{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.757679180887372, "eval_steps": 500, "global_step": 3330, "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 } ], "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": 14805564383232.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }