{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "eval_steps": 500, "global_step": 5796, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003450655624568668, "grad_norm": 2.7905118465423584, "learning_rate": 4.993098688750863e-05, "loss": 1.7703, "step": 10 }, { "epoch": 0.006901311249137336, "grad_norm": 1.9586756229400635, "learning_rate": 4.984472049689442e-05, "loss": 0.9224, "step": 20 }, { "epoch": 0.010351966873706004, "grad_norm": 1.6500365734100342, "learning_rate": 4.9758454106280194e-05, "loss": 0.7301, "step": 30 }, { "epoch": 0.013802622498274672, "grad_norm": 1.4835381507873535, "learning_rate": 4.967218771566598e-05, "loss": 0.6398, "step": 40 }, { "epoch": 0.01725327812284334, "grad_norm": 1.4787698984146118, "learning_rate": 4.958592132505176e-05, "loss": 0.6192, "step": 50 }, { "epoch": 0.020703933747412008, "grad_norm": 1.2620238065719604, "learning_rate": 4.949965493443754e-05, "loss": 0.5913, "step": 60 }, { "epoch": 0.024154589371980676, "grad_norm": 1.2006996870040894, "learning_rate": 4.941338854382333e-05, "loss": 0.5527, "step": 70 }, { "epoch": 0.027605244996549344, "grad_norm": 1.395799160003662, "learning_rate": 4.932712215320911e-05, "loss": 0.5465, "step": 80 }, { "epoch": 0.031055900621118012, "grad_norm": 1.2285804748535156, "learning_rate": 4.92408557625949e-05, "loss": 0.5123, "step": 90 }, { "epoch": 0.03450655624568668, "grad_norm": 1.2713665962219238, "learning_rate": 4.915458937198068e-05, "loss": 0.5129, "step": 100 }, { "epoch": 0.03795721187025535, "grad_norm": 1.1799707412719727, "learning_rate": 4.906832298136646e-05, "loss": 0.5046, "step": 110 }, { "epoch": 0.041407867494824016, "grad_norm": 1.0650718212127686, "learning_rate": 4.8982056590752246e-05, "loss": 0.4909, "step": 120 }, { "epoch": 0.04485852311939269, "grad_norm": 1.0782257318496704, "learning_rate": 4.889579020013803e-05, "loss": 0.4822, "step": 130 }, { "epoch": 0.04830917874396135, "grad_norm": 1.0218424797058105, "learning_rate": 4.880952380952381e-05, "loss": 0.4791, "step": 140 }, { "epoch": 0.051759834368530024, "grad_norm": 0.9384006261825562, "learning_rate": 4.8723257418909594e-05, "loss": 0.4811, "step": 150 }, { "epoch": 0.05521048999309869, "grad_norm": 1.0220593214035034, "learning_rate": 4.8636991028295384e-05, "loss": 0.4718, "step": 160 }, { "epoch": 0.05866114561766736, "grad_norm": 1.1225941181182861, "learning_rate": 4.855072463768116e-05, "loss": 0.449, "step": 170 }, { "epoch": 0.062111801242236024, "grad_norm": 0.9393426775932312, "learning_rate": 4.846445824706695e-05, "loss": 0.4644, "step": 180 }, { "epoch": 0.06556245686680469, "grad_norm": 0.9687050580978394, "learning_rate": 4.8378191856452725e-05, "loss": 0.4554, "step": 190 }, { "epoch": 0.06901311249137336, "grad_norm": 0.9987794160842896, "learning_rate": 4.829192546583851e-05, "loss": 0.4487, "step": 200 }, { "epoch": 0.07246376811594203, "grad_norm": 0.8862098455429077, "learning_rate": 4.82056590752243e-05, "loss": 0.4507, "step": 210 }, { "epoch": 0.0759144237405107, "grad_norm": 0.8340906500816345, "learning_rate": 4.8119392684610074e-05, "loss": 0.438, "step": 220 }, { "epoch": 0.07936507936507936, "grad_norm": 0.8599636554718018, "learning_rate": 4.803312629399586e-05, "loss": 0.4406, "step": 230 }, { "epoch": 0.08281573498964803, "grad_norm": 0.8934054970741272, "learning_rate": 4.7946859903381646e-05, "loss": 0.4351, "step": 240 }, { "epoch": 0.0862663906142167, "grad_norm": 0.7076365947723389, "learning_rate": 4.786059351276743e-05, "loss": 0.434, "step": 250 }, { "epoch": 0.08971704623878538, "grad_norm": 0.8420814871788025, "learning_rate": 4.777432712215321e-05, "loss": 0.434, "step": 260 }, { "epoch": 0.09316770186335403, "grad_norm": 0.768744170665741, "learning_rate": 4.7688060731538994e-05, "loss": 0.4317, "step": 270 }, { "epoch": 0.0966183574879227, "grad_norm": 0.8374963998794556, "learning_rate": 4.760179434092478e-05, "loss": 0.4297, "step": 280 }, { "epoch": 0.10006901311249138, "grad_norm": 0.7579261064529419, "learning_rate": 4.751552795031056e-05, "loss": 0.4304, "step": 290 }, { "epoch": 0.10351966873706005, "grad_norm": 0.7822970151901245, "learning_rate": 4.742926155969634e-05, "loss": 0.4323, "step": 300 }, { "epoch": 0.1069703243616287, "grad_norm": 0.7832859754562378, "learning_rate": 4.7342995169082125e-05, "loss": 0.4235, "step": 310 }, { "epoch": 0.11042097998619738, "grad_norm": 0.8547880053520203, "learning_rate": 4.7256728778467915e-05, "loss": 0.4295, "step": 320 }, { "epoch": 0.11387163561076605, "grad_norm": 0.7879784107208252, "learning_rate": 4.717046238785369e-05, "loss": 0.4281, "step": 330 }, { "epoch": 0.11732229123533472, "grad_norm": 0.7858608961105347, "learning_rate": 4.708419599723948e-05, "loss": 0.4162, "step": 340 }, { "epoch": 0.12077294685990338, "grad_norm": 0.730283796787262, "learning_rate": 4.699792960662526e-05, "loss": 0.4167, "step": 350 }, { "epoch": 0.12422360248447205, "grad_norm": 0.7356073260307312, "learning_rate": 4.6911663216011046e-05, "loss": 0.4156, "step": 360 }, { "epoch": 0.1276742581090407, "grad_norm": 0.6808834075927734, "learning_rate": 4.682539682539683e-05, "loss": 0.4095, "step": 370 }, { "epoch": 0.13112491373360938, "grad_norm": 0.7014838457107544, "learning_rate": 4.673913043478261e-05, "loss": 0.4163, "step": 380 }, { "epoch": 0.13457556935817805, "grad_norm": 0.8114644289016724, "learning_rate": 4.6652864044168394e-05, "loss": 0.403, "step": 390 }, { "epoch": 0.13802622498274672, "grad_norm": 0.7766273021697998, "learning_rate": 4.656659765355418e-05, "loss": 0.4144, "step": 400 }, { "epoch": 0.1414768806073154, "grad_norm": 0.7870580554008484, "learning_rate": 4.6480331262939966e-05, "loss": 0.4079, "step": 410 }, { "epoch": 0.14492753623188406, "grad_norm": 0.6845405101776123, "learning_rate": 4.639406487232574e-05, "loss": 0.4108, "step": 420 }, { "epoch": 0.14837819185645273, "grad_norm": 0.725813627243042, "learning_rate": 4.6307798481711525e-05, "loss": 0.4085, "step": 430 }, { "epoch": 0.1518288474810214, "grad_norm": 0.7153050899505615, "learning_rate": 4.622153209109731e-05, "loss": 0.4059, "step": 440 }, { "epoch": 0.15527950310559005, "grad_norm": 0.7423162460327148, "learning_rate": 4.613526570048309e-05, "loss": 0.4022, "step": 450 }, { "epoch": 0.15873015873015872, "grad_norm": 0.7227168083190918, "learning_rate": 4.604899930986888e-05, "loss": 0.4103, "step": 460 }, { "epoch": 0.1621808143547274, "grad_norm": 0.7159332633018494, "learning_rate": 4.5962732919254656e-05, "loss": 0.4114, "step": 470 }, { "epoch": 0.16563146997929606, "grad_norm": 0.7028360366821289, "learning_rate": 4.5876466528640446e-05, "loss": 0.4052, "step": 480 }, { "epoch": 0.16908212560386474, "grad_norm": 0.6726347208023071, "learning_rate": 4.579020013802623e-05, "loss": 0.4114, "step": 490 }, { "epoch": 0.1725327812284334, "grad_norm": 0.694276750087738, "learning_rate": 4.570393374741201e-05, "loss": 0.3981, "step": 500 }, { "epoch": 0.17598343685300208, "grad_norm": 0.6861996650695801, "learning_rate": 4.5617667356797794e-05, "loss": 0.391, "step": 510 }, { "epoch": 0.17943409247757075, "grad_norm": 0.6844497919082642, "learning_rate": 4.553140096618358e-05, "loss": 0.389, "step": 520 }, { "epoch": 0.1828847481021394, "grad_norm": 0.7536503076553345, "learning_rate": 4.544513457556936e-05, "loss": 0.4013, "step": 530 }, { "epoch": 0.18633540372670807, "grad_norm": 0.6106601357460022, "learning_rate": 4.535886818495514e-05, "loss": 0.3913, "step": 540 }, { "epoch": 0.18978605935127674, "grad_norm": 0.6462461948394775, "learning_rate": 4.5272601794340925e-05, "loss": 0.3887, "step": 550 }, { "epoch": 0.1932367149758454, "grad_norm": 0.6759409308433533, "learning_rate": 4.518633540372671e-05, "loss": 0.3972, "step": 560 }, { "epoch": 0.19668737060041408, "grad_norm": 0.6945746541023254, "learning_rate": 4.51000690131125e-05, "loss": 0.3865, "step": 570 }, { "epoch": 0.20013802622498275, "grad_norm": 0.6554960608482361, "learning_rate": 4.5013802622498274e-05, "loss": 0.3902, "step": 580 }, { "epoch": 0.20358868184955142, "grad_norm": 0.6467313170433044, "learning_rate": 4.492753623188406e-05, "loss": 0.3889, "step": 590 }, { "epoch": 0.2070393374741201, "grad_norm": 0.6078060865402222, "learning_rate": 4.4841269841269846e-05, "loss": 0.3898, "step": 600 }, { "epoch": 0.21048999309868874, "grad_norm": 0.8312433362007141, "learning_rate": 4.475500345065562e-05, "loss": 0.3897, "step": 610 }, { "epoch": 0.2139406487232574, "grad_norm": 0.5776800513267517, "learning_rate": 4.466873706004141e-05, "loss": 0.3885, "step": 620 }, { "epoch": 0.21739130434782608, "grad_norm": 0.6104868054389954, "learning_rate": 4.4582470669427194e-05, "loss": 0.3992, "step": 630 }, { "epoch": 0.22084195997239475, "grad_norm": 0.6720201373100281, "learning_rate": 4.449620427881298e-05, "loss": 0.3865, "step": 640 }, { "epoch": 0.22429261559696342, "grad_norm": 0.6955760717391968, "learning_rate": 4.440993788819876e-05, "loss": 0.3878, "step": 650 }, { "epoch": 0.2277432712215321, "grad_norm": 0.6438873410224915, "learning_rate": 4.432367149758454e-05, "loss": 0.3881, "step": 660 }, { "epoch": 0.23119392684610077, "grad_norm": 0.622840940952301, "learning_rate": 4.4237405106970325e-05, "loss": 0.3939, "step": 670 }, { "epoch": 0.23464458247066944, "grad_norm": 0.6797952055931091, "learning_rate": 4.415113871635611e-05, "loss": 0.401, "step": 680 }, { "epoch": 0.23809523809523808, "grad_norm": 0.6357337832450867, "learning_rate": 4.406487232574189e-05, "loss": 0.391, "step": 690 }, { "epoch": 0.24154589371980675, "grad_norm": 0.612765908241272, "learning_rate": 4.3978605935127674e-05, "loss": 0.389, "step": 700 }, { "epoch": 0.24499654934437542, "grad_norm": 0.5818207859992981, "learning_rate": 4.389233954451346e-05, "loss": 0.3835, "step": 710 }, { "epoch": 0.2484472049689441, "grad_norm": 0.530667245388031, "learning_rate": 4.380607315389924e-05, "loss": 0.3789, "step": 720 }, { "epoch": 0.25189786059351277, "grad_norm": 0.6901742815971375, "learning_rate": 4.371980676328503e-05, "loss": 0.3832, "step": 730 }, { "epoch": 0.2553485162180814, "grad_norm": 0.6087227463722229, "learning_rate": 4.363354037267081e-05, "loss": 0.3766, "step": 740 }, { "epoch": 0.2587991718426501, "grad_norm": 0.5868121981620789, "learning_rate": 4.3547273982056594e-05, "loss": 0.3717, "step": 750 }, { "epoch": 0.26224982746721875, "grad_norm": 0.646168053150177, "learning_rate": 4.346100759144238e-05, "loss": 0.3882, "step": 760 }, { "epoch": 0.26570048309178745, "grad_norm": 0.6688451170921326, "learning_rate": 4.337474120082816e-05, "loss": 0.3791, "step": 770 }, { "epoch": 0.2691511387163561, "grad_norm": 0.5901142358779907, "learning_rate": 4.328847481021394e-05, "loss": 0.3854, "step": 780 }, { "epoch": 0.2726017943409248, "grad_norm": 0.6211585402488708, "learning_rate": 4.3202208419599725e-05, "loss": 0.3829, "step": 790 }, { "epoch": 0.27605244996549344, "grad_norm": 0.6081684231758118, "learning_rate": 4.3115942028985515e-05, "loss": 0.3859, "step": 800 }, { "epoch": 0.2795031055900621, "grad_norm": 0.6079261898994446, "learning_rate": 4.302967563837129e-05, "loss": 0.3766, "step": 810 }, { "epoch": 0.2829537612146308, "grad_norm": 0.5618354082107544, "learning_rate": 4.294340924775708e-05, "loss": 0.3846, "step": 820 }, { "epoch": 0.2864044168391994, "grad_norm": 0.5210685729980469, "learning_rate": 4.2857142857142856e-05, "loss": 0.3706, "step": 830 }, { "epoch": 0.2898550724637681, "grad_norm": 0.587040901184082, "learning_rate": 4.277087646652864e-05, "loss": 0.3771, "step": 840 }, { "epoch": 0.29330572808833677, "grad_norm": 0.6124225854873657, "learning_rate": 4.268461007591443e-05, "loss": 0.3818, "step": 850 }, { "epoch": 0.29675638371290547, "grad_norm": 0.6286453008651733, "learning_rate": 4.2598343685300205e-05, "loss": 0.3712, "step": 860 }, { "epoch": 0.3002070393374741, "grad_norm": 0.5031912922859192, "learning_rate": 4.2512077294685994e-05, "loss": 0.3737, "step": 870 }, { "epoch": 0.3036576949620428, "grad_norm": 0.56924968957901, "learning_rate": 4.242581090407178e-05, "loss": 0.3675, "step": 880 }, { "epoch": 0.30710835058661146, "grad_norm": 0.5326061248779297, "learning_rate": 4.233954451345756e-05, "loss": 0.3724, "step": 890 }, { "epoch": 0.3105590062111801, "grad_norm": 0.6027249097824097, "learning_rate": 4.225327812284334e-05, "loss": 0.3752, "step": 900 }, { "epoch": 0.3140096618357488, "grad_norm": 0.48393580317497253, "learning_rate": 4.2167011732229125e-05, "loss": 0.3696, "step": 910 }, { "epoch": 0.31746031746031744, "grad_norm": 0.5405918955802917, "learning_rate": 4.208074534161491e-05, "loss": 0.3798, "step": 920 }, { "epoch": 0.32091097308488614, "grad_norm": 0.5802543759346008, "learning_rate": 4.199447895100069e-05, "loss": 0.3644, "step": 930 }, { "epoch": 0.3243616287094548, "grad_norm": 0.5545613169670105, "learning_rate": 4.1908212560386474e-05, "loss": 0.3627, "step": 940 }, { "epoch": 0.3278122843340235, "grad_norm": 0.536221444606781, "learning_rate": 4.1821946169772256e-05, "loss": 0.3648, "step": 950 }, { "epoch": 0.33126293995859213, "grad_norm": 0.5735536217689514, "learning_rate": 4.1735679779158046e-05, "loss": 0.3694, "step": 960 }, { "epoch": 0.33471359558316083, "grad_norm": 0.5893762707710266, "learning_rate": 4.164941338854382e-05, "loss": 0.3691, "step": 970 }, { "epoch": 0.33816425120772947, "grad_norm": 0.5472971796989441, "learning_rate": 4.156314699792961e-05, "loss": 0.3697, "step": 980 }, { "epoch": 0.3416149068322981, "grad_norm": 0.5291191935539246, "learning_rate": 4.1476880607315394e-05, "loss": 0.378, "step": 990 }, { "epoch": 0.3450655624568668, "grad_norm": 0.5894433259963989, "learning_rate": 4.139061421670117e-05, "loss": 0.3676, "step": 1000 }, { "epoch": 0.34851621808143546, "grad_norm": 0.519848644733429, "learning_rate": 4.130434782608696e-05, "loss": 0.366, "step": 1010 }, { "epoch": 0.35196687370600416, "grad_norm": 0.620188295841217, "learning_rate": 4.121808143547274e-05, "loss": 0.3569, "step": 1020 }, { "epoch": 0.3554175293305728, "grad_norm": 0.589487612247467, "learning_rate": 4.1131815044858525e-05, "loss": 0.3599, "step": 1030 }, { "epoch": 0.3588681849551415, "grad_norm": 0.5304994583129883, "learning_rate": 4.104554865424431e-05, "loss": 0.3622, "step": 1040 }, { "epoch": 0.36231884057971014, "grad_norm": 0.5731646418571472, "learning_rate": 4.09592822636301e-05, "loss": 0.3641, "step": 1050 }, { "epoch": 0.3657694962042788, "grad_norm": 0.527119517326355, "learning_rate": 4.0873015873015874e-05, "loss": 0.3639, "step": 1060 }, { "epoch": 0.3692201518288475, "grad_norm": 0.4605042636394501, "learning_rate": 4.0786749482401656e-05, "loss": 0.3685, "step": 1070 }, { "epoch": 0.37267080745341613, "grad_norm": 0.5230356454849243, "learning_rate": 4.070048309178744e-05, "loss": 0.369, "step": 1080 }, { "epoch": 0.37612146307798483, "grad_norm": 0.5160967111587524, "learning_rate": 4.061421670117322e-05, "loss": 0.3577, "step": 1090 }, { "epoch": 0.3795721187025535, "grad_norm": 0.5491387248039246, "learning_rate": 4.052795031055901e-05, "loss": 0.3786, "step": 1100 }, { "epoch": 0.3830227743271222, "grad_norm": 0.5626661777496338, "learning_rate": 4.044168391994479e-05, "loss": 0.3676, "step": 1110 }, { "epoch": 0.3864734299516908, "grad_norm": 0.5415998697280884, "learning_rate": 4.035541752933058e-05, "loss": 0.3608, "step": 1120 }, { "epoch": 0.3899240855762595, "grad_norm": 0.6332689523696899, "learning_rate": 4.026915113871636e-05, "loss": 0.3707, "step": 1130 }, { "epoch": 0.39337474120082816, "grad_norm": 0.5432307124137878, "learning_rate": 4.018288474810214e-05, "loss": 0.364, "step": 1140 }, { "epoch": 0.3968253968253968, "grad_norm": 0.498445987701416, "learning_rate": 4.0096618357487925e-05, "loss": 0.3609, "step": 1150 }, { "epoch": 0.4002760524499655, "grad_norm": 0.5753563046455383, "learning_rate": 4.001035196687371e-05, "loss": 0.3625, "step": 1160 }, { "epoch": 0.40372670807453415, "grad_norm": 0.5175043940544128, "learning_rate": 3.992408557625949e-05, "loss": 0.3602, "step": 1170 }, { "epoch": 0.40717736369910285, "grad_norm": 0.521827220916748, "learning_rate": 3.9837819185645274e-05, "loss": 0.3523, "step": 1180 }, { "epoch": 0.4106280193236715, "grad_norm": 0.538490355014801, "learning_rate": 3.9751552795031056e-05, "loss": 0.3702, "step": 1190 }, { "epoch": 0.4140786749482402, "grad_norm": 0.5001350045204163, "learning_rate": 3.966528640441684e-05, "loss": 0.3455, "step": 1200 }, { "epoch": 0.41752933057280883, "grad_norm": 0.47586187720298767, "learning_rate": 3.957902001380263e-05, "loss": 0.3569, "step": 1210 }, { "epoch": 0.4209799861973775, "grad_norm": 0.5283018946647644, "learning_rate": 3.9492753623188405e-05, "loss": 0.3691, "step": 1220 }, { "epoch": 0.4244306418219462, "grad_norm": 0.6038020253181458, "learning_rate": 3.940648723257419e-05, "loss": 0.3611, "step": 1230 }, { "epoch": 0.4278812974465148, "grad_norm": 0.4514884054660797, "learning_rate": 3.932022084195998e-05, "loss": 0.367, "step": 1240 }, { "epoch": 0.4313319530710835, "grad_norm": 0.5189118385314941, "learning_rate": 3.923395445134575e-05, "loss": 0.3553, "step": 1250 }, { "epoch": 0.43478260869565216, "grad_norm": 0.5897815227508545, "learning_rate": 3.914768806073154e-05, "loss": 0.3658, "step": 1260 }, { "epoch": 0.43823326432022086, "grad_norm": 0.5077459812164307, "learning_rate": 3.9061421670117325e-05, "loss": 0.3734, "step": 1270 }, { "epoch": 0.4416839199447895, "grad_norm": 0.5408049821853638, "learning_rate": 3.897515527950311e-05, "loss": 0.3516, "step": 1280 }, { "epoch": 0.4451345755693582, "grad_norm": 0.5270346999168396, "learning_rate": 3.888888888888889e-05, "loss": 0.3645, "step": 1290 }, { "epoch": 0.44858523119392685, "grad_norm": 0.5303778052330017, "learning_rate": 3.8802622498274674e-05, "loss": 0.3657, "step": 1300 }, { "epoch": 0.4520358868184955, "grad_norm": 0.47278884053230286, "learning_rate": 3.8716356107660456e-05, "loss": 0.3535, "step": 1310 }, { "epoch": 0.4554865424430642, "grad_norm": 0.5036910772323608, "learning_rate": 3.863008971704624e-05, "loss": 0.3541, "step": 1320 }, { "epoch": 0.45893719806763283, "grad_norm": 0.4585624635219574, "learning_rate": 3.854382332643202e-05, "loss": 0.3623, "step": 1330 }, { "epoch": 0.46238785369220153, "grad_norm": 0.5310036540031433, "learning_rate": 3.8457556935817805e-05, "loss": 0.3586, "step": 1340 }, { "epoch": 0.4658385093167702, "grad_norm": 0.5218350291252136, "learning_rate": 3.8371290545203594e-05, "loss": 0.3412, "step": 1350 }, { "epoch": 0.4692891649413389, "grad_norm": 0.4900505244731903, "learning_rate": 3.828502415458937e-05, "loss": 0.3589, "step": 1360 }, { "epoch": 0.4727398205659075, "grad_norm": 0.526491105556488, "learning_rate": 3.819875776397516e-05, "loss": 0.3509, "step": 1370 }, { "epoch": 0.47619047619047616, "grad_norm": 0.4933297634124756, "learning_rate": 3.811249137336094e-05, "loss": 0.3551, "step": 1380 }, { "epoch": 0.47964113181504486, "grad_norm": 0.5509902834892273, "learning_rate": 3.8026224982746725e-05, "loss": 0.3455, "step": 1390 }, { "epoch": 0.4830917874396135, "grad_norm": 0.4903262257575989, "learning_rate": 3.793995859213251e-05, "loss": 0.3604, "step": 1400 }, { "epoch": 0.4865424430641822, "grad_norm": 0.5923327207565308, "learning_rate": 3.785369220151829e-05, "loss": 0.3507, "step": 1410 }, { "epoch": 0.48999309868875085, "grad_norm": 0.4920748472213745, "learning_rate": 3.7767425810904074e-05, "loss": 0.3544, "step": 1420 }, { "epoch": 0.49344375431331955, "grad_norm": 0.5345553755760193, "learning_rate": 3.7681159420289856e-05, "loss": 0.3551, "step": 1430 }, { "epoch": 0.4968944099378882, "grad_norm": 0.532013475894928, "learning_rate": 3.759489302967564e-05, "loss": 0.3457, "step": 1440 }, { "epoch": 0.5003450655624568, "grad_norm": 0.5193171501159668, "learning_rate": 3.750862663906142e-05, "loss": 0.3527, "step": 1450 }, { "epoch": 0.5037957211870255, "grad_norm": 0.4619022309780121, "learning_rate": 3.742236024844721e-05, "loss": 0.363, "step": 1460 }, { "epoch": 0.5072463768115942, "grad_norm": 0.5606441497802734, "learning_rate": 3.733609385783299e-05, "loss": 0.3547, "step": 1470 }, { "epoch": 0.5106970324361628, "grad_norm": 0.5041471719741821, "learning_rate": 3.724982746721877e-05, "loss": 0.3504, "step": 1480 }, { "epoch": 0.5141476880607315, "grad_norm": 0.46617457270622253, "learning_rate": 3.716356107660456e-05, "loss": 0.3492, "step": 1490 }, { "epoch": 0.5175983436853002, "grad_norm": 0.4503689706325531, "learning_rate": 3.7077294685990336e-05, "loss": 0.3585, "step": 1500 }, { "epoch": 0.5210489993098689, "grad_norm": 0.4509599506855011, "learning_rate": 3.6991028295376125e-05, "loss": 0.351, "step": 1510 }, { "epoch": 0.5244996549344375, "grad_norm": 0.46619853377342224, "learning_rate": 3.690476190476191e-05, "loss": 0.3588, "step": 1520 }, { "epoch": 0.5279503105590062, "grad_norm": 0.528289794921875, "learning_rate": 3.681849551414769e-05, "loss": 0.3499, "step": 1530 }, { "epoch": 0.5314009661835749, "grad_norm": 0.5260491371154785, "learning_rate": 3.6732229123533474e-05, "loss": 0.3487, "step": 1540 }, { "epoch": 0.5348516218081435, "grad_norm": 0.5708985924720764, "learning_rate": 3.6645962732919256e-05, "loss": 0.3587, "step": 1550 }, { "epoch": 0.5383022774327122, "grad_norm": 0.4738638401031494, "learning_rate": 3.655969634230504e-05, "loss": 0.3527, "step": 1560 }, { "epoch": 0.5417529330572809, "grad_norm": 0.447976291179657, "learning_rate": 3.647342995169082e-05, "loss": 0.3457, "step": 1570 }, { "epoch": 0.5452035886818496, "grad_norm": 0.44627007842063904, "learning_rate": 3.6387163561076605e-05, "loss": 0.3524, "step": 1580 }, { "epoch": 0.5486542443064182, "grad_norm": 0.4650660753250122, "learning_rate": 3.630089717046239e-05, "loss": 0.3569, "step": 1590 }, { "epoch": 0.5521048999309869, "grad_norm": 0.4500880837440491, "learning_rate": 3.621463077984818e-05, "loss": 0.341, "step": 1600 }, { "epoch": 0.5555555555555556, "grad_norm": 0.5702126622200012, "learning_rate": 3.612836438923395e-05, "loss": 0.3519, "step": 1610 }, { "epoch": 0.5590062111801242, "grad_norm": 0.4988945424556732, "learning_rate": 3.604209799861974e-05, "loss": 0.3427, "step": 1620 }, { "epoch": 0.5624568668046929, "grad_norm": 0.5189987421035767, "learning_rate": 3.5955831608005525e-05, "loss": 0.3473, "step": 1630 }, { "epoch": 0.5659075224292616, "grad_norm": 0.45712926983833313, "learning_rate": 3.58695652173913e-05, "loss": 0.3612, "step": 1640 }, { "epoch": 0.5693581780538303, "grad_norm": 0.5472447276115417, "learning_rate": 3.578329882677709e-05, "loss": 0.3456, "step": 1650 }, { "epoch": 0.5728088336783989, "grad_norm": 0.4873930811882019, "learning_rate": 3.5697032436162874e-05, "loss": 0.3692, "step": 1660 }, { "epoch": 0.5762594893029676, "grad_norm": 0.503449022769928, "learning_rate": 3.5610766045548656e-05, "loss": 0.3771, "step": 1670 }, { "epoch": 0.5797101449275363, "grad_norm": 0.5346298217773438, "learning_rate": 3.552449965493444e-05, "loss": 0.3543, "step": 1680 }, { "epoch": 0.583160800552105, "grad_norm": 0.44163742661476135, "learning_rate": 3.543823326432023e-05, "loss": 0.3439, "step": 1690 }, { "epoch": 0.5866114561766735, "grad_norm": 0.46102967858314514, "learning_rate": 3.5351966873706005e-05, "loss": 0.3433, "step": 1700 }, { "epoch": 0.5900621118012422, "grad_norm": 0.4693676233291626, "learning_rate": 3.526570048309179e-05, "loss": 0.3445, "step": 1710 }, { "epoch": 0.5935127674258109, "grad_norm": 0.4793757200241089, "learning_rate": 3.517943409247757e-05, "loss": 0.3594, "step": 1720 }, { "epoch": 0.5969634230503795, "grad_norm": 0.5180776119232178, "learning_rate": 3.509316770186335e-05, "loss": 0.3379, "step": 1730 }, { "epoch": 0.6004140786749482, "grad_norm": 0.48090022802352905, "learning_rate": 3.500690131124914e-05, "loss": 0.3528, "step": 1740 }, { "epoch": 0.6038647342995169, "grad_norm": 0.5032299757003784, "learning_rate": 3.492063492063492e-05, "loss": 0.35, "step": 1750 }, { "epoch": 0.6073153899240856, "grad_norm": 0.4861588478088379, "learning_rate": 3.483436853002071e-05, "loss": 0.3417, "step": 1760 }, { "epoch": 0.6107660455486542, "grad_norm": 0.43041691184043884, "learning_rate": 3.474810213940649e-05, "loss": 0.3456, "step": 1770 }, { "epoch": 0.6142167011732229, "grad_norm": 0.4854036569595337, "learning_rate": 3.4661835748792274e-05, "loss": 0.348, "step": 1780 }, { "epoch": 0.6176673567977916, "grad_norm": 0.4714268445968628, "learning_rate": 3.4575569358178056e-05, "loss": 0.3384, "step": 1790 }, { "epoch": 0.6211180124223602, "grad_norm": 0.4720168113708496, "learning_rate": 3.448930296756384e-05, "loss": 0.3437, "step": 1800 }, { "epoch": 0.6245686680469289, "grad_norm": 0.4411432147026062, "learning_rate": 3.440303657694962e-05, "loss": 0.352, "step": 1810 }, { "epoch": 0.6280193236714976, "grad_norm": 0.4829714000225067, "learning_rate": 3.4316770186335405e-05, "loss": 0.3246, "step": 1820 }, { "epoch": 0.6314699792960663, "grad_norm": 0.5485419034957886, "learning_rate": 3.423050379572119e-05, "loss": 0.3462, "step": 1830 }, { "epoch": 0.6349206349206349, "grad_norm": 0.43877512216567993, "learning_rate": 3.414423740510697e-05, "loss": 0.3441, "step": 1840 }, { "epoch": 0.6383712905452036, "grad_norm": 0.5296888947486877, "learning_rate": 3.405797101449276e-05, "loss": 0.3419, "step": 1850 }, { "epoch": 0.6418219461697723, "grad_norm": 0.4342154860496521, "learning_rate": 3.3971704623878536e-05, "loss": 0.3479, "step": 1860 }, { "epoch": 0.6452726017943409, "grad_norm": 0.5657359957695007, "learning_rate": 3.388543823326432e-05, "loss": 0.3461, "step": 1870 }, { "epoch": 0.6487232574189096, "grad_norm": 0.48667094111442566, "learning_rate": 3.379917184265011e-05, "loss": 0.3489, "step": 1880 }, { "epoch": 0.6521739130434783, "grad_norm": 0.5117901563644409, "learning_rate": 3.3712905452035884e-05, "loss": 0.3412, "step": 1890 }, { "epoch": 0.655624568668047, "grad_norm": 0.5122795104980469, "learning_rate": 3.3626639061421674e-05, "loss": 0.3464, "step": 1900 }, { "epoch": 0.6590752242926156, "grad_norm": 0.44816821813583374, "learning_rate": 3.3540372670807456e-05, "loss": 0.3404, "step": 1910 }, { "epoch": 0.6625258799171843, "grad_norm": 0.5073073506355286, "learning_rate": 3.345410628019324e-05, "loss": 0.3435, "step": 1920 }, { "epoch": 0.665976535541753, "grad_norm": 0.506574273109436, "learning_rate": 3.336783988957902e-05, "loss": 0.3419, "step": 1930 }, { "epoch": 0.6694271911663217, "grad_norm": 0.47071290016174316, "learning_rate": 3.3281573498964805e-05, "loss": 0.3357, "step": 1940 }, { "epoch": 0.6728778467908902, "grad_norm": 0.4610745310783386, "learning_rate": 3.319530710835059e-05, "loss": 0.3529, "step": 1950 }, { "epoch": 0.6763285024154589, "grad_norm": 0.496146023273468, "learning_rate": 3.310904071773637e-05, "loss": 0.3315, "step": 1960 }, { "epoch": 0.6797791580400276, "grad_norm": 0.5223260521888733, "learning_rate": 3.302277432712215e-05, "loss": 0.3409, "step": 1970 }, { "epoch": 0.6832298136645962, "grad_norm": 0.46770042181015015, "learning_rate": 3.2936507936507936e-05, "loss": 0.3323, "step": 1980 }, { "epoch": 0.6866804692891649, "grad_norm": 0.43604782223701477, "learning_rate": 3.2850241545893725e-05, "loss": 0.3395, "step": 1990 }, { "epoch": 0.6901311249137336, "grad_norm": 0.48399487137794495, "learning_rate": 3.27639751552795e-05, "loss": 0.338, "step": 2000 }, { "epoch": 0.6935817805383023, "grad_norm": 0.4736705720424652, "learning_rate": 3.267770876466529e-05, "loss": 0.3461, "step": 2010 }, { "epoch": 0.6970324361628709, "grad_norm": 0.42121419310569763, "learning_rate": 3.2591442374051074e-05, "loss": 0.3438, "step": 2020 }, { "epoch": 0.7004830917874396, "grad_norm": 0.4169461131095886, "learning_rate": 3.2505175983436856e-05, "loss": 0.335, "step": 2030 }, { "epoch": 0.7039337474120083, "grad_norm": 0.4755675792694092, "learning_rate": 3.241890959282264e-05, "loss": 0.3434, "step": 2040 }, { "epoch": 0.7073844030365769, "grad_norm": 0.4242471754550934, "learning_rate": 3.233264320220842e-05, "loss": 0.3409, "step": 2050 }, { "epoch": 0.7108350586611456, "grad_norm": 0.488776296377182, "learning_rate": 3.2246376811594205e-05, "loss": 0.3427, "step": 2060 }, { "epoch": 0.7142857142857143, "grad_norm": 0.45658791065216064, "learning_rate": 3.216011042097999e-05, "loss": 0.3405, "step": 2070 }, { "epoch": 0.717736369910283, "grad_norm": 0.5127744078636169, "learning_rate": 3.207384403036577e-05, "loss": 0.3377, "step": 2080 }, { "epoch": 0.7211870255348516, "grad_norm": 0.44740578532218933, "learning_rate": 3.198757763975155e-05, "loss": 0.3439, "step": 2090 }, { "epoch": 0.7246376811594203, "grad_norm": 0.49259352684020996, "learning_rate": 3.1901311249137336e-05, "loss": 0.3413, "step": 2100 }, { "epoch": 0.728088336783989, "grad_norm": 0.48841023445129395, "learning_rate": 3.181504485852312e-05, "loss": 0.3574, "step": 2110 }, { "epoch": 0.7315389924085576, "grad_norm": 0.4867888391017914, "learning_rate": 3.17287784679089e-05, "loss": 0.3462, "step": 2120 }, { "epoch": 0.7349896480331263, "grad_norm": 0.49781835079193115, "learning_rate": 3.164251207729469e-05, "loss": 0.3275, "step": 2130 }, { "epoch": 0.738440303657695, "grad_norm": 0.4472957253456116, "learning_rate": 3.155624568668047e-05, "loss": 0.3486, "step": 2140 }, { "epoch": 0.7418909592822637, "grad_norm": 0.47612202167510986, "learning_rate": 3.1469979296066256e-05, "loss": 0.3534, "step": 2150 }, { "epoch": 0.7453416149068323, "grad_norm": 0.4207702577114105, "learning_rate": 3.138371290545204e-05, "loss": 0.3426, "step": 2160 }, { "epoch": 0.748792270531401, "grad_norm": 0.49038171768188477, "learning_rate": 3.129744651483782e-05, "loss": 0.3371, "step": 2170 }, { "epoch": 0.7522429261559697, "grad_norm": 0.503973662853241, "learning_rate": 3.1211180124223605e-05, "loss": 0.3304, "step": 2180 }, { "epoch": 0.7556935817805382, "grad_norm": 0.3980702757835388, "learning_rate": 3.112491373360939e-05, "loss": 0.3374, "step": 2190 }, { "epoch": 0.759144237405107, "grad_norm": 0.4417726397514343, "learning_rate": 3.103864734299517e-05, "loss": 0.3359, "step": 2200 }, { "epoch": 0.7625948930296756, "grad_norm": 0.47529205679893494, "learning_rate": 3.095238095238095e-05, "loss": 0.334, "step": 2210 }, { "epoch": 0.7660455486542443, "grad_norm": 0.44546717405319214, "learning_rate": 3.0866114561766736e-05, "loss": 0.3338, "step": 2220 }, { "epoch": 0.7694962042788129, "grad_norm": 0.45016613602638245, "learning_rate": 3.077984817115252e-05, "loss": 0.3457, "step": 2230 }, { "epoch": 0.7729468599033816, "grad_norm": 0.48992106318473816, "learning_rate": 3.069358178053831e-05, "loss": 0.3372, "step": 2240 }, { "epoch": 0.7763975155279503, "grad_norm": 0.47713831067085266, "learning_rate": 3.0607315389924084e-05, "loss": 0.3491, "step": 2250 }, { "epoch": 0.779848171152519, "grad_norm": 0.4510743319988251, "learning_rate": 3.0521048999309874e-05, "loss": 0.3348, "step": 2260 }, { "epoch": 0.7832988267770876, "grad_norm": 0.43214091658592224, "learning_rate": 3.0434782608695656e-05, "loss": 0.3406, "step": 2270 }, { "epoch": 0.7867494824016563, "grad_norm": 0.4679895341396332, "learning_rate": 3.0348516218081436e-05, "loss": 0.3369, "step": 2280 }, { "epoch": 0.790200138026225, "grad_norm": 0.5620041489601135, "learning_rate": 3.0262249827467222e-05, "loss": 0.3399, "step": 2290 }, { "epoch": 0.7936507936507936, "grad_norm": 0.4891965091228485, "learning_rate": 3.0175983436853005e-05, "loss": 0.3323, "step": 2300 }, { "epoch": 0.7971014492753623, "grad_norm": 0.44797247648239136, "learning_rate": 3.0089717046238784e-05, "loss": 0.3369, "step": 2310 }, { "epoch": 0.800552104899931, "grad_norm": 0.4653630554676056, "learning_rate": 3.000345065562457e-05, "loss": 0.3304, "step": 2320 }, { "epoch": 0.8040027605244997, "grad_norm": 0.4232509434223175, "learning_rate": 2.991718426501035e-05, "loss": 0.3361, "step": 2330 }, { "epoch": 0.8074534161490683, "grad_norm": 0.42916807532310486, "learning_rate": 2.9830917874396136e-05, "loss": 0.338, "step": 2340 }, { "epoch": 0.810904071773637, "grad_norm": 0.44847798347473145, "learning_rate": 2.9744651483781922e-05, "loss": 0.3388, "step": 2350 }, { "epoch": 0.8143547273982057, "grad_norm": 0.41225308179855347, "learning_rate": 2.96583850931677e-05, "loss": 0.3352, "step": 2360 }, { "epoch": 0.8178053830227743, "grad_norm": 0.5075202584266663, "learning_rate": 2.9572118702553487e-05, "loss": 0.333, "step": 2370 }, { "epoch": 0.821256038647343, "grad_norm": 0.42286792397499084, "learning_rate": 2.948585231193927e-05, "loss": 0.3324, "step": 2380 }, { "epoch": 0.8247066942719117, "grad_norm": 0.4466201663017273, "learning_rate": 2.9399585921325053e-05, "loss": 0.3338, "step": 2390 }, { "epoch": 0.8281573498964804, "grad_norm": 0.3980891704559326, "learning_rate": 2.9313319530710836e-05, "loss": 0.3414, "step": 2400 }, { "epoch": 0.831608005521049, "grad_norm": 0.4684590697288513, "learning_rate": 2.9227053140096622e-05, "loss": 0.34, "step": 2410 }, { "epoch": 0.8350586611456177, "grad_norm": 0.43318429589271545, "learning_rate": 2.91407867494824e-05, "loss": 0.3221, "step": 2420 }, { "epoch": 0.8385093167701864, "grad_norm": 0.4576144218444824, "learning_rate": 2.9054520358868187e-05, "loss": 0.341, "step": 2430 }, { "epoch": 0.841959972394755, "grad_norm": 0.4950040280818939, "learning_rate": 2.8968253968253974e-05, "loss": 0.3331, "step": 2440 }, { "epoch": 0.8454106280193237, "grad_norm": 0.5012326836585999, "learning_rate": 2.8881987577639753e-05, "loss": 0.3269, "step": 2450 }, { "epoch": 0.8488612836438924, "grad_norm": 0.4127346873283386, "learning_rate": 2.8795721187025536e-05, "loss": 0.3313, "step": 2460 }, { "epoch": 0.852311939268461, "grad_norm": 0.4913102388381958, "learning_rate": 2.870945479641132e-05, "loss": 0.3329, "step": 2470 }, { "epoch": 0.8557625948930296, "grad_norm": 0.4684026837348938, "learning_rate": 2.86231884057971e-05, "loss": 0.3374, "step": 2480 }, { "epoch": 0.8592132505175983, "grad_norm": 0.4605260491371155, "learning_rate": 2.8536922015182887e-05, "loss": 0.3398, "step": 2490 }, { "epoch": 0.862663906142167, "grad_norm": 0.4554629325866699, "learning_rate": 2.8450655624568667e-05, "loss": 0.333, "step": 2500 }, { "epoch": 0.8661145617667356, "grad_norm": 0.4555531442165375, "learning_rate": 2.8364389233954453e-05, "loss": 0.3503, "step": 2510 }, { "epoch": 0.8695652173913043, "grad_norm": 0.46253401041030884, "learning_rate": 2.827812284334024e-05, "loss": 0.3282, "step": 2520 }, { "epoch": 0.873015873015873, "grad_norm": 0.523013174533844, "learning_rate": 2.819185645272602e-05, "loss": 0.336, "step": 2530 }, { "epoch": 0.8764665286404417, "grad_norm": 0.4696257710456848, "learning_rate": 2.8105590062111805e-05, "loss": 0.346, "step": 2540 }, { "epoch": 0.8799171842650103, "grad_norm": 0.44945546984672546, "learning_rate": 2.8019323671497587e-05, "loss": 0.3297, "step": 2550 }, { "epoch": 0.883367839889579, "grad_norm": 0.40458500385284424, "learning_rate": 2.7933057280883367e-05, "loss": 0.3391, "step": 2560 }, { "epoch": 0.8868184955141477, "grad_norm": 0.4970335364341736, "learning_rate": 2.7846790890269153e-05, "loss": 0.342, "step": 2570 }, { "epoch": 0.8902691511387164, "grad_norm": 0.4015578627586365, "learning_rate": 2.776052449965494e-05, "loss": 0.33, "step": 2580 }, { "epoch": 0.893719806763285, "grad_norm": 0.482063889503479, "learning_rate": 2.767425810904072e-05, "loss": 0.3366, "step": 2590 }, { "epoch": 0.8971704623878537, "grad_norm": 0.3847281336784363, "learning_rate": 2.7587991718426505e-05, "loss": 0.3528, "step": 2600 }, { "epoch": 0.9006211180124224, "grad_norm": 0.4189392626285553, "learning_rate": 2.7501725327812284e-05, "loss": 0.3346, "step": 2610 }, { "epoch": 0.904071773636991, "grad_norm": 0.44489413499832153, "learning_rate": 2.741545893719807e-05, "loss": 0.3279, "step": 2620 }, { "epoch": 0.9075224292615597, "grad_norm": 0.43097779154777527, "learning_rate": 2.7329192546583853e-05, "loss": 0.3318, "step": 2630 }, { "epoch": 0.9109730848861284, "grad_norm": 0.4077935516834259, "learning_rate": 2.7242926155969632e-05, "loss": 0.3248, "step": 2640 }, { "epoch": 0.9144237405106971, "grad_norm": 0.39787349104881287, "learning_rate": 2.715665976535542e-05, "loss": 0.3398, "step": 2650 }, { "epoch": 0.9178743961352657, "grad_norm": 0.4495849609375, "learning_rate": 2.7070393374741205e-05, "loss": 0.3406, "step": 2660 }, { "epoch": 0.9213250517598344, "grad_norm": 0.4294304847717285, "learning_rate": 2.6984126984126984e-05, "loss": 0.3375, "step": 2670 }, { "epoch": 0.9247757073844031, "grad_norm": 0.40841785073280334, "learning_rate": 2.689786059351277e-05, "loss": 0.3198, "step": 2680 }, { "epoch": 0.9282263630089717, "grad_norm": 0.44023633003234863, "learning_rate": 2.6811594202898553e-05, "loss": 0.3344, "step": 2690 }, { "epoch": 0.9316770186335404, "grad_norm": 0.4349792003631592, "learning_rate": 2.6725327812284336e-05, "loss": 0.3357, "step": 2700 }, { "epoch": 0.935127674258109, "grad_norm": 0.4780796766281128, "learning_rate": 2.663906142167012e-05, "loss": 0.3297, "step": 2710 }, { "epoch": 0.9385783298826778, "grad_norm": 0.45114386081695557, "learning_rate": 2.6552795031055898e-05, "loss": 0.3289, "step": 2720 }, { "epoch": 0.9420289855072463, "grad_norm": 0.4433158040046692, "learning_rate": 2.6466528640441684e-05, "loss": 0.3241, "step": 2730 }, { "epoch": 0.945479641131815, "grad_norm": 0.4354463219642639, "learning_rate": 2.638026224982747e-05, "loss": 0.3249, "step": 2740 }, { "epoch": 0.9489302967563837, "grad_norm": 0.4203226864337921, "learning_rate": 2.629399585921325e-05, "loss": 0.329, "step": 2750 }, { "epoch": 0.9523809523809523, "grad_norm": 0.42780405282974243, "learning_rate": 2.6207729468599036e-05, "loss": 0.337, "step": 2760 }, { "epoch": 0.955831608005521, "grad_norm": 0.44693320989608765, "learning_rate": 2.6121463077984822e-05, "loss": 0.3307, "step": 2770 }, { "epoch": 0.9592822636300897, "grad_norm": 0.3886381983757019, "learning_rate": 2.60351966873706e-05, "loss": 0.3332, "step": 2780 }, { "epoch": 0.9627329192546584, "grad_norm": 0.40549805760383606, "learning_rate": 2.5948930296756384e-05, "loss": 0.3259, "step": 2790 }, { "epoch": 0.966183574879227, "grad_norm": 0.4425898492336273, "learning_rate": 2.586266390614217e-05, "loss": 0.3337, "step": 2800 }, { "epoch": 0.9696342305037957, "grad_norm": 0.4614519774913788, "learning_rate": 2.577639751552795e-05, "loss": 0.3223, "step": 2810 }, { "epoch": 0.9730848861283644, "grad_norm": 0.47300979495048523, "learning_rate": 2.5690131124913736e-05, "loss": 0.3382, "step": 2820 }, { "epoch": 0.9765355417529331, "grad_norm": 0.39775508642196655, "learning_rate": 2.5603864734299522e-05, "loss": 0.335, "step": 2830 }, { "epoch": 0.9799861973775017, "grad_norm": 0.4713541865348816, "learning_rate": 2.55175983436853e-05, "loss": 0.3312, "step": 2840 }, { "epoch": 0.9834368530020704, "grad_norm": 0.41885653138160706, "learning_rate": 2.5431331953071087e-05, "loss": 0.3256, "step": 2850 }, { "epoch": 0.9868875086266391, "grad_norm": 0.41549861431121826, "learning_rate": 2.5345065562456867e-05, "loss": 0.3389, "step": 2860 }, { "epoch": 0.9903381642512077, "grad_norm": 0.4274648427963257, "learning_rate": 2.525879917184265e-05, "loss": 0.3285, "step": 2870 }, { "epoch": 0.9937888198757764, "grad_norm": 0.48823943734169006, "learning_rate": 2.5172532781228436e-05, "loss": 0.3225, "step": 2880 }, { "epoch": 0.9972394755003451, "grad_norm": 0.4019075930118561, "learning_rate": 2.5086266390614215e-05, "loss": 0.3258, "step": 2890 }, { "epoch": 1.0006901311249137, "grad_norm": 0.40041089057922363, "learning_rate": 2.5e-05, "loss": 0.3275, "step": 2900 }, { "epoch": 1.0041407867494825, "grad_norm": 0.43522876501083374, "learning_rate": 2.4913733609385784e-05, "loss": 0.3229, "step": 2910 }, { "epoch": 1.007591442374051, "grad_norm": 0.4160279333591461, "learning_rate": 2.482746721877157e-05, "loss": 0.3241, "step": 2920 }, { "epoch": 1.0110420979986197, "grad_norm": 0.4505654573440552, "learning_rate": 2.4741200828157353e-05, "loss": 0.3302, "step": 2930 }, { "epoch": 1.0144927536231885, "grad_norm": 0.4337722957134247, "learning_rate": 2.4654934437543136e-05, "loss": 0.3232, "step": 2940 }, { "epoch": 1.017943409247757, "grad_norm": 0.44694915413856506, "learning_rate": 2.4568668046928915e-05, "loss": 0.3167, "step": 2950 }, { "epoch": 1.0213940648723256, "grad_norm": 0.4304366707801819, "learning_rate": 2.44824016563147e-05, "loss": 0.3247, "step": 2960 }, { "epoch": 1.0248447204968945, "grad_norm": 0.3924005329608917, "learning_rate": 2.4396135265700484e-05, "loss": 0.3203, "step": 2970 }, { "epoch": 1.028295376121463, "grad_norm": 0.4276975393295288, "learning_rate": 2.4309868875086267e-05, "loss": 0.3222, "step": 2980 }, { "epoch": 1.0317460317460316, "grad_norm": 0.41446569561958313, "learning_rate": 2.4223602484472053e-05, "loss": 0.3219, "step": 2990 }, { "epoch": 1.0351966873706004, "grad_norm": 0.4318532347679138, "learning_rate": 2.4137336093857836e-05, "loss": 0.3315, "step": 3000 }, { "epoch": 1.038647342995169, "grad_norm": 0.48855313658714294, "learning_rate": 2.405106970324362e-05, "loss": 0.3213, "step": 3010 }, { "epoch": 1.0420979986197378, "grad_norm": 0.46925896406173706, "learning_rate": 2.39648033126294e-05, "loss": 0.3296, "step": 3020 }, { "epoch": 1.0455486542443064, "grad_norm": 0.435977041721344, "learning_rate": 2.3878536922015184e-05, "loss": 0.3159, "step": 3030 }, { "epoch": 1.048999309868875, "grad_norm": 0.5048639178276062, "learning_rate": 2.3792270531400967e-05, "loss": 0.3195, "step": 3040 }, { "epoch": 1.0524499654934438, "grad_norm": 0.43628114461898804, "learning_rate": 2.370600414078675e-05, "loss": 0.3288, "step": 3050 }, { "epoch": 1.0559006211180124, "grad_norm": 0.49825018644332886, "learning_rate": 2.3619737750172532e-05, "loss": 0.3182, "step": 3060 }, { "epoch": 1.059351276742581, "grad_norm": 0.4279234707355499, "learning_rate": 2.353347135955832e-05, "loss": 0.3196, "step": 3070 }, { "epoch": 1.0628019323671498, "grad_norm": 0.45079633593559265, "learning_rate": 2.34472049689441e-05, "loss": 0.3252, "step": 3080 }, { "epoch": 1.0662525879917184, "grad_norm": 0.43655019998550415, "learning_rate": 2.3360938578329884e-05, "loss": 0.3163, "step": 3090 }, { "epoch": 1.069703243616287, "grad_norm": 0.4613361954689026, "learning_rate": 2.3274672187715667e-05, "loss": 0.3281, "step": 3100 }, { "epoch": 1.0731538992408558, "grad_norm": 0.4135753810405731, "learning_rate": 2.318840579710145e-05, "loss": 0.3157, "step": 3110 }, { "epoch": 1.0766045548654244, "grad_norm": 0.4026079773902893, "learning_rate": 2.3102139406487232e-05, "loss": 0.3188, "step": 3120 }, { "epoch": 1.0800552104899932, "grad_norm": 0.4352892339229584, "learning_rate": 2.3015873015873015e-05, "loss": 0.3104, "step": 3130 }, { "epoch": 1.0835058661145618, "grad_norm": 0.3982253074645996, "learning_rate": 2.29296066252588e-05, "loss": 0.3245, "step": 3140 }, { "epoch": 1.0869565217391304, "grad_norm": 0.3897712230682373, "learning_rate": 2.2843340234644584e-05, "loss": 0.3136, "step": 3150 }, { "epoch": 1.0904071773636992, "grad_norm": 0.37550920248031616, "learning_rate": 2.2757073844030367e-05, "loss": 0.3256, "step": 3160 }, { "epoch": 1.0938578329882678, "grad_norm": 0.41109180450439453, "learning_rate": 2.2670807453416153e-05, "loss": 0.317, "step": 3170 }, { "epoch": 1.0973084886128364, "grad_norm": 0.46931174397468567, "learning_rate": 2.2584541062801932e-05, "loss": 0.318, "step": 3180 }, { "epoch": 1.1007591442374052, "grad_norm": 0.41200053691864014, "learning_rate": 2.2498274672187715e-05, "loss": 0.3292, "step": 3190 }, { "epoch": 1.1042097998619738, "grad_norm": 0.45627737045288086, "learning_rate": 2.2412008281573498e-05, "loss": 0.3235, "step": 3200 }, { "epoch": 1.1076604554865424, "grad_norm": 0.4548512101173401, "learning_rate": 2.2325741890959284e-05, "loss": 0.3239, "step": 3210 }, { "epoch": 1.1111111111111112, "grad_norm": 0.37763795256614685, "learning_rate": 2.2239475500345067e-05, "loss": 0.3183, "step": 3220 }, { "epoch": 1.1145617667356797, "grad_norm": 0.4083724319934845, "learning_rate": 2.215320910973085e-05, "loss": 0.32, "step": 3230 }, { "epoch": 1.1180124223602483, "grad_norm": 0.3961089551448822, "learning_rate": 2.2066942719116636e-05, "loss": 0.309, "step": 3240 }, { "epoch": 1.1214630779848171, "grad_norm": 0.44931307435035706, "learning_rate": 2.198067632850242e-05, "loss": 0.312, "step": 3250 }, { "epoch": 1.1249137336093857, "grad_norm": 0.360643595457077, "learning_rate": 2.1894409937888198e-05, "loss": 0.3279, "step": 3260 }, { "epoch": 1.1283643892339545, "grad_norm": 0.37495937943458557, "learning_rate": 2.180814354727398e-05, "loss": 0.3123, "step": 3270 }, { "epoch": 1.1318150448585231, "grad_norm": 0.3907097280025482, "learning_rate": 2.1721877156659767e-05, "loss": 0.3256, "step": 3280 }, { "epoch": 1.1352657004830917, "grad_norm": 0.3720877468585968, "learning_rate": 2.163561076604555e-05, "loss": 0.329, "step": 3290 }, { "epoch": 1.1387163561076605, "grad_norm": 0.4681912660598755, "learning_rate": 2.1549344375431332e-05, "loss": 0.3267, "step": 3300 }, { "epoch": 1.1421670117322291, "grad_norm": 0.5137665271759033, "learning_rate": 2.146307798481712e-05, "loss": 0.3228, "step": 3310 }, { "epoch": 1.1456176673567977, "grad_norm": 0.41128090023994446, "learning_rate": 2.13768115942029e-05, "loss": 0.3209, "step": 3320 }, { "epoch": 1.1490683229813665, "grad_norm": 0.3766096532344818, "learning_rate": 2.1290545203588684e-05, "loss": 0.3228, "step": 3330 }, { "epoch": 1.152518978605935, "grad_norm": 0.37945806980133057, "learning_rate": 2.1204278812974467e-05, "loss": 0.3179, "step": 3340 }, { "epoch": 1.155969634230504, "grad_norm": 0.46118223667144775, "learning_rate": 2.111801242236025e-05, "loss": 0.3281, "step": 3350 }, { "epoch": 1.1594202898550725, "grad_norm": 0.3962058126926422, "learning_rate": 2.1031746031746032e-05, "loss": 0.3263, "step": 3360 }, { "epoch": 1.162870945479641, "grad_norm": 0.4236689805984497, "learning_rate": 2.0945479641131815e-05, "loss": 0.3198, "step": 3370 }, { "epoch": 1.16632160110421, "grad_norm": 0.39470767974853516, "learning_rate": 2.0859213250517598e-05, "loss": 0.3177, "step": 3380 }, { "epoch": 1.1697722567287785, "grad_norm": 0.3914918005466461, "learning_rate": 2.0772946859903384e-05, "loss": 0.3149, "step": 3390 }, { "epoch": 1.173222912353347, "grad_norm": 0.4191271960735321, "learning_rate": 2.0686680469289167e-05, "loss": 0.3234, "step": 3400 }, { "epoch": 1.176673567977916, "grad_norm": 0.40341272950172424, "learning_rate": 2.060041407867495e-05, "loss": 0.3314, "step": 3410 }, { "epoch": 1.1801242236024845, "grad_norm": 0.4099895656108856, "learning_rate": 2.0514147688060732e-05, "loss": 0.3251, "step": 3420 }, { "epoch": 1.183574879227053, "grad_norm": 0.4049317240715027, "learning_rate": 2.0427881297446515e-05, "loss": 0.3151, "step": 3430 }, { "epoch": 1.1870255348516219, "grad_norm": 0.4270973205566406, "learning_rate": 2.0341614906832298e-05, "loss": 0.3314, "step": 3440 }, { "epoch": 1.1904761904761905, "grad_norm": 0.40859755873680115, "learning_rate": 2.025534851621808e-05, "loss": 0.3204, "step": 3450 }, { "epoch": 1.193926846100759, "grad_norm": 0.4711279273033142, "learning_rate": 2.0169082125603867e-05, "loss": 0.3175, "step": 3460 }, { "epoch": 1.1973775017253279, "grad_norm": 0.458126038312912, "learning_rate": 2.008281573498965e-05, "loss": 0.3215, "step": 3470 }, { "epoch": 1.2008281573498965, "grad_norm": 0.39996790885925293, "learning_rate": 1.9996549344375432e-05, "loss": 0.3231, "step": 3480 }, { "epoch": 1.204278812974465, "grad_norm": 0.43717774748802185, "learning_rate": 1.991028295376122e-05, "loss": 0.3096, "step": 3490 }, { "epoch": 1.2077294685990339, "grad_norm": 0.4430061876773834, "learning_rate": 1.9824016563146998e-05, "loss": 0.3145, "step": 3500 }, { "epoch": 1.2111801242236024, "grad_norm": 0.4095977544784546, "learning_rate": 1.973775017253278e-05, "loss": 0.3203, "step": 3510 }, { "epoch": 1.2146307798481712, "grad_norm": 0.3994373679161072, "learning_rate": 1.9651483781918563e-05, "loss": 0.3237, "step": 3520 }, { "epoch": 1.2180814354727398, "grad_norm": 0.39703670144081116, "learning_rate": 1.956521739130435e-05, "loss": 0.323, "step": 3530 }, { "epoch": 1.2215320910973084, "grad_norm": 0.49238279461860657, "learning_rate": 1.9478951000690132e-05, "loss": 0.3201, "step": 3540 }, { "epoch": 1.2249827467218772, "grad_norm": 0.4226069450378418, "learning_rate": 1.9392684610075915e-05, "loss": 0.3182, "step": 3550 }, { "epoch": 1.2284334023464458, "grad_norm": 0.3932894170284271, "learning_rate": 1.93064182194617e-05, "loss": 0.3076, "step": 3560 }, { "epoch": 1.2318840579710144, "grad_norm": 0.41728973388671875, "learning_rate": 1.9220151828847484e-05, "loss": 0.3146, "step": 3570 }, { "epoch": 1.2353347135955832, "grad_norm": 0.37457820773124695, "learning_rate": 1.9133885438233263e-05, "loss": 0.3214, "step": 3580 }, { "epoch": 1.2387853692201518, "grad_norm": 0.4085715711116791, "learning_rate": 1.9047619047619046e-05, "loss": 0.3136, "step": 3590 }, { "epoch": 1.2422360248447206, "grad_norm": 0.3940805196762085, "learning_rate": 1.8961352657004832e-05, "loss": 0.319, "step": 3600 }, { "epoch": 1.2456866804692892, "grad_norm": 0.43771472573280334, "learning_rate": 1.8875086266390615e-05, "loss": 0.3206, "step": 3610 }, { "epoch": 1.2491373360938578, "grad_norm": 0.43656811118125916, "learning_rate": 1.8788819875776398e-05, "loss": 0.3143, "step": 3620 }, { "epoch": 1.2525879917184266, "grad_norm": 0.44114479422569275, "learning_rate": 1.8702553485162184e-05, "loss": 0.3213, "step": 3630 }, { "epoch": 1.2560386473429952, "grad_norm": 0.5253772139549255, "learning_rate": 1.8616287094547967e-05, "loss": 0.3144, "step": 3640 }, { "epoch": 1.2594893029675638, "grad_norm": 0.4206532835960388, "learning_rate": 1.853002070393375e-05, "loss": 0.3111, "step": 3650 }, { "epoch": 1.2629399585921326, "grad_norm": 0.4311043620109558, "learning_rate": 1.8443754313319532e-05, "loss": 0.3158, "step": 3660 }, { "epoch": 1.2663906142167012, "grad_norm": 0.41564512252807617, "learning_rate": 1.8357487922705315e-05, "loss": 0.3353, "step": 3670 }, { "epoch": 1.2698412698412698, "grad_norm": 0.43433383107185364, "learning_rate": 1.8271221532091098e-05, "loss": 0.3224, "step": 3680 }, { "epoch": 1.2732919254658386, "grad_norm": 0.436788946390152, "learning_rate": 1.818495514147688e-05, "loss": 0.3232, "step": 3690 }, { "epoch": 1.2767425810904072, "grad_norm": 0.42806386947631836, "learning_rate": 1.8098688750862663e-05, "loss": 0.3124, "step": 3700 }, { "epoch": 1.2801932367149758, "grad_norm": 0.40641355514526367, "learning_rate": 1.801242236024845e-05, "loss": 0.3203, "step": 3710 }, { "epoch": 1.2836438923395446, "grad_norm": 0.418039470911026, "learning_rate": 1.7926155969634232e-05, "loss": 0.3125, "step": 3720 }, { "epoch": 1.2870945479641132, "grad_norm": 0.41426384449005127, "learning_rate": 1.7839889579020015e-05, "loss": 0.3225, "step": 3730 }, { "epoch": 1.2905452035886817, "grad_norm": 0.40283921360969543, "learning_rate": 1.7753623188405798e-05, "loss": 0.3084, "step": 3740 }, { "epoch": 1.2939958592132506, "grad_norm": 0.42750898003578186, "learning_rate": 1.766735679779158e-05, "loss": 0.3135, "step": 3750 }, { "epoch": 1.2974465148378191, "grad_norm": 0.4350372552871704, "learning_rate": 1.7581090407177363e-05, "loss": 0.3161, "step": 3760 }, { "epoch": 1.3008971704623877, "grad_norm": 0.3901418447494507, "learning_rate": 1.7494824016563146e-05, "loss": 0.3192, "step": 3770 }, { "epoch": 1.3043478260869565, "grad_norm": 0.4102122187614441, "learning_rate": 1.7408557625948932e-05, "loss": 0.3168, "step": 3780 }, { "epoch": 1.3077984817115251, "grad_norm": 0.40640777349472046, "learning_rate": 1.7322291235334715e-05, "loss": 0.3288, "step": 3790 }, { "epoch": 1.311249137336094, "grad_norm": 0.39230260252952576, "learning_rate": 1.7236024844720498e-05, "loss": 0.3179, "step": 3800 }, { "epoch": 1.3146997929606625, "grad_norm": 0.462086945772171, "learning_rate": 1.714975845410628e-05, "loss": 0.31, "step": 3810 }, { "epoch": 1.3181504485852311, "grad_norm": 0.4115633964538574, "learning_rate": 1.7063492063492063e-05, "loss": 0.324, "step": 3820 }, { "epoch": 1.3216011042098, "grad_norm": 0.37690791487693787, "learning_rate": 1.6977225672877846e-05, "loss": 0.3082, "step": 3830 }, { "epoch": 1.3250517598343685, "grad_norm": 0.4259274899959564, "learning_rate": 1.689095928226363e-05, "loss": 0.3205, "step": 3840 }, { "epoch": 1.3285024154589373, "grad_norm": 0.39965760707855225, "learning_rate": 1.6804692891649415e-05, "loss": 0.3288, "step": 3850 }, { "epoch": 1.331953071083506, "grad_norm": 0.4190365970134735, "learning_rate": 1.6718426501035198e-05, "loss": 0.3144, "step": 3860 }, { "epoch": 1.3354037267080745, "grad_norm": 0.3663720488548279, "learning_rate": 1.663216011042098e-05, "loss": 0.3167, "step": 3870 }, { "epoch": 1.3388543823326433, "grad_norm": 0.4190429747104645, "learning_rate": 1.6545893719806767e-05, "loss": 0.3091, "step": 3880 }, { "epoch": 1.342305037957212, "grad_norm": 0.42641833424568176, "learning_rate": 1.645962732919255e-05, "loss": 0.3108, "step": 3890 }, { "epoch": 1.3457556935817805, "grad_norm": 0.39976075291633606, "learning_rate": 1.637336093857833e-05, "loss": 0.3066, "step": 3900 }, { "epoch": 1.3492063492063493, "grad_norm": 0.41375914216041565, "learning_rate": 1.628709454796411e-05, "loss": 0.3193, "step": 3910 }, { "epoch": 1.3526570048309179, "grad_norm": 0.3997116982936859, "learning_rate": 1.6200828157349898e-05, "loss": 0.3126, "step": 3920 }, { "epoch": 1.3561076604554865, "grad_norm": 0.3594307601451874, "learning_rate": 1.611456176673568e-05, "loss": 0.315, "step": 3930 }, { "epoch": 1.3595583160800553, "grad_norm": 0.3919355273246765, "learning_rate": 1.6028295376121463e-05, "loss": 0.313, "step": 3940 }, { "epoch": 1.3630089717046239, "grad_norm": 0.436056911945343, "learning_rate": 1.5942028985507246e-05, "loss": 0.3185, "step": 3950 }, { "epoch": 1.3664596273291925, "grad_norm": 0.41378381848335266, "learning_rate": 1.5855762594893032e-05, "loss": 0.3174, "step": 3960 }, { "epoch": 1.3699102829537613, "grad_norm": 0.4355195164680481, "learning_rate": 1.5769496204278815e-05, "loss": 0.3227, "step": 3970 }, { "epoch": 1.3733609385783299, "grad_norm": 0.41318216919898987, "learning_rate": 1.5683229813664594e-05, "loss": 0.3169, "step": 3980 }, { "epoch": 1.3768115942028984, "grad_norm": 0.47554564476013184, "learning_rate": 1.559696342305038e-05, "loss": 0.3034, "step": 3990 }, { "epoch": 1.3802622498274673, "grad_norm": 0.4184069037437439, "learning_rate": 1.5510697032436163e-05, "loss": 0.3205, "step": 4000 }, { "epoch": 1.3837129054520358, "grad_norm": 0.5330416560173035, "learning_rate": 1.5424430641821946e-05, "loss": 0.3282, "step": 4010 }, { "epoch": 1.3871635610766044, "grad_norm": 0.43783238530158997, "learning_rate": 1.533816425120773e-05, "loss": 0.3246, "step": 4020 }, { "epoch": 1.3906142167011732, "grad_norm": 0.41294294595718384, "learning_rate": 1.5251897860593515e-05, "loss": 0.3202, "step": 4030 }, { "epoch": 1.3940648723257418, "grad_norm": 0.34258291125297546, "learning_rate": 1.5165631469979296e-05, "loss": 0.3195, "step": 4040 }, { "epoch": 1.3975155279503104, "grad_norm": 0.3889847695827484, "learning_rate": 1.5079365079365079e-05, "loss": 0.3142, "step": 4050 }, { "epoch": 1.4009661835748792, "grad_norm": 0.3962327539920807, "learning_rate": 1.4993098688750865e-05, "loss": 0.3208, "step": 4060 }, { "epoch": 1.4044168391994478, "grad_norm": 0.4308435022830963, "learning_rate": 1.4906832298136648e-05, "loss": 0.3176, "step": 4070 }, { "epoch": 1.4078674948240166, "grad_norm": 0.4349764585494995, "learning_rate": 1.4820565907522429e-05, "loss": 0.3258, "step": 4080 }, { "epoch": 1.4113181504485852, "grad_norm": 0.44505828619003296, "learning_rate": 1.4734299516908212e-05, "loss": 0.3211, "step": 4090 }, { "epoch": 1.414768806073154, "grad_norm": 0.4607914984226227, "learning_rate": 1.4648033126293998e-05, "loss": 0.3166, "step": 4100 }, { "epoch": 1.4182194616977226, "grad_norm": 0.40675053000450134, "learning_rate": 1.456176673567978e-05, "loss": 0.3171, "step": 4110 }, { "epoch": 1.4216701173222912, "grad_norm": 0.36485347151756287, "learning_rate": 1.4475500345065562e-05, "loss": 0.3143, "step": 4120 }, { "epoch": 1.42512077294686, "grad_norm": 0.39989733695983887, "learning_rate": 1.4389233954451348e-05, "loss": 0.3176, "step": 4130 }, { "epoch": 1.4285714285714286, "grad_norm": 0.4142070412635803, "learning_rate": 1.430296756383713e-05, "loss": 0.3232, "step": 4140 }, { "epoch": 1.4320220841959972, "grad_norm": 0.41662535071372986, "learning_rate": 1.4216701173222913e-05, "loss": 0.3154, "step": 4150 }, { "epoch": 1.435472739820566, "grad_norm": 0.39152923226356506, "learning_rate": 1.4130434782608694e-05, "loss": 0.3141, "step": 4160 }, { "epoch": 1.4389233954451346, "grad_norm": 0.41523277759552, "learning_rate": 1.404416839199448e-05, "loss": 0.3175, "step": 4170 }, { "epoch": 1.4423740510697032, "grad_norm": 0.40196311473846436, "learning_rate": 1.3957902001380263e-05, "loss": 0.3161, "step": 4180 }, { "epoch": 1.445824706694272, "grad_norm": 0.441524863243103, "learning_rate": 1.3871635610766046e-05, "loss": 0.3147, "step": 4190 }, { "epoch": 1.4492753623188406, "grad_norm": 0.4273381531238556, "learning_rate": 1.378536922015183e-05, "loss": 0.3195, "step": 4200 }, { "epoch": 1.4527260179434092, "grad_norm": 0.41128963232040405, "learning_rate": 1.3699102829537613e-05, "loss": 0.3301, "step": 4210 }, { "epoch": 1.456176673567978, "grad_norm": 0.39311736822128296, "learning_rate": 1.3612836438923396e-05, "loss": 0.3242, "step": 4220 }, { "epoch": 1.4596273291925466, "grad_norm": 0.4288855493068695, "learning_rate": 1.3526570048309179e-05, "loss": 0.3328, "step": 4230 }, { "epoch": 1.4630779848171152, "grad_norm": 0.38766637444496155, "learning_rate": 1.3440303657694963e-05, "loss": 0.3178, "step": 4240 }, { "epoch": 1.466528640441684, "grad_norm": 0.48711350560188293, "learning_rate": 1.3354037267080746e-05, "loss": 0.3231, "step": 4250 }, { "epoch": 1.4699792960662525, "grad_norm": 0.3988294303417206, "learning_rate": 1.3267770876466529e-05, "loss": 0.314, "step": 4260 }, { "epoch": 1.4734299516908211, "grad_norm": 0.46785667538642883, "learning_rate": 1.3181504485852312e-05, "loss": 0.3071, "step": 4270 }, { "epoch": 1.47688060731539, "grad_norm": 0.413526326417923, "learning_rate": 1.3095238095238096e-05, "loss": 0.3253, "step": 4280 }, { "epoch": 1.4803312629399585, "grad_norm": 0.38375306129455566, "learning_rate": 1.3008971704623879e-05, "loss": 0.3172, "step": 4290 }, { "epoch": 1.4837819185645271, "grad_norm": 0.4046788513660431, "learning_rate": 1.2922705314009662e-05, "loss": 0.3189, "step": 4300 }, { "epoch": 1.487232574189096, "grad_norm": 0.4205913245677948, "learning_rate": 1.2836438923395446e-05, "loss": 0.3103, "step": 4310 }, { "epoch": 1.4906832298136645, "grad_norm": 0.36907699704170227, "learning_rate": 1.2750172532781229e-05, "loss": 0.3134, "step": 4320 }, { "epoch": 1.4941338854382333, "grad_norm": 0.48446428775787354, "learning_rate": 1.2663906142167012e-05, "loss": 0.3148, "step": 4330 }, { "epoch": 1.497584541062802, "grad_norm": 0.4103284776210785, "learning_rate": 1.2577639751552794e-05, "loss": 0.3091, "step": 4340 }, { "epoch": 1.5010351966873707, "grad_norm": 0.44808661937713623, "learning_rate": 1.2491373360938579e-05, "loss": 0.3182, "step": 4350 }, { "epoch": 1.5044858523119393, "grad_norm": 0.3726388216018677, "learning_rate": 1.2405106970324362e-05, "loss": 0.3189, "step": 4360 }, { "epoch": 1.507936507936508, "grad_norm": 0.41934582591056824, "learning_rate": 1.2318840579710146e-05, "loss": 0.3045, "step": 4370 }, { "epoch": 1.5113871635610767, "grad_norm": 0.4093073904514313, "learning_rate": 1.2232574189095929e-05, "loss": 0.3122, "step": 4380 }, { "epoch": 1.5148378191856453, "grad_norm": 0.4008806049823761, "learning_rate": 1.2146307798481713e-05, "loss": 0.3199, "step": 4390 }, { "epoch": 1.518288474810214, "grad_norm": 0.4053995609283447, "learning_rate": 1.2060041407867494e-05, "loss": 0.3151, "step": 4400 }, { "epoch": 1.5217391304347827, "grad_norm": 0.3917613923549652, "learning_rate": 1.1973775017253279e-05, "loss": 0.3179, "step": 4410 }, { "epoch": 1.5251897860593513, "grad_norm": 0.40496399998664856, "learning_rate": 1.1887508626639062e-05, "loss": 0.3139, "step": 4420 }, { "epoch": 1.5286404416839199, "grad_norm": 0.46660828590393066, "learning_rate": 1.1801242236024846e-05, "loss": 0.3246, "step": 4430 }, { "epoch": 1.5320910973084887, "grad_norm": 0.4034900963306427, "learning_rate": 1.1714975845410629e-05, "loss": 0.3151, "step": 4440 }, { "epoch": 1.5355417529330573, "grad_norm": 0.4394216537475586, "learning_rate": 1.1628709454796412e-05, "loss": 0.3139, "step": 4450 }, { "epoch": 1.5389924085576259, "grad_norm": 0.390280157327652, "learning_rate": 1.1542443064182196e-05, "loss": 0.3066, "step": 4460 }, { "epoch": 1.5424430641821947, "grad_norm": 0.4169969856739044, "learning_rate": 1.1456176673567979e-05, "loss": 0.306, "step": 4470 }, { "epoch": 1.5458937198067633, "grad_norm": 0.41443514823913574, "learning_rate": 1.1369910282953762e-05, "loss": 0.3162, "step": 4480 }, { "epoch": 1.5493443754313319, "grad_norm": 0.3956752121448517, "learning_rate": 1.1283643892339544e-05, "loss": 0.3119, "step": 4490 }, { "epoch": 1.5527950310559007, "grad_norm": 0.5450183153152466, "learning_rate": 1.1197377501725329e-05, "loss": 0.3179, "step": 4500 }, { "epoch": 1.5562456866804693, "grad_norm": 0.3975428640842438, "learning_rate": 1.1111111111111112e-05, "loss": 0.3236, "step": 4510 }, { "epoch": 1.5596963423050378, "grad_norm": 0.40076789259910583, "learning_rate": 1.1024844720496894e-05, "loss": 0.3149, "step": 4520 }, { "epoch": 1.5631469979296067, "grad_norm": 0.3667759597301483, "learning_rate": 1.0938578329882679e-05, "loss": 0.3225, "step": 4530 }, { "epoch": 1.5665976535541752, "grad_norm": 0.38204970955848694, "learning_rate": 1.0852311939268462e-05, "loss": 0.3028, "step": 4540 }, { "epoch": 1.5700483091787438, "grad_norm": 0.39774495363235474, "learning_rate": 1.0766045548654244e-05, "loss": 0.3172, "step": 4550 }, { "epoch": 1.5734989648033126, "grad_norm": 0.4597463011741638, "learning_rate": 1.0679779158040027e-05, "loss": 0.3094, "step": 4560 }, { "epoch": 1.5769496204278814, "grad_norm": 0.4128337800502777, "learning_rate": 1.0593512767425812e-05, "loss": 0.3104, "step": 4570 }, { "epoch": 1.5804002760524498, "grad_norm": 0.3733654022216797, "learning_rate": 1.0507246376811594e-05, "loss": 0.3183, "step": 4580 }, { "epoch": 1.5838509316770186, "grad_norm": 0.4163801670074463, "learning_rate": 1.0420979986197379e-05, "loss": 0.3103, "step": 4590 }, { "epoch": 1.5873015873015874, "grad_norm": 0.39067670702934265, "learning_rate": 1.0334713595583162e-05, "loss": 0.3174, "step": 4600 }, { "epoch": 1.590752242926156, "grad_norm": 0.4127908945083618, "learning_rate": 1.0248447204968944e-05, "loss": 0.3164, "step": 4610 }, { "epoch": 1.5942028985507246, "grad_norm": 0.43183374404907227, "learning_rate": 1.0162180814354729e-05, "loss": 0.3158, "step": 4620 }, { "epoch": 1.5976535541752934, "grad_norm": 0.419487863779068, "learning_rate": 1.0075914423740512e-05, "loss": 0.3131, "step": 4630 }, { "epoch": 1.601104209799862, "grad_norm": 0.38240429759025574, "learning_rate": 9.989648033126294e-06, "loss": 0.3186, "step": 4640 }, { "epoch": 1.6045548654244306, "grad_norm": 0.36253178119659424, "learning_rate": 9.903381642512077e-06, "loss": 0.3185, "step": 4650 }, { "epoch": 1.6080055210489994, "grad_norm": 0.43411344289779663, "learning_rate": 9.817115251897862e-06, "loss": 0.3097, "step": 4660 }, { "epoch": 1.611456176673568, "grad_norm": 0.4432518482208252, "learning_rate": 9.730848861283644e-06, "loss": 0.3216, "step": 4670 }, { "epoch": 1.6149068322981366, "grad_norm": 0.4278890788555145, "learning_rate": 9.644582470669427e-06, "loss": 0.3057, "step": 4680 }, { "epoch": 1.6183574879227054, "grad_norm": 0.4062660336494446, "learning_rate": 9.558316080055212e-06, "loss": 0.3157, "step": 4690 }, { "epoch": 1.621808143547274, "grad_norm": 0.4200899302959442, "learning_rate": 9.472049689440994e-06, "loss": 0.3277, "step": 4700 }, { "epoch": 1.6252587991718426, "grad_norm": 0.464057058095932, "learning_rate": 9.385783298826777e-06, "loss": 0.3145, "step": 4710 }, { "epoch": 1.6287094547964114, "grad_norm": 0.47834956645965576, "learning_rate": 9.29951690821256e-06, "loss": 0.3194, "step": 4720 }, { "epoch": 1.63216011042098, "grad_norm": 0.37987402081489563, "learning_rate": 9.213250517598344e-06, "loss": 0.311, "step": 4730 }, { "epoch": 1.6356107660455486, "grad_norm": 0.3906196653842926, "learning_rate": 9.126984126984127e-06, "loss": 0.3143, "step": 4740 }, { "epoch": 1.6390614216701174, "grad_norm": 0.4006254971027374, "learning_rate": 9.040717736369912e-06, "loss": 0.3211, "step": 4750 }, { "epoch": 1.642512077294686, "grad_norm": 0.3769705593585968, "learning_rate": 8.954451345755693e-06, "loss": 0.3111, "step": 4760 }, { "epoch": 1.6459627329192545, "grad_norm": 0.416151762008667, "learning_rate": 8.868184955141477e-06, "loss": 0.3132, "step": 4770 }, { "epoch": 1.6494133885438234, "grad_norm": 0.3826710879802704, "learning_rate": 8.781918564527262e-06, "loss": 0.3072, "step": 4780 }, { "epoch": 1.652864044168392, "grad_norm": 0.4245678782463074, "learning_rate": 8.695652173913044e-06, "loss": 0.3173, "step": 4790 }, { "epoch": 1.6563146997929605, "grad_norm": 0.4384019374847412, "learning_rate": 8.609385783298827e-06, "loss": 0.3248, "step": 4800 }, { "epoch": 1.6597653554175293, "grad_norm": 0.3771851360797882, "learning_rate": 8.52311939268461e-06, "loss": 0.3183, "step": 4810 }, { "epoch": 1.6632160110420982, "grad_norm": 0.3873949944972992, "learning_rate": 8.436853002070394e-06, "loss": 0.3041, "step": 4820 }, { "epoch": 1.6666666666666665, "grad_norm": 0.3559800684452057, "learning_rate": 8.350586611456177e-06, "loss": 0.3044, "step": 4830 }, { "epoch": 1.6701173222912353, "grad_norm": 0.3813839852809906, "learning_rate": 8.26432022084196e-06, "loss": 0.3239, "step": 4840 }, { "epoch": 1.6735679779158041, "grad_norm": 0.3791593015193939, "learning_rate": 8.178053830227744e-06, "loss": 0.3135, "step": 4850 }, { "epoch": 1.6770186335403725, "grad_norm": 0.4418281018733978, "learning_rate": 8.091787439613527e-06, "loss": 0.3097, "step": 4860 }, { "epoch": 1.6804692891649413, "grad_norm": 0.3764910101890564, "learning_rate": 8.00552104899931e-06, "loss": 0.3183, "step": 4870 }, { "epoch": 1.6839199447895101, "grad_norm": 0.3997323513031006, "learning_rate": 7.919254658385093e-06, "loss": 0.3011, "step": 4880 }, { "epoch": 1.6873706004140787, "grad_norm": 0.3994773030281067, "learning_rate": 7.832988267770877e-06, "loss": 0.306, "step": 4890 }, { "epoch": 1.6908212560386473, "grad_norm": 0.4935210943222046, "learning_rate": 7.74672187715666e-06, "loss": 0.313, "step": 4900 }, { "epoch": 1.6942719116632161, "grad_norm": 0.4665015935897827, "learning_rate": 7.660455486542443e-06, "loss": 0.3174, "step": 4910 }, { "epoch": 1.6977225672877847, "grad_norm": 0.376712828874588, "learning_rate": 7.574189095928226e-06, "loss": 0.3196, "step": 4920 }, { "epoch": 1.7011732229123533, "grad_norm": 0.49038994312286377, "learning_rate": 7.48792270531401e-06, "loss": 0.3106, "step": 4930 }, { "epoch": 1.704623878536922, "grad_norm": 0.3488839566707611, "learning_rate": 7.4016563146997936e-06, "loss": 0.3168, "step": 4940 }, { "epoch": 1.7080745341614907, "grad_norm": 0.523395836353302, "learning_rate": 7.315389924085576e-06, "loss": 0.3145, "step": 4950 }, { "epoch": 1.7115251897860593, "grad_norm": 0.37245577573776245, "learning_rate": 7.22912353347136e-06, "loss": 0.3145, "step": 4960 }, { "epoch": 1.714975845410628, "grad_norm": 0.37012121081352234, "learning_rate": 7.142857142857143e-06, "loss": 0.3063, "step": 4970 }, { "epoch": 1.7184265010351967, "grad_norm": 0.42040547728538513, "learning_rate": 7.056590752242926e-06, "loss": 0.3175, "step": 4980 }, { "epoch": 1.7218771566597653, "grad_norm": 0.40232813358306885, "learning_rate": 6.970324361628709e-06, "loss": 0.3244, "step": 4990 }, { "epoch": 1.725327812284334, "grad_norm": 0.39817917346954346, "learning_rate": 6.884057971014493e-06, "loss": 0.3105, "step": 5000 }, { "epoch": 1.7287784679089027, "grad_norm": 0.3830488622188568, "learning_rate": 6.797791580400277e-06, "loss": 0.3149, "step": 5010 }, { "epoch": 1.7322291235334712, "grad_norm": 0.3827120363712311, "learning_rate": 6.71152518978606e-06, "loss": 0.325, "step": 5020 }, { "epoch": 1.73567977915804, "grad_norm": 0.39774030447006226, "learning_rate": 6.6252587991718436e-06, "loss": 0.3075, "step": 5030 }, { "epoch": 1.7391304347826086, "grad_norm": 0.3926418423652649, "learning_rate": 6.538992408557626e-06, "loss": 0.3181, "step": 5040 }, { "epoch": 1.7425810904071772, "grad_norm": 0.44530630111694336, "learning_rate": 6.45272601794341e-06, "loss": 0.3221, "step": 5050 }, { "epoch": 1.746031746031746, "grad_norm": 0.465223103761673, "learning_rate": 6.366459627329193e-06, "loss": 0.3081, "step": 5060 }, { "epoch": 1.7494824016563149, "grad_norm": 0.37095168232917786, "learning_rate": 6.280193236714976e-06, "loss": 0.3143, "step": 5070 }, { "epoch": 1.7529330572808832, "grad_norm": 0.38144221901893616, "learning_rate": 6.193926846100759e-06, "loss": 0.309, "step": 5080 }, { "epoch": 1.756383712905452, "grad_norm": 0.3875439763069153, "learning_rate": 6.107660455486543e-06, "loss": 0.3164, "step": 5090 }, { "epoch": 1.7598343685300208, "grad_norm": 0.3756479322910309, "learning_rate": 6.021394064872326e-06, "loss": 0.3034, "step": 5100 }, { "epoch": 1.7632850241545892, "grad_norm": 0.3956107795238495, "learning_rate": 5.935127674258109e-06, "loss": 0.3159, "step": 5110 }, { "epoch": 1.766735679779158, "grad_norm": 0.42773470282554626, "learning_rate": 5.848861283643893e-06, "loss": 0.3121, "step": 5120 }, { "epoch": 1.7701863354037268, "grad_norm": 0.36983245611190796, "learning_rate": 5.762594893029676e-06, "loss": 0.3186, "step": 5130 }, { "epoch": 1.7736369910282954, "grad_norm": 0.3716372549533844, "learning_rate": 5.676328502415459e-06, "loss": 0.3081, "step": 5140 }, { "epoch": 1.777087646652864, "grad_norm": 0.44651734828948975, "learning_rate": 5.590062111801243e-06, "loss": 0.3059, "step": 5150 }, { "epoch": 1.7805383022774328, "grad_norm": 0.4011802077293396, "learning_rate": 5.5037957211870255e-06, "loss": 0.3184, "step": 5160 }, { "epoch": 1.7839889579020014, "grad_norm": 0.4085049331188202, "learning_rate": 5.417529330572809e-06, "loss": 0.3108, "step": 5170 }, { "epoch": 1.78743961352657, "grad_norm": 0.42412784695625305, "learning_rate": 5.331262939958592e-06, "loss": 0.309, "step": 5180 }, { "epoch": 1.7908902691511388, "grad_norm": 0.39434853196144104, "learning_rate": 5.2449965493443755e-06, "loss": 0.31, "step": 5190 }, { "epoch": 1.7943409247757074, "grad_norm": 0.40047556161880493, "learning_rate": 5.158730158730159e-06, "loss": 0.3137, "step": 5200 }, { "epoch": 1.797791580400276, "grad_norm": 0.37244585156440735, "learning_rate": 5.072463768115943e-06, "loss": 0.3142, "step": 5210 }, { "epoch": 1.8012422360248448, "grad_norm": 0.4294069707393646, "learning_rate": 4.9861973775017255e-06, "loss": 0.3165, "step": 5220 }, { "epoch": 1.8046928916494134, "grad_norm": 0.4544520974159241, "learning_rate": 4.899930986887509e-06, "loss": 0.3102, "step": 5230 }, { "epoch": 1.808143547273982, "grad_norm": 0.4094906151294708, "learning_rate": 4.813664596273292e-06, "loss": 0.3139, "step": 5240 }, { "epoch": 1.8115942028985508, "grad_norm": 0.39041367173194885, "learning_rate": 4.7273982056590755e-06, "loss": 0.3166, "step": 5250 }, { "epoch": 1.8150448585231194, "grad_norm": 0.3962461054325104, "learning_rate": 4.641131815044858e-06, "loss": 0.3143, "step": 5260 }, { "epoch": 1.818495514147688, "grad_norm": 0.377625972032547, "learning_rate": 4.554865424430642e-06, "loss": 0.3049, "step": 5270 }, { "epoch": 1.8219461697722568, "grad_norm": 0.3768768012523651, "learning_rate": 4.468599033816425e-06, "loss": 0.3037, "step": 5280 }, { "epoch": 1.8253968253968254, "grad_norm": 0.4099718928337097, "learning_rate": 4.382332643202209e-06, "loss": 0.3005, "step": 5290 }, { "epoch": 1.828847481021394, "grad_norm": 0.4395196735858917, "learning_rate": 4.296066252587992e-06, "loss": 0.3204, "step": 5300 }, { "epoch": 1.8322981366459627, "grad_norm": 0.42731696367263794, "learning_rate": 4.2097998619737755e-06, "loss": 0.3068, "step": 5310 }, { "epoch": 1.8357487922705316, "grad_norm": 0.4241114854812622, "learning_rate": 4.123533471359558e-06, "loss": 0.3161, "step": 5320 }, { "epoch": 1.8391994478951, "grad_norm": 0.42963647842407227, "learning_rate": 4.037267080745342e-06, "loss": 0.3165, "step": 5330 }, { "epoch": 1.8426501035196687, "grad_norm": 0.3871797025203705, "learning_rate": 3.951000690131125e-06, "loss": 0.3134, "step": 5340 }, { "epoch": 1.8461007591442375, "grad_norm": 0.42624005675315857, "learning_rate": 3.864734299516908e-06, "loss": 0.2984, "step": 5350 }, { "epoch": 1.849551414768806, "grad_norm": 0.41070929169654846, "learning_rate": 3.7784679089026914e-06, "loss": 0.3188, "step": 5360 }, { "epoch": 1.8530020703933747, "grad_norm": 0.3936339020729065, "learning_rate": 3.6922015182884755e-06, "loss": 0.3197, "step": 5370 }, { "epoch": 1.8564527260179435, "grad_norm": 0.36642739176750183, "learning_rate": 3.6059351276742587e-06, "loss": 0.3144, "step": 5380 }, { "epoch": 1.8599033816425121, "grad_norm": 0.3942832350730896, "learning_rate": 3.519668737060042e-06, "loss": 0.3074, "step": 5390 }, { "epoch": 1.8633540372670807, "grad_norm": 0.40915966033935547, "learning_rate": 3.433402346445825e-06, "loss": 0.3113, "step": 5400 }, { "epoch": 1.8668046928916495, "grad_norm": 0.4114089608192444, "learning_rate": 3.3471359558316083e-06, "loss": 0.3126, "step": 5410 }, { "epoch": 1.870255348516218, "grad_norm": 0.39819249510765076, "learning_rate": 3.2608695652173914e-06, "loss": 0.3096, "step": 5420 }, { "epoch": 1.8737060041407867, "grad_norm": 0.36793217062950134, "learning_rate": 3.1746031746031746e-06, "loss": 0.3044, "step": 5430 }, { "epoch": 1.8771566597653555, "grad_norm": 0.37002408504486084, "learning_rate": 3.0883367839889582e-06, "loss": 0.3048, "step": 5440 }, { "epoch": 1.880607315389924, "grad_norm": 0.3860247731208801, "learning_rate": 3.0020703933747414e-06, "loss": 0.322, "step": 5450 }, { "epoch": 1.8840579710144927, "grad_norm": 0.3954015374183655, "learning_rate": 2.9158040027605246e-06, "loss": 0.3123, "step": 5460 }, { "epoch": 1.8875086266390615, "grad_norm": 0.38128572702407837, "learning_rate": 2.829537612146308e-06, "loss": 0.3137, "step": 5470 }, { "epoch": 1.89095928226363, "grad_norm": 0.3847271502017975, "learning_rate": 2.7432712215320914e-06, "loss": 0.3088, "step": 5480 }, { "epoch": 1.8944099378881987, "grad_norm": 0.3671834170818329, "learning_rate": 2.6570048309178746e-06, "loss": 0.3183, "step": 5490 }, { "epoch": 1.8978605935127675, "grad_norm": 0.37888962030410767, "learning_rate": 2.570738440303658e-06, "loss": 0.3082, "step": 5500 }, { "epoch": 1.901311249137336, "grad_norm": 0.44054657220840454, "learning_rate": 2.484472049689441e-06, "loss": 0.3183, "step": 5510 }, { "epoch": 1.9047619047619047, "grad_norm": 0.3805292844772339, "learning_rate": 2.3982056590752246e-06, "loss": 0.3132, "step": 5520 }, { "epoch": 1.9082125603864735, "grad_norm": 0.435813844203949, "learning_rate": 2.311939268461008e-06, "loss": 0.3112, "step": 5530 }, { "epoch": 1.911663216011042, "grad_norm": 0.43616998195648193, "learning_rate": 2.225672877846791e-06, "loss": 0.3012, "step": 5540 }, { "epoch": 1.9151138716356106, "grad_norm": 0.3860926032066345, "learning_rate": 2.139406487232574e-06, "loss": 0.3107, "step": 5550 }, { "epoch": 1.9185645272601795, "grad_norm": 0.362621009349823, "learning_rate": 2.053140096618358e-06, "loss": 0.318, "step": 5560 }, { "epoch": 1.9220151828847483, "grad_norm": 0.36867082118988037, "learning_rate": 1.966873706004141e-06, "loss": 0.3058, "step": 5570 }, { "epoch": 1.9254658385093166, "grad_norm": 0.411702036857605, "learning_rate": 1.880607315389924e-06, "loss": 0.3176, "step": 5580 }, { "epoch": 1.9289164941338854, "grad_norm": 0.38144487142562866, "learning_rate": 1.7943409247757074e-06, "loss": 0.3185, "step": 5590 }, { "epoch": 1.9323671497584543, "grad_norm": 0.4448649287223816, "learning_rate": 1.7080745341614908e-06, "loss": 0.31, "step": 5600 }, { "epoch": 1.9358178053830226, "grad_norm": 0.4115146994590759, "learning_rate": 1.621808143547274e-06, "loss": 0.306, "step": 5610 }, { "epoch": 1.9392684610075914, "grad_norm": 0.42539817094802856, "learning_rate": 1.5355417529330572e-06, "loss": 0.3113, "step": 5620 }, { "epoch": 1.9427191166321602, "grad_norm": 0.4134668707847595, "learning_rate": 1.4492753623188406e-06, "loss": 0.3105, "step": 5630 }, { "epoch": 1.9461697722567288, "grad_norm": 0.3997168242931366, "learning_rate": 1.3630089717046238e-06, "loss": 0.3131, "step": 5640 }, { "epoch": 1.9496204278812974, "grad_norm": 0.38052627444267273, "learning_rate": 1.2767425810904072e-06, "loss": 0.3221, "step": 5650 }, { "epoch": 1.9530710835058662, "grad_norm": 0.38866785168647766, "learning_rate": 1.1904761904761904e-06, "loss": 0.3075, "step": 5660 }, { "epoch": 1.9565217391304348, "grad_norm": 0.43774518370628357, "learning_rate": 1.1042097998619738e-06, "loss": 0.313, "step": 5670 }, { "epoch": 1.9599723947550034, "grad_norm": 0.3998454511165619, "learning_rate": 1.017943409247757e-06, "loss": 0.3201, "step": 5680 }, { "epoch": 1.9634230503795722, "grad_norm": 0.4336646795272827, "learning_rate": 9.316770186335405e-07, "loss": 0.3028, "step": 5690 }, { "epoch": 1.9668737060041408, "grad_norm": 0.44575753808021545, "learning_rate": 8.454106280193237e-07, "loss": 0.2932, "step": 5700 }, { "epoch": 1.9703243616287094, "grad_norm": 0.3937937319278717, "learning_rate": 7.591442374051071e-07, "loss": 0.3154, "step": 5710 }, { "epoch": 1.9737750172532782, "grad_norm": 0.36204755306243896, "learning_rate": 6.728778467908903e-07, "loss": 0.3036, "step": 5720 }, { "epoch": 1.9772256728778468, "grad_norm": 0.346011757850647, "learning_rate": 5.866114561766736e-07, "loss": 0.3113, "step": 5730 }, { "epoch": 1.9806763285024154, "grad_norm": 0.4556833803653717, "learning_rate": 5.003450655624569e-07, "loss": 0.3087, "step": 5740 }, { "epoch": 1.9841269841269842, "grad_norm": 0.35639697313308716, "learning_rate": 4.140786749482402e-07, "loss": 0.3016, "step": 5750 }, { "epoch": 1.9875776397515528, "grad_norm": 0.38764461874961853, "learning_rate": 3.2781228433402347e-07, "loss": 0.3147, "step": 5760 }, { "epoch": 1.9910282953761214, "grad_norm": 0.45488855242729187, "learning_rate": 2.4154589371980677e-07, "loss": 0.3146, "step": 5770 }, { "epoch": 1.9944789510006902, "grad_norm": 0.38507527112960815, "learning_rate": 1.5527950310559006e-07, "loss": 0.3222, "step": 5780 }, { "epoch": 1.9979296066252588, "grad_norm": 0.3804410994052887, "learning_rate": 6.901311249137336e-08, "loss": 0.312, "step": 5790 } ], "logging_steps": 10, "max_steps": 5796, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.444959504309944e+17, "train_batch_size": 8, "trial_name": null, "trial_params": null }