diff --git "a/checkpoint-1800/trainer_state.json" "b/checkpoint-1800/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-1800/trainer_state.json" @@ -0,0 +1,12634 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.45302963568866794, + "eval_steps": 500, + "global_step": 1800, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0002516831309381489, + "grad_norm": 0.329476922750473, + "learning_rate": 0.0, + "loss": 0.8311, + "step": 1 + }, + { + "epoch": 0.0005033662618762978, + "grad_norm": 0.4849923849105835, + "learning_rate": 4e-05, + "loss": 0.8748, + "step": 2 + }, + { + "epoch": 0.0007550493928144466, + "grad_norm": 0.3723175823688507, + "learning_rate": 8e-05, + "loss": 0.836, + "step": 3 + }, + { + "epoch": 0.0010067325237525955, + "grad_norm": 0.39188194274902344, + "learning_rate": 0.00012, + "loss": 0.8132, + "step": 4 + }, + { + "epoch": 0.0012584156546907442, + "grad_norm": 0.35436731576919556, + "learning_rate": 0.00016, + "loss": 0.7304, + "step": 5 + }, + { + "epoch": 0.0015100987856288932, + "grad_norm": 0.3973938524723053, + "learning_rate": 0.0002, + "loss": 0.6529, + "step": 6 + }, + { + "epoch": 0.0017617819165670421, + "grad_norm": 0.6035579442977905, + "learning_rate": 0.00019984555984555987, + "loss": 0.5394, + "step": 7 + }, + { + "epoch": 0.002013465047505191, + "grad_norm": 0.5570532083511353, + "learning_rate": 0.00019969111969111972, + "loss": 0.4813, + "step": 8 + }, + { + "epoch": 0.0022651481784433398, + "grad_norm": 1.1288173198699951, + "learning_rate": 0.00019953667953667953, + "loss": 0.467, + "step": 9 + }, + { + "epoch": 0.0025168313093814885, + "grad_norm": 0.39599481225013733, + "learning_rate": 0.00019938223938223938, + "loss": 0.4542, + "step": 10 + }, + { + "epoch": 0.0027685144403196377, + "grad_norm": 0.3398708999156952, + "learning_rate": 0.00019922779922779924, + "loss": 0.3725, + "step": 11 + }, + { + "epoch": 0.0030201975712577864, + "grad_norm": 0.2688748240470886, + "learning_rate": 0.0001990733590733591, + "loss": 0.4224, + "step": 12 + }, + { + "epoch": 0.0032718807021959355, + "grad_norm": 0.42461395263671875, + "learning_rate": 0.00019891891891891895, + "loss": 0.2745, + "step": 13 + }, + { + "epoch": 0.0035235638331340842, + "grad_norm": 0.24118675291538239, + "learning_rate": 0.00019876447876447876, + "loss": 0.2708, + "step": 14 + }, + { + "epoch": 0.003775246964072233, + "grad_norm": 0.22975586354732513, + "learning_rate": 0.0001986100386100386, + "loss": 0.3141, + "step": 15 + }, + { + "epoch": 0.004026930095010382, + "grad_norm": 0.19872848689556122, + "learning_rate": 0.00019845559845559847, + "loss": 0.3394, + "step": 16 + }, + { + "epoch": 0.00427861322594853, + "grad_norm": 0.20219211280345917, + "learning_rate": 0.00019830115830115833, + "loss": 0.2715, + "step": 17 + }, + { + "epoch": 0.0045302963568866796, + "grad_norm": 0.1769620180130005, + "learning_rate": 0.00019814671814671816, + "loss": 0.442, + "step": 18 + }, + { + "epoch": 0.004781979487824829, + "grad_norm": 0.21095411479473114, + "learning_rate": 0.00019799227799227798, + "loss": 0.2622, + "step": 19 + }, + { + "epoch": 0.005033662618762977, + "grad_norm": 0.197563037276268, + "learning_rate": 0.00019783783783783784, + "loss": 0.3574, + "step": 20 + }, + { + "epoch": 0.005285345749701126, + "grad_norm": 0.181350976228714, + "learning_rate": 0.0001976833976833977, + "loss": 0.3314, + "step": 21 + }, + { + "epoch": 0.005537028880639275, + "grad_norm": 0.1490330845117569, + "learning_rate": 0.00019752895752895756, + "loss": 0.2269, + "step": 22 + }, + { + "epoch": 0.005788712011577424, + "grad_norm": 0.15000289678573608, + "learning_rate": 0.00019737451737451738, + "loss": 0.2589, + "step": 23 + }, + { + "epoch": 0.006040395142515573, + "grad_norm": 0.1449330598115921, + "learning_rate": 0.00019722007722007721, + "loss": 0.2023, + "step": 24 + }, + { + "epoch": 0.006292078273453722, + "grad_norm": 0.15910927951335907, + "learning_rate": 0.00019706563706563707, + "loss": 0.2229, + "step": 25 + }, + { + "epoch": 0.006543761404391871, + "grad_norm": 0.1857304573059082, + "learning_rate": 0.00019691119691119693, + "loss": 0.2961, + "step": 26 + }, + { + "epoch": 0.006795444535330019, + "grad_norm": 0.1554359644651413, + "learning_rate": 0.00019675675675675678, + "loss": 0.2986, + "step": 27 + }, + { + "epoch": 0.0070471276662681685, + "grad_norm": 0.17084558308124542, + "learning_rate": 0.00019660231660231661, + "loss": 0.2899, + "step": 28 + }, + { + "epoch": 0.007298810797206318, + "grad_norm": 0.13841071724891663, + "learning_rate": 0.00019644787644787644, + "loss": 0.1713, + "step": 29 + }, + { + "epoch": 0.007550493928144466, + "grad_norm": 0.1402069330215454, + "learning_rate": 0.0001962934362934363, + "loss": 0.2147, + "step": 30 + }, + { + "epoch": 0.007802177059082615, + "grad_norm": 0.15434876084327698, + "learning_rate": 0.00019613899613899616, + "loss": 0.2271, + "step": 31 + }, + { + "epoch": 0.008053860190020764, + "grad_norm": 0.16629579663276672, + "learning_rate": 0.000195984555984556, + "loss": 0.2762, + "step": 32 + }, + { + "epoch": 0.008305543320958913, + "grad_norm": 0.16793154180049896, + "learning_rate": 0.00019583011583011584, + "loss": 0.2658, + "step": 33 + }, + { + "epoch": 0.00855722645189706, + "grad_norm": 0.15022817254066467, + "learning_rate": 0.00019567567567567567, + "loss": 0.2035, + "step": 34 + }, + { + "epoch": 0.00880890958283521, + "grad_norm": 0.16670134663581848, + "learning_rate": 0.00019552123552123553, + "loss": 0.2569, + "step": 35 + }, + { + "epoch": 0.009060592713773359, + "grad_norm": 0.17107737064361572, + "learning_rate": 0.0001953667953667954, + "loss": 0.2186, + "step": 36 + }, + { + "epoch": 0.009312275844711507, + "grad_norm": 0.1448991596698761, + "learning_rate": 0.00019521235521235522, + "loss": 0.2115, + "step": 37 + }, + { + "epoch": 0.009563958975649657, + "grad_norm": 0.15753597021102905, + "learning_rate": 0.00019505791505791507, + "loss": 0.2423, + "step": 38 + }, + { + "epoch": 0.009815642106587806, + "grad_norm": 0.16168172657489777, + "learning_rate": 0.0001949034749034749, + "loss": 0.2227, + "step": 39 + }, + { + "epoch": 0.010067325237525954, + "grad_norm": 0.14396736025810242, + "learning_rate": 0.00019474903474903476, + "loss": 0.19, + "step": 40 + }, + { + "epoch": 0.010319008368464104, + "grad_norm": 0.13932976126670837, + "learning_rate": 0.00019459459459459462, + "loss": 0.1788, + "step": 41 + }, + { + "epoch": 0.010570691499402252, + "grad_norm": 0.14287018775939941, + "learning_rate": 0.00019444015444015445, + "loss": 0.2332, + "step": 42 + }, + { + "epoch": 0.0108223746303404, + "grad_norm": 0.13390415906906128, + "learning_rate": 0.0001942857142857143, + "loss": 0.2339, + "step": 43 + }, + { + "epoch": 0.01107405776127855, + "grad_norm": 0.14630131423473358, + "learning_rate": 0.00019413127413127413, + "loss": 0.3043, + "step": 44 + }, + { + "epoch": 0.011325740892216699, + "grad_norm": 0.13791146874427795, + "learning_rate": 0.000193976833976834, + "loss": 0.2397, + "step": 45 + }, + { + "epoch": 0.011577424023154847, + "grad_norm": 0.13436608016490936, + "learning_rate": 0.00019382239382239382, + "loss": 0.2316, + "step": 46 + }, + { + "epoch": 0.011829107154092997, + "grad_norm": 0.1335318386554718, + "learning_rate": 0.00019366795366795368, + "loss": 0.2208, + "step": 47 + }, + { + "epoch": 0.012080790285031145, + "grad_norm": 0.1479828953742981, + "learning_rate": 0.00019351351351351353, + "loss": 0.2994, + "step": 48 + }, + { + "epoch": 0.012332473415969296, + "grad_norm": 0.14406880736351013, + "learning_rate": 0.00019335907335907336, + "loss": 0.2315, + "step": 49 + }, + { + "epoch": 0.012584156546907444, + "grad_norm": 0.12098966538906097, + "learning_rate": 0.00019320463320463322, + "loss": 0.2009, + "step": 50 + }, + { + "epoch": 0.012835839677845592, + "grad_norm": 0.13052137196063995, + "learning_rate": 0.00019305019305019305, + "loss": 0.2352, + "step": 51 + }, + { + "epoch": 0.013087522808783742, + "grad_norm": 0.1393144726753235, + "learning_rate": 0.0001928957528957529, + "loss": 0.2432, + "step": 52 + }, + { + "epoch": 0.01333920593972189, + "grad_norm": 0.14476315677165985, + "learning_rate": 0.00019274131274131276, + "loss": 0.3344, + "step": 53 + }, + { + "epoch": 0.013590889070660039, + "grad_norm": 0.1502915769815445, + "learning_rate": 0.0001925868725868726, + "loss": 0.3052, + "step": 54 + }, + { + "epoch": 0.013842572201598189, + "grad_norm": 0.11713853478431702, + "learning_rate": 0.00019243243243243245, + "loss": 0.1926, + "step": 55 + }, + { + "epoch": 0.014094255332536337, + "grad_norm": 0.13639037311077118, + "learning_rate": 0.00019227799227799228, + "loss": 0.1879, + "step": 56 + }, + { + "epoch": 0.014345938463474485, + "grad_norm": 0.16080759465694427, + "learning_rate": 0.00019212355212355214, + "loss": 0.3227, + "step": 57 + }, + { + "epoch": 0.014597621594412635, + "grad_norm": 0.13372309505939484, + "learning_rate": 0.000191969111969112, + "loss": 0.2201, + "step": 58 + }, + { + "epoch": 0.014849304725350784, + "grad_norm": 0.14892010390758514, + "learning_rate": 0.00019181467181467182, + "loss": 0.2576, + "step": 59 + }, + { + "epoch": 0.015100987856288932, + "grad_norm": 0.1428636759519577, + "learning_rate": 0.00019166023166023165, + "loss": 0.2194, + "step": 60 + }, + { + "epoch": 0.015352670987227082, + "grad_norm": 0.15054349601268768, + "learning_rate": 0.0001915057915057915, + "loss": 0.2535, + "step": 61 + }, + { + "epoch": 0.01560435411816523, + "grad_norm": 0.19205592572689056, + "learning_rate": 0.00019135135135135137, + "loss": 0.2969, + "step": 62 + }, + { + "epoch": 0.01585603724910338, + "grad_norm": 0.14666880667209625, + "learning_rate": 0.00019119691119691122, + "loss": 0.3299, + "step": 63 + }, + { + "epoch": 0.01610772038004153, + "grad_norm": 0.1472829133272171, + "learning_rate": 0.00019104247104247105, + "loss": 0.2171, + "step": 64 + }, + { + "epoch": 0.016359403510979677, + "grad_norm": 0.14535754919052124, + "learning_rate": 0.00019088803088803088, + "loss": 0.2093, + "step": 65 + }, + { + "epoch": 0.016611086641917825, + "grad_norm": 0.16431409120559692, + "learning_rate": 0.00019073359073359074, + "loss": 0.4131, + "step": 66 + }, + { + "epoch": 0.016862769772855973, + "grad_norm": 0.14110589027404785, + "learning_rate": 0.0001905791505791506, + "loss": 0.2163, + "step": 67 + }, + { + "epoch": 0.01711445290379412, + "grad_norm": 0.15895260870456696, + "learning_rate": 0.00019042471042471045, + "loss": 0.3282, + "step": 68 + }, + { + "epoch": 0.017366136034732273, + "grad_norm": 0.1170768290758133, + "learning_rate": 0.00019027027027027028, + "loss": 0.1889, + "step": 69 + }, + { + "epoch": 0.01761781916567042, + "grad_norm": 0.14574691653251648, + "learning_rate": 0.0001901158301158301, + "loss": 0.3353, + "step": 70 + }, + { + "epoch": 0.01786950229660857, + "grad_norm": 0.12267708778381348, + "learning_rate": 0.00018996138996138997, + "loss": 0.2362, + "step": 71 + }, + { + "epoch": 0.018121185427546718, + "grad_norm": 0.12682929635047913, + "learning_rate": 0.00018980694980694982, + "loss": 0.2762, + "step": 72 + }, + { + "epoch": 0.018372868558484867, + "grad_norm": 0.12169703841209412, + "learning_rate": 0.00018965250965250968, + "loss": 0.2159, + "step": 73 + }, + { + "epoch": 0.018624551689423015, + "grad_norm": 0.1468794047832489, + "learning_rate": 0.00018949806949806948, + "loss": 0.2406, + "step": 74 + }, + { + "epoch": 0.018876234820361167, + "grad_norm": 0.1189325675368309, + "learning_rate": 0.00018934362934362934, + "loss": 0.1767, + "step": 75 + }, + { + "epoch": 0.019127917951299315, + "grad_norm": 0.13012659549713135, + "learning_rate": 0.0001891891891891892, + "loss": 0.1991, + "step": 76 + }, + { + "epoch": 0.019379601082237463, + "grad_norm": 0.13789086043834686, + "learning_rate": 0.00018903474903474905, + "loss": 0.3025, + "step": 77 + }, + { + "epoch": 0.01963128421317561, + "grad_norm": 0.12848761677742004, + "learning_rate": 0.0001888803088803089, + "loss": 0.2031, + "step": 78 + }, + { + "epoch": 0.01988296734411376, + "grad_norm": 0.1437838077545166, + "learning_rate": 0.00018872586872586871, + "loss": 0.2383, + "step": 79 + }, + { + "epoch": 0.020134650475051908, + "grad_norm": 0.15037506818771362, + "learning_rate": 0.00018857142857142857, + "loss": 0.2149, + "step": 80 + }, + { + "epoch": 0.02038633360599006, + "grad_norm": 0.13283689320087433, + "learning_rate": 0.00018841698841698843, + "loss": 0.2278, + "step": 81 + }, + { + "epoch": 0.020638016736928208, + "grad_norm": 0.1339232474565506, + "learning_rate": 0.00018826254826254828, + "loss": 0.2373, + "step": 82 + }, + { + "epoch": 0.020889699867866356, + "grad_norm": 0.1281827837228775, + "learning_rate": 0.00018810810810810814, + "loss": 0.2211, + "step": 83 + }, + { + "epoch": 0.021141382998804505, + "grad_norm": 0.1303407847881317, + "learning_rate": 0.00018795366795366794, + "loss": 0.1825, + "step": 84 + }, + { + "epoch": 0.021393066129742653, + "grad_norm": 0.13296560943126678, + "learning_rate": 0.0001877992277992278, + "loss": 0.2207, + "step": 85 + }, + { + "epoch": 0.0216447492606808, + "grad_norm": 0.1432134509086609, + "learning_rate": 0.00018764478764478766, + "loss": 0.3297, + "step": 86 + }, + { + "epoch": 0.021896432391618953, + "grad_norm": 0.13179218769073486, + "learning_rate": 0.00018749034749034751, + "loss": 0.2509, + "step": 87 + }, + { + "epoch": 0.0221481155225571, + "grad_norm": 0.1309555023908615, + "learning_rate": 0.00018733590733590734, + "loss": 0.2553, + "step": 88 + }, + { + "epoch": 0.02239979865349525, + "grad_norm": 0.14680728316307068, + "learning_rate": 0.00018718146718146717, + "loss": 0.3495, + "step": 89 + }, + { + "epoch": 0.022651481784433398, + "grad_norm": 0.15065547823905945, + "learning_rate": 0.00018702702702702703, + "loss": 0.3184, + "step": 90 + }, + { + "epoch": 0.022903164915371546, + "grad_norm": 0.13301771879196167, + "learning_rate": 0.0001868725868725869, + "loss": 0.2472, + "step": 91 + }, + { + "epoch": 0.023154848046309694, + "grad_norm": 0.12873731553554535, + "learning_rate": 0.00018671814671814674, + "loss": 0.2491, + "step": 92 + }, + { + "epoch": 0.023406531177247846, + "grad_norm": 0.12557490170001984, + "learning_rate": 0.00018656370656370657, + "loss": 0.2428, + "step": 93 + }, + { + "epoch": 0.023658214308185994, + "grad_norm": 0.14100699126720428, + "learning_rate": 0.0001864092664092664, + "loss": 0.2205, + "step": 94 + }, + { + "epoch": 0.023909897439124143, + "grad_norm": 0.12088359147310257, + "learning_rate": 0.00018625482625482626, + "loss": 0.1887, + "step": 95 + }, + { + "epoch": 0.02416158057006229, + "grad_norm": 0.14802686870098114, + "learning_rate": 0.00018610038610038612, + "loss": 0.3071, + "step": 96 + }, + { + "epoch": 0.02441326370100044, + "grad_norm": 0.1482524573802948, + "learning_rate": 0.00018594594594594597, + "loss": 0.2294, + "step": 97 + }, + { + "epoch": 0.02466494683193859, + "grad_norm": 0.15035012364387512, + "learning_rate": 0.0001857915057915058, + "loss": 0.2458, + "step": 98 + }, + { + "epoch": 0.02491662996287674, + "grad_norm": 0.13679851591587067, + "learning_rate": 0.00018563706563706563, + "loss": 0.2594, + "step": 99 + }, + { + "epoch": 0.025168313093814888, + "grad_norm": 0.12072622030973434, + "learning_rate": 0.0001854826254826255, + "loss": 0.1825, + "step": 100 + }, + { + "epoch": 0.025419996224753036, + "grad_norm": 0.13065776228904724, + "learning_rate": 0.00018532818532818535, + "loss": 0.2248, + "step": 101 + }, + { + "epoch": 0.025671679355691184, + "grad_norm": 0.12138719111680984, + "learning_rate": 0.00018517374517374518, + "loss": 0.2321, + "step": 102 + }, + { + "epoch": 0.025923362486629332, + "grad_norm": 0.1442973017692566, + "learning_rate": 0.00018501930501930503, + "loss": 0.3118, + "step": 103 + }, + { + "epoch": 0.026175045617567484, + "grad_norm": 0.11479207873344421, + "learning_rate": 0.00018486486486486486, + "loss": 0.1835, + "step": 104 + }, + { + "epoch": 0.026426728748505632, + "grad_norm": 0.11119953542947769, + "learning_rate": 0.00018471042471042472, + "loss": 0.1696, + "step": 105 + }, + { + "epoch": 0.02667841187944378, + "grad_norm": 0.11810140311717987, + "learning_rate": 0.00018455598455598458, + "loss": 0.1959, + "step": 106 + }, + { + "epoch": 0.02693009501038193, + "grad_norm": 0.1518896073102951, + "learning_rate": 0.0001844015444015444, + "loss": 0.2411, + "step": 107 + }, + { + "epoch": 0.027181778141320077, + "grad_norm": 0.12609782814979553, + "learning_rate": 0.00018424710424710426, + "loss": 0.1985, + "step": 108 + }, + { + "epoch": 0.027433461272258226, + "grad_norm": 0.15130166709423065, + "learning_rate": 0.0001840926640926641, + "loss": 0.2749, + "step": 109 + }, + { + "epoch": 0.027685144403196377, + "grad_norm": 0.1306765377521515, + "learning_rate": 0.00018393822393822395, + "loss": 0.2079, + "step": 110 + }, + { + "epoch": 0.027936827534134526, + "grad_norm": 0.1174439787864685, + "learning_rate": 0.0001837837837837838, + "loss": 0.1793, + "step": 111 + }, + { + "epoch": 0.028188510665072674, + "grad_norm": 0.13786830008029938, + "learning_rate": 0.00018362934362934363, + "loss": 0.2002, + "step": 112 + }, + { + "epoch": 0.028440193796010822, + "grad_norm": 0.13419173657894135, + "learning_rate": 0.0001834749034749035, + "loss": 0.2141, + "step": 113 + }, + { + "epoch": 0.02869187692694897, + "grad_norm": 0.11032861471176147, + "learning_rate": 0.00018332046332046332, + "loss": 0.1904, + "step": 114 + }, + { + "epoch": 0.02894356005788712, + "grad_norm": 0.11321274191141129, + "learning_rate": 0.00018316602316602318, + "loss": 0.1933, + "step": 115 + }, + { + "epoch": 0.02919524318882527, + "grad_norm": 0.12911099195480347, + "learning_rate": 0.000183011583011583, + "loss": 0.1936, + "step": 116 + }, + { + "epoch": 0.02944692631976342, + "grad_norm": 0.1368657350540161, + "learning_rate": 0.00018285714285714286, + "loss": 0.26, + "step": 117 + }, + { + "epoch": 0.029698609450701567, + "grad_norm": 0.13694381713867188, + "learning_rate": 0.00018270270270270272, + "loss": 0.2252, + "step": 118 + }, + { + "epoch": 0.029950292581639715, + "grad_norm": 0.17163509130477905, + "learning_rate": 0.00018254826254826255, + "loss": 0.1868, + "step": 119 + }, + { + "epoch": 0.030201975712577864, + "grad_norm": 0.16179172694683075, + "learning_rate": 0.0001823938223938224, + "loss": 0.3332, + "step": 120 + }, + { + "epoch": 0.030453658843516012, + "grad_norm": 0.1263120323419571, + "learning_rate": 0.00018223938223938224, + "loss": 0.2215, + "step": 121 + }, + { + "epoch": 0.030705341974454164, + "grad_norm": 0.13037261366844177, + "learning_rate": 0.0001820849420849421, + "loss": 0.2128, + "step": 122 + }, + { + "epoch": 0.030957025105392312, + "grad_norm": 0.12833549082279205, + "learning_rate": 0.00018193050193050195, + "loss": 0.1963, + "step": 123 + }, + { + "epoch": 0.03120870823633046, + "grad_norm": 0.1527228206396103, + "learning_rate": 0.00018177606177606178, + "loss": 0.3337, + "step": 124 + }, + { + "epoch": 0.03146039136726861, + "grad_norm": 0.12143786996603012, + "learning_rate": 0.00018162162162162164, + "loss": 0.1958, + "step": 125 + }, + { + "epoch": 0.03171207449820676, + "grad_norm": 0.13796019554138184, + "learning_rate": 0.00018146718146718147, + "loss": 0.2835, + "step": 126 + }, + { + "epoch": 0.031963757629144905, + "grad_norm": 0.13006722927093506, + "learning_rate": 0.00018131274131274132, + "loss": 0.2613, + "step": 127 + }, + { + "epoch": 0.03221544076008306, + "grad_norm": 0.11909051239490509, + "learning_rate": 0.00018115830115830118, + "loss": 0.1952, + "step": 128 + }, + { + "epoch": 0.0324671238910212, + "grad_norm": 0.12578924000263214, + "learning_rate": 0.000181003861003861, + "loss": 0.2194, + "step": 129 + }, + { + "epoch": 0.032718807021959354, + "grad_norm": 0.12145200371742249, + "learning_rate": 0.00018084942084942084, + "loss": 0.2155, + "step": 130 + }, + { + "epoch": 0.032970490152897505, + "grad_norm": 0.13555808365345, + "learning_rate": 0.0001806949806949807, + "loss": 0.2401, + "step": 131 + }, + { + "epoch": 0.03322217328383565, + "grad_norm": 0.12870927155017853, + "learning_rate": 0.00018054054054054055, + "loss": 0.2209, + "step": 132 + }, + { + "epoch": 0.0334738564147738, + "grad_norm": 0.15006543695926666, + "learning_rate": 0.0001803861003861004, + "loss": 0.3276, + "step": 133 + }, + { + "epoch": 0.03372553954571195, + "grad_norm": 0.1290162354707718, + "learning_rate": 0.00018023166023166024, + "loss": 0.3612, + "step": 134 + }, + { + "epoch": 0.0339772226766501, + "grad_norm": 0.1316012740135193, + "learning_rate": 0.00018007722007722007, + "loss": 0.22, + "step": 135 + }, + { + "epoch": 0.03422890580758824, + "grad_norm": 0.15092524886131287, + "learning_rate": 0.00017992277992277993, + "loss": 0.251, + "step": 136 + }, + { + "epoch": 0.034480588938526395, + "grad_norm": 0.12077280133962631, + "learning_rate": 0.00017976833976833978, + "loss": 0.2806, + "step": 137 + }, + { + "epoch": 0.03473227206946455, + "grad_norm": 0.13874058425426483, + "learning_rate": 0.00017961389961389964, + "loss": 0.2467, + "step": 138 + }, + { + "epoch": 0.03498395520040269, + "grad_norm": 0.11702004075050354, + "learning_rate": 0.00017945945945945947, + "loss": 0.2004, + "step": 139 + }, + { + "epoch": 0.03523563833134084, + "grad_norm": 0.10300116240978241, + "learning_rate": 0.0001793050193050193, + "loss": 0.1773, + "step": 140 + }, + { + "epoch": 0.03548732146227899, + "grad_norm": 0.14277376234531403, + "learning_rate": 0.00017915057915057916, + "loss": 0.2708, + "step": 141 + }, + { + "epoch": 0.03573900459321714, + "grad_norm": 0.17798127233982086, + "learning_rate": 0.000178996138996139, + "loss": 0.2128, + "step": 142 + }, + { + "epoch": 0.03599068772415529, + "grad_norm": 0.1409713327884674, + "learning_rate": 0.00017884169884169887, + "loss": 0.3095, + "step": 143 + }, + { + "epoch": 0.036242370855093436, + "grad_norm": 0.1393904983997345, + "learning_rate": 0.00017868725868725867, + "loss": 0.2049, + "step": 144 + }, + { + "epoch": 0.03649405398603159, + "grad_norm": 0.11879459768533707, + "learning_rate": 0.00017853281853281853, + "loss": 0.1933, + "step": 145 + }, + { + "epoch": 0.03674573711696973, + "grad_norm": 0.1240384429693222, + "learning_rate": 0.00017837837837837839, + "loss": 0.2328, + "step": 146 + }, + { + "epoch": 0.036997420247907885, + "grad_norm": 0.12959986925125122, + "learning_rate": 0.00017822393822393824, + "loss": 0.1893, + "step": 147 + }, + { + "epoch": 0.03724910337884603, + "grad_norm": 0.1437711864709854, + "learning_rate": 0.0001780694980694981, + "loss": 0.2927, + "step": 148 + }, + { + "epoch": 0.03750078650978418, + "grad_norm": 0.1455400139093399, + "learning_rate": 0.0001779150579150579, + "loss": 0.2359, + "step": 149 + }, + { + "epoch": 0.03775246964072233, + "grad_norm": 0.13055048882961273, + "learning_rate": 0.00017776061776061776, + "loss": 0.2546, + "step": 150 + }, + { + "epoch": 0.03800415277166048, + "grad_norm": 0.14180779457092285, + "learning_rate": 0.00017760617760617762, + "loss": 0.2913, + "step": 151 + }, + { + "epoch": 0.03825583590259863, + "grad_norm": 0.14623427391052246, + "learning_rate": 0.00017745173745173747, + "loss": 0.2033, + "step": 152 + }, + { + "epoch": 0.038507519033536775, + "grad_norm": 0.14906173944473267, + "learning_rate": 0.00017729729729729733, + "loss": 0.2934, + "step": 153 + }, + { + "epoch": 0.038759202164474926, + "grad_norm": 0.1549226939678192, + "learning_rate": 0.00017714285714285713, + "loss": 0.283, + "step": 154 + }, + { + "epoch": 0.03901088529541308, + "grad_norm": 0.13560202717781067, + "learning_rate": 0.000176988416988417, + "loss": 0.2977, + "step": 155 + }, + { + "epoch": 0.03926256842635122, + "grad_norm": 0.11980538070201874, + "learning_rate": 0.00017683397683397684, + "loss": 0.1923, + "step": 156 + }, + { + "epoch": 0.039514251557289375, + "grad_norm": 0.12517070770263672, + "learning_rate": 0.0001766795366795367, + "loss": 0.2006, + "step": 157 + }, + { + "epoch": 0.03976593468822752, + "grad_norm": 0.12278347462415695, + "learning_rate": 0.00017652509652509653, + "loss": 0.2104, + "step": 158 + }, + { + "epoch": 0.04001761781916567, + "grad_norm": 0.1440463364124298, + "learning_rate": 0.00017637065637065636, + "loss": 0.3079, + "step": 159 + }, + { + "epoch": 0.040269300950103816, + "grad_norm": 0.1633518487215042, + "learning_rate": 0.00017621621621621622, + "loss": 0.2346, + "step": 160 + }, + { + "epoch": 0.04052098408104197, + "grad_norm": 0.1538175791501999, + "learning_rate": 0.00017606177606177607, + "loss": 0.2297, + "step": 161 + }, + { + "epoch": 0.04077266721198012, + "grad_norm": 0.12572993338108063, + "learning_rate": 0.00017590733590733593, + "loss": 0.3292, + "step": 162 + }, + { + "epoch": 0.041024350342918264, + "grad_norm": 0.1295558214187622, + "learning_rate": 0.00017575289575289576, + "loss": 0.3377, + "step": 163 + }, + { + "epoch": 0.041276033473856416, + "grad_norm": 0.1346619874238968, + "learning_rate": 0.0001755984555984556, + "loss": 0.2098, + "step": 164 + }, + { + "epoch": 0.04152771660479456, + "grad_norm": 0.13213655352592468, + "learning_rate": 0.00017544401544401545, + "loss": 0.3119, + "step": 165 + }, + { + "epoch": 0.04177939973573271, + "grad_norm": 0.13211074471473694, + "learning_rate": 0.0001752895752895753, + "loss": 0.3585, + "step": 166 + }, + { + "epoch": 0.042031082866670864, + "grad_norm": 0.11970951408147812, + "learning_rate": 0.00017513513513513516, + "loss": 0.2239, + "step": 167 + }, + { + "epoch": 0.04228276599760901, + "grad_norm": 0.1309015303850174, + "learning_rate": 0.000174980694980695, + "loss": 0.1908, + "step": 168 + }, + { + "epoch": 0.04253444912854716, + "grad_norm": 0.11883830279111862, + "learning_rate": 0.00017482625482625482, + "loss": 0.2014, + "step": 169 + }, + { + "epoch": 0.042786132259485306, + "grad_norm": 0.12086677551269531, + "learning_rate": 0.00017467181467181468, + "loss": 0.1732, + "step": 170 + }, + { + "epoch": 0.04303781539042346, + "grad_norm": 0.12822604179382324, + "learning_rate": 0.00017451737451737453, + "loss": 0.2079, + "step": 171 + }, + { + "epoch": 0.0432894985213616, + "grad_norm": 0.13641922175884247, + "learning_rate": 0.00017436293436293436, + "loss": 0.2337, + "step": 172 + }, + { + "epoch": 0.043541181652299754, + "grad_norm": 0.12764209508895874, + "learning_rate": 0.00017420849420849422, + "loss": 0.2021, + "step": 173 + }, + { + "epoch": 0.043792864783237906, + "grad_norm": 0.14252115786075592, + "learning_rate": 0.00017405405405405405, + "loss": 0.2874, + "step": 174 + }, + { + "epoch": 0.04404454791417605, + "grad_norm": 0.1169133186340332, + "learning_rate": 0.0001738996138996139, + "loss": 0.1717, + "step": 175 + }, + { + "epoch": 0.0442962310451142, + "grad_norm": 0.11252434551715851, + "learning_rate": 0.00017374517374517376, + "loss": 0.1944, + "step": 176 + }, + { + "epoch": 0.04454791417605235, + "grad_norm": 0.1344681680202484, + "learning_rate": 0.0001735907335907336, + "loss": 0.2589, + "step": 177 + }, + { + "epoch": 0.0447995973069905, + "grad_norm": 0.11424121260643005, + "learning_rate": 0.00017343629343629345, + "loss": 0.1582, + "step": 178 + }, + { + "epoch": 0.04505128043792865, + "grad_norm": 0.13600754737854004, + "learning_rate": 0.00017328185328185328, + "loss": 0.3491, + "step": 179 + }, + { + "epoch": 0.045302963568866796, + "grad_norm": 0.12664635479450226, + "learning_rate": 0.00017312741312741314, + "loss": 0.2, + "step": 180 + }, + { + "epoch": 0.04555464669980495, + "grad_norm": 0.12114012241363525, + "learning_rate": 0.000172972972972973, + "loss": 0.205, + "step": 181 + }, + { + "epoch": 0.04580632983074309, + "grad_norm": 0.12991607189178467, + "learning_rate": 0.00017281853281853282, + "loss": 0.2114, + "step": 182 + }, + { + "epoch": 0.046058012961681244, + "grad_norm": 0.13017401099205017, + "learning_rate": 0.00017266409266409268, + "loss": 0.1884, + "step": 183 + }, + { + "epoch": 0.04630969609261939, + "grad_norm": 0.12459751218557358, + "learning_rate": 0.0001725096525096525, + "loss": 0.1899, + "step": 184 + }, + { + "epoch": 0.04656137922355754, + "grad_norm": 0.15462429821491241, + "learning_rate": 0.00017235521235521237, + "loss": 0.2084, + "step": 185 + }, + { + "epoch": 0.04681306235449569, + "grad_norm": 0.14958007633686066, + "learning_rate": 0.0001722007722007722, + "loss": 0.275, + "step": 186 + }, + { + "epoch": 0.04706474548543384, + "grad_norm": 0.12161766737699509, + "learning_rate": 0.00017204633204633205, + "loss": 0.19, + "step": 187 + }, + { + "epoch": 0.04731642861637199, + "grad_norm": 0.13190779089927673, + "learning_rate": 0.0001718918918918919, + "loss": 0.2285, + "step": 188 + }, + { + "epoch": 0.047568111747310134, + "grad_norm": 0.1161525547504425, + "learning_rate": 0.00017173745173745174, + "loss": 0.2082, + "step": 189 + }, + { + "epoch": 0.047819794878248285, + "grad_norm": 0.12745632231235504, + "learning_rate": 0.0001715830115830116, + "loss": 0.2195, + "step": 190 + }, + { + "epoch": 0.04807147800918644, + "grad_norm": 0.12998801469802856, + "learning_rate": 0.00017142857142857143, + "loss": 0.1844, + "step": 191 + }, + { + "epoch": 0.04832316114012458, + "grad_norm": 0.1154274195432663, + "learning_rate": 0.00017127413127413128, + "loss": 0.1905, + "step": 192 + }, + { + "epoch": 0.048574844271062734, + "grad_norm": 0.14668631553649902, + "learning_rate": 0.00017111969111969114, + "loss": 0.2191, + "step": 193 + }, + { + "epoch": 0.04882652740200088, + "grad_norm": 0.13747969269752502, + "learning_rate": 0.00017096525096525097, + "loss": 0.2624, + "step": 194 + }, + { + "epoch": 0.04907821053293903, + "grad_norm": 0.1509721875190735, + "learning_rate": 0.00017081081081081083, + "loss": 0.342, + "step": 195 + }, + { + "epoch": 0.04932989366387718, + "grad_norm": 0.12830862402915955, + "learning_rate": 0.00017065637065637065, + "loss": 0.2193, + "step": 196 + }, + { + "epoch": 0.04958157679481533, + "grad_norm": 0.13495340943336487, + "learning_rate": 0.0001705019305019305, + "loss": 0.2112, + "step": 197 + }, + { + "epoch": 0.04983325992575348, + "grad_norm": 0.11948922276496887, + "learning_rate": 0.00017034749034749037, + "loss": 0.1937, + "step": 198 + }, + { + "epoch": 0.05008494305669162, + "grad_norm": 0.12280004471540451, + "learning_rate": 0.0001701930501930502, + "loss": 0.2016, + "step": 199 + }, + { + "epoch": 0.050336626187629775, + "grad_norm": 0.15302538871765137, + "learning_rate": 0.00017003861003861005, + "loss": 0.3085, + "step": 200 + }, + { + "epoch": 0.05058830931856792, + "grad_norm": 0.11357412487268448, + "learning_rate": 0.00016988416988416988, + "loss": 0.2051, + "step": 201 + }, + { + "epoch": 0.05083999244950607, + "grad_norm": 0.12694703042507172, + "learning_rate": 0.00016972972972972974, + "loss": 0.2221, + "step": 202 + }, + { + "epoch": 0.051091675580444224, + "grad_norm": 0.1424601823091507, + "learning_rate": 0.0001695752895752896, + "loss": 0.2268, + "step": 203 + }, + { + "epoch": 0.05134335871138237, + "grad_norm": 0.13991156220436096, + "learning_rate": 0.00016942084942084943, + "loss": 0.2186, + "step": 204 + }, + { + "epoch": 0.05159504184232052, + "grad_norm": 0.1291760504245758, + "learning_rate": 0.00016926640926640926, + "loss": 0.3011, + "step": 205 + }, + { + "epoch": 0.051846724973258665, + "grad_norm": 0.13830232620239258, + "learning_rate": 0.00016911196911196911, + "loss": 0.2822, + "step": 206 + }, + { + "epoch": 0.05209840810419682, + "grad_norm": 0.1272973120212555, + "learning_rate": 0.00016895752895752897, + "loss": 0.1819, + "step": 207 + }, + { + "epoch": 0.05235009123513497, + "grad_norm": 0.12307526171207428, + "learning_rate": 0.00016880308880308883, + "loss": 0.1999, + "step": 208 + }, + { + "epoch": 0.05260177436607311, + "grad_norm": 0.1312309354543686, + "learning_rate": 0.00016864864864864866, + "loss": 0.2222, + "step": 209 + }, + { + "epoch": 0.052853457497011265, + "grad_norm": 0.14754608273506165, + "learning_rate": 0.0001684942084942085, + "loss": 0.1731, + "step": 210 + }, + { + "epoch": 0.05310514062794941, + "grad_norm": 0.14864255487918854, + "learning_rate": 0.00016833976833976834, + "loss": 0.2797, + "step": 211 + }, + { + "epoch": 0.05335682375888756, + "grad_norm": 0.1260637491941452, + "learning_rate": 0.0001681853281853282, + "loss": 0.1864, + "step": 212 + }, + { + "epoch": 0.053608506889825706, + "grad_norm": 0.12886057794094086, + "learning_rate": 0.00016803088803088806, + "loss": 0.2182, + "step": 213 + }, + { + "epoch": 0.05386019002076386, + "grad_norm": 0.14718972146511078, + "learning_rate": 0.0001678764478764479, + "loss": 0.22, + "step": 214 + }, + { + "epoch": 0.05411187315170201, + "grad_norm": 0.13571980595588684, + "learning_rate": 0.00016772200772200772, + "loss": 0.2447, + "step": 215 + }, + { + "epoch": 0.054363556282640155, + "grad_norm": 0.11550338566303253, + "learning_rate": 0.00016756756756756757, + "loss": 0.1639, + "step": 216 + }, + { + "epoch": 0.054615239413578306, + "grad_norm": 0.14989754557609558, + "learning_rate": 0.00016741312741312743, + "loss": 0.3188, + "step": 217 + }, + { + "epoch": 0.05486692254451645, + "grad_norm": 0.13520431518554688, + "learning_rate": 0.0001672586872586873, + "loss": 0.2864, + "step": 218 + }, + { + "epoch": 0.0551186056754546, + "grad_norm": 0.12954586744308472, + "learning_rate": 0.0001671042471042471, + "loss": 0.1944, + "step": 219 + }, + { + "epoch": 0.055370288806392755, + "grad_norm": 0.15795505046844482, + "learning_rate": 0.00016694980694980695, + "loss": 0.2733, + "step": 220 + }, + { + "epoch": 0.0556219719373309, + "grad_norm": 0.12952439486980438, + "learning_rate": 0.0001667953667953668, + "loss": 0.2573, + "step": 221 + }, + { + "epoch": 0.05587365506826905, + "grad_norm": 0.11759003251791, + "learning_rate": 0.00016664092664092666, + "loss": 0.1674, + "step": 222 + }, + { + "epoch": 0.056125338199207196, + "grad_norm": 0.11585386097431183, + "learning_rate": 0.00016648648648648652, + "loss": 0.2013, + "step": 223 + }, + { + "epoch": 0.05637702133014535, + "grad_norm": 0.11154724657535553, + "learning_rate": 0.00016633204633204632, + "loss": 0.1814, + "step": 224 + }, + { + "epoch": 0.05662870446108349, + "grad_norm": 0.1463312804698944, + "learning_rate": 0.00016617760617760618, + "loss": 0.2198, + "step": 225 + }, + { + "epoch": 0.056880387592021645, + "grad_norm": 0.149198517203331, + "learning_rate": 0.00016602316602316603, + "loss": 0.3223, + "step": 226 + }, + { + "epoch": 0.057132070722959796, + "grad_norm": 0.1325506716966629, + "learning_rate": 0.0001658687258687259, + "loss": 0.2718, + "step": 227 + }, + { + "epoch": 0.05738375385389794, + "grad_norm": 0.15414372086524963, + "learning_rate": 0.00016571428571428575, + "loss": 0.2661, + "step": 228 + }, + { + "epoch": 0.05763543698483609, + "grad_norm": 0.12570172548294067, + "learning_rate": 0.00016555984555984555, + "loss": 0.2167, + "step": 229 + }, + { + "epoch": 0.05788712011577424, + "grad_norm": 0.1413959413766861, + "learning_rate": 0.0001654054054054054, + "loss": 0.2491, + "step": 230 + }, + { + "epoch": 0.05813880324671239, + "grad_norm": 0.13764715194702148, + "learning_rate": 0.00016525096525096526, + "loss": 0.2967, + "step": 231 + }, + { + "epoch": 0.05839048637765054, + "grad_norm": 0.1098836362361908, + "learning_rate": 0.00016509652509652512, + "loss": 0.1922, + "step": 232 + }, + { + "epoch": 0.058642169508588686, + "grad_norm": 0.11060483753681183, + "learning_rate": 0.00016494208494208495, + "loss": 0.1972, + "step": 233 + }, + { + "epoch": 0.05889385263952684, + "grad_norm": 0.11183371394872665, + "learning_rate": 0.00016478764478764478, + "loss": 0.2155, + "step": 234 + }, + { + "epoch": 0.05914553577046498, + "grad_norm": 0.14082889258861542, + "learning_rate": 0.00016463320463320464, + "loss": 0.1822, + "step": 235 + }, + { + "epoch": 0.059397218901403134, + "grad_norm": 0.1429906189441681, + "learning_rate": 0.0001644787644787645, + "loss": 0.262, + "step": 236 + }, + { + "epoch": 0.05964890203234128, + "grad_norm": 0.12406495213508606, + "learning_rate": 0.00016432432432432435, + "loss": 0.1712, + "step": 237 + }, + { + "epoch": 0.05990058516327943, + "grad_norm": 0.1340409517288208, + "learning_rate": 0.00016416988416988418, + "loss": 0.2401, + "step": 238 + }, + { + "epoch": 0.06015226829421758, + "grad_norm": 0.12884925305843353, + "learning_rate": 0.000164015444015444, + "loss": 0.2122, + "step": 239 + }, + { + "epoch": 0.06040395142515573, + "grad_norm": 0.12663012742996216, + "learning_rate": 0.00016386100386100386, + "loss": 0.2131, + "step": 240 + }, + { + "epoch": 0.06065563455609388, + "grad_norm": 0.13400785624980927, + "learning_rate": 0.00016370656370656372, + "loss": 0.2337, + "step": 241 + }, + { + "epoch": 0.060907317687032024, + "grad_norm": 0.14849476516246796, + "learning_rate": 0.00016355212355212358, + "loss": 0.2776, + "step": 242 + }, + { + "epoch": 0.061159000817970176, + "grad_norm": 0.12700201570987701, + "learning_rate": 0.0001633976833976834, + "loss": 0.1762, + "step": 243 + }, + { + "epoch": 0.06141068394890833, + "grad_norm": 0.11481945961713791, + "learning_rate": 0.00016324324324324324, + "loss": 0.2156, + "step": 244 + }, + { + "epoch": 0.06166236707984647, + "grad_norm": 0.12354963272809982, + "learning_rate": 0.0001630888030888031, + "loss": 0.1927, + "step": 245 + }, + { + "epoch": 0.061914050210784624, + "grad_norm": 0.12551970779895782, + "learning_rate": 0.00016293436293436295, + "loss": 0.2067, + "step": 246 + }, + { + "epoch": 0.06216573334172277, + "grad_norm": 0.1234751045703888, + "learning_rate": 0.00016277992277992278, + "loss": 0.2009, + "step": 247 + }, + { + "epoch": 0.06241741647266092, + "grad_norm": 0.10748082399368286, + "learning_rate": 0.00016262548262548264, + "loss": 0.1657, + "step": 248 + }, + { + "epoch": 0.06266909960359907, + "grad_norm": 0.14514018595218658, + "learning_rate": 0.00016247104247104247, + "loss": 0.2835, + "step": 249 + }, + { + "epoch": 0.06292078273453722, + "grad_norm": 0.1324596405029297, + "learning_rate": 0.00016231660231660232, + "loss": 0.2286, + "step": 250 + }, + { + "epoch": 0.06317246586547537, + "grad_norm": 0.12410667538642883, + "learning_rate": 0.00016216216216216218, + "loss": 0.244, + "step": 251 + }, + { + "epoch": 0.06342414899641352, + "grad_norm": 0.12856566905975342, + "learning_rate": 0.000162007722007722, + "loss": 0.2299, + "step": 252 + }, + { + "epoch": 0.06367583212735166, + "grad_norm": 0.1232319325208664, + "learning_rate": 0.00016185328185328187, + "loss": 0.2668, + "step": 253 + }, + { + "epoch": 0.06392751525828981, + "grad_norm": 0.13640795648097992, + "learning_rate": 0.0001616988416988417, + "loss": 0.2423, + "step": 254 + }, + { + "epoch": 0.06417919838922796, + "grad_norm": 0.13468897342681885, + "learning_rate": 0.00016154440154440155, + "loss": 0.2312, + "step": 255 + }, + { + "epoch": 0.06443088152016611, + "grad_norm": 0.11596588045358658, + "learning_rate": 0.0001613899613899614, + "loss": 0.168, + "step": 256 + }, + { + "epoch": 0.06468256465110427, + "grad_norm": 0.118974968791008, + "learning_rate": 0.00016123552123552124, + "loss": 0.189, + "step": 257 + }, + { + "epoch": 0.0649342477820424, + "grad_norm": 0.1390894055366516, + "learning_rate": 0.0001610810810810811, + "loss": 0.2975, + "step": 258 + }, + { + "epoch": 0.06518593091298056, + "grad_norm": 0.13288350403308868, + "learning_rate": 0.00016092664092664093, + "loss": 0.2517, + "step": 259 + }, + { + "epoch": 0.06543761404391871, + "grad_norm": 0.12513595819473267, + "learning_rate": 0.00016077220077220078, + "loss": 0.2558, + "step": 260 + }, + { + "epoch": 0.06568929717485686, + "grad_norm": 0.13168072700500488, + "learning_rate": 0.0001606177606177606, + "loss": 0.1936, + "step": 261 + }, + { + "epoch": 0.06594098030579501, + "grad_norm": 0.12524648010730743, + "learning_rate": 0.00016046332046332047, + "loss": 0.2277, + "step": 262 + }, + { + "epoch": 0.06619266343673315, + "grad_norm": 0.11089562624692917, + "learning_rate": 0.00016030888030888033, + "loss": 0.1843, + "step": 263 + }, + { + "epoch": 0.0664443465676713, + "grad_norm": 0.09994406998157501, + "learning_rate": 0.00016015444015444016, + "loss": 0.1565, + "step": 264 + }, + { + "epoch": 0.06669602969860945, + "grad_norm": 0.13970190286636353, + "learning_rate": 0.00016, + "loss": 0.2534, + "step": 265 + }, + { + "epoch": 0.0669477128295476, + "grad_norm": 0.12788502871990204, + "learning_rate": 0.00015984555984555984, + "loss": 0.2384, + "step": 266 + }, + { + "epoch": 0.06719939596048574, + "grad_norm": 0.12347918003797531, + "learning_rate": 0.0001596911196911197, + "loss": 0.2095, + "step": 267 + }, + { + "epoch": 0.0674510790914239, + "grad_norm": 0.1359950602054596, + "learning_rate": 0.00015953667953667956, + "loss": 0.2677, + "step": 268 + }, + { + "epoch": 0.06770276222236205, + "grad_norm": 0.12793773412704468, + "learning_rate": 0.00015938223938223939, + "loss": 0.1827, + "step": 269 + }, + { + "epoch": 0.0679544453533002, + "grad_norm": 0.1240396648645401, + "learning_rate": 0.00015922779922779924, + "loss": 0.214, + "step": 270 + }, + { + "epoch": 0.06820612848423835, + "grad_norm": 0.14594678580760956, + "learning_rate": 0.00015907335907335907, + "loss": 0.2814, + "step": 271 + }, + { + "epoch": 0.06845781161517649, + "grad_norm": 0.12793748080730438, + "learning_rate": 0.00015891891891891893, + "loss": 0.2221, + "step": 272 + }, + { + "epoch": 0.06870949474611464, + "grad_norm": 0.11719738692045212, + "learning_rate": 0.00015876447876447879, + "loss": 0.1852, + "step": 273 + }, + { + "epoch": 0.06896117787705279, + "grad_norm": 0.12055392563343048, + "learning_rate": 0.00015861003861003862, + "loss": 0.1868, + "step": 274 + }, + { + "epoch": 0.06921286100799094, + "grad_norm": 0.12247222661972046, + "learning_rate": 0.00015845559845559845, + "loss": 0.2001, + "step": 275 + }, + { + "epoch": 0.0694645441389291, + "grad_norm": 0.1345289647579193, + "learning_rate": 0.0001583011583011583, + "loss": 0.1731, + "step": 276 + }, + { + "epoch": 0.06971622726986723, + "grad_norm": 0.1287080943584442, + "learning_rate": 0.00015814671814671816, + "loss": 0.2373, + "step": 277 + }, + { + "epoch": 0.06996791040080538, + "grad_norm": 0.1344958245754242, + "learning_rate": 0.00015799227799227802, + "loss": 0.2439, + "step": 278 + }, + { + "epoch": 0.07021959353174353, + "grad_norm": 0.11678414791822433, + "learning_rate": 0.00015783783783783785, + "loss": 0.191, + "step": 279 + }, + { + "epoch": 0.07047127666268169, + "grad_norm": 0.16685999929904938, + "learning_rate": 0.00015768339768339767, + "loss": 0.2591, + "step": 280 + }, + { + "epoch": 0.07072295979361984, + "grad_norm": 0.1351940929889679, + "learning_rate": 0.00015752895752895753, + "loss": 0.194, + "step": 281 + }, + { + "epoch": 0.07097464292455798, + "grad_norm": 0.16132301092147827, + "learning_rate": 0.0001573745173745174, + "loss": 0.2753, + "step": 282 + }, + { + "epoch": 0.07122632605549613, + "grad_norm": 0.14595595002174377, + "learning_rate": 0.00015722007722007725, + "loss": 0.2545, + "step": 283 + }, + { + "epoch": 0.07147800918643428, + "grad_norm": 0.14386332035064697, + "learning_rate": 0.00015706563706563707, + "loss": 0.3216, + "step": 284 + }, + { + "epoch": 0.07172969231737243, + "grad_norm": 0.13627417385578156, + "learning_rate": 0.0001569111969111969, + "loss": 0.2481, + "step": 285 + }, + { + "epoch": 0.07198137544831058, + "grad_norm": 0.12509195506572723, + "learning_rate": 0.00015675675675675676, + "loss": 0.2708, + "step": 286 + }, + { + "epoch": 0.07223305857924872, + "grad_norm": 0.10608310252428055, + "learning_rate": 0.00015660231660231662, + "loss": 0.1495, + "step": 287 + }, + { + "epoch": 0.07248474171018687, + "grad_norm": 0.12942315638065338, + "learning_rate": 0.00015644787644787647, + "loss": 0.2043, + "step": 288 + }, + { + "epoch": 0.07273642484112502, + "grad_norm": 0.10886157304048538, + "learning_rate": 0.00015629343629343628, + "loss": 0.1886, + "step": 289 + }, + { + "epoch": 0.07298810797206318, + "grad_norm": 0.12388721108436584, + "learning_rate": 0.00015613899613899613, + "loss": 0.2647, + "step": 290 + }, + { + "epoch": 0.07323979110300133, + "grad_norm": 0.1376280039548874, + "learning_rate": 0.000155984555984556, + "loss": 0.226, + "step": 291 + }, + { + "epoch": 0.07349147423393947, + "grad_norm": 0.13396234810352325, + "learning_rate": 0.00015583011583011585, + "loss": 0.2593, + "step": 292 + }, + { + "epoch": 0.07374315736487762, + "grad_norm": 0.13066333532333374, + "learning_rate": 0.0001556756756756757, + "loss": 0.2203, + "step": 293 + }, + { + "epoch": 0.07399484049581577, + "grad_norm": 0.13232409954071045, + "learning_rate": 0.0001555212355212355, + "loss": 0.2662, + "step": 294 + }, + { + "epoch": 0.07424652362675392, + "grad_norm": 0.12771448493003845, + "learning_rate": 0.00015536679536679536, + "loss": 0.2022, + "step": 295 + }, + { + "epoch": 0.07449820675769206, + "grad_norm": 0.12821722030639648, + "learning_rate": 0.00015521235521235522, + "loss": 0.1917, + "step": 296 + }, + { + "epoch": 0.07474988988863021, + "grad_norm": 0.1343013495206833, + "learning_rate": 0.00015505791505791508, + "loss": 0.2246, + "step": 297 + }, + { + "epoch": 0.07500157301956836, + "grad_norm": 0.14461186528205872, + "learning_rate": 0.00015490347490347493, + "loss": 0.2294, + "step": 298 + }, + { + "epoch": 0.07525325615050651, + "grad_norm": 0.11262819170951843, + "learning_rate": 0.00015474903474903474, + "loss": 0.1733, + "step": 299 + }, + { + "epoch": 0.07550493928144467, + "grad_norm": 0.12715961039066315, + "learning_rate": 0.0001545945945945946, + "loss": 0.2303, + "step": 300 + }, + { + "epoch": 0.0757566224123828, + "grad_norm": 0.1332138478755951, + "learning_rate": 0.00015444015444015445, + "loss": 0.3789, + "step": 301 + }, + { + "epoch": 0.07600830554332096, + "grad_norm": 0.11156772077083588, + "learning_rate": 0.0001542857142857143, + "loss": 0.1807, + "step": 302 + }, + { + "epoch": 0.07625998867425911, + "grad_norm": 0.10366728156805038, + "learning_rate": 0.00015413127413127414, + "loss": 0.1558, + "step": 303 + }, + { + "epoch": 0.07651167180519726, + "grad_norm": 0.1493196338415146, + "learning_rate": 0.00015397683397683397, + "loss": 0.34, + "step": 304 + }, + { + "epoch": 0.07676335493613541, + "grad_norm": 0.11116340756416321, + "learning_rate": 0.00015382239382239382, + "loss": 0.2014, + "step": 305 + }, + { + "epoch": 0.07701503806707355, + "grad_norm": 0.11475585401058197, + "learning_rate": 0.00015366795366795368, + "loss": 0.1824, + "step": 306 + }, + { + "epoch": 0.0772667211980117, + "grad_norm": 0.14417363703250885, + "learning_rate": 0.00015351351351351354, + "loss": 0.323, + "step": 307 + }, + { + "epoch": 0.07751840432894985, + "grad_norm": 0.1290784329175949, + "learning_rate": 0.00015335907335907337, + "loss": 0.2086, + "step": 308 + }, + { + "epoch": 0.077770087459888, + "grad_norm": 0.158465176820755, + "learning_rate": 0.0001532046332046332, + "loss": 0.2001, + "step": 309 + }, + { + "epoch": 0.07802177059082616, + "grad_norm": 0.10680986940860748, + "learning_rate": 0.00015305019305019305, + "loss": 0.1683, + "step": 310 + }, + { + "epoch": 0.0782734537217643, + "grad_norm": 0.12347064912319183, + "learning_rate": 0.0001528957528957529, + "loss": 0.2438, + "step": 311 + }, + { + "epoch": 0.07852513685270245, + "grad_norm": 0.1270691603422165, + "learning_rate": 0.00015274131274131277, + "loss": 0.2307, + "step": 312 + }, + { + "epoch": 0.0787768199836406, + "grad_norm": 0.12614499032497406, + "learning_rate": 0.0001525868725868726, + "loss": 0.3015, + "step": 313 + }, + { + "epoch": 0.07902850311457875, + "grad_norm": 0.11198928952217102, + "learning_rate": 0.00015243243243243243, + "loss": 0.1948, + "step": 314 + }, + { + "epoch": 0.0792801862455169, + "grad_norm": 0.11850691586732864, + "learning_rate": 0.00015227799227799228, + "loss": 0.2059, + "step": 315 + }, + { + "epoch": 0.07953186937645504, + "grad_norm": 0.13141778111457825, + "learning_rate": 0.00015212355212355214, + "loss": 0.2173, + "step": 316 + }, + { + "epoch": 0.07978355250739319, + "grad_norm": 0.12298451364040375, + "learning_rate": 0.00015196911196911197, + "loss": 0.2187, + "step": 317 + }, + { + "epoch": 0.08003523563833134, + "grad_norm": 0.13927814364433289, + "learning_rate": 0.00015181467181467183, + "loss": 0.2932, + "step": 318 + }, + { + "epoch": 0.0802869187692695, + "grad_norm": 0.13508445024490356, + "learning_rate": 0.00015166023166023166, + "loss": 0.2317, + "step": 319 + }, + { + "epoch": 0.08053860190020763, + "grad_norm": 0.11320673674345016, + "learning_rate": 0.0001515057915057915, + "loss": 0.1927, + "step": 320 + }, + { + "epoch": 0.08079028503114578, + "grad_norm": 0.11665111035108566, + "learning_rate": 0.00015135135135135137, + "loss": 0.2072, + "step": 321 + }, + { + "epoch": 0.08104196816208394, + "grad_norm": 0.11465509235858917, + "learning_rate": 0.0001511969111969112, + "loss": 0.194, + "step": 322 + }, + { + "epoch": 0.08129365129302209, + "grad_norm": 0.14008402824401855, + "learning_rate": 0.00015104247104247106, + "loss": 0.3015, + "step": 323 + }, + { + "epoch": 0.08154533442396024, + "grad_norm": 0.12534737586975098, + "learning_rate": 0.00015088803088803088, + "loss": 0.2716, + "step": 324 + }, + { + "epoch": 0.08179701755489838, + "grad_norm": 0.1350509226322174, + "learning_rate": 0.00015073359073359074, + "loss": 0.2485, + "step": 325 + }, + { + "epoch": 0.08204870068583653, + "grad_norm": 0.13865070044994354, + "learning_rate": 0.0001505791505791506, + "loss": 0.2999, + "step": 326 + }, + { + "epoch": 0.08230038381677468, + "grad_norm": 0.13422353565692902, + "learning_rate": 0.00015042471042471043, + "loss": 0.3132, + "step": 327 + }, + { + "epoch": 0.08255206694771283, + "grad_norm": 0.0934542641043663, + "learning_rate": 0.00015027027027027028, + "loss": 0.1455, + "step": 328 + }, + { + "epoch": 0.08280375007865098, + "grad_norm": 0.10283880680799484, + "learning_rate": 0.00015011583011583011, + "loss": 0.1679, + "step": 329 + }, + { + "epoch": 0.08305543320958912, + "grad_norm": 0.11898140609264374, + "learning_rate": 0.00014996138996138997, + "loss": 0.1911, + "step": 330 + }, + { + "epoch": 0.08330711634052727, + "grad_norm": 0.11024250090122223, + "learning_rate": 0.0001498069498069498, + "loss": 0.1841, + "step": 331 + }, + { + "epoch": 0.08355879947146543, + "grad_norm": 0.14024034142494202, + "learning_rate": 0.00014965250965250966, + "loss": 0.1819, + "step": 332 + }, + { + "epoch": 0.08381048260240358, + "grad_norm": 0.13840073347091675, + "learning_rate": 0.00014949806949806951, + "loss": 0.2853, + "step": 333 + }, + { + "epoch": 0.08406216573334173, + "grad_norm": 0.1443752646446228, + "learning_rate": 0.00014934362934362934, + "loss": 0.3836, + "step": 334 + }, + { + "epoch": 0.08431384886427987, + "grad_norm": 0.10499413311481476, + "learning_rate": 0.0001491891891891892, + "loss": 0.1666, + "step": 335 + }, + { + "epoch": 0.08456553199521802, + "grad_norm": 0.11997657269239426, + "learning_rate": 0.00014903474903474903, + "loss": 0.1694, + "step": 336 + }, + { + "epoch": 0.08481721512615617, + "grad_norm": 0.13212887942790985, + "learning_rate": 0.0001488803088803089, + "loss": 0.3409, + "step": 337 + }, + { + "epoch": 0.08506889825709432, + "grad_norm": 0.140362948179245, + "learning_rate": 0.00014872586872586874, + "loss": 0.2972, + "step": 338 + }, + { + "epoch": 0.08532058138803247, + "grad_norm": 0.1353239268064499, + "learning_rate": 0.00014857142857142857, + "loss": 0.2951, + "step": 339 + }, + { + "epoch": 0.08557226451897061, + "grad_norm": 0.11341475695371628, + "learning_rate": 0.00014841698841698843, + "loss": 0.1579, + "step": 340 + }, + { + "epoch": 0.08582394764990876, + "grad_norm": 0.13959603011608124, + "learning_rate": 0.00014826254826254826, + "loss": 0.169, + "step": 341 + }, + { + "epoch": 0.08607563078084692, + "grad_norm": 0.11672891676425934, + "learning_rate": 0.00014810810810810812, + "loss": 0.1564, + "step": 342 + }, + { + "epoch": 0.08632731391178507, + "grad_norm": 0.1251240372657776, + "learning_rate": 0.00014795366795366797, + "loss": 0.2113, + "step": 343 + }, + { + "epoch": 0.0865789970427232, + "grad_norm": 0.13342410326004028, + "learning_rate": 0.0001477992277992278, + "loss": 0.1964, + "step": 344 + }, + { + "epoch": 0.08683068017366136, + "grad_norm": 0.1275550276041031, + "learning_rate": 0.00014764478764478763, + "loss": 0.2899, + "step": 345 + }, + { + "epoch": 0.08708236330459951, + "grad_norm": 0.12599168717861176, + "learning_rate": 0.0001474903474903475, + "loss": 0.172, + "step": 346 + }, + { + "epoch": 0.08733404643553766, + "grad_norm": 0.11375829577445984, + "learning_rate": 0.00014733590733590735, + "loss": 0.1865, + "step": 347 + }, + { + "epoch": 0.08758572956647581, + "grad_norm": 0.13509558141231537, + "learning_rate": 0.0001471814671814672, + "loss": 0.2954, + "step": 348 + }, + { + "epoch": 0.08783741269741395, + "grad_norm": 0.1293509155511856, + "learning_rate": 0.00014702702702702703, + "loss": 0.2597, + "step": 349 + }, + { + "epoch": 0.0880890958283521, + "grad_norm": 0.12318065017461777, + "learning_rate": 0.00014687258687258686, + "loss": 0.1809, + "step": 350 + }, + { + "epoch": 0.08834077895929025, + "grad_norm": 0.1373419165611267, + "learning_rate": 0.00014671814671814672, + "loss": 0.3222, + "step": 351 + }, + { + "epoch": 0.0885924620902284, + "grad_norm": 0.12368537485599518, + "learning_rate": 0.00014656370656370658, + "loss": 0.1968, + "step": 352 + }, + { + "epoch": 0.08884414522116656, + "grad_norm": 0.1034383699297905, + "learning_rate": 0.00014640926640926643, + "loss": 0.1579, + "step": 353 + }, + { + "epoch": 0.0890958283521047, + "grad_norm": 0.12372983247041702, + "learning_rate": 0.00014625482625482626, + "loss": 0.1748, + "step": 354 + }, + { + "epoch": 0.08934751148304285, + "grad_norm": 0.12052860110998154, + "learning_rate": 0.0001461003861003861, + "loss": 0.2091, + "step": 355 + }, + { + "epoch": 0.089599194613981, + "grad_norm": 0.1348121613264084, + "learning_rate": 0.00014594594594594595, + "loss": 0.2714, + "step": 356 + }, + { + "epoch": 0.08985087774491915, + "grad_norm": 0.12104601413011551, + "learning_rate": 0.0001457915057915058, + "loss": 0.2022, + "step": 357 + }, + { + "epoch": 0.0901025608758573, + "grad_norm": 0.13189780712127686, + "learning_rate": 0.00014563706563706566, + "loss": 0.2654, + "step": 358 + }, + { + "epoch": 0.09035424400679544, + "grad_norm": 0.11207465827465057, + "learning_rate": 0.00014548262548262547, + "loss": 0.1875, + "step": 359 + }, + { + "epoch": 0.09060592713773359, + "grad_norm": 0.12804561853408813, + "learning_rate": 0.00014532818532818532, + "loss": 0.2049, + "step": 360 + }, + { + "epoch": 0.09085761026867174, + "grad_norm": 0.12313145399093628, + "learning_rate": 0.00014517374517374518, + "loss": 0.1626, + "step": 361 + }, + { + "epoch": 0.0911092933996099, + "grad_norm": 0.1248214989900589, + "learning_rate": 0.00014501930501930504, + "loss": 0.2413, + "step": 362 + }, + { + "epoch": 0.09136097653054805, + "grad_norm": 0.12343775480985641, + "learning_rate": 0.0001448648648648649, + "loss": 0.166, + "step": 363 + }, + { + "epoch": 0.09161265966148618, + "grad_norm": 0.12935423851013184, + "learning_rate": 0.0001447104247104247, + "loss": 0.2848, + "step": 364 + }, + { + "epoch": 0.09186434279242434, + "grad_norm": 0.11857825517654419, + "learning_rate": 0.00014455598455598455, + "loss": 0.1879, + "step": 365 + }, + { + "epoch": 0.09211602592336249, + "grad_norm": 0.12521713972091675, + "learning_rate": 0.0001444015444015444, + "loss": 0.2061, + "step": 366 + }, + { + "epoch": 0.09236770905430064, + "grad_norm": 0.15852835774421692, + "learning_rate": 0.00014424710424710427, + "loss": 0.3303, + "step": 367 + }, + { + "epoch": 0.09261939218523878, + "grad_norm": 0.11650358885526657, + "learning_rate": 0.00014409266409266412, + "loss": 0.2002, + "step": 368 + }, + { + "epoch": 0.09287107531617693, + "grad_norm": 0.1254633069038391, + "learning_rate": 0.00014393822393822392, + "loss": 0.2705, + "step": 369 + }, + { + "epoch": 0.09312275844711508, + "grad_norm": 0.1139984056353569, + "learning_rate": 0.00014378378378378378, + "loss": 0.1862, + "step": 370 + }, + { + "epoch": 0.09337444157805323, + "grad_norm": 0.12517507374286652, + "learning_rate": 0.00014362934362934364, + "loss": 0.1893, + "step": 371 + }, + { + "epoch": 0.09362612470899138, + "grad_norm": 0.13221777975559235, + "learning_rate": 0.0001434749034749035, + "loss": 0.2496, + "step": 372 + }, + { + "epoch": 0.09387780783992952, + "grad_norm": 0.1270243227481842, + "learning_rate": 0.00014332046332046332, + "loss": 0.2862, + "step": 373 + }, + { + "epoch": 0.09412949097086767, + "grad_norm": 0.11766324937343597, + "learning_rate": 0.00014316602316602315, + "loss": 0.1933, + "step": 374 + }, + { + "epoch": 0.09438117410180583, + "grad_norm": 0.1286013275384903, + "learning_rate": 0.000143011583011583, + "loss": 0.2742, + "step": 375 + }, + { + "epoch": 0.09463285723274398, + "grad_norm": 0.1172587051987648, + "learning_rate": 0.00014285714285714287, + "loss": 0.2012, + "step": 376 + }, + { + "epoch": 0.09488454036368213, + "grad_norm": 0.13332095742225647, + "learning_rate": 0.00014270270270270272, + "loss": 0.2442, + "step": 377 + }, + { + "epoch": 0.09513622349462027, + "grad_norm": 0.11457065492868423, + "learning_rate": 0.00014254826254826255, + "loss": 0.178, + "step": 378 + }, + { + "epoch": 0.09538790662555842, + "grad_norm": 0.1414901614189148, + "learning_rate": 0.00014239382239382238, + "loss": 0.2906, + "step": 379 + }, + { + "epoch": 0.09563958975649657, + "grad_norm": 0.1505407691001892, + "learning_rate": 0.00014223938223938224, + "loss": 0.3076, + "step": 380 + }, + { + "epoch": 0.09589127288743472, + "grad_norm": 0.12012407183647156, + "learning_rate": 0.0001420849420849421, + "loss": 0.2062, + "step": 381 + }, + { + "epoch": 0.09614295601837287, + "grad_norm": 0.1284686028957367, + "learning_rate": 0.00014193050193050195, + "loss": 0.2361, + "step": 382 + }, + { + "epoch": 0.09639463914931101, + "grad_norm": 0.11715728789567947, + "learning_rate": 0.00014177606177606178, + "loss": 0.1804, + "step": 383 + }, + { + "epoch": 0.09664632228024916, + "grad_norm": 0.13079042732715607, + "learning_rate": 0.00014162162162162161, + "loss": 0.2358, + "step": 384 + }, + { + "epoch": 0.09689800541118732, + "grad_norm": 0.10525856167078018, + "learning_rate": 0.00014146718146718147, + "loss": 0.1604, + "step": 385 + }, + { + "epoch": 0.09714968854212547, + "grad_norm": 0.12698408961296082, + "learning_rate": 0.00014131274131274133, + "loss": 0.1978, + "step": 386 + }, + { + "epoch": 0.09740137167306362, + "grad_norm": 0.13074257969856262, + "learning_rate": 0.00014115830115830118, + "loss": 0.2839, + "step": 387 + }, + { + "epoch": 0.09765305480400176, + "grad_norm": 0.11895257234573364, + "learning_rate": 0.00014100386100386101, + "loss": 0.1817, + "step": 388 + }, + { + "epoch": 0.09790473793493991, + "grad_norm": 0.14683619141578674, + "learning_rate": 0.00014084942084942084, + "loss": 0.1711, + "step": 389 + }, + { + "epoch": 0.09815642106587806, + "grad_norm": 0.14826332032680511, + "learning_rate": 0.0001406949806949807, + "loss": 0.3603, + "step": 390 + }, + { + "epoch": 0.09840810419681621, + "grad_norm": 0.13553692400455475, + "learning_rate": 0.00014054054054054056, + "loss": 0.2651, + "step": 391 + }, + { + "epoch": 0.09865978732775436, + "grad_norm": 0.13319283723831177, + "learning_rate": 0.0001403861003861004, + "loss": 0.2308, + "step": 392 + }, + { + "epoch": 0.0989114704586925, + "grad_norm": 0.11546744406223297, + "learning_rate": 0.00014023166023166024, + "loss": 0.2007, + "step": 393 + }, + { + "epoch": 0.09916315358963065, + "grad_norm": 0.1334381252527237, + "learning_rate": 0.00014007722007722007, + "loss": 0.2167, + "step": 394 + }, + { + "epoch": 0.0994148367205688, + "grad_norm": 0.11112739145755768, + "learning_rate": 0.00013992277992277993, + "loss": 0.1792, + "step": 395 + }, + { + "epoch": 0.09966651985150696, + "grad_norm": 0.13586269319057465, + "learning_rate": 0.00013976833976833979, + "loss": 0.2589, + "step": 396 + }, + { + "epoch": 0.0999182029824451, + "grad_norm": 0.13598954677581787, + "learning_rate": 0.00013961389961389962, + "loss": 0.3028, + "step": 397 + }, + { + "epoch": 0.10016988611338325, + "grad_norm": 0.11205045878887177, + "learning_rate": 0.00013945945945945947, + "loss": 0.2063, + "step": 398 + }, + { + "epoch": 0.1004215692443214, + "grad_norm": 0.12628450989723206, + "learning_rate": 0.0001393050193050193, + "loss": 0.1997, + "step": 399 + }, + { + "epoch": 0.10067325237525955, + "grad_norm": 0.12549594044685364, + "learning_rate": 0.00013915057915057916, + "loss": 0.2032, + "step": 400 + }, + { + "epoch": 0.1009249355061977, + "grad_norm": 0.13437925279140472, + "learning_rate": 0.00013899613899613902, + "loss": 0.1842, + "step": 401 + }, + { + "epoch": 0.10117661863713584, + "grad_norm": 0.12974749505519867, + "learning_rate": 0.00013884169884169885, + "loss": 0.2015, + "step": 402 + }, + { + "epoch": 0.10142830176807399, + "grad_norm": 0.11364075541496277, + "learning_rate": 0.0001386872586872587, + "loss": 0.1682, + "step": 403 + }, + { + "epoch": 0.10167998489901214, + "grad_norm": 0.12087921053171158, + "learning_rate": 0.00013853281853281853, + "loss": 0.286, + "step": 404 + }, + { + "epoch": 0.1019316680299503, + "grad_norm": 0.14084592461585999, + "learning_rate": 0.0001383783783783784, + "loss": 0.3225, + "step": 405 + }, + { + "epoch": 0.10218335116088845, + "grad_norm": 0.11849649250507355, + "learning_rate": 0.00013822393822393822, + "loss": 0.1813, + "step": 406 + }, + { + "epoch": 0.10243503429182658, + "grad_norm": 0.10580828785896301, + "learning_rate": 0.00013806949806949808, + "loss": 0.1658, + "step": 407 + }, + { + "epoch": 0.10268671742276474, + "grad_norm": 0.13645821809768677, + "learning_rate": 0.00013791505791505793, + "loss": 0.2622, + "step": 408 + }, + { + "epoch": 0.10293840055370289, + "grad_norm": 0.1310422271490097, + "learning_rate": 0.00013776061776061776, + "loss": 0.2926, + "step": 409 + }, + { + "epoch": 0.10319008368464104, + "grad_norm": 0.13212424516677856, + "learning_rate": 0.00013760617760617762, + "loss": 0.2507, + "step": 410 + }, + { + "epoch": 0.10344176681557919, + "grad_norm": 0.13919883966445923, + "learning_rate": 0.00013745173745173745, + "loss": 0.282, + "step": 411 + }, + { + "epoch": 0.10369344994651733, + "grad_norm": 0.11700534075498581, + "learning_rate": 0.0001372972972972973, + "loss": 0.1921, + "step": 412 + }, + { + "epoch": 0.10394513307745548, + "grad_norm": 0.12436141073703766, + "learning_rate": 0.00013714285714285716, + "loss": 0.2035, + "step": 413 + }, + { + "epoch": 0.10419681620839363, + "grad_norm": 0.14166633784770966, + "learning_rate": 0.000136988416988417, + "loss": 0.2772, + "step": 414 + }, + { + "epoch": 0.10444849933933179, + "grad_norm": 0.12872518599033356, + "learning_rate": 0.00013683397683397685, + "loss": 0.266, + "step": 415 + }, + { + "epoch": 0.10470018247026994, + "grad_norm": 0.1133652925491333, + "learning_rate": 0.00013667953667953668, + "loss": 0.1827, + "step": 416 + }, + { + "epoch": 0.10495186560120807, + "grad_norm": 0.12793788313865662, + "learning_rate": 0.00013652509652509653, + "loss": 0.2719, + "step": 417 + }, + { + "epoch": 0.10520354873214623, + "grad_norm": 0.11811885982751846, + "learning_rate": 0.0001363706563706564, + "loss": 0.1913, + "step": 418 + }, + { + "epoch": 0.10545523186308438, + "grad_norm": 0.12616224586963654, + "learning_rate": 0.00013621621621621622, + "loss": 0.2924, + "step": 419 + }, + { + "epoch": 0.10570691499402253, + "grad_norm": 0.13413816690444946, + "learning_rate": 0.00013606177606177605, + "loss": 0.2479, + "step": 420 + }, + { + "epoch": 0.10595859812496067, + "grad_norm": 0.12477762252092361, + "learning_rate": 0.0001359073359073359, + "loss": 0.2249, + "step": 421 + }, + { + "epoch": 0.10621028125589882, + "grad_norm": 0.11896514147520065, + "learning_rate": 0.00013575289575289576, + "loss": 0.1769, + "step": 422 + }, + { + "epoch": 0.10646196438683697, + "grad_norm": 0.11515690386295319, + "learning_rate": 0.00013559845559845562, + "loss": 0.1927, + "step": 423 + }, + { + "epoch": 0.10671364751777512, + "grad_norm": 0.11549720168113708, + "learning_rate": 0.00013544401544401545, + "loss": 0.2058, + "step": 424 + }, + { + "epoch": 0.10696533064871327, + "grad_norm": 0.12389887124300003, + "learning_rate": 0.00013528957528957528, + "loss": 0.1991, + "step": 425 + }, + { + "epoch": 0.10721701377965141, + "grad_norm": 0.11497371643781662, + "learning_rate": 0.00013513513513513514, + "loss": 0.1901, + "step": 426 + }, + { + "epoch": 0.10746869691058956, + "grad_norm": 0.11292821913957596, + "learning_rate": 0.000134980694980695, + "loss": 0.207, + "step": 427 + }, + { + "epoch": 0.10772038004152772, + "grad_norm": 0.11298840492963791, + "learning_rate": 0.00013482625482625485, + "loss": 0.1931, + "step": 428 + }, + { + "epoch": 0.10797206317246587, + "grad_norm": 0.16521763801574707, + "learning_rate": 0.00013467181467181468, + "loss": 0.2212, + "step": 429 + }, + { + "epoch": 0.10822374630340402, + "grad_norm": 0.11956854909658432, + "learning_rate": 0.0001345173745173745, + "loss": 0.1803, + "step": 430 + }, + { + "epoch": 0.10847542943434216, + "grad_norm": 0.13231535255908966, + "learning_rate": 0.00013436293436293437, + "loss": 0.1981, + "step": 431 + }, + { + "epoch": 0.10872711256528031, + "grad_norm": 0.12157753109931946, + "learning_rate": 0.00013420849420849422, + "loss": 0.2986, + "step": 432 + }, + { + "epoch": 0.10897879569621846, + "grad_norm": 0.13348908722400665, + "learning_rate": 0.00013405405405405408, + "loss": 0.2582, + "step": 433 + }, + { + "epoch": 0.10923047882715661, + "grad_norm": 0.13118185102939606, + "learning_rate": 0.00013389961389961388, + "loss": 0.2238, + "step": 434 + }, + { + "epoch": 0.10948216195809476, + "grad_norm": 0.15415893495082855, + "learning_rate": 0.00013374517374517374, + "loss": 0.2293, + "step": 435 + }, + { + "epoch": 0.1097338450890329, + "grad_norm": 0.1375678926706314, + "learning_rate": 0.0001335907335907336, + "loss": 0.3007, + "step": 436 + }, + { + "epoch": 0.10998552821997105, + "grad_norm": 0.11303649097681046, + "learning_rate": 0.00013343629343629345, + "loss": 0.2248, + "step": 437 + }, + { + "epoch": 0.1102372113509092, + "grad_norm": 0.12215966731309891, + "learning_rate": 0.0001332818532818533, + "loss": 0.2177, + "step": 438 + }, + { + "epoch": 0.11048889448184736, + "grad_norm": 0.14314697682857513, + "learning_rate": 0.0001331274131274131, + "loss": 0.2786, + "step": 439 + }, + { + "epoch": 0.11074057761278551, + "grad_norm": 0.115242138504982, + "learning_rate": 0.00013297297297297297, + "loss": 0.1939, + "step": 440 + }, + { + "epoch": 0.11099226074372365, + "grad_norm": 0.10042456537485123, + "learning_rate": 0.00013281853281853283, + "loss": 0.1601, + "step": 441 + }, + { + "epoch": 0.1112439438746618, + "grad_norm": 0.12134849280118942, + "learning_rate": 0.00013266409266409268, + "loss": 0.2214, + "step": 442 + }, + { + "epoch": 0.11149562700559995, + "grad_norm": 0.11193907260894775, + "learning_rate": 0.00013250965250965254, + "loss": 0.1901, + "step": 443 + }, + { + "epoch": 0.1117473101365381, + "grad_norm": 0.13861793279647827, + "learning_rate": 0.00013235521235521234, + "loss": 0.1967, + "step": 444 + }, + { + "epoch": 0.11199899326747624, + "grad_norm": 0.1239226907491684, + "learning_rate": 0.0001322007722007722, + "loss": 0.1925, + "step": 445 + }, + { + "epoch": 0.11225067639841439, + "grad_norm": 0.10150051862001419, + "learning_rate": 0.00013204633204633206, + "loss": 0.1664, + "step": 446 + }, + { + "epoch": 0.11250235952935254, + "grad_norm": 0.10890820622444153, + "learning_rate": 0.0001318918918918919, + "loss": 0.1669, + "step": 447 + }, + { + "epoch": 0.1127540426602907, + "grad_norm": 0.128633052110672, + "learning_rate": 0.00013173745173745174, + "loss": 0.2141, + "step": 448 + }, + { + "epoch": 0.11300572579122885, + "grad_norm": 0.12850677967071533, + "learning_rate": 0.00013158301158301157, + "loss": 0.22, + "step": 449 + }, + { + "epoch": 0.11325740892216699, + "grad_norm": 0.1250707060098648, + "learning_rate": 0.00013142857142857143, + "loss": 0.3382, + "step": 450 + }, + { + "epoch": 0.11350909205310514, + "grad_norm": 0.10860474407672882, + "learning_rate": 0.00013127413127413129, + "loss": 0.1487, + "step": 451 + }, + { + "epoch": 0.11376077518404329, + "grad_norm": 0.1171938106417656, + "learning_rate": 0.00013111969111969114, + "loss": 0.2145, + "step": 452 + }, + { + "epoch": 0.11401245831498144, + "grad_norm": 0.11092403531074524, + "learning_rate": 0.00013096525096525097, + "loss": 0.1749, + "step": 453 + }, + { + "epoch": 0.11426414144591959, + "grad_norm": 0.1231343001127243, + "learning_rate": 0.0001308108108108108, + "loss": 0.2133, + "step": 454 + }, + { + "epoch": 0.11451582457685773, + "grad_norm": 0.14924941956996918, + "learning_rate": 0.00013065637065637066, + "loss": 0.1876, + "step": 455 + }, + { + "epoch": 0.11476750770779588, + "grad_norm": 0.13826563954353333, + "learning_rate": 0.00013050193050193052, + "loss": 0.1785, + "step": 456 + }, + { + "epoch": 0.11501919083873403, + "grad_norm": 0.12644802033901215, + "learning_rate": 0.00013034749034749037, + "loss": 0.2797, + "step": 457 + }, + { + "epoch": 0.11527087396967219, + "grad_norm": 0.111385278403759, + "learning_rate": 0.0001301930501930502, + "loss": 0.1984, + "step": 458 + }, + { + "epoch": 0.11552255710061034, + "grad_norm": 0.12314785271883011, + "learning_rate": 0.00013003861003861003, + "loss": 0.2605, + "step": 459 + }, + { + "epoch": 0.11577424023154848, + "grad_norm": 0.14393912255764008, + "learning_rate": 0.0001298841698841699, + "loss": 0.1979, + "step": 460 + }, + { + "epoch": 0.11602592336248663, + "grad_norm": 0.11056330800056458, + "learning_rate": 0.00012972972972972974, + "loss": 0.147, + "step": 461 + }, + { + "epoch": 0.11627760649342478, + "grad_norm": 0.12650489807128906, + "learning_rate": 0.00012957528957528957, + "loss": 0.178, + "step": 462 + }, + { + "epoch": 0.11652928962436293, + "grad_norm": 0.09785740822553635, + "learning_rate": 0.00012942084942084943, + "loss": 0.1451, + "step": 463 + }, + { + "epoch": 0.11678097275530108, + "grad_norm": 0.1326170116662979, + "learning_rate": 0.00012926640926640926, + "loss": 0.1539, + "step": 464 + }, + { + "epoch": 0.11703265588623922, + "grad_norm": 0.12317286431789398, + "learning_rate": 0.00012911196911196912, + "loss": 0.2554, + "step": 465 + }, + { + "epoch": 0.11728433901717737, + "grad_norm": 0.12863805890083313, + "learning_rate": 0.00012895752895752897, + "loss": 0.2163, + "step": 466 + }, + { + "epoch": 0.11753602214811552, + "grad_norm": 0.12836167216300964, + "learning_rate": 0.0001288030888030888, + "loss": 0.2479, + "step": 467 + }, + { + "epoch": 0.11778770527905368, + "grad_norm": 0.1181485503911972, + "learning_rate": 0.00012864864864864866, + "loss": 0.1887, + "step": 468 + }, + { + "epoch": 0.11803938840999181, + "grad_norm": 0.12902015447616577, + "learning_rate": 0.0001284942084942085, + "loss": 0.2687, + "step": 469 + }, + { + "epoch": 0.11829107154092997, + "grad_norm": 0.11522433161735535, + "learning_rate": 0.00012833976833976835, + "loss": 0.1925, + "step": 470 + }, + { + "epoch": 0.11854275467186812, + "grad_norm": 0.1294202208518982, + "learning_rate": 0.0001281853281853282, + "loss": 0.2751, + "step": 471 + }, + { + "epoch": 0.11879443780280627, + "grad_norm": 0.10506186634302139, + "learning_rate": 0.00012803088803088803, + "loss": 0.1966, + "step": 472 + }, + { + "epoch": 0.11904612093374442, + "grad_norm": 0.1428360491991043, + "learning_rate": 0.0001278764478764479, + "loss": 0.3105, + "step": 473 + }, + { + "epoch": 0.11929780406468256, + "grad_norm": 0.11374489217996597, + "learning_rate": 0.00012772200772200772, + "loss": 0.1915, + "step": 474 + }, + { + "epoch": 0.11954948719562071, + "grad_norm": 0.12029575556516647, + "learning_rate": 0.00012756756756756758, + "loss": 0.1906, + "step": 475 + }, + { + "epoch": 0.11980117032655886, + "grad_norm": 0.13671739399433136, + "learning_rate": 0.0001274131274131274, + "loss": 0.2875, + "step": 476 + }, + { + "epoch": 0.12005285345749701, + "grad_norm": 0.12996806204319, + "learning_rate": 0.00012725868725868726, + "loss": 0.2602, + "step": 477 + }, + { + "epoch": 0.12030453658843517, + "grad_norm": 0.13968601822853088, + "learning_rate": 0.00012710424710424712, + "loss": 0.3111, + "step": 478 + }, + { + "epoch": 0.1205562197193733, + "grad_norm": 0.1273827999830246, + "learning_rate": 0.00012694980694980695, + "loss": 0.2575, + "step": 479 + }, + { + "epoch": 0.12080790285031145, + "grad_norm": 0.11352833360433578, + "learning_rate": 0.0001267953667953668, + "loss": 0.2238, + "step": 480 + }, + { + "epoch": 0.1210595859812496, + "grad_norm": 0.1088777706027031, + "learning_rate": 0.00012664092664092664, + "loss": 0.1864, + "step": 481 + }, + { + "epoch": 0.12131126911218776, + "grad_norm": 0.11757859587669373, + "learning_rate": 0.0001264864864864865, + "loss": 0.1789, + "step": 482 + }, + { + "epoch": 0.12156295224312591, + "grad_norm": 0.12215077877044678, + "learning_rate": 0.00012633204633204635, + "loss": 0.2352, + "step": 483 + }, + { + "epoch": 0.12181463537406405, + "grad_norm": 0.1086617186665535, + "learning_rate": 0.00012617760617760618, + "loss": 0.1807, + "step": 484 + }, + { + "epoch": 0.1220663185050022, + "grad_norm": 0.11662300676107407, + "learning_rate": 0.00012602316602316604, + "loss": 0.213, + "step": 485 + }, + { + "epoch": 0.12231800163594035, + "grad_norm": 0.1269250214099884, + "learning_rate": 0.00012586872586872587, + "loss": 0.2493, + "step": 486 + }, + { + "epoch": 0.1225696847668785, + "grad_norm": 0.11199555546045303, + "learning_rate": 0.00012571428571428572, + "loss": 0.1834, + "step": 487 + }, + { + "epoch": 0.12282136789781666, + "grad_norm": 0.11743030697107315, + "learning_rate": 0.00012555984555984558, + "loss": 0.2372, + "step": 488 + }, + { + "epoch": 0.12307305102875479, + "grad_norm": 0.11579969525337219, + "learning_rate": 0.0001254054054054054, + "loss": 0.1788, + "step": 489 + }, + { + "epoch": 0.12332473415969294, + "grad_norm": 0.15817004442214966, + "learning_rate": 0.00012525096525096524, + "loss": 0.2356, + "step": 490 + }, + { + "epoch": 0.1235764172906311, + "grad_norm": 0.10750343650579453, + "learning_rate": 0.0001250965250965251, + "loss": 0.1566, + "step": 491 + }, + { + "epoch": 0.12382810042156925, + "grad_norm": 0.11415287852287292, + "learning_rate": 0.00012494208494208495, + "loss": 0.1998, + "step": 492 + }, + { + "epoch": 0.1240797835525074, + "grad_norm": 0.13980022072792053, + "learning_rate": 0.0001247876447876448, + "loss": 0.2461, + "step": 493 + }, + { + "epoch": 0.12433146668344554, + "grad_norm": 0.12425754964351654, + "learning_rate": 0.00012463320463320464, + "loss": 0.2095, + "step": 494 + }, + { + "epoch": 0.12458314981438369, + "grad_norm": 0.12338211387395859, + "learning_rate": 0.00012447876447876447, + "loss": 0.2147, + "step": 495 + }, + { + "epoch": 0.12483483294532184, + "grad_norm": 0.13000816106796265, + "learning_rate": 0.00012432432432432433, + "loss": 0.3092, + "step": 496 + }, + { + "epoch": 0.12508651607625998, + "grad_norm": 0.12113019078969955, + "learning_rate": 0.00012416988416988418, + "loss": 0.1596, + "step": 497 + }, + { + "epoch": 0.12533819920719813, + "grad_norm": 0.12784144282341003, + "learning_rate": 0.00012401544401544404, + "loss": 0.277, + "step": 498 + }, + { + "epoch": 0.12558988233813628, + "grad_norm": 0.1454787701368332, + "learning_rate": 0.00012386100386100387, + "loss": 0.2897, + "step": 499 + }, + { + "epoch": 0.12584156546907443, + "grad_norm": 0.12137217074632645, + "learning_rate": 0.0001237065637065637, + "loss": 0.2139, + "step": 500 + }, + { + "epoch": 0.1260932486000126, + "grad_norm": 0.12419489026069641, + "learning_rate": 0.00012355212355212355, + "loss": 0.2132, + "step": 501 + }, + { + "epoch": 0.12634493173095074, + "grad_norm": 0.1173022985458374, + "learning_rate": 0.0001233976833976834, + "loss": 0.1981, + "step": 502 + }, + { + "epoch": 0.1265966148618889, + "grad_norm": 0.11166828125715256, + "learning_rate": 0.00012324324324324327, + "loss": 0.1911, + "step": 503 + }, + { + "epoch": 0.12684829799282704, + "grad_norm": 0.1224573478102684, + "learning_rate": 0.00012308880308880307, + "loss": 0.1899, + "step": 504 + }, + { + "epoch": 0.1270999811237652, + "grad_norm": 0.13685697317123413, + "learning_rate": 0.00012293436293436293, + "loss": 0.2231, + "step": 505 + }, + { + "epoch": 0.12735166425470332, + "grad_norm": 0.09964842349290848, + "learning_rate": 0.00012277992277992278, + "loss": 0.1634, + "step": 506 + }, + { + "epoch": 0.12760334738564147, + "grad_norm": 0.12324117124080658, + "learning_rate": 0.00012262548262548264, + "loss": 0.2187, + "step": 507 + }, + { + "epoch": 0.12785503051657962, + "grad_norm": 0.1302337795495987, + "learning_rate": 0.0001224710424710425, + "loss": 0.2077, + "step": 508 + }, + { + "epoch": 0.12810671364751777, + "grad_norm": 0.12188442796468735, + "learning_rate": 0.0001223166023166023, + "loss": 0.1791, + "step": 509 + }, + { + "epoch": 0.12835839677845592, + "grad_norm": 0.12415802478790283, + "learning_rate": 0.00012216216216216216, + "loss": 0.3054, + "step": 510 + }, + { + "epoch": 0.12861007990939408, + "grad_norm": 0.10483986884355545, + "learning_rate": 0.00012200772200772201, + "loss": 0.1783, + "step": 511 + }, + { + "epoch": 0.12886176304033223, + "grad_norm": 0.1163237914443016, + "learning_rate": 0.00012185328185328187, + "loss": 0.1775, + "step": 512 + }, + { + "epoch": 0.12911344617127038, + "grad_norm": 0.12157920747995377, + "learning_rate": 0.00012169884169884171, + "loss": 0.2146, + "step": 513 + }, + { + "epoch": 0.12936512930220853, + "grad_norm": 0.11034421622753143, + "learning_rate": 0.00012154440154440154, + "loss": 0.2072, + "step": 514 + }, + { + "epoch": 0.12961681243314666, + "grad_norm": 0.12620924413204193, + "learning_rate": 0.00012138996138996139, + "loss": 0.2414, + "step": 515 + }, + { + "epoch": 0.1298684955640848, + "grad_norm": 0.12227532267570496, + "learning_rate": 0.00012123552123552124, + "loss": 0.2444, + "step": 516 + }, + { + "epoch": 0.13012017869502296, + "grad_norm": 0.11898826062679291, + "learning_rate": 0.00012108108108108109, + "loss": 0.2041, + "step": 517 + }, + { + "epoch": 0.1303718618259611, + "grad_norm": 0.13300721347332, + "learning_rate": 0.00012092664092664092, + "loss": 0.2679, + "step": 518 + }, + { + "epoch": 0.13062354495689926, + "grad_norm": 0.11909513175487518, + "learning_rate": 0.00012077220077220077, + "loss": 0.2274, + "step": 519 + }, + { + "epoch": 0.13087522808783741, + "grad_norm": 0.10608676820993423, + "learning_rate": 0.00012061776061776062, + "loss": 0.2062, + "step": 520 + }, + { + "epoch": 0.13112691121877557, + "grad_norm": 0.11428362876176834, + "learning_rate": 0.00012046332046332047, + "loss": 0.195, + "step": 521 + }, + { + "epoch": 0.13137859434971372, + "grad_norm": 0.12112773209810257, + "learning_rate": 0.00012030888030888032, + "loss": 0.2005, + "step": 522 + }, + { + "epoch": 0.13163027748065187, + "grad_norm": 0.13230712711811066, + "learning_rate": 0.00012015444015444015, + "loss": 0.3307, + "step": 523 + }, + { + "epoch": 0.13188196061159002, + "grad_norm": 0.11997675895690918, + "learning_rate": 0.00012, + "loss": 0.1924, + "step": 524 + }, + { + "epoch": 0.13213364374252815, + "grad_norm": 0.12526918947696686, + "learning_rate": 0.00011984555984555985, + "loss": 0.1863, + "step": 525 + }, + { + "epoch": 0.1323853268734663, + "grad_norm": 0.11963635683059692, + "learning_rate": 0.0001196911196911197, + "loss": 0.1815, + "step": 526 + }, + { + "epoch": 0.13263701000440445, + "grad_norm": 0.10293988138437271, + "learning_rate": 0.00011953667953667955, + "loss": 0.1684, + "step": 527 + }, + { + "epoch": 0.1328886931353426, + "grad_norm": 0.13003425300121307, + "learning_rate": 0.00011938223938223938, + "loss": 0.2475, + "step": 528 + }, + { + "epoch": 0.13314037626628075, + "grad_norm": 0.1069263368844986, + "learning_rate": 0.00011922779922779923, + "loss": 0.1879, + "step": 529 + }, + { + "epoch": 0.1333920593972189, + "grad_norm": 0.105068638920784, + "learning_rate": 0.00011907335907335908, + "loss": 0.1868, + "step": 530 + }, + { + "epoch": 0.13364374252815706, + "grad_norm": 0.12748779356479645, + "learning_rate": 0.00011891891891891893, + "loss": 0.3147, + "step": 531 + }, + { + "epoch": 0.1338954256590952, + "grad_norm": 0.10573223233222961, + "learning_rate": 0.00011876447876447876, + "loss": 0.1626, + "step": 532 + }, + { + "epoch": 0.13414710879003336, + "grad_norm": 0.11097335815429688, + "learning_rate": 0.0001186100386100386, + "loss": 0.1651, + "step": 533 + }, + { + "epoch": 0.13439879192097148, + "grad_norm": 0.10889330506324768, + "learning_rate": 0.00011845559845559846, + "loss": 0.1708, + "step": 534 + }, + { + "epoch": 0.13465047505190963, + "grad_norm": 0.13869279623031616, + "learning_rate": 0.0001183011583011583, + "loss": 0.2957, + "step": 535 + }, + { + "epoch": 0.1349021581828478, + "grad_norm": 0.11425995826721191, + "learning_rate": 0.00011814671814671816, + "loss": 0.2006, + "step": 536 + }, + { + "epoch": 0.13515384131378594, + "grad_norm": 0.11701849102973938, + "learning_rate": 0.00011799227799227799, + "loss": 0.1917, + "step": 537 + }, + { + "epoch": 0.1354055244447241, + "grad_norm": 0.10709807276725769, + "learning_rate": 0.00011783783783783784, + "loss": 0.1842, + "step": 538 + }, + { + "epoch": 0.13565720757566224, + "grad_norm": 0.13820475339889526, + "learning_rate": 0.00011768339768339769, + "loss": 0.3169, + "step": 539 + }, + { + "epoch": 0.1359088907066004, + "grad_norm": 0.13134200870990753, + "learning_rate": 0.00011752895752895754, + "loss": 0.3117, + "step": 540 + }, + { + "epoch": 0.13616057383753855, + "grad_norm": 0.1296994835138321, + "learning_rate": 0.00011737451737451739, + "loss": 0.3081, + "step": 541 + }, + { + "epoch": 0.1364122569684767, + "grad_norm": 0.12059689313173294, + "learning_rate": 0.00011722007722007722, + "loss": 0.1882, + "step": 542 + }, + { + "epoch": 0.13666394009941485, + "grad_norm": 0.10846618562936783, + "learning_rate": 0.00011706563706563706, + "loss": 0.1827, + "step": 543 + }, + { + "epoch": 0.13691562323035297, + "grad_norm": 0.12945525348186493, + "learning_rate": 0.00011691119691119692, + "loss": 0.266, + "step": 544 + }, + { + "epoch": 0.13716730636129112, + "grad_norm": 0.12391635030508041, + "learning_rate": 0.00011675675675675676, + "loss": 0.1925, + "step": 545 + }, + { + "epoch": 0.13741898949222928, + "grad_norm": 0.11515142768621445, + "learning_rate": 0.0001166023166023166, + "loss": 0.2378, + "step": 546 + }, + { + "epoch": 0.13767067262316743, + "grad_norm": 0.11925746500492096, + "learning_rate": 0.00011644787644787645, + "loss": 0.2067, + "step": 547 + }, + { + "epoch": 0.13792235575410558, + "grad_norm": 0.12388575077056885, + "learning_rate": 0.0001162934362934363, + "loss": 0.2501, + "step": 548 + }, + { + "epoch": 0.13817403888504373, + "grad_norm": 0.12686868011951447, + "learning_rate": 0.00011613899613899615, + "loss": 0.1795, + "step": 549 + }, + { + "epoch": 0.13842572201598188, + "grad_norm": 0.13369601964950562, + "learning_rate": 0.000115984555984556, + "loss": 0.3065, + "step": 550 + }, + { + "epoch": 0.13867740514692004, + "grad_norm": 0.10592126846313477, + "learning_rate": 0.00011583011583011582, + "loss": 0.1887, + "step": 551 + }, + { + "epoch": 0.1389290882778582, + "grad_norm": 0.11250271648168564, + "learning_rate": 0.00011567567567567568, + "loss": 0.2147, + "step": 552 + }, + { + "epoch": 0.13918077140879634, + "grad_norm": 0.1211760938167572, + "learning_rate": 0.00011552123552123552, + "loss": 0.2072, + "step": 553 + }, + { + "epoch": 0.13943245453973446, + "grad_norm": 0.10624734312295914, + "learning_rate": 0.00011536679536679538, + "loss": 0.1914, + "step": 554 + }, + { + "epoch": 0.13968413767067261, + "grad_norm": 0.10755342245101929, + "learning_rate": 0.00011521235521235522, + "loss": 0.1607, + "step": 555 + }, + { + "epoch": 0.13993582080161077, + "grad_norm": 0.12237009406089783, + "learning_rate": 0.00011505791505791505, + "loss": 0.2001, + "step": 556 + }, + { + "epoch": 0.14018750393254892, + "grad_norm": 0.0914900004863739, + "learning_rate": 0.00011490347490347491, + "loss": 0.1392, + "step": 557 + }, + { + "epoch": 0.14043918706348707, + "grad_norm": 0.1298578530550003, + "learning_rate": 0.00011474903474903475, + "loss": 0.3164, + "step": 558 + }, + { + "epoch": 0.14069087019442522, + "grad_norm": 0.12323296070098877, + "learning_rate": 0.00011459459459459461, + "loss": 0.1792, + "step": 559 + }, + { + "epoch": 0.14094255332536337, + "grad_norm": 0.10546413064002991, + "learning_rate": 0.00011444015444015444, + "loss": 0.1961, + "step": 560 + }, + { + "epoch": 0.14119423645630153, + "grad_norm": 0.12417905777692795, + "learning_rate": 0.00011428571428571428, + "loss": 0.2055, + "step": 561 + }, + { + "epoch": 0.14144591958723968, + "grad_norm": 0.11603572219610214, + "learning_rate": 0.00011413127413127414, + "loss": 0.243, + "step": 562 + }, + { + "epoch": 0.1416976027181778, + "grad_norm": 0.11287704855203629, + "learning_rate": 0.00011397683397683398, + "loss": 0.1856, + "step": 563 + }, + { + "epoch": 0.14194928584911595, + "grad_norm": 0.10937439650297165, + "learning_rate": 0.00011382239382239384, + "loss": 0.1928, + "step": 564 + }, + { + "epoch": 0.1422009689800541, + "grad_norm": 0.11717220395803452, + "learning_rate": 0.00011366795366795367, + "loss": 0.1808, + "step": 565 + }, + { + "epoch": 0.14245265211099226, + "grad_norm": 0.11976751685142517, + "learning_rate": 0.00011351351351351351, + "loss": 0.1688, + "step": 566 + }, + { + "epoch": 0.1427043352419304, + "grad_norm": 0.11204173415899277, + "learning_rate": 0.00011335907335907337, + "loss": 0.1918, + "step": 567 + }, + { + "epoch": 0.14295601837286856, + "grad_norm": 0.10919293761253357, + "learning_rate": 0.00011320463320463321, + "loss": 0.2101, + "step": 568 + }, + { + "epoch": 0.1432077015038067, + "grad_norm": 0.0950295552611351, + "learning_rate": 0.00011305019305019307, + "loss": 0.1586, + "step": 569 + }, + { + "epoch": 0.14345938463474486, + "grad_norm": 0.10668401420116425, + "learning_rate": 0.0001128957528957529, + "loss": 0.1613, + "step": 570 + }, + { + "epoch": 0.14371106776568301, + "grad_norm": 0.13077902793884277, + "learning_rate": 0.00011274131274131274, + "loss": 0.2721, + "step": 571 + }, + { + "epoch": 0.14396275089662117, + "grad_norm": 0.12684080004692078, + "learning_rate": 0.0001125868725868726, + "loss": 0.2773, + "step": 572 + }, + { + "epoch": 0.1442144340275593, + "grad_norm": 0.12286555022001266, + "learning_rate": 0.00011243243243243244, + "loss": 0.2667, + "step": 573 + }, + { + "epoch": 0.14446611715849744, + "grad_norm": 0.10733581334352493, + "learning_rate": 0.0001122779922779923, + "loss": 0.1918, + "step": 574 + }, + { + "epoch": 0.1447178002894356, + "grad_norm": 0.10526907444000244, + "learning_rate": 0.00011212355212355213, + "loss": 0.1509, + "step": 575 + }, + { + "epoch": 0.14496948342037375, + "grad_norm": 0.12601731717586517, + "learning_rate": 0.00011196911196911197, + "loss": 0.3135, + "step": 576 + }, + { + "epoch": 0.1452211665513119, + "grad_norm": 0.12917165458202362, + "learning_rate": 0.00011181467181467183, + "loss": 0.3887, + "step": 577 + }, + { + "epoch": 0.14547284968225005, + "grad_norm": 0.12870042026042938, + "learning_rate": 0.00011166023166023167, + "loss": 0.2853, + "step": 578 + }, + { + "epoch": 0.1457245328131882, + "grad_norm": 0.11365857720375061, + "learning_rate": 0.0001115057915057915, + "loss": 0.1577, + "step": 579 + }, + { + "epoch": 0.14597621594412635, + "grad_norm": 0.10439585149288177, + "learning_rate": 0.00011135135135135135, + "loss": 0.1465, + "step": 580 + }, + { + "epoch": 0.1462278990750645, + "grad_norm": 0.12972363829612732, + "learning_rate": 0.0001111969111969112, + "loss": 0.2522, + "step": 581 + }, + { + "epoch": 0.14647958220600266, + "grad_norm": 0.12689651548862457, + "learning_rate": 0.00011104247104247106, + "loss": 0.1909, + "step": 582 + }, + { + "epoch": 0.14673126533694078, + "grad_norm": 0.12790319323539734, + "learning_rate": 0.0001108880308880309, + "loss": 0.1956, + "step": 583 + }, + { + "epoch": 0.14698294846787893, + "grad_norm": 0.1108984649181366, + "learning_rate": 0.00011073359073359073, + "loss": 0.1724, + "step": 584 + }, + { + "epoch": 0.14723463159881708, + "grad_norm": 0.1242600828409195, + "learning_rate": 0.00011057915057915057, + "loss": 0.1903, + "step": 585 + }, + { + "epoch": 0.14748631472975524, + "grad_norm": 0.1401650607585907, + "learning_rate": 0.00011042471042471043, + "loss": 0.2368, + "step": 586 + }, + { + "epoch": 0.1477379978606934, + "grad_norm": 0.1247275248169899, + "learning_rate": 0.00011027027027027029, + "loss": 0.269, + "step": 587 + }, + { + "epoch": 0.14798968099163154, + "grad_norm": 0.1258562058210373, + "learning_rate": 0.00011011583011583013, + "loss": 0.2514, + "step": 588 + }, + { + "epoch": 0.1482413641225697, + "grad_norm": 0.12861283123493195, + "learning_rate": 0.00010996138996138996, + "loss": 0.1674, + "step": 589 + }, + { + "epoch": 0.14849304725350784, + "grad_norm": 0.12524257600307465, + "learning_rate": 0.0001098069498069498, + "loss": 0.1987, + "step": 590 + }, + { + "epoch": 0.148744730384446, + "grad_norm": 0.11897309869527817, + "learning_rate": 0.00010965250965250966, + "loss": 0.1668, + "step": 591 + }, + { + "epoch": 0.14899641351538412, + "grad_norm": 0.1354551613330841, + "learning_rate": 0.0001094980694980695, + "loss": 0.1998, + "step": 592 + }, + { + "epoch": 0.14924809664632227, + "grad_norm": 0.10902054607868195, + "learning_rate": 0.00010934362934362933, + "loss": 0.1906, + "step": 593 + }, + { + "epoch": 0.14949977977726042, + "grad_norm": 0.11879126727581024, + "learning_rate": 0.00010918918918918919, + "loss": 0.1882, + "step": 594 + }, + { + "epoch": 0.14975146290819857, + "grad_norm": 0.128411203622818, + "learning_rate": 0.00010903474903474903, + "loss": 0.1486, + "step": 595 + }, + { + "epoch": 0.15000314603913673, + "grad_norm": 0.12789411842823029, + "learning_rate": 0.00010888030888030889, + "loss": 0.2916, + "step": 596 + }, + { + "epoch": 0.15025482917007488, + "grad_norm": 0.1334899216890335, + "learning_rate": 0.00010872586872586873, + "loss": 0.2437, + "step": 597 + }, + { + "epoch": 0.15050651230101303, + "grad_norm": 0.11723010987043381, + "learning_rate": 0.00010857142857142856, + "loss": 0.1788, + "step": 598 + }, + { + "epoch": 0.15075819543195118, + "grad_norm": 0.12522217631340027, + "learning_rate": 0.00010841698841698842, + "loss": 0.1779, + "step": 599 + }, + { + "epoch": 0.15100987856288933, + "grad_norm": 0.12590517103672028, + "learning_rate": 0.00010826254826254826, + "loss": 0.2202, + "step": 600 + }, + { + "epoch": 0.15126156169382748, + "grad_norm": 0.12168847024440765, + "learning_rate": 0.00010810810810810812, + "loss": 0.2106, + "step": 601 + }, + { + "epoch": 0.1515132448247656, + "grad_norm": 0.13695412874221802, + "learning_rate": 0.00010795366795366796, + "loss": 0.3031, + "step": 602 + }, + { + "epoch": 0.15176492795570376, + "grad_norm": 0.13396726548671722, + "learning_rate": 0.0001077992277992278, + "loss": 0.2745, + "step": 603 + }, + { + "epoch": 0.1520166110866419, + "grad_norm": 0.13887108862400055, + "learning_rate": 0.00010764478764478765, + "loss": 0.2121, + "step": 604 + }, + { + "epoch": 0.15226829421758006, + "grad_norm": 0.16270503401756287, + "learning_rate": 0.0001074903474903475, + "loss": 0.3188, + "step": 605 + }, + { + "epoch": 0.15251997734851822, + "grad_norm": 0.11081714928150177, + "learning_rate": 0.00010733590733590735, + "loss": 0.1663, + "step": 606 + }, + { + "epoch": 0.15277166047945637, + "grad_norm": 0.10872180014848709, + "learning_rate": 0.00010718146718146718, + "loss": 0.1541, + "step": 607 + }, + { + "epoch": 0.15302334361039452, + "grad_norm": 0.12311674654483795, + "learning_rate": 0.00010702702702702702, + "loss": 0.2491, + "step": 608 + }, + { + "epoch": 0.15327502674133267, + "grad_norm": 0.11725156754255295, + "learning_rate": 0.00010687258687258688, + "loss": 0.249, + "step": 609 + }, + { + "epoch": 0.15352670987227082, + "grad_norm": 0.13047930598258972, + "learning_rate": 0.00010671814671814672, + "loss": 0.2221, + "step": 610 + }, + { + "epoch": 0.15377839300320895, + "grad_norm": 0.14251331984996796, + "learning_rate": 0.00010656370656370658, + "loss": 0.2157, + "step": 611 + }, + { + "epoch": 0.1540300761341471, + "grad_norm": 0.14140063524246216, + "learning_rate": 0.00010640926640926641, + "loss": 0.2544, + "step": 612 + }, + { + "epoch": 0.15428175926508525, + "grad_norm": 0.12871818244457245, + "learning_rate": 0.00010625482625482625, + "loss": 0.2782, + "step": 613 + }, + { + "epoch": 0.1545334423960234, + "grad_norm": 0.11662749201059341, + "learning_rate": 0.00010610038610038611, + "loss": 0.2156, + "step": 614 + }, + { + "epoch": 0.15478512552696155, + "grad_norm": 0.12321148067712784, + "learning_rate": 0.00010594594594594595, + "loss": 0.2924, + "step": 615 + }, + { + "epoch": 0.1550368086578997, + "grad_norm": 0.15163083374500275, + "learning_rate": 0.00010579150579150581, + "loss": 0.2777, + "step": 616 + }, + { + "epoch": 0.15528849178883786, + "grad_norm": 0.12142910808324814, + "learning_rate": 0.00010563706563706564, + "loss": 0.3034, + "step": 617 + }, + { + "epoch": 0.155540174919776, + "grad_norm": 0.1033206582069397, + "learning_rate": 0.00010548262548262548, + "loss": 0.1883, + "step": 618 + }, + { + "epoch": 0.15579185805071416, + "grad_norm": 0.11347199231386185, + "learning_rate": 0.00010532818532818534, + "loss": 0.2067, + "step": 619 + }, + { + "epoch": 0.1560435411816523, + "grad_norm": 0.11548115313053131, + "learning_rate": 0.00010517374517374518, + "loss": 0.1789, + "step": 620 + }, + { + "epoch": 0.15629522431259044, + "grad_norm": 0.12020577490329742, + "learning_rate": 0.00010501930501930501, + "loss": 0.2287, + "step": 621 + }, + { + "epoch": 0.1565469074435286, + "grad_norm": 0.10243792831897736, + "learning_rate": 0.00010486486486486487, + "loss": 0.1624, + "step": 622 + }, + { + "epoch": 0.15679859057446674, + "grad_norm": 0.1149209588766098, + "learning_rate": 0.00010471042471042471, + "loss": 0.2107, + "step": 623 + }, + { + "epoch": 0.1570502737054049, + "grad_norm": 0.11209885030984879, + "learning_rate": 0.00010455598455598457, + "loss": 0.1743, + "step": 624 + }, + { + "epoch": 0.15730195683634304, + "grad_norm": 0.11851311475038528, + "learning_rate": 0.00010440154440154441, + "loss": 0.1947, + "step": 625 + }, + { + "epoch": 0.1575536399672812, + "grad_norm": 0.1332913339138031, + "learning_rate": 0.00010424710424710424, + "loss": 0.2571, + "step": 626 + }, + { + "epoch": 0.15780532309821935, + "grad_norm": 0.12531130015850067, + "learning_rate": 0.0001040926640926641, + "loss": 0.1936, + "step": 627 + }, + { + "epoch": 0.1580570062291575, + "grad_norm": 0.10874873399734497, + "learning_rate": 0.00010393822393822394, + "loss": 0.1816, + "step": 628 + }, + { + "epoch": 0.15830868936009565, + "grad_norm": 0.11458396911621094, + "learning_rate": 0.0001037837837837838, + "loss": 0.1932, + "step": 629 + }, + { + "epoch": 0.1585603724910338, + "grad_norm": 0.11653389036655426, + "learning_rate": 0.00010362934362934364, + "loss": 0.1692, + "step": 630 + }, + { + "epoch": 0.15881205562197193, + "grad_norm": 0.11083444207906723, + "learning_rate": 0.00010347490347490347, + "loss": 0.1794, + "step": 631 + }, + { + "epoch": 0.15906373875291008, + "grad_norm": 0.11074251681566238, + "learning_rate": 0.00010332046332046333, + "loss": 0.2096, + "step": 632 + }, + { + "epoch": 0.15931542188384823, + "grad_norm": 0.1104135736823082, + "learning_rate": 0.00010316602316602317, + "loss": 0.1795, + "step": 633 + }, + { + "epoch": 0.15956710501478638, + "grad_norm": 0.13042470812797546, + "learning_rate": 0.00010301158301158303, + "loss": 0.2023, + "step": 634 + }, + { + "epoch": 0.15981878814572453, + "grad_norm": 0.13127300143241882, + "learning_rate": 0.00010285714285714286, + "loss": 0.2671, + "step": 635 + }, + { + "epoch": 0.16007047127666268, + "grad_norm": 0.15113769471645355, + "learning_rate": 0.0001027027027027027, + "loss": 0.3285, + "step": 636 + }, + { + "epoch": 0.16032215440760084, + "grad_norm": 0.11295834183692932, + "learning_rate": 0.00010254826254826256, + "loss": 0.1903, + "step": 637 + }, + { + "epoch": 0.160573837538539, + "grad_norm": 0.11116720736026764, + "learning_rate": 0.0001023938223938224, + "loss": 0.2105, + "step": 638 + }, + { + "epoch": 0.16082552066947714, + "grad_norm": 0.12054213881492615, + "learning_rate": 0.00010223938223938226, + "loss": 0.3026, + "step": 639 + }, + { + "epoch": 0.16107720380041526, + "grad_norm": 0.11756197363138199, + "learning_rate": 0.00010208494208494209, + "loss": 0.2304, + "step": 640 + }, + { + "epoch": 0.16132888693135342, + "grad_norm": 0.13645577430725098, + "learning_rate": 0.00010193050193050193, + "loss": 0.1769, + "step": 641 + }, + { + "epoch": 0.16158057006229157, + "grad_norm": 0.10601159930229187, + "learning_rate": 0.00010177606177606179, + "loss": 0.1891, + "step": 642 + }, + { + "epoch": 0.16183225319322972, + "grad_norm": 0.12233410030603409, + "learning_rate": 0.00010162162162162163, + "loss": 0.2665, + "step": 643 + }, + { + "epoch": 0.16208393632416787, + "grad_norm": 0.13072800636291504, + "learning_rate": 0.00010146718146718149, + "loss": 0.2374, + "step": 644 + }, + { + "epoch": 0.16233561945510602, + "grad_norm": 0.1123851016163826, + "learning_rate": 0.00010131274131274132, + "loss": 0.2053, + "step": 645 + }, + { + "epoch": 0.16258730258604417, + "grad_norm": 0.12138116359710693, + "learning_rate": 0.00010115830115830116, + "loss": 0.2058, + "step": 646 + }, + { + "epoch": 0.16283898571698233, + "grad_norm": 0.1370951384305954, + "learning_rate": 0.00010100386100386102, + "loss": 0.2874, + "step": 647 + }, + { + "epoch": 0.16309066884792048, + "grad_norm": 0.1120346263051033, + "learning_rate": 0.00010084942084942086, + "loss": 0.1979, + "step": 648 + }, + { + "epoch": 0.16334235197885863, + "grad_norm": 0.1129554808139801, + "learning_rate": 0.00010069498069498069, + "loss": 0.1621, + "step": 649 + }, + { + "epoch": 0.16359403510979675, + "grad_norm": 0.12297677248716354, + "learning_rate": 0.00010054054054054053, + "loss": 0.2417, + "step": 650 + }, + { + "epoch": 0.1638457182407349, + "grad_norm": 0.1278134435415268, + "learning_rate": 0.00010038610038610039, + "loss": 0.3603, + "step": 651 + }, + { + "epoch": 0.16409740137167306, + "grad_norm": 0.11952311545610428, + "learning_rate": 0.00010023166023166025, + "loss": 0.1953, + "step": 652 + }, + { + "epoch": 0.1643490845026112, + "grad_norm": 0.11924976855516434, + "learning_rate": 0.00010007722007722009, + "loss": 0.2065, + "step": 653 + }, + { + "epoch": 0.16460076763354936, + "grad_norm": 0.10717752575874329, + "learning_rate": 9.992277992277993e-05, + "loss": 0.181, + "step": 654 + }, + { + "epoch": 0.1648524507644875, + "grad_norm": 0.11415696144104004, + "learning_rate": 9.976833976833976e-05, + "loss": 0.1981, + "step": 655 + }, + { + "epoch": 0.16510413389542566, + "grad_norm": 0.1112828329205513, + "learning_rate": 9.961389961389962e-05, + "loss": 0.1913, + "step": 656 + }, + { + "epoch": 0.16535581702636382, + "grad_norm": 0.10803753137588501, + "learning_rate": 9.945945945945948e-05, + "loss": 0.1872, + "step": 657 + }, + { + "epoch": 0.16560750015730197, + "grad_norm": 0.11290117353200912, + "learning_rate": 9.93050193050193e-05, + "loss": 0.1769, + "step": 658 + }, + { + "epoch": 0.16585918328824012, + "grad_norm": 0.1432167887687683, + "learning_rate": 9.915057915057916e-05, + "loss": 0.1765, + "step": 659 + }, + { + "epoch": 0.16611086641917824, + "grad_norm": 0.12793876230716705, + "learning_rate": 9.899613899613899e-05, + "loss": 0.285, + "step": 660 + }, + { + "epoch": 0.1663625495501164, + "grad_norm": 0.11001358926296234, + "learning_rate": 9.884169884169885e-05, + "loss": 0.1607, + "step": 661 + }, + { + "epoch": 0.16661423268105455, + "grad_norm": 0.11504772305488586, + "learning_rate": 9.868725868725869e-05, + "loss": 0.2131, + "step": 662 + }, + { + "epoch": 0.1668659158119927, + "grad_norm": 0.1218910664319992, + "learning_rate": 9.853281853281854e-05, + "loss": 0.1963, + "step": 663 + }, + { + "epoch": 0.16711759894293085, + "grad_norm": 0.12067963927984238, + "learning_rate": 9.837837837837839e-05, + "loss": 0.1987, + "step": 664 + }, + { + "epoch": 0.167369282073869, + "grad_norm": 0.1263176053762436, + "learning_rate": 9.822393822393822e-05, + "loss": 0.2294, + "step": 665 + }, + { + "epoch": 0.16762096520480715, + "grad_norm": 0.1330689638853073, + "learning_rate": 9.806949806949808e-05, + "loss": 0.3078, + "step": 666 + }, + { + "epoch": 0.1678726483357453, + "grad_norm": 0.12405157089233398, + "learning_rate": 9.791505791505792e-05, + "loss": 0.2171, + "step": 667 + }, + { + "epoch": 0.16812433146668346, + "grad_norm": 0.13401226699352264, + "learning_rate": 9.776061776061777e-05, + "loss": 0.1824, + "step": 668 + }, + { + "epoch": 0.16837601459762158, + "grad_norm": 0.14228777587413788, + "learning_rate": 9.760617760617761e-05, + "loss": 0.3368, + "step": 669 + }, + { + "epoch": 0.16862769772855973, + "grad_norm": 0.14598000049591064, + "learning_rate": 9.745173745173745e-05, + "loss": 0.2041, + "step": 670 + }, + { + "epoch": 0.16887938085949789, + "grad_norm": 0.11994544416666031, + "learning_rate": 9.729729729729731e-05, + "loss": 0.1955, + "step": 671 + }, + { + "epoch": 0.16913106399043604, + "grad_norm": 0.12644168734550476, + "learning_rate": 9.714285714285715e-05, + "loss": 0.2223, + "step": 672 + }, + { + "epoch": 0.1693827471213742, + "grad_norm": 0.13086773455142975, + "learning_rate": 9.6988416988417e-05, + "loss": 0.2678, + "step": 673 + }, + { + "epoch": 0.16963443025231234, + "grad_norm": 0.1267138570547104, + "learning_rate": 9.683397683397684e-05, + "loss": 0.3337, + "step": 674 + }, + { + "epoch": 0.1698861133832505, + "grad_norm": 0.11454298347234726, + "learning_rate": 9.667953667953668e-05, + "loss": 0.1799, + "step": 675 + }, + { + "epoch": 0.17013779651418864, + "grad_norm": 0.09887906163930893, + "learning_rate": 9.652509652509652e-05, + "loss": 0.1571, + "step": 676 + }, + { + "epoch": 0.1703894796451268, + "grad_norm": 0.12321919947862625, + "learning_rate": 9.637065637065638e-05, + "loss": 0.2585, + "step": 677 + }, + { + "epoch": 0.17064116277606495, + "grad_norm": 0.11580654978752136, + "learning_rate": 9.621621621621622e-05, + "loss": 0.2242, + "step": 678 + }, + { + "epoch": 0.17089284590700307, + "grad_norm": 0.10302761197090149, + "learning_rate": 9.606177606177607e-05, + "loss": 0.19, + "step": 679 + }, + { + "epoch": 0.17114452903794122, + "grad_norm": 0.11387483775615692, + "learning_rate": 9.590733590733591e-05, + "loss": 0.2058, + "step": 680 + }, + { + "epoch": 0.17139621216887937, + "grad_norm": 0.12242447584867477, + "learning_rate": 9.575289575289575e-05, + "loss": 0.297, + "step": 681 + }, + { + "epoch": 0.17164789529981753, + "grad_norm": 0.11938390880823135, + "learning_rate": 9.559845559845561e-05, + "loss": 0.1737, + "step": 682 + }, + { + "epoch": 0.17189957843075568, + "grad_norm": 0.10991702973842621, + "learning_rate": 9.544401544401544e-05, + "loss": 0.2098, + "step": 683 + }, + { + "epoch": 0.17215126156169383, + "grad_norm": 0.1246740072965622, + "learning_rate": 9.52895752895753e-05, + "loss": 0.1852, + "step": 684 + }, + { + "epoch": 0.17240294469263198, + "grad_norm": 0.13339395821094513, + "learning_rate": 9.513513513513514e-05, + "loss": 0.2528, + "step": 685 + }, + { + "epoch": 0.17265462782357013, + "grad_norm": 0.10641826689243317, + "learning_rate": 9.498069498069498e-05, + "loss": 0.19, + "step": 686 + }, + { + "epoch": 0.17290631095450829, + "grad_norm": 0.12051086127758026, + "learning_rate": 9.482625482625484e-05, + "loss": 0.2326, + "step": 687 + }, + { + "epoch": 0.1731579940854464, + "grad_norm": 0.11246494948863983, + "learning_rate": 9.467181467181467e-05, + "loss": 0.1961, + "step": 688 + }, + { + "epoch": 0.17340967721638456, + "grad_norm": 0.1181284710764885, + "learning_rate": 9.451737451737453e-05, + "loss": 0.1856, + "step": 689 + }, + { + "epoch": 0.1736613603473227, + "grad_norm": 0.10863402485847473, + "learning_rate": 9.436293436293436e-05, + "loss": 0.1759, + "step": 690 + }, + { + "epoch": 0.17391304347826086, + "grad_norm": 0.12081693857908249, + "learning_rate": 9.420849420849421e-05, + "loss": 0.2196, + "step": 691 + }, + { + "epoch": 0.17416472660919902, + "grad_norm": 0.09784593433141708, + "learning_rate": 9.405405405405407e-05, + "loss": 0.1539, + "step": 692 + }, + { + "epoch": 0.17441640974013717, + "grad_norm": 0.1435626894235611, + "learning_rate": 9.38996138996139e-05, + "loss": 0.3111, + "step": 693 + }, + { + "epoch": 0.17466809287107532, + "grad_norm": 0.16294117271900177, + "learning_rate": 9.374517374517376e-05, + "loss": 0.2604, + "step": 694 + }, + { + "epoch": 0.17491977600201347, + "grad_norm": 0.10859392583370209, + "learning_rate": 9.359073359073359e-05, + "loss": 0.2045, + "step": 695 + }, + { + "epoch": 0.17517145913295162, + "grad_norm": 0.11717144399881363, + "learning_rate": 9.343629343629344e-05, + "loss": 0.2068, + "step": 696 + }, + { + "epoch": 0.17542314226388978, + "grad_norm": 0.11331836134195328, + "learning_rate": 9.328185328185329e-05, + "loss": 0.2002, + "step": 697 + }, + { + "epoch": 0.1756748253948279, + "grad_norm": 0.11233112215995789, + "learning_rate": 9.312741312741313e-05, + "loss": 0.1808, + "step": 698 + }, + { + "epoch": 0.17592650852576605, + "grad_norm": 0.11876025795936584, + "learning_rate": 9.297297297297299e-05, + "loss": 0.2435, + "step": 699 + }, + { + "epoch": 0.1761781916567042, + "grad_norm": 0.13402500748634338, + "learning_rate": 9.281853281853282e-05, + "loss": 0.2077, + "step": 700 + }, + { + "epoch": 0.17642987478764235, + "grad_norm": 0.12083412706851959, + "learning_rate": 9.266409266409267e-05, + "loss": 0.173, + "step": 701 + }, + { + "epoch": 0.1766815579185805, + "grad_norm": 0.14151129126548767, + "learning_rate": 9.250965250965252e-05, + "loss": 0.2, + "step": 702 + }, + { + "epoch": 0.17693324104951866, + "grad_norm": 0.11531958729028702, + "learning_rate": 9.235521235521236e-05, + "loss": 0.2182, + "step": 703 + }, + { + "epoch": 0.1771849241804568, + "grad_norm": 0.13341595232486725, + "learning_rate": 9.22007722007722e-05, + "loss": 0.2048, + "step": 704 + }, + { + "epoch": 0.17743660731139496, + "grad_norm": 0.11693217605352402, + "learning_rate": 9.204633204633205e-05, + "loss": 0.243, + "step": 705 + }, + { + "epoch": 0.1776882904423331, + "grad_norm": 0.11376842111349106, + "learning_rate": 9.18918918918919e-05, + "loss": 0.205, + "step": 706 + }, + { + "epoch": 0.17793997357327127, + "grad_norm": 0.10769005119800568, + "learning_rate": 9.173745173745175e-05, + "loss": 0.1648, + "step": 707 + }, + { + "epoch": 0.1781916567042094, + "grad_norm": 0.10357316583395004, + "learning_rate": 9.158301158301159e-05, + "loss": 0.1586, + "step": 708 + }, + { + "epoch": 0.17844333983514754, + "grad_norm": 0.1224375069141388, + "learning_rate": 9.142857142857143e-05, + "loss": 0.235, + "step": 709 + }, + { + "epoch": 0.1786950229660857, + "grad_norm": 0.1266258805990219, + "learning_rate": 9.127413127413128e-05, + "loss": 0.2824, + "step": 710 + }, + { + "epoch": 0.17894670609702384, + "grad_norm": 0.13108877837657928, + "learning_rate": 9.111969111969112e-05, + "loss": 0.2135, + "step": 711 + }, + { + "epoch": 0.179198389227962, + "grad_norm": 0.12189647555351257, + "learning_rate": 9.096525096525098e-05, + "loss": 0.1985, + "step": 712 + }, + { + "epoch": 0.17945007235890015, + "grad_norm": 0.1352180391550064, + "learning_rate": 9.081081081081082e-05, + "loss": 0.2607, + "step": 713 + }, + { + "epoch": 0.1797017554898383, + "grad_norm": 0.11011572927236557, + "learning_rate": 9.065637065637066e-05, + "loss": 0.1816, + "step": 714 + }, + { + "epoch": 0.17995343862077645, + "grad_norm": 0.13449819386005402, + "learning_rate": 9.05019305019305e-05, + "loss": 0.2387, + "step": 715 + }, + { + "epoch": 0.1802051217517146, + "grad_norm": 0.11416808515787125, + "learning_rate": 9.034749034749035e-05, + "loss": 0.208, + "step": 716 + }, + { + "epoch": 0.18045680488265273, + "grad_norm": 0.11835426092147827, + "learning_rate": 9.01930501930502e-05, + "loss": 0.1997, + "step": 717 + }, + { + "epoch": 0.18070848801359088, + "grad_norm": 0.11853640526533127, + "learning_rate": 9.003861003861003e-05, + "loss": 0.2328, + "step": 718 + }, + { + "epoch": 0.18096017114452903, + "grad_norm": 0.13454513251781464, + "learning_rate": 8.988416988416989e-05, + "loss": 0.3601, + "step": 719 + }, + { + "epoch": 0.18121185427546718, + "grad_norm": 0.12746816873550415, + "learning_rate": 8.972972972972973e-05, + "loss": 0.301, + "step": 720 + }, + { + "epoch": 0.18146353740640533, + "grad_norm": 0.11179789900779724, + "learning_rate": 8.957528957528958e-05, + "loss": 0.213, + "step": 721 + }, + { + "epoch": 0.18171522053734349, + "grad_norm": 0.10550542175769806, + "learning_rate": 8.942084942084943e-05, + "loss": 0.1773, + "step": 722 + }, + { + "epoch": 0.18196690366828164, + "grad_norm": 0.12112996727228165, + "learning_rate": 8.926640926640926e-05, + "loss": 0.2032, + "step": 723 + }, + { + "epoch": 0.1822185867992198, + "grad_norm": 0.11407957971096039, + "learning_rate": 8.911196911196912e-05, + "loss": 0.1757, + "step": 724 + }, + { + "epoch": 0.18247026993015794, + "grad_norm": 0.11602102965116501, + "learning_rate": 8.895752895752895e-05, + "loss": 0.2475, + "step": 725 + }, + { + "epoch": 0.1827219530610961, + "grad_norm": 0.11990536004304886, + "learning_rate": 8.880308880308881e-05, + "loss": 0.2045, + "step": 726 + }, + { + "epoch": 0.18297363619203422, + "grad_norm": 0.10824867337942123, + "learning_rate": 8.864864864864866e-05, + "loss": 0.1778, + "step": 727 + }, + { + "epoch": 0.18322531932297237, + "grad_norm": 0.12675610184669495, + "learning_rate": 8.84942084942085e-05, + "loss": 0.2995, + "step": 728 + }, + { + "epoch": 0.18347700245391052, + "grad_norm": 0.1296304315328598, + "learning_rate": 8.833976833976835e-05, + "loss": 0.2115, + "step": 729 + }, + { + "epoch": 0.18372868558484867, + "grad_norm": 0.11376269161701202, + "learning_rate": 8.818532818532818e-05, + "loss": 0.2732, + "step": 730 + }, + { + "epoch": 0.18398036871578682, + "grad_norm": 0.09493126720190048, + "learning_rate": 8.803088803088804e-05, + "loss": 0.1703, + "step": 731 + }, + { + "epoch": 0.18423205184672498, + "grad_norm": 0.12636464834213257, + "learning_rate": 8.787644787644788e-05, + "loss": 0.3137, + "step": 732 + }, + { + "epoch": 0.18448373497766313, + "grad_norm": 0.12038034945726395, + "learning_rate": 8.772200772200772e-05, + "loss": 0.2051, + "step": 733 + }, + { + "epoch": 0.18473541810860128, + "grad_norm": 0.09699856489896774, + "learning_rate": 8.756756756756758e-05, + "loss": 0.1329, + "step": 734 + }, + { + "epoch": 0.18498710123953943, + "grad_norm": 0.12379298359155655, + "learning_rate": 8.741312741312741e-05, + "loss": 0.259, + "step": 735 + }, + { + "epoch": 0.18523878437047755, + "grad_norm": 0.09711078554391861, + "learning_rate": 8.725868725868727e-05, + "loss": 0.1326, + "step": 736 + }, + { + "epoch": 0.1854904675014157, + "grad_norm": 0.10566964000463486, + "learning_rate": 8.710424710424711e-05, + "loss": 0.1995, + "step": 737 + }, + { + "epoch": 0.18574215063235386, + "grad_norm": 0.10495832562446594, + "learning_rate": 8.694980694980695e-05, + "loss": 0.1551, + "step": 738 + }, + { + "epoch": 0.185993833763292, + "grad_norm": 0.12425988167524338, + "learning_rate": 8.67953667953668e-05, + "loss": 0.2847, + "step": 739 + }, + { + "epoch": 0.18624551689423016, + "grad_norm": 0.1310386210680008, + "learning_rate": 8.664092664092664e-05, + "loss": 0.1806, + "step": 740 + }, + { + "epoch": 0.1864972000251683, + "grad_norm": 0.12181442975997925, + "learning_rate": 8.64864864864865e-05, + "loss": 0.1899, + "step": 741 + }, + { + "epoch": 0.18674888315610647, + "grad_norm": 0.12717288732528687, + "learning_rate": 8.633204633204634e-05, + "loss": 0.351, + "step": 742 + }, + { + "epoch": 0.18700056628704462, + "grad_norm": 0.12999722361564636, + "learning_rate": 8.617760617760618e-05, + "loss": 0.2052, + "step": 743 + }, + { + "epoch": 0.18725224941798277, + "grad_norm": 0.1204526275396347, + "learning_rate": 8.602316602316603e-05, + "loss": 0.2155, + "step": 744 + }, + { + "epoch": 0.18750393254892092, + "grad_norm": 0.1277237832546234, + "learning_rate": 8.586872586872587e-05, + "loss": 0.2356, + "step": 745 + }, + { + "epoch": 0.18775561567985904, + "grad_norm": 0.11569196730852127, + "learning_rate": 8.571428571428571e-05, + "loss": 0.1954, + "step": 746 + }, + { + "epoch": 0.1880072988107972, + "grad_norm": 0.1201886534690857, + "learning_rate": 8.555984555984557e-05, + "loss": 0.2402, + "step": 747 + }, + { + "epoch": 0.18825898194173535, + "grad_norm": 0.1264549046754837, + "learning_rate": 8.540540540540541e-05, + "loss": 0.2015, + "step": 748 + }, + { + "epoch": 0.1885106650726735, + "grad_norm": 0.11811226606369019, + "learning_rate": 8.525096525096526e-05, + "loss": 0.1854, + "step": 749 + }, + { + "epoch": 0.18876234820361165, + "grad_norm": 0.1121421530842781, + "learning_rate": 8.50965250965251e-05, + "loss": 0.182, + "step": 750 + }, + { + "epoch": 0.1890140313345498, + "grad_norm": 0.10724892467260361, + "learning_rate": 8.494208494208494e-05, + "loss": 0.1663, + "step": 751 + }, + { + "epoch": 0.18926571446548796, + "grad_norm": 0.12521569430828094, + "learning_rate": 8.47876447876448e-05, + "loss": 0.1871, + "step": 752 + }, + { + "epoch": 0.1895173975964261, + "grad_norm": 0.13731469213962555, + "learning_rate": 8.463320463320463e-05, + "loss": 0.2015, + "step": 753 + }, + { + "epoch": 0.18976908072736426, + "grad_norm": 0.11995433270931244, + "learning_rate": 8.447876447876449e-05, + "loss": 0.2264, + "step": 754 + }, + { + "epoch": 0.1900207638583024, + "grad_norm": 0.12111391872167587, + "learning_rate": 8.432432432432433e-05, + "loss": 0.2035, + "step": 755 + }, + { + "epoch": 0.19027244698924053, + "grad_norm": 0.13403065502643585, + "learning_rate": 8.416988416988417e-05, + "loss": 0.178, + "step": 756 + }, + { + "epoch": 0.1905241301201787, + "grad_norm": 0.12914441525936127, + "learning_rate": 8.401544401544403e-05, + "loss": 0.2966, + "step": 757 + }, + { + "epoch": 0.19077581325111684, + "grad_norm": 0.1367819607257843, + "learning_rate": 8.386100386100386e-05, + "loss": 0.2672, + "step": 758 + }, + { + "epoch": 0.191027496382055, + "grad_norm": 0.13592910766601562, + "learning_rate": 8.370656370656372e-05, + "loss": 0.2255, + "step": 759 + }, + { + "epoch": 0.19127917951299314, + "grad_norm": 0.13383309543132782, + "learning_rate": 8.355212355212354e-05, + "loss": 0.285, + "step": 760 + }, + { + "epoch": 0.1915308626439313, + "grad_norm": 0.10186809301376343, + "learning_rate": 8.33976833976834e-05, + "loss": 0.1208, + "step": 761 + }, + { + "epoch": 0.19178254577486945, + "grad_norm": 0.16211846470832825, + "learning_rate": 8.324324324324326e-05, + "loss": 0.1956, + "step": 762 + }, + { + "epoch": 0.1920342289058076, + "grad_norm": 0.1024671196937561, + "learning_rate": 8.308880308880309e-05, + "loss": 0.1781, + "step": 763 + }, + { + "epoch": 0.19228591203674575, + "grad_norm": 0.1122509241104126, + "learning_rate": 8.293436293436294e-05, + "loss": 0.199, + "step": 764 + }, + { + "epoch": 0.19253759516768387, + "grad_norm": 0.11130458861589432, + "learning_rate": 8.277992277992277e-05, + "loss": 0.2311, + "step": 765 + }, + { + "epoch": 0.19278927829862202, + "grad_norm": 0.12414205074310303, + "learning_rate": 8.262548262548263e-05, + "loss": 0.3013, + "step": 766 + }, + { + "epoch": 0.19304096142956018, + "grad_norm": 0.13851679861545563, + "learning_rate": 8.247104247104247e-05, + "loss": 0.2744, + "step": 767 + }, + { + "epoch": 0.19329264456049833, + "grad_norm": 0.11790241301059723, + "learning_rate": 8.231660231660232e-05, + "loss": 0.1932, + "step": 768 + }, + { + "epoch": 0.19354432769143648, + "grad_norm": 0.10062595456838608, + "learning_rate": 8.216216216216217e-05, + "loss": 0.1663, + "step": 769 + }, + { + "epoch": 0.19379601082237463, + "grad_norm": 0.12206324189901352, + "learning_rate": 8.2007722007722e-05, + "loss": 0.214, + "step": 770 + }, + { + "epoch": 0.19404769395331278, + "grad_norm": 0.11738628149032593, + "learning_rate": 8.185328185328186e-05, + "loss": 0.1529, + "step": 771 + }, + { + "epoch": 0.19429937708425093, + "grad_norm": 0.10208728909492493, + "learning_rate": 8.16988416988417e-05, + "loss": 0.1877, + "step": 772 + }, + { + "epoch": 0.1945510602151891, + "grad_norm": 0.10993005335330963, + "learning_rate": 8.154440154440155e-05, + "loss": 0.181, + "step": 773 + }, + { + "epoch": 0.19480274334612724, + "grad_norm": 0.11965054273605347, + "learning_rate": 8.138996138996139e-05, + "loss": 0.1821, + "step": 774 + }, + { + "epoch": 0.19505442647706536, + "grad_norm": 0.1099630743265152, + "learning_rate": 8.123552123552123e-05, + "loss": 0.1931, + "step": 775 + }, + { + "epoch": 0.19530610960800351, + "grad_norm": 0.1116647720336914, + "learning_rate": 8.108108108108109e-05, + "loss": 0.1867, + "step": 776 + }, + { + "epoch": 0.19555779273894167, + "grad_norm": 0.11637338250875473, + "learning_rate": 8.092664092664093e-05, + "loss": 0.2831, + "step": 777 + }, + { + "epoch": 0.19580947586987982, + "grad_norm": 0.11428670585155487, + "learning_rate": 8.077220077220078e-05, + "loss": 0.1925, + "step": 778 + }, + { + "epoch": 0.19606115900081797, + "grad_norm": 0.09395647048950195, + "learning_rate": 8.061776061776062e-05, + "loss": 0.1446, + "step": 779 + }, + { + "epoch": 0.19631284213175612, + "grad_norm": 0.10720973461866379, + "learning_rate": 8.046332046332046e-05, + "loss": 0.2163, + "step": 780 + }, + { + "epoch": 0.19656452526269427, + "grad_norm": 0.11614006012678146, + "learning_rate": 8.03088803088803e-05, + "loss": 0.2011, + "step": 781 + }, + { + "epoch": 0.19681620839363242, + "grad_norm": 0.11895287781953812, + "learning_rate": 8.015444015444016e-05, + "loss": 0.2878, + "step": 782 + }, + { + "epoch": 0.19706789152457058, + "grad_norm": 0.12292661517858505, + "learning_rate": 8e-05, + "loss": 0.1807, + "step": 783 + }, + { + "epoch": 0.19731957465550873, + "grad_norm": 0.13356351852416992, + "learning_rate": 7.984555984555985e-05, + "loss": 0.2066, + "step": 784 + }, + { + "epoch": 0.19757125778644685, + "grad_norm": 0.13021701574325562, + "learning_rate": 7.969111969111969e-05, + "loss": 0.3049, + "step": 785 + }, + { + "epoch": 0.197822940917385, + "grad_norm": 0.10717355459928513, + "learning_rate": 7.953667953667954e-05, + "loss": 0.2006, + "step": 786 + }, + { + "epoch": 0.19807462404832316, + "grad_norm": 0.12250859290361404, + "learning_rate": 7.938223938223939e-05, + "loss": 0.1622, + "step": 787 + }, + { + "epoch": 0.1983263071792613, + "grad_norm": 0.12550446391105652, + "learning_rate": 7.922779922779922e-05, + "loss": 0.2137, + "step": 788 + }, + { + "epoch": 0.19857799031019946, + "grad_norm": 0.0988641008734703, + "learning_rate": 7.907335907335908e-05, + "loss": 0.1484, + "step": 789 + }, + { + "epoch": 0.1988296734411376, + "grad_norm": 0.11315549165010452, + "learning_rate": 7.891891891891892e-05, + "loss": 0.1915, + "step": 790 + }, + { + "epoch": 0.19908135657207576, + "grad_norm": 0.10526572912931442, + "learning_rate": 7.876447876447877e-05, + "loss": 0.1574, + "step": 791 + }, + { + "epoch": 0.19933303970301391, + "grad_norm": 0.12042815238237381, + "learning_rate": 7.861003861003862e-05, + "loss": 0.2781, + "step": 792 + }, + { + "epoch": 0.19958472283395207, + "grad_norm": 0.1455085426568985, + "learning_rate": 7.845559845559845e-05, + "loss": 0.2089, + "step": 793 + }, + { + "epoch": 0.1998364059648902, + "grad_norm": 0.1152004525065422, + "learning_rate": 7.830115830115831e-05, + "loss": 0.1896, + "step": 794 + }, + { + "epoch": 0.20008808909582834, + "grad_norm": 0.1096256673336029, + "learning_rate": 7.814671814671814e-05, + "loss": 0.19, + "step": 795 + }, + { + "epoch": 0.2003397722267665, + "grad_norm": 0.11116250604391098, + "learning_rate": 7.7992277992278e-05, + "loss": 0.1948, + "step": 796 + }, + { + "epoch": 0.20059145535770465, + "grad_norm": 0.11579190194606781, + "learning_rate": 7.783783783783785e-05, + "loss": 0.1929, + "step": 797 + }, + { + "epoch": 0.2008431384886428, + "grad_norm": 0.11348525434732437, + "learning_rate": 7.768339768339768e-05, + "loss": 0.1867, + "step": 798 + }, + { + "epoch": 0.20109482161958095, + "grad_norm": 0.12241964787244797, + "learning_rate": 7.752895752895754e-05, + "loss": 0.2096, + "step": 799 + }, + { + "epoch": 0.2013465047505191, + "grad_norm": 0.1293066143989563, + "learning_rate": 7.737451737451737e-05, + "loss": 0.3055, + "step": 800 + }, + { + "epoch": 0.20159818788145725, + "grad_norm": 0.12381697446107864, + "learning_rate": 7.722007722007723e-05, + "loss": 0.1725, + "step": 801 + }, + { + "epoch": 0.2018498710123954, + "grad_norm": 0.10360271483659744, + "learning_rate": 7.706563706563707e-05, + "loss": 0.1879, + "step": 802 + }, + { + "epoch": 0.20210155414333356, + "grad_norm": 0.1031133309006691, + "learning_rate": 7.691119691119691e-05, + "loss": 0.1523, + "step": 803 + }, + { + "epoch": 0.20235323727427168, + "grad_norm": 0.11991957575082779, + "learning_rate": 7.675675675675677e-05, + "loss": 0.2003, + "step": 804 + }, + { + "epoch": 0.20260492040520983, + "grad_norm": 0.13465255498886108, + "learning_rate": 7.66023166023166e-05, + "loss": 0.2861, + "step": 805 + }, + { + "epoch": 0.20285660353614798, + "grad_norm": 0.12251454591751099, + "learning_rate": 7.644787644787645e-05, + "loss": 0.2761, + "step": 806 + }, + { + "epoch": 0.20310828666708614, + "grad_norm": 0.12354443967342377, + "learning_rate": 7.62934362934363e-05, + "loss": 0.2901, + "step": 807 + }, + { + "epoch": 0.2033599697980243, + "grad_norm": 0.11855776607990265, + "learning_rate": 7.613899613899614e-05, + "loss": 0.1942, + "step": 808 + }, + { + "epoch": 0.20361165292896244, + "grad_norm": 0.10894890129566193, + "learning_rate": 7.598455598455598e-05, + "loss": 0.1823, + "step": 809 + }, + { + "epoch": 0.2038633360599006, + "grad_norm": 0.10938367992639542, + "learning_rate": 7.583011583011583e-05, + "loss": 0.1954, + "step": 810 + }, + { + "epoch": 0.20411501919083874, + "grad_norm": 0.10300524532794952, + "learning_rate": 7.567567567567568e-05, + "loss": 0.1544, + "step": 811 + }, + { + "epoch": 0.2043667023217769, + "grad_norm": 0.13688234984874725, + "learning_rate": 7.552123552123553e-05, + "loss": 0.2932, + "step": 812 + }, + { + "epoch": 0.20461838545271502, + "grad_norm": 0.12232501059770584, + "learning_rate": 7.536679536679537e-05, + "loss": 0.2582, + "step": 813 + }, + { + "epoch": 0.20487006858365317, + "grad_norm": 0.1373576819896698, + "learning_rate": 7.521235521235521e-05, + "loss": 0.3144, + "step": 814 + }, + { + "epoch": 0.20512175171459132, + "grad_norm": 0.11639385670423508, + "learning_rate": 7.505791505791506e-05, + "loss": 0.2176, + "step": 815 + }, + { + "epoch": 0.20537343484552947, + "grad_norm": 0.1113649383187294, + "learning_rate": 7.49034749034749e-05, + "loss": 0.2129, + "step": 816 + }, + { + "epoch": 0.20562511797646763, + "grad_norm": 0.13591669499874115, + "learning_rate": 7.474903474903476e-05, + "loss": 0.2749, + "step": 817 + }, + { + "epoch": 0.20587680110740578, + "grad_norm": 0.10466151684522629, + "learning_rate": 7.45945945945946e-05, + "loss": 0.1653, + "step": 818 + }, + { + "epoch": 0.20612848423834393, + "grad_norm": 0.10343621671199799, + "learning_rate": 7.444015444015444e-05, + "loss": 0.1837, + "step": 819 + }, + { + "epoch": 0.20638016736928208, + "grad_norm": 0.12417241185903549, + "learning_rate": 7.428571428571429e-05, + "loss": 0.211, + "step": 820 + }, + { + "epoch": 0.20663185050022023, + "grad_norm": 0.11205418407917023, + "learning_rate": 7.413127413127413e-05, + "loss": 0.159, + "step": 821 + }, + { + "epoch": 0.20688353363115838, + "grad_norm": 0.10791382938623428, + "learning_rate": 7.397683397683399e-05, + "loss": 0.2108, + "step": 822 + }, + { + "epoch": 0.2071352167620965, + "grad_norm": 0.12284215539693832, + "learning_rate": 7.382239382239382e-05, + "loss": 0.2149, + "step": 823 + }, + { + "epoch": 0.20738689989303466, + "grad_norm": 0.09839356690645218, + "learning_rate": 7.366795366795367e-05, + "loss": 0.1486, + "step": 824 + }, + { + "epoch": 0.2076385830239728, + "grad_norm": 0.10778332501649857, + "learning_rate": 7.351351351351352e-05, + "loss": 0.1828, + "step": 825 + }, + { + "epoch": 0.20789026615491096, + "grad_norm": 0.14378227293491364, + "learning_rate": 7.335907335907336e-05, + "loss": 0.2987, + "step": 826 + }, + { + "epoch": 0.20814194928584911, + "grad_norm": 0.1336110681295395, + "learning_rate": 7.320463320463322e-05, + "loss": 0.2666, + "step": 827 + }, + { + "epoch": 0.20839363241678727, + "grad_norm": 0.12167420983314514, + "learning_rate": 7.305019305019305e-05, + "loss": 0.2431, + "step": 828 + }, + { + "epoch": 0.20864531554772542, + "grad_norm": 0.11411017179489136, + "learning_rate": 7.28957528957529e-05, + "loss": 0.1648, + "step": 829 + }, + { + "epoch": 0.20889699867866357, + "grad_norm": 0.11440549045801163, + "learning_rate": 7.274131274131273e-05, + "loss": 0.2089, + "step": 830 + }, + { + "epoch": 0.20914868180960172, + "grad_norm": 0.13502226769924164, + "learning_rate": 7.258687258687259e-05, + "loss": 0.1743, + "step": 831 + }, + { + "epoch": 0.20940036494053987, + "grad_norm": 0.10134825110435486, + "learning_rate": 7.243243243243245e-05, + "loss": 0.1896, + "step": 832 + }, + { + "epoch": 0.209652048071478, + "grad_norm": 0.11911708116531372, + "learning_rate": 7.227799227799228e-05, + "loss": 0.2703, + "step": 833 + }, + { + "epoch": 0.20990373120241615, + "grad_norm": 0.122975654900074, + "learning_rate": 7.212355212355213e-05, + "loss": 0.2072, + "step": 834 + }, + { + "epoch": 0.2101554143333543, + "grad_norm": 0.1166321337223053, + "learning_rate": 7.196911196911196e-05, + "loss": 0.2147, + "step": 835 + }, + { + "epoch": 0.21040709746429245, + "grad_norm": 0.10795572400093079, + "learning_rate": 7.181467181467182e-05, + "loss": 0.1988, + "step": 836 + }, + { + "epoch": 0.2106587805952306, + "grad_norm": 0.11532330513000488, + "learning_rate": 7.166023166023166e-05, + "loss": 0.194, + "step": 837 + }, + { + "epoch": 0.21091046372616876, + "grad_norm": 0.11524844914674759, + "learning_rate": 7.15057915057915e-05, + "loss": 0.1776, + "step": 838 + }, + { + "epoch": 0.2111621468571069, + "grad_norm": 0.12468288838863373, + "learning_rate": 7.135135135135136e-05, + "loss": 0.2111, + "step": 839 + }, + { + "epoch": 0.21141382998804506, + "grad_norm": 0.11456232517957687, + "learning_rate": 7.119691119691119e-05, + "loss": 0.217, + "step": 840 + }, + { + "epoch": 0.2116655131189832, + "grad_norm": 0.11296792328357697, + "learning_rate": 7.104247104247105e-05, + "loss": 0.2073, + "step": 841 + }, + { + "epoch": 0.21191719624992134, + "grad_norm": 0.11052672564983368, + "learning_rate": 7.088803088803089e-05, + "loss": 0.1997, + "step": 842 + }, + { + "epoch": 0.2121688793808595, + "grad_norm": 0.12768107652664185, + "learning_rate": 7.073359073359074e-05, + "loss": 0.2371, + "step": 843 + }, + { + "epoch": 0.21242056251179764, + "grad_norm": 0.1103857234120369, + "learning_rate": 7.057915057915059e-05, + "loss": 0.1666, + "step": 844 + }, + { + "epoch": 0.2126722456427358, + "grad_norm": 0.11844266951084137, + "learning_rate": 7.042471042471042e-05, + "loss": 0.3073, + "step": 845 + }, + { + "epoch": 0.21292392877367394, + "grad_norm": 0.12780922651290894, + "learning_rate": 7.027027027027028e-05, + "loss": 0.2033, + "step": 846 + }, + { + "epoch": 0.2131756119046121, + "grad_norm": 0.11559857428073883, + "learning_rate": 7.011583011583012e-05, + "loss": 0.2162, + "step": 847 + }, + { + "epoch": 0.21342729503555025, + "grad_norm": 0.12350855022668839, + "learning_rate": 6.996138996138996e-05, + "loss": 0.178, + "step": 848 + }, + { + "epoch": 0.2136789781664884, + "grad_norm": 0.11439146846532822, + "learning_rate": 6.980694980694981e-05, + "loss": 0.1957, + "step": 849 + }, + { + "epoch": 0.21393066129742655, + "grad_norm": 0.10569050908088684, + "learning_rate": 6.965250965250965e-05, + "loss": 0.1788, + "step": 850 + }, + { + "epoch": 0.2141823444283647, + "grad_norm": 0.09878722578287125, + "learning_rate": 6.949806949806951e-05, + "loss": 0.1613, + "step": 851 + }, + { + "epoch": 0.21443402755930283, + "grad_norm": 0.12721562385559082, + "learning_rate": 6.934362934362935e-05, + "loss": 0.2728, + "step": 852 + }, + { + "epoch": 0.21468571069024098, + "grad_norm": 0.1345331072807312, + "learning_rate": 6.91891891891892e-05, + "loss": 0.2529, + "step": 853 + }, + { + "epoch": 0.21493739382117913, + "grad_norm": 0.1275331676006317, + "learning_rate": 6.903474903474904e-05, + "loss": 0.2961, + "step": 854 + }, + { + "epoch": 0.21518907695211728, + "grad_norm": 0.135565385222435, + "learning_rate": 6.888030888030888e-05, + "loss": 0.2719, + "step": 855 + }, + { + "epoch": 0.21544076008305543, + "grad_norm": 0.107664093375206, + "learning_rate": 6.872586872586872e-05, + "loss": 0.2193, + "step": 856 + }, + { + "epoch": 0.21569244321399358, + "grad_norm": 0.13568522036075592, + "learning_rate": 6.857142857142858e-05, + "loss": 0.3006, + "step": 857 + }, + { + "epoch": 0.21594412634493174, + "grad_norm": 0.1316979080438614, + "learning_rate": 6.841698841698842e-05, + "loss": 0.2839, + "step": 858 + }, + { + "epoch": 0.2161958094758699, + "grad_norm": 0.1251004934310913, + "learning_rate": 6.826254826254827e-05, + "loss": 0.1826, + "step": 859 + }, + { + "epoch": 0.21644749260680804, + "grad_norm": 0.12274324148893356, + "learning_rate": 6.810810810810811e-05, + "loss": 0.2829, + "step": 860 + }, + { + "epoch": 0.2166991757377462, + "grad_norm": 0.1308038979768753, + "learning_rate": 6.795366795366795e-05, + "loss": 0.2849, + "step": 861 + }, + { + "epoch": 0.21695085886868432, + "grad_norm": 0.11672983318567276, + "learning_rate": 6.779922779922781e-05, + "loss": 0.1756, + "step": 862 + }, + { + "epoch": 0.21720254199962247, + "grad_norm": 0.13185741007328033, + "learning_rate": 6.764478764478764e-05, + "loss": 0.2019, + "step": 863 + }, + { + "epoch": 0.21745422513056062, + "grad_norm": 0.10524323582649231, + "learning_rate": 6.74903474903475e-05, + "loss": 0.1693, + "step": 864 + }, + { + "epoch": 0.21770590826149877, + "grad_norm": 0.1012788638472557, + "learning_rate": 6.733590733590734e-05, + "loss": 0.1737, + "step": 865 + }, + { + "epoch": 0.21795759139243692, + "grad_norm": 0.14006006717681885, + "learning_rate": 6.718146718146718e-05, + "loss": 0.2057, + "step": 866 + }, + { + "epoch": 0.21820927452337507, + "grad_norm": 0.1357884407043457, + "learning_rate": 6.702702702702704e-05, + "loss": 0.2091, + "step": 867 + }, + { + "epoch": 0.21846095765431323, + "grad_norm": 0.13076336681842804, + "learning_rate": 6.687258687258687e-05, + "loss": 0.2354, + "step": 868 + }, + { + "epoch": 0.21871264078525138, + "grad_norm": 0.10596498847007751, + "learning_rate": 6.671814671814673e-05, + "loss": 0.1924, + "step": 869 + }, + { + "epoch": 0.21896432391618953, + "grad_norm": 0.11557600647211075, + "learning_rate": 6.656370656370656e-05, + "loss": 0.2115, + "step": 870 + }, + { + "epoch": 0.21921600704712765, + "grad_norm": 0.11588539928197861, + "learning_rate": 6.640926640926641e-05, + "loss": 0.1805, + "step": 871 + }, + { + "epoch": 0.2194676901780658, + "grad_norm": 0.10284875333309174, + "learning_rate": 6.625482625482627e-05, + "loss": 0.1946, + "step": 872 + }, + { + "epoch": 0.21971937330900396, + "grad_norm": 0.11946608871221542, + "learning_rate": 6.61003861003861e-05, + "loss": 0.254, + "step": 873 + }, + { + "epoch": 0.2199710564399421, + "grad_norm": 0.09503907710313797, + "learning_rate": 6.594594594594596e-05, + "loss": 0.1742, + "step": 874 + }, + { + "epoch": 0.22022273957088026, + "grad_norm": 0.11438967287540436, + "learning_rate": 6.579150579150579e-05, + "loss": 0.1804, + "step": 875 + }, + { + "epoch": 0.2204744227018184, + "grad_norm": 0.10719439387321472, + "learning_rate": 6.563706563706564e-05, + "loss": 0.1788, + "step": 876 + }, + { + "epoch": 0.22072610583275656, + "grad_norm": 0.1386880874633789, + "learning_rate": 6.548262548262549e-05, + "loss": 0.2026, + "step": 877 + }, + { + "epoch": 0.22097778896369472, + "grad_norm": 0.11157010495662689, + "learning_rate": 6.532818532818533e-05, + "loss": 0.2139, + "step": 878 + }, + { + "epoch": 0.22122947209463287, + "grad_norm": 0.1380106806755066, + "learning_rate": 6.517374517374519e-05, + "loss": 0.1641, + "step": 879 + }, + { + "epoch": 0.22148115522557102, + "grad_norm": 0.11523478478193283, + "learning_rate": 6.501930501930502e-05, + "loss": 0.2079, + "step": 880 + }, + { + "epoch": 0.22173283835650914, + "grad_norm": 0.12161862105131149, + "learning_rate": 6.486486486486487e-05, + "loss": 0.1756, + "step": 881 + }, + { + "epoch": 0.2219845214874473, + "grad_norm": 0.14189398288726807, + "learning_rate": 6.471042471042472e-05, + "loss": 0.3237, + "step": 882 + }, + { + "epoch": 0.22223620461838545, + "grad_norm": 0.11399094015359879, + "learning_rate": 6.455598455598456e-05, + "loss": 0.2346, + "step": 883 + }, + { + "epoch": 0.2224878877493236, + "grad_norm": 0.1166088730096817, + "learning_rate": 6.44015444015444e-05, + "loss": 0.2696, + "step": 884 + }, + { + "epoch": 0.22273957088026175, + "grad_norm": 0.1107102707028389, + "learning_rate": 6.424710424710425e-05, + "loss": 0.1893, + "step": 885 + }, + { + "epoch": 0.2229912540111999, + "grad_norm": 0.11463375389575958, + "learning_rate": 6.40926640926641e-05, + "loss": 0.2022, + "step": 886 + }, + { + "epoch": 0.22324293714213805, + "grad_norm": 0.11348404735326767, + "learning_rate": 6.393822393822395e-05, + "loss": 0.1998, + "step": 887 + }, + { + "epoch": 0.2234946202730762, + "grad_norm": 0.11484857648611069, + "learning_rate": 6.378378378378379e-05, + "loss": 0.2261, + "step": 888 + }, + { + "epoch": 0.22374630340401436, + "grad_norm": 0.12992657721042633, + "learning_rate": 6.362934362934363e-05, + "loss": 0.2644, + "step": 889 + }, + { + "epoch": 0.22399798653495248, + "grad_norm": 0.11617530137300491, + "learning_rate": 6.347490347490347e-05, + "loss": 0.1849, + "step": 890 + }, + { + "epoch": 0.22424966966589063, + "grad_norm": 0.1282077431678772, + "learning_rate": 6.332046332046332e-05, + "loss": 0.2497, + "step": 891 + }, + { + "epoch": 0.22450135279682878, + "grad_norm": 0.15674304962158203, + "learning_rate": 6.316602316602317e-05, + "loss": 0.269, + "step": 892 + }, + { + "epoch": 0.22475303592776694, + "grad_norm": 0.12478800117969513, + "learning_rate": 6.301158301158302e-05, + "loss": 0.2814, + "step": 893 + }, + { + "epoch": 0.2250047190587051, + "grad_norm": 0.11571218818426132, + "learning_rate": 6.285714285714286e-05, + "loss": 0.1436, + "step": 894 + }, + { + "epoch": 0.22525640218964324, + "grad_norm": 0.1241186335682869, + "learning_rate": 6.27027027027027e-05, + "loss": 0.1882, + "step": 895 + }, + { + "epoch": 0.2255080853205814, + "grad_norm": 0.11000651121139526, + "learning_rate": 6.254826254826255e-05, + "loss": 0.2101, + "step": 896 + }, + { + "epoch": 0.22575976845151954, + "grad_norm": 0.11669962853193283, + "learning_rate": 6.23938223938224e-05, + "loss": 0.2011, + "step": 897 + }, + { + "epoch": 0.2260114515824577, + "grad_norm": 0.12564580142498016, + "learning_rate": 6.223938223938223e-05, + "loss": 0.2734, + "step": 898 + }, + { + "epoch": 0.22626313471339585, + "grad_norm": 0.11603422462940216, + "learning_rate": 6.208494208494209e-05, + "loss": 0.2436, + "step": 899 + }, + { + "epoch": 0.22651481784433397, + "grad_norm": 0.11243658512830734, + "learning_rate": 6.193050193050193e-05, + "loss": 0.1952, + "step": 900 + }, + { + "epoch": 0.22676650097527212, + "grad_norm": 0.11419893056154251, + "learning_rate": 6.177606177606178e-05, + "loss": 0.1639, + "step": 901 + }, + { + "epoch": 0.22701818410621027, + "grad_norm": 0.12908916175365448, + "learning_rate": 6.162162162162163e-05, + "loss": 0.2544, + "step": 902 + }, + { + "epoch": 0.22726986723714843, + "grad_norm": 0.09652577340602875, + "learning_rate": 6.146718146718146e-05, + "loss": 0.1485, + "step": 903 + }, + { + "epoch": 0.22752155036808658, + "grad_norm": 0.12185376137495041, + "learning_rate": 6.131274131274132e-05, + "loss": 0.2263, + "step": 904 + }, + { + "epoch": 0.22777323349902473, + "grad_norm": 0.10491900146007538, + "learning_rate": 6.115830115830115e-05, + "loss": 0.1594, + "step": 905 + }, + { + "epoch": 0.22802491662996288, + "grad_norm": 0.12206004559993744, + "learning_rate": 6.100386100386101e-05, + "loss": 0.2818, + "step": 906 + }, + { + "epoch": 0.22827659976090103, + "grad_norm": 0.12587115168571472, + "learning_rate": 6.084942084942086e-05, + "loss": 0.2062, + "step": 907 + }, + { + "epoch": 0.22852828289183919, + "grad_norm": 0.10318470746278763, + "learning_rate": 6.0694980694980693e-05, + "loss": 0.1931, + "step": 908 + }, + { + "epoch": 0.22877996602277734, + "grad_norm": 0.11015098541975021, + "learning_rate": 6.0540540540540543e-05, + "loss": 0.1856, + "step": 909 + }, + { + "epoch": 0.22903164915371546, + "grad_norm": 0.11355525255203247, + "learning_rate": 6.038610038610039e-05, + "loss": 0.19, + "step": 910 + }, + { + "epoch": 0.2292833322846536, + "grad_norm": 0.12140153348445892, + "learning_rate": 6.023166023166024e-05, + "loss": 0.2126, + "step": 911 + }, + { + "epoch": 0.22953501541559176, + "grad_norm": 0.12254661321640015, + "learning_rate": 6.007722007722007e-05, + "loss": 0.1616, + "step": 912 + }, + { + "epoch": 0.22978669854652992, + "grad_norm": 0.1052892729640007, + "learning_rate": 5.992277992277992e-05, + "loss": 0.1796, + "step": 913 + }, + { + "epoch": 0.23003838167746807, + "grad_norm": 0.11111687868833542, + "learning_rate": 5.976833976833977e-05, + "loss": 0.2553, + "step": 914 + }, + { + "epoch": 0.23029006480840622, + "grad_norm": 0.11671284586191177, + "learning_rate": 5.9613899613899616e-05, + "loss": 0.1943, + "step": 915 + }, + { + "epoch": 0.23054174793934437, + "grad_norm": 0.11089365184307098, + "learning_rate": 5.9459459459459466e-05, + "loss": 0.1675, + "step": 916 + }, + { + "epoch": 0.23079343107028252, + "grad_norm": 0.12005841732025146, + "learning_rate": 5.93050193050193e-05, + "loss": 0.2225, + "step": 917 + }, + { + "epoch": 0.23104511420122067, + "grad_norm": 0.10551342368125916, + "learning_rate": 5.915057915057915e-05, + "loss": 0.2001, + "step": 918 + }, + { + "epoch": 0.2312967973321588, + "grad_norm": 0.13088083267211914, + "learning_rate": 5.8996138996138996e-05, + "loss": 0.2214, + "step": 919 + }, + { + "epoch": 0.23154848046309695, + "grad_norm": 0.1163930594921112, + "learning_rate": 5.8841698841698846e-05, + "loss": 0.1948, + "step": 920 + }, + { + "epoch": 0.2318001635940351, + "grad_norm": 0.12209770828485489, + "learning_rate": 5.8687258687258696e-05, + "loss": 0.3061, + "step": 921 + }, + { + "epoch": 0.23205184672497325, + "grad_norm": 0.10689982771873474, + "learning_rate": 5.853281853281853e-05, + "loss": 0.1761, + "step": 922 + }, + { + "epoch": 0.2323035298559114, + "grad_norm": 0.11450164765119553, + "learning_rate": 5.837837837837838e-05, + "loss": 0.2311, + "step": 923 + }, + { + "epoch": 0.23255521298684956, + "grad_norm": 0.09272663295269012, + "learning_rate": 5.8223938223938226e-05, + "loss": 0.1614, + "step": 924 + }, + { + "epoch": 0.2328068961177877, + "grad_norm": 0.12887226045131683, + "learning_rate": 5.8069498069498076e-05, + "loss": 0.2614, + "step": 925 + }, + { + "epoch": 0.23305857924872586, + "grad_norm": 0.15005137026309967, + "learning_rate": 5.791505791505791e-05, + "loss": 0.254, + "step": 926 + }, + { + "epoch": 0.233310262379664, + "grad_norm": 0.12406851351261139, + "learning_rate": 5.776061776061776e-05, + "loss": 0.2363, + "step": 927 + }, + { + "epoch": 0.23356194551060216, + "grad_norm": 0.11630482971668243, + "learning_rate": 5.760617760617761e-05, + "loss": 0.2065, + "step": 928 + }, + { + "epoch": 0.2338136286415403, + "grad_norm": 0.10879343748092651, + "learning_rate": 5.7451737451737455e-05, + "loss": 0.191, + "step": 929 + }, + { + "epoch": 0.23406531177247844, + "grad_norm": 0.1258416324853897, + "learning_rate": 5.7297297297297305e-05, + "loss": 0.2545, + "step": 930 + }, + { + "epoch": 0.2343169949034166, + "grad_norm": 0.13407813012599945, + "learning_rate": 5.714285714285714e-05, + "loss": 0.2601, + "step": 931 + }, + { + "epoch": 0.23456867803435474, + "grad_norm": 0.11370331794023514, + "learning_rate": 5.698841698841699e-05, + "loss": 0.185, + "step": 932 + }, + { + "epoch": 0.2348203611652929, + "grad_norm": 0.12003123760223389, + "learning_rate": 5.6833976833976835e-05, + "loss": 0.1964, + "step": 933 + }, + { + "epoch": 0.23507204429623105, + "grad_norm": 0.11761590093374252, + "learning_rate": 5.6679536679536685e-05, + "loss": 0.1902, + "step": 934 + }, + { + "epoch": 0.2353237274271692, + "grad_norm": 0.11095016449689865, + "learning_rate": 5.6525096525096535e-05, + "loss": 0.2197, + "step": 935 + }, + { + "epoch": 0.23557541055810735, + "grad_norm": 0.11618047952651978, + "learning_rate": 5.637065637065637e-05, + "loss": 0.2197, + "step": 936 + }, + { + "epoch": 0.2358270936890455, + "grad_norm": 0.10983315110206604, + "learning_rate": 5.621621621621622e-05, + "loss": 0.2007, + "step": 937 + }, + { + "epoch": 0.23607877681998363, + "grad_norm": 0.10982047766447067, + "learning_rate": 5.6061776061776065e-05, + "loss": 0.2031, + "step": 938 + }, + { + "epoch": 0.23633045995092178, + "grad_norm": 0.12462889403104782, + "learning_rate": 5.5907335907335915e-05, + "loss": 0.2436, + "step": 939 + }, + { + "epoch": 0.23658214308185993, + "grad_norm": 0.12027290463447571, + "learning_rate": 5.575289575289575e-05, + "loss": 0.283, + "step": 940 + }, + { + "epoch": 0.23683382621279808, + "grad_norm": 0.12372776120901108, + "learning_rate": 5.55984555984556e-05, + "loss": 0.2614, + "step": 941 + }, + { + "epoch": 0.23708550934373623, + "grad_norm": 0.11352698504924774, + "learning_rate": 5.544401544401545e-05, + "loss": 0.2116, + "step": 942 + }, + { + "epoch": 0.23733719247467439, + "grad_norm": 0.11478787660598755, + "learning_rate": 5.528957528957529e-05, + "loss": 0.1521, + "step": 943 + }, + { + "epoch": 0.23758887560561254, + "grad_norm": 0.11719299107789993, + "learning_rate": 5.5135135135135144e-05, + "loss": 0.177, + "step": 944 + }, + { + "epoch": 0.2378405587365507, + "grad_norm": 0.13160140812397003, + "learning_rate": 5.498069498069498e-05, + "loss": 0.2915, + "step": 945 + }, + { + "epoch": 0.23809224186748884, + "grad_norm": 0.1344316303730011, + "learning_rate": 5.482625482625483e-05, + "loss": 0.2428, + "step": 946 + }, + { + "epoch": 0.238343924998427, + "grad_norm": 0.1322656124830246, + "learning_rate": 5.467181467181467e-05, + "loss": 0.2133, + "step": 947 + }, + { + "epoch": 0.23859560812936512, + "grad_norm": 0.11685141175985336, + "learning_rate": 5.451737451737452e-05, + "loss": 0.1811, + "step": 948 + }, + { + "epoch": 0.23884729126030327, + "grad_norm": 0.10659373551607132, + "learning_rate": 5.436293436293437e-05, + "loss": 0.1772, + "step": 949 + }, + { + "epoch": 0.23909897439124142, + "grad_norm": 0.11862076818943024, + "learning_rate": 5.420849420849421e-05, + "loss": 0.1728, + "step": 950 + }, + { + "epoch": 0.23935065752217957, + "grad_norm": 0.1008882224559784, + "learning_rate": 5.405405405405406e-05, + "loss": 0.1658, + "step": 951 + }, + { + "epoch": 0.23960234065311772, + "grad_norm": 0.1069093868136406, + "learning_rate": 5.38996138996139e-05, + "loss": 0.1808, + "step": 952 + }, + { + "epoch": 0.23985402378405588, + "grad_norm": 0.12278815358877182, + "learning_rate": 5.374517374517375e-05, + "loss": 0.1975, + "step": 953 + }, + { + "epoch": 0.24010570691499403, + "grad_norm": 0.1403549462556839, + "learning_rate": 5.359073359073359e-05, + "loss": 0.2607, + "step": 954 + }, + { + "epoch": 0.24035739004593218, + "grad_norm": 0.1228654757142067, + "learning_rate": 5.343629343629344e-05, + "loss": 0.2108, + "step": 955 + }, + { + "epoch": 0.24060907317687033, + "grad_norm": 0.1225048378109932, + "learning_rate": 5.328185328185329e-05, + "loss": 0.2604, + "step": 956 + }, + { + "epoch": 0.24086075630780848, + "grad_norm": 0.1180121898651123, + "learning_rate": 5.3127413127413126e-05, + "loss": 0.3247, + "step": 957 + }, + { + "epoch": 0.2411124394387466, + "grad_norm": 0.11166714131832123, + "learning_rate": 5.2972972972972976e-05, + "loss": 0.1987, + "step": 958 + }, + { + "epoch": 0.24136412256968476, + "grad_norm": 0.1221095398068428, + "learning_rate": 5.281853281853282e-05, + "loss": 0.2362, + "step": 959 + }, + { + "epoch": 0.2416158057006229, + "grad_norm": 0.1138242557644844, + "learning_rate": 5.266409266409267e-05, + "loss": 0.1971, + "step": 960 + }, + { + "epoch": 0.24186748883156106, + "grad_norm": 0.10679937154054642, + "learning_rate": 5.2509652509652506e-05, + "loss": 0.1947, + "step": 961 + }, + { + "epoch": 0.2421191719624992, + "grad_norm": 0.11496231704950333, + "learning_rate": 5.2355212355212356e-05, + "loss": 0.1869, + "step": 962 + }, + { + "epoch": 0.24237085509343737, + "grad_norm": 0.10841769725084305, + "learning_rate": 5.2200772200772206e-05, + "loss": 0.1788, + "step": 963 + }, + { + "epoch": 0.24262253822437552, + "grad_norm": 0.11322899162769318, + "learning_rate": 5.204633204633205e-05, + "loss": 0.1774, + "step": 964 + }, + { + "epoch": 0.24287422135531367, + "grad_norm": 0.1382610946893692, + "learning_rate": 5.18918918918919e-05, + "loss": 0.2627, + "step": 965 + }, + { + "epoch": 0.24312590448625182, + "grad_norm": 0.1277177929878235, + "learning_rate": 5.1737451737451736e-05, + "loss": 0.2139, + "step": 966 + }, + { + "epoch": 0.24337758761718994, + "grad_norm": 0.13496394455432892, + "learning_rate": 5.1583011583011586e-05, + "loss": 0.2972, + "step": 967 + }, + { + "epoch": 0.2436292707481281, + "grad_norm": 0.0899004116654396, + "learning_rate": 5.142857142857143e-05, + "loss": 0.1317, + "step": 968 + }, + { + "epoch": 0.24388095387906625, + "grad_norm": 0.11566326767206192, + "learning_rate": 5.127413127413128e-05, + "loss": 0.2163, + "step": 969 + }, + { + "epoch": 0.2441326370100044, + "grad_norm": 0.14185094833374023, + "learning_rate": 5.111969111969113e-05, + "loss": 0.2879, + "step": 970 + }, + { + "epoch": 0.24438432014094255, + "grad_norm": 0.12760186195373535, + "learning_rate": 5.0965250965250965e-05, + "loss": 0.2838, + "step": 971 + }, + { + "epoch": 0.2446360032718807, + "grad_norm": 0.14247183501720428, + "learning_rate": 5.0810810810810815e-05, + "loss": 0.214, + "step": 972 + }, + { + "epoch": 0.24488768640281885, + "grad_norm": 0.12118181586265564, + "learning_rate": 5.065637065637066e-05, + "loss": 0.1591, + "step": 973 + }, + { + "epoch": 0.245139369533757, + "grad_norm": 0.11757013201713562, + "learning_rate": 5.050193050193051e-05, + "loss": 0.2389, + "step": 974 + }, + { + "epoch": 0.24539105266469516, + "grad_norm": 0.10428398102521896, + "learning_rate": 5.0347490347490345e-05, + "loss": 0.1821, + "step": 975 + }, + { + "epoch": 0.2456427357956333, + "grad_norm": 0.11219316720962524, + "learning_rate": 5.0193050193050195e-05, + "loss": 0.1693, + "step": 976 + }, + { + "epoch": 0.24589441892657143, + "grad_norm": 0.12479031831026077, + "learning_rate": 5.0038610038610045e-05, + "loss": 0.2121, + "step": 977 + }, + { + "epoch": 0.24614610205750959, + "grad_norm": 0.10959812998771667, + "learning_rate": 4.988416988416988e-05, + "loss": 0.1886, + "step": 978 + }, + { + "epoch": 0.24639778518844774, + "grad_norm": 0.13046742975711823, + "learning_rate": 4.972972972972974e-05, + "loss": 0.3019, + "step": 979 + }, + { + "epoch": 0.2466494683193859, + "grad_norm": 0.12494894117116928, + "learning_rate": 4.957528957528958e-05, + "loss": 0.1586, + "step": 980 + }, + { + "epoch": 0.24690115145032404, + "grad_norm": 0.1257307231426239, + "learning_rate": 4.9420849420849425e-05, + "loss": 0.225, + "step": 981 + }, + { + "epoch": 0.2471528345812622, + "grad_norm": 0.10836142301559448, + "learning_rate": 4.926640926640927e-05, + "loss": 0.1808, + "step": 982 + }, + { + "epoch": 0.24740451771220034, + "grad_norm": 0.14597921073436737, + "learning_rate": 4.911196911196911e-05, + "loss": 0.2758, + "step": 983 + }, + { + "epoch": 0.2476562008431385, + "grad_norm": 0.10853477567434311, + "learning_rate": 4.895752895752896e-05, + "loss": 0.1828, + "step": 984 + }, + { + "epoch": 0.24790788397407665, + "grad_norm": 0.10994373261928558, + "learning_rate": 4.8803088803088804e-05, + "loss": 0.1703, + "step": 985 + }, + { + "epoch": 0.2481595671050148, + "grad_norm": 0.11955688893795013, + "learning_rate": 4.8648648648648654e-05, + "loss": 0.1887, + "step": 986 + }, + { + "epoch": 0.24841125023595292, + "grad_norm": 0.12595857679843903, + "learning_rate": 4.84942084942085e-05, + "loss": 0.2445, + "step": 987 + }, + { + "epoch": 0.24866293336689108, + "grad_norm": 0.1134994626045227, + "learning_rate": 4.833976833976834e-05, + "loss": 0.2567, + "step": 988 + }, + { + "epoch": 0.24891461649782923, + "grad_norm": 0.10119155794382095, + "learning_rate": 4.818532818532819e-05, + "loss": 0.1677, + "step": 989 + }, + { + "epoch": 0.24916629962876738, + "grad_norm": 0.12207161635160446, + "learning_rate": 4.8030888030888034e-05, + "loss": 0.1904, + "step": 990 + }, + { + "epoch": 0.24941798275970553, + "grad_norm": 0.11913091689348221, + "learning_rate": 4.787644787644788e-05, + "loss": 0.2693, + "step": 991 + }, + { + "epoch": 0.24966966589064368, + "grad_norm": 0.120506651699543, + "learning_rate": 4.772200772200772e-05, + "loss": 0.2458, + "step": 992 + }, + { + "epoch": 0.24992134902158183, + "grad_norm": 0.10583953559398651, + "learning_rate": 4.756756756756757e-05, + "loss": 0.1755, + "step": 993 + }, + { + "epoch": 0.25017303215251996, + "grad_norm": 0.12434154003858566, + "learning_rate": 4.741312741312742e-05, + "loss": 0.2477, + "step": 994 + }, + { + "epoch": 0.2504247152834581, + "grad_norm": 0.1402542144060135, + "learning_rate": 4.7258687258687264e-05, + "loss": 0.2101, + "step": 995 + }, + { + "epoch": 0.25067639841439626, + "grad_norm": 0.10040777176618576, + "learning_rate": 4.710424710424711e-05, + "loss": 0.171, + "step": 996 + }, + { + "epoch": 0.2509280815453344, + "grad_norm": 0.11074397712945938, + "learning_rate": 4.694980694980695e-05, + "loss": 0.1682, + "step": 997 + }, + { + "epoch": 0.25117976467627257, + "grad_norm": 0.11190413683652878, + "learning_rate": 4.679536679536679e-05, + "loss": 0.1814, + "step": 998 + }, + { + "epoch": 0.2514314478072107, + "grad_norm": 0.1140817180275917, + "learning_rate": 4.664092664092664e-05, + "loss": 0.1956, + "step": 999 + }, + { + "epoch": 0.25168313093814887, + "grad_norm": 0.13233628869056702, + "learning_rate": 4.648648648648649e-05, + "loss": 0.2883, + "step": 1000 + }, + { + "epoch": 0.251934814069087, + "grad_norm": 0.13028374314308167, + "learning_rate": 4.6332046332046336e-05, + "loss": 0.2954, + "step": 1001 + }, + { + "epoch": 0.2521864972000252, + "grad_norm": 0.13944485783576965, + "learning_rate": 4.617760617760618e-05, + "loss": 0.2667, + "step": 1002 + }, + { + "epoch": 0.2524381803309633, + "grad_norm": 0.12177547812461853, + "learning_rate": 4.602316602316602e-05, + "loss": 0.1686, + "step": 1003 + }, + { + "epoch": 0.2526898634619015, + "grad_norm": 0.12015821039676666, + "learning_rate": 4.586872586872587e-05, + "loss": 0.1894, + "step": 1004 + }, + { + "epoch": 0.25294154659283963, + "grad_norm": 0.12894301116466522, + "learning_rate": 4.5714285714285716e-05, + "loss": 0.2275, + "step": 1005 + }, + { + "epoch": 0.2531932297237778, + "grad_norm": 0.11510162800550461, + "learning_rate": 4.555984555984556e-05, + "loss": 0.1897, + "step": 1006 + }, + { + "epoch": 0.25344491285471593, + "grad_norm": 0.12964919209480286, + "learning_rate": 4.540540540540541e-05, + "loss": 0.2756, + "step": 1007 + }, + { + "epoch": 0.2536965959856541, + "grad_norm": 0.1366814523935318, + "learning_rate": 4.525096525096525e-05, + "loss": 0.2933, + "step": 1008 + }, + { + "epoch": 0.25394827911659223, + "grad_norm": 0.12422391772270203, + "learning_rate": 4.50965250965251e-05, + "loss": 0.2126, + "step": 1009 + }, + { + "epoch": 0.2541999622475304, + "grad_norm": 0.0907035693526268, + "learning_rate": 4.4942084942084946e-05, + "loss": 0.1298, + "step": 1010 + }, + { + "epoch": 0.2544516453784685, + "grad_norm": 0.11650924384593964, + "learning_rate": 4.478764478764479e-05, + "loss": 0.1809, + "step": 1011 + }, + { + "epoch": 0.25470332850940663, + "grad_norm": 0.12858541309833527, + "learning_rate": 4.463320463320463e-05, + "loss": 0.2896, + "step": 1012 + }, + { + "epoch": 0.2549550116403448, + "grad_norm": 0.12392865866422653, + "learning_rate": 4.4478764478764475e-05, + "loss": 0.2254, + "step": 1013 + }, + { + "epoch": 0.25520669477128294, + "grad_norm": 0.1028500497341156, + "learning_rate": 4.432432432432433e-05, + "loss": 0.1751, + "step": 1014 + }, + { + "epoch": 0.2554583779022211, + "grad_norm": 0.10513060539960861, + "learning_rate": 4.4169884169884175e-05, + "loss": 0.2129, + "step": 1015 + }, + { + "epoch": 0.25571006103315924, + "grad_norm": 0.10929843783378601, + "learning_rate": 4.401544401544402e-05, + "loss": 0.16, + "step": 1016 + }, + { + "epoch": 0.2559617441640974, + "grad_norm": 0.11750845611095428, + "learning_rate": 4.386100386100386e-05, + "loss": 0.1741, + "step": 1017 + }, + { + "epoch": 0.25621342729503555, + "grad_norm": 0.13358919322490692, + "learning_rate": 4.3706563706563705e-05, + "loss": 0.1978, + "step": 1018 + }, + { + "epoch": 0.2564651104259737, + "grad_norm": 0.11565891653299332, + "learning_rate": 4.3552123552123555e-05, + "loss": 0.1914, + "step": 1019 + }, + { + "epoch": 0.25671679355691185, + "grad_norm": 0.13126766681671143, + "learning_rate": 4.33976833976834e-05, + "loss": 0.3173, + "step": 1020 + }, + { + "epoch": 0.25696847668785, + "grad_norm": 0.12171319872140884, + "learning_rate": 4.324324324324325e-05, + "loss": 0.2111, + "step": 1021 + }, + { + "epoch": 0.25722015981878815, + "grad_norm": 0.11284907162189484, + "learning_rate": 4.308880308880309e-05, + "loss": 0.1929, + "step": 1022 + }, + { + "epoch": 0.2574718429497263, + "grad_norm": 0.11088172346353531, + "learning_rate": 4.2934362934362935e-05, + "loss": 0.1978, + "step": 1023 + }, + { + "epoch": 0.25772352608066446, + "grad_norm": 0.11636247485876083, + "learning_rate": 4.2779922779922785e-05, + "loss": 0.2287, + "step": 1024 + }, + { + "epoch": 0.2579752092116026, + "grad_norm": 0.104326993227005, + "learning_rate": 4.262548262548263e-05, + "loss": 0.1667, + "step": 1025 + }, + { + "epoch": 0.25822689234254076, + "grad_norm": 0.1274103820323944, + "learning_rate": 4.247104247104247e-05, + "loss": 0.2782, + "step": 1026 + }, + { + "epoch": 0.2584785754734789, + "grad_norm": 0.10540760308504105, + "learning_rate": 4.2316602316602314e-05, + "loss": 0.1441, + "step": 1027 + }, + { + "epoch": 0.25873025860441706, + "grad_norm": 0.10421574115753174, + "learning_rate": 4.2162162162162164e-05, + "loss": 0.158, + "step": 1028 + }, + { + "epoch": 0.2589819417353552, + "grad_norm": 0.1166294738650322, + "learning_rate": 4.2007722007722014e-05, + "loss": 0.2506, + "step": 1029 + }, + { + "epoch": 0.2592336248662933, + "grad_norm": 0.10888038575649261, + "learning_rate": 4.185328185328186e-05, + "loss": 0.1857, + "step": 1030 + }, + { + "epoch": 0.25948530799723146, + "grad_norm": 0.16468745470046997, + "learning_rate": 4.16988416988417e-05, + "loss": 0.1869, + "step": 1031 + }, + { + "epoch": 0.2597369911281696, + "grad_norm": 0.12184029817581177, + "learning_rate": 4.1544401544401544e-05, + "loss": 0.215, + "step": 1032 + }, + { + "epoch": 0.25998867425910777, + "grad_norm": 0.12397003918886185, + "learning_rate": 4.138996138996139e-05, + "loss": 0.1808, + "step": 1033 + }, + { + "epoch": 0.2602403573900459, + "grad_norm": 0.11407697200775146, + "learning_rate": 4.123552123552124e-05, + "loss": 0.1656, + "step": 1034 + }, + { + "epoch": 0.26049204052098407, + "grad_norm": 0.11733423173427582, + "learning_rate": 4.108108108108109e-05, + "loss": 0.2123, + "step": 1035 + }, + { + "epoch": 0.2607437236519222, + "grad_norm": 0.12729212641716003, + "learning_rate": 4.092664092664093e-05, + "loss": 0.213, + "step": 1036 + }, + { + "epoch": 0.2609954067828604, + "grad_norm": 0.12634289264678955, + "learning_rate": 4.0772200772200774e-05, + "loss": 0.2124, + "step": 1037 + }, + { + "epoch": 0.2612470899137985, + "grad_norm": 0.15637880563735962, + "learning_rate": 4.061776061776062e-05, + "loss": 0.1908, + "step": 1038 + }, + { + "epoch": 0.2614987730447367, + "grad_norm": 0.12071269005537033, + "learning_rate": 4.046332046332047e-05, + "loss": 0.2043, + "step": 1039 + }, + { + "epoch": 0.26175045617567483, + "grad_norm": 0.1071467399597168, + "learning_rate": 4.030888030888031e-05, + "loss": 0.1806, + "step": 1040 + }, + { + "epoch": 0.262002139306613, + "grad_norm": 0.12470516562461853, + "learning_rate": 4.015444015444015e-05, + "loss": 0.1849, + "step": 1041 + }, + { + "epoch": 0.26225382243755113, + "grad_norm": 0.1274518519639969, + "learning_rate": 4e-05, + "loss": 0.2224, + "step": 1042 + }, + { + "epoch": 0.2625055055684893, + "grad_norm": 0.12582527101039886, + "learning_rate": 3.9845559845559846e-05, + "loss": 0.2578, + "step": 1043 + }, + { + "epoch": 0.26275718869942744, + "grad_norm": 0.10735000669956207, + "learning_rate": 3.9691119691119696e-05, + "loss": 0.1636, + "step": 1044 + }, + { + "epoch": 0.2630088718303656, + "grad_norm": 0.11986963450908661, + "learning_rate": 3.953667953667954e-05, + "loss": 0.208, + "step": 1045 + }, + { + "epoch": 0.26326055496130374, + "grad_norm": 0.12065369635820389, + "learning_rate": 3.938223938223938e-05, + "loss": 0.2018, + "step": 1046 + }, + { + "epoch": 0.2635122380922419, + "grad_norm": 0.14128756523132324, + "learning_rate": 3.9227799227799226e-05, + "loss": 0.18, + "step": 1047 + }, + { + "epoch": 0.26376392122318004, + "grad_norm": 0.12504349648952484, + "learning_rate": 3.907335907335907e-05, + "loss": 0.2789, + "step": 1048 + }, + { + "epoch": 0.26401560435411814, + "grad_norm": 0.10742396861314774, + "learning_rate": 3.8918918918918926e-05, + "loss": 0.1801, + "step": 1049 + }, + { + "epoch": 0.2642672874850563, + "grad_norm": 0.16062429547309875, + "learning_rate": 3.876447876447877e-05, + "loss": 0.1722, + "step": 1050 + }, + { + "epoch": 0.26451897061599444, + "grad_norm": 0.11531528830528259, + "learning_rate": 3.861003861003861e-05, + "loss": 0.1965, + "step": 1051 + }, + { + "epoch": 0.2647706537469326, + "grad_norm": 0.11625973135232925, + "learning_rate": 3.8455598455598456e-05, + "loss": 0.1837, + "step": 1052 + }, + { + "epoch": 0.26502233687787075, + "grad_norm": 0.11149144917726517, + "learning_rate": 3.83011583011583e-05, + "loss": 0.2426, + "step": 1053 + }, + { + "epoch": 0.2652740200088089, + "grad_norm": 0.10706043988466263, + "learning_rate": 3.814671814671815e-05, + "loss": 0.1851, + "step": 1054 + }, + { + "epoch": 0.26552570313974705, + "grad_norm": 0.11466892808675766, + "learning_rate": 3.799227799227799e-05, + "loss": 0.1901, + "step": 1055 + }, + { + "epoch": 0.2657773862706852, + "grad_norm": 0.11885873228311539, + "learning_rate": 3.783783783783784e-05, + "loss": 0.2046, + "step": 1056 + }, + { + "epoch": 0.26602906940162335, + "grad_norm": 0.1276128590106964, + "learning_rate": 3.7683397683397685e-05, + "loss": 0.2049, + "step": 1057 + }, + { + "epoch": 0.2662807525325615, + "grad_norm": 0.11042508482933044, + "learning_rate": 3.752895752895753e-05, + "loss": 0.1624, + "step": 1058 + }, + { + "epoch": 0.26653243566349966, + "grad_norm": 0.10830891132354736, + "learning_rate": 3.737451737451738e-05, + "loss": 0.2117, + "step": 1059 + }, + { + "epoch": 0.2667841187944378, + "grad_norm": 0.10290905088186264, + "learning_rate": 3.722007722007722e-05, + "loss": 0.1841, + "step": 1060 + }, + { + "epoch": 0.26703580192537596, + "grad_norm": 0.14225976169109344, + "learning_rate": 3.7065637065637065e-05, + "loss": 0.2795, + "step": 1061 + }, + { + "epoch": 0.2672874850563141, + "grad_norm": 0.13055454194545746, + "learning_rate": 3.691119691119691e-05, + "loss": 0.1753, + "step": 1062 + }, + { + "epoch": 0.26753916818725226, + "grad_norm": 0.12350232899188995, + "learning_rate": 3.675675675675676e-05, + "loss": 0.2056, + "step": 1063 + }, + { + "epoch": 0.2677908513181904, + "grad_norm": 0.12295592576265335, + "learning_rate": 3.660231660231661e-05, + "loss": 0.2627, + "step": 1064 + }, + { + "epoch": 0.26804253444912857, + "grad_norm": 0.11139633506536484, + "learning_rate": 3.644787644787645e-05, + "loss": 0.1929, + "step": 1065 + }, + { + "epoch": 0.2682942175800667, + "grad_norm": 0.0964319109916687, + "learning_rate": 3.6293436293436295e-05, + "loss": 0.1412, + "step": 1066 + }, + { + "epoch": 0.26854590071100487, + "grad_norm": 0.11253803223371506, + "learning_rate": 3.613899613899614e-05, + "loss": 0.1559, + "step": 1067 + }, + { + "epoch": 0.26879758384194297, + "grad_norm": 0.11267100274562836, + "learning_rate": 3.598455598455598e-05, + "loss": 0.1847, + "step": 1068 + }, + { + "epoch": 0.2690492669728811, + "grad_norm": 0.11491706222295761, + "learning_rate": 3.583011583011583e-05, + "loss": 0.1977, + "step": 1069 + }, + { + "epoch": 0.26930095010381927, + "grad_norm": 0.11298020929098129, + "learning_rate": 3.567567567567568e-05, + "loss": 0.2078, + "step": 1070 + }, + { + "epoch": 0.2695526332347574, + "grad_norm": 0.11329969763755798, + "learning_rate": 3.5521235521235524e-05, + "loss": 0.1991, + "step": 1071 + }, + { + "epoch": 0.2698043163656956, + "grad_norm": 0.13279412686824799, + "learning_rate": 3.536679536679537e-05, + "loss": 0.1957, + "step": 1072 + }, + { + "epoch": 0.2700559994966337, + "grad_norm": 0.1119299829006195, + "learning_rate": 3.521235521235521e-05, + "loss": 0.1968, + "step": 1073 + }, + { + "epoch": 0.2703076826275719, + "grad_norm": 0.14684811234474182, + "learning_rate": 3.505791505791506e-05, + "loss": 0.2674, + "step": 1074 + }, + { + "epoch": 0.27055936575851003, + "grad_norm": 0.12224563956260681, + "learning_rate": 3.4903474903474904e-05, + "loss": 0.2022, + "step": 1075 + }, + { + "epoch": 0.2708110488894482, + "grad_norm": 0.11459530889987946, + "learning_rate": 3.4749034749034754e-05, + "loss": 0.2055, + "step": 1076 + }, + { + "epoch": 0.27106273202038633, + "grad_norm": 0.11826267838478088, + "learning_rate": 3.45945945945946e-05, + "loss": 0.217, + "step": 1077 + }, + { + "epoch": 0.2713144151513245, + "grad_norm": 0.12009011954069138, + "learning_rate": 3.444015444015444e-05, + "loss": 0.1448, + "step": 1078 + }, + { + "epoch": 0.27156609828226264, + "grad_norm": 0.11043018847703934, + "learning_rate": 3.428571428571429e-05, + "loss": 0.1787, + "step": 1079 + }, + { + "epoch": 0.2718177814132008, + "grad_norm": 0.1335824429988861, + "learning_rate": 3.4131274131274134e-05, + "loss": 0.2584, + "step": 1080 + }, + { + "epoch": 0.27206946454413894, + "grad_norm": 0.14351008832454681, + "learning_rate": 3.397683397683398e-05, + "loss": 0.3061, + "step": 1081 + }, + { + "epoch": 0.2723211476750771, + "grad_norm": 0.11424921452999115, + "learning_rate": 3.382239382239382e-05, + "loss": 0.1813, + "step": 1082 + }, + { + "epoch": 0.27257283080601524, + "grad_norm": 0.13373470306396484, + "learning_rate": 3.366795366795367e-05, + "loss": 0.1789, + "step": 1083 + }, + { + "epoch": 0.2728245139369534, + "grad_norm": 0.13649781048297882, + "learning_rate": 3.351351351351352e-05, + "loss": 0.3204, + "step": 1084 + }, + { + "epoch": 0.27307619706789155, + "grad_norm": 0.12645111978054047, + "learning_rate": 3.335907335907336e-05, + "loss": 0.2888, + "step": 1085 + }, + { + "epoch": 0.2733278801988297, + "grad_norm": 0.11262895911931992, + "learning_rate": 3.3204633204633207e-05, + "loss": 0.2273, + "step": 1086 + }, + { + "epoch": 0.27357956332976785, + "grad_norm": 0.12918204069137573, + "learning_rate": 3.305019305019305e-05, + "loss": 0.2737, + "step": 1087 + }, + { + "epoch": 0.27383124646070595, + "grad_norm": 0.11092269420623779, + "learning_rate": 3.289575289575289e-05, + "loss": 0.1919, + "step": 1088 + }, + { + "epoch": 0.2740829295916441, + "grad_norm": 0.12466593086719513, + "learning_rate": 3.274131274131274e-05, + "loss": 0.2753, + "step": 1089 + }, + { + "epoch": 0.27433461272258225, + "grad_norm": 0.13460558652877808, + "learning_rate": 3.258687258687259e-05, + "loss": 0.3344, + "step": 1090 + }, + { + "epoch": 0.2745862958535204, + "grad_norm": 0.12301404029130936, + "learning_rate": 3.2432432432432436e-05, + "loss": 0.2199, + "step": 1091 + }, + { + "epoch": 0.27483797898445855, + "grad_norm": 0.1156139075756073, + "learning_rate": 3.227799227799228e-05, + "loss": 0.196, + "step": 1092 + }, + { + "epoch": 0.2750896621153967, + "grad_norm": 0.12295875698328018, + "learning_rate": 3.212355212355212e-05, + "loss": 0.2201, + "step": 1093 + }, + { + "epoch": 0.27534134524633486, + "grad_norm": 0.09860794991254807, + "learning_rate": 3.196911196911197e-05, + "loss": 0.1419, + "step": 1094 + }, + { + "epoch": 0.275593028377273, + "grad_norm": 0.11538466811180115, + "learning_rate": 3.1814671814671816e-05, + "loss": 0.1863, + "step": 1095 + }, + { + "epoch": 0.27584471150821116, + "grad_norm": 0.1137920543551445, + "learning_rate": 3.166023166023166e-05, + "loss": 0.1761, + "step": 1096 + }, + { + "epoch": 0.2760963946391493, + "grad_norm": 0.1273881494998932, + "learning_rate": 3.150579150579151e-05, + "loss": 0.3375, + "step": 1097 + }, + { + "epoch": 0.27634807777008746, + "grad_norm": 0.10999948531389236, + "learning_rate": 3.135135135135135e-05, + "loss": 0.1784, + "step": 1098 + }, + { + "epoch": 0.2765997609010256, + "grad_norm": 0.10580787062644958, + "learning_rate": 3.11969111969112e-05, + "loss": 0.1779, + "step": 1099 + }, + { + "epoch": 0.27685144403196377, + "grad_norm": 0.12471788376569748, + "learning_rate": 3.1042471042471045e-05, + "loss": 0.2171, + "step": 1100 + }, + { + "epoch": 0.2771031271629019, + "grad_norm": 0.11509516835212708, + "learning_rate": 3.088803088803089e-05, + "loss": 0.2189, + "step": 1101 + }, + { + "epoch": 0.27735481029384007, + "grad_norm": 0.12886741757392883, + "learning_rate": 3.073359073359073e-05, + "loss": 0.304, + "step": 1102 + }, + { + "epoch": 0.2776064934247782, + "grad_norm": 0.1041407361626625, + "learning_rate": 3.0579150579150575e-05, + "loss": 0.1663, + "step": 1103 + }, + { + "epoch": 0.2778581765557164, + "grad_norm": 0.12129200994968414, + "learning_rate": 3.042471042471043e-05, + "loss": 0.1743, + "step": 1104 + }, + { + "epoch": 0.2781098596866545, + "grad_norm": 0.1064925342798233, + "learning_rate": 3.0270270270270272e-05, + "loss": 0.1867, + "step": 1105 + }, + { + "epoch": 0.2783615428175927, + "grad_norm": 0.10880186408758163, + "learning_rate": 3.011583011583012e-05, + "loss": 0.171, + "step": 1106 + }, + { + "epoch": 0.2786132259485308, + "grad_norm": 0.10235221683979034, + "learning_rate": 2.996138996138996e-05, + "loss": 0.1826, + "step": 1107 + }, + { + "epoch": 0.2788649090794689, + "grad_norm": 0.12178605794906616, + "learning_rate": 2.9806949806949808e-05, + "loss": 0.1979, + "step": 1108 + }, + { + "epoch": 0.2791165922104071, + "grad_norm": 0.11338433623313904, + "learning_rate": 2.965250965250965e-05, + "loss": 0.1605, + "step": 1109 + }, + { + "epoch": 0.27936827534134523, + "grad_norm": 0.11511190235614777, + "learning_rate": 2.9498069498069498e-05, + "loss": 0.2119, + "step": 1110 + }, + { + "epoch": 0.2796199584722834, + "grad_norm": 0.11978597939014435, + "learning_rate": 2.9343629343629348e-05, + "loss": 0.1909, + "step": 1111 + }, + { + "epoch": 0.27987164160322153, + "grad_norm": 0.10626047104597092, + "learning_rate": 2.918918918918919e-05, + "loss": 0.1616, + "step": 1112 + }, + { + "epoch": 0.2801233247341597, + "grad_norm": 0.12202031165361404, + "learning_rate": 2.9034749034749038e-05, + "loss": 0.1987, + "step": 1113 + }, + { + "epoch": 0.28037500786509784, + "grad_norm": 0.10144208371639252, + "learning_rate": 2.888030888030888e-05, + "loss": 0.1748, + "step": 1114 + }, + { + "epoch": 0.280626690996036, + "grad_norm": 0.11613299697637558, + "learning_rate": 2.8725868725868728e-05, + "loss": 0.2086, + "step": 1115 + }, + { + "epoch": 0.28087837412697414, + "grad_norm": 0.11896048486232758, + "learning_rate": 2.857142857142857e-05, + "loss": 0.2033, + "step": 1116 + }, + { + "epoch": 0.2811300572579123, + "grad_norm": 0.12568190693855286, + "learning_rate": 2.8416988416988417e-05, + "loss": 0.1941, + "step": 1117 + }, + { + "epoch": 0.28138174038885044, + "grad_norm": 0.13146355748176575, + "learning_rate": 2.8262548262548267e-05, + "loss": 0.2298, + "step": 1118 + }, + { + "epoch": 0.2816334235197886, + "grad_norm": 0.11802760511636734, + "learning_rate": 2.810810810810811e-05, + "loss": 0.2136, + "step": 1119 + }, + { + "epoch": 0.28188510665072675, + "grad_norm": 0.11035969108343124, + "learning_rate": 2.7953667953667957e-05, + "loss": 0.1761, + "step": 1120 + }, + { + "epoch": 0.2821367897816649, + "grad_norm": 0.12584011256694794, + "learning_rate": 2.77992277992278e-05, + "loss": 0.2633, + "step": 1121 + }, + { + "epoch": 0.28238847291260305, + "grad_norm": 0.11829424649477005, + "learning_rate": 2.7644787644787644e-05, + "loss": 0.2321, + "step": 1122 + }, + { + "epoch": 0.2826401560435412, + "grad_norm": 0.12350109964609146, + "learning_rate": 2.749034749034749e-05, + "loss": 0.3102, + "step": 1123 + }, + { + "epoch": 0.28289183917447935, + "grad_norm": 0.10934913903474808, + "learning_rate": 2.7335907335907334e-05, + "loss": 0.1602, + "step": 1124 + }, + { + "epoch": 0.2831435223054175, + "grad_norm": 0.11251812428236008, + "learning_rate": 2.7181467181467184e-05, + "loss": 0.198, + "step": 1125 + }, + { + "epoch": 0.2833952054363556, + "grad_norm": 0.11449240148067474, + "learning_rate": 2.702702702702703e-05, + "loss": 0.1898, + "step": 1126 + }, + { + "epoch": 0.28364688856729375, + "grad_norm": 0.1228322982788086, + "learning_rate": 2.6872586872586873e-05, + "loss": 0.215, + "step": 1127 + }, + { + "epoch": 0.2838985716982319, + "grad_norm": 0.11673863232135773, + "learning_rate": 2.671814671814672e-05, + "loss": 0.1724, + "step": 1128 + }, + { + "epoch": 0.28415025482917006, + "grad_norm": 0.11994639039039612, + "learning_rate": 2.6563706563706563e-05, + "loss": 0.2286, + "step": 1129 + }, + { + "epoch": 0.2844019379601082, + "grad_norm": 0.13262636959552765, + "learning_rate": 2.640926640926641e-05, + "loss": 0.2558, + "step": 1130 + }, + { + "epoch": 0.28465362109104636, + "grad_norm": 0.10788439214229584, + "learning_rate": 2.6254826254826253e-05, + "loss": 0.1642, + "step": 1131 + }, + { + "epoch": 0.2849053042219845, + "grad_norm": 0.13081598281860352, + "learning_rate": 2.6100386100386103e-05, + "loss": 0.2936, + "step": 1132 + }, + { + "epoch": 0.28515698735292266, + "grad_norm": 0.11097247898578644, + "learning_rate": 2.594594594594595e-05, + "loss": 0.2002, + "step": 1133 + }, + { + "epoch": 0.2854086704838608, + "grad_norm": 0.11326999962329865, + "learning_rate": 2.5791505791505793e-05, + "loss": 0.1732, + "step": 1134 + }, + { + "epoch": 0.28566035361479897, + "grad_norm": 0.1345786303281784, + "learning_rate": 2.563706563706564e-05, + "loss": 0.2436, + "step": 1135 + }, + { + "epoch": 0.2859120367457371, + "grad_norm": 0.1398163139820099, + "learning_rate": 2.5482625482625483e-05, + "loss": 0.2196, + "step": 1136 + }, + { + "epoch": 0.28616371987667527, + "grad_norm": 0.1091829314827919, + "learning_rate": 2.532818532818533e-05, + "loss": 0.1729, + "step": 1137 + }, + { + "epoch": 0.2864154030076134, + "grad_norm": 0.11381202191114426, + "learning_rate": 2.5173745173745172e-05, + "loss": 0.1655, + "step": 1138 + }, + { + "epoch": 0.2866670861385516, + "grad_norm": 0.14988526701927185, + "learning_rate": 2.5019305019305022e-05, + "loss": 0.263, + "step": 1139 + }, + { + "epoch": 0.2869187692694897, + "grad_norm": 0.11513830721378326, + "learning_rate": 2.486486486486487e-05, + "loss": 0.1617, + "step": 1140 + }, + { + "epoch": 0.2871704524004279, + "grad_norm": 0.10990703850984573, + "learning_rate": 2.4710424710424712e-05, + "loss": 0.1692, + "step": 1141 + }, + { + "epoch": 0.28742213553136603, + "grad_norm": 0.13732917606830597, + "learning_rate": 2.4555984555984556e-05, + "loss": 0.2716, + "step": 1142 + }, + { + "epoch": 0.2876738186623042, + "grad_norm": 0.10894232243299484, + "learning_rate": 2.4401544401544402e-05, + "loss": 0.1739, + "step": 1143 + }, + { + "epoch": 0.28792550179324233, + "grad_norm": 0.11809521913528442, + "learning_rate": 2.424710424710425e-05, + "loss": 0.1501, + "step": 1144 + }, + { + "epoch": 0.28817718492418043, + "grad_norm": 0.12969113886356354, + "learning_rate": 2.4092664092664095e-05, + "loss": 0.2403, + "step": 1145 + }, + { + "epoch": 0.2884288680551186, + "grad_norm": 0.09926270693540573, + "learning_rate": 2.393822393822394e-05, + "loss": 0.1596, + "step": 1146 + }, + { + "epoch": 0.28868055118605673, + "grad_norm": 0.11534673720598221, + "learning_rate": 2.3783783783783785e-05, + "loss": 0.1842, + "step": 1147 + }, + { + "epoch": 0.2889322343169949, + "grad_norm": 0.10872384160757065, + "learning_rate": 2.3629343629343632e-05, + "loss": 0.1582, + "step": 1148 + }, + { + "epoch": 0.28918391744793304, + "grad_norm": 0.14010004699230194, + "learning_rate": 2.3474903474903475e-05, + "loss": 0.3286, + "step": 1149 + }, + { + "epoch": 0.2894356005788712, + "grad_norm": 0.12276498228311539, + "learning_rate": 2.332046332046332e-05, + "loss": 0.1778, + "step": 1150 + }, + { + "epoch": 0.28968728370980934, + "grad_norm": 0.11340644955635071, + "learning_rate": 2.3166023166023168e-05, + "loss": 0.1839, + "step": 1151 + }, + { + "epoch": 0.2899389668407475, + "grad_norm": 0.11367323249578476, + "learning_rate": 2.301158301158301e-05, + "loss": 0.1569, + "step": 1152 + }, + { + "epoch": 0.29019064997168564, + "grad_norm": 0.10651504993438721, + "learning_rate": 2.2857142857142858e-05, + "loss": 0.1751, + "step": 1153 + }, + { + "epoch": 0.2904423331026238, + "grad_norm": 0.10850241035223007, + "learning_rate": 2.2702702702702705e-05, + "loss": 0.1948, + "step": 1154 + }, + { + "epoch": 0.29069401623356195, + "grad_norm": 0.10044048726558685, + "learning_rate": 2.254826254826255e-05, + "loss": 0.1357, + "step": 1155 + }, + { + "epoch": 0.2909456993645001, + "grad_norm": 0.10623396188020706, + "learning_rate": 2.2393822393822394e-05, + "loss": 0.1618, + "step": 1156 + }, + { + "epoch": 0.29119738249543825, + "grad_norm": 0.10969967395067215, + "learning_rate": 2.2239382239382238e-05, + "loss": 0.1808, + "step": 1157 + }, + { + "epoch": 0.2914490656263764, + "grad_norm": 0.1167631521821022, + "learning_rate": 2.2084942084942088e-05, + "loss": 0.1531, + "step": 1158 + }, + { + "epoch": 0.29170074875731455, + "grad_norm": 0.12506425380706787, + "learning_rate": 2.193050193050193e-05, + "loss": 0.2705, + "step": 1159 + }, + { + "epoch": 0.2919524318882527, + "grad_norm": 0.10216249525547028, + "learning_rate": 2.1776061776061778e-05, + "loss": 0.1652, + "step": 1160 + }, + { + "epoch": 0.29220411501919086, + "grad_norm": 0.15487869083881378, + "learning_rate": 2.1621621621621624e-05, + "loss": 0.2346, + "step": 1161 + }, + { + "epoch": 0.292455798150129, + "grad_norm": 0.120756596326828, + "learning_rate": 2.1467181467181467e-05, + "loss": 0.2041, + "step": 1162 + }, + { + "epoch": 0.29270748128106716, + "grad_norm": 0.1200776919722557, + "learning_rate": 2.1312741312741314e-05, + "loss": 0.2739, + "step": 1163 + }, + { + "epoch": 0.2929591644120053, + "grad_norm": 0.11220280826091766, + "learning_rate": 2.1158301158301157e-05, + "loss": 0.2138, + "step": 1164 + }, + { + "epoch": 0.2932108475429434, + "grad_norm": 0.12042269110679626, + "learning_rate": 2.1003861003861007e-05, + "loss": 0.2959, + "step": 1165 + }, + { + "epoch": 0.29346253067388156, + "grad_norm": 0.11125637590885162, + "learning_rate": 2.084942084942085e-05, + "loss": 0.1978, + "step": 1166 + }, + { + "epoch": 0.2937142138048197, + "grad_norm": 0.1175675168633461, + "learning_rate": 2.0694980694980694e-05, + "loss": 0.1821, + "step": 1167 + }, + { + "epoch": 0.29396589693575786, + "grad_norm": 0.12963427603244781, + "learning_rate": 2.0540540540540544e-05, + "loss": 0.2667, + "step": 1168 + }, + { + "epoch": 0.294217580066696, + "grad_norm": 0.10838277637958527, + "learning_rate": 2.0386100386100387e-05, + "loss": 0.1712, + "step": 1169 + }, + { + "epoch": 0.29446926319763417, + "grad_norm": 0.1149011105298996, + "learning_rate": 2.0231660231660233e-05, + "loss": 0.1772, + "step": 1170 + }, + { + "epoch": 0.2947209463285723, + "grad_norm": 0.10764069855213165, + "learning_rate": 2.0077220077220077e-05, + "loss": 0.1862, + "step": 1171 + }, + { + "epoch": 0.29497262945951047, + "grad_norm": 0.10213491320610046, + "learning_rate": 1.9922779922779923e-05, + "loss": 0.1486, + "step": 1172 + }, + { + "epoch": 0.2952243125904486, + "grad_norm": 0.12230014055967331, + "learning_rate": 1.976833976833977e-05, + "loss": 0.1597, + "step": 1173 + }, + { + "epoch": 0.2954759957213868, + "grad_norm": 0.12827563285827637, + "learning_rate": 1.9613899613899613e-05, + "loss": 0.2108, + "step": 1174 + }, + { + "epoch": 0.2957276788523249, + "grad_norm": 0.10555461049079895, + "learning_rate": 1.9459459459459463e-05, + "loss": 0.1704, + "step": 1175 + }, + { + "epoch": 0.2959793619832631, + "grad_norm": 0.12792570888996124, + "learning_rate": 1.9305019305019306e-05, + "loss": 0.1998, + "step": 1176 + }, + { + "epoch": 0.29623104511420123, + "grad_norm": 0.13367794454097748, + "learning_rate": 1.915057915057915e-05, + "loss": 0.2641, + "step": 1177 + }, + { + "epoch": 0.2964827282451394, + "grad_norm": 0.1238659918308258, + "learning_rate": 1.8996138996138996e-05, + "loss": 0.1796, + "step": 1178 + }, + { + "epoch": 0.29673441137607753, + "grad_norm": 0.1328970342874527, + "learning_rate": 1.8841698841698843e-05, + "loss": 0.2828, + "step": 1179 + }, + { + "epoch": 0.2969860945070157, + "grad_norm": 0.12191575020551682, + "learning_rate": 1.868725868725869e-05, + "loss": 0.2626, + "step": 1180 + }, + { + "epoch": 0.29723777763795384, + "grad_norm": 0.10699662566184998, + "learning_rate": 1.8532818532818533e-05, + "loss": 0.1811, + "step": 1181 + }, + { + "epoch": 0.297489460768892, + "grad_norm": 0.10955280065536499, + "learning_rate": 1.837837837837838e-05, + "loss": 0.2065, + "step": 1182 + }, + { + "epoch": 0.29774114389983014, + "grad_norm": 0.12156296521425247, + "learning_rate": 1.8223938223938226e-05, + "loss": 0.2452, + "step": 1183 + }, + { + "epoch": 0.29799282703076824, + "grad_norm": 0.12040810286998749, + "learning_rate": 1.806949806949807e-05, + "loss": 0.155, + "step": 1184 + }, + { + "epoch": 0.2982445101617064, + "grad_norm": 0.12451086938381195, + "learning_rate": 1.7915057915057916e-05, + "loss": 0.2033, + "step": 1185 + }, + { + "epoch": 0.29849619329264454, + "grad_norm": 0.13171659409999847, + "learning_rate": 1.7760617760617762e-05, + "loss": 0.2662, + "step": 1186 + }, + { + "epoch": 0.2987478764235827, + "grad_norm": 0.11966963112354279, + "learning_rate": 1.7606177606177605e-05, + "loss": 0.1885, + "step": 1187 + }, + { + "epoch": 0.29899955955452084, + "grad_norm": 0.13331931829452515, + "learning_rate": 1.7451737451737452e-05, + "loss": 0.3094, + "step": 1188 + }, + { + "epoch": 0.299251242685459, + "grad_norm": 0.13429853320121765, + "learning_rate": 1.72972972972973e-05, + "loss": 0.2154, + "step": 1189 + }, + { + "epoch": 0.29950292581639715, + "grad_norm": 0.12839801609516144, + "learning_rate": 1.7142857142857145e-05, + "loss": 0.3101, + "step": 1190 + }, + { + "epoch": 0.2997546089473353, + "grad_norm": 0.13523225486278534, + "learning_rate": 1.698841698841699e-05, + "loss": 0.2952, + "step": 1191 + }, + { + "epoch": 0.30000629207827345, + "grad_norm": 0.12616321444511414, + "learning_rate": 1.6833976833976835e-05, + "loss": 0.2037, + "step": 1192 + }, + { + "epoch": 0.3002579752092116, + "grad_norm": 0.12214881926774979, + "learning_rate": 1.667953667953668e-05, + "loss": 0.2694, + "step": 1193 + }, + { + "epoch": 0.30050965834014975, + "grad_norm": 0.13034924864768982, + "learning_rate": 1.6525096525096525e-05, + "loss": 0.3363, + "step": 1194 + }, + { + "epoch": 0.3007613414710879, + "grad_norm": 0.12563075125217438, + "learning_rate": 1.637065637065637e-05, + "loss": 0.1907, + "step": 1195 + }, + { + "epoch": 0.30101302460202606, + "grad_norm": 0.1157495304942131, + "learning_rate": 1.6216216216216218e-05, + "loss": 0.1726, + "step": 1196 + }, + { + "epoch": 0.3012647077329642, + "grad_norm": 0.10682664811611176, + "learning_rate": 1.606177606177606e-05, + "loss": 0.163, + "step": 1197 + }, + { + "epoch": 0.30151639086390236, + "grad_norm": 0.13119934499263763, + "learning_rate": 1.5907335907335908e-05, + "loss": 0.2472, + "step": 1198 + }, + { + "epoch": 0.3017680739948405, + "grad_norm": 0.12205095589160919, + "learning_rate": 1.5752895752895755e-05, + "loss": 0.1765, + "step": 1199 + }, + { + "epoch": 0.30201975712577867, + "grad_norm": 0.11481821537017822, + "learning_rate": 1.55984555984556e-05, + "loss": 0.187, + "step": 1200 + }, + { + "epoch": 0.3022714402567168, + "grad_norm": 0.1158025711774826, + "learning_rate": 1.5444015444015444e-05, + "loss": 0.2702, + "step": 1201 + }, + { + "epoch": 0.30252312338765497, + "grad_norm": 0.11820419132709503, + "learning_rate": 1.5289575289575288e-05, + "loss": 0.2023, + "step": 1202 + }, + { + "epoch": 0.30277480651859306, + "grad_norm": 0.11605869978666306, + "learning_rate": 1.5135135135135136e-05, + "loss": 0.1818, + "step": 1203 + }, + { + "epoch": 0.3030264896495312, + "grad_norm": 0.1274358481168747, + "learning_rate": 1.498069498069498e-05, + "loss": 0.182, + "step": 1204 + }, + { + "epoch": 0.30327817278046937, + "grad_norm": 0.1191435381770134, + "learning_rate": 1.4826254826254826e-05, + "loss": 0.1919, + "step": 1205 + }, + { + "epoch": 0.3035298559114075, + "grad_norm": 0.12055885791778564, + "learning_rate": 1.4671814671814674e-05, + "loss": 0.196, + "step": 1206 + }, + { + "epoch": 0.30378153904234567, + "grad_norm": 0.10810515284538269, + "learning_rate": 1.4517374517374519e-05, + "loss": 0.1373, + "step": 1207 + }, + { + "epoch": 0.3040332221732838, + "grad_norm": 0.134585440158844, + "learning_rate": 1.4362934362934364e-05, + "loss": 0.1998, + "step": 1208 + }, + { + "epoch": 0.304284905304222, + "grad_norm": 0.11716397851705551, + "learning_rate": 1.4208494208494209e-05, + "loss": 0.2083, + "step": 1209 + }, + { + "epoch": 0.3045365884351601, + "grad_norm": 0.11314120888710022, + "learning_rate": 1.4054054054054055e-05, + "loss": 0.1682, + "step": 1210 + }, + { + "epoch": 0.3047882715660983, + "grad_norm": 0.12260904163122177, + "learning_rate": 1.38996138996139e-05, + "loss": 0.2873, + "step": 1211 + }, + { + "epoch": 0.30503995469703643, + "grad_norm": 0.11663558334112167, + "learning_rate": 1.3745173745173745e-05, + "loss": 0.2597, + "step": 1212 + }, + { + "epoch": 0.3052916378279746, + "grad_norm": 0.11076261848211288, + "learning_rate": 1.3590733590733592e-05, + "loss": 0.1931, + "step": 1213 + }, + { + "epoch": 0.30554332095891273, + "grad_norm": 0.12960609793663025, + "learning_rate": 1.3436293436293437e-05, + "loss": 0.2723, + "step": 1214 + }, + { + "epoch": 0.3057950040898509, + "grad_norm": 0.11710008978843689, + "learning_rate": 1.3281853281853282e-05, + "loss": 0.1785, + "step": 1215 + }, + { + "epoch": 0.30604668722078904, + "grad_norm": 0.1324790120124817, + "learning_rate": 1.3127413127413127e-05, + "loss": 0.2531, + "step": 1216 + }, + { + "epoch": 0.3062983703517272, + "grad_norm": 0.10366220027208328, + "learning_rate": 1.2972972972972975e-05, + "loss": 0.1709, + "step": 1217 + }, + { + "epoch": 0.30655005348266534, + "grad_norm": 0.12124229222536087, + "learning_rate": 1.281853281853282e-05, + "loss": 0.2091, + "step": 1218 + }, + { + "epoch": 0.3068017366136035, + "grad_norm": 0.12739507853984833, + "learning_rate": 1.2664092664092665e-05, + "loss": 0.1874, + "step": 1219 + }, + { + "epoch": 0.30705341974454164, + "grad_norm": 0.11585366725921631, + "learning_rate": 1.2509652509652511e-05, + "loss": 0.1994, + "step": 1220 + }, + { + "epoch": 0.3073051028754798, + "grad_norm": 0.11585409939289093, + "learning_rate": 1.2355212355212356e-05, + "loss": 0.1554, + "step": 1221 + }, + { + "epoch": 0.3075567860064179, + "grad_norm": 0.13356682658195496, + "learning_rate": 1.2200772200772201e-05, + "loss": 0.292, + "step": 1222 + }, + { + "epoch": 0.30780846913735604, + "grad_norm": 0.1229170560836792, + "learning_rate": 1.2046332046332048e-05, + "loss": 0.2054, + "step": 1223 + }, + { + "epoch": 0.3080601522682942, + "grad_norm": 0.12661148607730865, + "learning_rate": 1.1891891891891893e-05, + "loss": 0.2907, + "step": 1224 + }, + { + "epoch": 0.30831183539923235, + "grad_norm": 0.12302204221487045, + "learning_rate": 1.1737451737451738e-05, + "loss": 0.2161, + "step": 1225 + }, + { + "epoch": 0.3085635185301705, + "grad_norm": 0.10867659747600555, + "learning_rate": 1.1583011583011584e-05, + "loss": 0.1566, + "step": 1226 + }, + { + "epoch": 0.30881520166110865, + "grad_norm": 0.11101822555065155, + "learning_rate": 1.1428571428571429e-05, + "loss": 0.1808, + "step": 1227 + }, + { + "epoch": 0.3090668847920468, + "grad_norm": 0.12416107207536697, + "learning_rate": 1.1274131274131276e-05, + "loss": 0.2063, + "step": 1228 + }, + { + "epoch": 0.30931856792298495, + "grad_norm": 0.12677694857120514, + "learning_rate": 1.1119691119691119e-05, + "loss": 0.2124, + "step": 1229 + }, + { + "epoch": 0.3095702510539231, + "grad_norm": 0.11091936379671097, + "learning_rate": 1.0965250965250965e-05, + "loss": 0.1775, + "step": 1230 + }, + { + "epoch": 0.30982193418486126, + "grad_norm": 0.1285959929227829, + "learning_rate": 1.0810810810810812e-05, + "loss": 0.1781, + "step": 1231 + }, + { + "epoch": 0.3100736173157994, + "grad_norm": 0.12301789224147797, + "learning_rate": 1.0656370656370657e-05, + "loss": 0.2751, + "step": 1232 + }, + { + "epoch": 0.31032530044673756, + "grad_norm": 0.11078784614801407, + "learning_rate": 1.0501930501930504e-05, + "loss": 0.177, + "step": 1233 + }, + { + "epoch": 0.3105769835776757, + "grad_norm": 0.10243731737136841, + "learning_rate": 1.0347490347490347e-05, + "loss": 0.1627, + "step": 1234 + }, + { + "epoch": 0.31082866670861387, + "grad_norm": 0.11306975036859512, + "learning_rate": 1.0193050193050193e-05, + "loss": 0.1565, + "step": 1235 + }, + { + "epoch": 0.311080349839552, + "grad_norm": 0.14901570975780487, + "learning_rate": 1.0038610038610038e-05, + "loss": 0.2292, + "step": 1236 + }, + { + "epoch": 0.31133203297049017, + "grad_norm": 0.12266213446855545, + "learning_rate": 9.884169884169885e-06, + "loss": 0.1645, + "step": 1237 + }, + { + "epoch": 0.3115837161014283, + "grad_norm": 0.1003609299659729, + "learning_rate": 9.729729729729732e-06, + "loss": 0.1417, + "step": 1238 + }, + { + "epoch": 0.3118353992323665, + "grad_norm": 0.11278005689382553, + "learning_rate": 9.575289575289575e-06, + "loss": 0.2284, + "step": 1239 + }, + { + "epoch": 0.3120870823633046, + "grad_norm": 0.12975087761878967, + "learning_rate": 9.420849420849421e-06, + "loss": 0.1984, + "step": 1240 + }, + { + "epoch": 0.3123387654942428, + "grad_norm": 0.10069584846496582, + "learning_rate": 9.266409266409266e-06, + "loss": 0.1442, + "step": 1241 + }, + { + "epoch": 0.3125904486251809, + "grad_norm": 0.12141278386116028, + "learning_rate": 9.111969111969113e-06, + "loss": 0.1835, + "step": 1242 + }, + { + "epoch": 0.312842131756119, + "grad_norm": 0.1390591412782669, + "learning_rate": 8.957528957528958e-06, + "loss": 0.2455, + "step": 1243 + }, + { + "epoch": 0.3130938148870572, + "grad_norm": 0.13599443435668945, + "learning_rate": 8.803088803088803e-06, + "loss": 0.2671, + "step": 1244 + }, + { + "epoch": 0.3133454980179953, + "grad_norm": 0.10891294479370117, + "learning_rate": 8.64864864864865e-06, + "loss": 0.1854, + "step": 1245 + }, + { + "epoch": 0.3135971811489335, + "grad_norm": 0.12481709569692612, + "learning_rate": 8.494208494208494e-06, + "loss": 0.2603, + "step": 1246 + }, + { + "epoch": 0.31384886427987163, + "grad_norm": 0.1223716139793396, + "learning_rate": 8.33976833976834e-06, + "loss": 0.2172, + "step": 1247 + }, + { + "epoch": 0.3141005474108098, + "grad_norm": 0.11725403368473053, + "learning_rate": 8.185328185328186e-06, + "loss": 0.1812, + "step": 1248 + }, + { + "epoch": 0.31435223054174793, + "grad_norm": 0.1037382185459137, + "learning_rate": 8.03088803088803e-06, + "loss": 0.1726, + "step": 1249 + }, + { + "epoch": 0.3146039136726861, + "grad_norm": 0.10864757746458054, + "learning_rate": 7.876447876447877e-06, + "loss": 0.1774, + "step": 1250 + }, + { + "epoch": 0.31485559680362424, + "grad_norm": 0.13045260310173035, + "learning_rate": 7.722007722007722e-06, + "loss": 0.1964, + "step": 1251 + }, + { + "epoch": 0.3151072799345624, + "grad_norm": 0.11590087413787842, + "learning_rate": 7.567567567567568e-06, + "loss": 0.1642, + "step": 1252 + }, + { + "epoch": 0.31535896306550054, + "grad_norm": 0.1371401995420456, + "learning_rate": 7.413127413127413e-06, + "loss": 0.2655, + "step": 1253 + }, + { + "epoch": 0.3156106461964387, + "grad_norm": 0.1123264878988266, + "learning_rate": 7.2586872586872595e-06, + "loss": 0.2175, + "step": 1254 + }, + { + "epoch": 0.31586232932737685, + "grad_norm": 0.13261693716049194, + "learning_rate": 7.104247104247104e-06, + "loss": 0.2853, + "step": 1255 + }, + { + "epoch": 0.316114012458315, + "grad_norm": 0.12644068896770477, + "learning_rate": 6.94980694980695e-06, + "loss": 0.1902, + "step": 1256 + }, + { + "epoch": 0.31636569558925315, + "grad_norm": 0.12659354507923126, + "learning_rate": 6.795366795366796e-06, + "loss": 0.1894, + "step": 1257 + }, + { + "epoch": 0.3166173787201913, + "grad_norm": 0.12486696243286133, + "learning_rate": 6.640926640926641e-06, + "loss": 0.2347, + "step": 1258 + }, + { + "epoch": 0.31686906185112945, + "grad_norm": 0.11326190084218979, + "learning_rate": 6.486486486486487e-06, + "loss": 0.1862, + "step": 1259 + }, + { + "epoch": 0.3171207449820676, + "grad_norm": 0.13314907252788544, + "learning_rate": 6.332046332046332e-06, + "loss": 0.2987, + "step": 1260 + }, + { + "epoch": 0.3173724281130057, + "grad_norm": 0.11197976768016815, + "learning_rate": 6.177606177606178e-06, + "loss": 0.2069, + "step": 1261 + }, + { + "epoch": 0.31762411124394385, + "grad_norm": 0.12554554641246796, + "learning_rate": 6.023166023166024e-06, + "loss": 0.1725, + "step": 1262 + }, + { + "epoch": 0.317875794374882, + "grad_norm": 0.10719497501850128, + "learning_rate": 5.868725868725869e-06, + "loss": 0.1718, + "step": 1263 + }, + { + "epoch": 0.31812747750582016, + "grad_norm": 0.11357811838388443, + "learning_rate": 5.7142857142857145e-06, + "loss": 0.2049, + "step": 1264 + }, + { + "epoch": 0.3183791606367583, + "grad_norm": 0.11098509281873703, + "learning_rate": 5.5598455598455594e-06, + "loss": 0.175, + "step": 1265 + }, + { + "epoch": 0.31863084376769646, + "grad_norm": 0.10744933038949966, + "learning_rate": 5.405405405405406e-06, + "loss": 0.1786, + "step": 1266 + }, + { + "epoch": 0.3188825268986346, + "grad_norm": 0.11408790200948715, + "learning_rate": 5.250965250965252e-06, + "loss": 0.1976, + "step": 1267 + }, + { + "epoch": 0.31913421002957276, + "grad_norm": 0.11416409909725189, + "learning_rate": 5.096525096525097e-06, + "loss": 0.1789, + "step": 1268 + }, + { + "epoch": 0.3193858931605109, + "grad_norm": 0.1410493105649948, + "learning_rate": 4.9420849420849425e-06, + "loss": 0.239, + "step": 1269 + }, + { + "epoch": 0.31963757629144907, + "grad_norm": 0.113820381462574, + "learning_rate": 4.787644787644787e-06, + "loss": 0.1934, + "step": 1270 + }, + { + "epoch": 0.3198892594223872, + "grad_norm": 0.1171860620379448, + "learning_rate": 4.633204633204633e-06, + "loss": 0.1585, + "step": 1271 + }, + { + "epoch": 0.32014094255332537, + "grad_norm": 0.128767728805542, + "learning_rate": 4.478764478764479e-06, + "loss": 0.1946, + "step": 1272 + }, + { + "epoch": 0.3203926256842635, + "grad_norm": 0.11866892129182816, + "learning_rate": 4.324324324324325e-06, + "loss": 0.1516, + "step": 1273 + }, + { + "epoch": 0.3206443088152017, + "grad_norm": 0.11772280931472778, + "learning_rate": 4.16988416988417e-06, + "loss": 0.2034, + "step": 1274 + }, + { + "epoch": 0.3208959919461398, + "grad_norm": 0.12072817981243134, + "learning_rate": 4.015444015444015e-06, + "loss": 0.1563, + "step": 1275 + }, + { + "epoch": 0.321147675077078, + "grad_norm": 0.1165320873260498, + "learning_rate": 3.861003861003861e-06, + "loss": 0.2126, + "step": 1276 + }, + { + "epoch": 0.32139935820801613, + "grad_norm": 0.12448088079690933, + "learning_rate": 3.7065637065637064e-06, + "loss": 0.2792, + "step": 1277 + }, + { + "epoch": 0.3216510413389543, + "grad_norm": 0.11725261062383652, + "learning_rate": 3.552123552123552e-06, + "loss": 0.1817, + "step": 1278 + }, + { + "epoch": 0.32190272446989243, + "grad_norm": 0.13542520999908447, + "learning_rate": 3.397683397683398e-06, + "loss": 0.2303, + "step": 1279 + }, + { + "epoch": 0.32215440760083053, + "grad_norm": 0.12533719837665558, + "learning_rate": 3.2432432432432437e-06, + "loss": 0.2545, + "step": 1280 + }, + { + "epoch": 0.3224060907317687, + "grad_norm": 0.11446390300989151, + "learning_rate": 3.088803088803089e-06, + "loss": 0.2177, + "step": 1281 + }, + { + "epoch": 0.32265777386270683, + "grad_norm": 0.10495037585496902, + "learning_rate": 2.9343629343629344e-06, + "loss": 0.1658, + "step": 1282 + }, + { + "epoch": 0.322909456993645, + "grad_norm": 0.11402431130409241, + "learning_rate": 2.7799227799227797e-06, + "loss": 0.188, + "step": 1283 + }, + { + "epoch": 0.32316114012458313, + "grad_norm": 0.12307335436344147, + "learning_rate": 2.625482625482626e-06, + "loss": 0.2025, + "step": 1284 + }, + { + "epoch": 0.3234128232555213, + "grad_norm": 0.1155342161655426, + "learning_rate": 2.4710424710424712e-06, + "loss": 0.1907, + "step": 1285 + }, + { + "epoch": 0.32366450638645944, + "grad_norm": 0.11528854817152023, + "learning_rate": 2.3166023166023166e-06, + "loss": 0.1732, + "step": 1286 + }, + { + "epoch": 0.3239161895173976, + "grad_norm": 0.1581220179796219, + "learning_rate": 2.1621621621621623e-06, + "loss": 0.24, + "step": 1287 + }, + { + "epoch": 0.32416787264833574, + "grad_norm": 0.10157429426908493, + "learning_rate": 2.0077220077220077e-06, + "loss": 0.161, + "step": 1288 + }, + { + "epoch": 0.3244195557792739, + "grad_norm": 0.13096509873867035, + "learning_rate": 1.8532818532818532e-06, + "loss": 0.2106, + "step": 1289 + }, + { + "epoch": 0.32467123891021205, + "grad_norm": 0.12065386772155762, + "learning_rate": 1.698841698841699e-06, + "loss": 0.2677, + "step": 1290 + }, + { + "epoch": 0.3249229220411502, + "grad_norm": 0.12294468283653259, + "learning_rate": 1.5444015444015445e-06, + "loss": 0.2426, + "step": 1291 + }, + { + "epoch": 0.32517460517208835, + "grad_norm": 0.11715657263994217, + "learning_rate": 1.3899613899613899e-06, + "loss": 0.1966, + "step": 1292 + }, + { + "epoch": 0.3254262883030265, + "grad_norm": 0.12442471832036972, + "learning_rate": 1.2355212355212356e-06, + "loss": 0.1863, + "step": 1293 + }, + { + "epoch": 0.32567797143396465, + "grad_norm": 0.12084191292524338, + "learning_rate": 1.0810810810810812e-06, + "loss": 0.2246, + "step": 1294 + }, + { + "epoch": 0.3259296545649028, + "grad_norm": 0.12736566364765167, + "learning_rate": 9.266409266409266e-07, + "loss": 0.2233, + "step": 1295 + }, + { + "epoch": 0.32618133769584096, + "grad_norm": 0.11683889478445053, + "learning_rate": 7.722007722007723e-07, + "loss": 0.2107, + "step": 1296 + }, + { + "epoch": 0.3264330208267791, + "grad_norm": 0.12388931214809418, + "learning_rate": 6.177606177606178e-07, + "loss": 0.3402, + "step": 1297 + }, + { + "epoch": 0.32668470395771726, + "grad_norm": 0.10320279002189636, + "learning_rate": 4.633204633204633e-07, + "loss": 0.1598, + "step": 1298 + }, + { + "epoch": 0.32693638708865536, + "grad_norm": 0.12689346075057983, + "learning_rate": 3.088803088803089e-07, + "loss": 0.1905, + "step": 1299 + }, + { + "epoch": 0.3271880702195935, + "grad_norm": 0.13141871988773346, + "learning_rate": 1.5444015444015445e-07, + "loss": 0.1684, + "step": 1300 + }, + { + "epoch": 0.32743975335053166, + "grad_norm": 0.11880113929510117, + "learning_rate": 0.0, + "loss": 0.216, + "step": 1301 + }, + { + "epoch": 0.3276914364814698, + "grad_norm": 0.10346447676420212, + "learning_rate": 5.559888579387187e-05, + "loss": 0.1501, + "step": 1302 + }, + { + "epoch": 0.32794311961240796, + "grad_norm": 0.1102476716041565, + "learning_rate": 5.5487465181058494e-05, + "loss": 0.1779, + "step": 1303 + }, + { + "epoch": 0.3281948027433461, + "grad_norm": 0.11776607483625412, + "learning_rate": 5.537604456824513e-05, + "loss": 0.2563, + "step": 1304 + }, + { + "epoch": 0.32844648587428427, + "grad_norm": 0.1242268979549408, + "learning_rate": 5.526462395543176e-05, + "loss": 0.3328, + "step": 1305 + }, + { + "epoch": 0.3286981690052224, + "grad_norm": 0.11003847420215607, + "learning_rate": 5.5153203342618384e-05, + "loss": 0.1639, + "step": 1306 + }, + { + "epoch": 0.32894985213616057, + "grad_norm": 0.12507784366607666, + "learning_rate": 5.504178272980501e-05, + "loss": 0.2304, + "step": 1307 + }, + { + "epoch": 0.3292015352670987, + "grad_norm": 0.1260419338941574, + "learning_rate": 5.493036211699165e-05, + "loss": 0.1681, + "step": 1308 + }, + { + "epoch": 0.3294532183980369, + "grad_norm": 0.11337315291166306, + "learning_rate": 5.481894150417828e-05, + "loss": 0.2015, + "step": 1309 + }, + { + "epoch": 0.329704901528975, + "grad_norm": 0.13672634959220886, + "learning_rate": 5.47075208913649e-05, + "loss": 0.284, + "step": 1310 + }, + { + "epoch": 0.3299565846599132, + "grad_norm": 0.09925438463687897, + "learning_rate": 5.459610027855153e-05, + "loss": 0.154, + "step": 1311 + }, + { + "epoch": 0.33020826779085133, + "grad_norm": 0.12440837919712067, + "learning_rate": 5.448467966573816e-05, + "loss": 0.2726, + "step": 1312 + }, + { + "epoch": 0.3304599509217895, + "grad_norm": 0.12719741463661194, + "learning_rate": 5.43732590529248e-05, + "loss": 0.2737, + "step": 1313 + }, + { + "epoch": 0.33071163405272763, + "grad_norm": 0.11711200326681137, + "learning_rate": 5.426183844011142e-05, + "loss": 0.1807, + "step": 1314 + }, + { + "epoch": 0.3309633171836658, + "grad_norm": 0.15956875681877136, + "learning_rate": 5.415041782729805e-05, + "loss": 0.2553, + "step": 1315 + }, + { + "epoch": 0.33121500031460394, + "grad_norm": 0.10041430592536926, + "learning_rate": 5.4038997214484674e-05, + "loss": 0.1577, + "step": 1316 + }, + { + "epoch": 0.3314666834455421, + "grad_norm": 0.13205572962760925, + "learning_rate": 5.392757660167132e-05, + "loss": 0.2103, + "step": 1317 + }, + { + "epoch": 0.33171836657648024, + "grad_norm": 0.12297097593545914, + "learning_rate": 5.381615598885794e-05, + "loss": 0.2447, + "step": 1318 + }, + { + "epoch": 0.33197004970741834, + "grad_norm": 0.1405712366104126, + "learning_rate": 5.370473537604457e-05, + "loss": 0.2568, + "step": 1319 + }, + { + "epoch": 0.3322217328383565, + "grad_norm": 0.10418865084648132, + "learning_rate": 5.359331476323119e-05, + "loss": 0.1745, + "step": 1320 + }, + { + "epoch": 0.33247341596929464, + "grad_norm": 0.14331026375293732, + "learning_rate": 5.3481894150417836e-05, + "loss": 0.2055, + "step": 1321 + }, + { + "epoch": 0.3327250991002328, + "grad_norm": 0.10940337926149368, + "learning_rate": 5.337047353760446e-05, + "loss": 0.188, + "step": 1322 + }, + { + "epoch": 0.33297678223117094, + "grad_norm": 0.10605291277170181, + "learning_rate": 5.325905292479109e-05, + "loss": 0.1916, + "step": 1323 + }, + { + "epoch": 0.3332284653621091, + "grad_norm": 0.12298569083213806, + "learning_rate": 5.314763231197771e-05, + "loss": 0.2112, + "step": 1324 + }, + { + "epoch": 0.33348014849304725, + "grad_norm": 0.11341530084609985, + "learning_rate": 5.303621169916435e-05, + "loss": 0.1721, + "step": 1325 + }, + { + "epoch": 0.3337318316239854, + "grad_norm": 0.13144564628601074, + "learning_rate": 5.292479108635098e-05, + "loss": 0.2622, + "step": 1326 + }, + { + "epoch": 0.33398351475492355, + "grad_norm": 0.11916239559650421, + "learning_rate": 5.281337047353761e-05, + "loss": 0.1799, + "step": 1327 + }, + { + "epoch": 0.3342351978858617, + "grad_norm": 0.11598220467567444, + "learning_rate": 5.270194986072423e-05, + "loss": 0.1791, + "step": 1328 + }, + { + "epoch": 0.33448688101679985, + "grad_norm": 0.13085304200649261, + "learning_rate": 5.259052924791087e-05, + "loss": 0.2407, + "step": 1329 + }, + { + "epoch": 0.334738564147738, + "grad_norm": 0.13301348686218262, + "learning_rate": 5.2479108635097497e-05, + "loss": 0.2174, + "step": 1330 + }, + { + "epoch": 0.33499024727867616, + "grad_norm": 0.12722571194171906, + "learning_rate": 5.236768802228412e-05, + "loss": 0.2434, + "step": 1331 + }, + { + "epoch": 0.3352419304096143, + "grad_norm": 0.1095128282904625, + "learning_rate": 5.225626740947075e-05, + "loss": 0.1942, + "step": 1332 + }, + { + "epoch": 0.33549361354055246, + "grad_norm": 0.12538395822048187, + "learning_rate": 5.2144846796657386e-05, + "loss": 0.2522, + "step": 1333 + }, + { + "epoch": 0.3357452966714906, + "grad_norm": 0.12161744385957718, + "learning_rate": 5.2033426183844015e-05, + "loss": 0.2887, + "step": 1334 + }, + { + "epoch": 0.33599697980242876, + "grad_norm": 0.11880531162023544, + "learning_rate": 5.192200557103064e-05, + "loss": 0.2683, + "step": 1335 + }, + { + "epoch": 0.3362486629333669, + "grad_norm": 0.1349031776189804, + "learning_rate": 5.181058495821727e-05, + "loss": 0.1887, + "step": 1336 + }, + { + "epoch": 0.33650034606430507, + "grad_norm": 0.12587116658687592, + "learning_rate": 5.1699164345403904e-05, + "loss": 0.2782, + "step": 1337 + }, + { + "epoch": 0.33675202919524316, + "grad_norm": 0.11964056640863419, + "learning_rate": 5.1587743732590534e-05, + "loss": 0.1812, + "step": 1338 + }, + { + "epoch": 0.3370037123261813, + "grad_norm": 0.11750610917806625, + "learning_rate": 5.147632311977716e-05, + "loss": 0.2187, + "step": 1339 + }, + { + "epoch": 0.33725539545711947, + "grad_norm": 0.11665648221969604, + "learning_rate": 5.136490250696379e-05, + "loss": 0.1962, + "step": 1340 + }, + { + "epoch": 0.3375070785880576, + "grad_norm": 0.12791278958320618, + "learning_rate": 5.125348189415042e-05, + "loss": 0.2119, + "step": 1341 + }, + { + "epoch": 0.33775876171899577, + "grad_norm": 0.12369109690189362, + "learning_rate": 5.114206128133705e-05, + "loss": 0.2827, + "step": 1342 + }, + { + "epoch": 0.3380104448499339, + "grad_norm": 0.11490099132061005, + "learning_rate": 5.1030640668523676e-05, + "loss": 0.1918, + "step": 1343 + }, + { + "epoch": 0.3382621279808721, + "grad_norm": 0.10008670389652252, + "learning_rate": 5.0919220055710306e-05, + "loss": 0.1558, + "step": 1344 + }, + { + "epoch": 0.3385138111118102, + "grad_norm": 0.12538427114486694, + "learning_rate": 5.080779944289694e-05, + "loss": 0.1908, + "step": 1345 + }, + { + "epoch": 0.3387654942427484, + "grad_norm": 0.12679807841777802, + "learning_rate": 5.069637883008357e-05, + "loss": 0.183, + "step": 1346 + }, + { + "epoch": 0.33901717737368653, + "grad_norm": 0.09309987723827362, + "learning_rate": 5.0584958217270195e-05, + "loss": 0.1402, + "step": 1347 + }, + { + "epoch": 0.3392688605046247, + "grad_norm": 0.12294266372919083, + "learning_rate": 5.0473537604456824e-05, + "loss": 0.255, + "step": 1348 + }, + { + "epoch": 0.33952054363556283, + "grad_norm": 0.12824276089668274, + "learning_rate": 5.036211699164346e-05, + "loss": 0.155, + "step": 1349 + }, + { + "epoch": 0.339772226766501, + "grad_norm": 0.15534637868404388, + "learning_rate": 5.0250696378830084e-05, + "loss": 0.1949, + "step": 1350 + }, + { + "epoch": 0.34002390989743914, + "grad_norm": 0.12093646824359894, + "learning_rate": 5.0139275766016714e-05, + "loss": 0.2405, + "step": 1351 + }, + { + "epoch": 0.3402755930283773, + "grad_norm": 0.11619894951581955, + "learning_rate": 5.002785515320334e-05, + "loss": 0.1921, + "step": 1352 + }, + { + "epoch": 0.34052727615931544, + "grad_norm": 0.10668577253818512, + "learning_rate": 4.991643454038997e-05, + "loss": 0.1564, + "step": 1353 + }, + { + "epoch": 0.3407789592902536, + "grad_norm": 0.14439751207828522, + "learning_rate": 4.98050139275766e-05, + "loss": 0.2823, + "step": 1354 + }, + { + "epoch": 0.34103064242119174, + "grad_norm": 0.1141776368021965, + "learning_rate": 4.969359331476323e-05, + "loss": 0.2025, + "step": 1355 + }, + { + "epoch": 0.3412823255521299, + "grad_norm": 0.11166267096996307, + "learning_rate": 4.958217270194986e-05, + "loss": 0.2099, + "step": 1356 + }, + { + "epoch": 0.341534008683068, + "grad_norm": 0.1155567541718483, + "learning_rate": 4.947075208913649e-05, + "loss": 0.2155, + "step": 1357 + }, + { + "epoch": 0.34178569181400614, + "grad_norm": 0.12425469607114792, + "learning_rate": 4.935933147632312e-05, + "loss": 0.1995, + "step": 1358 + }, + { + "epoch": 0.3420373749449443, + "grad_norm": 0.1180310994386673, + "learning_rate": 4.924791086350975e-05, + "loss": 0.2052, + "step": 1359 + }, + { + "epoch": 0.34228905807588245, + "grad_norm": 0.12304294109344482, + "learning_rate": 4.913649025069638e-05, + "loss": 0.236, + "step": 1360 + }, + { + "epoch": 0.3425407412068206, + "grad_norm": 0.11097686737775803, + "learning_rate": 4.902506963788301e-05, + "loss": 0.2207, + "step": 1361 + }, + { + "epoch": 0.34279242433775875, + "grad_norm": 0.127012237906456, + "learning_rate": 4.891364902506964e-05, + "loss": 0.1691, + "step": 1362 + }, + { + "epoch": 0.3430441074686969, + "grad_norm": 0.13973818719387054, + "learning_rate": 4.880222841225627e-05, + "loss": 0.2201, + "step": 1363 + }, + { + "epoch": 0.34329579059963505, + "grad_norm": 0.14379915595054626, + "learning_rate": 4.86908077994429e-05, + "loss": 0.2793, + "step": 1364 + }, + { + "epoch": 0.3435474737305732, + "grad_norm": 0.12253068387508392, + "learning_rate": 4.857938718662953e-05, + "loss": 0.2179, + "step": 1365 + }, + { + "epoch": 0.34379915686151136, + "grad_norm": 0.12574009597301483, + "learning_rate": 4.846796657381616e-05, + "loss": 0.2732, + "step": 1366 + }, + { + "epoch": 0.3440508399924495, + "grad_norm": 0.13135932385921478, + "learning_rate": 4.835654596100279e-05, + "loss": 0.2407, + "step": 1367 + }, + { + "epoch": 0.34430252312338766, + "grad_norm": 0.13892509043216705, + "learning_rate": 4.824512534818942e-05, + "loss": 0.2472, + "step": 1368 + }, + { + "epoch": 0.3445542062543258, + "grad_norm": 0.11705227196216583, + "learning_rate": 4.813370473537605e-05, + "loss": 0.2153, + "step": 1369 + }, + { + "epoch": 0.34480588938526396, + "grad_norm": 0.1340564638376236, + "learning_rate": 4.802228412256268e-05, + "loss": 0.2504, + "step": 1370 + }, + { + "epoch": 0.3450575725162021, + "grad_norm": 0.12827801704406738, + "learning_rate": 4.79108635097493e-05, + "loss": 0.2376, + "step": 1371 + }, + { + "epoch": 0.34530925564714027, + "grad_norm": 0.11675853282213211, + "learning_rate": 4.779944289693594e-05, + "loss": 0.1723, + "step": 1372 + }, + { + "epoch": 0.3455609387780784, + "grad_norm": 0.11679848283529282, + "learning_rate": 4.768802228412256e-05, + "loss": 0.2632, + "step": 1373 + }, + { + "epoch": 0.34581262190901657, + "grad_norm": 0.11980115622282028, + "learning_rate": 4.75766016713092e-05, + "loss": 0.1969, + "step": 1374 + }, + { + "epoch": 0.3460643050399547, + "grad_norm": 0.12757889926433563, + "learning_rate": 4.746518105849582e-05, + "loss": 0.2792, + "step": 1375 + }, + { + "epoch": 0.3463159881708928, + "grad_norm": 0.14682187139987946, + "learning_rate": 4.7353760445682456e-05, + "loss": 0.2906, + "step": 1376 + }, + { + "epoch": 0.34656767130183097, + "grad_norm": 0.1247834637761116, + "learning_rate": 4.724233983286908e-05, + "loss": 0.233, + "step": 1377 + }, + { + "epoch": 0.3468193544327691, + "grad_norm": 0.11924496293067932, + "learning_rate": 4.7130919220055716e-05, + "loss": 0.1974, + "step": 1378 + }, + { + "epoch": 0.3470710375637073, + "grad_norm": 0.11510487645864487, + "learning_rate": 4.701949860724234e-05, + "loss": 0.2348, + "step": 1379 + }, + { + "epoch": 0.3473227206946454, + "grad_norm": 0.10272910445928574, + "learning_rate": 4.6908077994428975e-05, + "loss": 0.1599, + "step": 1380 + }, + { + "epoch": 0.3475744038255836, + "grad_norm": 0.13208091259002686, + "learning_rate": 4.67966573816156e-05, + "loss": 0.3091, + "step": 1381 + }, + { + "epoch": 0.34782608695652173, + "grad_norm": 0.12919732928276062, + "learning_rate": 4.6685236768802234e-05, + "loss": 0.2833, + "step": 1382 + }, + { + "epoch": 0.3480777700874599, + "grad_norm": 0.10964875668287277, + "learning_rate": 4.657381615598886e-05, + "loss": 0.1903, + "step": 1383 + }, + { + "epoch": 0.34832945321839803, + "grad_norm": 0.13498936593532562, + "learning_rate": 4.6462395543175494e-05, + "loss": 0.213, + "step": 1384 + }, + { + "epoch": 0.3485811363493362, + "grad_norm": 0.17722249031066895, + "learning_rate": 4.635097493036212e-05, + "loss": 0.3333, + "step": 1385 + }, + { + "epoch": 0.34883281948027434, + "grad_norm": 0.13401447236537933, + "learning_rate": 4.623955431754875e-05, + "loss": 0.1633, + "step": 1386 + }, + { + "epoch": 0.3490845026112125, + "grad_norm": 0.13066500425338745, + "learning_rate": 4.6128133704735376e-05, + "loss": 0.2958, + "step": 1387 + }, + { + "epoch": 0.34933618574215064, + "grad_norm": 0.1369560956954956, + "learning_rate": 4.6016713091922006e-05, + "loss": 0.1884, + "step": 1388 + }, + { + "epoch": 0.3495878688730888, + "grad_norm": 0.12027233093976974, + "learning_rate": 4.5905292479108636e-05, + "loss": 0.2813, + "step": 1389 + }, + { + "epoch": 0.34983955200402694, + "grad_norm": 0.11347846686840057, + "learning_rate": 4.5793871866295265e-05, + "loss": 0.1701, + "step": 1390 + }, + { + "epoch": 0.3500912351349651, + "grad_norm": 0.10369741171598434, + "learning_rate": 4.5682451253481895e-05, + "loss": 0.1636, + "step": 1391 + }, + { + "epoch": 0.35034291826590325, + "grad_norm": 0.10896562039852142, + "learning_rate": 4.5571030640668525e-05, + "loss": 0.1644, + "step": 1392 + }, + { + "epoch": 0.3505946013968414, + "grad_norm": 0.104862280189991, + "learning_rate": 4.5459610027855154e-05, + "loss": 0.1315, + "step": 1393 + }, + { + "epoch": 0.35084628452777955, + "grad_norm": 0.10331819951534271, + "learning_rate": 4.5348189415041784e-05, + "loss": 0.1648, + "step": 1394 + }, + { + "epoch": 0.35109796765871765, + "grad_norm": 0.11424372345209122, + "learning_rate": 4.5236768802228414e-05, + "loss": 0.1963, + "step": 1395 + }, + { + "epoch": 0.3513496507896558, + "grad_norm": 0.11450093239545822, + "learning_rate": 4.5125348189415044e-05, + "loss": 0.1952, + "step": 1396 + }, + { + "epoch": 0.35160133392059395, + "grad_norm": 0.11348758637905121, + "learning_rate": 4.501392757660167e-05, + "loss": 0.173, + "step": 1397 + }, + { + "epoch": 0.3518530170515321, + "grad_norm": 0.127132385969162, + "learning_rate": 4.49025069637883e-05, + "loss": 0.1851, + "step": 1398 + }, + { + "epoch": 0.35210470018247025, + "grad_norm": 0.11630198359489441, + "learning_rate": 4.479108635097493e-05, + "loss": 0.2175, + "step": 1399 + }, + { + "epoch": 0.3523563833134084, + "grad_norm": 0.12995584309101105, + "learning_rate": 4.467966573816156e-05, + "loss": 0.2824, + "step": 1400 + }, + { + "epoch": 0.35260806644434656, + "grad_norm": 0.11960501968860626, + "learning_rate": 4.456824512534819e-05, + "loss": 0.2951, + "step": 1401 + }, + { + "epoch": 0.3528597495752847, + "grad_norm": 0.11468775570392609, + "learning_rate": 4.445682451253482e-05, + "loss": 0.1626, + "step": 1402 + }, + { + "epoch": 0.35311143270622286, + "grad_norm": 0.12279869616031647, + "learning_rate": 4.434540389972145e-05, + "loss": 0.2633, + "step": 1403 + }, + { + "epoch": 0.353363115837161, + "grad_norm": 0.12615051865577698, + "learning_rate": 4.423398328690808e-05, + "loss": 0.2701, + "step": 1404 + }, + { + "epoch": 0.35361479896809916, + "grad_norm": 0.11709825694561005, + "learning_rate": 4.412256267409471e-05, + "loss": 0.1709, + "step": 1405 + }, + { + "epoch": 0.3538664820990373, + "grad_norm": 0.1304224729537964, + "learning_rate": 4.401114206128134e-05, + "loss": 0.2356, + "step": 1406 + }, + { + "epoch": 0.35411816522997547, + "grad_norm": 0.12506772577762604, + "learning_rate": 4.389972144846797e-05, + "loss": 0.2331, + "step": 1407 + }, + { + "epoch": 0.3543698483609136, + "grad_norm": 0.11886215209960938, + "learning_rate": 4.37883008356546e-05, + "loss": 0.1865, + "step": 1408 + }, + { + "epoch": 0.35462153149185177, + "grad_norm": 0.13724276423454285, + "learning_rate": 4.367688022284123e-05, + "loss": 0.2272, + "step": 1409 + }, + { + "epoch": 0.3548732146227899, + "grad_norm": 0.13399899005889893, + "learning_rate": 4.356545961002786e-05, + "loss": 0.344, + "step": 1410 + }, + { + "epoch": 0.3551248977537281, + "grad_norm": 0.14068353176116943, + "learning_rate": 4.345403899721448e-05, + "loss": 0.2251, + "step": 1411 + }, + { + "epoch": 0.3553765808846662, + "grad_norm": 0.11466100066900253, + "learning_rate": 4.334261838440112e-05, + "loss": 0.1888, + "step": 1412 + }, + { + "epoch": 0.3556282640156044, + "grad_norm": 0.1321353316307068, + "learning_rate": 4.323119777158774e-05, + "loss": 0.2291, + "step": 1413 + }, + { + "epoch": 0.35587994714654253, + "grad_norm": 0.1290498822927475, + "learning_rate": 4.311977715877438e-05, + "loss": 0.2677, + "step": 1414 + }, + { + "epoch": 0.3561316302774806, + "grad_norm": 0.1221776008605957, + "learning_rate": 4.3008356545961e-05, + "loss": 0.185, + "step": 1415 + }, + { + "epoch": 0.3563833134084188, + "grad_norm": 0.11119630187749863, + "learning_rate": 4.289693593314764e-05, + "loss": 0.242, + "step": 1416 + }, + { + "epoch": 0.35663499653935693, + "grad_norm": 0.1062801405787468, + "learning_rate": 4.278551532033426e-05, + "loss": 0.1638, + "step": 1417 + }, + { + "epoch": 0.3568866796702951, + "grad_norm": 0.11139536648988724, + "learning_rate": 4.26740947075209e-05, + "loss": 0.18, + "step": 1418 + }, + { + "epoch": 0.35713836280123323, + "grad_norm": 0.10590539127588272, + "learning_rate": 4.256267409470752e-05, + "loss": 0.1469, + "step": 1419 + }, + { + "epoch": 0.3573900459321714, + "grad_norm": 0.11620762944221497, + "learning_rate": 4.2451253481894157e-05, + "loss": 0.2087, + "step": 1420 + }, + { + "epoch": 0.35764172906310954, + "grad_norm": 0.10536862909793854, + "learning_rate": 4.233983286908078e-05, + "loss": 0.1684, + "step": 1421 + }, + { + "epoch": 0.3578934121940477, + "grad_norm": 0.11078710108995438, + "learning_rate": 4.2228412256267416e-05, + "loss": 0.2037, + "step": 1422 + }, + { + "epoch": 0.35814509532498584, + "grad_norm": 0.12910079956054688, + "learning_rate": 4.211699164345404e-05, + "loss": 0.2287, + "step": 1423 + }, + { + "epoch": 0.358396778455924, + "grad_norm": 0.10344430059194565, + "learning_rate": 4.2005571030640675e-05, + "loss": 0.1529, + "step": 1424 + }, + { + "epoch": 0.35864846158686214, + "grad_norm": 0.11895854026079178, + "learning_rate": 4.18941504178273e-05, + "loss": 0.1671, + "step": 1425 + }, + { + "epoch": 0.3589001447178003, + "grad_norm": 0.12069137394428253, + "learning_rate": 4.1782729805013935e-05, + "loss": 0.2049, + "step": 1426 + }, + { + "epoch": 0.35915182784873845, + "grad_norm": 0.13283878564834595, + "learning_rate": 4.167130919220056e-05, + "loss": 0.2004, + "step": 1427 + }, + { + "epoch": 0.3594035109796766, + "grad_norm": 0.10277939587831497, + "learning_rate": 4.155988857938719e-05, + "loss": 0.1425, + "step": 1428 + }, + { + "epoch": 0.35965519411061475, + "grad_norm": 0.13160032033920288, + "learning_rate": 4.144846796657382e-05, + "loss": 0.2482, + "step": 1429 + }, + { + "epoch": 0.3599068772415529, + "grad_norm": 0.15410944819450378, + "learning_rate": 4.133704735376045e-05, + "loss": 0.2453, + "step": 1430 + }, + { + "epoch": 0.36015856037249105, + "grad_norm": 0.11637213081121445, + "learning_rate": 4.1225626740947077e-05, + "loss": 0.1959, + "step": 1431 + }, + { + "epoch": 0.3604102435034292, + "grad_norm": 0.12155333161354065, + "learning_rate": 4.1114206128133706e-05, + "loss": 0.2081, + "step": 1432 + }, + { + "epoch": 0.36066192663436736, + "grad_norm": 0.11685843020677567, + "learning_rate": 4.1002785515320336e-05, + "loss": 0.1898, + "step": 1433 + }, + { + "epoch": 0.36091360976530545, + "grad_norm": 0.12555062770843506, + "learning_rate": 4.0891364902506966e-05, + "loss": 0.309, + "step": 1434 + }, + { + "epoch": 0.3611652928962436, + "grad_norm": 0.1283169835805893, + "learning_rate": 4.0779944289693595e-05, + "loss": 0.2227, + "step": 1435 + }, + { + "epoch": 0.36141697602718176, + "grad_norm": 0.1100114956498146, + "learning_rate": 4.0668523676880225e-05, + "loss": 0.1964, + "step": 1436 + }, + { + "epoch": 0.3616686591581199, + "grad_norm": 0.12097296863794327, + "learning_rate": 4.0557103064066855e-05, + "loss": 0.1889, + "step": 1437 + }, + { + "epoch": 0.36192034228905806, + "grad_norm": 0.12069032341241837, + "learning_rate": 4.0445682451253484e-05, + "loss": 0.2083, + "step": 1438 + }, + { + "epoch": 0.3621720254199962, + "grad_norm": 0.11620620638132095, + "learning_rate": 4.0334261838440114e-05, + "loss": 0.2844, + "step": 1439 + }, + { + "epoch": 0.36242370855093436, + "grad_norm": 0.1290942132472992, + "learning_rate": 4.0222841225626744e-05, + "loss": 0.1933, + "step": 1440 + }, + { + "epoch": 0.3626753916818725, + "grad_norm": 0.12494636327028275, + "learning_rate": 4.0111420612813374e-05, + "loss": 0.2239, + "step": 1441 + }, + { + "epoch": 0.36292707481281067, + "grad_norm": 0.12561176717281342, + "learning_rate": 4e-05, + "loss": 0.2554, + "step": 1442 + }, + { + "epoch": 0.3631787579437488, + "grad_norm": 0.1276828795671463, + "learning_rate": 3.988857938718663e-05, + "loss": 0.1947, + "step": 1443 + }, + { + "epoch": 0.36343044107468697, + "grad_norm": 0.12439251691102982, + "learning_rate": 3.977715877437326e-05, + "loss": 0.1855, + "step": 1444 + }, + { + "epoch": 0.3636821242056251, + "grad_norm": 0.11355730146169662, + "learning_rate": 3.966573816155989e-05, + "loss": 0.1825, + "step": 1445 + }, + { + "epoch": 0.3639338073365633, + "grad_norm": 0.10460755974054337, + "learning_rate": 3.955431754874652e-05, + "loss": 0.1616, + "step": 1446 + }, + { + "epoch": 0.3641854904675014, + "grad_norm": 0.12744562327861786, + "learning_rate": 3.944289693593315e-05, + "loss": 0.1907, + "step": 1447 + }, + { + "epoch": 0.3644371735984396, + "grad_norm": 0.11712396144866943, + "learning_rate": 3.933147632311978e-05, + "loss": 0.1978, + "step": 1448 + }, + { + "epoch": 0.36468885672937773, + "grad_norm": 0.11206866055727005, + "learning_rate": 3.922005571030641e-05, + "loss": 0.1831, + "step": 1449 + }, + { + "epoch": 0.3649405398603159, + "grad_norm": 0.12751565873622894, + "learning_rate": 3.910863509749304e-05, + "loss": 0.251, + "step": 1450 + }, + { + "epoch": 0.36519222299125403, + "grad_norm": 0.11813338100910187, + "learning_rate": 3.8997214484679664e-05, + "loss": 0.2015, + "step": 1451 + }, + { + "epoch": 0.3654439061221922, + "grad_norm": 0.1097281202673912, + "learning_rate": 3.88857938718663e-05, + "loss": 0.2004, + "step": 1452 + }, + { + "epoch": 0.3656955892531303, + "grad_norm": 0.12413442879915237, + "learning_rate": 3.877437325905292e-05, + "loss": 0.2844, + "step": 1453 + }, + { + "epoch": 0.36594727238406843, + "grad_norm": 0.12721839547157288, + "learning_rate": 3.866295264623956e-05, + "loss": 0.2568, + "step": 1454 + }, + { + "epoch": 0.3661989555150066, + "grad_norm": 0.10556313395500183, + "learning_rate": 3.855153203342618e-05, + "loss": 0.1621, + "step": 1455 + }, + { + "epoch": 0.36645063864594474, + "grad_norm": 0.10769017785787582, + "learning_rate": 3.844011142061282e-05, + "loss": 0.1919, + "step": 1456 + }, + { + "epoch": 0.3667023217768829, + "grad_norm": 0.12968257069587708, + "learning_rate": 3.832869080779944e-05, + "loss": 0.2739, + "step": 1457 + }, + { + "epoch": 0.36695400490782104, + "grad_norm": 0.14019334316253662, + "learning_rate": 3.821727019498608e-05, + "loss": 0.1953, + "step": 1458 + }, + { + "epoch": 0.3672056880387592, + "grad_norm": 0.13034367561340332, + "learning_rate": 3.81058495821727e-05, + "loss": 0.2945, + "step": 1459 + }, + { + "epoch": 0.36745737116969734, + "grad_norm": 0.11869984865188599, + "learning_rate": 3.799442896935934e-05, + "loss": 0.2077, + "step": 1460 + }, + { + "epoch": 0.3677090543006355, + "grad_norm": 0.1253458708524704, + "learning_rate": 3.788300835654596e-05, + "loss": 0.2249, + "step": 1461 + }, + { + "epoch": 0.36796073743157365, + "grad_norm": 0.11638636142015457, + "learning_rate": 3.77715877437326e-05, + "loss": 0.1667, + "step": 1462 + }, + { + "epoch": 0.3682124205625118, + "grad_norm": 0.11331362277269363, + "learning_rate": 3.766016713091922e-05, + "loss": 0.2732, + "step": 1463 + }, + { + "epoch": 0.36846410369344995, + "grad_norm": 0.11691166460514069, + "learning_rate": 3.754874651810586e-05, + "loss": 0.1812, + "step": 1464 + }, + { + "epoch": 0.3687157868243881, + "grad_norm": 0.13369520008563995, + "learning_rate": 3.743732590529248e-05, + "loss": 0.2803, + "step": 1465 + }, + { + "epoch": 0.36896746995532625, + "grad_norm": 0.10118907690048218, + "learning_rate": 3.7325905292479116e-05, + "loss": 0.1638, + "step": 1466 + }, + { + "epoch": 0.3692191530862644, + "grad_norm": 0.11917991191148758, + "learning_rate": 3.721448467966574e-05, + "loss": 0.1827, + "step": 1467 + }, + { + "epoch": 0.36947083621720256, + "grad_norm": 0.1294524371623993, + "learning_rate": 3.710306406685237e-05, + "loss": 0.2582, + "step": 1468 + }, + { + "epoch": 0.3697225193481407, + "grad_norm": 0.11105356365442276, + "learning_rate": 3.6991643454039e-05, + "loss": 0.1772, + "step": 1469 + }, + { + "epoch": 0.36997420247907886, + "grad_norm": 0.12181617319583893, + "learning_rate": 3.688022284122563e-05, + "loss": 0.2123, + "step": 1470 + }, + { + "epoch": 0.370225885610017, + "grad_norm": 0.12193644791841507, + "learning_rate": 3.676880222841226e-05, + "loss": 0.2018, + "step": 1471 + }, + { + "epoch": 0.3704775687409551, + "grad_norm": 0.11559071391820908, + "learning_rate": 3.665738161559889e-05, + "loss": 0.1557, + "step": 1472 + }, + { + "epoch": 0.37072925187189326, + "grad_norm": 0.1364925354719162, + "learning_rate": 3.654596100278552e-05, + "loss": 0.33, + "step": 1473 + }, + { + "epoch": 0.3709809350028314, + "grad_norm": 0.12118813395500183, + "learning_rate": 3.643454038997215e-05, + "loss": 0.1872, + "step": 1474 + }, + { + "epoch": 0.37123261813376957, + "grad_norm": 0.126297265291214, + "learning_rate": 3.632311977715878e-05, + "loss": 0.1948, + "step": 1475 + }, + { + "epoch": 0.3714843012647077, + "grad_norm": 0.10534999519586563, + "learning_rate": 3.62116991643454e-05, + "loss": 0.1691, + "step": 1476 + }, + { + "epoch": 0.37173598439564587, + "grad_norm": 0.13739937543869019, + "learning_rate": 3.6100278551532036e-05, + "loss": 0.2049, + "step": 1477 + }, + { + "epoch": 0.371987667526584, + "grad_norm": 0.113822802901268, + "learning_rate": 3.598885793871866e-05, + "loss": 0.1759, + "step": 1478 + }, + { + "epoch": 0.37223935065752217, + "grad_norm": 0.1225641593337059, + "learning_rate": 3.5877437325905296e-05, + "loss": 0.2003, + "step": 1479 + }, + { + "epoch": 0.3724910337884603, + "grad_norm": 0.10749515146017075, + "learning_rate": 3.576601671309192e-05, + "loss": 0.1339, + "step": 1480 + }, + { + "epoch": 0.3727427169193985, + "grad_norm": 0.10851122438907623, + "learning_rate": 3.5654596100278555e-05, + "loss": 0.1855, + "step": 1481 + }, + { + "epoch": 0.3729944000503366, + "grad_norm": 0.11161625385284424, + "learning_rate": 3.554317548746518e-05, + "loss": 0.1429, + "step": 1482 + }, + { + "epoch": 0.3732460831812748, + "grad_norm": 0.11973297595977783, + "learning_rate": 3.5431754874651814e-05, + "loss": 0.2336, + "step": 1483 + }, + { + "epoch": 0.37349776631221293, + "grad_norm": 0.12315451353788376, + "learning_rate": 3.532033426183844e-05, + "loss": 0.2422, + "step": 1484 + }, + { + "epoch": 0.3737494494431511, + "grad_norm": 0.10754470527172089, + "learning_rate": 3.5208913649025074e-05, + "loss": 0.151, + "step": 1485 + }, + { + "epoch": 0.37400113257408923, + "grad_norm": 0.13255761563777924, + "learning_rate": 3.50974930362117e-05, + "loss": 0.1996, + "step": 1486 + }, + { + "epoch": 0.3742528157050274, + "grad_norm": 0.10981941968202591, + "learning_rate": 3.498607242339833e-05, + "loss": 0.1722, + "step": 1487 + }, + { + "epoch": 0.37450449883596554, + "grad_norm": 0.13403695821762085, + "learning_rate": 3.4874651810584956e-05, + "loss": 0.2729, + "step": 1488 + }, + { + "epoch": 0.3747561819669037, + "grad_norm": 0.11870770901441574, + "learning_rate": 3.476323119777159e-05, + "loss": 0.2197, + "step": 1489 + }, + { + "epoch": 0.37500786509784184, + "grad_norm": 0.11082787066698074, + "learning_rate": 3.4651810584958216e-05, + "loss": 0.1909, + "step": 1490 + }, + { + "epoch": 0.37525954822878, + "grad_norm": 0.11936727166175842, + "learning_rate": 3.454038997214485e-05, + "loss": 0.1981, + "step": 1491 + }, + { + "epoch": 0.3755112313597181, + "grad_norm": 0.14362333714962006, + "learning_rate": 3.4428969359331475e-05, + "loss": 0.2624, + "step": 1492 + }, + { + "epoch": 0.37576291449065624, + "grad_norm": 0.11489781737327576, + "learning_rate": 3.4317548746518105e-05, + "loss": 0.1903, + "step": 1493 + }, + { + "epoch": 0.3760145976215944, + "grad_norm": 0.12070348858833313, + "learning_rate": 3.4206128133704734e-05, + "loss": 0.2063, + "step": 1494 + }, + { + "epoch": 0.37626628075253254, + "grad_norm": 0.11841773986816406, + "learning_rate": 3.4094707520891364e-05, + "loss": 0.1706, + "step": 1495 + }, + { + "epoch": 0.3765179638834707, + "grad_norm": 0.12456177175045013, + "learning_rate": 3.3983286908077994e-05, + "loss": 0.2042, + "step": 1496 + }, + { + "epoch": 0.37676964701440885, + "grad_norm": 0.11188086867332458, + "learning_rate": 3.3871866295264624e-05, + "loss": 0.1761, + "step": 1497 + }, + { + "epoch": 0.377021330145347, + "grad_norm": 0.10862310230731964, + "learning_rate": 3.376044568245125e-05, + "loss": 0.1579, + "step": 1498 + }, + { + "epoch": 0.37727301327628515, + "grad_norm": 0.12973307073116302, + "learning_rate": 3.364902506963788e-05, + "loss": 0.2508, + "step": 1499 + }, + { + "epoch": 0.3775246964072233, + "grad_norm": 0.11587046086788177, + "learning_rate": 3.353760445682451e-05, + "loss": 0.1877, + "step": 1500 + }, + { + "epoch": 0.37777637953816146, + "grad_norm": 0.11175840348005295, + "learning_rate": 3.342618384401114e-05, + "loss": 0.1577, + "step": 1501 + }, + { + "epoch": 0.3780280626690996, + "grad_norm": 0.13211950659751892, + "learning_rate": 3.331476323119777e-05, + "loss": 0.2055, + "step": 1502 + }, + { + "epoch": 0.37827974580003776, + "grad_norm": 0.11200255900621414, + "learning_rate": 3.32033426183844e-05, + "loss": 0.215, + "step": 1503 + }, + { + "epoch": 0.3785314289309759, + "grad_norm": 0.10731122642755508, + "learning_rate": 3.309192200557103e-05, + "loss": 0.1695, + "step": 1504 + }, + { + "epoch": 0.37878311206191406, + "grad_norm": 0.11924634128808975, + "learning_rate": 3.298050139275766e-05, + "loss": 0.1519, + "step": 1505 + }, + { + "epoch": 0.3790347951928522, + "grad_norm": 0.13024650514125824, + "learning_rate": 3.286908077994429e-05, + "loss": 0.2819, + "step": 1506 + }, + { + "epoch": 0.37928647832379037, + "grad_norm": 0.1128431037068367, + "learning_rate": 3.275766016713092e-05, + "loss": 0.2107, + "step": 1507 + }, + { + "epoch": 0.3795381614547285, + "grad_norm": 0.12745632231235504, + "learning_rate": 3.264623955431755e-05, + "loss": 0.2524, + "step": 1508 + }, + { + "epoch": 0.37978984458566667, + "grad_norm": 0.12261222302913666, + "learning_rate": 3.253481894150418e-05, + "loss": 0.2305, + "step": 1509 + }, + { + "epoch": 0.3800415277166048, + "grad_norm": 0.1230972483754158, + "learning_rate": 3.242339832869081e-05, + "loss": 0.2855, + "step": 1510 + }, + { + "epoch": 0.3802932108475429, + "grad_norm": 0.13548460602760315, + "learning_rate": 3.231197771587744e-05, + "loss": 0.2272, + "step": 1511 + }, + { + "epoch": 0.38054489397848107, + "grad_norm": 0.13061048090457916, + "learning_rate": 3.220055710306407e-05, + "loss": 0.2323, + "step": 1512 + }, + { + "epoch": 0.3807965771094192, + "grad_norm": 0.1231827586889267, + "learning_rate": 3.20891364902507e-05, + "loss": 0.2487, + "step": 1513 + }, + { + "epoch": 0.3810482602403574, + "grad_norm": 0.11537005752325058, + "learning_rate": 3.197771587743733e-05, + "loss": 0.1908, + "step": 1514 + }, + { + "epoch": 0.3812999433712955, + "grad_norm": 0.10413341969251633, + "learning_rate": 3.186629526462396e-05, + "loss": 0.2058, + "step": 1515 + }, + { + "epoch": 0.3815516265022337, + "grad_norm": 0.1326233446598053, + "learning_rate": 3.175487465181058e-05, + "loss": 0.2827, + "step": 1516 + }, + { + "epoch": 0.38180330963317183, + "grad_norm": 0.10836413502693176, + "learning_rate": 3.164345403899722e-05, + "loss": 0.1647, + "step": 1517 + }, + { + "epoch": 0.38205499276411, + "grad_norm": 0.11082088947296143, + "learning_rate": 3.153203342618384e-05, + "loss": 0.1986, + "step": 1518 + }, + { + "epoch": 0.38230667589504813, + "grad_norm": 0.11223185807466507, + "learning_rate": 3.142061281337048e-05, + "loss": 0.2326, + "step": 1519 + }, + { + "epoch": 0.3825583590259863, + "grad_norm": 0.12838886678218842, + "learning_rate": 3.13091922005571e-05, + "loss": 0.2201, + "step": 1520 + }, + { + "epoch": 0.38281004215692443, + "grad_norm": 0.11324883252382278, + "learning_rate": 3.1197771587743737e-05, + "loss": 0.2085, + "step": 1521 + }, + { + "epoch": 0.3830617252878626, + "grad_norm": 0.12703056633472443, + "learning_rate": 3.108635097493036e-05, + "loss": 0.2177, + "step": 1522 + }, + { + "epoch": 0.38331340841880074, + "grad_norm": 0.1312105357646942, + "learning_rate": 3.0974930362116996e-05, + "loss": 0.2657, + "step": 1523 + }, + { + "epoch": 0.3835650915497389, + "grad_norm": 0.13176077604293823, + "learning_rate": 3.086350974930362e-05, + "loss": 0.3421, + "step": 1524 + }, + { + "epoch": 0.38381677468067704, + "grad_norm": 0.12279617041349411, + "learning_rate": 3.0752089136490255e-05, + "loss": 0.2471, + "step": 1525 + }, + { + "epoch": 0.3840684578116152, + "grad_norm": 0.13655629754066467, + "learning_rate": 3.064066852367688e-05, + "loss": 0.2753, + "step": 1526 + }, + { + "epoch": 0.38432014094255335, + "grad_norm": 0.12528571486473083, + "learning_rate": 3.0529247910863515e-05, + "loss": 0.2151, + "step": 1527 + }, + { + "epoch": 0.3845718240734915, + "grad_norm": 0.1173616424202919, + "learning_rate": 3.0417827298050138e-05, + "loss": 0.1875, + "step": 1528 + }, + { + "epoch": 0.38482350720442965, + "grad_norm": 0.12106706202030182, + "learning_rate": 3.030640668523677e-05, + "loss": 0.1819, + "step": 1529 + }, + { + "epoch": 0.38507519033536775, + "grad_norm": 0.10126858949661255, + "learning_rate": 3.0194986072423397e-05, + "loss": 0.1745, + "step": 1530 + }, + { + "epoch": 0.3853268734663059, + "grad_norm": 0.1013520210981369, + "learning_rate": 3.008356545961003e-05, + "loss": 0.1495, + "step": 1531 + }, + { + "epoch": 0.38557855659724405, + "grad_norm": 0.13077495992183685, + "learning_rate": 2.9972144846796656e-05, + "loss": 0.1837, + "step": 1532 + }, + { + "epoch": 0.3858302397281822, + "grad_norm": 0.10864868760108948, + "learning_rate": 2.986072423398329e-05, + "loss": 0.1504, + "step": 1533 + }, + { + "epoch": 0.38608192285912035, + "grad_norm": 0.1056288480758667, + "learning_rate": 2.9749303621169916e-05, + "loss": 0.1414, + "step": 1534 + }, + { + "epoch": 0.3863336059900585, + "grad_norm": 0.11963026225566864, + "learning_rate": 2.963788300835655e-05, + "loss": 0.1927, + "step": 1535 + }, + { + "epoch": 0.38658528912099666, + "grad_norm": 0.10899466276168823, + "learning_rate": 2.9526462395543175e-05, + "loss": 0.1659, + "step": 1536 + }, + { + "epoch": 0.3868369722519348, + "grad_norm": 0.1271812468767166, + "learning_rate": 2.941504178272981e-05, + "loss": 0.2352, + "step": 1537 + }, + { + "epoch": 0.38708865538287296, + "grad_norm": 0.11509156227111816, + "learning_rate": 2.9303621169916435e-05, + "loss": 0.1702, + "step": 1538 + }, + { + "epoch": 0.3873403385138111, + "grad_norm": 0.11564816534519196, + "learning_rate": 2.9192200557103068e-05, + "loss": 0.1746, + "step": 1539 + }, + { + "epoch": 0.38759202164474926, + "grad_norm": 0.11589440703392029, + "learning_rate": 2.9080779944289694e-05, + "loss": 0.1709, + "step": 1540 + }, + { + "epoch": 0.3878437047756874, + "grad_norm": 0.11901594698429108, + "learning_rate": 2.8969359331476327e-05, + "loss": 0.1902, + "step": 1541 + }, + { + "epoch": 0.38809538790662557, + "grad_norm": 0.11014410108327866, + "learning_rate": 2.8857938718662954e-05, + "loss": 0.1668, + "step": 1542 + }, + { + "epoch": 0.3883470710375637, + "grad_norm": 0.12025776505470276, + "learning_rate": 2.8746518105849583e-05, + "loss": 0.1731, + "step": 1543 + }, + { + "epoch": 0.38859875416850187, + "grad_norm": 0.1301485002040863, + "learning_rate": 2.8635097493036213e-05, + "loss": 0.187, + "step": 1544 + }, + { + "epoch": 0.38885043729944, + "grad_norm": 0.11576654762029648, + "learning_rate": 2.8523676880222843e-05, + "loss": 0.1581, + "step": 1545 + }, + { + "epoch": 0.3891021204303782, + "grad_norm": 0.12987595796585083, + "learning_rate": 2.841225626740947e-05, + "loss": 0.2368, + "step": 1546 + }, + { + "epoch": 0.3893538035613163, + "grad_norm": 0.11712390184402466, + "learning_rate": 2.8300835654596102e-05, + "loss": 0.1717, + "step": 1547 + }, + { + "epoch": 0.3896054866922545, + "grad_norm": 0.12855015695095062, + "learning_rate": 2.818941504178273e-05, + "loss": 0.1717, + "step": 1548 + }, + { + "epoch": 0.3898571698231926, + "grad_norm": 0.10797053575515747, + "learning_rate": 2.807799442896936e-05, + "loss": 0.1996, + "step": 1549 + }, + { + "epoch": 0.3901088529541307, + "grad_norm": 0.12115349620580673, + "learning_rate": 2.7966573816155988e-05, + "loss": 0.1883, + "step": 1550 + }, + { + "epoch": 0.3903605360850689, + "grad_norm": 0.11512169241905212, + "learning_rate": 2.785515320334262e-05, + "loss": 0.1636, + "step": 1551 + }, + { + "epoch": 0.39061221921600703, + "grad_norm": 0.13461914658546448, + "learning_rate": 2.7743732590529247e-05, + "loss": 0.2417, + "step": 1552 + }, + { + "epoch": 0.3908639023469452, + "grad_norm": 0.112267404794693, + "learning_rate": 2.763231197771588e-05, + "loss": 0.1623, + "step": 1553 + }, + { + "epoch": 0.39111558547788333, + "grad_norm": 0.12404221296310425, + "learning_rate": 2.7520891364902507e-05, + "loss": 0.1693, + "step": 1554 + }, + { + "epoch": 0.3913672686088215, + "grad_norm": 0.1335454285144806, + "learning_rate": 2.740947075208914e-05, + "loss": 0.3109, + "step": 1555 + }, + { + "epoch": 0.39161895173975964, + "grad_norm": 0.1222451850771904, + "learning_rate": 2.7298050139275766e-05, + "loss": 0.2032, + "step": 1556 + }, + { + "epoch": 0.3918706348706978, + "grad_norm": 0.11775821447372437, + "learning_rate": 2.71866295264624e-05, + "loss": 0.168, + "step": 1557 + }, + { + "epoch": 0.39212231800163594, + "grad_norm": 0.10678533464670181, + "learning_rate": 2.7075208913649025e-05, + "loss": 0.156, + "step": 1558 + }, + { + "epoch": 0.3923740011325741, + "grad_norm": 0.12982624769210815, + "learning_rate": 2.696378830083566e-05, + "loss": 0.2234, + "step": 1559 + }, + { + "epoch": 0.39262568426351224, + "grad_norm": 0.11723200976848602, + "learning_rate": 2.6852367688022285e-05, + "loss": 0.1958, + "step": 1560 + }, + { + "epoch": 0.3928773673944504, + "grad_norm": 0.13143743574619293, + "learning_rate": 2.6740947075208918e-05, + "loss": 0.2304, + "step": 1561 + }, + { + "epoch": 0.39312905052538855, + "grad_norm": 0.14005522429943085, + "learning_rate": 2.6629526462395544e-05, + "loss": 0.2262, + "step": 1562 + }, + { + "epoch": 0.3933807336563267, + "grad_norm": 0.11448992043733597, + "learning_rate": 2.6518105849582174e-05, + "loss": 0.1966, + "step": 1563 + }, + { + "epoch": 0.39363241678726485, + "grad_norm": 0.13646821677684784, + "learning_rate": 2.6406685236768804e-05, + "loss": 0.2333, + "step": 1564 + }, + { + "epoch": 0.393884099918203, + "grad_norm": 0.13760502636432648, + "learning_rate": 2.6295264623955433e-05, + "loss": 0.1682, + "step": 1565 + }, + { + "epoch": 0.39413578304914115, + "grad_norm": 0.13530333340168, + "learning_rate": 2.618384401114206e-05, + "loss": 0.2557, + "step": 1566 + }, + { + "epoch": 0.3943874661800793, + "grad_norm": 0.12617719173431396, + "learning_rate": 2.6072423398328693e-05, + "loss": 0.1992, + "step": 1567 + }, + { + "epoch": 0.39463914931101746, + "grad_norm": 0.1105303093791008, + "learning_rate": 2.596100278551532e-05, + "loss": 0.2121, + "step": 1568 + }, + { + "epoch": 0.39489083244195555, + "grad_norm": 0.12541228532791138, + "learning_rate": 2.5849582172701952e-05, + "loss": 0.1973, + "step": 1569 + }, + { + "epoch": 0.3951425155728937, + "grad_norm": 0.13525298237800598, + "learning_rate": 2.573816155988858e-05, + "loss": 0.2608, + "step": 1570 + }, + { + "epoch": 0.39539419870383186, + "grad_norm": 0.1305062472820282, + "learning_rate": 2.562674094707521e-05, + "loss": 0.2147, + "step": 1571 + }, + { + "epoch": 0.39564588183477, + "grad_norm": 0.1400519162416458, + "learning_rate": 2.5515320334261838e-05, + "loss": 0.232, + "step": 1572 + }, + { + "epoch": 0.39589756496570816, + "grad_norm": 0.12001420557498932, + "learning_rate": 2.540389972144847e-05, + "loss": 0.2178, + "step": 1573 + }, + { + "epoch": 0.3961492480966463, + "grad_norm": 0.12808546423912048, + "learning_rate": 2.5292479108635097e-05, + "loss": 0.2797, + "step": 1574 + }, + { + "epoch": 0.39640093122758446, + "grad_norm": 0.1347254514694214, + "learning_rate": 2.518105849582173e-05, + "loss": 0.3186, + "step": 1575 + }, + { + "epoch": 0.3966526143585226, + "grad_norm": 0.12524932622909546, + "learning_rate": 2.5069637883008357e-05, + "loss": 0.1896, + "step": 1576 + }, + { + "epoch": 0.39690429748946077, + "grad_norm": 0.11158795654773712, + "learning_rate": 2.4958217270194986e-05, + "loss": 0.1796, + "step": 1577 + }, + { + "epoch": 0.3971559806203989, + "grad_norm": 0.11095944046974182, + "learning_rate": 2.4846796657381616e-05, + "loss": 0.1867, + "step": 1578 + }, + { + "epoch": 0.39740766375133707, + "grad_norm": 0.13477955758571625, + "learning_rate": 2.4735376044568246e-05, + "loss": 0.2201, + "step": 1579 + }, + { + "epoch": 0.3976593468822752, + "grad_norm": 0.12716251611709595, + "learning_rate": 2.4623955431754876e-05, + "loss": 0.2557, + "step": 1580 + }, + { + "epoch": 0.3979110300132134, + "grad_norm": 0.1331804394721985, + "learning_rate": 2.4512534818941505e-05, + "loss": 0.2259, + "step": 1581 + }, + { + "epoch": 0.3981627131441515, + "grad_norm": 0.12287537753582001, + "learning_rate": 2.4401114206128135e-05, + "loss": 0.212, + "step": 1582 + }, + { + "epoch": 0.3984143962750897, + "grad_norm": 0.12359605729579926, + "learning_rate": 2.4289693593314765e-05, + "loss": 0.2648, + "step": 1583 + }, + { + "epoch": 0.39866607940602783, + "grad_norm": 0.14068689942359924, + "learning_rate": 2.4178272980501394e-05, + "loss": 0.2807, + "step": 1584 + }, + { + "epoch": 0.398917762536966, + "grad_norm": 0.11963742226362228, + "learning_rate": 2.4066852367688024e-05, + "loss": 0.2248, + "step": 1585 + }, + { + "epoch": 0.39916944566790413, + "grad_norm": 0.11003325134515762, + "learning_rate": 2.395543175487465e-05, + "loss": 0.1755, + "step": 1586 + }, + { + "epoch": 0.3994211287988423, + "grad_norm": 0.12428192049264908, + "learning_rate": 2.384401114206128e-05, + "loss": 0.2641, + "step": 1587 + }, + { + "epoch": 0.3996728119297804, + "grad_norm": 0.12524215877056122, + "learning_rate": 2.373259052924791e-05, + "loss": 0.1352, + "step": 1588 + }, + { + "epoch": 0.39992449506071853, + "grad_norm": 0.11337368935346603, + "learning_rate": 2.362116991643454e-05, + "loss": 0.1868, + "step": 1589 + }, + { + "epoch": 0.4001761781916567, + "grad_norm": 0.13470786809921265, + "learning_rate": 2.350974930362117e-05, + "loss": 0.3166, + "step": 1590 + }, + { + "epoch": 0.40042786132259484, + "grad_norm": 0.11815305054187775, + "learning_rate": 2.33983286908078e-05, + "loss": 0.204, + "step": 1591 + }, + { + "epoch": 0.400679544453533, + "grad_norm": 0.1277095377445221, + "learning_rate": 2.328690807799443e-05, + "loss": 0.2084, + "step": 1592 + }, + { + "epoch": 0.40093122758447114, + "grad_norm": 0.1264360249042511, + "learning_rate": 2.317548746518106e-05, + "loss": 0.2645, + "step": 1593 + }, + { + "epoch": 0.4011829107154093, + "grad_norm": 0.12079417705535889, + "learning_rate": 2.3064066852367688e-05, + "loss": 0.2727, + "step": 1594 + }, + { + "epoch": 0.40143459384634744, + "grad_norm": 0.11793193221092224, + "learning_rate": 2.2952646239554318e-05, + "loss": 0.1665, + "step": 1595 + }, + { + "epoch": 0.4016862769772856, + "grad_norm": 0.13514013588428497, + "learning_rate": 2.2841225626740948e-05, + "loss": 0.3526, + "step": 1596 + }, + { + "epoch": 0.40193796010822375, + "grad_norm": 0.1337570697069168, + "learning_rate": 2.2729805013927577e-05, + "loss": 0.2644, + "step": 1597 + }, + { + "epoch": 0.4021896432391619, + "grad_norm": 0.1283758282661438, + "learning_rate": 2.2618384401114207e-05, + "loss": 0.1976, + "step": 1598 + }, + { + "epoch": 0.40244132637010005, + "grad_norm": 0.11114092916250229, + "learning_rate": 2.2506963788300837e-05, + "loss": 0.1732, + "step": 1599 + }, + { + "epoch": 0.4026930095010382, + "grad_norm": 0.12203486263751984, + "learning_rate": 2.2395543175487466e-05, + "loss": 0.2394, + "step": 1600 + }, + { + "epoch": 0.40294469263197635, + "grad_norm": 0.10410600900650024, + "learning_rate": 2.2284122562674096e-05, + "loss": 0.1468, + "step": 1601 + }, + { + "epoch": 0.4031963757629145, + "grad_norm": 0.12466426193714142, + "learning_rate": 2.2172701949860726e-05, + "loss": 0.1789, + "step": 1602 + }, + { + "epoch": 0.40344805889385266, + "grad_norm": 0.12807929515838623, + "learning_rate": 2.2061281337047355e-05, + "loss": 0.2996, + "step": 1603 + }, + { + "epoch": 0.4036997420247908, + "grad_norm": 0.134046271443367, + "learning_rate": 2.1949860724233985e-05, + "loss": 0.1677, + "step": 1604 + }, + { + "epoch": 0.40395142515572896, + "grad_norm": 0.11934178322553635, + "learning_rate": 2.1838440111420615e-05, + "loss": 0.1887, + "step": 1605 + }, + { + "epoch": 0.4042031082866671, + "grad_norm": 0.1264006793498993, + "learning_rate": 2.172701949860724e-05, + "loss": 0.2706, + "step": 1606 + }, + { + "epoch": 0.4044547914176052, + "grad_norm": 0.1394217163324356, + "learning_rate": 2.161559888579387e-05, + "loss": 0.2763, + "step": 1607 + }, + { + "epoch": 0.40470647454854336, + "grad_norm": 0.1118813082575798, + "learning_rate": 2.15041782729805e-05, + "loss": 0.1623, + "step": 1608 + }, + { + "epoch": 0.4049581576794815, + "grad_norm": 0.14862200617790222, + "learning_rate": 2.139275766016713e-05, + "loss": 0.3129, + "step": 1609 + }, + { + "epoch": 0.40520984081041966, + "grad_norm": 0.13348925113677979, + "learning_rate": 2.128133704735376e-05, + "loss": 0.2703, + "step": 1610 + }, + { + "epoch": 0.4054615239413578, + "grad_norm": 0.11746223270893097, + "learning_rate": 2.116991643454039e-05, + "loss": 0.2065, + "step": 1611 + }, + { + "epoch": 0.40571320707229597, + "grad_norm": 0.11947312951087952, + "learning_rate": 2.105849582172702e-05, + "loss": 0.2645, + "step": 1612 + }, + { + "epoch": 0.4059648902032341, + "grad_norm": 0.11651267856359482, + "learning_rate": 2.094707520891365e-05, + "loss": 0.2004, + "step": 1613 + }, + { + "epoch": 0.40621657333417227, + "grad_norm": 0.11361946910619736, + "learning_rate": 2.083565459610028e-05, + "loss": 0.1971, + "step": 1614 + }, + { + "epoch": 0.4064682564651104, + "grad_norm": 0.12196743488311768, + "learning_rate": 2.072423398328691e-05, + "loss": 0.1761, + "step": 1615 + }, + { + "epoch": 0.4067199395960486, + "grad_norm": 0.13875369727611542, + "learning_rate": 2.0612813370473538e-05, + "loss": 0.1955, + "step": 1616 + }, + { + "epoch": 0.4069716227269867, + "grad_norm": 0.10357370972633362, + "learning_rate": 2.0501392757660168e-05, + "loss": 0.1556, + "step": 1617 + }, + { + "epoch": 0.4072233058579249, + "grad_norm": 0.14564485847949982, + "learning_rate": 2.0389972144846798e-05, + "loss": 0.2164, + "step": 1618 + }, + { + "epoch": 0.40747498898886303, + "grad_norm": 0.1264670342206955, + "learning_rate": 2.0278551532033427e-05, + "loss": 0.2699, + "step": 1619 + }, + { + "epoch": 0.4077266721198012, + "grad_norm": 0.10554368048906326, + "learning_rate": 2.0167130919220057e-05, + "loss": 0.1586, + "step": 1620 + }, + { + "epoch": 0.40797835525073933, + "grad_norm": 0.126899853348732, + "learning_rate": 2.0055710306406687e-05, + "loss": 0.2447, + "step": 1621 + }, + { + "epoch": 0.4082300383816775, + "grad_norm": 0.11072602868080139, + "learning_rate": 1.9944289693593316e-05, + "loss": 0.2028, + "step": 1622 + }, + { + "epoch": 0.40848172151261564, + "grad_norm": 0.11283150315284729, + "learning_rate": 1.9832869080779946e-05, + "loss": 0.1888, + "step": 1623 + }, + { + "epoch": 0.4087334046435538, + "grad_norm": 0.09984546899795532, + "learning_rate": 1.9721448467966576e-05, + "loss": 0.1427, + "step": 1624 + }, + { + "epoch": 0.40898508777449194, + "grad_norm": 0.10653432458639145, + "learning_rate": 1.9610027855153206e-05, + "loss": 0.1613, + "step": 1625 + }, + { + "epoch": 0.40923677090543004, + "grad_norm": 0.1343163102865219, + "learning_rate": 1.9498607242339832e-05, + "loss": 0.2143, + "step": 1626 + }, + { + "epoch": 0.4094884540363682, + "grad_norm": 0.12554699182510376, + "learning_rate": 1.938718662952646e-05, + "loss": 0.2081, + "step": 1627 + }, + { + "epoch": 0.40974013716730634, + "grad_norm": 0.12273167073726654, + "learning_rate": 1.927576601671309e-05, + "loss": 0.2191, + "step": 1628 + }, + { + "epoch": 0.4099918202982445, + "grad_norm": 0.10582928359508514, + "learning_rate": 1.916434540389972e-05, + "loss": 0.1697, + "step": 1629 + }, + { + "epoch": 0.41024350342918264, + "grad_norm": 0.1342368721961975, + "learning_rate": 1.905292479108635e-05, + "loss": 0.2242, + "step": 1630 + }, + { + "epoch": 0.4104951865601208, + "grad_norm": 0.10649964213371277, + "learning_rate": 1.894150417827298e-05, + "loss": 0.1731, + "step": 1631 + }, + { + "epoch": 0.41074686969105895, + "grad_norm": 0.1511407345533371, + "learning_rate": 1.883008356545961e-05, + "loss": 0.2044, + "step": 1632 + }, + { + "epoch": 0.4109985528219971, + "grad_norm": 0.1291278600692749, + "learning_rate": 1.871866295264624e-05, + "loss": 0.2077, + "step": 1633 + }, + { + "epoch": 0.41125023595293525, + "grad_norm": 0.1145143136382103, + "learning_rate": 1.860724233983287e-05, + "loss": 0.1842, + "step": 1634 + }, + { + "epoch": 0.4115019190838734, + "grad_norm": 0.10685187578201294, + "learning_rate": 1.84958217270195e-05, + "loss": 0.1434, + "step": 1635 + }, + { + "epoch": 0.41175360221481155, + "grad_norm": 0.12742866575717926, + "learning_rate": 1.838440111420613e-05, + "loss": 0.1903, + "step": 1636 + }, + { + "epoch": 0.4120052853457497, + "grad_norm": 0.12571850419044495, + "learning_rate": 1.827298050139276e-05, + "loss": 0.2067, + "step": 1637 + }, + { + "epoch": 0.41225696847668786, + "grad_norm": 0.1156994178891182, + "learning_rate": 1.816155988857939e-05, + "loss": 0.1678, + "step": 1638 + }, + { + "epoch": 0.412508651607626, + "grad_norm": 0.13212306797504425, + "learning_rate": 1.8050139275766018e-05, + "loss": 0.2627, + "step": 1639 + }, + { + "epoch": 0.41276033473856416, + "grad_norm": 0.11550114303827286, + "learning_rate": 1.7938718662952648e-05, + "loss": 0.2172, + "step": 1640 + }, + { + "epoch": 0.4130120178695023, + "grad_norm": 0.12706896662712097, + "learning_rate": 1.7827298050139278e-05, + "loss": 0.2812, + "step": 1641 + }, + { + "epoch": 0.41326370100044046, + "grad_norm": 0.11760733276605606, + "learning_rate": 1.7715877437325907e-05, + "loss": 0.2175, + "step": 1642 + }, + { + "epoch": 0.4135153841313786, + "grad_norm": 0.12683291733264923, + "learning_rate": 1.7604456824512537e-05, + "loss": 0.2346, + "step": 1643 + }, + { + "epoch": 0.41376706726231677, + "grad_norm": 0.12913131713867188, + "learning_rate": 1.7493036211699167e-05, + "loss": 0.2289, + "step": 1644 + }, + { + "epoch": 0.4140187503932549, + "grad_norm": 0.11891523748636246, + "learning_rate": 1.7381615598885796e-05, + "loss": 0.2491, + "step": 1645 + }, + { + "epoch": 0.414270433524193, + "grad_norm": 0.11425753682851791, + "learning_rate": 1.7270194986072426e-05, + "loss": 0.1697, + "step": 1646 + }, + { + "epoch": 0.41452211665513117, + "grad_norm": 0.11220378428697586, + "learning_rate": 1.7158774373259052e-05, + "loss": 0.1674, + "step": 1647 + }, + { + "epoch": 0.4147737997860693, + "grad_norm": 0.13470561802387238, + "learning_rate": 1.7047353760445682e-05, + "loss": 0.2655, + "step": 1648 + }, + { + "epoch": 0.41502548291700747, + "grad_norm": 0.12833577394485474, + "learning_rate": 1.6935933147632312e-05, + "loss": 0.2963, + "step": 1649 + }, + { + "epoch": 0.4152771660479456, + "grad_norm": 0.11308885365724564, + "learning_rate": 1.682451253481894e-05, + "loss": 0.1554, + "step": 1650 + }, + { + "epoch": 0.4155288491788838, + "grad_norm": 0.11535003036260605, + "learning_rate": 1.671309192200557e-05, + "loss": 0.1969, + "step": 1651 + }, + { + "epoch": 0.4157805323098219, + "grad_norm": 0.12915892899036407, + "learning_rate": 1.66016713091922e-05, + "loss": 0.1807, + "step": 1652 + }, + { + "epoch": 0.4160322154407601, + "grad_norm": 0.11030598729848862, + "learning_rate": 1.649025069637883e-05, + "loss": 0.16, + "step": 1653 + }, + { + "epoch": 0.41628389857169823, + "grad_norm": 0.1220245510339737, + "learning_rate": 1.637883008356546e-05, + "loss": 0.1988, + "step": 1654 + }, + { + "epoch": 0.4165355817026364, + "grad_norm": 0.12646174430847168, + "learning_rate": 1.626740947075209e-05, + "loss": 0.2721, + "step": 1655 + }, + { + "epoch": 0.41678726483357453, + "grad_norm": 0.1344011276960373, + "learning_rate": 1.615598885793872e-05, + "loss": 0.1863, + "step": 1656 + }, + { + "epoch": 0.4170389479645127, + "grad_norm": 0.12318551540374756, + "learning_rate": 1.604456824512535e-05, + "loss": 0.1725, + "step": 1657 + }, + { + "epoch": 0.41729063109545084, + "grad_norm": 0.1099114865064621, + "learning_rate": 1.593314763231198e-05, + "loss": 0.1979, + "step": 1658 + }, + { + "epoch": 0.417542314226389, + "grad_norm": 0.10595951229333878, + "learning_rate": 1.582172701949861e-05, + "loss": 0.1857, + "step": 1659 + }, + { + "epoch": 0.41779399735732714, + "grad_norm": 0.1281132698059082, + "learning_rate": 1.571030640668524e-05, + "loss": 0.1894, + "step": 1660 + }, + { + "epoch": 0.4180456804882653, + "grad_norm": 0.11448536068201065, + "learning_rate": 1.5598885793871868e-05, + "loss": 0.1886, + "step": 1661 + }, + { + "epoch": 0.41829736361920344, + "grad_norm": 0.11868792772293091, + "learning_rate": 1.5487465181058498e-05, + "loss": 0.2074, + "step": 1662 + }, + { + "epoch": 0.4185490467501416, + "grad_norm": 0.12809720635414124, + "learning_rate": 1.5376044568245128e-05, + "loss": 0.1896, + "step": 1663 + }, + { + "epoch": 0.41880072988107975, + "grad_norm": 0.12570390105247498, + "learning_rate": 1.5264623955431757e-05, + "loss": 0.2134, + "step": 1664 + }, + { + "epoch": 0.41905241301201784, + "grad_norm": 0.13494408130645752, + "learning_rate": 1.5153203342618385e-05, + "loss": 0.2234, + "step": 1665 + }, + { + "epoch": 0.419304096142956, + "grad_norm": 0.12481972575187683, + "learning_rate": 1.5041782729805015e-05, + "loss": 0.2137, + "step": 1666 + }, + { + "epoch": 0.41955577927389415, + "grad_norm": 0.14138968288898468, + "learning_rate": 1.4930362116991645e-05, + "loss": 0.3497, + "step": 1667 + }, + { + "epoch": 0.4198074624048323, + "grad_norm": 0.12695787847042084, + "learning_rate": 1.4818941504178274e-05, + "loss": 0.2669, + "step": 1668 + }, + { + "epoch": 0.42005914553577045, + "grad_norm": 0.11878138780593872, + "learning_rate": 1.4707520891364904e-05, + "loss": 0.1857, + "step": 1669 + }, + { + "epoch": 0.4203108286667086, + "grad_norm": 0.13491526246070862, + "learning_rate": 1.4596100278551534e-05, + "loss": 0.2131, + "step": 1670 + }, + { + "epoch": 0.42056251179764675, + "grad_norm": 0.12049665302038193, + "learning_rate": 1.4484679665738164e-05, + "loss": 0.19, + "step": 1671 + }, + { + "epoch": 0.4208141949285849, + "grad_norm": 0.1288139969110489, + "learning_rate": 1.4373259052924792e-05, + "loss": 0.2593, + "step": 1672 + }, + { + "epoch": 0.42106587805952306, + "grad_norm": 0.12692755460739136, + "learning_rate": 1.4261838440111421e-05, + "loss": 0.2761, + "step": 1673 + }, + { + "epoch": 0.4213175611904612, + "grad_norm": 0.10401088744401932, + "learning_rate": 1.4150417827298051e-05, + "loss": 0.1522, + "step": 1674 + }, + { + "epoch": 0.42156924432139936, + "grad_norm": 0.13406498730182648, + "learning_rate": 1.403899721448468e-05, + "loss": 0.2071, + "step": 1675 + }, + { + "epoch": 0.4218209274523375, + "grad_norm": 0.10419892519712448, + "learning_rate": 1.392757660167131e-05, + "loss": 0.1513, + "step": 1676 + }, + { + "epoch": 0.42207261058327566, + "grad_norm": 0.11039672791957855, + "learning_rate": 1.381615598885794e-05, + "loss": 0.1476, + "step": 1677 + }, + { + "epoch": 0.4223242937142138, + "grad_norm": 0.11776840686798096, + "learning_rate": 1.370473537604457e-05, + "loss": 0.1847, + "step": 1678 + }, + { + "epoch": 0.42257597684515197, + "grad_norm": 0.13029669225215912, + "learning_rate": 1.35933147632312e-05, + "loss": 0.2574, + "step": 1679 + }, + { + "epoch": 0.4228276599760901, + "grad_norm": 0.11814728379249573, + "learning_rate": 1.348189415041783e-05, + "loss": 0.1606, + "step": 1680 + }, + { + "epoch": 0.42307934310702827, + "grad_norm": 0.10507238656282425, + "learning_rate": 1.3370473537604459e-05, + "loss": 0.1909, + "step": 1681 + }, + { + "epoch": 0.4233310262379664, + "grad_norm": 0.1118084043264389, + "learning_rate": 1.3259052924791087e-05, + "loss": 0.1664, + "step": 1682 + }, + { + "epoch": 0.4235827093689046, + "grad_norm": 0.11687816679477692, + "learning_rate": 1.3147632311977717e-05, + "loss": 0.175, + "step": 1683 + }, + { + "epoch": 0.42383439249984267, + "grad_norm": 0.11954981833696365, + "learning_rate": 1.3036211699164346e-05, + "loss": 0.2379, + "step": 1684 + }, + { + "epoch": 0.4240860756307808, + "grad_norm": 0.13958217203617096, + "learning_rate": 1.2924791086350976e-05, + "loss": 0.3296, + "step": 1685 + }, + { + "epoch": 0.424337758761719, + "grad_norm": 0.12187931686639786, + "learning_rate": 1.2813370473537606e-05, + "loss": 0.1836, + "step": 1686 + }, + { + "epoch": 0.4245894418926571, + "grad_norm": 0.1378149539232254, + "learning_rate": 1.2701949860724236e-05, + "loss": 0.1666, + "step": 1687 + }, + { + "epoch": 0.4248411250235953, + "grad_norm": 0.13840830326080322, + "learning_rate": 1.2590529247910865e-05, + "loss": 0.2931, + "step": 1688 + }, + { + "epoch": 0.42509280815453343, + "grad_norm": 0.12253736704587936, + "learning_rate": 1.2479108635097493e-05, + "loss": 0.2023, + "step": 1689 + }, + { + "epoch": 0.4253444912854716, + "grad_norm": 0.11376120895147324, + "learning_rate": 1.2367688022284123e-05, + "loss": 0.1682, + "step": 1690 + }, + { + "epoch": 0.42559617441640973, + "grad_norm": 0.12678204476833344, + "learning_rate": 1.2256267409470753e-05, + "loss": 0.2971, + "step": 1691 + }, + { + "epoch": 0.4258478575473479, + "grad_norm": 0.11819051951169968, + "learning_rate": 1.2144846796657382e-05, + "loss": 0.2291, + "step": 1692 + }, + { + "epoch": 0.42609954067828604, + "grad_norm": 0.09422247111797333, + "learning_rate": 1.2033426183844012e-05, + "loss": 0.1366, + "step": 1693 + }, + { + "epoch": 0.4263512238092242, + "grad_norm": 0.12087205797433853, + "learning_rate": 1.192200557103064e-05, + "loss": 0.2098, + "step": 1694 + }, + { + "epoch": 0.42660290694016234, + "grad_norm": 0.1267717331647873, + "learning_rate": 1.181058495821727e-05, + "loss": 0.3056, + "step": 1695 + }, + { + "epoch": 0.4268545900711005, + "grad_norm": 0.11077269166707993, + "learning_rate": 1.16991643454039e-05, + "loss": 0.1831, + "step": 1696 + }, + { + "epoch": 0.42710627320203864, + "grad_norm": 0.11711994558572769, + "learning_rate": 1.158774373259053e-05, + "loss": 0.1814, + "step": 1697 + }, + { + "epoch": 0.4273579563329768, + "grad_norm": 0.1243748590350151, + "learning_rate": 1.1476323119777159e-05, + "loss": 0.1727, + "step": 1698 + }, + { + "epoch": 0.42760963946391495, + "grad_norm": 0.10812552273273468, + "learning_rate": 1.1364902506963789e-05, + "loss": 0.1588, + "step": 1699 + }, + { + "epoch": 0.4278613225948531, + "grad_norm": 0.1100439578294754, + "learning_rate": 1.1253481894150418e-05, + "loss": 0.1658, + "step": 1700 + }, + { + "epoch": 0.42811300572579125, + "grad_norm": 0.11406370997428894, + "learning_rate": 1.1142061281337048e-05, + "loss": 0.1925, + "step": 1701 + }, + { + "epoch": 0.4283646888567294, + "grad_norm": 0.1281452625989914, + "learning_rate": 1.1030640668523678e-05, + "loss": 0.2421, + "step": 1702 + }, + { + "epoch": 0.4286163719876675, + "grad_norm": 0.12457474321126938, + "learning_rate": 1.0919220055710307e-05, + "loss": 0.2695, + "step": 1703 + }, + { + "epoch": 0.42886805511860565, + "grad_norm": 0.12564806640148163, + "learning_rate": 1.0807799442896935e-05, + "loss": 0.2272, + "step": 1704 + }, + { + "epoch": 0.4291197382495438, + "grad_norm": 0.10148760676383972, + "learning_rate": 1.0696378830083565e-05, + "loss": 0.1785, + "step": 1705 + }, + { + "epoch": 0.42937142138048195, + "grad_norm": 0.13082671165466309, + "learning_rate": 1.0584958217270195e-05, + "loss": 0.2078, + "step": 1706 + }, + { + "epoch": 0.4296231045114201, + "grad_norm": 0.10842859745025635, + "learning_rate": 1.0473537604456825e-05, + "loss": 0.1552, + "step": 1707 + }, + { + "epoch": 0.42987478764235826, + "grad_norm": 0.14441132545471191, + "learning_rate": 1.0362116991643454e-05, + "loss": 0.2473, + "step": 1708 + }, + { + "epoch": 0.4301264707732964, + "grad_norm": 0.09071933478116989, + "learning_rate": 1.0250696378830084e-05, + "loss": 0.1343, + "step": 1709 + }, + { + "epoch": 0.43037815390423456, + "grad_norm": 0.10789783298969269, + "learning_rate": 1.0139275766016714e-05, + "loss": 0.1676, + "step": 1710 + }, + { + "epoch": 0.4306298370351727, + "grad_norm": 0.12387588620185852, + "learning_rate": 1.0027855153203343e-05, + "loss": 0.1949, + "step": 1711 + }, + { + "epoch": 0.43088152016611087, + "grad_norm": 0.11823824048042297, + "learning_rate": 9.916434540389973e-06, + "loss": 0.1732, + "step": 1712 + }, + { + "epoch": 0.431133203297049, + "grad_norm": 0.11307192593812943, + "learning_rate": 9.805013927576603e-06, + "loss": 0.173, + "step": 1713 + }, + { + "epoch": 0.43138488642798717, + "grad_norm": 0.11619716137647629, + "learning_rate": 9.69359331476323e-06, + "loss": 0.1809, + "step": 1714 + }, + { + "epoch": 0.4316365695589253, + "grad_norm": 0.12573541700839996, + "learning_rate": 9.58217270194986e-06, + "loss": 0.1741, + "step": 1715 + }, + { + "epoch": 0.43188825268986347, + "grad_norm": 0.12677261233329773, + "learning_rate": 9.47075208913649e-06, + "loss": 0.2831, + "step": 1716 + }, + { + "epoch": 0.4321399358208016, + "grad_norm": 0.11583074182271957, + "learning_rate": 9.35933147632312e-06, + "loss": 0.223, + "step": 1717 + }, + { + "epoch": 0.4323916189517398, + "grad_norm": 0.12441383302211761, + "learning_rate": 9.24791086350975e-06, + "loss": 0.1897, + "step": 1718 + }, + { + "epoch": 0.4326433020826779, + "grad_norm": 0.11675354093313217, + "learning_rate": 9.13649025069638e-06, + "loss": 0.2039, + "step": 1719 + }, + { + "epoch": 0.4328949852136161, + "grad_norm": 0.1284424066543579, + "learning_rate": 9.025069637883009e-06, + "loss": 0.2687, + "step": 1720 + }, + { + "epoch": 0.43314666834455423, + "grad_norm": 0.13134267926216125, + "learning_rate": 8.913649025069639e-06, + "loss": 0.222, + "step": 1721 + }, + { + "epoch": 0.4333983514754924, + "grad_norm": 0.12908576428890228, + "learning_rate": 8.802228412256268e-06, + "loss": 0.2049, + "step": 1722 + }, + { + "epoch": 0.4336500346064305, + "grad_norm": 0.11427567154169083, + "learning_rate": 8.690807799442898e-06, + "loss": 0.1939, + "step": 1723 + }, + { + "epoch": 0.43390171773736863, + "grad_norm": 0.12355178594589233, + "learning_rate": 8.579387186629526e-06, + "loss": 0.164, + "step": 1724 + }, + { + "epoch": 0.4341534008683068, + "grad_norm": 0.10546781867742538, + "learning_rate": 8.467966573816156e-06, + "loss": 0.1715, + "step": 1725 + }, + { + "epoch": 0.43440508399924493, + "grad_norm": 0.13099981844425201, + "learning_rate": 8.356545961002786e-06, + "loss": 0.2949, + "step": 1726 + }, + { + "epoch": 0.4346567671301831, + "grad_norm": 0.13408686220645905, + "learning_rate": 8.245125348189415e-06, + "loss": 0.2555, + "step": 1727 + }, + { + "epoch": 0.43490845026112124, + "grad_norm": 0.11657360196113586, + "learning_rate": 8.133704735376045e-06, + "loss": 0.1474, + "step": 1728 + }, + { + "epoch": 0.4351601333920594, + "grad_norm": 0.11211398243904114, + "learning_rate": 8.022284122562675e-06, + "loss": 0.1829, + "step": 1729 + }, + { + "epoch": 0.43541181652299754, + "grad_norm": 0.11190121620893478, + "learning_rate": 7.910863509749304e-06, + "loss": 0.1689, + "step": 1730 + }, + { + "epoch": 0.4356634996539357, + "grad_norm": 0.11207761615514755, + "learning_rate": 7.799442896935934e-06, + "loss": 0.1904, + "step": 1731 + }, + { + "epoch": 0.43591518278487384, + "grad_norm": 0.1225392296910286, + "learning_rate": 7.688022284122564e-06, + "loss": 0.2609, + "step": 1732 + }, + { + "epoch": 0.436166865915812, + "grad_norm": 0.11663854867219925, + "learning_rate": 7.576601671309193e-06, + "loss": 0.1803, + "step": 1733 + }, + { + "epoch": 0.43641854904675015, + "grad_norm": 0.11473322659730911, + "learning_rate": 7.465181058495822e-06, + "loss": 0.1848, + "step": 1734 + }, + { + "epoch": 0.4366702321776883, + "grad_norm": 0.11451229453086853, + "learning_rate": 7.353760445682452e-06, + "loss": 0.155, + "step": 1735 + }, + { + "epoch": 0.43692191530862645, + "grad_norm": 0.1351272314786911, + "learning_rate": 7.242339832869082e-06, + "loss": 0.234, + "step": 1736 + }, + { + "epoch": 0.4371735984395646, + "grad_norm": 0.11213961988687515, + "learning_rate": 7.130919220055711e-06, + "loss": 0.1883, + "step": 1737 + }, + { + "epoch": 0.43742528157050276, + "grad_norm": 0.11221275478601456, + "learning_rate": 7.01949860724234e-06, + "loss": 0.1791, + "step": 1738 + }, + { + "epoch": 0.4376769647014409, + "grad_norm": 0.1166306883096695, + "learning_rate": 6.90807799442897e-06, + "loss": 0.1594, + "step": 1739 + }, + { + "epoch": 0.43792864783237906, + "grad_norm": 0.09530488401651382, + "learning_rate": 6.7966573816156e-06, + "loss": 0.1342, + "step": 1740 + }, + { + "epoch": 0.4381803309633172, + "grad_norm": 0.127318874001503, + "learning_rate": 6.6852367688022295e-06, + "loss": 0.2446, + "step": 1741 + }, + { + "epoch": 0.4384320140942553, + "grad_norm": 0.11398839205503464, + "learning_rate": 6.573816155988858e-06, + "loss": 0.1513, + "step": 1742 + }, + { + "epoch": 0.43868369722519346, + "grad_norm": 0.11196225136518478, + "learning_rate": 6.462395543175488e-06, + "loss": 0.1622, + "step": 1743 + }, + { + "epoch": 0.4389353803561316, + "grad_norm": 0.11288930475711823, + "learning_rate": 6.350974930362118e-06, + "loss": 0.1509, + "step": 1744 + }, + { + "epoch": 0.43918706348706976, + "grad_norm": 0.12640945613384247, + "learning_rate": 6.239554317548747e-06, + "loss": 0.2155, + "step": 1745 + }, + { + "epoch": 0.4394387466180079, + "grad_norm": 0.12675067782402039, + "learning_rate": 6.128133704735376e-06, + "loss": 0.1948, + "step": 1746 + }, + { + "epoch": 0.43969042974894607, + "grad_norm": 0.12437254935503006, + "learning_rate": 6.016713091922006e-06, + "loss": 0.2338, + "step": 1747 + }, + { + "epoch": 0.4399421128798842, + "grad_norm": 0.10237505286931992, + "learning_rate": 5.905292479108635e-06, + "loss": 0.1581, + "step": 1748 + }, + { + "epoch": 0.44019379601082237, + "grad_norm": 0.1185777559876442, + "learning_rate": 5.793871866295265e-06, + "loss": 0.1726, + "step": 1749 + }, + { + "epoch": 0.4404454791417605, + "grad_norm": 0.13407579064369202, + "learning_rate": 5.682451253481894e-06, + "loss": 0.2593, + "step": 1750 + }, + { + "epoch": 0.4406971622726987, + "grad_norm": 0.11823869496583939, + "learning_rate": 5.571030640668524e-06, + "loss": 0.1958, + "step": 1751 + }, + { + "epoch": 0.4409488454036368, + "grad_norm": 0.12665455043315887, + "learning_rate": 5.459610027855154e-06, + "loss": 0.2831, + "step": 1752 + }, + { + "epoch": 0.441200528534575, + "grad_norm": 0.11812491714954376, + "learning_rate": 5.348189415041783e-06, + "loss": 0.2452, + "step": 1753 + }, + { + "epoch": 0.44145221166551313, + "grad_norm": 0.10584802180528641, + "learning_rate": 5.236768802228412e-06, + "loss": 0.1685, + "step": 1754 + }, + { + "epoch": 0.4417038947964513, + "grad_norm": 0.11695295572280884, + "learning_rate": 5.125348189415042e-06, + "loss": 0.1804, + "step": 1755 + }, + { + "epoch": 0.44195557792738943, + "grad_norm": 0.12248878180980682, + "learning_rate": 5.013927576601672e-06, + "loss": 0.1745, + "step": 1756 + }, + { + "epoch": 0.4422072610583276, + "grad_norm": 0.1335570365190506, + "learning_rate": 4.902506963788301e-06, + "loss": 0.2801, + "step": 1757 + }, + { + "epoch": 0.44245894418926573, + "grad_norm": 0.12324695289134979, + "learning_rate": 4.79108635097493e-06, + "loss": 0.2076, + "step": 1758 + }, + { + "epoch": 0.4427106273202039, + "grad_norm": 0.10358007997274399, + "learning_rate": 4.67966573816156e-06, + "loss": 0.121, + "step": 1759 + }, + { + "epoch": 0.44296231045114204, + "grad_norm": 0.11502121388912201, + "learning_rate": 4.56824512534819e-06, + "loss": 0.1624, + "step": 1760 + }, + { + "epoch": 0.44321399358208013, + "grad_norm": 0.12248494476079941, + "learning_rate": 4.456824512534819e-06, + "loss": 0.2897, + "step": 1761 + }, + { + "epoch": 0.4434656767130183, + "grad_norm": 0.136197030544281, + "learning_rate": 4.345403899721449e-06, + "loss": 0.2271, + "step": 1762 + }, + { + "epoch": 0.44371735984395644, + "grad_norm": 0.12870381772518158, + "learning_rate": 4.233983286908078e-06, + "loss": 0.2786, + "step": 1763 + }, + { + "epoch": 0.4439690429748946, + "grad_norm": 0.10985758900642395, + "learning_rate": 4.122562674094708e-06, + "loss": 0.1573, + "step": 1764 + }, + { + "epoch": 0.44422072610583274, + "grad_norm": 0.1309724748134613, + "learning_rate": 4.011142061281337e-06, + "loss": 0.2581, + "step": 1765 + }, + { + "epoch": 0.4444724092367709, + "grad_norm": 0.13059131801128387, + "learning_rate": 3.899721448467967e-06, + "loss": 0.2535, + "step": 1766 + }, + { + "epoch": 0.44472409236770905, + "grad_norm": 0.11294321715831757, + "learning_rate": 3.7883008356545963e-06, + "loss": 0.1522, + "step": 1767 + }, + { + "epoch": 0.4449757754986472, + "grad_norm": 0.11899720132350922, + "learning_rate": 3.676880222841226e-06, + "loss": 0.1618, + "step": 1768 + }, + { + "epoch": 0.44522745862958535, + "grad_norm": 0.11886916309595108, + "learning_rate": 3.5654596100278553e-06, + "loss": 0.202, + "step": 1769 + }, + { + "epoch": 0.4454791417605235, + "grad_norm": 0.12279828637838364, + "learning_rate": 3.454038997214485e-06, + "loss": 0.1976, + "step": 1770 + }, + { + "epoch": 0.44573082489146165, + "grad_norm": 0.11792545765638351, + "learning_rate": 3.3426183844011147e-06, + "loss": 0.1635, + "step": 1771 + }, + { + "epoch": 0.4459825080223998, + "grad_norm": 0.1324690878391266, + "learning_rate": 3.231197771587744e-06, + "loss": 0.229, + "step": 1772 + }, + { + "epoch": 0.44623419115333796, + "grad_norm": 0.1192900538444519, + "learning_rate": 3.1197771587743733e-06, + "loss": 0.1919, + "step": 1773 + }, + { + "epoch": 0.4464858742842761, + "grad_norm": 0.13808223605155945, + "learning_rate": 3.008356545961003e-06, + "loss": 0.2583, + "step": 1774 + }, + { + "epoch": 0.44673755741521426, + "grad_norm": 0.11708478629589081, + "learning_rate": 2.8969359331476323e-06, + "loss": 0.2023, + "step": 1775 + }, + { + "epoch": 0.4469892405461524, + "grad_norm": 0.13183362782001495, + "learning_rate": 2.785515320334262e-06, + "loss": 0.3166, + "step": 1776 + }, + { + "epoch": 0.44724092367709056, + "grad_norm": 0.10360544919967651, + "learning_rate": 2.6740947075208913e-06, + "loss": 0.1578, + "step": 1777 + }, + { + "epoch": 0.4474926068080287, + "grad_norm": 0.13705076277256012, + "learning_rate": 2.562674094707521e-06, + "loss": 0.1659, + "step": 1778 + }, + { + "epoch": 0.44774428993896687, + "grad_norm": 0.12482505291700363, + "learning_rate": 2.4512534818941507e-06, + "loss": 0.2097, + "step": 1779 + }, + { + "epoch": 0.44799597306990496, + "grad_norm": 0.11935511976480484, + "learning_rate": 2.33983286908078e-06, + "loss": 0.194, + "step": 1780 + }, + { + "epoch": 0.4482476562008431, + "grad_norm": 0.10971649736166, + "learning_rate": 2.2284122562674097e-06, + "loss": 0.1733, + "step": 1781 + }, + { + "epoch": 0.44849933933178127, + "grad_norm": 0.12503021955490112, + "learning_rate": 2.116991643454039e-06, + "loss": 0.2356, + "step": 1782 + }, + { + "epoch": 0.4487510224627194, + "grad_norm": 0.12967701256275177, + "learning_rate": 2.0055710306406687e-06, + "loss": 0.193, + "step": 1783 + }, + { + "epoch": 0.44900270559365757, + "grad_norm": 0.12671984732151031, + "learning_rate": 1.8941504178272982e-06, + "loss": 0.2196, + "step": 1784 + }, + { + "epoch": 0.4492543887245957, + "grad_norm": 0.13080210983753204, + "learning_rate": 1.7827298050139277e-06, + "loss": 0.2125, + "step": 1785 + }, + { + "epoch": 0.4495060718555339, + "grad_norm": 0.10942334681749344, + "learning_rate": 1.6713091922005574e-06, + "loss": 0.1743, + "step": 1786 + }, + { + "epoch": 0.449757754986472, + "grad_norm": 0.13126231729984283, + "learning_rate": 1.5598885793871867e-06, + "loss": 0.2406, + "step": 1787 + }, + { + "epoch": 0.4500094381174102, + "grad_norm": 0.13587597012519836, + "learning_rate": 1.4484679665738162e-06, + "loss": 0.2156, + "step": 1788 + }, + { + "epoch": 0.45026112124834833, + "grad_norm": 0.1082504391670227, + "learning_rate": 1.3370473537604456e-06, + "loss": 0.1632, + "step": 1789 + }, + { + "epoch": 0.4505128043792865, + "grad_norm": 0.1333242952823639, + "learning_rate": 1.2256267409470754e-06, + "loss": 0.2403, + "step": 1790 + }, + { + "epoch": 0.45076448751022463, + "grad_norm": 0.11991868913173676, + "learning_rate": 1.1142061281337048e-06, + "loss": 0.1678, + "step": 1791 + }, + { + "epoch": 0.4510161706411628, + "grad_norm": 0.127787247300148, + "learning_rate": 1.0027855153203343e-06, + "loss": 0.3245, + "step": 1792 + }, + { + "epoch": 0.45126785377210094, + "grad_norm": 0.12396389245986938, + "learning_rate": 8.913649025069638e-07, + "loss": 0.2053, + "step": 1793 + }, + { + "epoch": 0.4515195369030391, + "grad_norm": 0.14221100509166718, + "learning_rate": 7.799442896935933e-07, + "loss": 0.1949, + "step": 1794 + }, + { + "epoch": 0.45177122003397724, + "grad_norm": 0.13420484960079193, + "learning_rate": 6.685236768802228e-07, + "loss": 0.2931, + "step": 1795 + }, + { + "epoch": 0.4520229031649154, + "grad_norm": 0.12022432684898376, + "learning_rate": 5.571030640668524e-07, + "loss": 0.1676, + "step": 1796 + }, + { + "epoch": 0.45227458629585354, + "grad_norm": 0.11700937896966934, + "learning_rate": 4.456824512534819e-07, + "loss": 0.2085, + "step": 1797 + }, + { + "epoch": 0.4525262694267917, + "grad_norm": 0.12383181601762772, + "learning_rate": 3.342618384401114e-07, + "loss": 0.1646, + "step": 1798 + }, + { + "epoch": 0.45277795255772985, + "grad_norm": 0.11531715095043182, + "learning_rate": 2.2284122562674096e-07, + "loss": 0.1877, + "step": 1799 + }, + { + "epoch": 0.45302963568866794, + "grad_norm": 0.12468922883272171, + "learning_rate": 1.1142061281337048e-07, + "loss": 0.2575, + "step": 1800 + } + ], + "logging_steps": 1, + "max_steps": 1800, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 4.193370706405048e+17, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}