diff --git "a/checkpoint-11100/trainer_state.json" "b/checkpoint-11100/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-11100/trainer_state.json" @@ -0,0 +1,77734 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.5255972696245736, + "eval_steps": 500, + "global_step": 11100, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00022753128555176336, + "grad_norm": 12.3071932545212, + "learning_rate": 1.25e-06, + "loss": 0.3186, + "step": 1 + }, + { + "epoch": 0.0004550625711035267, + "grad_norm": 11.959825961880057, + "learning_rate": 1.2499999936130725e-06, + "loss": 0.3776, + "step": 2 + }, + { + "epoch": 0.0006825938566552901, + "grad_norm": 5.4315221034586365, + "learning_rate": 1.2499999744522896e-06, + "loss": 0.4755, + "step": 3 + }, + { + "epoch": 0.0009101251422070534, + "grad_norm": 21.003860231065644, + "learning_rate": 1.2499999425176518e-06, + "loss": 0.3334, + "step": 4 + }, + { + "epoch": 0.0011376564277588168, + "grad_norm": 9.549170994835775, + "learning_rate": 1.2499998978091598e-06, + "loss": 0.375, + "step": 5 + }, + { + "epoch": 0.0013651877133105802, + "grad_norm": 3.400827392368318, + "learning_rate": 1.2499998403268147e-06, + "loss": 0.2286, + "step": 6 + }, + { + "epoch": 0.0015927189988623437, + "grad_norm": 8.451175634489234, + "learning_rate": 1.2499997700706173e-06, + "loss": 0.3216, + "step": 7 + }, + { + "epoch": 0.0018202502844141069, + "grad_norm": 7.494987211346803, + "learning_rate": 1.2499996870405692e-06, + "loss": 0.2339, + "step": 8 + }, + { + "epoch": 0.0020477815699658703, + "grad_norm": 9.138399201835718, + "learning_rate": 1.2499995912366722e-06, + "loss": 0.326, + "step": 9 + }, + { + "epoch": 0.0022753128555176336, + "grad_norm": 3.2188295955534123, + "learning_rate": 1.2499994826589282e-06, + "loss": 0.2514, + "step": 10 + }, + { + "epoch": 0.002502844141069397, + "grad_norm": 22.66663249526738, + "learning_rate": 1.2499993613073393e-06, + "loss": 0.4005, + "step": 11 + }, + { + "epoch": 0.0027303754266211604, + "grad_norm": 21.954799290073527, + "learning_rate": 1.2499992271819083e-06, + "loss": 0.1492, + "step": 12 + }, + { + "epoch": 0.0029579067121729237, + "grad_norm": 6.624298118045555, + "learning_rate": 1.2499990802826377e-06, + "loss": 0.3024, + "step": 13 + }, + { + "epoch": 0.0031854379977246873, + "grad_norm": 3.923975454400113, + "learning_rate": 1.2499989206095304e-06, + "loss": 0.2411, + "step": 14 + }, + { + "epoch": 0.0034129692832764505, + "grad_norm": 1.9398605915746092, + "learning_rate": 1.2499987481625899e-06, + "loss": 0.1849, + "step": 15 + }, + { + "epoch": 0.0036405005688282138, + "grad_norm": 5.482695785208493, + "learning_rate": 1.2499985629418195e-06, + "loss": 0.3122, + "step": 16 + }, + { + "epoch": 0.0038680318543799774, + "grad_norm": 5.552109300872593, + "learning_rate": 1.2499983649472233e-06, + "loss": 0.3393, + "step": 17 + }, + { + "epoch": 0.004095563139931741, + "grad_norm": 4.610318891370888, + "learning_rate": 1.249998154178805e-06, + "loss": 0.3, + "step": 18 + }, + { + "epoch": 0.004323094425483504, + "grad_norm": 8.793267315718285, + "learning_rate": 1.2499979306365692e-06, + "loss": 0.2266, + "step": 19 + }, + { + "epoch": 0.004550625711035267, + "grad_norm": 11.785540460314868, + "learning_rate": 1.2499976943205202e-06, + "loss": 0.258, + "step": 20 + }, + { + "epoch": 0.00477815699658703, + "grad_norm": 7.848333910468807, + "learning_rate": 1.249997445230663e-06, + "loss": 0.3733, + "step": 21 + }, + { + "epoch": 0.005005688282138794, + "grad_norm": 11.509651474854413, + "learning_rate": 1.2499971833670026e-06, + "loss": 0.3606, + "step": 22 + }, + { + "epoch": 0.005233219567690558, + "grad_norm": 8.662973783002895, + "learning_rate": 1.2499969087295443e-06, + "loss": 0.3884, + "step": 23 + }, + { + "epoch": 0.005460750853242321, + "grad_norm": 5.341258812295752, + "learning_rate": 1.249996621318294e-06, + "loss": 0.2677, + "step": 24 + }, + { + "epoch": 0.005688282138794084, + "grad_norm": 4.742018594757072, + "learning_rate": 1.2499963211332573e-06, + "loss": 0.3253, + "step": 25 + }, + { + "epoch": 0.005915813424345847, + "grad_norm": 2.4536573603250624, + "learning_rate": 1.2499960081744405e-06, + "loss": 0.2393, + "step": 26 + }, + { + "epoch": 0.0061433447098976105, + "grad_norm": 6.34705088291597, + "learning_rate": 1.24999568244185e-06, + "loss": 0.4326, + "step": 27 + }, + { + "epoch": 0.006370875995449375, + "grad_norm": 9.775833264439491, + "learning_rate": 1.249995343935492e-06, + "loss": 0.4252, + "step": 28 + }, + { + "epoch": 0.006598407281001138, + "grad_norm": 6.064212735225404, + "learning_rate": 1.2499949926553743e-06, + "loss": 0.2988, + "step": 29 + }, + { + "epoch": 0.006825938566552901, + "grad_norm": 4.4254830237015845, + "learning_rate": 1.2499946286015032e-06, + "loss": 0.2988, + "step": 30 + }, + { + "epoch": 0.007053469852104664, + "grad_norm": 4.883047495609927, + "learning_rate": 1.2499942517738867e-06, + "loss": 0.2285, + "step": 31 + }, + { + "epoch": 0.0072810011376564275, + "grad_norm": 8.135398866699179, + "learning_rate": 1.2499938621725322e-06, + "loss": 0.1529, + "step": 32 + }, + { + "epoch": 0.007508532423208191, + "grad_norm": 2.973365765084456, + "learning_rate": 1.2499934597974478e-06, + "loss": 0.2436, + "step": 33 + }, + { + "epoch": 0.007736063708759955, + "grad_norm": 5.612693729952574, + "learning_rate": 1.2499930446486416e-06, + "loss": 0.3466, + "step": 34 + }, + { + "epoch": 0.007963594994311717, + "grad_norm": 3.022290156639827, + "learning_rate": 1.2499926167261224e-06, + "loss": 0.2728, + "step": 35 + }, + { + "epoch": 0.008191126279863481, + "grad_norm": 3.1279992715224467, + "learning_rate": 1.2499921760298987e-06, + "loss": 0.2469, + "step": 36 + }, + { + "epoch": 0.008418657565415245, + "grad_norm": 14.845448376418034, + "learning_rate": 1.2499917225599796e-06, + "loss": 0.5145, + "step": 37 + }, + { + "epoch": 0.008646188850967008, + "grad_norm": 14.138433401115075, + "learning_rate": 1.2499912563163742e-06, + "loss": 0.2705, + "step": 38 + }, + { + "epoch": 0.008873720136518772, + "grad_norm": 4.324563647824762, + "learning_rate": 1.249990777299092e-06, + "loss": 0.1563, + "step": 39 + }, + { + "epoch": 0.009101251422070534, + "grad_norm": 11.315529959215173, + "learning_rate": 1.249990285508143e-06, + "loss": 0.4123, + "step": 40 + }, + { + "epoch": 0.009328782707622298, + "grad_norm": 6.3112839729366765, + "learning_rate": 1.2499897809435374e-06, + "loss": 0.1742, + "step": 41 + }, + { + "epoch": 0.00955631399317406, + "grad_norm": 8.25726966946455, + "learning_rate": 1.249989263605285e-06, + "loss": 0.3229, + "step": 42 + }, + { + "epoch": 0.009783845278725825, + "grad_norm": 6.3545712967505334, + "learning_rate": 1.249988733493397e-06, + "loss": 0.3055, + "step": 43 + }, + { + "epoch": 0.010011376564277589, + "grad_norm": 5.356373706603287, + "learning_rate": 1.2499881906078836e-06, + "loss": 0.2601, + "step": 44 + }, + { + "epoch": 0.010238907849829351, + "grad_norm": 1.9215795165819936, + "learning_rate": 1.2499876349487564e-06, + "loss": 0.1517, + "step": 45 + }, + { + "epoch": 0.010466439135381115, + "grad_norm": 8.506503892761648, + "learning_rate": 1.2499870665160262e-06, + "loss": 0.2831, + "step": 46 + }, + { + "epoch": 0.010693970420932878, + "grad_norm": 5.909503420571465, + "learning_rate": 1.2499864853097054e-06, + "loss": 0.2252, + "step": 47 + }, + { + "epoch": 0.010921501706484642, + "grad_norm": 5.488265194188453, + "learning_rate": 1.2499858913298053e-06, + "loss": 0.3466, + "step": 48 + }, + { + "epoch": 0.011149032992036406, + "grad_norm": 12.162427245650075, + "learning_rate": 1.249985284576338e-06, + "loss": 0.2426, + "step": 49 + }, + { + "epoch": 0.011376564277588168, + "grad_norm": 9.969211407495816, + "learning_rate": 1.2499846650493164e-06, + "loss": 0.2801, + "step": 50 + }, + { + "epoch": 0.011604095563139932, + "grad_norm": 5.741578552447352, + "learning_rate": 1.2499840327487528e-06, + "loss": 0.2664, + "step": 51 + }, + { + "epoch": 0.011831626848691695, + "grad_norm": 2.937767840084915, + "learning_rate": 1.24998338767466e-06, + "loss": 0.1834, + "step": 52 + }, + { + "epoch": 0.012059158134243459, + "grad_norm": 4.130655112830682, + "learning_rate": 1.2499827298270515e-06, + "loss": 0.2675, + "step": 53 + }, + { + "epoch": 0.012286689419795221, + "grad_norm": 4.5227789119131625, + "learning_rate": 1.2499820592059405e-06, + "loss": 0.3205, + "step": 54 + }, + { + "epoch": 0.012514220705346985, + "grad_norm": 4.653850683576537, + "learning_rate": 1.2499813758113409e-06, + "loss": 0.1921, + "step": 55 + }, + { + "epoch": 0.01274175199089875, + "grad_norm": 6.204991552012506, + "learning_rate": 1.2499806796432665e-06, + "loss": 0.1989, + "step": 56 + }, + { + "epoch": 0.012969283276450512, + "grad_norm": 7.81696538748595, + "learning_rate": 1.2499799707017315e-06, + "loss": 0.1301, + "step": 57 + }, + { + "epoch": 0.013196814562002276, + "grad_norm": 6.427887275035889, + "learning_rate": 1.2499792489867508e-06, + "loss": 0.3376, + "step": 58 + }, + { + "epoch": 0.013424345847554038, + "grad_norm": 4.713573539887475, + "learning_rate": 1.2499785144983386e-06, + "loss": 0.1673, + "step": 59 + }, + { + "epoch": 0.013651877133105802, + "grad_norm": 6.7169275734426055, + "learning_rate": 1.24997776723651e-06, + "loss": 0.2501, + "step": 60 + }, + { + "epoch": 0.013879408418657566, + "grad_norm": 11.702392641770421, + "learning_rate": 1.2499770072012809e-06, + "loss": 0.293, + "step": 61 + }, + { + "epoch": 0.014106939704209329, + "grad_norm": 5.86563350345107, + "learning_rate": 1.2499762343926661e-06, + "loss": 0.2346, + "step": 62 + }, + { + "epoch": 0.014334470989761093, + "grad_norm": 4.562933746130791, + "learning_rate": 1.2499754488106817e-06, + "loss": 0.1349, + "step": 63 + }, + { + "epoch": 0.014562002275312855, + "grad_norm": 16.935870758573948, + "learning_rate": 1.2499746504553436e-06, + "loss": 0.2869, + "step": 64 + }, + { + "epoch": 0.01478953356086462, + "grad_norm": 3.252674290241083, + "learning_rate": 1.2499738393266684e-06, + "loss": 0.2125, + "step": 65 + }, + { + "epoch": 0.015017064846416382, + "grad_norm": 3.767321260449828, + "learning_rate": 1.2499730154246726e-06, + "loss": 0.2049, + "step": 66 + }, + { + "epoch": 0.015244596131968146, + "grad_norm": 7.264091175555215, + "learning_rate": 1.2499721787493726e-06, + "loss": 0.2521, + "step": 67 + }, + { + "epoch": 0.01547212741751991, + "grad_norm": 2.846384337735166, + "learning_rate": 1.2499713293007862e-06, + "loss": 0.1745, + "step": 68 + }, + { + "epoch": 0.015699658703071672, + "grad_norm": 30.829215228751778, + "learning_rate": 1.2499704670789301e-06, + "loss": 0.1514, + "step": 69 + }, + { + "epoch": 0.015927189988623434, + "grad_norm": 7.168923083631056, + "learning_rate": 1.2499695920838225e-06, + "loss": 0.2393, + "step": 70 + }, + { + "epoch": 0.0161547212741752, + "grad_norm": 3.418723817035884, + "learning_rate": 1.2499687043154809e-06, + "loss": 0.1342, + "step": 71 + }, + { + "epoch": 0.016382252559726963, + "grad_norm": 6.316537441364383, + "learning_rate": 1.2499678037739235e-06, + "loss": 0.1698, + "step": 72 + }, + { + "epoch": 0.016609783845278725, + "grad_norm": 3.8561981086650596, + "learning_rate": 1.2499668904591688e-06, + "loss": 0.3104, + "step": 73 + }, + { + "epoch": 0.01683731513083049, + "grad_norm": 4.679806938064617, + "learning_rate": 1.2499659643712356e-06, + "loss": 0.2139, + "step": 74 + }, + { + "epoch": 0.017064846416382253, + "grad_norm": 4.26137230837329, + "learning_rate": 1.2499650255101425e-06, + "loss": 0.2433, + "step": 75 + }, + { + "epoch": 0.017292377701934016, + "grad_norm": 3.7227188471827914, + "learning_rate": 1.2499640738759088e-06, + "loss": 0.2334, + "step": 76 + }, + { + "epoch": 0.017519908987485778, + "grad_norm": 6.044525591826923, + "learning_rate": 1.249963109468554e-06, + "loss": 0.3106, + "step": 77 + }, + { + "epoch": 0.017747440273037544, + "grad_norm": 6.248705646938244, + "learning_rate": 1.2499621322880979e-06, + "loss": 0.2025, + "step": 78 + }, + { + "epoch": 0.017974971558589306, + "grad_norm": 2.8368621495357313, + "learning_rate": 1.2499611423345604e-06, + "loss": 0.1492, + "step": 79 + }, + { + "epoch": 0.01820250284414107, + "grad_norm": 5.049736361542706, + "learning_rate": 1.2499601396079617e-06, + "loss": 0.1341, + "step": 80 + }, + { + "epoch": 0.018430034129692834, + "grad_norm": 6.760221850362585, + "learning_rate": 1.2499591241083222e-06, + "loss": 0.2092, + "step": 81 + }, + { + "epoch": 0.018657565415244597, + "grad_norm": 6.630540720646431, + "learning_rate": 1.2499580958356628e-06, + "loss": 0.2181, + "step": 82 + }, + { + "epoch": 0.01888509670079636, + "grad_norm": 3.8482585047631863, + "learning_rate": 1.2499570547900045e-06, + "loss": 0.1613, + "step": 83 + }, + { + "epoch": 0.01911262798634812, + "grad_norm": 6.605304588968454, + "learning_rate": 1.2499560009713684e-06, + "loss": 0.2959, + "step": 84 + }, + { + "epoch": 0.019340159271899887, + "grad_norm": 6.012809221970948, + "learning_rate": 1.2499549343797764e-06, + "loss": 0.2393, + "step": 85 + }, + { + "epoch": 0.01956769055745165, + "grad_norm": 6.254621323206641, + "learning_rate": 1.24995385501525e-06, + "loss": 0.2285, + "step": 86 + }, + { + "epoch": 0.019795221843003412, + "grad_norm": 3.4046999226542733, + "learning_rate": 1.2499527628778116e-06, + "loss": 0.1187, + "step": 87 + }, + { + "epoch": 0.020022753128555178, + "grad_norm": 7.419781715158706, + "learning_rate": 1.2499516579674831e-06, + "loss": 0.2817, + "step": 88 + }, + { + "epoch": 0.02025028441410694, + "grad_norm": 21.819719933471735, + "learning_rate": 1.2499505402842872e-06, + "loss": 0.2469, + "step": 89 + }, + { + "epoch": 0.020477815699658702, + "grad_norm": 2.8418419055080766, + "learning_rate": 1.2499494098282469e-06, + "loss": 0.2955, + "step": 90 + }, + { + "epoch": 0.020705346985210465, + "grad_norm": 7.066317637431583, + "learning_rate": 1.2499482665993851e-06, + "loss": 0.2044, + "step": 91 + }, + { + "epoch": 0.02093287827076223, + "grad_norm": 5.925737098985834, + "learning_rate": 1.2499471105977252e-06, + "loss": 0.2335, + "step": 92 + }, + { + "epoch": 0.021160409556313993, + "grad_norm": 3.0480275776898473, + "learning_rate": 1.249945941823291e-06, + "loss": 0.3633, + "step": 93 + }, + { + "epoch": 0.021387940841865755, + "grad_norm": 2.946352549362824, + "learning_rate": 1.2499447602761063e-06, + "loss": 0.2011, + "step": 94 + }, + { + "epoch": 0.02161547212741752, + "grad_norm": 6.07129225638081, + "learning_rate": 1.2499435659561954e-06, + "loss": 0.2585, + "step": 95 + }, + { + "epoch": 0.021843003412969283, + "grad_norm": 4.592794032374342, + "learning_rate": 1.2499423588635823e-06, + "loss": 0.2336, + "step": 96 + }, + { + "epoch": 0.022070534698521046, + "grad_norm": 19.61835193566366, + "learning_rate": 1.2499411389982919e-06, + "loss": 0.2438, + "step": 97 + }, + { + "epoch": 0.02229806598407281, + "grad_norm": 4.697964666160796, + "learning_rate": 1.2499399063603492e-06, + "loss": 0.26, + "step": 98 + }, + { + "epoch": 0.022525597269624574, + "grad_norm": 6.831528796415563, + "learning_rate": 1.2499386609497793e-06, + "loss": 0.1291, + "step": 99 + }, + { + "epoch": 0.022753128555176336, + "grad_norm": 3.3770537551655653, + "learning_rate": 1.2499374027666078e-06, + "loss": 0.1919, + "step": 100 + }, + { + "epoch": 0.0229806598407281, + "grad_norm": 10.54402988548413, + "learning_rate": 1.2499361318108602e-06, + "loss": 0.2695, + "step": 101 + }, + { + "epoch": 0.023208191126279865, + "grad_norm": 6.4464740357818116, + "learning_rate": 1.2499348480825627e-06, + "loss": 0.1883, + "step": 102 + }, + { + "epoch": 0.023435722411831627, + "grad_norm": 5.7228283849137895, + "learning_rate": 1.2499335515817413e-06, + "loss": 0.225, + "step": 103 + }, + { + "epoch": 0.02366325369738339, + "grad_norm": 8.575195167369158, + "learning_rate": 1.2499322423084226e-06, + "loss": 0.1988, + "step": 104 + }, + { + "epoch": 0.023890784982935155, + "grad_norm": 5.524822469569831, + "learning_rate": 1.2499309202626336e-06, + "loss": 0.1362, + "step": 105 + }, + { + "epoch": 0.024118316268486917, + "grad_norm": 1.4259194554286314, + "learning_rate": 1.249929585444401e-06, + "loss": 0.1341, + "step": 106 + }, + { + "epoch": 0.02434584755403868, + "grad_norm": 5.569399731315438, + "learning_rate": 1.2499282378537522e-06, + "loss": 0.1823, + "step": 107 + }, + { + "epoch": 0.024573378839590442, + "grad_norm": 5.131038290322419, + "learning_rate": 1.2499268774907144e-06, + "loss": 0.1674, + "step": 108 + }, + { + "epoch": 0.024800910125142208, + "grad_norm": 2.9740215362829368, + "learning_rate": 1.249925504355316e-06, + "loss": 0.1443, + "step": 109 + }, + { + "epoch": 0.02502844141069397, + "grad_norm": 7.125610878241638, + "learning_rate": 1.2499241184475848e-06, + "loss": 0.1993, + "step": 110 + }, + { + "epoch": 0.025255972696245733, + "grad_norm": 3.5104920582246284, + "learning_rate": 1.249922719767549e-06, + "loss": 0.1387, + "step": 111 + }, + { + "epoch": 0.0254835039817975, + "grad_norm": 15.180689323576399, + "learning_rate": 1.2499213083152374e-06, + "loss": 0.1609, + "step": 112 + }, + { + "epoch": 0.02571103526734926, + "grad_norm": 2.6467486780240077, + "learning_rate": 1.2499198840906787e-06, + "loss": 0.0766, + "step": 113 + }, + { + "epoch": 0.025938566552901023, + "grad_norm": 6.947833673299234, + "learning_rate": 1.249918447093902e-06, + "loss": 0.1988, + "step": 114 + }, + { + "epoch": 0.026166097838452786, + "grad_norm": 3.236155694827761, + "learning_rate": 1.249916997324937e-06, + "loss": 0.2822, + "step": 115 + }, + { + "epoch": 0.02639362912400455, + "grad_norm": 4.424229361394889, + "learning_rate": 1.2499155347838129e-06, + "loss": 0.2639, + "step": 116 + }, + { + "epoch": 0.026621160409556314, + "grad_norm": 6.7125880752306, + "learning_rate": 1.2499140594705596e-06, + "loss": 0.1758, + "step": 117 + }, + { + "epoch": 0.026848691695108076, + "grad_norm": 12.978485247890044, + "learning_rate": 1.2499125713852076e-06, + "loss": 0.2966, + "step": 118 + }, + { + "epoch": 0.027076222980659842, + "grad_norm": 2.4562187666064297, + "learning_rate": 1.2499110705277869e-06, + "loss": 0.1317, + "step": 119 + }, + { + "epoch": 0.027303754266211604, + "grad_norm": 2.450514697648912, + "learning_rate": 1.2499095568983284e-06, + "loss": 0.2491, + "step": 120 + }, + { + "epoch": 0.027531285551763367, + "grad_norm": 2.962900989508568, + "learning_rate": 1.2499080304968634e-06, + "loss": 0.1782, + "step": 121 + }, + { + "epoch": 0.027758816837315133, + "grad_norm": 4.706451675787787, + "learning_rate": 1.2499064913234222e-06, + "loss": 0.2063, + "step": 122 + }, + { + "epoch": 0.027986348122866895, + "grad_norm": 4.848247166198472, + "learning_rate": 1.249904939378037e-06, + "loss": 0.1873, + "step": 123 + }, + { + "epoch": 0.028213879408418657, + "grad_norm": 5.57275566955423, + "learning_rate": 1.2499033746607395e-06, + "loss": 0.2362, + "step": 124 + }, + { + "epoch": 0.02844141069397042, + "grad_norm": 4.528761927217566, + "learning_rate": 1.2499017971715614e-06, + "loss": 0.2686, + "step": 125 + }, + { + "epoch": 0.028668941979522185, + "grad_norm": 7.35859467900191, + "learning_rate": 1.2499002069105348e-06, + "loss": 0.275, + "step": 126 + }, + { + "epoch": 0.028896473265073948, + "grad_norm": 4.494727686955716, + "learning_rate": 1.2498986038776926e-06, + "loss": 0.1759, + "step": 127 + }, + { + "epoch": 0.02912400455062571, + "grad_norm": 7.273216392666622, + "learning_rate": 1.2498969880730671e-06, + "loss": 0.2159, + "step": 128 + }, + { + "epoch": 0.029351535836177476, + "grad_norm": 4.955227920384567, + "learning_rate": 1.249895359496692e-06, + "loss": 0.1888, + "step": 129 + }, + { + "epoch": 0.02957906712172924, + "grad_norm": 6.321445200949685, + "learning_rate": 1.2498937181486e-06, + "loss": 0.3007, + "step": 130 + }, + { + "epoch": 0.029806598407281, + "grad_norm": 2.76312902269676, + "learning_rate": 1.2498920640288248e-06, + "loss": 0.2442, + "step": 131 + }, + { + "epoch": 0.030034129692832763, + "grad_norm": 56.774720129580295, + "learning_rate": 1.2498903971374005e-06, + "loss": 0.223, + "step": 132 + }, + { + "epoch": 0.03026166097838453, + "grad_norm": 3.9468490187056324, + "learning_rate": 1.2498887174743606e-06, + "loss": 0.2504, + "step": 133 + }, + { + "epoch": 0.03048919226393629, + "grad_norm": 3.9118814976883542, + "learning_rate": 1.24988702503974e-06, + "loss": 0.1939, + "step": 134 + }, + { + "epoch": 0.030716723549488054, + "grad_norm": 3.7837188268010506, + "learning_rate": 1.2498853198335728e-06, + "loss": 0.2199, + "step": 135 + }, + { + "epoch": 0.03094425483503982, + "grad_norm": 4.0297942240817175, + "learning_rate": 1.2498836018558942e-06, + "loss": 0.1566, + "step": 136 + }, + { + "epoch": 0.031171786120591582, + "grad_norm": 3.4754550482446698, + "learning_rate": 1.2498818711067392e-06, + "loss": 0.2666, + "step": 137 + }, + { + "epoch": 0.031399317406143344, + "grad_norm": 3.864651244769, + "learning_rate": 1.2498801275861433e-06, + "loss": 0.1173, + "step": 138 + }, + { + "epoch": 0.03162684869169511, + "grad_norm": 8.216814820623972, + "learning_rate": 1.2498783712941418e-06, + "loss": 0.1879, + "step": 139 + }, + { + "epoch": 0.03185437997724687, + "grad_norm": 3.637457358045326, + "learning_rate": 1.2498766022307709e-06, + "loss": 0.2047, + "step": 140 + }, + { + "epoch": 0.032081911262798635, + "grad_norm": 2.58051980801193, + "learning_rate": 1.2498748203960665e-06, + "loss": 0.1008, + "step": 141 + }, + { + "epoch": 0.0323094425483504, + "grad_norm": 3.8775724824241764, + "learning_rate": 1.2498730257900655e-06, + "loss": 0.2042, + "step": 142 + }, + { + "epoch": 0.03253697383390216, + "grad_norm": 5.772591680829651, + "learning_rate": 1.249871218412804e-06, + "loss": 0.2352, + "step": 143 + }, + { + "epoch": 0.032764505119453925, + "grad_norm": 2.210254874393301, + "learning_rate": 1.2498693982643192e-06, + "loss": 0.1803, + "step": 144 + }, + { + "epoch": 0.03299203640500569, + "grad_norm": 6.540771980552272, + "learning_rate": 1.2498675653446485e-06, + "loss": 0.2304, + "step": 145 + }, + { + "epoch": 0.03321956769055745, + "grad_norm": 2.904522388367919, + "learning_rate": 1.249865719653829e-06, + "loss": 0.1707, + "step": 146 + }, + { + "epoch": 0.033447098976109216, + "grad_norm": 9.318986716894935, + "learning_rate": 1.2498638611918985e-06, + "loss": 0.2038, + "step": 147 + }, + { + "epoch": 0.03367463026166098, + "grad_norm": 9.58516027118141, + "learning_rate": 1.249861989958895e-06, + "loss": 0.2357, + "step": 148 + }, + { + "epoch": 0.03390216154721274, + "grad_norm": 3.559770501878285, + "learning_rate": 1.2498601059548572e-06, + "loss": 0.1613, + "step": 149 + }, + { + "epoch": 0.034129692832764506, + "grad_norm": 3.348814329958542, + "learning_rate": 1.2498582091798228e-06, + "loss": 0.2016, + "step": 150 + }, + { + "epoch": 0.034357224118316265, + "grad_norm": 6.375342543891093, + "learning_rate": 1.2498562996338312e-06, + "loss": 0.2231, + "step": 151 + }, + { + "epoch": 0.03458475540386803, + "grad_norm": 7.488809251815451, + "learning_rate": 1.249854377316921e-06, + "loss": 0.1819, + "step": 152 + }, + { + "epoch": 0.0348122866894198, + "grad_norm": 2.508487580474721, + "learning_rate": 1.2498524422291319e-06, + "loss": 0.182, + "step": 153 + }, + { + "epoch": 0.035039817974971556, + "grad_norm": 3.656563964135558, + "learning_rate": 1.2498504943705033e-06, + "loss": 0.165, + "step": 154 + }, + { + "epoch": 0.03526734926052332, + "grad_norm": 2.771070563762278, + "learning_rate": 1.249848533741075e-06, + "loss": 0.2569, + "step": 155 + }, + { + "epoch": 0.03549488054607509, + "grad_norm": 5.610529774003187, + "learning_rate": 1.2498465603408865e-06, + "loss": 0.2873, + "step": 156 + }, + { + "epoch": 0.035722411831626846, + "grad_norm": 3.6657793262286638, + "learning_rate": 1.2498445741699792e-06, + "loss": 0.1086, + "step": 157 + }, + { + "epoch": 0.03594994311717861, + "grad_norm": 11.136381961854878, + "learning_rate": 1.249842575228393e-06, + "loss": 0.1653, + "step": 158 + }, + { + "epoch": 0.03617747440273038, + "grad_norm": 4.607920317694178, + "learning_rate": 1.249840563516169e-06, + "loss": 0.1816, + "step": 159 + }, + { + "epoch": 0.03640500568828214, + "grad_norm": 4.765507333684582, + "learning_rate": 1.249838539033348e-06, + "loss": 0.1735, + "step": 160 + }, + { + "epoch": 0.0366325369738339, + "grad_norm": 3.024559515436515, + "learning_rate": 1.2498365017799715e-06, + "loss": 0.0997, + "step": 161 + }, + { + "epoch": 0.03686006825938567, + "grad_norm": 3.0006086205585594, + "learning_rate": 1.2498344517560815e-06, + "loss": 0.2742, + "step": 162 + }, + { + "epoch": 0.03708759954493743, + "grad_norm": 4.390575337778858, + "learning_rate": 1.2498323889617198e-06, + "loss": 0.2112, + "step": 163 + }, + { + "epoch": 0.03731513083048919, + "grad_norm": 4.987032274568943, + "learning_rate": 1.2498303133969281e-06, + "loss": 0.2282, + "step": 164 + }, + { + "epoch": 0.03754266211604096, + "grad_norm": 3.813775711394782, + "learning_rate": 1.2498282250617492e-06, + "loss": 0.1944, + "step": 165 + }, + { + "epoch": 0.03777019340159272, + "grad_norm": 3.361678763128891, + "learning_rate": 1.2498261239562257e-06, + "loss": 0.2018, + "step": 166 + }, + { + "epoch": 0.037997724687144484, + "grad_norm": 4.992072192203259, + "learning_rate": 1.2498240100804005e-06, + "loss": 0.2089, + "step": 167 + }, + { + "epoch": 0.03822525597269624, + "grad_norm": 8.050790934059092, + "learning_rate": 1.249821883434317e-06, + "loss": 0.2696, + "step": 168 + }, + { + "epoch": 0.03845278725824801, + "grad_norm": 2.642297340192281, + "learning_rate": 1.2498197440180182e-06, + "loss": 0.2691, + "step": 169 + }, + { + "epoch": 0.038680318543799774, + "grad_norm": 3.35790306734272, + "learning_rate": 1.2498175918315484e-06, + "loss": 0.1851, + "step": 170 + }, + { + "epoch": 0.03890784982935153, + "grad_norm": 3.524642269348137, + "learning_rate": 1.2498154268749513e-06, + "loss": 0.2276, + "step": 171 + }, + { + "epoch": 0.0391353811149033, + "grad_norm": 2.188667506818875, + "learning_rate": 1.249813249148271e-06, + "loss": 0.1616, + "step": 172 + }, + { + "epoch": 0.039362912400455065, + "grad_norm": 5.1958946099491845, + "learning_rate": 1.2498110586515525e-06, + "loss": 0.1987, + "step": 173 + }, + { + "epoch": 0.039590443686006824, + "grad_norm": 5.09328084896296, + "learning_rate": 1.2498088553848398e-06, + "loss": 0.195, + "step": 174 + }, + { + "epoch": 0.03981797497155859, + "grad_norm": 2.8290595777512952, + "learning_rate": 1.2498066393481787e-06, + "loss": 0.1568, + "step": 175 + }, + { + "epoch": 0.040045506257110355, + "grad_norm": 2.360697357040943, + "learning_rate": 1.249804410541614e-06, + "loss": 0.2065, + "step": 176 + }, + { + "epoch": 0.040273037542662114, + "grad_norm": 4.718810327826489, + "learning_rate": 1.2498021689651916e-06, + "loss": 0.2003, + "step": 177 + }, + { + "epoch": 0.04050056882821388, + "grad_norm": 2.6458436624930237, + "learning_rate": 1.249799914618957e-06, + "loss": 0.1589, + "step": 178 + }, + { + "epoch": 0.040728100113765646, + "grad_norm": 3.289621635927127, + "learning_rate": 1.2497976475029566e-06, + "loss": 0.1905, + "step": 179 + }, + { + "epoch": 0.040955631399317405, + "grad_norm": 2.7547654896260028, + "learning_rate": 1.2497953676172364e-06, + "loss": 0.1538, + "step": 180 + }, + { + "epoch": 0.04118316268486917, + "grad_norm": 4.715970073162376, + "learning_rate": 1.2497930749618431e-06, + "loss": 0.1297, + "step": 181 + }, + { + "epoch": 0.04141069397042093, + "grad_norm": 13.147614048372157, + "learning_rate": 1.2497907695368238e-06, + "loss": 0.164, + "step": 182 + }, + { + "epoch": 0.041638225255972695, + "grad_norm": 2.692225418023433, + "learning_rate": 1.2497884513422253e-06, + "loss": 0.2537, + "step": 183 + }, + { + "epoch": 0.04186575654152446, + "grad_norm": 5.166049507007355, + "learning_rate": 1.249786120378095e-06, + "loss": 0.074, + "step": 184 + }, + { + "epoch": 0.04209328782707622, + "grad_norm": 3.0648916024092596, + "learning_rate": 1.2497837766444806e-06, + "loss": 0.1639, + "step": 185 + }, + { + "epoch": 0.042320819112627986, + "grad_norm": 4.567688921451397, + "learning_rate": 1.2497814201414304e-06, + "loss": 0.2905, + "step": 186 + }, + { + "epoch": 0.04254835039817975, + "grad_norm": 3.970377559361967, + "learning_rate": 1.249779050868992e-06, + "loss": 0.2001, + "step": 187 + }, + { + "epoch": 0.04277588168373151, + "grad_norm": 2.2768846909587763, + "learning_rate": 1.249776668827214e-06, + "loss": 0.0951, + "step": 188 + }, + { + "epoch": 0.043003412969283276, + "grad_norm": 6.438142708090974, + "learning_rate": 1.249774274016145e-06, + "loss": 0.203, + "step": 189 + }, + { + "epoch": 0.04323094425483504, + "grad_norm": 2.4175466744317977, + "learning_rate": 1.2497718664358341e-06, + "loss": 0.1713, + "step": 190 + }, + { + "epoch": 0.0434584755403868, + "grad_norm": 4.37204480901975, + "learning_rate": 1.2497694460863307e-06, + "loss": 0.2986, + "step": 191 + }, + { + "epoch": 0.04368600682593857, + "grad_norm": 3.2046762676937255, + "learning_rate": 1.2497670129676838e-06, + "loss": 0.1288, + "step": 192 + }, + { + "epoch": 0.04391353811149033, + "grad_norm": 3.901472238917995, + "learning_rate": 1.2497645670799436e-06, + "loss": 0.1291, + "step": 193 + }, + { + "epoch": 0.04414106939704209, + "grad_norm": 3.891177273974114, + "learning_rate": 1.2497621084231595e-06, + "loss": 0.1165, + "step": 194 + }, + { + "epoch": 0.04436860068259386, + "grad_norm": 3.831124951630966, + "learning_rate": 1.2497596369973823e-06, + "loss": 0.175, + "step": 195 + }, + { + "epoch": 0.04459613196814562, + "grad_norm": 7.137497588920377, + "learning_rate": 1.2497571528026623e-06, + "loss": 0.2319, + "step": 196 + }, + { + "epoch": 0.04482366325369738, + "grad_norm": 2.9787063992991256, + "learning_rate": 1.2497546558390503e-06, + "loss": 0.2044, + "step": 197 + }, + { + "epoch": 0.04505119453924915, + "grad_norm": 2.5728244375494413, + "learning_rate": 1.2497521461065973e-06, + "loss": 0.1395, + "step": 198 + }, + { + "epoch": 0.04527872582480091, + "grad_norm": 7.102221321561537, + "learning_rate": 1.2497496236053547e-06, + "loss": 0.1969, + "step": 199 + }, + { + "epoch": 0.04550625711035267, + "grad_norm": 2.579422809989494, + "learning_rate": 1.2497470883353738e-06, + "loss": 0.1019, + "step": 200 + }, + { + "epoch": 0.04573378839590444, + "grad_norm": 4.340132040430137, + "learning_rate": 1.2497445402967068e-06, + "loss": 0.241, + "step": 201 + }, + { + "epoch": 0.0459613196814562, + "grad_norm": 2.2195665044126276, + "learning_rate": 1.2497419794894053e-06, + "loss": 0.2059, + "step": 202 + }, + { + "epoch": 0.04618885096700796, + "grad_norm": 3.274345001247324, + "learning_rate": 1.249739405913522e-06, + "loss": 0.1328, + "step": 203 + }, + { + "epoch": 0.04641638225255973, + "grad_norm": 2.527264534705696, + "learning_rate": 1.2497368195691095e-06, + "loss": 0.1408, + "step": 204 + }, + { + "epoch": 0.04664391353811149, + "grad_norm": 3.306757570747259, + "learning_rate": 1.2497342204562205e-06, + "loss": 0.2233, + "step": 205 + }, + { + "epoch": 0.046871444823663254, + "grad_norm": 3.6647451852915336, + "learning_rate": 1.2497316085749081e-06, + "loss": 0.1239, + "step": 206 + }, + { + "epoch": 0.04709897610921502, + "grad_norm": 4.68508784917087, + "learning_rate": 1.249728983925226e-06, + "loss": 0.1707, + "step": 207 + }, + { + "epoch": 0.04732650739476678, + "grad_norm": 3.18438034976801, + "learning_rate": 1.2497263465072274e-06, + "loss": 0.1325, + "step": 208 + }, + { + "epoch": 0.047554038680318544, + "grad_norm": 2.665536371480516, + "learning_rate": 1.2497236963209663e-06, + "loss": 0.247, + "step": 209 + }, + { + "epoch": 0.04778156996587031, + "grad_norm": 3.6305897675111822, + "learning_rate": 1.2497210333664972e-06, + "loss": 0.1399, + "step": 210 + }, + { + "epoch": 0.04800910125142207, + "grad_norm": 3.427786312260657, + "learning_rate": 1.2497183576438743e-06, + "loss": 0.1595, + "step": 211 + }, + { + "epoch": 0.048236632536973835, + "grad_norm": 3.501593030667954, + "learning_rate": 1.2497156691531523e-06, + "loss": 0.1895, + "step": 212 + }, + { + "epoch": 0.048464163822525594, + "grad_norm": 2.29399983953313, + "learning_rate": 1.249712967894386e-06, + "loss": 0.1273, + "step": 213 + }, + { + "epoch": 0.04869169510807736, + "grad_norm": 4.248497703608046, + "learning_rate": 1.2497102538676308e-06, + "loss": 0.2118, + "step": 214 + }, + { + "epoch": 0.048919226393629126, + "grad_norm": 5.009911727752511, + "learning_rate": 1.249707527072942e-06, + "loss": 0.1533, + "step": 215 + }, + { + "epoch": 0.049146757679180884, + "grad_norm": 3.254064879259487, + "learning_rate": 1.2497047875103757e-06, + "loss": 0.3042, + "step": 216 + }, + { + "epoch": 0.04937428896473265, + "grad_norm": 2.700363753095535, + "learning_rate": 1.2497020351799875e-06, + "loss": 0.1933, + "step": 217 + }, + { + "epoch": 0.049601820250284416, + "grad_norm": 2.2159854350533763, + "learning_rate": 1.2496992700818335e-06, + "loss": 0.1733, + "step": 218 + }, + { + "epoch": 0.049829351535836175, + "grad_norm": 6.438623712108173, + "learning_rate": 1.249696492215971e-06, + "loss": 0.2233, + "step": 219 + }, + { + "epoch": 0.05005688282138794, + "grad_norm": 3.6403163135182552, + "learning_rate": 1.249693701582456e-06, + "loss": 0.1542, + "step": 220 + }, + { + "epoch": 0.05028441410693971, + "grad_norm": 3.280631643810882, + "learning_rate": 1.2496908981813458e-06, + "loss": 0.1799, + "step": 221 + }, + { + "epoch": 0.050511945392491465, + "grad_norm": 2.5684306853319687, + "learning_rate": 1.2496880820126977e-06, + "loss": 0.2051, + "step": 222 + }, + { + "epoch": 0.05073947667804323, + "grad_norm": 2.7401430199461108, + "learning_rate": 1.2496852530765695e-06, + "loss": 0.1828, + "step": 223 + }, + { + "epoch": 0.050967007963595, + "grad_norm": 2.95485123311806, + "learning_rate": 1.2496824113730186e-06, + "loss": 0.2602, + "step": 224 + }, + { + "epoch": 0.051194539249146756, + "grad_norm": 2.5679914292312738, + "learning_rate": 1.2496795569021033e-06, + "loss": 0.1838, + "step": 225 + }, + { + "epoch": 0.05142207053469852, + "grad_norm": 4.2106953289503055, + "learning_rate": 1.2496766896638819e-06, + "loss": 0.1831, + "step": 226 + }, + { + "epoch": 0.05164960182025029, + "grad_norm": 2.4133590857510603, + "learning_rate": 1.249673809658413e-06, + "loss": 0.1869, + "step": 227 + }, + { + "epoch": 0.05187713310580205, + "grad_norm": 2.009672236932174, + "learning_rate": 1.2496709168857555e-06, + "loss": 0.1297, + "step": 228 + }, + { + "epoch": 0.05210466439135381, + "grad_norm": 2.57569428799923, + "learning_rate": 1.2496680113459683e-06, + "loss": 0.1887, + "step": 229 + }, + { + "epoch": 0.05233219567690557, + "grad_norm": 3.3094428680937464, + "learning_rate": 1.2496650930391113e-06, + "loss": 0.2654, + "step": 230 + }, + { + "epoch": 0.05255972696245734, + "grad_norm": 2.847650693015463, + "learning_rate": 1.2496621619652435e-06, + "loss": 0.1704, + "step": 231 + }, + { + "epoch": 0.0527872582480091, + "grad_norm": 2.9888611972362167, + "learning_rate": 1.2496592181244253e-06, + "loss": 0.1601, + "step": 232 + }, + { + "epoch": 0.05301478953356086, + "grad_norm": 2.08648737949565, + "learning_rate": 1.249656261516717e-06, + "loss": 0.1953, + "step": 233 + }, + { + "epoch": 0.05324232081911263, + "grad_norm": 2.531082669247976, + "learning_rate": 1.2496532921421781e-06, + "loss": 0.1717, + "step": 234 + }, + { + "epoch": 0.053469852104664393, + "grad_norm": 2.7509933573597896, + "learning_rate": 1.2496503100008704e-06, + "loss": 0.2469, + "step": 235 + }, + { + "epoch": 0.05369738339021615, + "grad_norm": 3.5155091690123923, + "learning_rate": 1.249647315092854e-06, + "loss": 0.1314, + "step": 236 + }, + { + "epoch": 0.05392491467576792, + "grad_norm": 3.2336581137529135, + "learning_rate": 1.2496443074181905e-06, + "loss": 0.1479, + "step": 237 + }, + { + "epoch": 0.054152445961319684, + "grad_norm": 1.9727228995954271, + "learning_rate": 1.2496412869769415e-06, + "loss": 0.1072, + "step": 238 + }, + { + "epoch": 0.05437997724687144, + "grad_norm": 9.030280638699303, + "learning_rate": 1.2496382537691686e-06, + "loss": 0.1993, + "step": 239 + }, + { + "epoch": 0.05460750853242321, + "grad_norm": 2.012237999972146, + "learning_rate": 1.2496352077949336e-06, + "loss": 0.2021, + "step": 240 + }, + { + "epoch": 0.054835039817974975, + "grad_norm": 2.875480352440569, + "learning_rate": 1.249632149054299e-06, + "loss": 0.1071, + "step": 241 + }, + { + "epoch": 0.05506257110352673, + "grad_norm": 3.027078266755971, + "learning_rate": 1.249629077547327e-06, + "loss": 0.2081, + "step": 242 + }, + { + "epoch": 0.0552901023890785, + "grad_norm": 3.212706521917931, + "learning_rate": 1.2496259932740813e-06, + "loss": 0.235, + "step": 243 + }, + { + "epoch": 0.055517633674630265, + "grad_norm": 1.5899391805286471, + "learning_rate": 1.2496228962346236e-06, + "loss": 0.1498, + "step": 244 + }, + { + "epoch": 0.055745164960182024, + "grad_norm": 2.252897408154709, + "learning_rate": 1.249619786429018e-06, + "loss": 0.0875, + "step": 245 + }, + { + "epoch": 0.05597269624573379, + "grad_norm": 1.7851217439709355, + "learning_rate": 1.2496166638573278e-06, + "loss": 0.163, + "step": 246 + }, + { + "epoch": 0.05620022753128555, + "grad_norm": 4.076208180076855, + "learning_rate": 1.2496135285196172e-06, + "loss": 0.1298, + "step": 247 + }, + { + "epoch": 0.056427758816837315, + "grad_norm": 8.235783447081577, + "learning_rate": 1.2496103804159497e-06, + "loss": 0.1994, + "step": 248 + }, + { + "epoch": 0.05665529010238908, + "grad_norm": 4.224863516307238, + "learning_rate": 1.2496072195463904e-06, + "loss": 0.1917, + "step": 249 + }, + { + "epoch": 0.05688282138794084, + "grad_norm": 2.600108393969465, + "learning_rate": 1.249604045911003e-06, + "loss": 0.1728, + "step": 250 + }, + { + "epoch": 0.057110352673492605, + "grad_norm": 4.193154020881599, + "learning_rate": 1.249600859509853e-06, + "loss": 0.1469, + "step": 251 + }, + { + "epoch": 0.05733788395904437, + "grad_norm": 3.3023049454358957, + "learning_rate": 1.2495976603430054e-06, + "loss": 0.3015, + "step": 252 + }, + { + "epoch": 0.05756541524459613, + "grad_norm": 2.1335803404002815, + "learning_rate": 1.2495944484105254e-06, + "loss": 0.1237, + "step": 253 + }, + { + "epoch": 0.057792946530147896, + "grad_norm": 5.342229724882705, + "learning_rate": 1.2495912237124787e-06, + "loss": 0.1134, + "step": 254 + }, + { + "epoch": 0.05802047781569966, + "grad_norm": 4.8799722775641765, + "learning_rate": 1.2495879862489312e-06, + "loss": 0.1865, + "step": 255 + }, + { + "epoch": 0.05824800910125142, + "grad_norm": 5.731543371657422, + "learning_rate": 1.2495847360199495e-06, + "loss": 0.2008, + "step": 256 + }, + { + "epoch": 0.058475540386803186, + "grad_norm": 2.313924736001694, + "learning_rate": 1.2495814730255993e-06, + "loss": 0.1361, + "step": 257 + }, + { + "epoch": 0.05870307167235495, + "grad_norm": 1.3942403935107488, + "learning_rate": 1.2495781972659479e-06, + "loss": 0.1103, + "step": 258 + }, + { + "epoch": 0.05893060295790671, + "grad_norm": 1.8635600367271647, + "learning_rate": 1.2495749087410618e-06, + "loss": 0.1736, + "step": 259 + }, + { + "epoch": 0.05915813424345848, + "grad_norm": 3.934800507138662, + "learning_rate": 1.2495716074510087e-06, + "loss": 0.1706, + "step": 260 + }, + { + "epoch": 0.059385665529010236, + "grad_norm": 7.067913001607123, + "learning_rate": 1.2495682933958555e-06, + "loss": 0.1963, + "step": 261 + }, + { + "epoch": 0.059613196814562, + "grad_norm": 2.692944909371077, + "learning_rate": 1.2495649665756705e-06, + "loss": 0.2486, + "step": 262 + }, + { + "epoch": 0.05984072810011377, + "grad_norm": 2.4930462253175305, + "learning_rate": 1.2495616269905212e-06, + "loss": 0.1447, + "step": 263 + }, + { + "epoch": 0.060068259385665526, + "grad_norm": 1.7948148568482771, + "learning_rate": 1.2495582746404762e-06, + "loss": 0.0994, + "step": 264 + }, + { + "epoch": 0.06029579067121729, + "grad_norm": 2.021876252112372, + "learning_rate": 1.249554909525604e-06, + "loss": 0.1386, + "step": 265 + }, + { + "epoch": 0.06052332195676906, + "grad_norm": 2.069960058640526, + "learning_rate": 1.249551531645973e-06, + "loss": 0.1866, + "step": 266 + }, + { + "epoch": 0.06075085324232082, + "grad_norm": 8.549797598789278, + "learning_rate": 1.2495481410016527e-06, + "loss": 0.3426, + "step": 267 + }, + { + "epoch": 0.06097838452787258, + "grad_norm": 6.033524800668443, + "learning_rate": 1.2495447375927122e-06, + "loss": 0.2039, + "step": 268 + }, + { + "epoch": 0.06120591581342435, + "grad_norm": 3.3984019223631656, + "learning_rate": 1.2495413214192209e-06, + "loss": 0.1562, + "step": 269 + }, + { + "epoch": 0.06143344709897611, + "grad_norm": 2.78909231360363, + "learning_rate": 1.2495378924812486e-06, + "loss": 0.2056, + "step": 270 + }, + { + "epoch": 0.06166097838452787, + "grad_norm": 5.781877877875473, + "learning_rate": 1.2495344507788662e-06, + "loss": 0.2293, + "step": 271 + }, + { + "epoch": 0.06188850967007964, + "grad_norm": 2.3180826263300607, + "learning_rate": 1.249530996312143e-06, + "loss": 0.1489, + "step": 272 + }, + { + "epoch": 0.0621160409556314, + "grad_norm": 7.2617460886104475, + "learning_rate": 1.2495275290811499e-06, + "loss": 0.2172, + "step": 273 + }, + { + "epoch": 0.062343572241183164, + "grad_norm": 2.1316035699431173, + "learning_rate": 1.2495240490859581e-06, + "loss": 0.2176, + "step": 274 + }, + { + "epoch": 0.06257110352673492, + "grad_norm": 2.5542857532037235, + "learning_rate": 1.2495205563266384e-06, + "loss": 0.1521, + "step": 275 + }, + { + "epoch": 0.06279863481228669, + "grad_norm": 3.5696131149812644, + "learning_rate": 1.2495170508032624e-06, + "loss": 0.2817, + "step": 276 + }, + { + "epoch": 0.06302616609783845, + "grad_norm": 4.055804927691344, + "learning_rate": 1.2495135325159015e-06, + "loss": 0.1484, + "step": 277 + }, + { + "epoch": 0.06325369738339022, + "grad_norm": 2.830287596995614, + "learning_rate": 1.2495100014646277e-06, + "loss": 0.1714, + "step": 278 + }, + { + "epoch": 0.06348122866894199, + "grad_norm": 5.2323794095215685, + "learning_rate": 1.2495064576495134e-06, + "loss": 0.3121, + "step": 279 + }, + { + "epoch": 0.06370875995449374, + "grad_norm": 2.500465425444752, + "learning_rate": 1.2495029010706306e-06, + "loss": 0.1005, + "step": 280 + }, + { + "epoch": 0.0639362912400455, + "grad_norm": 2.7474098845449433, + "learning_rate": 1.2494993317280524e-06, + "loss": 0.1755, + "step": 281 + }, + { + "epoch": 0.06416382252559727, + "grad_norm": 3.1110646620479967, + "learning_rate": 1.2494957496218516e-06, + "loss": 0.194, + "step": 282 + }, + { + "epoch": 0.06439135381114904, + "grad_norm": 1.162926170243262, + "learning_rate": 1.2494921547521013e-06, + "loss": 0.1667, + "step": 283 + }, + { + "epoch": 0.0646188850967008, + "grad_norm": 2.034958588386092, + "learning_rate": 1.249488547118875e-06, + "loss": 0.1031, + "step": 284 + }, + { + "epoch": 0.06484641638225255, + "grad_norm": 2.8585727096596214, + "learning_rate": 1.2494849267222466e-06, + "loss": 0.1199, + "step": 285 + }, + { + "epoch": 0.06507394766780432, + "grad_norm": 2.3756686418598916, + "learning_rate": 1.24948129356229e-06, + "loss": 0.203, + "step": 286 + }, + { + "epoch": 0.06530147895335608, + "grad_norm": 6.080154909085321, + "learning_rate": 1.2494776476390793e-06, + "loss": 0.2723, + "step": 287 + }, + { + "epoch": 0.06552901023890785, + "grad_norm": 3.1578927707769684, + "learning_rate": 1.2494739889526894e-06, + "loss": 0.1218, + "step": 288 + }, + { + "epoch": 0.06575654152445962, + "grad_norm": 2.7745317736308373, + "learning_rate": 1.2494703175031946e-06, + "loss": 0.194, + "step": 289 + }, + { + "epoch": 0.06598407281001138, + "grad_norm": 2.872306438815133, + "learning_rate": 1.2494666332906702e-06, + "loss": 0.143, + "step": 290 + }, + { + "epoch": 0.06621160409556313, + "grad_norm": 2.2661659384858277, + "learning_rate": 1.2494629363151916e-06, + "loss": 0.1497, + "step": 291 + }, + { + "epoch": 0.0664391353811149, + "grad_norm": 2.7978250826969586, + "learning_rate": 1.2494592265768343e-06, + "loss": 0.1817, + "step": 292 + }, + { + "epoch": 0.06666666666666667, + "grad_norm": 2.9435086338480496, + "learning_rate": 1.2494555040756737e-06, + "loss": 0.1195, + "step": 293 + }, + { + "epoch": 0.06689419795221843, + "grad_norm": 2.525871560805257, + "learning_rate": 1.2494517688117867e-06, + "loss": 0.2054, + "step": 294 + }, + { + "epoch": 0.0671217292377702, + "grad_norm": 3.3530486331117126, + "learning_rate": 1.2494480207852489e-06, + "loss": 0.1186, + "step": 295 + }, + { + "epoch": 0.06734926052332196, + "grad_norm": 3.791549905681902, + "learning_rate": 1.249444259996137e-06, + "loss": 0.1616, + "step": 296 + }, + { + "epoch": 0.06757679180887372, + "grad_norm": 2.3603348366809236, + "learning_rate": 1.2494404864445284e-06, + "loss": 0.1392, + "step": 297 + }, + { + "epoch": 0.06780432309442548, + "grad_norm": 2.161901751847752, + "learning_rate": 1.2494367001304996e-06, + "loss": 0.1548, + "step": 298 + }, + { + "epoch": 0.06803185437997725, + "grad_norm": 2.3978175716297634, + "learning_rate": 1.2494329010541284e-06, + "loss": 0.1634, + "step": 299 + }, + { + "epoch": 0.06825938566552901, + "grad_norm": 5.413503442113624, + "learning_rate": 1.2494290892154922e-06, + "loss": 0.2876, + "step": 300 + }, + { + "epoch": 0.06848691695108078, + "grad_norm": 1.904095426332445, + "learning_rate": 1.2494252646146692e-06, + "loss": 0.1942, + "step": 301 + }, + { + "epoch": 0.06871444823663253, + "grad_norm": 2.0091735504190504, + "learning_rate": 1.249421427251737e-06, + "loss": 0.1403, + "step": 302 + }, + { + "epoch": 0.0689419795221843, + "grad_norm": 2.6001586830103123, + "learning_rate": 1.2494175771267748e-06, + "loss": 0.2376, + "step": 303 + }, + { + "epoch": 0.06916951080773606, + "grad_norm": 2.8009063420794265, + "learning_rate": 1.2494137142398607e-06, + "loss": 0.1877, + "step": 304 + }, + { + "epoch": 0.06939704209328783, + "grad_norm": 2.0648464255318517, + "learning_rate": 1.249409838591074e-06, + "loss": 0.1462, + "step": 305 + }, + { + "epoch": 0.0696245733788396, + "grad_norm": 2.6396516124770657, + "learning_rate": 1.2494059501804937e-06, + "loss": 0.256, + "step": 306 + }, + { + "epoch": 0.06985210466439136, + "grad_norm": 2.9901343092043837, + "learning_rate": 1.249402049008199e-06, + "loss": 0.1483, + "step": 307 + }, + { + "epoch": 0.07007963594994311, + "grad_norm": 3.0343546498099356, + "learning_rate": 1.2493981350742704e-06, + "loss": 0.1561, + "step": 308 + }, + { + "epoch": 0.07030716723549488, + "grad_norm": 3.2148889672864636, + "learning_rate": 1.2493942083787872e-06, + "loss": 0.1856, + "step": 309 + }, + { + "epoch": 0.07053469852104664, + "grad_norm": 2.795539793994042, + "learning_rate": 1.2493902689218299e-06, + "loss": 0.1294, + "step": 310 + }, + { + "epoch": 0.07076222980659841, + "grad_norm": 2.1866434219410307, + "learning_rate": 1.249386316703479e-06, + "loss": 0.1789, + "step": 311 + }, + { + "epoch": 0.07098976109215017, + "grad_norm": 4.93386744278198, + "learning_rate": 1.2493823517238154e-06, + "loss": 0.1529, + "step": 312 + }, + { + "epoch": 0.07121729237770194, + "grad_norm": 2.127480030167813, + "learning_rate": 1.2493783739829202e-06, + "loss": 0.1593, + "step": 313 + }, + { + "epoch": 0.07144482366325369, + "grad_norm": 2.565861378561538, + "learning_rate": 1.2493743834808741e-06, + "loss": 0.1442, + "step": 314 + }, + { + "epoch": 0.07167235494880546, + "grad_norm": 3.129314599970171, + "learning_rate": 1.2493703802177594e-06, + "loss": 0.1936, + "step": 315 + }, + { + "epoch": 0.07189988623435722, + "grad_norm": 4.26603531282599, + "learning_rate": 1.2493663641936576e-06, + "loss": 0.1343, + "step": 316 + }, + { + "epoch": 0.07212741751990899, + "grad_norm": 1.778626655821605, + "learning_rate": 1.2493623354086507e-06, + "loss": 0.1751, + "step": 317 + }, + { + "epoch": 0.07235494880546076, + "grad_norm": 2.576979617695665, + "learning_rate": 1.2493582938628213e-06, + "loss": 0.1405, + "step": 318 + }, + { + "epoch": 0.07258248009101251, + "grad_norm": 2.528946823784448, + "learning_rate": 1.2493542395562516e-06, + "loss": 0.1207, + "step": 319 + }, + { + "epoch": 0.07281001137656427, + "grad_norm": 1.7105561186222351, + "learning_rate": 1.2493501724890247e-06, + "loss": 0.1067, + "step": 320 + }, + { + "epoch": 0.07303754266211604, + "grad_norm": 3.0021555230652144, + "learning_rate": 1.249346092661224e-06, + "loss": 0.1769, + "step": 321 + }, + { + "epoch": 0.0732650739476678, + "grad_norm": 3.2473648686733787, + "learning_rate": 1.2493420000729322e-06, + "loss": 0.1797, + "step": 322 + }, + { + "epoch": 0.07349260523321957, + "grad_norm": 2.9141882965376644, + "learning_rate": 1.2493378947242336e-06, + "loss": 0.1936, + "step": 323 + }, + { + "epoch": 0.07372013651877134, + "grad_norm": 2.139000059452357, + "learning_rate": 1.2493337766152119e-06, + "loss": 0.1323, + "step": 324 + }, + { + "epoch": 0.07394766780432309, + "grad_norm": 3.7562365963393773, + "learning_rate": 1.249329645745951e-06, + "loss": 0.1521, + "step": 325 + }, + { + "epoch": 0.07417519908987485, + "grad_norm": 3.1427328506374343, + "learning_rate": 1.2493255021165357e-06, + "loss": 0.1426, + "step": 326 + }, + { + "epoch": 0.07440273037542662, + "grad_norm": 2.5928821859504225, + "learning_rate": 1.2493213457270504e-06, + "loss": 0.1492, + "step": 327 + }, + { + "epoch": 0.07463026166097839, + "grad_norm": 2.6116349350740773, + "learning_rate": 1.2493171765775804e-06, + "loss": 0.1079, + "step": 328 + }, + { + "epoch": 0.07485779294653015, + "grad_norm": 2.5063754100070796, + "learning_rate": 1.2493129946682107e-06, + "loss": 0.1449, + "step": 329 + }, + { + "epoch": 0.07508532423208192, + "grad_norm": 2.7029390289735247, + "learning_rate": 1.2493087999990263e-06, + "loss": 0.2012, + "step": 330 + }, + { + "epoch": 0.07531285551763367, + "grad_norm": 3.168250561710959, + "learning_rate": 1.249304592570114e-06, + "loss": 0.135, + "step": 331 + }, + { + "epoch": 0.07554038680318544, + "grad_norm": 3.358825282989208, + "learning_rate": 1.2493003723815588e-06, + "loss": 0.202, + "step": 332 + }, + { + "epoch": 0.0757679180887372, + "grad_norm": 3.4712230061099367, + "learning_rate": 1.2492961394334474e-06, + "loss": 0.1796, + "step": 333 + }, + { + "epoch": 0.07599544937428897, + "grad_norm": 2.7447934095202586, + "learning_rate": 1.2492918937258663e-06, + "loss": 0.1529, + "step": 334 + }, + { + "epoch": 0.07622298065984073, + "grad_norm": 4.884489478774658, + "learning_rate": 1.2492876352589024e-06, + "loss": 0.1983, + "step": 335 + }, + { + "epoch": 0.07645051194539249, + "grad_norm": 11.840111431867928, + "learning_rate": 1.2492833640326424e-06, + "loss": 0.1701, + "step": 336 + }, + { + "epoch": 0.07667804323094425, + "grad_norm": 3.6493332372043032, + "learning_rate": 1.2492790800471738e-06, + "loss": 0.1894, + "step": 337 + }, + { + "epoch": 0.07690557451649602, + "grad_norm": 2.2273861687776657, + "learning_rate": 1.249274783302584e-06, + "loss": 0.1168, + "step": 338 + }, + { + "epoch": 0.07713310580204778, + "grad_norm": 3.0155968100929016, + "learning_rate": 1.249270473798961e-06, + "loss": 0.1877, + "step": 339 + }, + { + "epoch": 0.07736063708759955, + "grad_norm": 3.6811309004263197, + "learning_rate": 1.249266151536393e-06, + "loss": 0.1841, + "step": 340 + }, + { + "epoch": 0.07758816837315131, + "grad_norm": 3.3318670131929355, + "learning_rate": 1.249261816514968e-06, + "loss": 0.1425, + "step": 341 + }, + { + "epoch": 0.07781569965870307, + "grad_norm": 1.542707864707429, + "learning_rate": 1.2492574687347747e-06, + "loss": 0.0954, + "step": 342 + }, + { + "epoch": 0.07804323094425483, + "grad_norm": 5.219514434003638, + "learning_rate": 1.249253108195902e-06, + "loss": 0.1523, + "step": 343 + }, + { + "epoch": 0.0782707622298066, + "grad_norm": 2.685054702258556, + "learning_rate": 1.249248734898439e-06, + "loss": 0.1932, + "step": 344 + }, + { + "epoch": 0.07849829351535836, + "grad_norm": 3.782143044532345, + "learning_rate": 1.2492443488424753e-06, + "loss": 0.1782, + "step": 345 + }, + { + "epoch": 0.07872582480091013, + "grad_norm": 2.987081909452687, + "learning_rate": 1.2492399500281002e-06, + "loss": 0.1174, + "step": 346 + }, + { + "epoch": 0.07895335608646188, + "grad_norm": 2.4163752446451667, + "learning_rate": 1.2492355384554039e-06, + "loss": 0.1864, + "step": 347 + }, + { + "epoch": 0.07918088737201365, + "grad_norm": 2.881696468020635, + "learning_rate": 1.2492311141244764e-06, + "loss": 0.1509, + "step": 348 + }, + { + "epoch": 0.07940841865756541, + "grad_norm": 4.2425549257036925, + "learning_rate": 1.249226677035408e-06, + "loss": 0.1384, + "step": 349 + }, + { + "epoch": 0.07963594994311718, + "grad_norm": 2.999886291999185, + "learning_rate": 1.2492222271882896e-06, + "loss": 0.1631, + "step": 350 + }, + { + "epoch": 0.07986348122866894, + "grad_norm": 4.681484131322112, + "learning_rate": 1.2492177645832121e-06, + "loss": 0.1752, + "step": 351 + }, + { + "epoch": 0.08009101251422071, + "grad_norm": 2.921704965075288, + "learning_rate": 1.2492132892202668e-06, + "loss": 0.1486, + "step": 352 + }, + { + "epoch": 0.08031854379977246, + "grad_norm": 5.592595582830648, + "learning_rate": 1.2492088010995449e-06, + "loss": 0.2707, + "step": 353 + }, + { + "epoch": 0.08054607508532423, + "grad_norm": 2.9440013961704823, + "learning_rate": 1.2492043002211385e-06, + "loss": 0.2054, + "step": 354 + }, + { + "epoch": 0.080773606370876, + "grad_norm": 2.2221784159000006, + "learning_rate": 1.2491997865851392e-06, + "loss": 0.1373, + "step": 355 + }, + { + "epoch": 0.08100113765642776, + "grad_norm": 1.7381570114572884, + "learning_rate": 1.2491952601916395e-06, + "loss": 0.0858, + "step": 356 + }, + { + "epoch": 0.08122866894197953, + "grad_norm": 2.930524510809462, + "learning_rate": 1.2491907210407319e-06, + "loss": 0.2179, + "step": 357 + }, + { + "epoch": 0.08145620022753129, + "grad_norm": 1.329914120982883, + "learning_rate": 1.249186169132509e-06, + "loss": 0.1839, + "step": 358 + }, + { + "epoch": 0.08168373151308304, + "grad_norm": 4.774637200381304, + "learning_rate": 1.2491816044670641e-06, + "loss": 0.1266, + "step": 359 + }, + { + "epoch": 0.08191126279863481, + "grad_norm": 3.0085506218930442, + "learning_rate": 1.24917702704449e-06, + "loss": 0.1813, + "step": 360 + }, + { + "epoch": 0.08213879408418658, + "grad_norm": 2.683588571853357, + "learning_rate": 1.2491724368648808e-06, + "loss": 0.1182, + "step": 361 + }, + { + "epoch": 0.08236632536973834, + "grad_norm": 4.142859587264675, + "learning_rate": 1.2491678339283303e-06, + "loss": 0.1213, + "step": 362 + }, + { + "epoch": 0.08259385665529011, + "grad_norm": 2.266538556877378, + "learning_rate": 1.249163218234932e-06, + "loss": 0.1669, + "step": 363 + }, + { + "epoch": 0.08282138794084186, + "grad_norm": 3.340308786527698, + "learning_rate": 1.249158589784781e-06, + "loss": 0.1449, + "step": 364 + }, + { + "epoch": 0.08304891922639362, + "grad_norm": 3.600922134824311, + "learning_rate": 1.2491539485779713e-06, + "loss": 0.1934, + "step": 365 + }, + { + "epoch": 0.08327645051194539, + "grad_norm": 2.5603148777390796, + "learning_rate": 1.2491492946145981e-06, + "loss": 0.1215, + "step": 366 + }, + { + "epoch": 0.08350398179749716, + "grad_norm": 1.4306937563740754, + "learning_rate": 1.2491446278947563e-06, + "loss": 0.1218, + "step": 367 + }, + { + "epoch": 0.08373151308304892, + "grad_norm": 6.514691076015768, + "learning_rate": 1.2491399484185413e-06, + "loss": 0.1723, + "step": 368 + }, + { + "epoch": 0.08395904436860069, + "grad_norm": 2.1513333963844214, + "learning_rate": 1.249135256186049e-06, + "loss": 0.242, + "step": 369 + }, + { + "epoch": 0.08418657565415244, + "grad_norm": 1.697947937157404, + "learning_rate": 1.249130551197375e-06, + "loss": 0.1045, + "step": 370 + }, + { + "epoch": 0.0844141069397042, + "grad_norm": 1.4338559958770856, + "learning_rate": 1.2491258334526155e-06, + "loss": 0.1671, + "step": 371 + }, + { + "epoch": 0.08464163822525597, + "grad_norm": 2.7532236684188773, + "learning_rate": 1.2491211029518672e-06, + "loss": 0.1034, + "step": 372 + }, + { + "epoch": 0.08486916951080774, + "grad_norm": 2.665642318134447, + "learning_rate": 1.2491163596952264e-06, + "loss": 0.1737, + "step": 373 + }, + { + "epoch": 0.0850967007963595, + "grad_norm": 1.5130437493435105, + "learning_rate": 1.2491116036827902e-06, + "loss": 0.0804, + "step": 374 + }, + { + "epoch": 0.08532423208191127, + "grad_norm": 1.3642320073282543, + "learning_rate": 1.2491068349146559e-06, + "loss": 0.1428, + "step": 375 + }, + { + "epoch": 0.08555176336746302, + "grad_norm": 2.1006895230964444, + "learning_rate": 1.249102053390921e-06, + "loss": 0.2759, + "step": 376 + }, + { + "epoch": 0.08577929465301479, + "grad_norm": 1.5335225229109515, + "learning_rate": 1.249097259111683e-06, + "loss": 0.1836, + "step": 377 + }, + { + "epoch": 0.08600682593856655, + "grad_norm": 4.09523641946509, + "learning_rate": 1.24909245207704e-06, + "loss": 0.2771, + "step": 378 + }, + { + "epoch": 0.08623435722411832, + "grad_norm": 2.2658393838403477, + "learning_rate": 1.2490876322870904e-06, + "loss": 0.1815, + "step": 379 + }, + { + "epoch": 0.08646188850967008, + "grad_norm": 3.053596441038967, + "learning_rate": 1.2490827997419325e-06, + "loss": 0.1183, + "step": 380 + }, + { + "epoch": 0.08668941979522184, + "grad_norm": 2.9366601199125153, + "learning_rate": 1.249077954441665e-06, + "loss": 0.1738, + "step": 381 + }, + { + "epoch": 0.0869169510807736, + "grad_norm": 1.9726593738442935, + "learning_rate": 1.249073096386387e-06, + "loss": 0.1427, + "step": 382 + }, + { + "epoch": 0.08714448236632537, + "grad_norm": 2.8452874204285985, + "learning_rate": 1.249068225576198e-06, + "loss": 0.2767, + "step": 383 + }, + { + "epoch": 0.08737201365187713, + "grad_norm": 4.292343700500067, + "learning_rate": 1.2490633420111974e-06, + "loss": 0.127, + "step": 384 + }, + { + "epoch": 0.0875995449374289, + "grad_norm": 4.105827667785258, + "learning_rate": 1.249058445691485e-06, + "loss": 0.1639, + "step": 385 + }, + { + "epoch": 0.08782707622298067, + "grad_norm": 4.310698395146462, + "learning_rate": 1.2490535366171607e-06, + "loss": 0.1289, + "step": 386 + }, + { + "epoch": 0.08805460750853242, + "grad_norm": 3.5788743602832795, + "learning_rate": 1.249048614788325e-06, + "loss": 0.1804, + "step": 387 + }, + { + "epoch": 0.08828213879408418, + "grad_norm": 2.6616942664445413, + "learning_rate": 1.249043680205079e-06, + "loss": 0.144, + "step": 388 + }, + { + "epoch": 0.08850967007963595, + "grad_norm": 2.989163897960478, + "learning_rate": 1.2490387328675226e-06, + "loss": 0.2016, + "step": 389 + }, + { + "epoch": 0.08873720136518772, + "grad_norm": 4.587176162210019, + "learning_rate": 1.2490337727757576e-06, + "loss": 0.2284, + "step": 390 + }, + { + "epoch": 0.08896473265073948, + "grad_norm": 2.794747809075531, + "learning_rate": 1.249028799929885e-06, + "loss": 0.2002, + "step": 391 + }, + { + "epoch": 0.08919226393629125, + "grad_norm": 2.0197262567230276, + "learning_rate": 1.2490238143300066e-06, + "loss": 0.1143, + "step": 392 + }, + { + "epoch": 0.089419795221843, + "grad_norm": 3.184614553894442, + "learning_rate": 1.2490188159762243e-06, + "loss": 0.1913, + "step": 393 + }, + { + "epoch": 0.08964732650739476, + "grad_norm": 2.518010477046937, + "learning_rate": 1.2490138048686405e-06, + "loss": 0.1981, + "step": 394 + }, + { + "epoch": 0.08987485779294653, + "grad_norm": 5.010077865699377, + "learning_rate": 1.249008781007357e-06, + "loss": 0.1423, + "step": 395 + }, + { + "epoch": 0.0901023890784983, + "grad_norm": 1.420461399090385, + "learning_rate": 1.2490037443924768e-06, + "loss": 0.1363, + "step": 396 + }, + { + "epoch": 0.09032992036405006, + "grad_norm": 2.5810652557759863, + "learning_rate": 1.2489986950241032e-06, + "loss": 0.1002, + "step": 397 + }, + { + "epoch": 0.09055745164960181, + "grad_norm": 1.8725706501255737, + "learning_rate": 1.2489936329023387e-06, + "loss": 0.1974, + "step": 398 + }, + { + "epoch": 0.09078498293515358, + "grad_norm": 3.2869147678539554, + "learning_rate": 1.2489885580272874e-06, + "loss": 0.1629, + "step": 399 + }, + { + "epoch": 0.09101251422070535, + "grad_norm": 1.7546095764098488, + "learning_rate": 1.2489834703990527e-06, + "loss": 0.1326, + "step": 400 + }, + { + "epoch": 0.09124004550625711, + "grad_norm": 3.0930989898336407, + "learning_rate": 1.2489783700177385e-06, + "loss": 0.2565, + "step": 401 + }, + { + "epoch": 0.09146757679180888, + "grad_norm": 4.363886237065706, + "learning_rate": 1.2489732568834492e-06, + "loss": 0.1425, + "step": 402 + }, + { + "epoch": 0.09169510807736064, + "grad_norm": 2.141413419957395, + "learning_rate": 1.2489681309962895e-06, + "loss": 0.1458, + "step": 403 + }, + { + "epoch": 0.0919226393629124, + "grad_norm": 4.5478526718009205, + "learning_rate": 1.2489629923563637e-06, + "loss": 0.1655, + "step": 404 + }, + { + "epoch": 0.09215017064846416, + "grad_norm": 5.253865415098631, + "learning_rate": 1.2489578409637774e-06, + "loss": 0.2702, + "step": 405 + }, + { + "epoch": 0.09237770193401593, + "grad_norm": 6.114423825591168, + "learning_rate": 1.2489526768186352e-06, + "loss": 0.1364, + "step": 406 + }, + { + "epoch": 0.09260523321956769, + "grad_norm": 2.4260049242900505, + "learning_rate": 1.2489474999210434e-06, + "loss": 0.1573, + "step": 407 + }, + { + "epoch": 0.09283276450511946, + "grad_norm": 6.696614155480106, + "learning_rate": 1.2489423102711068e-06, + "loss": 0.2365, + "step": 408 + }, + { + "epoch": 0.09306029579067122, + "grad_norm": 3.4093511525509848, + "learning_rate": 1.2489371078689326e-06, + "loss": 0.1552, + "step": 409 + }, + { + "epoch": 0.09328782707622298, + "grad_norm": 3.512014449058475, + "learning_rate": 1.2489318927146263e-06, + "loss": 0.1392, + "step": 410 + }, + { + "epoch": 0.09351535836177474, + "grad_norm": 4.385040034701264, + "learning_rate": 1.2489266648082951e-06, + "loss": 0.1184, + "step": 411 + }, + { + "epoch": 0.09374288964732651, + "grad_norm": 11.030038016242493, + "learning_rate": 1.2489214241500453e-06, + "loss": 0.2445, + "step": 412 + }, + { + "epoch": 0.09397042093287827, + "grad_norm": 3.8160488235069487, + "learning_rate": 1.2489161707399843e-06, + "loss": 0.2422, + "step": 413 + }, + { + "epoch": 0.09419795221843004, + "grad_norm": 2.5154081754915554, + "learning_rate": 1.2489109045782194e-06, + "loss": 0.1284, + "step": 414 + }, + { + "epoch": 0.09442548350398179, + "grad_norm": 2.186602019326803, + "learning_rate": 1.2489056256648582e-06, + "loss": 0.1387, + "step": 415 + }, + { + "epoch": 0.09465301478953356, + "grad_norm": 3.1244704898712223, + "learning_rate": 1.2489003340000089e-06, + "loss": 0.2695, + "step": 416 + }, + { + "epoch": 0.09488054607508532, + "grad_norm": 1.9015703147093774, + "learning_rate": 1.2488950295837792e-06, + "loss": 0.2029, + "step": 417 + }, + { + "epoch": 0.09510807736063709, + "grad_norm": 3.2255120343889523, + "learning_rate": 1.2488897124162777e-06, + "loss": 0.1708, + "step": 418 + }, + { + "epoch": 0.09533560864618885, + "grad_norm": 2.4361554392110354, + "learning_rate": 1.248884382497613e-06, + "loss": 0.237, + "step": 419 + }, + { + "epoch": 0.09556313993174062, + "grad_norm": 5.44904137240634, + "learning_rate": 1.2488790398278941e-06, + "loss": 0.2259, + "step": 420 + }, + { + "epoch": 0.09579067121729237, + "grad_norm": 2.5542725247665725, + "learning_rate": 1.2488736844072304e-06, + "loss": 0.1706, + "step": 421 + }, + { + "epoch": 0.09601820250284414, + "grad_norm": 3.3440828684749837, + "learning_rate": 1.248868316235731e-06, + "loss": 0.166, + "step": 422 + }, + { + "epoch": 0.0962457337883959, + "grad_norm": 2.837980086891423, + "learning_rate": 1.2488629353135059e-06, + "loss": 0.1974, + "step": 423 + }, + { + "epoch": 0.09647326507394767, + "grad_norm": 3.0821716156484413, + "learning_rate": 1.2488575416406649e-06, + "loss": 0.2029, + "step": 424 + }, + { + "epoch": 0.09670079635949944, + "grad_norm": 4.11082660525738, + "learning_rate": 1.2488521352173183e-06, + "loss": 0.1288, + "step": 425 + }, + { + "epoch": 0.09692832764505119, + "grad_norm": 2.792375492899653, + "learning_rate": 1.2488467160435765e-06, + "loss": 0.1318, + "step": 426 + }, + { + "epoch": 0.09715585893060295, + "grad_norm": 2.54978143800456, + "learning_rate": 1.2488412841195505e-06, + "loss": 0.2235, + "step": 427 + }, + { + "epoch": 0.09738339021615472, + "grad_norm": 1.8685713785223814, + "learning_rate": 1.2488358394453512e-06, + "loss": 0.1018, + "step": 428 + }, + { + "epoch": 0.09761092150170649, + "grad_norm": 2.19856597261874, + "learning_rate": 1.2488303820210897e-06, + "loss": 0.0955, + "step": 429 + }, + { + "epoch": 0.09783845278725825, + "grad_norm": 2.756460140283964, + "learning_rate": 1.2488249118468776e-06, + "loss": 0.161, + "step": 430 + }, + { + "epoch": 0.09806598407281002, + "grad_norm": 3.1658885878432446, + "learning_rate": 1.248819428922827e-06, + "loss": 0.1707, + "step": 431 + }, + { + "epoch": 0.09829351535836177, + "grad_norm": 3.574624372801338, + "learning_rate": 1.2488139332490495e-06, + "loss": 0.2412, + "step": 432 + }, + { + "epoch": 0.09852104664391353, + "grad_norm": 2.63473599121384, + "learning_rate": 1.248808424825658e-06, + "loss": 0.1195, + "step": 433 + }, + { + "epoch": 0.0987485779294653, + "grad_norm": 3.928170371490413, + "learning_rate": 1.2488029036527645e-06, + "loss": 0.1478, + "step": 434 + }, + { + "epoch": 0.09897610921501707, + "grad_norm": 2.0459697190569583, + "learning_rate": 1.2487973697304822e-06, + "loss": 0.0868, + "step": 435 + }, + { + "epoch": 0.09920364050056883, + "grad_norm": 2.2037192709560283, + "learning_rate": 1.248791823058924e-06, + "loss": 0.1911, + "step": 436 + }, + { + "epoch": 0.0994311717861206, + "grad_norm": 3.549121049187713, + "learning_rate": 1.2487862636382034e-06, + "loss": 0.1218, + "step": 437 + }, + { + "epoch": 0.09965870307167235, + "grad_norm": 1.4303061363329783, + "learning_rate": 1.248780691468434e-06, + "loss": 0.1116, + "step": 438 + }, + { + "epoch": 0.09988623435722412, + "grad_norm": 3.8141735085769746, + "learning_rate": 1.2487751065497296e-06, + "loss": 0.2179, + "step": 439 + }, + { + "epoch": 0.10011376564277588, + "grad_norm": 2.6329169063924986, + "learning_rate": 1.2487695088822044e-06, + "loss": 0.1492, + "step": 440 + }, + { + "epoch": 0.10034129692832765, + "grad_norm": 2.8773216855185635, + "learning_rate": 1.2487638984659729e-06, + "loss": 0.0988, + "step": 441 + }, + { + "epoch": 0.10056882821387941, + "grad_norm": 2.5448731857786284, + "learning_rate": 1.2487582753011496e-06, + "loss": 0.1023, + "step": 442 + }, + { + "epoch": 0.10079635949943117, + "grad_norm": 2.4399816480891445, + "learning_rate": 1.2487526393878497e-06, + "loss": 0.2015, + "step": 443 + }, + { + "epoch": 0.10102389078498293, + "grad_norm": 2.056202357783669, + "learning_rate": 1.248746990726188e-06, + "loss": 0.1376, + "step": 444 + }, + { + "epoch": 0.1012514220705347, + "grad_norm": 2.489946255383071, + "learning_rate": 1.2487413293162803e-06, + "loss": 0.1389, + "step": 445 + }, + { + "epoch": 0.10147895335608646, + "grad_norm": 2.3660691937468807, + "learning_rate": 1.2487356551582421e-06, + "loss": 0.2235, + "step": 446 + }, + { + "epoch": 0.10170648464163823, + "grad_norm": 2.5030375037996575, + "learning_rate": 1.2487299682521893e-06, + "loss": 0.2156, + "step": 447 + }, + { + "epoch": 0.10193401592719, + "grad_norm": 2.210721856008811, + "learning_rate": 1.2487242685982384e-06, + "loss": 0.1101, + "step": 448 + }, + { + "epoch": 0.10216154721274175, + "grad_norm": 2.250420318734035, + "learning_rate": 1.2487185561965057e-06, + "loss": 0.1241, + "step": 449 + }, + { + "epoch": 0.10238907849829351, + "grad_norm": 2.019413043508561, + "learning_rate": 1.248712831047108e-06, + "loss": 0.1217, + "step": 450 + }, + { + "epoch": 0.10261660978384528, + "grad_norm": 3.2295330442493713, + "learning_rate": 1.2487070931501624e-06, + "loss": 0.2304, + "step": 451 + }, + { + "epoch": 0.10284414106939704, + "grad_norm": 2.444299385213433, + "learning_rate": 1.2487013425057858e-06, + "loss": 0.2084, + "step": 452 + }, + { + "epoch": 0.10307167235494881, + "grad_norm": 2.8966369631126367, + "learning_rate": 1.2486955791140964e-06, + "loss": 0.1838, + "step": 453 + }, + { + "epoch": 0.10329920364050058, + "grad_norm": 2.0941566856763387, + "learning_rate": 1.2486898029752113e-06, + "loss": 0.1043, + "step": 454 + }, + { + "epoch": 0.10352673492605233, + "grad_norm": 2.3019250022426925, + "learning_rate": 1.248684014089249e-06, + "loss": 0.1189, + "step": 455 + }, + { + "epoch": 0.1037542662116041, + "grad_norm": 2.1349092143720387, + "learning_rate": 1.2486782124563277e-06, + "loss": 0.1708, + "step": 456 + }, + { + "epoch": 0.10398179749715586, + "grad_norm": 3.101054381668985, + "learning_rate": 1.2486723980765659e-06, + "loss": 0.1796, + "step": 457 + }, + { + "epoch": 0.10420932878270762, + "grad_norm": 1.9574694651381292, + "learning_rate": 1.2486665709500826e-06, + "loss": 0.1762, + "step": 458 + }, + { + "epoch": 0.10443686006825939, + "grad_norm": 1.9997685220641748, + "learning_rate": 1.2486607310769965e-06, + "loss": 0.1626, + "step": 459 + }, + { + "epoch": 0.10466439135381114, + "grad_norm": 1.4987645243428842, + "learning_rate": 1.2486548784574275e-06, + "loss": 0.1104, + "step": 460 + }, + { + "epoch": 0.10489192263936291, + "grad_norm": 3.0056305765303857, + "learning_rate": 1.2486490130914948e-06, + "loss": 0.1526, + "step": 461 + }, + { + "epoch": 0.10511945392491467, + "grad_norm": 1.6498658926200307, + "learning_rate": 1.2486431349793185e-06, + "loss": 0.1158, + "step": 462 + }, + { + "epoch": 0.10534698521046644, + "grad_norm": 2.8097802744351035, + "learning_rate": 1.2486372441210188e-06, + "loss": 0.174, + "step": 463 + }, + { + "epoch": 0.1055745164960182, + "grad_norm": 2.2295425114906955, + "learning_rate": 1.248631340516716e-06, + "loss": 0.0993, + "step": 464 + }, + { + "epoch": 0.10580204778156997, + "grad_norm": 1.7352971105344217, + "learning_rate": 1.2486254241665302e-06, + "loss": 0.1799, + "step": 465 + }, + { + "epoch": 0.10602957906712172, + "grad_norm": 3.37890451450669, + "learning_rate": 1.2486194950705831e-06, + "loss": 0.1456, + "step": 466 + }, + { + "epoch": 0.10625711035267349, + "grad_norm": 4.485196875503332, + "learning_rate": 1.248613553228996e-06, + "loss": 0.1509, + "step": 467 + }, + { + "epoch": 0.10648464163822526, + "grad_norm": 3.8128664414272833, + "learning_rate": 1.2486075986418896e-06, + "loss": 0.1217, + "step": 468 + }, + { + "epoch": 0.10671217292377702, + "grad_norm": 1.9049325746647565, + "learning_rate": 1.248601631309386e-06, + "loss": 0.1973, + "step": 469 + }, + { + "epoch": 0.10693970420932879, + "grad_norm": 1.9433225744575688, + "learning_rate": 1.2485956512316072e-06, + "loss": 0.1422, + "step": 470 + }, + { + "epoch": 0.10716723549488055, + "grad_norm": 1.7542185976103952, + "learning_rate": 1.2485896584086754e-06, + "loss": 0.1187, + "step": 471 + }, + { + "epoch": 0.1073947667804323, + "grad_norm": 0.985585738392577, + "learning_rate": 1.248583652840713e-06, + "loss": 0.1116, + "step": 472 + }, + { + "epoch": 0.10762229806598407, + "grad_norm": 6.520293791736507, + "learning_rate": 1.2485776345278427e-06, + "loss": 0.1634, + "step": 473 + }, + { + "epoch": 0.10784982935153584, + "grad_norm": 2.9958165676640935, + "learning_rate": 1.2485716034701876e-06, + "loss": 0.1468, + "step": 474 + }, + { + "epoch": 0.1080773606370876, + "grad_norm": 3.496540224028896, + "learning_rate": 1.2485655596678712e-06, + "loss": 0.1444, + "step": 475 + }, + { + "epoch": 0.10830489192263937, + "grad_norm": 2.6887910577996603, + "learning_rate": 1.2485595031210164e-06, + "loss": 0.2257, + "step": 476 + }, + { + "epoch": 0.10853242320819112, + "grad_norm": 2.210859712757279, + "learning_rate": 1.2485534338297475e-06, + "loss": 0.0858, + "step": 477 + }, + { + "epoch": 0.10875995449374289, + "grad_norm": 1.5912288577365465, + "learning_rate": 1.2485473517941884e-06, + "loss": 0.1021, + "step": 478 + }, + { + "epoch": 0.10898748577929465, + "grad_norm": 2.162920899638659, + "learning_rate": 1.2485412570144633e-06, + "loss": 0.2051, + "step": 479 + }, + { + "epoch": 0.10921501706484642, + "grad_norm": 2.3337569161162186, + "learning_rate": 1.2485351494906969e-06, + "loss": 0.1726, + "step": 480 + }, + { + "epoch": 0.10944254835039818, + "grad_norm": 1.6587972530161754, + "learning_rate": 1.2485290292230142e-06, + "loss": 0.1589, + "step": 481 + }, + { + "epoch": 0.10967007963594995, + "grad_norm": 2.549443212629399, + "learning_rate": 1.24852289621154e-06, + "loss": 0.1107, + "step": 482 + }, + { + "epoch": 0.1098976109215017, + "grad_norm": 1.9600173744992218, + "learning_rate": 1.2485167504563995e-06, + "loss": 0.1497, + "step": 483 + }, + { + "epoch": 0.11012514220705347, + "grad_norm": 2.914488733886043, + "learning_rate": 1.2485105919577187e-06, + "loss": 0.2242, + "step": 484 + }, + { + "epoch": 0.11035267349260523, + "grad_norm": 2.4334592724633475, + "learning_rate": 1.2485044207156233e-06, + "loss": 0.1326, + "step": 485 + }, + { + "epoch": 0.110580204778157, + "grad_norm": 2.1918094312708374, + "learning_rate": 1.2484982367302395e-06, + "loss": 0.1611, + "step": 486 + }, + { + "epoch": 0.11080773606370876, + "grad_norm": 2.2072766100880843, + "learning_rate": 1.2484920400016936e-06, + "loss": 0.1402, + "step": 487 + }, + { + "epoch": 0.11103526734926053, + "grad_norm": 1.6859469474720183, + "learning_rate": 1.2484858305301122e-06, + "loss": 0.1472, + "step": 488 + }, + { + "epoch": 0.11126279863481228, + "grad_norm": 1.590244696061809, + "learning_rate": 1.2484796083156222e-06, + "loss": 0.0824, + "step": 489 + }, + { + "epoch": 0.11149032992036405, + "grad_norm": 4.525638347888733, + "learning_rate": 1.2484733733583511e-06, + "loss": 0.1257, + "step": 490 + }, + { + "epoch": 0.11171786120591581, + "grad_norm": 2.6721724669454723, + "learning_rate": 1.248467125658426e-06, + "loss": 0.2084, + "step": 491 + }, + { + "epoch": 0.11194539249146758, + "grad_norm": 2.300055245713483, + "learning_rate": 1.2484608652159746e-06, + "loss": 0.1053, + "step": 492 + }, + { + "epoch": 0.11217292377701935, + "grad_norm": 3.273977920110333, + "learning_rate": 1.248454592031125e-06, + "loss": 0.1176, + "step": 493 + }, + { + "epoch": 0.1124004550625711, + "grad_norm": 2.101057790899636, + "learning_rate": 1.2484483061040054e-06, + "loss": 0.1277, + "step": 494 + }, + { + "epoch": 0.11262798634812286, + "grad_norm": 3.6133620556599984, + "learning_rate": 1.2484420074347441e-06, + "loss": 0.1845, + "step": 495 + }, + { + "epoch": 0.11285551763367463, + "grad_norm": 1.9619725915027257, + "learning_rate": 1.24843569602347e-06, + "loss": 0.1894, + "step": 496 + }, + { + "epoch": 0.1130830489192264, + "grad_norm": 2.636905846270966, + "learning_rate": 1.2484293718703119e-06, + "loss": 0.1874, + "step": 497 + }, + { + "epoch": 0.11331058020477816, + "grad_norm": 2.5593822043936125, + "learning_rate": 1.2484230349753994e-06, + "loss": 0.0927, + "step": 498 + }, + { + "epoch": 0.11353811149032993, + "grad_norm": 2.2440609982402715, + "learning_rate": 1.2484166853388617e-06, + "loss": 0.1381, + "step": 499 + }, + { + "epoch": 0.11376564277588168, + "grad_norm": 2.7232866925160506, + "learning_rate": 1.2484103229608288e-06, + "loss": 0.1758, + "step": 500 + }, + { + "epoch": 0.11399317406143344, + "grad_norm": 2.6484317978572816, + "learning_rate": 1.2484039478414305e-06, + "loss": 0.1259, + "step": 501 + }, + { + "epoch": 0.11422070534698521, + "grad_norm": 2.1058374053464464, + "learning_rate": 1.2483975599807972e-06, + "loss": 0.1369, + "step": 502 + }, + { + "epoch": 0.11444823663253698, + "grad_norm": 2.1458925241645903, + "learning_rate": 1.2483911593790595e-06, + "loss": 0.1004, + "step": 503 + }, + { + "epoch": 0.11467576791808874, + "grad_norm": 3.031837353586065, + "learning_rate": 1.2483847460363482e-06, + "loss": 0.154, + "step": 504 + }, + { + "epoch": 0.1149032992036405, + "grad_norm": 3.1297621875057544, + "learning_rate": 1.2483783199527943e-06, + "loss": 0.1071, + "step": 505 + }, + { + "epoch": 0.11513083048919226, + "grad_norm": 2.5407911203085787, + "learning_rate": 1.2483718811285296e-06, + "loss": 0.1744, + "step": 506 + }, + { + "epoch": 0.11535836177474403, + "grad_norm": 3.1175064627764377, + "learning_rate": 1.2483654295636848e-06, + "loss": 0.1072, + "step": 507 + }, + { + "epoch": 0.11558589306029579, + "grad_norm": 3.0988741009535667, + "learning_rate": 1.2483589652583924e-06, + "loss": 0.1753, + "step": 508 + }, + { + "epoch": 0.11581342434584756, + "grad_norm": 1.8808814641931946, + "learning_rate": 1.2483524882127846e-06, + "loss": 0.0859, + "step": 509 + }, + { + "epoch": 0.11604095563139932, + "grad_norm": 2.8937543802568158, + "learning_rate": 1.2483459984269933e-06, + "loss": 0.1816, + "step": 510 + }, + { + "epoch": 0.11626848691695107, + "grad_norm": 2.186370885841539, + "learning_rate": 1.2483394959011514e-06, + "loss": 0.0819, + "step": 511 + }, + { + "epoch": 0.11649601820250284, + "grad_norm": 1.8650801779387822, + "learning_rate": 1.248332980635392e-06, + "loss": 0.1436, + "step": 512 + }, + { + "epoch": 0.1167235494880546, + "grad_norm": 2.9270321544640994, + "learning_rate": 1.2483264526298478e-06, + "loss": 0.1308, + "step": 513 + }, + { + "epoch": 0.11695108077360637, + "grad_norm": 1.9942689645578024, + "learning_rate": 1.2483199118846525e-06, + "loss": 0.1656, + "step": 514 + }, + { + "epoch": 0.11717861205915814, + "grad_norm": 2.8104633311436116, + "learning_rate": 1.2483133583999399e-06, + "loss": 0.1681, + "step": 515 + }, + { + "epoch": 0.1174061433447099, + "grad_norm": 2.546169206593085, + "learning_rate": 1.2483067921758439e-06, + "loss": 0.0925, + "step": 516 + }, + { + "epoch": 0.11763367463026166, + "grad_norm": 2.0758430805982178, + "learning_rate": 1.2483002132124983e-06, + "loss": 0.203, + "step": 517 + }, + { + "epoch": 0.11786120591581342, + "grad_norm": 2.1497459150584386, + "learning_rate": 1.2482936215100382e-06, + "loss": 0.1056, + "step": 518 + }, + { + "epoch": 0.11808873720136519, + "grad_norm": 2.197584956184683, + "learning_rate": 1.2482870170685978e-06, + "loss": 0.0933, + "step": 519 + }, + { + "epoch": 0.11831626848691695, + "grad_norm": 4.944962250057973, + "learning_rate": 1.2482803998883122e-06, + "loss": 0.2129, + "step": 520 + }, + { + "epoch": 0.11854379977246872, + "grad_norm": 1.5333537239736301, + "learning_rate": 1.2482737699693168e-06, + "loss": 0.1729, + "step": 521 + }, + { + "epoch": 0.11877133105802047, + "grad_norm": 2.5556570479037948, + "learning_rate": 1.248267127311747e-06, + "loss": 0.1607, + "step": 522 + }, + { + "epoch": 0.11899886234357224, + "grad_norm": 2.0949542782407398, + "learning_rate": 1.2482604719157386e-06, + "loss": 0.1857, + "step": 523 + }, + { + "epoch": 0.119226393629124, + "grad_norm": 2.2586097350216385, + "learning_rate": 1.2482538037814277e-06, + "loss": 0.1258, + "step": 524 + }, + { + "epoch": 0.11945392491467577, + "grad_norm": 3.036602602741407, + "learning_rate": 1.2482471229089502e-06, + "loss": 0.161, + "step": 525 + }, + { + "epoch": 0.11968145620022753, + "grad_norm": 3.382002996482515, + "learning_rate": 1.2482404292984431e-06, + "loss": 0.1784, + "step": 526 + }, + { + "epoch": 0.1199089874857793, + "grad_norm": 1.571226708630226, + "learning_rate": 1.248233722950043e-06, + "loss": 0.1605, + "step": 527 + }, + { + "epoch": 0.12013651877133105, + "grad_norm": 3.0053996402943737, + "learning_rate": 1.2482270038638872e-06, + "loss": 0.1201, + "step": 528 + }, + { + "epoch": 0.12036405005688282, + "grad_norm": 4.663906907753179, + "learning_rate": 1.2482202720401128e-06, + "loss": 0.203, + "step": 529 + }, + { + "epoch": 0.12059158134243458, + "grad_norm": 2.107107186527039, + "learning_rate": 1.248213527478857e-06, + "loss": 0.1933, + "step": 530 + }, + { + "epoch": 0.12081911262798635, + "grad_norm": 2.191569921182264, + "learning_rate": 1.2482067701802583e-06, + "loss": 0.1735, + "step": 531 + }, + { + "epoch": 0.12104664391353812, + "grad_norm": 1.611611034864374, + "learning_rate": 1.2482000001444547e-06, + "loss": 0.1299, + "step": 532 + }, + { + "epoch": 0.12127417519908988, + "grad_norm": 1.9644367618752439, + "learning_rate": 1.2481932173715845e-06, + "loss": 0.0868, + "step": 533 + }, + { + "epoch": 0.12150170648464163, + "grad_norm": 1.7597689357542332, + "learning_rate": 1.2481864218617859e-06, + "loss": 0.1977, + "step": 534 + }, + { + "epoch": 0.1217292377701934, + "grad_norm": 1.0455766882042379, + "learning_rate": 1.2481796136151984e-06, + "loss": 0.0856, + "step": 535 + }, + { + "epoch": 0.12195676905574517, + "grad_norm": 3.2419347761543684, + "learning_rate": 1.2481727926319609e-06, + "loss": 0.2399, + "step": 536 + }, + { + "epoch": 0.12218430034129693, + "grad_norm": 3.339873316715719, + "learning_rate": 1.2481659589122127e-06, + "loss": 0.186, + "step": 537 + }, + { + "epoch": 0.1224118316268487, + "grad_norm": 3.4453888669974146, + "learning_rate": 1.2481591124560934e-06, + "loss": 0.2007, + "step": 538 + }, + { + "epoch": 0.12263936291240045, + "grad_norm": 3.4700673703521736, + "learning_rate": 1.2481522532637435e-06, + "loss": 0.1632, + "step": 539 + }, + { + "epoch": 0.12286689419795221, + "grad_norm": 2.355397510374851, + "learning_rate": 1.2481453813353026e-06, + "loss": 0.1212, + "step": 540 + }, + { + "epoch": 0.12309442548350398, + "grad_norm": 5.338957920220655, + "learning_rate": 1.2481384966709116e-06, + "loss": 0.1592, + "step": 541 + }, + { + "epoch": 0.12332195676905575, + "grad_norm": 2.990026650956376, + "learning_rate": 1.2481315992707104e-06, + "loss": 0.2656, + "step": 542 + }, + { + "epoch": 0.12354948805460751, + "grad_norm": 1.8798810865858828, + "learning_rate": 1.248124689134841e-06, + "loss": 0.1125, + "step": 543 + }, + { + "epoch": 0.12377701934015928, + "grad_norm": 1.6104299610891197, + "learning_rate": 1.2481177662634438e-06, + "loss": 0.1557, + "step": 544 + }, + { + "epoch": 0.12400455062571103, + "grad_norm": 3.302283676048537, + "learning_rate": 1.2481108306566609e-06, + "loss": 0.1799, + "step": 545 + }, + { + "epoch": 0.1242320819112628, + "grad_norm": 2.0532951352869513, + "learning_rate": 1.2481038823146338e-06, + "loss": 0.0815, + "step": 546 + }, + { + "epoch": 0.12445961319681456, + "grad_norm": 1.4326913794879275, + "learning_rate": 1.2480969212375043e-06, + "loss": 0.177, + "step": 547 + }, + { + "epoch": 0.12468714448236633, + "grad_norm": 3.5494676426295286, + "learning_rate": 1.2480899474254151e-06, + "loss": 0.136, + "step": 548 + }, + { + "epoch": 0.12491467576791809, + "grad_norm": 1.3410455744599155, + "learning_rate": 1.2480829608785085e-06, + "loss": 0.1078, + "step": 549 + }, + { + "epoch": 0.12514220705346984, + "grad_norm": 1.7709434217848017, + "learning_rate": 1.2480759615969273e-06, + "loss": 0.1114, + "step": 550 + }, + { + "epoch": 0.12536973833902162, + "grad_norm": 1.4865770903343614, + "learning_rate": 1.2480689495808144e-06, + "loss": 0.1377, + "step": 551 + }, + { + "epoch": 0.12559726962457338, + "grad_norm": 1.6211826207402742, + "learning_rate": 1.2480619248303133e-06, + "loss": 0.1873, + "step": 552 + }, + { + "epoch": 0.12582480091012513, + "grad_norm": 3.1755876159758794, + "learning_rate": 1.2480548873455675e-06, + "loss": 0.2135, + "step": 553 + }, + { + "epoch": 0.1260523321956769, + "grad_norm": 3.6986046315140952, + "learning_rate": 1.248047837126721e-06, + "loss": 0.3549, + "step": 554 + }, + { + "epoch": 0.12627986348122866, + "grad_norm": 2.782290781984551, + "learning_rate": 1.248040774173918e-06, + "loss": 0.1936, + "step": 555 + }, + { + "epoch": 0.12650739476678044, + "grad_norm": 2.329760734261347, + "learning_rate": 1.248033698487302e-06, + "loss": 0.1395, + "step": 556 + }, + { + "epoch": 0.1267349260523322, + "grad_norm": 2.258554836923121, + "learning_rate": 1.2480266100670189e-06, + "loss": 0.1605, + "step": 557 + }, + { + "epoch": 0.12696245733788397, + "grad_norm": 3.058041285297341, + "learning_rate": 1.2480195089132125e-06, + "loss": 0.1975, + "step": 558 + }, + { + "epoch": 0.12718998862343572, + "grad_norm": 2.406042057945949, + "learning_rate": 1.2480123950260284e-06, + "loss": 0.1405, + "step": 559 + }, + { + "epoch": 0.12741751990898748, + "grad_norm": 1.4634033865621767, + "learning_rate": 1.248005268405612e-06, + "loss": 0.0686, + "step": 560 + }, + { + "epoch": 0.12764505119453926, + "grad_norm": 1.1470288222889338, + "learning_rate": 1.2479981290521087e-06, + "loss": 0.0649, + "step": 561 + }, + { + "epoch": 0.127872582480091, + "grad_norm": 3.357158703331078, + "learning_rate": 1.2479909769656648e-06, + "loss": 0.1684, + "step": 562 + }, + { + "epoch": 0.1281001137656428, + "grad_norm": 2.4363436867877595, + "learning_rate": 1.2479838121464263e-06, + "loss": 0.2155, + "step": 563 + }, + { + "epoch": 0.12832764505119454, + "grad_norm": 4.051636355021599, + "learning_rate": 1.2479766345945395e-06, + "loss": 0.1853, + "step": 564 + }, + { + "epoch": 0.1285551763367463, + "grad_norm": 1.6707836764627593, + "learning_rate": 1.2479694443101513e-06, + "loss": 0.2261, + "step": 565 + }, + { + "epoch": 0.12878270762229807, + "grad_norm": 1.3008647546251737, + "learning_rate": 1.2479622412934087e-06, + "loss": 0.1606, + "step": 566 + }, + { + "epoch": 0.12901023890784982, + "grad_norm": 3.421202381350775, + "learning_rate": 1.2479550255444586e-06, + "loss": 0.147, + "step": 567 + }, + { + "epoch": 0.1292377701934016, + "grad_norm": 1.5157864652280186, + "learning_rate": 1.2479477970634487e-06, + "loss": 0.1536, + "step": 568 + }, + { + "epoch": 0.12946530147895335, + "grad_norm": 3.27856184412377, + "learning_rate": 1.2479405558505267e-06, + "loss": 0.1931, + "step": 569 + }, + { + "epoch": 0.1296928327645051, + "grad_norm": 2.5943823025048474, + "learning_rate": 1.247933301905841e-06, + "loss": 0.1384, + "step": 570 + }, + { + "epoch": 0.12992036405005689, + "grad_norm": 4.278003846990416, + "learning_rate": 1.2479260352295388e-06, + "loss": 0.1771, + "step": 571 + }, + { + "epoch": 0.13014789533560864, + "grad_norm": 3.446486195671729, + "learning_rate": 1.2479187558217697e-06, + "loss": 0.1323, + "step": 572 + }, + { + "epoch": 0.13037542662116042, + "grad_norm": 1.5099352019896337, + "learning_rate": 1.247911463682682e-06, + "loss": 0.1444, + "step": 573 + }, + { + "epoch": 0.13060295790671217, + "grad_norm": 3.798908546439363, + "learning_rate": 1.2479041588124247e-06, + "loss": 0.1504, + "step": 574 + }, + { + "epoch": 0.13083048919226395, + "grad_norm": 3.7532424433768754, + "learning_rate": 1.2478968412111471e-06, + "loss": 0.1518, + "step": 575 + }, + { + "epoch": 0.1310580204778157, + "grad_norm": 2.056630545760187, + "learning_rate": 1.247889510878999e-06, + "loss": 0.2708, + "step": 576 + }, + { + "epoch": 0.13128555176336745, + "grad_norm": 2.303355999452058, + "learning_rate": 1.24788216781613e-06, + "loss": 0.1662, + "step": 577 + }, + { + "epoch": 0.13151308304891923, + "grad_norm": 2.269104241548175, + "learning_rate": 1.2478748120226902e-06, + "loss": 0.1337, + "step": 578 + }, + { + "epoch": 0.13174061433447098, + "grad_norm": 3.0692597907642862, + "learning_rate": 1.2478674434988299e-06, + "loss": 0.1326, + "step": 579 + }, + { + "epoch": 0.13196814562002276, + "grad_norm": 1.6865202158454742, + "learning_rate": 1.2478600622447001e-06, + "loss": 0.1647, + "step": 580 + }, + { + "epoch": 0.13219567690557452, + "grad_norm": 2.939283703136826, + "learning_rate": 1.2478526682604512e-06, + "loss": 0.1303, + "step": 581 + }, + { + "epoch": 0.13242320819112627, + "grad_norm": 3.1064926411391713, + "learning_rate": 1.2478452615462345e-06, + "loss": 0.1409, + "step": 582 + }, + { + "epoch": 0.13265073947667805, + "grad_norm": 2.5571749562826485, + "learning_rate": 1.247837842102201e-06, + "loss": 0.1791, + "step": 583 + }, + { + "epoch": 0.1328782707622298, + "grad_norm": 2.795629539563545, + "learning_rate": 1.2478304099285031e-06, + "loss": 0.1567, + "step": 584 + }, + { + "epoch": 0.13310580204778158, + "grad_norm": 2.0832780528771466, + "learning_rate": 1.2478229650252921e-06, + "loss": 0.1639, + "step": 585 + }, + { + "epoch": 0.13333333333333333, + "grad_norm": 2.9969798024524117, + "learning_rate": 1.2478155073927204e-06, + "loss": 0.2444, + "step": 586 + }, + { + "epoch": 0.13356086461888508, + "grad_norm": 1.9274087851448982, + "learning_rate": 1.2478080370309404e-06, + "loss": 0.105, + "step": 587 + }, + { + "epoch": 0.13378839590443686, + "grad_norm": 4.021015627831867, + "learning_rate": 1.2478005539401046e-06, + "loss": 0.1734, + "step": 588 + }, + { + "epoch": 0.13401592718998862, + "grad_norm": 2.9342976021528027, + "learning_rate": 1.2477930581203663e-06, + "loss": 0.1465, + "step": 589 + }, + { + "epoch": 0.1342434584755404, + "grad_norm": 2.3242426333780632, + "learning_rate": 1.2477855495718782e-06, + "loss": 0.2241, + "step": 590 + }, + { + "epoch": 0.13447098976109215, + "grad_norm": 2.957504561813871, + "learning_rate": 1.2477780282947942e-06, + "loss": 0.1734, + "step": 591 + }, + { + "epoch": 0.13469852104664393, + "grad_norm": 1.8788696793522301, + "learning_rate": 1.2477704942892677e-06, + "loss": 0.1469, + "step": 592 + }, + { + "epoch": 0.13492605233219568, + "grad_norm": 2.339527187323086, + "learning_rate": 1.2477629475554532e-06, + "loss": 0.1312, + "step": 593 + }, + { + "epoch": 0.13515358361774743, + "grad_norm": 3.707567497860105, + "learning_rate": 1.2477553880935043e-06, + "loss": 0.1916, + "step": 594 + }, + { + "epoch": 0.1353811149032992, + "grad_norm": 3.2750827489523022, + "learning_rate": 1.2477478159035758e-06, + "loss": 0.1774, + "step": 595 + }, + { + "epoch": 0.13560864618885096, + "grad_norm": 2.777476705753077, + "learning_rate": 1.2477402309858226e-06, + "loss": 0.1789, + "step": 596 + }, + { + "epoch": 0.13583617747440274, + "grad_norm": 2.144596195630353, + "learning_rate": 1.2477326333403995e-06, + "loss": 0.147, + "step": 597 + }, + { + "epoch": 0.1360637087599545, + "grad_norm": 2.3685083837175935, + "learning_rate": 1.2477250229674618e-06, + "loss": 0.1831, + "step": 598 + }, + { + "epoch": 0.13629124004550625, + "grad_norm": 1.9843295041761948, + "learning_rate": 1.2477173998671653e-06, + "loss": 0.178, + "step": 599 + }, + { + "epoch": 0.13651877133105803, + "grad_norm": 3.434039497211011, + "learning_rate": 1.2477097640396655e-06, + "loss": 0.1235, + "step": 600 + }, + { + "epoch": 0.13674630261660978, + "grad_norm": 1.4586285890850859, + "learning_rate": 1.2477021154851185e-06, + "loss": 0.0977, + "step": 601 + }, + { + "epoch": 0.13697383390216156, + "grad_norm": 3.3913304667052198, + "learning_rate": 1.2476944542036806e-06, + "loss": 0.1786, + "step": 602 + }, + { + "epoch": 0.1372013651877133, + "grad_norm": 2.667804003182341, + "learning_rate": 1.2476867801955086e-06, + "loss": 0.1204, + "step": 603 + }, + { + "epoch": 0.13742889647326506, + "grad_norm": 2.4655446209984033, + "learning_rate": 1.247679093460759e-06, + "loss": 0.2298, + "step": 604 + }, + { + "epoch": 0.13765642775881684, + "grad_norm": 3.1521634114958816, + "learning_rate": 1.2476713939995895e-06, + "loss": 0.1264, + "step": 605 + }, + { + "epoch": 0.1378839590443686, + "grad_norm": 1.8219187381761075, + "learning_rate": 1.2476636818121568e-06, + "loss": 0.1028, + "step": 606 + }, + { + "epoch": 0.13811149032992037, + "grad_norm": 2.337156447435568, + "learning_rate": 1.247655956898619e-06, + "loss": 0.1946, + "step": 607 + }, + { + "epoch": 0.13833902161547212, + "grad_norm": 3.2562899945752966, + "learning_rate": 1.2476482192591335e-06, + "loss": 0.1465, + "step": 608 + }, + { + "epoch": 0.1385665529010239, + "grad_norm": 1.8250022998173558, + "learning_rate": 1.247640468893859e-06, + "loss": 0.1467, + "step": 609 + }, + { + "epoch": 0.13879408418657566, + "grad_norm": 3.5242803865119603, + "learning_rate": 1.2476327058029534e-06, + "loss": 0.1225, + "step": 610 + }, + { + "epoch": 0.1390216154721274, + "grad_norm": 3.027013883019154, + "learning_rate": 1.2476249299865757e-06, + "loss": 0.1595, + "step": 611 + }, + { + "epoch": 0.1392491467576792, + "grad_norm": 2.3807833370240843, + "learning_rate": 1.2476171414448847e-06, + "loss": 0.0984, + "step": 612 + }, + { + "epoch": 0.13947667804323094, + "grad_norm": 3.1119739781274416, + "learning_rate": 1.2476093401780397e-06, + "loss": 0.154, + "step": 613 + }, + { + "epoch": 0.13970420932878272, + "grad_norm": 3.4567643287811958, + "learning_rate": 1.2476015261861998e-06, + "loss": 0.1405, + "step": 614 + }, + { + "epoch": 0.13993174061433447, + "grad_norm": 2.6730132596017504, + "learning_rate": 1.247593699469525e-06, + "loss": 0.117, + "step": 615 + }, + { + "epoch": 0.14015927189988622, + "grad_norm": 2.78286071664722, + "learning_rate": 1.2475858600281754e-06, + "loss": 0.1504, + "step": 616 + }, + { + "epoch": 0.140386803185438, + "grad_norm": 2.0905809356248803, + "learning_rate": 1.247578007862311e-06, + "loss": 0.1221, + "step": 617 + }, + { + "epoch": 0.14061433447098975, + "grad_norm": 2.307570493464016, + "learning_rate": 1.2475701429720923e-06, + "loss": 0.1166, + "step": 618 + }, + { + "epoch": 0.14084186575654153, + "grad_norm": 1.2783682538203782, + "learning_rate": 1.24756226535768e-06, + "loss": 0.1346, + "step": 619 + }, + { + "epoch": 0.1410693970420933, + "grad_norm": 1.497656716954093, + "learning_rate": 1.2475543750192352e-06, + "loss": 0.2064, + "step": 620 + }, + { + "epoch": 0.14129692832764504, + "grad_norm": 3.79056695480817, + "learning_rate": 1.2475464719569192e-06, + "loss": 0.2673, + "step": 621 + }, + { + "epoch": 0.14152445961319682, + "grad_norm": 1.4805750856049538, + "learning_rate": 1.2475385561708934e-06, + "loss": 0.1992, + "step": 622 + }, + { + "epoch": 0.14175199089874857, + "grad_norm": 1.6748002073239907, + "learning_rate": 1.2475306276613194e-06, + "loss": 0.0979, + "step": 623 + }, + { + "epoch": 0.14197952218430035, + "grad_norm": 2.5674392190565736, + "learning_rate": 1.2475226864283596e-06, + "loss": 0.1337, + "step": 624 + }, + { + "epoch": 0.1422070534698521, + "grad_norm": 2.656075374063454, + "learning_rate": 1.2475147324721764e-06, + "loss": 0.2501, + "step": 625 + }, + { + "epoch": 0.14243458475540388, + "grad_norm": 2.03707084801983, + "learning_rate": 1.2475067657929319e-06, + "loss": 0.1673, + "step": 626 + }, + { + "epoch": 0.14266211604095563, + "grad_norm": 2.975904435297751, + "learning_rate": 1.2474987863907894e-06, + "loss": 0.135, + "step": 627 + }, + { + "epoch": 0.14288964732650739, + "grad_norm": 2.2205623276633295, + "learning_rate": 1.2474907942659116e-06, + "loss": 0.2149, + "step": 628 + }, + { + "epoch": 0.14311717861205916, + "grad_norm": 2.271865927518249, + "learning_rate": 1.247482789418462e-06, + "loss": 0.1519, + "step": 629 + }, + { + "epoch": 0.14334470989761092, + "grad_norm": 6.542697842484103, + "learning_rate": 1.2474747718486044e-06, + "loss": 0.1757, + "step": 630 + }, + { + "epoch": 0.1435722411831627, + "grad_norm": 1.8493295758356152, + "learning_rate": 1.2474667415565022e-06, + "loss": 0.096, + "step": 631 + }, + { + "epoch": 0.14379977246871445, + "grad_norm": 4.567549869753572, + "learning_rate": 1.24745869854232e-06, + "loss": 0.1745, + "step": 632 + }, + { + "epoch": 0.1440273037542662, + "grad_norm": 3.104479250541457, + "learning_rate": 1.2474506428062219e-06, + "loss": 0.14, + "step": 633 + }, + { + "epoch": 0.14425483503981798, + "grad_norm": 2.9519743566943464, + "learning_rate": 1.2474425743483726e-06, + "loss": 0.237, + "step": 634 + }, + { + "epoch": 0.14448236632536973, + "grad_norm": 1.4814831832284159, + "learning_rate": 1.2474344931689371e-06, + "loss": 0.0873, + "step": 635 + }, + { + "epoch": 0.1447098976109215, + "grad_norm": 2.0222816327136712, + "learning_rate": 1.2474263992680805e-06, + "loss": 0.155, + "step": 636 + }, + { + "epoch": 0.14493742889647326, + "grad_norm": 2.0590304829666914, + "learning_rate": 1.247418292645968e-06, + "loss": 0.107, + "step": 637 + }, + { + "epoch": 0.14516496018202502, + "grad_norm": 2.5562023131920633, + "learning_rate": 1.2474101733027659e-06, + "loss": 0.2256, + "step": 638 + }, + { + "epoch": 0.1453924914675768, + "grad_norm": 2.3833084873555195, + "learning_rate": 1.2474020412386395e-06, + "loss": 0.1087, + "step": 639 + }, + { + "epoch": 0.14562002275312855, + "grad_norm": 1.5076273114920544, + "learning_rate": 1.2473938964537551e-06, + "loss": 0.0893, + "step": 640 + }, + { + "epoch": 0.14584755403868033, + "grad_norm": 2.3708066851044887, + "learning_rate": 1.2473857389482797e-06, + "loss": 0.1247, + "step": 641 + }, + { + "epoch": 0.14607508532423208, + "grad_norm": 1.5590215080673084, + "learning_rate": 1.2473775687223794e-06, + "loss": 0.1504, + "step": 642 + }, + { + "epoch": 0.14630261660978386, + "grad_norm": 1.6107910166409294, + "learning_rate": 1.2473693857762215e-06, + "loss": 0.149, + "step": 643 + }, + { + "epoch": 0.1465301478953356, + "grad_norm": 1.7918533159116738, + "learning_rate": 1.247361190109973e-06, + "loss": 0.1104, + "step": 644 + }, + { + "epoch": 0.14675767918088736, + "grad_norm": 2.8984966135096566, + "learning_rate": 1.2473529817238016e-06, + "loss": 0.1755, + "step": 645 + }, + { + "epoch": 0.14698521046643914, + "grad_norm": 1.9091822418599347, + "learning_rate": 1.2473447606178754e-06, + "loss": 0.1077, + "step": 646 + }, + { + "epoch": 0.1472127417519909, + "grad_norm": 4.199288030915391, + "learning_rate": 1.2473365267923617e-06, + "loss": 0.2124, + "step": 647 + }, + { + "epoch": 0.14744027303754267, + "grad_norm": 2.331859473332942, + "learning_rate": 1.2473282802474293e-06, + "loss": 0.1576, + "step": 648 + }, + { + "epoch": 0.14766780432309443, + "grad_norm": 3.5722786659910577, + "learning_rate": 1.2473200209832465e-06, + "loss": 0.2027, + "step": 649 + }, + { + "epoch": 0.14789533560864618, + "grad_norm": 1.5390826591189062, + "learning_rate": 1.2473117489999823e-06, + "loss": 0.161, + "step": 650 + }, + { + "epoch": 0.14812286689419796, + "grad_norm": 2.741044883004237, + "learning_rate": 1.2473034642978057e-06, + "loss": 0.1656, + "step": 651 + }, + { + "epoch": 0.1483503981797497, + "grad_norm": 2.2681711762464034, + "learning_rate": 1.247295166876886e-06, + "loss": 0.1254, + "step": 652 + }, + { + "epoch": 0.1485779294653015, + "grad_norm": 2.2254637289761194, + "learning_rate": 1.2472868567373924e-06, + "loss": 0.1291, + "step": 653 + }, + { + "epoch": 0.14880546075085324, + "grad_norm": 2.213517163461755, + "learning_rate": 1.2472785338794953e-06, + "loss": 0.1541, + "step": 654 + }, + { + "epoch": 0.149032992036405, + "grad_norm": 1.6789308605390307, + "learning_rate": 1.247270198303365e-06, + "loss": 0.1316, + "step": 655 + }, + { + "epoch": 0.14926052332195677, + "grad_norm": 2.179149997459725, + "learning_rate": 1.247261850009171e-06, + "loss": 0.2437, + "step": 656 + }, + { + "epoch": 0.14948805460750852, + "grad_norm": 2.910894270371587, + "learning_rate": 1.2472534889970848e-06, + "loss": 0.2038, + "step": 657 + }, + { + "epoch": 0.1497155858930603, + "grad_norm": 1.751607816792672, + "learning_rate": 1.2472451152672766e-06, + "loss": 0.1164, + "step": 658 + }, + { + "epoch": 0.14994311717861206, + "grad_norm": 1.6602009490349432, + "learning_rate": 1.2472367288199177e-06, + "loss": 0.1193, + "step": 659 + }, + { + "epoch": 0.15017064846416384, + "grad_norm": 2.038150970938399, + "learning_rate": 1.2472283296551798e-06, + "loss": 0.102, + "step": 660 + }, + { + "epoch": 0.1503981797497156, + "grad_norm": 2.1439804373776936, + "learning_rate": 1.2472199177732346e-06, + "loss": 0.1502, + "step": 661 + }, + { + "epoch": 0.15062571103526734, + "grad_norm": 2.5777822840030358, + "learning_rate": 1.2472114931742537e-06, + "loss": 0.1168, + "step": 662 + }, + { + "epoch": 0.15085324232081912, + "grad_norm": 2.4175964563163177, + "learning_rate": 1.2472030558584093e-06, + "loss": 0.1035, + "step": 663 + }, + { + "epoch": 0.15108077360637087, + "grad_norm": 2.635267423704016, + "learning_rate": 1.2471946058258742e-06, + "loss": 0.1701, + "step": 664 + }, + { + "epoch": 0.15130830489192265, + "grad_norm": 1.9337561786859772, + "learning_rate": 1.2471861430768205e-06, + "loss": 0.1075, + "step": 665 + }, + { + "epoch": 0.1515358361774744, + "grad_norm": 1.7937795679496227, + "learning_rate": 1.2471776676114217e-06, + "loss": 0.1785, + "step": 666 + }, + { + "epoch": 0.15176336746302616, + "grad_norm": 3.0588710289274816, + "learning_rate": 1.2471691794298508e-06, + "loss": 0.1798, + "step": 667 + }, + { + "epoch": 0.15199089874857794, + "grad_norm": 2.638986072752188, + "learning_rate": 1.2471606785322814e-06, + "loss": 0.0878, + "step": 668 + }, + { + "epoch": 0.1522184300341297, + "grad_norm": 2.732712357601826, + "learning_rate": 1.247152164918887e-06, + "loss": 0.1267, + "step": 669 + }, + { + "epoch": 0.15244596131968147, + "grad_norm": 1.7481991977105777, + "learning_rate": 1.247143638589842e-06, + "loss": 0.1584, + "step": 670 + }, + { + "epoch": 0.15267349260523322, + "grad_norm": 2.794672743532085, + "learning_rate": 1.2471350995453203e-06, + "loss": 0.1584, + "step": 671 + }, + { + "epoch": 0.15290102389078497, + "grad_norm": 3.1279366528301633, + "learning_rate": 1.2471265477854966e-06, + "loss": 0.148, + "step": 672 + }, + { + "epoch": 0.15312855517633675, + "grad_norm": 3.920575109905724, + "learning_rate": 1.2471179833105454e-06, + "loss": 0.1732, + "step": 673 + }, + { + "epoch": 0.1533560864618885, + "grad_norm": 1.7916571238390178, + "learning_rate": 1.2471094061206422e-06, + "loss": 0.2336, + "step": 674 + }, + { + "epoch": 0.15358361774744028, + "grad_norm": 1.7363850632393116, + "learning_rate": 1.247100816215962e-06, + "loss": 0.1244, + "step": 675 + }, + { + "epoch": 0.15381114903299203, + "grad_norm": 2.504377712379844, + "learning_rate": 1.2470922135966806e-06, + "loss": 0.1674, + "step": 676 + }, + { + "epoch": 0.1540386803185438, + "grad_norm": 2.43043947984636, + "learning_rate": 1.2470835982629736e-06, + "loss": 0.1249, + "step": 677 + }, + { + "epoch": 0.15426621160409557, + "grad_norm": 3.950497364660697, + "learning_rate": 1.247074970215017e-06, + "loss": 0.2401, + "step": 678 + }, + { + "epoch": 0.15449374288964732, + "grad_norm": 3.1492013494233846, + "learning_rate": 1.2470663294529873e-06, + "loss": 0.1605, + "step": 679 + }, + { + "epoch": 0.1547212741751991, + "grad_norm": 1.80598204305421, + "learning_rate": 1.2470576759770612e-06, + "loss": 0.113, + "step": 680 + }, + { + "epoch": 0.15494880546075085, + "grad_norm": 2.0454054940402506, + "learning_rate": 1.2470490097874155e-06, + "loss": 0.1453, + "step": 681 + }, + { + "epoch": 0.15517633674630263, + "grad_norm": 3.6952564849548053, + "learning_rate": 1.247040330884227e-06, + "loss": 0.1581, + "step": 682 + }, + { + "epoch": 0.15540386803185438, + "grad_norm": 2.3655397835651075, + "learning_rate": 1.2470316392676738e-06, + "loss": 0.169, + "step": 683 + }, + { + "epoch": 0.15563139931740613, + "grad_norm": 3.416348712472315, + "learning_rate": 1.2470229349379326e-06, + "loss": 0.1347, + "step": 684 + }, + { + "epoch": 0.1558589306029579, + "grad_norm": 2.618995350775909, + "learning_rate": 1.2470142178951822e-06, + "loss": 0.1924, + "step": 685 + }, + { + "epoch": 0.15608646188850966, + "grad_norm": 1.344663220923034, + "learning_rate": 1.2470054881396002e-06, + "loss": 0.2013, + "step": 686 + }, + { + "epoch": 0.15631399317406144, + "grad_norm": 1.1568986493989724, + "learning_rate": 1.246996745671365e-06, + "loss": 0.131, + "step": 687 + }, + { + "epoch": 0.1565415244596132, + "grad_norm": 3.0558312091963473, + "learning_rate": 1.2469879904906556e-06, + "loss": 0.14, + "step": 688 + }, + { + "epoch": 0.15676905574516495, + "grad_norm": 4.767157427966137, + "learning_rate": 1.2469792225976507e-06, + "loss": 0.156, + "step": 689 + }, + { + "epoch": 0.15699658703071673, + "grad_norm": 1.9971770266956603, + "learning_rate": 1.2469704419925296e-06, + "loss": 0.1413, + "step": 690 + }, + { + "epoch": 0.15722411831626848, + "grad_norm": 3.560138993273607, + "learning_rate": 1.246961648675472e-06, + "loss": 0.2274, + "step": 691 + }, + { + "epoch": 0.15745164960182026, + "grad_norm": 1.8091873297743188, + "learning_rate": 1.246952842646657e-06, + "loss": 0.2606, + "step": 692 + }, + { + "epoch": 0.157679180887372, + "grad_norm": 1.9524492716137443, + "learning_rate": 1.2469440239062653e-06, + "loss": 0.1888, + "step": 693 + }, + { + "epoch": 0.15790671217292376, + "grad_norm": 1.978419283294589, + "learning_rate": 1.2469351924544766e-06, + "loss": 0.168, + "step": 694 + }, + { + "epoch": 0.15813424345847554, + "grad_norm": 1.909977232991382, + "learning_rate": 1.2469263482914716e-06, + "loss": 0.1302, + "step": 695 + }, + { + "epoch": 0.1583617747440273, + "grad_norm": 2.786836009335205, + "learning_rate": 1.246917491417431e-06, + "loss": 0.1603, + "step": 696 + }, + { + "epoch": 0.15858930602957907, + "grad_norm": 2.700038379786115, + "learning_rate": 1.246908621832536e-06, + "loss": 0.2268, + "step": 697 + }, + { + "epoch": 0.15881683731513083, + "grad_norm": 1.4116863857464026, + "learning_rate": 1.2468997395369677e-06, + "loss": 0.1761, + "step": 698 + }, + { + "epoch": 0.1590443686006826, + "grad_norm": 2.8928190492615133, + "learning_rate": 1.2468908445309077e-06, + "loss": 0.1789, + "step": 699 + }, + { + "epoch": 0.15927189988623436, + "grad_norm": 1.650749552825084, + "learning_rate": 1.2468819368145376e-06, + "loss": 0.1324, + "step": 700 + }, + { + "epoch": 0.1594994311717861, + "grad_norm": 2.3722473947353677, + "learning_rate": 1.2468730163880398e-06, + "loss": 0.1116, + "step": 701 + }, + { + "epoch": 0.1597269624573379, + "grad_norm": 2.879822957568519, + "learning_rate": 1.2468640832515962e-06, + "loss": 0.0564, + "step": 702 + }, + { + "epoch": 0.15995449374288964, + "grad_norm": 2.162764734574199, + "learning_rate": 1.24685513740539e-06, + "loss": 0.1739, + "step": 703 + }, + { + "epoch": 0.16018202502844142, + "grad_norm": 2.8968364936480206, + "learning_rate": 1.2468461788496036e-06, + "loss": 0.2091, + "step": 704 + }, + { + "epoch": 0.16040955631399317, + "grad_norm": 1.8559610510087743, + "learning_rate": 1.24683720758442e-06, + "loss": 0.1533, + "step": 705 + }, + { + "epoch": 0.16063708759954493, + "grad_norm": 2.184281056476426, + "learning_rate": 1.2468282236100226e-06, + "loss": 0.1582, + "step": 706 + }, + { + "epoch": 0.1608646188850967, + "grad_norm": 1.3209438595657337, + "learning_rate": 1.2468192269265955e-06, + "loss": 0.1914, + "step": 707 + }, + { + "epoch": 0.16109215017064846, + "grad_norm": 2.1470386790088174, + "learning_rate": 1.246810217534322e-06, + "loss": 0.0831, + "step": 708 + }, + { + "epoch": 0.16131968145620024, + "grad_norm": 1.594792083731403, + "learning_rate": 1.2468011954333864e-06, + "loss": 0.1349, + "step": 709 + }, + { + "epoch": 0.161547212741752, + "grad_norm": 1.9899900139983586, + "learning_rate": 1.2467921606239734e-06, + "loss": 0.1406, + "step": 710 + }, + { + "epoch": 0.16177474402730374, + "grad_norm": 2.161056989124219, + "learning_rate": 1.2467831131062672e-06, + "loss": 0.1186, + "step": 711 + }, + { + "epoch": 0.16200227531285552, + "grad_norm": 3.2786168252573438, + "learning_rate": 1.2467740528804528e-06, + "loss": 0.1525, + "step": 712 + }, + { + "epoch": 0.16222980659840727, + "grad_norm": 2.152367629184536, + "learning_rate": 1.2467649799467156e-06, + "loss": 0.1403, + "step": 713 + }, + { + "epoch": 0.16245733788395905, + "grad_norm": 2.658644939282435, + "learning_rate": 1.246755894305241e-06, + "loss": 0.1287, + "step": 714 + }, + { + "epoch": 0.1626848691695108, + "grad_norm": 1.8320157906526173, + "learning_rate": 1.2467467959562143e-06, + "loss": 0.1489, + "step": 715 + }, + { + "epoch": 0.16291240045506258, + "grad_norm": 3.0792158572997526, + "learning_rate": 1.2467376848998221e-06, + "loss": 0.1929, + "step": 716 + }, + { + "epoch": 0.16313993174061434, + "grad_norm": 2.592666663523021, + "learning_rate": 1.2467285611362501e-06, + "loss": 0.1198, + "step": 717 + }, + { + "epoch": 0.1633674630261661, + "grad_norm": 2.3270639642215123, + "learning_rate": 1.2467194246656851e-06, + "loss": 0.119, + "step": 718 + }, + { + "epoch": 0.16359499431171787, + "grad_norm": 1.5662096056295784, + "learning_rate": 1.2467102754883136e-06, + "loss": 0.1488, + "step": 719 + }, + { + "epoch": 0.16382252559726962, + "grad_norm": 2.0754259992407174, + "learning_rate": 1.2467011136043228e-06, + "loss": 0.1206, + "step": 720 + }, + { + "epoch": 0.1640500568828214, + "grad_norm": 2.377809704915352, + "learning_rate": 1.2466919390138995e-06, + "loss": 0.2349, + "step": 721 + }, + { + "epoch": 0.16427758816837315, + "grad_norm": 2.1373727350700205, + "learning_rate": 1.246682751717232e-06, + "loss": 0.1333, + "step": 722 + }, + { + "epoch": 0.1645051194539249, + "grad_norm": 3.8601459911234697, + "learning_rate": 1.2466735517145074e-06, + "loss": 0.3259, + "step": 723 + }, + { + "epoch": 0.16473265073947668, + "grad_norm": 2.1273982856593614, + "learning_rate": 1.2466643390059138e-06, + "loss": 0.199, + "step": 724 + }, + { + "epoch": 0.16496018202502843, + "grad_norm": 2.274158988300012, + "learning_rate": 1.2466551135916398e-06, + "loss": 0.1351, + "step": 725 + }, + { + "epoch": 0.16518771331058021, + "grad_norm": 2.1566789936379287, + "learning_rate": 1.2466458754718737e-06, + "loss": 0.219, + "step": 726 + }, + { + "epoch": 0.16541524459613197, + "grad_norm": 3.388462178150055, + "learning_rate": 1.2466366246468045e-06, + "loss": 0.1456, + "step": 727 + }, + { + "epoch": 0.16564277588168372, + "grad_norm": 2.792548754369155, + "learning_rate": 1.246627361116621e-06, + "loss": 0.2178, + "step": 728 + }, + { + "epoch": 0.1658703071672355, + "grad_norm": 1.7787275123381943, + "learning_rate": 1.246618084881513e-06, + "loss": 0.2584, + "step": 729 + }, + { + "epoch": 0.16609783845278725, + "grad_norm": 2.150845029279013, + "learning_rate": 1.2466087959416695e-06, + "loss": 0.1474, + "step": 730 + }, + { + "epoch": 0.16632536973833903, + "grad_norm": 3.4162019984229213, + "learning_rate": 1.2465994942972805e-06, + "loss": 0.1415, + "step": 731 + }, + { + "epoch": 0.16655290102389078, + "grad_norm": 3.5172418167047743, + "learning_rate": 1.2465901799485366e-06, + "loss": 0.2267, + "step": 732 + }, + { + "epoch": 0.16678043230944256, + "grad_norm": 1.9664520821504867, + "learning_rate": 1.2465808528956277e-06, + "loss": 0.1027, + "step": 733 + }, + { + "epoch": 0.1670079635949943, + "grad_norm": 2.053925645911197, + "learning_rate": 1.2465715131387446e-06, + "loss": 0.1405, + "step": 734 + }, + { + "epoch": 0.16723549488054607, + "grad_norm": 1.6417683696863474, + "learning_rate": 1.2465621606780778e-06, + "loss": 0.1804, + "step": 735 + }, + { + "epoch": 0.16746302616609784, + "grad_norm": 1.9532511665276102, + "learning_rate": 1.2465527955138191e-06, + "loss": 0.1438, + "step": 736 + }, + { + "epoch": 0.1676905574516496, + "grad_norm": 2.7978077296538295, + "learning_rate": 1.2465434176461596e-06, + "loss": 0.1806, + "step": 737 + }, + { + "epoch": 0.16791808873720138, + "grad_norm": 1.7861222447513503, + "learning_rate": 1.2465340270752908e-06, + "loss": 0.0953, + "step": 738 + }, + { + "epoch": 0.16814562002275313, + "grad_norm": 1.2545980680473232, + "learning_rate": 1.2465246238014047e-06, + "loss": 0.0881, + "step": 739 + }, + { + "epoch": 0.16837315130830488, + "grad_norm": 2.49195685975364, + "learning_rate": 1.2465152078246936e-06, + "loss": 0.1643, + "step": 740 + }, + { + "epoch": 0.16860068259385666, + "grad_norm": 2.0211233157427637, + "learning_rate": 1.24650577914535e-06, + "loss": 0.1263, + "step": 741 + }, + { + "epoch": 0.1688282138794084, + "grad_norm": 2.7858317155477317, + "learning_rate": 1.2464963377635667e-06, + "loss": 0.1547, + "step": 742 + }, + { + "epoch": 0.1690557451649602, + "grad_norm": 1.7097291360774547, + "learning_rate": 1.246486883679536e-06, + "loss": 0.2516, + "step": 743 + }, + { + "epoch": 0.16928327645051194, + "grad_norm": 3.9137648292026737, + "learning_rate": 1.246477416893452e-06, + "loss": 0.2036, + "step": 744 + }, + { + "epoch": 0.1695108077360637, + "grad_norm": 3.005605654107358, + "learning_rate": 1.2464679374055074e-06, + "loss": 0.1481, + "step": 745 + }, + { + "epoch": 0.16973833902161548, + "grad_norm": 3.401532765227879, + "learning_rate": 1.2464584452158968e-06, + "loss": 0.1841, + "step": 746 + }, + { + "epoch": 0.16996587030716723, + "grad_norm": 2.843140048954733, + "learning_rate": 1.2464489403248133e-06, + "loss": 0.184, + "step": 747 + }, + { + "epoch": 0.170193401592719, + "grad_norm": 1.515779223289782, + "learning_rate": 1.246439422732452e-06, + "loss": 0.1262, + "step": 748 + }, + { + "epoch": 0.17042093287827076, + "grad_norm": 2.618293101772126, + "learning_rate": 1.2464298924390066e-06, + "loss": 0.1415, + "step": 749 + }, + { + "epoch": 0.17064846416382254, + "grad_norm": 2.248269138511338, + "learning_rate": 1.2464203494446725e-06, + "loss": 0.185, + "step": 750 + }, + { + "epoch": 0.1708759954493743, + "grad_norm": 1.3558978429200024, + "learning_rate": 1.2464107937496444e-06, + "loss": 0.096, + "step": 751 + }, + { + "epoch": 0.17110352673492604, + "grad_norm": 1.8355286869437153, + "learning_rate": 1.246401225354118e-06, + "loss": 0.0936, + "step": 752 + }, + { + "epoch": 0.17133105802047782, + "grad_norm": 2.611386377303649, + "learning_rate": 1.2463916442582883e-06, + "loss": 0.2058, + "step": 753 + }, + { + "epoch": 0.17155858930602957, + "grad_norm": 1.81511526173022, + "learning_rate": 1.2463820504623516e-06, + "loss": 0.0722, + "step": 754 + }, + { + "epoch": 0.17178612059158135, + "grad_norm": 1.6836561465138316, + "learning_rate": 1.246372443966504e-06, + "loss": 0.1419, + "step": 755 + }, + { + "epoch": 0.1720136518771331, + "grad_norm": 3.189715404864015, + "learning_rate": 1.246362824770941e-06, + "loss": 0.1604, + "step": 756 + }, + { + "epoch": 0.17224118316268486, + "grad_norm": 2.8556456489625193, + "learning_rate": 1.2463531928758605e-06, + "loss": 0.1793, + "step": 757 + }, + { + "epoch": 0.17246871444823664, + "grad_norm": 2.1490228034084344, + "learning_rate": 1.2463435482814585e-06, + "loss": 0.1928, + "step": 758 + }, + { + "epoch": 0.1726962457337884, + "grad_norm": 1.866877451814791, + "learning_rate": 1.246333890987932e-06, + "loss": 0.2064, + "step": 759 + }, + { + "epoch": 0.17292377701934017, + "grad_norm": 2.7361601673612284, + "learning_rate": 1.246324220995479e-06, + "loss": 0.1024, + "step": 760 + }, + { + "epoch": 0.17315130830489192, + "grad_norm": 3.6715173407277004, + "learning_rate": 1.2463145383042966e-06, + "loss": 0.1741, + "step": 761 + }, + { + "epoch": 0.17337883959044367, + "grad_norm": 4.388914943676026, + "learning_rate": 1.2463048429145832e-06, + "loss": 0.2951, + "step": 762 + }, + { + "epoch": 0.17360637087599545, + "grad_norm": 3.0864567661578075, + "learning_rate": 1.2462951348265364e-06, + "loss": 0.1681, + "step": 763 + }, + { + "epoch": 0.1738339021615472, + "grad_norm": 2.2429137189515487, + "learning_rate": 1.2462854140403553e-06, + "loss": 0.1698, + "step": 764 + }, + { + "epoch": 0.17406143344709898, + "grad_norm": 3.7655750343422487, + "learning_rate": 1.2462756805562378e-06, + "loss": 0.1972, + "step": 765 + }, + { + "epoch": 0.17428896473265074, + "grad_norm": 1.4821109763148475, + "learning_rate": 1.2462659343743832e-06, + "loss": 0.1144, + "step": 766 + }, + { + "epoch": 0.17451649601820252, + "grad_norm": 2.9261323093043234, + "learning_rate": 1.2462561754949908e-06, + "loss": 0.1354, + "step": 767 + }, + { + "epoch": 0.17474402730375427, + "grad_norm": 2.021278631174851, + "learning_rate": 1.2462464039182598e-06, + "loss": 0.1158, + "step": 768 + }, + { + "epoch": 0.17497155858930602, + "grad_norm": 2.189903163956334, + "learning_rate": 1.2462366196443903e-06, + "loss": 0.1587, + "step": 769 + }, + { + "epoch": 0.1751990898748578, + "grad_norm": 3.7285174958892364, + "learning_rate": 1.246226822673582e-06, + "loss": 0.2024, + "step": 770 + }, + { + "epoch": 0.17542662116040955, + "grad_norm": 1.9007743093993184, + "learning_rate": 1.2462170130060351e-06, + "loss": 0.1025, + "step": 771 + }, + { + "epoch": 0.17565415244596133, + "grad_norm": 3.3341124392840134, + "learning_rate": 1.24620719064195e-06, + "loss": 0.1718, + "step": 772 + }, + { + "epoch": 0.17588168373151308, + "grad_norm": 2.271177623744295, + "learning_rate": 1.246197355581528e-06, + "loss": 0.1713, + "step": 773 + }, + { + "epoch": 0.17610921501706484, + "grad_norm": 2.631276315974309, + "learning_rate": 1.2461875078249694e-06, + "loss": 0.1769, + "step": 774 + }, + { + "epoch": 0.17633674630261661, + "grad_norm": 2.2924143983188765, + "learning_rate": 1.246177647372476e-06, + "loss": 0.1155, + "step": 775 + }, + { + "epoch": 0.17656427758816837, + "grad_norm": 4.145219852575127, + "learning_rate": 1.246167774224249e-06, + "loss": 0.1997, + "step": 776 + }, + { + "epoch": 0.17679180887372015, + "grad_norm": 3.5955716696986237, + "learning_rate": 1.2461578883804903e-06, + "loss": 0.1434, + "step": 777 + }, + { + "epoch": 0.1770193401592719, + "grad_norm": 3.5823237759342477, + "learning_rate": 1.246147989841402e-06, + "loss": 0.131, + "step": 778 + }, + { + "epoch": 0.17724687144482365, + "grad_norm": 1.7885388560764315, + "learning_rate": 1.2461380786071863e-06, + "loss": 0.0755, + "step": 779 + }, + { + "epoch": 0.17747440273037543, + "grad_norm": 2.362853335883513, + "learning_rate": 1.246128154678046e-06, + "loss": 0.1285, + "step": 780 + }, + { + "epoch": 0.17770193401592718, + "grad_norm": 2.826403481752188, + "learning_rate": 1.2461182180541835e-06, + "loss": 0.0898, + "step": 781 + }, + { + "epoch": 0.17792946530147896, + "grad_norm": 5.793503549962082, + "learning_rate": 1.2461082687358022e-06, + "loss": 0.0971, + "step": 782 + }, + { + "epoch": 0.1781569965870307, + "grad_norm": 1.8035940463938722, + "learning_rate": 1.2460983067231055e-06, + "loss": 0.1105, + "step": 783 + }, + { + "epoch": 0.1783845278725825, + "grad_norm": 2.3286047675537613, + "learning_rate": 1.246088332016297e-06, + "loss": 0.0997, + "step": 784 + }, + { + "epoch": 0.17861205915813425, + "grad_norm": 2.4331158536688067, + "learning_rate": 1.2460783446155802e-06, + "loss": 0.2145, + "step": 785 + }, + { + "epoch": 0.178839590443686, + "grad_norm": 2.4301917574272234, + "learning_rate": 1.2460683445211596e-06, + "loss": 0.1826, + "step": 786 + }, + { + "epoch": 0.17906712172923778, + "grad_norm": 3.191042960124482, + "learning_rate": 1.2460583317332395e-06, + "loss": 0.2224, + "step": 787 + }, + { + "epoch": 0.17929465301478953, + "grad_norm": 1.9281932990563415, + "learning_rate": 1.2460483062520246e-06, + "loss": 0.1012, + "step": 788 + }, + { + "epoch": 0.1795221843003413, + "grad_norm": 1.9401318974845003, + "learning_rate": 1.2460382680777196e-06, + "loss": 0.0761, + "step": 789 + }, + { + "epoch": 0.17974971558589306, + "grad_norm": 13.086161362963225, + "learning_rate": 1.2460282172105298e-06, + "loss": 0.2088, + "step": 790 + }, + { + "epoch": 0.1799772468714448, + "grad_norm": 1.4783130702588718, + "learning_rate": 1.2460181536506608e-06, + "loss": 0.2126, + "step": 791 + }, + { + "epoch": 0.1802047781569966, + "grad_norm": 2.4964786740518763, + "learning_rate": 1.2460080773983177e-06, + "loss": 0.1385, + "step": 792 + }, + { + "epoch": 0.18043230944254834, + "grad_norm": 2.7778972521749545, + "learning_rate": 1.2459979884537072e-06, + "loss": 0.1448, + "step": 793 + }, + { + "epoch": 0.18065984072810012, + "grad_norm": 2.167813491126184, + "learning_rate": 1.2459878868170348e-06, + "loss": 0.1379, + "step": 794 + }, + { + "epoch": 0.18088737201365188, + "grad_norm": 1.9654699615947284, + "learning_rate": 1.2459777724885075e-06, + "loss": 0.1314, + "step": 795 + }, + { + "epoch": 0.18111490329920363, + "grad_norm": 2.293952257528565, + "learning_rate": 1.2459676454683318e-06, + "loss": 0.1695, + "step": 796 + }, + { + "epoch": 0.1813424345847554, + "grad_norm": 3.9215044200778144, + "learning_rate": 1.2459575057567144e-06, + "loss": 0.2204, + "step": 797 + }, + { + "epoch": 0.18156996587030716, + "grad_norm": 2.8214133097210117, + "learning_rate": 1.245947353353863e-06, + "loss": 0.1558, + "step": 798 + }, + { + "epoch": 0.18179749715585894, + "grad_norm": 5.317020653859289, + "learning_rate": 1.245937188259985e-06, + "loss": 0.2603, + "step": 799 + }, + { + "epoch": 0.1820250284414107, + "grad_norm": 4.004955818619992, + "learning_rate": 1.245927010475288e-06, + "loss": 0.1196, + "step": 800 + }, + { + "epoch": 0.18225255972696247, + "grad_norm": 3.792524464667178, + "learning_rate": 1.24591681999998e-06, + "loss": 0.1821, + "step": 801 + }, + { + "epoch": 0.18248009101251422, + "grad_norm": 2.813011742342484, + "learning_rate": 1.2459066168342693e-06, + "loss": 0.1513, + "step": 802 + }, + { + "epoch": 0.18270762229806597, + "grad_norm": 3.511510747002315, + "learning_rate": 1.2458964009783646e-06, + "loss": 0.2163, + "step": 803 + }, + { + "epoch": 0.18293515358361775, + "grad_norm": 2.802158661308834, + "learning_rate": 1.2458861724324745e-06, + "loss": 0.1963, + "step": 804 + }, + { + "epoch": 0.1831626848691695, + "grad_norm": 3.64850186041969, + "learning_rate": 1.2458759311968084e-06, + "loss": 0.303, + "step": 805 + }, + { + "epoch": 0.1833902161547213, + "grad_norm": 2.6182595326596725, + "learning_rate": 1.245865677271575e-06, + "loss": 0.1456, + "step": 806 + }, + { + "epoch": 0.18361774744027304, + "grad_norm": 2.399741320725503, + "learning_rate": 1.2458554106569844e-06, + "loss": 0.2288, + "step": 807 + }, + { + "epoch": 0.1838452787258248, + "grad_norm": 1.252106549654472, + "learning_rate": 1.2458451313532463e-06, + "loss": 0.0801, + "step": 808 + }, + { + "epoch": 0.18407281001137657, + "grad_norm": 3.696224132577839, + "learning_rate": 1.2458348393605708e-06, + "loss": 0.2059, + "step": 809 + }, + { + "epoch": 0.18430034129692832, + "grad_norm": 1.3783330613855644, + "learning_rate": 1.2458245346791678e-06, + "loss": 0.1164, + "step": 810 + }, + { + "epoch": 0.1845278725824801, + "grad_norm": 1.5623432135982267, + "learning_rate": 1.2458142173092486e-06, + "loss": 0.176, + "step": 811 + }, + { + "epoch": 0.18475540386803185, + "grad_norm": 6.552053967433837, + "learning_rate": 1.2458038872510237e-06, + "loss": 0.118, + "step": 812 + }, + { + "epoch": 0.1849829351535836, + "grad_norm": 3.2237210845046964, + "learning_rate": 1.2457935445047042e-06, + "loss": 0.1875, + "step": 813 + }, + { + "epoch": 0.18521046643913538, + "grad_norm": 1.7463109516387256, + "learning_rate": 1.2457831890705018e-06, + "loss": 0.1945, + "step": 814 + }, + { + "epoch": 0.18543799772468714, + "grad_norm": 2.8292409598595953, + "learning_rate": 1.2457728209486279e-06, + "loss": 0.1711, + "step": 815 + }, + { + "epoch": 0.18566552901023892, + "grad_norm": 3.198074487753419, + "learning_rate": 1.2457624401392943e-06, + "loss": 0.2552, + "step": 816 + }, + { + "epoch": 0.18589306029579067, + "grad_norm": 3.2293783551138278, + "learning_rate": 1.2457520466427135e-06, + "loss": 0.1955, + "step": 817 + }, + { + "epoch": 0.18612059158134245, + "grad_norm": 2.5604778410965383, + "learning_rate": 1.2457416404590974e-06, + "loss": 0.1689, + "step": 818 + }, + { + "epoch": 0.1863481228668942, + "grad_norm": 2.4475267016374427, + "learning_rate": 1.2457312215886592e-06, + "loss": 0.1165, + "step": 819 + }, + { + "epoch": 0.18657565415244595, + "grad_norm": 1.9856047790588058, + "learning_rate": 1.2457207900316115e-06, + "loss": 0.195, + "step": 820 + }, + { + "epoch": 0.18680318543799773, + "grad_norm": 3.030251865029441, + "learning_rate": 1.245710345788168e-06, + "loss": 0.2233, + "step": 821 + }, + { + "epoch": 0.18703071672354948, + "grad_norm": 6.914472069589314, + "learning_rate": 1.2456998888585414e-06, + "loss": 0.1294, + "step": 822 + }, + { + "epoch": 0.18725824800910126, + "grad_norm": 1.5392801223632877, + "learning_rate": 1.245689419242946e-06, + "loss": 0.1031, + "step": 823 + }, + { + "epoch": 0.18748577929465302, + "grad_norm": 1.5563008585328006, + "learning_rate": 1.2456789369415955e-06, + "loss": 0.1233, + "step": 824 + }, + { + "epoch": 0.18771331058020477, + "grad_norm": 1.5005319006316646, + "learning_rate": 1.2456684419547044e-06, + "loss": 0.1698, + "step": 825 + }, + { + "epoch": 0.18794084186575655, + "grad_norm": 2.5311436309198245, + "learning_rate": 1.245657934282487e-06, + "loss": 0.1242, + "step": 826 + }, + { + "epoch": 0.1881683731513083, + "grad_norm": 1.3382771790085715, + "learning_rate": 1.245647413925158e-06, + "loss": 0.1173, + "step": 827 + }, + { + "epoch": 0.18839590443686008, + "grad_norm": 2.455502403566395, + "learning_rate": 1.2456368808829327e-06, + "loss": 0.0912, + "step": 828 + }, + { + "epoch": 0.18862343572241183, + "grad_norm": 2.9752303589937212, + "learning_rate": 1.2456263351560261e-06, + "loss": 0.2599, + "step": 829 + }, + { + "epoch": 0.18885096700796358, + "grad_norm": 5.043835077918359, + "learning_rate": 1.2456157767446538e-06, + "loss": 0.1609, + "step": 830 + }, + { + "epoch": 0.18907849829351536, + "grad_norm": 2.756359704558054, + "learning_rate": 1.245605205649032e-06, + "loss": 0.1323, + "step": 831 + }, + { + "epoch": 0.18930602957906711, + "grad_norm": 1.835440265718024, + "learning_rate": 1.245594621869376e-06, + "loss": 0.2094, + "step": 832 + }, + { + "epoch": 0.1895335608646189, + "grad_norm": 1.2880237601014817, + "learning_rate": 1.2455840254059026e-06, + "loss": 0.1085, + "step": 833 + }, + { + "epoch": 0.18976109215017065, + "grad_norm": 1.4808086873300856, + "learning_rate": 1.2455734162588282e-06, + "loss": 0.1067, + "step": 834 + }, + { + "epoch": 0.1899886234357224, + "grad_norm": 2.3351983872627597, + "learning_rate": 1.2455627944283697e-06, + "loss": 0.1493, + "step": 835 + }, + { + "epoch": 0.19021615472127418, + "grad_norm": 2.422722379821762, + "learning_rate": 1.245552159914744e-06, + "loss": 0.1387, + "step": 836 + }, + { + "epoch": 0.19044368600682593, + "grad_norm": 2.2005548282870477, + "learning_rate": 1.245541512718169e-06, + "loss": 0.1047, + "step": 837 + }, + { + "epoch": 0.1906712172923777, + "grad_norm": 2.379475571028047, + "learning_rate": 1.245530852838862e-06, + "loss": 0.1524, + "step": 838 + }, + { + "epoch": 0.19089874857792946, + "grad_norm": 1.669935289366072, + "learning_rate": 1.2455201802770405e-06, + "loss": 0.157, + "step": 839 + }, + { + "epoch": 0.19112627986348124, + "grad_norm": 2.357020791051429, + "learning_rate": 1.245509495032923e-06, + "loss": 0.2156, + "step": 840 + }, + { + "epoch": 0.191353811149033, + "grad_norm": 3.871602599108809, + "learning_rate": 1.2454987971067278e-06, + "loss": 0.1557, + "step": 841 + }, + { + "epoch": 0.19158134243458474, + "grad_norm": 2.5332197020943887, + "learning_rate": 1.2454880864986737e-06, + "loss": 0.1644, + "step": 842 + }, + { + "epoch": 0.19180887372013652, + "grad_norm": 3.1286962973408596, + "learning_rate": 1.2454773632089795e-06, + "loss": 0.0794, + "step": 843 + }, + { + "epoch": 0.19203640500568828, + "grad_norm": 2.3210649274985666, + "learning_rate": 1.2454666272378644e-06, + "loss": 0.129, + "step": 844 + }, + { + "epoch": 0.19226393629124006, + "grad_norm": 3.000200402253768, + "learning_rate": 1.2454558785855475e-06, + "loss": 0.1628, + "step": 845 + }, + { + "epoch": 0.1924914675767918, + "grad_norm": 2.3643323080869902, + "learning_rate": 1.245445117252249e-06, + "loss": 0.1345, + "step": 846 + }, + { + "epoch": 0.19271899886234356, + "grad_norm": 2.532625203594351, + "learning_rate": 1.2454343432381886e-06, + "loss": 0.2082, + "step": 847 + }, + { + "epoch": 0.19294653014789534, + "grad_norm": 1.9628657145639428, + "learning_rate": 1.2454235565435862e-06, + "loss": 0.0782, + "step": 848 + }, + { + "epoch": 0.1931740614334471, + "grad_norm": 1.609178421923729, + "learning_rate": 1.2454127571686629e-06, + "loss": 0.1405, + "step": 849 + }, + { + "epoch": 0.19340159271899887, + "grad_norm": 1.7728115247069527, + "learning_rate": 1.245401945113639e-06, + "loss": 0.203, + "step": 850 + }, + { + "epoch": 0.19362912400455062, + "grad_norm": 3.2450475274049118, + "learning_rate": 1.2453911203787355e-06, + "loss": 0.1524, + "step": 851 + }, + { + "epoch": 0.19385665529010238, + "grad_norm": 22.097060091469434, + "learning_rate": 1.2453802829641736e-06, + "loss": 0.2636, + "step": 852 + }, + { + "epoch": 0.19408418657565416, + "grad_norm": 2.5365065820289496, + "learning_rate": 1.2453694328701752e-06, + "loss": 0.1019, + "step": 853 + }, + { + "epoch": 0.1943117178612059, + "grad_norm": 2.090322149834491, + "learning_rate": 1.2453585700969614e-06, + "loss": 0.1498, + "step": 854 + }, + { + "epoch": 0.1945392491467577, + "grad_norm": 2.6606765925685787, + "learning_rate": 1.2453476946447547e-06, + "loss": 0.1398, + "step": 855 + }, + { + "epoch": 0.19476678043230944, + "grad_norm": 3.56083888144899, + "learning_rate": 1.2453368065137772e-06, + "loss": 0.1463, + "step": 856 + }, + { + "epoch": 0.19499431171786122, + "grad_norm": 2.1276836242796793, + "learning_rate": 1.2453259057042514e-06, + "loss": 0.1753, + "step": 857 + }, + { + "epoch": 0.19522184300341297, + "grad_norm": 2.5690977004159805, + "learning_rate": 1.2453149922164003e-06, + "loss": 0.1292, + "step": 858 + }, + { + "epoch": 0.19544937428896472, + "grad_norm": 4.345742784369693, + "learning_rate": 1.2453040660504468e-06, + "loss": 0.15, + "step": 859 + }, + { + "epoch": 0.1956769055745165, + "grad_norm": 3.118246879884093, + "learning_rate": 1.2452931272066141e-06, + "loss": 0.169, + "step": 860 + }, + { + "epoch": 0.19590443686006825, + "grad_norm": 2.68254786515319, + "learning_rate": 1.245282175685126e-06, + "loss": 0.157, + "step": 861 + }, + { + "epoch": 0.19613196814562003, + "grad_norm": 2.088476673647213, + "learning_rate": 1.2452712114862063e-06, + "loss": 0.1782, + "step": 862 + }, + { + "epoch": 0.19635949943117179, + "grad_norm": 1.568141769132608, + "learning_rate": 1.245260234610079e-06, + "loss": 0.1295, + "step": 863 + }, + { + "epoch": 0.19658703071672354, + "grad_norm": 2.186319656948205, + "learning_rate": 1.2452492450569682e-06, + "loss": 0.1734, + "step": 864 + }, + { + "epoch": 0.19681456200227532, + "grad_norm": 2.7655739546712135, + "learning_rate": 1.245238242827099e-06, + "loss": 0.1694, + "step": 865 + }, + { + "epoch": 0.19704209328782707, + "grad_norm": 3.0373302408208196, + "learning_rate": 1.245227227920696e-06, + "loss": 0.1356, + "step": 866 + }, + { + "epoch": 0.19726962457337885, + "grad_norm": 2.1820099415146914, + "learning_rate": 1.2452162003379842e-06, + "loss": 0.2082, + "step": 867 + }, + { + "epoch": 0.1974971558589306, + "grad_norm": 3.6721625065681827, + "learning_rate": 1.2452051600791891e-06, + "loss": 0.1915, + "step": 868 + }, + { + "epoch": 0.19772468714448235, + "grad_norm": 6.490462296454016, + "learning_rate": 1.2451941071445367e-06, + "loss": 0.1815, + "step": 869 + }, + { + "epoch": 0.19795221843003413, + "grad_norm": 3.246518762107006, + "learning_rate": 1.2451830415342524e-06, + "loss": 0.137, + "step": 870 + }, + { + "epoch": 0.19817974971558588, + "grad_norm": 2.7033364330836873, + "learning_rate": 1.2451719632485627e-06, + "loss": 0.1317, + "step": 871 + }, + { + "epoch": 0.19840728100113766, + "grad_norm": 3.30778551761739, + "learning_rate": 1.2451608722876938e-06, + "loss": 0.1099, + "step": 872 + }, + { + "epoch": 0.19863481228668942, + "grad_norm": 2.2687509460631294, + "learning_rate": 1.2451497686518722e-06, + "loss": 0.1361, + "step": 873 + }, + { + "epoch": 0.1988623435722412, + "grad_norm": 1.641721237453431, + "learning_rate": 1.2451386523413252e-06, + "loss": 0.1052, + "step": 874 + }, + { + "epoch": 0.19908987485779295, + "grad_norm": 2.206444085506852, + "learning_rate": 1.24512752335628e-06, + "loss": 0.1018, + "step": 875 + }, + { + "epoch": 0.1993174061433447, + "grad_norm": 2.210652731669232, + "learning_rate": 1.2451163816969639e-06, + "loss": 0.1879, + "step": 876 + }, + { + "epoch": 0.19954493742889648, + "grad_norm": 2.085600222270482, + "learning_rate": 1.2451052273636045e-06, + "loss": 0.127, + "step": 877 + }, + { + "epoch": 0.19977246871444823, + "grad_norm": 2.6309536592299705, + "learning_rate": 1.24509406035643e-06, + "loss": 0.1678, + "step": 878 + }, + { + "epoch": 0.2, + "grad_norm": 4.158698099165945, + "learning_rate": 1.2450828806756685e-06, + "loss": 0.2095, + "step": 879 + }, + { + "epoch": 0.20022753128555176, + "grad_norm": 2.602198490586786, + "learning_rate": 1.245071688321549e-06, + "loss": 0.1436, + "step": 880 + }, + { + "epoch": 0.20045506257110352, + "grad_norm": 2.252594865848713, + "learning_rate": 1.2450604832942991e-06, + "loss": 0.1231, + "step": 881 + }, + { + "epoch": 0.2006825938566553, + "grad_norm": 1.912453352899942, + "learning_rate": 1.245049265594149e-06, + "loss": 0.1408, + "step": 882 + }, + { + "epoch": 0.20091012514220705, + "grad_norm": 3.264942350461524, + "learning_rate": 1.2450380352213271e-06, + "loss": 0.1697, + "step": 883 + }, + { + "epoch": 0.20113765642775883, + "grad_norm": 2.415399674888119, + "learning_rate": 1.2450267921760636e-06, + "loss": 0.1331, + "step": 884 + }, + { + "epoch": 0.20136518771331058, + "grad_norm": 2.62867521080006, + "learning_rate": 1.2450155364585878e-06, + "loss": 0.1217, + "step": 885 + }, + { + "epoch": 0.20159271899886233, + "grad_norm": 2.3552959017058477, + "learning_rate": 1.2450042680691301e-06, + "loss": 0.1216, + "step": 886 + }, + { + "epoch": 0.2018202502844141, + "grad_norm": 1.4369969713280852, + "learning_rate": 1.2449929870079206e-06, + "loss": 0.1282, + "step": 887 + }, + { + "epoch": 0.20204778156996586, + "grad_norm": 2.305787931213179, + "learning_rate": 1.24498169327519e-06, + "loss": 0.1076, + "step": 888 + }, + { + "epoch": 0.20227531285551764, + "grad_norm": 1.7868835912702514, + "learning_rate": 1.2449703868711688e-06, + "loss": 0.1225, + "step": 889 + }, + { + "epoch": 0.2025028441410694, + "grad_norm": 2.1124657583403494, + "learning_rate": 1.2449590677960886e-06, + "loss": 0.1765, + "step": 890 + }, + { + "epoch": 0.20273037542662117, + "grad_norm": 1.6102832172606196, + "learning_rate": 1.2449477360501802e-06, + "loss": 0.0719, + "step": 891 + }, + { + "epoch": 0.20295790671217293, + "grad_norm": 3.8988824882283843, + "learning_rate": 1.2449363916336756e-06, + "loss": 0.1854, + "step": 892 + }, + { + "epoch": 0.20318543799772468, + "grad_norm": 3.2116126604298882, + "learning_rate": 1.2449250345468065e-06, + "loss": 0.2028, + "step": 893 + }, + { + "epoch": 0.20341296928327646, + "grad_norm": 2.083882159988442, + "learning_rate": 1.244913664789805e-06, + "loss": 0.1337, + "step": 894 + }, + { + "epoch": 0.2036405005688282, + "grad_norm": 1.8394649372022975, + "learning_rate": 1.2449022823629036e-06, + "loss": 0.1205, + "step": 895 + }, + { + "epoch": 0.20386803185438, + "grad_norm": 2.6323013014057004, + "learning_rate": 1.2448908872663347e-06, + "loss": 0.1133, + "step": 896 + }, + { + "epoch": 0.20409556313993174, + "grad_norm": 1.8291857038844686, + "learning_rate": 1.2448794795003313e-06, + "loss": 0.1142, + "step": 897 + }, + { + "epoch": 0.2043230944254835, + "grad_norm": 1.7184606914815217, + "learning_rate": 1.2448680590651269e-06, + "loss": 0.1222, + "step": 898 + }, + { + "epoch": 0.20455062571103527, + "grad_norm": 2.7034652156706716, + "learning_rate": 1.2448566259609543e-06, + "loss": 0.1991, + "step": 899 + }, + { + "epoch": 0.20477815699658702, + "grad_norm": 2.5930455129642653, + "learning_rate": 1.2448451801880476e-06, + "loss": 0.1085, + "step": 900 + }, + { + "epoch": 0.2050056882821388, + "grad_norm": 2.44560677998223, + "learning_rate": 1.2448337217466404e-06, + "loss": 0.1735, + "step": 901 + }, + { + "epoch": 0.20523321956769056, + "grad_norm": 2.257000828394708, + "learning_rate": 1.2448222506369675e-06, + "loss": 0.1118, + "step": 902 + }, + { + "epoch": 0.2054607508532423, + "grad_norm": 2.5459054260546323, + "learning_rate": 1.2448107668592626e-06, + "loss": 0.1975, + "step": 903 + }, + { + "epoch": 0.2056882821387941, + "grad_norm": 5.093888329917388, + "learning_rate": 1.244799270413761e-06, + "loss": 0.2277, + "step": 904 + }, + { + "epoch": 0.20591581342434584, + "grad_norm": 4.116266489839909, + "learning_rate": 1.2447877613006972e-06, + "loss": 0.2004, + "step": 905 + }, + { + "epoch": 0.20614334470989762, + "grad_norm": 1.8199951318249294, + "learning_rate": 1.244776239520307e-06, + "loss": 0.2131, + "step": 906 + }, + { + "epoch": 0.20637087599544937, + "grad_norm": 2.7663340604707267, + "learning_rate": 1.244764705072825e-06, + "loss": 0.2145, + "step": 907 + }, + { + "epoch": 0.20659840728100115, + "grad_norm": 1.8748872621346087, + "learning_rate": 1.2447531579584878e-06, + "loss": 0.1327, + "step": 908 + }, + { + "epoch": 0.2068259385665529, + "grad_norm": 3.4272822632320237, + "learning_rate": 1.2447415981775312e-06, + "loss": 0.2198, + "step": 909 + }, + { + "epoch": 0.20705346985210465, + "grad_norm": 3.1215491420073396, + "learning_rate": 1.2447300257301912e-06, + "loss": 0.1342, + "step": 910 + }, + { + "epoch": 0.20728100113765643, + "grad_norm": 2.5239722345332396, + "learning_rate": 1.2447184406167045e-06, + "loss": 0.1868, + "step": 911 + }, + { + "epoch": 0.2075085324232082, + "grad_norm": 1.9655955083845185, + "learning_rate": 1.2447068428373077e-06, + "loss": 0.1769, + "step": 912 + }, + { + "epoch": 0.20773606370875997, + "grad_norm": 3.157478086474276, + "learning_rate": 1.244695232392238e-06, + "loss": 0.1824, + "step": 913 + }, + { + "epoch": 0.20796359499431172, + "grad_norm": 1.9386984879122342, + "learning_rate": 1.2446836092817328e-06, + "loss": 0.1036, + "step": 914 + }, + { + "epoch": 0.20819112627986347, + "grad_norm": 2.2587342441489997, + "learning_rate": 1.2446719735060293e-06, + "loss": 0.2175, + "step": 915 + }, + { + "epoch": 0.20841865756541525, + "grad_norm": 2.3841098586953846, + "learning_rate": 1.2446603250653658e-06, + "loss": 0.1917, + "step": 916 + }, + { + "epoch": 0.208646188850967, + "grad_norm": 2.0643080194861496, + "learning_rate": 1.24464866395998e-06, + "loss": 0.1276, + "step": 917 + }, + { + "epoch": 0.20887372013651878, + "grad_norm": 1.1445975014034748, + "learning_rate": 1.2446369901901102e-06, + "loss": 0.0884, + "step": 918 + }, + { + "epoch": 0.20910125142207053, + "grad_norm": 3.359267538919808, + "learning_rate": 1.2446253037559952e-06, + "loss": 0.1214, + "step": 919 + }, + { + "epoch": 0.20932878270762229, + "grad_norm": 2.1583486474112927, + "learning_rate": 1.2446136046578739e-06, + "loss": 0.1093, + "step": 920 + }, + { + "epoch": 0.20955631399317406, + "grad_norm": 2.692763960200507, + "learning_rate": 1.2446018928959853e-06, + "loss": 0.2289, + "step": 921 + }, + { + "epoch": 0.20978384527872582, + "grad_norm": 2.356276890733175, + "learning_rate": 1.2445901684705685e-06, + "loss": 0.2222, + "step": 922 + }, + { + "epoch": 0.2100113765642776, + "grad_norm": 2.596476104334523, + "learning_rate": 1.2445784313818638e-06, + "loss": 0.1574, + "step": 923 + }, + { + "epoch": 0.21023890784982935, + "grad_norm": 2.788233818738729, + "learning_rate": 1.2445666816301102e-06, + "loss": 0.1303, + "step": 924 + }, + { + "epoch": 0.21046643913538113, + "grad_norm": 2.3013258694625245, + "learning_rate": 1.2445549192155487e-06, + "loss": 0.2232, + "step": 925 + }, + { + "epoch": 0.21069397042093288, + "grad_norm": 2.364410552617768, + "learning_rate": 1.244543144138419e-06, + "loss": 0.1967, + "step": 926 + }, + { + "epoch": 0.21092150170648463, + "grad_norm": 1.4320620142185012, + "learning_rate": 1.2445313563989624e-06, + "loss": 0.1533, + "step": 927 + }, + { + "epoch": 0.2111490329920364, + "grad_norm": 1.8979786639459473, + "learning_rate": 1.2445195559974194e-06, + "loss": 0.1494, + "step": 928 + }, + { + "epoch": 0.21137656427758816, + "grad_norm": 2.1174466003626446, + "learning_rate": 1.244507742934031e-06, + "loss": 0.1973, + "step": 929 + }, + { + "epoch": 0.21160409556313994, + "grad_norm": 2.164188059326067, + "learning_rate": 1.2444959172090393e-06, + "loss": 0.1336, + "step": 930 + }, + { + "epoch": 0.2118316268486917, + "grad_norm": 1.5503789009056947, + "learning_rate": 1.2444840788226854e-06, + "loss": 0.1948, + "step": 931 + }, + { + "epoch": 0.21205915813424345, + "grad_norm": 1.8654319466920093, + "learning_rate": 1.2444722277752114e-06, + "loss": 0.2043, + "step": 932 + }, + { + "epoch": 0.21228668941979523, + "grad_norm": 2.020474941013341, + "learning_rate": 1.2444603640668596e-06, + "loss": 0.2211, + "step": 933 + }, + { + "epoch": 0.21251422070534698, + "grad_norm": 2.0138343922511206, + "learning_rate": 1.2444484876978725e-06, + "loss": 0.1402, + "step": 934 + }, + { + "epoch": 0.21274175199089876, + "grad_norm": 1.5804379894073013, + "learning_rate": 1.2444365986684929e-06, + "loss": 0.1311, + "step": 935 + }, + { + "epoch": 0.2129692832764505, + "grad_norm": 2.2151819679335367, + "learning_rate": 1.2444246969789633e-06, + "loss": 0.0884, + "step": 936 + }, + { + "epoch": 0.21319681456200226, + "grad_norm": 2.4707341962723834, + "learning_rate": 1.2444127826295277e-06, + "loss": 0.1138, + "step": 937 + }, + { + "epoch": 0.21342434584755404, + "grad_norm": 2.142646726979162, + "learning_rate": 1.244400855620429e-06, + "loss": 0.1234, + "step": 938 + }, + { + "epoch": 0.2136518771331058, + "grad_norm": 1.3461044168942922, + "learning_rate": 1.2443889159519113e-06, + "loss": 0.0966, + "step": 939 + }, + { + "epoch": 0.21387940841865757, + "grad_norm": 2.824705608850421, + "learning_rate": 1.2443769636242185e-06, + "loss": 0.1736, + "step": 940 + }, + { + "epoch": 0.21410693970420933, + "grad_norm": 3.3926592270656526, + "learning_rate": 1.244364998637595e-06, + "loss": 0.102, + "step": 941 + }, + { + "epoch": 0.2143344709897611, + "grad_norm": 2.1478829302272278, + "learning_rate": 1.2443530209922848e-06, + "loss": 0.0958, + "step": 942 + }, + { + "epoch": 0.21456200227531286, + "grad_norm": 2.084791701381943, + "learning_rate": 1.2443410306885337e-06, + "loss": 0.128, + "step": 943 + }, + { + "epoch": 0.2147895335608646, + "grad_norm": 2.667044034523646, + "learning_rate": 1.244329027726586e-06, + "loss": 0.2088, + "step": 944 + }, + { + "epoch": 0.2150170648464164, + "grad_norm": 1.4354076627961647, + "learning_rate": 1.2443170121066872e-06, + "loss": 0.1295, + "step": 945 + }, + { + "epoch": 0.21524459613196814, + "grad_norm": 3.608014557262876, + "learning_rate": 1.2443049838290827e-06, + "loss": 0.1479, + "step": 946 + }, + { + "epoch": 0.21547212741751992, + "grad_norm": 2.4907426669888424, + "learning_rate": 1.2442929428940186e-06, + "loss": 0.2094, + "step": 947 + }, + { + "epoch": 0.21569965870307167, + "grad_norm": 1.889292577370491, + "learning_rate": 1.2442808893017414e-06, + "loss": 0.1182, + "step": 948 + }, + { + "epoch": 0.21592718998862342, + "grad_norm": 1.295703999044032, + "learning_rate": 1.2442688230524965e-06, + "loss": 0.1493, + "step": 949 + }, + { + "epoch": 0.2161547212741752, + "grad_norm": 3.010053578949512, + "learning_rate": 1.244256744146531e-06, + "loss": 0.1837, + "step": 950 + }, + { + "epoch": 0.21638225255972696, + "grad_norm": 2.2542440250817357, + "learning_rate": 1.244244652584092e-06, + "loss": 0.2011, + "step": 951 + }, + { + "epoch": 0.21660978384527874, + "grad_norm": 1.8471360091007536, + "learning_rate": 1.2442325483654263e-06, + "loss": 0.1529, + "step": 952 + }, + { + "epoch": 0.2168373151308305, + "grad_norm": 3.360264898638295, + "learning_rate": 1.2442204314907812e-06, + "loss": 0.1952, + "step": 953 + }, + { + "epoch": 0.21706484641638224, + "grad_norm": 2.2836983418694308, + "learning_rate": 1.2442083019604047e-06, + "loss": 0.2068, + "step": 954 + }, + { + "epoch": 0.21729237770193402, + "grad_norm": 2.534259478561885, + "learning_rate": 1.2441961597745447e-06, + "loss": 0.131, + "step": 955 + }, + { + "epoch": 0.21751990898748577, + "grad_norm": 2.116332324988344, + "learning_rate": 1.244184004933449e-06, + "loss": 0.1433, + "step": 956 + }, + { + "epoch": 0.21774744027303755, + "grad_norm": 1.9239447267712195, + "learning_rate": 1.2441718374373662e-06, + "loss": 0.1296, + "step": 957 + }, + { + "epoch": 0.2179749715585893, + "grad_norm": 3.11283517907892, + "learning_rate": 1.244159657286545e-06, + "loss": 0.1556, + "step": 958 + }, + { + "epoch": 0.21820250284414108, + "grad_norm": 2.1030310163998, + "learning_rate": 1.2441474644812345e-06, + "loss": 0.1398, + "step": 959 + }, + { + "epoch": 0.21843003412969283, + "grad_norm": 2.6301386027385734, + "learning_rate": 1.2441352590216836e-06, + "loss": 0.1328, + "step": 960 + }, + { + "epoch": 0.2186575654152446, + "grad_norm": 1.6843043929069075, + "learning_rate": 1.244123040908142e-06, + "loss": 0.2169, + "step": 961 + }, + { + "epoch": 0.21888509670079637, + "grad_norm": 2.021371056385805, + "learning_rate": 1.2441108101408592e-06, + "loss": 0.105, + "step": 962 + }, + { + "epoch": 0.21911262798634812, + "grad_norm": 2.932640255317413, + "learning_rate": 1.2440985667200853e-06, + "loss": 0.1186, + "step": 963 + }, + { + "epoch": 0.2193401592718999, + "grad_norm": 2.287879466073487, + "learning_rate": 1.2440863106460705e-06, + "loss": 0.1418, + "step": 964 + }, + { + "epoch": 0.21956769055745165, + "grad_norm": 2.4323172112890807, + "learning_rate": 1.2440740419190655e-06, + "loss": 0.2116, + "step": 965 + }, + { + "epoch": 0.2197952218430034, + "grad_norm": 2.906286752213052, + "learning_rate": 1.2440617605393208e-06, + "loss": 0.2029, + "step": 966 + }, + { + "epoch": 0.22002275312855518, + "grad_norm": 2.420234503572233, + "learning_rate": 1.2440494665070874e-06, + "loss": 0.2227, + "step": 967 + }, + { + "epoch": 0.22025028441410693, + "grad_norm": 2.1531642600457874, + "learning_rate": 1.2440371598226165e-06, + "loss": 0.1565, + "step": 968 + }, + { + "epoch": 0.2204778156996587, + "grad_norm": 1.7851844835265829, + "learning_rate": 1.2440248404861598e-06, + "loss": 0.1132, + "step": 969 + }, + { + "epoch": 0.22070534698521047, + "grad_norm": 2.2253443799094605, + "learning_rate": 1.2440125084979693e-06, + "loss": 0.1141, + "step": 970 + }, + { + "epoch": 0.22093287827076222, + "grad_norm": 3.491367387042196, + "learning_rate": 1.2440001638582965e-06, + "loss": 0.1678, + "step": 971 + }, + { + "epoch": 0.221160409556314, + "grad_norm": 2.6799332639547297, + "learning_rate": 1.2439878065673944e-06, + "loss": 0.1791, + "step": 972 + }, + { + "epoch": 0.22138794084186575, + "grad_norm": 0.9028117739016462, + "learning_rate": 1.2439754366255149e-06, + "loss": 0.0794, + "step": 973 + }, + { + "epoch": 0.22161547212741753, + "grad_norm": 1.6629358802939667, + "learning_rate": 1.2439630540329111e-06, + "loss": 0.1328, + "step": 974 + }, + { + "epoch": 0.22184300341296928, + "grad_norm": 2.734953415687441, + "learning_rate": 1.2439506587898358e-06, + "loss": 0.1168, + "step": 975 + }, + { + "epoch": 0.22207053469852106, + "grad_norm": 2.0986779517624745, + "learning_rate": 1.243938250896543e-06, + "loss": 0.1288, + "step": 976 + }, + { + "epoch": 0.2222980659840728, + "grad_norm": 2.4554262769941766, + "learning_rate": 1.2439258303532858e-06, + "loss": 0.1545, + "step": 977 + }, + { + "epoch": 0.22252559726962456, + "grad_norm": 1.7628888954012072, + "learning_rate": 1.243913397160318e-06, + "loss": 0.0967, + "step": 978 + }, + { + "epoch": 0.22275312855517634, + "grad_norm": 1.8371409568342896, + "learning_rate": 1.2439009513178938e-06, + "loss": 0.1184, + "step": 979 + }, + { + "epoch": 0.2229806598407281, + "grad_norm": 3.4838138279645103, + "learning_rate": 1.2438884928262678e-06, + "loss": 0.1686, + "step": 980 + }, + { + "epoch": 0.22320819112627988, + "grad_norm": 1.743212643613601, + "learning_rate": 1.2438760216856944e-06, + "loss": 0.1005, + "step": 981 + }, + { + "epoch": 0.22343572241183163, + "grad_norm": 2.2940811110233135, + "learning_rate": 1.2438635378964284e-06, + "loss": 0.1261, + "step": 982 + }, + { + "epoch": 0.22366325369738338, + "grad_norm": 3.306786589733754, + "learning_rate": 1.2438510414587251e-06, + "loss": 0.1057, + "step": 983 + }, + { + "epoch": 0.22389078498293516, + "grad_norm": 1.8312197926008273, + "learning_rate": 1.24383853237284e-06, + "loss": 0.1121, + "step": 984 + }, + { + "epoch": 0.2241183162684869, + "grad_norm": 1.375951456745173, + "learning_rate": 1.2438260106390285e-06, + "loss": 0.1137, + "step": 985 + }, + { + "epoch": 0.2243458475540387, + "grad_norm": 2.2850475547846507, + "learning_rate": 1.2438134762575467e-06, + "loss": 0.1528, + "step": 986 + }, + { + "epoch": 0.22457337883959044, + "grad_norm": 1.7811601291763544, + "learning_rate": 1.243800929228651e-06, + "loss": 0.114, + "step": 987 + }, + { + "epoch": 0.2248009101251422, + "grad_norm": 2.175503500486742, + "learning_rate": 1.2437883695525974e-06, + "loss": 0.2246, + "step": 988 + }, + { + "epoch": 0.22502844141069397, + "grad_norm": 2.5853887611675375, + "learning_rate": 1.2437757972296427e-06, + "loss": 0.2126, + "step": 989 + }, + { + "epoch": 0.22525597269624573, + "grad_norm": 2.4622729490723065, + "learning_rate": 1.2437632122600442e-06, + "loss": 0.1806, + "step": 990 + }, + { + "epoch": 0.2254835039817975, + "grad_norm": 2.2336859931017794, + "learning_rate": 1.2437506146440587e-06, + "loss": 0.1948, + "step": 991 + }, + { + "epoch": 0.22571103526734926, + "grad_norm": 2.388802906376772, + "learning_rate": 1.243738004381944e-06, + "loss": 0.1028, + "step": 992 + }, + { + "epoch": 0.225938566552901, + "grad_norm": 2.526457136508687, + "learning_rate": 1.2437253814739572e-06, + "loss": 0.1394, + "step": 993 + }, + { + "epoch": 0.2261660978384528, + "grad_norm": 2.282347439516019, + "learning_rate": 1.2437127459203572e-06, + "loss": 0.1678, + "step": 994 + }, + { + "epoch": 0.22639362912400454, + "grad_norm": 1.3050466119815518, + "learning_rate": 1.2437000977214015e-06, + "loss": 0.0753, + "step": 995 + }, + { + "epoch": 0.22662116040955632, + "grad_norm": 2.159334429482828, + "learning_rate": 1.243687436877349e-06, + "loss": 0.2767, + "step": 996 + }, + { + "epoch": 0.22684869169510807, + "grad_norm": 2.4741243617261617, + "learning_rate": 1.2436747633884583e-06, + "loss": 0.167, + "step": 997 + }, + { + "epoch": 0.22707622298065985, + "grad_norm": 2.522130011756034, + "learning_rate": 1.2436620772549885e-06, + "loss": 0.2229, + "step": 998 + }, + { + "epoch": 0.2273037542662116, + "grad_norm": 2.2654639871535873, + "learning_rate": 1.243649378477199e-06, + "loss": 0.1376, + "step": 999 + }, + { + "epoch": 0.22753128555176336, + "grad_norm": 2.737389406083516, + "learning_rate": 1.2436366670553491e-06, + "loss": 0.1672, + "step": 1000 + }, + { + "epoch": 0.22775881683731514, + "grad_norm": 2.497999857751637, + "learning_rate": 1.2436239429896988e-06, + "loss": 0.2831, + "step": 1001 + }, + { + "epoch": 0.2279863481228669, + "grad_norm": 2.3986139069373125, + "learning_rate": 1.2436112062805081e-06, + "loss": 0.1413, + "step": 1002 + }, + { + "epoch": 0.22821387940841867, + "grad_norm": 1.63194618315687, + "learning_rate": 1.2435984569280372e-06, + "loss": 0.1509, + "step": 1003 + }, + { + "epoch": 0.22844141069397042, + "grad_norm": 1.9884735218546312, + "learning_rate": 1.2435856949325467e-06, + "loss": 0.0909, + "step": 1004 + }, + { + "epoch": 0.22866894197952217, + "grad_norm": 3.7364717574130877, + "learning_rate": 1.2435729202942972e-06, + "loss": 0.1362, + "step": 1005 + }, + { + "epoch": 0.22889647326507395, + "grad_norm": 4.3498400339740595, + "learning_rate": 1.2435601330135506e-06, + "loss": 0.1364, + "step": 1006 + }, + { + "epoch": 0.2291240045506257, + "grad_norm": 1.468486521047109, + "learning_rate": 1.2435473330905674e-06, + "loss": 0.1902, + "step": 1007 + }, + { + "epoch": 0.22935153583617748, + "grad_norm": 2.602985360302298, + "learning_rate": 1.2435345205256097e-06, + "loss": 0.0947, + "step": 1008 + }, + { + "epoch": 0.22957906712172924, + "grad_norm": 2.117002790495142, + "learning_rate": 1.243521695318939e-06, + "loss": 0.1228, + "step": 1009 + }, + { + "epoch": 0.229806598407281, + "grad_norm": 2.0012843231226034, + "learning_rate": 1.2435088574708178e-06, + "loss": 0.1156, + "step": 1010 + }, + { + "epoch": 0.23003412969283277, + "grad_norm": 2.490148339748286, + "learning_rate": 1.2434960069815083e-06, + "loss": 0.164, + "step": 1011 + }, + { + "epoch": 0.23026166097838452, + "grad_norm": 2.450730689081713, + "learning_rate": 1.243483143851273e-06, + "loss": 0.138, + "step": 1012 + }, + { + "epoch": 0.2304891922639363, + "grad_norm": 2.892744061430906, + "learning_rate": 1.2434702680803751e-06, + "loss": 0.1061, + "step": 1013 + }, + { + "epoch": 0.23071672354948805, + "grad_norm": 2.790226387512928, + "learning_rate": 1.2434573796690774e-06, + "loss": 0.1957, + "step": 1014 + }, + { + "epoch": 0.23094425483503983, + "grad_norm": 2.4036726186705972, + "learning_rate": 1.2434444786176435e-06, + "loss": 0.1544, + "step": 1015 + }, + { + "epoch": 0.23117178612059158, + "grad_norm": 1.3271746602955339, + "learning_rate": 1.2434315649263372e-06, + "loss": 0.061, + "step": 1016 + }, + { + "epoch": 0.23139931740614333, + "grad_norm": 1.4063593684445947, + "learning_rate": 1.2434186385954225e-06, + "loss": 0.1068, + "step": 1017 + }, + { + "epoch": 0.23162684869169511, + "grad_norm": 2.9525793198909724, + "learning_rate": 1.243405699625163e-06, + "loss": 0.1067, + "step": 1018 + }, + { + "epoch": 0.23185437997724687, + "grad_norm": 2.7846219600282747, + "learning_rate": 1.243392748015824e-06, + "loss": 0.1435, + "step": 1019 + }, + { + "epoch": 0.23208191126279865, + "grad_norm": 1.5658061687677385, + "learning_rate": 1.2433797837676694e-06, + "loss": 0.1492, + "step": 1020 + }, + { + "epoch": 0.2323094425483504, + "grad_norm": 4.123388323133236, + "learning_rate": 1.2433668068809648e-06, + "loss": 0.1699, + "step": 1021 + }, + { + "epoch": 0.23253697383390215, + "grad_norm": 2.0976126762166403, + "learning_rate": 1.243353817355975e-06, + "loss": 0.1257, + "step": 1022 + }, + { + "epoch": 0.23276450511945393, + "grad_norm": 2.4116621601065296, + "learning_rate": 1.2433408151929655e-06, + "loss": 0.133, + "step": 1023 + }, + { + "epoch": 0.23299203640500568, + "grad_norm": 1.395623834578789, + "learning_rate": 1.2433278003922026e-06, + "loss": 0.0936, + "step": 1024 + }, + { + "epoch": 0.23321956769055746, + "grad_norm": 1.7768669244027402, + "learning_rate": 1.2433147729539514e-06, + "loss": 0.1264, + "step": 1025 + }, + { + "epoch": 0.2334470989761092, + "grad_norm": 2.489847520949891, + "learning_rate": 1.2433017328784788e-06, + "loss": 0.1714, + "step": 1026 + }, + { + "epoch": 0.23367463026166096, + "grad_norm": 1.722648702759186, + "learning_rate": 1.2432886801660513e-06, + "loss": 0.122, + "step": 1027 + }, + { + "epoch": 0.23390216154721274, + "grad_norm": 1.3061284883014919, + "learning_rate": 1.2432756148169354e-06, + "loss": 0.0726, + "step": 1028 + }, + { + "epoch": 0.2341296928327645, + "grad_norm": 2.807955909764041, + "learning_rate": 1.2432625368313983e-06, + "loss": 0.1667, + "step": 1029 + }, + { + "epoch": 0.23435722411831628, + "grad_norm": 1.9724601313774524, + "learning_rate": 1.2432494462097072e-06, + "loss": 0.1995, + "step": 1030 + }, + { + "epoch": 0.23458475540386803, + "grad_norm": 2.3943947067430895, + "learning_rate": 1.2432363429521295e-06, + "loss": 0.1625, + "step": 1031 + }, + { + "epoch": 0.2348122866894198, + "grad_norm": 1.5436408096888365, + "learning_rate": 1.2432232270589335e-06, + "loss": 0.076, + "step": 1032 + }, + { + "epoch": 0.23503981797497156, + "grad_norm": 1.1938881747627557, + "learning_rate": 1.2432100985303868e-06, + "loss": 0.1002, + "step": 1033 + }, + { + "epoch": 0.2352673492605233, + "grad_norm": 2.0446974564823304, + "learning_rate": 1.243196957366758e-06, + "loss": 0.1721, + "step": 1034 + }, + { + "epoch": 0.2354948805460751, + "grad_norm": 1.079879180238331, + "learning_rate": 1.2431838035683155e-06, + "loss": 0.1257, + "step": 1035 + }, + { + "epoch": 0.23572241183162684, + "grad_norm": 1.8378535292320874, + "learning_rate": 1.2431706371353282e-06, + "loss": 0.1821, + "step": 1036 + }, + { + "epoch": 0.23594994311717862, + "grad_norm": 1.969855842746801, + "learning_rate": 1.2431574580680653e-06, + "loss": 0.1436, + "step": 1037 + }, + { + "epoch": 0.23617747440273038, + "grad_norm": 3.058757707801488, + "learning_rate": 1.2431442663667958e-06, + "loss": 0.1605, + "step": 1038 + }, + { + "epoch": 0.23640500568828213, + "grad_norm": 1.2648716547694445, + "learning_rate": 1.2431310620317898e-06, + "loss": 0.1614, + "step": 1039 + }, + { + "epoch": 0.2366325369738339, + "grad_norm": 1.9610877034271015, + "learning_rate": 1.2431178450633168e-06, + "loss": 0.139, + "step": 1040 + }, + { + "epoch": 0.23686006825938566, + "grad_norm": 1.5919631273318544, + "learning_rate": 1.2431046154616473e-06, + "loss": 0.0888, + "step": 1041 + }, + { + "epoch": 0.23708759954493744, + "grad_norm": 1.791707313865184, + "learning_rate": 1.2430913732270512e-06, + "loss": 0.1087, + "step": 1042 + }, + { + "epoch": 0.2373151308304892, + "grad_norm": 3.1377911678690666, + "learning_rate": 1.2430781183597995e-06, + "loss": 0.1565, + "step": 1043 + }, + { + "epoch": 0.23754266211604094, + "grad_norm": 2.2837991793589607, + "learning_rate": 1.243064850860163e-06, + "loss": 0.1126, + "step": 1044 + }, + { + "epoch": 0.23777019340159272, + "grad_norm": 2.6823412767535246, + "learning_rate": 1.243051570728413e-06, + "loss": 0.2083, + "step": 1045 + }, + { + "epoch": 0.23799772468714447, + "grad_norm": 4.365244516577561, + "learning_rate": 1.2430382779648208e-06, + "loss": 0.1904, + "step": 1046 + }, + { + "epoch": 0.23822525597269625, + "grad_norm": 2.434739692035364, + "learning_rate": 1.243024972569658e-06, + "loss": 0.1347, + "step": 1047 + }, + { + "epoch": 0.238452787258248, + "grad_norm": 2.1595986496307384, + "learning_rate": 1.2430116545431966e-06, + "loss": 0.1926, + "step": 1048 + }, + { + "epoch": 0.23868031854379979, + "grad_norm": 2.2542031412662573, + "learning_rate": 1.2429983238857088e-06, + "loss": 0.1667, + "step": 1049 + }, + { + "epoch": 0.23890784982935154, + "grad_norm": 2.0405926385207787, + "learning_rate": 1.2429849805974673e-06, + "loss": 0.1872, + "step": 1050 + }, + { + "epoch": 0.2391353811149033, + "grad_norm": 2.2037085916589043, + "learning_rate": 1.2429716246787444e-06, + "loss": 0.0775, + "step": 1051 + }, + { + "epoch": 0.23936291240045507, + "grad_norm": 0.9628371959013814, + "learning_rate": 1.242958256129813e-06, + "loss": 0.1378, + "step": 1052 + }, + { + "epoch": 0.23959044368600682, + "grad_norm": 2.1187588487355424, + "learning_rate": 1.242944874950947e-06, + "loss": 0.159, + "step": 1053 + }, + { + "epoch": 0.2398179749715586, + "grad_norm": 1.9961766997876433, + "learning_rate": 1.2429314811424192e-06, + "loss": 0.1568, + "step": 1054 + }, + { + "epoch": 0.24004550625711035, + "grad_norm": 1.935471261024473, + "learning_rate": 1.242918074704504e-06, + "loss": 0.1596, + "step": 1055 + }, + { + "epoch": 0.2402730375426621, + "grad_norm": 1.4988665110908368, + "learning_rate": 1.2429046556374747e-06, + "loss": 0.0987, + "step": 1056 + }, + { + "epoch": 0.24050056882821388, + "grad_norm": 2.4283216098462015, + "learning_rate": 1.2428912239416057e-06, + "loss": 0.1127, + "step": 1057 + }, + { + "epoch": 0.24072810011376564, + "grad_norm": 2.3264824459084448, + "learning_rate": 1.242877779617172e-06, + "loss": 0.1274, + "step": 1058 + }, + { + "epoch": 0.24095563139931742, + "grad_norm": 2.159687331291489, + "learning_rate": 1.242864322664448e-06, + "loss": 0.1399, + "step": 1059 + }, + { + "epoch": 0.24118316268486917, + "grad_norm": 2.3632421336063087, + "learning_rate": 1.2428508530837088e-06, + "loss": 0.1751, + "step": 1060 + }, + { + "epoch": 0.24141069397042092, + "grad_norm": 4.564054038887482, + "learning_rate": 1.2428373708752298e-06, + "loss": 0.1623, + "step": 1061 + }, + { + "epoch": 0.2416382252559727, + "grad_norm": 2.913968751293169, + "learning_rate": 1.2428238760392862e-06, + "loss": 0.2404, + "step": 1062 + }, + { + "epoch": 0.24186575654152445, + "grad_norm": 2.375864551832549, + "learning_rate": 1.2428103685761543e-06, + "loss": 0.1551, + "step": 1063 + }, + { + "epoch": 0.24209328782707623, + "grad_norm": 2.773326434228427, + "learning_rate": 1.2427968484861097e-06, + "loss": 0.1129, + "step": 1064 + }, + { + "epoch": 0.24232081911262798, + "grad_norm": 3.440322207371564, + "learning_rate": 1.2427833157694292e-06, + "loss": 0.2312, + "step": 1065 + }, + { + "epoch": 0.24254835039817976, + "grad_norm": 2.09362609958651, + "learning_rate": 1.2427697704263892e-06, + "loss": 0.1047, + "step": 1066 + }, + { + "epoch": 0.24277588168373151, + "grad_norm": 2.0696892695320432, + "learning_rate": 1.2427562124572663e-06, + "loss": 0.1156, + "step": 1067 + }, + { + "epoch": 0.24300341296928327, + "grad_norm": 1.923568801452821, + "learning_rate": 1.2427426418623377e-06, + "loss": 0.1609, + "step": 1068 + }, + { + "epoch": 0.24323094425483505, + "grad_norm": 1.5158781630471698, + "learning_rate": 1.242729058641881e-06, + "loss": 0.094, + "step": 1069 + }, + { + "epoch": 0.2434584755403868, + "grad_norm": 2.2258107327352037, + "learning_rate": 1.2427154627961737e-06, + "loss": 0.2017, + "step": 1070 + }, + { + "epoch": 0.24368600682593858, + "grad_norm": 2.3481688305100645, + "learning_rate": 1.2427018543254935e-06, + "loss": 0.1535, + "step": 1071 + }, + { + "epoch": 0.24391353811149033, + "grad_norm": 2.148375299510445, + "learning_rate": 1.2426882332301187e-06, + "loss": 0.1812, + "step": 1072 + }, + { + "epoch": 0.24414106939704208, + "grad_norm": 1.6816805152718777, + "learning_rate": 1.2426745995103277e-06, + "loss": 0.1341, + "step": 1073 + }, + { + "epoch": 0.24436860068259386, + "grad_norm": 2.651811251817173, + "learning_rate": 1.242660953166399e-06, + "loss": 0.1318, + "step": 1074 + }, + { + "epoch": 0.2445961319681456, + "grad_norm": 2.473544844662378, + "learning_rate": 1.2426472941986117e-06, + "loss": 0.1972, + "step": 1075 + }, + { + "epoch": 0.2448236632536974, + "grad_norm": 1.3274925024741444, + "learning_rate": 1.2426336226072449e-06, + "loss": 0.1497, + "step": 1076 + }, + { + "epoch": 0.24505119453924915, + "grad_norm": 2.1014804926130277, + "learning_rate": 1.242619938392578e-06, + "loss": 0.1186, + "step": 1077 + }, + { + "epoch": 0.2452787258248009, + "grad_norm": 3.0260303106049973, + "learning_rate": 1.2426062415548907e-06, + "loss": 0.2506, + "step": 1078 + }, + { + "epoch": 0.24550625711035268, + "grad_norm": 1.2327761741993546, + "learning_rate": 1.2425925320944628e-06, + "loss": 0.117, + "step": 1079 + }, + { + "epoch": 0.24573378839590443, + "grad_norm": 3.2155457599215036, + "learning_rate": 1.2425788100115747e-06, + "loss": 0.1412, + "step": 1080 + }, + { + "epoch": 0.2459613196814562, + "grad_norm": 1.6672046307721682, + "learning_rate": 1.2425650753065065e-06, + "loss": 0.148, + "step": 1081 + }, + { + "epoch": 0.24618885096700796, + "grad_norm": 4.323033908726176, + "learning_rate": 1.2425513279795395e-06, + "loss": 0.1685, + "step": 1082 + }, + { + "epoch": 0.24641638225255974, + "grad_norm": 2.4128743686143146, + "learning_rate": 1.2425375680309543e-06, + "loss": 0.0992, + "step": 1083 + }, + { + "epoch": 0.2466439135381115, + "grad_norm": 2.0582783253443497, + "learning_rate": 1.2425237954610322e-06, + "loss": 0.1263, + "step": 1084 + }, + { + "epoch": 0.24687144482366324, + "grad_norm": 2.5810033905990637, + "learning_rate": 1.2425100102700547e-06, + "loss": 0.2102, + "step": 1085 + }, + { + "epoch": 0.24709897610921502, + "grad_norm": 2.269665820869707, + "learning_rate": 1.2424962124583033e-06, + "loss": 0.105, + "step": 1086 + }, + { + "epoch": 0.24732650739476678, + "grad_norm": 2.706182109515585, + "learning_rate": 1.2424824020260603e-06, + "loss": 0.1596, + "step": 1087 + }, + { + "epoch": 0.24755403868031856, + "grad_norm": 3.0056026517839016, + "learning_rate": 1.2424685789736077e-06, + "loss": 0.1809, + "step": 1088 + }, + { + "epoch": 0.2477815699658703, + "grad_norm": 2.2230272708907513, + "learning_rate": 1.2424547433012284e-06, + "loss": 0.1187, + "step": 1089 + }, + { + "epoch": 0.24800910125142206, + "grad_norm": 2.271631978747539, + "learning_rate": 1.2424408950092049e-06, + "loss": 0.1478, + "step": 1090 + }, + { + "epoch": 0.24823663253697384, + "grad_norm": 2.485671272218175, + "learning_rate": 1.2424270340978204e-06, + "loss": 0.1595, + "step": 1091 + }, + { + "epoch": 0.2484641638225256, + "grad_norm": 2.5242524420773087, + "learning_rate": 1.2424131605673582e-06, + "loss": 0.2519, + "step": 1092 + }, + { + "epoch": 0.24869169510807737, + "grad_norm": 2.6439941529662025, + "learning_rate": 1.2423992744181015e-06, + "loss": 0.1389, + "step": 1093 + }, + { + "epoch": 0.24891922639362912, + "grad_norm": 2.1610086973465417, + "learning_rate": 1.2423853756503343e-06, + "loss": 0.1017, + "step": 1094 + }, + { + "epoch": 0.24914675767918087, + "grad_norm": 1.8954846688503157, + "learning_rate": 1.2423714642643408e-06, + "loss": 0.2796, + "step": 1095 + }, + { + "epoch": 0.24937428896473265, + "grad_norm": 1.3124277359799683, + "learning_rate": 1.2423575402604051e-06, + "loss": 0.12, + "step": 1096 + }, + { + "epoch": 0.2496018202502844, + "grad_norm": 2.5234695537617444, + "learning_rate": 1.2423436036388122e-06, + "loss": 0.1242, + "step": 1097 + }, + { + "epoch": 0.24982935153583619, + "grad_norm": 2.044792039361886, + "learning_rate": 1.2423296543998465e-06, + "loss": 0.1743, + "step": 1098 + }, + { + "epoch": 0.25005688282138794, + "grad_norm": 3.6767614291561492, + "learning_rate": 1.2423156925437932e-06, + "loss": 0.2584, + "step": 1099 + }, + { + "epoch": 0.2502844141069397, + "grad_norm": 2.1397151355216506, + "learning_rate": 1.2423017180709376e-06, + "loss": 0.1586, + "step": 1100 + }, + { + "epoch": 0.25051194539249144, + "grad_norm": 1.670738860931536, + "learning_rate": 1.2422877309815656e-06, + "loss": 0.0821, + "step": 1101 + }, + { + "epoch": 0.25073947667804325, + "grad_norm": 2.3733300367714185, + "learning_rate": 1.242273731275963e-06, + "loss": 0.1335, + "step": 1102 + }, + { + "epoch": 0.250967007963595, + "grad_norm": 2.6954093027320534, + "learning_rate": 1.2422597189544155e-06, + "loss": 0.1244, + "step": 1103 + }, + { + "epoch": 0.25119453924914675, + "grad_norm": 2.17330712431736, + "learning_rate": 1.2422456940172101e-06, + "loss": 0.1799, + "step": 1104 + }, + { + "epoch": 0.2514220705346985, + "grad_norm": 2.4883101223722397, + "learning_rate": 1.2422316564646331e-06, + "loss": 0.0881, + "step": 1105 + }, + { + "epoch": 0.25164960182025026, + "grad_norm": 2.4975644528149528, + "learning_rate": 1.2422176062969713e-06, + "loss": 0.2376, + "step": 1106 + }, + { + "epoch": 0.25187713310580206, + "grad_norm": 2.242874102497345, + "learning_rate": 1.2422035435145121e-06, + "loss": 0.1117, + "step": 1107 + }, + { + "epoch": 0.2521046643913538, + "grad_norm": 2.1430334401000994, + "learning_rate": 1.2421894681175428e-06, + "loss": 0.1937, + "step": 1108 + }, + { + "epoch": 0.25233219567690557, + "grad_norm": 2.8329522904929796, + "learning_rate": 1.2421753801063511e-06, + "loss": 0.2192, + "step": 1109 + }, + { + "epoch": 0.2525597269624573, + "grad_norm": 2.7185072984242016, + "learning_rate": 1.2421612794812248e-06, + "loss": 0.1612, + "step": 1110 + }, + { + "epoch": 0.25278725824800913, + "grad_norm": 1.3607580813979583, + "learning_rate": 1.2421471662424525e-06, + "loss": 0.0967, + "step": 1111 + }, + { + "epoch": 0.2530147895335609, + "grad_norm": 5.202570048846043, + "learning_rate": 1.2421330403903222e-06, + "loss": 0.1696, + "step": 1112 + }, + { + "epoch": 0.25324232081911263, + "grad_norm": 5.6196714700914585, + "learning_rate": 1.2421189019251228e-06, + "loss": 0.1241, + "step": 1113 + }, + { + "epoch": 0.2534698521046644, + "grad_norm": 5.713500840284014, + "learning_rate": 1.2421047508471433e-06, + "loss": 0.1904, + "step": 1114 + }, + { + "epoch": 0.25369738339021614, + "grad_norm": 3.86015124804852, + "learning_rate": 1.242090587156673e-06, + "loss": 0.1078, + "step": 1115 + }, + { + "epoch": 0.25392491467576794, + "grad_norm": 5.247729743705811, + "learning_rate": 1.242076410854001e-06, + "loss": 0.209, + "step": 1116 + }, + { + "epoch": 0.2541524459613197, + "grad_norm": 2.568207486664039, + "learning_rate": 1.2420622219394174e-06, + "loss": 0.2007, + "step": 1117 + }, + { + "epoch": 0.25437997724687145, + "grad_norm": 2.258373006677131, + "learning_rate": 1.2420480204132117e-06, + "loss": 0.1081, + "step": 1118 + }, + { + "epoch": 0.2546075085324232, + "grad_norm": 2.124182108668539, + "learning_rate": 1.242033806275675e-06, + "loss": 0.1156, + "step": 1119 + }, + { + "epoch": 0.25483503981797495, + "grad_norm": 1.909249686410152, + "learning_rate": 1.2420195795270973e-06, + "loss": 0.1354, + "step": 1120 + }, + { + "epoch": 0.25506257110352676, + "grad_norm": 1.530124680560141, + "learning_rate": 1.2420053401677693e-06, + "loss": 0.1683, + "step": 1121 + }, + { + "epoch": 0.2552901023890785, + "grad_norm": 2.2436097952937595, + "learning_rate": 1.241991088197982e-06, + "loss": 0.1121, + "step": 1122 + }, + { + "epoch": 0.25551763367463026, + "grad_norm": 1.7733643321322672, + "learning_rate": 1.241976823618027e-06, + "loss": 0.2215, + "step": 1123 + }, + { + "epoch": 0.255745164960182, + "grad_norm": 1.9028486034678005, + "learning_rate": 1.241962546428196e-06, + "loss": 0.1741, + "step": 1124 + }, + { + "epoch": 0.25597269624573377, + "grad_norm": 2.225065438584496, + "learning_rate": 1.24194825662878e-06, + "loss": 0.1927, + "step": 1125 + }, + { + "epoch": 0.2562002275312856, + "grad_norm": 2.5921888880190096, + "learning_rate": 1.2419339542200715e-06, + "loss": 0.1912, + "step": 1126 + }, + { + "epoch": 0.2564277588168373, + "grad_norm": 1.450566741299779, + "learning_rate": 1.241919639202363e-06, + "loss": 0.1243, + "step": 1127 + }, + { + "epoch": 0.2566552901023891, + "grad_norm": 2.0976571865572895, + "learning_rate": 1.2419053115759468e-06, + "loss": 0.1071, + "step": 1128 + }, + { + "epoch": 0.25688282138794083, + "grad_norm": 2.7303120294152476, + "learning_rate": 1.2418909713411161e-06, + "loss": 0.1601, + "step": 1129 + }, + { + "epoch": 0.2571103526734926, + "grad_norm": 2.685114498491141, + "learning_rate": 1.2418766184981634e-06, + "loss": 0.1577, + "step": 1130 + }, + { + "epoch": 0.2573378839590444, + "grad_norm": 2.169271608320249, + "learning_rate": 1.2418622530473825e-06, + "loss": 0.1876, + "step": 1131 + }, + { + "epoch": 0.25756541524459614, + "grad_norm": 1.4396485167707906, + "learning_rate": 1.2418478749890672e-06, + "loss": 0.1051, + "step": 1132 + }, + { + "epoch": 0.2577929465301479, + "grad_norm": 1.3251283927245865, + "learning_rate": 1.2418334843235105e-06, + "loss": 0.127, + "step": 1133 + }, + { + "epoch": 0.25802047781569964, + "grad_norm": 1.7722655677157828, + "learning_rate": 1.2418190810510075e-06, + "loss": 0.0952, + "step": 1134 + }, + { + "epoch": 0.2582480091012514, + "grad_norm": 3.5363464228733976, + "learning_rate": 1.2418046651718518e-06, + "loss": 0.1442, + "step": 1135 + }, + { + "epoch": 0.2584755403868032, + "grad_norm": 1.4208162726414038, + "learning_rate": 1.2417902366863386e-06, + "loss": 0.1365, + "step": 1136 + }, + { + "epoch": 0.25870307167235496, + "grad_norm": 2.6384013164080367, + "learning_rate": 1.2417757955947623e-06, + "loss": 0.1884, + "step": 1137 + }, + { + "epoch": 0.2589306029579067, + "grad_norm": 2.0462327930247555, + "learning_rate": 1.2417613418974187e-06, + "loss": 0.1815, + "step": 1138 + }, + { + "epoch": 0.25915813424345846, + "grad_norm": 1.8956914891680519, + "learning_rate": 1.2417468755946025e-06, + "loss": 0.107, + "step": 1139 + }, + { + "epoch": 0.2593856655290102, + "grad_norm": 1.6865771627256954, + "learning_rate": 1.2417323966866097e-06, + "loss": 0.1254, + "step": 1140 + }, + { + "epoch": 0.259613196814562, + "grad_norm": 3.648733556829662, + "learning_rate": 1.2417179051737364e-06, + "loss": 0.1763, + "step": 1141 + }, + { + "epoch": 0.25984072810011377, + "grad_norm": 3.6358344752950518, + "learning_rate": 1.2417034010562784e-06, + "loss": 0.2671, + "step": 1142 + }, + { + "epoch": 0.2600682593856655, + "grad_norm": 1.8508374458484436, + "learning_rate": 1.2416888843345323e-06, + "loss": 0.1058, + "step": 1143 + }, + { + "epoch": 0.2602957906712173, + "grad_norm": 1.6379231265979228, + "learning_rate": 1.2416743550087951e-06, + "loss": 0.1959, + "step": 1144 + }, + { + "epoch": 0.2605233219567691, + "grad_norm": 2.7329577702359127, + "learning_rate": 1.241659813079363e-06, + "loss": 0.2185, + "step": 1145 + }, + { + "epoch": 0.26075085324232083, + "grad_norm": 2.1507497458601206, + "learning_rate": 1.2416452585465342e-06, + "loss": 0.1289, + "step": 1146 + }, + { + "epoch": 0.2609783845278726, + "grad_norm": 1.5858683873624775, + "learning_rate": 1.2416306914106053e-06, + "loss": 0.0789, + "step": 1147 + }, + { + "epoch": 0.26120591581342434, + "grad_norm": 0.9635476566506428, + "learning_rate": 1.2416161116718744e-06, + "loss": 0.0576, + "step": 1148 + }, + { + "epoch": 0.2614334470989761, + "grad_norm": 3.1902671512005747, + "learning_rate": 1.2416015193306397e-06, + "loss": 0.1745, + "step": 1149 + }, + { + "epoch": 0.2616609783845279, + "grad_norm": 2.4274328198153383, + "learning_rate": 1.241586914387199e-06, + "loss": 0.1837, + "step": 1150 + }, + { + "epoch": 0.26188850967007965, + "grad_norm": 2.1338082634804647, + "learning_rate": 1.2415722968418508e-06, + "loss": 0.1481, + "step": 1151 + }, + { + "epoch": 0.2621160409556314, + "grad_norm": 2.9280052924977973, + "learning_rate": 1.2415576666948945e-06, + "loss": 0.1869, + "step": 1152 + }, + { + "epoch": 0.26234357224118315, + "grad_norm": 2.4870407515539315, + "learning_rate": 1.2415430239466283e-06, + "loss": 0.1309, + "step": 1153 + }, + { + "epoch": 0.2625711035267349, + "grad_norm": 2.393155333098502, + "learning_rate": 1.241528368597352e-06, + "loss": 0.1292, + "step": 1154 + }, + { + "epoch": 0.2627986348122867, + "grad_norm": 2.327013036086297, + "learning_rate": 1.2415137006473649e-06, + "loss": 0.1714, + "step": 1155 + }, + { + "epoch": 0.26302616609783847, + "grad_norm": 1.3188871982753934, + "learning_rate": 1.241499020096967e-06, + "loss": 0.1637, + "step": 1156 + }, + { + "epoch": 0.2632536973833902, + "grad_norm": 2.588916025472251, + "learning_rate": 1.2414843269464579e-06, + "loss": 0.1409, + "step": 1157 + }, + { + "epoch": 0.26348122866894197, + "grad_norm": 3.771153920400626, + "learning_rate": 1.2414696211961386e-06, + "loss": 0.1883, + "step": 1158 + }, + { + "epoch": 0.2637087599544937, + "grad_norm": 4.499595516640575, + "learning_rate": 1.2414549028463087e-06, + "loss": 0.2953, + "step": 1159 + }, + { + "epoch": 0.26393629124004553, + "grad_norm": 2.2000868495180517, + "learning_rate": 1.2414401718972703e-06, + "loss": 0.11, + "step": 1160 + }, + { + "epoch": 0.2641638225255973, + "grad_norm": 1.9915564020272047, + "learning_rate": 1.2414254283493232e-06, + "loss": 0.15, + "step": 1161 + }, + { + "epoch": 0.26439135381114903, + "grad_norm": 1.7935865348857665, + "learning_rate": 1.2414106722027694e-06, + "loss": 0.1163, + "step": 1162 + }, + { + "epoch": 0.2646188850967008, + "grad_norm": 1.9493113396178239, + "learning_rate": 1.2413959034579104e-06, + "loss": 0.0895, + "step": 1163 + }, + { + "epoch": 0.26484641638225254, + "grad_norm": 2.164760508646002, + "learning_rate": 1.2413811221150478e-06, + "loss": 0.1221, + "step": 1164 + }, + { + "epoch": 0.26507394766780434, + "grad_norm": 2.7137559336575787, + "learning_rate": 1.2413663281744843e-06, + "loss": 0.1702, + "step": 1165 + }, + { + "epoch": 0.2653014789533561, + "grad_norm": 2.05578991760022, + "learning_rate": 1.2413515216365216e-06, + "loss": 0.1885, + "step": 1166 + }, + { + "epoch": 0.26552901023890785, + "grad_norm": 2.347644214662539, + "learning_rate": 1.2413367025014628e-06, + "loss": 0.1502, + "step": 1167 + }, + { + "epoch": 0.2657565415244596, + "grad_norm": 1.4755705417169032, + "learning_rate": 1.2413218707696103e-06, + "loss": 0.1539, + "step": 1168 + }, + { + "epoch": 0.26598407281001135, + "grad_norm": 3.019338825447027, + "learning_rate": 1.2413070264412677e-06, + "loss": 0.19, + "step": 1169 + }, + { + "epoch": 0.26621160409556316, + "grad_norm": 1.6254001938773606, + "learning_rate": 1.2412921695167381e-06, + "loss": 0.1523, + "step": 1170 + }, + { + "epoch": 0.2664391353811149, + "grad_norm": 1.440204109732841, + "learning_rate": 1.2412772999963253e-06, + "loss": 0.0929, + "step": 1171 + }, + { + "epoch": 0.26666666666666666, + "grad_norm": 1.8290728472456999, + "learning_rate": 1.2412624178803332e-06, + "loss": 0.1134, + "step": 1172 + }, + { + "epoch": 0.2668941979522184, + "grad_norm": 3.0096224925263084, + "learning_rate": 1.2412475231690656e-06, + "loss": 0.1381, + "step": 1173 + }, + { + "epoch": 0.26712172923777017, + "grad_norm": 2.199128798503439, + "learning_rate": 1.2412326158628275e-06, + "loss": 0.125, + "step": 1174 + }, + { + "epoch": 0.267349260523322, + "grad_norm": 2.319439781311619, + "learning_rate": 1.2412176959619232e-06, + "loss": 0.1491, + "step": 1175 + }, + { + "epoch": 0.2675767918088737, + "grad_norm": 2.941596562466676, + "learning_rate": 1.2412027634666578e-06, + "loss": 0.2338, + "step": 1176 + }, + { + "epoch": 0.2678043230944255, + "grad_norm": 2.7311199238732913, + "learning_rate": 1.2411878183773366e-06, + "loss": 0.2515, + "step": 1177 + }, + { + "epoch": 0.26803185437997723, + "grad_norm": 1.687064956485482, + "learning_rate": 1.2411728606942647e-06, + "loss": 0.0986, + "step": 1178 + }, + { + "epoch": 0.26825938566552904, + "grad_norm": 1.7930618105277882, + "learning_rate": 1.241157890417748e-06, + "loss": 0.1244, + "step": 1179 + }, + { + "epoch": 0.2684869169510808, + "grad_norm": 2.3927463592130698, + "learning_rate": 1.2411429075480923e-06, + "loss": 0.1089, + "step": 1180 + }, + { + "epoch": 0.26871444823663254, + "grad_norm": 1.5259649282053471, + "learning_rate": 1.2411279120856042e-06, + "loss": 0.0967, + "step": 1181 + }, + { + "epoch": 0.2689419795221843, + "grad_norm": 2.2250810785927535, + "learning_rate": 1.24111290403059e-06, + "loss": 0.1186, + "step": 1182 + }, + { + "epoch": 0.26916951080773605, + "grad_norm": 3.3967860062269963, + "learning_rate": 1.2410978833833564e-06, + "loss": 0.1935, + "step": 1183 + }, + { + "epoch": 0.26939704209328785, + "grad_norm": 2.1923686844506483, + "learning_rate": 1.2410828501442104e-06, + "loss": 0.1329, + "step": 1184 + }, + { + "epoch": 0.2696245733788396, + "grad_norm": 2.149373763512951, + "learning_rate": 1.2410678043134591e-06, + "loss": 0.1829, + "step": 1185 + }, + { + "epoch": 0.26985210466439136, + "grad_norm": 2.6757944701183, + "learning_rate": 1.2410527458914103e-06, + "loss": 0.2143, + "step": 1186 + }, + { + "epoch": 0.2700796359499431, + "grad_norm": 1.8260214393902483, + "learning_rate": 1.2410376748783714e-06, + "loss": 0.165, + "step": 1187 + }, + { + "epoch": 0.27030716723549486, + "grad_norm": 1.0916691334467872, + "learning_rate": 1.241022591274651e-06, + "loss": 0.1854, + "step": 1188 + }, + { + "epoch": 0.27053469852104667, + "grad_norm": 2.642360870928487, + "learning_rate": 1.2410074950805567e-06, + "loss": 0.174, + "step": 1189 + }, + { + "epoch": 0.2707622298065984, + "grad_norm": 2.1241327325526576, + "learning_rate": 1.2409923862963973e-06, + "loss": 0.1936, + "step": 1190 + }, + { + "epoch": 0.27098976109215017, + "grad_norm": 1.8969035428697545, + "learning_rate": 1.240977264922482e-06, + "loss": 0.1241, + "step": 1191 + }, + { + "epoch": 0.2712172923777019, + "grad_norm": 2.4081060354042743, + "learning_rate": 1.2409621309591195e-06, + "loss": 0.0908, + "step": 1192 + }, + { + "epoch": 0.2714448236632537, + "grad_norm": 3.1731526033525763, + "learning_rate": 1.2409469844066188e-06, + "loss": 0.1297, + "step": 1193 + }, + { + "epoch": 0.2716723549488055, + "grad_norm": 1.415839584125243, + "learning_rate": 1.2409318252652899e-06, + "loss": 0.1728, + "step": 1194 + }, + { + "epoch": 0.27189988623435724, + "grad_norm": 2.5230803349903366, + "learning_rate": 1.2409166535354428e-06, + "loss": 0.2229, + "step": 1195 + }, + { + "epoch": 0.272127417519909, + "grad_norm": 1.8982743663775448, + "learning_rate": 1.2409014692173872e-06, + "loss": 0.08, + "step": 1196 + }, + { + "epoch": 0.27235494880546074, + "grad_norm": 2.3845532894091943, + "learning_rate": 1.240886272311433e-06, + "loss": 0.1233, + "step": 1197 + }, + { + "epoch": 0.2725824800910125, + "grad_norm": 1.6059907271573468, + "learning_rate": 1.240871062817892e-06, + "loss": 0.141, + "step": 1198 + }, + { + "epoch": 0.2728100113765643, + "grad_norm": 2.2253019852544647, + "learning_rate": 1.240855840737074e-06, + "loss": 0.1019, + "step": 1199 + }, + { + "epoch": 0.27303754266211605, + "grad_norm": 1.8641328097177905, + "learning_rate": 1.2408406060692909e-06, + "loss": 0.1489, + "step": 1200 + }, + { + "epoch": 0.2732650739476678, + "grad_norm": 2.451518484264104, + "learning_rate": 1.2408253588148532e-06, + "loss": 0.1228, + "step": 1201 + }, + { + "epoch": 0.27349260523321955, + "grad_norm": 6.967560998726808, + "learning_rate": 1.2408100989740735e-06, + "loss": 0.1814, + "step": 1202 + }, + { + "epoch": 0.2737201365187713, + "grad_norm": 3.37131502042136, + "learning_rate": 1.2407948265472628e-06, + "loss": 0.1422, + "step": 1203 + }, + { + "epoch": 0.2739476678043231, + "grad_norm": 2.5918040941488, + "learning_rate": 1.2407795415347336e-06, + "loss": 0.1913, + "step": 1204 + }, + { + "epoch": 0.27417519908987487, + "grad_norm": 2.0851599516702106, + "learning_rate": 1.2407642439367986e-06, + "loss": 0.1295, + "step": 1205 + }, + { + "epoch": 0.2744027303754266, + "grad_norm": 2.384896134065583, + "learning_rate": 1.24074893375377e-06, + "loss": 0.1765, + "step": 1206 + }, + { + "epoch": 0.27463026166097837, + "grad_norm": 2.2263253984010087, + "learning_rate": 1.2407336109859607e-06, + "loss": 0.1677, + "step": 1207 + }, + { + "epoch": 0.2748577929465301, + "grad_norm": 4.0136041805844425, + "learning_rate": 1.2407182756336844e-06, + "loss": 0.2066, + "step": 1208 + }, + { + "epoch": 0.27508532423208193, + "grad_norm": 2.4218266443283794, + "learning_rate": 1.240702927697254e-06, + "loss": 0.1029, + "step": 1209 + }, + { + "epoch": 0.2753128555176337, + "grad_norm": 2.975498605777637, + "learning_rate": 1.2406875671769837e-06, + "loss": 0.2205, + "step": 1210 + }, + { + "epoch": 0.27554038680318543, + "grad_norm": 2.6168945120696634, + "learning_rate": 1.2406721940731866e-06, + "loss": 0.1641, + "step": 1211 + }, + { + "epoch": 0.2757679180887372, + "grad_norm": 2.189110929582555, + "learning_rate": 1.2406568083861776e-06, + "loss": 0.1369, + "step": 1212 + }, + { + "epoch": 0.27599544937428894, + "grad_norm": 3.1088766651076574, + "learning_rate": 1.2406414101162708e-06, + "loss": 0.2137, + "step": 1213 + }, + { + "epoch": 0.27622298065984074, + "grad_norm": 1.477236721867685, + "learning_rate": 1.2406259992637815e-06, + "loss": 0.108, + "step": 1214 + }, + { + "epoch": 0.2764505119453925, + "grad_norm": 3.1813357275785275, + "learning_rate": 1.240610575829024e-06, + "loss": 0.1376, + "step": 1215 + }, + { + "epoch": 0.27667804323094425, + "grad_norm": 2.636131080247815, + "learning_rate": 1.2405951398123136e-06, + "loss": 0.1964, + "step": 1216 + }, + { + "epoch": 0.276905574516496, + "grad_norm": 2.580878369476001, + "learning_rate": 1.2405796912139662e-06, + "loss": 0.1478, + "step": 1217 + }, + { + "epoch": 0.2771331058020478, + "grad_norm": 3.2969282409336564, + "learning_rate": 1.240564230034297e-06, + "loss": 0.1612, + "step": 1218 + }, + { + "epoch": 0.27736063708759956, + "grad_norm": 1.7096976219125466, + "learning_rate": 1.2405487562736226e-06, + "loss": 0.0684, + "step": 1219 + }, + { + "epoch": 0.2775881683731513, + "grad_norm": 1.5798461976853169, + "learning_rate": 1.240533269932259e-06, + "loss": 0.2019, + "step": 1220 + }, + { + "epoch": 0.27781569965870306, + "grad_norm": 1.7709663885713447, + "learning_rate": 1.2405177710105223e-06, + "loss": 0.0644, + "step": 1221 + }, + { + "epoch": 0.2780432309442548, + "grad_norm": 2.7039398210352172, + "learning_rate": 1.24050225950873e-06, + "loss": 0.1279, + "step": 1222 + }, + { + "epoch": 0.2782707622298066, + "grad_norm": 2.064393149635238, + "learning_rate": 1.2404867354271984e-06, + "loss": 0.2025, + "step": 1223 + }, + { + "epoch": 0.2784982935153584, + "grad_norm": 2.1454678847811506, + "learning_rate": 1.2404711987662452e-06, + "loss": 0.22, + "step": 1224 + }, + { + "epoch": 0.2787258248009101, + "grad_norm": 3.3818448317517706, + "learning_rate": 1.240455649526188e-06, + "loss": 0.1667, + "step": 1225 + }, + { + "epoch": 0.2789533560864619, + "grad_norm": 1.5959505587151988, + "learning_rate": 1.2404400877073446e-06, + "loss": 0.0904, + "step": 1226 + }, + { + "epoch": 0.27918088737201363, + "grad_norm": 3.2220961668197337, + "learning_rate": 1.2404245133100328e-06, + "loss": 0.1864, + "step": 1227 + }, + { + "epoch": 0.27940841865756544, + "grad_norm": 2.1405279662692744, + "learning_rate": 1.240408926334571e-06, + "loss": 0.1021, + "step": 1228 + }, + { + "epoch": 0.2796359499431172, + "grad_norm": 3.3416472980915315, + "learning_rate": 1.240393326781278e-06, + "loss": 0.2055, + "step": 1229 + }, + { + "epoch": 0.27986348122866894, + "grad_norm": 2.091589658288764, + "learning_rate": 1.2403777146504722e-06, + "loss": 0.126, + "step": 1230 + }, + { + "epoch": 0.2800910125142207, + "grad_norm": 1.416466049772803, + "learning_rate": 1.240362089942473e-06, + "loss": 0.1542, + "step": 1231 + }, + { + "epoch": 0.28031854379977245, + "grad_norm": 1.3058878302334813, + "learning_rate": 1.2403464526575997e-06, + "loss": 0.0475, + "step": 1232 + }, + { + "epoch": 0.28054607508532425, + "grad_norm": 4.125917323296369, + "learning_rate": 1.240330802796172e-06, + "loss": 0.181, + "step": 1233 + }, + { + "epoch": 0.280773606370876, + "grad_norm": 2.716562306766488, + "learning_rate": 1.2403151403585093e-06, + "loss": 0.202, + "step": 1234 + }, + { + "epoch": 0.28100113765642776, + "grad_norm": 3.3376947997511843, + "learning_rate": 1.240299465344932e-06, + "loss": 0.1384, + "step": 1235 + }, + { + "epoch": 0.2812286689419795, + "grad_norm": 1.2166897547300315, + "learning_rate": 1.2402837777557608e-06, + "loss": 0.143, + "step": 1236 + }, + { + "epoch": 0.28145620022753126, + "grad_norm": 2.6970364920835115, + "learning_rate": 1.240268077591316e-06, + "loss": 0.1991, + "step": 1237 + }, + { + "epoch": 0.28168373151308307, + "grad_norm": 1.2700401000964803, + "learning_rate": 1.2402523648519184e-06, + "loss": 0.0921, + "step": 1238 + }, + { + "epoch": 0.2819112627986348, + "grad_norm": 3.574248956011507, + "learning_rate": 1.2402366395378892e-06, + "loss": 0.1592, + "step": 1239 + }, + { + "epoch": 0.2821387940841866, + "grad_norm": 1.6046415887628118, + "learning_rate": 1.24022090164955e-06, + "loss": 0.1368, + "step": 1240 + }, + { + "epoch": 0.2823663253697383, + "grad_norm": 1.853669863286635, + "learning_rate": 1.240205151187222e-06, + "loss": 0.1363, + "step": 1241 + }, + { + "epoch": 0.2825938566552901, + "grad_norm": 2.166955108238533, + "learning_rate": 1.2401893881512278e-06, + "loss": 0.1211, + "step": 1242 + }, + { + "epoch": 0.2828213879408419, + "grad_norm": 2.523899316800894, + "learning_rate": 1.240173612541889e-06, + "loss": 0.1735, + "step": 1243 + }, + { + "epoch": 0.28304891922639364, + "grad_norm": 2.5063025861997046, + "learning_rate": 1.2401578243595281e-06, + "loss": 0.1227, + "step": 1244 + }, + { + "epoch": 0.2832764505119454, + "grad_norm": 1.8829939441618075, + "learning_rate": 1.2401420236044678e-06, + "loss": 0.0875, + "step": 1245 + }, + { + "epoch": 0.28350398179749714, + "grad_norm": 2.927257999123812, + "learning_rate": 1.2401262102770314e-06, + "loss": 0.2237, + "step": 1246 + }, + { + "epoch": 0.2837315130830489, + "grad_norm": 2.701605298598302, + "learning_rate": 1.2401103843775416e-06, + "loss": 0.2524, + "step": 1247 + }, + { + "epoch": 0.2839590443686007, + "grad_norm": 1.8917697620531524, + "learning_rate": 1.240094545906322e-06, + "loss": 0.1681, + "step": 1248 + }, + { + "epoch": 0.28418657565415245, + "grad_norm": 1.790645149029892, + "learning_rate": 1.2400786948636966e-06, + "loss": 0.1141, + "step": 1249 + }, + { + "epoch": 0.2844141069397042, + "grad_norm": 2.280105910927623, + "learning_rate": 1.2400628312499892e-06, + "loss": 0.1467, + "step": 1250 + }, + { + "epoch": 0.28464163822525596, + "grad_norm": 1.9642089313734676, + "learning_rate": 1.2400469550655239e-06, + "loss": 0.1475, + "step": 1251 + }, + { + "epoch": 0.28486916951080776, + "grad_norm": 2.788765935312485, + "learning_rate": 1.240031066310625e-06, + "loss": 0.1443, + "step": 1252 + }, + { + "epoch": 0.2850967007963595, + "grad_norm": 2.7906903727756007, + "learning_rate": 1.2400151649856176e-06, + "loss": 0.2541, + "step": 1253 + }, + { + "epoch": 0.28532423208191127, + "grad_norm": 1.7188739232283574, + "learning_rate": 1.2399992510908266e-06, + "loss": 0.1287, + "step": 1254 + }, + { + "epoch": 0.285551763367463, + "grad_norm": 2.448125831923515, + "learning_rate": 1.2399833246265772e-06, + "loss": 0.1727, + "step": 1255 + }, + { + "epoch": 0.28577929465301477, + "grad_norm": 1.670687696490714, + "learning_rate": 1.2399673855931951e-06, + "loss": 0.1783, + "step": 1256 + }, + { + "epoch": 0.2860068259385666, + "grad_norm": 1.680153916469469, + "learning_rate": 1.2399514339910058e-06, + "loss": 0.072, + "step": 1257 + }, + { + "epoch": 0.28623435722411833, + "grad_norm": 1.868751293838818, + "learning_rate": 1.2399354698203353e-06, + "loss": 0.2307, + "step": 1258 + }, + { + "epoch": 0.2864618885096701, + "grad_norm": 1.7025005817315042, + "learning_rate": 1.2399194930815103e-06, + "loss": 0.1668, + "step": 1259 + }, + { + "epoch": 0.28668941979522183, + "grad_norm": 2.300621243889103, + "learning_rate": 1.2399035037748567e-06, + "loss": 0.1402, + "step": 1260 + }, + { + "epoch": 0.2869169510807736, + "grad_norm": 1.9848277562608234, + "learning_rate": 1.2398875019007017e-06, + "loss": 0.0945, + "step": 1261 + }, + { + "epoch": 0.2871444823663254, + "grad_norm": 2.389217778082788, + "learning_rate": 1.2398714874593722e-06, + "loss": 0.1205, + "step": 1262 + }, + { + "epoch": 0.28737201365187715, + "grad_norm": 2.3311431375314733, + "learning_rate": 1.2398554604511958e-06, + "loss": 0.1375, + "step": 1263 + }, + { + "epoch": 0.2875995449374289, + "grad_norm": 1.6072927982231655, + "learning_rate": 1.2398394208764998e-06, + "loss": 0.1011, + "step": 1264 + }, + { + "epoch": 0.28782707622298065, + "grad_norm": 3.4081467363917772, + "learning_rate": 1.239823368735612e-06, + "loss": 0.2417, + "step": 1265 + }, + { + "epoch": 0.2880546075085324, + "grad_norm": 1.534207652635837, + "learning_rate": 1.2398073040288605e-06, + "loss": 0.102, + "step": 1266 + }, + { + "epoch": 0.2882821387940842, + "grad_norm": 1.5768713894474875, + "learning_rate": 1.2397912267565738e-06, + "loss": 0.1132, + "step": 1267 + }, + { + "epoch": 0.28850967007963596, + "grad_norm": 1.579072277023231, + "learning_rate": 1.2397751369190803e-06, + "loss": 0.0898, + "step": 1268 + }, + { + "epoch": 0.2887372013651877, + "grad_norm": 1.3149754797376219, + "learning_rate": 1.2397590345167088e-06, + "loss": 0.2011, + "step": 1269 + }, + { + "epoch": 0.28896473265073946, + "grad_norm": 1.6939565519433766, + "learning_rate": 1.2397429195497887e-06, + "loss": 0.1421, + "step": 1270 + }, + { + "epoch": 0.2891922639362912, + "grad_norm": 1.356052034301606, + "learning_rate": 1.239726792018649e-06, + "loss": 0.0458, + "step": 1271 + }, + { + "epoch": 0.289419795221843, + "grad_norm": 2.355272771080969, + "learning_rate": 1.2397106519236197e-06, + "loss": 0.1237, + "step": 1272 + }, + { + "epoch": 0.2896473265073948, + "grad_norm": 1.8270125880467827, + "learning_rate": 1.2396944992650303e-06, + "loss": 0.0967, + "step": 1273 + }, + { + "epoch": 0.2898748577929465, + "grad_norm": 1.6010935468127871, + "learning_rate": 1.2396783340432111e-06, + "loss": 0.1298, + "step": 1274 + }, + { + "epoch": 0.2901023890784983, + "grad_norm": 2.7512982520907685, + "learning_rate": 1.2396621562584925e-06, + "loss": 0.1489, + "step": 1275 + }, + { + "epoch": 0.29032992036405003, + "grad_norm": 2.7012778911705153, + "learning_rate": 1.2396459659112052e-06, + "loss": 0.1077, + "step": 1276 + }, + { + "epoch": 0.29055745164960184, + "grad_norm": 1.942476301895168, + "learning_rate": 1.23962976300168e-06, + "loss": 0.1705, + "step": 1277 + }, + { + "epoch": 0.2907849829351536, + "grad_norm": 1.9559220493458693, + "learning_rate": 1.2396135475302478e-06, + "loss": 0.194, + "step": 1278 + }, + { + "epoch": 0.29101251422070534, + "grad_norm": 3.484604084621699, + "learning_rate": 1.2395973194972406e-06, + "loss": 0.1654, + "step": 1279 + }, + { + "epoch": 0.2912400455062571, + "grad_norm": 3.078532329336957, + "learning_rate": 1.2395810789029898e-06, + "loss": 0.1096, + "step": 1280 + }, + { + "epoch": 0.29146757679180885, + "grad_norm": 2.1019316083307005, + "learning_rate": 1.2395648257478271e-06, + "loss": 0.1773, + "step": 1281 + }, + { + "epoch": 0.29169510807736065, + "grad_norm": 2.6622086192257646, + "learning_rate": 1.239548560032085e-06, + "loss": 0.1807, + "step": 1282 + }, + { + "epoch": 0.2919226393629124, + "grad_norm": 2.724085742068777, + "learning_rate": 1.239532281756096e-06, + "loss": 0.2633, + "step": 1283 + }, + { + "epoch": 0.29215017064846416, + "grad_norm": 3.476983854537522, + "learning_rate": 1.2395159909201924e-06, + "loss": 0.2206, + "step": 1284 + }, + { + "epoch": 0.2923777019340159, + "grad_norm": 2.4597634305484304, + "learning_rate": 1.2394996875247075e-06, + "loss": 0.173, + "step": 1285 + }, + { + "epoch": 0.2926052332195677, + "grad_norm": 2.382644730814897, + "learning_rate": 1.2394833715699743e-06, + "loss": 0.171, + "step": 1286 + }, + { + "epoch": 0.29283276450511947, + "grad_norm": 1.0179882332573014, + "learning_rate": 1.2394670430563264e-06, + "loss": 0.1506, + "step": 1287 + }, + { + "epoch": 0.2930602957906712, + "grad_norm": 2.0313503091521845, + "learning_rate": 1.2394507019840974e-06, + "loss": 0.1157, + "step": 1288 + }, + { + "epoch": 0.293287827076223, + "grad_norm": 2.1563562932885394, + "learning_rate": 1.2394343483536215e-06, + "loss": 0.1455, + "step": 1289 + }, + { + "epoch": 0.2935153583617747, + "grad_norm": 1.590464214144866, + "learning_rate": 1.2394179821652326e-06, + "loss": 0.1346, + "step": 1290 + }, + { + "epoch": 0.29374288964732653, + "grad_norm": 2.5288590185853885, + "learning_rate": 1.2394016034192654e-06, + "loss": 0.1699, + "step": 1291 + }, + { + "epoch": 0.2939704209328783, + "grad_norm": 1.3441395527785462, + "learning_rate": 1.2393852121160548e-06, + "loss": 0.1406, + "step": 1292 + }, + { + "epoch": 0.29419795221843004, + "grad_norm": 2.53263687980971, + "learning_rate": 1.2393688082559357e-06, + "loss": 0.1629, + "step": 1293 + }, + { + "epoch": 0.2944254835039818, + "grad_norm": 1.711669162319686, + "learning_rate": 1.2393523918392433e-06, + "loss": 0.0766, + "step": 1294 + }, + { + "epoch": 0.29465301478953354, + "grad_norm": 2.2138801021106214, + "learning_rate": 1.2393359628663133e-06, + "loss": 0.2713, + "step": 1295 + }, + { + "epoch": 0.29488054607508535, + "grad_norm": 1.7610053159614951, + "learning_rate": 1.239319521337481e-06, + "loss": 0.1454, + "step": 1296 + }, + { + "epoch": 0.2951080773606371, + "grad_norm": 1.1164148240700726, + "learning_rate": 1.2393030672530828e-06, + "loss": 0.0873, + "step": 1297 + }, + { + "epoch": 0.29533560864618885, + "grad_norm": 2.5581318128260264, + "learning_rate": 1.239286600613455e-06, + "loss": 0.1986, + "step": 1298 + }, + { + "epoch": 0.2955631399317406, + "grad_norm": 2.130651315015736, + "learning_rate": 1.2392701214189343e-06, + "loss": 0.1141, + "step": 1299 + }, + { + "epoch": 0.29579067121729236, + "grad_norm": 2.483114350460134, + "learning_rate": 1.2392536296698571e-06, + "loss": 0.1921, + "step": 1300 + }, + { + "epoch": 0.29601820250284416, + "grad_norm": 1.4378642619116868, + "learning_rate": 1.2392371253665605e-06, + "loss": 0.1495, + "step": 1301 + }, + { + "epoch": 0.2962457337883959, + "grad_norm": 1.4002315021023155, + "learning_rate": 1.2392206085093823e-06, + "loss": 0.1929, + "step": 1302 + }, + { + "epoch": 0.29647326507394767, + "grad_norm": 1.8673010049645025, + "learning_rate": 1.2392040790986594e-06, + "loss": 0.17, + "step": 1303 + }, + { + "epoch": 0.2967007963594994, + "grad_norm": 2.360098461281988, + "learning_rate": 1.2391875371347303e-06, + "loss": 0.1221, + "step": 1304 + }, + { + "epoch": 0.29692832764505117, + "grad_norm": 1.149049669073457, + "learning_rate": 1.2391709826179327e-06, + "loss": 0.0648, + "step": 1305 + }, + { + "epoch": 0.297155858930603, + "grad_norm": 2.7074403192946948, + "learning_rate": 1.239154415548605e-06, + "loss": 0.101, + "step": 1306 + }, + { + "epoch": 0.29738339021615473, + "grad_norm": 2.3404087696279006, + "learning_rate": 1.2391378359270859e-06, + "loss": 0.169, + "step": 1307 + }, + { + "epoch": 0.2976109215017065, + "grad_norm": 2.051588482824801, + "learning_rate": 1.2391212437537138e-06, + "loss": 0.1231, + "step": 1308 + }, + { + "epoch": 0.29783845278725823, + "grad_norm": 2.242825915734783, + "learning_rate": 1.2391046390288287e-06, + "loss": 0.0926, + "step": 1309 + }, + { + "epoch": 0.29806598407281, + "grad_norm": 1.544022609869059, + "learning_rate": 1.2390880217527692e-06, + "loss": 0.1112, + "step": 1310 + }, + { + "epoch": 0.2982935153583618, + "grad_norm": 2.1469992174478802, + "learning_rate": 1.2390713919258752e-06, + "loss": 0.142, + "step": 1311 + }, + { + "epoch": 0.29852104664391355, + "grad_norm": 1.7368698248867847, + "learning_rate": 1.2390547495484866e-06, + "loss": 0.173, + "step": 1312 + }, + { + "epoch": 0.2987485779294653, + "grad_norm": 1.935573701314408, + "learning_rate": 1.2390380946209436e-06, + "loss": 0.1042, + "step": 1313 + }, + { + "epoch": 0.29897610921501705, + "grad_norm": 2.312605974001612, + "learning_rate": 1.2390214271435863e-06, + "loss": 0.1058, + "step": 1314 + }, + { + "epoch": 0.2992036405005688, + "grad_norm": 2.0971420605328417, + "learning_rate": 1.2390047471167557e-06, + "loss": 0.1131, + "step": 1315 + }, + { + "epoch": 0.2994311717861206, + "grad_norm": 2.3687418924208568, + "learning_rate": 1.2389880545407926e-06, + "loss": 0.1197, + "step": 1316 + }, + { + "epoch": 0.29965870307167236, + "grad_norm": 2.1772555556434163, + "learning_rate": 1.2389713494160379e-06, + "loss": 0.1362, + "step": 1317 + }, + { + "epoch": 0.2998862343572241, + "grad_norm": 2.347318027018112, + "learning_rate": 1.2389546317428335e-06, + "loss": 0.115, + "step": 1318 + }, + { + "epoch": 0.30011376564277586, + "grad_norm": 2.393814516886665, + "learning_rate": 1.2389379015215208e-06, + "loss": 0.1366, + "step": 1319 + }, + { + "epoch": 0.3003412969283277, + "grad_norm": 3.2734975537480295, + "learning_rate": 1.2389211587524416e-06, + "loss": 0.1663, + "step": 1320 + }, + { + "epoch": 0.3005688282138794, + "grad_norm": 2.476288177865563, + "learning_rate": 1.2389044034359383e-06, + "loss": 0.1585, + "step": 1321 + }, + { + "epoch": 0.3007963594994312, + "grad_norm": 2.0801439225061094, + "learning_rate": 1.2388876355723533e-06, + "loss": 0.0896, + "step": 1322 + }, + { + "epoch": 0.30102389078498293, + "grad_norm": 3.3649680946347043, + "learning_rate": 1.2388708551620295e-06, + "loss": 0.1927, + "step": 1323 + }, + { + "epoch": 0.3012514220705347, + "grad_norm": 1.6798655722697455, + "learning_rate": 1.2388540622053095e-06, + "loss": 0.1514, + "step": 1324 + }, + { + "epoch": 0.3014789533560865, + "grad_norm": 2.2188046109768433, + "learning_rate": 1.2388372567025367e-06, + "loss": 0.1886, + "step": 1325 + }, + { + "epoch": 0.30170648464163824, + "grad_norm": 3.8153066915737437, + "learning_rate": 1.2388204386540546e-06, + "loss": 0.1965, + "step": 1326 + }, + { + "epoch": 0.30193401592719, + "grad_norm": 3.157753400659548, + "learning_rate": 1.238803608060207e-06, + "loss": 0.1571, + "step": 1327 + }, + { + "epoch": 0.30216154721274174, + "grad_norm": 3.5073308744349068, + "learning_rate": 1.2387867649213376e-06, + "loss": 0.1517, + "step": 1328 + }, + { + "epoch": 0.3023890784982935, + "grad_norm": 2.1553373677368017, + "learning_rate": 1.2387699092377908e-06, + "loss": 0.2075, + "step": 1329 + }, + { + "epoch": 0.3026166097838453, + "grad_norm": 2.282662404456656, + "learning_rate": 1.2387530410099113e-06, + "loss": 0.1414, + "step": 1330 + }, + { + "epoch": 0.30284414106939705, + "grad_norm": 1.9117313188558176, + "learning_rate": 1.2387361602380436e-06, + "loss": 0.1266, + "step": 1331 + }, + { + "epoch": 0.3030716723549488, + "grad_norm": 1.8130769390361627, + "learning_rate": 1.238719266922533e-06, + "loss": 0.1256, + "step": 1332 + }, + { + "epoch": 0.30329920364050056, + "grad_norm": 2.8536582395130865, + "learning_rate": 1.238702361063724e-06, + "loss": 0.1814, + "step": 1333 + }, + { + "epoch": 0.3035267349260523, + "grad_norm": 2.8179372897441057, + "learning_rate": 1.2386854426619633e-06, + "loss": 0.1992, + "step": 1334 + }, + { + "epoch": 0.3037542662116041, + "grad_norm": 2.8198093864971274, + "learning_rate": 1.2386685117175956e-06, + "loss": 0.1244, + "step": 1335 + }, + { + "epoch": 0.30398179749715587, + "grad_norm": 2.2011541075966514, + "learning_rate": 1.2386515682309676e-06, + "loss": 0.1803, + "step": 1336 + }, + { + "epoch": 0.3042093287827076, + "grad_norm": 2.1936077489701513, + "learning_rate": 1.2386346122024253e-06, + "loss": 0.1471, + "step": 1337 + }, + { + "epoch": 0.3044368600682594, + "grad_norm": 2.3414747287597586, + "learning_rate": 1.2386176436323154e-06, + "loss": 0.1968, + "step": 1338 + }, + { + "epoch": 0.3046643913538111, + "grad_norm": 2.4409570439965265, + "learning_rate": 1.2386006625209847e-06, + "loss": 0.1304, + "step": 1339 + }, + { + "epoch": 0.30489192263936293, + "grad_norm": 1.5611171851558396, + "learning_rate": 1.2385836688687802e-06, + "loss": 0.1826, + "step": 1340 + }, + { + "epoch": 0.3051194539249147, + "grad_norm": 2.1495118418095758, + "learning_rate": 1.2385666626760493e-06, + "loss": 0.1299, + "step": 1341 + }, + { + "epoch": 0.30534698521046644, + "grad_norm": 2.427709246358795, + "learning_rate": 1.2385496439431395e-06, + "loss": 0.2042, + "step": 1342 + }, + { + "epoch": 0.3055745164960182, + "grad_norm": 2.1582025174813477, + "learning_rate": 1.2385326126703986e-06, + "loss": 0.2137, + "step": 1343 + }, + { + "epoch": 0.30580204778156994, + "grad_norm": 2.692539956880222, + "learning_rate": 1.2385155688581746e-06, + "loss": 0.1421, + "step": 1344 + }, + { + "epoch": 0.30602957906712175, + "grad_norm": 1.7945869672607249, + "learning_rate": 1.238498512506816e-06, + "loss": 0.1111, + "step": 1345 + }, + { + "epoch": 0.3062571103526735, + "grad_norm": 2.4481827112208236, + "learning_rate": 1.2384814436166715e-06, + "loss": 0.198, + "step": 1346 + }, + { + "epoch": 0.30648464163822525, + "grad_norm": 1.515871383058607, + "learning_rate": 1.2384643621880898e-06, + "loss": 0.1038, + "step": 1347 + }, + { + "epoch": 0.306712172923777, + "grad_norm": 1.1946482530332898, + "learning_rate": 1.2384472682214201e-06, + "loss": 0.1369, + "step": 1348 + }, + { + "epoch": 0.30693970420932876, + "grad_norm": 2.487738709975654, + "learning_rate": 1.2384301617170116e-06, + "loss": 0.1327, + "step": 1349 + }, + { + "epoch": 0.30716723549488056, + "grad_norm": 2.2787395109301496, + "learning_rate": 1.2384130426752142e-06, + "loss": 0.2413, + "step": 1350 + }, + { + "epoch": 0.3073947667804323, + "grad_norm": 3.009302223374968, + "learning_rate": 1.2383959110963775e-06, + "loss": 0.1308, + "step": 1351 + }, + { + "epoch": 0.30762229806598407, + "grad_norm": 1.9025690830276472, + "learning_rate": 1.2383787669808518e-06, + "loss": 0.1292, + "step": 1352 + }, + { + "epoch": 0.3078498293515358, + "grad_norm": 2.478845494921755, + "learning_rate": 1.2383616103289871e-06, + "loss": 0.1388, + "step": 1353 + }, + { + "epoch": 0.3080773606370876, + "grad_norm": 2.378732608316692, + "learning_rate": 1.238344441141135e-06, + "loss": 0.1445, + "step": 1354 + }, + { + "epoch": 0.3083048919226394, + "grad_norm": 2.7865872438029604, + "learning_rate": 1.2383272594176454e-06, + "loss": 0.1987, + "step": 1355 + }, + { + "epoch": 0.30853242320819113, + "grad_norm": 4.941618358694522, + "learning_rate": 1.23831006515887e-06, + "loss": 0.2554, + "step": 1356 + }, + { + "epoch": 0.3087599544937429, + "grad_norm": 2.3648132845973433, + "learning_rate": 1.2382928583651601e-06, + "loss": 0.1316, + "step": 1357 + }, + { + "epoch": 0.30898748577929463, + "grad_norm": 4.506401473111362, + "learning_rate": 1.2382756390368674e-06, + "loss": 0.2064, + "step": 1358 + }, + { + "epoch": 0.30921501706484644, + "grad_norm": 2.9032765876327367, + "learning_rate": 1.2382584071743438e-06, + "loss": 0.1722, + "step": 1359 + }, + { + "epoch": 0.3094425483503982, + "grad_norm": 1.9393994140062751, + "learning_rate": 1.2382411627779414e-06, + "loss": 0.1198, + "step": 1360 + }, + { + "epoch": 0.30967007963594995, + "grad_norm": 2.9033108515884813, + "learning_rate": 1.2382239058480128e-06, + "loss": 0.1898, + "step": 1361 + }, + { + "epoch": 0.3098976109215017, + "grad_norm": 1.7474335493562614, + "learning_rate": 1.2382066363849106e-06, + "loss": 0.0873, + "step": 1362 + }, + { + "epoch": 0.31012514220705345, + "grad_norm": 2.7271456687130535, + "learning_rate": 1.2381893543889878e-06, + "loss": 0.1164, + "step": 1363 + }, + { + "epoch": 0.31035267349260526, + "grad_norm": 2.442928236122465, + "learning_rate": 1.2381720598605976e-06, + "loss": 0.1866, + "step": 1364 + }, + { + "epoch": 0.310580204778157, + "grad_norm": 2.2683129394489336, + "learning_rate": 1.2381547528000934e-06, + "loss": 0.1821, + "step": 1365 + }, + { + "epoch": 0.31080773606370876, + "grad_norm": 2.2249086535735665, + "learning_rate": 1.238137433207829e-06, + "loss": 0.1424, + "step": 1366 + }, + { + "epoch": 0.3110352673492605, + "grad_norm": 2.3904987570803424, + "learning_rate": 1.2381201010841585e-06, + "loss": 0.1241, + "step": 1367 + }, + { + "epoch": 0.31126279863481227, + "grad_norm": 3.3568266666034563, + "learning_rate": 1.2381027564294359e-06, + "loss": 0.2353, + "step": 1368 + }, + { + "epoch": 0.3114903299203641, + "grad_norm": 1.089454840869158, + "learning_rate": 1.238085399244016e-06, + "loss": 0.1225, + "step": 1369 + }, + { + "epoch": 0.3117178612059158, + "grad_norm": 1.9458300879080968, + "learning_rate": 1.2380680295282532e-06, + "loss": 0.1969, + "step": 1370 + }, + { + "epoch": 0.3119453924914676, + "grad_norm": 2.205384735367907, + "learning_rate": 1.2380506472825025e-06, + "loss": 0.1051, + "step": 1371 + }, + { + "epoch": 0.31217292377701933, + "grad_norm": 3.009662856212347, + "learning_rate": 1.2380332525071194e-06, + "loss": 0.1357, + "step": 1372 + }, + { + "epoch": 0.3124004550625711, + "grad_norm": 1.7090482855605478, + "learning_rate": 1.2380158452024595e-06, + "loss": 0.1093, + "step": 1373 + }, + { + "epoch": 0.3126279863481229, + "grad_norm": 1.534224124551681, + "learning_rate": 1.2379984253688783e-06, + "loss": 0.1668, + "step": 1374 + }, + { + "epoch": 0.31285551763367464, + "grad_norm": 3.5021434401932847, + "learning_rate": 1.237980993006732e-06, + "loss": 0.2277, + "step": 1375 + }, + { + "epoch": 0.3130830489192264, + "grad_norm": 2.262191330295691, + "learning_rate": 1.2379635481163768e-06, + "loss": 0.2239, + "step": 1376 + }, + { + "epoch": 0.31331058020477814, + "grad_norm": 2.053106476927726, + "learning_rate": 1.2379460906981692e-06, + "loss": 0.1531, + "step": 1377 + }, + { + "epoch": 0.3135381114903299, + "grad_norm": 1.5039843101917945, + "learning_rate": 1.237928620752466e-06, + "loss": 0.15, + "step": 1378 + }, + { + "epoch": 0.3137656427758817, + "grad_norm": 3.906363623322993, + "learning_rate": 1.2379111382796246e-06, + "loss": 0.1751, + "step": 1379 + }, + { + "epoch": 0.31399317406143346, + "grad_norm": 1.8885019656196067, + "learning_rate": 1.2378936432800017e-06, + "loss": 0.1706, + "step": 1380 + }, + { + "epoch": 0.3142207053469852, + "grad_norm": 2.868561218863271, + "learning_rate": 1.2378761357539554e-06, + "loss": 0.2075, + "step": 1381 + }, + { + "epoch": 0.31444823663253696, + "grad_norm": 3.0041888547459332, + "learning_rate": 1.2378586157018434e-06, + "loss": 0.1174, + "step": 1382 + }, + { + "epoch": 0.3146757679180887, + "grad_norm": 4.579577850116939, + "learning_rate": 1.2378410831240235e-06, + "loss": 0.0989, + "step": 1383 + }, + { + "epoch": 0.3149032992036405, + "grad_norm": 2.369378417403813, + "learning_rate": 1.2378235380208542e-06, + "loss": 0.1939, + "step": 1384 + }, + { + "epoch": 0.31513083048919227, + "grad_norm": 1.3212831618328642, + "learning_rate": 1.2378059803926941e-06, + "loss": 0.0939, + "step": 1385 + }, + { + "epoch": 0.315358361774744, + "grad_norm": 1.7303755529967304, + "learning_rate": 1.2377884102399023e-06, + "loss": 0.0969, + "step": 1386 + }, + { + "epoch": 0.3155858930602958, + "grad_norm": 2.1687520273562058, + "learning_rate": 1.2377708275628375e-06, + "loss": 0.1362, + "step": 1387 + }, + { + "epoch": 0.3158134243458475, + "grad_norm": 2.1690288580082115, + "learning_rate": 1.2377532323618593e-06, + "loss": 0.15, + "step": 1388 + }, + { + "epoch": 0.31604095563139933, + "grad_norm": 2.6772029483945987, + "learning_rate": 1.2377356246373271e-06, + "loss": 0.156, + "step": 1389 + }, + { + "epoch": 0.3162684869169511, + "grad_norm": 2.3434657796789335, + "learning_rate": 1.2377180043896012e-06, + "loss": 0.1605, + "step": 1390 + }, + { + "epoch": 0.31649601820250284, + "grad_norm": 0.6778545782857848, + "learning_rate": 1.2377003716190411e-06, + "loss": 0.0678, + "step": 1391 + }, + { + "epoch": 0.3167235494880546, + "grad_norm": 3.4250150700071598, + "learning_rate": 1.2376827263260078e-06, + "loss": 0.2121, + "step": 1392 + }, + { + "epoch": 0.3169510807736064, + "grad_norm": 2.212315886742257, + "learning_rate": 1.2376650685108612e-06, + "loss": 0.0926, + "step": 1393 + }, + { + "epoch": 0.31717861205915815, + "grad_norm": 2.544863602950953, + "learning_rate": 1.2376473981739632e-06, + "loss": 0.2237, + "step": 1394 + }, + { + "epoch": 0.3174061433447099, + "grad_norm": 3.814819174160117, + "learning_rate": 1.237629715315674e-06, + "loss": 0.1803, + "step": 1395 + }, + { + "epoch": 0.31763367463026165, + "grad_norm": 2.016517421139005, + "learning_rate": 1.2376120199363554e-06, + "loss": 0.1629, + "step": 1396 + }, + { + "epoch": 0.3178612059158134, + "grad_norm": 2.2015049586409736, + "learning_rate": 1.2375943120363692e-06, + "loss": 0.1774, + "step": 1397 + }, + { + "epoch": 0.3180887372013652, + "grad_norm": 1.4748569856435156, + "learning_rate": 1.2375765916160773e-06, + "loss": 0.0892, + "step": 1398 + }, + { + "epoch": 0.31831626848691696, + "grad_norm": 4.396707263537578, + "learning_rate": 1.2375588586758415e-06, + "loss": 0.1694, + "step": 1399 + }, + { + "epoch": 0.3185437997724687, + "grad_norm": 2.815879553875731, + "learning_rate": 1.2375411132160245e-06, + "loss": 0.1574, + "step": 1400 + }, + { + "epoch": 0.31877133105802047, + "grad_norm": 2.854288805845529, + "learning_rate": 1.2375233552369892e-06, + "loss": 0.21, + "step": 1401 + }, + { + "epoch": 0.3189988623435722, + "grad_norm": 2.4400815913032137, + "learning_rate": 1.237505584739098e-06, + "loss": 0.1222, + "step": 1402 + }, + { + "epoch": 0.31922639362912403, + "grad_norm": 2.448908720687653, + "learning_rate": 1.2374878017227147e-06, + "loss": 0.1494, + "step": 1403 + }, + { + "epoch": 0.3194539249146758, + "grad_norm": 1.687277139104771, + "learning_rate": 1.237470006188202e-06, + "loss": 0.0914, + "step": 1404 + }, + { + "epoch": 0.31968145620022753, + "grad_norm": 2.2925412031811776, + "learning_rate": 1.2374521981359245e-06, + "loss": 0.0881, + "step": 1405 + }, + { + "epoch": 0.3199089874857793, + "grad_norm": 4.072013951747903, + "learning_rate": 1.2374343775662456e-06, + "loss": 0.1696, + "step": 1406 + }, + { + "epoch": 0.32013651877133104, + "grad_norm": 2.5836523726802545, + "learning_rate": 1.2374165444795296e-06, + "loss": 0.1346, + "step": 1407 + }, + { + "epoch": 0.32036405005688284, + "grad_norm": 1.599365912466973, + "learning_rate": 1.237398698876141e-06, + "loss": 0.1517, + "step": 1408 + }, + { + "epoch": 0.3205915813424346, + "grad_norm": 3.0092798969280694, + "learning_rate": 1.2373808407564446e-06, + "loss": 0.1388, + "step": 1409 + }, + { + "epoch": 0.32081911262798635, + "grad_norm": 2.7681165920899917, + "learning_rate": 1.2373629701208053e-06, + "loss": 0.0903, + "step": 1410 + }, + { + "epoch": 0.3210466439135381, + "grad_norm": 2.33093444363794, + "learning_rate": 1.2373450869695883e-06, + "loss": 0.1258, + "step": 1411 + }, + { + "epoch": 0.32127417519908985, + "grad_norm": 1.3355433862599404, + "learning_rate": 1.2373271913031593e-06, + "loss": 0.0933, + "step": 1412 + }, + { + "epoch": 0.32150170648464166, + "grad_norm": 3.6953831980698326, + "learning_rate": 1.237309283121884e-06, + "loss": 0.1799, + "step": 1413 + }, + { + "epoch": 0.3217292377701934, + "grad_norm": 1.4551969737419015, + "learning_rate": 1.237291362426128e-06, + "loss": 0.132, + "step": 1414 + }, + { + "epoch": 0.32195676905574516, + "grad_norm": 1.3811005049648641, + "learning_rate": 1.2372734292162584e-06, + "loss": 0.1393, + "step": 1415 + }, + { + "epoch": 0.3221843003412969, + "grad_norm": 2.1330825424153894, + "learning_rate": 1.237255483492641e-06, + "loss": 0.1174, + "step": 1416 + }, + { + "epoch": 0.32241183162684867, + "grad_norm": 2.487852309027526, + "learning_rate": 1.2372375252556429e-06, + "loss": 0.1417, + "step": 1417 + }, + { + "epoch": 0.3226393629124005, + "grad_norm": 1.913453520107482, + "learning_rate": 1.2372195545056308e-06, + "loss": 0.0859, + "step": 1418 + }, + { + "epoch": 0.3228668941979522, + "grad_norm": 2.646487035772668, + "learning_rate": 1.2372015712429725e-06, + "loss": 0.1506, + "step": 1419 + }, + { + "epoch": 0.323094425483504, + "grad_norm": 3.245246064065534, + "learning_rate": 1.237183575468035e-06, + "loss": 0.1634, + "step": 1420 + }, + { + "epoch": 0.32332195676905573, + "grad_norm": 2.0793785188235714, + "learning_rate": 1.2371655671811866e-06, + "loss": 0.0893, + "step": 1421 + }, + { + "epoch": 0.3235494880546075, + "grad_norm": 3.3709416770572203, + "learning_rate": 1.237147546382795e-06, + "loss": 0.1863, + "step": 1422 + }, + { + "epoch": 0.3237770193401593, + "grad_norm": 1.9834221905090585, + "learning_rate": 1.237129513073229e-06, + "loss": 0.1022, + "step": 1423 + }, + { + "epoch": 0.32400455062571104, + "grad_norm": 1.5082121448900583, + "learning_rate": 1.2371114672528565e-06, + "loss": 0.1191, + "step": 1424 + }, + { + "epoch": 0.3242320819112628, + "grad_norm": 2.1618548110562728, + "learning_rate": 1.2370934089220466e-06, + "loss": 0.1036, + "step": 1425 + }, + { + "epoch": 0.32445961319681454, + "grad_norm": 3.0267433921571247, + "learning_rate": 1.2370753380811685e-06, + "loss": 0.1331, + "step": 1426 + }, + { + "epoch": 0.32468714448236635, + "grad_norm": 2.328447963796751, + "learning_rate": 1.2370572547305915e-06, + "loss": 0.0995, + "step": 1427 + }, + { + "epoch": 0.3249146757679181, + "grad_norm": 1.5043231729048063, + "learning_rate": 1.237039158870685e-06, + "loss": 0.0938, + "step": 1428 + }, + { + "epoch": 0.32514220705346986, + "grad_norm": 1.489228734233685, + "learning_rate": 1.237021050501819e-06, + "loss": 0.183, + "step": 1429 + }, + { + "epoch": 0.3253697383390216, + "grad_norm": 1.3957978981478696, + "learning_rate": 1.2370029296243638e-06, + "loss": 0.137, + "step": 1430 + }, + { + "epoch": 0.32559726962457336, + "grad_norm": 1.6002182368576523, + "learning_rate": 1.2369847962386893e-06, + "loss": 0.1383, + "step": 1431 + }, + { + "epoch": 0.32582480091012517, + "grad_norm": 1.4292210315358218, + "learning_rate": 1.2369666503451665e-06, + "loss": 0.1306, + "step": 1432 + }, + { + "epoch": 0.3260523321956769, + "grad_norm": 3.3664684804237153, + "learning_rate": 1.2369484919441662e-06, + "loss": 0.1495, + "step": 1433 + }, + { + "epoch": 0.32627986348122867, + "grad_norm": 2.213331961777345, + "learning_rate": 1.2369303210360592e-06, + "loss": 0.104, + "step": 1434 + }, + { + "epoch": 0.3265073947667804, + "grad_norm": 2.7142292072973224, + "learning_rate": 1.2369121376212174e-06, + "loss": 0.23, + "step": 1435 + }, + { + "epoch": 0.3267349260523322, + "grad_norm": 2.5989691526691905, + "learning_rate": 1.236893941700012e-06, + "loss": 0.164, + "step": 1436 + }, + { + "epoch": 0.326962457337884, + "grad_norm": 2.7909547799726666, + "learning_rate": 1.236875733272815e-06, + "loss": 0.1231, + "step": 1437 + }, + { + "epoch": 0.32718998862343573, + "grad_norm": 1.5171731115835614, + "learning_rate": 1.236857512339999e-06, + "loss": 0.1568, + "step": 1438 + }, + { + "epoch": 0.3274175199089875, + "grad_norm": 2.7299542263447987, + "learning_rate": 1.2368392789019356e-06, + "loss": 0.192, + "step": 1439 + }, + { + "epoch": 0.32764505119453924, + "grad_norm": 3.3727472288143074, + "learning_rate": 1.2368210329589982e-06, + "loss": 0.1387, + "step": 1440 + }, + { + "epoch": 0.327872582480091, + "grad_norm": 1.9170836491353773, + "learning_rate": 1.236802774511559e-06, + "loss": 0.1325, + "step": 1441 + }, + { + "epoch": 0.3281001137656428, + "grad_norm": 1.3065482462033158, + "learning_rate": 1.2367845035599919e-06, + "loss": 0.1347, + "step": 1442 + }, + { + "epoch": 0.32832764505119455, + "grad_norm": 1.170941186213132, + "learning_rate": 1.2367662201046698e-06, + "loss": 0.0776, + "step": 1443 + }, + { + "epoch": 0.3285551763367463, + "grad_norm": 2.650987144280419, + "learning_rate": 1.2367479241459666e-06, + "loss": 0.137, + "step": 1444 + }, + { + "epoch": 0.32878270762229805, + "grad_norm": 2.607148114564764, + "learning_rate": 1.2367296156842562e-06, + "loss": 0.1722, + "step": 1445 + }, + { + "epoch": 0.3290102389078498, + "grad_norm": 1.9899827485582995, + "learning_rate": 1.2367112947199128e-06, + "loss": 0.0736, + "step": 1446 + }, + { + "epoch": 0.3292377701934016, + "grad_norm": 1.7939831808853237, + "learning_rate": 1.2366929612533109e-06, + "loss": 0.1678, + "step": 1447 + }, + { + "epoch": 0.32946530147895337, + "grad_norm": 2.18373767390569, + "learning_rate": 1.2366746152848249e-06, + "loss": 0.1248, + "step": 1448 + }, + { + "epoch": 0.3296928327645051, + "grad_norm": 1.6786238142668934, + "learning_rate": 1.2366562568148301e-06, + "loss": 0.113, + "step": 1449 + }, + { + "epoch": 0.32992036405005687, + "grad_norm": 2.1261100071128216, + "learning_rate": 1.2366378858437016e-06, + "loss": 0.2123, + "step": 1450 + }, + { + "epoch": 0.3301478953356086, + "grad_norm": 2.428509398466611, + "learning_rate": 1.2366195023718152e-06, + "loss": 0.1528, + "step": 1451 + }, + { + "epoch": 0.33037542662116043, + "grad_norm": 2.0995327071550745, + "learning_rate": 1.2366011063995458e-06, + "loss": 0.1479, + "step": 1452 + }, + { + "epoch": 0.3306029579067122, + "grad_norm": 2.7571183190896544, + "learning_rate": 1.2365826979272702e-06, + "loss": 0.2723, + "step": 1453 + }, + { + "epoch": 0.33083048919226393, + "grad_norm": 1.37616338455632, + "learning_rate": 1.2365642769553644e-06, + "loss": 0.186, + "step": 1454 + }, + { + "epoch": 0.3310580204778157, + "grad_norm": 3.562357493800966, + "learning_rate": 1.2365458434842046e-06, + "loss": 0.1935, + "step": 1455 + }, + { + "epoch": 0.33128555176336744, + "grad_norm": 3.0055653446008757, + "learning_rate": 1.2365273975141675e-06, + "loss": 0.1349, + "step": 1456 + }, + { + "epoch": 0.33151308304891924, + "grad_norm": 1.9084405856232312, + "learning_rate": 1.2365089390456308e-06, + "loss": 0.1579, + "step": 1457 + }, + { + "epoch": 0.331740614334471, + "grad_norm": 2.509163343882828, + "learning_rate": 1.236490468078971e-06, + "loss": 0.1399, + "step": 1458 + }, + { + "epoch": 0.33196814562002275, + "grad_norm": 2.762892695226815, + "learning_rate": 1.2364719846145662e-06, + "loss": 0.1123, + "step": 1459 + }, + { + "epoch": 0.3321956769055745, + "grad_norm": 1.509507061278107, + "learning_rate": 1.2364534886527937e-06, + "loss": 0.0761, + "step": 1460 + }, + { + "epoch": 0.3324232081911263, + "grad_norm": 2.125250796251379, + "learning_rate": 1.2364349801940317e-06, + "loss": 0.1979, + "step": 1461 + }, + { + "epoch": 0.33265073947667806, + "grad_norm": 1.564090485896412, + "learning_rate": 1.2364164592386588e-06, + "loss": 0.1273, + "step": 1462 + }, + { + "epoch": 0.3328782707622298, + "grad_norm": 2.1634983990483825, + "learning_rate": 1.2363979257870528e-06, + "loss": 0.1942, + "step": 1463 + }, + { + "epoch": 0.33310580204778156, + "grad_norm": 3.3365080077721854, + "learning_rate": 1.2363793798395932e-06, + "loss": 0.1447, + "step": 1464 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 3.9561125952221525, + "learning_rate": 1.2363608213966588e-06, + "loss": 0.2634, + "step": 1465 + }, + { + "epoch": 0.3335608646188851, + "grad_norm": 2.251288332330216, + "learning_rate": 1.2363422504586286e-06, + "loss": 0.124, + "step": 1466 + }, + { + "epoch": 0.3337883959044369, + "grad_norm": 2.6458027444889844, + "learning_rate": 1.2363236670258827e-06, + "loss": 0.1211, + "step": 1467 + }, + { + "epoch": 0.3340159271899886, + "grad_norm": 2.256871006091209, + "learning_rate": 1.2363050710988003e-06, + "loss": 0.18, + "step": 1468 + }, + { + "epoch": 0.3342434584755404, + "grad_norm": 2.3666680923714973, + "learning_rate": 1.236286462677762e-06, + "loss": 0.1146, + "step": 1469 + }, + { + "epoch": 0.33447098976109213, + "grad_norm": 1.9163232744074985, + "learning_rate": 1.236267841763148e-06, + "loss": 0.2394, + "step": 1470 + }, + { + "epoch": 0.33469852104664394, + "grad_norm": 1.0000779857222588, + "learning_rate": 1.2362492083553387e-06, + "loss": 0.1034, + "step": 1471 + }, + { + "epoch": 0.3349260523321957, + "grad_norm": 4.037436172394108, + "learning_rate": 1.236230562454715e-06, + "loss": 0.1436, + "step": 1472 + }, + { + "epoch": 0.33515358361774744, + "grad_norm": 2.470429929609661, + "learning_rate": 1.2362119040616582e-06, + "loss": 0.2313, + "step": 1473 + }, + { + "epoch": 0.3353811149032992, + "grad_norm": 1.7641673534435953, + "learning_rate": 1.2361932331765492e-06, + "loss": 0.1131, + "step": 1474 + }, + { + "epoch": 0.33560864618885095, + "grad_norm": 2.283488425453049, + "learning_rate": 1.2361745497997702e-06, + "loss": 0.2092, + "step": 1475 + }, + { + "epoch": 0.33583617747440275, + "grad_norm": 2.44990291433213, + "learning_rate": 1.2361558539317023e-06, + "loss": 0.1677, + "step": 1476 + }, + { + "epoch": 0.3360637087599545, + "grad_norm": 2.9170619880191664, + "learning_rate": 1.2361371455727284e-06, + "loss": 0.1622, + "step": 1477 + }, + { + "epoch": 0.33629124004550626, + "grad_norm": 1.8904885782974936, + "learning_rate": 1.2361184247232302e-06, + "loss": 0.174, + "step": 1478 + }, + { + "epoch": 0.336518771331058, + "grad_norm": 2.729547155076976, + "learning_rate": 1.2360996913835907e-06, + "loss": 0.2034, + "step": 1479 + }, + { + "epoch": 0.33674630261660976, + "grad_norm": 1.5905963504881455, + "learning_rate": 1.2360809455541928e-06, + "loss": 0.1115, + "step": 1480 + }, + { + "epoch": 0.33697383390216157, + "grad_norm": 2.0824120193071987, + "learning_rate": 1.2360621872354195e-06, + "loss": 0.0922, + "step": 1481 + }, + { + "epoch": 0.3372013651877133, + "grad_norm": 1.7589178939868375, + "learning_rate": 1.236043416427654e-06, + "loss": 0.1538, + "step": 1482 + }, + { + "epoch": 0.33742889647326507, + "grad_norm": 2.2945363962104595, + "learning_rate": 1.2360246331312804e-06, + "loss": 0.102, + "step": 1483 + }, + { + "epoch": 0.3376564277588168, + "grad_norm": 2.6126607217811233, + "learning_rate": 1.2360058373466821e-06, + "loss": 0.1287, + "step": 1484 + }, + { + "epoch": 0.3378839590443686, + "grad_norm": 3.6250866550257776, + "learning_rate": 1.2359870290742437e-06, + "loss": 0.2267, + "step": 1485 + }, + { + "epoch": 0.3381114903299204, + "grad_norm": 1.9639418873270704, + "learning_rate": 1.2359682083143494e-06, + "loss": 0.1678, + "step": 1486 + }, + { + "epoch": 0.33833902161547214, + "grad_norm": 1.7491629271174762, + "learning_rate": 1.2359493750673835e-06, + "loss": 0.0793, + "step": 1487 + }, + { + "epoch": 0.3385665529010239, + "grad_norm": 2.9106462432976703, + "learning_rate": 1.2359305293337316e-06, + "loss": 0.1325, + "step": 1488 + }, + { + "epoch": 0.33879408418657564, + "grad_norm": 1.276937248770152, + "learning_rate": 1.2359116711137785e-06, + "loss": 0.0878, + "step": 1489 + }, + { + "epoch": 0.3390216154721274, + "grad_norm": 2.7341692722816355, + "learning_rate": 1.2358928004079095e-06, + "loss": 0.204, + "step": 1490 + }, + { + "epoch": 0.3392491467576792, + "grad_norm": 1.938938866233754, + "learning_rate": 1.2358739172165108e-06, + "loss": 0.2461, + "step": 1491 + }, + { + "epoch": 0.33947667804323095, + "grad_norm": 2.4567051660275325, + "learning_rate": 1.235855021539968e-06, + "loss": 0.108, + "step": 1492 + }, + { + "epoch": 0.3397042093287827, + "grad_norm": 1.7865319339113264, + "learning_rate": 1.2358361133786668e-06, + "loss": 0.1137, + "step": 1493 + }, + { + "epoch": 0.33993174061433445, + "grad_norm": 1.2860622511723663, + "learning_rate": 1.2358171927329946e-06, + "loss": 0.1066, + "step": 1494 + }, + { + "epoch": 0.34015927189988626, + "grad_norm": 2.3217877744819155, + "learning_rate": 1.2357982596033374e-06, + "loss": 0.1195, + "step": 1495 + }, + { + "epoch": 0.340386803185438, + "grad_norm": 1.405655578490845, + "learning_rate": 1.2357793139900823e-06, + "loss": 0.1675, + "step": 1496 + }, + { + "epoch": 0.34061433447098977, + "grad_norm": 1.590690138510479, + "learning_rate": 1.2357603558936168e-06, + "loss": 0.0715, + "step": 1497 + }, + { + "epoch": 0.3408418657565415, + "grad_norm": 3.6662652923931276, + "learning_rate": 1.2357413853143282e-06, + "loss": 0.2245, + "step": 1498 + }, + { + "epoch": 0.34106939704209327, + "grad_norm": 2.0102829011772885, + "learning_rate": 1.2357224022526041e-06, + "loss": 0.1646, + "step": 1499 + }, + { + "epoch": 0.3412969283276451, + "grad_norm": 2.0702589837056724, + "learning_rate": 1.2357034067088327e-06, + "loss": 0.1011, + "step": 1500 + }, + { + "epoch": 0.34152445961319683, + "grad_norm": 2.054708483083456, + "learning_rate": 1.235684398683402e-06, + "loss": 0.2029, + "step": 1501 + }, + { + "epoch": 0.3417519908987486, + "grad_norm": 3.1447213877676923, + "learning_rate": 1.2356653781767009e-06, + "loss": 0.1381, + "step": 1502 + }, + { + "epoch": 0.34197952218430033, + "grad_norm": 3.052797117025516, + "learning_rate": 1.2356463451891174e-06, + "loss": 0.1964, + "step": 1503 + }, + { + "epoch": 0.3422070534698521, + "grad_norm": 1.4103275005890912, + "learning_rate": 1.2356272997210414e-06, + "loss": 0.0826, + "step": 1504 + }, + { + "epoch": 0.3424345847554039, + "grad_norm": 1.5146376320654487, + "learning_rate": 1.2356082417728612e-06, + "loss": 0.1341, + "step": 1505 + }, + { + "epoch": 0.34266211604095564, + "grad_norm": 1.3688524329968725, + "learning_rate": 1.2355891713449672e-06, + "loss": 0.085, + "step": 1506 + }, + { + "epoch": 0.3428896473265074, + "grad_norm": 1.8416893881250955, + "learning_rate": 1.2355700884377485e-06, + "loss": 0.078, + "step": 1507 + }, + { + "epoch": 0.34311717861205915, + "grad_norm": 1.8673717383483042, + "learning_rate": 1.2355509930515958e-06, + "loss": 0.1891, + "step": 1508 + }, + { + "epoch": 0.3433447098976109, + "grad_norm": 1.6669190438483317, + "learning_rate": 1.2355318851868987e-06, + "loss": 0.1358, + "step": 1509 + }, + { + "epoch": 0.3435722411831627, + "grad_norm": 1.5281076865918781, + "learning_rate": 1.235512764844048e-06, + "loss": 0.0914, + "step": 1510 + }, + { + "epoch": 0.34379977246871446, + "grad_norm": 1.8126618025798658, + "learning_rate": 1.2354936320234345e-06, + "loss": 0.0873, + "step": 1511 + }, + { + "epoch": 0.3440273037542662, + "grad_norm": 1.5533427781258378, + "learning_rate": 1.2354744867254493e-06, + "loss": 0.1394, + "step": 1512 + }, + { + "epoch": 0.34425483503981796, + "grad_norm": 2.7412588250152417, + "learning_rate": 1.2354553289504836e-06, + "loss": 0.1263, + "step": 1513 + }, + { + "epoch": 0.3444823663253697, + "grad_norm": 1.9701757720367468, + "learning_rate": 1.2354361586989287e-06, + "loss": 0.1668, + "step": 1514 + }, + { + "epoch": 0.3447098976109215, + "grad_norm": 2.234561622535395, + "learning_rate": 1.235416975971177e-06, + "loss": 0.2347, + "step": 1515 + }, + { + "epoch": 0.3449374288964733, + "grad_norm": 1.640428050965421, + "learning_rate": 1.2353977807676205e-06, + "loss": 0.1271, + "step": 1516 + }, + { + "epoch": 0.345164960182025, + "grad_norm": 1.6546462677875515, + "learning_rate": 1.2353785730886506e-06, + "loss": 0.1331, + "step": 1517 + }, + { + "epoch": 0.3453924914675768, + "grad_norm": 2.5775712943835503, + "learning_rate": 1.235359352934661e-06, + "loss": 0.1179, + "step": 1518 + }, + { + "epoch": 0.34562002275312853, + "grad_norm": 1.7488081801237039, + "learning_rate": 1.235340120306044e-06, + "loss": 0.1999, + "step": 1519 + }, + { + "epoch": 0.34584755403868034, + "grad_norm": 2.313669007056231, + "learning_rate": 1.2353208752031925e-06, + "loss": 0.1494, + "step": 1520 + }, + { + "epoch": 0.3460750853242321, + "grad_norm": 2.9872932892021216, + "learning_rate": 1.2353016176265002e-06, + "loss": 0.1356, + "step": 1521 + }, + { + "epoch": 0.34630261660978384, + "grad_norm": 1.7333666870847044, + "learning_rate": 1.2352823475763603e-06, + "loss": 0.1156, + "step": 1522 + }, + { + "epoch": 0.3465301478953356, + "grad_norm": 2.1394646023986685, + "learning_rate": 1.2352630650531672e-06, + "loss": 0.1081, + "step": 1523 + }, + { + "epoch": 0.34675767918088735, + "grad_norm": 4.030786885073691, + "learning_rate": 1.2352437700573147e-06, + "loss": 0.1657, + "step": 1524 + }, + { + "epoch": 0.34698521046643915, + "grad_norm": 1.1100438837053874, + "learning_rate": 1.235224462589197e-06, + "loss": 0.0775, + "step": 1525 + }, + { + "epoch": 0.3472127417519909, + "grad_norm": 2.0720474848328188, + "learning_rate": 1.2352051426492089e-06, + "loss": 0.1879, + "step": 1526 + }, + { + "epoch": 0.34744027303754266, + "grad_norm": 2.7187132915415995, + "learning_rate": 1.2351858102377455e-06, + "loss": 0.1459, + "step": 1527 + }, + { + "epoch": 0.3476678043230944, + "grad_norm": 1.5373928030910593, + "learning_rate": 1.2351664653552012e-06, + "loss": 0.0697, + "step": 1528 + }, + { + "epoch": 0.34789533560864616, + "grad_norm": 14.12497611556846, + "learning_rate": 1.235147108001972e-06, + "loss": 0.0612, + "step": 1529 + }, + { + "epoch": 0.34812286689419797, + "grad_norm": 2.619961242647469, + "learning_rate": 1.2351277381784532e-06, + "loss": 0.1477, + "step": 1530 + }, + { + "epoch": 0.3483503981797497, + "grad_norm": 1.410493095940194, + "learning_rate": 1.235108355885041e-06, + "loss": 0.1716, + "step": 1531 + }, + { + "epoch": 0.3485779294653015, + "grad_norm": 1.8615160375782975, + "learning_rate": 1.2350889611221315e-06, + "loss": 0.1329, + "step": 1532 + }, + { + "epoch": 0.3488054607508532, + "grad_norm": 1.701233603216574, + "learning_rate": 1.2350695538901207e-06, + "loss": 0.1123, + "step": 1533 + }, + { + "epoch": 0.34903299203640503, + "grad_norm": 2.439441991830968, + "learning_rate": 1.2350501341894055e-06, + "loss": 0.1771, + "step": 1534 + }, + { + "epoch": 0.3492605233219568, + "grad_norm": 2.203771442439418, + "learning_rate": 1.235030702020383e-06, + "loss": 0.2083, + "step": 1535 + }, + { + "epoch": 0.34948805460750854, + "grad_norm": 1.69346918827836, + "learning_rate": 1.23501125738345e-06, + "loss": 0.1102, + "step": 1536 + }, + { + "epoch": 0.3497155858930603, + "grad_norm": 1.5090268058773, + "learning_rate": 1.2349918002790043e-06, + "loss": 0.1208, + "step": 1537 + }, + { + "epoch": 0.34994311717861204, + "grad_norm": 1.0946145352592496, + "learning_rate": 1.2349723307074432e-06, + "loss": 0.0996, + "step": 1538 + }, + { + "epoch": 0.35017064846416385, + "grad_norm": 2.301142408035351, + "learning_rate": 1.2349528486691648e-06, + "loss": 0.1756, + "step": 1539 + }, + { + "epoch": 0.3503981797497156, + "grad_norm": 2.4673785079867843, + "learning_rate": 1.2349333541645672e-06, + "loss": 0.1278, + "step": 1540 + }, + { + "epoch": 0.35062571103526735, + "grad_norm": 1.481886020980225, + "learning_rate": 1.2349138471940489e-06, + "loss": 0.1291, + "step": 1541 + }, + { + "epoch": 0.3508532423208191, + "grad_norm": 2.1610786496075063, + "learning_rate": 1.2348943277580086e-06, + "loss": 0.0983, + "step": 1542 + }, + { + "epoch": 0.35108077360637086, + "grad_norm": 1.3504864412021873, + "learning_rate": 1.2348747958568452e-06, + "loss": 0.0784, + "step": 1543 + }, + { + "epoch": 0.35130830489192266, + "grad_norm": 1.7504763814802757, + "learning_rate": 1.2348552514909579e-06, + "loss": 0.1448, + "step": 1544 + }, + { + "epoch": 0.3515358361774744, + "grad_norm": 2.9148856498776805, + "learning_rate": 1.2348356946607462e-06, + "loss": 0.1094, + "step": 1545 + }, + { + "epoch": 0.35176336746302617, + "grad_norm": 1.550887500201074, + "learning_rate": 1.2348161253666096e-06, + "loss": 0.0963, + "step": 1546 + }, + { + "epoch": 0.3519908987485779, + "grad_norm": 2.3639406464976824, + "learning_rate": 1.2347965436089484e-06, + "loss": 0.2153, + "step": 1547 + }, + { + "epoch": 0.35221843003412967, + "grad_norm": 4.095826967593627, + "learning_rate": 1.2347769493881625e-06, + "loss": 0.1153, + "step": 1548 + }, + { + "epoch": 0.3524459613196815, + "grad_norm": 1.277515077329962, + "learning_rate": 1.2347573427046527e-06, + "loss": 0.1465, + "step": 1549 + }, + { + "epoch": 0.35267349260523323, + "grad_norm": 1.9893288512267644, + "learning_rate": 1.2347377235588193e-06, + "loss": 0.267, + "step": 1550 + }, + { + "epoch": 0.352901023890785, + "grad_norm": 1.8718250436694113, + "learning_rate": 1.2347180919510637e-06, + "loss": 0.0973, + "step": 1551 + }, + { + "epoch": 0.35312855517633673, + "grad_norm": 3.36620771557597, + "learning_rate": 1.234698447881787e-06, + "loss": 0.1827, + "step": 1552 + }, + { + "epoch": 0.3533560864618885, + "grad_norm": 2.5974374464162717, + "learning_rate": 1.2346787913513904e-06, + "loss": 0.0969, + "step": 1553 + }, + { + "epoch": 0.3535836177474403, + "grad_norm": 2.360372812373607, + "learning_rate": 1.234659122360276e-06, + "loss": 0.1914, + "step": 1554 + }, + { + "epoch": 0.35381114903299204, + "grad_norm": 1.5248022456878274, + "learning_rate": 1.2346394409088457e-06, + "loss": 0.0785, + "step": 1555 + }, + { + "epoch": 0.3540386803185438, + "grad_norm": 1.2961890354523629, + "learning_rate": 1.2346197469975016e-06, + "loss": 0.1526, + "step": 1556 + }, + { + "epoch": 0.35426621160409555, + "grad_norm": 2.592957397339881, + "learning_rate": 1.2346000406266466e-06, + "loss": 0.1947, + "step": 1557 + }, + { + "epoch": 0.3544937428896473, + "grad_norm": 2.701062801999542, + "learning_rate": 1.2345803217966829e-06, + "loss": 0.1241, + "step": 1558 + }, + { + "epoch": 0.3547212741751991, + "grad_norm": 1.957155061679964, + "learning_rate": 1.2345605905080141e-06, + "loss": 0.0938, + "step": 1559 + }, + { + "epoch": 0.35494880546075086, + "grad_norm": 1.613669040940011, + "learning_rate": 1.234540846761043e-06, + "loss": 0.1728, + "step": 1560 + }, + { + "epoch": 0.3551763367463026, + "grad_norm": 3.0998396775844705, + "learning_rate": 1.2345210905561733e-06, + "loss": 0.1917, + "step": 1561 + }, + { + "epoch": 0.35540386803185436, + "grad_norm": 1.9882010530259928, + "learning_rate": 1.2345013218938089e-06, + "loss": 0.1309, + "step": 1562 + }, + { + "epoch": 0.3556313993174061, + "grad_norm": 1.5264967935473548, + "learning_rate": 1.2344815407743537e-06, + "loss": 0.0714, + "step": 1563 + }, + { + "epoch": 0.3558589306029579, + "grad_norm": 1.752943768112035, + "learning_rate": 1.2344617471982119e-06, + "loss": 0.1269, + "step": 1564 + }, + { + "epoch": 0.3560864618885097, + "grad_norm": 2.8753549426732405, + "learning_rate": 1.2344419411657883e-06, + "loss": 0.1418, + "step": 1565 + }, + { + "epoch": 0.3563139931740614, + "grad_norm": 2.8850586384790553, + "learning_rate": 1.2344221226774874e-06, + "loss": 0.1861, + "step": 1566 + }, + { + "epoch": 0.3565415244596132, + "grad_norm": 2.273061381543351, + "learning_rate": 1.2344022917337147e-06, + "loss": 0.1687, + "step": 1567 + }, + { + "epoch": 0.356769055745165, + "grad_norm": 1.6580131621970915, + "learning_rate": 1.234382448334875e-06, + "loss": 0.0952, + "step": 1568 + }, + { + "epoch": 0.35699658703071674, + "grad_norm": 1.453031205202229, + "learning_rate": 1.2343625924813741e-06, + "loss": 0.0823, + "step": 1569 + }, + { + "epoch": 0.3572241183162685, + "grad_norm": 2.9435349750610493, + "learning_rate": 1.234342724173618e-06, + "loss": 0.1503, + "step": 1570 + }, + { + "epoch": 0.35745164960182024, + "grad_norm": 1.6110988488812807, + "learning_rate": 1.2343228434120124e-06, + "loss": 0.1525, + "step": 1571 + }, + { + "epoch": 0.357679180887372, + "grad_norm": 1.6748300562704266, + "learning_rate": 1.2343029501969638e-06, + "loss": 0.1568, + "step": 1572 + }, + { + "epoch": 0.3579067121729238, + "grad_norm": 1.6603809620526928, + "learning_rate": 1.2342830445288788e-06, + "loss": 0.1728, + "step": 1573 + }, + { + "epoch": 0.35813424345847555, + "grad_norm": 1.1014107617300568, + "learning_rate": 1.2342631264081643e-06, + "loss": 0.0965, + "step": 1574 + }, + { + "epoch": 0.3583617747440273, + "grad_norm": 1.2732581681154054, + "learning_rate": 1.234243195835227e-06, + "loss": 0.1203, + "step": 1575 + }, + { + "epoch": 0.35858930602957906, + "grad_norm": 2.5995048914773053, + "learning_rate": 1.234223252810475e-06, + "loss": 0.1603, + "step": 1576 + }, + { + "epoch": 0.3588168373151308, + "grad_norm": 2.444792569525692, + "learning_rate": 1.2342032973343152e-06, + "loss": 0.149, + "step": 1577 + }, + { + "epoch": 0.3590443686006826, + "grad_norm": 3.7704137884261915, + "learning_rate": 1.2341833294071558e-06, + "loss": 0.2355, + "step": 1578 + }, + { + "epoch": 0.35927189988623437, + "grad_norm": 1.4763457287663684, + "learning_rate": 1.2341633490294046e-06, + "loss": 0.1272, + "step": 1579 + }, + { + "epoch": 0.3594994311717861, + "grad_norm": 2.020387401103254, + "learning_rate": 1.2341433562014705e-06, + "loss": 0.1376, + "step": 1580 + }, + { + "epoch": 0.3597269624573379, + "grad_norm": 2.0277864066708378, + "learning_rate": 1.2341233509237616e-06, + "loss": 0.1355, + "step": 1581 + }, + { + "epoch": 0.3599544937428896, + "grad_norm": 1.9258577073033267, + "learning_rate": 1.234103333196687e-06, + "loss": 0.1073, + "step": 1582 + }, + { + "epoch": 0.36018202502844143, + "grad_norm": 1.7051549966202966, + "learning_rate": 1.2340833030206558e-06, + "loss": 0.0829, + "step": 1583 + }, + { + "epoch": 0.3604095563139932, + "grad_norm": 3.206269578758429, + "learning_rate": 1.2340632603960774e-06, + "loss": 0.1355, + "step": 1584 + }, + { + "epoch": 0.36063708759954494, + "grad_norm": 2.1361631881752707, + "learning_rate": 1.2340432053233615e-06, + "loss": 0.0976, + "step": 1585 + }, + { + "epoch": 0.3608646188850967, + "grad_norm": 1.443174416772564, + "learning_rate": 1.2340231378029177e-06, + "loss": 0.1334, + "step": 1586 + }, + { + "epoch": 0.36109215017064844, + "grad_norm": 1.6382188653330987, + "learning_rate": 1.2340030578351564e-06, + "loss": 0.1066, + "step": 1587 + }, + { + "epoch": 0.36131968145620025, + "grad_norm": 2.198309549797981, + "learning_rate": 1.2339829654204878e-06, + "loss": 0.1208, + "step": 1588 + }, + { + "epoch": 0.361547212741752, + "grad_norm": 1.8122457718538585, + "learning_rate": 1.2339628605593229e-06, + "loss": 0.0885, + "step": 1589 + }, + { + "epoch": 0.36177474402730375, + "grad_norm": 2.132309804894969, + "learning_rate": 1.2339427432520722e-06, + "loss": 0.1064, + "step": 1590 + }, + { + "epoch": 0.3620022753128555, + "grad_norm": 2.3756453959322026, + "learning_rate": 1.2339226134991471e-06, + "loss": 0.2288, + "step": 1591 + }, + { + "epoch": 0.36222980659840726, + "grad_norm": 2.265871972514948, + "learning_rate": 1.2339024713009592e-06, + "loss": 0.1249, + "step": 1592 + }, + { + "epoch": 0.36245733788395906, + "grad_norm": 1.243536992034958, + "learning_rate": 1.2338823166579197e-06, + "loss": 0.1503, + "step": 1593 + }, + { + "epoch": 0.3626848691695108, + "grad_norm": 1.9724737870995313, + "learning_rate": 1.2338621495704409e-06, + "loss": 0.1493, + "step": 1594 + }, + { + "epoch": 0.36291240045506257, + "grad_norm": 2.065008978290742, + "learning_rate": 1.2338419700389349e-06, + "loss": 0.1542, + "step": 1595 + }, + { + "epoch": 0.3631399317406143, + "grad_norm": 2.958677711950205, + "learning_rate": 1.2338217780638137e-06, + "loss": 0.1661, + "step": 1596 + }, + { + "epoch": 0.36336746302616607, + "grad_norm": 1.7562621032856343, + "learning_rate": 1.2338015736454908e-06, + "loss": 0.2308, + "step": 1597 + }, + { + "epoch": 0.3635949943117179, + "grad_norm": 2.403965977937276, + "learning_rate": 1.2337813567843784e-06, + "loss": 0.1, + "step": 1598 + }, + { + "epoch": 0.36382252559726963, + "grad_norm": 2.1245624482048955, + "learning_rate": 1.2337611274808901e-06, + "loss": 0.1712, + "step": 1599 + }, + { + "epoch": 0.3640500568828214, + "grad_norm": 2.6993144740475326, + "learning_rate": 1.2337408857354394e-06, + "loss": 0.1376, + "step": 1600 + }, + { + "epoch": 0.36427758816837313, + "grad_norm": 1.8364645187073159, + "learning_rate": 1.2337206315484396e-06, + "loss": 0.1144, + "step": 1601 + }, + { + "epoch": 0.36450511945392494, + "grad_norm": 2.803014924439121, + "learning_rate": 1.2337003649203049e-06, + "loss": 0.2327, + "step": 1602 + }, + { + "epoch": 0.3647326507394767, + "grad_norm": 3.0921523919013425, + "learning_rate": 1.2336800858514498e-06, + "loss": 0.186, + "step": 1603 + }, + { + "epoch": 0.36496018202502845, + "grad_norm": 4.351581648162927, + "learning_rate": 1.2336597943422883e-06, + "loss": 0.1506, + "step": 1604 + }, + { + "epoch": 0.3651877133105802, + "grad_norm": 2.507983408971188, + "learning_rate": 1.2336394903932353e-06, + "loss": 0.1458, + "step": 1605 + }, + { + "epoch": 0.36541524459613195, + "grad_norm": 1.6768145183868413, + "learning_rate": 1.233619174004706e-06, + "loss": 0.137, + "step": 1606 + }, + { + "epoch": 0.36564277588168376, + "grad_norm": 1.324685955900903, + "learning_rate": 1.233598845177115e-06, + "loss": 0.1097, + "step": 1607 + }, + { + "epoch": 0.3658703071672355, + "grad_norm": 2.1338037745135443, + "learning_rate": 1.2335785039108787e-06, + "loss": 0.1784, + "step": 1608 + }, + { + "epoch": 0.36609783845278726, + "grad_norm": 2.3694737994247217, + "learning_rate": 1.233558150206412e-06, + "loss": 0.1991, + "step": 1609 + }, + { + "epoch": 0.366325369738339, + "grad_norm": 2.262254134148611, + "learning_rate": 1.2335377840641314e-06, + "loss": 0.1146, + "step": 1610 + }, + { + "epoch": 0.36655290102389076, + "grad_norm": 3.1588719561640985, + "learning_rate": 1.233517405484453e-06, + "loss": 0.1852, + "step": 1611 + }, + { + "epoch": 0.3667804323094426, + "grad_norm": 1.8006159718642254, + "learning_rate": 1.2334970144677929e-06, + "loss": 0.0956, + "step": 1612 + }, + { + "epoch": 0.3670079635949943, + "grad_norm": 2.563120316581933, + "learning_rate": 1.2334766110145684e-06, + "loss": 0.1088, + "step": 1613 + }, + { + "epoch": 0.3672354948805461, + "grad_norm": 2.235035261873256, + "learning_rate": 1.2334561951251967e-06, + "loss": 0.1607, + "step": 1614 + }, + { + "epoch": 0.36746302616609783, + "grad_norm": 2.3167227182441885, + "learning_rate": 1.2334357668000943e-06, + "loss": 0.1144, + "step": 1615 + }, + { + "epoch": 0.3676905574516496, + "grad_norm": 1.6955497804569648, + "learning_rate": 1.2334153260396795e-06, + "loss": 0.1868, + "step": 1616 + }, + { + "epoch": 0.3679180887372014, + "grad_norm": 3.4310526929941485, + "learning_rate": 1.2333948728443692e-06, + "loss": 0.173, + "step": 1617 + }, + { + "epoch": 0.36814562002275314, + "grad_norm": 1.5138266896493986, + "learning_rate": 1.2333744072145824e-06, + "loss": 0.0827, + "step": 1618 + }, + { + "epoch": 0.3683731513083049, + "grad_norm": 1.4784636330255214, + "learning_rate": 1.2333539291507365e-06, + "loss": 0.1569, + "step": 1619 + }, + { + "epoch": 0.36860068259385664, + "grad_norm": 2.704193421290089, + "learning_rate": 1.2333334386532507e-06, + "loss": 0.2538, + "step": 1620 + }, + { + "epoch": 0.3688282138794084, + "grad_norm": 1.742774697152776, + "learning_rate": 1.2333129357225434e-06, + "loss": 0.1055, + "step": 1621 + }, + { + "epoch": 0.3690557451649602, + "grad_norm": 2.9047485648418165, + "learning_rate": 1.2332924203590341e-06, + "loss": 0.1663, + "step": 1622 + }, + { + "epoch": 0.36928327645051195, + "grad_norm": 1.314984403604908, + "learning_rate": 1.2332718925631414e-06, + "loss": 0.1806, + "step": 1623 + }, + { + "epoch": 0.3695108077360637, + "grad_norm": 1.9510383304155832, + "learning_rate": 1.2332513523352853e-06, + "loss": 0.1242, + "step": 1624 + }, + { + "epoch": 0.36973833902161546, + "grad_norm": 2.559935844387003, + "learning_rate": 1.2332307996758854e-06, + "loss": 0.1478, + "step": 1625 + }, + { + "epoch": 0.3699658703071672, + "grad_norm": 2.88434227714402, + "learning_rate": 1.233210234585362e-06, + "loss": 0.2302, + "step": 1626 + }, + { + "epoch": 0.370193401592719, + "grad_norm": 2.1413595772930156, + "learning_rate": 1.2331896570641354e-06, + "loss": 0.1838, + "step": 1627 + }, + { + "epoch": 0.37042093287827077, + "grad_norm": 1.7922287054453576, + "learning_rate": 1.233169067112626e-06, + "loss": 0.0986, + "step": 1628 + }, + { + "epoch": 0.3706484641638225, + "grad_norm": 1.3323542239151287, + "learning_rate": 1.2331484647312545e-06, + "loss": 0.1167, + "step": 1629 + }, + { + "epoch": 0.3708759954493743, + "grad_norm": 2.46069963324228, + "learning_rate": 1.2331278499204423e-06, + "loss": 0.1553, + "step": 1630 + }, + { + "epoch": 0.371103526734926, + "grad_norm": 2.1291335035057104, + "learning_rate": 1.2331072226806107e-06, + "loss": 0.1213, + "step": 1631 + }, + { + "epoch": 0.37133105802047783, + "grad_norm": 2.2993804494788184, + "learning_rate": 1.233086583012181e-06, + "loss": 0.189, + "step": 1632 + }, + { + "epoch": 0.3715585893060296, + "grad_norm": 1.1991511815123632, + "learning_rate": 1.233065930915575e-06, + "loss": 0.1255, + "step": 1633 + }, + { + "epoch": 0.37178612059158134, + "grad_norm": 1.8590842904843217, + "learning_rate": 1.2330452663912155e-06, + "loss": 0.1271, + "step": 1634 + }, + { + "epoch": 0.3720136518771331, + "grad_norm": 2.373387407985538, + "learning_rate": 1.233024589439524e-06, + "loss": 0.2041, + "step": 1635 + }, + { + "epoch": 0.3722411831626849, + "grad_norm": 2.5782541251038333, + "learning_rate": 1.2330039000609233e-06, + "loss": 0.1329, + "step": 1636 + }, + { + "epoch": 0.37246871444823665, + "grad_norm": 2.671348916705849, + "learning_rate": 1.2329831982558365e-06, + "loss": 0.1618, + "step": 1637 + }, + { + "epoch": 0.3726962457337884, + "grad_norm": 3.5366418411008933, + "learning_rate": 1.2329624840246867e-06, + "loss": 0.1783, + "step": 1638 + }, + { + "epoch": 0.37292377701934015, + "grad_norm": 2.0065876352581062, + "learning_rate": 1.2329417573678974e-06, + "loss": 0.1156, + "step": 1639 + }, + { + "epoch": 0.3731513083048919, + "grad_norm": 1.975723329702582, + "learning_rate": 1.2329210182858915e-06, + "loss": 0.0976, + "step": 1640 + }, + { + "epoch": 0.3733788395904437, + "grad_norm": 1.7544914994439575, + "learning_rate": 1.2329002667790937e-06, + "loss": 0.1745, + "step": 1641 + }, + { + "epoch": 0.37360637087599546, + "grad_norm": 2.7660619782133193, + "learning_rate": 1.2328795028479275e-06, + "loss": 0.1375, + "step": 1642 + }, + { + "epoch": 0.3738339021615472, + "grad_norm": 1.9338461955860702, + "learning_rate": 1.2328587264928176e-06, + "loss": 0.1323, + "step": 1643 + }, + { + "epoch": 0.37406143344709897, + "grad_norm": 1.8716686944849417, + "learning_rate": 1.2328379377141885e-06, + "loss": 0.1937, + "step": 1644 + }, + { + "epoch": 0.3742889647326507, + "grad_norm": 2.1413388813058014, + "learning_rate": 1.2328171365124655e-06, + "loss": 0.1235, + "step": 1645 + }, + { + "epoch": 0.3745164960182025, + "grad_norm": 2.857793709939454, + "learning_rate": 1.2327963228880733e-06, + "loss": 0.1416, + "step": 1646 + }, + { + "epoch": 0.3747440273037543, + "grad_norm": 1.714384370948363, + "learning_rate": 1.2327754968414372e-06, + "loss": 0.1492, + "step": 1647 + }, + { + "epoch": 0.37497155858930603, + "grad_norm": 3.7781621231583724, + "learning_rate": 1.232754658372983e-06, + "loss": 0.1682, + "step": 1648 + }, + { + "epoch": 0.3751990898748578, + "grad_norm": 2.0029578055155244, + "learning_rate": 1.2327338074831366e-06, + "loss": 0.147, + "step": 1649 + }, + { + "epoch": 0.37542662116040953, + "grad_norm": 1.529791159285724, + "learning_rate": 1.2327129441723242e-06, + "loss": 0.1046, + "step": 1650 + }, + { + "epoch": 0.37565415244596134, + "grad_norm": 2.2299976733099482, + "learning_rate": 1.2326920684409724e-06, + "loss": 0.1617, + "step": 1651 + }, + { + "epoch": 0.3758816837315131, + "grad_norm": 3.398460731332957, + "learning_rate": 1.2326711802895077e-06, + "loss": 0.1554, + "step": 1652 + }, + { + "epoch": 0.37610921501706485, + "grad_norm": 2.130256184128101, + "learning_rate": 1.2326502797183568e-06, + "loss": 0.1392, + "step": 1653 + }, + { + "epoch": 0.3763367463026166, + "grad_norm": 1.741889659518372, + "learning_rate": 1.2326293667279472e-06, + "loss": 0.1305, + "step": 1654 + }, + { + "epoch": 0.37656427758816835, + "grad_norm": 2.3236871523378024, + "learning_rate": 1.232608441318706e-06, + "loss": 0.1208, + "step": 1655 + }, + { + "epoch": 0.37679180887372016, + "grad_norm": 1.9685344629470136, + "learning_rate": 1.232587503491061e-06, + "loss": 0.1238, + "step": 1656 + }, + { + "epoch": 0.3770193401592719, + "grad_norm": 3.0384258350922804, + "learning_rate": 1.2325665532454403e-06, + "loss": 0.205, + "step": 1657 + }, + { + "epoch": 0.37724687144482366, + "grad_norm": 1.997898726616711, + "learning_rate": 1.2325455905822719e-06, + "loss": 0.1299, + "step": 1658 + }, + { + "epoch": 0.3774744027303754, + "grad_norm": 1.5727587532720877, + "learning_rate": 1.2325246155019844e-06, + "loss": 0.0932, + "step": 1659 + }, + { + "epoch": 0.37770193401592717, + "grad_norm": 1.8064403636927258, + "learning_rate": 1.2325036280050063e-06, + "loss": 0.0926, + "step": 1660 + }, + { + "epoch": 0.377929465301479, + "grad_norm": 1.8737817001388601, + "learning_rate": 1.2324826280917664e-06, + "loss": 0.1008, + "step": 1661 + }, + { + "epoch": 0.3781569965870307, + "grad_norm": 2.459023941480425, + "learning_rate": 1.2324616157626943e-06, + "loss": 0.1157, + "step": 1662 + }, + { + "epoch": 0.3783845278725825, + "grad_norm": 2.3398340438241125, + "learning_rate": 1.2324405910182195e-06, + "loss": 0.2354, + "step": 1663 + }, + { + "epoch": 0.37861205915813423, + "grad_norm": 5.2055530727041175, + "learning_rate": 1.2324195538587713e-06, + "loss": 0.2025, + "step": 1664 + }, + { + "epoch": 0.378839590443686, + "grad_norm": 1.413281070506484, + "learning_rate": 1.23239850428478e-06, + "loss": 0.1108, + "step": 1665 + }, + { + "epoch": 0.3790671217292378, + "grad_norm": 1.7169043667269983, + "learning_rate": 1.2323774422966756e-06, + "loss": 0.0666, + "step": 1666 + }, + { + "epoch": 0.37929465301478954, + "grad_norm": 2.0458621353853834, + "learning_rate": 1.2323563678948885e-06, + "loss": 0.1231, + "step": 1667 + }, + { + "epoch": 0.3795221843003413, + "grad_norm": 1.5816448796619285, + "learning_rate": 1.2323352810798498e-06, + "loss": 0.0909, + "step": 1668 + }, + { + "epoch": 0.37974971558589304, + "grad_norm": 1.6282393884382793, + "learning_rate": 1.23231418185199e-06, + "loss": 0.1704, + "step": 1669 + }, + { + "epoch": 0.3799772468714448, + "grad_norm": 2.698679980786322, + "learning_rate": 1.2322930702117406e-06, + "loss": 0.1471, + "step": 1670 + }, + { + "epoch": 0.3802047781569966, + "grad_norm": 2.5205534849870093, + "learning_rate": 1.232271946159533e-06, + "loss": 0.1428, + "step": 1671 + }, + { + "epoch": 0.38043230944254836, + "grad_norm": 2.8663298664670402, + "learning_rate": 1.2322508096957992e-06, + "loss": 0.1582, + "step": 1672 + }, + { + "epoch": 0.3806598407281001, + "grad_norm": 3.606634062381101, + "learning_rate": 1.2322296608209709e-06, + "loss": 0.2957, + "step": 1673 + }, + { + "epoch": 0.38088737201365186, + "grad_norm": 3.3538078455291354, + "learning_rate": 1.2322084995354805e-06, + "loss": 0.1774, + "step": 1674 + }, + { + "epoch": 0.38111490329920367, + "grad_norm": 2.5478844933367513, + "learning_rate": 1.2321873258397602e-06, + "loss": 0.1554, + "step": 1675 + }, + { + "epoch": 0.3813424345847554, + "grad_norm": 1.540805779875073, + "learning_rate": 1.232166139734243e-06, + "loss": 0.0745, + "step": 1676 + }, + { + "epoch": 0.38156996587030717, + "grad_norm": 2.9923127366387146, + "learning_rate": 1.2321449412193622e-06, + "loss": 0.1788, + "step": 1677 + }, + { + "epoch": 0.3817974971558589, + "grad_norm": 2.6896661400495088, + "learning_rate": 1.2321237302955505e-06, + "loss": 0.1626, + "step": 1678 + }, + { + "epoch": 0.3820250284414107, + "grad_norm": 1.648292924059167, + "learning_rate": 1.2321025069632416e-06, + "loss": 0.0765, + "step": 1679 + }, + { + "epoch": 0.3822525597269625, + "grad_norm": 2.91462725493712, + "learning_rate": 1.2320812712228694e-06, + "loss": 0.2092, + "step": 1680 + }, + { + "epoch": 0.38248009101251423, + "grad_norm": 1.5588683602741686, + "learning_rate": 1.2320600230748677e-06, + "loss": 0.1276, + "step": 1681 + }, + { + "epoch": 0.382707622298066, + "grad_norm": 1.7369346142612618, + "learning_rate": 1.232038762519671e-06, + "loss": 0.0779, + "step": 1682 + }, + { + "epoch": 0.38293515358361774, + "grad_norm": 1.6511070866583553, + "learning_rate": 1.2320174895577138e-06, + "loss": 0.252, + "step": 1683 + }, + { + "epoch": 0.3831626848691695, + "grad_norm": 1.3968162606976187, + "learning_rate": 1.2319962041894307e-06, + "loss": 0.0834, + "step": 1684 + }, + { + "epoch": 0.3833902161547213, + "grad_norm": 2.653016956021873, + "learning_rate": 1.2319749064152569e-06, + "loss": 0.2423, + "step": 1685 + }, + { + "epoch": 0.38361774744027305, + "grad_norm": 3.176933366606272, + "learning_rate": 1.2319535962356277e-06, + "loss": 0.1773, + "step": 1686 + }, + { + "epoch": 0.3838452787258248, + "grad_norm": 1.800240710088086, + "learning_rate": 1.2319322736509784e-06, + "loss": 0.1566, + "step": 1687 + }, + { + "epoch": 0.38407281001137655, + "grad_norm": 2.2064113571623496, + "learning_rate": 1.2319109386617452e-06, + "loss": 0.1613, + "step": 1688 + }, + { + "epoch": 0.3843003412969283, + "grad_norm": 2.3087080662459387, + "learning_rate": 1.2318895912683638e-06, + "loss": 0.1545, + "step": 1689 + }, + { + "epoch": 0.3845278725824801, + "grad_norm": 1.837126601147182, + "learning_rate": 1.2318682314712706e-06, + "loss": 0.129, + "step": 1690 + }, + { + "epoch": 0.38475540386803186, + "grad_norm": 4.830481371425988, + "learning_rate": 1.2318468592709022e-06, + "loss": 0.266, + "step": 1691 + }, + { + "epoch": 0.3849829351535836, + "grad_norm": 2.441725289188966, + "learning_rate": 1.2318254746676954e-06, + "loss": 0.1554, + "step": 1692 + }, + { + "epoch": 0.38521046643913537, + "grad_norm": 1.6831246409088816, + "learning_rate": 1.2318040776620872e-06, + "loss": 0.1086, + "step": 1693 + }, + { + "epoch": 0.3854379977246871, + "grad_norm": 2.296389371722253, + "learning_rate": 1.231782668254515e-06, + "loss": 0.2497, + "step": 1694 + }, + { + "epoch": 0.3856655290102389, + "grad_norm": 2.8107479812520917, + "learning_rate": 1.2317612464454161e-06, + "loss": 0.1653, + "step": 1695 + }, + { + "epoch": 0.3858930602957907, + "grad_norm": 2.616629150374977, + "learning_rate": 1.2317398122352289e-06, + "loss": 0.1879, + "step": 1696 + }, + { + "epoch": 0.38612059158134243, + "grad_norm": 3.376176637253404, + "learning_rate": 1.2317183656243912e-06, + "loss": 0.3059, + "step": 1697 + }, + { + "epoch": 0.3863481228668942, + "grad_norm": 2.017884168409449, + "learning_rate": 1.231696906613341e-06, + "loss": 0.1138, + "step": 1698 + }, + { + "epoch": 0.38657565415244594, + "grad_norm": 2.383903076889702, + "learning_rate": 1.2316754352025173e-06, + "loss": 0.158, + "step": 1699 + }, + { + "epoch": 0.38680318543799774, + "grad_norm": 2.70005415107345, + "learning_rate": 1.2316539513923585e-06, + "loss": 0.1242, + "step": 1700 + }, + { + "epoch": 0.3870307167235495, + "grad_norm": 1.4233225917656729, + "learning_rate": 1.2316324551833042e-06, + "loss": 0.1382, + "step": 1701 + }, + { + "epoch": 0.38725824800910125, + "grad_norm": 1.3613270033001337, + "learning_rate": 1.2316109465757934e-06, + "loss": 0.0959, + "step": 1702 + }, + { + "epoch": 0.387485779294653, + "grad_norm": 1.7422130731580474, + "learning_rate": 1.231589425570266e-06, + "loss": 0.1694, + "step": 1703 + }, + { + "epoch": 0.38771331058020475, + "grad_norm": 1.1477314066351112, + "learning_rate": 1.2315678921671615e-06, + "loss": 0.1187, + "step": 1704 + }, + { + "epoch": 0.38794084186575656, + "grad_norm": 2.0773570604590708, + "learning_rate": 1.2315463463669202e-06, + "loss": 0.1594, + "step": 1705 + }, + { + "epoch": 0.3881683731513083, + "grad_norm": 1.3478575868276468, + "learning_rate": 1.2315247881699825e-06, + "loss": 0.0991, + "step": 1706 + }, + { + "epoch": 0.38839590443686006, + "grad_norm": 2.354162536662197, + "learning_rate": 1.2315032175767887e-06, + "loss": 0.1008, + "step": 1707 + }, + { + "epoch": 0.3886234357224118, + "grad_norm": 1.6538240421949093, + "learning_rate": 1.23148163458778e-06, + "loss": 0.1275, + "step": 1708 + }, + { + "epoch": 0.3888509670079636, + "grad_norm": 1.5311610006843637, + "learning_rate": 1.2314600392033974e-06, + "loss": 0.0968, + "step": 1709 + }, + { + "epoch": 0.3890784982935154, + "grad_norm": 2.4683314063286734, + "learning_rate": 1.2314384314240824e-06, + "loss": 0.1078, + "step": 1710 + }, + { + "epoch": 0.3893060295790671, + "grad_norm": 2.2142084705382636, + "learning_rate": 1.2314168112502765e-06, + "loss": 0.2684, + "step": 1711 + }, + { + "epoch": 0.3895335608646189, + "grad_norm": 1.2536342442255415, + "learning_rate": 1.2313951786824213e-06, + "loss": 0.1248, + "step": 1712 + }, + { + "epoch": 0.38976109215017063, + "grad_norm": 2.6941191297381293, + "learning_rate": 1.2313735337209593e-06, + "loss": 0.2081, + "step": 1713 + }, + { + "epoch": 0.38998862343572244, + "grad_norm": 2.30131500883838, + "learning_rate": 1.231351876366333e-06, + "loss": 0.1491, + "step": 1714 + }, + { + "epoch": 0.3902161547212742, + "grad_norm": 2.3841970341168204, + "learning_rate": 1.2313302066189846e-06, + "loss": 0.0996, + "step": 1715 + }, + { + "epoch": 0.39044368600682594, + "grad_norm": 2.6347474164019937, + "learning_rate": 1.2313085244793573e-06, + "loss": 0.1772, + "step": 1716 + }, + { + "epoch": 0.3906712172923777, + "grad_norm": 1.824344810986505, + "learning_rate": 1.2312868299478944e-06, + "loss": 0.0914, + "step": 1717 + }, + { + "epoch": 0.39089874857792944, + "grad_norm": 1.773023478059839, + "learning_rate": 1.2312651230250387e-06, + "loss": 0.182, + "step": 1718 + }, + { + "epoch": 0.39112627986348125, + "grad_norm": 2.558986871552195, + "learning_rate": 1.2312434037112345e-06, + "loss": 0.1102, + "step": 1719 + }, + { + "epoch": 0.391353811149033, + "grad_norm": 1.547999236334684, + "learning_rate": 1.2312216720069251e-06, + "loss": 0.1488, + "step": 1720 + }, + { + "epoch": 0.39158134243458476, + "grad_norm": 2.8418021765667247, + "learning_rate": 1.2311999279125552e-06, + "loss": 0.1658, + "step": 1721 + }, + { + "epoch": 0.3918088737201365, + "grad_norm": 5.673024628891561, + "learning_rate": 1.2311781714285689e-06, + "loss": 0.221, + "step": 1722 + }, + { + "epoch": 0.39203640500568826, + "grad_norm": 2.3720745639494494, + "learning_rate": 1.231156402555411e-06, + "loss": 0.1788, + "step": 1723 + }, + { + "epoch": 0.39226393629124007, + "grad_norm": 1.7411798788297947, + "learning_rate": 1.2311346212935262e-06, + "loss": 0.1318, + "step": 1724 + }, + { + "epoch": 0.3924914675767918, + "grad_norm": 1.8671484665662437, + "learning_rate": 1.23111282764336e-06, + "loss": 0.145, + "step": 1725 + }, + { + "epoch": 0.39271899886234357, + "grad_norm": 1.257862162167775, + "learning_rate": 1.2310910216053576e-06, + "loss": 0.1059, + "step": 1726 + }, + { + "epoch": 0.3929465301478953, + "grad_norm": 2.2794733420992275, + "learning_rate": 1.2310692031799646e-06, + "loss": 0.1938, + "step": 1727 + }, + { + "epoch": 0.3931740614334471, + "grad_norm": 2.1119176270114313, + "learning_rate": 1.2310473723676272e-06, + "loss": 0.1157, + "step": 1728 + }, + { + "epoch": 0.3934015927189989, + "grad_norm": 1.3881889344482166, + "learning_rate": 1.2310255291687913e-06, + "loss": 0.0832, + "step": 1729 + }, + { + "epoch": 0.39362912400455063, + "grad_norm": 2.023136430703965, + "learning_rate": 1.2310036735839037e-06, + "loss": 0.1242, + "step": 1730 + }, + { + "epoch": 0.3938566552901024, + "grad_norm": 2.624591745407245, + "learning_rate": 1.2309818056134108e-06, + "loss": 0.1507, + "step": 1731 + }, + { + "epoch": 0.39408418657565414, + "grad_norm": 1.2844255816526005, + "learning_rate": 1.2309599252577593e-06, + "loss": 0.074, + "step": 1732 + }, + { + "epoch": 0.3943117178612059, + "grad_norm": 2.43035532680509, + "learning_rate": 1.230938032517397e-06, + "loss": 0.2302, + "step": 1733 + }, + { + "epoch": 0.3945392491467577, + "grad_norm": 2.4967953244354817, + "learning_rate": 1.2309161273927708e-06, + "loss": 0.122, + "step": 1734 + }, + { + "epoch": 0.39476678043230945, + "grad_norm": 2.8201501794880643, + "learning_rate": 1.2308942098843289e-06, + "loss": 0.1314, + "step": 1735 + }, + { + "epoch": 0.3949943117178612, + "grad_norm": 1.377416323825032, + "learning_rate": 1.2308722799925188e-06, + "loss": 0.1068, + "step": 1736 + }, + { + "epoch": 0.39522184300341295, + "grad_norm": 3.157882549207335, + "learning_rate": 1.2308503377177887e-06, + "loss": 0.1845, + "step": 1737 + }, + { + "epoch": 0.3954493742889647, + "grad_norm": 2.8098606380126943, + "learning_rate": 1.2308283830605877e-06, + "loss": 0.1387, + "step": 1738 + }, + { + "epoch": 0.3956769055745165, + "grad_norm": 1.773927388786271, + "learning_rate": 1.230806416021364e-06, + "loss": 0.059, + "step": 1739 + }, + { + "epoch": 0.39590443686006827, + "grad_norm": 2.440949931918227, + "learning_rate": 1.2307844366005665e-06, + "loss": 0.1398, + "step": 1740 + }, + { + "epoch": 0.39613196814562, + "grad_norm": 1.5057479299724394, + "learning_rate": 1.2307624447986446e-06, + "loss": 0.1438, + "step": 1741 + }, + { + "epoch": 0.39635949943117177, + "grad_norm": 1.7016670626096497, + "learning_rate": 1.2307404406160476e-06, + "loss": 0.1971, + "step": 1742 + }, + { + "epoch": 0.3965870307167236, + "grad_norm": 2.1741262002450465, + "learning_rate": 1.2307184240532255e-06, + "loss": 0.1, + "step": 1743 + }, + { + "epoch": 0.39681456200227533, + "grad_norm": 1.6480951134265682, + "learning_rate": 1.2306963951106283e-06, + "loss": 0.144, + "step": 1744 + }, + { + "epoch": 0.3970420932878271, + "grad_norm": 1.4849711465690703, + "learning_rate": 1.2306743537887058e-06, + "loss": 0.2173, + "step": 1745 + }, + { + "epoch": 0.39726962457337883, + "grad_norm": 1.750478097654071, + "learning_rate": 1.2306523000879086e-06, + "loss": 0.0965, + "step": 1746 + }, + { + "epoch": 0.3974971558589306, + "grad_norm": 1.3984444468696073, + "learning_rate": 1.230630234008688e-06, + "loss": 0.1522, + "step": 1747 + }, + { + "epoch": 0.3977246871444824, + "grad_norm": 1.6859723238809896, + "learning_rate": 1.2306081555514942e-06, + "loss": 0.0826, + "step": 1748 + }, + { + "epoch": 0.39795221843003414, + "grad_norm": 2.580780832144414, + "learning_rate": 1.2305860647167792e-06, + "loss": 0.1193, + "step": 1749 + }, + { + "epoch": 0.3981797497155859, + "grad_norm": 2.4353217528542532, + "learning_rate": 1.2305639615049938e-06, + "loss": 0.1157, + "step": 1750 + }, + { + "epoch": 0.39840728100113765, + "grad_norm": 4.900966801210391, + "learning_rate": 1.2305418459165902e-06, + "loss": 0.1967, + "step": 1751 + }, + { + "epoch": 0.3986348122866894, + "grad_norm": 2.340473825012059, + "learning_rate": 1.2305197179520203e-06, + "loss": 0.1513, + "step": 1752 + }, + { + "epoch": 0.3988623435722412, + "grad_norm": 1.6987905011562776, + "learning_rate": 1.2304975776117362e-06, + "loss": 0.1526, + "step": 1753 + }, + { + "epoch": 0.39908987485779296, + "grad_norm": 1.6154688960203794, + "learning_rate": 1.2304754248961906e-06, + "loss": 0.1199, + "step": 1754 + }, + { + "epoch": 0.3993174061433447, + "grad_norm": 1.8354123927057464, + "learning_rate": 1.2304532598058363e-06, + "loss": 0.1382, + "step": 1755 + }, + { + "epoch": 0.39954493742889646, + "grad_norm": 2.2202210353934757, + "learning_rate": 1.230431082341126e-06, + "loss": 0.1486, + "step": 1756 + }, + { + "epoch": 0.3997724687144482, + "grad_norm": 2.1595657089748337, + "learning_rate": 1.2304088925025133e-06, + "loss": 0.1111, + "step": 1757 + }, + { + "epoch": 0.4, + "grad_norm": 2.051418292936696, + "learning_rate": 1.2303866902904515e-06, + "loss": 0.1125, + "step": 1758 + }, + { + "epoch": 0.4002275312855518, + "grad_norm": 1.4683812530262192, + "learning_rate": 1.2303644757053945e-06, + "loss": 0.121, + "step": 1759 + }, + { + "epoch": 0.4004550625711035, + "grad_norm": 2.9222870011841544, + "learning_rate": 1.2303422487477965e-06, + "loss": 0.1909, + "step": 1760 + }, + { + "epoch": 0.4006825938566553, + "grad_norm": 2.4890537658855485, + "learning_rate": 1.230320009418111e-06, + "loss": 0.1728, + "step": 1761 + }, + { + "epoch": 0.40091012514220703, + "grad_norm": 1.6296292199347002, + "learning_rate": 1.2302977577167937e-06, + "loss": 0.1238, + "step": 1762 + }, + { + "epoch": 0.40113765642775884, + "grad_norm": 1.9686403685728797, + "learning_rate": 1.2302754936442986e-06, + "loss": 0.1157, + "step": 1763 + }, + { + "epoch": 0.4013651877133106, + "grad_norm": 3.3666708406816315, + "learning_rate": 1.2302532172010809e-06, + "loss": 0.1081, + "step": 1764 + }, + { + "epoch": 0.40159271899886234, + "grad_norm": 1.2014023232311797, + "learning_rate": 1.2302309283875958e-06, + "loss": 0.0927, + "step": 1765 + }, + { + "epoch": 0.4018202502844141, + "grad_norm": 1.896842215979294, + "learning_rate": 1.230208627204299e-06, + "loss": 0.1408, + "step": 1766 + }, + { + "epoch": 0.40204778156996585, + "grad_norm": 2.6032820946847584, + "learning_rate": 1.2301863136516463e-06, + "loss": 0.2446, + "step": 1767 + }, + { + "epoch": 0.40227531285551765, + "grad_norm": 2.6105518283568214, + "learning_rate": 1.2301639877300937e-06, + "loss": 0.1439, + "step": 1768 + }, + { + "epoch": 0.4025028441410694, + "grad_norm": 1.3240539738643666, + "learning_rate": 1.2301416494400974e-06, + "loss": 0.106, + "step": 1769 + }, + { + "epoch": 0.40273037542662116, + "grad_norm": 3.0980518217773265, + "learning_rate": 1.2301192987821142e-06, + "loss": 0.1403, + "step": 1770 + }, + { + "epoch": 0.4029579067121729, + "grad_norm": 3.422689409153713, + "learning_rate": 1.2300969357566008e-06, + "loss": 0.1786, + "step": 1771 + }, + { + "epoch": 0.40318543799772466, + "grad_norm": 2.4443605233075725, + "learning_rate": 1.230074560364014e-06, + "loss": 0.1532, + "step": 1772 + }, + { + "epoch": 0.40341296928327647, + "grad_norm": 1.1122594786359905, + "learning_rate": 1.2300521726048114e-06, + "loss": 0.1026, + "step": 1773 + }, + { + "epoch": 0.4036405005688282, + "grad_norm": 2.4538948369426996, + "learning_rate": 1.2300297724794506e-06, + "loss": 0.1436, + "step": 1774 + }, + { + "epoch": 0.40386803185437997, + "grad_norm": 1.1549269839884009, + "learning_rate": 1.2300073599883892e-06, + "loss": 0.1081, + "step": 1775 + }, + { + "epoch": 0.4040955631399317, + "grad_norm": 2.8038987971188822, + "learning_rate": 1.2299849351320854e-06, + "loss": 0.1752, + "step": 1776 + }, + { + "epoch": 0.40432309442548353, + "grad_norm": 1.7076470891369002, + "learning_rate": 1.2299624979109976e-06, + "loss": 0.1199, + "step": 1777 + }, + { + "epoch": 0.4045506257110353, + "grad_norm": 1.2055567774564526, + "learning_rate": 1.229940048325584e-06, + "loss": 0.1149, + "step": 1778 + }, + { + "epoch": 0.40477815699658704, + "grad_norm": 3.58239703526112, + "learning_rate": 1.229917586376304e-06, + "loss": 0.1486, + "step": 1779 + }, + { + "epoch": 0.4050056882821388, + "grad_norm": 1.4724200301684793, + "learning_rate": 1.2298951120636163e-06, + "loss": 0.1487, + "step": 1780 + }, + { + "epoch": 0.40523321956769054, + "grad_norm": 2.180162524813868, + "learning_rate": 1.2298726253879802e-06, + "loss": 0.1407, + "step": 1781 + }, + { + "epoch": 0.40546075085324235, + "grad_norm": 2.5371024976546424, + "learning_rate": 1.2298501263498557e-06, + "loss": 0.2582, + "step": 1782 + }, + { + "epoch": 0.4056882821387941, + "grad_norm": 2.3758196386707136, + "learning_rate": 1.229827614949702e-06, + "loss": 0.153, + "step": 1783 + }, + { + "epoch": 0.40591581342434585, + "grad_norm": 1.7394821341749753, + "learning_rate": 1.22980509118798e-06, + "loss": 0.1211, + "step": 1784 + }, + { + "epoch": 0.4061433447098976, + "grad_norm": 1.8213502745824992, + "learning_rate": 1.2297825550651491e-06, + "loss": 0.1351, + "step": 1785 + }, + { + "epoch": 0.40637087599544935, + "grad_norm": 1.5366804604898765, + "learning_rate": 1.2297600065816707e-06, + "loss": 0.172, + "step": 1786 + }, + { + "epoch": 0.40659840728100116, + "grad_norm": 2.867711224850475, + "learning_rate": 1.229737445738005e-06, + "loss": 0.1413, + "step": 1787 + }, + { + "epoch": 0.4068259385665529, + "grad_norm": 2.8801832410299717, + "learning_rate": 1.2297148725346137e-06, + "loss": 0.1068, + "step": 1788 + }, + { + "epoch": 0.40705346985210467, + "grad_norm": 2.26302057191021, + "learning_rate": 1.229692286971958e-06, + "loss": 0.2012, + "step": 1789 + }, + { + "epoch": 0.4072810011376564, + "grad_norm": 2.025110259831313, + "learning_rate": 1.2296696890504992e-06, + "loss": 0.0936, + "step": 1790 + }, + { + "epoch": 0.40750853242320817, + "grad_norm": 2.038765984004919, + "learning_rate": 1.2296470787706993e-06, + "loss": 0.1893, + "step": 1791 + }, + { + "epoch": 0.40773606370876, + "grad_norm": 1.2472833998526156, + "learning_rate": 1.2296244561330206e-06, + "loss": 0.1077, + "step": 1792 + }, + { + "epoch": 0.40796359499431173, + "grad_norm": 1.652590726736059, + "learning_rate": 1.2296018211379253e-06, + "loss": 0.1657, + "step": 1793 + }, + { + "epoch": 0.4081911262798635, + "grad_norm": 1.9370986933072707, + "learning_rate": 1.229579173785876e-06, + "loss": 0.1483, + "step": 1794 + }, + { + "epoch": 0.40841865756541523, + "grad_norm": 2.1166679197631946, + "learning_rate": 1.2295565140773357e-06, + "loss": 0.097, + "step": 1795 + }, + { + "epoch": 0.408646188850967, + "grad_norm": 1.6183560555534648, + "learning_rate": 1.2295338420127673e-06, + "loss": 0.1116, + "step": 1796 + }, + { + "epoch": 0.4088737201365188, + "grad_norm": 4.024689412752317, + "learning_rate": 1.2295111575926344e-06, + "loss": 0.1566, + "step": 1797 + }, + { + "epoch": 0.40910125142207054, + "grad_norm": 1.645924937894827, + "learning_rate": 1.2294884608174007e-06, + "loss": 0.0893, + "step": 1798 + }, + { + "epoch": 0.4093287827076223, + "grad_norm": 2.1920700631018435, + "learning_rate": 1.2294657516875297e-06, + "loss": 0.1111, + "step": 1799 + }, + { + "epoch": 0.40955631399317405, + "grad_norm": 2.1814995815983105, + "learning_rate": 1.229443030203486e-06, + "loss": 0.0974, + "step": 1800 + }, + { + "epoch": 0.4097838452787258, + "grad_norm": 3.4034833350789553, + "learning_rate": 1.2294202963657335e-06, + "loss": 0.1251, + "step": 1801 + }, + { + "epoch": 0.4100113765642776, + "grad_norm": 2.072295418124302, + "learning_rate": 1.2293975501747372e-06, + "loss": 0.1566, + "step": 1802 + }, + { + "epoch": 0.41023890784982936, + "grad_norm": 3.023299474624079, + "learning_rate": 1.229374791630962e-06, + "loss": 0.1615, + "step": 1803 + }, + { + "epoch": 0.4104664391353811, + "grad_norm": 2.96052428349603, + "learning_rate": 1.2293520207348727e-06, + "loss": 0.1246, + "step": 1804 + }, + { + "epoch": 0.41069397042093286, + "grad_norm": 2.3232889792011404, + "learning_rate": 1.229329237486935e-06, + "loss": 0.119, + "step": 1805 + }, + { + "epoch": 0.4109215017064846, + "grad_norm": 1.680210318484648, + "learning_rate": 1.2293064418876145e-06, + "loss": 0.0759, + "step": 1806 + }, + { + "epoch": 0.4111490329920364, + "grad_norm": 2.2603194082450946, + "learning_rate": 1.2292836339373771e-06, + "loss": 0.1255, + "step": 1807 + }, + { + "epoch": 0.4113765642775882, + "grad_norm": 2.9474726641163396, + "learning_rate": 1.2292608136366887e-06, + "loss": 0.1411, + "step": 1808 + }, + { + "epoch": 0.4116040955631399, + "grad_norm": 2.7451292492288313, + "learning_rate": 1.2292379809860162e-06, + "loss": 0.1366, + "step": 1809 + }, + { + "epoch": 0.4118316268486917, + "grad_norm": 1.7748176376725044, + "learning_rate": 1.2292151359858258e-06, + "loss": 0.0838, + "step": 1810 + }, + { + "epoch": 0.4120591581342435, + "grad_norm": 1.9600147202758083, + "learning_rate": 1.2291922786365846e-06, + "loss": 0.0611, + "step": 1811 + }, + { + "epoch": 0.41228668941979524, + "grad_norm": 1.9430142401566255, + "learning_rate": 1.2291694089387599e-06, + "loss": 0.1047, + "step": 1812 + }, + { + "epoch": 0.412514220705347, + "grad_norm": 2.10075635319286, + "learning_rate": 1.2291465268928187e-06, + "loss": 0.1982, + "step": 1813 + }, + { + "epoch": 0.41274175199089874, + "grad_norm": 3.157474099750552, + "learning_rate": 1.2291236324992291e-06, + "loss": 0.2055, + "step": 1814 + }, + { + "epoch": 0.4129692832764505, + "grad_norm": 2.0954118584949657, + "learning_rate": 1.229100725758459e-06, + "loss": 0.1466, + "step": 1815 + }, + { + "epoch": 0.4131968145620023, + "grad_norm": 1.8718484159029354, + "learning_rate": 1.2290778066709763e-06, + "loss": 0.0949, + "step": 1816 + }, + { + "epoch": 0.41342434584755405, + "grad_norm": 2.2325202956764625, + "learning_rate": 1.2290548752372494e-06, + "loss": 0.1135, + "step": 1817 + }, + { + "epoch": 0.4136518771331058, + "grad_norm": 1.942418916883084, + "learning_rate": 1.2290319314577473e-06, + "loss": 0.153, + "step": 1818 + }, + { + "epoch": 0.41387940841865756, + "grad_norm": 1.252464692558786, + "learning_rate": 1.2290089753329386e-06, + "loss": 0.0687, + "step": 1819 + }, + { + "epoch": 0.4141069397042093, + "grad_norm": 2.285517486522719, + "learning_rate": 1.2289860068632929e-06, + "loss": 0.1432, + "step": 1820 + }, + { + "epoch": 0.4143344709897611, + "grad_norm": 2.752384746112051, + "learning_rate": 1.228963026049279e-06, + "loss": 0.1178, + "step": 1821 + }, + { + "epoch": 0.41456200227531287, + "grad_norm": 2.4041920784703175, + "learning_rate": 1.228940032891367e-06, + "loss": 0.1721, + "step": 1822 + }, + { + "epoch": 0.4147895335608646, + "grad_norm": 1.8795669463811113, + "learning_rate": 1.2289170273900272e-06, + "loss": 0.0916, + "step": 1823 + }, + { + "epoch": 0.4150170648464164, + "grad_norm": 1.1852055506021697, + "learning_rate": 1.228894009545729e-06, + "loss": 0.1252, + "step": 1824 + }, + { + "epoch": 0.4152445961319681, + "grad_norm": 1.9135382951198285, + "learning_rate": 1.2288709793589434e-06, + "loss": 0.2363, + "step": 1825 + }, + { + "epoch": 0.41547212741751993, + "grad_norm": 5.3507052059831555, + "learning_rate": 1.2288479368301408e-06, + "loss": 0.1991, + "step": 1826 + }, + { + "epoch": 0.4156996587030717, + "grad_norm": 2.0538462354836255, + "learning_rate": 1.2288248819597922e-06, + "loss": 0.1979, + "step": 1827 + }, + { + "epoch": 0.41592718998862344, + "grad_norm": 1.5733199096098156, + "learning_rate": 1.228801814748369e-06, + "loss": 0.0897, + "step": 1828 + }, + { + "epoch": 0.4161547212741752, + "grad_norm": 2.1146593845270916, + "learning_rate": 1.2287787351963427e-06, + "loss": 0.1143, + "step": 1829 + }, + { + "epoch": 0.41638225255972694, + "grad_norm": 1.9141333315182336, + "learning_rate": 1.2287556433041845e-06, + "loss": 0.1447, + "step": 1830 + }, + { + "epoch": 0.41660978384527875, + "grad_norm": 2.665142936614018, + "learning_rate": 1.2287325390723669e-06, + "loss": 0.1329, + "step": 1831 + }, + { + "epoch": 0.4168373151308305, + "grad_norm": 2.19183346029768, + "learning_rate": 1.2287094225013618e-06, + "loss": 0.2281, + "step": 1832 + }, + { + "epoch": 0.41706484641638225, + "grad_norm": 2.127380551764798, + "learning_rate": 1.2286862935916416e-06, + "loss": 0.1784, + "step": 1833 + }, + { + "epoch": 0.417292377701934, + "grad_norm": 2.3095877045434, + "learning_rate": 1.2286631523436793e-06, + "loss": 0.1159, + "step": 1834 + }, + { + "epoch": 0.41751990898748575, + "grad_norm": 3.9175539200968386, + "learning_rate": 1.2286399987579478e-06, + "loss": 0.2352, + "step": 1835 + }, + { + "epoch": 0.41774744027303756, + "grad_norm": 1.5984970371973604, + "learning_rate": 1.2286168328349202e-06, + "loss": 0.106, + "step": 1836 + }, + { + "epoch": 0.4179749715585893, + "grad_norm": 1.3035202220513378, + "learning_rate": 1.2285936545750698e-06, + "loss": 0.09, + "step": 1837 + }, + { + "epoch": 0.41820250284414107, + "grad_norm": 2.7776624038109707, + "learning_rate": 1.2285704639788707e-06, + "loss": 0.0956, + "step": 1838 + }, + { + "epoch": 0.4184300341296928, + "grad_norm": 2.176123621315213, + "learning_rate": 1.2285472610467969e-06, + "loss": 0.0822, + "step": 1839 + }, + { + "epoch": 0.41865756541524457, + "grad_norm": 3.1080618782037064, + "learning_rate": 1.228524045779322e-06, + "loss": 0.2253, + "step": 1840 + }, + { + "epoch": 0.4188850967007964, + "grad_norm": 3.107268409079957, + "learning_rate": 1.2285008181769212e-06, + "loss": 0.1601, + "step": 1841 + }, + { + "epoch": 0.41911262798634813, + "grad_norm": 2.166037873871064, + "learning_rate": 1.228477578240069e-06, + "loss": 0.1026, + "step": 1842 + }, + { + "epoch": 0.4193401592718999, + "grad_norm": 2.4389639267330607, + "learning_rate": 1.22845432596924e-06, + "loss": 0.1287, + "step": 1843 + }, + { + "epoch": 0.41956769055745163, + "grad_norm": 2.5484705835141583, + "learning_rate": 1.22843106136491e-06, + "loss": 0.1095, + "step": 1844 + }, + { + "epoch": 0.4197952218430034, + "grad_norm": 2.908823301090787, + "learning_rate": 1.2284077844275543e-06, + "loss": 0.1573, + "step": 1845 + }, + { + "epoch": 0.4200227531285552, + "grad_norm": 3.0404493249523386, + "learning_rate": 1.2283844951576484e-06, + "loss": 0.1426, + "step": 1846 + }, + { + "epoch": 0.42025028441410694, + "grad_norm": 2.187557277461029, + "learning_rate": 1.2283611935556686e-06, + "loss": 0.1289, + "step": 1847 + }, + { + "epoch": 0.4204778156996587, + "grad_norm": 2.0438695218752154, + "learning_rate": 1.2283378796220909e-06, + "loss": 0.1464, + "step": 1848 + }, + { + "epoch": 0.42070534698521045, + "grad_norm": 2.973625989974048, + "learning_rate": 1.228314553357392e-06, + "loss": 0.1576, + "step": 1849 + }, + { + "epoch": 0.42093287827076226, + "grad_norm": 2.065578760595499, + "learning_rate": 1.2282912147620483e-06, + "loss": 0.1228, + "step": 1850 + }, + { + "epoch": 0.421160409556314, + "grad_norm": 1.087351915862162, + "learning_rate": 1.2282678638365373e-06, + "loss": 0.1229, + "step": 1851 + }, + { + "epoch": 0.42138794084186576, + "grad_norm": 2.683413990345536, + "learning_rate": 1.2282445005813359e-06, + "loss": 0.175, + "step": 1852 + }, + { + "epoch": 0.4216154721274175, + "grad_norm": 2.6502454120286743, + "learning_rate": 1.2282211249969217e-06, + "loss": 0.1437, + "step": 1853 + }, + { + "epoch": 0.42184300341296926, + "grad_norm": 2.546457693607218, + "learning_rate": 1.2281977370837725e-06, + "loss": 0.1637, + "step": 1854 + }, + { + "epoch": 0.42207053469852107, + "grad_norm": 1.7164185473669251, + "learning_rate": 1.2281743368423662e-06, + "loss": 0.0716, + "step": 1855 + }, + { + "epoch": 0.4222980659840728, + "grad_norm": 1.1259934133581069, + "learning_rate": 1.2281509242731813e-06, + "loss": 0.1674, + "step": 1856 + }, + { + "epoch": 0.4225255972696246, + "grad_norm": 1.824612068932172, + "learning_rate": 1.228127499376696e-06, + "loss": 0.1473, + "step": 1857 + }, + { + "epoch": 0.4227531285551763, + "grad_norm": 3.052950993125296, + "learning_rate": 1.228104062153389e-06, + "loss": 0.145, + "step": 1858 + }, + { + "epoch": 0.4229806598407281, + "grad_norm": 0.890302498925592, + "learning_rate": 1.2280806126037396e-06, + "loss": 0.0978, + "step": 1859 + }, + { + "epoch": 0.4232081911262799, + "grad_norm": 1.936861534299759, + "learning_rate": 1.2280571507282272e-06, + "loss": 0.1774, + "step": 1860 + }, + { + "epoch": 0.42343572241183164, + "grad_norm": 1.7891865799019433, + "learning_rate": 1.2280336765273309e-06, + "loss": 0.1108, + "step": 1861 + }, + { + "epoch": 0.4236632536973834, + "grad_norm": 1.8865955624918218, + "learning_rate": 1.2280101900015306e-06, + "loss": 0.1417, + "step": 1862 + }, + { + "epoch": 0.42389078498293514, + "grad_norm": 1.8128448248353213, + "learning_rate": 1.2279866911513064e-06, + "loss": 0.1428, + "step": 1863 + }, + { + "epoch": 0.4241183162684869, + "grad_norm": 2.901810032716033, + "learning_rate": 1.2279631799771386e-06, + "loss": 0.1846, + "step": 1864 + }, + { + "epoch": 0.4243458475540387, + "grad_norm": 2.262057679433472, + "learning_rate": 1.2279396564795077e-06, + "loss": 0.2503, + "step": 1865 + }, + { + "epoch": 0.42457337883959045, + "grad_norm": 3.0660607522891485, + "learning_rate": 1.2279161206588944e-06, + "loss": 0.1034, + "step": 1866 + }, + { + "epoch": 0.4248009101251422, + "grad_norm": 2.3354669669388852, + "learning_rate": 1.2278925725157798e-06, + "loss": 0.1624, + "step": 1867 + }, + { + "epoch": 0.42502844141069396, + "grad_norm": 1.5871965476689007, + "learning_rate": 1.2278690120506451e-06, + "loss": 0.1101, + "step": 1868 + }, + { + "epoch": 0.4252559726962457, + "grad_norm": 2.709885114592366, + "learning_rate": 1.2278454392639722e-06, + "loss": 0.1347, + "step": 1869 + }, + { + "epoch": 0.4254835039817975, + "grad_norm": 1.413512267705387, + "learning_rate": 1.2278218541562422e-06, + "loss": 0.0734, + "step": 1870 + }, + { + "epoch": 0.42571103526734927, + "grad_norm": 1.7321041587522092, + "learning_rate": 1.2277982567279377e-06, + "loss": 0.1911, + "step": 1871 + }, + { + "epoch": 0.425938566552901, + "grad_norm": 2.066073629245216, + "learning_rate": 1.2277746469795407e-06, + "loss": 0.1626, + "step": 1872 + }, + { + "epoch": 0.4261660978384528, + "grad_norm": 2.0535165144225056, + "learning_rate": 1.227751024911534e-06, + "loss": 0.0855, + "step": 1873 + }, + { + "epoch": 0.4263936291240045, + "grad_norm": 3.005758093068664, + "learning_rate": 1.2277273905244002e-06, + "loss": 0.1643, + "step": 1874 + }, + { + "epoch": 0.42662116040955633, + "grad_norm": 2.3459498044051137, + "learning_rate": 1.2277037438186224e-06, + "loss": 0.1279, + "step": 1875 + }, + { + "epoch": 0.4268486916951081, + "grad_norm": 1.5882472504054521, + "learning_rate": 1.2276800847946839e-06, + "loss": 0.0926, + "step": 1876 + }, + { + "epoch": 0.42707622298065984, + "grad_norm": 1.7112545971646451, + "learning_rate": 1.227656413453068e-06, + "loss": 0.1147, + "step": 1877 + }, + { + "epoch": 0.4273037542662116, + "grad_norm": 2.282517318806134, + "learning_rate": 1.227632729794259e-06, + "loss": 0.1405, + "step": 1878 + }, + { + "epoch": 0.42753128555176334, + "grad_norm": 2.782117068218183, + "learning_rate": 1.2276090338187403e-06, + "loss": 0.2318, + "step": 1879 + }, + { + "epoch": 0.42775881683731515, + "grad_norm": 5.886268196659077, + "learning_rate": 1.2275853255269967e-06, + "loss": 0.1079, + "step": 1880 + }, + { + "epoch": 0.4279863481228669, + "grad_norm": 2.3326199322014394, + "learning_rate": 1.2275616049195129e-06, + "loss": 0.1448, + "step": 1881 + }, + { + "epoch": 0.42821387940841865, + "grad_norm": 1.8816717252042165, + "learning_rate": 1.2275378719967733e-06, + "loss": 0.0942, + "step": 1882 + }, + { + "epoch": 0.4284414106939704, + "grad_norm": 1.7985569079604016, + "learning_rate": 1.227514126759263e-06, + "loss": 0.112, + "step": 1883 + }, + { + "epoch": 0.4286689419795222, + "grad_norm": 3.0259876193386646, + "learning_rate": 1.2274903692074674e-06, + "loss": 0.1872, + "step": 1884 + }, + { + "epoch": 0.42889647326507396, + "grad_norm": 3.1916760297795417, + "learning_rate": 1.2274665993418722e-06, + "loss": 0.1792, + "step": 1885 + }, + { + "epoch": 0.4291240045506257, + "grad_norm": 2.5906641441781613, + "learning_rate": 1.227442817162963e-06, + "loss": 0.1644, + "step": 1886 + }, + { + "epoch": 0.42935153583617747, + "grad_norm": 3.430163291900366, + "learning_rate": 1.227419022671226e-06, + "loss": 0.1339, + "step": 1887 + }, + { + "epoch": 0.4295790671217292, + "grad_norm": 3.041561020625739, + "learning_rate": 1.2273952158671472e-06, + "loss": 0.1736, + "step": 1888 + }, + { + "epoch": 0.429806598407281, + "grad_norm": 3.0955843388329347, + "learning_rate": 1.2273713967512137e-06, + "loss": 0.1665, + "step": 1889 + }, + { + "epoch": 0.4300341296928328, + "grad_norm": 6.520120337578789, + "learning_rate": 1.227347565323912e-06, + "loss": 0.1274, + "step": 1890 + }, + { + "epoch": 0.43026166097838453, + "grad_norm": 3.436953118763558, + "learning_rate": 1.2273237215857293e-06, + "loss": 0.1299, + "step": 1891 + }, + { + "epoch": 0.4304891922639363, + "grad_norm": 3.4857387853802364, + "learning_rate": 1.2272998655371526e-06, + "loss": 0.1726, + "step": 1892 + }, + { + "epoch": 0.43071672354948803, + "grad_norm": 3.6540817320958228, + "learning_rate": 1.22727599717867e-06, + "loss": 0.1921, + "step": 1893 + }, + { + "epoch": 0.43094425483503984, + "grad_norm": 1.8546702242131714, + "learning_rate": 1.2272521165107687e-06, + "loss": 0.1954, + "step": 1894 + }, + { + "epoch": 0.4311717861205916, + "grad_norm": 2.906647877087954, + "learning_rate": 1.2272282235339372e-06, + "loss": 0.1436, + "step": 1895 + }, + { + "epoch": 0.43139931740614335, + "grad_norm": 2.2434613722644863, + "learning_rate": 1.2272043182486638e-06, + "loss": 0.1086, + "step": 1896 + }, + { + "epoch": 0.4316268486916951, + "grad_norm": 3.00916232200625, + "learning_rate": 1.227180400655437e-06, + "loss": 0.1753, + "step": 1897 + }, + { + "epoch": 0.43185437997724685, + "grad_norm": 2.818766317391872, + "learning_rate": 1.2271564707547457e-06, + "loss": 0.3096, + "step": 1898 + }, + { + "epoch": 0.43208191126279866, + "grad_norm": 1.1542786147828004, + "learning_rate": 1.227132528547079e-06, + "loss": 0.1049, + "step": 1899 + }, + { + "epoch": 0.4323094425483504, + "grad_norm": 2.4796477207464207, + "learning_rate": 1.2271085740329261e-06, + "loss": 0.2142, + "step": 1900 + }, + { + "epoch": 0.43253697383390216, + "grad_norm": 1.9623610484322256, + "learning_rate": 1.2270846072127764e-06, + "loss": 0.1611, + "step": 1901 + }, + { + "epoch": 0.4327645051194539, + "grad_norm": 2.554393698647313, + "learning_rate": 1.2270606280871205e-06, + "loss": 0.0902, + "step": 1902 + }, + { + "epoch": 0.43299203640500566, + "grad_norm": 2.0537100903512364, + "learning_rate": 1.2270366366564476e-06, + "loss": 0.1265, + "step": 1903 + }, + { + "epoch": 0.43321956769055747, + "grad_norm": 1.9963013061766377, + "learning_rate": 1.2270126329212486e-06, + "loss": 0.1739, + "step": 1904 + }, + { + "epoch": 0.4334470989761092, + "grad_norm": 1.1842645267471326, + "learning_rate": 1.2269886168820138e-06, + "loss": 0.0839, + "step": 1905 + }, + { + "epoch": 0.433674630261661, + "grad_norm": 1.7356609118289905, + "learning_rate": 1.2269645885392342e-06, + "loss": 0.1932, + "step": 1906 + }, + { + "epoch": 0.43390216154721273, + "grad_norm": 2.7617703436331835, + "learning_rate": 1.226940547893401e-06, + "loss": 0.1339, + "step": 1907 + }, + { + "epoch": 0.4341296928327645, + "grad_norm": 1.5346172373445122, + "learning_rate": 1.2269164949450052e-06, + "loss": 0.0955, + "step": 1908 + }, + { + "epoch": 0.4343572241183163, + "grad_norm": 3.5636615948411, + "learning_rate": 1.2268924296945387e-06, + "loss": 0.1829, + "step": 1909 + }, + { + "epoch": 0.43458475540386804, + "grad_norm": 1.67295814200305, + "learning_rate": 1.2268683521424932e-06, + "loss": 0.0815, + "step": 1910 + }, + { + "epoch": 0.4348122866894198, + "grad_norm": 2.6841445059138054, + "learning_rate": 1.226844262289361e-06, + "loss": 0.1003, + "step": 1911 + }, + { + "epoch": 0.43503981797497154, + "grad_norm": 2.8366861529632206, + "learning_rate": 1.2268201601356342e-06, + "loss": 0.1634, + "step": 1912 + }, + { + "epoch": 0.4352673492605233, + "grad_norm": 1.6905197116385013, + "learning_rate": 1.2267960456818054e-06, + "loss": 0.0796, + "step": 1913 + }, + { + "epoch": 0.4354948805460751, + "grad_norm": 3.505560593587699, + "learning_rate": 1.2267719189283676e-06, + "loss": 0.1547, + "step": 1914 + }, + { + "epoch": 0.43572241183162685, + "grad_norm": 8.776691522016812, + "learning_rate": 1.2267477798758141e-06, + "loss": 0.151, + "step": 1915 + }, + { + "epoch": 0.4359499431171786, + "grad_norm": 2.3591467713222807, + "learning_rate": 1.2267236285246376e-06, + "loss": 0.1631, + "step": 1916 + }, + { + "epoch": 0.43617747440273036, + "grad_norm": 3.0422098264181887, + "learning_rate": 1.2266994648753325e-06, + "loss": 0.1776, + "step": 1917 + }, + { + "epoch": 0.43640500568828217, + "grad_norm": 1.537478908875297, + "learning_rate": 1.2266752889283923e-06, + "loss": 0.1669, + "step": 1918 + }, + { + "epoch": 0.4366325369738339, + "grad_norm": 1.8827801159749011, + "learning_rate": 1.226651100684311e-06, + "loss": 0.1161, + "step": 1919 + }, + { + "epoch": 0.43686006825938567, + "grad_norm": 3.9329321893042124, + "learning_rate": 1.2266269001435829e-06, + "loss": 0.1344, + "step": 1920 + }, + { + "epoch": 0.4370875995449374, + "grad_norm": 3.1389052206999666, + "learning_rate": 1.226602687306703e-06, + "loss": 0.1233, + "step": 1921 + }, + { + "epoch": 0.4373151308304892, + "grad_norm": 2.998637701201584, + "learning_rate": 1.226578462174166e-06, + "loss": 0.1134, + "step": 1922 + }, + { + "epoch": 0.437542662116041, + "grad_norm": 1.8514157151948887, + "learning_rate": 1.2265542247464668e-06, + "loss": 0.1484, + "step": 1923 + }, + { + "epoch": 0.43777019340159273, + "grad_norm": 2.304538146026998, + "learning_rate": 1.226529975024101e-06, + "loss": 0.2034, + "step": 1924 + }, + { + "epoch": 0.4379977246871445, + "grad_norm": 1.744388776164245, + "learning_rate": 1.2265057130075641e-06, + "loss": 0.1854, + "step": 1925 + }, + { + "epoch": 0.43822525597269624, + "grad_norm": 2.6565426243412165, + "learning_rate": 1.2264814386973523e-06, + "loss": 0.1572, + "step": 1926 + }, + { + "epoch": 0.438452787258248, + "grad_norm": 6.9733970699782954, + "learning_rate": 1.2264571520939612e-06, + "loss": 0.0721, + "step": 1927 + }, + { + "epoch": 0.4386803185437998, + "grad_norm": 1.9165403272554284, + "learning_rate": 1.2264328531978875e-06, + "loss": 0.1644, + "step": 1928 + }, + { + "epoch": 0.43890784982935155, + "grad_norm": 2.125664734049219, + "learning_rate": 1.2264085420096277e-06, + "loss": 0.1061, + "step": 1929 + }, + { + "epoch": 0.4391353811149033, + "grad_norm": 1.905504057373527, + "learning_rate": 1.226384218529679e-06, + "loss": 0.153, + "step": 1930 + }, + { + "epoch": 0.43936291240045505, + "grad_norm": 1.6818760874849463, + "learning_rate": 1.2263598827585379e-06, + "loss": 0.1771, + "step": 1931 + }, + { + "epoch": 0.4395904436860068, + "grad_norm": 0.7483156859971863, + "learning_rate": 1.2263355346967023e-06, + "loss": 0.0657, + "step": 1932 + }, + { + "epoch": 0.4398179749715586, + "grad_norm": 2.8922141342339915, + "learning_rate": 1.2263111743446697e-06, + "loss": 0.1127, + "step": 1933 + }, + { + "epoch": 0.44004550625711036, + "grad_norm": 2.254842475988348, + "learning_rate": 1.2262868017029377e-06, + "loss": 0.1213, + "step": 1934 + }, + { + "epoch": 0.4402730375426621, + "grad_norm": 2.5250433582131477, + "learning_rate": 1.226262416772005e-06, + "loss": 0.1912, + "step": 1935 + }, + { + "epoch": 0.44050056882821387, + "grad_norm": 1.7654279959465413, + "learning_rate": 1.2262380195523696e-06, + "loss": 0.0965, + "step": 1936 + }, + { + "epoch": 0.4407281001137656, + "grad_norm": 1.9288183994078796, + "learning_rate": 1.2262136100445303e-06, + "loss": 0.1442, + "step": 1937 + }, + { + "epoch": 0.4409556313993174, + "grad_norm": 3.208052375402599, + "learning_rate": 1.2261891882489855e-06, + "loss": 0.1476, + "step": 1938 + }, + { + "epoch": 0.4411831626848692, + "grad_norm": 1.9111076610363271, + "learning_rate": 1.226164754166235e-06, + "loss": 0.1038, + "step": 1939 + }, + { + "epoch": 0.44141069397042093, + "grad_norm": 2.755324988033307, + "learning_rate": 1.2261403077967778e-06, + "loss": 0.1473, + "step": 1940 + }, + { + "epoch": 0.4416382252559727, + "grad_norm": 9.71923868520851, + "learning_rate": 1.2261158491411136e-06, + "loss": 0.124, + "step": 1941 + }, + { + "epoch": 0.44186575654152443, + "grad_norm": 2.7390582242923887, + "learning_rate": 1.2260913781997425e-06, + "loss": 0.1482, + "step": 1942 + }, + { + "epoch": 0.44209328782707624, + "grad_norm": 2.2870752287666964, + "learning_rate": 1.2260668949731644e-06, + "loss": 0.1578, + "step": 1943 + }, + { + "epoch": 0.442320819112628, + "grad_norm": 3.234690008586509, + "learning_rate": 1.2260423994618798e-06, + "loss": 0.1443, + "step": 1944 + }, + { + "epoch": 0.44254835039817975, + "grad_norm": 3.0788870765568865, + "learning_rate": 1.2260178916663892e-06, + "loss": 0.1467, + "step": 1945 + }, + { + "epoch": 0.4427758816837315, + "grad_norm": 1.1800230173128983, + "learning_rate": 1.2259933715871935e-06, + "loss": 0.1334, + "step": 1946 + }, + { + "epoch": 0.44300341296928325, + "grad_norm": 1.9016957352422574, + "learning_rate": 1.2259688392247942e-06, + "loss": 0.0917, + "step": 1947 + }, + { + "epoch": 0.44323094425483506, + "grad_norm": 3.0210466868569505, + "learning_rate": 1.2259442945796926e-06, + "loss": 0.2328, + "step": 1948 + }, + { + "epoch": 0.4434584755403868, + "grad_norm": 1.5788241191031314, + "learning_rate": 1.2259197376523898e-06, + "loss": 0.1697, + "step": 1949 + }, + { + "epoch": 0.44368600682593856, + "grad_norm": 1.998070514722676, + "learning_rate": 1.2258951684433883e-06, + "loss": 0.1603, + "step": 1950 + }, + { + "epoch": 0.4439135381114903, + "grad_norm": 2.6869498999380674, + "learning_rate": 1.2258705869531901e-06, + "loss": 0.1949, + "step": 1951 + }, + { + "epoch": 0.4441410693970421, + "grad_norm": 2.1632784019049325, + "learning_rate": 1.2258459931822974e-06, + "loss": 0.205, + "step": 1952 + }, + { + "epoch": 0.4443686006825939, + "grad_norm": 2.5366287820014897, + "learning_rate": 1.225821387131213e-06, + "loss": 0.0819, + "step": 1953 + }, + { + "epoch": 0.4445961319681456, + "grad_norm": 1.069933102487221, + "learning_rate": 1.22579676880044e-06, + "loss": 0.0658, + "step": 1954 + }, + { + "epoch": 0.4448236632536974, + "grad_norm": 2.159006669286408, + "learning_rate": 1.2257721381904811e-06, + "loss": 0.1409, + "step": 1955 + }, + { + "epoch": 0.44505119453924913, + "grad_norm": 2.4876680890845644, + "learning_rate": 1.22574749530184e-06, + "loss": 0.1376, + "step": 1956 + }, + { + "epoch": 0.44527872582480094, + "grad_norm": 2.824828278975413, + "learning_rate": 1.2257228401350205e-06, + "loss": 0.1554, + "step": 1957 + }, + { + "epoch": 0.4455062571103527, + "grad_norm": 1.021893819715831, + "learning_rate": 1.2256981726905262e-06, + "loss": 0.0644, + "step": 1958 + }, + { + "epoch": 0.44573378839590444, + "grad_norm": 2.3996775721551034, + "learning_rate": 1.2256734929688612e-06, + "loss": 0.1624, + "step": 1959 + }, + { + "epoch": 0.4459613196814562, + "grad_norm": 1.6628903100272334, + "learning_rate": 1.2256488009705303e-06, + "loss": 0.0861, + "step": 1960 + }, + { + "epoch": 0.44618885096700794, + "grad_norm": 1.7833579074549226, + "learning_rate": 1.225624096696038e-06, + "loss": 0.0887, + "step": 1961 + }, + { + "epoch": 0.44641638225255975, + "grad_norm": 0.9965946886914974, + "learning_rate": 1.225599380145889e-06, + "loss": 0.0664, + "step": 1962 + }, + { + "epoch": 0.4466439135381115, + "grad_norm": 3.226735520877417, + "learning_rate": 1.2255746513205889e-06, + "loss": 0.1774, + "step": 1963 + }, + { + "epoch": 0.44687144482366326, + "grad_norm": 1.7796640096245704, + "learning_rate": 1.2255499102206423e-06, + "loss": 0.0894, + "step": 1964 + }, + { + "epoch": 0.447098976109215, + "grad_norm": 4.62578287316141, + "learning_rate": 1.2255251568465558e-06, + "loss": 0.179, + "step": 1965 + }, + { + "epoch": 0.44732650739476676, + "grad_norm": 1.4148303981446813, + "learning_rate": 1.2255003911988348e-06, + "loss": 0.1264, + "step": 1966 + }, + { + "epoch": 0.44755403868031857, + "grad_norm": 2.0366633980384785, + "learning_rate": 1.2254756132779855e-06, + "loss": 0.1721, + "step": 1967 + }, + { + "epoch": 0.4477815699658703, + "grad_norm": 1.571525400347696, + "learning_rate": 1.2254508230845144e-06, + "loss": 0.1231, + "step": 1968 + }, + { + "epoch": 0.44800910125142207, + "grad_norm": 1.8759159261783616, + "learning_rate": 1.2254260206189283e-06, + "loss": 0.1027, + "step": 1969 + }, + { + "epoch": 0.4482366325369738, + "grad_norm": 1.3834676958640604, + "learning_rate": 1.2254012058817337e-06, + "loss": 0.1492, + "step": 1970 + }, + { + "epoch": 0.4484641638225256, + "grad_norm": 2.6467245289779733, + "learning_rate": 1.2253763788734384e-06, + "loss": 0.1474, + "step": 1971 + }, + { + "epoch": 0.4486916951080774, + "grad_norm": 1.5990498535799054, + "learning_rate": 1.2253515395945492e-06, + "loss": 0.1472, + "step": 1972 + }, + { + "epoch": 0.44891922639362913, + "grad_norm": 1.437136755746043, + "learning_rate": 1.2253266880455742e-06, + "loss": 0.1196, + "step": 1973 + }, + { + "epoch": 0.4491467576791809, + "grad_norm": 1.2239358879138267, + "learning_rate": 1.225301824227021e-06, + "loss": 0.0703, + "step": 1974 + }, + { + "epoch": 0.44937428896473264, + "grad_norm": 1.6035752105793946, + "learning_rate": 1.2252769481393979e-06, + "loss": 0.0868, + "step": 1975 + }, + { + "epoch": 0.4496018202502844, + "grad_norm": 3.2371327667206096, + "learning_rate": 1.2252520597832132e-06, + "loss": 0.1704, + "step": 1976 + }, + { + "epoch": 0.4498293515358362, + "grad_norm": 1.6184063611102666, + "learning_rate": 1.2252271591589759e-06, + "loss": 0.1344, + "step": 1977 + }, + { + "epoch": 0.45005688282138795, + "grad_norm": 2.7294883577821833, + "learning_rate": 1.2252022462671947e-06, + "loss": 0.1336, + "step": 1978 + }, + { + "epoch": 0.4502844141069397, + "grad_norm": 1.5187257522965836, + "learning_rate": 1.2251773211083789e-06, + "loss": 0.1412, + "step": 1979 + }, + { + "epoch": 0.45051194539249145, + "grad_norm": 1.3575191114107326, + "learning_rate": 1.225152383683038e-06, + "loss": 0.1915, + "step": 1980 + }, + { + "epoch": 0.4507394766780432, + "grad_norm": 3.573777818440175, + "learning_rate": 1.225127433991681e-06, + "loss": 0.2149, + "step": 1981 + }, + { + "epoch": 0.450967007963595, + "grad_norm": 1.9064246989112013, + "learning_rate": 1.2251024720348186e-06, + "loss": 0.131, + "step": 1982 + }, + { + "epoch": 0.45119453924914676, + "grad_norm": 2.60894949438164, + "learning_rate": 1.2250774978129606e-06, + "loss": 0.1191, + "step": 1983 + }, + { + "epoch": 0.4514220705346985, + "grad_norm": 2.0893541262163655, + "learning_rate": 1.2250525113266175e-06, + "loss": 0.1246, + "step": 1984 + }, + { + "epoch": 0.45164960182025027, + "grad_norm": 3.058414234549828, + "learning_rate": 1.2250275125763002e-06, + "loss": 0.1562, + "step": 1985 + }, + { + "epoch": 0.451877133105802, + "grad_norm": 2.7361733850127288, + "learning_rate": 1.2250025015625194e-06, + "loss": 0.1386, + "step": 1986 + }, + { + "epoch": 0.4521046643913538, + "grad_norm": 3.551478539562137, + "learning_rate": 1.2249774782857863e-06, + "loss": 0.1592, + "step": 1987 + }, + { + "epoch": 0.4523321956769056, + "grad_norm": 1.7962854205539769, + "learning_rate": 1.2249524427466123e-06, + "loss": 0.0993, + "step": 1988 + }, + { + "epoch": 0.45255972696245733, + "grad_norm": 1.7250148293655851, + "learning_rate": 1.2249273949455092e-06, + "loss": 0.1935, + "step": 1989 + }, + { + "epoch": 0.4527872582480091, + "grad_norm": 2.798595683002779, + "learning_rate": 1.224902334882989e-06, + "loss": 0.1092, + "step": 1990 + }, + { + "epoch": 0.4530147895335609, + "grad_norm": 1.5382254747774728, + "learning_rate": 1.2248772625595633e-06, + "loss": 0.1337, + "step": 1991 + }, + { + "epoch": 0.45324232081911264, + "grad_norm": 2.0009706394613227, + "learning_rate": 1.2248521779757455e-06, + "loss": 0.1426, + "step": 1992 + }, + { + "epoch": 0.4534698521046644, + "grad_norm": 1.9370488492395648, + "learning_rate": 1.2248270811320473e-06, + "loss": 0.119, + "step": 1993 + }, + { + "epoch": 0.45369738339021615, + "grad_norm": 2.5525945012477345, + "learning_rate": 1.2248019720289826e-06, + "loss": 0.1389, + "step": 1994 + }, + { + "epoch": 0.4539249146757679, + "grad_norm": 1.218140356587049, + "learning_rate": 1.2247768506670639e-06, + "loss": 0.1386, + "step": 1995 + }, + { + "epoch": 0.4541524459613197, + "grad_norm": 2.258231427971832, + "learning_rate": 1.2247517170468045e-06, + "loss": 0.2193, + "step": 1996 + }, + { + "epoch": 0.45437997724687146, + "grad_norm": 2.1071468220391996, + "learning_rate": 1.2247265711687186e-06, + "loss": 0.1245, + "step": 1997 + }, + { + "epoch": 0.4546075085324232, + "grad_norm": 1.873057521166509, + "learning_rate": 1.2247014130333198e-06, + "loss": 0.1437, + "step": 1998 + }, + { + "epoch": 0.45483503981797496, + "grad_norm": 1.9807431832192035, + "learning_rate": 1.2246762426411227e-06, + "loss": 0.1464, + "step": 1999 + }, + { + "epoch": 0.4550625711035267, + "grad_norm": 1.6280778924472747, + "learning_rate": 1.2246510599926412e-06, + "loss": 0.121, + "step": 2000 + }, + { + "epoch": 0.4552901023890785, + "grad_norm": 2.8593046291182826, + "learning_rate": 1.2246258650883904e-06, + "loss": 0.1397, + "step": 2001 + }, + { + "epoch": 0.4555176336746303, + "grad_norm": 2.1546576597681906, + "learning_rate": 1.224600657928885e-06, + "loss": 0.1574, + "step": 2002 + }, + { + "epoch": 0.455745164960182, + "grad_norm": 2.023369886602771, + "learning_rate": 1.2245754385146402e-06, + "loss": 0.1274, + "step": 2003 + }, + { + "epoch": 0.4559726962457338, + "grad_norm": 2.1812153004323, + "learning_rate": 1.2245502068461718e-06, + "loss": 0.0905, + "step": 2004 + }, + { + "epoch": 0.45620022753128553, + "grad_norm": 2.3877035049857342, + "learning_rate": 1.224524962923995e-06, + "loss": 0.1017, + "step": 2005 + }, + { + "epoch": 0.45642775881683734, + "grad_norm": 1.5777250547989092, + "learning_rate": 1.2244997067486258e-06, + "loss": 0.2139, + "step": 2006 + }, + { + "epoch": 0.4566552901023891, + "grad_norm": 2.9772883997052997, + "learning_rate": 1.2244744383205807e-06, + "loss": 0.1471, + "step": 2007 + }, + { + "epoch": 0.45688282138794084, + "grad_norm": 3.180677452250655, + "learning_rate": 1.224449157640376e-06, + "loss": 0.1183, + "step": 2008 + }, + { + "epoch": 0.4571103526734926, + "grad_norm": 2.3851861401264145, + "learning_rate": 1.2244238647085283e-06, + "loss": 0.1519, + "step": 2009 + }, + { + "epoch": 0.45733788395904434, + "grad_norm": 3.1889875080379615, + "learning_rate": 1.2243985595255546e-06, + "loss": 0.1488, + "step": 2010 + }, + { + "epoch": 0.45756541524459615, + "grad_norm": 3.3499971452048825, + "learning_rate": 1.224373242091972e-06, + "loss": 0.1913, + "step": 2011 + }, + { + "epoch": 0.4577929465301479, + "grad_norm": 2.114007694687429, + "learning_rate": 1.224347912408298e-06, + "loss": 0.1205, + "step": 2012 + }, + { + "epoch": 0.45802047781569966, + "grad_norm": 3.372165092943733, + "learning_rate": 1.2243225704750506e-06, + "loss": 0.1729, + "step": 2013 + }, + { + "epoch": 0.4582480091012514, + "grad_norm": 2.7776776213730234, + "learning_rate": 1.2242972162927472e-06, + "loss": 0.1789, + "step": 2014 + }, + { + "epoch": 0.45847554038680316, + "grad_norm": 1.5378497020029431, + "learning_rate": 1.2242718498619063e-06, + "loss": 0.1046, + "step": 2015 + }, + { + "epoch": 0.45870307167235497, + "grad_norm": 1.9245485247734166, + "learning_rate": 1.2242464711830463e-06, + "loss": 0.1296, + "step": 2016 + }, + { + "epoch": 0.4589306029579067, + "grad_norm": 3.571578005575834, + "learning_rate": 1.224221080256686e-06, + "loss": 0.1493, + "step": 2017 + }, + { + "epoch": 0.45915813424345847, + "grad_norm": 3.0687747713314506, + "learning_rate": 1.2241956770833444e-06, + "loss": 0.1298, + "step": 2018 + }, + { + "epoch": 0.4593856655290102, + "grad_norm": 1.4963685838255636, + "learning_rate": 1.2241702616635403e-06, + "loss": 0.0692, + "step": 2019 + }, + { + "epoch": 0.459613196814562, + "grad_norm": 3.5237300833320346, + "learning_rate": 1.2241448339977931e-06, + "loss": 0.213, + "step": 2020 + }, + { + "epoch": 0.4598407281001138, + "grad_norm": 1.5853111616974174, + "learning_rate": 1.2241193940866234e-06, + "loss": 0.0529, + "step": 2021 + }, + { + "epoch": 0.46006825938566553, + "grad_norm": 1.5166819743159363, + "learning_rate": 1.22409394193055e-06, + "loss": 0.1126, + "step": 2022 + }, + { + "epoch": 0.4602957906712173, + "grad_norm": 2.7441332092646924, + "learning_rate": 1.224068477530094e-06, + "loss": 0.1397, + "step": 2023 + }, + { + "epoch": 0.46052332195676904, + "grad_norm": 2.324643380132358, + "learning_rate": 1.2240430008857751e-06, + "loss": 0.2038, + "step": 2024 + }, + { + "epoch": 0.46075085324232085, + "grad_norm": 1.0420451784612332, + "learning_rate": 1.2240175119981146e-06, + "loss": 0.1546, + "step": 2025 + }, + { + "epoch": 0.4609783845278726, + "grad_norm": 2.3221839473002692, + "learning_rate": 1.223992010867633e-06, + "loss": 0.1177, + "step": 2026 + }, + { + "epoch": 0.46120591581342435, + "grad_norm": 0.7809955408321576, + "learning_rate": 1.223966497494852e-06, + "loss": 0.0408, + "step": 2027 + }, + { + "epoch": 0.4614334470989761, + "grad_norm": 2.944141184764414, + "learning_rate": 1.2239409718802927e-06, + "loss": 0.196, + "step": 2028 + }, + { + "epoch": 0.46166097838452785, + "grad_norm": 2.485798960402695, + "learning_rate": 1.2239154340244766e-06, + "loss": 0.1661, + "step": 2029 + }, + { + "epoch": 0.46188850967007966, + "grad_norm": 2.204789648433421, + "learning_rate": 1.2238898839279262e-06, + "loss": 0.1651, + "step": 2030 + }, + { + "epoch": 0.4621160409556314, + "grad_norm": 1.7965983672501258, + "learning_rate": 1.223864321591163e-06, + "loss": 0.0958, + "step": 2031 + }, + { + "epoch": 0.46234357224118316, + "grad_norm": 1.663374784429922, + "learning_rate": 1.2238387470147103e-06, + "loss": 0.1122, + "step": 2032 + }, + { + "epoch": 0.4625711035267349, + "grad_norm": 2.7681620279158974, + "learning_rate": 1.22381316019909e-06, + "loss": 0.1038, + "step": 2033 + }, + { + "epoch": 0.46279863481228667, + "grad_norm": 0.941266935569426, + "learning_rate": 1.2237875611448254e-06, + "loss": 0.1504, + "step": 2034 + }, + { + "epoch": 0.4630261660978385, + "grad_norm": 1.5788004965037967, + "learning_rate": 1.2237619498524397e-06, + "loss": 0.0884, + "step": 2035 + }, + { + "epoch": 0.46325369738339023, + "grad_norm": 1.7335921265291445, + "learning_rate": 1.2237363263224564e-06, + "loss": 0.1805, + "step": 2036 + }, + { + "epoch": 0.463481228668942, + "grad_norm": 1.7271541406470394, + "learning_rate": 1.2237106905553991e-06, + "loss": 0.1099, + "step": 2037 + }, + { + "epoch": 0.46370875995449373, + "grad_norm": 0.9884238734669359, + "learning_rate": 1.2236850425517918e-06, + "loss": 0.0713, + "step": 2038 + }, + { + "epoch": 0.4639362912400455, + "grad_norm": 5.002976224886838, + "learning_rate": 1.2236593823121586e-06, + "loss": 0.2421, + "step": 2039 + }, + { + "epoch": 0.4641638225255973, + "grad_norm": 3.663063876653759, + "learning_rate": 1.223633709837024e-06, + "loss": 0.126, + "step": 2040 + }, + { + "epoch": 0.46439135381114904, + "grad_norm": 2.3948083765141526, + "learning_rate": 1.2236080251269129e-06, + "loss": 0.2088, + "step": 2041 + }, + { + "epoch": 0.4646188850967008, + "grad_norm": 2.728688530304909, + "learning_rate": 1.2235823281823498e-06, + "loss": 0.1795, + "step": 2042 + }, + { + "epoch": 0.46484641638225255, + "grad_norm": 2.7033015325769054, + "learning_rate": 1.2235566190038602e-06, + "loss": 0.2331, + "step": 2043 + }, + { + "epoch": 0.4650739476678043, + "grad_norm": 1.4528538728854552, + "learning_rate": 1.2235308975919697e-06, + "loss": 0.0993, + "step": 2044 + }, + { + "epoch": 0.4653014789533561, + "grad_norm": 3.747453107446523, + "learning_rate": 1.2235051639472037e-06, + "loss": 0.1748, + "step": 2045 + }, + { + "epoch": 0.46552901023890786, + "grad_norm": 3.8624694024749027, + "learning_rate": 1.2234794180700883e-06, + "loss": 0.279, + "step": 2046 + }, + { + "epoch": 0.4657565415244596, + "grad_norm": 2.899667813559601, + "learning_rate": 1.2234536599611496e-06, + "loss": 0.1747, + "step": 2047 + }, + { + "epoch": 0.46598407281001136, + "grad_norm": 1.6765710152355593, + "learning_rate": 1.2234278896209142e-06, + "loss": 0.1285, + "step": 2048 + }, + { + "epoch": 0.4662116040955631, + "grad_norm": 1.7242970700826221, + "learning_rate": 1.2234021070499086e-06, + "loss": 0.1298, + "step": 2049 + }, + { + "epoch": 0.4664391353811149, + "grad_norm": 1.677717106231001, + "learning_rate": 1.22337631224866e-06, + "loss": 0.1136, + "step": 2050 + }, + { + "epoch": 0.4666666666666667, + "grad_norm": 2.5864911073545533, + "learning_rate": 1.2233505052176952e-06, + "loss": 0.1429, + "step": 2051 + }, + { + "epoch": 0.4668941979522184, + "grad_norm": 2.8514324884488667, + "learning_rate": 1.223324685957542e-06, + "loss": 0.1401, + "step": 2052 + }, + { + "epoch": 0.4671217292377702, + "grad_norm": 1.1457675992429823, + "learning_rate": 1.2232988544687282e-06, + "loss": 0.1492, + "step": 2053 + }, + { + "epoch": 0.46734926052332193, + "grad_norm": 1.049329465937079, + "learning_rate": 1.2232730107517813e-06, + "loss": 0.0931, + "step": 2054 + }, + { + "epoch": 0.46757679180887374, + "grad_norm": 3.4990351755141864, + "learning_rate": 1.22324715480723e-06, + "loss": 0.155, + "step": 2055 + }, + { + "epoch": 0.4678043230944255, + "grad_norm": 1.915743720906956, + "learning_rate": 1.2232212866356022e-06, + "loss": 0.1268, + "step": 2056 + }, + { + "epoch": 0.46803185437997724, + "grad_norm": 2.1421674895019907, + "learning_rate": 1.2231954062374272e-06, + "loss": 0.1223, + "step": 2057 + }, + { + "epoch": 0.468259385665529, + "grad_norm": 2.422880416406602, + "learning_rate": 1.2231695136132333e-06, + "loss": 0.1921, + "step": 2058 + }, + { + "epoch": 0.4684869169510808, + "grad_norm": 13.197158009983157, + "learning_rate": 1.2231436087635504e-06, + "loss": 0.1583, + "step": 2059 + }, + { + "epoch": 0.46871444823663255, + "grad_norm": 2.1847258298952767, + "learning_rate": 1.2231176916889072e-06, + "loss": 0.1362, + "step": 2060 + }, + { + "epoch": 0.4689419795221843, + "grad_norm": 1.444519439515334, + "learning_rate": 1.223091762389834e-06, + "loss": 0.1375, + "step": 2061 + }, + { + "epoch": 0.46916951080773606, + "grad_norm": 1.5232732099653936, + "learning_rate": 1.2230658208668606e-06, + "loss": 0.0855, + "step": 2062 + }, + { + "epoch": 0.4693970420932878, + "grad_norm": 1.8870106277847236, + "learning_rate": 1.223039867120517e-06, + "loss": 0.1874, + "step": 2063 + }, + { + "epoch": 0.4696245733788396, + "grad_norm": 3.196966642834261, + "learning_rate": 1.2230139011513338e-06, + "loss": 0.125, + "step": 2064 + }, + { + "epoch": 0.46985210466439137, + "grad_norm": 1.5297510970560153, + "learning_rate": 1.2229879229598416e-06, + "loss": 0.0823, + "step": 2065 + }, + { + "epoch": 0.4700796359499431, + "grad_norm": 2.8720469940024027, + "learning_rate": 1.2229619325465717e-06, + "loss": 0.1788, + "step": 2066 + }, + { + "epoch": 0.47030716723549487, + "grad_norm": 1.3771574147519354, + "learning_rate": 1.2229359299120548e-06, + "loss": 0.1141, + "step": 2067 + }, + { + "epoch": 0.4705346985210466, + "grad_norm": 3.590921354687459, + "learning_rate": 1.2229099150568224e-06, + "loss": 0.1523, + "step": 2068 + }, + { + "epoch": 0.47076222980659843, + "grad_norm": 2.7712958332833653, + "learning_rate": 1.2228838879814065e-06, + "loss": 0.1356, + "step": 2069 + }, + { + "epoch": 0.4709897610921502, + "grad_norm": 2.71072596202505, + "learning_rate": 1.222857848686339e-06, + "loss": 0.2181, + "step": 2070 + }, + { + "epoch": 0.47121729237770194, + "grad_norm": 1.555950330379804, + "learning_rate": 1.2228317971721518e-06, + "loss": 0.1346, + "step": 2071 + }, + { + "epoch": 0.4714448236632537, + "grad_norm": 1.642590700515521, + "learning_rate": 1.2228057334393777e-06, + "loss": 0.0998, + "step": 2072 + }, + { + "epoch": 0.47167235494880544, + "grad_norm": 1.943206327307487, + "learning_rate": 1.2227796574885493e-06, + "loss": 0.1742, + "step": 2073 + }, + { + "epoch": 0.47189988623435725, + "grad_norm": 1.8704660178513137, + "learning_rate": 1.2227535693201994e-06, + "loss": 0.0758, + "step": 2074 + }, + { + "epoch": 0.472127417519909, + "grad_norm": 2.657426724814391, + "learning_rate": 1.2227274689348612e-06, + "loss": 0.1379, + "step": 2075 + }, + { + "epoch": 0.47235494880546075, + "grad_norm": 4.70640052962396, + "learning_rate": 1.2227013563330684e-06, + "loss": 0.2874, + "step": 2076 + }, + { + "epoch": 0.4725824800910125, + "grad_norm": 1.6118104022489745, + "learning_rate": 1.2226752315153543e-06, + "loss": 0.0888, + "step": 2077 + }, + { + "epoch": 0.47281001137656425, + "grad_norm": 2.1906220939459287, + "learning_rate": 1.2226490944822533e-06, + "loss": 0.1162, + "step": 2078 + }, + { + "epoch": 0.47303754266211606, + "grad_norm": 1.5961395424007108, + "learning_rate": 1.2226229452342991e-06, + "loss": 0.0706, + "step": 2079 + }, + { + "epoch": 0.4732650739476678, + "grad_norm": 2.752087193860096, + "learning_rate": 1.2225967837720265e-06, + "loss": 0.1306, + "step": 2080 + }, + { + "epoch": 0.47349260523321957, + "grad_norm": 1.519564459692136, + "learning_rate": 1.2225706100959702e-06, + "loss": 0.144, + "step": 2081 + }, + { + "epoch": 0.4737201365187713, + "grad_norm": 2.5145870829703068, + "learning_rate": 1.2225444242066648e-06, + "loss": 0.1257, + "step": 2082 + }, + { + "epoch": 0.47394766780432307, + "grad_norm": 2.625233347705291, + "learning_rate": 1.2225182261046458e-06, + "loss": 0.2017, + "step": 2083 + }, + { + "epoch": 0.4741751990898749, + "grad_norm": 1.5951025219500239, + "learning_rate": 1.2224920157904487e-06, + "loss": 0.2029, + "step": 2084 + }, + { + "epoch": 0.47440273037542663, + "grad_norm": 1.882372906280183, + "learning_rate": 1.222465793264609e-06, + "loss": 0.1386, + "step": 2085 + }, + { + "epoch": 0.4746302616609784, + "grad_norm": 2.312787899846543, + "learning_rate": 1.2224395585276626e-06, + "loss": 0.21, + "step": 2086 + }, + { + "epoch": 0.47485779294653013, + "grad_norm": 2.7346134087318985, + "learning_rate": 1.2224133115801458e-06, + "loss": 0.1635, + "step": 2087 + }, + { + "epoch": 0.4750853242320819, + "grad_norm": 1.9967379262226481, + "learning_rate": 1.222387052422595e-06, + "loss": 0.1022, + "step": 2088 + }, + { + "epoch": 0.4753128555176337, + "grad_norm": 2.567172930928881, + "learning_rate": 1.222360781055547e-06, + "loss": 0.1497, + "step": 2089 + }, + { + "epoch": 0.47554038680318544, + "grad_norm": 1.4244534395274109, + "learning_rate": 1.2223344974795386e-06, + "loss": 0.2069, + "step": 2090 + }, + { + "epoch": 0.4757679180887372, + "grad_norm": 2.156175658506119, + "learning_rate": 1.2223082016951071e-06, + "loss": 0.1239, + "step": 2091 + }, + { + "epoch": 0.47599544937428895, + "grad_norm": 2.834630845876437, + "learning_rate": 1.2222818937027898e-06, + "loss": 0.0872, + "step": 2092 + }, + { + "epoch": 0.47622298065984076, + "grad_norm": 1.8380588142136158, + "learning_rate": 1.2222555735031245e-06, + "loss": 0.1091, + "step": 2093 + }, + { + "epoch": 0.4764505119453925, + "grad_norm": 2.5737283640174122, + "learning_rate": 1.222229241096649e-06, + "loss": 0.2635, + "step": 2094 + }, + { + "epoch": 0.47667804323094426, + "grad_norm": 1.4566454597308833, + "learning_rate": 1.2222028964839017e-06, + "loss": 0.1348, + "step": 2095 + }, + { + "epoch": 0.476905574516496, + "grad_norm": 1.9697307912334634, + "learning_rate": 1.2221765396654209e-06, + "loss": 0.1558, + "step": 2096 + }, + { + "epoch": 0.47713310580204776, + "grad_norm": 1.451409908091677, + "learning_rate": 1.2221501706417454e-06, + "loss": 0.12, + "step": 2097 + }, + { + "epoch": 0.47736063708759957, + "grad_norm": 2.303809415774566, + "learning_rate": 1.2221237894134138e-06, + "loss": 0.1487, + "step": 2098 + }, + { + "epoch": 0.4775881683731513, + "grad_norm": 2.6908048074102426, + "learning_rate": 1.222097395980966e-06, + "loss": 0.2478, + "step": 2099 + }, + { + "epoch": 0.4778156996587031, + "grad_norm": 1.9153149162558512, + "learning_rate": 1.2220709903449403e-06, + "loss": 0.1633, + "step": 2100 + }, + { + "epoch": 0.4780432309442548, + "grad_norm": 1.5349309710920869, + "learning_rate": 1.2220445725058775e-06, + "loss": 0.111, + "step": 2101 + }, + { + "epoch": 0.4782707622298066, + "grad_norm": 1.8132178697871648, + "learning_rate": 1.222018142464317e-06, + "loss": 0.2154, + "step": 2102 + }, + { + "epoch": 0.4784982935153584, + "grad_norm": 1.7599313333363966, + "learning_rate": 1.221991700220799e-06, + "loss": 0.1222, + "step": 2103 + }, + { + "epoch": 0.47872582480091014, + "grad_norm": 2.029241242931466, + "learning_rate": 1.221965245775864e-06, + "loss": 0.1307, + "step": 2104 + }, + { + "epoch": 0.4789533560864619, + "grad_norm": 2.960462408034195, + "learning_rate": 1.2219387791300527e-06, + "loss": 0.1382, + "step": 2105 + }, + { + "epoch": 0.47918088737201364, + "grad_norm": 2.347329357814124, + "learning_rate": 1.2219123002839058e-06, + "loss": 0.1531, + "step": 2106 + }, + { + "epoch": 0.4794084186575654, + "grad_norm": 1.4488221446198866, + "learning_rate": 1.2218858092379648e-06, + "loss": 0.1309, + "step": 2107 + }, + { + "epoch": 0.4796359499431172, + "grad_norm": 4.331492523033653, + "learning_rate": 1.221859305992771e-06, + "loss": 0.1518, + "step": 2108 + }, + { + "epoch": 0.47986348122866895, + "grad_norm": 2.2907841977007695, + "learning_rate": 1.2218327905488662e-06, + "loss": 0.2048, + "step": 2109 + }, + { + "epoch": 0.4800910125142207, + "grad_norm": 3.226955121067652, + "learning_rate": 1.221806262906792e-06, + "loss": 0.1442, + "step": 2110 + }, + { + "epoch": 0.48031854379977246, + "grad_norm": 3.004082420754334, + "learning_rate": 1.221779723067091e-06, + "loss": 0.1558, + "step": 2111 + }, + { + "epoch": 0.4805460750853242, + "grad_norm": 1.2580166705137232, + "learning_rate": 1.2217531710303053e-06, + "loss": 0.108, + "step": 2112 + }, + { + "epoch": 0.480773606370876, + "grad_norm": 2.219396640494861, + "learning_rate": 1.2217266067969778e-06, + "loss": 0.2107, + "step": 2113 + }, + { + "epoch": 0.48100113765642777, + "grad_norm": 1.497489945400175, + "learning_rate": 1.221700030367651e-06, + "loss": 0.1004, + "step": 2114 + }, + { + "epoch": 0.4812286689419795, + "grad_norm": 2.1517574031007523, + "learning_rate": 1.2216734417428686e-06, + "loss": 0.2276, + "step": 2115 + }, + { + "epoch": 0.4814562002275313, + "grad_norm": 1.3524221649994943, + "learning_rate": 1.2216468409231738e-06, + "loss": 0.1168, + "step": 2116 + }, + { + "epoch": 0.481683731513083, + "grad_norm": 2.8589810452438913, + "learning_rate": 1.2216202279091104e-06, + "loss": 0.1496, + "step": 2117 + }, + { + "epoch": 0.48191126279863483, + "grad_norm": 3.14821100455351, + "learning_rate": 1.2215936027012221e-06, + "loss": 0.1713, + "step": 2118 + }, + { + "epoch": 0.4821387940841866, + "grad_norm": 3.0269143310768363, + "learning_rate": 1.2215669653000532e-06, + "loss": 0.1253, + "step": 2119 + }, + { + "epoch": 0.48236632536973834, + "grad_norm": 1.620917497810277, + "learning_rate": 1.2215403157061478e-06, + "loss": 0.1045, + "step": 2120 + }, + { + "epoch": 0.4825938566552901, + "grad_norm": 4.356411863912081, + "learning_rate": 1.2215136539200512e-06, + "loss": 0.1898, + "step": 2121 + }, + { + "epoch": 0.48282138794084184, + "grad_norm": 2.0989757571956296, + "learning_rate": 1.2214869799423078e-06, + "loss": 0.0811, + "step": 2122 + }, + { + "epoch": 0.48304891922639365, + "grad_norm": 2.7274929171067592, + "learning_rate": 1.2214602937734632e-06, + "loss": 0.1439, + "step": 2123 + }, + { + "epoch": 0.4832764505119454, + "grad_norm": 2.424703742996382, + "learning_rate": 1.2214335954140624e-06, + "loss": 0.1519, + "step": 2124 + }, + { + "epoch": 0.48350398179749715, + "grad_norm": 1.890465690299916, + "learning_rate": 1.221406884864651e-06, + "loss": 0.0998, + "step": 2125 + }, + { + "epoch": 0.4837315130830489, + "grad_norm": 2.326638607730014, + "learning_rate": 1.2213801621257754e-06, + "loss": 0.2102, + "step": 2126 + }, + { + "epoch": 0.48395904436860065, + "grad_norm": 2.066734671660341, + "learning_rate": 1.2213534271979815e-06, + "loss": 0.1374, + "step": 2127 + }, + { + "epoch": 0.48418657565415246, + "grad_norm": 2.172208059040071, + "learning_rate": 1.2213266800818158e-06, + "loss": 0.1312, + "step": 2128 + }, + { + "epoch": 0.4844141069397042, + "grad_norm": 1.8499994829117998, + "learning_rate": 1.2212999207778246e-06, + "loss": 0.1154, + "step": 2129 + }, + { + "epoch": 0.48464163822525597, + "grad_norm": 2.6816295680240256, + "learning_rate": 1.2212731492865552e-06, + "loss": 0.0923, + "step": 2130 + }, + { + "epoch": 0.4848691695108077, + "grad_norm": 3.0796575929883354, + "learning_rate": 1.2212463656085548e-06, + "loss": 0.1856, + "step": 2131 + }, + { + "epoch": 0.4850967007963595, + "grad_norm": 1.684680537758439, + "learning_rate": 1.2212195697443704e-06, + "loss": 0.1398, + "step": 2132 + }, + { + "epoch": 0.4853242320819113, + "grad_norm": 2.25319089168003, + "learning_rate": 1.2211927616945502e-06, + "loss": 0.1027, + "step": 2133 + }, + { + "epoch": 0.48555176336746303, + "grad_norm": 3.0215260043450307, + "learning_rate": 1.2211659414596417e-06, + "loss": 0.1564, + "step": 2134 + }, + { + "epoch": 0.4857792946530148, + "grad_norm": 3.420547138984465, + "learning_rate": 1.2211391090401931e-06, + "loss": 0.1629, + "step": 2135 + }, + { + "epoch": 0.48600682593856653, + "grad_norm": 2.7151912053510223, + "learning_rate": 1.2211122644367531e-06, + "loss": 0.1562, + "step": 2136 + }, + { + "epoch": 0.48623435722411834, + "grad_norm": 3.168254535519415, + "learning_rate": 1.2210854076498699e-06, + "loss": 0.2003, + "step": 2137 + }, + { + "epoch": 0.4864618885096701, + "grad_norm": 2.4342182669126435, + "learning_rate": 1.2210585386800927e-06, + "loss": 0.142, + "step": 2138 + }, + { + "epoch": 0.48668941979522184, + "grad_norm": 2.365252540519636, + "learning_rate": 1.2210316575279707e-06, + "loss": 0.1409, + "step": 2139 + }, + { + "epoch": 0.4869169510807736, + "grad_norm": 1.1372019419508703, + "learning_rate": 1.221004764194053e-06, + "loss": 0.1547, + "step": 2140 + }, + { + "epoch": 0.48714448236632535, + "grad_norm": 3.75609449121758, + "learning_rate": 1.22097785867889e-06, + "loss": 0.1512, + "step": 2141 + }, + { + "epoch": 0.48737201365187716, + "grad_norm": 1.7473221878185898, + "learning_rate": 1.2209509409830304e-06, + "loss": 0.2868, + "step": 2142 + }, + { + "epoch": 0.4875995449374289, + "grad_norm": 2.0082173427217382, + "learning_rate": 1.2209240111070254e-06, + "loss": 0.1812, + "step": 2143 + }, + { + "epoch": 0.48782707622298066, + "grad_norm": 2.8863675080606415, + "learning_rate": 1.2208970690514247e-06, + "loss": 0.1066, + "step": 2144 + }, + { + "epoch": 0.4880546075085324, + "grad_norm": 1.4857978026455614, + "learning_rate": 1.2208701148167795e-06, + "loss": 0.0811, + "step": 2145 + }, + { + "epoch": 0.48828213879408416, + "grad_norm": 3.2054667060001383, + "learning_rate": 1.2208431484036405e-06, + "loss": 0.194, + "step": 2146 + }, + { + "epoch": 0.48850967007963597, + "grad_norm": 2.7115726560617133, + "learning_rate": 1.2208161698125584e-06, + "loss": 0.1733, + "step": 2147 + }, + { + "epoch": 0.4887372013651877, + "grad_norm": 3.272715328933079, + "learning_rate": 1.2207891790440852e-06, + "loss": 0.2408, + "step": 2148 + }, + { + "epoch": 0.4889647326507395, + "grad_norm": 2.7359042131162714, + "learning_rate": 1.2207621760987723e-06, + "loss": 0.1235, + "step": 2149 + }, + { + "epoch": 0.4891922639362912, + "grad_norm": 3.3510438031026646, + "learning_rate": 1.2207351609771718e-06, + "loss": 0.1649, + "step": 2150 + }, + { + "epoch": 0.489419795221843, + "grad_norm": 1.914954577682318, + "learning_rate": 1.2207081336798352e-06, + "loss": 0.2042, + "step": 2151 + }, + { + "epoch": 0.4896473265073948, + "grad_norm": 3.0994780231949215, + "learning_rate": 1.2206810942073158e-06, + "loss": 0.0953, + "step": 2152 + }, + { + "epoch": 0.48987485779294654, + "grad_norm": 1.927931591193398, + "learning_rate": 1.2206540425601653e-06, + "loss": 0.1011, + "step": 2153 + }, + { + "epoch": 0.4901023890784983, + "grad_norm": 1.23851496034413, + "learning_rate": 1.2206269787389374e-06, + "loss": 0.0811, + "step": 2154 + }, + { + "epoch": 0.49032992036405004, + "grad_norm": 2.31064364212512, + "learning_rate": 1.220599902744185e-06, + "loss": 0.1341, + "step": 2155 + }, + { + "epoch": 0.4905574516496018, + "grad_norm": 1.7800269493495693, + "learning_rate": 1.220572814576461e-06, + "loss": 0.0611, + "step": 2156 + }, + { + "epoch": 0.4907849829351536, + "grad_norm": 1.4294481100641636, + "learning_rate": 1.2205457142363197e-06, + "loss": 0.1242, + "step": 2157 + }, + { + "epoch": 0.49101251422070535, + "grad_norm": 3.32236728544913, + "learning_rate": 1.2205186017243146e-06, + "loss": 0.1439, + "step": 2158 + }, + { + "epoch": 0.4912400455062571, + "grad_norm": 2.1161347164300737, + "learning_rate": 1.2204914770409999e-06, + "loss": 0.1312, + "step": 2159 + }, + { + "epoch": 0.49146757679180886, + "grad_norm": 1.6760298562325886, + "learning_rate": 1.22046434018693e-06, + "loss": 0.1367, + "step": 2160 + }, + { + "epoch": 0.4916951080773606, + "grad_norm": 2.5053031123422844, + "learning_rate": 1.2204371911626597e-06, + "loss": 0.1004, + "step": 2161 + }, + { + "epoch": 0.4919226393629124, + "grad_norm": 1.8628349503031747, + "learning_rate": 1.2204100299687436e-06, + "loss": 0.2272, + "step": 2162 + }, + { + "epoch": 0.49215017064846417, + "grad_norm": 2.9254446518246207, + "learning_rate": 1.2203828566057368e-06, + "loss": 0.1306, + "step": 2163 + }, + { + "epoch": 0.4923777019340159, + "grad_norm": 2.156502914233145, + "learning_rate": 1.2203556710741948e-06, + "loss": 0.1165, + "step": 2164 + }, + { + "epoch": 0.4926052332195677, + "grad_norm": 1.765376175489098, + "learning_rate": 1.2203284733746734e-06, + "loss": 0.1386, + "step": 2165 + }, + { + "epoch": 0.4928327645051195, + "grad_norm": 2.680620034477845, + "learning_rate": 1.2203012635077283e-06, + "loss": 0.1641, + "step": 2166 + }, + { + "epoch": 0.49306029579067123, + "grad_norm": 2.096581132566952, + "learning_rate": 1.2202740414739153e-06, + "loss": 0.1464, + "step": 2167 + }, + { + "epoch": 0.493287827076223, + "grad_norm": 2.330371824149322, + "learning_rate": 1.2202468072737914e-06, + "loss": 0.0977, + "step": 2168 + }, + { + "epoch": 0.49351535836177474, + "grad_norm": 2.736555613980781, + "learning_rate": 1.2202195609079128e-06, + "loss": 0.1212, + "step": 2169 + }, + { + "epoch": 0.4937428896473265, + "grad_norm": 1.370889091477706, + "learning_rate": 1.2201923023768365e-06, + "loss": 0.1147, + "step": 2170 + }, + { + "epoch": 0.4939704209328783, + "grad_norm": 1.9931953231084023, + "learning_rate": 1.2201650316811194e-06, + "loss": 0.1843, + "step": 2171 + }, + { + "epoch": 0.49419795221843005, + "grad_norm": 1.9873229116599542, + "learning_rate": 1.2201377488213195e-06, + "loss": 0.2833, + "step": 2172 + }, + { + "epoch": 0.4944254835039818, + "grad_norm": 2.766859520035512, + "learning_rate": 1.2201104537979934e-06, + "loss": 0.1065, + "step": 2173 + }, + { + "epoch": 0.49465301478953355, + "grad_norm": 2.8257426123215947, + "learning_rate": 1.2200831466117e-06, + "loss": 0.171, + "step": 2174 + }, + { + "epoch": 0.4948805460750853, + "grad_norm": 2.416755735243659, + "learning_rate": 1.2200558272629967e-06, + "loss": 0.1049, + "step": 2175 + }, + { + "epoch": 0.4951080773606371, + "grad_norm": 3.847603230089162, + "learning_rate": 1.2200284957524421e-06, + "loss": 0.1543, + "step": 2176 + }, + { + "epoch": 0.49533560864618886, + "grad_norm": 1.808070223158287, + "learning_rate": 1.2200011520805947e-06, + "loss": 0.1589, + "step": 2177 + }, + { + "epoch": 0.4955631399317406, + "grad_norm": 1.4634936792045132, + "learning_rate": 1.2199737962480136e-06, + "loss": 0.1005, + "step": 2178 + }, + { + "epoch": 0.49579067121729237, + "grad_norm": 2.0415979620186433, + "learning_rate": 1.2199464282552576e-06, + "loss": 0.1504, + "step": 2179 + }, + { + "epoch": 0.4960182025028441, + "grad_norm": 1.8340897113272197, + "learning_rate": 1.2199190481028864e-06, + "loss": 0.1168, + "step": 2180 + }, + { + "epoch": 0.4962457337883959, + "grad_norm": 1.9694309627394886, + "learning_rate": 1.2198916557914592e-06, + "loss": 0.0702, + "step": 2181 + }, + { + "epoch": 0.4964732650739477, + "grad_norm": 1.757184086126526, + "learning_rate": 1.2198642513215362e-06, + "loss": 0.0998, + "step": 2182 + }, + { + "epoch": 0.49670079635949943, + "grad_norm": 1.2539776130317066, + "learning_rate": 1.2198368346936774e-06, + "loss": 0.0843, + "step": 2183 + }, + { + "epoch": 0.4969283276450512, + "grad_norm": 1.3188419596574201, + "learning_rate": 1.219809405908443e-06, + "loss": 0.0788, + "step": 2184 + }, + { + "epoch": 0.49715585893060293, + "grad_norm": 1.523980855726603, + "learning_rate": 1.2197819649663934e-06, + "loss": 0.1011, + "step": 2185 + }, + { + "epoch": 0.49738339021615474, + "grad_norm": 2.409527037495328, + "learning_rate": 1.21975451186809e-06, + "loss": 0.0966, + "step": 2186 + }, + { + "epoch": 0.4976109215017065, + "grad_norm": 1.3037564823590608, + "learning_rate": 1.2197270466140936e-06, + "loss": 0.1047, + "step": 2187 + }, + { + "epoch": 0.49783845278725825, + "grad_norm": 1.789706258984112, + "learning_rate": 1.2196995692049656e-06, + "loss": 0.072, + "step": 2188 + }, + { + "epoch": 0.49806598407281, + "grad_norm": 2.4890313700185382, + "learning_rate": 1.2196720796412675e-06, + "loss": 0.1822, + "step": 2189 + }, + { + "epoch": 0.49829351535836175, + "grad_norm": 3.43016997911828, + "learning_rate": 1.219644577923561e-06, + "loss": 0.1781, + "step": 2190 + }, + { + "epoch": 0.49852104664391356, + "grad_norm": 1.6871311149806307, + "learning_rate": 1.2196170640524087e-06, + "loss": 0.1616, + "step": 2191 + }, + { + "epoch": 0.4987485779294653, + "grad_norm": 2.2297408965863132, + "learning_rate": 1.2195895380283725e-06, + "loss": 0.2451, + "step": 2192 + }, + { + "epoch": 0.49897610921501706, + "grad_norm": 1.389848256901662, + "learning_rate": 1.219561999852015e-06, + "loss": 0.0917, + "step": 2193 + }, + { + "epoch": 0.4992036405005688, + "grad_norm": 3.041396119145658, + "learning_rate": 1.2195344495238993e-06, + "loss": 0.1329, + "step": 2194 + }, + { + "epoch": 0.49943117178612056, + "grad_norm": 2.216904135088197, + "learning_rate": 1.2195068870445878e-06, + "loss": 0.1071, + "step": 2195 + }, + { + "epoch": 0.49965870307167237, + "grad_norm": 1.9688477353283649, + "learning_rate": 1.2194793124146447e-06, + "loss": 0.0971, + "step": 2196 + }, + { + "epoch": 0.4998862343572241, + "grad_norm": 2.164052002018189, + "learning_rate": 1.2194517256346333e-06, + "loss": 0.1602, + "step": 2197 + }, + { + "epoch": 0.5001137656427759, + "grad_norm": 1.388280113458421, + "learning_rate": 1.2194241267051172e-06, + "loss": 0.0848, + "step": 2198 + }, + { + "epoch": 0.5003412969283276, + "grad_norm": 1.9472805252357173, + "learning_rate": 1.2193965156266605e-06, + "loss": 0.1418, + "step": 2199 + }, + { + "epoch": 0.5005688282138794, + "grad_norm": 4.465127702734097, + "learning_rate": 1.2193688923998277e-06, + "loss": 0.1968, + "step": 2200 + }, + { + "epoch": 0.5007963594994311, + "grad_norm": 2.8204878994002147, + "learning_rate": 1.219341257025183e-06, + "loss": 0.1636, + "step": 2201 + }, + { + "epoch": 0.5010238907849829, + "grad_norm": 4.7787213981946834, + "learning_rate": 1.2193136095032918e-06, + "loss": 0.1761, + "step": 2202 + }, + { + "epoch": 0.5012514220705347, + "grad_norm": 2.2115998556881116, + "learning_rate": 1.2192859498347189e-06, + "loss": 0.1736, + "step": 2203 + }, + { + "epoch": 0.5014789533560865, + "grad_norm": 2.3947307414083907, + "learning_rate": 1.2192582780200293e-06, + "loss": 0.1562, + "step": 2204 + }, + { + "epoch": 0.5017064846416383, + "grad_norm": 1.137850645365097, + "learning_rate": 1.219230594059789e-06, + "loss": 0.1245, + "step": 2205 + }, + { + "epoch": 0.50193401592719, + "grad_norm": 1.3050370676179703, + "learning_rate": 1.2192028979545637e-06, + "loss": 0.144, + "step": 2206 + }, + { + "epoch": 0.5021615472127418, + "grad_norm": 1.5689422022860104, + "learning_rate": 1.2191751897049192e-06, + "loss": 0.1109, + "step": 2207 + }, + { + "epoch": 0.5023890784982935, + "grad_norm": 1.1140260533548134, + "learning_rate": 1.2191474693114223e-06, + "loss": 0.0729, + "step": 2208 + }, + { + "epoch": 0.5026166097838453, + "grad_norm": 3.1668244036808404, + "learning_rate": 1.2191197367746389e-06, + "loss": 0.1086, + "step": 2209 + }, + { + "epoch": 0.502844141069397, + "grad_norm": 1.2112158809642222, + "learning_rate": 1.2190919920951363e-06, + "loss": 0.0925, + "step": 2210 + }, + { + "epoch": 0.5030716723549488, + "grad_norm": 3.140103243583768, + "learning_rate": 1.2190642352734814e-06, + "loss": 0.1309, + "step": 2211 + }, + { + "epoch": 0.5032992036405005, + "grad_norm": 2.20383956504207, + "learning_rate": 1.2190364663102417e-06, + "loss": 0.1261, + "step": 2212 + }, + { + "epoch": 0.5035267349260524, + "grad_norm": 4.70556372527374, + "learning_rate": 1.2190086852059844e-06, + "loss": 0.1457, + "step": 2213 + }, + { + "epoch": 0.5037542662116041, + "grad_norm": 1.5904720691715128, + "learning_rate": 1.2189808919612773e-06, + "loss": 0.117, + "step": 2214 + }, + { + "epoch": 0.5039817974971559, + "grad_norm": 4.614372733382792, + "learning_rate": 1.2189530865766888e-06, + "loss": 0.2221, + "step": 2215 + }, + { + "epoch": 0.5042093287827076, + "grad_norm": 2.1481536506375143, + "learning_rate": 1.218925269052787e-06, + "loss": 0.1181, + "step": 2216 + }, + { + "epoch": 0.5044368600682594, + "grad_norm": 3.0881852694485157, + "learning_rate": 1.2188974393901404e-06, + "loss": 0.1824, + "step": 2217 + }, + { + "epoch": 0.5046643913538111, + "grad_norm": 2.38294561984735, + "learning_rate": 1.2188695975893177e-06, + "loss": 0.1518, + "step": 2218 + }, + { + "epoch": 0.5048919226393629, + "grad_norm": 2.1925852844845783, + "learning_rate": 1.218841743650888e-06, + "loss": 0.172, + "step": 2219 + }, + { + "epoch": 0.5051194539249146, + "grad_norm": 1.3374825429070463, + "learning_rate": 1.2188138775754207e-06, + "loss": 0.0952, + "step": 2220 + }, + { + "epoch": 0.5053469852104664, + "grad_norm": 1.7224866619431942, + "learning_rate": 1.2187859993634854e-06, + "loss": 0.1261, + "step": 2221 + }, + { + "epoch": 0.5055745164960183, + "grad_norm": 1.4650603873711627, + "learning_rate": 1.2187581090156518e-06, + "loss": 0.0962, + "step": 2222 + }, + { + "epoch": 0.50580204778157, + "grad_norm": 2.2148606700249736, + "learning_rate": 1.2187302065324896e-06, + "loss": 0.2173, + "step": 2223 + }, + { + "epoch": 0.5060295790671218, + "grad_norm": 2.3828933955331864, + "learning_rate": 1.2187022919145695e-06, + "loss": 0.133, + "step": 2224 + }, + { + "epoch": 0.5062571103526735, + "grad_norm": 2.529903147299925, + "learning_rate": 1.2186743651624617e-06, + "loss": 0.1873, + "step": 2225 + }, + { + "epoch": 0.5064846416382253, + "grad_norm": 1.0288567052421398, + "learning_rate": 1.2186464262767372e-06, + "loss": 0.1036, + "step": 2226 + }, + { + "epoch": 0.506712172923777, + "grad_norm": 2.7669101294284904, + "learning_rate": 1.2186184752579671e-06, + "loss": 0.1161, + "step": 2227 + }, + { + "epoch": 0.5069397042093288, + "grad_norm": 2.613686872517249, + "learning_rate": 1.2185905121067223e-06, + "loss": 0.1654, + "step": 2228 + }, + { + "epoch": 0.5071672354948805, + "grad_norm": 1.5815125330195172, + "learning_rate": 1.218562536823575e-06, + "loss": 0.161, + "step": 2229 + }, + { + "epoch": 0.5073947667804323, + "grad_norm": 2.5849262983185803, + "learning_rate": 1.2185345494090959e-06, + "loss": 0.1458, + "step": 2230 + }, + { + "epoch": 0.507622298065984, + "grad_norm": 1.7848106830254082, + "learning_rate": 1.218506549863858e-06, + "loss": 0.0774, + "step": 2231 + }, + { + "epoch": 0.5078498293515359, + "grad_norm": 2.0041648260692617, + "learning_rate": 1.2184785381884332e-06, + "loss": 0.1407, + "step": 2232 + }, + { + "epoch": 0.5080773606370876, + "grad_norm": 2.6691932199960813, + "learning_rate": 1.218450514383394e-06, + "loss": 0.2101, + "step": 2233 + }, + { + "epoch": 0.5083048919226394, + "grad_norm": 0.8618265034013133, + "learning_rate": 1.2184224784493127e-06, + "loss": 0.1081, + "step": 2234 + }, + { + "epoch": 0.5085324232081911, + "grad_norm": 2.7387414496667972, + "learning_rate": 1.218394430386763e-06, + "loss": 0.2138, + "step": 2235 + }, + { + "epoch": 0.5087599544937429, + "grad_norm": 3.077864568174803, + "learning_rate": 1.2183663701963181e-06, + "loss": 0.1218, + "step": 2236 + }, + { + "epoch": 0.5089874857792946, + "grad_norm": 2.689017818647376, + "learning_rate": 1.218338297878551e-06, + "loss": 0.1825, + "step": 2237 + }, + { + "epoch": 0.5092150170648464, + "grad_norm": 3.296215386410284, + "learning_rate": 1.2183102134340361e-06, + "loss": 0.1954, + "step": 2238 + }, + { + "epoch": 0.5094425483503981, + "grad_norm": 3.577016938326506, + "learning_rate": 1.2182821168633466e-06, + "loss": 0.1965, + "step": 2239 + }, + { + "epoch": 0.5096700796359499, + "grad_norm": 1.4427654273235515, + "learning_rate": 1.2182540081670574e-06, + "loss": 0.1356, + "step": 2240 + }, + { + "epoch": 0.5098976109215017, + "grad_norm": 2.246195129100446, + "learning_rate": 1.2182258873457428e-06, + "loss": 0.2082, + "step": 2241 + }, + { + "epoch": 0.5101251422070535, + "grad_norm": 1.8974029135655925, + "learning_rate": 1.2181977543999776e-06, + "loss": 0.1416, + "step": 2242 + }, + { + "epoch": 0.5103526734926053, + "grad_norm": 1.1656107173765236, + "learning_rate": 1.2181696093303363e-06, + "loss": 0.0992, + "step": 2243 + }, + { + "epoch": 0.510580204778157, + "grad_norm": 1.7060747898678588, + "learning_rate": 1.218141452137395e-06, + "loss": 0.1836, + "step": 2244 + }, + { + "epoch": 0.5108077360637088, + "grad_norm": 2.7425212256939204, + "learning_rate": 1.2181132828217284e-06, + "loss": 0.1172, + "step": 2245 + }, + { + "epoch": 0.5110352673492605, + "grad_norm": 3.4984114514747735, + "learning_rate": 1.2180851013839127e-06, + "loss": 0.1448, + "step": 2246 + }, + { + "epoch": 0.5112627986348123, + "grad_norm": 1.0541301873031321, + "learning_rate": 1.2180569078245236e-06, + "loss": 0.1137, + "step": 2247 + }, + { + "epoch": 0.511490329920364, + "grad_norm": 1.9754045421213982, + "learning_rate": 1.2180287021441372e-06, + "loss": 0.1268, + "step": 2248 + }, + { + "epoch": 0.5117178612059158, + "grad_norm": 2.2348913132386135, + "learning_rate": 1.2180004843433305e-06, + "loss": 0.151, + "step": 2249 + }, + { + "epoch": 0.5119453924914675, + "grad_norm": 2.69280556200464, + "learning_rate": 1.21797225442268e-06, + "loss": 0.1318, + "step": 2250 + }, + { + "epoch": 0.5121729237770194, + "grad_norm": 2.8889424540900004, + "learning_rate": 1.2179440123827625e-06, + "loss": 0.2037, + "step": 2251 + }, + { + "epoch": 0.5124004550625711, + "grad_norm": 1.7014653966715638, + "learning_rate": 1.2179157582241554e-06, + "loss": 0.1011, + "step": 2252 + }, + { + "epoch": 0.5126279863481229, + "grad_norm": 1.2440635472766928, + "learning_rate": 1.2178874919474359e-06, + "loss": 0.0976, + "step": 2253 + }, + { + "epoch": 0.5128555176336747, + "grad_norm": 1.9265304355947193, + "learning_rate": 1.217859213553182e-06, + "loss": 0.1251, + "step": 2254 + }, + { + "epoch": 0.5130830489192264, + "grad_norm": 2.692342674859063, + "learning_rate": 1.2178309230419714e-06, + "loss": 0.1062, + "step": 2255 + }, + { + "epoch": 0.5133105802047782, + "grad_norm": 2.851487861587013, + "learning_rate": 1.2178026204143827e-06, + "loss": 0.1802, + "step": 2256 + }, + { + "epoch": 0.5135381114903299, + "grad_norm": 2.593347894957166, + "learning_rate": 1.217774305670994e-06, + "loss": 0.1785, + "step": 2257 + }, + { + "epoch": 0.5137656427758817, + "grad_norm": 3.0650056170459696, + "learning_rate": 1.2177459788123841e-06, + "loss": 0.134, + "step": 2258 + }, + { + "epoch": 0.5139931740614334, + "grad_norm": 1.4532346128286195, + "learning_rate": 1.217717639839132e-06, + "loss": 0.1381, + "step": 2259 + }, + { + "epoch": 0.5142207053469852, + "grad_norm": 1.7176834033996482, + "learning_rate": 1.2176892887518166e-06, + "loss": 0.1169, + "step": 2260 + }, + { + "epoch": 0.514448236632537, + "grad_norm": 1.7341224196467013, + "learning_rate": 1.2176609255510176e-06, + "loss": 0.1231, + "step": 2261 + }, + { + "epoch": 0.5146757679180888, + "grad_norm": 1.629200246930976, + "learning_rate": 1.217632550237315e-06, + "loss": 0.0887, + "step": 2262 + }, + { + "epoch": 0.5149032992036405, + "grad_norm": 1.30225190150621, + "learning_rate": 1.2176041628112884e-06, + "loss": 0.1014, + "step": 2263 + }, + { + "epoch": 0.5151308304891923, + "grad_norm": 2.250042339940742, + "learning_rate": 1.217575763273518e-06, + "loss": 0.1716, + "step": 2264 + }, + { + "epoch": 0.515358361774744, + "grad_norm": 4.277366750571248, + "learning_rate": 1.217547351624584e-06, + "loss": 0.1202, + "step": 2265 + }, + { + "epoch": 0.5155858930602958, + "grad_norm": 2.1130660104324037, + "learning_rate": 1.2175189278650677e-06, + "loss": 0.1429, + "step": 2266 + }, + { + "epoch": 0.5158134243458475, + "grad_norm": 1.897460974800647, + "learning_rate": 1.2174904919955493e-06, + "loss": 0.1326, + "step": 2267 + }, + { + "epoch": 0.5160409556313993, + "grad_norm": 1.55299514699498, + "learning_rate": 1.2174620440166105e-06, + "loss": 0.1039, + "step": 2268 + }, + { + "epoch": 0.516268486916951, + "grad_norm": 1.4624259135108133, + "learning_rate": 1.2174335839288326e-06, + "loss": 0.1051, + "step": 2269 + }, + { + "epoch": 0.5164960182025028, + "grad_norm": 2.4152486054409628, + "learning_rate": 1.2174051117327972e-06, + "loss": 0.1334, + "step": 2270 + }, + { + "epoch": 0.5167235494880547, + "grad_norm": 2.7224350669895383, + "learning_rate": 1.2173766274290864e-06, + "loss": 0.1522, + "step": 2271 + }, + { + "epoch": 0.5169510807736064, + "grad_norm": 2.6898213271514364, + "learning_rate": 1.217348131018282e-06, + "loss": 0.1178, + "step": 2272 + }, + { + "epoch": 0.5171786120591582, + "grad_norm": 2.17692457415729, + "learning_rate": 1.2173196225009666e-06, + "loss": 0.1005, + "step": 2273 + }, + { + "epoch": 0.5174061433447099, + "grad_norm": 2.183356804193101, + "learning_rate": 1.217291101877723e-06, + "loss": 0.1492, + "step": 2274 + }, + { + "epoch": 0.5176336746302617, + "grad_norm": 1.8408045884703919, + "learning_rate": 1.2172625691491342e-06, + "loss": 0.1157, + "step": 2275 + }, + { + "epoch": 0.5178612059158134, + "grad_norm": 1.182594251283049, + "learning_rate": 1.217234024315783e-06, + "loss": 0.0909, + "step": 2276 + }, + { + "epoch": 0.5180887372013652, + "grad_norm": 2.3264197436803267, + "learning_rate": 1.2172054673782527e-06, + "loss": 0.1377, + "step": 2277 + }, + { + "epoch": 0.5183162684869169, + "grad_norm": 1.8746855825776054, + "learning_rate": 1.2171768983371276e-06, + "loss": 0.1614, + "step": 2278 + }, + { + "epoch": 0.5185437997724687, + "grad_norm": 2.681521889311292, + "learning_rate": 1.217148317192991e-06, + "loss": 0.1334, + "step": 2279 + }, + { + "epoch": 0.5187713310580204, + "grad_norm": 1.4510905197833242, + "learning_rate": 1.2171197239464274e-06, + "loss": 0.0984, + "step": 2280 + }, + { + "epoch": 0.5189988623435723, + "grad_norm": 2.7722645937153088, + "learning_rate": 1.217091118598021e-06, + "loss": 0.1025, + "step": 2281 + }, + { + "epoch": 0.519226393629124, + "grad_norm": 2.018163494797303, + "learning_rate": 1.2170625011483565e-06, + "loss": 0.115, + "step": 2282 + }, + { + "epoch": 0.5194539249146758, + "grad_norm": 3.7374281896722232, + "learning_rate": 1.2170338715980187e-06, + "loss": 0.1673, + "step": 2283 + }, + { + "epoch": 0.5196814562002275, + "grad_norm": 2.8429408206378035, + "learning_rate": 1.217005229947593e-06, + "loss": 0.1083, + "step": 2284 + }, + { + "epoch": 0.5199089874857793, + "grad_norm": 1.7024003746323926, + "learning_rate": 1.2169765761976646e-06, + "loss": 0.1382, + "step": 2285 + }, + { + "epoch": 0.520136518771331, + "grad_norm": 2.1767008349124892, + "learning_rate": 1.216947910348819e-06, + "loss": 0.1101, + "step": 2286 + }, + { + "epoch": 0.5203640500568828, + "grad_norm": 2.9882776274660254, + "learning_rate": 1.2169192324016423e-06, + "loss": 0.1401, + "step": 2287 + }, + { + "epoch": 0.5205915813424346, + "grad_norm": 1.8462104806775126, + "learning_rate": 1.2168905423567205e-06, + "loss": 0.2272, + "step": 2288 + }, + { + "epoch": 0.5208191126279863, + "grad_norm": 3.1696493697631962, + "learning_rate": 1.21686184021464e-06, + "loss": 0.296, + "step": 2289 + }, + { + "epoch": 0.5210466439135382, + "grad_norm": 2.0242733158944315, + "learning_rate": 1.2168331259759875e-06, + "loss": 0.1394, + "step": 2290 + }, + { + "epoch": 0.5212741751990899, + "grad_norm": 3.175915917997839, + "learning_rate": 1.2168043996413497e-06, + "loss": 0.1532, + "step": 2291 + }, + { + "epoch": 0.5215017064846417, + "grad_norm": 1.8835065784179783, + "learning_rate": 1.216775661211314e-06, + "loss": 0.1239, + "step": 2292 + }, + { + "epoch": 0.5217292377701934, + "grad_norm": 1.9703941833763796, + "learning_rate": 1.2167469106864673e-06, + "loss": 0.218, + "step": 2293 + }, + { + "epoch": 0.5219567690557452, + "grad_norm": 1.7379669142562126, + "learning_rate": 1.2167181480673977e-06, + "loss": 0.138, + "step": 2294 + }, + { + "epoch": 0.5221843003412969, + "grad_norm": 1.0392802621173225, + "learning_rate": 1.2166893733546927e-06, + "loss": 0.1262, + "step": 2295 + }, + { + "epoch": 0.5224118316268487, + "grad_norm": 1.8936366842951171, + "learning_rate": 1.2166605865489406e-06, + "loss": 0.2228, + "step": 2296 + }, + { + "epoch": 0.5226393629124004, + "grad_norm": 1.6981025004152621, + "learning_rate": 1.2166317876507296e-06, + "loss": 0.1228, + "step": 2297 + }, + { + "epoch": 0.5228668941979522, + "grad_norm": 2.1909663135538695, + "learning_rate": 1.2166029766606486e-06, + "loss": 0.1409, + "step": 2298 + }, + { + "epoch": 0.5230944254835039, + "grad_norm": 2.1158950054230634, + "learning_rate": 1.2165741535792861e-06, + "loss": 0.1606, + "step": 2299 + }, + { + "epoch": 0.5233219567690558, + "grad_norm": 1.4135780574595918, + "learning_rate": 1.2165453184072312e-06, + "loss": 0.1169, + "step": 2300 + }, + { + "epoch": 0.5235494880546075, + "grad_norm": 2.1057935187163737, + "learning_rate": 1.2165164711450735e-06, + "loss": 0.2336, + "step": 2301 + }, + { + "epoch": 0.5237770193401593, + "grad_norm": 2.248987100684698, + "learning_rate": 1.2164876117934024e-06, + "loss": 0.1074, + "step": 2302 + }, + { + "epoch": 0.524004550625711, + "grad_norm": 3.3409227566362385, + "learning_rate": 1.2164587403528078e-06, + "loss": 0.1466, + "step": 2303 + }, + { + "epoch": 0.5242320819112628, + "grad_norm": 2.662759790566142, + "learning_rate": 1.2164298568238797e-06, + "loss": 0.2104, + "step": 2304 + }, + { + "epoch": 0.5244596131968146, + "grad_norm": 1.830035008905779, + "learning_rate": 1.2164009612072085e-06, + "loss": 0.1283, + "step": 2305 + }, + { + "epoch": 0.5246871444823663, + "grad_norm": 1.8990428054871098, + "learning_rate": 1.2163720535033845e-06, + "loss": 0.1127, + "step": 2306 + }, + { + "epoch": 0.5249146757679181, + "grad_norm": 2.9544210428909, + "learning_rate": 1.2163431337129993e-06, + "loss": 0.131, + "step": 2307 + }, + { + "epoch": 0.5251422070534698, + "grad_norm": 1.5069043409936649, + "learning_rate": 1.216314201836643e-06, + "loss": 0.1278, + "step": 2308 + }, + { + "epoch": 0.5253697383390216, + "grad_norm": 3.795761311329223, + "learning_rate": 1.2162852578749076e-06, + "loss": 0.1419, + "step": 2309 + }, + { + "epoch": 0.5255972696245734, + "grad_norm": 2.967399207541332, + "learning_rate": 1.2162563018283843e-06, + "loss": 0.1838, + "step": 2310 + }, + { + "epoch": 0.5258248009101252, + "grad_norm": 1.1243093423898234, + "learning_rate": 1.216227333697665e-06, + "loss": 0.0939, + "step": 2311 + }, + { + "epoch": 0.5260523321956769, + "grad_norm": 1.6963647749955448, + "learning_rate": 1.216198353483342e-06, + "loss": 0.0794, + "step": 2312 + }, + { + "epoch": 0.5262798634812287, + "grad_norm": 2.0805603023570023, + "learning_rate": 1.2161693611860072e-06, + "loss": 0.1097, + "step": 2313 + }, + { + "epoch": 0.5265073947667804, + "grad_norm": 1.7107342546351574, + "learning_rate": 1.2161403568062533e-06, + "loss": 0.0994, + "step": 2314 + }, + { + "epoch": 0.5267349260523322, + "grad_norm": 1.9862157408748995, + "learning_rate": 1.2161113403446733e-06, + "loss": 0.1259, + "step": 2315 + }, + { + "epoch": 0.5269624573378839, + "grad_norm": 2.482118779617471, + "learning_rate": 1.21608231180186e-06, + "loss": 0.1643, + "step": 2316 + }, + { + "epoch": 0.5271899886234357, + "grad_norm": 2.0726950491499423, + "learning_rate": 1.2160532711784066e-06, + "loss": 0.1502, + "step": 2317 + }, + { + "epoch": 0.5274175199089874, + "grad_norm": 1.4883744563296912, + "learning_rate": 1.2160242184749069e-06, + "loss": 0.1251, + "step": 2318 + }, + { + "epoch": 0.5276450511945392, + "grad_norm": 2.304698494036364, + "learning_rate": 1.2159951536919547e-06, + "loss": 0.1392, + "step": 2319 + }, + { + "epoch": 0.5278725824800911, + "grad_norm": 3.8449144926935035, + "learning_rate": 1.2159660768301438e-06, + "loss": 0.1403, + "step": 2320 + }, + { + "epoch": 0.5281001137656428, + "grad_norm": 2.9094583253548225, + "learning_rate": 1.2159369878900687e-06, + "loss": 0.129, + "step": 2321 + }, + { + "epoch": 0.5283276450511946, + "grad_norm": 4.3378705361683, + "learning_rate": 1.2159078868723238e-06, + "loss": 0.2085, + "step": 2322 + }, + { + "epoch": 0.5285551763367463, + "grad_norm": 1.7024973102380594, + "learning_rate": 1.2158787737775037e-06, + "loss": 0.1317, + "step": 2323 + }, + { + "epoch": 0.5287827076222981, + "grad_norm": 1.3655076255928897, + "learning_rate": 1.2158496486062039e-06, + "loss": 0.1327, + "step": 2324 + }, + { + "epoch": 0.5290102389078498, + "grad_norm": 2.8976980348396433, + "learning_rate": 1.215820511359019e-06, + "loss": 0.1634, + "step": 2325 + }, + { + "epoch": 0.5292377701934016, + "grad_norm": 3.3802109886132867, + "learning_rate": 1.215791362036545e-06, + "loss": 0.1762, + "step": 2326 + }, + { + "epoch": 0.5294653014789533, + "grad_norm": 1.8540359020863921, + "learning_rate": 1.2157622006393777e-06, + "loss": 0.1649, + "step": 2327 + }, + { + "epoch": 0.5296928327645051, + "grad_norm": 2.841457854773911, + "learning_rate": 1.2157330271681129e-06, + "loss": 0.1474, + "step": 2328 + }, + { + "epoch": 0.5299203640500569, + "grad_norm": 4.1952377999594, + "learning_rate": 1.215703841623347e-06, + "loss": 0.1585, + "step": 2329 + }, + { + "epoch": 0.5301478953356087, + "grad_norm": 1.4873684025860323, + "learning_rate": 1.2156746440056762e-06, + "loss": 0.1165, + "step": 2330 + }, + { + "epoch": 0.5303754266211604, + "grad_norm": 2.5751493361989657, + "learning_rate": 1.2156454343156976e-06, + "loss": 0.1522, + "step": 2331 + }, + { + "epoch": 0.5306029579067122, + "grad_norm": 1.3229989088243255, + "learning_rate": 1.2156162125540078e-06, + "loss": 0.061, + "step": 2332 + }, + { + "epoch": 0.5308304891922639, + "grad_norm": 1.0015280999927394, + "learning_rate": 1.2155869787212046e-06, + "loss": 0.0875, + "step": 2333 + }, + { + "epoch": 0.5310580204778157, + "grad_norm": 1.7236628455878646, + "learning_rate": 1.215557732817885e-06, + "loss": 0.1357, + "step": 2334 + }, + { + "epoch": 0.5312855517633674, + "grad_norm": 2.143419078889929, + "learning_rate": 1.2155284748446469e-06, + "loss": 0.1297, + "step": 2335 + }, + { + "epoch": 0.5315130830489192, + "grad_norm": 1.6815351249254196, + "learning_rate": 1.2154992048020882e-06, + "loss": 0.1414, + "step": 2336 + }, + { + "epoch": 0.531740614334471, + "grad_norm": 2.133575319364267, + "learning_rate": 1.2154699226908072e-06, + "loss": 0.1321, + "step": 2337 + }, + { + "epoch": 0.5319681456200227, + "grad_norm": 1.139669924038715, + "learning_rate": 1.2154406285114025e-06, + "loss": 0.0903, + "step": 2338 + }, + { + "epoch": 0.5321956769055746, + "grad_norm": 1.7837908668506595, + "learning_rate": 1.2154113222644727e-06, + "loss": 0.1191, + "step": 2339 + }, + { + "epoch": 0.5324232081911263, + "grad_norm": 4.103642831661849, + "learning_rate": 1.2153820039506167e-06, + "loss": 0.1628, + "step": 2340 + }, + { + "epoch": 0.5326507394766781, + "grad_norm": 1.6530276808213893, + "learning_rate": 1.2153526735704337e-06, + "loss": 0.1123, + "step": 2341 + }, + { + "epoch": 0.5328782707622298, + "grad_norm": 1.6519117615379917, + "learning_rate": 1.2153233311245234e-06, + "loss": 0.1547, + "step": 2342 + }, + { + "epoch": 0.5331058020477816, + "grad_norm": 1.8534786765494804, + "learning_rate": 1.2152939766134852e-06, + "loss": 0.1322, + "step": 2343 + }, + { + "epoch": 0.5333333333333333, + "grad_norm": 2.0466229249542747, + "learning_rate": 1.2152646100379193e-06, + "loss": 0.1735, + "step": 2344 + }, + { + "epoch": 0.5335608646188851, + "grad_norm": 1.7038078681513562, + "learning_rate": 1.2152352313984257e-06, + "loss": 0.0974, + "step": 2345 + }, + { + "epoch": 0.5337883959044368, + "grad_norm": 2.924087638349447, + "learning_rate": 1.2152058406956049e-06, + "loss": 0.1973, + "step": 2346 + }, + { + "epoch": 0.5340159271899886, + "grad_norm": 2.7275068217190723, + "learning_rate": 1.2151764379300578e-06, + "loss": 0.1237, + "step": 2347 + }, + { + "epoch": 0.5342434584755403, + "grad_norm": 1.424984124254766, + "learning_rate": 1.2151470231023851e-06, + "loss": 0.1238, + "step": 2348 + }, + { + "epoch": 0.5344709897610922, + "grad_norm": 1.5436354239233119, + "learning_rate": 1.2151175962131881e-06, + "loss": 0.1482, + "step": 2349 + }, + { + "epoch": 0.534698521046644, + "grad_norm": 1.3973484809650978, + "learning_rate": 1.215088157263068e-06, + "loss": 0.08, + "step": 2350 + }, + { + "epoch": 0.5349260523321957, + "grad_norm": 1.0011406880829632, + "learning_rate": 1.2150587062526267e-06, + "loss": 0.0743, + "step": 2351 + }, + { + "epoch": 0.5351535836177475, + "grad_norm": 1.681422253336755, + "learning_rate": 1.2150292431824663e-06, + "loss": 0.2121, + "step": 2352 + }, + { + "epoch": 0.5353811149032992, + "grad_norm": 2.071137301181157, + "learning_rate": 1.2149997680531886e-06, + "loss": 0.1329, + "step": 2353 + }, + { + "epoch": 0.535608646188851, + "grad_norm": 2.7731157561747932, + "learning_rate": 1.214970280865396e-06, + "loss": 0.1483, + "step": 2354 + }, + { + "epoch": 0.5358361774744027, + "grad_norm": 1.825127983213737, + "learning_rate": 1.2149407816196917e-06, + "loss": 0.2009, + "step": 2355 + }, + { + "epoch": 0.5360637087599545, + "grad_norm": 1.5271969230637406, + "learning_rate": 1.214911270316678e-06, + "loss": 0.109, + "step": 2356 + }, + { + "epoch": 0.5362912400455062, + "grad_norm": 1.1546891312653111, + "learning_rate": 1.2148817469569584e-06, + "loss": 0.152, + "step": 2357 + }, + { + "epoch": 0.5365187713310581, + "grad_norm": 1.201446827839472, + "learning_rate": 1.214852211541136e-06, + "loss": 0.0874, + "step": 2358 + }, + { + "epoch": 0.5367463026166098, + "grad_norm": 1.5029157511372113, + "learning_rate": 1.2148226640698148e-06, + "loss": 0.0655, + "step": 2359 + }, + { + "epoch": 0.5369738339021616, + "grad_norm": 1.3717547829576855, + "learning_rate": 1.2147931045435988e-06, + "loss": 0.1699, + "step": 2360 + }, + { + "epoch": 0.5372013651877133, + "grad_norm": 1.3834832471577652, + "learning_rate": 1.2147635329630916e-06, + "loss": 0.1904, + "step": 2361 + }, + { + "epoch": 0.5374288964732651, + "grad_norm": 2.35690597522568, + "learning_rate": 1.214733949328898e-06, + "loss": 0.1102, + "step": 2362 + }, + { + "epoch": 0.5376564277588168, + "grad_norm": 2.0463590798211033, + "learning_rate": 1.2147043536416226e-06, + "loss": 0.1773, + "step": 2363 + }, + { + "epoch": 0.5378839590443686, + "grad_norm": 2.5222590318625384, + "learning_rate": 1.21467474590187e-06, + "loss": 0.2526, + "step": 2364 + }, + { + "epoch": 0.5381114903299203, + "grad_norm": 3.296081575537043, + "learning_rate": 1.2146451261102458e-06, + "loss": 0.1592, + "step": 2365 + }, + { + "epoch": 0.5383390216154721, + "grad_norm": 2.6706214296804336, + "learning_rate": 1.2146154942673548e-06, + "loss": 0.1196, + "step": 2366 + }, + { + "epoch": 0.5385665529010238, + "grad_norm": 2.9249686559098795, + "learning_rate": 1.2145858503738032e-06, + "loss": 0.1877, + "step": 2367 + }, + { + "epoch": 0.5387940841865757, + "grad_norm": 2.2953281999149593, + "learning_rate": 1.2145561944301963e-06, + "loss": 0.1972, + "step": 2368 + }, + { + "epoch": 0.5390216154721275, + "grad_norm": 1.4636027746143558, + "learning_rate": 1.2145265264371406e-06, + "loss": 0.162, + "step": 2369 + }, + { + "epoch": 0.5392491467576792, + "grad_norm": 2.161526587594214, + "learning_rate": 1.2144968463952425e-06, + "loss": 0.0907, + "step": 2370 + }, + { + "epoch": 0.539476678043231, + "grad_norm": 2.5292055598654875, + "learning_rate": 1.2144671543051085e-06, + "loss": 0.1697, + "step": 2371 + }, + { + "epoch": 0.5397042093287827, + "grad_norm": 1.129436740546767, + "learning_rate": 1.2144374501673454e-06, + "loss": 0.0512, + "step": 2372 + }, + { + "epoch": 0.5399317406143345, + "grad_norm": 3.143413485445767, + "learning_rate": 1.2144077339825603e-06, + "loss": 0.2778, + "step": 2373 + }, + { + "epoch": 0.5401592718998862, + "grad_norm": 1.9332739586254322, + "learning_rate": 1.2143780057513605e-06, + "loss": 0.1697, + "step": 2374 + }, + { + "epoch": 0.540386803185438, + "grad_norm": 2.328675689936947, + "learning_rate": 1.2143482654743535e-06, + "loss": 0.1321, + "step": 2375 + }, + { + "epoch": 0.5406143344709897, + "grad_norm": 1.2157576056753703, + "learning_rate": 1.2143185131521475e-06, + "loss": 0.1111, + "step": 2376 + }, + { + "epoch": 0.5408418657565415, + "grad_norm": 2.407893236147383, + "learning_rate": 1.2142887487853503e-06, + "loss": 0.1563, + "step": 2377 + }, + { + "epoch": 0.5410693970420933, + "grad_norm": 2.9189820741426376, + "learning_rate": 1.2142589723745705e-06, + "loss": 0.1164, + "step": 2378 + }, + { + "epoch": 0.5412969283276451, + "grad_norm": 2.396152562102295, + "learning_rate": 1.2142291839204163e-06, + "loss": 0.1382, + "step": 2379 + }, + { + "epoch": 0.5415244596131968, + "grad_norm": 1.2309543853665252, + "learning_rate": 1.2141993834234967e-06, + "loss": 0.1599, + "step": 2380 + }, + { + "epoch": 0.5417519908987486, + "grad_norm": 1.4815415846583502, + "learning_rate": 1.2141695708844209e-06, + "loss": 0.0808, + "step": 2381 + }, + { + "epoch": 0.5419795221843003, + "grad_norm": 1.7440274898378374, + "learning_rate": 1.214139746303798e-06, + "loss": 0.1357, + "step": 2382 + }, + { + "epoch": 0.5422070534698521, + "grad_norm": 2.501198944799206, + "learning_rate": 1.2141099096822376e-06, + "loss": 0.1619, + "step": 2383 + }, + { + "epoch": 0.5424345847554038, + "grad_norm": 1.830304404422833, + "learning_rate": 1.2140800610203497e-06, + "loss": 0.225, + "step": 2384 + }, + { + "epoch": 0.5426621160409556, + "grad_norm": 1.997345142772124, + "learning_rate": 1.214050200318744e-06, + "loss": 0.1129, + "step": 2385 + }, + { + "epoch": 0.5428896473265074, + "grad_norm": 2.6541224516196724, + "learning_rate": 1.214020327578031e-06, + "loss": 0.1106, + "step": 2386 + }, + { + "epoch": 0.5431171786120591, + "grad_norm": 1.3841215073352757, + "learning_rate": 1.2139904427988213e-06, + "loss": 0.0714, + "step": 2387 + }, + { + "epoch": 0.543344709897611, + "grad_norm": 1.3495168470574121, + "learning_rate": 1.2139605459817259e-06, + "loss": 0.1468, + "step": 2388 + }, + { + "epoch": 0.5435722411831627, + "grad_norm": 1.2198752931302614, + "learning_rate": 1.2139306371273552e-06, + "loss": 0.1431, + "step": 2389 + }, + { + "epoch": 0.5437997724687145, + "grad_norm": 1.685309275084401, + "learning_rate": 1.213900716236321e-06, + "loss": 0.0955, + "step": 2390 + }, + { + "epoch": 0.5440273037542662, + "grad_norm": 1.9310383949097503, + "learning_rate": 1.2138707833092345e-06, + "loss": 0.1583, + "step": 2391 + }, + { + "epoch": 0.544254835039818, + "grad_norm": 1.3819179331884706, + "learning_rate": 1.213840838346708e-06, + "loss": 0.145, + "step": 2392 + }, + { + "epoch": 0.5444823663253697, + "grad_norm": 1.9093745108266076, + "learning_rate": 1.213810881349353e-06, + "loss": 0.1082, + "step": 2393 + }, + { + "epoch": 0.5447098976109215, + "grad_norm": 4.561592933421953, + "learning_rate": 1.213780912317782e-06, + "loss": 0.1692, + "step": 2394 + }, + { + "epoch": 0.5449374288964732, + "grad_norm": 2.0605709034236868, + "learning_rate": 1.2137509312526074e-06, + "loss": 0.1049, + "step": 2395 + }, + { + "epoch": 0.545164960182025, + "grad_norm": 1.4601179535219537, + "learning_rate": 1.213720938154442e-06, + "loss": 0.1286, + "step": 2396 + }, + { + "epoch": 0.5453924914675768, + "grad_norm": 2.3642573322951894, + "learning_rate": 1.2136909330238988e-06, + "loss": 0.1955, + "step": 2397 + }, + { + "epoch": 0.5456200227531286, + "grad_norm": 2.2512568505689403, + "learning_rate": 1.213660915861591e-06, + "loss": 0.2434, + "step": 2398 + }, + { + "epoch": 0.5458475540386803, + "grad_norm": 1.992478458605465, + "learning_rate": 1.2136308866681323e-06, + "loss": 0.1372, + "step": 2399 + }, + { + "epoch": 0.5460750853242321, + "grad_norm": 1.8111222939771605, + "learning_rate": 1.2136008454441363e-06, + "loss": 0.197, + "step": 2400 + }, + { + "epoch": 0.5463026166097839, + "grad_norm": 1.3777489796507147, + "learning_rate": 1.213570792190217e-06, + "loss": 0.0908, + "step": 2401 + }, + { + "epoch": 0.5465301478953356, + "grad_norm": 2.6063902777425105, + "learning_rate": 1.2135407269069885e-06, + "loss": 0.1692, + "step": 2402 + }, + { + "epoch": 0.5467576791808874, + "grad_norm": 1.9011415061442813, + "learning_rate": 1.2135106495950655e-06, + "loss": 0.0622, + "step": 2403 + }, + { + "epoch": 0.5469852104664391, + "grad_norm": 2.3440439149150287, + "learning_rate": 1.2134805602550625e-06, + "loss": 0.1539, + "step": 2404 + }, + { + "epoch": 0.5472127417519909, + "grad_norm": 2.50409667066478, + "learning_rate": 1.2134504588875948e-06, + "loss": 0.1633, + "step": 2405 + }, + { + "epoch": 0.5474402730375426, + "grad_norm": 3.3789456844433285, + "learning_rate": 1.2134203454932772e-06, + "loss": 0.2344, + "step": 2406 + }, + { + "epoch": 0.5476678043230945, + "grad_norm": 2.1006436694907333, + "learning_rate": 1.2133902200727256e-06, + "loss": 0.1003, + "step": 2407 + }, + { + "epoch": 0.5478953356086462, + "grad_norm": 2.1646662277721638, + "learning_rate": 1.2133600826265555e-06, + "loss": 0.1144, + "step": 2408 + }, + { + "epoch": 0.548122866894198, + "grad_norm": 2.48456294214119, + "learning_rate": 1.2133299331553826e-06, + "loss": 0.0843, + "step": 2409 + }, + { + "epoch": 0.5483503981797497, + "grad_norm": 2.445609950491129, + "learning_rate": 1.2132997716598236e-06, + "loss": 0.1332, + "step": 2410 + }, + { + "epoch": 0.5485779294653015, + "grad_norm": 1.607891388562304, + "learning_rate": 1.2132695981404943e-06, + "loss": 0.0809, + "step": 2411 + }, + { + "epoch": 0.5488054607508532, + "grad_norm": 3.0954255254232175, + "learning_rate": 1.2132394125980122e-06, + "loss": 0.1231, + "step": 2412 + }, + { + "epoch": 0.549032992036405, + "grad_norm": 1.6018022170636927, + "learning_rate": 1.2132092150329936e-06, + "loss": 0.1073, + "step": 2413 + }, + { + "epoch": 0.5492605233219567, + "grad_norm": 3.405787206112886, + "learning_rate": 1.213179005446056e-06, + "loss": 0.1619, + "step": 2414 + }, + { + "epoch": 0.5494880546075085, + "grad_norm": 1.6685361812270334, + "learning_rate": 1.2131487838378167e-06, + "loss": 0.0928, + "step": 2415 + }, + { + "epoch": 0.5497155858930602, + "grad_norm": 2.2102317570381405, + "learning_rate": 1.2131185502088932e-06, + "loss": 0.0894, + "step": 2416 + }, + { + "epoch": 0.5499431171786121, + "grad_norm": 2.796381970818353, + "learning_rate": 1.213088304559904e-06, + "loss": 0.1034, + "step": 2417 + }, + { + "epoch": 0.5501706484641639, + "grad_norm": 2.7540797302904996, + "learning_rate": 1.2130580468914665e-06, + "loss": 0.1926, + "step": 2418 + }, + { + "epoch": 0.5503981797497156, + "grad_norm": 2.1184433609657303, + "learning_rate": 1.2130277772041999e-06, + "loss": 0.1385, + "step": 2419 + }, + { + "epoch": 0.5506257110352674, + "grad_norm": 1.5929408368756812, + "learning_rate": 1.2129974954987222e-06, + "loss": 0.0825, + "step": 2420 + }, + { + "epoch": 0.5508532423208191, + "grad_norm": 1.7635977427083063, + "learning_rate": 1.2129672017756524e-06, + "loss": 0.1236, + "step": 2421 + }, + { + "epoch": 0.5510807736063709, + "grad_norm": 1.71615894112007, + "learning_rate": 1.2129368960356102e-06, + "loss": 0.1382, + "step": 2422 + }, + { + "epoch": 0.5513083048919226, + "grad_norm": 1.937244032383735, + "learning_rate": 1.2129065782792142e-06, + "loss": 0.1291, + "step": 2423 + }, + { + "epoch": 0.5515358361774744, + "grad_norm": 3.1352304902387385, + "learning_rate": 1.2128762485070848e-06, + "loss": 0.1544, + "step": 2424 + }, + { + "epoch": 0.5517633674630261, + "grad_norm": 2.1008217911515046, + "learning_rate": 1.2128459067198414e-06, + "loss": 0.1314, + "step": 2425 + }, + { + "epoch": 0.5519908987485779, + "grad_norm": 4.964728468745166, + "learning_rate": 1.2128155529181042e-06, + "loss": 0.1358, + "step": 2426 + }, + { + "epoch": 0.5522184300341297, + "grad_norm": 2.8349612253194088, + "learning_rate": 1.2127851871024937e-06, + "loss": 0.2205, + "step": 2427 + }, + { + "epoch": 0.5524459613196815, + "grad_norm": 2.389556505433841, + "learning_rate": 1.2127548092736305e-06, + "loss": 0.1562, + "step": 2428 + }, + { + "epoch": 0.5526734926052332, + "grad_norm": 1.9336117163609754, + "learning_rate": 1.2127244194321353e-06, + "loss": 0.1665, + "step": 2429 + }, + { + "epoch": 0.552901023890785, + "grad_norm": 2.3714953779733223, + "learning_rate": 1.2126940175786294e-06, + "loss": 0.1747, + "step": 2430 + }, + { + "epoch": 0.5531285551763367, + "grad_norm": 1.6953148543099945, + "learning_rate": 1.212663603713734e-06, + "loss": 0.0945, + "step": 2431 + }, + { + "epoch": 0.5533560864618885, + "grad_norm": 1.737943346359041, + "learning_rate": 1.212633177838071e-06, + "loss": 0.1451, + "step": 2432 + }, + { + "epoch": 0.5535836177474402, + "grad_norm": 2.211570479238446, + "learning_rate": 1.2126027399522617e-06, + "loss": 0.1064, + "step": 2433 + }, + { + "epoch": 0.553811149032992, + "grad_norm": 1.7857339852544778, + "learning_rate": 1.2125722900569288e-06, + "loss": 0.1806, + "step": 2434 + }, + { + "epoch": 0.5540386803185438, + "grad_norm": 3.0559649821603245, + "learning_rate": 1.2125418281526944e-06, + "loss": 0.2878, + "step": 2435 + }, + { + "epoch": 0.5542662116040956, + "grad_norm": 3.234604579214076, + "learning_rate": 1.212511354240181e-06, + "loss": 0.1531, + "step": 2436 + }, + { + "epoch": 0.5544937428896474, + "grad_norm": 2.0773489813309967, + "learning_rate": 1.2124808683200113e-06, + "loss": 0.1326, + "step": 2437 + }, + { + "epoch": 0.5547212741751991, + "grad_norm": 1.2340445485090055, + "learning_rate": 1.2124503703928088e-06, + "loss": 0.1116, + "step": 2438 + }, + { + "epoch": 0.5549488054607509, + "grad_norm": 3.5232591527626824, + "learning_rate": 1.2124198604591965e-06, + "loss": 0.1686, + "step": 2439 + }, + { + "epoch": 0.5551763367463026, + "grad_norm": 2.8726622730449503, + "learning_rate": 1.2123893385197982e-06, + "loss": 0.1791, + "step": 2440 + }, + { + "epoch": 0.5554038680318544, + "grad_norm": 2.091853211246227, + "learning_rate": 1.2123588045752373e-06, + "loss": 0.1481, + "step": 2441 + }, + { + "epoch": 0.5556313993174061, + "grad_norm": 2.3690335770551294, + "learning_rate": 1.2123282586261384e-06, + "loss": 0.1708, + "step": 2442 + }, + { + "epoch": 0.5558589306029579, + "grad_norm": 2.233061824072559, + "learning_rate": 1.2122977006731256e-06, + "loss": 0.1571, + "step": 2443 + }, + { + "epoch": 0.5560864618885096, + "grad_norm": 1.9900055676960977, + "learning_rate": 1.2122671307168232e-06, + "loss": 0.0985, + "step": 2444 + }, + { + "epoch": 0.5563139931740614, + "grad_norm": 1.669304123397253, + "learning_rate": 1.212236548757856e-06, + "loss": 0.2418, + "step": 2445 + }, + { + "epoch": 0.5565415244596132, + "grad_norm": 2.0839688820404247, + "learning_rate": 1.2122059547968496e-06, + "loss": 0.11, + "step": 2446 + }, + { + "epoch": 0.556769055745165, + "grad_norm": 2.3322979495915543, + "learning_rate": 1.2121753488344286e-06, + "loss": 0.1758, + "step": 2447 + }, + { + "epoch": 0.5569965870307167, + "grad_norm": 3.079512271260833, + "learning_rate": 1.212144730871219e-06, + "loss": 0.1408, + "step": 2448 + }, + { + "epoch": 0.5572241183162685, + "grad_norm": 3.057705857554608, + "learning_rate": 1.2121141009078462e-06, + "loss": 0.1547, + "step": 2449 + }, + { + "epoch": 0.5574516496018203, + "grad_norm": 2.754720342311703, + "learning_rate": 1.2120834589449368e-06, + "loss": 0.1227, + "step": 2450 + }, + { + "epoch": 0.557679180887372, + "grad_norm": 2.0039807911619696, + "learning_rate": 1.2120528049831165e-06, + "loss": 0.1627, + "step": 2451 + }, + { + "epoch": 0.5579067121729238, + "grad_norm": 2.152939979711796, + "learning_rate": 1.212022139023012e-06, + "loss": 0.224, + "step": 2452 + }, + { + "epoch": 0.5581342434584755, + "grad_norm": 3.2411118966017964, + "learning_rate": 1.21199146106525e-06, + "loss": 0.1193, + "step": 2453 + }, + { + "epoch": 0.5583617747440273, + "grad_norm": 1.5945262516296992, + "learning_rate": 1.2119607711104574e-06, + "loss": 0.1496, + "step": 2454 + }, + { + "epoch": 0.558589306029579, + "grad_norm": 1.5388366931624589, + "learning_rate": 1.211930069159262e-06, + "loss": 0.1171, + "step": 2455 + }, + { + "epoch": 0.5588168373151309, + "grad_norm": 1.4040745534326453, + "learning_rate": 1.2118993552122907e-06, + "loss": 0.13, + "step": 2456 + }, + { + "epoch": 0.5590443686006826, + "grad_norm": 1.4683917259528867, + "learning_rate": 1.2118686292701715e-06, + "loss": 0.0933, + "step": 2457 + }, + { + "epoch": 0.5592718998862344, + "grad_norm": 2.3779110713148492, + "learning_rate": 1.211837891333532e-06, + "loss": 0.131, + "step": 2458 + }, + { + "epoch": 0.5594994311717861, + "grad_norm": 1.9444812735417465, + "learning_rate": 1.211807141403001e-06, + "loss": 0.197, + "step": 2459 + }, + { + "epoch": 0.5597269624573379, + "grad_norm": 1.7785495979173072, + "learning_rate": 1.211776379479207e-06, + "loss": 0.0819, + "step": 2460 + }, + { + "epoch": 0.5599544937428896, + "grad_norm": 2.8186861224308446, + "learning_rate": 1.211745605562778e-06, + "loss": 0.1185, + "step": 2461 + }, + { + "epoch": 0.5601820250284414, + "grad_norm": 1.36323819201261, + "learning_rate": 1.2117148196543436e-06, + "loss": 0.1724, + "step": 2462 + }, + { + "epoch": 0.5604095563139931, + "grad_norm": 4.03855027641924, + "learning_rate": 1.2116840217545329e-06, + "loss": 0.2283, + "step": 2463 + }, + { + "epoch": 0.5606370875995449, + "grad_norm": 2.123596815339297, + "learning_rate": 1.211653211863975e-06, + "loss": 0.163, + "step": 2464 + }, + { + "epoch": 0.5608646188850968, + "grad_norm": 2.591456710871039, + "learning_rate": 1.2116223899833e-06, + "loss": 0.1665, + "step": 2465 + }, + { + "epoch": 0.5610921501706485, + "grad_norm": 2.277844579413468, + "learning_rate": 1.2115915561131376e-06, + "loss": 0.1107, + "step": 2466 + }, + { + "epoch": 0.5613196814562003, + "grad_norm": 2.6316966240447526, + "learning_rate": 1.211560710254118e-06, + "loss": 0.1288, + "step": 2467 + }, + { + "epoch": 0.561547212741752, + "grad_norm": 2.2537204656115493, + "learning_rate": 1.211529852406872e-06, + "loss": 0.1563, + "step": 2468 + }, + { + "epoch": 0.5617747440273038, + "grad_norm": 2.1713954845528676, + "learning_rate": 1.2114989825720298e-06, + "loss": 0.1442, + "step": 2469 + }, + { + "epoch": 0.5620022753128555, + "grad_norm": 2.693618014034486, + "learning_rate": 1.2114681007502227e-06, + "loss": 0.132, + "step": 2470 + }, + { + "epoch": 0.5622298065984073, + "grad_norm": 1.342955875247325, + "learning_rate": 1.2114372069420815e-06, + "loss": 0.1661, + "step": 2471 + }, + { + "epoch": 0.562457337883959, + "grad_norm": 2.295134838913565, + "learning_rate": 1.2114063011482378e-06, + "loss": 0.1068, + "step": 2472 + }, + { + "epoch": 0.5626848691695108, + "grad_norm": 2.2269684435425594, + "learning_rate": 1.2113753833693234e-06, + "loss": 0.0884, + "step": 2473 + }, + { + "epoch": 0.5629124004550625, + "grad_norm": 1.8251560760537473, + "learning_rate": 1.2113444536059699e-06, + "loss": 0.1569, + "step": 2474 + }, + { + "epoch": 0.5631399317406144, + "grad_norm": 1.2373817468358461, + "learning_rate": 1.2113135118588096e-06, + "loss": 0.1253, + "step": 2475 + }, + { + "epoch": 0.5633674630261661, + "grad_norm": 1.5518735019134444, + "learning_rate": 1.2112825581284752e-06, + "loss": 0.1655, + "step": 2476 + }, + { + "epoch": 0.5635949943117179, + "grad_norm": 2.821927704907384, + "learning_rate": 1.2112515924155987e-06, + "loss": 0.1968, + "step": 2477 + }, + { + "epoch": 0.5638225255972696, + "grad_norm": 2.6034604032683424, + "learning_rate": 1.2112206147208134e-06, + "loss": 0.1579, + "step": 2478 + }, + { + "epoch": 0.5640500568828214, + "grad_norm": 3.005350448867513, + "learning_rate": 1.2111896250447525e-06, + "loss": 0.1824, + "step": 2479 + }, + { + "epoch": 0.5642775881683731, + "grad_norm": 2.0319627346292846, + "learning_rate": 1.211158623388049e-06, + "loss": 0.1139, + "step": 2480 + }, + { + "epoch": 0.5645051194539249, + "grad_norm": 1.6717615590204162, + "learning_rate": 1.2111276097513369e-06, + "loss": 0.1229, + "step": 2481 + }, + { + "epoch": 0.5647326507394766, + "grad_norm": 1.0856506812146736, + "learning_rate": 1.2110965841352498e-06, + "loss": 0.1326, + "step": 2482 + }, + { + "epoch": 0.5649601820250284, + "grad_norm": 2.1610866648004623, + "learning_rate": 1.211065546540422e-06, + "loss": 0.1317, + "step": 2483 + }, + { + "epoch": 0.5651877133105802, + "grad_norm": 1.6835926280266125, + "learning_rate": 1.2110344969674877e-06, + "loss": 0.0876, + "step": 2484 + }, + { + "epoch": 0.565415244596132, + "grad_norm": 1.8868431281124944, + "learning_rate": 1.2110034354170816e-06, + "loss": 0.2136, + "step": 2485 + }, + { + "epoch": 0.5656427758816838, + "grad_norm": 1.2849513459232036, + "learning_rate": 1.2109723618898383e-06, + "loss": 0.1176, + "step": 2486 + }, + { + "epoch": 0.5658703071672355, + "grad_norm": 2.225412733583598, + "learning_rate": 1.2109412763863933e-06, + "loss": 0.2349, + "step": 2487 + }, + { + "epoch": 0.5660978384527873, + "grad_norm": 1.7394007182133215, + "learning_rate": 1.2109101789073815e-06, + "loss": 0.1889, + "step": 2488 + }, + { + "epoch": 0.566325369738339, + "grad_norm": 1.0417756307418755, + "learning_rate": 1.2108790694534389e-06, + "loss": 0.1295, + "step": 2489 + }, + { + "epoch": 0.5665529010238908, + "grad_norm": 2.8490023502098913, + "learning_rate": 1.2108479480252011e-06, + "loss": 0.2094, + "step": 2490 + }, + { + "epoch": 0.5667804323094425, + "grad_norm": 1.4969528469131224, + "learning_rate": 1.210816814623304e-06, + "loss": 0.189, + "step": 2491 + }, + { + "epoch": 0.5670079635949943, + "grad_norm": 2.272829112716015, + "learning_rate": 1.2107856692483843e-06, + "loss": 0.1269, + "step": 2492 + }, + { + "epoch": 0.567235494880546, + "grad_norm": 2.108513421489345, + "learning_rate": 1.2107545119010783e-06, + "loss": 0.1345, + "step": 2493 + }, + { + "epoch": 0.5674630261660978, + "grad_norm": 2.201970892175689, + "learning_rate": 1.2107233425820229e-06, + "loss": 0.225, + "step": 2494 + }, + { + "epoch": 0.5676905574516496, + "grad_norm": 3.4046924059583548, + "learning_rate": 1.2106921612918549e-06, + "loss": 0.0863, + "step": 2495 + }, + { + "epoch": 0.5679180887372014, + "grad_norm": 2.18740259779597, + "learning_rate": 1.2106609680312117e-06, + "loss": 0.1988, + "step": 2496 + }, + { + "epoch": 0.5681456200227532, + "grad_norm": 1.5283955361497255, + "learning_rate": 1.210629762800731e-06, + "loss": 0.1219, + "step": 2497 + }, + { + "epoch": 0.5683731513083049, + "grad_norm": 1.3678120343081321, + "learning_rate": 1.2105985456010506e-06, + "loss": 0.0836, + "step": 2498 + }, + { + "epoch": 0.5686006825938567, + "grad_norm": 1.8601115221660813, + "learning_rate": 1.2105673164328081e-06, + "loss": 0.1723, + "step": 2499 + }, + { + "epoch": 0.5688282138794084, + "grad_norm": 1.7842260253011806, + "learning_rate": 1.2105360752966424e-06, + "loss": 0.118, + "step": 2500 + }, + { + "epoch": 0.5690557451649602, + "grad_norm": 2.221443756494453, + "learning_rate": 1.2105048221931915e-06, + "loss": 0.109, + "step": 2501 + }, + { + "epoch": 0.5692832764505119, + "grad_norm": 1.961751419767975, + "learning_rate": 1.2104735571230944e-06, + "loss": 0.0913, + "step": 2502 + }, + { + "epoch": 0.5695108077360637, + "grad_norm": 1.8181307632990853, + "learning_rate": 1.21044228008699e-06, + "loss": 0.1361, + "step": 2503 + }, + { + "epoch": 0.5697383390216155, + "grad_norm": 1.9118720943978658, + "learning_rate": 1.2104109910855176e-06, + "loss": 0.1627, + "step": 2504 + }, + { + "epoch": 0.5699658703071673, + "grad_norm": 1.5414220442454112, + "learning_rate": 1.2103796901193166e-06, + "loss": 0.1813, + "step": 2505 + }, + { + "epoch": 0.570193401592719, + "grad_norm": 2.8373335682103336, + "learning_rate": 1.210348377189027e-06, + "loss": 0.1289, + "step": 2506 + }, + { + "epoch": 0.5704209328782708, + "grad_norm": 3.083061389684008, + "learning_rate": 1.2103170522952885e-06, + "loss": 0.2029, + "step": 2507 + }, + { + "epoch": 0.5706484641638225, + "grad_norm": 1.8935818147405985, + "learning_rate": 1.2102857154387413e-06, + "loss": 0.1268, + "step": 2508 + }, + { + "epoch": 0.5708759954493743, + "grad_norm": 2.1706672948764685, + "learning_rate": 1.210254366620026e-06, + "loss": 0.146, + "step": 2509 + }, + { + "epoch": 0.571103526734926, + "grad_norm": 1.779554233102803, + "learning_rate": 1.2102230058397832e-06, + "loss": 0.1588, + "step": 2510 + }, + { + "epoch": 0.5713310580204778, + "grad_norm": 1.8784546833055895, + "learning_rate": 1.210191633098654e-06, + "loss": 0.1458, + "step": 2511 + }, + { + "epoch": 0.5715585893060295, + "grad_norm": 1.8629674419414763, + "learning_rate": 1.2101602483972797e-06, + "loss": 0.1051, + "step": 2512 + }, + { + "epoch": 0.5717861205915813, + "grad_norm": 2.262837973386215, + "learning_rate": 1.2101288517363016e-06, + "loss": 0.0856, + "step": 2513 + }, + { + "epoch": 0.5720136518771332, + "grad_norm": 1.123397475538139, + "learning_rate": 1.2100974431163614e-06, + "loss": 0.0578, + "step": 2514 + }, + { + "epoch": 0.5722411831626849, + "grad_norm": 2.735408428715252, + "learning_rate": 1.2100660225381008e-06, + "loss": 0.112, + "step": 2515 + }, + { + "epoch": 0.5724687144482367, + "grad_norm": 1.5475871665351315, + "learning_rate": 1.2100345900021624e-06, + "loss": 0.1233, + "step": 2516 + }, + { + "epoch": 0.5726962457337884, + "grad_norm": 2.598039263582099, + "learning_rate": 1.2100031455091883e-06, + "loss": 0.2105, + "step": 2517 + }, + { + "epoch": 0.5729237770193402, + "grad_norm": 1.5283635000190452, + "learning_rate": 1.2099716890598212e-06, + "loss": 0.1105, + "step": 2518 + }, + { + "epoch": 0.5731513083048919, + "grad_norm": 2.3919998558423092, + "learning_rate": 1.2099402206547042e-06, + "loss": 0.1701, + "step": 2519 + }, + { + "epoch": 0.5733788395904437, + "grad_norm": 1.704386974569275, + "learning_rate": 1.2099087402944805e-06, + "loss": 0.1535, + "step": 2520 + }, + { + "epoch": 0.5736063708759954, + "grad_norm": 2.1761169232651105, + "learning_rate": 1.2098772479797933e-06, + "loss": 0.0741, + "step": 2521 + }, + { + "epoch": 0.5738339021615472, + "grad_norm": 2.2691369150088305, + "learning_rate": 1.2098457437112862e-06, + "loss": 0.1228, + "step": 2522 + }, + { + "epoch": 0.5740614334470989, + "grad_norm": 3.3253296009015556, + "learning_rate": 1.2098142274896033e-06, + "loss": 0.257, + "step": 2523 + }, + { + "epoch": 0.5742889647326508, + "grad_norm": 2.266880563348357, + "learning_rate": 1.2097826993153886e-06, + "loss": 0.1444, + "step": 2524 + }, + { + "epoch": 0.5745164960182025, + "grad_norm": 3.1698627124068297, + "learning_rate": 1.2097511591892863e-06, + "loss": 0.1051, + "step": 2525 + }, + { + "epoch": 0.5747440273037543, + "grad_norm": 2.8868768890738834, + "learning_rate": 1.2097196071119415e-06, + "loss": 0.0974, + "step": 2526 + }, + { + "epoch": 0.574971558589306, + "grad_norm": 1.9086447497414667, + "learning_rate": 1.2096880430839985e-06, + "loss": 0.127, + "step": 2527 + }, + { + "epoch": 0.5751990898748578, + "grad_norm": 1.9202653666719622, + "learning_rate": 1.2096564671061031e-06, + "loss": 0.1134, + "step": 2528 + }, + { + "epoch": 0.5754266211604095, + "grad_norm": 1.8265247037292291, + "learning_rate": 1.2096248791789e-06, + "loss": 0.1667, + "step": 2529 + }, + { + "epoch": 0.5756541524459613, + "grad_norm": 1.5150790673231787, + "learning_rate": 1.2095932793030353e-06, + "loss": 0.0952, + "step": 2530 + }, + { + "epoch": 0.575881683731513, + "grad_norm": 1.960225707664033, + "learning_rate": 1.2095616674791543e-06, + "loss": 0.1164, + "step": 2531 + }, + { + "epoch": 0.5761092150170648, + "grad_norm": 1.6828800349037405, + "learning_rate": 1.2095300437079034e-06, + "loss": 0.1587, + "step": 2532 + }, + { + "epoch": 0.5763367463026167, + "grad_norm": 1.4207974667109202, + "learning_rate": 1.2094984079899292e-06, + "loss": 0.0687, + "step": 2533 + }, + { + "epoch": 0.5765642775881684, + "grad_norm": 2.334850140119816, + "learning_rate": 1.2094667603258779e-06, + "loss": 0.1068, + "step": 2534 + }, + { + "epoch": 0.5767918088737202, + "grad_norm": 4.35410919897498, + "learning_rate": 1.2094351007163962e-06, + "loss": 0.1886, + "step": 2535 + }, + { + "epoch": 0.5770193401592719, + "grad_norm": 1.6200824503597024, + "learning_rate": 1.2094034291621315e-06, + "loss": 0.1223, + "step": 2536 + }, + { + "epoch": 0.5772468714448237, + "grad_norm": 2.362865286588217, + "learning_rate": 1.2093717456637311e-06, + "loss": 0.1375, + "step": 2537 + }, + { + "epoch": 0.5774744027303754, + "grad_norm": 2.308641909521365, + "learning_rate": 1.2093400502218422e-06, + "loss": 0.2275, + "step": 2538 + }, + { + "epoch": 0.5777019340159272, + "grad_norm": 2.463338738431784, + "learning_rate": 1.209308342837113e-06, + "loss": 0.1802, + "step": 2539 + }, + { + "epoch": 0.5779294653014789, + "grad_norm": 1.8971271189178798, + "learning_rate": 1.2092766235101917e-06, + "loss": 0.1305, + "step": 2540 + }, + { + "epoch": 0.5781569965870307, + "grad_norm": 2.1589012912488044, + "learning_rate": 1.2092448922417258e-06, + "loss": 0.1804, + "step": 2541 + }, + { + "epoch": 0.5783845278725824, + "grad_norm": 1.551935572792547, + "learning_rate": 1.2092131490323644e-06, + "loss": 0.1, + "step": 2542 + }, + { + "epoch": 0.5786120591581343, + "grad_norm": 1.868154527807695, + "learning_rate": 1.2091813938827563e-06, + "loss": 0.1133, + "step": 2543 + }, + { + "epoch": 0.578839590443686, + "grad_norm": 1.989931342962644, + "learning_rate": 1.2091496267935502e-06, + "loss": 0.163, + "step": 2544 + }, + { + "epoch": 0.5790671217292378, + "grad_norm": 0.9701995130312155, + "learning_rate": 1.2091178477653957e-06, + "loss": 0.1338, + "step": 2545 + }, + { + "epoch": 0.5792946530147896, + "grad_norm": 1.45039507553725, + "learning_rate": 1.209086056798942e-06, + "loss": 0.097, + "step": 2546 + }, + { + "epoch": 0.5795221843003413, + "grad_norm": 1.955389923714404, + "learning_rate": 1.2090542538948392e-06, + "loss": 0.1777, + "step": 2547 + }, + { + "epoch": 0.579749715585893, + "grad_norm": 1.4334415366285986, + "learning_rate": 1.209022439053737e-06, + "loss": 0.0969, + "step": 2548 + }, + { + "epoch": 0.5799772468714448, + "grad_norm": 2.3655923077610312, + "learning_rate": 1.2089906122762859e-06, + "loss": 0.1219, + "step": 2549 + }, + { + "epoch": 0.5802047781569966, + "grad_norm": 2.243979043940131, + "learning_rate": 1.208958773563136e-06, + "loss": 0.1543, + "step": 2550 + }, + { + "epoch": 0.5804323094425483, + "grad_norm": 2.87195064677509, + "learning_rate": 1.2089269229149383e-06, + "loss": 0.2538, + "step": 2551 + }, + { + "epoch": 0.5806598407281001, + "grad_norm": 1.6061469454663684, + "learning_rate": 1.2088950603323438e-06, + "loss": 0.1271, + "step": 2552 + }, + { + "epoch": 0.5808873720136519, + "grad_norm": 2.7705649289893666, + "learning_rate": 1.2088631858160033e-06, + "loss": 0.1284, + "step": 2553 + }, + { + "epoch": 0.5811149032992037, + "grad_norm": 2.2423139977954203, + "learning_rate": 1.2088312993665689e-06, + "loss": 0.1697, + "step": 2554 + }, + { + "epoch": 0.5813424345847554, + "grad_norm": 1.5090404390691508, + "learning_rate": 1.208799400984692e-06, + "loss": 0.1707, + "step": 2555 + }, + { + "epoch": 0.5815699658703072, + "grad_norm": 1.8137511001931925, + "learning_rate": 1.2087674906710242e-06, + "loss": 0.0793, + "step": 2556 + }, + { + "epoch": 0.5817974971558589, + "grad_norm": 1.2084657511133172, + "learning_rate": 1.2087355684262183e-06, + "loss": 0.1209, + "step": 2557 + }, + { + "epoch": 0.5820250284414107, + "grad_norm": 1.6868878164350085, + "learning_rate": 1.2087036342509265e-06, + "loss": 0.0798, + "step": 2558 + }, + { + "epoch": 0.5822525597269624, + "grad_norm": 2.4820318013238345, + "learning_rate": 1.208671688145801e-06, + "loss": 0.1487, + "step": 2559 + }, + { + "epoch": 0.5824800910125142, + "grad_norm": 1.956438161841264, + "learning_rate": 1.2086397301114955e-06, + "loss": 0.2435, + "step": 2560 + }, + { + "epoch": 0.5827076222980659, + "grad_norm": 1.8160672855837017, + "learning_rate": 1.2086077601486627e-06, + "loss": 0.1238, + "step": 2561 + }, + { + "epoch": 0.5829351535836177, + "grad_norm": 1.893090448329992, + "learning_rate": 1.2085757782579562e-06, + "loss": 0.2005, + "step": 2562 + }, + { + "epoch": 0.5831626848691696, + "grad_norm": 2.6162511578342778, + "learning_rate": 1.2085437844400293e-06, + "loss": 0.1268, + "step": 2563 + }, + { + "epoch": 0.5833902161547213, + "grad_norm": 1.4268922855878878, + "learning_rate": 1.2085117786955366e-06, + "loss": 0.1082, + "step": 2564 + }, + { + "epoch": 0.5836177474402731, + "grad_norm": 1.7555654121905828, + "learning_rate": 1.2084797610251314e-06, + "loss": 0.1628, + "step": 2565 + }, + { + "epoch": 0.5838452787258248, + "grad_norm": 2.7712208071139197, + "learning_rate": 1.2084477314294688e-06, + "loss": 0.1259, + "step": 2566 + }, + { + "epoch": 0.5840728100113766, + "grad_norm": 2.2477692009821886, + "learning_rate": 1.2084156899092028e-06, + "loss": 0.1151, + "step": 2567 + }, + { + "epoch": 0.5843003412969283, + "grad_norm": 1.4787985647224473, + "learning_rate": 1.2083836364649888e-06, + "loss": 0.1413, + "step": 2568 + }, + { + "epoch": 0.5845278725824801, + "grad_norm": 2.6543995947788397, + "learning_rate": 1.2083515710974813e-06, + "loss": 0.1298, + "step": 2569 + }, + { + "epoch": 0.5847554038680318, + "grad_norm": 2.211923597091862, + "learning_rate": 1.2083194938073363e-06, + "loss": 0.1284, + "step": 2570 + }, + { + "epoch": 0.5849829351535836, + "grad_norm": 3.9781424289807363, + "learning_rate": 1.2082874045952094e-06, + "loss": 0.1472, + "step": 2571 + }, + { + "epoch": 0.5852104664391354, + "grad_norm": 5.160608841853167, + "learning_rate": 1.2082553034617559e-06, + "loss": 0.1942, + "step": 2572 + }, + { + "epoch": 0.5854379977246872, + "grad_norm": 2.160091359131514, + "learning_rate": 1.2082231904076323e-06, + "loss": 0.1089, + "step": 2573 + }, + { + "epoch": 0.5856655290102389, + "grad_norm": 4.892526132201958, + "learning_rate": 1.2081910654334948e-06, + "loss": 0.2815, + "step": 2574 + }, + { + "epoch": 0.5858930602957907, + "grad_norm": 1.5921622052362032, + "learning_rate": 1.2081589285399998e-06, + "loss": 0.0935, + "step": 2575 + }, + { + "epoch": 0.5861205915813424, + "grad_norm": 1.9760107843414554, + "learning_rate": 1.2081267797278043e-06, + "loss": 0.1158, + "step": 2576 + }, + { + "epoch": 0.5863481228668942, + "grad_norm": 1.402210261553332, + "learning_rate": 1.2080946189975656e-06, + "loss": 0.1351, + "step": 2577 + }, + { + "epoch": 0.586575654152446, + "grad_norm": 2.437290370611529, + "learning_rate": 1.2080624463499407e-06, + "loss": 0.1348, + "step": 2578 + }, + { + "epoch": 0.5868031854379977, + "grad_norm": 1.569823831308918, + "learning_rate": 1.2080302617855874e-06, + "loss": 0.2161, + "step": 2579 + }, + { + "epoch": 0.5870307167235495, + "grad_norm": 2.0597197185846032, + "learning_rate": 1.2079980653051629e-06, + "loss": 0.0922, + "step": 2580 + }, + { + "epoch": 0.5872582480091012, + "grad_norm": 1.7138455088041311, + "learning_rate": 1.207965856909326e-06, + "loss": 0.1029, + "step": 2581 + }, + { + "epoch": 0.5874857792946531, + "grad_norm": 1.2897243188730496, + "learning_rate": 1.2079336365987345e-06, + "loss": 0.1481, + "step": 2582 + }, + { + "epoch": 0.5877133105802048, + "grad_norm": 1.8906043234612928, + "learning_rate": 1.2079014043740471e-06, + "loss": 0.1316, + "step": 2583 + }, + { + "epoch": 0.5879408418657566, + "grad_norm": 1.7419270898501555, + "learning_rate": 1.2078691602359224e-06, + "loss": 0.1193, + "step": 2584 + }, + { + "epoch": 0.5881683731513083, + "grad_norm": 1.819163624808354, + "learning_rate": 1.2078369041850197e-06, + "loss": 0.1174, + "step": 2585 + }, + { + "epoch": 0.5883959044368601, + "grad_norm": 1.877871761831476, + "learning_rate": 1.207804636221998e-06, + "loss": 0.1372, + "step": 2586 + }, + { + "epoch": 0.5886234357224118, + "grad_norm": 2.0626590360093133, + "learning_rate": 1.207772356347517e-06, + "loss": 0.095, + "step": 2587 + }, + { + "epoch": 0.5888509670079636, + "grad_norm": 2.456994619660604, + "learning_rate": 1.2077400645622363e-06, + "loss": 0.1224, + "step": 2588 + }, + { + "epoch": 0.5890784982935153, + "grad_norm": 1.5356214520040972, + "learning_rate": 1.207707760866816e-06, + "loss": 0.1166, + "step": 2589 + }, + { + "epoch": 0.5893060295790671, + "grad_norm": 5.018978969106035, + "learning_rate": 1.207675445261916e-06, + "loss": 0.1457, + "step": 2590 + }, + { + "epoch": 0.5895335608646188, + "grad_norm": 2.359188809516013, + "learning_rate": 1.207643117748197e-06, + "loss": 0.2648, + "step": 2591 + }, + { + "epoch": 0.5897610921501707, + "grad_norm": 2.4510393228560647, + "learning_rate": 1.2076107783263199e-06, + "loss": 0.1325, + "step": 2592 + }, + { + "epoch": 0.5899886234357224, + "grad_norm": 1.934483134443095, + "learning_rate": 1.2075784269969457e-06, + "loss": 0.1164, + "step": 2593 + }, + { + "epoch": 0.5902161547212742, + "grad_norm": 1.6195788093768093, + "learning_rate": 1.2075460637607351e-06, + "loss": 0.0678, + "step": 2594 + }, + { + "epoch": 0.590443686006826, + "grad_norm": 1.7633862699028207, + "learning_rate": 1.20751368861835e-06, + "loss": 0.1427, + "step": 2595 + }, + { + "epoch": 0.5906712172923777, + "grad_norm": 2.8589827615450503, + "learning_rate": 1.2074813015704518e-06, + "loss": 0.1232, + "step": 2596 + }, + { + "epoch": 0.5908987485779295, + "grad_norm": 2.237662265223862, + "learning_rate": 1.2074489026177024e-06, + "loss": 0.1414, + "step": 2597 + }, + { + "epoch": 0.5911262798634812, + "grad_norm": 1.6741078305495773, + "learning_rate": 1.2074164917607644e-06, + "loss": 0.1525, + "step": 2598 + }, + { + "epoch": 0.591353811149033, + "grad_norm": 1.7511984201836153, + "learning_rate": 1.2073840690003e-06, + "loss": 0.1283, + "step": 2599 + }, + { + "epoch": 0.5915813424345847, + "grad_norm": 0.9823739657729703, + "learning_rate": 1.2073516343369717e-06, + "loss": 0.1253, + "step": 2600 + }, + { + "epoch": 0.5918088737201365, + "grad_norm": 1.426041876858941, + "learning_rate": 1.2073191877714424e-06, + "loss": 0.1472, + "step": 2601 + }, + { + "epoch": 0.5920364050056883, + "grad_norm": 1.5467081035937693, + "learning_rate": 1.2072867293043755e-06, + "loss": 0.1129, + "step": 2602 + }, + { + "epoch": 0.5922639362912401, + "grad_norm": 2.4176519282492372, + "learning_rate": 1.2072542589364343e-06, + "loss": 0.1302, + "step": 2603 + }, + { + "epoch": 0.5924914675767918, + "grad_norm": 1.5431029441975226, + "learning_rate": 1.2072217766682822e-06, + "loss": 0.1259, + "step": 2604 + }, + { + "epoch": 0.5927189988623436, + "grad_norm": 2.1679653478901075, + "learning_rate": 1.2071892825005835e-06, + "loss": 0.0746, + "step": 2605 + }, + { + "epoch": 0.5929465301478953, + "grad_norm": 1.581497780473167, + "learning_rate": 1.207156776434002e-06, + "loss": 0.1668, + "step": 2606 + }, + { + "epoch": 0.5931740614334471, + "grad_norm": 1.9548539073094118, + "learning_rate": 1.2071242584692022e-06, + "loss": 0.1719, + "step": 2607 + }, + { + "epoch": 0.5934015927189988, + "grad_norm": 1.2461220647331583, + "learning_rate": 1.2070917286068486e-06, + "loss": 0.1041, + "step": 2608 + }, + { + "epoch": 0.5936291240045506, + "grad_norm": 1.3095375732805556, + "learning_rate": 1.2070591868476062e-06, + "loss": 0.0823, + "step": 2609 + }, + { + "epoch": 0.5938566552901023, + "grad_norm": 2.1017791753609347, + "learning_rate": 1.20702663319214e-06, + "loss": 0.098, + "step": 2610 + }, + { + "epoch": 0.5940841865756542, + "grad_norm": 1.3808889393646218, + "learning_rate": 1.2069940676411154e-06, + "loss": 0.1266, + "step": 2611 + }, + { + "epoch": 0.594311717861206, + "grad_norm": 1.7153990981225686, + "learning_rate": 1.2069614901951978e-06, + "loss": 0.0988, + "step": 2612 + }, + { + "epoch": 0.5945392491467577, + "grad_norm": 2.130551381955189, + "learning_rate": 1.2069289008550533e-06, + "loss": 0.1193, + "step": 2613 + }, + { + "epoch": 0.5947667804323095, + "grad_norm": 2.178000241862761, + "learning_rate": 1.2068962996213476e-06, + "loss": 0.125, + "step": 2614 + }, + { + "epoch": 0.5949943117178612, + "grad_norm": 3.035579264083657, + "learning_rate": 1.2068636864947475e-06, + "loss": 0.1585, + "step": 2615 + }, + { + "epoch": 0.595221843003413, + "grad_norm": 2.5260264361065476, + "learning_rate": 1.206831061475919e-06, + "loss": 0.1217, + "step": 2616 + }, + { + "epoch": 0.5954493742889647, + "grad_norm": 1.949452015122984, + "learning_rate": 1.2067984245655292e-06, + "loss": 0.068, + "step": 2617 + }, + { + "epoch": 0.5956769055745165, + "grad_norm": 1.911993378147549, + "learning_rate": 1.2067657757642453e-06, + "loss": 0.134, + "step": 2618 + }, + { + "epoch": 0.5959044368600682, + "grad_norm": 1.6307557736669274, + "learning_rate": 1.2067331150727343e-06, + "loss": 0.1071, + "step": 2619 + }, + { + "epoch": 0.59613196814562, + "grad_norm": 2.3641591554636165, + "learning_rate": 1.206700442491664e-06, + "loss": 0.1015, + "step": 2620 + }, + { + "epoch": 0.5963594994311718, + "grad_norm": 3.0546256940872776, + "learning_rate": 1.2066677580217018e-06, + "loss": 0.1353, + "step": 2621 + }, + { + "epoch": 0.5965870307167236, + "grad_norm": 2.0389636631107706, + "learning_rate": 1.2066350616635159e-06, + "loss": 0.0953, + "step": 2622 + }, + { + "epoch": 0.5968145620022753, + "grad_norm": 3.626384851780054, + "learning_rate": 1.2066023534177746e-06, + "loss": 0.2982, + "step": 2623 + }, + { + "epoch": 0.5970420932878271, + "grad_norm": 3.692303324562613, + "learning_rate": 1.2065696332851463e-06, + "loss": 0.1916, + "step": 2624 + }, + { + "epoch": 0.5972696245733788, + "grad_norm": 1.9357626525567908, + "learning_rate": 1.2065369012662999e-06, + "loss": 0.1351, + "step": 2625 + }, + { + "epoch": 0.5974971558589306, + "grad_norm": 1.690947384729753, + "learning_rate": 1.206504157361904e-06, + "loss": 0.1787, + "step": 2626 + }, + { + "epoch": 0.5977246871444823, + "grad_norm": 1.5180922357697746, + "learning_rate": 1.2064714015726283e-06, + "loss": 0.1399, + "step": 2627 + }, + { + "epoch": 0.5979522184300341, + "grad_norm": 1.9498622882163583, + "learning_rate": 1.2064386338991423e-06, + "loss": 0.1524, + "step": 2628 + }, + { + "epoch": 0.5981797497155859, + "grad_norm": 1.5948413120183693, + "learning_rate": 1.2064058543421152e-06, + "loss": 0.127, + "step": 2629 + }, + { + "epoch": 0.5984072810011376, + "grad_norm": 2.564122919945674, + "learning_rate": 1.2063730629022173e-06, + "loss": 0.1033, + "step": 2630 + }, + { + "epoch": 0.5986348122866895, + "grad_norm": 3.1247451680105227, + "learning_rate": 1.2063402595801187e-06, + "loss": 0.1595, + "step": 2631 + }, + { + "epoch": 0.5988623435722412, + "grad_norm": 1.1117218292823476, + "learning_rate": 1.2063074443764897e-06, + "loss": 0.1578, + "step": 2632 + }, + { + "epoch": 0.599089874857793, + "grad_norm": 3.166944891339004, + "learning_rate": 1.2062746172920014e-06, + "loss": 0.2975, + "step": 2633 + }, + { + "epoch": 0.5993174061433447, + "grad_norm": 3.3367189368333787, + "learning_rate": 1.2062417783273246e-06, + "loss": 0.155, + "step": 2634 + }, + { + "epoch": 0.5995449374288965, + "grad_norm": 2.581770735788427, + "learning_rate": 1.2062089274831299e-06, + "loss": 0.14, + "step": 2635 + }, + { + "epoch": 0.5997724687144482, + "grad_norm": 1.6918956017307545, + "learning_rate": 1.2061760647600894e-06, + "loss": 0.0787, + "step": 2636 + }, + { + "epoch": 0.6, + "grad_norm": 2.4426051263794752, + "learning_rate": 1.2061431901588747e-06, + "loss": 0.1761, + "step": 2637 + }, + { + "epoch": 0.6002275312855517, + "grad_norm": 1.5852858580417466, + "learning_rate": 1.2061103036801573e-06, + "loss": 0.0991, + "step": 2638 + }, + { + "epoch": 0.6004550625711035, + "grad_norm": 2.9361978747865205, + "learning_rate": 1.2060774053246096e-06, + "loss": 0.1966, + "step": 2639 + }, + { + "epoch": 0.6006825938566553, + "grad_norm": 1.1851482113806004, + "learning_rate": 1.2060444950929038e-06, + "loss": 0.1298, + "step": 2640 + }, + { + "epoch": 0.6009101251422071, + "grad_norm": 1.5583103094664725, + "learning_rate": 1.2060115729857128e-06, + "loss": 0.1305, + "step": 2641 + }, + { + "epoch": 0.6011376564277588, + "grad_norm": 2.0530153143627103, + "learning_rate": 1.2059786390037093e-06, + "loss": 0.1194, + "step": 2642 + }, + { + "epoch": 0.6013651877133106, + "grad_norm": 1.3816354544320564, + "learning_rate": 1.2059456931475663e-06, + "loss": 0.0506, + "step": 2643 + }, + { + "epoch": 0.6015927189988624, + "grad_norm": 1.8067957499098637, + "learning_rate": 1.2059127354179573e-06, + "loss": 0.1829, + "step": 2644 + }, + { + "epoch": 0.6018202502844141, + "grad_norm": 2.2404814486972278, + "learning_rate": 1.205879765815556e-06, + "loss": 0.135, + "step": 2645 + }, + { + "epoch": 0.6020477815699659, + "grad_norm": 2.0019027614965728, + "learning_rate": 1.205846784341036e-06, + "loss": 0.1412, + "step": 2646 + }, + { + "epoch": 0.6022753128555176, + "grad_norm": 2.3427613212776186, + "learning_rate": 1.2058137909950719e-06, + "loss": 0.1094, + "step": 2647 + }, + { + "epoch": 0.6025028441410694, + "grad_norm": 2.551361820189807, + "learning_rate": 1.2057807857783371e-06, + "loss": 0.1278, + "step": 2648 + }, + { + "epoch": 0.6027303754266211, + "grad_norm": 2.4318258413096108, + "learning_rate": 1.205747768691507e-06, + "loss": 0.1063, + "step": 2649 + }, + { + "epoch": 0.602957906712173, + "grad_norm": 1.5810203336032662, + "learning_rate": 1.2057147397352559e-06, + "loss": 0.1529, + "step": 2650 + }, + { + "epoch": 0.6031854379977247, + "grad_norm": 1.5838470813817764, + "learning_rate": 1.2056816989102591e-06, + "loss": 0.1376, + "step": 2651 + }, + { + "epoch": 0.6034129692832765, + "grad_norm": 3.9369996093181014, + "learning_rate": 1.2056486462171918e-06, + "loss": 0.2464, + "step": 2652 + }, + { + "epoch": 0.6036405005688282, + "grad_norm": 1.7557943491318282, + "learning_rate": 1.2056155816567297e-06, + "loss": 0.1506, + "step": 2653 + }, + { + "epoch": 0.60386803185438, + "grad_norm": 1.4902613853193347, + "learning_rate": 1.2055825052295486e-06, + "loss": 0.1356, + "step": 2654 + }, + { + "epoch": 0.6040955631399317, + "grad_norm": 1.6515131979163473, + "learning_rate": 1.205549416936324e-06, + "loss": 0.1648, + "step": 2655 + }, + { + "epoch": 0.6043230944254835, + "grad_norm": 1.4640203024864866, + "learning_rate": 1.2055163167777328e-06, + "loss": 0.1356, + "step": 2656 + }, + { + "epoch": 0.6045506257110352, + "grad_norm": 3.266508891752542, + "learning_rate": 1.2054832047544512e-06, + "loss": 0.173, + "step": 2657 + }, + { + "epoch": 0.604778156996587, + "grad_norm": 1.3631002432598898, + "learning_rate": 1.205450080867156e-06, + "loss": 0.1096, + "step": 2658 + }, + { + "epoch": 0.6050056882821387, + "grad_norm": 2.242984662905767, + "learning_rate": 1.205416945116524e-06, + "loss": 0.1335, + "step": 2659 + }, + { + "epoch": 0.6052332195676906, + "grad_norm": 2.3400648593117745, + "learning_rate": 1.2053837975032328e-06, + "loss": 0.0818, + "step": 2660 + }, + { + "epoch": 0.6054607508532424, + "grad_norm": 2.283977245108534, + "learning_rate": 1.2053506380279597e-06, + "loss": 0.1264, + "step": 2661 + }, + { + "epoch": 0.6056882821387941, + "grad_norm": 1.5198638190414868, + "learning_rate": 1.2053174666913826e-06, + "loss": 0.1303, + "step": 2662 + }, + { + "epoch": 0.6059158134243459, + "grad_norm": 1.3638719142798548, + "learning_rate": 1.2052842834941791e-06, + "loss": 0.0996, + "step": 2663 + }, + { + "epoch": 0.6061433447098976, + "grad_norm": 2.0164471068005305, + "learning_rate": 1.2052510884370274e-06, + "loss": 0.1338, + "step": 2664 + }, + { + "epoch": 0.6063708759954494, + "grad_norm": 1.7349041320402774, + "learning_rate": 1.2052178815206064e-06, + "loss": 0.1023, + "step": 2665 + }, + { + "epoch": 0.6065984072810011, + "grad_norm": 1.4876292213814133, + "learning_rate": 1.2051846627455946e-06, + "loss": 0.1226, + "step": 2666 + }, + { + "epoch": 0.6068259385665529, + "grad_norm": 1.6427948767677636, + "learning_rate": 1.2051514321126705e-06, + "loss": 0.141, + "step": 2667 + }, + { + "epoch": 0.6070534698521046, + "grad_norm": 2.0737972304672083, + "learning_rate": 1.2051181896225139e-06, + "loss": 0.1221, + "step": 2668 + }, + { + "epoch": 0.6072810011376564, + "grad_norm": 1.9014861477421603, + "learning_rate": 1.205084935275804e-06, + "loss": 0.1461, + "step": 2669 + }, + { + "epoch": 0.6075085324232082, + "grad_norm": 2.80889588671251, + "learning_rate": 1.20505166907322e-06, + "loss": 0.1772, + "step": 2670 + }, + { + "epoch": 0.60773606370876, + "grad_norm": 1.5524251793071988, + "learning_rate": 1.2050183910154425e-06, + "loss": 0.1063, + "step": 2671 + }, + { + "epoch": 0.6079635949943117, + "grad_norm": 2.282729570815144, + "learning_rate": 1.2049851011031514e-06, + "loss": 0.1548, + "step": 2672 + }, + { + "epoch": 0.6081911262798635, + "grad_norm": 2.1806480444286747, + "learning_rate": 1.2049517993370269e-06, + "loss": 0.1622, + "step": 2673 + }, + { + "epoch": 0.6084186575654152, + "grad_norm": 2.3188525896490906, + "learning_rate": 1.2049184857177498e-06, + "loss": 0.2115, + "step": 2674 + }, + { + "epoch": 0.608646188850967, + "grad_norm": 1.700801852435792, + "learning_rate": 1.204885160246001e-06, + "loss": 0.2375, + "step": 2675 + }, + { + "epoch": 0.6088737201365187, + "grad_norm": 3.07988581510915, + "learning_rate": 1.2048518229224613e-06, + "loss": 0.1452, + "step": 2676 + }, + { + "epoch": 0.6091012514220705, + "grad_norm": 1.579825688333633, + "learning_rate": 1.2048184737478124e-06, + "loss": 0.1165, + "step": 2677 + }, + { + "epoch": 0.6093287827076223, + "grad_norm": 1.7635258179947948, + "learning_rate": 1.2047851127227358e-06, + "loss": 0.1015, + "step": 2678 + }, + { + "epoch": 0.6095563139931741, + "grad_norm": 1.9171716270170973, + "learning_rate": 1.2047517398479135e-06, + "loss": 0.1093, + "step": 2679 + }, + { + "epoch": 0.6097838452787259, + "grad_norm": 1.6246200869035246, + "learning_rate": 1.204718355124027e-06, + "loss": 0.1477, + "step": 2680 + }, + { + "epoch": 0.6100113765642776, + "grad_norm": 2.033819677297613, + "learning_rate": 1.2046849585517595e-06, + "loss": 0.1112, + "step": 2681 + }, + { + "epoch": 0.6102389078498294, + "grad_norm": 1.5057733369483088, + "learning_rate": 1.2046515501317927e-06, + "loss": 0.1144, + "step": 2682 + }, + { + "epoch": 0.6104664391353811, + "grad_norm": 1.9207619842965977, + "learning_rate": 1.2046181298648101e-06, + "loss": 0.165, + "step": 2683 + }, + { + "epoch": 0.6106939704209329, + "grad_norm": 2.1068795511508536, + "learning_rate": 1.2045846977514943e-06, + "loss": 0.1119, + "step": 2684 + }, + { + "epoch": 0.6109215017064846, + "grad_norm": 1.5765181731137923, + "learning_rate": 1.2045512537925287e-06, + "loss": 0.0938, + "step": 2685 + }, + { + "epoch": 0.6111490329920364, + "grad_norm": 1.652132408051153, + "learning_rate": 1.2045177979885969e-06, + "loss": 0.1119, + "step": 2686 + }, + { + "epoch": 0.6113765642775881, + "grad_norm": 1.9933401024918733, + "learning_rate": 1.2044843303403827e-06, + "loss": 0.1127, + "step": 2687 + }, + { + "epoch": 0.6116040955631399, + "grad_norm": 1.2553964688013253, + "learning_rate": 1.20445085084857e-06, + "loss": 0.117, + "step": 2688 + }, + { + "epoch": 0.6118316268486917, + "grad_norm": 1.9328161675026851, + "learning_rate": 1.204417359513843e-06, + "loss": 0.1055, + "step": 2689 + }, + { + "epoch": 0.6120591581342435, + "grad_norm": 1.6135864750332363, + "learning_rate": 1.2043838563368865e-06, + "loss": 0.1311, + "step": 2690 + }, + { + "epoch": 0.6122866894197952, + "grad_norm": 2.8366202101928386, + "learning_rate": 1.204350341318385e-06, + "loss": 0.1576, + "step": 2691 + }, + { + "epoch": 0.612514220705347, + "grad_norm": 1.9115878659986092, + "learning_rate": 1.2043168144590237e-06, + "loss": 0.098, + "step": 2692 + }, + { + "epoch": 0.6127417519908988, + "grad_norm": 1.6553419799458462, + "learning_rate": 1.2042832757594875e-06, + "loss": 0.1482, + "step": 2693 + }, + { + "epoch": 0.6129692832764505, + "grad_norm": 1.3298128063922734, + "learning_rate": 1.204249725220462e-06, + "loss": 0.0801, + "step": 2694 + }, + { + "epoch": 0.6131968145620023, + "grad_norm": 2.112063522190015, + "learning_rate": 1.204216162842633e-06, + "loss": 0.1534, + "step": 2695 + }, + { + "epoch": 0.613424345847554, + "grad_norm": 1.9789341169381445, + "learning_rate": 1.2041825886266866e-06, + "loss": 0.0911, + "step": 2696 + }, + { + "epoch": 0.6136518771331058, + "grad_norm": 2.121057060579873, + "learning_rate": 1.2041490025733089e-06, + "loss": 0.1465, + "step": 2697 + }, + { + "epoch": 0.6138794084186575, + "grad_norm": 1.8588673717447004, + "learning_rate": 1.2041154046831859e-06, + "loss": 0.1035, + "step": 2698 + }, + { + "epoch": 0.6141069397042094, + "grad_norm": 1.734446448883455, + "learning_rate": 1.2040817949570046e-06, + "loss": 0.1563, + "step": 2699 + }, + { + "epoch": 0.6143344709897611, + "grad_norm": 1.9933118037524564, + "learning_rate": 1.2040481733954523e-06, + "loss": 0.1024, + "step": 2700 + }, + { + "epoch": 0.6145620022753129, + "grad_norm": 3.17488856013429, + "learning_rate": 1.2040145399992157e-06, + "loss": 0.1226, + "step": 2701 + }, + { + "epoch": 0.6147895335608646, + "grad_norm": 1.591059020243372, + "learning_rate": 1.203980894768982e-06, + "loss": 0.1195, + "step": 2702 + }, + { + "epoch": 0.6150170648464164, + "grad_norm": 1.601828253256089, + "learning_rate": 1.2039472377054395e-06, + "loss": 0.1385, + "step": 2703 + }, + { + "epoch": 0.6152445961319681, + "grad_norm": 3.476222784517199, + "learning_rate": 1.2039135688092757e-06, + "loss": 0.2026, + "step": 2704 + }, + { + "epoch": 0.6154721274175199, + "grad_norm": 2.5679221489922814, + "learning_rate": 1.203879888081179e-06, + "loss": 0.1373, + "step": 2705 + }, + { + "epoch": 0.6156996587030716, + "grad_norm": 2.1049982546040584, + "learning_rate": 1.203846195521837e-06, + "loss": 0.1353, + "step": 2706 + }, + { + "epoch": 0.6159271899886234, + "grad_norm": 2.449412628808893, + "learning_rate": 1.2038124911319393e-06, + "loss": 0.136, + "step": 2707 + }, + { + "epoch": 0.6161547212741753, + "grad_norm": 2.8316029320784857, + "learning_rate": 1.2037787749121741e-06, + "loss": 0.1887, + "step": 2708 + }, + { + "epoch": 0.616382252559727, + "grad_norm": 2.0505324298226517, + "learning_rate": 1.203745046863231e-06, + "loss": 0.1824, + "step": 2709 + }, + { + "epoch": 0.6166097838452788, + "grad_norm": 2.354091965640521, + "learning_rate": 1.203711306985799e-06, + "loss": 0.1024, + "step": 2710 + }, + { + "epoch": 0.6168373151308305, + "grad_norm": 1.497422096890729, + "learning_rate": 1.2036775552805674e-06, + "loss": 0.1001, + "step": 2711 + }, + { + "epoch": 0.6170648464163823, + "grad_norm": 1.6367535767277195, + "learning_rate": 1.2036437917482267e-06, + "loss": 0.1951, + "step": 2712 + }, + { + "epoch": 0.617292377701934, + "grad_norm": 1.311702192960621, + "learning_rate": 1.2036100163894665e-06, + "loss": 0.0991, + "step": 2713 + }, + { + "epoch": 0.6175199089874858, + "grad_norm": 1.1374509331150917, + "learning_rate": 1.2035762292049772e-06, + "loss": 0.1279, + "step": 2714 + }, + { + "epoch": 0.6177474402730375, + "grad_norm": 2.0690667114321437, + "learning_rate": 1.2035424301954496e-06, + "loss": 0.1213, + "step": 2715 + }, + { + "epoch": 0.6179749715585893, + "grad_norm": 1.6164688794833326, + "learning_rate": 1.2035086193615743e-06, + "loss": 0.0896, + "step": 2716 + }, + { + "epoch": 0.618202502844141, + "grad_norm": 1.4130291059742248, + "learning_rate": 1.203474796704042e-06, + "loss": 0.1107, + "step": 2717 + }, + { + "epoch": 0.6184300341296929, + "grad_norm": 2.317711426962595, + "learning_rate": 1.2034409622235444e-06, + "loss": 0.1119, + "step": 2718 + }, + { + "epoch": 0.6186575654152446, + "grad_norm": 1.70184913449098, + "learning_rate": 1.203407115920773e-06, + "loss": 0.0904, + "step": 2719 + }, + { + "epoch": 0.6188850967007964, + "grad_norm": 1.6503172233844747, + "learning_rate": 1.2033732577964194e-06, + "loss": 0.1266, + "step": 2720 + }, + { + "epoch": 0.6191126279863481, + "grad_norm": 2.4366209863100523, + "learning_rate": 1.2033393878511756e-06, + "loss": 0.1257, + "step": 2721 + }, + { + "epoch": 0.6193401592718999, + "grad_norm": 1.582050028709684, + "learning_rate": 1.203305506085734e-06, + "loss": 0.0962, + "step": 2722 + }, + { + "epoch": 0.6195676905574516, + "grad_norm": 1.2773158456960036, + "learning_rate": 1.2032716125007868e-06, + "loss": 0.1288, + "step": 2723 + }, + { + "epoch": 0.6197952218430034, + "grad_norm": 2.7972062325966607, + "learning_rate": 1.2032377070970268e-06, + "loss": 0.1169, + "step": 2724 + }, + { + "epoch": 0.6200227531285551, + "grad_norm": 1.7121884786477586, + "learning_rate": 1.2032037898751475e-06, + "loss": 0.0864, + "step": 2725 + }, + { + "epoch": 0.6202502844141069, + "grad_norm": 1.8870447056565103, + "learning_rate": 1.2031698608358414e-06, + "loss": 0.0722, + "step": 2726 + }, + { + "epoch": 0.6204778156996587, + "grad_norm": 1.8796137296419013, + "learning_rate": 1.2031359199798021e-06, + "loss": 0.1111, + "step": 2727 + }, + { + "epoch": 0.6207053469852105, + "grad_norm": 1.6206526861878807, + "learning_rate": 1.2031019673077237e-06, + "loss": 0.1443, + "step": 2728 + }, + { + "epoch": 0.6209328782707623, + "grad_norm": 1.359835269149287, + "learning_rate": 1.2030680028202995e-06, + "loss": 0.1357, + "step": 2729 + }, + { + "epoch": 0.621160409556314, + "grad_norm": 2.7444994804468434, + "learning_rate": 1.2030340265182242e-06, + "loss": 0.123, + "step": 2730 + }, + { + "epoch": 0.6213879408418658, + "grad_norm": 1.601609743110609, + "learning_rate": 1.2030000384021919e-06, + "loss": 0.2195, + "step": 2731 + }, + { + "epoch": 0.6216154721274175, + "grad_norm": 3.331047575038035, + "learning_rate": 1.2029660384728974e-06, + "loss": 0.1182, + "step": 2732 + }, + { + "epoch": 0.6218430034129693, + "grad_norm": 2.934494529592981, + "learning_rate": 1.2029320267310359e-06, + "loss": 0.1967, + "step": 2733 + }, + { + "epoch": 0.622070534698521, + "grad_norm": 3.0682381227751043, + "learning_rate": 1.2028980031773018e-06, + "loss": 0.1732, + "step": 2734 + }, + { + "epoch": 0.6222980659840728, + "grad_norm": 1.8442015868919528, + "learning_rate": 1.202863967812391e-06, + "loss": 0.1799, + "step": 2735 + }, + { + "epoch": 0.6225255972696245, + "grad_norm": 1.8593176813243222, + "learning_rate": 1.2028299206369991e-06, + "loss": 0.1908, + "step": 2736 + }, + { + "epoch": 0.6227531285551763, + "grad_norm": 2.728968371383416, + "learning_rate": 1.2027958616518218e-06, + "loss": 0.152, + "step": 2737 + }, + { + "epoch": 0.6229806598407281, + "grad_norm": 2.6619054180942214, + "learning_rate": 1.2027617908575553e-06, + "loss": 0.1727, + "step": 2738 + }, + { + "epoch": 0.6232081911262799, + "grad_norm": 2.0414592133205303, + "learning_rate": 1.2027277082548958e-06, + "loss": 0.1424, + "step": 2739 + }, + { + "epoch": 0.6234357224118316, + "grad_norm": 3.1585074562667583, + "learning_rate": 1.20269361384454e-06, + "loss": 0.1412, + "step": 2740 + }, + { + "epoch": 0.6236632536973834, + "grad_norm": 2.520373823420793, + "learning_rate": 1.2026595076271848e-06, + "loss": 0.1181, + "step": 2741 + }, + { + "epoch": 0.6238907849829352, + "grad_norm": 1.8119986570748203, + "learning_rate": 1.2026253896035273e-06, + "loss": 0.1567, + "step": 2742 + }, + { + "epoch": 0.6241183162684869, + "grad_norm": 1.5487296455776507, + "learning_rate": 1.2025912597742646e-06, + "loss": 0.1557, + "step": 2743 + }, + { + "epoch": 0.6243458475540387, + "grad_norm": 2.5231360246572656, + "learning_rate": 1.2025571181400944e-06, + "loss": 0.1992, + "step": 2744 + }, + { + "epoch": 0.6245733788395904, + "grad_norm": 3.0625635018387585, + "learning_rate": 1.2025229647017145e-06, + "loss": 0.1716, + "step": 2745 + }, + { + "epoch": 0.6248009101251422, + "grad_norm": 3.1177129975506657, + "learning_rate": 1.2024887994598227e-06, + "loss": 0.1792, + "step": 2746 + }, + { + "epoch": 0.625028441410694, + "grad_norm": 2.1464563875652187, + "learning_rate": 1.2024546224151176e-06, + "loss": 0.0871, + "step": 2747 + }, + { + "epoch": 0.6252559726962458, + "grad_norm": 2.0265120439546136, + "learning_rate": 1.2024204335682977e-06, + "loss": 0.1361, + "step": 2748 + }, + { + "epoch": 0.6254835039817975, + "grad_norm": 3.7277830645236354, + "learning_rate": 1.2023862329200613e-06, + "loss": 0.1569, + "step": 2749 + }, + { + "epoch": 0.6257110352673493, + "grad_norm": 1.4107955390021432, + "learning_rate": 1.2023520204711078e-06, + "loss": 0.0982, + "step": 2750 + }, + { + "epoch": 0.625938566552901, + "grad_norm": 1.802956258585525, + "learning_rate": 1.2023177962221366e-06, + "loss": 0.202, + "step": 2751 + }, + { + "epoch": 0.6261660978384528, + "grad_norm": 2.570407091125862, + "learning_rate": 1.2022835601738467e-06, + "loss": 0.1389, + "step": 2752 + }, + { + "epoch": 0.6263936291240045, + "grad_norm": 1.9988049268762371, + "learning_rate": 1.2022493123269383e-06, + "loss": 0.1377, + "step": 2753 + }, + { + "epoch": 0.6266211604095563, + "grad_norm": 2.191923581547263, + "learning_rate": 1.202215052682111e-06, + "loss": 0.2517, + "step": 2754 + }, + { + "epoch": 0.626848691695108, + "grad_norm": 1.851526943619937, + "learning_rate": 1.2021807812400652e-06, + "loss": 0.1705, + "step": 2755 + }, + { + "epoch": 0.6270762229806598, + "grad_norm": 1.6140773681120666, + "learning_rate": 1.2021464980015014e-06, + "loss": 0.1198, + "step": 2756 + }, + { + "epoch": 0.6273037542662117, + "grad_norm": 1.5559727173563562, + "learning_rate": 1.20211220296712e-06, + "loss": 0.145, + "step": 2757 + }, + { + "epoch": 0.6275312855517634, + "grad_norm": 2.8683941887361746, + "learning_rate": 1.2020778961376223e-06, + "loss": 0.0834, + "step": 2758 + }, + { + "epoch": 0.6277588168373152, + "grad_norm": 2.155047837786128, + "learning_rate": 1.202043577513709e-06, + "loss": 0.1452, + "step": 2759 + }, + { + "epoch": 0.6279863481228669, + "grad_norm": 2.433600771864086, + "learning_rate": 1.202009247096082e-06, + "loss": 0.1903, + "step": 2760 + }, + { + "epoch": 0.6282138794084187, + "grad_norm": 1.9186109054985514, + "learning_rate": 1.2019749048854426e-06, + "loss": 0.1177, + "step": 2761 + }, + { + "epoch": 0.6284414106939704, + "grad_norm": 1.109450269371729, + "learning_rate": 1.2019405508824927e-06, + "loss": 0.0703, + "step": 2762 + }, + { + "epoch": 0.6286689419795222, + "grad_norm": 1.5591658788371794, + "learning_rate": 1.201906185087935e-06, + "loss": 0.101, + "step": 2763 + }, + { + "epoch": 0.6288964732650739, + "grad_norm": 2.141050899035466, + "learning_rate": 1.201871807502471e-06, + "loss": 0.1387, + "step": 2764 + }, + { + "epoch": 0.6291240045506257, + "grad_norm": 2.7514196386904746, + "learning_rate": 1.2018374181268039e-06, + "loss": 0.1553, + "step": 2765 + }, + { + "epoch": 0.6293515358361774, + "grad_norm": 2.535759371521438, + "learning_rate": 1.2018030169616363e-06, + "loss": 0.1287, + "step": 2766 + }, + { + "epoch": 0.6295790671217293, + "grad_norm": 2.770098292987776, + "learning_rate": 1.2017686040076715e-06, + "loss": 0.1689, + "step": 2767 + }, + { + "epoch": 0.629806598407281, + "grad_norm": 1.5676839925835293, + "learning_rate": 1.2017341792656129e-06, + "loss": 0.1506, + "step": 2768 + }, + { + "epoch": 0.6300341296928328, + "grad_norm": 2.568669185611603, + "learning_rate": 1.2016997427361635e-06, + "loss": 0.1183, + "step": 2769 + }, + { + "epoch": 0.6302616609783845, + "grad_norm": 1.4579610759738018, + "learning_rate": 1.201665294420028e-06, + "loss": 0.1407, + "step": 2770 + }, + { + "epoch": 0.6304891922639363, + "grad_norm": 2.951412169453364, + "learning_rate": 1.2016308343179098e-06, + "loss": 0.1178, + "step": 2771 + }, + { + "epoch": 0.630716723549488, + "grad_norm": 2.0276568694084953, + "learning_rate": 1.2015963624305132e-06, + "loss": 0.1649, + "step": 2772 + }, + { + "epoch": 0.6309442548350398, + "grad_norm": 2.280981687544379, + "learning_rate": 1.201561878758543e-06, + "loss": 0.1413, + "step": 2773 + }, + { + "epoch": 0.6311717861205915, + "grad_norm": 2.1266711894469736, + "learning_rate": 1.2015273833027041e-06, + "loss": 0.2162, + "step": 2774 + }, + { + "epoch": 0.6313993174061433, + "grad_norm": 3.8642578300165384, + "learning_rate": 1.2014928760637014e-06, + "loss": 0.2027, + "step": 2775 + }, + { + "epoch": 0.631626848691695, + "grad_norm": 3.216392739870665, + "learning_rate": 1.20145835704224e-06, + "loss": 0.1329, + "step": 2776 + }, + { + "epoch": 0.6318543799772469, + "grad_norm": 2.0044712366152053, + "learning_rate": 1.2014238262390254e-06, + "loss": 0.1391, + "step": 2777 + }, + { + "epoch": 0.6320819112627987, + "grad_norm": 3.2153887606485982, + "learning_rate": 1.2013892836547635e-06, + "loss": 0.2441, + "step": 2778 + }, + { + "epoch": 0.6323094425483504, + "grad_norm": 1.3533799512884244, + "learning_rate": 1.2013547292901605e-06, + "loss": 0.1688, + "step": 2779 + }, + { + "epoch": 0.6325369738339022, + "grad_norm": 1.751487029561619, + "learning_rate": 1.2013201631459222e-06, + "loss": 0.207, + "step": 2780 + }, + { + "epoch": 0.6327645051194539, + "grad_norm": 3.9713087691369635, + "learning_rate": 1.201285585222755e-06, + "loss": 0.2355, + "step": 2781 + }, + { + "epoch": 0.6329920364050057, + "grad_norm": 2.953556617774688, + "learning_rate": 1.2012509955213664e-06, + "loss": 0.1799, + "step": 2782 + }, + { + "epoch": 0.6332195676905574, + "grad_norm": 1.430149829222375, + "learning_rate": 1.2012163940424624e-06, + "loss": 0.1034, + "step": 2783 + }, + { + "epoch": 0.6334470989761092, + "grad_norm": 2.0923129272368164, + "learning_rate": 1.2011817807867507e-06, + "loss": 0.1246, + "step": 2784 + }, + { + "epoch": 0.6336746302616609, + "grad_norm": 3.407094572241824, + "learning_rate": 1.2011471557549387e-06, + "loss": 0.1728, + "step": 2785 + }, + { + "epoch": 0.6339021615472128, + "grad_norm": 2.0611288538144197, + "learning_rate": 1.2011125189477339e-06, + "loss": 0.1469, + "step": 2786 + }, + { + "epoch": 0.6341296928327645, + "grad_norm": 2.2966058332401587, + "learning_rate": 1.2010778703658441e-06, + "loss": 0.1559, + "step": 2787 + }, + { + "epoch": 0.6343572241183163, + "grad_norm": 1.545873831830917, + "learning_rate": 1.2010432100099781e-06, + "loss": 0.1571, + "step": 2788 + }, + { + "epoch": 0.634584755403868, + "grad_norm": 3.1456139874110156, + "learning_rate": 1.2010085378808437e-06, + "loss": 0.1906, + "step": 2789 + }, + { + "epoch": 0.6348122866894198, + "grad_norm": 2.2942374687881317, + "learning_rate": 1.2009738539791497e-06, + "loss": 0.2298, + "step": 2790 + }, + { + "epoch": 0.6350398179749716, + "grad_norm": 1.8483321215374322, + "learning_rate": 1.2009391583056048e-06, + "loss": 0.1769, + "step": 2791 + }, + { + "epoch": 0.6352673492605233, + "grad_norm": 2.097634425130467, + "learning_rate": 1.2009044508609182e-06, + "loss": 0.1369, + "step": 2792 + }, + { + "epoch": 0.6354948805460751, + "grad_norm": 2.181756009813163, + "learning_rate": 1.2008697316457997e-06, + "loss": 0.1671, + "step": 2793 + }, + { + "epoch": 0.6357224118316268, + "grad_norm": 2.4427121188314853, + "learning_rate": 1.2008350006609584e-06, + "loss": 0.1311, + "step": 2794 + }, + { + "epoch": 0.6359499431171786, + "grad_norm": 1.9343877225744588, + "learning_rate": 1.2008002579071043e-06, + "loss": 0.1908, + "step": 2795 + }, + { + "epoch": 0.6361774744027304, + "grad_norm": 1.8080071533035844, + "learning_rate": 1.2007655033849474e-06, + "loss": 0.1446, + "step": 2796 + }, + { + "epoch": 0.6364050056882822, + "grad_norm": 2.0360388556721034, + "learning_rate": 1.2007307370951983e-06, + "loss": 0.1262, + "step": 2797 + }, + { + "epoch": 0.6366325369738339, + "grad_norm": 1.6254628406869118, + "learning_rate": 1.200695959038567e-06, + "loss": 0.1256, + "step": 2798 + }, + { + "epoch": 0.6368600682593857, + "grad_norm": 1.9561465852797444, + "learning_rate": 1.2006611692157648e-06, + "loss": 0.1436, + "step": 2799 + }, + { + "epoch": 0.6370875995449374, + "grad_norm": 1.8226458153348386, + "learning_rate": 1.2006263676275026e-06, + "loss": 0.1756, + "step": 2800 + }, + { + "epoch": 0.6373151308304892, + "grad_norm": 1.9581606556638822, + "learning_rate": 1.2005915542744915e-06, + "loss": 0.1508, + "step": 2801 + }, + { + "epoch": 0.6375426621160409, + "grad_norm": 2.042581734647661, + "learning_rate": 1.2005567291574434e-06, + "loss": 0.1042, + "step": 2802 + }, + { + "epoch": 0.6377701934015927, + "grad_norm": 2.1726322807786067, + "learning_rate": 1.2005218922770695e-06, + "loss": 0.1318, + "step": 2803 + }, + { + "epoch": 0.6379977246871444, + "grad_norm": 1.9412456138719576, + "learning_rate": 1.2004870436340824e-06, + "loss": 0.1893, + "step": 2804 + }, + { + "epoch": 0.6382252559726962, + "grad_norm": 2.0304542382347113, + "learning_rate": 1.2004521832291943e-06, + "loss": 0.0981, + "step": 2805 + }, + { + "epoch": 0.6384527872582481, + "grad_norm": 1.6325066329724485, + "learning_rate": 1.200417311063117e-06, + "loss": 0.1384, + "step": 2806 + }, + { + "epoch": 0.6386803185437998, + "grad_norm": 1.8108206373995843, + "learning_rate": 1.200382427136564e-06, + "loss": 0.1196, + "step": 2807 + }, + { + "epoch": 0.6389078498293516, + "grad_norm": 2.5409921749520086, + "learning_rate": 1.2003475314502477e-06, + "loss": 0.1537, + "step": 2808 + }, + { + "epoch": 0.6391353811149033, + "grad_norm": 1.2748621426670415, + "learning_rate": 1.200312624004882e-06, + "loss": 0.0795, + "step": 2809 + }, + { + "epoch": 0.6393629124004551, + "grad_norm": 1.5734235185818852, + "learning_rate": 1.2002777048011794e-06, + "loss": 0.1221, + "step": 2810 + }, + { + "epoch": 0.6395904436860068, + "grad_norm": 2.8473432803680416, + "learning_rate": 1.2002427738398543e-06, + "loss": 0.1685, + "step": 2811 + }, + { + "epoch": 0.6398179749715586, + "grad_norm": 2.5860893642979956, + "learning_rate": 1.2002078311216205e-06, + "loss": 0.1225, + "step": 2812 + }, + { + "epoch": 0.6400455062571103, + "grad_norm": 2.173062793872002, + "learning_rate": 1.2001728766471919e-06, + "loss": 0.1418, + "step": 2813 + }, + { + "epoch": 0.6402730375426621, + "grad_norm": 2.341929374014376, + "learning_rate": 1.2001379104172832e-06, + "loss": 0.1501, + "step": 2814 + }, + { + "epoch": 0.6405005688282139, + "grad_norm": 1.9717123924624265, + "learning_rate": 1.2001029324326087e-06, + "loss": 0.1034, + "step": 2815 + }, + { + "epoch": 0.6407281001137657, + "grad_norm": 1.5832378059806158, + "learning_rate": 1.2000679426938838e-06, + "loss": 0.0655, + "step": 2816 + }, + { + "epoch": 0.6409556313993174, + "grad_norm": 2.5340372908959856, + "learning_rate": 1.2000329412018233e-06, + "loss": 0.1382, + "step": 2817 + }, + { + "epoch": 0.6411831626848692, + "grad_norm": 1.221154684672452, + "learning_rate": 1.1999979279571425e-06, + "loss": 0.1122, + "step": 2818 + }, + { + "epoch": 0.6414106939704209, + "grad_norm": 2.0522471249218666, + "learning_rate": 1.1999629029605572e-06, + "loss": 0.0947, + "step": 2819 + }, + { + "epoch": 0.6416382252559727, + "grad_norm": 1.2769878053312456, + "learning_rate": 1.1999278662127832e-06, + "loss": 0.0864, + "step": 2820 + }, + { + "epoch": 0.6418657565415244, + "grad_norm": 2.2645164597272216, + "learning_rate": 1.1998928177145363e-06, + "loss": 0.1158, + "step": 2821 + }, + { + "epoch": 0.6420932878270762, + "grad_norm": 2.1530974045268034, + "learning_rate": 1.1998577574665334e-06, + "loss": 0.1692, + "step": 2822 + }, + { + "epoch": 0.642320819112628, + "grad_norm": 1.962935725540055, + "learning_rate": 1.1998226854694906e-06, + "loss": 0.1575, + "step": 2823 + }, + { + "epoch": 0.6425483503981797, + "grad_norm": 2.392603352353799, + "learning_rate": 1.1997876017241248e-06, + "loss": 0.1013, + "step": 2824 + }, + { + "epoch": 0.6427758816837316, + "grad_norm": 1.7118896384760522, + "learning_rate": 1.199752506231153e-06, + "loss": 0.1192, + "step": 2825 + }, + { + "epoch": 0.6430034129692833, + "grad_norm": 2.563449128448653, + "learning_rate": 1.1997173989912928e-06, + "loss": 0.1531, + "step": 2826 + }, + { + "epoch": 0.6432309442548351, + "grad_norm": 2.3688066675487196, + "learning_rate": 1.1996822800052614e-06, + "loss": 0.1159, + "step": 2827 + }, + { + "epoch": 0.6434584755403868, + "grad_norm": 1.8278311039923627, + "learning_rate": 1.1996471492737767e-06, + "loss": 0.141, + "step": 2828 + }, + { + "epoch": 0.6436860068259386, + "grad_norm": 1.4624967632207888, + "learning_rate": 1.1996120067975568e-06, + "loss": 0.1154, + "step": 2829 + }, + { + "epoch": 0.6439135381114903, + "grad_norm": 2.2364175356014333, + "learning_rate": 1.1995768525773195e-06, + "loss": 0.1283, + "step": 2830 + }, + { + "epoch": 0.6441410693970421, + "grad_norm": 3.2767392401427062, + "learning_rate": 1.1995416866137837e-06, + "loss": 0.1808, + "step": 2831 + }, + { + "epoch": 0.6443686006825938, + "grad_norm": 2.586327741034491, + "learning_rate": 1.1995065089076682e-06, + "loss": 0.154, + "step": 2832 + }, + { + "epoch": 0.6445961319681456, + "grad_norm": 1.2535014504137996, + "learning_rate": 1.199471319459692e-06, + "loss": 0.1147, + "step": 2833 + }, + { + "epoch": 0.6448236632536973, + "grad_norm": 1.7122556783292748, + "learning_rate": 1.199436118270574e-06, + "loss": 0.1146, + "step": 2834 + }, + { + "epoch": 0.6450511945392492, + "grad_norm": 1.8732917825303554, + "learning_rate": 1.1994009053410336e-06, + "loss": 0.0699, + "step": 2835 + }, + { + "epoch": 0.645278725824801, + "grad_norm": 1.6917793123575706, + "learning_rate": 1.1993656806717906e-06, + "loss": 0.1035, + "step": 2836 + }, + { + "epoch": 0.6455062571103527, + "grad_norm": 1.7398376655862977, + "learning_rate": 1.199330444263565e-06, + "loss": 0.1377, + "step": 2837 + }, + { + "epoch": 0.6457337883959045, + "grad_norm": 3.6823862200780524, + "learning_rate": 1.1992951961170774e-06, + "loss": 0.2285, + "step": 2838 + }, + { + "epoch": 0.6459613196814562, + "grad_norm": 2.7082289619854807, + "learning_rate": 1.1992599362330474e-06, + "loss": 0.1617, + "step": 2839 + }, + { + "epoch": 0.646188850967008, + "grad_norm": 1.911147339712701, + "learning_rate": 1.199224664612196e-06, + "loss": 0.1181, + "step": 2840 + }, + { + "epoch": 0.6464163822525597, + "grad_norm": 2.6044902420279006, + "learning_rate": 1.199189381255244e-06, + "loss": 0.1986, + "step": 2841 + }, + { + "epoch": 0.6466439135381115, + "grad_norm": 3.1518629789469284, + "learning_rate": 1.199154086162913e-06, + "loss": 0.2413, + "step": 2842 + }, + { + "epoch": 0.6468714448236632, + "grad_norm": 1.64619969788916, + "learning_rate": 1.1991187793359239e-06, + "loss": 0.1469, + "step": 2843 + }, + { + "epoch": 0.647098976109215, + "grad_norm": 3.4988801724543994, + "learning_rate": 1.1990834607749981e-06, + "loss": 0.0906, + "step": 2844 + }, + { + "epoch": 0.6473265073947668, + "grad_norm": 1.4716329070831264, + "learning_rate": 1.199048130480858e-06, + "loss": 0.1628, + "step": 2845 + }, + { + "epoch": 0.6475540386803186, + "grad_norm": 1.7378731147169513, + "learning_rate": 1.1990127884542251e-06, + "loss": 0.1317, + "step": 2846 + }, + { + "epoch": 0.6477815699658703, + "grad_norm": 2.249048740116717, + "learning_rate": 1.1989774346958225e-06, + "loss": 0.0967, + "step": 2847 + }, + { + "epoch": 0.6480091012514221, + "grad_norm": 1.1255899263681175, + "learning_rate": 1.1989420692063723e-06, + "loss": 0.1335, + "step": 2848 + }, + { + "epoch": 0.6482366325369738, + "grad_norm": 1.8914124945151989, + "learning_rate": 1.198906691986597e-06, + "loss": 0.1352, + "step": 2849 + }, + { + "epoch": 0.6484641638225256, + "grad_norm": 1.8849598926256956, + "learning_rate": 1.1988713030372202e-06, + "loss": 0.1649, + "step": 2850 + }, + { + "epoch": 0.6486916951080773, + "grad_norm": 2.417708669547903, + "learning_rate": 1.198835902358965e-06, + "loss": 0.111, + "step": 2851 + }, + { + "epoch": 0.6489192263936291, + "grad_norm": 0.8906821438798999, + "learning_rate": 1.1988004899525547e-06, + "loss": 0.0551, + "step": 2852 + }, + { + "epoch": 0.6491467576791808, + "grad_norm": 2.3567391474624837, + "learning_rate": 1.1987650658187133e-06, + "loss": 0.1235, + "step": 2853 + }, + { + "epoch": 0.6493742889647327, + "grad_norm": 1.0836164298505087, + "learning_rate": 1.1987296299581648e-06, + "loss": 0.1158, + "step": 2854 + }, + { + "epoch": 0.6496018202502845, + "grad_norm": 2.4559644918599246, + "learning_rate": 1.1986941823716333e-06, + "loss": 0.1147, + "step": 2855 + }, + { + "epoch": 0.6498293515358362, + "grad_norm": 2.2390322585917604, + "learning_rate": 1.1986587230598437e-06, + "loss": 0.09, + "step": 2856 + }, + { + "epoch": 0.650056882821388, + "grad_norm": 1.344402365377071, + "learning_rate": 1.19862325202352e-06, + "loss": 0.1436, + "step": 2857 + }, + { + "epoch": 0.6502844141069397, + "grad_norm": 1.426926920816382, + "learning_rate": 1.198587769263388e-06, + "loss": 0.087, + "step": 2858 + }, + { + "epoch": 0.6505119453924915, + "grad_norm": 2.0719268939322713, + "learning_rate": 1.198552274780172e-06, + "loss": 0.1156, + "step": 2859 + }, + { + "epoch": 0.6507394766780432, + "grad_norm": 1.4286233376888462, + "learning_rate": 1.1985167685745982e-06, + "loss": 0.107, + "step": 2860 + }, + { + "epoch": 0.650967007963595, + "grad_norm": 1.5231976360847568, + "learning_rate": 1.198481250647392e-06, + "loss": 0.0846, + "step": 2861 + }, + { + "epoch": 0.6511945392491467, + "grad_norm": 1.595972966209539, + "learning_rate": 1.1984457209992792e-06, + "loss": 0.1834, + "step": 2862 + }, + { + "epoch": 0.6514220705346985, + "grad_norm": 1.7030298326486626, + "learning_rate": 1.1984101796309862e-06, + "loss": 0.1557, + "step": 2863 + }, + { + "epoch": 0.6516496018202503, + "grad_norm": 2.6294697295392324, + "learning_rate": 1.1983746265432392e-06, + "loss": 0.206, + "step": 2864 + }, + { + "epoch": 0.6518771331058021, + "grad_norm": 1.9056557077654461, + "learning_rate": 1.1983390617367649e-06, + "loss": 0.1442, + "step": 2865 + }, + { + "epoch": 0.6521046643913538, + "grad_norm": 3.364370208484872, + "learning_rate": 1.1983034852122902e-06, + "loss": 0.2494, + "step": 2866 + }, + { + "epoch": 0.6523321956769056, + "grad_norm": 1.5505126489167045, + "learning_rate": 1.1982678969705425e-06, + "loss": 0.0815, + "step": 2867 + }, + { + "epoch": 0.6525597269624573, + "grad_norm": 2.3551775420223597, + "learning_rate": 1.1982322970122485e-06, + "loss": 0.1385, + "step": 2868 + }, + { + "epoch": 0.6527872582480091, + "grad_norm": 2.411065323169576, + "learning_rate": 1.1981966853381364e-06, + "loss": 0.1974, + "step": 2869 + }, + { + "epoch": 0.6530147895335608, + "grad_norm": 2.242797006771464, + "learning_rate": 1.1981610619489337e-06, + "loss": 0.1136, + "step": 2870 + }, + { + "epoch": 0.6532423208191126, + "grad_norm": 1.510260522760697, + "learning_rate": 1.1981254268453684e-06, + "loss": 0.0959, + "step": 2871 + }, + { + "epoch": 0.6534698521046644, + "grad_norm": 2.904454351577721, + "learning_rate": 1.1980897800281694e-06, + "loss": 0.2106, + "step": 2872 + }, + { + "epoch": 0.6536973833902161, + "grad_norm": 2.6314570278686444, + "learning_rate": 1.1980541214980646e-06, + "loss": 0.1762, + "step": 2873 + }, + { + "epoch": 0.653924914675768, + "grad_norm": 2.5532198759232023, + "learning_rate": 1.1980184512557833e-06, + "loss": 0.1413, + "step": 2874 + }, + { + "epoch": 0.6541524459613197, + "grad_norm": 2.3410192204822318, + "learning_rate": 1.1979827693020541e-06, + "loss": 0.1382, + "step": 2875 + }, + { + "epoch": 0.6543799772468715, + "grad_norm": 1.7142557174328752, + "learning_rate": 1.1979470756376064e-06, + "loss": 0.1048, + "step": 2876 + }, + { + "epoch": 0.6546075085324232, + "grad_norm": 1.0889093964458614, + "learning_rate": 1.1979113702631697e-06, + "loss": 0.1071, + "step": 2877 + }, + { + "epoch": 0.654835039817975, + "grad_norm": 1.8563209707215407, + "learning_rate": 1.197875653179474e-06, + "loss": 0.088, + "step": 2878 + }, + { + "epoch": 0.6550625711035267, + "grad_norm": 2.7001720187434572, + "learning_rate": 1.1978399243872492e-06, + "loss": 0.1437, + "step": 2879 + }, + { + "epoch": 0.6552901023890785, + "grad_norm": 2.258002197022518, + "learning_rate": 1.1978041838872253e-06, + "loss": 0.1762, + "step": 2880 + }, + { + "epoch": 0.6555176336746302, + "grad_norm": 1.6515838649974548, + "learning_rate": 1.197768431680133e-06, + "loss": 0.1046, + "step": 2881 + }, + { + "epoch": 0.655745164960182, + "grad_norm": 2.4946899822007507, + "learning_rate": 1.197732667766703e-06, + "loss": 0.1262, + "step": 2882 + }, + { + "epoch": 0.6559726962457337, + "grad_norm": 4.602680866014901, + "learning_rate": 1.1976968921476662e-06, + "loss": 0.1744, + "step": 2883 + }, + { + "epoch": 0.6562002275312856, + "grad_norm": 2.360092535792725, + "learning_rate": 1.1976611048237534e-06, + "loss": 0.0931, + "step": 2884 + }, + { + "epoch": 0.6564277588168373, + "grad_norm": 1.483983811262198, + "learning_rate": 1.1976253057956968e-06, + "loss": 0.0807, + "step": 2885 + }, + { + "epoch": 0.6566552901023891, + "grad_norm": 2.842807886661957, + "learning_rate": 1.1975894950642276e-06, + "loss": 0.1619, + "step": 2886 + }, + { + "epoch": 0.6568828213879409, + "grad_norm": 2.486726037530193, + "learning_rate": 1.1975536726300776e-06, + "loss": 0.1901, + "step": 2887 + }, + { + "epoch": 0.6571103526734926, + "grad_norm": 2.049571239290376, + "learning_rate": 1.1975178384939793e-06, + "loss": 0.1047, + "step": 2888 + }, + { + "epoch": 0.6573378839590444, + "grad_norm": 1.7524729733051039, + "learning_rate": 1.197481992656665e-06, + "loss": 0.1674, + "step": 2889 + }, + { + "epoch": 0.6575654152445961, + "grad_norm": 2.080752113172955, + "learning_rate": 1.1974461351188668e-06, + "loss": 0.11, + "step": 2890 + }, + { + "epoch": 0.6577929465301479, + "grad_norm": 1.8490641173720084, + "learning_rate": 1.1974102658813183e-06, + "loss": 0.0879, + "step": 2891 + }, + { + "epoch": 0.6580204778156996, + "grad_norm": 1.5276572210559733, + "learning_rate": 1.1973743849447522e-06, + "loss": 0.1732, + "step": 2892 + }, + { + "epoch": 0.6582480091012515, + "grad_norm": 2.25951432823213, + "learning_rate": 1.197338492309902e-06, + "loss": 0.144, + "step": 2893 + }, + { + "epoch": 0.6584755403868032, + "grad_norm": 1.6131229701226704, + "learning_rate": 1.1973025879775011e-06, + "loss": 0.1721, + "step": 2894 + }, + { + "epoch": 0.658703071672355, + "grad_norm": 1.9025169260113142, + "learning_rate": 1.1972666719482833e-06, + "loss": 0.1233, + "step": 2895 + }, + { + "epoch": 0.6589306029579067, + "grad_norm": 1.9930955290074996, + "learning_rate": 1.197230744222983e-06, + "loss": 0.0945, + "step": 2896 + }, + { + "epoch": 0.6591581342434585, + "grad_norm": 1.6272211241544459, + "learning_rate": 1.1971948048023343e-06, + "loss": 0.1287, + "step": 2897 + }, + { + "epoch": 0.6593856655290102, + "grad_norm": 1.7105829399675778, + "learning_rate": 1.1971588536870717e-06, + "loss": 0.1636, + "step": 2898 + }, + { + "epoch": 0.659613196814562, + "grad_norm": 1.6038784432273976, + "learning_rate": 1.19712289087793e-06, + "loss": 0.1555, + "step": 2899 + }, + { + "epoch": 0.6598407281001137, + "grad_norm": 2.189018947914395, + "learning_rate": 1.197086916375644e-06, + "loss": 0.1999, + "step": 2900 + }, + { + "epoch": 0.6600682593856655, + "grad_norm": 2.1933170276472493, + "learning_rate": 1.1970509301809493e-06, + "loss": 0.0573, + "step": 2901 + }, + { + "epoch": 0.6602957906712172, + "grad_norm": 1.3496992062217439, + "learning_rate": 1.1970149322945812e-06, + "loss": 0.1642, + "step": 2902 + }, + { + "epoch": 0.6605233219567691, + "grad_norm": 1.2111324014689482, + "learning_rate": 1.1969789227172755e-06, + "loss": 0.0847, + "step": 2903 + }, + { + "epoch": 0.6607508532423209, + "grad_norm": 1.5516745653341248, + "learning_rate": 1.1969429014497684e-06, + "loss": 0.1287, + "step": 2904 + }, + { + "epoch": 0.6609783845278726, + "grad_norm": 4.0240616277853904, + "learning_rate": 1.1969068684927954e-06, + "loss": 0.1983, + "step": 2905 + }, + { + "epoch": 0.6612059158134244, + "grad_norm": 2.3056372453856953, + "learning_rate": 1.1968708238470936e-06, + "loss": 0.1888, + "step": 2906 + }, + { + "epoch": 0.6614334470989761, + "grad_norm": 2.8003705602565643, + "learning_rate": 1.1968347675133995e-06, + "loss": 0.1145, + "step": 2907 + }, + { + "epoch": 0.6616609783845279, + "grad_norm": 1.8246199835217014, + "learning_rate": 1.19679869949245e-06, + "loss": 0.212, + "step": 2908 + }, + { + "epoch": 0.6618885096700796, + "grad_norm": 3.15585157724945, + "learning_rate": 1.1967626197849824e-06, + "loss": 0.1361, + "step": 2909 + }, + { + "epoch": 0.6621160409556314, + "grad_norm": 2.387258120107012, + "learning_rate": 1.1967265283917339e-06, + "loss": 0.1856, + "step": 2910 + }, + { + "epoch": 0.6623435722411831, + "grad_norm": 2.1362133210531575, + "learning_rate": 1.1966904253134422e-06, + "loss": 0.1408, + "step": 2911 + }, + { + "epoch": 0.6625711035267349, + "grad_norm": 3.166323287444489, + "learning_rate": 1.1966543105508454e-06, + "loss": 0.1418, + "step": 2912 + }, + { + "epoch": 0.6627986348122867, + "grad_norm": 3.551198157075063, + "learning_rate": 1.1966181841046812e-06, + "loss": 0.1901, + "step": 2913 + }, + { + "epoch": 0.6630261660978385, + "grad_norm": 1.4633198936003367, + "learning_rate": 1.1965820459756882e-06, + "loss": 0.0924, + "step": 2914 + }, + { + "epoch": 0.6632536973833902, + "grad_norm": 2.2531543446585616, + "learning_rate": 1.1965458961646051e-06, + "loss": 0.1933, + "step": 2915 + }, + { + "epoch": 0.663481228668942, + "grad_norm": 2.0085037030014132, + "learning_rate": 1.1965097346721707e-06, + "loss": 0.1298, + "step": 2916 + }, + { + "epoch": 0.6637087599544937, + "grad_norm": 1.25673364297024, + "learning_rate": 1.1964735614991237e-06, + "loss": 0.1956, + "step": 2917 + }, + { + "epoch": 0.6639362912400455, + "grad_norm": 2.8836800293112352, + "learning_rate": 1.196437376646204e-06, + "loss": 0.1327, + "step": 2918 + }, + { + "epoch": 0.6641638225255972, + "grad_norm": 3.1266651348730172, + "learning_rate": 1.1964011801141505e-06, + "loss": 0.2037, + "step": 2919 + }, + { + "epoch": 0.664391353811149, + "grad_norm": 2.619788668251123, + "learning_rate": 1.1963649719037037e-06, + "loss": 0.1558, + "step": 2920 + }, + { + "epoch": 0.6646188850967008, + "grad_norm": 1.7667541052371545, + "learning_rate": 1.196328752015603e-06, + "loss": 0.1748, + "step": 2921 + }, + { + "epoch": 0.6648464163822526, + "grad_norm": 1.3147225620045102, + "learning_rate": 1.1962925204505894e-06, + "loss": 0.0753, + "step": 2922 + }, + { + "epoch": 0.6650739476678044, + "grad_norm": 2.131649606321363, + "learning_rate": 1.1962562772094024e-06, + "loss": 0.2134, + "step": 2923 + }, + { + "epoch": 0.6653014789533561, + "grad_norm": 1.6622774393728745, + "learning_rate": 1.1962200222927836e-06, + "loss": 0.0936, + "step": 2924 + }, + { + "epoch": 0.6655290102389079, + "grad_norm": 1.585708645419218, + "learning_rate": 1.1961837557014736e-06, + "loss": 0.1026, + "step": 2925 + }, + { + "epoch": 0.6657565415244596, + "grad_norm": 1.146941765850051, + "learning_rate": 1.196147477436214e-06, + "loss": 0.1825, + "step": 2926 + }, + { + "epoch": 0.6659840728100114, + "grad_norm": 2.0034538246710287, + "learning_rate": 1.1961111874977455e-06, + "loss": 0.1406, + "step": 2927 + }, + { + "epoch": 0.6662116040955631, + "grad_norm": 2.0528707029612137, + "learning_rate": 1.1960748858868104e-06, + "loss": 0.1129, + "step": 2928 + }, + { + "epoch": 0.6664391353811149, + "grad_norm": 2.2619718630209387, + "learning_rate": 1.1960385726041507e-06, + "loss": 0.2007, + "step": 2929 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 1.5600095714292874, + "learning_rate": 1.1960022476505082e-06, + "loss": 0.0834, + "step": 2930 + }, + { + "epoch": 0.6668941979522184, + "grad_norm": 1.80261641322924, + "learning_rate": 1.1959659110266256e-06, + "loss": 0.105, + "step": 2931 + }, + { + "epoch": 0.6671217292377702, + "grad_norm": 1.7004492925954158, + "learning_rate": 1.1959295627332454e-06, + "loss": 0.1787, + "step": 2932 + }, + { + "epoch": 0.667349260523322, + "grad_norm": 5.737730720018478, + "learning_rate": 1.1958932027711106e-06, + "loss": 0.2202, + "step": 2933 + }, + { + "epoch": 0.6675767918088737, + "grad_norm": 3.2248265845939916, + "learning_rate": 1.1958568311409643e-06, + "loss": 0.1432, + "step": 2934 + }, + { + "epoch": 0.6678043230944255, + "grad_norm": 1.611673867911579, + "learning_rate": 1.1958204478435497e-06, + "loss": 0.1198, + "step": 2935 + }, + { + "epoch": 0.6680318543799773, + "grad_norm": 1.8159369259356373, + "learning_rate": 1.195784052879611e-06, + "loss": 0.0952, + "step": 2936 + }, + { + "epoch": 0.668259385665529, + "grad_norm": 3.609361552599488, + "learning_rate": 1.195747646249891e-06, + "loss": 0.0806, + "step": 2937 + }, + { + "epoch": 0.6684869169510808, + "grad_norm": 1.9817411361850743, + "learning_rate": 1.1957112279551347e-06, + "loss": 0.1723, + "step": 2938 + }, + { + "epoch": 0.6687144482366325, + "grad_norm": 3.475296677730135, + "learning_rate": 1.195674797996086e-06, + "loss": 0.1626, + "step": 2939 + }, + { + "epoch": 0.6689419795221843, + "grad_norm": 1.0280004164275525, + "learning_rate": 1.1956383563734897e-06, + "loss": 0.0776, + "step": 2940 + }, + { + "epoch": 0.669169510807736, + "grad_norm": 2.0465094431015025, + "learning_rate": 1.1956019030880902e-06, + "loss": 0.2099, + "step": 2941 + }, + { + "epoch": 0.6693970420932879, + "grad_norm": 1.520908408928458, + "learning_rate": 1.1955654381406331e-06, + "loss": 0.1446, + "step": 2942 + }, + { + "epoch": 0.6696245733788396, + "grad_norm": 1.8121961010303413, + "learning_rate": 1.1955289615318632e-06, + "loss": 0.1224, + "step": 2943 + }, + { + "epoch": 0.6698521046643914, + "grad_norm": 2.0001610941499375, + "learning_rate": 1.1954924732625264e-06, + "loss": 0.1547, + "step": 2944 + }, + { + "epoch": 0.6700796359499431, + "grad_norm": 1.6700743442908508, + "learning_rate": 1.1954559733333681e-06, + "loss": 0.0778, + "step": 2945 + }, + { + "epoch": 0.6703071672354949, + "grad_norm": 2.3975672802355428, + "learning_rate": 1.1954194617451345e-06, + "loss": 0.139, + "step": 2946 + }, + { + "epoch": 0.6705346985210466, + "grad_norm": 1.8024278725077454, + "learning_rate": 1.1953829384985716e-06, + "loss": 0.113, + "step": 2947 + }, + { + "epoch": 0.6707622298065984, + "grad_norm": 3.145969371692759, + "learning_rate": 1.1953464035944262e-06, + "loss": 0.1883, + "step": 2948 + }, + { + "epoch": 0.6709897610921501, + "grad_norm": 14.466120052896613, + "learning_rate": 1.195309857033445e-06, + "loss": 0.1307, + "step": 2949 + }, + { + "epoch": 0.6712172923777019, + "grad_norm": 1.5503689850377973, + "learning_rate": 1.1952732988163745e-06, + "loss": 0.1009, + "step": 2950 + }, + { + "epoch": 0.6714448236632536, + "grad_norm": 1.9518492837230381, + "learning_rate": 1.1952367289439624e-06, + "loss": 0.1653, + "step": 2951 + }, + { + "epoch": 0.6716723549488055, + "grad_norm": 2.3329213171041734, + "learning_rate": 1.1952001474169558e-06, + "loss": 0.138, + "step": 2952 + }, + { + "epoch": 0.6718998862343573, + "grad_norm": 1.7820092144519726, + "learning_rate": 1.1951635542361025e-06, + "loss": 0.14, + "step": 2953 + }, + { + "epoch": 0.672127417519909, + "grad_norm": 3.0377950533141385, + "learning_rate": 1.1951269494021503e-06, + "loss": 0.1433, + "step": 2954 + }, + { + "epoch": 0.6723549488054608, + "grad_norm": 1.3355396724116602, + "learning_rate": 1.1950903329158475e-06, + "loss": 0.0866, + "step": 2955 + }, + { + "epoch": 0.6725824800910125, + "grad_norm": 1.593734083800843, + "learning_rate": 1.1950537047779424e-06, + "loss": 0.1, + "step": 2956 + }, + { + "epoch": 0.6728100113765643, + "grad_norm": 1.5275734610546747, + "learning_rate": 1.1950170649891836e-06, + "loss": 0.079, + "step": 2957 + }, + { + "epoch": 0.673037542662116, + "grad_norm": 1.8841850472581299, + "learning_rate": 1.1949804135503196e-06, + "loss": 0.1615, + "step": 2958 + }, + { + "epoch": 0.6732650739476678, + "grad_norm": 1.1481325273828376, + "learning_rate": 1.1949437504621e-06, + "loss": 0.0845, + "step": 2959 + }, + { + "epoch": 0.6734926052332195, + "grad_norm": 1.6878277522438039, + "learning_rate": 1.194907075725274e-06, + "loss": 0.1545, + "step": 2960 + }, + { + "epoch": 0.6737201365187714, + "grad_norm": 2.1451059139863413, + "learning_rate": 1.1948703893405911e-06, + "loss": 0.1137, + "step": 2961 + }, + { + "epoch": 0.6739476678043231, + "grad_norm": 2.215251205467777, + "learning_rate": 1.194833691308801e-06, + "loss": 0.1348, + "step": 2962 + }, + { + "epoch": 0.6741751990898749, + "grad_norm": 2.906612033660968, + "learning_rate": 1.194796981630654e-06, + "loss": 0.1405, + "step": 2963 + }, + { + "epoch": 0.6744027303754266, + "grad_norm": 1.6397552216723599, + "learning_rate": 1.1947602603069002e-06, + "loss": 0.1068, + "step": 2964 + }, + { + "epoch": 0.6746302616609784, + "grad_norm": 3.1336175876135868, + "learning_rate": 1.19472352733829e-06, + "loss": 0.1591, + "step": 2965 + }, + { + "epoch": 0.6748577929465301, + "grad_norm": 1.5562237586152252, + "learning_rate": 1.1946867827255744e-06, + "loss": 0.11, + "step": 2966 + }, + { + "epoch": 0.6750853242320819, + "grad_norm": 2.282197724231165, + "learning_rate": 1.1946500264695044e-06, + "loss": 0.2207, + "step": 2967 + }, + { + "epoch": 0.6753128555176336, + "grad_norm": 1.9433531473768753, + "learning_rate": 1.194613258570831e-06, + "loss": 0.1495, + "step": 2968 + }, + { + "epoch": 0.6755403868031854, + "grad_norm": 1.6130949175423062, + "learning_rate": 1.1945764790303059e-06, + "loss": 0.1114, + "step": 2969 + }, + { + "epoch": 0.6757679180887372, + "grad_norm": 2.2200546809210087, + "learning_rate": 1.1945396878486805e-06, + "loss": 0.1359, + "step": 2970 + }, + { + "epoch": 0.675995449374289, + "grad_norm": 2.3411085380950403, + "learning_rate": 1.194502885026707e-06, + "loss": 0.174, + "step": 2971 + }, + { + "epoch": 0.6762229806598408, + "grad_norm": 1.0390908830744956, + "learning_rate": 1.1944660705651375e-06, + "loss": 0.1238, + "step": 2972 + }, + { + "epoch": 0.6764505119453925, + "grad_norm": 1.7477836354627838, + "learning_rate": 1.1944292444647248e-06, + "loss": 0.1505, + "step": 2973 + }, + { + "epoch": 0.6766780432309443, + "grad_norm": 1.5463097681692215, + "learning_rate": 1.1943924067262208e-06, + "loss": 0.0973, + "step": 2974 + }, + { + "epoch": 0.676905574516496, + "grad_norm": 2.0006077731515184, + "learning_rate": 1.194355557350379e-06, + "loss": 0.1278, + "step": 2975 + }, + { + "epoch": 0.6771331058020478, + "grad_norm": 2.428149037379935, + "learning_rate": 1.1943186963379522e-06, + "loss": 0.2203, + "step": 2976 + }, + { + "epoch": 0.6773606370875995, + "grad_norm": 1.691046625500558, + "learning_rate": 1.194281823689694e-06, + "loss": 0.127, + "step": 2977 + }, + { + "epoch": 0.6775881683731513, + "grad_norm": 3.951131260976323, + "learning_rate": 1.194244939406358e-06, + "loss": 0.1884, + "step": 2978 + }, + { + "epoch": 0.677815699658703, + "grad_norm": 2.579071481165006, + "learning_rate": 1.1942080434886978e-06, + "loss": 0.1256, + "step": 2979 + }, + { + "epoch": 0.6780432309442548, + "grad_norm": 1.4931981146612179, + "learning_rate": 1.1941711359374678e-06, + "loss": 0.1169, + "step": 2980 + }, + { + "epoch": 0.6782707622298066, + "grad_norm": 1.2787114819440706, + "learning_rate": 1.194134216753422e-06, + "loss": 0.089, + "step": 2981 + }, + { + "epoch": 0.6784982935153584, + "grad_norm": 2.7311666892685196, + "learning_rate": 1.1940972859373151e-06, + "loss": 0.1461, + "step": 2982 + }, + { + "epoch": 0.6787258248009101, + "grad_norm": 1.8785492720960564, + "learning_rate": 1.194060343489902e-06, + "loss": 0.1885, + "step": 2983 + }, + { + "epoch": 0.6789533560864619, + "grad_norm": 2.724028446550519, + "learning_rate": 1.1940233894119377e-06, + "loss": 0.1386, + "step": 2984 + }, + { + "epoch": 0.6791808873720137, + "grad_norm": 1.6619664728697439, + "learning_rate": 1.1939864237041774e-06, + "loss": 0.1606, + "step": 2985 + }, + { + "epoch": 0.6794084186575654, + "grad_norm": 1.8345758168918145, + "learning_rate": 1.1939494463673767e-06, + "loss": 0.1092, + "step": 2986 + }, + { + "epoch": 0.6796359499431172, + "grad_norm": 2.981218053944757, + "learning_rate": 1.1939124574022914e-06, + "loss": 0.1748, + "step": 2987 + }, + { + "epoch": 0.6798634812286689, + "grad_norm": 2.7677489635883123, + "learning_rate": 1.1938754568096771e-06, + "loss": 0.2211, + "step": 2988 + }, + { + "epoch": 0.6800910125142207, + "grad_norm": 0.9903369979817118, + "learning_rate": 1.1938384445902905e-06, + "loss": 0.085, + "step": 2989 + }, + { + "epoch": 0.6803185437997725, + "grad_norm": 1.219724518175553, + "learning_rate": 1.1938014207448877e-06, + "loss": 0.0765, + "step": 2990 + }, + { + "epoch": 0.6805460750853243, + "grad_norm": 1.4640170722241062, + "learning_rate": 1.1937643852742258e-06, + "loss": 0.0597, + "step": 2991 + }, + { + "epoch": 0.680773606370876, + "grad_norm": 3.2799154475490773, + "learning_rate": 1.1937273381790615e-06, + "loss": 0.1073, + "step": 2992 + }, + { + "epoch": 0.6810011376564278, + "grad_norm": 1.545857022318275, + "learning_rate": 1.1936902794601518e-06, + "loss": 0.09, + "step": 2993 + }, + { + "epoch": 0.6812286689419795, + "grad_norm": 2.248965391138622, + "learning_rate": 1.1936532091182544e-06, + "loss": 0.1727, + "step": 2994 + }, + { + "epoch": 0.6814562002275313, + "grad_norm": 1.298847302563234, + "learning_rate": 1.1936161271541268e-06, + "loss": 0.1177, + "step": 2995 + }, + { + "epoch": 0.681683731513083, + "grad_norm": 1.7590140590212042, + "learning_rate": 1.1935790335685272e-06, + "loss": 0.1594, + "step": 2996 + }, + { + "epoch": 0.6819112627986348, + "grad_norm": 2.558222426527528, + "learning_rate": 1.193541928362213e-06, + "loss": 0.1225, + "step": 2997 + }, + { + "epoch": 0.6821387940841865, + "grad_norm": 2.283950297920294, + "learning_rate": 1.1935048115359432e-06, + "loss": 0.1604, + "step": 2998 + }, + { + "epoch": 0.6823663253697383, + "grad_norm": 1.7491142515816718, + "learning_rate": 1.1934676830904763e-06, + "loss": 0.155, + "step": 2999 + }, + { + "epoch": 0.6825938566552902, + "grad_norm": 2.984227416216763, + "learning_rate": 1.193430543026571e-06, + "loss": 0.169, + "step": 3000 + }, + { + "epoch": 0.6828213879408419, + "grad_norm": 2.1040763156780544, + "learning_rate": 1.1933933913449867e-06, + "loss": 0.1525, + "step": 3001 + }, + { + "epoch": 0.6830489192263937, + "grad_norm": 2.1389125690606186, + "learning_rate": 1.193356228046482e-06, + "loss": 0.1844, + "step": 3002 + }, + { + "epoch": 0.6832764505119454, + "grad_norm": 3.0465279486690147, + "learning_rate": 1.1933190531318172e-06, + "loss": 0.1603, + "step": 3003 + }, + { + "epoch": 0.6835039817974972, + "grad_norm": 1.9423392499102479, + "learning_rate": 1.1932818666017516e-06, + "loss": 0.1198, + "step": 3004 + }, + { + "epoch": 0.6837315130830489, + "grad_norm": 2.7426424574978308, + "learning_rate": 1.1932446684570455e-06, + "loss": 0.1811, + "step": 3005 + }, + { + "epoch": 0.6839590443686007, + "grad_norm": 2.3988027673451424, + "learning_rate": 1.1932074586984592e-06, + "loss": 0.1173, + "step": 3006 + }, + { + "epoch": 0.6841865756541524, + "grad_norm": 1.757923475127038, + "learning_rate": 1.1931702373267527e-06, + "loss": 0.0871, + "step": 3007 + }, + { + "epoch": 0.6844141069397042, + "grad_norm": 1.9363103132107797, + "learning_rate": 1.1931330043426872e-06, + "loss": 0.212, + "step": 3008 + }, + { + "epoch": 0.6846416382252559, + "grad_norm": 1.9686025748503317, + "learning_rate": 1.1930957597470238e-06, + "loss": 0.1664, + "step": 3009 + }, + { + "epoch": 0.6848691695108078, + "grad_norm": 1.726511800870621, + "learning_rate": 1.1930585035405235e-06, + "loss": 0.1566, + "step": 3010 + }, + { + "epoch": 0.6850967007963595, + "grad_norm": 2.6585618793727046, + "learning_rate": 1.1930212357239475e-06, + "loss": 0.1212, + "step": 3011 + }, + { + "epoch": 0.6853242320819113, + "grad_norm": 1.910724842176571, + "learning_rate": 1.1929839562980579e-06, + "loss": 0.1618, + "step": 3012 + }, + { + "epoch": 0.685551763367463, + "grad_norm": 1.903571460689554, + "learning_rate": 1.1929466652636164e-06, + "loss": 0.1138, + "step": 3013 + }, + { + "epoch": 0.6857792946530148, + "grad_norm": 2.3424987856505846, + "learning_rate": 1.1929093626213852e-06, + "loss": 0.208, + "step": 3014 + }, + { + "epoch": 0.6860068259385665, + "grad_norm": 1.7442259260132478, + "learning_rate": 1.1928720483721269e-06, + "loss": 0.1205, + "step": 3015 + }, + { + "epoch": 0.6862343572241183, + "grad_norm": 2.3431955332980343, + "learning_rate": 1.1928347225166035e-06, + "loss": 0.0955, + "step": 3016 + }, + { + "epoch": 0.68646188850967, + "grad_norm": 2.3675636221872574, + "learning_rate": 1.1927973850555785e-06, + "loss": 0.1031, + "step": 3017 + }, + { + "epoch": 0.6866894197952218, + "grad_norm": 2.3230512432118755, + "learning_rate": 1.192760035989815e-06, + "loss": 0.1114, + "step": 3018 + }, + { + "epoch": 0.6869169510807736, + "grad_norm": 2.2869670912635818, + "learning_rate": 1.192722675320076e-06, + "loss": 0.1193, + "step": 3019 + }, + { + "epoch": 0.6871444823663254, + "grad_norm": 2.077036790853124, + "learning_rate": 1.1926853030471253e-06, + "loss": 0.1072, + "step": 3020 + }, + { + "epoch": 0.6873720136518772, + "grad_norm": 1.745378445366266, + "learning_rate": 1.1926479191717267e-06, + "loss": 0.1406, + "step": 3021 + }, + { + "epoch": 0.6875995449374289, + "grad_norm": 1.0136707384794403, + "learning_rate": 1.1926105236946443e-06, + "loss": 0.0706, + "step": 3022 + }, + { + "epoch": 0.6878270762229807, + "grad_norm": 2.3015181855491256, + "learning_rate": 1.192573116616642e-06, + "loss": 0.1252, + "step": 3023 + }, + { + "epoch": 0.6880546075085324, + "grad_norm": 2.1547079811494805, + "learning_rate": 1.192535697938485e-06, + "loss": 0.1187, + "step": 3024 + }, + { + "epoch": 0.6882821387940842, + "grad_norm": 1.822837821217239, + "learning_rate": 1.1924982676609377e-06, + "loss": 0.1658, + "step": 3025 + }, + { + "epoch": 0.6885096700796359, + "grad_norm": 2.124153777097937, + "learning_rate": 1.1924608257847651e-06, + "loss": 0.1041, + "step": 3026 + }, + { + "epoch": 0.6887372013651877, + "grad_norm": 1.0872503933411688, + "learning_rate": 1.1924233723107322e-06, + "loss": 0.1241, + "step": 3027 + }, + { + "epoch": 0.6889647326507394, + "grad_norm": 1.29773367128732, + "learning_rate": 1.1923859072396051e-06, + "loss": 0.1299, + "step": 3028 + }, + { + "epoch": 0.6891922639362913, + "grad_norm": 1.0511026164299604, + "learning_rate": 1.1923484305721489e-06, + "loss": 0.0724, + "step": 3029 + }, + { + "epoch": 0.689419795221843, + "grad_norm": 1.990350741929456, + "learning_rate": 1.19231094230913e-06, + "loss": 0.1034, + "step": 3030 + }, + { + "epoch": 0.6896473265073948, + "grad_norm": 1.3035401289261104, + "learning_rate": 1.1922734424513144e-06, + "loss": 0.1207, + "step": 3031 + }, + { + "epoch": 0.6898748577929465, + "grad_norm": 1.9244132444178317, + "learning_rate": 1.1922359309994685e-06, + "loss": 0.0979, + "step": 3032 + }, + { + "epoch": 0.6901023890784983, + "grad_norm": 1.1055660037716095, + "learning_rate": 1.192198407954359e-06, + "loss": 0.0855, + "step": 3033 + }, + { + "epoch": 0.69032992036405, + "grad_norm": 2.862275522322784, + "learning_rate": 1.192160873316753e-06, + "loss": 0.1768, + "step": 3034 + }, + { + "epoch": 0.6905574516496018, + "grad_norm": 2.1121426133206027, + "learning_rate": 1.1921233270874174e-06, + "loss": 0.1465, + "step": 3035 + }, + { + "epoch": 0.6907849829351536, + "grad_norm": 1.4975621183670067, + "learning_rate": 1.1920857692671196e-06, + "loss": 0.1224, + "step": 3036 + }, + { + "epoch": 0.6910125142207053, + "grad_norm": 1.4957707166752725, + "learning_rate": 1.192048199856627e-06, + "loss": 0.1212, + "step": 3037 + }, + { + "epoch": 0.6912400455062571, + "grad_norm": 1.9181007196013113, + "learning_rate": 1.192010618856708e-06, + "loss": 0.1744, + "step": 3038 + }, + { + "epoch": 0.6914675767918089, + "grad_norm": 1.689194988103543, + "learning_rate": 1.1919730262681304e-06, + "loss": 0.1563, + "step": 3039 + }, + { + "epoch": 0.6916951080773607, + "grad_norm": 1.989563122278693, + "learning_rate": 1.1919354220916624e-06, + "loss": 0.1061, + "step": 3040 + }, + { + "epoch": 0.6919226393629124, + "grad_norm": 3.094413273343531, + "learning_rate": 1.1918978063280726e-06, + "loss": 0.2185, + "step": 3041 + }, + { + "epoch": 0.6921501706484642, + "grad_norm": 1.9118711208773385, + "learning_rate": 1.1918601789781299e-06, + "loss": 0.1, + "step": 3042 + }, + { + "epoch": 0.6923777019340159, + "grad_norm": 2.826520192849369, + "learning_rate": 1.1918225400426032e-06, + "loss": 0.1389, + "step": 3043 + }, + { + "epoch": 0.6926052332195677, + "grad_norm": 2.32332931486, + "learning_rate": 1.191784889522262e-06, + "loss": 0.124, + "step": 3044 + }, + { + "epoch": 0.6928327645051194, + "grad_norm": 3.3162758751367267, + "learning_rate": 1.1917472274178757e-06, + "loss": 0.1755, + "step": 3045 + }, + { + "epoch": 0.6930602957906712, + "grad_norm": 1.7652192594849543, + "learning_rate": 1.191709553730214e-06, + "loss": 0.136, + "step": 3046 + }, + { + "epoch": 0.6932878270762229, + "grad_norm": 1.9194876138624921, + "learning_rate": 1.1916718684600469e-06, + "loss": 0.1409, + "step": 3047 + }, + { + "epoch": 0.6935153583617747, + "grad_norm": 1.5068265619508907, + "learning_rate": 1.1916341716081446e-06, + "loss": 0.1221, + "step": 3048 + }, + { + "epoch": 0.6937428896473266, + "grad_norm": 2.5710593050636157, + "learning_rate": 1.1915964631752775e-06, + "loss": 0.0891, + "step": 3049 + }, + { + "epoch": 0.6939704209328783, + "grad_norm": 1.8315542955459863, + "learning_rate": 1.1915587431622164e-06, + "loss": 0.1503, + "step": 3050 + }, + { + "epoch": 0.6941979522184301, + "grad_norm": 1.98668392468283, + "learning_rate": 1.1915210115697324e-06, + "loss": 0.1622, + "step": 3051 + }, + { + "epoch": 0.6944254835039818, + "grad_norm": 1.5107124398660574, + "learning_rate": 1.1914832683985962e-06, + "loss": 0.0928, + "step": 3052 + }, + { + "epoch": 0.6946530147895336, + "grad_norm": 3.73846147303143, + "learning_rate": 1.1914455136495796e-06, + "loss": 0.2613, + "step": 3053 + }, + { + "epoch": 0.6948805460750853, + "grad_norm": 2.8035148694959453, + "learning_rate": 1.191407747323454e-06, + "loss": 0.1517, + "step": 3054 + }, + { + "epoch": 0.6951080773606371, + "grad_norm": 1.895552274543866, + "learning_rate": 1.1913699694209914e-06, + "loss": 0.0942, + "step": 3055 + }, + { + "epoch": 0.6953356086461888, + "grad_norm": 1.7459829471709587, + "learning_rate": 1.191332179942964e-06, + "loss": 0.1396, + "step": 3056 + }, + { + "epoch": 0.6955631399317406, + "grad_norm": 3.815224090825608, + "learning_rate": 1.1912943788901438e-06, + "loss": 0.2051, + "step": 3057 + }, + { + "epoch": 0.6957906712172923, + "grad_norm": 1.6661183563844684, + "learning_rate": 1.191256566263304e-06, + "loss": 0.1971, + "step": 3058 + }, + { + "epoch": 0.6960182025028442, + "grad_norm": 1.8000907111995663, + "learning_rate": 1.1912187420632165e-06, + "loss": 0.1623, + "step": 3059 + }, + { + "epoch": 0.6962457337883959, + "grad_norm": 1.6926787312264402, + "learning_rate": 1.1911809062906552e-06, + "loss": 0.1535, + "step": 3060 + }, + { + "epoch": 0.6964732650739477, + "grad_norm": 2.146640045867537, + "learning_rate": 1.1911430589463931e-06, + "loss": 0.1584, + "step": 3061 + }, + { + "epoch": 0.6967007963594994, + "grad_norm": 2.573487768453337, + "learning_rate": 1.1911052000312038e-06, + "loss": 0.1839, + "step": 3062 + }, + { + "epoch": 0.6969283276450512, + "grad_norm": 1.2062041272330437, + "learning_rate": 1.1910673295458607e-06, + "loss": 0.1174, + "step": 3063 + }, + { + "epoch": 0.697155858930603, + "grad_norm": 2.2446368133366015, + "learning_rate": 1.1910294474911382e-06, + "loss": 0.1186, + "step": 3064 + }, + { + "epoch": 0.6973833902161547, + "grad_norm": 1.832728201805095, + "learning_rate": 1.1909915538678105e-06, + "loss": 0.201, + "step": 3065 + }, + { + "epoch": 0.6976109215017064, + "grad_norm": 0.8873694701683021, + "learning_rate": 1.1909536486766522e-06, + "loss": 0.0817, + "step": 3066 + }, + { + "epoch": 0.6978384527872582, + "grad_norm": 1.6140383266242742, + "learning_rate": 1.1909157319184373e-06, + "loss": 0.1116, + "step": 3067 + }, + { + "epoch": 0.6980659840728101, + "grad_norm": 1.571905366138055, + "learning_rate": 1.1908778035939416e-06, + "loss": 0.1295, + "step": 3068 + }, + { + "epoch": 0.6982935153583618, + "grad_norm": 3.001382201945936, + "learning_rate": 1.1908398637039398e-06, + "loss": 0.0721, + "step": 3069 + }, + { + "epoch": 0.6985210466439136, + "grad_norm": 2.750456856631494, + "learning_rate": 1.1908019122492077e-06, + "loss": 0.1422, + "step": 3070 + }, + { + "epoch": 0.6987485779294653, + "grad_norm": 2.7469566413459425, + "learning_rate": 1.1907639492305205e-06, + "loss": 0.1605, + "step": 3071 + }, + { + "epoch": 0.6989761092150171, + "grad_norm": 1.573510557079222, + "learning_rate": 1.1907259746486547e-06, + "loss": 0.0679, + "step": 3072 + }, + { + "epoch": 0.6992036405005688, + "grad_norm": 1.535165170922043, + "learning_rate": 1.1906879885043856e-06, + "loss": 0.1303, + "step": 3073 + }, + { + "epoch": 0.6994311717861206, + "grad_norm": 1.51597365006242, + "learning_rate": 1.1906499907984903e-06, + "loss": 0.1091, + "step": 3074 + }, + { + "epoch": 0.6996587030716723, + "grad_norm": 1.9433457342834564, + "learning_rate": 1.190611981531745e-06, + "loss": 0.1875, + "step": 3075 + }, + { + "epoch": 0.6998862343572241, + "grad_norm": 1.3852493508640962, + "learning_rate": 1.1905739607049267e-06, + "loss": 0.11, + "step": 3076 + }, + { + "epoch": 0.7001137656427758, + "grad_norm": 1.6634488720193932, + "learning_rate": 1.1905359283188126e-06, + "loss": 0.0649, + "step": 3077 + }, + { + "epoch": 0.7003412969283277, + "grad_norm": 1.5075356462841907, + "learning_rate": 1.1904978843741796e-06, + "loss": 0.0771, + "step": 3078 + }, + { + "epoch": 0.7005688282138794, + "grad_norm": 2.8860526665616706, + "learning_rate": 1.1904598288718055e-06, + "loss": 0.2098, + "step": 3079 + }, + { + "epoch": 0.7007963594994312, + "grad_norm": 1.3469428884610732, + "learning_rate": 1.1904217618124684e-06, + "loss": 0.1217, + "step": 3080 + }, + { + "epoch": 0.701023890784983, + "grad_norm": 1.534126034264905, + "learning_rate": 1.1903836831969458e-06, + "loss": 0.1307, + "step": 3081 + }, + { + "epoch": 0.7012514220705347, + "grad_norm": 1.036473399987068, + "learning_rate": 1.190345593026016e-06, + "loss": 0.0627, + "step": 3082 + }, + { + "epoch": 0.7014789533560865, + "grad_norm": 2.8331577249200866, + "learning_rate": 1.1903074913004579e-06, + "loss": 0.2588, + "step": 3083 + }, + { + "epoch": 0.7017064846416382, + "grad_norm": 2.6619142781192973, + "learning_rate": 1.19026937802105e-06, + "loss": 0.2616, + "step": 3084 + }, + { + "epoch": 0.70193401592719, + "grad_norm": 1.5193886249818858, + "learning_rate": 1.1902312531885712e-06, + "loss": 0.0836, + "step": 3085 + }, + { + "epoch": 0.7021615472127417, + "grad_norm": 1.9617101099614265, + "learning_rate": 1.1901931168038007e-06, + "loss": 0.1756, + "step": 3086 + }, + { + "epoch": 0.7023890784982935, + "grad_norm": 1.5141276225062847, + "learning_rate": 1.190154968867518e-06, + "loss": 0.0929, + "step": 3087 + }, + { + "epoch": 0.7026166097838453, + "grad_norm": 1.4372083252193162, + "learning_rate": 1.190116809380503e-06, + "loss": 0.0679, + "step": 3088 + }, + { + "epoch": 0.7028441410693971, + "grad_norm": 1.5581282033103077, + "learning_rate": 1.190078638343535e-06, + "loss": 0.1009, + "step": 3089 + }, + { + "epoch": 0.7030716723549488, + "grad_norm": 1.6889378087890292, + "learning_rate": 1.1900404557573948e-06, + "loss": 0.0775, + "step": 3090 + }, + { + "epoch": 0.7032992036405006, + "grad_norm": 2.245465693257155, + "learning_rate": 1.1900022616228624e-06, + "loss": 0.205, + "step": 3091 + }, + { + "epoch": 0.7035267349260523, + "grad_norm": 1.8215824095731956, + "learning_rate": 1.1899640559407186e-06, + "loss": 0.2364, + "step": 3092 + }, + { + "epoch": 0.7037542662116041, + "grad_norm": 1.6915516378179793, + "learning_rate": 1.189925838711744e-06, + "loss": 0.1389, + "step": 3093 + }, + { + "epoch": 0.7039817974971558, + "grad_norm": 1.6928362600822655, + "learning_rate": 1.18988760993672e-06, + "loss": 0.1829, + "step": 3094 + }, + { + "epoch": 0.7042093287827076, + "grad_norm": 2.931008813048908, + "learning_rate": 1.1898493696164279e-06, + "loss": 0.0923, + "step": 3095 + }, + { + "epoch": 0.7044368600682593, + "grad_norm": 1.1154029557797203, + "learning_rate": 1.1898111177516488e-06, + "loss": 0.0757, + "step": 3096 + }, + { + "epoch": 0.7046643913538112, + "grad_norm": 2.711712559556853, + "learning_rate": 1.1897728543431653e-06, + "loss": 0.2867, + "step": 3097 + }, + { + "epoch": 0.704891922639363, + "grad_norm": 2.705535470633257, + "learning_rate": 1.1897345793917589e-06, + "loss": 0.1299, + "step": 3098 + }, + { + "epoch": 0.7051194539249147, + "grad_norm": 1.9634050602340667, + "learning_rate": 1.1896962928982116e-06, + "loss": 0.0929, + "step": 3099 + }, + { + "epoch": 0.7053469852104665, + "grad_norm": 1.925806274179465, + "learning_rate": 1.1896579948633067e-06, + "loss": 0.1962, + "step": 3100 + }, + { + "epoch": 0.7055745164960182, + "grad_norm": 1.4954397243767268, + "learning_rate": 1.1896196852878262e-06, + "loss": 0.1621, + "step": 3101 + }, + { + "epoch": 0.70580204778157, + "grad_norm": 2.3069003202986362, + "learning_rate": 1.1895813641725535e-06, + "loss": 0.106, + "step": 3102 + }, + { + "epoch": 0.7060295790671217, + "grad_norm": 1.7569589217138821, + "learning_rate": 1.1895430315182719e-06, + "loss": 0.0941, + "step": 3103 + }, + { + "epoch": 0.7062571103526735, + "grad_norm": 1.6341215878240687, + "learning_rate": 1.1895046873257644e-06, + "loss": 0.0992, + "step": 3104 + }, + { + "epoch": 0.7064846416382252, + "grad_norm": 2.0620355437898654, + "learning_rate": 1.189466331595815e-06, + "loss": 0.1256, + "step": 3105 + }, + { + "epoch": 0.706712172923777, + "grad_norm": 2.6829703904365045, + "learning_rate": 1.1894279643292074e-06, + "loss": 0.1548, + "step": 3106 + }, + { + "epoch": 0.7069397042093288, + "grad_norm": 2.0525899216404313, + "learning_rate": 1.1893895855267262e-06, + "loss": 0.1323, + "step": 3107 + }, + { + "epoch": 0.7071672354948806, + "grad_norm": 2.3112715545673983, + "learning_rate": 1.1893511951891553e-06, + "loss": 0.1359, + "step": 3108 + }, + { + "epoch": 0.7073947667804323, + "grad_norm": 2.473665918442311, + "learning_rate": 1.1893127933172794e-06, + "loss": 0.1766, + "step": 3109 + }, + { + "epoch": 0.7076222980659841, + "grad_norm": 2.6324961135431195, + "learning_rate": 1.1892743799118838e-06, + "loss": 0.2513, + "step": 3110 + }, + { + "epoch": 0.7078498293515358, + "grad_norm": 2.0428028154854663, + "learning_rate": 1.189235954973753e-06, + "loss": 0.1409, + "step": 3111 + }, + { + "epoch": 0.7080773606370876, + "grad_norm": 1.4954878901352442, + "learning_rate": 1.189197518503673e-06, + "loss": 0.1985, + "step": 3112 + }, + { + "epoch": 0.7083048919226393, + "grad_norm": 1.7220766086751687, + "learning_rate": 1.1891590705024288e-06, + "loss": 0.1868, + "step": 3113 + }, + { + "epoch": 0.7085324232081911, + "grad_norm": 1.5383065815934964, + "learning_rate": 1.1891206109708065e-06, + "loss": 0.1796, + "step": 3114 + }, + { + "epoch": 0.7087599544937428, + "grad_norm": 1.3010401003361285, + "learning_rate": 1.1890821399095917e-06, + "loss": 0.1327, + "step": 3115 + }, + { + "epoch": 0.7089874857792946, + "grad_norm": 2.550340002997386, + "learning_rate": 1.1890436573195714e-06, + "loss": 0.2581, + "step": 3116 + }, + { + "epoch": 0.7092150170648465, + "grad_norm": 5.680850722665896, + "learning_rate": 1.1890051632015315e-06, + "loss": 0.1181, + "step": 3117 + }, + { + "epoch": 0.7094425483503982, + "grad_norm": 2.055782957788213, + "learning_rate": 1.1889666575562593e-06, + "loss": 0.1056, + "step": 3118 + }, + { + "epoch": 0.70967007963595, + "grad_norm": 1.8000182980206194, + "learning_rate": 1.1889281403845413e-06, + "loss": 0.1555, + "step": 3119 + }, + { + "epoch": 0.7098976109215017, + "grad_norm": 1.4086799175857945, + "learning_rate": 1.1888896116871649e-06, + "loss": 0.1702, + "step": 3120 + }, + { + "epoch": 0.7101251422070535, + "grad_norm": 2.825675781255187, + "learning_rate": 1.1888510714649176e-06, + "loss": 0.1559, + "step": 3121 + }, + { + "epoch": 0.7103526734926052, + "grad_norm": 1.516178187565781, + "learning_rate": 1.1888125197185867e-06, + "loss": 0.1457, + "step": 3122 + }, + { + "epoch": 0.710580204778157, + "grad_norm": 2.146464562262945, + "learning_rate": 1.1887739564489608e-06, + "loss": 0.1336, + "step": 3123 + }, + { + "epoch": 0.7108077360637087, + "grad_norm": 2.326773332109297, + "learning_rate": 1.1887353816568277e-06, + "loss": 0.1892, + "step": 3124 + }, + { + "epoch": 0.7110352673492605, + "grad_norm": 2.7537460461611363, + "learning_rate": 1.188696795342976e-06, + "loss": 0.1751, + "step": 3125 + }, + { + "epoch": 0.7112627986348122, + "grad_norm": 1.6542807081780677, + "learning_rate": 1.188658197508194e-06, + "loss": 0.1314, + "step": 3126 + }, + { + "epoch": 0.7114903299203641, + "grad_norm": 2.275977763130455, + "learning_rate": 1.1886195881532705e-06, + "loss": 0.1304, + "step": 3127 + }, + { + "epoch": 0.7117178612059158, + "grad_norm": 2.741548691257323, + "learning_rate": 1.1885809672789953e-06, + "loss": 0.1057, + "step": 3128 + }, + { + "epoch": 0.7119453924914676, + "grad_norm": 2.576016774238181, + "learning_rate": 1.188542334886157e-06, + "loss": 0.144, + "step": 3129 + }, + { + "epoch": 0.7121729237770194, + "grad_norm": 2.17965083103891, + "learning_rate": 1.1885036909755454e-06, + "loss": 0.1345, + "step": 3130 + }, + { + "epoch": 0.7124004550625711, + "grad_norm": 3.896449546672936, + "learning_rate": 1.1884650355479505e-06, + "loss": 0.1726, + "step": 3131 + }, + { + "epoch": 0.7126279863481229, + "grad_norm": 1.7651769113033686, + "learning_rate": 1.1884263686041622e-06, + "loss": 0.1649, + "step": 3132 + }, + { + "epoch": 0.7128555176336746, + "grad_norm": 3.270644942804106, + "learning_rate": 1.1883876901449707e-06, + "loss": 0.1106, + "step": 3133 + }, + { + "epoch": 0.7130830489192264, + "grad_norm": 0.9540720474270146, + "learning_rate": 1.1883490001711667e-06, + "loss": 0.0479, + "step": 3134 + }, + { + "epoch": 0.7133105802047781, + "grad_norm": 1.5281389835207844, + "learning_rate": 1.1883102986835408e-06, + "loss": 0.1265, + "step": 3135 + }, + { + "epoch": 0.71353811149033, + "grad_norm": 1.7275445515308603, + "learning_rate": 1.1882715856828842e-06, + "loss": 0.143, + "step": 3136 + }, + { + "epoch": 0.7137656427758817, + "grad_norm": 1.4760935486211386, + "learning_rate": 1.1882328611699879e-06, + "loss": 0.1286, + "step": 3137 + }, + { + "epoch": 0.7139931740614335, + "grad_norm": 2.061027793910494, + "learning_rate": 1.1881941251456434e-06, + "loss": 0.1199, + "step": 3138 + }, + { + "epoch": 0.7142207053469852, + "grad_norm": 1.6504084382846145, + "learning_rate": 1.1881553776106423e-06, + "loss": 0.1572, + "step": 3139 + }, + { + "epoch": 0.714448236632537, + "grad_norm": 2.871038430361009, + "learning_rate": 1.1881166185657765e-06, + "loss": 0.1957, + "step": 3140 + }, + { + "epoch": 0.7146757679180887, + "grad_norm": 1.7942243510846807, + "learning_rate": 1.1880778480118388e-06, + "loss": 0.0828, + "step": 3141 + }, + { + "epoch": 0.7149032992036405, + "grad_norm": 1.4934742711037803, + "learning_rate": 1.1880390659496207e-06, + "loss": 0.1775, + "step": 3142 + }, + { + "epoch": 0.7151308304891922, + "grad_norm": 1.5497263135986263, + "learning_rate": 1.1880002723799155e-06, + "loss": 0.1242, + "step": 3143 + }, + { + "epoch": 0.715358361774744, + "grad_norm": 1.9759475029193152, + "learning_rate": 1.1879614673035158e-06, + "loss": 0.1099, + "step": 3144 + }, + { + "epoch": 0.7155858930602957, + "grad_norm": 3.273086972853465, + "learning_rate": 1.1879226507212146e-06, + "loss": 0.1725, + "step": 3145 + }, + { + "epoch": 0.7158134243458476, + "grad_norm": 2.692056389406711, + "learning_rate": 1.1878838226338054e-06, + "loss": 0.1295, + "step": 3146 + }, + { + "epoch": 0.7160409556313994, + "grad_norm": 3.0726031872665662, + "learning_rate": 1.187844983042082e-06, + "loss": 0.2144, + "step": 3147 + }, + { + "epoch": 0.7162684869169511, + "grad_norm": 2.296547635428459, + "learning_rate": 1.1878061319468376e-06, + "loss": 0.1033, + "step": 3148 + }, + { + "epoch": 0.7164960182025029, + "grad_norm": 1.109110237947707, + "learning_rate": 1.1877672693488669e-06, + "loss": 0.1335, + "step": 3149 + }, + { + "epoch": 0.7167235494880546, + "grad_norm": 2.1949665597550783, + "learning_rate": 1.1877283952489636e-06, + "loss": 0.1547, + "step": 3150 + }, + { + "epoch": 0.7169510807736064, + "grad_norm": 1.7458101584938244, + "learning_rate": 1.1876895096479226e-06, + "loss": 0.1027, + "step": 3151 + }, + { + "epoch": 0.7171786120591581, + "grad_norm": 2.5253782285983917, + "learning_rate": 1.1876506125465386e-06, + "loss": 0.1677, + "step": 3152 + }, + { + "epoch": 0.7174061433447099, + "grad_norm": 1.7720829535802851, + "learning_rate": 1.1876117039456065e-06, + "loss": 0.0918, + "step": 3153 + }, + { + "epoch": 0.7176336746302616, + "grad_norm": 3.444761369547945, + "learning_rate": 1.1875727838459213e-06, + "loss": 0.1905, + "step": 3154 + }, + { + "epoch": 0.7178612059158134, + "grad_norm": 2.7071394507983033, + "learning_rate": 1.187533852248279e-06, + "loss": 0.1088, + "step": 3155 + }, + { + "epoch": 0.7180887372013652, + "grad_norm": 1.389997018526983, + "learning_rate": 1.1874949091534749e-06, + "loss": 0.1409, + "step": 3156 + }, + { + "epoch": 0.718316268486917, + "grad_norm": 2.4375044265295345, + "learning_rate": 1.1874559545623049e-06, + "loss": 0.1502, + "step": 3157 + }, + { + "epoch": 0.7185437997724687, + "grad_norm": 1.6111336105377863, + "learning_rate": 1.1874169884755654e-06, + "loss": 0.0958, + "step": 3158 + }, + { + "epoch": 0.7187713310580205, + "grad_norm": 2.4785884630082182, + "learning_rate": 1.1873780108940527e-06, + "loss": 0.0979, + "step": 3159 + }, + { + "epoch": 0.7189988623435722, + "grad_norm": 2.309500047954281, + "learning_rate": 1.1873390218185636e-06, + "loss": 0.0857, + "step": 3160 + }, + { + "epoch": 0.719226393629124, + "grad_norm": 1.9974160071993916, + "learning_rate": 1.1873000212498942e-06, + "loss": 0.1217, + "step": 3161 + }, + { + "epoch": 0.7194539249146757, + "grad_norm": 1.2977760267071896, + "learning_rate": 1.1872610091888426e-06, + "loss": 0.0946, + "step": 3162 + }, + { + "epoch": 0.7196814562002275, + "grad_norm": 3.3747565831591646, + "learning_rate": 1.1872219856362057e-06, + "loss": 0.1854, + "step": 3163 + }, + { + "epoch": 0.7199089874857793, + "grad_norm": 1.8430938679177409, + "learning_rate": 1.187182950592781e-06, + "loss": 0.1771, + "step": 3164 + }, + { + "epoch": 0.7201365187713311, + "grad_norm": 1.3979587929565391, + "learning_rate": 1.1871439040593663e-06, + "loss": 0.178, + "step": 3165 + }, + { + "epoch": 0.7203640500568829, + "grad_norm": 1.9035217199652052, + "learning_rate": 1.1871048460367598e-06, + "loss": 0.1594, + "step": 3166 + }, + { + "epoch": 0.7205915813424346, + "grad_norm": 3.1235740653373747, + "learning_rate": 1.1870657765257595e-06, + "loss": 0.1065, + "step": 3167 + }, + { + "epoch": 0.7208191126279864, + "grad_norm": 2.0971860071500523, + "learning_rate": 1.1870266955271645e-06, + "loss": 0.1457, + "step": 3168 + }, + { + "epoch": 0.7210466439135381, + "grad_norm": 2.773067534867043, + "learning_rate": 1.186987603041773e-06, + "loss": 0.2245, + "step": 3169 + }, + { + "epoch": 0.7212741751990899, + "grad_norm": 1.6672498177883812, + "learning_rate": 1.1869484990703839e-06, + "loss": 0.1552, + "step": 3170 + }, + { + "epoch": 0.7215017064846416, + "grad_norm": 1.037913927891429, + "learning_rate": 1.1869093836137968e-06, + "loss": 0.0904, + "step": 3171 + }, + { + "epoch": 0.7217292377701934, + "grad_norm": 2.496116324041325, + "learning_rate": 1.186870256672811e-06, + "loss": 0.1904, + "step": 3172 + }, + { + "epoch": 0.7219567690557451, + "grad_norm": 2.7658500953788, + "learning_rate": 1.1868311182482262e-06, + "loss": 0.1968, + "step": 3173 + }, + { + "epoch": 0.7221843003412969, + "grad_norm": 2.2669129231199143, + "learning_rate": 1.1867919683408421e-06, + "loss": 0.0807, + "step": 3174 + }, + { + "epoch": 0.7224118316268487, + "grad_norm": 2.139798819306396, + "learning_rate": 1.1867528069514591e-06, + "loss": 0.2389, + "step": 3175 + }, + { + "epoch": 0.7226393629124005, + "grad_norm": 1.8865928408275647, + "learning_rate": 1.1867136340808778e-06, + "loss": 0.0814, + "step": 3176 + }, + { + "epoch": 0.7228668941979522, + "grad_norm": 2.0538790027428595, + "learning_rate": 1.1866744497298982e-06, + "loss": 0.1128, + "step": 3177 + }, + { + "epoch": 0.723094425483504, + "grad_norm": 2.3275182667078638, + "learning_rate": 1.1866352538993216e-06, + "loss": 0.1125, + "step": 3178 + }, + { + "epoch": 0.7233219567690558, + "grad_norm": 2.5368864511406133, + "learning_rate": 1.1865960465899492e-06, + "loss": 0.1257, + "step": 3179 + }, + { + "epoch": 0.7235494880546075, + "grad_norm": 2.208159889999602, + "learning_rate": 1.186556827802582e-06, + "loss": 0.1483, + "step": 3180 + }, + { + "epoch": 0.7237770193401593, + "grad_norm": 1.3791871408723215, + "learning_rate": 1.1865175975380218e-06, + "loss": 0.1221, + "step": 3181 + }, + { + "epoch": 0.724004550625711, + "grad_norm": 1.977376136232679, + "learning_rate": 1.18647835579707e-06, + "loss": 0.1394, + "step": 3182 + }, + { + "epoch": 0.7242320819112628, + "grad_norm": 2.4697637699058537, + "learning_rate": 1.186439102580529e-06, + "loss": 0.1847, + "step": 3183 + }, + { + "epoch": 0.7244596131968145, + "grad_norm": 1.8882424053371538, + "learning_rate": 1.1863998378892011e-06, + "loss": 0.0924, + "step": 3184 + }, + { + "epoch": 0.7246871444823664, + "grad_norm": 3.5850536595775604, + "learning_rate": 1.1863605617238885e-06, + "loss": 0.2122, + "step": 3185 + }, + { + "epoch": 0.7249146757679181, + "grad_norm": 1.511139760494177, + "learning_rate": 1.1863212740853941e-06, + "loss": 0.1896, + "step": 3186 + }, + { + "epoch": 0.7251422070534699, + "grad_norm": 2.5563678633022837, + "learning_rate": 1.1862819749745212e-06, + "loss": 0.1247, + "step": 3187 + }, + { + "epoch": 0.7253697383390216, + "grad_norm": 2.3301556693527123, + "learning_rate": 1.1862426643920722e-06, + "loss": 0.1177, + "step": 3188 + }, + { + "epoch": 0.7255972696245734, + "grad_norm": 2.33880572168744, + "learning_rate": 1.1862033423388513e-06, + "loss": 0.177, + "step": 3189 + }, + { + "epoch": 0.7258248009101251, + "grad_norm": 1.3976182704723772, + "learning_rate": 1.1861640088156617e-06, + "loss": 0.0935, + "step": 3190 + }, + { + "epoch": 0.7260523321956769, + "grad_norm": 1.6758681316019886, + "learning_rate": 1.1861246638233077e-06, + "loss": 0.1094, + "step": 3191 + }, + { + "epoch": 0.7262798634812286, + "grad_norm": 1.600929618729876, + "learning_rate": 1.1860853073625931e-06, + "loss": 0.1332, + "step": 3192 + }, + { + "epoch": 0.7265073947667804, + "grad_norm": 1.8021701788400442, + "learning_rate": 1.1860459394343223e-06, + "loss": 0.1689, + "step": 3193 + }, + { + "epoch": 0.7267349260523321, + "grad_norm": 2.1221863719292786, + "learning_rate": 1.1860065600393002e-06, + "loss": 0.1398, + "step": 3194 + }, + { + "epoch": 0.726962457337884, + "grad_norm": 3.5848223977448335, + "learning_rate": 1.1859671691783315e-06, + "loss": 0.1911, + "step": 3195 + }, + { + "epoch": 0.7271899886234358, + "grad_norm": 1.4735878220175993, + "learning_rate": 1.1859277668522209e-06, + "loss": 0.0927, + "step": 3196 + }, + { + "epoch": 0.7274175199089875, + "grad_norm": 2.6836187261265048, + "learning_rate": 1.1858883530617743e-06, + "loss": 0.3431, + "step": 3197 + }, + { + "epoch": 0.7276450511945393, + "grad_norm": 2.8853311096517023, + "learning_rate": 1.185848927807797e-06, + "loss": 0.16, + "step": 3198 + }, + { + "epoch": 0.727872582480091, + "grad_norm": 1.9070608581564725, + "learning_rate": 1.1858094910910945e-06, + "loss": 0.0933, + "step": 3199 + }, + { + "epoch": 0.7281001137656428, + "grad_norm": 1.636161711581521, + "learning_rate": 1.1857700429124733e-06, + "loss": 0.0921, + "step": 3200 + }, + { + "epoch": 0.7283276450511945, + "grad_norm": 1.163604378500458, + "learning_rate": 1.1857305832727395e-06, + "loss": 0.0819, + "step": 3201 + }, + { + "epoch": 0.7285551763367463, + "grad_norm": 1.9828883754527078, + "learning_rate": 1.1856911121726993e-06, + "loss": 0.1576, + "step": 3202 + }, + { + "epoch": 0.728782707622298, + "grad_norm": 3.2149916841430355, + "learning_rate": 1.1856516296131596e-06, + "loss": 0.1322, + "step": 3203 + }, + { + "epoch": 0.7290102389078499, + "grad_norm": 2.0052514996312136, + "learning_rate": 1.1856121355949276e-06, + "loss": 0.214, + "step": 3204 + }, + { + "epoch": 0.7292377701934016, + "grad_norm": 2.5233532450432206, + "learning_rate": 1.18557263011881e-06, + "loss": 0.126, + "step": 3205 + }, + { + "epoch": 0.7294653014789534, + "grad_norm": 1.6359218526238526, + "learning_rate": 1.1855331131856146e-06, + "loss": 0.1913, + "step": 3206 + }, + { + "epoch": 0.7296928327645051, + "grad_norm": 1.6181503958338594, + "learning_rate": 1.185493584796149e-06, + "loss": 0.0989, + "step": 3207 + }, + { + "epoch": 0.7299203640500569, + "grad_norm": 1.9166380253206605, + "learning_rate": 1.185454044951221e-06, + "loss": 0.0894, + "step": 3208 + }, + { + "epoch": 0.7301478953356086, + "grad_norm": 1.3700990896238152, + "learning_rate": 1.1854144936516388e-06, + "loss": 0.0844, + "step": 3209 + }, + { + "epoch": 0.7303754266211604, + "grad_norm": 2.6450641956403476, + "learning_rate": 1.1853749308982107e-06, + "loss": 0.1549, + "step": 3210 + }, + { + "epoch": 0.7306029579067121, + "grad_norm": 2.1756090269876314, + "learning_rate": 1.1853353566917452e-06, + "loss": 0.1004, + "step": 3211 + }, + { + "epoch": 0.7308304891922639, + "grad_norm": 1.2519258630474208, + "learning_rate": 1.1852957710330511e-06, + "loss": 0.0763, + "step": 3212 + }, + { + "epoch": 0.7310580204778157, + "grad_norm": 18.651298184865464, + "learning_rate": 1.1852561739229377e-06, + "loss": 0.1599, + "step": 3213 + }, + { + "epoch": 0.7312855517633675, + "grad_norm": 3.12678058720072, + "learning_rate": 1.1852165653622141e-06, + "loss": 0.3179, + "step": 3214 + }, + { + "epoch": 0.7315130830489193, + "grad_norm": 1.2066479080845354, + "learning_rate": 1.18517694535169e-06, + "loss": 0.0688, + "step": 3215 + }, + { + "epoch": 0.731740614334471, + "grad_norm": 1.80310804012313, + "learning_rate": 1.1851373138921752e-06, + "loss": 0.1103, + "step": 3216 + }, + { + "epoch": 0.7319681456200228, + "grad_norm": 1.366328303092527, + "learning_rate": 1.1850976709844792e-06, + "loss": 0.1247, + "step": 3217 + }, + { + "epoch": 0.7321956769055745, + "grad_norm": 1.9493249225988984, + "learning_rate": 1.1850580166294127e-06, + "loss": 0.0962, + "step": 3218 + }, + { + "epoch": 0.7324232081911263, + "grad_norm": 1.1976897071979897, + "learning_rate": 1.1850183508277862e-06, + "loss": 0.1775, + "step": 3219 + }, + { + "epoch": 0.732650739476678, + "grad_norm": 3.0289421262567084, + "learning_rate": 1.18497867358041e-06, + "loss": 0.2875, + "step": 3220 + }, + { + "epoch": 0.7328782707622298, + "grad_norm": 1.6390476846704543, + "learning_rate": 1.1849389848880955e-06, + "loss": 0.099, + "step": 3221 + }, + { + "epoch": 0.7331058020477815, + "grad_norm": 1.8644621568477064, + "learning_rate": 1.1848992847516535e-06, + "loss": 0.0925, + "step": 3222 + }, + { + "epoch": 0.7333333333333333, + "grad_norm": 2.158548498055859, + "learning_rate": 1.1848595731718955e-06, + "loss": 0.0937, + "step": 3223 + }, + { + "epoch": 0.7335608646188851, + "grad_norm": 1.6735365679375431, + "learning_rate": 1.1848198501496331e-06, + "loss": 0.1359, + "step": 3224 + }, + { + "epoch": 0.7337883959044369, + "grad_norm": 1.3225066929770561, + "learning_rate": 1.1847801156856783e-06, + "loss": 0.1176, + "step": 3225 + }, + { + "epoch": 0.7340159271899886, + "grad_norm": 2.1054486369243106, + "learning_rate": 1.1847403697808433e-06, + "loss": 0.1962, + "step": 3226 + }, + { + "epoch": 0.7342434584755404, + "grad_norm": 1.7746017108416965, + "learning_rate": 1.18470061243594e-06, + "loss": 0.1681, + "step": 3227 + }, + { + "epoch": 0.7344709897610922, + "grad_norm": 2.9480584375351646, + "learning_rate": 1.1846608436517813e-06, + "loss": 0.1256, + "step": 3228 + }, + { + "epoch": 0.7346985210466439, + "grad_norm": 2.623755146702941, + "learning_rate": 1.1846210634291799e-06, + "loss": 0.1623, + "step": 3229 + }, + { + "epoch": 0.7349260523321957, + "grad_norm": 2.1291378433809114, + "learning_rate": 1.184581271768949e-06, + "loss": 0.1612, + "step": 3230 + }, + { + "epoch": 0.7351535836177474, + "grad_norm": 2.1608869767979457, + "learning_rate": 1.1845414686719014e-06, + "loss": 0.2046, + "step": 3231 + }, + { + "epoch": 0.7353811149032992, + "grad_norm": 1.3738772434928717, + "learning_rate": 1.1845016541388513e-06, + "loss": 0.1205, + "step": 3232 + }, + { + "epoch": 0.7356086461888509, + "grad_norm": 2.046143944672607, + "learning_rate": 1.184461828170612e-06, + "loss": 0.165, + "step": 3233 + }, + { + "epoch": 0.7358361774744028, + "grad_norm": 1.732382614564042, + "learning_rate": 1.1844219907679973e-06, + "loss": 0.1074, + "step": 3234 + }, + { + "epoch": 0.7360637087599545, + "grad_norm": 2.1175601609684698, + "learning_rate": 1.184382141931822e-06, + "loss": 0.1152, + "step": 3235 + }, + { + "epoch": 0.7362912400455063, + "grad_norm": 2.735577181836857, + "learning_rate": 1.1843422816628998e-06, + "loss": 0.1559, + "step": 3236 + }, + { + "epoch": 0.736518771331058, + "grad_norm": 1.551253205529087, + "learning_rate": 1.184302409962046e-06, + "loss": 0.0872, + "step": 3237 + }, + { + "epoch": 0.7367463026166098, + "grad_norm": 1.6777152959131918, + "learning_rate": 1.1842625268300754e-06, + "loss": 0.098, + "step": 3238 + }, + { + "epoch": 0.7369738339021615, + "grad_norm": 1.9068533548096838, + "learning_rate": 1.1842226322678028e-06, + "loss": 0.076, + "step": 3239 + }, + { + "epoch": 0.7372013651877133, + "grad_norm": 3.027756278397639, + "learning_rate": 1.1841827262760436e-06, + "loss": 0.1511, + "step": 3240 + }, + { + "epoch": 0.737428896473265, + "grad_norm": 3.2145156468158245, + "learning_rate": 1.1841428088556137e-06, + "loss": 0.1693, + "step": 3241 + }, + { + "epoch": 0.7376564277588168, + "grad_norm": 1.640064862711897, + "learning_rate": 1.184102880007329e-06, + "loss": 0.1045, + "step": 3242 + }, + { + "epoch": 0.7378839590443687, + "grad_norm": 2.2190011378264556, + "learning_rate": 1.1840629397320052e-06, + "loss": 0.1467, + "step": 3243 + }, + { + "epoch": 0.7381114903299204, + "grad_norm": 2.0452904656567608, + "learning_rate": 1.1840229880304589e-06, + "loss": 0.1435, + "step": 3244 + }, + { + "epoch": 0.7383390216154722, + "grad_norm": 2.8918271598629346, + "learning_rate": 1.1839830249035062e-06, + "loss": 0.1346, + "step": 3245 + }, + { + "epoch": 0.7385665529010239, + "grad_norm": 2.342345834795013, + "learning_rate": 1.1839430503519645e-06, + "loss": 0.1392, + "step": 3246 + }, + { + "epoch": 0.7387940841865757, + "grad_norm": 3.264140845800353, + "learning_rate": 1.1839030643766505e-06, + "loss": 0.2778, + "step": 3247 + }, + { + "epoch": 0.7390216154721274, + "grad_norm": 2.4607222935859494, + "learning_rate": 1.1838630669783814e-06, + "loss": 0.1624, + "step": 3248 + }, + { + "epoch": 0.7392491467576792, + "grad_norm": 1.9445345822890676, + "learning_rate": 1.1838230581579746e-06, + "loss": 0.0794, + "step": 3249 + }, + { + "epoch": 0.7394766780432309, + "grad_norm": 2.1685275554098613, + "learning_rate": 1.183783037916248e-06, + "loss": 0.1796, + "step": 3250 + }, + { + "epoch": 0.7397042093287827, + "grad_norm": 3.2158244936505236, + "learning_rate": 1.1837430062540196e-06, + "loss": 0.1964, + "step": 3251 + }, + { + "epoch": 0.7399317406143344, + "grad_norm": 1.7184491733113405, + "learning_rate": 1.1837029631721072e-06, + "loss": 0.1352, + "step": 3252 + }, + { + "epoch": 0.7401592718998863, + "grad_norm": 1.7778593056510787, + "learning_rate": 1.1836629086713296e-06, + "loss": 0.0873, + "step": 3253 + }, + { + "epoch": 0.740386803185438, + "grad_norm": 1.7203195667874172, + "learning_rate": 1.1836228427525054e-06, + "loss": 0.0676, + "step": 3254 + }, + { + "epoch": 0.7406143344709898, + "grad_norm": 2.922732576924441, + "learning_rate": 1.183582765416453e-06, + "loss": 0.172, + "step": 3255 + }, + { + "epoch": 0.7408418657565415, + "grad_norm": 2.039786020178238, + "learning_rate": 1.1835426766639923e-06, + "loss": 0.135, + "step": 3256 + }, + { + "epoch": 0.7410693970420933, + "grad_norm": 2.2399538962699017, + "learning_rate": 1.183502576495942e-06, + "loss": 0.116, + "step": 3257 + }, + { + "epoch": 0.741296928327645, + "grad_norm": 1.3632994430066812, + "learning_rate": 1.1834624649131218e-06, + "loss": 0.1134, + "step": 3258 + }, + { + "epoch": 0.7415244596131968, + "grad_norm": 1.2918804407041258, + "learning_rate": 1.1834223419163518e-06, + "loss": 0.112, + "step": 3259 + }, + { + "epoch": 0.7417519908987485, + "grad_norm": 2.6887841828247234, + "learning_rate": 1.1833822075064517e-06, + "loss": 0.1191, + "step": 3260 + }, + { + "epoch": 0.7419795221843003, + "grad_norm": 2.197636783508214, + "learning_rate": 1.183342061684242e-06, + "loss": 0.1279, + "step": 3261 + }, + { + "epoch": 0.742207053469852, + "grad_norm": 1.6494974684467505, + "learning_rate": 1.183301904450543e-06, + "loss": 0.1166, + "step": 3262 + }, + { + "epoch": 0.7424345847554039, + "grad_norm": 2.828124132396838, + "learning_rate": 1.1832617358061756e-06, + "loss": 0.1858, + "step": 3263 + }, + { + "epoch": 0.7426621160409557, + "grad_norm": 2.6045427793496727, + "learning_rate": 1.1832215557519608e-06, + "loss": 0.1324, + "step": 3264 + }, + { + "epoch": 0.7428896473265074, + "grad_norm": 2.127063173518277, + "learning_rate": 1.1831813642887196e-06, + "loss": 0.1959, + "step": 3265 + }, + { + "epoch": 0.7431171786120592, + "grad_norm": 2.7387306057680267, + "learning_rate": 1.1831411614172735e-06, + "loss": 0.1821, + "step": 3266 + }, + { + "epoch": 0.7433447098976109, + "grad_norm": 2.940295950006921, + "learning_rate": 1.1831009471384445e-06, + "loss": 0.1243, + "step": 3267 + }, + { + "epoch": 0.7435722411831627, + "grad_norm": 1.4605085364781514, + "learning_rate": 1.1830607214530543e-06, + "loss": 0.1861, + "step": 3268 + }, + { + "epoch": 0.7437997724687144, + "grad_norm": 3.3366956312883143, + "learning_rate": 1.1830204843619248e-06, + "loss": 0.1764, + "step": 3269 + }, + { + "epoch": 0.7440273037542662, + "grad_norm": 2.2222642240615094, + "learning_rate": 1.1829802358658785e-06, + "loss": 0.215, + "step": 3270 + }, + { + "epoch": 0.7442548350398179, + "grad_norm": 1.5340999546765406, + "learning_rate": 1.1829399759657383e-06, + "loss": 0.1379, + "step": 3271 + }, + { + "epoch": 0.7444823663253698, + "grad_norm": 1.553840846268558, + "learning_rate": 1.1828997046623267e-06, + "loss": 0.1392, + "step": 3272 + }, + { + "epoch": 0.7447098976109215, + "grad_norm": 2.597096922534138, + "learning_rate": 1.1828594219564669e-06, + "loss": 0.1466, + "step": 3273 + }, + { + "epoch": 0.7449374288964733, + "grad_norm": 1.3052925878692923, + "learning_rate": 1.182819127848982e-06, + "loss": 0.0779, + "step": 3274 + }, + { + "epoch": 0.745164960182025, + "grad_norm": 3.485548486880467, + "learning_rate": 1.1827788223406959e-06, + "loss": 0.2671, + "step": 3275 + }, + { + "epoch": 0.7453924914675768, + "grad_norm": 2.3473098076936907, + "learning_rate": 1.1827385054324323e-06, + "loss": 0.1415, + "step": 3276 + }, + { + "epoch": 0.7456200227531286, + "grad_norm": 4.3291248044860815, + "learning_rate": 1.1826981771250148e-06, + "loss": 0.2232, + "step": 3277 + }, + { + "epoch": 0.7458475540386803, + "grad_norm": 3.7337389140136463, + "learning_rate": 1.1826578374192681e-06, + "loss": 0.2382, + "step": 3278 + }, + { + "epoch": 0.7460750853242321, + "grad_norm": 3.3043996653981846, + "learning_rate": 1.1826174863160168e-06, + "loss": 0.1457, + "step": 3279 + }, + { + "epoch": 0.7463026166097838, + "grad_norm": 1.984573811971679, + "learning_rate": 1.182577123816085e-06, + "loss": 0.1282, + "step": 3280 + }, + { + "epoch": 0.7465301478953356, + "grad_norm": 2.5055449072250084, + "learning_rate": 1.1825367499202978e-06, + "loss": 0.152, + "step": 3281 + }, + { + "epoch": 0.7467576791808874, + "grad_norm": 2.566420978783162, + "learning_rate": 1.1824963646294806e-06, + "loss": 0.1344, + "step": 3282 + }, + { + "epoch": 0.7469852104664392, + "grad_norm": 3.030207356215811, + "learning_rate": 1.1824559679444588e-06, + "loss": 0.1618, + "step": 3283 + }, + { + "epoch": 0.7472127417519909, + "grad_norm": 1.1348070316131165, + "learning_rate": 1.182415559866058e-06, + "loss": 0.118, + "step": 3284 + }, + { + "epoch": 0.7474402730375427, + "grad_norm": 1.0639490098963196, + "learning_rate": 1.182375140395104e-06, + "loss": 0.0784, + "step": 3285 + }, + { + "epoch": 0.7476678043230944, + "grad_norm": 1.6785139016792132, + "learning_rate": 1.1823347095324228e-06, + "loss": 0.1616, + "step": 3286 + }, + { + "epoch": 0.7478953356086462, + "grad_norm": 1.4785752433396468, + "learning_rate": 1.1822942672788409e-06, + "loss": 0.065, + "step": 3287 + }, + { + "epoch": 0.7481228668941979, + "grad_norm": 2.3836521148460688, + "learning_rate": 1.1822538136351849e-06, + "loss": 0.1318, + "step": 3288 + }, + { + "epoch": 0.7483503981797497, + "grad_norm": 1.9833187363432578, + "learning_rate": 1.1822133486022815e-06, + "loss": 0.1365, + "step": 3289 + }, + { + "epoch": 0.7485779294653014, + "grad_norm": 2.2065405901117074, + "learning_rate": 1.1821728721809577e-06, + "loss": 0.1553, + "step": 3290 + }, + { + "epoch": 0.7488054607508532, + "grad_norm": 3.0349788681713017, + "learning_rate": 1.1821323843720408e-06, + "loss": 0.1817, + "step": 3291 + }, + { + "epoch": 0.749032992036405, + "grad_norm": 1.9694888646938773, + "learning_rate": 1.1820918851763582e-06, + "loss": 0.125, + "step": 3292 + }, + { + "epoch": 0.7492605233219568, + "grad_norm": 1.5124900798168812, + "learning_rate": 1.182051374594738e-06, + "loss": 0.1254, + "step": 3293 + }, + { + "epoch": 0.7494880546075086, + "grad_norm": 3.67837925062901, + "learning_rate": 1.1820108526280076e-06, + "loss": 0.1798, + "step": 3294 + }, + { + "epoch": 0.7497155858930603, + "grad_norm": 3.3239018085434786, + "learning_rate": 1.1819703192769955e-06, + "loss": 0.1614, + "step": 3295 + }, + { + "epoch": 0.7499431171786121, + "grad_norm": 2.8547697014823945, + "learning_rate": 1.1819297745425304e-06, + "loss": 0.1254, + "step": 3296 + }, + { + "epoch": 0.7501706484641638, + "grad_norm": 1.615743914055625, + "learning_rate": 1.1818892184254404e-06, + "loss": 0.1423, + "step": 3297 + }, + { + "epoch": 0.7503981797497156, + "grad_norm": 2.3926722908231683, + "learning_rate": 1.1818486509265547e-06, + "loss": 0.1005, + "step": 3298 + }, + { + "epoch": 0.7506257110352673, + "grad_norm": 1.659490567501026, + "learning_rate": 1.1818080720467026e-06, + "loss": 0.093, + "step": 3299 + }, + { + "epoch": 0.7508532423208191, + "grad_norm": 1.4636043042566025, + "learning_rate": 1.1817674817867131e-06, + "loss": 0.1194, + "step": 3300 + }, + { + "epoch": 0.7510807736063708, + "grad_norm": 1.859124523017859, + "learning_rate": 1.181726880147416e-06, + "loss": 0.1264, + "step": 3301 + }, + { + "epoch": 0.7513083048919227, + "grad_norm": 0.7197511547852126, + "learning_rate": 1.181686267129641e-06, + "loss": 0.0518, + "step": 3302 + }, + { + "epoch": 0.7515358361774744, + "grad_norm": 2.2033546472464542, + "learning_rate": 1.1816456427342181e-06, + "loss": 0.1269, + "step": 3303 + }, + { + "epoch": 0.7517633674630262, + "grad_norm": 1.6378674348907998, + "learning_rate": 1.181605006961978e-06, + "loss": 0.0921, + "step": 3304 + }, + { + "epoch": 0.7519908987485779, + "grad_norm": 1.7124569436254014, + "learning_rate": 1.1815643598137507e-06, + "loss": 0.1369, + "step": 3305 + }, + { + "epoch": 0.7522184300341297, + "grad_norm": 1.5770049353688607, + "learning_rate": 1.1815237012903675e-06, + "loss": 0.1099, + "step": 3306 + }, + { + "epoch": 0.7524459613196814, + "grad_norm": 2.571994096255603, + "learning_rate": 1.1814830313926589e-06, + "loss": 0.175, + "step": 3307 + }, + { + "epoch": 0.7526734926052332, + "grad_norm": 1.823553996843364, + "learning_rate": 1.1814423501214562e-06, + "loss": 0.1043, + "step": 3308 + }, + { + "epoch": 0.752901023890785, + "grad_norm": 2.167497133328913, + "learning_rate": 1.1814016574775909e-06, + "loss": 0.1003, + "step": 3309 + }, + { + "epoch": 0.7531285551763367, + "grad_norm": 1.3244656811190443, + "learning_rate": 1.1813609534618948e-06, + "loss": 0.089, + "step": 3310 + }, + { + "epoch": 0.7533560864618886, + "grad_norm": 1.9888469659716759, + "learning_rate": 1.1813202380751998e-06, + "loss": 0.0601, + "step": 3311 + }, + { + "epoch": 0.7535836177474403, + "grad_norm": 1.765491538331343, + "learning_rate": 1.1812795113183378e-06, + "loss": 0.1248, + "step": 3312 + }, + { + "epoch": 0.7538111490329921, + "grad_norm": 2.0478968271407383, + "learning_rate": 1.1812387731921415e-06, + "loss": 0.1382, + "step": 3313 + }, + { + "epoch": 0.7540386803185438, + "grad_norm": 3.2784426559764985, + "learning_rate": 1.1811980236974435e-06, + "loss": 0.146, + "step": 3314 + }, + { + "epoch": 0.7542662116040956, + "grad_norm": 1.4734056038714625, + "learning_rate": 1.1811572628350764e-06, + "loss": 0.127, + "step": 3315 + }, + { + "epoch": 0.7544937428896473, + "grad_norm": 2.0370268043177115, + "learning_rate": 1.1811164906058735e-06, + "loss": 0.1445, + "step": 3316 + }, + { + "epoch": 0.7547212741751991, + "grad_norm": 2.96415685198712, + "learning_rate": 1.1810757070106678e-06, + "loss": 0.2705, + "step": 3317 + }, + { + "epoch": 0.7549488054607508, + "grad_norm": 2.2971763536780943, + "learning_rate": 1.1810349120502932e-06, + "loss": 0.1491, + "step": 3318 + }, + { + "epoch": 0.7551763367463026, + "grad_norm": 2.681036370900885, + "learning_rate": 1.1809941057255834e-06, + "loss": 0.1695, + "step": 3319 + }, + { + "epoch": 0.7554038680318543, + "grad_norm": 2.724601949111482, + "learning_rate": 1.1809532880373721e-06, + "loss": 0.2026, + "step": 3320 + }, + { + "epoch": 0.7556313993174062, + "grad_norm": 1.8617956455441396, + "learning_rate": 1.180912458986494e-06, + "loss": 0.0897, + "step": 3321 + }, + { + "epoch": 0.755858930602958, + "grad_norm": 2.53226962131568, + "learning_rate": 1.180871618573783e-06, + "loss": 0.1572, + "step": 3322 + }, + { + "epoch": 0.7560864618885097, + "grad_norm": 2.21274255033511, + "learning_rate": 1.1808307668000745e-06, + "loss": 0.1583, + "step": 3323 + }, + { + "epoch": 0.7563139931740614, + "grad_norm": 2.125706306020661, + "learning_rate": 1.180789903666203e-06, + "loss": 0.1883, + "step": 3324 + }, + { + "epoch": 0.7565415244596132, + "grad_norm": 1.6246651739254347, + "learning_rate": 1.1807490291730036e-06, + "loss": 0.1668, + "step": 3325 + }, + { + "epoch": 0.756769055745165, + "grad_norm": 1.8005307284189929, + "learning_rate": 1.1807081433213122e-06, + "loss": 0.1145, + "step": 3326 + }, + { + "epoch": 0.7569965870307167, + "grad_norm": 1.405317196271188, + "learning_rate": 1.1806672461119637e-06, + "loss": 0.1173, + "step": 3327 + }, + { + "epoch": 0.7572241183162685, + "grad_norm": 1.8080670242907921, + "learning_rate": 1.1806263375457947e-06, + "loss": 0.0874, + "step": 3328 + }, + { + "epoch": 0.7574516496018202, + "grad_norm": 2.4167979568785842, + "learning_rate": 1.1805854176236406e-06, + "loss": 0.1695, + "step": 3329 + }, + { + "epoch": 0.757679180887372, + "grad_norm": 2.3063952560271015, + "learning_rate": 1.1805444863463384e-06, + "loss": 0.1849, + "step": 3330 + }, + { + "epoch": 0.7579067121729238, + "grad_norm": 2.079072885402329, + "learning_rate": 1.180503543714724e-06, + "loss": 0.0683, + "step": 3331 + }, + { + "epoch": 0.7581342434584756, + "grad_norm": 1.5746175237173137, + "learning_rate": 1.1804625897296345e-06, + "loss": 0.0965, + "step": 3332 + }, + { + "epoch": 0.7583617747440273, + "grad_norm": 3.530110679680489, + "learning_rate": 1.1804216243919074e-06, + "loss": 0.1614, + "step": 3333 + }, + { + "epoch": 0.7585893060295791, + "grad_norm": 2.0332642196040753, + "learning_rate": 1.1803806477023792e-06, + "loss": 0.0879, + "step": 3334 + }, + { + "epoch": 0.7588168373151308, + "grad_norm": 1.3656933323492333, + "learning_rate": 1.1803396596618878e-06, + "loss": 0.0995, + "step": 3335 + }, + { + "epoch": 0.7590443686006826, + "grad_norm": 2.0902834793608087, + "learning_rate": 1.1802986602712705e-06, + "loss": 0.1337, + "step": 3336 + }, + { + "epoch": 0.7592718998862343, + "grad_norm": 2.5965116123759007, + "learning_rate": 1.1802576495313657e-06, + "loss": 0.0907, + "step": 3337 + }, + { + "epoch": 0.7594994311717861, + "grad_norm": 4.142087858458215, + "learning_rate": 1.1802166274430116e-06, + "loss": 0.1953, + "step": 3338 + }, + { + "epoch": 0.7597269624573378, + "grad_norm": 1.7222237359967432, + "learning_rate": 1.1801755940070464e-06, + "loss": 0.0783, + "step": 3339 + }, + { + "epoch": 0.7599544937428896, + "grad_norm": 3.05894156690617, + "learning_rate": 1.1801345492243087e-06, + "loss": 0.1581, + "step": 3340 + }, + { + "epoch": 0.7601820250284415, + "grad_norm": 2.5343054135059213, + "learning_rate": 1.1800934930956378e-06, + "loss": 0.1767, + "step": 3341 + }, + { + "epoch": 0.7604095563139932, + "grad_norm": 1.3193612113140212, + "learning_rate": 1.1800524256218724e-06, + "loss": 0.1118, + "step": 3342 + }, + { + "epoch": 0.760637087599545, + "grad_norm": 2.2549990933213406, + "learning_rate": 1.1800113468038518e-06, + "loss": 0.1354, + "step": 3343 + }, + { + "epoch": 0.7608646188850967, + "grad_norm": 2.3490442108286347, + "learning_rate": 1.1799702566424159e-06, + "loss": 0.2573, + "step": 3344 + }, + { + "epoch": 0.7610921501706485, + "grad_norm": 1.535894674316497, + "learning_rate": 1.1799291551384042e-06, + "loss": 0.1606, + "step": 3345 + }, + { + "epoch": 0.7613196814562002, + "grad_norm": 2.14792184820641, + "learning_rate": 1.1798880422926568e-06, + "loss": 0.1214, + "step": 3346 + }, + { + "epoch": 0.761547212741752, + "grad_norm": 1.5845109192049205, + "learning_rate": 1.1798469181060143e-06, + "loss": 0.1501, + "step": 3347 + }, + { + "epoch": 0.7617747440273037, + "grad_norm": 2.711691703624, + "learning_rate": 1.1798057825793167e-06, + "loss": 0.1688, + "step": 3348 + }, + { + "epoch": 0.7620022753128555, + "grad_norm": 2.4863555470792056, + "learning_rate": 1.179764635713405e-06, + "loss": 0.156, + "step": 3349 + }, + { + "epoch": 0.7622298065984073, + "grad_norm": 1.7061185359263562, + "learning_rate": 1.1797234775091204e-06, + "loss": 0.1129, + "step": 3350 + }, + { + "epoch": 0.7624573378839591, + "grad_norm": 2.706753314731583, + "learning_rate": 1.1796823079673036e-06, + "loss": 0.1696, + "step": 3351 + }, + { + "epoch": 0.7626848691695108, + "grad_norm": 1.9907259016451502, + "learning_rate": 1.1796411270887965e-06, + "loss": 0.1619, + "step": 3352 + }, + { + "epoch": 0.7629124004550626, + "grad_norm": 2.4653009277692552, + "learning_rate": 1.1795999348744403e-06, + "loss": 0.1482, + "step": 3353 + }, + { + "epoch": 0.7631399317406143, + "grad_norm": 2.15969706994595, + "learning_rate": 1.1795587313250773e-06, + "loss": 0.0883, + "step": 3354 + }, + { + "epoch": 0.7633674630261661, + "grad_norm": 1.7222779007665892, + "learning_rate": 1.1795175164415493e-06, + "loss": 0.1562, + "step": 3355 + }, + { + "epoch": 0.7635949943117178, + "grad_norm": 1.9770931086408174, + "learning_rate": 1.1794762902246988e-06, + "loss": 0.1936, + "step": 3356 + }, + { + "epoch": 0.7638225255972696, + "grad_norm": 2.787868720119073, + "learning_rate": 1.1794350526753688e-06, + "loss": 0.2025, + "step": 3357 + }, + { + "epoch": 0.7640500568828213, + "grad_norm": 2.238724793825616, + "learning_rate": 1.1793938037944014e-06, + "loss": 0.1931, + "step": 3358 + }, + { + "epoch": 0.7642775881683731, + "grad_norm": 2.0044557011415143, + "learning_rate": 1.1793525435826399e-06, + "loss": 0.1247, + "step": 3359 + }, + { + "epoch": 0.764505119453925, + "grad_norm": 2.692377197275603, + "learning_rate": 1.1793112720409277e-06, + "loss": 0.2121, + "step": 3360 + }, + { + "epoch": 0.7647326507394767, + "grad_norm": 3.1721291331762855, + "learning_rate": 1.1792699891701085e-06, + "loss": 0.1208, + "step": 3361 + }, + { + "epoch": 0.7649601820250285, + "grad_norm": 1.814950863168666, + "learning_rate": 1.1792286949710254e-06, + "loss": 0.1618, + "step": 3362 + }, + { + "epoch": 0.7651877133105802, + "grad_norm": 2.468945582085837, + "learning_rate": 1.1791873894445233e-06, + "loss": 0.127, + "step": 3363 + }, + { + "epoch": 0.765415244596132, + "grad_norm": 3.033669581494895, + "learning_rate": 1.1791460725914455e-06, + "loss": 0.1724, + "step": 3364 + }, + { + "epoch": 0.7656427758816837, + "grad_norm": 1.3064478725478224, + "learning_rate": 1.179104744412637e-06, + "loss": 0.0949, + "step": 3365 + }, + { + "epoch": 0.7658703071672355, + "grad_norm": 2.499744061538605, + "learning_rate": 1.1790634049089425e-06, + "loss": 0.168, + "step": 3366 + }, + { + "epoch": 0.7660978384527872, + "grad_norm": 2.2054261208451416, + "learning_rate": 1.1790220540812063e-06, + "loss": 0.1099, + "step": 3367 + }, + { + "epoch": 0.766325369738339, + "grad_norm": 1.9138510381699176, + "learning_rate": 1.1789806919302743e-06, + "loss": 0.1629, + "step": 3368 + }, + { + "epoch": 0.7665529010238907, + "grad_norm": 1.5530188387515713, + "learning_rate": 1.1789393184569914e-06, + "loss": 0.0733, + "step": 3369 + }, + { + "epoch": 0.7667804323094426, + "grad_norm": 2.4731819444524525, + "learning_rate": 1.1788979336622034e-06, + "loss": 0.0997, + "step": 3370 + }, + { + "epoch": 0.7670079635949943, + "grad_norm": 2.039023451665474, + "learning_rate": 1.178856537546756e-06, + "loss": 0.1742, + "step": 3371 + }, + { + "epoch": 0.7672354948805461, + "grad_norm": 2.295655111835599, + "learning_rate": 1.1788151301114952e-06, + "loss": 0.1817, + "step": 3372 + }, + { + "epoch": 0.7674630261660979, + "grad_norm": 3.0168578047526435, + "learning_rate": 1.1787737113572678e-06, + "loss": 0.1219, + "step": 3373 + }, + { + "epoch": 0.7676905574516496, + "grad_norm": 1.9623836795766205, + "learning_rate": 1.1787322812849196e-06, + "loss": 0.1471, + "step": 3374 + }, + { + "epoch": 0.7679180887372014, + "grad_norm": 3.1623467825524125, + "learning_rate": 1.1786908398952977e-06, + "loss": 0.2429, + "step": 3375 + }, + { + "epoch": 0.7681456200227531, + "grad_norm": 2.41967387734086, + "learning_rate": 1.1786493871892491e-06, + "loss": 0.0984, + "step": 3376 + }, + { + "epoch": 0.7683731513083049, + "grad_norm": 1.47807577415724, + "learning_rate": 1.178607923167621e-06, + "loss": 0.1533, + "step": 3377 + }, + { + "epoch": 0.7686006825938566, + "grad_norm": 2.0022276093033873, + "learning_rate": 1.1785664478312607e-06, + "loss": 0.1213, + "step": 3378 + }, + { + "epoch": 0.7688282138794085, + "grad_norm": 1.3538482719717833, + "learning_rate": 1.1785249611810163e-06, + "loss": 0.1202, + "step": 3379 + }, + { + "epoch": 0.7690557451649602, + "grad_norm": 2.09633182774713, + "learning_rate": 1.1784834632177352e-06, + "loss": 0.1063, + "step": 3380 + }, + { + "epoch": 0.769283276450512, + "grad_norm": 2.0767741041017485, + "learning_rate": 1.178441953942266e-06, + "loss": 0.1268, + "step": 3381 + }, + { + "epoch": 0.7695108077360637, + "grad_norm": 2.1500043301721465, + "learning_rate": 1.1784004333554565e-06, + "loss": 0.0931, + "step": 3382 + }, + { + "epoch": 0.7697383390216155, + "grad_norm": 2.3874439937858396, + "learning_rate": 1.178358901458156e-06, + "loss": 0.2008, + "step": 3383 + }, + { + "epoch": 0.7699658703071672, + "grad_norm": 1.8338484321709367, + "learning_rate": 1.1783173582512127e-06, + "loss": 0.128, + "step": 3384 + }, + { + "epoch": 0.770193401592719, + "grad_norm": 1.5910093680360542, + "learning_rate": 1.178275803735476e-06, + "loss": 0.1111, + "step": 3385 + }, + { + "epoch": 0.7704209328782707, + "grad_norm": 2.064832606530375, + "learning_rate": 1.1782342379117954e-06, + "loss": 0.1496, + "step": 3386 + }, + { + "epoch": 0.7706484641638225, + "grad_norm": 2.599028903859917, + "learning_rate": 1.17819266078102e-06, + "loss": 0.1673, + "step": 3387 + }, + { + "epoch": 0.7708759954493742, + "grad_norm": 1.6264002152784236, + "learning_rate": 1.1781510723439995e-06, + "loss": 0.1288, + "step": 3388 + }, + { + "epoch": 0.7711035267349261, + "grad_norm": 2.0707608895341196, + "learning_rate": 1.1781094726015842e-06, + "loss": 0.1472, + "step": 3389 + }, + { + "epoch": 0.7713310580204779, + "grad_norm": 2.4106800417919096, + "learning_rate": 1.1780678615546245e-06, + "loss": 0.1807, + "step": 3390 + }, + { + "epoch": 0.7715585893060296, + "grad_norm": 2.819472416476872, + "learning_rate": 1.1780262392039706e-06, + "loss": 0.1614, + "step": 3391 + }, + { + "epoch": 0.7717861205915814, + "grad_norm": 2.2332176948627787, + "learning_rate": 1.177984605550473e-06, + "loss": 0.1236, + "step": 3392 + }, + { + "epoch": 0.7720136518771331, + "grad_norm": 1.711407204599878, + "learning_rate": 1.177942960594983e-06, + "loss": 0.1284, + "step": 3393 + }, + { + "epoch": 0.7722411831626849, + "grad_norm": 1.6565829770786178, + "learning_rate": 1.1779013043383516e-06, + "loss": 0.0791, + "step": 3394 + }, + { + "epoch": 0.7724687144482366, + "grad_norm": 1.6450273322349465, + "learning_rate": 1.17785963678143e-06, + "loss": 0.1579, + "step": 3395 + }, + { + "epoch": 0.7726962457337884, + "grad_norm": 2.731296051334521, + "learning_rate": 1.1778179579250699e-06, + "loss": 0.1343, + "step": 3396 + }, + { + "epoch": 0.7729237770193401, + "grad_norm": 2.374318965488984, + "learning_rate": 1.1777762677701232e-06, + "loss": 0.1052, + "step": 3397 + }, + { + "epoch": 0.7731513083048919, + "grad_norm": 1.450879572931303, + "learning_rate": 1.1777345663174419e-06, + "loss": 0.1993, + "step": 3398 + }, + { + "epoch": 0.7733788395904437, + "grad_norm": 2.0577588079276925, + "learning_rate": 1.1776928535678784e-06, + "loss": 0.1257, + "step": 3399 + }, + { + "epoch": 0.7736063708759955, + "grad_norm": 1.7629557940982252, + "learning_rate": 1.1776511295222852e-06, + "loss": 0.1317, + "step": 3400 + }, + { + "epoch": 0.7738339021615472, + "grad_norm": 2.0710954801105452, + "learning_rate": 1.177609394181515e-06, + "loss": 0.0998, + "step": 3401 + }, + { + "epoch": 0.774061433447099, + "grad_norm": 2.777274274681113, + "learning_rate": 1.177567647546421e-06, + "loss": 0.2848, + "step": 3402 + }, + { + "epoch": 0.7742889647326507, + "grad_norm": 2.0921178488829977, + "learning_rate": 1.177525889617856e-06, + "loss": 0.1265, + "step": 3403 + }, + { + "epoch": 0.7745164960182025, + "grad_norm": 1.8373289479738266, + "learning_rate": 1.177484120396674e-06, + "loss": 0.1216, + "step": 3404 + }, + { + "epoch": 0.7747440273037542, + "grad_norm": 1.6538813916238742, + "learning_rate": 1.1774423398837282e-06, + "loss": 0.0639, + "step": 3405 + }, + { + "epoch": 0.774971558589306, + "grad_norm": 1.2496723907667286, + "learning_rate": 1.177400548079873e-06, + "loss": 0.1251, + "step": 3406 + }, + { + "epoch": 0.7751990898748577, + "grad_norm": 1.7383502930197228, + "learning_rate": 1.177358744985962e-06, + "loss": 0.1146, + "step": 3407 + }, + { + "epoch": 0.7754266211604095, + "grad_norm": 2.110624002292149, + "learning_rate": 1.1773169306028498e-06, + "loss": 0.087, + "step": 3408 + }, + { + "epoch": 0.7756541524459614, + "grad_norm": 2.0104258894056675, + "learning_rate": 1.1772751049313911e-06, + "loss": 0.1155, + "step": 3409 + }, + { + "epoch": 0.7758816837315131, + "grad_norm": 1.6597706808667767, + "learning_rate": 1.1772332679724408e-06, + "loss": 0.1155, + "step": 3410 + }, + { + "epoch": 0.7761092150170649, + "grad_norm": 1.5975357267288846, + "learning_rate": 1.1771914197268538e-06, + "loss": 0.0954, + "step": 3411 + }, + { + "epoch": 0.7763367463026166, + "grad_norm": 1.5910280646678994, + "learning_rate": 1.1771495601954856e-06, + "loss": 0.1215, + "step": 3412 + }, + { + "epoch": 0.7765642775881684, + "grad_norm": 2.619906793593659, + "learning_rate": 1.1771076893791914e-06, + "loss": 0.1167, + "step": 3413 + }, + { + "epoch": 0.7767918088737201, + "grad_norm": 3.1996885670881867, + "learning_rate": 1.1770658072788272e-06, + "loss": 0.1344, + "step": 3414 + }, + { + "epoch": 0.7770193401592719, + "grad_norm": 2.8478691570571137, + "learning_rate": 1.1770239138952492e-06, + "loss": 0.1774, + "step": 3415 + }, + { + "epoch": 0.7772468714448236, + "grad_norm": 3.2605307459396853, + "learning_rate": 1.1769820092293132e-06, + "loss": 0.2085, + "step": 3416 + }, + { + "epoch": 0.7774744027303754, + "grad_norm": 1.634390217637281, + "learning_rate": 1.1769400932818758e-06, + "loss": 0.1419, + "step": 3417 + }, + { + "epoch": 0.7777019340159272, + "grad_norm": 1.1507493581000052, + "learning_rate": 1.1768981660537938e-06, + "loss": 0.1707, + "step": 3418 + }, + { + "epoch": 0.777929465301479, + "grad_norm": 1.09109698933865, + "learning_rate": 1.1768562275459242e-06, + "loss": 0.0952, + "step": 3419 + }, + { + "epoch": 0.7781569965870307, + "grad_norm": 1.9193041776951303, + "learning_rate": 1.1768142777591235e-06, + "loss": 0.163, + "step": 3420 + }, + { + "epoch": 0.7783845278725825, + "grad_norm": 2.660447855208393, + "learning_rate": 1.17677231669425e-06, + "loss": 0.1946, + "step": 3421 + }, + { + "epoch": 0.7786120591581343, + "grad_norm": 2.1193355335950743, + "learning_rate": 1.1767303443521608e-06, + "loss": 0.1059, + "step": 3422 + }, + { + "epoch": 0.778839590443686, + "grad_norm": 1.5289759590491452, + "learning_rate": 1.1766883607337137e-06, + "loss": 0.1464, + "step": 3423 + }, + { + "epoch": 0.7790671217292378, + "grad_norm": 1.3577624261424397, + "learning_rate": 1.176646365839767e-06, + "loss": 0.0896, + "step": 3424 + }, + { + "epoch": 0.7792946530147895, + "grad_norm": 1.3596876247439247, + "learning_rate": 1.1766043596711787e-06, + "loss": 0.11, + "step": 3425 + }, + { + "epoch": 0.7795221843003413, + "grad_norm": 1.678223461795012, + "learning_rate": 1.1765623422288078e-06, + "loss": 0.1154, + "step": 3426 + }, + { + "epoch": 0.779749715585893, + "grad_norm": 1.728616847327067, + "learning_rate": 1.1765203135135126e-06, + "loss": 0.1138, + "step": 3427 + }, + { + "epoch": 0.7799772468714449, + "grad_norm": 1.4891809627493324, + "learning_rate": 1.176478273526152e-06, + "loss": 0.0999, + "step": 3428 + }, + { + "epoch": 0.7802047781569966, + "grad_norm": 1.778994205604812, + "learning_rate": 1.1764362222675857e-06, + "loss": 0.1816, + "step": 3429 + }, + { + "epoch": 0.7804323094425484, + "grad_norm": 2.0271370931761963, + "learning_rate": 1.176394159738673e-06, + "loss": 0.1178, + "step": 3430 + }, + { + "epoch": 0.7806598407281001, + "grad_norm": 2.237166839400963, + "learning_rate": 1.1763520859402735e-06, + "loss": 0.1421, + "step": 3431 + }, + { + "epoch": 0.7808873720136519, + "grad_norm": 1.50673747208918, + "learning_rate": 1.176310000873247e-06, + "loss": 0.0839, + "step": 3432 + }, + { + "epoch": 0.7811149032992036, + "grad_norm": 1.8131829377270645, + "learning_rate": 1.1762679045384537e-06, + "loss": 0.1395, + "step": 3433 + }, + { + "epoch": 0.7813424345847554, + "grad_norm": 2.4892969634569333, + "learning_rate": 1.1762257969367543e-06, + "loss": 0.1181, + "step": 3434 + }, + { + "epoch": 0.7815699658703071, + "grad_norm": 2.411363769583913, + "learning_rate": 1.176183678069009e-06, + "loss": 0.1729, + "step": 3435 + }, + { + "epoch": 0.7817974971558589, + "grad_norm": 1.8263989621523369, + "learning_rate": 1.1761415479360784e-06, + "loss": 0.1787, + "step": 3436 + }, + { + "epoch": 0.7820250284414106, + "grad_norm": 2.173624893203865, + "learning_rate": 1.1760994065388246e-06, + "loss": 0.1014, + "step": 3437 + }, + { + "epoch": 0.7822525597269625, + "grad_norm": 2.5061481612728587, + "learning_rate": 1.1760572538781077e-06, + "loss": 0.1105, + "step": 3438 + }, + { + "epoch": 0.7824800910125143, + "grad_norm": 1.5904833725277863, + "learning_rate": 1.17601508995479e-06, + "loss": 0.075, + "step": 3439 + }, + { + "epoch": 0.782707622298066, + "grad_norm": 2.3323087090668846, + "learning_rate": 1.175972914769733e-06, + "loss": 0.1044, + "step": 3440 + }, + { + "epoch": 0.7829351535836178, + "grad_norm": 2.103149597484196, + "learning_rate": 1.1759307283237986e-06, + "loss": 0.1347, + "step": 3441 + }, + { + "epoch": 0.7831626848691695, + "grad_norm": 1.6169147832378625, + "learning_rate": 1.175888530617849e-06, + "loss": 0.0846, + "step": 3442 + }, + { + "epoch": 0.7833902161547213, + "grad_norm": 2.899804826982188, + "learning_rate": 1.175846321652747e-06, + "loss": 0.1551, + "step": 3443 + }, + { + "epoch": 0.783617747440273, + "grad_norm": 1.3458740703284953, + "learning_rate": 1.1758041014293548e-06, + "loss": 0.0923, + "step": 3444 + }, + { + "epoch": 0.7838452787258248, + "grad_norm": 1.493122071904463, + "learning_rate": 1.1757618699485353e-06, + "loss": 0.141, + "step": 3445 + }, + { + "epoch": 0.7840728100113765, + "grad_norm": 2.1402488719534074, + "learning_rate": 1.1757196272111524e-06, + "loss": 0.1485, + "step": 3446 + }, + { + "epoch": 0.7843003412969284, + "grad_norm": 2.152687176574438, + "learning_rate": 1.1756773732180684e-06, + "loss": 0.0962, + "step": 3447 + }, + { + "epoch": 0.7845278725824801, + "grad_norm": 1.974274187209729, + "learning_rate": 1.1756351079701477e-06, + "loss": 0.107, + "step": 3448 + }, + { + "epoch": 0.7847554038680319, + "grad_norm": 2.0072732316781283, + "learning_rate": 1.1755928314682537e-06, + "loss": 0.0983, + "step": 3449 + }, + { + "epoch": 0.7849829351535836, + "grad_norm": 1.880475722204832, + "learning_rate": 1.1755505437132507e-06, + "loss": 0.1217, + "step": 3450 + }, + { + "epoch": 0.7852104664391354, + "grad_norm": 1.863619823885818, + "learning_rate": 1.1755082447060029e-06, + "loss": 0.1714, + "step": 3451 + }, + { + "epoch": 0.7854379977246871, + "grad_norm": 2.2078473549037123, + "learning_rate": 1.1754659344473747e-06, + "loss": 0.1717, + "step": 3452 + }, + { + "epoch": 0.7856655290102389, + "grad_norm": 0.9604209487774992, + "learning_rate": 1.175423612938231e-06, + "loss": 0.1206, + "step": 3453 + }, + { + "epoch": 0.7858930602957906, + "grad_norm": 2.8777141127877566, + "learning_rate": 1.1753812801794368e-06, + "loss": 0.1956, + "step": 3454 + }, + { + "epoch": 0.7861205915813424, + "grad_norm": 2.2683260816233624, + "learning_rate": 1.175338936171857e-06, + "loss": 0.0813, + "step": 3455 + }, + { + "epoch": 0.7863481228668942, + "grad_norm": 2.2494833127836555, + "learning_rate": 1.1752965809163574e-06, + "loss": 0.1326, + "step": 3456 + }, + { + "epoch": 0.786575654152446, + "grad_norm": 1.0250772681759766, + "learning_rate": 1.1752542144138033e-06, + "loss": 0.1171, + "step": 3457 + }, + { + "epoch": 0.7868031854379978, + "grad_norm": 2.0754576696406155, + "learning_rate": 1.1752118366650608e-06, + "loss": 0.1096, + "step": 3458 + }, + { + "epoch": 0.7870307167235495, + "grad_norm": 1.4736818380845116, + "learning_rate": 1.1751694476709962e-06, + "loss": 0.0629, + "step": 3459 + }, + { + "epoch": 0.7872582480091013, + "grad_norm": 1.7459278393261928, + "learning_rate": 1.1751270474324757e-06, + "loss": 0.1304, + "step": 3460 + }, + { + "epoch": 0.787485779294653, + "grad_norm": 1.8293046683083407, + "learning_rate": 1.1750846359503657e-06, + "loss": 0.1377, + "step": 3461 + }, + { + "epoch": 0.7877133105802048, + "grad_norm": 1.4740555170674776, + "learning_rate": 1.1750422132255334e-06, + "loss": 0.101, + "step": 3462 + }, + { + "epoch": 0.7879408418657565, + "grad_norm": 2.506831345458617, + "learning_rate": 1.1749997792588453e-06, + "loss": 0.15, + "step": 3463 + }, + { + "epoch": 0.7881683731513083, + "grad_norm": 1.3945678529440881, + "learning_rate": 1.1749573340511693e-06, + "loss": 0.1358, + "step": 3464 + }, + { + "epoch": 0.78839590443686, + "grad_norm": 2.1814793309718485, + "learning_rate": 1.1749148776033723e-06, + "loss": 0.1633, + "step": 3465 + }, + { + "epoch": 0.7886234357224118, + "grad_norm": 2.641721026309912, + "learning_rate": 1.1748724099163225e-06, + "loss": 0.1939, + "step": 3466 + }, + { + "epoch": 0.7888509670079636, + "grad_norm": 1.2748767276799335, + "learning_rate": 1.1748299309908878e-06, + "loss": 0.0653, + "step": 3467 + }, + { + "epoch": 0.7890784982935154, + "grad_norm": 2.211160099467529, + "learning_rate": 1.174787440827936e-06, + "loss": 0.1336, + "step": 3468 + }, + { + "epoch": 0.7893060295790671, + "grad_norm": 2.3953771038471743, + "learning_rate": 1.1747449394283361e-06, + "loss": 0.13, + "step": 3469 + }, + { + "epoch": 0.7895335608646189, + "grad_norm": 2.515025121426822, + "learning_rate": 1.174702426792956e-06, + "loss": 0.188, + "step": 3470 + }, + { + "epoch": 0.7897610921501707, + "grad_norm": 1.8273081736073256, + "learning_rate": 1.1746599029226656e-06, + "loss": 0.1129, + "step": 3471 + }, + { + "epoch": 0.7899886234357224, + "grad_norm": 1.749015243103347, + "learning_rate": 1.174617367818333e-06, + "loss": 0.1096, + "step": 3472 + }, + { + "epoch": 0.7902161547212742, + "grad_norm": 1.202000950355586, + "learning_rate": 1.174574821480828e-06, + "loss": 0.153, + "step": 3473 + }, + { + "epoch": 0.7904436860068259, + "grad_norm": 1.7678029264916482, + "learning_rate": 1.1745322639110203e-06, + "loss": 0.1284, + "step": 3474 + }, + { + "epoch": 0.7906712172923777, + "grad_norm": 1.0898438896362213, + "learning_rate": 1.1744896951097798e-06, + "loss": 0.1178, + "step": 3475 + }, + { + "epoch": 0.7908987485779294, + "grad_norm": 1.6287384221021175, + "learning_rate": 1.1744471150779758e-06, + "loss": 0.0908, + "step": 3476 + }, + { + "epoch": 0.7911262798634813, + "grad_norm": 2.020655633683389, + "learning_rate": 1.1744045238164793e-06, + "loss": 0.1178, + "step": 3477 + }, + { + "epoch": 0.791353811149033, + "grad_norm": 3.13062068048053, + "learning_rate": 1.1743619213261604e-06, + "loss": 0.1702, + "step": 3478 + }, + { + "epoch": 0.7915813424345848, + "grad_norm": 1.9211729488264053, + "learning_rate": 1.1743193076078901e-06, + "loss": 0.1438, + "step": 3479 + }, + { + "epoch": 0.7918088737201365, + "grad_norm": 2.4542917746271145, + "learning_rate": 1.174276682662539e-06, + "loss": 0.1783, + "step": 3480 + }, + { + "epoch": 0.7920364050056883, + "grad_norm": 2.000737339074988, + "learning_rate": 1.1742340464909786e-06, + "loss": 0.1256, + "step": 3481 + }, + { + "epoch": 0.79226393629124, + "grad_norm": 2.354494981818762, + "learning_rate": 1.1741913990940801e-06, + "loss": 0.1332, + "step": 3482 + }, + { + "epoch": 0.7924914675767918, + "grad_norm": 1.6008608978921464, + "learning_rate": 1.174148740472715e-06, + "loss": 0.1235, + "step": 3483 + }, + { + "epoch": 0.7927189988623435, + "grad_norm": 3.3922095426630468, + "learning_rate": 1.1741060706277557e-06, + "loss": 0.1673, + "step": 3484 + }, + { + "epoch": 0.7929465301478953, + "grad_norm": 1.493369989814107, + "learning_rate": 1.1740633895600738e-06, + "loss": 0.1465, + "step": 3485 + }, + { + "epoch": 0.7931740614334472, + "grad_norm": 1.8432157550864956, + "learning_rate": 1.1740206972705418e-06, + "loss": 0.1209, + "step": 3486 + }, + { + "epoch": 0.7934015927189989, + "grad_norm": 1.7889600196785538, + "learning_rate": 1.1739779937600323e-06, + "loss": 0.0642, + "step": 3487 + }, + { + "epoch": 0.7936291240045507, + "grad_norm": 1.4593061124512618, + "learning_rate": 1.1739352790294177e-06, + "loss": 0.0828, + "step": 3488 + }, + { + "epoch": 0.7938566552901024, + "grad_norm": 2.134125735004945, + "learning_rate": 1.1738925530795716e-06, + "loss": 0.1497, + "step": 3489 + }, + { + "epoch": 0.7940841865756542, + "grad_norm": 3.0082402968183244, + "learning_rate": 1.173849815911367e-06, + "loss": 0.1692, + "step": 3490 + }, + { + "epoch": 0.7943117178612059, + "grad_norm": 2.384918860515077, + "learning_rate": 1.1738070675256773e-06, + "loss": 0.1934, + "step": 3491 + }, + { + "epoch": 0.7945392491467577, + "grad_norm": 1.8153820624252357, + "learning_rate": 1.1737643079233763e-06, + "loss": 0.0786, + "step": 3492 + }, + { + "epoch": 0.7947667804323094, + "grad_norm": 1.7671049533475054, + "learning_rate": 1.1737215371053376e-06, + "loss": 0.0866, + "step": 3493 + }, + { + "epoch": 0.7949943117178612, + "grad_norm": 1.4867713437127692, + "learning_rate": 1.1736787550724357e-06, + "loss": 0.1448, + "step": 3494 + }, + { + "epoch": 0.7952218430034129, + "grad_norm": 2.749928140643179, + "learning_rate": 1.1736359618255452e-06, + "loss": 0.16, + "step": 3495 + }, + { + "epoch": 0.7954493742889648, + "grad_norm": 3.118689887471958, + "learning_rate": 1.1735931573655402e-06, + "loss": 0.2245, + "step": 3496 + }, + { + "epoch": 0.7956769055745165, + "grad_norm": 4.92304830305448, + "learning_rate": 1.1735503416932957e-06, + "loss": 0.1682, + "step": 3497 + }, + { + "epoch": 0.7959044368600683, + "grad_norm": 0.7816490382403714, + "learning_rate": 1.1735075148096869e-06, + "loss": 0.0881, + "step": 3498 + }, + { + "epoch": 0.79613196814562, + "grad_norm": 2.83082513127168, + "learning_rate": 1.173464676715589e-06, + "loss": 0.1574, + "step": 3499 + }, + { + "epoch": 0.7963594994311718, + "grad_norm": 1.8736840297636916, + "learning_rate": 1.1734218274118775e-06, + "loss": 0.1357, + "step": 3500 + }, + { + "epoch": 0.7965870307167235, + "grad_norm": 1.5238621219584085, + "learning_rate": 1.1733789668994285e-06, + "loss": 0.188, + "step": 3501 + }, + { + "epoch": 0.7968145620022753, + "grad_norm": 1.990641449188006, + "learning_rate": 1.1733360951791176e-06, + "loss": 0.1838, + "step": 3502 + }, + { + "epoch": 0.797042093287827, + "grad_norm": 1.4319575898148043, + "learning_rate": 1.173293212251821e-06, + "loss": 0.1096, + "step": 3503 + }, + { + "epoch": 0.7972696245733788, + "grad_norm": 2.2086542380222944, + "learning_rate": 1.1732503181184155e-06, + "loss": 0.1326, + "step": 3504 + }, + { + "epoch": 0.7974971558589306, + "grad_norm": 3.1095746415107235, + "learning_rate": 1.1732074127797773e-06, + "loss": 0.1399, + "step": 3505 + }, + { + "epoch": 0.7977246871444824, + "grad_norm": 1.5737796542808713, + "learning_rate": 1.1731644962367838e-06, + "loss": 0.1047, + "step": 3506 + }, + { + "epoch": 0.7979522184300342, + "grad_norm": 1.912201296923179, + "learning_rate": 1.173121568490312e-06, + "loss": 0.1626, + "step": 3507 + }, + { + "epoch": 0.7981797497155859, + "grad_norm": 1.6232060275987237, + "learning_rate": 1.173078629541239e-06, + "loss": 0.2191, + "step": 3508 + }, + { + "epoch": 0.7984072810011377, + "grad_norm": 2.7876686082536333, + "learning_rate": 1.1730356793904426e-06, + "loss": 0.1472, + "step": 3509 + }, + { + "epoch": 0.7986348122866894, + "grad_norm": 3.0524433393481125, + "learning_rate": 1.1729927180388008e-06, + "loss": 0.1744, + "step": 3510 + }, + { + "epoch": 0.7988623435722412, + "grad_norm": 2.533079498985637, + "learning_rate": 1.172949745487191e-06, + "loss": 0.1341, + "step": 3511 + }, + { + "epoch": 0.7990898748577929, + "grad_norm": 2.8158615912420037, + "learning_rate": 1.1729067617364923e-06, + "loss": 0.1398, + "step": 3512 + }, + { + "epoch": 0.7993174061433447, + "grad_norm": 3.810931501836283, + "learning_rate": 1.1728637667875825e-06, + "loss": 0.1603, + "step": 3513 + }, + { + "epoch": 0.7995449374288964, + "grad_norm": 1.3831236821227444, + "learning_rate": 1.172820760641341e-06, + "loss": 0.1667, + "step": 3514 + }, + { + "epoch": 0.7997724687144482, + "grad_norm": 1.5087369009672673, + "learning_rate": 1.1727777432986465e-06, + "loss": 0.0937, + "step": 3515 + }, + { + "epoch": 0.8, + "grad_norm": 1.8545109020075539, + "learning_rate": 1.1727347147603778e-06, + "loss": 0.1271, + "step": 3516 + }, + { + "epoch": 0.8002275312855518, + "grad_norm": 1.5875557587618756, + "learning_rate": 1.1726916750274148e-06, + "loss": 0.1111, + "step": 3517 + }, + { + "epoch": 0.8004550625711035, + "grad_norm": 1.7651224057238886, + "learning_rate": 1.172648624100637e-06, + "loss": 0.1151, + "step": 3518 + }, + { + "epoch": 0.8006825938566553, + "grad_norm": 2.522491282937047, + "learning_rate": 1.1726055619809245e-06, + "loss": 0.1259, + "step": 3519 + }, + { + "epoch": 0.800910125142207, + "grad_norm": 2.9242004594958853, + "learning_rate": 1.172562488669157e-06, + "loss": 0.1981, + "step": 3520 + }, + { + "epoch": 0.8011376564277588, + "grad_norm": 2.4349195989881958, + "learning_rate": 1.172519404166215e-06, + "loss": 0.1747, + "step": 3521 + }, + { + "epoch": 0.8013651877133106, + "grad_norm": 2.262185722438123, + "learning_rate": 1.1724763084729792e-06, + "loss": 0.0748, + "step": 3522 + }, + { + "epoch": 0.8015927189988623, + "grad_norm": 1.7293354987916565, + "learning_rate": 1.1724332015903303e-06, + "loss": 0.0951, + "step": 3523 + }, + { + "epoch": 0.8018202502844141, + "grad_norm": 3.6362391316593756, + "learning_rate": 1.1723900835191494e-06, + "loss": 0.1691, + "step": 3524 + }, + { + "epoch": 0.8020477815699659, + "grad_norm": 1.4941805820798755, + "learning_rate": 1.1723469542603174e-06, + "loss": 0.1029, + "step": 3525 + }, + { + "epoch": 0.8022753128555177, + "grad_norm": 2.142785774100463, + "learning_rate": 1.1723038138147165e-06, + "loss": 0.1449, + "step": 3526 + }, + { + "epoch": 0.8025028441410694, + "grad_norm": 5.041596350366937, + "learning_rate": 1.1722606621832278e-06, + "loss": 0.2721, + "step": 3527 + }, + { + "epoch": 0.8027303754266212, + "grad_norm": 3.309987840845101, + "learning_rate": 1.1722174993667335e-06, + "loss": 0.1324, + "step": 3528 + }, + { + "epoch": 0.8029579067121729, + "grad_norm": 2.4375248585456535, + "learning_rate": 1.1721743253661155e-06, + "loss": 0.1308, + "step": 3529 + }, + { + "epoch": 0.8031854379977247, + "grad_norm": 1.7485109343470884, + "learning_rate": 1.1721311401822566e-06, + "loss": 0.0598, + "step": 3530 + }, + { + "epoch": 0.8034129692832764, + "grad_norm": 2.8466501164494047, + "learning_rate": 1.1720879438160391e-06, + "loss": 0.1275, + "step": 3531 + }, + { + "epoch": 0.8036405005688282, + "grad_norm": 2.1948137095546514, + "learning_rate": 1.1720447362683458e-06, + "loss": 0.1537, + "step": 3532 + }, + { + "epoch": 0.8038680318543799, + "grad_norm": 2.5761273249687355, + "learning_rate": 1.1720015175400602e-06, + "loss": 0.2656, + "step": 3533 + }, + { + "epoch": 0.8040955631399317, + "grad_norm": 1.35303524636575, + "learning_rate": 1.1719582876320655e-06, + "loss": 0.0725, + "step": 3534 + }, + { + "epoch": 0.8043230944254836, + "grad_norm": 1.8582104202289196, + "learning_rate": 1.1719150465452447e-06, + "loss": 0.0731, + "step": 3535 + }, + { + "epoch": 0.8045506257110353, + "grad_norm": 2.3070370218822993, + "learning_rate": 1.1718717942804822e-06, + "loss": 0.182, + "step": 3536 + }, + { + "epoch": 0.8047781569965871, + "grad_norm": 2.9363741121384734, + "learning_rate": 1.1718285308386618e-06, + "loss": 0.1945, + "step": 3537 + }, + { + "epoch": 0.8050056882821388, + "grad_norm": 2.1830272314645605, + "learning_rate": 1.1717852562206678e-06, + "loss": 0.1165, + "step": 3538 + }, + { + "epoch": 0.8052332195676906, + "grad_norm": 3.245711083253953, + "learning_rate": 1.1717419704273844e-06, + "loss": 0.1591, + "step": 3539 + }, + { + "epoch": 0.8054607508532423, + "grad_norm": 2.5490845759819, + "learning_rate": 1.1716986734596963e-06, + "loss": 0.1237, + "step": 3540 + }, + { + "epoch": 0.8056882821387941, + "grad_norm": 2.9048546128584203, + "learning_rate": 1.1716553653184887e-06, + "loss": 0.149, + "step": 3541 + }, + { + "epoch": 0.8059158134243458, + "grad_norm": 3.4610037458517735, + "learning_rate": 1.1716120460046464e-06, + "loss": 0.116, + "step": 3542 + }, + { + "epoch": 0.8061433447098976, + "grad_norm": 1.2797618581011017, + "learning_rate": 1.1715687155190553e-06, + "loss": 0.15, + "step": 3543 + }, + { + "epoch": 0.8063708759954493, + "grad_norm": 2.799222931592063, + "learning_rate": 1.1715253738626005e-06, + "loss": 0.2169, + "step": 3544 + }, + { + "epoch": 0.8065984072810012, + "grad_norm": 2.3387405024192827, + "learning_rate": 1.171482021036168e-06, + "loss": 0.1816, + "step": 3545 + }, + { + "epoch": 0.8068259385665529, + "grad_norm": 1.3614682853594804, + "learning_rate": 1.1714386570406439e-06, + "loss": 0.0795, + "step": 3546 + }, + { + "epoch": 0.8070534698521047, + "grad_norm": 2.35431862689593, + "learning_rate": 1.1713952818769142e-06, + "loss": 0.1802, + "step": 3547 + }, + { + "epoch": 0.8072810011376564, + "grad_norm": 4.073744850230153, + "learning_rate": 1.1713518955458657e-06, + "loss": 0.1367, + "step": 3548 + }, + { + "epoch": 0.8075085324232082, + "grad_norm": 2.2417307320503315, + "learning_rate": 1.1713084980483849e-06, + "loss": 0.1117, + "step": 3549 + }, + { + "epoch": 0.8077360637087599, + "grad_norm": 1.8171121725967154, + "learning_rate": 1.1712650893853591e-06, + "loss": 0.1116, + "step": 3550 + }, + { + "epoch": 0.8079635949943117, + "grad_norm": 1.1273822389879684, + "learning_rate": 1.1712216695576753e-06, + "loss": 0.1209, + "step": 3551 + }, + { + "epoch": 0.8081911262798634, + "grad_norm": 3.0048491196415985, + "learning_rate": 1.171178238566221e-06, + "loss": 0.1695, + "step": 3552 + }, + { + "epoch": 0.8084186575654152, + "grad_norm": 2.968285056264807, + "learning_rate": 1.1711347964118837e-06, + "loss": 0.1703, + "step": 3553 + }, + { + "epoch": 0.8086461888509671, + "grad_norm": 1.6533852634693318, + "learning_rate": 1.1710913430955514e-06, + "loss": 0.1568, + "step": 3554 + }, + { + "epoch": 0.8088737201365188, + "grad_norm": 1.966029753790106, + "learning_rate": 1.171047878618112e-06, + "loss": 0.132, + "step": 3555 + }, + { + "epoch": 0.8091012514220706, + "grad_norm": 1.8837334268649408, + "learning_rate": 1.1710044029804542e-06, + "loss": 0.1739, + "step": 3556 + }, + { + "epoch": 0.8093287827076223, + "grad_norm": 1.5944840560397648, + "learning_rate": 1.1709609161834663e-06, + "loss": 0.1447, + "step": 3557 + }, + { + "epoch": 0.8095563139931741, + "grad_norm": 3.3750416113176547, + "learning_rate": 1.1709174182280371e-06, + "loss": 0.2264, + "step": 3558 + }, + { + "epoch": 0.8097838452787258, + "grad_norm": 1.42052223693585, + "learning_rate": 1.1708739091150557e-06, + "loss": 0.0966, + "step": 3559 + }, + { + "epoch": 0.8100113765642776, + "grad_norm": 1.6640096543568796, + "learning_rate": 1.1708303888454113e-06, + "loss": 0.0897, + "step": 3560 + }, + { + "epoch": 0.8102389078498293, + "grad_norm": 2.489262801628844, + "learning_rate": 1.1707868574199934e-06, + "loss": 0.2036, + "step": 3561 + }, + { + "epoch": 0.8104664391353811, + "grad_norm": 1.7044828952839948, + "learning_rate": 1.1707433148396918e-06, + "loss": 0.112, + "step": 3562 + }, + { + "epoch": 0.8106939704209328, + "grad_norm": 3.36001160307574, + "learning_rate": 1.1706997611053963e-06, + "loss": 0.1259, + "step": 3563 + }, + { + "epoch": 0.8109215017064847, + "grad_norm": 1.6392717376534625, + "learning_rate": 1.1706561962179968e-06, + "loss": 0.1526, + "step": 3564 + }, + { + "epoch": 0.8111490329920364, + "grad_norm": 1.7155808753624902, + "learning_rate": 1.1706126201783844e-06, + "loss": 0.1241, + "step": 3565 + }, + { + "epoch": 0.8113765642775882, + "grad_norm": 2.5616579159310144, + "learning_rate": 1.170569032987449e-06, + "loss": 0.1258, + "step": 3566 + }, + { + "epoch": 0.81160409556314, + "grad_norm": 2.769894221468227, + "learning_rate": 1.1705254346460818e-06, + "loss": 0.0991, + "step": 3567 + }, + { + "epoch": 0.8118316268486917, + "grad_norm": 1.477469197427597, + "learning_rate": 1.170481825155174e-06, + "loss": 0.1278, + "step": 3568 + }, + { + "epoch": 0.8120591581342435, + "grad_norm": 1.9002053859746235, + "learning_rate": 1.1704382045156162e-06, + "loss": 0.1467, + "step": 3569 + }, + { + "epoch": 0.8122866894197952, + "grad_norm": 2.5583118029634395, + "learning_rate": 1.1703945727283008e-06, + "loss": 0.1011, + "step": 3570 + }, + { + "epoch": 0.812514220705347, + "grad_norm": 0.9971154304856114, + "learning_rate": 1.1703509297941193e-06, + "loss": 0.073, + "step": 3571 + }, + { + "epoch": 0.8127417519908987, + "grad_norm": 2.2480762900084996, + "learning_rate": 1.1703072757139633e-06, + "loss": 0.1251, + "step": 3572 + }, + { + "epoch": 0.8129692832764505, + "grad_norm": 3.0737759248208745, + "learning_rate": 1.1702636104887253e-06, + "loss": 0.1322, + "step": 3573 + }, + { + "epoch": 0.8131968145620023, + "grad_norm": 1.8198486335000044, + "learning_rate": 1.1702199341192977e-06, + "loss": 0.1625, + "step": 3574 + }, + { + "epoch": 0.8134243458475541, + "grad_norm": 1.0495962275251278, + "learning_rate": 1.1701762466065733e-06, + "loss": 0.0854, + "step": 3575 + }, + { + "epoch": 0.8136518771331058, + "grad_norm": 1.8541583708777531, + "learning_rate": 1.1701325479514446e-06, + "loss": 0.1218, + "step": 3576 + }, + { + "epoch": 0.8138794084186576, + "grad_norm": 1.6740410358627995, + "learning_rate": 1.1700888381548052e-06, + "loss": 0.1061, + "step": 3577 + }, + { + "epoch": 0.8141069397042093, + "grad_norm": 2.030287883816433, + "learning_rate": 1.1700451172175482e-06, + "loss": 0.1824, + "step": 3578 + }, + { + "epoch": 0.8143344709897611, + "grad_norm": 1.3733587978383417, + "learning_rate": 1.1700013851405673e-06, + "loss": 0.0824, + "step": 3579 + }, + { + "epoch": 0.8145620022753128, + "grad_norm": 2.7300886954563777, + "learning_rate": 1.169957641924756e-06, + "loss": 0.1499, + "step": 3580 + }, + { + "epoch": 0.8147895335608646, + "grad_norm": 2.702060988424549, + "learning_rate": 1.1699138875710087e-06, + "loss": 0.1538, + "step": 3581 + }, + { + "epoch": 0.8150170648464163, + "grad_norm": 2.0159475716435207, + "learning_rate": 1.1698701220802194e-06, + "loss": 0.1436, + "step": 3582 + }, + { + "epoch": 0.8152445961319681, + "grad_norm": 1.9482556961132234, + "learning_rate": 1.1698263454532827e-06, + "loss": 0.1107, + "step": 3583 + }, + { + "epoch": 0.81547212741752, + "grad_norm": 3.3242671905337766, + "learning_rate": 1.1697825576910933e-06, + "loss": 0.1605, + "step": 3584 + }, + { + "epoch": 0.8156996587030717, + "grad_norm": 4.406286080021148, + "learning_rate": 1.1697387587945461e-06, + "loss": 0.1412, + "step": 3585 + }, + { + "epoch": 0.8159271899886235, + "grad_norm": 3.521875860215498, + "learning_rate": 1.1696949487645365e-06, + "loss": 0.1904, + "step": 3586 + }, + { + "epoch": 0.8161547212741752, + "grad_norm": 2.09054171831565, + "learning_rate": 1.1696511276019595e-06, + "loss": 0.2462, + "step": 3587 + }, + { + "epoch": 0.816382252559727, + "grad_norm": 2.27420287387407, + "learning_rate": 1.169607295307711e-06, + "loss": 0.1244, + "step": 3588 + }, + { + "epoch": 0.8166097838452787, + "grad_norm": 2.2747464420359145, + "learning_rate": 1.169563451882687e-06, + "loss": 0.1764, + "step": 3589 + }, + { + "epoch": 0.8168373151308305, + "grad_norm": 2.7182648228191653, + "learning_rate": 1.1695195973277831e-06, + "loss": 0.1134, + "step": 3590 + }, + { + "epoch": 0.8170648464163822, + "grad_norm": 1.5728156279782024, + "learning_rate": 1.169475731643896e-06, + "loss": 0.0988, + "step": 3591 + }, + { + "epoch": 0.817292377701934, + "grad_norm": 1.8757192624992662, + "learning_rate": 1.169431854831922e-06, + "loss": 0.1089, + "step": 3592 + }, + { + "epoch": 0.8175199089874858, + "grad_norm": 1.1856883759314114, + "learning_rate": 1.169387966892758e-06, + "loss": 0.1476, + "step": 3593 + }, + { + "epoch": 0.8177474402730376, + "grad_norm": 1.915281752164245, + "learning_rate": 1.169344067827301e-06, + "loss": 0.1859, + "step": 3594 + }, + { + "epoch": 0.8179749715585893, + "grad_norm": 1.641433371476581, + "learning_rate": 1.1693001576364483e-06, + "loss": 0.147, + "step": 3595 + }, + { + "epoch": 0.8182025028441411, + "grad_norm": 1.9995302113667346, + "learning_rate": 1.1692562363210969e-06, + "loss": 0.1197, + "step": 3596 + }, + { + "epoch": 0.8184300341296928, + "grad_norm": 1.6487131000503967, + "learning_rate": 1.169212303882145e-06, + "loss": 0.0894, + "step": 3597 + }, + { + "epoch": 0.8186575654152446, + "grad_norm": 1.0718237003787452, + "learning_rate": 1.1691683603204901e-06, + "loss": 0.1471, + "step": 3598 + }, + { + "epoch": 0.8188850967007963, + "grad_norm": 1.9872022609356077, + "learning_rate": 1.1691244056370307e-06, + "loss": 0.0775, + "step": 3599 + }, + { + "epoch": 0.8191126279863481, + "grad_norm": 1.675911881257451, + "learning_rate": 1.169080439832665e-06, + "loss": 0.0875, + "step": 3600 + }, + { + "epoch": 0.8193401592718998, + "grad_norm": 2.8463715468246105, + "learning_rate": 1.1690364629082915e-06, + "loss": 0.1676, + "step": 3601 + }, + { + "epoch": 0.8195676905574516, + "grad_norm": 1.860062069008861, + "learning_rate": 1.168992474864809e-06, + "loss": 0.0992, + "step": 3602 + }, + { + "epoch": 0.8197952218430035, + "grad_norm": 2.4111719114216763, + "learning_rate": 1.1689484757031167e-06, + "loss": 0.1242, + "step": 3603 + }, + { + "epoch": 0.8200227531285552, + "grad_norm": 1.804420569976691, + "learning_rate": 1.1689044654241136e-06, + "loss": 0.1382, + "step": 3604 + }, + { + "epoch": 0.820250284414107, + "grad_norm": 1.5941220037037336, + "learning_rate": 1.1688604440286994e-06, + "loss": 0.1924, + "step": 3605 + }, + { + "epoch": 0.8204778156996587, + "grad_norm": 1.5263579220058776, + "learning_rate": 1.1688164115177737e-06, + "loss": 0.1021, + "step": 3606 + }, + { + "epoch": 0.8207053469852105, + "grad_norm": 2.8075035415219505, + "learning_rate": 1.1687723678922367e-06, + "loss": 0.1988, + "step": 3607 + }, + { + "epoch": 0.8209328782707622, + "grad_norm": 1.7318635873604538, + "learning_rate": 1.1687283131529882e-06, + "loss": 0.1329, + "step": 3608 + }, + { + "epoch": 0.821160409556314, + "grad_norm": 1.5626567981027735, + "learning_rate": 1.1686842473009286e-06, + "loss": 0.089, + "step": 3609 + }, + { + "epoch": 0.8213879408418657, + "grad_norm": 1.7420355931507356, + "learning_rate": 1.168640170336959e-06, + "loss": 0.0914, + "step": 3610 + }, + { + "epoch": 0.8216154721274175, + "grad_norm": 1.9844192517974875, + "learning_rate": 1.16859608226198e-06, + "loss": 0.1082, + "step": 3611 + }, + { + "epoch": 0.8218430034129692, + "grad_norm": 1.9480179532884792, + "learning_rate": 1.1685519830768923e-06, + "loss": 0.1706, + "step": 3612 + }, + { + "epoch": 0.8220705346985211, + "grad_norm": 2.335868396841245, + "learning_rate": 1.1685078727825978e-06, + "loss": 0.1348, + "step": 3613 + }, + { + "epoch": 0.8222980659840728, + "grad_norm": 2.2509068386811113, + "learning_rate": 1.1684637513799976e-06, + "loss": 0.1105, + "step": 3614 + }, + { + "epoch": 0.8225255972696246, + "grad_norm": 1.385199024522045, + "learning_rate": 1.1684196188699936e-06, + "loss": 0.1107, + "step": 3615 + }, + { + "epoch": 0.8227531285551763, + "grad_norm": 2.721417628687412, + "learning_rate": 1.168375475253488e-06, + "loss": 0.2143, + "step": 3616 + }, + { + "epoch": 0.8229806598407281, + "grad_norm": 2.1327255768939564, + "learning_rate": 1.168331320531383e-06, + "loss": 0.1029, + "step": 3617 + }, + { + "epoch": 0.8232081911262799, + "grad_norm": 2.2052514638307312, + "learning_rate": 1.1682871547045805e-06, + "loss": 0.1301, + "step": 3618 + }, + { + "epoch": 0.8234357224118316, + "grad_norm": 2.295794037048131, + "learning_rate": 1.1682429777739836e-06, + "loss": 0.1319, + "step": 3619 + }, + { + "epoch": 0.8236632536973834, + "grad_norm": 1.7346497245119692, + "learning_rate": 1.168198789740495e-06, + "loss": 0.1474, + "step": 3620 + }, + { + "epoch": 0.8238907849829351, + "grad_norm": 1.4867035615045876, + "learning_rate": 1.1681545906050184e-06, + "loss": 0.0918, + "step": 3621 + }, + { + "epoch": 0.824118316268487, + "grad_norm": 1.5853265849774527, + "learning_rate": 1.1681103803684564e-06, + "loss": 0.1508, + "step": 3622 + }, + { + "epoch": 0.8243458475540387, + "grad_norm": 2.6635300104778876, + "learning_rate": 1.168066159031713e-06, + "loss": 0.1551, + "step": 3623 + }, + { + "epoch": 0.8245733788395905, + "grad_norm": 3.10455139314672, + "learning_rate": 1.1680219265956918e-06, + "loss": 0.186, + "step": 3624 + }, + { + "epoch": 0.8248009101251422, + "grad_norm": 2.6599031485707387, + "learning_rate": 1.167977683061297e-06, + "loss": 0.0953, + "step": 3625 + }, + { + "epoch": 0.825028441410694, + "grad_norm": 2.0712725542677957, + "learning_rate": 1.1679334284294328e-06, + "loss": 0.1206, + "step": 3626 + }, + { + "epoch": 0.8252559726962457, + "grad_norm": 1.339098894932053, + "learning_rate": 1.1678891627010036e-06, + "loss": 0.1296, + "step": 3627 + }, + { + "epoch": 0.8254835039817975, + "grad_norm": 1.3418346832410328, + "learning_rate": 1.1678448858769139e-06, + "loss": 0.1069, + "step": 3628 + }, + { + "epoch": 0.8257110352673492, + "grad_norm": 2.300704895731321, + "learning_rate": 1.1678005979580694e-06, + "loss": 0.232, + "step": 3629 + }, + { + "epoch": 0.825938566552901, + "grad_norm": 2.565692079837086, + "learning_rate": 1.1677562989453745e-06, + "loss": 0.1685, + "step": 3630 + }, + { + "epoch": 0.8261660978384527, + "grad_norm": 2.7439979743790177, + "learning_rate": 1.1677119888397348e-06, + "loss": 0.2008, + "step": 3631 + }, + { + "epoch": 0.8263936291240046, + "grad_norm": 2.3355350842028133, + "learning_rate": 1.1676676676420562e-06, + "loss": 0.1449, + "step": 3632 + }, + { + "epoch": 0.8266211604095564, + "grad_norm": 2.6558322573348927, + "learning_rate": 1.1676233353532442e-06, + "loss": 0.1475, + "step": 3633 + }, + { + "epoch": 0.8268486916951081, + "grad_norm": 2.694804298395181, + "learning_rate": 1.167578991974205e-06, + "loss": 0.1267, + "step": 3634 + }, + { + "epoch": 0.8270762229806599, + "grad_norm": 1.3844578486757833, + "learning_rate": 1.167534637505845e-06, + "loss": 0.1051, + "step": 3635 + }, + { + "epoch": 0.8273037542662116, + "grad_norm": 3.8021262153636513, + "learning_rate": 1.1674902719490704e-06, + "loss": 0.1409, + "step": 3636 + }, + { + "epoch": 0.8275312855517634, + "grad_norm": 1.9081376942490276, + "learning_rate": 1.1674458953047883e-06, + "loss": 0.1318, + "step": 3637 + }, + { + "epoch": 0.8277588168373151, + "grad_norm": 1.5941119751883612, + "learning_rate": 1.1674015075739054e-06, + "loss": 0.1643, + "step": 3638 + }, + { + "epoch": 0.8279863481228669, + "grad_norm": 4.574871513487587, + "learning_rate": 1.1673571087573293e-06, + "loss": 0.166, + "step": 3639 + }, + { + "epoch": 0.8282138794084186, + "grad_norm": 3.608726841283483, + "learning_rate": 1.167312698855967e-06, + "loss": 0.1647, + "step": 3640 + }, + { + "epoch": 0.8284414106939704, + "grad_norm": 1.5402864670776968, + "learning_rate": 1.1672682778707262e-06, + "loss": 0.0943, + "step": 3641 + }, + { + "epoch": 0.8286689419795222, + "grad_norm": 1.1120726912524128, + "learning_rate": 1.167223845802515e-06, + "loss": 0.1138, + "step": 3642 + }, + { + "epoch": 0.828896473265074, + "grad_norm": 2.078866420634302, + "learning_rate": 1.1671794026522417e-06, + "loss": 0.1944, + "step": 3643 + }, + { + "epoch": 0.8291240045506257, + "grad_norm": 2.041196689513367, + "learning_rate": 1.1671349484208142e-06, + "loss": 0.1759, + "step": 3644 + }, + { + "epoch": 0.8293515358361775, + "grad_norm": 3.0074049329788144, + "learning_rate": 1.1670904831091412e-06, + "loss": 0.1606, + "step": 3645 + }, + { + "epoch": 0.8295790671217292, + "grad_norm": 2.414856061249191, + "learning_rate": 1.1670460067181313e-06, + "loss": 0.1622, + "step": 3646 + }, + { + "epoch": 0.829806598407281, + "grad_norm": 1.6822279974303718, + "learning_rate": 1.167001519248694e-06, + "loss": 0.2434, + "step": 3647 + }, + { + "epoch": 0.8300341296928327, + "grad_norm": 2.6099405415593475, + "learning_rate": 1.1669570207017384e-06, + "loss": 0.1426, + "step": 3648 + }, + { + "epoch": 0.8302616609783845, + "grad_norm": 2.396884053282652, + "learning_rate": 1.1669125110781737e-06, + "loss": 0.1275, + "step": 3649 + }, + { + "epoch": 0.8304891922639362, + "grad_norm": 1.9075049230967023, + "learning_rate": 1.1668679903789095e-06, + "loss": 0.1403, + "step": 3650 + }, + { + "epoch": 0.830716723549488, + "grad_norm": 2.2255645987859727, + "learning_rate": 1.1668234586048562e-06, + "loss": 0.11, + "step": 3651 + }, + { + "epoch": 0.8309442548350399, + "grad_norm": 1.1833888824062964, + "learning_rate": 1.1667789157569237e-06, + "loss": 0.1392, + "step": 3652 + }, + { + "epoch": 0.8311717861205916, + "grad_norm": 2.4325462275271423, + "learning_rate": 1.1667343618360224e-06, + "loss": 0.1786, + "step": 3653 + }, + { + "epoch": 0.8313993174061434, + "grad_norm": 1.3710217277398862, + "learning_rate": 1.166689796843063e-06, + "loss": 0.1462, + "step": 3654 + }, + { + "epoch": 0.8316268486916951, + "grad_norm": 1.5198212298672198, + "learning_rate": 1.166645220778956e-06, + "loss": 0.1459, + "step": 3655 + }, + { + "epoch": 0.8318543799772469, + "grad_norm": 1.628569489693494, + "learning_rate": 1.1666006336446127e-06, + "loss": 0.1202, + "step": 3656 + }, + { + "epoch": 0.8320819112627986, + "grad_norm": 3.30227173439614, + "learning_rate": 1.1665560354409444e-06, + "loss": 0.1609, + "step": 3657 + }, + { + "epoch": 0.8323094425483504, + "grad_norm": 1.4038774964979965, + "learning_rate": 1.1665114261688625e-06, + "loss": 0.1544, + "step": 3658 + }, + { + "epoch": 0.8325369738339021, + "grad_norm": 1.821635366527404, + "learning_rate": 1.1664668058292789e-06, + "loss": 0.0919, + "step": 3659 + }, + { + "epoch": 0.8327645051194539, + "grad_norm": 2.0806661475154695, + "learning_rate": 1.1664221744231052e-06, + "loss": 0.1351, + "step": 3660 + }, + { + "epoch": 0.8329920364050057, + "grad_norm": 2.629918294593467, + "learning_rate": 1.1663775319512539e-06, + "loss": 0.1379, + "step": 3661 + }, + { + "epoch": 0.8332195676905575, + "grad_norm": 1.8266189295452462, + "learning_rate": 1.1663328784146375e-06, + "loss": 0.1339, + "step": 3662 + }, + { + "epoch": 0.8334470989761092, + "grad_norm": 2.567173056748027, + "learning_rate": 1.1662882138141684e-06, + "loss": 0.1347, + "step": 3663 + }, + { + "epoch": 0.833674630261661, + "grad_norm": 1.8059501719212914, + "learning_rate": 1.1662435381507595e-06, + "loss": 0.1445, + "step": 3664 + }, + { + "epoch": 0.8339021615472128, + "grad_norm": 1.28119552202134, + "learning_rate": 1.1661988514253237e-06, + "loss": 0.1317, + "step": 3665 + }, + { + "epoch": 0.8341296928327645, + "grad_norm": 0.7832646833508663, + "learning_rate": 1.1661541536387748e-06, + "loss": 0.0444, + "step": 3666 + }, + { + "epoch": 0.8343572241183163, + "grad_norm": 2.4326662770770673, + "learning_rate": 1.166109444792026e-06, + "loss": 0.183, + "step": 3667 + }, + { + "epoch": 0.834584755403868, + "grad_norm": 1.5964670632363234, + "learning_rate": 1.1660647248859913e-06, + "loss": 0.1303, + "step": 3668 + }, + { + "epoch": 0.8348122866894198, + "grad_norm": 1.9117342096036827, + "learning_rate": 1.1660199939215845e-06, + "loss": 0.1007, + "step": 3669 + }, + { + "epoch": 0.8350398179749715, + "grad_norm": 2.605848649739405, + "learning_rate": 1.1659752518997197e-06, + "loss": 0.1419, + "step": 3670 + }, + { + "epoch": 0.8352673492605234, + "grad_norm": 1.7438484870911297, + "learning_rate": 1.1659304988213115e-06, + "loss": 0.2068, + "step": 3671 + }, + { + "epoch": 0.8354948805460751, + "grad_norm": 1.0600642685383936, + "learning_rate": 1.1658857346872745e-06, + "loss": 0.1193, + "step": 3672 + }, + { + "epoch": 0.8357224118316269, + "grad_norm": 1.9518189045588334, + "learning_rate": 1.165840959498524e-06, + "loss": 0.1262, + "step": 3673 + }, + { + "epoch": 0.8359499431171786, + "grad_norm": 1.7335154761632459, + "learning_rate": 1.1657961732559745e-06, + "loss": 0.1009, + "step": 3674 + }, + { + "epoch": 0.8361774744027304, + "grad_norm": 2.2309685306176017, + "learning_rate": 1.1657513759605417e-06, + "loss": 0.1784, + "step": 3675 + }, + { + "epoch": 0.8364050056882821, + "grad_norm": 2.2907312733247194, + "learning_rate": 1.1657065676131412e-06, + "loss": 0.1291, + "step": 3676 + }, + { + "epoch": 0.8366325369738339, + "grad_norm": 1.6294624150585804, + "learning_rate": 1.1656617482146886e-06, + "loss": 0.1442, + "step": 3677 + }, + { + "epoch": 0.8368600682593856, + "grad_norm": 2.8212610764914707, + "learning_rate": 1.1656169177661e-06, + "loss": 0.1711, + "step": 3678 + }, + { + "epoch": 0.8370875995449374, + "grad_norm": 2.076450457598849, + "learning_rate": 1.165572076268292e-06, + "loss": 0.2057, + "step": 3679 + }, + { + "epoch": 0.8373151308304891, + "grad_norm": 3.366007033872947, + "learning_rate": 1.1655272237221804e-06, + "loss": 0.1565, + "step": 3680 + }, + { + "epoch": 0.837542662116041, + "grad_norm": 2.2261279746795846, + "learning_rate": 1.1654823601286826e-06, + "loss": 0.1642, + "step": 3681 + }, + { + "epoch": 0.8377701934015928, + "grad_norm": 1.81724632660024, + "learning_rate": 1.165437485488715e-06, + "loss": 0.0943, + "step": 3682 + }, + { + "epoch": 0.8379977246871445, + "grad_norm": 3.1269822903807225, + "learning_rate": 1.165392599803195e-06, + "loss": 0.1449, + "step": 3683 + }, + { + "epoch": 0.8382252559726963, + "grad_norm": 2.379708780504192, + "learning_rate": 1.16534770307304e-06, + "loss": 0.153, + "step": 3684 + }, + { + "epoch": 0.838452787258248, + "grad_norm": 1.8603168211221506, + "learning_rate": 1.1653027952991676e-06, + "loss": 0.1143, + "step": 3685 + }, + { + "epoch": 0.8386803185437998, + "grad_norm": 2.2831231162185417, + "learning_rate": 1.1652578764824953e-06, + "loss": 0.1459, + "step": 3686 + }, + { + "epoch": 0.8389078498293515, + "grad_norm": 2.8451139022978085, + "learning_rate": 1.1652129466239417e-06, + "loss": 0.1634, + "step": 3687 + }, + { + "epoch": 0.8391353811149033, + "grad_norm": 2.3009781892443573, + "learning_rate": 1.1651680057244247e-06, + "loss": 0.1378, + "step": 3688 + }, + { + "epoch": 0.839362912400455, + "grad_norm": 4.0903919882686814, + "learning_rate": 1.165123053784863e-06, + "loss": 0.1721, + "step": 3689 + }, + { + "epoch": 0.8395904436860068, + "grad_norm": 1.7433135774699924, + "learning_rate": 1.1650780908061753e-06, + "loss": 0.1282, + "step": 3690 + }, + { + "epoch": 0.8398179749715586, + "grad_norm": 1.3400959169617475, + "learning_rate": 1.1650331167892805e-06, + "loss": 0.1631, + "step": 3691 + }, + { + "epoch": 0.8400455062571104, + "grad_norm": 2.56860477659448, + "learning_rate": 1.1649881317350979e-06, + "loss": 0.1444, + "step": 3692 + }, + { + "epoch": 0.8402730375426621, + "grad_norm": 1.7498028154005396, + "learning_rate": 1.1649431356445467e-06, + "loss": 0.1098, + "step": 3693 + }, + { + "epoch": 0.8405005688282139, + "grad_norm": 1.4621069776019282, + "learning_rate": 1.1648981285185464e-06, + "loss": 0.1369, + "step": 3694 + }, + { + "epoch": 0.8407281001137656, + "grad_norm": 2.254007653069776, + "learning_rate": 1.1648531103580178e-06, + "loss": 0.2232, + "step": 3695 + }, + { + "epoch": 0.8409556313993174, + "grad_norm": 2.03443979756766, + "learning_rate": 1.1648080811638797e-06, + "loss": 0.1291, + "step": 3696 + }, + { + "epoch": 0.8411831626848691, + "grad_norm": 1.3415702204257411, + "learning_rate": 1.1647630409370533e-06, + "loss": 0.1113, + "step": 3697 + }, + { + "epoch": 0.8414106939704209, + "grad_norm": 2.0728006792206086, + "learning_rate": 1.1647179896784588e-06, + "loss": 0.1298, + "step": 3698 + }, + { + "epoch": 0.8416382252559726, + "grad_norm": 2.7540588294232253, + "learning_rate": 1.1646729273890172e-06, + "loss": 0.1053, + "step": 3699 + }, + { + "epoch": 0.8418657565415245, + "grad_norm": 2.407303374360263, + "learning_rate": 1.1646278540696493e-06, + "loss": 0.1237, + "step": 3700 + }, + { + "epoch": 0.8420932878270763, + "grad_norm": 1.6348359597782052, + "learning_rate": 1.1645827697212762e-06, + "loss": 0.0736, + "step": 3701 + }, + { + "epoch": 0.842320819112628, + "grad_norm": 2.1217116646699474, + "learning_rate": 1.1645376743448194e-06, + "loss": 0.1372, + "step": 3702 + }, + { + "epoch": 0.8425483503981798, + "grad_norm": 1.4266722248368366, + "learning_rate": 1.1644925679412008e-06, + "loss": 0.124, + "step": 3703 + }, + { + "epoch": 0.8427758816837315, + "grad_norm": 0.988882296235461, + "learning_rate": 1.1644474505113423e-06, + "loss": 0.1114, + "step": 3704 + }, + { + "epoch": 0.8430034129692833, + "grad_norm": 1.4548572895040972, + "learning_rate": 1.1644023220561657e-06, + "loss": 0.1811, + "step": 3705 + }, + { + "epoch": 0.843230944254835, + "grad_norm": 2.0960066842624427, + "learning_rate": 1.1643571825765935e-06, + "loss": 0.1543, + "step": 3706 + }, + { + "epoch": 0.8434584755403868, + "grad_norm": 3.0082397132223155, + "learning_rate": 1.1643120320735481e-06, + "loss": 0.2195, + "step": 3707 + }, + { + "epoch": 0.8436860068259385, + "grad_norm": 2.7601349795873378, + "learning_rate": 1.164266870547953e-06, + "loss": 0.1608, + "step": 3708 + }, + { + "epoch": 0.8439135381114903, + "grad_norm": 3.232912460236744, + "learning_rate": 1.16422169800073e-06, + "loss": 0.1857, + "step": 3709 + }, + { + "epoch": 0.8441410693970421, + "grad_norm": 3.7374128521838665, + "learning_rate": 1.1641765144328035e-06, + "loss": 0.2689, + "step": 3710 + }, + { + "epoch": 0.8443686006825939, + "grad_norm": 1.5630753818611163, + "learning_rate": 1.1641313198450966e-06, + "loss": 0.1372, + "step": 3711 + }, + { + "epoch": 0.8445961319681456, + "grad_norm": 1.8779453375789192, + "learning_rate": 1.1640861142385326e-06, + "loss": 0.0862, + "step": 3712 + }, + { + "epoch": 0.8448236632536974, + "grad_norm": 3.2220024645366028, + "learning_rate": 1.1640408976140358e-06, + "loss": 0.1589, + "step": 3713 + }, + { + "epoch": 0.8450511945392492, + "grad_norm": 2.5580803791843376, + "learning_rate": 1.1639956699725303e-06, + "loss": 0.1293, + "step": 3714 + }, + { + "epoch": 0.8452787258248009, + "grad_norm": 2.096403600241952, + "learning_rate": 1.1639504313149403e-06, + "loss": 0.0954, + "step": 3715 + }, + { + "epoch": 0.8455062571103527, + "grad_norm": 2.080786160223831, + "learning_rate": 1.1639051816421906e-06, + "loss": 0.13, + "step": 3716 + }, + { + "epoch": 0.8457337883959044, + "grad_norm": 1.6759172842352954, + "learning_rate": 1.163859920955206e-06, + "loss": 0.1121, + "step": 3717 + }, + { + "epoch": 0.8459613196814562, + "grad_norm": 2.0642665653257346, + "learning_rate": 1.1638146492549112e-06, + "loss": 0.1178, + "step": 3718 + }, + { + "epoch": 0.8461888509670079, + "grad_norm": 1.5058802030094345, + "learning_rate": 1.163769366542232e-06, + "loss": 0.0927, + "step": 3719 + }, + { + "epoch": 0.8464163822525598, + "grad_norm": 1.308746050377977, + "learning_rate": 1.1637240728180937e-06, + "loss": 0.0819, + "step": 3720 + }, + { + "epoch": 0.8466439135381115, + "grad_norm": 2.6006877199397747, + "learning_rate": 1.163678768083422e-06, + "loss": 0.1516, + "step": 3721 + }, + { + "epoch": 0.8468714448236633, + "grad_norm": 1.8224971272064627, + "learning_rate": 1.1636334523391426e-06, + "loss": 0.09, + "step": 3722 + }, + { + "epoch": 0.847098976109215, + "grad_norm": 1.6032671914851704, + "learning_rate": 1.1635881255861821e-06, + "loss": 0.1375, + "step": 3723 + }, + { + "epoch": 0.8473265073947668, + "grad_norm": 1.726720538161939, + "learning_rate": 1.1635427878254667e-06, + "loss": 0.1176, + "step": 3724 + }, + { + "epoch": 0.8475540386803185, + "grad_norm": 3.7139895273320915, + "learning_rate": 1.163497439057923e-06, + "loss": 0.1631, + "step": 3725 + }, + { + "epoch": 0.8477815699658703, + "grad_norm": 1.1481004566162971, + "learning_rate": 1.1634520792844778e-06, + "loss": 0.1331, + "step": 3726 + }, + { + "epoch": 0.848009101251422, + "grad_norm": 1.947322963642618, + "learning_rate": 1.1634067085060583e-06, + "loss": 0.1086, + "step": 3727 + }, + { + "epoch": 0.8482366325369738, + "grad_norm": 1.9494034977033496, + "learning_rate": 1.1633613267235915e-06, + "loss": 0.1835, + "step": 3728 + }, + { + "epoch": 0.8484641638225257, + "grad_norm": 2.408589703999215, + "learning_rate": 1.1633159339380054e-06, + "loss": 0.1851, + "step": 3729 + }, + { + "epoch": 0.8486916951080774, + "grad_norm": 2.4119649285649123, + "learning_rate": 1.1632705301502272e-06, + "loss": 0.1233, + "step": 3730 + }, + { + "epoch": 0.8489192263936292, + "grad_norm": 3.340579250378617, + "learning_rate": 1.1632251153611855e-06, + "loss": 0.1362, + "step": 3731 + }, + { + "epoch": 0.8491467576791809, + "grad_norm": 2.23461351462565, + "learning_rate": 1.163179689571808e-06, + "loss": 0.1592, + "step": 3732 + }, + { + "epoch": 0.8493742889647327, + "grad_norm": 1.9169594273914996, + "learning_rate": 1.1631342527830234e-06, + "loss": 0.0894, + "step": 3733 + }, + { + "epoch": 0.8496018202502844, + "grad_norm": 1.706614485101184, + "learning_rate": 1.16308880499576e-06, + "loss": 0.0964, + "step": 3734 + }, + { + "epoch": 0.8498293515358362, + "grad_norm": 2.212965703656509, + "learning_rate": 1.163043346210947e-06, + "loss": 0.146, + "step": 3735 + }, + { + "epoch": 0.8500568828213879, + "grad_norm": 1.6891386815491374, + "learning_rate": 1.1629978764295133e-06, + "loss": 0.1352, + "step": 3736 + }, + { + "epoch": 0.8502844141069397, + "grad_norm": 2.1697592534447034, + "learning_rate": 1.1629523956523883e-06, + "loss": 0.0917, + "step": 3737 + }, + { + "epoch": 0.8505119453924914, + "grad_norm": 3.1025926867971876, + "learning_rate": 1.1629069038805018e-06, + "loss": 0.1329, + "step": 3738 + }, + { + "epoch": 0.8507394766780433, + "grad_norm": 1.942888178882637, + "learning_rate": 1.162861401114783e-06, + "loss": 0.1248, + "step": 3739 + }, + { + "epoch": 0.850967007963595, + "grad_norm": 1.5004249323078944, + "learning_rate": 1.1628158873561624e-06, + "loss": 0.085, + "step": 3740 + }, + { + "epoch": 0.8511945392491468, + "grad_norm": 2.0290099909974333, + "learning_rate": 1.16277036260557e-06, + "loss": 0.1727, + "step": 3741 + }, + { + "epoch": 0.8514220705346985, + "grad_norm": 1.3697183727422024, + "learning_rate": 1.1627248268639363e-06, + "loss": 0.0708, + "step": 3742 + }, + { + "epoch": 0.8516496018202503, + "grad_norm": 1.4963457644894143, + "learning_rate": 1.1626792801321917e-06, + "loss": 0.1481, + "step": 3743 + }, + { + "epoch": 0.851877133105802, + "grad_norm": 2.250464954679021, + "learning_rate": 1.1626337224112676e-06, + "loss": 0.1561, + "step": 3744 + }, + { + "epoch": 0.8521046643913538, + "grad_norm": 2.1869910077111823, + "learning_rate": 1.1625881537020948e-06, + "loss": 0.1157, + "step": 3745 + }, + { + "epoch": 0.8523321956769055, + "grad_norm": 2.428910593262319, + "learning_rate": 1.1625425740056046e-06, + "loss": 0.1068, + "step": 3746 + }, + { + "epoch": 0.8525597269624573, + "grad_norm": 1.5214154740818184, + "learning_rate": 1.1624969833227287e-06, + "loss": 0.1403, + "step": 3747 + }, + { + "epoch": 0.852787258248009, + "grad_norm": 2.467402534163389, + "learning_rate": 1.1624513816543988e-06, + "loss": 0.138, + "step": 3748 + }, + { + "epoch": 0.8530147895335609, + "grad_norm": 3.3573512145729842, + "learning_rate": 1.162405769001547e-06, + "loss": 0.1799, + "step": 3749 + }, + { + "epoch": 0.8532423208191127, + "grad_norm": 1.5381915258615733, + "learning_rate": 1.1623601453651053e-06, + "loss": 0.106, + "step": 3750 + }, + { + "epoch": 0.8534698521046644, + "grad_norm": 2.0574854880892697, + "learning_rate": 1.1623145107460065e-06, + "loss": 0.0945, + "step": 3751 + }, + { + "epoch": 0.8536973833902162, + "grad_norm": 1.0351595320260978, + "learning_rate": 1.1622688651451833e-06, + "loss": 0.078, + "step": 3752 + }, + { + "epoch": 0.8539249146757679, + "grad_norm": 2.6677954818832332, + "learning_rate": 1.1622232085635683e-06, + "loss": 0.2526, + "step": 3753 + }, + { + "epoch": 0.8541524459613197, + "grad_norm": 2.6487501278609398, + "learning_rate": 1.162177541002095e-06, + "loss": 0.1601, + "step": 3754 + }, + { + "epoch": 0.8543799772468714, + "grad_norm": 0.9301475208416561, + "learning_rate": 1.162131862461696e-06, + "loss": 0.0665, + "step": 3755 + }, + { + "epoch": 0.8546075085324232, + "grad_norm": 1.763957033542155, + "learning_rate": 1.1620861729433062e-06, + "loss": 0.1202, + "step": 3756 + }, + { + "epoch": 0.8548350398179749, + "grad_norm": 2.176479505741929, + "learning_rate": 1.1620404724478582e-06, + "loss": 0.1798, + "step": 3757 + }, + { + "epoch": 0.8550625711035267, + "grad_norm": 2.945225816531013, + "learning_rate": 1.1619947609762867e-06, + "loss": 0.1396, + "step": 3758 + }, + { + "epoch": 0.8552901023890785, + "grad_norm": 2.562908702733841, + "learning_rate": 1.1619490385295255e-06, + "loss": 0.1973, + "step": 3759 + }, + { + "epoch": 0.8555176336746303, + "grad_norm": 0.9401287921099067, + "learning_rate": 1.1619033051085094e-06, + "loss": 0.1027, + "step": 3760 + }, + { + "epoch": 0.855745164960182, + "grad_norm": 1.7243955273221043, + "learning_rate": 1.161857560714173e-06, + "loss": 0.1261, + "step": 3761 + }, + { + "epoch": 0.8559726962457338, + "grad_norm": 2.2572939153056666, + "learning_rate": 1.1618118053474514e-06, + "loss": 0.1806, + "step": 3762 + }, + { + "epoch": 0.8562002275312856, + "grad_norm": 1.6576632747588675, + "learning_rate": 1.1617660390092794e-06, + "loss": 0.1136, + "step": 3763 + }, + { + "epoch": 0.8564277588168373, + "grad_norm": 2.8436603000222784, + "learning_rate": 1.161720261700593e-06, + "loss": 0.1964, + "step": 3764 + }, + { + "epoch": 0.856655290102389, + "grad_norm": 3.1381523860209155, + "learning_rate": 1.161674473422327e-06, + "loss": 0.1301, + "step": 3765 + }, + { + "epoch": 0.8568828213879408, + "grad_norm": 2.469303248081205, + "learning_rate": 1.1616286741754178e-06, + "loss": 0.1002, + "step": 3766 + }, + { + "epoch": 0.8571103526734926, + "grad_norm": 2.8285714025814492, + "learning_rate": 1.1615828639608013e-06, + "loss": 0.1487, + "step": 3767 + }, + { + "epoch": 0.8573378839590444, + "grad_norm": 1.796917081972481, + "learning_rate": 1.1615370427794138e-06, + "loss": 0.2538, + "step": 3768 + }, + { + "epoch": 0.8575654152445962, + "grad_norm": 1.2104132513749966, + "learning_rate": 1.1614912106321916e-06, + "loss": 0.1111, + "step": 3769 + }, + { + "epoch": 0.8577929465301479, + "grad_norm": 2.240630387028681, + "learning_rate": 1.1614453675200718e-06, + "loss": 0.1593, + "step": 3770 + }, + { + "epoch": 0.8580204778156997, + "grad_norm": 2.4065649472522033, + "learning_rate": 1.1613995134439912e-06, + "loss": 0.202, + "step": 3771 + }, + { + "epoch": 0.8582480091012514, + "grad_norm": 2.424583488709643, + "learning_rate": 1.1613536484048866e-06, + "loss": 0.1912, + "step": 3772 + }, + { + "epoch": 0.8584755403868032, + "grad_norm": 2.3925333932617905, + "learning_rate": 1.161307772403696e-06, + "loss": 0.1897, + "step": 3773 + }, + { + "epoch": 0.8587030716723549, + "grad_norm": 2.3199334542035475, + "learning_rate": 1.1612618854413566e-06, + "loss": 0.1913, + "step": 3774 + }, + { + "epoch": 0.8589306029579067, + "grad_norm": 2.3317040406753895, + "learning_rate": 1.1612159875188065e-06, + "loss": 0.1689, + "step": 3775 + }, + { + "epoch": 0.8591581342434584, + "grad_norm": 1.7725857904878792, + "learning_rate": 1.1611700786369835e-06, + "loss": 0.1673, + "step": 3776 + }, + { + "epoch": 0.8593856655290102, + "grad_norm": 1.6084308537687173, + "learning_rate": 1.161124158796826e-06, + "loss": 0.141, + "step": 3777 + }, + { + "epoch": 0.859613196814562, + "grad_norm": 1.2596148838555763, + "learning_rate": 1.1610782279992728e-06, + "loss": 0.0778, + "step": 3778 + }, + { + "epoch": 0.8598407281001138, + "grad_norm": 2.02054838296593, + "learning_rate": 1.161032286245262e-06, + "loss": 0.1876, + "step": 3779 + }, + { + "epoch": 0.8600682593856656, + "grad_norm": 2.687762249569876, + "learning_rate": 1.1609863335357332e-06, + "loss": 0.1482, + "step": 3780 + }, + { + "epoch": 0.8602957906712173, + "grad_norm": 2.3935859487409408, + "learning_rate": 1.1609403698716255e-06, + "loss": 0.1692, + "step": 3781 + }, + { + "epoch": 0.8605233219567691, + "grad_norm": 2.508017368514477, + "learning_rate": 1.160894395253878e-06, + "loss": 0.134, + "step": 3782 + }, + { + "epoch": 0.8607508532423208, + "grad_norm": 1.8303186156957627, + "learning_rate": 1.1608484096834306e-06, + "loss": 0.2355, + "step": 3783 + }, + { + "epoch": 0.8609783845278726, + "grad_norm": 1.8831463075858677, + "learning_rate": 1.1608024131612231e-06, + "loss": 0.1362, + "step": 3784 + }, + { + "epoch": 0.8612059158134243, + "grad_norm": 2.11822052966638, + "learning_rate": 1.1607564056881953e-06, + "loss": 0.0832, + "step": 3785 + }, + { + "epoch": 0.8614334470989761, + "grad_norm": 2.5355097758095386, + "learning_rate": 1.160710387265288e-06, + "loss": 0.1257, + "step": 3786 + }, + { + "epoch": 0.8616609783845278, + "grad_norm": 2.315866021639714, + "learning_rate": 1.1606643578934414e-06, + "loss": 0.1048, + "step": 3787 + }, + { + "epoch": 0.8618885096700797, + "grad_norm": 2.428912559884676, + "learning_rate": 1.1606183175735963e-06, + "loss": 0.1784, + "step": 3788 + }, + { + "epoch": 0.8621160409556314, + "grad_norm": 1.6349014413744898, + "learning_rate": 1.1605722663066938e-06, + "loss": 0.1257, + "step": 3789 + }, + { + "epoch": 0.8623435722411832, + "grad_norm": 1.4193212111174016, + "learning_rate": 1.1605262040936752e-06, + "loss": 0.1551, + "step": 3790 + }, + { + "epoch": 0.8625711035267349, + "grad_norm": 1.8487470971360136, + "learning_rate": 1.1604801309354815e-06, + "loss": 0.1162, + "step": 3791 + }, + { + "epoch": 0.8627986348122867, + "grad_norm": 1.346999735527141, + "learning_rate": 1.1604340468330546e-06, + "loss": 0.1075, + "step": 3792 + }, + { + "epoch": 0.8630261660978384, + "grad_norm": 3.178320117753975, + "learning_rate": 1.1603879517873366e-06, + "loss": 0.1917, + "step": 3793 + }, + { + "epoch": 0.8632536973833902, + "grad_norm": 2.007629059273734, + "learning_rate": 1.160341845799269e-06, + "loss": 0.1435, + "step": 3794 + }, + { + "epoch": 0.863481228668942, + "grad_norm": 1.777320793531153, + "learning_rate": 1.160295728869795e-06, + "loss": 0.165, + "step": 3795 + }, + { + "epoch": 0.8637087599544937, + "grad_norm": 1.0573583615388784, + "learning_rate": 1.1602496009998562e-06, + "loss": 0.1049, + "step": 3796 + }, + { + "epoch": 0.8639362912400455, + "grad_norm": 2.667616509532965, + "learning_rate": 1.160203462190396e-06, + "loss": 0.1379, + "step": 3797 + }, + { + "epoch": 0.8641638225255973, + "grad_norm": 1.6159958386357638, + "learning_rate": 1.1601573124423573e-06, + "loss": 0.1306, + "step": 3798 + }, + { + "epoch": 0.8643913538111491, + "grad_norm": 1.5656967132317656, + "learning_rate": 1.1601111517566831e-06, + "loss": 0.1139, + "step": 3799 + }, + { + "epoch": 0.8646188850967008, + "grad_norm": 1.3820258167187487, + "learning_rate": 1.1600649801343173e-06, + "loss": 0.1014, + "step": 3800 + }, + { + "epoch": 0.8648464163822526, + "grad_norm": 1.39843503895015, + "learning_rate": 1.1600187975762029e-06, + "loss": 0.1087, + "step": 3801 + }, + { + "epoch": 0.8650739476678043, + "grad_norm": 2.1408742589304737, + "learning_rate": 1.159972604083284e-06, + "loss": 0.0972, + "step": 3802 + }, + { + "epoch": 0.8653014789533561, + "grad_norm": 2.26397998281963, + "learning_rate": 1.159926399656505e-06, + "loss": 0.1448, + "step": 3803 + }, + { + "epoch": 0.8655290102389078, + "grad_norm": 2.490895204679328, + "learning_rate": 1.1598801842968103e-06, + "loss": 0.1195, + "step": 3804 + }, + { + "epoch": 0.8657565415244596, + "grad_norm": 1.6194440157886663, + "learning_rate": 1.1598339580051439e-06, + "loss": 0.1018, + "step": 3805 + }, + { + "epoch": 0.8659840728100113, + "grad_norm": 1.3490011154721018, + "learning_rate": 1.159787720782451e-06, + "loss": 0.0966, + "step": 3806 + }, + { + "epoch": 0.8662116040955632, + "grad_norm": 2.507900361953006, + "learning_rate": 1.1597414726296764e-06, + "loss": 0.2012, + "step": 3807 + }, + { + "epoch": 0.8664391353811149, + "grad_norm": 1.2833341524803188, + "learning_rate": 1.1596952135477656e-06, + "loss": 0.1121, + "step": 3808 + }, + { + "epoch": 0.8666666666666667, + "grad_norm": 2.2151027996183927, + "learning_rate": 1.1596489435376638e-06, + "loss": 0.2105, + "step": 3809 + }, + { + "epoch": 0.8668941979522184, + "grad_norm": 1.1996595882366885, + "learning_rate": 1.1596026626003168e-06, + "loss": 0.1457, + "step": 3810 + }, + { + "epoch": 0.8671217292377702, + "grad_norm": 1.4820717336888578, + "learning_rate": 1.1595563707366705e-06, + "loss": 0.1781, + "step": 3811 + }, + { + "epoch": 0.867349260523322, + "grad_norm": 2.635221308330241, + "learning_rate": 1.1595100679476707e-06, + "loss": 0.1454, + "step": 3812 + }, + { + "epoch": 0.8675767918088737, + "grad_norm": 1.6108295995494901, + "learning_rate": 1.1594637542342644e-06, + "loss": 0.106, + "step": 3813 + }, + { + "epoch": 0.8678043230944255, + "grad_norm": 1.8221220205121134, + "learning_rate": 1.1594174295973976e-06, + "loss": 0.1006, + "step": 3814 + }, + { + "epoch": 0.8680318543799772, + "grad_norm": 3.0101316748683664, + "learning_rate": 1.1593710940380172e-06, + "loss": 0.121, + "step": 3815 + }, + { + "epoch": 0.868259385665529, + "grad_norm": 2.735583899461839, + "learning_rate": 1.1593247475570704e-06, + "loss": 0.209, + "step": 3816 + }, + { + "epoch": 0.8684869169510808, + "grad_norm": 0.9832960186884339, + "learning_rate": 1.1592783901555043e-06, + "loss": 0.0808, + "step": 3817 + }, + { + "epoch": 0.8687144482366326, + "grad_norm": 1.4074790278159912, + "learning_rate": 1.1592320218342665e-06, + "loss": 0.0947, + "step": 3818 + }, + { + "epoch": 0.8689419795221843, + "grad_norm": 1.4230263763515514, + "learning_rate": 1.1591856425943044e-06, + "loss": 0.1207, + "step": 3819 + }, + { + "epoch": 0.8691695108077361, + "grad_norm": 1.269490471422441, + "learning_rate": 1.159139252436566e-06, + "loss": 0.1045, + "step": 3820 + }, + { + "epoch": 0.8693970420932878, + "grad_norm": 2.550687247141249, + "learning_rate": 1.1590928513619997e-06, + "loss": 0.1525, + "step": 3821 + }, + { + "epoch": 0.8696245733788396, + "grad_norm": 1.7947776642256512, + "learning_rate": 1.1590464393715536e-06, + "loss": 0.0921, + "step": 3822 + }, + { + "epoch": 0.8698521046643913, + "grad_norm": 1.7311193251335109, + "learning_rate": 1.1590000164661763e-06, + "loss": 0.1531, + "step": 3823 + }, + { + "epoch": 0.8700796359499431, + "grad_norm": 1.5146472578814127, + "learning_rate": 1.1589535826468168e-06, + "loss": 0.1648, + "step": 3824 + }, + { + "epoch": 0.8703071672354948, + "grad_norm": 1.5015916222628989, + "learning_rate": 1.158907137914424e-06, + "loss": 0.1389, + "step": 3825 + }, + { + "epoch": 0.8705346985210466, + "grad_norm": 1.8122124285787728, + "learning_rate": 1.158860682269947e-06, + "loss": 0.1037, + "step": 3826 + }, + { + "epoch": 0.8707622298065985, + "grad_norm": 2.833296008494366, + "learning_rate": 1.1588142157143353e-06, + "loss": 0.1188, + "step": 3827 + }, + { + "epoch": 0.8709897610921502, + "grad_norm": 2.216618266323227, + "learning_rate": 1.1587677382485386e-06, + "loss": 0.1465, + "step": 3828 + }, + { + "epoch": 0.871217292377702, + "grad_norm": 1.2793900862334322, + "learning_rate": 1.158721249873507e-06, + "loss": 0.1095, + "step": 3829 + }, + { + "epoch": 0.8714448236632537, + "grad_norm": 1.8941138487312816, + "learning_rate": 1.1586747505901904e-06, + "loss": 0.1681, + "step": 3830 + }, + { + "epoch": 0.8716723549488055, + "grad_norm": 2.3145356477245405, + "learning_rate": 1.1586282403995395e-06, + "loss": 0.2006, + "step": 3831 + }, + { + "epoch": 0.8718998862343572, + "grad_norm": 2.6600034233732677, + "learning_rate": 1.1585817193025046e-06, + "loss": 0.1121, + "step": 3832 + }, + { + "epoch": 0.872127417519909, + "grad_norm": 1.725320598895434, + "learning_rate": 1.1585351873000365e-06, + "loss": 0.082, + "step": 3833 + }, + { + "epoch": 0.8723549488054607, + "grad_norm": 1.8752752336053349, + "learning_rate": 1.1584886443930863e-06, + "loss": 0.0895, + "step": 3834 + }, + { + "epoch": 0.8725824800910125, + "grad_norm": 2.751725576098527, + "learning_rate": 1.1584420905826051e-06, + "loss": 0.1928, + "step": 3835 + }, + { + "epoch": 0.8728100113765643, + "grad_norm": 1.9424867364532703, + "learning_rate": 1.1583955258695447e-06, + "loss": 0.1673, + "step": 3836 + }, + { + "epoch": 0.8730375426621161, + "grad_norm": 1.4261787932440506, + "learning_rate": 1.1583489502548566e-06, + "loss": 0.2017, + "step": 3837 + }, + { + "epoch": 0.8732650739476678, + "grad_norm": 2.6521619288755085, + "learning_rate": 1.1583023637394928e-06, + "loss": 0.1423, + "step": 3838 + }, + { + "epoch": 0.8734926052332196, + "grad_norm": 1.4138016233960111, + "learning_rate": 1.1582557663244052e-06, + "loss": 0.1816, + "step": 3839 + }, + { + "epoch": 0.8737201365187713, + "grad_norm": 1.6328060502150679, + "learning_rate": 1.1582091580105464e-06, + "loss": 0.1092, + "step": 3840 + }, + { + "epoch": 0.8739476678043231, + "grad_norm": 2.337312865938115, + "learning_rate": 1.158162538798869e-06, + "loss": 0.1281, + "step": 3841 + }, + { + "epoch": 0.8741751990898748, + "grad_norm": 1.98329071301972, + "learning_rate": 1.158115908690326e-06, + "loss": 0.2402, + "step": 3842 + }, + { + "epoch": 0.8744027303754266, + "grad_norm": 1.685782018846633, + "learning_rate": 1.1580692676858699e-06, + "loss": 0.1349, + "step": 3843 + }, + { + "epoch": 0.8746302616609783, + "grad_norm": 2.185499848922571, + "learning_rate": 1.1580226157864542e-06, + "loss": 0.1586, + "step": 3844 + }, + { + "epoch": 0.8748577929465301, + "grad_norm": 1.8272041921707005, + "learning_rate": 1.1579759529930324e-06, + "loss": 0.0838, + "step": 3845 + }, + { + "epoch": 0.875085324232082, + "grad_norm": 2.854267276269833, + "learning_rate": 1.1579292793065583e-06, + "loss": 0.1751, + "step": 3846 + }, + { + "epoch": 0.8753128555176337, + "grad_norm": 1.8839936940679116, + "learning_rate": 1.157882594727986e-06, + "loss": 0.118, + "step": 3847 + }, + { + "epoch": 0.8755403868031855, + "grad_norm": 1.2659058041051248, + "learning_rate": 1.1578358992582689e-06, + "loss": 0.1039, + "step": 3848 + }, + { + "epoch": 0.8757679180887372, + "grad_norm": 1.81301741951695, + "learning_rate": 1.1577891928983622e-06, + "loss": 0.2074, + "step": 3849 + }, + { + "epoch": 0.875995449374289, + "grad_norm": 1.651298203571477, + "learning_rate": 1.15774247564922e-06, + "loss": 0.1129, + "step": 3850 + }, + { + "epoch": 0.8762229806598407, + "grad_norm": 1.2180184462913173, + "learning_rate": 1.1576957475117973e-06, + "loss": 0.1528, + "step": 3851 + }, + { + "epoch": 0.8764505119453925, + "grad_norm": 2.1631750813889763, + "learning_rate": 1.1576490084870493e-06, + "loss": 0.1133, + "step": 3852 + }, + { + "epoch": 0.8766780432309442, + "grad_norm": 1.7995773049554316, + "learning_rate": 1.1576022585759308e-06, + "loss": 0.1789, + "step": 3853 + }, + { + "epoch": 0.876905574516496, + "grad_norm": 1.8726002975048064, + "learning_rate": 1.1575554977793975e-06, + "loss": 0.1175, + "step": 3854 + }, + { + "epoch": 0.8771331058020477, + "grad_norm": 1.4099458224848984, + "learning_rate": 1.1575087260984056e-06, + "loss": 0.0726, + "step": 3855 + }, + { + "epoch": 0.8773606370875996, + "grad_norm": 2.647213314256987, + "learning_rate": 1.1574619435339101e-06, + "loss": 0.1726, + "step": 3856 + }, + { + "epoch": 0.8775881683731513, + "grad_norm": 0.9740327969556655, + "learning_rate": 1.157415150086868e-06, + "loss": 0.0632, + "step": 3857 + }, + { + "epoch": 0.8778156996587031, + "grad_norm": 1.4207473326991096, + "learning_rate": 1.1573683457582349e-06, + "loss": 0.1026, + "step": 3858 + }, + { + "epoch": 0.8780432309442548, + "grad_norm": 2.003642185189255, + "learning_rate": 1.157321530548968e-06, + "loss": 0.1118, + "step": 3859 + }, + { + "epoch": 0.8782707622298066, + "grad_norm": 1.706947541711292, + "learning_rate": 1.157274704460024e-06, + "loss": 0.0931, + "step": 3860 + }, + { + "epoch": 0.8784982935153584, + "grad_norm": 2.5323455895431586, + "learning_rate": 1.1572278674923598e-06, + "loss": 0.1391, + "step": 3861 + }, + { + "epoch": 0.8787258248009101, + "grad_norm": 1.8894906354825123, + "learning_rate": 1.1571810196469326e-06, + "loss": 0.1403, + "step": 3862 + }, + { + "epoch": 0.8789533560864619, + "grad_norm": 7.156443582789994, + "learning_rate": 1.1571341609247003e-06, + "loss": 0.2029, + "step": 3863 + }, + { + "epoch": 0.8791808873720136, + "grad_norm": 1.6220204212083171, + "learning_rate": 1.1570872913266202e-06, + "loss": 0.1553, + "step": 3864 + }, + { + "epoch": 0.8794084186575654, + "grad_norm": 1.0193269833748249, + "learning_rate": 1.1570404108536501e-06, + "loss": 0.047, + "step": 3865 + }, + { + "epoch": 0.8796359499431172, + "grad_norm": 2.675021704289578, + "learning_rate": 1.1569935195067487e-06, + "loss": 0.1267, + "step": 3866 + }, + { + "epoch": 0.879863481228669, + "grad_norm": 1.730509204885054, + "learning_rate": 1.1569466172868737e-06, + "loss": 0.0736, + "step": 3867 + }, + { + "epoch": 0.8800910125142207, + "grad_norm": 1.1047910151414873, + "learning_rate": 1.1568997041949843e-06, + "loss": 0.0584, + "step": 3868 + }, + { + "epoch": 0.8803185437997725, + "grad_norm": 1.496404259499306, + "learning_rate": 1.156852780232039e-06, + "loss": 0.1104, + "step": 3869 + }, + { + "epoch": 0.8805460750853242, + "grad_norm": 3.3560356704270617, + "learning_rate": 1.156805845398997e-06, + "loss": 0.1684, + "step": 3870 + }, + { + "epoch": 0.880773606370876, + "grad_norm": 1.561882218368479, + "learning_rate": 1.1567588996968173e-06, + "loss": 0.0996, + "step": 3871 + }, + { + "epoch": 0.8810011376564277, + "grad_norm": 1.9936186582009299, + "learning_rate": 1.1567119431264598e-06, + "loss": 0.1882, + "step": 3872 + }, + { + "epoch": 0.8812286689419795, + "grad_norm": 1.8110134738488533, + "learning_rate": 1.156664975688884e-06, + "loss": 0.101, + "step": 3873 + }, + { + "epoch": 0.8814562002275312, + "grad_norm": 2.229579875203127, + "learning_rate": 1.1566179973850496e-06, + "loss": 0.21, + "step": 3874 + }, + { + "epoch": 0.8816837315130831, + "grad_norm": 1.2621397772431622, + "learning_rate": 1.156571008215917e-06, + "loss": 0.0986, + "step": 3875 + }, + { + "epoch": 0.8819112627986349, + "grad_norm": 0.9768697935292275, + "learning_rate": 1.1565240081824466e-06, + "loss": 0.07, + "step": 3876 + }, + { + "epoch": 0.8821387940841866, + "grad_norm": 3.6517858861658117, + "learning_rate": 1.1564769972855987e-06, + "loss": 0.1368, + "step": 3877 + }, + { + "epoch": 0.8823663253697384, + "grad_norm": 2.247760853958852, + "learning_rate": 1.1564299755263345e-06, + "loss": 0.1956, + "step": 3878 + }, + { + "epoch": 0.8825938566552901, + "grad_norm": 1.5076798581031285, + "learning_rate": 1.1563829429056148e-06, + "loss": 0.1544, + "step": 3879 + }, + { + "epoch": 0.8828213879408419, + "grad_norm": 1.4938750757570904, + "learning_rate": 1.156335899424401e-06, + "loss": 0.0977, + "step": 3880 + }, + { + "epoch": 0.8830489192263936, + "grad_norm": 1.4303887183395276, + "learning_rate": 1.1562888450836544e-06, + "loss": 0.1261, + "step": 3881 + }, + { + "epoch": 0.8832764505119454, + "grad_norm": 2.561231231387681, + "learning_rate": 1.156241779884337e-06, + "loss": 0.1923, + "step": 3882 + }, + { + "epoch": 0.8835039817974971, + "grad_norm": 1.7108276568530942, + "learning_rate": 1.1561947038274104e-06, + "loss": 0.0778, + "step": 3883 + }, + { + "epoch": 0.8837315130830489, + "grad_norm": 2.065780133370591, + "learning_rate": 1.156147616913837e-06, + "loss": 0.1611, + "step": 3884 + }, + { + "epoch": 0.8839590443686007, + "grad_norm": 1.470042341002447, + "learning_rate": 1.156100519144579e-06, + "loss": 0.1056, + "step": 3885 + }, + { + "epoch": 0.8841865756541525, + "grad_norm": 1.2016930431615216, + "learning_rate": 1.156053410520599e-06, + "loss": 0.0606, + "step": 3886 + }, + { + "epoch": 0.8844141069397042, + "grad_norm": 2.282484813653413, + "learning_rate": 1.15600629104286e-06, + "loss": 0.1176, + "step": 3887 + }, + { + "epoch": 0.884641638225256, + "grad_norm": 2.299338905769147, + "learning_rate": 1.1559591607123248e-06, + "loss": 0.1981, + "step": 3888 + }, + { + "epoch": 0.8848691695108077, + "grad_norm": 2.099693518455378, + "learning_rate": 1.1559120195299566e-06, + "loss": 0.202, + "step": 3889 + }, + { + "epoch": 0.8850967007963595, + "grad_norm": 2.0753613579757944, + "learning_rate": 1.1558648674967191e-06, + "loss": 0.1623, + "step": 3890 + }, + { + "epoch": 0.8853242320819112, + "grad_norm": 1.4830462940755922, + "learning_rate": 1.1558177046135761e-06, + "loss": 0.0907, + "step": 3891 + }, + { + "epoch": 0.885551763367463, + "grad_norm": 2.9005721227848795, + "learning_rate": 1.1557705308814914e-06, + "loss": 0.1057, + "step": 3892 + }, + { + "epoch": 0.8857792946530147, + "grad_norm": 1.8592154177143236, + "learning_rate": 1.155723346301429e-06, + "loss": 0.1055, + "step": 3893 + }, + { + "epoch": 0.8860068259385665, + "grad_norm": 1.7039048379439774, + "learning_rate": 1.1556761508743532e-06, + "loss": 0.142, + "step": 3894 + }, + { + "epoch": 0.8862343572241184, + "grad_norm": 2.956045848513432, + "learning_rate": 1.1556289446012292e-06, + "loss": 0.1218, + "step": 3895 + }, + { + "epoch": 0.8864618885096701, + "grad_norm": 3.1858139191729675, + "learning_rate": 1.155581727483021e-06, + "loss": 0.1639, + "step": 3896 + }, + { + "epoch": 0.8866894197952219, + "grad_norm": 1.5140961525493326, + "learning_rate": 1.1555344995206941e-06, + "loss": 0.1113, + "step": 3897 + }, + { + "epoch": 0.8869169510807736, + "grad_norm": 3.442091854151577, + "learning_rate": 1.1554872607152138e-06, + "loss": 0.0723, + "step": 3898 + }, + { + "epoch": 0.8871444823663254, + "grad_norm": 1.6965582449202878, + "learning_rate": 1.1554400110675453e-06, + "loss": 0.1209, + "step": 3899 + }, + { + "epoch": 0.8873720136518771, + "grad_norm": 1.3831082327687239, + "learning_rate": 1.1553927505786543e-06, + "loss": 0.0651, + "step": 3900 + }, + { + "epoch": 0.8875995449374289, + "grad_norm": 2.663260132328584, + "learning_rate": 1.1553454792495072e-06, + "loss": 0.1917, + "step": 3901 + }, + { + "epoch": 0.8878270762229806, + "grad_norm": 2.0976954738551927, + "learning_rate": 1.1552981970810694e-06, + "loss": 0.1899, + "step": 3902 + }, + { + "epoch": 0.8880546075085324, + "grad_norm": 1.5870071722314034, + "learning_rate": 1.1552509040743078e-06, + "loss": 0.1131, + "step": 3903 + }, + { + "epoch": 0.8882821387940842, + "grad_norm": 2.128086406305553, + "learning_rate": 1.1552036002301891e-06, + "loss": 0.1998, + "step": 3904 + }, + { + "epoch": 0.888509670079636, + "grad_norm": 1.883247951676848, + "learning_rate": 1.1551562855496796e-06, + "loss": 0.1259, + "step": 3905 + }, + { + "epoch": 0.8887372013651877, + "grad_norm": 2.1360180179379187, + "learning_rate": 1.1551089600337465e-06, + "loss": 0.1888, + "step": 3906 + }, + { + "epoch": 0.8889647326507395, + "grad_norm": 2.217339619643206, + "learning_rate": 1.1550616236833574e-06, + "loss": 0.1299, + "step": 3907 + }, + { + "epoch": 0.8891922639362912, + "grad_norm": 2.282198282828942, + "learning_rate": 1.155014276499479e-06, + "loss": 0.1434, + "step": 3908 + }, + { + "epoch": 0.889419795221843, + "grad_norm": 1.6341018967015024, + "learning_rate": 1.1549669184830796e-06, + "loss": 0.0931, + "step": 3909 + }, + { + "epoch": 0.8896473265073948, + "grad_norm": 1.7540331216445133, + "learning_rate": 1.1549195496351271e-06, + "loss": 0.2306, + "step": 3910 + }, + { + "epoch": 0.8898748577929465, + "grad_norm": 2.3260297313053115, + "learning_rate": 1.1548721699565896e-06, + "loss": 0.1459, + "step": 3911 + }, + { + "epoch": 0.8901023890784983, + "grad_norm": 2.9134592795255685, + "learning_rate": 1.1548247794484353e-06, + "loss": 0.2359, + "step": 3912 + }, + { + "epoch": 0.89032992036405, + "grad_norm": 2.014218612686331, + "learning_rate": 1.1547773781116326e-06, + "loss": 0.1021, + "step": 3913 + }, + { + "epoch": 0.8905574516496019, + "grad_norm": 1.9756799447867912, + "learning_rate": 1.1547299659471509e-06, + "loss": 0.1223, + "step": 3914 + }, + { + "epoch": 0.8907849829351536, + "grad_norm": 1.1758644708180022, + "learning_rate": 1.1546825429559585e-06, + "loss": 0.1228, + "step": 3915 + }, + { + "epoch": 0.8910125142207054, + "grad_norm": 2.545752499921782, + "learning_rate": 1.1546351091390253e-06, + "loss": 0.1292, + "step": 3916 + }, + { + "epoch": 0.8912400455062571, + "grad_norm": 1.13014018897145, + "learning_rate": 1.1545876644973202e-06, + "loss": 0.0827, + "step": 3917 + }, + { + "epoch": 0.8914675767918089, + "grad_norm": 1.3952469100874085, + "learning_rate": 1.1545402090318133e-06, + "loss": 0.1256, + "step": 3918 + }, + { + "epoch": 0.8916951080773606, + "grad_norm": 1.4005307819920794, + "learning_rate": 1.1544927427434743e-06, + "loss": 0.1093, + "step": 3919 + }, + { + "epoch": 0.8919226393629124, + "grad_norm": 2.9964531845500453, + "learning_rate": 1.1544452656332733e-06, + "loss": 0.1892, + "step": 3920 + }, + { + "epoch": 0.8921501706484641, + "grad_norm": 1.8000119548590816, + "learning_rate": 1.1543977777021808e-06, + "loss": 0.1624, + "step": 3921 + }, + { + "epoch": 0.8923777019340159, + "grad_norm": 2.004857047650623, + "learning_rate": 1.1543502789511671e-06, + "loss": 0.1227, + "step": 3922 + }, + { + "epoch": 0.8926052332195676, + "grad_norm": 1.5469848933341204, + "learning_rate": 1.1543027693812033e-06, + "loss": 0.0991, + "step": 3923 + }, + { + "epoch": 0.8928327645051195, + "grad_norm": 1.7189673141936324, + "learning_rate": 1.15425524899326e-06, + "loss": 0.1411, + "step": 3924 + }, + { + "epoch": 0.8930602957906713, + "grad_norm": 2.516228330966882, + "learning_rate": 1.154207717788309e-06, + "loss": 0.1137, + "step": 3925 + }, + { + "epoch": 0.893287827076223, + "grad_norm": 2.8926955599289315, + "learning_rate": 1.1541601757673216e-06, + "loss": 0.1127, + "step": 3926 + }, + { + "epoch": 0.8935153583617748, + "grad_norm": 2.429413658560354, + "learning_rate": 1.154112622931269e-06, + "loss": 0.1738, + "step": 3927 + }, + { + "epoch": 0.8937428896473265, + "grad_norm": 1.9238214340170672, + "learning_rate": 1.1540650592811233e-06, + "loss": 0.1376, + "step": 3928 + }, + { + "epoch": 0.8939704209328783, + "grad_norm": 2.4395643175891055, + "learning_rate": 1.154017484817857e-06, + "loss": 0.1679, + "step": 3929 + }, + { + "epoch": 0.89419795221843, + "grad_norm": 1.9381655876966044, + "learning_rate": 1.1539698995424423e-06, + "loss": 0.2045, + "step": 3930 + }, + { + "epoch": 0.8944254835039818, + "grad_norm": 1.436692153329724, + "learning_rate": 1.1539223034558513e-06, + "loss": 0.1339, + "step": 3931 + }, + { + "epoch": 0.8946530147895335, + "grad_norm": 1.706342058684486, + "learning_rate": 1.1538746965590572e-06, + "loss": 0.1282, + "step": 3932 + }, + { + "epoch": 0.8948805460750853, + "grad_norm": 1.6727216071468962, + "learning_rate": 1.153827078853033e-06, + "loss": 0.1372, + "step": 3933 + }, + { + "epoch": 0.8951080773606371, + "grad_norm": 1.3595940949373464, + "learning_rate": 1.1537794503387516e-06, + "loss": 0.1, + "step": 3934 + }, + { + "epoch": 0.8953356086461889, + "grad_norm": 2.5018232688405684, + "learning_rate": 1.1537318110171867e-06, + "loss": 0.1883, + "step": 3935 + }, + { + "epoch": 0.8955631399317406, + "grad_norm": 1.843425187540069, + "learning_rate": 1.153684160889312e-06, + "loss": 0.1922, + "step": 3936 + }, + { + "epoch": 0.8957906712172924, + "grad_norm": 2.4952621408315387, + "learning_rate": 1.1536364999561011e-06, + "loss": 0.1519, + "step": 3937 + }, + { + "epoch": 0.8960182025028441, + "grad_norm": 2.32279003930844, + "learning_rate": 1.1535888282185283e-06, + "loss": 0.2344, + "step": 3938 + }, + { + "epoch": 0.8962457337883959, + "grad_norm": 2.0547686374255054, + "learning_rate": 1.1535411456775682e-06, + "loss": 0.1401, + "step": 3939 + }, + { + "epoch": 0.8964732650739476, + "grad_norm": 1.798639150183759, + "learning_rate": 1.1534934523341952e-06, + "loss": 0.0823, + "step": 3940 + }, + { + "epoch": 0.8967007963594994, + "grad_norm": 2.1804994768684014, + "learning_rate": 1.1534457481893834e-06, + "loss": 0.1315, + "step": 3941 + }, + { + "epoch": 0.8969283276450511, + "grad_norm": 1.2797290557497885, + "learning_rate": 1.1533980332441085e-06, + "loss": 0.0734, + "step": 3942 + }, + { + "epoch": 0.897155858930603, + "grad_norm": 2.829413645850169, + "learning_rate": 1.1533503074993455e-06, + "loss": 0.1007, + "step": 3943 + }, + { + "epoch": 0.8973833902161548, + "grad_norm": 2.0371130013204763, + "learning_rate": 1.15330257095607e-06, + "loss": 0.1432, + "step": 3944 + }, + { + "epoch": 0.8976109215017065, + "grad_norm": 1.9028884616413433, + "learning_rate": 1.1532548236152574e-06, + "loss": 0.0673, + "step": 3945 + }, + { + "epoch": 0.8978384527872583, + "grad_norm": 1.7454338393426052, + "learning_rate": 1.1532070654778838e-06, + "loss": 0.1829, + "step": 3946 + }, + { + "epoch": 0.89806598407281, + "grad_norm": 2.3355341925065356, + "learning_rate": 1.1531592965449249e-06, + "loss": 0.1586, + "step": 3947 + }, + { + "epoch": 0.8982935153583618, + "grad_norm": 2.763480949546953, + "learning_rate": 1.1531115168173574e-06, + "loss": 0.1884, + "step": 3948 + }, + { + "epoch": 0.8985210466439135, + "grad_norm": 1.4510805507342073, + "learning_rate": 1.1530637262961574e-06, + "loss": 0.1425, + "step": 3949 + }, + { + "epoch": 0.8987485779294653, + "grad_norm": 1.968506252851868, + "learning_rate": 1.1530159249823022e-06, + "loss": 0.1893, + "step": 3950 + }, + { + "epoch": 0.898976109215017, + "grad_norm": 2.324903096410494, + "learning_rate": 1.1529681128767686e-06, + "loss": 0.1095, + "step": 3951 + }, + { + "epoch": 0.8992036405005688, + "grad_norm": 2.545094901563708, + "learning_rate": 1.1529202899805336e-06, + "loss": 0.1262, + "step": 3952 + }, + { + "epoch": 0.8994311717861206, + "grad_norm": 2.15492457068193, + "learning_rate": 1.1528724562945748e-06, + "loss": 0.185, + "step": 3953 + }, + { + "epoch": 0.8996587030716724, + "grad_norm": 2.492930770654846, + "learning_rate": 1.1528246118198697e-06, + "loss": 0.104, + "step": 3954 + }, + { + "epoch": 0.8998862343572241, + "grad_norm": 2.3443014851416697, + "learning_rate": 1.1527767565573961e-06, + "loss": 0.1056, + "step": 3955 + }, + { + "epoch": 0.9001137656427759, + "grad_norm": 2.3122789121507803, + "learning_rate": 1.152728890508132e-06, + "loss": 0.1805, + "step": 3956 + }, + { + "epoch": 0.9003412969283277, + "grad_norm": 1.655678977924301, + "learning_rate": 1.1526810136730562e-06, + "loss": 0.1392, + "step": 3957 + }, + { + "epoch": 0.9005688282138794, + "grad_norm": 2.107224890787166, + "learning_rate": 1.1526331260531467e-06, + "loss": 0.1126, + "step": 3958 + }, + { + "epoch": 0.9007963594994312, + "grad_norm": 1.8474653043517228, + "learning_rate": 1.1525852276493825e-06, + "loss": 0.1033, + "step": 3959 + }, + { + "epoch": 0.9010238907849829, + "grad_norm": 2.1340619481623704, + "learning_rate": 1.1525373184627426e-06, + "loss": 0.1245, + "step": 3960 + }, + { + "epoch": 0.9012514220705347, + "grad_norm": 1.1501325031888365, + "learning_rate": 1.1524893984942059e-06, + "loss": 0.0947, + "step": 3961 + }, + { + "epoch": 0.9014789533560864, + "grad_norm": 1.997782703140848, + "learning_rate": 1.152441467744752e-06, + "loss": 0.1392, + "step": 3962 + }, + { + "epoch": 0.9017064846416383, + "grad_norm": 1.8412903765538933, + "learning_rate": 1.1523935262153604e-06, + "loss": 0.0778, + "step": 3963 + }, + { + "epoch": 0.90193401592719, + "grad_norm": 3.1212109332297477, + "learning_rate": 1.152345573907011e-06, + "loss": 0.1328, + "step": 3964 + }, + { + "epoch": 0.9021615472127418, + "grad_norm": 1.7203762023569207, + "learning_rate": 1.1522976108206838e-06, + "loss": 0.114, + "step": 3965 + }, + { + "epoch": 0.9023890784982935, + "grad_norm": 3.173805185587451, + "learning_rate": 1.1522496369573592e-06, + "loss": 0.1798, + "step": 3966 + }, + { + "epoch": 0.9026166097838453, + "grad_norm": 2.473918490691335, + "learning_rate": 1.1522016523180177e-06, + "loss": 0.1495, + "step": 3967 + }, + { + "epoch": 0.902844141069397, + "grad_norm": 2.038623249613075, + "learning_rate": 1.15215365690364e-06, + "loss": 0.1534, + "step": 3968 + }, + { + "epoch": 0.9030716723549488, + "grad_norm": 2.24172548077322, + "learning_rate": 1.1521056507152068e-06, + "loss": 0.1472, + "step": 3969 + }, + { + "epoch": 0.9032992036405005, + "grad_norm": 1.886243794797654, + "learning_rate": 1.1520576337536995e-06, + "loss": 0.1659, + "step": 3970 + }, + { + "epoch": 0.9035267349260523, + "grad_norm": 1.64432508648709, + "learning_rate": 1.1520096060200995e-06, + "loss": 0.1822, + "step": 3971 + }, + { + "epoch": 0.903754266211604, + "grad_norm": 1.1043959428327752, + "learning_rate": 1.1519615675153884e-06, + "loss": 0.1708, + "step": 3972 + }, + { + "epoch": 0.9039817974971559, + "grad_norm": 3.077639264235174, + "learning_rate": 1.1519135182405477e-06, + "loss": 0.1232, + "step": 3973 + }, + { + "epoch": 0.9042093287827077, + "grad_norm": 2.3583973301893395, + "learning_rate": 1.1518654581965597e-06, + "loss": 0.1644, + "step": 3974 + }, + { + "epoch": 0.9044368600682594, + "grad_norm": 1.931055818227307, + "learning_rate": 1.1518173873844068e-06, + "loss": 0.0957, + "step": 3975 + }, + { + "epoch": 0.9046643913538112, + "grad_norm": 2.302484759884308, + "learning_rate": 1.1517693058050714e-06, + "loss": 0.1308, + "step": 3976 + }, + { + "epoch": 0.9048919226393629, + "grad_norm": 2.0541876018634118, + "learning_rate": 1.151721213459536e-06, + "loss": 0.1398, + "step": 3977 + }, + { + "epoch": 0.9051194539249147, + "grad_norm": 1.8604819838615643, + "learning_rate": 1.1516731103487836e-06, + "loss": 0.0897, + "step": 3978 + }, + { + "epoch": 0.9053469852104664, + "grad_norm": 2.3469141053218965, + "learning_rate": 1.1516249964737974e-06, + "loss": 0.1314, + "step": 3979 + }, + { + "epoch": 0.9055745164960182, + "grad_norm": 1.3279642669148322, + "learning_rate": 1.1515768718355607e-06, + "loss": 0.1102, + "step": 3980 + }, + { + "epoch": 0.9058020477815699, + "grad_norm": 1.8053577348674945, + "learning_rate": 1.1515287364350573e-06, + "loss": 0.1213, + "step": 3981 + }, + { + "epoch": 0.9060295790671218, + "grad_norm": 2.871234378558095, + "learning_rate": 1.1514805902732706e-06, + "loss": 0.1212, + "step": 3982 + }, + { + "epoch": 0.9062571103526735, + "grad_norm": 2.156112604782932, + "learning_rate": 1.151432433351185e-06, + "loss": 0.1114, + "step": 3983 + }, + { + "epoch": 0.9064846416382253, + "grad_norm": 2.6297498949498053, + "learning_rate": 1.1513842656697844e-06, + "loss": 0.1919, + "step": 3984 + }, + { + "epoch": 0.906712172923777, + "grad_norm": 1.8042623931688218, + "learning_rate": 1.1513360872300535e-06, + "loss": 0.1075, + "step": 3985 + }, + { + "epoch": 0.9069397042093288, + "grad_norm": 3.2391253841492182, + "learning_rate": 1.1512878980329771e-06, + "loss": 0.1631, + "step": 3986 + }, + { + "epoch": 0.9071672354948805, + "grad_norm": 1.1000651357067122, + "learning_rate": 1.1512396980795399e-06, + "loss": 0.116, + "step": 3987 + }, + { + "epoch": 0.9073947667804323, + "grad_norm": 2.1033192967440746, + "learning_rate": 1.1511914873707269e-06, + "loss": 0.1053, + "step": 3988 + }, + { + "epoch": 0.907622298065984, + "grad_norm": 2.2451810625182556, + "learning_rate": 1.1511432659075234e-06, + "loss": 0.1043, + "step": 3989 + }, + { + "epoch": 0.9078498293515358, + "grad_norm": 1.4658968735063378, + "learning_rate": 1.1510950336909154e-06, + "loss": 0.1242, + "step": 3990 + }, + { + "epoch": 0.9080773606370875, + "grad_norm": 2.05962070131195, + "learning_rate": 1.1510467907218883e-06, + "loss": 0.1355, + "step": 3991 + }, + { + "epoch": 0.9083048919226394, + "grad_norm": 1.4653271062892934, + "learning_rate": 1.1509985370014283e-06, + "loss": 0.0894, + "step": 3992 + }, + { + "epoch": 0.9085324232081912, + "grad_norm": 3.091085962402644, + "learning_rate": 1.1509502725305214e-06, + "loss": 0.136, + "step": 3993 + }, + { + "epoch": 0.9087599544937429, + "grad_norm": 2.8986186181498774, + "learning_rate": 1.1509019973101542e-06, + "loss": 0.1416, + "step": 3994 + }, + { + "epoch": 0.9089874857792947, + "grad_norm": 1.2981522628859017, + "learning_rate": 1.1508537113413134e-06, + "loss": 0.0676, + "step": 3995 + }, + { + "epoch": 0.9092150170648464, + "grad_norm": 1.5433048194105081, + "learning_rate": 1.1508054146249858e-06, + "loss": 0.1052, + "step": 3996 + }, + { + "epoch": 0.9094425483503982, + "grad_norm": 1.4606601789320433, + "learning_rate": 1.1507571071621585e-06, + "loss": 0.0704, + "step": 3997 + }, + { + "epoch": 0.9096700796359499, + "grad_norm": 2.1790916444676864, + "learning_rate": 1.1507087889538186e-06, + "loss": 0.145, + "step": 3998 + }, + { + "epoch": 0.9098976109215017, + "grad_norm": 1.80575976747422, + "learning_rate": 1.1506604600009542e-06, + "loss": 0.1376, + "step": 3999 + }, + { + "epoch": 0.9101251422070534, + "grad_norm": 1.4156080512130724, + "learning_rate": 1.1506121203045524e-06, + "loss": 0.1384, + "step": 4000 + }, + { + "epoch": 0.9103526734926052, + "grad_norm": 1.384363765889395, + "learning_rate": 1.1505637698656014e-06, + "loss": 0.1224, + "step": 4001 + }, + { + "epoch": 0.910580204778157, + "grad_norm": 2.331610054158199, + "learning_rate": 1.1505154086850898e-06, + "loss": 0.1111, + "step": 4002 + }, + { + "epoch": 0.9108077360637088, + "grad_norm": 1.6098503805284052, + "learning_rate": 1.1504670367640054e-06, + "loss": 0.087, + "step": 4003 + }, + { + "epoch": 0.9110352673492605, + "grad_norm": 1.3999525978643317, + "learning_rate": 1.1504186541033371e-06, + "loss": 0.1418, + "step": 4004 + }, + { + "epoch": 0.9112627986348123, + "grad_norm": 2.2558646882092823, + "learning_rate": 1.150370260704074e-06, + "loss": 0.1381, + "step": 4005 + }, + { + "epoch": 0.911490329920364, + "grad_norm": 1.7065927040440496, + "learning_rate": 1.1503218565672047e-06, + "loss": 0.0844, + "step": 4006 + }, + { + "epoch": 0.9117178612059158, + "grad_norm": 2.1447300443618253, + "learning_rate": 1.1502734416937188e-06, + "loss": 0.2181, + "step": 4007 + }, + { + "epoch": 0.9119453924914676, + "grad_norm": 1.680142723606261, + "learning_rate": 1.150225016084606e-06, + "loss": 0.1837, + "step": 4008 + }, + { + "epoch": 0.9121729237770193, + "grad_norm": 2.283553663622, + "learning_rate": 1.1501765797408558e-06, + "loss": 0.1612, + "step": 4009 + }, + { + "epoch": 0.9124004550625711, + "grad_norm": 2.1243190108046868, + "learning_rate": 1.1501281326634578e-06, + "loss": 0.1769, + "step": 4010 + }, + { + "epoch": 0.9126279863481229, + "grad_norm": 1.5630837662985877, + "learning_rate": 1.1500796748534026e-06, + "loss": 0.1129, + "step": 4011 + }, + { + "epoch": 0.9128555176336747, + "grad_norm": 1.1110498761701864, + "learning_rate": 1.1500312063116803e-06, + "loss": 0.0506, + "step": 4012 + }, + { + "epoch": 0.9130830489192264, + "grad_norm": 2.454314887166611, + "learning_rate": 1.149982727039282e-06, + "loss": 0.2378, + "step": 4013 + }, + { + "epoch": 0.9133105802047782, + "grad_norm": 1.6145277367237205, + "learning_rate": 1.149934237037198e-06, + "loss": 0.1419, + "step": 4014 + }, + { + "epoch": 0.9135381114903299, + "grad_norm": 2.4477607721641332, + "learning_rate": 1.1498857363064198e-06, + "loss": 0.1204, + "step": 4015 + }, + { + "epoch": 0.9137656427758817, + "grad_norm": 1.5769253132902517, + "learning_rate": 1.1498372248479383e-06, + "loss": 0.212, + "step": 4016 + }, + { + "epoch": 0.9139931740614334, + "grad_norm": 1.6710945747403774, + "learning_rate": 1.1497887026627451e-06, + "loss": 0.1535, + "step": 4017 + }, + { + "epoch": 0.9142207053469852, + "grad_norm": 2.8361936305649467, + "learning_rate": 1.1497401697518318e-06, + "loss": 0.1207, + "step": 4018 + }, + { + "epoch": 0.9144482366325369, + "grad_norm": 1.6904198391156955, + "learning_rate": 1.1496916261161908e-06, + "loss": 0.1918, + "step": 4019 + }, + { + "epoch": 0.9146757679180887, + "grad_norm": 1.2927293111341114, + "learning_rate": 1.1496430717568136e-06, + "loss": 0.0644, + "step": 4020 + }, + { + "epoch": 0.9149032992036406, + "grad_norm": 1.1501478000489853, + "learning_rate": 1.149594506674693e-06, + "loss": 0.087, + "step": 4021 + }, + { + "epoch": 0.9151308304891923, + "grad_norm": 1.2191834115791784, + "learning_rate": 1.1495459308708212e-06, + "loss": 0.0694, + "step": 4022 + }, + { + "epoch": 0.9153583617747441, + "grad_norm": 1.41962768386301, + "learning_rate": 1.1494973443461915e-06, + "loss": 0.0914, + "step": 4023 + }, + { + "epoch": 0.9155858930602958, + "grad_norm": 0.8951074706326217, + "learning_rate": 1.1494487471017965e-06, + "loss": 0.0556, + "step": 4024 + }, + { + "epoch": 0.9158134243458476, + "grad_norm": 1.9498245721877319, + "learning_rate": 1.1494001391386298e-06, + "loss": 0.1551, + "step": 4025 + }, + { + "epoch": 0.9160409556313993, + "grad_norm": 2.530813661494047, + "learning_rate": 1.1493515204576844e-06, + "loss": 0.1453, + "step": 4026 + }, + { + "epoch": 0.9162684869169511, + "grad_norm": 2.734946404831388, + "learning_rate": 1.1493028910599544e-06, + "loss": 0.2283, + "step": 4027 + }, + { + "epoch": 0.9164960182025028, + "grad_norm": 2.0385212703852127, + "learning_rate": 1.1492542509464333e-06, + "loss": 0.1215, + "step": 4028 + }, + { + "epoch": 0.9167235494880546, + "grad_norm": 2.0433495208554375, + "learning_rate": 1.1492056001181157e-06, + "loss": 0.1181, + "step": 4029 + }, + { + "epoch": 0.9169510807736063, + "grad_norm": 1.4433586254013528, + "learning_rate": 1.1491569385759953e-06, + "loss": 0.0923, + "step": 4030 + }, + { + "epoch": 0.9171786120591582, + "grad_norm": 1.9753580690191006, + "learning_rate": 1.1491082663210675e-06, + "loss": 0.1565, + "step": 4031 + }, + { + "epoch": 0.9174061433447099, + "grad_norm": 2.667070125143639, + "learning_rate": 1.1490595833543263e-06, + "loss": 0.1626, + "step": 4032 + }, + { + "epoch": 0.9176336746302617, + "grad_norm": 2.269844760737215, + "learning_rate": 1.1490108896767672e-06, + "loss": 0.1313, + "step": 4033 + }, + { + "epoch": 0.9178612059158134, + "grad_norm": 2.1613808031038673, + "learning_rate": 1.1489621852893849e-06, + "loss": 0.1564, + "step": 4034 + }, + { + "epoch": 0.9180887372013652, + "grad_norm": 1.9130661258416666, + "learning_rate": 1.1489134701931753e-06, + "loss": 0.1327, + "step": 4035 + }, + { + "epoch": 0.9183162684869169, + "grad_norm": 1.669381111217745, + "learning_rate": 1.1488647443891339e-06, + "loss": 0.1389, + "step": 4036 + }, + { + "epoch": 0.9185437997724687, + "grad_norm": 1.989793112795475, + "learning_rate": 1.1488160078782565e-06, + "loss": 0.1622, + "step": 4037 + }, + { + "epoch": 0.9187713310580204, + "grad_norm": 2.076351484775596, + "learning_rate": 1.148767260661539e-06, + "loss": 0.098, + "step": 4038 + }, + { + "epoch": 0.9189988623435722, + "grad_norm": 3.2089411401806487, + "learning_rate": 1.1487185027399783e-06, + "loss": 0.2277, + "step": 4039 + }, + { + "epoch": 0.919226393629124, + "grad_norm": 2.6063043974899838, + "learning_rate": 1.1486697341145703e-06, + "loss": 0.1371, + "step": 4040 + }, + { + "epoch": 0.9194539249146758, + "grad_norm": 1.7360554006678817, + "learning_rate": 1.148620954786312e-06, + "loss": 0.1237, + "step": 4041 + }, + { + "epoch": 0.9196814562002276, + "grad_norm": 2.1507897047848146, + "learning_rate": 1.1485721647562005e-06, + "loss": 0.2027, + "step": 4042 + }, + { + "epoch": 0.9199089874857793, + "grad_norm": 1.8157430648391701, + "learning_rate": 1.1485233640252328e-06, + "loss": 0.1071, + "step": 4043 + }, + { + "epoch": 0.9201365187713311, + "grad_norm": 2.013896650858182, + "learning_rate": 1.1484745525944063e-06, + "loss": 0.1488, + "step": 4044 + }, + { + "epoch": 0.9203640500568828, + "grad_norm": 2.005039763917703, + "learning_rate": 1.1484257304647187e-06, + "loss": 0.1545, + "step": 4045 + }, + { + "epoch": 0.9205915813424346, + "grad_norm": 3.0226451865233464, + "learning_rate": 1.1483768976371677e-06, + "loss": 0.167, + "step": 4046 + }, + { + "epoch": 0.9208191126279863, + "grad_norm": 2.3586368879355626, + "learning_rate": 1.1483280541127513e-06, + "loss": 0.167, + "step": 4047 + }, + { + "epoch": 0.9210466439135381, + "grad_norm": 2.3505247603935584, + "learning_rate": 1.1482791998924681e-06, + "loss": 0.1048, + "step": 4048 + }, + { + "epoch": 0.9212741751990898, + "grad_norm": 1.8400187101895282, + "learning_rate": 1.1482303349773164e-06, + "loss": 0.1606, + "step": 4049 + }, + { + "epoch": 0.9215017064846417, + "grad_norm": 3.0228280199249102, + "learning_rate": 1.1481814593682946e-06, + "loss": 0.1459, + "step": 4050 + }, + { + "epoch": 0.9217292377701934, + "grad_norm": 1.2188213501070726, + "learning_rate": 1.1481325730664023e-06, + "loss": 0.1301, + "step": 4051 + }, + { + "epoch": 0.9219567690557452, + "grad_norm": 1.8672928736173242, + "learning_rate": 1.148083676072638e-06, + "loss": 0.1316, + "step": 4052 + }, + { + "epoch": 0.922184300341297, + "grad_norm": 1.8403390593858633, + "learning_rate": 1.1480347683880016e-06, + "loss": 0.1136, + "step": 4053 + }, + { + "epoch": 0.9224118316268487, + "grad_norm": 2.132830111753058, + "learning_rate": 1.1479858500134924e-06, + "loss": 0.0883, + "step": 4054 + }, + { + "epoch": 0.9226393629124005, + "grad_norm": 1.2880475838091152, + "learning_rate": 1.14793692095011e-06, + "loss": 0.118, + "step": 4055 + }, + { + "epoch": 0.9228668941979522, + "grad_norm": 1.9377333599879183, + "learning_rate": 1.147887981198855e-06, + "loss": 0.1256, + "step": 4056 + }, + { + "epoch": 0.923094425483504, + "grad_norm": 2.56950028002691, + "learning_rate": 1.147839030760727e-06, + "loss": 0.1674, + "step": 4057 + }, + { + "epoch": 0.9233219567690557, + "grad_norm": 3.3720583293840103, + "learning_rate": 1.1477900696367269e-06, + "loss": 0.1903, + "step": 4058 + }, + { + "epoch": 0.9235494880546075, + "grad_norm": 1.4802557956584177, + "learning_rate": 1.147741097827855e-06, + "loss": 0.0574, + "step": 4059 + }, + { + "epoch": 0.9237770193401593, + "grad_norm": 1.3196731818666811, + "learning_rate": 1.1476921153351126e-06, + "loss": 0.1187, + "step": 4060 + }, + { + "epoch": 0.9240045506257111, + "grad_norm": 2.3841783165877453, + "learning_rate": 1.1476431221595005e-06, + "loss": 0.1433, + "step": 4061 + }, + { + "epoch": 0.9242320819112628, + "grad_norm": 2.159370750846227, + "learning_rate": 1.1475941183020203e-06, + "loss": 0.1423, + "step": 4062 + }, + { + "epoch": 0.9244596131968146, + "grad_norm": 1.9272353361429937, + "learning_rate": 1.1475451037636733e-06, + "loss": 0.111, + "step": 4063 + }, + { + "epoch": 0.9246871444823663, + "grad_norm": 1.5766120669779804, + "learning_rate": 1.1474960785454615e-06, + "loss": 0.1207, + "step": 4064 + }, + { + "epoch": 0.9249146757679181, + "grad_norm": 1.7214929778176364, + "learning_rate": 1.1474470426483868e-06, + "loss": 0.1049, + "step": 4065 + }, + { + "epoch": 0.9251422070534698, + "grad_norm": 1.9077781728883594, + "learning_rate": 1.1473979960734513e-06, + "loss": 0.099, + "step": 4066 + }, + { + "epoch": 0.9253697383390216, + "grad_norm": 1.4279277010662257, + "learning_rate": 1.1473489388216574e-06, + "loss": 0.1325, + "step": 4067 + }, + { + "epoch": 0.9255972696245733, + "grad_norm": 1.2402250408372297, + "learning_rate": 1.1472998708940079e-06, + "loss": 0.1266, + "step": 4068 + }, + { + "epoch": 0.9258248009101251, + "grad_norm": 1.7364107091551009, + "learning_rate": 1.1472507922915056e-06, + "loss": 0.0807, + "step": 4069 + }, + { + "epoch": 0.926052332195677, + "grad_norm": 1.9921546594650124, + "learning_rate": 1.1472017030151536e-06, + "loss": 0.1734, + "step": 4070 + }, + { + "epoch": 0.9262798634812287, + "grad_norm": 1.8248980083404165, + "learning_rate": 1.147152603065955e-06, + "loss": 0.1147, + "step": 4071 + }, + { + "epoch": 0.9265073947667805, + "grad_norm": 2.0270199058560348, + "learning_rate": 1.1471034924449133e-06, + "loss": 0.1316, + "step": 4072 + }, + { + "epoch": 0.9267349260523322, + "grad_norm": 2.317502793742025, + "learning_rate": 1.1470543711530328e-06, + "loss": 0.1377, + "step": 4073 + }, + { + "epoch": 0.926962457337884, + "grad_norm": 2.198374934269782, + "learning_rate": 1.147005239191317e-06, + "loss": 0.2035, + "step": 4074 + }, + { + "epoch": 0.9271899886234357, + "grad_norm": 1.1810521503732803, + "learning_rate": 1.1469560965607699e-06, + "loss": 0.0786, + "step": 4075 + }, + { + "epoch": 0.9274175199089875, + "grad_norm": 1.4292953433454851, + "learning_rate": 1.1469069432623965e-06, + "loss": 0.0914, + "step": 4076 + }, + { + "epoch": 0.9276450511945392, + "grad_norm": 2.7112408608897365, + "learning_rate": 1.1468577792972004e-06, + "loss": 0.1557, + "step": 4077 + }, + { + "epoch": 0.927872582480091, + "grad_norm": 1.2029939421333486, + "learning_rate": 1.1468086046661874e-06, + "loss": 0.1556, + "step": 4078 + }, + { + "epoch": 0.9281001137656428, + "grad_norm": 2.0892087550113234, + "learning_rate": 1.146759419370362e-06, + "loss": 0.2167, + "step": 4079 + }, + { + "epoch": 0.9283276450511946, + "grad_norm": 1.7305291008510646, + "learning_rate": 1.14671022341073e-06, + "loss": 0.0964, + "step": 4080 + }, + { + "epoch": 0.9285551763367463, + "grad_norm": 1.6688027122772393, + "learning_rate": 1.1466610167882963e-06, + "loss": 0.1447, + "step": 4081 + }, + { + "epoch": 0.9287827076222981, + "grad_norm": 3.0784918422868373, + "learning_rate": 1.1466117995040666e-06, + "loss": 0.1506, + "step": 4082 + }, + { + "epoch": 0.9290102389078498, + "grad_norm": 2.124545290642094, + "learning_rate": 1.1465625715590473e-06, + "loss": 0.1008, + "step": 4083 + }, + { + "epoch": 0.9292377701934016, + "grad_norm": 1.2578652782716944, + "learning_rate": 1.146513332954244e-06, + "loss": 0.1283, + "step": 4084 + }, + { + "epoch": 0.9294653014789533, + "grad_norm": 2.4153588158676182, + "learning_rate": 1.1464640836906635e-06, + "loss": 0.1172, + "step": 4085 + }, + { + "epoch": 0.9296928327645051, + "grad_norm": 2.3544484292723804, + "learning_rate": 1.146414823769312e-06, + "loss": 0.1101, + "step": 4086 + }, + { + "epoch": 0.9299203640500568, + "grad_norm": 2.134969325323832, + "learning_rate": 1.1463655531911963e-06, + "loss": 0.0966, + "step": 4087 + }, + { + "epoch": 0.9301478953356086, + "grad_norm": 4.133663465987289, + "learning_rate": 1.1463162719573236e-06, + "loss": 0.1581, + "step": 4088 + }, + { + "epoch": 0.9303754266211605, + "grad_norm": 1.8487089633836562, + "learning_rate": 1.1462669800687012e-06, + "loss": 0.1634, + "step": 4089 + }, + { + "epoch": 0.9306029579067122, + "grad_norm": 1.7975377039405993, + "learning_rate": 1.1462176775263365e-06, + "loss": 0.0946, + "step": 4090 + }, + { + "epoch": 0.930830489192264, + "grad_norm": 2.373976736841564, + "learning_rate": 1.1461683643312366e-06, + "loss": 0.1869, + "step": 4091 + }, + { + "epoch": 0.9310580204778157, + "grad_norm": 1.5868652532544276, + "learning_rate": 1.1461190404844103e-06, + "loss": 0.1333, + "step": 4092 + }, + { + "epoch": 0.9312855517633675, + "grad_norm": 1.6337419551444752, + "learning_rate": 1.1460697059868648e-06, + "loss": 0.1835, + "step": 4093 + }, + { + "epoch": 0.9315130830489192, + "grad_norm": 2.125031365520933, + "learning_rate": 1.146020360839609e-06, + "loss": 0.1, + "step": 4094 + }, + { + "epoch": 0.931740614334471, + "grad_norm": 1.3164350885665195, + "learning_rate": 1.1459710050436513e-06, + "loss": 0.1065, + "step": 4095 + }, + { + "epoch": 0.9319681456200227, + "grad_norm": 1.9989494690793859, + "learning_rate": 1.1459216385999999e-06, + "loss": 0.1275, + "step": 4096 + }, + { + "epoch": 0.9321956769055745, + "grad_norm": 2.185423641353373, + "learning_rate": 1.1458722615096648e-06, + "loss": 0.0737, + "step": 4097 + }, + { + "epoch": 0.9324232081911262, + "grad_norm": 2.4202341523485007, + "learning_rate": 1.1458228737736542e-06, + "loss": 0.1073, + "step": 4098 + }, + { + "epoch": 0.9326507394766781, + "grad_norm": 1.66652378551551, + "learning_rate": 1.145773475392978e-06, + "loss": 0.1211, + "step": 4099 + }, + { + "epoch": 0.9328782707622298, + "grad_norm": 1.8441447680982124, + "learning_rate": 1.145724066368646e-06, + "loss": 0.1297, + "step": 4100 + }, + { + "epoch": 0.9331058020477816, + "grad_norm": 2.2269635218641426, + "learning_rate": 1.1456746467016675e-06, + "loss": 0.0988, + "step": 4101 + }, + { + "epoch": 0.9333333333333333, + "grad_norm": 2.121237671716207, + "learning_rate": 1.1456252163930528e-06, + "loss": 0.1355, + "step": 4102 + }, + { + "epoch": 0.9335608646188851, + "grad_norm": 2.7856152827774143, + "learning_rate": 1.1455757754438122e-06, + "loss": 0.1413, + "step": 4103 + }, + { + "epoch": 0.9337883959044369, + "grad_norm": 1.3498604957526619, + "learning_rate": 1.1455263238549563e-06, + "loss": 0.1072, + "step": 4104 + }, + { + "epoch": 0.9340159271899886, + "grad_norm": 1.6746673810268238, + "learning_rate": 1.1454768616274955e-06, + "loss": 0.1455, + "step": 4105 + }, + { + "epoch": 0.9342434584755404, + "grad_norm": 1.5273623252161623, + "learning_rate": 1.1454273887624407e-06, + "loss": 0.0891, + "step": 4106 + }, + { + "epoch": 0.9344709897610921, + "grad_norm": 2.0414032969179776, + "learning_rate": 1.1453779052608032e-06, + "loss": 0.0906, + "step": 4107 + }, + { + "epoch": 0.9346985210466439, + "grad_norm": 1.954781728427514, + "learning_rate": 1.1453284111235947e-06, + "loss": 0.2374, + "step": 4108 + }, + { + "epoch": 0.9349260523321957, + "grad_norm": 2.2715359787453213, + "learning_rate": 1.145278906351826e-06, + "loss": 0.1218, + "step": 4109 + }, + { + "epoch": 0.9351535836177475, + "grad_norm": 1.5696301603122902, + "learning_rate": 1.1452293909465095e-06, + "loss": 0.123, + "step": 4110 + }, + { + "epoch": 0.9353811149032992, + "grad_norm": 2.19325395449243, + "learning_rate": 1.145179864908657e-06, + "loss": 0.1683, + "step": 4111 + }, + { + "epoch": 0.935608646188851, + "grad_norm": 2.5751019735499963, + "learning_rate": 1.1451303282392808e-06, + "loss": 0.1256, + "step": 4112 + }, + { + "epoch": 0.9358361774744027, + "grad_norm": 1.8716699205619842, + "learning_rate": 1.145080780939393e-06, + "loss": 0.0974, + "step": 4113 + }, + { + "epoch": 0.9360637087599545, + "grad_norm": 1.2789688026695687, + "learning_rate": 1.145031223010007e-06, + "loss": 0.0736, + "step": 4114 + }, + { + "epoch": 0.9362912400455062, + "grad_norm": 1.5985054428399412, + "learning_rate": 1.1449816544521347e-06, + "loss": 0.1484, + "step": 4115 + }, + { + "epoch": 0.936518771331058, + "grad_norm": 1.6163529234144407, + "learning_rate": 1.1449320752667898e-06, + "loss": 0.1358, + "step": 4116 + }, + { + "epoch": 0.9367463026166097, + "grad_norm": 1.0948746671297045, + "learning_rate": 1.1448824854549856e-06, + "loss": 0.0662, + "step": 4117 + }, + { + "epoch": 0.9369738339021616, + "grad_norm": 2.0429510432349103, + "learning_rate": 1.1448328850177356e-06, + "loss": 0.1023, + "step": 4118 + }, + { + "epoch": 0.9372013651877134, + "grad_norm": 2.909739539143929, + "learning_rate": 1.1447832739560533e-06, + "loss": 0.2704, + "step": 4119 + }, + { + "epoch": 0.9374288964732651, + "grad_norm": 1.8669293896102523, + "learning_rate": 1.1447336522709528e-06, + "loss": 0.1643, + "step": 4120 + }, + { + "epoch": 0.9376564277588169, + "grad_norm": 1.98051617462352, + "learning_rate": 1.1446840199634483e-06, + "loss": 0.1212, + "step": 4121 + }, + { + "epoch": 0.9378839590443686, + "grad_norm": 2.575814064863609, + "learning_rate": 1.1446343770345544e-06, + "loss": 0.1351, + "step": 4122 + }, + { + "epoch": 0.9381114903299204, + "grad_norm": 1.4793880550121636, + "learning_rate": 1.1445847234852853e-06, + "loss": 0.1332, + "step": 4123 + }, + { + "epoch": 0.9383390216154721, + "grad_norm": 2.6986765190100876, + "learning_rate": 1.1445350593166559e-06, + "loss": 0.215, + "step": 4124 + }, + { + "epoch": 0.9385665529010239, + "grad_norm": 13.432551035472212, + "learning_rate": 1.1444853845296816e-06, + "loss": 0.153, + "step": 4125 + }, + { + "epoch": 0.9387940841865756, + "grad_norm": 2.7787794431563353, + "learning_rate": 1.1444356991253774e-06, + "loss": 0.1385, + "step": 4126 + }, + { + "epoch": 0.9390216154721274, + "grad_norm": 2.844290881868777, + "learning_rate": 1.1443860031047589e-06, + "loss": 0.1564, + "step": 4127 + }, + { + "epoch": 0.9392491467576792, + "grad_norm": 1.3610957169917577, + "learning_rate": 1.1443362964688416e-06, + "loss": 0.143, + "step": 4128 + }, + { + "epoch": 0.939476678043231, + "grad_norm": 2.1082592980578174, + "learning_rate": 1.1442865792186413e-06, + "loss": 0.16, + "step": 4129 + }, + { + "epoch": 0.9397042093287827, + "grad_norm": 1.9753673499965854, + "learning_rate": 1.1442368513551746e-06, + "loss": 0.0904, + "step": 4130 + }, + { + "epoch": 0.9399317406143345, + "grad_norm": 1.479785764805608, + "learning_rate": 1.1441871128794576e-06, + "loss": 0.0822, + "step": 4131 + }, + { + "epoch": 0.9401592718998862, + "grad_norm": 1.577368771739526, + "learning_rate": 1.1441373637925068e-06, + "loss": 0.168, + "step": 4132 + }, + { + "epoch": 0.940386803185438, + "grad_norm": 1.1502270382236108, + "learning_rate": 1.1440876040953392e-06, + "loss": 0.1633, + "step": 4133 + }, + { + "epoch": 0.9406143344709897, + "grad_norm": 2.8101793735585034, + "learning_rate": 1.1440378337889713e-06, + "loss": 0.156, + "step": 4134 + }, + { + "epoch": 0.9408418657565415, + "grad_norm": 1.207307845030043, + "learning_rate": 1.143988052874421e-06, + "loss": 0.0777, + "step": 4135 + }, + { + "epoch": 0.9410693970420932, + "grad_norm": 2.773572388788643, + "learning_rate": 1.143938261352705e-06, + "loss": 0.1961, + "step": 4136 + }, + { + "epoch": 0.941296928327645, + "grad_norm": 2.619426570540865, + "learning_rate": 1.1438884592248416e-06, + "loss": 0.0814, + "step": 4137 + }, + { + "epoch": 0.9415244596131969, + "grad_norm": 1.8476195592110707, + "learning_rate": 1.1438386464918483e-06, + "loss": 0.0998, + "step": 4138 + }, + { + "epoch": 0.9417519908987486, + "grad_norm": 2.5320903121338376, + "learning_rate": 1.1437888231547434e-06, + "loss": 0.1897, + "step": 4139 + }, + { + "epoch": 0.9419795221843004, + "grad_norm": 1.8341313571748092, + "learning_rate": 1.143738989214545e-06, + "loss": 0.111, + "step": 4140 + }, + { + "epoch": 0.9422070534698521, + "grad_norm": 2.0680871615443657, + "learning_rate": 1.1436891446722718e-06, + "loss": 0.147, + "step": 4141 + }, + { + "epoch": 0.9424345847554039, + "grad_norm": 2.938990966912184, + "learning_rate": 1.1436392895289423e-06, + "loss": 0.2542, + "step": 4142 + }, + { + "epoch": 0.9426621160409556, + "grad_norm": 2.162441432206442, + "learning_rate": 1.1435894237855754e-06, + "loss": 0.1437, + "step": 4143 + }, + { + "epoch": 0.9428896473265074, + "grad_norm": 2.1600113975804045, + "learning_rate": 1.1435395474431906e-06, + "loss": 0.1397, + "step": 4144 + }, + { + "epoch": 0.9431171786120591, + "grad_norm": 2.136599526696913, + "learning_rate": 1.143489660502807e-06, + "loss": 0.1394, + "step": 4145 + }, + { + "epoch": 0.9433447098976109, + "grad_norm": 1.8658722635830092, + "learning_rate": 1.1434397629654445e-06, + "loss": 0.2155, + "step": 4146 + }, + { + "epoch": 0.9435722411831626, + "grad_norm": 1.4701919591216344, + "learning_rate": 1.1433898548321226e-06, + "loss": 0.0903, + "step": 4147 + }, + { + "epoch": 0.9437997724687145, + "grad_norm": 2.2540515992681525, + "learning_rate": 1.1433399361038614e-06, + "loss": 0.1212, + "step": 4148 + }, + { + "epoch": 0.9440273037542662, + "grad_norm": 1.4986748683203326, + "learning_rate": 1.1432900067816813e-06, + "loss": 0.1015, + "step": 4149 + }, + { + "epoch": 0.944254835039818, + "grad_norm": 1.7538742820647613, + "learning_rate": 1.1432400668666028e-06, + "loss": 0.1078, + "step": 4150 + }, + { + "epoch": 0.9444823663253697, + "grad_norm": 2.7575185503123407, + "learning_rate": 1.1431901163596462e-06, + "loss": 0.2745, + "step": 4151 + }, + { + "epoch": 0.9447098976109215, + "grad_norm": 2.0879360274125998, + "learning_rate": 1.1431401552618327e-06, + "loss": 0.1636, + "step": 4152 + }, + { + "epoch": 0.9449374288964733, + "grad_norm": 1.3069902569614682, + "learning_rate": 1.1430901835741833e-06, + "loss": 0.0736, + "step": 4153 + }, + { + "epoch": 0.945164960182025, + "grad_norm": 2.257465630700085, + "learning_rate": 1.1430402012977195e-06, + "loss": 0.141, + "step": 4154 + }, + { + "epoch": 0.9453924914675768, + "grad_norm": 1.1487328455761492, + "learning_rate": 1.1429902084334627e-06, + "loss": 0.0854, + "step": 4155 + }, + { + "epoch": 0.9456200227531285, + "grad_norm": 1.7462901983970547, + "learning_rate": 1.1429402049824348e-06, + "loss": 0.1554, + "step": 4156 + }, + { + "epoch": 0.9458475540386804, + "grad_norm": 2.1404630984856987, + "learning_rate": 1.1428901909456575e-06, + "loss": 0.1513, + "step": 4157 + }, + { + "epoch": 0.9460750853242321, + "grad_norm": 2.160428099601572, + "learning_rate": 1.1428401663241533e-06, + "loss": 0.1132, + "step": 4158 + }, + { + "epoch": 0.9463026166097839, + "grad_norm": 1.01649187642893, + "learning_rate": 1.1427901311189444e-06, + "loss": 0.0906, + "step": 4159 + }, + { + "epoch": 0.9465301478953356, + "grad_norm": 1.5167730970382043, + "learning_rate": 1.1427400853310536e-06, + "loss": 0.0898, + "step": 4160 + }, + { + "epoch": 0.9467576791808874, + "grad_norm": 1.05791105281377, + "learning_rate": 1.1426900289615034e-06, + "loss": 0.0774, + "step": 4161 + }, + { + "epoch": 0.9469852104664391, + "grad_norm": 1.9510241509956734, + "learning_rate": 1.1426399620113174e-06, + "loss": 0.1024, + "step": 4162 + }, + { + "epoch": 0.9472127417519909, + "grad_norm": 1.818663643287264, + "learning_rate": 1.1425898844815183e-06, + "loss": 0.1049, + "step": 4163 + }, + { + "epoch": 0.9474402730375426, + "grad_norm": 2.0321761096429847, + "learning_rate": 1.1425397963731303e-06, + "loss": 0.1718, + "step": 4164 + }, + { + "epoch": 0.9476678043230944, + "grad_norm": 2.810440066300498, + "learning_rate": 1.1424896976871763e-06, + "loss": 0.1967, + "step": 4165 + }, + { + "epoch": 0.9478953356086461, + "grad_norm": 1.844076725117748, + "learning_rate": 1.1424395884246808e-06, + "loss": 0.1809, + "step": 4166 + }, + { + "epoch": 0.948122866894198, + "grad_norm": 2.62312758907365, + "learning_rate": 1.1423894685866677e-06, + "loss": 0.1068, + "step": 4167 + }, + { + "epoch": 0.9483503981797498, + "grad_norm": 3.383530309668352, + "learning_rate": 1.1423393381741614e-06, + "loss": 0.1795, + "step": 4168 + }, + { + "epoch": 0.9485779294653015, + "grad_norm": 2.8737766321935343, + "learning_rate": 1.1422891971881867e-06, + "loss": 0.226, + "step": 4169 + }, + { + "epoch": 0.9488054607508533, + "grad_norm": 1.492334368814069, + "learning_rate": 1.142239045629768e-06, + "loss": 0.1521, + "step": 4170 + }, + { + "epoch": 0.949032992036405, + "grad_norm": 1.6985507812410747, + "learning_rate": 1.1421888834999306e-06, + "loss": 0.1272, + "step": 4171 + }, + { + "epoch": 0.9492605233219568, + "grad_norm": 2.6594628914223204, + "learning_rate": 1.1421387107996993e-06, + "loss": 0.175, + "step": 4172 + }, + { + "epoch": 0.9494880546075085, + "grad_norm": 1.6600808249427008, + "learning_rate": 1.1420885275301001e-06, + "loss": 0.1736, + "step": 4173 + }, + { + "epoch": 0.9497155858930603, + "grad_norm": 1.5070094807377818, + "learning_rate": 1.1420383336921583e-06, + "loss": 0.1116, + "step": 4174 + }, + { + "epoch": 0.949943117178612, + "grad_norm": 2.976695679711648, + "learning_rate": 1.1419881292869e-06, + "loss": 0.1242, + "step": 4175 + }, + { + "epoch": 0.9501706484641638, + "grad_norm": 3.24134019331233, + "learning_rate": 1.1419379143153511e-06, + "loss": 0.1451, + "step": 4176 + }, + { + "epoch": 0.9503981797497156, + "grad_norm": 2.5026189974890283, + "learning_rate": 1.1418876887785379e-06, + "loss": 0.1261, + "step": 4177 + }, + { + "epoch": 0.9506257110352674, + "grad_norm": 3.516403926919639, + "learning_rate": 1.1418374526774872e-06, + "loss": 0.1354, + "step": 4178 + }, + { + "epoch": 0.9508532423208191, + "grad_norm": 2.1467294760401123, + "learning_rate": 1.1417872060132251e-06, + "loss": 0.1613, + "step": 4179 + }, + { + "epoch": 0.9510807736063709, + "grad_norm": 1.7910764537754473, + "learning_rate": 1.1417369487867793e-06, + "loss": 0.1552, + "step": 4180 + }, + { + "epoch": 0.9513083048919226, + "grad_norm": 1.261796346597939, + "learning_rate": 1.1416866809991763e-06, + "loss": 0.1095, + "step": 4181 + }, + { + "epoch": 0.9515358361774744, + "grad_norm": 1.5378712086295998, + "learning_rate": 1.1416364026514443e-06, + "loss": 0.0908, + "step": 4182 + }, + { + "epoch": 0.9517633674630261, + "grad_norm": 2.3404103222924673, + "learning_rate": 1.1415861137446099e-06, + "loss": 0.1253, + "step": 4183 + }, + { + "epoch": 0.9519908987485779, + "grad_norm": 1.927348026595231, + "learning_rate": 1.1415358142797018e-06, + "loss": 0.1227, + "step": 4184 + }, + { + "epoch": 0.9522184300341296, + "grad_norm": 1.4109091755193108, + "learning_rate": 1.1414855042577474e-06, + "loss": 0.1501, + "step": 4185 + }, + { + "epoch": 0.9524459613196815, + "grad_norm": 1.7942241015528582, + "learning_rate": 1.1414351836797755e-06, + "loss": 0.1365, + "step": 4186 + }, + { + "epoch": 0.9526734926052333, + "grad_norm": 1.3105709011841327, + "learning_rate": 1.1413848525468139e-06, + "loss": 0.1223, + "step": 4187 + }, + { + "epoch": 0.952901023890785, + "grad_norm": 1.750535419659169, + "learning_rate": 1.1413345108598916e-06, + "loss": 0.1227, + "step": 4188 + }, + { + "epoch": 0.9531285551763368, + "grad_norm": 2.565908749432796, + "learning_rate": 1.1412841586200378e-06, + "loss": 0.1156, + "step": 4189 + }, + { + "epoch": 0.9533560864618885, + "grad_norm": 3.0056891740775318, + "learning_rate": 1.1412337958282812e-06, + "loss": 0.1786, + "step": 4190 + }, + { + "epoch": 0.9535836177474403, + "grad_norm": 1.5893153387746444, + "learning_rate": 1.1411834224856514e-06, + "loss": 0.1434, + "step": 4191 + }, + { + "epoch": 0.953811149032992, + "grad_norm": 1.3984147057203036, + "learning_rate": 1.1411330385931776e-06, + "loss": 0.1293, + "step": 4192 + }, + { + "epoch": 0.9540386803185438, + "grad_norm": 1.2592930853587179, + "learning_rate": 1.1410826441518898e-06, + "loss": 0.1555, + "step": 4193 + }, + { + "epoch": 0.9542662116040955, + "grad_norm": 2.687102327685415, + "learning_rate": 1.1410322391628179e-06, + "loss": 0.1676, + "step": 4194 + }, + { + "epoch": 0.9544937428896473, + "grad_norm": 2.0680105884093845, + "learning_rate": 1.140981823626992e-06, + "loss": 0.0851, + "step": 4195 + }, + { + "epoch": 0.9547212741751991, + "grad_norm": 1.1971146998473337, + "learning_rate": 1.1409313975454429e-06, + "loss": 0.1242, + "step": 4196 + }, + { + "epoch": 0.9549488054607509, + "grad_norm": 0.9679023425689602, + "learning_rate": 1.1408809609192007e-06, + "loss": 0.078, + "step": 4197 + }, + { + "epoch": 0.9551763367463026, + "grad_norm": 2.949152525431351, + "learning_rate": 1.1408305137492963e-06, + "loss": 0.3056, + "step": 4198 + }, + { + "epoch": 0.9554038680318544, + "grad_norm": 2.3564769577645817, + "learning_rate": 1.1407800560367612e-06, + "loss": 0.1402, + "step": 4199 + }, + { + "epoch": 0.9556313993174061, + "grad_norm": 2.2384378241835376, + "learning_rate": 1.140729587782626e-06, + "loss": 0.143, + "step": 4200 + }, + { + "epoch": 0.9558589306029579, + "grad_norm": 1.947160996821623, + "learning_rate": 1.1406791089879229e-06, + "loss": 0.1644, + "step": 4201 + }, + { + "epoch": 0.9560864618885097, + "grad_norm": 3.4892873782238296, + "learning_rate": 1.1406286196536832e-06, + "loss": 0.1897, + "step": 4202 + }, + { + "epoch": 0.9563139931740614, + "grad_norm": 1.9038405626772368, + "learning_rate": 1.1405781197809388e-06, + "loss": 0.115, + "step": 4203 + }, + { + "epoch": 0.9565415244596132, + "grad_norm": 1.5068217662014, + "learning_rate": 1.1405276093707218e-06, + "loss": 0.0991, + "step": 4204 + }, + { + "epoch": 0.9567690557451649, + "grad_norm": 2.796879901668464, + "learning_rate": 1.1404770884240645e-06, + "loss": 0.1362, + "step": 4205 + }, + { + "epoch": 0.9569965870307168, + "grad_norm": 2.3007498141980585, + "learning_rate": 1.1404265569419998e-06, + "loss": 0.1599, + "step": 4206 + }, + { + "epoch": 0.9572241183162685, + "grad_norm": 1.799583084349667, + "learning_rate": 1.14037601492556e-06, + "loss": 0.1147, + "step": 4207 + }, + { + "epoch": 0.9574516496018203, + "grad_norm": 1.3651569176318479, + "learning_rate": 1.1403254623757785e-06, + "loss": 0.0413, + "step": 4208 + }, + { + "epoch": 0.957679180887372, + "grad_norm": 3.035407426201199, + "learning_rate": 1.1402748992936881e-06, + "loss": 0.1924, + "step": 4209 + }, + { + "epoch": 0.9579067121729238, + "grad_norm": 1.467754212052448, + "learning_rate": 1.1402243256803228e-06, + "loss": 0.0743, + "step": 4210 + }, + { + "epoch": 0.9581342434584755, + "grad_norm": 1.9336209918045806, + "learning_rate": 1.1401737415367157e-06, + "loss": 0.1342, + "step": 4211 + }, + { + "epoch": 0.9583617747440273, + "grad_norm": 1.680828937378028, + "learning_rate": 1.1401231468639008e-06, + "loss": 0.1099, + "step": 4212 + }, + { + "epoch": 0.958589306029579, + "grad_norm": 1.5684810967016425, + "learning_rate": 1.140072541662912e-06, + "loss": 0.1596, + "step": 4213 + }, + { + "epoch": 0.9588168373151308, + "grad_norm": 1.97644029701773, + "learning_rate": 1.1400219259347842e-06, + "loss": 0.1719, + "step": 4214 + }, + { + "epoch": 0.9590443686006825, + "grad_norm": 1.6091477688636493, + "learning_rate": 1.139971299680551e-06, + "loss": 0.1682, + "step": 4215 + }, + { + "epoch": 0.9592718998862344, + "grad_norm": 1.5649145183454007, + "learning_rate": 1.1399206629012478e-06, + "loss": 0.0748, + "step": 4216 + }, + { + "epoch": 0.9594994311717862, + "grad_norm": 1.168564720112399, + "learning_rate": 1.1398700155979092e-06, + "loss": 0.1224, + "step": 4217 + }, + { + "epoch": 0.9597269624573379, + "grad_norm": 1.5060255960765723, + "learning_rate": 1.1398193577715705e-06, + "loss": 0.1202, + "step": 4218 + }, + { + "epoch": 0.9599544937428897, + "grad_norm": 2.200763885887003, + "learning_rate": 1.1397686894232671e-06, + "loss": 0.1255, + "step": 4219 + }, + { + "epoch": 0.9601820250284414, + "grad_norm": 2.777826300232821, + "learning_rate": 1.1397180105540343e-06, + "loss": 0.1323, + "step": 4220 + }, + { + "epoch": 0.9604095563139932, + "grad_norm": 1.9394526192366455, + "learning_rate": 1.1396673211649078e-06, + "loss": 0.1286, + "step": 4221 + }, + { + "epoch": 0.9606370875995449, + "grad_norm": 2.4079011934215533, + "learning_rate": 1.139616621256924e-06, + "loss": 0.0972, + "step": 4222 + }, + { + "epoch": 0.9608646188850967, + "grad_norm": 1.5717434706985411, + "learning_rate": 1.1395659108311192e-06, + "loss": 0.1546, + "step": 4223 + }, + { + "epoch": 0.9610921501706484, + "grad_norm": 1.891681115479138, + "learning_rate": 1.1395151898885293e-06, + "loss": 0.1113, + "step": 4224 + }, + { + "epoch": 0.9613196814562003, + "grad_norm": 1.569786603078423, + "learning_rate": 1.1394644584301912e-06, + "loss": 0.1161, + "step": 4225 + }, + { + "epoch": 0.961547212741752, + "grad_norm": 3.0048493135744043, + "learning_rate": 1.1394137164571418e-06, + "loss": 0.167, + "step": 4226 + }, + { + "epoch": 0.9617747440273038, + "grad_norm": 2.7038083428526716, + "learning_rate": 1.1393629639704182e-06, + "loss": 0.1764, + "step": 4227 + }, + { + "epoch": 0.9620022753128555, + "grad_norm": 1.6628649624730723, + "learning_rate": 1.1393122009710575e-06, + "loss": 0.1079, + "step": 4228 + }, + { + "epoch": 0.9622298065984073, + "grad_norm": 1.7102127604416588, + "learning_rate": 1.1392614274600975e-06, + "loss": 0.0968, + "step": 4229 + }, + { + "epoch": 0.962457337883959, + "grad_norm": 1.3187473775319558, + "learning_rate": 1.1392106434385754e-06, + "loss": 0.1167, + "step": 4230 + }, + { + "epoch": 0.9626848691695108, + "grad_norm": 1.8678205589578505, + "learning_rate": 1.1391598489075298e-06, + "loss": 0.2621, + "step": 4231 + }, + { + "epoch": 0.9629124004550625, + "grad_norm": 1.7996446274602325, + "learning_rate": 1.1391090438679986e-06, + "loss": 0.1643, + "step": 4232 + }, + { + "epoch": 0.9631399317406143, + "grad_norm": 2.1682833518232405, + "learning_rate": 1.1390582283210199e-06, + "loss": 0.2173, + "step": 4233 + }, + { + "epoch": 0.963367463026166, + "grad_norm": 2.6867113923181387, + "learning_rate": 1.1390074022676325e-06, + "loss": 0.0996, + "step": 4234 + }, + { + "epoch": 0.9635949943117179, + "grad_norm": 2.011700216724661, + "learning_rate": 1.1389565657088752e-06, + "loss": 0.1241, + "step": 4235 + }, + { + "epoch": 0.9638225255972697, + "grad_norm": 1.5044579516502963, + "learning_rate": 1.1389057186457868e-06, + "loss": 0.2565, + "step": 4236 + }, + { + "epoch": 0.9640500568828214, + "grad_norm": 1.9887001444404104, + "learning_rate": 1.1388548610794069e-06, + "loss": 0.1277, + "step": 4237 + }, + { + "epoch": 0.9642775881683732, + "grad_norm": 2.285745992348972, + "learning_rate": 1.1388039930107747e-06, + "loss": 0.1067, + "step": 4238 + }, + { + "epoch": 0.9645051194539249, + "grad_norm": 1.6891307072134218, + "learning_rate": 1.1387531144409297e-06, + "loss": 0.0772, + "step": 4239 + }, + { + "epoch": 0.9647326507394767, + "grad_norm": 1.337769022595117, + "learning_rate": 1.138702225370912e-06, + "loss": 0.0804, + "step": 4240 + }, + { + "epoch": 0.9649601820250284, + "grad_norm": 2.7570690521663934, + "learning_rate": 1.1386513258017617e-06, + "loss": 0.1543, + "step": 4241 + }, + { + "epoch": 0.9651877133105802, + "grad_norm": 2.5030514235096657, + "learning_rate": 1.138600415734519e-06, + "loss": 0.0915, + "step": 4242 + }, + { + "epoch": 0.9654152445961319, + "grad_norm": 1.1507439636656707, + "learning_rate": 1.1385494951702245e-06, + "loss": 0.1138, + "step": 4243 + }, + { + "epoch": 0.9656427758816837, + "grad_norm": 1.4979011428477753, + "learning_rate": 1.1384985641099187e-06, + "loss": 0.1466, + "step": 4244 + }, + { + "epoch": 0.9658703071672355, + "grad_norm": 1.9623811949903942, + "learning_rate": 1.1384476225546426e-06, + "loss": 0.1219, + "step": 4245 + }, + { + "epoch": 0.9660978384527873, + "grad_norm": 1.84621487590758, + "learning_rate": 1.1383966705054377e-06, + "loss": 0.1323, + "step": 4246 + }, + { + "epoch": 0.966325369738339, + "grad_norm": 2.8495103716366983, + "learning_rate": 1.1383457079633448e-06, + "loss": 0.1594, + "step": 4247 + }, + { + "epoch": 0.9665529010238908, + "grad_norm": 1.655415187692044, + "learning_rate": 1.138294734929406e-06, + "loss": 0.141, + "step": 4248 + }, + { + "epoch": 0.9667804323094426, + "grad_norm": 2.375785741891734, + "learning_rate": 1.1382437514046627e-06, + "loss": 0.1236, + "step": 4249 + }, + { + "epoch": 0.9670079635949943, + "grad_norm": 2.001867889330283, + "learning_rate": 1.1381927573901572e-06, + "loss": 0.1337, + "step": 4250 + }, + { + "epoch": 0.967235494880546, + "grad_norm": 0.9396950060200854, + "learning_rate": 1.1381417528869316e-06, + "loss": 0.0702, + "step": 4251 + }, + { + "epoch": 0.9674630261660978, + "grad_norm": 1.462486126163728, + "learning_rate": 1.1380907378960282e-06, + "loss": 0.1029, + "step": 4252 + }, + { + "epoch": 0.9676905574516496, + "grad_norm": 1.6071269347895911, + "learning_rate": 1.13803971241849e-06, + "loss": 0.1696, + "step": 4253 + }, + { + "epoch": 0.9679180887372013, + "grad_norm": 1.628532405088259, + "learning_rate": 1.1379886764553596e-06, + "loss": 0.1051, + "step": 4254 + }, + { + "epoch": 0.9681456200227532, + "grad_norm": 2.923835193493409, + "learning_rate": 1.1379376300076803e-06, + "loss": 0.1836, + "step": 4255 + }, + { + "epoch": 0.9683731513083049, + "grad_norm": 2.5536251649436794, + "learning_rate": 1.137886573076495e-06, + "loss": 0.1119, + "step": 4256 + }, + { + "epoch": 0.9686006825938567, + "grad_norm": 1.4150265596545728, + "learning_rate": 1.1378355056628474e-06, + "loss": 0.1061, + "step": 4257 + }, + { + "epoch": 0.9688282138794084, + "grad_norm": 3.0955486444699916, + "learning_rate": 1.1377844277677815e-06, + "loss": 0.2038, + "step": 4258 + }, + { + "epoch": 0.9690557451649602, + "grad_norm": 1.6087312990432219, + "learning_rate": 1.1377333393923408e-06, + "loss": 0.1059, + "step": 4259 + }, + { + "epoch": 0.9692832764505119, + "grad_norm": 1.1513657828053114, + "learning_rate": 1.1376822405375698e-06, + "loss": 0.0679, + "step": 4260 + }, + { + "epoch": 0.9695108077360637, + "grad_norm": 1.623397897479566, + "learning_rate": 1.1376311312045128e-06, + "loss": 0.0915, + "step": 4261 + }, + { + "epoch": 0.9697383390216154, + "grad_norm": 2.3944554360529886, + "learning_rate": 1.1375800113942144e-06, + "loss": 0.115, + "step": 4262 + }, + { + "epoch": 0.9699658703071672, + "grad_norm": 1.9802197111942301, + "learning_rate": 1.137528881107719e-06, + "loss": 0.1113, + "step": 4263 + }, + { + "epoch": 0.970193401592719, + "grad_norm": 3.0310932910862145, + "learning_rate": 1.137477740346072e-06, + "loss": 0.105, + "step": 4264 + }, + { + "epoch": 0.9704209328782708, + "grad_norm": 2.8664093923483946, + "learning_rate": 1.1374265891103187e-06, + "loss": 0.1464, + "step": 4265 + }, + { + "epoch": 0.9706484641638226, + "grad_norm": 2.241954646362492, + "learning_rate": 1.1373754274015044e-06, + "loss": 0.1171, + "step": 4266 + }, + { + "epoch": 0.9708759954493743, + "grad_norm": 2.0471078332382358, + "learning_rate": 1.1373242552206744e-06, + "loss": 0.1427, + "step": 4267 + }, + { + "epoch": 0.9711035267349261, + "grad_norm": 2.246513733887583, + "learning_rate": 1.1372730725688754e-06, + "loss": 0.1524, + "step": 4268 + }, + { + "epoch": 0.9713310580204778, + "grad_norm": 3.6522298247730585, + "learning_rate": 1.1372218794471527e-06, + "loss": 0.1175, + "step": 4269 + }, + { + "epoch": 0.9715585893060296, + "grad_norm": 2.095106143688212, + "learning_rate": 1.1371706758565529e-06, + "loss": 0.1154, + "step": 4270 + }, + { + "epoch": 0.9717861205915813, + "grad_norm": 1.3788867318393225, + "learning_rate": 1.1371194617981224e-06, + "loss": 0.0995, + "step": 4271 + }, + { + "epoch": 0.9720136518771331, + "grad_norm": 1.8027774578469935, + "learning_rate": 1.137068237272908e-06, + "loss": 0.1189, + "step": 4272 + }, + { + "epoch": 0.9722411831626848, + "grad_norm": 2.092762888855007, + "learning_rate": 1.1370170022819569e-06, + "loss": 0.0788, + "step": 4273 + }, + { + "epoch": 0.9724687144482367, + "grad_norm": 2.4272221864712558, + "learning_rate": 1.1369657568263157e-06, + "loss": 0.1111, + "step": 4274 + }, + { + "epoch": 0.9726962457337884, + "grad_norm": 1.629457983537679, + "learning_rate": 1.1369145009070323e-06, + "loss": 0.1251, + "step": 4275 + }, + { + "epoch": 0.9729237770193402, + "grad_norm": 1.8676345743374245, + "learning_rate": 1.1368632345251538e-06, + "loss": 0.0962, + "step": 4276 + }, + { + "epoch": 0.9731513083048919, + "grad_norm": 2.0581929804225907, + "learning_rate": 1.1368119576817283e-06, + "loss": 0.1381, + "step": 4277 + }, + { + "epoch": 0.9733788395904437, + "grad_norm": 1.3640113222245722, + "learning_rate": 1.1367606703778037e-06, + "loss": 0.0782, + "step": 4278 + }, + { + "epoch": 0.9736063708759954, + "grad_norm": 2.2722595301486783, + "learning_rate": 1.1367093726144283e-06, + "loss": 0.134, + "step": 4279 + }, + { + "epoch": 0.9738339021615472, + "grad_norm": 1.4359397100942914, + "learning_rate": 1.1366580643926506e-06, + "loss": 0.09, + "step": 4280 + }, + { + "epoch": 0.974061433447099, + "grad_norm": 1.6832451491075302, + "learning_rate": 1.1366067457135188e-06, + "loss": 0.1221, + "step": 4281 + }, + { + "epoch": 0.9742889647326507, + "grad_norm": 2.130004538111451, + "learning_rate": 1.1365554165780823e-06, + "loss": 0.126, + "step": 4282 + }, + { + "epoch": 0.9745164960182024, + "grad_norm": 3.1871453915863475, + "learning_rate": 1.13650407698739e-06, + "loss": 0.1402, + "step": 4283 + }, + { + "epoch": 0.9747440273037543, + "grad_norm": 2.183678937207992, + "learning_rate": 1.136452726942491e-06, + "loss": 0.1173, + "step": 4284 + }, + { + "epoch": 0.9749715585893061, + "grad_norm": 1.9710667666023438, + "learning_rate": 1.1364013664444351e-06, + "loss": 0.1596, + "step": 4285 + }, + { + "epoch": 0.9751990898748578, + "grad_norm": 2.163665685442942, + "learning_rate": 1.1363499954942717e-06, + "loss": 0.1537, + "step": 4286 + }, + { + "epoch": 0.9754266211604096, + "grad_norm": 2.534636527671189, + "learning_rate": 1.1362986140930509e-06, + "loss": 0.1157, + "step": 4287 + }, + { + "epoch": 0.9756541524459613, + "grad_norm": 1.636399149395774, + "learning_rate": 1.1362472222418228e-06, + "loss": 0.1538, + "step": 4288 + }, + { + "epoch": 0.9758816837315131, + "grad_norm": 1.826529942423519, + "learning_rate": 1.1361958199416378e-06, + "loss": 0.116, + "step": 4289 + }, + { + "epoch": 0.9761092150170648, + "grad_norm": 3.576678848053755, + "learning_rate": 1.1361444071935467e-06, + "loss": 0.1821, + "step": 4290 + }, + { + "epoch": 0.9763367463026166, + "grad_norm": 1.713382015938669, + "learning_rate": 1.1360929839985998e-06, + "loss": 0.1742, + "step": 4291 + }, + { + "epoch": 0.9765642775881683, + "grad_norm": 1.609348815275912, + "learning_rate": 1.1360415503578485e-06, + "loss": 0.1446, + "step": 4292 + }, + { + "epoch": 0.9767918088737202, + "grad_norm": 1.0924323312589486, + "learning_rate": 1.1359901062723437e-06, + "loss": 0.0912, + "step": 4293 + }, + { + "epoch": 0.9770193401592719, + "grad_norm": 2.3332736667282146, + "learning_rate": 1.1359386517431366e-06, + "loss": 0.127, + "step": 4294 + }, + { + "epoch": 0.9772468714448237, + "grad_norm": 2.2011673248048975, + "learning_rate": 1.1358871867712797e-06, + "loss": 0.2629, + "step": 4295 + }, + { + "epoch": 0.9774744027303754, + "grad_norm": 1.9212910975034734, + "learning_rate": 1.1358357113578242e-06, + "loss": 0.1228, + "step": 4296 + }, + { + "epoch": 0.9777019340159272, + "grad_norm": 2.3260282738895226, + "learning_rate": 1.1357842255038222e-06, + "loss": 0.1067, + "step": 4297 + }, + { + "epoch": 0.977929465301479, + "grad_norm": 1.475524850603775, + "learning_rate": 1.1357327292103266e-06, + "loss": 0.1522, + "step": 4298 + }, + { + "epoch": 0.9781569965870307, + "grad_norm": 1.8612552057655247, + "learning_rate": 1.135681222478389e-06, + "loss": 0.1375, + "step": 4299 + }, + { + "epoch": 0.9783845278725825, + "grad_norm": 1.8283279905566892, + "learning_rate": 1.1356297053090623e-06, + "loss": 0.1235, + "step": 4300 + }, + { + "epoch": 0.9786120591581342, + "grad_norm": 1.8357637664074558, + "learning_rate": 1.1355781777033998e-06, + "loss": 0.138, + "step": 4301 + }, + { + "epoch": 0.978839590443686, + "grad_norm": 1.2631401818522685, + "learning_rate": 1.1355266396624545e-06, + "loss": 0.1161, + "step": 4302 + }, + { + "epoch": 0.9790671217292378, + "grad_norm": 1.2730308216200366, + "learning_rate": 1.1354750911872795e-06, + "loss": 0.119, + "step": 4303 + }, + { + "epoch": 0.9792946530147896, + "grad_norm": 1.9518448337670264, + "learning_rate": 1.1354235322789286e-06, + "loss": 0.1945, + "step": 4304 + }, + { + "epoch": 0.9795221843003413, + "grad_norm": 1.2218564664195801, + "learning_rate": 1.1353719629384554e-06, + "loss": 0.1329, + "step": 4305 + }, + { + "epoch": 0.9797497155858931, + "grad_norm": 1.3504543016276331, + "learning_rate": 1.135320383166914e-06, + "loss": 0.1504, + "step": 4306 + }, + { + "epoch": 0.9799772468714448, + "grad_norm": 3.1228187183100946, + "learning_rate": 1.1352687929653586e-06, + "loss": 0.2716, + "step": 4307 + }, + { + "epoch": 0.9802047781569966, + "grad_norm": 2.4100339134625264, + "learning_rate": 1.1352171923348438e-06, + "loss": 0.1911, + "step": 4308 + }, + { + "epoch": 0.9804323094425483, + "grad_norm": 1.8961258255968092, + "learning_rate": 1.1351655812764236e-06, + "loss": 0.1552, + "step": 4309 + }, + { + "epoch": 0.9806598407281001, + "grad_norm": 1.758739411124323, + "learning_rate": 1.1351139597911536e-06, + "loss": 0.1091, + "step": 4310 + }, + { + "epoch": 0.9808873720136518, + "grad_norm": 2.257284438813299, + "learning_rate": 1.1350623278800884e-06, + "loss": 0.1427, + "step": 4311 + }, + { + "epoch": 0.9811149032992036, + "grad_norm": 1.5193987741131023, + "learning_rate": 1.1350106855442833e-06, + "loss": 0.1043, + "step": 4312 + }, + { + "epoch": 0.9813424345847555, + "grad_norm": 3.0839542265345194, + "learning_rate": 1.134959032784794e-06, + "loss": 0.1315, + "step": 4313 + }, + { + "epoch": 0.9815699658703072, + "grad_norm": 2.6269195704445725, + "learning_rate": 1.1349073696026759e-06, + "loss": 0.132, + "step": 4314 + }, + { + "epoch": 0.981797497155859, + "grad_norm": 2.410755923938861, + "learning_rate": 1.1348556959989848e-06, + "loss": 0.0962, + "step": 4315 + }, + { + "epoch": 0.9820250284414107, + "grad_norm": 1.8676901583023484, + "learning_rate": 1.1348040119747771e-06, + "loss": 0.108, + "step": 4316 + }, + { + "epoch": 0.9822525597269625, + "grad_norm": 1.7889496982318338, + "learning_rate": 1.1347523175311092e-06, + "loss": 0.1453, + "step": 4317 + }, + { + "epoch": 0.9824800910125142, + "grad_norm": 1.4259289185918111, + "learning_rate": 1.1347006126690377e-06, + "loss": 0.1674, + "step": 4318 + }, + { + "epoch": 0.982707622298066, + "grad_norm": 2.7410660853435838, + "learning_rate": 1.1346488973896188e-06, + "loss": 0.1252, + "step": 4319 + }, + { + "epoch": 0.9829351535836177, + "grad_norm": 1.6223968424245374, + "learning_rate": 1.13459717169391e-06, + "loss": 0.2018, + "step": 4320 + }, + { + "epoch": 0.9831626848691695, + "grad_norm": 1.3310882897622611, + "learning_rate": 1.1345454355829682e-06, + "loss": 0.1641, + "step": 4321 + }, + { + "epoch": 0.9833902161547212, + "grad_norm": 1.781413030603326, + "learning_rate": 1.1344936890578508e-06, + "loss": 0.1044, + "step": 4322 + }, + { + "epoch": 0.9836177474402731, + "grad_norm": 1.447371024861841, + "learning_rate": 1.1344419321196156e-06, + "loss": 0.1488, + "step": 4323 + }, + { + "epoch": 0.9838452787258248, + "grad_norm": 1.4980956875745954, + "learning_rate": 1.1343901647693204e-06, + "loss": 0.1309, + "step": 4324 + }, + { + "epoch": 0.9840728100113766, + "grad_norm": 1.2811922093105048, + "learning_rate": 1.134338387008023e-06, + "loss": 0.0593, + "step": 4325 + }, + { + "epoch": 0.9843003412969283, + "grad_norm": 1.3642128367264363, + "learning_rate": 1.134286598836782e-06, + "loss": 0.1017, + "step": 4326 + }, + { + "epoch": 0.9845278725824801, + "grad_norm": 1.7177876329128943, + "learning_rate": 1.1342348002566553e-06, + "loss": 0.0867, + "step": 4327 + }, + { + "epoch": 0.9847554038680318, + "grad_norm": 1.3703182672414826, + "learning_rate": 1.1341829912687023e-06, + "loss": 0.1109, + "step": 4328 + }, + { + "epoch": 0.9849829351535836, + "grad_norm": 1.5958337506813505, + "learning_rate": 1.134131171873981e-06, + "loss": 0.1292, + "step": 4329 + }, + { + "epoch": 0.9852104664391353, + "grad_norm": 3.3711109781277444, + "learning_rate": 1.1340793420735514e-06, + "loss": 0.1689, + "step": 4330 + }, + { + "epoch": 0.9854379977246871, + "grad_norm": 2.000675388129801, + "learning_rate": 1.1340275018684722e-06, + "loss": 0.1109, + "step": 4331 + }, + { + "epoch": 0.985665529010239, + "grad_norm": 2.0533711499371288, + "learning_rate": 1.133975651259803e-06, + "loss": 0.1052, + "step": 4332 + }, + { + "epoch": 0.9858930602957907, + "grad_norm": 1.7529387892804091, + "learning_rate": 1.1339237902486037e-06, + "loss": 0.11, + "step": 4333 + }, + { + "epoch": 0.9861205915813425, + "grad_norm": 1.6889892694894149, + "learning_rate": 1.1338719188359343e-06, + "loss": 0.1184, + "step": 4334 + }, + { + "epoch": 0.9863481228668942, + "grad_norm": 3.5878456222264887, + "learning_rate": 1.1338200370228546e-06, + "loss": 0.1742, + "step": 4335 + }, + { + "epoch": 0.986575654152446, + "grad_norm": 1.6673173056055803, + "learning_rate": 1.1337681448104254e-06, + "loss": 0.1423, + "step": 4336 + }, + { + "epoch": 0.9868031854379977, + "grad_norm": 1.7330563337658174, + "learning_rate": 1.1337162421997072e-06, + "loss": 0.1696, + "step": 4337 + }, + { + "epoch": 0.9870307167235495, + "grad_norm": 2.0663054470234115, + "learning_rate": 1.1336643291917604e-06, + "loss": 0.1273, + "step": 4338 + }, + { + "epoch": 0.9872582480091012, + "grad_norm": 2.11842399087793, + "learning_rate": 1.1336124057876464e-06, + "loss": 0.1032, + "step": 4339 + }, + { + "epoch": 0.987485779294653, + "grad_norm": 2.4883273260756105, + "learning_rate": 1.1335604719884264e-06, + "loss": 0.135, + "step": 4340 + }, + { + "epoch": 0.9877133105802047, + "grad_norm": 1.731780640656524, + "learning_rate": 1.1335085277951616e-06, + "loss": 0.1621, + "step": 4341 + }, + { + "epoch": 0.9879408418657566, + "grad_norm": 1.203909326079841, + "learning_rate": 1.1334565732089138e-06, + "loss": 0.14, + "step": 4342 + }, + { + "epoch": 0.9881683731513083, + "grad_norm": 1.601403569944671, + "learning_rate": 1.133404608230745e-06, + "loss": 0.2358, + "step": 4343 + }, + { + "epoch": 0.9883959044368601, + "grad_norm": 1.964021565396351, + "learning_rate": 1.1333526328617168e-06, + "loss": 0.1895, + "step": 4344 + }, + { + "epoch": 0.9886234357224118, + "grad_norm": 2.3437207735796366, + "learning_rate": 1.133300647102892e-06, + "loss": 0.1144, + "step": 4345 + }, + { + "epoch": 0.9888509670079636, + "grad_norm": 3.5296942681172943, + "learning_rate": 1.1332486509553328e-06, + "loss": 0.1699, + "step": 4346 + }, + { + "epoch": 0.9890784982935154, + "grad_norm": 2.4413901735081134, + "learning_rate": 1.133196644420102e-06, + "loss": 0.1381, + "step": 4347 + }, + { + "epoch": 0.9893060295790671, + "grad_norm": 1.841207848166938, + "learning_rate": 1.1331446274982625e-06, + "loss": 0.132, + "step": 4348 + }, + { + "epoch": 0.9895335608646189, + "grad_norm": 2.662793499305453, + "learning_rate": 1.1330926001908777e-06, + "loss": 0.1421, + "step": 4349 + }, + { + "epoch": 0.9897610921501706, + "grad_norm": 1.809157807855444, + "learning_rate": 1.1330405624990104e-06, + "loss": 0.0863, + "step": 4350 + }, + { + "epoch": 0.9899886234357224, + "grad_norm": 1.999436579625728, + "learning_rate": 1.1329885144237243e-06, + "loss": 0.1119, + "step": 4351 + }, + { + "epoch": 0.9902161547212742, + "grad_norm": 2.0702270524123705, + "learning_rate": 1.1329364559660836e-06, + "loss": 0.1301, + "step": 4352 + }, + { + "epoch": 0.990443686006826, + "grad_norm": 2.094037825552476, + "learning_rate": 1.132884387127152e-06, + "loss": 0.1169, + "step": 4353 + }, + { + "epoch": 0.9906712172923777, + "grad_norm": 1.8657162382122898, + "learning_rate": 1.1328323079079934e-06, + "loss": 0.1845, + "step": 4354 + }, + { + "epoch": 0.9908987485779295, + "grad_norm": 2.0171570859055494, + "learning_rate": 1.1327802183096725e-06, + "loss": 0.1207, + "step": 4355 + }, + { + "epoch": 0.9911262798634812, + "grad_norm": 1.7790517405477932, + "learning_rate": 1.1327281183332542e-06, + "loss": 0.0978, + "step": 4356 + }, + { + "epoch": 0.991353811149033, + "grad_norm": 3.3828371556295216, + "learning_rate": 1.1326760079798027e-06, + "loss": 0.1455, + "step": 4357 + }, + { + "epoch": 0.9915813424345847, + "grad_norm": 1.9916518111295805, + "learning_rate": 1.1326238872503837e-06, + "loss": 0.1303, + "step": 4358 + }, + { + "epoch": 0.9918088737201365, + "grad_norm": 2.3902240797390735, + "learning_rate": 1.1325717561460617e-06, + "loss": 0.113, + "step": 4359 + }, + { + "epoch": 0.9920364050056882, + "grad_norm": 2.543784637415317, + "learning_rate": 1.132519614667903e-06, + "loss": 0.1722, + "step": 4360 + }, + { + "epoch": 0.9922639362912401, + "grad_norm": 1.9595219175974092, + "learning_rate": 1.1324674628169725e-06, + "loss": 0.2108, + "step": 4361 + }, + { + "epoch": 0.9924914675767919, + "grad_norm": 2.5982341940178837, + "learning_rate": 1.1324153005943367e-06, + "loss": 0.1161, + "step": 4362 + }, + { + "epoch": 0.9927189988623436, + "grad_norm": 1.7123301218970444, + "learning_rate": 1.1323631280010611e-06, + "loss": 0.1137, + "step": 4363 + }, + { + "epoch": 0.9929465301478954, + "grad_norm": 1.9745516784082002, + "learning_rate": 1.1323109450382128e-06, + "loss": 0.1954, + "step": 4364 + }, + { + "epoch": 0.9931740614334471, + "grad_norm": 1.4852701391065133, + "learning_rate": 1.1322587517068576e-06, + "loss": 0.092, + "step": 4365 + }, + { + "epoch": 0.9934015927189989, + "grad_norm": 1.2880529332410495, + "learning_rate": 1.1322065480080625e-06, + "loss": 0.1119, + "step": 4366 + }, + { + "epoch": 0.9936291240045506, + "grad_norm": 3.5272975673478877, + "learning_rate": 1.1321543339428946e-06, + "loss": 0.2384, + "step": 4367 + }, + { + "epoch": 0.9938566552901024, + "grad_norm": 1.3569450896326727, + "learning_rate": 1.132102109512421e-06, + "loss": 0.1211, + "step": 4368 + }, + { + "epoch": 0.9940841865756541, + "grad_norm": 4.000179282059702, + "learning_rate": 1.1320498747177088e-06, + "loss": 0.1658, + "step": 4369 + }, + { + "epoch": 0.9943117178612059, + "grad_norm": 1.282888080346011, + "learning_rate": 1.1319976295598258e-06, + "loss": 0.0889, + "step": 4370 + }, + { + "epoch": 0.9945392491467577, + "grad_norm": 1.7469871671415789, + "learning_rate": 1.1319453740398397e-06, + "loss": 0.1074, + "step": 4371 + }, + { + "epoch": 0.9947667804323095, + "grad_norm": 2.332358467010625, + "learning_rate": 1.1318931081588188e-06, + "loss": 0.1113, + "step": 4372 + }, + { + "epoch": 0.9949943117178612, + "grad_norm": 2.4141042022127968, + "learning_rate": 1.1318408319178308e-06, + "loss": 0.2397, + "step": 4373 + }, + { + "epoch": 0.995221843003413, + "grad_norm": 2.536120347644097, + "learning_rate": 1.1317885453179448e-06, + "loss": 0.0825, + "step": 4374 + }, + { + "epoch": 0.9954493742889647, + "grad_norm": 2.758829896595987, + "learning_rate": 1.131736248360229e-06, + "loss": 0.1675, + "step": 4375 + }, + { + "epoch": 0.9956769055745165, + "grad_norm": 1.9724252102552065, + "learning_rate": 1.1316839410457523e-06, + "loss": 0.1328, + "step": 4376 + }, + { + "epoch": 0.9959044368600682, + "grad_norm": 2.2407432243866454, + "learning_rate": 1.1316316233755837e-06, + "loss": 0.1014, + "step": 4377 + }, + { + "epoch": 0.99613196814562, + "grad_norm": 1.5373618862518739, + "learning_rate": 1.1315792953507924e-06, + "loss": 0.0816, + "step": 4378 + }, + { + "epoch": 0.9963594994311717, + "grad_norm": 1.1171814219877398, + "learning_rate": 1.1315269569724483e-06, + "loss": 0.1068, + "step": 4379 + }, + { + "epoch": 0.9965870307167235, + "grad_norm": 1.3551021899004305, + "learning_rate": 1.131474608241621e-06, + "loss": 0.1286, + "step": 4380 + }, + { + "epoch": 0.9968145620022754, + "grad_norm": 1.6680324675811677, + "learning_rate": 1.1314222491593798e-06, + "loss": 0.0832, + "step": 4381 + }, + { + "epoch": 0.9970420932878271, + "grad_norm": 2.6677504525045803, + "learning_rate": 1.1313698797267958e-06, + "loss": 0.1544, + "step": 4382 + }, + { + "epoch": 0.9972696245733789, + "grad_norm": 2.224612434800048, + "learning_rate": 1.1313174999449384e-06, + "loss": 0.1407, + "step": 4383 + }, + { + "epoch": 0.9974971558589306, + "grad_norm": 1.5495425486172278, + "learning_rate": 1.1312651098148788e-06, + "loss": 0.0819, + "step": 4384 + }, + { + "epoch": 0.9977246871444824, + "grad_norm": 3.56356215123795, + "learning_rate": 1.1312127093376876e-06, + "loss": 0.1517, + "step": 4385 + }, + { + "epoch": 0.9979522184300341, + "grad_norm": 2.1877397253991417, + "learning_rate": 1.1311602985144358e-06, + "loss": 0.1212, + "step": 4386 + }, + { + "epoch": 0.9981797497155859, + "grad_norm": 1.3671801752420045, + "learning_rate": 1.1311078773461942e-06, + "loss": 0.1401, + "step": 4387 + }, + { + "epoch": 0.9984072810011376, + "grad_norm": 1.7649784039691725, + "learning_rate": 1.1310554458340345e-06, + "loss": 0.1155, + "step": 4388 + }, + { + "epoch": 0.9986348122866894, + "grad_norm": 2.4503650421883116, + "learning_rate": 1.1310030039790285e-06, + "loss": 0.2011, + "step": 4389 + }, + { + "epoch": 0.9988623435722411, + "grad_norm": 1.0225634979283895, + "learning_rate": 1.1309505517822476e-06, + "loss": 0.0906, + "step": 4390 + }, + { + "epoch": 0.999089874857793, + "grad_norm": 1.6113750565499732, + "learning_rate": 1.1308980892447641e-06, + "loss": 0.1153, + "step": 4391 + }, + { + "epoch": 0.9993174061433447, + "grad_norm": 2.5556355560493413, + "learning_rate": 1.1308456163676501e-06, + "loss": 0.2161, + "step": 4392 + }, + { + "epoch": 0.9995449374288965, + "grad_norm": 1.5875119927953458, + "learning_rate": 1.1307931331519783e-06, + "loss": 0.1184, + "step": 4393 + }, + { + "epoch": 0.9997724687144482, + "grad_norm": 1.8164623960583735, + "learning_rate": 1.1307406395988211e-06, + "loss": 0.1784, + "step": 4394 + }, + { + "epoch": 1.0, + "grad_norm": 1.4833104582392516, + "learning_rate": 1.1306881357092513e-06, + "loss": 0.0764, + "step": 4395 + }, + { + "epoch": 1.0002275312855518, + "grad_norm": 1.4747819154618664, + "learning_rate": 1.1306356214843423e-06, + "loss": 0.0695, + "step": 4396 + }, + { + "epoch": 1.0004550625711035, + "grad_norm": 1.996181577315539, + "learning_rate": 1.1305830969251672e-06, + "loss": 0.0754, + "step": 4397 + }, + { + "epoch": 1.0006825938566553, + "grad_norm": 2.4403368221950883, + "learning_rate": 1.1305305620327994e-06, + "loss": 0.1301, + "step": 4398 + }, + { + "epoch": 1.000910125142207, + "grad_norm": 2.5103859341625836, + "learning_rate": 1.1304780168083128e-06, + "loss": 0.1432, + "step": 4399 + }, + { + "epoch": 1.0011376564277588, + "grad_norm": 2.2576121489705274, + "learning_rate": 1.1304254612527815e-06, + "loss": 0.0847, + "step": 4400 + }, + { + "epoch": 1.0013651877133105, + "grad_norm": 0.6505949983538077, + "learning_rate": 1.130372895367279e-06, + "loss": 0.0411, + "step": 4401 + }, + { + "epoch": 1.0015927189988623, + "grad_norm": 1.5672813299417419, + "learning_rate": 1.1303203191528803e-06, + "loss": 0.0546, + "step": 4402 + }, + { + "epoch": 1.001820250284414, + "grad_norm": 1.8844107765714966, + "learning_rate": 1.1302677326106598e-06, + "loss": 0.0755, + "step": 4403 + }, + { + "epoch": 1.0020477815699658, + "grad_norm": 2.3441263756887456, + "learning_rate": 1.130215135741692e-06, + "loss": 0.0932, + "step": 4404 + }, + { + "epoch": 1.0022753128555177, + "grad_norm": 0.9320352106293828, + "learning_rate": 1.1301625285470522e-06, + "loss": 0.0492, + "step": 4405 + }, + { + "epoch": 1.0025028441410695, + "grad_norm": 1.120280020876937, + "learning_rate": 1.1301099110278156e-06, + "loss": 0.069, + "step": 4406 + }, + { + "epoch": 1.0027303754266212, + "grad_norm": 1.7643601297911453, + "learning_rate": 1.1300572831850574e-06, + "loss": 0.0617, + "step": 4407 + }, + { + "epoch": 1.002957906712173, + "grad_norm": 1.7223702795031912, + "learning_rate": 1.1300046450198532e-06, + "loss": 0.0975, + "step": 4408 + }, + { + "epoch": 1.0031854379977247, + "grad_norm": 1.4345188013436136, + "learning_rate": 1.1299519965332791e-06, + "loss": 0.093, + "step": 4409 + }, + { + "epoch": 1.0034129692832765, + "grad_norm": 2.2514745321091003, + "learning_rate": 1.1298993377264108e-06, + "loss": 0.1432, + "step": 4410 + }, + { + "epoch": 1.0036405005688283, + "grad_norm": 1.5123230129730694, + "learning_rate": 1.129846668600325e-06, + "loss": 0.0897, + "step": 4411 + }, + { + "epoch": 1.00386803185438, + "grad_norm": 1.1854568683295474, + "learning_rate": 1.1297939891560975e-06, + "loss": 0.0795, + "step": 4412 + }, + { + "epoch": 1.0040955631399318, + "grad_norm": 2.1331405358880065, + "learning_rate": 1.1297412993948054e-06, + "loss": 0.1339, + "step": 4413 + }, + { + "epoch": 1.0043230944254835, + "grad_norm": 2.1403450032317006, + "learning_rate": 1.1296885993175255e-06, + "loss": 0.0914, + "step": 4414 + }, + { + "epoch": 1.0045506257110353, + "grad_norm": 2.0722720568137825, + "learning_rate": 1.1296358889253351e-06, + "loss": 0.1073, + "step": 4415 + }, + { + "epoch": 1.004778156996587, + "grad_norm": 1.6803936020610233, + "learning_rate": 1.1295831682193115e-06, + "loss": 0.0737, + "step": 4416 + }, + { + "epoch": 1.0050056882821388, + "grad_norm": 1.6133414130965238, + "learning_rate": 1.1295304372005316e-06, + "loss": 0.0951, + "step": 4417 + }, + { + "epoch": 1.0052332195676905, + "grad_norm": 1.011782483121823, + "learning_rate": 1.129477695870074e-06, + "loss": 0.0989, + "step": 4418 + }, + { + "epoch": 1.0054607508532423, + "grad_norm": 1.697633317040223, + "learning_rate": 1.129424944229016e-06, + "loss": 0.0689, + "step": 4419 + }, + { + "epoch": 1.005688282138794, + "grad_norm": 1.3459567859085924, + "learning_rate": 1.1293721822784359e-06, + "loss": 0.0781, + "step": 4420 + }, + { + "epoch": 1.0059158134243458, + "grad_norm": 1.7935976771238544, + "learning_rate": 1.1293194100194121e-06, + "loss": 0.0679, + "step": 4421 + }, + { + "epoch": 1.0061433447098975, + "grad_norm": 1.4206028858761883, + "learning_rate": 1.1292666274530232e-06, + "loss": 0.0604, + "step": 4422 + }, + { + "epoch": 1.0063708759954493, + "grad_norm": 1.8737473801514988, + "learning_rate": 1.129213834580348e-06, + "loss": 0.0892, + "step": 4423 + }, + { + "epoch": 1.006598407281001, + "grad_norm": 1.4562994747544142, + "learning_rate": 1.1291610314024653e-06, + "loss": 0.0621, + "step": 4424 + }, + { + "epoch": 1.006825938566553, + "grad_norm": 1.6144715931341527, + "learning_rate": 1.1291082179204548e-06, + "loss": 0.0891, + "step": 4425 + }, + { + "epoch": 1.0070534698521048, + "grad_norm": 1.417358558553153, + "learning_rate": 1.1290553941353954e-06, + "loss": 0.0469, + "step": 4426 + }, + { + "epoch": 1.0072810011376565, + "grad_norm": 1.2388288333880253, + "learning_rate": 1.1290025600483667e-06, + "loss": 0.1067, + "step": 4427 + }, + { + "epoch": 1.0075085324232083, + "grad_norm": 1.2758605963996141, + "learning_rate": 1.1289497156604487e-06, + "loss": 0.1237, + "step": 4428 + }, + { + "epoch": 1.00773606370876, + "grad_norm": 1.4052553927682834, + "learning_rate": 1.1288968609727216e-06, + "loss": 0.0604, + "step": 4429 + }, + { + "epoch": 1.0079635949943118, + "grad_norm": 1.119682904381322, + "learning_rate": 1.1288439959862654e-06, + "loss": 0.0902, + "step": 4430 + }, + { + "epoch": 1.0081911262798635, + "grad_norm": 1.7203969017671572, + "learning_rate": 1.128791120702161e-06, + "loss": 0.1059, + "step": 4431 + }, + { + "epoch": 1.0084186575654153, + "grad_norm": 2.6485605967013686, + "learning_rate": 1.1287382351214884e-06, + "loss": 0.1169, + "step": 4432 + }, + { + "epoch": 1.008646188850967, + "grad_norm": 1.2027084176698186, + "learning_rate": 1.128685339245329e-06, + "loss": 0.0624, + "step": 4433 + }, + { + "epoch": 1.0088737201365188, + "grad_norm": 1.933143757940641, + "learning_rate": 1.1286324330747637e-06, + "loss": 0.1785, + "step": 4434 + }, + { + "epoch": 1.0091012514220705, + "grad_norm": 1.8514912698937183, + "learning_rate": 1.1285795166108735e-06, + "loss": 0.0624, + "step": 4435 + }, + { + "epoch": 1.0093287827076223, + "grad_norm": 1.8965468725111816, + "learning_rate": 1.1285265898547406e-06, + "loss": 0.0993, + "step": 4436 + }, + { + "epoch": 1.009556313993174, + "grad_norm": 2.2625515852101232, + "learning_rate": 1.1284736528074464e-06, + "loss": 0.1312, + "step": 4437 + }, + { + "epoch": 1.0097838452787258, + "grad_norm": 0.9513962126598605, + "learning_rate": 1.1284207054700727e-06, + "loss": 0.042, + "step": 4438 + }, + { + "epoch": 1.0100113765642775, + "grad_norm": 1.463114771314281, + "learning_rate": 1.1283677478437016e-06, + "loss": 0.0763, + "step": 4439 + }, + { + "epoch": 1.0102389078498293, + "grad_norm": 2.354851541830737, + "learning_rate": 1.1283147799294158e-06, + "loss": 0.1046, + "step": 4440 + }, + { + "epoch": 1.010466439135381, + "grad_norm": 1.1766333820955324, + "learning_rate": 1.1282618017282977e-06, + "loss": 0.1248, + "step": 4441 + }, + { + "epoch": 1.0106939704209328, + "grad_norm": 1.5422549437499427, + "learning_rate": 1.1282088132414297e-06, + "loss": 0.114, + "step": 4442 + }, + { + "epoch": 1.0109215017064845, + "grad_norm": 2.2298271761237483, + "learning_rate": 1.1281558144698956e-06, + "loss": 0.0924, + "step": 4443 + }, + { + "epoch": 1.0111490329920365, + "grad_norm": 2.519798954114394, + "learning_rate": 1.128102805414778e-06, + "loss": 0.157, + "step": 4444 + }, + { + "epoch": 1.0113765642775883, + "grad_norm": 1.4415500014186144, + "learning_rate": 1.1280497860771603e-06, + "loss": 0.1013, + "step": 4445 + }, + { + "epoch": 1.01160409556314, + "grad_norm": 2.0125789101308205, + "learning_rate": 1.1279967564581264e-06, + "loss": 0.0848, + "step": 4446 + }, + { + "epoch": 1.0118316268486918, + "grad_norm": 2.251550332027883, + "learning_rate": 1.12794371655876e-06, + "loss": 0.0646, + "step": 4447 + }, + { + "epoch": 1.0120591581342435, + "grad_norm": 1.0526890868708527, + "learning_rate": 1.127890666380145e-06, + "loss": 0.0767, + "step": 4448 + }, + { + "epoch": 1.0122866894197953, + "grad_norm": 1.0816163534208596, + "learning_rate": 1.1278376059233658e-06, + "loss": 0.0563, + "step": 4449 + }, + { + "epoch": 1.012514220705347, + "grad_norm": 1.6202591756823337, + "learning_rate": 1.127784535189507e-06, + "loss": 0.082, + "step": 4450 + }, + { + "epoch": 1.0127417519908988, + "grad_norm": 2.114784758301838, + "learning_rate": 1.127731454179653e-06, + "loss": 0.1543, + "step": 4451 + }, + { + "epoch": 1.0129692832764505, + "grad_norm": 1.5200937072938314, + "learning_rate": 1.1276783628948887e-06, + "loss": 0.0926, + "step": 4452 + }, + { + "epoch": 1.0131968145620023, + "grad_norm": 0.8832036316733904, + "learning_rate": 1.1276252613362995e-06, + "loss": 0.051, + "step": 4453 + }, + { + "epoch": 1.013424345847554, + "grad_norm": 1.791095170570671, + "learning_rate": 1.1275721495049702e-06, + "loss": 0.126, + "step": 4454 + }, + { + "epoch": 1.0136518771331058, + "grad_norm": 1.4567099222693494, + "learning_rate": 1.1275190274019867e-06, + "loss": 0.0937, + "step": 4455 + }, + { + "epoch": 1.0138794084186575, + "grad_norm": 1.8661960896015608, + "learning_rate": 1.1274658950284347e-06, + "loss": 0.0746, + "step": 4456 + }, + { + "epoch": 1.0141069397042093, + "grad_norm": 2.054547499620593, + "learning_rate": 1.1274127523854e-06, + "loss": 0.098, + "step": 4457 + }, + { + "epoch": 1.014334470989761, + "grad_norm": 1.4935646424282278, + "learning_rate": 1.1273595994739688e-06, + "loss": 0.096, + "step": 4458 + }, + { + "epoch": 1.0145620022753128, + "grad_norm": 1.8465602968348116, + "learning_rate": 1.1273064362952272e-06, + "loss": 0.0927, + "step": 4459 + }, + { + "epoch": 1.0147895335608645, + "grad_norm": 1.921602489637173, + "learning_rate": 1.1272532628502621e-06, + "loss": 0.0645, + "step": 4460 + }, + { + "epoch": 1.0150170648464163, + "grad_norm": 1.383709505060379, + "learning_rate": 1.1272000791401602e-06, + "loss": 0.1155, + "step": 4461 + }, + { + "epoch": 1.015244596131968, + "grad_norm": 1.1642883714545766, + "learning_rate": 1.1271468851660084e-06, + "loss": 0.0924, + "step": 4462 + }, + { + "epoch": 1.01547212741752, + "grad_norm": 1.2445933138998786, + "learning_rate": 1.127093680928894e-06, + "loss": 0.052, + "step": 4463 + }, + { + "epoch": 1.0156996587030718, + "grad_norm": 3.4339049713427667, + "learning_rate": 1.1270404664299042e-06, + "loss": 0.1109, + "step": 4464 + }, + { + "epoch": 1.0159271899886235, + "grad_norm": 1.6198738915791715, + "learning_rate": 1.1269872416701267e-06, + "loss": 0.1034, + "step": 4465 + }, + { + "epoch": 1.0161547212741753, + "grad_norm": 1.8925845714324792, + "learning_rate": 1.1269340066506493e-06, + "loss": 0.0613, + "step": 4466 + }, + { + "epoch": 1.016382252559727, + "grad_norm": 1.4139672561252017, + "learning_rate": 1.12688076137256e-06, + "loss": 0.0875, + "step": 4467 + }, + { + "epoch": 1.0166097838452788, + "grad_norm": 1.6115204329476527, + "learning_rate": 1.1268275058369472e-06, + "loss": 0.089, + "step": 4468 + }, + { + "epoch": 1.0168373151308305, + "grad_norm": 2.0290217444617853, + "learning_rate": 1.1267742400448992e-06, + "loss": 0.1111, + "step": 4469 + }, + { + "epoch": 1.0170648464163823, + "grad_norm": 1.779159979414333, + "learning_rate": 1.1267209639975046e-06, + "loss": 0.0839, + "step": 4470 + }, + { + "epoch": 1.017292377701934, + "grad_norm": 1.3519160870633005, + "learning_rate": 1.1266676776958523e-06, + "loss": 0.0885, + "step": 4471 + }, + { + "epoch": 1.0175199089874858, + "grad_norm": 1.4586296010653232, + "learning_rate": 1.1266143811410317e-06, + "loss": 0.0668, + "step": 4472 + }, + { + "epoch": 1.0177474402730375, + "grad_norm": 2.016744342105442, + "learning_rate": 1.1265610743341316e-06, + "loss": 0.1435, + "step": 4473 + }, + { + "epoch": 1.0179749715585893, + "grad_norm": 1.1369067026133264, + "learning_rate": 1.1265077572762418e-06, + "loss": 0.108, + "step": 4474 + }, + { + "epoch": 1.018202502844141, + "grad_norm": 8.903995088813302, + "learning_rate": 1.1264544299684518e-06, + "loss": 0.1271, + "step": 4475 + }, + { + "epoch": 1.0184300341296928, + "grad_norm": 1.6336221990915103, + "learning_rate": 1.1264010924118518e-06, + "loss": 0.0878, + "step": 4476 + }, + { + "epoch": 1.0186575654152445, + "grad_norm": 1.658678946132026, + "learning_rate": 1.1263477446075315e-06, + "loss": 0.1076, + "step": 4477 + }, + { + "epoch": 1.0188850967007963, + "grad_norm": 1.707584314379802, + "learning_rate": 1.1262943865565818e-06, + "loss": 0.0964, + "step": 4478 + }, + { + "epoch": 1.019112627986348, + "grad_norm": 1.9120240803965922, + "learning_rate": 1.1262410182600927e-06, + "loss": 0.1555, + "step": 4479 + }, + { + "epoch": 1.0193401592718998, + "grad_norm": 0.9282735076245712, + "learning_rate": 1.1261876397191554e-06, + "loss": 0.0626, + "step": 4480 + }, + { + "epoch": 1.0195676905574516, + "grad_norm": 1.6682216698884167, + "learning_rate": 1.1261342509348604e-06, + "loss": 0.1009, + "step": 4481 + }, + { + "epoch": 1.0197952218430033, + "grad_norm": 1.2851007188914807, + "learning_rate": 1.126080851908299e-06, + "loss": 0.0596, + "step": 4482 + }, + { + "epoch": 1.0200227531285553, + "grad_norm": 1.7254355793639742, + "learning_rate": 1.1260274426405629e-06, + "loss": 0.1048, + "step": 4483 + }, + { + "epoch": 1.020250284414107, + "grad_norm": 1.4883296224579121, + "learning_rate": 1.1259740231327434e-06, + "loss": 0.1188, + "step": 4484 + }, + { + "epoch": 1.0204778156996588, + "grad_norm": 1.4466839237766165, + "learning_rate": 1.1259205933859325e-06, + "loss": 0.0551, + "step": 4485 + }, + { + "epoch": 1.0207053469852105, + "grad_norm": 1.6088130949320203, + "learning_rate": 1.1258671534012216e-06, + "loss": 0.1246, + "step": 4486 + }, + { + "epoch": 1.0209328782707623, + "grad_norm": 1.9306073145993348, + "learning_rate": 1.1258137031797037e-06, + "loss": 0.0852, + "step": 4487 + }, + { + "epoch": 1.021160409556314, + "grad_norm": 1.4681528767570764, + "learning_rate": 1.125760242722471e-06, + "loss": 0.073, + "step": 4488 + }, + { + "epoch": 1.0213879408418658, + "grad_norm": 1.8686422771397933, + "learning_rate": 1.1257067720306159e-06, + "loss": 0.0845, + "step": 4489 + }, + { + "epoch": 1.0216154721274175, + "grad_norm": 1.1681457319700357, + "learning_rate": 1.1256532911052313e-06, + "loss": 0.1197, + "step": 4490 + }, + { + "epoch": 1.0218430034129693, + "grad_norm": 1.1190800102290346, + "learning_rate": 1.1255997999474105e-06, + "loss": 0.0398, + "step": 4491 + }, + { + "epoch": 1.022070534698521, + "grad_norm": 1.42435477981386, + "learning_rate": 1.1255462985582465e-06, + "loss": 0.071, + "step": 4492 + }, + { + "epoch": 1.0222980659840728, + "grad_norm": 1.8224520226771979, + "learning_rate": 1.125492786938833e-06, + "loss": 0.1101, + "step": 4493 + }, + { + "epoch": 1.0225255972696246, + "grad_norm": 1.4069476965043848, + "learning_rate": 1.1254392650902633e-06, + "loss": 0.0753, + "step": 4494 + }, + { + "epoch": 1.0227531285551763, + "grad_norm": 1.5861436800658038, + "learning_rate": 1.1253857330136316e-06, + "loss": 0.1617, + "step": 4495 + }, + { + "epoch": 1.022980659840728, + "grad_norm": 1.2272660070201709, + "learning_rate": 1.125332190710032e-06, + "loss": 0.0804, + "step": 4496 + }, + { + "epoch": 1.0232081911262798, + "grad_norm": 1.5725673439218046, + "learning_rate": 1.125278638180559e-06, + "loss": 0.0666, + "step": 4497 + }, + { + "epoch": 1.0234357224118316, + "grad_norm": 2.002782257227273, + "learning_rate": 1.1252250754263064e-06, + "loss": 0.0899, + "step": 4498 + }, + { + "epoch": 1.0236632536973833, + "grad_norm": 1.456920740957027, + "learning_rate": 1.1251715024483695e-06, + "loss": 0.0697, + "step": 4499 + }, + { + "epoch": 1.023890784982935, + "grad_norm": 1.108094546855012, + "learning_rate": 1.125117919247843e-06, + "loss": 0.0665, + "step": 4500 + }, + { + "epoch": 1.0241183162684868, + "grad_norm": 1.4026920191631007, + "learning_rate": 1.1250643258258225e-06, + "loss": 0.1499, + "step": 4501 + }, + { + "epoch": 1.0243458475540388, + "grad_norm": 2.029837399777728, + "learning_rate": 1.1250107221834027e-06, + "loss": 0.1546, + "step": 4502 + }, + { + "epoch": 1.0245733788395905, + "grad_norm": 0.887255926728237, + "learning_rate": 1.12495710832168e-06, + "loss": 0.0648, + "step": 4503 + }, + { + "epoch": 1.0248009101251423, + "grad_norm": 1.615916482107172, + "learning_rate": 1.1249034842417489e-06, + "loss": 0.097, + "step": 4504 + }, + { + "epoch": 1.025028441410694, + "grad_norm": 1.402342600220676, + "learning_rate": 1.1248498499447065e-06, + "loss": 0.0915, + "step": 4505 + }, + { + "epoch": 1.0252559726962458, + "grad_norm": 1.592211294130619, + "learning_rate": 1.1247962054316485e-06, + "loss": 0.1552, + "step": 4506 + }, + { + "epoch": 1.0254835039817976, + "grad_norm": 1.7740489402500772, + "learning_rate": 1.1247425507036715e-06, + "loss": 0.1013, + "step": 4507 + }, + { + "epoch": 1.0257110352673493, + "grad_norm": 2.236483074679014, + "learning_rate": 1.1246888857618719e-06, + "loss": 0.1269, + "step": 4508 + }, + { + "epoch": 1.025938566552901, + "grad_norm": 2.026475728815122, + "learning_rate": 1.1246352106073466e-06, + "loss": 0.1492, + "step": 4509 + }, + { + "epoch": 1.0261660978384528, + "grad_norm": 1.4043085831160411, + "learning_rate": 1.1245815252411928e-06, + "loss": 0.1606, + "step": 4510 + }, + { + "epoch": 1.0263936291240046, + "grad_norm": 1.3478102522875726, + "learning_rate": 1.1245278296645073e-06, + "loss": 0.138, + "step": 4511 + }, + { + "epoch": 1.0266211604095563, + "grad_norm": 1.5130340363330543, + "learning_rate": 1.124474123878388e-06, + "loss": 0.0584, + "step": 4512 + }, + { + "epoch": 1.026848691695108, + "grad_norm": 3.0044980775672867, + "learning_rate": 1.1244204078839325e-06, + "loss": 0.1119, + "step": 4513 + }, + { + "epoch": 1.0270762229806598, + "grad_norm": 1.7841919834467284, + "learning_rate": 1.1243666816822382e-06, + "loss": 0.1234, + "step": 4514 + }, + { + "epoch": 1.0273037542662116, + "grad_norm": 2.0156152758572996, + "learning_rate": 1.1243129452744036e-06, + "loss": 0.1077, + "step": 4515 + }, + { + "epoch": 1.0275312855517633, + "grad_norm": 1.3091943644643798, + "learning_rate": 1.1242591986615268e-06, + "loss": 0.1099, + "step": 4516 + }, + { + "epoch": 1.027758816837315, + "grad_norm": 1.9516196995491382, + "learning_rate": 1.1242054418447063e-06, + "loss": 0.0734, + "step": 4517 + }, + { + "epoch": 1.0279863481228668, + "grad_norm": 1.8054250637811027, + "learning_rate": 1.1241516748250408e-06, + "loss": 0.1087, + "step": 4518 + }, + { + "epoch": 1.0282138794084186, + "grad_norm": 2.874562035518301, + "learning_rate": 1.1240978976036294e-06, + "loss": 0.2136, + "step": 4519 + }, + { + "epoch": 1.0284414106939703, + "grad_norm": 1.0770500977157564, + "learning_rate": 1.124044110181571e-06, + "loss": 0.0895, + "step": 4520 + }, + { + "epoch": 1.028668941979522, + "grad_norm": 1.5287898441796675, + "learning_rate": 1.1239903125599648e-06, + "loss": 0.1935, + "step": 4521 + }, + { + "epoch": 1.028896473265074, + "grad_norm": 1.2141770039320132, + "learning_rate": 1.1239365047399106e-06, + "loss": 0.1316, + "step": 4522 + }, + { + "epoch": 1.0291240045506258, + "grad_norm": 1.7076130924026178, + "learning_rate": 1.1238826867225077e-06, + "loss": 0.1006, + "step": 4523 + }, + { + "epoch": 1.0293515358361776, + "grad_norm": 2.172750884502955, + "learning_rate": 1.1238288585088567e-06, + "loss": 0.0846, + "step": 4524 + }, + { + "epoch": 1.0295790671217293, + "grad_norm": 1.5673811074944972, + "learning_rate": 1.1237750201000574e-06, + "loss": 0.1357, + "step": 4525 + }, + { + "epoch": 1.029806598407281, + "grad_norm": 2.361397974753223, + "learning_rate": 1.1237211714972098e-06, + "loss": 0.1862, + "step": 4526 + }, + { + "epoch": 1.0300341296928328, + "grad_norm": 1.59606082155671, + "learning_rate": 1.1236673127014152e-06, + "loss": 0.1032, + "step": 4527 + }, + { + "epoch": 1.0302616609783846, + "grad_norm": 1.4456707646110476, + "learning_rate": 1.1236134437137738e-06, + "loss": 0.0556, + "step": 4528 + }, + { + "epoch": 1.0304891922639363, + "grad_norm": 1.9510921040673854, + "learning_rate": 1.1235595645353869e-06, + "loss": 0.128, + "step": 4529 + }, + { + "epoch": 1.030716723549488, + "grad_norm": 1.1503668274652719, + "learning_rate": 1.1235056751673554e-06, + "loss": 0.0674, + "step": 4530 + }, + { + "epoch": 1.0309442548350398, + "grad_norm": 1.9465124654240242, + "learning_rate": 1.123451775610781e-06, + "loss": 0.1208, + "step": 4531 + }, + { + "epoch": 1.0311717861205916, + "grad_norm": 0.8598640038575219, + "learning_rate": 1.1233978658667651e-06, + "loss": 0.0863, + "step": 4532 + }, + { + "epoch": 1.0313993174061433, + "grad_norm": 1.311506374530133, + "learning_rate": 1.1233439459364097e-06, + "loss": 0.0974, + "step": 4533 + }, + { + "epoch": 1.031626848691695, + "grad_norm": 2.3989951849111546, + "learning_rate": 1.1232900158208166e-06, + "loss": 0.0865, + "step": 4534 + }, + { + "epoch": 1.0318543799772468, + "grad_norm": 2.0112852237633376, + "learning_rate": 1.1232360755210883e-06, + "loss": 0.0935, + "step": 4535 + }, + { + "epoch": 1.0320819112627986, + "grad_norm": 1.4723457241660713, + "learning_rate": 1.123182125038327e-06, + "loss": 0.1059, + "step": 4536 + }, + { + "epoch": 1.0323094425483503, + "grad_norm": 2.086137912197496, + "learning_rate": 1.1231281643736353e-06, + "loss": 0.1119, + "step": 4537 + }, + { + "epoch": 1.032536973833902, + "grad_norm": 1.2936224119616597, + "learning_rate": 1.1230741935281163e-06, + "loss": 0.1498, + "step": 4538 + }, + { + "epoch": 1.0327645051194538, + "grad_norm": 1.9307986409707742, + "learning_rate": 1.123020212502873e-06, + "loss": 0.0861, + "step": 4539 + }, + { + "epoch": 1.0329920364050056, + "grad_norm": 1.9937770717754, + "learning_rate": 1.1229662212990088e-06, + "loss": 0.2239, + "step": 4540 + }, + { + "epoch": 1.0332195676905576, + "grad_norm": 1.5346898654618122, + "learning_rate": 1.1229122199176268e-06, + "loss": 0.0611, + "step": 4541 + }, + { + "epoch": 1.0334470989761093, + "grad_norm": 1.1175713863887276, + "learning_rate": 1.1228582083598311e-06, + "loss": 0.0794, + "step": 4542 + }, + { + "epoch": 1.033674630261661, + "grad_norm": 2.2328550567917484, + "learning_rate": 1.122804186626725e-06, + "loss": 0.0804, + "step": 4543 + }, + { + "epoch": 1.0339021615472128, + "grad_norm": 1.1528552818810556, + "learning_rate": 1.1227501547194133e-06, + "loss": 0.0955, + "step": 4544 + }, + { + "epoch": 1.0341296928327646, + "grad_norm": 2.9382553353367706, + "learning_rate": 1.1226961126390001e-06, + "loss": 0.1691, + "step": 4545 + }, + { + "epoch": 1.0343572241183163, + "grad_norm": 1.9931887866365103, + "learning_rate": 1.1226420603865898e-06, + "loss": 0.1017, + "step": 4546 + }, + { + "epoch": 1.034584755403868, + "grad_norm": 1.6486532012729544, + "learning_rate": 1.122587997963287e-06, + "loss": 0.121, + "step": 4547 + }, + { + "epoch": 1.0348122866894198, + "grad_norm": 2.2996104242216595, + "learning_rate": 1.122533925370197e-06, + "loss": 0.08, + "step": 4548 + }, + { + "epoch": 1.0350398179749716, + "grad_norm": 2.05667419327721, + "learning_rate": 1.1224798426084246e-06, + "loss": 0.0905, + "step": 4549 + }, + { + "epoch": 1.0352673492605233, + "grad_norm": 1.6750893821513704, + "learning_rate": 1.1224257496790756e-06, + "loss": 0.1614, + "step": 4550 + }, + { + "epoch": 1.035494880546075, + "grad_norm": 1.8124862022379826, + "learning_rate": 1.122371646583255e-06, + "loss": 0.1094, + "step": 4551 + }, + { + "epoch": 1.0357224118316268, + "grad_norm": 1.1789521290469518, + "learning_rate": 1.1223175333220688e-06, + "loss": 0.0544, + "step": 4552 + }, + { + "epoch": 1.0359499431171786, + "grad_norm": 1.4726290884229478, + "learning_rate": 1.122263409896623e-06, + "loss": 0.0415, + "step": 4553 + }, + { + "epoch": 1.0361774744027303, + "grad_norm": 1.5983976452101323, + "learning_rate": 1.1222092763080242e-06, + "loss": 0.1389, + "step": 4554 + }, + { + "epoch": 1.036405005688282, + "grad_norm": 1.7064510686199956, + "learning_rate": 1.1221551325573779e-06, + "loss": 0.1995, + "step": 4555 + }, + { + "epoch": 1.0366325369738338, + "grad_norm": 1.8831480696901932, + "learning_rate": 1.1221009786457914e-06, + "loss": 0.1159, + "step": 4556 + }, + { + "epoch": 1.0368600682593856, + "grad_norm": 1.8251977038998766, + "learning_rate": 1.1220468145743713e-06, + "loss": 0.0593, + "step": 4557 + }, + { + "epoch": 1.0370875995449373, + "grad_norm": 2.250994931720387, + "learning_rate": 1.1219926403442247e-06, + "loss": 0.1241, + "step": 4558 + }, + { + "epoch": 1.037315130830489, + "grad_norm": 2.149745732751104, + "learning_rate": 1.1219384559564587e-06, + "loss": 0.0641, + "step": 4559 + }, + { + "epoch": 1.0375426621160408, + "grad_norm": 1.481812364607165, + "learning_rate": 1.1218842614121806e-06, + "loss": 0.1053, + "step": 4560 + }, + { + "epoch": 1.0377701934015928, + "grad_norm": 3.1982872252107017, + "learning_rate": 1.1218300567124983e-06, + "loss": 0.0885, + "step": 4561 + }, + { + "epoch": 1.0379977246871446, + "grad_norm": 1.5207099396961141, + "learning_rate": 1.1217758418585195e-06, + "loss": 0.1485, + "step": 4562 + }, + { + "epoch": 1.0382252559726963, + "grad_norm": 1.357320820979069, + "learning_rate": 1.1217216168513522e-06, + "loss": 0.0863, + "step": 4563 + }, + { + "epoch": 1.038452787258248, + "grad_norm": 1.7245587056307685, + "learning_rate": 1.1216673816921048e-06, + "loss": 0.1392, + "step": 4564 + }, + { + "epoch": 1.0386803185437998, + "grad_norm": 1.3388736695631331, + "learning_rate": 1.1216131363818859e-06, + "loss": 0.0712, + "step": 4565 + }, + { + "epoch": 1.0389078498293516, + "grad_norm": 1.6991747662454542, + "learning_rate": 1.1215588809218038e-06, + "loss": 0.1135, + "step": 4566 + }, + { + "epoch": 1.0391353811149033, + "grad_norm": 1.7543912071841141, + "learning_rate": 1.1215046153129678e-06, + "loss": 0.0656, + "step": 4567 + }, + { + "epoch": 1.039362912400455, + "grad_norm": 1.3584849508462316, + "learning_rate": 1.1214503395564866e-06, + "loss": 0.0536, + "step": 4568 + }, + { + "epoch": 1.0395904436860068, + "grad_norm": 0.9240388411832862, + "learning_rate": 1.1213960536534698e-06, + "loss": 0.0332, + "step": 4569 + }, + { + "epoch": 1.0398179749715586, + "grad_norm": 1.6546290657725133, + "learning_rate": 1.1213417576050267e-06, + "loss": 0.1005, + "step": 4570 + }, + { + "epoch": 1.0400455062571103, + "grad_norm": 1.3885969596375243, + "learning_rate": 1.1212874514122669e-06, + "loss": 0.1047, + "step": 4571 + }, + { + "epoch": 1.040273037542662, + "grad_norm": 2.017166875593496, + "learning_rate": 1.1212331350763007e-06, + "loss": 0.1456, + "step": 4572 + }, + { + "epoch": 1.0405005688282138, + "grad_norm": 1.444948131842632, + "learning_rate": 1.1211788085982381e-06, + "loss": 0.0579, + "step": 4573 + }, + { + "epoch": 1.0407281001137656, + "grad_norm": 2.406951373648501, + "learning_rate": 1.1211244719791892e-06, + "loss": 0.1583, + "step": 4574 + }, + { + "epoch": 1.0409556313993173, + "grad_norm": 2.197884382980465, + "learning_rate": 1.1210701252202647e-06, + "loss": 0.1243, + "step": 4575 + }, + { + "epoch": 1.041183162684869, + "grad_norm": 1.4181821249909716, + "learning_rate": 1.1210157683225753e-06, + "loss": 0.0774, + "step": 4576 + }, + { + "epoch": 1.0414106939704209, + "grad_norm": 1.6354137634346002, + "learning_rate": 1.1209614012872323e-06, + "loss": 0.0879, + "step": 4577 + }, + { + "epoch": 1.0416382252559726, + "grad_norm": 1.0303262963114748, + "learning_rate": 1.1209070241153462e-06, + "loss": 0.104, + "step": 4578 + }, + { + "epoch": 1.0418657565415244, + "grad_norm": 2.1044621651764723, + "learning_rate": 1.1208526368080288e-06, + "loss": 0.1179, + "step": 4579 + }, + { + "epoch": 1.0420932878270763, + "grad_norm": 1.2487244960480326, + "learning_rate": 1.120798239366392e-06, + "loss": 0.0926, + "step": 4580 + }, + { + "epoch": 1.042320819112628, + "grad_norm": 1.224762304659932, + "learning_rate": 1.1207438317915468e-06, + "loss": 0.0584, + "step": 4581 + }, + { + "epoch": 1.0425483503981798, + "grad_norm": 1.4128646708017816, + "learning_rate": 1.1206894140846055e-06, + "loss": 0.0498, + "step": 4582 + }, + { + "epoch": 1.0427758816837316, + "grad_norm": 2.6167846385891567, + "learning_rate": 1.1206349862466807e-06, + "loss": 0.1441, + "step": 4583 + }, + { + "epoch": 1.0430034129692833, + "grad_norm": 1.3737857439282581, + "learning_rate": 1.1205805482788846e-06, + "loss": 0.0846, + "step": 4584 + }, + { + "epoch": 1.043230944254835, + "grad_norm": 1.4524682790222339, + "learning_rate": 1.1205261001823293e-06, + "loss": 0.1068, + "step": 4585 + }, + { + "epoch": 1.0434584755403868, + "grad_norm": 1.4167951979507394, + "learning_rate": 1.1204716419581281e-06, + "loss": 0.0669, + "step": 4586 + }, + { + "epoch": 1.0436860068259386, + "grad_norm": 1.9626607976722603, + "learning_rate": 1.1204171736073942e-06, + "loss": 0.0596, + "step": 4587 + }, + { + "epoch": 1.0439135381114903, + "grad_norm": 0.965468994366607, + "learning_rate": 1.1203626951312405e-06, + "loss": 0.0621, + "step": 4588 + }, + { + "epoch": 1.044141069397042, + "grad_norm": 1.0583807127720506, + "learning_rate": 1.1203082065307805e-06, + "loss": 0.0468, + "step": 4589 + }, + { + "epoch": 1.0443686006825939, + "grad_norm": 1.353491198352836, + "learning_rate": 1.1202537078071277e-06, + "loss": 0.0972, + "step": 4590 + }, + { + "epoch": 1.0445961319681456, + "grad_norm": 2.135114467092782, + "learning_rate": 1.1201991989613963e-06, + "loss": 0.0813, + "step": 4591 + }, + { + "epoch": 1.0448236632536974, + "grad_norm": 1.7028217782174653, + "learning_rate": 1.1201446799947003e-06, + "loss": 0.0968, + "step": 4592 + }, + { + "epoch": 1.045051194539249, + "grad_norm": 1.1670795075657887, + "learning_rate": 1.1200901509081537e-06, + "loss": 0.0451, + "step": 4593 + }, + { + "epoch": 1.0452787258248009, + "grad_norm": 0.9396890176885602, + "learning_rate": 1.120035611702871e-06, + "loss": 0.0931, + "step": 4594 + }, + { + "epoch": 1.0455062571103526, + "grad_norm": 1.7193114496757371, + "learning_rate": 1.1199810623799673e-06, + "loss": 0.1, + "step": 4595 + }, + { + "epoch": 1.0457337883959044, + "grad_norm": 1.6764946052943284, + "learning_rate": 1.119926502940557e-06, + "loss": 0.0632, + "step": 4596 + }, + { + "epoch": 1.0459613196814561, + "grad_norm": 1.7852800076694235, + "learning_rate": 1.1198719333857555e-06, + "loss": 0.0833, + "step": 4597 + }, + { + "epoch": 1.0461888509670079, + "grad_norm": 2.3234885020060094, + "learning_rate": 1.119817353716678e-06, + "loss": 0.0766, + "step": 4598 + }, + { + "epoch": 1.0464163822525596, + "grad_norm": 1.4797020097779663, + "learning_rate": 1.11976276393444e-06, + "loss": 0.055, + "step": 4599 + }, + { + "epoch": 1.0466439135381116, + "grad_norm": 1.634600829488423, + "learning_rate": 1.1197081640401572e-06, + "loss": 0.1146, + "step": 4600 + }, + { + "epoch": 1.0468714448236633, + "grad_norm": 1.8746220464401466, + "learning_rate": 1.1196535540349453e-06, + "loss": 0.1229, + "step": 4601 + }, + { + "epoch": 1.047098976109215, + "grad_norm": 2.3559999951268624, + "learning_rate": 1.119598933919921e-06, + "loss": 0.1277, + "step": 4602 + }, + { + "epoch": 1.0473265073947668, + "grad_norm": 1.0083505534534876, + "learning_rate": 1.1195443036962002e-06, + "loss": 0.067, + "step": 4603 + }, + { + "epoch": 1.0475540386803186, + "grad_norm": 0.8680772321300652, + "learning_rate": 1.1194896633648996e-06, + "loss": 0.0546, + "step": 4604 + }, + { + "epoch": 1.0477815699658704, + "grad_norm": 2.427145418238581, + "learning_rate": 1.1194350129271358e-06, + "loss": 0.1341, + "step": 4605 + }, + { + "epoch": 1.048009101251422, + "grad_norm": 2.3008553524730933, + "learning_rate": 1.119380352384026e-06, + "loss": 0.1438, + "step": 4606 + }, + { + "epoch": 1.0482366325369739, + "grad_norm": 2.2562366914461376, + "learning_rate": 1.1193256817366871e-06, + "loss": 0.0817, + "step": 4607 + }, + { + "epoch": 1.0484641638225256, + "grad_norm": 2.020216249876879, + "learning_rate": 1.1192710009862365e-06, + "loss": 0.0862, + "step": 4608 + }, + { + "epoch": 1.0486916951080774, + "grad_norm": 1.2123623538229429, + "learning_rate": 1.1192163101337921e-06, + "loss": 0.0589, + "step": 4609 + }, + { + "epoch": 1.0489192263936291, + "grad_norm": 0.7241673359429386, + "learning_rate": 1.1191616091804712e-06, + "loss": 0.0505, + "step": 4610 + }, + { + "epoch": 1.0491467576791809, + "grad_norm": 1.552740662211025, + "learning_rate": 1.1191068981273919e-06, + "loss": 0.09, + "step": 4611 + }, + { + "epoch": 1.0493742889647326, + "grad_norm": 1.036959423467112, + "learning_rate": 1.1190521769756729e-06, + "loss": 0.0897, + "step": 4612 + }, + { + "epoch": 1.0496018202502844, + "grad_norm": 1.3205129521300534, + "learning_rate": 1.118997445726432e-06, + "loss": 0.1081, + "step": 4613 + }, + { + "epoch": 1.0498293515358361, + "grad_norm": 2.0114633754619655, + "learning_rate": 1.118942704380788e-06, + "loss": 0.1166, + "step": 4614 + }, + { + "epoch": 1.0500568828213879, + "grad_norm": 2.426555596440509, + "learning_rate": 1.11888795293986e-06, + "loss": 0.1125, + "step": 4615 + }, + { + "epoch": 1.0502844141069396, + "grad_norm": 1.8060961422070485, + "learning_rate": 1.1188331914047666e-06, + "loss": 0.0715, + "step": 4616 + }, + { + "epoch": 1.0505119453924914, + "grad_norm": 1.7144346832473707, + "learning_rate": 1.1187784197766269e-06, + "loss": 0.0914, + "step": 4617 + }, + { + "epoch": 1.0507394766780431, + "grad_norm": 1.3408767090161975, + "learning_rate": 1.1187236380565608e-06, + "loss": 0.1256, + "step": 4618 + }, + { + "epoch": 1.050967007963595, + "grad_norm": 1.5665147033583955, + "learning_rate": 1.1186688462456879e-06, + "loss": 0.0745, + "step": 4619 + }, + { + "epoch": 1.0511945392491469, + "grad_norm": 1.3016358691203176, + "learning_rate": 1.118614044345128e-06, + "loss": 0.0716, + "step": 4620 + }, + { + "epoch": 1.0514220705346986, + "grad_norm": 2.045746614166097, + "learning_rate": 1.1185592323560006e-06, + "loss": 0.1064, + "step": 4621 + }, + { + "epoch": 1.0516496018202504, + "grad_norm": 1.7391658273056083, + "learning_rate": 1.1185044102794267e-06, + "loss": 0.1341, + "step": 4622 + }, + { + "epoch": 1.051877133105802, + "grad_norm": 2.219721072354361, + "learning_rate": 1.1184495781165263e-06, + "loss": 0.133, + "step": 4623 + }, + { + "epoch": 1.0521046643913539, + "grad_norm": 3.8225784468179542, + "learning_rate": 1.1183947358684203e-06, + "loss": 0.0639, + "step": 4624 + }, + { + "epoch": 1.0523321956769056, + "grad_norm": 1.2881445992416602, + "learning_rate": 1.1183398835362298e-06, + "loss": 0.0725, + "step": 4625 + }, + { + "epoch": 1.0525597269624574, + "grad_norm": 2.304920176139941, + "learning_rate": 1.1182850211210752e-06, + "loss": 0.0916, + "step": 4626 + }, + { + "epoch": 1.0527872582480091, + "grad_norm": 1.837079698598604, + "learning_rate": 1.1182301486240782e-06, + "loss": 0.1102, + "step": 4627 + }, + { + "epoch": 1.0530147895335609, + "grad_norm": 1.8399238573897383, + "learning_rate": 1.1181752660463604e-06, + "loss": 0.0992, + "step": 4628 + }, + { + "epoch": 1.0532423208191126, + "grad_norm": 1.1035148907341084, + "learning_rate": 1.1181203733890433e-06, + "loss": 0.0705, + "step": 4629 + }, + { + "epoch": 1.0534698521046644, + "grad_norm": 1.284026604785757, + "learning_rate": 1.118065470653249e-06, + "loss": 0.1644, + "step": 4630 + }, + { + "epoch": 1.0536973833902161, + "grad_norm": 1.390049950226029, + "learning_rate": 1.1180105578400993e-06, + "loss": 0.1564, + "step": 4631 + }, + { + "epoch": 1.0539249146757679, + "grad_norm": 2.007397098246824, + "learning_rate": 1.117955634950717e-06, + "loss": 0.1507, + "step": 4632 + }, + { + "epoch": 1.0541524459613196, + "grad_norm": 2.3767977233264457, + "learning_rate": 1.117900701986224e-06, + "loss": 0.2561, + "step": 4633 + }, + { + "epoch": 1.0543799772468714, + "grad_norm": 1.4294941525905513, + "learning_rate": 1.1178457589477434e-06, + "loss": 0.0564, + "step": 4634 + }, + { + "epoch": 1.0546075085324231, + "grad_norm": 1.5732365617118949, + "learning_rate": 1.1177908058363984e-06, + "loss": 0.0443, + "step": 4635 + }, + { + "epoch": 1.0548350398179749, + "grad_norm": 2.193495953897933, + "learning_rate": 1.1177358426533115e-06, + "loss": 0.1299, + "step": 4636 + }, + { + "epoch": 1.0550625711035266, + "grad_norm": 2.2874034231596765, + "learning_rate": 1.1176808693996067e-06, + "loss": 0.1652, + "step": 4637 + }, + { + "epoch": 1.0552901023890784, + "grad_norm": 1.9529291419382124, + "learning_rate": 1.117625886076407e-06, + "loss": 0.0731, + "step": 4638 + }, + { + "epoch": 1.0555176336746304, + "grad_norm": 2.759454689334842, + "learning_rate": 1.1175708926848363e-06, + "loss": 0.0728, + "step": 4639 + }, + { + "epoch": 1.0557451649601821, + "grad_norm": 1.5338896104841968, + "learning_rate": 1.1175158892260187e-06, + "loss": 0.1026, + "step": 4640 + }, + { + "epoch": 1.0559726962457339, + "grad_norm": 1.4378649009540245, + "learning_rate": 1.1174608757010785e-06, + "loss": 0.0894, + "step": 4641 + }, + { + "epoch": 1.0562002275312856, + "grad_norm": 1.5134147490681178, + "learning_rate": 1.1174058521111398e-06, + "loss": 0.0924, + "step": 4642 + }, + { + "epoch": 1.0564277588168374, + "grad_norm": 1.917950365533836, + "learning_rate": 1.1173508184573273e-06, + "loss": 0.0958, + "step": 4643 + }, + { + "epoch": 1.0566552901023891, + "grad_norm": 1.3789830119471285, + "learning_rate": 1.1172957747407657e-06, + "loss": 0.0816, + "step": 4644 + }, + { + "epoch": 1.0568828213879409, + "grad_norm": 1.2790394403403582, + "learning_rate": 1.11724072096258e-06, + "loss": 0.1122, + "step": 4645 + }, + { + "epoch": 1.0571103526734926, + "grad_norm": 1.0135911519508813, + "learning_rate": 1.1171856571238958e-06, + "loss": 0.0646, + "step": 4646 + }, + { + "epoch": 1.0573378839590444, + "grad_norm": 1.8567305234547644, + "learning_rate": 1.1171305832258378e-06, + "loss": 0.1032, + "step": 4647 + }, + { + "epoch": 1.0575654152445961, + "grad_norm": 2.4359360899436098, + "learning_rate": 1.117075499269532e-06, + "loss": 0.0981, + "step": 4648 + }, + { + "epoch": 1.0577929465301479, + "grad_norm": 1.5734015373702166, + "learning_rate": 1.1170204052561045e-06, + "loss": 0.0963, + "step": 4649 + }, + { + "epoch": 1.0580204778156996, + "grad_norm": 1.515846702271059, + "learning_rate": 1.1169653011866806e-06, + "loss": 0.1188, + "step": 4650 + }, + { + "epoch": 1.0582480091012514, + "grad_norm": 1.4229982355792368, + "learning_rate": 1.1169101870623872e-06, + "loss": 0.1127, + "step": 4651 + }, + { + "epoch": 1.0584755403868031, + "grad_norm": 1.6554966357036334, + "learning_rate": 1.1168550628843506e-06, + "loss": 0.0789, + "step": 4652 + }, + { + "epoch": 1.058703071672355, + "grad_norm": 1.5192728010457008, + "learning_rate": 1.116799928653697e-06, + "loss": 0.1155, + "step": 4653 + }, + { + "epoch": 1.0589306029579066, + "grad_norm": 2.2933314149526898, + "learning_rate": 1.1167447843715536e-06, + "loss": 0.1181, + "step": 4654 + }, + { + "epoch": 1.0591581342434584, + "grad_norm": 1.2625384296045075, + "learning_rate": 1.1166896300390475e-06, + "loss": 0.0763, + "step": 4655 + }, + { + "epoch": 1.0593856655290101, + "grad_norm": 1.2148579431146604, + "learning_rate": 1.1166344656573058e-06, + "loss": 0.0759, + "step": 4656 + }, + { + "epoch": 1.059613196814562, + "grad_norm": 1.2800038842620556, + "learning_rate": 1.116579291227456e-06, + "loss": 0.0766, + "step": 4657 + }, + { + "epoch": 1.0598407281001139, + "grad_norm": 1.8062981732112615, + "learning_rate": 1.1165241067506258e-06, + "loss": 0.1418, + "step": 4658 + }, + { + "epoch": 1.0600682593856656, + "grad_norm": 1.4847559035451179, + "learning_rate": 1.116468912227943e-06, + "loss": 0.1106, + "step": 4659 + }, + { + "epoch": 1.0602957906712174, + "grad_norm": 1.8952358982035167, + "learning_rate": 1.1164137076605359e-06, + "loss": 0.0678, + "step": 4660 + }, + { + "epoch": 1.0605233219567691, + "grad_norm": 1.4381274024858606, + "learning_rate": 1.1163584930495323e-06, + "loss": 0.1189, + "step": 4661 + }, + { + "epoch": 1.0607508532423209, + "grad_norm": 2.7251709950763447, + "learning_rate": 1.1163032683960612e-06, + "loss": 0.1013, + "step": 4662 + }, + { + "epoch": 1.0609783845278726, + "grad_norm": 1.0847987069830758, + "learning_rate": 1.116248033701251e-06, + "loss": 0.0759, + "step": 4663 + }, + { + "epoch": 1.0612059158134244, + "grad_norm": 1.8578929914423556, + "learning_rate": 1.1161927889662307e-06, + "loss": 0.0724, + "step": 4664 + }, + { + "epoch": 1.0614334470989761, + "grad_norm": 1.1291431011699526, + "learning_rate": 1.1161375341921293e-06, + "loss": 0.0809, + "step": 4665 + }, + { + "epoch": 1.0616609783845279, + "grad_norm": 1.9028105528288284, + "learning_rate": 1.1160822693800761e-06, + "loss": 0.0788, + "step": 4666 + }, + { + "epoch": 1.0618885096700796, + "grad_norm": 1.5906930716104764, + "learning_rate": 1.116026994531201e-06, + "loss": 0.1806, + "step": 4667 + }, + { + "epoch": 1.0621160409556314, + "grad_norm": 1.4806073894025749, + "learning_rate": 1.1159717096466332e-06, + "loss": 0.0917, + "step": 4668 + }, + { + "epoch": 1.0623435722411831, + "grad_norm": 1.7555748075182174, + "learning_rate": 1.1159164147275026e-06, + "loss": 0.1221, + "step": 4669 + }, + { + "epoch": 1.062571103526735, + "grad_norm": 1.9388103998351425, + "learning_rate": 1.11586110977494e-06, + "loss": 0.101, + "step": 4670 + }, + { + "epoch": 1.0627986348122866, + "grad_norm": 2.54858320636725, + "learning_rate": 1.1158057947900749e-06, + "loss": 0.0826, + "step": 4671 + }, + { + "epoch": 1.0630261660978384, + "grad_norm": 1.7593201330973203, + "learning_rate": 1.1157504697740384e-06, + "loss": 0.0921, + "step": 4672 + }, + { + "epoch": 1.0632536973833902, + "grad_norm": 1.3601408281439702, + "learning_rate": 1.1156951347279612e-06, + "loss": 0.0909, + "step": 4673 + }, + { + "epoch": 1.063481228668942, + "grad_norm": 2.4921469575180923, + "learning_rate": 1.1156397896529739e-06, + "loss": 0.1152, + "step": 4674 + }, + { + "epoch": 1.0637087599544937, + "grad_norm": 2.61680060550728, + "learning_rate": 1.1155844345502079e-06, + "loss": 0.1493, + "step": 4675 + }, + { + "epoch": 1.0639362912400454, + "grad_norm": 1.2304215041389976, + "learning_rate": 1.1155290694207946e-06, + "loss": 0.0948, + "step": 4676 + }, + { + "epoch": 1.0641638225255972, + "grad_norm": 1.499666016464916, + "learning_rate": 1.1154736942658655e-06, + "loss": 0.0505, + "step": 4677 + }, + { + "epoch": 1.0643913538111491, + "grad_norm": 1.5895614920307246, + "learning_rate": 1.1154183090865523e-06, + "loss": 0.102, + "step": 4678 + }, + { + "epoch": 1.0646188850967009, + "grad_norm": 2.115605875809715, + "learning_rate": 1.1153629138839869e-06, + "loss": 0.1225, + "step": 4679 + }, + { + "epoch": 1.0648464163822526, + "grad_norm": 2.3561395200123747, + "learning_rate": 1.115307508659302e-06, + "loss": 0.1031, + "step": 4680 + }, + { + "epoch": 1.0650739476678044, + "grad_norm": 1.1554650460211398, + "learning_rate": 1.115252093413629e-06, + "loss": 0.1216, + "step": 4681 + }, + { + "epoch": 1.0653014789533561, + "grad_norm": 1.4823986847762305, + "learning_rate": 1.1151966681481013e-06, + "loss": 0.0594, + "step": 4682 + }, + { + "epoch": 1.065529010238908, + "grad_norm": 1.6251842447572193, + "learning_rate": 1.1151412328638516e-06, + "loss": 0.0907, + "step": 4683 + }, + { + "epoch": 1.0657565415244596, + "grad_norm": 1.1507819894156535, + "learning_rate": 1.1150857875620129e-06, + "loss": 0.0472, + "step": 4684 + }, + { + "epoch": 1.0659840728100114, + "grad_norm": 1.3805150312721923, + "learning_rate": 1.1150303322437179e-06, + "loss": 0.0771, + "step": 4685 + }, + { + "epoch": 1.0662116040955631, + "grad_norm": 1.358932906756745, + "learning_rate": 1.1149748669101005e-06, + "loss": 0.0513, + "step": 4686 + }, + { + "epoch": 1.066439135381115, + "grad_norm": 5.71481532720155, + "learning_rate": 1.1149193915622942e-06, + "loss": 0.1539, + "step": 4687 + }, + { + "epoch": 1.0666666666666667, + "grad_norm": 2.6297199766031865, + "learning_rate": 1.1148639062014325e-06, + "loss": 0.1656, + "step": 4688 + }, + { + "epoch": 1.0668941979522184, + "grad_norm": 1.9080409573155523, + "learning_rate": 1.11480841082865e-06, + "loss": 0.0762, + "step": 4689 + }, + { + "epoch": 1.0671217292377702, + "grad_norm": 1.7438819672239148, + "learning_rate": 1.1147529054450805e-06, + "loss": 0.0885, + "step": 4690 + }, + { + "epoch": 1.067349260523322, + "grad_norm": 2.0574645907461955, + "learning_rate": 1.1146973900518587e-06, + "loss": 0.1442, + "step": 4691 + }, + { + "epoch": 1.0675767918088737, + "grad_norm": 2.6252395753862294, + "learning_rate": 1.1146418646501189e-06, + "loss": 0.1878, + "step": 4692 + }, + { + "epoch": 1.0678043230944254, + "grad_norm": 1.0913326597489377, + "learning_rate": 1.114586329240996e-06, + "loss": 0.0378, + "step": 4693 + }, + { + "epoch": 1.0680318543799772, + "grad_norm": 1.9442927380841992, + "learning_rate": 1.1145307838256255e-06, + "loss": 0.0636, + "step": 4694 + }, + { + "epoch": 1.068259385665529, + "grad_norm": 1.2746222339178395, + "learning_rate": 1.1144752284051422e-06, + "loss": 0.0999, + "step": 4695 + }, + { + "epoch": 1.068486916951081, + "grad_norm": 1.5469092535365685, + "learning_rate": 1.1144196629806817e-06, + "loss": 0.1484, + "step": 4696 + }, + { + "epoch": 1.0687144482366326, + "grad_norm": 1.8149614702646155, + "learning_rate": 1.1143640875533795e-06, + "loss": 0.0656, + "step": 4697 + }, + { + "epoch": 1.0689419795221844, + "grad_norm": 1.952760861877504, + "learning_rate": 1.1143085021243717e-06, + "loss": 0.116, + "step": 4698 + }, + { + "epoch": 1.0691695108077361, + "grad_norm": 1.5095899894504656, + "learning_rate": 1.1142529066947941e-06, + "loss": 0.1141, + "step": 4699 + }, + { + "epoch": 1.069397042093288, + "grad_norm": 1.2954537974569234, + "learning_rate": 1.1141973012657834e-06, + "loss": 0.0521, + "step": 4700 + }, + { + "epoch": 1.0696245733788396, + "grad_norm": 1.8070083827165602, + "learning_rate": 1.1141416858384753e-06, + "loss": 0.1679, + "step": 4701 + }, + { + "epoch": 1.0698521046643914, + "grad_norm": 1.1006465944591577, + "learning_rate": 1.1140860604140076e-06, + "loss": 0.1058, + "step": 4702 + }, + { + "epoch": 1.0700796359499432, + "grad_norm": 1.1929894713879348, + "learning_rate": 1.114030424993516e-06, + "loss": 0.0839, + "step": 4703 + }, + { + "epoch": 1.070307167235495, + "grad_norm": 1.8098296886220957, + "learning_rate": 1.1139747795781382e-06, + "loss": 0.0751, + "step": 4704 + }, + { + "epoch": 1.0705346985210467, + "grad_norm": 1.3608491383009365, + "learning_rate": 1.1139191241690116e-06, + "loss": 0.0636, + "step": 4705 + }, + { + "epoch": 1.0707622298065984, + "grad_norm": 0.7968044916321982, + "learning_rate": 1.1138634587672734e-06, + "loss": 0.0196, + "step": 4706 + }, + { + "epoch": 1.0709897610921502, + "grad_norm": 1.1618137365747236, + "learning_rate": 1.1138077833740616e-06, + "loss": 0.0543, + "step": 4707 + }, + { + "epoch": 1.071217292377702, + "grad_norm": 1.6636727239611144, + "learning_rate": 1.1137520979905138e-06, + "loss": 0.078, + "step": 4708 + }, + { + "epoch": 1.0714448236632537, + "grad_norm": 1.3224939982871424, + "learning_rate": 1.1136964026177683e-06, + "loss": 0.081, + "step": 4709 + }, + { + "epoch": 1.0716723549488054, + "grad_norm": 2.102064361920128, + "learning_rate": 1.113640697256963e-06, + "loss": 0.0756, + "step": 4710 + }, + { + "epoch": 1.0718998862343572, + "grad_norm": 1.5869491881927147, + "learning_rate": 1.113584981909237e-06, + "loss": 0.0977, + "step": 4711 + }, + { + "epoch": 1.072127417519909, + "grad_norm": 2.528352862438852, + "learning_rate": 1.1135292565757288e-06, + "loss": 0.0922, + "step": 4712 + }, + { + "epoch": 1.0723549488054607, + "grad_norm": 2.1890348604827152, + "learning_rate": 1.1134735212575772e-06, + "loss": 0.0748, + "step": 4713 + }, + { + "epoch": 1.0725824800910124, + "grad_norm": 2.721031491992015, + "learning_rate": 1.1134177759559216e-06, + "loss": 0.166, + "step": 4714 + }, + { + "epoch": 1.0728100113765642, + "grad_norm": 1.2699457735676007, + "learning_rate": 1.1133620206719011e-06, + "loss": 0.1368, + "step": 4715 + }, + { + "epoch": 1.073037542662116, + "grad_norm": 1.4567637517882654, + "learning_rate": 1.1133062554066551e-06, + "loss": 0.0957, + "step": 4716 + }, + { + "epoch": 1.073265073947668, + "grad_norm": 1.1981840138040833, + "learning_rate": 1.1132504801613237e-06, + "loss": 0.0996, + "step": 4717 + }, + { + "epoch": 1.0734926052332197, + "grad_norm": 2.5605965751924016, + "learning_rate": 1.1131946949370467e-06, + "loss": 0.1288, + "step": 4718 + }, + { + "epoch": 1.0737201365187714, + "grad_norm": 2.2199917939261424, + "learning_rate": 1.113138899734964e-06, + "loss": 0.1176, + "step": 4719 + }, + { + "epoch": 1.0739476678043232, + "grad_norm": 1.8349868522475592, + "learning_rate": 1.1130830945562165e-06, + "loss": 0.1546, + "step": 4720 + }, + { + "epoch": 1.074175199089875, + "grad_norm": 2.835134050444113, + "learning_rate": 1.1130272794019442e-06, + "loss": 0.1291, + "step": 4721 + }, + { + "epoch": 1.0744027303754267, + "grad_norm": 1.009967448557906, + "learning_rate": 1.1129714542732882e-06, + "loss": 0.0669, + "step": 4722 + }, + { + "epoch": 1.0746302616609784, + "grad_norm": 1.529305478280692, + "learning_rate": 1.1129156191713893e-06, + "loss": 0.0727, + "step": 4723 + }, + { + "epoch": 1.0748577929465302, + "grad_norm": 1.3960850231566015, + "learning_rate": 1.1128597740973886e-06, + "loss": 0.1294, + "step": 4724 + }, + { + "epoch": 1.075085324232082, + "grad_norm": 1.5662589297405252, + "learning_rate": 1.1128039190524278e-06, + "loss": 0.0441, + "step": 4725 + }, + { + "epoch": 1.0753128555176337, + "grad_norm": 2.147391485773349, + "learning_rate": 1.112748054037648e-06, + "loss": 0.0715, + "step": 4726 + }, + { + "epoch": 1.0755403868031854, + "grad_norm": 1.9537494356887783, + "learning_rate": 1.1126921790541915e-06, + "loss": 0.1266, + "step": 4727 + }, + { + "epoch": 1.0757679180887372, + "grad_norm": 1.807870084752669, + "learning_rate": 1.1126362941032e-06, + "loss": 0.0663, + "step": 4728 + }, + { + "epoch": 1.075995449374289, + "grad_norm": 1.2731792994483404, + "learning_rate": 1.1125803991858156e-06, + "loss": 0.0777, + "step": 4729 + }, + { + "epoch": 1.0762229806598407, + "grad_norm": 1.9875691409773595, + "learning_rate": 1.1125244943031809e-06, + "loss": 0.0595, + "step": 4730 + }, + { + "epoch": 1.0764505119453924, + "grad_norm": 1.3445874567355482, + "learning_rate": 1.1124685794564383e-06, + "loss": 0.1031, + "step": 4731 + }, + { + "epoch": 1.0766780432309442, + "grad_norm": 1.2246092409806042, + "learning_rate": 1.1124126546467307e-06, + "loss": 0.0939, + "step": 4732 + }, + { + "epoch": 1.076905574516496, + "grad_norm": 1.431137650749334, + "learning_rate": 1.1123567198752012e-06, + "loss": 0.1475, + "step": 4733 + }, + { + "epoch": 1.0771331058020477, + "grad_norm": 1.8994402495648997, + "learning_rate": 1.1123007751429928e-06, + "loss": 0.1429, + "step": 4734 + }, + { + "epoch": 1.0773606370875997, + "grad_norm": 1.4085164315858043, + "learning_rate": 1.1122448204512493e-06, + "loss": 0.1428, + "step": 4735 + }, + { + "epoch": 1.0775881683731514, + "grad_norm": 1.2999518219276889, + "learning_rate": 1.1121888558011136e-06, + "loss": 0.0662, + "step": 4736 + }, + { + "epoch": 1.0778156996587032, + "grad_norm": 1.7946331927576185, + "learning_rate": 1.11213288119373e-06, + "loss": 0.0765, + "step": 4737 + }, + { + "epoch": 1.078043230944255, + "grad_norm": 1.6859375154233005, + "learning_rate": 1.1120768966302424e-06, + "loss": 0.0686, + "step": 4738 + }, + { + "epoch": 1.0782707622298067, + "grad_norm": 0.9077920170925504, + "learning_rate": 1.1120209021117953e-06, + "loss": 0.0337, + "step": 4739 + }, + { + "epoch": 1.0784982935153584, + "grad_norm": 1.5316954108177494, + "learning_rate": 1.111964897639533e-06, + "loss": 0.0663, + "step": 4740 + }, + { + "epoch": 1.0787258248009102, + "grad_norm": 1.4238197221056483, + "learning_rate": 1.1119088832145999e-06, + "loss": 0.0609, + "step": 4741 + }, + { + "epoch": 1.078953356086462, + "grad_norm": 1.8622699266174956, + "learning_rate": 1.1118528588381408e-06, + "loss": 0.0967, + "step": 4742 + }, + { + "epoch": 1.0791808873720137, + "grad_norm": 2.132771649775128, + "learning_rate": 1.111796824511301e-06, + "loss": 0.0949, + "step": 4743 + }, + { + "epoch": 1.0794084186575654, + "grad_norm": 1.4389905550232829, + "learning_rate": 1.1117407802352257e-06, + "loss": 0.0592, + "step": 4744 + }, + { + "epoch": 1.0796359499431172, + "grad_norm": 2.313842970080908, + "learning_rate": 1.11168472601106e-06, + "loss": 0.11, + "step": 4745 + }, + { + "epoch": 1.079863481228669, + "grad_norm": 1.189342164666975, + "learning_rate": 1.1116286618399502e-06, + "loss": 0.1139, + "step": 4746 + }, + { + "epoch": 1.0800910125142207, + "grad_norm": 1.5878813013229367, + "learning_rate": 1.1115725877230416e-06, + "loss": 0.0941, + "step": 4747 + }, + { + "epoch": 1.0803185437997724, + "grad_norm": 1.5708323312118704, + "learning_rate": 1.1115165036614803e-06, + "loss": 0.0805, + "step": 4748 + }, + { + "epoch": 1.0805460750853242, + "grad_norm": 1.7412513143173347, + "learning_rate": 1.1114604096564128e-06, + "loss": 0.0927, + "step": 4749 + }, + { + "epoch": 1.080773606370876, + "grad_norm": 2.1069394060272932, + "learning_rate": 1.1114043057089855e-06, + "loss": 0.0689, + "step": 4750 + }, + { + "epoch": 1.0810011376564277, + "grad_norm": 1.584268007677841, + "learning_rate": 1.1113481918203447e-06, + "loss": 0.1189, + "step": 4751 + }, + { + "epoch": 1.0812286689419794, + "grad_norm": 1.6608102806453529, + "learning_rate": 1.111292067991638e-06, + "loss": 0.0689, + "step": 4752 + }, + { + "epoch": 1.0814562002275312, + "grad_norm": 1.5725937241650605, + "learning_rate": 1.1112359342240118e-06, + "loss": 0.0956, + "step": 4753 + }, + { + "epoch": 1.081683731513083, + "grad_norm": 3.407327630763881, + "learning_rate": 1.1111797905186137e-06, + "loss": 0.1318, + "step": 4754 + }, + { + "epoch": 1.0819112627986347, + "grad_norm": 2.3742849467462017, + "learning_rate": 1.111123636876591e-06, + "loss": 0.1672, + "step": 4755 + }, + { + "epoch": 1.0821387940841867, + "grad_norm": 2.436154151939714, + "learning_rate": 1.1110674732990915e-06, + "loss": 0.0952, + "step": 4756 + }, + { + "epoch": 1.0823663253697384, + "grad_norm": 1.9025140854878257, + "learning_rate": 1.1110112997872627e-06, + "loss": 0.0782, + "step": 4757 + }, + { + "epoch": 1.0825938566552902, + "grad_norm": 1.8868683604933614, + "learning_rate": 1.1109551163422535e-06, + "loss": 0.0948, + "step": 4758 + }, + { + "epoch": 1.082821387940842, + "grad_norm": 1.3681510858841002, + "learning_rate": 1.1108989229652115e-06, + "loss": 0.0732, + "step": 4759 + }, + { + "epoch": 1.0830489192263937, + "grad_norm": 2.282564072440466, + "learning_rate": 1.1108427196572854e-06, + "loss": 0.0941, + "step": 4760 + }, + { + "epoch": 1.0832764505119454, + "grad_norm": 1.185382354142743, + "learning_rate": 1.110786506419624e-06, + "loss": 0.1061, + "step": 4761 + }, + { + "epoch": 1.0835039817974972, + "grad_norm": 1.2750556200490737, + "learning_rate": 1.110730283253376e-06, + "loss": 0.0983, + "step": 4762 + }, + { + "epoch": 1.083731513083049, + "grad_norm": 0.9432019459962149, + "learning_rate": 1.1106740501596904e-06, + "loss": 0.0719, + "step": 4763 + }, + { + "epoch": 1.0839590443686007, + "grad_norm": 2.399313506752063, + "learning_rate": 1.110617807139717e-06, + "loss": 0.1592, + "step": 4764 + }, + { + "epoch": 1.0841865756541524, + "grad_norm": 2.0585624525235673, + "learning_rate": 1.1105615541946049e-06, + "loss": 0.1019, + "step": 4765 + }, + { + "epoch": 1.0844141069397042, + "grad_norm": 1.5509608506343646, + "learning_rate": 1.1105052913255038e-06, + "loss": 0.063, + "step": 4766 + }, + { + "epoch": 1.084641638225256, + "grad_norm": 2.3875790847090896, + "learning_rate": 1.1104490185335638e-06, + "loss": 0.0617, + "step": 4767 + }, + { + "epoch": 1.0848691695108077, + "grad_norm": 1.5526617818466677, + "learning_rate": 1.1103927358199349e-06, + "loss": 0.0636, + "step": 4768 + }, + { + "epoch": 1.0850967007963594, + "grad_norm": 2.070040492410767, + "learning_rate": 1.1103364431857672e-06, + "loss": 0.1196, + "step": 4769 + }, + { + "epoch": 1.0853242320819112, + "grad_norm": 0.9037860659580849, + "learning_rate": 1.1102801406322118e-06, + "loss": 0.0788, + "step": 4770 + }, + { + "epoch": 1.085551763367463, + "grad_norm": 2.51446051197986, + "learning_rate": 1.110223828160419e-06, + "loss": 0.1056, + "step": 4771 + }, + { + "epoch": 1.0857792946530147, + "grad_norm": 1.5056718047772528, + "learning_rate": 1.1101675057715396e-06, + "loss": 0.0553, + "step": 4772 + }, + { + "epoch": 1.0860068259385665, + "grad_norm": 1.3450576783641792, + "learning_rate": 1.110111173466725e-06, + "loss": 0.114, + "step": 4773 + }, + { + "epoch": 1.0862343572241184, + "grad_norm": 2.3282598728110564, + "learning_rate": 1.1100548312471266e-06, + "loss": 0.1398, + "step": 4774 + }, + { + "epoch": 1.0864618885096702, + "grad_norm": 1.5835969433641823, + "learning_rate": 1.1099984791138957e-06, + "loss": 0.0563, + "step": 4775 + }, + { + "epoch": 1.086689419795222, + "grad_norm": 1.7042160190222275, + "learning_rate": 1.109942117068184e-06, + "loss": 0.0595, + "step": 4776 + }, + { + "epoch": 1.0869169510807737, + "grad_norm": 1.8015845399975852, + "learning_rate": 1.1098857451111437e-06, + "loss": 0.1093, + "step": 4777 + }, + { + "epoch": 1.0871444823663254, + "grad_norm": 1.4929033187735148, + "learning_rate": 1.1098293632439267e-06, + "loss": 0.0568, + "step": 4778 + }, + { + "epoch": 1.0873720136518772, + "grad_norm": 1.819585409380252, + "learning_rate": 1.1097729714676855e-06, + "loss": 0.0784, + "step": 4779 + }, + { + "epoch": 1.087599544937429, + "grad_norm": 1.6093861745924347, + "learning_rate": 1.1097165697835726e-06, + "loss": 0.0815, + "step": 4780 + }, + { + "epoch": 1.0878270762229807, + "grad_norm": 2.3915450012148947, + "learning_rate": 1.1096601581927407e-06, + "loss": 0.0843, + "step": 4781 + }, + { + "epoch": 1.0880546075085324, + "grad_norm": 1.9249173368001073, + "learning_rate": 1.109603736696343e-06, + "loss": 0.1324, + "step": 4782 + }, + { + "epoch": 1.0882821387940842, + "grad_norm": 1.8896040569960282, + "learning_rate": 1.1095473052955322e-06, + "loss": 0.1745, + "step": 4783 + }, + { + "epoch": 1.088509670079636, + "grad_norm": 1.7358772955397874, + "learning_rate": 1.1094908639914617e-06, + "loss": 0.0989, + "step": 4784 + }, + { + "epoch": 1.0887372013651877, + "grad_norm": 1.5331795064228024, + "learning_rate": 1.1094344127852855e-06, + "loss": 0.1397, + "step": 4785 + }, + { + "epoch": 1.0889647326507395, + "grad_norm": 1.3003409141659286, + "learning_rate": 1.1093779516781571e-06, + "loss": 0.0485, + "step": 4786 + }, + { + "epoch": 1.0891922639362912, + "grad_norm": 1.6700571373067126, + "learning_rate": 1.1093214806712305e-06, + "loss": 0.097, + "step": 4787 + }, + { + "epoch": 1.089419795221843, + "grad_norm": 1.9278343656449526, + "learning_rate": 1.1092649997656597e-06, + "loss": 0.0512, + "step": 4788 + }, + { + "epoch": 1.0896473265073947, + "grad_norm": 1.577450330940266, + "learning_rate": 1.1092085089625992e-06, + "loss": 0.0609, + "step": 4789 + }, + { + "epoch": 1.0898748577929465, + "grad_norm": 2.1296342852219263, + "learning_rate": 1.1091520082632037e-06, + "loss": 0.0782, + "step": 4790 + }, + { + "epoch": 1.0901023890784982, + "grad_norm": 1.0783695776737507, + "learning_rate": 1.1090954976686277e-06, + "loss": 0.0638, + "step": 4791 + }, + { + "epoch": 1.09032992036405, + "grad_norm": 2.154161769490288, + "learning_rate": 1.1090389771800264e-06, + "loss": 0.1817, + "step": 4792 + }, + { + "epoch": 1.0905574516496017, + "grad_norm": 1.5962990513677988, + "learning_rate": 1.1089824467985549e-06, + "loss": 0.0733, + "step": 4793 + }, + { + "epoch": 1.0907849829351535, + "grad_norm": 1.4177139267549546, + "learning_rate": 1.1089259065253684e-06, + "loss": 0.0906, + "step": 4794 + }, + { + "epoch": 1.0910125142207054, + "grad_norm": 1.5152987959640343, + "learning_rate": 1.1088693563616226e-06, + "loss": 0.0775, + "step": 4795 + }, + { + "epoch": 1.0912400455062572, + "grad_norm": 1.0343892482754946, + "learning_rate": 1.1088127963084736e-06, + "loss": 0.092, + "step": 4796 + }, + { + "epoch": 1.091467576791809, + "grad_norm": 1.3749328966275285, + "learning_rate": 1.108756226367077e-06, + "loss": 0.0552, + "step": 4797 + }, + { + "epoch": 1.0916951080773607, + "grad_norm": 1.8028526464917305, + "learning_rate": 1.108699646538589e-06, + "loss": 0.1074, + "step": 4798 + }, + { + "epoch": 1.0919226393629125, + "grad_norm": 1.511012234308768, + "learning_rate": 1.108643056824166e-06, + "loss": 0.1182, + "step": 4799 + }, + { + "epoch": 1.0921501706484642, + "grad_norm": 1.8866591465514666, + "learning_rate": 1.108586457224965e-06, + "loss": 0.0742, + "step": 4800 + }, + { + "epoch": 1.092377701934016, + "grad_norm": 2.1241623003632992, + "learning_rate": 1.1085298477421421e-06, + "loss": 0.0651, + "step": 4801 + }, + { + "epoch": 1.0926052332195677, + "grad_norm": 1.9682663221505046, + "learning_rate": 1.1084732283768548e-06, + "loss": 0.0589, + "step": 4802 + }, + { + "epoch": 1.0928327645051195, + "grad_norm": 1.7770689808885827, + "learning_rate": 1.1084165991302601e-06, + "loss": 0.0624, + "step": 4803 + }, + { + "epoch": 1.0930602957906712, + "grad_norm": 1.9487102696775727, + "learning_rate": 1.1083599600035155e-06, + "loss": 0.2643, + "step": 4804 + }, + { + "epoch": 1.093287827076223, + "grad_norm": 1.3820178073566898, + "learning_rate": 1.1083033109977787e-06, + "loss": 0.1225, + "step": 4805 + }, + { + "epoch": 1.0935153583617747, + "grad_norm": 2.6803891273727563, + "learning_rate": 1.1082466521142072e-06, + "loss": 0.0959, + "step": 4806 + }, + { + "epoch": 1.0937428896473265, + "grad_norm": 2.7679540832371603, + "learning_rate": 1.1081899833539592e-06, + "loss": 0.1187, + "step": 4807 + }, + { + "epoch": 1.0939704209328782, + "grad_norm": 1.4189385820383673, + "learning_rate": 1.1081333047181928e-06, + "loss": 0.0551, + "step": 4808 + }, + { + "epoch": 1.09419795221843, + "grad_norm": 2.407210728799108, + "learning_rate": 1.1080766162080664e-06, + "loss": 0.2183, + "step": 4809 + }, + { + "epoch": 1.0944254835039817, + "grad_norm": 2.190691343555159, + "learning_rate": 1.1080199178247388e-06, + "loss": 0.1554, + "step": 4810 + }, + { + "epoch": 1.0946530147895335, + "grad_norm": 1.9158986499607824, + "learning_rate": 1.1079632095693688e-06, + "loss": 0.1014, + "step": 4811 + }, + { + "epoch": 1.0948805460750852, + "grad_norm": 2.1376153598636334, + "learning_rate": 1.107906491443115e-06, + "loss": 0.0867, + "step": 4812 + }, + { + "epoch": 1.0951080773606372, + "grad_norm": 1.640434083574093, + "learning_rate": 1.1078497634471373e-06, + "loss": 0.119, + "step": 4813 + }, + { + "epoch": 1.095335608646189, + "grad_norm": 1.953361284918329, + "learning_rate": 1.1077930255825944e-06, + "loss": 0.0685, + "step": 4814 + }, + { + "epoch": 1.0955631399317407, + "grad_norm": 2.2145384548118496, + "learning_rate": 1.1077362778506464e-06, + "loss": 0.0773, + "step": 4815 + }, + { + "epoch": 1.0957906712172925, + "grad_norm": 1.3681951995706059, + "learning_rate": 1.107679520252453e-06, + "loss": 0.0696, + "step": 4816 + }, + { + "epoch": 1.0960182025028442, + "grad_norm": 1.720391413339496, + "learning_rate": 1.107622752789174e-06, + "loss": 0.0927, + "step": 4817 + }, + { + "epoch": 1.096245733788396, + "grad_norm": 2.535960750420784, + "learning_rate": 1.10756597546197e-06, + "loss": 0.1101, + "step": 4818 + }, + { + "epoch": 1.0964732650739477, + "grad_norm": 1.3722685641226562, + "learning_rate": 1.1075091882720012e-06, + "loss": 0.0326, + "step": 4819 + }, + { + "epoch": 1.0967007963594995, + "grad_norm": 2.7001720504024527, + "learning_rate": 1.1074523912204282e-06, + "loss": 0.0871, + "step": 4820 + }, + { + "epoch": 1.0969283276450512, + "grad_norm": 1.9075168583854094, + "learning_rate": 1.107395584308412e-06, + "loss": 0.1269, + "step": 4821 + }, + { + "epoch": 1.097155858930603, + "grad_norm": 2.205540615959917, + "learning_rate": 1.1073387675371134e-06, + "loss": 0.1026, + "step": 4822 + }, + { + "epoch": 1.0973833902161547, + "grad_norm": 1.0718686815433365, + "learning_rate": 1.1072819409076937e-06, + "loss": 0.0546, + "step": 4823 + }, + { + "epoch": 1.0976109215017065, + "grad_norm": 1.2739063256637186, + "learning_rate": 1.1072251044213146e-06, + "loss": 0.1489, + "step": 4824 + }, + { + "epoch": 1.0978384527872582, + "grad_norm": 1.6711917421442544, + "learning_rate": 1.1071682580791375e-06, + "loss": 0.1218, + "step": 4825 + }, + { + "epoch": 1.09806598407281, + "grad_norm": 1.0808000148116539, + "learning_rate": 1.107111401882324e-06, + "loss": 0.0937, + "step": 4826 + }, + { + "epoch": 1.0982935153583617, + "grad_norm": 1.3984069618118502, + "learning_rate": 1.1070545358320367e-06, + "loss": 0.081, + "step": 4827 + }, + { + "epoch": 1.0985210466439135, + "grad_norm": 1.6817172368352458, + "learning_rate": 1.1069976599294374e-06, + "loss": 0.0446, + "step": 4828 + }, + { + "epoch": 1.0987485779294652, + "grad_norm": 2.9953765468149873, + "learning_rate": 1.1069407741756884e-06, + "loss": 0.121, + "step": 4829 + }, + { + "epoch": 1.098976109215017, + "grad_norm": 1.5977627847753106, + "learning_rate": 1.106883878571953e-06, + "loss": 0.1372, + "step": 4830 + }, + { + "epoch": 1.0992036405005687, + "grad_norm": 1.473455640746034, + "learning_rate": 1.1068269731193936e-06, + "loss": 0.0628, + "step": 4831 + }, + { + "epoch": 1.0994311717861205, + "grad_norm": 2.574413671871508, + "learning_rate": 1.106770057819173e-06, + "loss": 0.0773, + "step": 4832 + }, + { + "epoch": 1.0996587030716722, + "grad_norm": 1.7305196200030186, + "learning_rate": 1.1067131326724551e-06, + "loss": 0.1119, + "step": 4833 + }, + { + "epoch": 1.0998862343572242, + "grad_norm": 1.9296560592971308, + "learning_rate": 1.106656197680403e-06, + "loss": 0.0751, + "step": 4834 + }, + { + "epoch": 1.100113765642776, + "grad_norm": 1.2661901992123141, + "learning_rate": 1.10659925284418e-06, + "loss": 0.1527, + "step": 4835 + }, + { + "epoch": 1.1003412969283277, + "grad_norm": 1.4346695408957206, + "learning_rate": 1.1065422981649506e-06, + "loss": 0.0821, + "step": 4836 + }, + { + "epoch": 1.1005688282138795, + "grad_norm": 1.522864316512361, + "learning_rate": 1.1064853336438782e-06, + "loss": 0.1456, + "step": 4837 + }, + { + "epoch": 1.1007963594994312, + "grad_norm": 1.674140320495389, + "learning_rate": 1.1064283592821276e-06, + "loss": 0.0824, + "step": 4838 + }, + { + "epoch": 1.101023890784983, + "grad_norm": 1.8199138679313875, + "learning_rate": 1.106371375080863e-06, + "loss": 0.0821, + "step": 4839 + }, + { + "epoch": 1.1012514220705347, + "grad_norm": 1.0988975008672435, + "learning_rate": 1.106314381041249e-06, + "loss": 0.0747, + "step": 4840 + }, + { + "epoch": 1.1014789533560865, + "grad_norm": 2.2473800940435344, + "learning_rate": 1.1062573771644506e-06, + "loss": 0.1135, + "step": 4841 + }, + { + "epoch": 1.1017064846416382, + "grad_norm": 2.173377398488759, + "learning_rate": 1.106200363451633e-06, + "loss": 0.1225, + "step": 4842 + }, + { + "epoch": 1.10193401592719, + "grad_norm": 1.0237409844466847, + "learning_rate": 1.1061433399039608e-06, + "loss": 0.0913, + "step": 4843 + }, + { + "epoch": 1.1021615472127417, + "grad_norm": 1.4883802918021127, + "learning_rate": 1.1060863065226002e-06, + "loss": 0.0718, + "step": 4844 + }, + { + "epoch": 1.1023890784982935, + "grad_norm": 2.0993117460282273, + "learning_rate": 1.1060292633087167e-06, + "loss": 0.1518, + "step": 4845 + }, + { + "epoch": 1.1026166097838452, + "grad_norm": 2.228132250689892, + "learning_rate": 1.1059722102634756e-06, + "loss": 0.092, + "step": 4846 + }, + { + "epoch": 1.102844141069397, + "grad_norm": 2.3453851765604954, + "learning_rate": 1.1059151473880439e-06, + "loss": 0.2292, + "step": 4847 + }, + { + "epoch": 1.1030716723549487, + "grad_norm": 2.1780310675452053, + "learning_rate": 1.105858074683587e-06, + "loss": 0.0824, + "step": 4848 + }, + { + "epoch": 1.1032992036405005, + "grad_norm": 2.3050276311535005, + "learning_rate": 1.1058009921512717e-06, + "loss": 0.0875, + "step": 4849 + }, + { + "epoch": 1.1035267349260522, + "grad_norm": 1.9644576717502578, + "learning_rate": 1.1057438997922648e-06, + "loss": 0.0938, + "step": 4850 + }, + { + "epoch": 1.103754266211604, + "grad_norm": 1.910478387408807, + "learning_rate": 1.105686797607733e-06, + "loss": 0.1052, + "step": 4851 + }, + { + "epoch": 1.103981797497156, + "grad_norm": 1.4309370002920787, + "learning_rate": 1.1056296855988432e-06, + "loss": 0.0575, + "step": 4852 + }, + { + "epoch": 1.1042093287827077, + "grad_norm": 1.4145434776922763, + "learning_rate": 1.105572563766763e-06, + "loss": 0.1068, + "step": 4853 + }, + { + "epoch": 1.1044368600682595, + "grad_norm": 1.6308808817252087, + "learning_rate": 1.1055154321126597e-06, + "loss": 0.1075, + "step": 4854 + }, + { + "epoch": 1.1046643913538112, + "grad_norm": 1.261091948999157, + "learning_rate": 1.105458290637701e-06, + "loss": 0.0849, + "step": 4855 + }, + { + "epoch": 1.104891922639363, + "grad_norm": 1.3215074790429198, + "learning_rate": 1.105401139343055e-06, + "loss": 0.0793, + "step": 4856 + }, + { + "epoch": 1.1051194539249147, + "grad_norm": 1.6985162376269767, + "learning_rate": 1.105343978229889e-06, + "loss": 0.1139, + "step": 4857 + }, + { + "epoch": 1.1053469852104665, + "grad_norm": 2.510127015289047, + "learning_rate": 1.1052868072993723e-06, + "loss": 0.1154, + "step": 4858 + }, + { + "epoch": 1.1055745164960182, + "grad_norm": 1.4859308252397974, + "learning_rate": 1.1052296265526726e-06, + "loss": 0.0858, + "step": 4859 + }, + { + "epoch": 1.10580204778157, + "grad_norm": 1.5618239549892472, + "learning_rate": 1.105172435990959e-06, + "loss": 0.1463, + "step": 4860 + }, + { + "epoch": 1.1060295790671217, + "grad_norm": 2.5953777991374887, + "learning_rate": 1.1051152356154e-06, + "loss": 0.1413, + "step": 4861 + }, + { + "epoch": 1.1062571103526735, + "grad_norm": 1.7585866088824806, + "learning_rate": 1.105058025427165e-06, + "loss": 0.0762, + "step": 4862 + }, + { + "epoch": 1.1064846416382252, + "grad_norm": 1.480281395602001, + "learning_rate": 1.105000805427423e-06, + "loss": 0.0773, + "step": 4863 + }, + { + "epoch": 1.106712172923777, + "grad_norm": 1.0381473831114791, + "learning_rate": 1.1049435756173439e-06, + "loss": 0.0798, + "step": 4864 + }, + { + "epoch": 1.1069397042093287, + "grad_norm": 1.1214002462532684, + "learning_rate": 1.104886335998097e-06, + "loss": 0.0458, + "step": 4865 + }, + { + "epoch": 1.1071672354948805, + "grad_norm": 1.3717563405461803, + "learning_rate": 1.104829086570852e-06, + "loss": 0.0844, + "step": 4866 + }, + { + "epoch": 1.1073947667804322, + "grad_norm": 1.5047887960440787, + "learning_rate": 1.1047718273367794e-06, + "loss": 0.0806, + "step": 4867 + }, + { + "epoch": 1.107622298065984, + "grad_norm": 1.3342232875764293, + "learning_rate": 1.1047145582970494e-06, + "loss": 0.0361, + "step": 4868 + }, + { + "epoch": 1.1078498293515358, + "grad_norm": 1.3652330753382278, + "learning_rate": 1.1046572794528324e-06, + "loss": 0.1085, + "step": 4869 + }, + { + "epoch": 1.1080773606370875, + "grad_norm": 1.3247715704309602, + "learning_rate": 1.104599990805299e-06, + "loss": 0.0697, + "step": 4870 + }, + { + "epoch": 1.1083048919226393, + "grad_norm": 1.8015325943851643, + "learning_rate": 1.1045426923556198e-06, + "loss": 0.0804, + "step": 4871 + }, + { + "epoch": 1.108532423208191, + "grad_norm": 1.7563135798847476, + "learning_rate": 1.1044853841049668e-06, + "loss": 0.0822, + "step": 4872 + }, + { + "epoch": 1.108759954493743, + "grad_norm": 0.8710308246253357, + "learning_rate": 1.1044280660545103e-06, + "loss": 0.0206, + "step": 4873 + }, + { + "epoch": 1.1089874857792947, + "grad_norm": 2.005582754489557, + "learning_rate": 1.1043707382054223e-06, + "loss": 0.0684, + "step": 4874 + }, + { + "epoch": 1.1092150170648465, + "grad_norm": 0.869109478461898, + "learning_rate": 1.1043134005588743e-06, + "loss": 0.0478, + "step": 4875 + }, + { + "epoch": 1.1094425483503982, + "grad_norm": 1.298794429145746, + "learning_rate": 1.1042560531160381e-06, + "loss": 0.105, + "step": 4876 + }, + { + "epoch": 1.10967007963595, + "grad_norm": 2.4479802256736893, + "learning_rate": 1.104198695878086e-06, + "loss": 0.0964, + "step": 4877 + }, + { + "epoch": 1.1098976109215017, + "grad_norm": 1.2851994456273843, + "learning_rate": 1.1041413288461903e-06, + "loss": 0.0592, + "step": 4878 + }, + { + "epoch": 1.1101251422070535, + "grad_norm": 2.100558559057352, + "learning_rate": 1.1040839520215233e-06, + "loss": 0.0617, + "step": 4879 + }, + { + "epoch": 1.1103526734926052, + "grad_norm": 1.160798247725018, + "learning_rate": 1.1040265654052575e-06, + "loss": 0.0901, + "step": 4880 + }, + { + "epoch": 1.110580204778157, + "grad_norm": 1.120661149863947, + "learning_rate": 1.103969168998566e-06, + "loss": 0.121, + "step": 4881 + }, + { + "epoch": 1.1108077360637088, + "grad_norm": 4.950864682787037, + "learning_rate": 1.1039117628026222e-06, + "loss": 0.1655, + "step": 4882 + }, + { + "epoch": 1.1110352673492605, + "grad_norm": 1.364699813740917, + "learning_rate": 1.1038543468185988e-06, + "loss": 0.0854, + "step": 4883 + }, + { + "epoch": 1.1112627986348123, + "grad_norm": 1.1250871225008297, + "learning_rate": 1.1037969210476696e-06, + "loss": 0.1017, + "step": 4884 + }, + { + "epoch": 1.111490329920364, + "grad_norm": 2.983700929393962, + "learning_rate": 1.1037394854910082e-06, + "loss": 0.1118, + "step": 4885 + }, + { + "epoch": 1.1117178612059158, + "grad_norm": 1.8781961530183755, + "learning_rate": 1.1036820401497884e-06, + "loss": 0.0519, + "step": 4886 + }, + { + "epoch": 1.1119453924914675, + "grad_norm": 2.069070944004058, + "learning_rate": 1.1036245850251844e-06, + "loss": 0.0479, + "step": 4887 + }, + { + "epoch": 1.1121729237770193, + "grad_norm": 1.712184545237977, + "learning_rate": 1.1035671201183706e-06, + "loss": 0.0939, + "step": 4888 + }, + { + "epoch": 1.112400455062571, + "grad_norm": 2.189959357572597, + "learning_rate": 1.1035096454305213e-06, + "loss": 0.151, + "step": 4889 + }, + { + "epoch": 1.1126279863481228, + "grad_norm": 1.3078629490348257, + "learning_rate": 1.103452160962811e-06, + "loss": 0.0482, + "step": 4890 + }, + { + "epoch": 1.1128555176336747, + "grad_norm": 2.860153538637598, + "learning_rate": 1.103394666716415e-06, + "loss": 0.1244, + "step": 4891 + }, + { + "epoch": 1.1130830489192265, + "grad_norm": 1.722875065107273, + "learning_rate": 1.1033371626925079e-06, + "loss": 0.0584, + "step": 4892 + }, + { + "epoch": 1.1133105802047782, + "grad_norm": 1.801918674856795, + "learning_rate": 1.1032796488922653e-06, + "loss": 0.0536, + "step": 4893 + }, + { + "epoch": 1.11353811149033, + "grad_norm": 1.779178738655481, + "learning_rate": 1.103222125316863e-06, + "loss": 0.0569, + "step": 4894 + }, + { + "epoch": 1.1137656427758817, + "grad_norm": 2.4375614426254204, + "learning_rate": 1.1031645919674758e-06, + "loss": 0.1431, + "step": 4895 + }, + { + "epoch": 1.1139931740614335, + "grad_norm": 0.9387408943236797, + "learning_rate": 1.10310704884528e-06, + "loss": 0.0793, + "step": 4896 + }, + { + "epoch": 1.1142207053469853, + "grad_norm": 1.7562606828790148, + "learning_rate": 1.1030494959514521e-06, + "loss": 0.0955, + "step": 4897 + }, + { + "epoch": 1.114448236632537, + "grad_norm": 1.6211572762717632, + "learning_rate": 1.1029919332871678e-06, + "loss": 0.1339, + "step": 4898 + }, + { + "epoch": 1.1146757679180888, + "grad_norm": 1.682319009818202, + "learning_rate": 1.102934360853604e-06, + "loss": 0.092, + "step": 4899 + }, + { + "epoch": 1.1149032992036405, + "grad_norm": 1.2468699954267968, + "learning_rate": 1.1028767786519368e-06, + "loss": 0.0494, + "step": 4900 + }, + { + "epoch": 1.1151308304891923, + "grad_norm": 1.152573256589869, + "learning_rate": 1.1028191866833438e-06, + "loss": 0.0447, + "step": 4901 + }, + { + "epoch": 1.115358361774744, + "grad_norm": 1.2871716357390797, + "learning_rate": 1.1027615849490014e-06, + "loss": 0.0545, + "step": 4902 + }, + { + "epoch": 1.1155858930602958, + "grad_norm": 2.0888356191297532, + "learning_rate": 1.1027039734500872e-06, + "loss": 0.1307, + "step": 4903 + }, + { + "epoch": 1.1158134243458475, + "grad_norm": 1.5339979897809541, + "learning_rate": 1.102646352187779e-06, + "loss": 0.1462, + "step": 4904 + }, + { + "epoch": 1.1160409556313993, + "grad_norm": 1.6652585638651645, + "learning_rate": 1.1025887211632538e-06, + "loss": 0.0727, + "step": 4905 + }, + { + "epoch": 1.116268486916951, + "grad_norm": 1.585173064754058, + "learning_rate": 1.1025310803776898e-06, + "loss": 0.1625, + "step": 4906 + }, + { + "epoch": 1.1164960182025028, + "grad_norm": 1.888781057343968, + "learning_rate": 1.1024734298322655e-06, + "loss": 0.1054, + "step": 4907 + }, + { + "epoch": 1.1167235494880545, + "grad_norm": 1.425278772221824, + "learning_rate": 1.1024157695281582e-06, + "loss": 0.119, + "step": 4908 + }, + { + "epoch": 1.1169510807736063, + "grad_norm": 1.8240200320037065, + "learning_rate": 1.1023580994665472e-06, + "loss": 0.0669, + "step": 4909 + }, + { + "epoch": 1.117178612059158, + "grad_norm": 1.5457442923166032, + "learning_rate": 1.1023004196486108e-06, + "loss": 0.1373, + "step": 4910 + }, + { + "epoch": 1.11740614334471, + "grad_norm": 2.094866405122345, + "learning_rate": 1.102242730075528e-06, + "loss": 0.1202, + "step": 4911 + }, + { + "epoch": 1.1176336746302618, + "grad_norm": 2.445145034318404, + "learning_rate": 1.1021850307484776e-06, + "loss": 0.133, + "step": 4912 + }, + { + "epoch": 1.1178612059158135, + "grad_norm": 1.7994599245951428, + "learning_rate": 1.1021273216686397e-06, + "loss": 0.0912, + "step": 4913 + }, + { + "epoch": 1.1180887372013653, + "grad_norm": 3.0642622839592706, + "learning_rate": 1.1020696028371926e-06, + "loss": 0.1026, + "step": 4914 + }, + { + "epoch": 1.118316268486917, + "grad_norm": 1.5494532571576725, + "learning_rate": 1.1020118742553166e-06, + "loss": 0.1116, + "step": 4915 + }, + { + "epoch": 1.1185437997724688, + "grad_norm": 2.504252666932219, + "learning_rate": 1.1019541359241917e-06, + "loss": 0.0939, + "step": 4916 + }, + { + "epoch": 1.1187713310580205, + "grad_norm": 2.331022199171125, + "learning_rate": 1.1018963878449976e-06, + "loss": 0.0996, + "step": 4917 + }, + { + "epoch": 1.1189988623435723, + "grad_norm": 1.7064165985683355, + "learning_rate": 1.1018386300189148e-06, + "loss": 0.0452, + "step": 4918 + }, + { + "epoch": 1.119226393629124, + "grad_norm": 1.495724359531087, + "learning_rate": 1.1017808624471237e-06, + "loss": 0.0973, + "step": 4919 + }, + { + "epoch": 1.1194539249146758, + "grad_norm": 1.3478931502531566, + "learning_rate": 1.101723085130805e-06, + "loss": 0.0929, + "step": 4920 + }, + { + "epoch": 1.1196814562002275, + "grad_norm": 1.8942149767063106, + "learning_rate": 1.1016652980711392e-06, + "loss": 0.0939, + "step": 4921 + }, + { + "epoch": 1.1199089874857793, + "grad_norm": 1.1254758558845417, + "learning_rate": 1.1016075012693082e-06, + "loss": 0.0788, + "step": 4922 + }, + { + "epoch": 1.120136518771331, + "grad_norm": 1.2756441216407504, + "learning_rate": 1.1015496947264923e-06, + "loss": 0.0834, + "step": 4923 + }, + { + "epoch": 1.1203640500568828, + "grad_norm": 2.911374051671654, + "learning_rate": 1.1014918784438736e-06, + "loss": 0.0624, + "step": 4924 + }, + { + "epoch": 1.1205915813424345, + "grad_norm": 1.3868160433273433, + "learning_rate": 1.1014340524226337e-06, + "loss": 0.1127, + "step": 4925 + }, + { + "epoch": 1.1208191126279863, + "grad_norm": 1.3222460616762193, + "learning_rate": 1.101376216663954e-06, + "loss": 0.0991, + "step": 4926 + }, + { + "epoch": 1.121046643913538, + "grad_norm": 1.8165175156132298, + "learning_rate": 1.101318371169017e-06, + "loss": 0.204, + "step": 4927 + }, + { + "epoch": 1.1212741751990898, + "grad_norm": 1.1690083635578878, + "learning_rate": 1.1012605159390048e-06, + "loss": 0.0939, + "step": 4928 + }, + { + "epoch": 1.1215017064846415, + "grad_norm": 1.5839171650253496, + "learning_rate": 1.1012026509751e-06, + "loss": 0.0811, + "step": 4929 + }, + { + "epoch": 1.1217292377701935, + "grad_norm": 1.3688530836003974, + "learning_rate": 1.1011447762784849e-06, + "loss": 0.0628, + "step": 4930 + }, + { + "epoch": 1.1219567690557453, + "grad_norm": 1.682646501606769, + "learning_rate": 1.1010868918503429e-06, + "loss": 0.1209, + "step": 4931 + }, + { + "epoch": 1.122184300341297, + "grad_norm": 1.1263472398197065, + "learning_rate": 1.1010289976918565e-06, + "loss": 0.0512, + "step": 4932 + }, + { + "epoch": 1.1224118316268488, + "grad_norm": 1.8786745580953335, + "learning_rate": 1.1009710938042093e-06, + "loss": 0.0829, + "step": 4933 + }, + { + "epoch": 1.1226393629124005, + "grad_norm": 1.7691231961554044, + "learning_rate": 1.1009131801885848e-06, + "loss": 0.1705, + "step": 4934 + }, + { + "epoch": 1.1228668941979523, + "grad_norm": 1.2152009575549836, + "learning_rate": 1.100855256846166e-06, + "loss": 0.0517, + "step": 4935 + }, + { + "epoch": 1.123094425483504, + "grad_norm": 1.4082011663537433, + "learning_rate": 1.1007973237781377e-06, + "loss": 0.0915, + "step": 4936 + }, + { + "epoch": 1.1233219567690558, + "grad_norm": 1.7851853742678743, + "learning_rate": 1.1007393809856834e-06, + "loss": 0.0666, + "step": 4937 + }, + { + "epoch": 1.1235494880546075, + "grad_norm": 1.3679603490528933, + "learning_rate": 1.1006814284699873e-06, + "loss": 0.187, + "step": 4938 + }, + { + "epoch": 1.1237770193401593, + "grad_norm": 1.0869679752100778, + "learning_rate": 1.100623466232234e-06, + "loss": 0.0979, + "step": 4939 + }, + { + "epoch": 1.124004550625711, + "grad_norm": 1.8324197313337576, + "learning_rate": 1.1005654942736082e-06, + "loss": 0.0991, + "step": 4940 + }, + { + "epoch": 1.1242320819112628, + "grad_norm": 0.8916797514793634, + "learning_rate": 1.1005075125952946e-06, + "loss": 0.0449, + "step": 4941 + }, + { + "epoch": 1.1244596131968145, + "grad_norm": 1.4439535008292297, + "learning_rate": 1.1004495211984783e-06, + "loss": 0.1386, + "step": 4942 + }, + { + "epoch": 1.1246871444823663, + "grad_norm": 1.7320674419620372, + "learning_rate": 1.1003915200843446e-06, + "loss": 0.1045, + "step": 4943 + }, + { + "epoch": 1.124914675767918, + "grad_norm": 2.086445312550206, + "learning_rate": 1.1003335092540787e-06, + "loss": 0.0756, + "step": 4944 + }, + { + "epoch": 1.1251422070534698, + "grad_norm": 1.188570368927323, + "learning_rate": 1.1002754887088665e-06, + "loss": 0.0578, + "step": 4945 + }, + { + "epoch": 1.1253697383390215, + "grad_norm": 1.7752270711874487, + "learning_rate": 1.1002174584498938e-06, + "loss": 0.0948, + "step": 4946 + }, + { + "epoch": 1.1255972696245733, + "grad_norm": 2.3256358668512687, + "learning_rate": 1.1001594184783464e-06, + "loss": 0.1521, + "step": 4947 + }, + { + "epoch": 1.125824800910125, + "grad_norm": 1.5836370395195216, + "learning_rate": 1.1001013687954109e-06, + "loss": 0.0716, + "step": 4948 + }, + { + "epoch": 1.1260523321956768, + "grad_norm": 0.8926228009452495, + "learning_rate": 1.1000433094022735e-06, + "loss": 0.0785, + "step": 4949 + }, + { + "epoch": 1.1262798634812285, + "grad_norm": 1.2143047544290266, + "learning_rate": 1.0999852403001208e-06, + "loss": 0.0755, + "step": 4950 + }, + { + "epoch": 1.1265073947667805, + "grad_norm": 1.6120702109468077, + "learning_rate": 1.0999271614901396e-06, + "loss": 0.079, + "step": 4951 + }, + { + "epoch": 1.1267349260523323, + "grad_norm": 1.0700572670291029, + "learning_rate": 1.099869072973517e-06, + "loss": 0.0459, + "step": 4952 + }, + { + "epoch": 1.126962457337884, + "grad_norm": 1.406661725254882, + "learning_rate": 1.0998109747514404e-06, + "loss": 0.1368, + "step": 4953 + }, + { + "epoch": 1.1271899886234358, + "grad_norm": 1.8428127725262113, + "learning_rate": 1.099752866825097e-06, + "loss": 0.0559, + "step": 4954 + }, + { + "epoch": 1.1274175199089875, + "grad_norm": 1.6669400062763435, + "learning_rate": 1.0996947491956745e-06, + "loss": 0.0721, + "step": 4955 + }, + { + "epoch": 1.1276450511945393, + "grad_norm": 1.563884717527578, + "learning_rate": 1.0996366218643607e-06, + "loss": 0.0622, + "step": 4956 + }, + { + "epoch": 1.127872582480091, + "grad_norm": 1.8893173163715162, + "learning_rate": 1.0995784848323434e-06, + "loss": 0.0927, + "step": 4957 + }, + { + "epoch": 1.1281001137656428, + "grad_norm": 1.631214871572094, + "learning_rate": 1.0995203381008112e-06, + "loss": 0.0658, + "step": 4958 + }, + { + "epoch": 1.1283276450511945, + "grad_norm": 2.057566048669985, + "learning_rate": 1.099462181670952e-06, + "loss": 0.1078, + "step": 4959 + }, + { + "epoch": 1.1285551763367463, + "grad_norm": 2.9901712732479093, + "learning_rate": 1.0994040155439553e-06, + "loss": 0.1499, + "step": 4960 + }, + { + "epoch": 1.128782707622298, + "grad_norm": 1.8173197795146472, + "learning_rate": 1.0993458397210092e-06, + "loss": 0.0909, + "step": 4961 + }, + { + "epoch": 1.1290102389078498, + "grad_norm": 1.4406862960819018, + "learning_rate": 1.0992876542033026e-06, + "loss": 0.0679, + "step": 4962 + }, + { + "epoch": 1.1292377701934015, + "grad_norm": 2.0147220621119724, + "learning_rate": 1.0992294589920252e-06, + "loss": 0.0979, + "step": 4963 + }, + { + "epoch": 1.1294653014789533, + "grad_norm": 1.7427158658250603, + "learning_rate": 1.099171254088366e-06, + "loss": 0.0578, + "step": 4964 + }, + { + "epoch": 1.129692832764505, + "grad_norm": 2.3163769351990084, + "learning_rate": 1.0991130394935148e-06, + "loss": 0.0706, + "step": 4965 + }, + { + "epoch": 1.1299203640500568, + "grad_norm": 2.9626845696330824, + "learning_rate": 1.0990548152086616e-06, + "loss": 0.1327, + "step": 4966 + }, + { + "epoch": 1.1301478953356086, + "grad_norm": 1.4455808915658892, + "learning_rate": 1.098996581234996e-06, + "loss": 0.0883, + "step": 4967 + }, + { + "epoch": 1.1303754266211605, + "grad_norm": 1.6512236234603601, + "learning_rate": 1.0989383375737081e-06, + "loss": 0.0543, + "step": 4968 + }, + { + "epoch": 1.1306029579067123, + "grad_norm": 1.533650029665724, + "learning_rate": 1.098880084225989e-06, + "loss": 0.0964, + "step": 4969 + }, + { + "epoch": 1.130830489192264, + "grad_norm": 1.604433883847564, + "learning_rate": 1.0988218211930285e-06, + "loss": 0.0791, + "step": 4970 + }, + { + "epoch": 1.1310580204778158, + "grad_norm": 1.7409770933966728, + "learning_rate": 1.0987635484760178e-06, + "loss": 0.13, + "step": 4971 + }, + { + "epoch": 1.1312855517633675, + "grad_norm": 2.7884517331400325, + "learning_rate": 1.098705266076148e-06, + "loss": 0.0671, + "step": 4972 + }, + { + "epoch": 1.1315130830489193, + "grad_norm": 1.630625307393665, + "learning_rate": 1.0986469739946102e-06, + "loss": 0.0855, + "step": 4973 + }, + { + "epoch": 1.131740614334471, + "grad_norm": 1.2094350075349947, + "learning_rate": 1.0985886722325954e-06, + "loss": 0.0368, + "step": 4974 + }, + { + "epoch": 1.1319681456200228, + "grad_norm": 1.3111484263137263, + "learning_rate": 1.0985303607912956e-06, + "loss": 0.0511, + "step": 4975 + }, + { + "epoch": 1.1321956769055745, + "grad_norm": 2.657021119411178, + "learning_rate": 1.0984720396719024e-06, + "loss": 0.1313, + "step": 4976 + }, + { + "epoch": 1.1324232081911263, + "grad_norm": 1.891785343782106, + "learning_rate": 1.098413708875608e-06, + "loss": 0.0948, + "step": 4977 + }, + { + "epoch": 1.132650739476678, + "grad_norm": 1.5850598764256127, + "learning_rate": 1.098355368403604e-06, + "loss": 0.0744, + "step": 4978 + }, + { + "epoch": 1.1328782707622298, + "grad_norm": 1.8115705441515912, + "learning_rate": 1.0982970182570837e-06, + "loss": 0.0817, + "step": 4979 + }, + { + "epoch": 1.1331058020477816, + "grad_norm": 1.8766599987971402, + "learning_rate": 1.098238658437239e-06, + "loss": 0.0903, + "step": 4980 + }, + { + "epoch": 1.1333333333333333, + "grad_norm": 1.4992947077118386, + "learning_rate": 1.0981802889452627e-06, + "loss": 0.075, + "step": 4981 + }, + { + "epoch": 1.133560864618885, + "grad_norm": 3.497579577636057, + "learning_rate": 1.0981219097823479e-06, + "loss": 0.1198, + "step": 4982 + }, + { + "epoch": 1.1337883959044368, + "grad_norm": 1.7734853138248883, + "learning_rate": 1.0980635209496878e-06, + "loss": 0.0982, + "step": 4983 + }, + { + "epoch": 1.1340159271899886, + "grad_norm": 2.5972880470148616, + "learning_rate": 1.0980051224484756e-06, + "loss": 0.1034, + "step": 4984 + }, + { + "epoch": 1.1342434584755403, + "grad_norm": 1.6807793940071352, + "learning_rate": 1.0979467142799052e-06, + "loss": 0.1123, + "step": 4985 + }, + { + "epoch": 1.134470989761092, + "grad_norm": 1.167512015914663, + "learning_rate": 1.0978882964451698e-06, + "loss": 0.1098, + "step": 4986 + }, + { + "epoch": 1.1346985210466438, + "grad_norm": 1.1705606733035332, + "learning_rate": 1.097829868945464e-06, + "loss": 0.0673, + "step": 4987 + }, + { + "epoch": 1.1349260523321956, + "grad_norm": 2.2907339208973925, + "learning_rate": 1.0977714317819812e-06, + "loss": 0.1176, + "step": 4988 + }, + { + "epoch": 1.1351535836177473, + "grad_norm": 2.121070478697396, + "learning_rate": 1.0977129849559165e-06, + "loss": 0.0845, + "step": 4989 + }, + { + "epoch": 1.1353811149032993, + "grad_norm": 1.806954543632207, + "learning_rate": 1.0976545284684642e-06, + "loss": 0.0599, + "step": 4990 + }, + { + "epoch": 1.135608646188851, + "grad_norm": 2.5598944741988734, + "learning_rate": 1.0975960623208188e-06, + "loss": 0.1371, + "step": 4991 + }, + { + "epoch": 1.1358361774744028, + "grad_norm": 2.0091610638634583, + "learning_rate": 1.0975375865141753e-06, + "loss": 0.0653, + "step": 4992 + }, + { + "epoch": 1.1360637087599545, + "grad_norm": 1.7957636193915902, + "learning_rate": 1.097479101049729e-06, + "loss": 0.0551, + "step": 4993 + }, + { + "epoch": 1.1362912400455063, + "grad_norm": 1.1892795380116103, + "learning_rate": 1.0974206059286752e-06, + "loss": 0.1151, + "step": 4994 + }, + { + "epoch": 1.136518771331058, + "grad_norm": 2.1464705127422543, + "learning_rate": 1.0973621011522096e-06, + "loss": 0.0761, + "step": 4995 + }, + { + "epoch": 1.1367463026166098, + "grad_norm": 1.9966944001252394, + "learning_rate": 1.0973035867215276e-06, + "loss": 0.1168, + "step": 4996 + }, + { + "epoch": 1.1369738339021616, + "grad_norm": 1.020794945372424, + "learning_rate": 1.0972450626378254e-06, + "loss": 0.0485, + "step": 4997 + }, + { + "epoch": 1.1372013651877133, + "grad_norm": 1.64511125117566, + "learning_rate": 1.0971865289022988e-06, + "loss": 0.0694, + "step": 4998 + }, + { + "epoch": 1.137428896473265, + "grad_norm": 1.175913889162089, + "learning_rate": 1.0971279855161442e-06, + "loss": 0.084, + "step": 4999 + }, + { + "epoch": 1.1376564277588168, + "grad_norm": 2.8231213648251714, + "learning_rate": 1.0970694324805586e-06, + "loss": 0.1428, + "step": 5000 + }, + { + "epoch": 1.1378839590443686, + "grad_norm": 2.6932229520099846, + "learning_rate": 1.0970108697967382e-06, + "loss": 0.1014, + "step": 5001 + }, + { + "epoch": 1.1381114903299203, + "grad_norm": 1.3526729503758106, + "learning_rate": 1.09695229746588e-06, + "loss": 0.0988, + "step": 5002 + }, + { + "epoch": 1.138339021615472, + "grad_norm": 3.6679630171563193, + "learning_rate": 1.0968937154891812e-06, + "loss": 0.1297, + "step": 5003 + }, + { + "epoch": 1.1385665529010238, + "grad_norm": 2.0787582994902976, + "learning_rate": 1.096835123867839e-06, + "loss": 0.0786, + "step": 5004 + }, + { + "epoch": 1.1387940841865756, + "grad_norm": 2.0738995997206207, + "learning_rate": 1.0967765226030512e-06, + "loss": 0.096, + "step": 5005 + }, + { + "epoch": 1.1390216154721273, + "grad_norm": 1.792981831800975, + "learning_rate": 1.0967179116960153e-06, + "loss": 0.108, + "step": 5006 + }, + { + "epoch": 1.1392491467576793, + "grad_norm": 1.7850160079944348, + "learning_rate": 1.096659291147929e-06, + "loss": 0.1949, + "step": 5007 + }, + { + "epoch": 1.139476678043231, + "grad_norm": 1.4390443615551238, + "learning_rate": 1.0966006609599908e-06, + "loss": 0.05, + "step": 5008 + }, + { + "epoch": 1.1397042093287828, + "grad_norm": 2.2005543377996886, + "learning_rate": 1.0965420211333984e-06, + "loss": 0.0938, + "step": 5009 + }, + { + "epoch": 1.1399317406143346, + "grad_norm": 1.3374633377814356, + "learning_rate": 1.0964833716693512e-06, + "loss": 0.1373, + "step": 5010 + }, + { + "epoch": 1.1401592718998863, + "grad_norm": 2.1028444721908586, + "learning_rate": 1.096424712569047e-06, + "loss": 0.0684, + "step": 5011 + }, + { + "epoch": 1.140386803185438, + "grad_norm": 1.7106971142516763, + "learning_rate": 1.0963660438336851e-06, + "loss": 0.1593, + "step": 5012 + }, + { + "epoch": 1.1406143344709898, + "grad_norm": 1.2864010761726647, + "learning_rate": 1.0963073654644645e-06, + "loss": 0.0538, + "step": 5013 + }, + { + "epoch": 1.1408418657565416, + "grad_norm": 1.3129074175693134, + "learning_rate": 1.0962486774625847e-06, + "loss": 0.0949, + "step": 5014 + }, + { + "epoch": 1.1410693970420933, + "grad_norm": 1.4503139865982977, + "learning_rate": 1.096189979829245e-06, + "loss": 0.0662, + "step": 5015 + }, + { + "epoch": 1.141296928327645, + "grad_norm": 1.3609039708489337, + "learning_rate": 1.096131272565645e-06, + "loss": 0.0492, + "step": 5016 + }, + { + "epoch": 1.1415244596131968, + "grad_norm": 1.5885653226656653, + "learning_rate": 1.0960725556729845e-06, + "loss": 0.047, + "step": 5017 + }, + { + "epoch": 1.1417519908987486, + "grad_norm": 1.4277301607595434, + "learning_rate": 1.0960138291524637e-06, + "loss": 0.0681, + "step": 5018 + }, + { + "epoch": 1.1419795221843003, + "grad_norm": 1.7749885462374024, + "learning_rate": 1.095955093005283e-06, + "loss": 0.1022, + "step": 5019 + }, + { + "epoch": 1.142207053469852, + "grad_norm": 1.9945055261306845, + "learning_rate": 1.0958963472326426e-06, + "loss": 0.0992, + "step": 5020 + }, + { + "epoch": 1.1424345847554038, + "grad_norm": 1.7159767521687925, + "learning_rate": 1.0958375918357433e-06, + "loss": 0.0699, + "step": 5021 + }, + { + "epoch": 1.1426621160409556, + "grad_norm": 0.9252880714466132, + "learning_rate": 1.095778826815786e-06, + "loss": 0.0426, + "step": 5022 + }, + { + "epoch": 1.1428896473265073, + "grad_norm": 2.3010291974952604, + "learning_rate": 1.0957200521739715e-06, + "loss": 0.1163, + "step": 5023 + }, + { + "epoch": 1.143117178612059, + "grad_norm": 1.7201270965579583, + "learning_rate": 1.0956612679115012e-06, + "loss": 0.174, + "step": 5024 + }, + { + "epoch": 1.1433447098976108, + "grad_norm": 1.2170020491679054, + "learning_rate": 1.0956024740295767e-06, + "loss": 0.1368, + "step": 5025 + }, + { + "epoch": 1.1435722411831626, + "grad_norm": 1.6673945996132118, + "learning_rate": 1.0955436705293996e-06, + "loss": 0.1039, + "step": 5026 + }, + { + "epoch": 1.1437997724687143, + "grad_norm": 1.4526736263426159, + "learning_rate": 1.0954848574121715e-06, + "loss": 0.148, + "step": 5027 + }, + { + "epoch": 1.144027303754266, + "grad_norm": 1.000929358914613, + "learning_rate": 1.0954260346790944e-06, + "loss": 0.063, + "step": 5028 + }, + { + "epoch": 1.144254835039818, + "grad_norm": 2.004003133169068, + "learning_rate": 1.0953672023313709e-06, + "loss": 0.1182, + "step": 5029 + }, + { + "epoch": 1.1444823663253698, + "grad_norm": 2.8747884612525074, + "learning_rate": 1.0953083603702031e-06, + "loss": 0.0799, + "step": 5030 + }, + { + "epoch": 1.1447098976109216, + "grad_norm": 2.616654201453664, + "learning_rate": 1.0952495087967939e-06, + "loss": 0.0952, + "step": 5031 + }, + { + "epoch": 1.1449374288964733, + "grad_norm": 1.1926571603629328, + "learning_rate": 1.095190647612346e-06, + "loss": 0.1146, + "step": 5032 + }, + { + "epoch": 1.145164960182025, + "grad_norm": 2.012002754656629, + "learning_rate": 1.0951317768180623e-06, + "loss": 0.0832, + "step": 5033 + }, + { + "epoch": 1.1453924914675768, + "grad_norm": 1.7498100999489232, + "learning_rate": 1.0950728964151457e-06, + "loss": 0.1023, + "step": 5034 + }, + { + "epoch": 1.1456200227531286, + "grad_norm": 1.1549230270434567, + "learning_rate": 1.0950140064048005e-06, + "loss": 0.0307, + "step": 5035 + }, + { + "epoch": 1.1458475540386803, + "grad_norm": 1.3442692312701563, + "learning_rate": 1.0949551067882297e-06, + "loss": 0.0996, + "step": 5036 + }, + { + "epoch": 1.146075085324232, + "grad_norm": 1.2566411523687606, + "learning_rate": 1.094896197566637e-06, + "loss": 0.0358, + "step": 5037 + }, + { + "epoch": 1.1463026166097838, + "grad_norm": 1.5156306393603887, + "learning_rate": 1.0948372787412267e-06, + "loss": 0.0861, + "step": 5038 + }, + { + "epoch": 1.1465301478953356, + "grad_norm": 1.8535248219358293, + "learning_rate": 1.094778350313203e-06, + "loss": 0.0702, + "step": 5039 + }, + { + "epoch": 1.1467576791808873, + "grad_norm": 1.2766493100453777, + "learning_rate": 1.09471941228377e-06, + "loss": 0.0863, + "step": 5040 + }, + { + "epoch": 1.146985210466439, + "grad_norm": 1.798776404954717, + "learning_rate": 1.0946604646541327e-06, + "loss": 0.1255, + "step": 5041 + }, + { + "epoch": 1.1472127417519908, + "grad_norm": 1.5012306399931543, + "learning_rate": 1.0946015074254957e-06, + "loss": 0.0504, + "step": 5042 + }, + { + "epoch": 1.1474402730375426, + "grad_norm": 1.7885741561851027, + "learning_rate": 1.0945425405990636e-06, + "loss": 0.1188, + "step": 5043 + }, + { + "epoch": 1.1476678043230943, + "grad_norm": 1.830862543786872, + "learning_rate": 1.094483564176042e-06, + "loss": 0.151, + "step": 5044 + }, + { + "epoch": 1.147895335608646, + "grad_norm": 2.6040357792438322, + "learning_rate": 1.0944245781576363e-06, + "loss": 0.1065, + "step": 5045 + }, + { + "epoch": 1.148122866894198, + "grad_norm": 2.0581342138073175, + "learning_rate": 1.0943655825450517e-06, + "loss": 0.0851, + "step": 5046 + }, + { + "epoch": 1.1483503981797498, + "grad_norm": 2.143603515071713, + "learning_rate": 1.0943065773394943e-06, + "loss": 0.137, + "step": 5047 + }, + { + "epoch": 1.1485779294653016, + "grad_norm": 1.352482485940918, + "learning_rate": 1.0942475625421701e-06, + "loss": 0.0726, + "step": 5048 + }, + { + "epoch": 1.1488054607508533, + "grad_norm": 1.6964232784625224, + "learning_rate": 1.094188538154285e-06, + "loss": 0.0994, + "step": 5049 + }, + { + "epoch": 1.149032992036405, + "grad_norm": 2.775135542145203, + "learning_rate": 1.0941295041770453e-06, + "loss": 0.099, + "step": 5050 + }, + { + "epoch": 1.1492605233219568, + "grad_norm": 1.5050400089796847, + "learning_rate": 1.0940704606116578e-06, + "loss": 0.0574, + "step": 5051 + }, + { + "epoch": 1.1494880546075086, + "grad_norm": 2.721899773786621, + "learning_rate": 1.0940114074593292e-06, + "loss": 0.0886, + "step": 5052 + }, + { + "epoch": 1.1497155858930603, + "grad_norm": 1.356790292993199, + "learning_rate": 1.0939523447212665e-06, + "loss": 0.0632, + "step": 5053 + }, + { + "epoch": 1.149943117178612, + "grad_norm": 1.3503374996306448, + "learning_rate": 1.0938932723986766e-06, + "loss": 0.0816, + "step": 5054 + }, + { + "epoch": 1.1501706484641638, + "grad_norm": 1.481659233232767, + "learning_rate": 1.0938341904927669e-06, + "loss": 0.0719, + "step": 5055 + }, + { + "epoch": 1.1503981797497156, + "grad_norm": 0.8694013815410049, + "learning_rate": 1.093775099004745e-06, + "loss": 0.0424, + "step": 5056 + }, + { + "epoch": 1.1506257110352673, + "grad_norm": 1.919365341975163, + "learning_rate": 1.0937159979358186e-06, + "loss": 0.1127, + "step": 5057 + }, + { + "epoch": 1.150853242320819, + "grad_norm": 2.0981902421175773, + "learning_rate": 1.0936568872871958e-06, + "loss": 0.0914, + "step": 5058 + }, + { + "epoch": 1.1510807736063708, + "grad_norm": 1.5675678492995555, + "learning_rate": 1.0935977670600843e-06, + "loss": 0.1131, + "step": 5059 + }, + { + "epoch": 1.1513083048919226, + "grad_norm": 1.9064072576465214, + "learning_rate": 1.0935386372556928e-06, + "loss": 0.0772, + "step": 5060 + }, + { + "epoch": 1.1515358361774743, + "grad_norm": 1.4104845636818022, + "learning_rate": 1.0934794978752295e-06, + "loss": 0.0684, + "step": 5061 + }, + { + "epoch": 1.151763367463026, + "grad_norm": 2.5853584445718676, + "learning_rate": 1.0934203489199033e-06, + "loss": 0.0998, + "step": 5062 + }, + { + "epoch": 1.1519908987485779, + "grad_norm": 1.9121176842785628, + "learning_rate": 1.093361190390923e-06, + "loss": 0.1038, + "step": 5063 + }, + { + "epoch": 1.1522184300341296, + "grad_norm": 2.0243437314386616, + "learning_rate": 1.0933020222894978e-06, + "loss": 0.1361, + "step": 5064 + }, + { + "epoch": 1.1524459613196814, + "grad_norm": 1.7972592627695714, + "learning_rate": 1.0932428446168369e-06, + "loss": 0.0427, + "step": 5065 + }, + { + "epoch": 1.152673492605233, + "grad_norm": 1.6741471355620121, + "learning_rate": 1.0931836573741498e-06, + "loss": 0.1559, + "step": 5066 + }, + { + "epoch": 1.1529010238907849, + "grad_norm": 2.3765148471218462, + "learning_rate": 1.093124460562646e-06, + "loss": 0.0764, + "step": 5067 + }, + { + "epoch": 1.1531285551763368, + "grad_norm": 1.5942557601662923, + "learning_rate": 1.0930652541835357e-06, + "loss": 0.1573, + "step": 5068 + }, + { + "epoch": 1.1533560864618886, + "grad_norm": 3.08185190716079, + "learning_rate": 1.093006038238029e-06, + "loss": 0.141, + "step": 5069 + }, + { + "epoch": 1.1535836177474403, + "grad_norm": 1.414273885302686, + "learning_rate": 1.0929468127273357e-06, + "loss": 0.1006, + "step": 5070 + }, + { + "epoch": 1.153811149032992, + "grad_norm": 2.080210442470953, + "learning_rate": 1.0928875776526667e-06, + "loss": 0.1276, + "step": 5071 + }, + { + "epoch": 1.1540386803185438, + "grad_norm": 1.6679057714706864, + "learning_rate": 1.0928283330152325e-06, + "loss": 0.1001, + "step": 5072 + }, + { + "epoch": 1.1542662116040956, + "grad_norm": 1.4223736080513678, + "learning_rate": 1.092769078816244e-06, + "loss": 0.1408, + "step": 5073 + }, + { + "epoch": 1.1544937428896473, + "grad_norm": 1.8966004842746413, + "learning_rate": 1.092709815056912e-06, + "loss": 0.0953, + "step": 5074 + }, + { + "epoch": 1.154721274175199, + "grad_norm": 1.1682597216102555, + "learning_rate": 1.0926505417384482e-06, + "loss": 0.0513, + "step": 5075 + }, + { + "epoch": 1.1549488054607508, + "grad_norm": 2.2910856841946154, + "learning_rate": 1.0925912588620637e-06, + "loss": 0.0945, + "step": 5076 + }, + { + "epoch": 1.1551763367463026, + "grad_norm": 1.0720514021970478, + "learning_rate": 1.0925319664289703e-06, + "loss": 0.0764, + "step": 5077 + }, + { + "epoch": 1.1554038680318544, + "grad_norm": 1.5743332758446995, + "learning_rate": 1.0924726644403797e-06, + "loss": 0.059, + "step": 5078 + }, + { + "epoch": 1.155631399317406, + "grad_norm": 1.3255429568097516, + "learning_rate": 1.0924133528975039e-06, + "loss": 0.0919, + "step": 5079 + }, + { + "epoch": 1.1558589306029579, + "grad_norm": 1.5332950391438218, + "learning_rate": 1.0923540318015552e-06, + "loss": 0.0913, + "step": 5080 + }, + { + "epoch": 1.1560864618885096, + "grad_norm": 2.6893499621057306, + "learning_rate": 1.092294701153746e-06, + "loss": 0.126, + "step": 5081 + }, + { + "epoch": 1.1563139931740614, + "grad_norm": 0.9148314458809074, + "learning_rate": 1.092235360955289e-06, + "loss": 0.1184, + "step": 5082 + }, + { + "epoch": 1.1565415244596131, + "grad_norm": 1.1437794949654425, + "learning_rate": 1.092176011207397e-06, + "loss": 0.0367, + "step": 5083 + }, + { + "epoch": 1.1567690557451649, + "grad_norm": 1.8172588347866085, + "learning_rate": 1.0921166519112828e-06, + "loss": 0.1471, + "step": 5084 + }, + { + "epoch": 1.1569965870307168, + "grad_norm": 1.1447836349023166, + "learning_rate": 1.0920572830681597e-06, + "loss": 0.0537, + "step": 5085 + }, + { + "epoch": 1.1572241183162686, + "grad_norm": 1.5357107672237025, + "learning_rate": 1.0919979046792411e-06, + "loss": 0.0537, + "step": 5086 + }, + { + "epoch": 1.1574516496018203, + "grad_norm": 1.0688283394230311, + "learning_rate": 1.0919385167457408e-06, + "loss": 0.0354, + "step": 5087 + }, + { + "epoch": 1.157679180887372, + "grad_norm": 2.0639977539582595, + "learning_rate": 1.0918791192688722e-06, + "loss": 0.09, + "step": 5088 + }, + { + "epoch": 1.1579067121729238, + "grad_norm": 1.3518105035209564, + "learning_rate": 1.0918197122498495e-06, + "loss": 0.0798, + "step": 5089 + }, + { + "epoch": 1.1581342434584756, + "grad_norm": 1.5164367559828507, + "learning_rate": 1.0917602956898867e-06, + "loss": 0.0943, + "step": 5090 + }, + { + "epoch": 1.1583617747440274, + "grad_norm": 1.792644442256943, + "learning_rate": 1.0917008695901985e-06, + "loss": 0.0656, + "step": 5091 + }, + { + "epoch": 1.158589306029579, + "grad_norm": 2.1479763014207953, + "learning_rate": 1.091641433951999e-06, + "loss": 0.1433, + "step": 5092 + }, + { + "epoch": 1.1588168373151309, + "grad_norm": 2.053907620353767, + "learning_rate": 1.0915819887765034e-06, + "loss": 0.1345, + "step": 5093 + }, + { + "epoch": 1.1590443686006826, + "grad_norm": 1.2848859703626283, + "learning_rate": 1.0915225340649264e-06, + "loss": 0.0808, + "step": 5094 + }, + { + "epoch": 1.1592718998862344, + "grad_norm": 1.6760046409461768, + "learning_rate": 1.091463069818483e-06, + "loss": 0.0932, + "step": 5095 + }, + { + "epoch": 1.159499431171786, + "grad_norm": 1.9797530016618736, + "learning_rate": 1.091403596038389e-06, + "loss": 0.1296, + "step": 5096 + }, + { + "epoch": 1.1597269624573379, + "grad_norm": 1.372540390789414, + "learning_rate": 1.0913441127258596e-06, + "loss": 0.0416, + "step": 5097 + }, + { + "epoch": 1.1599544937428896, + "grad_norm": 1.770026041242371, + "learning_rate": 1.0912846198821105e-06, + "loss": 0.1109, + "step": 5098 + }, + { + "epoch": 1.1601820250284414, + "grad_norm": 2.3255312369703427, + "learning_rate": 1.091225117508358e-06, + "loss": 0.1967, + "step": 5099 + }, + { + "epoch": 1.1604095563139931, + "grad_norm": 1.922123825453258, + "learning_rate": 1.0911656056058175e-06, + "loss": 0.1104, + "step": 5100 + }, + { + "epoch": 1.1606370875995449, + "grad_norm": 1.5508865361545927, + "learning_rate": 1.0911060841757063e-06, + "loss": 0.1008, + "step": 5101 + }, + { + "epoch": 1.1608646188850966, + "grad_norm": 3.3431646625962825, + "learning_rate": 1.09104655321924e-06, + "loss": 0.1382, + "step": 5102 + }, + { + "epoch": 1.1610921501706484, + "grad_norm": 1.699504224482545, + "learning_rate": 1.0909870127376358e-06, + "loss": 0.1004, + "step": 5103 + }, + { + "epoch": 1.1613196814562001, + "grad_norm": 2.462992773726134, + "learning_rate": 1.0909274627321106e-06, + "loss": 0.1523, + "step": 5104 + }, + { + "epoch": 1.1615472127417519, + "grad_norm": 1.4732070097539725, + "learning_rate": 1.090867903203881e-06, + "loss": 0.0655, + "step": 5105 + }, + { + "epoch": 1.1617747440273036, + "grad_norm": 1.85057272922146, + "learning_rate": 1.090808334154165e-06, + "loss": 0.1804, + "step": 5106 + }, + { + "epoch": 1.1620022753128556, + "grad_norm": 1.4961803957728892, + "learning_rate": 1.0907487555841797e-06, + "loss": 0.0582, + "step": 5107 + }, + { + "epoch": 1.1622298065984074, + "grad_norm": 2.43515890824142, + "learning_rate": 1.0906891674951426e-06, + "loss": 0.1159, + "step": 5108 + }, + { + "epoch": 1.162457337883959, + "grad_norm": 2.4092226646171033, + "learning_rate": 1.090629569888272e-06, + "loss": 0.1874, + "step": 5109 + }, + { + "epoch": 1.1626848691695109, + "grad_norm": 1.430724702642088, + "learning_rate": 1.0905699627647857e-06, + "loss": 0.0464, + "step": 5110 + }, + { + "epoch": 1.1629124004550626, + "grad_norm": 2.0043701924808364, + "learning_rate": 1.090510346125902e-06, + "loss": 0.1058, + "step": 5111 + }, + { + "epoch": 1.1631399317406144, + "grad_norm": 1.6881125359192517, + "learning_rate": 1.0904507199728392e-06, + "loss": 0.0975, + "step": 5112 + }, + { + "epoch": 1.1633674630261661, + "grad_norm": 2.020904439641956, + "learning_rate": 1.0903910843068163e-06, + "loss": 0.1164, + "step": 5113 + }, + { + "epoch": 1.1635949943117179, + "grad_norm": 1.4662243703806388, + "learning_rate": 1.090331439129052e-06, + "loss": 0.0449, + "step": 5114 + }, + { + "epoch": 1.1638225255972696, + "grad_norm": 2.3936294807872502, + "learning_rate": 1.0902717844407651e-06, + "loss": 0.1056, + "step": 5115 + }, + { + "epoch": 1.1640500568828214, + "grad_norm": 1.402339373672487, + "learning_rate": 1.0902121202431754e-06, + "loss": 0.0984, + "step": 5116 + }, + { + "epoch": 1.1642775881683731, + "grad_norm": 1.1601403443745117, + "learning_rate": 1.0901524465375015e-06, + "loss": 0.0618, + "step": 5117 + }, + { + "epoch": 1.1645051194539249, + "grad_norm": 2.3796698078699827, + "learning_rate": 1.0900927633249638e-06, + "loss": 0.2121, + "step": 5118 + }, + { + "epoch": 1.1647326507394766, + "grad_norm": 2.111092663401694, + "learning_rate": 1.0900330706067818e-06, + "loss": 0.1236, + "step": 5119 + }, + { + "epoch": 1.1649601820250284, + "grad_norm": 2.4073374503765286, + "learning_rate": 1.0899733683841753e-06, + "loss": 0.064, + "step": 5120 + }, + { + "epoch": 1.1651877133105801, + "grad_norm": 1.4765861177514648, + "learning_rate": 1.0899136566583647e-06, + "loss": 0.0787, + "step": 5121 + }, + { + "epoch": 1.1654152445961319, + "grad_norm": 1.8206625324312107, + "learning_rate": 1.0898539354305706e-06, + "loss": 0.0589, + "step": 5122 + }, + { + "epoch": 1.1656427758816836, + "grad_norm": 1.2742725949565343, + "learning_rate": 1.0897942047020131e-06, + "loss": 0.0823, + "step": 5123 + }, + { + "epoch": 1.1658703071672356, + "grad_norm": 2.317344383776149, + "learning_rate": 1.0897344644739139e-06, + "loss": 0.0802, + "step": 5124 + }, + { + "epoch": 1.1660978384527874, + "grad_norm": 2.082018147741833, + "learning_rate": 1.089674714747493e-06, + "loss": 0.0879, + "step": 5125 + }, + { + "epoch": 1.1663253697383391, + "grad_norm": 1.017078316760799, + "learning_rate": 1.0896149555239717e-06, + "loss": 0.0675, + "step": 5126 + }, + { + "epoch": 1.1665529010238909, + "grad_norm": 1.1599218032685075, + "learning_rate": 1.0895551868045718e-06, + "loss": 0.0357, + "step": 5127 + }, + { + "epoch": 1.1667804323094426, + "grad_norm": 1.3348293242411355, + "learning_rate": 1.0894954085905147e-06, + "loss": 0.0986, + "step": 5128 + }, + { + "epoch": 1.1670079635949944, + "grad_norm": 1.1294710535163477, + "learning_rate": 1.0894356208830223e-06, + "loss": 0.098, + "step": 5129 + }, + { + "epoch": 1.1672354948805461, + "grad_norm": 1.6825900072808717, + "learning_rate": 1.089375823683316e-06, + "loss": 0.0856, + "step": 5130 + }, + { + "epoch": 1.1674630261660979, + "grad_norm": 1.4078748724249117, + "learning_rate": 1.0893160169926186e-06, + "loss": 0.0397, + "step": 5131 + }, + { + "epoch": 1.1676905574516496, + "grad_norm": 2.2607172341356963, + "learning_rate": 1.0892562008121522e-06, + "loss": 0.1146, + "step": 5132 + }, + { + "epoch": 1.1679180887372014, + "grad_norm": 2.137903831657451, + "learning_rate": 1.0891963751431392e-06, + "loss": 0.0831, + "step": 5133 + }, + { + "epoch": 1.1681456200227531, + "grad_norm": 1.4391610604694447, + "learning_rate": 1.0891365399868022e-06, + "loss": 0.1239, + "step": 5134 + }, + { + "epoch": 1.1683731513083049, + "grad_norm": 1.3650904209209243, + "learning_rate": 1.0890766953443646e-06, + "loss": 0.0555, + "step": 5135 + }, + { + "epoch": 1.1686006825938566, + "grad_norm": 1.0371856273982027, + "learning_rate": 1.0890168412170493e-06, + "loss": 0.0951, + "step": 5136 + }, + { + "epoch": 1.1688282138794084, + "grad_norm": 1.5565752560490373, + "learning_rate": 1.0889569776060796e-06, + "loss": 0.0835, + "step": 5137 + }, + { + "epoch": 1.1690557451649601, + "grad_norm": 1.6651305013672462, + "learning_rate": 1.088897104512679e-06, + "loss": 0.1987, + "step": 5138 + }, + { + "epoch": 1.1692832764505119, + "grad_norm": 2.3900689357963825, + "learning_rate": 1.0888372219380709e-06, + "loss": 0.0795, + "step": 5139 + }, + { + "epoch": 1.1695108077360636, + "grad_norm": 2.432864936838184, + "learning_rate": 1.0887773298834798e-06, + "loss": 0.1243, + "step": 5140 + }, + { + "epoch": 1.1697383390216154, + "grad_norm": 1.185319864549767, + "learning_rate": 1.0887174283501293e-06, + "loss": 0.0477, + "step": 5141 + }, + { + "epoch": 1.1699658703071671, + "grad_norm": 1.8287546810857624, + "learning_rate": 1.0886575173392435e-06, + "loss": 0.1251, + "step": 5142 + }, + { + "epoch": 1.170193401592719, + "grad_norm": 2.177633518642662, + "learning_rate": 1.0885975968520476e-06, + "loss": 0.0893, + "step": 5143 + }, + { + "epoch": 1.1704209328782706, + "grad_norm": 1.3130463339730047, + "learning_rate": 1.0885376668897656e-06, + "loss": 0.0855, + "step": 5144 + }, + { + "epoch": 1.1706484641638226, + "grad_norm": 1.5628072719826096, + "learning_rate": 1.0884777274536228e-06, + "loss": 0.0856, + "step": 5145 + }, + { + "epoch": 1.1708759954493744, + "grad_norm": 2.4402271387318017, + "learning_rate": 1.0884177785448441e-06, + "loss": 0.1261, + "step": 5146 + }, + { + "epoch": 1.1711035267349261, + "grad_norm": 1.1284689217760129, + "learning_rate": 1.0883578201646546e-06, + "loss": 0.0861, + "step": 5147 + }, + { + "epoch": 1.1713310580204779, + "grad_norm": 1.808467621183764, + "learning_rate": 1.08829785231428e-06, + "loss": 0.0825, + "step": 5148 + }, + { + "epoch": 1.1715585893060296, + "grad_norm": 2.542285399099055, + "learning_rate": 1.0882378749949456e-06, + "loss": 0.1083, + "step": 5149 + }, + { + "epoch": 1.1717861205915814, + "grad_norm": 2.021913721897118, + "learning_rate": 1.0881778882078774e-06, + "loss": 0.0901, + "step": 5150 + }, + { + "epoch": 1.1720136518771331, + "grad_norm": 1.406045953578652, + "learning_rate": 1.0881178919543016e-06, + "loss": 0.0773, + "step": 5151 + }, + { + "epoch": 1.1722411831626849, + "grad_norm": 2.0040086875362944, + "learning_rate": 1.0880578862354444e-06, + "loss": 0.1644, + "step": 5152 + }, + { + "epoch": 1.1724687144482366, + "grad_norm": 1.8294788787626037, + "learning_rate": 1.087997871052532e-06, + "loss": 0.0948, + "step": 5153 + }, + { + "epoch": 1.1726962457337884, + "grad_norm": 1.736926598359245, + "learning_rate": 1.0879378464067906e-06, + "loss": 0.1223, + "step": 5154 + }, + { + "epoch": 1.1729237770193401, + "grad_norm": 1.91009580821313, + "learning_rate": 1.0878778122994477e-06, + "loss": 0.1093, + "step": 5155 + }, + { + "epoch": 1.173151308304892, + "grad_norm": 1.6708279487986484, + "learning_rate": 1.0878177687317302e-06, + "loss": 0.0982, + "step": 5156 + }, + { + "epoch": 1.1733788395904436, + "grad_norm": 1.7121086391611304, + "learning_rate": 1.0877577157048648e-06, + "loss": 0.0841, + "step": 5157 + }, + { + "epoch": 1.1736063708759954, + "grad_norm": 1.5932913712350176, + "learning_rate": 1.0876976532200797e-06, + "loss": 0.1199, + "step": 5158 + }, + { + "epoch": 1.1738339021615471, + "grad_norm": 1.9929519320083202, + "learning_rate": 1.0876375812786017e-06, + "loss": 0.094, + "step": 5159 + }, + { + "epoch": 1.174061433447099, + "grad_norm": 1.0817988622931909, + "learning_rate": 1.0875774998816586e-06, + "loss": 0.0748, + "step": 5160 + }, + { + "epoch": 1.1742889647326507, + "grad_norm": 1.2130835603422332, + "learning_rate": 1.087517409030479e-06, + "loss": 0.0383, + "step": 5161 + }, + { + "epoch": 1.1745164960182026, + "grad_norm": 1.7517362332183557, + "learning_rate": 1.0874573087262902e-06, + "loss": 0.0986, + "step": 5162 + }, + { + "epoch": 1.1747440273037544, + "grad_norm": 1.0166108394008488, + "learning_rate": 1.087397198970321e-06, + "loss": 0.0846, + "step": 5163 + }, + { + "epoch": 1.1749715585893061, + "grad_norm": 1.9628550032769965, + "learning_rate": 1.0873370797638002e-06, + "loss": 0.1405, + "step": 5164 + }, + { + "epoch": 1.1751990898748579, + "grad_norm": 1.8776527771081926, + "learning_rate": 1.0872769511079561e-06, + "loss": 0.1075, + "step": 5165 + }, + { + "epoch": 1.1754266211604096, + "grad_norm": 1.0899004955073874, + "learning_rate": 1.0872168130040175e-06, + "loss": 0.0422, + "step": 5166 + }, + { + "epoch": 1.1756541524459614, + "grad_norm": 1.3832930528440743, + "learning_rate": 1.087156665453214e-06, + "loss": 0.0768, + "step": 5167 + }, + { + "epoch": 1.1758816837315131, + "grad_norm": 1.9657143389773364, + "learning_rate": 1.0870965084567748e-06, + "loss": 0.1274, + "step": 5168 + }, + { + "epoch": 1.176109215017065, + "grad_norm": 1.5211869967581584, + "learning_rate": 1.087036342015929e-06, + "loss": 0.0845, + "step": 5169 + }, + { + "epoch": 1.1763367463026166, + "grad_norm": 1.143246238134219, + "learning_rate": 1.0869761661319064e-06, + "loss": 0.0973, + "step": 5170 + }, + { + "epoch": 1.1765642775881684, + "grad_norm": 1.9263291809785696, + "learning_rate": 1.0869159808059373e-06, + "loss": 0.0769, + "step": 5171 + }, + { + "epoch": 1.1767918088737201, + "grad_norm": 2.2676787630517787, + "learning_rate": 1.0868557860392516e-06, + "loss": 0.1097, + "step": 5172 + }, + { + "epoch": 1.177019340159272, + "grad_norm": 2.0019023188667764, + "learning_rate": 1.0867955818330792e-06, + "loss": 0.1312, + "step": 5173 + }, + { + "epoch": 1.1772468714448237, + "grad_norm": 1.9525647770660846, + "learning_rate": 1.086735368188651e-06, + "loss": 0.1152, + "step": 5174 + }, + { + "epoch": 1.1774744027303754, + "grad_norm": 1.096166997406837, + "learning_rate": 1.0866751451071974e-06, + "loss": 0.0966, + "step": 5175 + }, + { + "epoch": 1.1777019340159272, + "grad_norm": 1.796427540212248, + "learning_rate": 1.0866149125899495e-06, + "loss": 0.1346, + "step": 5176 + }, + { + "epoch": 1.177929465301479, + "grad_norm": 2.0245892091737177, + "learning_rate": 1.086554670638138e-06, + "loss": 0.1421, + "step": 5177 + }, + { + "epoch": 1.1781569965870307, + "grad_norm": 1.7782356641667965, + "learning_rate": 1.0864944192529946e-06, + "loss": 0.1021, + "step": 5178 + }, + { + "epoch": 1.1783845278725824, + "grad_norm": 1.1463385215583979, + "learning_rate": 1.08643415843575e-06, + "loss": 0.0348, + "step": 5179 + }, + { + "epoch": 1.1786120591581342, + "grad_norm": 2.176529725496333, + "learning_rate": 1.0863738881876369e-06, + "loss": 0.1201, + "step": 5180 + }, + { + "epoch": 1.178839590443686, + "grad_norm": 2.5057338760003276, + "learning_rate": 1.0863136085098862e-06, + "loss": 0.2077, + "step": 5181 + }, + { + "epoch": 1.1790671217292377, + "grad_norm": 1.5428397804224192, + "learning_rate": 1.08625331940373e-06, + "loss": 0.0771, + "step": 5182 + }, + { + "epoch": 1.1792946530147894, + "grad_norm": 1.8434351161312814, + "learning_rate": 1.086193020870401e-06, + "loss": 0.1054, + "step": 5183 + }, + { + "epoch": 1.1795221843003414, + "grad_norm": 1.9271362170030948, + "learning_rate": 1.0861327129111313e-06, + "loss": 0.0839, + "step": 5184 + }, + { + "epoch": 1.1797497155858931, + "grad_norm": 1.5531813526055755, + "learning_rate": 1.0860723955271533e-06, + "loss": 0.0622, + "step": 5185 + }, + { + "epoch": 1.179977246871445, + "grad_norm": 1.759159281092021, + "learning_rate": 1.0860120687196998e-06, + "loss": 0.1115, + "step": 5186 + }, + { + "epoch": 1.1802047781569966, + "grad_norm": 1.8647371747744599, + "learning_rate": 1.0859517324900042e-06, + "loss": 0.0861, + "step": 5187 + }, + { + "epoch": 1.1804323094425484, + "grad_norm": 1.805897808366817, + "learning_rate": 1.0858913868392993e-06, + "loss": 0.1431, + "step": 5188 + }, + { + "epoch": 1.1806598407281002, + "grad_norm": 1.6704627128473584, + "learning_rate": 1.0858310317688184e-06, + "loss": 0.0873, + "step": 5189 + }, + { + "epoch": 1.180887372013652, + "grad_norm": 1.343094360546439, + "learning_rate": 1.0857706672797954e-06, + "loss": 0.0859, + "step": 5190 + }, + { + "epoch": 1.1811149032992037, + "grad_norm": 2.066147911559209, + "learning_rate": 1.0857102933734636e-06, + "loss": 0.1149, + "step": 5191 + }, + { + "epoch": 1.1813424345847554, + "grad_norm": 1.4544321870028036, + "learning_rate": 1.0856499100510575e-06, + "loss": 0.0601, + "step": 5192 + }, + { + "epoch": 1.1815699658703072, + "grad_norm": 2.1031693123515574, + "learning_rate": 1.0855895173138107e-06, + "loss": 0.1274, + "step": 5193 + }, + { + "epoch": 1.181797497155859, + "grad_norm": 1.7573489864066676, + "learning_rate": 1.0855291151629576e-06, + "loss": 0.0615, + "step": 5194 + }, + { + "epoch": 1.1820250284414107, + "grad_norm": 1.9854988027981146, + "learning_rate": 1.0854687035997329e-06, + "loss": 0.097, + "step": 5195 + }, + { + "epoch": 1.1822525597269624, + "grad_norm": 1.5268785940107066, + "learning_rate": 1.0854082826253712e-06, + "loss": 0.0491, + "step": 5196 + }, + { + "epoch": 1.1824800910125142, + "grad_norm": 1.1410295227600973, + "learning_rate": 1.0853478522411072e-06, + "loss": 0.0546, + "step": 5197 + }, + { + "epoch": 1.182707622298066, + "grad_norm": 1.4884685353979168, + "learning_rate": 1.0852874124481764e-06, + "loss": 0.0544, + "step": 5198 + }, + { + "epoch": 1.1829351535836177, + "grad_norm": 1.3165143158682053, + "learning_rate": 1.0852269632478138e-06, + "loss": 0.0916, + "step": 5199 + }, + { + "epoch": 1.1831626848691694, + "grad_norm": 1.3572510385460108, + "learning_rate": 1.085166504641255e-06, + "loss": 0.0509, + "step": 5200 + }, + { + "epoch": 1.1833902161547214, + "grad_norm": 0.9661066398138327, + "learning_rate": 1.0851060366297356e-06, + "loss": 0.0875, + "step": 5201 + }, + { + "epoch": 1.1836177474402731, + "grad_norm": 2.333836932538073, + "learning_rate": 1.0850455592144916e-06, + "loss": 0.1679, + "step": 5202 + }, + { + "epoch": 1.183845278725825, + "grad_norm": 1.445056359388779, + "learning_rate": 1.0849850723967585e-06, + "loss": 0.071, + "step": 5203 + }, + { + "epoch": 1.1840728100113767, + "grad_norm": 1.6295801416897442, + "learning_rate": 1.0849245761777733e-06, + "loss": 0.0938, + "step": 5204 + }, + { + "epoch": 1.1843003412969284, + "grad_norm": 2.3549017722464316, + "learning_rate": 1.0848640705587718e-06, + "loss": 0.124, + "step": 5205 + }, + { + "epoch": 1.1845278725824802, + "grad_norm": 2.050952800493849, + "learning_rate": 1.0848035555409911e-06, + "loss": 0.0691, + "step": 5206 + }, + { + "epoch": 1.184755403868032, + "grad_norm": 2.077817689629389, + "learning_rate": 1.0847430311256676e-06, + "loss": 0.0818, + "step": 5207 + }, + { + "epoch": 1.1849829351535837, + "grad_norm": 2.253900685722683, + "learning_rate": 1.0846824973140388e-06, + "loss": 0.1127, + "step": 5208 + }, + { + "epoch": 1.1852104664391354, + "grad_norm": 1.9840592846831089, + "learning_rate": 1.0846219541073417e-06, + "loss": 0.0618, + "step": 5209 + }, + { + "epoch": 1.1854379977246872, + "grad_norm": 1.6704261322608704, + "learning_rate": 1.084561401506813e-06, + "loss": 0.1531, + "step": 5210 + }, + { + "epoch": 1.185665529010239, + "grad_norm": 2.1725141580654896, + "learning_rate": 1.0845008395136915e-06, + "loss": 0.2325, + "step": 5211 + }, + { + "epoch": 1.1858930602957907, + "grad_norm": 2.393322470314106, + "learning_rate": 1.0844402681292144e-06, + "loss": 0.2173, + "step": 5212 + }, + { + "epoch": 1.1861205915813424, + "grad_norm": 1.6261443486893132, + "learning_rate": 1.084379687354619e-06, + "loss": 0.0728, + "step": 5213 + }, + { + "epoch": 1.1863481228668942, + "grad_norm": 2.2205983994574523, + "learning_rate": 1.0843190971911447e-06, + "loss": 0.1077, + "step": 5214 + }, + { + "epoch": 1.186575654152446, + "grad_norm": 1.6223311298818315, + "learning_rate": 1.0842584976400292e-06, + "loss": 0.169, + "step": 5215 + }, + { + "epoch": 1.1868031854379977, + "grad_norm": 1.5664039317333338, + "learning_rate": 1.084197888702511e-06, + "loss": 0.1082, + "step": 5216 + }, + { + "epoch": 1.1870307167235494, + "grad_norm": 2.1952187183985354, + "learning_rate": 1.0841372703798289e-06, + "loss": 0.1031, + "step": 5217 + }, + { + "epoch": 1.1872582480091012, + "grad_norm": 1.5272178345552438, + "learning_rate": 1.0840766426732219e-06, + "loss": 0.0822, + "step": 5218 + }, + { + "epoch": 1.187485779294653, + "grad_norm": 1.4542703091629854, + "learning_rate": 1.0840160055839291e-06, + "loss": 0.1352, + "step": 5219 + }, + { + "epoch": 1.1877133105802047, + "grad_norm": 1.6566468578059188, + "learning_rate": 1.08395535911319e-06, + "loss": 0.0599, + "step": 5220 + }, + { + "epoch": 1.1879408418657564, + "grad_norm": 1.4369561966167148, + "learning_rate": 1.0838947032622436e-06, + "loss": 0.0588, + "step": 5221 + }, + { + "epoch": 1.1881683731513082, + "grad_norm": 1.6911609691995355, + "learning_rate": 1.0838340380323297e-06, + "loss": 0.107, + "step": 5222 + }, + { + "epoch": 1.1883959044368602, + "grad_norm": 2.7120228637912085, + "learning_rate": 1.083773363424689e-06, + "loss": 0.1183, + "step": 5223 + }, + { + "epoch": 1.188623435722412, + "grad_norm": 1.4822450492657908, + "learning_rate": 1.0837126794405603e-06, + "loss": 0.0733, + "step": 5224 + }, + { + "epoch": 1.1888509670079637, + "grad_norm": 2.189919934484601, + "learning_rate": 1.083651986081185e-06, + "loss": 0.1001, + "step": 5225 + }, + { + "epoch": 1.1890784982935154, + "grad_norm": 1.3261475366877535, + "learning_rate": 1.0835912833478029e-06, + "loss": 0.1536, + "step": 5226 + }, + { + "epoch": 1.1893060295790672, + "grad_norm": 2.0355476056994553, + "learning_rate": 1.0835305712416546e-06, + "loss": 0.0915, + "step": 5227 + }, + { + "epoch": 1.189533560864619, + "grad_norm": 1.9337187468158625, + "learning_rate": 1.0834698497639817e-06, + "loss": 0.1797, + "step": 5228 + }, + { + "epoch": 1.1897610921501707, + "grad_norm": 1.528534377609125, + "learning_rate": 1.0834091189160243e-06, + "loss": 0.0623, + "step": 5229 + }, + { + "epoch": 1.1899886234357224, + "grad_norm": 1.7978417621819656, + "learning_rate": 1.0833483786990243e-06, + "loss": 0.1616, + "step": 5230 + }, + { + "epoch": 1.1902161547212742, + "grad_norm": 1.3180132857072435, + "learning_rate": 1.0832876291142228e-06, + "loss": 0.0656, + "step": 5231 + }, + { + "epoch": 1.190443686006826, + "grad_norm": 1.5694544783138118, + "learning_rate": 1.0832268701628613e-06, + "loss": 0.1214, + "step": 5232 + }, + { + "epoch": 1.1906712172923777, + "grad_norm": 2.7018835662805047, + "learning_rate": 1.083166101846182e-06, + "loss": 0.2433, + "step": 5233 + }, + { + "epoch": 1.1908987485779294, + "grad_norm": 1.2165773857409203, + "learning_rate": 1.0831053241654265e-06, + "loss": 0.0302, + "step": 5234 + }, + { + "epoch": 1.1911262798634812, + "grad_norm": 1.4873073484241224, + "learning_rate": 1.0830445371218374e-06, + "loss": 0.1109, + "step": 5235 + }, + { + "epoch": 1.191353811149033, + "grad_norm": 2.169001416121012, + "learning_rate": 1.0829837407166565e-06, + "loss": 0.1875, + "step": 5236 + }, + { + "epoch": 1.1915813424345847, + "grad_norm": 2.2992957388980924, + "learning_rate": 1.082922934951127e-06, + "loss": 0.1762, + "step": 5237 + }, + { + "epoch": 1.1918088737201364, + "grad_norm": 3.0098436793577776, + "learning_rate": 1.082862119826491e-06, + "loss": 0.1161, + "step": 5238 + }, + { + "epoch": 1.1920364050056882, + "grad_norm": 1.0473761856310173, + "learning_rate": 1.0828012953439921e-06, + "loss": 0.151, + "step": 5239 + }, + { + "epoch": 1.1922639362912402, + "grad_norm": 2.200991609795066, + "learning_rate": 1.082740461504873e-06, + "loss": 0.049, + "step": 5240 + }, + { + "epoch": 1.192491467576792, + "grad_norm": 2.1341044427621654, + "learning_rate": 1.0826796183103774e-06, + "loss": 0.0995, + "step": 5241 + }, + { + "epoch": 1.1927189988623437, + "grad_norm": 1.797138107236956, + "learning_rate": 1.0826187657617484e-06, + "loss": 0.0836, + "step": 5242 + }, + { + "epoch": 1.1929465301478954, + "grad_norm": 1.5361734716911093, + "learning_rate": 1.0825579038602298e-06, + "loss": 0.0956, + "step": 5243 + }, + { + "epoch": 1.1931740614334472, + "grad_norm": 1.3830245145453257, + "learning_rate": 1.082497032607066e-06, + "loss": 0.0859, + "step": 5244 + }, + { + "epoch": 1.193401592718999, + "grad_norm": 1.653730545070457, + "learning_rate": 1.0824361520035004e-06, + "loss": 0.1367, + "step": 5245 + }, + { + "epoch": 1.1936291240045507, + "grad_norm": 1.5316553149024767, + "learning_rate": 1.0823752620507778e-06, + "loss": 0.0923, + "step": 5246 + }, + { + "epoch": 1.1938566552901024, + "grad_norm": 1.2318860231960242, + "learning_rate": 1.0823143627501423e-06, + "loss": 0.0465, + "step": 5247 + }, + { + "epoch": 1.1940841865756542, + "grad_norm": 1.368501769833903, + "learning_rate": 1.082253454102839e-06, + "loss": 0.0527, + "step": 5248 + }, + { + "epoch": 1.194311717861206, + "grad_norm": 1.688804518608022, + "learning_rate": 1.0821925361101124e-06, + "loss": 0.0794, + "step": 5249 + }, + { + "epoch": 1.1945392491467577, + "grad_norm": 1.350823772473527, + "learning_rate": 1.0821316087732075e-06, + "loss": 0.0687, + "step": 5250 + }, + { + "epoch": 1.1947667804323094, + "grad_norm": 1.1567294535302972, + "learning_rate": 1.0820706720933698e-06, + "loss": 0.104, + "step": 5251 + }, + { + "epoch": 1.1949943117178612, + "grad_norm": 1.6015117566025172, + "learning_rate": 1.0820097260718448e-06, + "loss": 0.0855, + "step": 5252 + }, + { + "epoch": 1.195221843003413, + "grad_norm": 2.133718516574482, + "learning_rate": 1.081948770709878e-06, + "loss": 0.1237, + "step": 5253 + }, + { + "epoch": 1.1954493742889647, + "grad_norm": 2.192934701634061, + "learning_rate": 1.0818878060087151e-06, + "loss": 0.1113, + "step": 5254 + }, + { + "epoch": 1.1956769055745164, + "grad_norm": 1.5826077427944494, + "learning_rate": 1.081826831969602e-06, + "loss": 0.122, + "step": 5255 + }, + { + "epoch": 1.1959044368600682, + "grad_norm": 1.9016171603713237, + "learning_rate": 1.0817658485937855e-06, + "loss": 0.1475, + "step": 5256 + }, + { + "epoch": 1.19613196814562, + "grad_norm": 1.574823856287606, + "learning_rate": 1.0817048558825114e-06, + "loss": 0.1361, + "step": 5257 + }, + { + "epoch": 1.1963594994311717, + "grad_norm": 1.8062796035655573, + "learning_rate": 1.0816438538370262e-06, + "loss": 0.0838, + "step": 5258 + }, + { + "epoch": 1.1965870307167235, + "grad_norm": 2.150669189308933, + "learning_rate": 1.0815828424585772e-06, + "loss": 0.0727, + "step": 5259 + }, + { + "epoch": 1.1968145620022752, + "grad_norm": 2.37299043188769, + "learning_rate": 1.081521821748411e-06, + "loss": 0.0777, + "step": 5260 + }, + { + "epoch": 1.197042093287827, + "grad_norm": 1.5748977460726736, + "learning_rate": 1.081460791707775e-06, + "loss": 0.0713, + "step": 5261 + }, + { + "epoch": 1.197269624573379, + "grad_norm": 1.845039914221183, + "learning_rate": 1.0813997523379163e-06, + "loss": 0.1783, + "step": 5262 + }, + { + "epoch": 1.1974971558589307, + "grad_norm": 1.2418153199696063, + "learning_rate": 1.0813387036400825e-06, + "loss": 0.0838, + "step": 5263 + }, + { + "epoch": 1.1977246871444824, + "grad_norm": 1.601138886177542, + "learning_rate": 1.081277645615521e-06, + "loss": 0.0561, + "step": 5264 + }, + { + "epoch": 1.1979522184300342, + "grad_norm": 2.477792000861364, + "learning_rate": 1.0812165782654806e-06, + "loss": 0.0802, + "step": 5265 + }, + { + "epoch": 1.198179749715586, + "grad_norm": 1.8339107520724027, + "learning_rate": 1.0811555015912086e-06, + "loss": 0.0638, + "step": 5266 + }, + { + "epoch": 1.1984072810011377, + "grad_norm": 1.1686725493823096, + "learning_rate": 1.0810944155939536e-06, + "loss": 0.059, + "step": 5267 + }, + { + "epoch": 1.1986348122866894, + "grad_norm": 1.9602369962972666, + "learning_rate": 1.081033320274964e-06, + "loss": 0.1053, + "step": 5268 + }, + { + "epoch": 1.1988623435722412, + "grad_norm": 1.1474755283420095, + "learning_rate": 1.0809722156354884e-06, + "loss": 0.1005, + "step": 5269 + }, + { + "epoch": 1.199089874857793, + "grad_norm": 0.9917236908020932, + "learning_rate": 1.0809111016767762e-06, + "loss": 0.0622, + "step": 5270 + }, + { + "epoch": 1.1993174061433447, + "grad_norm": 1.4830981024424121, + "learning_rate": 1.0808499784000756e-06, + "loss": 0.0561, + "step": 5271 + }, + { + "epoch": 1.1995449374288965, + "grad_norm": 1.2794179990446428, + "learning_rate": 1.0807888458066364e-06, + "loss": 0.0547, + "step": 5272 + }, + { + "epoch": 1.1997724687144482, + "grad_norm": 2.4510983663131745, + "learning_rate": 1.0807277038977083e-06, + "loss": 0.2331, + "step": 5273 + }, + { + "epoch": 1.2, + "grad_norm": 3.2059009973371646, + "learning_rate": 1.0806665526745403e-06, + "loss": 0.2011, + "step": 5274 + }, + { + "epoch": 1.2002275312855517, + "grad_norm": 1.603764392751713, + "learning_rate": 1.0806053921383823e-06, + "loss": 0.0736, + "step": 5275 + }, + { + "epoch": 1.2004550625711035, + "grad_norm": 1.098547269397925, + "learning_rate": 1.0805442222904846e-06, + "loss": 0.071, + "step": 5276 + }, + { + "epoch": 1.2006825938566552, + "grad_norm": 2.2506163203297875, + "learning_rate": 1.0804830431320972e-06, + "loss": 0.1489, + "step": 5277 + }, + { + "epoch": 1.200910125142207, + "grad_norm": 1.904831313890037, + "learning_rate": 1.0804218546644708e-06, + "loss": 0.128, + "step": 5278 + }, + { + "epoch": 1.201137656427759, + "grad_norm": 1.3427616236849924, + "learning_rate": 1.0803606568888557e-06, + "loss": 0.0641, + "step": 5279 + }, + { + "epoch": 1.2013651877133107, + "grad_norm": 2.695821206487897, + "learning_rate": 1.0802994498065027e-06, + "loss": 0.1282, + "step": 5280 + }, + { + "epoch": 1.2015927189988624, + "grad_norm": 2.0080613955222235, + "learning_rate": 1.0802382334186627e-06, + "loss": 0.0821, + "step": 5281 + }, + { + "epoch": 1.2018202502844142, + "grad_norm": 1.6911329459727755, + "learning_rate": 1.080177007726587e-06, + "loss": 0.1019, + "step": 5282 + }, + { + "epoch": 1.202047781569966, + "grad_norm": 1.6381634754276007, + "learning_rate": 1.080115772731527e-06, + "loss": 0.0434, + "step": 5283 + }, + { + "epoch": 1.2022753128555177, + "grad_norm": 1.1969099245269164, + "learning_rate": 1.0800545284347338e-06, + "loss": 0.0616, + "step": 5284 + }, + { + "epoch": 1.2025028441410694, + "grad_norm": 1.7741375061533398, + "learning_rate": 1.0799932748374598e-06, + "loss": 0.0436, + "step": 5285 + }, + { + "epoch": 1.2027303754266212, + "grad_norm": 2.264571806309411, + "learning_rate": 1.0799320119409562e-06, + "loss": 0.104, + "step": 5286 + }, + { + "epoch": 1.202957906712173, + "grad_norm": 1.848273839594112, + "learning_rate": 1.0798707397464756e-06, + "loss": 0.117, + "step": 5287 + }, + { + "epoch": 1.2031854379977247, + "grad_norm": 2.262303139744276, + "learning_rate": 1.0798094582552703e-06, + "loss": 0.1028, + "step": 5288 + }, + { + "epoch": 1.2034129692832765, + "grad_norm": 2.0589025509516263, + "learning_rate": 1.0797481674685925e-06, + "loss": 0.2736, + "step": 5289 + }, + { + "epoch": 1.2036405005688282, + "grad_norm": 1.7761896760428244, + "learning_rate": 1.0796868673876947e-06, + "loss": 0.087, + "step": 5290 + }, + { + "epoch": 1.20386803185438, + "grad_norm": 1.7782779357046952, + "learning_rate": 1.0796255580138303e-06, + "loss": 0.0899, + "step": 5291 + }, + { + "epoch": 1.2040955631399317, + "grad_norm": 1.1500037983509879, + "learning_rate": 1.0795642393482523e-06, + "loss": 0.118, + "step": 5292 + }, + { + "epoch": 1.2043230944254835, + "grad_norm": 1.842925587308416, + "learning_rate": 1.0795029113922136e-06, + "loss": 0.0426, + "step": 5293 + }, + { + "epoch": 1.2045506257110352, + "grad_norm": 2.1630846104313783, + "learning_rate": 1.0794415741469677e-06, + "loss": 0.0854, + "step": 5294 + }, + { + "epoch": 1.204778156996587, + "grad_norm": 1.9984928427035924, + "learning_rate": 1.0793802276137683e-06, + "loss": 0.0758, + "step": 5295 + }, + { + "epoch": 1.2050056882821387, + "grad_norm": 1.7270916728186407, + "learning_rate": 1.0793188717938693e-06, + "loss": 0.0984, + "step": 5296 + }, + { + "epoch": 1.2052332195676905, + "grad_norm": 2.562446895041978, + "learning_rate": 1.0792575066885245e-06, + "loss": 0.0848, + "step": 5297 + }, + { + "epoch": 1.2054607508532422, + "grad_norm": 1.0211875489183757, + "learning_rate": 1.0791961322989882e-06, + "loss": 0.0809, + "step": 5298 + }, + { + "epoch": 1.205688282138794, + "grad_norm": 1.4354948603513233, + "learning_rate": 1.0791347486265147e-06, + "loss": 0.0578, + "step": 5299 + }, + { + "epoch": 1.2059158134243457, + "grad_norm": 1.7350825872789628, + "learning_rate": 1.0790733556723589e-06, + "loss": 0.0941, + "step": 5300 + }, + { + "epoch": 1.2061433447098977, + "grad_norm": 1.8738498200816718, + "learning_rate": 1.079011953437775e-06, + "loss": 0.0693, + "step": 5301 + }, + { + "epoch": 1.2063708759954495, + "grad_norm": 1.3923030676619426, + "learning_rate": 1.0789505419240185e-06, + "loss": 0.0453, + "step": 5302 + }, + { + "epoch": 1.2065984072810012, + "grad_norm": 1.5273840730643755, + "learning_rate": 1.0788891211323442e-06, + "loss": 0.1353, + "step": 5303 + }, + { + "epoch": 1.206825938566553, + "grad_norm": 1.33456271047386, + "learning_rate": 1.0788276910640074e-06, + "loss": 0.0269, + "step": 5304 + }, + { + "epoch": 1.2070534698521047, + "grad_norm": 1.3783261991509468, + "learning_rate": 1.0787662517202641e-06, + "loss": 0.0564, + "step": 5305 + }, + { + "epoch": 1.2072810011376565, + "grad_norm": 1.2413093608926025, + "learning_rate": 1.0787048031023693e-06, + "loss": 0.0514, + "step": 5306 + }, + { + "epoch": 1.2075085324232082, + "grad_norm": 2.067632414807228, + "learning_rate": 1.0786433452115794e-06, + "loss": 0.0852, + "step": 5307 + }, + { + "epoch": 1.20773606370876, + "grad_norm": 1.6493210193338594, + "learning_rate": 1.0785818780491502e-06, + "loss": 0.0957, + "step": 5308 + }, + { + "epoch": 1.2079635949943117, + "grad_norm": 2.335430233427145, + "learning_rate": 1.0785204016163384e-06, + "loss": 0.1113, + "step": 5309 + }, + { + "epoch": 1.2081911262798635, + "grad_norm": 1.8914362366278195, + "learning_rate": 1.0784589159143999e-06, + "loss": 0.0587, + "step": 5310 + }, + { + "epoch": 1.2084186575654152, + "grad_norm": 1.8389745120373866, + "learning_rate": 1.0783974209445915e-06, + "loss": 0.0582, + "step": 5311 + }, + { + "epoch": 1.208646188850967, + "grad_norm": 1.2870683840945183, + "learning_rate": 1.0783359167081705e-06, + "loss": 0.0769, + "step": 5312 + }, + { + "epoch": 1.2088737201365187, + "grad_norm": 1.9375950528547237, + "learning_rate": 1.0782744032063935e-06, + "loss": 0.1055, + "step": 5313 + }, + { + "epoch": 1.2091012514220705, + "grad_norm": 2.9149019557260667, + "learning_rate": 1.078212880440518e-06, + "loss": 0.0749, + "step": 5314 + }, + { + "epoch": 1.2093287827076222, + "grad_norm": 1.4032344304927438, + "learning_rate": 1.0781513484118008e-06, + "loss": 0.0953, + "step": 5315 + }, + { + "epoch": 1.209556313993174, + "grad_norm": 2.021712881413012, + "learning_rate": 1.0780898071215004e-06, + "loss": 0.074, + "step": 5316 + }, + { + "epoch": 1.2097838452787257, + "grad_norm": 1.295514380352357, + "learning_rate": 1.0780282565708738e-06, + "loss": 0.1227, + "step": 5317 + }, + { + "epoch": 1.2100113765642777, + "grad_norm": 2.7681513279773555, + "learning_rate": 1.0779666967611796e-06, + "loss": 0.187, + "step": 5318 + }, + { + "epoch": 1.2102389078498295, + "grad_norm": 1.8640645938184524, + "learning_rate": 1.0779051276936755e-06, + "loss": 0.0711, + "step": 5319 + }, + { + "epoch": 1.2104664391353812, + "grad_norm": 1.3066357408288654, + "learning_rate": 1.0778435493696202e-06, + "loss": 0.1392, + "step": 5320 + }, + { + "epoch": 1.210693970420933, + "grad_norm": 2.2361490684160663, + "learning_rate": 1.0777819617902718e-06, + "loss": 0.1587, + "step": 5321 + }, + { + "epoch": 1.2109215017064847, + "grad_norm": 2.28890369536384, + "learning_rate": 1.0777203649568896e-06, + "loss": 0.1427, + "step": 5322 + }, + { + "epoch": 1.2111490329920365, + "grad_norm": 2.443743866077574, + "learning_rate": 1.077658758870732e-06, + "loss": 0.0844, + "step": 5323 + }, + { + "epoch": 1.2113765642775882, + "grad_norm": 2.5045475861456246, + "learning_rate": 1.0775971435330588e-06, + "loss": 0.0822, + "step": 5324 + }, + { + "epoch": 1.21160409556314, + "grad_norm": 1.4419614896477466, + "learning_rate": 1.0775355189451286e-06, + "loss": 0.1231, + "step": 5325 + }, + { + "epoch": 1.2118316268486917, + "grad_norm": 2.1607954202798574, + "learning_rate": 1.0774738851082011e-06, + "loss": 0.0656, + "step": 5326 + }, + { + "epoch": 1.2120591581342435, + "grad_norm": 2.0076172738726563, + "learning_rate": 1.0774122420235363e-06, + "loss": 0.1213, + "step": 5327 + }, + { + "epoch": 1.2122866894197952, + "grad_norm": 2.243065656502911, + "learning_rate": 1.0773505896923936e-06, + "loss": 0.2355, + "step": 5328 + }, + { + "epoch": 1.212514220705347, + "grad_norm": 1.5464308593607745, + "learning_rate": 1.0772889281160335e-06, + "loss": 0.1193, + "step": 5329 + }, + { + "epoch": 1.2127417519908987, + "grad_norm": 2.0341247889416456, + "learning_rate": 1.0772272572957158e-06, + "loss": 0.0715, + "step": 5330 + }, + { + "epoch": 1.2129692832764505, + "grad_norm": 1.638874906761963, + "learning_rate": 1.0771655772327013e-06, + "loss": 0.0915, + "step": 5331 + }, + { + "epoch": 1.2131968145620022, + "grad_norm": 1.8261492962825066, + "learning_rate": 1.0771038879282505e-06, + "loss": 0.1693, + "step": 5332 + }, + { + "epoch": 1.213424345847554, + "grad_norm": 1.6573900903132563, + "learning_rate": 1.0770421893836243e-06, + "loss": 0.1321, + "step": 5333 + }, + { + "epoch": 1.2136518771331057, + "grad_norm": 1.0977836462853496, + "learning_rate": 1.0769804816000835e-06, + "loss": 0.0332, + "step": 5334 + }, + { + "epoch": 1.2138794084186575, + "grad_norm": 2.5558904100663935, + "learning_rate": 1.0769187645788895e-06, + "loss": 0.1563, + "step": 5335 + }, + { + "epoch": 1.2141069397042092, + "grad_norm": 1.6661784494218759, + "learning_rate": 1.0768570383213035e-06, + "loss": 0.0554, + "step": 5336 + }, + { + "epoch": 1.214334470989761, + "grad_norm": 2.081264112380425, + "learning_rate": 1.0767953028285872e-06, + "loss": 0.0795, + "step": 5337 + }, + { + "epoch": 1.2145620022753127, + "grad_norm": 1.3727735429346972, + "learning_rate": 1.0767335581020024e-06, + "loss": 0.0684, + "step": 5338 + }, + { + "epoch": 1.2147895335608645, + "grad_norm": 1.2415036393799737, + "learning_rate": 1.076671804142811e-06, + "loss": 0.065, + "step": 5339 + }, + { + "epoch": 1.2150170648464165, + "grad_norm": 2.1447181222162635, + "learning_rate": 1.076610040952275e-06, + "loss": 0.1577, + "step": 5340 + }, + { + "epoch": 1.2152445961319682, + "grad_norm": 1.8197546232930755, + "learning_rate": 1.076548268531657e-06, + "loss": 0.0763, + "step": 5341 + }, + { + "epoch": 1.21547212741752, + "grad_norm": 1.5817686653534704, + "learning_rate": 1.0764864868822194e-06, + "loss": 0.1499, + "step": 5342 + }, + { + "epoch": 1.2156996587030717, + "grad_norm": 1.7955937776313586, + "learning_rate": 1.0764246960052247e-06, + "loss": 0.0699, + "step": 5343 + }, + { + "epoch": 1.2159271899886235, + "grad_norm": 1.5998182027119985, + "learning_rate": 1.0763628959019359e-06, + "loss": 0.0873, + "step": 5344 + }, + { + "epoch": 1.2161547212741752, + "grad_norm": 2.0206122711914953, + "learning_rate": 1.076301086573616e-06, + "loss": 0.0591, + "step": 5345 + }, + { + "epoch": 1.216382252559727, + "grad_norm": 1.7209121866538086, + "learning_rate": 1.076239268021529e-06, + "loss": 0.1182, + "step": 5346 + }, + { + "epoch": 1.2166097838452787, + "grad_norm": 1.6012738381710532, + "learning_rate": 1.0761774402469375e-06, + "loss": 0.1265, + "step": 5347 + }, + { + "epoch": 1.2168373151308305, + "grad_norm": 1.5190529580939407, + "learning_rate": 1.0761156032511052e-06, + "loss": 0.0509, + "step": 5348 + }, + { + "epoch": 1.2170648464163822, + "grad_norm": 1.2081251429678452, + "learning_rate": 1.0760537570352963e-06, + "loss": 0.0699, + "step": 5349 + }, + { + "epoch": 1.217292377701934, + "grad_norm": 1.270701433256153, + "learning_rate": 1.0759919016007747e-06, + "loss": 0.0988, + "step": 5350 + }, + { + "epoch": 1.2175199089874857, + "grad_norm": 1.7911755711480872, + "learning_rate": 1.0759300369488046e-06, + "loss": 0.1166, + "step": 5351 + }, + { + "epoch": 1.2177474402730375, + "grad_norm": 1.9513325042672363, + "learning_rate": 1.0758681630806502e-06, + "loss": 0.1296, + "step": 5352 + }, + { + "epoch": 1.2179749715585892, + "grad_norm": 1.7136195442158118, + "learning_rate": 1.0758062799975765e-06, + "loss": 0.1126, + "step": 5353 + }, + { + "epoch": 1.218202502844141, + "grad_norm": 1.744883740836297, + "learning_rate": 1.075744387700848e-06, + "loss": 0.1597, + "step": 5354 + }, + { + "epoch": 1.2184300341296928, + "grad_norm": 1.417163512342875, + "learning_rate": 1.0756824861917294e-06, + "loss": 0.0616, + "step": 5355 + }, + { + "epoch": 1.2186575654152445, + "grad_norm": 2.5446233172568244, + "learning_rate": 1.0756205754714866e-06, + "loss": 0.1226, + "step": 5356 + }, + { + "epoch": 1.2188850967007965, + "grad_norm": 1.2377188221566995, + "learning_rate": 1.0755586555413845e-06, + "loss": 0.0689, + "step": 5357 + }, + { + "epoch": 1.2191126279863482, + "grad_norm": 0.8909679121651022, + "learning_rate": 1.0754967264026882e-06, + "loss": 0.0289, + "step": 5358 + }, + { + "epoch": 1.2193401592719, + "grad_norm": 2.0352234100842055, + "learning_rate": 1.0754347880566643e-06, + "loss": 0.0828, + "step": 5359 + }, + { + "epoch": 1.2195676905574517, + "grad_norm": 1.0940962413492108, + "learning_rate": 1.075372840504578e-06, + "loss": 0.0727, + "step": 5360 + }, + { + "epoch": 1.2197952218430035, + "grad_norm": 2.270597989691131, + "learning_rate": 1.0753108837476958e-06, + "loss": 0.0725, + "step": 5361 + }, + { + "epoch": 1.2200227531285552, + "grad_norm": 3.189546972379516, + "learning_rate": 1.0752489177872839e-06, + "loss": 0.1068, + "step": 5362 + }, + { + "epoch": 1.220250284414107, + "grad_norm": 1.2750167089262343, + "learning_rate": 1.0751869426246086e-06, + "loss": 0.1154, + "step": 5363 + }, + { + "epoch": 1.2204778156996587, + "grad_norm": 2.135052433897321, + "learning_rate": 1.0751249582609368e-06, + "loss": 0.0865, + "step": 5364 + }, + { + "epoch": 1.2207053469852105, + "grad_norm": 1.244270521244428, + "learning_rate": 1.075062964697535e-06, + "loss": 0.0383, + "step": 5365 + }, + { + "epoch": 1.2209328782707622, + "grad_norm": 2.7250882706023827, + "learning_rate": 1.0750009619356706e-06, + "loss": 0.1476, + "step": 5366 + }, + { + "epoch": 1.221160409556314, + "grad_norm": 1.9790334045682223, + "learning_rate": 1.0749389499766106e-06, + "loss": 0.1329, + "step": 5367 + }, + { + "epoch": 1.2213879408418657, + "grad_norm": 2.648242876474357, + "learning_rate": 1.0748769288216226e-06, + "loss": 0.1221, + "step": 5368 + }, + { + "epoch": 1.2216154721274175, + "grad_norm": 1.8748774083050568, + "learning_rate": 1.074814898471974e-06, + "loss": 0.1707, + "step": 5369 + }, + { + "epoch": 1.2218430034129693, + "grad_norm": 1.8555297714441998, + "learning_rate": 1.0747528589289327e-06, + "loss": 0.0599, + "step": 5370 + }, + { + "epoch": 1.222070534698521, + "grad_norm": 0.9556357019620195, + "learning_rate": 1.0746908101937666e-06, + "loss": 0.0234, + "step": 5371 + }, + { + "epoch": 1.2222980659840728, + "grad_norm": 1.5483602186289163, + "learning_rate": 1.0746287522677439e-06, + "loss": 0.098, + "step": 5372 + }, + { + "epoch": 1.2225255972696245, + "grad_norm": 1.484110142108006, + "learning_rate": 1.074566685152133e-06, + "loss": 0.1136, + "step": 5373 + }, + { + "epoch": 1.2227531285551763, + "grad_norm": 0.860555011969557, + "learning_rate": 1.0745046088482025e-06, + "loss": 0.0212, + "step": 5374 + }, + { + "epoch": 1.222980659840728, + "grad_norm": 2.0013686983091783, + "learning_rate": 1.074442523357221e-06, + "loss": 0.0731, + "step": 5375 + }, + { + "epoch": 1.2232081911262798, + "grad_norm": 2.4325343000739887, + "learning_rate": 1.0743804286804573e-06, + "loss": 0.1035, + "step": 5376 + }, + { + "epoch": 1.2234357224118315, + "grad_norm": 1.9106606707808065, + "learning_rate": 1.0743183248191806e-06, + "loss": 0.0576, + "step": 5377 + }, + { + "epoch": 1.2236632536973833, + "grad_norm": 0.5769559312003509, + "learning_rate": 1.0742562117746604e-06, + "loss": 0.0339, + "step": 5378 + }, + { + "epoch": 1.2238907849829352, + "grad_norm": 1.2801763987646924, + "learning_rate": 1.074194089548166e-06, + "loss": 0.0796, + "step": 5379 + }, + { + "epoch": 1.224118316268487, + "grad_norm": 1.2678528066846348, + "learning_rate": 1.0741319581409667e-06, + "loss": 0.0785, + "step": 5380 + }, + { + "epoch": 1.2243458475540387, + "grad_norm": 1.866996912283029, + "learning_rate": 1.0740698175543332e-06, + "loss": 0.0697, + "step": 5381 + }, + { + "epoch": 1.2245733788395905, + "grad_norm": 1.5828941718449743, + "learning_rate": 1.0740076677895348e-06, + "loss": 0.095, + "step": 5382 + }, + { + "epoch": 1.2248009101251423, + "grad_norm": 2.0245338544598757, + "learning_rate": 1.0739455088478422e-06, + "loss": 0.0893, + "step": 5383 + }, + { + "epoch": 1.225028441410694, + "grad_norm": 1.9376486203134127, + "learning_rate": 1.0738833407305254e-06, + "loss": 0.125, + "step": 5384 + }, + { + "epoch": 1.2252559726962458, + "grad_norm": 1.2831894786317846, + "learning_rate": 1.0738211634388554e-06, + "loss": 0.061, + "step": 5385 + }, + { + "epoch": 1.2254835039817975, + "grad_norm": 2.28610050767493, + "learning_rate": 1.0737589769741025e-06, + "loss": 0.0829, + "step": 5386 + }, + { + "epoch": 1.2257110352673493, + "grad_norm": 2.079256703978429, + "learning_rate": 1.0736967813375382e-06, + "loss": 0.0933, + "step": 5387 + }, + { + "epoch": 1.225938566552901, + "grad_norm": 1.4133775667119193, + "learning_rate": 1.0736345765304335e-06, + "loss": 0.1665, + "step": 5388 + }, + { + "epoch": 1.2261660978384528, + "grad_norm": 1.8025993666918807, + "learning_rate": 1.0735723625540596e-06, + "loss": 0.1953, + "step": 5389 + }, + { + "epoch": 1.2263936291240045, + "grad_norm": 1.6168058770721976, + "learning_rate": 1.073510139409688e-06, + "loss": 0.0958, + "step": 5390 + }, + { + "epoch": 1.2266211604095563, + "grad_norm": 0.8413217564117436, + "learning_rate": 1.0734479070985908e-06, + "loss": 0.0231, + "step": 5391 + }, + { + "epoch": 1.226848691695108, + "grad_norm": 1.7545268518006023, + "learning_rate": 1.0733856656220396e-06, + "loss": 0.1106, + "step": 5392 + }, + { + "epoch": 1.2270762229806598, + "grad_norm": 0.9684484368779872, + "learning_rate": 1.0733234149813065e-06, + "loss": 0.0525, + "step": 5393 + }, + { + "epoch": 1.2273037542662115, + "grad_norm": 1.118301840198287, + "learning_rate": 1.0732611551776639e-06, + "loss": 0.0526, + "step": 5394 + }, + { + "epoch": 1.2275312855517633, + "grad_norm": 2.4499203280689064, + "learning_rate": 1.0731988862123841e-06, + "loss": 0.0808, + "step": 5395 + }, + { + "epoch": 1.2277588168373152, + "grad_norm": 1.8673149754872067, + "learning_rate": 1.07313660808674e-06, + "loss": 0.081, + "step": 5396 + }, + { + "epoch": 1.227986348122867, + "grad_norm": 1.187493720900378, + "learning_rate": 1.0730743208020044e-06, + "loss": 0.0467, + "step": 5397 + }, + { + "epoch": 1.2282138794084188, + "grad_norm": 1.5388729878525773, + "learning_rate": 1.0730120243594504e-06, + "loss": 0.0801, + "step": 5398 + }, + { + "epoch": 1.2284414106939705, + "grad_norm": 1.3665636981078553, + "learning_rate": 1.0729497187603508e-06, + "loss": 0.1399, + "step": 5399 + }, + { + "epoch": 1.2286689419795223, + "grad_norm": 2.1297488683384715, + "learning_rate": 1.0728874040059798e-06, + "loss": 0.0788, + "step": 5400 + }, + { + "epoch": 1.228896473265074, + "grad_norm": 2.7263173378574153, + "learning_rate": 1.07282508009761e-06, + "loss": 0.0686, + "step": 5401 + }, + { + "epoch": 1.2291240045506258, + "grad_norm": 1.2951500192928536, + "learning_rate": 1.072762747036516e-06, + "loss": 0.0842, + "step": 5402 + }, + { + "epoch": 1.2293515358361775, + "grad_norm": 1.952069542680283, + "learning_rate": 1.0727004048239715e-06, + "loss": 0.1264, + "step": 5403 + }, + { + "epoch": 1.2295790671217293, + "grad_norm": 1.6010547640917303, + "learning_rate": 1.0726380534612507e-06, + "loss": 0.0853, + "step": 5404 + }, + { + "epoch": 1.229806598407281, + "grad_norm": 1.9483808000546532, + "learning_rate": 1.0725756929496277e-06, + "loss": 0.0829, + "step": 5405 + }, + { + "epoch": 1.2300341296928328, + "grad_norm": 2.4565704934042647, + "learning_rate": 1.0725133232903773e-06, + "loss": 0.2031, + "step": 5406 + }, + { + "epoch": 1.2302616609783845, + "grad_norm": 1.1049855430303344, + "learning_rate": 1.0724509444847741e-06, + "loss": 0.1009, + "step": 5407 + }, + { + "epoch": 1.2304891922639363, + "grad_norm": 1.9135004329441547, + "learning_rate": 1.0723885565340933e-06, + "loss": 0.1154, + "step": 5408 + }, + { + "epoch": 1.230716723549488, + "grad_norm": 1.914997345530264, + "learning_rate": 1.0723261594396095e-06, + "loss": 0.1121, + "step": 5409 + }, + { + "epoch": 1.2309442548350398, + "grad_norm": 1.443690506471369, + "learning_rate": 1.0722637532025984e-06, + "loss": 0.1022, + "step": 5410 + }, + { + "epoch": 1.2311717861205915, + "grad_norm": 2.6024216163733582, + "learning_rate": 1.0722013378243354e-06, + "loss": 0.1794, + "step": 5411 + }, + { + "epoch": 1.2313993174061433, + "grad_norm": 3.20516486682456, + "learning_rate": 1.0721389133060958e-06, + "loss": 0.0982, + "step": 5412 + }, + { + "epoch": 1.231626848691695, + "grad_norm": 2.3908403467267445, + "learning_rate": 1.0720764796491559e-06, + "loss": 0.0561, + "step": 5413 + }, + { + "epoch": 1.2318543799772468, + "grad_norm": 2.498040600232711, + "learning_rate": 1.0720140368547915e-06, + "loss": 0.0705, + "step": 5414 + }, + { + "epoch": 1.2320819112627985, + "grad_norm": 1.9676099266840996, + "learning_rate": 1.0719515849242787e-06, + "loss": 0.1109, + "step": 5415 + }, + { + "epoch": 1.2323094425483503, + "grad_norm": 1.3576426657142715, + "learning_rate": 1.0718891238588943e-06, + "loss": 0.0617, + "step": 5416 + }, + { + "epoch": 1.232536973833902, + "grad_norm": 1.8753810587976327, + "learning_rate": 1.0718266536599145e-06, + "loss": 0.1394, + "step": 5417 + }, + { + "epoch": 1.232764505119454, + "grad_norm": 1.8597389433553986, + "learning_rate": 1.0717641743286163e-06, + "loss": 0.0849, + "step": 5418 + }, + { + "epoch": 1.2329920364050058, + "grad_norm": 1.922637481586932, + "learning_rate": 1.0717016858662766e-06, + "loss": 0.1043, + "step": 5419 + }, + { + "epoch": 1.2332195676905575, + "grad_norm": 1.620832607357832, + "learning_rate": 1.0716391882741722e-06, + "loss": 0.1114, + "step": 5420 + }, + { + "epoch": 1.2334470989761093, + "grad_norm": 2.0014578199306343, + "learning_rate": 1.0715766815535812e-06, + "loss": 0.1206, + "step": 5421 + }, + { + "epoch": 1.233674630261661, + "grad_norm": 1.2240310028768215, + "learning_rate": 1.0715141657057805e-06, + "loss": 0.137, + "step": 5422 + }, + { + "epoch": 1.2339021615472128, + "grad_norm": 2.138110039795511, + "learning_rate": 1.0714516407320482e-06, + "loss": 0.1787, + "step": 5423 + }, + { + "epoch": 1.2341296928327645, + "grad_norm": 2.316015876424966, + "learning_rate": 1.0713891066336619e-06, + "loss": 0.1035, + "step": 5424 + }, + { + "epoch": 1.2343572241183163, + "grad_norm": 1.9765888798589069, + "learning_rate": 1.0713265634118998e-06, + "loss": 0.1094, + "step": 5425 + }, + { + "epoch": 1.234584755403868, + "grad_norm": 1.7144437579313214, + "learning_rate": 1.0712640110680398e-06, + "loss": 0.0621, + "step": 5426 + }, + { + "epoch": 1.2348122866894198, + "grad_norm": 1.4697474263995853, + "learning_rate": 1.071201449603361e-06, + "loss": 0.1143, + "step": 5427 + }, + { + "epoch": 1.2350398179749715, + "grad_norm": 2.4457858543323154, + "learning_rate": 1.0711388790191418e-06, + "loss": 0.0991, + "step": 5428 + }, + { + "epoch": 1.2352673492605233, + "grad_norm": 1.4119744703675479, + "learning_rate": 1.071076299316661e-06, + "loss": 0.0973, + "step": 5429 + }, + { + "epoch": 1.235494880546075, + "grad_norm": 1.4595760481057491, + "learning_rate": 1.0710137104971973e-06, + "loss": 0.0999, + "step": 5430 + }, + { + "epoch": 1.2357224118316268, + "grad_norm": 1.5583570046816027, + "learning_rate": 1.0709511125620306e-06, + "loss": 0.124, + "step": 5431 + }, + { + "epoch": 1.2359499431171785, + "grad_norm": 1.843599923959497, + "learning_rate": 1.0708885055124396e-06, + "loss": 0.0858, + "step": 5432 + }, + { + "epoch": 1.2361774744027303, + "grad_norm": 1.1999122894890224, + "learning_rate": 1.0708258893497043e-06, + "loss": 0.0479, + "step": 5433 + }, + { + "epoch": 1.236405005688282, + "grad_norm": 1.457239578244025, + "learning_rate": 1.0707632640751042e-06, + "loss": 0.0449, + "step": 5434 + }, + { + "epoch": 1.236632536973834, + "grad_norm": 1.6945912200555417, + "learning_rate": 1.0707006296899194e-06, + "loss": 0.081, + "step": 5435 + }, + { + "epoch": 1.2368600682593858, + "grad_norm": 2.6001020024335815, + "learning_rate": 1.0706379861954299e-06, + "loss": 0.2526, + "step": 5436 + }, + { + "epoch": 1.2370875995449375, + "grad_norm": 1.3492096197419317, + "learning_rate": 1.0705753335929162e-06, + "loss": 0.1496, + "step": 5437 + }, + { + "epoch": 1.2373151308304893, + "grad_norm": 1.6010914508078664, + "learning_rate": 1.070512671883659e-06, + "loss": 0.0601, + "step": 5438 + }, + { + "epoch": 1.237542662116041, + "grad_norm": 0.9988092205638144, + "learning_rate": 1.0704500010689383e-06, + "loss": 0.0708, + "step": 5439 + }, + { + "epoch": 1.2377701934015928, + "grad_norm": 2.0272592593408856, + "learning_rate": 1.0703873211500356e-06, + "loss": 0.0981, + "step": 5440 + }, + { + "epoch": 1.2379977246871445, + "grad_norm": 1.5814448790855402, + "learning_rate": 1.0703246321282316e-06, + "loss": 0.0647, + "step": 5441 + }, + { + "epoch": 1.2382252559726963, + "grad_norm": 1.3583333002765134, + "learning_rate": 1.0702619340048077e-06, + "loss": 0.0937, + "step": 5442 + }, + { + "epoch": 1.238452787258248, + "grad_norm": 1.6595104474776738, + "learning_rate": 1.0701992267810454e-06, + "loss": 0.0559, + "step": 5443 + }, + { + "epoch": 1.2386803185437998, + "grad_norm": 1.6197295051794525, + "learning_rate": 1.0701365104582262e-06, + "loss": 0.154, + "step": 5444 + }, + { + "epoch": 1.2389078498293515, + "grad_norm": 1.2894140286449491, + "learning_rate": 1.070073785037632e-06, + "loss": 0.0704, + "step": 5445 + }, + { + "epoch": 1.2391353811149033, + "grad_norm": 1.5941117277473875, + "learning_rate": 1.0700110505205447e-06, + "loss": 0.0839, + "step": 5446 + }, + { + "epoch": 1.239362912400455, + "grad_norm": 1.7878642941689529, + "learning_rate": 1.0699483069082468e-06, + "loss": 0.0649, + "step": 5447 + }, + { + "epoch": 1.2395904436860068, + "grad_norm": 2.599750455430904, + "learning_rate": 1.0698855542020201e-06, + "loss": 0.1301, + "step": 5448 + }, + { + "epoch": 1.2398179749715585, + "grad_norm": 2.0450552321843025, + "learning_rate": 1.0698227924031474e-06, + "loss": 0.1969, + "step": 5449 + }, + { + "epoch": 1.2400455062571103, + "grad_norm": 1.2281153831660558, + "learning_rate": 1.0697600215129113e-06, + "loss": 0.0613, + "step": 5450 + }, + { + "epoch": 1.240273037542662, + "grad_norm": 2.143751448330337, + "learning_rate": 1.0696972415325954e-06, + "loss": 0.0722, + "step": 5451 + }, + { + "epoch": 1.2405005688282138, + "grad_norm": 1.322311765084235, + "learning_rate": 1.069634452463482e-06, + "loss": 0.1549, + "step": 5452 + }, + { + "epoch": 1.2407281001137656, + "grad_norm": 1.6004401205919885, + "learning_rate": 1.0695716543068548e-06, + "loss": 0.1183, + "step": 5453 + }, + { + "epoch": 1.2409556313993173, + "grad_norm": 1.3942986796389523, + "learning_rate": 1.0695088470639973e-06, + "loss": 0.0864, + "step": 5454 + }, + { + "epoch": 1.241183162684869, + "grad_norm": 2.578779304276581, + "learning_rate": 1.069446030736193e-06, + "loss": 0.1039, + "step": 5455 + }, + { + "epoch": 1.2414106939704208, + "grad_norm": 1.504902639716046, + "learning_rate": 1.0693832053247256e-06, + "loss": 0.0512, + "step": 5456 + }, + { + "epoch": 1.2416382252559728, + "grad_norm": 2.6715606976374646, + "learning_rate": 1.0693203708308792e-06, + "loss": 0.0914, + "step": 5457 + }, + { + "epoch": 1.2418657565415245, + "grad_norm": 1.8134513201348859, + "learning_rate": 1.0692575272559385e-06, + "loss": 0.0939, + "step": 5458 + }, + { + "epoch": 1.2420932878270763, + "grad_norm": 2.353744451889051, + "learning_rate": 1.0691946746011874e-06, + "loss": 0.0906, + "step": 5459 + }, + { + "epoch": 1.242320819112628, + "grad_norm": 2.066531174712776, + "learning_rate": 1.0691318128679107e-06, + "loss": 0.1075, + "step": 5460 + }, + { + "epoch": 1.2425483503981798, + "grad_norm": 1.8513813952978166, + "learning_rate": 1.0690689420573933e-06, + "loss": 0.1163, + "step": 5461 + }, + { + "epoch": 1.2427758816837315, + "grad_norm": 2.4822445733516143, + "learning_rate": 1.0690060621709198e-06, + "loss": 0.0855, + "step": 5462 + }, + { + "epoch": 1.2430034129692833, + "grad_norm": 1.9977462685240035, + "learning_rate": 1.0689431732097754e-06, + "loss": 0.1, + "step": 5463 + }, + { + "epoch": 1.243230944254835, + "grad_norm": 1.6816887574312476, + "learning_rate": 1.0688802751752458e-06, + "loss": 0.1155, + "step": 5464 + }, + { + "epoch": 1.2434584755403868, + "grad_norm": 1.6544785097209609, + "learning_rate": 1.0688173680686164e-06, + "loss": 0.0555, + "step": 5465 + }, + { + "epoch": 1.2436860068259386, + "grad_norm": 4.858334551394226, + "learning_rate": 1.0687544518911726e-06, + "loss": 0.0815, + "step": 5466 + }, + { + "epoch": 1.2439135381114903, + "grad_norm": 2.2928222165455883, + "learning_rate": 1.0686915266442005e-06, + "loss": 0.0874, + "step": 5467 + }, + { + "epoch": 1.244141069397042, + "grad_norm": 1.7968322866387734, + "learning_rate": 1.0686285923289863e-06, + "loss": 0.1786, + "step": 5468 + }, + { + "epoch": 1.2443686006825938, + "grad_norm": 1.8341969945123524, + "learning_rate": 1.0685656489468161e-06, + "loss": 0.0728, + "step": 5469 + }, + { + "epoch": 1.2445961319681456, + "grad_norm": 1.9808057129983412, + "learning_rate": 1.0685026964989764e-06, + "loss": 0.0614, + "step": 5470 + }, + { + "epoch": 1.2448236632536973, + "grad_norm": 2.3603917978939686, + "learning_rate": 1.0684397349867537e-06, + "loss": 0.12, + "step": 5471 + }, + { + "epoch": 1.245051194539249, + "grad_norm": 1.3367535002517599, + "learning_rate": 1.068376764411435e-06, + "loss": 0.0885, + "step": 5472 + }, + { + "epoch": 1.2452787258248008, + "grad_norm": 1.397713582074191, + "learning_rate": 1.0683137847743076e-06, + "loss": 0.142, + "step": 5473 + }, + { + "epoch": 1.2455062571103528, + "grad_norm": 1.6484425230118132, + "learning_rate": 1.0682507960766578e-06, + "loss": 0.0859, + "step": 5474 + }, + { + "epoch": 1.2457337883959045, + "grad_norm": 1.4069345626600442, + "learning_rate": 1.0681877983197738e-06, + "loss": 0.1161, + "step": 5475 + }, + { + "epoch": 1.2459613196814563, + "grad_norm": 1.4998871720536815, + "learning_rate": 1.0681247915049428e-06, + "loss": 0.0772, + "step": 5476 + }, + { + "epoch": 1.246188850967008, + "grad_norm": 1.9389078200268701, + "learning_rate": 1.0680617756334527e-06, + "loss": 0.0716, + "step": 5477 + }, + { + "epoch": 1.2464163822525598, + "grad_norm": 2.7660431376565695, + "learning_rate": 1.0679987507065912e-06, + "loss": 0.1567, + "step": 5478 + }, + { + "epoch": 1.2466439135381115, + "grad_norm": 2.4466457466496205, + "learning_rate": 1.0679357167256465e-06, + "loss": 0.0861, + "step": 5479 + }, + { + "epoch": 1.2468714448236633, + "grad_norm": 2.243817072525832, + "learning_rate": 1.067872673691907e-06, + "loss": 0.06, + "step": 5480 + }, + { + "epoch": 1.247098976109215, + "grad_norm": 1.1074974938970048, + "learning_rate": 1.0678096216066611e-06, + "loss": 0.0617, + "step": 5481 + }, + { + "epoch": 1.2473265073947668, + "grad_norm": 1.5816842034349576, + "learning_rate": 1.0677465604711975e-06, + "loss": 0.1165, + "step": 5482 + }, + { + "epoch": 1.2475540386803186, + "grad_norm": 1.6815250673847648, + "learning_rate": 1.0676834902868051e-06, + "loss": 0.0966, + "step": 5483 + }, + { + "epoch": 1.2477815699658703, + "grad_norm": 1.8948162149491237, + "learning_rate": 1.067620411054773e-06, + "loss": 0.1998, + "step": 5484 + }, + { + "epoch": 1.248009101251422, + "grad_norm": 1.5218119585260081, + "learning_rate": 1.0675573227763903e-06, + "loss": 0.1333, + "step": 5485 + }, + { + "epoch": 1.2482366325369738, + "grad_norm": 1.894461051388274, + "learning_rate": 1.0674942254529463e-06, + "loss": 0.1003, + "step": 5486 + }, + { + "epoch": 1.2484641638225256, + "grad_norm": 2.2504857522339297, + "learning_rate": 1.0674311190857308e-06, + "loss": 0.1372, + "step": 5487 + }, + { + "epoch": 1.2486916951080773, + "grad_norm": 1.5857559144795927, + "learning_rate": 1.0673680036760333e-06, + "loss": 0.1743, + "step": 5488 + }, + { + "epoch": 1.248919226393629, + "grad_norm": 1.7267319127098426, + "learning_rate": 1.0673048792251443e-06, + "loss": 0.0452, + "step": 5489 + }, + { + "epoch": 1.2491467576791808, + "grad_norm": 1.7855570682733017, + "learning_rate": 1.0672417457343535e-06, + "loss": 0.0699, + "step": 5490 + }, + { + "epoch": 1.2493742889647326, + "grad_norm": 1.775936905731107, + "learning_rate": 1.0671786032049512e-06, + "loss": 0.0854, + "step": 5491 + }, + { + "epoch": 1.2496018202502843, + "grad_norm": 2.2449973755366126, + "learning_rate": 1.0671154516382283e-06, + "loss": 0.0941, + "step": 5492 + }, + { + "epoch": 1.249829351535836, + "grad_norm": 1.4681093724936798, + "learning_rate": 1.067052291035475e-06, + "loss": 0.0744, + "step": 5493 + }, + { + "epoch": 1.2500568828213878, + "grad_norm": 1.4102566771482665, + "learning_rate": 1.0669891213979826e-06, + "loss": 0.0794, + "step": 5494 + }, + { + "epoch": 1.2502844141069396, + "grad_norm": 1.7355330983101092, + "learning_rate": 1.066925942727042e-06, + "loss": 0.0578, + "step": 5495 + }, + { + "epoch": 1.2505119453924913, + "grad_norm": 2.3854746229810497, + "learning_rate": 1.0668627550239444e-06, + "loss": 0.0943, + "step": 5496 + }, + { + "epoch": 1.2507394766780433, + "grad_norm": 1.5833507298430867, + "learning_rate": 1.0667995582899815e-06, + "loss": 0.0441, + "step": 5497 + }, + { + "epoch": 1.250967007963595, + "grad_norm": 2.0083629771890057, + "learning_rate": 1.0667363525264446e-06, + "loss": 0.0733, + "step": 5498 + }, + { + "epoch": 1.2511945392491468, + "grad_norm": 1.2971360177807068, + "learning_rate": 1.0666731377346257e-06, + "loss": 0.1036, + "step": 5499 + }, + { + "epoch": 1.2514220705346986, + "grad_norm": 2.800119901797372, + "learning_rate": 1.0666099139158168e-06, + "loss": 0.1237, + "step": 5500 + }, + { + "epoch": 1.2516496018202503, + "grad_norm": 1.8870897807924603, + "learning_rate": 1.06654668107131e-06, + "loss": 0.086, + "step": 5501 + }, + { + "epoch": 1.251877133105802, + "grad_norm": 2.290616338788931, + "learning_rate": 1.0664834392023975e-06, + "loss": 0.1499, + "step": 5502 + }, + { + "epoch": 1.2521046643913538, + "grad_norm": 2.0803939807047613, + "learning_rate": 1.0664201883103722e-06, + "loss": 0.1016, + "step": 5503 + }, + { + "epoch": 1.2523321956769056, + "grad_norm": 1.9461188419368005, + "learning_rate": 1.066356928396527e-06, + "loss": 0.0614, + "step": 5504 + }, + { + "epoch": 1.2525597269624573, + "grad_norm": 2.2124110257952405, + "learning_rate": 1.066293659462154e-06, + "loss": 0.0948, + "step": 5505 + }, + { + "epoch": 1.252787258248009, + "grad_norm": 2.0599721227236443, + "learning_rate": 1.066230381508547e-06, + "loss": 0.1591, + "step": 5506 + }, + { + "epoch": 1.2530147895335608, + "grad_norm": 1.8004176359521435, + "learning_rate": 1.0661670945369991e-06, + "loss": 0.1559, + "step": 5507 + }, + { + "epoch": 1.2532423208191126, + "grad_norm": 1.1888930303857044, + "learning_rate": 1.0661037985488037e-06, + "loss": 0.0518, + "step": 5508 + }, + { + "epoch": 1.2534698521046643, + "grad_norm": 1.8289874247099218, + "learning_rate": 1.0660404935452545e-06, + "loss": 0.0546, + "step": 5509 + }, + { + "epoch": 1.253697383390216, + "grad_norm": 1.1560218852925717, + "learning_rate": 1.0659771795276451e-06, + "loss": 0.1254, + "step": 5510 + }, + { + "epoch": 1.253924914675768, + "grad_norm": 1.2249808857764073, + "learning_rate": 1.06591385649727e-06, + "loss": 0.0676, + "step": 5511 + }, + { + "epoch": 1.2541524459613198, + "grad_norm": 0.85050849951928, + "learning_rate": 1.0658505244554233e-06, + "loss": 0.093, + "step": 5512 + }, + { + "epoch": 1.2543799772468716, + "grad_norm": 1.5438548423850245, + "learning_rate": 1.0657871834033992e-06, + "loss": 0.125, + "step": 5513 + }, + { + "epoch": 1.2546075085324233, + "grad_norm": 0.9078275355366267, + "learning_rate": 1.0657238333424922e-06, + "loss": 0.0376, + "step": 5514 + }, + { + "epoch": 1.254835039817975, + "grad_norm": 2.2143429786465063, + "learning_rate": 1.0656604742739974e-06, + "loss": 0.0724, + "step": 5515 + }, + { + "epoch": 1.2550625711035268, + "grad_norm": 2.0486216037329146, + "learning_rate": 1.0655971061992093e-06, + "loss": 0.1584, + "step": 5516 + }, + { + "epoch": 1.2552901023890786, + "grad_norm": 1.997450488802889, + "learning_rate": 1.0655337291194235e-06, + "loss": 0.0802, + "step": 5517 + }, + { + "epoch": 1.2555176336746303, + "grad_norm": 1.1050951081736293, + "learning_rate": 1.0654703430359348e-06, + "loss": 0.077, + "step": 5518 + }, + { + "epoch": 1.255745164960182, + "grad_norm": 2.039605618688286, + "learning_rate": 1.065406947950039e-06, + "loss": 0.0795, + "step": 5519 + }, + { + "epoch": 1.2559726962457338, + "grad_norm": 1.518327874025393, + "learning_rate": 1.065343543863032e-06, + "loss": 0.0779, + "step": 5520 + }, + { + "epoch": 1.2562002275312856, + "grad_norm": 1.3387099786697993, + "learning_rate": 1.0652801307762093e-06, + "loss": 0.0931, + "step": 5521 + }, + { + "epoch": 1.2564277588168373, + "grad_norm": 3.272032880936236, + "learning_rate": 1.065216708690867e-06, + "loss": 0.1039, + "step": 5522 + }, + { + "epoch": 1.256655290102389, + "grad_norm": 1.9746026723150696, + "learning_rate": 1.0651532776083014e-06, + "loss": 0.1066, + "step": 5523 + }, + { + "epoch": 1.2568828213879408, + "grad_norm": 1.5931847333482678, + "learning_rate": 1.0650898375298088e-06, + "loss": 0.0763, + "step": 5524 + }, + { + "epoch": 1.2571103526734926, + "grad_norm": 1.8732431674607817, + "learning_rate": 1.0650263884566863e-06, + "loss": 0.0763, + "step": 5525 + }, + { + "epoch": 1.2573378839590443, + "grad_norm": 1.8693038945153335, + "learning_rate": 1.06496293039023e-06, + "loss": 0.1149, + "step": 5526 + }, + { + "epoch": 1.257565415244596, + "grad_norm": 1.4481248834894889, + "learning_rate": 1.0648994633317373e-06, + "loss": 0.1457, + "step": 5527 + }, + { + "epoch": 1.2577929465301478, + "grad_norm": 1.862432920570466, + "learning_rate": 1.064835987282505e-06, + "loss": 0.1352, + "step": 5528 + }, + { + "epoch": 1.2580204778156996, + "grad_norm": 1.2450473886682687, + "learning_rate": 1.0647725022438307e-06, + "loss": 0.0673, + "step": 5529 + }, + { + "epoch": 1.2582480091012513, + "grad_norm": 1.752937419466306, + "learning_rate": 1.0647090082170118e-06, + "loss": 0.0803, + "step": 5530 + }, + { + "epoch": 1.258475540386803, + "grad_norm": 1.7576675405029405, + "learning_rate": 1.0646455052033463e-06, + "loss": 0.1399, + "step": 5531 + }, + { + "epoch": 1.2587030716723548, + "grad_norm": 1.3799958444972384, + "learning_rate": 1.0645819932041317e-06, + "loss": 0.1049, + "step": 5532 + }, + { + "epoch": 1.2589306029579066, + "grad_norm": 1.0385789823026492, + "learning_rate": 1.064518472220666e-06, + "loss": 0.0457, + "step": 5533 + }, + { + "epoch": 1.2591581342434583, + "grad_norm": 2.019081092458579, + "learning_rate": 1.064454942254248e-06, + "loss": 0.0607, + "step": 5534 + }, + { + "epoch": 1.25938566552901, + "grad_norm": 2.6223562992885108, + "learning_rate": 1.0643914033061757e-06, + "loss": 0.0976, + "step": 5535 + }, + { + "epoch": 1.259613196814562, + "grad_norm": 1.1669553650719042, + "learning_rate": 1.0643278553777477e-06, + "loss": 0.0836, + "step": 5536 + }, + { + "epoch": 1.2598407281001138, + "grad_norm": 3.0375633482678013, + "learning_rate": 1.0642642984702632e-06, + "loss": 0.1243, + "step": 5537 + }, + { + "epoch": 1.2600682593856656, + "grad_norm": 1.168545824136467, + "learning_rate": 1.0642007325850207e-06, + "loss": 0.083, + "step": 5538 + }, + { + "epoch": 1.2602957906712173, + "grad_norm": 2.2256486202114676, + "learning_rate": 1.0641371577233197e-06, + "loss": 0.1316, + "step": 5539 + }, + { + "epoch": 1.260523321956769, + "grad_norm": 2.136519087012331, + "learning_rate": 1.0640735738864594e-06, + "loss": 0.071, + "step": 5540 + }, + { + "epoch": 1.2607508532423208, + "grad_norm": 1.6283300874870004, + "learning_rate": 1.0640099810757394e-06, + "loss": 0.0911, + "step": 5541 + }, + { + "epoch": 1.2609783845278726, + "grad_norm": 1.1416528868312568, + "learning_rate": 1.0639463792924592e-06, + "loss": 0.0657, + "step": 5542 + }, + { + "epoch": 1.2612059158134243, + "grad_norm": 3.4581463177480702, + "learning_rate": 1.0638827685379191e-06, + "loss": 0.0751, + "step": 5543 + }, + { + "epoch": 1.261433447098976, + "grad_norm": 1.6347126585276193, + "learning_rate": 1.063819148813419e-06, + "loss": 0.0988, + "step": 5544 + }, + { + "epoch": 1.2616609783845278, + "grad_norm": 1.9297479475314918, + "learning_rate": 1.063755520120259e-06, + "loss": 0.1323, + "step": 5545 + }, + { + "epoch": 1.2618885096700796, + "grad_norm": 1.4945227351451722, + "learning_rate": 1.0636918824597397e-06, + "loss": 0.074, + "step": 5546 + }, + { + "epoch": 1.2621160409556313, + "grad_norm": 1.2720191090234823, + "learning_rate": 1.063628235833162e-06, + "loss": 0.0211, + "step": 5547 + }, + { + "epoch": 1.262343572241183, + "grad_norm": 2.0454700040014284, + "learning_rate": 1.0635645802418263e-06, + "loss": 0.1092, + "step": 5548 + }, + { + "epoch": 1.2625711035267349, + "grad_norm": 3.526245551948375, + "learning_rate": 1.0635009156870338e-06, + "loss": 0.122, + "step": 5549 + }, + { + "epoch": 1.2627986348122868, + "grad_norm": 1.0127431108984934, + "learning_rate": 1.0634372421700858e-06, + "loss": 0.0495, + "step": 5550 + }, + { + "epoch": 1.2630261660978386, + "grad_norm": 1.5378058015868816, + "learning_rate": 1.0633735596922834e-06, + "loss": 0.1009, + "step": 5551 + }, + { + "epoch": 1.2632536973833903, + "grad_norm": 1.0282183974082975, + "learning_rate": 1.0633098682549282e-06, + "loss": 0.0469, + "step": 5552 + }, + { + "epoch": 1.263481228668942, + "grad_norm": 1.402805359819934, + "learning_rate": 1.063246167859322e-06, + "loss": 0.0961, + "step": 5553 + }, + { + "epoch": 1.2637087599544938, + "grad_norm": 1.6230015516881557, + "learning_rate": 1.0631824585067668e-06, + "loss": 0.0797, + "step": 5554 + }, + { + "epoch": 1.2639362912400456, + "grad_norm": 1.2909297831561177, + "learning_rate": 1.0631187401985647e-06, + "loss": 0.0845, + "step": 5555 + }, + { + "epoch": 1.2641638225255973, + "grad_norm": 1.6465161376952524, + "learning_rate": 1.0630550129360179e-06, + "loss": 0.0516, + "step": 5556 + }, + { + "epoch": 1.264391353811149, + "grad_norm": 1.3793066474762286, + "learning_rate": 1.062991276720429e-06, + "loss": 0.0632, + "step": 5557 + }, + { + "epoch": 1.2646188850967008, + "grad_norm": 1.8076204975362697, + "learning_rate": 1.0629275315531005e-06, + "loss": 0.0838, + "step": 5558 + }, + { + "epoch": 1.2648464163822526, + "grad_norm": 1.4158184088750396, + "learning_rate": 1.0628637774353351e-06, + "loss": 0.0835, + "step": 5559 + }, + { + "epoch": 1.2650739476678043, + "grad_norm": 1.4013673117505225, + "learning_rate": 1.062800014368436e-06, + "loss": 0.0642, + "step": 5560 + }, + { + "epoch": 1.265301478953356, + "grad_norm": 1.391664380207402, + "learning_rate": 1.0627362423537065e-06, + "loss": 0.0977, + "step": 5561 + }, + { + "epoch": 1.2655290102389078, + "grad_norm": 2.030166330980807, + "learning_rate": 1.06267246139245e-06, + "loss": 0.1186, + "step": 5562 + }, + { + "epoch": 1.2657565415244596, + "grad_norm": 1.677270778893888, + "learning_rate": 1.06260867148597e-06, + "loss": 0.0582, + "step": 5563 + }, + { + "epoch": 1.2659840728100114, + "grad_norm": 2.386010353958561, + "learning_rate": 1.06254487263557e-06, + "loss": 0.0916, + "step": 5564 + }, + { + "epoch": 1.266211604095563, + "grad_norm": 1.2929316919365015, + "learning_rate": 1.062481064842554e-06, + "loss": 0.0915, + "step": 5565 + }, + { + "epoch": 1.2664391353811149, + "grad_norm": 1.8947340385343405, + "learning_rate": 1.0624172481082265e-06, + "loss": 0.073, + "step": 5566 + }, + { + "epoch": 1.2666666666666666, + "grad_norm": 1.9806391604348212, + "learning_rate": 1.0623534224338916e-06, + "loss": 0.1069, + "step": 5567 + }, + { + "epoch": 1.2668941979522184, + "grad_norm": 2.0768859821030805, + "learning_rate": 1.0622895878208535e-06, + "loss": 0.1271, + "step": 5568 + }, + { + "epoch": 1.2671217292377701, + "grad_norm": 1.7245435012523789, + "learning_rate": 1.0622257442704174e-06, + "loss": 0.1509, + "step": 5569 + }, + { + "epoch": 1.2673492605233219, + "grad_norm": 1.7895301814757045, + "learning_rate": 1.0621618917838874e-06, + "loss": 0.1585, + "step": 5570 + }, + { + "epoch": 1.2675767918088736, + "grad_norm": 1.6904234075905629, + "learning_rate": 1.062098030362569e-06, + "loss": 0.0695, + "step": 5571 + }, + { + "epoch": 1.2678043230944254, + "grad_norm": 2.4457978209810913, + "learning_rate": 1.0620341600077675e-06, + "loss": 0.1008, + "step": 5572 + }, + { + "epoch": 1.2680318543799771, + "grad_norm": 2.161084791259615, + "learning_rate": 1.0619702807207881e-06, + "loss": 0.0829, + "step": 5573 + }, + { + "epoch": 1.268259385665529, + "grad_norm": 2.2600901965292937, + "learning_rate": 1.0619063925029367e-06, + "loss": 0.068, + "step": 5574 + }, + { + "epoch": 1.2684869169510808, + "grad_norm": 1.0840331512965, + "learning_rate": 1.0618424953555186e-06, + "loss": 0.0616, + "step": 5575 + }, + { + "epoch": 1.2687144482366326, + "grad_norm": 1.2319354830540028, + "learning_rate": 1.0617785892798399e-06, + "loss": 0.0385, + "step": 5576 + }, + { + "epoch": 1.2689419795221843, + "grad_norm": 1.5534769449009878, + "learning_rate": 1.061714674277207e-06, + "loss": 0.0855, + "step": 5577 + }, + { + "epoch": 1.269169510807736, + "grad_norm": 2.4306350480501657, + "learning_rate": 1.0616507503489255e-06, + "loss": 0.0877, + "step": 5578 + }, + { + "epoch": 1.2693970420932879, + "grad_norm": 1.7942644238062382, + "learning_rate": 1.0615868174963025e-06, + "loss": 0.0649, + "step": 5579 + }, + { + "epoch": 1.2696245733788396, + "grad_norm": 2.5654321765997885, + "learning_rate": 1.0615228757206448e-06, + "loss": 0.1688, + "step": 5580 + }, + { + "epoch": 1.2698521046643914, + "grad_norm": 1.6812385027743317, + "learning_rate": 1.0614589250232588e-06, + "loss": 0.1631, + "step": 5581 + }, + { + "epoch": 1.270079635949943, + "grad_norm": 1.4665264164731382, + "learning_rate": 1.0613949654054518e-06, + "loss": 0.0838, + "step": 5582 + }, + { + "epoch": 1.2703071672354949, + "grad_norm": 2.560767700352155, + "learning_rate": 1.0613309968685308e-06, + "loss": 0.1086, + "step": 5583 + }, + { + "epoch": 1.2705346985210466, + "grad_norm": 1.5508544451796749, + "learning_rate": 1.0612670194138033e-06, + "loss": 0.0702, + "step": 5584 + }, + { + "epoch": 1.2707622298065984, + "grad_norm": 2.1461634004729837, + "learning_rate": 1.061203033042577e-06, + "loss": 0.0986, + "step": 5585 + }, + { + "epoch": 1.2709897610921501, + "grad_norm": 2.4218915402318193, + "learning_rate": 1.0611390377561596e-06, + "loss": 0.1215, + "step": 5586 + }, + { + "epoch": 1.2712172923777019, + "grad_norm": 0.6658643110079683, + "learning_rate": 1.0610750335558589e-06, + "loss": 0.0471, + "step": 5587 + }, + { + "epoch": 1.2714448236632536, + "grad_norm": 1.3141548979843396, + "learning_rate": 1.0610110204429832e-06, + "loss": 0.071, + "step": 5588 + }, + { + "epoch": 1.2716723549488056, + "grad_norm": 1.7639431438675843, + "learning_rate": 1.060946998418841e-06, + "loss": 0.1114, + "step": 5589 + }, + { + "epoch": 1.2718998862343573, + "grad_norm": 1.506178749665914, + "learning_rate": 1.06088296748474e-06, + "loss": 0.0562, + "step": 5590 + }, + { + "epoch": 1.272127417519909, + "grad_norm": 1.9579583267558633, + "learning_rate": 1.0608189276419898e-06, + "loss": 0.0871, + "step": 5591 + }, + { + "epoch": 1.2723549488054609, + "grad_norm": 2.0110213837918134, + "learning_rate": 1.0607548788918989e-06, + "loss": 0.1181, + "step": 5592 + }, + { + "epoch": 1.2725824800910126, + "grad_norm": 2.0737318256837542, + "learning_rate": 1.0606908212357764e-06, + "loss": 0.1301, + "step": 5593 + }, + { + "epoch": 1.2728100113765644, + "grad_norm": 1.9325897325011154, + "learning_rate": 1.0606267546749312e-06, + "loss": 0.0946, + "step": 5594 + }, + { + "epoch": 1.273037542662116, + "grad_norm": 1.9916873026614195, + "learning_rate": 1.0605626792106729e-06, + "loss": 0.0904, + "step": 5595 + }, + { + "epoch": 1.2732650739476679, + "grad_norm": 2.584223969309277, + "learning_rate": 1.0604985948443115e-06, + "loss": 0.2029, + "step": 5596 + }, + { + "epoch": 1.2734926052332196, + "grad_norm": 2.407891422443583, + "learning_rate": 1.0604345015771561e-06, + "loss": 0.097, + "step": 5597 + }, + { + "epoch": 1.2737201365187714, + "grad_norm": 1.8811406288698242, + "learning_rate": 1.060370399410517e-06, + "loss": 0.1304, + "step": 5598 + }, + { + "epoch": 1.2739476678043231, + "grad_norm": 1.0291908406249004, + "learning_rate": 1.0603062883457044e-06, + "loss": 0.0576, + "step": 5599 + }, + { + "epoch": 1.2741751990898749, + "grad_norm": 1.760148573333007, + "learning_rate": 1.0602421683840283e-06, + "loss": 0.0556, + "step": 5600 + }, + { + "epoch": 1.2744027303754266, + "grad_norm": 3.241467227070589, + "learning_rate": 1.0601780395267997e-06, + "loss": 0.0584, + "step": 5601 + }, + { + "epoch": 1.2746302616609784, + "grad_norm": 1.5240599185793644, + "learning_rate": 1.0601139017753286e-06, + "loss": 0.0954, + "step": 5602 + }, + { + "epoch": 1.2748577929465301, + "grad_norm": 1.5993856935822532, + "learning_rate": 1.0600497551309263e-06, + "loss": 0.0654, + "step": 5603 + }, + { + "epoch": 1.2750853242320819, + "grad_norm": 1.174588757527884, + "learning_rate": 1.0599855995949038e-06, + "loss": 0.1085, + "step": 5604 + }, + { + "epoch": 1.2753128555176336, + "grad_norm": 1.5969271473918978, + "learning_rate": 1.0599214351685724e-06, + "loss": 0.0532, + "step": 5605 + }, + { + "epoch": 1.2755403868031854, + "grad_norm": 1.2741242300770832, + "learning_rate": 1.0598572618532433e-06, + "loss": 0.0927, + "step": 5606 + }, + { + "epoch": 1.2757679180887371, + "grad_norm": 1.970426944694558, + "learning_rate": 1.0597930796502282e-06, + "loss": 0.059, + "step": 5607 + }, + { + "epoch": 1.2759954493742889, + "grad_norm": 1.7974688284252667, + "learning_rate": 1.059728888560839e-06, + "loss": 0.0546, + "step": 5608 + }, + { + "epoch": 1.2762229806598406, + "grad_norm": 2.347519835833904, + "learning_rate": 1.059664688586387e-06, + "loss": 0.0678, + "step": 5609 + }, + { + "epoch": 1.2764505119453924, + "grad_norm": 1.9966964077823932, + "learning_rate": 1.0596004797281853e-06, + "loss": 0.1171, + "step": 5610 + }, + { + "epoch": 1.2766780432309441, + "grad_norm": 2.3005969263882973, + "learning_rate": 1.0595362619875455e-06, + "loss": 0.1085, + "step": 5611 + }, + { + "epoch": 1.276905574516496, + "grad_norm": 1.6333485760567956, + "learning_rate": 1.0594720353657802e-06, + "loss": 0.117, + "step": 5612 + }, + { + "epoch": 1.2771331058020479, + "grad_norm": 1.9147011409795431, + "learning_rate": 1.0594077998642025e-06, + "loss": 0.0874, + "step": 5613 + }, + { + "epoch": 1.2773606370875996, + "grad_norm": 2.1078634316924187, + "learning_rate": 1.0593435554841247e-06, + "loss": 0.2271, + "step": 5614 + }, + { + "epoch": 1.2775881683731514, + "grad_norm": 2.022126628668214, + "learning_rate": 1.0592793022268603e-06, + "loss": 0.0842, + "step": 5615 + }, + { + "epoch": 1.2778156996587031, + "grad_norm": 1.9052082685253122, + "learning_rate": 1.0592150400937222e-06, + "loss": 0.099, + "step": 5616 + }, + { + "epoch": 1.2780432309442549, + "grad_norm": 1.4084668092686947, + "learning_rate": 1.059150769086024e-06, + "loss": 0.1706, + "step": 5617 + }, + { + "epoch": 1.2782707622298066, + "grad_norm": 2.5181732763637887, + "learning_rate": 1.059086489205079e-06, + "loss": 0.1101, + "step": 5618 + }, + { + "epoch": 1.2784982935153584, + "grad_norm": 1.438424800867154, + "learning_rate": 1.0590222004522012e-06, + "loss": 0.0597, + "step": 5619 + }, + { + "epoch": 1.2787258248009101, + "grad_norm": 2.542196122167591, + "learning_rate": 1.0589579028287045e-06, + "loss": 0.1012, + "step": 5620 + }, + { + "epoch": 1.2789533560864619, + "grad_norm": 1.9099915285362725, + "learning_rate": 1.0588935963359032e-06, + "loss": 0.123, + "step": 5621 + }, + { + "epoch": 1.2791808873720136, + "grad_norm": 1.409727510364223, + "learning_rate": 1.0588292809751112e-06, + "loss": 0.1016, + "step": 5622 + }, + { + "epoch": 1.2794084186575654, + "grad_norm": 1.3690280829539982, + "learning_rate": 1.0587649567476434e-06, + "loss": 0.0966, + "step": 5623 + }, + { + "epoch": 1.2796359499431171, + "grad_norm": 1.5389585621222681, + "learning_rate": 1.0587006236548142e-06, + "loss": 0.0743, + "step": 5624 + }, + { + "epoch": 1.2798634812286689, + "grad_norm": 1.7026203900857915, + "learning_rate": 1.0586362816979383e-06, + "loss": 0.0732, + "step": 5625 + }, + { + "epoch": 1.2800910125142206, + "grad_norm": 1.5673530833581029, + "learning_rate": 1.0585719308783316e-06, + "loss": 0.0803, + "step": 5626 + }, + { + "epoch": 1.2803185437997724, + "grad_norm": 2.7308896344498383, + "learning_rate": 1.058507571197308e-06, + "loss": 0.1711, + "step": 5627 + }, + { + "epoch": 1.2805460750853244, + "grad_norm": 0.9517054919299557, + "learning_rate": 1.058443202656184e-06, + "loss": 0.0311, + "step": 5628 + }, + { + "epoch": 1.2807736063708761, + "grad_norm": 1.891198334894284, + "learning_rate": 1.0583788252562745e-06, + "loss": 0.1027, + "step": 5629 + }, + { + "epoch": 1.2810011376564279, + "grad_norm": 2.3500902017479754, + "learning_rate": 1.058314438998896e-06, + "loss": 0.2345, + "step": 5630 + }, + { + "epoch": 1.2812286689419796, + "grad_norm": 1.5797405508748195, + "learning_rate": 1.0582500438853631e-06, + "loss": 0.1152, + "step": 5631 + }, + { + "epoch": 1.2814562002275314, + "grad_norm": 1.5413701487828582, + "learning_rate": 1.0581856399169934e-06, + "loss": 0.0678, + "step": 5632 + }, + { + "epoch": 1.2816837315130831, + "grad_norm": 1.078485797497606, + "learning_rate": 1.0581212270951024e-06, + "loss": 0.0311, + "step": 5633 + }, + { + "epoch": 1.2819112627986349, + "grad_norm": 2.2121037992989, + "learning_rate": 1.058056805421007e-06, + "loss": 0.1267, + "step": 5634 + }, + { + "epoch": 1.2821387940841866, + "grad_norm": 2.0746165641618486, + "learning_rate": 1.057992374896023e-06, + "loss": 0.0901, + "step": 5635 + }, + { + "epoch": 1.2823663253697384, + "grad_norm": 1.2046274771811845, + "learning_rate": 1.0579279355214683e-06, + "loss": 0.0671, + "step": 5636 + }, + { + "epoch": 1.2825938566552901, + "grad_norm": 2.0273686949332292, + "learning_rate": 1.0578634872986592e-06, + "loss": 0.0875, + "step": 5637 + }, + { + "epoch": 1.2828213879408419, + "grad_norm": 1.8866997216058947, + "learning_rate": 1.0577990302289136e-06, + "loss": 0.0966, + "step": 5638 + }, + { + "epoch": 1.2830489192263936, + "grad_norm": 1.5093028228498133, + "learning_rate": 1.0577345643135482e-06, + "loss": 0.1298, + "step": 5639 + }, + { + "epoch": 1.2832764505119454, + "grad_norm": 1.9771803724128194, + "learning_rate": 1.0576700895538809e-06, + "loss": 0.1155, + "step": 5640 + }, + { + "epoch": 1.2835039817974971, + "grad_norm": 1.7222932217295666, + "learning_rate": 1.0576056059512292e-06, + "loss": 0.0774, + "step": 5641 + }, + { + "epoch": 1.283731513083049, + "grad_norm": 1.3538690259349349, + "learning_rate": 1.057541113506911e-06, + "loss": 0.0756, + "step": 5642 + }, + { + "epoch": 1.2839590443686006, + "grad_norm": 1.8641486336878552, + "learning_rate": 1.057476612222245e-06, + "loss": 0.0695, + "step": 5643 + }, + { + "epoch": 1.2841865756541524, + "grad_norm": 1.2831640838979976, + "learning_rate": 1.057412102098549e-06, + "loss": 0.1428, + "step": 5644 + }, + { + "epoch": 1.2844141069397041, + "grad_norm": 1.3642696261409177, + "learning_rate": 1.0573475831371416e-06, + "loss": 0.0862, + "step": 5645 + }, + { + "epoch": 1.284641638225256, + "grad_norm": 1.9245965484657077, + "learning_rate": 1.0572830553393412e-06, + "loss": 0.0904, + "step": 5646 + }, + { + "epoch": 1.2848691695108077, + "grad_norm": 1.2171403811385195, + "learning_rate": 1.057218518706467e-06, + "loss": 0.0336, + "step": 5647 + }, + { + "epoch": 1.2850967007963594, + "grad_norm": 1.3095189728109227, + "learning_rate": 1.0571539732398378e-06, + "loss": 0.0418, + "step": 5648 + }, + { + "epoch": 1.2853242320819112, + "grad_norm": 1.918484794100106, + "learning_rate": 1.057089418940773e-06, + "loss": 0.0619, + "step": 5649 + }, + { + "epoch": 1.285551763367463, + "grad_norm": 1.1443578318237353, + "learning_rate": 1.0570248558105915e-06, + "loss": 0.0674, + "step": 5650 + }, + { + "epoch": 1.2857792946530147, + "grad_norm": 1.606788398963902, + "learning_rate": 1.0569602838506136e-06, + "loss": 0.0614, + "step": 5651 + }, + { + "epoch": 1.2860068259385666, + "grad_norm": 1.5390260189373348, + "learning_rate": 1.0568957030621582e-06, + "loss": 0.0453, + "step": 5652 + }, + { + "epoch": 1.2862343572241184, + "grad_norm": 1.4702477528553395, + "learning_rate": 1.0568311134465457e-06, + "loss": 0.1226, + "step": 5653 + }, + { + "epoch": 1.2864618885096701, + "grad_norm": 1.46344076229504, + "learning_rate": 1.056766515005096e-06, + "loss": 0.0922, + "step": 5654 + }, + { + "epoch": 1.286689419795222, + "grad_norm": 1.9431330097306456, + "learning_rate": 1.0567019077391296e-06, + "loss": 0.063, + "step": 5655 + }, + { + "epoch": 1.2869169510807736, + "grad_norm": 1.444515304458717, + "learning_rate": 1.056637291649967e-06, + "loss": 0.0745, + "step": 5656 + }, + { + "epoch": 1.2871444823663254, + "grad_norm": 2.0390282770012718, + "learning_rate": 1.0565726667389284e-06, + "loss": 0.1238, + "step": 5657 + }, + { + "epoch": 1.2873720136518771, + "grad_norm": 1.6617432624086061, + "learning_rate": 1.056508033007335e-06, + "loss": 0.0546, + "step": 5658 + }, + { + "epoch": 1.287599544937429, + "grad_norm": 2.318932450133174, + "learning_rate": 1.0564433904565078e-06, + "loss": 0.1079, + "step": 5659 + }, + { + "epoch": 1.2878270762229806, + "grad_norm": 1.650902114497156, + "learning_rate": 1.0563787390877677e-06, + "loss": 0.0884, + "step": 5660 + }, + { + "epoch": 1.2880546075085324, + "grad_norm": 2.147322191567513, + "learning_rate": 1.0563140789024363e-06, + "loss": 0.0977, + "step": 5661 + }, + { + "epoch": 1.2882821387940842, + "grad_norm": 1.4183546752619831, + "learning_rate": 1.0562494099018346e-06, + "loss": 0.0749, + "step": 5662 + }, + { + "epoch": 1.288509670079636, + "grad_norm": 1.8061494369968136, + "learning_rate": 1.0561847320872853e-06, + "loss": 0.093, + "step": 5663 + }, + { + "epoch": 1.2887372013651877, + "grad_norm": 2.081733636480595, + "learning_rate": 1.0561200454601097e-06, + "loss": 0.1204, + "step": 5664 + }, + { + "epoch": 1.2889647326507394, + "grad_norm": 1.2291622298614695, + "learning_rate": 1.0560553500216298e-06, + "loss": 0.0756, + "step": 5665 + }, + { + "epoch": 1.2891922639362912, + "grad_norm": 2.9727470638088787, + "learning_rate": 1.0559906457731678e-06, + "loss": 0.0822, + "step": 5666 + }, + { + "epoch": 1.2894197952218431, + "grad_norm": 2.481036835543893, + "learning_rate": 1.0559259327160465e-06, + "loss": 0.0894, + "step": 5667 + }, + { + "epoch": 1.2896473265073949, + "grad_norm": 0.96010237753671, + "learning_rate": 1.0558612108515883e-06, + "loss": 0.0232, + "step": 5668 + }, + { + "epoch": 1.2898748577929466, + "grad_norm": 1.2978970896725612, + "learning_rate": 1.0557964801811162e-06, + "loss": 0.0443, + "step": 5669 + }, + { + "epoch": 1.2901023890784984, + "grad_norm": 1.7666859988255192, + "learning_rate": 1.0557317407059529e-06, + "loss": 0.0789, + "step": 5670 + }, + { + "epoch": 1.2903299203640501, + "grad_norm": 1.9566224899598423, + "learning_rate": 1.0556669924274217e-06, + "loss": 0.0995, + "step": 5671 + }, + { + "epoch": 1.290557451649602, + "grad_norm": 2.3581648947367713, + "learning_rate": 1.0556022353468459e-06, + "loss": 0.0985, + "step": 5672 + }, + { + "epoch": 1.2907849829351536, + "grad_norm": 1.8768821502934563, + "learning_rate": 1.055537469465549e-06, + "loss": 0.1041, + "step": 5673 + }, + { + "epoch": 1.2910125142207054, + "grad_norm": 1.0645494201548409, + "learning_rate": 1.0554726947848545e-06, + "loss": 0.0558, + "step": 5674 + }, + { + "epoch": 1.2912400455062572, + "grad_norm": 1.56718896783486, + "learning_rate": 1.0554079113060869e-06, + "loss": 0.0923, + "step": 5675 + }, + { + "epoch": 1.291467576791809, + "grad_norm": 1.488761679619322, + "learning_rate": 1.0553431190305695e-06, + "loss": 0.082, + "step": 5676 + }, + { + "epoch": 1.2916951080773607, + "grad_norm": 1.5383291552422151, + "learning_rate": 1.055278317959627e-06, + "loss": 0.0508, + "step": 5677 + }, + { + "epoch": 1.2919226393629124, + "grad_norm": 1.3268434182473585, + "learning_rate": 1.0552135080945839e-06, + "loss": 0.0828, + "step": 5678 + }, + { + "epoch": 1.2921501706484642, + "grad_norm": 1.2228706582710298, + "learning_rate": 1.0551486894367643e-06, + "loss": 0.1055, + "step": 5679 + }, + { + "epoch": 1.292377701934016, + "grad_norm": 0.9920593257927844, + "learning_rate": 1.0550838619874933e-06, + "loss": 0.0755, + "step": 5680 + }, + { + "epoch": 1.2926052332195677, + "grad_norm": 2.1725969925810866, + "learning_rate": 1.055019025748096e-06, + "loss": 0.1201, + "step": 5681 + }, + { + "epoch": 1.2928327645051194, + "grad_norm": 1.660596292271756, + "learning_rate": 1.0549541807198974e-06, + "loss": 0.0754, + "step": 5682 + }, + { + "epoch": 1.2930602957906712, + "grad_norm": 2.723536633803254, + "learning_rate": 1.0548893269042226e-06, + "loss": 0.0805, + "step": 5683 + }, + { + "epoch": 1.293287827076223, + "grad_norm": 1.5224476485394507, + "learning_rate": 1.0548244643023972e-06, + "loss": 0.1975, + "step": 5684 + }, + { + "epoch": 1.2935153583617747, + "grad_norm": 1.7245323067588136, + "learning_rate": 1.054759592915747e-06, + "loss": 0.0894, + "step": 5685 + }, + { + "epoch": 1.2937428896473264, + "grad_norm": 2.351366087052394, + "learning_rate": 1.054694712745598e-06, + "loss": 0.0762, + "step": 5686 + }, + { + "epoch": 1.2939704209328782, + "grad_norm": 2.481113435758225, + "learning_rate": 1.0546298237932757e-06, + "loss": 0.1448, + "step": 5687 + }, + { + "epoch": 1.29419795221843, + "grad_norm": 1.7881104717575749, + "learning_rate": 1.0545649260601068e-06, + "loss": 0.1129, + "step": 5688 + }, + { + "epoch": 1.2944254835039817, + "grad_norm": 1.6533591614039695, + "learning_rate": 1.0545000195474175e-06, + "loss": 0.0931, + "step": 5689 + }, + { + "epoch": 1.2946530147895334, + "grad_norm": 2.977524456984651, + "learning_rate": 1.0544351042565344e-06, + "loss": 0.0972, + "step": 5690 + }, + { + "epoch": 1.2948805460750854, + "grad_norm": 2.6291144709864818, + "learning_rate": 1.0543701801887842e-06, + "loss": 0.1104, + "step": 5691 + }, + { + "epoch": 1.2951080773606372, + "grad_norm": 2.1587512273916354, + "learning_rate": 1.054305247345494e-06, + "loss": 0.0771, + "step": 5692 + }, + { + "epoch": 1.295335608646189, + "grad_norm": 2.151718306148393, + "learning_rate": 1.0542403057279907e-06, + "loss": 0.0545, + "step": 5693 + }, + { + "epoch": 1.2955631399317407, + "grad_norm": 1.2060203503251004, + "learning_rate": 1.0541753553376016e-06, + "loss": 0.1352, + "step": 5694 + }, + { + "epoch": 1.2957906712172924, + "grad_norm": 1.6546320164361477, + "learning_rate": 1.0541103961756543e-06, + "loss": 0.0775, + "step": 5695 + }, + { + "epoch": 1.2960182025028442, + "grad_norm": 1.5435451270919913, + "learning_rate": 1.0540454282434765e-06, + "loss": 0.1312, + "step": 5696 + }, + { + "epoch": 1.296245733788396, + "grad_norm": 3.3085895333931234, + "learning_rate": 1.0539804515423955e-06, + "loss": 0.2124, + "step": 5697 + }, + { + "epoch": 1.2964732650739477, + "grad_norm": 1.2152294810518844, + "learning_rate": 1.0539154660737401e-06, + "loss": 0.1194, + "step": 5698 + }, + { + "epoch": 1.2967007963594994, + "grad_norm": 1.885666654668208, + "learning_rate": 1.053850471838838e-06, + "loss": 0.0736, + "step": 5699 + }, + { + "epoch": 1.2969283276450512, + "grad_norm": 1.0678926873042414, + "learning_rate": 1.0537854688390175e-06, + "loss": 0.0816, + "step": 5700 + }, + { + "epoch": 1.297155858930603, + "grad_norm": 1.56587703849235, + "learning_rate": 1.0537204570756076e-06, + "loss": 0.062, + "step": 5701 + }, + { + "epoch": 1.2973833902161547, + "grad_norm": 1.806050287456521, + "learning_rate": 1.0536554365499367e-06, + "loss": 0.0823, + "step": 5702 + }, + { + "epoch": 1.2976109215017064, + "grad_norm": 3.053384552284598, + "learning_rate": 1.0535904072633334e-06, + "loss": 0.2154, + "step": 5703 + }, + { + "epoch": 1.2978384527872582, + "grad_norm": 2.837347986125031, + "learning_rate": 1.0535253692171273e-06, + "loss": 0.1194, + "step": 5704 + }, + { + "epoch": 1.29806598407281, + "grad_norm": 1.0783174429070619, + "learning_rate": 1.0534603224126474e-06, + "loss": 0.1224, + "step": 5705 + }, + { + "epoch": 1.298293515358362, + "grad_norm": 1.4126873741655113, + "learning_rate": 1.0533952668512231e-06, + "loss": 0.0402, + "step": 5706 + }, + { + "epoch": 1.2985210466439137, + "grad_norm": 1.2900165683757403, + "learning_rate": 1.0533302025341843e-06, + "loss": 0.0818, + "step": 5707 + }, + { + "epoch": 1.2987485779294654, + "grad_norm": 1.5354168050539385, + "learning_rate": 1.0532651294628607e-06, + "loss": 0.1133, + "step": 5708 + }, + { + "epoch": 1.2989761092150172, + "grad_norm": 2.3982642364346876, + "learning_rate": 1.053200047638582e-06, + "loss": 0.134, + "step": 5709 + }, + { + "epoch": 1.299203640500569, + "grad_norm": 2.049947095929528, + "learning_rate": 1.0531349570626787e-06, + "loss": 0.129, + "step": 5710 + }, + { + "epoch": 1.2994311717861207, + "grad_norm": 1.530794000421886, + "learning_rate": 1.0530698577364807e-06, + "loss": 0.0717, + "step": 5711 + }, + { + "epoch": 1.2996587030716724, + "grad_norm": 1.3695000986945034, + "learning_rate": 1.053004749661319e-06, + "loss": 0.0917, + "step": 5712 + }, + { + "epoch": 1.2998862343572242, + "grad_norm": 1.382838780567727, + "learning_rate": 1.0529396328385238e-06, + "loss": 0.0727, + "step": 5713 + }, + { + "epoch": 1.300113765642776, + "grad_norm": 2.002717365583164, + "learning_rate": 1.0528745072694266e-06, + "loss": 0.237, + "step": 5714 + }, + { + "epoch": 1.3003412969283277, + "grad_norm": 2.643373983806965, + "learning_rate": 1.052809372955358e-06, + "loss": 0.0887, + "step": 5715 + }, + { + "epoch": 1.3005688282138794, + "grad_norm": 2.232032108317541, + "learning_rate": 1.0527442298976492e-06, + "loss": 0.0973, + "step": 5716 + }, + { + "epoch": 1.3007963594994312, + "grad_norm": 1.776755863040562, + "learning_rate": 1.0526790780976318e-06, + "loss": 0.1761, + "step": 5717 + }, + { + "epoch": 1.301023890784983, + "grad_norm": 2.101698342878711, + "learning_rate": 1.052613917556637e-06, + "loss": 0.099, + "step": 5718 + }, + { + "epoch": 1.3012514220705347, + "grad_norm": 2.301450668536455, + "learning_rate": 1.0525487482759975e-06, + "loss": 0.1094, + "step": 5719 + }, + { + "epoch": 1.3014789533560864, + "grad_norm": 1.226004654267289, + "learning_rate": 1.052483570257044e-06, + "loss": 0.1364, + "step": 5720 + }, + { + "epoch": 1.3017064846416382, + "grad_norm": 2.0134123240196744, + "learning_rate": 1.0524183835011095e-06, + "loss": 0.0869, + "step": 5721 + }, + { + "epoch": 1.30193401592719, + "grad_norm": 1.534286561621179, + "learning_rate": 1.052353188009526e-06, + "loss": 0.0985, + "step": 5722 + }, + { + "epoch": 1.3021615472127417, + "grad_norm": 1.9316225583331454, + "learning_rate": 1.052287983783626e-06, + "loss": 0.101, + "step": 5723 + }, + { + "epoch": 1.3023890784982934, + "grad_norm": 2.2788899795164443, + "learning_rate": 1.052222770824742e-06, + "loss": 0.1353, + "step": 5724 + }, + { + "epoch": 1.3026166097838452, + "grad_norm": 1.8692663859425822, + "learning_rate": 1.0521575491342074e-06, + "loss": 0.0766, + "step": 5725 + }, + { + "epoch": 1.302844141069397, + "grad_norm": 1.8953396577295798, + "learning_rate": 1.0520923187133544e-06, + "loss": 0.1077, + "step": 5726 + }, + { + "epoch": 1.3030716723549487, + "grad_norm": 2.517469477681662, + "learning_rate": 1.0520270795635167e-06, + "loss": 0.1913, + "step": 5727 + }, + { + "epoch": 1.3032992036405004, + "grad_norm": 1.4886467817121716, + "learning_rate": 1.0519618316860274e-06, + "loss": 0.0373, + "step": 5728 + }, + { + "epoch": 1.3035267349260522, + "grad_norm": 2.0300452591888134, + "learning_rate": 1.0518965750822204e-06, + "loss": 0.1017, + "step": 5729 + }, + { + "epoch": 1.3037542662116042, + "grad_norm": 1.2890598526750778, + "learning_rate": 1.0518313097534292e-06, + "loss": 0.0604, + "step": 5730 + }, + { + "epoch": 1.303981797497156, + "grad_norm": 0.9524365472372076, + "learning_rate": 1.0517660357009877e-06, + "loss": 0.1005, + "step": 5731 + }, + { + "epoch": 1.3042093287827077, + "grad_norm": 1.2925513698786057, + "learning_rate": 1.0517007529262301e-06, + "loss": 0.0721, + "step": 5732 + }, + { + "epoch": 1.3044368600682594, + "grad_norm": 1.969070459589164, + "learning_rate": 1.0516354614304905e-06, + "loss": 0.1578, + "step": 5733 + }, + { + "epoch": 1.3046643913538112, + "grad_norm": 1.9412894909614369, + "learning_rate": 1.0515701612151035e-06, + "loss": 0.1085, + "step": 5734 + }, + { + "epoch": 1.304891922639363, + "grad_norm": 2.1131197697011195, + "learning_rate": 1.0515048522814034e-06, + "loss": 0.0863, + "step": 5735 + }, + { + "epoch": 1.3051194539249147, + "grad_norm": 1.2923421707711995, + "learning_rate": 1.0514395346307254e-06, + "loss": 0.073, + "step": 5736 + }, + { + "epoch": 1.3053469852104664, + "grad_norm": 1.8321089482558766, + "learning_rate": 1.0513742082644043e-06, + "loss": 0.1271, + "step": 5737 + }, + { + "epoch": 1.3055745164960182, + "grad_norm": 1.485078710256677, + "learning_rate": 1.0513088731837753e-06, + "loss": 0.0854, + "step": 5738 + }, + { + "epoch": 1.30580204778157, + "grad_norm": 1.3330940238964653, + "learning_rate": 1.0512435293901737e-06, + "loss": 0.079, + "step": 5739 + }, + { + "epoch": 1.3060295790671217, + "grad_norm": 1.536444213687335, + "learning_rate": 1.051178176884935e-06, + "loss": 0.0992, + "step": 5740 + }, + { + "epoch": 1.3062571103526734, + "grad_norm": 2.2452413636951554, + "learning_rate": 1.0511128156693947e-06, + "loss": 0.1075, + "step": 5741 + }, + { + "epoch": 1.3064846416382252, + "grad_norm": 1.4528356683489017, + "learning_rate": 1.0510474457448888e-06, + "loss": 0.0695, + "step": 5742 + }, + { + "epoch": 1.306712172923777, + "grad_norm": 1.7443487526246482, + "learning_rate": 1.0509820671127535e-06, + "loss": 0.08, + "step": 5743 + }, + { + "epoch": 1.3069397042093287, + "grad_norm": 1.5210003023160206, + "learning_rate": 1.050916679774325e-06, + "loss": 0.0447, + "step": 5744 + }, + { + "epoch": 1.3071672354948807, + "grad_norm": 1.5217878952043518, + "learning_rate": 1.0508512837309394e-06, + "loss": 0.0469, + "step": 5745 + }, + { + "epoch": 1.3073947667804324, + "grad_norm": 1.569608575831104, + "learning_rate": 1.0507858789839336e-06, + "loss": 0.0461, + "step": 5746 + }, + { + "epoch": 1.3076222980659842, + "grad_norm": 1.7294471484004552, + "learning_rate": 1.0507204655346442e-06, + "loss": 0.132, + "step": 5747 + }, + { + "epoch": 1.307849829351536, + "grad_norm": 1.6430994302512345, + "learning_rate": 1.050655043384408e-06, + "loss": 0.1076, + "step": 5748 + }, + { + "epoch": 1.3080773606370877, + "grad_norm": 1.2577545749106074, + "learning_rate": 1.0505896125345624e-06, + "loss": 0.0643, + "step": 5749 + }, + { + "epoch": 1.3083048919226394, + "grad_norm": 2.2746902540206464, + "learning_rate": 1.0505241729864446e-06, + "loss": 0.1662, + "step": 5750 + }, + { + "epoch": 1.3085324232081912, + "grad_norm": 1.7990261754234909, + "learning_rate": 1.050458724741392e-06, + "loss": 0.0451, + "step": 5751 + }, + { + "epoch": 1.308759954493743, + "grad_norm": 2.0834042000512145, + "learning_rate": 1.0503932678007423e-06, + "loss": 0.147, + "step": 5752 + }, + { + "epoch": 1.3089874857792947, + "grad_norm": 1.7939490285705315, + "learning_rate": 1.0503278021658331e-06, + "loss": 0.0655, + "step": 5753 + }, + { + "epoch": 1.3092150170648464, + "grad_norm": 1.5956435196651289, + "learning_rate": 1.0502623278380025e-06, + "loss": 0.0505, + "step": 5754 + }, + { + "epoch": 1.3094425483503982, + "grad_norm": 1.936354732487317, + "learning_rate": 1.050196844818589e-06, + "loss": 0.081, + "step": 5755 + }, + { + "epoch": 1.30967007963595, + "grad_norm": 1.912194809038067, + "learning_rate": 1.0501313531089306e-06, + "loss": 0.0677, + "step": 5756 + }, + { + "epoch": 1.3098976109215017, + "grad_norm": 2.168014313975063, + "learning_rate": 1.050065852710366e-06, + "loss": 0.1209, + "step": 5757 + }, + { + "epoch": 1.3101251422070535, + "grad_norm": 2.32550846320349, + "learning_rate": 1.0500003436242338e-06, + "loss": 0.1123, + "step": 5758 + }, + { + "epoch": 1.3103526734926052, + "grad_norm": 2.358664055020327, + "learning_rate": 1.049934825851873e-06, + "loss": 0.1254, + "step": 5759 + }, + { + "epoch": 1.310580204778157, + "grad_norm": 1.8343721616257782, + "learning_rate": 1.0498692993946225e-06, + "loss": 0.1464, + "step": 5760 + }, + { + "epoch": 1.3108077360637087, + "grad_norm": 1.9635602038856228, + "learning_rate": 1.0498037642538215e-06, + "loss": 0.0844, + "step": 5761 + }, + { + "epoch": 1.3110352673492605, + "grad_norm": 2.0496799670315857, + "learning_rate": 1.0497382204308099e-06, + "loss": 0.0733, + "step": 5762 + }, + { + "epoch": 1.3112627986348122, + "grad_norm": 1.210777013561021, + "learning_rate": 1.0496726679269265e-06, + "loss": 0.085, + "step": 5763 + }, + { + "epoch": 1.311490329920364, + "grad_norm": 1.7241240680481036, + "learning_rate": 1.0496071067435118e-06, + "loss": 0.0803, + "step": 5764 + }, + { + "epoch": 1.3117178612059157, + "grad_norm": 2.0878916046381977, + "learning_rate": 1.0495415368819057e-06, + "loss": 0.09, + "step": 5765 + }, + { + "epoch": 1.3119453924914675, + "grad_norm": 1.7632105499111685, + "learning_rate": 1.0494759583434478e-06, + "loss": 0.1059, + "step": 5766 + }, + { + "epoch": 1.3121729237770192, + "grad_norm": 1.5672705509207427, + "learning_rate": 1.0494103711294786e-06, + "loss": 0.0851, + "step": 5767 + }, + { + "epoch": 1.312400455062571, + "grad_norm": 1.4938195763871733, + "learning_rate": 1.049344775241339e-06, + "loss": 0.0979, + "step": 5768 + }, + { + "epoch": 1.312627986348123, + "grad_norm": 1.7176031546340367, + "learning_rate": 1.049279170680369e-06, + "loss": 0.0921, + "step": 5769 + }, + { + "epoch": 1.3128555176336747, + "grad_norm": 1.1557472694595412, + "learning_rate": 1.0492135574479097e-06, + "loss": 0.161, + "step": 5770 + }, + { + "epoch": 1.3130830489192264, + "grad_norm": 2.1783183998983797, + "learning_rate": 1.0491479355453026e-06, + "loss": 0.1091, + "step": 5771 + }, + { + "epoch": 1.3133105802047782, + "grad_norm": 1.8021717680419203, + "learning_rate": 1.0490823049738884e-06, + "loss": 0.0778, + "step": 5772 + }, + { + "epoch": 1.31353811149033, + "grad_norm": 1.3721710815692068, + "learning_rate": 1.0490166657350084e-06, + "loss": 0.0703, + "step": 5773 + }, + { + "epoch": 1.3137656427758817, + "grad_norm": 2.074316848775226, + "learning_rate": 1.0489510178300043e-06, + "loss": 0.0835, + "step": 5774 + }, + { + "epoch": 1.3139931740614335, + "grad_norm": 3.4881842631245656, + "learning_rate": 1.0488853612602178e-06, + "loss": 0.0772, + "step": 5775 + }, + { + "epoch": 1.3142207053469852, + "grad_norm": 1.8006810241587943, + "learning_rate": 1.0488196960269912e-06, + "loss": 0.114, + "step": 5776 + }, + { + "epoch": 1.314448236632537, + "grad_norm": 1.7856100748843726, + "learning_rate": 1.0487540221316659e-06, + "loss": 0.0555, + "step": 5777 + }, + { + "epoch": 1.3146757679180887, + "grad_norm": 1.067332294646937, + "learning_rate": 1.0486883395755845e-06, + "loss": 0.1084, + "step": 5778 + }, + { + "epoch": 1.3149032992036405, + "grad_norm": 1.70673215150491, + "learning_rate": 1.0486226483600894e-06, + "loss": 0.1335, + "step": 5779 + }, + { + "epoch": 1.3151308304891922, + "grad_norm": 3.4711332840947757, + "learning_rate": 1.0485569484865231e-06, + "loss": 0.1033, + "step": 5780 + }, + { + "epoch": 1.315358361774744, + "grad_norm": 2.0387320114180114, + "learning_rate": 1.0484912399562285e-06, + "loss": 0.076, + "step": 5781 + }, + { + "epoch": 1.3155858930602957, + "grad_norm": 1.8189138137721752, + "learning_rate": 1.0484255227705487e-06, + "loss": 0.0809, + "step": 5782 + }, + { + "epoch": 1.3158134243458475, + "grad_norm": 2.0535694365802715, + "learning_rate": 1.0483597969308266e-06, + "loss": 0.0621, + "step": 5783 + }, + { + "epoch": 1.3160409556313994, + "grad_norm": 1.946344026400018, + "learning_rate": 1.0482940624384054e-06, + "loss": 0.0685, + "step": 5784 + }, + { + "epoch": 1.3162684869169512, + "grad_norm": 1.1032176541225551, + "learning_rate": 1.048228319294629e-06, + "loss": 0.0996, + "step": 5785 + }, + { + "epoch": 1.316496018202503, + "grad_norm": 1.7898000777406582, + "learning_rate": 1.0481625675008409e-06, + "loss": 0.1119, + "step": 5786 + }, + { + "epoch": 1.3167235494880547, + "grad_norm": 2.206016130040065, + "learning_rate": 1.048096807058385e-06, + "loss": 0.1098, + "step": 5787 + }, + { + "epoch": 1.3169510807736065, + "grad_norm": 1.8133767218889647, + "learning_rate": 1.0480310379686048e-06, + "loss": 0.1035, + "step": 5788 + }, + { + "epoch": 1.3171786120591582, + "grad_norm": 1.4796571765990723, + "learning_rate": 1.0479652602328453e-06, + "loss": 0.0605, + "step": 5789 + }, + { + "epoch": 1.31740614334471, + "grad_norm": 1.1968819875341754, + "learning_rate": 1.0478994738524504e-06, + "loss": 0.0683, + "step": 5790 + }, + { + "epoch": 1.3176336746302617, + "grad_norm": 1.2404482616708883, + "learning_rate": 1.047833678828765e-06, + "loss": 0.0509, + "step": 5791 + }, + { + "epoch": 1.3178612059158135, + "grad_norm": 2.1533741321795046, + "learning_rate": 1.0477678751631332e-06, + "loss": 0.1336, + "step": 5792 + }, + { + "epoch": 1.3180887372013652, + "grad_norm": 1.5517989824699339, + "learning_rate": 1.0477020628569005e-06, + "loss": 0.0844, + "step": 5793 + }, + { + "epoch": 1.318316268486917, + "grad_norm": 1.4786711835589954, + "learning_rate": 1.0476362419114117e-06, + "loss": 0.1779, + "step": 5794 + }, + { + "epoch": 1.3185437997724687, + "grad_norm": 1.6768383320832863, + "learning_rate": 1.047570412328012e-06, + "loss": 0.0618, + "step": 5795 + }, + { + "epoch": 1.3187713310580205, + "grad_norm": 1.5663335122379476, + "learning_rate": 1.0475045741080473e-06, + "loss": 0.1298, + "step": 5796 + }, + { + "epoch": 1.3189988623435722, + "grad_norm": 2.4337391656673364, + "learning_rate": 1.0474387272528627e-06, + "loss": 0.1788, + "step": 5797 + }, + { + "epoch": 1.319226393629124, + "grad_norm": 2.66491053649857, + "learning_rate": 1.047372871763804e-06, + "loss": 0.0773, + "step": 5798 + }, + { + "epoch": 1.3194539249146757, + "grad_norm": 1.8928878472670687, + "learning_rate": 1.0473070076422176e-06, + "loss": 0.0908, + "step": 5799 + }, + { + "epoch": 1.3196814562002275, + "grad_norm": 2.276868624611035, + "learning_rate": 1.0472411348894492e-06, + "loss": 0.1277, + "step": 5800 + }, + { + "epoch": 1.3199089874857792, + "grad_norm": 1.5594575132733213, + "learning_rate": 1.0471752535068455e-06, + "loss": 0.1108, + "step": 5801 + }, + { + "epoch": 1.320136518771331, + "grad_norm": 1.3384265244459481, + "learning_rate": 1.0471093634957528e-06, + "loss": 0.0729, + "step": 5802 + }, + { + "epoch": 1.3203640500568827, + "grad_norm": 1.931742246823199, + "learning_rate": 1.0470434648575175e-06, + "loss": 0.0799, + "step": 5803 + }, + { + "epoch": 1.3205915813424345, + "grad_norm": 1.2351662386947617, + "learning_rate": 1.046977557593487e-06, + "loss": 0.0879, + "step": 5804 + }, + { + "epoch": 1.3208191126279862, + "grad_norm": 1.5595235915936758, + "learning_rate": 1.0469116417050078e-06, + "loss": 0.1352, + "step": 5805 + }, + { + "epoch": 1.321046643913538, + "grad_norm": 3.0284116645875083, + "learning_rate": 1.0468457171934276e-06, + "loss": 0.1001, + "step": 5806 + }, + { + "epoch": 1.3212741751990897, + "grad_norm": 1.1006401738487668, + "learning_rate": 1.0467797840600934e-06, + "loss": 0.1256, + "step": 5807 + }, + { + "epoch": 1.3215017064846417, + "grad_norm": 2.048683668619229, + "learning_rate": 1.0467138423063529e-06, + "loss": 0.1206, + "step": 5808 + }, + { + "epoch": 1.3217292377701935, + "grad_norm": 2.721602472652372, + "learning_rate": 1.0466478919335538e-06, + "loss": 0.1185, + "step": 5809 + }, + { + "epoch": 1.3219567690557452, + "grad_norm": 1.4264521873334444, + "learning_rate": 1.0465819329430439e-06, + "loss": 0.1311, + "step": 5810 + }, + { + "epoch": 1.322184300341297, + "grad_norm": 1.6730255927051096, + "learning_rate": 1.0465159653361716e-06, + "loss": 0.0815, + "step": 5811 + }, + { + "epoch": 1.3224118316268487, + "grad_norm": 1.2282421184268042, + "learning_rate": 1.0464499891142847e-06, + "loss": 0.0617, + "step": 5812 + }, + { + "epoch": 1.3226393629124005, + "grad_norm": 1.54847311265808, + "learning_rate": 1.046384004278732e-06, + "loss": 0.0799, + "step": 5813 + }, + { + "epoch": 1.3228668941979522, + "grad_norm": 1.621833713328949, + "learning_rate": 1.0463180108308615e-06, + "loss": 0.0712, + "step": 5814 + }, + { + "epoch": 1.323094425483504, + "grad_norm": 1.9271344527036025, + "learning_rate": 1.0462520087720231e-06, + "loss": 0.0931, + "step": 5815 + }, + { + "epoch": 1.3233219567690557, + "grad_norm": 1.2022404430865519, + "learning_rate": 1.0461859981035649e-06, + "loss": 0.054, + "step": 5816 + }, + { + "epoch": 1.3235494880546075, + "grad_norm": 1.3792412126497369, + "learning_rate": 1.0461199788268364e-06, + "loss": 0.0906, + "step": 5817 + }, + { + "epoch": 1.3237770193401592, + "grad_norm": 1.3530300025052608, + "learning_rate": 1.0460539509431865e-06, + "loss": 0.0482, + "step": 5818 + }, + { + "epoch": 1.324004550625711, + "grad_norm": 1.3311479250314509, + "learning_rate": 1.045987914453965e-06, + "loss": 0.0517, + "step": 5819 + }, + { + "epoch": 1.3242320819112627, + "grad_norm": 2.197848133474945, + "learning_rate": 1.0459218693605216e-06, + "loss": 0.1283, + "step": 5820 + }, + { + "epoch": 1.3244596131968145, + "grad_norm": 1.7362479038886882, + "learning_rate": 1.0458558156642063e-06, + "loss": 0.088, + "step": 5821 + }, + { + "epoch": 1.3246871444823665, + "grad_norm": 1.8362008981229307, + "learning_rate": 1.0457897533663686e-06, + "loss": 0.117, + "step": 5822 + }, + { + "epoch": 1.3249146757679182, + "grad_norm": 1.4992468656529518, + "learning_rate": 1.0457236824683592e-06, + "loss": 0.0961, + "step": 5823 + }, + { + "epoch": 1.32514220705347, + "grad_norm": 2.0624079614706985, + "learning_rate": 1.045657602971528e-06, + "loss": 0.1186, + "step": 5824 + }, + { + "epoch": 1.3253697383390217, + "grad_norm": 1.387226086927801, + "learning_rate": 1.0455915148772262e-06, + "loss": 0.0644, + "step": 5825 + }, + { + "epoch": 1.3255972696245735, + "grad_norm": 1.451657775323959, + "learning_rate": 1.0455254181868037e-06, + "loss": 0.084, + "step": 5826 + }, + { + "epoch": 1.3258248009101252, + "grad_norm": 3.203594309996943, + "learning_rate": 1.0454593129016121e-06, + "loss": 0.1307, + "step": 5827 + }, + { + "epoch": 1.326052332195677, + "grad_norm": 1.2359629926153486, + "learning_rate": 1.045393199023002e-06, + "loss": 0.071, + "step": 5828 + }, + { + "epoch": 1.3262798634812287, + "grad_norm": 2.09626696688845, + "learning_rate": 1.0453270765523247e-06, + "loss": 0.1149, + "step": 5829 + }, + { + "epoch": 1.3265073947667805, + "grad_norm": 1.1510394557592603, + "learning_rate": 1.045260945490932e-06, + "loss": 0.074, + "step": 5830 + }, + { + "epoch": 1.3267349260523322, + "grad_norm": 2.8998180438743857, + "learning_rate": 1.045194805840175e-06, + "loss": 0.1232, + "step": 5831 + }, + { + "epoch": 1.326962457337884, + "grad_norm": 2.0371831357255625, + "learning_rate": 1.045128657601406e-06, + "loss": 0.148, + "step": 5832 + }, + { + "epoch": 1.3271899886234357, + "grad_norm": 1.6823112018670245, + "learning_rate": 1.0450625007759765e-06, + "loss": 0.0794, + "step": 5833 + }, + { + "epoch": 1.3274175199089875, + "grad_norm": 1.378507230734654, + "learning_rate": 1.044996335365239e-06, + "loss": 0.0473, + "step": 5834 + }, + { + "epoch": 1.3276450511945392, + "grad_norm": 2.501713620892651, + "learning_rate": 1.0449301613705453e-06, + "loss": 0.1352, + "step": 5835 + }, + { + "epoch": 1.327872582480091, + "grad_norm": 1.310982383924543, + "learning_rate": 1.0448639787932482e-06, + "loss": 0.0629, + "step": 5836 + }, + { + "epoch": 1.3281001137656427, + "grad_norm": 1.4961669986775104, + "learning_rate": 1.0447977876347005e-06, + "loss": 0.0505, + "step": 5837 + }, + { + "epoch": 1.3283276450511945, + "grad_norm": 2.0946957506049677, + "learning_rate": 1.0447315878962547e-06, + "loss": 0.0564, + "step": 5838 + }, + { + "epoch": 1.3285551763367462, + "grad_norm": 1.3448895548736886, + "learning_rate": 1.044665379579264e-06, + "loss": 0.072, + "step": 5839 + }, + { + "epoch": 1.328782707622298, + "grad_norm": 1.6449819772089789, + "learning_rate": 1.0445991626850816e-06, + "loss": 0.1275, + "step": 5840 + }, + { + "epoch": 1.3290102389078498, + "grad_norm": 1.3217135874023973, + "learning_rate": 1.0445329372150607e-06, + "loss": 0.1001, + "step": 5841 + }, + { + "epoch": 1.3292377701934015, + "grad_norm": 1.3083326572244676, + "learning_rate": 1.0444667031705549e-06, + "loss": 0.0633, + "step": 5842 + }, + { + "epoch": 1.3294653014789533, + "grad_norm": 1.6851774826419728, + "learning_rate": 1.0444004605529178e-06, + "loss": 0.1603, + "step": 5843 + }, + { + "epoch": 1.329692832764505, + "grad_norm": 1.4502489521313568, + "learning_rate": 1.0443342093635036e-06, + "loss": 0.0826, + "step": 5844 + }, + { + "epoch": 1.3299203640500568, + "grad_norm": 1.9821203054145495, + "learning_rate": 1.044267949603666e-06, + "loss": 0.1617, + "step": 5845 + }, + { + "epoch": 1.3301478953356085, + "grad_norm": 1.8051039972539313, + "learning_rate": 1.0442016812747594e-06, + "loss": 0.0716, + "step": 5846 + }, + { + "epoch": 1.3303754266211605, + "grad_norm": 2.129526007883718, + "learning_rate": 1.0441354043781381e-06, + "loss": 0.0649, + "step": 5847 + }, + { + "epoch": 1.3306029579067122, + "grad_norm": 1.4233191957308409, + "learning_rate": 1.0440691189151567e-06, + "loss": 0.0607, + "step": 5848 + }, + { + "epoch": 1.330830489192264, + "grad_norm": 1.6736105046051897, + "learning_rate": 1.0440028248871702e-06, + "loss": 0.0712, + "step": 5849 + }, + { + "epoch": 1.3310580204778157, + "grad_norm": 1.4967468018831813, + "learning_rate": 1.0439365222955332e-06, + "loss": 0.0454, + "step": 5850 + }, + { + "epoch": 1.3312855517633675, + "grad_norm": 0.9536612302057513, + "learning_rate": 1.043870211141601e-06, + "loss": 0.0591, + "step": 5851 + }, + { + "epoch": 1.3315130830489192, + "grad_norm": 2.1762186577819103, + "learning_rate": 1.0438038914267287e-06, + "loss": 0.141, + "step": 5852 + }, + { + "epoch": 1.331740614334471, + "grad_norm": 1.4591366909505499, + "learning_rate": 1.043737563152272e-06, + "loss": 0.0627, + "step": 5853 + }, + { + "epoch": 1.3319681456200227, + "grad_norm": 4.350121279354156, + "learning_rate": 1.0436712263195862e-06, + "loss": 0.1319, + "step": 5854 + }, + { + "epoch": 1.3321956769055745, + "grad_norm": 1.7016647523273083, + "learning_rate": 1.0436048809300273e-06, + "loss": 0.1901, + "step": 5855 + }, + { + "epoch": 1.3324232081911263, + "grad_norm": 1.1948459627791288, + "learning_rate": 1.0435385269849515e-06, + "loss": 0.0885, + "step": 5856 + }, + { + "epoch": 1.332650739476678, + "grad_norm": 2.231584916495476, + "learning_rate": 1.0434721644857146e-06, + "loss": 0.0842, + "step": 5857 + }, + { + "epoch": 1.3328782707622298, + "grad_norm": 1.3919515812269696, + "learning_rate": 1.043405793433673e-06, + "loss": 0.0733, + "step": 5858 + }, + { + "epoch": 1.3331058020477815, + "grad_norm": 0.9859935017445751, + "learning_rate": 1.0433394138301835e-06, + "loss": 0.0439, + "step": 5859 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 2.1136452812594873, + "learning_rate": 1.0432730256766022e-06, + "loss": 0.1696, + "step": 5860 + }, + { + "epoch": 1.3335608646188852, + "grad_norm": 1.7286776962693724, + "learning_rate": 1.0432066289742864e-06, + "loss": 0.1255, + "step": 5861 + }, + { + "epoch": 1.333788395904437, + "grad_norm": 1.8643295717261203, + "learning_rate": 1.0431402237245932e-06, + "loss": 0.077, + "step": 5862 + }, + { + "epoch": 1.3340159271899887, + "grad_norm": 1.0528362516615255, + "learning_rate": 1.0430738099288794e-06, + "loss": 0.0466, + "step": 5863 + }, + { + "epoch": 1.3342434584755405, + "grad_norm": 1.51341037677569, + "learning_rate": 1.0430073875885026e-06, + "loss": 0.0691, + "step": 5864 + }, + { + "epoch": 1.3344709897610922, + "grad_norm": 1.699780878797244, + "learning_rate": 1.0429409567048205e-06, + "loss": 0.0593, + "step": 5865 + }, + { + "epoch": 1.334698521046644, + "grad_norm": 0.9013097690218285, + "learning_rate": 1.0428745172791905e-06, + "loss": 0.1108, + "step": 5866 + }, + { + "epoch": 1.3349260523321957, + "grad_norm": 2.5759017069295944, + "learning_rate": 1.0428080693129708e-06, + "loss": 0.1563, + "step": 5867 + }, + { + "epoch": 1.3351535836177475, + "grad_norm": 1.3870463099298531, + "learning_rate": 1.0427416128075192e-06, + "loss": 0.0459, + "step": 5868 + }, + { + "epoch": 1.3353811149032992, + "grad_norm": 1.220961536090874, + "learning_rate": 1.0426751477641941e-06, + "loss": 0.1204, + "step": 5869 + }, + { + "epoch": 1.335608646188851, + "grad_norm": 1.2909291825710791, + "learning_rate": 1.042608674184354e-06, + "loss": 0.0746, + "step": 5870 + }, + { + "epoch": 1.3358361774744028, + "grad_norm": 2.606205725169108, + "learning_rate": 1.0425421920693575e-06, + "loss": 0.179, + "step": 5871 + }, + { + "epoch": 1.3360637087599545, + "grad_norm": 0.8052339430093359, + "learning_rate": 1.042475701420563e-06, + "loss": 0.0446, + "step": 5872 + }, + { + "epoch": 1.3362912400455063, + "grad_norm": 1.5627987415782247, + "learning_rate": 1.04240920223933e-06, + "loss": 0.0587, + "step": 5873 + }, + { + "epoch": 1.336518771331058, + "grad_norm": 1.5997474951882902, + "learning_rate": 1.0423426945270174e-06, + "loss": 0.0876, + "step": 5874 + }, + { + "epoch": 1.3367463026166098, + "grad_norm": 1.1449459131825908, + "learning_rate": 1.0422761782849842e-06, + "loss": 0.0739, + "step": 5875 + }, + { + "epoch": 1.3369738339021615, + "grad_norm": 1.851697392588764, + "learning_rate": 1.0422096535145902e-06, + "loss": 0.1552, + "step": 5876 + }, + { + "epoch": 1.3372013651877133, + "grad_norm": 2.3632246540349593, + "learning_rate": 1.042143120217195e-06, + "loss": 0.1272, + "step": 5877 + }, + { + "epoch": 1.337428896473265, + "grad_norm": 1.6754416909482417, + "learning_rate": 1.0420765783941586e-06, + "loss": 0.0454, + "step": 5878 + }, + { + "epoch": 1.3376564277588168, + "grad_norm": 2.32343084336057, + "learning_rate": 1.0420100280468404e-06, + "loss": 0.1528, + "step": 5879 + }, + { + "epoch": 1.3378839590443685, + "grad_norm": 1.9816599470429406, + "learning_rate": 1.0419434691766012e-06, + "loss": 0.1561, + "step": 5880 + }, + { + "epoch": 1.3381114903299203, + "grad_norm": 1.5497562642603393, + "learning_rate": 1.041876901784801e-06, + "loss": 0.0758, + "step": 5881 + }, + { + "epoch": 1.338339021615472, + "grad_norm": 1.9049260514033595, + "learning_rate": 1.0418103258728001e-06, + "loss": 0.0979, + "step": 5882 + }, + { + "epoch": 1.3385665529010238, + "grad_norm": 1.3282399491488064, + "learning_rate": 1.04174374144196e-06, + "loss": 0.0717, + "step": 5883 + }, + { + "epoch": 1.3387940841865755, + "grad_norm": 1.4546413694820024, + "learning_rate": 1.0416771484936409e-06, + "loss": 0.0364, + "step": 5884 + }, + { + "epoch": 1.3390216154721273, + "grad_norm": 1.380966775798661, + "learning_rate": 1.041610547029204e-06, + "loss": 0.1469, + "step": 5885 + }, + { + "epoch": 1.3392491467576793, + "grad_norm": 1.4838765642178313, + "learning_rate": 1.0415439370500104e-06, + "loss": 0.095, + "step": 5886 + }, + { + "epoch": 1.339476678043231, + "grad_norm": 0.9893478877017928, + "learning_rate": 1.0414773185574214e-06, + "loss": 0.0988, + "step": 5887 + }, + { + "epoch": 1.3397042093287828, + "grad_norm": 2.552307381411876, + "learning_rate": 1.0414106915527991e-06, + "loss": 0.0769, + "step": 5888 + }, + { + "epoch": 1.3399317406143345, + "grad_norm": 1.5341685999322272, + "learning_rate": 1.041344056037505e-06, + "loss": 0.163, + "step": 5889 + }, + { + "epoch": 1.3401592718998863, + "grad_norm": 2.050981202480822, + "learning_rate": 1.0412774120129004e-06, + "loss": 0.1326, + "step": 5890 + }, + { + "epoch": 1.340386803185438, + "grad_norm": 1.0449449099860342, + "learning_rate": 1.0412107594803484e-06, + "loss": 0.1196, + "step": 5891 + }, + { + "epoch": 1.3406143344709898, + "grad_norm": 1.9352276977320042, + "learning_rate": 1.0411440984412103e-06, + "loss": 0.1095, + "step": 5892 + }, + { + "epoch": 1.3408418657565415, + "grad_norm": 2.5332880240233955, + "learning_rate": 1.0410774288968492e-06, + "loss": 0.1012, + "step": 5893 + }, + { + "epoch": 1.3410693970420933, + "grad_norm": 1.6964443954344153, + "learning_rate": 1.0410107508486272e-06, + "loss": 0.0618, + "step": 5894 + }, + { + "epoch": 1.341296928327645, + "grad_norm": 1.6754493818964415, + "learning_rate": 1.0409440642979077e-06, + "loss": 0.0646, + "step": 5895 + }, + { + "epoch": 1.3415244596131968, + "grad_norm": 1.2216423929052722, + "learning_rate": 1.040877369246053e-06, + "loss": 0.1081, + "step": 5896 + }, + { + "epoch": 1.3417519908987485, + "grad_norm": 1.5997557392502906, + "learning_rate": 1.0408106656944267e-06, + "loss": 0.0664, + "step": 5897 + }, + { + "epoch": 1.3419795221843003, + "grad_norm": 1.8788278907156983, + "learning_rate": 1.0407439536443919e-06, + "loss": 0.1048, + "step": 5898 + }, + { + "epoch": 1.342207053469852, + "grad_norm": 1.7938675169169382, + "learning_rate": 1.040677233097312e-06, + "loss": 0.0697, + "step": 5899 + }, + { + "epoch": 1.342434584755404, + "grad_norm": 0.9355624853300721, + "learning_rate": 1.0406105040545509e-06, + "loss": 0.0284, + "step": 5900 + }, + { + "epoch": 1.3426621160409558, + "grad_norm": 1.5042553303858235, + "learning_rate": 1.0405437665174719e-06, + "loss": 0.1187, + "step": 5901 + }, + { + "epoch": 1.3428896473265075, + "grad_norm": 1.4026846157680415, + "learning_rate": 1.0404770204874396e-06, + "loss": 0.0765, + "step": 5902 + }, + { + "epoch": 1.3431171786120593, + "grad_norm": 1.8439416331349125, + "learning_rate": 1.040410265965818e-06, + "loss": 0.0685, + "step": 5903 + }, + { + "epoch": 1.343344709897611, + "grad_norm": 1.315124752696636, + "learning_rate": 1.0403435029539711e-06, + "loss": 0.076, + "step": 5904 + }, + { + "epoch": 1.3435722411831628, + "grad_norm": 2.8348421300082554, + "learning_rate": 1.0402767314532638e-06, + "loss": 0.1332, + "step": 5905 + }, + { + "epoch": 1.3437997724687145, + "grad_norm": 1.2357365501581734, + "learning_rate": 1.0402099514650607e-06, + "loss": 0.0785, + "step": 5906 + }, + { + "epoch": 1.3440273037542663, + "grad_norm": 1.062333481303275, + "learning_rate": 1.0401431629907267e-06, + "loss": 0.0453, + "step": 5907 + }, + { + "epoch": 1.344254835039818, + "grad_norm": 1.9191232064798869, + "learning_rate": 1.0400763660316265e-06, + "loss": 0.0642, + "step": 5908 + }, + { + "epoch": 1.3444823663253698, + "grad_norm": 1.8190539631092906, + "learning_rate": 1.0400095605891258e-06, + "loss": 0.1386, + "step": 5909 + }, + { + "epoch": 1.3447098976109215, + "grad_norm": 2.321732624283394, + "learning_rate": 1.0399427466645895e-06, + "loss": 0.0609, + "step": 5910 + }, + { + "epoch": 1.3449374288964733, + "grad_norm": 2.707788483644483, + "learning_rate": 1.0398759242593834e-06, + "loss": 0.1499, + "step": 5911 + }, + { + "epoch": 1.345164960182025, + "grad_norm": 1.664041570190974, + "learning_rate": 1.0398090933748733e-06, + "loss": 0.1073, + "step": 5912 + }, + { + "epoch": 1.3453924914675768, + "grad_norm": 2.4540612257121603, + "learning_rate": 1.039742254012425e-06, + "loss": 0.099, + "step": 5913 + }, + { + "epoch": 1.3456200227531285, + "grad_norm": 1.219808928274102, + "learning_rate": 1.0396754061734047e-06, + "loss": 0.1104, + "step": 5914 + }, + { + "epoch": 1.3458475540386803, + "grad_norm": 2.7416891076755023, + "learning_rate": 1.0396085498591785e-06, + "loss": 0.1198, + "step": 5915 + }, + { + "epoch": 1.346075085324232, + "grad_norm": 1.578816398295786, + "learning_rate": 1.0395416850711127e-06, + "loss": 0.1303, + "step": 5916 + }, + { + "epoch": 1.3463026166097838, + "grad_norm": 1.6764518514726592, + "learning_rate": 1.0394748118105743e-06, + "loss": 0.0917, + "step": 5917 + }, + { + "epoch": 1.3465301478953355, + "grad_norm": 2.803240136856215, + "learning_rate": 1.0394079300789296e-06, + "loss": 0.0922, + "step": 5918 + }, + { + "epoch": 1.3467576791808873, + "grad_norm": 2.438798620238522, + "learning_rate": 1.0393410398775459e-06, + "loss": 0.148, + "step": 5919 + }, + { + "epoch": 1.346985210466439, + "grad_norm": 2.025396852926161, + "learning_rate": 1.03927414120779e-06, + "loss": 0.0781, + "step": 5920 + }, + { + "epoch": 1.3472127417519908, + "grad_norm": 1.7022905945417979, + "learning_rate": 1.0392072340710296e-06, + "loss": 0.1126, + "step": 5921 + }, + { + "epoch": 1.3474402730375425, + "grad_norm": 1.820096766632113, + "learning_rate": 1.0391403184686318e-06, + "loss": 0.113, + "step": 5922 + }, + { + "epoch": 1.3476678043230943, + "grad_norm": 1.445494347851026, + "learning_rate": 1.0390733944019645e-06, + "loss": 0.0577, + "step": 5923 + }, + { + "epoch": 1.347895335608646, + "grad_norm": 1.6969833902189329, + "learning_rate": 1.0390064618723952e-06, + "loss": 0.0563, + "step": 5924 + }, + { + "epoch": 1.348122866894198, + "grad_norm": 2.5942714651292698, + "learning_rate": 1.038939520881292e-06, + "loss": 0.0867, + "step": 5925 + }, + { + "epoch": 1.3483503981797498, + "grad_norm": 1.5855460417612113, + "learning_rate": 1.038872571430023e-06, + "loss": 0.1554, + "step": 5926 + }, + { + "epoch": 1.3485779294653015, + "grad_norm": 1.6885697182726835, + "learning_rate": 1.038805613519957e-06, + "loss": 0.1151, + "step": 5927 + }, + { + "epoch": 1.3488054607508533, + "grad_norm": 1.1799583949182342, + "learning_rate": 1.038738647152462e-06, + "loss": 0.0321, + "step": 5928 + }, + { + "epoch": 1.349032992036405, + "grad_norm": 1.5007571959530954, + "learning_rate": 1.0386716723289063e-06, + "loss": 0.0903, + "step": 5929 + }, + { + "epoch": 1.3492605233219568, + "grad_norm": 1.0886205302970793, + "learning_rate": 1.0386046890506596e-06, + "loss": 0.0503, + "step": 5930 + }, + { + "epoch": 1.3494880546075085, + "grad_norm": 2.5513820854170914, + "learning_rate": 1.0385376973190906e-06, + "loss": 0.0841, + "step": 5931 + }, + { + "epoch": 1.3497155858930603, + "grad_norm": 1.6875766939167716, + "learning_rate": 1.0384706971355683e-06, + "loss": 0.1215, + "step": 5932 + }, + { + "epoch": 1.349943117178612, + "grad_norm": 5.082685707073952, + "learning_rate": 1.038403688501462e-06, + "loss": 0.1221, + "step": 5933 + }, + { + "epoch": 1.3501706484641638, + "grad_norm": 1.5607491563614442, + "learning_rate": 1.0383366714181419e-06, + "loss": 0.0603, + "step": 5934 + }, + { + "epoch": 1.3503981797497155, + "grad_norm": 1.6201503065498213, + "learning_rate": 1.038269645886977e-06, + "loss": 0.0782, + "step": 5935 + }, + { + "epoch": 1.3506257110352673, + "grad_norm": 2.0996414301663333, + "learning_rate": 1.0382026119093372e-06, + "loss": 0.1101, + "step": 5936 + }, + { + "epoch": 1.350853242320819, + "grad_norm": 2.1247499282829443, + "learning_rate": 1.038135569486593e-06, + "loss": 0.0509, + "step": 5937 + }, + { + "epoch": 1.3510807736063708, + "grad_norm": 3.3212216068998774, + "learning_rate": 1.0380685186201143e-06, + "loss": 0.1659, + "step": 5938 + }, + { + "epoch": 1.3513083048919228, + "grad_norm": 1.7561604700268703, + "learning_rate": 1.0380014593112714e-06, + "loss": 0.1308, + "step": 5939 + }, + { + "epoch": 1.3515358361774745, + "grad_norm": 1.1220194869116855, + "learning_rate": 1.0379343915614354e-06, + "loss": 0.065, + "step": 5940 + }, + { + "epoch": 1.3517633674630263, + "grad_norm": 1.83273858306615, + "learning_rate": 1.0378673153719764e-06, + "loss": 0.1116, + "step": 5941 + }, + { + "epoch": 1.351990898748578, + "grad_norm": 2.0716912267355942, + "learning_rate": 1.0378002307442659e-06, + "loss": 0.0669, + "step": 5942 + }, + { + "epoch": 1.3522184300341298, + "grad_norm": 0.8977133455560301, + "learning_rate": 1.0377331376796745e-06, + "loss": 0.0999, + "step": 5943 + }, + { + "epoch": 1.3524459613196815, + "grad_norm": 1.2511184360608965, + "learning_rate": 1.0376660361795738e-06, + "loss": 0.1479, + "step": 5944 + }, + { + "epoch": 1.3526734926052333, + "grad_norm": 1.3086756806996873, + "learning_rate": 1.0375989262453348e-06, + "loss": 0.1577, + "step": 5945 + }, + { + "epoch": 1.352901023890785, + "grad_norm": 1.4033184337593902, + "learning_rate": 1.0375318078783294e-06, + "loss": 0.1543, + "step": 5946 + }, + { + "epoch": 1.3531285551763368, + "grad_norm": 1.3177544568938675, + "learning_rate": 1.0374646810799297e-06, + "loss": 0.1321, + "step": 5947 + }, + { + "epoch": 1.3533560864618885, + "grad_norm": 1.6253035269611373, + "learning_rate": 1.037397545851507e-06, + "loss": 0.1215, + "step": 5948 + }, + { + "epoch": 1.3535836177474403, + "grad_norm": 2.0350270154931365, + "learning_rate": 1.0373304021944338e-06, + "loss": 0.0774, + "step": 5949 + }, + { + "epoch": 1.353811149032992, + "grad_norm": 1.5562479173929296, + "learning_rate": 1.0372632501100826e-06, + "loss": 0.0434, + "step": 5950 + }, + { + "epoch": 1.3540386803185438, + "grad_norm": 1.9639000027131648, + "learning_rate": 1.0371960895998252e-06, + "loss": 0.1173, + "step": 5951 + }, + { + "epoch": 1.3542662116040955, + "grad_norm": 2.83162310264471, + "learning_rate": 1.0371289206650349e-06, + "loss": 0.1132, + "step": 5952 + }, + { + "epoch": 1.3544937428896473, + "grad_norm": 1.6006614306610998, + "learning_rate": 1.0370617433070842e-06, + "loss": 0.0796, + "step": 5953 + }, + { + "epoch": 1.354721274175199, + "grad_norm": 1.2529908090115869, + "learning_rate": 1.036994557527346e-06, + "loss": 0.0691, + "step": 5954 + }, + { + "epoch": 1.3549488054607508, + "grad_norm": 3.8878650458910258, + "learning_rate": 1.0369273633271936e-06, + "loss": 0.1068, + "step": 5955 + }, + { + "epoch": 1.3551763367463026, + "grad_norm": 1.504373255205242, + "learning_rate": 1.0368601607080004e-06, + "loss": 0.0542, + "step": 5956 + }, + { + "epoch": 1.3554038680318543, + "grad_norm": 2.1975103651204004, + "learning_rate": 1.0367929496711397e-06, + "loss": 0.1272, + "step": 5957 + }, + { + "epoch": 1.355631399317406, + "grad_norm": 1.4461832329122462, + "learning_rate": 1.0367257302179853e-06, + "loss": 0.0621, + "step": 5958 + }, + { + "epoch": 1.3558589306029578, + "grad_norm": 1.8573245201747381, + "learning_rate": 1.036658502349911e-06, + "loss": 0.101, + "step": 5959 + }, + { + "epoch": 1.3560864618885096, + "grad_norm": 1.567187644910704, + "learning_rate": 1.0365912660682908e-06, + "loss": 0.0984, + "step": 5960 + }, + { + "epoch": 1.3563139931740613, + "grad_norm": 2.162745505227273, + "learning_rate": 1.036524021374499e-06, + "loss": 0.0802, + "step": 5961 + }, + { + "epoch": 1.356541524459613, + "grad_norm": 2.3326169965757884, + "learning_rate": 1.0364567682699098e-06, + "loss": 0.1185, + "step": 5962 + }, + { + "epoch": 1.356769055745165, + "grad_norm": 1.5145485778181955, + "learning_rate": 1.036389506755898e-06, + "loss": 0.0715, + "step": 5963 + }, + { + "epoch": 1.3569965870307168, + "grad_norm": 1.3551941984066445, + "learning_rate": 1.036322236833838e-06, + "loss": 0.0768, + "step": 5964 + }, + { + "epoch": 1.3572241183162685, + "grad_norm": 1.1211006521431286, + "learning_rate": 1.0362549585051046e-06, + "loss": 0.0894, + "step": 5965 + }, + { + "epoch": 1.3574516496018203, + "grad_norm": 1.9996402931539106, + "learning_rate": 1.0361876717710731e-06, + "loss": 0.1432, + "step": 5966 + }, + { + "epoch": 1.357679180887372, + "grad_norm": 1.522756728382017, + "learning_rate": 1.0361203766331187e-06, + "loss": 0.0979, + "step": 5967 + }, + { + "epoch": 1.3579067121729238, + "grad_norm": 1.1331201405305724, + "learning_rate": 1.036053073092617e-06, + "loss": 0.0397, + "step": 5968 + }, + { + "epoch": 1.3581342434584756, + "grad_norm": 2.497612271094558, + "learning_rate": 1.0359857611509428e-06, + "loss": 0.1579, + "step": 5969 + }, + { + "epoch": 1.3583617747440273, + "grad_norm": 1.9306593556926885, + "learning_rate": 1.0359184408094726e-06, + "loss": 0.089, + "step": 5970 + }, + { + "epoch": 1.358589306029579, + "grad_norm": 1.6257772134491746, + "learning_rate": 1.0358511120695819e-06, + "loss": 0.0781, + "step": 5971 + }, + { + "epoch": 1.3588168373151308, + "grad_norm": 2.0008785362216304, + "learning_rate": 1.0357837749326471e-06, + "loss": 0.0939, + "step": 5972 + }, + { + "epoch": 1.3590443686006826, + "grad_norm": 1.510087734989364, + "learning_rate": 1.0357164294000442e-06, + "loss": 0.0556, + "step": 5973 + }, + { + "epoch": 1.3592718998862343, + "grad_norm": 1.329932710456839, + "learning_rate": 1.0356490754731496e-06, + "loss": 0.1131, + "step": 5974 + }, + { + "epoch": 1.359499431171786, + "grad_norm": 2.409170005936539, + "learning_rate": 1.03558171315334e-06, + "loss": 0.0818, + "step": 5975 + }, + { + "epoch": 1.3597269624573378, + "grad_norm": 1.2096360812807305, + "learning_rate": 1.0355143424419922e-06, + "loss": 0.0421, + "step": 5976 + }, + { + "epoch": 1.3599544937428896, + "grad_norm": 1.6720121080102, + "learning_rate": 1.035446963340483e-06, + "loss": 0.1711, + "step": 5977 + }, + { + "epoch": 1.3601820250284415, + "grad_norm": 1.4457125424363366, + "learning_rate": 1.0353795758501894e-06, + "loss": 0.1127, + "step": 5978 + }, + { + "epoch": 1.3604095563139933, + "grad_norm": 2.109294137925049, + "learning_rate": 1.0353121799724892e-06, + "loss": 0.1251, + "step": 5979 + }, + { + "epoch": 1.360637087599545, + "grad_norm": 1.463533989091872, + "learning_rate": 1.0352447757087592e-06, + "loss": 0.0521, + "step": 5980 + }, + { + "epoch": 1.3608646188850968, + "grad_norm": 2.139182529319199, + "learning_rate": 1.0351773630603774e-06, + "loss": 0.0889, + "step": 5981 + }, + { + "epoch": 1.3610921501706486, + "grad_norm": 1.6032438033790135, + "learning_rate": 1.0351099420287213e-06, + "loss": 0.0731, + "step": 5982 + }, + { + "epoch": 1.3613196814562003, + "grad_norm": 1.5891744053001757, + "learning_rate": 1.0350425126151694e-06, + "loss": 0.1427, + "step": 5983 + }, + { + "epoch": 1.361547212741752, + "grad_norm": 1.7389302044250665, + "learning_rate": 1.0349750748210994e-06, + "loss": 0.0719, + "step": 5984 + }, + { + "epoch": 1.3617747440273038, + "grad_norm": 1.716288886034002, + "learning_rate": 1.0349076286478897e-06, + "loss": 0.0841, + "step": 5985 + }, + { + "epoch": 1.3620022753128556, + "grad_norm": 1.5231208210255824, + "learning_rate": 1.0348401740969188e-06, + "loss": 0.0888, + "step": 5986 + }, + { + "epoch": 1.3622298065984073, + "grad_norm": 1.9618341869885971, + "learning_rate": 1.0347727111695652e-06, + "loss": 0.0703, + "step": 5987 + }, + { + "epoch": 1.362457337883959, + "grad_norm": 2.119675238920378, + "learning_rate": 1.0347052398672079e-06, + "loss": 0.1164, + "step": 5988 + }, + { + "epoch": 1.3626848691695108, + "grad_norm": 1.4923771093419005, + "learning_rate": 1.034637760191226e-06, + "loss": 0.1033, + "step": 5989 + }, + { + "epoch": 1.3629124004550626, + "grad_norm": 1.4764363477830744, + "learning_rate": 1.0345702721429982e-06, + "loss": 0.0973, + "step": 5990 + }, + { + "epoch": 1.3631399317406143, + "grad_norm": 1.070134987056995, + "learning_rate": 1.0345027757239044e-06, + "loss": 0.053, + "step": 5991 + }, + { + "epoch": 1.363367463026166, + "grad_norm": 1.1915968190738013, + "learning_rate": 1.0344352709353237e-06, + "loss": 0.081, + "step": 5992 + }, + { + "epoch": 1.3635949943117178, + "grad_norm": 1.0769956353353518, + "learning_rate": 1.034367757778636e-06, + "loss": 0.0696, + "step": 5993 + }, + { + "epoch": 1.3638225255972696, + "grad_norm": 1.928908351422348, + "learning_rate": 1.034300236255221e-06, + "loss": 0.0729, + "step": 5994 + }, + { + "epoch": 1.3640500568828213, + "grad_norm": 2.8515506900628194, + "learning_rate": 1.0342327063664587e-06, + "loss": 0.0955, + "step": 5995 + }, + { + "epoch": 1.364277588168373, + "grad_norm": 1.8960944981081354, + "learning_rate": 1.0341651681137293e-06, + "loss": 0.0725, + "step": 5996 + }, + { + "epoch": 1.3645051194539248, + "grad_norm": 1.4003406696741232, + "learning_rate": 1.0340976214984136e-06, + "loss": 0.0408, + "step": 5997 + }, + { + "epoch": 1.3647326507394766, + "grad_norm": 1.6929433228108035, + "learning_rate": 1.0340300665218913e-06, + "loss": 0.0823, + "step": 5998 + }, + { + "epoch": 1.3649601820250283, + "grad_norm": 1.3416521952685851, + "learning_rate": 1.0339625031855438e-06, + "loss": 0.0755, + "step": 5999 + }, + { + "epoch": 1.36518771331058, + "grad_norm": 1.3736008241682673, + "learning_rate": 1.0338949314907515e-06, + "loss": 0.1204, + "step": 6000 + }, + { + "epoch": 1.3654152445961318, + "grad_norm": 1.1265827930392238, + "learning_rate": 1.0338273514388958e-06, + "loss": 0.0705, + "step": 6001 + }, + { + "epoch": 1.3656427758816838, + "grad_norm": 2.0430120786187835, + "learning_rate": 1.0337597630313578e-06, + "loss": 0.1022, + "step": 6002 + }, + { + "epoch": 1.3658703071672356, + "grad_norm": 1.556682501798468, + "learning_rate": 1.0336921662695188e-06, + "loss": 0.1093, + "step": 6003 + }, + { + "epoch": 1.3660978384527873, + "grad_norm": 1.7795318504483018, + "learning_rate": 1.0336245611547605e-06, + "loss": 0.0873, + "step": 6004 + }, + { + "epoch": 1.366325369738339, + "grad_norm": 0.7929012709354717, + "learning_rate": 1.0335569476884643e-06, + "loss": 0.0532, + "step": 6005 + }, + { + "epoch": 1.3665529010238908, + "grad_norm": 1.6984974466407434, + "learning_rate": 1.0334893258720124e-06, + "loss": 0.122, + "step": 6006 + }, + { + "epoch": 1.3667804323094426, + "grad_norm": 1.7720537171200441, + "learning_rate": 1.033421695706787e-06, + "loss": 0.0573, + "step": 6007 + }, + { + "epoch": 1.3670079635949943, + "grad_norm": 1.915680831871843, + "learning_rate": 1.03335405719417e-06, + "loss": 0.0656, + "step": 6008 + }, + { + "epoch": 1.367235494880546, + "grad_norm": 1.7862258212947102, + "learning_rate": 1.0332864103355438e-06, + "loss": 0.0953, + "step": 6009 + }, + { + "epoch": 1.3674630261660978, + "grad_norm": 1.6810109121162469, + "learning_rate": 1.0332187551322914e-06, + "loss": 0.0922, + "step": 6010 + }, + { + "epoch": 1.3676905574516496, + "grad_norm": 1.7563798308384426, + "learning_rate": 1.0331510915857951e-06, + "loss": 0.0441, + "step": 6011 + }, + { + "epoch": 1.3679180887372013, + "grad_norm": 1.605413277057477, + "learning_rate": 1.0330834196974378e-06, + "loss": 0.0719, + "step": 6012 + }, + { + "epoch": 1.368145620022753, + "grad_norm": 1.2335738036590032, + "learning_rate": 1.0330157394686031e-06, + "loss": 0.1298, + "step": 6013 + }, + { + "epoch": 1.3683731513083048, + "grad_norm": 2.883850918158107, + "learning_rate": 1.0329480509006738e-06, + "loss": 0.1038, + "step": 6014 + }, + { + "epoch": 1.3686006825938566, + "grad_norm": 2.689802752529953, + "learning_rate": 1.0328803539950332e-06, + "loss": 0.0876, + "step": 6015 + }, + { + "epoch": 1.3688282138794083, + "grad_norm": 4.58606537572728, + "learning_rate": 1.0328126487530657e-06, + "loss": 0.0673, + "step": 6016 + }, + { + "epoch": 1.3690557451649603, + "grad_norm": 1.6526437618261232, + "learning_rate": 1.0327449351761542e-06, + "loss": 0.0868, + "step": 6017 + }, + { + "epoch": 1.369283276450512, + "grad_norm": 1.6946562162827543, + "learning_rate": 1.0326772132656828e-06, + "loss": 0.0715, + "step": 6018 + }, + { + "epoch": 1.3695108077360638, + "grad_norm": 3.007213425143118, + "learning_rate": 1.0326094830230362e-06, + "loss": 0.1021, + "step": 6019 + }, + { + "epoch": 1.3697383390216156, + "grad_norm": 2.611585168793522, + "learning_rate": 1.032541744449598e-06, + "loss": 0.2578, + "step": 6020 + }, + { + "epoch": 1.3699658703071673, + "grad_norm": 1.4761408630263304, + "learning_rate": 1.0324739975467529e-06, + "loss": 0.1518, + "step": 6021 + }, + { + "epoch": 1.370193401592719, + "grad_norm": 1.3284428317452652, + "learning_rate": 1.0324062423158857e-06, + "loss": 0.0949, + "step": 6022 + }, + { + "epoch": 1.3704209328782708, + "grad_norm": 1.2642919336112104, + "learning_rate": 1.0323384787583809e-06, + "loss": 0.087, + "step": 6023 + }, + { + "epoch": 1.3706484641638226, + "grad_norm": 2.504112434393499, + "learning_rate": 1.0322707068756238e-06, + "loss": 0.0953, + "step": 6024 + }, + { + "epoch": 1.3708759954493743, + "grad_norm": 1.1108553042003824, + "learning_rate": 1.0322029266689992e-06, + "loss": 0.112, + "step": 6025 + }, + { + "epoch": 1.371103526734926, + "grad_norm": 1.9589141312824296, + "learning_rate": 1.0321351381398926e-06, + "loss": 0.1449, + "step": 6026 + }, + { + "epoch": 1.3713310580204778, + "grad_norm": 1.5284346282963852, + "learning_rate": 1.0320673412896891e-06, + "loss": 0.0704, + "step": 6027 + }, + { + "epoch": 1.3715585893060296, + "grad_norm": 1.6360808959018123, + "learning_rate": 1.0319995361197752e-06, + "loss": 0.1075, + "step": 6028 + }, + { + "epoch": 1.3717861205915813, + "grad_norm": 1.0651159743295313, + "learning_rate": 1.0319317226315358e-06, + "loss": 0.1076, + "step": 6029 + }, + { + "epoch": 1.372013651877133, + "grad_norm": 1.4465076881759498, + "learning_rate": 1.0318639008263572e-06, + "loss": 0.0904, + "step": 6030 + }, + { + "epoch": 1.3722411831626848, + "grad_norm": 1.5828337267735475, + "learning_rate": 1.0317960707056256e-06, + "loss": 0.106, + "step": 6031 + }, + { + "epoch": 1.3724687144482366, + "grad_norm": 1.3535896802540421, + "learning_rate": 1.0317282322707275e-06, + "loss": 0.0623, + "step": 6032 + }, + { + "epoch": 1.3726962457337883, + "grad_norm": 0.6064410036946437, + "learning_rate": 1.0316603855230492e-06, + "loss": 0.0251, + "step": 6033 + }, + { + "epoch": 1.37292377701934, + "grad_norm": 1.9326954205683844, + "learning_rate": 1.0315925304639773e-06, + "loss": 0.0851, + "step": 6034 + }, + { + "epoch": 1.3731513083048918, + "grad_norm": 1.6547627103908698, + "learning_rate": 1.0315246670948988e-06, + "loss": 0.066, + "step": 6035 + }, + { + "epoch": 1.3733788395904436, + "grad_norm": 1.521791506208634, + "learning_rate": 1.0314567954172006e-06, + "loss": 0.1252, + "step": 6036 + }, + { + "epoch": 1.3736063708759954, + "grad_norm": 1.5173057147246578, + "learning_rate": 1.03138891543227e-06, + "loss": 0.1001, + "step": 6037 + }, + { + "epoch": 1.373833902161547, + "grad_norm": 1.6935760456434947, + "learning_rate": 1.031321027141494e-06, + "loss": 0.0764, + "step": 6038 + }, + { + "epoch": 1.3740614334470989, + "grad_norm": 2.1148518007647645, + "learning_rate": 1.0312531305462607e-06, + "loss": 0.2322, + "step": 6039 + }, + { + "epoch": 1.3742889647326506, + "grad_norm": 1.5431113869692263, + "learning_rate": 1.031185225647957e-06, + "loss": 0.0815, + "step": 6040 + }, + { + "epoch": 1.3745164960182026, + "grad_norm": 1.7571930584978857, + "learning_rate": 1.0311173124479715e-06, + "loss": 0.0729, + "step": 6041 + }, + { + "epoch": 1.3747440273037543, + "grad_norm": 1.6709560157545251, + "learning_rate": 1.0310493909476916e-06, + "loss": 0.0888, + "step": 6042 + }, + { + "epoch": 1.374971558589306, + "grad_norm": 0.9656044017250746, + "learning_rate": 1.0309814611485062e-06, + "loss": 0.0697, + "step": 6043 + }, + { + "epoch": 1.3751990898748578, + "grad_norm": 1.7934133654712006, + "learning_rate": 1.0309135230518028e-06, + "loss": 0.0923, + "step": 6044 + }, + { + "epoch": 1.3754266211604096, + "grad_norm": 1.7460975815711577, + "learning_rate": 1.0308455766589706e-06, + "loss": 0.0788, + "step": 6045 + }, + { + "epoch": 1.3756541524459613, + "grad_norm": 2.3162330853750124, + "learning_rate": 1.030777621971398e-06, + "loss": 0.066, + "step": 6046 + }, + { + "epoch": 1.375881683731513, + "grad_norm": 2.2361579800366034, + "learning_rate": 1.0307096589904742e-06, + "loss": 0.0959, + "step": 6047 + }, + { + "epoch": 1.3761092150170648, + "grad_norm": 2.2482212896632454, + "learning_rate": 1.030641687717588e-06, + "loss": 0.0972, + "step": 6048 + }, + { + "epoch": 1.3763367463026166, + "grad_norm": 3.063555572971071, + "learning_rate": 1.0305737081541283e-06, + "loss": 0.0879, + "step": 6049 + }, + { + "epoch": 1.3765642775881684, + "grad_norm": 1.1688880694396802, + "learning_rate": 1.0305057203014848e-06, + "loss": 0.0986, + "step": 6050 + }, + { + "epoch": 1.37679180887372, + "grad_norm": 1.220418623684574, + "learning_rate": 1.0304377241610472e-06, + "loss": 0.1351, + "step": 6051 + }, + { + "epoch": 1.3770193401592719, + "grad_norm": 1.6807893089278623, + "learning_rate": 1.030369719734205e-06, + "loss": 0.1322, + "step": 6052 + }, + { + "epoch": 1.3772468714448236, + "grad_norm": 1.9365626811701335, + "learning_rate": 1.0303017070223482e-06, + "loss": 0.1129, + "step": 6053 + }, + { + "epoch": 1.3774744027303754, + "grad_norm": 0.8533626037721147, + "learning_rate": 1.0302336860268667e-06, + "loss": 0.0496, + "step": 6054 + }, + { + "epoch": 1.377701934015927, + "grad_norm": 1.5261483411582581, + "learning_rate": 1.0301656567491507e-06, + "loss": 0.121, + "step": 6055 + }, + { + "epoch": 1.377929465301479, + "grad_norm": 2.511979075018236, + "learning_rate": 1.0300976191905907e-06, + "loss": 0.0824, + "step": 6056 + }, + { + "epoch": 1.3781569965870308, + "grad_norm": 1.38134265855808, + "learning_rate": 1.0300295733525774e-06, + "loss": 0.1248, + "step": 6057 + }, + { + "epoch": 1.3783845278725826, + "grad_norm": 1.5095526065881761, + "learning_rate": 1.0299615192365015e-06, + "loss": 0.0656, + "step": 6058 + }, + { + "epoch": 1.3786120591581343, + "grad_norm": 1.244350640420765, + "learning_rate": 1.0298934568437535e-06, + "loss": 0.0809, + "step": 6059 + }, + { + "epoch": 1.378839590443686, + "grad_norm": 2.00551764047578, + "learning_rate": 1.029825386175725e-06, + "loss": 0.1414, + "step": 6060 + }, + { + "epoch": 1.3790671217292378, + "grad_norm": 0.9661969179329034, + "learning_rate": 1.029757307233807e-06, + "loss": 0.0969, + "step": 6061 + }, + { + "epoch": 1.3792946530147896, + "grad_norm": 2.2025332944612495, + "learning_rate": 1.0296892200193908e-06, + "loss": 0.169, + "step": 6062 + }, + { + "epoch": 1.3795221843003413, + "grad_norm": 1.7254179219751133, + "learning_rate": 1.029621124533868e-06, + "loss": 0.0718, + "step": 6063 + }, + { + "epoch": 1.379749715585893, + "grad_norm": 1.7478260517346846, + "learning_rate": 1.0295530207786307e-06, + "loss": 0.0525, + "step": 6064 + }, + { + "epoch": 1.3799772468714449, + "grad_norm": 2.139388267720325, + "learning_rate": 1.0294849087550703e-06, + "loss": 0.1528, + "step": 6065 + }, + { + "epoch": 1.3802047781569966, + "grad_norm": 2.303671631702034, + "learning_rate": 1.0294167884645795e-06, + "loss": 0.1523, + "step": 6066 + }, + { + "epoch": 1.3804323094425484, + "grad_norm": 3.0251883633272167, + "learning_rate": 1.02934865990855e-06, + "loss": 0.1039, + "step": 6067 + }, + { + "epoch": 1.3806598407281, + "grad_norm": 1.845259206551197, + "learning_rate": 1.0292805230883743e-06, + "loss": 0.0584, + "step": 6068 + }, + { + "epoch": 1.3808873720136519, + "grad_norm": 2.085602227170942, + "learning_rate": 1.0292123780054452e-06, + "loss": 0.0759, + "step": 6069 + }, + { + "epoch": 1.3811149032992036, + "grad_norm": 1.5485285983889232, + "learning_rate": 1.0291442246611555e-06, + "loss": 0.1108, + "step": 6070 + }, + { + "epoch": 1.3813424345847554, + "grad_norm": 1.9756390644683244, + "learning_rate": 1.029076063056898e-06, + "loss": 0.1089, + "step": 6071 + }, + { + "epoch": 1.3815699658703071, + "grad_norm": 1.210855828233612, + "learning_rate": 1.0290078931940656e-06, + "loss": 0.0958, + "step": 6072 + }, + { + "epoch": 1.3817974971558589, + "grad_norm": 0.8802448365353132, + "learning_rate": 1.028939715074052e-06, + "loss": 0.0334, + "step": 6073 + }, + { + "epoch": 1.3820250284414106, + "grad_norm": 1.3463530588489843, + "learning_rate": 1.0288715286982504e-06, + "loss": 0.0677, + "step": 6074 + }, + { + "epoch": 1.3822525597269624, + "grad_norm": 0.7421571324570394, + "learning_rate": 1.0288033340680543e-06, + "loss": 0.026, + "step": 6075 + }, + { + "epoch": 1.3824800910125141, + "grad_norm": 1.0235714825965652, + "learning_rate": 1.0287351311848574e-06, + "loss": 0.0874, + "step": 6076 + }, + { + "epoch": 1.3827076222980659, + "grad_norm": 2.778457322941933, + "learning_rate": 1.028666920050054e-06, + "loss": 0.0906, + "step": 6077 + }, + { + "epoch": 1.3829351535836176, + "grad_norm": 1.0504053551900818, + "learning_rate": 1.0285987006650381e-06, + "loss": 0.0382, + "step": 6078 + }, + { + "epoch": 1.3831626848691694, + "grad_norm": 1.5439276989638069, + "learning_rate": 1.028530473031204e-06, + "loss": 0.1201, + "step": 6079 + }, + { + "epoch": 1.3833902161547214, + "grad_norm": 2.2472197379141927, + "learning_rate": 1.0284622371499457e-06, + "loss": 0.0875, + "step": 6080 + }, + { + "epoch": 1.383617747440273, + "grad_norm": 1.3277316421883985, + "learning_rate": 1.0283939930226584e-06, + "loss": 0.1003, + "step": 6081 + }, + { + "epoch": 1.3838452787258249, + "grad_norm": 1.8061757451329605, + "learning_rate": 1.0283257406507366e-06, + "loss": 0.0628, + "step": 6082 + }, + { + "epoch": 1.3840728100113766, + "grad_norm": 1.2434351612293517, + "learning_rate": 1.0282574800355755e-06, + "loss": 0.0762, + "step": 6083 + }, + { + "epoch": 1.3843003412969284, + "grad_norm": 2.1571207801279955, + "learning_rate": 1.0281892111785699e-06, + "loss": 0.1653, + "step": 6084 + }, + { + "epoch": 1.3845278725824801, + "grad_norm": 0.7925456249329077, + "learning_rate": 1.0281209340811151e-06, + "loss": 0.0503, + "step": 6085 + }, + { + "epoch": 1.3847554038680319, + "grad_norm": 1.6782816202693709, + "learning_rate": 1.0280526487446069e-06, + "loss": 0.0482, + "step": 6086 + }, + { + "epoch": 1.3849829351535836, + "grad_norm": 2.1389563232253104, + "learning_rate": 1.0279843551704409e-06, + "loss": 0.0586, + "step": 6087 + }, + { + "epoch": 1.3852104664391354, + "grad_norm": 1.5865791625267798, + "learning_rate": 1.0279160533600121e-06, + "loss": 0.1161, + "step": 6088 + }, + { + "epoch": 1.3854379977246871, + "grad_norm": 1.4237771848423213, + "learning_rate": 1.0278477433147176e-06, + "loss": 0.1359, + "step": 6089 + }, + { + "epoch": 1.3856655290102389, + "grad_norm": 1.5778953854419226, + "learning_rate": 1.0277794250359529e-06, + "loss": 0.0863, + "step": 6090 + }, + { + "epoch": 1.3858930602957906, + "grad_norm": 2.5592814368264274, + "learning_rate": 1.0277110985251142e-06, + "loss": 0.1515, + "step": 6091 + }, + { + "epoch": 1.3861205915813424, + "grad_norm": 1.5625510631280042, + "learning_rate": 1.0276427637835984e-06, + "loss": 0.1457, + "step": 6092 + }, + { + "epoch": 1.3863481228668941, + "grad_norm": 1.4218575035452607, + "learning_rate": 1.0275744208128019e-06, + "loss": 0.0434, + "step": 6093 + }, + { + "epoch": 1.3865756541524459, + "grad_norm": 1.3700974293267705, + "learning_rate": 1.0275060696141215e-06, + "loss": 0.0574, + "step": 6094 + }, + { + "epoch": 1.3868031854379979, + "grad_norm": 2.074735787061168, + "learning_rate": 1.027437710188954e-06, + "loss": 0.078, + "step": 6095 + }, + { + "epoch": 1.3870307167235496, + "grad_norm": 2.2617424180316648, + "learning_rate": 1.027369342538697e-06, + "loss": 0.16, + "step": 6096 + }, + { + "epoch": 1.3872582480091014, + "grad_norm": 1.6506933315691088, + "learning_rate": 1.0273009666647472e-06, + "loss": 0.1009, + "step": 6097 + }, + { + "epoch": 1.387485779294653, + "grad_norm": 2.093753106006641, + "learning_rate": 1.0272325825685028e-06, + "loss": 0.0834, + "step": 6098 + }, + { + "epoch": 1.3877133105802049, + "grad_norm": 1.4578100498874582, + "learning_rate": 1.027164190251361e-06, + "loss": 0.1295, + "step": 6099 + }, + { + "epoch": 1.3879408418657566, + "grad_norm": 0.8839198324440494, + "learning_rate": 1.0270957897147196e-06, + "loss": 0.0752, + "step": 6100 + }, + { + "epoch": 1.3881683731513084, + "grad_norm": 1.8268726950343663, + "learning_rate": 1.0270273809599764e-06, + "loss": 0.0755, + "step": 6101 + }, + { + "epoch": 1.3883959044368601, + "grad_norm": 1.7221146859777103, + "learning_rate": 1.0269589639885302e-06, + "loss": 0.0634, + "step": 6102 + }, + { + "epoch": 1.3886234357224119, + "grad_norm": 2.2959019107378036, + "learning_rate": 1.0268905388017788e-06, + "loss": 0.1049, + "step": 6103 + }, + { + "epoch": 1.3888509670079636, + "grad_norm": 0.9734172611334406, + "learning_rate": 1.0268221054011208e-06, + "loss": 0.0407, + "step": 6104 + }, + { + "epoch": 1.3890784982935154, + "grad_norm": 1.6445095549398978, + "learning_rate": 1.026753663787955e-06, + "loss": 0.0779, + "step": 6105 + }, + { + "epoch": 1.3893060295790671, + "grad_norm": 1.7481612724663955, + "learning_rate": 1.0266852139636799e-06, + "loss": 0.0655, + "step": 6106 + }, + { + "epoch": 1.3895335608646189, + "grad_norm": 2.387058614741269, + "learning_rate": 1.0266167559296946e-06, + "loss": 0.1849, + "step": 6107 + }, + { + "epoch": 1.3897610921501706, + "grad_norm": 1.749430126521713, + "learning_rate": 1.0265482896873986e-06, + "loss": 0.0571, + "step": 6108 + }, + { + "epoch": 1.3899886234357224, + "grad_norm": 2.154234273156198, + "learning_rate": 1.0264798152381907e-06, + "loss": 0.0949, + "step": 6109 + }, + { + "epoch": 1.3902161547212741, + "grad_norm": 2.0717344001521942, + "learning_rate": 1.026411332583471e-06, + "loss": 0.0803, + "step": 6110 + }, + { + "epoch": 1.3904436860068259, + "grad_norm": 1.9152543686027883, + "learning_rate": 1.0263428417246385e-06, + "loss": 0.109, + "step": 6111 + }, + { + "epoch": 1.3906712172923776, + "grad_norm": 1.4284369217746564, + "learning_rate": 1.0262743426630935e-06, + "loss": 0.0646, + "step": 6112 + }, + { + "epoch": 1.3908987485779294, + "grad_norm": 1.5184843974473854, + "learning_rate": 1.0262058354002357e-06, + "loss": 0.1204, + "step": 6113 + }, + { + "epoch": 1.3911262798634811, + "grad_norm": 1.4056897180706915, + "learning_rate": 1.0261373199374655e-06, + "loss": 0.0607, + "step": 6114 + }, + { + "epoch": 1.391353811149033, + "grad_norm": 2.348101018413372, + "learning_rate": 1.026068796276183e-06, + "loss": 0.066, + "step": 6115 + }, + { + "epoch": 1.3915813424345846, + "grad_norm": 1.9137612245987456, + "learning_rate": 1.0260002644177892e-06, + "loss": 0.0845, + "step": 6116 + }, + { + "epoch": 1.3918088737201364, + "grad_norm": 2.3856609440966827, + "learning_rate": 1.025931724363684e-06, + "loss": 0.1256, + "step": 6117 + }, + { + "epoch": 1.3920364050056881, + "grad_norm": 1.3816079520128584, + "learning_rate": 1.0258631761152687e-06, + "loss": 0.094, + "step": 6118 + }, + { + "epoch": 1.3922639362912401, + "grad_norm": 1.433005286962213, + "learning_rate": 1.0257946196739444e-06, + "loss": 0.1512, + "step": 6119 + }, + { + "epoch": 1.3924914675767919, + "grad_norm": 1.2323580910038312, + "learning_rate": 1.025726055041112e-06, + "loss": 0.0493, + "step": 6120 + }, + { + "epoch": 1.3927189988623436, + "grad_norm": 1.7407247244496606, + "learning_rate": 1.0256574822181727e-06, + "loss": 0.1078, + "step": 6121 + }, + { + "epoch": 1.3929465301478954, + "grad_norm": 1.5994369802036565, + "learning_rate": 1.0255889012065285e-06, + "loss": 0.0655, + "step": 6122 + }, + { + "epoch": 1.3931740614334471, + "grad_norm": 1.7930383624127975, + "learning_rate": 1.025520312007581e-06, + "loss": 0.0875, + "step": 6123 + }, + { + "epoch": 1.3934015927189989, + "grad_norm": 2.1120758522904453, + "learning_rate": 1.0254517146227314e-06, + "loss": 0.1572, + "step": 6124 + }, + { + "epoch": 1.3936291240045506, + "grad_norm": 2.5678979874544856, + "learning_rate": 1.0253831090533823e-06, + "loss": 0.1329, + "step": 6125 + }, + { + "epoch": 1.3938566552901024, + "grad_norm": 2.1764102691534175, + "learning_rate": 1.0253144953009357e-06, + "loss": 0.1409, + "step": 6126 + }, + { + "epoch": 1.3940841865756541, + "grad_norm": 1.3223191843329276, + "learning_rate": 1.025245873366794e-06, + "loss": 0.0666, + "step": 6127 + }, + { + "epoch": 1.394311717861206, + "grad_norm": 1.9410808222093632, + "learning_rate": 1.0251772432523596e-06, + "loss": 0.2062, + "step": 6128 + }, + { + "epoch": 1.3945392491467576, + "grad_norm": 1.8157179540741168, + "learning_rate": 1.0251086049590355e-06, + "loss": 0.1529, + "step": 6129 + }, + { + "epoch": 1.3947667804323094, + "grad_norm": 1.7518458655719975, + "learning_rate": 1.0250399584882239e-06, + "loss": 0.077, + "step": 6130 + }, + { + "epoch": 1.3949943117178611, + "grad_norm": 2.1802078459114016, + "learning_rate": 1.0249713038413285e-06, + "loss": 0.0698, + "step": 6131 + }, + { + "epoch": 1.395221843003413, + "grad_norm": 1.4485958979135534, + "learning_rate": 1.024902641019752e-06, + "loss": 0.0684, + "step": 6132 + }, + { + "epoch": 1.3954493742889647, + "grad_norm": 1.9433569593797686, + "learning_rate": 1.024833970024898e-06, + "loss": 0.101, + "step": 6133 + }, + { + "epoch": 1.3956769055745166, + "grad_norm": 2.3198698152365664, + "learning_rate": 1.0247652908581697e-06, + "loss": 0.1356, + "step": 6134 + }, + { + "epoch": 1.3959044368600684, + "grad_norm": 1.9234892150048948, + "learning_rate": 1.0246966035209712e-06, + "loss": 0.1432, + "step": 6135 + }, + { + "epoch": 1.3961319681456201, + "grad_norm": 2.227200428647969, + "learning_rate": 1.024627908014706e-06, + "loss": 0.0873, + "step": 6136 + }, + { + "epoch": 1.3963594994311719, + "grad_norm": 1.3512910037945858, + "learning_rate": 1.0245592043407784e-06, + "loss": 0.0561, + "step": 6137 + }, + { + "epoch": 1.3965870307167236, + "grad_norm": 1.5050664712520612, + "learning_rate": 1.0244904925005924e-06, + "loss": 0.0631, + "step": 6138 + }, + { + "epoch": 1.3968145620022754, + "grad_norm": 1.3276673190831771, + "learning_rate": 1.0244217724955523e-06, + "loss": 0.1134, + "step": 6139 + }, + { + "epoch": 1.3970420932878271, + "grad_norm": 2.11597419896245, + "learning_rate": 1.0243530443270627e-06, + "loss": 0.0511, + "step": 6140 + }, + { + "epoch": 1.3972696245733789, + "grad_norm": 1.383620506367811, + "learning_rate": 1.0242843079965281e-06, + "loss": 0.0396, + "step": 6141 + }, + { + "epoch": 1.3974971558589306, + "grad_norm": 2.0579977081129304, + "learning_rate": 1.024215563505354e-06, + "loss": 0.0488, + "step": 6142 + }, + { + "epoch": 1.3977246871444824, + "grad_norm": 1.6176146828457385, + "learning_rate": 1.0241468108549443e-06, + "loss": 0.0657, + "step": 6143 + }, + { + "epoch": 1.3979522184300341, + "grad_norm": 1.1605364666028937, + "learning_rate": 1.024078050046705e-06, + "loss": 0.1009, + "step": 6144 + }, + { + "epoch": 1.398179749715586, + "grad_norm": 1.9390494144680035, + "learning_rate": 1.0240092810820412e-06, + "loss": 0.1228, + "step": 6145 + }, + { + "epoch": 1.3984072810011376, + "grad_norm": 2.2751796101471444, + "learning_rate": 1.0239405039623585e-06, + "loss": 0.1679, + "step": 6146 + }, + { + "epoch": 1.3986348122866894, + "grad_norm": 1.4446668035390078, + "learning_rate": 1.0238717186890625e-06, + "loss": 0.047, + "step": 6147 + }, + { + "epoch": 1.3988623435722412, + "grad_norm": 1.2522080333454766, + "learning_rate": 1.0238029252635591e-06, + "loss": 0.0541, + "step": 6148 + }, + { + "epoch": 1.399089874857793, + "grad_norm": 1.206361926946461, + "learning_rate": 1.0237341236872544e-06, + "loss": 0.0262, + "step": 6149 + }, + { + "epoch": 1.3993174061433447, + "grad_norm": 1.3900768598387714, + "learning_rate": 1.0236653139615542e-06, + "loss": 0.0632, + "step": 6150 + }, + { + "epoch": 1.3995449374288964, + "grad_norm": 1.446506693950247, + "learning_rate": 1.0235964960878655e-06, + "loss": 0.135, + "step": 6151 + }, + { + "epoch": 1.3997724687144482, + "grad_norm": 1.2859328173017583, + "learning_rate": 1.023527670067594e-06, + "loss": 0.0906, + "step": 6152 + }, + { + "epoch": 1.4, + "grad_norm": 1.8510910035348387, + "learning_rate": 1.023458835902147e-06, + "loss": 0.1158, + "step": 6153 + }, + { + "epoch": 1.4002275312855517, + "grad_norm": 2.006291037320859, + "learning_rate": 1.0233899935929311e-06, + "loss": 0.0972, + "step": 6154 + }, + { + "epoch": 1.4004550625711034, + "grad_norm": 1.9171493600614367, + "learning_rate": 1.0233211431413534e-06, + "loss": 0.1664, + "step": 6155 + }, + { + "epoch": 1.4006825938566552, + "grad_norm": 1.9981434882673672, + "learning_rate": 1.023252284548821e-06, + "loss": 0.1034, + "step": 6156 + }, + { + "epoch": 1.400910125142207, + "grad_norm": 1.6198664782069099, + "learning_rate": 1.0231834178167412e-06, + "loss": 0.049, + "step": 6157 + }, + { + "epoch": 1.401137656427759, + "grad_norm": 2.1959481687542817, + "learning_rate": 1.0231145429465216e-06, + "loss": 0.0896, + "step": 6158 + }, + { + "epoch": 1.4013651877133106, + "grad_norm": 1.5186673742926442, + "learning_rate": 1.02304565993957e-06, + "loss": 0.0876, + "step": 6159 + }, + { + "epoch": 1.4015927189988624, + "grad_norm": 1.6726004894471778, + "learning_rate": 1.022976768797294e-06, + "loss": 0.0649, + "step": 6160 + }, + { + "epoch": 1.4018202502844141, + "grad_norm": 1.7616787253173187, + "learning_rate": 1.0229078695211015e-06, + "loss": 0.0689, + "step": 6161 + }, + { + "epoch": 1.402047781569966, + "grad_norm": 2.101238409018286, + "learning_rate": 1.0228389621124011e-06, + "loss": 0.0962, + "step": 6162 + }, + { + "epoch": 1.4022753128555177, + "grad_norm": 18.423779024002584, + "learning_rate": 1.022770046572601e-06, + "loss": 0.1132, + "step": 6163 + }, + { + "epoch": 1.4025028441410694, + "grad_norm": 1.3626382690966223, + "learning_rate": 1.0227011229031095e-06, + "loss": 0.0923, + "step": 6164 + }, + { + "epoch": 1.4027303754266212, + "grad_norm": 1.5664192058369393, + "learning_rate": 1.0226321911053353e-06, + "loss": 0.0883, + "step": 6165 + }, + { + "epoch": 1.402957906712173, + "grad_norm": 2.1407280846260557, + "learning_rate": 1.0225632511806873e-06, + "loss": 0.0813, + "step": 6166 + }, + { + "epoch": 1.4031854379977247, + "grad_norm": 1.9126432331266539, + "learning_rate": 1.0224943031305747e-06, + "loss": 0.083, + "step": 6167 + }, + { + "epoch": 1.4034129692832764, + "grad_norm": 1.2929386744191405, + "learning_rate": 1.0224253469564067e-06, + "loss": 0.0842, + "step": 6168 + }, + { + "epoch": 1.4036405005688282, + "grad_norm": 1.5870857265163638, + "learning_rate": 1.0223563826595923e-06, + "loss": 0.1172, + "step": 6169 + }, + { + "epoch": 1.40386803185438, + "grad_norm": 1.1823753570324391, + "learning_rate": 1.0222874102415412e-06, + "loss": 0.0551, + "step": 6170 + }, + { + "epoch": 1.4040955631399317, + "grad_norm": 2.996937816725621, + "learning_rate": 1.0222184297036628e-06, + "loss": 0.1564, + "step": 6171 + }, + { + "epoch": 1.4043230944254836, + "grad_norm": 2.2927668992732997, + "learning_rate": 1.0221494410473674e-06, + "loss": 0.1045, + "step": 6172 + }, + { + "epoch": 1.4045506257110354, + "grad_norm": 1.1281259385923086, + "learning_rate": 1.0220804442740648e-06, + "loss": 0.042, + "step": 6173 + }, + { + "epoch": 1.4047781569965871, + "grad_norm": 1.9469775916421443, + "learning_rate": 1.022011439385165e-06, + "loss": 0.0999, + "step": 6174 + }, + { + "epoch": 1.405005688282139, + "grad_norm": 1.6529622749650812, + "learning_rate": 1.0219424263820784e-06, + "loss": 0.0777, + "step": 6175 + }, + { + "epoch": 1.4052332195676907, + "grad_norm": 1.803447728801388, + "learning_rate": 1.0218734052662158e-06, + "loss": 0.0877, + "step": 6176 + }, + { + "epoch": 1.4054607508532424, + "grad_norm": 1.0877665384054194, + "learning_rate": 1.0218043760389875e-06, + "loss": 0.1214, + "step": 6177 + }, + { + "epoch": 1.4056882821387942, + "grad_norm": 3.036932254264831, + "learning_rate": 1.0217353387018045e-06, + "loss": 0.1388, + "step": 6178 + }, + { + "epoch": 1.405915813424346, + "grad_norm": 1.6432118346236038, + "learning_rate": 1.0216662932560779e-06, + "loss": 0.091, + "step": 6179 + }, + { + "epoch": 1.4061433447098977, + "grad_norm": 1.4641635382861418, + "learning_rate": 1.0215972397032185e-06, + "loss": 0.0498, + "step": 6180 + }, + { + "epoch": 1.4063708759954494, + "grad_norm": 2.2055527595278366, + "learning_rate": 1.0215281780446378e-06, + "loss": 0.1207, + "step": 6181 + }, + { + "epoch": 1.4065984072810012, + "grad_norm": 1.3294688000691555, + "learning_rate": 1.0214591082817477e-06, + "loss": 0.1173, + "step": 6182 + }, + { + "epoch": 1.406825938566553, + "grad_norm": 2.179512895860987, + "learning_rate": 1.0213900304159592e-06, + "loss": 0.0584, + "step": 6183 + }, + { + "epoch": 1.4070534698521047, + "grad_norm": 1.6222476017901606, + "learning_rate": 1.0213209444486844e-06, + "loss": 0.0875, + "step": 6184 + }, + { + "epoch": 1.4072810011376564, + "grad_norm": 1.93817268709916, + "learning_rate": 1.0212518503813356e-06, + "loss": 0.1459, + "step": 6185 + }, + { + "epoch": 1.4075085324232082, + "grad_norm": 1.741064264886961, + "learning_rate": 1.0211827482153244e-06, + "loss": 0.0728, + "step": 6186 + }, + { + "epoch": 1.40773606370876, + "grad_norm": 2.5576686669016864, + "learning_rate": 1.0211136379520636e-06, + "loss": 0.2464, + "step": 6187 + }, + { + "epoch": 1.4079635949943117, + "grad_norm": 1.187363532717576, + "learning_rate": 1.0210445195929653e-06, + "loss": 0.0429, + "step": 6188 + }, + { + "epoch": 1.4081911262798634, + "grad_norm": 1.7031117991008564, + "learning_rate": 1.0209753931394426e-06, + "loss": 0.1466, + "step": 6189 + }, + { + "epoch": 1.4084186575654152, + "grad_norm": 1.4426123688536225, + "learning_rate": 1.0209062585929077e-06, + "loss": 0.0296, + "step": 6190 + }, + { + "epoch": 1.408646188850967, + "grad_norm": 1.9144448784443266, + "learning_rate": 1.0208371159547742e-06, + "loss": 0.1058, + "step": 6191 + }, + { + "epoch": 1.4088737201365187, + "grad_norm": 1.244136782652026, + "learning_rate": 1.020767965226455e-06, + "loss": 0.1711, + "step": 6192 + }, + { + "epoch": 1.4091012514220704, + "grad_norm": 1.452605807632957, + "learning_rate": 1.0206988064093633e-06, + "loss": 0.0815, + "step": 6193 + }, + { + "epoch": 1.4093287827076222, + "grad_norm": 1.8442597312009432, + "learning_rate": 1.0206296395049128e-06, + "loss": 0.0674, + "step": 6194 + }, + { + "epoch": 1.409556313993174, + "grad_norm": 1.7262145580460375, + "learning_rate": 1.020560464514517e-06, + "loss": 0.0657, + "step": 6195 + }, + { + "epoch": 1.4097838452787257, + "grad_norm": 1.5207284157163985, + "learning_rate": 1.0204912814395898e-06, + "loss": 0.105, + "step": 6196 + }, + { + "epoch": 1.4100113765642777, + "grad_norm": 2.3815556969730314, + "learning_rate": 1.020422090281545e-06, + "loss": 0.0772, + "step": 6197 + }, + { + "epoch": 1.4102389078498294, + "grad_norm": 2.268950151122507, + "learning_rate": 1.0203528910417967e-06, + "loss": 0.1148, + "step": 6198 + }, + { + "epoch": 1.4104664391353812, + "grad_norm": 1.7550413594498013, + "learning_rate": 1.0202836837217597e-06, + "loss": 0.0923, + "step": 6199 + }, + { + "epoch": 1.410693970420933, + "grad_norm": 1.1355924309874208, + "learning_rate": 1.0202144683228478e-06, + "loss": 0.065, + "step": 6200 + }, + { + "epoch": 1.4109215017064847, + "grad_norm": 1.3913974009246857, + "learning_rate": 1.0201452448464762e-06, + "loss": 0.073, + "step": 6201 + }, + { + "epoch": 1.4111490329920364, + "grad_norm": 5.592058527828976, + "learning_rate": 1.0200760132940597e-06, + "loss": 0.2201, + "step": 6202 + }, + { + "epoch": 1.4113765642775882, + "grad_norm": 1.1062997370223235, + "learning_rate": 1.0200067736670125e-06, + "loss": 0.0628, + "step": 6203 + }, + { + "epoch": 1.41160409556314, + "grad_norm": 0.8021573016810204, + "learning_rate": 1.0199375259667505e-06, + "loss": 0.0557, + "step": 6204 + }, + { + "epoch": 1.4118316268486917, + "grad_norm": 1.344669498905401, + "learning_rate": 1.0198682701946889e-06, + "loss": 0.1438, + "step": 6205 + }, + { + "epoch": 1.4120591581342434, + "grad_norm": 2.0484954391129833, + "learning_rate": 1.0197990063522428e-06, + "loss": 0.0764, + "step": 6206 + }, + { + "epoch": 1.4122866894197952, + "grad_norm": 1.1107361142156864, + "learning_rate": 1.0197297344408284e-06, + "loss": 0.0598, + "step": 6207 + }, + { + "epoch": 1.412514220705347, + "grad_norm": 1.818289355783258, + "learning_rate": 1.0196604544618607e-06, + "loss": 0.0911, + "step": 6208 + }, + { + "epoch": 1.4127417519908987, + "grad_norm": 2.3440454116714324, + "learning_rate": 1.0195911664167562e-06, + "loss": 0.075, + "step": 6209 + }, + { + "epoch": 1.4129692832764504, + "grad_norm": 1.4981628212882372, + "learning_rate": 1.0195218703069311e-06, + "loss": 0.1064, + "step": 6210 + }, + { + "epoch": 1.4131968145620024, + "grad_norm": 1.6748847246596112, + "learning_rate": 1.0194525661338014e-06, + "loss": 0.0789, + "step": 6211 + }, + { + "epoch": 1.4134243458475542, + "grad_norm": 1.6005524221534388, + "learning_rate": 1.0193832538987838e-06, + "loss": 0.0713, + "step": 6212 + }, + { + "epoch": 1.413651877133106, + "grad_norm": 1.737186728394442, + "learning_rate": 1.0193139336032945e-06, + "loss": 0.0587, + "step": 6213 + }, + { + "epoch": 1.4138794084186577, + "grad_norm": 1.0915457381923228, + "learning_rate": 1.0192446052487505e-06, + "loss": 0.0754, + "step": 6214 + }, + { + "epoch": 1.4141069397042094, + "grad_norm": 1.7435000498468016, + "learning_rate": 1.0191752688365691e-06, + "loss": 0.0765, + "step": 6215 + }, + { + "epoch": 1.4143344709897612, + "grad_norm": 1.8752909733484666, + "learning_rate": 1.019105924368167e-06, + "loss": 0.0922, + "step": 6216 + }, + { + "epoch": 1.414562002275313, + "grad_norm": 1.4392973922132166, + "learning_rate": 1.0190365718449616e-06, + "loss": 0.0568, + "step": 6217 + }, + { + "epoch": 1.4147895335608647, + "grad_norm": 2.2754532688396485, + "learning_rate": 1.0189672112683704e-06, + "loss": 0.1008, + "step": 6218 + }, + { + "epoch": 1.4150170648464164, + "grad_norm": 2.431173380429492, + "learning_rate": 1.0188978426398107e-06, + "loss": 0.099, + "step": 6219 + }, + { + "epoch": 1.4152445961319682, + "grad_norm": 1.5991864100743545, + "learning_rate": 1.0188284659607007e-06, + "loss": 0.143, + "step": 6220 + }, + { + "epoch": 1.41547212741752, + "grad_norm": 1.3594759675442543, + "learning_rate": 1.018759081232458e-06, + "loss": 0.1335, + "step": 6221 + }, + { + "epoch": 1.4156996587030717, + "grad_norm": 0.7702332161662481, + "learning_rate": 1.0186896884565005e-06, + "loss": 0.031, + "step": 6222 + }, + { + "epoch": 1.4159271899886234, + "grad_norm": 1.7864404602384039, + "learning_rate": 1.0186202876342473e-06, + "loss": 0.0831, + "step": 6223 + }, + { + "epoch": 1.4161547212741752, + "grad_norm": 1.484634480862312, + "learning_rate": 1.0185508787671162e-06, + "loss": 0.0801, + "step": 6224 + }, + { + "epoch": 1.416382252559727, + "grad_norm": 1.7693126707625348, + "learning_rate": 1.0184814618565257e-06, + "loss": 0.0877, + "step": 6225 + }, + { + "epoch": 1.4166097838452787, + "grad_norm": 2.12446577736148, + "learning_rate": 1.0184120369038948e-06, + "loss": 0.1375, + "step": 6226 + }, + { + "epoch": 1.4168373151308304, + "grad_norm": 2.71536098313585, + "learning_rate": 1.0183426039106425e-06, + "loss": 0.0992, + "step": 6227 + }, + { + "epoch": 1.4170648464163822, + "grad_norm": 2.3403843742524977, + "learning_rate": 1.0182731628781876e-06, + "loss": 0.1272, + "step": 6228 + }, + { + "epoch": 1.417292377701934, + "grad_norm": 0.8873199147810173, + "learning_rate": 1.0182037138079494e-06, + "loss": 0.0724, + "step": 6229 + }, + { + "epoch": 1.4175199089874857, + "grad_norm": 2.1962588500843285, + "learning_rate": 1.0181342567013477e-06, + "loss": 0.0874, + "step": 6230 + }, + { + "epoch": 1.4177474402730375, + "grad_norm": 1.046799205589163, + "learning_rate": 1.0180647915598017e-06, + "loss": 0.0615, + "step": 6231 + }, + { + "epoch": 1.4179749715585892, + "grad_norm": 1.559125762785821, + "learning_rate": 1.017995318384731e-06, + "loss": 0.0967, + "step": 6232 + }, + { + "epoch": 1.418202502844141, + "grad_norm": 2.221222240769814, + "learning_rate": 1.017925837177556e-06, + "loss": 0.1052, + "step": 6233 + }, + { + "epoch": 1.4184300341296927, + "grad_norm": 1.148681278966676, + "learning_rate": 1.0178563479396964e-06, + "loss": 0.0888, + "step": 6234 + }, + { + "epoch": 1.4186575654152445, + "grad_norm": 2.368699801178719, + "learning_rate": 1.0177868506725725e-06, + "loss": 0.0786, + "step": 6235 + }, + { + "epoch": 1.4188850967007964, + "grad_norm": 1.8803457207584942, + "learning_rate": 1.0177173453776047e-06, + "loss": 0.1253, + "step": 6236 + }, + { + "epoch": 1.4191126279863482, + "grad_norm": 1.8625616067005561, + "learning_rate": 1.0176478320562136e-06, + "loss": 0.1561, + "step": 6237 + }, + { + "epoch": 1.4193401592719, + "grad_norm": 1.4241783140991238, + "learning_rate": 1.01757831070982e-06, + "loss": 0.1423, + "step": 6238 + }, + { + "epoch": 1.4195676905574517, + "grad_norm": 1.1462264802545075, + "learning_rate": 1.0175087813398446e-06, + "loss": 0.1293, + "step": 6239 + }, + { + "epoch": 1.4197952218430034, + "grad_norm": 1.7660807457711354, + "learning_rate": 1.0174392439477087e-06, + "loss": 0.1749, + "step": 6240 + }, + { + "epoch": 1.4200227531285552, + "grad_norm": 1.731636472675248, + "learning_rate": 1.0173696985348333e-06, + "loss": 0.2122, + "step": 6241 + }, + { + "epoch": 1.420250284414107, + "grad_norm": 3.206125566573037, + "learning_rate": 1.0173001451026396e-06, + "loss": 0.176, + "step": 6242 + }, + { + "epoch": 1.4204778156996587, + "grad_norm": 1.7861193001189515, + "learning_rate": 1.0172305836525498e-06, + "loss": 0.1038, + "step": 6243 + }, + { + "epoch": 1.4207053469852104, + "grad_norm": 1.564303100927579, + "learning_rate": 1.017161014185985e-06, + "loss": 0.0623, + "step": 6244 + }, + { + "epoch": 1.4209328782707622, + "grad_norm": 2.39397601371307, + "learning_rate": 1.0170914367043672e-06, + "loss": 0.0967, + "step": 6245 + }, + { + "epoch": 1.421160409556314, + "grad_norm": 2.2487565032595995, + "learning_rate": 1.0170218512091188e-06, + "loss": 0.1099, + "step": 6246 + }, + { + "epoch": 1.4213879408418657, + "grad_norm": 2.4116148271697786, + "learning_rate": 1.0169522577016614e-06, + "loss": 0.1188, + "step": 6247 + }, + { + "epoch": 1.4216154721274175, + "grad_norm": 1.2570520555601248, + "learning_rate": 1.016882656183418e-06, + "loss": 0.0385, + "step": 6248 + }, + { + "epoch": 1.4218430034129692, + "grad_norm": 2.6411908340836767, + "learning_rate": 1.0168130466558106e-06, + "loss": 0.0763, + "step": 6249 + }, + { + "epoch": 1.4220705346985212, + "grad_norm": 1.7982694079050958, + "learning_rate": 1.0167434291202622e-06, + "loss": 0.0911, + "step": 6250 + }, + { + "epoch": 1.422298065984073, + "grad_norm": 1.3577959316062924, + "learning_rate": 1.0166738035781954e-06, + "loss": 0.0727, + "step": 6251 + }, + { + "epoch": 1.4225255972696247, + "grad_norm": 1.8897772791113623, + "learning_rate": 1.0166041700310334e-06, + "loss": 0.0667, + "step": 6252 + }, + { + "epoch": 1.4227531285551764, + "grad_norm": 1.2512416221815659, + "learning_rate": 1.0165345284801995e-06, + "loss": 0.0558, + "step": 6253 + }, + { + "epoch": 1.4229806598407282, + "grad_norm": 1.1610234015826426, + "learning_rate": 1.0164648789271167e-06, + "loss": 0.1102, + "step": 6254 + }, + { + "epoch": 1.42320819112628, + "grad_norm": 2.7364696648328106, + "learning_rate": 1.016395221373209e-06, + "loss": 0.1659, + "step": 6255 + }, + { + "epoch": 1.4234357224118317, + "grad_norm": 2.492044091165344, + "learning_rate": 1.0163255558198995e-06, + "loss": 0.1253, + "step": 6256 + }, + { + "epoch": 1.4236632536973834, + "grad_norm": 1.4141996293268364, + "learning_rate": 1.0162558822686123e-06, + "loss": 0.063, + "step": 6257 + }, + { + "epoch": 1.4238907849829352, + "grad_norm": 2.1623801632900137, + "learning_rate": 1.0161862007207715e-06, + "loss": 0.2044, + "step": 6258 + }, + { + "epoch": 1.424118316268487, + "grad_norm": 3.0022631890111215, + "learning_rate": 1.0161165111778013e-06, + "loss": 0.071, + "step": 6259 + }, + { + "epoch": 1.4243458475540387, + "grad_norm": 1.5430074698777752, + "learning_rate": 1.0160468136411258e-06, + "loss": 0.0437, + "step": 6260 + }, + { + "epoch": 1.4245733788395905, + "grad_norm": 1.4418875571900045, + "learning_rate": 1.0159771081121697e-06, + "loss": 0.1592, + "step": 6261 + }, + { + "epoch": 1.4248009101251422, + "grad_norm": 1.1023568443010037, + "learning_rate": 1.0159073945923575e-06, + "loss": 0.0544, + "step": 6262 + }, + { + "epoch": 1.425028441410694, + "grad_norm": 2.2051068207070386, + "learning_rate": 1.015837673083114e-06, + "loss": 0.1451, + "step": 6263 + }, + { + "epoch": 1.4252559726962457, + "grad_norm": 1.8003514085030992, + "learning_rate": 1.0157679435858643e-06, + "loss": 0.1134, + "step": 6264 + }, + { + "epoch": 1.4254835039817975, + "grad_norm": 1.671768891780443, + "learning_rate": 1.0156982061020335e-06, + "loss": 0.0697, + "step": 6265 + }, + { + "epoch": 1.4257110352673492, + "grad_norm": 1.691638560811478, + "learning_rate": 1.0156284606330468e-06, + "loss": 0.1116, + "step": 6266 + }, + { + "epoch": 1.425938566552901, + "grad_norm": 1.668218037713917, + "learning_rate": 1.0155587071803298e-06, + "loss": 0.0734, + "step": 6267 + }, + { + "epoch": 1.4261660978384527, + "grad_norm": 1.4322553363365562, + "learning_rate": 1.0154889457453082e-06, + "loss": 0.0602, + "step": 6268 + }, + { + "epoch": 1.4263936291240045, + "grad_norm": 1.15268242289466, + "learning_rate": 1.0154191763294077e-06, + "loss": 0.0818, + "step": 6269 + }, + { + "epoch": 1.4266211604095562, + "grad_norm": 1.80984807928571, + "learning_rate": 1.015349398934054e-06, + "loss": 0.0739, + "step": 6270 + }, + { + "epoch": 1.426848691695108, + "grad_norm": 2.6026819179736687, + "learning_rate": 1.0152796135606739e-06, + "loss": 0.1506, + "step": 6271 + }, + { + "epoch": 1.4270762229806597, + "grad_norm": 1.5071088896936038, + "learning_rate": 1.015209820210693e-06, + "loss": 0.1376, + "step": 6272 + }, + { + "epoch": 1.4273037542662115, + "grad_norm": 1.4586819449421182, + "learning_rate": 1.015140018885538e-06, + "loss": 0.1053, + "step": 6273 + }, + { + "epoch": 1.4275312855517632, + "grad_norm": 1.1403111332425095, + "learning_rate": 1.0150702095866354e-06, + "loss": 0.0494, + "step": 6274 + }, + { + "epoch": 1.4277588168373152, + "grad_norm": 1.626188422438232, + "learning_rate": 1.0150003923154124e-06, + "loss": 0.1072, + "step": 6275 + }, + { + "epoch": 1.427986348122867, + "grad_norm": 2.510844293567412, + "learning_rate": 1.0149305670732953e-06, + "loss": 0.1233, + "step": 6276 + }, + { + "epoch": 1.4282138794084187, + "grad_norm": 1.1258865254836934, + "learning_rate": 1.0148607338617118e-06, + "loss": 0.0837, + "step": 6277 + }, + { + "epoch": 1.4284414106939705, + "grad_norm": 2.059538582647301, + "learning_rate": 1.0147908926820887e-06, + "loss": 0.1354, + "step": 6278 + }, + { + "epoch": 1.4286689419795222, + "grad_norm": 2.2333199746531194, + "learning_rate": 1.014721043535854e-06, + "loss": 0.0965, + "step": 6279 + }, + { + "epoch": 1.428896473265074, + "grad_norm": 2.348990396070376, + "learning_rate": 1.0146511864244344e-06, + "loss": 0.117, + "step": 6280 + }, + { + "epoch": 1.4291240045506257, + "grad_norm": 2.579726909586119, + "learning_rate": 1.0145813213492587e-06, + "loss": 0.0875, + "step": 6281 + }, + { + "epoch": 1.4293515358361775, + "grad_norm": 2.491278810072405, + "learning_rate": 1.0145114483117539e-06, + "loss": 0.0861, + "step": 6282 + }, + { + "epoch": 1.4295790671217292, + "grad_norm": 1.785513438132783, + "learning_rate": 1.0144415673133485e-06, + "loss": 0.0973, + "step": 6283 + }, + { + "epoch": 1.429806598407281, + "grad_norm": 1.43857963983858, + "learning_rate": 1.0143716783554709e-06, + "loss": 0.0474, + "step": 6284 + }, + { + "epoch": 1.4300341296928327, + "grad_norm": 1.3085731530788816, + "learning_rate": 1.0143017814395489e-06, + "loss": 0.084, + "step": 6285 + }, + { + "epoch": 1.4302616609783845, + "grad_norm": 1.8064962742353994, + "learning_rate": 1.0142318765670117e-06, + "loss": 0.077, + "step": 6286 + }, + { + "epoch": 1.4304891922639362, + "grad_norm": 1.878771887521387, + "learning_rate": 1.0141619637392878e-06, + "loss": 0.1095, + "step": 6287 + }, + { + "epoch": 1.430716723549488, + "grad_norm": 2.343408941947069, + "learning_rate": 1.0140920429578061e-06, + "loss": 0.1102, + "step": 6288 + }, + { + "epoch": 1.43094425483504, + "grad_norm": 2.0560219852914607, + "learning_rate": 1.0140221142239957e-06, + "loss": 0.0991, + "step": 6289 + }, + { + "epoch": 1.4311717861205917, + "grad_norm": 0.795792793486454, + "learning_rate": 1.0139521775392856e-06, + "loss": 0.0553, + "step": 6290 + }, + { + "epoch": 1.4313993174061435, + "grad_norm": 1.8826815884172625, + "learning_rate": 1.0138822329051052e-06, + "loss": 0.1291, + "step": 6291 + }, + { + "epoch": 1.4316268486916952, + "grad_norm": 1.8726101517248754, + "learning_rate": 1.0138122803228843e-06, + "loss": 0.0669, + "step": 6292 + }, + { + "epoch": 1.431854379977247, + "grad_norm": 2.3660511817894196, + "learning_rate": 1.0137423197940527e-06, + "loss": 0.1365, + "step": 6293 + }, + { + "epoch": 1.4320819112627987, + "grad_norm": 1.7241008303823981, + "learning_rate": 1.0136723513200396e-06, + "loss": 0.1627, + "step": 6294 + }, + { + "epoch": 1.4323094425483505, + "grad_norm": 1.4719757229629116, + "learning_rate": 1.0136023749022759e-06, + "loss": 0.157, + "step": 6295 + }, + { + "epoch": 1.4325369738339022, + "grad_norm": 1.5457067737199763, + "learning_rate": 1.013532390542191e-06, + "loss": 0.0758, + "step": 6296 + }, + { + "epoch": 1.432764505119454, + "grad_norm": 1.563734167251608, + "learning_rate": 1.0134623982412156e-06, + "loss": 0.0911, + "step": 6297 + }, + { + "epoch": 1.4329920364050057, + "grad_norm": 1.9585239556119394, + "learning_rate": 1.0133923980007804e-06, + "loss": 0.0684, + "step": 6298 + }, + { + "epoch": 1.4332195676905575, + "grad_norm": 1.3489415515735705, + "learning_rate": 1.013322389822316e-06, + "loss": 0.05, + "step": 6299 + }, + { + "epoch": 1.4334470989761092, + "grad_norm": 1.7767579179072022, + "learning_rate": 1.0132523737072528e-06, + "loss": 0.1488, + "step": 6300 + }, + { + "epoch": 1.433674630261661, + "grad_norm": 1.8711312931697477, + "learning_rate": 1.0131823496570222e-06, + "loss": 0.0709, + "step": 6301 + }, + { + "epoch": 1.4339021615472127, + "grad_norm": 1.9155497814506361, + "learning_rate": 1.0131123176730555e-06, + "loss": 0.0941, + "step": 6302 + }, + { + "epoch": 1.4341296928327645, + "grad_norm": 2.3525073153549143, + "learning_rate": 1.0130422777567837e-06, + "loss": 0.1123, + "step": 6303 + }, + { + "epoch": 1.4343572241183162, + "grad_norm": 1.7137222313794656, + "learning_rate": 1.0129722299096385e-06, + "loss": 0.0805, + "step": 6304 + }, + { + "epoch": 1.434584755403868, + "grad_norm": 1.7017112708859887, + "learning_rate": 1.0129021741330514e-06, + "loss": 0.0906, + "step": 6305 + }, + { + "epoch": 1.4348122866894197, + "grad_norm": 2.550190726939763, + "learning_rate": 1.0128321104284541e-06, + "loss": 0.1724, + "step": 6306 + }, + { + "epoch": 1.4350398179749715, + "grad_norm": 1.950569258241582, + "learning_rate": 1.0127620387972789e-06, + "loss": 0.0888, + "step": 6307 + }, + { + "epoch": 1.4352673492605232, + "grad_norm": 1.0716771624591401, + "learning_rate": 1.0126919592409578e-06, + "loss": 0.0818, + "step": 6308 + }, + { + "epoch": 1.435494880546075, + "grad_norm": 2.719604541403, + "learning_rate": 1.012621871760923e-06, + "loss": 0.1134, + "step": 6309 + }, + { + "epoch": 1.4357224118316267, + "grad_norm": 1.5418381863106374, + "learning_rate": 1.012551776358607e-06, + "loss": 0.0622, + "step": 6310 + }, + { + "epoch": 1.4359499431171785, + "grad_norm": 1.094940212737539, + "learning_rate": 1.0124816730354428e-06, + "loss": 0.0349, + "step": 6311 + }, + { + "epoch": 1.4361774744027302, + "grad_norm": 2.214067429631255, + "learning_rate": 1.0124115617928626e-06, + "loss": 0.1129, + "step": 6312 + }, + { + "epoch": 1.4364050056882822, + "grad_norm": 1.9161547207577005, + "learning_rate": 1.0123414426322995e-06, + "loss": 0.0811, + "step": 6313 + }, + { + "epoch": 1.436632536973834, + "grad_norm": 1.4061438449920354, + "learning_rate": 1.0122713155551867e-06, + "loss": 0.1189, + "step": 6314 + }, + { + "epoch": 1.4368600682593857, + "grad_norm": 2.1150900833574315, + "learning_rate": 1.0122011805629574e-06, + "loss": 0.1468, + "step": 6315 + }, + { + "epoch": 1.4370875995449375, + "grad_norm": 1.3331813768009924, + "learning_rate": 1.0121310376570454e-06, + "loss": 0.1308, + "step": 6316 + }, + { + "epoch": 1.4373151308304892, + "grad_norm": 2.505967190170033, + "learning_rate": 1.0120608868388837e-06, + "loss": 0.1062, + "step": 6317 + }, + { + "epoch": 1.437542662116041, + "grad_norm": 1.9193512163717878, + "learning_rate": 1.0119907281099066e-06, + "loss": 0.1135, + "step": 6318 + }, + { + "epoch": 1.4377701934015927, + "grad_norm": 1.6143484552181144, + "learning_rate": 1.0119205614715476e-06, + "loss": 0.0827, + "step": 6319 + }, + { + "epoch": 1.4379977246871445, + "grad_norm": 1.6262807068571428, + "learning_rate": 1.011850386925241e-06, + "loss": 0.065, + "step": 6320 + }, + { + "epoch": 1.4382252559726962, + "grad_norm": 3.4130385631807187, + "learning_rate": 1.011780204472421e-06, + "loss": 0.0972, + "step": 6321 + }, + { + "epoch": 1.438452787258248, + "grad_norm": 1.7183448758361486, + "learning_rate": 1.0117100141145217e-06, + "loss": 0.087, + "step": 6322 + }, + { + "epoch": 1.4386803185437997, + "grad_norm": 1.7894363102296018, + "learning_rate": 1.0116398158529782e-06, + "loss": 0.0868, + "step": 6323 + }, + { + "epoch": 1.4389078498293515, + "grad_norm": 1.3683813014203625, + "learning_rate": 1.0115696096892247e-06, + "loss": 0.1473, + "step": 6324 + }, + { + "epoch": 1.4391353811149032, + "grad_norm": 2.167526729040326, + "learning_rate": 1.0114993956246968e-06, + "loss": 0.0716, + "step": 6325 + }, + { + "epoch": 1.439362912400455, + "grad_norm": 1.9967306110570626, + "learning_rate": 1.0114291736608289e-06, + "loss": 0.1219, + "step": 6326 + }, + { + "epoch": 1.4395904436860067, + "grad_norm": 1.9523121423220702, + "learning_rate": 1.0113589437990562e-06, + "loss": 0.1192, + "step": 6327 + }, + { + "epoch": 1.4398179749715587, + "grad_norm": 1.4074354798186564, + "learning_rate": 1.0112887060408144e-06, + "loss": 0.1657, + "step": 6328 + }, + { + "epoch": 1.4400455062571105, + "grad_norm": 2.224946954285757, + "learning_rate": 1.0112184603875387e-06, + "loss": 0.1528, + "step": 6329 + }, + { + "epoch": 1.4402730375426622, + "grad_norm": 1.7404203745348572, + "learning_rate": 1.0111482068406655e-06, + "loss": 0.0488, + "step": 6330 + }, + { + "epoch": 1.440500568828214, + "grad_norm": 2.139255550161412, + "learning_rate": 1.0110779454016299e-06, + "loss": 0.0723, + "step": 6331 + }, + { + "epoch": 1.4407281001137657, + "grad_norm": 0.9791070299648054, + "learning_rate": 1.011007676071868e-06, + "loss": 0.0234, + "step": 6332 + }, + { + "epoch": 1.4409556313993175, + "grad_norm": 1.255456216854933, + "learning_rate": 1.0109373988528161e-06, + "loss": 0.0931, + "step": 6333 + }, + { + "epoch": 1.4411831626848692, + "grad_norm": 2.415122633007749, + "learning_rate": 1.010867113745911e-06, + "loss": 0.0976, + "step": 6334 + }, + { + "epoch": 1.441410693970421, + "grad_norm": 1.1072274876147699, + "learning_rate": 1.0107968207525884e-06, + "loss": 0.0799, + "step": 6335 + }, + { + "epoch": 1.4416382252559727, + "grad_norm": 2.294103829148771, + "learning_rate": 1.0107265198742855e-06, + "loss": 0.1874, + "step": 6336 + }, + { + "epoch": 1.4418657565415245, + "grad_norm": 1.6789846212867021, + "learning_rate": 1.010656211112439e-06, + "loss": 0.127, + "step": 6337 + }, + { + "epoch": 1.4420932878270762, + "grad_norm": 1.7387215182234683, + "learning_rate": 1.0105858944684856e-06, + "loss": 0.1075, + "step": 6338 + }, + { + "epoch": 1.442320819112628, + "grad_norm": 1.758435671842687, + "learning_rate": 1.010515569943863e-06, + "loss": 0.1135, + "step": 6339 + }, + { + "epoch": 1.4425483503981797, + "grad_norm": 1.875372603637353, + "learning_rate": 1.0104452375400078e-06, + "loss": 0.1024, + "step": 6340 + }, + { + "epoch": 1.4427758816837315, + "grad_norm": 1.1327411263887621, + "learning_rate": 1.0103748972583582e-06, + "loss": 0.1037, + "step": 6341 + }, + { + "epoch": 1.4430034129692833, + "grad_norm": 1.6265307975280838, + "learning_rate": 1.0103045491003514e-06, + "loss": 0.0978, + "step": 6342 + }, + { + "epoch": 1.443230944254835, + "grad_norm": 0.9532939564545074, + "learning_rate": 1.0102341930674252e-06, + "loss": 0.0985, + "step": 6343 + }, + { + "epoch": 1.4434584755403868, + "grad_norm": 2.1579266950046216, + "learning_rate": 1.0101638291610176e-06, + "loss": 0.0841, + "step": 6344 + }, + { + "epoch": 1.4436860068259385, + "grad_norm": 1.912319337135995, + "learning_rate": 1.0100934573825668e-06, + "loss": 0.159, + "step": 6345 + }, + { + "epoch": 1.4439135381114903, + "grad_norm": 1.6177628381389577, + "learning_rate": 1.010023077733511e-06, + "loss": 0.118, + "step": 6346 + }, + { + "epoch": 1.444141069397042, + "grad_norm": 2.01085232773475, + "learning_rate": 1.0099526902152886e-06, + "loss": 0.1013, + "step": 6347 + }, + { + "epoch": 1.4443686006825938, + "grad_norm": 1.8693082354619368, + "learning_rate": 1.0098822948293382e-06, + "loss": 0.1215, + "step": 6348 + }, + { + "epoch": 1.4445961319681455, + "grad_norm": 1.9616406323151174, + "learning_rate": 1.0098118915770985e-06, + "loss": 0.1025, + "step": 6349 + }, + { + "epoch": 1.4448236632536973, + "grad_norm": 1.74247396817358, + "learning_rate": 1.0097414804600087e-06, + "loss": 0.1232, + "step": 6350 + }, + { + "epoch": 1.445051194539249, + "grad_norm": 1.706851733043719, + "learning_rate": 1.0096710614795077e-06, + "loss": 0.0759, + "step": 6351 + }, + { + "epoch": 1.445278725824801, + "grad_norm": 2.4531940640602996, + "learning_rate": 1.0096006346370344e-06, + "loss": 0.1526, + "step": 6352 + }, + { + "epoch": 1.4455062571103527, + "grad_norm": 2.3321099387624473, + "learning_rate": 1.0095301999340285e-06, + "loss": 0.1587, + "step": 6353 + }, + { + "epoch": 1.4457337883959045, + "grad_norm": 1.5497257332095864, + "learning_rate": 1.0094597573719297e-06, + "loss": 0.1221, + "step": 6354 + }, + { + "epoch": 1.4459613196814562, + "grad_norm": 1.7921451438519347, + "learning_rate": 1.0093893069521777e-06, + "loss": 0.0806, + "step": 6355 + }, + { + "epoch": 1.446188850967008, + "grad_norm": 1.4730470898269366, + "learning_rate": 1.009318848676212e-06, + "loss": 0.0486, + "step": 6356 + }, + { + "epoch": 1.4464163822525598, + "grad_norm": 1.5931102826996517, + "learning_rate": 1.0092483825454729e-06, + "loss": 0.1791, + "step": 6357 + }, + { + "epoch": 1.4466439135381115, + "grad_norm": 1.882918083630657, + "learning_rate": 1.0091779085614006e-06, + "loss": 0.0707, + "step": 6358 + }, + { + "epoch": 1.4468714448236633, + "grad_norm": 1.5885687355433298, + "learning_rate": 1.0091074267254355e-06, + "loss": 0.0796, + "step": 6359 + }, + { + "epoch": 1.447098976109215, + "grad_norm": 2.2821278057106476, + "learning_rate": 1.009036937039018e-06, + "loss": 0.1604, + "step": 6360 + }, + { + "epoch": 1.4473265073947668, + "grad_norm": 1.4470125949831585, + "learning_rate": 1.008966439503589e-06, + "loss": 0.0735, + "step": 6361 + }, + { + "epoch": 1.4475540386803185, + "grad_norm": 1.4464111232881574, + "learning_rate": 1.008895934120589e-06, + "loss": 0.048, + "step": 6362 + }, + { + "epoch": 1.4477815699658703, + "grad_norm": 1.7415639336355635, + "learning_rate": 1.0088254208914593e-06, + "loss": 0.0542, + "step": 6363 + }, + { + "epoch": 1.448009101251422, + "grad_norm": 2.709852589489752, + "learning_rate": 1.0087548998176409e-06, + "loss": 0.0975, + "step": 6364 + }, + { + "epoch": 1.4482366325369738, + "grad_norm": 1.6286911486133107, + "learning_rate": 1.008684370900575e-06, + "loss": 0.141, + "step": 6365 + }, + { + "epoch": 1.4484641638225255, + "grad_norm": 2.054520820963295, + "learning_rate": 1.0086138341417035e-06, + "loss": 0.0991, + "step": 6366 + }, + { + "epoch": 1.4486916951080775, + "grad_norm": 1.511113359825543, + "learning_rate": 1.0085432895424674e-06, + "loss": 0.0743, + "step": 6367 + }, + { + "epoch": 1.4489192263936292, + "grad_norm": 1.7275913423025957, + "learning_rate": 1.0084727371043094e-06, + "loss": 0.0957, + "step": 6368 + }, + { + "epoch": 1.449146757679181, + "grad_norm": 2.4101408073521893, + "learning_rate": 1.0084021768286706e-06, + "loss": 0.0752, + "step": 6369 + }, + { + "epoch": 1.4493742889647327, + "grad_norm": 1.139924427258152, + "learning_rate": 1.0083316087169935e-06, + "loss": 0.1332, + "step": 6370 + }, + { + "epoch": 1.4496018202502845, + "grad_norm": 1.4397211555456666, + "learning_rate": 1.0082610327707204e-06, + "loss": 0.0654, + "step": 6371 + }, + { + "epoch": 1.4498293515358363, + "grad_norm": 1.7349452038054234, + "learning_rate": 1.0081904489912937e-06, + "loss": 0.1151, + "step": 6372 + }, + { + "epoch": 1.450056882821388, + "grad_norm": 2.177613333168079, + "learning_rate": 1.008119857380156e-06, + "loss": 0.1433, + "step": 6373 + }, + { + "epoch": 1.4502844141069398, + "grad_norm": 2.313688898466905, + "learning_rate": 1.0080492579387503e-06, + "loss": 0.1393, + "step": 6374 + }, + { + "epoch": 1.4505119453924915, + "grad_norm": 2.2444922001845287, + "learning_rate": 1.007978650668519e-06, + "loss": 0.0651, + "step": 6375 + }, + { + "epoch": 1.4507394766780433, + "grad_norm": 1.1098239666150351, + "learning_rate": 1.0079080355709057e-06, + "loss": 0.0886, + "step": 6376 + }, + { + "epoch": 1.450967007963595, + "grad_norm": 2.197433600973728, + "learning_rate": 1.007837412647353e-06, + "loss": 0.0504, + "step": 6377 + }, + { + "epoch": 1.4511945392491468, + "grad_norm": 1.9050357892618024, + "learning_rate": 1.007766781899305e-06, + "loss": 0.0663, + "step": 6378 + }, + { + "epoch": 1.4514220705346985, + "grad_norm": 2.020215521457231, + "learning_rate": 1.0076961433282051e-06, + "loss": 0.0671, + "step": 6379 + }, + { + "epoch": 1.4516496018202503, + "grad_norm": 1.6555451394956837, + "learning_rate": 1.0076254969354967e-06, + "loss": 0.052, + "step": 6380 + }, + { + "epoch": 1.451877133105802, + "grad_norm": 1.0205024219802799, + "learning_rate": 1.0075548427226241e-06, + "loss": 0.0818, + "step": 6381 + }, + { + "epoch": 1.4521046643913538, + "grad_norm": 1.110373279767029, + "learning_rate": 1.007484180691031e-06, + "loss": 0.0618, + "step": 6382 + }, + { + "epoch": 1.4523321956769055, + "grad_norm": 1.1903813445076674, + "learning_rate": 1.0074135108421616e-06, + "loss": 0.0493, + "step": 6383 + }, + { + "epoch": 1.4525597269624573, + "grad_norm": 1.4692733192683496, + "learning_rate": 1.0073428331774605e-06, + "loss": 0.0586, + "step": 6384 + }, + { + "epoch": 1.452787258248009, + "grad_norm": 1.7369951256720448, + "learning_rate": 1.007272147698372e-06, + "loss": 0.1068, + "step": 6385 + }, + { + "epoch": 1.4530147895335608, + "grad_norm": 3.0366934736471203, + "learning_rate": 1.0072014544063412e-06, + "loss": 0.1184, + "step": 6386 + }, + { + "epoch": 1.4532423208191125, + "grad_norm": 1.438235770345755, + "learning_rate": 1.0071307533028125e-06, + "loss": 0.1207, + "step": 6387 + }, + { + "epoch": 1.4534698521046643, + "grad_norm": 1.5692102048952972, + "learning_rate": 1.007060044389231e-06, + "loss": 0.1218, + "step": 6388 + }, + { + "epoch": 1.453697383390216, + "grad_norm": 1.3218250523299544, + "learning_rate": 1.006989327667042e-06, + "loss": 0.0731, + "step": 6389 + }, + { + "epoch": 1.4539249146757678, + "grad_norm": 2.189807535289241, + "learning_rate": 1.0069186031376906e-06, + "loss": 0.0947, + "step": 6390 + }, + { + "epoch": 1.4541524459613198, + "grad_norm": 1.9105068604592736, + "learning_rate": 1.0068478708026224e-06, + "loss": 0.1026, + "step": 6391 + }, + { + "epoch": 1.4543799772468715, + "grad_norm": 0.5964990943736891, + "learning_rate": 1.0067771306632832e-06, + "loss": 0.0295, + "step": 6392 + }, + { + "epoch": 1.4546075085324233, + "grad_norm": 1.8016037599952333, + "learning_rate": 1.0067063827211184e-06, + "loss": 0.0669, + "step": 6393 + }, + { + "epoch": 1.454835039817975, + "grad_norm": 1.125517173679015, + "learning_rate": 1.0066356269775744e-06, + "loss": 0.0479, + "step": 6394 + }, + { + "epoch": 1.4550625711035268, + "grad_norm": 1.4902876720368317, + "learning_rate": 1.0065648634340971e-06, + "loss": 0.0559, + "step": 6395 + }, + { + "epoch": 1.4552901023890785, + "grad_norm": 3.2050237634656478, + "learning_rate": 1.0064940920921328e-06, + "loss": 0.0683, + "step": 6396 + }, + { + "epoch": 1.4555176336746303, + "grad_norm": 1.5099892690669368, + "learning_rate": 1.006423312953128e-06, + "loss": 0.1018, + "step": 6397 + }, + { + "epoch": 1.455745164960182, + "grad_norm": 2.861502444994587, + "learning_rate": 1.0063525260185288e-06, + "loss": 0.1235, + "step": 6398 + }, + { + "epoch": 1.4559726962457338, + "grad_norm": 1.5797883956515744, + "learning_rate": 1.0062817312897826e-06, + "loss": 0.0837, + "step": 6399 + }, + { + "epoch": 1.4562002275312855, + "grad_norm": 2.0656400096946554, + "learning_rate": 1.0062109287683364e-06, + "loss": 0.1421, + "step": 6400 + }, + { + "epoch": 1.4564277588168373, + "grad_norm": 2.3153617321211897, + "learning_rate": 1.0061401184556366e-06, + "loss": 0.0708, + "step": 6401 + }, + { + "epoch": 1.456655290102389, + "grad_norm": 1.5512424090621548, + "learning_rate": 1.006069300353131e-06, + "loss": 0.0609, + "step": 6402 + }, + { + "epoch": 1.4568828213879408, + "grad_norm": 2.546713550355769, + "learning_rate": 1.005998474462267e-06, + "loss": 0.1593, + "step": 6403 + }, + { + "epoch": 1.4571103526734925, + "grad_norm": 1.691309762382103, + "learning_rate": 1.0059276407844915e-06, + "loss": 0.0659, + "step": 6404 + }, + { + "epoch": 1.4573378839590443, + "grad_norm": 1.7505170604113123, + "learning_rate": 1.0058567993212528e-06, + "loss": 0.099, + "step": 6405 + }, + { + "epoch": 1.4575654152445963, + "grad_norm": 1.1854966704428127, + "learning_rate": 1.005785950073999e-06, + "loss": 0.0727, + "step": 6406 + }, + { + "epoch": 1.457792946530148, + "grad_norm": 1.0488854883043695, + "learning_rate": 1.0057150930441772e-06, + "loss": 0.0772, + "step": 6407 + }, + { + "epoch": 1.4580204778156998, + "grad_norm": 1.513523792992345, + "learning_rate": 1.0056442282332365e-06, + "loss": 0.0548, + "step": 6408 + }, + { + "epoch": 1.4582480091012515, + "grad_norm": 1.1397568424774098, + "learning_rate": 1.0055733556426248e-06, + "loss": 0.0925, + "step": 6409 + }, + { + "epoch": 1.4584755403868033, + "grad_norm": 1.595106295663973, + "learning_rate": 1.0055024752737906e-06, + "loss": 0.1603, + "step": 6410 + }, + { + "epoch": 1.458703071672355, + "grad_norm": 2.1292594035873136, + "learning_rate": 1.0054315871281828e-06, + "loss": 0.073, + "step": 6411 + }, + { + "epoch": 1.4589306029579068, + "grad_norm": 2.1734889912318645, + "learning_rate": 1.00536069120725e-06, + "loss": 0.1034, + "step": 6412 + }, + { + "epoch": 1.4591581342434585, + "grad_norm": 2.4603522129310105, + "learning_rate": 1.0052897875124412e-06, + "loss": 0.0854, + "step": 6413 + }, + { + "epoch": 1.4593856655290103, + "grad_norm": 2.085924463649985, + "learning_rate": 1.005218876045206e-06, + "loss": 0.124, + "step": 6414 + }, + { + "epoch": 1.459613196814562, + "grad_norm": 2.2584471880345847, + "learning_rate": 1.0051479568069927e-06, + "loss": 0.0859, + "step": 6415 + }, + { + "epoch": 1.4598407281001138, + "grad_norm": 1.6411154403946306, + "learning_rate": 1.0050770297992517e-06, + "loss": 0.0827, + "step": 6416 + }, + { + "epoch": 1.4600682593856655, + "grad_norm": 0.9421738537038823, + "learning_rate": 1.0050060950234324e-06, + "loss": 0.0385, + "step": 6417 + }, + { + "epoch": 1.4602957906712173, + "grad_norm": 1.9520741759536568, + "learning_rate": 1.0049351524809842e-06, + "loss": 0.0855, + "step": 6418 + }, + { + "epoch": 1.460523321956769, + "grad_norm": 1.1983158344001326, + "learning_rate": 1.0048642021733576e-06, + "loss": 0.0365, + "step": 6419 + }, + { + "epoch": 1.4607508532423208, + "grad_norm": 2.441857013431579, + "learning_rate": 1.0047932441020022e-06, + "loss": 0.0925, + "step": 6420 + }, + { + "epoch": 1.4609783845278725, + "grad_norm": 1.5665363909548544, + "learning_rate": 1.0047222782683686e-06, + "loss": 0.0977, + "step": 6421 + }, + { + "epoch": 1.4612059158134243, + "grad_norm": 2.264590681075332, + "learning_rate": 1.0046513046739069e-06, + "loss": 0.148, + "step": 6422 + }, + { + "epoch": 1.461433447098976, + "grad_norm": 1.6526202515649944, + "learning_rate": 1.0045803233200679e-06, + "loss": 0.041, + "step": 6423 + }, + { + "epoch": 1.4616609783845278, + "grad_norm": 1.7626217119136827, + "learning_rate": 1.0045093342083022e-06, + "loss": 0.1285, + "step": 6424 + }, + { + "epoch": 1.4618885096700796, + "grad_norm": 2.2562587860636873, + "learning_rate": 1.0044383373400608e-06, + "loss": 0.0864, + "step": 6425 + }, + { + "epoch": 1.4621160409556313, + "grad_norm": 1.4449976996724352, + "learning_rate": 1.0043673327167946e-06, + "loss": 0.1135, + "step": 6426 + }, + { + "epoch": 1.462343572241183, + "grad_norm": 1.2114903089701574, + "learning_rate": 1.004296320339955e-06, + "loss": 0.0497, + "step": 6427 + }, + { + "epoch": 1.4625711035267348, + "grad_norm": 2.3353065793883965, + "learning_rate": 1.0042253002109933e-06, + "loss": 0.0995, + "step": 6428 + }, + { + "epoch": 1.4627986348122866, + "grad_norm": 1.213988405070411, + "learning_rate": 1.004154272331361e-06, + "loss": 0.0522, + "step": 6429 + }, + { + "epoch": 1.4630261660978385, + "grad_norm": 2.1531614083113406, + "learning_rate": 1.0040832367025097e-06, + "loss": 0.0997, + "step": 6430 + }, + { + "epoch": 1.4632536973833903, + "grad_norm": 2.413926356303523, + "learning_rate": 1.0040121933258912e-06, + "loss": 0.0909, + "step": 6431 + }, + { + "epoch": 1.463481228668942, + "grad_norm": 2.957734256860313, + "learning_rate": 1.0039411422029578e-06, + "loss": 0.1351, + "step": 6432 + }, + { + "epoch": 1.4637087599544938, + "grad_norm": 1.5243086926860814, + "learning_rate": 1.0038700833351612e-06, + "loss": 0.0928, + "step": 6433 + }, + { + "epoch": 1.4639362912400455, + "grad_norm": 1.2116195117085764, + "learning_rate": 1.0037990167239542e-06, + "loss": 0.0571, + "step": 6434 + }, + { + "epoch": 1.4641638225255973, + "grad_norm": 2.615635797890528, + "learning_rate": 1.0037279423707888e-06, + "loss": 0.0696, + "step": 6435 + }, + { + "epoch": 1.464391353811149, + "grad_norm": 1.7244140162748254, + "learning_rate": 1.0036568602771179e-06, + "loss": 0.189, + "step": 6436 + }, + { + "epoch": 1.4646188850967008, + "grad_norm": 2.6958961651306486, + "learning_rate": 1.0035857704443944e-06, + "loss": 0.1115, + "step": 6437 + }, + { + "epoch": 1.4648464163822525, + "grad_norm": 1.2190590501428018, + "learning_rate": 1.0035146728740712e-06, + "loss": 0.1424, + "step": 6438 + }, + { + "epoch": 1.4650739476678043, + "grad_norm": 1.5054931397105094, + "learning_rate": 1.0034435675676012e-06, + "loss": 0.0908, + "step": 6439 + }, + { + "epoch": 1.465301478953356, + "grad_norm": 1.6291323643009734, + "learning_rate": 1.0033724545264378e-06, + "loss": 0.089, + "step": 6440 + }, + { + "epoch": 1.4655290102389078, + "grad_norm": 1.341085989404762, + "learning_rate": 1.0033013337520342e-06, + "loss": 0.0486, + "step": 6441 + }, + { + "epoch": 1.4657565415244596, + "grad_norm": 1.011164152814645, + "learning_rate": 1.0032302052458443e-06, + "loss": 0.0499, + "step": 6442 + }, + { + "epoch": 1.4659840728100113, + "grad_norm": 1.0948910983582214, + "learning_rate": 1.0031590690093214e-06, + "loss": 0.0425, + "step": 6443 + }, + { + "epoch": 1.466211604095563, + "grad_norm": 1.7286333048940579, + "learning_rate": 1.0030879250439203e-06, + "loss": 0.0568, + "step": 6444 + }, + { + "epoch": 1.466439135381115, + "grad_norm": 0.9290542248967091, + "learning_rate": 1.003016773351094e-06, + "loss": 0.0223, + "step": 6445 + }, + { + "epoch": 1.4666666666666668, + "grad_norm": 2.5778455131993376, + "learning_rate": 1.0029456139322973e-06, + "loss": 0.0923, + "step": 6446 + }, + { + "epoch": 1.4668941979522185, + "grad_norm": 2.2511687310056416, + "learning_rate": 1.0028744467889845e-06, + "loss": 0.0967, + "step": 6447 + }, + { + "epoch": 1.4671217292377703, + "grad_norm": 2.725626168515591, + "learning_rate": 1.0028032719226097e-06, + "loss": 0.1895, + "step": 6448 + }, + { + "epoch": 1.467349260523322, + "grad_norm": 2.179394448530092, + "learning_rate": 1.0027320893346284e-06, + "loss": 0.1575, + "step": 6449 + }, + { + "epoch": 1.4675767918088738, + "grad_norm": 2.714923875843345, + "learning_rate": 1.0026608990264946e-06, + "loss": 0.1003, + "step": 6450 + }, + { + "epoch": 1.4678043230944255, + "grad_norm": 2.2884422798139217, + "learning_rate": 1.0025897009996639e-06, + "loss": 0.1281, + "step": 6451 + }, + { + "epoch": 1.4680318543799773, + "grad_norm": 2.13186315358863, + "learning_rate": 1.0025184952555911e-06, + "loss": 0.168, + "step": 6452 + }, + { + "epoch": 1.468259385665529, + "grad_norm": 2.088602911180568, + "learning_rate": 1.0024472817957318e-06, + "loss": 0.0728, + "step": 6453 + }, + { + "epoch": 1.4684869169510808, + "grad_norm": 2.0115528940586342, + "learning_rate": 1.0023760606215412e-06, + "loss": 0.1641, + "step": 6454 + }, + { + "epoch": 1.4687144482366326, + "grad_norm": 1.5082622000114985, + "learning_rate": 1.0023048317344752e-06, + "loss": 0.0494, + "step": 6455 + }, + { + "epoch": 1.4689419795221843, + "grad_norm": 2.0733909322591004, + "learning_rate": 1.0022335951359894e-06, + "loss": 0.1157, + "step": 6456 + }, + { + "epoch": 1.469169510807736, + "grad_norm": 1.854647627348043, + "learning_rate": 1.0021623508275397e-06, + "loss": 0.1409, + "step": 6457 + }, + { + "epoch": 1.4693970420932878, + "grad_norm": 2.3064994581234277, + "learning_rate": 1.0020910988105822e-06, + "loss": 0.0987, + "step": 6458 + }, + { + "epoch": 1.4696245733788396, + "grad_norm": 1.529342248309303, + "learning_rate": 1.0020198390865735e-06, + "loss": 0.0995, + "step": 6459 + }, + { + "epoch": 1.4698521046643913, + "grad_norm": 1.7759972116630178, + "learning_rate": 1.0019485716569696e-06, + "loss": 0.0773, + "step": 6460 + }, + { + "epoch": 1.470079635949943, + "grad_norm": 1.5298887299228214, + "learning_rate": 1.0018772965232272e-06, + "loss": 0.0803, + "step": 6461 + }, + { + "epoch": 1.4703071672354948, + "grad_norm": 1.7926988565702198, + "learning_rate": 1.0018060136868033e-06, + "loss": 0.1503, + "step": 6462 + }, + { + "epoch": 1.4705346985210466, + "grad_norm": 1.95957874155669, + "learning_rate": 1.0017347231491544e-06, + "loss": 0.0749, + "step": 6463 + }, + { + "epoch": 1.4707622298065983, + "grad_norm": 1.8565870188101044, + "learning_rate": 1.0016634249117378e-06, + "loss": 0.0903, + "step": 6464 + }, + { + "epoch": 1.47098976109215, + "grad_norm": 2.3785869049832487, + "learning_rate": 1.0015921189760105e-06, + "loss": 0.1752, + "step": 6465 + }, + { + "epoch": 1.4712172923777018, + "grad_norm": 1.08120840897355, + "learning_rate": 1.00152080534343e-06, + "loss": 0.0837, + "step": 6466 + }, + { + "epoch": 1.4714448236632536, + "grad_norm": 1.2600015453121518, + "learning_rate": 1.001449484015454e-06, + "loss": 0.0938, + "step": 6467 + }, + { + "epoch": 1.4716723549488053, + "grad_norm": 1.1675909506811741, + "learning_rate": 1.0013781549935396e-06, + "loss": 0.0802, + "step": 6468 + }, + { + "epoch": 1.4718998862343573, + "grad_norm": 2.5320154156973995, + "learning_rate": 1.0013068182791454e-06, + "loss": 0.0838, + "step": 6469 + }, + { + "epoch": 1.472127417519909, + "grad_norm": 1.2168783337222795, + "learning_rate": 1.0012354738737288e-06, + "loss": 0.1312, + "step": 6470 + }, + { + "epoch": 1.4723549488054608, + "grad_norm": 2.040097512167998, + "learning_rate": 1.0011641217787481e-06, + "loss": 0.1133, + "step": 6471 + }, + { + "epoch": 1.4725824800910126, + "grad_norm": 1.5937269991105336, + "learning_rate": 1.001092761995662e-06, + "loss": 0.1045, + "step": 6472 + }, + { + "epoch": 1.4728100113765643, + "grad_norm": 1.6603935998287558, + "learning_rate": 1.0010213945259282e-06, + "loss": 0.0628, + "step": 6473 + }, + { + "epoch": 1.473037542662116, + "grad_norm": 1.7229506631941387, + "learning_rate": 1.000950019371006e-06, + "loss": 0.0624, + "step": 6474 + }, + { + "epoch": 1.4732650739476678, + "grad_norm": 1.2010677843671187, + "learning_rate": 1.0008786365323539e-06, + "loss": 0.0706, + "step": 6475 + }, + { + "epoch": 1.4734926052332196, + "grad_norm": 1.8751575864533414, + "learning_rate": 1.0008072460114308e-06, + "loss": 0.0628, + "step": 6476 + }, + { + "epoch": 1.4737201365187713, + "grad_norm": 1.73050953802294, + "learning_rate": 1.0007358478096959e-06, + "loss": 0.0981, + "step": 6477 + }, + { + "epoch": 1.473947667804323, + "grad_norm": 1.5313330145032578, + "learning_rate": 1.0006644419286084e-06, + "loss": 0.0871, + "step": 6478 + }, + { + "epoch": 1.4741751990898748, + "grad_norm": 1.7211307536072435, + "learning_rate": 1.0005930283696277e-06, + "loss": 0.0804, + "step": 6479 + }, + { + "epoch": 1.4744027303754266, + "grad_norm": 1.041669030198631, + "learning_rate": 1.0005216071342133e-06, + "loss": 0.11, + "step": 6480 + }, + { + "epoch": 1.4746302616609783, + "grad_norm": 1.9063568947913965, + "learning_rate": 1.000450178223825e-06, + "loss": 0.1378, + "step": 6481 + }, + { + "epoch": 1.47485779294653, + "grad_norm": 1.4363869120455537, + "learning_rate": 1.0003787416399226e-06, + "loss": 0.1435, + "step": 6482 + }, + { + "epoch": 1.4750853242320818, + "grad_norm": 1.656907400406917, + "learning_rate": 1.0003072973839665e-06, + "loss": 0.067, + "step": 6483 + }, + { + "epoch": 1.4753128555176338, + "grad_norm": 1.9091492086095017, + "learning_rate": 1.0002358454574163e-06, + "loss": 0.0559, + "step": 6484 + }, + { + "epoch": 1.4755403868031856, + "grad_norm": 1.933307503414605, + "learning_rate": 1.0001643858617326e-06, + "loss": 0.199, + "step": 6485 + }, + { + "epoch": 1.4757679180887373, + "grad_norm": 1.524360431972506, + "learning_rate": 1.0000929185983762e-06, + "loss": 0.0514, + "step": 6486 + }, + { + "epoch": 1.475995449374289, + "grad_norm": 2.558602268977592, + "learning_rate": 1.0000214436688074e-06, + "loss": 0.1992, + "step": 6487 + }, + { + "epoch": 1.4762229806598408, + "grad_norm": 1.5057019130864449, + "learning_rate": 9.99949961074487e-07, + "loss": 0.1788, + "step": 6488 + }, + { + "epoch": 1.4764505119453926, + "grad_norm": 1.9974779073044902, + "learning_rate": 9.998784708168762e-07, + "loss": 0.1041, + "step": 6489 + }, + { + "epoch": 1.4766780432309443, + "grad_norm": 2.146999426549233, + "learning_rate": 9.998069728974357e-07, + "loss": 0.1536, + "step": 6490 + }, + { + "epoch": 1.476905574516496, + "grad_norm": 1.9204261446003483, + "learning_rate": 9.997354673176273e-07, + "loss": 0.12, + "step": 6491 + }, + { + "epoch": 1.4771331058020478, + "grad_norm": 1.7900450156383576, + "learning_rate": 9.996639540789124e-07, + "loss": 0.0709, + "step": 6492 + }, + { + "epoch": 1.4773606370875996, + "grad_norm": 2.706054567060273, + "learning_rate": 9.995924331827521e-07, + "loss": 0.1134, + "step": 6493 + }, + { + "epoch": 1.4775881683731513, + "grad_norm": 2.1430131676281143, + "learning_rate": 9.99520904630609e-07, + "loss": 0.1307, + "step": 6494 + }, + { + "epoch": 1.477815699658703, + "grad_norm": 1.3831116860997594, + "learning_rate": 9.994493684239443e-07, + "loss": 0.1063, + "step": 6495 + }, + { + "epoch": 1.4780432309442548, + "grad_norm": 0.8455695567282069, + "learning_rate": 9.993778245642202e-07, + "loss": 0.0408, + "step": 6496 + }, + { + "epoch": 1.4782707622298066, + "grad_norm": 2.0481093026394936, + "learning_rate": 9.99306273052899e-07, + "loss": 0.0507, + "step": 6497 + }, + { + "epoch": 1.4784982935153583, + "grad_norm": 2.701397993282172, + "learning_rate": 9.99234713891443e-07, + "loss": 0.0766, + "step": 6498 + }, + { + "epoch": 1.47872582480091, + "grad_norm": 2.060007223620782, + "learning_rate": 9.99163147081315e-07, + "loss": 0.0827, + "step": 6499 + }, + { + "epoch": 1.4789533560864618, + "grad_norm": 2.1178800053991895, + "learning_rate": 9.990915726239774e-07, + "loss": 0.0989, + "step": 6500 + }, + { + "epoch": 1.4791808873720136, + "grad_norm": 1.2820131971427713, + "learning_rate": 9.990199905208933e-07, + "loss": 0.1165, + "step": 6501 + }, + { + "epoch": 1.4794084186575653, + "grad_norm": 1.455912203082024, + "learning_rate": 9.989484007735256e-07, + "loss": 0.1711, + "step": 6502 + }, + { + "epoch": 1.479635949943117, + "grad_norm": 1.3882286358374694, + "learning_rate": 9.988768033833374e-07, + "loss": 0.1038, + "step": 6503 + }, + { + "epoch": 1.4798634812286688, + "grad_norm": 1.586658752493086, + "learning_rate": 9.98805198351792e-07, + "loss": 0.1503, + "step": 6504 + }, + { + "epoch": 1.4800910125142206, + "grad_norm": 1.216647170795259, + "learning_rate": 9.98733585680353e-07, + "loss": 0.0537, + "step": 6505 + }, + { + "epoch": 1.4803185437997723, + "grad_norm": 1.3218892161161693, + "learning_rate": 9.986619653704838e-07, + "loss": 0.0938, + "step": 6506 + }, + { + "epoch": 1.480546075085324, + "grad_norm": 2.0695991159537237, + "learning_rate": 9.985903374236487e-07, + "loss": 0.0963, + "step": 6507 + }, + { + "epoch": 1.480773606370876, + "grad_norm": 1.392307307265326, + "learning_rate": 9.985187018413108e-07, + "loss": 0.1009, + "step": 6508 + }, + { + "epoch": 1.4810011376564278, + "grad_norm": 1.322014214217586, + "learning_rate": 9.98447058624935e-07, + "loss": 0.052, + "step": 6509 + }, + { + "epoch": 1.4812286689419796, + "grad_norm": 1.5462590599779562, + "learning_rate": 9.983754077759852e-07, + "loss": 0.1301, + "step": 6510 + }, + { + "epoch": 1.4814562002275313, + "grad_norm": 1.767717707012614, + "learning_rate": 9.98303749295926e-07, + "loss": 0.137, + "step": 6511 + }, + { + "epoch": 1.481683731513083, + "grad_norm": 1.7186132236718707, + "learning_rate": 9.982320831862217e-07, + "loss": 0.0881, + "step": 6512 + }, + { + "epoch": 1.4819112627986348, + "grad_norm": 1.5236204963454782, + "learning_rate": 9.981604094483374e-07, + "loss": 0.1413, + "step": 6513 + }, + { + "epoch": 1.4821387940841866, + "grad_norm": 1.9188472431082706, + "learning_rate": 9.980887280837377e-07, + "loss": 0.2172, + "step": 6514 + }, + { + "epoch": 1.4823663253697383, + "grad_norm": 2.4374555651002483, + "learning_rate": 9.980170390938873e-07, + "loss": 0.1176, + "step": 6515 + }, + { + "epoch": 1.48259385665529, + "grad_norm": 2.1093125480487527, + "learning_rate": 9.979453424802522e-07, + "loss": 0.1106, + "step": 6516 + }, + { + "epoch": 1.4828213879408418, + "grad_norm": 1.915584556202085, + "learning_rate": 9.978736382442969e-07, + "loss": 0.1678, + "step": 6517 + }, + { + "epoch": 1.4830489192263936, + "grad_norm": 1.8641625340477204, + "learning_rate": 9.978019263874875e-07, + "loss": 0.0863, + "step": 6518 + }, + { + "epoch": 1.4832764505119453, + "grad_norm": 2.7441229909928615, + "learning_rate": 9.977302069112896e-07, + "loss": 0.1262, + "step": 6519 + }, + { + "epoch": 1.483503981797497, + "grad_norm": 1.5055670521242785, + "learning_rate": 9.97658479817169e-07, + "loss": 0.049, + "step": 6520 + }, + { + "epoch": 1.4837315130830488, + "grad_norm": 2.122217127690761, + "learning_rate": 9.975867451065913e-07, + "loss": 0.0481, + "step": 6521 + }, + { + "epoch": 1.4839590443686006, + "grad_norm": 2.4305735770475234, + "learning_rate": 9.97515002781023e-07, + "loss": 0.1932, + "step": 6522 + }, + { + "epoch": 1.4841865756541526, + "grad_norm": 1.826505366433851, + "learning_rate": 9.9744325284193e-07, + "loss": 0.1239, + "step": 6523 + }, + { + "epoch": 1.4844141069397043, + "grad_norm": 2.0963354537171797, + "learning_rate": 9.973714952907792e-07, + "loss": 0.1266, + "step": 6524 + }, + { + "epoch": 1.484641638225256, + "grad_norm": 2.3608423699852055, + "learning_rate": 9.97299730129037e-07, + "loss": 0.1219, + "step": 6525 + }, + { + "epoch": 1.4848691695108078, + "grad_norm": 2.1044457678017245, + "learning_rate": 9.972279573581705e-07, + "loss": 0.0771, + "step": 6526 + }, + { + "epoch": 1.4850967007963596, + "grad_norm": 1.7779259303890518, + "learning_rate": 9.97156176979646e-07, + "loss": 0.0842, + "step": 6527 + }, + { + "epoch": 1.4853242320819113, + "grad_norm": 1.6092438981897077, + "learning_rate": 9.970843889949305e-07, + "loss": 0.0797, + "step": 6528 + }, + { + "epoch": 1.485551763367463, + "grad_norm": 1.866981974786664, + "learning_rate": 9.97012593405492e-07, + "loss": 0.1233, + "step": 6529 + }, + { + "epoch": 1.4857792946530148, + "grad_norm": 2.8790996399351703, + "learning_rate": 9.969407902127972e-07, + "loss": 0.1261, + "step": 6530 + }, + { + "epoch": 1.4860068259385666, + "grad_norm": 1.9010114215145653, + "learning_rate": 9.968689794183137e-07, + "loss": 0.0963, + "step": 6531 + }, + { + "epoch": 1.4862343572241183, + "grad_norm": 1.4621814472728394, + "learning_rate": 9.967971610235095e-07, + "loss": 0.0952, + "step": 6532 + }, + { + "epoch": 1.48646188850967, + "grad_norm": 2.057756155886733, + "learning_rate": 9.967253350298522e-07, + "loss": 0.0779, + "step": 6533 + }, + { + "epoch": 1.4866894197952218, + "grad_norm": 2.397744809721191, + "learning_rate": 9.966535014388098e-07, + "loss": 0.0818, + "step": 6534 + }, + { + "epoch": 1.4869169510807736, + "grad_norm": 1.6045481156972197, + "learning_rate": 9.965816602518505e-07, + "loss": 0.066, + "step": 6535 + }, + { + "epoch": 1.4871444823663253, + "grad_norm": 1.3037900416564931, + "learning_rate": 9.965098114704425e-07, + "loss": 0.1013, + "step": 6536 + }, + { + "epoch": 1.487372013651877, + "grad_norm": 1.9894750167479516, + "learning_rate": 9.964379550960544e-07, + "loss": 0.086, + "step": 6537 + }, + { + "epoch": 1.4875995449374289, + "grad_norm": 1.3172380324912327, + "learning_rate": 9.96366091130155e-07, + "loss": 0.1158, + "step": 6538 + }, + { + "epoch": 1.4878270762229806, + "grad_norm": 1.4623068919383868, + "learning_rate": 9.962942195742125e-07, + "loss": 0.1483, + "step": 6539 + }, + { + "epoch": 1.4880546075085324, + "grad_norm": 2.1750563920153283, + "learning_rate": 9.96222340429696e-07, + "loss": 0.0826, + "step": 6540 + }, + { + "epoch": 1.488282138794084, + "grad_norm": 1.2779440444083219, + "learning_rate": 9.96150453698075e-07, + "loss": 0.0834, + "step": 6541 + }, + { + "epoch": 1.4885096700796359, + "grad_norm": 2.4346203961684267, + "learning_rate": 9.960785593808187e-07, + "loss": 0.0693, + "step": 6542 + }, + { + "epoch": 1.4887372013651876, + "grad_norm": 2.4068679198331844, + "learning_rate": 9.960066574793959e-07, + "loss": 0.1503, + "step": 6543 + }, + { + "epoch": 1.4889647326507394, + "grad_norm": 1.7234434466434458, + "learning_rate": 9.959347479952764e-07, + "loss": 0.0743, + "step": 6544 + }, + { + "epoch": 1.4891922639362911, + "grad_norm": 1.992864343711084, + "learning_rate": 9.958628309299303e-07, + "loss": 0.1348, + "step": 6545 + }, + { + "epoch": 1.4894197952218429, + "grad_norm": 1.6688641478519832, + "learning_rate": 9.95790906284827e-07, + "loss": 0.0975, + "step": 6546 + }, + { + "epoch": 1.4896473265073948, + "grad_norm": 1.461032294107854, + "learning_rate": 9.957189740614364e-07, + "loss": 0.08, + "step": 6547 + }, + { + "epoch": 1.4898748577929466, + "grad_norm": 1.6301476658725782, + "learning_rate": 9.956470342612292e-07, + "loss": 0.1803, + "step": 6548 + }, + { + "epoch": 1.4901023890784983, + "grad_norm": 1.8707264579028058, + "learning_rate": 9.955750868856753e-07, + "loss": 0.1212, + "step": 6549 + }, + { + "epoch": 1.49032992036405, + "grad_norm": 1.1641651322644342, + "learning_rate": 9.955031319362455e-07, + "loss": 0.092, + "step": 6550 + }, + { + "epoch": 1.4905574516496019, + "grad_norm": 1.0019288135962998, + "learning_rate": 9.954311694144101e-07, + "loss": 0.0959, + "step": 6551 + }, + { + "epoch": 1.4907849829351536, + "grad_norm": 2.052231836856024, + "learning_rate": 9.9535919932164e-07, + "loss": 0.0883, + "step": 6552 + }, + { + "epoch": 1.4910125142207054, + "grad_norm": 2.915207302536033, + "learning_rate": 9.952872216594062e-07, + "loss": 0.1058, + "step": 6553 + }, + { + "epoch": 1.491240045506257, + "grad_norm": 1.920148074541123, + "learning_rate": 9.952152364291795e-07, + "loss": 0.1474, + "step": 6554 + }, + { + "epoch": 1.4914675767918089, + "grad_norm": 1.285894577133101, + "learning_rate": 9.951432436324317e-07, + "loss": 0.2073, + "step": 6555 + }, + { + "epoch": 1.4916951080773606, + "grad_norm": 1.8218299718178723, + "learning_rate": 9.950712432706338e-07, + "loss": 0.0898, + "step": 6556 + }, + { + "epoch": 1.4919226393629124, + "grad_norm": 1.4323762701705023, + "learning_rate": 9.949992353452575e-07, + "loss": 0.0552, + "step": 6557 + }, + { + "epoch": 1.4921501706484641, + "grad_norm": 1.3545291611167616, + "learning_rate": 9.949272198577741e-07, + "loss": 0.1064, + "step": 6558 + }, + { + "epoch": 1.4923777019340159, + "grad_norm": 1.807647397449778, + "learning_rate": 9.948551968096562e-07, + "loss": 0.1034, + "step": 6559 + }, + { + "epoch": 1.4926052332195676, + "grad_norm": 1.4266173320216142, + "learning_rate": 9.947831662023751e-07, + "loss": 0.0439, + "step": 6560 + }, + { + "epoch": 1.4928327645051196, + "grad_norm": 1.9762698521372128, + "learning_rate": 9.947111280374036e-07, + "loss": 0.1725, + "step": 6561 + }, + { + "epoch": 1.4930602957906713, + "grad_norm": 1.851925457742424, + "learning_rate": 9.946390823162136e-07, + "loss": 0.2263, + "step": 6562 + }, + { + "epoch": 1.493287827076223, + "grad_norm": 1.5693473014259545, + "learning_rate": 9.945670290402778e-07, + "loss": 0.0571, + "step": 6563 + }, + { + "epoch": 1.4935153583617748, + "grad_norm": 2.2381860291552544, + "learning_rate": 9.944949682110689e-07, + "loss": 0.0704, + "step": 6564 + }, + { + "epoch": 1.4937428896473266, + "grad_norm": 2.894165644581643, + "learning_rate": 9.944228998300592e-07, + "loss": 0.1759, + "step": 6565 + }, + { + "epoch": 1.4939704209328784, + "grad_norm": 2.2783175420383595, + "learning_rate": 9.943508238987223e-07, + "loss": 0.1116, + "step": 6566 + }, + { + "epoch": 1.49419795221843, + "grad_norm": 2.3625175390423623, + "learning_rate": 9.942787404185307e-07, + "loss": 0.0811, + "step": 6567 + }, + { + "epoch": 1.4944254835039819, + "grad_norm": 1.3641102146894355, + "learning_rate": 9.942066493909582e-07, + "loss": 0.1175, + "step": 6568 + }, + { + "epoch": 1.4946530147895336, + "grad_norm": 1.962191132038736, + "learning_rate": 9.941345508174778e-07, + "loss": 0.1533, + "step": 6569 + }, + { + "epoch": 1.4948805460750854, + "grad_norm": 1.7938107718153122, + "learning_rate": 9.940624446995633e-07, + "loss": 0.0799, + "step": 6570 + }, + { + "epoch": 1.4951080773606371, + "grad_norm": 2.0893704746422386, + "learning_rate": 9.939903310386883e-07, + "loss": 0.1693, + "step": 6571 + }, + { + "epoch": 1.4953356086461889, + "grad_norm": 1.218836143766395, + "learning_rate": 9.939182098363267e-07, + "loss": 0.0514, + "step": 6572 + }, + { + "epoch": 1.4955631399317406, + "grad_norm": 2.225528007582617, + "learning_rate": 9.938460810939526e-07, + "loss": 0.0524, + "step": 6573 + }, + { + "epoch": 1.4957906712172924, + "grad_norm": 1.0703565767955194, + "learning_rate": 9.9377394481304e-07, + "loss": 0.0441, + "step": 6574 + }, + { + "epoch": 1.4960182025028441, + "grad_norm": 0.9831353955217996, + "learning_rate": 9.937018009950637e-07, + "loss": 0.1142, + "step": 6575 + }, + { + "epoch": 1.4962457337883959, + "grad_norm": 1.7656731882413514, + "learning_rate": 9.936296496414974e-07, + "loss": 0.1091, + "step": 6576 + }, + { + "epoch": 1.4964732650739476, + "grad_norm": 1.6503005620914835, + "learning_rate": 9.935574907538162e-07, + "loss": 0.0796, + "step": 6577 + }, + { + "epoch": 1.4967007963594994, + "grad_norm": 1.4721719805155578, + "learning_rate": 9.93485324333495e-07, + "loss": 0.1494, + "step": 6578 + }, + { + "epoch": 1.4969283276450511, + "grad_norm": 1.587186305176549, + "learning_rate": 9.934131503820086e-07, + "loss": 0.0994, + "step": 6579 + }, + { + "epoch": 1.4971558589306029, + "grad_norm": 2.077685939286771, + "learning_rate": 9.933409689008323e-07, + "loss": 0.1364, + "step": 6580 + }, + { + "epoch": 1.4973833902161546, + "grad_norm": 1.705800226622494, + "learning_rate": 9.932687798914408e-07, + "loss": 0.0697, + "step": 6581 + }, + { + "epoch": 1.4976109215017064, + "grad_norm": 0.8984678868158584, + "learning_rate": 9.9319658335531e-07, + "loss": 0.0444, + "step": 6582 + }, + { + "epoch": 1.4978384527872581, + "grad_norm": 1.340663582582159, + "learning_rate": 9.931243792939157e-07, + "loss": 0.0691, + "step": 6583 + }, + { + "epoch": 1.4980659840728099, + "grad_norm": 1.3634631218721727, + "learning_rate": 9.93052167708733e-07, + "loss": 0.0617, + "step": 6584 + }, + { + "epoch": 1.4982935153583616, + "grad_norm": 2.2638220421161956, + "learning_rate": 9.929799486012381e-07, + "loss": 0.1215, + "step": 6585 + }, + { + "epoch": 1.4985210466439136, + "grad_norm": 1.7511442172371627, + "learning_rate": 9.92907721972907e-07, + "loss": 0.0777, + "step": 6586 + }, + { + "epoch": 1.4987485779294654, + "grad_norm": 1.5992381545154368, + "learning_rate": 9.928354878252156e-07, + "loss": 0.1163, + "step": 6587 + }, + { + "epoch": 1.4989761092150171, + "grad_norm": 2.502483423033871, + "learning_rate": 9.927632461596409e-07, + "loss": 0.0938, + "step": 6588 + }, + { + "epoch": 1.4992036405005689, + "grad_norm": 2.7902929599302206, + "learning_rate": 9.926909969776588e-07, + "loss": 0.1, + "step": 6589 + }, + { + "epoch": 1.4994311717861206, + "grad_norm": 1.6546051994345607, + "learning_rate": 9.926187402807461e-07, + "loss": 0.0455, + "step": 6590 + }, + { + "epoch": 1.4996587030716724, + "grad_norm": 1.4386855095650237, + "learning_rate": 9.925464760703796e-07, + "loss": 0.1305, + "step": 6591 + }, + { + "epoch": 1.4998862343572241, + "grad_norm": 2.1550963420089855, + "learning_rate": 9.924742043480361e-07, + "loss": 0.1383, + "step": 6592 + }, + { + "epoch": 1.5001137656427759, + "grad_norm": 1.6050123025797802, + "learning_rate": 9.924019251151932e-07, + "loss": 0.0692, + "step": 6593 + }, + { + "epoch": 1.5003412969283276, + "grad_norm": 1.9173243071797523, + "learning_rate": 9.923296383733274e-07, + "loss": 0.1308, + "step": 6594 + }, + { + "epoch": 1.5005688282138794, + "grad_norm": 1.7215501934103334, + "learning_rate": 9.92257344123917e-07, + "loss": 0.0748, + "step": 6595 + }, + { + "epoch": 1.5007963594994311, + "grad_norm": 1.4572833583837446, + "learning_rate": 9.921850423684387e-07, + "loss": 0.1466, + "step": 6596 + }, + { + "epoch": 1.5010238907849829, + "grad_norm": 1.6185551035449188, + "learning_rate": 9.921127331083708e-07, + "loss": 0.0662, + "step": 6597 + }, + { + "epoch": 1.5012514220705349, + "grad_norm": 1.768377172897687, + "learning_rate": 9.92040416345191e-07, + "loss": 0.0475, + "step": 6598 + }, + { + "epoch": 1.5014789533560866, + "grad_norm": 1.3879959304267988, + "learning_rate": 9.91968092080377e-07, + "loss": 0.1134, + "step": 6599 + }, + { + "epoch": 1.5017064846416384, + "grad_norm": 3.2549540579687686, + "learning_rate": 9.918957603154076e-07, + "loss": 0.1008, + "step": 6600 + }, + { + "epoch": 1.5019340159271901, + "grad_norm": 2.547758885220799, + "learning_rate": 9.918234210517606e-07, + "loss": 0.066, + "step": 6601 + }, + { + "epoch": 1.5021615472127419, + "grad_norm": 1.542292781810862, + "learning_rate": 9.917510742909147e-07, + "loss": 0.1515, + "step": 6602 + }, + { + "epoch": 1.5023890784982936, + "grad_norm": 1.9701921831117433, + "learning_rate": 9.916787200343487e-07, + "loss": 0.0722, + "step": 6603 + }, + { + "epoch": 1.5026166097838454, + "grad_norm": 1.4978983128830725, + "learning_rate": 9.91606358283541e-07, + "loss": 0.0415, + "step": 6604 + }, + { + "epoch": 1.5028441410693971, + "grad_norm": 1.5436341118379642, + "learning_rate": 9.915339890399707e-07, + "loss": 0.1133, + "step": 6605 + }, + { + "epoch": 1.5030716723549489, + "grad_norm": 2.011492185812262, + "learning_rate": 9.914616123051172e-07, + "loss": 0.0556, + "step": 6606 + }, + { + "epoch": 1.5032992036405006, + "grad_norm": 2.038059208259661, + "learning_rate": 9.913892280804593e-07, + "loss": 0.0631, + "step": 6607 + }, + { + "epoch": 1.5035267349260524, + "grad_norm": 1.9867057182536725, + "learning_rate": 9.913168363674768e-07, + "loss": 0.1106, + "step": 6608 + }, + { + "epoch": 1.5037542662116041, + "grad_norm": 2.195537896793738, + "learning_rate": 9.912444371676489e-07, + "loss": 0.0736, + "step": 6609 + }, + { + "epoch": 1.5039817974971559, + "grad_norm": 1.523328848008561, + "learning_rate": 9.911720304824555e-07, + "loss": 0.0933, + "step": 6610 + }, + { + "epoch": 1.5042093287827076, + "grad_norm": 2.3802529025066637, + "learning_rate": 9.910996163133762e-07, + "loss": 0.0734, + "step": 6611 + }, + { + "epoch": 1.5044368600682594, + "grad_norm": 1.5938524758852861, + "learning_rate": 9.910271946618913e-07, + "loss": 0.0709, + "step": 6612 + }, + { + "epoch": 1.5046643913538111, + "grad_norm": 2.3867229570238675, + "learning_rate": 9.90954765529481e-07, + "loss": 0.1531, + "step": 6613 + }, + { + "epoch": 1.5048919226393629, + "grad_norm": 1.493728906940114, + "learning_rate": 9.908823289176255e-07, + "loss": 0.1798, + "step": 6614 + }, + { + "epoch": 1.5051194539249146, + "grad_norm": 1.5985296880996096, + "learning_rate": 9.90809884827805e-07, + "loss": 0.1794, + "step": 6615 + }, + { + "epoch": 1.5053469852104664, + "grad_norm": 1.570406258704178, + "learning_rate": 9.907374332615007e-07, + "loss": 0.0955, + "step": 6616 + }, + { + "epoch": 1.5055745164960181, + "grad_norm": 1.9969551241681303, + "learning_rate": 9.90664974220193e-07, + "loss": 0.0821, + "step": 6617 + }, + { + "epoch": 1.50580204778157, + "grad_norm": 1.7680267450626428, + "learning_rate": 9.90592507705363e-07, + "loss": 0.0656, + "step": 6618 + }, + { + "epoch": 1.5060295790671216, + "grad_norm": 1.4735623654469852, + "learning_rate": 9.905200337184915e-07, + "loss": 0.1243, + "step": 6619 + }, + { + "epoch": 1.5062571103526734, + "grad_norm": 2.3922194742332845, + "learning_rate": 9.904475522610602e-07, + "loss": 0.1074, + "step": 6620 + }, + { + "epoch": 1.5064846416382252, + "grad_norm": 1.4767746353973168, + "learning_rate": 9.9037506333455e-07, + "loss": 0.1425, + "step": 6621 + }, + { + "epoch": 1.506712172923777, + "grad_norm": 1.91538419707602, + "learning_rate": 9.90302566940443e-07, + "loss": 0.191, + "step": 6622 + }, + { + "epoch": 1.5069397042093287, + "grad_norm": 1.2549638165273682, + "learning_rate": 9.902300630802201e-07, + "loss": 0.0748, + "step": 6623 + }, + { + "epoch": 1.5071672354948804, + "grad_norm": 1.7074464037284505, + "learning_rate": 9.901575517553636e-07, + "loss": 0.0994, + "step": 6624 + }, + { + "epoch": 1.5073947667804322, + "grad_norm": 1.9037982472501616, + "learning_rate": 9.900850329673559e-07, + "loss": 0.0547, + "step": 6625 + }, + { + "epoch": 1.507622298065984, + "grad_norm": 2.0585060161102726, + "learning_rate": 9.900125067176782e-07, + "loss": 0.1257, + "step": 6626 + }, + { + "epoch": 1.5078498293515359, + "grad_norm": 0.8030202615841675, + "learning_rate": 9.899399730078138e-07, + "loss": 0.0853, + "step": 6627 + }, + { + "epoch": 1.5080773606370876, + "grad_norm": 1.6767887588871828, + "learning_rate": 9.898674318392446e-07, + "loss": 0.051, + "step": 6628 + }, + { + "epoch": 1.5083048919226394, + "grad_norm": 2.054186286357012, + "learning_rate": 9.897948832134532e-07, + "loss": 0.0668, + "step": 6629 + }, + { + "epoch": 1.5085324232081911, + "grad_norm": 1.5401946815877716, + "learning_rate": 9.897223271319227e-07, + "loss": 0.0701, + "step": 6630 + }, + { + "epoch": 1.508759954493743, + "grad_norm": 1.2266849927770183, + "learning_rate": 9.896497635961357e-07, + "loss": 0.064, + "step": 6631 + }, + { + "epoch": 1.5089874857792946, + "grad_norm": 1.893883359862433, + "learning_rate": 9.895771926075753e-07, + "loss": 0.1201, + "step": 6632 + }, + { + "epoch": 1.5092150170648464, + "grad_norm": 2.2124782416693165, + "learning_rate": 9.895046141677248e-07, + "loss": 0.1076, + "step": 6633 + }, + { + "epoch": 1.5094425483503981, + "grad_norm": 1.4193547479482487, + "learning_rate": 9.894320282780675e-07, + "loss": 0.0629, + "step": 6634 + }, + { + "epoch": 1.50967007963595, + "grad_norm": 1.4318969910886052, + "learning_rate": 9.89359434940087e-07, + "loss": 0.072, + "step": 6635 + }, + { + "epoch": 1.5098976109215017, + "grad_norm": 2.3903951577867364, + "learning_rate": 9.89286834155267e-07, + "loss": 0.1012, + "step": 6636 + }, + { + "epoch": 1.5101251422070536, + "grad_norm": 1.183723374024596, + "learning_rate": 9.89214225925091e-07, + "loss": 0.0379, + "step": 6637 + }, + { + "epoch": 1.5103526734926054, + "grad_norm": 1.5171515744161823, + "learning_rate": 9.891416102510436e-07, + "loss": 0.0696, + "step": 6638 + }, + { + "epoch": 1.5105802047781571, + "grad_norm": 1.7171746933495065, + "learning_rate": 9.890689871346084e-07, + "loss": 0.1177, + "step": 6639 + }, + { + "epoch": 1.5108077360637089, + "grad_norm": 1.7717611457403284, + "learning_rate": 9.8899635657727e-07, + "loss": 0.1159, + "step": 6640 + }, + { + "epoch": 1.5110352673492606, + "grad_norm": 2.677419290783526, + "learning_rate": 9.889237185805126e-07, + "loss": 0.0774, + "step": 6641 + }, + { + "epoch": 1.5112627986348124, + "grad_norm": 1.7854353991829401, + "learning_rate": 9.88851073145821e-07, + "loss": 0.1097, + "step": 6642 + }, + { + "epoch": 1.5114903299203641, + "grad_norm": 1.6833529414717199, + "learning_rate": 9.887784202746797e-07, + "loss": 0.1238, + "step": 6643 + }, + { + "epoch": 1.511717861205916, + "grad_norm": 1.532779466839382, + "learning_rate": 9.887057599685735e-07, + "loss": 0.1802, + "step": 6644 + }, + { + "epoch": 1.5119453924914676, + "grad_norm": 2.045124714315863, + "learning_rate": 9.88633092228988e-07, + "loss": 0.1381, + "step": 6645 + }, + { + "epoch": 1.5121729237770194, + "grad_norm": 1.2599289970561616, + "learning_rate": 9.885604170574081e-07, + "loss": 0.0993, + "step": 6646 + }, + { + "epoch": 1.5124004550625711, + "grad_norm": 1.7609732931837845, + "learning_rate": 9.884877344553189e-07, + "loss": 0.0797, + "step": 6647 + }, + { + "epoch": 1.512627986348123, + "grad_norm": 2.4222684044480105, + "learning_rate": 9.884150444242063e-07, + "loss": 0.1511, + "step": 6648 + }, + { + "epoch": 1.5128555176336747, + "grad_norm": 2.7535858833714357, + "learning_rate": 9.883423469655553e-07, + "loss": 0.124, + "step": 6649 + }, + { + "epoch": 1.5130830489192264, + "grad_norm": 2.1990285652154173, + "learning_rate": 9.882696420808526e-07, + "loss": 0.1464, + "step": 6650 + }, + { + "epoch": 1.5133105802047782, + "grad_norm": 1.3786062317925045, + "learning_rate": 9.881969297715836e-07, + "loss": 0.1032, + "step": 6651 + }, + { + "epoch": 1.51353811149033, + "grad_norm": 1.5534353780953807, + "learning_rate": 9.881242100392346e-07, + "loss": 0.0711, + "step": 6652 + }, + { + "epoch": 1.5137656427758817, + "grad_norm": 1.3604382033229832, + "learning_rate": 9.880514828852916e-07, + "loss": 0.0569, + "step": 6653 + }, + { + "epoch": 1.5139931740614334, + "grad_norm": 1.169340702023143, + "learning_rate": 9.879787483112413e-07, + "loss": 0.1239, + "step": 6654 + }, + { + "epoch": 1.5142207053469852, + "grad_norm": 2.2308897011011575, + "learning_rate": 9.879060063185702e-07, + "loss": 0.1562, + "step": 6655 + }, + { + "epoch": 1.514448236632537, + "grad_norm": 1.3587637173777, + "learning_rate": 9.878332569087647e-07, + "loss": 0.0329, + "step": 6656 + }, + { + "epoch": 1.5146757679180887, + "grad_norm": 2.545733827960348, + "learning_rate": 9.877605000833122e-07, + "loss": 0.1105, + "step": 6657 + }, + { + "epoch": 1.5149032992036404, + "grad_norm": 1.5953663605342474, + "learning_rate": 9.876877358436996e-07, + "loss": 0.057, + "step": 6658 + }, + { + "epoch": 1.5151308304891922, + "grad_norm": 1.1323787710966986, + "learning_rate": 9.876149641914135e-07, + "loss": 0.0752, + "step": 6659 + }, + { + "epoch": 1.515358361774744, + "grad_norm": 2.450063405723873, + "learning_rate": 9.875421851279419e-07, + "loss": 0.1737, + "step": 6660 + }, + { + "epoch": 1.5155858930602957, + "grad_norm": 1.3511044892138, + "learning_rate": 9.874693986547717e-07, + "loss": 0.0854, + "step": 6661 + }, + { + "epoch": 1.5158134243458474, + "grad_norm": 1.4537383245165247, + "learning_rate": 9.87396604773391e-07, + "loss": 0.0981, + "step": 6662 + }, + { + "epoch": 1.5160409556313992, + "grad_norm": 2.150257842110415, + "learning_rate": 9.873238034852875e-07, + "loss": 0.054, + "step": 6663 + }, + { + "epoch": 1.516268486916951, + "grad_norm": 1.6009149655756139, + "learning_rate": 9.872509947919489e-07, + "loss": 0.0934, + "step": 6664 + }, + { + "epoch": 1.5164960182025027, + "grad_norm": 2.69299983979343, + "learning_rate": 9.871781786948636e-07, + "loss": 0.0956, + "step": 6665 + }, + { + "epoch": 1.5167235494880547, + "grad_norm": 1.9231658149059578, + "learning_rate": 9.871053551955194e-07, + "loss": 0.0938, + "step": 6666 + }, + { + "epoch": 1.5169510807736064, + "grad_norm": 1.5996264354941863, + "learning_rate": 9.87032524295405e-07, + "loss": 0.0851, + "step": 6667 + }, + { + "epoch": 1.5171786120591582, + "grad_norm": 2.3602848121961135, + "learning_rate": 9.869596859960087e-07, + "loss": 0.0878, + "step": 6668 + }, + { + "epoch": 1.51740614334471, + "grad_norm": 2.49783971066406, + "learning_rate": 9.868868402988194e-07, + "loss": 0.1241, + "step": 6669 + }, + { + "epoch": 1.5176336746302617, + "grad_norm": 1.9002930155631617, + "learning_rate": 9.86813987205326e-07, + "loss": 0.1825, + "step": 6670 + }, + { + "epoch": 1.5178612059158134, + "grad_norm": 2.175376286906289, + "learning_rate": 9.867411267170171e-07, + "loss": 0.237, + "step": 6671 + }, + { + "epoch": 1.5180887372013652, + "grad_norm": 2.136172560112039, + "learning_rate": 9.866682588353823e-07, + "loss": 0.1025, + "step": 6672 + }, + { + "epoch": 1.518316268486917, + "grad_norm": 2.2409744987674522, + "learning_rate": 9.865953835619105e-07, + "loss": 0.1267, + "step": 6673 + }, + { + "epoch": 1.5185437997724687, + "grad_norm": 1.1677610227530477, + "learning_rate": 9.865225008980913e-07, + "loss": 0.077, + "step": 6674 + }, + { + "epoch": 1.5187713310580204, + "grad_norm": 1.6686212093967283, + "learning_rate": 9.864496108454142e-07, + "loss": 0.0827, + "step": 6675 + }, + { + "epoch": 1.5189988623435724, + "grad_norm": 1.4606352032280105, + "learning_rate": 9.863767134053691e-07, + "loss": 0.0657, + "step": 6676 + }, + { + "epoch": 1.5192263936291241, + "grad_norm": 1.6472537173662813, + "learning_rate": 9.86303808579446e-07, + "loss": 0.1342, + "step": 6677 + }, + { + "epoch": 1.519453924914676, + "grad_norm": 1.7887052016764249, + "learning_rate": 9.862308963691344e-07, + "loss": 0.1469, + "step": 6678 + }, + { + "epoch": 1.5196814562002277, + "grad_norm": 2.0564231913518194, + "learning_rate": 9.86157976775925e-07, + "loss": 0.2687, + "step": 6679 + }, + { + "epoch": 1.5199089874857794, + "grad_norm": 2.173256072420634, + "learning_rate": 9.86085049801308e-07, + "loss": 0.1331, + "step": 6680 + }, + { + "epoch": 1.5201365187713312, + "grad_norm": 1.828606013658809, + "learning_rate": 9.860121154467738e-07, + "loss": 0.082, + "step": 6681 + }, + { + "epoch": 1.520364050056883, + "grad_norm": 1.433527391061732, + "learning_rate": 9.859391737138132e-07, + "loss": 0.1203, + "step": 6682 + }, + { + "epoch": 1.5205915813424347, + "grad_norm": 2.462623028096727, + "learning_rate": 9.85866224603917e-07, + "loss": 0.1017, + "step": 6683 + }, + { + "epoch": 1.5208191126279864, + "grad_norm": 1.5185381292209394, + "learning_rate": 9.85793268118576e-07, + "loss": 0.1266, + "step": 6684 + }, + { + "epoch": 1.5210466439135382, + "grad_norm": 1.68148981523328, + "learning_rate": 9.857203042592813e-07, + "loss": 0.0575, + "step": 6685 + }, + { + "epoch": 1.52127417519909, + "grad_norm": 2.3064553659517126, + "learning_rate": 9.856473330275243e-07, + "loss": 0.0862, + "step": 6686 + }, + { + "epoch": 1.5215017064846417, + "grad_norm": 1.9203644918552756, + "learning_rate": 9.855743544247962e-07, + "loss": 0.1983, + "step": 6687 + }, + { + "epoch": 1.5217292377701934, + "grad_norm": 1.828129246833364, + "learning_rate": 9.855013684525888e-07, + "loss": 0.0802, + "step": 6688 + }, + { + "epoch": 1.5219567690557452, + "grad_norm": 2.5899944461839497, + "learning_rate": 9.854283751123935e-07, + "loss": 0.0769, + "step": 6689 + }, + { + "epoch": 1.522184300341297, + "grad_norm": 1.1795018537600235, + "learning_rate": 9.853553744057023e-07, + "loss": 0.0245, + "step": 6690 + }, + { + "epoch": 1.5224118316268487, + "grad_norm": 1.8602779860567926, + "learning_rate": 9.852823663340074e-07, + "loss": 0.0683, + "step": 6691 + }, + { + "epoch": 1.5226393629124004, + "grad_norm": 2.468331705268543, + "learning_rate": 9.852093508988006e-07, + "loss": 0.1311, + "step": 6692 + }, + { + "epoch": 1.5228668941979522, + "grad_norm": 3.072633064169901, + "learning_rate": 9.851363281015745e-07, + "loss": 0.1717, + "step": 6693 + }, + { + "epoch": 1.523094425483504, + "grad_norm": 1.7646833258471915, + "learning_rate": 9.850632979438211e-07, + "loss": 0.0604, + "step": 6694 + }, + { + "epoch": 1.5233219567690557, + "grad_norm": 1.5840900489872247, + "learning_rate": 9.849902604270335e-07, + "loss": 0.1526, + "step": 6695 + }, + { + "epoch": 1.5235494880546074, + "grad_norm": 1.9100983840195138, + "learning_rate": 9.849172155527044e-07, + "loss": 0.0927, + "step": 6696 + }, + { + "epoch": 1.5237770193401592, + "grad_norm": 1.7968210113673053, + "learning_rate": 9.848441633223266e-07, + "loss": 0.1103, + "step": 6697 + }, + { + "epoch": 1.524004550625711, + "grad_norm": 1.7268833657807332, + "learning_rate": 9.847711037373928e-07, + "loss": 0.1204, + "step": 6698 + }, + { + "epoch": 1.5242320819112627, + "grad_norm": 1.1093468674675278, + "learning_rate": 9.846980367993968e-07, + "loss": 0.1065, + "step": 6699 + }, + { + "epoch": 1.5244596131968144, + "grad_norm": 0.8658636255558, + "learning_rate": 9.846249625098317e-07, + "loss": 0.0542, + "step": 6700 + }, + { + "epoch": 1.5246871444823662, + "grad_norm": 1.333652555504289, + "learning_rate": 9.845518808701906e-07, + "loss": 0.1258, + "step": 6701 + }, + { + "epoch": 1.524914675767918, + "grad_norm": 1.910866305841753, + "learning_rate": 9.84478791881968e-07, + "loss": 0.1404, + "step": 6702 + }, + { + "epoch": 1.5251422070534697, + "grad_norm": 1.6361267289033996, + "learning_rate": 9.844056955466571e-07, + "loss": 0.0944, + "step": 6703 + }, + { + "epoch": 1.5253697383390215, + "grad_norm": 1.5309137515954483, + "learning_rate": 9.84332591865752e-07, + "loss": 0.0522, + "step": 6704 + }, + { + "epoch": 1.5255972696245734, + "grad_norm": 3.572339727844091, + "learning_rate": 9.842594808407467e-07, + "loss": 0.1908, + "step": 6705 + }, + { + "epoch": 1.5258248009101252, + "grad_norm": 3.4734172827801526, + "learning_rate": 9.841863624731358e-07, + "loss": 0.1023, + "step": 6706 + }, + { + "epoch": 1.526052332195677, + "grad_norm": 1.7120902324796543, + "learning_rate": 9.841132367644133e-07, + "loss": 0.0958, + "step": 6707 + }, + { + "epoch": 1.5262798634812287, + "grad_norm": 1.95257483276106, + "learning_rate": 9.840401037160737e-07, + "loss": 0.087, + "step": 6708 + }, + { + "epoch": 1.5265073947667804, + "grad_norm": 1.839958674531052, + "learning_rate": 9.839669633296122e-07, + "loss": 0.1274, + "step": 6709 + }, + { + "epoch": 1.5267349260523322, + "grad_norm": 1.1867164810093782, + "learning_rate": 9.838938156065236e-07, + "loss": 0.0737, + "step": 6710 + }, + { + "epoch": 1.526962457337884, + "grad_norm": 1.6245966674135188, + "learning_rate": 9.838206605483024e-07, + "loss": 0.0657, + "step": 6711 + }, + { + "epoch": 1.5271899886234357, + "grad_norm": 1.9555111968047711, + "learning_rate": 9.83747498156444e-07, + "loss": 0.0641, + "step": 6712 + }, + { + "epoch": 1.5274175199089874, + "grad_norm": 2.2335184965178723, + "learning_rate": 9.836743284324438e-07, + "loss": 0.1571, + "step": 6713 + }, + { + "epoch": 1.5276450511945392, + "grad_norm": 1.2440844987147237, + "learning_rate": 9.836011513777975e-07, + "loss": 0.0987, + "step": 6714 + }, + { + "epoch": 1.5278725824800912, + "grad_norm": 1.774129851231337, + "learning_rate": 9.835279669940002e-07, + "loss": 0.1022, + "step": 6715 + }, + { + "epoch": 1.528100113765643, + "grad_norm": 3.0157041807810834, + "learning_rate": 9.834547752825477e-07, + "loss": 0.1205, + "step": 6716 + }, + { + "epoch": 1.5283276450511947, + "grad_norm": 1.426750353390932, + "learning_rate": 9.833815762449364e-07, + "loss": 0.0717, + "step": 6717 + }, + { + "epoch": 1.5285551763367464, + "grad_norm": 2.4046750983405674, + "learning_rate": 9.833083698826618e-07, + "loss": 0.1465, + "step": 6718 + }, + { + "epoch": 1.5287827076222982, + "grad_norm": 1.0940696422129481, + "learning_rate": 9.832351561972204e-07, + "loss": 0.0456, + "step": 6719 + }, + { + "epoch": 1.52901023890785, + "grad_norm": 1.62018463307765, + "learning_rate": 9.831619351901083e-07, + "loss": 0.0318, + "step": 6720 + }, + { + "epoch": 1.5292377701934017, + "grad_norm": 1.7579436507766677, + "learning_rate": 9.830887068628223e-07, + "loss": 0.074, + "step": 6721 + }, + { + "epoch": 1.5294653014789534, + "grad_norm": 1.3534465665094793, + "learning_rate": 9.830154712168591e-07, + "loss": 0.0873, + "step": 6722 + }, + { + "epoch": 1.5296928327645052, + "grad_norm": 2.061404616403052, + "learning_rate": 9.829422282537152e-07, + "loss": 0.072, + "step": 6723 + }, + { + "epoch": 1.529920364050057, + "grad_norm": 1.5775828869406339, + "learning_rate": 9.828689779748877e-07, + "loss": 0.0909, + "step": 6724 + }, + { + "epoch": 1.5301478953356087, + "grad_norm": 2.3455959707506664, + "learning_rate": 9.827957203818737e-07, + "loss": 0.0916, + "step": 6725 + }, + { + "epoch": 1.5303754266211604, + "grad_norm": 2.5840806262089835, + "learning_rate": 9.827224554761705e-07, + "loss": 0.1354, + "step": 6726 + }, + { + "epoch": 1.5306029579067122, + "grad_norm": 1.0556116471214028, + "learning_rate": 9.826491832592754e-07, + "loss": 0.1295, + "step": 6727 + }, + { + "epoch": 1.530830489192264, + "grad_norm": 2.196821539587016, + "learning_rate": 9.825759037326861e-07, + "loss": 0.1352, + "step": 6728 + }, + { + "epoch": 1.5310580204778157, + "grad_norm": 1.7033480648148973, + "learning_rate": 9.825026168979001e-07, + "loss": 0.1256, + "step": 6729 + }, + { + "epoch": 1.5312855517633674, + "grad_norm": 1.314871959729546, + "learning_rate": 9.824293227564154e-07, + "loss": 0.1147, + "step": 6730 + }, + { + "epoch": 1.5315130830489192, + "grad_norm": 2.295215182711722, + "learning_rate": 9.8235602130973e-07, + "loss": 0.087, + "step": 6731 + }, + { + "epoch": 1.531740614334471, + "grad_norm": 1.7613060587120912, + "learning_rate": 9.822827125593417e-07, + "loss": 0.0816, + "step": 6732 + }, + { + "epoch": 1.5319681456200227, + "grad_norm": 3.6189711422913757, + "learning_rate": 9.822093965067492e-07, + "loss": 0.217, + "step": 6733 + }, + { + "epoch": 1.5321956769055745, + "grad_norm": 1.5407713472072504, + "learning_rate": 9.821360731534512e-07, + "loss": 0.049, + "step": 6734 + }, + { + "epoch": 1.5324232081911262, + "grad_norm": 1.6479162072423494, + "learning_rate": 9.820627425009455e-07, + "loss": 0.0688, + "step": 6735 + }, + { + "epoch": 1.532650739476678, + "grad_norm": 0.7948885817082926, + "learning_rate": 9.819894045507315e-07, + "loss": 0.0599, + "step": 6736 + }, + { + "epoch": 1.5328782707622297, + "grad_norm": 1.1151560312631288, + "learning_rate": 9.819160593043076e-07, + "loss": 0.1409, + "step": 6737 + }, + { + "epoch": 1.5331058020477815, + "grad_norm": 1.923769297389666, + "learning_rate": 9.818427067631733e-07, + "loss": 0.0986, + "step": 6738 + }, + { + "epoch": 1.5333333333333332, + "grad_norm": 1.2519127952741742, + "learning_rate": 9.817693469288276e-07, + "loss": 0.1078, + "step": 6739 + }, + { + "epoch": 1.533560864618885, + "grad_norm": 1.046024360494148, + "learning_rate": 9.8169597980277e-07, + "loss": 0.0381, + "step": 6740 + }, + { + "epoch": 1.5337883959044367, + "grad_norm": 1.1028938926163987, + "learning_rate": 9.816226053864996e-07, + "loss": 0.1161, + "step": 6741 + }, + { + "epoch": 1.5340159271899885, + "grad_norm": 1.7266193972711186, + "learning_rate": 9.815492236815163e-07, + "loss": 0.0788, + "step": 6742 + }, + { + "epoch": 1.5342434584755402, + "grad_norm": 1.6081871572113473, + "learning_rate": 9.8147583468932e-07, + "loss": 0.1339, + "step": 6743 + }, + { + "epoch": 1.5344709897610922, + "grad_norm": 1.402571830332417, + "learning_rate": 9.814024384114102e-07, + "loss": 0.0378, + "step": 6744 + }, + { + "epoch": 1.534698521046644, + "grad_norm": 1.846203157196796, + "learning_rate": 9.813290348492874e-07, + "loss": 0.0693, + "step": 6745 + }, + { + "epoch": 1.5349260523321957, + "grad_norm": 1.6339574311431553, + "learning_rate": 9.812556240044518e-07, + "loss": 0.0803, + "step": 6746 + }, + { + "epoch": 1.5351535836177475, + "grad_norm": 1.3328670448791076, + "learning_rate": 9.811822058784038e-07, + "loss": 0.1181, + "step": 6747 + }, + { + "epoch": 1.5353811149032992, + "grad_norm": 1.4665000621481399, + "learning_rate": 9.811087804726436e-07, + "loss": 0.1358, + "step": 6748 + }, + { + "epoch": 1.535608646188851, + "grad_norm": 1.7442320100495479, + "learning_rate": 9.810353477886722e-07, + "loss": 0.0473, + "step": 6749 + }, + { + "epoch": 1.5358361774744027, + "grad_norm": 1.8155220142705994, + "learning_rate": 9.809619078279904e-07, + "loss": 0.0756, + "step": 6750 + }, + { + "epoch": 1.5360637087599545, + "grad_norm": 1.2458858709598846, + "learning_rate": 9.80888460592099e-07, + "loss": 0.1036, + "step": 6751 + }, + { + "epoch": 1.5362912400455062, + "grad_norm": 1.4447722998575065, + "learning_rate": 9.808150060824995e-07, + "loss": 0.1606, + "step": 6752 + }, + { + "epoch": 1.5365187713310582, + "grad_norm": 1.5230317243883573, + "learning_rate": 9.807415443006926e-07, + "loss": 0.092, + "step": 6753 + }, + { + "epoch": 1.53674630261661, + "grad_norm": 1.6478604856615626, + "learning_rate": 9.806680752481803e-07, + "loss": 0.0621, + "step": 6754 + }, + { + "epoch": 1.5369738339021617, + "grad_norm": 2.437908355394251, + "learning_rate": 9.805945989264638e-07, + "loss": 0.081, + "step": 6755 + }, + { + "epoch": 1.5372013651877134, + "grad_norm": 1.266952056682482, + "learning_rate": 9.805211153370448e-07, + "loss": 0.0891, + "step": 6756 + }, + { + "epoch": 1.5374288964732652, + "grad_norm": 2.1205351302476227, + "learning_rate": 9.804476244814253e-07, + "loss": 0.0786, + "step": 6757 + }, + { + "epoch": 1.537656427758817, + "grad_norm": 1.3261225574966184, + "learning_rate": 9.803741263611076e-07, + "loss": 0.1225, + "step": 6758 + }, + { + "epoch": 1.5378839590443687, + "grad_norm": 2.0857772110578683, + "learning_rate": 9.803006209775936e-07, + "loss": 0.1329, + "step": 6759 + }, + { + "epoch": 1.5381114903299204, + "grad_norm": 1.6409653695306137, + "learning_rate": 9.802271083323855e-07, + "loss": 0.1046, + "step": 6760 + }, + { + "epoch": 1.5383390216154722, + "grad_norm": 1.9140415782586275, + "learning_rate": 9.801535884269859e-07, + "loss": 0.0583, + "step": 6761 + }, + { + "epoch": 1.538566552901024, + "grad_norm": 1.5742717342382722, + "learning_rate": 9.800800612628971e-07, + "loss": 0.1474, + "step": 6762 + }, + { + "epoch": 1.5387940841865757, + "grad_norm": 2.587740921277969, + "learning_rate": 9.800065268416226e-07, + "loss": 0.1413, + "step": 6763 + }, + { + "epoch": 1.5390216154721275, + "grad_norm": 1.9113702724685222, + "learning_rate": 9.799329851646643e-07, + "loss": 0.0734, + "step": 6764 + }, + { + "epoch": 1.5392491467576792, + "grad_norm": 1.9971326212884126, + "learning_rate": 9.798594362335265e-07, + "loss": 0.0665, + "step": 6765 + }, + { + "epoch": 1.539476678043231, + "grad_norm": 1.9851093003136897, + "learning_rate": 9.797858800497112e-07, + "loss": 0.1075, + "step": 6766 + }, + { + "epoch": 1.5397042093287827, + "grad_norm": 2.0738730426091574, + "learning_rate": 9.797123166147224e-07, + "loss": 0.1353, + "step": 6767 + }, + { + "epoch": 1.5399317406143345, + "grad_norm": 2.7095905938056988, + "learning_rate": 9.796387459300635e-07, + "loss": 0.0688, + "step": 6768 + }, + { + "epoch": 1.5401592718998862, + "grad_norm": 1.239920336046195, + "learning_rate": 9.795651679972382e-07, + "loss": 0.1014, + "step": 6769 + }, + { + "epoch": 1.540386803185438, + "grad_norm": 1.607016101863712, + "learning_rate": 9.7949158281775e-07, + "loss": 0.0505, + "step": 6770 + }, + { + "epoch": 1.5406143344709897, + "grad_norm": 1.386117850600802, + "learning_rate": 9.794179903931035e-07, + "loss": 0.0687, + "step": 6771 + }, + { + "epoch": 1.5408418657565415, + "grad_norm": 1.236749985995557, + "learning_rate": 9.79344390724802e-07, + "loss": 0.0327, + "step": 6772 + }, + { + "epoch": 1.5410693970420932, + "grad_norm": 1.6797050069106503, + "learning_rate": 9.7927078381435e-07, + "loss": 0.0795, + "step": 6773 + }, + { + "epoch": 1.541296928327645, + "grad_norm": 2.9730659229323093, + "learning_rate": 9.791971696632523e-07, + "loss": 0.0906, + "step": 6774 + }, + { + "epoch": 1.5415244596131967, + "grad_norm": 1.8652340188388545, + "learning_rate": 9.79123548273013e-07, + "loss": 0.1489, + "step": 6775 + }, + { + "epoch": 1.5417519908987485, + "grad_norm": 1.744502185028443, + "learning_rate": 9.79049919645137e-07, + "loss": 0.0638, + "step": 6776 + }, + { + "epoch": 1.5419795221843002, + "grad_norm": 2.096956448280094, + "learning_rate": 9.78976283781129e-07, + "loss": 0.0925, + "step": 6777 + }, + { + "epoch": 1.542207053469852, + "grad_norm": 1.9057593286282604, + "learning_rate": 9.78902640682494e-07, + "loss": 0.0763, + "step": 6778 + }, + { + "epoch": 1.5424345847554037, + "grad_norm": 1.634924562850608, + "learning_rate": 9.788289903507373e-07, + "loss": 0.1582, + "step": 6779 + }, + { + "epoch": 1.5426621160409555, + "grad_norm": 2.052440866489605, + "learning_rate": 9.787553327873637e-07, + "loss": 0.0615, + "step": 6780 + }, + { + "epoch": 1.5428896473265072, + "grad_norm": 1.5557781540863105, + "learning_rate": 9.786816679938794e-07, + "loss": 0.1339, + "step": 6781 + }, + { + "epoch": 1.543117178612059, + "grad_norm": 1.6106987587996306, + "learning_rate": 9.786079959717892e-07, + "loss": 0.0947, + "step": 6782 + }, + { + "epoch": 1.543344709897611, + "grad_norm": 1.5775713242046356, + "learning_rate": 9.785343167225994e-07, + "loss": 0.0633, + "step": 6783 + }, + { + "epoch": 1.5435722411831627, + "grad_norm": 1.6760823520678827, + "learning_rate": 9.784606302478155e-07, + "loss": 0.0872, + "step": 6784 + }, + { + "epoch": 1.5437997724687145, + "grad_norm": 2.696297293597503, + "learning_rate": 9.783869365489437e-07, + "loss": 0.0909, + "step": 6785 + }, + { + "epoch": 1.5440273037542662, + "grad_norm": 1.1664906375290016, + "learning_rate": 9.783132356274901e-07, + "loss": 0.042, + "step": 6786 + }, + { + "epoch": 1.544254835039818, + "grad_norm": 1.9040250600939805, + "learning_rate": 9.78239527484961e-07, + "loss": 0.1065, + "step": 6787 + }, + { + "epoch": 1.5444823663253697, + "grad_norm": 1.198846442121248, + "learning_rate": 9.781658121228628e-07, + "loss": 0.044, + "step": 6788 + }, + { + "epoch": 1.5447098976109215, + "grad_norm": 1.245850403558456, + "learning_rate": 9.780920895427025e-07, + "loss": 0.0615, + "step": 6789 + }, + { + "epoch": 1.5449374288964732, + "grad_norm": 1.8321354791933504, + "learning_rate": 9.780183597459864e-07, + "loss": 0.0969, + "step": 6790 + }, + { + "epoch": 1.545164960182025, + "grad_norm": 2.222629959611143, + "learning_rate": 9.779446227342216e-07, + "loss": 0.2203, + "step": 6791 + }, + { + "epoch": 1.545392491467577, + "grad_norm": 2.458814173048237, + "learning_rate": 9.77870878508915e-07, + "loss": 0.1028, + "step": 6792 + }, + { + "epoch": 1.5456200227531287, + "grad_norm": 1.9091345644157072, + "learning_rate": 9.77797127071574e-07, + "loss": 0.0589, + "step": 6793 + }, + { + "epoch": 1.5458475540386805, + "grad_norm": 1.789562373939949, + "learning_rate": 9.777233684237056e-07, + "loss": 0.0805, + "step": 6794 + }, + { + "epoch": 1.5460750853242322, + "grad_norm": 1.7950526990429894, + "learning_rate": 9.776496025668174e-07, + "loss": 0.049, + "step": 6795 + }, + { + "epoch": 1.546302616609784, + "grad_norm": 1.2634480571691555, + "learning_rate": 9.775758295024177e-07, + "loss": 0.0752, + "step": 6796 + }, + { + "epoch": 1.5465301478953357, + "grad_norm": 1.7316859014052663, + "learning_rate": 9.775020492320136e-07, + "loss": 0.0714, + "step": 6797 + }, + { + "epoch": 1.5467576791808875, + "grad_norm": 1.7667136366773373, + "learning_rate": 9.774282617571129e-07, + "loss": 0.1049, + "step": 6798 + }, + { + "epoch": 1.5469852104664392, + "grad_norm": 1.8930266011594143, + "learning_rate": 9.773544670792243e-07, + "loss": 0.1935, + "step": 6799 + }, + { + "epoch": 1.547212741751991, + "grad_norm": 1.7813371718575919, + "learning_rate": 9.772806651998555e-07, + "loss": 0.1241, + "step": 6800 + }, + { + "epoch": 1.5474402730375427, + "grad_norm": 1.504169607795343, + "learning_rate": 9.772068561205152e-07, + "loss": 0.0828, + "step": 6801 + }, + { + "epoch": 1.5476678043230945, + "grad_norm": 1.3029079334587395, + "learning_rate": 9.771330398427118e-07, + "loss": 0.1542, + "step": 6802 + }, + { + "epoch": 1.5478953356086462, + "grad_norm": 1.4723702531747105, + "learning_rate": 9.770592163679539e-07, + "loss": 0.0748, + "step": 6803 + }, + { + "epoch": 1.548122866894198, + "grad_norm": 1.5167655147301062, + "learning_rate": 9.769853856977503e-07, + "loss": 0.0596, + "step": 6804 + }, + { + "epoch": 1.5483503981797497, + "grad_norm": 1.2483226690375377, + "learning_rate": 9.769115478336102e-07, + "loss": 0.0558, + "step": 6805 + }, + { + "epoch": 1.5485779294653015, + "grad_norm": 2.610342632868789, + "learning_rate": 9.768377027770427e-07, + "loss": 0.2088, + "step": 6806 + }, + { + "epoch": 1.5488054607508532, + "grad_norm": 0.948433081408228, + "learning_rate": 9.767638505295566e-07, + "loss": 0.0764, + "step": 6807 + }, + { + "epoch": 1.549032992036405, + "grad_norm": 1.110321529246112, + "learning_rate": 9.766899910926617e-07, + "loss": 0.1255, + "step": 6808 + }, + { + "epoch": 1.5492605233219567, + "grad_norm": 2.1808152587794467, + "learning_rate": 9.766161244678675e-07, + "loss": 0.1556, + "step": 6809 + }, + { + "epoch": 1.5494880546075085, + "grad_norm": 1.9798598047343043, + "learning_rate": 9.765422506566837e-07, + "loss": 0.0551, + "step": 6810 + }, + { + "epoch": 1.5497155858930602, + "grad_norm": 1.6743554420444267, + "learning_rate": 9.7646836966062e-07, + "loss": 0.1103, + "step": 6811 + }, + { + "epoch": 1.549943117178612, + "grad_norm": 2.2413105664096924, + "learning_rate": 9.763944814811866e-07, + "loss": 0.1012, + "step": 6812 + }, + { + "epoch": 1.5501706484641637, + "grad_norm": 1.447294551212359, + "learning_rate": 9.763205861198935e-07, + "loss": 0.0626, + "step": 6813 + }, + { + "epoch": 1.5503981797497155, + "grad_norm": 1.1736663175641113, + "learning_rate": 9.76246683578251e-07, + "loss": 0.0856, + "step": 6814 + }, + { + "epoch": 1.5506257110352673, + "grad_norm": 1.5996826005862175, + "learning_rate": 9.761727738577698e-07, + "loss": 0.0855, + "step": 6815 + }, + { + "epoch": 1.550853242320819, + "grad_norm": 3.3557607688406166, + "learning_rate": 9.760988569599602e-07, + "loss": 0.2147, + "step": 6816 + }, + { + "epoch": 1.5510807736063708, + "grad_norm": 2.009826047910445, + "learning_rate": 9.760249328863328e-07, + "loss": 0.1139, + "step": 6817 + }, + { + "epoch": 1.5513083048919225, + "grad_norm": 2.5898350691364675, + "learning_rate": 9.759510016383987e-07, + "loss": 0.0993, + "step": 6818 + }, + { + "epoch": 1.5515358361774743, + "grad_norm": 1.7623781439792425, + "learning_rate": 9.758770632176688e-07, + "loss": 0.1161, + "step": 6819 + }, + { + "epoch": 1.551763367463026, + "grad_norm": 3.6161571986406833, + "learning_rate": 9.758031176256543e-07, + "loss": 0.1231, + "step": 6820 + }, + { + "epoch": 1.5519908987485778, + "grad_norm": 3.205748656453346, + "learning_rate": 9.757291648638666e-07, + "loss": 0.1015, + "step": 6821 + }, + { + "epoch": 1.5522184300341297, + "grad_norm": 1.9988303163461136, + "learning_rate": 9.756552049338174e-07, + "loss": 0.0921, + "step": 6822 + }, + { + "epoch": 1.5524459613196815, + "grad_norm": 2.203100020710235, + "learning_rate": 9.755812378370177e-07, + "loss": 0.1434, + "step": 6823 + }, + { + "epoch": 1.5526734926052332, + "grad_norm": 1.6940970795845784, + "learning_rate": 9.755072635749795e-07, + "loss": 0.1209, + "step": 6824 + }, + { + "epoch": 1.552901023890785, + "grad_norm": 2.3652238714277196, + "learning_rate": 9.754332821492148e-07, + "loss": 0.0594, + "step": 6825 + }, + { + "epoch": 1.5531285551763367, + "grad_norm": 1.330771904775176, + "learning_rate": 9.753592935612358e-07, + "loss": 0.0703, + "step": 6826 + }, + { + "epoch": 1.5533560864618885, + "grad_norm": 1.3972072191448723, + "learning_rate": 9.752852978125544e-07, + "loss": 0.0647, + "step": 6827 + }, + { + "epoch": 1.5535836177474402, + "grad_norm": 1.3360409284842407, + "learning_rate": 9.75211294904683e-07, + "loss": 0.112, + "step": 6828 + }, + { + "epoch": 1.553811149032992, + "grad_norm": 2.137018392575463, + "learning_rate": 9.75137284839134e-07, + "loss": 0.129, + "step": 6829 + }, + { + "epoch": 1.5540386803185438, + "grad_norm": 2.097333377582825, + "learning_rate": 9.750632676174201e-07, + "loss": 0.1122, + "step": 6830 + }, + { + "epoch": 1.5542662116040957, + "grad_norm": 2.657275222687091, + "learning_rate": 9.749892432410544e-07, + "loss": 0.1439, + "step": 6831 + }, + { + "epoch": 1.5544937428896475, + "grad_norm": 4.847015432196619, + "learning_rate": 9.749152117115494e-07, + "loss": 0.1597, + "step": 6832 + }, + { + "epoch": 1.5547212741751992, + "grad_norm": 1.2808956168274448, + "learning_rate": 9.748411730304184e-07, + "loss": 0.0966, + "step": 6833 + }, + { + "epoch": 1.554948805460751, + "grad_norm": 1.1006280066879472, + "learning_rate": 9.747671271991746e-07, + "loss": 0.0373, + "step": 6834 + }, + { + "epoch": 1.5551763367463027, + "grad_norm": 1.5930242822758494, + "learning_rate": 9.746930742193307e-07, + "loss": 0.064, + "step": 6835 + }, + { + "epoch": 1.5554038680318545, + "grad_norm": 1.6233957248439397, + "learning_rate": 9.746190140924014e-07, + "loss": 0.1068, + "step": 6836 + }, + { + "epoch": 1.5556313993174062, + "grad_norm": 1.6523988807255852, + "learning_rate": 9.745449468198997e-07, + "loss": 0.0785, + "step": 6837 + }, + { + "epoch": 1.555858930602958, + "grad_norm": 2.005971472506433, + "learning_rate": 9.744708724033393e-07, + "loss": 0.1289, + "step": 6838 + }, + { + "epoch": 1.5560864618885097, + "grad_norm": 0.9656668110662248, + "learning_rate": 9.743967908442343e-07, + "loss": 0.0903, + "step": 6839 + }, + { + "epoch": 1.5563139931740615, + "grad_norm": 1.4731049586322165, + "learning_rate": 9.743227021440988e-07, + "loss": 0.0798, + "step": 6840 + }, + { + "epoch": 1.5565415244596132, + "grad_norm": 2.445080003949355, + "learning_rate": 9.74248606304447e-07, + "loss": 0.0935, + "step": 6841 + }, + { + "epoch": 1.556769055745165, + "grad_norm": 1.7224040698601348, + "learning_rate": 9.741745033267932e-07, + "loss": 0.0714, + "step": 6842 + }, + { + "epoch": 1.5569965870307167, + "grad_norm": 2.235660729949979, + "learning_rate": 9.741003932126522e-07, + "loss": 0.1476, + "step": 6843 + }, + { + "epoch": 1.5572241183162685, + "grad_norm": 1.7270359857667794, + "learning_rate": 9.740262759635386e-07, + "loss": 0.1641, + "step": 6844 + }, + { + "epoch": 1.5574516496018203, + "grad_norm": 1.5709024234225488, + "learning_rate": 9.739521515809669e-07, + "loss": 0.0562, + "step": 6845 + }, + { + "epoch": 1.557679180887372, + "grad_norm": 1.6021103887440078, + "learning_rate": 9.738780200664525e-07, + "loss": 0.0635, + "step": 6846 + }, + { + "epoch": 1.5579067121729238, + "grad_norm": 1.5543318427321744, + "learning_rate": 9.738038814215102e-07, + "loss": 0.0687, + "step": 6847 + }, + { + "epoch": 1.5581342434584755, + "grad_norm": 1.9060796993827795, + "learning_rate": 9.737297356476554e-07, + "loss": 0.0939, + "step": 6848 + }, + { + "epoch": 1.5583617747440273, + "grad_norm": 1.2877821231918085, + "learning_rate": 9.736555827464034e-07, + "loss": 0.1147, + "step": 6849 + }, + { + "epoch": 1.558589306029579, + "grad_norm": 2.1612992390892454, + "learning_rate": 9.7358142271927e-07, + "loss": 0.0986, + "step": 6850 + }, + { + "epoch": 1.5588168373151308, + "grad_norm": 2.139466326520036, + "learning_rate": 9.735072555677705e-07, + "loss": 0.079, + "step": 6851 + }, + { + "epoch": 1.5590443686006825, + "grad_norm": 1.690342851607921, + "learning_rate": 9.73433081293421e-07, + "loss": 0.0702, + "step": 6852 + }, + { + "epoch": 1.5592718998862343, + "grad_norm": 2.288411560849199, + "learning_rate": 9.733588998977376e-07, + "loss": 0.1057, + "step": 6853 + }, + { + "epoch": 1.559499431171786, + "grad_norm": 1.4867185756863013, + "learning_rate": 9.73284711382236e-07, + "loss": 0.0917, + "step": 6854 + }, + { + "epoch": 1.5597269624573378, + "grad_norm": 2.007357013936276, + "learning_rate": 9.732105157484332e-07, + "loss": 0.1453, + "step": 6855 + }, + { + "epoch": 1.5599544937428895, + "grad_norm": 2.2567812237722786, + "learning_rate": 9.731363129978447e-07, + "loss": 0.0901, + "step": 6856 + }, + { + "epoch": 1.5601820250284413, + "grad_norm": 2.386291564760578, + "learning_rate": 9.730621031319878e-07, + "loss": 0.1216, + "step": 6857 + }, + { + "epoch": 1.560409556313993, + "grad_norm": 2.2353388504604723, + "learning_rate": 9.729878861523788e-07, + "loss": 0.0815, + "step": 6858 + }, + { + "epoch": 1.5606370875995448, + "grad_norm": 2.077499479091852, + "learning_rate": 9.729136620605347e-07, + "loss": 0.0838, + "step": 6859 + }, + { + "epoch": 1.5608646188850968, + "grad_norm": 2.2800936343714078, + "learning_rate": 9.728394308579727e-07, + "loss": 0.1203, + "step": 6860 + }, + { + "epoch": 1.5610921501706485, + "grad_norm": 2.1406140087060828, + "learning_rate": 9.727651925462098e-07, + "loss": 0.1006, + "step": 6861 + }, + { + "epoch": 1.5613196814562003, + "grad_norm": 1.5127165240822227, + "learning_rate": 9.726909471267632e-07, + "loss": 0.0694, + "step": 6862 + }, + { + "epoch": 1.561547212741752, + "grad_norm": 1.6310699035842238, + "learning_rate": 9.726166946011503e-07, + "loss": 0.0844, + "step": 6863 + }, + { + "epoch": 1.5617747440273038, + "grad_norm": 1.7691994331743215, + "learning_rate": 9.72542434970889e-07, + "loss": 0.1549, + "step": 6864 + }, + { + "epoch": 1.5620022753128555, + "grad_norm": 1.8051651007893625, + "learning_rate": 9.724681682374965e-07, + "loss": 0.0917, + "step": 6865 + }, + { + "epoch": 1.5622298065984073, + "grad_norm": 2.047807046008497, + "learning_rate": 9.723938944024913e-07, + "loss": 0.1174, + "step": 6866 + }, + { + "epoch": 1.562457337883959, + "grad_norm": 1.6345050861116583, + "learning_rate": 9.72319613467391e-07, + "loss": 0.0487, + "step": 6867 + }, + { + "epoch": 1.5626848691695108, + "grad_norm": 2.9889942861711254, + "learning_rate": 9.722453254337139e-07, + "loss": 0.2182, + "step": 6868 + }, + { + "epoch": 1.5629124004550625, + "grad_norm": 2.086195248205776, + "learning_rate": 9.721710303029783e-07, + "loss": 0.0827, + "step": 6869 + }, + { + "epoch": 1.5631399317406145, + "grad_norm": 1.7647791997831161, + "learning_rate": 9.720967280767026e-07, + "loss": 0.0589, + "step": 6870 + }, + { + "epoch": 1.5633674630261662, + "grad_norm": 2.0543205301100684, + "learning_rate": 9.720224187564057e-07, + "loss": 0.0689, + "step": 6871 + }, + { + "epoch": 1.563594994311718, + "grad_norm": 1.6868564624508946, + "learning_rate": 9.719481023436059e-07, + "loss": 0.0729, + "step": 6872 + }, + { + "epoch": 1.5638225255972698, + "grad_norm": 1.51683971139059, + "learning_rate": 9.718737788398223e-07, + "loss": 0.0938, + "step": 6873 + }, + { + "epoch": 1.5640500568828215, + "grad_norm": 2.696283500518205, + "learning_rate": 9.71799448246574e-07, + "loss": 0.2259, + "step": 6874 + }, + { + "epoch": 1.5642775881683733, + "grad_norm": 1.3830068949348764, + "learning_rate": 9.717251105653799e-07, + "loss": 0.0391, + "step": 6875 + }, + { + "epoch": 1.564505119453925, + "grad_norm": 2.369926309898731, + "learning_rate": 9.716507657977597e-07, + "loss": 0.1235, + "step": 6876 + }, + { + "epoch": 1.5647326507394768, + "grad_norm": 2.801317707906041, + "learning_rate": 9.715764139452327e-07, + "loss": 0.1541, + "step": 6877 + }, + { + "epoch": 1.5649601820250285, + "grad_norm": 1.8375888059655026, + "learning_rate": 9.715020550093185e-07, + "loss": 0.1213, + "step": 6878 + }, + { + "epoch": 1.5651877133105803, + "grad_norm": 1.6745414210291196, + "learning_rate": 9.71427688991537e-07, + "loss": 0.0851, + "step": 6879 + }, + { + "epoch": 1.565415244596132, + "grad_norm": 1.8212934551331674, + "learning_rate": 9.713533158934079e-07, + "loss": 0.0878, + "step": 6880 + }, + { + "epoch": 1.5656427758816838, + "grad_norm": 1.6967874312270814, + "learning_rate": 9.712789357164512e-07, + "loss": 0.0521, + "step": 6881 + }, + { + "epoch": 1.5658703071672355, + "grad_norm": 2.0607765520963945, + "learning_rate": 9.712045484621874e-07, + "loss": 0.1194, + "step": 6882 + }, + { + "epoch": 1.5660978384527873, + "grad_norm": 1.5593897600597133, + "learning_rate": 9.711301541321365e-07, + "loss": 0.0679, + "step": 6883 + }, + { + "epoch": 1.566325369738339, + "grad_norm": 1.7857655925492695, + "learning_rate": 9.710557527278195e-07, + "loss": 0.0687, + "step": 6884 + }, + { + "epoch": 1.5665529010238908, + "grad_norm": 2.032305412937479, + "learning_rate": 9.709813442507565e-07, + "loss": 0.0935, + "step": 6885 + }, + { + "epoch": 1.5667804323094425, + "grad_norm": 1.983196083400438, + "learning_rate": 9.709069287024684e-07, + "loss": 0.1605, + "step": 6886 + }, + { + "epoch": 1.5670079635949943, + "grad_norm": 1.5093436712574515, + "learning_rate": 9.708325060844761e-07, + "loss": 0.0709, + "step": 6887 + }, + { + "epoch": 1.567235494880546, + "grad_norm": 1.9807068159490688, + "learning_rate": 9.707580763983008e-07, + "loss": 0.0753, + "step": 6888 + }, + { + "epoch": 1.5674630261660978, + "grad_norm": 1.5396665395951017, + "learning_rate": 9.706836396454638e-07, + "loss": 0.0645, + "step": 6889 + }, + { + "epoch": 1.5676905574516495, + "grad_norm": 1.8268486418358298, + "learning_rate": 9.70609195827486e-07, + "loss": 0.0563, + "step": 6890 + }, + { + "epoch": 1.5679180887372013, + "grad_norm": 1.643220539321596, + "learning_rate": 9.705347449458896e-07, + "loss": 0.0881, + "step": 6891 + }, + { + "epoch": 1.568145620022753, + "grad_norm": 1.442144111123451, + "learning_rate": 9.704602870021954e-07, + "loss": 0.1025, + "step": 6892 + }, + { + "epoch": 1.5683731513083048, + "grad_norm": 1.576138061818805, + "learning_rate": 9.70385821997926e-07, + "loss": 0.169, + "step": 6893 + }, + { + "epoch": 1.5686006825938565, + "grad_norm": 2.5159254541918026, + "learning_rate": 9.703113499346026e-07, + "loss": 0.1124, + "step": 6894 + }, + { + "epoch": 1.5688282138794083, + "grad_norm": 0.9136715615253127, + "learning_rate": 9.70236870813748e-07, + "loss": 0.0634, + "step": 6895 + }, + { + "epoch": 1.56905574516496, + "grad_norm": 1.3668108620677517, + "learning_rate": 9.701623846368836e-07, + "loss": 0.0427, + "step": 6896 + }, + { + "epoch": 1.5692832764505118, + "grad_norm": 1.6270016341076774, + "learning_rate": 9.700878914055325e-07, + "loss": 0.17, + "step": 6897 + }, + { + "epoch": 1.5695108077360636, + "grad_norm": 2.350790731713196, + "learning_rate": 9.700133911212168e-07, + "loss": 0.1382, + "step": 6898 + }, + { + "epoch": 1.5697383390216155, + "grad_norm": 2.0493315460254795, + "learning_rate": 9.699388837854593e-07, + "loss": 0.0823, + "step": 6899 + }, + { + "epoch": 1.5699658703071673, + "grad_norm": 1.6522265372325529, + "learning_rate": 9.698643693997827e-07, + "loss": 0.1321, + "step": 6900 + }, + { + "epoch": 1.570193401592719, + "grad_norm": 1.348507321966463, + "learning_rate": 9.697898479657098e-07, + "loss": 0.0652, + "step": 6901 + }, + { + "epoch": 1.5704209328782708, + "grad_norm": 1.4026345515667455, + "learning_rate": 9.697153194847641e-07, + "loss": 0.15, + "step": 6902 + }, + { + "epoch": 1.5706484641638225, + "grad_norm": 1.5516065262403587, + "learning_rate": 9.696407839584684e-07, + "loss": 0.1302, + "step": 6903 + }, + { + "epoch": 1.5708759954493743, + "grad_norm": 1.736103408391572, + "learning_rate": 9.695662413883466e-07, + "loss": 0.0623, + "step": 6904 + }, + { + "epoch": 1.571103526734926, + "grad_norm": 1.0606800077660063, + "learning_rate": 9.694916917759218e-07, + "loss": 0.0757, + "step": 6905 + }, + { + "epoch": 1.5713310580204778, + "grad_norm": 1.8230964381380002, + "learning_rate": 9.694171351227175e-07, + "loss": 0.1306, + "step": 6906 + }, + { + "epoch": 1.5715585893060295, + "grad_norm": 2.1893081391718052, + "learning_rate": 9.693425714302577e-07, + "loss": 0.066, + "step": 6907 + }, + { + "epoch": 1.5717861205915813, + "grad_norm": 1.2965655169374837, + "learning_rate": 9.692680007000663e-07, + "loss": 0.03, + "step": 6908 + }, + { + "epoch": 1.5720136518771333, + "grad_norm": 1.9214288656500598, + "learning_rate": 9.691934229336677e-07, + "loss": 0.058, + "step": 6909 + }, + { + "epoch": 1.572241183162685, + "grad_norm": 2.0891285566817372, + "learning_rate": 9.69118838132586e-07, + "loss": 0.0738, + "step": 6910 + }, + { + "epoch": 1.5724687144482368, + "grad_norm": 1.4112326396809503, + "learning_rate": 9.69044246298345e-07, + "loss": 0.0964, + "step": 6911 + }, + { + "epoch": 1.5726962457337885, + "grad_norm": 1.1066871989030969, + "learning_rate": 9.689696474324703e-07, + "loss": 0.0537, + "step": 6912 + }, + { + "epoch": 1.5729237770193403, + "grad_norm": 1.128146729687468, + "learning_rate": 9.688950415364855e-07, + "loss": 0.102, + "step": 6913 + }, + { + "epoch": 1.573151308304892, + "grad_norm": 1.5871840992559776, + "learning_rate": 9.68820428611916e-07, + "loss": 0.2003, + "step": 6914 + }, + { + "epoch": 1.5733788395904438, + "grad_norm": 2.315116795851502, + "learning_rate": 9.687458086602866e-07, + "loss": 0.1135, + "step": 6915 + }, + { + "epoch": 1.5736063708759955, + "grad_norm": 1.214880530889042, + "learning_rate": 9.686711816831226e-07, + "loss": 0.0717, + "step": 6916 + }, + { + "epoch": 1.5738339021615473, + "grad_norm": 1.8811606454809604, + "learning_rate": 9.68596547681949e-07, + "loss": 0.0666, + "step": 6917 + }, + { + "epoch": 1.574061433447099, + "grad_norm": 1.4841201037319591, + "learning_rate": 9.68521906658291e-07, + "loss": 0.1282, + "step": 6918 + }, + { + "epoch": 1.5742889647326508, + "grad_norm": 1.6912614659007157, + "learning_rate": 9.684472586136745e-07, + "loss": 0.0507, + "step": 6919 + }, + { + "epoch": 1.5745164960182025, + "grad_norm": 1.100232031105278, + "learning_rate": 9.68372603549625e-07, + "loss": 0.0753, + "step": 6920 + }, + { + "epoch": 1.5747440273037543, + "grad_norm": 2.0618384339064644, + "learning_rate": 9.682979414676682e-07, + "loss": 0.0743, + "step": 6921 + }, + { + "epoch": 1.574971558589306, + "grad_norm": 3.092052885379181, + "learning_rate": 9.682232723693305e-07, + "loss": 0.0985, + "step": 6922 + }, + { + "epoch": 1.5751990898748578, + "grad_norm": 1.26881499733035, + "learning_rate": 9.681485962561377e-07, + "loss": 0.0757, + "step": 6923 + }, + { + "epoch": 1.5754266211604095, + "grad_norm": 1.2213566354259484, + "learning_rate": 9.680739131296158e-07, + "loss": 0.0995, + "step": 6924 + }, + { + "epoch": 1.5756541524459613, + "grad_norm": 1.8835777410462962, + "learning_rate": 9.679992229912914e-07, + "loss": 0.1492, + "step": 6925 + }, + { + "epoch": 1.575881683731513, + "grad_norm": 2.5455892384333185, + "learning_rate": 9.67924525842691e-07, + "loss": 0.1431, + "step": 6926 + }, + { + "epoch": 1.5761092150170648, + "grad_norm": 1.7592564778289752, + "learning_rate": 9.678498216853416e-07, + "loss": 0.0809, + "step": 6927 + }, + { + "epoch": 1.5763367463026166, + "grad_norm": 2.5212143965708287, + "learning_rate": 9.677751105207696e-07, + "loss": 0.0813, + "step": 6928 + }, + { + "epoch": 1.5765642775881683, + "grad_norm": 1.6364687457943778, + "learning_rate": 9.677003923505025e-07, + "loss": 0.1237, + "step": 6929 + }, + { + "epoch": 1.57679180887372, + "grad_norm": 1.9567829536702195, + "learning_rate": 9.676256671760665e-07, + "loss": 0.1214, + "step": 6930 + }, + { + "epoch": 1.5770193401592718, + "grad_norm": 1.8218655104471277, + "learning_rate": 9.675509349989896e-07, + "loss": 0.108, + "step": 6931 + }, + { + "epoch": 1.5772468714448236, + "grad_norm": 2.0740005049352344, + "learning_rate": 9.674761958207986e-07, + "loss": 0.0797, + "step": 6932 + }, + { + "epoch": 1.5774744027303753, + "grad_norm": 2.0550059922871786, + "learning_rate": 9.674014496430218e-07, + "loss": 0.0698, + "step": 6933 + }, + { + "epoch": 1.577701934015927, + "grad_norm": 1.636582063223626, + "learning_rate": 9.673266964671862e-07, + "loss": 0.0513, + "step": 6934 + }, + { + "epoch": 1.5779294653014788, + "grad_norm": 3.082308543745888, + "learning_rate": 9.6725193629482e-07, + "loss": 0.0955, + "step": 6935 + }, + { + "epoch": 1.5781569965870306, + "grad_norm": 1.7853501137033265, + "learning_rate": 9.671771691274508e-07, + "loss": 0.0976, + "step": 6936 + }, + { + "epoch": 1.5783845278725823, + "grad_norm": 3.0256879779417534, + "learning_rate": 9.671023949666073e-07, + "loss": 0.0987, + "step": 6937 + }, + { + "epoch": 1.5786120591581343, + "grad_norm": 1.3498446575744398, + "learning_rate": 9.67027613813817e-07, + "loss": 0.071, + "step": 6938 + }, + { + "epoch": 1.578839590443686, + "grad_norm": 1.7436971564140509, + "learning_rate": 9.66952825670609e-07, + "loss": 0.1357, + "step": 6939 + }, + { + "epoch": 1.5790671217292378, + "grad_norm": 1.5228285224595741, + "learning_rate": 9.66878030538511e-07, + "loss": 0.0502, + "step": 6940 + }, + { + "epoch": 1.5792946530147896, + "grad_norm": 2.9682409454518948, + "learning_rate": 9.668032284190529e-07, + "loss": 0.0779, + "step": 6941 + }, + { + "epoch": 1.5795221843003413, + "grad_norm": 1.6246373361607298, + "learning_rate": 9.667284193137622e-07, + "loss": 0.0805, + "step": 6942 + }, + { + "epoch": 1.579749715585893, + "grad_norm": 1.3560913793817733, + "learning_rate": 9.666536032241687e-07, + "loss": 0.1614, + "step": 6943 + }, + { + "epoch": 1.5799772468714448, + "grad_norm": 1.6031076552379593, + "learning_rate": 9.66578780151801e-07, + "loss": 0.1452, + "step": 6944 + }, + { + "epoch": 1.5802047781569966, + "grad_norm": 2.14630884242415, + "learning_rate": 9.66503950098189e-07, + "loss": 0.0938, + "step": 6945 + }, + { + "epoch": 1.5804323094425483, + "grad_norm": 2.300918036454428, + "learning_rate": 9.664291130648616e-07, + "loss": 0.1036, + "step": 6946 + }, + { + "epoch": 1.5806598407281, + "grad_norm": 2.076711835486533, + "learning_rate": 9.663542690533485e-07, + "loss": 0.0978, + "step": 6947 + }, + { + "epoch": 1.580887372013652, + "grad_norm": 1.9858408778781345, + "learning_rate": 9.66279418065179e-07, + "loss": 0.1235, + "step": 6948 + }, + { + "epoch": 1.5811149032992038, + "grad_norm": 2.8134635569970556, + "learning_rate": 9.662045601018834e-07, + "loss": 0.0959, + "step": 6949 + }, + { + "epoch": 1.5813424345847555, + "grad_norm": 2.412041671272594, + "learning_rate": 9.661296951649914e-07, + "loss": 0.1581, + "step": 6950 + }, + { + "epoch": 1.5815699658703073, + "grad_norm": 2.212966835054143, + "learning_rate": 9.660548232560331e-07, + "loss": 0.1705, + "step": 6951 + }, + { + "epoch": 1.581797497155859, + "grad_norm": 1.0167103004961158, + "learning_rate": 9.659799443765392e-07, + "loss": 0.1194, + "step": 6952 + }, + { + "epoch": 1.5820250284414108, + "grad_norm": 1.5945837819374222, + "learning_rate": 9.659050585280394e-07, + "loss": 0.1693, + "step": 6953 + }, + { + "epoch": 1.5822525597269625, + "grad_norm": 1.401656682392233, + "learning_rate": 9.658301657120646e-07, + "loss": 0.1292, + "step": 6954 + }, + { + "epoch": 1.5824800910125143, + "grad_norm": 1.2110702975067411, + "learning_rate": 9.657552659301455e-07, + "loss": 0.0494, + "step": 6955 + }, + { + "epoch": 1.582707622298066, + "grad_norm": 2.610040851944766, + "learning_rate": 9.656803591838126e-07, + "loss": 0.1847, + "step": 6956 + }, + { + "epoch": 1.5829351535836178, + "grad_norm": 0.863503979710921, + "learning_rate": 9.656054454745973e-07, + "loss": 0.0815, + "step": 6957 + }, + { + "epoch": 1.5831626848691696, + "grad_norm": 1.3367436245719442, + "learning_rate": 9.655305248040302e-07, + "loss": 0.1176, + "step": 6958 + }, + { + "epoch": 1.5833902161547213, + "grad_norm": 1.8868194318678535, + "learning_rate": 9.654555971736431e-07, + "loss": 0.0984, + "step": 6959 + }, + { + "epoch": 1.583617747440273, + "grad_norm": 1.4598869301223507, + "learning_rate": 9.653806625849671e-07, + "loss": 0.1053, + "step": 6960 + }, + { + "epoch": 1.5838452787258248, + "grad_norm": 0.9993222033681556, + "learning_rate": 9.653057210395338e-07, + "loss": 0.0589, + "step": 6961 + }, + { + "epoch": 1.5840728100113766, + "grad_norm": 1.4700139285725813, + "learning_rate": 9.652307725388746e-07, + "loss": 0.0704, + "step": 6962 + }, + { + "epoch": 1.5843003412969283, + "grad_norm": 1.499868989401943, + "learning_rate": 9.651558170845216e-07, + "loss": 0.0495, + "step": 6963 + }, + { + "epoch": 1.58452787258248, + "grad_norm": 2.7042028681340655, + "learning_rate": 9.650808546780068e-07, + "loss": 0.1304, + "step": 6964 + }, + { + "epoch": 1.5847554038680318, + "grad_norm": 2.9874751883735424, + "learning_rate": 9.65005885320862e-07, + "loss": 0.0828, + "step": 6965 + }, + { + "epoch": 1.5849829351535836, + "grad_norm": 1.655538505482436, + "learning_rate": 9.649309090146194e-07, + "loss": 0.0925, + "step": 6966 + }, + { + "epoch": 1.5852104664391353, + "grad_norm": 2.0441507873994125, + "learning_rate": 9.64855925760812e-07, + "loss": 0.0516, + "step": 6967 + }, + { + "epoch": 1.585437997724687, + "grad_norm": 1.1883886573945852, + "learning_rate": 9.647809355609715e-07, + "loss": 0.1587, + "step": 6968 + }, + { + "epoch": 1.5856655290102388, + "grad_norm": 1.7093461262727119, + "learning_rate": 9.647059384166314e-07, + "loss": 0.0608, + "step": 6969 + }, + { + "epoch": 1.5858930602957906, + "grad_norm": 1.8247176331808013, + "learning_rate": 9.646309343293237e-07, + "loss": 0.1467, + "step": 6970 + }, + { + "epoch": 1.5861205915813423, + "grad_norm": 1.8193934335145596, + "learning_rate": 9.645559233005819e-07, + "loss": 0.0931, + "step": 6971 + }, + { + "epoch": 1.586348122866894, + "grad_norm": 2.347813013209486, + "learning_rate": 9.644809053319388e-07, + "loss": 0.1074, + "step": 6972 + }, + { + "epoch": 1.5865756541524458, + "grad_norm": 1.9276522554322204, + "learning_rate": 9.644058804249276e-07, + "loss": 0.1117, + "step": 6973 + }, + { + "epoch": 1.5868031854379976, + "grad_norm": 1.1894784473133884, + "learning_rate": 9.64330848581082e-07, + "loss": 0.1122, + "step": 6974 + }, + { + "epoch": 1.5870307167235493, + "grad_norm": 1.381690182100237, + "learning_rate": 9.642558098019353e-07, + "loss": 0.0875, + "step": 6975 + }, + { + "epoch": 1.587258248009101, + "grad_norm": 1.379264433533777, + "learning_rate": 9.641807640890212e-07, + "loss": 0.0586, + "step": 6976 + }, + { + "epoch": 1.587485779294653, + "grad_norm": 1.3421330124356945, + "learning_rate": 9.64105711443873e-07, + "loss": 0.0991, + "step": 6977 + }, + { + "epoch": 1.5877133105802048, + "grad_norm": 1.517449553301864, + "learning_rate": 9.640306518680257e-07, + "loss": 0.0792, + "step": 6978 + }, + { + "epoch": 1.5879408418657566, + "grad_norm": 1.4648923960371314, + "learning_rate": 9.639555853630126e-07, + "loss": 0.0689, + "step": 6979 + }, + { + "epoch": 1.5881683731513083, + "grad_norm": 2.1326027621988772, + "learning_rate": 9.63880511930368e-07, + "loss": 0.0876, + "step": 6980 + }, + { + "epoch": 1.58839590443686, + "grad_norm": 1.5553369120899478, + "learning_rate": 9.638054315716264e-07, + "loss": 0.0973, + "step": 6981 + }, + { + "epoch": 1.5886234357224118, + "grad_norm": 1.440135727652256, + "learning_rate": 9.637303442883223e-07, + "loss": 0.0428, + "step": 6982 + }, + { + "epoch": 1.5888509670079636, + "grad_norm": 1.756091503554684, + "learning_rate": 9.636552500819903e-07, + "loss": 0.0636, + "step": 6983 + }, + { + "epoch": 1.5890784982935153, + "grad_norm": 1.0526721397744394, + "learning_rate": 9.635801489541652e-07, + "loss": 0.0631, + "step": 6984 + }, + { + "epoch": 1.589306029579067, + "grad_norm": 1.9369758747288157, + "learning_rate": 9.635050409063818e-07, + "loss": 0.0821, + "step": 6985 + }, + { + "epoch": 1.5895335608646188, + "grad_norm": 1.4734142882640557, + "learning_rate": 9.634299259401756e-07, + "loss": 0.1045, + "step": 6986 + }, + { + "epoch": 1.5897610921501708, + "grad_norm": 1.7504083792088394, + "learning_rate": 9.633548040570815e-07, + "loss": 0.1305, + "step": 6987 + }, + { + "epoch": 1.5899886234357226, + "grad_norm": 2.0842740380498563, + "learning_rate": 9.632796752586345e-07, + "loss": 0.2348, + "step": 6988 + }, + { + "epoch": 1.5902161547212743, + "grad_norm": 1.5809150377120453, + "learning_rate": 9.632045395463708e-07, + "loss": 0.1034, + "step": 6989 + }, + { + "epoch": 1.590443686006826, + "grad_norm": 2.2274737894742227, + "learning_rate": 9.631293969218256e-07, + "loss": 0.1022, + "step": 6990 + }, + { + "epoch": 1.5906712172923778, + "grad_norm": 1.4521152592202264, + "learning_rate": 9.63054247386535e-07, + "loss": 0.1371, + "step": 6991 + }, + { + "epoch": 1.5908987485779296, + "grad_norm": 2.0326995835147565, + "learning_rate": 9.629790909420344e-07, + "loss": 0.0669, + "step": 6992 + }, + { + "epoch": 1.5911262798634813, + "grad_norm": 1.6732679871311429, + "learning_rate": 9.629039275898603e-07, + "loss": 0.0411, + "step": 6993 + }, + { + "epoch": 1.591353811149033, + "grad_norm": 1.682539766220766, + "learning_rate": 9.628287573315488e-07, + "loss": 0.0717, + "step": 6994 + }, + { + "epoch": 1.5915813424345848, + "grad_norm": 1.883899405259209, + "learning_rate": 9.627535801686359e-07, + "loss": 0.1088, + "step": 6995 + }, + { + "epoch": 1.5918088737201366, + "grad_norm": 1.765171359290216, + "learning_rate": 9.626783961026586e-07, + "loss": 0.0915, + "step": 6996 + }, + { + "epoch": 1.5920364050056883, + "grad_norm": 1.5215975735336589, + "learning_rate": 9.626032051351534e-07, + "loss": 0.135, + "step": 6997 + }, + { + "epoch": 1.59226393629124, + "grad_norm": 1.3707195749768784, + "learning_rate": 9.62528007267657e-07, + "loss": 0.0686, + "step": 6998 + }, + { + "epoch": 1.5924914675767918, + "grad_norm": 1.7331076088661026, + "learning_rate": 9.62452802501706e-07, + "loss": 0.0607, + "step": 6999 + }, + { + "epoch": 1.5927189988623436, + "grad_norm": 1.7262671826564764, + "learning_rate": 9.62377590838838e-07, + "loss": 0.086, + "step": 7000 + }, + { + "epoch": 1.5929465301478953, + "grad_norm": 2.108256338543314, + "learning_rate": 9.623023722805898e-07, + "loss": 0.2314, + "step": 7001 + }, + { + "epoch": 1.593174061433447, + "grad_norm": 1.6648982855436203, + "learning_rate": 9.62227146828499e-07, + "loss": 0.1015, + "step": 7002 + }, + { + "epoch": 1.5934015927189988, + "grad_norm": 1.764553597167271, + "learning_rate": 9.621519144841028e-07, + "loss": 0.1638, + "step": 7003 + }, + { + "epoch": 1.5936291240045506, + "grad_norm": 1.8331089224965664, + "learning_rate": 9.62076675248939e-07, + "loss": 0.0737, + "step": 7004 + }, + { + "epoch": 1.5938566552901023, + "grad_norm": 1.9442080447235466, + "learning_rate": 9.620014291245452e-07, + "loss": 0.1642, + "step": 7005 + }, + { + "epoch": 1.594084186575654, + "grad_norm": 1.6777130725526501, + "learning_rate": 9.619261761124592e-07, + "loss": 0.0947, + "step": 7006 + }, + { + "epoch": 1.5943117178612058, + "grad_norm": 2.4255255413124663, + "learning_rate": 9.618509162142196e-07, + "loss": 0.1277, + "step": 7007 + }, + { + "epoch": 1.5945392491467576, + "grad_norm": 1.1516566125989556, + "learning_rate": 9.61775649431364e-07, + "loss": 0.0717, + "step": 7008 + }, + { + "epoch": 1.5947667804323093, + "grad_norm": 1.6950557748553574, + "learning_rate": 9.617003757654309e-07, + "loss": 0.1676, + "step": 7009 + }, + { + "epoch": 1.594994311717861, + "grad_norm": 1.4209377855156373, + "learning_rate": 9.616250952179586e-07, + "loss": 0.0595, + "step": 7010 + }, + { + "epoch": 1.5952218430034129, + "grad_norm": 1.3998940240833262, + "learning_rate": 9.615498077904865e-07, + "loss": 0.0571, + "step": 7011 + }, + { + "epoch": 1.5954493742889646, + "grad_norm": 2.4228825196425334, + "learning_rate": 9.614745134845518e-07, + "loss": 0.0836, + "step": 7012 + }, + { + "epoch": 1.5956769055745164, + "grad_norm": 2.5226928466592855, + "learning_rate": 9.613992123016948e-07, + "loss": 0.0989, + "step": 7013 + }, + { + "epoch": 1.595904436860068, + "grad_norm": 2.828698420545608, + "learning_rate": 9.61323904243454e-07, + "loss": 0.1082, + "step": 7014 + }, + { + "epoch": 1.5961319681456199, + "grad_norm": 0.658379542494829, + "learning_rate": 9.612485893113682e-07, + "loss": 0.0617, + "step": 7015 + }, + { + "epoch": 1.5963594994311718, + "grad_norm": 2.043515550440299, + "learning_rate": 9.611732675069773e-07, + "loss": 0.0592, + "step": 7016 + }, + { + "epoch": 1.5965870307167236, + "grad_norm": 1.420713270979496, + "learning_rate": 9.610979388318206e-07, + "loss": 0.0522, + "step": 7017 + }, + { + "epoch": 1.5968145620022753, + "grad_norm": 1.8286267570389563, + "learning_rate": 9.610226032874374e-07, + "loss": 0.0704, + "step": 7018 + }, + { + "epoch": 1.597042093287827, + "grad_norm": 1.9593689633156057, + "learning_rate": 9.609472608753676e-07, + "loss": 0.0873, + "step": 7019 + }, + { + "epoch": 1.5972696245733788, + "grad_norm": 3.025411157925881, + "learning_rate": 9.60871911597151e-07, + "loss": 0.1941, + "step": 7020 + }, + { + "epoch": 1.5974971558589306, + "grad_norm": 1.1756630023838337, + "learning_rate": 9.607965554543276e-07, + "loss": 0.0605, + "step": 7021 + }, + { + "epoch": 1.5977246871444823, + "grad_norm": 1.98439511132966, + "learning_rate": 9.607211924484378e-07, + "loss": 0.0863, + "step": 7022 + }, + { + "epoch": 1.597952218430034, + "grad_norm": 1.4953595829239505, + "learning_rate": 9.606458225810214e-07, + "loss": 0.1295, + "step": 7023 + }, + { + "epoch": 1.5981797497155859, + "grad_norm": 2.7182642433005006, + "learning_rate": 9.605704458536193e-07, + "loss": 0.1064, + "step": 7024 + }, + { + "epoch": 1.5984072810011376, + "grad_norm": 2.3527859069877817, + "learning_rate": 9.604950622677717e-07, + "loss": 0.1657, + "step": 7025 + }, + { + "epoch": 1.5986348122866896, + "grad_norm": 1.036550268362939, + "learning_rate": 9.604196718250197e-07, + "loss": 0.0853, + "step": 7026 + }, + { + "epoch": 1.5988623435722413, + "grad_norm": 1.8246053867576546, + "learning_rate": 9.603442745269036e-07, + "loss": 0.0692, + "step": 7027 + }, + { + "epoch": 1.599089874857793, + "grad_norm": 1.657164284314806, + "learning_rate": 9.602688703749648e-07, + "loss": 0.1082, + "step": 7028 + }, + { + "epoch": 1.5993174061433448, + "grad_norm": 1.6135815088813417, + "learning_rate": 9.601934593707444e-07, + "loss": 0.1634, + "step": 7029 + }, + { + "epoch": 1.5995449374288966, + "grad_norm": 2.5168553017718325, + "learning_rate": 9.601180415157834e-07, + "loss": 0.1249, + "step": 7030 + }, + { + "epoch": 1.5997724687144483, + "grad_norm": 1.9332350220212315, + "learning_rate": 9.600426168116237e-07, + "loss": 0.0954, + "step": 7031 + }, + { + "epoch": 1.6, + "grad_norm": 2.779016388521835, + "learning_rate": 9.599671852598062e-07, + "loss": 0.1163, + "step": 7032 + }, + { + "epoch": 1.6002275312855518, + "grad_norm": 1.709991621608222, + "learning_rate": 9.59891746861873e-07, + "loss": 0.103, + "step": 7033 + }, + { + "epoch": 1.6004550625711036, + "grad_norm": 1.601744659007227, + "learning_rate": 9.598163016193656e-07, + "loss": 0.128, + "step": 7034 + }, + { + "epoch": 1.6006825938566553, + "grad_norm": 1.427848052455716, + "learning_rate": 9.597408495338266e-07, + "loss": 0.085, + "step": 7035 + }, + { + "epoch": 1.600910125142207, + "grad_norm": 1.7114236276422017, + "learning_rate": 9.596653906067974e-07, + "loss": 0.0751, + "step": 7036 + }, + { + "epoch": 1.6011376564277588, + "grad_norm": 1.099324963342583, + "learning_rate": 9.595899248398209e-07, + "loss": 0.0571, + "step": 7037 + }, + { + "epoch": 1.6013651877133106, + "grad_norm": 1.138796067907748, + "learning_rate": 9.595144522344386e-07, + "loss": 0.1072, + "step": 7038 + }, + { + "epoch": 1.6015927189988624, + "grad_norm": 1.25580073682027, + "learning_rate": 9.594389727921937e-07, + "loss": 0.1145, + "step": 7039 + }, + { + "epoch": 1.601820250284414, + "grad_norm": 2.2093517995718455, + "learning_rate": 9.593634865146286e-07, + "loss": 0.0875, + "step": 7040 + }, + { + "epoch": 1.6020477815699659, + "grad_norm": 1.1512438683598305, + "learning_rate": 9.592879934032864e-07, + "loss": 0.1016, + "step": 7041 + }, + { + "epoch": 1.6022753128555176, + "grad_norm": 2.151102117346863, + "learning_rate": 9.5921249345971e-07, + "loss": 0.081, + "step": 7042 + }, + { + "epoch": 1.6025028441410694, + "grad_norm": 1.8116958423155012, + "learning_rate": 9.59136986685442e-07, + "loss": 0.1139, + "step": 7043 + }, + { + "epoch": 1.6027303754266211, + "grad_norm": 1.4728303591830172, + "learning_rate": 9.59061473082026e-07, + "loss": 0.0792, + "step": 7044 + }, + { + "epoch": 1.6029579067121729, + "grad_norm": 1.348123202641275, + "learning_rate": 9.589859526510053e-07, + "loss": 0.1171, + "step": 7045 + }, + { + "epoch": 1.6031854379977246, + "grad_norm": 1.652518244804472, + "learning_rate": 9.589104253939233e-07, + "loss": 0.0981, + "step": 7046 + }, + { + "epoch": 1.6034129692832764, + "grad_norm": 1.6491842823027245, + "learning_rate": 9.588348913123238e-07, + "loss": 0.152, + "step": 7047 + }, + { + "epoch": 1.6036405005688281, + "grad_norm": 2.3757077558379214, + "learning_rate": 9.587593504077506e-07, + "loss": 0.1021, + "step": 7048 + }, + { + "epoch": 1.6038680318543799, + "grad_norm": 1.7408054303248572, + "learning_rate": 9.586838026817475e-07, + "loss": 0.0624, + "step": 7049 + }, + { + "epoch": 1.6040955631399316, + "grad_norm": 1.4722586598022023, + "learning_rate": 9.586082481358587e-07, + "loss": 0.0811, + "step": 7050 + }, + { + "epoch": 1.6043230944254834, + "grad_norm": 1.209641522630536, + "learning_rate": 9.58532686771628e-07, + "loss": 0.084, + "step": 7051 + }, + { + "epoch": 1.6045506257110351, + "grad_norm": 1.5600339093769533, + "learning_rate": 9.584571185906002e-07, + "loss": 0.1278, + "step": 7052 + }, + { + "epoch": 1.6047781569965869, + "grad_norm": 2.2242771765608502, + "learning_rate": 9.583815435943195e-07, + "loss": 0.1165, + "step": 7053 + }, + { + "epoch": 1.6050056882821386, + "grad_norm": 2.2003878293393844, + "learning_rate": 9.583059617843306e-07, + "loss": 0.1519, + "step": 7054 + }, + { + "epoch": 1.6052332195676906, + "grad_norm": 2.030595699092688, + "learning_rate": 9.582303731621784e-07, + "loss": 0.1063, + "step": 7055 + }, + { + "epoch": 1.6054607508532424, + "grad_norm": 1.3464973730328764, + "learning_rate": 9.581547777294076e-07, + "loss": 0.0988, + "step": 7056 + }, + { + "epoch": 1.605688282138794, + "grad_norm": 1.7661580546805442, + "learning_rate": 9.580791754875631e-07, + "loss": 0.0608, + "step": 7057 + }, + { + "epoch": 1.6059158134243459, + "grad_norm": 1.7729834089831826, + "learning_rate": 9.580035664381905e-07, + "loss": 0.0704, + "step": 7058 + }, + { + "epoch": 1.6061433447098976, + "grad_norm": 4.823012303438253, + "learning_rate": 9.579279505828348e-07, + "loss": 0.0881, + "step": 7059 + }, + { + "epoch": 1.6063708759954494, + "grad_norm": 1.9764505036357716, + "learning_rate": 9.578523279230414e-07, + "loss": 0.1049, + "step": 7060 + }, + { + "epoch": 1.6065984072810011, + "grad_norm": 2.0091897106409458, + "learning_rate": 9.577766984603562e-07, + "loss": 0.1441, + "step": 7061 + }, + { + "epoch": 1.6068259385665529, + "grad_norm": 1.2132768642438971, + "learning_rate": 9.577010621963249e-07, + "loss": 0.0754, + "step": 7062 + }, + { + "epoch": 1.6070534698521046, + "grad_norm": 1.9676241704559914, + "learning_rate": 9.576254191324926e-07, + "loss": 0.058, + "step": 7063 + }, + { + "epoch": 1.6072810011376564, + "grad_norm": 1.2672826972829434, + "learning_rate": 9.575497692704063e-07, + "loss": 0.0846, + "step": 7064 + }, + { + "epoch": 1.6075085324232083, + "grad_norm": 1.5216162690213744, + "learning_rate": 9.574741126116118e-07, + "loss": 0.0742, + "step": 7065 + }, + { + "epoch": 1.60773606370876, + "grad_norm": 2.2158930745759986, + "learning_rate": 9.573984491576554e-07, + "loss": 0.081, + "step": 7066 + }, + { + "epoch": 1.6079635949943119, + "grad_norm": 1.2776286929824625, + "learning_rate": 9.573227789100833e-07, + "loss": 0.05, + "step": 7067 + }, + { + "epoch": 1.6081911262798636, + "grad_norm": 1.5841000629893869, + "learning_rate": 9.572471018704422e-07, + "loss": 0.0628, + "step": 7068 + }, + { + "epoch": 1.6084186575654154, + "grad_norm": 1.5963700008271497, + "learning_rate": 9.57171418040279e-07, + "loss": 0.0905, + "step": 7069 + }, + { + "epoch": 1.608646188850967, + "grad_norm": 1.3320973228563622, + "learning_rate": 9.570957274211399e-07, + "loss": 0.0436, + "step": 7070 + }, + { + "epoch": 1.6088737201365189, + "grad_norm": 1.7565713941984569, + "learning_rate": 9.570200300145727e-07, + "loss": 0.1369, + "step": 7071 + }, + { + "epoch": 1.6091012514220706, + "grad_norm": 1.4051062508002854, + "learning_rate": 9.569443258221243e-07, + "loss": 0.1311, + "step": 7072 + }, + { + "epoch": 1.6093287827076224, + "grad_norm": 1.5003808219816102, + "learning_rate": 9.568686148453412e-07, + "loss": 0.0552, + "step": 7073 + }, + { + "epoch": 1.6095563139931741, + "grad_norm": 1.136113330681752, + "learning_rate": 9.56792897085772e-07, + "loss": 0.1055, + "step": 7074 + }, + { + "epoch": 1.6097838452787259, + "grad_norm": 1.6147782764940433, + "learning_rate": 9.567171725449635e-07, + "loss": 0.084, + "step": 7075 + }, + { + "epoch": 1.6100113765642776, + "grad_norm": 1.4354474070417866, + "learning_rate": 9.566414412244635e-07, + "loss": 0.0394, + "step": 7076 + }, + { + "epoch": 1.6102389078498294, + "grad_norm": 1.7446646500466492, + "learning_rate": 9.565657031258196e-07, + "loss": 0.0676, + "step": 7077 + }, + { + "epoch": 1.6104664391353811, + "grad_norm": 2.4698974949806565, + "learning_rate": 9.564899582505802e-07, + "loss": 0.0857, + "step": 7078 + }, + { + "epoch": 1.6106939704209329, + "grad_norm": 2.489527530624285, + "learning_rate": 9.56414206600293e-07, + "loss": 0.092, + "step": 7079 + }, + { + "epoch": 1.6109215017064846, + "grad_norm": 1.563596833456884, + "learning_rate": 9.563384481765064e-07, + "loss": 0.0412, + "step": 7080 + }, + { + "epoch": 1.6111490329920364, + "grad_norm": 2.5558659608398955, + "learning_rate": 9.562626829807689e-07, + "loss": 0.1769, + "step": 7081 + }, + { + "epoch": 1.6113765642775881, + "grad_norm": 1.8816520646545773, + "learning_rate": 9.561869110146288e-07, + "loss": 0.127, + "step": 7082 + }, + { + "epoch": 1.6116040955631399, + "grad_norm": 1.6643867089828204, + "learning_rate": 9.561111322796346e-07, + "loss": 0.0764, + "step": 7083 + }, + { + "epoch": 1.6118316268486916, + "grad_norm": 1.5341770274431337, + "learning_rate": 9.560353467773354e-07, + "loss": 0.054, + "step": 7084 + }, + { + "epoch": 1.6120591581342434, + "grad_norm": 2.2467041325463004, + "learning_rate": 9.5595955450928e-07, + "loss": 0.1787, + "step": 7085 + }, + { + "epoch": 1.6122866894197951, + "grad_norm": 2.4601590783176164, + "learning_rate": 9.558837554770173e-07, + "loss": 0.0728, + "step": 7086 + }, + { + "epoch": 1.612514220705347, + "grad_norm": 1.4520641015196234, + "learning_rate": 9.558079496820965e-07, + "loss": 0.093, + "step": 7087 + }, + { + "epoch": 1.6127417519908986, + "grad_norm": 1.2574529969406754, + "learning_rate": 9.557321371260675e-07, + "loss": 0.1536, + "step": 7088 + }, + { + "epoch": 1.6129692832764504, + "grad_norm": 4.775882995639071, + "learning_rate": 9.55656317810479e-07, + "loss": 0.0628, + "step": 7089 + }, + { + "epoch": 1.6131968145620021, + "grad_norm": 1.2500051050487235, + "learning_rate": 9.555804917368808e-07, + "loss": 0.1272, + "step": 7090 + }, + { + "epoch": 1.613424345847554, + "grad_norm": 1.3616233079512923, + "learning_rate": 9.55504658906823e-07, + "loss": 0.0395, + "step": 7091 + }, + { + "epoch": 1.6136518771331056, + "grad_norm": 1.4428559066579976, + "learning_rate": 9.554288193218552e-07, + "loss": 0.0638, + "step": 7092 + }, + { + "epoch": 1.6138794084186574, + "grad_norm": 2.157620490776601, + "learning_rate": 9.553529729835275e-07, + "loss": 0.1196, + "step": 7093 + }, + { + "epoch": 1.6141069397042094, + "grad_norm": 1.4309452846628332, + "learning_rate": 9.552771198933903e-07, + "loss": 0.117, + "step": 7094 + }, + { + "epoch": 1.6143344709897611, + "grad_norm": 1.3480658748775887, + "learning_rate": 9.552012600529934e-07, + "loss": 0.0449, + "step": 7095 + }, + { + "epoch": 1.6145620022753129, + "grad_norm": 1.171421665597667, + "learning_rate": 9.551253934638874e-07, + "loss": 0.0657, + "step": 7096 + }, + { + "epoch": 1.6147895335608646, + "grad_norm": 1.8892686669171361, + "learning_rate": 9.550495201276231e-07, + "loss": 0.0705, + "step": 7097 + }, + { + "epoch": 1.6150170648464164, + "grad_norm": 1.3180111718701897, + "learning_rate": 9.54973640045751e-07, + "loss": 0.071, + "step": 7098 + }, + { + "epoch": 1.6152445961319681, + "grad_norm": 2.238229572587678, + "learning_rate": 9.54897753219822e-07, + "loss": 0.1586, + "step": 7099 + }, + { + "epoch": 1.6154721274175199, + "grad_norm": 3.299443853412907, + "learning_rate": 9.548218596513871e-07, + "loss": 0.1078, + "step": 7100 + }, + { + "epoch": 1.6156996587030716, + "grad_norm": 1.5809997114750205, + "learning_rate": 9.547459593419975e-07, + "loss": 0.0908, + "step": 7101 + }, + { + "epoch": 1.6159271899886234, + "grad_norm": 1.5987662948298578, + "learning_rate": 9.546700522932042e-07, + "loss": 0.0794, + "step": 7102 + }, + { + "epoch": 1.6161547212741754, + "grad_norm": 1.6890028281579352, + "learning_rate": 9.54594138506559e-07, + "loss": 0.0551, + "step": 7103 + }, + { + "epoch": 1.6163822525597271, + "grad_norm": 1.132133276504026, + "learning_rate": 9.545182179836132e-07, + "loss": 0.0503, + "step": 7104 + }, + { + "epoch": 1.6166097838452789, + "grad_norm": 1.5304822279177308, + "learning_rate": 9.544422907259185e-07, + "loss": 0.0649, + "step": 7105 + }, + { + "epoch": 1.6168373151308306, + "grad_norm": 2.002856930667923, + "learning_rate": 9.543663567350267e-07, + "loss": 0.1061, + "step": 7106 + }, + { + "epoch": 1.6170648464163824, + "grad_norm": 1.2661716612665568, + "learning_rate": 9.542904160124896e-07, + "loss": 0.058, + "step": 7107 + }, + { + "epoch": 1.6172923777019341, + "grad_norm": 1.294677148035939, + "learning_rate": 9.542144685598598e-07, + "loss": 0.1243, + "step": 7108 + }, + { + "epoch": 1.6175199089874859, + "grad_norm": 1.537674166999775, + "learning_rate": 9.54138514378689e-07, + "loss": 0.1044, + "step": 7109 + }, + { + "epoch": 1.6177474402730376, + "grad_norm": 1.1942736822351498, + "learning_rate": 9.540625534705297e-07, + "loss": 0.0983, + "step": 7110 + }, + { + "epoch": 1.6179749715585894, + "grad_norm": 1.6860430625022544, + "learning_rate": 9.539865858369347e-07, + "loss": 0.074, + "step": 7111 + }, + { + "epoch": 1.6182025028441411, + "grad_norm": 2.7741366323041934, + "learning_rate": 9.539106114794564e-07, + "loss": 0.1305, + "step": 7112 + }, + { + "epoch": 1.6184300341296929, + "grad_norm": 1.5967947474659638, + "learning_rate": 9.538346303996472e-07, + "loss": 0.1135, + "step": 7113 + }, + { + "epoch": 1.6186575654152446, + "grad_norm": 1.9548245891512768, + "learning_rate": 9.537586425990604e-07, + "loss": 0.0773, + "step": 7114 + }, + { + "epoch": 1.6188850967007964, + "grad_norm": 2.2461488842288366, + "learning_rate": 9.536826480792493e-07, + "loss": 0.1846, + "step": 7115 + }, + { + "epoch": 1.6191126279863481, + "grad_norm": 1.9145559325768369, + "learning_rate": 9.53606646841767e-07, + "loss": 0.0559, + "step": 7116 + }, + { + "epoch": 1.6193401592719, + "grad_norm": 1.5353243471583073, + "learning_rate": 9.535306388881663e-07, + "loss": 0.1401, + "step": 7117 + }, + { + "epoch": 1.6195676905574516, + "grad_norm": 1.2965207126861003, + "learning_rate": 9.534546242200012e-07, + "loss": 0.0991, + "step": 7118 + }, + { + "epoch": 1.6197952218430034, + "grad_norm": 2.39739525863025, + "learning_rate": 9.53378602838825e-07, + "loss": 0.1789, + "step": 7119 + }, + { + "epoch": 1.6200227531285551, + "grad_norm": 2.3161329141037346, + "learning_rate": 9.533025747461916e-07, + "loss": 0.1097, + "step": 7120 + }, + { + "epoch": 1.620250284414107, + "grad_norm": 1.2769845925916703, + "learning_rate": 9.532265399436549e-07, + "loss": 0.0564, + "step": 7121 + }, + { + "epoch": 1.6204778156996587, + "grad_norm": 1.5848254675436142, + "learning_rate": 9.531504984327688e-07, + "loss": 0.0901, + "step": 7122 + }, + { + "epoch": 1.6207053469852104, + "grad_norm": 1.8035770243114464, + "learning_rate": 9.530744502150874e-07, + "loss": 0.089, + "step": 7123 + }, + { + "epoch": 1.6209328782707622, + "grad_norm": 2.0544769768196316, + "learning_rate": 9.529983952921652e-07, + "loss": 0.0696, + "step": 7124 + }, + { + "epoch": 1.621160409556314, + "grad_norm": 2.415000920373807, + "learning_rate": 9.529223336655565e-07, + "loss": 0.1083, + "step": 7125 + }, + { + "epoch": 1.6213879408418657, + "grad_norm": 1.7883493546660625, + "learning_rate": 9.528462653368158e-07, + "loss": 0.0644, + "step": 7126 + }, + { + "epoch": 1.6216154721274174, + "grad_norm": 0.9354416438272177, + "learning_rate": 9.527701903074977e-07, + "loss": 0.0451, + "step": 7127 + }, + { + "epoch": 1.6218430034129692, + "grad_norm": 1.9495521336227608, + "learning_rate": 9.526941085791574e-07, + "loss": 0.0657, + "step": 7128 + }, + { + "epoch": 1.622070534698521, + "grad_norm": 2.1414259183644795, + "learning_rate": 9.526180201533497e-07, + "loss": 0.1335, + "step": 7129 + }, + { + "epoch": 1.6222980659840727, + "grad_norm": 2.081666693341058, + "learning_rate": 9.525419250316295e-07, + "loss": 0.1073, + "step": 7130 + }, + { + "epoch": 1.6225255972696244, + "grad_norm": 1.298777524655821, + "learning_rate": 9.524658232155524e-07, + "loss": 0.0812, + "step": 7131 + }, + { + "epoch": 1.6227531285551762, + "grad_norm": 2.266832318052867, + "learning_rate": 9.523897147066735e-07, + "loss": 0.132, + "step": 7132 + }, + { + "epoch": 1.6229806598407281, + "grad_norm": 1.7503424667417475, + "learning_rate": 9.523135995065483e-07, + "loss": 0.1555, + "step": 7133 + }, + { + "epoch": 1.62320819112628, + "grad_norm": 1.4805919827253016, + "learning_rate": 9.522374776167328e-07, + "loss": 0.0802, + "step": 7134 + }, + { + "epoch": 1.6234357224118316, + "grad_norm": 1.3424048205748553, + "learning_rate": 9.521613490387824e-07, + "loss": 0.1015, + "step": 7135 + }, + { + "epoch": 1.6236632536973834, + "grad_norm": 1.8135164351802546, + "learning_rate": 9.520852137742534e-07, + "loss": 0.0866, + "step": 7136 + }, + { + "epoch": 1.6238907849829352, + "grad_norm": 2.0423720356000206, + "learning_rate": 9.520090718247016e-07, + "loss": 0.0555, + "step": 7137 + }, + { + "epoch": 1.624118316268487, + "grad_norm": 2.1346845240346686, + "learning_rate": 9.519329231916831e-07, + "loss": 0.1197, + "step": 7138 + }, + { + "epoch": 1.6243458475540387, + "grad_norm": 1.7824098626608995, + "learning_rate": 9.518567678767546e-07, + "loss": 0.0743, + "step": 7139 + }, + { + "epoch": 1.6245733788395904, + "grad_norm": 0.9942649821530546, + "learning_rate": 9.51780605881472e-07, + "loss": 0.0776, + "step": 7140 + }, + { + "epoch": 1.6248009101251422, + "grad_norm": 1.240154830498896, + "learning_rate": 9.517044372073927e-07, + "loss": 0.1216, + "step": 7141 + }, + { + "epoch": 1.6250284414106941, + "grad_norm": 1.4726407561890795, + "learning_rate": 9.516282618560727e-07, + "loss": 0.0645, + "step": 7142 + }, + { + "epoch": 1.6252559726962459, + "grad_norm": 1.4504003744629235, + "learning_rate": 9.515520798290695e-07, + "loss": 0.1242, + "step": 7143 + }, + { + "epoch": 1.6254835039817976, + "grad_norm": 1.9699918214488714, + "learning_rate": 9.514758911279398e-07, + "loss": 0.1662, + "step": 7144 + }, + { + "epoch": 1.6257110352673494, + "grad_norm": 2.0421004880361626, + "learning_rate": 9.513996957542407e-07, + "loss": 0.0782, + "step": 7145 + }, + { + "epoch": 1.6259385665529011, + "grad_norm": 1.0829443140069515, + "learning_rate": 9.513234937095296e-07, + "loss": 0.0864, + "step": 7146 + }, + { + "epoch": 1.626166097838453, + "grad_norm": 1.905465586127579, + "learning_rate": 9.512472849953639e-07, + "loss": 0.0566, + "step": 7147 + }, + { + "epoch": 1.6263936291240046, + "grad_norm": 1.6659429452735535, + "learning_rate": 9.511710696133012e-07, + "loss": 0.1531, + "step": 7148 + }, + { + "epoch": 1.6266211604095564, + "grad_norm": 1.128530637007164, + "learning_rate": 9.510948475648993e-07, + "loss": 0.0476, + "step": 7149 + }, + { + "epoch": 1.6268486916951082, + "grad_norm": 1.7859175113663965, + "learning_rate": 9.510186188517159e-07, + "loss": 0.084, + "step": 7150 + }, + { + "epoch": 1.62707622298066, + "grad_norm": 2.2211450363669547, + "learning_rate": 9.50942383475309e-07, + "loss": 0.1057, + "step": 7151 + }, + { + "epoch": 1.6273037542662117, + "grad_norm": 1.7188290771119823, + "learning_rate": 9.508661414372367e-07, + "loss": 0.0628, + "step": 7152 + }, + { + "epoch": 1.6275312855517634, + "grad_norm": 2.33047844943163, + "learning_rate": 9.507898927390571e-07, + "loss": 0.0941, + "step": 7153 + }, + { + "epoch": 1.6277588168373152, + "grad_norm": 1.9604490750514882, + "learning_rate": 9.507136373823289e-07, + "loss": 0.0727, + "step": 7154 + }, + { + "epoch": 1.627986348122867, + "grad_norm": 0.988973424139529, + "learning_rate": 9.506373753686104e-07, + "loss": 0.0795, + "step": 7155 + }, + { + "epoch": 1.6282138794084187, + "grad_norm": 1.7604699945221596, + "learning_rate": 9.505611066994605e-07, + "loss": 0.1175, + "step": 7156 + }, + { + "epoch": 1.6284414106939704, + "grad_norm": 2.014909081415411, + "learning_rate": 9.504848313764375e-07, + "loss": 0.0645, + "step": 7157 + }, + { + "epoch": 1.6286689419795222, + "grad_norm": 1.2736611112774021, + "learning_rate": 9.50408549401101e-07, + "loss": 0.1051, + "step": 7158 + }, + { + "epoch": 1.628896473265074, + "grad_norm": 1.2474664294183557, + "learning_rate": 9.503322607750096e-07, + "loss": 0.0771, + "step": 7159 + }, + { + "epoch": 1.6291240045506257, + "grad_norm": 1.2307537903220425, + "learning_rate": 9.502559654997226e-07, + "loss": 0.0545, + "step": 7160 + }, + { + "epoch": 1.6293515358361774, + "grad_norm": 1.6777703584025532, + "learning_rate": 9.501796635767992e-07, + "loss": 0.0956, + "step": 7161 + }, + { + "epoch": 1.6295790671217292, + "grad_norm": 3.3940231093078364, + "learning_rate": 9.501033550077993e-07, + "loss": 0.0723, + "step": 7162 + }, + { + "epoch": 1.629806598407281, + "grad_norm": 1.4027176613897487, + "learning_rate": 9.500270397942819e-07, + "loss": 0.1389, + "step": 7163 + }, + { + "epoch": 1.6300341296928327, + "grad_norm": 1.4510463438526642, + "learning_rate": 9.499507179378072e-07, + "loss": 0.0988, + "step": 7164 + }, + { + "epoch": 1.6302616609783844, + "grad_norm": 1.740850173674952, + "learning_rate": 9.498743894399348e-07, + "loss": 0.0695, + "step": 7165 + }, + { + "epoch": 1.6304891922639362, + "grad_norm": 1.147692771315242, + "learning_rate": 9.497980543022251e-07, + "loss": 0.0267, + "step": 7166 + }, + { + "epoch": 1.630716723549488, + "grad_norm": 1.3365492534347532, + "learning_rate": 9.497217125262378e-07, + "loss": 0.1274, + "step": 7167 + }, + { + "epoch": 1.6309442548350397, + "grad_norm": 1.6871961606908568, + "learning_rate": 9.496453641135337e-07, + "loss": 0.1147, + "step": 7168 + }, + { + "epoch": 1.6311717861205914, + "grad_norm": 2.1846968460485083, + "learning_rate": 9.495690090656728e-07, + "loss": 0.0436, + "step": 7169 + }, + { + "epoch": 1.6313993174061432, + "grad_norm": 1.358235913812706, + "learning_rate": 9.494926473842156e-07, + "loss": 0.0722, + "step": 7170 + }, + { + "epoch": 1.631626848691695, + "grad_norm": 1.120586198535551, + "learning_rate": 9.494162790707232e-07, + "loss": 0.0273, + "step": 7171 + }, + { + "epoch": 1.631854379977247, + "grad_norm": 1.5332225602063558, + "learning_rate": 9.493399041267559e-07, + "loss": 0.043, + "step": 7172 + }, + { + "epoch": 1.6320819112627987, + "grad_norm": 2.609985215164489, + "learning_rate": 9.492635225538751e-07, + "loss": 0.1224, + "step": 7173 + }, + { + "epoch": 1.6323094425483504, + "grad_norm": 1.2149222886137578, + "learning_rate": 9.491871343536418e-07, + "loss": 0.0824, + "step": 7174 + }, + { + "epoch": 1.6325369738339022, + "grad_norm": 2.0312203841057213, + "learning_rate": 9.491107395276172e-07, + "loss": 0.0969, + "step": 7175 + }, + { + "epoch": 1.632764505119454, + "grad_norm": 2.3293223801623895, + "learning_rate": 9.490343380773629e-07, + "loss": 0.108, + "step": 7176 + }, + { + "epoch": 1.6329920364050057, + "grad_norm": 2.367637061172044, + "learning_rate": 9.489579300044396e-07, + "loss": 0.0863, + "step": 7177 + }, + { + "epoch": 1.6332195676905574, + "grad_norm": 1.5769317594673877, + "learning_rate": 9.4888151531041e-07, + "loss": 0.0798, + "step": 7178 + }, + { + "epoch": 1.6334470989761092, + "grad_norm": 1.4455661912823792, + "learning_rate": 9.488050939968351e-07, + "loss": 0.1161, + "step": 7179 + }, + { + "epoch": 1.633674630261661, + "grad_norm": 2.8524695247450853, + "learning_rate": 9.487286660652773e-07, + "loss": 0.077, + "step": 7180 + }, + { + "epoch": 1.633902161547213, + "grad_norm": 2.7487698553006075, + "learning_rate": 9.486522315172983e-07, + "loss": 0.1453, + "step": 7181 + }, + { + "epoch": 1.6341296928327647, + "grad_norm": 1.4587274045828376, + "learning_rate": 9.485757903544606e-07, + "loss": 0.0651, + "step": 7182 + }, + { + "epoch": 1.6343572241183164, + "grad_norm": 2.04298506693863, + "learning_rate": 9.484993425783262e-07, + "loss": 0.0962, + "step": 7183 + }, + { + "epoch": 1.6345847554038682, + "grad_norm": 1.7933342325194643, + "learning_rate": 9.484228881904577e-07, + "loss": 0.1191, + "step": 7184 + }, + { + "epoch": 1.63481228668942, + "grad_norm": 2.2774414909954506, + "learning_rate": 9.483464271924177e-07, + "loss": 0.0589, + "step": 7185 + }, + { + "epoch": 1.6350398179749717, + "grad_norm": 1.4174058926031983, + "learning_rate": 9.482699595857689e-07, + "loss": 0.0845, + "step": 7186 + }, + { + "epoch": 1.6352673492605234, + "grad_norm": 1.2239168361023993, + "learning_rate": 9.481934853720742e-07, + "loss": 0.0601, + "step": 7187 + }, + { + "epoch": 1.6354948805460752, + "grad_norm": 3.0956636955260706, + "learning_rate": 9.481170045528968e-07, + "loss": 0.1153, + "step": 7188 + }, + { + "epoch": 1.635722411831627, + "grad_norm": 1.9187976135861455, + "learning_rate": 9.480405171297992e-07, + "loss": 0.0955, + "step": 7189 + }, + { + "epoch": 1.6359499431171787, + "grad_norm": 1.4924440579295828, + "learning_rate": 9.479640231043453e-07, + "loss": 0.0999, + "step": 7190 + }, + { + "epoch": 1.6361774744027304, + "grad_norm": 1.0007987751530008, + "learning_rate": 9.47887522478098e-07, + "loss": 0.0598, + "step": 7191 + }, + { + "epoch": 1.6364050056882822, + "grad_norm": 2.3399915938395885, + "learning_rate": 9.478110152526212e-07, + "loss": 0.0647, + "step": 7192 + }, + { + "epoch": 1.636632536973834, + "grad_norm": 1.305800819783954, + "learning_rate": 9.477345014294787e-07, + "loss": 0.0852, + "step": 7193 + }, + { + "epoch": 1.6368600682593857, + "grad_norm": 1.6920843391163056, + "learning_rate": 9.476579810102337e-07, + "loss": 0.1114, + "step": 7194 + }, + { + "epoch": 1.6370875995449374, + "grad_norm": 2.1814109460896094, + "learning_rate": 9.475814539964506e-07, + "loss": 0.1856, + "step": 7195 + }, + { + "epoch": 1.6373151308304892, + "grad_norm": 1.7859494589363305, + "learning_rate": 9.475049203896934e-07, + "loss": 0.1159, + "step": 7196 + }, + { + "epoch": 1.637542662116041, + "grad_norm": 1.8121948843839317, + "learning_rate": 9.474283801915262e-07, + "loss": 0.0515, + "step": 7197 + }, + { + "epoch": 1.6377701934015927, + "grad_norm": 1.4848641666208762, + "learning_rate": 9.473518334035134e-07, + "loss": 0.0964, + "step": 7198 + }, + { + "epoch": 1.6379977246871444, + "grad_norm": 2.035419836700452, + "learning_rate": 9.472752800272194e-07, + "loss": 0.0666, + "step": 7199 + }, + { + "epoch": 1.6382252559726962, + "grad_norm": 1.5897835103223596, + "learning_rate": 9.471987200642093e-07, + "loss": 0.1136, + "step": 7200 + }, + { + "epoch": 1.638452787258248, + "grad_norm": 1.1177849069853134, + "learning_rate": 9.471221535160471e-07, + "loss": 0.0528, + "step": 7201 + }, + { + "epoch": 1.6386803185437997, + "grad_norm": 1.4149779560776885, + "learning_rate": 9.470455803842982e-07, + "loss": 0.0672, + "step": 7202 + }, + { + "epoch": 1.6389078498293514, + "grad_norm": 2.1597142960821345, + "learning_rate": 9.469690006705274e-07, + "loss": 0.0728, + "step": 7203 + }, + { + "epoch": 1.6391353811149032, + "grad_norm": 1.573868562708633, + "learning_rate": 9.468924143762996e-07, + "loss": 0.0609, + "step": 7204 + }, + { + "epoch": 1.639362912400455, + "grad_norm": 1.3969158589202828, + "learning_rate": 9.468158215031805e-07, + "loss": 0.0595, + "step": 7205 + }, + { + "epoch": 1.6395904436860067, + "grad_norm": 1.9871624130035885, + "learning_rate": 9.467392220527358e-07, + "loss": 0.1169, + "step": 7206 + }, + { + "epoch": 1.6398179749715585, + "grad_norm": 1.6143675451430033, + "learning_rate": 9.4666261602653e-07, + "loss": 0.1121, + "step": 7207 + }, + { + "epoch": 1.6400455062571102, + "grad_norm": 2.4786692040030887, + "learning_rate": 9.465860034261298e-07, + "loss": 0.062, + "step": 7208 + }, + { + "epoch": 1.640273037542662, + "grad_norm": 1.651797912353839, + "learning_rate": 9.465093842531007e-07, + "loss": 0.0734, + "step": 7209 + }, + { + "epoch": 1.640500568828214, + "grad_norm": 1.4933908025967075, + "learning_rate": 9.464327585090084e-07, + "loss": 0.0681, + "step": 7210 + }, + { + "epoch": 1.6407281001137657, + "grad_norm": 2.46724683222329, + "learning_rate": 9.463561261954192e-07, + "loss": 0.129, + "step": 7211 + }, + { + "epoch": 1.6409556313993174, + "grad_norm": 1.4709863716034994, + "learning_rate": 9.462794873138995e-07, + "loss": 0.0914, + "step": 7212 + }, + { + "epoch": 1.6411831626848692, + "grad_norm": 3.2781700122611643, + "learning_rate": 9.462028418660155e-07, + "loss": 0.154, + "step": 7213 + }, + { + "epoch": 1.641410693970421, + "grad_norm": 1.2959743830109027, + "learning_rate": 9.461261898533333e-07, + "loss": 0.0627, + "step": 7214 + }, + { + "epoch": 1.6416382252559727, + "grad_norm": 1.9860376233888792, + "learning_rate": 9.460495312774203e-07, + "loss": 0.169, + "step": 7215 + }, + { + "epoch": 1.6418657565415244, + "grad_norm": 2.818835837267242, + "learning_rate": 9.459728661398427e-07, + "loss": 0.0847, + "step": 7216 + }, + { + "epoch": 1.6420932878270762, + "grad_norm": 1.389674709889107, + "learning_rate": 9.458961944421674e-07, + "loss": 0.085, + "step": 7217 + }, + { + "epoch": 1.642320819112628, + "grad_norm": 1.876929962467865, + "learning_rate": 9.458195161859618e-07, + "loss": 0.0596, + "step": 7218 + }, + { + "epoch": 1.6425483503981797, + "grad_norm": 1.786147919520637, + "learning_rate": 9.457428313727927e-07, + "loss": 0.089, + "step": 7219 + }, + { + "epoch": 1.6427758816837317, + "grad_norm": 1.3401964491024454, + "learning_rate": 9.456661400042278e-07, + "loss": 0.1182, + "step": 7220 + }, + { + "epoch": 1.6430034129692834, + "grad_norm": 1.5281082687017675, + "learning_rate": 9.455894420818339e-07, + "loss": 0.1498, + "step": 7221 + }, + { + "epoch": 1.6432309442548352, + "grad_norm": 2.235395462768349, + "learning_rate": 9.455127376071791e-07, + "loss": 0.1527, + "step": 7222 + }, + { + "epoch": 1.643458475540387, + "grad_norm": 1.6316478922952569, + "learning_rate": 9.454360265818309e-07, + "loss": 0.0635, + "step": 7223 + }, + { + "epoch": 1.6436860068259387, + "grad_norm": 2.0316717048016084, + "learning_rate": 9.453593090073571e-07, + "loss": 0.1008, + "step": 7224 + }, + { + "epoch": 1.6439135381114904, + "grad_norm": 1.8222797124219188, + "learning_rate": 9.45282584885326e-07, + "loss": 0.1174, + "step": 7225 + }, + { + "epoch": 1.6441410693970422, + "grad_norm": 1.4162521237098797, + "learning_rate": 9.452058542173054e-07, + "loss": 0.0531, + "step": 7226 + }, + { + "epoch": 1.644368600682594, + "grad_norm": 1.4806445425094952, + "learning_rate": 9.451291170048632e-07, + "loss": 0.0976, + "step": 7227 + }, + { + "epoch": 1.6445961319681457, + "grad_norm": 2.04356178536637, + "learning_rate": 9.450523732495684e-07, + "loss": 0.1338, + "step": 7228 + }, + { + "epoch": 1.6448236632536974, + "grad_norm": 1.592647023562932, + "learning_rate": 9.449756229529893e-07, + "loss": 0.0617, + "step": 7229 + }, + { + "epoch": 1.6450511945392492, + "grad_norm": 1.2827950381449074, + "learning_rate": 9.448988661166944e-07, + "loss": 0.0932, + "step": 7230 + }, + { + "epoch": 1.645278725824801, + "grad_norm": 1.9362527509370229, + "learning_rate": 9.448221027422525e-07, + "loss": 0.088, + "step": 7231 + }, + { + "epoch": 1.6455062571103527, + "grad_norm": 2.150147523402948, + "learning_rate": 9.447453328312325e-07, + "loss": 0.0864, + "step": 7232 + }, + { + "epoch": 1.6457337883959045, + "grad_norm": 1.436369267098461, + "learning_rate": 9.446685563852037e-07, + "loss": 0.1491, + "step": 7233 + }, + { + "epoch": 1.6459613196814562, + "grad_norm": 2.431568244659582, + "learning_rate": 9.445917734057349e-07, + "loss": 0.1234, + "step": 7234 + }, + { + "epoch": 1.646188850967008, + "grad_norm": 2.398693177911805, + "learning_rate": 9.445149838943955e-07, + "loss": 0.117, + "step": 7235 + }, + { + "epoch": 1.6464163822525597, + "grad_norm": 1.3252800678869499, + "learning_rate": 9.44438187852755e-07, + "loss": 0.0532, + "step": 7236 + }, + { + "epoch": 1.6466439135381115, + "grad_norm": 1.6130505488960083, + "learning_rate": 9.443613852823832e-07, + "loss": 0.0498, + "step": 7237 + }, + { + "epoch": 1.6468714448236632, + "grad_norm": 1.5445975785219161, + "learning_rate": 9.442845761848493e-07, + "loss": 0.0905, + "step": 7238 + }, + { + "epoch": 1.647098976109215, + "grad_norm": 1.6214589044623604, + "learning_rate": 9.442077605617236e-07, + "loss": 0.1482, + "step": 7239 + }, + { + "epoch": 1.6473265073947667, + "grad_norm": 2.5934304521178926, + "learning_rate": 9.441309384145758e-07, + "loss": 0.1356, + "step": 7240 + }, + { + "epoch": 1.6475540386803185, + "grad_norm": 2.1821198275071145, + "learning_rate": 9.440541097449759e-07, + "loss": 0.1202, + "step": 7241 + }, + { + "epoch": 1.6477815699658702, + "grad_norm": 1.3783319298593066, + "learning_rate": 9.439772745544945e-07, + "loss": 0.0809, + "step": 7242 + }, + { + "epoch": 1.648009101251422, + "grad_norm": 1.9235495347033278, + "learning_rate": 9.439004328447019e-07, + "loss": 0.0793, + "step": 7243 + }, + { + "epoch": 1.6482366325369737, + "grad_norm": 1.5695626343986404, + "learning_rate": 9.438235846171684e-07, + "loss": 0.073, + "step": 7244 + }, + { + "epoch": 1.6484641638225255, + "grad_norm": 2.3576206439809275, + "learning_rate": 9.437467298734646e-07, + "loss": 0.1557, + "step": 7245 + }, + { + "epoch": 1.6486916951080772, + "grad_norm": 1.4663932891903617, + "learning_rate": 9.436698686151616e-07, + "loss": 0.1639, + "step": 7246 + }, + { + "epoch": 1.648919226393629, + "grad_norm": 1.2759564481247696, + "learning_rate": 9.435930008438299e-07, + "loss": 0.0874, + "step": 7247 + }, + { + "epoch": 1.6491467576791807, + "grad_norm": 3.0126912789420293, + "learning_rate": 9.435161265610407e-07, + "loss": 0.1002, + "step": 7248 + }, + { + "epoch": 1.6493742889647327, + "grad_norm": 3.268767822166679, + "learning_rate": 9.434392457683653e-07, + "loss": 0.0983, + "step": 7249 + }, + { + "epoch": 1.6496018202502845, + "grad_norm": 1.6367805701152656, + "learning_rate": 9.433623584673751e-07, + "loss": 0.0473, + "step": 7250 + }, + { + "epoch": 1.6498293515358362, + "grad_norm": 1.7890854849904998, + "learning_rate": 9.432854646596412e-07, + "loss": 0.0534, + "step": 7251 + }, + { + "epoch": 1.650056882821388, + "grad_norm": 1.3726332380580277, + "learning_rate": 9.432085643467352e-07, + "loss": 0.1425, + "step": 7252 + }, + { + "epoch": 1.6502844141069397, + "grad_norm": 1.9235963258967004, + "learning_rate": 9.43131657530229e-07, + "loss": 0.1119, + "step": 7253 + }, + { + "epoch": 1.6505119453924915, + "grad_norm": 2.962111856390726, + "learning_rate": 9.430547442116944e-07, + "loss": 0.0984, + "step": 7254 + }, + { + "epoch": 1.6507394766780432, + "grad_norm": 2.1556391413872777, + "learning_rate": 9.429778243927031e-07, + "loss": 0.1026, + "step": 7255 + }, + { + "epoch": 1.650967007963595, + "grad_norm": 2.341832429516187, + "learning_rate": 9.429008980748279e-07, + "loss": 0.1612, + "step": 7256 + }, + { + "epoch": 1.6511945392491467, + "grad_norm": 2.6524619584140723, + "learning_rate": 9.428239652596402e-07, + "loss": 0.1038, + "step": 7257 + }, + { + "epoch": 1.6514220705346985, + "grad_norm": 1.8037046803671937, + "learning_rate": 9.427470259487127e-07, + "loss": 0.0648, + "step": 7258 + }, + { + "epoch": 1.6516496018202504, + "grad_norm": 1.3920666963111505, + "learning_rate": 9.426700801436181e-07, + "loss": 0.0887, + "step": 7259 + }, + { + "epoch": 1.6518771331058022, + "grad_norm": 2.6323345473885693, + "learning_rate": 9.425931278459287e-07, + "loss": 0.1154, + "step": 7260 + }, + { + "epoch": 1.652104664391354, + "grad_norm": 1.235526720771683, + "learning_rate": 9.425161690572174e-07, + "loss": 0.0673, + "step": 7261 + }, + { + "epoch": 1.6523321956769057, + "grad_norm": 1.2599138951236994, + "learning_rate": 9.42439203779057e-07, + "loss": 0.0922, + "step": 7262 + }, + { + "epoch": 1.6525597269624575, + "grad_norm": 2.8618213081398993, + "learning_rate": 9.42362232013021e-07, + "loss": 0.1017, + "step": 7263 + }, + { + "epoch": 1.6527872582480092, + "grad_norm": 1.351501595228745, + "learning_rate": 9.422852537606819e-07, + "loss": 0.0715, + "step": 7264 + }, + { + "epoch": 1.653014789533561, + "grad_norm": 1.3482567491714648, + "learning_rate": 9.422082690236134e-07, + "loss": 0.0969, + "step": 7265 + }, + { + "epoch": 1.6532423208191127, + "grad_norm": 1.6556273596729216, + "learning_rate": 9.421312778033889e-07, + "loss": 0.0983, + "step": 7266 + }, + { + "epoch": 1.6534698521046645, + "grad_norm": 2.480720701978015, + "learning_rate": 9.420542801015817e-07, + "loss": 0.0726, + "step": 7267 + }, + { + "epoch": 1.6536973833902162, + "grad_norm": 1.360633062993953, + "learning_rate": 9.419772759197656e-07, + "loss": 0.0536, + "step": 7268 + }, + { + "epoch": 1.653924914675768, + "grad_norm": 2.1757990592385634, + "learning_rate": 9.41900265259515e-07, + "loss": 0.0717, + "step": 7269 + }, + { + "epoch": 1.6541524459613197, + "grad_norm": 2.2391168303019064, + "learning_rate": 9.418232481224029e-07, + "loss": 0.0974, + "step": 7270 + }, + { + "epoch": 1.6543799772468715, + "grad_norm": 1.371405310607541, + "learning_rate": 9.417462245100038e-07, + "loss": 0.1253, + "step": 7271 + }, + { + "epoch": 1.6546075085324232, + "grad_norm": 2.1960817643284862, + "learning_rate": 9.416691944238922e-07, + "loss": 0.1268, + "step": 7272 + }, + { + "epoch": 1.654835039817975, + "grad_norm": 1.6102464810078585, + "learning_rate": 9.415921578656422e-07, + "loss": 0.077, + "step": 7273 + }, + { + "epoch": 1.6550625711035267, + "grad_norm": 1.7189966445140032, + "learning_rate": 9.415151148368282e-07, + "loss": 0.0674, + "step": 7274 + }, + { + "epoch": 1.6552901023890785, + "grad_norm": 2.5040526861285923, + "learning_rate": 9.41438065339025e-07, + "loss": 0.0667, + "step": 7275 + }, + { + "epoch": 1.6555176336746302, + "grad_norm": 1.5933575864388763, + "learning_rate": 9.413610093738072e-07, + "loss": 0.2265, + "step": 7276 + }, + { + "epoch": 1.655745164960182, + "grad_norm": 2.086475507868105, + "learning_rate": 9.412839469427498e-07, + "loss": 0.0996, + "step": 7277 + }, + { + "epoch": 1.6559726962457337, + "grad_norm": 1.5557715465268633, + "learning_rate": 9.412068780474277e-07, + "loss": 0.0458, + "step": 7278 + }, + { + "epoch": 1.6562002275312855, + "grad_norm": 1.6837503079650977, + "learning_rate": 9.411298026894161e-07, + "loss": 0.1504, + "step": 7279 + }, + { + "epoch": 1.6564277588168372, + "grad_norm": 2.6116115808263762, + "learning_rate": 9.410527208702905e-07, + "loss": 0.1527, + "step": 7280 + }, + { + "epoch": 1.656655290102389, + "grad_norm": 2.393680145786846, + "learning_rate": 9.409756325916259e-07, + "loss": 0.0633, + "step": 7281 + }, + { + "epoch": 1.6568828213879407, + "grad_norm": 1.5217677015722841, + "learning_rate": 9.408985378549981e-07, + "loss": 0.0723, + "step": 7282 + }, + { + "epoch": 1.6571103526734925, + "grad_norm": 1.637542506575522, + "learning_rate": 9.408214366619829e-07, + "loss": 0.0583, + "step": 7283 + }, + { + "epoch": 1.6573378839590442, + "grad_norm": 3.660853802611821, + "learning_rate": 9.407443290141557e-07, + "loss": 0.0881, + "step": 7284 + }, + { + "epoch": 1.657565415244596, + "grad_norm": 1.9416812297062425, + "learning_rate": 9.406672149130928e-07, + "loss": 0.1396, + "step": 7285 + }, + { + "epoch": 1.6577929465301477, + "grad_norm": 1.4345822033154445, + "learning_rate": 9.405900943603702e-07, + "loss": 0.053, + "step": 7286 + }, + { + "epoch": 1.6580204778156995, + "grad_norm": 2.4200211765437682, + "learning_rate": 9.405129673575639e-07, + "loss": 0.1002, + "step": 7287 + }, + { + "epoch": 1.6582480091012515, + "grad_norm": 1.3927335425180136, + "learning_rate": 9.404358339062505e-07, + "loss": 0.1223, + "step": 7288 + }, + { + "epoch": 1.6584755403868032, + "grad_norm": 1.4509667046510046, + "learning_rate": 9.403586940080063e-07, + "loss": 0.0495, + "step": 7289 + }, + { + "epoch": 1.658703071672355, + "grad_norm": 1.3336177413806338, + "learning_rate": 9.40281547664408e-07, + "loss": 0.078, + "step": 7290 + }, + { + "epoch": 1.6589306029579067, + "grad_norm": 1.5111731661957328, + "learning_rate": 9.402043948770321e-07, + "loss": 0.1164, + "step": 7291 + }, + { + "epoch": 1.6591581342434585, + "grad_norm": 1.614767504803358, + "learning_rate": 9.401272356474557e-07, + "loss": 0.0956, + "step": 7292 + }, + { + "epoch": 1.6593856655290102, + "grad_norm": 1.9721878434789486, + "learning_rate": 9.400500699772558e-07, + "loss": 0.1389, + "step": 7293 + }, + { + "epoch": 1.659613196814562, + "grad_norm": 2.342810678020302, + "learning_rate": 9.399728978680094e-07, + "loss": 0.0946, + "step": 7294 + }, + { + "epoch": 1.6598407281001137, + "grad_norm": 2.2416455071278922, + "learning_rate": 9.39895719321294e-07, + "loss": 0.1539, + "step": 7295 + }, + { + "epoch": 1.6600682593856655, + "grad_norm": 1.5789748727066666, + "learning_rate": 9.398185343386866e-07, + "loss": 0.0489, + "step": 7296 + }, + { + "epoch": 1.6602957906712172, + "grad_norm": 1.6124613457863823, + "learning_rate": 9.39741342921765e-07, + "loss": 0.1324, + "step": 7297 + }, + { + "epoch": 1.6605233219567692, + "grad_norm": 2.159172767256166, + "learning_rate": 9.396641450721067e-07, + "loss": 0.0725, + "step": 7298 + }, + { + "epoch": 1.660750853242321, + "grad_norm": 2.003015363340268, + "learning_rate": 9.395869407912897e-07, + "loss": 0.1397, + "step": 7299 + }, + { + "epoch": 1.6609783845278727, + "grad_norm": 1.3324354785637096, + "learning_rate": 9.395097300808916e-07, + "loss": 0.0889, + "step": 7300 + }, + { + "epoch": 1.6612059158134245, + "grad_norm": 3.478146164044341, + "learning_rate": 9.394325129424906e-07, + "loss": 0.1451, + "step": 7301 + }, + { + "epoch": 1.6614334470989762, + "grad_norm": 2.0254463245059036, + "learning_rate": 9.39355289377665e-07, + "loss": 0.1064, + "step": 7302 + }, + { + "epoch": 1.661660978384528, + "grad_norm": 1.6595497039564182, + "learning_rate": 9.392780593879932e-07, + "loss": 0.0654, + "step": 7303 + }, + { + "epoch": 1.6618885096700797, + "grad_norm": 2.7480531876747705, + "learning_rate": 9.392008229750533e-07, + "loss": 0.1202, + "step": 7304 + }, + { + "epoch": 1.6621160409556315, + "grad_norm": 1.2471523479013982, + "learning_rate": 9.391235801404236e-07, + "loss": 0.0977, + "step": 7305 + }, + { + "epoch": 1.6623435722411832, + "grad_norm": 2.0642567756103642, + "learning_rate": 9.390463308856837e-07, + "loss": 0.1131, + "step": 7306 + }, + { + "epoch": 1.662571103526735, + "grad_norm": 1.2924016185822214, + "learning_rate": 9.389690752124118e-07, + "loss": 0.0654, + "step": 7307 + }, + { + "epoch": 1.6627986348122867, + "grad_norm": 0.8691007635102995, + "learning_rate": 9.388918131221869e-07, + "loss": 0.0454, + "step": 7308 + }, + { + "epoch": 1.6630261660978385, + "grad_norm": 1.2818332030580666, + "learning_rate": 9.388145446165884e-07, + "loss": 0.056, + "step": 7309 + }, + { + "epoch": 1.6632536973833902, + "grad_norm": 2.6473266198798937, + "learning_rate": 9.387372696971952e-07, + "loss": 0.0676, + "step": 7310 + }, + { + "epoch": 1.663481228668942, + "grad_norm": 1.721467410153477, + "learning_rate": 9.386599883655869e-07, + "loss": 0.0943, + "step": 7311 + }, + { + "epoch": 1.6637087599544937, + "grad_norm": 1.6056959376364102, + "learning_rate": 9.385827006233426e-07, + "loss": 0.0786, + "step": 7312 + }, + { + "epoch": 1.6639362912400455, + "grad_norm": 1.756717640480748, + "learning_rate": 9.385054064720425e-07, + "loss": 0.0915, + "step": 7313 + }, + { + "epoch": 1.6641638225255972, + "grad_norm": 2.2466102544864404, + "learning_rate": 9.38428105913266e-07, + "loss": 0.0912, + "step": 7314 + }, + { + "epoch": 1.664391353811149, + "grad_norm": 1.5791598649367002, + "learning_rate": 9.38350798948593e-07, + "loss": 0.1095, + "step": 7315 + }, + { + "epoch": 1.6646188850967008, + "grad_norm": 1.3715623153388081, + "learning_rate": 9.382734855796036e-07, + "loss": 0.0893, + "step": 7316 + }, + { + "epoch": 1.6648464163822525, + "grad_norm": 2.0142585633039354, + "learning_rate": 9.381961658078778e-07, + "loss": 0.1221, + "step": 7317 + }, + { + "epoch": 1.6650739476678043, + "grad_norm": 1.5505453307871044, + "learning_rate": 9.381188396349958e-07, + "loss": 0.0655, + "step": 7318 + }, + { + "epoch": 1.665301478953356, + "grad_norm": 2.0244620285601327, + "learning_rate": 9.380415070625384e-07, + "loss": 0.0513, + "step": 7319 + }, + { + "epoch": 1.6655290102389078, + "grad_norm": 1.7938433474249436, + "learning_rate": 9.379641680920859e-07, + "loss": 0.1331, + "step": 7320 + }, + { + "epoch": 1.6657565415244595, + "grad_norm": 1.3409299619748931, + "learning_rate": 9.378868227252188e-07, + "loss": 0.0918, + "step": 7321 + }, + { + "epoch": 1.6659840728100113, + "grad_norm": 1.4700377964483387, + "learning_rate": 9.378094709635183e-07, + "loss": 0.0877, + "step": 7322 + }, + { + "epoch": 1.666211604095563, + "grad_norm": 2.2568506457344193, + "learning_rate": 9.377321128085651e-07, + "loss": 0.0944, + "step": 7323 + }, + { + "epoch": 1.6664391353811148, + "grad_norm": 0.9908159172763481, + "learning_rate": 9.3765474826194e-07, + "loss": 0.0691, + "step": 7324 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 1.2584129227145304, + "learning_rate": 9.375773773252248e-07, + "loss": 0.0852, + "step": 7325 + }, + { + "epoch": 1.6668941979522183, + "grad_norm": 2.003696311286243, + "learning_rate": 9.375000000000001e-07, + "loss": 0.0563, + "step": 7326 + }, + { + "epoch": 1.6671217292377702, + "grad_norm": 3.3423358537743937, + "learning_rate": 9.374226162878478e-07, + "loss": 0.1533, + "step": 7327 + }, + { + "epoch": 1.667349260523322, + "grad_norm": 1.9621823491467638, + "learning_rate": 9.373452261903495e-07, + "loss": 0.0976, + "step": 7328 + }, + { + "epoch": 1.6675767918088737, + "grad_norm": 1.4020914530896604, + "learning_rate": 9.372678297090867e-07, + "loss": 0.0445, + "step": 7329 + }, + { + "epoch": 1.6678043230944255, + "grad_norm": 2.0898847741309274, + "learning_rate": 9.371904268456415e-07, + "loss": 0.1187, + "step": 7330 + }, + { + "epoch": 1.6680318543799773, + "grad_norm": 0.8770107550240693, + "learning_rate": 9.371130176015958e-07, + "loss": 0.0985, + "step": 7331 + }, + { + "epoch": 1.668259385665529, + "grad_norm": 1.859585957876411, + "learning_rate": 9.370356019785315e-07, + "loss": 0.1819, + "step": 7332 + }, + { + "epoch": 1.6684869169510808, + "grad_norm": 1.7703722436235858, + "learning_rate": 9.369581799780308e-07, + "loss": 0.0685, + "step": 7333 + }, + { + "epoch": 1.6687144482366325, + "grad_norm": 1.4134381222208843, + "learning_rate": 9.368807516016764e-07, + "loss": 0.0582, + "step": 7334 + }, + { + "epoch": 1.6689419795221843, + "grad_norm": 1.1064915462918847, + "learning_rate": 9.368033168510506e-07, + "loss": 0.051, + "step": 7335 + }, + { + "epoch": 1.669169510807736, + "grad_norm": 1.2109521599260762, + "learning_rate": 9.36725875727736e-07, + "loss": 0.1537, + "step": 7336 + }, + { + "epoch": 1.669397042093288, + "grad_norm": 2.3473344419643385, + "learning_rate": 9.366484282333155e-07, + "loss": 0.0998, + "step": 7337 + }, + { + "epoch": 1.6696245733788397, + "grad_norm": 2.038359008920555, + "learning_rate": 9.365709743693718e-07, + "loss": 0.064, + "step": 7338 + }, + { + "epoch": 1.6698521046643915, + "grad_norm": 2.3072660154667686, + "learning_rate": 9.364935141374881e-07, + "loss": 0.12, + "step": 7339 + }, + { + "epoch": 1.6700796359499432, + "grad_norm": 1.2461166565952482, + "learning_rate": 9.364160475392472e-07, + "loss": 0.0792, + "step": 7340 + }, + { + "epoch": 1.670307167235495, + "grad_norm": 1.682321979793571, + "learning_rate": 9.363385745762329e-07, + "loss": 0.0823, + "step": 7341 + }, + { + "epoch": 1.6705346985210467, + "grad_norm": 1.9310846458137203, + "learning_rate": 9.362610952500281e-07, + "loss": 0.0765, + "step": 7342 + }, + { + "epoch": 1.6707622298065985, + "grad_norm": 1.2871026073090324, + "learning_rate": 9.361836095622167e-07, + "loss": 0.095, + "step": 7343 + }, + { + "epoch": 1.6709897610921502, + "grad_norm": 1.8987569542659957, + "learning_rate": 9.361061175143823e-07, + "loss": 0.1196, + "step": 7344 + }, + { + "epoch": 1.671217292377702, + "grad_norm": 1.515884398351274, + "learning_rate": 9.360286191081085e-07, + "loss": 0.1374, + "step": 7345 + }, + { + "epoch": 1.6714448236632538, + "grad_norm": 1.7492424501998036, + "learning_rate": 9.359511143449794e-07, + "loss": 0.0774, + "step": 7346 + }, + { + "epoch": 1.6716723549488055, + "grad_norm": 0.9697929852387172, + "learning_rate": 9.358736032265788e-07, + "loss": 0.0666, + "step": 7347 + }, + { + "epoch": 1.6718998862343573, + "grad_norm": 2.5780166734661876, + "learning_rate": 9.357960857544912e-07, + "loss": 0.1432, + "step": 7348 + }, + { + "epoch": 1.672127417519909, + "grad_norm": 2.0838034914374783, + "learning_rate": 9.357185619303009e-07, + "loss": 0.0782, + "step": 7349 + }, + { + "epoch": 1.6723549488054608, + "grad_norm": 1.3919641574753976, + "learning_rate": 9.356410317555922e-07, + "loss": 0.1299, + "step": 7350 + }, + { + "epoch": 1.6725824800910125, + "grad_norm": 1.9809306658771628, + "learning_rate": 9.355634952319498e-07, + "loss": 0.0992, + "step": 7351 + }, + { + "epoch": 1.6728100113765643, + "grad_norm": 2.3969878568860263, + "learning_rate": 9.354859523609583e-07, + "loss": 0.1136, + "step": 7352 + }, + { + "epoch": 1.673037542662116, + "grad_norm": 1.1943562344041239, + "learning_rate": 9.354084031442027e-07, + "loss": 0.0872, + "step": 7353 + }, + { + "epoch": 1.6732650739476678, + "grad_norm": 1.3707784823503284, + "learning_rate": 9.353308475832676e-07, + "loss": 0.0509, + "step": 7354 + }, + { + "epoch": 1.6734926052332195, + "grad_norm": 1.239340089454578, + "learning_rate": 9.352532856797382e-07, + "loss": 0.0434, + "step": 7355 + }, + { + "epoch": 1.6737201365187713, + "grad_norm": 2.0296765650090456, + "learning_rate": 9.351757174352e-07, + "loss": 0.0672, + "step": 7356 + }, + { + "epoch": 1.673947667804323, + "grad_norm": 1.404142150592023, + "learning_rate": 9.350981428512383e-07, + "loss": 0.042, + "step": 7357 + }, + { + "epoch": 1.6741751990898748, + "grad_norm": 1.7723519707770459, + "learning_rate": 9.350205619294383e-07, + "loss": 0.1413, + "step": 7358 + }, + { + "epoch": 1.6744027303754265, + "grad_norm": 1.969222231883204, + "learning_rate": 9.349429746713859e-07, + "loss": 0.0994, + "step": 7359 + }, + { + "epoch": 1.6746302616609783, + "grad_norm": 2.2660444367482264, + "learning_rate": 9.348653810786667e-07, + "loss": 0.1153, + "step": 7360 + }, + { + "epoch": 1.67485779294653, + "grad_norm": 1.3311999975188635, + "learning_rate": 9.347877811528666e-07, + "loss": 0.1224, + "step": 7361 + }, + { + "epoch": 1.6750853242320818, + "grad_norm": 1.9003716373130064, + "learning_rate": 9.347101748955715e-07, + "loss": 0.1709, + "step": 7362 + }, + { + "epoch": 1.6753128555176335, + "grad_norm": 2.57799394374302, + "learning_rate": 9.346325623083679e-07, + "loss": 0.0849, + "step": 7363 + }, + { + "epoch": 1.6755403868031853, + "grad_norm": 1.3887719065547302, + "learning_rate": 9.345549433928416e-07, + "loss": 0.1361, + "step": 7364 + }, + { + "epoch": 1.675767918088737, + "grad_norm": 1.6924992160904613, + "learning_rate": 9.344773181505793e-07, + "loss": 0.1183, + "step": 7365 + }, + { + "epoch": 1.675995449374289, + "grad_norm": 1.8383687373119448, + "learning_rate": 9.343996865831673e-07, + "loss": 0.1812, + "step": 7366 + }, + { + "epoch": 1.6762229806598408, + "grad_norm": 1.8831689141554542, + "learning_rate": 9.343220486921924e-07, + "loss": 0.0607, + "step": 7367 + }, + { + "epoch": 1.6764505119453925, + "grad_norm": 1.4982669170761194, + "learning_rate": 9.342444044792412e-07, + "loss": 0.0613, + "step": 7368 + }, + { + "epoch": 1.6766780432309443, + "grad_norm": 1.3553553248957957, + "learning_rate": 9.341667539459006e-07, + "loss": 0.0816, + "step": 7369 + }, + { + "epoch": 1.676905574516496, + "grad_norm": 2.062071006752431, + "learning_rate": 9.340890970937583e-07, + "loss": 0.0679, + "step": 7370 + }, + { + "epoch": 1.6771331058020478, + "grad_norm": 1.6215383369901835, + "learning_rate": 9.340114339244006e-07, + "loss": 0.0387, + "step": 7371 + }, + { + "epoch": 1.6773606370875995, + "grad_norm": 1.8044914050783736, + "learning_rate": 9.33933764439415e-07, + "loss": 0.0636, + "step": 7372 + }, + { + "epoch": 1.6775881683731513, + "grad_norm": 1.1985423875980186, + "learning_rate": 9.338560886403891e-07, + "loss": 0.1181, + "step": 7373 + }, + { + "epoch": 1.677815699658703, + "grad_norm": 1.8611780061417367, + "learning_rate": 9.337784065289104e-07, + "loss": 0.1201, + "step": 7374 + }, + { + "epoch": 1.6780432309442548, + "grad_norm": 2.48197193723222, + "learning_rate": 9.337007181065667e-07, + "loss": 0.214, + "step": 7375 + }, + { + "epoch": 1.6782707622298068, + "grad_norm": 2.226764853413825, + "learning_rate": 9.336230233749456e-07, + "loss": 0.0789, + "step": 7376 + }, + { + "epoch": 1.6784982935153585, + "grad_norm": 1.878909636526424, + "learning_rate": 9.335453223356351e-07, + "loss": 0.0894, + "step": 7377 + }, + { + "epoch": 1.6787258248009103, + "grad_norm": 1.7374008228680462, + "learning_rate": 9.334676149902233e-07, + "loss": 0.1085, + "step": 7378 + }, + { + "epoch": 1.678953356086462, + "grad_norm": 1.3233416996257326, + "learning_rate": 9.333899013402984e-07, + "loss": 0.0782, + "step": 7379 + }, + { + "epoch": 1.6791808873720138, + "grad_norm": 2.4254202349146428, + "learning_rate": 9.333121813874486e-07, + "loss": 0.0991, + "step": 7380 + }, + { + "epoch": 1.6794084186575655, + "grad_norm": 1.8306572114812147, + "learning_rate": 9.332344551332626e-07, + "loss": 0.1109, + "step": 7381 + }, + { + "epoch": 1.6796359499431173, + "grad_norm": 1.6849794506796347, + "learning_rate": 9.331567225793288e-07, + "loss": 0.0925, + "step": 7382 + }, + { + "epoch": 1.679863481228669, + "grad_norm": 1.789407849888052, + "learning_rate": 9.330789837272359e-07, + "loss": 0.0914, + "step": 7383 + }, + { + "epoch": 1.6800910125142208, + "grad_norm": 1.9047852958348788, + "learning_rate": 9.330012385785729e-07, + "loss": 0.1122, + "step": 7384 + }, + { + "epoch": 1.6803185437997725, + "grad_norm": 1.4475225500771003, + "learning_rate": 9.329234871349285e-07, + "loss": 0.0829, + "step": 7385 + }, + { + "epoch": 1.6805460750853243, + "grad_norm": 2.9593046800656366, + "learning_rate": 9.328457293978921e-07, + "loss": 0.0867, + "step": 7386 + }, + { + "epoch": 1.680773606370876, + "grad_norm": 1.496333388049632, + "learning_rate": 9.327679653690527e-07, + "loss": 0.0758, + "step": 7387 + }, + { + "epoch": 1.6810011376564278, + "grad_norm": 1.9637310115501778, + "learning_rate": 9.326901950499997e-07, + "loss": 0.148, + "step": 7388 + }, + { + "epoch": 1.6812286689419795, + "grad_norm": 1.7570532890404065, + "learning_rate": 9.326124184423228e-07, + "loss": 0.0996, + "step": 7389 + }, + { + "epoch": 1.6814562002275313, + "grad_norm": 1.562562274864593, + "learning_rate": 9.325346355476114e-07, + "loss": 0.0752, + "step": 7390 + }, + { + "epoch": 1.681683731513083, + "grad_norm": 1.146072588137425, + "learning_rate": 9.324568463674552e-07, + "loss": 0.0484, + "step": 7391 + }, + { + "epoch": 1.6819112627986348, + "grad_norm": 2.2430769759037266, + "learning_rate": 9.323790509034441e-07, + "loss": 0.0917, + "step": 7392 + }, + { + "epoch": 1.6821387940841865, + "grad_norm": 2.3561795405892627, + "learning_rate": 9.323012491571682e-07, + "loss": 0.1339, + "step": 7393 + }, + { + "epoch": 1.6823663253697383, + "grad_norm": 1.8302136240158402, + "learning_rate": 9.322234411302175e-07, + "loss": 0.0976, + "step": 7394 + }, + { + "epoch": 1.68259385665529, + "grad_norm": 1.6372820312897203, + "learning_rate": 9.321456268241824e-07, + "loss": 0.0698, + "step": 7395 + }, + { + "epoch": 1.6828213879408418, + "grad_norm": 1.9643540747530492, + "learning_rate": 9.320678062406531e-07, + "loss": 0.066, + "step": 7396 + }, + { + "epoch": 1.6830489192263935, + "grad_norm": 1.838968404769983, + "learning_rate": 9.319899793812204e-07, + "loss": 0.0824, + "step": 7397 + }, + { + "epoch": 1.6832764505119453, + "grad_norm": 2.5218651261530125, + "learning_rate": 9.319121462474745e-07, + "loss": 0.1007, + "step": 7398 + }, + { + "epoch": 1.683503981797497, + "grad_norm": 1.2388968231459465, + "learning_rate": 9.318343068410065e-07, + "loss": 0.1034, + "step": 7399 + }, + { + "epoch": 1.6837315130830488, + "grad_norm": 1.336541589569829, + "learning_rate": 9.317564611634072e-07, + "loss": 0.0576, + "step": 7400 + }, + { + "epoch": 1.6839590443686006, + "grad_norm": 2.107630718393057, + "learning_rate": 9.316786092162678e-07, + "loss": 0.0993, + "step": 7401 + }, + { + "epoch": 1.6841865756541523, + "grad_norm": 1.766479652983476, + "learning_rate": 9.31600751001179e-07, + "loss": 0.2175, + "step": 7402 + }, + { + "epoch": 1.684414106939704, + "grad_norm": 2.4056863315167263, + "learning_rate": 9.315228865197326e-07, + "loss": 0.1928, + "step": 7403 + }, + { + "epoch": 1.6846416382252558, + "grad_norm": 3.2428738679985614, + "learning_rate": 9.314450157735197e-07, + "loss": 0.1178, + "step": 7404 + }, + { + "epoch": 1.6848691695108078, + "grad_norm": 1.678043329452889, + "learning_rate": 9.313671387641318e-07, + "loss": 0.142, + "step": 7405 + }, + { + "epoch": 1.6850967007963595, + "grad_norm": 2.58564835703974, + "learning_rate": 9.312892554931608e-07, + "loss": 0.1441, + "step": 7406 + }, + { + "epoch": 1.6853242320819113, + "grad_norm": 1.355706143918413, + "learning_rate": 9.312113659621984e-07, + "loss": 0.156, + "step": 7407 + }, + { + "epoch": 1.685551763367463, + "grad_norm": 1.9664685627593343, + "learning_rate": 9.311334701728364e-07, + "loss": 0.0824, + "step": 7408 + }, + { + "epoch": 1.6857792946530148, + "grad_norm": 1.3219939112430976, + "learning_rate": 9.310555681266668e-07, + "loss": 0.0582, + "step": 7409 + }, + { + "epoch": 1.6860068259385665, + "grad_norm": 2.142310689486401, + "learning_rate": 9.30977659825282e-07, + "loss": 0.0998, + "step": 7410 + }, + { + "epoch": 1.6862343572241183, + "grad_norm": 1.3473183389737833, + "learning_rate": 9.308997452702743e-07, + "loss": 0.069, + "step": 7411 + }, + { + "epoch": 1.68646188850967, + "grad_norm": 1.3004701099778404, + "learning_rate": 9.308218244632358e-07, + "loss": 0.0812, + "step": 7412 + }, + { + "epoch": 1.6866894197952218, + "grad_norm": 1.8468339197188928, + "learning_rate": 9.307438974057595e-07, + "loss": 0.0723, + "step": 7413 + }, + { + "epoch": 1.6869169510807736, + "grad_norm": 1.5261190981362804, + "learning_rate": 9.306659640994381e-07, + "loss": 0.0696, + "step": 7414 + }, + { + "epoch": 1.6871444823663255, + "grad_norm": 1.1719461672581963, + "learning_rate": 9.305880245458637e-07, + "loss": 0.1161, + "step": 7415 + }, + { + "epoch": 1.6873720136518773, + "grad_norm": 1.7890461179460164, + "learning_rate": 9.305100787466301e-07, + "loss": 0.1343, + "step": 7416 + }, + { + "epoch": 1.687599544937429, + "grad_norm": 3.307144150549738, + "learning_rate": 9.304321267033298e-07, + "loss": 0.1053, + "step": 7417 + }, + { + "epoch": 1.6878270762229808, + "grad_norm": 1.4387304873939588, + "learning_rate": 9.303541684175563e-07, + "loss": 0.0766, + "step": 7418 + }, + { + "epoch": 1.6880546075085325, + "grad_norm": 1.5682681746734135, + "learning_rate": 9.302762038909028e-07, + "loss": 0.1224, + "step": 7419 + }, + { + "epoch": 1.6882821387940843, + "grad_norm": 1.7379756393731887, + "learning_rate": 9.301982331249629e-07, + "loss": 0.0902, + "step": 7420 + }, + { + "epoch": 1.688509670079636, + "grad_norm": 1.788536899485545, + "learning_rate": 9.301202561213298e-07, + "loss": 0.0872, + "step": 7421 + }, + { + "epoch": 1.6887372013651878, + "grad_norm": 1.4193679489333435, + "learning_rate": 9.300422728815976e-07, + "loss": 0.0646, + "step": 7422 + }, + { + "epoch": 1.6889647326507395, + "grad_norm": 1.4684247824398127, + "learning_rate": 9.2996428340736e-07, + "loss": 0.0653, + "step": 7423 + }, + { + "epoch": 1.6891922639362913, + "grad_norm": 1.1675186706745337, + "learning_rate": 9.29886287700211e-07, + "loss": 0.0772, + "step": 7424 + }, + { + "epoch": 1.689419795221843, + "grad_norm": 1.2408519690279203, + "learning_rate": 9.298082857617446e-07, + "loss": 0.0949, + "step": 7425 + }, + { + "epoch": 1.6896473265073948, + "grad_norm": 1.6086428034451206, + "learning_rate": 9.297302775935552e-07, + "loss": 0.0828, + "step": 7426 + }, + { + "epoch": 1.6898748577929465, + "grad_norm": 1.879258209189917, + "learning_rate": 9.296522631972368e-07, + "loss": 0.1485, + "step": 7427 + }, + { + "epoch": 1.6901023890784983, + "grad_norm": 1.1973586410272825, + "learning_rate": 9.295742425743842e-07, + "loss": 0.0956, + "step": 7428 + }, + { + "epoch": 1.69032992036405, + "grad_norm": 1.4457121713578762, + "learning_rate": 9.294962157265918e-07, + "loss": 0.0554, + "step": 7429 + }, + { + "epoch": 1.6905574516496018, + "grad_norm": 1.4104396381813507, + "learning_rate": 9.294181826554547e-07, + "loss": 0.0562, + "step": 7430 + }, + { + "epoch": 1.6907849829351536, + "grad_norm": 0.9262969737219219, + "learning_rate": 9.29340143362567e-07, + "loss": 0.0855, + "step": 7431 + }, + { + "epoch": 1.6910125142207053, + "grad_norm": 1.453364971460863, + "learning_rate": 9.292620978495246e-07, + "loss": 0.0564, + "step": 7432 + }, + { + "epoch": 1.691240045506257, + "grad_norm": 1.551913385880135, + "learning_rate": 9.291840461179219e-07, + "loss": 0.1061, + "step": 7433 + }, + { + "epoch": 1.6914675767918088, + "grad_norm": 1.5546023867739276, + "learning_rate": 9.291059881693544e-07, + "loss": 0.0446, + "step": 7434 + }, + { + "epoch": 1.6916951080773606, + "grad_norm": 3.377966899747525, + "learning_rate": 9.290279240054176e-07, + "loss": 0.1998, + "step": 7435 + }, + { + "epoch": 1.6919226393629123, + "grad_norm": 1.7385496523204151, + "learning_rate": 9.289498536277066e-07, + "loss": 0.1234, + "step": 7436 + }, + { + "epoch": 1.692150170648464, + "grad_norm": 2.114012284187751, + "learning_rate": 9.288717770378172e-07, + "loss": 0.0982, + "step": 7437 + }, + { + "epoch": 1.6923777019340158, + "grad_norm": 1.351731477417337, + "learning_rate": 9.287936942373454e-07, + "loss": 0.0787, + "step": 7438 + }, + { + "epoch": 1.6926052332195676, + "grad_norm": 1.8739219968172143, + "learning_rate": 9.287156052278869e-07, + "loss": 0.0679, + "step": 7439 + }, + { + "epoch": 1.6928327645051193, + "grad_norm": 1.5864485237863282, + "learning_rate": 9.286375100110376e-07, + "loss": 0.0902, + "step": 7440 + }, + { + "epoch": 1.693060295790671, + "grad_norm": 1.0252685039976748, + "learning_rate": 9.285594085883937e-07, + "loss": 0.0306, + "step": 7441 + }, + { + "epoch": 1.6932878270762228, + "grad_norm": 1.921676451564313, + "learning_rate": 9.284813009615512e-07, + "loss": 0.1212, + "step": 7442 + }, + { + "epoch": 1.6935153583617746, + "grad_norm": 1.2815356918274832, + "learning_rate": 9.28403187132107e-07, + "loss": 0.1666, + "step": 7443 + }, + { + "epoch": 1.6937428896473266, + "grad_norm": 1.5595077639365074, + "learning_rate": 9.28325067101657e-07, + "loss": 0.1326, + "step": 7444 + }, + { + "epoch": 1.6939704209328783, + "grad_norm": 1.0162405891241884, + "learning_rate": 9.282469408717984e-07, + "loss": 0.0482, + "step": 7445 + }, + { + "epoch": 1.69419795221843, + "grad_norm": 1.3773979926189792, + "learning_rate": 9.281688084441277e-07, + "loss": 0.1231, + "step": 7446 + }, + { + "epoch": 1.6944254835039818, + "grad_norm": 1.104581808250732, + "learning_rate": 9.280906698202417e-07, + "loss": 0.1338, + "step": 7447 + }, + { + "epoch": 1.6946530147895336, + "grad_norm": 2.2721944255174846, + "learning_rate": 9.280125250017375e-07, + "loss": 0.1168, + "step": 7448 + }, + { + "epoch": 1.6948805460750853, + "grad_norm": 1.942154363545402, + "learning_rate": 9.279343739902122e-07, + "loss": 0.0765, + "step": 7449 + }, + { + "epoch": 1.695108077360637, + "grad_norm": 2.4606208794929048, + "learning_rate": 9.278562167872632e-07, + "loss": 0.0913, + "step": 7450 + }, + { + "epoch": 1.6953356086461888, + "grad_norm": 2.3818147882797263, + "learning_rate": 9.277780533944879e-07, + "loss": 0.0863, + "step": 7451 + }, + { + "epoch": 1.6955631399317406, + "grad_norm": 1.3677925854977118, + "learning_rate": 9.276998838134834e-07, + "loss": 0.0586, + "step": 7452 + }, + { + "epoch": 1.6957906712172923, + "grad_norm": 1.8230493817169902, + "learning_rate": 9.276217080458478e-07, + "loss": 0.0808, + "step": 7453 + }, + { + "epoch": 1.6960182025028443, + "grad_norm": 1.8503778421300845, + "learning_rate": 9.275435260931786e-07, + "loss": 0.0816, + "step": 7454 + }, + { + "epoch": 1.696245733788396, + "grad_norm": 2.000832926112369, + "learning_rate": 9.274653379570739e-07, + "loss": 0.169, + "step": 7455 + }, + { + "epoch": 1.6964732650739478, + "grad_norm": 1.0146915198809943, + "learning_rate": 9.273871436391315e-07, + "loss": 0.0714, + "step": 7456 + }, + { + "epoch": 1.6967007963594996, + "grad_norm": 1.2304016971464222, + "learning_rate": 9.273089431409499e-07, + "loss": 0.0421, + "step": 7457 + }, + { + "epoch": 1.6969283276450513, + "grad_norm": 2.047942404694497, + "learning_rate": 9.272307364641271e-07, + "loss": 0.0755, + "step": 7458 + }, + { + "epoch": 1.697155858930603, + "grad_norm": 2.6383743586012156, + "learning_rate": 9.271525236102614e-07, + "loss": 0.107, + "step": 7459 + }, + { + "epoch": 1.6973833902161548, + "grad_norm": 2.7056667767669076, + "learning_rate": 9.270743045809516e-07, + "loss": 0.0826, + "step": 7460 + }, + { + "epoch": 1.6976109215017066, + "grad_norm": 1.7563129007136633, + "learning_rate": 9.269960793777963e-07, + "loss": 0.0915, + "step": 7461 + }, + { + "epoch": 1.6978384527872583, + "grad_norm": 1.956944676427302, + "learning_rate": 9.26917848002394e-07, + "loss": 0.0903, + "step": 7462 + }, + { + "epoch": 1.69806598407281, + "grad_norm": 2.4161778851437843, + "learning_rate": 9.26839610456344e-07, + "loss": 0.0969, + "step": 7463 + }, + { + "epoch": 1.6982935153583618, + "grad_norm": 2.068452668994588, + "learning_rate": 9.267613667412453e-07, + "loss": 0.0607, + "step": 7464 + }, + { + "epoch": 1.6985210466439136, + "grad_norm": 2.0151967606843457, + "learning_rate": 9.266831168586968e-07, + "loss": 0.0696, + "step": 7465 + }, + { + "epoch": 1.6987485779294653, + "grad_norm": 1.1856067501750585, + "learning_rate": 9.266048608102978e-07, + "loss": 0.0574, + "step": 7466 + }, + { + "epoch": 1.698976109215017, + "grad_norm": 2.688891725537148, + "learning_rate": 9.265265985976478e-07, + "loss": 0.0888, + "step": 7467 + }, + { + "epoch": 1.6992036405005688, + "grad_norm": 1.4125321029456797, + "learning_rate": 9.264483302223464e-07, + "loss": 0.1077, + "step": 7468 + }, + { + "epoch": 1.6994311717861206, + "grad_norm": 1.9819253790117193, + "learning_rate": 9.263700556859931e-07, + "loss": 0.1409, + "step": 7469 + }, + { + "epoch": 1.6996587030716723, + "grad_norm": 1.4363926897897454, + "learning_rate": 9.26291774990188e-07, + "loss": 0.0711, + "step": 7470 + }, + { + "epoch": 1.699886234357224, + "grad_norm": 0.8563946587398912, + "learning_rate": 9.262134881365307e-07, + "loss": 0.0416, + "step": 7471 + }, + { + "epoch": 1.7001137656427758, + "grad_norm": 2.4762136004144315, + "learning_rate": 9.261351951266211e-07, + "loss": 0.0967, + "step": 7472 + }, + { + "epoch": 1.7003412969283276, + "grad_norm": 2.4079001396424813, + "learning_rate": 9.2605689596206e-07, + "loss": 0.1825, + "step": 7473 + }, + { + "epoch": 1.7005688282138793, + "grad_norm": 1.8068877491506772, + "learning_rate": 9.259785906444473e-07, + "loss": 0.197, + "step": 7474 + }, + { + "epoch": 1.700796359499431, + "grad_norm": 1.5866555515774667, + "learning_rate": 9.259002791753832e-07, + "loss": 0.087, + "step": 7475 + }, + { + "epoch": 1.7010238907849828, + "grad_norm": 1.6481904048653275, + "learning_rate": 9.258219615564684e-07, + "loss": 0.1001, + "step": 7476 + }, + { + "epoch": 1.7012514220705346, + "grad_norm": 1.2942037721759625, + "learning_rate": 9.25743637789304e-07, + "loss": 0.0349, + "step": 7477 + }, + { + "epoch": 1.7014789533560863, + "grad_norm": 1.4716549549075366, + "learning_rate": 9.256653078754903e-07, + "loss": 0.0827, + "step": 7478 + }, + { + "epoch": 1.701706484641638, + "grad_norm": 2.6593429187871367, + "learning_rate": 9.255869718166281e-07, + "loss": 0.2166, + "step": 7479 + }, + { + "epoch": 1.7019340159271898, + "grad_norm": 1.9648955811870432, + "learning_rate": 9.255086296143189e-07, + "loss": 0.1262, + "step": 7480 + }, + { + "epoch": 1.7021615472127416, + "grad_norm": 1.7478426250374688, + "learning_rate": 9.254302812701636e-07, + "loss": 0.0644, + "step": 7481 + }, + { + "epoch": 1.7023890784982934, + "grad_norm": 2.987380058350133, + "learning_rate": 9.253519267857637e-07, + "loss": 0.1216, + "step": 7482 + }, + { + "epoch": 1.7026166097838453, + "grad_norm": 2.3484831882762935, + "learning_rate": 9.252735661627204e-07, + "loss": 0.0812, + "step": 7483 + }, + { + "epoch": 1.702844141069397, + "grad_norm": 2.1512910669945664, + "learning_rate": 9.251951994026353e-07, + "loss": 0.1304, + "step": 7484 + }, + { + "epoch": 1.7030716723549488, + "grad_norm": 1.528584307846986, + "learning_rate": 9.251168265071101e-07, + "loss": 0.1446, + "step": 7485 + }, + { + "epoch": 1.7032992036405006, + "grad_norm": 1.523165761907841, + "learning_rate": 9.250384474777465e-07, + "loss": 0.1226, + "step": 7486 + }, + { + "epoch": 1.7035267349260523, + "grad_norm": 1.0735981813202926, + "learning_rate": 9.249600623161467e-07, + "loss": 0.0632, + "step": 7487 + }, + { + "epoch": 1.703754266211604, + "grad_norm": 1.7693977564744923, + "learning_rate": 9.248816710239125e-07, + "loss": 0.1213, + "step": 7488 + }, + { + "epoch": 1.7039817974971558, + "grad_norm": 1.5156085599699003, + "learning_rate": 9.248032736026463e-07, + "loss": 0.0664, + "step": 7489 + }, + { + "epoch": 1.7042093287827076, + "grad_norm": 1.8444395238269065, + "learning_rate": 9.247248700539502e-07, + "loss": 0.0733, + "step": 7490 + }, + { + "epoch": 1.7044368600682593, + "grad_norm": 2.282489530965188, + "learning_rate": 9.246464603794266e-07, + "loss": 0.1731, + "step": 7491 + }, + { + "epoch": 1.7046643913538113, + "grad_norm": 1.1135323424692893, + "learning_rate": 9.245680445806782e-07, + "loss": 0.0622, + "step": 7492 + }, + { + "epoch": 1.704891922639363, + "grad_norm": 1.5015562370177542, + "learning_rate": 9.244896226593074e-07, + "loss": 0.0755, + "step": 7493 + }, + { + "epoch": 1.7051194539249148, + "grad_norm": 2.380047859800059, + "learning_rate": 9.244111946169173e-07, + "loss": 0.1285, + "step": 7494 + }, + { + "epoch": 1.7053469852104666, + "grad_norm": 1.4197630339019536, + "learning_rate": 9.243327604551109e-07, + "loss": 0.0716, + "step": 7495 + }, + { + "epoch": 1.7055745164960183, + "grad_norm": 1.8467214333007922, + "learning_rate": 9.242543201754908e-07, + "loss": 0.1151, + "step": 7496 + }, + { + "epoch": 1.70580204778157, + "grad_norm": 2.3732110067695302, + "learning_rate": 9.241758737796608e-07, + "loss": 0.0923, + "step": 7497 + }, + { + "epoch": 1.7060295790671218, + "grad_norm": 1.7194839267259672, + "learning_rate": 9.240974212692235e-07, + "loss": 0.0953, + "step": 7498 + }, + { + "epoch": 1.7062571103526736, + "grad_norm": 0.8677031615537418, + "learning_rate": 9.240189626457828e-07, + "loss": 0.0689, + "step": 7499 + }, + { + "epoch": 1.7064846416382253, + "grad_norm": 1.7933546109365042, + "learning_rate": 9.239404979109422e-07, + "loss": 0.0692, + "step": 7500 + }, + { + "epoch": 1.706712172923777, + "grad_norm": 1.5534447964205367, + "learning_rate": 9.238620270663053e-07, + "loss": 0.0825, + "step": 7501 + }, + { + "epoch": 1.7069397042093288, + "grad_norm": 1.981780715242399, + "learning_rate": 9.237835501134759e-07, + "loss": 0.1216, + "step": 7502 + }, + { + "epoch": 1.7071672354948806, + "grad_norm": 1.95930269394389, + "learning_rate": 9.237050670540579e-07, + "loss": 0.0715, + "step": 7503 + }, + { + "epoch": 1.7073947667804323, + "grad_norm": 1.7374402364968724, + "learning_rate": 9.236265778896554e-07, + "loss": 0.0453, + "step": 7504 + }, + { + "epoch": 1.707622298065984, + "grad_norm": 1.7191098911921983, + "learning_rate": 9.235480826218726e-07, + "loss": 0.117, + "step": 7505 + }, + { + "epoch": 1.7078498293515358, + "grad_norm": 1.3374036497567463, + "learning_rate": 9.234695812523137e-07, + "loss": 0.0876, + "step": 7506 + }, + { + "epoch": 1.7080773606370876, + "grad_norm": 1.700497077975526, + "learning_rate": 9.233910737825831e-07, + "loss": 0.1571, + "step": 7507 + }, + { + "epoch": 1.7083048919226393, + "grad_norm": 2.4095573392593144, + "learning_rate": 9.233125602142856e-07, + "loss": 0.0849, + "step": 7508 + }, + { + "epoch": 1.708532423208191, + "grad_norm": 1.8893150630704287, + "learning_rate": 9.232340405490256e-07, + "loss": 0.0968, + "step": 7509 + }, + { + "epoch": 1.7087599544937428, + "grad_norm": 1.8421406492720631, + "learning_rate": 9.23155514788408e-07, + "loss": 0.0758, + "step": 7510 + }, + { + "epoch": 1.7089874857792946, + "grad_norm": 2.521534449539196, + "learning_rate": 9.230769829340378e-07, + "loss": 0.1002, + "step": 7511 + }, + { + "epoch": 1.7092150170648464, + "grad_norm": 1.896914834149343, + "learning_rate": 9.229984449875199e-07, + "loss": 0.095, + "step": 7512 + }, + { + "epoch": 1.709442548350398, + "grad_norm": 1.7323554177641725, + "learning_rate": 9.229199009504594e-07, + "loss": 0.0693, + "step": 7513 + }, + { + "epoch": 1.7096700796359499, + "grad_norm": 1.5400684620058087, + "learning_rate": 9.228413508244621e-07, + "loss": 0.0503, + "step": 7514 + }, + { + "epoch": 1.7098976109215016, + "grad_norm": 1.4653262523601092, + "learning_rate": 9.227627946111328e-07, + "loss": 0.1178, + "step": 7515 + }, + { + "epoch": 1.7101251422070534, + "grad_norm": 2.5594902657863297, + "learning_rate": 9.226842323120773e-07, + "loss": 0.0907, + "step": 7516 + }, + { + "epoch": 1.7103526734926051, + "grad_norm": 3.080947181053405, + "learning_rate": 9.226056639289013e-07, + "loss": 0.1352, + "step": 7517 + }, + { + "epoch": 1.7105802047781569, + "grad_norm": 1.8768778604470615, + "learning_rate": 9.225270894632107e-07, + "loss": 0.0519, + "step": 7518 + }, + { + "epoch": 1.7108077360637086, + "grad_norm": 1.2966109435086275, + "learning_rate": 9.224485089166111e-07, + "loss": 0.0787, + "step": 7519 + }, + { + "epoch": 1.7110352673492604, + "grad_norm": 1.4658236968406182, + "learning_rate": 9.223699222907088e-07, + "loss": 0.0926, + "step": 7520 + }, + { + "epoch": 1.7112627986348121, + "grad_norm": 2.9174819569607338, + "learning_rate": 9.222913295871101e-07, + "loss": 0.0919, + "step": 7521 + }, + { + "epoch": 1.711490329920364, + "grad_norm": 1.2475180743268615, + "learning_rate": 9.22212730807421e-07, + "loss": 0.1199, + "step": 7522 + }, + { + "epoch": 1.7117178612059158, + "grad_norm": 1.9562927246156157, + "learning_rate": 9.221341259532476e-07, + "loss": 0.1351, + "step": 7523 + }, + { + "epoch": 1.7119453924914676, + "grad_norm": 1.34154787076131, + "learning_rate": 9.220555150261973e-07, + "loss": 0.0829, + "step": 7524 + }, + { + "epoch": 1.7121729237770194, + "grad_norm": 2.239735065898415, + "learning_rate": 9.219768980278762e-07, + "loss": 0.152, + "step": 7525 + }, + { + "epoch": 1.712400455062571, + "grad_norm": 1.5791759662482927, + "learning_rate": 9.218982749598911e-07, + "loss": 0.1274, + "step": 7526 + }, + { + "epoch": 1.7126279863481229, + "grad_norm": 1.413781534828615, + "learning_rate": 9.21819645823849e-07, + "loss": 0.0514, + "step": 7527 + }, + { + "epoch": 1.7128555176336746, + "grad_norm": 1.702854036720936, + "learning_rate": 9.21741010621357e-07, + "loss": 0.0738, + "step": 7528 + }, + { + "epoch": 1.7130830489192264, + "grad_norm": 2.74163381590818, + "learning_rate": 9.216623693540222e-07, + "loss": 0.1665, + "step": 7529 + }, + { + "epoch": 1.713310580204778, + "grad_norm": 0.9798309074213785, + "learning_rate": 9.215837220234518e-07, + "loss": 0.0385, + "step": 7530 + }, + { + "epoch": 1.71353811149033, + "grad_norm": 3.196701294240453, + "learning_rate": 9.215050686312534e-07, + "loss": 0.1205, + "step": 7531 + }, + { + "epoch": 1.7137656427758818, + "grad_norm": 1.7687344142577266, + "learning_rate": 9.214264091790343e-07, + "loss": 0.0513, + "step": 7532 + }, + { + "epoch": 1.7139931740614336, + "grad_norm": 2.5907793564802146, + "learning_rate": 9.213477436684025e-07, + "loss": 0.1354, + "step": 7533 + }, + { + "epoch": 1.7142207053469853, + "grad_norm": 1.6937574383193232, + "learning_rate": 9.212690721009654e-07, + "loss": 0.1259, + "step": 7534 + }, + { + "epoch": 1.714448236632537, + "grad_norm": 1.3495979297592327, + "learning_rate": 9.21190394478331e-07, + "loss": 0.128, + "step": 7535 + }, + { + "epoch": 1.7146757679180888, + "grad_norm": 2.476516605961478, + "learning_rate": 9.211117108021074e-07, + "loss": 0.1085, + "step": 7536 + }, + { + "epoch": 1.7149032992036406, + "grad_norm": 1.9707282922873293, + "learning_rate": 9.210330210739029e-07, + "loss": 0.0851, + "step": 7537 + }, + { + "epoch": 1.7151308304891923, + "grad_norm": 1.4845383213085293, + "learning_rate": 9.209543252953254e-07, + "loss": 0.0607, + "step": 7538 + }, + { + "epoch": 1.715358361774744, + "grad_norm": 2.0048158522616553, + "learning_rate": 9.208756234679836e-07, + "loss": 0.0919, + "step": 7539 + }, + { + "epoch": 1.7155858930602959, + "grad_norm": 1.4218673743727959, + "learning_rate": 9.20796915593486e-07, + "loss": 0.1921, + "step": 7540 + }, + { + "epoch": 1.7158134243458476, + "grad_norm": 1.3584412932163983, + "learning_rate": 9.20718201673441e-07, + "loss": 0.1622, + "step": 7541 + }, + { + "epoch": 1.7160409556313994, + "grad_norm": 1.0608020578801372, + "learning_rate": 9.206394817094577e-07, + "loss": 0.0509, + "step": 7542 + }, + { + "epoch": 1.716268486916951, + "grad_norm": 1.6070514150158068, + "learning_rate": 9.205607557031446e-07, + "loss": 0.0854, + "step": 7543 + }, + { + "epoch": 1.7164960182025029, + "grad_norm": 1.254528380272989, + "learning_rate": 9.204820236561111e-07, + "loss": 0.0653, + "step": 7544 + }, + { + "epoch": 1.7167235494880546, + "grad_norm": 1.5303193243898203, + "learning_rate": 9.204032855699663e-07, + "loss": 0.0626, + "step": 7545 + }, + { + "epoch": 1.7169510807736064, + "grad_norm": 2.479100939226992, + "learning_rate": 9.203245414463192e-07, + "loss": 0.092, + "step": 7546 + }, + { + "epoch": 1.7171786120591581, + "grad_norm": 1.9985993200466214, + "learning_rate": 9.202457912867795e-07, + "loss": 0.0841, + "step": 7547 + }, + { + "epoch": 1.7174061433447099, + "grad_norm": 1.846036514822781, + "learning_rate": 9.201670350929564e-07, + "loss": 0.1315, + "step": 7548 + }, + { + "epoch": 1.7176336746302616, + "grad_norm": 2.0790408771999718, + "learning_rate": 9.200882728664598e-07, + "loss": 0.0576, + "step": 7549 + }, + { + "epoch": 1.7178612059158134, + "grad_norm": 1.7949883449506443, + "learning_rate": 9.20009504608899e-07, + "loss": 0.0988, + "step": 7550 + }, + { + "epoch": 1.7180887372013651, + "grad_norm": 1.5998800056946103, + "learning_rate": 9.199307303218844e-07, + "loss": 0.0432, + "step": 7551 + }, + { + "epoch": 1.7183162684869169, + "grad_norm": 1.7524050987728221, + "learning_rate": 9.198519500070261e-07, + "loss": 0.0697, + "step": 7552 + }, + { + "epoch": 1.7185437997724686, + "grad_norm": 1.2703856925744228, + "learning_rate": 9.197731636659335e-07, + "loss": 0.0854, + "step": 7553 + }, + { + "epoch": 1.7187713310580204, + "grad_norm": 1.8783737724172556, + "learning_rate": 9.196943713002177e-07, + "loss": 0.1068, + "step": 7554 + }, + { + "epoch": 1.7189988623435721, + "grad_norm": 1.0936436524766402, + "learning_rate": 9.196155729114883e-07, + "loss": 0.0411, + "step": 7555 + }, + { + "epoch": 1.7192263936291239, + "grad_norm": 2.0836304634461746, + "learning_rate": 9.195367685013564e-07, + "loss": 0.0702, + "step": 7556 + }, + { + "epoch": 1.7194539249146756, + "grad_norm": 1.6019233848518821, + "learning_rate": 9.19457958071432e-07, + "loss": 0.1408, + "step": 7557 + }, + { + "epoch": 1.7196814562002274, + "grad_norm": 1.4860268396568896, + "learning_rate": 9.193791416233266e-07, + "loss": 0.0727, + "step": 7558 + }, + { + "epoch": 1.7199089874857791, + "grad_norm": 1.5916056892498054, + "learning_rate": 9.193003191586507e-07, + "loss": 0.124, + "step": 7559 + }, + { + "epoch": 1.7201365187713311, + "grad_norm": 1.489766796980958, + "learning_rate": 9.192214906790149e-07, + "loss": 0.0671, + "step": 7560 + }, + { + "epoch": 1.7203640500568829, + "grad_norm": 2.0567448731874647, + "learning_rate": 9.191426561860308e-07, + "loss": 0.1081, + "step": 7561 + }, + { + "epoch": 1.7205915813424346, + "grad_norm": 1.281462849688222, + "learning_rate": 9.190638156813097e-07, + "loss": 0.0748, + "step": 7562 + }, + { + "epoch": 1.7208191126279864, + "grad_norm": 2.0271807019330947, + "learning_rate": 9.189849691664626e-07, + "loss": 0.0908, + "step": 7563 + }, + { + "epoch": 1.7210466439135381, + "grad_norm": 1.500561474240595, + "learning_rate": 9.189061166431012e-07, + "loss": 0.1661, + "step": 7564 + }, + { + "epoch": 1.7212741751990899, + "grad_norm": 1.2723138560334935, + "learning_rate": 9.188272581128372e-07, + "loss": 0.0645, + "step": 7565 + }, + { + "epoch": 1.7215017064846416, + "grad_norm": 1.9257600489647957, + "learning_rate": 9.187483935772818e-07, + "loss": 0.1276, + "step": 7566 + }, + { + "epoch": 1.7217292377701934, + "grad_norm": 2.0566117894728513, + "learning_rate": 9.186695230380474e-07, + "loss": 0.0831, + "step": 7567 + }, + { + "epoch": 1.7219567690557451, + "grad_norm": 1.5340647500171707, + "learning_rate": 9.185906464967459e-07, + "loss": 0.0623, + "step": 7568 + }, + { + "epoch": 1.7221843003412969, + "grad_norm": 2.1190698924957134, + "learning_rate": 9.185117639549891e-07, + "loss": 0.1153, + "step": 7569 + }, + { + "epoch": 1.7224118316268489, + "grad_norm": 1.7292016066869216, + "learning_rate": 9.184328754143893e-07, + "loss": 0.0518, + "step": 7570 + }, + { + "epoch": 1.7226393629124006, + "grad_norm": 1.1964881114058432, + "learning_rate": 9.183539808765591e-07, + "loss": 0.0673, + "step": 7571 + }, + { + "epoch": 1.7228668941979524, + "grad_norm": 1.1565136922837351, + "learning_rate": 9.182750803431109e-07, + "loss": 0.0851, + "step": 7572 + }, + { + "epoch": 1.723094425483504, + "grad_norm": 1.8740841010678184, + "learning_rate": 9.181961738156568e-07, + "loss": 0.1463, + "step": 7573 + }, + { + "epoch": 1.7233219567690559, + "grad_norm": 2.1491852106368943, + "learning_rate": 9.181172612958101e-07, + "loss": 0.0839, + "step": 7574 + }, + { + "epoch": 1.7235494880546076, + "grad_norm": 1.8165011826984763, + "learning_rate": 9.180383427851834e-07, + "loss": 0.1487, + "step": 7575 + }, + { + "epoch": 1.7237770193401594, + "grad_norm": 1.9293967119537, + "learning_rate": 9.179594182853898e-07, + "loss": 0.0593, + "step": 7576 + }, + { + "epoch": 1.7240045506257111, + "grad_norm": 1.697160909654997, + "learning_rate": 9.17880487798042e-07, + "loss": 0.1196, + "step": 7577 + }, + { + "epoch": 1.7242320819112629, + "grad_norm": 1.4096567598186796, + "learning_rate": 9.178015513247534e-07, + "loss": 0.0752, + "step": 7578 + }, + { + "epoch": 1.7244596131968146, + "grad_norm": 1.8342504525775487, + "learning_rate": 9.177226088671375e-07, + "loss": 0.0839, + "step": 7579 + }, + { + "epoch": 1.7246871444823664, + "grad_norm": 1.5778205964395018, + "learning_rate": 9.176436604268073e-07, + "loss": 0.0839, + "step": 7580 + }, + { + "epoch": 1.7249146757679181, + "grad_norm": 1.1085482431210012, + "learning_rate": 9.175647060053767e-07, + "loss": 0.0974, + "step": 7581 + }, + { + "epoch": 1.7251422070534699, + "grad_norm": 1.8092570257001175, + "learning_rate": 9.174857456044595e-07, + "loss": 0.0599, + "step": 7582 + }, + { + "epoch": 1.7253697383390216, + "grad_norm": 1.0296866950786054, + "learning_rate": 9.174067792256693e-07, + "loss": 0.0395, + "step": 7583 + }, + { + "epoch": 1.7255972696245734, + "grad_norm": 1.216069418979427, + "learning_rate": 9.1732780687062e-07, + "loss": 0.0317, + "step": 7584 + }, + { + "epoch": 1.7258248009101251, + "grad_norm": 1.7995583471787395, + "learning_rate": 9.172488285409256e-07, + "loss": 0.0624, + "step": 7585 + }, + { + "epoch": 1.7260523321956769, + "grad_norm": 2.18914611889024, + "learning_rate": 9.171698442382005e-07, + "loss": 0.1457, + "step": 7586 + }, + { + "epoch": 1.7262798634812286, + "grad_norm": 2.1762080793122105, + "learning_rate": 9.170908539640587e-07, + "loss": 0.0768, + "step": 7587 + }, + { + "epoch": 1.7265073947667804, + "grad_norm": 2.1161315095336413, + "learning_rate": 9.170118577201149e-07, + "loss": 0.1035, + "step": 7588 + }, + { + "epoch": 1.7267349260523321, + "grad_norm": 1.20294784797812, + "learning_rate": 9.169328555079836e-07, + "loss": 0.0765, + "step": 7589 + }, + { + "epoch": 1.726962457337884, + "grad_norm": 1.3058798502727609, + "learning_rate": 9.168538473292793e-07, + "loss": 0.0459, + "step": 7590 + }, + { + "epoch": 1.7271899886234356, + "grad_norm": 2.3609082153012855, + "learning_rate": 9.167748331856169e-07, + "loss": 0.1301, + "step": 7591 + }, + { + "epoch": 1.7274175199089874, + "grad_norm": 1.5176580933563264, + "learning_rate": 9.166958130786113e-07, + "loss": 0.0417, + "step": 7592 + }, + { + "epoch": 1.7276450511945391, + "grad_norm": 2.861613410192663, + "learning_rate": 9.166167870098773e-07, + "loss": 0.1977, + "step": 7593 + }, + { + "epoch": 1.727872582480091, + "grad_norm": 1.6362586317221, + "learning_rate": 9.165377549810305e-07, + "loss": 0.0739, + "step": 7594 + }, + { + "epoch": 1.7281001137656427, + "grad_norm": 1.3870134636338667, + "learning_rate": 9.164587169936858e-07, + "loss": 0.0564, + "step": 7595 + }, + { + "epoch": 1.7283276450511944, + "grad_norm": 1.77736625821223, + "learning_rate": 9.163796730494587e-07, + "loss": 0.0844, + "step": 7596 + }, + { + "epoch": 1.7285551763367462, + "grad_norm": 1.5654068200172355, + "learning_rate": 9.163006231499647e-07, + "loss": 0.0794, + "step": 7597 + }, + { + "epoch": 1.728782707622298, + "grad_norm": 1.6324365372014396, + "learning_rate": 9.162215672968194e-07, + "loss": 0.0715, + "step": 7598 + }, + { + "epoch": 1.7290102389078499, + "grad_norm": 1.775487028133511, + "learning_rate": 9.161425054916388e-07, + "loss": 0.0984, + "step": 7599 + }, + { + "epoch": 1.7292377701934016, + "grad_norm": 2.596742115531444, + "learning_rate": 9.160634377360383e-07, + "loss": 0.0943, + "step": 7600 + }, + { + "epoch": 1.7294653014789534, + "grad_norm": 1.2799569142806528, + "learning_rate": 9.159843640316345e-07, + "loss": 0.1087, + "step": 7601 + }, + { + "epoch": 1.7296928327645051, + "grad_norm": 1.3882731828318262, + "learning_rate": 9.159052843800431e-07, + "loss": 0.0969, + "step": 7602 + }, + { + "epoch": 1.729920364050057, + "grad_norm": 1.7312412523501368, + "learning_rate": 9.158261987828804e-07, + "loss": 0.054, + "step": 7603 + }, + { + "epoch": 1.7301478953356086, + "grad_norm": 1.5319719763944186, + "learning_rate": 9.157471072417629e-07, + "loss": 0.1203, + "step": 7604 + }, + { + "epoch": 1.7303754266211604, + "grad_norm": 1.5104710044764469, + "learning_rate": 9.156680097583071e-07, + "loss": 0.0367, + "step": 7605 + }, + { + "epoch": 1.7306029579067121, + "grad_norm": 1.3247706686865452, + "learning_rate": 9.155889063341293e-07, + "loss": 0.0843, + "step": 7606 + }, + { + "epoch": 1.730830489192264, + "grad_norm": 1.2796887606358964, + "learning_rate": 9.155097969708464e-07, + "loss": 0.1399, + "step": 7607 + }, + { + "epoch": 1.7310580204778157, + "grad_norm": 1.3776077953235888, + "learning_rate": 9.154306816700755e-07, + "loss": 0.1421, + "step": 7608 + }, + { + "epoch": 1.7312855517633676, + "grad_norm": 2.131102346008641, + "learning_rate": 9.153515604334334e-07, + "loss": 0.0511, + "step": 7609 + }, + { + "epoch": 1.7315130830489194, + "grad_norm": 1.1597126843073358, + "learning_rate": 9.152724332625369e-07, + "loss": 0.0577, + "step": 7610 + }, + { + "epoch": 1.7317406143344711, + "grad_norm": 1.8881270454348769, + "learning_rate": 9.151933001590035e-07, + "loss": 0.0532, + "step": 7611 + }, + { + "epoch": 1.7319681456200229, + "grad_norm": 1.383856972416893, + "learning_rate": 9.151141611244507e-07, + "loss": 0.1214, + "step": 7612 + }, + { + "epoch": 1.7321956769055746, + "grad_norm": 1.4899115457796293, + "learning_rate": 9.150350161604957e-07, + "loss": 0.135, + "step": 7613 + }, + { + "epoch": 1.7324232081911264, + "grad_norm": 2.132261288734918, + "learning_rate": 9.149558652687561e-07, + "loss": 0.1575, + "step": 7614 + }, + { + "epoch": 1.7326507394766781, + "grad_norm": 1.6673093075964234, + "learning_rate": 9.148767084508497e-07, + "loss": 0.0531, + "step": 7615 + }, + { + "epoch": 1.7328782707622299, + "grad_norm": 1.894698040329308, + "learning_rate": 9.147975457083943e-07, + "loss": 0.062, + "step": 7616 + }, + { + "epoch": 1.7331058020477816, + "grad_norm": 1.6934329179996055, + "learning_rate": 9.147183770430076e-07, + "loss": 0.0803, + "step": 7617 + }, + { + "epoch": 1.7333333333333334, + "grad_norm": 2.055058833041149, + "learning_rate": 9.146392024563081e-07, + "loss": 0.1421, + "step": 7618 + }, + { + "epoch": 1.7335608646188851, + "grad_norm": 1.2194880106920962, + "learning_rate": 9.145600219499137e-07, + "loss": 0.0652, + "step": 7619 + }, + { + "epoch": 1.733788395904437, + "grad_norm": 3.787036965512966, + "learning_rate": 9.144808355254426e-07, + "loss": 0.1356, + "step": 7620 + }, + { + "epoch": 1.7340159271899886, + "grad_norm": 1.7698576527312129, + "learning_rate": 9.144016431845136e-07, + "loss": 0.0539, + "step": 7621 + }, + { + "epoch": 1.7342434584755404, + "grad_norm": 1.656528845656429, + "learning_rate": 9.143224449287449e-07, + "loss": 0.076, + "step": 7622 + }, + { + "epoch": 1.7344709897610922, + "grad_norm": 2.206917546512681, + "learning_rate": 9.142432407597552e-07, + "loss": 0.1051, + "step": 7623 + }, + { + "epoch": 1.734698521046644, + "grad_norm": 3.318813955989774, + "learning_rate": 9.141640306791635e-07, + "loss": 0.062, + "step": 7624 + }, + { + "epoch": 1.7349260523321957, + "grad_norm": 1.19670148986426, + "learning_rate": 9.140848146885888e-07, + "loss": 0.1542, + "step": 7625 + }, + { + "epoch": 1.7351535836177474, + "grad_norm": 1.6504370070610002, + "learning_rate": 9.140055927896497e-07, + "loss": 0.0701, + "step": 7626 + }, + { + "epoch": 1.7353811149032992, + "grad_norm": 2.215620986397485, + "learning_rate": 9.139263649839654e-07, + "loss": 0.1286, + "step": 7627 + }, + { + "epoch": 1.735608646188851, + "grad_norm": 1.6530715852937048, + "learning_rate": 9.138471312731558e-07, + "loss": 0.0639, + "step": 7628 + }, + { + "epoch": 1.7358361774744027, + "grad_norm": 1.199037402415222, + "learning_rate": 9.137678916588395e-07, + "loss": 0.0884, + "step": 7629 + }, + { + "epoch": 1.7360637087599544, + "grad_norm": 2.332708041188332, + "learning_rate": 9.136886461426363e-07, + "loss": 0.1075, + "step": 7630 + }, + { + "epoch": 1.7362912400455062, + "grad_norm": 1.9520991750590064, + "learning_rate": 9.136093947261659e-07, + "loss": 0.2023, + "step": 7631 + }, + { + "epoch": 1.736518771331058, + "grad_norm": 3.361016751047298, + "learning_rate": 9.135301374110482e-07, + "loss": 0.1957, + "step": 7632 + }, + { + "epoch": 1.7367463026166097, + "grad_norm": 2.0303541910091347, + "learning_rate": 9.134508741989028e-07, + "loss": 0.0697, + "step": 7633 + }, + { + "epoch": 1.7369738339021614, + "grad_norm": 2.0687313535754317, + "learning_rate": 9.133716050913499e-07, + "loss": 0.0699, + "step": 7634 + }, + { + "epoch": 1.7372013651877132, + "grad_norm": 1.8929361088138872, + "learning_rate": 9.132923300900096e-07, + "loss": 0.087, + "step": 7635 + }, + { + "epoch": 1.737428896473265, + "grad_norm": 1.82190683609358, + "learning_rate": 9.132130491965019e-07, + "loss": 0.0963, + "step": 7636 + }, + { + "epoch": 1.7376564277588167, + "grad_norm": 1.784844620846255, + "learning_rate": 9.131337624124473e-07, + "loss": 0.1103, + "step": 7637 + }, + { + "epoch": 1.7378839590443687, + "grad_norm": 1.6832982433867738, + "learning_rate": 9.130544697394662e-07, + "loss": 0.0598, + "step": 7638 + }, + { + "epoch": 1.7381114903299204, + "grad_norm": 1.6649544579559832, + "learning_rate": 9.129751711791796e-07, + "loss": 0.0494, + "step": 7639 + }, + { + "epoch": 1.7383390216154722, + "grad_norm": 1.910777998489734, + "learning_rate": 9.128958667332076e-07, + "loss": 0.0807, + "step": 7640 + }, + { + "epoch": 1.738566552901024, + "grad_norm": 2.7566149417687336, + "learning_rate": 9.128165564031715e-07, + "loss": 0.1526, + "step": 7641 + }, + { + "epoch": 1.7387940841865757, + "grad_norm": 1.9829806241230596, + "learning_rate": 9.127372401906919e-07, + "loss": 0.0676, + "step": 7642 + }, + { + "epoch": 1.7390216154721274, + "grad_norm": 1.2411223705529266, + "learning_rate": 9.126579180973904e-07, + "loss": 0.1025, + "step": 7643 + }, + { + "epoch": 1.7392491467576792, + "grad_norm": 1.9871055456257984, + "learning_rate": 9.125785901248875e-07, + "loss": 0.0926, + "step": 7644 + }, + { + "epoch": 1.739476678043231, + "grad_norm": 1.3381728325764048, + "learning_rate": 9.124992562748051e-07, + "loss": 0.1002, + "step": 7645 + }, + { + "epoch": 1.7397042093287827, + "grad_norm": 2.028450569544906, + "learning_rate": 9.124199165487646e-07, + "loss": 0.1095, + "step": 7646 + }, + { + "epoch": 1.7399317406143344, + "grad_norm": 1.5375677076815077, + "learning_rate": 9.12340570948387e-07, + "loss": 0.1787, + "step": 7647 + }, + { + "epoch": 1.7401592718998864, + "grad_norm": 1.3000928333153594, + "learning_rate": 9.122612194752947e-07, + "loss": 0.0964, + "step": 7648 + }, + { + "epoch": 1.7403868031854381, + "grad_norm": 2.4207627845936783, + "learning_rate": 9.12181862131109e-07, + "loss": 0.0617, + "step": 7649 + }, + { + "epoch": 1.74061433447099, + "grad_norm": 1.9625432494712964, + "learning_rate": 9.121024989174521e-07, + "loss": 0.0534, + "step": 7650 + }, + { + "epoch": 1.7408418657565417, + "grad_norm": 1.671691037993077, + "learning_rate": 9.120231298359458e-07, + "loss": 0.0661, + "step": 7651 + }, + { + "epoch": 1.7410693970420934, + "grad_norm": 1.62650810819884, + "learning_rate": 9.119437548882125e-07, + "loss": 0.1592, + "step": 7652 + }, + { + "epoch": 1.7412969283276452, + "grad_norm": 2.0886290403537813, + "learning_rate": 9.118643740758744e-07, + "loss": 0.0891, + "step": 7653 + }, + { + "epoch": 1.741524459613197, + "grad_norm": 1.0202798082642592, + "learning_rate": 9.117849874005537e-07, + "loss": 0.0609, + "step": 7654 + }, + { + "epoch": 1.7417519908987487, + "grad_norm": 1.1087960114356423, + "learning_rate": 9.117055948638731e-07, + "loss": 0.1328, + "step": 7655 + }, + { + "epoch": 1.7419795221843004, + "grad_norm": 2.337569118992264, + "learning_rate": 9.116261964674553e-07, + "loss": 0.1774, + "step": 7656 + }, + { + "epoch": 1.7422070534698522, + "grad_norm": 0.833425953569515, + "learning_rate": 9.115467922129229e-07, + "loss": 0.0644, + "step": 7657 + }, + { + "epoch": 1.742434584755404, + "grad_norm": 1.8459128416919384, + "learning_rate": 9.114673821018987e-07, + "loss": 0.1487, + "step": 7658 + }, + { + "epoch": 1.7426621160409557, + "grad_norm": 3.0180621961746463, + "learning_rate": 9.113879661360063e-07, + "loss": 0.0997, + "step": 7659 + }, + { + "epoch": 1.7428896473265074, + "grad_norm": 2.365356291433293, + "learning_rate": 9.11308544316868e-07, + "loss": 0.0899, + "step": 7660 + }, + { + "epoch": 1.7431171786120592, + "grad_norm": 0.9947468958069099, + "learning_rate": 9.112291166461076e-07, + "loss": 0.0648, + "step": 7661 + }, + { + "epoch": 1.743344709897611, + "grad_norm": 2.0342009053766645, + "learning_rate": 9.111496831253481e-07, + "loss": 0.1004, + "step": 7662 + }, + { + "epoch": 1.7435722411831627, + "grad_norm": 1.9020869768225852, + "learning_rate": 9.110702437562132e-07, + "loss": 0.0512, + "step": 7663 + }, + { + "epoch": 1.7437997724687144, + "grad_norm": 3.361205986166927, + "learning_rate": 9.109907985403265e-07, + "loss": 0.0713, + "step": 7664 + }, + { + "epoch": 1.7440273037542662, + "grad_norm": 2.2716223567748863, + "learning_rate": 9.109113474793116e-07, + "loss": 0.1792, + "step": 7665 + }, + { + "epoch": 1.744254835039818, + "grad_norm": 1.4167288377703477, + "learning_rate": 9.108318905747924e-07, + "loss": 0.041, + "step": 7666 + }, + { + "epoch": 1.7444823663253697, + "grad_norm": 1.5931326292015675, + "learning_rate": 9.107524278283928e-07, + "loss": 0.0769, + "step": 7667 + }, + { + "epoch": 1.7447098976109214, + "grad_norm": 2.1749947008693757, + "learning_rate": 9.106729592417368e-07, + "loss": 0.0683, + "step": 7668 + }, + { + "epoch": 1.7449374288964732, + "grad_norm": 2.4611060336156023, + "learning_rate": 9.105934848164488e-07, + "loss": 0.0866, + "step": 7669 + }, + { + "epoch": 1.745164960182025, + "grad_norm": 1.4394833957099606, + "learning_rate": 9.105140045541532e-07, + "loss": 0.128, + "step": 7670 + }, + { + "epoch": 1.7453924914675767, + "grad_norm": 1.2704249019967093, + "learning_rate": 9.10434518456474e-07, + "loss": 0.0899, + "step": 7671 + }, + { + "epoch": 1.7456200227531284, + "grad_norm": 1.8382274357405695, + "learning_rate": 9.10355026525036e-07, + "loss": 0.0707, + "step": 7672 + }, + { + "epoch": 1.7458475540386802, + "grad_norm": 1.6516294862787348, + "learning_rate": 9.102755287614639e-07, + "loss": 0.0628, + "step": 7673 + }, + { + "epoch": 1.746075085324232, + "grad_norm": 1.4687435497605381, + "learning_rate": 9.101960251673825e-07, + "loss": 0.0736, + "step": 7674 + }, + { + "epoch": 1.7463026166097837, + "grad_norm": 2.19983896423635, + "learning_rate": 9.101165157444166e-07, + "loss": 0.0943, + "step": 7675 + }, + { + "epoch": 1.7465301478953354, + "grad_norm": 0.9785091789738013, + "learning_rate": 9.100370004941912e-07, + "loss": 0.0403, + "step": 7676 + }, + { + "epoch": 1.7467576791808874, + "grad_norm": 1.5092521481168648, + "learning_rate": 9.099574794183317e-07, + "loss": 0.0863, + "step": 7677 + }, + { + "epoch": 1.7469852104664392, + "grad_norm": 1.543615972124766, + "learning_rate": 9.098779525184631e-07, + "loss": 0.0558, + "step": 7678 + }, + { + "epoch": 1.747212741751991, + "grad_norm": 1.8708202868409067, + "learning_rate": 9.097984197962109e-07, + "loss": 0.0766, + "step": 7679 + }, + { + "epoch": 1.7474402730375427, + "grad_norm": 2.326961905782801, + "learning_rate": 9.097188812532006e-07, + "loss": 0.1163, + "step": 7680 + }, + { + "epoch": 1.7476678043230944, + "grad_norm": 1.221197109664306, + "learning_rate": 9.096393368910578e-07, + "loss": 0.051, + "step": 7681 + }, + { + "epoch": 1.7478953356086462, + "grad_norm": 2.481725149573211, + "learning_rate": 9.095597867114082e-07, + "loss": 0.0864, + "step": 7682 + }, + { + "epoch": 1.748122866894198, + "grad_norm": 1.5652542104869682, + "learning_rate": 9.094802307158777e-07, + "loss": 0.1056, + "step": 7683 + }, + { + "epoch": 1.7483503981797497, + "grad_norm": 1.9462963975521113, + "learning_rate": 9.094006689060924e-07, + "loss": 0.1019, + "step": 7684 + }, + { + "epoch": 1.7485779294653014, + "grad_norm": 1.5304924444892434, + "learning_rate": 9.093211012836782e-07, + "loss": 0.0718, + "step": 7685 + }, + { + "epoch": 1.7488054607508532, + "grad_norm": 1.2722942476418655, + "learning_rate": 9.092415278502614e-07, + "loss": 0.1876, + "step": 7686 + }, + { + "epoch": 1.7490329920364052, + "grad_norm": 1.619476736827669, + "learning_rate": 9.091619486074684e-07, + "loss": 0.0692, + "step": 7687 + }, + { + "epoch": 1.749260523321957, + "grad_norm": 1.4809674215317568, + "learning_rate": 9.090823635569254e-07, + "loss": 0.1654, + "step": 7688 + }, + { + "epoch": 1.7494880546075087, + "grad_norm": 1.9547753713840321, + "learning_rate": 9.090027727002594e-07, + "loss": 0.0753, + "step": 7689 + }, + { + "epoch": 1.7497155858930604, + "grad_norm": 1.205130127056155, + "learning_rate": 9.089231760390968e-07, + "loss": 0.0785, + "step": 7690 + }, + { + "epoch": 1.7499431171786122, + "grad_norm": 1.8649502044332449, + "learning_rate": 9.088435735750643e-07, + "loss": 0.1367, + "step": 7691 + }, + { + "epoch": 1.750170648464164, + "grad_norm": 1.9758421731913736, + "learning_rate": 9.087639653097892e-07, + "loss": 0.1531, + "step": 7692 + }, + { + "epoch": 1.7503981797497157, + "grad_norm": 2.0250963128586807, + "learning_rate": 9.086843512448983e-07, + "loss": 0.1064, + "step": 7693 + }, + { + "epoch": 1.7506257110352674, + "grad_norm": 1.0986110331885017, + "learning_rate": 9.086047313820186e-07, + "loss": 0.0717, + "step": 7694 + }, + { + "epoch": 1.7508532423208192, + "grad_norm": 1.575315504512503, + "learning_rate": 9.085251057227777e-07, + "loss": 0.1608, + "step": 7695 + }, + { + "epoch": 1.751080773606371, + "grad_norm": 1.8499617034128235, + "learning_rate": 9.08445474268803e-07, + "loss": 0.0711, + "step": 7696 + }, + { + "epoch": 1.7513083048919227, + "grad_norm": 1.302390760544348, + "learning_rate": 9.083658370217219e-07, + "loss": 0.1105, + "step": 7697 + }, + { + "epoch": 1.7515358361774744, + "grad_norm": 1.5550174209700023, + "learning_rate": 9.082861939831619e-07, + "loss": 0.174, + "step": 7698 + }, + { + "epoch": 1.7517633674630262, + "grad_norm": 1.9470548649485333, + "learning_rate": 9.08206545154751e-07, + "loss": 0.1269, + "step": 7699 + }, + { + "epoch": 1.751990898748578, + "grad_norm": 1.4523391683511102, + "learning_rate": 9.08126890538117e-07, + "loss": 0.0811, + "step": 7700 + }, + { + "epoch": 1.7522184300341297, + "grad_norm": 1.6124064298453191, + "learning_rate": 9.080472301348878e-07, + "loss": 0.0569, + "step": 7701 + }, + { + "epoch": 1.7524459613196814, + "grad_norm": 1.051165491939012, + "learning_rate": 9.079675639466918e-07, + "loss": 0.0766, + "step": 7702 + }, + { + "epoch": 1.7526734926052332, + "grad_norm": 1.2298198234925792, + "learning_rate": 9.078878919751569e-07, + "loss": 0.1004, + "step": 7703 + }, + { + "epoch": 1.752901023890785, + "grad_norm": 1.1952869279452107, + "learning_rate": 9.078082142219114e-07, + "loss": 0.1095, + "step": 7704 + }, + { + "epoch": 1.7531285551763367, + "grad_norm": 1.6588372426049813, + "learning_rate": 9.077285306885842e-07, + "loss": 0.0828, + "step": 7705 + }, + { + "epoch": 1.7533560864618885, + "grad_norm": 1.8863766574519831, + "learning_rate": 9.076488413768035e-07, + "loss": 0.085, + "step": 7706 + }, + { + "epoch": 1.7535836177474402, + "grad_norm": 1.868998557716498, + "learning_rate": 9.075691462881981e-07, + "loss": 0.1284, + "step": 7707 + }, + { + "epoch": 1.753811149032992, + "grad_norm": 2.575061219358169, + "learning_rate": 9.074894454243968e-07, + "loss": 0.0961, + "step": 7708 + }, + { + "epoch": 1.7540386803185437, + "grad_norm": 2.5451709842460453, + "learning_rate": 9.074097387870289e-07, + "loss": 0.1331, + "step": 7709 + }, + { + "epoch": 1.7542662116040955, + "grad_norm": 3.0446870928807233, + "learning_rate": 9.07330026377723e-07, + "loss": 0.0673, + "step": 7710 + }, + { + "epoch": 1.7544937428896472, + "grad_norm": 1.6906044696739255, + "learning_rate": 9.072503081981081e-07, + "loss": 0.1958, + "step": 7711 + }, + { + "epoch": 1.754721274175199, + "grad_norm": 2.123375136911464, + "learning_rate": 9.07170584249814e-07, + "loss": 0.0578, + "step": 7712 + }, + { + "epoch": 1.7549488054607507, + "grad_norm": 1.7145675039121357, + "learning_rate": 9.070908545344702e-07, + "loss": 0.1375, + "step": 7713 + }, + { + "epoch": 1.7551763367463025, + "grad_norm": 1.8471429747593309, + "learning_rate": 9.070111190537057e-07, + "loss": 0.178, + "step": 7714 + }, + { + "epoch": 1.7554038680318542, + "grad_norm": 1.1550274210194234, + "learning_rate": 9.069313778091504e-07, + "loss": 0.1264, + "step": 7715 + }, + { + "epoch": 1.7556313993174062, + "grad_norm": 2.2122684459057758, + "learning_rate": 9.068516308024343e-07, + "loss": 0.1094, + "step": 7716 + }, + { + "epoch": 1.755858930602958, + "grad_norm": 2.0615514422002703, + "learning_rate": 9.067718780351867e-07, + "loss": 0.1321, + "step": 7717 + }, + { + "epoch": 1.7560864618885097, + "grad_norm": 1.2306617245131113, + "learning_rate": 9.066921195090383e-07, + "loss": 0.0888, + "step": 7718 + }, + { + "epoch": 1.7563139931740614, + "grad_norm": 1.9009024264638439, + "learning_rate": 9.066123552256187e-07, + "loss": 0.1886, + "step": 7719 + }, + { + "epoch": 1.7565415244596132, + "grad_norm": 2.0336857043554883, + "learning_rate": 9.065325851865583e-07, + "loss": 0.0602, + "step": 7720 + }, + { + "epoch": 1.756769055745165, + "grad_norm": 1.716514487313978, + "learning_rate": 9.064528093934874e-07, + "loss": 0.045, + "step": 7721 + }, + { + "epoch": 1.7569965870307167, + "grad_norm": 1.4130334728309684, + "learning_rate": 9.063730278480368e-07, + "loss": 0.0688, + "step": 7722 + }, + { + "epoch": 1.7572241183162685, + "grad_norm": 1.7148947194848858, + "learning_rate": 9.062932405518365e-07, + "loss": 0.067, + "step": 7723 + }, + { + "epoch": 1.7574516496018202, + "grad_norm": 1.1702296450916967, + "learning_rate": 9.062134475065176e-07, + "loss": 0.1344, + "step": 7724 + }, + { + "epoch": 1.757679180887372, + "grad_norm": 2.1482513621808828, + "learning_rate": 9.06133648713711e-07, + "loss": 0.1423, + "step": 7725 + }, + { + "epoch": 1.757906712172924, + "grad_norm": 2.284486576088291, + "learning_rate": 9.060538441750475e-07, + "loss": 0.1495, + "step": 7726 + }, + { + "epoch": 1.7581342434584757, + "grad_norm": 2.191973827420181, + "learning_rate": 9.05974033892158e-07, + "loss": 0.0847, + "step": 7727 + }, + { + "epoch": 1.7583617747440274, + "grad_norm": 1.8911279223088515, + "learning_rate": 9.058942178666738e-07, + "loss": 0.1093, + "step": 7728 + }, + { + "epoch": 1.7585893060295792, + "grad_norm": 1.607789254154104, + "learning_rate": 9.058143961002263e-07, + "loss": 0.0697, + "step": 7729 + }, + { + "epoch": 1.758816837315131, + "grad_norm": 2.531179289862926, + "learning_rate": 9.05734568594447e-07, + "loss": 0.1071, + "step": 7730 + }, + { + "epoch": 1.7590443686006827, + "grad_norm": 1.761839393673622, + "learning_rate": 9.05654735350967e-07, + "loss": 0.0738, + "step": 7731 + }, + { + "epoch": 1.7592718998862344, + "grad_norm": 1.9432446627659616, + "learning_rate": 9.055748963714183e-07, + "loss": 0.0888, + "step": 7732 + }, + { + "epoch": 1.7594994311717862, + "grad_norm": 2.477380296176341, + "learning_rate": 9.054950516574327e-07, + "loss": 0.1222, + "step": 7733 + }, + { + "epoch": 1.759726962457338, + "grad_norm": 2.1426177085938822, + "learning_rate": 9.054152012106417e-07, + "loss": 0.1153, + "step": 7734 + }, + { + "epoch": 1.7599544937428897, + "grad_norm": 1.7579957725375353, + "learning_rate": 9.053353450326777e-07, + "loss": 0.1293, + "step": 7735 + }, + { + "epoch": 1.7601820250284415, + "grad_norm": 1.7277103983923738, + "learning_rate": 9.052554831251725e-07, + "loss": 0.142, + "step": 7736 + }, + { + "epoch": 1.7604095563139932, + "grad_norm": 2.3482069939578434, + "learning_rate": 9.051756154897587e-07, + "loss": 0.0941, + "step": 7737 + }, + { + "epoch": 1.760637087599545, + "grad_norm": 2.2017522392952644, + "learning_rate": 9.050957421280683e-07, + "loss": 0.0769, + "step": 7738 + }, + { + "epoch": 1.7608646188850967, + "grad_norm": 1.2463989976132068, + "learning_rate": 9.05015863041734e-07, + "loss": 0.0541, + "step": 7739 + }, + { + "epoch": 1.7610921501706485, + "grad_norm": 2.1184900749319957, + "learning_rate": 9.049359782323881e-07, + "loss": 0.1823, + "step": 7740 + }, + { + "epoch": 1.7613196814562002, + "grad_norm": 2.343042918418259, + "learning_rate": 9.048560877016637e-07, + "loss": 0.1149, + "step": 7741 + }, + { + "epoch": 1.761547212741752, + "grad_norm": 2.4441312268439574, + "learning_rate": 9.047761914511933e-07, + "loss": 0.0749, + "step": 7742 + }, + { + "epoch": 1.7617747440273037, + "grad_norm": 1.8406444859552782, + "learning_rate": 9.0469628948261e-07, + "loss": 0.0456, + "step": 7743 + }, + { + "epoch": 1.7620022753128555, + "grad_norm": 2.528726454922943, + "learning_rate": 9.046163817975466e-07, + "loss": 0.0819, + "step": 7744 + }, + { + "epoch": 1.7622298065984072, + "grad_norm": 2.233801008165811, + "learning_rate": 9.045364683976366e-07, + "loss": 0.0814, + "step": 7745 + }, + { + "epoch": 1.762457337883959, + "grad_norm": 1.386317934062558, + "learning_rate": 9.044565492845131e-07, + "loss": 0.1425, + "step": 7746 + }, + { + "epoch": 1.7626848691695107, + "grad_norm": 2.3550014134454798, + "learning_rate": 9.043766244598096e-07, + "loss": 0.2095, + "step": 7747 + }, + { + "epoch": 1.7629124004550625, + "grad_norm": 1.7521921266846439, + "learning_rate": 9.042966939251595e-07, + "loss": 0.1088, + "step": 7748 + }, + { + "epoch": 1.7631399317406142, + "grad_norm": 1.5105909426282509, + "learning_rate": 9.042167576821964e-07, + "loss": 0.0839, + "step": 7749 + }, + { + "epoch": 1.763367463026166, + "grad_norm": 1.346098449702512, + "learning_rate": 9.041368157325543e-07, + "loss": 0.1096, + "step": 7750 + }, + { + "epoch": 1.7635949943117177, + "grad_norm": 2.230968633211587, + "learning_rate": 9.040568680778668e-07, + "loss": 0.1296, + "step": 7751 + }, + { + "epoch": 1.7638225255972695, + "grad_norm": 2.008287198164812, + "learning_rate": 9.03976914719768e-07, + "loss": 0.1533, + "step": 7752 + }, + { + "epoch": 1.7640500568828212, + "grad_norm": 2.0286358209472213, + "learning_rate": 9.03896955659892e-07, + "loss": 0.1248, + "step": 7753 + }, + { + "epoch": 1.764277588168373, + "grad_norm": 1.4410538366212453, + "learning_rate": 9.03816990899873e-07, + "loss": 0.1253, + "step": 7754 + }, + { + "epoch": 1.764505119453925, + "grad_norm": 2.3003285393100126, + "learning_rate": 9.037370204413452e-07, + "loss": 0.1108, + "step": 7755 + }, + { + "epoch": 1.7647326507394767, + "grad_norm": 1.8094108627109409, + "learning_rate": 9.036570442859433e-07, + "loss": 0.0957, + "step": 7756 + }, + { + "epoch": 1.7649601820250285, + "grad_norm": 1.8456575539371738, + "learning_rate": 9.035770624353018e-07, + "loss": 0.1355, + "step": 7757 + }, + { + "epoch": 1.7651877133105802, + "grad_norm": 2.063327024936909, + "learning_rate": 9.034970748910552e-07, + "loss": 0.1718, + "step": 7758 + }, + { + "epoch": 1.765415244596132, + "grad_norm": 1.4051691223639915, + "learning_rate": 9.034170816548387e-07, + "loss": 0.0486, + "step": 7759 + }, + { + "epoch": 1.7656427758816837, + "grad_norm": 1.7755553859713349, + "learning_rate": 9.033370827282868e-07, + "loss": 0.1381, + "step": 7760 + }, + { + "epoch": 1.7658703071672355, + "grad_norm": 2.3606055646893367, + "learning_rate": 9.032570781130346e-07, + "loss": 0.0858, + "step": 7761 + }, + { + "epoch": 1.7660978384527872, + "grad_norm": 1.577299531099599, + "learning_rate": 9.031770678107174e-07, + "loss": 0.1253, + "step": 7762 + }, + { + "epoch": 1.766325369738339, + "grad_norm": 1.582327546524722, + "learning_rate": 9.030970518229704e-07, + "loss": 0.1605, + "step": 7763 + }, + { + "epoch": 1.7665529010238907, + "grad_norm": 2.0474199246300224, + "learning_rate": 9.030170301514289e-07, + "loss": 0.0884, + "step": 7764 + }, + { + "epoch": 1.7667804323094427, + "grad_norm": 2.0688842173234083, + "learning_rate": 9.029370027977284e-07, + "loss": 0.0616, + "step": 7765 + }, + { + "epoch": 1.7670079635949945, + "grad_norm": 1.700530417390062, + "learning_rate": 9.028569697635047e-07, + "loss": 0.1148, + "step": 7766 + }, + { + "epoch": 1.7672354948805462, + "grad_norm": 1.2065680171495423, + "learning_rate": 9.027769310503935e-07, + "loss": 0.0862, + "step": 7767 + }, + { + "epoch": 1.767463026166098, + "grad_norm": 2.763378345600242, + "learning_rate": 9.026968866600304e-07, + "loss": 0.0837, + "step": 7768 + }, + { + "epoch": 1.7676905574516497, + "grad_norm": 1.5979987877916608, + "learning_rate": 9.026168365940516e-07, + "loss": 0.1767, + "step": 7769 + }, + { + "epoch": 1.7679180887372015, + "grad_norm": 2.679596696381199, + "learning_rate": 9.02536780854093e-07, + "loss": 0.0819, + "step": 7770 + }, + { + "epoch": 1.7681456200227532, + "grad_norm": 1.8230684306888139, + "learning_rate": 9.024567194417911e-07, + "loss": 0.0637, + "step": 7771 + }, + { + "epoch": 1.768373151308305, + "grad_norm": 1.9631162966685984, + "learning_rate": 9.023766523587817e-07, + "loss": 0.0615, + "step": 7772 + }, + { + "epoch": 1.7686006825938567, + "grad_norm": 3.0762963295111554, + "learning_rate": 9.022965796067016e-07, + "loss": 0.0891, + "step": 7773 + }, + { + "epoch": 1.7688282138794085, + "grad_norm": 1.5944641746686694, + "learning_rate": 9.022165011871873e-07, + "loss": 0.0994, + "step": 7774 + }, + { + "epoch": 1.7690557451649602, + "grad_norm": 1.4729360968816696, + "learning_rate": 9.021364171018754e-07, + "loss": 0.1397, + "step": 7775 + }, + { + "epoch": 1.769283276450512, + "grad_norm": 1.6493723281696528, + "learning_rate": 9.020563273524027e-07, + "loss": 0.0886, + "step": 7776 + }, + { + "epoch": 1.7695108077360637, + "grad_norm": 1.9381157892220884, + "learning_rate": 9.019762319404061e-07, + "loss": 0.1045, + "step": 7777 + }, + { + "epoch": 1.7697383390216155, + "grad_norm": 1.5331658549824572, + "learning_rate": 9.018961308675225e-07, + "loss": 0.097, + "step": 7778 + }, + { + "epoch": 1.7699658703071672, + "grad_norm": 1.7149391038670783, + "learning_rate": 9.018160241353893e-07, + "loss": 0.0906, + "step": 7779 + }, + { + "epoch": 1.770193401592719, + "grad_norm": 1.8887009132890562, + "learning_rate": 9.017359117456434e-07, + "loss": 0.1742, + "step": 7780 + }, + { + "epoch": 1.7704209328782707, + "grad_norm": 2.5088063317551983, + "learning_rate": 9.016557936999221e-07, + "loss": 0.0995, + "step": 7781 + }, + { + "epoch": 1.7706484641638225, + "grad_norm": 1.088562934787203, + "learning_rate": 9.015756699998632e-07, + "loss": 0.1064, + "step": 7782 + }, + { + "epoch": 1.7708759954493742, + "grad_norm": 1.7457599624243116, + "learning_rate": 9.014955406471041e-07, + "loss": 0.0916, + "step": 7783 + }, + { + "epoch": 1.771103526734926, + "grad_norm": 0.990592771673969, + "learning_rate": 9.014154056432828e-07, + "loss": 0.0636, + "step": 7784 + }, + { + "epoch": 1.7713310580204777, + "grad_norm": 1.5216263430612997, + "learning_rate": 9.013352649900365e-07, + "loss": 0.0734, + "step": 7785 + }, + { + "epoch": 1.7715585893060295, + "grad_norm": 1.6761526514507648, + "learning_rate": 9.012551186890037e-07, + "loss": 0.1128, + "step": 7786 + }, + { + "epoch": 1.7717861205915812, + "grad_norm": 2.406746693881849, + "learning_rate": 9.011749667418221e-07, + "loss": 0.108, + "step": 7787 + }, + { + "epoch": 1.772013651877133, + "grad_norm": 2.088337573823908, + "learning_rate": 9.010948091501298e-07, + "loss": 0.104, + "step": 7788 + }, + { + "epoch": 1.7722411831626848, + "grad_norm": 1.2096744129268449, + "learning_rate": 9.010146459155654e-07, + "loss": 0.0491, + "step": 7789 + }, + { + "epoch": 1.7724687144482365, + "grad_norm": 2.1724696253694398, + "learning_rate": 9.009344770397671e-07, + "loss": 0.0906, + "step": 7790 + }, + { + "epoch": 1.7726962457337883, + "grad_norm": 2.720624059626655, + "learning_rate": 9.008543025243735e-07, + "loss": 0.1334, + "step": 7791 + }, + { + "epoch": 1.77292377701934, + "grad_norm": 1.501000526881006, + "learning_rate": 9.007741223710232e-07, + "loss": 0.0542, + "step": 7792 + }, + { + "epoch": 1.7731513083048918, + "grad_norm": 1.8774662298947806, + "learning_rate": 9.006939365813549e-07, + "loss": 0.1842, + "step": 7793 + }, + { + "epoch": 1.7733788395904437, + "grad_norm": 2.182592030310091, + "learning_rate": 9.006137451570074e-07, + "loss": 0.0597, + "step": 7794 + }, + { + "epoch": 1.7736063708759955, + "grad_norm": 2.253307611268069, + "learning_rate": 9.005335480996196e-07, + "loss": 0.1503, + "step": 7795 + }, + { + "epoch": 1.7738339021615472, + "grad_norm": 1.3152085878713347, + "learning_rate": 9.004533454108308e-07, + "loss": 0.0859, + "step": 7796 + }, + { + "epoch": 1.774061433447099, + "grad_norm": 2.3123934197708262, + "learning_rate": 9.0037313709228e-07, + "loss": 0.0783, + "step": 7797 + }, + { + "epoch": 1.7742889647326507, + "grad_norm": 1.257329892595704, + "learning_rate": 9.002929231456067e-07, + "loss": 0.1411, + "step": 7798 + }, + { + "epoch": 1.7745164960182025, + "grad_norm": 1.3867909744443634, + "learning_rate": 9.002127035724502e-07, + "loss": 0.1619, + "step": 7799 + }, + { + "epoch": 1.7747440273037542, + "grad_norm": 1.7551261198500565, + "learning_rate": 9.001324783744501e-07, + "loss": 0.1414, + "step": 7800 + }, + { + "epoch": 1.774971558589306, + "grad_norm": 2.17055077900065, + "learning_rate": 9.000522475532461e-07, + "loss": 0.1175, + "step": 7801 + }, + { + "epoch": 1.7751990898748577, + "grad_norm": 2.261645285178795, + "learning_rate": 8.999720111104776e-07, + "loss": 0.0675, + "step": 7802 + }, + { + "epoch": 1.7754266211604095, + "grad_norm": 1.8841771418200437, + "learning_rate": 8.99891769047785e-07, + "loss": 0.067, + "step": 7803 + }, + { + "epoch": 1.7756541524459615, + "grad_norm": 2.0197317634528975, + "learning_rate": 8.998115213668082e-07, + "loss": 0.0977, + "step": 7804 + }, + { + "epoch": 1.7758816837315132, + "grad_norm": 2.0191667764223644, + "learning_rate": 8.997312680691869e-07, + "loss": 0.0602, + "step": 7805 + }, + { + "epoch": 1.776109215017065, + "grad_norm": 1.3290974204701784, + "learning_rate": 8.996510091565618e-07, + "loss": 0.0557, + "step": 7806 + }, + { + "epoch": 1.7763367463026167, + "grad_norm": 2.1763864519173968, + "learning_rate": 8.99570744630573e-07, + "loss": 0.0754, + "step": 7807 + }, + { + "epoch": 1.7765642775881685, + "grad_norm": 1.4389036394707273, + "learning_rate": 8.99490474492861e-07, + "loss": 0.0845, + "step": 7808 + }, + { + "epoch": 1.7767918088737202, + "grad_norm": 1.9050671159870316, + "learning_rate": 8.994101987450665e-07, + "loss": 0.0838, + "step": 7809 + }, + { + "epoch": 1.777019340159272, + "grad_norm": 1.0723806211617144, + "learning_rate": 8.993299173888302e-07, + "loss": 0.06, + "step": 7810 + }, + { + "epoch": 1.7772468714448237, + "grad_norm": 1.9747451991197136, + "learning_rate": 8.992496304257926e-07, + "loss": 0.0992, + "step": 7811 + }, + { + "epoch": 1.7774744027303755, + "grad_norm": 1.7176621767274736, + "learning_rate": 8.99169337857595e-07, + "loss": 0.0905, + "step": 7812 + }, + { + "epoch": 1.7777019340159272, + "grad_norm": 1.0753173089734889, + "learning_rate": 8.990890396858781e-07, + "loss": 0.033, + "step": 7813 + }, + { + "epoch": 1.777929465301479, + "grad_norm": 1.7152862439179297, + "learning_rate": 8.990087359122832e-07, + "loss": 0.1307, + "step": 7814 + }, + { + "epoch": 1.7781569965870307, + "grad_norm": 1.95552270558147, + "learning_rate": 8.989284265384515e-07, + "loss": 0.1575, + "step": 7815 + }, + { + "epoch": 1.7783845278725825, + "grad_norm": 1.8848727548966473, + "learning_rate": 8.988481115660247e-07, + "loss": 0.0718, + "step": 7816 + }, + { + "epoch": 1.7786120591581343, + "grad_norm": 1.3525083622427532, + "learning_rate": 8.987677909966439e-07, + "loss": 0.1044, + "step": 7817 + }, + { + "epoch": 1.778839590443686, + "grad_norm": 2.2268446862759537, + "learning_rate": 8.986874648319507e-07, + "loss": 0.0853, + "step": 7818 + }, + { + "epoch": 1.7790671217292378, + "grad_norm": 2.8297802361696567, + "learning_rate": 8.986071330735872e-07, + "loss": 0.0838, + "step": 7819 + }, + { + "epoch": 1.7792946530147895, + "grad_norm": 2.612831240136297, + "learning_rate": 8.985267957231947e-07, + "loss": 0.1672, + "step": 7820 + }, + { + "epoch": 1.7795221843003413, + "grad_norm": 1.1944218658112027, + "learning_rate": 8.984464527824157e-07, + "loss": 0.0988, + "step": 7821 + }, + { + "epoch": 1.779749715585893, + "grad_norm": 1.9194825227954213, + "learning_rate": 8.983661042528917e-07, + "loss": 0.0917, + "step": 7822 + }, + { + "epoch": 1.7799772468714448, + "grad_norm": 2.19105148559906, + "learning_rate": 8.982857501362655e-07, + "loss": 0.0869, + "step": 7823 + }, + { + "epoch": 1.7802047781569965, + "grad_norm": 1.6695883304194064, + "learning_rate": 8.982053904341789e-07, + "loss": 0.0941, + "step": 7824 + }, + { + "epoch": 1.7804323094425483, + "grad_norm": 1.5502404185987204, + "learning_rate": 8.981250251482746e-07, + "loss": 0.0646, + "step": 7825 + }, + { + "epoch": 1.7806598407281, + "grad_norm": 2.256945452811573, + "learning_rate": 8.980446542801947e-07, + "loss": 0.0921, + "step": 7826 + }, + { + "epoch": 1.7808873720136518, + "grad_norm": 1.4129414140934224, + "learning_rate": 8.979642778315824e-07, + "loss": 0.094, + "step": 7827 + }, + { + "epoch": 1.7811149032992035, + "grad_norm": 1.2346827183794675, + "learning_rate": 8.9788389580408e-07, + "loss": 0.0791, + "step": 7828 + }, + { + "epoch": 1.7813424345847553, + "grad_norm": 2.0776026778754493, + "learning_rate": 8.978035081993307e-07, + "loss": 0.1026, + "step": 7829 + }, + { + "epoch": 1.781569965870307, + "grad_norm": 1.648641965139214, + "learning_rate": 8.977231150189772e-07, + "loss": 0.0896, + "step": 7830 + }, + { + "epoch": 1.7817974971558588, + "grad_norm": 1.7606901431693627, + "learning_rate": 8.976427162646628e-07, + "loss": 0.1071, + "step": 7831 + }, + { + "epoch": 1.7820250284414105, + "grad_norm": 1.0437463347997784, + "learning_rate": 8.975623119380304e-07, + "loss": 0.0864, + "step": 7832 + }, + { + "epoch": 1.7822525597269625, + "grad_norm": 1.4175356073598935, + "learning_rate": 8.974819020407237e-07, + "loss": 0.0533, + "step": 7833 + }, + { + "epoch": 1.7824800910125143, + "grad_norm": 1.7466844691413699, + "learning_rate": 8.974014865743859e-07, + "loss": 0.0661, + "step": 7834 + }, + { + "epoch": 1.782707622298066, + "grad_norm": 1.7236750667746046, + "learning_rate": 8.973210655406605e-07, + "loss": 0.1125, + "step": 7835 + }, + { + "epoch": 1.7829351535836178, + "grad_norm": 1.6443849825820265, + "learning_rate": 8.972406389411915e-07, + "loss": 0.1002, + "step": 7836 + }, + { + "epoch": 1.7831626848691695, + "grad_norm": 1.6149835042504506, + "learning_rate": 8.971602067776222e-07, + "loss": 0.0743, + "step": 7837 + }, + { + "epoch": 1.7833902161547213, + "grad_norm": 1.2815799918515915, + "learning_rate": 8.970797690515967e-07, + "loss": 0.0743, + "step": 7838 + }, + { + "epoch": 1.783617747440273, + "grad_norm": 1.9971918347069777, + "learning_rate": 8.969993257647591e-07, + "loss": 0.0738, + "step": 7839 + }, + { + "epoch": 1.7838452787258248, + "grad_norm": 2.003342173446247, + "learning_rate": 8.969188769187534e-07, + "loss": 0.114, + "step": 7840 + }, + { + "epoch": 1.7840728100113765, + "grad_norm": 4.793773386642825, + "learning_rate": 8.96838422515224e-07, + "loss": 0.0991, + "step": 7841 + }, + { + "epoch": 1.7843003412969285, + "grad_norm": 2.2809709648649803, + "learning_rate": 8.967579625558148e-07, + "loss": 0.0549, + "step": 7842 + }, + { + "epoch": 1.7845278725824802, + "grad_norm": 1.7635893669409994, + "learning_rate": 8.966774970421708e-07, + "loss": 0.1327, + "step": 7843 + }, + { + "epoch": 1.784755403868032, + "grad_norm": 2.394614637747171, + "learning_rate": 8.965970259759363e-07, + "loss": 0.1148, + "step": 7844 + }, + { + "epoch": 1.7849829351535837, + "grad_norm": 1.930436802586443, + "learning_rate": 8.965165493587557e-07, + "loss": 0.0819, + "step": 7845 + }, + { + "epoch": 1.7852104664391355, + "grad_norm": 1.4801002678171222, + "learning_rate": 8.964360671922743e-07, + "loss": 0.1262, + "step": 7846 + }, + { + "epoch": 1.7854379977246873, + "grad_norm": 1.7059687688784455, + "learning_rate": 8.963555794781369e-07, + "loss": 0.0745, + "step": 7847 + }, + { + "epoch": 1.785665529010239, + "grad_norm": 1.5711566617884094, + "learning_rate": 8.962750862179883e-07, + "loss": 0.1109, + "step": 7848 + }, + { + "epoch": 1.7858930602957908, + "grad_norm": 1.706377672983305, + "learning_rate": 8.961945874134738e-07, + "loss": 0.0554, + "step": 7849 + }, + { + "epoch": 1.7861205915813425, + "grad_norm": 2.039101020748713, + "learning_rate": 8.961140830662386e-07, + "loss": 0.131, + "step": 7850 + }, + { + "epoch": 1.7863481228668943, + "grad_norm": 1.59291616283066, + "learning_rate": 8.960335731779281e-07, + "loss": 0.0962, + "step": 7851 + }, + { + "epoch": 1.786575654152446, + "grad_norm": 2.174229002035782, + "learning_rate": 8.959530577501875e-07, + "loss": 0.1294, + "step": 7852 + }, + { + "epoch": 1.7868031854379978, + "grad_norm": 1.2599519473724012, + "learning_rate": 8.958725367846628e-07, + "loss": 0.0621, + "step": 7853 + }, + { + "epoch": 1.7870307167235495, + "grad_norm": 2.5449372656277425, + "learning_rate": 8.957920102829997e-07, + "loss": 0.0728, + "step": 7854 + }, + { + "epoch": 1.7872582480091013, + "grad_norm": 2.952238728396471, + "learning_rate": 8.957114782468436e-07, + "loss": 0.1664, + "step": 7855 + }, + { + "epoch": 1.787485779294653, + "grad_norm": 1.5952903547857578, + "learning_rate": 8.956309406778407e-07, + "loss": 0.0595, + "step": 7856 + }, + { + "epoch": 1.7877133105802048, + "grad_norm": 1.7272271436452793, + "learning_rate": 8.955503975776371e-07, + "loss": 0.0651, + "step": 7857 + }, + { + "epoch": 1.7879408418657565, + "grad_norm": 1.8654300815282825, + "learning_rate": 8.954698489478788e-07, + "loss": 0.0871, + "step": 7858 + }, + { + "epoch": 1.7881683731513083, + "grad_norm": 1.318743783757732, + "learning_rate": 8.953892947902121e-07, + "loss": 0.079, + "step": 7859 + }, + { + "epoch": 1.78839590443686, + "grad_norm": 2.0132943524891567, + "learning_rate": 8.953087351062835e-07, + "loss": 0.0739, + "step": 7860 + }, + { + "epoch": 1.7886234357224118, + "grad_norm": 2.289347746053406, + "learning_rate": 8.952281698977394e-07, + "loss": 0.1019, + "step": 7861 + }, + { + "epoch": 1.7888509670079635, + "grad_norm": 1.7022126119195158, + "learning_rate": 8.951475991662263e-07, + "loss": 0.1015, + "step": 7862 + }, + { + "epoch": 1.7890784982935153, + "grad_norm": 2.650537981470732, + "learning_rate": 8.950670229133912e-07, + "loss": 0.1347, + "step": 7863 + }, + { + "epoch": 1.789306029579067, + "grad_norm": 5.940113863866019, + "learning_rate": 8.949864411408807e-07, + "loss": 0.2453, + "step": 7864 + }, + { + "epoch": 1.7895335608646188, + "grad_norm": 1.1622668271144168, + "learning_rate": 8.949058538503416e-07, + "loss": 0.0779, + "step": 7865 + }, + { + "epoch": 1.7897610921501705, + "grad_norm": 0.9652518854804722, + "learning_rate": 8.948252610434213e-07, + "loss": 0.0557, + "step": 7866 + }, + { + "epoch": 1.7899886234357223, + "grad_norm": 2.246060745645692, + "learning_rate": 8.947446627217669e-07, + "loss": 0.0764, + "step": 7867 + }, + { + "epoch": 1.790216154721274, + "grad_norm": 1.3878867152062166, + "learning_rate": 8.946640588870254e-07, + "loss": 0.1098, + "step": 7868 + }, + { + "epoch": 1.7904436860068258, + "grad_norm": 1.5529127906610505, + "learning_rate": 8.945834495408447e-07, + "loss": 0.0889, + "step": 7869 + }, + { + "epoch": 1.7906712172923775, + "grad_norm": 1.2415713797537098, + "learning_rate": 8.945028346848718e-07, + "loss": 0.1012, + "step": 7870 + }, + { + "epoch": 1.7908987485779293, + "grad_norm": 1.7786519501363127, + "learning_rate": 8.944222143207545e-07, + "loss": 0.099, + "step": 7871 + }, + { + "epoch": 1.7911262798634813, + "grad_norm": 1.4685412189747176, + "learning_rate": 8.943415884501407e-07, + "loss": 0.0719, + "step": 7872 + }, + { + "epoch": 1.791353811149033, + "grad_norm": 1.8654576674097252, + "learning_rate": 8.942609570746781e-07, + "loss": 0.1275, + "step": 7873 + }, + { + "epoch": 1.7915813424345848, + "grad_norm": 1.591036046248092, + "learning_rate": 8.941803201960146e-07, + "loss": 0.0888, + "step": 7874 + }, + { + "epoch": 1.7918088737201365, + "grad_norm": 1.6317920806584023, + "learning_rate": 8.940996778157983e-07, + "loss": 0.1372, + "step": 7875 + }, + { + "epoch": 1.7920364050056883, + "grad_norm": 1.6429130913792436, + "learning_rate": 8.940190299356774e-07, + "loss": 0.128, + "step": 7876 + }, + { + "epoch": 1.79226393629124, + "grad_norm": 2.1205434237500875, + "learning_rate": 8.939383765573004e-07, + "loss": 0.1155, + "step": 7877 + }, + { + "epoch": 1.7924914675767918, + "grad_norm": 1.0015403399612886, + "learning_rate": 8.938577176823154e-07, + "loss": 0.062, + "step": 7878 + }, + { + "epoch": 1.7927189988623435, + "grad_norm": 1.5861665019909275, + "learning_rate": 8.93777053312371e-07, + "loss": 0.0942, + "step": 7879 + }, + { + "epoch": 1.7929465301478953, + "grad_norm": 2.0196755272152225, + "learning_rate": 8.936963834491161e-07, + "loss": 0.0852, + "step": 7880 + }, + { + "epoch": 1.7931740614334473, + "grad_norm": 1.5719758145617422, + "learning_rate": 8.93615708094199e-07, + "loss": 0.0635, + "step": 7881 + }, + { + "epoch": 1.793401592718999, + "grad_norm": 2.7461205577430485, + "learning_rate": 8.935350272492687e-07, + "loss": 0.1522, + "step": 7882 + }, + { + "epoch": 1.7936291240045508, + "grad_norm": 1.5693797211524632, + "learning_rate": 8.934543409159743e-07, + "loss": 0.0636, + "step": 7883 + }, + { + "epoch": 1.7938566552901025, + "grad_norm": 1.2564947826522308, + "learning_rate": 8.933736490959649e-07, + "loss": 0.0618, + "step": 7884 + }, + { + "epoch": 1.7940841865756543, + "grad_norm": 1.1241818055947628, + "learning_rate": 8.932929517908896e-07, + "loss": 0.0482, + "step": 7885 + }, + { + "epoch": 1.794311717861206, + "grad_norm": 1.4114770467917908, + "learning_rate": 8.932122490023977e-07, + "loss": 0.096, + "step": 7886 + }, + { + "epoch": 1.7945392491467578, + "grad_norm": 2.1205223878064157, + "learning_rate": 8.931315407321387e-07, + "loss": 0.117, + "step": 7887 + }, + { + "epoch": 1.7947667804323095, + "grad_norm": 1.5491352131872387, + "learning_rate": 8.93050826981762e-07, + "loss": 0.1023, + "step": 7888 + }, + { + "epoch": 1.7949943117178613, + "grad_norm": 1.2732550601865582, + "learning_rate": 8.929701077529173e-07, + "loss": 0.0614, + "step": 7889 + }, + { + "epoch": 1.795221843003413, + "grad_norm": 2.7047939410685093, + "learning_rate": 8.928893830472544e-07, + "loss": 0.0976, + "step": 7890 + }, + { + "epoch": 1.7954493742889648, + "grad_norm": 1.5563267855594476, + "learning_rate": 8.92808652866423e-07, + "loss": 0.0975, + "step": 7891 + }, + { + "epoch": 1.7956769055745165, + "grad_norm": 1.3613707002326496, + "learning_rate": 8.927279172120734e-07, + "loss": 0.0786, + "step": 7892 + }, + { + "epoch": 1.7959044368600683, + "grad_norm": 2.8711013984874154, + "learning_rate": 8.926471760858554e-07, + "loss": 0.1526, + "step": 7893 + }, + { + "epoch": 1.79613196814562, + "grad_norm": 1.5532830349646967, + "learning_rate": 8.925664294894193e-07, + "loss": 0.1193, + "step": 7894 + }, + { + "epoch": 1.7963594994311718, + "grad_norm": 1.5363693463316634, + "learning_rate": 8.924856774244154e-07, + "loss": 0.124, + "step": 7895 + }, + { + "epoch": 1.7965870307167235, + "grad_norm": 1.4901127201141373, + "learning_rate": 8.92404919892494e-07, + "loss": 0.0796, + "step": 7896 + }, + { + "epoch": 1.7968145620022753, + "grad_norm": 1.433497031131219, + "learning_rate": 8.923241568953061e-07, + "loss": 0.13, + "step": 7897 + }, + { + "epoch": 1.797042093287827, + "grad_norm": 1.9449335414827038, + "learning_rate": 8.922433884345018e-07, + "loss": 0.0619, + "step": 7898 + }, + { + "epoch": 1.7972696245733788, + "grad_norm": 1.275474603473048, + "learning_rate": 8.921626145117321e-07, + "loss": 0.1032, + "step": 7899 + }, + { + "epoch": 1.7974971558589306, + "grad_norm": 2.121545544925719, + "learning_rate": 8.920818351286479e-07, + "loss": 0.1586, + "step": 7900 + }, + { + "epoch": 1.7977246871444823, + "grad_norm": 2.4854523522884606, + "learning_rate": 8.920010502868999e-07, + "loss": 0.1011, + "step": 7901 + }, + { + "epoch": 1.797952218430034, + "grad_norm": 1.8256396602848413, + "learning_rate": 8.919202599881395e-07, + "loss": 0.0729, + "step": 7902 + }, + { + "epoch": 1.7981797497155858, + "grad_norm": 2.1225175531463947, + "learning_rate": 8.918394642340179e-07, + "loss": 0.0888, + "step": 7903 + }, + { + "epoch": 1.7984072810011376, + "grad_norm": 1.1052579331476957, + "learning_rate": 8.917586630261864e-07, + "loss": 0.052, + "step": 7904 + }, + { + "epoch": 1.7986348122866893, + "grad_norm": 2.6200628722142207, + "learning_rate": 8.916778563662963e-07, + "loss": 0.1277, + "step": 7905 + }, + { + "epoch": 1.798862343572241, + "grad_norm": 1.9287472476517964, + "learning_rate": 8.915970442559993e-07, + "loss": 0.0981, + "step": 7906 + }, + { + "epoch": 1.7990898748577928, + "grad_norm": 1.189291751694118, + "learning_rate": 8.915162266969469e-07, + "loss": 0.0909, + "step": 7907 + }, + { + "epoch": 1.7993174061433446, + "grad_norm": 1.618233610564045, + "learning_rate": 8.91435403690791e-07, + "loss": 0.0687, + "step": 7908 + }, + { + "epoch": 1.7995449374288963, + "grad_norm": 2.8755569609856084, + "learning_rate": 8.913545752391832e-07, + "loss": 0.0838, + "step": 7909 + }, + { + "epoch": 1.799772468714448, + "grad_norm": 3.768457228350488, + "learning_rate": 8.912737413437758e-07, + "loss": 0.0996, + "step": 7910 + }, + { + "epoch": 1.8, + "grad_norm": 1.2436444240934987, + "learning_rate": 8.91192902006221e-07, + "loss": 0.0592, + "step": 7911 + }, + { + "epoch": 1.8002275312855518, + "grad_norm": 1.3041810375702834, + "learning_rate": 8.911120572281705e-07, + "loss": 0.0754, + "step": 7912 + }, + { + "epoch": 1.8004550625711035, + "grad_norm": 1.6467738523949054, + "learning_rate": 8.91031207011277e-07, + "loss": 0.073, + "step": 7913 + }, + { + "epoch": 1.8006825938566553, + "grad_norm": 1.8420486910028369, + "learning_rate": 8.90950351357193e-07, + "loss": 0.054, + "step": 7914 + }, + { + "epoch": 1.800910125142207, + "grad_norm": 1.916477915303434, + "learning_rate": 8.908694902675706e-07, + "loss": 0.1207, + "step": 7915 + }, + { + "epoch": 1.8011376564277588, + "grad_norm": 1.0329327989450314, + "learning_rate": 8.907886237440627e-07, + "loss": 0.0659, + "step": 7916 + }, + { + "epoch": 1.8013651877133106, + "grad_norm": 2.55890915319135, + "learning_rate": 8.907077517883225e-07, + "loss": 0.1255, + "step": 7917 + }, + { + "epoch": 1.8015927189988623, + "grad_norm": 1.0708527594861654, + "learning_rate": 8.906268744020022e-07, + "loss": 0.1292, + "step": 7918 + }, + { + "epoch": 1.801820250284414, + "grad_norm": 1.768678572289282, + "learning_rate": 8.905459915867551e-07, + "loss": 0.1178, + "step": 7919 + }, + { + "epoch": 1.802047781569966, + "grad_norm": 1.7701801384501021, + "learning_rate": 8.904651033442342e-07, + "loss": 0.0865, + "step": 7920 + }, + { + "epoch": 1.8022753128555178, + "grad_norm": 1.209791177338225, + "learning_rate": 8.903842096760929e-07, + "loss": 0.13, + "step": 7921 + }, + { + "epoch": 1.8025028441410695, + "grad_norm": 1.9309745239711726, + "learning_rate": 8.903033105839842e-07, + "loss": 0.1219, + "step": 7922 + }, + { + "epoch": 1.8027303754266213, + "grad_norm": 1.791549375354808, + "learning_rate": 8.902224060695619e-07, + "loss": 0.1401, + "step": 7923 + }, + { + "epoch": 1.802957906712173, + "grad_norm": 2.2309786352695107, + "learning_rate": 8.901414961344792e-07, + "loss": 0.1285, + "step": 7924 + }, + { + "epoch": 1.8031854379977248, + "grad_norm": 1.5413508883349476, + "learning_rate": 8.900605807803901e-07, + "loss": 0.0976, + "step": 7925 + }, + { + "epoch": 1.8034129692832765, + "grad_norm": 2.2230731549669596, + "learning_rate": 8.89979660008948e-07, + "loss": 0.1132, + "step": 7926 + }, + { + "epoch": 1.8036405005688283, + "grad_norm": 1.6901971188862963, + "learning_rate": 8.898987338218069e-07, + "loss": 0.161, + "step": 7927 + }, + { + "epoch": 1.80386803185438, + "grad_norm": 1.562612297701134, + "learning_rate": 8.89817802220621e-07, + "loss": 0.0824, + "step": 7928 + }, + { + "epoch": 1.8040955631399318, + "grad_norm": 1.886126267735738, + "learning_rate": 8.89736865207044e-07, + "loss": 0.1123, + "step": 7929 + }, + { + "epoch": 1.8043230944254836, + "grad_norm": 1.2601843398994992, + "learning_rate": 8.896559227827305e-07, + "loss": 0.0966, + "step": 7930 + }, + { + "epoch": 1.8045506257110353, + "grad_norm": 1.6232214146708215, + "learning_rate": 8.895749749493346e-07, + "loss": 0.0818, + "step": 7931 + }, + { + "epoch": 1.804778156996587, + "grad_norm": 1.398048235791363, + "learning_rate": 8.894940217085106e-07, + "loss": 0.048, + "step": 7932 + }, + { + "epoch": 1.8050056882821388, + "grad_norm": 1.5243613828322313, + "learning_rate": 8.894130630619133e-07, + "loss": 0.0731, + "step": 7933 + }, + { + "epoch": 1.8052332195676906, + "grad_norm": 1.1676191057862337, + "learning_rate": 8.893320990111972e-07, + "loss": 0.0715, + "step": 7934 + }, + { + "epoch": 1.8054607508532423, + "grad_norm": 1.5462430254914166, + "learning_rate": 8.892511295580172e-07, + "loss": 0.1072, + "step": 7935 + }, + { + "epoch": 1.805688282138794, + "grad_norm": 1.9019514172502328, + "learning_rate": 8.891701547040281e-07, + "loss": 0.1753, + "step": 7936 + }, + { + "epoch": 1.8059158134243458, + "grad_norm": 1.9237291621109978, + "learning_rate": 8.890891744508847e-07, + "loss": 0.0644, + "step": 7937 + }, + { + "epoch": 1.8061433447098976, + "grad_norm": 2.5070157979610697, + "learning_rate": 8.890081888002424e-07, + "loss": 0.1119, + "step": 7938 + }, + { + "epoch": 1.8063708759954493, + "grad_norm": 1.9063518086360378, + "learning_rate": 8.88927197753756e-07, + "loss": 0.069, + "step": 7939 + }, + { + "epoch": 1.806598407281001, + "grad_norm": 1.2918093421228658, + "learning_rate": 8.888462013130811e-07, + "loss": 0.0747, + "step": 7940 + }, + { + "epoch": 1.8068259385665528, + "grad_norm": 1.445252862411054, + "learning_rate": 8.887651994798732e-07, + "loss": 0.1199, + "step": 7941 + }, + { + "epoch": 1.8070534698521046, + "grad_norm": 2.3338068209044995, + "learning_rate": 8.886841922557876e-07, + "loss": 0.0807, + "step": 7942 + }, + { + "epoch": 1.8072810011376563, + "grad_norm": 2.4976203732381865, + "learning_rate": 8.8860317964248e-07, + "loss": 0.0824, + "step": 7943 + }, + { + "epoch": 1.807508532423208, + "grad_norm": 1.4425565983008646, + "learning_rate": 8.885221616416063e-07, + "loss": 0.1075, + "step": 7944 + }, + { + "epoch": 1.8077360637087598, + "grad_norm": 1.3669743032172474, + "learning_rate": 8.884411382548221e-07, + "loss": 0.0789, + "step": 7945 + }, + { + "epoch": 1.8079635949943116, + "grad_norm": 2.624662078998394, + "learning_rate": 8.883601094837835e-07, + "loss": 0.1702, + "step": 7946 + }, + { + "epoch": 1.8081911262798633, + "grad_norm": 2.4385008822736336, + "learning_rate": 8.882790753301465e-07, + "loss": 0.1041, + "step": 7947 + }, + { + "epoch": 1.808418657565415, + "grad_norm": 1.8857003325098818, + "learning_rate": 8.881980357955676e-07, + "loss": 0.0705, + "step": 7948 + }, + { + "epoch": 1.808646188850967, + "grad_norm": 2.2324107717841972, + "learning_rate": 8.881169908817028e-07, + "loss": 0.1169, + "step": 7949 + }, + { + "epoch": 1.8088737201365188, + "grad_norm": 1.98085870686822, + "learning_rate": 8.880359405902085e-07, + "loss": 0.0931, + "step": 7950 + }, + { + "epoch": 1.8091012514220706, + "grad_norm": 1.7089610091975043, + "learning_rate": 8.879548849227413e-07, + "loss": 0.0488, + "step": 7951 + }, + { + "epoch": 1.8093287827076223, + "grad_norm": 2.205016197511881, + "learning_rate": 8.87873823880958e-07, + "loss": 0.1264, + "step": 7952 + }, + { + "epoch": 1.809556313993174, + "grad_norm": 2.425531264655183, + "learning_rate": 8.877927574665149e-07, + "loss": 0.0676, + "step": 7953 + }, + { + "epoch": 1.8097838452787258, + "grad_norm": 1.9803598255840258, + "learning_rate": 8.877116856810693e-07, + "loss": 0.1153, + "step": 7954 + }, + { + "epoch": 1.8100113765642776, + "grad_norm": 1.49947563040702, + "learning_rate": 8.876306085262781e-07, + "loss": 0.0548, + "step": 7955 + }, + { + "epoch": 1.8102389078498293, + "grad_norm": 2.0261667389538323, + "learning_rate": 8.875495260037979e-07, + "loss": 0.1146, + "step": 7956 + }, + { + "epoch": 1.810466439135381, + "grad_norm": 2.1813142108063803, + "learning_rate": 8.874684381152865e-07, + "loss": 0.0813, + "step": 7957 + }, + { + "epoch": 1.8106939704209328, + "grad_norm": 1.4894661526116098, + "learning_rate": 8.873873448624008e-07, + "loss": 0.071, + "step": 7958 + }, + { + "epoch": 1.8109215017064848, + "grad_norm": 1.9943222444261415, + "learning_rate": 8.873062462467983e-07, + "loss": 0.0804, + "step": 7959 + }, + { + "epoch": 1.8111490329920366, + "grad_norm": 1.5808149503128233, + "learning_rate": 8.872251422701366e-07, + "loss": 0.066, + "step": 7960 + }, + { + "epoch": 1.8113765642775883, + "grad_norm": 1.6329075256511885, + "learning_rate": 8.871440329340733e-07, + "loss": 0.0919, + "step": 7961 + }, + { + "epoch": 1.81160409556314, + "grad_norm": 1.374945952782231, + "learning_rate": 8.870629182402659e-07, + "loss": 0.1439, + "step": 7962 + }, + { + "epoch": 1.8118316268486918, + "grad_norm": 1.819629299014224, + "learning_rate": 8.869817981903725e-07, + "loss": 0.1289, + "step": 7963 + }, + { + "epoch": 1.8120591581342436, + "grad_norm": 1.9037894605610242, + "learning_rate": 8.869006727860508e-07, + "loss": 0.1254, + "step": 7964 + }, + { + "epoch": 1.8122866894197953, + "grad_norm": 2.6305291415119134, + "learning_rate": 8.868195420289591e-07, + "loss": 0.0888, + "step": 7965 + }, + { + "epoch": 1.812514220705347, + "grad_norm": 1.2718606094342204, + "learning_rate": 8.867384059207554e-07, + "loss": 0.0732, + "step": 7966 + }, + { + "epoch": 1.8127417519908988, + "grad_norm": 1.7677749741199136, + "learning_rate": 8.86657264463098e-07, + "loss": 0.0579, + "step": 7967 + }, + { + "epoch": 1.8129692832764506, + "grad_norm": 2.6051676714289633, + "learning_rate": 8.865761176576457e-07, + "loss": 0.1816, + "step": 7968 + }, + { + "epoch": 1.8131968145620023, + "grad_norm": 2.5904520494446706, + "learning_rate": 8.864949655060562e-07, + "loss": 0.068, + "step": 7969 + }, + { + "epoch": 1.813424345847554, + "grad_norm": 2.402814540167061, + "learning_rate": 8.864138080099886e-07, + "loss": 0.1039, + "step": 7970 + }, + { + "epoch": 1.8136518771331058, + "grad_norm": 1.413902540697436, + "learning_rate": 8.863326451711015e-07, + "loss": 0.0425, + "step": 7971 + }, + { + "epoch": 1.8138794084186576, + "grad_norm": 1.9231350711585855, + "learning_rate": 8.862514769910538e-07, + "loss": 0.0771, + "step": 7972 + }, + { + "epoch": 1.8141069397042093, + "grad_norm": 1.4378931294990538, + "learning_rate": 8.861703034715042e-07, + "loss": 0.0451, + "step": 7973 + }, + { + "epoch": 1.814334470989761, + "grad_norm": 1.5674244634740826, + "learning_rate": 8.860891246141123e-07, + "loss": 0.068, + "step": 7974 + }, + { + "epoch": 1.8145620022753128, + "grad_norm": 1.8947060156558089, + "learning_rate": 8.860079404205367e-07, + "loss": 0.06, + "step": 7975 + }, + { + "epoch": 1.8147895335608646, + "grad_norm": 1.404481266144173, + "learning_rate": 8.859267508924366e-07, + "loss": 0.0756, + "step": 7976 + }, + { + "epoch": 1.8150170648464163, + "grad_norm": 2.189011388774002, + "learning_rate": 8.858455560314718e-07, + "loss": 0.0879, + "step": 7977 + }, + { + "epoch": 1.815244596131968, + "grad_norm": 1.5985167291388305, + "learning_rate": 8.857643558393015e-07, + "loss": 0.0925, + "step": 7978 + }, + { + "epoch": 1.8154721274175198, + "grad_norm": 1.6872714253032772, + "learning_rate": 8.856831503175852e-07, + "loss": 0.1196, + "step": 7979 + }, + { + "epoch": 1.8156996587030716, + "grad_norm": 2.412942583341033, + "learning_rate": 8.856019394679828e-07, + "loss": 0.088, + "step": 7980 + }, + { + "epoch": 1.8159271899886233, + "grad_norm": 2.059832124378441, + "learning_rate": 8.855207232921541e-07, + "loss": 0.1585, + "step": 7981 + }, + { + "epoch": 1.816154721274175, + "grad_norm": 1.8426611964824577, + "learning_rate": 8.854395017917588e-07, + "loss": 0.1427, + "step": 7982 + }, + { + "epoch": 1.8163822525597269, + "grad_norm": 1.3887679776448114, + "learning_rate": 8.853582749684571e-07, + "loss": 0.0922, + "step": 7983 + }, + { + "epoch": 1.8166097838452786, + "grad_norm": 2.2636010215981512, + "learning_rate": 8.852770428239091e-07, + "loss": 0.0895, + "step": 7984 + }, + { + "epoch": 1.8168373151308304, + "grad_norm": 1.921207164884614, + "learning_rate": 8.851958053597751e-07, + "loss": 0.117, + "step": 7985 + }, + { + "epoch": 1.817064846416382, + "grad_norm": 1.586916399892704, + "learning_rate": 8.851145625777153e-07, + "loss": 0.0935, + "step": 7986 + }, + { + "epoch": 1.8172923777019339, + "grad_norm": 1.35753517250393, + "learning_rate": 8.850333144793903e-07, + "loss": 0.0597, + "step": 7987 + }, + { + "epoch": 1.8175199089874858, + "grad_norm": 1.5035937327144726, + "learning_rate": 8.849520610664605e-07, + "loss": 0.0486, + "step": 7988 + }, + { + "epoch": 1.8177474402730376, + "grad_norm": 1.9056294709428654, + "learning_rate": 8.848708023405866e-07, + "loss": 0.1184, + "step": 7989 + }, + { + "epoch": 1.8179749715585893, + "grad_norm": 2.2830969435991455, + "learning_rate": 8.847895383034294e-07, + "loss": 0.0697, + "step": 7990 + }, + { + "epoch": 1.818202502844141, + "grad_norm": 1.758115942769437, + "learning_rate": 8.847082689566499e-07, + "loss": 0.0525, + "step": 7991 + }, + { + "epoch": 1.8184300341296928, + "grad_norm": 2.331603106136869, + "learning_rate": 8.846269943019091e-07, + "loss": 0.0888, + "step": 7992 + }, + { + "epoch": 1.8186575654152446, + "grad_norm": 1.5640397677378846, + "learning_rate": 8.84545714340868e-07, + "loss": 0.1167, + "step": 7993 + }, + { + "epoch": 1.8188850967007963, + "grad_norm": 2.090933926102918, + "learning_rate": 8.844644290751877e-07, + "loss": 0.1279, + "step": 7994 + }, + { + "epoch": 1.819112627986348, + "grad_norm": 1.8004666049038967, + "learning_rate": 8.843831385065298e-07, + "loss": 0.0983, + "step": 7995 + }, + { + "epoch": 1.8193401592718998, + "grad_norm": 1.6621509552129172, + "learning_rate": 8.843018426365555e-07, + "loss": 0.1476, + "step": 7996 + }, + { + "epoch": 1.8195676905574516, + "grad_norm": 1.3210004176459789, + "learning_rate": 8.842205414669264e-07, + "loss": 0.0324, + "step": 7997 + }, + { + "epoch": 1.8197952218430036, + "grad_norm": 1.7428824299417647, + "learning_rate": 8.841392349993041e-07, + "loss": 0.0792, + "step": 7998 + }, + { + "epoch": 1.8200227531285553, + "grad_norm": 0.9945653597861828, + "learning_rate": 8.840579232353506e-07, + "loss": 0.0595, + "step": 7999 + }, + { + "epoch": 1.820250284414107, + "grad_norm": 1.518949966756168, + "learning_rate": 8.839766061767277e-07, + "loss": 0.1414, + "step": 8000 + }, + { + "epoch": 1.8204778156996588, + "grad_norm": 1.2545089380132042, + "learning_rate": 8.838952838250971e-07, + "loss": 0.0652, + "step": 8001 + }, + { + "epoch": 1.8207053469852106, + "grad_norm": 1.682902077980011, + "learning_rate": 8.83813956182121e-07, + "loss": 0.116, + "step": 8002 + }, + { + "epoch": 1.8209328782707623, + "grad_norm": 1.7098678771979696, + "learning_rate": 8.837326232494616e-07, + "loss": 0.0622, + "step": 8003 + }, + { + "epoch": 1.821160409556314, + "grad_norm": 2.246667661732211, + "learning_rate": 8.836512850287812e-07, + "loss": 0.0899, + "step": 8004 + }, + { + "epoch": 1.8213879408418658, + "grad_norm": 2.090778727283746, + "learning_rate": 8.835699415217425e-07, + "loss": 0.1351, + "step": 8005 + }, + { + "epoch": 1.8216154721274176, + "grad_norm": 1.9646684357539137, + "learning_rate": 8.834885927300075e-07, + "loss": 0.0804, + "step": 8006 + }, + { + "epoch": 1.8218430034129693, + "grad_norm": 2.5359712036374513, + "learning_rate": 8.834072386552392e-07, + "loss": 0.1094, + "step": 8007 + }, + { + "epoch": 1.822070534698521, + "grad_norm": 1.79114517493048, + "learning_rate": 8.833258792991001e-07, + "loss": 0.1098, + "step": 8008 + }, + { + "epoch": 1.8222980659840728, + "grad_norm": 2.1882338427516426, + "learning_rate": 8.832445146632531e-07, + "loss": 0.0914, + "step": 8009 + }, + { + "epoch": 1.8225255972696246, + "grad_norm": 1.6716849762117383, + "learning_rate": 8.831631447493612e-07, + "loss": 0.137, + "step": 8010 + }, + { + "epoch": 1.8227531285551763, + "grad_norm": 1.9983528698245776, + "learning_rate": 8.830817695590875e-07, + "loss": 0.119, + "step": 8011 + }, + { + "epoch": 1.822980659840728, + "grad_norm": 1.1983879720021489, + "learning_rate": 8.830003890940953e-07, + "loss": 0.045, + "step": 8012 + }, + { + "epoch": 1.8232081911262799, + "grad_norm": 2.504799737236628, + "learning_rate": 8.829190033560473e-07, + "loss": 0.0951, + "step": 8013 + }, + { + "epoch": 1.8234357224118316, + "grad_norm": 2.326452595122347, + "learning_rate": 8.828376123466072e-07, + "loss": 0.0797, + "step": 8014 + }, + { + "epoch": 1.8236632536973834, + "grad_norm": 1.3881241943485159, + "learning_rate": 8.827562160674387e-07, + "loss": 0.0842, + "step": 8015 + }, + { + "epoch": 1.823890784982935, + "grad_norm": 1.8278684346169267, + "learning_rate": 8.826748145202052e-07, + "loss": 0.1022, + "step": 8016 + }, + { + "epoch": 1.8241183162684869, + "grad_norm": 1.147248074416315, + "learning_rate": 8.8259340770657e-07, + "loss": 0.0953, + "step": 8017 + }, + { + "epoch": 1.8243458475540386, + "grad_norm": 2.0274129863643586, + "learning_rate": 8.82511995628198e-07, + "loss": 0.1525, + "step": 8018 + }, + { + "epoch": 1.8245733788395904, + "grad_norm": 1.6139035468352776, + "learning_rate": 8.824305782867521e-07, + "loss": 0.0783, + "step": 8019 + }, + { + "epoch": 1.8248009101251421, + "grad_norm": 1.6086703437171743, + "learning_rate": 8.823491556838964e-07, + "loss": 0.121, + "step": 8020 + }, + { + "epoch": 1.8250284414106939, + "grad_norm": 2.12903072394807, + "learning_rate": 8.822677278212955e-07, + "loss": 0.0896, + "step": 8021 + }, + { + "epoch": 1.8252559726962456, + "grad_norm": 1.955779110294756, + "learning_rate": 8.821862947006134e-07, + "loss": 0.0823, + "step": 8022 + }, + { + "epoch": 1.8254835039817974, + "grad_norm": 2.025028737283517, + "learning_rate": 8.821048563235143e-07, + "loss": 0.1314, + "step": 8023 + }, + { + "epoch": 1.8257110352673491, + "grad_norm": 1.3606475345993916, + "learning_rate": 8.820234126916631e-07, + "loss": 0.0718, + "step": 8024 + }, + { + "epoch": 1.8259385665529009, + "grad_norm": 1.588216088081773, + "learning_rate": 8.81941963806724e-07, + "loss": 0.125, + "step": 8025 + }, + { + "epoch": 1.8261660978384526, + "grad_norm": 1.006748908249921, + "learning_rate": 8.818605096703614e-07, + "loss": 0.0426, + "step": 8026 + }, + { + "epoch": 1.8263936291240046, + "grad_norm": 2.261848902386564, + "learning_rate": 8.817790502842406e-07, + "loss": 0.0784, + "step": 8027 + }, + { + "epoch": 1.8266211604095564, + "grad_norm": 1.3516735433864935, + "learning_rate": 8.816975856500264e-07, + "loss": 0.1547, + "step": 8028 + }, + { + "epoch": 1.826848691695108, + "grad_norm": 1.9302802634188536, + "learning_rate": 8.816161157693837e-07, + "loss": 0.1135, + "step": 8029 + }, + { + "epoch": 1.8270762229806599, + "grad_norm": 2.551159549795896, + "learning_rate": 8.815346406439773e-07, + "loss": 0.1004, + "step": 8030 + }, + { + "epoch": 1.8273037542662116, + "grad_norm": 3.3378461756304882, + "learning_rate": 8.814531602754728e-07, + "loss": 0.0821, + "step": 8031 + }, + { + "epoch": 1.8275312855517634, + "grad_norm": 1.7083657595140067, + "learning_rate": 8.813716746655354e-07, + "loss": 0.0905, + "step": 8032 + }, + { + "epoch": 1.8277588168373151, + "grad_norm": 1.9588276728965104, + "learning_rate": 8.812901838158304e-07, + "loss": 0.1355, + "step": 8033 + }, + { + "epoch": 1.8279863481228669, + "grad_norm": 1.266204130784061, + "learning_rate": 8.812086877280234e-07, + "loss": 0.0404, + "step": 8034 + }, + { + "epoch": 1.8282138794084186, + "grad_norm": 2.178652141034485, + "learning_rate": 8.811271864037802e-07, + "loss": 0.0816, + "step": 8035 + }, + { + "epoch": 1.8284414106939704, + "grad_norm": 1.7132494402020864, + "learning_rate": 8.810456798447664e-07, + "loss": 0.0716, + "step": 8036 + }, + { + "epoch": 1.8286689419795223, + "grad_norm": 1.8796367243061582, + "learning_rate": 8.809641680526477e-07, + "loss": 0.1086, + "step": 8037 + }, + { + "epoch": 1.828896473265074, + "grad_norm": 2.8230478404980235, + "learning_rate": 8.808826510290904e-07, + "loss": 0.0811, + "step": 8038 + }, + { + "epoch": 1.8291240045506258, + "grad_norm": 1.4886710711072335, + "learning_rate": 8.808011287757601e-07, + "loss": 0.0494, + "step": 8039 + }, + { + "epoch": 1.8293515358361776, + "grad_norm": 1.4082184039834797, + "learning_rate": 8.807196012943231e-07, + "loss": 0.0808, + "step": 8040 + }, + { + "epoch": 1.8295790671217294, + "grad_norm": 1.8820887098503434, + "learning_rate": 8.80638068586446e-07, + "loss": 0.0786, + "step": 8041 + }, + { + "epoch": 1.829806598407281, + "grad_norm": 1.791583784109299, + "learning_rate": 8.805565306537949e-07, + "loss": 0.1724, + "step": 8042 + }, + { + "epoch": 1.8300341296928329, + "grad_norm": 2.0567022001992714, + "learning_rate": 8.804749874980364e-07, + "loss": 0.1747, + "step": 8043 + }, + { + "epoch": 1.8302616609783846, + "grad_norm": 1.5407941639122054, + "learning_rate": 8.80393439120837e-07, + "loss": 0.0866, + "step": 8044 + }, + { + "epoch": 1.8304891922639364, + "grad_norm": 3.2108287944549856, + "learning_rate": 8.803118855238635e-07, + "loss": 0.139, + "step": 8045 + }, + { + "epoch": 1.8307167235494881, + "grad_norm": 1.571261699714713, + "learning_rate": 8.802303267087825e-07, + "loss": 0.1423, + "step": 8046 + }, + { + "epoch": 1.8309442548350399, + "grad_norm": 2.204022760838511, + "learning_rate": 8.801487626772611e-07, + "loss": 0.0613, + "step": 8047 + }, + { + "epoch": 1.8311717861205916, + "grad_norm": 1.4603157962210571, + "learning_rate": 8.800671934309663e-07, + "loss": 0.1155, + "step": 8048 + }, + { + "epoch": 1.8313993174061434, + "grad_norm": 2.439806649731172, + "learning_rate": 8.799856189715653e-07, + "loss": 0.0698, + "step": 8049 + }, + { + "epoch": 1.8316268486916951, + "grad_norm": 3.8365577402784767, + "learning_rate": 8.79904039300725e-07, + "loss": 0.095, + "step": 8050 + }, + { + "epoch": 1.8318543799772469, + "grad_norm": 2.871749940142669, + "learning_rate": 8.798224544201132e-07, + "loss": 0.0761, + "step": 8051 + }, + { + "epoch": 1.8320819112627986, + "grad_norm": 3.2286627056288624, + "learning_rate": 8.79740864331397e-07, + "loss": 0.1399, + "step": 8052 + }, + { + "epoch": 1.8323094425483504, + "grad_norm": 1.04856166067485, + "learning_rate": 8.796592690362439e-07, + "loss": 0.0368, + "step": 8053 + }, + { + "epoch": 1.8325369738339021, + "grad_norm": 1.2119852843619046, + "learning_rate": 8.795776685363219e-07, + "loss": 0.0695, + "step": 8054 + }, + { + "epoch": 1.8327645051194539, + "grad_norm": 1.6382039063021019, + "learning_rate": 8.794960628332986e-07, + "loss": 0.0906, + "step": 8055 + }, + { + "epoch": 1.8329920364050056, + "grad_norm": 1.2671269042796347, + "learning_rate": 8.794144519288419e-07, + "loss": 0.0773, + "step": 8056 + }, + { + "epoch": 1.8332195676905574, + "grad_norm": 1.8428690789141955, + "learning_rate": 8.793328358246198e-07, + "loss": 0.2364, + "step": 8057 + }, + { + "epoch": 1.8334470989761091, + "grad_norm": 1.4422639160578108, + "learning_rate": 8.792512145223002e-07, + "loss": 0.0852, + "step": 8058 + }, + { + "epoch": 1.8336746302616609, + "grad_norm": 1.0749626632834002, + "learning_rate": 8.791695880235515e-07, + "loss": 0.0813, + "step": 8059 + }, + { + "epoch": 1.8339021615472126, + "grad_norm": 1.8125034413887882, + "learning_rate": 8.790879563300416e-07, + "loss": 0.2156, + "step": 8060 + }, + { + "epoch": 1.8341296928327644, + "grad_norm": 1.5235577300099732, + "learning_rate": 8.790063194434397e-07, + "loss": 0.0545, + "step": 8061 + }, + { + "epoch": 1.8343572241183161, + "grad_norm": 2.187927437051437, + "learning_rate": 8.789246773654136e-07, + "loss": 0.1461, + "step": 8062 + }, + { + "epoch": 1.834584755403868, + "grad_norm": 1.2562342000446287, + "learning_rate": 8.788430300976321e-07, + "loss": 0.0845, + "step": 8063 + }, + { + "epoch": 1.8348122866894196, + "grad_norm": 2.5351377175024266, + "learning_rate": 8.787613776417639e-07, + "loss": 0.1025, + "step": 8064 + }, + { + "epoch": 1.8350398179749714, + "grad_norm": 1.6372928328601253, + "learning_rate": 8.78679719999478e-07, + "loss": 0.0626, + "step": 8065 + }, + { + "epoch": 1.8352673492605234, + "grad_norm": 1.9498306238851026, + "learning_rate": 8.785980571724433e-07, + "loss": 0.0868, + "step": 8066 + }, + { + "epoch": 1.8354948805460751, + "grad_norm": 1.1706345924977686, + "learning_rate": 8.785163891623284e-07, + "loss": 0.0475, + "step": 8067 + }, + { + "epoch": 1.8357224118316269, + "grad_norm": 1.5229949811289203, + "learning_rate": 8.784347159708033e-07, + "loss": 0.0921, + "step": 8068 + }, + { + "epoch": 1.8359499431171786, + "grad_norm": 2.0738291750855704, + "learning_rate": 8.783530375995364e-07, + "loss": 0.0811, + "step": 8069 + }, + { + "epoch": 1.8361774744027304, + "grad_norm": 2.767989298322564, + "learning_rate": 8.782713540501975e-07, + "loss": 0.0767, + "step": 8070 + }, + { + "epoch": 1.8364050056882821, + "grad_norm": 1.8211505086205266, + "learning_rate": 8.78189665324456e-07, + "loss": 0.1038, + "step": 8071 + }, + { + "epoch": 1.8366325369738339, + "grad_norm": 1.1072562267315125, + "learning_rate": 8.781079714239812e-07, + "loss": 0.1055, + "step": 8072 + }, + { + "epoch": 1.8368600682593856, + "grad_norm": 2.080814912516335, + "learning_rate": 8.780262723504434e-07, + "loss": 0.1409, + "step": 8073 + }, + { + "epoch": 1.8370875995449374, + "grad_norm": 2.72947192873807, + "learning_rate": 8.779445681055115e-07, + "loss": 0.0946, + "step": 8074 + }, + { + "epoch": 1.8373151308304891, + "grad_norm": 1.5289441377360764, + "learning_rate": 8.778628586908563e-07, + "loss": 0.1339, + "step": 8075 + }, + { + "epoch": 1.8375426621160411, + "grad_norm": 1.8774735399981626, + "learning_rate": 8.777811441081475e-07, + "loss": 0.1135, + "step": 8076 + }, + { + "epoch": 1.8377701934015929, + "grad_norm": 1.7574623462253978, + "learning_rate": 8.776994243590545e-07, + "loss": 0.0762, + "step": 8077 + }, + { + "epoch": 1.8379977246871446, + "grad_norm": 1.6150293370965112, + "learning_rate": 8.776176994452485e-07, + "loss": 0.0845, + "step": 8078 + }, + { + "epoch": 1.8382252559726964, + "grad_norm": 2.5156319072151168, + "learning_rate": 8.775359693683991e-07, + "loss": 0.1245, + "step": 8079 + }, + { + "epoch": 1.8384527872582481, + "grad_norm": 1.4177266066688612, + "learning_rate": 8.77454234130177e-07, + "loss": 0.1527, + "step": 8080 + }, + { + "epoch": 1.8386803185437999, + "grad_norm": 1.528124365633614, + "learning_rate": 8.773724937322531e-07, + "loss": 0.1792, + "step": 8081 + }, + { + "epoch": 1.8389078498293516, + "grad_norm": 1.6384604180830433, + "learning_rate": 8.772907481762973e-07, + "loss": 0.1178, + "step": 8082 + }, + { + "epoch": 1.8391353811149034, + "grad_norm": 1.4534282416732556, + "learning_rate": 8.772089974639806e-07, + "loss": 0.1404, + "step": 8083 + }, + { + "epoch": 1.8393629124004551, + "grad_norm": 2.137834308161238, + "learning_rate": 8.77127241596974e-07, + "loss": 0.0846, + "step": 8084 + }, + { + "epoch": 1.8395904436860069, + "grad_norm": 1.3227030583321073, + "learning_rate": 8.770454805769482e-07, + "loss": 0.0343, + "step": 8085 + }, + { + "epoch": 1.8398179749715586, + "grad_norm": 2.437576573768376, + "learning_rate": 8.769637144055745e-07, + "loss": 0.1037, + "step": 8086 + }, + { + "epoch": 1.8400455062571104, + "grad_norm": 2.089350321464631, + "learning_rate": 8.768819430845241e-07, + "loss": 0.0672, + "step": 8087 + }, + { + "epoch": 1.8402730375426621, + "grad_norm": 1.185033433378712, + "learning_rate": 8.768001666154678e-07, + "loss": 0.1027, + "step": 8088 + }, + { + "epoch": 1.840500568828214, + "grad_norm": 1.5535944254783345, + "learning_rate": 8.767183850000774e-07, + "loss": 0.0863, + "step": 8089 + }, + { + "epoch": 1.8407281001137656, + "grad_norm": 1.9827049694014909, + "learning_rate": 8.76636598240024e-07, + "loss": 0.057, + "step": 8090 + }, + { + "epoch": 1.8409556313993174, + "grad_norm": 1.883304759862909, + "learning_rate": 8.765548063369795e-07, + "loss": 0.1045, + "step": 8091 + }, + { + "epoch": 1.8411831626848691, + "grad_norm": 2.435086253809783, + "learning_rate": 8.764730092926156e-07, + "loss": 0.14, + "step": 8092 + }, + { + "epoch": 1.841410693970421, + "grad_norm": 2.169725599820099, + "learning_rate": 8.763912071086039e-07, + "loss": 0.0668, + "step": 8093 + }, + { + "epoch": 1.8416382252559726, + "grad_norm": 3.0719306843376613, + "learning_rate": 8.763093997866162e-07, + "loss": 0.1174, + "step": 8094 + }, + { + "epoch": 1.8418657565415244, + "grad_norm": 2.123605048024908, + "learning_rate": 8.762275873283248e-07, + "loss": 0.0827, + "step": 8095 + }, + { + "epoch": 1.8420932878270762, + "grad_norm": 1.2168845416302847, + "learning_rate": 8.761457697354015e-07, + "loss": 0.0356, + "step": 8096 + }, + { + "epoch": 1.842320819112628, + "grad_norm": 1.9241216472684115, + "learning_rate": 8.760639470095185e-07, + "loss": 0.0668, + "step": 8097 + }, + { + "epoch": 1.8425483503981797, + "grad_norm": 1.9828801227804087, + "learning_rate": 8.759821191523483e-07, + "loss": 0.1372, + "step": 8098 + }, + { + "epoch": 1.8427758816837314, + "grad_norm": 1.3275853472693522, + "learning_rate": 8.759002861655633e-07, + "loss": 0.1272, + "step": 8099 + }, + { + "epoch": 1.8430034129692832, + "grad_norm": 3.054475774822718, + "learning_rate": 8.75818448050836e-07, + "loss": 0.0964, + "step": 8100 + }, + { + "epoch": 1.843230944254835, + "grad_norm": 2.6397589735313844, + "learning_rate": 8.757366048098389e-07, + "loss": 0.1831, + "step": 8101 + }, + { + "epoch": 1.8434584755403867, + "grad_norm": 2.8524811972808144, + "learning_rate": 8.756547564442449e-07, + "loss": 0.0668, + "step": 8102 + }, + { + "epoch": 1.8436860068259384, + "grad_norm": 1.8411565631383815, + "learning_rate": 8.755729029557266e-07, + "loss": 0.0599, + "step": 8103 + }, + { + "epoch": 1.8439135381114902, + "grad_norm": 1.7505721085259205, + "learning_rate": 8.75491044345957e-07, + "loss": 0.1314, + "step": 8104 + }, + { + "epoch": 1.8441410693970421, + "grad_norm": 1.427770337506339, + "learning_rate": 8.754091806166092e-07, + "loss": 0.0457, + "step": 8105 + }, + { + "epoch": 1.844368600682594, + "grad_norm": 1.062378521807983, + "learning_rate": 8.753273117693567e-07, + "loss": 0.1159, + "step": 8106 + }, + { + "epoch": 1.8445961319681456, + "grad_norm": 2.1411890853403746, + "learning_rate": 8.752454378058721e-07, + "loss": 0.1575, + "step": 8107 + }, + { + "epoch": 1.8448236632536974, + "grad_norm": 1.9368703837820778, + "learning_rate": 8.751635587278291e-07, + "loss": 0.1211, + "step": 8108 + }, + { + "epoch": 1.8450511945392492, + "grad_norm": 1.4777317556999319, + "learning_rate": 8.750816745369012e-07, + "loss": 0.1091, + "step": 8109 + }, + { + "epoch": 1.845278725824801, + "grad_norm": 2.2453409816350494, + "learning_rate": 8.749997852347619e-07, + "loss": 0.133, + "step": 8110 + }, + { + "epoch": 1.8455062571103527, + "grad_norm": 2.5939468446279057, + "learning_rate": 8.749178908230845e-07, + "loss": 0.0728, + "step": 8111 + }, + { + "epoch": 1.8457337883959044, + "grad_norm": 2.175685537074212, + "learning_rate": 8.748359913035436e-07, + "loss": 0.1728, + "step": 8112 + }, + { + "epoch": 1.8459613196814562, + "grad_norm": 1.0204058719803912, + "learning_rate": 8.747540866778124e-07, + "loss": 0.0245, + "step": 8113 + }, + { + "epoch": 1.846188850967008, + "grad_norm": 1.629186033060192, + "learning_rate": 8.74672176947565e-07, + "loss": 0.0789, + "step": 8114 + }, + { + "epoch": 1.8464163822525599, + "grad_norm": 1.7530913750858943, + "learning_rate": 8.745902621144755e-07, + "loss": 0.0708, + "step": 8115 + }, + { + "epoch": 1.8466439135381116, + "grad_norm": 2.02757935650334, + "learning_rate": 8.745083421802183e-07, + "loss": 0.0856, + "step": 8116 + }, + { + "epoch": 1.8468714448236634, + "grad_norm": 1.527710679169608, + "learning_rate": 8.744264171464673e-07, + "loss": 0.0879, + "step": 8117 + }, + { + "epoch": 1.8470989761092151, + "grad_norm": 1.3835096798236566, + "learning_rate": 8.743444870148974e-07, + "loss": 0.0809, + "step": 8118 + }, + { + "epoch": 1.847326507394767, + "grad_norm": 1.3878336575616186, + "learning_rate": 8.742625517871828e-07, + "loss": 0.1416, + "step": 8119 + }, + { + "epoch": 1.8475540386803186, + "grad_norm": 1.5434914988676236, + "learning_rate": 8.74180611464998e-07, + "loss": 0.0582, + "step": 8120 + }, + { + "epoch": 1.8477815699658704, + "grad_norm": 1.6401975680980627, + "learning_rate": 8.74098666050018e-07, + "loss": 0.0453, + "step": 8121 + }, + { + "epoch": 1.8480091012514221, + "grad_norm": 1.1470988287444377, + "learning_rate": 8.740167155439173e-07, + "loss": 0.0505, + "step": 8122 + }, + { + "epoch": 1.848236632536974, + "grad_norm": 1.2615553854964674, + "learning_rate": 8.739347599483712e-07, + "loss": 0.0671, + "step": 8123 + }, + { + "epoch": 1.8484641638225257, + "grad_norm": 1.751915429159614, + "learning_rate": 8.738527992650542e-07, + "loss": 0.1325, + "step": 8124 + }, + { + "epoch": 1.8486916951080774, + "grad_norm": 1.7203620218025986, + "learning_rate": 8.737708334956421e-07, + "loss": 0.0897, + "step": 8125 + }, + { + "epoch": 1.8489192263936292, + "grad_norm": 2.006545073891573, + "learning_rate": 8.736888626418097e-07, + "loss": 0.0572, + "step": 8126 + }, + { + "epoch": 1.849146757679181, + "grad_norm": 1.7854996069190847, + "learning_rate": 8.736068867052322e-07, + "loss": 0.1387, + "step": 8127 + }, + { + "epoch": 1.8493742889647327, + "grad_norm": 1.5870892229388842, + "learning_rate": 8.735249056875852e-07, + "loss": 0.1861, + "step": 8128 + }, + { + "epoch": 1.8496018202502844, + "grad_norm": 2.1956369428640157, + "learning_rate": 8.734429195905446e-07, + "loss": 0.0882, + "step": 8129 + }, + { + "epoch": 1.8498293515358362, + "grad_norm": 2.278245974868355, + "learning_rate": 8.733609284157855e-07, + "loss": 0.1547, + "step": 8130 + }, + { + "epoch": 1.850056882821388, + "grad_norm": 2.010801737068543, + "learning_rate": 8.73278932164984e-07, + "loss": 0.066, + "step": 8131 + }, + { + "epoch": 1.8502844141069397, + "grad_norm": 2.251868297942089, + "learning_rate": 8.731969308398158e-07, + "loss": 0.0541, + "step": 8132 + }, + { + "epoch": 1.8505119453924914, + "grad_norm": 0.9874246845573058, + "learning_rate": 8.731149244419568e-07, + "loss": 0.0544, + "step": 8133 + }, + { + "epoch": 1.8507394766780432, + "grad_norm": 2.054501652787974, + "learning_rate": 8.73032912973083e-07, + "loss": 0.1195, + "step": 8134 + }, + { + "epoch": 1.850967007963595, + "grad_norm": 1.407276108240153, + "learning_rate": 8.72950896434871e-07, + "loss": 0.0751, + "step": 8135 + }, + { + "epoch": 1.8511945392491467, + "grad_norm": 1.3362315246063559, + "learning_rate": 8.728688748289966e-07, + "loss": 0.1127, + "step": 8136 + }, + { + "epoch": 1.8514220705346984, + "grad_norm": 2.4612186065451125, + "learning_rate": 8.727868481571365e-07, + "loss": 0.2134, + "step": 8137 + }, + { + "epoch": 1.8516496018202502, + "grad_norm": 2.1682087913507306, + "learning_rate": 8.72704816420967e-07, + "loss": 0.1247, + "step": 8138 + }, + { + "epoch": 1.851877133105802, + "grad_norm": 1.7255568864832886, + "learning_rate": 8.726227796221646e-07, + "loss": 0.0508, + "step": 8139 + }, + { + "epoch": 1.8521046643913537, + "grad_norm": 2.407003427099894, + "learning_rate": 8.725407377624062e-07, + "loss": 0.0689, + "step": 8140 + }, + { + "epoch": 1.8523321956769054, + "grad_norm": 2.082513378607011, + "learning_rate": 8.724586908433682e-07, + "loss": 0.0903, + "step": 8141 + }, + { + "epoch": 1.8525597269624572, + "grad_norm": 1.359523606105818, + "learning_rate": 8.72376638866728e-07, + "loss": 0.056, + "step": 8142 + }, + { + "epoch": 1.852787258248009, + "grad_norm": 1.5305548882571467, + "learning_rate": 8.722945818341624e-07, + "loss": 0.1265, + "step": 8143 + }, + { + "epoch": 1.853014789533561, + "grad_norm": 1.5834807080288225, + "learning_rate": 8.722125197473483e-07, + "loss": 0.0803, + "step": 8144 + }, + { + "epoch": 1.8532423208191127, + "grad_norm": 1.2918790662143784, + "learning_rate": 8.721304526079631e-07, + "loss": 0.094, + "step": 8145 + }, + { + "epoch": 1.8534698521046644, + "grad_norm": 1.8366297023243192, + "learning_rate": 8.72048380417684e-07, + "loss": 0.124, + "step": 8146 + }, + { + "epoch": 1.8536973833902162, + "grad_norm": 1.2472104565569448, + "learning_rate": 8.719663031781884e-07, + "loss": 0.0875, + "step": 8147 + }, + { + "epoch": 1.853924914675768, + "grad_norm": 1.9034596129118144, + "learning_rate": 8.71884220891154e-07, + "loss": 0.1596, + "step": 8148 + }, + { + "epoch": 1.8541524459613197, + "grad_norm": 2.470990486257438, + "learning_rate": 8.718021335582583e-07, + "loss": 0.1083, + "step": 8149 + }, + { + "epoch": 1.8543799772468714, + "grad_norm": 0.8588122126598416, + "learning_rate": 8.71720041181179e-07, + "loss": 0.0813, + "step": 8150 + }, + { + "epoch": 1.8546075085324232, + "grad_norm": 1.3830193645898874, + "learning_rate": 8.716379437615937e-07, + "loss": 0.068, + "step": 8151 + }, + { + "epoch": 1.854835039817975, + "grad_norm": 4.054893512600613, + "learning_rate": 8.715558413011807e-07, + "loss": 0.0993, + "step": 8152 + }, + { + "epoch": 1.8550625711035267, + "grad_norm": 1.9216151652590312, + "learning_rate": 8.714737338016178e-07, + "loss": 0.1017, + "step": 8153 + }, + { + "epoch": 1.8552901023890787, + "grad_norm": 1.797804120202323, + "learning_rate": 8.713916212645832e-07, + "loss": 0.0759, + "step": 8154 + }, + { + "epoch": 1.8555176336746304, + "grad_norm": 1.9435258221897045, + "learning_rate": 8.713095036917551e-07, + "loss": 0.0612, + "step": 8155 + }, + { + "epoch": 1.8557451649601822, + "grad_norm": 1.961110798616376, + "learning_rate": 8.712273810848118e-07, + "loss": 0.1092, + "step": 8156 + }, + { + "epoch": 1.855972696245734, + "grad_norm": 1.7601247902120083, + "learning_rate": 8.711452534454318e-07, + "loss": 0.1661, + "step": 8157 + }, + { + "epoch": 1.8562002275312857, + "grad_norm": 2.2464868593263554, + "learning_rate": 8.710631207752936e-07, + "loss": 0.0679, + "step": 8158 + }, + { + "epoch": 1.8564277588168374, + "grad_norm": 3.337665078096119, + "learning_rate": 8.709809830760759e-07, + "loss": 0.1932, + "step": 8159 + }, + { + "epoch": 1.8566552901023892, + "grad_norm": 1.5630086426802838, + "learning_rate": 8.708988403494572e-07, + "loss": 0.0462, + "step": 8160 + }, + { + "epoch": 1.856882821387941, + "grad_norm": 1.7733085293738995, + "learning_rate": 8.708166925971168e-07, + "loss": 0.0939, + "step": 8161 + }, + { + "epoch": 1.8571103526734927, + "grad_norm": 1.979119367608335, + "learning_rate": 8.707345398207332e-07, + "loss": 0.0723, + "step": 8162 + }, + { + "epoch": 1.8573378839590444, + "grad_norm": 1.5396517304164394, + "learning_rate": 8.706523820219858e-07, + "loss": 0.1391, + "step": 8163 + }, + { + "epoch": 1.8575654152445962, + "grad_norm": 2.3463502557091895, + "learning_rate": 8.705702192025537e-07, + "loss": 0.1303, + "step": 8164 + }, + { + "epoch": 1.857792946530148, + "grad_norm": 2.063524500898958, + "learning_rate": 8.704880513641158e-07, + "loss": 0.0778, + "step": 8165 + }, + { + "epoch": 1.8580204778156997, + "grad_norm": 2.4892268289881105, + "learning_rate": 8.704058785083519e-07, + "loss": 0.1097, + "step": 8166 + }, + { + "epoch": 1.8582480091012514, + "grad_norm": 1.6821843849689957, + "learning_rate": 8.703237006369411e-07, + "loss": 0.0654, + "step": 8167 + }, + { + "epoch": 1.8584755403868032, + "grad_norm": 1.779146938217964, + "learning_rate": 8.702415177515633e-07, + "loss": 0.1096, + "step": 8168 + }, + { + "epoch": 1.858703071672355, + "grad_norm": 1.8893828203527572, + "learning_rate": 8.70159329853898e-07, + "loss": 0.0759, + "step": 8169 + }, + { + "epoch": 1.8589306029579067, + "grad_norm": 1.1992146238336574, + "learning_rate": 8.700771369456249e-07, + "loss": 0.096, + "step": 8170 + }, + { + "epoch": 1.8591581342434584, + "grad_norm": 2.788348084768204, + "learning_rate": 8.69994939028424e-07, + "loss": 0.1728, + "step": 8171 + }, + { + "epoch": 1.8593856655290102, + "grad_norm": 3.6376334733716953, + "learning_rate": 8.699127361039753e-07, + "loss": 0.1694, + "step": 8172 + }, + { + "epoch": 1.859613196814562, + "grad_norm": 2.2945377327150513, + "learning_rate": 8.698305281739589e-07, + "loss": 0.0817, + "step": 8173 + }, + { + "epoch": 1.8598407281001137, + "grad_norm": 1.486317017024812, + "learning_rate": 8.697483152400546e-07, + "loss": 0.0676, + "step": 8174 + }, + { + "epoch": 1.8600682593856654, + "grad_norm": 2.280025037072071, + "learning_rate": 8.696660973039432e-07, + "loss": 0.1797, + "step": 8175 + }, + { + "epoch": 1.8602957906712172, + "grad_norm": 1.5402396104562603, + "learning_rate": 8.695838743673048e-07, + "loss": 0.13, + "step": 8176 + }, + { + "epoch": 1.860523321956769, + "grad_norm": 1.6290630688565892, + "learning_rate": 8.695016464318199e-07, + "loss": 0.1265, + "step": 8177 + }, + { + "epoch": 1.8607508532423207, + "grad_norm": 1.8565183527521674, + "learning_rate": 8.69419413499169e-07, + "loss": 0.0746, + "step": 8178 + }, + { + "epoch": 1.8609783845278725, + "grad_norm": 1.3125831060654516, + "learning_rate": 8.693371755710332e-07, + "loss": 0.035, + "step": 8179 + }, + { + "epoch": 1.8612059158134242, + "grad_norm": 1.540824128915698, + "learning_rate": 8.692549326490929e-07, + "loss": 0.0591, + "step": 8180 + }, + { + "epoch": 1.861433447098976, + "grad_norm": 3.5123038339194967, + "learning_rate": 8.69172684735029e-07, + "loss": 0.1065, + "step": 8181 + }, + { + "epoch": 1.8616609783845277, + "grad_norm": 1.3882683308527415, + "learning_rate": 8.690904318305228e-07, + "loss": 0.0453, + "step": 8182 + }, + { + "epoch": 1.8618885096700797, + "grad_norm": 1.3620991320658384, + "learning_rate": 8.690081739372553e-07, + "loss": 0.076, + "step": 8183 + }, + { + "epoch": 1.8621160409556314, + "grad_norm": 1.5799180062922773, + "learning_rate": 8.689259110569072e-07, + "loss": 0.0779, + "step": 8184 + }, + { + "epoch": 1.8623435722411832, + "grad_norm": 1.2709498182853995, + "learning_rate": 8.688436431911604e-07, + "loss": 0.0463, + "step": 8185 + }, + { + "epoch": 1.862571103526735, + "grad_norm": 2.3365313574511855, + "learning_rate": 8.687613703416962e-07, + "loss": 0.1141, + "step": 8186 + }, + { + "epoch": 1.8627986348122867, + "grad_norm": 1.5747577544042832, + "learning_rate": 8.686790925101959e-07, + "loss": 0.0706, + "step": 8187 + }, + { + "epoch": 1.8630261660978384, + "grad_norm": 1.2396408494465165, + "learning_rate": 8.685968096983413e-07, + "loss": 0.1196, + "step": 8188 + }, + { + "epoch": 1.8632536973833902, + "grad_norm": 1.8487311158961117, + "learning_rate": 8.685145219078141e-07, + "loss": 0.1178, + "step": 8189 + }, + { + "epoch": 1.863481228668942, + "grad_norm": 1.4459826163690286, + "learning_rate": 8.684322291402959e-07, + "loss": 0.1308, + "step": 8190 + }, + { + "epoch": 1.8637087599544937, + "grad_norm": 2.1654731303154646, + "learning_rate": 8.683499313974687e-07, + "loss": 0.1005, + "step": 8191 + }, + { + "epoch": 1.8639362912400455, + "grad_norm": 1.5116070135857027, + "learning_rate": 8.682676286810147e-07, + "loss": 0.0352, + "step": 8192 + }, + { + "epoch": 1.8641638225255974, + "grad_norm": 1.2489503106020365, + "learning_rate": 8.68185320992616e-07, + "loss": 0.0629, + "step": 8193 + }, + { + "epoch": 1.8643913538111492, + "grad_norm": 1.2554800050773864, + "learning_rate": 8.681030083339545e-07, + "loss": 0.0884, + "step": 8194 + }, + { + "epoch": 1.864618885096701, + "grad_norm": 1.6767342107089667, + "learning_rate": 8.680206907067129e-07, + "loss": 0.0727, + "step": 8195 + }, + { + "epoch": 1.8648464163822527, + "grad_norm": 2.3704922722420436, + "learning_rate": 8.679383681125735e-07, + "loss": 0.0883, + "step": 8196 + }, + { + "epoch": 1.8650739476678044, + "grad_norm": 1.7957176242096409, + "learning_rate": 8.678560405532186e-07, + "loss": 0.0678, + "step": 8197 + }, + { + "epoch": 1.8653014789533562, + "grad_norm": 1.5453285496923461, + "learning_rate": 8.67773708030331e-07, + "loss": 0.0763, + "step": 8198 + }, + { + "epoch": 1.865529010238908, + "grad_norm": 1.7530154720544147, + "learning_rate": 8.676913705455935e-07, + "loss": 0.119, + "step": 8199 + }, + { + "epoch": 1.8657565415244597, + "grad_norm": 1.7747425306638418, + "learning_rate": 8.676090281006889e-07, + "loss": 0.0616, + "step": 8200 + }, + { + "epoch": 1.8659840728100114, + "grad_norm": 1.4450167421454763, + "learning_rate": 8.675266806972999e-07, + "loss": 0.0934, + "step": 8201 + }, + { + "epoch": 1.8662116040955632, + "grad_norm": 1.4191089951936058, + "learning_rate": 8.674443283371099e-07, + "loss": 0.1293, + "step": 8202 + }, + { + "epoch": 1.866439135381115, + "grad_norm": 2.2180210581811153, + "learning_rate": 8.673619710218018e-07, + "loss": 0.1191, + "step": 8203 + }, + { + "epoch": 1.8666666666666667, + "grad_norm": 0.9959512273970105, + "learning_rate": 8.672796087530589e-07, + "loss": 0.0607, + "step": 8204 + }, + { + "epoch": 1.8668941979522184, + "grad_norm": 1.4140679911460456, + "learning_rate": 8.671972415325644e-07, + "loss": 0.0895, + "step": 8205 + }, + { + "epoch": 1.8671217292377702, + "grad_norm": 2.4634709849571124, + "learning_rate": 8.671148693620019e-07, + "loss": 0.1481, + "step": 8206 + }, + { + "epoch": 1.867349260523322, + "grad_norm": 1.3604140438233554, + "learning_rate": 8.67032492243055e-07, + "loss": 0.0791, + "step": 8207 + }, + { + "epoch": 1.8675767918088737, + "grad_norm": 1.4725557827579294, + "learning_rate": 8.66950110177407e-07, + "loss": 0.054, + "step": 8208 + }, + { + "epoch": 1.8678043230944255, + "grad_norm": 0.9421659920369739, + "learning_rate": 8.668677231667422e-07, + "loss": 0.0192, + "step": 8209 + }, + { + "epoch": 1.8680318543799772, + "grad_norm": 2.439861863056658, + "learning_rate": 8.667853312127439e-07, + "loss": 0.0727, + "step": 8210 + }, + { + "epoch": 1.868259385665529, + "grad_norm": 1.1163814864932045, + "learning_rate": 8.66702934317096e-07, + "loss": 0.0754, + "step": 8211 + }, + { + "epoch": 1.8684869169510807, + "grad_norm": 1.4707594486588582, + "learning_rate": 8.66620532481483e-07, + "loss": 0.1072, + "step": 8212 + }, + { + "epoch": 1.8687144482366325, + "grad_norm": 1.195295377767991, + "learning_rate": 8.66538125707589e-07, + "loss": 0.067, + "step": 8213 + }, + { + "epoch": 1.8689419795221842, + "grad_norm": 1.0068905553577698, + "learning_rate": 8.664557139970978e-07, + "loss": 0.0577, + "step": 8214 + }, + { + "epoch": 1.869169510807736, + "grad_norm": 1.745424099718077, + "learning_rate": 8.663732973516942e-07, + "loss": 0.1158, + "step": 8215 + }, + { + "epoch": 1.8693970420932877, + "grad_norm": 1.9544779077531358, + "learning_rate": 8.662908757730623e-07, + "loss": 0.1273, + "step": 8216 + }, + { + "epoch": 1.8696245733788395, + "grad_norm": 1.9233550606345866, + "learning_rate": 8.66208449262887e-07, + "loss": 0.0765, + "step": 8217 + }, + { + "epoch": 1.8698521046643912, + "grad_norm": 1.7145468545538933, + "learning_rate": 8.661260178228524e-07, + "loss": 0.0641, + "step": 8218 + }, + { + "epoch": 1.870079635949943, + "grad_norm": 1.6994600505497042, + "learning_rate": 8.660435814546439e-07, + "loss": 0.1338, + "step": 8219 + }, + { + "epoch": 1.8703071672354947, + "grad_norm": 1.7977587549342686, + "learning_rate": 8.65961140159946e-07, + "loss": 0.0887, + "step": 8220 + }, + { + "epoch": 1.8705346985210465, + "grad_norm": 1.6649953525826708, + "learning_rate": 8.658786939404435e-07, + "loss": 0.0825, + "step": 8221 + }, + { + "epoch": 1.8707622298065985, + "grad_norm": 2.091920476011299, + "learning_rate": 8.657962427978219e-07, + "loss": 0.0728, + "step": 8222 + }, + { + "epoch": 1.8709897610921502, + "grad_norm": 2.332037808044974, + "learning_rate": 8.657137867337659e-07, + "loss": 0.1176, + "step": 8223 + }, + { + "epoch": 1.871217292377702, + "grad_norm": 1.4201121364278315, + "learning_rate": 8.65631325749961e-07, + "loss": 0.0411, + "step": 8224 + }, + { + "epoch": 1.8714448236632537, + "grad_norm": 1.586031908405511, + "learning_rate": 8.655488598480925e-07, + "loss": 0.1204, + "step": 8225 + }, + { + "epoch": 1.8716723549488055, + "grad_norm": 2.1514924491921326, + "learning_rate": 8.65466389029846e-07, + "loss": 0.1011, + "step": 8226 + }, + { + "epoch": 1.8718998862343572, + "grad_norm": 2.2594262919793526, + "learning_rate": 8.65383913296907e-07, + "loss": 0.0681, + "step": 8227 + }, + { + "epoch": 1.872127417519909, + "grad_norm": 1.2654000276886403, + "learning_rate": 8.653014326509605e-07, + "loss": 0.0517, + "step": 8228 + }, + { + "epoch": 1.8723549488054607, + "grad_norm": 1.3899124560393632, + "learning_rate": 8.652189470936932e-07, + "loss": 0.1193, + "step": 8229 + }, + { + "epoch": 1.8725824800910125, + "grad_norm": 3.2563307631193505, + "learning_rate": 8.651364566267906e-07, + "loss": 0.1597, + "step": 8230 + }, + { + "epoch": 1.8728100113765644, + "grad_norm": 2.6756362088820844, + "learning_rate": 8.650539612519385e-07, + "loss": 0.1332, + "step": 8231 + }, + { + "epoch": 1.8730375426621162, + "grad_norm": 1.3476498608915706, + "learning_rate": 8.64971460970823e-07, + "loss": 0.0375, + "step": 8232 + }, + { + "epoch": 1.873265073947668, + "grad_norm": 1.9234146586811676, + "learning_rate": 8.648889557851306e-07, + "loss": 0.1022, + "step": 8233 + }, + { + "epoch": 1.8734926052332197, + "grad_norm": 2.189947855250575, + "learning_rate": 8.64806445696547e-07, + "loss": 0.0671, + "step": 8234 + }, + { + "epoch": 1.8737201365187715, + "grad_norm": 1.9894848543680852, + "learning_rate": 8.647239307067588e-07, + "loss": 0.1531, + "step": 8235 + }, + { + "epoch": 1.8739476678043232, + "grad_norm": 2.011337725001851, + "learning_rate": 8.646414108174527e-07, + "loss": 0.0964, + "step": 8236 + }, + { + "epoch": 1.874175199089875, + "grad_norm": 2.5605691507555073, + "learning_rate": 8.64558886030315e-07, + "loss": 0.0987, + "step": 8237 + }, + { + "epoch": 1.8744027303754267, + "grad_norm": 2.7815286281993874, + "learning_rate": 8.644763563470324e-07, + "loss": 0.1361, + "step": 8238 + }, + { + "epoch": 1.8746302616609785, + "grad_norm": 3.559819049652831, + "learning_rate": 8.643938217692916e-07, + "loss": 0.1083, + "step": 8239 + }, + { + "epoch": 1.8748577929465302, + "grad_norm": 2.0734532160282946, + "learning_rate": 8.643112822987795e-07, + "loss": 0.1501, + "step": 8240 + }, + { + "epoch": 1.875085324232082, + "grad_norm": 1.796695049035969, + "learning_rate": 8.642287379371831e-07, + "loss": 0.0804, + "step": 8241 + }, + { + "epoch": 1.8753128555176337, + "grad_norm": 2.0981891551441327, + "learning_rate": 8.641461886861893e-07, + "loss": 0.1276, + "step": 8242 + }, + { + "epoch": 1.8755403868031855, + "grad_norm": 2.030917202859018, + "learning_rate": 8.640636345474857e-07, + "loss": 0.1723, + "step": 8243 + }, + { + "epoch": 1.8757679180887372, + "grad_norm": 2.676745078762152, + "learning_rate": 8.639810755227591e-07, + "loss": 0.1073, + "step": 8244 + }, + { + "epoch": 1.875995449374289, + "grad_norm": 3.0043687715907628, + "learning_rate": 8.638985116136968e-07, + "loss": 0.1064, + "step": 8245 + }, + { + "epoch": 1.8762229806598407, + "grad_norm": 1.7613446237818156, + "learning_rate": 8.638159428219866e-07, + "loss": 0.0647, + "step": 8246 + }, + { + "epoch": 1.8764505119453925, + "grad_norm": 2.2085611676970736, + "learning_rate": 8.637333691493159e-07, + "loss": 0.102, + "step": 8247 + }, + { + "epoch": 1.8766780432309442, + "grad_norm": 1.9702990168659118, + "learning_rate": 8.636507905973722e-07, + "loss": 0.0993, + "step": 8248 + }, + { + "epoch": 1.876905574516496, + "grad_norm": 1.7498654561923823, + "learning_rate": 8.635682071678437e-07, + "loss": 0.1733, + "step": 8249 + }, + { + "epoch": 1.8771331058020477, + "grad_norm": 1.830647843822453, + "learning_rate": 8.634856188624177e-07, + "loss": 0.1777, + "step": 8250 + }, + { + "epoch": 1.8773606370875995, + "grad_norm": 1.8382522826065273, + "learning_rate": 8.634030256827825e-07, + "loss": 0.0899, + "step": 8251 + }, + { + "epoch": 1.8775881683731512, + "grad_norm": 2.634208451273832, + "learning_rate": 8.633204276306261e-07, + "loss": 0.0802, + "step": 8252 + }, + { + "epoch": 1.877815699658703, + "grad_norm": 1.8398339448793228, + "learning_rate": 8.632378247076366e-07, + "loss": 0.1009, + "step": 8253 + }, + { + "epoch": 1.8780432309442547, + "grad_norm": 1.6217919155242353, + "learning_rate": 8.631552169155023e-07, + "loss": 0.0379, + "step": 8254 + }, + { + "epoch": 1.8782707622298065, + "grad_norm": 1.3190326043007634, + "learning_rate": 8.630726042559115e-07, + "loss": 0.0573, + "step": 8255 + }, + { + "epoch": 1.8784982935153582, + "grad_norm": 1.8760347738026566, + "learning_rate": 8.629899867305526e-07, + "loss": 0.0821, + "step": 8256 + }, + { + "epoch": 1.87872582480091, + "grad_norm": 1.5065217532584043, + "learning_rate": 8.629073643411145e-07, + "loss": 0.1825, + "step": 8257 + }, + { + "epoch": 1.8789533560864617, + "grad_norm": 1.3124918608997533, + "learning_rate": 8.628247370892853e-07, + "loss": 0.0848, + "step": 8258 + }, + { + "epoch": 1.8791808873720135, + "grad_norm": 1.8729970019792939, + "learning_rate": 8.627421049767541e-07, + "loss": 0.1306, + "step": 8259 + }, + { + "epoch": 1.8794084186575652, + "grad_norm": 1.212882147308011, + "learning_rate": 8.626594680052097e-07, + "loss": 0.0804, + "step": 8260 + }, + { + "epoch": 1.8796359499431172, + "grad_norm": 3.189924617267235, + "learning_rate": 8.62576826176341e-07, + "loss": 0.0968, + "step": 8261 + }, + { + "epoch": 1.879863481228669, + "grad_norm": 1.395245332753133, + "learning_rate": 8.62494179491837e-07, + "loss": 0.1108, + "step": 8262 + }, + { + "epoch": 1.8800910125142207, + "grad_norm": 3.442166694172348, + "learning_rate": 8.624115279533872e-07, + "loss": 0.1112, + "step": 8263 + }, + { + "epoch": 1.8803185437997725, + "grad_norm": 1.4344145011506018, + "learning_rate": 8.623288715626804e-07, + "loss": 0.062, + "step": 8264 + }, + { + "epoch": 1.8805460750853242, + "grad_norm": 2.1182252172781135, + "learning_rate": 8.62246210321406e-07, + "loss": 0.1575, + "step": 8265 + }, + { + "epoch": 1.880773606370876, + "grad_norm": 2.061685485576038, + "learning_rate": 8.621635442312537e-07, + "loss": 0.0696, + "step": 8266 + }, + { + "epoch": 1.8810011376564277, + "grad_norm": 2.3448425692841015, + "learning_rate": 8.620808732939129e-07, + "loss": 0.0739, + "step": 8267 + }, + { + "epoch": 1.8812286689419795, + "grad_norm": 2.4492682818191076, + "learning_rate": 8.619981975110731e-07, + "loss": 0.1094, + "step": 8268 + }, + { + "epoch": 1.8814562002275312, + "grad_norm": 1.4997812182923556, + "learning_rate": 8.619155168844243e-07, + "loss": 0.125, + "step": 8269 + }, + { + "epoch": 1.8816837315130832, + "grad_norm": 1.8390918655122415, + "learning_rate": 8.618328314156564e-07, + "loss": 0.0683, + "step": 8270 + }, + { + "epoch": 1.881911262798635, + "grad_norm": 2.11266005662525, + "learning_rate": 8.617501411064588e-07, + "loss": 0.0787, + "step": 8271 + }, + { + "epoch": 1.8821387940841867, + "grad_norm": 1.5676917774588235, + "learning_rate": 8.616674459585221e-07, + "loss": 0.1448, + "step": 8272 + }, + { + "epoch": 1.8823663253697385, + "grad_norm": 1.7133435613588885, + "learning_rate": 8.615847459735363e-07, + "loss": 0.0943, + "step": 8273 + }, + { + "epoch": 1.8825938566552902, + "grad_norm": 1.7691964747757152, + "learning_rate": 8.615020411531915e-07, + "loss": 0.0799, + "step": 8274 + }, + { + "epoch": 1.882821387940842, + "grad_norm": 1.2155475313958286, + "learning_rate": 8.61419331499178e-07, + "loss": 0.09, + "step": 8275 + }, + { + "epoch": 1.8830489192263937, + "grad_norm": 1.9364963071364532, + "learning_rate": 8.613366170131867e-07, + "loss": 0.1157, + "step": 8276 + }, + { + "epoch": 1.8832764505119455, + "grad_norm": 1.3310492341567115, + "learning_rate": 8.612538976969074e-07, + "loss": 0.0838, + "step": 8277 + }, + { + "epoch": 1.8835039817974972, + "grad_norm": 1.3381461293338204, + "learning_rate": 8.611711735520312e-07, + "loss": 0.0681, + "step": 8278 + }, + { + "epoch": 1.883731513083049, + "grad_norm": 1.46469419190077, + "learning_rate": 8.610884445802488e-07, + "loss": 0.166, + "step": 8279 + }, + { + "epoch": 1.8839590443686007, + "grad_norm": 1.4746271910348123, + "learning_rate": 8.610057107832509e-07, + "loss": 0.1038, + "step": 8280 + }, + { + "epoch": 1.8841865756541525, + "grad_norm": 1.8227506034405914, + "learning_rate": 8.609229721627287e-07, + "loss": 0.1171, + "step": 8281 + }, + { + "epoch": 1.8844141069397042, + "grad_norm": 1.2338686150597393, + "learning_rate": 8.608402287203728e-07, + "loss": 0.0647, + "step": 8282 + }, + { + "epoch": 1.884641638225256, + "grad_norm": 1.5525017423068463, + "learning_rate": 8.607574804578747e-07, + "loss": 0.0886, + "step": 8283 + }, + { + "epoch": 1.8848691695108077, + "grad_norm": 1.619647694383806, + "learning_rate": 8.606747273769253e-07, + "loss": 0.0578, + "step": 8284 + }, + { + "epoch": 1.8850967007963595, + "grad_norm": 2.2442457789709587, + "learning_rate": 8.605919694792161e-07, + "loss": 0.1067, + "step": 8285 + }, + { + "epoch": 1.8853242320819112, + "grad_norm": 2.0046325858807466, + "learning_rate": 8.605092067664386e-07, + "loss": 0.1072, + "step": 8286 + }, + { + "epoch": 1.885551763367463, + "grad_norm": 1.7278131172869977, + "learning_rate": 8.604264392402842e-07, + "loss": 0.1248, + "step": 8287 + }, + { + "epoch": 1.8857792946530147, + "grad_norm": 2.0397944094622704, + "learning_rate": 8.603436669024446e-07, + "loss": 0.1081, + "step": 8288 + }, + { + "epoch": 1.8860068259385665, + "grad_norm": 1.6612267301198451, + "learning_rate": 8.602608897546115e-07, + "loss": 0.0954, + "step": 8289 + }, + { + "epoch": 1.8862343572241183, + "grad_norm": 1.3468640416438875, + "learning_rate": 8.601781077984767e-07, + "loss": 0.0784, + "step": 8290 + }, + { + "epoch": 1.88646188850967, + "grad_norm": 1.505824481848141, + "learning_rate": 8.600953210357319e-07, + "loss": 0.0931, + "step": 8291 + }, + { + "epoch": 1.8866894197952218, + "grad_norm": 2.2113765085413775, + "learning_rate": 8.600125294680692e-07, + "loss": 0.1949, + "step": 8292 + }, + { + "epoch": 1.8869169510807735, + "grad_norm": 1.8238650189070063, + "learning_rate": 8.59929733097181e-07, + "loss": 0.074, + "step": 8293 + }, + { + "epoch": 1.8871444823663253, + "grad_norm": 1.9762933225786676, + "learning_rate": 8.598469319247593e-07, + "loss": 0.1055, + "step": 8294 + }, + { + "epoch": 1.887372013651877, + "grad_norm": 2.8785714836056293, + "learning_rate": 8.597641259524965e-07, + "loss": 0.1279, + "step": 8295 + }, + { + "epoch": 1.8875995449374288, + "grad_norm": 1.5172357183887974, + "learning_rate": 8.596813151820849e-07, + "loss": 0.0631, + "step": 8296 + }, + { + "epoch": 1.8878270762229805, + "grad_norm": 2.014306640694779, + "learning_rate": 8.595984996152168e-07, + "loss": 0.0833, + "step": 8297 + }, + { + "epoch": 1.8880546075085323, + "grad_norm": 1.1602096020075132, + "learning_rate": 8.595156792535852e-07, + "loss": 0.1066, + "step": 8298 + }, + { + "epoch": 1.8882821387940842, + "grad_norm": 3.568651228615464, + "learning_rate": 8.594328540988825e-07, + "loss": 0.126, + "step": 8299 + }, + { + "epoch": 1.888509670079636, + "grad_norm": 2.0351080172870013, + "learning_rate": 8.593500241528016e-07, + "loss": 0.0598, + "step": 8300 + }, + { + "epoch": 1.8887372013651877, + "grad_norm": 1.5173767046842237, + "learning_rate": 8.592671894170356e-07, + "loss": 0.1236, + "step": 8301 + }, + { + "epoch": 1.8889647326507395, + "grad_norm": 2.776342054790528, + "learning_rate": 8.59184349893277e-07, + "loss": 0.1044, + "step": 8302 + }, + { + "epoch": 1.8891922639362912, + "grad_norm": 1.6990791967005991, + "learning_rate": 8.591015055832195e-07, + "loss": 0.0793, + "step": 8303 + }, + { + "epoch": 1.889419795221843, + "grad_norm": 2.23139948985411, + "learning_rate": 8.590186564885557e-07, + "loss": 0.1156, + "step": 8304 + }, + { + "epoch": 1.8896473265073948, + "grad_norm": 1.9638463611645929, + "learning_rate": 8.589358026109792e-07, + "loss": 0.1422, + "step": 8305 + }, + { + "epoch": 1.8898748577929465, + "grad_norm": 2.2979610951844545, + "learning_rate": 8.588529439521834e-07, + "loss": 0.0898, + "step": 8306 + }, + { + "epoch": 1.8901023890784983, + "grad_norm": 1.839046186571141, + "learning_rate": 8.587700805138617e-07, + "loss": 0.0831, + "step": 8307 + }, + { + "epoch": 1.89032992036405, + "grad_norm": 2.7089181775515137, + "learning_rate": 8.586872122977076e-07, + "loss": 0.2059, + "step": 8308 + }, + { + "epoch": 1.890557451649602, + "grad_norm": 1.688883265529711, + "learning_rate": 8.58604339305415e-07, + "loss": 0.1291, + "step": 8309 + }, + { + "epoch": 1.8907849829351537, + "grad_norm": 1.460898448096465, + "learning_rate": 8.585214615386773e-07, + "loss": 0.0999, + "step": 8310 + }, + { + "epoch": 1.8910125142207055, + "grad_norm": 1.2953094489885983, + "learning_rate": 8.584385789991887e-07, + "loss": 0.0776, + "step": 8311 + }, + { + "epoch": 1.8912400455062572, + "grad_norm": 2.2075985131237137, + "learning_rate": 8.583556916886432e-07, + "loss": 0.0963, + "step": 8312 + }, + { + "epoch": 1.891467576791809, + "grad_norm": 1.7744142385673198, + "learning_rate": 8.582727996087345e-07, + "loss": 0.0495, + "step": 8313 + }, + { + "epoch": 1.8916951080773607, + "grad_norm": 1.6762850317162457, + "learning_rate": 8.581899027611574e-07, + "loss": 0.0882, + "step": 8314 + }, + { + "epoch": 1.8919226393629125, + "grad_norm": 2.3529920868513408, + "learning_rate": 8.581070011476053e-07, + "loss": 0.0679, + "step": 8315 + }, + { + "epoch": 1.8921501706484642, + "grad_norm": 2.100364972939074, + "learning_rate": 8.580240947697732e-07, + "loss": 0.0748, + "step": 8316 + }, + { + "epoch": 1.892377701934016, + "grad_norm": 1.9244535455022826, + "learning_rate": 8.579411836293555e-07, + "loss": 0.081, + "step": 8317 + }, + { + "epoch": 1.8926052332195678, + "grad_norm": 1.2639776790010304, + "learning_rate": 8.578582677280464e-07, + "loss": 0.1304, + "step": 8318 + }, + { + "epoch": 1.8928327645051195, + "grad_norm": 1.710773752904337, + "learning_rate": 8.577753470675408e-07, + "loss": 0.0803, + "step": 8319 + }, + { + "epoch": 1.8930602957906713, + "grad_norm": 2.111304301029889, + "learning_rate": 8.576924216495336e-07, + "loss": 0.1025, + "step": 8320 + }, + { + "epoch": 1.893287827076223, + "grad_norm": 1.3924061339613205, + "learning_rate": 8.576094914757194e-07, + "loss": 0.1797, + "step": 8321 + }, + { + "epoch": 1.8935153583617748, + "grad_norm": 1.9426094569182117, + "learning_rate": 8.575265565477931e-07, + "loss": 0.0829, + "step": 8322 + }, + { + "epoch": 1.8937428896473265, + "grad_norm": 2.0189174721635945, + "learning_rate": 8.574436168674498e-07, + "loss": 0.102, + "step": 8323 + }, + { + "epoch": 1.8939704209328783, + "grad_norm": 1.5399383265730866, + "learning_rate": 8.573606724363848e-07, + "loss": 0.1369, + "step": 8324 + }, + { + "epoch": 1.89419795221843, + "grad_norm": 1.7288081091530902, + "learning_rate": 8.572777232562929e-07, + "loss": 0.0725, + "step": 8325 + }, + { + "epoch": 1.8944254835039818, + "grad_norm": 1.730717895188744, + "learning_rate": 8.571947693288702e-07, + "loss": 0.1152, + "step": 8326 + }, + { + "epoch": 1.8946530147895335, + "grad_norm": 1.6527431481293537, + "learning_rate": 8.571118106558114e-07, + "loss": 0.1009, + "step": 8327 + }, + { + "epoch": 1.8948805460750853, + "grad_norm": 0.9789189145952566, + "learning_rate": 8.570288472388122e-07, + "loss": 0.0415, + "step": 8328 + }, + { + "epoch": 1.895108077360637, + "grad_norm": 2.4637644623031347, + "learning_rate": 8.569458790795685e-07, + "loss": 0.2276, + "step": 8329 + }, + { + "epoch": 1.8953356086461888, + "grad_norm": 1.626466255646907, + "learning_rate": 8.568629061797757e-07, + "loss": 0.0409, + "step": 8330 + }, + { + "epoch": 1.8955631399317405, + "grad_norm": 2.6103787346739282, + "learning_rate": 8.567799285411298e-07, + "loss": 0.1717, + "step": 8331 + }, + { + "epoch": 1.8957906712172923, + "grad_norm": 1.7252251553086313, + "learning_rate": 8.566969461653266e-07, + "loss": 0.1094, + "step": 8332 + }, + { + "epoch": 1.896018202502844, + "grad_norm": 1.20125019285821, + "learning_rate": 8.566139590540622e-07, + "loss": 0.0526, + "step": 8333 + }, + { + "epoch": 1.8962457337883958, + "grad_norm": 1.4917481767592597, + "learning_rate": 8.565309672090328e-07, + "loss": 0.1327, + "step": 8334 + }, + { + "epoch": 1.8964732650739475, + "grad_norm": 1.8357499566231195, + "learning_rate": 8.564479706319339e-07, + "loss": 0.0629, + "step": 8335 + }, + { + "epoch": 1.8967007963594993, + "grad_norm": 1.1409227083946447, + "learning_rate": 8.563649693244629e-07, + "loss": 0.0815, + "step": 8336 + }, + { + "epoch": 1.896928327645051, + "grad_norm": 2.0492970568832356, + "learning_rate": 8.562819632883155e-07, + "loss": 0.0546, + "step": 8337 + }, + { + "epoch": 1.897155858930603, + "grad_norm": 1.1938915913631698, + "learning_rate": 8.561989525251883e-07, + "loss": 0.1295, + "step": 8338 + }, + { + "epoch": 1.8973833902161548, + "grad_norm": 1.7629521515316806, + "learning_rate": 8.56115937036778e-07, + "loss": 0.1166, + "step": 8339 + }, + { + "epoch": 1.8976109215017065, + "grad_norm": 2.7816434755427077, + "learning_rate": 8.560329168247812e-07, + "loss": 0.1039, + "step": 8340 + }, + { + "epoch": 1.8978384527872583, + "grad_norm": 1.2398296197062553, + "learning_rate": 8.559498918908948e-07, + "loss": 0.0941, + "step": 8341 + }, + { + "epoch": 1.89806598407281, + "grad_norm": 2.056526288023806, + "learning_rate": 8.558668622368154e-07, + "loss": 0.0783, + "step": 8342 + }, + { + "epoch": 1.8982935153583618, + "grad_norm": 1.439365910121674, + "learning_rate": 8.557838278642401e-07, + "loss": 0.1154, + "step": 8343 + }, + { + "epoch": 1.8985210466439135, + "grad_norm": 1.7162223115697652, + "learning_rate": 8.557007887748661e-07, + "loss": 0.0905, + "step": 8344 + }, + { + "epoch": 1.8987485779294653, + "grad_norm": 1.9842439019793214, + "learning_rate": 8.556177449703906e-07, + "loss": 0.1486, + "step": 8345 + }, + { + "epoch": 1.898976109215017, + "grad_norm": 1.755920878322768, + "learning_rate": 8.555346964525107e-07, + "loss": 0.1004, + "step": 8346 + }, + { + "epoch": 1.8992036405005688, + "grad_norm": 1.7757725128098303, + "learning_rate": 8.554516432229238e-07, + "loss": 0.0709, + "step": 8347 + }, + { + "epoch": 1.8994311717861208, + "grad_norm": 1.2197949009977749, + "learning_rate": 8.553685852833274e-07, + "loss": 0.044, + "step": 8348 + }, + { + "epoch": 1.8996587030716725, + "grad_norm": 1.6457838920903318, + "learning_rate": 8.552855226354187e-07, + "loss": 0.1477, + "step": 8349 + }, + { + "epoch": 1.8998862343572243, + "grad_norm": 1.6996799716523001, + "learning_rate": 8.55202455280896e-07, + "loss": 0.0671, + "step": 8350 + }, + { + "epoch": 1.900113765642776, + "grad_norm": 2.286339837192545, + "learning_rate": 8.551193832214567e-07, + "loss": 0.0783, + "step": 8351 + }, + { + "epoch": 1.9003412969283278, + "grad_norm": 1.9501381206018091, + "learning_rate": 8.550363064587985e-07, + "loss": 0.0886, + "step": 8352 + }, + { + "epoch": 1.9005688282138795, + "grad_norm": 2.3841156996218578, + "learning_rate": 8.549532249946197e-07, + "loss": 0.1172, + "step": 8353 + }, + { + "epoch": 1.9007963594994313, + "grad_norm": 1.5973692879748291, + "learning_rate": 8.548701388306179e-07, + "loss": 0.0549, + "step": 8354 + }, + { + "epoch": 1.901023890784983, + "grad_norm": 1.727772482080952, + "learning_rate": 8.547870479684916e-07, + "loss": 0.2124, + "step": 8355 + }, + { + "epoch": 1.9012514220705348, + "grad_norm": 3.0525620283617956, + "learning_rate": 8.547039524099387e-07, + "loss": 0.0762, + "step": 8356 + }, + { + "epoch": 1.9014789533560865, + "grad_norm": 1.7064844536754875, + "learning_rate": 8.546208521566578e-07, + "loss": 0.0973, + "step": 8357 + }, + { + "epoch": 1.9017064846416383, + "grad_norm": 1.78821288482924, + "learning_rate": 8.545377472103474e-07, + "loss": 0.073, + "step": 8358 + }, + { + "epoch": 1.90193401592719, + "grad_norm": 2.822565788897465, + "learning_rate": 8.544546375727055e-07, + "loss": 0.1166, + "step": 8359 + }, + { + "epoch": 1.9021615472127418, + "grad_norm": 2.033612171758109, + "learning_rate": 8.543715232454311e-07, + "loss": 0.1454, + "step": 8360 + }, + { + "epoch": 1.9023890784982935, + "grad_norm": 1.6519078958373479, + "learning_rate": 8.54288404230223e-07, + "loss": 0.1722, + "step": 8361 + }, + { + "epoch": 1.9026166097838453, + "grad_norm": 1.6349924643849787, + "learning_rate": 8.542052805287797e-07, + "loss": 0.1117, + "step": 8362 + }, + { + "epoch": 1.902844141069397, + "grad_norm": 1.7230199638939674, + "learning_rate": 8.541221521428003e-07, + "loss": 0.0933, + "step": 8363 + }, + { + "epoch": 1.9030716723549488, + "grad_norm": 2.128008748714492, + "learning_rate": 8.540390190739839e-07, + "loss": 0.1449, + "step": 8364 + }, + { + "epoch": 1.9032992036405005, + "grad_norm": 2.520208975230359, + "learning_rate": 8.539558813240291e-07, + "loss": 0.0579, + "step": 8365 + }, + { + "epoch": 1.9035267349260523, + "grad_norm": 1.8327988679527187, + "learning_rate": 8.538727388946356e-07, + "loss": 0.0816, + "step": 8366 + }, + { + "epoch": 1.903754266211604, + "grad_norm": 1.0131312440967444, + "learning_rate": 8.537895917875023e-07, + "loss": 0.081, + "step": 8367 + }, + { + "epoch": 1.9039817974971558, + "grad_norm": 2.525664618164771, + "learning_rate": 8.537064400043289e-07, + "loss": 0.1376, + "step": 8368 + }, + { + "epoch": 1.9042093287827075, + "grad_norm": 2.0270036303350527, + "learning_rate": 8.536232835468145e-07, + "loss": 0.1755, + "step": 8369 + }, + { + "epoch": 1.9044368600682593, + "grad_norm": 1.5781167079311442, + "learning_rate": 8.535401224166593e-07, + "loss": 0.0664, + "step": 8370 + }, + { + "epoch": 1.904664391353811, + "grad_norm": 1.7575018943937446, + "learning_rate": 8.534569566155623e-07, + "loss": 0.112, + "step": 8371 + }, + { + "epoch": 1.9048919226393628, + "grad_norm": 1.0596850759981111, + "learning_rate": 8.533737861452235e-07, + "loss": 0.1052, + "step": 8372 + }, + { + "epoch": 1.9051194539249146, + "grad_norm": 1.0803329294529274, + "learning_rate": 8.532906110073427e-07, + "loss": 0.054, + "step": 8373 + }, + { + "epoch": 1.9053469852104663, + "grad_norm": 1.4269226543498574, + "learning_rate": 8.5320743120362e-07, + "loss": 0.0906, + "step": 8374 + }, + { + "epoch": 1.905574516496018, + "grad_norm": 1.8968299723220357, + "learning_rate": 8.531242467357555e-07, + "loss": 0.1619, + "step": 8375 + }, + { + "epoch": 1.9058020477815698, + "grad_norm": 1.7596520806000042, + "learning_rate": 8.530410576054489e-07, + "loss": 0.0861, + "step": 8376 + }, + { + "epoch": 1.9060295790671218, + "grad_norm": 1.7964392119029702, + "learning_rate": 8.52957863814401e-07, + "loss": 0.0651, + "step": 8377 + }, + { + "epoch": 1.9062571103526735, + "grad_norm": 2.288788767440748, + "learning_rate": 8.528746653643116e-07, + "loss": 0.0977, + "step": 8378 + }, + { + "epoch": 1.9064846416382253, + "grad_norm": 1.9994178831489846, + "learning_rate": 8.527914622568814e-07, + "loss": 0.1369, + "step": 8379 + }, + { + "epoch": 1.906712172923777, + "grad_norm": 1.260689905620287, + "learning_rate": 8.527082544938111e-07, + "loss": 0.0359, + "step": 8380 + }, + { + "epoch": 1.9069397042093288, + "grad_norm": 1.7946739368224318, + "learning_rate": 8.526250420768009e-07, + "loss": 0.0514, + "step": 8381 + }, + { + "epoch": 1.9071672354948805, + "grad_norm": 1.423877072534363, + "learning_rate": 8.525418250075518e-07, + "loss": 0.1304, + "step": 8382 + }, + { + "epoch": 1.9073947667804323, + "grad_norm": 1.4892533610710477, + "learning_rate": 8.524586032877645e-07, + "loss": 0.1109, + "step": 8383 + }, + { + "epoch": 1.907622298065984, + "grad_norm": 1.8983697284600995, + "learning_rate": 8.523753769191399e-07, + "loss": 0.0886, + "step": 8384 + }, + { + "epoch": 1.9078498293515358, + "grad_norm": 1.7364272953887896, + "learning_rate": 8.522921459033791e-07, + "loss": 0.0632, + "step": 8385 + }, + { + "epoch": 1.9080773606370875, + "grad_norm": 1.984748396350035, + "learning_rate": 8.522089102421829e-07, + "loss": 0.117, + "step": 8386 + }, + { + "epoch": 1.9083048919226395, + "grad_norm": 1.5206341771574357, + "learning_rate": 8.52125669937253e-07, + "loss": 0.0739, + "step": 8387 + }, + { + "epoch": 1.9085324232081913, + "grad_norm": 1.4813839884119313, + "learning_rate": 8.5204242499029e-07, + "loss": 0.1032, + "step": 8388 + }, + { + "epoch": 1.908759954493743, + "grad_norm": 2.8830784784360004, + "learning_rate": 8.51959175402996e-07, + "loss": 0.195, + "step": 8389 + }, + { + "epoch": 1.9089874857792948, + "grad_norm": 1.3322631129748252, + "learning_rate": 8.518759211770719e-07, + "loss": 0.1141, + "step": 8390 + }, + { + "epoch": 1.9092150170648465, + "grad_norm": 2.383725391131221, + "learning_rate": 8.517926623142196e-07, + "loss": 0.078, + "step": 8391 + }, + { + "epoch": 1.9094425483503983, + "grad_norm": 3.5156731006712048, + "learning_rate": 8.517093988161404e-07, + "loss": 0.1994, + "step": 8392 + }, + { + "epoch": 1.90967007963595, + "grad_norm": 1.9960590154522282, + "learning_rate": 8.516261306845365e-07, + "loss": 0.0986, + "step": 8393 + }, + { + "epoch": 1.9098976109215018, + "grad_norm": 1.4147014281353942, + "learning_rate": 8.515428579211095e-07, + "loss": 0.0571, + "step": 8394 + }, + { + "epoch": 1.9101251422070535, + "grad_norm": 1.3103127236232746, + "learning_rate": 8.514595805275614e-07, + "loss": 0.0536, + "step": 8395 + }, + { + "epoch": 1.9103526734926053, + "grad_norm": 1.1912098943499887, + "learning_rate": 8.513762985055942e-07, + "loss": 0.1332, + "step": 8396 + }, + { + "epoch": 1.910580204778157, + "grad_norm": 2.0496364865781658, + "learning_rate": 8.5129301185691e-07, + "loss": 0.0697, + "step": 8397 + }, + { + "epoch": 1.9108077360637088, + "grad_norm": 0.9461185817769251, + "learning_rate": 8.512097205832111e-07, + "loss": 0.0478, + "step": 8398 + }, + { + "epoch": 1.9110352673492605, + "grad_norm": 1.7546955788450125, + "learning_rate": 8.511264246861997e-07, + "loss": 0.0874, + "step": 8399 + }, + { + "epoch": 1.9112627986348123, + "grad_norm": 2.3319144466335544, + "learning_rate": 8.510431241675784e-07, + "loss": 0.1562, + "step": 8400 + }, + { + "epoch": 1.911490329920364, + "grad_norm": 2.4271876266800483, + "learning_rate": 8.509598190290497e-07, + "loss": 0.101, + "step": 8401 + }, + { + "epoch": 1.9117178612059158, + "grad_norm": 1.8427885222115377, + "learning_rate": 8.508765092723159e-07, + "loss": 0.0702, + "step": 8402 + }, + { + "epoch": 1.9119453924914676, + "grad_norm": 1.5707748325621465, + "learning_rate": 8.507931948990801e-07, + "loss": 0.1112, + "step": 8403 + }, + { + "epoch": 1.9121729237770193, + "grad_norm": 1.0774601536928494, + "learning_rate": 8.507098759110449e-07, + "loss": 0.0883, + "step": 8404 + }, + { + "epoch": 1.912400455062571, + "grad_norm": 2.134705751837509, + "learning_rate": 8.506265523099133e-07, + "loss": 0.1139, + "step": 8405 + }, + { + "epoch": 1.9126279863481228, + "grad_norm": 1.9067676708528498, + "learning_rate": 8.50543224097388e-07, + "loss": 0.0732, + "step": 8406 + }, + { + "epoch": 1.9128555176336746, + "grad_norm": 0.9727668156299778, + "learning_rate": 8.504598912751722e-07, + "loss": 0.0331, + "step": 8407 + }, + { + "epoch": 1.9130830489192263, + "grad_norm": 2.350308805559537, + "learning_rate": 8.503765538449695e-07, + "loss": 0.1554, + "step": 8408 + }, + { + "epoch": 1.913310580204778, + "grad_norm": 1.4015398689059053, + "learning_rate": 8.502932118084825e-07, + "loss": 0.0634, + "step": 8409 + }, + { + "epoch": 1.9135381114903298, + "grad_norm": 1.8963424951873085, + "learning_rate": 8.502098651674148e-07, + "loss": 0.1863, + "step": 8410 + }, + { + "epoch": 1.9137656427758816, + "grad_norm": 1.3398651655669072, + "learning_rate": 8.501265139234702e-07, + "loss": 0.0709, + "step": 8411 + }, + { + "epoch": 1.9139931740614333, + "grad_norm": 2.2695420225526757, + "learning_rate": 8.500431580783518e-07, + "loss": 0.0719, + "step": 8412 + }, + { + "epoch": 1.914220705346985, + "grad_norm": 0.9721125468019496, + "learning_rate": 8.499597976337632e-07, + "loss": 0.0514, + "step": 8413 + }, + { + "epoch": 1.9144482366325368, + "grad_norm": 2.1322236230579223, + "learning_rate": 8.498764325914087e-07, + "loss": 0.1084, + "step": 8414 + }, + { + "epoch": 1.9146757679180886, + "grad_norm": 2.35934344607237, + "learning_rate": 8.497930629529916e-07, + "loss": 0.0763, + "step": 8415 + }, + { + "epoch": 1.9149032992036406, + "grad_norm": 1.4798203631279319, + "learning_rate": 8.497096887202158e-07, + "loss": 0.0588, + "step": 8416 + }, + { + "epoch": 1.9151308304891923, + "grad_norm": 1.6849690708017626, + "learning_rate": 8.496263098947857e-07, + "loss": 0.1134, + "step": 8417 + }, + { + "epoch": 1.915358361774744, + "grad_norm": 1.9929702536635638, + "learning_rate": 8.495429264784051e-07, + "loss": 0.126, + "step": 8418 + }, + { + "epoch": 1.9155858930602958, + "grad_norm": 2.242324964257082, + "learning_rate": 8.494595384727783e-07, + "loss": 0.1326, + "step": 8419 + }, + { + "epoch": 1.9158134243458476, + "grad_norm": 1.199433028857461, + "learning_rate": 8.493761458796098e-07, + "loss": 0.1425, + "step": 8420 + }, + { + "epoch": 1.9160409556313993, + "grad_norm": 1.5634859845022955, + "learning_rate": 8.492927487006037e-07, + "loss": 0.064, + "step": 8421 + }, + { + "epoch": 1.916268486916951, + "grad_norm": 2.247087348394463, + "learning_rate": 8.492093469374646e-07, + "loss": 0.0961, + "step": 8422 + }, + { + "epoch": 1.9164960182025028, + "grad_norm": 1.6277470161077132, + "learning_rate": 8.491259405918969e-07, + "loss": 0.0823, + "step": 8423 + }, + { + "epoch": 1.9167235494880546, + "grad_norm": 1.258836642443553, + "learning_rate": 8.490425296656057e-07, + "loss": 0.073, + "step": 8424 + }, + { + "epoch": 1.9169510807736063, + "grad_norm": 0.9943141297680885, + "learning_rate": 8.489591141602954e-07, + "loss": 0.0679, + "step": 8425 + }, + { + "epoch": 1.9171786120591583, + "grad_norm": 1.5514946605389608, + "learning_rate": 8.488756940776709e-07, + "loss": 0.0557, + "step": 8426 + }, + { + "epoch": 1.91740614334471, + "grad_norm": 3.338604708256682, + "learning_rate": 8.487922694194374e-07, + "loss": 0.2102, + "step": 8427 + }, + { + "epoch": 1.9176336746302618, + "grad_norm": 1.792222041676389, + "learning_rate": 8.487088401872997e-07, + "loss": 0.1272, + "step": 8428 + }, + { + "epoch": 1.9178612059158135, + "grad_norm": 1.8876701986161428, + "learning_rate": 8.486254063829628e-07, + "loss": 0.0605, + "step": 8429 + }, + { + "epoch": 1.9180887372013653, + "grad_norm": 2.2659142728739736, + "learning_rate": 8.485419680081324e-07, + "loss": 0.1049, + "step": 8430 + }, + { + "epoch": 1.918316268486917, + "grad_norm": 1.9475951010122616, + "learning_rate": 8.484585250645134e-07, + "loss": 0.1072, + "step": 8431 + }, + { + "epoch": 1.9185437997724688, + "grad_norm": 1.8729195424533605, + "learning_rate": 8.483750775538116e-07, + "loss": 0.0868, + "step": 8432 + }, + { + "epoch": 1.9187713310580206, + "grad_norm": 1.2718749744993825, + "learning_rate": 8.482916254777321e-07, + "loss": 0.0838, + "step": 8433 + }, + { + "epoch": 1.9189988623435723, + "grad_norm": 3.179250927334045, + "learning_rate": 8.482081688379809e-07, + "loss": 0.1408, + "step": 8434 + }, + { + "epoch": 1.919226393629124, + "grad_norm": 0.9912202600373338, + "learning_rate": 8.481247076362633e-07, + "loss": 0.1012, + "step": 8435 + }, + { + "epoch": 1.9194539249146758, + "grad_norm": 2.1647864863915554, + "learning_rate": 8.480412418742855e-07, + "loss": 0.1023, + "step": 8436 + }, + { + "epoch": 1.9196814562002276, + "grad_norm": 1.2022709576605757, + "learning_rate": 8.479577715537531e-07, + "loss": 0.0424, + "step": 8437 + }, + { + "epoch": 1.9199089874857793, + "grad_norm": 2.668259844445578, + "learning_rate": 8.478742966763721e-07, + "loss": 0.0619, + "step": 8438 + }, + { + "epoch": 1.920136518771331, + "grad_norm": 1.3071073149878354, + "learning_rate": 8.477908172438488e-07, + "loss": 0.1575, + "step": 8439 + }, + { + "epoch": 1.9203640500568828, + "grad_norm": 1.8530288632841168, + "learning_rate": 8.477073332578892e-07, + "loss": 0.0684, + "step": 8440 + }, + { + "epoch": 1.9205915813424346, + "grad_norm": 1.232503648261482, + "learning_rate": 8.476238447201995e-07, + "loss": 0.0728, + "step": 8441 + }, + { + "epoch": 1.9208191126279863, + "grad_norm": 2.497572414815773, + "learning_rate": 8.475403516324863e-07, + "loss": 0.0822, + "step": 8442 + }, + { + "epoch": 1.921046643913538, + "grad_norm": 1.9091586596017254, + "learning_rate": 8.474568539964556e-07, + "loss": 0.1432, + "step": 8443 + }, + { + "epoch": 1.9212741751990898, + "grad_norm": 1.7922258981875185, + "learning_rate": 8.473733518138145e-07, + "loss": 0.0858, + "step": 8444 + }, + { + "epoch": 1.9215017064846416, + "grad_norm": 1.6463268320552324, + "learning_rate": 8.472898450862691e-07, + "loss": 0.0483, + "step": 8445 + }, + { + "epoch": 1.9217292377701933, + "grad_norm": 2.5348026255674987, + "learning_rate": 8.472063338155265e-07, + "loss": 0.1395, + "step": 8446 + }, + { + "epoch": 1.921956769055745, + "grad_norm": 1.736325113713475, + "learning_rate": 8.471228180032934e-07, + "loss": 0.0768, + "step": 8447 + }, + { + "epoch": 1.9221843003412968, + "grad_norm": 2.1050598157100544, + "learning_rate": 8.470392976512768e-07, + "loss": 0.1608, + "step": 8448 + }, + { + "epoch": 1.9224118316268486, + "grad_norm": 1.6835896578771998, + "learning_rate": 8.469557727611833e-07, + "loss": 0.1586, + "step": 8449 + }, + { + "epoch": 1.9226393629124003, + "grad_norm": 1.9175114617751532, + "learning_rate": 8.468722433347204e-07, + "loss": 0.1146, + "step": 8450 + }, + { + "epoch": 1.922866894197952, + "grad_norm": 3.4114979818235165, + "learning_rate": 8.467887093735953e-07, + "loss": 0.0854, + "step": 8451 + }, + { + "epoch": 1.9230944254835038, + "grad_norm": 1.9034177264033771, + "learning_rate": 8.467051708795152e-07, + "loss": 0.0824, + "step": 8452 + }, + { + "epoch": 1.9233219567690556, + "grad_norm": 1.8754113697450077, + "learning_rate": 8.466216278541874e-07, + "loss": 0.0891, + "step": 8453 + }, + { + "epoch": 1.9235494880546073, + "grad_norm": 2.15705770201668, + "learning_rate": 8.465380802993193e-07, + "loss": 0.0708, + "step": 8454 + }, + { + "epoch": 1.9237770193401593, + "grad_norm": 1.809327681436963, + "learning_rate": 8.464545282166187e-07, + "loss": 0.0702, + "step": 8455 + }, + { + "epoch": 1.924004550625711, + "grad_norm": 1.0962609004853898, + "learning_rate": 8.463709716077929e-07, + "loss": 0.1039, + "step": 8456 + }, + { + "epoch": 1.9242320819112628, + "grad_norm": 1.5206808018082445, + "learning_rate": 8.4628741047455e-07, + "loss": 0.0496, + "step": 8457 + }, + { + "epoch": 1.9244596131968146, + "grad_norm": 3.0651136015171705, + "learning_rate": 8.462038448185977e-07, + "loss": 0.1708, + "step": 8458 + }, + { + "epoch": 1.9246871444823663, + "grad_norm": 1.0623543651630516, + "learning_rate": 8.461202746416442e-07, + "loss": 0.0663, + "step": 8459 + }, + { + "epoch": 1.924914675767918, + "grad_norm": 1.665256204666068, + "learning_rate": 8.460366999453968e-07, + "loss": 0.0957, + "step": 8460 + }, + { + "epoch": 1.9251422070534698, + "grad_norm": 1.7811657834442234, + "learning_rate": 8.459531207315644e-07, + "loss": 0.0912, + "step": 8461 + }, + { + "epoch": 1.9253697383390216, + "grad_norm": 2.319950791748146, + "learning_rate": 8.458695370018546e-07, + "loss": 0.0839, + "step": 8462 + }, + { + "epoch": 1.9255972696245733, + "grad_norm": 2.2330640736837437, + "learning_rate": 8.457859487579762e-07, + "loss": 0.1652, + "step": 8463 + }, + { + "epoch": 1.925824800910125, + "grad_norm": 2.522709981070815, + "learning_rate": 8.457023560016371e-07, + "loss": 0.1151, + "step": 8464 + }, + { + "epoch": 1.926052332195677, + "grad_norm": 1.8640678015129717, + "learning_rate": 8.456187587345463e-07, + "loss": 0.0459, + "step": 8465 + }, + { + "epoch": 1.9262798634812288, + "grad_norm": 2.50091870403617, + "learning_rate": 8.455351569584119e-07, + "loss": 0.0564, + "step": 8466 + }, + { + "epoch": 1.9265073947667806, + "grad_norm": 2.5819766855842956, + "learning_rate": 8.454515506749431e-07, + "loss": 0.083, + "step": 8467 + }, + { + "epoch": 1.9267349260523323, + "grad_norm": 1.6796689480749127, + "learning_rate": 8.453679398858481e-07, + "loss": 0.0531, + "step": 8468 + }, + { + "epoch": 1.926962457337884, + "grad_norm": 1.677839119494002, + "learning_rate": 8.452843245928359e-07, + "loss": 0.0613, + "step": 8469 + }, + { + "epoch": 1.9271899886234358, + "grad_norm": 1.8973529305250347, + "learning_rate": 8.452007047976155e-07, + "loss": 0.078, + "step": 8470 + }, + { + "epoch": 1.9274175199089876, + "grad_norm": 2.3986730572979265, + "learning_rate": 8.451170805018964e-07, + "loss": 0.0893, + "step": 8471 + }, + { + "epoch": 1.9276450511945393, + "grad_norm": 1.3191923575313789, + "learning_rate": 8.45033451707387e-07, + "loss": 0.0711, + "step": 8472 + }, + { + "epoch": 1.927872582480091, + "grad_norm": 2.919992540617028, + "learning_rate": 8.449498184157968e-07, + "loss": 0.087, + "step": 8473 + }, + { + "epoch": 1.9281001137656428, + "grad_norm": 2.1736715695739366, + "learning_rate": 8.448661806288352e-07, + "loss": 0.1122, + "step": 8474 + }, + { + "epoch": 1.9283276450511946, + "grad_norm": 1.7246493796250064, + "learning_rate": 8.447825383482116e-07, + "loss": 0.1058, + "step": 8475 + }, + { + "epoch": 1.9285551763367463, + "grad_norm": 1.9251531278033478, + "learning_rate": 8.446988915756353e-07, + "loss": 0.0836, + "step": 8476 + }, + { + "epoch": 1.928782707622298, + "grad_norm": 1.4968384792196872, + "learning_rate": 8.446152403128161e-07, + "loss": 0.0998, + "step": 8477 + }, + { + "epoch": 1.9290102389078498, + "grad_norm": 1.3499858437839942, + "learning_rate": 8.445315845614636e-07, + "loss": 0.1275, + "step": 8478 + }, + { + "epoch": 1.9292377701934016, + "grad_norm": 1.971157495492609, + "learning_rate": 8.444479243232875e-07, + "loss": 0.0787, + "step": 8479 + }, + { + "epoch": 1.9294653014789533, + "grad_norm": 1.2783668702913251, + "learning_rate": 8.443642595999977e-07, + "loss": 0.0503, + "step": 8480 + }, + { + "epoch": 1.929692832764505, + "grad_norm": 2.068704739883542, + "learning_rate": 8.442805903933041e-07, + "loss": 0.0918, + "step": 8481 + }, + { + "epoch": 1.9299203640500568, + "grad_norm": 1.4539311831790112, + "learning_rate": 8.441969167049171e-07, + "loss": 0.0937, + "step": 8482 + }, + { + "epoch": 1.9301478953356086, + "grad_norm": 2.423333261668445, + "learning_rate": 8.441132385365462e-07, + "loss": 0.0632, + "step": 8483 + }, + { + "epoch": 1.9303754266211604, + "grad_norm": 1.816600565728022, + "learning_rate": 8.440295558899024e-07, + "loss": 0.0642, + "step": 8484 + }, + { + "epoch": 1.930602957906712, + "grad_norm": 1.9843915366698888, + "learning_rate": 8.439458687666954e-07, + "loss": 0.0949, + "step": 8485 + }, + { + "epoch": 1.9308304891922639, + "grad_norm": 1.8704383168228935, + "learning_rate": 8.438621771686358e-07, + "loss": 0.1104, + "step": 8486 + }, + { + "epoch": 1.9310580204778156, + "grad_norm": 2.089140082871353, + "learning_rate": 8.437784810974343e-07, + "loss": 0.1113, + "step": 8487 + }, + { + "epoch": 1.9312855517633674, + "grad_norm": 1.5628222585539076, + "learning_rate": 8.43694780554801e-07, + "loss": 0.0848, + "step": 8488 + }, + { + "epoch": 1.931513083048919, + "grad_norm": 1.731060896808239, + "learning_rate": 8.436110755424472e-07, + "loss": 0.0731, + "step": 8489 + }, + { + "epoch": 1.9317406143344709, + "grad_norm": 1.846587230420918, + "learning_rate": 8.435273660620833e-07, + "loss": 0.059, + "step": 8490 + }, + { + "epoch": 1.9319681456200226, + "grad_norm": 1.896878928250611, + "learning_rate": 8.434436521154202e-07, + "loss": 0.0574, + "step": 8491 + }, + { + "epoch": 1.9321956769055744, + "grad_norm": 1.3544988049376627, + "learning_rate": 8.43359933704169e-07, + "loss": 0.0535, + "step": 8492 + }, + { + "epoch": 1.9324232081911261, + "grad_norm": 1.740141894716064, + "learning_rate": 8.432762108300404e-07, + "loss": 0.0891, + "step": 8493 + }, + { + "epoch": 1.932650739476678, + "grad_norm": 0.9830501187953077, + "learning_rate": 8.431924834947461e-07, + "loss": 0.1051, + "step": 8494 + }, + { + "epoch": 1.9328782707622298, + "grad_norm": 2.4380986003963083, + "learning_rate": 8.431087516999969e-07, + "loss": 0.1076, + "step": 8495 + }, + { + "epoch": 1.9331058020477816, + "grad_norm": 1.4384427780874542, + "learning_rate": 8.430250154475044e-07, + "loss": 0.1277, + "step": 8496 + }, + { + "epoch": 1.9333333333333333, + "grad_norm": 2.3763608180998066, + "learning_rate": 8.429412747389798e-07, + "loss": 0.1468, + "step": 8497 + }, + { + "epoch": 1.933560864618885, + "grad_norm": 2.7126993024948796, + "learning_rate": 8.428575295761346e-07, + "loss": 0.0923, + "step": 8498 + }, + { + "epoch": 1.9337883959044369, + "grad_norm": 2.1813475835249982, + "learning_rate": 8.427737799606806e-07, + "loss": 0.0782, + "step": 8499 + }, + { + "epoch": 1.9340159271899886, + "grad_norm": 2.6373903019450813, + "learning_rate": 8.426900258943292e-07, + "loss": 0.0889, + "step": 8500 + }, + { + "epoch": 1.9342434584755404, + "grad_norm": 1.4708943806049117, + "learning_rate": 8.426062673787926e-07, + "loss": 0.0483, + "step": 8501 + }, + { + "epoch": 1.934470989761092, + "grad_norm": 1.1381297165096618, + "learning_rate": 8.425225044157823e-07, + "loss": 0.0428, + "step": 8502 + }, + { + "epoch": 1.9346985210466439, + "grad_norm": 1.7545261529169938, + "learning_rate": 8.424387370070102e-07, + "loss": 0.1039, + "step": 8503 + }, + { + "epoch": 1.9349260523321958, + "grad_norm": 1.4167186825796818, + "learning_rate": 8.423549651541889e-07, + "loss": 0.1195, + "step": 8504 + }, + { + "epoch": 1.9351535836177476, + "grad_norm": 1.5331925096031622, + "learning_rate": 8.4227118885903e-07, + "loss": 0.0939, + "step": 8505 + }, + { + "epoch": 1.9353811149032993, + "grad_norm": 1.3675225762983716, + "learning_rate": 8.421874081232459e-07, + "loss": 0.1108, + "step": 8506 + }, + { + "epoch": 1.935608646188851, + "grad_norm": 2.024739386960217, + "learning_rate": 8.421036229485489e-07, + "loss": 0.0882, + "step": 8507 + }, + { + "epoch": 1.9358361774744028, + "grad_norm": 1.9821151920385132, + "learning_rate": 8.420198333366513e-07, + "loss": 0.0854, + "step": 8508 + }, + { + "epoch": 1.9360637087599546, + "grad_norm": 1.0951514220743803, + "learning_rate": 8.419360392892663e-07, + "loss": 0.0594, + "step": 8509 + }, + { + "epoch": 1.9362912400455063, + "grad_norm": 2.2169726519990904, + "learning_rate": 8.418522408081054e-07, + "loss": 0.1, + "step": 8510 + }, + { + "epoch": 1.936518771331058, + "grad_norm": 2.1220228348519505, + "learning_rate": 8.417684378948822e-07, + "loss": 0.0438, + "step": 8511 + }, + { + "epoch": 1.9367463026166098, + "grad_norm": 0.8841115425271152, + "learning_rate": 8.416846305513089e-07, + "loss": 0.0205, + "step": 8512 + }, + { + "epoch": 1.9369738339021616, + "grad_norm": 1.0874199804171312, + "learning_rate": 8.416008187790986e-07, + "loss": 0.0487, + "step": 8513 + }, + { + "epoch": 1.9372013651877134, + "grad_norm": 1.6181231078970064, + "learning_rate": 8.415170025799644e-07, + "loss": 0.1301, + "step": 8514 + }, + { + "epoch": 1.937428896473265, + "grad_norm": 1.1611858386666436, + "learning_rate": 8.414331819556193e-07, + "loss": 0.0285, + "step": 8515 + }, + { + "epoch": 1.9376564277588169, + "grad_norm": 1.2191047345391055, + "learning_rate": 8.41349356907776e-07, + "loss": 0.0758, + "step": 8516 + }, + { + "epoch": 1.9378839590443686, + "grad_norm": 1.8008450035039854, + "learning_rate": 8.412655274381481e-07, + "loss": 0.0816, + "step": 8517 + }, + { + "epoch": 1.9381114903299204, + "grad_norm": 2.549335874160235, + "learning_rate": 8.411816935484491e-07, + "loss": 0.1047, + "step": 8518 + }, + { + "epoch": 1.9383390216154721, + "grad_norm": 1.863164888167891, + "learning_rate": 8.410978552403923e-07, + "loss": 0.1242, + "step": 8519 + }, + { + "epoch": 1.9385665529010239, + "grad_norm": 1.9511545134715371, + "learning_rate": 8.410140125156907e-07, + "loss": 0.0776, + "step": 8520 + }, + { + "epoch": 1.9387940841865756, + "grad_norm": 1.4616590047656322, + "learning_rate": 8.409301653760587e-07, + "loss": 0.0534, + "step": 8521 + }, + { + "epoch": 1.9390216154721274, + "grad_norm": 1.5355929638697312, + "learning_rate": 8.408463138232094e-07, + "loss": 0.0876, + "step": 8522 + }, + { + "epoch": 1.9392491467576791, + "grad_norm": 1.4389045256078594, + "learning_rate": 8.407624578588566e-07, + "loss": 0.0846, + "step": 8523 + }, + { + "epoch": 1.9394766780432309, + "grad_norm": 2.931059465647058, + "learning_rate": 8.406785974847145e-07, + "loss": 0.1378, + "step": 8524 + }, + { + "epoch": 1.9397042093287826, + "grad_norm": 1.4601601928244312, + "learning_rate": 8.405947327024968e-07, + "loss": 0.0566, + "step": 8525 + }, + { + "epoch": 1.9399317406143344, + "grad_norm": 2.200201035887927, + "learning_rate": 8.405108635139178e-07, + "loss": 0.0842, + "step": 8526 + }, + { + "epoch": 1.9401592718998861, + "grad_norm": 2.0423363286701677, + "learning_rate": 8.404269899206911e-07, + "loss": 0.0823, + "step": 8527 + }, + { + "epoch": 1.9403868031854379, + "grad_norm": 1.8410916810358255, + "learning_rate": 8.403431119245316e-07, + "loss": 0.1669, + "step": 8528 + }, + { + "epoch": 1.9406143344709896, + "grad_norm": 1.8658327967514299, + "learning_rate": 8.40259229527153e-07, + "loss": 0.112, + "step": 8529 + }, + { + "epoch": 1.9408418657565414, + "grad_norm": 1.646429046748771, + "learning_rate": 8.4017534273027e-07, + "loss": 0.0939, + "step": 8530 + }, + { + "epoch": 1.9410693970420931, + "grad_norm": 1.9699312522483354, + "learning_rate": 8.400914515355972e-07, + "loss": 0.0713, + "step": 8531 + }, + { + "epoch": 1.9412969283276449, + "grad_norm": 3.035616092105102, + "learning_rate": 8.40007555944849e-07, + "loss": 0.1617, + "step": 8532 + }, + { + "epoch": 1.9415244596131969, + "grad_norm": 1.1459056528449258, + "learning_rate": 8.399236559597403e-07, + "loss": 0.0407, + "step": 8533 + }, + { + "epoch": 1.9417519908987486, + "grad_norm": 1.6129092079250968, + "learning_rate": 8.398397515819855e-07, + "loss": 0.0355, + "step": 8534 + }, + { + "epoch": 1.9419795221843004, + "grad_norm": 1.6025860195926533, + "learning_rate": 8.397558428132995e-07, + "loss": 0.0769, + "step": 8535 + }, + { + "epoch": 1.9422070534698521, + "grad_norm": 1.7837472740584903, + "learning_rate": 8.396719296553976e-07, + "loss": 0.0558, + "step": 8536 + }, + { + "epoch": 1.9424345847554039, + "grad_norm": 2.8544033907639945, + "learning_rate": 8.395880121099944e-07, + "loss": 0.2276, + "step": 8537 + }, + { + "epoch": 1.9426621160409556, + "grad_norm": 2.226030766319897, + "learning_rate": 8.395040901788056e-07, + "loss": 0.163, + "step": 8538 + }, + { + "epoch": 1.9428896473265074, + "grad_norm": 2.4601269585424066, + "learning_rate": 8.394201638635458e-07, + "loss": 0.1241, + "step": 8539 + }, + { + "epoch": 1.9431171786120591, + "grad_norm": 1.6894697055252035, + "learning_rate": 8.393362331659305e-07, + "loss": 0.1729, + "step": 8540 + }, + { + "epoch": 1.9433447098976109, + "grad_norm": 1.089462098183616, + "learning_rate": 8.392522980876752e-07, + "loss": 0.0458, + "step": 8541 + }, + { + "epoch": 1.9435722411831626, + "grad_norm": 1.4212217862469188, + "learning_rate": 8.391683586304954e-07, + "loss": 0.0483, + "step": 8542 + }, + { + "epoch": 1.9437997724687146, + "grad_norm": 2.0508094063197895, + "learning_rate": 8.390844147961064e-07, + "loss": 0.0837, + "step": 8543 + }, + { + "epoch": 1.9440273037542664, + "grad_norm": 2.6999732469953663, + "learning_rate": 8.39000466586224e-07, + "loss": 0.0963, + "step": 8544 + }, + { + "epoch": 1.944254835039818, + "grad_norm": 1.4407700122396472, + "learning_rate": 8.389165140025642e-07, + "loss": 0.167, + "step": 8545 + }, + { + "epoch": 1.9444823663253699, + "grad_norm": 2.4557818509444247, + "learning_rate": 8.388325570468425e-07, + "loss": 0.0731, + "step": 8546 + }, + { + "epoch": 1.9447098976109216, + "grad_norm": 1.354782697339982, + "learning_rate": 8.38748595720775e-07, + "loss": 0.1182, + "step": 8547 + }, + { + "epoch": 1.9449374288964734, + "grad_norm": 0.7537613438324625, + "learning_rate": 8.386646300260777e-07, + "loss": 0.0589, + "step": 8548 + }, + { + "epoch": 1.9451649601820251, + "grad_norm": 1.4623107199177439, + "learning_rate": 8.385806599644666e-07, + "loss": 0.0587, + "step": 8549 + }, + { + "epoch": 1.9453924914675769, + "grad_norm": 1.892414732824123, + "learning_rate": 8.384966855376579e-07, + "loss": 0.0533, + "step": 8550 + }, + { + "epoch": 1.9456200227531286, + "grad_norm": 2.19221426746065, + "learning_rate": 8.384127067473681e-07, + "loss": 0.0642, + "step": 8551 + }, + { + "epoch": 1.9458475540386804, + "grad_norm": 1.5719041762146422, + "learning_rate": 8.383287235953133e-07, + "loss": 0.0261, + "step": 8552 + }, + { + "epoch": 1.9460750853242321, + "grad_norm": 2.376314018674937, + "learning_rate": 8.382447360832102e-07, + "loss": 0.2128, + "step": 8553 + }, + { + "epoch": 1.9463026166097839, + "grad_norm": 2.631107798034157, + "learning_rate": 8.381607442127753e-07, + "loss": 0.0754, + "step": 8554 + }, + { + "epoch": 1.9465301478953356, + "grad_norm": 1.8455788499407606, + "learning_rate": 8.38076747985725e-07, + "loss": 0.1417, + "step": 8555 + }, + { + "epoch": 1.9467576791808874, + "grad_norm": 0.9453191903679632, + "learning_rate": 8.379927474037762e-07, + "loss": 0.1107, + "step": 8556 + }, + { + "epoch": 1.9469852104664391, + "grad_norm": 2.383443887327382, + "learning_rate": 8.379087424686458e-07, + "loss": 0.1473, + "step": 8557 + }, + { + "epoch": 1.9472127417519909, + "grad_norm": 2.248092514465891, + "learning_rate": 8.378247331820504e-07, + "loss": 0.114, + "step": 8558 + }, + { + "epoch": 1.9474402730375426, + "grad_norm": 1.7609556991713609, + "learning_rate": 8.377407195457077e-07, + "loss": 0.1203, + "step": 8559 + }, + { + "epoch": 1.9476678043230944, + "grad_norm": 2.468550221729246, + "learning_rate": 8.376567015613339e-07, + "loss": 0.0881, + "step": 8560 + }, + { + "epoch": 1.9478953356086461, + "grad_norm": 1.0006009709287198, + "learning_rate": 8.375726792306467e-07, + "loss": 0.0568, + "step": 8561 + }, + { + "epoch": 1.948122866894198, + "grad_norm": 1.5722493388504248, + "learning_rate": 8.374886525553632e-07, + "loss": 0.0684, + "step": 8562 + }, + { + "epoch": 1.9483503981797496, + "grad_norm": 1.7833042384805906, + "learning_rate": 8.374046215372011e-07, + "loss": 0.0784, + "step": 8563 + }, + { + "epoch": 1.9485779294653014, + "grad_norm": 2.2718997282755162, + "learning_rate": 8.37320586177877e-07, + "loss": 0.1032, + "step": 8564 + }, + { + "epoch": 1.9488054607508531, + "grad_norm": 2.021110083073223, + "learning_rate": 8.372365464791094e-07, + "loss": 0.1369, + "step": 8565 + }, + { + "epoch": 1.949032992036405, + "grad_norm": 1.893991429842655, + "learning_rate": 8.371525024426153e-07, + "loss": 0.1258, + "step": 8566 + }, + { + "epoch": 1.9492605233219567, + "grad_norm": 1.8229655764102353, + "learning_rate": 8.370684540701126e-07, + "loss": 0.1047, + "step": 8567 + }, + { + "epoch": 1.9494880546075084, + "grad_norm": 1.9936427878273355, + "learning_rate": 8.369844013633191e-07, + "loss": 0.1409, + "step": 8568 + }, + { + "epoch": 1.9497155858930602, + "grad_norm": 1.2484127425000624, + "learning_rate": 8.369003443239528e-07, + "loss": 0.0634, + "step": 8569 + }, + { + "epoch": 1.949943117178612, + "grad_norm": 2.4458940117366152, + "learning_rate": 8.368162829537313e-07, + "loss": 0.0893, + "step": 8570 + }, + { + "epoch": 1.9501706484641637, + "grad_norm": 1.3735546349913483, + "learning_rate": 8.367322172543729e-07, + "loss": 0.0971, + "step": 8571 + }, + { + "epoch": 1.9503981797497156, + "grad_norm": 1.0857680420382734, + "learning_rate": 8.366481472275961e-07, + "loss": 0.0363, + "step": 8572 + }, + { + "epoch": 1.9506257110352674, + "grad_norm": 2.9551954213167075, + "learning_rate": 8.365640728751184e-07, + "loss": 0.1919, + "step": 8573 + }, + { + "epoch": 1.9508532423208191, + "grad_norm": 1.8609193821716674, + "learning_rate": 8.364799941986587e-07, + "loss": 0.0689, + "step": 8574 + }, + { + "epoch": 1.9510807736063709, + "grad_norm": 2.45267786058061, + "learning_rate": 8.363959111999352e-07, + "loss": 0.1152, + "step": 8575 + }, + { + "epoch": 1.9513083048919226, + "grad_norm": 1.9628407395323646, + "learning_rate": 8.363118238806663e-07, + "loss": 0.0626, + "step": 8576 + }, + { + "epoch": 1.9515358361774744, + "grad_norm": 1.8859253724260385, + "learning_rate": 8.362277322425709e-07, + "loss": 0.0701, + "step": 8577 + }, + { + "epoch": 1.9517633674630261, + "grad_norm": 1.8757889787827413, + "learning_rate": 8.361436362873676e-07, + "loss": 0.1373, + "step": 8578 + }, + { + "epoch": 1.951990898748578, + "grad_norm": 1.6478100869469023, + "learning_rate": 8.360595360167748e-07, + "loss": 0.1116, + "step": 8579 + }, + { + "epoch": 1.9522184300341296, + "grad_norm": 2.7164270881373214, + "learning_rate": 8.359754314325117e-07, + "loss": 0.2087, + "step": 8580 + }, + { + "epoch": 1.9524459613196816, + "grad_norm": 1.6867400681793747, + "learning_rate": 8.358913225362972e-07, + "loss": 0.0968, + "step": 8581 + }, + { + "epoch": 1.9526734926052334, + "grad_norm": 1.6519218488263627, + "learning_rate": 8.358072093298503e-07, + "loss": 0.0727, + "step": 8582 + }, + { + "epoch": 1.9529010238907851, + "grad_norm": 1.833600292382146, + "learning_rate": 8.357230918148901e-07, + "loss": 0.1168, + "step": 8583 + }, + { + "epoch": 1.9531285551763369, + "grad_norm": 1.3802649701297778, + "learning_rate": 8.356389699931359e-07, + "loss": 0.1326, + "step": 8584 + }, + { + "epoch": 1.9533560864618886, + "grad_norm": 2.21485988798864, + "learning_rate": 8.35554843866307e-07, + "loss": 0.0858, + "step": 8585 + }, + { + "epoch": 1.9535836177474404, + "grad_norm": 2.0919450611286816, + "learning_rate": 8.354707134361224e-07, + "loss": 0.1045, + "step": 8586 + }, + { + "epoch": 1.9538111490329921, + "grad_norm": 2.8589127146761113, + "learning_rate": 8.35386578704302e-07, + "loss": 0.0908, + "step": 8587 + }, + { + "epoch": 1.9540386803185439, + "grad_norm": 1.4441818031879883, + "learning_rate": 8.353024396725653e-07, + "loss": 0.0924, + "step": 8588 + }, + { + "epoch": 1.9542662116040956, + "grad_norm": 2.2692155542041657, + "learning_rate": 8.352182963426317e-07, + "loss": 0.1268, + "step": 8589 + }, + { + "epoch": 1.9544937428896474, + "grad_norm": 1.7174063991169783, + "learning_rate": 8.351341487162214e-07, + "loss": 0.1364, + "step": 8590 + }, + { + "epoch": 1.9547212741751991, + "grad_norm": 2.47929661092712, + "learning_rate": 8.350499967950538e-07, + "loss": 0.183, + "step": 8591 + }, + { + "epoch": 1.954948805460751, + "grad_norm": 0.9115396121552741, + "learning_rate": 8.349658405808489e-07, + "loss": 0.047, + "step": 8592 + }, + { + "epoch": 1.9551763367463026, + "grad_norm": 1.6855778243414008, + "learning_rate": 8.348816800753269e-07, + "loss": 0.0586, + "step": 8593 + }, + { + "epoch": 1.9554038680318544, + "grad_norm": 1.410626305008107, + "learning_rate": 8.347975152802075e-07, + "loss": 0.0732, + "step": 8594 + }, + { + "epoch": 1.9556313993174061, + "grad_norm": 2.3121651096996807, + "learning_rate": 8.347133461972112e-07, + "loss": 0.1034, + "step": 8595 + }, + { + "epoch": 1.955858930602958, + "grad_norm": 2.1842489314084768, + "learning_rate": 8.346291728280584e-07, + "loss": 0.1254, + "step": 8596 + }, + { + "epoch": 1.9560864618885097, + "grad_norm": 1.2125108398522104, + "learning_rate": 8.345449951744692e-07, + "loss": 0.0821, + "step": 8597 + }, + { + "epoch": 1.9563139931740614, + "grad_norm": 2.32704489508761, + "learning_rate": 8.344608132381637e-07, + "loss": 0.0857, + "step": 8598 + }, + { + "epoch": 1.9565415244596132, + "grad_norm": 1.476103066454572, + "learning_rate": 8.343766270208631e-07, + "loss": 0.0973, + "step": 8599 + }, + { + "epoch": 1.956769055745165, + "grad_norm": 2.1631455138407243, + "learning_rate": 8.342924365242878e-07, + "loss": 0.0974, + "step": 8600 + }, + { + "epoch": 1.9569965870307167, + "grad_norm": 1.4889235511311651, + "learning_rate": 8.342082417501579e-07, + "loss": 0.0426, + "step": 8601 + }, + { + "epoch": 1.9572241183162684, + "grad_norm": 1.3914138563247866, + "learning_rate": 8.341240427001951e-07, + "loss": 0.0676, + "step": 8602 + }, + { + "epoch": 1.9574516496018202, + "grad_norm": 2.746354690345646, + "learning_rate": 8.340398393761199e-07, + "loss": 0.096, + "step": 8603 + }, + { + "epoch": 1.957679180887372, + "grad_norm": 1.7950527438701973, + "learning_rate": 8.339556317796529e-07, + "loss": 0.0526, + "step": 8604 + }, + { + "epoch": 1.9579067121729237, + "grad_norm": 1.4717345104063482, + "learning_rate": 8.338714199125157e-07, + "loss": 0.0565, + "step": 8605 + }, + { + "epoch": 1.9581342434584754, + "grad_norm": 1.9075516581715017, + "learning_rate": 8.337872037764292e-07, + "loss": 0.0599, + "step": 8606 + }, + { + "epoch": 1.9583617747440272, + "grad_norm": 2.264996914994366, + "learning_rate": 8.337029833731145e-07, + "loss": 0.0812, + "step": 8607 + }, + { + "epoch": 1.958589306029579, + "grad_norm": 1.3912222017896145, + "learning_rate": 8.336187587042932e-07, + "loss": 0.042, + "step": 8608 + }, + { + "epoch": 1.9588168373151307, + "grad_norm": 1.2403469619206942, + "learning_rate": 8.335345297716863e-07, + "loss": 0.0808, + "step": 8609 + }, + { + "epoch": 1.9590443686006824, + "grad_norm": 1.7039712007849193, + "learning_rate": 8.334502965770158e-07, + "loss": 0.102, + "step": 8610 + }, + { + "epoch": 1.9592718998862344, + "grad_norm": 2.188588003821983, + "learning_rate": 8.33366059122003e-07, + "loss": 0.0744, + "step": 8611 + }, + { + "epoch": 1.9594994311717862, + "grad_norm": 1.5541202044831102, + "learning_rate": 8.332818174083694e-07, + "loss": 0.0423, + "step": 8612 + }, + { + "epoch": 1.959726962457338, + "grad_norm": 2.35657294378755, + "learning_rate": 8.331975714378369e-07, + "loss": 0.0667, + "step": 8613 + }, + { + "epoch": 1.9599544937428897, + "grad_norm": 2.9190505646170095, + "learning_rate": 8.331133212121273e-07, + "loss": 0.1226, + "step": 8614 + }, + { + "epoch": 1.9601820250284414, + "grad_norm": 1.074796870429527, + "learning_rate": 8.330290667329627e-07, + "loss": 0.0824, + "step": 8615 + }, + { + "epoch": 1.9604095563139932, + "grad_norm": 1.6689032764316707, + "learning_rate": 8.329448080020651e-07, + "loss": 0.15, + "step": 8616 + }, + { + "epoch": 1.960637087599545, + "grad_norm": 1.3409500924708422, + "learning_rate": 8.32860545021156e-07, + "loss": 0.0867, + "step": 8617 + }, + { + "epoch": 1.9608646188850967, + "grad_norm": 3.7050291604272387, + "learning_rate": 8.327762777919585e-07, + "loss": 0.0839, + "step": 8618 + }, + { + "epoch": 1.9610921501706484, + "grad_norm": 0.8218866668948319, + "learning_rate": 8.326920063161942e-07, + "loss": 0.0467, + "step": 8619 + }, + { + "epoch": 1.9613196814562004, + "grad_norm": 1.0653092132142425, + "learning_rate": 8.326077305955857e-07, + "loss": 0.0754, + "step": 8620 + }, + { + "epoch": 1.9615472127417521, + "grad_norm": 1.8128981252159282, + "learning_rate": 8.325234506318553e-07, + "loss": 0.0821, + "step": 8621 + }, + { + "epoch": 1.961774744027304, + "grad_norm": 1.6899341568403192, + "learning_rate": 8.324391664267257e-07, + "loss": 0.0888, + "step": 8622 + }, + { + "epoch": 1.9620022753128556, + "grad_norm": 2.3338714314112, + "learning_rate": 8.323548779819194e-07, + "loss": 0.0576, + "step": 8623 + }, + { + "epoch": 1.9622298065984074, + "grad_norm": 1.4170197276530796, + "learning_rate": 8.322705852991592e-07, + "loss": 0.1294, + "step": 8624 + }, + { + "epoch": 1.9624573378839592, + "grad_norm": 1.6082562993048413, + "learning_rate": 8.321862883801677e-07, + "loss": 0.0561, + "step": 8625 + }, + { + "epoch": 1.962684869169511, + "grad_norm": 2.067958132163914, + "learning_rate": 8.321019872266682e-07, + "loss": 0.1137, + "step": 8626 + }, + { + "epoch": 1.9629124004550627, + "grad_norm": 1.5830819168127874, + "learning_rate": 8.320176818403831e-07, + "loss": 0.0839, + "step": 8627 + }, + { + "epoch": 1.9631399317406144, + "grad_norm": 1.9808658987468761, + "learning_rate": 8.319333722230359e-07, + "loss": 0.1352, + "step": 8628 + }, + { + "epoch": 1.9633674630261662, + "grad_norm": 1.2815991943265168, + "learning_rate": 8.318490583763494e-07, + "loss": 0.0971, + "step": 8629 + }, + { + "epoch": 1.963594994311718, + "grad_norm": 1.9747071861845462, + "learning_rate": 8.317647403020472e-07, + "loss": 0.229, + "step": 8630 + }, + { + "epoch": 1.9638225255972697, + "grad_norm": 1.3506976246768094, + "learning_rate": 8.316804180018522e-07, + "loss": 0.0412, + "step": 8631 + }, + { + "epoch": 1.9640500568828214, + "grad_norm": 1.4891462190541658, + "learning_rate": 8.315960914774878e-07, + "loss": 0.0689, + "step": 8632 + }, + { + "epoch": 1.9642775881683732, + "grad_norm": 1.6604876475552375, + "learning_rate": 8.31511760730678e-07, + "loss": 0.1113, + "step": 8633 + }, + { + "epoch": 1.964505119453925, + "grad_norm": 1.7505201034146314, + "learning_rate": 8.314274257631458e-07, + "loss": 0.0753, + "step": 8634 + }, + { + "epoch": 1.9647326507394767, + "grad_norm": 2.223169669274098, + "learning_rate": 8.313430865766154e-07, + "loss": 0.0839, + "step": 8635 + }, + { + "epoch": 1.9649601820250284, + "grad_norm": 2.581921503540097, + "learning_rate": 8.312587431728098e-07, + "loss": 0.071, + "step": 8636 + }, + { + "epoch": 1.9651877133105802, + "grad_norm": 1.866436868374168, + "learning_rate": 8.311743955534533e-07, + "loss": 0.1104, + "step": 8637 + }, + { + "epoch": 1.965415244596132, + "grad_norm": 1.222648355666239, + "learning_rate": 8.310900437202699e-07, + "loss": 0.0565, + "step": 8638 + }, + { + "epoch": 1.9656427758816837, + "grad_norm": 1.990385424763967, + "learning_rate": 8.310056876749833e-07, + "loss": 0.1203, + "step": 8639 + }, + { + "epoch": 1.9658703071672354, + "grad_norm": 1.5209309282517698, + "learning_rate": 8.309213274193179e-07, + "loss": 0.138, + "step": 8640 + }, + { + "epoch": 1.9660978384527872, + "grad_norm": 2.370651074382666, + "learning_rate": 8.308369629549976e-07, + "loss": 0.0579, + "step": 8641 + }, + { + "epoch": 1.966325369738339, + "grad_norm": 1.6617922953603113, + "learning_rate": 8.307525942837468e-07, + "loss": 0.1192, + "step": 8642 + }, + { + "epoch": 1.9665529010238907, + "grad_norm": 2.0286903030901375, + "learning_rate": 8.306682214072897e-07, + "loss": 0.2664, + "step": 8643 + }, + { + "epoch": 1.9667804323094424, + "grad_norm": 1.3145498331251388, + "learning_rate": 8.305838443273508e-07, + "loss": 0.1039, + "step": 8644 + }, + { + "epoch": 1.9670079635949942, + "grad_norm": 2.292862361880169, + "learning_rate": 8.304994630456546e-07, + "loss": 0.1654, + "step": 8645 + }, + { + "epoch": 1.967235494880546, + "grad_norm": 2.1230943172193157, + "learning_rate": 8.304150775639258e-07, + "loss": 0.0782, + "step": 8646 + }, + { + "epoch": 1.9674630261660977, + "grad_norm": 2.0206529007810183, + "learning_rate": 8.303306878838892e-07, + "loss": 0.1018, + "step": 8647 + }, + { + "epoch": 1.9676905574516494, + "grad_norm": 1.0586918202520623, + "learning_rate": 8.302462940072691e-07, + "loss": 0.0425, + "step": 8648 + }, + { + "epoch": 1.9679180887372012, + "grad_norm": 1.7279422628295729, + "learning_rate": 8.301618959357908e-07, + "loss": 0.1064, + "step": 8649 + }, + { + "epoch": 1.9681456200227532, + "grad_norm": 1.864109156077631, + "learning_rate": 8.300774936711792e-07, + "loss": 0.1265, + "step": 8650 + }, + { + "epoch": 1.968373151308305, + "grad_norm": 11.346775621475524, + "learning_rate": 8.299930872151589e-07, + "loss": 0.1063, + "step": 8651 + }, + { + "epoch": 1.9686006825938567, + "grad_norm": 4.708817423035009, + "learning_rate": 8.299086765694554e-07, + "loss": 0.1209, + "step": 8652 + }, + { + "epoch": 1.9688282138794084, + "grad_norm": 1.6750224470810209, + "learning_rate": 8.298242617357939e-07, + "loss": 0.0439, + "step": 8653 + }, + { + "epoch": 1.9690557451649602, + "grad_norm": 1.7314343639955407, + "learning_rate": 8.297398427158996e-07, + "loss": 0.1048, + "step": 8654 + }, + { + "epoch": 1.969283276450512, + "grad_norm": 2.9630009476977377, + "learning_rate": 8.296554195114979e-07, + "loss": 0.1042, + "step": 8655 + }, + { + "epoch": 1.9695108077360637, + "grad_norm": 1.8404252877953569, + "learning_rate": 8.295709921243143e-07, + "loss": 0.1182, + "step": 8656 + }, + { + "epoch": 1.9697383390216154, + "grad_norm": 1.2324941962130138, + "learning_rate": 8.294865605560743e-07, + "loss": 0.0501, + "step": 8657 + }, + { + "epoch": 1.9699658703071672, + "grad_norm": 1.4366065854180936, + "learning_rate": 8.294021248085032e-07, + "loss": 0.082, + "step": 8658 + }, + { + "epoch": 1.9701934015927192, + "grad_norm": 1.478003504693769, + "learning_rate": 8.293176848833273e-07, + "loss": 0.1013, + "step": 8659 + }, + { + "epoch": 1.970420932878271, + "grad_norm": 1.5789170948783189, + "learning_rate": 8.292332407822721e-07, + "loss": 0.0883, + "step": 8660 + }, + { + "epoch": 1.9706484641638227, + "grad_norm": 1.7038465627576111, + "learning_rate": 8.291487925070634e-07, + "loss": 0.0388, + "step": 8661 + }, + { + "epoch": 1.9708759954493744, + "grad_norm": 1.6944943552866523, + "learning_rate": 8.290643400594273e-07, + "loss": 0.0665, + "step": 8662 + }, + { + "epoch": 1.9711035267349262, + "grad_norm": 1.631425973583011, + "learning_rate": 8.289798834410898e-07, + "loss": 0.0694, + "step": 8663 + }, + { + "epoch": 1.971331058020478, + "grad_norm": 0.9533971656815422, + "learning_rate": 8.288954226537768e-07, + "loss": 0.0392, + "step": 8664 + }, + { + "epoch": 1.9715585893060297, + "grad_norm": 1.5917535714667914, + "learning_rate": 8.288109576992151e-07, + "loss": 0.1289, + "step": 8665 + }, + { + "epoch": 1.9717861205915814, + "grad_norm": 1.809822780479361, + "learning_rate": 8.287264885791308e-07, + "loss": 0.0719, + "step": 8666 + }, + { + "epoch": 1.9720136518771332, + "grad_norm": 1.578200781080261, + "learning_rate": 8.286420152952499e-07, + "loss": 0.0866, + "step": 8667 + }, + { + "epoch": 1.972241183162685, + "grad_norm": 2.8878884989148244, + "learning_rate": 8.285575378492992e-07, + "loss": 0.1145, + "step": 8668 + }, + { + "epoch": 1.9724687144482367, + "grad_norm": 1.5615581436014474, + "learning_rate": 8.284730562430053e-07, + "loss": 0.1209, + "step": 8669 + }, + { + "epoch": 1.9726962457337884, + "grad_norm": 1.3844281684842732, + "learning_rate": 8.283885704780947e-07, + "loss": 0.0842, + "step": 8670 + }, + { + "epoch": 1.9729237770193402, + "grad_norm": 2.387948296712115, + "learning_rate": 8.283040805562942e-07, + "loss": 0.0832, + "step": 8671 + }, + { + "epoch": 1.973151308304892, + "grad_norm": 1.7102712047888355, + "learning_rate": 8.282195864793307e-07, + "loss": 0.1447, + "step": 8672 + }, + { + "epoch": 1.9733788395904437, + "grad_norm": 1.4547768267132164, + "learning_rate": 8.281350882489311e-07, + "loss": 0.0597, + "step": 8673 + }, + { + "epoch": 1.9736063708759954, + "grad_norm": 1.8652494247457443, + "learning_rate": 8.280505858668222e-07, + "loss": 0.1626, + "step": 8674 + }, + { + "epoch": 1.9738339021615472, + "grad_norm": 2.1832358441304534, + "learning_rate": 8.279660793347314e-07, + "loss": 0.1992, + "step": 8675 + }, + { + "epoch": 1.974061433447099, + "grad_norm": 1.4624364468009274, + "learning_rate": 8.278815686543854e-07, + "loss": 0.1492, + "step": 8676 + }, + { + "epoch": 1.9742889647326507, + "grad_norm": 2.2089614656107153, + "learning_rate": 8.277970538275118e-07, + "loss": 0.067, + "step": 8677 + }, + { + "epoch": 1.9745164960182024, + "grad_norm": 1.3650501445700982, + "learning_rate": 8.277125348558376e-07, + "loss": 0.0592, + "step": 8678 + }, + { + "epoch": 1.9747440273037542, + "grad_norm": 3.0771392361969356, + "learning_rate": 8.276280117410909e-07, + "loss": 0.0427, + "step": 8679 + }, + { + "epoch": 1.974971558589306, + "grad_norm": 1.2119852743527402, + "learning_rate": 8.275434844849985e-07, + "loss": 0.0656, + "step": 8680 + }, + { + "epoch": 1.9751990898748577, + "grad_norm": 2.2806841761919965, + "learning_rate": 8.274589530892881e-07, + "loss": 0.0588, + "step": 8681 + }, + { + "epoch": 1.9754266211604095, + "grad_norm": 2.55583141126003, + "learning_rate": 8.273744175556877e-07, + "loss": 0.1209, + "step": 8682 + }, + { + "epoch": 1.9756541524459612, + "grad_norm": 1.4254587602708517, + "learning_rate": 8.272898778859247e-07, + "loss": 0.0723, + "step": 8683 + }, + { + "epoch": 1.975881683731513, + "grad_norm": 2.673288718810534, + "learning_rate": 8.272053340817271e-07, + "loss": 0.0835, + "step": 8684 + }, + { + "epoch": 1.9761092150170647, + "grad_norm": 0.9119109227202765, + "learning_rate": 8.271207861448227e-07, + "loss": 0.0867, + "step": 8685 + }, + { + "epoch": 1.9763367463026165, + "grad_norm": 1.5946016036601556, + "learning_rate": 8.270362340769397e-07, + "loss": 0.1654, + "step": 8686 + }, + { + "epoch": 1.9765642775881682, + "grad_norm": 1.4365129045779053, + "learning_rate": 8.269516778798062e-07, + "loss": 0.1164, + "step": 8687 + }, + { + "epoch": 1.9767918088737202, + "grad_norm": 1.4889068357273756, + "learning_rate": 8.268671175551499e-07, + "loss": 0.0874, + "step": 8688 + }, + { + "epoch": 1.977019340159272, + "grad_norm": 1.886486084945335, + "learning_rate": 8.267825531046997e-07, + "loss": 0.0753, + "step": 8689 + }, + { + "epoch": 1.9772468714448237, + "grad_norm": 2.4827800257148493, + "learning_rate": 8.266979845301837e-07, + "loss": 0.1679, + "step": 8690 + }, + { + "epoch": 1.9774744027303754, + "grad_norm": 1.3579765440263711, + "learning_rate": 8.266134118333302e-07, + "loss": 0.0974, + "step": 8691 + }, + { + "epoch": 1.9777019340159272, + "grad_norm": 2.7996405259060584, + "learning_rate": 8.265288350158677e-07, + "loss": 0.1313, + "step": 8692 + }, + { + "epoch": 1.977929465301479, + "grad_norm": 1.7784935591636652, + "learning_rate": 8.26444254079525e-07, + "loss": 0.0611, + "step": 8693 + }, + { + "epoch": 1.9781569965870307, + "grad_norm": 2.153563806186732, + "learning_rate": 8.263596690260306e-07, + "loss": 0.1406, + "step": 8694 + }, + { + "epoch": 1.9783845278725825, + "grad_norm": 1.4147170215746951, + "learning_rate": 8.262750798571134e-07, + "loss": 0.0482, + "step": 8695 + }, + { + "epoch": 1.9786120591581342, + "grad_norm": 1.7619991838965041, + "learning_rate": 8.261904865745022e-07, + "loss": 0.1815, + "step": 8696 + }, + { + "epoch": 1.978839590443686, + "grad_norm": 1.1331623903514012, + "learning_rate": 8.26105889179926e-07, + "loss": 0.0692, + "step": 8697 + }, + { + "epoch": 1.979067121729238, + "grad_norm": 1.253627258938659, + "learning_rate": 8.260212876751137e-07, + "loss": 0.1328, + "step": 8698 + }, + { + "epoch": 1.9792946530147897, + "grad_norm": 2.0095171447623783, + "learning_rate": 8.259366820617943e-07, + "loss": 0.0824, + "step": 8699 + }, + { + "epoch": 1.9795221843003414, + "grad_norm": 1.7847243273657507, + "learning_rate": 8.258520723416972e-07, + "loss": 0.0839, + "step": 8700 + }, + { + "epoch": 1.9797497155858932, + "grad_norm": 1.1223884821522332, + "learning_rate": 8.257674585165515e-07, + "loss": 0.0501, + "step": 8701 + }, + { + "epoch": 1.979977246871445, + "grad_norm": 0.7103291026724687, + "learning_rate": 8.256828405880868e-07, + "loss": 0.0882, + "step": 8702 + }, + { + "epoch": 1.9802047781569967, + "grad_norm": 1.529906844117204, + "learning_rate": 8.255982185580323e-07, + "loss": 0.053, + "step": 8703 + }, + { + "epoch": 1.9804323094425484, + "grad_norm": 1.51643427579225, + "learning_rate": 8.255135924281175e-07, + "loss": 0.0542, + "step": 8704 + }, + { + "epoch": 1.9806598407281002, + "grad_norm": 1.93250010433533, + "learning_rate": 8.254289622000724e-07, + "loss": 0.0969, + "step": 8705 + }, + { + "epoch": 1.980887372013652, + "grad_norm": 2.222065038425281, + "learning_rate": 8.253443278756261e-07, + "loss": 0.0797, + "step": 8706 + }, + { + "epoch": 1.9811149032992037, + "grad_norm": 1.6875050782240095, + "learning_rate": 8.252596894565088e-07, + "loss": 0.1117, + "step": 8707 + }, + { + "epoch": 1.9813424345847555, + "grad_norm": 1.6021453556306384, + "learning_rate": 8.251750469444498e-07, + "loss": 0.062, + "step": 8708 + }, + { + "epoch": 1.9815699658703072, + "grad_norm": 1.3794430805592486, + "learning_rate": 8.250904003411798e-07, + "loss": 0.1158, + "step": 8709 + }, + { + "epoch": 1.981797497155859, + "grad_norm": 2.223816644362845, + "learning_rate": 8.250057496484285e-07, + "loss": 0.1028, + "step": 8710 + }, + { + "epoch": 1.9820250284414107, + "grad_norm": 1.902049328564107, + "learning_rate": 8.24921094867926e-07, + "loss": 0.0457, + "step": 8711 + }, + { + "epoch": 1.9822525597269625, + "grad_norm": 2.159891289436667, + "learning_rate": 8.248364360014023e-07, + "loss": 0.0958, + "step": 8712 + }, + { + "epoch": 1.9824800910125142, + "grad_norm": 2.5114187966886097, + "learning_rate": 8.247517730505879e-07, + "loss": 0.1056, + "step": 8713 + }, + { + "epoch": 1.982707622298066, + "grad_norm": 2.2504820631273086, + "learning_rate": 8.24667106017213e-07, + "loss": 0.0671, + "step": 8714 + }, + { + "epoch": 1.9829351535836177, + "grad_norm": 1.2019567535363975, + "learning_rate": 8.245824349030082e-07, + "loss": 0.0722, + "step": 8715 + }, + { + "epoch": 1.9831626848691695, + "grad_norm": 1.6629097763213867, + "learning_rate": 8.244977597097039e-07, + "loss": 0.1164, + "step": 8716 + }, + { + "epoch": 1.9833902161547212, + "grad_norm": 1.2374423438478135, + "learning_rate": 8.244130804390311e-07, + "loss": 0.051, + "step": 8717 + }, + { + "epoch": 1.983617747440273, + "grad_norm": 1.5062111724253495, + "learning_rate": 8.243283970927196e-07, + "loss": 0.1171, + "step": 8718 + }, + { + "epoch": 1.9838452787258247, + "grad_norm": 2.3205296578644656, + "learning_rate": 8.242437096725009e-07, + "loss": 0.0971, + "step": 8719 + }, + { + "epoch": 1.9840728100113765, + "grad_norm": 1.1268486989632436, + "learning_rate": 8.241590181801059e-07, + "loss": 0.0205, + "step": 8720 + }, + { + "epoch": 1.9843003412969282, + "grad_norm": 2.539102510420942, + "learning_rate": 8.240743226172651e-07, + "loss": 0.1488, + "step": 8721 + }, + { + "epoch": 1.98452787258248, + "grad_norm": 1.3846449421864462, + "learning_rate": 8.239896229857096e-07, + "loss": 0.1043, + "step": 8722 + }, + { + "epoch": 1.9847554038680317, + "grad_norm": 1.9851920729714756, + "learning_rate": 8.23904919287171e-07, + "loss": 0.2263, + "step": 8723 + }, + { + "epoch": 1.9849829351535835, + "grad_norm": 2.221709013257648, + "learning_rate": 8.2382021152338e-07, + "loss": 0.0977, + "step": 8724 + }, + { + "epoch": 1.9852104664391352, + "grad_norm": 1.6218224330242628, + "learning_rate": 8.237354996960678e-07, + "loss": 0.0854, + "step": 8725 + }, + { + "epoch": 1.985437997724687, + "grad_norm": 1.5029991830245644, + "learning_rate": 8.23650783806966e-07, + "loss": 0.1117, + "step": 8726 + }, + { + "epoch": 1.985665529010239, + "grad_norm": 2.9301067848601767, + "learning_rate": 8.235660638578061e-07, + "loss": 0.1432, + "step": 8727 + }, + { + "epoch": 1.9858930602957907, + "grad_norm": 3.2845770288624614, + "learning_rate": 8.234813398503194e-07, + "loss": 0.1129, + "step": 8728 + }, + { + "epoch": 1.9861205915813425, + "grad_norm": 0.8591511147372719, + "learning_rate": 8.233966117862378e-07, + "loss": 0.0442, + "step": 8729 + }, + { + "epoch": 1.9863481228668942, + "grad_norm": 1.2599805504047201, + "learning_rate": 8.233118796672929e-07, + "loss": 0.0706, + "step": 8730 + }, + { + "epoch": 1.986575654152446, + "grad_norm": 1.832085019662094, + "learning_rate": 8.23227143495216e-07, + "loss": 0.0444, + "step": 8731 + }, + { + "epoch": 1.9868031854379977, + "grad_norm": 2.725371399737415, + "learning_rate": 8.231424032717395e-07, + "loss": 0.1039, + "step": 8732 + }, + { + "epoch": 1.9870307167235495, + "grad_norm": 1.0608242847643243, + "learning_rate": 8.230576589985951e-07, + "loss": 0.0605, + "step": 8733 + }, + { + "epoch": 1.9872582480091012, + "grad_norm": 2.265480959394933, + "learning_rate": 8.22972910677515e-07, + "loss": 0.0924, + "step": 8734 + }, + { + "epoch": 1.987485779294653, + "grad_norm": 1.5908850088022235, + "learning_rate": 8.22888158310231e-07, + "loss": 0.084, + "step": 8735 + }, + { + "epoch": 1.9877133105802047, + "grad_norm": 1.5149168873070165, + "learning_rate": 8.228034018984757e-07, + "loss": 0.1045, + "step": 8736 + }, + { + "epoch": 1.9879408418657567, + "grad_norm": 1.601832771127956, + "learning_rate": 8.227186414439812e-07, + "loss": 0.1341, + "step": 8737 + }, + { + "epoch": 1.9881683731513085, + "grad_norm": 1.8254229787531513, + "learning_rate": 8.226338769484793e-07, + "loss": 0.07, + "step": 8738 + }, + { + "epoch": 1.9883959044368602, + "grad_norm": 1.364448391816041, + "learning_rate": 8.225491084137032e-07, + "loss": 0.0783, + "step": 8739 + }, + { + "epoch": 1.988623435722412, + "grad_norm": 2.3290448585354575, + "learning_rate": 8.22464335841385e-07, + "loss": 0.0726, + "step": 8740 + }, + { + "epoch": 1.9888509670079637, + "grad_norm": 1.5552876815848529, + "learning_rate": 8.223795592332575e-07, + "loss": 0.0713, + "step": 8741 + }, + { + "epoch": 1.9890784982935155, + "grad_norm": 1.5238339414276638, + "learning_rate": 8.222947785910534e-07, + "loss": 0.0605, + "step": 8742 + }, + { + "epoch": 1.9893060295790672, + "grad_norm": 1.8533393839233188, + "learning_rate": 8.222099939165053e-07, + "loss": 0.0658, + "step": 8743 + }, + { + "epoch": 1.989533560864619, + "grad_norm": 1.636769542071038, + "learning_rate": 8.221252052113461e-07, + "loss": 0.0686, + "step": 8744 + }, + { + "epoch": 1.9897610921501707, + "grad_norm": 1.1784979000680322, + "learning_rate": 8.220404124773084e-07, + "loss": 0.0454, + "step": 8745 + }, + { + "epoch": 1.9899886234357225, + "grad_norm": 1.207734253241908, + "learning_rate": 8.21955615716126e-07, + "loss": 0.0862, + "step": 8746 + }, + { + "epoch": 1.9902161547212742, + "grad_norm": 1.4545277819121272, + "learning_rate": 8.218708149295312e-07, + "loss": 0.0549, + "step": 8747 + }, + { + "epoch": 1.990443686006826, + "grad_norm": 2.7172887991799524, + "learning_rate": 8.217860101192578e-07, + "loss": 0.1243, + "step": 8748 + }, + { + "epoch": 1.9906712172923777, + "grad_norm": 1.702889527991951, + "learning_rate": 8.217012012870384e-07, + "loss": 0.1054, + "step": 8749 + }, + { + "epoch": 1.9908987485779295, + "grad_norm": 1.271613927088692, + "learning_rate": 8.21616388434607e-07, + "loss": 0.1486, + "step": 8750 + }, + { + "epoch": 1.9911262798634812, + "grad_norm": 1.8988243187589482, + "learning_rate": 8.215315715636965e-07, + "loss": 0.1423, + "step": 8751 + }, + { + "epoch": 1.991353811149033, + "grad_norm": 2.055876805575002, + "learning_rate": 8.214467506760407e-07, + "loss": 0.115, + "step": 8752 + }, + { + "epoch": 1.9915813424345847, + "grad_norm": 2.638508117319233, + "learning_rate": 8.213619257733729e-07, + "loss": 0.0905, + "step": 8753 + }, + { + "epoch": 1.9918088737201365, + "grad_norm": 2.567328517530897, + "learning_rate": 8.212770968574274e-07, + "loss": 0.1084, + "step": 8754 + }, + { + "epoch": 1.9920364050056882, + "grad_norm": 1.7075683935276236, + "learning_rate": 8.211922639299372e-07, + "loss": 0.1013, + "step": 8755 + }, + { + "epoch": 1.99226393629124, + "grad_norm": 1.648301967423068, + "learning_rate": 8.211074269926364e-07, + "loss": 0.0839, + "step": 8756 + }, + { + "epoch": 1.9924914675767917, + "grad_norm": 1.434701677591572, + "learning_rate": 8.21022586047259e-07, + "loss": 0.1214, + "step": 8757 + }, + { + "epoch": 1.9927189988623435, + "grad_norm": 1.8420000043262064, + "learning_rate": 8.209377410955388e-07, + "loss": 0.1467, + "step": 8758 + }, + { + "epoch": 1.9929465301478952, + "grad_norm": 2.52652405311473, + "learning_rate": 8.208528921392101e-07, + "loss": 0.1096, + "step": 8759 + }, + { + "epoch": 1.993174061433447, + "grad_norm": 1.1848532764347495, + "learning_rate": 8.207680391800071e-07, + "loss": 0.0315, + "step": 8760 + }, + { + "epoch": 1.9934015927189987, + "grad_norm": 1.6017535978855324, + "learning_rate": 8.206831822196639e-07, + "loss": 0.1236, + "step": 8761 + }, + { + "epoch": 1.9936291240045505, + "grad_norm": 1.649573992216395, + "learning_rate": 8.205983212599147e-07, + "loss": 0.0656, + "step": 8762 + }, + { + "epoch": 1.9938566552901023, + "grad_norm": 2.890250389013289, + "learning_rate": 8.205134563024942e-07, + "loss": 0.1369, + "step": 8763 + }, + { + "epoch": 1.994084186575654, + "grad_norm": 2.140472700627925, + "learning_rate": 8.204285873491366e-07, + "loss": 0.1049, + "step": 8764 + }, + { + "epoch": 1.9943117178612058, + "grad_norm": 2.7508207881108455, + "learning_rate": 8.203437144015766e-07, + "loss": 0.15, + "step": 8765 + }, + { + "epoch": 1.9945392491467577, + "grad_norm": 1.5774029621418546, + "learning_rate": 8.202588374615489e-07, + "loss": 0.0621, + "step": 8766 + }, + { + "epoch": 1.9947667804323095, + "grad_norm": 4.257497987165835, + "learning_rate": 8.201739565307881e-07, + "loss": 0.1181, + "step": 8767 + }, + { + "epoch": 1.9949943117178612, + "grad_norm": 1.54072123867852, + "learning_rate": 8.200890716110291e-07, + "loss": 0.0576, + "step": 8768 + }, + { + "epoch": 1.995221843003413, + "grad_norm": 1.3992445871747745, + "learning_rate": 8.200041827040067e-07, + "loss": 0.1309, + "step": 8769 + }, + { + "epoch": 1.9954493742889647, + "grad_norm": 2.107248853389515, + "learning_rate": 8.19919289811456e-07, + "loss": 0.074, + "step": 8770 + }, + { + "epoch": 1.9956769055745165, + "grad_norm": 1.7026980270034522, + "learning_rate": 8.19834392935112e-07, + "loss": 0.1173, + "step": 8771 + }, + { + "epoch": 1.9959044368600682, + "grad_norm": 1.8419011010512982, + "learning_rate": 8.197494920767098e-07, + "loss": 0.1412, + "step": 8772 + }, + { + "epoch": 1.99613196814562, + "grad_norm": 1.3318756882986866, + "learning_rate": 8.196645872379847e-07, + "loss": 0.0862, + "step": 8773 + }, + { + "epoch": 1.9963594994311717, + "grad_norm": 1.5624969864715754, + "learning_rate": 8.19579678420672e-07, + "loss": 0.0824, + "step": 8774 + }, + { + "epoch": 1.9965870307167235, + "grad_norm": 2.530990456405567, + "learning_rate": 8.194947656265068e-07, + "loss": 0.1119, + "step": 8775 + }, + { + "epoch": 1.9968145620022755, + "grad_norm": 1.0514160033556272, + "learning_rate": 8.19409848857225e-07, + "loss": 0.0535, + "step": 8776 + }, + { + "epoch": 1.9970420932878272, + "grad_norm": 1.4347125548012525, + "learning_rate": 8.193249281145618e-07, + "loss": 0.0278, + "step": 8777 + }, + { + "epoch": 1.997269624573379, + "grad_norm": 1.3797341750105707, + "learning_rate": 8.19240003400253e-07, + "loss": 0.0479, + "step": 8778 + }, + { + "epoch": 1.9974971558589307, + "grad_norm": 2.0241199844319366, + "learning_rate": 8.191550747160343e-07, + "loss": 0.1545, + "step": 8779 + }, + { + "epoch": 1.9977246871444825, + "grad_norm": 1.5886296816722167, + "learning_rate": 8.190701420636415e-07, + "loss": 0.1636, + "step": 8780 + }, + { + "epoch": 1.9979522184300342, + "grad_norm": 2.8866673579296904, + "learning_rate": 8.189852054448104e-07, + "loss": 0.0926, + "step": 8781 + }, + { + "epoch": 1.998179749715586, + "grad_norm": 1.8291149977747132, + "learning_rate": 8.189002648612768e-07, + "loss": 0.0636, + "step": 8782 + }, + { + "epoch": 1.9984072810011377, + "grad_norm": 1.891164270992027, + "learning_rate": 8.188153203147769e-07, + "loss": 0.1051, + "step": 8783 + }, + { + "epoch": 1.9986348122866895, + "grad_norm": 1.8788867320034368, + "learning_rate": 8.18730371807047e-07, + "loss": 0.0737, + "step": 8784 + }, + { + "epoch": 1.9988623435722412, + "grad_norm": 1.7781478435152995, + "learning_rate": 8.18645419339823e-07, + "loss": 0.0693, + "step": 8785 + }, + { + "epoch": 1.999089874857793, + "grad_norm": 2.3742153897723957, + "learning_rate": 8.185604629148413e-07, + "loss": 0.1433, + "step": 8786 + }, + { + "epoch": 1.9993174061433447, + "grad_norm": 1.5317856218091108, + "learning_rate": 8.184755025338384e-07, + "loss": 0.1055, + "step": 8787 + }, + { + "epoch": 1.9995449374288965, + "grad_norm": 1.9943110303869713, + "learning_rate": 8.183905381985503e-07, + "loss": 0.1895, + "step": 8788 + }, + { + "epoch": 1.9997724687144482, + "grad_norm": 1.6863939890780129, + "learning_rate": 8.183055699107139e-07, + "loss": 0.0721, + "step": 8789 + }, + { + "epoch": 2.0, + "grad_norm": 1.1227135701228614, + "learning_rate": 8.182205976720656e-07, + "loss": 0.0323, + "step": 8790 + }, + { + "epoch": 2.0002275312855518, + "grad_norm": 1.639248317330326, + "learning_rate": 8.181356214843422e-07, + "loss": 0.0745, + "step": 8791 + }, + { + "epoch": 2.0004550625711035, + "grad_norm": 1.1149408311128595, + "learning_rate": 8.180506413492804e-07, + "loss": 0.0451, + "step": 8792 + }, + { + "epoch": 2.0006825938566553, + "grad_norm": 1.6891048132315853, + "learning_rate": 8.17965657268617e-07, + "loss": 0.0988, + "step": 8793 + }, + { + "epoch": 2.000910125142207, + "grad_norm": 1.27443946664208, + "learning_rate": 8.178806692440891e-07, + "loss": 0.0492, + "step": 8794 + }, + { + "epoch": 2.0011376564277588, + "grad_norm": 1.2380161041122009, + "learning_rate": 8.177956772774334e-07, + "loss": 0.0448, + "step": 8795 + }, + { + "epoch": 2.0013651877133105, + "grad_norm": 1.7709227835300692, + "learning_rate": 8.177106813703872e-07, + "loss": 0.0489, + "step": 8796 + }, + { + "epoch": 2.0015927189988623, + "grad_norm": 1.524886744541349, + "learning_rate": 8.176256815246878e-07, + "loss": 0.058, + "step": 8797 + }, + { + "epoch": 2.001820250284414, + "grad_norm": 1.5196347556733598, + "learning_rate": 8.175406777420721e-07, + "loss": 0.0764, + "step": 8798 + }, + { + "epoch": 2.0020477815699658, + "grad_norm": 0.8129924881751084, + "learning_rate": 8.174556700242775e-07, + "loss": 0.0473, + "step": 8799 + }, + { + "epoch": 2.0022753128555175, + "grad_norm": 0.9728735219017393, + "learning_rate": 8.173706583730414e-07, + "loss": 0.0632, + "step": 8800 + }, + { + "epoch": 2.0025028441410693, + "grad_norm": 0.825161854835932, + "learning_rate": 8.172856427901015e-07, + "loss": 0.0627, + "step": 8801 + }, + { + "epoch": 2.002730375426621, + "grad_norm": 1.1071985751788505, + "learning_rate": 8.17200623277195e-07, + "loss": 0.0827, + "step": 8802 + }, + { + "epoch": 2.0029579067121728, + "grad_norm": 0.9408215131976463, + "learning_rate": 8.171155998360601e-07, + "loss": 0.0691, + "step": 8803 + }, + { + "epoch": 2.0031854379977245, + "grad_norm": 1.2887530621852699, + "learning_rate": 8.170305724684341e-07, + "loss": 0.0919, + "step": 8804 + }, + { + "epoch": 2.0034129692832763, + "grad_norm": 1.6152645826005223, + "learning_rate": 8.169455411760547e-07, + "loss": 0.0437, + "step": 8805 + }, + { + "epoch": 2.003640500568828, + "grad_norm": 1.8198581728956609, + "learning_rate": 8.168605059606601e-07, + "loss": 0.1383, + "step": 8806 + }, + { + "epoch": 2.00386803185438, + "grad_norm": 1.4179904940921069, + "learning_rate": 8.167754668239883e-07, + "loss": 0.0828, + "step": 8807 + }, + { + "epoch": 2.0040955631399315, + "grad_norm": 1.5926951533212381, + "learning_rate": 8.16690423767777e-07, + "loss": 0.1072, + "step": 8808 + }, + { + "epoch": 2.0043230944254833, + "grad_norm": 1.2223116840843713, + "learning_rate": 8.166053767937643e-07, + "loss": 0.0738, + "step": 8809 + }, + { + "epoch": 2.0045506257110355, + "grad_norm": 1.445344991442188, + "learning_rate": 8.165203259036888e-07, + "loss": 0.0428, + "step": 8810 + }, + { + "epoch": 2.0047781569965872, + "grad_norm": 0.9090576750966644, + "learning_rate": 8.164352710992887e-07, + "loss": 0.0741, + "step": 8811 + }, + { + "epoch": 2.005005688282139, + "grad_norm": 1.7276762171111015, + "learning_rate": 8.163502123823021e-07, + "loss": 0.0435, + "step": 8812 + }, + { + "epoch": 2.0052332195676907, + "grad_norm": 1.492141595193191, + "learning_rate": 8.162651497544677e-07, + "loss": 0.0351, + "step": 8813 + }, + { + "epoch": 2.0054607508532425, + "grad_norm": 1.3715675122970237, + "learning_rate": 8.161800832175239e-07, + "loss": 0.0424, + "step": 8814 + }, + { + "epoch": 2.0056882821387942, + "grad_norm": 1.7638913521749375, + "learning_rate": 8.160950127732093e-07, + "loss": 0.0776, + "step": 8815 + }, + { + "epoch": 2.005915813424346, + "grad_norm": 0.735937669675378, + "learning_rate": 8.160099384232625e-07, + "loss": 0.0524, + "step": 8816 + }, + { + "epoch": 2.0061433447098977, + "grad_norm": 1.4131055981317602, + "learning_rate": 8.159248601694226e-07, + "loss": 0.0509, + "step": 8817 + }, + { + "epoch": 2.0063708759954495, + "grad_norm": 1.4558188081430434, + "learning_rate": 8.158397780134281e-07, + "loss": 0.0303, + "step": 8818 + }, + { + "epoch": 2.0065984072810013, + "grad_norm": 2.320515554826066, + "learning_rate": 8.157546919570181e-07, + "loss": 0.0373, + "step": 8819 + }, + { + "epoch": 2.006825938566553, + "grad_norm": 2.027702445818353, + "learning_rate": 8.156696020019314e-07, + "loss": 0.0575, + "step": 8820 + }, + { + "epoch": 2.0070534698521048, + "grad_norm": 2.6298986980382972, + "learning_rate": 8.155845081499074e-07, + "loss": 0.1308, + "step": 8821 + }, + { + "epoch": 2.0072810011376565, + "grad_norm": 1.694566079523619, + "learning_rate": 8.154994104026849e-07, + "loss": 0.0606, + "step": 8822 + }, + { + "epoch": 2.0075085324232083, + "grad_norm": 1.4030073007426522, + "learning_rate": 8.154143087620035e-07, + "loss": 0.0316, + "step": 8823 + }, + { + "epoch": 2.00773606370876, + "grad_norm": 1.6338153612852564, + "learning_rate": 8.153292032296025e-07, + "loss": 0.0435, + "step": 8824 + }, + { + "epoch": 2.0079635949943118, + "grad_norm": 1.4876129806903389, + "learning_rate": 8.152440938072208e-07, + "loss": 0.1292, + "step": 8825 + }, + { + "epoch": 2.0081911262798635, + "grad_norm": 1.9032960662440848, + "learning_rate": 8.151589804965984e-07, + "loss": 0.0339, + "step": 8826 + }, + { + "epoch": 2.0084186575654153, + "grad_norm": 1.4548831058798386, + "learning_rate": 8.150738632994748e-07, + "loss": 0.1377, + "step": 8827 + }, + { + "epoch": 2.008646188850967, + "grad_norm": 2.4131691841176623, + "learning_rate": 8.149887422175895e-07, + "loss": 0.0382, + "step": 8828 + }, + { + "epoch": 2.0088737201365188, + "grad_norm": 1.3122056977883902, + "learning_rate": 8.149036172526821e-07, + "loss": 0.0725, + "step": 8829 + }, + { + "epoch": 2.0091012514220705, + "grad_norm": 1.0099883818234812, + "learning_rate": 8.148184884064928e-07, + "loss": 0.0446, + "step": 8830 + }, + { + "epoch": 2.0093287827076223, + "grad_norm": 1.8582106873286008, + "learning_rate": 8.14733355680761e-07, + "loss": 0.0424, + "step": 8831 + }, + { + "epoch": 2.009556313993174, + "grad_norm": 1.6924905733400868, + "learning_rate": 8.146482190772271e-07, + "loss": 0.1153, + "step": 8832 + }, + { + "epoch": 2.0097838452787258, + "grad_norm": 1.8432234472230091, + "learning_rate": 8.145630785976307e-07, + "loss": 0.0517, + "step": 8833 + }, + { + "epoch": 2.0100113765642775, + "grad_norm": 1.7375043175991889, + "learning_rate": 8.144779342437123e-07, + "loss": 0.0457, + "step": 8834 + }, + { + "epoch": 2.0102389078498293, + "grad_norm": 1.342260023405727, + "learning_rate": 8.143927860172118e-07, + "loss": 0.0304, + "step": 8835 + }, + { + "epoch": 2.010466439135381, + "grad_norm": 0.911422332247528, + "learning_rate": 8.143076339198698e-07, + "loss": 0.0395, + "step": 8836 + }, + { + "epoch": 2.010693970420933, + "grad_norm": 3.5170367489675023, + "learning_rate": 8.142224779534263e-07, + "loss": 0.0322, + "step": 8837 + }, + { + "epoch": 2.0109215017064845, + "grad_norm": 2.261756172333368, + "learning_rate": 8.14137318119622e-07, + "loss": 0.0608, + "step": 8838 + }, + { + "epoch": 2.0111490329920363, + "grad_norm": 1.2494008164346324, + "learning_rate": 8.140521544201972e-07, + "loss": 0.0542, + "step": 8839 + }, + { + "epoch": 2.011376564277588, + "grad_norm": 0.9523857994791479, + "learning_rate": 8.139669868568927e-07, + "loss": 0.0295, + "step": 8840 + }, + { + "epoch": 2.01160409556314, + "grad_norm": 1.838077742603398, + "learning_rate": 8.13881815431449e-07, + "loss": 0.16, + "step": 8841 + }, + { + "epoch": 2.0118316268486915, + "grad_norm": 0.8507938682088919, + "learning_rate": 8.13796640145607e-07, + "loss": 0.0488, + "step": 8842 + }, + { + "epoch": 2.0120591581342433, + "grad_norm": 1.9073506063694765, + "learning_rate": 8.137114610011074e-07, + "loss": 0.1008, + "step": 8843 + }, + { + "epoch": 2.012286689419795, + "grad_norm": 2.059551748427677, + "learning_rate": 8.136262779996912e-07, + "loss": 0.0701, + "step": 8844 + }, + { + "epoch": 2.012514220705347, + "grad_norm": 1.713946499618614, + "learning_rate": 8.135410911430992e-07, + "loss": 0.0454, + "step": 8845 + }, + { + "epoch": 2.0127417519908986, + "grad_norm": 1.8394902270979199, + "learning_rate": 8.134559004330725e-07, + "loss": 0.0409, + "step": 8846 + }, + { + "epoch": 2.0129692832764503, + "grad_norm": 1.276868434763315, + "learning_rate": 8.133707058713525e-07, + "loss": 0.0334, + "step": 8847 + }, + { + "epoch": 2.013196814562002, + "grad_norm": 1.2184236082444977, + "learning_rate": 8.132855074596803e-07, + "loss": 0.0253, + "step": 8848 + }, + { + "epoch": 2.0134243458475543, + "grad_norm": 1.5754488122016854, + "learning_rate": 8.132003051997972e-07, + "loss": 0.0844, + "step": 8849 + }, + { + "epoch": 2.013651877133106, + "grad_norm": 1.491733201506274, + "learning_rate": 8.131150990934445e-07, + "loss": 0.0533, + "step": 8850 + }, + { + "epoch": 2.0138794084186578, + "grad_norm": 1.3402104757934845, + "learning_rate": 8.130298891423636e-07, + "loss": 0.0915, + "step": 8851 + }, + { + "epoch": 2.0141069397042095, + "grad_norm": 1.1312502357304939, + "learning_rate": 8.12944675348296e-07, + "loss": 0.0338, + "step": 8852 + }, + { + "epoch": 2.0143344709897613, + "grad_norm": 1.67824625782941, + "learning_rate": 8.128594577129836e-07, + "loss": 0.0259, + "step": 8853 + }, + { + "epoch": 2.014562002275313, + "grad_norm": 1.785006820022791, + "learning_rate": 8.12774236238168e-07, + "loss": 0.0629, + "step": 8854 + }, + { + "epoch": 2.0147895335608648, + "grad_norm": 1.5476841430485058, + "learning_rate": 8.126890109255908e-07, + "loss": 0.0813, + "step": 8855 + }, + { + "epoch": 2.0150170648464165, + "grad_norm": 1.3137594059313582, + "learning_rate": 8.126037817769939e-07, + "loss": 0.0803, + "step": 8856 + }, + { + "epoch": 2.0152445961319683, + "grad_norm": 1.0884237412764168, + "learning_rate": 8.125185487941195e-07, + "loss": 0.0542, + "step": 8857 + }, + { + "epoch": 2.01547212741752, + "grad_norm": 3.13843937228986, + "learning_rate": 8.124333119787093e-07, + "loss": 0.1141, + "step": 8858 + }, + { + "epoch": 2.0156996587030718, + "grad_norm": 1.3093146134482476, + "learning_rate": 8.123480713325053e-07, + "loss": 0.1013, + "step": 8859 + }, + { + "epoch": 2.0159271899886235, + "grad_norm": 0.9028993616744687, + "learning_rate": 8.122628268572499e-07, + "loss": 0.046, + "step": 8860 + }, + { + "epoch": 2.0161547212741753, + "grad_norm": 1.532618420082054, + "learning_rate": 8.121775785546855e-07, + "loss": 0.0421, + "step": 8861 + }, + { + "epoch": 2.016382252559727, + "grad_norm": 1.2710114861205974, + "learning_rate": 8.120923264265539e-07, + "loss": 0.0201, + "step": 8862 + }, + { + "epoch": 2.016609783845279, + "grad_norm": 1.1152451222395094, + "learning_rate": 8.120070704745979e-07, + "loss": 0.0302, + "step": 8863 + }, + { + "epoch": 2.0168373151308305, + "grad_norm": 1.5461921042299296, + "learning_rate": 8.1192181070056e-07, + "loss": 0.1264, + "step": 8864 + }, + { + "epoch": 2.0170648464163823, + "grad_norm": 1.7138961394140304, + "learning_rate": 8.118365471061825e-07, + "loss": 0.0841, + "step": 8865 + }, + { + "epoch": 2.017292377701934, + "grad_norm": 3.0348422801696717, + "learning_rate": 8.117512796932079e-07, + "loss": 0.0509, + "step": 8866 + }, + { + "epoch": 2.017519908987486, + "grad_norm": 1.4907253023289126, + "learning_rate": 8.116660084633796e-07, + "loss": 0.0468, + "step": 8867 + }, + { + "epoch": 2.0177474402730375, + "grad_norm": 1.3836711005142381, + "learning_rate": 8.115807334184398e-07, + "loss": 0.0216, + "step": 8868 + }, + { + "epoch": 2.0179749715585893, + "grad_norm": 1.1496152961442507, + "learning_rate": 8.114954545601314e-07, + "loss": 0.0308, + "step": 8869 + }, + { + "epoch": 2.018202502844141, + "grad_norm": 1.3218587437148022, + "learning_rate": 8.114101718901976e-07, + "loss": 0.04, + "step": 8870 + }, + { + "epoch": 2.018430034129693, + "grad_norm": 1.7977004770876162, + "learning_rate": 8.113248854103811e-07, + "loss": 0.1233, + "step": 8871 + }, + { + "epoch": 2.0186575654152445, + "grad_norm": 1.0030429934118974, + "learning_rate": 8.112395951224254e-07, + "loss": 0.0318, + "step": 8872 + }, + { + "epoch": 2.0188850967007963, + "grad_norm": 1.3374556340738952, + "learning_rate": 8.111543010280733e-07, + "loss": 0.1082, + "step": 8873 + }, + { + "epoch": 2.019112627986348, + "grad_norm": 1.2754626107915659, + "learning_rate": 8.110690031290683e-07, + "loss": 0.0238, + "step": 8874 + }, + { + "epoch": 2.0193401592719, + "grad_norm": 1.1302821237073781, + "learning_rate": 8.109837014271536e-07, + "loss": 0.0518, + "step": 8875 + }, + { + "epoch": 2.0195676905574516, + "grad_norm": 2.3241974039177973, + "learning_rate": 8.108983959240726e-07, + "loss": 0.0477, + "step": 8876 + }, + { + "epoch": 2.0197952218430033, + "grad_norm": 2.012356940370276, + "learning_rate": 8.108130866215689e-07, + "loss": 0.0759, + "step": 8877 + }, + { + "epoch": 2.020022753128555, + "grad_norm": 1.2928306795197322, + "learning_rate": 8.107277735213861e-07, + "loss": 0.0985, + "step": 8878 + }, + { + "epoch": 2.020250284414107, + "grad_norm": 1.616751693476569, + "learning_rate": 8.106424566252675e-07, + "loss": 0.1537, + "step": 8879 + }, + { + "epoch": 2.0204778156996586, + "grad_norm": 1.1982048192000676, + "learning_rate": 8.105571359349575e-07, + "loss": 0.0761, + "step": 8880 + }, + { + "epoch": 2.0207053469852103, + "grad_norm": 1.5225924618859668, + "learning_rate": 8.104718114521993e-07, + "loss": 0.097, + "step": 8881 + }, + { + "epoch": 2.020932878270762, + "grad_norm": 1.5669304640435022, + "learning_rate": 8.103864831787367e-07, + "loss": 0.0511, + "step": 8882 + }, + { + "epoch": 2.021160409556314, + "grad_norm": 1.810778789132058, + "learning_rate": 8.103011511163141e-07, + "loss": 0.0375, + "step": 8883 + }, + { + "epoch": 2.0213879408418656, + "grad_norm": 1.557172516994359, + "learning_rate": 8.102158152666753e-07, + "loss": 0.0917, + "step": 8884 + }, + { + "epoch": 2.0216154721274173, + "grad_norm": 1.3301375818507017, + "learning_rate": 8.101304756315645e-07, + "loss": 0.0776, + "step": 8885 + }, + { + "epoch": 2.021843003412969, + "grad_norm": 1.760767637387474, + "learning_rate": 8.100451322127257e-07, + "loss": 0.0576, + "step": 8886 + }, + { + "epoch": 2.022070534698521, + "grad_norm": 1.265408011042625, + "learning_rate": 8.099597850119035e-07, + "loss": 0.0131, + "step": 8887 + }, + { + "epoch": 2.022298065984073, + "grad_norm": 1.2303519344090885, + "learning_rate": 8.098744340308419e-07, + "loss": 0.0284, + "step": 8888 + }, + { + "epoch": 2.0225255972696248, + "grad_norm": 0.7748711148160067, + "learning_rate": 8.097890792712853e-07, + "loss": 0.0354, + "step": 8889 + }, + { + "epoch": 2.0227531285551765, + "grad_norm": 1.5478258333880945, + "learning_rate": 8.097037207349785e-07, + "loss": 0.1163, + "step": 8890 + }, + { + "epoch": 2.0229806598407283, + "grad_norm": 1.0405193819995862, + "learning_rate": 8.096183584236659e-07, + "loss": 0.1061, + "step": 8891 + }, + { + "epoch": 2.02320819112628, + "grad_norm": 2.1481594666987185, + "learning_rate": 8.095329923390924e-07, + "loss": 0.0545, + "step": 8892 + }, + { + "epoch": 2.023435722411832, + "grad_norm": 1.1859826176533887, + "learning_rate": 8.094476224830022e-07, + "loss": 0.0936, + "step": 8893 + }, + { + "epoch": 2.0236632536973835, + "grad_norm": 3.5992574346212485, + "learning_rate": 8.093622488571405e-07, + "loss": 0.1488, + "step": 8894 + }, + { + "epoch": 2.0238907849829353, + "grad_norm": 1.8619071779147156, + "learning_rate": 8.09276871463252e-07, + "loss": 0.1311, + "step": 8895 + }, + { + "epoch": 2.024118316268487, + "grad_norm": 1.9699566241656805, + "learning_rate": 8.091914903030818e-07, + "loss": 0.0539, + "step": 8896 + }, + { + "epoch": 2.024345847554039, + "grad_norm": 1.4290101274644949, + "learning_rate": 8.091061053783748e-07, + "loss": 0.1257, + "step": 8897 + }, + { + "epoch": 2.0245733788395905, + "grad_norm": 1.63019832829888, + "learning_rate": 8.090207166908763e-07, + "loss": 0.0497, + "step": 8898 + }, + { + "epoch": 2.0248009101251423, + "grad_norm": 1.3568352935794203, + "learning_rate": 8.089353242423313e-07, + "loss": 0.0555, + "step": 8899 + }, + { + "epoch": 2.025028441410694, + "grad_norm": 1.8337008340406296, + "learning_rate": 8.088499280344851e-07, + "loss": 0.0439, + "step": 8900 + }, + { + "epoch": 2.025255972696246, + "grad_norm": 0.8472940461819305, + "learning_rate": 8.087645280690831e-07, + "loss": 0.0156, + "step": 8901 + }, + { + "epoch": 2.0254835039817976, + "grad_norm": 1.5355922509172337, + "learning_rate": 8.086791243478709e-07, + "loss": 0.0257, + "step": 8902 + }, + { + "epoch": 2.0257110352673493, + "grad_norm": 1.8813496487292052, + "learning_rate": 8.085937168725934e-07, + "loss": 0.0322, + "step": 8903 + }, + { + "epoch": 2.025938566552901, + "grad_norm": 1.324514290999542, + "learning_rate": 8.085083056449968e-07, + "loss": 0.0517, + "step": 8904 + }, + { + "epoch": 2.026166097838453, + "grad_norm": 1.123634941418051, + "learning_rate": 8.084228906668267e-07, + "loss": 0.0402, + "step": 8905 + }, + { + "epoch": 2.0263936291240046, + "grad_norm": 1.984050494640443, + "learning_rate": 8.083374719398282e-07, + "loss": 0.0454, + "step": 8906 + }, + { + "epoch": 2.0266211604095563, + "grad_norm": 1.0585652976995195, + "learning_rate": 8.082520494657478e-07, + "loss": 0.1032, + "step": 8907 + }, + { + "epoch": 2.026848691695108, + "grad_norm": 2.0027178523200737, + "learning_rate": 8.08166623246331e-07, + "loss": 0.1118, + "step": 8908 + }, + { + "epoch": 2.02707622298066, + "grad_norm": 1.0927974697343035, + "learning_rate": 8.08081193283324e-07, + "loss": 0.0853, + "step": 8909 + }, + { + "epoch": 2.0273037542662116, + "grad_norm": 1.9775497901190555, + "learning_rate": 8.079957595784727e-07, + "loss": 0.0341, + "step": 8910 + }, + { + "epoch": 2.0275312855517633, + "grad_norm": 0.7191262298610074, + "learning_rate": 8.07910322133523e-07, + "loss": 0.0447, + "step": 8911 + }, + { + "epoch": 2.027758816837315, + "grad_norm": 1.1461280136279222, + "learning_rate": 8.078248809502215e-07, + "loss": 0.0261, + "step": 8912 + }, + { + "epoch": 2.027986348122867, + "grad_norm": 1.4199751658771598, + "learning_rate": 8.077394360303143e-07, + "loss": 0.0391, + "step": 8913 + }, + { + "epoch": 2.0282138794084186, + "grad_norm": 2.4467428915964144, + "learning_rate": 8.076539873755476e-07, + "loss": 0.0847, + "step": 8914 + }, + { + "epoch": 2.0284414106939703, + "grad_norm": 1.2357965996381937, + "learning_rate": 8.07568534987668e-07, + "loss": 0.0474, + "step": 8915 + }, + { + "epoch": 2.028668941979522, + "grad_norm": 1.2427640347071134, + "learning_rate": 8.074830788684218e-07, + "loss": 0.0401, + "step": 8916 + }, + { + "epoch": 2.028896473265074, + "grad_norm": 2.4817322255843157, + "learning_rate": 8.073976190195557e-07, + "loss": 0.069, + "step": 8917 + }, + { + "epoch": 2.0291240045506256, + "grad_norm": 1.9727789886388352, + "learning_rate": 8.073121554428165e-07, + "loss": 0.0611, + "step": 8918 + }, + { + "epoch": 2.0293515358361773, + "grad_norm": 1.3750702848213363, + "learning_rate": 8.072266881399504e-07, + "loss": 0.0209, + "step": 8919 + }, + { + "epoch": 2.029579067121729, + "grad_norm": 2.008802936258141, + "learning_rate": 8.071412171127047e-07, + "loss": 0.0649, + "step": 8920 + }, + { + "epoch": 2.029806598407281, + "grad_norm": 2.1349094497728998, + "learning_rate": 8.070557423628262e-07, + "loss": 0.083, + "step": 8921 + }, + { + "epoch": 2.0300341296928326, + "grad_norm": 0.9866346052842123, + "learning_rate": 8.069702638920615e-07, + "loss": 0.018, + "step": 8922 + }, + { + "epoch": 2.0302616609783843, + "grad_norm": 1.324608361219327, + "learning_rate": 8.06884781702158e-07, + "loss": 0.0222, + "step": 8923 + }, + { + "epoch": 2.030489192263936, + "grad_norm": 0.7576778134924638, + "learning_rate": 8.067992957948628e-07, + "loss": 0.0521, + "step": 8924 + }, + { + "epoch": 2.030716723549488, + "grad_norm": 0.8122934536776792, + "learning_rate": 8.067138061719227e-07, + "loss": 0.019, + "step": 8925 + }, + { + "epoch": 2.03094425483504, + "grad_norm": 1.3291243418602547, + "learning_rate": 8.066283128350854e-07, + "loss": 0.0298, + "step": 8926 + }, + { + "epoch": 2.031171786120592, + "grad_norm": 2.042240688145885, + "learning_rate": 8.065428157860978e-07, + "loss": 0.0386, + "step": 8927 + }, + { + "epoch": 2.0313993174061435, + "grad_norm": 1.640167432849453, + "learning_rate": 8.064573150267077e-07, + "loss": 0.0307, + "step": 8928 + }, + { + "epoch": 2.0316268486916953, + "grad_norm": 1.6054240756394935, + "learning_rate": 8.063718105586624e-07, + "loss": 0.0256, + "step": 8929 + }, + { + "epoch": 2.031854379977247, + "grad_norm": 4.192156659893677, + "learning_rate": 8.062863023837093e-07, + "loss": 0.0681, + "step": 8930 + }, + { + "epoch": 2.032081911262799, + "grad_norm": 1.572181011752906, + "learning_rate": 8.062007905035965e-07, + "loss": 0.0228, + "step": 8931 + }, + { + "epoch": 2.0323094425483506, + "grad_norm": 1.943062464709226, + "learning_rate": 8.061152749200713e-07, + "loss": 0.0923, + "step": 8932 + }, + { + "epoch": 2.0325369738339023, + "grad_norm": 1.5336636120970002, + "learning_rate": 8.060297556348812e-07, + "loss": 0.0834, + "step": 8933 + }, + { + "epoch": 2.032764505119454, + "grad_norm": 2.282119148292395, + "learning_rate": 8.059442326497748e-07, + "loss": 0.0549, + "step": 8934 + }, + { + "epoch": 2.032992036405006, + "grad_norm": 2.624530880383298, + "learning_rate": 8.058587059664996e-07, + "loss": 0.0505, + "step": 8935 + }, + { + "epoch": 2.0332195676905576, + "grad_norm": 1.1785663028124431, + "learning_rate": 8.057731755868036e-07, + "loss": 0.095, + "step": 8936 + }, + { + "epoch": 2.0334470989761093, + "grad_norm": 1.4395348877574188, + "learning_rate": 8.056876415124352e-07, + "loss": 0.0292, + "step": 8937 + }, + { + "epoch": 2.033674630261661, + "grad_norm": 1.2287277310831204, + "learning_rate": 8.056021037451422e-07, + "loss": 0.0334, + "step": 8938 + }, + { + "epoch": 2.033902161547213, + "grad_norm": 1.8255268955888289, + "learning_rate": 8.055165622866726e-07, + "loss": 0.0331, + "step": 8939 + }, + { + "epoch": 2.0341296928327646, + "grad_norm": 1.5879690663346466, + "learning_rate": 8.054310171387756e-07, + "loss": 0.0242, + "step": 8940 + }, + { + "epoch": 2.0343572241183163, + "grad_norm": 2.0235019064610724, + "learning_rate": 8.053454683031987e-07, + "loss": 0.1372, + "step": 8941 + }, + { + "epoch": 2.034584755403868, + "grad_norm": 1.181125885769763, + "learning_rate": 8.052599157816908e-07, + "loss": 0.0227, + "step": 8942 + }, + { + "epoch": 2.03481228668942, + "grad_norm": 3.9382854212462934, + "learning_rate": 8.051743595760005e-07, + "loss": 0.0995, + "step": 8943 + }, + { + "epoch": 2.0350398179749716, + "grad_norm": 2.1906054815912035, + "learning_rate": 8.050887996878761e-07, + "loss": 0.0406, + "step": 8944 + }, + { + "epoch": 2.0352673492605233, + "grad_norm": 1.8522349512738712, + "learning_rate": 8.050032361190666e-07, + "loss": 0.0627, + "step": 8945 + }, + { + "epoch": 2.035494880546075, + "grad_norm": 1.6617262239448711, + "learning_rate": 8.049176688713203e-07, + "loss": 0.0573, + "step": 8946 + }, + { + "epoch": 2.035722411831627, + "grad_norm": 2.872919050675009, + "learning_rate": 8.048320979463867e-07, + "loss": 0.1393, + "step": 8947 + }, + { + "epoch": 2.0359499431171786, + "grad_norm": 1.6577751858636431, + "learning_rate": 8.047465233460141e-07, + "loss": 0.1273, + "step": 8948 + }, + { + "epoch": 2.0361774744027303, + "grad_norm": 0.9839107736775068, + "learning_rate": 8.04660945071952e-07, + "loss": 0.0172, + "step": 8949 + }, + { + "epoch": 2.036405005688282, + "grad_norm": 1.6974546856016852, + "learning_rate": 8.045753631259491e-07, + "loss": 0.0249, + "step": 8950 + }, + { + "epoch": 2.036632536973834, + "grad_norm": 1.2769205319660486, + "learning_rate": 8.044897775097548e-07, + "loss": 0.0307, + "step": 8951 + }, + { + "epoch": 2.0368600682593856, + "grad_norm": 1.2296353160559101, + "learning_rate": 8.04404188225118e-07, + "loss": 0.0148, + "step": 8952 + }, + { + "epoch": 2.0370875995449373, + "grad_norm": 1.4095932818883932, + "learning_rate": 8.043185952737881e-07, + "loss": 0.1281, + "step": 8953 + }, + { + "epoch": 2.037315130830489, + "grad_norm": 1.6892658920866102, + "learning_rate": 8.042329986575145e-07, + "loss": 0.0239, + "step": 8954 + }, + { + "epoch": 2.037542662116041, + "grad_norm": 1.7860108163689632, + "learning_rate": 8.041473983780467e-07, + "loss": 0.0833, + "step": 8955 + }, + { + "epoch": 2.0377701934015926, + "grad_norm": 1.3974190234214017, + "learning_rate": 8.040617944371343e-07, + "loss": 0.0225, + "step": 8956 + }, + { + "epoch": 2.0379977246871444, + "grad_norm": 1.7684063697572843, + "learning_rate": 8.039761868365267e-07, + "loss": 0.0471, + "step": 8957 + }, + { + "epoch": 2.038225255972696, + "grad_norm": 2.016680107128214, + "learning_rate": 8.038905755779737e-07, + "loss": 0.0653, + "step": 8958 + }, + { + "epoch": 2.038452787258248, + "grad_norm": 1.1038807275490525, + "learning_rate": 8.038049606632248e-07, + "loss": 0.0609, + "step": 8959 + }, + { + "epoch": 2.0386803185437996, + "grad_norm": 1.2448753957934051, + "learning_rate": 8.0371934209403e-07, + "loss": 0.0176, + "step": 8960 + }, + { + "epoch": 2.0389078498293514, + "grad_norm": 1.741936926501403, + "learning_rate": 8.036337198721392e-07, + "loss": 0.0457, + "step": 8961 + }, + { + "epoch": 2.039135381114903, + "grad_norm": 1.3952274993028444, + "learning_rate": 8.035480939993025e-07, + "loss": 0.0676, + "step": 8962 + }, + { + "epoch": 2.039362912400455, + "grad_norm": 1.0455820280644919, + "learning_rate": 8.034624644772694e-07, + "loss": 0.0438, + "step": 8963 + }, + { + "epoch": 2.0395904436860066, + "grad_norm": 1.139554219107509, + "learning_rate": 8.033768313077905e-07, + "loss": 0.0235, + "step": 8964 + }, + { + "epoch": 2.039817974971559, + "grad_norm": 1.9335144343525827, + "learning_rate": 8.03291194492616e-07, + "loss": 0.0556, + "step": 8965 + }, + { + "epoch": 2.0400455062571106, + "grad_norm": 1.1042452402293887, + "learning_rate": 8.032055540334958e-07, + "loss": 0.0378, + "step": 8966 + }, + { + "epoch": 2.0402730375426623, + "grad_norm": 2.009877223396931, + "learning_rate": 8.031199099321804e-07, + "loss": 0.0416, + "step": 8967 + }, + { + "epoch": 2.040500568828214, + "grad_norm": 1.848101717007317, + "learning_rate": 8.030342621904205e-07, + "loss": 0.125, + "step": 8968 + }, + { + "epoch": 2.040728100113766, + "grad_norm": 1.5200120739584286, + "learning_rate": 8.029486108099663e-07, + "loss": 0.0732, + "step": 8969 + }, + { + "epoch": 2.0409556313993176, + "grad_norm": 0.9588206533321003, + "learning_rate": 8.028629557925683e-07, + "loss": 0.0376, + "step": 8970 + }, + { + "epoch": 2.0411831626848693, + "grad_norm": 1.8713116880826133, + "learning_rate": 8.027772971399773e-07, + "loss": 0.0575, + "step": 8971 + }, + { + "epoch": 2.041410693970421, + "grad_norm": 1.4634893162573441, + "learning_rate": 8.026916348539438e-07, + "loss": 0.0571, + "step": 8972 + }, + { + "epoch": 2.041638225255973, + "grad_norm": 2.215319961433251, + "learning_rate": 8.026059689362186e-07, + "loss": 0.1205, + "step": 8973 + }, + { + "epoch": 2.0418657565415246, + "grad_norm": 0.8852613105021321, + "learning_rate": 8.025202993885528e-07, + "loss": 0.0347, + "step": 8974 + }, + { + "epoch": 2.0420932878270763, + "grad_norm": 2.0572452023597747, + "learning_rate": 8.024346262126976e-07, + "loss": 0.0445, + "step": 8975 + }, + { + "epoch": 2.042320819112628, + "grad_norm": 0.8397897258684369, + "learning_rate": 8.02348949410403e-07, + "loss": 0.0304, + "step": 8976 + }, + { + "epoch": 2.04254835039818, + "grad_norm": 1.60407005281846, + "learning_rate": 8.022632689834209e-07, + "loss": 0.0969, + "step": 8977 + }, + { + "epoch": 2.0427758816837316, + "grad_norm": 0.9333464874995563, + "learning_rate": 8.021775849335023e-07, + "loss": 0.0318, + "step": 8978 + }, + { + "epoch": 2.0430034129692833, + "grad_norm": 2.263807067343164, + "learning_rate": 8.020918972623983e-07, + "loss": 0.0944, + "step": 8979 + }, + { + "epoch": 2.043230944254835, + "grad_norm": 1.9712926683450096, + "learning_rate": 8.0200620597186e-07, + "loss": 0.0317, + "step": 8980 + }, + { + "epoch": 2.043458475540387, + "grad_norm": 1.6767964098327561, + "learning_rate": 8.019205110636394e-07, + "loss": 0.0765, + "step": 8981 + }, + { + "epoch": 2.0436860068259386, + "grad_norm": 1.3875052253841036, + "learning_rate": 8.018348125394874e-07, + "loss": 0.1099, + "step": 8982 + }, + { + "epoch": 2.0439135381114903, + "grad_norm": 2.3749906044977758, + "learning_rate": 8.017491104011557e-07, + "loss": 0.044, + "step": 8983 + }, + { + "epoch": 2.044141069397042, + "grad_norm": 1.7636354504511285, + "learning_rate": 8.016634046503958e-07, + "loss": 0.0402, + "step": 8984 + }, + { + "epoch": 2.044368600682594, + "grad_norm": 1.4913896041513928, + "learning_rate": 8.015776952889596e-07, + "loss": 0.0517, + "step": 8985 + }, + { + "epoch": 2.0445961319681456, + "grad_norm": 2.511436444195971, + "learning_rate": 8.014919823185987e-07, + "loss": 0.101, + "step": 8986 + }, + { + "epoch": 2.0448236632536974, + "grad_norm": 0.9461858874889127, + "learning_rate": 8.014062657410648e-07, + "loss": 0.0146, + "step": 8987 + }, + { + "epoch": 2.045051194539249, + "grad_norm": 1.6882002541921113, + "learning_rate": 8.0132054555811e-07, + "loss": 0.0279, + "step": 8988 + }, + { + "epoch": 2.045278725824801, + "grad_norm": 2.164694201901671, + "learning_rate": 8.012348217714861e-07, + "loss": 0.0687, + "step": 8989 + }, + { + "epoch": 2.0455062571103526, + "grad_norm": 0.8913950871887149, + "learning_rate": 8.011490943829451e-07, + "loss": 0.0545, + "step": 8990 + }, + { + "epoch": 2.0457337883959044, + "grad_norm": 1.7822662572491148, + "learning_rate": 8.010633633942394e-07, + "loss": 0.082, + "step": 8991 + }, + { + "epoch": 2.045961319681456, + "grad_norm": 1.6849998182193533, + "learning_rate": 8.00977628807121e-07, + "loss": 0.1276, + "step": 8992 + }, + { + "epoch": 2.046188850967008, + "grad_norm": 0.9499780954546638, + "learning_rate": 8.008918906233421e-07, + "loss": 0.0252, + "step": 8993 + }, + { + "epoch": 2.0464163822525596, + "grad_norm": 1.1148052194865308, + "learning_rate": 8.00806148844655e-07, + "loss": 0.0184, + "step": 8994 + }, + { + "epoch": 2.0466439135381114, + "grad_norm": 1.4157448520587377, + "learning_rate": 8.007204034728123e-07, + "loss": 0.0119, + "step": 8995 + }, + { + "epoch": 2.046871444823663, + "grad_norm": 1.8161237313706846, + "learning_rate": 8.006346545095664e-07, + "loss": 0.0361, + "step": 8996 + }, + { + "epoch": 2.047098976109215, + "grad_norm": 1.6145593767095079, + "learning_rate": 8.005489019566697e-07, + "loss": 0.0853, + "step": 8997 + }, + { + "epoch": 2.0473265073947666, + "grad_norm": 1.5070509103543634, + "learning_rate": 8.004631458158749e-07, + "loss": 0.0219, + "step": 8998 + }, + { + "epoch": 2.0475540386803184, + "grad_norm": 1.5212533805532744, + "learning_rate": 8.00377386088935e-07, + "loss": 0.0308, + "step": 8999 + }, + { + "epoch": 2.04778156996587, + "grad_norm": 1.7683804948094972, + "learning_rate": 8.002916227776023e-07, + "loss": 0.061, + "step": 9000 + }, + { + "epoch": 2.048009101251422, + "grad_norm": 1.1676965297854693, + "learning_rate": 8.002058558836298e-07, + "loss": 0.0179, + "step": 9001 + }, + { + "epoch": 2.0482366325369736, + "grad_norm": 1.1725218533720503, + "learning_rate": 8.001200854087707e-07, + "loss": 0.0619, + "step": 9002 + }, + { + "epoch": 2.0484641638225254, + "grad_norm": 1.5277769596056463, + "learning_rate": 8.000343113547777e-07, + "loss": 0.1207, + "step": 9003 + }, + { + "epoch": 2.0486916951080776, + "grad_norm": 1.2876329386633896, + "learning_rate": 7.999485337234038e-07, + "loss": 0.129, + "step": 9004 + }, + { + "epoch": 2.0489192263936293, + "grad_norm": 1.4316387359992495, + "learning_rate": 7.998627525164024e-07, + "loss": 0.0308, + "step": 9005 + }, + { + "epoch": 2.049146757679181, + "grad_norm": 2.0391400130019366, + "learning_rate": 7.997769677355266e-07, + "loss": 0.146, + "step": 9006 + }, + { + "epoch": 2.049374288964733, + "grad_norm": 2.5254020354583044, + "learning_rate": 7.996911793825296e-07, + "loss": 0.0958, + "step": 9007 + }, + { + "epoch": 2.0496018202502846, + "grad_norm": 1.859705882278126, + "learning_rate": 7.996053874591649e-07, + "loss": 0.034, + "step": 9008 + }, + { + "epoch": 2.0498293515358363, + "grad_norm": 1.4013398825771965, + "learning_rate": 7.995195919671858e-07, + "loss": 0.0355, + "step": 9009 + }, + { + "epoch": 2.050056882821388, + "grad_norm": 1.8066008322109752, + "learning_rate": 7.994337929083458e-07, + "loss": 0.0322, + "step": 9010 + }, + { + "epoch": 2.05028441410694, + "grad_norm": 1.3387518936152152, + "learning_rate": 7.993479902843987e-07, + "loss": 0.0192, + "step": 9011 + }, + { + "epoch": 2.0505119453924916, + "grad_norm": 1.3836163125157097, + "learning_rate": 7.99262184097098e-07, + "loss": 0.0723, + "step": 9012 + }, + { + "epoch": 2.0507394766780433, + "grad_norm": 2.182460664922858, + "learning_rate": 7.991763743481971e-07, + "loss": 0.0427, + "step": 9013 + }, + { + "epoch": 2.050967007963595, + "grad_norm": 3.715086981999057, + "learning_rate": 7.990905610394503e-07, + "loss": 0.0436, + "step": 9014 + }, + { + "epoch": 2.051194539249147, + "grad_norm": 1.3497563155973114, + "learning_rate": 7.990047441726114e-07, + "loss": 0.1029, + "step": 9015 + }, + { + "epoch": 2.0514220705346986, + "grad_norm": 1.8241562674210596, + "learning_rate": 7.989189237494339e-07, + "loss": 0.1718, + "step": 9016 + }, + { + "epoch": 2.0516496018202504, + "grad_norm": 2.1318260927651576, + "learning_rate": 7.988330997716723e-07, + "loss": 0.0298, + "step": 9017 + }, + { + "epoch": 2.051877133105802, + "grad_norm": 1.625880671121143, + "learning_rate": 7.987472722410805e-07, + "loss": 0.1317, + "step": 9018 + }, + { + "epoch": 2.052104664391354, + "grad_norm": 3.0419852460456105, + "learning_rate": 7.986614411594126e-07, + "loss": 0.0902, + "step": 9019 + }, + { + "epoch": 2.0523321956769056, + "grad_norm": 1.1566723855079115, + "learning_rate": 7.98575606528423e-07, + "loss": 0.0287, + "step": 9020 + }, + { + "epoch": 2.0525597269624574, + "grad_norm": 2.173986589113051, + "learning_rate": 7.984897683498658e-07, + "loss": 0.1862, + "step": 9021 + }, + { + "epoch": 2.052787258248009, + "grad_norm": 2.657098105490752, + "learning_rate": 7.984039266254955e-07, + "loss": 0.0605, + "step": 9022 + }, + { + "epoch": 2.053014789533561, + "grad_norm": 1.3439299079894502, + "learning_rate": 7.983180813570665e-07, + "loss": 0.0776, + "step": 9023 + }, + { + "epoch": 2.0532423208191126, + "grad_norm": 1.8443990231321319, + "learning_rate": 7.982322325463331e-07, + "loss": 0.0335, + "step": 9024 + }, + { + "epoch": 2.0534698521046644, + "grad_norm": 1.6303694233983472, + "learning_rate": 7.981463801950507e-07, + "loss": 0.0379, + "step": 9025 + }, + { + "epoch": 2.053697383390216, + "grad_norm": 1.7700467233785628, + "learning_rate": 7.98060524304973e-07, + "loss": 0.0474, + "step": 9026 + }, + { + "epoch": 2.053924914675768, + "grad_norm": 1.7842786850819012, + "learning_rate": 7.97974664877855e-07, + "loss": 0.0705, + "step": 9027 + }, + { + "epoch": 2.0541524459613196, + "grad_norm": 1.608702386328513, + "learning_rate": 7.978888019154518e-07, + "loss": 0.0466, + "step": 9028 + }, + { + "epoch": 2.0543799772468714, + "grad_norm": 1.2148934424447984, + "learning_rate": 7.978029354195182e-07, + "loss": 0.0286, + "step": 9029 + }, + { + "epoch": 2.054607508532423, + "grad_norm": 1.9444035598114653, + "learning_rate": 7.977170653918088e-07, + "loss": 0.1644, + "step": 9030 + }, + { + "epoch": 2.054835039817975, + "grad_norm": 1.1480733728097936, + "learning_rate": 7.976311918340792e-07, + "loss": 0.0166, + "step": 9031 + }, + { + "epoch": 2.0550625711035266, + "grad_norm": 1.6450968119134726, + "learning_rate": 7.975453147480843e-07, + "loss": 0.0259, + "step": 9032 + }, + { + "epoch": 2.0552901023890784, + "grad_norm": 1.8067992560101058, + "learning_rate": 7.974594341355787e-07, + "loss": 0.0429, + "step": 9033 + }, + { + "epoch": 2.05551763367463, + "grad_norm": 1.7079755352057893, + "learning_rate": 7.973735499983185e-07, + "loss": 0.0405, + "step": 9034 + }, + { + "epoch": 2.055745164960182, + "grad_norm": 1.494102609790233, + "learning_rate": 7.972876623380585e-07, + "loss": 0.067, + "step": 9035 + }, + { + "epoch": 2.0559726962457336, + "grad_norm": 1.0921037707533194, + "learning_rate": 7.972017711565543e-07, + "loss": 0.082, + "step": 9036 + }, + { + "epoch": 2.0562002275312854, + "grad_norm": 2.564234211322546, + "learning_rate": 7.971158764555611e-07, + "loss": 0.0424, + "step": 9037 + }, + { + "epoch": 2.056427758816837, + "grad_norm": 1.3145382795708485, + "learning_rate": 7.970299782368347e-07, + "loss": 0.0835, + "step": 9038 + }, + { + "epoch": 2.056655290102389, + "grad_norm": 2.060414155748251, + "learning_rate": 7.969440765021306e-07, + "loss": 0.0792, + "step": 9039 + }, + { + "epoch": 2.0568828213879407, + "grad_norm": 0.9272262973137503, + "learning_rate": 7.968581712532044e-07, + "loss": 0.0751, + "step": 9040 + }, + { + "epoch": 2.0571103526734924, + "grad_norm": 0.9430514246449326, + "learning_rate": 7.96772262491812e-07, + "loss": 0.0422, + "step": 9041 + }, + { + "epoch": 2.057337883959044, + "grad_norm": 1.817014778048195, + "learning_rate": 7.966863502197092e-07, + "loss": 0.0684, + "step": 9042 + }, + { + "epoch": 2.0575654152445964, + "grad_norm": 1.372358221318876, + "learning_rate": 7.966004344386517e-07, + "loss": 0.0597, + "step": 9043 + }, + { + "epoch": 2.057792946530148, + "grad_norm": 2.5726795542399232, + "learning_rate": 7.965145151503957e-07, + "loss": 0.1362, + "step": 9044 + }, + { + "epoch": 2.0580204778157, + "grad_norm": 2.0362358762589654, + "learning_rate": 7.964285923566971e-07, + "loss": 0.0824, + "step": 9045 + }, + { + "epoch": 2.0582480091012516, + "grad_norm": 1.3864717525542436, + "learning_rate": 7.963426660593121e-07, + "loss": 0.0541, + "step": 9046 + }, + { + "epoch": 2.0584755403868034, + "grad_norm": 1.2558444766460526, + "learning_rate": 7.962567362599965e-07, + "loss": 0.0847, + "step": 9047 + }, + { + "epoch": 2.058703071672355, + "grad_norm": 1.9079873952492432, + "learning_rate": 7.961708029605072e-07, + "loss": 0.0994, + "step": 9048 + }, + { + "epoch": 2.058930602957907, + "grad_norm": 2.486622958190756, + "learning_rate": 7.960848661626e-07, + "loss": 0.0611, + "step": 9049 + }, + { + "epoch": 2.0591581342434586, + "grad_norm": 1.786473375243381, + "learning_rate": 7.959989258680314e-07, + "loss": 0.1021, + "step": 9050 + }, + { + "epoch": 2.0593856655290104, + "grad_norm": 1.2647734446477317, + "learning_rate": 7.959129820785581e-07, + "loss": 0.0246, + "step": 9051 + }, + { + "epoch": 2.059613196814562, + "grad_norm": 2.055755652392172, + "learning_rate": 7.958270347959365e-07, + "loss": 0.0318, + "step": 9052 + }, + { + "epoch": 2.059840728100114, + "grad_norm": 1.7331547478611729, + "learning_rate": 7.95741084021923e-07, + "loss": 0.0504, + "step": 9053 + }, + { + "epoch": 2.0600682593856656, + "grad_norm": 2.0810800670078553, + "learning_rate": 7.956551297582744e-07, + "loss": 0.1526, + "step": 9054 + }, + { + "epoch": 2.0602957906712174, + "grad_norm": 1.2473878307051924, + "learning_rate": 7.955691720067476e-07, + "loss": 0.068, + "step": 9055 + }, + { + "epoch": 2.060523321956769, + "grad_norm": 1.1892730990813287, + "learning_rate": 7.954832107690994e-07, + "loss": 0.0193, + "step": 9056 + }, + { + "epoch": 2.060750853242321, + "grad_norm": 2.625497516100211, + "learning_rate": 7.953972460470865e-07, + "loss": 0.081, + "step": 9057 + }, + { + "epoch": 2.0609783845278726, + "grad_norm": 1.0465781093855349, + "learning_rate": 7.953112778424658e-07, + "loss": 0.0323, + "step": 9058 + }, + { + "epoch": 2.0612059158134244, + "grad_norm": 1.9502856330253195, + "learning_rate": 7.952253061569946e-07, + "loss": 0.0778, + "step": 9059 + }, + { + "epoch": 2.061433447098976, + "grad_norm": 1.2985376747533437, + "learning_rate": 7.951393309924299e-07, + "loss": 0.0355, + "step": 9060 + }, + { + "epoch": 2.061660978384528, + "grad_norm": 2.0180691734194545, + "learning_rate": 7.950533523505288e-07, + "loss": 0.0283, + "step": 9061 + }, + { + "epoch": 2.0618885096700796, + "grad_norm": 1.3043138217729549, + "learning_rate": 7.949673702330487e-07, + "loss": 0.0212, + "step": 9062 + }, + { + "epoch": 2.0621160409556314, + "grad_norm": 2.2842920182686943, + "learning_rate": 7.948813846417469e-07, + "loss": 0.0585, + "step": 9063 + }, + { + "epoch": 2.062343572241183, + "grad_norm": 1.0500746839458064, + "learning_rate": 7.947953955783808e-07, + "loss": 0.0598, + "step": 9064 + }, + { + "epoch": 2.062571103526735, + "grad_norm": 1.3277227542346568, + "learning_rate": 7.947094030447077e-07, + "loss": 0.0845, + "step": 9065 + }, + { + "epoch": 2.0627986348122866, + "grad_norm": 1.253186260522492, + "learning_rate": 7.946234070424852e-07, + "loss": 0.0546, + "step": 9066 + }, + { + "epoch": 2.0630261660978384, + "grad_norm": 3.1941381013094934, + "learning_rate": 7.945374075734706e-07, + "loss": 0.0531, + "step": 9067 + }, + { + "epoch": 2.06325369738339, + "grad_norm": 2.170365099074321, + "learning_rate": 7.944514046394222e-07, + "loss": 0.0558, + "step": 9068 + }, + { + "epoch": 2.063481228668942, + "grad_norm": 1.267710151283297, + "learning_rate": 7.943653982420976e-07, + "loss": 0.1233, + "step": 9069 + }, + { + "epoch": 2.0637087599544937, + "grad_norm": 2.934322988147777, + "learning_rate": 7.942793883832541e-07, + "loss": 0.0118, + "step": 9070 + }, + { + "epoch": 2.0639362912400454, + "grad_norm": 1.7674927965665588, + "learning_rate": 7.9419337506465e-07, + "loss": 0.037, + "step": 9071 + }, + { + "epoch": 2.064163822525597, + "grad_norm": 1.5262962717343906, + "learning_rate": 7.941073582880431e-07, + "loss": 0.0296, + "step": 9072 + }, + { + "epoch": 2.064391353811149, + "grad_norm": 2.4846929499037236, + "learning_rate": 7.940213380551918e-07, + "loss": 0.0798, + "step": 9073 + }, + { + "epoch": 2.0646188850967007, + "grad_norm": 1.7093651494630786, + "learning_rate": 7.939353143678535e-07, + "loss": 0.068, + "step": 9074 + }, + { + "epoch": 2.0648464163822524, + "grad_norm": 0.947770878057256, + "learning_rate": 7.938492872277872e-07, + "loss": 0.0332, + "step": 9075 + }, + { + "epoch": 2.065073947667804, + "grad_norm": 3.037640012456654, + "learning_rate": 7.937632566367505e-07, + "loss": 0.1178, + "step": 9076 + }, + { + "epoch": 2.065301478953356, + "grad_norm": 1.4987170595019004, + "learning_rate": 7.936772225965018e-07, + "loss": 0.0143, + "step": 9077 + }, + { + "epoch": 2.0655290102389077, + "grad_norm": 1.2376842513542412, + "learning_rate": 7.935911851087996e-07, + "loss": 0.0199, + "step": 9078 + }, + { + "epoch": 2.0657565415244594, + "grad_norm": 1.6224691256067159, + "learning_rate": 7.935051441754024e-07, + "loss": 0.0436, + "step": 9079 + }, + { + "epoch": 2.065984072810011, + "grad_norm": 2.088654020758353, + "learning_rate": 7.934190997980687e-07, + "loss": 0.1661, + "step": 9080 + }, + { + "epoch": 2.066211604095563, + "grad_norm": 1.42065406988388, + "learning_rate": 7.933330519785569e-07, + "loss": 0.0226, + "step": 9081 + }, + { + "epoch": 2.066439135381115, + "grad_norm": 1.3806344487018214, + "learning_rate": 7.93247000718626e-07, + "loss": 0.0488, + "step": 9082 + }, + { + "epoch": 2.066666666666667, + "grad_norm": 1.063319184761991, + "learning_rate": 7.931609460200345e-07, + "loss": 0.0771, + "step": 9083 + }, + { + "epoch": 2.0668941979522186, + "grad_norm": 1.0127117965184074, + "learning_rate": 7.930748878845411e-07, + "loss": 0.0173, + "step": 9084 + }, + { + "epoch": 2.0671217292377704, + "grad_norm": 1.4019625398505184, + "learning_rate": 7.929888263139049e-07, + "loss": 0.0336, + "step": 9085 + }, + { + "epoch": 2.067349260523322, + "grad_norm": 1.1131020167883046, + "learning_rate": 7.929027613098848e-07, + "loss": 0.0456, + "step": 9086 + }, + { + "epoch": 2.067576791808874, + "grad_norm": 1.733949004627717, + "learning_rate": 7.928166928742398e-07, + "loss": 0.1259, + "step": 9087 + }, + { + "epoch": 2.0678043230944256, + "grad_norm": 1.043896409482535, + "learning_rate": 7.927306210087287e-07, + "loss": 0.0583, + "step": 9088 + }, + { + "epoch": 2.0680318543799774, + "grad_norm": 1.1197448277741158, + "learning_rate": 7.926445457151111e-07, + "loss": 0.0828, + "step": 9089 + }, + { + "epoch": 2.068259385665529, + "grad_norm": 1.9678162359488363, + "learning_rate": 7.925584669951459e-07, + "loss": 0.0623, + "step": 9090 + }, + { + "epoch": 2.068486916951081, + "grad_norm": 1.5132352396941313, + "learning_rate": 7.924723848505925e-07, + "loss": 0.0237, + "step": 9091 + }, + { + "epoch": 2.0687144482366326, + "grad_norm": 1.646682794518182, + "learning_rate": 7.923862992832103e-07, + "loss": 0.1909, + "step": 9092 + }, + { + "epoch": 2.0689419795221844, + "grad_norm": 2.3077472062054585, + "learning_rate": 7.923002102947587e-07, + "loss": 0.0745, + "step": 9093 + }, + { + "epoch": 2.069169510807736, + "grad_norm": 0.9780808903063914, + "learning_rate": 7.922141178869973e-07, + "loss": 0.0152, + "step": 9094 + }, + { + "epoch": 2.069397042093288, + "grad_norm": 1.4315578858964255, + "learning_rate": 7.921280220616855e-07, + "loss": 0.1481, + "step": 9095 + }, + { + "epoch": 2.0696245733788396, + "grad_norm": 1.2423627879635764, + "learning_rate": 7.920419228205829e-07, + "loss": 0.1512, + "step": 9096 + }, + { + "epoch": 2.0698521046643914, + "grad_norm": 2.285717832479959, + "learning_rate": 7.919558201654493e-07, + "loss": 0.0467, + "step": 9097 + }, + { + "epoch": 2.070079635949943, + "grad_norm": 1.7444028206997109, + "learning_rate": 7.918697140980447e-07, + "loss": 0.0286, + "step": 9098 + }, + { + "epoch": 2.070307167235495, + "grad_norm": 1.9074989101870283, + "learning_rate": 7.917836046201285e-07, + "loss": 0.0639, + "step": 9099 + }, + { + "epoch": 2.0705346985210467, + "grad_norm": 1.3169246364354064, + "learning_rate": 7.916974917334612e-07, + "loss": 0.0769, + "step": 9100 + }, + { + "epoch": 2.0707622298065984, + "grad_norm": 1.8006086624089455, + "learning_rate": 7.916113754398022e-07, + "loss": 0.1236, + "step": 9101 + }, + { + "epoch": 2.07098976109215, + "grad_norm": 1.1154065737416012, + "learning_rate": 7.91525255740912e-07, + "loss": 0.0396, + "step": 9102 + }, + { + "epoch": 2.071217292377702, + "grad_norm": 1.6098922197697982, + "learning_rate": 7.914391326385504e-07, + "loss": 0.1823, + "step": 9103 + }, + { + "epoch": 2.0714448236632537, + "grad_norm": 1.7142868450675366, + "learning_rate": 7.913530061344778e-07, + "loss": 0.043, + "step": 9104 + }, + { + "epoch": 2.0716723549488054, + "grad_norm": 1.5874061597081095, + "learning_rate": 7.912668762304544e-07, + "loss": 0.0792, + "step": 9105 + }, + { + "epoch": 2.071899886234357, + "grad_norm": 1.509108721934602, + "learning_rate": 7.911807429282406e-07, + "loss": 0.0848, + "step": 9106 + }, + { + "epoch": 2.072127417519909, + "grad_norm": 2.749770216447884, + "learning_rate": 7.910946062295967e-07, + "loss": 0.0584, + "step": 9107 + }, + { + "epoch": 2.0723549488054607, + "grad_norm": 0.7878630269766755, + "learning_rate": 7.910084661362832e-07, + "loss": 0.0688, + "step": 9108 + }, + { + "epoch": 2.0725824800910124, + "grad_norm": 1.2346259917745928, + "learning_rate": 7.909223226500609e-07, + "loss": 0.0327, + "step": 9109 + }, + { + "epoch": 2.072810011376564, + "grad_norm": 1.3686367464991218, + "learning_rate": 7.908361757726901e-07, + "loss": 0.0668, + "step": 9110 + }, + { + "epoch": 2.073037542662116, + "grad_norm": 1.8352613867965457, + "learning_rate": 7.907500255059314e-07, + "loss": 0.0672, + "step": 9111 + }, + { + "epoch": 2.0732650739476677, + "grad_norm": 2.697406308758313, + "learning_rate": 7.906638718515458e-07, + "loss": 0.0489, + "step": 9112 + }, + { + "epoch": 2.0734926052332194, + "grad_norm": 1.098914714863621, + "learning_rate": 7.905777148112943e-07, + "loss": 0.031, + "step": 9113 + }, + { + "epoch": 2.073720136518771, + "grad_norm": 3.2621670369013933, + "learning_rate": 7.904915543869372e-07, + "loss": 0.0521, + "step": 9114 + }, + { + "epoch": 2.073947667804323, + "grad_norm": 2.479039575608758, + "learning_rate": 7.904053905802362e-07, + "loss": 0.0878, + "step": 9115 + }, + { + "epoch": 2.0741751990898747, + "grad_norm": 1.418821664800718, + "learning_rate": 7.903192233929515e-07, + "loss": 0.0567, + "step": 9116 + }, + { + "epoch": 2.0744027303754264, + "grad_norm": 1.9336354771265998, + "learning_rate": 7.90233052826845e-07, + "loss": 0.0442, + "step": 9117 + }, + { + "epoch": 2.074630261660978, + "grad_norm": 1.637993635401336, + "learning_rate": 7.901468788836771e-07, + "loss": 0.1124, + "step": 9118 + }, + { + "epoch": 2.07485779294653, + "grad_norm": 1.1804507773489084, + "learning_rate": 7.900607015652101e-07, + "loss": 0.0512, + "step": 9119 + }, + { + "epoch": 2.0750853242320817, + "grad_norm": 1.5299470508494952, + "learning_rate": 7.899745208732043e-07, + "loss": 0.0354, + "step": 9120 + }, + { + "epoch": 2.075312855517634, + "grad_norm": 2.053834900698496, + "learning_rate": 7.898883368094213e-07, + "loss": 0.0438, + "step": 9121 + }, + { + "epoch": 2.0755403868031856, + "grad_norm": 1.6512024163216108, + "learning_rate": 7.898021493756228e-07, + "loss": 0.1283, + "step": 9122 + }, + { + "epoch": 2.0757679180887374, + "grad_norm": 1.2485011334703044, + "learning_rate": 7.897159585735702e-07, + "loss": 0.0559, + "step": 9123 + }, + { + "epoch": 2.075995449374289, + "grad_norm": 1.718240711541217, + "learning_rate": 7.896297644050249e-07, + "loss": 0.0463, + "step": 9124 + }, + { + "epoch": 2.076222980659841, + "grad_norm": 1.9656083707282161, + "learning_rate": 7.895435668717488e-07, + "loss": 0.0604, + "step": 9125 + }, + { + "epoch": 2.0764505119453927, + "grad_norm": 2.0888372710238703, + "learning_rate": 7.894573659755038e-07, + "loss": 0.0895, + "step": 9126 + }, + { + "epoch": 2.0766780432309444, + "grad_norm": 1.8353650592615016, + "learning_rate": 7.89371161718051e-07, + "loss": 0.0522, + "step": 9127 + }, + { + "epoch": 2.076905574516496, + "grad_norm": 2.5236097562873403, + "learning_rate": 7.892849541011531e-07, + "loss": 0.0367, + "step": 9128 + }, + { + "epoch": 2.077133105802048, + "grad_norm": 1.1616657344858807, + "learning_rate": 7.891987431265714e-07, + "loss": 0.0174, + "step": 9129 + }, + { + "epoch": 2.0773606370875997, + "grad_norm": 1.578144325050575, + "learning_rate": 7.891125287960682e-07, + "loss": 0.0812, + "step": 9130 + }, + { + "epoch": 2.0775881683731514, + "grad_norm": 1.1517242687449327, + "learning_rate": 7.890263111114052e-07, + "loss": 0.0452, + "step": 9131 + }, + { + "epoch": 2.077815699658703, + "grad_norm": 1.8810114460736644, + "learning_rate": 7.889400900743452e-07, + "loss": 0.0528, + "step": 9132 + }, + { + "epoch": 2.078043230944255, + "grad_norm": 2.0473626545484116, + "learning_rate": 7.888538656866498e-07, + "loss": 0.0368, + "step": 9133 + }, + { + "epoch": 2.0782707622298067, + "grad_norm": 1.8958294920231127, + "learning_rate": 7.887676379500814e-07, + "loss": 0.1566, + "step": 9134 + }, + { + "epoch": 2.0784982935153584, + "grad_norm": 2.387957751750927, + "learning_rate": 7.886814068664025e-07, + "loss": 0.1222, + "step": 9135 + }, + { + "epoch": 2.07872582480091, + "grad_norm": 1.7538970479706861, + "learning_rate": 7.885951724373754e-07, + "loss": 0.1517, + "step": 9136 + }, + { + "epoch": 2.078953356086462, + "grad_norm": 2.3686037145553342, + "learning_rate": 7.885089346647625e-07, + "loss": 0.153, + "step": 9137 + }, + { + "epoch": 2.0791808873720137, + "grad_norm": 2.0979094686177095, + "learning_rate": 7.884226935503267e-07, + "loss": 0.0604, + "step": 9138 + }, + { + "epoch": 2.0794084186575654, + "grad_norm": 1.3953638862022424, + "learning_rate": 7.8833644909583e-07, + "loss": 0.0337, + "step": 9139 + }, + { + "epoch": 2.079635949943117, + "grad_norm": 1.179189164442287, + "learning_rate": 7.882502013030358e-07, + "loss": 0.0312, + "step": 9140 + }, + { + "epoch": 2.079863481228669, + "grad_norm": 1.182582252605836, + "learning_rate": 7.881639501737059e-07, + "loss": 0.0649, + "step": 9141 + }, + { + "epoch": 2.0800910125142207, + "grad_norm": 0.6775958909095499, + "learning_rate": 7.880776957096041e-07, + "loss": 0.0095, + "step": 9142 + }, + { + "epoch": 2.0803185437997724, + "grad_norm": 1.2891614755595928, + "learning_rate": 7.879914379124928e-07, + "loss": 0.0367, + "step": 9143 + }, + { + "epoch": 2.080546075085324, + "grad_norm": 1.4798323666152837, + "learning_rate": 7.879051767841351e-07, + "loss": 0.0229, + "step": 9144 + }, + { + "epoch": 2.080773606370876, + "grad_norm": 1.575524970401019, + "learning_rate": 7.878189123262937e-07, + "loss": 0.059, + "step": 9145 + }, + { + "epoch": 2.0810011376564277, + "grad_norm": 1.0135897054386669, + "learning_rate": 7.877326445407321e-07, + "loss": 0.0563, + "step": 9146 + }, + { + "epoch": 2.0812286689419794, + "grad_norm": 2.891970697742299, + "learning_rate": 7.876463734292133e-07, + "loss": 0.0909, + "step": 9147 + }, + { + "epoch": 2.081456200227531, + "grad_norm": 1.3653078247569157, + "learning_rate": 7.875600989935004e-07, + "loss": 0.0474, + "step": 9148 + }, + { + "epoch": 2.081683731513083, + "grad_norm": 2.5159777114813062, + "learning_rate": 7.874738212353567e-07, + "loss": 0.047, + "step": 9149 + }, + { + "epoch": 2.0819112627986347, + "grad_norm": 1.4073584073202479, + "learning_rate": 7.873875401565458e-07, + "loss": 0.0439, + "step": 9150 + }, + { + "epoch": 2.0821387940841865, + "grad_norm": 1.1811930490134919, + "learning_rate": 7.873012557588309e-07, + "loss": 0.036, + "step": 9151 + }, + { + "epoch": 2.082366325369738, + "grad_norm": 1.4703966938214825, + "learning_rate": 7.872149680439756e-07, + "loss": 0.1374, + "step": 9152 + }, + { + "epoch": 2.08259385665529, + "grad_norm": 1.5401455211601165, + "learning_rate": 7.871286770137434e-07, + "loss": 0.0349, + "step": 9153 + }, + { + "epoch": 2.0828213879408417, + "grad_norm": 2.0287849754461793, + "learning_rate": 7.870423826698981e-07, + "loss": 0.0689, + "step": 9154 + }, + { + "epoch": 2.0830489192263935, + "grad_norm": 1.2804858837836415, + "learning_rate": 7.869560850142031e-07, + "loss": 0.0316, + "step": 9155 + }, + { + "epoch": 2.083276450511945, + "grad_norm": 2.730136139700032, + "learning_rate": 7.868697840484225e-07, + "loss": 0.0411, + "step": 9156 + }, + { + "epoch": 2.083503981797497, + "grad_norm": 1.4714407773423903, + "learning_rate": 7.867834797743199e-07, + "loss": 0.0537, + "step": 9157 + }, + { + "epoch": 2.0837315130830487, + "grad_norm": 2.2903325365713587, + "learning_rate": 7.866971721936592e-07, + "loss": 0.0266, + "step": 9158 + }, + { + "epoch": 2.0839590443686005, + "grad_norm": 2.387960948734622, + "learning_rate": 7.866108613082045e-07, + "loss": 0.0897, + "step": 9159 + }, + { + "epoch": 2.0841865756541527, + "grad_norm": 2.0120140017901584, + "learning_rate": 7.865245471197198e-07, + "loss": 0.0986, + "step": 9160 + }, + { + "epoch": 2.0844141069397044, + "grad_norm": 1.1629231276556602, + "learning_rate": 7.864382296299689e-07, + "loss": 0.1165, + "step": 9161 + }, + { + "epoch": 2.084641638225256, + "grad_norm": 1.2142308313846102, + "learning_rate": 7.863519088407164e-07, + "loss": 0.0208, + "step": 9162 + }, + { + "epoch": 2.084869169510808, + "grad_norm": 2.4135273275384908, + "learning_rate": 7.862655847537265e-07, + "loss": 0.0456, + "step": 9163 + }, + { + "epoch": 2.0850967007963597, + "grad_norm": 1.793821084419417, + "learning_rate": 7.861792573707634e-07, + "loss": 0.0241, + "step": 9164 + }, + { + "epoch": 2.0853242320819114, + "grad_norm": 1.4462244269890792, + "learning_rate": 7.860929266935914e-07, + "loss": 0.0216, + "step": 9165 + }, + { + "epoch": 2.085551763367463, + "grad_norm": 1.8524138928143818, + "learning_rate": 7.860065927239752e-07, + "loss": 0.0659, + "step": 9166 + }, + { + "epoch": 2.085779294653015, + "grad_norm": 1.5037523706054334, + "learning_rate": 7.859202554636788e-07, + "loss": 0.0533, + "step": 9167 + }, + { + "epoch": 2.0860068259385667, + "grad_norm": 1.269697638544373, + "learning_rate": 7.858339149144671e-07, + "loss": 0.0749, + "step": 9168 + }, + { + "epoch": 2.0862343572241184, + "grad_norm": 1.8488925262658285, + "learning_rate": 7.857475710781048e-07, + "loss": 0.0428, + "step": 9169 + }, + { + "epoch": 2.08646188850967, + "grad_norm": 1.1660865685004795, + "learning_rate": 7.856612239563568e-07, + "loss": 0.0405, + "step": 9170 + }, + { + "epoch": 2.086689419795222, + "grad_norm": 1.1594702561901014, + "learning_rate": 7.855748735509873e-07, + "loss": 0.0301, + "step": 9171 + }, + { + "epoch": 2.0869169510807737, + "grad_norm": 1.7929484121993156, + "learning_rate": 7.854885198637615e-07, + "loss": 0.031, + "step": 9172 + }, + { + "epoch": 2.0871444823663254, + "grad_norm": 1.9650678687188834, + "learning_rate": 7.854021628964445e-07, + "loss": 0.0825, + "step": 9173 + }, + { + "epoch": 2.087372013651877, + "grad_norm": 1.5320972901943375, + "learning_rate": 7.853158026508009e-07, + "loss": 0.042, + "step": 9174 + }, + { + "epoch": 2.087599544937429, + "grad_norm": 1.125421549574689, + "learning_rate": 7.852294391285959e-07, + "loss": 0.0277, + "step": 9175 + }, + { + "epoch": 2.0878270762229807, + "grad_norm": 1.8538364240405865, + "learning_rate": 7.851430723315946e-07, + "loss": 0.0852, + "step": 9176 + }, + { + "epoch": 2.0880546075085324, + "grad_norm": 0.9958089698704002, + "learning_rate": 7.850567022615623e-07, + "loss": 0.0159, + "step": 9177 + }, + { + "epoch": 2.088282138794084, + "grad_norm": 2.4409117643496354, + "learning_rate": 7.84970328920264e-07, + "loss": 0.1737, + "step": 9178 + }, + { + "epoch": 2.088509670079636, + "grad_norm": 1.6812234925733471, + "learning_rate": 7.848839523094651e-07, + "loss": 0.0558, + "step": 9179 + }, + { + "epoch": 2.0887372013651877, + "grad_norm": 1.9036372254475549, + "learning_rate": 7.847975724309312e-07, + "loss": 0.0563, + "step": 9180 + }, + { + "epoch": 2.0889647326507395, + "grad_norm": 2.337524718407092, + "learning_rate": 7.847111892864275e-07, + "loss": 0.1269, + "step": 9181 + }, + { + "epoch": 2.089192263936291, + "grad_norm": 1.1593951789956227, + "learning_rate": 7.846248028777195e-07, + "loss": 0.0389, + "step": 9182 + }, + { + "epoch": 2.089419795221843, + "grad_norm": 2.7307935985707568, + "learning_rate": 7.845384132065731e-07, + "loss": 0.127, + "step": 9183 + }, + { + "epoch": 2.0896473265073947, + "grad_norm": 1.9933713250868286, + "learning_rate": 7.844520202747534e-07, + "loss": 0.0581, + "step": 9184 + }, + { + "epoch": 2.0898748577929465, + "grad_norm": 0.960440095213362, + "learning_rate": 7.843656240840265e-07, + "loss": 0.0621, + "step": 9185 + }, + { + "epoch": 2.090102389078498, + "grad_norm": 1.3040238998096036, + "learning_rate": 7.842792246361583e-07, + "loss": 0.0401, + "step": 9186 + }, + { + "epoch": 2.09032992036405, + "grad_norm": 2.9926707670526156, + "learning_rate": 7.841928219329144e-07, + "loss": 0.143, + "step": 9187 + }, + { + "epoch": 2.0905574516496017, + "grad_norm": 1.8199233717870194, + "learning_rate": 7.841064159760605e-07, + "loss": 0.065, + "step": 9188 + }, + { + "epoch": 2.0907849829351535, + "grad_norm": 1.310924972378196, + "learning_rate": 7.840200067673632e-07, + "loss": 0.0489, + "step": 9189 + }, + { + "epoch": 2.091012514220705, + "grad_norm": 1.3121302997582052, + "learning_rate": 7.839335943085882e-07, + "loss": 0.019, + "step": 9190 + }, + { + "epoch": 2.091240045506257, + "grad_norm": 1.0244602061787904, + "learning_rate": 7.838471786015013e-07, + "loss": 0.0361, + "step": 9191 + }, + { + "epoch": 2.0914675767918087, + "grad_norm": 2.161668507744929, + "learning_rate": 7.837607596478691e-07, + "loss": 0.0548, + "step": 9192 + }, + { + "epoch": 2.0916951080773605, + "grad_norm": 1.287337927794009, + "learning_rate": 7.836743374494579e-07, + "loss": 0.072, + "step": 9193 + }, + { + "epoch": 2.0919226393629122, + "grad_norm": 1.3083130207692248, + "learning_rate": 7.835879120080337e-07, + "loss": 0.0175, + "step": 9194 + }, + { + "epoch": 2.092150170648464, + "grad_norm": 1.4580985938386846, + "learning_rate": 7.83501483325363e-07, + "loss": 0.0385, + "step": 9195 + }, + { + "epoch": 2.0923777019340157, + "grad_norm": 1.7573973967743006, + "learning_rate": 7.834150514032124e-07, + "loss": 0.0702, + "step": 9196 + }, + { + "epoch": 2.0926052332195675, + "grad_norm": 0.9961034152160794, + "learning_rate": 7.833286162433481e-07, + "loss": 0.0192, + "step": 9197 + }, + { + "epoch": 2.0928327645051192, + "grad_norm": 1.8957486861488144, + "learning_rate": 7.832421778475369e-07, + "loss": 0.056, + "step": 9198 + }, + { + "epoch": 2.0930602957906714, + "grad_norm": 1.352577067070104, + "learning_rate": 7.831557362175454e-07, + "loss": 0.1217, + "step": 9199 + }, + { + "epoch": 2.093287827076223, + "grad_norm": 1.5536317526274144, + "learning_rate": 7.830692913551403e-07, + "loss": 0.1074, + "step": 9200 + }, + { + "epoch": 2.093515358361775, + "grad_norm": 1.8333157486267224, + "learning_rate": 7.829828432620884e-07, + "loss": 0.0442, + "step": 9201 + }, + { + "epoch": 2.0937428896473267, + "grad_norm": 1.3639882761041908, + "learning_rate": 7.828963919401567e-07, + "loss": 0.0254, + "step": 9202 + }, + { + "epoch": 2.0939704209328784, + "grad_norm": 1.171910362710783, + "learning_rate": 7.828099373911116e-07, + "loss": 0.063, + "step": 9203 + }, + { + "epoch": 2.09419795221843, + "grad_norm": 1.1519817333073965, + "learning_rate": 7.827234796167206e-07, + "loss": 0.0642, + "step": 9204 + }, + { + "epoch": 2.094425483503982, + "grad_norm": 1.486828533039071, + "learning_rate": 7.826370186187503e-07, + "loss": 0.0555, + "step": 9205 + }, + { + "epoch": 2.0946530147895337, + "grad_norm": 1.4307695048245372, + "learning_rate": 7.825505543989682e-07, + "loss": 0.0521, + "step": 9206 + }, + { + "epoch": 2.0948805460750854, + "grad_norm": 2.2027508824528765, + "learning_rate": 7.824640869591412e-07, + "loss": 0.1103, + "step": 9207 + }, + { + "epoch": 2.095108077360637, + "grad_norm": 2.1915433152702546, + "learning_rate": 7.823776163010369e-07, + "loss": 0.0485, + "step": 9208 + }, + { + "epoch": 2.095335608646189, + "grad_norm": 1.1490720786451287, + "learning_rate": 7.822911424264222e-07, + "loss": 0.0243, + "step": 9209 + }, + { + "epoch": 2.0955631399317407, + "grad_norm": 1.3690667315155065, + "learning_rate": 7.822046653370647e-07, + "loss": 0.0666, + "step": 9210 + }, + { + "epoch": 2.0957906712172925, + "grad_norm": 1.9260718490822129, + "learning_rate": 7.821181850347316e-07, + "loss": 0.0282, + "step": 9211 + }, + { + "epoch": 2.096018202502844, + "grad_norm": 2.13769511073873, + "learning_rate": 7.820317015211905e-07, + "loss": 0.1475, + "step": 9212 + }, + { + "epoch": 2.096245733788396, + "grad_norm": 1.6462915109688485, + "learning_rate": 7.819452147982091e-07, + "loss": 0.0544, + "step": 9213 + }, + { + "epoch": 2.0964732650739477, + "grad_norm": 1.4228717351251425, + "learning_rate": 7.81858724867555e-07, + "loss": 0.0302, + "step": 9214 + }, + { + "epoch": 2.0967007963594995, + "grad_norm": 2.0362478269213815, + "learning_rate": 7.817722317309958e-07, + "loss": 0.0908, + "step": 9215 + }, + { + "epoch": 2.096928327645051, + "grad_norm": 3.2784385787483243, + "learning_rate": 7.816857353902993e-07, + "loss": 0.0597, + "step": 9216 + }, + { + "epoch": 2.097155858930603, + "grad_norm": 1.5586908834050757, + "learning_rate": 7.815992358472332e-07, + "loss": 0.0292, + "step": 9217 + }, + { + "epoch": 2.0973833902161547, + "grad_norm": 1.547293545985965, + "learning_rate": 7.815127331035656e-07, + "loss": 0.0405, + "step": 9218 + }, + { + "epoch": 2.0976109215017065, + "grad_norm": 1.532042401197724, + "learning_rate": 7.814262271610645e-07, + "loss": 0.0261, + "step": 9219 + }, + { + "epoch": 2.0978384527872582, + "grad_norm": 2.2259657887934465, + "learning_rate": 7.813397180214978e-07, + "loss": 0.1, + "step": 9220 + }, + { + "epoch": 2.09806598407281, + "grad_norm": 1.656783599799285, + "learning_rate": 7.812532056866334e-07, + "loss": 0.1656, + "step": 9221 + }, + { + "epoch": 2.0982935153583617, + "grad_norm": 1.9188569759547522, + "learning_rate": 7.811666901582397e-07, + "loss": 0.1067, + "step": 9222 + }, + { + "epoch": 2.0985210466439135, + "grad_norm": 1.6568588645485047, + "learning_rate": 7.81080171438085e-07, + "loss": 0.0211, + "step": 9223 + }, + { + "epoch": 2.0987485779294652, + "grad_norm": 1.6822198634989907, + "learning_rate": 7.809936495279373e-07, + "loss": 0.035, + "step": 9224 + }, + { + "epoch": 2.098976109215017, + "grad_norm": 1.7191687731002951, + "learning_rate": 7.80907124429565e-07, + "loss": 0.042, + "step": 9225 + }, + { + "epoch": 2.0992036405005687, + "grad_norm": 1.8710917610254834, + "learning_rate": 7.808205961447368e-07, + "loss": 0.0381, + "step": 9226 + }, + { + "epoch": 2.0994311717861205, + "grad_norm": 2.3408149009690513, + "learning_rate": 7.80734064675221e-07, + "loss": 0.1432, + "step": 9227 + }, + { + "epoch": 2.0996587030716722, + "grad_norm": 1.9889802010588338, + "learning_rate": 7.806475300227859e-07, + "loss": 0.056, + "step": 9228 + }, + { + "epoch": 2.099886234357224, + "grad_norm": 2.674385690194074, + "learning_rate": 7.805609921892006e-07, + "loss": 0.0483, + "step": 9229 + }, + { + "epoch": 2.1001137656427757, + "grad_norm": 1.02491445767166, + "learning_rate": 7.804744511762334e-07, + "loss": 0.0736, + "step": 9230 + }, + { + "epoch": 2.1003412969283275, + "grad_norm": 1.117188844229928, + "learning_rate": 7.803879069856532e-07, + "loss": 0.0221, + "step": 9231 + }, + { + "epoch": 2.1005688282138792, + "grad_norm": 1.3889245732873237, + "learning_rate": 7.803013596192285e-07, + "loss": 0.1145, + "step": 9232 + }, + { + "epoch": 2.100796359499431, + "grad_norm": 2.7664038135593243, + "learning_rate": 7.802148090787289e-07, + "loss": 0.0757, + "step": 9233 + }, + { + "epoch": 2.1010238907849828, + "grad_norm": 1.717634586587934, + "learning_rate": 7.801282553659225e-07, + "loss": 0.0867, + "step": 9234 + }, + { + "epoch": 2.1012514220705345, + "grad_norm": 1.604396344801769, + "learning_rate": 7.800416984825787e-07, + "loss": 0.0786, + "step": 9235 + }, + { + "epoch": 2.1014789533560863, + "grad_norm": 1.3900155303859807, + "learning_rate": 7.799551384304666e-07, + "loss": 0.0755, + "step": 9236 + }, + { + "epoch": 2.101706484641638, + "grad_norm": 1.6256090464414028, + "learning_rate": 7.798685752113553e-07, + "loss": 0.0839, + "step": 9237 + }, + { + "epoch": 2.10193401592719, + "grad_norm": 1.504365372068615, + "learning_rate": 7.797820088270141e-07, + "loss": 0.0649, + "step": 9238 + }, + { + "epoch": 2.102161547212742, + "grad_norm": 1.9296763686094016, + "learning_rate": 7.796954392792119e-07, + "loss": 0.1081, + "step": 9239 + }, + { + "epoch": 2.1023890784982937, + "grad_norm": 1.52904792022275, + "learning_rate": 7.796088665697183e-07, + "loss": 0.1283, + "step": 9240 + }, + { + "epoch": 2.1026166097838455, + "grad_norm": 1.089545196735548, + "learning_rate": 7.795222907003025e-07, + "loss": 0.0625, + "step": 9241 + }, + { + "epoch": 2.102844141069397, + "grad_norm": 1.2066058203186265, + "learning_rate": 7.794357116727342e-07, + "loss": 0.092, + "step": 9242 + }, + { + "epoch": 2.103071672354949, + "grad_norm": 1.3844205195318093, + "learning_rate": 7.793491294887827e-07, + "loss": 0.0141, + "step": 9243 + }, + { + "epoch": 2.1032992036405007, + "grad_norm": 1.3781651085528592, + "learning_rate": 7.792625441502178e-07, + "loss": 0.1123, + "step": 9244 + }, + { + "epoch": 2.1035267349260525, + "grad_norm": 1.5402665201462948, + "learning_rate": 7.791759556588088e-07, + "loss": 0.1873, + "step": 9245 + }, + { + "epoch": 2.103754266211604, + "grad_norm": 2.3134223307321857, + "learning_rate": 7.790893640163258e-07, + "loss": 0.1023, + "step": 9246 + }, + { + "epoch": 2.103981797497156, + "grad_norm": 1.9875973334417596, + "learning_rate": 7.790027692245383e-07, + "loss": 0.0567, + "step": 9247 + }, + { + "epoch": 2.1042093287827077, + "grad_norm": 2.7137902344214337, + "learning_rate": 7.789161712852163e-07, + "loss": 0.068, + "step": 9248 + }, + { + "epoch": 2.1044368600682595, + "grad_norm": 1.054420198034236, + "learning_rate": 7.788295702001296e-07, + "loss": 0.0787, + "step": 9249 + }, + { + "epoch": 2.1046643913538112, + "grad_norm": 1.5842620339906497, + "learning_rate": 7.787429659710483e-07, + "loss": 0.0612, + "step": 9250 + }, + { + "epoch": 2.104891922639363, + "grad_norm": 1.1935488217069934, + "learning_rate": 7.786563585997422e-07, + "loss": 0.028, + "step": 9251 + }, + { + "epoch": 2.1051194539249147, + "grad_norm": 1.6264059074191337, + "learning_rate": 7.785697480879817e-07, + "loss": 0.0296, + "step": 9252 + }, + { + "epoch": 2.1053469852104665, + "grad_norm": 1.945992851236711, + "learning_rate": 7.784831344375368e-07, + "loss": 0.0292, + "step": 9253 + }, + { + "epoch": 2.1055745164960182, + "grad_norm": 1.64066328910728, + "learning_rate": 7.783965176501776e-07, + "loss": 0.0496, + "step": 9254 + }, + { + "epoch": 2.10580204778157, + "grad_norm": 1.7098989672937894, + "learning_rate": 7.783098977276746e-07, + "loss": 0.1689, + "step": 9255 + }, + { + "epoch": 2.1060295790671217, + "grad_norm": 2.1049711830611346, + "learning_rate": 7.78223274671798e-07, + "loss": 0.0274, + "step": 9256 + }, + { + "epoch": 2.1062571103526735, + "grad_norm": 1.1594944674223808, + "learning_rate": 7.781366484843184e-07, + "loss": 0.0183, + "step": 9257 + }, + { + "epoch": 2.1064846416382252, + "grad_norm": 1.8663119260676362, + "learning_rate": 7.78050019167006e-07, + "loss": 0.0367, + "step": 9258 + }, + { + "epoch": 2.106712172923777, + "grad_norm": 2.013908637721363, + "learning_rate": 7.779633867216316e-07, + "loss": 0.0723, + "step": 9259 + }, + { + "epoch": 2.1069397042093287, + "grad_norm": 1.8907967348439711, + "learning_rate": 7.778767511499657e-07, + "loss": 0.1024, + "step": 9260 + }, + { + "epoch": 2.1071672354948805, + "grad_norm": 2.7819586912495864, + "learning_rate": 7.77790112453779e-07, + "loss": 0.0349, + "step": 9261 + }, + { + "epoch": 2.1073947667804322, + "grad_norm": 1.8492682898613977, + "learning_rate": 7.77703470634842e-07, + "loss": 0.0492, + "step": 9262 + }, + { + "epoch": 2.107622298065984, + "grad_norm": 1.6053819092913735, + "learning_rate": 7.77616825694926e-07, + "loss": 0.089, + "step": 9263 + }, + { + "epoch": 2.1078498293515358, + "grad_norm": 1.3193114208099739, + "learning_rate": 7.775301776358017e-07, + "loss": 0.1046, + "step": 9264 + }, + { + "epoch": 2.1080773606370875, + "grad_norm": 2.1875301567197623, + "learning_rate": 7.774435264592396e-07, + "loss": 0.121, + "step": 9265 + }, + { + "epoch": 2.1083048919226393, + "grad_norm": 0.6687916654549803, + "learning_rate": 7.773568721670111e-07, + "loss": 0.0259, + "step": 9266 + }, + { + "epoch": 2.108532423208191, + "grad_norm": 1.8008293534548026, + "learning_rate": 7.772702147608873e-07, + "loss": 0.0264, + "step": 9267 + }, + { + "epoch": 2.1087599544937428, + "grad_norm": 1.933657792193932, + "learning_rate": 7.771835542426389e-07, + "loss": 0.0621, + "step": 9268 + }, + { + "epoch": 2.1089874857792945, + "grad_norm": 1.9376188439217816, + "learning_rate": 7.770968906140376e-07, + "loss": 0.1101, + "step": 9269 + }, + { + "epoch": 2.1092150170648463, + "grad_norm": 1.5472534560706894, + "learning_rate": 7.770102238768543e-07, + "loss": 0.0526, + "step": 9270 + }, + { + "epoch": 2.109442548350398, + "grad_norm": 0.7997691185086151, + "learning_rate": 7.769235540328607e-07, + "loss": 0.0095, + "step": 9271 + }, + { + "epoch": 2.1096700796359498, + "grad_norm": 1.267434805419255, + "learning_rate": 7.768368810838275e-07, + "loss": 0.0345, + "step": 9272 + }, + { + "epoch": 2.1098976109215015, + "grad_norm": 2.3345124217744986, + "learning_rate": 7.767502050315266e-07, + "loss": 0.0527, + "step": 9273 + }, + { + "epoch": 2.1101251422070533, + "grad_norm": 1.6574598892540606, + "learning_rate": 7.766635258777296e-07, + "loss": 0.0457, + "step": 9274 + }, + { + "epoch": 2.110352673492605, + "grad_norm": 1.8446041861565192, + "learning_rate": 7.765768436242077e-07, + "loss": 0.0545, + "step": 9275 + }, + { + "epoch": 2.1105802047781568, + "grad_norm": 1.9922484224033918, + "learning_rate": 7.764901582727328e-07, + "loss": 0.1015, + "step": 9276 + }, + { + "epoch": 2.110807736063709, + "grad_norm": 1.4663333368461369, + "learning_rate": 7.764034698250767e-07, + "loss": 0.0403, + "step": 9277 + }, + { + "epoch": 2.1110352673492607, + "grad_norm": 1.7587593839166777, + "learning_rate": 7.763167782830107e-07, + "loss": 0.0442, + "step": 9278 + }, + { + "epoch": 2.1112627986348125, + "grad_norm": 1.5505472155812654, + "learning_rate": 7.762300836483069e-07, + "loss": 0.0297, + "step": 9279 + }, + { + "epoch": 2.1114903299203642, + "grad_norm": 4.552718774084342, + "learning_rate": 7.761433859227373e-07, + "loss": 0.1291, + "step": 9280 + }, + { + "epoch": 2.111717861205916, + "grad_norm": 1.7308443259341895, + "learning_rate": 7.760566851080736e-07, + "loss": 0.1098, + "step": 9281 + }, + { + "epoch": 2.1119453924914677, + "grad_norm": 1.4546359242636488, + "learning_rate": 7.759699812060878e-07, + "loss": 0.0327, + "step": 9282 + }, + { + "epoch": 2.1121729237770195, + "grad_norm": 1.3589593002123062, + "learning_rate": 7.758832742185523e-07, + "loss": 0.0461, + "step": 9283 + }, + { + "epoch": 2.1124004550625712, + "grad_norm": 1.7187535172779498, + "learning_rate": 7.75796564147239e-07, + "loss": 0.0301, + "step": 9284 + }, + { + "epoch": 2.112627986348123, + "grad_norm": 1.8877137255874747, + "learning_rate": 7.757098509939198e-07, + "loss": 0.1051, + "step": 9285 + }, + { + "epoch": 2.1128555176336747, + "grad_norm": 1.6167586830480725, + "learning_rate": 7.756231347603674e-07, + "loss": 0.0257, + "step": 9286 + }, + { + "epoch": 2.1130830489192265, + "grad_norm": 4.788451240481285, + "learning_rate": 7.755364154483541e-07, + "loss": 0.0366, + "step": 9287 + }, + { + "epoch": 2.1133105802047782, + "grad_norm": 1.6206474647519005, + "learning_rate": 7.754496930596521e-07, + "loss": 0.062, + "step": 9288 + }, + { + "epoch": 2.11353811149033, + "grad_norm": 1.907574892689514, + "learning_rate": 7.753629675960339e-07, + "loss": 0.0288, + "step": 9289 + }, + { + "epoch": 2.1137656427758817, + "grad_norm": 0.9073674986740545, + "learning_rate": 7.752762390592721e-07, + "loss": 0.0232, + "step": 9290 + }, + { + "epoch": 2.1139931740614335, + "grad_norm": 1.880735312372201, + "learning_rate": 7.751895074511391e-07, + "loss": 0.0821, + "step": 9291 + }, + { + "epoch": 2.1142207053469853, + "grad_norm": 1.3482497632602575, + "learning_rate": 7.751027727734075e-07, + "loss": 0.0833, + "step": 9292 + }, + { + "epoch": 2.114448236632537, + "grad_norm": 1.0427717471233335, + "learning_rate": 7.750160350278503e-07, + "loss": 0.0175, + "step": 9293 + }, + { + "epoch": 2.1146757679180888, + "grad_norm": 2.3483958958765068, + "learning_rate": 7.7492929421624e-07, + "loss": 0.0537, + "step": 9294 + }, + { + "epoch": 2.1149032992036405, + "grad_norm": 2.161396815830777, + "learning_rate": 7.748425503403497e-07, + "loss": 0.0859, + "step": 9295 + }, + { + "epoch": 2.1151308304891923, + "grad_norm": 1.8146383508840411, + "learning_rate": 7.747558034019518e-07, + "loss": 0.0833, + "step": 9296 + }, + { + "epoch": 2.115358361774744, + "grad_norm": 2.5248523466150554, + "learning_rate": 7.746690534028198e-07, + "loss": 0.0817, + "step": 9297 + }, + { + "epoch": 2.1155858930602958, + "grad_norm": 1.9570064726403718, + "learning_rate": 7.745823003447262e-07, + "loss": 0.0172, + "step": 9298 + }, + { + "epoch": 2.1158134243458475, + "grad_norm": 1.5616533451594679, + "learning_rate": 7.744955442294444e-07, + "loss": 0.0845, + "step": 9299 + }, + { + "epoch": 2.1160409556313993, + "grad_norm": 1.114550302147854, + "learning_rate": 7.744087850587476e-07, + "loss": 0.0609, + "step": 9300 + }, + { + "epoch": 2.116268486916951, + "grad_norm": 1.4756084667763665, + "learning_rate": 7.743220228344088e-07, + "loss": 0.0667, + "step": 9301 + }, + { + "epoch": 2.1164960182025028, + "grad_norm": 0.5628312222017164, + "learning_rate": 7.742352575582014e-07, + "loss": 0.006, + "step": 9302 + }, + { + "epoch": 2.1167235494880545, + "grad_norm": 1.2782286449244458, + "learning_rate": 7.741484892318986e-07, + "loss": 0.0482, + "step": 9303 + }, + { + "epoch": 2.1169510807736063, + "grad_norm": 1.3831854833490138, + "learning_rate": 7.740617178572738e-07, + "loss": 0.0322, + "step": 9304 + }, + { + "epoch": 2.117178612059158, + "grad_norm": 1.9633493269534374, + "learning_rate": 7.739749434361003e-07, + "loss": 0.2001, + "step": 9305 + }, + { + "epoch": 2.11740614334471, + "grad_norm": 1.6257913797625625, + "learning_rate": 7.738881659701521e-07, + "loss": 0.0571, + "step": 9306 + }, + { + "epoch": 2.1176336746302615, + "grad_norm": 1.9442813740192841, + "learning_rate": 7.738013854612023e-07, + "loss": 0.0408, + "step": 9307 + }, + { + "epoch": 2.1178612059158133, + "grad_norm": 1.2053114079894778, + "learning_rate": 7.737146019110247e-07, + "loss": 0.0525, + "step": 9308 + }, + { + "epoch": 2.118088737201365, + "grad_norm": 0.9481311933193868, + "learning_rate": 7.736278153213932e-07, + "loss": 0.0736, + "step": 9309 + }, + { + "epoch": 2.118316268486917, + "grad_norm": 1.4583880459240584, + "learning_rate": 7.735410256940812e-07, + "loss": 0.0236, + "step": 9310 + }, + { + "epoch": 2.1185437997724685, + "grad_norm": 1.6654794619107185, + "learning_rate": 7.734542330308626e-07, + "loss": 0.0201, + "step": 9311 + }, + { + "epoch": 2.1187713310580203, + "grad_norm": 1.9559494752862399, + "learning_rate": 7.733674373335112e-07, + "loss": 0.0497, + "step": 9312 + }, + { + "epoch": 2.118998862343572, + "grad_norm": 1.6291429217634994, + "learning_rate": 7.732806386038013e-07, + "loss": 0.076, + "step": 9313 + }, + { + "epoch": 2.119226393629124, + "grad_norm": 1.6769086835362415, + "learning_rate": 7.731938368435069e-07, + "loss": 0.1245, + "step": 9314 + }, + { + "epoch": 2.1194539249146755, + "grad_norm": 1.7261062993980787, + "learning_rate": 7.731070320544017e-07, + "loss": 0.0645, + "step": 9315 + }, + { + "epoch": 2.1196814562002277, + "grad_norm": 1.354263270160637, + "learning_rate": 7.7302022423826e-07, + "loss": 0.0473, + "step": 9316 + }, + { + "epoch": 2.1199089874857795, + "grad_norm": 1.144116211948735, + "learning_rate": 7.72933413396856e-07, + "loss": 0.0403, + "step": 9317 + }, + { + "epoch": 2.1201365187713312, + "grad_norm": 0.7540237291195536, + "learning_rate": 7.728465995319641e-07, + "loss": 0.0095, + "step": 9318 + }, + { + "epoch": 2.120364050056883, + "grad_norm": 1.9035585388260063, + "learning_rate": 7.727597826453583e-07, + "loss": 0.0668, + "step": 9319 + }, + { + "epoch": 2.1205915813424348, + "grad_norm": 1.8238624012381768, + "learning_rate": 7.726729627388132e-07, + "loss": 0.0504, + "step": 9320 + }, + { + "epoch": 2.1208191126279865, + "grad_norm": 1.9086678345242842, + "learning_rate": 7.725861398141032e-07, + "loss": 0.047, + "step": 9321 + }, + { + "epoch": 2.1210466439135383, + "grad_norm": 1.5033310708941, + "learning_rate": 7.724993138730027e-07, + "loss": 0.0843, + "step": 9322 + }, + { + "epoch": 2.12127417519909, + "grad_norm": 1.1496840426498787, + "learning_rate": 7.724124849172865e-07, + "loss": 0.0972, + "step": 9323 + }, + { + "epoch": 2.1215017064846418, + "grad_norm": 2.3831184390276308, + "learning_rate": 7.723256529487291e-07, + "loss": 0.0543, + "step": 9324 + }, + { + "epoch": 2.1217292377701935, + "grad_norm": 1.3729829894259622, + "learning_rate": 7.722388179691051e-07, + "loss": 0.1203, + "step": 9325 + }, + { + "epoch": 2.1219567690557453, + "grad_norm": 1.5554020234867307, + "learning_rate": 7.721519799801892e-07, + "loss": 0.1493, + "step": 9326 + }, + { + "epoch": 2.122184300341297, + "grad_norm": 1.5741715519315977, + "learning_rate": 7.720651389837566e-07, + "loss": 0.0454, + "step": 9327 + }, + { + "epoch": 2.1224118316268488, + "grad_norm": 2.105704821929588, + "learning_rate": 7.719782949815817e-07, + "loss": 0.0964, + "step": 9328 + }, + { + "epoch": 2.1226393629124005, + "grad_norm": 1.6909820738720005, + "learning_rate": 7.718914479754395e-07, + "loss": 0.0381, + "step": 9329 + }, + { + "epoch": 2.1228668941979523, + "grad_norm": 1.3862966404078314, + "learning_rate": 7.718045979671054e-07, + "loss": 0.03, + "step": 9330 + }, + { + "epoch": 2.123094425483504, + "grad_norm": 1.4632245048086763, + "learning_rate": 7.717177449583541e-07, + "loss": 0.0194, + "step": 9331 + }, + { + "epoch": 2.1233219567690558, + "grad_norm": 1.7957475628487232, + "learning_rate": 7.716308889509608e-07, + "loss": 0.0962, + "step": 9332 + }, + { + "epoch": 2.1235494880546075, + "grad_norm": 1.4308951859223038, + "learning_rate": 7.715440299467006e-07, + "loss": 0.075, + "step": 9333 + }, + { + "epoch": 2.1237770193401593, + "grad_norm": 1.827115383065602, + "learning_rate": 7.714571679473489e-07, + "loss": 0.079, + "step": 9334 + }, + { + "epoch": 2.124004550625711, + "grad_norm": 1.7496968559853514, + "learning_rate": 7.713703029546809e-07, + "loss": 0.0694, + "step": 9335 + }, + { + "epoch": 2.124232081911263, + "grad_norm": 1.3700989172510394, + "learning_rate": 7.712834349704718e-07, + "loss": 0.0216, + "step": 9336 + }, + { + "epoch": 2.1244596131968145, + "grad_norm": 1.5251962233140202, + "learning_rate": 7.711965639964971e-07, + "loss": 0.0336, + "step": 9337 + }, + { + "epoch": 2.1246871444823663, + "grad_norm": 2.07406916221973, + "learning_rate": 7.711096900345327e-07, + "loss": 0.0501, + "step": 9338 + }, + { + "epoch": 2.124914675767918, + "grad_norm": 1.5684792220083636, + "learning_rate": 7.710228130863537e-07, + "loss": 0.1003, + "step": 9339 + }, + { + "epoch": 2.12514220705347, + "grad_norm": 1.1999205767233307, + "learning_rate": 7.70935933153736e-07, + "loss": 0.0676, + "step": 9340 + }, + { + "epoch": 2.1253697383390215, + "grad_norm": 0.9945934713928603, + "learning_rate": 7.70849050238455e-07, + "loss": 0.0183, + "step": 9341 + }, + { + "epoch": 2.1255972696245733, + "grad_norm": 1.5438017475505683, + "learning_rate": 7.707621643422862e-07, + "loss": 0.0237, + "step": 9342 + }, + { + "epoch": 2.125824800910125, + "grad_norm": 1.334350766967659, + "learning_rate": 7.706752754670061e-07, + "loss": 0.1407, + "step": 9343 + }, + { + "epoch": 2.126052332195677, + "grad_norm": 1.1002643363414366, + "learning_rate": 7.7058838361439e-07, + "loss": 0.0318, + "step": 9344 + }, + { + "epoch": 2.1262798634812285, + "grad_norm": 1.1860539424250727, + "learning_rate": 7.70501488786214e-07, + "loss": 0.0162, + "step": 9345 + }, + { + "epoch": 2.1265073947667803, + "grad_norm": 0.8553559475992251, + "learning_rate": 7.70414590984254e-07, + "loss": 0.0175, + "step": 9346 + }, + { + "epoch": 2.126734926052332, + "grad_norm": 1.4510114726435523, + "learning_rate": 7.703276902102863e-07, + "loss": 0.0439, + "step": 9347 + }, + { + "epoch": 2.126962457337884, + "grad_norm": 1.91645784490155, + "learning_rate": 7.702407864660865e-07, + "loss": 0.0397, + "step": 9348 + }, + { + "epoch": 2.1271899886234356, + "grad_norm": 2.0171474300527814, + "learning_rate": 7.701538797534311e-07, + "loss": 0.0188, + "step": 9349 + }, + { + "epoch": 2.1274175199089873, + "grad_norm": 2.6771909067136868, + "learning_rate": 7.700669700740962e-07, + "loss": 0.0433, + "step": 9350 + }, + { + "epoch": 2.127645051194539, + "grad_norm": 2.3833218184332954, + "learning_rate": 7.699800574298582e-07, + "loss": 0.0888, + "step": 9351 + }, + { + "epoch": 2.127872582480091, + "grad_norm": 1.2784843894285451, + "learning_rate": 7.698931418224934e-07, + "loss": 0.0384, + "step": 9352 + }, + { + "epoch": 2.128100113765643, + "grad_norm": 1.0439730225104034, + "learning_rate": 7.69806223253778e-07, + "loss": 0.0149, + "step": 9353 + }, + { + "epoch": 2.1283276450511943, + "grad_norm": 1.1905001337091723, + "learning_rate": 7.697193017254886e-07, + "loss": 0.0552, + "step": 9354 + }, + { + "epoch": 2.1285551763367465, + "grad_norm": 2.263266920896213, + "learning_rate": 7.696323772394018e-07, + "loss": 0.1084, + "step": 9355 + }, + { + "epoch": 2.1287827076222983, + "grad_norm": 2.2829384441719722, + "learning_rate": 7.69545449797294e-07, + "loss": 0.0593, + "step": 9356 + }, + { + "epoch": 2.12901023890785, + "grad_norm": 1.2589596813152582, + "learning_rate": 7.694585194009419e-07, + "loss": 0.0679, + "step": 9357 + }, + { + "epoch": 2.1292377701934018, + "grad_norm": 2.1008274118717063, + "learning_rate": 7.693715860521224e-07, + "loss": 0.0494, + "step": 9358 + }, + { + "epoch": 2.1294653014789535, + "grad_norm": 2.093089778207093, + "learning_rate": 7.69284649752612e-07, + "loss": 0.0417, + "step": 9359 + }, + { + "epoch": 2.1296928327645053, + "grad_norm": 1.6706339428594954, + "learning_rate": 7.691977105041876e-07, + "loss": 0.0568, + "step": 9360 + }, + { + "epoch": 2.129920364050057, + "grad_norm": 1.61865596608991, + "learning_rate": 7.69110768308626e-07, + "loss": 0.0462, + "step": 9361 + }, + { + "epoch": 2.1301478953356088, + "grad_norm": 1.3800817073437315, + "learning_rate": 7.690238231677045e-07, + "loss": 0.0691, + "step": 9362 + }, + { + "epoch": 2.1303754266211605, + "grad_norm": 1.4972416473429389, + "learning_rate": 7.689368750831994e-07, + "loss": 0.0413, + "step": 9363 + }, + { + "epoch": 2.1306029579067123, + "grad_norm": 1.467718571259731, + "learning_rate": 7.688499240568883e-07, + "loss": 0.0178, + "step": 9364 + }, + { + "epoch": 2.130830489192264, + "grad_norm": 1.1384642957667104, + "learning_rate": 7.687629700905485e-07, + "loss": 0.019, + "step": 9365 + }, + { + "epoch": 2.131058020477816, + "grad_norm": 1.495499693656789, + "learning_rate": 7.686760131859566e-07, + "loss": 0.0816, + "step": 9366 + }, + { + "epoch": 2.1312855517633675, + "grad_norm": 1.0987974074095415, + "learning_rate": 7.685890533448901e-07, + "loss": 0.0739, + "step": 9367 + }, + { + "epoch": 2.1315130830489193, + "grad_norm": 1.328379571268298, + "learning_rate": 7.685020905691265e-07, + "loss": 0.0311, + "step": 9368 + }, + { + "epoch": 2.131740614334471, + "grad_norm": 1.8858406859241228, + "learning_rate": 7.684151248604428e-07, + "loss": 0.086, + "step": 9369 + }, + { + "epoch": 2.131968145620023, + "grad_norm": 2.1099407081741903, + "learning_rate": 7.683281562206167e-07, + "loss": 0.0636, + "step": 9370 + }, + { + "epoch": 2.1321956769055745, + "grad_norm": 1.8010112143230375, + "learning_rate": 7.682411846514258e-07, + "loss": 0.1258, + "step": 9371 + }, + { + "epoch": 2.1324232081911263, + "grad_norm": 0.908234135058555, + "learning_rate": 7.68154210154647e-07, + "loss": 0.0164, + "step": 9372 + }, + { + "epoch": 2.132650739476678, + "grad_norm": 2.1527881845029926, + "learning_rate": 7.680672327320586e-07, + "loss": 0.1047, + "step": 9373 + }, + { + "epoch": 2.13287827076223, + "grad_norm": 1.6232277831727786, + "learning_rate": 7.679802523854379e-07, + "loss": 0.0372, + "step": 9374 + }, + { + "epoch": 2.1331058020477816, + "grad_norm": 1.4897316167472505, + "learning_rate": 7.678932691165628e-07, + "loss": 0.0354, + "step": 9375 + }, + { + "epoch": 2.1333333333333333, + "grad_norm": 0.8040641557234913, + "learning_rate": 7.678062829272107e-07, + "loss": 0.0386, + "step": 9376 + }, + { + "epoch": 2.133560864618885, + "grad_norm": 2.1506293627331585, + "learning_rate": 7.677192938191599e-07, + "loss": 0.0708, + "step": 9377 + }, + { + "epoch": 2.133788395904437, + "grad_norm": 1.217271987394615, + "learning_rate": 7.676323017941882e-07, + "loss": 0.0444, + "step": 9378 + }, + { + "epoch": 2.1340159271899886, + "grad_norm": 1.640052554857382, + "learning_rate": 7.675453068540733e-07, + "loss": 0.0304, + "step": 9379 + }, + { + "epoch": 2.1342434584755403, + "grad_norm": 1.396330359096103, + "learning_rate": 7.674583090005934e-07, + "loss": 0.0766, + "step": 9380 + }, + { + "epoch": 2.134470989761092, + "grad_norm": 1.2593153400616426, + "learning_rate": 7.673713082355267e-07, + "loss": 0.0612, + "step": 9381 + }, + { + "epoch": 2.134698521046644, + "grad_norm": 1.9196581659153655, + "learning_rate": 7.672843045606511e-07, + "loss": 0.0925, + "step": 9382 + }, + { + "epoch": 2.1349260523321956, + "grad_norm": 2.424430667024588, + "learning_rate": 7.671972979777448e-07, + "loss": 0.0443, + "step": 9383 + }, + { + "epoch": 2.1351535836177473, + "grad_norm": 1.789095012568431, + "learning_rate": 7.671102884885865e-07, + "loss": 0.1368, + "step": 9384 + }, + { + "epoch": 2.135381114903299, + "grad_norm": 1.9908795772804533, + "learning_rate": 7.67023276094954e-07, + "loss": 0.0493, + "step": 9385 + }, + { + "epoch": 2.135608646188851, + "grad_norm": 1.544089389991087, + "learning_rate": 7.669362607986256e-07, + "loss": 0.0939, + "step": 9386 + }, + { + "epoch": 2.1358361774744026, + "grad_norm": 1.5223363268950432, + "learning_rate": 7.668492426013802e-07, + "loss": 0.0738, + "step": 9387 + }, + { + "epoch": 2.1360637087599543, + "grad_norm": 2.590422659932699, + "learning_rate": 7.667622215049959e-07, + "loss": 0.094, + "step": 9388 + }, + { + "epoch": 2.136291240045506, + "grad_norm": 0.9533778360174558, + "learning_rate": 7.666751975112515e-07, + "loss": 0.0747, + "step": 9389 + }, + { + "epoch": 2.136518771331058, + "grad_norm": 1.6869229827903554, + "learning_rate": 7.665881706219256e-07, + "loss": 0.1335, + "step": 9390 + }, + { + "epoch": 2.1367463026166096, + "grad_norm": 1.3404189139443545, + "learning_rate": 7.665011408387968e-07, + "loss": 0.0379, + "step": 9391 + }, + { + "epoch": 2.136973833902162, + "grad_norm": 0.90948302588563, + "learning_rate": 7.664141081636438e-07, + "loss": 0.0367, + "step": 9392 + }, + { + "epoch": 2.137201365187713, + "grad_norm": 1.7494800758004867, + "learning_rate": 7.663270725982452e-07, + "loss": 0.053, + "step": 9393 + }, + { + "epoch": 2.1374288964732653, + "grad_norm": 1.1693153351230774, + "learning_rate": 7.662400341443804e-07, + "loss": 0.0583, + "step": 9394 + }, + { + "epoch": 2.137656427758817, + "grad_norm": 1.557485317691199, + "learning_rate": 7.661529928038277e-07, + "loss": 0.015, + "step": 9395 + }, + { + "epoch": 2.137883959044369, + "grad_norm": 1.6827033189928462, + "learning_rate": 7.660659485783667e-07, + "loss": 0.0508, + "step": 9396 + }, + { + "epoch": 2.1381114903299205, + "grad_norm": 1.1102334222602452, + "learning_rate": 7.659789014697758e-07, + "loss": 0.0271, + "step": 9397 + }, + { + "epoch": 2.1383390216154723, + "grad_norm": 1.774247472400488, + "learning_rate": 7.658918514798345e-07, + "loss": 0.0445, + "step": 9398 + }, + { + "epoch": 2.138566552901024, + "grad_norm": 1.5178141737496575, + "learning_rate": 7.658047986103215e-07, + "loss": 0.1275, + "step": 9399 + }, + { + "epoch": 2.138794084186576, + "grad_norm": 1.2540145922087464, + "learning_rate": 7.657177428630164e-07, + "loss": 0.0362, + "step": 9400 + }, + { + "epoch": 2.1390216154721275, + "grad_norm": 1.5881320882876477, + "learning_rate": 7.656306842396985e-07, + "loss": 0.0329, + "step": 9401 + }, + { + "epoch": 2.1392491467576793, + "grad_norm": 1.0347622447579183, + "learning_rate": 7.655436227421468e-07, + "loss": 0.0442, + "step": 9402 + }, + { + "epoch": 2.139476678043231, + "grad_norm": 0.9942062765597185, + "learning_rate": 7.65456558372141e-07, + "loss": 0.0444, + "step": 9403 + }, + { + "epoch": 2.139704209328783, + "grad_norm": 1.4894334597650998, + "learning_rate": 7.653694911314602e-07, + "loss": 0.0939, + "step": 9404 + }, + { + "epoch": 2.1399317406143346, + "grad_norm": 2.5553012707365985, + "learning_rate": 7.652824210218842e-07, + "loss": 0.0582, + "step": 9405 + }, + { + "epoch": 2.1401592718998863, + "grad_norm": 2.2207688664464396, + "learning_rate": 7.651953480451923e-07, + "loss": 0.0822, + "step": 9406 + }, + { + "epoch": 2.140386803185438, + "grad_norm": 1.459662417091181, + "learning_rate": 7.651082722031644e-07, + "loss": 0.0415, + "step": 9407 + }, + { + "epoch": 2.14061433447099, + "grad_norm": 0.9799504903348215, + "learning_rate": 7.6502119349758e-07, + "loss": 0.0176, + "step": 9408 + }, + { + "epoch": 2.1408418657565416, + "grad_norm": 2.452607093339156, + "learning_rate": 7.649341119302188e-07, + "loss": 0.0475, + "step": 9409 + }, + { + "epoch": 2.1410693970420933, + "grad_norm": 2.091931225218873, + "learning_rate": 7.648470275028607e-07, + "loss": 0.0558, + "step": 9410 + }, + { + "epoch": 2.141296928327645, + "grad_norm": 1.5658581552329647, + "learning_rate": 7.647599402172854e-07, + "loss": 0.0998, + "step": 9411 + }, + { + "epoch": 2.141524459613197, + "grad_norm": 1.823426513047552, + "learning_rate": 7.646728500752729e-07, + "loss": 0.0354, + "step": 9412 + }, + { + "epoch": 2.1417519908987486, + "grad_norm": 2.2777896486497293, + "learning_rate": 7.645857570786029e-07, + "loss": 0.0759, + "step": 9413 + }, + { + "epoch": 2.1419795221843003, + "grad_norm": 0.6840226560483477, + "learning_rate": 7.64498661229056e-07, + "loss": 0.0283, + "step": 9414 + }, + { + "epoch": 2.142207053469852, + "grad_norm": 3.3359064141459296, + "learning_rate": 7.644115625284119e-07, + "loss": 0.0575, + "step": 9415 + }, + { + "epoch": 2.142434584755404, + "grad_norm": 1.2843331590612495, + "learning_rate": 7.643244609784506e-07, + "loss": 0.0796, + "step": 9416 + }, + { + "epoch": 2.1426621160409556, + "grad_norm": 1.9411472132833747, + "learning_rate": 7.642373565809527e-07, + "loss": 0.0293, + "step": 9417 + }, + { + "epoch": 2.1428896473265073, + "grad_norm": 1.769227281674406, + "learning_rate": 7.641502493376981e-07, + "loss": 0.0649, + "step": 9418 + }, + { + "epoch": 2.143117178612059, + "grad_norm": 2.6692341252820286, + "learning_rate": 7.640631392504673e-07, + "loss": 0.0838, + "step": 9419 + }, + { + "epoch": 2.143344709897611, + "grad_norm": 1.375710303841949, + "learning_rate": 7.639760263210405e-07, + "loss": 0.0745, + "step": 9420 + }, + { + "epoch": 2.1435722411831626, + "grad_norm": 1.1020386493301202, + "learning_rate": 7.638889105511983e-07, + "loss": 0.0374, + "step": 9421 + }, + { + "epoch": 2.1437997724687143, + "grad_norm": 1.5165134244613883, + "learning_rate": 7.638017919427212e-07, + "loss": 0.0962, + "step": 9422 + }, + { + "epoch": 2.144027303754266, + "grad_norm": 2.263419238613544, + "learning_rate": 7.637146704973897e-07, + "loss": 0.0475, + "step": 9423 + }, + { + "epoch": 2.144254835039818, + "grad_norm": 1.6063544321457297, + "learning_rate": 7.636275462169843e-07, + "loss": 0.0676, + "step": 9424 + }, + { + "epoch": 2.1444823663253696, + "grad_norm": 1.535827142260173, + "learning_rate": 7.635404191032858e-07, + "loss": 0.044, + "step": 9425 + }, + { + "epoch": 2.1447098976109213, + "grad_norm": 1.3474563003971842, + "learning_rate": 7.634532891580748e-07, + "loss": 0.0468, + "step": 9426 + }, + { + "epoch": 2.144937428896473, + "grad_norm": 1.225183736294941, + "learning_rate": 7.633661563831321e-07, + "loss": 0.0297, + "step": 9427 + }, + { + "epoch": 2.145164960182025, + "grad_norm": 2.238719040829829, + "learning_rate": 7.632790207802388e-07, + "loss": 0.0631, + "step": 9428 + }, + { + "epoch": 2.1453924914675766, + "grad_norm": 1.395162871308499, + "learning_rate": 7.631918823511751e-07, + "loss": 0.0335, + "step": 9429 + }, + { + "epoch": 2.1456200227531284, + "grad_norm": 2.092423508040525, + "learning_rate": 7.631047410977227e-07, + "loss": 0.0465, + "step": 9430 + }, + { + "epoch": 2.1458475540386805, + "grad_norm": 1.4291616728321297, + "learning_rate": 7.630175970216625e-07, + "loss": 0.0265, + "step": 9431 + }, + { + "epoch": 2.146075085324232, + "grad_norm": 1.6015952917569425, + "learning_rate": 7.629304501247751e-07, + "loss": 0.0778, + "step": 9432 + }, + { + "epoch": 2.146302616609784, + "grad_norm": 1.2256299952561287, + "learning_rate": 7.628433004088419e-07, + "loss": 0.0905, + "step": 9433 + }, + { + "epoch": 2.146530147895336, + "grad_norm": 2.5100871022265974, + "learning_rate": 7.627561478756443e-07, + "loss": 0.0732, + "step": 9434 + }, + { + "epoch": 2.1467576791808876, + "grad_norm": 1.29392860753592, + "learning_rate": 7.626689925269633e-07, + "loss": 0.0405, + "step": 9435 + }, + { + "epoch": 2.1469852104664393, + "grad_norm": 1.498969777012569, + "learning_rate": 7.625818343645799e-07, + "loss": 0.0371, + "step": 9436 + }, + { + "epoch": 2.147212741751991, + "grad_norm": 1.3141756876340118, + "learning_rate": 7.624946733902762e-07, + "loss": 0.0516, + "step": 9437 + }, + { + "epoch": 2.147440273037543, + "grad_norm": 2.3200031750786443, + "learning_rate": 7.624075096058329e-07, + "loss": 0.0871, + "step": 9438 + }, + { + "epoch": 2.1476678043230946, + "grad_norm": 1.9264770280478787, + "learning_rate": 7.623203430130319e-07, + "loss": 0.0878, + "step": 9439 + }, + { + "epoch": 2.1478953356086463, + "grad_norm": 1.444710945875574, + "learning_rate": 7.622331736136546e-07, + "loss": 0.0194, + "step": 9440 + }, + { + "epoch": 2.148122866894198, + "grad_norm": 2.1435596203430145, + "learning_rate": 7.621460014094825e-07, + "loss": 0.1218, + "step": 9441 + }, + { + "epoch": 2.14835039817975, + "grad_norm": 2.1048982631125717, + "learning_rate": 7.620588264022973e-07, + "loss": 0.1983, + "step": 9442 + }, + { + "epoch": 2.1485779294653016, + "grad_norm": 1.1427311898981718, + "learning_rate": 7.619716485938805e-07, + "loss": 0.0926, + "step": 9443 + }, + { + "epoch": 2.1488054607508533, + "grad_norm": 3.2663065997266867, + "learning_rate": 7.618844679860144e-07, + "loss": 0.06, + "step": 9444 + }, + { + "epoch": 2.149032992036405, + "grad_norm": 1.6766682228640024, + "learning_rate": 7.617972845804802e-07, + "loss": 0.0417, + "step": 9445 + }, + { + "epoch": 2.149260523321957, + "grad_norm": 0.8605363751316158, + "learning_rate": 7.617100983790603e-07, + "loss": 0.0167, + "step": 9446 + }, + { + "epoch": 2.1494880546075086, + "grad_norm": 1.7274126311028568, + "learning_rate": 7.616229093835361e-07, + "loss": 0.0464, + "step": 9447 + }, + { + "epoch": 2.1497155858930603, + "grad_norm": 1.4531075823883306, + "learning_rate": 7.6153571759569e-07, + "loss": 0.0901, + "step": 9448 + }, + { + "epoch": 2.149943117178612, + "grad_norm": 1.2017254026999795, + "learning_rate": 7.61448523017304e-07, + "loss": 0.0454, + "step": 9449 + }, + { + "epoch": 2.150170648464164, + "grad_norm": 1.9031272001920252, + "learning_rate": 7.613613256501598e-07, + "loss": 0.0372, + "step": 9450 + }, + { + "epoch": 2.1503981797497156, + "grad_norm": 1.7048614482256514, + "learning_rate": 7.612741254960398e-07, + "loss": 0.0214, + "step": 9451 + }, + { + "epoch": 2.1506257110352673, + "grad_norm": 1.6154795712268546, + "learning_rate": 7.611869225567266e-07, + "loss": 0.0722, + "step": 9452 + }, + { + "epoch": 2.150853242320819, + "grad_norm": 1.6394253082305872, + "learning_rate": 7.610997168340019e-07, + "loss": 0.0546, + "step": 9453 + }, + { + "epoch": 2.151080773606371, + "grad_norm": 1.8045044276515763, + "learning_rate": 7.610125083296482e-07, + "loss": 0.0614, + "step": 9454 + }, + { + "epoch": 2.1513083048919226, + "grad_norm": 1.4902043316498548, + "learning_rate": 7.609252970454481e-07, + "loss": 0.023, + "step": 9455 + }, + { + "epoch": 2.1515358361774743, + "grad_norm": 1.5827470563432342, + "learning_rate": 7.608380829831838e-07, + "loss": 0.0824, + "step": 9456 + }, + { + "epoch": 2.151763367463026, + "grad_norm": 1.3928939708222616, + "learning_rate": 7.607508661446378e-07, + "loss": 0.0377, + "step": 9457 + }, + { + "epoch": 2.151990898748578, + "grad_norm": 4.359928182637225, + "learning_rate": 7.606636465315927e-07, + "loss": 0.0445, + "step": 9458 + }, + { + "epoch": 2.1522184300341296, + "grad_norm": 2.136122219824162, + "learning_rate": 7.605764241458313e-07, + "loss": 0.0309, + "step": 9459 + }, + { + "epoch": 2.1524459613196814, + "grad_norm": 1.0921428417713035, + "learning_rate": 7.604891989891358e-07, + "loss": 0.0478, + "step": 9460 + }, + { + "epoch": 2.152673492605233, + "grad_norm": 1.7350343198023745, + "learning_rate": 7.604019710632895e-07, + "loss": 0.0412, + "step": 9461 + }, + { + "epoch": 2.152901023890785, + "grad_norm": 1.8683402579419655, + "learning_rate": 7.603147403700746e-07, + "loss": 0.0324, + "step": 9462 + }, + { + "epoch": 2.1531285551763366, + "grad_norm": 2.03049621141726, + "learning_rate": 7.602275069112742e-07, + "loss": 0.0318, + "step": 9463 + }, + { + "epoch": 2.1533560864618884, + "grad_norm": 1.6146820302973948, + "learning_rate": 7.601402706886716e-07, + "loss": 0.0887, + "step": 9464 + }, + { + "epoch": 2.15358361774744, + "grad_norm": 1.4186738617417172, + "learning_rate": 7.60053031704049e-07, + "loss": 0.0231, + "step": 9465 + }, + { + "epoch": 2.153811149032992, + "grad_norm": 1.8502692538793657, + "learning_rate": 7.5996578995919e-07, + "loss": 0.0447, + "step": 9466 + }, + { + "epoch": 2.1540386803185436, + "grad_norm": 1.869073291535232, + "learning_rate": 7.598785454558773e-07, + "loss": 0.0502, + "step": 9467 + }, + { + "epoch": 2.1542662116040954, + "grad_norm": 1.934302737222525, + "learning_rate": 7.597912981958943e-07, + "loss": 0.1231, + "step": 9468 + }, + { + "epoch": 2.154493742889647, + "grad_norm": 1.4767373950442406, + "learning_rate": 7.597040481810239e-07, + "loss": 0.0249, + "step": 9469 + }, + { + "epoch": 2.1547212741751993, + "grad_norm": 1.675161104911878, + "learning_rate": 7.596167954130493e-07, + "loss": 0.1211, + "step": 9470 + }, + { + "epoch": 2.1549488054607506, + "grad_norm": 2.1630884838781133, + "learning_rate": 7.595295398937541e-07, + "loss": 0.0713, + "step": 9471 + }, + { + "epoch": 2.155176336746303, + "grad_norm": 1.1951201714125987, + "learning_rate": 7.594422816249217e-07, + "loss": 0.0165, + "step": 9472 + }, + { + "epoch": 2.1554038680318546, + "grad_norm": 1.4566850582274007, + "learning_rate": 7.59355020608335e-07, + "loss": 0.075, + "step": 9473 + }, + { + "epoch": 2.1556313993174063, + "grad_norm": 1.1604749857711565, + "learning_rate": 7.592677568457778e-07, + "loss": 0.1078, + "step": 9474 + }, + { + "epoch": 2.155858930602958, + "grad_norm": 1.6373713082842598, + "learning_rate": 7.591804903390336e-07, + "loss": 0.0598, + "step": 9475 + }, + { + "epoch": 2.15608646188851, + "grad_norm": 1.6959815284412367, + "learning_rate": 7.590932210898859e-07, + "loss": 0.0874, + "step": 9476 + }, + { + "epoch": 2.1563139931740616, + "grad_norm": 1.8014682740885186, + "learning_rate": 7.590059491001183e-07, + "loss": 0.0382, + "step": 9477 + }, + { + "epoch": 2.1565415244596133, + "grad_norm": 1.1796467197904394, + "learning_rate": 7.589186743715146e-07, + "loss": 0.0794, + "step": 9478 + }, + { + "epoch": 2.156769055745165, + "grad_norm": 1.756926717056566, + "learning_rate": 7.588313969058584e-07, + "loss": 0.0816, + "step": 9479 + }, + { + "epoch": 2.156996587030717, + "grad_norm": 2.6269960117902325, + "learning_rate": 7.587441167049335e-07, + "loss": 0.1693, + "step": 9480 + }, + { + "epoch": 2.1572241183162686, + "grad_norm": 0.42069891550323985, + "learning_rate": 7.586568337705239e-07, + "loss": 0.0029, + "step": 9481 + }, + { + "epoch": 2.1574516496018203, + "grad_norm": 1.6964780873318768, + "learning_rate": 7.585695481044133e-07, + "loss": 0.05, + "step": 9482 + }, + { + "epoch": 2.157679180887372, + "grad_norm": 2.3486378716222305, + "learning_rate": 7.584822597083859e-07, + "loss": 0.0782, + "step": 9483 + }, + { + "epoch": 2.157906712172924, + "grad_norm": 1.7234854015521162, + "learning_rate": 7.583949685842253e-07, + "loss": 0.0746, + "step": 9484 + }, + { + "epoch": 2.1581342434584756, + "grad_norm": 1.3851929183291392, + "learning_rate": 7.583076747337162e-07, + "loss": 0.0769, + "step": 9485 + }, + { + "epoch": 2.1583617747440274, + "grad_norm": 1.042012257312453, + "learning_rate": 7.582203781586422e-07, + "loss": 0.0309, + "step": 9486 + }, + { + "epoch": 2.158589306029579, + "grad_norm": 1.4771989092824398, + "learning_rate": 7.581330788607875e-07, + "loss": 0.0618, + "step": 9487 + }, + { + "epoch": 2.158816837315131, + "grad_norm": 0.7828197948837077, + "learning_rate": 7.580457768419367e-07, + "loss": 0.0172, + "step": 9488 + }, + { + "epoch": 2.1590443686006826, + "grad_norm": 1.7129026868418487, + "learning_rate": 7.579584721038738e-07, + "loss": 0.0445, + "step": 9489 + }, + { + "epoch": 2.1592718998862344, + "grad_norm": 1.0927664918168947, + "learning_rate": 7.578711646483832e-07, + "loss": 0.081, + "step": 9490 + }, + { + "epoch": 2.159499431171786, + "grad_norm": 1.8048150160951877, + "learning_rate": 7.577838544772495e-07, + "loss": 0.0855, + "step": 9491 + }, + { + "epoch": 2.159726962457338, + "grad_norm": 1.8082573615440634, + "learning_rate": 7.576965415922569e-07, + "loss": 0.0604, + "step": 9492 + }, + { + "epoch": 2.1599544937428896, + "grad_norm": 1.0695862916240448, + "learning_rate": 7.576092259951899e-07, + "loss": 0.0177, + "step": 9493 + }, + { + "epoch": 2.1601820250284414, + "grad_norm": 1.3988895145486373, + "learning_rate": 7.575219076878332e-07, + "loss": 0.0829, + "step": 9494 + }, + { + "epoch": 2.160409556313993, + "grad_norm": 3.021161121575875, + "learning_rate": 7.574345866719715e-07, + "loss": 0.0731, + "step": 9495 + }, + { + "epoch": 2.160637087599545, + "grad_norm": 1.2035602500367737, + "learning_rate": 7.573472629493894e-07, + "loss": 0.0161, + "step": 9496 + }, + { + "epoch": 2.1608646188850966, + "grad_norm": 1.9335725739239158, + "learning_rate": 7.572599365218716e-07, + "loss": 0.0783, + "step": 9497 + }, + { + "epoch": 2.1610921501706484, + "grad_norm": 2.508763547609667, + "learning_rate": 7.571726073912031e-07, + "loss": 0.1002, + "step": 9498 + }, + { + "epoch": 2.1613196814562, + "grad_norm": 1.279431099983586, + "learning_rate": 7.570852755591683e-07, + "loss": 0.1574, + "step": 9499 + }, + { + "epoch": 2.161547212741752, + "grad_norm": 2.664533934222619, + "learning_rate": 7.569979410275525e-07, + "loss": 0.143, + "step": 9500 + }, + { + "epoch": 2.1617747440273036, + "grad_norm": 2.137344675810386, + "learning_rate": 7.569106037981403e-07, + "loss": 0.0362, + "step": 9501 + }, + { + "epoch": 2.1620022753128554, + "grad_norm": 1.8263233406423385, + "learning_rate": 7.568232638727173e-07, + "loss": 0.0452, + "step": 9502 + }, + { + "epoch": 2.162229806598407, + "grad_norm": 1.7908284046425997, + "learning_rate": 7.56735921253068e-07, + "loss": 0.0948, + "step": 9503 + }, + { + "epoch": 2.162457337883959, + "grad_norm": 1.1055638980278877, + "learning_rate": 7.566485759409778e-07, + "loss": 0.0671, + "step": 9504 + }, + { + "epoch": 2.1626848691695106, + "grad_norm": 1.966283016527124, + "learning_rate": 7.565612279382318e-07, + "loss": 0.0748, + "step": 9505 + }, + { + "epoch": 2.1629124004550624, + "grad_norm": 0.9699618369480882, + "learning_rate": 7.564738772466153e-07, + "loss": 0.0355, + "step": 9506 + }, + { + "epoch": 2.163139931740614, + "grad_norm": 2.234936145616755, + "learning_rate": 7.563865238679133e-07, + "loss": 0.0425, + "step": 9507 + }, + { + "epoch": 2.163367463026166, + "grad_norm": 0.6953923227624036, + "learning_rate": 7.562991678039116e-07, + "loss": 0.0053, + "step": 9508 + }, + { + "epoch": 2.163594994311718, + "grad_norm": 1.4927692917033228, + "learning_rate": 7.562118090563953e-07, + "loss": 0.0753, + "step": 9509 + }, + { + "epoch": 2.1638225255972694, + "grad_norm": 1.4045146878025545, + "learning_rate": 7.5612444762715e-07, + "loss": 0.034, + "step": 9510 + }, + { + "epoch": 2.1640500568828216, + "grad_norm": 1.2359805545350997, + "learning_rate": 7.560370835179611e-07, + "loss": 0.051, + "step": 9511 + }, + { + "epoch": 2.1642775881683733, + "grad_norm": 1.299803781802618, + "learning_rate": 7.559497167306141e-07, + "loss": 0.0344, + "step": 9512 + }, + { + "epoch": 2.164505119453925, + "grad_norm": 1.5909800541101666, + "learning_rate": 7.558623472668948e-07, + "loss": 0.0332, + "step": 9513 + }, + { + "epoch": 2.164732650739477, + "grad_norm": 1.7244695675540986, + "learning_rate": 7.557749751285887e-07, + "loss": 0.0317, + "step": 9514 + }, + { + "epoch": 2.1649601820250286, + "grad_norm": 1.2855169129998656, + "learning_rate": 7.556876003174816e-07, + "loss": 0.055, + "step": 9515 + }, + { + "epoch": 2.1651877133105804, + "grad_norm": 1.4606586922640823, + "learning_rate": 7.556002228353595e-07, + "loss": 0.0864, + "step": 9516 + }, + { + "epoch": 2.165415244596132, + "grad_norm": 1.9120924782453184, + "learning_rate": 7.555128426840078e-07, + "loss": 0.0369, + "step": 9517 + }, + { + "epoch": 2.165642775881684, + "grad_norm": 1.0365258067432568, + "learning_rate": 7.554254598652127e-07, + "loss": 0.0748, + "step": 9518 + }, + { + "epoch": 2.1658703071672356, + "grad_norm": 2.1208575506873366, + "learning_rate": 7.553380743807602e-07, + "loss": 0.0278, + "step": 9519 + }, + { + "epoch": 2.1660978384527874, + "grad_norm": 1.5769455246998723, + "learning_rate": 7.552506862324358e-07, + "loss": 0.0658, + "step": 9520 + }, + { + "epoch": 2.166325369738339, + "grad_norm": 1.7764530341674318, + "learning_rate": 7.551632954220263e-07, + "loss": 0.0241, + "step": 9521 + }, + { + "epoch": 2.166552901023891, + "grad_norm": 2.0564282685963278, + "learning_rate": 7.550759019513173e-07, + "loss": 0.1122, + "step": 9522 + }, + { + "epoch": 2.1667804323094426, + "grad_norm": 1.4509585231688706, + "learning_rate": 7.54988505822095e-07, + "loss": 0.0771, + "step": 9523 + }, + { + "epoch": 2.1670079635949944, + "grad_norm": 1.4519579808787662, + "learning_rate": 7.549011070361459e-07, + "loss": 0.1164, + "step": 9524 + }, + { + "epoch": 2.167235494880546, + "grad_norm": 1.6549730773752167, + "learning_rate": 7.548137055952559e-07, + "loss": 0.1115, + "step": 9525 + }, + { + "epoch": 2.167463026166098, + "grad_norm": 2.924801535098329, + "learning_rate": 7.547263015012116e-07, + "loss": 0.0509, + "step": 9526 + }, + { + "epoch": 2.1676905574516496, + "grad_norm": 2.1844739743189523, + "learning_rate": 7.54638894755799e-07, + "loss": 0.1027, + "step": 9527 + }, + { + "epoch": 2.1679180887372014, + "grad_norm": 1.3046122618011156, + "learning_rate": 7.54551485360805e-07, + "loss": 0.0093, + "step": 9528 + }, + { + "epoch": 2.168145620022753, + "grad_norm": 1.856004275183145, + "learning_rate": 7.544640733180161e-07, + "loss": 0.0433, + "step": 9529 + }, + { + "epoch": 2.168373151308305, + "grad_norm": 1.451278219696552, + "learning_rate": 7.543766586292185e-07, + "loss": 0.0844, + "step": 9530 + }, + { + "epoch": 2.1686006825938566, + "grad_norm": 1.594240571498456, + "learning_rate": 7.542892412961988e-07, + "loss": 0.0425, + "step": 9531 + }, + { + "epoch": 2.1688282138794084, + "grad_norm": 1.1423295038645647, + "learning_rate": 7.54201821320744e-07, + "loss": 0.037, + "step": 9532 + }, + { + "epoch": 2.16905574516496, + "grad_norm": 1.558015800490758, + "learning_rate": 7.541143987046406e-07, + "loss": 0.0509, + "step": 9533 + }, + { + "epoch": 2.169283276450512, + "grad_norm": 1.59035521399315, + "learning_rate": 7.540269734496751e-07, + "loss": 0.0237, + "step": 9534 + }, + { + "epoch": 2.1695108077360636, + "grad_norm": 2.0807067699266133, + "learning_rate": 7.539395455576348e-07, + "loss": 0.0415, + "step": 9535 + }, + { + "epoch": 2.1697383390216154, + "grad_norm": 2.067919182776939, + "learning_rate": 7.538521150303063e-07, + "loss": 0.1002, + "step": 9536 + }, + { + "epoch": 2.169965870307167, + "grad_norm": 1.9992834288059516, + "learning_rate": 7.537646818694764e-07, + "loss": 0.1106, + "step": 9537 + }, + { + "epoch": 2.170193401592719, + "grad_norm": 1.4123785047184512, + "learning_rate": 7.536772460769324e-07, + "loss": 0.0233, + "step": 9538 + }, + { + "epoch": 2.1704209328782706, + "grad_norm": 1.665422220667401, + "learning_rate": 7.535898076544611e-07, + "loss": 0.0888, + "step": 9539 + }, + { + "epoch": 2.1706484641638224, + "grad_norm": 1.5148200586824807, + "learning_rate": 7.535023666038497e-07, + "loss": 0.0309, + "step": 9540 + }, + { + "epoch": 2.170875995449374, + "grad_norm": 1.482270971190226, + "learning_rate": 7.534149229268852e-07, + "loss": 0.0813, + "step": 9541 + }, + { + "epoch": 2.171103526734926, + "grad_norm": 1.291244523440277, + "learning_rate": 7.533274766253548e-07, + "loss": 0.0724, + "step": 9542 + }, + { + "epoch": 2.1713310580204777, + "grad_norm": 2.126739299406321, + "learning_rate": 7.532400277010458e-07, + "loss": 0.0552, + "step": 9543 + }, + { + "epoch": 2.1715585893060294, + "grad_norm": 1.6666769692922412, + "learning_rate": 7.531525761557454e-07, + "loss": 0.0782, + "step": 9544 + }, + { + "epoch": 2.171786120591581, + "grad_norm": 1.6514959116034478, + "learning_rate": 7.530651219912413e-07, + "loss": 0.0491, + "step": 9545 + }, + { + "epoch": 2.172013651877133, + "grad_norm": 2.1927974860136437, + "learning_rate": 7.529776652093204e-07, + "loss": 0.0727, + "step": 9546 + }, + { + "epoch": 2.1722411831626847, + "grad_norm": 1.9098009631719026, + "learning_rate": 7.528902058117707e-07, + "loss": 0.0791, + "step": 9547 + }, + { + "epoch": 2.172468714448237, + "grad_norm": 1.3068065250915974, + "learning_rate": 7.528027438003792e-07, + "loss": 0.1249, + "step": 9548 + }, + { + "epoch": 2.172696245733788, + "grad_norm": 2.503945304655778, + "learning_rate": 7.527152791769338e-07, + "loss": 0.1367, + "step": 9549 + }, + { + "epoch": 2.1729237770193404, + "grad_norm": 1.991208013783355, + "learning_rate": 7.526278119432219e-07, + "loss": 0.0383, + "step": 9550 + }, + { + "epoch": 2.173151308304892, + "grad_norm": 1.6803154280209256, + "learning_rate": 7.52540342101031e-07, + "loss": 0.0183, + "step": 9551 + }, + { + "epoch": 2.173378839590444, + "grad_norm": 1.7516963775448646, + "learning_rate": 7.524528696521495e-07, + "loss": 0.0816, + "step": 9552 + }, + { + "epoch": 2.1736063708759956, + "grad_norm": 1.257819638263749, + "learning_rate": 7.523653945983645e-07, + "loss": 0.0534, + "step": 9553 + }, + { + "epoch": 2.1738339021615474, + "grad_norm": 0.8986766894755777, + "learning_rate": 7.522779169414643e-07, + "loss": 0.0686, + "step": 9554 + }, + { + "epoch": 2.174061433447099, + "grad_norm": 2.0461962583918147, + "learning_rate": 7.521904366832365e-07, + "loss": 0.1134, + "step": 9555 + }, + { + "epoch": 2.174288964732651, + "grad_norm": 2.114234582962323, + "learning_rate": 7.521029538254692e-07, + "loss": 0.0189, + "step": 9556 + }, + { + "epoch": 2.1745164960182026, + "grad_norm": 1.2764140322075164, + "learning_rate": 7.520154683699501e-07, + "loss": 0.0358, + "step": 9557 + }, + { + "epoch": 2.1747440273037544, + "grad_norm": 1.7779424078634884, + "learning_rate": 7.519279803184676e-07, + "loss": 0.0713, + "step": 9558 + }, + { + "epoch": 2.174971558589306, + "grad_norm": 2.405996992965834, + "learning_rate": 7.518404896728096e-07, + "loss": 0.1248, + "step": 9559 + }, + { + "epoch": 2.175199089874858, + "grad_norm": 2.1106260889167503, + "learning_rate": 7.517529964347642e-07, + "loss": 0.0597, + "step": 9560 + }, + { + "epoch": 2.1754266211604096, + "grad_norm": 1.8519619466574628, + "learning_rate": 7.516655006061198e-07, + "loss": 0.1083, + "step": 9561 + }, + { + "epoch": 2.1756541524459614, + "grad_norm": 1.2364667323648664, + "learning_rate": 7.515780021886646e-07, + "loss": 0.0205, + "step": 9562 + }, + { + "epoch": 2.175881683731513, + "grad_norm": 1.3925058952327758, + "learning_rate": 7.514905011841867e-07, + "loss": 0.0338, + "step": 9563 + }, + { + "epoch": 2.176109215017065, + "grad_norm": 1.3204649550810612, + "learning_rate": 7.514029975944746e-07, + "loss": 0.0548, + "step": 9564 + }, + { + "epoch": 2.1763367463026166, + "grad_norm": 1.711066372302418, + "learning_rate": 7.513154914213168e-07, + "loss": 0.104, + "step": 9565 + }, + { + "epoch": 2.1765642775881684, + "grad_norm": 1.8364607608657266, + "learning_rate": 7.512279826665018e-07, + "loss": 0.065, + "step": 9566 + }, + { + "epoch": 2.17679180887372, + "grad_norm": 1.7375124140504703, + "learning_rate": 7.511404713318178e-07, + "loss": 0.087, + "step": 9567 + }, + { + "epoch": 2.177019340159272, + "grad_norm": 1.1281803318894696, + "learning_rate": 7.510529574190537e-07, + "loss": 0.1356, + "step": 9568 + }, + { + "epoch": 2.1772468714448237, + "grad_norm": 1.6726980185155025, + "learning_rate": 7.50965440929998e-07, + "loss": 0.0254, + "step": 9569 + }, + { + "epoch": 2.1774744027303754, + "grad_norm": 2.3018997634205527, + "learning_rate": 7.508779218664395e-07, + "loss": 0.0494, + "step": 9570 + }, + { + "epoch": 2.177701934015927, + "grad_norm": 1.7798912942310836, + "learning_rate": 7.507904002301665e-07, + "loss": 0.1451, + "step": 9571 + }, + { + "epoch": 2.177929465301479, + "grad_norm": 1.3289332852047717, + "learning_rate": 7.507028760229683e-07, + "loss": 0.0204, + "step": 9572 + }, + { + "epoch": 2.1781569965870307, + "grad_norm": 2.1303411735158977, + "learning_rate": 7.506153492466337e-07, + "loss": 0.0386, + "step": 9573 + }, + { + "epoch": 2.1783845278725824, + "grad_norm": 1.664882393547048, + "learning_rate": 7.505278199029511e-07, + "loss": 0.0678, + "step": 9574 + }, + { + "epoch": 2.178612059158134, + "grad_norm": 1.7529702135397573, + "learning_rate": 7.504402879937098e-07, + "loss": 0.0297, + "step": 9575 + }, + { + "epoch": 2.178839590443686, + "grad_norm": 1.2339158207363576, + "learning_rate": 7.503527535206989e-07, + "loss": 0.0203, + "step": 9576 + }, + { + "epoch": 2.1790671217292377, + "grad_norm": 1.3485489170953888, + "learning_rate": 7.502652164857072e-07, + "loss": 0.0769, + "step": 9577 + }, + { + "epoch": 2.1792946530147894, + "grad_norm": 1.3897911338084257, + "learning_rate": 7.501776768905238e-07, + "loss": 0.1447, + "step": 9578 + }, + { + "epoch": 2.179522184300341, + "grad_norm": 1.6876611981736775, + "learning_rate": 7.500901347369382e-07, + "loss": 0.0485, + "step": 9579 + }, + { + "epoch": 2.179749715585893, + "grad_norm": 1.0360630778350237, + "learning_rate": 7.500025900267391e-07, + "loss": 0.0107, + "step": 9580 + }, + { + "epoch": 2.1799772468714447, + "grad_norm": 1.2384392901934855, + "learning_rate": 7.49915042761716e-07, + "loss": 0.0607, + "step": 9581 + }, + { + "epoch": 2.1802047781569964, + "grad_norm": 2.594760956019517, + "learning_rate": 7.498274929436583e-07, + "loss": 0.0766, + "step": 9582 + }, + { + "epoch": 2.180432309442548, + "grad_norm": 2.04775190126401, + "learning_rate": 7.497399405743552e-07, + "loss": 0.0527, + "step": 9583 + }, + { + "epoch": 2.1806598407281, + "grad_norm": 2.3937978714203854, + "learning_rate": 7.49652385655596e-07, + "loss": 0.0353, + "step": 9584 + }, + { + "epoch": 2.1808873720136517, + "grad_norm": 1.8812082815224427, + "learning_rate": 7.495648281891707e-07, + "loss": 0.0713, + "step": 9585 + }, + { + "epoch": 2.1811149032992034, + "grad_norm": 2.6575784223226364, + "learning_rate": 7.494772681768683e-07, + "loss": 0.0716, + "step": 9586 + }, + { + "epoch": 2.1813424345847556, + "grad_norm": 1.165442203890328, + "learning_rate": 7.493897056204784e-07, + "loss": 0.0208, + "step": 9587 + }, + { + "epoch": 2.181569965870307, + "grad_norm": 1.371548163191165, + "learning_rate": 7.493021405217907e-07, + "loss": 0.0672, + "step": 9588 + }, + { + "epoch": 2.181797497155859, + "grad_norm": 1.6820420633054833, + "learning_rate": 7.49214572882595e-07, + "loss": 0.0641, + "step": 9589 + }, + { + "epoch": 2.182025028441411, + "grad_norm": 2.7580081413468744, + "learning_rate": 7.49127002704681e-07, + "loss": 0.0575, + "step": 9590 + }, + { + "epoch": 2.1822525597269626, + "grad_norm": 0.8828889868952942, + "learning_rate": 7.490394299898382e-07, + "loss": 0.0146, + "step": 9591 + }, + { + "epoch": 2.1824800910125144, + "grad_norm": 1.2354119944929676, + "learning_rate": 7.489518547398568e-07, + "loss": 0.0524, + "step": 9592 + }, + { + "epoch": 2.182707622298066, + "grad_norm": 1.8210810705906149, + "learning_rate": 7.488642769565264e-07, + "loss": 0.0098, + "step": 9593 + }, + { + "epoch": 2.182935153583618, + "grad_norm": 2.141462464205735, + "learning_rate": 7.48776696641637e-07, + "loss": 0.0781, + "step": 9594 + }, + { + "epoch": 2.1831626848691696, + "grad_norm": 1.8866884388185365, + "learning_rate": 7.486891137969786e-07, + "loss": 0.1051, + "step": 9595 + }, + { + "epoch": 2.1833902161547214, + "grad_norm": 1.2232011243579046, + "learning_rate": 7.486015284243413e-07, + "loss": 0.0779, + "step": 9596 + }, + { + "epoch": 2.183617747440273, + "grad_norm": 1.3069532595453608, + "learning_rate": 7.485139405255151e-07, + "loss": 0.0205, + "step": 9597 + }, + { + "epoch": 2.183845278725825, + "grad_norm": 1.7750761019098342, + "learning_rate": 7.484263501022902e-07, + "loss": 0.0922, + "step": 9598 + }, + { + "epoch": 2.1840728100113767, + "grad_norm": 2.067377867721793, + "learning_rate": 7.483387571564567e-07, + "loss": 0.0567, + "step": 9599 + }, + { + "epoch": 2.1843003412969284, + "grad_norm": 1.0342518095213404, + "learning_rate": 7.48251161689805e-07, + "loss": 0.0337, + "step": 9600 + }, + { + "epoch": 2.18452787258248, + "grad_norm": 1.7104776462824303, + "learning_rate": 7.48163563704125e-07, + "loss": 0.0797, + "step": 9601 + }, + { + "epoch": 2.184755403868032, + "grad_norm": 2.8112183343579664, + "learning_rate": 7.480759632012074e-07, + "loss": 0.0406, + "step": 9602 + }, + { + "epoch": 2.1849829351535837, + "grad_norm": 2.4812297741799116, + "learning_rate": 7.479883601828428e-07, + "loss": 0.0405, + "step": 9603 + }, + { + "epoch": 2.1852104664391354, + "grad_norm": 1.2978523279780194, + "learning_rate": 7.479007546508211e-07, + "loss": 0.0556, + "step": 9604 + }, + { + "epoch": 2.185437997724687, + "grad_norm": 1.4854935371422038, + "learning_rate": 7.478131466069331e-07, + "loss": 0.0575, + "step": 9605 + }, + { + "epoch": 2.185665529010239, + "grad_norm": 1.6313942260549827, + "learning_rate": 7.477255360529695e-07, + "loss": 0.1904, + "step": 9606 + }, + { + "epoch": 2.1858930602957907, + "grad_norm": 1.216152900468775, + "learning_rate": 7.476379229907205e-07, + "loss": 0.0696, + "step": 9607 + }, + { + "epoch": 2.1861205915813424, + "grad_norm": 0.6737250650335149, + "learning_rate": 7.475503074219769e-07, + "loss": 0.0123, + "step": 9608 + }, + { + "epoch": 2.186348122866894, + "grad_norm": 1.422171400837565, + "learning_rate": 7.474626893485295e-07, + "loss": 0.0241, + "step": 9609 + }, + { + "epoch": 2.186575654152446, + "grad_norm": 1.7517935884683415, + "learning_rate": 7.473750687721692e-07, + "loss": 0.0414, + "step": 9610 + }, + { + "epoch": 2.1868031854379977, + "grad_norm": 1.5468806772704229, + "learning_rate": 7.472874456946865e-07, + "loss": 0.0736, + "step": 9611 + }, + { + "epoch": 2.1870307167235494, + "grad_norm": 2.435102851240498, + "learning_rate": 7.471998201178724e-07, + "loss": 0.0396, + "step": 9612 + }, + { + "epoch": 2.187258248009101, + "grad_norm": 0.9156841203689913, + "learning_rate": 7.471121920435176e-07, + "loss": 0.013, + "step": 9613 + }, + { + "epoch": 2.187485779294653, + "grad_norm": 1.3462951849139937, + "learning_rate": 7.470245614734132e-07, + "loss": 0.0366, + "step": 9614 + }, + { + "epoch": 2.1877133105802047, + "grad_norm": 2.113392361520175, + "learning_rate": 7.469369284093504e-07, + "loss": 0.074, + "step": 9615 + }, + { + "epoch": 2.1879408418657564, + "grad_norm": 1.6015486857455614, + "learning_rate": 7.468492928531201e-07, + "loss": 0.0441, + "step": 9616 + }, + { + "epoch": 2.188168373151308, + "grad_norm": 1.8594041276412403, + "learning_rate": 7.467616548065134e-07, + "loss": 0.0563, + "step": 9617 + }, + { + "epoch": 2.18839590443686, + "grad_norm": 1.479593231636543, + "learning_rate": 7.466740142713217e-07, + "loss": 0.0785, + "step": 9618 + }, + { + "epoch": 2.1886234357224117, + "grad_norm": 1.5060560248711237, + "learning_rate": 7.465863712493357e-07, + "loss": 0.069, + "step": 9619 + }, + { + "epoch": 2.1888509670079634, + "grad_norm": 1.9353817486116598, + "learning_rate": 7.46498725742347e-07, + "loss": 0.0429, + "step": 9620 + }, + { + "epoch": 2.189078498293515, + "grad_norm": 1.9847954336755138, + "learning_rate": 7.464110777521467e-07, + "loss": 0.0418, + "step": 9621 + }, + { + "epoch": 2.189306029579067, + "grad_norm": 1.7208996320826648, + "learning_rate": 7.463234272805265e-07, + "loss": 0.0688, + "step": 9622 + }, + { + "epoch": 2.1895335608646187, + "grad_norm": 0.9602382872558913, + "learning_rate": 7.462357743292778e-07, + "loss": 0.0472, + "step": 9623 + }, + { + "epoch": 2.1897610921501705, + "grad_norm": 1.7044560878181323, + "learning_rate": 7.461481189001917e-07, + "loss": 0.0338, + "step": 9624 + }, + { + "epoch": 2.189988623435722, + "grad_norm": 1.2760275994057422, + "learning_rate": 7.460604609950599e-07, + "loss": 0.0439, + "step": 9625 + }, + { + "epoch": 2.1902161547212744, + "grad_norm": 1.6892058616447903, + "learning_rate": 7.459728006156741e-07, + "loss": 0.0128, + "step": 9626 + }, + { + "epoch": 2.1904436860068257, + "grad_norm": 2.0382940083903316, + "learning_rate": 7.458851377638257e-07, + "loss": 0.1025, + "step": 9627 + }, + { + "epoch": 2.190671217292378, + "grad_norm": 1.068888687313207, + "learning_rate": 7.457974724413065e-07, + "loss": 0.0652, + "step": 9628 + }, + { + "epoch": 2.1908987485779297, + "grad_norm": 1.3272239815035083, + "learning_rate": 7.457098046499084e-07, + "loss": 0.0216, + "step": 9629 + }, + { + "epoch": 2.1911262798634814, + "grad_norm": 3.1492661389940086, + "learning_rate": 7.456221343914228e-07, + "loss": 0.1145, + "step": 9630 + }, + { + "epoch": 2.191353811149033, + "grad_norm": 2.5980285749900074, + "learning_rate": 7.455344616676415e-07, + "loss": 0.0466, + "step": 9631 + }, + { + "epoch": 2.191581342434585, + "grad_norm": 0.9492209432938257, + "learning_rate": 7.454467864803567e-07, + "loss": 0.0564, + "step": 9632 + }, + { + "epoch": 2.1918088737201367, + "grad_norm": 0.9923431282879376, + "learning_rate": 7.453591088313603e-07, + "loss": 0.0216, + "step": 9633 + }, + { + "epoch": 2.1920364050056884, + "grad_norm": 1.8548331462860232, + "learning_rate": 7.45271428722444e-07, + "loss": 0.046, + "step": 9634 + }, + { + "epoch": 2.19226393629124, + "grad_norm": 2.08701457162056, + "learning_rate": 7.451837461553998e-07, + "loss": 0.0339, + "step": 9635 + }, + { + "epoch": 2.192491467576792, + "grad_norm": 1.6404266103350365, + "learning_rate": 7.450960611320204e-07, + "loss": 0.0795, + "step": 9636 + }, + { + "epoch": 2.1927189988623437, + "grad_norm": 1.7091531647594718, + "learning_rate": 7.450083736540972e-07, + "loss": 0.0306, + "step": 9637 + }, + { + "epoch": 2.1929465301478954, + "grad_norm": 1.2599120065586638, + "learning_rate": 7.449206837234224e-07, + "loss": 0.0403, + "step": 9638 + }, + { + "epoch": 2.193174061433447, + "grad_norm": 3.36104823873805, + "learning_rate": 7.448329913417887e-07, + "loss": 0.014, + "step": 9639 + }, + { + "epoch": 2.193401592718999, + "grad_norm": 1.0454678044508545, + "learning_rate": 7.447452965109882e-07, + "loss": 0.0705, + "step": 9640 + }, + { + "epoch": 2.1936291240045507, + "grad_norm": 1.2907620588951814, + "learning_rate": 7.446575992328128e-07, + "loss": 0.0724, + "step": 9641 + }, + { + "epoch": 2.1938566552901024, + "grad_norm": 6.017704364504656, + "learning_rate": 7.445698995090557e-07, + "loss": 0.0277, + "step": 9642 + }, + { + "epoch": 2.194084186575654, + "grad_norm": 1.486887890857098, + "learning_rate": 7.444821973415086e-07, + "loss": 0.0952, + "step": 9643 + }, + { + "epoch": 2.194311717861206, + "grad_norm": 1.145657339951393, + "learning_rate": 7.443944927319641e-07, + "loss": 0.0685, + "step": 9644 + }, + { + "epoch": 2.1945392491467577, + "grad_norm": 1.3397259790508476, + "learning_rate": 7.443067856822149e-07, + "loss": 0.0506, + "step": 9645 + }, + { + "epoch": 2.1947667804323094, + "grad_norm": 1.7411937518722584, + "learning_rate": 7.442190761940535e-07, + "loss": 0.0531, + "step": 9646 + }, + { + "epoch": 2.194994311717861, + "grad_norm": 1.192381975108253, + "learning_rate": 7.441313642692726e-07, + "loss": 0.0218, + "step": 9647 + }, + { + "epoch": 2.195221843003413, + "grad_norm": 1.6976397972723314, + "learning_rate": 7.440436499096647e-07, + "loss": 0.0816, + "step": 9648 + }, + { + "epoch": 2.1954493742889647, + "grad_norm": 1.6314851417235372, + "learning_rate": 7.439559331170226e-07, + "loss": 0.0651, + "step": 9649 + }, + { + "epoch": 2.1956769055745164, + "grad_norm": 1.2445680459975024, + "learning_rate": 7.438682138931393e-07, + "loss": 0.0537, + "step": 9650 + }, + { + "epoch": 2.195904436860068, + "grad_norm": 1.2745143884595345, + "learning_rate": 7.43780492239807e-07, + "loss": 0.0983, + "step": 9651 + }, + { + "epoch": 2.19613196814562, + "grad_norm": 1.8918216438837654, + "learning_rate": 7.436927681588192e-07, + "loss": 0.0729, + "step": 9652 + }, + { + "epoch": 2.1963594994311717, + "grad_norm": 1.4902332980583612, + "learning_rate": 7.436050416519687e-07, + "loss": 0.1101, + "step": 9653 + }, + { + "epoch": 2.1965870307167235, + "grad_norm": 1.698538000206574, + "learning_rate": 7.435173127210482e-07, + "loss": 0.0906, + "step": 9654 + }, + { + "epoch": 2.196814562002275, + "grad_norm": 2.441854587325201, + "learning_rate": 7.43429581367851e-07, + "loss": 0.0519, + "step": 9655 + }, + { + "epoch": 2.197042093287827, + "grad_norm": 2.0217096794105065, + "learning_rate": 7.4334184759417e-07, + "loss": 0.0802, + "step": 9656 + }, + { + "epoch": 2.1972696245733787, + "grad_norm": 0.5667236791158886, + "learning_rate": 7.432541114017984e-07, + "loss": 0.009, + "step": 9657 + }, + { + "epoch": 2.1974971558589305, + "grad_norm": 1.2024057232721754, + "learning_rate": 7.431663727925293e-07, + "loss": 0.0705, + "step": 9658 + }, + { + "epoch": 2.197724687144482, + "grad_norm": 1.4816985098359061, + "learning_rate": 7.430786317681559e-07, + "loss": 0.0331, + "step": 9659 + }, + { + "epoch": 2.197952218430034, + "grad_norm": 1.2982682390365592, + "learning_rate": 7.429908883304716e-07, + "loss": 0.1017, + "step": 9660 + }, + { + "epoch": 2.1981797497155857, + "grad_norm": 1.479198664737689, + "learning_rate": 7.429031424812697e-07, + "loss": 0.0379, + "step": 9661 + }, + { + "epoch": 2.1984072810011375, + "grad_norm": 2.452143558116467, + "learning_rate": 7.428153942223433e-07, + "loss": 0.0716, + "step": 9662 + }, + { + "epoch": 2.198634812286689, + "grad_norm": 1.8634538906957014, + "learning_rate": 7.427276435554861e-07, + "loss": 0.0993, + "step": 9663 + }, + { + "epoch": 2.198862343572241, + "grad_norm": 1.9471649792562606, + "learning_rate": 7.426398904824916e-07, + "loss": 0.1173, + "step": 9664 + }, + { + "epoch": 2.199089874857793, + "grad_norm": 2.1771342194370216, + "learning_rate": 7.425521350051529e-07, + "loss": 0.1255, + "step": 9665 + }, + { + "epoch": 2.1993174061433445, + "grad_norm": 2.055664142746377, + "learning_rate": 7.42464377125264e-07, + "loss": 0.1055, + "step": 9666 + }, + { + "epoch": 2.1995449374288967, + "grad_norm": 0.56759498734142, + "learning_rate": 7.423766168446187e-07, + "loss": 0.007, + "step": 9667 + }, + { + "epoch": 2.1997724687144484, + "grad_norm": 2.608154903722345, + "learning_rate": 7.422888541650097e-07, + "loss": 0.0689, + "step": 9668 + }, + { + "epoch": 2.2, + "grad_norm": 1.7413005487049034, + "learning_rate": 7.422010890882317e-07, + "loss": 0.0467, + "step": 9669 + }, + { + "epoch": 2.200227531285552, + "grad_norm": 1.0901352604030778, + "learning_rate": 7.421133216160781e-07, + "loss": 0.0136, + "step": 9670 + }, + { + "epoch": 2.2004550625711037, + "grad_norm": 1.4326097632701875, + "learning_rate": 7.420255517503424e-07, + "loss": 0.0294, + "step": 9671 + }, + { + "epoch": 2.2006825938566554, + "grad_norm": 1.7809393341390753, + "learning_rate": 7.41937779492819e-07, + "loss": 0.0249, + "step": 9672 + }, + { + "epoch": 2.200910125142207, + "grad_norm": 1.6838160667034856, + "learning_rate": 7.418500048453016e-07, + "loss": 0.0326, + "step": 9673 + }, + { + "epoch": 2.201137656427759, + "grad_norm": 1.3809458996390551, + "learning_rate": 7.417622278095838e-07, + "loss": 0.0794, + "step": 9674 + }, + { + "epoch": 2.2013651877133107, + "grad_norm": 2.4547422069992657, + "learning_rate": 7.416744483874602e-07, + "loss": 0.0687, + "step": 9675 + }, + { + "epoch": 2.2015927189988624, + "grad_norm": 1.0275726858368945, + "learning_rate": 7.415866665807245e-07, + "loss": 0.0722, + "step": 9676 + }, + { + "epoch": 2.201820250284414, + "grad_norm": 1.2538058634007505, + "learning_rate": 7.414988823911708e-07, + "loss": 0.1062, + "step": 9677 + }, + { + "epoch": 2.202047781569966, + "grad_norm": 1.1193489447576097, + "learning_rate": 7.41411095820593e-07, + "loss": 0.0275, + "step": 9678 + }, + { + "epoch": 2.2022753128555177, + "grad_norm": 1.6408552393319376, + "learning_rate": 7.41323306870786e-07, + "loss": 0.1016, + "step": 9679 + }, + { + "epoch": 2.2025028441410694, + "grad_norm": 1.474030368936237, + "learning_rate": 7.412355155435437e-07, + "loss": 0.0171, + "step": 9680 + }, + { + "epoch": 2.202730375426621, + "grad_norm": 1.6839679677626955, + "learning_rate": 7.411477218406602e-07, + "loss": 0.0297, + "step": 9681 + }, + { + "epoch": 2.202957906712173, + "grad_norm": 1.7100779752934014, + "learning_rate": 7.410599257639299e-07, + "loss": 0.0816, + "step": 9682 + }, + { + "epoch": 2.2031854379977247, + "grad_norm": 1.7401050895523356, + "learning_rate": 7.409721273151473e-07, + "loss": 0.0679, + "step": 9683 + }, + { + "epoch": 2.2034129692832765, + "grad_norm": 1.0215883331858133, + "learning_rate": 7.408843264961068e-07, + "loss": 0.0356, + "step": 9684 + }, + { + "epoch": 2.203640500568828, + "grad_norm": 1.85268010108362, + "learning_rate": 7.407965233086029e-07, + "loss": 0.1243, + "step": 9685 + }, + { + "epoch": 2.20386803185438, + "grad_norm": 2.511210838270463, + "learning_rate": 7.407087177544304e-07, + "loss": 0.0713, + "step": 9686 + }, + { + "epoch": 2.2040955631399317, + "grad_norm": 2.284806877904906, + "learning_rate": 7.406209098353834e-07, + "loss": 0.072, + "step": 9687 + }, + { + "epoch": 2.2043230944254835, + "grad_norm": 1.3709918922264874, + "learning_rate": 7.405330995532566e-07, + "loss": 0.1123, + "step": 9688 + }, + { + "epoch": 2.204550625711035, + "grad_norm": 1.7365254377542252, + "learning_rate": 7.40445286909845e-07, + "loss": 0.0289, + "step": 9689 + }, + { + "epoch": 2.204778156996587, + "grad_norm": 2.709186003181355, + "learning_rate": 7.403574719069431e-07, + "loss": 0.0781, + "step": 9690 + }, + { + "epoch": 2.2050056882821387, + "grad_norm": 1.6901995066750213, + "learning_rate": 7.402696545463458e-07, + "loss": 0.0449, + "step": 9691 + }, + { + "epoch": 2.2052332195676905, + "grad_norm": 1.0710209839388836, + "learning_rate": 7.401818348298478e-07, + "loss": 0.0376, + "step": 9692 + }, + { + "epoch": 2.2054607508532422, + "grad_norm": 1.7365012428305326, + "learning_rate": 7.400940127592441e-07, + "loss": 0.0298, + "step": 9693 + }, + { + "epoch": 2.205688282138794, + "grad_norm": 1.4249477504420347, + "learning_rate": 7.400061883363297e-07, + "loss": 0.0229, + "step": 9694 + }, + { + "epoch": 2.2059158134243457, + "grad_norm": 2.0061713425028365, + "learning_rate": 7.399183615628991e-07, + "loss": 0.0284, + "step": 9695 + }, + { + "epoch": 2.2061433447098975, + "grad_norm": 2.028487499424723, + "learning_rate": 7.398305324407479e-07, + "loss": 0.0797, + "step": 9696 + }, + { + "epoch": 2.2063708759954492, + "grad_norm": 0.8043899684970321, + "learning_rate": 7.397427009716709e-07, + "loss": 0.0031, + "step": 9697 + }, + { + "epoch": 2.206598407281001, + "grad_norm": 2.1450208855852666, + "learning_rate": 7.396548671574632e-07, + "loss": 0.0562, + "step": 9698 + }, + { + "epoch": 2.2068259385665527, + "grad_norm": 1.1377903816480621, + "learning_rate": 7.395670309999201e-07, + "loss": 0.0342, + "step": 9699 + }, + { + "epoch": 2.2070534698521045, + "grad_norm": 1.8889917881251588, + "learning_rate": 7.394791925008366e-07, + "loss": 0.023, + "step": 9700 + }, + { + "epoch": 2.2072810011376562, + "grad_norm": 1.4591739416194378, + "learning_rate": 7.39391351662008e-07, + "loss": 0.0581, + "step": 9701 + }, + { + "epoch": 2.207508532423208, + "grad_norm": 2.1030733777122066, + "learning_rate": 7.393035084852296e-07, + "loss": 0.0693, + "step": 9702 + }, + { + "epoch": 2.2077360637087597, + "grad_norm": 1.3391074219072032, + "learning_rate": 7.39215662972297e-07, + "loss": 0.0739, + "step": 9703 + }, + { + "epoch": 2.207963594994312, + "grad_norm": 1.8189202859693807, + "learning_rate": 7.391278151250053e-07, + "loss": 0.0673, + "step": 9704 + }, + { + "epoch": 2.2081911262798632, + "grad_norm": 1.8331683339714793, + "learning_rate": 7.390399649451501e-07, + "loss": 0.0411, + "step": 9705 + }, + { + "epoch": 2.2084186575654154, + "grad_norm": 2.013278806253988, + "learning_rate": 7.389521124345271e-07, + "loss": 0.0931, + "step": 9706 + }, + { + "epoch": 2.208646188850967, + "grad_norm": 1.5027122811390399, + "learning_rate": 7.388642575949315e-07, + "loss": 0.0791, + "step": 9707 + }, + { + "epoch": 2.208873720136519, + "grad_norm": 2.2217244601224015, + "learning_rate": 7.387764004281588e-07, + "loss": 0.0684, + "step": 9708 + }, + { + "epoch": 2.2091012514220707, + "grad_norm": 0.5824954666368191, + "learning_rate": 7.38688540936005e-07, + "loss": 0.0079, + "step": 9709 + }, + { + "epoch": 2.2093287827076225, + "grad_norm": 1.642309980878137, + "learning_rate": 7.386006791202656e-07, + "loss": 0.078, + "step": 9710 + }, + { + "epoch": 2.209556313993174, + "grad_norm": 1.86901387440969, + "learning_rate": 7.385128149827364e-07, + "loss": 0.1119, + "step": 9711 + }, + { + "epoch": 2.209783845278726, + "grad_norm": 1.4169879392035163, + "learning_rate": 7.384249485252132e-07, + "loss": 0.0482, + "step": 9712 + }, + { + "epoch": 2.2100113765642777, + "grad_norm": 1.9100739304666876, + "learning_rate": 7.383370797494918e-07, + "loss": 0.0434, + "step": 9713 + }, + { + "epoch": 2.2102389078498295, + "grad_norm": 2.399990436019343, + "learning_rate": 7.382492086573679e-07, + "loss": 0.0694, + "step": 9714 + }, + { + "epoch": 2.210466439135381, + "grad_norm": 1.812050241754321, + "learning_rate": 7.381613352506376e-07, + "loss": 0.1388, + "step": 9715 + }, + { + "epoch": 2.210693970420933, + "grad_norm": 1.5722994482086958, + "learning_rate": 7.380734595310969e-07, + "loss": 0.0942, + "step": 9716 + }, + { + "epoch": 2.2109215017064847, + "grad_norm": 1.9028886715125728, + "learning_rate": 7.37985581500542e-07, + "loss": 0.038, + "step": 9717 + }, + { + "epoch": 2.2111490329920365, + "grad_norm": 1.6522125900547038, + "learning_rate": 7.378977011607684e-07, + "loss": 0.0784, + "step": 9718 + }, + { + "epoch": 2.211376564277588, + "grad_norm": 1.353064902067639, + "learning_rate": 7.378098185135727e-07, + "loss": 0.0707, + "step": 9719 + }, + { + "epoch": 2.21160409556314, + "grad_norm": 1.6426673688489113, + "learning_rate": 7.377219335607509e-07, + "loss": 0.0498, + "step": 9720 + }, + { + "epoch": 2.2118316268486917, + "grad_norm": 1.589301263115234, + "learning_rate": 7.376340463040993e-07, + "loss": 0.1053, + "step": 9721 + }, + { + "epoch": 2.2120591581342435, + "grad_norm": 1.9195414892558205, + "learning_rate": 7.375461567454138e-07, + "loss": 0.0299, + "step": 9722 + }, + { + "epoch": 2.2122866894197952, + "grad_norm": 1.9133285965390077, + "learning_rate": 7.374582648864912e-07, + "loss": 0.0623, + "step": 9723 + }, + { + "epoch": 2.212514220705347, + "grad_norm": 1.384011575068315, + "learning_rate": 7.373703707291277e-07, + "loss": 0.0974, + "step": 9724 + }, + { + "epoch": 2.2127417519908987, + "grad_norm": 1.2982385454864285, + "learning_rate": 7.372824742751194e-07, + "loss": 0.0605, + "step": 9725 + }, + { + "epoch": 2.2129692832764505, + "grad_norm": 1.7065357308487878, + "learning_rate": 7.37194575526263e-07, + "loss": 0.1313, + "step": 9726 + }, + { + "epoch": 2.2131968145620022, + "grad_norm": 1.374205231380851, + "learning_rate": 7.371066744843551e-07, + "loss": 0.0545, + "step": 9727 + }, + { + "epoch": 2.213424345847554, + "grad_norm": 1.689381913821839, + "learning_rate": 7.37018771151192e-07, + "loss": 0.0475, + "step": 9728 + }, + { + "epoch": 2.2136518771331057, + "grad_norm": 1.9299964339226994, + "learning_rate": 7.369308655285702e-07, + "loss": 0.0377, + "step": 9729 + }, + { + "epoch": 2.2138794084186575, + "grad_norm": 1.126734400343911, + "learning_rate": 7.368429576182869e-07, + "loss": 0.0373, + "step": 9730 + }, + { + "epoch": 2.2141069397042092, + "grad_norm": 2.1669924188436376, + "learning_rate": 7.367550474221381e-07, + "loss": 0.0428, + "step": 9731 + }, + { + "epoch": 2.214334470989761, + "grad_norm": 1.3739401667507931, + "learning_rate": 7.366671349419207e-07, + "loss": 0.0837, + "step": 9732 + }, + { + "epoch": 2.2145620022753127, + "grad_norm": 1.4806133911538504, + "learning_rate": 7.365792201794318e-07, + "loss": 0.0595, + "step": 9733 + }, + { + "epoch": 2.2147895335608645, + "grad_norm": 1.8386823423252745, + "learning_rate": 7.364913031364679e-07, + "loss": 0.0383, + "step": 9734 + }, + { + "epoch": 2.2150170648464163, + "grad_norm": 1.1325561103587034, + "learning_rate": 7.364033838148258e-07, + "loss": 0.1147, + "step": 9735 + }, + { + "epoch": 2.215244596131968, + "grad_norm": 1.5302834735325364, + "learning_rate": 7.363154622163029e-07, + "loss": 0.0474, + "step": 9736 + }, + { + "epoch": 2.2154721274175198, + "grad_norm": 0.9544798037125929, + "learning_rate": 7.362275383426956e-07, + "loss": 0.0199, + "step": 9737 + }, + { + "epoch": 2.2156996587030715, + "grad_norm": 1.5671573786227384, + "learning_rate": 7.361396121958011e-07, + "loss": 0.0574, + "step": 9738 + }, + { + "epoch": 2.2159271899886233, + "grad_norm": 0.8337324725004281, + "learning_rate": 7.360516837774165e-07, + "loss": 0.0488, + "step": 9739 + }, + { + "epoch": 2.216154721274175, + "grad_norm": 1.7156210086056303, + "learning_rate": 7.359637530893389e-07, + "loss": 0.0481, + "step": 9740 + }, + { + "epoch": 2.2163822525597268, + "grad_norm": 1.8430166484430528, + "learning_rate": 7.358758201333654e-07, + "loss": 0.1091, + "step": 9741 + }, + { + "epoch": 2.2166097838452785, + "grad_norm": 1.7462111277595131, + "learning_rate": 7.357878849112932e-07, + "loss": 0.1213, + "step": 9742 + }, + { + "epoch": 2.2168373151308307, + "grad_norm": 0.9610959896146984, + "learning_rate": 7.356999474249196e-07, + "loss": 0.0181, + "step": 9743 + }, + { + "epoch": 2.217064846416382, + "grad_norm": 2.0783946435552605, + "learning_rate": 7.356120076760417e-07, + "loss": 0.0559, + "step": 9744 + }, + { + "epoch": 2.217292377701934, + "grad_norm": 1.207894693989277, + "learning_rate": 7.35524065666457e-07, + "loss": 0.0635, + "step": 9745 + }, + { + "epoch": 2.217519908987486, + "grad_norm": 1.4518995713090144, + "learning_rate": 7.354361213979627e-07, + "loss": 0.102, + "step": 9746 + }, + { + "epoch": 2.2177474402730377, + "grad_norm": 2.0477258265422487, + "learning_rate": 7.353481748723565e-07, + "loss": 0.0964, + "step": 9747 + }, + { + "epoch": 2.2179749715585895, + "grad_norm": 1.5116308436141248, + "learning_rate": 7.352602260914356e-07, + "loss": 0.024, + "step": 9748 + }, + { + "epoch": 2.218202502844141, + "grad_norm": 1.1216515051005054, + "learning_rate": 7.351722750569976e-07, + "loss": 0.0968, + "step": 9749 + }, + { + "epoch": 2.218430034129693, + "grad_norm": 1.1057340754528662, + "learning_rate": 7.350843217708402e-07, + "loss": 0.0489, + "step": 9750 + }, + { + "epoch": 2.2186575654152447, + "grad_norm": 1.6239021685625052, + "learning_rate": 7.349963662347608e-07, + "loss": 0.0491, + "step": 9751 + }, + { + "epoch": 2.2188850967007965, + "grad_norm": 1.5597764106867107, + "learning_rate": 7.349084084505569e-07, + "loss": 0.0729, + "step": 9752 + }, + { + "epoch": 2.2191126279863482, + "grad_norm": 1.6577340113661923, + "learning_rate": 7.348204484200267e-07, + "loss": 0.0583, + "step": 9753 + }, + { + "epoch": 2.2193401592719, + "grad_norm": 1.4670468899875693, + "learning_rate": 7.347324861449676e-07, + "loss": 0.0628, + "step": 9754 + }, + { + "epoch": 2.2195676905574517, + "grad_norm": 1.272090927322844, + "learning_rate": 7.346445216271773e-07, + "loss": 0.0516, + "step": 9755 + }, + { + "epoch": 2.2197952218430035, + "grad_norm": 1.6793841175002284, + "learning_rate": 7.34556554868454e-07, + "loss": 0.0401, + "step": 9756 + }, + { + "epoch": 2.2200227531285552, + "grad_norm": 1.1523066967480309, + "learning_rate": 7.344685858705952e-07, + "loss": 0.0843, + "step": 9757 + }, + { + "epoch": 2.220250284414107, + "grad_norm": 1.2505093583404552, + "learning_rate": 7.343806146353991e-07, + "loss": 0.0474, + "step": 9758 + }, + { + "epoch": 2.2204778156996587, + "grad_norm": 1.5414448891525965, + "learning_rate": 7.342926411646634e-07, + "loss": 0.0872, + "step": 9759 + }, + { + "epoch": 2.2207053469852105, + "grad_norm": 1.1602468762362055, + "learning_rate": 7.342046654601864e-07, + "loss": 0.0455, + "step": 9760 + }, + { + "epoch": 2.2209328782707622, + "grad_norm": 1.6372674271227037, + "learning_rate": 7.341166875237661e-07, + "loss": 0.0357, + "step": 9761 + }, + { + "epoch": 2.221160409556314, + "grad_norm": 1.3260888967521394, + "learning_rate": 7.340287073572004e-07, + "loss": 0.072, + "step": 9762 + }, + { + "epoch": 2.2213879408418657, + "grad_norm": 2.102233929435973, + "learning_rate": 7.339407249622878e-07, + "loss": 0.0621, + "step": 9763 + }, + { + "epoch": 2.2216154721274175, + "grad_norm": 2.1229429716481145, + "learning_rate": 7.33852740340826e-07, + "loss": 0.0383, + "step": 9764 + }, + { + "epoch": 2.2218430034129693, + "grad_norm": 1.1120423681299842, + "learning_rate": 7.337647534946137e-07, + "loss": 0.0634, + "step": 9765 + }, + { + "epoch": 2.222070534698521, + "grad_norm": 1.2874234004712588, + "learning_rate": 7.33676764425449e-07, + "loss": 0.0163, + "step": 9766 + }, + { + "epoch": 2.2222980659840728, + "grad_norm": 1.6403793909671274, + "learning_rate": 7.335887731351303e-07, + "loss": 0.1012, + "step": 9767 + }, + { + "epoch": 2.2225255972696245, + "grad_norm": 1.4790826574202585, + "learning_rate": 7.33500779625456e-07, + "loss": 0.0641, + "step": 9768 + }, + { + "epoch": 2.2227531285551763, + "grad_norm": 1.831512684653163, + "learning_rate": 7.334127838982244e-07, + "loss": 0.0289, + "step": 9769 + }, + { + "epoch": 2.222980659840728, + "grad_norm": 1.1976119278530164, + "learning_rate": 7.333247859552343e-07, + "loss": 0.1138, + "step": 9770 + }, + { + "epoch": 2.2232081911262798, + "grad_norm": 1.42079083380393, + "learning_rate": 7.332367857982836e-07, + "loss": 0.0193, + "step": 9771 + }, + { + "epoch": 2.2234357224118315, + "grad_norm": 0.8551325789309228, + "learning_rate": 7.331487834291712e-07, + "loss": 0.0093, + "step": 9772 + }, + { + "epoch": 2.2236632536973833, + "grad_norm": 1.5245253810997095, + "learning_rate": 7.33060778849696e-07, + "loss": 0.064, + "step": 9773 + }, + { + "epoch": 2.223890784982935, + "grad_norm": 2.3375577939766847, + "learning_rate": 7.329727720616564e-07, + "loss": 0.0725, + "step": 9774 + }, + { + "epoch": 2.2241183162684868, + "grad_norm": 1.6557009780459495, + "learning_rate": 7.328847630668508e-07, + "loss": 0.1068, + "step": 9775 + }, + { + "epoch": 2.2243458475540385, + "grad_norm": 1.5981644410565476, + "learning_rate": 7.327967518670784e-07, + "loss": 0.1051, + "step": 9776 + }, + { + "epoch": 2.2245733788395903, + "grad_norm": 1.3488788647667986, + "learning_rate": 7.327087384641378e-07, + "loss": 0.0914, + "step": 9777 + }, + { + "epoch": 2.224800910125142, + "grad_norm": 2.0507278721482987, + "learning_rate": 7.326207228598278e-07, + "loss": 0.0369, + "step": 9778 + }, + { + "epoch": 2.225028441410694, + "grad_norm": 1.8708311122404728, + "learning_rate": 7.325327050559473e-07, + "loss": 0.1559, + "step": 9779 + }, + { + "epoch": 2.2252559726962455, + "grad_norm": 1.341988589812003, + "learning_rate": 7.324446850542954e-07, + "loss": 0.0766, + "step": 9780 + }, + { + "epoch": 2.2254835039817973, + "grad_norm": 1.654650910486216, + "learning_rate": 7.32356662856671e-07, + "loss": 0.0753, + "step": 9781 + }, + { + "epoch": 2.2257110352673495, + "grad_norm": 1.8807522907636354, + "learning_rate": 7.322686384648727e-07, + "loss": 0.0791, + "step": 9782 + }, + { + "epoch": 2.225938566552901, + "grad_norm": 1.5036944036892335, + "learning_rate": 7.321806118807e-07, + "loss": 0.035, + "step": 9783 + }, + { + "epoch": 2.226166097838453, + "grad_norm": 1.4208447880112791, + "learning_rate": 7.320925831059519e-07, + "loss": 0.0778, + "step": 9784 + }, + { + "epoch": 2.2263936291240047, + "grad_norm": 1.7838232242465537, + "learning_rate": 7.320045521424277e-07, + "loss": 0.0627, + "step": 9785 + }, + { + "epoch": 2.2266211604095565, + "grad_norm": 1.803026894813878, + "learning_rate": 7.31916518991926e-07, + "loss": 0.086, + "step": 9786 + }, + { + "epoch": 2.2268486916951082, + "grad_norm": 3.3696694661970064, + "learning_rate": 7.318284836562469e-07, + "loss": 0.1026, + "step": 9787 + }, + { + "epoch": 2.22707622298066, + "grad_norm": 2.4538791249337257, + "learning_rate": 7.31740446137189e-07, + "loss": 0.0489, + "step": 9788 + }, + { + "epoch": 2.2273037542662117, + "grad_norm": 1.3973711922977448, + "learning_rate": 7.31652406436552e-07, + "loss": 0.0868, + "step": 9789 + }, + { + "epoch": 2.2275312855517635, + "grad_norm": 1.0066511313590052, + "learning_rate": 7.315643645561351e-07, + "loss": 0.0544, + "step": 9790 + }, + { + "epoch": 2.2277588168373152, + "grad_norm": 1.0743552925218625, + "learning_rate": 7.314763204977376e-07, + "loss": 0.0299, + "step": 9791 + }, + { + "epoch": 2.227986348122867, + "grad_norm": 2.483939066987129, + "learning_rate": 7.313882742631594e-07, + "loss": 0.1224, + "step": 9792 + }, + { + "epoch": 2.2282138794084188, + "grad_norm": 2.3064273686519816, + "learning_rate": 7.313002258541994e-07, + "loss": 0.0925, + "step": 9793 + }, + { + "epoch": 2.2284414106939705, + "grad_norm": 1.2767050376656814, + "learning_rate": 7.312121752726577e-07, + "loss": 0.0508, + "step": 9794 + }, + { + "epoch": 2.2286689419795223, + "grad_norm": 1.0868279370785345, + "learning_rate": 7.311241225203336e-07, + "loss": 0.0548, + "step": 9795 + }, + { + "epoch": 2.228896473265074, + "grad_norm": 1.1202440281984138, + "learning_rate": 7.310360675990266e-07, + "loss": 0.0295, + "step": 9796 + }, + { + "epoch": 2.2291240045506258, + "grad_norm": 1.502226360489142, + "learning_rate": 7.309480105105368e-07, + "loss": 0.0578, + "step": 9797 + }, + { + "epoch": 2.2293515358361775, + "grad_norm": 1.4792232919585468, + "learning_rate": 7.308599512566636e-07, + "loss": 0.0384, + "step": 9798 + }, + { + "epoch": 2.2295790671217293, + "grad_norm": 1.3290497926242055, + "learning_rate": 7.30771889839207e-07, + "loss": 0.0159, + "step": 9799 + }, + { + "epoch": 2.229806598407281, + "grad_norm": 1.2580296624029728, + "learning_rate": 7.306838262599666e-07, + "loss": 0.0841, + "step": 9800 + }, + { + "epoch": 2.2300341296928328, + "grad_norm": 2.081060327990041, + "learning_rate": 7.305957605207423e-07, + "loss": 0.0494, + "step": 9801 + }, + { + "epoch": 2.2302616609783845, + "grad_norm": 2.374571970889199, + "learning_rate": 7.30507692623334e-07, + "loss": 0.0498, + "step": 9802 + }, + { + "epoch": 2.2304891922639363, + "grad_norm": 1.4096838482715424, + "learning_rate": 7.304196225695417e-07, + "loss": 0.033, + "step": 9803 + }, + { + "epoch": 2.230716723549488, + "grad_norm": 1.2938763635639914, + "learning_rate": 7.303315503611654e-07, + "loss": 0.0559, + "step": 9804 + }, + { + "epoch": 2.2309442548350398, + "grad_norm": 2.2710267938224122, + "learning_rate": 7.302434760000053e-07, + "loss": 0.1357, + "step": 9805 + }, + { + "epoch": 2.2311717861205915, + "grad_norm": 1.2615744956314672, + "learning_rate": 7.301553994878613e-07, + "loss": 0.035, + "step": 9806 + }, + { + "epoch": 2.2313993174061433, + "grad_norm": 1.9084302832787015, + "learning_rate": 7.300673208265332e-07, + "loss": 0.0298, + "step": 9807 + }, + { + "epoch": 2.231626848691695, + "grad_norm": 1.80520200352483, + "learning_rate": 7.299792400178219e-07, + "loss": 0.0513, + "step": 9808 + }, + { + "epoch": 2.231854379977247, + "grad_norm": 2.324964360887839, + "learning_rate": 7.298911570635267e-07, + "loss": 0.1249, + "step": 9809 + }, + { + "epoch": 2.2320819112627985, + "grad_norm": 0.5890416085092136, + "learning_rate": 7.298030719654487e-07, + "loss": 0.0085, + "step": 9810 + }, + { + "epoch": 2.2323094425483503, + "grad_norm": 1.2366273798903231, + "learning_rate": 7.297149847253877e-07, + "loss": 0.0566, + "step": 9811 + }, + { + "epoch": 2.232536973833902, + "grad_norm": 1.321596022173302, + "learning_rate": 7.296268953451443e-07, + "loss": 0.063, + "step": 9812 + }, + { + "epoch": 2.232764505119454, + "grad_norm": 1.6276366578385928, + "learning_rate": 7.295388038265188e-07, + "loss": 0.0917, + "step": 9813 + }, + { + "epoch": 2.2329920364050055, + "grad_norm": 1.6677993307189287, + "learning_rate": 7.294507101713115e-07, + "loss": 0.042, + "step": 9814 + }, + { + "epoch": 2.2332195676905573, + "grad_norm": 1.6822351005486869, + "learning_rate": 7.29362614381323e-07, + "loss": 0.0311, + "step": 9815 + }, + { + "epoch": 2.233447098976109, + "grad_norm": 1.3732093547178756, + "learning_rate": 7.292745164583537e-07, + "loss": 0.0491, + "step": 9816 + }, + { + "epoch": 2.233674630261661, + "grad_norm": 1.2500554936588006, + "learning_rate": 7.291864164042042e-07, + "loss": 0.0308, + "step": 9817 + }, + { + "epoch": 2.2339021615472126, + "grad_norm": 1.857020354428366, + "learning_rate": 7.290983142206756e-07, + "loss": 0.0297, + "step": 9818 + }, + { + "epoch": 2.2341296928327643, + "grad_norm": 1.9841408180487183, + "learning_rate": 7.290102099095676e-07, + "loss": 0.1278, + "step": 9819 + }, + { + "epoch": 2.234357224118316, + "grad_norm": 1.4379399634472614, + "learning_rate": 7.289221034726816e-07, + "loss": 0.0671, + "step": 9820 + }, + { + "epoch": 2.2345847554038683, + "grad_norm": 1.473057794580955, + "learning_rate": 7.288339949118182e-07, + "loss": 0.1103, + "step": 9821 + }, + { + "epoch": 2.23481228668942, + "grad_norm": 1.0442658715530824, + "learning_rate": 7.287458842287781e-07, + "loss": 0.0531, + "step": 9822 + }, + { + "epoch": 2.2350398179749718, + "grad_norm": 2.5068843676249326, + "learning_rate": 7.286577714253619e-07, + "loss": 0.1222, + "step": 9823 + }, + { + "epoch": 2.2352673492605235, + "grad_norm": 1.4522134035301333, + "learning_rate": 7.285696565033711e-07, + "loss": 0.0874, + "step": 9824 + }, + { + "epoch": 2.2354948805460753, + "grad_norm": 1.2481878278233278, + "learning_rate": 7.284815394646058e-07, + "loss": 0.0815, + "step": 9825 + }, + { + "epoch": 2.235722411831627, + "grad_norm": 1.5249838750004165, + "learning_rate": 7.283934203108675e-07, + "loss": 0.1, + "step": 9826 + }, + { + "epoch": 2.2359499431171788, + "grad_norm": 1.3030221063732124, + "learning_rate": 7.283052990439571e-07, + "loss": 0.0594, + "step": 9827 + }, + { + "epoch": 2.2361774744027305, + "grad_norm": 2.3216748010911283, + "learning_rate": 7.282171756656756e-07, + "loss": 0.0255, + "step": 9828 + }, + { + "epoch": 2.2364050056882823, + "grad_norm": 1.9443385917771203, + "learning_rate": 7.281290501778238e-07, + "loss": 0.0457, + "step": 9829 + }, + { + "epoch": 2.236632536973834, + "grad_norm": 1.0056931152888324, + "learning_rate": 7.280409225822033e-07, + "loss": 0.0233, + "step": 9830 + }, + { + "epoch": 2.2368600682593858, + "grad_norm": 1.3155469661761876, + "learning_rate": 7.279527928806152e-07, + "loss": 0.1135, + "step": 9831 + }, + { + "epoch": 2.2370875995449375, + "grad_norm": 0.7904031792767765, + "learning_rate": 7.278646610748602e-07, + "loss": 0.0064, + "step": 9832 + }, + { + "epoch": 2.2373151308304893, + "grad_norm": 1.9904079589452548, + "learning_rate": 7.277765271667402e-07, + "loss": 0.1661, + "step": 9833 + }, + { + "epoch": 2.237542662116041, + "grad_norm": 1.4170094295444529, + "learning_rate": 7.276883911580561e-07, + "loss": 0.0539, + "step": 9834 + }, + { + "epoch": 2.2377701934015928, + "grad_norm": 1.4207997584092764, + "learning_rate": 7.276002530506094e-07, + "loss": 0.0577, + "step": 9835 + }, + { + "epoch": 2.2379977246871445, + "grad_norm": 1.2605270035527005, + "learning_rate": 7.275121128462012e-07, + "loss": 0.0456, + "step": 9836 + }, + { + "epoch": 2.2382252559726963, + "grad_norm": 2.3260293746886114, + "learning_rate": 7.274239705466336e-07, + "loss": 0.0451, + "step": 9837 + }, + { + "epoch": 2.238452787258248, + "grad_norm": 1.3997759929972946, + "learning_rate": 7.273358261537074e-07, + "loss": 0.0354, + "step": 9838 + }, + { + "epoch": 2.2386803185438, + "grad_norm": 1.2172453709257616, + "learning_rate": 7.272476796692242e-07, + "loss": 0.0561, + "step": 9839 + }, + { + "epoch": 2.2389078498293515, + "grad_norm": 1.4994253324301205, + "learning_rate": 7.271595310949858e-07, + "loss": 0.0566, + "step": 9840 + }, + { + "epoch": 2.2391353811149033, + "grad_norm": 1.4451968807486255, + "learning_rate": 7.270713804327937e-07, + "loss": 0.0622, + "step": 9841 + }, + { + "epoch": 2.239362912400455, + "grad_norm": 0.8515032871637725, + "learning_rate": 7.269832276844494e-07, + "loss": 0.0179, + "step": 9842 + }, + { + "epoch": 2.239590443686007, + "grad_norm": 1.3621068936112914, + "learning_rate": 7.26895072851755e-07, + "loss": 0.0986, + "step": 9843 + }, + { + "epoch": 2.2398179749715585, + "grad_norm": 1.5511523626297565, + "learning_rate": 7.268069159365117e-07, + "loss": 0.0241, + "step": 9844 + }, + { + "epoch": 2.2400455062571103, + "grad_norm": 1.9676231000096438, + "learning_rate": 7.267187569405215e-07, + "loss": 0.0417, + "step": 9845 + }, + { + "epoch": 2.240273037542662, + "grad_norm": 1.6958828791299405, + "learning_rate": 7.266305958655861e-07, + "loss": 0.0585, + "step": 9846 + }, + { + "epoch": 2.240500568828214, + "grad_norm": 0.9741324819867013, + "learning_rate": 7.265424327135077e-07, + "loss": 0.0231, + "step": 9847 + }, + { + "epoch": 2.2407281001137656, + "grad_norm": 1.630118897926355, + "learning_rate": 7.264542674860878e-07, + "loss": 0.0695, + "step": 9848 + }, + { + "epoch": 2.2409556313993173, + "grad_norm": 1.9962775184840824, + "learning_rate": 7.263661001851284e-07, + "loss": 0.0488, + "step": 9849 + }, + { + "epoch": 2.241183162684869, + "grad_norm": 0.9182636614486046, + "learning_rate": 7.262779308124317e-07, + "loss": 0.0788, + "step": 9850 + }, + { + "epoch": 2.241410693970421, + "grad_norm": 1.3622488518858606, + "learning_rate": 7.261897593697995e-07, + "loss": 0.0771, + "step": 9851 + }, + { + "epoch": 2.2416382252559726, + "grad_norm": 1.3539481154987774, + "learning_rate": 7.26101585859034e-07, + "loss": 0.075, + "step": 9852 + }, + { + "epoch": 2.2418657565415243, + "grad_norm": 1.7429889114509627, + "learning_rate": 7.26013410281937e-07, + "loss": 0.0889, + "step": 9853 + }, + { + "epoch": 2.242093287827076, + "grad_norm": 2.5125690754750236, + "learning_rate": 7.259252326403111e-07, + "loss": 0.0732, + "step": 9854 + }, + { + "epoch": 2.242320819112628, + "grad_norm": 2.6576290109840848, + "learning_rate": 7.258370529359583e-07, + "loss": 0.0606, + "step": 9855 + }, + { + "epoch": 2.2425483503981796, + "grad_norm": 1.1706931835111798, + "learning_rate": 7.257488711706808e-07, + "loss": 0.0737, + "step": 9856 + }, + { + "epoch": 2.2427758816837313, + "grad_norm": 1.6960326501685519, + "learning_rate": 7.256606873462808e-07, + "loss": 0.0368, + "step": 9857 + }, + { + "epoch": 2.243003412969283, + "grad_norm": 2.6125585430856084, + "learning_rate": 7.255725014645608e-07, + "loss": 0.0674, + "step": 9858 + }, + { + "epoch": 2.243230944254835, + "grad_norm": 2.1835334473959938, + "learning_rate": 7.254843135273229e-07, + "loss": 0.0862, + "step": 9859 + }, + { + "epoch": 2.243458475540387, + "grad_norm": 2.3970061321251235, + "learning_rate": 7.253961235363699e-07, + "loss": 0.0495, + "step": 9860 + }, + { + "epoch": 2.2436860068259388, + "grad_norm": 2.074108416187049, + "learning_rate": 7.253079314935038e-07, + "loss": 0.0454, + "step": 9861 + }, + { + "epoch": 2.2439135381114905, + "grad_norm": 1.328079819897513, + "learning_rate": 7.252197374005273e-07, + "loss": 0.0191, + "step": 9862 + }, + { + "epoch": 2.2441410693970423, + "grad_norm": 1.2939855825394784, + "learning_rate": 7.251315412592431e-07, + "loss": 0.0578, + "step": 9863 + }, + { + "epoch": 2.244368600682594, + "grad_norm": 2.2509045818859548, + "learning_rate": 7.250433430714534e-07, + "loss": 0.1045, + "step": 9864 + }, + { + "epoch": 2.244596131968146, + "grad_norm": 1.4898212729805753, + "learning_rate": 7.249551428389612e-07, + "loss": 0.1267, + "step": 9865 + }, + { + "epoch": 2.2448236632536975, + "grad_norm": 1.6614659712378599, + "learning_rate": 7.248669405635686e-07, + "loss": 0.0553, + "step": 9866 + }, + { + "epoch": 2.2450511945392493, + "grad_norm": 1.8869151203162269, + "learning_rate": 7.247787362470789e-07, + "loss": 0.0916, + "step": 9867 + }, + { + "epoch": 2.245278725824801, + "grad_norm": 1.5458943538465553, + "learning_rate": 7.246905298912945e-07, + "loss": 0.0554, + "step": 9868 + }, + { + "epoch": 2.245506257110353, + "grad_norm": 0.7392169094288412, + "learning_rate": 7.246023214980183e-07, + "loss": 0.0524, + "step": 9869 + }, + { + "epoch": 2.2457337883959045, + "grad_norm": 1.8209684923561502, + "learning_rate": 7.245141110690529e-07, + "loss": 0.0322, + "step": 9870 + }, + { + "epoch": 2.2459613196814563, + "grad_norm": 1.522212197488696, + "learning_rate": 7.244258986062015e-07, + "loss": 0.0645, + "step": 9871 + }, + { + "epoch": 2.246188850967008, + "grad_norm": 1.3876316248473302, + "learning_rate": 7.243376841112668e-07, + "loss": 0.0603, + "step": 9872 + }, + { + "epoch": 2.24641638225256, + "grad_norm": 1.5317231710378987, + "learning_rate": 7.242494675860515e-07, + "loss": 0.1296, + "step": 9873 + }, + { + "epoch": 2.2466439135381115, + "grad_norm": 1.1426129960504483, + "learning_rate": 7.241612490323591e-07, + "loss": 0.0823, + "step": 9874 + }, + { + "epoch": 2.2468714448236633, + "grad_norm": 2.3052770604857344, + "learning_rate": 7.240730284519924e-07, + "loss": 0.1218, + "step": 9875 + }, + { + "epoch": 2.247098976109215, + "grad_norm": 1.5940210980654792, + "learning_rate": 7.239848058467544e-07, + "loss": 0.1389, + "step": 9876 + }, + { + "epoch": 2.247326507394767, + "grad_norm": 1.7269644276056755, + "learning_rate": 7.238965812184482e-07, + "loss": 0.0697, + "step": 9877 + }, + { + "epoch": 2.2475540386803186, + "grad_norm": 1.0238053294267258, + "learning_rate": 7.23808354568877e-07, + "loss": 0.0301, + "step": 9878 + }, + { + "epoch": 2.2477815699658703, + "grad_norm": 1.7187898246569568, + "learning_rate": 7.23720125899844e-07, + "loss": 0.1406, + "step": 9879 + }, + { + "epoch": 2.248009101251422, + "grad_norm": 1.5000066250827442, + "learning_rate": 7.236318952131524e-07, + "loss": 0.1786, + "step": 9880 + }, + { + "epoch": 2.248236632536974, + "grad_norm": 1.463942565947484, + "learning_rate": 7.235436625106057e-07, + "loss": 0.0793, + "step": 9881 + }, + { + "epoch": 2.2484641638225256, + "grad_norm": 1.7880537918522657, + "learning_rate": 7.234554277940067e-07, + "loss": 0.0448, + "step": 9882 + }, + { + "epoch": 2.2486916951080773, + "grad_norm": 2.260998293452545, + "learning_rate": 7.233671910651592e-07, + "loss": 0.0529, + "step": 9883 + }, + { + "epoch": 2.248919226393629, + "grad_norm": 1.2807372862338948, + "learning_rate": 7.232789523258665e-07, + "loss": 0.059, + "step": 9884 + }, + { + "epoch": 2.249146757679181, + "grad_norm": 1.4950384427415302, + "learning_rate": 7.23190711577932e-07, + "loss": 0.0574, + "step": 9885 + }, + { + "epoch": 2.2493742889647326, + "grad_norm": 1.5598319789485426, + "learning_rate": 7.23102468823159e-07, + "loss": 0.047, + "step": 9886 + }, + { + "epoch": 2.2496018202502843, + "grad_norm": 1.9796818456053953, + "learning_rate": 7.230142240633515e-07, + "loss": 0.0707, + "step": 9887 + }, + { + "epoch": 2.249829351535836, + "grad_norm": 1.9752382025732782, + "learning_rate": 7.229259773003128e-07, + "loss": 0.0386, + "step": 9888 + }, + { + "epoch": 2.250056882821388, + "grad_norm": 7.475740126394822, + "learning_rate": 7.228377285358461e-07, + "loss": 0.1604, + "step": 9889 + }, + { + "epoch": 2.2502844141069396, + "grad_norm": 1.7369815637099226, + "learning_rate": 7.227494777717555e-07, + "loss": 0.1107, + "step": 9890 + }, + { + "epoch": 2.2505119453924913, + "grad_norm": 1.7058053834872264, + "learning_rate": 7.226612250098449e-07, + "loss": 0.0415, + "step": 9891 + }, + { + "epoch": 2.250739476678043, + "grad_norm": 1.9920451641621941, + "learning_rate": 7.225729702519174e-07, + "loss": 0.0864, + "step": 9892 + }, + { + "epoch": 2.250967007963595, + "grad_norm": 1.5859261163302527, + "learning_rate": 7.224847134997772e-07, + "loss": 0.0746, + "step": 9893 + }, + { + "epoch": 2.2511945392491466, + "grad_norm": 3.0570926687797613, + "learning_rate": 7.223964547552281e-07, + "loss": 0.1796, + "step": 9894 + }, + { + "epoch": 2.2514220705346983, + "grad_norm": 2.1411531385665046, + "learning_rate": 7.223081940200738e-07, + "loss": 0.1186, + "step": 9895 + }, + { + "epoch": 2.25164960182025, + "grad_norm": 2.5605029239540857, + "learning_rate": 7.22219931296118e-07, + "loss": 0.0903, + "step": 9896 + }, + { + "epoch": 2.2518771331058023, + "grad_norm": 1.4942403697676752, + "learning_rate": 7.22131666585165e-07, + "loss": 0.0936, + "step": 9897 + }, + { + "epoch": 2.2521046643913536, + "grad_norm": 3.5156967237443735, + "learning_rate": 7.220433998890188e-07, + "loss": 0.0587, + "step": 9898 + }, + { + "epoch": 2.252332195676906, + "grad_norm": 1.0127705571222378, + "learning_rate": 7.21955131209483e-07, + "loss": 0.0144, + "step": 9899 + }, + { + "epoch": 2.252559726962457, + "grad_norm": 1.8410734362290129, + "learning_rate": 7.21866860548362e-07, + "loss": 0.0524, + "step": 9900 + }, + { + "epoch": 2.2527872582480093, + "grad_norm": 1.80027236056254, + "learning_rate": 7.217785879074597e-07, + "loss": 0.0522, + "step": 9901 + }, + { + "epoch": 2.253014789533561, + "grad_norm": 2.5337764729931918, + "learning_rate": 7.216903132885803e-07, + "loss": 0.0441, + "step": 9902 + }, + { + "epoch": 2.253242320819113, + "grad_norm": 1.5470448377651977, + "learning_rate": 7.216020366935279e-07, + "loss": 0.0217, + "step": 9903 + }, + { + "epoch": 2.2534698521046646, + "grad_norm": 1.2354501880350004, + "learning_rate": 7.215137581241071e-07, + "loss": 0.0748, + "step": 9904 + }, + { + "epoch": 2.2536973833902163, + "grad_norm": 1.0999499177289918, + "learning_rate": 7.214254775821216e-07, + "loss": 0.035, + "step": 9905 + }, + { + "epoch": 2.253924914675768, + "grad_norm": 1.0690591094007649, + "learning_rate": 7.213371950693759e-07, + "loss": 0.0284, + "step": 9906 + }, + { + "epoch": 2.25415244596132, + "grad_norm": 0.6173208377977818, + "learning_rate": 7.212489105876745e-07, + "loss": 0.0074, + "step": 9907 + }, + { + "epoch": 2.2543799772468716, + "grad_norm": 1.7166506684911709, + "learning_rate": 7.211606241388217e-07, + "loss": 0.0831, + "step": 9908 + }, + { + "epoch": 2.2546075085324233, + "grad_norm": 2.062793189329472, + "learning_rate": 7.210723357246216e-07, + "loss": 0.0573, + "step": 9909 + }, + { + "epoch": 2.254835039817975, + "grad_norm": 1.0112757233221028, + "learning_rate": 7.20984045346879e-07, + "loss": 0.0392, + "step": 9910 + }, + { + "epoch": 2.255062571103527, + "grad_norm": 1.6933376215077947, + "learning_rate": 7.208957530073983e-07, + "loss": 0.0544, + "step": 9911 + }, + { + "epoch": 2.2552901023890786, + "grad_norm": 1.7129631404596137, + "learning_rate": 7.208074587079841e-07, + "loss": 0.0418, + "step": 9912 + }, + { + "epoch": 2.2555176336746303, + "grad_norm": 1.2546125790474387, + "learning_rate": 7.207191624504409e-07, + "loss": 0.0552, + "step": 9913 + }, + { + "epoch": 2.255745164960182, + "grad_norm": 2.2119715662073527, + "learning_rate": 7.206308642365733e-07, + "loss": 0.0487, + "step": 9914 + }, + { + "epoch": 2.255972696245734, + "grad_norm": 1.4917211182583106, + "learning_rate": 7.20542564068186e-07, + "loss": 0.0244, + "step": 9915 + }, + { + "epoch": 2.2562002275312856, + "grad_norm": 2.0171286803125867, + "learning_rate": 7.204542619470837e-07, + "loss": 0.1563, + "step": 9916 + }, + { + "epoch": 2.2564277588168373, + "grad_norm": 1.1436411224819762, + "learning_rate": 7.203659578750709e-07, + "loss": 0.0513, + "step": 9917 + }, + { + "epoch": 2.256655290102389, + "grad_norm": 1.8735502255584853, + "learning_rate": 7.202776518539527e-07, + "loss": 0.0469, + "step": 9918 + }, + { + "epoch": 2.256882821387941, + "grad_norm": 1.4066962208218972, + "learning_rate": 7.20189343885534e-07, + "loss": 0.0474, + "step": 9919 + }, + { + "epoch": 2.2571103526734926, + "grad_norm": 1.9691489628955297, + "learning_rate": 7.201010339716191e-07, + "loss": 0.1539, + "step": 9920 + }, + { + "epoch": 2.2573378839590443, + "grad_norm": 1.7835434199911864, + "learning_rate": 7.200127221140134e-07, + "loss": 0.0579, + "step": 9921 + }, + { + "epoch": 2.257565415244596, + "grad_norm": 1.0668305086393692, + "learning_rate": 7.199244083145217e-07, + "loss": 0.0534, + "step": 9922 + }, + { + "epoch": 2.257792946530148, + "grad_norm": 2.2770530807974465, + "learning_rate": 7.198360925749488e-07, + "loss": 0.044, + "step": 9923 + }, + { + "epoch": 2.2580204778156996, + "grad_norm": 1.7399756467597929, + "learning_rate": 7.197477748971e-07, + "loss": 0.0153, + "step": 9924 + }, + { + "epoch": 2.2582480091012513, + "grad_norm": 1.7431788996917812, + "learning_rate": 7.196594552827803e-07, + "loss": 0.1163, + "step": 9925 + }, + { + "epoch": 2.258475540386803, + "grad_norm": 1.5634561554047466, + "learning_rate": 7.195711337337943e-07, + "loss": 0.0367, + "step": 9926 + }, + { + "epoch": 2.258703071672355, + "grad_norm": 1.3915344740514322, + "learning_rate": 7.194828102519479e-07, + "loss": 0.0284, + "step": 9927 + }, + { + "epoch": 2.2589306029579066, + "grad_norm": 1.6271005234824392, + "learning_rate": 7.193944848390458e-07, + "loss": 0.0229, + "step": 9928 + }, + { + "epoch": 2.2591581342434583, + "grad_norm": 1.928042580660145, + "learning_rate": 7.193061574968932e-07, + "loss": 0.0689, + "step": 9929 + }, + { + "epoch": 2.25938566552901, + "grad_norm": 2.177865049462624, + "learning_rate": 7.192178282272955e-07, + "loss": 0.0598, + "step": 9930 + }, + { + "epoch": 2.259613196814562, + "grad_norm": 1.151007882661663, + "learning_rate": 7.191294970320581e-07, + "loss": 0.0597, + "step": 9931 + }, + { + "epoch": 2.2598407281001136, + "grad_norm": 1.6218107645944073, + "learning_rate": 7.19041163912986e-07, + "loss": 0.0395, + "step": 9932 + }, + { + "epoch": 2.2600682593856654, + "grad_norm": 1.673841673978979, + "learning_rate": 7.189528288718846e-07, + "loss": 0.0161, + "step": 9933 + }, + { + "epoch": 2.260295790671217, + "grad_norm": 1.0345413033273134, + "learning_rate": 7.188644919105597e-07, + "loss": 0.0244, + "step": 9934 + }, + { + "epoch": 2.260523321956769, + "grad_norm": 1.906104579654973, + "learning_rate": 7.187761530308164e-07, + "loss": 0.0512, + "step": 9935 + }, + { + "epoch": 2.260750853242321, + "grad_norm": 1.7059121788224236, + "learning_rate": 7.186878122344602e-07, + "loss": 0.0422, + "step": 9936 + }, + { + "epoch": 2.2609783845278724, + "grad_norm": 1.6416412531110258, + "learning_rate": 7.185994695232967e-07, + "loss": 0.0401, + "step": 9937 + }, + { + "epoch": 2.2612059158134246, + "grad_norm": 1.4075819784417, + "learning_rate": 7.185111248991318e-07, + "loss": 0.067, + "step": 9938 + }, + { + "epoch": 2.261433447098976, + "grad_norm": 2.1174815038953616, + "learning_rate": 7.184227783637705e-07, + "loss": 0.1327, + "step": 9939 + }, + { + "epoch": 2.261660978384528, + "grad_norm": 1.542504117878399, + "learning_rate": 7.183344299190186e-07, + "loss": 0.1173, + "step": 9940 + }, + { + "epoch": 2.26188850967008, + "grad_norm": 2.101680236434913, + "learning_rate": 7.182460795666821e-07, + "loss": 0.0369, + "step": 9941 + }, + { + "epoch": 2.2621160409556316, + "grad_norm": 1.356841882599635, + "learning_rate": 7.181577273085663e-07, + "loss": 0.0405, + "step": 9942 + }, + { + "epoch": 2.2623435722411833, + "grad_norm": 1.5454684769120097, + "learning_rate": 7.180693731464773e-07, + "loss": 0.0643, + "step": 9943 + }, + { + "epoch": 2.262571103526735, + "grad_norm": 1.3510060039314944, + "learning_rate": 7.179810170822208e-07, + "loss": 0.1395, + "step": 9944 + }, + { + "epoch": 2.262798634812287, + "grad_norm": 1.4931031761170002, + "learning_rate": 7.178926591176025e-07, + "loss": 0.0457, + "step": 9945 + }, + { + "epoch": 2.2630261660978386, + "grad_norm": 1.7016546510133408, + "learning_rate": 7.178042992544284e-07, + "loss": 0.1187, + "step": 9946 + }, + { + "epoch": 2.2632536973833903, + "grad_norm": 1.9574623874457986, + "learning_rate": 7.177159374945043e-07, + "loss": 0.2011, + "step": 9947 + }, + { + "epoch": 2.263481228668942, + "grad_norm": 2.460061761843094, + "learning_rate": 7.176275738396363e-07, + "loss": 0.0683, + "step": 9948 + }, + { + "epoch": 2.263708759954494, + "grad_norm": 1.4173305961331562, + "learning_rate": 7.175392082916305e-07, + "loss": 0.0494, + "step": 9949 + }, + { + "epoch": 2.2639362912400456, + "grad_norm": 1.8854550657499727, + "learning_rate": 7.174508408522926e-07, + "loss": 0.1446, + "step": 9950 + }, + { + "epoch": 2.2641638225255973, + "grad_norm": 1.5115714147812573, + "learning_rate": 7.173624715234288e-07, + "loss": 0.063, + "step": 9951 + }, + { + "epoch": 2.264391353811149, + "grad_norm": 1.870748040561277, + "learning_rate": 7.172741003068454e-07, + "loss": 0.0529, + "step": 9952 + }, + { + "epoch": 2.264618885096701, + "grad_norm": 3.4700874302820917, + "learning_rate": 7.171857272043481e-07, + "loss": 0.054, + "step": 9953 + }, + { + "epoch": 2.2648464163822526, + "grad_norm": 3.077671443557571, + "learning_rate": 7.170973522177435e-07, + "loss": 0.0668, + "step": 9954 + }, + { + "epoch": 2.2650739476678043, + "grad_norm": 1.2234363200663947, + "learning_rate": 7.170089753488378e-07, + "loss": 0.0145, + "step": 9955 + }, + { + "epoch": 2.265301478953356, + "grad_norm": 2.615122626044104, + "learning_rate": 7.169205965994371e-07, + "loss": 0.111, + "step": 9956 + }, + { + "epoch": 2.265529010238908, + "grad_norm": 1.455531441061188, + "learning_rate": 7.168322159713477e-07, + "loss": 0.0388, + "step": 9957 + }, + { + "epoch": 2.2657565415244596, + "grad_norm": 1.8868133759128798, + "learning_rate": 7.16743833466376e-07, + "loss": 0.0208, + "step": 9958 + }, + { + "epoch": 2.2659840728100114, + "grad_norm": 1.8986978689090965, + "learning_rate": 7.166554490863283e-07, + "loss": 0.0548, + "step": 9959 + }, + { + "epoch": 2.266211604095563, + "grad_norm": 2.0257615128365547, + "learning_rate": 7.165670628330112e-07, + "loss": 0.0377, + "step": 9960 + }, + { + "epoch": 2.266439135381115, + "grad_norm": 1.221895596184263, + "learning_rate": 7.16478674708231e-07, + "loss": 0.0228, + "step": 9961 + }, + { + "epoch": 2.2666666666666666, + "grad_norm": 2.0120406197398744, + "learning_rate": 7.163902847137942e-07, + "loss": 0.0733, + "step": 9962 + }, + { + "epoch": 2.2668941979522184, + "grad_norm": 1.1094453517774197, + "learning_rate": 7.163018928515074e-07, + "loss": 0.1032, + "step": 9963 + }, + { + "epoch": 2.26712172923777, + "grad_norm": 1.641632508142861, + "learning_rate": 7.16213499123177e-07, + "loss": 0.1367, + "step": 9964 + }, + { + "epoch": 2.267349260523322, + "grad_norm": 1.175803720791843, + "learning_rate": 7.161251035306099e-07, + "loss": 0.0277, + "step": 9965 + }, + { + "epoch": 2.2675767918088736, + "grad_norm": 1.6048600629963754, + "learning_rate": 7.160367060756125e-07, + "loss": 0.124, + "step": 9966 + }, + { + "epoch": 2.2678043230944254, + "grad_norm": 0.9414032827252117, + "learning_rate": 7.159483067599913e-07, + "loss": 0.0664, + "step": 9967 + }, + { + "epoch": 2.268031854379977, + "grad_norm": 1.5401276895833678, + "learning_rate": 7.158599055855536e-07, + "loss": 0.0298, + "step": 9968 + }, + { + "epoch": 2.268259385665529, + "grad_norm": 1.8654507290663085, + "learning_rate": 7.157715025541059e-07, + "loss": 0.1168, + "step": 9969 + }, + { + "epoch": 2.2684869169510806, + "grad_norm": 2.888492273928279, + "learning_rate": 7.156830976674547e-07, + "loss": 0.0642, + "step": 9970 + }, + { + "epoch": 2.2687144482366324, + "grad_norm": 1.4216840942395765, + "learning_rate": 7.155946909274071e-07, + "loss": 0.0355, + "step": 9971 + }, + { + "epoch": 2.268941979522184, + "grad_norm": 1.9071513514995047, + "learning_rate": 7.1550628233577e-07, + "loss": 0.1619, + "step": 9972 + }, + { + "epoch": 2.269169510807736, + "grad_norm": 2.4450924226682402, + "learning_rate": 7.154178718943502e-07, + "loss": 0.0472, + "step": 9973 + }, + { + "epoch": 2.2693970420932876, + "grad_norm": 1.6468107756072237, + "learning_rate": 7.153294596049547e-07, + "loss": 0.0422, + "step": 9974 + }, + { + "epoch": 2.26962457337884, + "grad_norm": 1.8138360151761403, + "learning_rate": 7.152410454693905e-07, + "loss": 0.0327, + "step": 9975 + }, + { + "epoch": 2.269852104664391, + "grad_norm": 2.0459442632750062, + "learning_rate": 7.151526294894646e-07, + "loss": 0.0352, + "step": 9976 + }, + { + "epoch": 2.2700796359499433, + "grad_norm": 1.982952688424646, + "learning_rate": 7.150642116669839e-07, + "loss": 0.0648, + "step": 9977 + }, + { + "epoch": 2.2703071672354946, + "grad_norm": 2.55458641638833, + "learning_rate": 7.149757920037558e-07, + "loss": 0.0707, + "step": 9978 + }, + { + "epoch": 2.270534698521047, + "grad_norm": 1.7074553739199554, + "learning_rate": 7.148873705015875e-07, + "loss": 0.0593, + "step": 9979 + }, + { + "epoch": 2.2707622298065986, + "grad_norm": 2.074424873267346, + "learning_rate": 7.147989471622856e-07, + "loss": 0.0227, + "step": 9980 + }, + { + "epoch": 2.2709897610921503, + "grad_norm": 1.3083781772002405, + "learning_rate": 7.147105219876578e-07, + "loss": 0.0459, + "step": 9981 + }, + { + "epoch": 2.271217292377702, + "grad_norm": 1.4915122509404122, + "learning_rate": 7.146220949795114e-07, + "loss": 0.0341, + "step": 9982 + }, + { + "epoch": 2.271444823663254, + "grad_norm": 2.8184334181010637, + "learning_rate": 7.145336661396532e-07, + "loss": 0.1324, + "step": 9983 + }, + { + "epoch": 2.2716723549488056, + "grad_norm": 1.6353015872148624, + "learning_rate": 7.144452354698911e-07, + "loss": 0.0517, + "step": 9984 + }, + { + "epoch": 2.2718998862343573, + "grad_norm": 2.9208976878906863, + "learning_rate": 7.143568029720321e-07, + "loss": 0.0848, + "step": 9985 + }, + { + "epoch": 2.272127417519909, + "grad_norm": 2.0928691734337086, + "learning_rate": 7.142683686478838e-07, + "loss": 0.0609, + "step": 9986 + }, + { + "epoch": 2.272354948805461, + "grad_norm": 1.8743162895870351, + "learning_rate": 7.141799324992532e-07, + "loss": 0.1895, + "step": 9987 + }, + { + "epoch": 2.2725824800910126, + "grad_norm": 2.3999558451914784, + "learning_rate": 7.140914945279486e-07, + "loss": 0.0508, + "step": 9988 + }, + { + "epoch": 2.2728100113765644, + "grad_norm": 2.049360168071473, + "learning_rate": 7.140030547357768e-07, + "loss": 0.1587, + "step": 9989 + }, + { + "epoch": 2.273037542662116, + "grad_norm": 1.0111727363641771, + "learning_rate": 7.139146131245453e-07, + "loss": 0.0275, + "step": 9990 + }, + { + "epoch": 2.273265073947668, + "grad_norm": 1.438736486549141, + "learning_rate": 7.138261696960624e-07, + "loss": 0.0405, + "step": 9991 + }, + { + "epoch": 2.2734926052332196, + "grad_norm": 1.4944431094548225, + "learning_rate": 7.137377244521348e-07, + "loss": 0.0337, + "step": 9992 + }, + { + "epoch": 2.2737201365187714, + "grad_norm": 1.9316880879496054, + "learning_rate": 7.136492773945711e-07, + "loss": 0.0879, + "step": 9993 + }, + { + "epoch": 2.273947667804323, + "grad_norm": 1.9559043924097677, + "learning_rate": 7.135608285251782e-07, + "loss": 0.1422, + "step": 9994 + }, + { + "epoch": 2.274175199089875, + "grad_norm": 1.5340336317413974, + "learning_rate": 7.134723778457643e-07, + "loss": 0.1182, + "step": 9995 + }, + { + "epoch": 2.2744027303754266, + "grad_norm": 1.9683779705330084, + "learning_rate": 7.13383925358137e-07, + "loss": 0.0946, + "step": 9996 + }, + { + "epoch": 2.2746302616609784, + "grad_norm": 1.4407703401361611, + "learning_rate": 7.13295471064104e-07, + "loss": 0.047, + "step": 9997 + }, + { + "epoch": 2.27485779294653, + "grad_norm": 1.3620126880714782, + "learning_rate": 7.132070149654734e-07, + "loss": 0.0754, + "step": 9998 + }, + { + "epoch": 2.275085324232082, + "grad_norm": 1.323713417182486, + "learning_rate": 7.131185570640529e-07, + "loss": 0.0945, + "step": 9999 + }, + { + "epoch": 2.2753128555176336, + "grad_norm": 1.1379200276931136, + "learning_rate": 7.130300973616506e-07, + "loss": 0.0491, + "step": 10000 + }, + { + "epoch": 2.2755403868031854, + "grad_norm": 2.159440972890442, + "learning_rate": 7.129416358600742e-07, + "loss": 0.0566, + "step": 10001 + }, + { + "epoch": 2.275767918088737, + "grad_norm": 1.0553680750834689, + "learning_rate": 7.12853172561132e-07, + "loss": 0.0398, + "step": 10002 + }, + { + "epoch": 2.275995449374289, + "grad_norm": 1.4740891265096314, + "learning_rate": 7.127647074666316e-07, + "loss": 0.0803, + "step": 10003 + }, + { + "epoch": 2.2762229806598406, + "grad_norm": 1.683324086486543, + "learning_rate": 7.126762405783813e-07, + "loss": 0.0685, + "step": 10004 + }, + { + "epoch": 2.2764505119453924, + "grad_norm": 1.3073877630721384, + "learning_rate": 7.125877718981894e-07, + "loss": 0.0517, + "step": 10005 + }, + { + "epoch": 2.276678043230944, + "grad_norm": 1.2100626161922907, + "learning_rate": 7.124993014278639e-07, + "loss": 0.0368, + "step": 10006 + }, + { + "epoch": 2.276905574516496, + "grad_norm": 2.799869689973577, + "learning_rate": 7.124108291692128e-07, + "loss": 0.0783, + "step": 10007 + }, + { + "epoch": 2.2771331058020476, + "grad_norm": 1.0624859753841573, + "learning_rate": 7.123223551240445e-07, + "loss": 0.0171, + "step": 10008 + }, + { + "epoch": 2.2773606370875994, + "grad_norm": 1.6051031330142305, + "learning_rate": 7.122338792941671e-07, + "loss": 0.0452, + "step": 10009 + }, + { + "epoch": 2.277588168373151, + "grad_norm": 1.8050367214300154, + "learning_rate": 7.121454016813889e-07, + "loss": 0.0238, + "step": 10010 + }, + { + "epoch": 2.277815699658703, + "grad_norm": 1.1666362583854997, + "learning_rate": 7.120569222875184e-07, + "loss": 0.0867, + "step": 10011 + }, + { + "epoch": 2.2780432309442546, + "grad_norm": 1.2099871046712096, + "learning_rate": 7.119684411143638e-07, + "loss": 0.0876, + "step": 10012 + }, + { + "epoch": 2.2782707622298064, + "grad_norm": 1.6031122595119038, + "learning_rate": 7.118799581637336e-07, + "loss": 0.0441, + "step": 10013 + }, + { + "epoch": 2.2784982935153586, + "grad_norm": 1.6220347650096594, + "learning_rate": 7.117914734374362e-07, + "loss": 0.0662, + "step": 10014 + }, + { + "epoch": 2.27872582480091, + "grad_norm": 2.308298130308238, + "learning_rate": 7.1170298693728e-07, + "loss": 0.0595, + "step": 10015 + }, + { + "epoch": 2.278953356086462, + "grad_norm": 1.402203199616304, + "learning_rate": 7.116144986650736e-07, + "loss": 0.0963, + "step": 10016 + }, + { + "epoch": 2.2791808873720134, + "grad_norm": 1.5311548749634378, + "learning_rate": 7.115260086226253e-07, + "loss": 0.0623, + "step": 10017 + }, + { + "epoch": 2.2794084186575656, + "grad_norm": 1.7957851850677895, + "learning_rate": 7.114375168117439e-07, + "loss": 0.1566, + "step": 10018 + }, + { + "epoch": 2.2796359499431174, + "grad_norm": 2.6418268330958146, + "learning_rate": 7.113490232342381e-07, + "loss": 0.0325, + "step": 10019 + }, + { + "epoch": 2.279863481228669, + "grad_norm": 1.8574341568779602, + "learning_rate": 7.112605278919163e-07, + "loss": 0.0733, + "step": 10020 + }, + { + "epoch": 2.280091012514221, + "grad_norm": 1.3796562818632219, + "learning_rate": 7.111720307865872e-07, + "loss": 0.1017, + "step": 10021 + }, + { + "epoch": 2.2803185437997726, + "grad_norm": 3.0821744563477265, + "learning_rate": 7.110835319200598e-07, + "loss": 0.0544, + "step": 10022 + }, + { + "epoch": 2.2805460750853244, + "grad_norm": 1.669297258209566, + "learning_rate": 7.109950312941426e-07, + "loss": 0.0952, + "step": 10023 + }, + { + "epoch": 2.280773606370876, + "grad_norm": 2.0627172751780956, + "learning_rate": 7.109065289106443e-07, + "loss": 0.1216, + "step": 10024 + }, + { + "epoch": 2.281001137656428, + "grad_norm": 1.207721341919061, + "learning_rate": 7.10818024771374e-07, + "loss": 0.0216, + "step": 10025 + }, + { + "epoch": 2.2812286689419796, + "grad_norm": 1.6718852900111274, + "learning_rate": 7.107295188781406e-07, + "loss": 0.0838, + "step": 10026 + }, + { + "epoch": 2.2814562002275314, + "grad_norm": 1.139793068897223, + "learning_rate": 7.106410112327526e-07, + "loss": 0.0447, + "step": 10027 + }, + { + "epoch": 2.281683731513083, + "grad_norm": 1.3475479150898466, + "learning_rate": 7.105525018370193e-07, + "loss": 0.0773, + "step": 10028 + }, + { + "epoch": 2.281911262798635, + "grad_norm": 1.1890249790286738, + "learning_rate": 7.104639906927495e-07, + "loss": 0.0419, + "step": 10029 + }, + { + "epoch": 2.2821387940841866, + "grad_norm": 0.9030696064001631, + "learning_rate": 7.103754778017522e-07, + "loss": 0.0243, + "step": 10030 + }, + { + "epoch": 2.2823663253697384, + "grad_norm": 1.9060237921462815, + "learning_rate": 7.102869631658366e-07, + "loss": 0.0857, + "step": 10031 + }, + { + "epoch": 2.28259385665529, + "grad_norm": 1.1331184828802838, + "learning_rate": 7.101984467868117e-07, + "loss": 0.0765, + "step": 10032 + }, + { + "epoch": 2.282821387940842, + "grad_norm": 4.987682033694193, + "learning_rate": 7.101099286664864e-07, + "loss": 0.1121, + "step": 10033 + }, + { + "epoch": 2.2830489192263936, + "grad_norm": 1.3189295840667192, + "learning_rate": 7.100214088066701e-07, + "loss": 0.0292, + "step": 10034 + }, + { + "epoch": 2.2832764505119454, + "grad_norm": 1.992738152444656, + "learning_rate": 7.09932887209172e-07, + "loss": 0.0824, + "step": 10035 + }, + { + "epoch": 2.283503981797497, + "grad_norm": 1.2439271811185566, + "learning_rate": 7.098443638758011e-07, + "loss": 0.0374, + "step": 10036 + }, + { + "epoch": 2.283731513083049, + "grad_norm": 1.449787819190392, + "learning_rate": 7.097558388083669e-07, + "loss": 0.0513, + "step": 10037 + }, + { + "epoch": 2.2839590443686006, + "grad_norm": 2.03306034401581, + "learning_rate": 7.096673120086786e-07, + "loss": 0.0923, + "step": 10038 + }, + { + "epoch": 2.2841865756541524, + "grad_norm": 1.2820139309970926, + "learning_rate": 7.095787834785454e-07, + "loss": 0.0831, + "step": 10039 + }, + { + "epoch": 2.284414106939704, + "grad_norm": 2.3143883924302924, + "learning_rate": 7.094902532197768e-07, + "loss": 0.0592, + "step": 10040 + }, + { + "epoch": 2.284641638225256, + "grad_norm": 1.96797682709101, + "learning_rate": 7.094017212341821e-07, + "loss": 0.0428, + "step": 10041 + }, + { + "epoch": 2.2848691695108077, + "grad_norm": 2.0363815858478587, + "learning_rate": 7.093131875235709e-07, + "loss": 0.0329, + "step": 10042 + }, + { + "epoch": 2.2850967007963594, + "grad_norm": 1.5373732609424149, + "learning_rate": 7.092246520897525e-07, + "loss": 0.0434, + "step": 10043 + }, + { + "epoch": 2.285324232081911, + "grad_norm": 1.1273123585609461, + "learning_rate": 7.091361149345364e-07, + "loss": 0.0703, + "step": 10044 + }, + { + "epoch": 2.285551763367463, + "grad_norm": 1.7123497664607472, + "learning_rate": 7.090475760597323e-07, + "loss": 0.0404, + "step": 10045 + }, + { + "epoch": 2.2857792946530147, + "grad_norm": 1.6485543468584964, + "learning_rate": 7.089590354671496e-07, + "loss": 0.0305, + "step": 10046 + }, + { + "epoch": 2.2860068259385664, + "grad_norm": 1.430496169992461, + "learning_rate": 7.088704931585981e-07, + "loss": 0.0712, + "step": 10047 + }, + { + "epoch": 2.286234357224118, + "grad_norm": 1.5768349646484414, + "learning_rate": 7.087819491358871e-07, + "loss": 0.0426, + "step": 10048 + }, + { + "epoch": 2.28646188850967, + "grad_norm": 1.0949635311954438, + "learning_rate": 7.086934034008268e-07, + "loss": 0.0229, + "step": 10049 + }, + { + "epoch": 2.2866894197952217, + "grad_norm": 1.493275805313581, + "learning_rate": 7.086048559552265e-07, + "loss": 0.0467, + "step": 10050 + }, + { + "epoch": 2.2869169510807734, + "grad_norm": 1.1998635039426, + "learning_rate": 7.08516306800896e-07, + "loss": 0.0803, + "step": 10051 + }, + { + "epoch": 2.287144482366325, + "grad_norm": 1.1485306499221362, + "learning_rate": 7.084277559396452e-07, + "loss": 0.0274, + "step": 10052 + }, + { + "epoch": 2.2873720136518774, + "grad_norm": 1.4765522299693974, + "learning_rate": 7.083392033732839e-07, + "loss": 0.0816, + "step": 10053 + }, + { + "epoch": 2.2875995449374287, + "grad_norm": 3.201735581076717, + "learning_rate": 7.082506491036216e-07, + "loss": 0.1219, + "step": 10054 + }, + { + "epoch": 2.287827076222981, + "grad_norm": 1.8492532949073477, + "learning_rate": 7.081620931324687e-07, + "loss": 0.0745, + "step": 10055 + }, + { + "epoch": 2.288054607508532, + "grad_norm": 1.0218501279370673, + "learning_rate": 7.080735354616349e-07, + "loss": 0.1123, + "step": 10056 + }, + { + "epoch": 2.2882821387940844, + "grad_norm": 1.5532113854842495, + "learning_rate": 7.079849760929304e-07, + "loss": 0.0358, + "step": 10057 + }, + { + "epoch": 2.288509670079636, + "grad_norm": 0.9743767439398741, + "learning_rate": 7.078964150281647e-07, + "loss": 0.0769, + "step": 10058 + }, + { + "epoch": 2.288737201365188, + "grad_norm": 1.3212847161018841, + "learning_rate": 7.078078522691482e-07, + "loss": 0.0921, + "step": 10059 + }, + { + "epoch": 2.2889647326507396, + "grad_norm": 1.9147415766072364, + "learning_rate": 7.07719287817691e-07, + "loss": 0.0441, + "step": 10060 + }, + { + "epoch": 2.2891922639362914, + "grad_norm": 1.1039619524571742, + "learning_rate": 7.076307216756028e-07, + "loss": 0.0322, + "step": 10061 + }, + { + "epoch": 2.289419795221843, + "grad_norm": 2.3783154417331445, + "learning_rate": 7.07542153844694e-07, + "loss": 0.0593, + "step": 10062 + }, + { + "epoch": 2.289647326507395, + "grad_norm": 2.422361220063779, + "learning_rate": 7.074535843267749e-07, + "loss": 0.1003, + "step": 10063 + }, + { + "epoch": 2.2898748577929466, + "grad_norm": 1.7141078067367503, + "learning_rate": 7.073650131236555e-07, + "loss": 0.1274, + "step": 10064 + }, + { + "epoch": 2.2901023890784984, + "grad_norm": 1.7069402366544915, + "learning_rate": 7.07276440237146e-07, + "loss": 0.0991, + "step": 10065 + }, + { + "epoch": 2.29032992036405, + "grad_norm": 1.6129815915034778, + "learning_rate": 7.071878656690567e-07, + "loss": 0.0363, + "step": 10066 + }, + { + "epoch": 2.290557451649602, + "grad_norm": 1.6072817085648465, + "learning_rate": 7.070992894211981e-07, + "loss": 0.141, + "step": 10067 + }, + { + "epoch": 2.2907849829351536, + "grad_norm": 2.398030301241516, + "learning_rate": 7.070107114953802e-07, + "loss": 0.0602, + "step": 10068 + }, + { + "epoch": 2.2910125142207054, + "grad_norm": 1.9998561195581437, + "learning_rate": 7.069221318934137e-07, + "loss": 0.0511, + "step": 10069 + }, + { + "epoch": 2.291240045506257, + "grad_norm": 2.8922879413023708, + "learning_rate": 7.068335506171089e-07, + "loss": 0.0468, + "step": 10070 + }, + { + "epoch": 2.291467576791809, + "grad_norm": 0.8405600450433354, + "learning_rate": 7.067449676682761e-07, + "loss": 0.0325, + "step": 10071 + }, + { + "epoch": 2.2916951080773607, + "grad_norm": 1.147158665440208, + "learning_rate": 7.066563830487259e-07, + "loss": 0.0425, + "step": 10072 + }, + { + "epoch": 2.2919226393629124, + "grad_norm": 1.054979192201, + "learning_rate": 7.065677967602688e-07, + "loss": 0.1022, + "step": 10073 + }, + { + "epoch": 2.292150170648464, + "grad_norm": 3.2704257511897694, + "learning_rate": 7.064792088047151e-07, + "loss": 0.0712, + "step": 10074 + }, + { + "epoch": 2.292377701934016, + "grad_norm": 1.3241895742777792, + "learning_rate": 7.063906191838758e-07, + "loss": 0.1161, + "step": 10075 + }, + { + "epoch": 2.2926052332195677, + "grad_norm": 2.170950985447108, + "learning_rate": 7.063020278995615e-07, + "loss": 0.0512, + "step": 10076 + }, + { + "epoch": 2.2928327645051194, + "grad_norm": 5.0787296093196295, + "learning_rate": 7.062134349535822e-07, + "loss": 0.0417, + "step": 10077 + }, + { + "epoch": 2.293060295790671, + "grad_norm": 1.450519148716169, + "learning_rate": 7.061248403477493e-07, + "loss": 0.1035, + "step": 10078 + }, + { + "epoch": 2.293287827076223, + "grad_norm": 1.0335660522453471, + "learning_rate": 7.060362440838732e-07, + "loss": 0.0682, + "step": 10079 + }, + { + "epoch": 2.2935153583617747, + "grad_norm": 1.7219680743403527, + "learning_rate": 7.059476461637647e-07, + "loss": 0.0501, + "step": 10080 + }, + { + "epoch": 2.2937428896473264, + "grad_norm": 1.3395445975809865, + "learning_rate": 7.058590465892344e-07, + "loss": 0.0524, + "step": 10081 + }, + { + "epoch": 2.293970420932878, + "grad_norm": 1.6637046562372, + "learning_rate": 7.057704453620934e-07, + "loss": 0.1022, + "step": 10082 + }, + { + "epoch": 2.29419795221843, + "grad_norm": 1.458889969476899, + "learning_rate": 7.056818424841526e-07, + "loss": 0.0257, + "step": 10083 + }, + { + "epoch": 2.2944254835039817, + "grad_norm": 1.7554511970810964, + "learning_rate": 7.055932379572225e-07, + "loss": 0.0065, + "step": 10084 + }, + { + "epoch": 2.2946530147895334, + "grad_norm": 2.1773005283315436, + "learning_rate": 7.055046317831142e-07, + "loss": 0.0702, + "step": 10085 + }, + { + "epoch": 2.294880546075085, + "grad_norm": 4.317779856586926, + "learning_rate": 7.054160239636387e-07, + "loss": 0.0618, + "step": 10086 + }, + { + "epoch": 2.295108077360637, + "grad_norm": 1.7896226389962315, + "learning_rate": 7.05327414500607e-07, + "loss": 0.0288, + "step": 10087 + }, + { + "epoch": 2.2953356086461887, + "grad_norm": 1.3495030196993232, + "learning_rate": 7.052388033958299e-07, + "loss": 0.1267, + "step": 10088 + }, + { + "epoch": 2.2955631399317404, + "grad_norm": 3.1537967131140094, + "learning_rate": 7.051501906511189e-07, + "loss": 0.077, + "step": 10089 + }, + { + "epoch": 2.295790671217292, + "grad_norm": 1.3866776838236015, + "learning_rate": 7.050615762682845e-07, + "loss": 0.0212, + "step": 10090 + }, + { + "epoch": 2.296018202502844, + "grad_norm": 1.712567216083561, + "learning_rate": 7.049729602491381e-07, + "loss": 0.0691, + "step": 10091 + }, + { + "epoch": 2.296245733788396, + "grad_norm": 1.8720464388897193, + "learning_rate": 7.048843425954911e-07, + "loss": 0.0773, + "step": 10092 + }, + { + "epoch": 2.2964732650739474, + "grad_norm": 1.7376347225843973, + "learning_rate": 7.047957233091543e-07, + "loss": 0.0369, + "step": 10093 + }, + { + "epoch": 2.2967007963594996, + "grad_norm": 1.5352656131792628, + "learning_rate": 7.047071023919391e-07, + "loss": 0.0437, + "step": 10094 + }, + { + "epoch": 2.296928327645051, + "grad_norm": 1.4800851511735222, + "learning_rate": 7.046184798456566e-07, + "loss": 0.0513, + "step": 10095 + }, + { + "epoch": 2.297155858930603, + "grad_norm": 0.9481631934344069, + "learning_rate": 7.045298556721184e-07, + "loss": 0.0245, + "step": 10096 + }, + { + "epoch": 2.297383390216155, + "grad_norm": 2.059395352873947, + "learning_rate": 7.044412298731354e-07, + "loss": 0.0654, + "step": 10097 + }, + { + "epoch": 2.2976109215017066, + "grad_norm": 1.566533347687481, + "learning_rate": 7.043526024505191e-07, + "loss": 0.0959, + "step": 10098 + }, + { + "epoch": 2.2978384527872584, + "grad_norm": 1.98758901008803, + "learning_rate": 7.042639734060811e-07, + "loss": 0.0572, + "step": 10099 + }, + { + "epoch": 2.29806598407281, + "grad_norm": 0.8716641372376241, + "learning_rate": 7.041753427416326e-07, + "loss": 0.0165, + "step": 10100 + }, + { + "epoch": 2.298293515358362, + "grad_norm": 1.7701070819210463, + "learning_rate": 7.040867104589852e-07, + "loss": 0.044, + "step": 10101 + }, + { + "epoch": 2.2985210466439137, + "grad_norm": 1.191527553532166, + "learning_rate": 7.0399807655995e-07, + "loss": 0.0244, + "step": 10102 + }, + { + "epoch": 2.2987485779294654, + "grad_norm": 1.6399352190142107, + "learning_rate": 7.03909441046339e-07, + "loss": 0.0504, + "step": 10103 + }, + { + "epoch": 2.298976109215017, + "grad_norm": 1.9999448132226203, + "learning_rate": 7.038208039199634e-07, + "loss": 0.0501, + "step": 10104 + }, + { + "epoch": 2.299203640500569, + "grad_norm": 1.3417488814515268, + "learning_rate": 7.037321651826351e-07, + "loss": 0.0403, + "step": 10105 + }, + { + "epoch": 2.2994311717861207, + "grad_norm": 3.9039504510784613, + "learning_rate": 7.036435248361655e-07, + "loss": 0.0347, + "step": 10106 + }, + { + "epoch": 2.2996587030716724, + "grad_norm": 2.3893293146461345, + "learning_rate": 7.035548828823662e-07, + "loss": 0.1475, + "step": 10107 + }, + { + "epoch": 2.299886234357224, + "grad_norm": 1.6916934633424234, + "learning_rate": 7.03466239323049e-07, + "loss": 0.0383, + "step": 10108 + }, + { + "epoch": 2.300113765642776, + "grad_norm": 3.4172866070423247, + "learning_rate": 7.033775941600257e-07, + "loss": 0.0622, + "step": 10109 + }, + { + "epoch": 2.3003412969283277, + "grad_norm": 1.6726203425653483, + "learning_rate": 7.032889473951078e-07, + "loss": 0.039, + "step": 10110 + }, + { + "epoch": 2.3005688282138794, + "grad_norm": 2.33938310576764, + "learning_rate": 7.032002990301071e-07, + "loss": 0.0599, + "step": 10111 + }, + { + "epoch": 2.300796359499431, + "grad_norm": 4.091327884543059, + "learning_rate": 7.031116490668355e-07, + "loss": 0.0838, + "step": 10112 + }, + { + "epoch": 2.301023890784983, + "grad_norm": 2.5933876046357214, + "learning_rate": 7.030229975071049e-07, + "loss": 0.0486, + "step": 10113 + }, + { + "epoch": 2.3012514220705347, + "grad_norm": 1.874560356299034, + "learning_rate": 7.029343443527273e-07, + "loss": 0.0466, + "step": 10114 + }, + { + "epoch": 2.3014789533560864, + "grad_norm": 2.2935452149335642, + "learning_rate": 7.028456896055143e-07, + "loss": 0.0768, + "step": 10115 + }, + { + "epoch": 2.301706484641638, + "grad_norm": 2.688109074386162, + "learning_rate": 7.02757033267278e-07, + "loss": 0.0786, + "step": 10116 + }, + { + "epoch": 2.30193401592719, + "grad_norm": 1.1151270718575637, + "learning_rate": 7.026683753398303e-07, + "loss": 0.0631, + "step": 10117 + }, + { + "epoch": 2.3021615472127417, + "grad_norm": 1.3473426685731544, + "learning_rate": 7.02579715824983e-07, + "loss": 0.0905, + "step": 10118 + }, + { + "epoch": 2.3023890784982934, + "grad_norm": 1.1242956105928719, + "learning_rate": 7.024910547245488e-07, + "loss": 0.0143, + "step": 10119 + }, + { + "epoch": 2.302616609783845, + "grad_norm": 1.870730120407611, + "learning_rate": 7.024023920403395e-07, + "loss": 0.024, + "step": 10120 + }, + { + "epoch": 2.302844141069397, + "grad_norm": 1.1529219808523674, + "learning_rate": 7.023137277741665e-07, + "loss": 0.0558, + "step": 10121 + }, + { + "epoch": 2.3030716723549487, + "grad_norm": 1.839284993413365, + "learning_rate": 7.022250619278428e-07, + "loss": 0.0594, + "step": 10122 + }, + { + "epoch": 2.3032992036405004, + "grad_norm": 1.0954860678635077, + "learning_rate": 7.021363945031804e-07, + "loss": 0.0236, + "step": 10123 + }, + { + "epoch": 2.303526734926052, + "grad_norm": 1.5779464396709175, + "learning_rate": 7.020477255019911e-07, + "loss": 0.0506, + "step": 10124 + }, + { + "epoch": 2.303754266211604, + "grad_norm": 1.9392962932491264, + "learning_rate": 7.019590549260874e-07, + "loss": 0.0737, + "step": 10125 + }, + { + "epoch": 2.3039817974971557, + "grad_norm": 1.8043832425297857, + "learning_rate": 7.018703827772816e-07, + "loss": 0.0635, + "step": 10126 + }, + { + "epoch": 2.3042093287827075, + "grad_norm": 1.4711456805325058, + "learning_rate": 7.017817090573863e-07, + "loss": 0.0183, + "step": 10127 + }, + { + "epoch": 2.304436860068259, + "grad_norm": 2.2512359681063963, + "learning_rate": 7.016930337682131e-07, + "loss": 0.1504, + "step": 10128 + }, + { + "epoch": 2.304664391353811, + "grad_norm": 1.677884166750546, + "learning_rate": 7.016043569115747e-07, + "loss": 0.0738, + "step": 10129 + }, + { + "epoch": 2.3048919226393627, + "grad_norm": 1.854098351346858, + "learning_rate": 7.015156784892838e-07, + "loss": 0.0667, + "step": 10130 + }, + { + "epoch": 2.305119453924915, + "grad_norm": 1.693603994712442, + "learning_rate": 7.014269985031523e-07, + "loss": 0.0725, + "step": 10131 + }, + { + "epoch": 2.305346985210466, + "grad_norm": 2.06112326539677, + "learning_rate": 7.01338316954993e-07, + "loss": 0.0728, + "step": 10132 + }, + { + "epoch": 2.3055745164960184, + "grad_norm": 1.2319994558308323, + "learning_rate": 7.012496338466186e-07, + "loss": 0.0208, + "step": 10133 + }, + { + "epoch": 2.3058020477815697, + "grad_norm": 1.1966728922907925, + "learning_rate": 7.01160949179841e-07, + "loss": 0.0136, + "step": 10134 + }, + { + "epoch": 2.306029579067122, + "grad_norm": 1.595694119966933, + "learning_rate": 7.010722629564732e-07, + "loss": 0.0335, + "step": 10135 + }, + { + "epoch": 2.3062571103526737, + "grad_norm": 1.3753962254891392, + "learning_rate": 7.009835751783277e-07, + "loss": 0.0374, + "step": 10136 + }, + { + "epoch": 2.3064846416382254, + "grad_norm": 3.0798061098172087, + "learning_rate": 7.00894885847217e-07, + "loss": 0.0793, + "step": 10137 + }, + { + "epoch": 2.306712172923777, + "grad_norm": 0.9753511578079058, + "learning_rate": 7.008061949649537e-07, + "loss": 0.0082, + "step": 10138 + }, + { + "epoch": 2.306939704209329, + "grad_norm": 1.1312077766122042, + "learning_rate": 7.00717502533351e-07, + "loss": 0.0657, + "step": 10139 + }, + { + "epoch": 2.3071672354948807, + "grad_norm": 1.2848887656888683, + "learning_rate": 7.00628808554221e-07, + "loss": 0.0403, + "step": 10140 + }, + { + "epoch": 2.3073947667804324, + "grad_norm": 1.832572487338452, + "learning_rate": 7.005401130293765e-07, + "loss": 0.0457, + "step": 10141 + }, + { + "epoch": 2.307622298065984, + "grad_norm": 1.0083600291135606, + "learning_rate": 7.004514159606307e-07, + "loss": 0.0603, + "step": 10142 + }, + { + "epoch": 2.307849829351536, + "grad_norm": 1.4021911649167007, + "learning_rate": 7.00362717349796e-07, + "loss": 0.0664, + "step": 10143 + }, + { + "epoch": 2.3080773606370877, + "grad_norm": 1.7458432050105481, + "learning_rate": 7.002740171986853e-07, + "loss": 0.0351, + "step": 10144 + }, + { + "epoch": 2.3083048919226394, + "grad_norm": 1.124439338286419, + "learning_rate": 7.001853155091117e-07, + "loss": 0.0584, + "step": 10145 + }, + { + "epoch": 2.308532423208191, + "grad_norm": 1.138275465729119, + "learning_rate": 7.00096612282888e-07, + "loss": 0.0856, + "step": 10146 + }, + { + "epoch": 2.308759954493743, + "grad_norm": 2.0102108957457134, + "learning_rate": 7.000079075218269e-07, + "loss": 0.0256, + "step": 10147 + }, + { + "epoch": 2.3089874857792947, + "grad_norm": 1.9643995708564213, + "learning_rate": 6.999192012277416e-07, + "loss": 0.0562, + "step": 10148 + }, + { + "epoch": 2.3092150170648464, + "grad_norm": 2.025784661414506, + "learning_rate": 6.99830493402445e-07, + "loss": 0.0352, + "step": 10149 + }, + { + "epoch": 2.309442548350398, + "grad_norm": 2.254375944695357, + "learning_rate": 6.997417840477502e-07, + "loss": 0.0765, + "step": 10150 + }, + { + "epoch": 2.30967007963595, + "grad_norm": 1.7845894886209006, + "learning_rate": 6.996530731654704e-07, + "loss": 0.0879, + "step": 10151 + }, + { + "epoch": 2.3098976109215017, + "grad_norm": 1.4978389960366234, + "learning_rate": 6.995643607574182e-07, + "loss": 0.0643, + "step": 10152 + }, + { + "epoch": 2.3101251422070535, + "grad_norm": 3.0634914492451153, + "learning_rate": 6.994756468254072e-07, + "loss": 0.0674, + "step": 10153 + }, + { + "epoch": 2.310352673492605, + "grad_norm": 1.553686778765929, + "learning_rate": 6.993869313712504e-07, + "loss": 0.1393, + "step": 10154 + }, + { + "epoch": 2.310580204778157, + "grad_norm": 0.8618737168268622, + "learning_rate": 6.992982143967607e-07, + "loss": 0.0284, + "step": 10155 + }, + { + "epoch": 2.3108077360637087, + "grad_norm": 0.7854142579646235, + "learning_rate": 6.992094959037518e-07, + "loss": 0.0595, + "step": 10156 + }, + { + "epoch": 2.3110352673492605, + "grad_norm": 1.3907922375502866, + "learning_rate": 6.991207758940367e-07, + "loss": 0.0631, + "step": 10157 + }, + { + "epoch": 2.311262798634812, + "grad_norm": 1.3702732962873834, + "learning_rate": 6.990320543694287e-07, + "loss": 0.0381, + "step": 10158 + }, + { + "epoch": 2.311490329920364, + "grad_norm": 1.048321945459076, + "learning_rate": 6.98943331331741e-07, + "loss": 0.1012, + "step": 10159 + }, + { + "epoch": 2.3117178612059157, + "grad_norm": 1.8533975651346535, + "learning_rate": 6.988546067827872e-07, + "loss": 0.1625, + "step": 10160 + }, + { + "epoch": 2.3119453924914675, + "grad_norm": 1.0225727774122166, + "learning_rate": 6.987658807243803e-07, + "loss": 0.0508, + "step": 10161 + }, + { + "epoch": 2.312172923777019, + "grad_norm": 1.6201709012239547, + "learning_rate": 6.986771531583339e-07, + "loss": 0.1398, + "step": 10162 + }, + { + "epoch": 2.312400455062571, + "grad_norm": 1.8130798331051952, + "learning_rate": 6.985884240864614e-07, + "loss": 0.0778, + "step": 10163 + }, + { + "epoch": 2.3126279863481227, + "grad_norm": 1.774029908656697, + "learning_rate": 6.984996935105765e-07, + "loss": 0.0541, + "step": 10164 + }, + { + "epoch": 2.3128555176336745, + "grad_norm": 1.4729286003642776, + "learning_rate": 6.984109614324923e-07, + "loss": 0.0911, + "step": 10165 + }, + { + "epoch": 2.3130830489192262, + "grad_norm": 2.3564239816606407, + "learning_rate": 6.983222278540225e-07, + "loss": 0.1191, + "step": 10166 + }, + { + "epoch": 2.313310580204778, + "grad_norm": 1.2995952701598417, + "learning_rate": 6.982334927769807e-07, + "loss": 0.0581, + "step": 10167 + }, + { + "epoch": 2.3135381114903297, + "grad_norm": 2.1883321448395154, + "learning_rate": 6.981447562031804e-07, + "loss": 0.1078, + "step": 10168 + }, + { + "epoch": 2.3137656427758815, + "grad_norm": 2.292356538303416, + "learning_rate": 6.980560181344352e-07, + "loss": 0.1524, + "step": 10169 + }, + { + "epoch": 2.3139931740614337, + "grad_norm": 1.1786745483805585, + "learning_rate": 6.979672785725588e-07, + "loss": 0.0543, + "step": 10170 + }, + { + "epoch": 2.314220705346985, + "grad_norm": 1.628133287353302, + "learning_rate": 6.97878537519365e-07, + "loss": 0.0975, + "step": 10171 + }, + { + "epoch": 2.314448236632537, + "grad_norm": 1.115136308714282, + "learning_rate": 6.977897949766673e-07, + "loss": 0.0918, + "step": 10172 + }, + { + "epoch": 2.3146757679180885, + "grad_norm": 1.097723287992714, + "learning_rate": 6.977010509462795e-07, + "loss": 0.0925, + "step": 10173 + }, + { + "epoch": 2.3149032992036407, + "grad_norm": 2.046566107351678, + "learning_rate": 6.976123054300153e-07, + "loss": 0.0507, + "step": 10174 + }, + { + "epoch": 2.3151308304891924, + "grad_norm": 1.4087649609051183, + "learning_rate": 6.975235584296884e-07, + "loss": 0.1052, + "step": 10175 + }, + { + "epoch": 2.315358361774744, + "grad_norm": 1.6218734642931316, + "learning_rate": 6.97434809947113e-07, + "loss": 0.0525, + "step": 10176 + }, + { + "epoch": 2.315585893060296, + "grad_norm": 1.8035795155258856, + "learning_rate": 6.973460599841029e-07, + "loss": 0.0809, + "step": 10177 + }, + { + "epoch": 2.3158134243458477, + "grad_norm": 1.6433672023606338, + "learning_rate": 6.972573085424715e-07, + "loss": 0.0698, + "step": 10178 + }, + { + "epoch": 2.3160409556313994, + "grad_norm": 2.0510882021247077, + "learning_rate": 6.971685556240331e-07, + "loss": 0.0596, + "step": 10179 + }, + { + "epoch": 2.316268486916951, + "grad_norm": 1.2412693951098324, + "learning_rate": 6.970798012306018e-07, + "loss": 0.0868, + "step": 10180 + }, + { + "epoch": 2.316496018202503, + "grad_norm": 1.528722328660584, + "learning_rate": 6.969910453639912e-07, + "loss": 0.1172, + "step": 10181 + }, + { + "epoch": 2.3167235494880547, + "grad_norm": 2.2801723786674217, + "learning_rate": 6.969022880260155e-07, + "loss": 0.0233, + "step": 10182 + }, + { + "epoch": 2.3169510807736065, + "grad_norm": 2.743946029399533, + "learning_rate": 6.968135292184889e-07, + "loss": 0.0475, + "step": 10183 + }, + { + "epoch": 2.317178612059158, + "grad_norm": 1.1287020117603226, + "learning_rate": 6.967247689432252e-07, + "loss": 0.0601, + "step": 10184 + }, + { + "epoch": 2.31740614334471, + "grad_norm": 2.6164682812761852, + "learning_rate": 6.966360072020384e-07, + "loss": 0.0769, + "step": 10185 + }, + { + "epoch": 2.3176336746302617, + "grad_norm": 1.4709909524809845, + "learning_rate": 6.965472439967428e-07, + "loss": 0.032, + "step": 10186 + }, + { + "epoch": 2.3178612059158135, + "grad_norm": 1.334787807805681, + "learning_rate": 6.964584793291527e-07, + "loss": 0.0662, + "step": 10187 + }, + { + "epoch": 2.318088737201365, + "grad_norm": 1.2806736191782186, + "learning_rate": 6.963697132010822e-07, + "loss": 0.0579, + "step": 10188 + }, + { + "epoch": 2.318316268486917, + "grad_norm": 2.1529014904395396, + "learning_rate": 6.962809456143453e-07, + "loss": 0.0572, + "step": 10189 + }, + { + "epoch": 2.3185437997724687, + "grad_norm": 1.6323391990780214, + "learning_rate": 6.961921765707567e-07, + "loss": 0.0922, + "step": 10190 + }, + { + "epoch": 2.3187713310580205, + "grad_norm": 1.138389230455129, + "learning_rate": 6.961034060721303e-07, + "loss": 0.0391, + "step": 10191 + }, + { + "epoch": 2.318998862343572, + "grad_norm": 1.3181284758375673, + "learning_rate": 6.960146341202802e-07, + "loss": 0.1295, + "step": 10192 + }, + { + "epoch": 2.319226393629124, + "grad_norm": 1.6807350928785871, + "learning_rate": 6.959258607170213e-07, + "loss": 0.026, + "step": 10193 + }, + { + "epoch": 2.3194539249146757, + "grad_norm": 1.26778168705304, + "learning_rate": 6.958370858641676e-07, + "loss": 0.078, + "step": 10194 + }, + { + "epoch": 2.3196814562002275, + "grad_norm": 1.2176158924301097, + "learning_rate": 6.957483095635335e-07, + "loss": 0.0882, + "step": 10195 + }, + { + "epoch": 2.3199089874857792, + "grad_norm": 1.257297282404605, + "learning_rate": 6.956595318169339e-07, + "loss": 0.1025, + "step": 10196 + }, + { + "epoch": 2.320136518771331, + "grad_norm": 1.9543429712640359, + "learning_rate": 6.955707526261826e-07, + "loss": 0.047, + "step": 10197 + }, + { + "epoch": 2.3203640500568827, + "grad_norm": 1.6798414963335275, + "learning_rate": 6.954819719930944e-07, + "loss": 0.0603, + "step": 10198 + }, + { + "epoch": 2.3205915813424345, + "grad_norm": 1.687521388398465, + "learning_rate": 6.953931899194838e-07, + "loss": 0.0584, + "step": 10199 + }, + { + "epoch": 2.3208191126279862, + "grad_norm": 1.8006313152407738, + "learning_rate": 6.953044064071653e-07, + "loss": 0.0943, + "step": 10200 + }, + { + "epoch": 2.321046643913538, + "grad_norm": 2.0515418147475764, + "learning_rate": 6.952156214579535e-07, + "loss": 0.0726, + "step": 10201 + }, + { + "epoch": 2.3212741751990897, + "grad_norm": 1.4252738231772164, + "learning_rate": 6.95126835073663e-07, + "loss": 0.0378, + "step": 10202 + }, + { + "epoch": 2.3215017064846415, + "grad_norm": 1.4816934761450637, + "learning_rate": 6.950380472561084e-07, + "loss": 0.0311, + "step": 10203 + }, + { + "epoch": 2.3217292377701932, + "grad_norm": 2.201113295909613, + "learning_rate": 6.949492580071044e-07, + "loss": 0.0399, + "step": 10204 + }, + { + "epoch": 2.321956769055745, + "grad_norm": 1.951993390816971, + "learning_rate": 6.948604673284655e-07, + "loss": 0.1165, + "step": 10205 + }, + { + "epoch": 2.3221843003412967, + "grad_norm": 1.9014797322372625, + "learning_rate": 6.947716752220069e-07, + "loss": 0.0814, + "step": 10206 + }, + { + "epoch": 2.3224118316268485, + "grad_norm": 1.7386795623318834, + "learning_rate": 6.946828816895428e-07, + "loss": 0.0365, + "step": 10207 + }, + { + "epoch": 2.3226393629124003, + "grad_norm": 1.4151791465942025, + "learning_rate": 6.945940867328883e-07, + "loss": 0.0396, + "step": 10208 + }, + { + "epoch": 2.3228668941979524, + "grad_norm": 2.0806909803120726, + "learning_rate": 6.945052903538582e-07, + "loss": 0.0498, + "step": 10209 + }, + { + "epoch": 2.3230944254835038, + "grad_norm": 1.3075092377123654, + "learning_rate": 6.944164925542672e-07, + "loss": 0.0909, + "step": 10210 + }, + { + "epoch": 2.323321956769056, + "grad_norm": 2.0010878949395643, + "learning_rate": 6.943276933359302e-07, + "loss": 0.0453, + "step": 10211 + }, + { + "epoch": 2.3235494880546073, + "grad_norm": 0.947341885880175, + "learning_rate": 6.94238892700662e-07, + "loss": 0.0202, + "step": 10212 + }, + { + "epoch": 2.3237770193401595, + "grad_norm": 1.1837389422504694, + "learning_rate": 6.941500906502778e-07, + "loss": 0.0261, + "step": 10213 + }, + { + "epoch": 2.324004550625711, + "grad_norm": 2.0267488048839684, + "learning_rate": 6.940612871865924e-07, + "loss": 0.0339, + "step": 10214 + }, + { + "epoch": 2.324232081911263, + "grad_norm": 1.1447099379950065, + "learning_rate": 6.939724823114206e-07, + "loss": 0.0521, + "step": 10215 + }, + { + "epoch": 2.3244596131968147, + "grad_norm": 1.1693330043500185, + "learning_rate": 6.938836760265778e-07, + "loss": 0.0981, + "step": 10216 + }, + { + "epoch": 2.3246871444823665, + "grad_norm": 1.6897289722976172, + "learning_rate": 6.937948683338787e-07, + "loss": 0.0825, + "step": 10217 + }, + { + "epoch": 2.324914675767918, + "grad_norm": 1.2395248507723478, + "learning_rate": 6.937060592351386e-07, + "loss": 0.0201, + "step": 10218 + }, + { + "epoch": 2.32514220705347, + "grad_norm": 1.9277027392111943, + "learning_rate": 6.936172487321722e-07, + "loss": 0.1542, + "step": 10219 + }, + { + "epoch": 2.3253697383390217, + "grad_norm": 1.2088750775925792, + "learning_rate": 6.935284368267951e-07, + "loss": 0.0565, + "step": 10220 + }, + { + "epoch": 2.3255972696245735, + "grad_norm": 1.8880837573650477, + "learning_rate": 6.934396235208224e-07, + "loss": 0.1229, + "step": 10221 + }, + { + "epoch": 2.3258248009101252, + "grad_norm": 1.1402610991841549, + "learning_rate": 6.933508088160689e-07, + "loss": 0.0786, + "step": 10222 + }, + { + "epoch": 2.326052332195677, + "grad_norm": 1.4819716197064237, + "learning_rate": 6.932619927143501e-07, + "loss": 0.0479, + "step": 10223 + }, + { + "epoch": 2.3262798634812287, + "grad_norm": 1.8724274607064795, + "learning_rate": 6.931731752174813e-07, + "loss": 0.0522, + "step": 10224 + }, + { + "epoch": 2.3265073947667805, + "grad_norm": 2.7618583509610133, + "learning_rate": 6.930843563272774e-07, + "loss": 0.0369, + "step": 10225 + }, + { + "epoch": 2.3267349260523322, + "grad_norm": 1.4961723432110914, + "learning_rate": 6.929955360455542e-07, + "loss": 0.0653, + "step": 10226 + }, + { + "epoch": 2.326962457337884, + "grad_norm": 1.2489142942883422, + "learning_rate": 6.929067143741267e-07, + "loss": 0.0176, + "step": 10227 + }, + { + "epoch": 2.3271899886234357, + "grad_norm": 1.8688671407267392, + "learning_rate": 6.928178913148101e-07, + "loss": 0.151, + "step": 10228 + }, + { + "epoch": 2.3274175199089875, + "grad_norm": 1.2547087381362572, + "learning_rate": 6.927290668694203e-07, + "loss": 0.0447, + "step": 10229 + }, + { + "epoch": 2.3276450511945392, + "grad_norm": 2.547317737539808, + "learning_rate": 6.926402410397723e-07, + "loss": 0.073, + "step": 10230 + }, + { + "epoch": 2.327872582480091, + "grad_norm": 2.212436434130153, + "learning_rate": 6.925514138276816e-07, + "loss": 0.1046, + "step": 10231 + }, + { + "epoch": 2.3281001137656427, + "grad_norm": 1.5109396462937263, + "learning_rate": 6.924625852349637e-07, + "loss": 0.0639, + "step": 10232 + }, + { + "epoch": 2.3283276450511945, + "grad_norm": 1.1343876043701355, + "learning_rate": 6.923737552634342e-07, + "loss": 0.0993, + "step": 10233 + }, + { + "epoch": 2.3285551763367462, + "grad_norm": 2.1883740880583913, + "learning_rate": 6.922849239149087e-07, + "loss": 0.0581, + "step": 10234 + }, + { + "epoch": 2.328782707622298, + "grad_norm": 2.3715039601045573, + "learning_rate": 6.921960911912024e-07, + "loss": 0.0779, + "step": 10235 + }, + { + "epoch": 2.3290102389078498, + "grad_norm": 1.1629067708971725, + "learning_rate": 6.921072570941311e-07, + "loss": 0.0381, + "step": 10236 + }, + { + "epoch": 2.3292377701934015, + "grad_norm": 1.3144806707499532, + "learning_rate": 6.920184216255102e-07, + "loss": 0.1268, + "step": 10237 + }, + { + "epoch": 2.3294653014789533, + "grad_norm": 1.3928998391045477, + "learning_rate": 6.919295847871557e-07, + "loss": 0.0693, + "step": 10238 + }, + { + "epoch": 2.329692832764505, + "grad_norm": 2.8158021702923874, + "learning_rate": 6.918407465808828e-07, + "loss": 0.0889, + "step": 10239 + }, + { + "epoch": 2.3299203640500568, + "grad_norm": 1.2226379251498047, + "learning_rate": 6.917519070085078e-07, + "loss": 0.0782, + "step": 10240 + }, + { + "epoch": 2.3301478953356085, + "grad_norm": 1.1831458582331267, + "learning_rate": 6.91663066071846e-07, + "loss": 0.0294, + "step": 10241 + }, + { + "epoch": 2.3303754266211603, + "grad_norm": 1.5032665133997754, + "learning_rate": 6.91574223772713e-07, + "loss": 0.0239, + "step": 10242 + }, + { + "epoch": 2.330602957906712, + "grad_norm": 1.7402310819181557, + "learning_rate": 6.914853801129249e-07, + "loss": 0.0997, + "step": 10243 + }, + { + "epoch": 2.3308304891922638, + "grad_norm": 0.9440754659422651, + "learning_rate": 6.913965350942975e-07, + "loss": 0.0656, + "step": 10244 + }, + { + "epoch": 2.3310580204778155, + "grad_norm": 1.5156113119001404, + "learning_rate": 6.913076887186465e-07, + "loss": 0.0386, + "step": 10245 + }, + { + "epoch": 2.3312855517633673, + "grad_norm": 1.1621760685679863, + "learning_rate": 6.912188409877876e-07, + "loss": 0.0593, + "step": 10246 + }, + { + "epoch": 2.331513083048919, + "grad_norm": 1.3907589276753458, + "learning_rate": 6.91129991903537e-07, + "loss": 0.11, + "step": 10247 + }, + { + "epoch": 2.331740614334471, + "grad_norm": 1.3054115370213695, + "learning_rate": 6.910411414677105e-07, + "loss": 0.0994, + "step": 10248 + }, + { + "epoch": 2.3319681456200225, + "grad_norm": 1.5335187923286828, + "learning_rate": 6.909522896821239e-07, + "loss": 0.0358, + "step": 10249 + }, + { + "epoch": 2.3321956769055747, + "grad_norm": 1.8568287158029235, + "learning_rate": 6.908634365485933e-07, + "loss": 0.0282, + "step": 10250 + }, + { + "epoch": 2.3324232081911265, + "grad_norm": 1.2864840099993982, + "learning_rate": 6.907745820689349e-07, + "loss": 0.0214, + "step": 10251 + }, + { + "epoch": 2.3326507394766782, + "grad_norm": 1.634727551940298, + "learning_rate": 6.906857262449642e-07, + "loss": 0.0866, + "step": 10252 + }, + { + "epoch": 2.33287827076223, + "grad_norm": 1.2768484301977154, + "learning_rate": 6.905968690784978e-07, + "loss": 0.0346, + "step": 10253 + }, + { + "epoch": 2.3331058020477817, + "grad_norm": 1.2471433914175802, + "learning_rate": 6.905080105713514e-07, + "loss": 0.0364, + "step": 10254 + }, + { + "epoch": 2.3333333333333335, + "grad_norm": 1.683253554991136, + "learning_rate": 6.904191507253412e-07, + "loss": 0.148, + "step": 10255 + }, + { + "epoch": 2.3335608646188852, + "grad_norm": 1.1810479667758818, + "learning_rate": 6.903302895422835e-07, + "loss": 0.1034, + "step": 10256 + }, + { + "epoch": 2.333788395904437, + "grad_norm": 1.518684420162794, + "learning_rate": 6.902414270239942e-07, + "loss": 0.1036, + "step": 10257 + }, + { + "epoch": 2.3340159271899887, + "grad_norm": 1.767095021711705, + "learning_rate": 6.901525631722896e-07, + "loss": 0.0405, + "step": 10258 + }, + { + "epoch": 2.3342434584755405, + "grad_norm": 1.2109482043628774, + "learning_rate": 6.900636979889861e-07, + "loss": 0.0558, + "step": 10259 + }, + { + "epoch": 2.3344709897610922, + "grad_norm": 1.2183266640422594, + "learning_rate": 6.899748314758997e-07, + "loss": 0.0386, + "step": 10260 + }, + { + "epoch": 2.334698521046644, + "grad_norm": 1.74579132112299, + "learning_rate": 6.898859636348468e-07, + "loss": 0.0557, + "step": 10261 + }, + { + "epoch": 2.3349260523321957, + "grad_norm": 2.504356833529423, + "learning_rate": 6.897970944676434e-07, + "loss": 0.0518, + "step": 10262 + }, + { + "epoch": 2.3351535836177475, + "grad_norm": 1.5461574518753687, + "learning_rate": 6.897082239761063e-07, + "loss": 0.1372, + "step": 10263 + }, + { + "epoch": 2.3353811149032992, + "grad_norm": 1.6835284694355497, + "learning_rate": 6.896193521620514e-07, + "loss": 0.032, + "step": 10264 + }, + { + "epoch": 2.335608646188851, + "grad_norm": 2.1364155279289823, + "learning_rate": 6.895304790272956e-07, + "loss": 0.0638, + "step": 10265 + }, + { + "epoch": 2.3358361774744028, + "grad_norm": 1.3660435833531537, + "learning_rate": 6.894416045736547e-07, + "loss": 0.0973, + "step": 10266 + }, + { + "epoch": 2.3360637087599545, + "grad_norm": 1.4432499747109002, + "learning_rate": 6.893527288029456e-07, + "loss": 0.0747, + "step": 10267 + }, + { + "epoch": 2.3362912400455063, + "grad_norm": 1.6956561497830758, + "learning_rate": 6.892638517169844e-07, + "loss": 0.1693, + "step": 10268 + }, + { + "epoch": 2.336518771331058, + "grad_norm": 1.6549834682911633, + "learning_rate": 6.891749733175879e-07, + "loss": 0.0667, + "step": 10269 + }, + { + "epoch": 2.3367463026166098, + "grad_norm": 1.4037229719452258, + "learning_rate": 6.890860936065724e-07, + "loss": 0.0892, + "step": 10270 + }, + { + "epoch": 2.3369738339021615, + "grad_norm": 1.2477080327455767, + "learning_rate": 6.889972125857547e-07, + "loss": 0.0572, + "step": 10271 + }, + { + "epoch": 2.3372013651877133, + "grad_norm": 1.813293932133548, + "learning_rate": 6.889083302569511e-07, + "loss": 0.1296, + "step": 10272 + }, + { + "epoch": 2.337428896473265, + "grad_norm": 1.719958752451596, + "learning_rate": 6.88819446621978e-07, + "loss": 0.1362, + "step": 10273 + }, + { + "epoch": 2.3376564277588168, + "grad_norm": 0.9069044556525777, + "learning_rate": 6.887305616826525e-07, + "loss": 0.0344, + "step": 10274 + }, + { + "epoch": 2.3378839590443685, + "grad_norm": 2.093450119488748, + "learning_rate": 6.886416754407912e-07, + "loss": 0.0539, + "step": 10275 + }, + { + "epoch": 2.3381114903299203, + "grad_norm": 1.4251842254677067, + "learning_rate": 6.885527878982103e-07, + "loss": 0.0466, + "step": 10276 + }, + { + "epoch": 2.338339021615472, + "grad_norm": 1.0877267410202085, + "learning_rate": 6.88463899056727e-07, + "loss": 0.0587, + "step": 10277 + }, + { + "epoch": 2.3385665529010238, + "grad_norm": 1.4686012432471598, + "learning_rate": 6.883750089181579e-07, + "loss": 0.0338, + "step": 10278 + }, + { + "epoch": 2.3387940841865755, + "grad_norm": 1.3154641679217756, + "learning_rate": 6.882861174843194e-07, + "loss": 0.0797, + "step": 10279 + }, + { + "epoch": 2.3390216154721273, + "grad_norm": 1.3542498985255194, + "learning_rate": 6.881972247570288e-07, + "loss": 0.0888, + "step": 10280 + }, + { + "epoch": 2.339249146757679, + "grad_norm": 2.303195591240399, + "learning_rate": 6.881083307381026e-07, + "loss": 0.0479, + "step": 10281 + }, + { + "epoch": 2.339476678043231, + "grad_norm": 1.4884105157363325, + "learning_rate": 6.880194354293577e-07, + "loss": 0.1355, + "step": 10282 + }, + { + "epoch": 2.3397042093287825, + "grad_norm": 1.459211571114225, + "learning_rate": 6.879305388326109e-07, + "loss": 0.0221, + "step": 10283 + }, + { + "epoch": 2.3399317406143343, + "grad_norm": 1.8900595358093086, + "learning_rate": 6.878416409496793e-07, + "loss": 0.1154, + "step": 10284 + }, + { + "epoch": 2.3401592718998865, + "grad_norm": 1.7133699210458022, + "learning_rate": 6.877527417823795e-07, + "loss": 0.0684, + "step": 10285 + }, + { + "epoch": 2.340386803185438, + "grad_norm": 2.351694662533878, + "learning_rate": 6.876638413325286e-07, + "loss": 0.0569, + "step": 10286 + }, + { + "epoch": 2.34061433447099, + "grad_norm": 2.4489783252942687, + "learning_rate": 6.875749396019435e-07, + "loss": 0.0939, + "step": 10287 + }, + { + "epoch": 2.3408418657565413, + "grad_norm": 2.620683841381813, + "learning_rate": 6.874860365924415e-07, + "loss": 0.1409, + "step": 10288 + }, + { + "epoch": 2.3410693970420935, + "grad_norm": 1.9696176755661552, + "learning_rate": 6.873971323058389e-07, + "loss": 0.1135, + "step": 10289 + }, + { + "epoch": 2.3412969283276452, + "grad_norm": 1.666515835233714, + "learning_rate": 6.873082267439536e-07, + "loss": 0.019, + "step": 10290 + }, + { + "epoch": 2.341524459613197, + "grad_norm": 0.9954620044682266, + "learning_rate": 6.87219319908602e-07, + "loss": 0.0415, + "step": 10291 + }, + { + "epoch": 2.3417519908987487, + "grad_norm": 1.2820102800365247, + "learning_rate": 6.871304118016015e-07, + "loss": 0.0491, + "step": 10292 + }, + { + "epoch": 2.3419795221843005, + "grad_norm": 1.7366173682816122, + "learning_rate": 6.870415024247693e-07, + "loss": 0.0473, + "step": 10293 + }, + { + "epoch": 2.3422070534698523, + "grad_norm": 2.1789347674413646, + "learning_rate": 6.869525917799223e-07, + "loss": 0.0835, + "step": 10294 + }, + { + "epoch": 2.342434584755404, + "grad_norm": 1.9513803531887899, + "learning_rate": 6.868636798688778e-07, + "loss": 0.0376, + "step": 10295 + }, + { + "epoch": 2.3426621160409558, + "grad_norm": 1.7117804503129268, + "learning_rate": 6.867747666934529e-07, + "loss": 0.0462, + "step": 10296 + }, + { + "epoch": 2.3428896473265075, + "grad_norm": 0.9348923089631415, + "learning_rate": 6.866858522554652e-07, + "loss": 0.0246, + "step": 10297 + }, + { + "epoch": 2.3431171786120593, + "grad_norm": 1.9836947542230896, + "learning_rate": 6.865969365567314e-07, + "loss": 0.0833, + "step": 10298 + }, + { + "epoch": 2.343344709897611, + "grad_norm": 2.112972121619996, + "learning_rate": 6.865080195990691e-07, + "loss": 0.0271, + "step": 10299 + }, + { + "epoch": 2.3435722411831628, + "grad_norm": 1.292114056225606, + "learning_rate": 6.864191013842955e-07, + "loss": 0.0773, + "step": 10300 + }, + { + "epoch": 2.3437997724687145, + "grad_norm": 1.1551357054172011, + "learning_rate": 6.863301819142279e-07, + "loss": 0.0215, + "step": 10301 + }, + { + "epoch": 2.3440273037542663, + "grad_norm": 2.0761561059966183, + "learning_rate": 6.862412611906838e-07, + "loss": 0.1052, + "step": 10302 + }, + { + "epoch": 2.344254835039818, + "grad_norm": 1.7150075711051953, + "learning_rate": 6.861523392154805e-07, + "loss": 0.0392, + "step": 10303 + }, + { + "epoch": 2.3444823663253698, + "grad_norm": 2.0297239687712425, + "learning_rate": 6.860634159904354e-07, + "loss": 0.0412, + "step": 10304 + }, + { + "epoch": 2.3447098976109215, + "grad_norm": 2.3473129194214954, + "learning_rate": 6.859744915173658e-07, + "loss": 0.1759, + "step": 10305 + }, + { + "epoch": 2.3449374288964733, + "grad_norm": 0.9927135538329924, + "learning_rate": 6.858855657980891e-07, + "loss": 0.0623, + "step": 10306 + }, + { + "epoch": 2.345164960182025, + "grad_norm": 1.0502860816941773, + "learning_rate": 6.857966388344232e-07, + "loss": 0.0986, + "step": 10307 + }, + { + "epoch": 2.345392491467577, + "grad_norm": 1.8586572672920925, + "learning_rate": 6.857077106281852e-07, + "loss": 0.0868, + "step": 10308 + }, + { + "epoch": 2.3456200227531285, + "grad_norm": 2.8671632145889974, + "learning_rate": 6.856187811811929e-07, + "loss": 0.05, + "step": 10309 + }, + { + "epoch": 2.3458475540386803, + "grad_norm": 1.6482087827889116, + "learning_rate": 6.855298504952637e-07, + "loss": 0.0471, + "step": 10310 + }, + { + "epoch": 2.346075085324232, + "grad_norm": 1.5900866985357793, + "learning_rate": 6.854409185722152e-07, + "loss": 0.0516, + "step": 10311 + }, + { + "epoch": 2.346302616609784, + "grad_norm": 1.0992061635391959, + "learning_rate": 6.853519854138652e-07, + "loss": 0.0721, + "step": 10312 + }, + { + "epoch": 2.3465301478953355, + "grad_norm": 1.297847017352353, + "learning_rate": 6.852630510220308e-07, + "loss": 0.0503, + "step": 10313 + }, + { + "epoch": 2.3467576791808873, + "grad_norm": 1.2115690449850958, + "learning_rate": 6.851741153985302e-07, + "loss": 0.039, + "step": 10314 + }, + { + "epoch": 2.346985210466439, + "grad_norm": 1.9199319473735892, + "learning_rate": 6.850851785451809e-07, + "loss": 0.0471, + "step": 10315 + }, + { + "epoch": 2.347212741751991, + "grad_norm": 1.5680436429643736, + "learning_rate": 6.849962404638005e-07, + "loss": 0.0197, + "step": 10316 + }, + { + "epoch": 2.3474402730375425, + "grad_norm": 3.0613039449943518, + "learning_rate": 6.849073011562069e-07, + "loss": 0.0542, + "step": 10317 + }, + { + "epoch": 2.3476678043230943, + "grad_norm": 1.502510452484883, + "learning_rate": 6.848183606242177e-07, + "loss": 0.1118, + "step": 10318 + }, + { + "epoch": 2.347895335608646, + "grad_norm": 1.273131133360515, + "learning_rate": 6.847294188696507e-07, + "loss": 0.0514, + "step": 10319 + }, + { + "epoch": 2.348122866894198, + "grad_norm": 1.731481576304848, + "learning_rate": 6.84640475894324e-07, + "loss": 0.0805, + "step": 10320 + }, + { + "epoch": 2.3483503981797496, + "grad_norm": 2.4925389495427104, + "learning_rate": 6.845515317000551e-07, + "loss": 0.0865, + "step": 10321 + }, + { + "epoch": 2.3485779294653013, + "grad_norm": 1.7572616660739149, + "learning_rate": 6.844625862886618e-07, + "loss": 0.1315, + "step": 10322 + }, + { + "epoch": 2.348805460750853, + "grad_norm": 1.8615518240805375, + "learning_rate": 6.843736396619622e-07, + "loss": 0.0476, + "step": 10323 + }, + { + "epoch": 2.3490329920364053, + "grad_norm": 1.978938155796057, + "learning_rate": 6.842846918217743e-07, + "loss": 0.0427, + "step": 10324 + }, + { + "epoch": 2.3492605233219566, + "grad_norm": 1.796764703808653, + "learning_rate": 6.841957427699158e-07, + "loss": 0.107, + "step": 10325 + }, + { + "epoch": 2.3494880546075088, + "grad_norm": 2.291946513424975, + "learning_rate": 6.841067925082046e-07, + "loss": 0.0971, + "step": 10326 + }, + { + "epoch": 2.34971558589306, + "grad_norm": 1.3626579338306246, + "learning_rate": 6.840178410384591e-07, + "loss": 0.0281, + "step": 10327 + }, + { + "epoch": 2.3499431171786123, + "grad_norm": 1.508767760007627, + "learning_rate": 6.839288883624969e-07, + "loss": 0.1248, + "step": 10328 + }, + { + "epoch": 2.350170648464164, + "grad_norm": 1.5004316046603756, + "learning_rate": 6.838399344821359e-07, + "loss": 0.0231, + "step": 10329 + }, + { + "epoch": 2.3503981797497158, + "grad_norm": 1.601205357241486, + "learning_rate": 6.837509793991946e-07, + "loss": 0.1039, + "step": 10330 + }, + { + "epoch": 2.3506257110352675, + "grad_norm": 1.3673686063479995, + "learning_rate": 6.836620231154908e-07, + "loss": 0.0402, + "step": 10331 + }, + { + "epoch": 2.3508532423208193, + "grad_norm": 1.6027769574096573, + "learning_rate": 6.835730656328429e-07, + "loss": 0.0382, + "step": 10332 + }, + { + "epoch": 2.351080773606371, + "grad_norm": 1.3406223894071378, + "learning_rate": 6.834841069530686e-07, + "loss": 0.0342, + "step": 10333 + }, + { + "epoch": 2.3513083048919228, + "grad_norm": 1.3144468817365327, + "learning_rate": 6.833951470779864e-07, + "loss": 0.0631, + "step": 10334 + }, + { + "epoch": 2.3515358361774745, + "grad_norm": 2.0359626840176595, + "learning_rate": 6.833061860094142e-07, + "loss": 0.065, + "step": 10335 + }, + { + "epoch": 2.3517633674630263, + "grad_norm": 1.6705499922072669, + "learning_rate": 6.832172237491703e-07, + "loss": 0.1039, + "step": 10336 + }, + { + "epoch": 2.351990898748578, + "grad_norm": 1.5039483741569506, + "learning_rate": 6.831282602990731e-07, + "loss": 0.0581, + "step": 10337 + }, + { + "epoch": 2.35221843003413, + "grad_norm": 0.8204124970938125, + "learning_rate": 6.830392956609406e-07, + "loss": 0.0966, + "step": 10338 + }, + { + "epoch": 2.3524459613196815, + "grad_norm": 2.162296188550802, + "learning_rate": 6.829503298365913e-07, + "loss": 0.1249, + "step": 10339 + }, + { + "epoch": 2.3526734926052333, + "grad_norm": 1.4952317784774078, + "learning_rate": 6.828613628278433e-07, + "loss": 0.0304, + "step": 10340 + }, + { + "epoch": 2.352901023890785, + "grad_norm": 1.802162035842996, + "learning_rate": 6.827723946365153e-07, + "loss": 0.1375, + "step": 10341 + }, + { + "epoch": 2.353128555176337, + "grad_norm": 1.5797018260595177, + "learning_rate": 6.826834252644251e-07, + "loss": 0.02, + "step": 10342 + }, + { + "epoch": 2.3533560864618885, + "grad_norm": 1.172831570434394, + "learning_rate": 6.825944547133913e-07, + "loss": 0.0199, + "step": 10343 + }, + { + "epoch": 2.3535836177474403, + "grad_norm": 1.5883685356610204, + "learning_rate": 6.825054829852323e-07, + "loss": 0.0408, + "step": 10344 + }, + { + "epoch": 2.353811149032992, + "grad_norm": 1.6460657275530246, + "learning_rate": 6.824165100817667e-07, + "loss": 0.106, + "step": 10345 + }, + { + "epoch": 2.354038680318544, + "grad_norm": 1.590402614110204, + "learning_rate": 6.823275360048126e-07, + "loss": 0.0361, + "step": 10346 + }, + { + "epoch": 2.3542662116040955, + "grad_norm": 1.6117724208524986, + "learning_rate": 6.822385607561889e-07, + "loss": 0.0455, + "step": 10347 + }, + { + "epoch": 2.3544937428896473, + "grad_norm": 1.5289786611280218, + "learning_rate": 6.821495843377138e-07, + "loss": 0.0348, + "step": 10348 + }, + { + "epoch": 2.354721274175199, + "grad_norm": 1.3615083313526894, + "learning_rate": 6.820606067512056e-07, + "loss": 0.0348, + "step": 10349 + }, + { + "epoch": 2.354948805460751, + "grad_norm": 1.1035158333778545, + "learning_rate": 6.819716279984833e-07, + "loss": 0.0224, + "step": 10350 + }, + { + "epoch": 2.3551763367463026, + "grad_norm": 1.6127580019064336, + "learning_rate": 6.818826480813652e-07, + "loss": 0.0402, + "step": 10351 + }, + { + "epoch": 2.3554038680318543, + "grad_norm": 1.0067043003137657, + "learning_rate": 6.8179366700167e-07, + "loss": 0.0143, + "step": 10352 + }, + { + "epoch": 2.355631399317406, + "grad_norm": 2.734268855996695, + "learning_rate": 6.817046847612164e-07, + "loss": 0.033, + "step": 10353 + }, + { + "epoch": 2.355858930602958, + "grad_norm": 1.4410312234429459, + "learning_rate": 6.816157013618227e-07, + "loss": 0.0384, + "step": 10354 + }, + { + "epoch": 2.3560864618885096, + "grad_norm": 1.3855904304090667, + "learning_rate": 6.815267168053078e-07, + "loss": 0.0453, + "step": 10355 + }, + { + "epoch": 2.3563139931740613, + "grad_norm": 2.7515947683773003, + "learning_rate": 6.814377310934901e-07, + "loss": 0.0574, + "step": 10356 + }, + { + "epoch": 2.356541524459613, + "grad_norm": 1.4031745709918233, + "learning_rate": 6.813487442281888e-07, + "loss": 0.0352, + "step": 10357 + }, + { + "epoch": 2.356769055745165, + "grad_norm": 2.8484808815150218, + "learning_rate": 6.812597562112223e-07, + "loss": 0.0626, + "step": 10358 + }, + { + "epoch": 2.3569965870307166, + "grad_norm": 1.534668915501644, + "learning_rate": 6.811707670444095e-07, + "loss": 0.0402, + "step": 10359 + }, + { + "epoch": 2.3572241183162683, + "grad_norm": 1.765601543389788, + "learning_rate": 6.810817767295691e-07, + "loss": 0.103, + "step": 10360 + }, + { + "epoch": 2.35745164960182, + "grad_norm": 1.7072836316197535, + "learning_rate": 6.809927852685198e-07, + "loss": 0.0764, + "step": 10361 + }, + { + "epoch": 2.357679180887372, + "grad_norm": 1.1410970983956845, + "learning_rate": 6.809037926630806e-07, + "loss": 0.0179, + "step": 10362 + }, + { + "epoch": 2.357906712172924, + "grad_norm": 0.8803541930038186, + "learning_rate": 6.808147989150701e-07, + "loss": 0.0799, + "step": 10363 + }, + { + "epoch": 2.3581342434584753, + "grad_norm": 1.6949544656573008, + "learning_rate": 6.807258040263075e-07, + "loss": 0.0574, + "step": 10364 + }, + { + "epoch": 2.3583617747440275, + "grad_norm": 2.6125927975379435, + "learning_rate": 6.806368079986114e-07, + "loss": 0.0772, + "step": 10365 + }, + { + "epoch": 2.358589306029579, + "grad_norm": 1.667257128844991, + "learning_rate": 6.805478108338009e-07, + "loss": 0.1122, + "step": 10366 + }, + { + "epoch": 2.358816837315131, + "grad_norm": 1.856152188586075, + "learning_rate": 6.80458812533695e-07, + "loss": 0.0294, + "step": 10367 + }, + { + "epoch": 2.359044368600683, + "grad_norm": 1.8307329732533564, + "learning_rate": 6.803698131001124e-07, + "loss": 0.0588, + "step": 10368 + }, + { + "epoch": 2.3592718998862345, + "grad_norm": 1.489190243200775, + "learning_rate": 6.802808125348722e-07, + "loss": 0.0632, + "step": 10369 + }, + { + "epoch": 2.3594994311717863, + "grad_norm": 1.2333545950135139, + "learning_rate": 6.801918108397934e-07, + "loss": 0.0393, + "step": 10370 + }, + { + "epoch": 2.359726962457338, + "grad_norm": 2.519109848016765, + "learning_rate": 6.801028080166952e-07, + "loss": 0.0418, + "step": 10371 + }, + { + "epoch": 2.35995449374289, + "grad_norm": 1.6748861718812718, + "learning_rate": 6.800138040673964e-07, + "loss": 0.054, + "step": 10372 + }, + { + "epoch": 2.3601820250284415, + "grad_norm": 1.822331215936714, + "learning_rate": 6.799247989937163e-07, + "loss": 0.0662, + "step": 10373 + }, + { + "epoch": 2.3604095563139933, + "grad_norm": 1.4346477535407098, + "learning_rate": 6.798357927974739e-07, + "loss": 0.0922, + "step": 10374 + }, + { + "epoch": 2.360637087599545, + "grad_norm": 1.252795785587945, + "learning_rate": 6.797467854804884e-07, + "loss": 0.0326, + "step": 10375 + }, + { + "epoch": 2.360864618885097, + "grad_norm": 1.9896922356110807, + "learning_rate": 6.796577770445785e-07, + "loss": 0.046, + "step": 10376 + }, + { + "epoch": 2.3610921501706486, + "grad_norm": 1.064666659792519, + "learning_rate": 6.795687674915639e-07, + "loss": 0.0476, + "step": 10377 + }, + { + "epoch": 2.3613196814562003, + "grad_norm": 1.0207304433420497, + "learning_rate": 6.794797568232639e-07, + "loss": 0.0257, + "step": 10378 + }, + { + "epoch": 2.361547212741752, + "grad_norm": 1.2416918419794614, + "learning_rate": 6.793907450414972e-07, + "loss": 0.0181, + "step": 10379 + }, + { + "epoch": 2.361774744027304, + "grad_norm": 1.4477653807207993, + "learning_rate": 6.793017321480834e-07, + "loss": 0.0874, + "step": 10380 + }, + { + "epoch": 2.3620022753128556, + "grad_norm": 1.1817989268786084, + "learning_rate": 6.792127181448415e-07, + "loss": 0.0337, + "step": 10381 + }, + { + "epoch": 2.3622298065984073, + "grad_norm": 1.0375656397551531, + "learning_rate": 6.791237030335911e-07, + "loss": 0.0421, + "step": 10382 + }, + { + "epoch": 2.362457337883959, + "grad_norm": 1.9502506528063932, + "learning_rate": 6.790346868161511e-07, + "loss": 0.0506, + "step": 10383 + }, + { + "epoch": 2.362684869169511, + "grad_norm": 1.5800476766186669, + "learning_rate": 6.789456694943413e-07, + "loss": 0.1179, + "step": 10384 + }, + { + "epoch": 2.3629124004550626, + "grad_norm": 0.9469539194251811, + "learning_rate": 6.788566510699808e-07, + "loss": 0.0319, + "step": 10385 + }, + { + "epoch": 2.3631399317406143, + "grad_norm": 1.674166648976265, + "learning_rate": 6.787676315448887e-07, + "loss": 0.046, + "step": 10386 + }, + { + "epoch": 2.363367463026166, + "grad_norm": 1.3057095334506177, + "learning_rate": 6.78678610920885e-07, + "loss": 0.0331, + "step": 10387 + }, + { + "epoch": 2.363594994311718, + "grad_norm": 1.2221485169323887, + "learning_rate": 6.785895891997887e-07, + "loss": 0.0575, + "step": 10388 + }, + { + "epoch": 2.3638225255972696, + "grad_norm": 2.47200812817322, + "learning_rate": 6.785005663834193e-07, + "loss": 0.0595, + "step": 10389 + }, + { + "epoch": 2.3640500568828213, + "grad_norm": 1.2902949523847893, + "learning_rate": 6.784115424735962e-07, + "loss": 0.0295, + "step": 10390 + }, + { + "epoch": 2.364277588168373, + "grad_norm": 2.026623907790253, + "learning_rate": 6.783225174721393e-07, + "loss": 0.0785, + "step": 10391 + }, + { + "epoch": 2.364505119453925, + "grad_norm": 1.7188511270636622, + "learning_rate": 6.782334913808678e-07, + "loss": 0.0746, + "step": 10392 + }, + { + "epoch": 2.3647326507394766, + "grad_norm": 1.7316607200710685, + "learning_rate": 6.781444642016008e-07, + "loss": 0.0309, + "step": 10393 + }, + { + "epoch": 2.3649601820250283, + "grad_norm": 1.555419843322222, + "learning_rate": 6.780554359361585e-07, + "loss": 0.0505, + "step": 10394 + }, + { + "epoch": 2.36518771331058, + "grad_norm": 1.0243836687399601, + "learning_rate": 6.779664065863605e-07, + "loss": 0.0602, + "step": 10395 + }, + { + "epoch": 2.365415244596132, + "grad_norm": 1.349428242935904, + "learning_rate": 6.778773761540258e-07, + "loss": 0.0535, + "step": 10396 + }, + { + "epoch": 2.3656427758816836, + "grad_norm": 1.3894434057799359, + "learning_rate": 6.777883446409746e-07, + "loss": 0.0621, + "step": 10397 + }, + { + "epoch": 2.3658703071672353, + "grad_norm": 1.4582759889005104, + "learning_rate": 6.776993120490262e-07, + "loss": 0.0301, + "step": 10398 + }, + { + "epoch": 2.366097838452787, + "grad_norm": 1.42088956232628, + "learning_rate": 6.776102783800005e-07, + "loss": 0.1321, + "step": 10399 + }, + { + "epoch": 2.366325369738339, + "grad_norm": 1.040231240478876, + "learning_rate": 6.775212436357167e-07, + "loss": 0.0135, + "step": 10400 + }, + { + "epoch": 2.3665529010238906, + "grad_norm": 1.441762410375783, + "learning_rate": 6.774322078179953e-07, + "loss": 0.0632, + "step": 10401 + }, + { + "epoch": 2.366780432309443, + "grad_norm": 1.7463501720699834, + "learning_rate": 6.773431709286554e-07, + "loss": 0.0482, + "step": 10402 + }, + { + "epoch": 2.367007963594994, + "grad_norm": 2.313903667162793, + "learning_rate": 6.77254132969517e-07, + "loss": 0.0684, + "step": 10403 + }, + { + "epoch": 2.3672354948805463, + "grad_norm": 1.2809218776804336, + "learning_rate": 6.771650939423999e-07, + "loss": 0.043, + "step": 10404 + }, + { + "epoch": 2.3674630261660976, + "grad_norm": 1.405519046760732, + "learning_rate": 6.770760538491236e-07, + "loss": 0.0632, + "step": 10405 + }, + { + "epoch": 2.36769055745165, + "grad_norm": 1.18910682344233, + "learning_rate": 6.769870126915083e-07, + "loss": 0.0183, + "step": 10406 + }, + { + "epoch": 2.3679180887372016, + "grad_norm": 1.5451868911236262, + "learning_rate": 6.768979704713735e-07, + "loss": 0.1525, + "step": 10407 + }, + { + "epoch": 2.3681456200227533, + "grad_norm": 1.954012861557462, + "learning_rate": 6.768089271905394e-07, + "loss": 0.0648, + "step": 10408 + }, + { + "epoch": 2.368373151308305, + "grad_norm": 1.9295472963366158, + "learning_rate": 6.767198828508257e-07, + "loss": 0.0324, + "step": 10409 + }, + { + "epoch": 2.368600682593857, + "grad_norm": 1.660240641429707, + "learning_rate": 6.766308374540523e-07, + "loss": 0.0779, + "step": 10410 + }, + { + "epoch": 2.3688282138794086, + "grad_norm": 1.169389631628708, + "learning_rate": 6.76541791002039e-07, + "loss": 0.0397, + "step": 10411 + }, + { + "epoch": 2.3690557451649603, + "grad_norm": 1.8021577785523133, + "learning_rate": 6.76452743496606e-07, + "loss": 0.0365, + "step": 10412 + }, + { + "epoch": 2.369283276450512, + "grad_norm": 1.6652405511680066, + "learning_rate": 6.763636949395731e-07, + "loss": 0.0882, + "step": 10413 + }, + { + "epoch": 2.369510807736064, + "grad_norm": 1.5091100026402833, + "learning_rate": 6.762746453327604e-07, + "loss": 0.0486, + "step": 10414 + }, + { + "epoch": 2.3697383390216156, + "grad_norm": 1.8699053258256377, + "learning_rate": 6.761855946779879e-07, + "loss": 0.0707, + "step": 10415 + }, + { + "epoch": 2.3699658703071673, + "grad_norm": 1.5328100039491475, + "learning_rate": 6.760965429770755e-07, + "loss": 0.0689, + "step": 10416 + }, + { + "epoch": 2.370193401592719, + "grad_norm": 1.314123434522282, + "learning_rate": 6.760074902318435e-07, + "loss": 0.0383, + "step": 10417 + }, + { + "epoch": 2.370420932878271, + "grad_norm": 1.4855379347463653, + "learning_rate": 6.759184364441117e-07, + "loss": 0.0482, + "step": 10418 + }, + { + "epoch": 2.3706484641638226, + "grad_norm": 2.670435917113975, + "learning_rate": 6.758293816157003e-07, + "loss": 0.0643, + "step": 10419 + }, + { + "epoch": 2.3708759954493743, + "grad_norm": 1.9280635575174545, + "learning_rate": 6.757403257484293e-07, + "loss": 0.0199, + "step": 10420 + }, + { + "epoch": 2.371103526734926, + "grad_norm": 0.9549318126905393, + "learning_rate": 6.756512688441191e-07, + "loss": 0.0355, + "step": 10421 + }, + { + "epoch": 2.371331058020478, + "grad_norm": 1.332552454350489, + "learning_rate": 6.7556221090459e-07, + "loss": 0.031, + "step": 10422 + }, + { + "epoch": 2.3715585893060296, + "grad_norm": 2.215723961040164, + "learning_rate": 6.754731519316615e-07, + "loss": 0.0522, + "step": 10423 + }, + { + "epoch": 2.3717861205915813, + "grad_norm": 0.8980813897551289, + "learning_rate": 6.753840919271542e-07, + "loss": 0.0179, + "step": 10424 + }, + { + "epoch": 2.372013651877133, + "grad_norm": 2.180420797793543, + "learning_rate": 6.752950308928887e-07, + "loss": 0.0434, + "step": 10425 + }, + { + "epoch": 2.372241183162685, + "grad_norm": 1.0978645616605351, + "learning_rate": 6.752059688306846e-07, + "loss": 0.0436, + "step": 10426 + }, + { + "epoch": 2.3724687144482366, + "grad_norm": 0.9931325799878876, + "learning_rate": 6.751169057423625e-07, + "loss": 0.0767, + "step": 10427 + }, + { + "epoch": 2.3726962457337883, + "grad_norm": 1.5451840499172698, + "learning_rate": 6.750278416297426e-07, + "loss": 0.1591, + "step": 10428 + }, + { + "epoch": 2.37292377701934, + "grad_norm": 1.5924635594374823, + "learning_rate": 6.749387764946454e-07, + "loss": 0.0657, + "step": 10429 + }, + { + "epoch": 2.373151308304892, + "grad_norm": 2.1149792343170857, + "learning_rate": 6.748497103388908e-07, + "loss": 0.0922, + "step": 10430 + }, + { + "epoch": 2.3733788395904436, + "grad_norm": 1.7817157981695644, + "learning_rate": 6.747606431642996e-07, + "loss": 0.0616, + "step": 10431 + }, + { + "epoch": 2.3736063708759954, + "grad_norm": 1.1096939510401005, + "learning_rate": 6.746715749726921e-07, + "loss": 0.0148, + "step": 10432 + }, + { + "epoch": 2.373833902161547, + "grad_norm": 2.0019869696931742, + "learning_rate": 6.745825057658884e-07, + "loss": 0.0595, + "step": 10433 + }, + { + "epoch": 2.374061433447099, + "grad_norm": 1.2718307936644717, + "learning_rate": 6.744934355457089e-07, + "loss": 0.093, + "step": 10434 + }, + { + "epoch": 2.3742889647326506, + "grad_norm": 1.3919586686320002, + "learning_rate": 6.744043643139746e-07, + "loss": 0.0363, + "step": 10435 + }, + { + "epoch": 2.3745164960182024, + "grad_norm": 1.5809566609156156, + "learning_rate": 6.743152920725054e-07, + "loss": 0.0811, + "step": 10436 + }, + { + "epoch": 2.374744027303754, + "grad_norm": 3.036678376933095, + "learning_rate": 6.742262188231219e-07, + "loss": 0.0378, + "step": 10437 + }, + { + "epoch": 2.374971558589306, + "grad_norm": 1.8720444851573026, + "learning_rate": 6.741371445676448e-07, + "loss": 0.12, + "step": 10438 + }, + { + "epoch": 2.3751990898748576, + "grad_norm": 1.5379917535938616, + "learning_rate": 6.740480693078944e-07, + "loss": 0.0351, + "step": 10439 + }, + { + "epoch": 2.3754266211604094, + "grad_norm": 0.9430200124299626, + "learning_rate": 6.739589930456911e-07, + "loss": 0.0501, + "step": 10440 + }, + { + "epoch": 2.3756541524459616, + "grad_norm": 1.9896257764938758, + "learning_rate": 6.738699157828558e-07, + "loss": 0.1009, + "step": 10441 + }, + { + "epoch": 2.375881683731513, + "grad_norm": 2.018242344845083, + "learning_rate": 6.737808375212091e-07, + "loss": 0.053, + "step": 10442 + }, + { + "epoch": 2.376109215017065, + "grad_norm": 1.584315374444164, + "learning_rate": 6.73691758262571e-07, + "loss": 0.0605, + "step": 10443 + }, + { + "epoch": 2.3763367463026164, + "grad_norm": 1.6702742440286638, + "learning_rate": 6.736026780087627e-07, + "loss": 0.0228, + "step": 10444 + }, + { + "epoch": 2.3765642775881686, + "grad_norm": 1.6115908004236295, + "learning_rate": 6.735135967616048e-07, + "loss": 0.0353, + "step": 10445 + }, + { + "epoch": 2.3767918088737203, + "grad_norm": 2.2811883410106764, + "learning_rate": 6.734245145229179e-07, + "loss": 0.0676, + "step": 10446 + }, + { + "epoch": 2.377019340159272, + "grad_norm": 2.3498223584151003, + "learning_rate": 6.733354312945223e-07, + "loss": 0.0503, + "step": 10447 + }, + { + "epoch": 2.377246871444824, + "grad_norm": 1.445057700877811, + "learning_rate": 6.732463470782394e-07, + "loss": 0.1034, + "step": 10448 + }, + { + "epoch": 2.3774744027303756, + "grad_norm": 2.1756792209653697, + "learning_rate": 6.731572618758893e-07, + "loss": 0.0487, + "step": 10449 + }, + { + "epoch": 2.3777019340159273, + "grad_norm": 2.2820082997636826, + "learning_rate": 6.730681756892929e-07, + "loss": 0.0959, + "step": 10450 + }, + { + "epoch": 2.377929465301479, + "grad_norm": 1.5032639741412177, + "learning_rate": 6.729790885202712e-07, + "loss": 0.0295, + "step": 10451 + }, + { + "epoch": 2.378156996587031, + "grad_norm": 1.780461004855868, + "learning_rate": 6.728900003706446e-07, + "loss": 0.0498, + "step": 10452 + }, + { + "epoch": 2.3783845278725826, + "grad_norm": 2.2662441775377253, + "learning_rate": 6.728009112422341e-07, + "loss": 0.0367, + "step": 10453 + }, + { + "epoch": 2.3786120591581343, + "grad_norm": 2.067235913331479, + "learning_rate": 6.727118211368607e-07, + "loss": 0.0735, + "step": 10454 + }, + { + "epoch": 2.378839590443686, + "grad_norm": 1.4605393416349512, + "learning_rate": 6.72622730056345e-07, + "loss": 0.0544, + "step": 10455 + }, + { + "epoch": 2.379067121729238, + "grad_norm": 2.390078839111463, + "learning_rate": 6.725336380025078e-07, + "loss": 0.0907, + "step": 10456 + }, + { + "epoch": 2.3792946530147896, + "grad_norm": 1.2800789058267308, + "learning_rate": 6.724445449771702e-07, + "loss": 0.045, + "step": 10457 + }, + { + "epoch": 2.3795221843003413, + "grad_norm": 1.2874977725325683, + "learning_rate": 6.72355450982153e-07, + "loss": 0.0929, + "step": 10458 + }, + { + "epoch": 2.379749715585893, + "grad_norm": 1.4661100617335616, + "learning_rate": 6.72266356019277e-07, + "loss": 0.1, + "step": 10459 + }, + { + "epoch": 2.379977246871445, + "grad_norm": 1.396837121599532, + "learning_rate": 6.721772600903634e-07, + "loss": 0.1244, + "step": 10460 + }, + { + "epoch": 2.3802047781569966, + "grad_norm": 1.2305615808874142, + "learning_rate": 6.720881631972328e-07, + "loss": 0.0425, + "step": 10461 + }, + { + "epoch": 2.3804323094425484, + "grad_norm": 1.3978490997285564, + "learning_rate": 6.719990653417066e-07, + "loss": 0.0828, + "step": 10462 + }, + { + "epoch": 2.3806598407281, + "grad_norm": 0.8497783443439443, + "learning_rate": 6.719099665256056e-07, + "loss": 0.0226, + "step": 10463 + }, + { + "epoch": 2.380887372013652, + "grad_norm": 1.4514177373432697, + "learning_rate": 6.718208667507506e-07, + "loss": 0.0559, + "step": 10464 + }, + { + "epoch": 2.3811149032992036, + "grad_norm": 2.480240492112608, + "learning_rate": 6.717317660189629e-07, + "loss": 0.048, + "step": 10465 + }, + { + "epoch": 2.3813424345847554, + "grad_norm": 2.270491728647551, + "learning_rate": 6.716426643320635e-07, + "loss": 0.1405, + "step": 10466 + }, + { + "epoch": 2.381569965870307, + "grad_norm": 2.518922614754743, + "learning_rate": 6.715535616918735e-07, + "loss": 0.1303, + "step": 10467 + }, + { + "epoch": 2.381797497155859, + "grad_norm": 2.4748624828029095, + "learning_rate": 6.714644581002139e-07, + "loss": 0.0945, + "step": 10468 + }, + { + "epoch": 2.3820250284414106, + "grad_norm": 1.2874037769841746, + "learning_rate": 6.71375353558906e-07, + "loss": 0.0298, + "step": 10469 + }, + { + "epoch": 2.3822525597269624, + "grad_norm": 1.781652751677099, + "learning_rate": 6.712862480697705e-07, + "loss": 0.0323, + "step": 10470 + }, + { + "epoch": 2.382480091012514, + "grad_norm": 1.3399476686901257, + "learning_rate": 6.711971416346291e-07, + "loss": 0.0251, + "step": 10471 + }, + { + "epoch": 2.382707622298066, + "grad_norm": 1.089469592931769, + "learning_rate": 6.711080342553027e-07, + "loss": 0.0643, + "step": 10472 + }, + { + "epoch": 2.3829351535836176, + "grad_norm": 1.4013619938263189, + "learning_rate": 6.710189259336125e-07, + "loss": 0.0922, + "step": 10473 + }, + { + "epoch": 2.3831626848691694, + "grad_norm": 3.376435801233772, + "learning_rate": 6.709298166713799e-07, + "loss": 0.1387, + "step": 10474 + }, + { + "epoch": 2.383390216154721, + "grad_norm": 1.9844579581141586, + "learning_rate": 6.708407064704258e-07, + "loss": 0.0767, + "step": 10475 + }, + { + "epoch": 2.383617747440273, + "grad_norm": 1.9453254227786483, + "learning_rate": 6.707515953325716e-07, + "loss": 0.0635, + "step": 10476 + }, + { + "epoch": 2.3838452787258246, + "grad_norm": 1.452221262310961, + "learning_rate": 6.706624832596385e-07, + "loss": 0.0494, + "step": 10477 + }, + { + "epoch": 2.3840728100113764, + "grad_norm": 1.38605309294856, + "learning_rate": 6.70573370253448e-07, + "loss": 0.0276, + "step": 10478 + }, + { + "epoch": 2.384300341296928, + "grad_norm": 1.7034378285942493, + "learning_rate": 6.704842563158214e-07, + "loss": 0.0224, + "step": 10479 + }, + { + "epoch": 2.3845278725824803, + "grad_norm": 1.021182659891657, + "learning_rate": 6.703951414485796e-07, + "loss": 0.0162, + "step": 10480 + }, + { + "epoch": 2.3847554038680316, + "grad_norm": 1.3913943329450087, + "learning_rate": 6.703060256535445e-07, + "loss": 0.0785, + "step": 10481 + }, + { + "epoch": 2.384982935153584, + "grad_norm": 1.5931390005406347, + "learning_rate": 6.702169089325371e-07, + "loss": 0.1032, + "step": 10482 + }, + { + "epoch": 2.385210466439135, + "grad_norm": 1.8369869104956416, + "learning_rate": 6.70127791287379e-07, + "loss": 0.0667, + "step": 10483 + }, + { + "epoch": 2.3854379977246873, + "grad_norm": 1.7301862270354487, + "learning_rate": 6.700386727198911e-07, + "loss": 0.041, + "step": 10484 + }, + { + "epoch": 2.385665529010239, + "grad_norm": 2.1651615466332315, + "learning_rate": 6.699495532318957e-07, + "loss": 0.0995, + "step": 10485 + }, + { + "epoch": 2.385893060295791, + "grad_norm": 1.3938195707799108, + "learning_rate": 6.698604328252137e-07, + "loss": 0.0856, + "step": 10486 + }, + { + "epoch": 2.3861205915813426, + "grad_norm": 1.7030373623693367, + "learning_rate": 6.697713115016663e-07, + "loss": 0.0881, + "step": 10487 + }, + { + "epoch": 2.3863481228668944, + "grad_norm": 1.8946035407113386, + "learning_rate": 6.696821892630754e-07, + "loss": 0.0395, + "step": 10488 + }, + { + "epoch": 2.386575654152446, + "grad_norm": 1.7859831641223316, + "learning_rate": 6.695930661112625e-07, + "loss": 0.0516, + "step": 10489 + }, + { + "epoch": 2.386803185437998, + "grad_norm": 2.086999207048147, + "learning_rate": 6.69503942048049e-07, + "loss": 0.0306, + "step": 10490 + }, + { + "epoch": 2.3870307167235496, + "grad_norm": 1.2645238222756374, + "learning_rate": 6.694148170752562e-07, + "loss": 0.0535, + "step": 10491 + }, + { + "epoch": 2.3872582480091014, + "grad_norm": 1.28426810874403, + "learning_rate": 6.693256911947063e-07, + "loss": 0.0998, + "step": 10492 + }, + { + "epoch": 2.387485779294653, + "grad_norm": 1.3537016819729306, + "learning_rate": 6.692365644082202e-07, + "loss": 0.0531, + "step": 10493 + }, + { + "epoch": 2.387713310580205, + "grad_norm": 2.83060355995732, + "learning_rate": 6.691474367176195e-07, + "loss": 0.102, + "step": 10494 + }, + { + "epoch": 2.3879408418657566, + "grad_norm": 1.4814180587134658, + "learning_rate": 6.690583081247264e-07, + "loss": 0.0652, + "step": 10495 + }, + { + "epoch": 2.3881683731513084, + "grad_norm": 1.7368911499047102, + "learning_rate": 6.68969178631362e-07, + "loss": 0.0802, + "step": 10496 + }, + { + "epoch": 2.38839590443686, + "grad_norm": 2.580003746707147, + "learning_rate": 6.688800482393481e-07, + "loss": 0.0751, + "step": 10497 + }, + { + "epoch": 2.388623435722412, + "grad_norm": 1.6310291695861174, + "learning_rate": 6.687909169505066e-07, + "loss": 0.038, + "step": 10498 + }, + { + "epoch": 2.3888509670079636, + "grad_norm": 2.1723786955879385, + "learning_rate": 6.687017847666588e-07, + "loss": 0.1669, + "step": 10499 + }, + { + "epoch": 2.3890784982935154, + "grad_norm": 1.666439442594346, + "learning_rate": 6.686126516896266e-07, + "loss": 0.0962, + "step": 10500 + }, + { + "epoch": 2.389306029579067, + "grad_norm": 2.4965699284748966, + "learning_rate": 6.685235177212315e-07, + "loss": 0.1345, + "step": 10501 + }, + { + "epoch": 2.389533560864619, + "grad_norm": 1.2904438567954284, + "learning_rate": 6.684343828632957e-07, + "loss": 0.0433, + "step": 10502 + }, + { + "epoch": 2.3897610921501706, + "grad_norm": 1.6373753924522154, + "learning_rate": 6.683452471176405e-07, + "loss": 0.1001, + "step": 10503 + }, + { + "epoch": 2.3899886234357224, + "grad_norm": 1.7248619499584223, + "learning_rate": 6.682561104860878e-07, + "loss": 0.1004, + "step": 10504 + }, + { + "epoch": 2.390216154721274, + "grad_norm": 1.4465268197949492, + "learning_rate": 6.681669729704595e-07, + "loss": 0.0779, + "step": 10505 + }, + { + "epoch": 2.390443686006826, + "grad_norm": 1.7436657794943984, + "learning_rate": 6.680778345725773e-07, + "loss": 0.0303, + "step": 10506 + }, + { + "epoch": 2.3906712172923776, + "grad_norm": 1.5225262043163035, + "learning_rate": 6.679886952942629e-07, + "loss": 0.0795, + "step": 10507 + }, + { + "epoch": 2.3908987485779294, + "grad_norm": 1.329757770233145, + "learning_rate": 6.678995551373385e-07, + "loss": 0.0397, + "step": 10508 + }, + { + "epoch": 2.391126279863481, + "grad_norm": 1.339207113075806, + "learning_rate": 6.678104141036257e-07, + "loss": 0.0317, + "step": 10509 + }, + { + "epoch": 2.391353811149033, + "grad_norm": 1.2227312515171715, + "learning_rate": 6.677212721949464e-07, + "loss": 0.0318, + "step": 10510 + }, + { + "epoch": 2.3915813424345846, + "grad_norm": 1.3724822590412142, + "learning_rate": 6.676321294131226e-07, + "loss": 0.0742, + "step": 10511 + }, + { + "epoch": 2.3918088737201364, + "grad_norm": 1.560545283718502, + "learning_rate": 6.675429857599762e-07, + "loss": 0.1334, + "step": 10512 + }, + { + "epoch": 2.392036405005688, + "grad_norm": 1.8795699309375529, + "learning_rate": 6.674538412373289e-07, + "loss": 0.032, + "step": 10513 + }, + { + "epoch": 2.39226393629124, + "grad_norm": 2.829895580966993, + "learning_rate": 6.673646958470029e-07, + "loss": 0.153, + "step": 10514 + }, + { + "epoch": 2.3924914675767917, + "grad_norm": 1.569501158041787, + "learning_rate": 6.672755495908202e-07, + "loss": 0.0261, + "step": 10515 + }, + { + "epoch": 2.3927189988623434, + "grad_norm": 1.522202430622725, + "learning_rate": 6.671864024706027e-07, + "loss": 0.0793, + "step": 10516 + }, + { + "epoch": 2.392946530147895, + "grad_norm": 4.090585940030713, + "learning_rate": 6.670972544881723e-07, + "loss": 0.0648, + "step": 10517 + }, + { + "epoch": 2.393174061433447, + "grad_norm": 1.495096016038974, + "learning_rate": 6.670081056453512e-07, + "loss": 0.0507, + "step": 10518 + }, + { + "epoch": 2.393401592718999, + "grad_norm": 2.71488569789709, + "learning_rate": 6.669189559439613e-07, + "loss": 0.0832, + "step": 10519 + }, + { + "epoch": 2.3936291240045504, + "grad_norm": 2.179825439345257, + "learning_rate": 6.668298053858248e-07, + "loss": 0.0469, + "step": 10520 + }, + { + "epoch": 2.3938566552901026, + "grad_norm": 1.2228033131349032, + "learning_rate": 6.667406539727634e-07, + "loss": 0.091, + "step": 10521 + }, + { + "epoch": 2.394084186575654, + "grad_norm": 2.352962433677147, + "learning_rate": 6.666515017065997e-07, + "loss": 0.0412, + "step": 10522 + }, + { + "epoch": 2.394311717861206, + "grad_norm": 2.3997551364175105, + "learning_rate": 6.665623485891558e-07, + "loss": 0.0582, + "step": 10523 + }, + { + "epoch": 2.394539249146758, + "grad_norm": 1.1027134450061242, + "learning_rate": 6.664731946222531e-07, + "loss": 0.0206, + "step": 10524 + }, + { + "epoch": 2.3947667804323096, + "grad_norm": 2.4549027128997642, + "learning_rate": 6.663840398077146e-07, + "loss": 0.0472, + "step": 10525 + }, + { + "epoch": 2.3949943117178614, + "grad_norm": 2.285795803069614, + "learning_rate": 6.662948841473621e-07, + "loss": 0.1502, + "step": 10526 + }, + { + "epoch": 2.395221843003413, + "grad_norm": 2.1733632351365633, + "learning_rate": 6.662057276430179e-07, + "loss": 0.1195, + "step": 10527 + }, + { + "epoch": 2.395449374288965, + "grad_norm": 1.6885301323046775, + "learning_rate": 6.661165702965037e-07, + "loss": 0.1168, + "step": 10528 + }, + { + "epoch": 2.3956769055745166, + "grad_norm": 1.5015771246663587, + "learning_rate": 6.660274121096425e-07, + "loss": 0.0375, + "step": 10529 + }, + { + "epoch": 2.3959044368600684, + "grad_norm": 3.086439361778395, + "learning_rate": 6.659382530842561e-07, + "loss": 0.0481, + "step": 10530 + }, + { + "epoch": 2.39613196814562, + "grad_norm": 0.9911636574779332, + "learning_rate": 6.658490932221664e-07, + "loss": 0.0445, + "step": 10531 + }, + { + "epoch": 2.396359499431172, + "grad_norm": 1.0139124781259454, + "learning_rate": 6.657599325251964e-07, + "loss": 0.0703, + "step": 10532 + }, + { + "epoch": 2.3965870307167236, + "grad_norm": 2.627123214882737, + "learning_rate": 6.656707709951679e-07, + "loss": 0.1473, + "step": 10533 + }, + { + "epoch": 2.3968145620022754, + "grad_norm": 1.5488525003364622, + "learning_rate": 6.655816086339033e-07, + "loss": 0.0444, + "step": 10534 + }, + { + "epoch": 2.397042093287827, + "grad_norm": 1.7357048098095746, + "learning_rate": 6.654924454432251e-07, + "loss": 0.1388, + "step": 10535 + }, + { + "epoch": 2.397269624573379, + "grad_norm": 1.8321173442713896, + "learning_rate": 6.654032814249554e-07, + "loss": 0.0781, + "step": 10536 + }, + { + "epoch": 2.3974971558589306, + "grad_norm": 1.5942316650306552, + "learning_rate": 6.653141165809166e-07, + "loss": 0.0406, + "step": 10537 + }, + { + "epoch": 2.3977246871444824, + "grad_norm": 2.6834686068069757, + "learning_rate": 6.65224950912931e-07, + "loss": 0.1056, + "step": 10538 + }, + { + "epoch": 2.397952218430034, + "grad_norm": 1.769237068832356, + "learning_rate": 6.651357844228213e-07, + "loss": 0.0813, + "step": 10539 + }, + { + "epoch": 2.398179749715586, + "grad_norm": 1.5233688630082565, + "learning_rate": 6.650466171124094e-07, + "loss": 0.0386, + "step": 10540 + }, + { + "epoch": 2.3984072810011376, + "grad_norm": 2.008063039797502, + "learning_rate": 6.649574489835181e-07, + "loss": 0.0273, + "step": 10541 + }, + { + "epoch": 2.3986348122866894, + "grad_norm": 1.6343752435695706, + "learning_rate": 6.648682800379698e-07, + "loss": 0.0331, + "step": 10542 + }, + { + "epoch": 2.398862343572241, + "grad_norm": 1.1926677364645177, + "learning_rate": 6.647791102775869e-07, + "loss": 0.0404, + "step": 10543 + }, + { + "epoch": 2.399089874857793, + "grad_norm": 1.5172629638216661, + "learning_rate": 6.646899397041915e-07, + "loss": 0.0912, + "step": 10544 + }, + { + "epoch": 2.3993174061433447, + "grad_norm": 1.5887104653749333, + "learning_rate": 6.646007683196068e-07, + "loss": 0.097, + "step": 10545 + }, + { + "epoch": 2.3995449374288964, + "grad_norm": 1.52999208817766, + "learning_rate": 6.645115961256549e-07, + "loss": 0.0364, + "step": 10546 + }, + { + "epoch": 2.399772468714448, + "grad_norm": 1.6407741258520294, + "learning_rate": 6.644224231241582e-07, + "loss": 0.1035, + "step": 10547 + }, + { + "epoch": 2.4, + "grad_norm": 1.3227732207379344, + "learning_rate": 6.643332493169393e-07, + "loss": 0.0594, + "step": 10548 + }, + { + "epoch": 2.4002275312855517, + "grad_norm": 1.4032556973087806, + "learning_rate": 6.642440747058209e-07, + "loss": 0.0276, + "step": 10549 + }, + { + "epoch": 2.4004550625711034, + "grad_norm": 1.5628829654489371, + "learning_rate": 6.641548992926256e-07, + "loss": 0.071, + "step": 10550 + }, + { + "epoch": 2.400682593856655, + "grad_norm": 2.642593705218311, + "learning_rate": 6.640657230791757e-07, + "loss": 0.0606, + "step": 10551 + }, + { + "epoch": 2.400910125142207, + "grad_norm": 2.257245286182545, + "learning_rate": 6.639765460672941e-07, + "loss": 0.0452, + "step": 10552 + }, + { + "epoch": 2.4011376564277587, + "grad_norm": 1.09570693200126, + "learning_rate": 6.638873682588032e-07, + "loss": 0.0459, + "step": 10553 + }, + { + "epoch": 2.4013651877133104, + "grad_norm": 1.3975264910232386, + "learning_rate": 6.637981896555257e-07, + "loss": 0.0388, + "step": 10554 + }, + { + "epoch": 2.401592718998862, + "grad_norm": 1.835536891122451, + "learning_rate": 6.637090102592843e-07, + "loss": 0.1115, + "step": 10555 + }, + { + "epoch": 2.401820250284414, + "grad_norm": 2.040239846492686, + "learning_rate": 6.636198300719017e-07, + "loss": 0.0594, + "step": 10556 + }, + { + "epoch": 2.4020477815699657, + "grad_norm": 1.9975693704452562, + "learning_rate": 6.635306490952003e-07, + "loss": 0.1658, + "step": 10557 + }, + { + "epoch": 2.402275312855518, + "grad_norm": 1.923275401920876, + "learning_rate": 6.63441467331003e-07, + "loss": 0.1388, + "step": 10558 + }, + { + "epoch": 2.402502844141069, + "grad_norm": 1.2983548889372092, + "learning_rate": 6.633522847811327e-07, + "loss": 0.0808, + "step": 10559 + }, + { + "epoch": 2.4027303754266214, + "grad_norm": 1.236313620739703, + "learning_rate": 6.63263101447412e-07, + "loss": 0.091, + "step": 10560 + }, + { + "epoch": 2.4029579067121727, + "grad_norm": 1.5806673071769493, + "learning_rate": 6.631739173316634e-07, + "loss": 0.0402, + "step": 10561 + }, + { + "epoch": 2.403185437997725, + "grad_norm": 1.6003664986803758, + "learning_rate": 6.630847324357098e-07, + "loss": 0.1065, + "step": 10562 + }, + { + "epoch": 2.4034129692832766, + "grad_norm": 1.2945758422502862, + "learning_rate": 6.629955467613741e-07, + "loss": 0.0235, + "step": 10563 + }, + { + "epoch": 2.4036405005688284, + "grad_norm": 1.5211462068444157, + "learning_rate": 6.629063603104789e-07, + "loss": 0.0418, + "step": 10564 + }, + { + "epoch": 2.40386803185438, + "grad_norm": 1.2126875345668893, + "learning_rate": 6.628171730848474e-07, + "loss": 0.0695, + "step": 10565 + }, + { + "epoch": 2.404095563139932, + "grad_norm": 2.57120009931569, + "learning_rate": 6.62727985086302e-07, + "loss": 0.0576, + "step": 10566 + }, + { + "epoch": 2.4043230944254836, + "grad_norm": 3.2348915583257583, + "learning_rate": 6.626387963166655e-07, + "loss": 0.0391, + "step": 10567 + }, + { + "epoch": 2.4045506257110354, + "grad_norm": 1.7454433517441201, + "learning_rate": 6.625496067777612e-07, + "loss": 0.0921, + "step": 10568 + }, + { + "epoch": 2.404778156996587, + "grad_norm": 2.321262013510454, + "learning_rate": 6.624604164714115e-07, + "loss": 0.0483, + "step": 10569 + }, + { + "epoch": 2.405005688282139, + "grad_norm": 1.7997980777352869, + "learning_rate": 6.623712253994397e-07, + "loss": 0.031, + "step": 10570 + }, + { + "epoch": 2.4052332195676907, + "grad_norm": 1.3828194024201597, + "learning_rate": 6.622820335636683e-07, + "loss": 0.0326, + "step": 10571 + }, + { + "epoch": 2.4054607508532424, + "grad_norm": 1.7488919911490366, + "learning_rate": 6.621928409659204e-07, + "loss": 0.0156, + "step": 10572 + }, + { + "epoch": 2.405688282138794, + "grad_norm": 1.7382412068729043, + "learning_rate": 6.621036476080191e-07, + "loss": 0.0979, + "step": 10573 + }, + { + "epoch": 2.405915813424346, + "grad_norm": 1.6281716241666897, + "learning_rate": 6.620144534917872e-07, + "loss": 0.0827, + "step": 10574 + }, + { + "epoch": 2.4061433447098977, + "grad_norm": 1.2394446083517534, + "learning_rate": 6.619252586190477e-07, + "loss": 0.0504, + "step": 10575 + }, + { + "epoch": 2.4063708759954494, + "grad_norm": 1.1996299911059907, + "learning_rate": 6.618360629916233e-07, + "loss": 0.0089, + "step": 10576 + }, + { + "epoch": 2.406598407281001, + "grad_norm": 1.6445535839940941, + "learning_rate": 6.617468666113375e-07, + "loss": 0.0632, + "step": 10577 + }, + { + "epoch": 2.406825938566553, + "grad_norm": 1.7208377523448997, + "learning_rate": 6.616576694800126e-07, + "loss": 0.0335, + "step": 10578 + }, + { + "epoch": 2.4070534698521047, + "grad_norm": 1.2991586455146025, + "learning_rate": 6.615684715994725e-07, + "loss": 0.0504, + "step": 10579 + }, + { + "epoch": 2.4072810011376564, + "grad_norm": 2.387057251902132, + "learning_rate": 6.614792729715398e-07, + "loss": 0.0278, + "step": 10580 + }, + { + "epoch": 2.407508532423208, + "grad_norm": 1.1634268950453206, + "learning_rate": 6.613900735980374e-07, + "loss": 0.0124, + "step": 10581 + }, + { + "epoch": 2.40773606370876, + "grad_norm": 1.49682262983672, + "learning_rate": 6.613008734807886e-07, + "loss": 0.1172, + "step": 10582 + }, + { + "epoch": 2.4079635949943117, + "grad_norm": 1.4685646699587809, + "learning_rate": 6.612116726216164e-07, + "loss": 0.0661, + "step": 10583 + }, + { + "epoch": 2.4081911262798634, + "grad_norm": 1.5513479030600477, + "learning_rate": 6.611224710223441e-07, + "loss": 0.103, + "step": 10584 + }, + { + "epoch": 2.408418657565415, + "grad_norm": 1.466169056624238, + "learning_rate": 6.610332686847944e-07, + "loss": 0.0416, + "step": 10585 + }, + { + "epoch": 2.408646188850967, + "grad_norm": 2.137838763287238, + "learning_rate": 6.609440656107909e-07, + "loss": 0.0266, + "step": 10586 + }, + { + "epoch": 2.4088737201365187, + "grad_norm": 1.5612872060850345, + "learning_rate": 6.608548618021563e-07, + "loss": 0.0277, + "step": 10587 + }, + { + "epoch": 2.4091012514220704, + "grad_norm": 1.2888637789751431, + "learning_rate": 6.60765657260714e-07, + "loss": 0.0692, + "step": 10588 + }, + { + "epoch": 2.409328782707622, + "grad_norm": 2.2910864201785692, + "learning_rate": 6.606764519882874e-07, + "loss": 0.0475, + "step": 10589 + }, + { + "epoch": 2.409556313993174, + "grad_norm": 1.5458940848749017, + "learning_rate": 6.605872459866993e-07, + "loss": 0.0357, + "step": 10590 + }, + { + "epoch": 2.4097838452787257, + "grad_norm": 1.8092381758797624, + "learning_rate": 6.604980392577729e-07, + "loss": 0.0737, + "step": 10591 + }, + { + "epoch": 2.4100113765642774, + "grad_norm": 0.8472770721012022, + "learning_rate": 6.604088318033321e-07, + "loss": 0.0302, + "step": 10592 + }, + { + "epoch": 2.410238907849829, + "grad_norm": 1.4934119759903768, + "learning_rate": 6.603196236251993e-07, + "loss": 0.0567, + "step": 10593 + }, + { + "epoch": 2.410466439135381, + "grad_norm": 1.37423121528308, + "learning_rate": 6.60230414725198e-07, + "loss": 0.0387, + "step": 10594 + }, + { + "epoch": 2.4106939704209327, + "grad_norm": 1.9350754144067868, + "learning_rate": 6.601412051051516e-07, + "loss": 0.0237, + "step": 10595 + }, + { + "epoch": 2.4109215017064844, + "grad_norm": 2.5387441409983396, + "learning_rate": 6.600519947668835e-07, + "loss": 0.093, + "step": 10596 + }, + { + "epoch": 2.4111490329920366, + "grad_norm": 1.3363002317558488, + "learning_rate": 6.599627837122167e-07, + "loss": 0.0593, + "step": 10597 + }, + { + "epoch": 2.411376564277588, + "grad_norm": 2.3969188016027974, + "learning_rate": 6.598735719429744e-07, + "loss": 0.0512, + "step": 10598 + }, + { + "epoch": 2.41160409556314, + "grad_norm": 1.609682451610226, + "learning_rate": 6.597843594609806e-07, + "loss": 0.0519, + "step": 10599 + }, + { + "epoch": 2.4118316268486915, + "grad_norm": 2.072723128432861, + "learning_rate": 6.59695146268058e-07, + "loss": 0.1232, + "step": 10600 + }, + { + "epoch": 2.4120591581342437, + "grad_norm": 1.817786471435142, + "learning_rate": 6.596059323660299e-07, + "loss": 0.1908, + "step": 10601 + }, + { + "epoch": 2.4122866894197954, + "grad_norm": 1.4728070130617423, + "learning_rate": 6.595167177567203e-07, + "loss": 0.0191, + "step": 10602 + }, + { + "epoch": 2.412514220705347, + "grad_norm": 1.842650549871717, + "learning_rate": 6.594275024419521e-07, + "loss": 0.0615, + "step": 10603 + }, + { + "epoch": 2.412741751990899, + "grad_norm": 1.130832928406942, + "learning_rate": 6.593382864235487e-07, + "loss": 0.0198, + "step": 10604 + }, + { + "epoch": 2.4129692832764507, + "grad_norm": 1.4138956609023778, + "learning_rate": 6.592490697033337e-07, + "loss": 0.0588, + "step": 10605 + }, + { + "epoch": 2.4131968145620024, + "grad_norm": 1.6261557637810253, + "learning_rate": 6.591598522831303e-07, + "loss": 0.0368, + "step": 10606 + }, + { + "epoch": 2.413424345847554, + "grad_norm": 0.803953439992238, + "learning_rate": 6.590706341647623e-07, + "loss": 0.0214, + "step": 10607 + }, + { + "epoch": 2.413651877133106, + "grad_norm": 1.45102004082111, + "learning_rate": 6.589814153500527e-07, + "loss": 0.0688, + "step": 10608 + }, + { + "epoch": 2.4138794084186577, + "grad_norm": 1.3071350386826537, + "learning_rate": 6.588921958408254e-07, + "loss": 0.1016, + "step": 10609 + }, + { + "epoch": 2.4141069397042094, + "grad_norm": 1.561782085640104, + "learning_rate": 6.588029756389037e-07, + "loss": 0.0725, + "step": 10610 + }, + { + "epoch": 2.414334470989761, + "grad_norm": 1.1935765970410221, + "learning_rate": 6.587137547461108e-07, + "loss": 0.0375, + "step": 10611 + }, + { + "epoch": 2.414562002275313, + "grad_norm": 1.9599412696011829, + "learning_rate": 6.586245331642707e-07, + "loss": 0.1582, + "step": 10612 + }, + { + "epoch": 2.4147895335608647, + "grad_norm": 2.8304494698055827, + "learning_rate": 6.585353108952068e-07, + "loss": 0.0454, + "step": 10613 + }, + { + "epoch": 2.4150170648464164, + "grad_norm": 1.7453458925600094, + "learning_rate": 6.584460879407425e-07, + "loss": 0.0909, + "step": 10614 + }, + { + "epoch": 2.415244596131968, + "grad_norm": 1.759146367940718, + "learning_rate": 6.583568643027012e-07, + "loss": 0.0695, + "step": 10615 + }, + { + "epoch": 2.41547212741752, + "grad_norm": 1.621949265713317, + "learning_rate": 6.58267639982907e-07, + "loss": 0.0415, + "step": 10616 + }, + { + "epoch": 2.4156996587030717, + "grad_norm": 2.39749211543786, + "learning_rate": 6.58178414983183e-07, + "loss": 0.0464, + "step": 10617 + }, + { + "epoch": 2.4159271899886234, + "grad_norm": 1.6038931260049425, + "learning_rate": 6.58089189305353e-07, + "loss": 0.0539, + "step": 10618 + }, + { + "epoch": 2.416154721274175, + "grad_norm": 1.0756461489983145, + "learning_rate": 6.579999629512407e-07, + "loss": 0.0148, + "step": 10619 + }, + { + "epoch": 2.416382252559727, + "grad_norm": 1.540697181124578, + "learning_rate": 6.579107359226695e-07, + "loss": 0.0189, + "step": 10620 + }, + { + "epoch": 2.4166097838452787, + "grad_norm": 1.3854580522572384, + "learning_rate": 6.578215082214629e-07, + "loss": 0.0533, + "step": 10621 + }, + { + "epoch": 2.4168373151308304, + "grad_norm": 1.4341453049106634, + "learning_rate": 6.577322798494449e-07, + "loss": 0.0296, + "step": 10622 + }, + { + "epoch": 2.417064846416382, + "grad_norm": 1.5702105258088734, + "learning_rate": 6.576430508084393e-07, + "loss": 0.031, + "step": 10623 + }, + { + "epoch": 2.417292377701934, + "grad_norm": 1.3171292579547644, + "learning_rate": 6.575538211002693e-07, + "loss": 0.03, + "step": 10624 + }, + { + "epoch": 2.4175199089874857, + "grad_norm": 1.4878630861481952, + "learning_rate": 6.57464590726759e-07, + "loss": 0.0325, + "step": 10625 + }, + { + "epoch": 2.4177474402730375, + "grad_norm": 1.2318300992441433, + "learning_rate": 6.573753596897318e-07, + "loss": 0.0255, + "step": 10626 + }, + { + "epoch": 2.417974971558589, + "grad_norm": 1.1438230000394354, + "learning_rate": 6.572861279910114e-07, + "loss": 0.0191, + "step": 10627 + }, + { + "epoch": 2.418202502844141, + "grad_norm": 1.9748913576549123, + "learning_rate": 6.571968956324218e-07, + "loss": 0.0579, + "step": 10628 + }, + { + "epoch": 2.4184300341296927, + "grad_norm": 2.548692414052215, + "learning_rate": 6.571076626157866e-07, + "loss": 0.0563, + "step": 10629 + }, + { + "epoch": 2.4186575654152445, + "grad_norm": 1.841220556087748, + "learning_rate": 6.570184289429297e-07, + "loss": 0.1044, + "step": 10630 + }, + { + "epoch": 2.418885096700796, + "grad_norm": 2.417142875175133, + "learning_rate": 6.569291946156746e-07, + "loss": 0.1108, + "step": 10631 + }, + { + "epoch": 2.419112627986348, + "grad_norm": 1.3672577776495731, + "learning_rate": 6.568399596358453e-07, + "loss": 0.0347, + "step": 10632 + }, + { + "epoch": 2.4193401592718997, + "grad_norm": 1.1848394405550018, + "learning_rate": 6.567507240052655e-07, + "loss": 0.0157, + "step": 10633 + }, + { + "epoch": 2.4195676905574515, + "grad_norm": 0.7408360795309923, + "learning_rate": 6.566614877257591e-07, + "loss": 0.0515, + "step": 10634 + }, + { + "epoch": 2.419795221843003, + "grad_norm": 3.5702580931507053, + "learning_rate": 6.565722507991497e-07, + "loss": 0.058, + "step": 10635 + }, + { + "epoch": 2.4200227531285554, + "grad_norm": 1.0847262663520345, + "learning_rate": 6.564830132272617e-07, + "loss": 0.0718, + "step": 10636 + }, + { + "epoch": 2.4202502844141067, + "grad_norm": 2.887173513609022, + "learning_rate": 6.563937750119183e-07, + "loss": 0.0981, + "step": 10637 + }, + { + "epoch": 2.420477815699659, + "grad_norm": 1.5870916114857796, + "learning_rate": 6.563045361549436e-07, + "loss": 0.0485, + "step": 10638 + }, + { + "epoch": 2.4207053469852102, + "grad_norm": 1.0970076404749562, + "learning_rate": 6.562152966581615e-07, + "loss": 0.1013, + "step": 10639 + }, + { + "epoch": 2.4209328782707624, + "grad_norm": 2.0280376001713885, + "learning_rate": 6.561260565233961e-07, + "loss": 0.0529, + "step": 10640 + }, + { + "epoch": 2.421160409556314, + "grad_norm": 1.1391187745397657, + "learning_rate": 6.56036815752471e-07, + "loss": 0.0148, + "step": 10641 + }, + { + "epoch": 2.421387940841866, + "grad_norm": 1.7302436040811533, + "learning_rate": 6.559475743472101e-07, + "loss": 0.0572, + "step": 10642 + }, + { + "epoch": 2.4216154721274177, + "grad_norm": 1.6632283107436028, + "learning_rate": 6.558583323094378e-07, + "loss": 0.0401, + "step": 10643 + }, + { + "epoch": 2.4218430034129694, + "grad_norm": 1.0592254181343634, + "learning_rate": 6.557690896409774e-07, + "loss": 0.0568, + "step": 10644 + }, + { + "epoch": 2.422070534698521, + "grad_norm": 2.3529914876801574, + "learning_rate": 6.556798463436531e-07, + "loss": 0.0776, + "step": 10645 + }, + { + "epoch": 2.422298065984073, + "grad_norm": 1.6577698001492, + "learning_rate": 6.555906024192892e-07, + "loss": 0.0757, + "step": 10646 + }, + { + "epoch": 2.4225255972696247, + "grad_norm": 0.9053635418106493, + "learning_rate": 6.555013578697092e-07, + "loss": 0.0306, + "step": 10647 + }, + { + "epoch": 2.4227531285551764, + "grad_norm": 1.5058578957353632, + "learning_rate": 6.554121126967375e-07, + "loss": 0.131, + "step": 10648 + }, + { + "epoch": 2.422980659840728, + "grad_norm": 1.870336040572531, + "learning_rate": 6.553228669021977e-07, + "loss": 0.1092, + "step": 10649 + }, + { + "epoch": 2.42320819112628, + "grad_norm": 1.9542945047991342, + "learning_rate": 6.552336204879142e-07, + "loss": 0.0669, + "step": 10650 + }, + { + "epoch": 2.4234357224118317, + "grad_norm": 1.7999047473753527, + "learning_rate": 6.551443734557108e-07, + "loss": 0.0424, + "step": 10651 + }, + { + "epoch": 2.4236632536973834, + "grad_norm": 0.9827538108341685, + "learning_rate": 6.550551258074115e-07, + "loss": 0.0203, + "step": 10652 + }, + { + "epoch": 2.423890784982935, + "grad_norm": 1.4152410262112376, + "learning_rate": 6.549658775448406e-07, + "loss": 0.0705, + "step": 10653 + }, + { + "epoch": 2.424118316268487, + "grad_norm": 1.08826764352521, + "learning_rate": 6.54876628669822e-07, + "loss": 0.0267, + "step": 10654 + }, + { + "epoch": 2.4243458475540387, + "grad_norm": 1.5384408674930627, + "learning_rate": 6.547873791841799e-07, + "loss": 0.0475, + "step": 10655 + }, + { + "epoch": 2.4245733788395905, + "grad_norm": 1.6667793947535348, + "learning_rate": 6.546981290897383e-07, + "loss": 0.0392, + "step": 10656 + }, + { + "epoch": 2.424800910125142, + "grad_norm": 1.5860211298246367, + "learning_rate": 6.546088783883215e-07, + "loss": 0.0739, + "step": 10657 + }, + { + "epoch": 2.425028441410694, + "grad_norm": 1.8038406137544518, + "learning_rate": 6.545196270817531e-07, + "loss": 0.0512, + "step": 10658 + }, + { + "epoch": 2.4252559726962457, + "grad_norm": 2.1572391868774745, + "learning_rate": 6.544303751718577e-07, + "loss": 0.134, + "step": 10659 + }, + { + "epoch": 2.4254835039817975, + "grad_norm": 1.8006479014105914, + "learning_rate": 6.543411226604595e-07, + "loss": 0.0238, + "step": 10660 + }, + { + "epoch": 2.425711035267349, + "grad_norm": 2.208639376759731, + "learning_rate": 6.542518695493823e-07, + "loss": 0.0769, + "step": 10661 + }, + { + "epoch": 2.425938566552901, + "grad_norm": 1.9443418278552014, + "learning_rate": 6.541626158404506e-07, + "loss": 0.1386, + "step": 10662 + }, + { + "epoch": 2.4261660978384527, + "grad_norm": 1.5754241156737616, + "learning_rate": 6.540733615354882e-07, + "loss": 0.0689, + "step": 10663 + }, + { + "epoch": 2.4263936291240045, + "grad_norm": 2.078032308936215, + "learning_rate": 6.539841066363198e-07, + "loss": 0.0436, + "step": 10664 + }, + { + "epoch": 2.426621160409556, + "grad_norm": 1.8687785949756106, + "learning_rate": 6.538948511447692e-07, + "loss": 0.0917, + "step": 10665 + }, + { + "epoch": 2.426848691695108, + "grad_norm": 1.3398527107964846, + "learning_rate": 6.538055950626608e-07, + "loss": 0.0846, + "step": 10666 + }, + { + "epoch": 2.4270762229806597, + "grad_norm": 1.2825273691433272, + "learning_rate": 6.537163383918188e-07, + "loss": 0.1239, + "step": 10667 + }, + { + "epoch": 2.4273037542662115, + "grad_norm": 1.9280214421112567, + "learning_rate": 6.536270811340674e-07, + "loss": 0.0986, + "step": 10668 + }, + { + "epoch": 2.4275312855517632, + "grad_norm": 1.4151310712289835, + "learning_rate": 6.53537823291231e-07, + "loss": 0.1475, + "step": 10669 + }, + { + "epoch": 2.427758816837315, + "grad_norm": 1.8422843240846454, + "learning_rate": 6.534485648651337e-07, + "loss": 0.0652, + "step": 10670 + }, + { + "epoch": 2.4279863481228667, + "grad_norm": 1.819895056833755, + "learning_rate": 6.533593058575997e-07, + "loss": 0.0312, + "step": 10671 + }, + { + "epoch": 2.4282138794084185, + "grad_norm": 2.060412485749707, + "learning_rate": 6.532700462704534e-07, + "loss": 0.0885, + "step": 10672 + }, + { + "epoch": 2.4284414106939702, + "grad_norm": 1.3171707458026929, + "learning_rate": 6.531807861055194e-07, + "loss": 0.0472, + "step": 10673 + }, + { + "epoch": 2.428668941979522, + "grad_norm": 1.6014971080885667, + "learning_rate": 6.530915253646219e-07, + "loss": 0.0236, + "step": 10674 + }, + { + "epoch": 2.428896473265074, + "grad_norm": 0.991228100862655, + "learning_rate": 6.530022640495845e-07, + "loss": 0.0325, + "step": 10675 + }, + { + "epoch": 2.4291240045506255, + "grad_norm": 1.7731158077990916, + "learning_rate": 6.529130021622324e-07, + "loss": 0.08, + "step": 10676 + }, + { + "epoch": 2.4293515358361777, + "grad_norm": 1.8359169565368136, + "learning_rate": 6.528237397043896e-07, + "loss": 0.119, + "step": 10677 + }, + { + "epoch": 2.429579067121729, + "grad_norm": 1.2173501994298692, + "learning_rate": 6.527344766778806e-07, + "loss": 0.0519, + "step": 10678 + }, + { + "epoch": 2.429806598407281, + "grad_norm": 0.8681978705300911, + "learning_rate": 6.526452130845296e-07, + "loss": 0.0122, + "step": 10679 + }, + { + "epoch": 2.430034129692833, + "grad_norm": 1.4590692338752178, + "learning_rate": 6.525559489261612e-07, + "loss": 0.0723, + "step": 10680 + }, + { + "epoch": 2.4302616609783847, + "grad_norm": 2.0795092949387683, + "learning_rate": 6.524666842045997e-07, + "loss": 0.1308, + "step": 10681 + }, + { + "epoch": 2.4304891922639364, + "grad_norm": 1.772161161399589, + "learning_rate": 6.523774189216692e-07, + "loss": 0.0634, + "step": 10682 + }, + { + "epoch": 2.430716723549488, + "grad_norm": 2.1812446826159766, + "learning_rate": 6.522881530791945e-07, + "loss": 0.1379, + "step": 10683 + }, + { + "epoch": 2.43094425483504, + "grad_norm": 1.749884881161601, + "learning_rate": 6.521988866790001e-07, + "loss": 0.0363, + "step": 10684 + }, + { + "epoch": 2.4311717861205917, + "grad_norm": 2.262473775430507, + "learning_rate": 6.5210961972291e-07, + "loss": 0.0289, + "step": 10685 + }, + { + "epoch": 2.4313993174061435, + "grad_norm": 2.106821787850952, + "learning_rate": 6.520203522127492e-07, + "loss": 0.038, + "step": 10686 + }, + { + "epoch": 2.431626848691695, + "grad_norm": 1.4823470626075748, + "learning_rate": 6.519310841503419e-07, + "loss": 0.0172, + "step": 10687 + }, + { + "epoch": 2.431854379977247, + "grad_norm": 1.7654943667039744, + "learning_rate": 6.518418155375123e-07, + "loss": 0.0702, + "step": 10688 + }, + { + "epoch": 2.4320819112627987, + "grad_norm": 1.6415241948042603, + "learning_rate": 6.517525463760852e-07, + "loss": 0.1197, + "step": 10689 + }, + { + "epoch": 2.4323094425483505, + "grad_norm": 1.8510138218220162, + "learning_rate": 6.516632766678853e-07, + "loss": 0.0277, + "step": 10690 + }, + { + "epoch": 2.432536973833902, + "grad_norm": 1.1099310953847858, + "learning_rate": 6.515740064147366e-07, + "loss": 0.0162, + "step": 10691 + }, + { + "epoch": 2.432764505119454, + "grad_norm": 1.1581248515271256, + "learning_rate": 6.514847356184639e-07, + "loss": 0.0431, + "step": 10692 + }, + { + "epoch": 2.4329920364050057, + "grad_norm": 1.1123917836464456, + "learning_rate": 6.513954642808919e-07, + "loss": 0.1107, + "step": 10693 + }, + { + "epoch": 2.4332195676905575, + "grad_norm": 1.2671981504873233, + "learning_rate": 6.513061924038448e-07, + "loss": 0.0316, + "step": 10694 + }, + { + "epoch": 2.4334470989761092, + "grad_norm": 1.2347140093114832, + "learning_rate": 6.512169199891473e-07, + "loss": 0.1102, + "step": 10695 + }, + { + "epoch": 2.433674630261661, + "grad_norm": 1.7608166360822675, + "learning_rate": 6.51127647038624e-07, + "loss": 0.0927, + "step": 10696 + }, + { + "epoch": 2.4339021615472127, + "grad_norm": 2.2439651714843816, + "learning_rate": 6.510383735540994e-07, + "loss": 0.0517, + "step": 10697 + }, + { + "epoch": 2.4341296928327645, + "grad_norm": 2.016108585740252, + "learning_rate": 6.509490995373983e-07, + "loss": 0.026, + "step": 10698 + }, + { + "epoch": 2.4343572241183162, + "grad_norm": 1.3594466625214487, + "learning_rate": 6.50859824990345e-07, + "loss": 0.063, + "step": 10699 + }, + { + "epoch": 2.434584755403868, + "grad_norm": 1.5356254084154082, + "learning_rate": 6.507705499147641e-07, + "loss": 0.0333, + "step": 10700 + }, + { + "epoch": 2.4348122866894197, + "grad_norm": 1.247460061236974, + "learning_rate": 6.506812743124806e-07, + "loss": 0.0434, + "step": 10701 + }, + { + "epoch": 2.4350398179749715, + "grad_norm": 1.540275392451105, + "learning_rate": 6.505919981853187e-07, + "loss": 0.1154, + "step": 10702 + }, + { + "epoch": 2.4352673492605232, + "grad_norm": 2.1244752670567317, + "learning_rate": 6.505027215351033e-07, + "loss": 0.0556, + "step": 10703 + }, + { + "epoch": 2.435494880546075, + "grad_norm": 1.8341931503670963, + "learning_rate": 6.504134443636591e-07, + "loss": 0.1113, + "step": 10704 + }, + { + "epoch": 2.4357224118316267, + "grad_norm": 2.3116116212807407, + "learning_rate": 6.503241666728105e-07, + "loss": 0.0521, + "step": 10705 + }, + { + "epoch": 2.4359499431171785, + "grad_norm": 1.6178987248860397, + "learning_rate": 6.502348884643824e-07, + "loss": 0.0427, + "step": 10706 + }, + { + "epoch": 2.4361774744027302, + "grad_norm": 1.9281110882142614, + "learning_rate": 6.501456097401992e-07, + "loss": 0.0455, + "step": 10707 + }, + { + "epoch": 2.436405005688282, + "grad_norm": 1.8559225960543262, + "learning_rate": 6.50056330502086e-07, + "loss": 0.0454, + "step": 10708 + }, + { + "epoch": 2.4366325369738338, + "grad_norm": 2.3483742537162424, + "learning_rate": 6.499670507518671e-07, + "loss": 0.0527, + "step": 10709 + }, + { + "epoch": 2.4368600682593855, + "grad_norm": 1.4458951216701315, + "learning_rate": 6.498777704913675e-07, + "loss": 0.0354, + "step": 10710 + }, + { + "epoch": 2.4370875995449373, + "grad_norm": 2.0427295101794942, + "learning_rate": 6.497884897224119e-07, + "loss": 0.0343, + "step": 10711 + }, + { + "epoch": 2.437315130830489, + "grad_norm": 2.405025995931995, + "learning_rate": 6.49699208446825e-07, + "loss": 0.0823, + "step": 10712 + }, + { + "epoch": 2.4375426621160408, + "grad_norm": 2.0468929323916676, + "learning_rate": 6.496099266664314e-07, + "loss": 0.0459, + "step": 10713 + }, + { + "epoch": 2.437770193401593, + "grad_norm": 1.7648377287593435, + "learning_rate": 6.495206443830561e-07, + "loss": 0.0714, + "step": 10714 + }, + { + "epoch": 2.4379977246871443, + "grad_norm": 1.8283934517937341, + "learning_rate": 6.494313615985235e-07, + "loss": 0.0499, + "step": 10715 + }, + { + "epoch": 2.4382252559726965, + "grad_norm": 1.7208692980239855, + "learning_rate": 6.493420783146587e-07, + "loss": 0.1212, + "step": 10716 + }, + { + "epoch": 2.4384527872582478, + "grad_norm": 1.716858466875207, + "learning_rate": 6.492527945332865e-07, + "loss": 0.1226, + "step": 10717 + }, + { + "epoch": 2.4386803185438, + "grad_norm": 1.7351072488692236, + "learning_rate": 6.491635102562315e-07, + "loss": 0.082, + "step": 10718 + }, + { + "epoch": 2.4389078498293517, + "grad_norm": 1.4982946300057065, + "learning_rate": 6.490742254853187e-07, + "loss": 0.0204, + "step": 10719 + }, + { + "epoch": 2.4391353811149035, + "grad_norm": 0.9082073619983685, + "learning_rate": 6.489849402223729e-07, + "loss": 0.0261, + "step": 10720 + }, + { + "epoch": 2.439362912400455, + "grad_norm": 1.6593591409181783, + "learning_rate": 6.488956544692187e-07, + "loss": 0.0362, + "step": 10721 + }, + { + "epoch": 2.439590443686007, + "grad_norm": 1.5444429079340354, + "learning_rate": 6.48806368227681e-07, + "loss": 0.1274, + "step": 10722 + }, + { + "epoch": 2.4398179749715587, + "grad_norm": 2.258854235516451, + "learning_rate": 6.487170814995849e-07, + "loss": 0.0541, + "step": 10723 + }, + { + "epoch": 2.4400455062571105, + "grad_norm": 1.375819855271847, + "learning_rate": 6.48627794286755e-07, + "loss": 0.0255, + "step": 10724 + }, + { + "epoch": 2.4402730375426622, + "grad_norm": 1.1788933902842234, + "learning_rate": 6.485385065910163e-07, + "loss": 0.0438, + "step": 10725 + }, + { + "epoch": 2.440500568828214, + "grad_norm": 1.5680757015826552, + "learning_rate": 6.484492184141937e-07, + "loss": 0.1305, + "step": 10726 + }, + { + "epoch": 2.4407281001137657, + "grad_norm": 1.476838607043402, + "learning_rate": 6.48359929758112e-07, + "loss": 0.0746, + "step": 10727 + }, + { + "epoch": 2.4409556313993175, + "grad_norm": 2.3685586806747745, + "learning_rate": 6.482706406245961e-07, + "loss": 0.1306, + "step": 10728 + }, + { + "epoch": 2.4411831626848692, + "grad_norm": 1.879831131236823, + "learning_rate": 6.481813510154706e-07, + "loss": 0.0547, + "step": 10729 + }, + { + "epoch": 2.441410693970421, + "grad_norm": 1.3697853468516277, + "learning_rate": 6.480920609325611e-07, + "loss": 0.0208, + "step": 10730 + }, + { + "epoch": 2.4416382252559727, + "grad_norm": 1.4657681407617524, + "learning_rate": 6.480027703776923e-07, + "loss": 0.0602, + "step": 10731 + }, + { + "epoch": 2.4418657565415245, + "grad_norm": 1.9042919050218716, + "learning_rate": 6.479134793526887e-07, + "loss": 0.0388, + "step": 10732 + }, + { + "epoch": 2.4420932878270762, + "grad_norm": 1.6406896281331167, + "learning_rate": 6.478241878593755e-07, + "loss": 0.0228, + "step": 10733 + }, + { + "epoch": 2.442320819112628, + "grad_norm": 1.2483716306531623, + "learning_rate": 6.47734895899578e-07, + "loss": 0.0421, + "step": 10734 + }, + { + "epoch": 2.4425483503981797, + "grad_norm": 2.785750578757238, + "learning_rate": 6.476456034751207e-07, + "loss": 0.0323, + "step": 10735 + }, + { + "epoch": 2.4427758816837315, + "grad_norm": 1.876625132138733, + "learning_rate": 6.475563105878285e-07, + "loss": 0.0584, + "step": 10736 + }, + { + "epoch": 2.4430034129692833, + "grad_norm": 1.3935791176038599, + "learning_rate": 6.474670172395271e-07, + "loss": 0.0377, + "step": 10737 + }, + { + "epoch": 2.443230944254835, + "grad_norm": 1.275301134582811, + "learning_rate": 6.473777234320408e-07, + "loss": 0.0678, + "step": 10738 + }, + { + "epoch": 2.4434584755403868, + "grad_norm": 1.5153471324070782, + "learning_rate": 6.472884291671947e-07, + "loss": 0.1395, + "step": 10739 + }, + { + "epoch": 2.4436860068259385, + "grad_norm": 2.0215771823642883, + "learning_rate": 6.47199134446814e-07, + "loss": 0.0277, + "step": 10740 + }, + { + "epoch": 2.4439135381114903, + "grad_norm": 1.248726231168551, + "learning_rate": 6.471098392727238e-07, + "loss": 0.0418, + "step": 10741 + }, + { + "epoch": 2.444141069397042, + "grad_norm": 1.9909115972775195, + "learning_rate": 6.470205436467487e-07, + "loss": 0.0956, + "step": 10742 + }, + { + "epoch": 2.4443686006825938, + "grad_norm": 1.1606253216543372, + "learning_rate": 6.469312475707141e-07, + "loss": 0.0451, + "step": 10743 + }, + { + "epoch": 2.4445961319681455, + "grad_norm": 1.274108763143027, + "learning_rate": 6.468419510464452e-07, + "loss": 0.0842, + "step": 10744 + }, + { + "epoch": 2.4448236632536973, + "grad_norm": 1.5163402041606309, + "learning_rate": 6.467526540757666e-07, + "loss": 0.0604, + "step": 10745 + }, + { + "epoch": 2.445051194539249, + "grad_norm": 2.029191898687564, + "learning_rate": 6.466633566605036e-07, + "loss": 0.0587, + "step": 10746 + }, + { + "epoch": 2.4452787258248008, + "grad_norm": 1.758333279331689, + "learning_rate": 6.465740588024813e-07, + "loss": 0.0511, + "step": 10747 + }, + { + "epoch": 2.4455062571103525, + "grad_norm": 1.6379300702116257, + "learning_rate": 6.464847605035247e-07, + "loss": 0.1883, + "step": 10748 + }, + { + "epoch": 2.4457337883959043, + "grad_norm": 1.9636421701081885, + "learning_rate": 6.46395461765459e-07, + "loss": 0.0359, + "step": 10749 + }, + { + "epoch": 2.445961319681456, + "grad_norm": 1.9142765817445715, + "learning_rate": 6.463061625901093e-07, + "loss": 0.0755, + "step": 10750 + }, + { + "epoch": 2.4461888509670078, + "grad_norm": 0.8841446043054305, + "learning_rate": 6.462168629793008e-07, + "loss": 0.013, + "step": 10751 + }, + { + "epoch": 2.4464163822525595, + "grad_norm": 2.1394249546323945, + "learning_rate": 6.461275629348581e-07, + "loss": 0.1495, + "step": 10752 + }, + { + "epoch": 2.4466439135381117, + "grad_norm": 0.7270738165419912, + "learning_rate": 6.460382624586069e-07, + "loss": 0.0074, + "step": 10753 + }, + { + "epoch": 2.446871444823663, + "grad_norm": 1.15289704204101, + "learning_rate": 6.45948961552372e-07, + "loss": 0.0562, + "step": 10754 + }, + { + "epoch": 2.4470989761092152, + "grad_norm": 1.6473230116533466, + "learning_rate": 6.45859660217979e-07, + "loss": 0.116, + "step": 10755 + }, + { + "epoch": 2.4473265073947665, + "grad_norm": 2.547342183522831, + "learning_rate": 6.457703584572525e-07, + "loss": 0.0521, + "step": 10756 + }, + { + "epoch": 2.4475540386803187, + "grad_norm": 1.6838663706265002, + "learning_rate": 6.45681056272018e-07, + "loss": 0.0455, + "step": 10757 + }, + { + "epoch": 2.4477815699658705, + "grad_norm": 1.2212377724665087, + "learning_rate": 6.455917536641006e-07, + "loss": 0.04, + "step": 10758 + }, + { + "epoch": 2.4480091012514222, + "grad_norm": 1.490475672931126, + "learning_rate": 6.455024506353252e-07, + "loss": 0.0671, + "step": 10759 + }, + { + "epoch": 2.448236632536974, + "grad_norm": 1.5184591661740638, + "learning_rate": 6.454131471875176e-07, + "loss": 0.0294, + "step": 10760 + }, + { + "epoch": 2.4484641638225257, + "grad_norm": 1.3938123308936536, + "learning_rate": 6.453238433225026e-07, + "loss": 0.1112, + "step": 10761 + }, + { + "epoch": 2.4486916951080775, + "grad_norm": 1.252816498717164, + "learning_rate": 6.452345390421054e-07, + "loss": 0.0199, + "step": 10762 + }, + { + "epoch": 2.4489192263936292, + "grad_norm": 2.3025806568067817, + "learning_rate": 6.451452343481512e-07, + "loss": 0.0483, + "step": 10763 + }, + { + "epoch": 2.449146757679181, + "grad_norm": 2.2006329204755817, + "learning_rate": 6.450559292424655e-07, + "loss": 0.0595, + "step": 10764 + }, + { + "epoch": 2.4493742889647327, + "grad_norm": 2.14628033721118, + "learning_rate": 6.449666237268733e-07, + "loss": 0.0297, + "step": 10765 + }, + { + "epoch": 2.4496018202502845, + "grad_norm": 1.617248945564638, + "learning_rate": 6.448773178031996e-07, + "loss": 0.0448, + "step": 10766 + }, + { + "epoch": 2.4498293515358363, + "grad_norm": 2.103623179671804, + "learning_rate": 6.447880114732702e-07, + "loss": 0.0973, + "step": 10767 + }, + { + "epoch": 2.450056882821388, + "grad_norm": 1.0954074293093925, + "learning_rate": 6.4469870473891e-07, + "loss": 0.04, + "step": 10768 + }, + { + "epoch": 2.4502844141069398, + "grad_norm": 1.2354183827654832, + "learning_rate": 6.446093976019443e-07, + "loss": 0.0806, + "step": 10769 + }, + { + "epoch": 2.4505119453924915, + "grad_norm": 1.0217733571217007, + "learning_rate": 6.445200900641986e-07, + "loss": 0.0274, + "step": 10770 + }, + { + "epoch": 2.4507394766780433, + "grad_norm": 1.0159088874973368, + "learning_rate": 6.444307821274979e-07, + "loss": 0.0392, + "step": 10771 + }, + { + "epoch": 2.450967007963595, + "grad_norm": 1.9220407075099928, + "learning_rate": 6.443414737936677e-07, + "loss": 0.1153, + "step": 10772 + }, + { + "epoch": 2.4511945392491468, + "grad_norm": 1.4222736350031056, + "learning_rate": 6.442521650645329e-07, + "loss": 0.025, + "step": 10773 + }, + { + "epoch": 2.4514220705346985, + "grad_norm": 1.6802408563288893, + "learning_rate": 6.441628559419194e-07, + "loss": 0.0609, + "step": 10774 + }, + { + "epoch": 2.4516496018202503, + "grad_norm": 1.2694793613626496, + "learning_rate": 6.440735464276524e-07, + "loss": 0.0164, + "step": 10775 + }, + { + "epoch": 2.451877133105802, + "grad_norm": 3.622568264399461, + "learning_rate": 6.439842365235566e-07, + "loss": 0.0614, + "step": 10776 + }, + { + "epoch": 2.4521046643913538, + "grad_norm": 1.0645642623729403, + "learning_rate": 6.43894926231458e-07, + "loss": 0.0541, + "step": 10777 + }, + { + "epoch": 2.4523321956769055, + "grad_norm": 1.1049432103300538, + "learning_rate": 6.438056155531816e-07, + "loss": 0.0316, + "step": 10778 + }, + { + "epoch": 2.4525597269624573, + "grad_norm": 2.6035304274976063, + "learning_rate": 6.437163044905528e-07, + "loss": 0.1614, + "step": 10779 + }, + { + "epoch": 2.452787258248009, + "grad_norm": 1.6884683257989475, + "learning_rate": 6.436269930453971e-07, + "loss": 0.0496, + "step": 10780 + }, + { + "epoch": 2.453014789533561, + "grad_norm": 1.3339295587517084, + "learning_rate": 6.4353768121954e-07, + "loss": 0.0516, + "step": 10781 + }, + { + "epoch": 2.4532423208191125, + "grad_norm": 1.4020930989417688, + "learning_rate": 6.434483690148063e-07, + "loss": 0.0577, + "step": 10782 + }, + { + "epoch": 2.4534698521046643, + "grad_norm": 1.3912072581744486, + "learning_rate": 6.43359056433022e-07, + "loss": 0.0385, + "step": 10783 + }, + { + "epoch": 2.453697383390216, + "grad_norm": 1.38982276338953, + "learning_rate": 6.43269743476012e-07, + "loss": 0.0631, + "step": 10784 + }, + { + "epoch": 2.453924914675768, + "grad_norm": 1.5348886513779152, + "learning_rate": 6.43180430145602e-07, + "loss": 0.0274, + "step": 10785 + }, + { + "epoch": 2.4541524459613195, + "grad_norm": 3.7506311156621175, + "learning_rate": 6.430911164436172e-07, + "loss": 0.1153, + "step": 10786 + }, + { + "epoch": 2.4543799772468713, + "grad_norm": 1.6448626628839786, + "learning_rate": 6.430018023718833e-07, + "loss": 0.0584, + "step": 10787 + }, + { + "epoch": 2.454607508532423, + "grad_norm": 1.0534019966424395, + "learning_rate": 6.429124879322256e-07, + "loss": 0.0314, + "step": 10788 + }, + { + "epoch": 2.454835039817975, + "grad_norm": 1.4640930397866614, + "learning_rate": 6.42823173126469e-07, + "loss": 0.1025, + "step": 10789 + }, + { + "epoch": 2.4550625711035265, + "grad_norm": 1.8435906346136508, + "learning_rate": 6.427338579564397e-07, + "loss": 0.0702, + "step": 10790 + }, + { + "epoch": 2.4552901023890783, + "grad_norm": 1.5390586109838404, + "learning_rate": 6.426445424239629e-07, + "loss": 0.0507, + "step": 10791 + }, + { + "epoch": 2.4555176336746305, + "grad_norm": 1.792875825383981, + "learning_rate": 6.425552265308639e-07, + "loss": 0.0292, + "step": 10792 + }, + { + "epoch": 2.455745164960182, + "grad_norm": 1.3350294846723614, + "learning_rate": 6.424659102789681e-07, + "loss": 0.0387, + "step": 10793 + }, + { + "epoch": 2.455972696245734, + "grad_norm": 1.1981536555136512, + "learning_rate": 6.423765936701012e-07, + "loss": 0.0798, + "step": 10794 + }, + { + "epoch": 2.4562002275312853, + "grad_norm": 1.5112908389859945, + "learning_rate": 6.422872767060886e-07, + "loss": 0.0387, + "step": 10795 + }, + { + "epoch": 2.4564277588168375, + "grad_norm": 2.4397855264389885, + "learning_rate": 6.421979593887555e-07, + "loss": 0.0549, + "step": 10796 + }, + { + "epoch": 2.4566552901023893, + "grad_norm": 1.7392320043842686, + "learning_rate": 6.421086417199277e-07, + "loss": 0.026, + "step": 10797 + }, + { + "epoch": 2.456882821387941, + "grad_norm": 1.5748479447667432, + "learning_rate": 6.420193237014306e-07, + "loss": 0.0789, + "step": 10798 + }, + { + "epoch": 2.4571103526734928, + "grad_norm": 1.275351095421082, + "learning_rate": 6.419300053350898e-07, + "loss": 0.0476, + "step": 10799 + }, + { + "epoch": 2.4573378839590445, + "grad_norm": 1.5009957426345688, + "learning_rate": 6.418406866227306e-07, + "loss": 0.0412, + "step": 10800 + }, + { + "epoch": 2.4575654152445963, + "grad_norm": 1.6812093111025637, + "learning_rate": 6.417513675661787e-07, + "loss": 0.0414, + "step": 10801 + }, + { + "epoch": 2.457792946530148, + "grad_norm": 1.3673637868393416, + "learning_rate": 6.416620481672595e-07, + "loss": 0.0399, + "step": 10802 + }, + { + "epoch": 2.4580204778156998, + "grad_norm": 1.8916693927031183, + "learning_rate": 6.415727284277984e-07, + "loss": 0.0895, + "step": 10803 + }, + { + "epoch": 2.4582480091012515, + "grad_norm": 2.524607698907239, + "learning_rate": 6.414834083496212e-07, + "loss": 0.0465, + "step": 10804 + }, + { + "epoch": 2.4584755403868033, + "grad_norm": 2.2817091471153605, + "learning_rate": 6.413940879345533e-07, + "loss": 0.1208, + "step": 10805 + }, + { + "epoch": 2.458703071672355, + "grad_norm": 1.4010689281603355, + "learning_rate": 6.413047671844203e-07, + "loss": 0.082, + "step": 10806 + }, + { + "epoch": 2.4589306029579068, + "grad_norm": 1.0169717727837417, + "learning_rate": 6.412154461010477e-07, + "loss": 0.0663, + "step": 10807 + }, + { + "epoch": 2.4591581342434585, + "grad_norm": 1.6391566607979509, + "learning_rate": 6.411261246862611e-07, + "loss": 0.0499, + "step": 10808 + }, + { + "epoch": 2.4593856655290103, + "grad_norm": 1.148975078128176, + "learning_rate": 6.410368029418859e-07, + "loss": 0.0659, + "step": 10809 + }, + { + "epoch": 2.459613196814562, + "grad_norm": 1.99952874926377, + "learning_rate": 6.40947480869748e-07, + "loss": 0.0997, + "step": 10810 + }, + { + "epoch": 2.459840728100114, + "grad_norm": 2.9458744665851273, + "learning_rate": 6.408581584716728e-07, + "loss": 0.0409, + "step": 10811 + }, + { + "epoch": 2.4600682593856655, + "grad_norm": 1.5741732885923228, + "learning_rate": 6.407688357494858e-07, + "loss": 0.0375, + "step": 10812 + }, + { + "epoch": 2.4602957906712173, + "grad_norm": 1.7425425192381223, + "learning_rate": 6.406795127050126e-07, + "loss": 0.0264, + "step": 10813 + }, + { + "epoch": 2.460523321956769, + "grad_norm": 1.6044876619606474, + "learning_rate": 6.405901893400791e-07, + "loss": 0.0309, + "step": 10814 + }, + { + "epoch": 2.460750853242321, + "grad_norm": 1.6192559761779082, + "learning_rate": 6.405008656565105e-07, + "loss": 0.062, + "step": 10815 + }, + { + "epoch": 2.4609783845278725, + "grad_norm": 2.0358550989694773, + "learning_rate": 6.404115416561326e-07, + "loss": 0.0363, + "step": 10816 + }, + { + "epoch": 2.4612059158134243, + "grad_norm": 1.9579573986723964, + "learning_rate": 6.403222173407711e-07, + "loss": 0.1106, + "step": 10817 + }, + { + "epoch": 2.461433447098976, + "grad_norm": 1.103167386812476, + "learning_rate": 6.402328927122514e-07, + "loss": 0.0529, + "step": 10818 + }, + { + "epoch": 2.461660978384528, + "grad_norm": 1.5010142939431699, + "learning_rate": 6.401435677723995e-07, + "loss": 0.0554, + "step": 10819 + }, + { + "epoch": 2.4618885096700796, + "grad_norm": 2.3307306230733, + "learning_rate": 6.400542425230407e-07, + "loss": 0.0508, + "step": 10820 + }, + { + "epoch": 2.4621160409556313, + "grad_norm": 0.7516582829505136, + "learning_rate": 6.399649169660007e-07, + "loss": 0.0305, + "step": 10821 + }, + { + "epoch": 2.462343572241183, + "grad_norm": 1.4594580184464123, + "learning_rate": 6.398755911031053e-07, + "loss": 0.0645, + "step": 10822 + }, + { + "epoch": 2.462571103526735, + "grad_norm": 2.1346944993056276, + "learning_rate": 6.397862649361798e-07, + "loss": 0.1862, + "step": 10823 + }, + { + "epoch": 2.4627986348122866, + "grad_norm": 0.8743918396347649, + "learning_rate": 6.396969384670504e-07, + "loss": 0.0241, + "step": 10824 + }, + { + "epoch": 2.4630261660978383, + "grad_norm": 1.6970079992379572, + "learning_rate": 6.396076116975426e-07, + "loss": 0.1211, + "step": 10825 + }, + { + "epoch": 2.46325369738339, + "grad_norm": 2.2413093568023945, + "learning_rate": 6.395182846294816e-07, + "loss": 0.0887, + "step": 10826 + }, + { + "epoch": 2.463481228668942, + "grad_norm": 1.9700672436219968, + "learning_rate": 6.394289572646938e-07, + "loss": 0.0618, + "step": 10827 + }, + { + "epoch": 2.4637087599544936, + "grad_norm": 1.550471476068397, + "learning_rate": 6.393396296050043e-07, + "loss": 0.0597, + "step": 10828 + }, + { + "epoch": 2.4639362912400453, + "grad_norm": 1.2971818428546398, + "learning_rate": 6.392503016522392e-07, + "loss": 0.0551, + "step": 10829 + }, + { + "epoch": 2.464163822525597, + "grad_norm": 1.8438709238425102, + "learning_rate": 6.391609734082238e-07, + "loss": 0.1326, + "step": 10830 + }, + { + "epoch": 2.4643913538111493, + "grad_norm": 2.29719579436221, + "learning_rate": 6.390716448747841e-07, + "loss": 0.0585, + "step": 10831 + }, + { + "epoch": 2.4646188850967006, + "grad_norm": 1.1431082350176027, + "learning_rate": 6.38982316053746e-07, + "loss": 0.0283, + "step": 10832 + }, + { + "epoch": 2.4648464163822528, + "grad_norm": 1.5486039126328415, + "learning_rate": 6.388929869469348e-07, + "loss": 0.118, + "step": 10833 + }, + { + "epoch": 2.465073947667804, + "grad_norm": 1.2811949077748084, + "learning_rate": 6.388036575561764e-07, + "loss": 0.1024, + "step": 10834 + }, + { + "epoch": 2.4653014789533563, + "grad_norm": 1.9528449906451266, + "learning_rate": 6.387143278832964e-07, + "loss": 0.0336, + "step": 10835 + }, + { + "epoch": 2.465529010238908, + "grad_norm": 1.4934876792933103, + "learning_rate": 6.386249979301207e-07, + "loss": 0.0866, + "step": 10836 + }, + { + "epoch": 2.4657565415244598, + "grad_norm": 1.346713413935413, + "learning_rate": 6.385356676984751e-07, + "loss": 0.0435, + "step": 10837 + }, + { + "epoch": 2.4659840728100115, + "grad_norm": 1.8334525389627976, + "learning_rate": 6.384463371901853e-07, + "loss": 0.0259, + "step": 10838 + }, + { + "epoch": 2.4662116040955633, + "grad_norm": 1.474866635603728, + "learning_rate": 6.383570064070768e-07, + "loss": 0.0437, + "step": 10839 + }, + { + "epoch": 2.466439135381115, + "grad_norm": 1.0519673060463886, + "learning_rate": 6.382676753509756e-07, + "loss": 0.0566, + "step": 10840 + }, + { + "epoch": 2.466666666666667, + "grad_norm": 1.7430158133885028, + "learning_rate": 6.381783440237076e-07, + "loss": 0.0712, + "step": 10841 + }, + { + "epoch": 2.4668941979522185, + "grad_norm": 1.6182931250989727, + "learning_rate": 6.380890124270982e-07, + "loss": 0.0406, + "step": 10842 + }, + { + "epoch": 2.4671217292377703, + "grad_norm": 1.4382384768853405, + "learning_rate": 6.379996805629733e-07, + "loss": 0.1288, + "step": 10843 + }, + { + "epoch": 2.467349260523322, + "grad_norm": 1.2369613458375917, + "learning_rate": 6.37910348433159e-07, + "loss": 0.0774, + "step": 10844 + }, + { + "epoch": 2.467576791808874, + "grad_norm": 0.8755195888088544, + "learning_rate": 6.378210160394807e-07, + "loss": 0.044, + "step": 10845 + }, + { + "epoch": 2.4678043230944255, + "grad_norm": 1.7498352397362558, + "learning_rate": 6.37731683383764e-07, + "loss": 0.0733, + "step": 10846 + }, + { + "epoch": 2.4680318543799773, + "grad_norm": 3.0243093650135178, + "learning_rate": 6.376423504678354e-07, + "loss": 0.0525, + "step": 10847 + }, + { + "epoch": 2.468259385665529, + "grad_norm": 1.9088213266639864, + "learning_rate": 6.375530172935203e-07, + "loss": 0.0791, + "step": 10848 + }, + { + "epoch": 2.468486916951081, + "grad_norm": 0.9999765427526403, + "learning_rate": 6.374636838626444e-07, + "loss": 0.0554, + "step": 10849 + }, + { + "epoch": 2.4687144482366326, + "grad_norm": 1.3301437955659614, + "learning_rate": 6.373743501770335e-07, + "loss": 0.0396, + "step": 10850 + }, + { + "epoch": 2.4689419795221843, + "grad_norm": 1.3990404816959723, + "learning_rate": 6.372850162385139e-07, + "loss": 0.0358, + "step": 10851 + }, + { + "epoch": 2.469169510807736, + "grad_norm": 1.4934607337372596, + "learning_rate": 6.371956820489107e-07, + "loss": 0.0678, + "step": 10852 + }, + { + "epoch": 2.469397042093288, + "grad_norm": 1.9751544493176358, + "learning_rate": 6.371063476100501e-07, + "loss": 0.0439, + "step": 10853 + }, + { + "epoch": 2.4696245733788396, + "grad_norm": 1.9053408951195272, + "learning_rate": 6.370170129237582e-07, + "loss": 0.0944, + "step": 10854 + }, + { + "epoch": 2.4698521046643913, + "grad_norm": 0.7055216183332879, + "learning_rate": 6.369276779918604e-07, + "loss": 0.0146, + "step": 10855 + }, + { + "epoch": 2.470079635949943, + "grad_norm": 1.332404056281277, + "learning_rate": 6.368383428161829e-07, + "loss": 0.0499, + "step": 10856 + }, + { + "epoch": 2.470307167235495, + "grad_norm": 1.5552345749572618, + "learning_rate": 6.36749007398551e-07, + "loss": 0.0394, + "step": 10857 + }, + { + "epoch": 2.4705346985210466, + "grad_norm": 1.9573735317377305, + "learning_rate": 6.366596717407912e-07, + "loss": 0.0359, + "step": 10858 + }, + { + "epoch": 2.4707622298065983, + "grad_norm": 2.142830137816076, + "learning_rate": 6.36570335844729e-07, + "loss": 0.0571, + "step": 10859 + }, + { + "epoch": 2.47098976109215, + "grad_norm": 1.9239905053259085, + "learning_rate": 6.364809997121901e-07, + "loss": 0.067, + "step": 10860 + }, + { + "epoch": 2.471217292377702, + "grad_norm": 1.8631922090993536, + "learning_rate": 6.363916633450009e-07, + "loss": 0.0458, + "step": 10861 + }, + { + "epoch": 2.4714448236632536, + "grad_norm": 1.437313273971417, + "learning_rate": 6.363023267449868e-07, + "loss": 0.0345, + "step": 10862 + }, + { + "epoch": 2.4716723549488053, + "grad_norm": 2.2565291195120363, + "learning_rate": 6.362129899139739e-07, + "loss": 0.091, + "step": 10863 + }, + { + "epoch": 2.471899886234357, + "grad_norm": 0.8965433620198081, + "learning_rate": 6.36123652853788e-07, + "loss": 0.0101, + "step": 10864 + }, + { + "epoch": 2.472127417519909, + "grad_norm": 1.563430697175068, + "learning_rate": 6.360343155662551e-07, + "loss": 0.1114, + "step": 10865 + }, + { + "epoch": 2.4723549488054606, + "grad_norm": 1.6041844994917043, + "learning_rate": 6.359449780532008e-07, + "loss": 0.0993, + "step": 10866 + }, + { + "epoch": 2.4725824800910123, + "grad_norm": 1.9128389674410409, + "learning_rate": 6.358556403164513e-07, + "loss": 0.1061, + "step": 10867 + }, + { + "epoch": 2.472810011376564, + "grad_norm": 1.9540243726959918, + "learning_rate": 6.357663023578324e-07, + "loss": 0.1383, + "step": 10868 + }, + { + "epoch": 2.473037542662116, + "grad_norm": 1.3820552685882872, + "learning_rate": 6.3567696417917e-07, + "loss": 0.0701, + "step": 10869 + }, + { + "epoch": 2.473265073947668, + "grad_norm": 1.8594423898077386, + "learning_rate": 6.3558762578229e-07, + "loss": 0.1291, + "step": 10870 + }, + { + "epoch": 2.4734926052332193, + "grad_norm": 2.1688244377617814, + "learning_rate": 6.354982871690184e-07, + "loss": 0.0756, + "step": 10871 + }, + { + "epoch": 2.4737201365187715, + "grad_norm": 1.567508337160604, + "learning_rate": 6.35408948341181e-07, + "loss": 0.0662, + "step": 10872 + }, + { + "epoch": 2.473947667804323, + "grad_norm": 1.3056905061664779, + "learning_rate": 6.353196093006035e-07, + "loss": 0.0383, + "step": 10873 + }, + { + "epoch": 2.474175199089875, + "grad_norm": 1.3423014901373584, + "learning_rate": 6.352302700491124e-07, + "loss": 0.0144, + "step": 10874 + }, + { + "epoch": 2.474402730375427, + "grad_norm": 1.9806297606684156, + "learning_rate": 6.351409305885332e-07, + "loss": 0.0388, + "step": 10875 + }, + { + "epoch": 2.4746302616609785, + "grad_norm": 1.987514408078129, + "learning_rate": 6.35051590920692e-07, + "loss": 0.0446, + "step": 10876 + }, + { + "epoch": 2.4748577929465303, + "grad_norm": 1.466224620623108, + "learning_rate": 6.349622510474146e-07, + "loss": 0.0676, + "step": 10877 + }, + { + "epoch": 2.475085324232082, + "grad_norm": 1.7727006116703508, + "learning_rate": 6.348729109705272e-07, + "loss": 0.1048, + "step": 10878 + }, + { + "epoch": 2.475312855517634, + "grad_norm": 1.0548915177972773, + "learning_rate": 6.347835706918555e-07, + "loss": 0.0135, + "step": 10879 + }, + { + "epoch": 2.4755403868031856, + "grad_norm": 1.2244378985072317, + "learning_rate": 6.346942302132253e-07, + "loss": 0.0216, + "step": 10880 + }, + { + "epoch": 2.4757679180887373, + "grad_norm": 1.5448357180995995, + "learning_rate": 6.34604889536463e-07, + "loss": 0.0784, + "step": 10881 + }, + { + "epoch": 2.475995449374289, + "grad_norm": 2.3915536157049977, + "learning_rate": 6.345155486633946e-07, + "loss": 0.0529, + "step": 10882 + }, + { + "epoch": 2.476222980659841, + "grad_norm": 5.283236051471409, + "learning_rate": 6.344262075958454e-07, + "loss": 0.1059, + "step": 10883 + }, + { + "epoch": 2.4764505119453926, + "grad_norm": 2.9812241962005013, + "learning_rate": 6.343368663356419e-07, + "loss": 0.1676, + "step": 10884 + }, + { + "epoch": 2.4766780432309443, + "grad_norm": 1.7392568804038129, + "learning_rate": 6.3424752488461e-07, + "loss": 0.1147, + "step": 10885 + }, + { + "epoch": 2.476905574516496, + "grad_norm": 1.7778725744664248, + "learning_rate": 6.341581832445757e-07, + "loss": 0.0638, + "step": 10886 + }, + { + "epoch": 2.477133105802048, + "grad_norm": 2.1746736380009914, + "learning_rate": 6.340688414173647e-07, + "loss": 0.0356, + "step": 10887 + }, + { + "epoch": 2.4773606370875996, + "grad_norm": 1.75293729766791, + "learning_rate": 6.339794994048035e-07, + "loss": 0.0609, + "step": 10888 + }, + { + "epoch": 2.4775881683731513, + "grad_norm": 2.3019638704234717, + "learning_rate": 6.338901572087177e-07, + "loss": 0.0762, + "step": 10889 + }, + { + "epoch": 2.477815699658703, + "grad_norm": 1.7872490771912772, + "learning_rate": 6.338008148309329e-07, + "loss": 0.0384, + "step": 10890 + }, + { + "epoch": 2.478043230944255, + "grad_norm": 1.5527157273794523, + "learning_rate": 6.337114722732761e-07, + "loss": 0.0206, + "step": 10891 + }, + { + "epoch": 2.4782707622298066, + "grad_norm": 2.471283722500934, + "learning_rate": 6.336221295375726e-07, + "loss": 0.0387, + "step": 10892 + }, + { + "epoch": 2.4784982935153583, + "grad_norm": 2.1810401201668084, + "learning_rate": 6.335327866256486e-07, + "loss": 0.0407, + "step": 10893 + }, + { + "epoch": 2.47872582480091, + "grad_norm": 1.2269915474957105, + "learning_rate": 6.334434435393298e-07, + "loss": 0.0837, + "step": 10894 + }, + { + "epoch": 2.478953356086462, + "grad_norm": 1.5026685542699592, + "learning_rate": 6.333541002804429e-07, + "loss": 0.0519, + "step": 10895 + }, + { + "epoch": 2.4791808873720136, + "grad_norm": 1.4966079948366215, + "learning_rate": 6.332647568508132e-07, + "loss": 0.0354, + "step": 10896 + }, + { + "epoch": 2.4794084186575653, + "grad_norm": 1.3340971483650939, + "learning_rate": 6.33175413252267e-07, + "loss": 0.0632, + "step": 10897 + }, + { + "epoch": 2.479635949943117, + "grad_norm": 1.364128001145437, + "learning_rate": 6.330860694866305e-07, + "loss": 0.0884, + "step": 10898 + }, + { + "epoch": 2.479863481228669, + "grad_norm": 1.4279572290887816, + "learning_rate": 6.329967255557294e-07, + "loss": 0.0211, + "step": 10899 + }, + { + "epoch": 2.4800910125142206, + "grad_norm": 1.1514963524259667, + "learning_rate": 6.329073814613899e-07, + "loss": 0.0317, + "step": 10900 + }, + { + "epoch": 2.4803185437997723, + "grad_norm": 1.2091953209581847, + "learning_rate": 6.328180372054382e-07, + "loss": 0.0967, + "step": 10901 + }, + { + "epoch": 2.480546075085324, + "grad_norm": 1.7577632697097967, + "learning_rate": 6.327286927897e-07, + "loss": 0.0314, + "step": 10902 + }, + { + "epoch": 2.480773606370876, + "grad_norm": 1.621642503371384, + "learning_rate": 6.326393482160013e-07, + "loss": 0.0417, + "step": 10903 + }, + { + "epoch": 2.4810011376564276, + "grad_norm": 1.2998934288929127, + "learning_rate": 6.325500034861684e-07, + "loss": 0.023, + "step": 10904 + }, + { + "epoch": 2.4812286689419794, + "grad_norm": 0.950097751804634, + "learning_rate": 6.324606586020274e-07, + "loss": 0.0219, + "step": 10905 + }, + { + "epoch": 2.481456200227531, + "grad_norm": 1.2672036485273066, + "learning_rate": 6.323713135654041e-07, + "loss": 0.0761, + "step": 10906 + }, + { + "epoch": 2.481683731513083, + "grad_norm": 1.6457683504653216, + "learning_rate": 6.322819683781248e-07, + "loss": 0.0964, + "step": 10907 + }, + { + "epoch": 2.4819112627986346, + "grad_norm": 1.1300112488301512, + "learning_rate": 6.321926230420153e-07, + "loss": 0.063, + "step": 10908 + }, + { + "epoch": 2.482138794084187, + "grad_norm": 1.7676597566346592, + "learning_rate": 6.321032775589018e-07, + "loss": 0.1228, + "step": 10909 + }, + { + "epoch": 2.482366325369738, + "grad_norm": 1.3995083312884067, + "learning_rate": 6.3201393193061e-07, + "loss": 0.0316, + "step": 10910 + }, + { + "epoch": 2.4825938566552903, + "grad_norm": 1.6485337964218663, + "learning_rate": 6.319245861589666e-07, + "loss": 0.0685, + "step": 10911 + }, + { + "epoch": 2.4828213879408416, + "grad_norm": 1.8539161700493796, + "learning_rate": 6.318352402457973e-07, + "loss": 0.0371, + "step": 10912 + }, + { + "epoch": 2.483048919226394, + "grad_norm": 2.07337472637515, + "learning_rate": 6.317458941929281e-07, + "loss": 0.0618, + "step": 10913 + }, + { + "epoch": 2.4832764505119456, + "grad_norm": 2.0706240220245706, + "learning_rate": 6.316565480021854e-07, + "loss": 0.0853, + "step": 10914 + }, + { + "epoch": 2.4835039817974973, + "grad_norm": 1.5878454449522577, + "learning_rate": 6.315672016753949e-07, + "loss": 0.0775, + "step": 10915 + }, + { + "epoch": 2.483731513083049, + "grad_norm": 1.284417858069104, + "learning_rate": 6.314778552143827e-07, + "loss": 0.0382, + "step": 10916 + }, + { + "epoch": 2.483959044368601, + "grad_norm": 1.4501725974250523, + "learning_rate": 6.31388508620975e-07, + "loss": 0.0689, + "step": 10917 + }, + { + "epoch": 2.4841865756541526, + "grad_norm": 1.8863046952116718, + "learning_rate": 6.312991618969981e-07, + "loss": 0.051, + "step": 10918 + }, + { + "epoch": 2.4844141069397043, + "grad_norm": 2.1319011616217893, + "learning_rate": 6.312098150442777e-07, + "loss": 0.0827, + "step": 10919 + }, + { + "epoch": 2.484641638225256, + "grad_norm": 1.5243712806420018, + "learning_rate": 6.311204680646403e-07, + "loss": 0.0254, + "step": 10920 + }, + { + "epoch": 2.484869169510808, + "grad_norm": 1.6283093342850699, + "learning_rate": 6.310311209599115e-07, + "loss": 0.0401, + "step": 10921 + }, + { + "epoch": 2.4850967007963596, + "grad_norm": 1.5825092963348164, + "learning_rate": 6.309417737319178e-07, + "loss": 0.0759, + "step": 10922 + }, + { + "epoch": 2.4853242320819113, + "grad_norm": 1.3152219871589137, + "learning_rate": 6.30852426382485e-07, + "loss": 0.0655, + "step": 10923 + }, + { + "epoch": 2.485551763367463, + "grad_norm": 1.4591531408229925, + "learning_rate": 6.307630789134393e-07, + "loss": 0.0527, + "step": 10924 + }, + { + "epoch": 2.485779294653015, + "grad_norm": 1.8254513711273366, + "learning_rate": 6.306737313266069e-07, + "loss": 0.0831, + "step": 10925 + }, + { + "epoch": 2.4860068259385666, + "grad_norm": 1.552719127965666, + "learning_rate": 6.305843836238139e-07, + "loss": 0.0803, + "step": 10926 + }, + { + "epoch": 2.4862343572241183, + "grad_norm": 1.5200975335161506, + "learning_rate": 6.304950358068862e-07, + "loss": 0.0346, + "step": 10927 + }, + { + "epoch": 2.48646188850967, + "grad_norm": 1.589587217277866, + "learning_rate": 6.3040568787765e-07, + "loss": 0.1075, + "step": 10928 + }, + { + "epoch": 2.486689419795222, + "grad_norm": 1.5307377554485049, + "learning_rate": 6.303163398379316e-07, + "loss": 0.1117, + "step": 10929 + }, + { + "epoch": 2.4869169510807736, + "grad_norm": 1.829675622819192, + "learning_rate": 6.302269916895566e-07, + "loss": 0.0528, + "step": 10930 + }, + { + "epoch": 2.4871444823663253, + "grad_norm": 1.0612215711711004, + "learning_rate": 6.301376434343517e-07, + "loss": 0.0203, + "step": 10931 + }, + { + "epoch": 2.487372013651877, + "grad_norm": 1.7034236852627376, + "learning_rate": 6.300482950741431e-07, + "loss": 0.0446, + "step": 10932 + }, + { + "epoch": 2.487599544937429, + "grad_norm": 0.8577980679304394, + "learning_rate": 6.299589466107561e-07, + "loss": 0.0459, + "step": 10933 + }, + { + "epoch": 2.4878270762229806, + "grad_norm": 1.3297674705334859, + "learning_rate": 6.298695980460174e-07, + "loss": 0.0834, + "step": 10934 + }, + { + "epoch": 2.4880546075085324, + "grad_norm": 1.0911005408658943, + "learning_rate": 6.297802493817533e-07, + "loss": 0.0273, + "step": 10935 + }, + { + "epoch": 2.488282138794084, + "grad_norm": 1.5025784372982494, + "learning_rate": 6.296909006197895e-07, + "loss": 0.0378, + "step": 10936 + }, + { + "epoch": 2.488509670079636, + "grad_norm": 1.3372925029531704, + "learning_rate": 6.296015517619522e-07, + "loss": 0.0306, + "step": 10937 + }, + { + "epoch": 2.4887372013651876, + "grad_norm": 1.2720465753095207, + "learning_rate": 6.295122028100677e-07, + "loss": 0.0563, + "step": 10938 + }, + { + "epoch": 2.4889647326507394, + "grad_norm": 1.1184442752655528, + "learning_rate": 6.294228537659622e-07, + "loss": 0.075, + "step": 10939 + }, + { + "epoch": 2.489192263936291, + "grad_norm": 2.7369299494537453, + "learning_rate": 6.293335046314612e-07, + "loss": 0.0516, + "step": 10940 + }, + { + "epoch": 2.489419795221843, + "grad_norm": 1.3520238857070834, + "learning_rate": 6.292441554083917e-07, + "loss": 0.031, + "step": 10941 + }, + { + "epoch": 2.4896473265073946, + "grad_norm": 1.3526759700850182, + "learning_rate": 6.291548060985793e-07, + "loss": 0.0347, + "step": 10942 + }, + { + "epoch": 2.4898748577929464, + "grad_norm": 1.9408977059615136, + "learning_rate": 6.290654567038504e-07, + "loss": 0.0943, + "step": 10943 + }, + { + "epoch": 2.490102389078498, + "grad_norm": 3.2539159300907556, + "learning_rate": 6.289761072260307e-07, + "loss": 0.1325, + "step": 10944 + }, + { + "epoch": 2.49032992036405, + "grad_norm": 1.813088070705479, + "learning_rate": 6.28886757666947e-07, + "loss": 0.0787, + "step": 10945 + }, + { + "epoch": 2.4905574516496016, + "grad_norm": 1.3320876525299923, + "learning_rate": 6.287974080284251e-07, + "loss": 0.0338, + "step": 10946 + }, + { + "epoch": 2.4907849829351534, + "grad_norm": 2.2645943301286873, + "learning_rate": 6.287080583122908e-07, + "loss": 0.0898, + "step": 10947 + }, + { + "epoch": 2.4910125142207056, + "grad_norm": 1.5394760949896429, + "learning_rate": 6.286187085203707e-07, + "loss": 0.0369, + "step": 10948 + }, + { + "epoch": 2.491240045506257, + "grad_norm": 1.1452440568877147, + "learning_rate": 6.28529358654491e-07, + "loss": 0.0302, + "step": 10949 + }, + { + "epoch": 2.491467576791809, + "grad_norm": 2.0395924077660297, + "learning_rate": 6.284400087164776e-07, + "loss": 0.0699, + "step": 10950 + }, + { + "epoch": 2.4916951080773604, + "grad_norm": 1.262492283073084, + "learning_rate": 6.283506587081568e-07, + "loss": 0.0341, + "step": 10951 + }, + { + "epoch": 2.4919226393629126, + "grad_norm": 2.1215436901960634, + "learning_rate": 6.282613086313546e-07, + "loss": 0.0458, + "step": 10952 + }, + { + "epoch": 2.4921501706484643, + "grad_norm": 0.7259583506962458, + "learning_rate": 6.28171958487897e-07, + "loss": 0.0074, + "step": 10953 + }, + { + "epoch": 2.492377701934016, + "grad_norm": 1.9498742458947604, + "learning_rate": 6.280826082796104e-07, + "loss": 0.0792, + "step": 10954 + }, + { + "epoch": 2.492605233219568, + "grad_norm": 1.221683181960274, + "learning_rate": 6.279932580083212e-07, + "loss": 0.0995, + "step": 10955 + }, + { + "epoch": 2.4928327645051196, + "grad_norm": 1.5853967716985577, + "learning_rate": 6.279039076758551e-07, + "loss": 0.0601, + "step": 10956 + }, + { + "epoch": 2.4930602957906713, + "grad_norm": 1.567758312242494, + "learning_rate": 6.278145572840385e-07, + "loss": 0.0474, + "step": 10957 + }, + { + "epoch": 2.493287827076223, + "grad_norm": 1.6941242657058169, + "learning_rate": 6.277252068346977e-07, + "loss": 0.028, + "step": 10958 + }, + { + "epoch": 2.493515358361775, + "grad_norm": 1.784161313779922, + "learning_rate": 6.276358563296585e-07, + "loss": 0.0373, + "step": 10959 + }, + { + "epoch": 2.4937428896473266, + "grad_norm": 1.8811344626005924, + "learning_rate": 6.275465057707471e-07, + "loss": 0.0683, + "step": 10960 + }, + { + "epoch": 2.4939704209328784, + "grad_norm": 2.1599851143667963, + "learning_rate": 6.274571551597899e-07, + "loss": 0.1277, + "step": 10961 + }, + { + "epoch": 2.49419795221843, + "grad_norm": 2.0660868345046706, + "learning_rate": 6.273678044986129e-07, + "loss": 0.0356, + "step": 10962 + }, + { + "epoch": 2.494425483503982, + "grad_norm": 2.311701878971303, + "learning_rate": 6.272784537890425e-07, + "loss": 0.1128, + "step": 10963 + }, + { + "epoch": 2.4946530147895336, + "grad_norm": 1.487582948660883, + "learning_rate": 6.271891030329046e-07, + "loss": 0.0848, + "step": 10964 + }, + { + "epoch": 2.4948805460750854, + "grad_norm": 1.5999148595393617, + "learning_rate": 6.270997522320254e-07, + "loss": 0.0484, + "step": 10965 + }, + { + "epoch": 2.495108077360637, + "grad_norm": 1.436429889972859, + "learning_rate": 6.270104013882311e-07, + "loss": 0.1179, + "step": 10966 + }, + { + "epoch": 2.495335608646189, + "grad_norm": 0.8154230969363158, + "learning_rate": 6.269210505033476e-07, + "loss": 0.0568, + "step": 10967 + }, + { + "epoch": 2.4955631399317406, + "grad_norm": 1.7445634835749073, + "learning_rate": 6.268316995792017e-07, + "loss": 0.053, + "step": 10968 + }, + { + "epoch": 2.4957906712172924, + "grad_norm": 2.3433361379413835, + "learning_rate": 6.267423486176191e-07, + "loss": 0.0349, + "step": 10969 + }, + { + "epoch": 2.496018202502844, + "grad_norm": 1.1080018783863554, + "learning_rate": 6.266529976204263e-07, + "loss": 0.0295, + "step": 10970 + }, + { + "epoch": 2.496245733788396, + "grad_norm": 1.093452768701886, + "learning_rate": 6.26563646589449e-07, + "loss": 0.0196, + "step": 10971 + }, + { + "epoch": 2.4964732650739476, + "grad_norm": 1.0864518857444965, + "learning_rate": 6.264742955265138e-07, + "loss": 0.0117, + "step": 10972 + }, + { + "epoch": 2.4967007963594994, + "grad_norm": 1.1478405442019546, + "learning_rate": 6.263849444334464e-07, + "loss": 0.0523, + "step": 10973 + }, + { + "epoch": 2.496928327645051, + "grad_norm": 2.0314005977256975, + "learning_rate": 6.262955933120735e-07, + "loss": 0.1608, + "step": 10974 + }, + { + "epoch": 2.497155858930603, + "grad_norm": 0.9983054994281034, + "learning_rate": 6.262062421642211e-07, + "loss": 0.0211, + "step": 10975 + }, + { + "epoch": 2.4973833902161546, + "grad_norm": 1.2974449810419029, + "learning_rate": 6.261168909917154e-07, + "loss": 0.137, + "step": 10976 + }, + { + "epoch": 2.4976109215017064, + "grad_norm": 1.2650681890286097, + "learning_rate": 6.26027539796382e-07, + "loss": 0.0311, + "step": 10977 + }, + { + "epoch": 2.497838452787258, + "grad_norm": 1.1881429020101082, + "learning_rate": 6.259381885800481e-07, + "loss": 0.0151, + "step": 10978 + }, + { + "epoch": 2.49806598407281, + "grad_norm": 1.4571595873025454, + "learning_rate": 6.258488373445391e-07, + "loss": 0.1114, + "step": 10979 + }, + { + "epoch": 2.4982935153583616, + "grad_norm": 1.4081994726939846, + "learning_rate": 6.257594860916815e-07, + "loss": 0.0828, + "step": 10980 + }, + { + "epoch": 2.4985210466439134, + "grad_norm": 1.3771773689730369, + "learning_rate": 6.256701348233012e-07, + "loss": 0.0478, + "step": 10981 + }, + { + "epoch": 2.498748577929465, + "grad_norm": 2.5167708605910204, + "learning_rate": 6.255807835412248e-07, + "loss": 0.0338, + "step": 10982 + }, + { + "epoch": 2.498976109215017, + "grad_norm": 1.389196305428987, + "learning_rate": 6.254914322472783e-07, + "loss": 0.0865, + "step": 10983 + }, + { + "epoch": 2.4992036405005686, + "grad_norm": 2.1680229620137004, + "learning_rate": 6.254020809432876e-07, + "loss": 0.044, + "step": 10984 + }, + { + "epoch": 2.4994311717861204, + "grad_norm": 2.0580800123487752, + "learning_rate": 6.253127296310791e-07, + "loss": 0.0313, + "step": 10985 + }, + { + "epoch": 2.499658703071672, + "grad_norm": 1.7368747569550809, + "learning_rate": 6.252233783124792e-07, + "loss": 0.0316, + "step": 10986 + }, + { + "epoch": 2.4998862343572243, + "grad_norm": 1.6387854826028714, + "learning_rate": 6.251340269893138e-07, + "loss": 0.0608, + "step": 10987 + }, + { + "epoch": 2.5001137656427757, + "grad_norm": 1.9209892270794555, + "learning_rate": 6.250446756634089e-07, + "loss": 0.0994, + "step": 10988 + }, + { + "epoch": 2.500341296928328, + "grad_norm": 2.557587086759991, + "learning_rate": 6.249553243365913e-07, + "loss": 0.0674, + "step": 10989 + }, + { + "epoch": 2.500568828213879, + "grad_norm": 1.1622729488191976, + "learning_rate": 6.248659730106863e-07, + "loss": 0.0382, + "step": 10990 + }, + { + "epoch": 2.5007963594994314, + "grad_norm": 1.1474335317567095, + "learning_rate": 6.24776621687521e-07, + "loss": 0.0347, + "step": 10991 + }, + { + "epoch": 2.5010238907849827, + "grad_norm": 1.493008113130724, + "learning_rate": 6.246872703689212e-07, + "loss": 0.0793, + "step": 10992 + }, + { + "epoch": 2.501251422070535, + "grad_norm": 1.3947752174446955, + "learning_rate": 6.245979190567126e-07, + "loss": 0.0981, + "step": 10993 + }, + { + "epoch": 2.5014789533560866, + "grad_norm": 1.9942647265938565, + "learning_rate": 6.245085677527219e-07, + "loss": 0.1302, + "step": 10994 + }, + { + "epoch": 2.5017064846416384, + "grad_norm": 1.4226297792199358, + "learning_rate": 6.244192164587753e-07, + "loss": 0.0287, + "step": 10995 + }, + { + "epoch": 2.50193401592719, + "grad_norm": 0.9751331692519019, + "learning_rate": 6.243298651766989e-07, + "loss": 0.0485, + "step": 10996 + }, + { + "epoch": 2.502161547212742, + "grad_norm": 2.659026624075969, + "learning_rate": 6.242405139083186e-07, + "loss": 0.1096, + "step": 10997 + }, + { + "epoch": 2.5023890784982936, + "grad_norm": 1.7746402395082745, + "learning_rate": 6.241511626554611e-07, + "loss": 0.0366, + "step": 10998 + }, + { + "epoch": 2.5026166097838454, + "grad_norm": 1.286081332148587, + "learning_rate": 6.240618114199522e-07, + "loss": 0.0455, + "step": 10999 + }, + { + "epoch": 2.502844141069397, + "grad_norm": 0.8516668622858481, + "learning_rate": 6.239724602036181e-07, + "loss": 0.0411, + "step": 11000 + }, + { + "epoch": 2.503071672354949, + "grad_norm": 1.2262068528065933, + "learning_rate": 6.23883109008285e-07, + "loss": 0.0322, + "step": 11001 + }, + { + "epoch": 2.5032992036405006, + "grad_norm": 2.1785208442001096, + "learning_rate": 6.23793757835779e-07, + "loss": 0.0402, + "step": 11002 + }, + { + "epoch": 2.5035267349260524, + "grad_norm": 2.379186611403098, + "learning_rate": 6.237044066879268e-07, + "loss": 0.0661, + "step": 11003 + }, + { + "epoch": 2.503754266211604, + "grad_norm": 2.0080939224737073, + "learning_rate": 6.236150555665537e-07, + "loss": 0.0328, + "step": 11004 + }, + { + "epoch": 2.503981797497156, + "grad_norm": 0.9759513029314878, + "learning_rate": 6.235257044734864e-07, + "loss": 0.0316, + "step": 11005 + }, + { + "epoch": 2.5042093287827076, + "grad_norm": 1.534420311364177, + "learning_rate": 6.234363534105513e-07, + "loss": 0.0572, + "step": 11006 + }, + { + "epoch": 2.5044368600682594, + "grad_norm": 1.2612808303628475, + "learning_rate": 6.23347002379574e-07, + "loss": 0.0407, + "step": 11007 + }, + { + "epoch": 2.504664391353811, + "grad_norm": 0.8566965253614552, + "learning_rate": 6.23257651382381e-07, + "loss": 0.0561, + "step": 11008 + }, + { + "epoch": 2.504891922639363, + "grad_norm": 2.268877775388729, + "learning_rate": 6.231683004207984e-07, + "loss": 0.1698, + "step": 11009 + }, + { + "epoch": 2.5051194539249146, + "grad_norm": 2.887640350074492, + "learning_rate": 6.230789494966525e-07, + "loss": 0.109, + "step": 11010 + }, + { + "epoch": 2.5053469852104664, + "grad_norm": 1.7785382762183093, + "learning_rate": 6.229895986117693e-07, + "loss": 0.0309, + "step": 11011 + }, + { + "epoch": 2.505574516496018, + "grad_norm": 2.0159140124925257, + "learning_rate": 6.229002477679749e-07, + "loss": 0.1146, + "step": 11012 + }, + { + "epoch": 2.50580204778157, + "grad_norm": 1.9841187845844523, + "learning_rate": 6.228108969670959e-07, + "loss": 0.0599, + "step": 11013 + }, + { + "epoch": 2.5060295790671216, + "grad_norm": 1.5881943757830714, + "learning_rate": 6.227215462109577e-07, + "loss": 0.0476, + "step": 11014 + }, + { + "epoch": 2.5062571103526734, + "grad_norm": 1.6265444658160035, + "learning_rate": 6.226321955013872e-07, + "loss": 0.0401, + "step": 11015 + }, + { + "epoch": 2.506484641638225, + "grad_norm": 1.2011087711776314, + "learning_rate": 6.225428448402102e-07, + "loss": 0.0126, + "step": 11016 + }, + { + "epoch": 2.506712172923777, + "grad_norm": 1.6469431396233152, + "learning_rate": 6.224534942292531e-07, + "loss": 0.0204, + "step": 11017 + }, + { + "epoch": 2.5069397042093287, + "grad_norm": 1.6526004117766206, + "learning_rate": 6.223641436703418e-07, + "loss": 0.1061, + "step": 11018 + }, + { + "epoch": 2.5071672354948804, + "grad_norm": 1.8611610896655753, + "learning_rate": 6.222747931653025e-07, + "loss": 0.0516, + "step": 11019 + }, + { + "epoch": 2.507394766780432, + "grad_norm": 1.422459754183788, + "learning_rate": 6.221854427159617e-07, + "loss": 0.1223, + "step": 11020 + }, + { + "epoch": 2.507622298065984, + "grad_norm": 1.5164008463616914, + "learning_rate": 6.220960923241448e-07, + "loss": 0.0451, + "step": 11021 + }, + { + "epoch": 2.507849829351536, + "grad_norm": 2.9298653444322147, + "learning_rate": 6.22006741991679e-07, + "loss": 0.0487, + "step": 11022 + }, + { + "epoch": 2.5080773606370874, + "grad_norm": 1.6393400485845175, + "learning_rate": 6.219173917203896e-07, + "loss": 0.0535, + "step": 11023 + }, + { + "epoch": 2.5083048919226396, + "grad_norm": 2.476678108434716, + "learning_rate": 6.218280415121032e-07, + "loss": 0.042, + "step": 11024 + }, + { + "epoch": 2.508532423208191, + "grad_norm": 1.7759330588507165, + "learning_rate": 6.217386913686459e-07, + "loss": 0.0368, + "step": 11025 + }, + { + "epoch": 2.508759954493743, + "grad_norm": 2.0112445175558067, + "learning_rate": 6.216493412918436e-07, + "loss": 0.117, + "step": 11026 + }, + { + "epoch": 2.5089874857792944, + "grad_norm": 2.0848083537206223, + "learning_rate": 6.215599912835226e-07, + "loss": 0.0397, + "step": 11027 + }, + { + "epoch": 2.5092150170648466, + "grad_norm": 1.2578285990640168, + "learning_rate": 6.214706413455091e-07, + "loss": 0.0435, + "step": 11028 + }, + { + "epoch": 2.509442548350398, + "grad_norm": 1.7198229490467627, + "learning_rate": 6.213812914796294e-07, + "loss": 0.0328, + "step": 11029 + }, + { + "epoch": 2.50967007963595, + "grad_norm": 1.9867695825918112, + "learning_rate": 6.212919416877094e-07, + "loss": 0.0574, + "step": 11030 + }, + { + "epoch": 2.5098976109215014, + "grad_norm": 2.7105225526160477, + "learning_rate": 6.212025919715751e-07, + "loss": 0.1359, + "step": 11031 + }, + { + "epoch": 2.5101251422070536, + "grad_norm": 2.342574023145511, + "learning_rate": 6.211132423330533e-07, + "loss": 0.0358, + "step": 11032 + }, + { + "epoch": 2.5103526734926054, + "grad_norm": 1.412072555498323, + "learning_rate": 6.210238927739693e-07, + "loss": 0.0506, + "step": 11033 + }, + { + "epoch": 2.510580204778157, + "grad_norm": 1.7655034110485555, + "learning_rate": 6.209345432961498e-07, + "loss": 0.0708, + "step": 11034 + }, + { + "epoch": 2.510807736063709, + "grad_norm": 2.1054380939463195, + "learning_rate": 6.208451939014207e-07, + "loss": 0.0433, + "step": 11035 + }, + { + "epoch": 2.5110352673492606, + "grad_norm": 2.8617859785515947, + "learning_rate": 6.207558445916085e-07, + "loss": 0.1186, + "step": 11036 + }, + { + "epoch": 2.5112627986348124, + "grad_norm": 2.1323967775729966, + "learning_rate": 6.206664953685389e-07, + "loss": 0.0782, + "step": 11037 + }, + { + "epoch": 2.511490329920364, + "grad_norm": 1.5742435506765042, + "learning_rate": 6.205771462340381e-07, + "loss": 0.0468, + "step": 11038 + }, + { + "epoch": 2.511717861205916, + "grad_norm": 1.3050726174092304, + "learning_rate": 6.204877971899325e-07, + "loss": 0.0569, + "step": 11039 + }, + { + "epoch": 2.5119453924914676, + "grad_norm": 1.7443565804290286, + "learning_rate": 6.203984482380479e-07, + "loss": 0.0798, + "step": 11040 + }, + { + "epoch": 2.5121729237770194, + "grad_norm": 2.5078676652413234, + "learning_rate": 6.203090993802107e-07, + "loss": 0.0677, + "step": 11041 + }, + { + "epoch": 2.512400455062571, + "grad_norm": 1.9131801027057873, + "learning_rate": 6.202197506182468e-07, + "loss": 0.073, + "step": 11042 + }, + { + "epoch": 2.512627986348123, + "grad_norm": 1.884257864387025, + "learning_rate": 6.201304019539827e-07, + "loss": 0.0436, + "step": 11043 + }, + { + "epoch": 2.5128555176336747, + "grad_norm": 1.7798345635565676, + "learning_rate": 6.200410533892441e-07, + "loss": 0.0258, + "step": 11044 + }, + { + "epoch": 2.5130830489192264, + "grad_norm": 2.4024235445276743, + "learning_rate": 6.199517049258571e-07, + "loss": 0.0535, + "step": 11045 + }, + { + "epoch": 2.513310580204778, + "grad_norm": 2.033907517798713, + "learning_rate": 6.198623565656484e-07, + "loss": 0.0354, + "step": 11046 + }, + { + "epoch": 2.51353811149033, + "grad_norm": 1.6898910861106502, + "learning_rate": 6.197730083104433e-07, + "loss": 0.0477, + "step": 11047 + }, + { + "epoch": 2.5137656427758817, + "grad_norm": 1.1103854740403931, + "learning_rate": 6.196836601620686e-07, + "loss": 0.0162, + "step": 11048 + }, + { + "epoch": 2.5139931740614334, + "grad_norm": 1.143845607126968, + "learning_rate": 6.195943121223503e-07, + "loss": 0.0227, + "step": 11049 + }, + { + "epoch": 2.514220705346985, + "grad_norm": 1.6919805089429434, + "learning_rate": 6.19504964193114e-07, + "loss": 0.081, + "step": 11050 + }, + { + "epoch": 2.514448236632537, + "grad_norm": 2.336081671533437, + "learning_rate": 6.194156163761863e-07, + "loss": 0.0382, + "step": 11051 + }, + { + "epoch": 2.5146757679180887, + "grad_norm": 1.589334318516393, + "learning_rate": 6.193262686733931e-07, + "loss": 0.1608, + "step": 11052 + }, + { + "epoch": 2.5149032992036404, + "grad_norm": 1.9718749666129884, + "learning_rate": 6.192369210865609e-07, + "loss": 0.018, + "step": 11053 + }, + { + "epoch": 2.515130830489192, + "grad_norm": 0.9533565122642095, + "learning_rate": 6.191475736175152e-07, + "loss": 0.0134, + "step": 11054 + }, + { + "epoch": 2.515358361774744, + "grad_norm": 1.2672994305420475, + "learning_rate": 6.190582262680824e-07, + "loss": 0.0425, + "step": 11055 + }, + { + "epoch": 2.5155858930602957, + "grad_norm": 1.0244631647287852, + "learning_rate": 6.189688790400888e-07, + "loss": 0.0155, + "step": 11056 + }, + { + "epoch": 2.5158134243458474, + "grad_norm": 2.049403246130174, + "learning_rate": 6.188795319353599e-07, + "loss": 0.0472, + "step": 11057 + }, + { + "epoch": 2.516040955631399, + "grad_norm": 1.1568270284298823, + "learning_rate": 6.187901849557224e-07, + "loss": 0.011, + "step": 11058 + }, + { + "epoch": 2.516268486916951, + "grad_norm": 2.041916449530491, + "learning_rate": 6.187008381030019e-07, + "loss": 0.075, + "step": 11059 + }, + { + "epoch": 2.5164960182025027, + "grad_norm": 0.9417216816686345, + "learning_rate": 6.186114913790251e-07, + "loss": 0.0699, + "step": 11060 + }, + { + "epoch": 2.516723549488055, + "grad_norm": 1.7052735078244157, + "learning_rate": 6.185221447856175e-07, + "loss": 0.0391, + "step": 11061 + }, + { + "epoch": 2.516951080773606, + "grad_norm": 2.6500799341978363, + "learning_rate": 6.184327983246054e-07, + "loss": 0.0442, + "step": 11062 + }, + { + "epoch": 2.5171786120591584, + "grad_norm": 1.5574220750337953, + "learning_rate": 6.18343451997815e-07, + "loss": 0.0255, + "step": 11063 + }, + { + "epoch": 2.5174061433447097, + "grad_norm": 0.9291304127062001, + "learning_rate": 6.182541058070721e-07, + "loss": 0.0796, + "step": 11064 + }, + { + "epoch": 2.517633674630262, + "grad_norm": 1.598773313826786, + "learning_rate": 6.181647597542029e-07, + "loss": 0.0455, + "step": 11065 + }, + { + "epoch": 2.517861205915813, + "grad_norm": 1.5361754454204313, + "learning_rate": 6.180754138410334e-07, + "loss": 0.0425, + "step": 11066 + }, + { + "epoch": 2.5180887372013654, + "grad_norm": 1.6102930226415857, + "learning_rate": 6.179860680693902e-07, + "loss": 0.1083, + "step": 11067 + }, + { + "epoch": 2.5183162684869167, + "grad_norm": 1.3999155840529558, + "learning_rate": 6.178967224410987e-07, + "loss": 0.032, + "step": 11068 + }, + { + "epoch": 2.518543799772469, + "grad_norm": 2.057394070295082, + "learning_rate": 6.17807376957985e-07, + "loss": 0.0937, + "step": 11069 + }, + { + "epoch": 2.51877133105802, + "grad_norm": 1.6919586826668034, + "learning_rate": 6.177180316218756e-07, + "loss": 0.0917, + "step": 11070 + }, + { + "epoch": 2.5189988623435724, + "grad_norm": 2.209998623716619, + "learning_rate": 6.17628686434596e-07, + "loss": 0.0893, + "step": 11071 + }, + { + "epoch": 2.519226393629124, + "grad_norm": 2.143952188375939, + "learning_rate": 6.175393413979728e-07, + "loss": 0.0489, + "step": 11072 + }, + { + "epoch": 2.519453924914676, + "grad_norm": 1.3893039312898101, + "learning_rate": 6.174499965138316e-07, + "loss": 0.0335, + "step": 11073 + }, + { + "epoch": 2.5196814562002277, + "grad_norm": 1.5638689869103837, + "learning_rate": 6.173606517839989e-07, + "loss": 0.0725, + "step": 11074 + }, + { + "epoch": 2.5199089874857794, + "grad_norm": 2.4618631078788065, + "learning_rate": 6.172713072103004e-07, + "loss": 0.0872, + "step": 11075 + }, + { + "epoch": 2.520136518771331, + "grad_norm": 1.4188905367615328, + "learning_rate": 6.17181962794562e-07, + "loss": 0.0431, + "step": 11076 + }, + { + "epoch": 2.520364050056883, + "grad_norm": 1.0415385112590472, + "learning_rate": 6.170926185386102e-07, + "loss": 0.0171, + "step": 11077 + }, + { + "epoch": 2.5205915813424347, + "grad_norm": 1.948232303654507, + "learning_rate": 6.170032744442706e-07, + "loss": 0.1255, + "step": 11078 + }, + { + "epoch": 2.5208191126279864, + "grad_norm": 0.7708876825095176, + "learning_rate": 6.169139305133697e-07, + "loss": 0.0683, + "step": 11079 + }, + { + "epoch": 2.521046643913538, + "grad_norm": 1.5834734224473694, + "learning_rate": 6.168245867477333e-07, + "loss": 0.0281, + "step": 11080 + }, + { + "epoch": 2.52127417519909, + "grad_norm": 2.088965532941998, + "learning_rate": 6.16735243149187e-07, + "loss": 0.0335, + "step": 11081 + }, + { + "epoch": 2.5215017064846417, + "grad_norm": 1.3640471096775282, + "learning_rate": 6.166458997195575e-07, + "loss": 0.1846, + "step": 11082 + }, + { + "epoch": 2.5217292377701934, + "grad_norm": 2.4088616521647324, + "learning_rate": 6.165565564606702e-07, + "loss": 0.0522, + "step": 11083 + }, + { + "epoch": 2.521956769055745, + "grad_norm": 2.207807190861441, + "learning_rate": 6.164672133743516e-07, + "loss": 0.0704, + "step": 11084 + }, + { + "epoch": 2.522184300341297, + "grad_norm": 1.6334687157525936, + "learning_rate": 6.163778704624275e-07, + "loss": 0.0884, + "step": 11085 + }, + { + "epoch": 2.5224118316268487, + "grad_norm": 1.1711488252341609, + "learning_rate": 6.162885277267241e-07, + "loss": 0.0123, + "step": 11086 + }, + { + "epoch": 2.5226393629124004, + "grad_norm": 1.4051759111261901, + "learning_rate": 6.161991851690672e-07, + "loss": 0.0564, + "step": 11087 + }, + { + "epoch": 2.522866894197952, + "grad_norm": 1.8550063678903088, + "learning_rate": 6.161098427912827e-07, + "loss": 0.1002, + "step": 11088 + }, + { + "epoch": 2.523094425483504, + "grad_norm": 2.5240498530430613, + "learning_rate": 6.160205005951969e-07, + "loss": 0.0563, + "step": 11089 + }, + { + "epoch": 2.5233219567690557, + "grad_norm": 1.6237095838440487, + "learning_rate": 6.159311585826353e-07, + "loss": 0.1429, + "step": 11090 + }, + { + "epoch": 2.5235494880546074, + "grad_norm": 2.0992388835763656, + "learning_rate": 6.158418167554245e-07, + "loss": 0.0927, + "step": 11091 + }, + { + "epoch": 2.523777019340159, + "grad_norm": 1.7069591270863522, + "learning_rate": 6.1575247511539e-07, + "loss": 0.0321, + "step": 11092 + }, + { + "epoch": 2.524004550625711, + "grad_norm": 1.3658787264675727, + "learning_rate": 6.156631336643582e-07, + "loss": 0.0368, + "step": 11093 + }, + { + "epoch": 2.5242320819112627, + "grad_norm": 1.57359365698972, + "learning_rate": 6.155737924041548e-07, + "loss": 0.0888, + "step": 11094 + }, + { + "epoch": 2.5244596131968144, + "grad_norm": 1.8229817642784567, + "learning_rate": 6.154844513366056e-07, + "loss": 0.0567, + "step": 11095 + }, + { + "epoch": 2.524687144482366, + "grad_norm": 1.252903518572975, + "learning_rate": 6.153951104635371e-07, + "loss": 0.066, + "step": 11096 + }, + { + "epoch": 2.524914675767918, + "grad_norm": 1.5752843940414576, + "learning_rate": 6.153057697867746e-07, + "loss": 0.0302, + "step": 11097 + }, + { + "epoch": 2.5251422070534697, + "grad_norm": 0.8352223227098137, + "learning_rate": 6.152164293081447e-07, + "loss": 0.027, + "step": 11098 + }, + { + "epoch": 2.5253697383390215, + "grad_norm": 1.2872165203116492, + "learning_rate": 6.151270890294731e-07, + "loss": 0.0266, + "step": 11099 + }, + { + "epoch": 2.5255972696245736, + "grad_norm": 1.6736483992310536, + "learning_rate": 6.150377489525855e-07, + "loss": 0.0843, + "step": 11100 + } + ], + "logging_steps": 1, + "max_steps": 21975, + "num_input_tokens_seen": 0, + "num_train_epochs": 5, + "save_steps": 1110, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 49274590224384.0, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}