{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.2627986348122868, "eval_steps": 500, "global_step": 5550, "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 } ], "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": 24592345989120.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }