{ "best_metric": 0.16163970530033112, "best_model_checkpoint": "./fine_tuned_mistral/checkpoint-2502", "epoch": 1.0, "eval_steps": 500, "global_step": 2502, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003996802557953637, "grad_norm": 187.7821502685547, "learning_rate": 0.00019968025579536372, "loss": 8.0944, "step": 10 }, { "epoch": 0.007993605115907274, "grad_norm": 0.9809202551841736, "learning_rate": 0.00019888089528377297, "loss": 0.5985, "step": 20 }, { "epoch": 0.011990407673860911, "grad_norm": 1.3078651428222656, "learning_rate": 0.00019808153477218226, "loss": 0.2975, "step": 30 }, { "epoch": 0.01598721023181455, "grad_norm": 1.6677987575531006, "learning_rate": 0.00019728217426059154, "loss": 0.4915, "step": 40 }, { "epoch": 0.019984012789768184, "grad_norm": 1.3078293800354004, "learning_rate": 0.00019648281374900082, "loss": 0.1599, "step": 50 }, { "epoch": 0.023980815347721823, "grad_norm": 0.7252454161643982, "learning_rate": 0.00019568345323741008, "loss": 0.2224, "step": 60 }, { "epoch": 0.027977617905675458, "grad_norm": 0.7883257865905762, "learning_rate": 0.00019488409272581936, "loss": 0.2254, "step": 70 }, { "epoch": 0.0319744204636291, "grad_norm": 0.2993629276752472, "learning_rate": 0.00019408473221422861, "loss": 0.2191, "step": 80 }, { "epoch": 0.03597122302158273, "grad_norm": 1.1389100551605225, "learning_rate": 0.0001932853717026379, "loss": 0.2474, "step": 90 }, { "epoch": 0.03996802557953637, "grad_norm": 0.1363106220960617, "learning_rate": 0.00019248601119104715, "loss": 0.0983, "step": 100 }, { "epoch": 0.04396482813749001, "grad_norm": 0.2188768833875656, "learning_rate": 0.00019168665067945644, "loss": 0.1767, "step": 110 }, { "epoch": 0.047961630695443645, "grad_norm": 0.5281928777694702, "learning_rate": 0.00019088729016786572, "loss": 0.2052, "step": 120 }, { "epoch": 0.05195843325339728, "grad_norm": 0.716292679309845, "learning_rate": 0.000190087929656275, "loss": 0.2613, "step": 130 }, { "epoch": 0.055955235811350916, "grad_norm": 0.6606540679931641, "learning_rate": 0.00018928856914468426, "loss": 0.2156, "step": 140 }, { "epoch": 0.05995203836930456, "grad_norm": 0.7895898222923279, "learning_rate": 0.00018848920863309354, "loss": 0.1875, "step": 150 }, { "epoch": 0.0639488409272582, "grad_norm": 1.9056642055511475, "learning_rate": 0.00018768984812150282, "loss": 0.1542, "step": 160 }, { "epoch": 0.06794564348521183, "grad_norm": 0.6912965774536133, "learning_rate": 0.00018689048760991208, "loss": 0.1603, "step": 170 }, { "epoch": 0.07194244604316546, "grad_norm": 0.33341923356056213, "learning_rate": 0.00018609112709832136, "loss": 0.199, "step": 180 }, { "epoch": 0.0759392486011191, "grad_norm": 0.30143648386001587, "learning_rate": 0.00018529176658673062, "loss": 0.1626, "step": 190 }, { "epoch": 0.07993605115907274, "grad_norm": 0.5107619166374207, "learning_rate": 0.0001844924060751399, "loss": 0.2186, "step": 200 }, { "epoch": 0.08393285371702638, "grad_norm": 0.6437661647796631, "learning_rate": 0.00018369304556354915, "loss": 0.1858, "step": 210 }, { "epoch": 0.08792965627498002, "grad_norm": 1.3543767929077148, "learning_rate": 0.00018289368505195846, "loss": 0.2575, "step": 220 }, { "epoch": 0.09192645883293366, "grad_norm": 0.10899212956428528, "learning_rate": 0.00018209432454036772, "loss": 0.236, "step": 230 }, { "epoch": 0.09592326139088729, "grad_norm": 0.8948803544044495, "learning_rate": 0.000181294964028777, "loss": 0.2171, "step": 240 }, { "epoch": 0.09992006394884093, "grad_norm": 0.8165796995162964, "learning_rate": 0.00018049560351718626, "loss": 0.1654, "step": 250 }, { "epoch": 0.10391686650679456, "grad_norm": 0.8842654824256897, "learning_rate": 0.00017969624300559554, "loss": 0.2257, "step": 260 }, { "epoch": 0.1079136690647482, "grad_norm": 0.6311301589012146, "learning_rate": 0.0001788968824940048, "loss": 0.1598, "step": 270 }, { "epoch": 0.11191047162270183, "grad_norm": 0.19081318378448486, "learning_rate": 0.00017809752198241408, "loss": 0.2578, "step": 280 }, { "epoch": 0.11590727418065548, "grad_norm": 0.7785463333129883, "learning_rate": 0.00017729816147082333, "loss": 0.2017, "step": 290 }, { "epoch": 0.11990407673860912, "grad_norm": 0.5954427123069763, "learning_rate": 0.00017649880095923262, "loss": 0.1798, "step": 300 }, { "epoch": 0.12390087929656275, "grad_norm": 0.9389680027961731, "learning_rate": 0.0001756994404476419, "loss": 0.2168, "step": 310 }, { "epoch": 0.1278976818545164, "grad_norm": 0.741294264793396, "learning_rate": 0.00017490007993605118, "loss": 0.1646, "step": 320 }, { "epoch": 0.13189448441247004, "grad_norm": 0.6614203453063965, "learning_rate": 0.00017410071942446044, "loss": 0.1309, "step": 330 }, { "epoch": 0.13589128697042366, "grad_norm": 0.31702834367752075, "learning_rate": 0.00017330135891286972, "loss": 0.25, "step": 340 }, { "epoch": 0.1398880895283773, "grad_norm": 0.9767606854438782, "learning_rate": 0.00017250199840127898, "loss": 0.223, "step": 350 }, { "epoch": 0.14388489208633093, "grad_norm": 0.7476068139076233, "learning_rate": 0.00017170263788968826, "loss": 0.2379, "step": 360 }, { "epoch": 0.14788169464428458, "grad_norm": 0.5389412045478821, "learning_rate": 0.00017090327737809751, "loss": 0.1281, "step": 370 }, { "epoch": 0.1518784972022382, "grad_norm": 0.3029794991016388, "learning_rate": 0.0001701039168665068, "loss": 0.1842, "step": 380 }, { "epoch": 0.15587529976019185, "grad_norm": 0.7127607464790344, "learning_rate": 0.00016930455635491608, "loss": 0.1654, "step": 390 }, { "epoch": 0.15987210231814547, "grad_norm": 0.6440016031265259, "learning_rate": 0.00016850519584332536, "loss": 0.2095, "step": 400 }, { "epoch": 0.16386890487609912, "grad_norm": 0.7137667536735535, "learning_rate": 0.00016770583533173462, "loss": 0.2133, "step": 410 }, { "epoch": 0.16786570743405277, "grad_norm": 0.5804216861724854, "learning_rate": 0.0001669064748201439, "loss": 0.1748, "step": 420 }, { "epoch": 0.1718625099920064, "grad_norm": 0.8831233978271484, "learning_rate": 0.00016610711430855316, "loss": 0.2371, "step": 430 }, { "epoch": 0.17585931254996004, "grad_norm": 0.47154247760772705, "learning_rate": 0.00016530775379696244, "loss": 0.1595, "step": 440 }, { "epoch": 0.17985611510791366, "grad_norm": 0.5801379680633545, "learning_rate": 0.00016450839328537172, "loss": 0.2059, "step": 450 }, { "epoch": 0.1838529176658673, "grad_norm": 0.6498193144798279, "learning_rate": 0.00016370903277378098, "loss": 0.1884, "step": 460 }, { "epoch": 0.18784972022382093, "grad_norm": 0.6178190112113953, "learning_rate": 0.00016290967226219026, "loss": 0.1875, "step": 470 }, { "epoch": 0.19184652278177458, "grad_norm": 0.4150198698043823, "learning_rate": 0.00016211031175059952, "loss": 0.163, "step": 480 }, { "epoch": 0.19584332533972823, "grad_norm": 0.6352556943893433, "learning_rate": 0.00016131095123900883, "loss": 0.2324, "step": 490 }, { "epoch": 0.19984012789768185, "grad_norm": 0.6494005918502808, "learning_rate": 0.00016051159072741808, "loss": 0.2352, "step": 500 }, { "epoch": 0.2038369304556355, "grad_norm": 0.41544780135154724, "learning_rate": 0.00015971223021582736, "loss": 0.132, "step": 510 }, { "epoch": 0.20783373301358912, "grad_norm": 0.5191593170166016, "learning_rate": 0.00015891286970423662, "loss": 0.1876, "step": 520 }, { "epoch": 0.21183053557154277, "grad_norm": 0.8121716976165771, "learning_rate": 0.0001581135091926459, "loss": 0.0978, "step": 530 }, { "epoch": 0.2158273381294964, "grad_norm": 0.8463426828384399, "learning_rate": 0.00015731414868105516, "loss": 0.1591, "step": 540 }, { "epoch": 0.21982414068745004, "grad_norm": 0.7256947159767151, "learning_rate": 0.00015651478816946444, "loss": 0.2144, "step": 550 }, { "epoch": 0.22382094324540366, "grad_norm": 0.817537248134613, "learning_rate": 0.0001557154276578737, "loss": 0.1262, "step": 560 }, { "epoch": 0.2278177458033573, "grad_norm": 0.6075506806373596, "learning_rate": 0.00015491606714628298, "loss": 0.1707, "step": 570 }, { "epoch": 0.23181454836131096, "grad_norm": 0.4158308804035187, "learning_rate": 0.00015411670663469223, "loss": 0.1992, "step": 580 }, { "epoch": 0.23581135091926458, "grad_norm": 0.8251807689666748, "learning_rate": 0.00015331734612310154, "loss": 0.188, "step": 590 }, { "epoch": 0.23980815347721823, "grad_norm": 0.3950844705104828, "learning_rate": 0.0001525179856115108, "loss": 0.1346, "step": 600 }, { "epoch": 0.24380495603517185, "grad_norm": 0.7561228275299072, "learning_rate": 0.00015171862509992008, "loss": 0.2248, "step": 610 }, { "epoch": 0.2478017585931255, "grad_norm": 1.5557297468185425, "learning_rate": 0.00015091926458832934, "loss": 0.1497, "step": 620 }, { "epoch": 0.2517985611510791, "grad_norm": 0.5314338207244873, "learning_rate": 0.00015011990407673862, "loss": 0.1718, "step": 630 }, { "epoch": 0.2557953637090328, "grad_norm": 0.622250497341156, "learning_rate": 0.00014932054356514788, "loss": 0.1844, "step": 640 }, { "epoch": 0.2597921662669864, "grad_norm": 0.7836592197418213, "learning_rate": 0.00014852118305355716, "loss": 0.1056, "step": 650 }, { "epoch": 0.2637889688249401, "grad_norm": 0.608813464641571, "learning_rate": 0.00014772182254196641, "loss": 0.2293, "step": 660 }, { "epoch": 0.26778577138289367, "grad_norm": 0.7594510316848755, "learning_rate": 0.0001469224620303757, "loss": 0.1408, "step": 670 }, { "epoch": 0.2717825739408473, "grad_norm": 0.7624852061271667, "learning_rate": 0.00014612310151878498, "loss": 0.1451, "step": 680 }, { "epoch": 0.27577937649880097, "grad_norm": 0.29632413387298584, "learning_rate": 0.00014532374100719426, "loss": 0.2376, "step": 690 }, { "epoch": 0.2797761790567546, "grad_norm": 0.5400770306587219, "learning_rate": 0.00014452438049560352, "loss": 0.1359, "step": 700 }, { "epoch": 0.2837729816147082, "grad_norm": 0.6905292272567749, "learning_rate": 0.0001437250199840128, "loss": 0.1188, "step": 710 }, { "epoch": 0.28776978417266186, "grad_norm": 0.43142080307006836, "learning_rate": 0.00014292565947242206, "loss": 0.1628, "step": 720 }, { "epoch": 0.2917665867306155, "grad_norm": 0.5245963335037231, "learning_rate": 0.00014212629896083134, "loss": 0.2371, "step": 730 }, { "epoch": 0.29576338928856916, "grad_norm": 0.2653980553150177, "learning_rate": 0.00014132693844924062, "loss": 0.1396, "step": 740 }, { "epoch": 0.2997601918465228, "grad_norm": 0.4195268154144287, "learning_rate": 0.00014052757793764988, "loss": 0.1438, "step": 750 }, { "epoch": 0.3037569944044764, "grad_norm": 0.48287633061408997, "learning_rate": 0.00013972821742605916, "loss": 0.1826, "step": 760 }, { "epoch": 0.30775379696243005, "grad_norm": 0.3496572971343994, "learning_rate": 0.00013892885691446844, "loss": 0.1042, "step": 770 }, { "epoch": 0.3117505995203837, "grad_norm": 0.4998539388179779, "learning_rate": 0.00013812949640287772, "loss": 0.1642, "step": 780 }, { "epoch": 0.31574740207833735, "grad_norm": 0.3192286789417267, "learning_rate": 0.00013733013589128698, "loss": 0.1511, "step": 790 }, { "epoch": 0.31974420463629094, "grad_norm": 0.19637304544448853, "learning_rate": 0.00013653077537969626, "loss": 0.1756, "step": 800 }, { "epoch": 0.3237410071942446, "grad_norm": 0.3817007839679718, "learning_rate": 0.00013573141486810552, "loss": 0.1729, "step": 810 }, { "epoch": 0.32773780975219824, "grad_norm": 0.5161091685295105, "learning_rate": 0.0001349320543565148, "loss": 0.1724, "step": 820 }, { "epoch": 0.3317346123101519, "grad_norm": 0.41696417331695557, "learning_rate": 0.00013413269384492406, "loss": 0.2164, "step": 830 }, { "epoch": 0.33573141486810554, "grad_norm": 0.7698529362678528, "learning_rate": 0.00013333333333333334, "loss": 0.1703, "step": 840 }, { "epoch": 0.33972821742605913, "grad_norm": 0.6981766819953918, "learning_rate": 0.0001325339728217426, "loss": 0.1563, "step": 850 }, { "epoch": 0.3437250199840128, "grad_norm": 0.8318856358528137, "learning_rate": 0.0001317346123101519, "loss": 0.185, "step": 860 }, { "epoch": 0.34772182254196643, "grad_norm": 0.7456117272377014, "learning_rate": 0.00013093525179856116, "loss": 0.1273, "step": 870 }, { "epoch": 0.3517186250999201, "grad_norm": 0.32372888922691345, "learning_rate": 0.00013013589128697044, "loss": 0.2206, "step": 880 }, { "epoch": 0.35571542765787373, "grad_norm": 0.2670097351074219, "learning_rate": 0.0001293365307753797, "loss": 0.1818, "step": 890 }, { "epoch": 0.3597122302158273, "grad_norm": 0.38958919048309326, "learning_rate": 0.00012853717026378898, "loss": 0.1137, "step": 900 }, { "epoch": 0.36370903277378097, "grad_norm": 0.47033756971359253, "learning_rate": 0.00012773780975219824, "loss": 0.1577, "step": 910 }, { "epoch": 0.3677058353317346, "grad_norm": 0.9784656167030334, "learning_rate": 0.00012693844924060752, "loss": 0.2, "step": 920 }, { "epoch": 0.37170263788968827, "grad_norm": 0.5877514481544495, "learning_rate": 0.00012613908872901678, "loss": 0.1277, "step": 930 }, { "epoch": 0.37569944044764186, "grad_norm": 0.5953680276870728, "learning_rate": 0.00012533972821742606, "loss": 0.1541, "step": 940 }, { "epoch": 0.3796962430055955, "grad_norm": 1.5499800443649292, "learning_rate": 0.00012454036770583534, "loss": 0.2633, "step": 950 }, { "epoch": 0.38369304556354916, "grad_norm": 0.8279550075531006, "learning_rate": 0.00012374100719424462, "loss": 0.2287, "step": 960 }, { "epoch": 0.3876898481215028, "grad_norm": 0.6293567419052124, "learning_rate": 0.00012294164668265388, "loss": 0.1654, "step": 970 }, { "epoch": 0.39168665067945646, "grad_norm": 0.897535502910614, "learning_rate": 0.00012214228617106316, "loss": 0.1163, "step": 980 }, { "epoch": 0.39568345323741005, "grad_norm": 0.2581837773323059, "learning_rate": 0.00012134292565947243, "loss": 0.1061, "step": 990 }, { "epoch": 0.3996802557953637, "grad_norm": 0.3749452531337738, "learning_rate": 0.0001205435651478817, "loss": 0.1198, "step": 1000 }, { "epoch": 0.40367705835331735, "grad_norm": 1.0581694841384888, "learning_rate": 0.00011974420463629097, "loss": 0.2201, "step": 1010 }, { "epoch": 0.407673860911271, "grad_norm": 0.5319008827209473, "learning_rate": 0.00011894484412470024, "loss": 0.1652, "step": 1020 }, { "epoch": 0.4116706634692246, "grad_norm": 0.5321987867355347, "learning_rate": 0.00011814548361310951, "loss": 0.1187, "step": 1030 }, { "epoch": 0.41566746602717825, "grad_norm": 0.764324426651001, "learning_rate": 0.0001173461231015188, "loss": 0.24, "step": 1040 }, { "epoch": 0.4196642685851319, "grad_norm": 0.19670064747333527, "learning_rate": 0.00011654676258992807, "loss": 0.1465, "step": 1050 }, { "epoch": 0.42366107114308554, "grad_norm": 0.6624041795730591, "learning_rate": 0.00011574740207833734, "loss": 0.1659, "step": 1060 }, { "epoch": 0.4276578737010392, "grad_norm": 0.3703235983848572, "learning_rate": 0.00011494804156674661, "loss": 0.1845, "step": 1070 }, { "epoch": 0.4316546762589928, "grad_norm": 0.7150025367736816, "learning_rate": 0.00011414868105515588, "loss": 0.1728, "step": 1080 }, { "epoch": 0.43565147881694644, "grad_norm": 0.46187496185302734, "learning_rate": 0.00011334932054356515, "loss": 0.1614, "step": 1090 }, { "epoch": 0.4396482813749001, "grad_norm": 0.614830493927002, "learning_rate": 0.00011254996003197442, "loss": 0.1669, "step": 1100 }, { "epoch": 0.44364508393285373, "grad_norm": 0.5666422247886658, "learning_rate": 0.00011175059952038369, "loss": 0.1387, "step": 1110 }, { "epoch": 0.44764188649080733, "grad_norm": 0.38144662976264954, "learning_rate": 0.00011095123900879296, "loss": 0.1492, "step": 1120 }, { "epoch": 0.451638689048761, "grad_norm": 0.46162304282188416, "learning_rate": 0.00011015187849720225, "loss": 0.1946, "step": 1130 }, { "epoch": 0.4556354916067146, "grad_norm": 0.305127888917923, "learning_rate": 0.00010935251798561152, "loss": 0.1498, "step": 1140 }, { "epoch": 0.4596322941646683, "grad_norm": 0.5906249284744263, "learning_rate": 0.00010855315747402079, "loss": 0.119, "step": 1150 }, { "epoch": 0.4636290967226219, "grad_norm": 0.10765030235052109, "learning_rate": 0.00010775379696243006, "loss": 0.1357, "step": 1160 }, { "epoch": 0.4676258992805755, "grad_norm": 0.6704568862915039, "learning_rate": 0.00010695443645083933, "loss": 0.1257, "step": 1170 }, { "epoch": 0.47162270183852917, "grad_norm": 0.5508949756622314, "learning_rate": 0.0001061550759392486, "loss": 0.1837, "step": 1180 }, { "epoch": 0.4756195043964828, "grad_norm": 0.5210421085357666, "learning_rate": 0.00010535571542765788, "loss": 0.2463, "step": 1190 }, { "epoch": 0.47961630695443647, "grad_norm": 0.8721128702163696, "learning_rate": 0.00010455635491606715, "loss": 0.1465, "step": 1200 }, { "epoch": 0.48361310951239006, "grad_norm": 1.4497110843658447, "learning_rate": 0.00010375699440447642, "loss": 0.1743, "step": 1210 }, { "epoch": 0.4876099120703437, "grad_norm": 0.46403270959854126, "learning_rate": 0.00010295763389288569, "loss": 0.1449, "step": 1220 }, { "epoch": 0.49160671462829736, "grad_norm": 0.6765905618667603, "learning_rate": 0.00010215827338129497, "loss": 0.1533, "step": 1230 }, { "epoch": 0.495603517186251, "grad_norm": 0.6220802068710327, "learning_rate": 0.00010135891286970425, "loss": 0.2103, "step": 1240 }, { "epoch": 0.49960031974420466, "grad_norm": 0.26665735244750977, "learning_rate": 0.00010055955235811352, "loss": 0.1589, "step": 1250 }, { "epoch": 0.5035971223021583, "grad_norm": 0.6205161213874817, "learning_rate": 9.976019184652279e-05, "loss": 0.1655, "step": 1260 }, { "epoch": 0.5075939248601119, "grad_norm": 0.6908559799194336, "learning_rate": 9.896083133493206e-05, "loss": 0.1659, "step": 1270 }, { "epoch": 0.5115907274180655, "grad_norm": 0.7626561522483826, "learning_rate": 9.816147082334133e-05, "loss": 0.1768, "step": 1280 }, { "epoch": 0.5155875299760192, "grad_norm": 0.6223479509353638, "learning_rate": 9.736211031175061e-05, "loss": 0.1931, "step": 1290 }, { "epoch": 0.5195843325339728, "grad_norm": 1.0529000759124756, "learning_rate": 9.656274980015988e-05, "loss": 0.2085, "step": 1300 }, { "epoch": 0.5235811350919265, "grad_norm": 1.199170708656311, "learning_rate": 9.576338928856915e-05, "loss": 0.1647, "step": 1310 }, { "epoch": 0.5275779376498801, "grad_norm": 0.8140775561332703, "learning_rate": 9.496402877697842e-05, "loss": 0.1954, "step": 1320 }, { "epoch": 0.5315747402078337, "grad_norm": 0.5082156658172607, "learning_rate": 9.416466826538769e-05, "loss": 0.211, "step": 1330 }, { "epoch": 0.5355715427657873, "grad_norm": 0.7366045117378235, "learning_rate": 9.336530775379697e-05, "loss": 0.2314, "step": 1340 }, { "epoch": 0.539568345323741, "grad_norm": 0.45685720443725586, "learning_rate": 9.256594724220624e-05, "loss": 0.1322, "step": 1350 }, { "epoch": 0.5435651478816946, "grad_norm": 0.9116412401199341, "learning_rate": 9.176658673061551e-05, "loss": 0.2288, "step": 1360 }, { "epoch": 0.5475619504396483, "grad_norm": 0.04045610502362251, "learning_rate": 9.096722621902478e-05, "loss": 0.1568, "step": 1370 }, { "epoch": 0.5515587529976019, "grad_norm": 0.5173507928848267, "learning_rate": 9.016786570743405e-05, "loss": 0.1515, "step": 1380 }, { "epoch": 0.5555555555555556, "grad_norm": 0.9522683620452881, "learning_rate": 8.936850519584333e-05, "loss": 0.2538, "step": 1390 }, { "epoch": 0.5595523581135092, "grad_norm": 0.5445271134376526, "learning_rate": 8.85691446842526e-05, "loss": 0.1693, "step": 1400 }, { "epoch": 0.5635491606714629, "grad_norm": 0.4531286954879761, "learning_rate": 8.776978417266187e-05, "loss": 0.1973, "step": 1410 }, { "epoch": 0.5675459632294164, "grad_norm": 0.7141970992088318, "learning_rate": 8.697042366107114e-05, "loss": 0.1563, "step": 1420 }, { "epoch": 0.5715427657873701, "grad_norm": 0.3411566913127899, "learning_rate": 8.617106314948042e-05, "loss": 0.1146, "step": 1430 }, { "epoch": 0.5755395683453237, "grad_norm": 0.33047667145729065, "learning_rate": 8.537170263788969e-05, "loss": 0.1565, "step": 1440 }, { "epoch": 0.5795363709032774, "grad_norm": 0.9432806372642517, "learning_rate": 8.457234212629896e-05, "loss": 0.1469, "step": 1450 }, { "epoch": 0.583533173461231, "grad_norm": 0.379395455121994, "learning_rate": 8.377298161470823e-05, "loss": 0.1181, "step": 1460 }, { "epoch": 0.5875299760191847, "grad_norm": 1.003814697265625, "learning_rate": 8.29736211031175e-05, "loss": 0.1335, "step": 1470 }, { "epoch": 0.5915267785771383, "grad_norm": 0.6143353581428528, "learning_rate": 8.217426059152678e-05, "loss": 0.1922, "step": 1480 }, { "epoch": 0.595523581135092, "grad_norm": 1.0933538675308228, "learning_rate": 8.137490007993605e-05, "loss": 0.1876, "step": 1490 }, { "epoch": 0.5995203836930456, "grad_norm": 0.7725152373313904, "learning_rate": 8.057553956834533e-05, "loss": 0.1944, "step": 1500 }, { "epoch": 0.6035171862509991, "grad_norm": 0.7263298034667969, "learning_rate": 7.97761790567546e-05, "loss": 0.2342, "step": 1510 }, { "epoch": 0.6075139888089528, "grad_norm": 0.9801928400993347, "learning_rate": 7.897681854516387e-05, "loss": 0.1932, "step": 1520 }, { "epoch": 0.6115107913669064, "grad_norm": 0.8123881816864014, "learning_rate": 7.817745803357315e-05, "loss": 0.1853, "step": 1530 }, { "epoch": 0.6155075939248601, "grad_norm": 0.19386546313762665, "learning_rate": 7.737809752198242e-05, "loss": 0.1773, "step": 1540 }, { "epoch": 0.6195043964828137, "grad_norm": 0.6336228251457214, "learning_rate": 7.657873701039169e-05, "loss": 0.1537, "step": 1550 }, { "epoch": 0.6235011990407674, "grad_norm": 0.5702576041221619, "learning_rate": 7.577937649880096e-05, "loss": 0.1771, "step": 1560 }, { "epoch": 0.627498001598721, "grad_norm": 0.35963866114616394, "learning_rate": 7.498001598721024e-05, "loss": 0.1688, "step": 1570 }, { "epoch": 0.6314948041566747, "grad_norm": 0.5332326889038086, "learning_rate": 7.418065547561951e-05, "loss": 0.169, "step": 1580 }, { "epoch": 0.6354916067146283, "grad_norm": 0.6524598598480225, "learning_rate": 7.338129496402878e-05, "loss": 0.1548, "step": 1590 }, { "epoch": 0.6394884092725819, "grad_norm": 0.5481794476509094, "learning_rate": 7.258193445243805e-05, "loss": 0.1428, "step": 1600 }, { "epoch": 0.6434852118305355, "grad_norm": 0.3422311842441559, "learning_rate": 7.178257394084733e-05, "loss": 0.1782, "step": 1610 }, { "epoch": 0.6474820143884892, "grad_norm": 0.3071247935295105, "learning_rate": 7.09832134292566e-05, "loss": 0.1486, "step": 1620 }, { "epoch": 0.6514788169464428, "grad_norm": 0.5833027958869934, "learning_rate": 7.018385291766587e-05, "loss": 0.239, "step": 1630 }, { "epoch": 0.6554756195043965, "grad_norm": 0.6470035910606384, "learning_rate": 6.938449240607514e-05, "loss": 0.1646, "step": 1640 }, { "epoch": 0.6594724220623501, "grad_norm": 0.8131123185157776, "learning_rate": 6.858513189448441e-05, "loss": 0.1891, "step": 1650 }, { "epoch": 0.6634692246203038, "grad_norm": 0.281305193901062, "learning_rate": 6.778577138289369e-05, "loss": 0.1694, "step": 1660 }, { "epoch": 0.6674660271782574, "grad_norm": 0.4203711748123169, "learning_rate": 6.698641087130296e-05, "loss": 0.1765, "step": 1670 }, { "epoch": 0.6714628297362111, "grad_norm": 0.3434304893016815, "learning_rate": 6.618705035971223e-05, "loss": 0.2065, "step": 1680 }, { "epoch": 0.6754596322941646, "grad_norm": 1.1291146278381348, "learning_rate": 6.53876898481215e-05, "loss": 0.2437, "step": 1690 }, { "epoch": 0.6794564348521183, "grad_norm": 0.5707736015319824, "learning_rate": 6.458832933653078e-05, "loss": 0.1552, "step": 1700 }, { "epoch": 0.6834532374100719, "grad_norm": 0.15203942358493805, "learning_rate": 6.378896882494005e-05, "loss": 0.1222, "step": 1710 }, { "epoch": 0.6874500399680256, "grad_norm": 1.4667842388153076, "learning_rate": 6.298960831334932e-05, "loss": 0.2537, "step": 1720 }, { "epoch": 0.6914468425259792, "grad_norm": 0.5294522643089294, "learning_rate": 6.219024780175859e-05, "loss": 0.1668, "step": 1730 }, { "epoch": 0.6954436450839329, "grad_norm": 0.7321768999099731, "learning_rate": 6.139088729016786e-05, "loss": 0.2347, "step": 1740 }, { "epoch": 0.6994404476418865, "grad_norm": 0.6722360849380493, "learning_rate": 6.059152677857714e-05, "loss": 0.1461, "step": 1750 }, { "epoch": 0.7034372501998402, "grad_norm": 0.6050254702568054, "learning_rate": 5.979216626698642e-05, "loss": 0.1743, "step": 1760 }, { "epoch": 0.7074340527577938, "grad_norm": 0.6438426971435547, "learning_rate": 5.899280575539569e-05, "loss": 0.134, "step": 1770 }, { "epoch": 0.7114308553157475, "grad_norm": 0.3928489685058594, "learning_rate": 5.8193445243804957e-05, "loss": 0.1741, "step": 1780 }, { "epoch": 0.715427657873701, "grad_norm": 0.2181692123413086, "learning_rate": 5.7394084732214226e-05, "loss": 0.1901, "step": 1790 }, { "epoch": 0.7194244604316546, "grad_norm": 0.17083604633808136, "learning_rate": 5.659472422062351e-05, "loss": 0.1611, "step": 1800 }, { "epoch": 0.7234212629896083, "grad_norm": 0.8438991904258728, "learning_rate": 5.579536370903278e-05, "loss": 0.1516, "step": 1810 }, { "epoch": 0.7274180655475619, "grad_norm": 0.5326571464538574, "learning_rate": 5.4996003197442047e-05, "loss": 0.1638, "step": 1820 }, { "epoch": 0.7314148681055156, "grad_norm": 0.17157304286956787, "learning_rate": 5.4196642685851316e-05, "loss": 0.1372, "step": 1830 }, { "epoch": 0.7354116706634692, "grad_norm": 0.5487836003303528, "learning_rate": 5.33972821742606e-05, "loss": 0.1297, "step": 1840 }, { "epoch": 0.7394084732214229, "grad_norm": 0.5392123460769653, "learning_rate": 5.259792166266987e-05, "loss": 0.1717, "step": 1850 }, { "epoch": 0.7434052757793765, "grad_norm": 0.9712696671485901, "learning_rate": 5.179856115107914e-05, "loss": 0.1538, "step": 1860 }, { "epoch": 0.7474020783373302, "grad_norm": 0.0985872820019722, "learning_rate": 5.0999200639488406e-05, "loss": 0.1218, "step": 1870 }, { "epoch": 0.7513988808952837, "grad_norm": 0.5596745014190674, "learning_rate": 5.019984012789768e-05, "loss": 0.1305, "step": 1880 }, { "epoch": 0.7553956834532374, "grad_norm": 0.6789805889129639, "learning_rate": 4.940047961630696e-05, "loss": 0.1586, "step": 1890 }, { "epoch": 0.759392486011191, "grad_norm": 0.6722508668899536, "learning_rate": 4.8601119104716234e-05, "loss": 0.1545, "step": 1900 }, { "epoch": 0.7633892885691447, "grad_norm": 0.9916704893112183, "learning_rate": 4.78017585931255e-05, "loss": 0.1132, "step": 1910 }, { "epoch": 0.7673860911270983, "grad_norm": 0.558546781539917, "learning_rate": 4.700239808153478e-05, "loss": 0.1211, "step": 1920 }, { "epoch": 0.771382893685052, "grad_norm": 0.5568097233772278, "learning_rate": 4.620303756994405e-05, "loss": 0.1415, "step": 1930 }, { "epoch": 0.7753796962430056, "grad_norm": 1.2026703357696533, "learning_rate": 4.5403677058353324e-05, "loss": 0.2147, "step": 1940 }, { "epoch": 0.7793764988009593, "grad_norm": 0.3532833456993103, "learning_rate": 4.460431654676259e-05, "loss": 0.1856, "step": 1950 }, { "epoch": 0.7833733013589129, "grad_norm": 0.7475394010543823, "learning_rate": 4.380495603517186e-05, "loss": 0.1259, "step": 1960 }, { "epoch": 0.7873701039168665, "grad_norm": 0.584240734577179, "learning_rate": 4.300559552358114e-05, "loss": 0.1643, "step": 1970 }, { "epoch": 0.7913669064748201, "grad_norm": 0.9373254179954529, "learning_rate": 4.220623501199041e-05, "loss": 0.1365, "step": 1980 }, { "epoch": 0.7953637090327738, "grad_norm": 0.09720321744680405, "learning_rate": 4.140687450039968e-05, "loss": 0.1569, "step": 1990 }, { "epoch": 0.7993605115907274, "grad_norm": 0.3333890438079834, "learning_rate": 4.060751398880895e-05, "loss": 0.1588, "step": 2000 }, { "epoch": 0.8033573141486811, "grad_norm": 0.19799380004405975, "learning_rate": 3.980815347721823e-05, "loss": 0.1463, "step": 2010 }, { "epoch": 0.8073541167066347, "grad_norm": 0.37868112325668335, "learning_rate": 3.90087929656275e-05, "loss": 0.1312, "step": 2020 }, { "epoch": 0.8113509192645884, "grad_norm": 0.48458245396614075, "learning_rate": 3.820943245403677e-05, "loss": 0.1252, "step": 2030 }, { "epoch": 0.815347721822542, "grad_norm": 0.14216120541095734, "learning_rate": 3.741007194244605e-05, "loss": 0.1453, "step": 2040 }, { "epoch": 0.8193445243804957, "grad_norm": 0.7967089414596558, "learning_rate": 3.661071143085532e-05, "loss": 0.139, "step": 2050 }, { "epoch": 0.8233413269384492, "grad_norm": 1.3931658267974854, "learning_rate": 3.5811350919264594e-05, "loss": 0.202, "step": 2060 }, { "epoch": 0.8273381294964028, "grad_norm": 0.6114959120750427, "learning_rate": 3.501199040767386e-05, "loss": 0.1687, "step": 2070 }, { "epoch": 0.8313349320543565, "grad_norm": 0.8499948382377625, "learning_rate": 3.421262989608314e-05, "loss": 0.1564, "step": 2080 }, { "epoch": 0.8353317346123101, "grad_norm": 1.8267698287963867, "learning_rate": 3.341326938449241e-05, "loss": 0.2304, "step": 2090 }, { "epoch": 0.8393285371702638, "grad_norm": 0.572331428527832, "learning_rate": 3.2613908872901684e-05, "loss": 0.0897, "step": 2100 }, { "epoch": 0.8433253397282174, "grad_norm": 0.6183590888977051, "learning_rate": 3.181454836131095e-05, "loss": 0.127, "step": 2110 }, { "epoch": 0.8473221422861711, "grad_norm": 0.44028234481811523, "learning_rate": 3.101518784972022e-05, "loss": 0.2057, "step": 2120 }, { "epoch": 0.8513189448441247, "grad_norm": 0.6345455050468445, "learning_rate": 3.0215827338129498e-05, "loss": 0.1591, "step": 2130 }, { "epoch": 0.8553157474020784, "grad_norm": 1.1488350629806519, "learning_rate": 2.9416466826538767e-05, "loss": 0.2077, "step": 2140 }, { "epoch": 0.8593125499600319, "grad_norm": 0.7199774384498596, "learning_rate": 2.8617106314948043e-05, "loss": 0.1261, "step": 2150 }, { "epoch": 0.8633093525179856, "grad_norm": 0.8149317502975464, "learning_rate": 2.7817745803357316e-05, "loss": 0.1791, "step": 2160 }, { "epoch": 0.8673061550759392, "grad_norm": 0.33090534806251526, "learning_rate": 2.701838529176659e-05, "loss": 0.2044, "step": 2170 }, { "epoch": 0.8713029576338929, "grad_norm": 1.0773526430130005, "learning_rate": 2.621902478017586e-05, "loss": 0.1776, "step": 2180 }, { "epoch": 0.8752997601918465, "grad_norm": 0.5906376838684082, "learning_rate": 2.541966426858513e-05, "loss": 0.1361, "step": 2190 }, { "epoch": 0.8792965627498002, "grad_norm": 0.6298534870147705, "learning_rate": 2.4620303756994406e-05, "loss": 0.178, "step": 2200 }, { "epoch": 0.8832933653077538, "grad_norm": 1.040358066558838, "learning_rate": 2.3820943245403678e-05, "loss": 0.2245, "step": 2210 }, { "epoch": 0.8872901678657075, "grad_norm": 0.6028202176094055, "learning_rate": 2.302158273381295e-05, "loss": 0.0818, "step": 2220 }, { "epoch": 0.8912869704236611, "grad_norm": 0.43753504753112793, "learning_rate": 2.2222222222222223e-05, "loss": 0.1271, "step": 2230 }, { "epoch": 0.8952837729816147, "grad_norm": 0.34733837842941284, "learning_rate": 2.1422861710631496e-05, "loss": 0.1148, "step": 2240 }, { "epoch": 0.8992805755395683, "grad_norm": 0.33790159225463867, "learning_rate": 2.062350119904077e-05, "loss": 0.2148, "step": 2250 }, { "epoch": 0.903277378097522, "grad_norm": 1.4219551086425781, "learning_rate": 1.982414068745004e-05, "loss": 0.181, "step": 2260 }, { "epoch": 0.9072741806554756, "grad_norm": 1.1721121072769165, "learning_rate": 1.9024780175859313e-05, "loss": 0.2161, "step": 2270 }, { "epoch": 0.9112709832134293, "grad_norm": 1.0592964887619019, "learning_rate": 1.8225419664268586e-05, "loss": 0.1566, "step": 2280 }, { "epoch": 0.9152677857713829, "grad_norm": 0.2788325250148773, "learning_rate": 1.742605915267786e-05, "loss": 0.1555, "step": 2290 }, { "epoch": 0.9192645883293366, "grad_norm": 0.7817983031272888, "learning_rate": 1.662669864108713e-05, "loss": 0.1883, "step": 2300 }, { "epoch": 0.9232613908872902, "grad_norm": 0.26448529958724976, "learning_rate": 1.5827338129496403e-05, "loss": 0.2078, "step": 2310 }, { "epoch": 0.9272581934452439, "grad_norm": 0.92426997423172, "learning_rate": 1.5027977617905676e-05, "loss": 0.1933, "step": 2320 }, { "epoch": 0.9312549960031974, "grad_norm": 0.7494958639144897, "learning_rate": 1.4228617106314948e-05, "loss": 0.1466, "step": 2330 }, { "epoch": 0.935251798561151, "grad_norm": 0.3624798357486725, "learning_rate": 1.3429256594724221e-05, "loss": 0.1308, "step": 2340 }, { "epoch": 0.9392486011191047, "grad_norm": 0.6994137763977051, "learning_rate": 1.2629896083133494e-05, "loss": 0.1665, "step": 2350 }, { "epoch": 0.9432454036770583, "grad_norm": 0.9269047975540161, "learning_rate": 1.1830535571542766e-05, "loss": 0.193, "step": 2360 }, { "epoch": 0.947242206235012, "grad_norm": 0.5760440230369568, "learning_rate": 1.1031175059952039e-05, "loss": 0.109, "step": 2370 }, { "epoch": 0.9512390087929656, "grad_norm": 0.9416897296905518, "learning_rate": 1.0231814548361311e-05, "loss": 0.1273, "step": 2380 }, { "epoch": 0.9552358113509193, "grad_norm": 0.7053970098495483, "learning_rate": 9.432454036770584e-06, "loss": 0.1861, "step": 2390 }, { "epoch": 0.9592326139088729, "grad_norm": 1.9960613250732422, "learning_rate": 8.633093525179858e-06, "loss": 0.1811, "step": 2400 }, { "epoch": 0.9632294164668266, "grad_norm": 1.1505160331726074, "learning_rate": 7.833733013589129e-06, "loss": 0.2093, "step": 2410 }, { "epoch": 0.9672262190247801, "grad_norm": 0.7323939204216003, "learning_rate": 7.034372501998401e-06, "loss": 0.1221, "step": 2420 }, { "epoch": 0.9712230215827338, "grad_norm": 0.810534656047821, "learning_rate": 6.2350119904076745e-06, "loss": 0.1261, "step": 2430 }, { "epoch": 0.9752198241406874, "grad_norm": 0.25207674503326416, "learning_rate": 5.435651478816946e-06, "loss": 0.1268, "step": 2440 }, { "epoch": 0.9792166266986411, "grad_norm": 0.6386179327964783, "learning_rate": 4.6362909672262196e-06, "loss": 0.1437, "step": 2450 }, { "epoch": 0.9832134292565947, "grad_norm": 0.9403632879257202, "learning_rate": 3.836930455635491e-06, "loss": 0.2184, "step": 2460 }, { "epoch": 0.9872102318145484, "grad_norm": 0.7406390905380249, "learning_rate": 3.0375699440447646e-06, "loss": 0.1826, "step": 2470 }, { "epoch": 0.991207034372502, "grad_norm": 0.5971857309341431, "learning_rate": 2.238209432454037e-06, "loss": 0.1409, "step": 2480 }, { "epoch": 0.9952038369304557, "grad_norm": 0.8005123734474182, "learning_rate": 1.4388489208633094e-06, "loss": 0.18, "step": 2490 }, { "epoch": 0.9992006394884093, "grad_norm": 0.33795028924942017, "learning_rate": 6.394884092725819e-07, "loss": 0.1293, "step": 2500 }, { "epoch": 1.0, "eval_loss": 0.16163970530033112, "eval_runtime": 143.7398, "eval_samples_per_second": 1.934, "eval_steps_per_second": 1.934, "step": 2502 } ], "logging_steps": 10, "max_steps": 2502, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 5.467975082783539e+16, "train_batch_size": 1, "trial_name": null, "trial_params": null }