": 0.00037073608617594256,
"loss": 2.000037431716919,
"step": 2676
},
{
"epoch": 0.565023604188095,
"grad_norm": 0.10618351399898529,
"learning_rate": 0.0003703770197486535,
"loss": 1.9426666498184204,
"step": 2678
},
{
"epoch": 0.5654455785004088,
"grad_norm": 0.09471061080694199,
"learning_rate": 0.0003700179533213645,
"loss": 1.974961757659912,
"step": 2680
},
{
"epoch": 0.5658675528127225,
"grad_norm": 0.10357743501663208,
"learning_rate": 0.00036965888689407546,
"loss": 1.9704082012176514,
"step": 2682
},
{
"epoch": 0.5662895271250362,
"grad_norm": 0.09940458089113235,
"learning_rate": 0.0003692998204667864,
"loss": 2.000023126602173,
"step": 2684
},
{
"epoch": 0.56671150143735,
"grad_norm": 0.10300898551940918,
"learning_rate": 0.0003689407540394973,
"loss": 1.933704137802124,
"step": 2686
},
{
"epoch": 0.5671334757496638,
"grad_norm": 0.13511933386325836,
"learning_rate": 0.00036858168761220826,
"loss": 1.9599982500076294,
"step": 2688
},
{
"epoch": 0.5675554500619775,
"grad_norm": 0.09282592684030533,
"learning_rate": 0.00036822262118491923,
"loss": 1.9441190958023071,
"step": 2690
},
{
"epoch": 0.5679774243742912,
"grad_norm": 0.11319000273942947,
"learning_rate": 0.00036786355475763014,
"loss": 2.0718369483947754,
"step": 2692
},
{
"epoch": 0.5683993986866049,
"grad_norm": 0.10511026531457901,
"learning_rate": 0.0003675044883303411,
"loss": 1.9435533285140991,
"step": 2694
},
{
"epoch": 0.5688213729989187,
"grad_norm": 0.12270218878984451,
"learning_rate": 0.0003671454219030521,
"loss": 2.0179195404052734,
"step": 2696
},
{
"epoch": 0.5692433473112324,
"grad_norm": 0.11975985020399094,
"learning_rate": 0.00036678635547576305,
"loss": 1.955338716506958,
"step": 2698
},
{
"epoch": 0.5696653216235462,
"grad_norm": 0.10499905049800873,
"learning_rate": 0.00036642728904847397,
"loss": 2.017014503479004,
"step": 2700
},
{
"epoch": 0.5700872959358599,
"grad_norm": 0.10907460004091263,
"learning_rate": 0.00036606822262118493,
"loss": 2.0178041458129883,
"step": 2702
},
{
"epoch": 0.5705092702481737,
"grad_norm": 0.10897300392389297,
"learning_rate": 0.0003657091561938959,
"loss": 2.012464761734009,
"step": 2704
},
{
"epoch": 0.5709312445604874,
"grad_norm": 0.11563042551279068,
"learning_rate": 0.00036535008976660687,
"loss": 1.962874174118042,
"step": 2706
},
{
"epoch": 0.5713532188728011,
"grad_norm": 0.09930475801229477,
"learning_rate": 0.0003649910233393178,
"loss": 1.9258553981781006,
"step": 2708
},
{
"epoch": 0.5717751931851148,
"grad_norm": 0.10951299965381622,
"learning_rate": 0.00036463195691202875,
"loss": 2.0046613216400146,
"step": 2710
},
{
"epoch": 0.5721971674974285,
"grad_norm": 0.12109868228435516,
"learning_rate": 0.0003642728904847397,
"loss": 1.957732915878296,
"step": 2712
},
{
"epoch": 0.5726191418097424,
"grad_norm": 0.10252423584461212,
"learning_rate": 0.00036391382405745064,
"loss": 2.0265414714813232,
"step": 2714
},
{
"epoch": 0.5730411161220561,
"grad_norm": 0.10182531177997589,
"learning_rate": 0.0003635547576301616,
"loss": 1.9260196685791016,
"step": 2716
},
{
"epoch": 0.5734630904343698,
"grad_norm": 0.09731879830360413,
"learning_rate": 0.0003631956912028725,
"loss": 1.9630494117736816,
"step": 2718
},
{
"epoch": 0.5738850647466835,
"grad_norm": 0.10727456212043762,
"learning_rate": 0.0003628366247755835,
"loss": 1.9906947612762451,
"step": 2720
},
{
"epoch": 0.5743070390589973,
"grad_norm": 0.1124085932970047,
"learning_rate": 0.00036247755834829446,
"loss": 1.9550503492355347,
"step": 2722
},
{
"epoch": 0.574729013371311,
"grad_norm": 0.10417519509792328,
"learning_rate": 0.00036211849192100537,
"loss": 1.9608229398727417,
"step": 2724
},
{
"epoch": 0.5751509876836247,
"grad_norm": 0.11434590071439743,
"learning_rate": 0.00036175942549371634,
"loss": 1.9783860445022583,
"step": 2726
},
{
"epoch": 0.5755729619959385,
"grad_norm": 0.1182977482676506,
"learning_rate": 0.0003614003590664273,
"loss": 1.9148451089859009,
"step": 2728
},
{
"epoch": 0.5759949363082523,
"grad_norm": 0.10166190564632416,
"learning_rate": 0.0003610412926391383,
"loss": 2.037506580352783,
"step": 2730
},
{
"epoch": 0.576416910620566,
"grad_norm": 0.10537232458591461,
"learning_rate": 0.0003606822262118492,
"loss": 2.0285391807556152,
"step": 2732
},
{
"epoch": 0.5768388849328797,
"grad_norm": 0.10557826608419418,
"learning_rate": 0.00036032315978456016,
"loss": 1.9756265878677368,
"step": 2734
},
{
"epoch": 0.5772608592451934,
"grad_norm": 0.116741843521595,
"learning_rate": 0.00035996409335727113,
"loss": 2.0573348999023438,
"step": 2736
},
{
"epoch": 0.5776828335575072,
"grad_norm": 0.10908560454845428,
"learning_rate": 0.0003596050269299821,
"loss": 1.9679235219955444,
"step": 2738
},
{
"epoch": 0.5781048078698209,
"grad_norm": 0.10668747872114182,
"learning_rate": 0.000359245960502693,
"loss": 1.9711685180664062,
"step": 2740
},
{
"epoch": 0.5785267821821347,
"grad_norm": 0.10489673912525177,
"learning_rate": 0.00035888689407540393,
"loss": 2.012512683868408,
"step": 2742
},
{
"epoch": 0.5789487564944484,
"grad_norm": 0.10706515610218048,
"learning_rate": 0.0003585278276481149,
"loss": 1.9759628772735596,
"step": 2744
},
{
"epoch": 0.5793707308067622,
"grad_norm": 0.10948754101991653,
"learning_rate": 0.00035816876122082587,
"loss": 1.8934252262115479,
"step": 2746
},
{
"epoch": 0.5797927051190759,
"grad_norm": 0.12581317126750946,
"learning_rate": 0.00035780969479353683,
"loss": 2.0331382751464844,
"step": 2748
},
{
"epoch": 0.5802146794313896,
"grad_norm": 0.10614161193370819,
"learning_rate": 0.00035745062836624775,
"loss": 1.9832097291946411,
"step": 2750
},
{
"epoch": 0.5806366537437033,
"grad_norm": 0.11300469934940338,
"learning_rate": 0.0003570915619389587,
"loss": 1.9601486921310425,
"step": 2752
},
{
"epoch": 0.581058628056017,
"grad_norm": 0.11584898829460144,
"learning_rate": 0.0003567324955116697,
"loss": 1.9560260772705078,
"step": 2754
},
{
"epoch": 0.5814806023683309,
"grad_norm": 0.10741697996854782,
"learning_rate": 0.00035637342908438065,
"loss": 1.934539556503296,
"step": 2756
},
{
"epoch": 0.5819025766806446,
"grad_norm": 0.10334764420986176,
"learning_rate": 0.00035601436265709157,
"loss": 1.9862955808639526,
"step": 2758
},
{
"epoch": 0.5823245509929583,
"grad_norm": 0.10299753397703171,
"learning_rate": 0.00035565529622980254,
"loss": 1.9381566047668457,
"step": 2760
},
{
"epoch": 0.582746525305272,
"grad_norm": 0.11009559035301208,
"learning_rate": 0.0003552962298025135,
"loss": 1.977015495300293,
"step": 2762
},
{
"epoch": 0.5831684996175858,
"grad_norm": 0.11602216958999634,
"learning_rate": 0.0003549371633752244,
"loss": 2.032754421234131,
"step": 2764
},
{
"epoch": 0.5835904739298995,
"grad_norm": 0.115249402821064,
"learning_rate": 0.0003545780969479354,
"loss": 2.0061745643615723,
"step": 2766
},
{
"epoch": 0.5840124482422132,
"grad_norm": 0.10980018228292465,
"learning_rate": 0.00035421903052064636,
"loss": 1.9890724420547485,
"step": 2768
},
{
"epoch": 0.584434422554527,
"grad_norm": 0.1161719411611557,
"learning_rate": 0.0003538599640933573,
"loss": 1.916581153869629,
"step": 2770
},
{
"epoch": 0.5848563968668408,
"grad_norm": 0.10902439057826996,
"learning_rate": 0.00035350089766606824,
"loss": 2.0256941318511963,
"step": 2772
},
{
"epoch": 0.5852783711791545,
"grad_norm": 0.10459676384925842,
"learning_rate": 0.00035314183123877916,
"loss": 1.9137017726898193,
"step": 2774
},
{
"epoch": 0.5857003454914682,
"grad_norm": 0.1014266386628151,
"learning_rate": 0.0003527827648114901,
"loss": 1.9476536512374878,
"step": 2776
},
{
"epoch": 0.5861223198037819,
"grad_norm": 0.11206267774105072,
"learning_rate": 0.0003524236983842011,
"loss": 2.0409858226776123,
"step": 2778
},
{
"epoch": 0.5865442941160957,
"grad_norm": 0.10134723782539368,
"learning_rate": 0.00035206463195691206,
"loss": 1.9624097347259521,
"step": 2780
},
{
"epoch": 0.5869662684284094,
"grad_norm": 0.10287471860647202,
"learning_rate": 0.000351705565529623,
"loss": 1.9929996728897095,
"step": 2782
},
{
"epoch": 0.5873882427407232,
"grad_norm": 0.10775137692689896,
"learning_rate": 0.00035134649910233395,
"loss": 2.008909225463867,
"step": 2784
},
{
"epoch": 0.5878102170530369,
"grad_norm": 0.10946634411811829,
"learning_rate": 0.0003509874326750449,
"loss": 2.0038068294525146,
"step": 2786
},
{
"epoch": 0.5882321913653507,
"grad_norm": 0.11175736784934998,
"learning_rate": 0.0003506283662477559,
"loss": 1.9708478450775146,
"step": 2788
},
{
"epoch": 0.5886541656776644,
"grad_norm": 0.10642441362142563,
"learning_rate": 0.0003502692998204668,
"loss": 1.967581033706665,
"step": 2790
},
{
"epoch": 0.5890761399899781,
"grad_norm": 0.10104036331176758,
"learning_rate": 0.00034991023339317777,
"loss": 1.9650471210479736,
"step": 2792
},
{
"epoch": 0.5894981143022918,
"grad_norm": 0.1050703376531601,
"learning_rate": 0.00034955116696588873,
"loss": 1.8978252410888672,
"step": 2794
},
{
"epoch": 0.5899200886146055,
"grad_norm": 0.10825706273317337,
"learning_rate": 0.00034919210053859965,
"loss": 1.9784396886825562,
"step": 2796
},
{
"epoch": 0.5903420629269194,
"grad_norm": 0.10270261764526367,
"learning_rate": 0.0003488330341113106,
"loss": 1.8945974111557007,
"step": 2798
},
{
"epoch": 0.5907640372392331,
"grad_norm": 0.10796843469142914,
"learning_rate": 0.00034847396768402153,
"loss": 1.8728725910186768,
"step": 2800
},
{
"epoch": 0.5911860115515468,
"grad_norm": 0.10691454261541367,
"learning_rate": 0.0003481149012567325,
"loss": 1.9897109270095825,
"step": 2802
},
{
"epoch": 0.5916079858638605,
"grad_norm": 0.11669714003801346,
"learning_rate": 0.00034775583482944347,
"loss": 1.9384316205978394,
"step": 2804
},
{
"epoch": 0.5920299601761743,
"grad_norm": 0.10912837088108063,
"learning_rate": 0.0003473967684021544,
"loss": 2.0111641883850098,
"step": 2806
},
{
"epoch": 0.592451934488488,
"grad_norm": 0.10557878017425537,
"learning_rate": 0.00034703770197486535,
"loss": 2.0158820152282715,
"step": 2808
},
{
"epoch": 0.5928739088008017,
"grad_norm": 0.10258884727954865,
"learning_rate": 0.0003466786355475763,
"loss": 1.9931654930114746,
"step": 2810
},
{
"epoch": 0.5932958831131155,
"grad_norm": 0.12349168956279755,
"learning_rate": 0.0003463195691202873,
"loss": 1.9360041618347168,
"step": 2812
},
{
"epoch": 0.5937178574254293,
"grad_norm": 0.12279972434043884,
"learning_rate": 0.0003459605026929982,
"loss": 1.9354610443115234,
"step": 2814
},
{
"epoch": 0.594139831737743,
"grad_norm": 0.10583202540874481,
"learning_rate": 0.0003456014362657092,
"loss": 1.9927340745925903,
"step": 2816
},
{
"epoch": 0.5945618060500567,
"grad_norm": 0.09568478912115097,
"learning_rate": 0.00034524236983842014,
"loss": 1.9943923950195312,
"step": 2818
},
{
"epoch": 0.5949837803623704,
"grad_norm": 0.10615766048431396,
"learning_rate": 0.0003448833034111311,
"loss": 1.9315276145935059,
"step": 2820
},
{
"epoch": 0.5954057546746842,
"grad_norm": 0.10695148259401321,
"learning_rate": 0.000344524236983842,
"loss": 1.9721901416778564,
"step": 2822
},
{
"epoch": 0.5958277289869979,
"grad_norm": 0.11504478007555008,
"learning_rate": 0.000344165170556553,
"loss": 2.0175111293792725,
"step": 2824
},
{
"epoch": 0.5962497032993117,
"grad_norm": 0.10889242589473724,
"learning_rate": 0.00034380610412926396,
"loss": 2.0752899646759033,
"step": 2826
},
{
"epoch": 0.5966716776116254,
"grad_norm": 0.1275372952222824,
"learning_rate": 0.0003434470377019749,
"loss": 1.9487941265106201,
"step": 2828
},
{
"epoch": 0.5970936519239392,
"grad_norm": 0.10920364409685135,
"learning_rate": 0.0003430879712746858,
"loss": 2.006545066833496,
"step": 2830
},
{
"epoch": 0.5975156262362529,
"grad_norm": 0.12148303538560867,
"learning_rate": 0.00034272890484739676,
"loss": 1.9555474519729614,
"step": 2832
},
{
"epoch": 0.5979376005485666,
"grad_norm": 0.11855153739452362,
"learning_rate": 0.00034236983842010773,
"loss": 1.9516345262527466,
"step": 2834
},
{
"epoch": 0.5983595748608803,
"grad_norm": 0.11642732471227646,
"learning_rate": 0.0003420107719928187,
"loss": 1.976448655128479,
"step": 2836
},
{
"epoch": 0.598781549173194,
"grad_norm": 0.12256433069705963,
"learning_rate": 0.0003416517055655296,
"loss": 1.9159197807312012,
"step": 2838
},
{
"epoch": 0.5992035234855079,
"grad_norm": 0.13907897472381592,
"learning_rate": 0.0003412926391382406,
"loss": 1.9876967668533325,
"step": 2840
},
{
"epoch": 0.5996254977978216,
"grad_norm": 0.11222678422927856,
"learning_rate": 0.00034093357271095155,
"loss": 2.0520358085632324,
"step": 2842
},
{
"epoch": 0.6000474721101353,
"grad_norm": 0.10357218980789185,
"learning_rate": 0.0003405745062836625,
"loss": 1.9905318021774292,
"step": 2844
},
{
"epoch": 0.6000474721101353,
"eval_loss": 1.9790700674057007,
"eval_runtime": 85.876,
"eval_samples_per_second": 35.679,
"eval_steps_per_second": 4.46,
"step": 2844
},
{
"epoch": 0.600469446422449,
"grad_norm": 0.10071981698274612,
"learning_rate": 0.00034021543985637343,
"loss": 2.0466420650482178,
"step": 2846
},
{
"epoch": 0.6008914207347628,
"grad_norm": 0.10100208967924118,
"learning_rate": 0.0003398563734290844,
"loss": 2.012871503829956,
"step": 2848
},
{
"epoch": 0.6013133950470765,
"grad_norm": 0.11836077272891998,
"learning_rate": 0.00033949730700179537,
"loss": 1.8980517387390137,
"step": 2850
},
{
"epoch": 0.6017353693593902,
"grad_norm": 0.11475014686584473,
"learning_rate": 0.00033913824057450634,
"loss": 1.9632195234298706,
"step": 2852
},
{
"epoch": 0.602157343671704,
"grad_norm": 0.10185011476278305,
"learning_rate": 0.00033877917414721725,
"loss": 1.9732929468154907,
"step": 2854
},
{
"epoch": 0.6025793179840178,
"grad_norm": 0.10145900398492813,
"learning_rate": 0.00033842010771992817,
"loss": 2.0089735984802246,
"step": 2856
},
{
"epoch": 0.6030012922963315,
"grad_norm": 0.10380379110574722,
"learning_rate": 0.00033806104129263914,
"loss": 1.9751780033111572,
"step": 2858
},
{
"epoch": 0.6034232666086452,
"grad_norm": 0.11150907725095749,
"learning_rate": 0.0003377019748653501,
"loss": 1.961464524269104,
"step": 2860
},
{
"epoch": 0.6038452409209589,
"grad_norm": 0.10287733376026154,
"learning_rate": 0.000337342908438061,
"loss": 1.9359874725341797,
"step": 2862
},
{
"epoch": 0.6042672152332726,
"grad_norm": 0.09571874886751175,
"learning_rate": 0.000336983842010772,
"loss": 1.9991509914398193,
"step": 2864
},
{
"epoch": 0.6046891895455864,
"grad_norm": 0.11020208895206451,
"learning_rate": 0.00033662477558348296,
"loss": 1.9465867280960083,
"step": 2866
},
{
"epoch": 0.6051111638579002,
"grad_norm": 0.1111195832490921,
"learning_rate": 0.0003362657091561939,
"loss": 1.9863793849945068,
"step": 2868
},
{
"epoch": 0.6055331381702139,
"grad_norm": 0.10025550425052643,
"learning_rate": 0.00033590664272890484,
"loss": 1.992165446281433,
"step": 2870
},
{
"epoch": 0.6059551124825276,
"grad_norm": 0.11705262959003448,
"learning_rate": 0.0003355475763016158,
"loss": 2.00213360786438,
"step": 2872
},
{
"epoch": 0.6063770867948414,
"grad_norm": 0.12702955305576324,
"learning_rate": 0.0003351885098743268,
"loss": 1.9955682754516602,
"step": 2874
},
{
"epoch": 0.6067990611071551,
"grad_norm": 0.10410808026790619,
"learning_rate": 0.00033482944344703775,
"loss": 1.987066626548767,
"step": 2876
},
{
"epoch": 0.6072210354194688,
"grad_norm": 0.1193886548280716,
"learning_rate": 0.00033447037701974866,
"loss": 1.9293038845062256,
"step": 2878
},
{
"epoch": 0.6076430097317825,
"grad_norm": 0.11874879151582718,
"learning_rate": 0.00033411131059245963,
"loss": 2.02607798576355,
"step": 2880
},
{
"epoch": 0.6080649840440964,
"grad_norm": 0.10892920196056366,
"learning_rate": 0.0003337522441651706,
"loss": 1.9464612007141113,
"step": 2882
},
{
"epoch": 0.6084869583564101,
"grad_norm": 0.11205556243658066,
"learning_rate": 0.00033339317773788157,
"loss": 2.024155855178833,
"step": 2884
},
{
"epoch": 0.6089089326687238,
"grad_norm": 0.11130602657794952,
"learning_rate": 0.0003330341113105925,
"loss": 2.01861834526062,
"step": 2886
},
{
"epoch": 0.6093309069810375,
"grad_norm": 0.11614452302455902,
"learning_rate": 0.0003326750448833034,
"loss": 1.9795364141464233,
"step": 2888
},
{
"epoch": 0.6097528812933513,
"grad_norm": 0.11269234120845795,
"learning_rate": 0.00033231597845601436,
"loss": 1.991916537284851,
"step": 2890
},
{
"epoch": 0.610174855605665,
"grad_norm": 0.12460962682962418,
"learning_rate": 0.00033195691202872533,
"loss": 1.9635121822357178,
"step": 2892
},
{
"epoch": 0.6105968299179787,
"grad_norm": 0.10501442104578018,
"learning_rate": 0.0003315978456014363,
"loss": 1.9993352890014648,
"step": 2894
},
{
"epoch": 0.6110188042302925,
"grad_norm": 0.12648485600948334,
"learning_rate": 0.0003312387791741472,
"loss": 1.9440288543701172,
"step": 2896
},
{
"epoch": 0.6114407785426063,
"grad_norm": 0.1058330088853836,
"learning_rate": 0.0003308797127468582,
"loss": 1.9708038568496704,
"step": 2898
},
{
"epoch": 0.61186275285492,
"grad_norm": 0.11296425759792328,
"learning_rate": 0.00033052064631956915,
"loss": 1.9509849548339844,
"step": 2900
},
{
"epoch": 0.6122847271672337,
"grad_norm": 0.10511868447065353,
"learning_rate": 0.00033016157989228007,
"loss": 1.9812482595443726,
"step": 2902
},
{
"epoch": 0.6127067014795474,
"grad_norm": 0.11557283997535706,
"learning_rate": 0.00032980251346499104,
"loss": 2.0157735347747803,
"step": 2904
},
{
"epoch": 0.6131286757918611,
"grad_norm": 0.10249584913253784,
"learning_rate": 0.000329443447037702,
"loss": 1.9411280155181885,
"step": 2906
},
{
"epoch": 0.6135506501041749,
"grad_norm": 0.11655893921852112,
"learning_rate": 0.000329084380610413,
"loss": 1.990997552871704,
"step": 2908
},
{
"epoch": 0.6139726244164887,
"grad_norm": 0.11614349484443665,
"learning_rate": 0.0003287253141831239,
"loss": 2.0188918113708496,
"step": 2910
},
{
"epoch": 0.6143945987288024,
"grad_norm": 0.09903915971517563,
"learning_rate": 0.00032836624775583486,
"loss": 1.9720869064331055,
"step": 2912
},
{
"epoch": 0.6148165730411161,
"grad_norm": 0.10472875088453293,
"learning_rate": 0.00032800718132854577,
"loss": 2.0398974418640137,
"step": 2914
},
{
"epoch": 0.6152385473534299,
"grad_norm": 0.10625752806663513,
"learning_rate": 0.00032764811490125674,
"loss": 1.9260084629058838,
"step": 2916
},
{
"epoch": 0.6156605216657436,
"grad_norm": 0.12201958894729614,
"learning_rate": 0.0003272890484739677,
"loss": 1.9209898710250854,
"step": 2918
},
{
"epoch": 0.6160824959780573,
"grad_norm": 0.11803163588047028,
"learning_rate": 0.0003269299820466786,
"loss": 1.9309139251708984,
"step": 2920
},
{
"epoch": 0.616504470290371,
"grad_norm": 0.12253094464540482,
"learning_rate": 0.0003265709156193896,
"loss": 2.005728006362915,
"step": 2922
},
{
"epoch": 0.6169264446026849,
"grad_norm": 0.10268803685903549,
"learning_rate": 0.00032621184919210056,
"loss": 1.9783984422683716,
"step": 2924
},
{
"epoch": 0.6173484189149986,
"grad_norm": 0.1152615025639534,
"learning_rate": 0.00032585278276481153,
"loss": 1.970442533493042,
"step": 2926
},
{
"epoch": 0.6177703932273123,
"grad_norm": 0.11511047184467316,
"learning_rate": 0.00032549371633752244,
"loss": 2.0116357803344727,
"step": 2928
},
{
"epoch": 0.618192367539626,
"grad_norm": 0.09982717037200928,
"learning_rate": 0.0003251346499102334,
"loss": 1.949166178703308,
"step": 2930
},
{
"epoch": 0.6186143418519398,
"grad_norm": 0.11270047724246979,
"learning_rate": 0.0003247755834829444,
"loss": 1.9565280675888062,
"step": 2932
},
{
"epoch": 0.6190363161642535,
"grad_norm": 0.10591725260019302,
"learning_rate": 0.0003244165170556553,
"loss": 1.9975123405456543,
"step": 2934
},
{
"epoch": 0.6194582904765672,
"grad_norm": 0.10390783101320267,
"learning_rate": 0.00032405745062836626,
"loss": 1.9380301237106323,
"step": 2936
},
{
"epoch": 0.619880264788881,
"grad_norm": 0.10181626677513123,
"learning_rate": 0.00032369838420107723,
"loss": 1.9853233098983765,
"step": 2938
},
{
"epoch": 0.6203022391011948,
"grad_norm": 0.1023334339261055,
"learning_rate": 0.0003233393177737882,
"loss": 1.9666825532913208,
"step": 2940
},
{
"epoch": 0.6207242134135085,
"grad_norm": 0.11403056234121323,
"learning_rate": 0.0003229802513464991,
"loss": 1.9822144508361816,
"step": 2942
},
{
"epoch": 0.6211461877258222,
"grad_norm": 0.10533489286899567,
"learning_rate": 0.00032262118491921003,
"loss": 2.022979974746704,
"step": 2944
},
{
"epoch": 0.6215681620381359,
"grad_norm": 0.10714052617549896,
"learning_rate": 0.000322262118491921,
"loss": 1.9644060134887695,
"step": 2946
},
{
"epoch": 0.6219901363504496,
"grad_norm": 0.09586666524410248,
"learning_rate": 0.00032190305206463197,
"loss": 1.973467469215393,
"step": 2948
},
{
"epoch": 0.6224121106627634,
"grad_norm": 0.10649166256189346,
"learning_rate": 0.00032154398563734294,
"loss": 1.959669828414917,
"step": 2950
},
{
"epoch": 0.6228340849750772,
"grad_norm": 0.11121958494186401,
"learning_rate": 0.00032118491921005385,
"loss": 1.8828858137130737,
"step": 2952
},
{
"epoch": 0.6232560592873909,
"grad_norm": 0.10249504446983337,
"learning_rate": 0.0003208258527827648,
"loss": 1.9781749248504639,
"step": 2954
},
{
"epoch": 0.6236780335997046,
"grad_norm": 0.1331326812505722,
"learning_rate": 0.0003204667863554758,
"loss": 2.0065133571624756,
"step": 2956
},
{
"epoch": 0.6241000079120184,
"grad_norm": 0.10772155970335007,
"learning_rate": 0.00032010771992818676,
"loss": 2.0177903175354004,
"step": 2958
},
{
"epoch": 0.6245219822243321,
"grad_norm": 0.10949335992336273,
"learning_rate": 0.00031974865350089767,
"loss": 1.968597650527954,
"step": 2960
},
{
"epoch": 0.6249439565366458,
"grad_norm": 0.11088529229164124,
"learning_rate": 0.00031938958707360864,
"loss": 2.063671350479126,
"step": 2962
},
{
"epoch": 0.6253659308489595,
"grad_norm": 0.11360663175582886,
"learning_rate": 0.0003190305206463196,
"loss": 1.9647870063781738,
"step": 2964
},
{
"epoch": 0.6257879051612734,
"grad_norm": 0.11286015063524246,
"learning_rate": 0.0003186714542190306,
"loss": 1.9654697179794312,
"step": 2966
},
{
"epoch": 0.6262098794735871,
"grad_norm": 0.10906553268432617,
"learning_rate": 0.0003183123877917415,
"loss": 2.0088539123535156,
"step": 2968
},
{
"epoch": 0.6266318537859008,
"grad_norm": 0.11842130869626999,
"learning_rate": 0.00031795332136445246,
"loss": 2.0344645977020264,
"step": 2970
},
{
"epoch": 0.6270538280982145,
"grad_norm": 0.10166054219007492,
"learning_rate": 0.0003175942549371634,
"loss": 2.016946315765381,
"step": 2972
},
{
"epoch": 0.6274758024105282,
"grad_norm": 0.1492660492658615,
"learning_rate": 0.00031723518850987434,
"loss": 2.0029025077819824,
"step": 2974
},
{
"epoch": 0.627897776722842,
"grad_norm": 0.10254668444395065,
"learning_rate": 0.00031687612208258526,
"loss": 1.9945354461669922,
"step": 2976
},
{
"epoch": 0.6283197510351557,
"grad_norm": 0.10376980900764465,
"learning_rate": 0.00031651705565529623,
"loss": 2.0055949687957764,
"step": 2978
},
{
"epoch": 0.6287417253474695,
"grad_norm": 0.10943156480789185,
"learning_rate": 0.0003161579892280072,
"loss": 1.945372223854065,
"step": 2980
},
{
"epoch": 0.6291636996597832,
"grad_norm": 0.1095069870352745,
"learning_rate": 0.00031579892280071816,
"loss": 1.9886653423309326,
"step": 2982
},
{
"epoch": 0.629585673972097,
"grad_norm": 0.11309268325567245,
"learning_rate": 0.0003154398563734291,
"loss": 1.9457988739013672,
"step": 2984
},
{
"epoch": 0.6300076482844107,
"grad_norm": 0.10724062472581863,
"learning_rate": 0.00031508078994614005,
"loss": 1.9849629402160645,
"step": 2986
},
{
"epoch": 0.6304296225967244,
"grad_norm": 0.11063455045223236,
"learning_rate": 0.000314721723518851,
"loss": 1.993377923965454,
"step": 2988
},
{
"epoch": 0.6308515969090381,
"grad_norm": 0.10344305634498596,
"learning_rate": 0.000314362657091562,
"loss": 1.9942773580551147,
"step": 2990
},
{
"epoch": 0.6312735712213519,
"grad_norm": 0.11132953315973282,
"learning_rate": 0.0003140035906642729,
"loss": 1.998533844947815,
"step": 2992
},
{
"epoch": 0.6316955455336657,
"grad_norm": 0.10856340825557709,
"learning_rate": 0.00031364452423698387,
"loss": 1.9747415781021118,
"step": 2994
},
{
"epoch": 0.6321175198459794,
"grad_norm": 0.10894513875246048,
"learning_rate": 0.00031328545780969484,
"loss": 2.042167901992798,
"step": 2996
},
{
"epoch": 0.6325394941582931,
"grad_norm": 0.10031988471746445,
"learning_rate": 0.0003129263913824058,
"loss": 1.9914487600326538,
"step": 2998
},
{
"epoch": 0.6329614684706069,
"grad_norm": 0.1088893935084343,
"learning_rate": 0.00031256732495511667,
"loss": 1.9473166465759277,
"step": 3000
},
{
"epoch": 0.6333834427829206,
"grad_norm": 0.10366523265838623,
"learning_rate": 0.00031220825852782763,
"loss": 1.9425256252288818,
"step": 3002
},
{
"epoch": 0.6338054170952343,
"grad_norm": 0.09686245024204254,
"learning_rate": 0.0003118491921005386,
"loss": 2.0071468353271484,
"step": 3004
},
{
"epoch": 0.634227391407548,
"grad_norm": 0.12878945469856262,
"learning_rate": 0.00031149012567324957,
"loss": 1.9694502353668213,
"step": 3006
},
{
"epoch": 0.6346493657198619,
"grad_norm": 0.10989037901163101,
"learning_rate": 0.0003111310592459605,
"loss": 1.9868390560150146,
"step": 3008
},
{
"epoch": 0.6350713400321756,
"grad_norm": 0.10989672690629959,
"learning_rate": 0.00031077199281867146,
"loss": 1.9732420444488525,
"step": 3010
},
{
"epoch": 0.6354933143444893,
"grad_norm": 0.1232403963804245,
"learning_rate": 0.0003104129263913824,
"loss": 2.003066301345825,
"step": 3012
},
{
"epoch": 0.635915288656803,
"grad_norm": 0.10542385280132294,
"learning_rate": 0.0003100538599640934,
"loss": 1.9185007810592651,
"step": 3014
},
{
"epoch": 0.6363372629691167,
"grad_norm": 0.10183856636285782,
"learning_rate": 0.0003096947935368043,
"loss": 1.9941368103027344,
"step": 3016
},
{
"epoch": 0.6367592372814305,
"grad_norm": 0.11390897631645203,
"learning_rate": 0.0003093357271095153,
"loss": 2.0033700466156006,
"step": 3018
},
{
"epoch": 0.6371812115937442,
"grad_norm": 0.10460637509822845,
"learning_rate": 0.00030897666068222624,
"loss": 1.9403115510940552,
"step": 3020
},
{
"epoch": 0.637603185906058,
"grad_norm": 0.10978318005800247,
"learning_rate": 0.0003086175942549372,
"loss": 1.9816153049468994,
"step": 3022
},
{
"epoch": 0.6380251602183717,
"grad_norm": 0.10029929131269455,
"learning_rate": 0.00030825852782764813,
"loss": 2.0122594833374023,
"step": 3024
},
{
"epoch": 0.6384471345306855,
"grad_norm": 0.10439856350421906,
"learning_rate": 0.0003078994614003591,
"loss": 1.9712491035461426,
"step": 3026
},
{
"epoch": 0.6388691088429992,
"grad_norm": 0.11405131965875626,
"learning_rate": 0.00030754039497307,
"loss": 1.9926965236663818,
"step": 3028
},
{
"epoch": 0.6392910831553129,
"grad_norm": 0.1223796159029007,
"learning_rate": 0.000307181328545781,
"loss": 1.9890563488006592,
"step": 3030
},
{
"epoch": 0.6397130574676266,
"grad_norm": 0.10618758946657181,
"learning_rate": 0.00030682226211849195,
"loss": 1.957329273223877,
"step": 3032
},
{
"epoch": 0.6401350317799404,
"grad_norm": 0.09942900389432907,
"learning_rate": 0.00030646319569120286,
"loss": 1.9920308589935303,
"step": 3034
},
{
"epoch": 0.6405570060922542,
"grad_norm": 0.11500733345746994,
"learning_rate": 0.00030610412926391383,
"loss": 2.0110859870910645,
"step": 3036
},
{
"epoch": 0.6409789804045679,
"grad_norm": 0.11476679146289825,
"learning_rate": 0.0003057450628366248,
"loss": 2.0004546642303467,
"step": 3038
},
{
"epoch": 0.6414009547168816,
"grad_norm": 0.1065887063741684,
"learning_rate": 0.0003053859964093357,
"loss": 1.9897539615631104,
"step": 3040
},
{
"epoch": 0.6418229290291954,
"grad_norm": 0.10964153707027435,
"learning_rate": 0.0003050269299820467,
"loss": 2.084834575653076,
"step": 3042
},
{
"epoch": 0.6422449033415091,
"grad_norm": 0.10541189461946487,
"learning_rate": 0.00030466786355475765,
"loss": 1.9767937660217285,
"step": 3044
},
{
"epoch": 0.6426668776538228,
"grad_norm": 0.10653765499591827,
"learning_rate": 0.0003043087971274686,
"loss": 1.9557280540466309,
"step": 3046
},
{
"epoch": 0.6430888519661365,
"grad_norm": 0.11697057634592056,
"learning_rate": 0.00030394973070017954,
"loss": 1.9890249967575073,
"step": 3048
},
{
"epoch": 0.6435108262784504,
"grad_norm": 0.10633612424135208,
"learning_rate": 0.0003035906642728905,
"loss": 1.9734883308410645,
"step": 3050
},
{
"epoch": 0.6439328005907641,
"grad_norm": 0.1019359678030014,
"learning_rate": 0.00030323159784560147,
"loss": 1.959587574005127,
"step": 3052
},
{
"epoch": 0.6443547749030778,
"grad_norm": 0.11090607196092606,
"learning_rate": 0.00030287253141831244,
"loss": 1.9707558155059814,
"step": 3054
},
{
"epoch": 0.6447767492153915,
"grad_norm": 0.10671336948871613,
"learning_rate": 0.00030251346499102336,
"loss": 2.031674385070801,
"step": 3056
},
{
"epoch": 0.6451987235277052,
"grad_norm": 0.11499840766191483,
"learning_rate": 0.00030215439856373427,
"loss": 1.9761170148849487,
"step": 3058
},
{
"epoch": 0.645620697840019,
"grad_norm": 0.09488421678543091,
"learning_rate": 0.00030179533213644524,
"loss": 1.9715839624404907,
"step": 3060
},
{
"epoch": 0.6460426721523327,
"grad_norm": 0.11742760241031647,
"learning_rate": 0.0003014362657091562,
"loss": 2.0545847415924072,
"step": 3062
},
{
"epoch": 0.6464646464646465,
"grad_norm": 0.10011886060237885,
"learning_rate": 0.0003010771992818672,
"loss": 1.973273754119873,
"step": 3064
},
{
"epoch": 0.6468866207769602,
"grad_norm": 0.1166694164276123,
"learning_rate": 0.0003007181328545781,
"loss": 2.045506238937378,
"step": 3066
},
{
"epoch": 0.647308595089274,
"grad_norm": 0.11640986055135727,
"learning_rate": 0.00030035906642728906,
"loss": 1.9989346265792847,
"step": 3068
},
{
"epoch": 0.6477305694015877,
"grad_norm": 0.11542612314224243,
"learning_rate": 0.00030000000000000003,
"loss": 1.9720537662506104,
"step": 3070
},
{
"epoch": 0.6481525437139014,
"grad_norm": 0.1030670627951622,
"learning_rate": 0.00029964093357271094,
"loss": 1.9584332704544067,
"step": 3072
},
{
"epoch": 0.6485745180262151,
"grad_norm": 0.10160288214683533,
"learning_rate": 0.0002992818671454219,
"loss": 1.9767011404037476,
"step": 3074
},
{
"epoch": 0.6489964923385289,
"grad_norm": 0.10808681696653366,
"learning_rate": 0.0002989228007181329,
"loss": 2.066359281539917,
"step": 3076
},
{
"epoch": 0.6494184666508427,
"grad_norm": 0.11602358520030975,
"learning_rate": 0.00029856373429084385,
"loss": 1.9115577936172485,
"step": 3078
},
{
"epoch": 0.6498404409631564,
"grad_norm": 0.12067712098360062,
"learning_rate": 0.00029820466786355476,
"loss": 1.9628145694732666,
"step": 3080
},
{
"epoch": 0.6502624152754701,
"grad_norm": 0.10651858896017075,
"learning_rate": 0.00029784560143626573,
"loss": 1.9922733306884766,
"step": 3082
},
{
"epoch": 0.6506843895877839,
"grad_norm": 0.10165168344974518,
"learning_rate": 0.0002974865350089767,
"loss": 1.9735554456710815,
"step": 3084
},
{
"epoch": 0.6511063639000976,
"grad_norm": 0.10648030787706375,
"learning_rate": 0.0002971274685816876,
"loss": 2.052121162414551,
"step": 3086
},
{
"epoch": 0.6515283382124113,
"grad_norm": 0.10655137896537781,
"learning_rate": 0.0002967684021543986,
"loss": 1.9688538312911987,
"step": 3088
},
{
"epoch": 0.651950312524725,
"grad_norm": 0.09834729135036469,
"learning_rate": 0.0002964093357271095,
"loss": 2.0202112197875977,
"step": 3090
},
{
"epoch": 0.6523722868370389,
"grad_norm": 0.11170980334281921,
"learning_rate": 0.00029605026929982047,
"loss": 2.0028228759765625,
"step": 3092
},
{
"epoch": 0.6527942611493526,
"grad_norm": 0.10852988064289093,
"learning_rate": 0.00029569120287253144,
"loss": 2.0273876190185547,
"step": 3094
},
{
"epoch": 0.6532162354616663,
"grad_norm": 0.09860529005527496,
"learning_rate": 0.0002953321364452424,
"loss": 1.9203295707702637,
"step": 3096
},
{
"epoch": 0.65363820977398,
"grad_norm": 0.1063685193657875,
"learning_rate": 0.0002949730700179533,
"loss": 1.9676445722579956,
"step": 3098
},
{
"epoch": 0.6540601840862937,
"grad_norm": 0.1150081679224968,
"learning_rate": 0.0002946140035906643,
"loss": 1.9306654930114746,
"step": 3100
},
{
"epoch": 0.6544821583986075,
"grad_norm": 0.11419858038425446,
"learning_rate": 0.00029425493716337526,
"loss": 2.0084142684936523,
"step": 3102
},
{
"epoch": 0.6549041327109212,
"grad_norm": 0.10884551703929901,
"learning_rate": 0.0002938958707360862,
"loss": 1.9549603462219238,
"step": 3104
},
{
"epoch": 0.655326107023235,
"grad_norm": 0.11228428781032562,
"learning_rate": 0.00029353680430879714,
"loss": 2.051999568939209,
"step": 3106
},
{
"epoch": 0.6557480813355487,
"grad_norm": 0.11050952970981598,
"learning_rate": 0.0002931777378815081,
"loss": 1.9858561754226685,
"step": 3108
},
{
"epoch": 0.6561700556478625,
"grad_norm": 0.09800341725349426,
"learning_rate": 0.0002928186714542191,
"loss": 1.9380171298980713,
"step": 3110
},
{
"epoch": 0.6565920299601762,
"grad_norm": 0.10719987004995346,
"learning_rate": 0.00029245960502693,
"loss": 1.9005001783370972,
"step": 3112
},
{
"epoch": 0.6570140042724899,
"grad_norm": 0.1078876405954361,
"learning_rate": 0.0002921005385996409,
"loss": 1.957391619682312,
"step": 3114
},
{
"epoch": 0.6574359785848036,
"grad_norm": 0.11697430163621902,
"learning_rate": 0.0002917414721723519,
"loss": 2.0177061557769775,
"step": 3116
},
{
"epoch": 0.6578579528971173,
"grad_norm": 0.1197165921330452,
"learning_rate": 0.00029138240574506284,
"loss": 2.0308010578155518,
"step": 3118
},
{
"epoch": 0.6582799272094311,
"grad_norm": 0.10472776740789413,
"learning_rate": 0.0002910233393177738,
"loss": 2.017258644104004,
"step": 3120
},
{
"epoch": 0.6587019015217449,
"grad_norm": 0.11442355811595917,
"learning_rate": 0.0002906642728904847,
"loss": 1.869146466255188,
"step": 3122
},
{
"epoch": 0.6591238758340586,
"grad_norm": 0.13063965737819672,
"learning_rate": 0.0002903052064631957,
"loss": 2.0101873874664307,
"step": 3124
},
{
"epoch": 0.6595458501463723,
"grad_norm": 0.10162826627492905,
"learning_rate": 0.00028994614003590666,
"loss": 2.0032289028167725,
"step": 3126
},
{
"epoch": 0.6599678244586861,
"grad_norm": 0.10672543197870255,
"learning_rate": 0.00028958707360861763,
"loss": 2.015963077545166,
"step": 3128
},
{
"epoch": 0.6603897987709998,
"grad_norm": 0.11390912532806396,
"learning_rate": 0.00028922800718132855,
"loss": 1.8277218341827393,
"step": 3130
},
{
"epoch": 0.6608117730833135,
"grad_norm": 0.10441584885120392,
"learning_rate": 0.0002888689407540395,
"loss": 1.9052979946136475,
"step": 3132
},
{
"epoch": 0.6612337473956272,
"grad_norm": 0.10032348334789276,
"learning_rate": 0.0002885098743267505,
"loss": 2.0039725303649902,
"step": 3134
},
{
"epoch": 0.6616557217079411,
"grad_norm": 0.09926117956638336,
"learning_rate": 0.00028815080789946145,
"loss": 1.9438047409057617,
"step": 3136
},
{
"epoch": 0.6620776960202548,
"grad_norm": 0.10437823832035065,
"learning_rate": 0.00028779174147217237,
"loss": 1.9770876169204712,
"step": 3138
},
{
"epoch": 0.6624996703325685,
"grad_norm": 0.12294481694698334,
"learning_rate": 0.00028743267504488334,
"loss": 1.9200589656829834,
"step": 3140
},
{
"epoch": 0.6629216446448822,
"grad_norm": 0.101417675614357,
"learning_rate": 0.00028707360861759425,
"loss": 1.9717966318130493,
"step": 3142
},
{
"epoch": 0.663343618957196,
"grad_norm": 0.11127351969480515,
"learning_rate": 0.0002867145421903052,
"loss": 2.0202882289886475,
"step": 3144
},
{
"epoch": 0.6637655932695097,
"grad_norm": 0.1103253960609436,
"learning_rate": 0.00028635547576301613,
"loss": 1.9596904516220093,
"step": 3146
},
{
"epoch": 0.6641875675818234,
"grad_norm": 0.10551868379116058,
"learning_rate": 0.0002859964093357271,
"loss": 2.001769542694092,
"step": 3148
},
{
"epoch": 0.6646095418941372,
"grad_norm": 0.10602214932441711,
"learning_rate": 0.00028563734290843807,
"loss": 1.9972647428512573,
"step": 3150
},
{
"epoch": 0.665031516206451,
"grad_norm": 0.1001637801527977,
"learning_rate": 0.00028527827648114904,
"loss": 1.9813357591629028,
"step": 3152
},
{
"epoch": 0.6654534905187647,
"grad_norm": 0.10740935802459717,
"learning_rate": 0.00028491921005385995,
"loss": 1.9325469732284546,
"step": 3154
},
{
"epoch": 0.6658754648310784,
"grad_norm": 0.10302960127592087,
"learning_rate": 0.0002845601436265709,
"loss": 1.9399449825286865,
"step": 3156
},
{
"epoch": 0.6662974391433921,
"grad_norm": 0.09835318475961685,
"learning_rate": 0.0002842010771992819,
"loss": 1.9897425174713135,
"step": 3158
},
{
"epoch": 0.6667194134557058,
"grad_norm": 0.1110355332493782,
"learning_rate": 0.00028384201077199286,
"loss": 1.9343997240066528,
"step": 3160
},
{
"epoch": 0.6671413877680196,
"grad_norm": 0.12437155097723007,
"learning_rate": 0.0002834829443447038,
"loss": 2.0056262016296387,
"step": 3162
},
{
"epoch": 0.6675633620803334,
"grad_norm": 0.11256714165210724,
"learning_rate": 0.00028312387791741474,
"loss": 1.9349945783615112,
"step": 3164
},
{
"epoch": 0.6679853363926471,
"grad_norm": 0.11093205213546753,
"learning_rate": 0.0002827648114901257,
"loss": 1.992722511291504,
"step": 3166
},
{
"epoch": 0.6684073107049608,
"grad_norm": 0.1201973408460617,
"learning_rate": 0.0002824057450628367,
"loss": 2.030259847640991,
"step": 3168
},
{
"epoch": 0.6688292850172746,
"grad_norm": 0.11056402325630188,
"learning_rate": 0.00028204667863554754,
"loss": 2.0019443035125732,
"step": 3170
},
{
"epoch": 0.6692512593295883,
"grad_norm": 0.11181090772151947,
"learning_rate": 0.0002816876122082585,
"loss": 1.9864492416381836,
"step": 3172
},
{
"epoch": 0.669673233641902,
"grad_norm": 0.11066945642232895,
"learning_rate": 0.0002813285457809695,
"loss": 2.005100965499878,
"step": 3174
},
{
"epoch": 0.6700952079542157,
"grad_norm": 0.09222931414842606,
"learning_rate": 0.00028096947935368045,
"loss": 1.984084129333496,
"step": 3176
},
{
"epoch": 0.6705171822665296,
"grad_norm": 0.11407238990068436,
"learning_rate": 0.00028061041292639136,
"loss": 1.9762685298919678,
"step": 3178
},
{
"epoch": 0.6709391565788433,
"grad_norm": 0.11664014309644699,
"learning_rate": 0.00028025134649910233,
"loss": 1.9102507829666138,
"step": 3180
},
{
"epoch": 0.671361130891157,
"grad_norm": 0.10872922837734222,
"learning_rate": 0.0002798922800718133,
"loss": 1.9498902559280396,
"step": 3182
},
{
"epoch": 0.6717831052034707,
"grad_norm": 0.10012578964233398,
"learning_rate": 0.00027953321364452427,
"loss": 2.013627052307129,
"step": 3184
},
{
"epoch": 0.6722050795157845,
"grad_norm": 0.11567337065935135,
"learning_rate": 0.0002791741472172352,
"loss": 2.023460865020752,
"step": 3186
},
{
"epoch": 0.6726270538280982,
"grad_norm": 0.1118452325463295,
"learning_rate": 0.00027881508078994615,
"loss": 1.9477510452270508,
"step": 3188
},
{
"epoch": 0.6730490281404119,
"grad_norm": 0.09833332896232605,
"learning_rate": 0.0002784560143626571,
"loss": 1.9331953525543213,
"step": 3190
},
{
"epoch": 0.6734710024527257,
"grad_norm": 0.10791114717721939,
"learning_rate": 0.0002780969479353681,
"loss": 1.9648985862731934,
"step": 3192
},
{
"epoch": 0.6738929767650395,
"grad_norm": 0.10593262314796448,
"learning_rate": 0.000277737881508079,
"loss": 2.0172150135040283,
"step": 3194
},
{
"epoch": 0.6743149510773532,
"grad_norm": 0.09685990959405899,
"learning_rate": 0.00027737881508078997,
"loss": 1.9128156900405884,
"step": 3196
},
{
"epoch": 0.6747369253896669,
"grad_norm": 0.12871618568897247,
"learning_rate": 0.00027701974865350094,
"loss": 1.99347984790802,
"step": 3198
},
{
"epoch": 0.6751588997019806,
"grad_norm": 0.1120900884270668,
"learning_rate": 0.00027666068222621185,
"loss": 1.997772455215454,
"step": 3200
},
{
"epoch": 0.6755808740142943,
"grad_norm": 0.09697538614273071,
"learning_rate": 0.0002763016157989228,
"loss": 1.9803364276885986,
"step": 3202
},
{
"epoch": 0.6760028483266081,
"grad_norm": 0.10590265691280365,
"learning_rate": 0.00027594254937163374,
"loss": 2.0186538696289062,
"step": 3204
},
{
"epoch": 0.6764248226389219,
"grad_norm": 0.09822264313697815,
"learning_rate": 0.0002755834829443447,
"loss": 1.9837347269058228,
"step": 3206
},
{
"epoch": 0.6768467969512356,
"grad_norm": 0.10471075028181076,
"learning_rate": 0.0002752244165170557,
"loss": 1.991766095161438,
"step": 3208
},
{
"epoch": 0.6772687712635493,
"grad_norm": 0.10055899620056152,
"learning_rate": 0.0002748653500897666,
"loss": 1.9920276403427124,
"step": 3210
},
{
"epoch": 0.6776907455758631,
"grad_norm": 0.10824410617351532,
"learning_rate": 0.00027450628366247756,
"loss": 1.9547581672668457,
"step": 3212
},
{
"epoch": 0.6781127198881768,
"grad_norm": 0.10399824380874634,
"learning_rate": 0.0002741472172351885,
"loss": 2.0019216537475586,
"step": 3214
},
{
"epoch": 0.6785346942004905,
"grad_norm": 0.12088403850793839,
"learning_rate": 0.0002737881508078995,
"loss": 1.9607950448989868,
"step": 3216
},
{
"epoch": 0.6789566685128042,
"grad_norm": 0.11124017089605331,
"learning_rate": 0.0002734290843806104,
"loss": 1.9928278923034668,
"step": 3218
},
{
"epoch": 0.6793786428251181,
"grad_norm": 0.09391073137521744,
"learning_rate": 0.0002730700179533214,
"loss": 1.9680840969085693,
"step": 3220
},
{
"epoch": 0.6798006171374318,
"grad_norm": 0.13161370158195496,
"learning_rate": 0.00027271095152603235,
"loss": 1.9610979557037354,
"step": 3222
},
{
"epoch": 0.6802225914497455,
"grad_norm": 0.11042996495962143,
"learning_rate": 0.0002723518850987433,
"loss": 2.0584731101989746,
"step": 3224
},
{
"epoch": 0.6806445657620592,
"grad_norm": 0.1079232320189476,
"learning_rate": 0.00027199281867145423,
"loss": 1.9513826370239258,
"step": 3226
},
{
"epoch": 0.681066540074373,
"grad_norm": 0.12318006902933121,
"learning_rate": 0.00027163375224416514,
"loss": 2.0483458042144775,
"step": 3228
},
{
"epoch": 0.6814885143866867,
"grad_norm": 0.11562740057706833,
"learning_rate": 0.0002712746858168761,
"loss": 2.0148608684539795,
"step": 3230
},
{
"epoch": 0.6819104886990004,
"grad_norm": 0.10717836767435074,
"learning_rate": 0.0002709156193895871,
"loss": 1.9613111019134521,
"step": 3232
},
{
"epoch": 0.6823324630113142,
"grad_norm": 0.1033383384346962,
"learning_rate": 0.00027055655296229805,
"loss": 1.9776865243911743,
"step": 3234
},
{
"epoch": 0.682754437323628,
"grad_norm": 0.11299113184213638,
"learning_rate": 0.00027019748653500897,
"loss": 2.0406699180603027,
"step": 3236
},
{
"epoch": 0.6831764116359417,
"grad_norm": 0.10743097215890884,
"learning_rate": 0.00026983842010771993,
"loss": 1.9806272983551025,
"step": 3238
},
{
"epoch": 0.6835983859482554,
"grad_norm": 0.11589135229587555,
"learning_rate": 0.0002694793536804309,
"loss": 1.9889198541641235,
"step": 3240
},
{
"epoch": 0.6840203602605691,
"grad_norm": 0.10776390135288239,
"learning_rate": 0.00026912028725314187,
"loss": 1.9237390756607056,
"step": 3242
},
{
"epoch": 0.6844423345728828,
"grad_norm": 0.10813435912132263,
"learning_rate": 0.0002687612208258528,
"loss": 2.0432019233703613,
"step": 3244
},
{
"epoch": 0.6848643088851966,
"grad_norm": 0.12461636960506439,
"learning_rate": 0.00026840215439856375,
"loss": 2.036771297454834,
"step": 3246
},
{
"epoch": 0.6852862831975104,
"grad_norm": 0.1076311469078064,
"learning_rate": 0.0002680430879712747,
"loss": 1.969459056854248,
"step": 3248
},
{
"epoch": 0.6857082575098241,
"grad_norm": 0.12014276534318924,
"learning_rate": 0.00026768402154398564,
"loss": 1.9760980606079102,
"step": 3250
},
{
"epoch": 0.6861302318221378,
"grad_norm": 0.12151424586772919,
"learning_rate": 0.0002673249551166966,
"loss": 1.953811764717102,
"step": 3252
},
{
"epoch": 0.6865522061344516,
"grad_norm": 0.12037377804517746,
"learning_rate": 0.0002669658886894076,
"loss": 1.9828217029571533,
"step": 3254
},
{
"epoch": 0.6869741804467653,
"grad_norm": 0.10588803887367249,
"learning_rate": 0.0002666068222621185,
"loss": 1.946938157081604,
"step": 3256
},
{
"epoch": 0.687396154759079,
"grad_norm": 0.09934581071138382,
"learning_rate": 0.00026624775583482946,
"loss": 2.014741897583008,
"step": 3258
},
{
"epoch": 0.6878181290713927,
"grad_norm": 0.11692436784505844,
"learning_rate": 0.00026588868940754037,
"loss": 1.9555954933166504,
"step": 3260
},
{
"epoch": 0.6882401033837066,
"grad_norm": 0.10826175659894943,
"learning_rate": 0.00026552962298025134,
"loss": 1.958064079284668,
"step": 3262
},
{
"epoch": 0.6886620776960203,
"grad_norm": 0.14348508417606354,
"learning_rate": 0.0002651705565529623,
"loss": 1.9756920337677002,
"step": 3264
},
{
"epoch": 0.689084052008334,
"grad_norm": 0.11485713720321655,
"learning_rate": 0.0002648114901256733,
"loss": 2.0419607162475586,
"step": 3266
},
{
"epoch": 0.6895060263206477,
"grad_norm": 0.11305432766675949,
"learning_rate": 0.0002644524236983842,
"loss": 1.9264963865280151,
"step": 3268
},
{
"epoch": 0.6899280006329614,
"grad_norm": 0.10322091728448868,
"learning_rate": 0.00026409335727109516,
"loss": 1.9949519634246826,
"step": 3270
},
{
"epoch": 0.6903499749452752,
"grad_norm": 0.10511468350887299,
"learning_rate": 0.00026373429084380613,
"loss": 2.029182195663452,
"step": 3272
},
{
"epoch": 0.6907719492575889,
"grad_norm": 0.10895533859729767,
"learning_rate": 0.0002633752244165171,
"loss": 1.9597563743591309,
"step": 3274
},
{
"epoch": 0.6911939235699027,
"grad_norm": 0.11771010607481003,
"learning_rate": 0.000263016157989228,
"loss": 1.990113377571106,
"step": 3276
},
{
"epoch": 0.6916158978822164,
"grad_norm": 0.10242278128862381,
"learning_rate": 0.000262657091561939,
"loss": 2.0089271068573,
"step": 3278
},
{
"epoch": 0.6920378721945302,
"grad_norm": 0.11202660202980042,
"learning_rate": 0.00026229802513464995,
"loss": 1.978929042816162,
"step": 3280
},
{
"epoch": 0.6924598465068439,
"grad_norm": 0.09460461139678955,
"learning_rate": 0.0002619389587073609,
"loss": 1.9332740306854248,
"step": 3282
},
{
"epoch": 0.6928818208191576,
"grad_norm": 0.09914391487836838,
"learning_rate": 0.00026157989228007183,
"loss": 1.9953382015228271,
"step": 3284
},
{
"epoch": 0.6933037951314713,
"grad_norm": 0.129542276263237,
"learning_rate": 0.00026122082585278275,
"loss": 2.0398082733154297,
"step": 3286
},
{
"epoch": 0.6937257694437851,
"grad_norm": 0.09624647349119186,
"learning_rate": 0.0002608617594254937,
"loss": 1.980837345123291,
"step": 3288
},
{
"epoch": 0.6941477437560989,
"grad_norm": 0.1069541648030281,
"learning_rate": 0.0002605026929982047,
"loss": 1.9964889287948608,
"step": 3290
},
{
"epoch": 0.6945697180684126,
"grad_norm": 0.09792979806661606,
"learning_rate": 0.0002601436265709156,
"loss": 1.92133629322052,
"step": 3292
},
{
"epoch": 0.6949916923807263,
"grad_norm": 0.10546477138996124,
"learning_rate": 0.00025978456014362657,
"loss": 1.9602383375167847,
"step": 3294
},
{
"epoch": 0.6954136666930401,
"grad_norm": 0.12030339986085892,
"learning_rate": 0.00025942549371633754,
"loss": 1.964606761932373,
"step": 3296
},
{
"epoch": 0.6958356410053538,
"grad_norm": 0.1270943135023117,
"learning_rate": 0.0002590664272890485,
"loss": 1.9512083530426025,
"step": 3298
},
{
"epoch": 0.6962576153176675,
"grad_norm": 0.14737571775913239,
"learning_rate": 0.0002587073608617594,
"loss": 2.012361764907837,
"step": 3300
},
{
"epoch": 0.6966795896299812,
"grad_norm": 0.11572889983654022,
"learning_rate": 0.0002583482944344704,
"loss": 2.010521173477173,
"step": 3302
},
{
"epoch": 0.697101563942295,
"grad_norm": 0.10504786670207977,
"learning_rate": 0.00025798922800718136,
"loss": 1.9199132919311523,
"step": 3304
},
{
"epoch": 0.6975235382546088,
"grad_norm": 0.12738004326820374,
"learning_rate": 0.00025763016157989233,
"loss": 2.0301239490509033,
"step": 3306
},
{
"epoch": 0.6979455125669225,
"grad_norm": 0.11718247085809708,
"learning_rate": 0.00025727109515260324,
"loss": 1.9508535861968994,
"step": 3308
},
{
"epoch": 0.6983674868792362,
"grad_norm": 0.10549486428499222,
"learning_rate": 0.0002569120287253142,
"loss": 1.9171465635299683,
"step": 3310
},
{
"epoch": 0.6987894611915499,
"grad_norm": 0.11088002473115921,
"learning_rate": 0.0002565529622980252,
"loss": 2.0184812545776367,
"step": 3312
},
{
"epoch": 0.6992114355038637,
"grad_norm": 0.10309018194675446,
"learning_rate": 0.0002561938958707361,
"loss": 1.9870266914367676,
"step": 3314
},
{
"epoch": 0.6996334098161774,
"grad_norm": 0.10589032620191574,
"learning_rate": 0.000255834829443447,
"loss": 1.9722516536712646,
"step": 3316
},
{
"epoch": 0.7000553841284912,
"grad_norm": 0.10752245783805847,
"learning_rate": 0.000255475763016158,
"loss": 1.9834229946136475,
"step": 3318
},
{
"epoch": 0.7004773584408049,
"grad_norm": 0.10278623551130295,
"learning_rate": 0.00025511669658886895,
"loss": 2.000675678253174,
"step": 3320
},
{
"epoch": 0.7008993327531187,
"grad_norm": 0.11228831857442856,
"learning_rate": 0.0002547576301615799,
"loss": 1.960915207862854,
"step": 3322
},
{
"epoch": 0.7013213070654324,
"grad_norm": 0.09654724597930908,
"learning_rate": 0.00025439856373429083,
"loss": 1.979228138923645,
"step": 3324
},
{
"epoch": 0.7017432813777461,
"grad_norm": 0.11637706309556961,
"learning_rate": 0.0002540394973070018,
"loss": 1.8917217254638672,
"step": 3326
},
{
"epoch": 0.7021652556900598,
"grad_norm": 0.11561064422130585,
"learning_rate": 0.00025368043087971277,
"loss": 1.9652663469314575,
"step": 3328
},
{
"epoch": 0.7025872300023736,
"grad_norm": 0.10369198769330978,
"learning_rate": 0.00025332136445242373,
"loss": 1.9404369592666626,
"step": 3330
},
{
"epoch": 0.7030092043146874,
"grad_norm": 0.12422200292348862,
"learning_rate": 0.00025296229802513465,
"loss": 1.931297779083252,
"step": 3332
},
{
"epoch": 0.7034311786270011,
"grad_norm": 0.11538632214069366,
"learning_rate": 0.0002526032315978456,
"loss": 2.0307869911193848,
"step": 3334
},
{
"epoch": 0.7038531529393148,
"grad_norm": 0.10798940807580948,
"learning_rate": 0.0002522441651705566,
"loss": 1.992327094078064,
"step": 3336
},
{
"epoch": 0.7042751272516286,
"grad_norm": 0.10671303421258926,
"learning_rate": 0.00025188509874326756,
"loss": 1.9387906789779663,
"step": 3338
},
{
"epoch": 0.7046971015639423,
"grad_norm": 0.1062905415892601,
"learning_rate": 0.00025152603231597847,
"loss": 2.0574240684509277,
"step": 3340
},
{
"epoch": 0.705119075876256,
"grad_norm": 0.10800740122795105,
"learning_rate": 0.0002511669658886894,
"loss": 1.9664337635040283,
"step": 3342
},
{
"epoch": 0.7055410501885697,
"grad_norm": 0.10134493559598923,
"learning_rate": 0.00025080789946140035,
"loss": 1.9446991682052612,
"step": 3344
},
{
"epoch": 0.7059630245008836,
"grad_norm": 0.20048022270202637,
"learning_rate": 0.0002504488330341113,
"loss": 1.95692777633667,
"step": 3346
},
{
"epoch": 0.7063849988131973,
"grad_norm": 0.11488698422908783,
"learning_rate": 0.00025008976660682224,
"loss": 1.9315844774246216,
"step": 3348
},
{
"epoch": 0.706806973125511,
"grad_norm": 0.10052983462810516,
"learning_rate": 0.0002497307001795332,
"loss": 1.9756077527999878,
"step": 3350
},
{
"epoch": 0.7072289474378247,
"grad_norm": 0.10937532037496567,
"learning_rate": 0.0002493716337522442,
"loss": 1.9646583795547485,
"step": 3352
},
{
"epoch": 0.7076509217501384,
"grad_norm": 0.10485361516475677,
"learning_rate": 0.00024901256732495514,
"loss": 1.969605565071106,
"step": 3354
},
{
"epoch": 0.7080728960624522,
"grad_norm": 0.1051955372095108,
"learning_rate": 0.00024865350089766606,
"loss": 1.8856544494628906,
"step": 3356
},
{
"epoch": 0.7084948703747659,
"grad_norm": 0.10765553265810013,
"learning_rate": 0.000248294434470377,
"loss": 1.958080530166626,
"step": 3358
},
{
"epoch": 0.7089168446870797,
"grad_norm": 0.11328635364770889,
"learning_rate": 0.000247935368043088,
"loss": 2.016552448272705,
"step": 3360
},
{
"epoch": 0.7093388189993934,
"grad_norm": 0.10600387305021286,
"learning_rate": 0.00024757630161579896,
"loss": 2.0450572967529297,
"step": 3362
},
{
"epoch": 0.7097607933117072,
"grad_norm": 0.10306272655725479,
"learning_rate": 0.0002472172351885099,
"loss": 2.014941692352295,
"step": 3364
},
{
"epoch": 0.7101827676240209,
"grad_norm": 0.10871893912553787,
"learning_rate": 0.00024685816876122085,
"loss": 2.0122649669647217,
"step": 3366
},
{
"epoch": 0.7106047419363346,
"grad_norm": 0.10582491010427475,
"learning_rate": 0.0002464991023339318,
"loss": 1.9494285583496094,
"step": 3368
},
{
"epoch": 0.7110267162486483,
"grad_norm": 0.120091013610363,
"learning_rate": 0.00024614003590664273,
"loss": 1.8860846757888794,
"step": 3370
},
{
"epoch": 0.711448690560962,
"grad_norm": 0.11734078824520111,
"learning_rate": 0.0002457809694793537,
"loss": 1.9980287551879883,
"step": 3372
},
{
"epoch": 0.7118706648732759,
"grad_norm": 0.10194410383701324,
"learning_rate": 0.0002454219030520646,
"loss": 1.9670648574829102,
"step": 3374
},
{
"epoch": 0.7122926391855896,
"grad_norm": 0.10699601471424103,
"learning_rate": 0.0002450628366247756,
"loss": 1.9585990905761719,
"step": 3376
},
{
"epoch": 0.7127146134979033,
"grad_norm": 0.10585588216781616,
"learning_rate": 0.00024470377019748655,
"loss": 1.9906713962554932,
"step": 3378
},
{
"epoch": 0.713136587810217,
"grad_norm": 0.11541510373353958,
"learning_rate": 0.0002443447037701975,
"loss": 1.952163815498352,
"step": 3380
},
{
"epoch": 0.7135585621225308,
"grad_norm": 0.09901058673858643,
"learning_rate": 0.00024398563734290846,
"loss": 1.9524643421173096,
"step": 3382
},
{
"epoch": 0.7139805364348445,
"grad_norm": 0.1152893453836441,
"learning_rate": 0.0002436265709156194,
"loss": 1.9829622507095337,
"step": 3384
},
{
"epoch": 0.7144025107471582,
"grad_norm": 0.11904725432395935,
"learning_rate": 0.00024326750448833034,
"loss": 2.0015158653259277,
"step": 3386
},
{
"epoch": 0.714824485059472,
"grad_norm": 0.10806585848331451,
"learning_rate": 0.0002429084380610413,
"loss": 1.9670894145965576,
"step": 3388
},
{
"epoch": 0.7152464593717858,
"grad_norm": 0.11711076647043228,
"learning_rate": 0.00024254937163375225,
"loss": 1.9742376804351807,
"step": 3390
},
{
"epoch": 0.7156684336840995,
"grad_norm": 0.10387764126062393,
"learning_rate": 0.00024219030520646322,
"loss": 1.9554024934768677,
"step": 3392
},
{
"epoch": 0.7160904079964132,
"grad_norm": 0.12585341930389404,
"learning_rate": 0.00024183123877917416,
"loss": 2.029963970184326,
"step": 3394
},
{
"epoch": 0.7165123823087269,
"grad_norm": 0.10207269340753555,
"learning_rate": 0.00024147217235188513,
"loss": 1.9371564388275146,
"step": 3396
},
{
"epoch": 0.7169343566210407,
"grad_norm": 0.10948192328214645,
"learning_rate": 0.00024111310592459607,
"loss": 1.960700511932373,
"step": 3398
},
{
"epoch": 0.7173563309333544,
"grad_norm": 0.10689511150121689,
"learning_rate": 0.000240754039497307,
"loss": 1.975806474685669,
"step": 3400
},
{
"epoch": 0.7177783052456682,
"grad_norm": 0.10742465406656265,
"learning_rate": 0.00024039497307001796,
"loss": 1.9533134698867798,
"step": 3402
},
{
"epoch": 0.7182002795579819,
"grad_norm": 0.10114706307649612,
"learning_rate": 0.0002400359066427289,
"loss": 2.021449327468872,
"step": 3404
},
{
"epoch": 0.7186222538702957,
"grad_norm": 0.10994695127010345,
"learning_rate": 0.00023967684021543987,
"loss": 1.9949291944503784,
"step": 3406
},
{
"epoch": 0.7190442281826094,
"grad_norm": 0.10153011232614517,
"learning_rate": 0.0002393177737881508,
"loss": 1.9540281295776367,
"step": 3408
},
{
"epoch": 0.7194662024949231,
"grad_norm": 0.10419446974992752,
"learning_rate": 0.00023895870736086178,
"loss": 1.9885565042495728,
"step": 3410
},
{
"epoch": 0.7198881768072368,
"grad_norm": 0.11679089814424515,
"learning_rate": 0.00023859964093357272,
"loss": 1.9555705785751343,
"step": 3412
},
{
"epoch": 0.7203101511195505,
"grad_norm": 0.10403481870889664,
"learning_rate": 0.0002382405745062837,
"loss": 1.9669502973556519,
"step": 3414
},
{
"epoch": 0.7207321254318644,
"grad_norm": 0.10146000981330872,
"learning_rate": 0.00023788150807899463,
"loss": 1.957541584968567,
"step": 3416
},
{
"epoch": 0.7211540997441781,
"grad_norm": 0.10877972096204758,
"learning_rate": 0.0002375224416517056,
"loss": 1.9599831104278564,
"step": 3418
},
{
"epoch": 0.7215760740564918,
"grad_norm": 0.1018085777759552,
"learning_rate": 0.00023716337522441654,
"loss": 1.951562523841858,
"step": 3420
},
{
"epoch": 0.7219980483688055,
"grad_norm": 0.10491974651813507,
"learning_rate": 0.00023680430879712748,
"loss": 1.900702953338623,
"step": 3422
},
{
"epoch": 0.7224200226811193,
"grad_norm": 0.11076642572879791,
"learning_rate": 0.00023644524236983845,
"loss": 1.9911231994628906,
"step": 3424
},
{
"epoch": 0.722841996993433,
"grad_norm": 0.1153176799416542,
"learning_rate": 0.0002360861759425494,
"loss": 2.008988380432129,
"step": 3426
},
{
"epoch": 0.7232639713057467,
"grad_norm": 0.11141860485076904,
"learning_rate": 0.0002357271095152603,
"loss": 2.0262913703918457,
"step": 3428
},
{
"epoch": 0.7236859456180605,
"grad_norm": 0.10677400231361389,
"learning_rate": 0.00023536804308797127,
"loss": 2.0297980308532715,
"step": 3430
},
{
"epoch": 0.7241079199303743,
"grad_norm": 0.11610022187232971,
"learning_rate": 0.00023500897666068222,
"loss": 2.015465021133423,
"step": 3432
},
{
"epoch": 0.724529894242688,
"grad_norm": 0.10522017627954483,
"learning_rate": 0.00023464991023339318,
"loss": 1.961220383644104,
"step": 3434
},
{
"epoch": 0.7249518685550017,
"grad_norm": 0.12547391653060913,
"learning_rate": 0.00023429084380610413,
"loss": 1.912192463874817,
"step": 3436
},
{
"epoch": 0.7253738428673154,
"grad_norm": 0.1278771013021469,
"learning_rate": 0.0002339317773788151,
"loss": 1.9481815099716187,
"step": 3438
},
{
"epoch": 0.7257958171796292,
"grad_norm": 0.1072755977511406,
"learning_rate": 0.00023357271095152604,
"loss": 1.9688409566879272,
"step": 3440
},
{
"epoch": 0.7262177914919429,
"grad_norm": 0.10849331319332123,
"learning_rate": 0.000233213644524237,
"loss": 2.003361701965332,
"step": 3442
},
{
"epoch": 0.7266397658042567,
"grad_norm": 0.10375521332025528,
"learning_rate": 0.00023285457809694795,
"loss": 1.9480583667755127,
"step": 3444
},
{
"epoch": 0.7270617401165704,
"grad_norm": 0.10477391630411148,
"learning_rate": 0.00023249551166965892,
"loss": 1.9900933504104614,
"step": 3446
},
{
"epoch": 0.7274837144288842,
"grad_norm": 0.10969476401805878,
"learning_rate": 0.00023213644524236986,
"loss": 2.015200138092041,
"step": 3448
},
{
"epoch": 0.7279056887411979,
"grad_norm": 0.10198700428009033,
"learning_rate": 0.00023177737881508083,
"loss": 1.95029878616333,
"step": 3450
},
{
"epoch": 0.7283276630535116,
"grad_norm": 0.11973521113395691,
"learning_rate": 0.00023141831238779177,
"loss": 1.933430552482605,
"step": 3452
},
{
"epoch": 0.7287496373658253,
"grad_norm": 0.10923260450363159,
"learning_rate": 0.00023105924596050274,
"loss": 2.0197134017944336,
"step": 3454
},
{
"epoch": 0.729171611678139,
"grad_norm": 0.1203337237238884,
"learning_rate": 0.00023070017953321362,
"loss": 1.8956019878387451,
"step": 3456
},
{
"epoch": 0.7295935859904529,
"grad_norm": 0.10758564621210098,
"learning_rate": 0.0002303411131059246,
"loss": 2.0047976970672607,
"step": 3458
},
{
"epoch": 0.7300155603027666,
"grad_norm": 0.11126770824193954,
"learning_rate": 0.00022998204667863553,
"loss": 2.0122556686401367,
"step": 3460
},
{
"epoch": 0.7304375346150803,
"grad_norm": 0.12188728153705597,
"learning_rate": 0.0002296229802513465,
"loss": 1.9030721187591553,
"step": 3462
},
{
"epoch": 0.730859508927394,
"grad_norm": 0.12843221426010132,
"learning_rate": 0.00022926391382405744,
"loss": 2.009836435317993,
"step": 3464
},
{
"epoch": 0.7312814832397078,
"grad_norm": 0.12043307721614838,
"learning_rate": 0.0002289048473967684,
"loss": 1.9517019987106323,
"step": 3466
},
{
"epoch": 0.7317034575520215,
"grad_norm": 0.10166627913713455,
"learning_rate": 0.00022854578096947935,
"loss": 1.9195466041564941,
"step": 3468
},
{
"epoch": 0.7321254318643352,
"grad_norm": 0.11254235357046127,
"learning_rate": 0.00022818671454219032,
"loss": 1.9569883346557617,
"step": 3470
},
{
"epoch": 0.732547406176649,
"grad_norm": 0.10294336825609207,
"learning_rate": 0.00022782764811490126,
"loss": 1.9870234727859497,
"step": 3472
},
{
"epoch": 0.7329693804889628,
"grad_norm": 0.0982080027461052,
"learning_rate": 0.00022746858168761223,
"loss": 2.0036308765411377,
"step": 3474
},
{
"epoch": 0.7333913548012765,
"grad_norm": 0.10740730166435242,
"learning_rate": 0.00022710951526032317,
"loss": 2.003549098968506,
"step": 3476
},
{
"epoch": 0.7338133291135902,
"grad_norm": 0.10025101155042648,
"learning_rate": 0.00022675044883303414,
"loss": 1.9435092210769653,
"step": 3478
},
{
"epoch": 0.7342353034259039,
"grad_norm": 0.10168825089931488,
"learning_rate": 0.00022639138240574509,
"loss": 1.9101588726043701,
"step": 3480
},
{
"epoch": 0.7346572777382177,
"grad_norm": 0.10846757888793945,
"learning_rate": 0.00022603231597845605,
"loss": 1.9815661907196045,
"step": 3482
},
{
"epoch": 0.7350792520505314,
"grad_norm": 0.10572932660579681,
"learning_rate": 0.000225673249551167,
"loss": 1.9580128192901611,
"step": 3484
},
{
"epoch": 0.7355012263628452,
"grad_norm": 0.10809822380542755,
"learning_rate": 0.0002253141831238779,
"loss": 2.00286602973938,
"step": 3486
},
{
"epoch": 0.7359232006751589,
"grad_norm": 0.12528178095817566,
"learning_rate": 0.00022495511669658885,
"loss": 1.9087705612182617,
"step": 3488
},
{
"epoch": 0.7363451749874727,
"grad_norm": 0.11756446212530136,
"learning_rate": 0.00022459605026929982,
"loss": 1.9461390972137451,
"step": 3490
},
{
"epoch": 0.7367671492997864,
"grad_norm": 0.10442161560058594,
"learning_rate": 0.00022423698384201076,
"loss": 2.007925510406494,
"step": 3492
},
{
"epoch": 0.7371891236121001,
"grad_norm": 0.10405369102954865,
"learning_rate": 0.00022387791741472173,
"loss": 1.972633957862854,
"step": 3494
},
{
"epoch": 0.7376110979244138,
"grad_norm": 0.12382562458515167,
"learning_rate": 0.00022351885098743267,
"loss": 1.8730112314224243,
"step": 3496
},
{
"epoch": 0.7380330722367275,
"grad_norm": 0.12369999289512634,
"learning_rate": 0.00022315978456014364,
"loss": 1.9955965280532837,
"step": 3498
},
{
"epoch": 0.7384550465490414,
"grad_norm": 0.10261324048042297,
"learning_rate": 0.00022280071813285458,
"loss": 1.9928297996520996,
"step": 3500
},
{
"epoch": 0.7388770208613551,
"grad_norm": 0.1183568611741066,
"learning_rate": 0.00022244165170556555,
"loss": 1.9984290599822998,
"step": 3502
},
{
"epoch": 0.7392989951736688,
"grad_norm": 0.11929517984390259,
"learning_rate": 0.0002220825852782765,
"loss": 1.9656211137771606,
"step": 3504
},
{
"epoch": 0.7397209694859825,
"grad_norm": 0.12298134714365005,
"learning_rate": 0.00022172351885098746,
"loss": 1.9360415935516357,
"step": 3506
},
{
"epoch": 0.7401429437982963,
"grad_norm": 0.10918302088975906,
"learning_rate": 0.0002213644524236984,
"loss": 1.9081302881240845,
"step": 3508
},
{
"epoch": 0.74056491811061,
"grad_norm": 0.10028845071792603,
"learning_rate": 0.00022100538599640937,
"loss": 1.9460699558258057,
"step": 3510
},
{
"epoch": 0.7409868924229237,
"grad_norm": 0.10435082018375397,
"learning_rate": 0.0002206463195691203,
"loss": 2.016641139984131,
"step": 3512
},
{
"epoch": 0.7414088667352375,
"grad_norm": 0.11152128875255585,
"learning_rate": 0.00022028725314183123,
"loss": 1.9294965267181396,
"step": 3514
},
{
"epoch": 0.7418308410475513,
"grad_norm": 0.10832561552524567,
"learning_rate": 0.0002199281867145422,
"loss": 2.009050130844116,
"step": 3516
},
{
"epoch": 0.742252815359865,
"grad_norm": 0.10893785953521729,
"learning_rate": 0.00021956912028725314,
"loss": 1.946255087852478,
"step": 3518
},
{
"epoch": 0.7426747896721787,
"grad_norm": 0.11153287440538406,
"learning_rate": 0.0002192100538599641,
"loss": 2.011038064956665,
"step": 3520
},
{
"epoch": 0.7430967639844924,
"grad_norm": 0.11881986260414124,
"learning_rate": 0.00021885098743267505,
"loss": 1.9285609722137451,
"step": 3522
},
{
"epoch": 0.7435187382968061,
"grad_norm": 0.10677717626094818,
"learning_rate": 0.000218491921005386,
"loss": 1.9684569835662842,
"step": 3524
},
{
"epoch": 0.7439407126091199,
"grad_norm": 0.11704882979393005,
"learning_rate": 0.00021813285457809696,
"loss": 1.975084662437439,
"step": 3526
},
{
"epoch": 0.7443626869214337,
"grad_norm": 0.1041966900229454,
"learning_rate": 0.0002177737881508079,
"loss": 1.9624989032745361,
"step": 3528
},
{
"epoch": 0.7447846612337474,
"grad_norm": 0.11179523169994354,
"learning_rate": 0.00021741472172351887,
"loss": 1.9390603303909302,
"step": 3530
},
{
"epoch": 0.7452066355460611,
"grad_norm": 0.10088254511356354,
"learning_rate": 0.0002170556552962298,
"loss": 2.016110420227051,
"step": 3532
},
{
"epoch": 0.7456286098583749,
"grad_norm": 0.11715737730264664,
"learning_rate": 0.00021669658886894078,
"loss": 1.9743874073028564,
"step": 3534
},
{
"epoch": 0.7460505841706886,
"grad_norm": 0.12583717703819275,
"learning_rate": 0.00021633752244165172,
"loss": 2.009941577911377,
"step": 3536
},
{
"epoch": 0.7464725584830023,
"grad_norm": 0.10543058812618256,
"learning_rate": 0.0002159784560143627,
"loss": 2.004218578338623,
"step": 3538
},
{
"epoch": 0.746894532795316,
"grad_norm": 0.11218339204788208,
"learning_rate": 0.00021561938958707363,
"loss": 2.0461535453796387,
"step": 3540
},
{
"epoch": 0.7473165071076299,
"grad_norm": 0.1123022735118866,
"learning_rate": 0.00021526032315978455,
"loss": 1.927854299545288,
"step": 3542
},
{
"epoch": 0.7477384814199436,
"grad_norm": 0.10767406225204468,
"learning_rate": 0.00021490125673249551,
"loss": 1.9685540199279785,
"step": 3544
},
{
"epoch": 0.7481604557322573,
"grad_norm": 0.11115457117557526,
"learning_rate": 0.00021454219030520646,
"loss": 1.9466699361801147,
"step": 3546
},
{
"epoch": 0.748582430044571,
"grad_norm": 0.11124234646558762,
"learning_rate": 0.00021418312387791742,
"loss": 1.9735201597213745,
"step": 3548
},
{
"epoch": 0.7490044043568848,
"grad_norm": 0.11646565049886703,
"learning_rate": 0.00021382405745062837,
"loss": 1.941564917564392,
"step": 3550
},
{
"epoch": 0.7494263786691985,
"grad_norm": 0.12785542011260986,
"learning_rate": 0.00021346499102333933,
"loss": 1.9586703777313232,
"step": 3552
},
{
"epoch": 0.7498483529815122,
"grad_norm": 0.10879122465848923,
"learning_rate": 0.00021310592459605028,
"loss": 1.993349313735962,
"step": 3554
},
{
"epoch": 0.750270327293826,
"grad_norm": 0.1141117587685585,
"learning_rate": 0.00021274685816876124,
"loss": 1.9418561458587646,
"step": 3556
},
{
"epoch": 0.7506923016061398,
"grad_norm": 0.11264986544847488,
"learning_rate": 0.00021238779174147219,
"loss": 1.9159716367721558,
"step": 3558
},
{
"epoch": 0.7511142759184535,
"grad_norm": 0.11546836048364639,
"learning_rate": 0.00021202872531418313,
"loss": 1.9348828792572021,
"step": 3560
},
{
"epoch": 0.7515362502307672,
"grad_norm": 0.10498052835464478,
"learning_rate": 0.0002116696588868941,
"loss": 1.9074705839157104,
"step": 3562
},
{
"epoch": 0.7519582245430809,
"grad_norm": 0.09647338837385178,
"learning_rate": 0.00021131059245960504,
"loss": 1.9550995826721191,
"step": 3564
},
{
"epoch": 0.7523801988553946,
"grad_norm": 0.11199572682380676,
"learning_rate": 0.000210951526032316,
"loss": 1.9723422527313232,
"step": 3566
},
{
"epoch": 0.7528021731677084,
"grad_norm": 0.1104026585817337,
"learning_rate": 0.00021059245960502695,
"loss": 1.995705485343933,
"step": 3568
},
{
"epoch": 0.7532241474800222,
"grad_norm": 0.11171594262123108,
"learning_rate": 0.00021023339317773786,
"loss": 1.9952317476272583,
"step": 3570
},
{
"epoch": 0.7536461217923359,
"grad_norm": 0.11041028052568436,
"learning_rate": 0.00020987432675044883,
"loss": 2.0292041301727295,
"step": 3572
},
{
"epoch": 0.7540680961046496,
"grad_norm": 0.13038399815559387,
"learning_rate": 0.00020951526032315977,
"loss": 1.9572522640228271,
"step": 3574
},
{
"epoch": 0.7544900704169634,
"grad_norm": 0.10326897352933884,
"learning_rate": 0.00020915619389587074,
"loss": 1.9925856590270996,
"step": 3576
},
{
"epoch": 0.7549120447292771,
"grad_norm": 0.11976686120033264,
"learning_rate": 0.00020879712746858168,
"loss": 1.9388549327850342,
"step": 3578
},
{
"epoch": 0.7553340190415908,
"grad_norm": 0.10051357746124268,
"learning_rate": 0.00020843806104129265,
"loss": 2.0152170658111572,
"step": 3580
},
{
"epoch": 0.7557559933539045,
"grad_norm": 0.10236506164073944,
"learning_rate": 0.0002080789946140036,
"loss": 1.9863691329956055,
"step": 3582
},
{
"epoch": 0.7561779676662184,
"grad_norm": 0.10837408155202866,
"learning_rate": 0.00020771992818671456,
"loss": 1.9722026586532593,
"step": 3584
},
{
"epoch": 0.7565999419785321,
"grad_norm": 0.1103808656334877,
"learning_rate": 0.0002073608617594255,
"loss": 1.914790153503418,
"step": 3586
},
{
"epoch": 0.7570219162908458,
"grad_norm": 0.10499264299869537,
"learning_rate": 0.00020700179533213647,
"loss": 2.012113332748413,
"step": 3588
},
{
"epoch": 0.7574438906031595,
"grad_norm": 0.10326241701841354,
"learning_rate": 0.00020664272890484741,
"loss": 1.989783763885498,
"step": 3590
},
{
"epoch": 0.7578658649154733,
"grad_norm": 0.10843243449926376,
"learning_rate": 0.00020628366247755838,
"loss": 1.9778681993484497,
"step": 3592
},
{
"epoch": 0.758287839227787,
"grad_norm": 0.1043754518032074,
"learning_rate": 0.00020592459605026932,
"loss": 1.9813368320465088,
"step": 3594
},
{
"epoch": 0.7587098135401007,
"grad_norm": 0.10203973948955536,
"learning_rate": 0.0002055655296229803,
"loss": 1.9706125259399414,
"step": 3596
},
{
"epoch": 0.7591317878524145,
"grad_norm": 0.10393171012401581,
"learning_rate": 0.00020520646319569123,
"loss": 1.9564557075500488,
"step": 3598
},
{
"epoch": 0.7595537621647283,
"grad_norm": 0.11524060368537903,
"learning_rate": 0.00020484739676840215,
"loss": 2.0136327743530273,
"step": 3600
},
{
"epoch": 0.759975736477042,
"grad_norm": 0.10547072440385818,
"learning_rate": 0.0002044883303411131,
"loss": 2.008685827255249,
"step": 3602
},
{
"epoch": 0.7603977107893557,
"grad_norm": 0.1118752509355545,
"learning_rate": 0.00020412926391382406,
"loss": 1.9472808837890625,
"step": 3604
},
{
"epoch": 0.7608196851016694,
"grad_norm": 0.12786750495433807,
"learning_rate": 0.000203770197486535,
"loss": 2.009627342224121,
"step": 3606
},
{
"epoch": 0.7612416594139831,
"grad_norm": 0.1082206666469574,
"learning_rate": 0.00020341113105924597,
"loss": 1.944786787033081,
"step": 3608
},
{
"epoch": 0.7616636337262969,
"grad_norm": 0.10475423187017441,
"learning_rate": 0.0002030520646319569,
"loss": 1.9330317974090576,
"step": 3610
},
{
"epoch": 0.7620856080386107,
"grad_norm": 0.11662724614143372,
"learning_rate": 0.00020269299820466788,
"loss": 1.9484472274780273,
"step": 3612
},
{
"epoch": 0.7625075823509244,
"grad_norm": 0.1409110575914383,
"learning_rate": 0.00020233393177737882,
"loss": 2.013054609298706,
"step": 3614
},
{
"epoch": 0.7629295566632381,
"grad_norm": 0.11422674357891083,
"learning_rate": 0.0002019748653500898,
"loss": 1.973104476928711,
"step": 3616
},
{
"epoch": 0.7633515309755519,
"grad_norm": 0.1183161586523056,
"learning_rate": 0.00020161579892280073,
"loss": 1.9923210144042969,
"step": 3618
},
{
"epoch": 0.7637735052878656,
"grad_norm": 0.11443831026554108,
"learning_rate": 0.0002012567324955117,
"loss": 1.975514531135559,
"step": 3620
},
{
"epoch": 0.7641954796001793,
"grad_norm": 0.11664769798517227,
"learning_rate": 0.00020089766606822264,
"loss": 1.979424238204956,
"step": 3622
},
{
"epoch": 0.764617453912493,
"grad_norm": 0.10687406361103058,
"learning_rate": 0.0002005385996409336,
"loss": 2.008500337600708,
"step": 3624
},
{
"epoch": 0.7650394282248069,
"grad_norm": 0.1044531837105751,
"learning_rate": 0.00020017953321364455,
"loss": 1.9567381143569946,
"step": 3626
},
{
"epoch": 0.7654614025371206,
"grad_norm": 0.0970902144908905,
"learning_rate": 0.0001998204667863555,
"loss": 1.947081208229065,
"step": 3628
},
{
"epoch": 0.7658833768494343,
"grad_norm": 0.10819247364997864,
"learning_rate": 0.00019946140035906644,
"loss": 1.9638906717300415,
"step": 3630
},
{
"epoch": 0.766305351161748,
"grad_norm": 0.11260806769132614,
"learning_rate": 0.0001991023339317774,
"loss": 2.0301337242126465,
"step": 3632
},
{
"epoch": 0.7667273254740617,
"grad_norm": 0.11109425127506256,
"learning_rate": 0.00019874326750448835,
"loss": 2.0169131755828857,
"step": 3634
},
{
"epoch": 0.7671492997863755,
"grad_norm": 0.10612336546182632,
"learning_rate": 0.0001983842010771993,
"loss": 1.917722225189209,
"step": 3636
},
{
"epoch": 0.7675712740986892,
"grad_norm": 0.10526350885629654,
"learning_rate": 0.00019802513464991023,
"loss": 1.9953590631484985,
"step": 3638
},
{
"epoch": 0.767993248411003,
"grad_norm": 0.14485600590705872,
"learning_rate": 0.0001976660682226212,
"loss": 1.9435768127441406,
"step": 3640
},
{
"epoch": 0.7684152227233167,
"grad_norm": 0.11948107182979584,
"learning_rate": 0.00019730700179533214,
"loss": 2.0030925273895264,
"step": 3642
},
{
"epoch": 0.7688371970356305,
"grad_norm": 0.10552872717380524,
"learning_rate": 0.0001969479353680431,
"loss": 1.9689610004425049,
"step": 3644
},
{
"epoch": 0.7692591713479442,
"grad_norm": 0.09945666044950485,
"learning_rate": 0.00019658886894075405,
"loss": 2.0131256580352783,
"step": 3646
},
{
"epoch": 0.7696811456602579,
"grad_norm": 0.10466771572828293,
"learning_rate": 0.00019622980251346502,
"loss": 2.0326340198516846,
"step": 3648
},
{
"epoch": 0.7701031199725716,
"grad_norm": 0.10571710765361786,
"learning_rate": 0.00019587073608617593,
"loss": 2.020745038986206,
"step": 3650
},
{
"epoch": 0.7705250942848854,
"grad_norm": 0.11482807993888855,
"learning_rate": 0.0001955116696588869,
"loss": 1.9489401578903198,
"step": 3652
},
{
"epoch": 0.7709470685971992,
"grad_norm": 0.10952337086200714,
"learning_rate": 0.00019515260323159784,
"loss": 1.9974874258041382,
"step": 3654
},
{
"epoch": 0.7713690429095129,
"grad_norm": 0.10379087179899216,
"learning_rate": 0.0001947935368043088,
"loss": 1.8910999298095703,
"step": 3656
},
{
"epoch": 0.7717910172218266,
"grad_norm": 0.10733562707901001,
"learning_rate": 0.00019443447037701975,
"loss": 1.9567420482635498,
"step": 3658
},
{
"epoch": 0.7722129915341404,
"grad_norm": 0.10968147218227386,
"learning_rate": 0.00019407540394973072,
"loss": 1.9608248472213745,
"step": 3660
},
{
"epoch": 0.7726349658464541,
"grad_norm": 0.10474677383899689,
"learning_rate": 0.00019371633752244166,
"loss": 2.0230579376220703,
"step": 3662
},
{
"epoch": 0.7730569401587678,
"grad_norm": 0.09309519082307816,
"learning_rate": 0.0001933572710951526,
"loss": 1.9471436738967896,
"step": 3664
},
{
"epoch": 0.7734789144710815,
"grad_norm": 0.1076444759964943,
"learning_rate": 0.00019299820466786355,
"loss": 1.973122477531433,
"step": 3666
},
{
"epoch": 0.7739008887833954,
"grad_norm": 0.1062762439250946,
"learning_rate": 0.00019263913824057452,
"loss": 1.9883229732513428,
"step": 3668
},
{
"epoch": 0.7743228630957091,
"grad_norm": 0.10461197793483734,
"learning_rate": 0.00019228007181328546,
"loss": 2.0011940002441406,
"step": 3670
},
{
"epoch": 0.7747448374080228,
"grad_norm": 0.10479731112718582,
"learning_rate": 0.00019192100538599643,
"loss": 2.011416435241699,
"step": 3672
},
{
"epoch": 0.7751668117203365,
"grad_norm": 0.11142764985561371,
"learning_rate": 0.00019156193895870737,
"loss": 2.001660108566284,
"step": 3674
},
{
"epoch": 0.7755887860326502,
"grad_norm": 0.1089906245470047,
"learning_rate": 0.00019120287253141834,
"loss": 2.006784677505493,
"step": 3676
},
{
"epoch": 0.776010760344964,
"grad_norm": 0.11439166963100433,
"learning_rate": 0.00019084380610412925,
"loss": 1.9675712585449219,
"step": 3678
},
{
"epoch": 0.7764327346572777,
"grad_norm": 0.10382472723722458,
"learning_rate": 0.00019048473967684022,
"loss": 1.948840618133545,
"step": 3680
},
{
"epoch": 0.7768547089695915,
"grad_norm": 0.10940327495336533,
"learning_rate": 0.00019012567324955116,
"loss": 1.9841792583465576,
"step": 3682
},
{
"epoch": 0.7772766832819052,
"grad_norm": 0.12075155973434448,
"learning_rate": 0.00018976660682226213,
"loss": 1.9691226482391357,
"step": 3684
},
{
"epoch": 0.777698657594219,
"grad_norm": 0.11114417761564255,
"learning_rate": 0.00018940754039497307,
"loss": 1.9727588891983032,
"step": 3686
},
{
"epoch": 0.7781206319065327,
"grad_norm": 0.11416927725076675,
"learning_rate": 0.00018904847396768404,
"loss": 1.9640755653381348,
"step": 3688
},
{
"epoch": 0.7785426062188464,
"grad_norm": 0.102940633893013,
"learning_rate": 0.00018868940754039498,
"loss": 1.9593833684921265,
"step": 3690
},
{
"epoch": 0.7789645805311601,
"grad_norm": 0.12227679044008255,
"learning_rate": 0.00018833034111310592,
"loss": 1.9931222200393677,
"step": 3692
},
{
"epoch": 0.7793865548434739,
"grad_norm": 0.11781952530145645,
"learning_rate": 0.0001879712746858169,
"loss": 1.9936915636062622,
"step": 3694
},
{
"epoch": 0.7798085291557877,
"grad_norm": 0.1044093668460846,
"learning_rate": 0.00018761220825852783,
"loss": 1.9988365173339844,
"step": 3696
},
{
"epoch": 0.7802305034681014,
"grad_norm": 0.11550258100032806,
"learning_rate": 0.00018725314183123877,
"loss": 1.9739943742752075,
"step": 3698
},
{
"epoch": 0.7806524777804151,
"grad_norm": 0.10740198194980621,
"learning_rate": 0.00018689407540394974,
"loss": 1.9282251596450806,
"step": 3700
},
{
"epoch": 0.7810744520927289,
"grad_norm": 0.1079789400100708,
"learning_rate": 0.00018653500897666068,
"loss": 1.9978529214859009,
"step": 3702
},
{
"epoch": 0.7814964264050426,
"grad_norm": 0.10162225365638733,
"learning_rate": 0.00018617594254937165,
"loss": 1.9327689409255981,
"step": 3704
},
{
"epoch": 0.7819184007173563,
"grad_norm": 0.10244615375995636,
"learning_rate": 0.0001858168761220826,
"loss": 1.9205598831176758,
"step": 3706
},
{
"epoch": 0.78234037502967,
"grad_norm": 0.10141397267580032,
"learning_rate": 0.00018545780969479354,
"loss": 2.0372838973999023,
"step": 3708
},
{
"epoch": 0.7827623493419839,
"grad_norm": 0.10559744387865067,
"learning_rate": 0.0001850987432675045,
"loss": 1.943070411682129,
"step": 3710
},
{
"epoch": 0.7831843236542976,
"grad_norm": 0.10123951733112335,
"learning_rate": 0.00018473967684021545,
"loss": 2.039090633392334,
"step": 3712
},
{
"epoch": 0.7836062979666113,
"grad_norm": 0.09855969995260239,
"learning_rate": 0.00018438061041292642,
"loss": 1.9554226398468018,
"step": 3714
},
{
"epoch": 0.784028272278925,
"grad_norm": 0.104286327958107,
"learning_rate": 0.00018402154398563736,
"loss": 1.8781663179397583,
"step": 3716
},
{
"epoch": 0.7844502465912387,
"grad_norm": 0.11810514330863953,
"learning_rate": 0.0001836624775583483,
"loss": 1.983089804649353,
"step": 3718
},
{
"epoch": 0.7848722209035525,
"grad_norm": 0.10730182379484177,
"learning_rate": 0.00018330341113105924,
"loss": 2.0369110107421875,
"step": 3720
},
{
"epoch": 0.7852941952158662,
"grad_norm": 0.1488947719335556,
"learning_rate": 0.0001829443447037702,
"loss": 2.0092146396636963,
"step": 3722
},
{
"epoch": 0.78571616952818,
"grad_norm": 0.10151592642068863,
"learning_rate": 0.00018258527827648115,
"loss": 1.9893856048583984,
"step": 3724
},
{
"epoch": 0.7861381438404937,
"grad_norm": 0.11504636704921722,
"learning_rate": 0.00018222621184919212,
"loss": 1.9818978309631348,
"step": 3726
},
{
"epoch": 0.7865601181528075,
"grad_norm": 0.11524399369955063,
"learning_rate": 0.00018186714542190306,
"loss": 2.0468122959136963,
"step": 3728
},
{
"epoch": 0.7869820924651212,
"grad_norm": 0.11578702181577682,
"learning_rate": 0.00018150807899461403,
"loss": 2.0041885375976562,
"step": 3730
},
{
"epoch": 0.7874040667774349,
"grad_norm": 0.10229094326496124,
"learning_rate": 0.00018114901256732497,
"loss": 1.9992265701293945,
"step": 3732
},
{
"epoch": 0.7878260410897486,
"grad_norm": 0.10223471373319626,
"learning_rate": 0.0001807899461400359,
"loss": 1.998656153678894,
"step": 3734
},
{
"epoch": 0.7882480154020624,
"grad_norm": 0.11094488203525543,
"learning_rate": 0.00018043087971274685,
"loss": 1.9252221584320068,
"step": 3736
},
{
"epoch": 0.7886699897143762,
"grad_norm": 0.1004694476723671,
"learning_rate": 0.00018007181328545782,
"loss": 1.9413923025131226,
"step": 3738
},
{
"epoch": 0.7890919640266899,
"grad_norm": 0.12001891434192657,
"learning_rate": 0.00017971274685816876,
"loss": 1.925302267074585,
"step": 3740
},
{
"epoch": 0.7895139383390036,
"grad_norm": 0.10260952264070511,
"learning_rate": 0.00017935368043087973,
"loss": 1.994821548461914,
"step": 3742
},
{
"epoch": 0.7899359126513174,
"grad_norm": 0.10674403607845306,
"learning_rate": 0.00017899461400359067,
"loss": 2.0333096981048584,
"step": 3744
},
{
"epoch": 0.7903578869636311,
"grad_norm": 0.12484647333621979,
"learning_rate": 0.00017863554757630164,
"loss": 1.941068172454834,
"step": 3746
},
{
"epoch": 0.7907798612759448,
"grad_norm": 0.11063817143440247,
"learning_rate": 0.00017827648114901259,
"loss": 2.0024468898773193,
"step": 3748
},
{
"epoch": 0.7912018355882585,
"grad_norm": 0.12851890921592712,
"learning_rate": 0.00017791741472172353,
"loss": 1.9324982166290283,
"step": 3750
},
{
"epoch": 0.7916238099005724,
"grad_norm": 0.1201174259185791,
"learning_rate": 0.00017755834829443447,
"loss": 2.037789821624756,
"step": 3752
},
{
"epoch": 0.7920457842128861,
"grad_norm": 0.11184491217136383,
"learning_rate": 0.00017719928186714544,
"loss": 1.9129966497421265,
"step": 3754
},
{
"epoch": 0.7924677585251998,
"grad_norm": 0.11286237835884094,
"learning_rate": 0.00017684021543985638,
"loss": 1.9537498950958252,
"step": 3756
},
{
"epoch": 0.7928897328375135,
"grad_norm": 0.10127856582403183,
"learning_rate": 0.00017648114901256735,
"loss": 1.9376894235610962,
"step": 3758
},
{
"epoch": 0.7933117071498272,
"grad_norm": 0.09692607074975967,
"learning_rate": 0.0001761220825852783,
"loss": 1.9800184965133667,
"step": 3760
},
{
"epoch": 0.793733681462141,
"grad_norm": 0.11241001635789871,
"learning_rate": 0.00017576301615798926,
"loss": 1.9580527544021606,
"step": 3762
},
{
"epoch": 0.7941556557744547,
"grad_norm": 0.1285153180360794,
"learning_rate": 0.00017540394973070017,
"loss": 2.015336751937866,
"step": 3764
},
{
"epoch": 0.7945776300867685,
"grad_norm": 0.10907573997974396,
"learning_rate": 0.00017504488330341114,
"loss": 1.9198459386825562,
"step": 3766
},
{
"epoch": 0.7949996043990822,
"grad_norm": 0.10614390671253204,
"learning_rate": 0.00017468581687612208,
"loss": 2.0223240852355957,
"step": 3768
},
{
"epoch": 0.795421578711396,
"grad_norm": 0.10694452375173569,
"learning_rate": 0.00017432675044883305,
"loss": 1.9959814548492432,
"step": 3770
},
{
"epoch": 0.7958435530237097,
"grad_norm": 0.10280728340148926,
"learning_rate": 0.000173967684021544,
"loss": 1.9737062454223633,
"step": 3772
},
{
"epoch": 0.7962655273360234,
"grad_norm": 0.12359616905450821,
"learning_rate": 0.00017360861759425496,
"loss": 2.0026590824127197,
"step": 3774
},
{
"epoch": 0.7966875016483371,
"grad_norm": 0.11161927133798599,
"learning_rate": 0.0001732495511669659,
"loss": 1.965014934539795,
"step": 3776
},
{
"epoch": 0.7971094759606508,
"grad_norm": 0.10933814197778702,
"learning_rate": 0.00017289048473967684,
"loss": 2.017759084701538,
"step": 3778
},
{
"epoch": 0.7975314502729647,
"grad_norm": 0.10623813420534134,
"learning_rate": 0.00017253141831238779,
"loss": 1.96980881690979,
"step": 3780
},
{
"epoch": 0.7979534245852784,
"grad_norm": 0.10798842459917068,
"learning_rate": 0.00017217235188509875,
"loss": 1.9675899744033813,
"step": 3782
},
{
"epoch": 0.7983753988975921,
"grad_norm": 0.10603086650371552,
"learning_rate": 0.0001718132854578097,
"loss": 1.965511679649353,
"step": 3784
},
{
"epoch": 0.7987973732099058,
"grad_norm": 0.12332173436880112,
"learning_rate": 0.00017145421903052066,
"loss": 1.9538955688476562,
"step": 3786
},
{
"epoch": 0.7992193475222196,
"grad_norm": 0.131839781999588,
"learning_rate": 0.0001710951526032316,
"loss": 2.0402355194091797,
"step": 3788
},
{
"epoch": 0.7996413218345333,
"grad_norm": 0.11338990926742554,
"learning_rate": 0.00017073608617594258,
"loss": 1.932624340057373,
"step": 3790
},
{
"epoch": 0.800063296146847,
"grad_norm": 0.10707061737775803,
"learning_rate": 0.0001703770197486535,
"loss": 1.9825648069381714,
"step": 3792
},
{
"epoch": 0.800063296146847,
"eval_loss": 1.9726277589797974,
"eval_runtime": 87.4577,
"eval_samples_per_second": 35.034,
"eval_steps_per_second": 4.379,
"step": 3792
},
{
"epoch": 0.8004852704591608,
"grad_norm": 0.10969210416078568,
"learning_rate": 0.00017001795332136446,
"loss": 2.002253532409668,
"step": 3794
},
{
"epoch": 0.8009072447714746,
"grad_norm": 0.11934729665517807,
"learning_rate": 0.0001696588868940754,
"loss": 1.9382480382919312,
"step": 3796
},
{
"epoch": 0.8013292190837883,
"grad_norm": 0.10215841978788376,
"learning_rate": 0.00016929982046678637,
"loss": 1.9855782985687256,
"step": 3798
},
{
"epoch": 0.801751193396102,
"grad_norm": 0.11482446640729904,
"learning_rate": 0.0001689407540394973,
"loss": 1.9739387035369873,
"step": 3800
},
{
"epoch": 0.8021731677084157,
"grad_norm": 0.11387546360492706,
"learning_rate": 0.00016858168761220828,
"loss": 1.833898901939392,
"step": 3802
},
{
"epoch": 0.8025951420207295,
"grad_norm": 0.10798979550600052,
"learning_rate": 0.00016822262118491922,
"loss": 1.9892562627792358,
"step": 3804
},
{
"epoch": 0.8030171163330432,
"grad_norm": 0.12746727466583252,
"learning_rate": 0.00016786355475763016,
"loss": 1.9931572675704956,
"step": 3806
},
{
"epoch": 0.803439090645357,
"grad_norm": 0.10256224125623703,
"learning_rate": 0.0001675044883303411,
"loss": 1.9989410638809204,
"step": 3808
},
{
"epoch": 0.8038610649576707,
"grad_norm": 0.1065383329987526,
"learning_rate": 0.00016714542190305207,
"loss": 2.022404670715332,
"step": 3810
},
{
"epoch": 0.8042830392699845,
"grad_norm": 0.11076526343822479,
"learning_rate": 0.00016678635547576301,
"loss": 1.9794429540634155,
"step": 3812
},
{
"epoch": 0.8047050135822982,
"grad_norm": 0.10274726152420044,
"learning_rate": 0.00016642728904847398,
"loss": 1.996565341949463,
"step": 3814
},
{
"epoch": 0.8051269878946119,
"grad_norm": 0.10509247332811356,
"learning_rate": 0.00016606822262118492,
"loss": 1.9526735544204712,
"step": 3816
},
{
"epoch": 0.8055489622069256,
"grad_norm": 0.10226453840732574,
"learning_rate": 0.0001657091561938959,
"loss": 1.9855551719665527,
"step": 3818
},
{
"epoch": 0.8059709365192393,
"grad_norm": 0.11684133857488632,
"learning_rate": 0.0001653500897666068,
"loss": 1.9855530261993408,
"step": 3820
},
{
"epoch": 0.8063929108315532,
"grad_norm": 0.10857287794351578,
"learning_rate": 0.00016499102333931778,
"loss": 1.9979839324951172,
"step": 3822
},
{
"epoch": 0.8068148851438669,
"grad_norm": 0.10480791330337524,
"learning_rate": 0.00016463195691202872,
"loss": 1.9737151861190796,
"step": 3824
},
{
"epoch": 0.8072368594561806,
"grad_norm": 0.10021811723709106,
"learning_rate": 0.00016427289048473969,
"loss": 1.9424771070480347,
"step": 3826
},
{
"epoch": 0.8076588337684943,
"grad_norm": 0.10852868109941483,
"learning_rate": 0.00016391382405745063,
"loss": 1.9553169012069702,
"step": 3828
},
{
"epoch": 0.8080808080808081,
"grad_norm": 0.11382559686899185,
"learning_rate": 0.0001635547576301616,
"loss": 1.958644151687622,
"step": 3830
},
{
"epoch": 0.8085027823931218,
"grad_norm": 0.1157197579741478,
"learning_rate": 0.00016319569120287254,
"loss": 1.9857683181762695,
"step": 3832
},
{
"epoch": 0.8089247567054355,
"grad_norm": 0.1109665185213089,
"learning_rate": 0.0001628366247755835,
"loss": 1.9512416124343872,
"step": 3834
},
{
"epoch": 0.8093467310177493,
"grad_norm": 0.10457272827625275,
"learning_rate": 0.00016247755834829442,
"loss": 1.978639006614685,
"step": 3836
},
{
"epoch": 0.8097687053300631,
"grad_norm": 0.10810290277004242,
"learning_rate": 0.0001621184919210054,
"loss": 1.980324387550354,
"step": 3838
},
{
"epoch": 0.8101906796423768,
"grad_norm": 0.13540904223918915,
"learning_rate": 0.00016175942549371633,
"loss": 1.9555515050888062,
"step": 3840
},
{
"epoch": 0.8106126539546905,
"grad_norm": 0.10845515131950378,
"learning_rate": 0.0001614003590664273,
"loss": 1.904801607131958,
"step": 3842
},
{
"epoch": 0.8110346282670042,
"grad_norm": 0.10743451118469238,
"learning_rate": 0.00016104129263913824,
"loss": 1.9551976919174194,
"step": 3844
},
{
"epoch": 0.811456602579318,
"grad_norm": 0.11844830960035324,
"learning_rate": 0.0001606822262118492,
"loss": 1.9656059741973877,
"step": 3846
},
{
"epoch": 0.8118785768916317,
"grad_norm": 0.10831819474697113,
"learning_rate": 0.00016032315978456015,
"loss": 1.9628654718399048,
"step": 3848
},
{
"epoch": 0.8123005512039455,
"grad_norm": 0.09924059361219406,
"learning_rate": 0.0001599640933572711,
"loss": 1.9888867139816284,
"step": 3850
},
{
"epoch": 0.8127225255162592,
"grad_norm": 0.1120271310210228,
"learning_rate": 0.00015960502692998206,
"loss": 1.9073693752288818,
"step": 3852
},
{
"epoch": 0.813144499828573,
"grad_norm": 0.11557640135288239,
"learning_rate": 0.000159245960502693,
"loss": 1.9355658292770386,
"step": 3854
},
{
"epoch": 0.8135664741408867,
"grad_norm": 0.1026308611035347,
"learning_rate": 0.00015888689407540395,
"loss": 1.956674337387085,
"step": 3856
},
{
"epoch": 0.8139884484532004,
"grad_norm": 0.09923417866230011,
"learning_rate": 0.00015852782764811491,
"loss": 2.0172617435455322,
"step": 3858
},
{
"epoch": 0.8144104227655141,
"grad_norm": 0.11352282017469406,
"learning_rate": 0.00015816876122082586,
"loss": 1.9894676208496094,
"step": 3860
},
{
"epoch": 0.8148323970778278,
"grad_norm": 0.11508678644895554,
"learning_rate": 0.00015780969479353682,
"loss": 1.9896647930145264,
"step": 3862
},
{
"epoch": 0.8152543713901417,
"grad_norm": 0.09686668962240219,
"learning_rate": 0.00015745062836624777,
"loss": 1.988989233970642,
"step": 3864
},
{
"epoch": 0.8156763457024554,
"grad_norm": 0.10609009116888046,
"learning_rate": 0.0001570915619389587,
"loss": 2.00748348236084,
"step": 3866
},
{
"epoch": 0.8160983200147691,
"grad_norm": 0.09999443590641022,
"learning_rate": 0.00015673249551166968,
"loss": 1.97810959815979,
"step": 3868
},
{
"epoch": 0.8165202943270828,
"grad_norm": 0.10757757723331451,
"learning_rate": 0.00015637342908438062,
"loss": 1.9655009508132935,
"step": 3870
},
{
"epoch": 0.8169422686393966,
"grad_norm": 0.10482610762119293,
"learning_rate": 0.00015601436265709156,
"loss": 2.0134100914001465,
"step": 3872
},
{
"epoch": 0.8173642429517103,
"grad_norm": 0.11153223365545273,
"learning_rate": 0.00015565529622980253,
"loss": 1.9545120000839233,
"step": 3874
},
{
"epoch": 0.817786217264024,
"grad_norm": 0.10973742604255676,
"learning_rate": 0.00015529622980251347,
"loss": 1.9370464086532593,
"step": 3876
},
{
"epoch": 0.8182081915763378,
"grad_norm": 0.09853997826576233,
"learning_rate": 0.0001549371633752244,
"loss": 2.0483956336975098,
"step": 3878
},
{
"epoch": 0.8186301658886516,
"grad_norm": 0.12285112589597702,
"learning_rate": 0.00015457809694793538,
"loss": 1.9881095886230469,
"step": 3880
},
{
"epoch": 0.8190521402009653,
"grad_norm": 0.10819984972476959,
"learning_rate": 0.00015421903052064632,
"loss": 1.9612258672714233,
"step": 3882
},
{
"epoch": 0.819474114513279,
"grad_norm": 0.11285301297903061,
"learning_rate": 0.0001538599640933573,
"loss": 1.9877020120620728,
"step": 3884
},
{
"epoch": 0.8198960888255927,
"grad_norm": 0.1077396348118782,
"learning_rate": 0.00015350089766606823,
"loss": 1.99911367893219,
"step": 3886
},
{
"epoch": 0.8203180631379065,
"grad_norm": 0.1047162190079689,
"learning_rate": 0.0001531418312387792,
"loss": 1.9960293769836426,
"step": 3888
},
{
"epoch": 0.8207400374502202,
"grad_norm": 0.11399483680725098,
"learning_rate": 0.00015278276481149014,
"loss": 1.942463994026184,
"step": 3890
},
{
"epoch": 0.821162011762534,
"grad_norm": 0.13342727720737457,
"learning_rate": 0.00015242369838420108,
"loss": 1.9544308185577393,
"step": 3892
},
{
"epoch": 0.8215839860748477,
"grad_norm": 0.10347689688205719,
"learning_rate": 0.00015206463195691203,
"loss": 1.9599984884262085,
"step": 3894
},
{
"epoch": 0.8220059603871614,
"grad_norm": 0.11319886147975922,
"learning_rate": 0.000151705565529623,
"loss": 2.025538444519043,
"step": 3896
},
{
"epoch": 0.8224279346994752,
"grad_norm": 0.1163545623421669,
"learning_rate": 0.00015134649910233394,
"loss": 1.9560476541519165,
"step": 3898
},
{
"epoch": 0.8228499090117889,
"grad_norm": 0.12468685209751129,
"learning_rate": 0.0001509874326750449,
"loss": 2.0232272148132324,
"step": 3900
},
{
"epoch": 0.8232718833241026,
"grad_norm": 0.10980268567800522,
"learning_rate": 0.00015062836624775585,
"loss": 2.0195233821868896,
"step": 3902
},
{
"epoch": 0.8236938576364163,
"grad_norm": 0.10719434171915054,
"learning_rate": 0.00015026929982046681,
"loss": 1.9837669134140015,
"step": 3904
},
{
"epoch": 0.8241158319487302,
"grad_norm": 0.10639816522598267,
"learning_rate": 0.00014991023339317773,
"loss": 1.9651360511779785,
"step": 3906
},
{
"epoch": 0.8245378062610439,
"grad_norm": 0.10504397004842758,
"learning_rate": 0.0001495511669658887,
"loss": 1.8785459995269775,
"step": 3908
},
{
"epoch": 0.8249597805733576,
"grad_norm": 0.10565751791000366,
"learning_rate": 0.00014919210053859964,
"loss": 2.0334479808807373,
"step": 3910
},
{
"epoch": 0.8253817548856713,
"grad_norm": 0.11844756454229355,
"learning_rate": 0.0001488330341113106,
"loss": 1.9758542776107788,
"step": 3912
},
{
"epoch": 0.8258037291979851,
"grad_norm": 0.11583025753498077,
"learning_rate": 0.00014847396768402155,
"loss": 2.0450258255004883,
"step": 3914
},
{
"epoch": 0.8262257035102988,
"grad_norm": 0.1050402820110321,
"learning_rate": 0.00014811490125673252,
"loss": 1.972878336906433,
"step": 3916
},
{
"epoch": 0.8266476778226125,
"grad_norm": 0.1121138483285904,
"learning_rate": 0.00014775583482944346,
"loss": 1.9859211444854736,
"step": 3918
},
{
"epoch": 0.8270696521349263,
"grad_norm": 0.10710328072309494,
"learning_rate": 0.0001473967684021544,
"loss": 1.9839608669281006,
"step": 3920
},
{
"epoch": 0.8274916264472401,
"grad_norm": 0.10813052952289581,
"learning_rate": 0.00014703770197486534,
"loss": 2.0114970207214355,
"step": 3922
},
{
"epoch": 0.8279136007595538,
"grad_norm": 0.10946498066186905,
"learning_rate": 0.0001466786355475763,
"loss": 1.9662872552871704,
"step": 3924
},
{
"epoch": 0.8283355750718675,
"grad_norm": 0.09399878978729248,
"learning_rate": 0.00014631956912028725,
"loss": 2.0013427734375,
"step": 3926
},
{
"epoch": 0.8287575493841812,
"grad_norm": 0.09642825275659561,
"learning_rate": 0.00014596050269299822,
"loss": 1.9572128057479858,
"step": 3928
},
{
"epoch": 0.829179523696495,
"grad_norm": 0.10715366899967194,
"learning_rate": 0.00014560143626570916,
"loss": 1.9087451696395874,
"step": 3930
},
{
"epoch": 0.8296014980088087,
"grad_norm": 0.10492988675832748,
"learning_rate": 0.00014524236983842013,
"loss": 1.9729474782943726,
"step": 3932
},
{
"epoch": 0.8300234723211225,
"grad_norm": 0.13220976293087006,
"learning_rate": 0.00014488330341113107,
"loss": 1.9855540990829468,
"step": 3934
},
{
"epoch": 0.8304454466334362,
"grad_norm": 0.10412155091762543,
"learning_rate": 0.00014452423698384202,
"loss": 1.9694067239761353,
"step": 3936
},
{
"epoch": 0.83086742094575,
"grad_norm": 0.1190498098731041,
"learning_rate": 0.00014416517055655296,
"loss": 2.044698476791382,
"step": 3938
},
{
"epoch": 0.8312893952580637,
"grad_norm": 0.10366658866405487,
"learning_rate": 0.00014380610412926393,
"loss": 1.942368745803833,
"step": 3940
},
{
"epoch": 0.8317113695703774,
"grad_norm": 0.10109183192253113,
"learning_rate": 0.00014344703770197487,
"loss": 2.008997917175293,
"step": 3942
},
{
"epoch": 0.8321333438826911,
"grad_norm": 0.105455681681633,
"learning_rate": 0.00014308797127468584,
"loss": 1.9734221696853638,
"step": 3944
},
{
"epoch": 0.8325553181950048,
"grad_norm": 0.13118450343608856,
"learning_rate": 0.00014272890484739678,
"loss": 2.0029873847961426,
"step": 3946
},
{
"epoch": 0.8329772925073187,
"grad_norm": 0.10281118005514145,
"learning_rate": 0.00014236983842010775,
"loss": 2.0388238430023193,
"step": 3948
},
{
"epoch": 0.8333992668196324,
"grad_norm": 0.10240864753723145,
"learning_rate": 0.00014201077199281866,
"loss": 1.9518684148788452,
"step": 3950
},
{
"epoch": 0.8338212411319461,
"grad_norm": 0.13563206791877747,
"learning_rate": 0.00014165170556552963,
"loss": 2.045661687850952,
"step": 3952
},
{
"epoch": 0.8342432154442598,
"grad_norm": 0.11986995488405228,
"learning_rate": 0.00014129263913824057,
"loss": 1.9527664184570312,
"step": 3954
},
{
"epoch": 0.8346651897565736,
"grad_norm": 0.09651095420122147,
"learning_rate": 0.00014093357271095154,
"loss": 1.9852243661880493,
"step": 3956
},
{
"epoch": 0.8350871640688873,
"grad_norm": 0.10007069259881973,
"learning_rate": 0.00014057450628366248,
"loss": 2.0071771144866943,
"step": 3958
},
{
"epoch": 0.835509138381201,
"grad_norm": 0.10559628903865814,
"learning_rate": 0.00014021543985637345,
"loss": 1.97202730178833,
"step": 3960
},
{
"epoch": 0.8359311126935148,
"grad_norm": 0.10255084186792374,
"learning_rate": 0.0001398563734290844,
"loss": 1.9617817401885986,
"step": 3962
},
{
"epoch": 0.8363530870058286,
"grad_norm": 0.11901061236858368,
"learning_rate": 0.00013949730700179533,
"loss": 2.0128912925720215,
"step": 3964
},
{
"epoch": 0.8367750613181423,
"grad_norm": 0.09807386994361877,
"learning_rate": 0.00013913824057450627,
"loss": 1.9158368110656738,
"step": 3966
},
{
"epoch": 0.837197035630456,
"grad_norm": 0.10064463317394257,
"learning_rate": 0.00013877917414721724,
"loss": 2.0165905952453613,
"step": 3968
},
{
"epoch": 0.8376190099427697,
"grad_norm": 0.13565891981124878,
"learning_rate": 0.00013842010771992818,
"loss": 1.9716047048568726,
"step": 3970
},
{
"epoch": 0.8380409842550834,
"grad_norm": 0.12023929506540298,
"learning_rate": 0.00013806104129263915,
"loss": 1.9705606698989868,
"step": 3972
},
{
"epoch": 0.8384629585673972,
"grad_norm": 0.10239725559949875,
"learning_rate": 0.0001377019748653501,
"loss": 1.9629578590393066,
"step": 3974
},
{
"epoch": 0.838884932879711,
"grad_norm": 0.1014062762260437,
"learning_rate": 0.00013734290843806106,
"loss": 1.9195724725723267,
"step": 3976
},
{
"epoch": 0.8393069071920247,
"grad_norm": 0.11347910016775131,
"learning_rate": 0.00013698384201077198,
"loss": 1.998403549194336,
"step": 3978
},
{
"epoch": 0.8397288815043384,
"grad_norm": 0.11289577931165695,
"learning_rate": 0.00013662477558348295,
"loss": 1.9242366552352905,
"step": 3980
},
{
"epoch": 0.8401508558166522,
"grad_norm": 0.10800612717866898,
"learning_rate": 0.0001362657091561939,
"loss": 2.0245115756988525,
"step": 3982
},
{
"epoch": 0.8405728301289659,
"grad_norm": 0.0984434187412262,
"learning_rate": 0.00013590664272890486,
"loss": 1.9462246894836426,
"step": 3984
},
{
"epoch": 0.8409948044412796,
"grad_norm": 0.10353617370128632,
"learning_rate": 0.0001355475763016158,
"loss": 1.9554603099822998,
"step": 3986
},
{
"epoch": 0.8414167787535933,
"grad_norm": 0.10475564002990723,
"learning_rate": 0.00013518850987432677,
"loss": 1.9545459747314453,
"step": 3988
},
{
"epoch": 0.8418387530659072,
"grad_norm": 0.10349466651678085,
"learning_rate": 0.0001348294434470377,
"loss": 1.9966001510620117,
"step": 3990
},
{
"epoch": 0.8422607273782209,
"grad_norm": 0.09848958253860474,
"learning_rate": 0.00013447037701974865,
"loss": 2.006046772003174,
"step": 3992
},
{
"epoch": 0.8426827016905346,
"grad_norm": 0.11327902972698212,
"learning_rate": 0.0001341113105924596,
"loss": 2.0462801456451416,
"step": 3994
},
{
"epoch": 0.8431046760028483,
"grad_norm": 0.12294293195009232,
"learning_rate": 0.00013375224416517056,
"loss": 1.9671330451965332,
"step": 3996
},
{
"epoch": 0.843526650315162,
"grad_norm": 0.10709433257579803,
"learning_rate": 0.0001333931777378815,
"loss": 2.005004405975342,
"step": 3998
},
{
"epoch": 0.8439486246274758,
"grad_norm": 0.1040915697813034,
"learning_rate": 0.00013303411131059247,
"loss": 1.940953254699707,
"step": 4000
},
{
"epoch": 0.8443705989397895,
"grad_norm": 0.1030125617980957,
"learning_rate": 0.0001326750448833034,
"loss": 1.9784326553344727,
"step": 4002
},
{
"epoch": 0.8447925732521033,
"grad_norm": 0.1263287216424942,
"learning_rate": 0.00013231597845601438,
"loss": 2.0198214054107666,
"step": 4004
},
{
"epoch": 0.845214547564417,
"grad_norm": 0.11853793263435364,
"learning_rate": 0.00013195691202872532,
"loss": 2.007014274597168,
"step": 4006
},
{
"epoch": 0.8456365218767308,
"grad_norm": 0.10487543046474457,
"learning_rate": 0.00013159784560143626,
"loss": 2.001098394393921,
"step": 4008
},
{
"epoch": 0.8460584961890445,
"grad_norm": 0.13676878809928894,
"learning_rate": 0.0001312387791741472,
"loss": 2.000941038131714,
"step": 4010
},
{
"epoch": 0.8464804705013582,
"grad_norm": 0.10252530872821808,
"learning_rate": 0.00013087971274685817,
"loss": 1.9954538345336914,
"step": 4012
},
{
"epoch": 0.8469024448136719,
"grad_norm": 0.11510323733091354,
"learning_rate": 0.00013052064631956912,
"loss": 1.9998911619186401,
"step": 4014
},
{
"epoch": 0.8473244191259857,
"grad_norm": 0.10545968264341354,
"learning_rate": 0.00013016157989228009,
"loss": 1.9323265552520752,
"step": 4016
},
{
"epoch": 0.8477463934382995,
"grad_norm": 0.11139178276062012,
"learning_rate": 0.00012980251346499103,
"loss": 1.9420866966247559,
"step": 4018
},
{
"epoch": 0.8481683677506132,
"grad_norm": 0.11030557751655579,
"learning_rate": 0.00012944344703770197,
"loss": 1.950394868850708,
"step": 4020
},
{
"epoch": 0.8485903420629269,
"grad_norm": 0.11337947100400925,
"learning_rate": 0.00012908438061041294,
"loss": 1.9704276323318481,
"step": 4022
},
{
"epoch": 0.8490123163752407,
"grad_norm": 0.10561737418174744,
"learning_rate": 0.00012872531418312388,
"loss": 1.905476450920105,
"step": 4024
},
{
"epoch": 0.8494342906875544,
"grad_norm": 0.10871019959449768,
"learning_rate": 0.00012836624775583485,
"loss": 1.988250970840454,
"step": 4026
},
{
"epoch": 0.8498562649998681,
"grad_norm": 0.096436507999897,
"learning_rate": 0.0001280071813285458,
"loss": 1.9566881656646729,
"step": 4028
},
{
"epoch": 0.8502782393121818,
"grad_norm": 0.10243004560470581,
"learning_rate": 0.00012764811490125673,
"loss": 1.9743250608444214,
"step": 4030
},
{
"epoch": 0.8507002136244957,
"grad_norm": 0.10555164515972137,
"learning_rate": 0.0001272890484739677,
"loss": 1.977521300315857,
"step": 4032
},
{
"epoch": 0.8511221879368094,
"grad_norm": 0.10545128583908081,
"learning_rate": 0.00012692998204667864,
"loss": 1.8971844911575317,
"step": 4034
},
{
"epoch": 0.8515441622491231,
"grad_norm": 0.11702746897935867,
"learning_rate": 0.00012657091561938958,
"loss": 1.9810165166854858,
"step": 4036
},
{
"epoch": 0.8519661365614368,
"grad_norm": 0.10837048292160034,
"learning_rate": 0.00012621184919210055,
"loss": 2.0539441108703613,
"step": 4038
},
{
"epoch": 0.8523881108737505,
"grad_norm": 0.10196461528539658,
"learning_rate": 0.0001258527827648115,
"loss": 1.9302244186401367,
"step": 4040
},
{
"epoch": 0.8528100851860643,
"grad_norm": 0.10470853000879288,
"learning_rate": 0.00012549371633752246,
"loss": 2.014164686203003,
"step": 4042
},
{
"epoch": 0.853232059498378,
"grad_norm": 0.1329105645418167,
"learning_rate": 0.0001251346499102334,
"loss": 1.9430943727493286,
"step": 4044
},
{
"epoch": 0.8536540338106918,
"grad_norm": 0.11608070135116577,
"learning_rate": 0.00012477558348294437,
"loss": 1.979904294013977,
"step": 4046
},
{
"epoch": 0.8540760081230055,
"grad_norm": 0.10431086272001266,
"learning_rate": 0.0001244165170556553,
"loss": 2.0531728267669678,
"step": 4048
},
{
"epoch": 0.8544979824353193,
"grad_norm": 0.10427115857601166,
"learning_rate": 0.00012405745062836625,
"loss": 1.9721683263778687,
"step": 4050
},
{
"epoch": 0.854919956747633,
"grad_norm": 0.10791371762752533,
"learning_rate": 0.0001236983842010772,
"loss": 1.9694339036941528,
"step": 4052
},
{
"epoch": 0.8553419310599467,
"grad_norm": 0.10225558280944824,
"learning_rate": 0.00012333931777378817,
"loss": 1.958451509475708,
"step": 4054
},
{
"epoch": 0.8557639053722604,
"grad_norm": 0.10280132293701172,
"learning_rate": 0.0001229802513464991,
"loss": 1.9418867826461792,
"step": 4056
},
{
"epoch": 0.8561858796845742,
"grad_norm": 0.10905694216489792,
"learning_rate": 0.00012262118491921008,
"loss": 1.92428719997406,
"step": 4058
},
{
"epoch": 0.856607853996888,
"grad_norm": 0.11398683488368988,
"learning_rate": 0.00012226211849192102,
"loss": 2.0300872325897217,
"step": 4060
},
{
"epoch": 0.8570298283092017,
"grad_norm": 0.1170513853430748,
"learning_rate": 0.00012190305206463197,
"loss": 1.9878041744232178,
"step": 4062
},
{
"epoch": 0.8574518026215154,
"grad_norm": 0.1165793314576149,
"learning_rate": 0.0001215439856373429,
"loss": 1.9732407331466675,
"step": 4064
},
{
"epoch": 0.8578737769338292,
"grad_norm": 0.0937754362821579,
"learning_rate": 0.00012118491921005386,
"loss": 1.9634145498275757,
"step": 4066
},
{
"epoch": 0.8582957512461429,
"grad_norm": 0.1188102439045906,
"learning_rate": 0.00012082585278276481,
"loss": 1.9644101858139038,
"step": 4068
},
{
"epoch": 0.8587177255584566,
"grad_norm": 0.1256292462348938,
"learning_rate": 0.00012046678635547577,
"loss": 2.0683329105377197,
"step": 4070
},
{
"epoch": 0.8591396998707703,
"grad_norm": 0.11852488666772842,
"learning_rate": 0.00012010771992818672,
"loss": 2.035352945327759,
"step": 4072
},
{
"epoch": 0.8595616741830842,
"grad_norm": 0.10779717564582825,
"learning_rate": 0.00011974865350089768,
"loss": 1.9677391052246094,
"step": 4074
},
{
"epoch": 0.8599836484953979,
"grad_norm": 0.11177530884742737,
"learning_rate": 0.00011938958707360863,
"loss": 1.9674873352050781,
"step": 4076
},
{
"epoch": 0.8604056228077116,
"grad_norm": 0.12280061841011047,
"learning_rate": 0.00011903052064631956,
"loss": 2.027679443359375,
"step": 4078
},
{
"epoch": 0.8608275971200253,
"grad_norm": 0.1015930324792862,
"learning_rate": 0.00011867145421903051,
"loss": 1.9724328517913818,
"step": 4080
},
{
"epoch": 0.861249571432339,
"grad_norm": 0.11449313908815384,
"learning_rate": 0.00011831238779174147,
"loss": 1.953128695487976,
"step": 4082
},
{
"epoch": 0.8616715457446528,
"grad_norm": 0.10541561245918274,
"learning_rate": 0.00011795332136445242,
"loss": 1.9996706247329712,
"step": 4084
},
{
"epoch": 0.8620935200569665,
"grad_norm": 0.10451189428567886,
"learning_rate": 0.00011759425493716338,
"loss": 1.979792833328247,
"step": 4086
},
{
"epoch": 0.8625154943692803,
"grad_norm": 0.1051163598895073,
"learning_rate": 0.00011723518850987433,
"loss": 2.032721519470215,
"step": 4088
},
{
"epoch": 0.862937468681594,
"grad_norm": 0.10321579873561859,
"learning_rate": 0.00011687612208258529,
"loss": 2.037062883377075,
"step": 4090
},
{
"epoch": 0.8633594429939078,
"grad_norm": 0.09791623055934906,
"learning_rate": 0.00011651705565529623,
"loss": 1.9651076793670654,
"step": 4092
},
{
"epoch": 0.8637814173062215,
"grad_norm": 0.1162537932395935,
"learning_rate": 0.00011615798922800719,
"loss": 1.9559674263000488,
"step": 4094
},
{
"epoch": 0.8642033916185352,
"grad_norm": 0.10345719754695892,
"learning_rate": 0.00011579892280071814,
"loss": 2.002694606781006,
"step": 4096
},
{
"epoch": 0.8646253659308489,
"grad_norm": 0.10408767312765121,
"learning_rate": 0.00011543985637342908,
"loss": 1.9696884155273438,
"step": 4098
},
{
"epoch": 0.8650473402431627,
"grad_norm": 0.10374891012907028,
"learning_rate": 0.00011508078994614004,
"loss": 2.006002187728882,
"step": 4100
},
{
"epoch": 0.8654693145554765,
"grad_norm": 0.11908835172653198,
"learning_rate": 0.000114721723518851,
"loss": 1.9251043796539307,
"step": 4102
},
{
"epoch": 0.8658912888677902,
"grad_norm": 0.10296973586082458,
"learning_rate": 0.00011436265709156195,
"loss": 1.9771960973739624,
"step": 4104
},
{
"epoch": 0.8663132631801039,
"grad_norm": 0.10220454633235931,
"learning_rate": 0.00011400359066427289,
"loss": 1.9760771989822388,
"step": 4106
},
{
"epoch": 0.8667352374924177,
"grad_norm": 0.09764791280031204,
"learning_rate": 0.00011364452423698385,
"loss": 1.9355459213256836,
"step": 4108
},
{
"epoch": 0.8671572118047314,
"grad_norm": 0.10642227530479431,
"learning_rate": 0.0001132854578096948,
"loss": 2.0034024715423584,
"step": 4110
},
{
"epoch": 0.8675791861170451,
"grad_norm": 0.11727109551429749,
"learning_rate": 0.00011292639138240576,
"loss": 1.989803671836853,
"step": 4112
},
{
"epoch": 0.8680011604293588,
"grad_norm": 0.1131054013967514,
"learning_rate": 0.00011256732495511671,
"loss": 2.017568349838257,
"step": 4114
},
{
"epoch": 0.8684231347416727,
"grad_norm": 0.1025441512465477,
"learning_rate": 0.00011220825852782765,
"loss": 1.9782181978225708,
"step": 4116
},
{
"epoch": 0.8688451090539864,
"grad_norm": 0.10363882035017014,
"learning_rate": 0.00011184919210053861,
"loss": 1.9786701202392578,
"step": 4118
},
{
"epoch": 0.8692670833663001,
"grad_norm": 0.10100357979536057,
"learning_rate": 0.00011149012567324955,
"loss": 1.9932900667190552,
"step": 4120
},
{
"epoch": 0.8696890576786138,
"grad_norm": 0.09827663749456406,
"learning_rate": 0.0001111310592459605,
"loss": 1.9554555416107178,
"step": 4122
},
{
"epoch": 0.8701110319909275,
"grad_norm": 0.12574602663516998,
"learning_rate": 0.00011077199281867146,
"loss": 2.06023907661438,
"step": 4124
},
{
"epoch": 0.8705330063032413,
"grad_norm": 0.10325512290000916,
"learning_rate": 0.00011041292639138241,
"loss": 2.0091071128845215,
"step": 4126
},
{
"epoch": 0.870954980615555,
"grad_norm": 0.13387209177017212,
"learning_rate": 0.00011005385996409337,
"loss": 1.981414794921875,
"step": 4128
},
{
"epoch": 0.8713769549278688,
"grad_norm": 0.10491533577442169,
"learning_rate": 0.00010969479353680432,
"loss": 1.9464714527130127,
"step": 4130
},
{
"epoch": 0.8717989292401825,
"grad_norm": 0.10984890908002853,
"learning_rate": 0.00010933572710951528,
"loss": 1.8622952699661255,
"step": 4132
},
{
"epoch": 0.8722209035524963,
"grad_norm": 0.0975826159119606,
"learning_rate": 0.00010897666068222623,
"loss": 1.9768540859222412,
"step": 4134
},
{
"epoch": 0.87264287786481,
"grad_norm": 0.10530994832515717,
"learning_rate": 0.00010861759425493716,
"loss": 1.923965573310852,
"step": 4136
},
{
"epoch": 0.8730648521771237,
"grad_norm": 0.10517559200525284,
"learning_rate": 0.00010825852782764812,
"loss": 1.9396735429763794,
"step": 4138
},
{
"epoch": 0.8734868264894374,
"grad_norm": 0.11980796605348587,
"learning_rate": 0.00010789946140035907,
"loss": 1.9914312362670898,
"step": 4140
},
{
"epoch": 0.8739088008017512,
"grad_norm": 0.10633966326713562,
"learning_rate": 0.00010754039497307003,
"loss": 1.9428434371948242,
"step": 4142
},
{
"epoch": 0.874330775114065,
"grad_norm": 0.10705510526895523,
"learning_rate": 0.00010718132854578098,
"loss": 1.9225833415985107,
"step": 4144
},
{
"epoch": 0.8747527494263787,
"grad_norm": 0.12202207744121552,
"learning_rate": 0.00010682226211849194,
"loss": 1.9532277584075928,
"step": 4146
},
{
"epoch": 0.8751747237386924,
"grad_norm": 0.1029386818408966,
"learning_rate": 0.0001064631956912029,
"loss": 2.003518581390381,
"step": 4148
},
{
"epoch": 0.8755966980510062,
"grad_norm": 0.10091995447874069,
"learning_rate": 0.00010610412926391382,
"loss": 1.9390450716018677,
"step": 4150
},
{
"epoch": 0.8760186723633199,
"grad_norm": 0.10615351051092148,
"learning_rate": 0.00010574506283662478,
"loss": 2.0119433403015137,
"step": 4152
},
{
"epoch": 0.8764406466756336,
"grad_norm": 0.10447157174348831,
"learning_rate": 0.00010538599640933573,
"loss": 1.9368691444396973,
"step": 4154
},
{
"epoch": 0.8768626209879473,
"grad_norm": 0.10399413853883743,
"learning_rate": 0.00010502692998204669,
"loss": 1.9585068225860596,
"step": 4156
},
{
"epoch": 0.8772845953002611,
"grad_norm": 0.11146201193332672,
"learning_rate": 0.00010466786355475764,
"loss": 2.018137216567993,
"step": 4158
},
{
"epoch": 0.8777065696125749,
"grad_norm": 0.09950372576713562,
"learning_rate": 0.0001043087971274686,
"loss": 1.9181655645370483,
"step": 4160
},
{
"epoch": 0.8781285439248886,
"grad_norm": 0.11151466518640518,
"learning_rate": 0.00010394973070017955,
"loss": 2.0159413814544678,
"step": 4162
},
{
"epoch": 0.8785505182372023,
"grad_norm": 0.10773728042840958,
"learning_rate": 0.00010359066427289048,
"loss": 1.9867284297943115,
"step": 4164
},
{
"epoch": 0.878972492549516,
"grad_norm": 0.09941746294498444,
"learning_rate": 0.00010323159784560144,
"loss": 1.986660122871399,
"step": 4166
},
{
"epoch": 0.8793944668618298,
"grad_norm": 0.12539422512054443,
"learning_rate": 0.00010287253141831239,
"loss": 1.9660176038742065,
"step": 4168
},
{
"epoch": 0.8798164411741435,
"grad_norm": 0.154988095164299,
"learning_rate": 0.00010251346499102335,
"loss": 1.8545987606048584,
"step": 4170
},
{
"epoch": 0.8802384154864573,
"grad_norm": 0.10810447484254837,
"learning_rate": 0.0001021543985637343,
"loss": 1.9924931526184082,
"step": 4172
},
{
"epoch": 0.880660389798771,
"grad_norm": 0.11465612053871155,
"learning_rate": 0.00010179533213644526,
"loss": 1.9525902271270752,
"step": 4174
},
{
"epoch": 0.8810823641110848,
"grad_norm": 0.12277161329984665,
"learning_rate": 0.00010143626570915621,
"loss": 1.9586793184280396,
"step": 4176
},
{
"epoch": 0.8815043384233985,
"grad_norm": 0.10205686092376709,
"learning_rate": 0.00010107719928186714,
"loss": 1.9789749383926392,
"step": 4178
},
{
"epoch": 0.8819263127357122,
"grad_norm": 0.09653634577989578,
"learning_rate": 0.0001007181328545781,
"loss": 1.9354426860809326,
"step": 4180
},
{
"epoch": 0.8823482870480259,
"grad_norm": 0.1114269569516182,
"learning_rate": 0.00010035906642728905,
"loss": 1.9049197435379028,
"step": 4182
},
{
"epoch": 0.8827702613603396,
"grad_norm": 0.10992174595594406,
"learning_rate": 0.0001,
"loss": 1.987058401107788,
"step": 4184
},
{
"epoch": 0.8831922356726535,
"grad_norm": 0.10310900211334229,
"learning_rate": 9.964093357271096e-05,
"loss": 2.058398723602295,
"step": 4186
},
{
"epoch": 0.8836142099849672,
"grad_norm": 0.09827963262796402,
"learning_rate": 9.92818671454219e-05,
"loss": 2.018256664276123,
"step": 4188
},
{
"epoch": 0.8840361842972809,
"grad_norm": 0.10367213189601898,
"learning_rate": 9.892280071813286e-05,
"loss": 1.933269739151001,
"step": 4190
},
{
"epoch": 0.8844581586095946,
"grad_norm": 0.10718783736228943,
"learning_rate": 9.856373429084381e-05,
"loss": 1.9890934228897095,
"step": 4192
},
{
"epoch": 0.8848801329219084,
"grad_norm": 0.10876277089118958,
"learning_rate": 9.820466786355477e-05,
"loss": 2.008486747741699,
"step": 4194
},
{
"epoch": 0.8853021072342221,
"grad_norm": 0.11397836357355118,
"learning_rate": 9.784560143626571e-05,
"loss": 1.8652293682098389,
"step": 4196
},
{
"epoch": 0.8857240815465358,
"grad_norm": 0.09931036829948425,
"learning_rate": 9.748653500897666e-05,
"loss": 1.9758901596069336,
"step": 4198
},
{
"epoch": 0.8861460558588496,
"grad_norm": 0.1357850879430771,
"learning_rate": 9.712746858168762e-05,
"loss": 1.9757301807403564,
"step": 4200
},
{
"epoch": 0.8865680301711634,
"grad_norm": 0.10815994441509247,
"learning_rate": 9.676840215439856e-05,
"loss": 1.9674029350280762,
"step": 4202
},
{
"epoch": 0.8869900044834771,
"grad_norm": 0.10157132148742676,
"learning_rate": 9.640933572710952e-05,
"loss": 1.956359624862671,
"step": 4204
},
{
"epoch": 0.8874119787957908,
"grad_norm": 0.10470298677682877,
"learning_rate": 9.605026929982047e-05,
"loss": 1.964519739151001,
"step": 4206
},
{
"epoch": 0.8878339531081045,
"grad_norm": 0.1094106137752533,
"learning_rate": 9.569120287253143e-05,
"loss": 1.9606529474258423,
"step": 4208
},
{
"epoch": 0.8882559274204183,
"grad_norm": 0.09847893565893173,
"learning_rate": 9.533213644524237e-05,
"loss": 1.9122717380523682,
"step": 4210
},
{
"epoch": 0.888677901732732,
"grad_norm": 0.10363010317087173,
"learning_rate": 9.497307001795332e-05,
"loss": 1.9435197114944458,
"step": 4212
},
{
"epoch": 0.8890998760450458,
"grad_norm": 0.116825170814991,
"learning_rate": 9.461400359066428e-05,
"loss": 1.9593642950057983,
"step": 4214
},
{
"epoch": 0.8895218503573595,
"grad_norm": 0.11186099797487259,
"learning_rate": 9.425493716337523e-05,
"loss": 1.9593404531478882,
"step": 4216
},
{
"epoch": 0.8899438246696733,
"grad_norm": 0.0962444618344307,
"learning_rate": 9.389587073608617e-05,
"loss": 1.9644770622253418,
"step": 4218
},
{
"epoch": 0.890365798981987,
"grad_norm": 0.11035660654306412,
"learning_rate": 9.353680430879713e-05,
"loss": 2.001209259033203,
"step": 4220
},
{
"epoch": 0.8907877732943007,
"grad_norm": 0.14517180621623993,
"learning_rate": 9.317773788150808e-05,
"loss": 2.0199880599975586,
"step": 4222
},
{
"epoch": 0.8912097476066144,
"grad_norm": 0.11026109755039215,
"learning_rate": 9.281867145421903e-05,
"loss": 1.976463794708252,
"step": 4224
},
{
"epoch": 0.8916317219189281,
"grad_norm": 0.10343006253242493,
"learning_rate": 9.245960502692998e-05,
"loss": 1.9763832092285156,
"step": 4226
},
{
"epoch": 0.892053696231242,
"grad_norm": 0.10587551444768906,
"learning_rate": 9.210053859964094e-05,
"loss": 2.028472900390625,
"step": 4228
},
{
"epoch": 0.8924756705435557,
"grad_norm": 0.09772776812314987,
"learning_rate": 9.174147217235189e-05,
"loss": 1.958916425704956,
"step": 4230
},
{
"epoch": 0.8928976448558694,
"grad_norm": 0.10985676944255829,
"learning_rate": 9.138240574506283e-05,
"loss": 1.9664442539215088,
"step": 4232
},
{
"epoch": 0.8933196191681831,
"grad_norm": 0.10186974704265594,
"learning_rate": 9.102333931777379e-05,
"loss": 1.922724962234497,
"step": 4234
},
{
"epoch": 0.8937415934804969,
"grad_norm": 0.10821647942066193,
"learning_rate": 9.066427289048474e-05,
"loss": 1.9519538879394531,
"step": 4236
},
{
"epoch": 0.8941635677928106,
"grad_norm": 0.10424809902906418,
"learning_rate": 9.030520646319568e-05,
"loss": 2.0254831314086914,
"step": 4238
},
{
"epoch": 0.8945855421051243,
"grad_norm": 0.11156921088695526,
"learning_rate": 8.994614003590664e-05,
"loss": 1.9796853065490723,
"step": 4240
},
{
"epoch": 0.8950075164174381,
"grad_norm": 0.10358042269945145,
"learning_rate": 8.95870736086176e-05,
"loss": 1.9777542352676392,
"step": 4242
},
{
"epoch": 0.8954294907297519,
"grad_norm": 0.09888289868831635,
"learning_rate": 8.922800718132855e-05,
"loss": 1.9858468770980835,
"step": 4244
},
{
"epoch": 0.8958514650420656,
"grad_norm": 0.10647071897983551,
"learning_rate": 8.886894075403949e-05,
"loss": 2.0609874725341797,
"step": 4246
},
{
"epoch": 0.8962734393543793,
"grad_norm": 0.10038262605667114,
"learning_rate": 8.850987432675045e-05,
"loss": 1.943960189819336,
"step": 4248
},
{
"epoch": 0.896695413666693,
"grad_norm": 0.1254945993423462,
"learning_rate": 8.81508078994614e-05,
"loss": 1.9855732917785645,
"step": 4250
},
{
"epoch": 0.8971173879790068,
"grad_norm": 0.10052505135536194,
"learning_rate": 8.779174147217236e-05,
"loss": 1.9286285638809204,
"step": 4252
},
{
"epoch": 0.8975393622913205,
"grad_norm": 0.10480878502130508,
"learning_rate": 8.74326750448833e-05,
"loss": 1.932157278060913,
"step": 4254
},
{
"epoch": 0.8979613366036343,
"grad_norm": 0.11183223128318787,
"learning_rate": 8.707360861759425e-05,
"loss": 1.9971097707748413,
"step": 4256
},
{
"epoch": 0.898383310915948,
"grad_norm": 0.11327829211950302,
"learning_rate": 8.671454219030521e-05,
"loss": 1.9844368696212769,
"step": 4258
},
{
"epoch": 0.8988052852282618,
"grad_norm": 0.10075780749320984,
"learning_rate": 8.635547576301616e-05,
"loss": 1.9415298700332642,
"step": 4260
},
{
"epoch": 0.8992272595405755,
"grad_norm": 0.11096285283565521,
"learning_rate": 8.599640933572712e-05,
"loss": 2.013035774230957,
"step": 4262
},
{
"epoch": 0.8996492338528892,
"grad_norm": 0.11807211488485336,
"learning_rate": 8.563734290843806e-05,
"loss": 1.9533275365829468,
"step": 4264
},
{
"epoch": 0.9000712081652029,
"grad_norm": 0.11440528929233551,
"learning_rate": 8.527827648114902e-05,
"loss": 1.9751338958740234,
"step": 4266
},
{
"epoch": 0.9004931824775166,
"grad_norm": 0.10884717106819153,
"learning_rate": 8.491921005385997e-05,
"loss": 1.9218369722366333,
"step": 4268
},
{
"epoch": 0.9009151567898305,
"grad_norm": 0.10807453095912933,
"learning_rate": 8.456014362657093e-05,
"loss": 1.9406864643096924,
"step": 4270
},
{
"epoch": 0.9013371311021442,
"grad_norm": 0.10760314017534256,
"learning_rate": 8.420107719928187e-05,
"loss": 2.030202627182007,
"step": 4272
},
{
"epoch": 0.9017591054144579,
"grad_norm": 0.10374918580055237,
"learning_rate": 8.384201077199282e-05,
"loss": 1.944858193397522,
"step": 4274
},
{
"epoch": 0.9021810797267716,
"grad_norm": 0.11201688647270203,
"learning_rate": 8.348294434470378e-05,
"loss": 1.99154794216156,
"step": 4276
},
{
"epoch": 0.9026030540390854,
"grad_norm": 0.09922835230827332,
"learning_rate": 8.312387791741473e-05,
"loss": 1.9297547340393066,
"step": 4278
},
{
"epoch": 0.9030250283513991,
"grad_norm": 0.10309100151062012,
"learning_rate": 8.276481149012569e-05,
"loss": 1.9827690124511719,
"step": 4280
},
{
"epoch": 0.9034470026637128,
"grad_norm": 0.10575927048921585,
"learning_rate": 8.240574506283663e-05,
"loss": 2.0133469104766846,
"step": 4282
},
{
"epoch": 0.9038689769760266,
"grad_norm": 0.11572188884019852,
"learning_rate": 8.204667863554759e-05,
"loss": 1.942476749420166,
"step": 4284
},
{
"epoch": 0.9042909512883404,
"grad_norm": 0.10649754852056503,
"learning_rate": 8.168761220825854e-05,
"loss": 1.9683701992034912,
"step": 4286
},
{
"epoch": 0.9047129256006541,
"grad_norm": 0.11013828217983246,
"learning_rate": 8.132854578096948e-05,
"loss": 1.9389806985855103,
"step": 4288
},
{
"epoch": 0.9051348999129678,
"grad_norm": 0.09807085990905762,
"learning_rate": 8.096947935368044e-05,
"loss": 1.9025166034698486,
"step": 4290
},
{
"epoch": 0.9055568742252815,
"grad_norm": 0.10324486345052719,
"learning_rate": 8.061041292639139e-05,
"loss": 2.0151143074035645,
"step": 4292
},
{
"epoch": 0.9059788485375952,
"grad_norm": 0.11508278548717499,
"learning_rate": 8.025134649910235e-05,
"loss": 1.9715490341186523,
"step": 4294
},
{
"epoch": 0.906400822849909,
"grad_norm": 0.10184572637081146,
"learning_rate": 7.989228007181329e-05,
"loss": 1.9775052070617676,
"step": 4296
},
{
"epoch": 0.9068227971622228,
"grad_norm": 0.10934049636125565,
"learning_rate": 7.953321364452424e-05,
"loss": 1.9419037103652954,
"step": 4298
},
{
"epoch": 0.9072447714745365,
"grad_norm": 0.11035382747650146,
"learning_rate": 7.91741472172352e-05,
"loss": 2.0245814323425293,
"step": 4300
},
{
"epoch": 0.9076667457868502,
"grad_norm": 0.1122780591249466,
"learning_rate": 7.881508078994614e-05,
"loss": 1.9756752252578735,
"step": 4302
},
{
"epoch": 0.908088720099164,
"grad_norm": 0.10943111032247543,
"learning_rate": 7.84560143626571e-05,
"loss": 1.9726569652557373,
"step": 4304
},
{
"epoch": 0.9085106944114777,
"grad_norm": 0.11035355925559998,
"learning_rate": 7.809694793536805e-05,
"loss": 2.0262718200683594,
"step": 4306
},
{
"epoch": 0.9089326687237914,
"grad_norm": 0.10763981193304062,
"learning_rate": 7.7737881508079e-05,
"loss": 1.9832899570465088,
"step": 4308
},
{
"epoch": 0.9093546430361051,
"grad_norm": 0.10239090770483017,
"learning_rate": 7.737881508078995e-05,
"loss": 1.9645538330078125,
"step": 4310
},
{
"epoch": 0.909776617348419,
"grad_norm": 0.109678253531456,
"learning_rate": 7.70197486535009e-05,
"loss": 1.9875560998916626,
"step": 4312
},
{
"epoch": 0.9101985916607327,
"grad_norm": 0.12480778247117996,
"learning_rate": 7.666068222621186e-05,
"loss": 1.9042906761169434,
"step": 4314
},
{
"epoch": 0.9106205659730464,
"grad_norm": 0.10408678650856018,
"learning_rate": 7.630161579892281e-05,
"loss": 2.029926300048828,
"step": 4316
},
{
"epoch": 0.9110425402853601,
"grad_norm": 0.1103670671582222,
"learning_rate": 7.594254937163375e-05,
"loss": 1.989965558052063,
"step": 4318
},
{
"epoch": 0.9114645145976739,
"grad_norm": 0.10089827328920364,
"learning_rate": 7.558348294434471e-05,
"loss": 1.9067894220352173,
"step": 4320
},
{
"epoch": 0.9118864889099876,
"grad_norm": 0.10455714166164398,
"learning_rate": 7.522441651705567e-05,
"loss": 1.9374706745147705,
"step": 4322
},
{
"epoch": 0.9123084632223013,
"grad_norm": 0.12277254462242126,
"learning_rate": 7.48653500897666e-05,
"loss": 1.9924230575561523,
"step": 4324
},
{
"epoch": 0.9127304375346151,
"grad_norm": 0.10296517610549927,
"learning_rate": 7.450628366247756e-05,
"loss": 1.976456880569458,
"step": 4326
},
{
"epoch": 0.9131524118469289,
"grad_norm": 0.09838514775037766,
"learning_rate": 7.414721723518852e-05,
"loss": 1.8516685962677002,
"step": 4328
},
{
"epoch": 0.9135743861592426,
"grad_norm": 0.10393622517585754,
"learning_rate": 7.378815080789947e-05,
"loss": 1.9571295976638794,
"step": 4330
},
{
"epoch": 0.9139963604715563,
"grad_norm": 0.10954293608665466,
"learning_rate": 7.342908438061041e-05,
"loss": 1.977062463760376,
"step": 4332
},
{
"epoch": 0.91441833478387,
"grad_norm": 0.11800585687160492,
"learning_rate": 7.307001795332137e-05,
"loss": 1.9611639976501465,
"step": 4334
},
{
"epoch": 0.9148403090961837,
"grad_norm": 0.11380811035633087,
"learning_rate": 7.271095152603232e-05,
"loss": 1.9709993600845337,
"step": 4336
},
{
"epoch": 0.9152622834084975,
"grad_norm": 0.1055958941578865,
"learning_rate": 7.235188509874327e-05,
"loss": 2.0279035568237305,
"step": 4338
},
{
"epoch": 0.9156842577208113,
"grad_norm": 0.10484731197357178,
"learning_rate": 7.199281867145422e-05,
"loss": 2.00527286529541,
"step": 4340
},
{
"epoch": 0.916106232033125,
"grad_norm": 0.10349317640066147,
"learning_rate": 7.163375224416518e-05,
"loss": 1.9561049938201904,
"step": 4342
},
{
"epoch": 0.9165282063454387,
"grad_norm": 0.10680350661277771,
"learning_rate": 7.127468581687613e-05,
"loss": 2.0026931762695312,
"step": 4344
},
{
"epoch": 0.9169501806577525,
"grad_norm": 0.12091751396656036,
"learning_rate": 7.091561938958707e-05,
"loss": 1.9692943096160889,
"step": 4346
},
{
"epoch": 0.9173721549700662,
"grad_norm": 0.12316521257162094,
"learning_rate": 7.055655296229803e-05,
"loss": 1.918934941291809,
"step": 4348
},
{
"epoch": 0.9177941292823799,
"grad_norm": 0.10428337007761002,
"learning_rate": 7.019748653500898e-05,
"loss": 2.015747547149658,
"step": 4350
},
{
"epoch": 0.9182161035946936,
"grad_norm": 0.11164549738168716,
"learning_rate": 6.983842010771992e-05,
"loss": 2.01839017868042,
"step": 4352
},
{
"epoch": 0.9186380779070074,
"grad_norm": 0.10518519580364227,
"learning_rate": 6.947935368043088e-05,
"loss": 1.9499198198318481,
"step": 4354
},
{
"epoch": 0.9190600522193212,
"grad_norm": 0.10779891163110733,
"learning_rate": 6.912028725314183e-05,
"loss": 1.9888851642608643,
"step": 4356
},
{
"epoch": 0.9194820265316349,
"grad_norm": 0.10748085379600525,
"learning_rate": 6.876122082585279e-05,
"loss": 2.006067991256714,
"step": 4358
},
{
"epoch": 0.9199040008439486,
"grad_norm": 0.11628947407007217,
"learning_rate": 6.840215439856373e-05,
"loss": 1.9224172830581665,
"step": 4360
},
{
"epoch": 0.9203259751562624,
"grad_norm": 0.10497194528579712,
"learning_rate": 6.804308797127469e-05,
"loss": 1.9339981079101562,
"step": 4362
},
{
"epoch": 0.9207479494685761,
"grad_norm": 0.1023145467042923,
"learning_rate": 6.768402154398564e-05,
"loss": 1.9809895753860474,
"step": 4364
},
{
"epoch": 0.9211699237808898,
"grad_norm": 0.1136220321059227,
"learning_rate": 6.73249551166966e-05,
"loss": 2.0420050621032715,
"step": 4366
},
{
"epoch": 0.9215918980932035,
"grad_norm": 0.09812942147254944,
"learning_rate": 6.696588868940754e-05,
"loss": 1.9726718664169312,
"step": 4368
},
{
"epoch": 0.9220138724055174,
"grad_norm": 0.11689987778663635,
"learning_rate": 6.66068222621185e-05,
"loss": 2.007542371749878,
"step": 4370
},
{
"epoch": 0.9224358467178311,
"grad_norm": 0.1088201105594635,
"learning_rate": 6.624775583482945e-05,
"loss": 1.9895540475845337,
"step": 4372
},
{
"epoch": 0.9228578210301448,
"grad_norm": 0.11217056214809418,
"learning_rate": 6.588868940754039e-05,
"loss": 1.9780479669570923,
"step": 4374
},
{
"epoch": 0.9232797953424585,
"grad_norm": 0.11308541148900986,
"learning_rate": 6.552962298025135e-05,
"loss": 1.9641718864440918,
"step": 4376
},
{
"epoch": 0.9237017696547722,
"grad_norm": 0.10051889717578888,
"learning_rate": 6.51705565529623e-05,
"loss": 1.9247889518737793,
"step": 4378
},
{
"epoch": 0.924123743967086,
"grad_norm": 0.11950184404850006,
"learning_rate": 6.481149012567326e-05,
"loss": 1.9736049175262451,
"step": 4380
},
{
"epoch": 0.9245457182793997,
"grad_norm": 0.09528274834156036,
"learning_rate": 6.44524236983842e-05,
"loss": 1.9767588376998901,
"step": 4382
},
{
"epoch": 0.9249676925917135,
"grad_norm": 0.11856667697429657,
"learning_rate": 6.409335727109515e-05,
"loss": 1.9190655946731567,
"step": 4384
},
{
"epoch": 0.9253896669040272,
"grad_norm": 0.09872009605169296,
"learning_rate": 6.373429084380611e-05,
"loss": 2.026851177215576,
"step": 4386
},
{
"epoch": 0.925811641216341,
"grad_norm": 0.11413022875785828,
"learning_rate": 6.337522441651705e-05,
"loss": 1.9873367547988892,
"step": 4388
},
{
"epoch": 0.9262336155286547,
"grad_norm": 0.1007830798625946,
"learning_rate": 6.3016157989228e-05,
"loss": 1.9649474620819092,
"step": 4390
},
{
"epoch": 0.9266555898409684,
"grad_norm": 0.11300705373287201,
"learning_rate": 6.265709156193896e-05,
"loss": 1.9788988828659058,
"step": 4392
},
{
"epoch": 0.9270775641532821,
"grad_norm": 0.10901802778244019,
"learning_rate": 6.229802513464991e-05,
"loss": 1.8743146657943726,
"step": 4394
},
{
"epoch": 0.9274995384655959,
"grad_norm": 0.10092721879482269,
"learning_rate": 6.193895870736086e-05,
"loss": 1.9390665292739868,
"step": 4396
},
{
"epoch": 0.9279215127779097,
"grad_norm": 0.12116295844316483,
"learning_rate": 6.157989228007181e-05,
"loss": 1.8761729001998901,
"step": 4398
},
{
"epoch": 0.9283434870902234,
"grad_norm": 0.11722904443740845,
"learning_rate": 6.122082585278277e-05,
"loss": 1.9988369941711426,
"step": 4400
},
{
"epoch": 0.9287654614025371,
"grad_norm": 0.11201320588588715,
"learning_rate": 6.0861759425493715e-05,
"loss": 1.963629961013794,
"step": 4402
},
{
"epoch": 0.9291874357148509,
"grad_norm": 0.1068640798330307,
"learning_rate": 6.050269299820467e-05,
"loss": 1.992313027381897,
"step": 4404
},
{
"epoch": 0.9296094100271646,
"grad_norm": 0.09950323402881622,
"learning_rate": 6.0143626570915625e-05,
"loss": 1.986338496208191,
"step": 4406
},
{
"epoch": 0.9300313843394783,
"grad_norm": 0.10620663315057755,
"learning_rate": 5.978456014362658e-05,
"loss": 1.8992366790771484,
"step": 4408
},
{
"epoch": 0.930453358651792,
"grad_norm": 0.10992123186588287,
"learning_rate": 5.942549371633752e-05,
"loss": 1.971193552017212,
"step": 4410
},
{
"epoch": 0.9308753329641059,
"grad_norm": 0.11310093104839325,
"learning_rate": 5.906642728904848e-05,
"loss": 1.9978446960449219,
"step": 4412
},
{
"epoch": 0.9312973072764196,
"grad_norm": 0.10823677480220795,
"learning_rate": 5.870736086175943e-05,
"loss": 1.9941303730010986,
"step": 4414
},
{
"epoch": 0.9317192815887333,
"grad_norm": 0.10390577465295792,
"learning_rate": 5.834829443447039e-05,
"loss": 1.983245611190796,
"step": 4416
},
{
"epoch": 0.932141255901047,
"grad_norm": 0.1217908039689064,
"learning_rate": 5.798922800718133e-05,
"loss": 2.067887544631958,
"step": 4418
},
{
"epoch": 0.9325632302133607,
"grad_norm": 0.1062224730849266,
"learning_rate": 5.7630161579892284e-05,
"loss": 1.9372761249542236,
"step": 4420
},
{
"epoch": 0.9329852045256745,
"grad_norm": 0.11350486427545547,
"learning_rate": 5.727109515260324e-05,
"loss": 1.9175602197647095,
"step": 4422
},
{
"epoch": 0.9334071788379882,
"grad_norm": 0.11135926097631454,
"learning_rate": 5.691202872531418e-05,
"loss": 2.0014498233795166,
"step": 4424
},
{
"epoch": 0.933829153150302,
"grad_norm": 0.10471297800540924,
"learning_rate": 5.6552962298025136e-05,
"loss": 1.9817594289779663,
"step": 4426
},
{
"epoch": 0.9342511274626157,
"grad_norm": 0.10763347893953323,
"learning_rate": 5.619389587073609e-05,
"loss": 1.969205379486084,
"step": 4428
},
{
"epoch": 0.9346731017749295,
"grad_norm": 0.10730740427970886,
"learning_rate": 5.5834829443447046e-05,
"loss": 1.9536798000335693,
"step": 4430
},
{
"epoch": 0.9350950760872432,
"grad_norm": 0.11071805655956268,
"learning_rate": 5.547576301615799e-05,
"loss": 1.995632290840149,
"step": 4432
},
{
"epoch": 0.9355170503995569,
"grad_norm": 0.1172584667801857,
"learning_rate": 5.511669658886894e-05,
"loss": 1.9432746171951294,
"step": 4434
},
{
"epoch": 0.9359390247118706,
"grad_norm": 0.10111786425113678,
"learning_rate": 5.47576301615799e-05,
"loss": 1.9635210037231445,
"step": 4436
},
{
"epoch": 0.9363609990241843,
"grad_norm": 0.1115003377199173,
"learning_rate": 5.439856373429084e-05,
"loss": 1.9342033863067627,
"step": 4438
},
{
"epoch": 0.9367829733364982,
"grad_norm": 0.11682593077421188,
"learning_rate": 5.4039497307001794e-05,
"loss": 1.9311261177062988,
"step": 4440
},
{
"epoch": 0.9372049476488119,
"grad_norm": 0.10311296582221985,
"learning_rate": 5.368043087971275e-05,
"loss": 1.9469839334487915,
"step": 4442
},
{
"epoch": 0.9376269219611256,
"grad_norm": 0.105555459856987,
"learning_rate": 5.3321364452423705e-05,
"loss": 1.9866645336151123,
"step": 4444
},
{
"epoch": 0.9380488962734393,
"grad_norm": 0.10601339489221573,
"learning_rate": 5.2962298025134646e-05,
"loss": 2.021909475326538,
"step": 4446
},
{
"epoch": 0.9384708705857531,
"grad_norm": 0.10167314112186432,
"learning_rate": 5.26032315978456e-05,
"loss": 1.940483570098877,
"step": 4448
},
{
"epoch": 0.9388928448980668,
"grad_norm": 0.11554458737373352,
"learning_rate": 5.2244165170556557e-05,
"loss": 2.0332086086273193,
"step": 4450
},
{
"epoch": 0.9393148192103805,
"grad_norm": 0.10473091155290604,
"learning_rate": 5.1885098743267505e-05,
"loss": 1.9359129667282104,
"step": 4452
},
{
"epoch": 0.9397367935226943,
"grad_norm": 0.10217129439115524,
"learning_rate": 5.152603231597845e-05,
"loss": 2.0070266723632812,
"step": 4454
},
{
"epoch": 0.9401587678350081,
"grad_norm": 0.1090213879942894,
"learning_rate": 5.116696588868941e-05,
"loss": 1.93622624874115,
"step": 4456
},
{
"epoch": 0.9405807421473218,
"grad_norm": 0.10054320096969604,
"learning_rate": 5.0807899461400364e-05,
"loss": 1.9456861019134521,
"step": 4458
},
{
"epoch": 0.9410027164596355,
"grad_norm": 0.1286977082490921,
"learning_rate": 5.044883303411131e-05,
"loss": 1.9709253311157227,
"step": 4460
},
{
"epoch": 0.9414246907719492,
"grad_norm": 0.11071281880140305,
"learning_rate": 5.008976660682227e-05,
"loss": 1.9578603506088257,
"step": 4462
},
{
"epoch": 0.941846665084263,
"grad_norm": 0.10554347932338715,
"learning_rate": 4.9730700179533215e-05,
"loss": 1.9619337320327759,
"step": 4464
},
{
"epoch": 0.9422686393965767,
"grad_norm": 0.1077689677476883,
"learning_rate": 4.937163375224417e-05,
"loss": 1.9693071842193604,
"step": 4466
},
{
"epoch": 0.9426906137088905,
"grad_norm": 0.10029803216457367,
"learning_rate": 4.901256732495512e-05,
"loss": 1.9795218706130981,
"step": 4468
},
{
"epoch": 0.9431125880212042,
"grad_norm": 0.09988780319690704,
"learning_rate": 4.8653500897666074e-05,
"loss": 1.8700910806655884,
"step": 4470
},
{
"epoch": 0.943534562333518,
"grad_norm": 0.11808252334594727,
"learning_rate": 4.829443447037702e-05,
"loss": 1.9545581340789795,
"step": 4472
},
{
"epoch": 0.9439565366458317,
"grad_norm": 0.10970522463321686,
"learning_rate": 4.793536804308798e-05,
"loss": 1.9331252574920654,
"step": 4474
},
{
"epoch": 0.9443785109581454,
"grad_norm": 0.11361062526702881,
"learning_rate": 4.7576301615798926e-05,
"loss": 1.9802976846694946,
"step": 4476
},
{
"epoch": 0.9448004852704591,
"grad_norm": 0.10900110751390457,
"learning_rate": 4.721723518850988e-05,
"loss": 1.9315359592437744,
"step": 4478
},
{
"epoch": 0.9452224595827728,
"grad_norm": 0.0988907590508461,
"learning_rate": 4.685816876122083e-05,
"loss": 1.9867444038391113,
"step": 4480
},
{
"epoch": 0.9456444338950867,
"grad_norm": 0.10760052502155304,
"learning_rate": 4.6499102333931784e-05,
"loss": 1.968553900718689,
"step": 4482
},
{
"epoch": 0.9460664082074004,
"grad_norm": 0.11219218373298645,
"learning_rate": 4.614003590664273e-05,
"loss": 1.9957237243652344,
"step": 4484
},
{
"epoch": 0.9464883825197141,
"grad_norm": 0.1125212013721466,
"learning_rate": 4.578096947935368e-05,
"loss": 1.9893512725830078,
"step": 4486
},
{
"epoch": 0.9469103568320278,
"grad_norm": 0.10519126057624817,
"learning_rate": 4.5421903052064636e-05,
"loss": 1.9717342853546143,
"step": 4488
},
{
"epoch": 0.9473323311443416,
"grad_norm": 0.10098539292812347,
"learning_rate": 4.5062836624775585e-05,
"loss": 1.9890072345733643,
"step": 4490
},
{
"epoch": 0.9477543054566553,
"grad_norm": 0.10550697147846222,
"learning_rate": 4.470377019748654e-05,
"loss": 2.011085033416748,
"step": 4492
},
{
"epoch": 0.948176279768969,
"grad_norm": 0.11036745458841324,
"learning_rate": 4.434470377019749e-05,
"loss": 1.934899926185608,
"step": 4494
},
{
"epoch": 0.9485982540812828,
"grad_norm": 0.09455365687608719,
"learning_rate": 4.398563734290844e-05,
"loss": 2.016329050064087,
"step": 4496
},
{
"epoch": 0.9490202283935966,
"grad_norm": 0.15500092506408691,
"learning_rate": 4.362657091561939e-05,
"loss": 1.994697093963623,
"step": 4498
},
{
"epoch": 0.9494422027059103,
"grad_norm": 0.106009341776371,
"learning_rate": 4.326750448833034e-05,
"loss": 1.9954835176467896,
"step": 4500
},
{
"epoch": 0.949864177018224,
"grad_norm": 0.10341881960630417,
"learning_rate": 4.2908438061041295e-05,
"loss": 1.8403732776641846,
"step": 4502
},
{
"epoch": 0.9502861513305377,
"grad_norm": 0.10378289222717285,
"learning_rate": 4.2549371633752244e-05,
"loss": 2.003725528717041,
"step": 4504
},
{
"epoch": 0.9507081256428515,
"grad_norm": 0.11462432146072388,
"learning_rate": 4.21903052064632e-05,
"loss": 1.9470783472061157,
"step": 4506
},
{
"epoch": 0.9511300999551652,
"grad_norm": 0.09759571403265,
"learning_rate": 4.183123877917415e-05,
"loss": 1.948257565498352,
"step": 4508
},
{
"epoch": 0.951552074267479,
"grad_norm": 0.1090717762708664,
"learning_rate": 4.14721723518851e-05,
"loss": 1.9400941133499146,
"step": 4510
},
{
"epoch": 0.9519740485797927,
"grad_norm": 0.10372298210859299,
"learning_rate": 4.111310592459605e-05,
"loss": 1.8825671672821045,
"step": 4512
},
{
"epoch": 0.9523960228921065,
"grad_norm": 0.10764601826667786,
"learning_rate": 4.0754039497307006e-05,
"loss": 1.8693889379501343,
"step": 4514
},
{
"epoch": 0.9528179972044202,
"grad_norm": 0.10617853701114655,
"learning_rate": 4.0394973070017954e-05,
"loss": 1.9854592084884644,
"step": 4516
},
{
"epoch": 0.9532399715167339,
"grad_norm": 0.10256534069776535,
"learning_rate": 4.00359066427289e-05,
"loss": 1.898912787437439,
"step": 4518
},
{
"epoch": 0.9536619458290476,
"grad_norm": 0.10808218270540237,
"learning_rate": 3.967684021543986e-05,
"loss": 1.93253493309021,
"step": 4520
},
{
"epoch": 0.9540839201413613,
"grad_norm": 0.11648982018232346,
"learning_rate": 3.9317773788150806e-05,
"loss": 1.9380801916122437,
"step": 4522
},
{
"epoch": 0.9545058944536752,
"grad_norm": 0.09977971762418747,
"learning_rate": 3.895870736086176e-05,
"loss": 1.9927090406417847,
"step": 4524
},
{
"epoch": 0.9549278687659889,
"grad_norm": 0.10238093137741089,
"learning_rate": 3.859964093357271e-05,
"loss": 1.9616984128952026,
"step": 4526
},
{
"epoch": 0.9553498430783026,
"grad_norm": 0.10674754530191422,
"learning_rate": 3.8240574506283664e-05,
"loss": 1.979896903038025,
"step": 4528
},
{
"epoch": 0.9557718173906163,
"grad_norm": 0.10306009650230408,
"learning_rate": 3.788150807899461e-05,
"loss": 1.8982036113739014,
"step": 4530
},
{
"epoch": 0.9561937917029301,
"grad_norm": 0.11265920847654343,
"learning_rate": 3.752244165170557e-05,
"loss": 1.9744625091552734,
"step": 4532
},
{
"epoch": 0.9566157660152438,
"grad_norm": 0.11647281795740128,
"learning_rate": 3.7163375224416516e-05,
"loss": 1.9784642457962036,
"step": 4534
},
{
"epoch": 0.9570377403275575,
"grad_norm": 0.11191514134407043,
"learning_rate": 3.6804308797127465e-05,
"loss": 1.9690855741500854,
"step": 4536
},
{
"epoch": 0.9574597146398713,
"grad_norm": 0.1162935197353363,
"learning_rate": 3.644524236983842e-05,
"loss": 1.9823527336120605,
"step": 4538
},
{
"epoch": 0.9578816889521851,
"grad_norm": 0.10934937745332718,
"learning_rate": 3.6086175942549375e-05,
"loss": 1.9947012662887573,
"step": 4540
},
{
"epoch": 0.9583036632644988,
"grad_norm": 0.10031798481941223,
"learning_rate": 3.572710951526032e-05,
"loss": 1.9506831169128418,
"step": 4542
},
{
"epoch": 0.9587256375768125,
"grad_norm": 0.13115912675857544,
"learning_rate": 3.536804308797128e-05,
"loss": 1.993109107017517,
"step": 4544
},
{
"epoch": 0.9591476118891262,
"grad_norm": 0.11059731245040894,
"learning_rate": 3.500897666068223e-05,
"loss": 1.9159808158874512,
"step": 4546
},
{
"epoch": 0.95956958620144,
"grad_norm": 0.10525340586900711,
"learning_rate": 3.464991023339318e-05,
"loss": 1.9480326175689697,
"step": 4548
},
{
"epoch": 0.9599915605137537,
"grad_norm": 0.11978833377361298,
"learning_rate": 3.429084380610413e-05,
"loss": 1.9725160598754883,
"step": 4550
},
{
"epoch": 0.9604135348260675,
"grad_norm": 0.10969995707273483,
"learning_rate": 3.3931777378815085e-05,
"loss": 2.004603147506714,
"step": 4552
},
{
"epoch": 0.9608355091383812,
"grad_norm": 0.10642252117395401,
"learning_rate": 3.3572710951526034e-05,
"loss": 2.0019936561584473,
"step": 4554
},
{
"epoch": 0.961257483450695,
"grad_norm": 0.11324192583560944,
"learning_rate": 3.321364452423699e-05,
"loss": 1.9790276288986206,
"step": 4556
},
{
"epoch": 0.9616794577630087,
"grad_norm": 0.10698672384023666,
"learning_rate": 3.285457809694794e-05,
"loss": 2.0339720249176025,
"step": 4558
},
{
"epoch": 0.9621014320753224,
"grad_norm": 0.12763382494449615,
"learning_rate": 3.249551166965889e-05,
"loss": 1.988168716430664,
"step": 4560
},
{
"epoch": 0.9625234063876361,
"grad_norm": 0.11609788239002228,
"learning_rate": 3.213644524236984e-05,
"loss": 2.009946346282959,
"step": 4562
},
{
"epoch": 0.9629453806999498,
"grad_norm": 0.10954302549362183,
"learning_rate": 3.1777378815080796e-05,
"loss": 1.935221791267395,
"step": 4564
},
{
"epoch": 0.9633673550122637,
"grad_norm": 0.09643895924091339,
"learning_rate": 3.1418312387791744e-05,
"loss": 2.0139708518981934,
"step": 4566
},
{
"epoch": 0.9637893293245774,
"grad_norm": 0.10529980808496475,
"learning_rate": 3.105924596050269e-05,
"loss": 1.9843734502792358,
"step": 4568
},
{
"epoch": 0.9642113036368911,
"grad_norm": 0.10969899594783783,
"learning_rate": 3.070017953321365e-05,
"loss": 2.025578737258911,
"step": 4570
},
{
"epoch": 0.9646332779492048,
"grad_norm": 0.10554074496030807,
"learning_rate": 3.0341113105924596e-05,
"loss": 1.946641445159912,
"step": 4572
},
{
"epoch": 0.9650552522615186,
"grad_norm": 0.11176867038011551,
"learning_rate": 2.998204667863555e-05,
"loss": 1.9374874830245972,
"step": 4574
},
{
"epoch": 0.9654772265738323,
"grad_norm": 0.11251261830329895,
"learning_rate": 2.96229802513465e-05,
"loss": 1.8668239116668701,
"step": 4576
},
{
"epoch": 0.965899200886146,
"grad_norm": 0.1133284792304039,
"learning_rate": 2.9263913824057455e-05,
"loss": 1.98781156539917,
"step": 4578
},
{
"epoch": 0.9663211751984598,
"grad_norm": 0.10248855501413345,
"learning_rate": 2.8904847396768403e-05,
"loss": 2.0073838233947754,
"step": 4580
},
{
"epoch": 0.9667431495107736,
"grad_norm": 0.12220220267772675,
"learning_rate": 2.8545780969479358e-05,
"loss": 2.0011730194091797,
"step": 4582
},
{
"epoch": 0.9671651238230873,
"grad_norm": 0.10336542874574661,
"learning_rate": 2.8186714542190307e-05,
"loss": 1.9912590980529785,
"step": 4584
},
{
"epoch": 0.967587098135401,
"grad_norm": 0.10533387213945389,
"learning_rate": 2.7827648114901255e-05,
"loss": 1.966177225112915,
"step": 4586
},
{
"epoch": 0.9680090724477147,
"grad_norm": 0.10882163792848587,
"learning_rate": 2.746858168761221e-05,
"loss": 1.9780558347702026,
"step": 4588
},
{
"epoch": 0.9684310467600284,
"grad_norm": 0.10504578799009323,
"learning_rate": 2.710951526032316e-05,
"loss": 1.9288051128387451,
"step": 4590
},
{
"epoch": 0.9688530210723422,
"grad_norm": 0.10803225636482239,
"learning_rate": 2.6750448833034114e-05,
"loss": 1.9908766746520996,
"step": 4592
},
{
"epoch": 0.969274995384656,
"grad_norm": 0.11677280813455582,
"learning_rate": 2.6391382405745062e-05,
"loss": 2.025902032852173,
"step": 4594
},
{
"epoch": 0.9696969696969697,
"grad_norm": 0.110410176217556,
"learning_rate": 2.6032315978456017e-05,
"loss": 2.0003957748413086,
"step": 4596
},
{
"epoch": 0.9701189440092834,
"grad_norm": 0.10641536116600037,
"learning_rate": 2.5673249551166965e-05,
"loss": 1.9546582698822021,
"step": 4598
},
{
"epoch": 0.9705409183215972,
"grad_norm": 0.1778954565525055,
"learning_rate": 2.5314183123877917e-05,
"loss": 1.9980928897857666,
"step": 4600
},
{
"epoch": 0.9709628926339109,
"grad_norm": 0.11043521016836166,
"learning_rate": 2.495511669658887e-05,
"loss": 2.0226240158081055,
"step": 4602
},
{
"epoch": 0.9713848669462246,
"grad_norm": 0.12288088351488113,
"learning_rate": 2.4596050269299824e-05,
"loss": 1.9354668855667114,
"step": 4604
},
{
"epoch": 0.9718068412585383,
"grad_norm": 0.11611411720514297,
"learning_rate": 2.4236983842010776e-05,
"loss": 1.9873931407928467,
"step": 4606
},
{
"epoch": 0.9722288155708522,
"grad_norm": 0.10628348588943481,
"learning_rate": 2.3877917414721724e-05,
"loss": 1.9681874513626099,
"step": 4608
},
{
"epoch": 0.9726507898831659,
"grad_norm": 0.10231490433216095,
"learning_rate": 2.3518850987432676e-05,
"loss": 2.0011558532714844,
"step": 4610
},
{
"epoch": 0.9730727641954796,
"grad_norm": 0.11022770404815674,
"learning_rate": 2.3159784560143628e-05,
"loss": 1.9592645168304443,
"step": 4612
},
{
"epoch": 0.9734947385077933,
"grad_norm": 0.10896950215101242,
"learning_rate": 2.280071813285458e-05,
"loss": 1.9518122673034668,
"step": 4614
},
{
"epoch": 0.9739167128201071,
"grad_norm": 0.10292213410139084,
"learning_rate": 2.244165170556553e-05,
"loss": 2.0716841220855713,
"step": 4616
},
{
"epoch": 0.9743386871324208,
"grad_norm": 0.10210495442152023,
"learning_rate": 2.2082585278276483e-05,
"loss": 1.9209129810333252,
"step": 4618
},
{
"epoch": 0.9747606614447345,
"grad_norm": 0.10981930792331696,
"learning_rate": 2.1723518850987435e-05,
"loss": 2.009765148162842,
"step": 4620
},
{
"epoch": 0.9751826357570483,
"grad_norm": 0.13394322991371155,
"learning_rate": 2.1364452423698386e-05,
"loss": 1.9774208068847656,
"step": 4622
},
{
"epoch": 0.975604610069362,
"grad_norm": 0.10786944627761841,
"learning_rate": 2.1005385996409338e-05,
"loss": 1.996282696723938,
"step": 4624
},
{
"epoch": 0.9760265843816758,
"grad_norm": 0.09994073212146759,
"learning_rate": 2.0646319569120286e-05,
"loss": 1.9573900699615479,
"step": 4626
},
{
"epoch": 0.9764485586939895,
"grad_norm": 0.10520496219396591,
"learning_rate": 2.0287253141831238e-05,
"loss": 1.9692916870117188,
"step": 4628
},
{
"epoch": 0.9768705330063032,
"grad_norm": 0.11642242968082428,
"learning_rate": 1.992818671454219e-05,
"loss": 1.9627984762191772,
"step": 4630
},
{
"epoch": 0.9772925073186169,
"grad_norm": 0.11063104122877121,
"learning_rate": 1.9569120287253142e-05,
"loss": 1.983302354812622,
"step": 4632
},
{
"epoch": 0.9777144816309307,
"grad_norm": 0.11088722944259644,
"learning_rate": 1.9210053859964093e-05,
"loss": 2.0288641452789307,
"step": 4634
},
{
"epoch": 0.9781364559432445,
"grad_norm": 0.10941636562347412,
"learning_rate": 1.8850987432675045e-05,
"loss": 1.9927992820739746,
"step": 4636
},
{
"epoch": 0.9785584302555582,
"grad_norm": 0.10383333265781403,
"learning_rate": 1.8491921005385997e-05,
"loss": 2.0099031925201416,
"step": 4638
},
{
"epoch": 0.9789804045678719,
"grad_norm": 0.10626033693552017,
"learning_rate": 1.813285457809695e-05,
"loss": 1.9769357442855835,
"step": 4640
},
{
"epoch": 0.9794023788801857,
"grad_norm": 0.12086393684148788,
"learning_rate": 1.77737881508079e-05,
"loss": 2.0133321285247803,
"step": 4642
},
{
"epoch": 0.9798243531924994,
"grad_norm": 0.10567011684179306,
"learning_rate": 1.7414721723518852e-05,
"loss": 2.023261308670044,
"step": 4644
},
{
"epoch": 0.9802463275048131,
"grad_norm": 0.10585828125476837,
"learning_rate": 1.7055655296229804e-05,
"loss": 1.963270902633667,
"step": 4646
},
{
"epoch": 0.9806683018171268,
"grad_norm": 0.10608888417482376,
"learning_rate": 1.6696588868940756e-05,
"loss": 1.9312522411346436,
"step": 4648
},
{
"epoch": 0.9810902761294407,
"grad_norm": 0.1118353009223938,
"learning_rate": 1.6337522441651707e-05,
"loss": 1.9820836782455444,
"step": 4650
},
{
"epoch": 0.9815122504417544,
"grad_norm": 0.11541516333818436,
"learning_rate": 1.597845601436266e-05,
"loss": 1.9589990377426147,
"step": 4652
},
{
"epoch": 0.9819342247540681,
"grad_norm": 0.12044358253479004,
"learning_rate": 1.561938958707361e-05,
"loss": 2.009178638458252,
"step": 4654
},
{
"epoch": 0.9823561990663818,
"grad_norm": 0.11771586537361145,
"learning_rate": 1.5260323159784563e-05,
"loss": 1.905807614326477,
"step": 4656
},
{
"epoch": 0.9827781733786956,
"grad_norm": 0.09837169200181961,
"learning_rate": 1.4901256732495511e-05,
"loss": 2.0050885677337646,
"step": 4658
},
{
"epoch": 0.9832001476910093,
"grad_norm": 0.10805536806583405,
"learning_rate": 1.4542190305206463e-05,
"loss": 1.9015088081359863,
"step": 4660
},
{
"epoch": 0.983622122003323,
"grad_norm": 0.13526605069637299,
"learning_rate": 1.4183123877917415e-05,
"loss": 1.9025412797927856,
"step": 4662
},
{
"epoch": 0.9840440963156368,
"grad_norm": 0.11049235612154007,
"learning_rate": 1.3824057450628366e-05,
"loss": 1.9939558506011963,
"step": 4664
},
{
"epoch": 0.9844660706279506,
"grad_norm": 0.11036085337400436,
"learning_rate": 1.3464991023339318e-05,
"loss": 1.9329862594604492,
"step": 4666
},
{
"epoch": 0.9848880449402643,
"grad_norm": 0.10211043059825897,
"learning_rate": 1.310592459605027e-05,
"loss": 1.9791620969772339,
"step": 4668
},
{
"epoch": 0.985310019252578,
"grad_norm": 0.10017786175012589,
"learning_rate": 1.2746858168761221e-05,
"loss": 2.033644437789917,
"step": 4670
},
{
"epoch": 0.9857319935648917,
"grad_norm": 0.11860562115907669,
"learning_rate": 1.2387791741472173e-05,
"loss": 1.9536200761795044,
"step": 4672
},
{
"epoch": 0.9861539678772054,
"grad_norm": 0.1107722818851471,
"learning_rate": 1.2028725314183125e-05,
"loss": 1.9566007852554321,
"step": 4674
},
{
"epoch": 0.9865759421895192,
"grad_norm": 0.09551042318344116,
"learning_rate": 1.1669658886894077e-05,
"loss": 2.0149271488189697,
"step": 4676
},
{
"epoch": 0.986997916501833,
"grad_norm": 0.10458600521087646,
"learning_rate": 1.1310592459605028e-05,
"loss": 1.9659817218780518,
"step": 4678
},
{
"epoch": 0.9874198908141467,
"grad_norm": 0.10398183763027191,
"learning_rate": 1.0951526032315979e-05,
"loss": 2.0331315994262695,
"step": 4680
},
{
"epoch": 0.9878418651264604,
"grad_norm": 0.10857747495174408,
"learning_rate": 1.059245960502693e-05,
"loss": 1.999959111213684,
"step": 4682
},
{
"epoch": 0.9882638394387742,
"grad_norm": 0.09563376754522324,
"learning_rate": 1.0233393177737882e-05,
"loss": 2.0447683334350586,
"step": 4684
},
{
"epoch": 0.9886858137510879,
"grad_norm": 0.12259466201066971,
"learning_rate": 9.874326750448834e-06,
"loss": 1.9301782846450806,
"step": 4686
},
{
"epoch": 0.9891077880634016,
"grad_norm": 0.11285020411014557,
"learning_rate": 9.515260323159784e-06,
"loss": 2.033548355102539,
"step": 4688
},
{
"epoch": 0.9895297623757153,
"grad_norm": 0.10428531467914581,
"learning_rate": 9.156193895870736e-06,
"loss": 1.9834178686141968,
"step": 4690
},
{
"epoch": 0.9899517366880292,
"grad_norm": 0.10745920985937119,
"learning_rate": 8.797127468581689e-06,
"loss": 1.9925299882888794,
"step": 4692
},
{
"epoch": 0.9903737110003429,
"grad_norm": 0.09838896244764328,
"learning_rate": 8.43806104129264e-06,
"loss": 1.9267971515655518,
"step": 4694
},
{
"epoch": 0.9907956853126566,
"grad_norm": 0.11232728511095047,
"learning_rate": 8.07899461400359e-06,
"loss": 1.9885969161987305,
"step": 4696
},
{
"epoch": 0.9912176596249703,
"grad_norm": 0.10153519362211227,
"learning_rate": 7.719928186714543e-06,
"loss": 1.965550184249878,
"step": 4698
},
{
"epoch": 0.991639633937284,
"grad_norm": 0.10343014448881149,
"learning_rate": 7.360861759425494e-06,
"loss": 1.9578814506530762,
"step": 4700
},
{
"epoch": 0.9920616082495978,
"grad_norm": 0.11567633599042892,
"learning_rate": 7.001795332136446e-06,
"loss": 2.0264432430267334,
"step": 4702
},
{
"epoch": 0.9924835825619115,
"grad_norm": 0.11094778776168823,
"learning_rate": 6.642728904847396e-06,
"loss": 1.9655907154083252,
"step": 4704
},
{
"epoch": 0.9929055568742253,
"grad_norm": 0.10449371486902237,
"learning_rate": 6.283662477558349e-06,
"loss": 1.949964165687561,
"step": 4706
},
{
"epoch": 0.993327531186539,
"grad_norm": 0.11244763433933258,
"learning_rate": 5.9245960502693004e-06,
"loss": 2.024744749069214,
"step": 4708
},
{
"epoch": 0.9937495054988528,
"grad_norm": 0.10339177399873734,
"learning_rate": 5.565529622980251e-06,
"loss": 2.0018577575683594,
"step": 4710
},
{
"epoch": 0.9941714798111665,
"grad_norm": 0.11212055385112762,
"learning_rate": 5.206463195691203e-06,
"loss": 1.9637234210968018,
"step": 4712
},
{
"epoch": 0.9945934541234802,
"grad_norm": 0.1078438013792038,
"learning_rate": 4.847396768402154e-06,
"loss": 2.0055603981018066,
"step": 4714
},
{
"epoch": 0.9950154284357939,
"grad_norm": 0.1065535768866539,
"learning_rate": 4.488330341113106e-06,
"loss": 1.9720178842544556,
"step": 4716
},
{
"epoch": 0.9954374027481077,
"grad_norm": 0.10319648683071136,
"learning_rate": 4.129263913824058e-06,
"loss": 2.0201807022094727,
"step": 4718
},
{
"epoch": 0.9958593770604215,
"grad_norm": 0.10053250938653946,
"learning_rate": 3.770197486535009e-06,
"loss": 1.988191843032837,
"step": 4720
},
{
"epoch": 0.9962813513727352,
"grad_norm": 0.11511623859405518,
"learning_rate": 3.411131059245961e-06,
"loss": 1.93172287940979,
"step": 4722
},
{
"epoch": 0.9967033256850489,
"grad_norm": 0.1131865456700325,
"learning_rate": 3.0520646319569123e-06,
"loss": 1.9372680187225342,
"step": 4724
},
{
"epoch": 0.9971252999973627,
"grad_norm": 0.09997057169675827,
"learning_rate": 2.6929982046678636e-06,
"loss": 1.9457156658172607,
"step": 4726
},
{
"epoch": 0.9975472743096764,
"grad_norm": 0.09570550173521042,
"learning_rate": 2.333931777378815e-06,
"loss": 1.9646587371826172,
"step": 4728
},
{
"epoch": 0.9979692486219901,
"grad_norm": 0.10176803916692734,
"learning_rate": 1.9748653500897667e-06,
"loss": 1.9580981731414795,
"step": 4730
},
{
"epoch": 0.9983912229343038,
"grad_norm": 0.10773016512393951,
"learning_rate": 1.6157989228007182e-06,
"loss": 1.9681389331817627,
"step": 4732
},
{
"epoch": 0.9988131972466177,
"grad_norm": 0.11025076359510422,
"learning_rate": 1.2567324955116697e-06,
"loss": 2.0080111026763916,
"step": 4734
},
{
"epoch": 0.9992351715589314,
"grad_norm": 0.10569436848163605,
"learning_rate": 8.976660682226213e-07,
"loss": 1.917080283164978,
"step": 4736
},
{
"epoch": 0.9996571458712451,
"grad_norm": 0.11203278601169586,
"learning_rate": 5.385996409335728e-07,
"loss": 1.8714354038238525,
"step": 4738
},
{
"epoch": 1.0,
"grad_norm": 0.11829442530870438,
"learning_rate": 1.7953321364452426e-07,
"loss": 1.9940471649169922,
"step": 4740
},
{
"epoch": 1.0,
"eval_loss": 1.9716895818710327,
"eval_runtime": 85.9928,
"eval_samples_per_second": 35.631,
"eval_steps_per_second": 4.454,
"step": 4740
}
],
"logging_steps": 2,
"max_steps": 4740,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 474,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.2140322727104123e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}