{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.2627986348122868, "eval_steps": 500, "global_step": 5550, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00022753128555176336, "grad_norm": 12.279616595766896, "learning_rate": 1.25e-06, "loss": 0.1112, "step": 1 }, { "epoch": 0.0004550625711035267, "grad_norm": 7.2629005990045785, "learning_rate": 1.2499999936130725e-06, "loss": 0.1394, "step": 2 }, { "epoch": 0.0006825938566552901, "grad_norm": 9.061755966937882, "learning_rate": 1.2499999744522896e-06, "loss": 0.266, "step": 3 }, { "epoch": 0.0009101251422070534, "grad_norm": 4.366742138714415, "learning_rate": 1.2499999425176518e-06, "loss": 0.1258, "step": 4 }, { "epoch": 0.0011376564277588168, "grad_norm": 8.875041328214682, "learning_rate": 1.2499998978091598e-06, "loss": 0.1749, "step": 5 }, { "epoch": 0.0013651877133105802, "grad_norm": 8.800851302952648, "learning_rate": 1.2499998403268147e-06, "loss": 0.2, "step": 6 }, { "epoch": 0.0015927189988623437, "grad_norm": 3.6087293622719416, "learning_rate": 1.2499997700706173e-06, "loss": 0.1398, "step": 7 }, { "epoch": 0.0018202502844141069, "grad_norm": 3.42318784466345, "learning_rate": 1.2499996870405692e-06, "loss": 0.1226, "step": 8 }, { "epoch": 0.0020477815699658703, "grad_norm": 10.818031536677639, "learning_rate": 1.2499995912366722e-06, "loss": 0.171, "step": 9 }, { "epoch": 0.0022753128555176336, "grad_norm": 4.021373220825307, "learning_rate": 1.2499994826589282e-06, "loss": 0.1326, "step": 10 }, { "epoch": 0.002502844141069397, "grad_norm": 23.4957115965293, "learning_rate": 1.2499993613073393e-06, "loss": 0.1235, "step": 11 }, { "epoch": 0.0027303754266211604, "grad_norm": 21.545758763913824, "learning_rate": 1.2499992271819083e-06, "loss": 0.0925, "step": 12 }, { "epoch": 0.0029579067121729237, "grad_norm": 3.510359233156002, "learning_rate": 1.2499990802826377e-06, "loss": 0.0985, "step": 13 }, { "epoch": 0.0031854379977246873, "grad_norm": 3.9794392758691632, "learning_rate": 1.2499989206095304e-06, "loss": 0.0906, "step": 14 }, { "epoch": 0.0034129692832764505, "grad_norm": 2.4759900062021503, "learning_rate": 1.2499987481625899e-06, "loss": 0.0746, "step": 15 }, { "epoch": 0.0036405005688282138, "grad_norm": 7.91228269690243, "learning_rate": 1.2499985629418195e-06, "loss": 0.1381, "step": 16 }, { "epoch": 0.0038680318543799774, "grad_norm": 3.050431958605509, "learning_rate": 1.2499983649472233e-06, "loss": 0.0509, "step": 17 }, { "epoch": 0.004095563139931741, "grad_norm": 9.537504830414553, "learning_rate": 1.249998154178805e-06, "loss": 0.0658, "step": 18 }, { "epoch": 0.004323094425483504, "grad_norm": 5.663160798723299, "learning_rate": 1.2499979306365692e-06, "loss": 0.097, "step": 19 }, { "epoch": 0.004550625711035267, "grad_norm": 13.411158998212743, "learning_rate": 1.2499976943205202e-06, "loss": 0.1, "step": 20 }, { "epoch": 0.00477815699658703, "grad_norm": 3.6621965264316194, "learning_rate": 1.249997445230663e-06, "loss": 0.0698, "step": 21 }, { "epoch": 0.005005688282138794, "grad_norm": 31.354306922977063, "learning_rate": 1.2499971833670026e-06, "loss": 0.1129, "step": 22 }, { "epoch": 0.005233219567690558, "grad_norm": 4.025175447301565, "learning_rate": 1.2499969087295443e-06, "loss": 0.1076, "step": 23 }, { "epoch": 0.005460750853242321, "grad_norm": 3.7989823695443863, "learning_rate": 1.249996621318294e-06, "loss": 0.0902, "step": 24 }, { "epoch": 0.005688282138794084, "grad_norm": 2.2185511399815545, "learning_rate": 1.2499963211332573e-06, "loss": 0.0398, "step": 25 }, { "epoch": 0.005915813424345847, "grad_norm": 3.775038573479949, "learning_rate": 1.2499960081744405e-06, "loss": 0.1145, "step": 26 }, { "epoch": 0.0061433447098976105, "grad_norm": 7.295114622894012, "learning_rate": 1.24999568244185e-06, "loss": 0.1379, "step": 27 }, { "epoch": 0.006370875995449375, "grad_norm": 21.290899699279613, "learning_rate": 1.249995343935492e-06, "loss": 0.0995, "step": 28 }, { "epoch": 0.006598407281001138, "grad_norm": 3.593142181899087, "learning_rate": 1.2499949926553743e-06, "loss": 0.0973, "step": 29 }, { "epoch": 0.006825938566552901, "grad_norm": 7.737004888367107, "learning_rate": 1.2499946286015032e-06, "loss": 0.09, "step": 30 }, { "epoch": 0.007053469852104664, "grad_norm": 2.955413576195516, "learning_rate": 1.2499942517738867e-06, "loss": 0.0625, "step": 31 }, { "epoch": 0.0072810011376564275, "grad_norm": 4.138230802608221, "learning_rate": 1.2499938621725322e-06, "loss": 0.0749, "step": 32 }, { "epoch": 0.007508532423208191, "grad_norm": 2.2958240484880115, "learning_rate": 1.2499934597974478e-06, "loss": 0.0986, "step": 33 }, { "epoch": 0.007736063708759955, "grad_norm": 3.9060226417706256, "learning_rate": 1.2499930446486416e-06, "loss": 0.1296, "step": 34 }, { "epoch": 0.007963594994311717, "grad_norm": 1.5544518249405173, "learning_rate": 1.2499926167261224e-06, "loss": 0.0598, "step": 35 }, { "epoch": 0.008191126279863481, "grad_norm": 5.2971839668544884, "learning_rate": 1.2499921760298987e-06, "loss": 0.1216, "step": 36 }, { "epoch": 0.008418657565415245, "grad_norm": 6.168021178604936, "learning_rate": 1.2499917225599796e-06, "loss": 0.1257, "step": 37 }, { "epoch": 0.008646188850967008, "grad_norm": 5.087771081836934, "learning_rate": 1.2499912563163742e-06, "loss": 0.0695, "step": 38 }, { "epoch": 0.008873720136518772, "grad_norm": 7.633032375201886, "learning_rate": 1.249990777299092e-06, "loss": 0.0522, "step": 39 }, { "epoch": 0.009101251422070534, "grad_norm": 2.9414497498297734, "learning_rate": 1.249990285508143e-06, "loss": 0.081, "step": 40 }, { "epoch": 0.009328782707622298, "grad_norm": 4.449234015706431, "learning_rate": 1.2499897809435374e-06, "loss": 0.1104, "step": 41 }, { "epoch": 0.00955631399317406, "grad_norm": 17.29783348718493, "learning_rate": 1.249989263605285e-06, "loss": 0.0952, "step": 42 }, { "epoch": 0.009783845278725825, "grad_norm": 4.357226856628825, "learning_rate": 1.249988733493397e-06, "loss": 0.0639, "step": 43 }, { "epoch": 0.010011376564277589, "grad_norm": 4.248351802911978, "learning_rate": 1.2499881906078836e-06, "loss": 0.0845, "step": 44 }, { "epoch": 0.010238907849829351, "grad_norm": 2.230593274135571, "learning_rate": 1.2499876349487564e-06, "loss": 0.1031, "step": 45 }, { "epoch": 0.010466439135381115, "grad_norm": 5.02333552530202, "learning_rate": 1.2499870665160262e-06, "loss": 0.0763, "step": 46 }, { "epoch": 0.010693970420932878, "grad_norm": 1.7336741239976143, "learning_rate": 1.2499864853097054e-06, "loss": 0.0495, "step": 47 }, { "epoch": 0.010921501706484642, "grad_norm": 4.9863935956267325, "learning_rate": 1.2499858913298053e-06, "loss": 0.1674, "step": 48 }, { "epoch": 0.011149032992036406, "grad_norm": 2.8583968641628967, "learning_rate": 1.249985284576338e-06, "loss": 0.0728, "step": 49 }, { "epoch": 0.011376564277588168, "grad_norm": 5.82977520269559, "learning_rate": 1.2499846650493164e-06, "loss": 0.1076, "step": 50 }, { "epoch": 0.011604095563139932, "grad_norm": 2.407346961463521, "learning_rate": 1.2499840327487528e-06, "loss": 0.0542, "step": 51 }, { "epoch": 0.011831626848691695, "grad_norm": 1.7581308929371289, "learning_rate": 1.24998338767466e-06, "loss": 0.0851, "step": 52 }, { "epoch": 0.012059158134243459, "grad_norm": 3.6069730541490315, "learning_rate": 1.2499827298270515e-06, "loss": 0.0771, "step": 53 }, { "epoch": 0.012286689419795221, "grad_norm": 2.6384830282527965, "learning_rate": 1.2499820592059405e-06, "loss": 0.1336, "step": 54 }, { "epoch": 0.012514220705346985, "grad_norm": 3.1162890015900793, "learning_rate": 1.2499813758113409e-06, "loss": 0.1058, "step": 55 }, { "epoch": 0.01274175199089875, "grad_norm": 3.5247536707308966, "learning_rate": 1.2499806796432665e-06, "loss": 0.0717, "step": 56 }, { "epoch": 0.012969283276450512, "grad_norm": 2.2371436604921584, "learning_rate": 1.2499799707017315e-06, "loss": 0.07, "step": 57 }, { "epoch": 0.013196814562002276, "grad_norm": 5.912440334168246, "learning_rate": 1.2499792489867508e-06, "loss": 0.1588, "step": 58 }, { "epoch": 0.013424345847554038, "grad_norm": 1.6415965725415542, "learning_rate": 1.2499785144983386e-06, "loss": 0.0587, "step": 59 }, { "epoch": 0.013651877133105802, "grad_norm": 3.921979088226587, "learning_rate": 1.24997776723651e-06, "loss": 0.1403, "step": 60 }, { "epoch": 0.013879408418657566, "grad_norm": 4.854246136713843, "learning_rate": 1.2499770072012809e-06, "loss": 0.0706, "step": 61 }, { "epoch": 0.014106939704209329, "grad_norm": 2.249851975045142, "learning_rate": 1.2499762343926661e-06, "loss": 0.0988, "step": 62 }, { "epoch": 0.014334470989761093, "grad_norm": 2.5833858090734956, "learning_rate": 1.2499754488106817e-06, "loss": 0.0618, "step": 63 }, { "epoch": 0.014562002275312855, "grad_norm": 3.523489637935598, "learning_rate": 1.2499746504553436e-06, "loss": 0.0788, "step": 64 }, { "epoch": 0.01478953356086462, "grad_norm": 1.4862582066571077, "learning_rate": 1.2499738393266684e-06, "loss": 0.0574, "step": 65 }, { "epoch": 0.015017064846416382, "grad_norm": 3.5354676602669506, "learning_rate": 1.2499730154246726e-06, "loss": 0.0742, "step": 66 }, { "epoch": 0.015244596131968146, "grad_norm": 4.702124240777191, "learning_rate": 1.2499721787493726e-06, "loss": 0.0994, "step": 67 }, { "epoch": 0.01547212741751991, "grad_norm": 2.2735257693436854, "learning_rate": 1.2499713293007862e-06, "loss": 0.0728, "step": 68 }, { "epoch": 0.015699658703071672, "grad_norm": 1.7727324277447247, "learning_rate": 1.2499704670789301e-06, "loss": 0.0492, "step": 69 }, { "epoch": 0.015927189988623434, "grad_norm": 2.0842450205223977, "learning_rate": 1.2499695920838225e-06, "loss": 0.0552, "step": 70 }, { "epoch": 0.0161547212741752, "grad_norm": 3.5228642040697116, "learning_rate": 1.2499687043154809e-06, "loss": 0.0556, "step": 71 }, { "epoch": 0.016382252559726963, "grad_norm": 2.797328413443058, "learning_rate": 1.2499678037739235e-06, "loss": 0.0519, "step": 72 }, { "epoch": 0.016609783845278725, "grad_norm": 3.349987636757497, "learning_rate": 1.2499668904591688e-06, "loss": 0.1053, "step": 73 }, { "epoch": 0.01683731513083049, "grad_norm": 4.71488578944105, "learning_rate": 1.2499659643712356e-06, "loss": 0.1072, "step": 74 }, { "epoch": 0.017064846416382253, "grad_norm": 1.9864905040446388, "learning_rate": 1.2499650255101425e-06, "loss": 0.1129, "step": 75 }, { "epoch": 0.017292377701934016, "grad_norm": 3.9786000792659073, "learning_rate": 1.2499640738759088e-06, "loss": 0.0798, "step": 76 }, { "epoch": 0.017519908987485778, "grad_norm": 3.293845214395531, "learning_rate": 1.249963109468554e-06, "loss": 0.1235, "step": 77 }, { "epoch": 0.017747440273037544, "grad_norm": 5.489256171335804, "learning_rate": 1.2499621322880979e-06, "loss": 0.0439, "step": 78 }, { "epoch": 0.017974971558589306, "grad_norm": 5.65488778928505, "learning_rate": 1.2499611423345604e-06, "loss": 0.0715, "step": 79 }, { "epoch": 0.01820250284414107, "grad_norm": 1.7907281324160922, "learning_rate": 1.2499601396079617e-06, "loss": 0.0668, "step": 80 }, { "epoch": 0.018430034129692834, "grad_norm": 2.5432248323972377, "learning_rate": 1.2499591241083222e-06, "loss": 0.0836, "step": 81 }, { "epoch": 0.018657565415244597, "grad_norm": 2.6549098221779146, "learning_rate": 1.2499580958356628e-06, "loss": 0.0612, "step": 82 }, { "epoch": 0.01888509670079636, "grad_norm": 3.6428276701171685, "learning_rate": 1.2499570547900045e-06, "loss": 0.0713, "step": 83 }, { "epoch": 0.01911262798634812, "grad_norm": 3.8990952294988994, "learning_rate": 1.2499560009713684e-06, "loss": 0.1046, "step": 84 }, { "epoch": 0.019340159271899887, "grad_norm": 1.7241928848576353, "learning_rate": 1.2499549343797764e-06, "loss": 0.0759, "step": 85 }, { "epoch": 0.01956769055745165, "grad_norm": 2.2613238963696545, "learning_rate": 1.24995385501525e-06, "loss": 0.0931, "step": 86 }, { "epoch": 0.019795221843003412, "grad_norm": 2.3167104475270492, "learning_rate": 1.2499527628778116e-06, "loss": 0.0775, "step": 87 }, { "epoch": 0.020022753128555178, "grad_norm": 2.631127335335558, "learning_rate": 1.2499516579674831e-06, "loss": 0.0911, "step": 88 }, { "epoch": 0.02025028441410694, "grad_norm": 3.902893778838773, "learning_rate": 1.2499505402842872e-06, "loss": 0.1129, "step": 89 }, { "epoch": 0.020477815699658702, "grad_norm": 2.6988246720898905, "learning_rate": 1.2499494098282469e-06, "loss": 0.088, "step": 90 }, { "epoch": 0.020705346985210465, "grad_norm": 1.4451215893923708, "learning_rate": 1.2499482665993851e-06, "loss": 0.0521, "step": 91 }, { "epoch": 0.02093287827076223, "grad_norm": 3.920423356576455, "learning_rate": 1.2499471105977252e-06, "loss": 0.079, "step": 92 }, { "epoch": 0.021160409556313993, "grad_norm": 3.1520828274033486, "learning_rate": 1.249945941823291e-06, "loss": 0.1206, "step": 93 }, { "epoch": 0.021387940841865755, "grad_norm": 6.767303206340345, "learning_rate": 1.2499447602761063e-06, "loss": 0.2231, "step": 94 }, { "epoch": 0.02161547212741752, "grad_norm": 8.857984803581953, "learning_rate": 1.2499435659561954e-06, "loss": 0.2288, "step": 95 }, { "epoch": 0.021843003412969283, "grad_norm": 3.2203370791462302, "learning_rate": 1.2499423588635823e-06, "loss": 0.1181, "step": 96 }, { "epoch": 0.022070534698521046, "grad_norm": 2.3890790390173864, "learning_rate": 1.2499411389982919e-06, "loss": 0.0546, "step": 97 }, { "epoch": 0.02229806598407281, "grad_norm": 3.9314083232645953, "learning_rate": 1.2499399063603492e-06, "loss": 0.1202, "step": 98 }, { "epoch": 0.022525597269624574, "grad_norm": 1.534409734668097, "learning_rate": 1.2499386609497793e-06, "loss": 0.0575, "step": 99 }, { "epoch": 0.022753128555176336, "grad_norm": 2.6915769569708163, "learning_rate": 1.2499374027666078e-06, "loss": 0.0865, "step": 100 }, { "epoch": 0.0229806598407281, "grad_norm": 1.9363626293552272, "learning_rate": 1.2499361318108602e-06, "loss": 0.0691, "step": 101 }, { "epoch": 0.023208191126279865, "grad_norm": 1.706755813278197, "learning_rate": 1.2499348480825627e-06, "loss": 0.0694, "step": 102 }, { "epoch": 0.023435722411831627, "grad_norm": 2.4058707201091902, "learning_rate": 1.2499335515817413e-06, "loss": 0.0873, "step": 103 }, { "epoch": 0.02366325369738339, "grad_norm": 1.4799661409619025, "learning_rate": 1.2499322423084226e-06, "loss": 0.0489, "step": 104 }, { "epoch": 0.023890784982935155, "grad_norm": 1.5529996293769865, "learning_rate": 1.2499309202626336e-06, "loss": 0.0489, "step": 105 }, { "epoch": 0.024118316268486917, "grad_norm": 1.2094533521149893, "learning_rate": 1.249929585444401e-06, "loss": 0.0555, "step": 106 }, { "epoch": 0.02434584755403868, "grad_norm": 3.3515777105454805, "learning_rate": 1.2499282378537522e-06, "loss": 0.0869, "step": 107 }, { "epoch": 0.024573378839590442, "grad_norm": 3.049734475824735, "learning_rate": 1.2499268774907144e-06, "loss": 0.0436, "step": 108 }, { "epoch": 0.024800910125142208, "grad_norm": 1.466216427195955, "learning_rate": 1.249925504355316e-06, "loss": 0.0602, "step": 109 }, { "epoch": 0.02502844141069397, "grad_norm": 1.7889692240619717, "learning_rate": 1.2499241184475848e-06, "loss": 0.0485, "step": 110 }, { "epoch": 0.025255972696245733, "grad_norm": 1.601570429523166, "learning_rate": 1.249922719767549e-06, "loss": 0.0657, "step": 111 }, { "epoch": 0.0254835039817975, "grad_norm": 1.9895197648670602, "learning_rate": 1.2499213083152374e-06, "loss": 0.0613, "step": 112 }, { "epoch": 0.02571103526734926, "grad_norm": 1.559725191147961, "learning_rate": 1.2499198840906787e-06, "loss": 0.0376, "step": 113 }, { "epoch": 0.025938566552901023, "grad_norm": 1.8823702931097668, "learning_rate": 1.249918447093902e-06, "loss": 0.0441, "step": 114 }, { "epoch": 0.026166097838452786, "grad_norm": 2.8320457267761707, "learning_rate": 1.249916997324937e-06, "loss": 0.1287, "step": 115 }, { "epoch": 0.02639362912400455, "grad_norm": 4.382916227720041, "learning_rate": 1.2499155347838129e-06, "loss": 0.0828, "step": 116 }, { "epoch": 0.026621160409556314, "grad_norm": 3.200832910369845, "learning_rate": 1.2499140594705596e-06, "loss": 0.0621, "step": 117 }, { "epoch": 0.026848691695108076, "grad_norm": 1.6914803041014228, "learning_rate": 1.2499125713852076e-06, "loss": 0.0778, "step": 118 }, { "epoch": 0.027076222980659842, "grad_norm": 1.192120885936281, "learning_rate": 1.2499110705277869e-06, "loss": 0.0505, "step": 119 }, { "epoch": 0.027303754266211604, "grad_norm": 2.0573836463025716, "learning_rate": 1.2499095568983284e-06, "loss": 0.0802, "step": 120 }, { "epoch": 0.027531285551763367, "grad_norm": 1.4902545114620074, "learning_rate": 1.2499080304968634e-06, "loss": 0.0565, "step": 121 }, { "epoch": 0.027758816837315133, "grad_norm": 5.341407524309934, "learning_rate": 1.2499064913234222e-06, "loss": 0.0556, "step": 122 }, { "epoch": 0.027986348122866895, "grad_norm": 1.9959451584520156, "learning_rate": 1.249904939378037e-06, "loss": 0.0425, "step": 123 }, { "epoch": 0.028213879408418657, "grad_norm": 4.8411742944908225, "learning_rate": 1.2499033746607395e-06, "loss": 0.1189, "step": 124 }, { "epoch": 0.02844141069397042, "grad_norm": 2.6968922160166624, "learning_rate": 1.2499017971715614e-06, "loss": 0.0952, "step": 125 }, { "epoch": 0.028668941979522185, "grad_norm": 2.498689273522412, "learning_rate": 1.2499002069105348e-06, "loss": 0.0609, "step": 126 }, { "epoch": 0.028896473265073948, "grad_norm": 1.7926923107505222, "learning_rate": 1.2498986038776926e-06, "loss": 0.0674, "step": 127 }, { "epoch": 0.02912400455062571, "grad_norm": 2.951592426546495, "learning_rate": 1.2498969880730671e-06, "loss": 0.0581, "step": 128 }, { "epoch": 0.029351535836177476, "grad_norm": 2.546012031191214, "learning_rate": 1.249895359496692e-06, "loss": 0.0604, "step": 129 }, { "epoch": 0.02957906712172924, "grad_norm": 2.982654740949648, "learning_rate": 1.2498937181486e-06, "loss": 0.1317, "step": 130 }, { "epoch": 0.029806598407281, "grad_norm": 3.35692876580473, "learning_rate": 1.2498920640288248e-06, "loss": 0.1357, "step": 131 }, { "epoch": 0.030034129692832763, "grad_norm": 3.0581666005631694, "learning_rate": 1.2498903971374005e-06, "loss": 0.1404, "step": 132 }, { "epoch": 0.03026166097838453, "grad_norm": 2.8194208797307936, "learning_rate": 1.2498887174743606e-06, "loss": 0.1139, "step": 133 }, { "epoch": 0.03048919226393629, "grad_norm": 1.307812448511231, "learning_rate": 1.24988702503974e-06, "loss": 0.0439, "step": 134 }, { "epoch": 0.030716723549488054, "grad_norm": 2.1828107869380178, "learning_rate": 1.2498853198335728e-06, "loss": 0.0714, "step": 135 }, { "epoch": 0.03094425483503982, "grad_norm": 2.7687164066652588, "learning_rate": 1.2498836018558942e-06, "loss": 0.0588, "step": 136 }, { "epoch": 0.031171786120591582, "grad_norm": 2.0398223571287124, "learning_rate": 1.2498818711067392e-06, "loss": 0.0788, "step": 137 }, { "epoch": 0.031399317406143344, "grad_norm": 1.544887443922682, "learning_rate": 1.2498801275861433e-06, "loss": 0.0424, "step": 138 }, { "epoch": 0.03162684869169511, "grad_norm": 1.9943988491119002, "learning_rate": 1.2498783712941418e-06, "loss": 0.0509, "step": 139 }, { "epoch": 0.03185437997724687, "grad_norm": 2.2756991800533726, "learning_rate": 1.2498766022307709e-06, "loss": 0.077, "step": 140 }, { "epoch": 0.032081911262798635, "grad_norm": 2.111047377421262, "learning_rate": 1.2498748203960665e-06, "loss": 0.0698, "step": 141 }, { "epoch": 0.0323094425483504, "grad_norm": 5.455732324109059, "learning_rate": 1.2498730257900655e-06, "loss": 0.0464, "step": 142 }, { "epoch": 0.03253697383390216, "grad_norm": 5.8400695805570475, "learning_rate": 1.249871218412804e-06, "loss": 0.0808, "step": 143 }, { "epoch": 0.032764505119453925, "grad_norm": 1.6154990753356673, "learning_rate": 1.2498693982643192e-06, "loss": 0.0579, "step": 144 }, { "epoch": 0.03299203640500569, "grad_norm": 2.6179374503343644, "learning_rate": 1.2498675653446485e-06, "loss": 0.0539, "step": 145 }, { "epoch": 0.03321956769055745, "grad_norm": 1.8802237147910983, "learning_rate": 1.249865719653829e-06, "loss": 0.0562, "step": 146 }, { "epoch": 0.033447098976109216, "grad_norm": 1.9601893311780516, "learning_rate": 1.2498638611918985e-06, "loss": 0.0842, "step": 147 }, { "epoch": 0.03367463026166098, "grad_norm": 3.2621758627603814, "learning_rate": 1.249861989958895e-06, "loss": 0.0707, "step": 148 }, { "epoch": 0.03390216154721274, "grad_norm": 1.7359031480428844, "learning_rate": 1.2498601059548572e-06, "loss": 0.0552, "step": 149 }, { "epoch": 0.034129692832764506, "grad_norm": 1.9031742914383765, "learning_rate": 1.2498582091798228e-06, "loss": 0.0551, "step": 150 }, { "epoch": 0.034357224118316265, "grad_norm": 1.6839469596143435, "learning_rate": 1.2498562996338312e-06, "loss": 0.0549, "step": 151 }, { "epoch": 0.03458475540386803, "grad_norm": 4.770073312573169, "learning_rate": 1.249854377316921e-06, "loss": 0.0922, "step": 152 }, { "epoch": 0.0348122866894198, "grad_norm": 9.77865071194089, "learning_rate": 1.2498524422291319e-06, "loss": 0.0656, "step": 153 }, { "epoch": 0.035039817974971556, "grad_norm": 3.279720923464348, "learning_rate": 1.2498504943705033e-06, "loss": 0.0628, "step": 154 }, { "epoch": 0.03526734926052332, "grad_norm": 2.5593319197975086, "learning_rate": 1.249848533741075e-06, "loss": 0.1116, "step": 155 }, { "epoch": 0.03549488054607509, "grad_norm": 2.7020759561258796, "learning_rate": 1.2498465603408865e-06, "loss": 0.0818, "step": 156 }, { "epoch": 0.035722411831626846, "grad_norm": 1.1827259380181383, "learning_rate": 1.2498445741699792e-06, "loss": 0.0412, "step": 157 }, { "epoch": 0.03594994311717861, "grad_norm": 3.090835953132798, "learning_rate": 1.249842575228393e-06, "loss": 0.0922, "step": 158 }, { "epoch": 0.03617747440273038, "grad_norm": 4.286688880757478, "learning_rate": 1.249840563516169e-06, "loss": 0.0921, "step": 159 }, { "epoch": 0.03640500568828214, "grad_norm": 3.826111531217773, "learning_rate": 1.249838539033348e-06, "loss": 0.0711, "step": 160 }, { "epoch": 0.0366325369738339, "grad_norm": 1.286547340104366, "learning_rate": 1.2498365017799715e-06, "loss": 0.0363, "step": 161 }, { "epoch": 0.03686006825938567, "grad_norm": 0.9115912875215112, "learning_rate": 1.2498344517560815e-06, "loss": 0.0499, "step": 162 }, { "epoch": 0.03708759954493743, "grad_norm": 1.967900659157127, "learning_rate": 1.2498323889617198e-06, "loss": 0.0642, "step": 163 }, { "epoch": 0.03731513083048919, "grad_norm": 20.150143075771535, "learning_rate": 1.2498303133969281e-06, "loss": 0.0764, "step": 164 }, { "epoch": 0.03754266211604096, "grad_norm": 1.8715348615633682, "learning_rate": 1.2498282250617492e-06, "loss": 0.0756, "step": 165 }, { "epoch": 0.03777019340159272, "grad_norm": 1.4158902127506805, "learning_rate": 1.2498261239562257e-06, "loss": 0.0669, "step": 166 }, { "epoch": 0.037997724687144484, "grad_norm": 2.5083436605654583, "learning_rate": 1.2498240100804005e-06, "loss": 0.0642, "step": 167 }, { "epoch": 0.03822525597269624, "grad_norm": 4.201645953801033, "learning_rate": 1.249821883434317e-06, "loss": 0.0804, "step": 168 }, { "epoch": 0.03845278725824801, "grad_norm": 1.8056911269518585, "learning_rate": 1.2498197440180182e-06, "loss": 0.0826, "step": 169 }, { "epoch": 0.038680318543799774, "grad_norm": 2.5577509171659174, "learning_rate": 1.2498175918315484e-06, "loss": 0.0782, "step": 170 }, { "epoch": 0.03890784982935153, "grad_norm": 1.6347218920613042, "learning_rate": 1.2498154268749513e-06, "loss": 0.0526, "step": 171 }, { "epoch": 0.0391353811149033, "grad_norm": 2.089011530912283, "learning_rate": 1.249813249148271e-06, "loss": 0.0845, "step": 172 }, { "epoch": 0.039362912400455065, "grad_norm": 2.0840328778691557, "learning_rate": 1.2498110586515525e-06, "loss": 0.0642, "step": 173 }, { "epoch": 0.039590443686006824, "grad_norm": 2.63166064853352, "learning_rate": 1.2498088553848398e-06, "loss": 0.0896, "step": 174 }, { "epoch": 0.03981797497155859, "grad_norm": 1.7790239100772818, "learning_rate": 1.2498066393481787e-06, "loss": 0.0887, "step": 175 }, { "epoch": 0.040045506257110355, "grad_norm": 2.0484275395799374, "learning_rate": 1.249804410541614e-06, "loss": 0.0757, "step": 176 }, { "epoch": 0.040273037542662114, "grad_norm": 2.7339764522222674, "learning_rate": 1.2498021689651916e-06, "loss": 0.0752, "step": 177 }, { "epoch": 0.04050056882821388, "grad_norm": 1.7202036539272256, "learning_rate": 1.249799914618957e-06, "loss": 0.0571, "step": 178 }, { "epoch": 0.040728100113765646, "grad_norm": 0.9595120325924671, "learning_rate": 1.2497976475029566e-06, "loss": 0.0431, "step": 179 }, { "epoch": 0.040955631399317405, "grad_norm": 2.2881459955255283, "learning_rate": 1.2497953676172364e-06, "loss": 0.0714, "step": 180 }, { "epoch": 0.04118316268486917, "grad_norm": 1.1818626309445535, "learning_rate": 1.2497930749618431e-06, "loss": 0.0352, "step": 181 }, { "epoch": 0.04141069397042093, "grad_norm": 1.7233391740704784, "learning_rate": 1.2497907695368238e-06, "loss": 0.0469, "step": 182 }, { "epoch": 0.041638225255972695, "grad_norm": 1.260961347917145, "learning_rate": 1.2497884513422253e-06, "loss": 0.0545, "step": 183 }, { "epoch": 0.04186575654152446, "grad_norm": 1.350181559062471, "learning_rate": 1.249786120378095e-06, "loss": 0.0579, "step": 184 }, { "epoch": 0.04209328782707622, "grad_norm": 1.4312208218600422, "learning_rate": 1.2497837766444806e-06, "loss": 0.0671, "step": 185 }, { "epoch": 0.042320819112627986, "grad_norm": 3.0862470606278527, "learning_rate": 1.2497814201414304e-06, "loss": 0.1552, "step": 186 }, { "epoch": 0.04254835039817975, "grad_norm": 2.010075620155332, "learning_rate": 1.249779050868992e-06, "loss": 0.0819, "step": 187 }, { "epoch": 0.04277588168373151, "grad_norm": 1.5041508185690229, "learning_rate": 1.249776668827214e-06, "loss": 0.0555, "step": 188 }, { "epoch": 0.043003412969283276, "grad_norm": 15.160154885670842, "learning_rate": 1.249774274016145e-06, "loss": 0.1377, "step": 189 }, { "epoch": 0.04323094425483504, "grad_norm": 1.3900253472134594, "learning_rate": 1.2497718664358341e-06, "loss": 0.0573, "step": 190 }, { "epoch": 0.0434584755403868, "grad_norm": 1.5383922685465743, "learning_rate": 1.2497694460863307e-06, "loss": 0.0986, "step": 191 }, { "epoch": 0.04368600682593857, "grad_norm": 2.2430359603043266, "learning_rate": 1.2497670129676838e-06, "loss": 0.0771, "step": 192 }, { "epoch": 0.04391353811149033, "grad_norm": 2.1439327799346333, "learning_rate": 1.2497645670799436e-06, "loss": 0.0436, "step": 193 }, { "epoch": 0.04414106939704209, "grad_norm": 1.771166715500092, "learning_rate": 1.2497621084231595e-06, "loss": 0.0523, "step": 194 }, { "epoch": 0.04436860068259386, "grad_norm": 1.4299897341433283, "learning_rate": 1.2497596369973823e-06, "loss": 0.0401, "step": 195 }, { "epoch": 0.04459613196814562, "grad_norm": 2.3614356932574654, "learning_rate": 1.2497571528026623e-06, "loss": 0.0888, "step": 196 }, { "epoch": 0.04482366325369738, "grad_norm": 1.8304520866772858, "learning_rate": 1.2497546558390503e-06, "loss": 0.065, "step": 197 }, { "epoch": 0.04505119453924915, "grad_norm": 1.8912148609925916, "learning_rate": 1.2497521461065973e-06, "loss": 0.0601, "step": 198 }, { "epoch": 0.04527872582480091, "grad_norm": 3.522073583380884, "learning_rate": 1.2497496236053547e-06, "loss": 0.0841, "step": 199 }, { "epoch": 0.04550625711035267, "grad_norm": 1.3143196144497622, "learning_rate": 1.2497470883353738e-06, "loss": 0.0485, "step": 200 }, { "epoch": 0.04573378839590444, "grad_norm": 2.251264141887164, "learning_rate": 1.2497445402967068e-06, "loss": 0.1068, "step": 201 }, { "epoch": 0.0459613196814562, "grad_norm": 1.5072412944239943, "learning_rate": 1.2497419794894053e-06, "loss": 0.0685, "step": 202 }, { "epoch": 0.04618885096700796, "grad_norm": 3.468383203818734, "learning_rate": 1.249739405913522e-06, "loss": 0.0815, "step": 203 }, { "epoch": 0.04641638225255973, "grad_norm": 2.799795848526847, "learning_rate": 1.2497368195691095e-06, "loss": 0.0611, "step": 204 }, { "epoch": 0.04664391353811149, "grad_norm": 1.8317350709529971, "learning_rate": 1.2497342204562205e-06, "loss": 0.0645, "step": 205 }, { "epoch": 0.046871444823663254, "grad_norm": 1.8552065397519177, "learning_rate": 1.2497316085749081e-06, "loss": 0.0475, "step": 206 }, { "epoch": 0.04709897610921502, "grad_norm": 33.648509264650215, "learning_rate": 1.249728983925226e-06, "loss": 0.4463, "step": 207 }, { "epoch": 0.04732650739476678, "grad_norm": 3.6371540108131555, "learning_rate": 1.2497263465072274e-06, "loss": 0.1261, "step": 208 }, { "epoch": 0.047554038680318544, "grad_norm": 4.579077765212189, "learning_rate": 1.2497236963209663e-06, "loss": 0.1537, "step": 209 }, { "epoch": 0.04778156996587031, "grad_norm": 1.8349256873988717, "learning_rate": 1.2497210333664972e-06, "loss": 0.0905, "step": 210 }, { "epoch": 0.04800910125142207, "grad_norm": 1.0349470677126553, "learning_rate": 1.2497183576438743e-06, "loss": 0.0383, "step": 211 }, { "epoch": 0.048236632536973835, "grad_norm": 1.7552990958997892, "learning_rate": 1.2497156691531523e-06, "loss": 0.0667, "step": 212 }, { "epoch": 0.048464163822525594, "grad_norm": 1.1378619448565346, "learning_rate": 1.249712967894386e-06, "loss": 0.0494, "step": 213 }, { "epoch": 0.04869169510807736, "grad_norm": 2.098843760453871, "learning_rate": 1.2497102538676308e-06, "loss": 0.0683, "step": 214 }, { "epoch": 0.048919226393629126, "grad_norm": 2.4023627244072996, "learning_rate": 1.249707527072942e-06, "loss": 0.0752, "step": 215 }, { "epoch": 0.049146757679180884, "grad_norm": 3.0741883094346405, "learning_rate": 1.2497047875103757e-06, "loss": 0.1576, "step": 216 }, { "epoch": 0.04937428896473265, "grad_norm": 2.109066858277886, "learning_rate": 1.2497020351799875e-06, "loss": 0.0834, "step": 217 }, { "epoch": 0.049601820250284416, "grad_norm": 1.1860658004774445, "learning_rate": 1.2496992700818335e-06, "loss": 0.0487, "step": 218 }, { "epoch": 0.049829351535836175, "grad_norm": 2.78284096499592, "learning_rate": 1.249696492215971e-06, "loss": 0.0743, "step": 219 }, { "epoch": 0.05005688282138794, "grad_norm": 2.111446571336187, "learning_rate": 1.249693701582456e-06, "loss": 0.0381, "step": 220 }, { "epoch": 0.05028441410693971, "grad_norm": 1.8439980435313363, "learning_rate": 1.2496908981813458e-06, "loss": 0.0821, "step": 221 }, { "epoch": 0.050511945392491465, "grad_norm": 2.1292431200987165, "learning_rate": 1.2496880820126977e-06, "loss": 0.102, "step": 222 }, { "epoch": 0.05073947667804323, "grad_norm": 1.58572513687618, "learning_rate": 1.2496852530765695e-06, "loss": 0.0451, "step": 223 }, { "epoch": 0.050967007963595, "grad_norm": 3.028551047012946, "learning_rate": 1.2496824113730186e-06, "loss": 0.1259, "step": 224 }, { "epoch": 0.051194539249146756, "grad_norm": 1.2656190346220688, "learning_rate": 1.2496795569021033e-06, "loss": 0.0593, "step": 225 }, { "epoch": 0.05142207053469852, "grad_norm": 2.5365255203279444, "learning_rate": 1.2496766896638819e-06, "loss": 0.0754, "step": 226 }, { "epoch": 0.05164960182025029, "grad_norm": 2.246271311279298, "learning_rate": 1.249673809658413e-06, "loss": 0.0689, "step": 227 }, { "epoch": 0.05187713310580205, "grad_norm": 1.212864625758916, "learning_rate": 1.2496709168857555e-06, "loss": 0.0451, "step": 228 }, { "epoch": 0.05210466439135381, "grad_norm": 2.806330657303488, "learning_rate": 1.2496680113459683e-06, "loss": 0.1473, "step": 229 }, { "epoch": 0.05233219567690557, "grad_norm": 2.652722899111948, "learning_rate": 1.2496650930391113e-06, "loss": 0.1155, "step": 230 }, { "epoch": 0.05255972696245734, "grad_norm": 2.7863683658107696, "learning_rate": 1.2496621619652435e-06, "loss": 0.0939, "step": 231 }, { "epoch": 0.0527872582480091, "grad_norm": 2.0852803666211925, "learning_rate": 1.2496592181244253e-06, "loss": 0.0385, "step": 232 }, { "epoch": 0.05301478953356086, "grad_norm": 2.3109089889052274, "learning_rate": 1.249656261516717e-06, "loss": 0.0591, "step": 233 }, { "epoch": 0.05324232081911263, "grad_norm": 1.657658325406227, "learning_rate": 1.2496532921421781e-06, "loss": 0.0666, "step": 234 }, { "epoch": 0.053469852104664393, "grad_norm": 1.8341397873796956, "learning_rate": 1.2496503100008704e-06, "loss": 0.0893, "step": 235 }, { "epoch": 0.05369738339021615, "grad_norm": 1.3722825156667726, "learning_rate": 1.249647315092854e-06, "loss": 0.0445, "step": 236 }, { "epoch": 0.05392491467576792, "grad_norm": 2.4225397650308187, "learning_rate": 1.2496443074181905e-06, "loss": 0.0783, "step": 237 }, { "epoch": 0.054152445961319684, "grad_norm": 1.27450689566742, "learning_rate": 1.2496412869769415e-06, "loss": 0.0482, "step": 238 }, { "epoch": 0.05437997724687144, "grad_norm": 1.5788523588867425, "learning_rate": 1.2496382537691686e-06, "loss": 0.0559, "step": 239 }, { "epoch": 0.05460750853242321, "grad_norm": 2.455674592198026, "learning_rate": 1.2496352077949336e-06, "loss": 0.0686, "step": 240 }, { "epoch": 0.054835039817974975, "grad_norm": 0.9822051281919313, "learning_rate": 1.249632149054299e-06, "loss": 0.0333, "step": 241 }, { "epoch": 0.05506257110352673, "grad_norm": 1.4918328919143296, "learning_rate": 1.249629077547327e-06, "loss": 0.0564, "step": 242 }, { "epoch": 0.0552901023890785, "grad_norm": 2.2431233119775973, "learning_rate": 1.2496259932740813e-06, "loss": 0.0998, "step": 243 }, { "epoch": 0.055517633674630265, "grad_norm": 1.0547810468664673, "learning_rate": 1.2496228962346236e-06, "loss": 0.054, "step": 244 }, { "epoch": 0.055745164960182024, "grad_norm": 4.2608959679243785, "learning_rate": 1.249619786429018e-06, "loss": 0.1324, "step": 245 }, { "epoch": 0.05597269624573379, "grad_norm": 0.9597443879271953, "learning_rate": 1.2496166638573278e-06, "loss": 0.0545, "step": 246 }, { "epoch": 0.05620022753128555, "grad_norm": 1.0564009660406621, "learning_rate": 1.2496135285196172e-06, "loss": 0.0409, "step": 247 }, { "epoch": 0.056427758816837315, "grad_norm": 5.181656955922701, "learning_rate": 1.2496103804159497e-06, "loss": 0.0514, "step": 248 }, { "epoch": 0.05665529010238908, "grad_norm": 31.956127145863782, "learning_rate": 1.2496072195463904e-06, "loss": 0.5089, "step": 249 }, { "epoch": 0.05688282138794084, "grad_norm": 1.6118693718642072, "learning_rate": 1.249604045911003e-06, "loss": 0.0601, "step": 250 }, { "epoch": 0.057110352673492605, "grad_norm": 1.8753798893546312, "learning_rate": 1.249600859509853e-06, "loss": 0.0781, "step": 251 }, { "epoch": 0.05733788395904437, "grad_norm": 1.9173456172296954, "learning_rate": 1.2495976603430054e-06, "loss": 0.0998, "step": 252 }, { "epoch": 0.05756541524459613, "grad_norm": 1.7446295444585516, "learning_rate": 1.2495944484105254e-06, "loss": 0.0348, "step": 253 }, { "epoch": 0.057792946530147896, "grad_norm": 3.1221291176703816, "learning_rate": 1.2495912237124787e-06, "loss": 0.0467, "step": 254 }, { "epoch": 0.05802047781569966, "grad_norm": 3.8018086885799454, "learning_rate": 1.2495879862489312e-06, "loss": 0.0893, "step": 255 }, { "epoch": 0.05824800910125142, "grad_norm": 7.114525761671455, "learning_rate": 1.2495847360199495e-06, "loss": 0.0647, "step": 256 }, { "epoch": 0.058475540386803186, "grad_norm": 1.861881748798064, "learning_rate": 1.2495814730255993e-06, "loss": 0.0832, "step": 257 }, { "epoch": 0.05870307167235495, "grad_norm": 1.920791237896296, "learning_rate": 1.2495781972659479e-06, "loss": 0.0346, "step": 258 }, { "epoch": 0.05893060295790671, "grad_norm": 1.9123959618493658, "learning_rate": 1.2495749087410618e-06, "loss": 0.0924, "step": 259 }, { "epoch": 0.05915813424345848, "grad_norm": 4.29696428710133, "learning_rate": 1.2495716074510087e-06, "loss": 0.0503, "step": 260 }, { "epoch": 0.059385665529010236, "grad_norm": 1.518204196410266, "learning_rate": 1.2495682933958555e-06, "loss": 0.0516, "step": 261 }, { "epoch": 0.059613196814562, "grad_norm": 4.632688881661792, "learning_rate": 1.2495649665756705e-06, "loss": 0.1211, "step": 262 }, { "epoch": 0.05984072810011377, "grad_norm": 2.486723056340893, "learning_rate": 1.2495616269905212e-06, "loss": 0.0811, "step": 263 }, { "epoch": 0.060068259385665526, "grad_norm": 1.581969386319132, "learning_rate": 1.2495582746404762e-06, "loss": 0.0589, "step": 264 }, { "epoch": 0.06029579067121729, "grad_norm": 1.5286635653264549, "learning_rate": 1.249554909525604e-06, "loss": 0.0638, "step": 265 }, { "epoch": 0.06052332195676906, "grad_norm": 1.2992103388623448, "learning_rate": 1.249551531645973e-06, "loss": 0.0309, "step": 266 }, { "epoch": 0.06075085324232082, "grad_norm": 1.4096677819832404, "learning_rate": 1.2495481410016527e-06, "loss": 0.0779, "step": 267 }, { "epoch": 0.06097838452787258, "grad_norm": 2.5850002425545586, "learning_rate": 1.2495447375927122e-06, "loss": 0.0718, "step": 268 }, { "epoch": 0.06120591581342435, "grad_norm": 2.2171422082649155, "learning_rate": 1.2495413214192209e-06, "loss": 0.0761, "step": 269 }, { "epoch": 0.06143344709897611, "grad_norm": 1.8061938933671926, "learning_rate": 1.2495378924812486e-06, "loss": 0.068, "step": 270 }, { "epoch": 0.06166097838452787, "grad_norm": 1.805390591341637, "learning_rate": 1.2495344507788662e-06, "loss": 0.0589, "step": 271 }, { "epoch": 0.06188850967007964, "grad_norm": 28.567737212004268, "learning_rate": 1.249530996312143e-06, "loss": 0.2672, "step": 272 }, { "epoch": 0.0621160409556314, "grad_norm": 2.023457290937467, "learning_rate": 1.2495275290811499e-06, "loss": 0.0762, "step": 273 }, { "epoch": 0.062343572241183164, "grad_norm": 1.726293880974768, "learning_rate": 1.2495240490859581e-06, "loss": 0.1124, "step": 274 }, { "epoch": 0.06257110352673492, "grad_norm": 3.5654068745258094, "learning_rate": 1.2495205563266384e-06, "loss": 0.08, "step": 275 }, { "epoch": 0.06279863481228669, "grad_norm": 3.330534204865081, "learning_rate": 1.2495170508032624e-06, "loss": 0.1063, "step": 276 }, { "epoch": 0.06302616609783845, "grad_norm": 2.7347891521994265, "learning_rate": 1.2495135325159015e-06, "loss": 0.0544, "step": 277 }, { "epoch": 0.06325369738339022, "grad_norm": 1.6191657189271829, "learning_rate": 1.2495100014646277e-06, "loss": 0.0399, "step": 278 }, { "epoch": 0.06348122866894199, "grad_norm": 2.2760450280718163, "learning_rate": 1.2495064576495134e-06, "loss": 0.0842, "step": 279 }, { "epoch": 0.06370875995449374, "grad_norm": 1.26417795343513, "learning_rate": 1.2495029010706306e-06, "loss": 0.0396, "step": 280 }, { "epoch": 0.0639362912400455, "grad_norm": 2.452124440380249, "learning_rate": 1.2494993317280524e-06, "loss": 0.057, "step": 281 }, { "epoch": 0.06416382252559727, "grad_norm": 1.6768179224480646, "learning_rate": 1.2494957496218516e-06, "loss": 0.0778, "step": 282 }, { "epoch": 0.06439135381114904, "grad_norm": 1.148511080426964, "learning_rate": 1.2494921547521013e-06, "loss": 0.0593, "step": 283 }, { "epoch": 0.0646188850967008, "grad_norm": 2.6954847712989687, "learning_rate": 1.249488547118875e-06, "loss": 0.0581, "step": 284 }, { "epoch": 0.06484641638225255, "grad_norm": 1.7038701082720327, "learning_rate": 1.2494849267222466e-06, "loss": 0.0467, "step": 285 }, { "epoch": 0.06507394766780432, "grad_norm": 2.0205984871282654, "learning_rate": 1.24948129356229e-06, "loss": 0.08, "step": 286 }, { "epoch": 0.06530147895335608, "grad_norm": 2.4430734451001093, "learning_rate": 1.2494776476390793e-06, "loss": 0.0689, "step": 287 }, { "epoch": 0.06552901023890785, "grad_norm": 2.2074548221975223, "learning_rate": 1.2494739889526894e-06, "loss": 0.036, "step": 288 }, { "epoch": 0.06575654152445962, "grad_norm": 2.3839227269087115, "learning_rate": 1.2494703175031946e-06, "loss": 0.1017, "step": 289 }, { "epoch": 0.06598407281001138, "grad_norm": 1.297764082661735, "learning_rate": 1.2494666332906702e-06, "loss": 0.0428, "step": 290 }, { "epoch": 0.06621160409556313, "grad_norm": 2.29700118159146, "learning_rate": 1.2494629363151916e-06, "loss": 0.0564, "step": 291 }, { "epoch": 0.0664391353811149, "grad_norm": 10.343219429302799, "learning_rate": 1.2494592265768343e-06, "loss": 0.2299, "step": 292 }, { "epoch": 0.06666666666666667, "grad_norm": 1.7119116189327344, "learning_rate": 1.2494555040756737e-06, "loss": 0.0491, "step": 293 }, { "epoch": 0.06689419795221843, "grad_norm": 1.5996315416913807, "learning_rate": 1.2494517688117867e-06, "loss": 0.0569, "step": 294 }, { "epoch": 0.0671217292377702, "grad_norm": 2.2996053958249068, "learning_rate": 1.2494480207852489e-06, "loss": 0.0561, "step": 295 }, { "epoch": 0.06734926052332196, "grad_norm": 2.4074335741025346, "learning_rate": 1.249444259996137e-06, "loss": 0.0578, "step": 296 }, { "epoch": 0.06757679180887372, "grad_norm": 3.415732508941554, "learning_rate": 1.2494404864445284e-06, "loss": 0.0617, "step": 297 }, { "epoch": 0.06780432309442548, "grad_norm": 2.777678894153543, "learning_rate": 1.2494367001304996e-06, "loss": 0.081, "step": 298 }, { "epoch": 0.06803185437997725, "grad_norm": 1.6387141538383663, "learning_rate": 1.2494329010541284e-06, "loss": 0.0387, "step": 299 }, { "epoch": 0.06825938566552901, "grad_norm": 2.069756645203544, "learning_rate": 1.2494290892154922e-06, "loss": 0.0742, "step": 300 }, { "epoch": 0.06848691695108078, "grad_norm": 3.4307396136508332, "learning_rate": 1.2494252646146692e-06, "loss": 0.0957, "step": 301 }, { "epoch": 0.06871444823663253, "grad_norm": 2.0232162640823974, "learning_rate": 1.249421427251737e-06, "loss": 0.0522, "step": 302 }, { "epoch": 0.0689419795221843, "grad_norm": 2.146160892723777, "learning_rate": 1.2494175771267748e-06, "loss": 0.0896, "step": 303 }, { "epoch": 0.06916951080773606, "grad_norm": 2.940508414612204, "learning_rate": 1.2494137142398607e-06, "loss": 0.0776, "step": 304 }, { "epoch": 0.06939704209328783, "grad_norm": 4.306654470652337, "learning_rate": 1.249409838591074e-06, "loss": 0.112, "step": 305 }, { "epoch": 0.0696245733788396, "grad_norm": 2.698854439380438, "learning_rate": 1.2494059501804937e-06, "loss": 0.0922, "step": 306 }, { "epoch": 0.06985210466439136, "grad_norm": 2.684055328227073, "learning_rate": 1.249402049008199e-06, "loss": 0.0603, "step": 307 }, { "epoch": 0.07007963594994311, "grad_norm": 1.6527179950360487, "learning_rate": 1.2493981350742704e-06, "loss": 0.035, "step": 308 }, { "epoch": 0.07030716723549488, "grad_norm": 2.632569906843673, "learning_rate": 1.2493942083787872e-06, "loss": 0.0483, "step": 309 }, { "epoch": 0.07053469852104664, "grad_norm": 1.8419184740112247, "learning_rate": 1.2493902689218299e-06, "loss": 0.0592, "step": 310 }, { "epoch": 0.07076222980659841, "grad_norm": 4.098447223521742, "learning_rate": 1.249386316703479e-06, "loss": 0.0647, "step": 311 }, { "epoch": 0.07098976109215017, "grad_norm": 3.425692782797119, "learning_rate": 1.2493823517238154e-06, "loss": 0.0304, "step": 312 }, { "epoch": 0.07121729237770194, "grad_norm": 2.09845005259606, "learning_rate": 1.2493783739829202e-06, "loss": 0.0841, "step": 313 }, { "epoch": 0.07144482366325369, "grad_norm": 1.3953874008324176, "learning_rate": 1.2493743834808741e-06, "loss": 0.0572, "step": 314 }, { "epoch": 0.07167235494880546, "grad_norm": 2.623892824594455, "learning_rate": 1.2493703802177594e-06, "loss": 0.0625, "step": 315 }, { "epoch": 0.07189988623435722, "grad_norm": 1.2851580205849038, "learning_rate": 1.2493663641936576e-06, "loss": 0.0321, "step": 316 }, { "epoch": 0.07212741751990899, "grad_norm": 1.9343201552647253, "learning_rate": 1.2493623354086507e-06, "loss": 0.067, "step": 317 }, { "epoch": 0.07235494880546076, "grad_norm": 2.2726283028135286, "learning_rate": 1.2493582938628213e-06, "loss": 0.0797, "step": 318 }, { "epoch": 0.07258248009101251, "grad_norm": 1.669835013133446, "learning_rate": 1.2493542395562516e-06, "loss": 0.0534, "step": 319 }, { "epoch": 0.07281001137656427, "grad_norm": 1.684538229380955, "learning_rate": 1.2493501724890247e-06, "loss": 0.0728, "step": 320 }, { "epoch": 0.07303754266211604, "grad_norm": 1.9071354849813447, "learning_rate": 1.249346092661224e-06, "loss": 0.0678, "step": 321 }, { "epoch": 0.0732650739476678, "grad_norm": 1.7042139773464355, "learning_rate": 1.2493420000729322e-06, "loss": 0.0784, "step": 322 }, { "epoch": 0.07349260523321957, "grad_norm": 1.6735213517836012, "learning_rate": 1.2493378947242336e-06, "loss": 0.0776, "step": 323 }, { "epoch": 0.07372013651877134, "grad_norm": 1.5646112601537545, "learning_rate": 1.2493337766152119e-06, "loss": 0.0582, "step": 324 }, { "epoch": 0.07394766780432309, "grad_norm": 1.5543937889365869, "learning_rate": 1.249329645745951e-06, "loss": 0.0512, "step": 325 }, { "epoch": 0.07417519908987485, "grad_norm": 1.982236801977647, "learning_rate": 1.2493255021165357e-06, "loss": 0.0804, "step": 326 }, { "epoch": 0.07440273037542662, "grad_norm": 1.2394712998486772, "learning_rate": 1.2493213457270504e-06, "loss": 0.0534, "step": 327 }, { "epoch": 0.07463026166097839, "grad_norm": 1.7803933921864556, "learning_rate": 1.2493171765775804e-06, "loss": 0.0687, "step": 328 }, { "epoch": 0.07485779294653015, "grad_norm": 2.379962120888543, "learning_rate": 1.2493129946682107e-06, "loss": 0.071, "step": 329 }, { "epoch": 0.07508532423208192, "grad_norm": 1.589657975033299, "learning_rate": 1.2493087999990263e-06, "loss": 0.0739, "step": 330 }, { "epoch": 0.07531285551763367, "grad_norm": 3.043606210421758, "learning_rate": 1.249304592570114e-06, "loss": 0.1518, "step": 331 }, { "epoch": 0.07554038680318544, "grad_norm": 1.4392385212589442, "learning_rate": 1.2493003723815588e-06, "loss": 0.0556, "step": 332 }, { "epoch": 0.0757679180887372, "grad_norm": 2.0188482312629397, "learning_rate": 1.2492961394334474e-06, "loss": 0.0662, "step": 333 }, { "epoch": 0.07599544937428897, "grad_norm": 1.3600493699697815, "learning_rate": 1.2492918937258663e-06, "loss": 0.042, "step": 334 }, { "epoch": 0.07622298065984073, "grad_norm": 2.2314121339792714, "learning_rate": 1.2492876352589024e-06, "loss": 0.0751, "step": 335 }, { "epoch": 0.07645051194539249, "grad_norm": 1.4434673698644476, "learning_rate": 1.2492833640326424e-06, "loss": 0.0359, "step": 336 }, { "epoch": 0.07667804323094425, "grad_norm": 1.9053597077248743, "learning_rate": 1.2492790800471738e-06, "loss": 0.0881, "step": 337 }, { "epoch": 0.07690557451649602, "grad_norm": 0.9804503872776514, "learning_rate": 1.249274783302584e-06, "loss": 0.0332, "step": 338 }, { "epoch": 0.07713310580204778, "grad_norm": 1.439933416412126, "learning_rate": 1.249270473798961e-06, "loss": 0.0363, "step": 339 }, { "epoch": 0.07736063708759955, "grad_norm": 1.8018763642808848, "learning_rate": 1.249266151536393e-06, "loss": 0.0652, "step": 340 }, { "epoch": 0.07758816837315131, "grad_norm": 1.605885586174771, "learning_rate": 1.249261816514968e-06, "loss": 0.0483, "step": 341 }, { "epoch": 0.07781569965870307, "grad_norm": 1.5475812110829932, "learning_rate": 1.2492574687347747e-06, "loss": 0.0518, "step": 342 }, { "epoch": 0.07804323094425483, "grad_norm": 1.3861334147739204, "learning_rate": 1.249253108195902e-06, "loss": 0.0438, "step": 343 }, { "epoch": 0.0782707622298066, "grad_norm": 1.6280816638823692, "learning_rate": 1.249248734898439e-06, "loss": 0.0804, "step": 344 }, { "epoch": 0.07849829351535836, "grad_norm": 1.945038730906618, "learning_rate": 1.2492443488424753e-06, "loss": 0.08, "step": 345 }, { "epoch": 0.07872582480091013, "grad_norm": 1.163435954016696, "learning_rate": 1.2492399500281002e-06, "loss": 0.0404, "step": 346 }, { "epoch": 0.07895335608646188, "grad_norm": 2.0650618862775225, "learning_rate": 1.2492355384554039e-06, "loss": 0.0965, "step": 347 }, { "epoch": 0.07918088737201365, "grad_norm": 1.2688387295944672, "learning_rate": 1.2492311141244764e-06, "loss": 0.0387, "step": 348 }, { "epoch": 0.07940841865756541, "grad_norm": 2.2968428653497472, "learning_rate": 1.249226677035408e-06, "loss": 0.0569, "step": 349 }, { "epoch": 0.07963594994311718, "grad_norm": 1.3014564625890543, "learning_rate": 1.2492222271882896e-06, "loss": 0.0418, "step": 350 }, { "epoch": 0.07986348122866894, "grad_norm": 2.3510870610479433, "learning_rate": 1.2492177645832121e-06, "loss": 0.0932, "step": 351 }, { "epoch": 0.08009101251422071, "grad_norm": 1.4510684428455591, "learning_rate": 1.2492132892202668e-06, "loss": 0.0513, "step": 352 }, { "epoch": 0.08031854379977246, "grad_norm": 4.594413722551582, "learning_rate": 1.2492088010995449e-06, "loss": 0.1235, "step": 353 }, { "epoch": 0.08054607508532423, "grad_norm": 2.137036810353962, "learning_rate": 1.2492043002211385e-06, "loss": 0.1026, "step": 354 }, { "epoch": 0.080773606370876, "grad_norm": 0.7577275839475693, "learning_rate": 1.2491997865851392e-06, "loss": 0.0368, "step": 355 }, { "epoch": 0.08100113765642776, "grad_norm": 1.0494266975406659, "learning_rate": 1.2491952601916395e-06, "loss": 0.0385, "step": 356 }, { "epoch": 0.08122866894197953, "grad_norm": 1.0423156372262818, "learning_rate": 1.2491907210407319e-06, "loss": 0.0424, "step": 357 }, { "epoch": 0.08145620022753129, "grad_norm": 1.150214646729361, "learning_rate": 1.249186169132509e-06, "loss": 0.0694, "step": 358 }, { "epoch": 0.08168373151308304, "grad_norm": 2.382439951131263, "learning_rate": 1.2491816044670641e-06, "loss": 0.0367, "step": 359 }, { "epoch": 0.08191126279863481, "grad_norm": 1.7484749214098252, "learning_rate": 1.24917702704449e-06, "loss": 0.0677, "step": 360 }, { "epoch": 0.08213879408418658, "grad_norm": 2.1328792490346227, "learning_rate": 1.2491724368648808e-06, "loss": 0.049, "step": 361 }, { "epoch": 0.08236632536973834, "grad_norm": 0.6514880974265386, "learning_rate": 1.2491678339283303e-06, "loss": 0.0217, "step": 362 }, { "epoch": 0.08259385665529011, "grad_norm": 1.3802700887569803, "learning_rate": 1.249163218234932e-06, "loss": 0.0406, "step": 363 }, { "epoch": 0.08282138794084186, "grad_norm": 1.7918385557024892, "learning_rate": 1.249158589784781e-06, "loss": 0.064, "step": 364 }, { "epoch": 0.08304891922639362, "grad_norm": 1.2060165323705336, "learning_rate": 1.2491539485779713e-06, "loss": 0.0367, "step": 365 }, { "epoch": 0.08327645051194539, "grad_norm": 1.2166249313910695, "learning_rate": 1.2491492946145981e-06, "loss": 0.0512, "step": 366 }, { "epoch": 0.08350398179749716, "grad_norm": 1.3048815510766842, "learning_rate": 1.2491446278947563e-06, "loss": 0.0554, "step": 367 }, { "epoch": 0.08373151308304892, "grad_norm": 1.9459326713846232, "learning_rate": 1.2491399484185413e-06, "loss": 0.0488, "step": 368 }, { "epoch": 0.08395904436860069, "grad_norm": 1.7915232664153333, "learning_rate": 1.249135256186049e-06, "loss": 0.1229, "step": 369 }, { "epoch": 0.08418657565415244, "grad_norm": 0.9354087864668412, "learning_rate": 1.249130551197375e-06, "loss": 0.0383, "step": 370 }, { "epoch": 0.0844141069397042, "grad_norm": 2.07264306868735, "learning_rate": 1.2491258334526155e-06, "loss": 0.0794, "step": 371 }, { "epoch": 0.08464163822525597, "grad_norm": 1.4534854537948092, "learning_rate": 1.2491211029518672e-06, "loss": 0.0282, "step": 372 }, { "epoch": 0.08486916951080774, "grad_norm": 2.0044431124055984, "learning_rate": 1.2491163596952264e-06, "loss": 0.0709, "step": 373 }, { "epoch": 0.0850967007963595, "grad_norm": 1.6305074169029203, "learning_rate": 1.2491116036827902e-06, "loss": 0.0407, "step": 374 }, { "epoch": 0.08532423208191127, "grad_norm": 1.223898789967783, "learning_rate": 1.2491068349146559e-06, "loss": 0.0595, "step": 375 }, { "epoch": 0.08555176336746302, "grad_norm": 1.7895390232916124, "learning_rate": 1.249102053390921e-06, "loss": 0.1034, "step": 376 }, { "epoch": 0.08577929465301479, "grad_norm": 1.6725786460952021, "learning_rate": 1.249097259111683e-06, "loss": 0.0955, "step": 377 }, { "epoch": 0.08600682593856655, "grad_norm": 2.4309538605657735, "learning_rate": 1.24909245207704e-06, "loss": 0.0794, "step": 378 }, { "epoch": 0.08623435722411832, "grad_norm": 0.8982937912678952, "learning_rate": 1.2490876322870904e-06, "loss": 0.0302, "step": 379 }, { "epoch": 0.08646188850967008, "grad_norm": 3.0756164983832, "learning_rate": 1.2490827997419325e-06, "loss": 0.0549, "step": 380 }, { "epoch": 0.08668941979522184, "grad_norm": 3.676872613008804, "learning_rate": 1.249077954441665e-06, "loss": 0.141, "step": 381 }, { "epoch": 0.0869169510807736, "grad_norm": 1.4373261550715828, "learning_rate": 1.249073096386387e-06, "loss": 0.0392, "step": 382 }, { "epoch": 0.08714448236632537, "grad_norm": 3.1988616398871006, "learning_rate": 1.249068225576198e-06, "loss": 0.0915, "step": 383 }, { "epoch": 0.08737201365187713, "grad_norm": 1.5619645893707617, "learning_rate": 1.2490633420111974e-06, "loss": 0.0541, "step": 384 }, { "epoch": 0.0875995449374289, "grad_norm": 0.9211933924050371, "learning_rate": 1.249058445691485e-06, "loss": 0.0337, "step": 385 }, { "epoch": 0.08782707622298067, "grad_norm": 1.1767496728905957, "learning_rate": 1.2490535366171607e-06, "loss": 0.039, "step": 386 }, { "epoch": 0.08805460750853242, "grad_norm": 2.9534787325055274, "learning_rate": 1.249048614788325e-06, "loss": 0.1048, "step": 387 }, { "epoch": 0.08828213879408418, "grad_norm": 6.651807262283111, "learning_rate": 1.249043680205079e-06, "loss": 0.0494, "step": 388 }, { "epoch": 0.08850967007963595, "grad_norm": 2.5123377668968745, "learning_rate": 1.2490387328675226e-06, "loss": 0.1199, "step": 389 }, { "epoch": 0.08873720136518772, "grad_norm": 1.5502594476140923, "learning_rate": 1.2490337727757576e-06, "loss": 0.0503, "step": 390 }, { "epoch": 0.08896473265073948, "grad_norm": 5.640796726611394, "learning_rate": 1.249028799929885e-06, "loss": 0.06, "step": 391 }, { "epoch": 0.08919226393629125, "grad_norm": 1.2047358188375967, "learning_rate": 1.2490238143300066e-06, "loss": 0.0427, "step": 392 }, { "epoch": 0.089419795221843, "grad_norm": 1.5816796650719427, "learning_rate": 1.2490188159762243e-06, "loss": 0.0665, "step": 393 }, { "epoch": 0.08964732650739476, "grad_norm": 2.8164102954322505, "learning_rate": 1.2490138048686405e-06, "loss": 0.076, "step": 394 }, { "epoch": 0.08987485779294653, "grad_norm": 2.0775035442513596, "learning_rate": 1.249008781007357e-06, "loss": 0.0857, "step": 395 }, { "epoch": 0.0901023890784983, "grad_norm": 1.3816134442939907, "learning_rate": 1.2490037443924768e-06, "loss": 0.0647, "step": 396 }, { "epoch": 0.09032992036405006, "grad_norm": 5.3602328587776205, "learning_rate": 1.2489986950241032e-06, "loss": 0.0857, "step": 397 }, { "epoch": 0.09055745164960181, "grad_norm": 1.3668732893642688, "learning_rate": 1.2489936329023387e-06, "loss": 0.0513, "step": 398 }, { "epoch": 0.09078498293515358, "grad_norm": 1.3818985376090422, "learning_rate": 1.2489885580272874e-06, "loss": 0.0643, "step": 399 }, { "epoch": 0.09101251422070535, "grad_norm": 1.1645733063202421, "learning_rate": 1.2489834703990527e-06, "loss": 0.0564, "step": 400 }, { "epoch": 0.09124004550625711, "grad_norm": 2.579861013447331, "learning_rate": 1.2489783700177385e-06, "loss": 0.0844, "step": 401 }, { "epoch": 0.09146757679180888, "grad_norm": 1.7416779142182774, "learning_rate": 1.2489732568834492e-06, "loss": 0.0874, "step": 402 }, { "epoch": 0.09169510807736064, "grad_norm": 1.2036980088831546, "learning_rate": 1.2489681309962895e-06, "loss": 0.0605, "step": 403 }, { "epoch": 0.0919226393629124, "grad_norm": 2.823337671795372, "learning_rate": 1.2489629923563637e-06, "loss": 0.0563, "step": 404 }, { "epoch": 0.09215017064846416, "grad_norm": 3.5147572243595406, "learning_rate": 1.2489578409637774e-06, "loss": 0.0962, "step": 405 }, { "epoch": 0.09237770193401593, "grad_norm": 2.441595167267241, "learning_rate": 1.2489526768186352e-06, "loss": 0.0409, "step": 406 }, { "epoch": 0.09260523321956769, "grad_norm": 1.606289452778594, "learning_rate": 1.2489474999210434e-06, "loss": 0.083, "step": 407 }, { "epoch": 0.09283276450511946, "grad_norm": 2.0108948875054358, "learning_rate": 1.2489423102711068e-06, "loss": 0.0424, "step": 408 }, { "epoch": 0.09306029579067122, "grad_norm": 1.9517497199537932, "learning_rate": 1.2489371078689326e-06, "loss": 0.063, "step": 409 }, { "epoch": 0.09328782707622298, "grad_norm": 2.1984153851855903, "learning_rate": 1.2489318927146263e-06, "loss": 0.0607, "step": 410 }, { "epoch": 0.09351535836177474, "grad_norm": 0.9821028313036854, "learning_rate": 1.2489266648082951e-06, "loss": 0.0269, "step": 411 }, { "epoch": 0.09374288964732651, "grad_norm": 1.4642895316806703, "learning_rate": 1.2489214241500453e-06, "loss": 0.0773, "step": 412 }, { "epoch": 0.09397042093287827, "grad_norm": 2.6315717094149598, "learning_rate": 1.2489161707399843e-06, "loss": 0.1042, "step": 413 }, { "epoch": 0.09419795221843004, "grad_norm": 1.2730067445859696, "learning_rate": 1.2489109045782194e-06, "loss": 0.0397, "step": 414 }, { "epoch": 0.09442548350398179, "grad_norm": 1.0214233244231261, "learning_rate": 1.2489056256648582e-06, "loss": 0.0297, "step": 415 }, { "epoch": 0.09465301478953356, "grad_norm": 1.8895414872949885, "learning_rate": 1.2489003340000089e-06, "loss": 0.1027, "step": 416 }, { "epoch": 0.09488054607508532, "grad_norm": 1.2431995135273388, "learning_rate": 1.2488950295837792e-06, "loss": 0.0694, "step": 417 }, { "epoch": 0.09510807736063709, "grad_norm": 2.741996145452892, "learning_rate": 1.2488897124162777e-06, "loss": 0.0815, "step": 418 }, { "epoch": 0.09533560864618885, "grad_norm": 1.711178715312897, "learning_rate": 1.248884382497613e-06, "loss": 0.0955, "step": 419 }, { "epoch": 0.09556313993174062, "grad_norm": 2.609523312381175, "learning_rate": 1.2488790398278941e-06, "loss": 0.1047, "step": 420 }, { "epoch": 0.09579067121729237, "grad_norm": 1.53212810794613, "learning_rate": 1.2488736844072304e-06, "loss": 0.0585, "step": 421 }, { "epoch": 0.09601820250284414, "grad_norm": 1.4174025848206742, "learning_rate": 1.248868316235731e-06, "loss": 0.0387, "step": 422 }, { "epoch": 0.0962457337883959, "grad_norm": 2.1766082548840644, "learning_rate": 1.2488629353135059e-06, "loss": 0.0883, "step": 423 }, { "epoch": 0.09647326507394767, "grad_norm": 1.2878325712311849, "learning_rate": 1.2488575416406649e-06, "loss": 0.0469, "step": 424 }, { "epoch": 0.09670079635949944, "grad_norm": 2.1138235879975222, "learning_rate": 1.2488521352173183e-06, "loss": 0.0531, "step": 425 }, { "epoch": 0.09692832764505119, "grad_norm": 1.6090227620325084, "learning_rate": 1.2488467160435765e-06, "loss": 0.0532, "step": 426 }, { "epoch": 0.09715585893060295, "grad_norm": 1.5021035863134842, "learning_rate": 1.2488412841195505e-06, "loss": 0.0805, "step": 427 }, { "epoch": 0.09738339021615472, "grad_norm": 2.0944361790568515, "learning_rate": 1.2488358394453512e-06, "loss": 0.0864, "step": 428 }, { "epoch": 0.09761092150170649, "grad_norm": 1.3336921136142907, "learning_rate": 1.2488303820210897e-06, "loss": 0.0405, "step": 429 }, { "epoch": 0.09783845278725825, "grad_norm": 1.4602632552266945, "learning_rate": 1.2488249118468776e-06, "loss": 0.0319, "step": 430 }, { "epoch": 0.09806598407281002, "grad_norm": 1.1300406263272467, "learning_rate": 1.248819428922827e-06, "loss": 0.0328, "step": 431 }, { "epoch": 0.09829351535836177, "grad_norm": 3.2436410504809006, "learning_rate": 1.2488139332490495e-06, "loss": 0.1052, "step": 432 }, { "epoch": 0.09852104664391353, "grad_norm": 1.1456421976312734, "learning_rate": 1.248808424825658e-06, "loss": 0.0455, "step": 433 }, { "epoch": 0.0987485779294653, "grad_norm": 1.8967128040132528, "learning_rate": 1.2488029036527645e-06, "loss": 0.0521, "step": 434 }, { "epoch": 0.09897610921501707, "grad_norm": 2.22293609570488, "learning_rate": 1.2487973697304822e-06, "loss": 0.0401, "step": 435 }, { "epoch": 0.09920364050056883, "grad_norm": 1.3201279140220556, "learning_rate": 1.248791823058924e-06, "loss": 0.0656, "step": 436 }, { "epoch": 0.0994311717861206, "grad_norm": 1.8801560325756854, "learning_rate": 1.2487862636382034e-06, "loss": 0.036, "step": 437 }, { "epoch": 0.09965870307167235, "grad_norm": 0.9413516301280102, "learning_rate": 1.248780691468434e-06, "loss": 0.0369, "step": 438 }, { "epoch": 0.09988623435722412, "grad_norm": 1.6280926081257134, "learning_rate": 1.2487751065497296e-06, "loss": 0.0679, "step": 439 }, { "epoch": 0.10011376564277588, "grad_norm": 1.1483666492978277, "learning_rate": 1.2487695088822044e-06, "loss": 0.0476, "step": 440 }, { "epoch": 0.10034129692832765, "grad_norm": 1.091509019893692, "learning_rate": 1.2487638984659729e-06, "loss": 0.0267, "step": 441 }, { "epoch": 0.10056882821387941, "grad_norm": 1.2894939135732226, "learning_rate": 1.2487582753011496e-06, "loss": 0.0379, "step": 442 }, { "epoch": 0.10079635949943117, "grad_norm": 1.5539892877637305, "learning_rate": 1.2487526393878497e-06, "loss": 0.0717, "step": 443 }, { "epoch": 0.10102389078498293, "grad_norm": 1.0799436334247106, "learning_rate": 1.248746990726188e-06, "loss": 0.0494, "step": 444 }, { "epoch": 0.1012514220705347, "grad_norm": 1.0482657304812906, "learning_rate": 1.2487413293162803e-06, "loss": 0.0328, "step": 445 }, { "epoch": 0.10147895335608646, "grad_norm": 2.974661390553103, "learning_rate": 1.2487356551582421e-06, "loss": 0.1436, "step": 446 }, { "epoch": 0.10170648464163823, "grad_norm": 2.6800805691171843, "learning_rate": 1.2487299682521893e-06, "loss": 0.1383, "step": 447 }, { "epoch": 0.10193401592719, "grad_norm": 2.2359167995553233, "learning_rate": 1.2487242685982384e-06, "loss": 0.0485, "step": 448 }, { "epoch": 0.10216154721274175, "grad_norm": 2.818745463327436, "learning_rate": 1.2487185561965057e-06, "loss": 0.0446, "step": 449 }, { "epoch": 0.10238907849829351, "grad_norm": 0.6315900981706886, "learning_rate": 1.248712831047108e-06, "loss": 0.0186, "step": 450 }, { "epoch": 0.10261660978384528, "grad_norm": 1.9674239723403535, "learning_rate": 1.2487070931501624e-06, "loss": 0.0673, "step": 451 }, { "epoch": 0.10284414106939704, "grad_norm": 1.6885325551942292, "learning_rate": 1.2487013425057858e-06, "loss": 0.0767, "step": 452 }, { "epoch": 0.10307167235494881, "grad_norm": 1.891051810210748, "learning_rate": 1.2486955791140964e-06, "loss": 0.063, "step": 453 }, { "epoch": 0.10329920364050058, "grad_norm": 1.1912814980714257, "learning_rate": 1.2486898029752113e-06, "loss": 0.0493, "step": 454 }, { "epoch": 0.10352673492605233, "grad_norm": 0.8883039975316582, "learning_rate": 1.248684014089249e-06, "loss": 0.0277, "step": 455 }, { "epoch": 0.1037542662116041, "grad_norm": 1.306066876530539, "learning_rate": 1.2486782124563277e-06, "loss": 0.0544, "step": 456 }, { "epoch": 0.10398179749715586, "grad_norm": 1.5420420577450527, "learning_rate": 1.2486723980765659e-06, "loss": 0.077, "step": 457 }, { "epoch": 0.10420932878270762, "grad_norm": 2.0367683349755823, "learning_rate": 1.2486665709500826e-06, "loss": 0.0649, "step": 458 }, { "epoch": 0.10443686006825939, "grad_norm": 1.1542715096556455, "learning_rate": 1.2486607310769965e-06, "loss": 0.0586, "step": 459 }, { "epoch": 0.10466439135381114, "grad_norm": 2.061046992152943, "learning_rate": 1.2486548784574275e-06, "loss": 0.0491, "step": 460 }, { "epoch": 0.10489192263936291, "grad_norm": 1.0629825041281826, "learning_rate": 1.2486490130914948e-06, "loss": 0.0445, "step": 461 }, { "epoch": 0.10511945392491467, "grad_norm": 1.415699131985968, "learning_rate": 1.2486431349793185e-06, "loss": 0.0679, "step": 462 }, { "epoch": 0.10534698521046644, "grad_norm": 1.8548752695365796, "learning_rate": 1.2486372441210188e-06, "loss": 0.0514, "step": 463 }, { "epoch": 0.1055745164960182, "grad_norm": 1.1301759724239804, "learning_rate": 1.248631340516716e-06, "loss": 0.0381, "step": 464 }, { "epoch": 0.10580204778156997, "grad_norm": 1.408994037607189, "learning_rate": 1.2486254241665302e-06, "loss": 0.0692, "step": 465 }, { "epoch": 0.10602957906712172, "grad_norm": 1.6631340659469183, "learning_rate": 1.2486194950705831e-06, "loss": 0.0454, "step": 466 }, { "epoch": 0.10625711035267349, "grad_norm": 4.446207788916147, "learning_rate": 1.248613553228996e-06, "loss": 0.051, "step": 467 }, { "epoch": 0.10648464163822526, "grad_norm": 2.558672760104107, "learning_rate": 1.2486075986418896e-06, "loss": 0.0581, "step": 468 }, { "epoch": 0.10671217292377702, "grad_norm": 1.1293994671100194, "learning_rate": 1.248601631309386e-06, "loss": 0.0601, "step": 469 }, { "epoch": 0.10693970420932879, "grad_norm": 0.9858665143030374, "learning_rate": 1.2485956512316072e-06, "loss": 0.042, "step": 470 }, { "epoch": 0.10716723549488055, "grad_norm": 1.0238223731711866, "learning_rate": 1.2485896584086754e-06, "loss": 0.0419, "step": 471 }, { "epoch": 0.1073947667804323, "grad_norm": 1.4636163379221103, "learning_rate": 1.248583652840713e-06, "loss": 0.0763, "step": 472 }, { "epoch": 0.10762229806598407, "grad_norm": 2.3052142467408534, "learning_rate": 1.2485776345278427e-06, "loss": 0.0577, "step": 473 }, { "epoch": 0.10784982935153584, "grad_norm": 1.797428495175526, "learning_rate": 1.2485716034701876e-06, "loss": 0.0664, "step": 474 }, { "epoch": 0.1080773606370876, "grad_norm": 0.9502391190499007, "learning_rate": 1.2485655596678712e-06, "loss": 0.032, "step": 475 }, { "epoch": 0.10830489192263937, "grad_norm": 2.0597597691340033, "learning_rate": 1.2485595031210164e-06, "loss": 0.0693, "step": 476 }, { "epoch": 0.10853242320819112, "grad_norm": 1.6422248848311762, "learning_rate": 1.2485534338297475e-06, "loss": 0.036, "step": 477 }, { "epoch": 0.10875995449374289, "grad_norm": 1.0205877451233056, "learning_rate": 1.2485473517941884e-06, "loss": 0.036, "step": 478 }, { "epoch": 0.10898748577929465, "grad_norm": 1.352496943979393, "learning_rate": 1.2485412570144633e-06, "loss": 0.0619, "step": 479 }, { "epoch": 0.10921501706484642, "grad_norm": 1.1460532851585836, "learning_rate": 1.2485351494906969e-06, "loss": 0.0457, "step": 480 }, { "epoch": 0.10944254835039818, "grad_norm": 2.136834742910293, "learning_rate": 1.2485290292230142e-06, "loss": 0.1229, "step": 481 }, { "epoch": 0.10967007963594995, "grad_norm": 1.2780826832089733, "learning_rate": 1.24852289621154e-06, "loss": 0.0295, "step": 482 }, { "epoch": 0.1098976109215017, "grad_norm": 1.5165261909270789, "learning_rate": 1.2485167504563995e-06, "loss": 0.0605, "step": 483 }, { "epoch": 0.11012514220705347, "grad_norm": 2.5598525552258153, "learning_rate": 1.2485105919577187e-06, "loss": 0.1343, "step": 484 }, { "epoch": 0.11035267349260523, "grad_norm": 1.2948078456387253, "learning_rate": 1.2485044207156233e-06, "loss": 0.0462, "step": 485 }, { "epoch": 0.110580204778157, "grad_norm": 2.2282842690273674, "learning_rate": 1.2484982367302395e-06, "loss": 0.0906, "step": 486 }, { "epoch": 0.11080773606370876, "grad_norm": 1.58668388631447, "learning_rate": 1.2484920400016936e-06, "loss": 0.0754, "step": 487 }, { "epoch": 0.11103526734926053, "grad_norm": 1.8486367636195544, "learning_rate": 1.2484858305301122e-06, "loss": 0.0858, "step": 488 }, { "epoch": 0.11126279863481228, "grad_norm": 1.2490242519124817, "learning_rate": 1.2484796083156222e-06, "loss": 0.0412, "step": 489 }, { "epoch": 0.11149032992036405, "grad_norm": 2.4293013886013024, "learning_rate": 1.2484733733583511e-06, "loss": 0.0347, "step": 490 }, { "epoch": 0.11171786120591581, "grad_norm": 1.3895430378196065, "learning_rate": 1.248467125658426e-06, "loss": 0.0508, "step": 491 }, { "epoch": 0.11194539249146758, "grad_norm": 1.0989498931762587, "learning_rate": 1.2484608652159746e-06, "loss": 0.0376, "step": 492 }, { "epoch": 0.11217292377701935, "grad_norm": 1.2816896914965858, "learning_rate": 1.248454592031125e-06, "loss": 0.0325, "step": 493 }, { "epoch": 0.1124004550625711, "grad_norm": 1.3834104899544186, "learning_rate": 1.2484483061040054e-06, "loss": 0.0502, "step": 494 }, { "epoch": 0.11262798634812286, "grad_norm": 1.8438437167351667, "learning_rate": 1.2484420074347441e-06, "loss": 0.0675, "step": 495 }, { "epoch": 0.11285551763367463, "grad_norm": 1.3847331523036337, "learning_rate": 1.24843569602347e-06, "loss": 0.0708, "step": 496 }, { "epoch": 0.1130830489192264, "grad_norm": 1.4695794221134664, "learning_rate": 1.2484293718703119e-06, "loss": 0.0677, "step": 497 }, { "epoch": 0.11331058020477816, "grad_norm": 1.4197704649286171, "learning_rate": 1.2484230349753994e-06, "loss": 0.0275, "step": 498 }, { "epoch": 0.11353811149032993, "grad_norm": 1.9758484599559376, "learning_rate": 1.2484166853388617e-06, "loss": 0.0846, "step": 499 }, { "epoch": 0.11376564277588168, "grad_norm": 1.4480733215282273, "learning_rate": 1.2484103229608288e-06, "loss": 0.0302, "step": 500 }, { "epoch": 0.11399317406143344, "grad_norm": 1.6803783059061204, "learning_rate": 1.2484039478414305e-06, "loss": 0.033, "step": 501 }, { "epoch": 0.11422070534698521, "grad_norm": 1.5093539895040036, "learning_rate": 1.2483975599807972e-06, "loss": 0.0592, "step": 502 }, { "epoch": 0.11444823663253698, "grad_norm": 1.2912092446277637, "learning_rate": 1.2483911593790595e-06, "loss": 0.0341, "step": 503 }, { "epoch": 0.11467576791808874, "grad_norm": 4.053757173255741, "learning_rate": 1.2483847460363482e-06, "loss": 0.08, "step": 504 }, { "epoch": 0.1149032992036405, "grad_norm": 1.8561008975587792, "learning_rate": 1.2483783199527943e-06, "loss": 0.0417, "step": 505 }, { "epoch": 0.11513083048919226, "grad_norm": 1.7954101019883197, "learning_rate": 1.2483718811285296e-06, "loss": 0.098, "step": 506 }, { "epoch": 0.11535836177474403, "grad_norm": 0.90376517227131, "learning_rate": 1.2483654295636848e-06, "loss": 0.0204, "step": 507 }, { "epoch": 0.11558589306029579, "grad_norm": 1.0402238740609373, "learning_rate": 1.2483589652583924e-06, "loss": 0.0393, "step": 508 }, { "epoch": 0.11581342434584756, "grad_norm": 1.2739463431329932, "learning_rate": 1.2483524882127846e-06, "loss": 0.0361, "step": 509 }, { "epoch": 0.11604095563139932, "grad_norm": 1.8542368928683455, "learning_rate": 1.2483459984269933e-06, "loss": 0.062, "step": 510 }, { "epoch": 0.11626848691695107, "grad_norm": 1.4901585211244064, "learning_rate": 1.2483394959011514e-06, "loss": 0.0362, "step": 511 }, { "epoch": 0.11649601820250284, "grad_norm": 1.655677974364157, "learning_rate": 1.248332980635392e-06, "loss": 0.068, "step": 512 }, { "epoch": 0.1167235494880546, "grad_norm": 1.40135662829084, "learning_rate": 1.2483264526298478e-06, "loss": 0.0467, "step": 513 }, { "epoch": 0.11695108077360637, "grad_norm": 0.9801745008839565, "learning_rate": 1.2483199118846525e-06, "loss": 0.0547, "step": 514 }, { "epoch": 0.11717861205915814, "grad_norm": 0.7956319000441687, "learning_rate": 1.2483133583999399e-06, "loss": 0.0292, "step": 515 }, { "epoch": 0.1174061433447099, "grad_norm": 1.5822262084465146, "learning_rate": 1.2483067921758439e-06, "loss": 0.0442, "step": 516 }, { "epoch": 0.11763367463026166, "grad_norm": 1.788875375154368, "learning_rate": 1.2483002132124983e-06, "loss": 0.1027, "step": 517 }, { "epoch": 0.11786120591581342, "grad_norm": 1.12691319203208, "learning_rate": 1.2482936215100382e-06, "loss": 0.0392, "step": 518 }, { "epoch": 0.11808873720136519, "grad_norm": 1.4278197951300633, "learning_rate": 1.2482870170685978e-06, "loss": 0.0309, "step": 519 }, { "epoch": 0.11831626848691695, "grad_norm": 2.3408918893379176, "learning_rate": 1.2482803998883122e-06, "loss": 0.0554, "step": 520 }, { "epoch": 0.11854379977246872, "grad_norm": 1.308272654811059, "learning_rate": 1.2482737699693168e-06, "loss": 0.0386, "step": 521 }, { "epoch": 0.11877133105802047, "grad_norm": 1.4643677550086747, "learning_rate": 1.248267127311747e-06, "loss": 0.0517, "step": 522 }, { "epoch": 0.11899886234357224, "grad_norm": 1.9375531511223079, "learning_rate": 1.2482604719157386e-06, "loss": 0.0547, "step": 523 }, { "epoch": 0.119226393629124, "grad_norm": 1.4368931736724269, "learning_rate": 1.2482538037814277e-06, "loss": 0.0594, "step": 524 }, { "epoch": 0.11945392491467577, "grad_norm": 2.3983593700848864, "learning_rate": 1.2482471229089502e-06, "loss": 0.0479, "step": 525 }, { "epoch": 0.11968145620022753, "grad_norm": 1.5406589025382933, "learning_rate": 1.2482404292984431e-06, "loss": 0.0468, "step": 526 }, { "epoch": 0.1199089874857793, "grad_norm": 2.7569107151011254, "learning_rate": 1.248233722950043e-06, "loss": 0.1653, "step": 527 }, { "epoch": 0.12013651877133105, "grad_norm": 1.2498018017145271, "learning_rate": 1.2482270038638872e-06, "loss": 0.0376, "step": 528 }, { "epoch": 0.12036405005688282, "grad_norm": 1.663550002347151, "learning_rate": 1.2482202720401128e-06, "loss": 0.0336, "step": 529 }, { "epoch": 0.12059158134243458, "grad_norm": 1.302350240596974, "learning_rate": 1.248213527478857e-06, "loss": 0.058, "step": 530 }, { "epoch": 0.12081911262798635, "grad_norm": 1.6402862563370133, "learning_rate": 1.2482067701802583e-06, "loss": 0.0826, "step": 531 }, { "epoch": 0.12104664391353812, "grad_norm": 1.1268914835455415, "learning_rate": 1.2482000001444547e-06, "loss": 0.0524, "step": 532 }, { "epoch": 0.12127417519908988, "grad_norm": 1.877525140915591, "learning_rate": 1.2481932173715845e-06, "loss": 0.0539, "step": 533 }, { "epoch": 0.12150170648464163, "grad_norm": 1.928222388008525, "learning_rate": 1.2481864218617859e-06, "loss": 0.1134, "step": 534 }, { "epoch": 0.1217292377701934, "grad_norm": 1.2293579919391775, "learning_rate": 1.2481796136151984e-06, "loss": 0.0516, "step": 535 }, { "epoch": 0.12195676905574517, "grad_norm": 1.5890038701436435, "learning_rate": 1.2481727926319609e-06, "loss": 0.0922, "step": 536 }, { "epoch": 0.12218430034129693, "grad_norm": 1.850040007143475, "learning_rate": 1.2481659589122127e-06, "loss": 0.075, "step": 537 }, { "epoch": 0.1224118316268487, "grad_norm": 2.5725083852053747, "learning_rate": 1.2481591124560934e-06, "loss": 0.0891, "step": 538 }, { "epoch": 0.12263936291240045, "grad_norm": 1.768076465759339, "learning_rate": 1.2481522532637435e-06, "loss": 0.0533, "step": 539 }, { "epoch": 0.12286689419795221, "grad_norm": 0.9070123837857468, "learning_rate": 1.2481453813353026e-06, "loss": 0.0294, "step": 540 }, { "epoch": 0.12309442548350398, "grad_norm": 1.625916805043753, "learning_rate": 1.2481384966709116e-06, "loss": 0.0323, "step": 541 }, { "epoch": 0.12332195676905575, "grad_norm": 2.291411200127928, "learning_rate": 1.2481315992707104e-06, "loss": 0.0964, "step": 542 }, { "epoch": 0.12354948805460751, "grad_norm": 1.7452732363406978, "learning_rate": 1.248124689134841e-06, "loss": 0.0555, "step": 543 }, { "epoch": 0.12377701934015928, "grad_norm": 1.4048824029212816, "learning_rate": 1.2481177662634438e-06, "loss": 0.0551, "step": 544 }, { "epoch": 0.12400455062571103, "grad_norm": 6.9359786939510695, "learning_rate": 1.2481108306566609e-06, "loss": 0.0746, "step": 545 }, { "epoch": 0.1242320819112628, "grad_norm": 1.2916063155011999, "learning_rate": 1.2481038823146338e-06, "loss": 0.0245, "step": 546 }, { "epoch": 0.12445961319681456, "grad_norm": 1.9432907622995745, "learning_rate": 1.2480969212375043e-06, "loss": 0.1119, "step": 547 }, { "epoch": 0.12468714448236633, "grad_norm": 2.9617568197379347, "learning_rate": 1.2480899474254151e-06, "loss": 0.0878, "step": 548 }, { "epoch": 0.12491467576791809, "grad_norm": 0.9954987625211155, "learning_rate": 1.2480829608785085e-06, "loss": 0.0397, "step": 549 }, { "epoch": 0.12514220705346984, "grad_norm": 1.2186721476872615, "learning_rate": 1.2480759615969273e-06, "loss": 0.0462, "step": 550 }, { "epoch": 0.12536973833902162, "grad_norm": 1.1723077588981092, "learning_rate": 1.2480689495808144e-06, "loss": 0.0646, "step": 551 }, { "epoch": 0.12559726962457338, "grad_norm": 1.4338215761686757, "learning_rate": 1.2480619248303133e-06, "loss": 0.0806, "step": 552 }, { "epoch": 0.12582480091012513, "grad_norm": 1.841000108666981, "learning_rate": 1.2480548873455675e-06, "loss": 0.0827, "step": 553 }, { "epoch": 0.1260523321956769, "grad_norm": 2.5765107503856144, "learning_rate": 1.248047837126721e-06, "loss": 0.1326, "step": 554 }, { "epoch": 0.12627986348122866, "grad_norm": 1.3867433794433537, "learning_rate": 1.248040774173918e-06, "loss": 0.0588, "step": 555 }, { "epoch": 0.12650739476678044, "grad_norm": 1.3761837680071665, "learning_rate": 1.248033698487302e-06, "loss": 0.0495, "step": 556 }, { "epoch": 0.1267349260523322, "grad_norm": 1.1704965217659915, "learning_rate": 1.2480266100670189e-06, "loss": 0.0467, "step": 557 }, { "epoch": 0.12696245733788397, "grad_norm": 2.106768640375489, "learning_rate": 1.2480195089132125e-06, "loss": 0.0426, "step": 558 }, { "epoch": 0.12718998862343572, "grad_norm": 1.1575398708268563, "learning_rate": 1.2480123950260284e-06, "loss": 0.0413, "step": 559 }, { "epoch": 0.12741751990898748, "grad_norm": 9.831393771519632, "learning_rate": 1.248005268405612e-06, "loss": 0.0434, "step": 560 }, { "epoch": 0.12764505119453926, "grad_norm": 1.0820393760670333, "learning_rate": 1.2479981290521087e-06, "loss": 0.034, "step": 561 }, { "epoch": 0.127872582480091, "grad_norm": 2.206902342168789, "learning_rate": 1.2479909769656648e-06, "loss": 0.073, "step": 562 }, { "epoch": 0.1281001137656428, "grad_norm": 1.5027865833496339, "learning_rate": 1.2479838121464263e-06, "loss": 0.0744, "step": 563 }, { "epoch": 0.12832764505119454, "grad_norm": 2.001324966678472, "learning_rate": 1.2479766345945395e-06, "loss": 0.0457, "step": 564 }, { "epoch": 0.1285551763367463, "grad_norm": 0.9498156004808115, "learning_rate": 1.2479694443101513e-06, "loss": 0.0442, "step": 565 }, { "epoch": 0.12878270762229807, "grad_norm": 1.427402300347758, "learning_rate": 1.2479622412934087e-06, "loss": 0.0774, "step": 566 }, { "epoch": 0.12901023890784982, "grad_norm": 1.338474573791641, "learning_rate": 1.2479550255444586e-06, "loss": 0.0498, "step": 567 }, { "epoch": 0.1292377701934016, "grad_norm": 1.0332062154173083, "learning_rate": 1.2479477970634487e-06, "loss": 0.0575, "step": 568 }, { "epoch": 0.12946530147895335, "grad_norm": 2.342085432201913, "learning_rate": 1.2479405558505267e-06, "loss": 0.0776, "step": 569 }, { "epoch": 0.1296928327645051, "grad_norm": 1.7614392055133243, "learning_rate": 1.247933301905841e-06, "loss": 0.0626, "step": 570 }, { "epoch": 0.12992036405005689, "grad_norm": 1.401551595719724, "learning_rate": 1.2479260352295388e-06, "loss": 0.0354, "step": 571 }, { "epoch": 0.13014789533560864, "grad_norm": 1.4755845320165697, "learning_rate": 1.2479187558217697e-06, "loss": 0.0407, "step": 572 }, { "epoch": 0.13037542662116042, "grad_norm": 0.7585780841566192, "learning_rate": 1.247911463682682e-06, "loss": 0.0376, "step": 573 }, { "epoch": 0.13060295790671217, "grad_norm": 2.2124015942087496, "learning_rate": 1.2479041588124247e-06, "loss": 0.0537, "step": 574 }, { "epoch": 0.13083048919226395, "grad_norm": 1.0766166316852028, "learning_rate": 1.2478968412111471e-06, "loss": 0.0292, "step": 575 }, { "epoch": 0.1310580204778157, "grad_norm": 0.995096761634457, "learning_rate": 1.247889510878999e-06, "loss": 0.0523, "step": 576 }, { "epoch": 0.13128555176336745, "grad_norm": 1.6016154355497043, "learning_rate": 1.24788216781613e-06, "loss": 0.0634, "step": 577 }, { "epoch": 0.13151308304891923, "grad_norm": 1.1448425600800458, "learning_rate": 1.2478748120226902e-06, "loss": 0.0449, "step": 578 }, { "epoch": 0.13174061433447098, "grad_norm": 1.1092744505453933, "learning_rate": 1.2478674434988299e-06, "loss": 0.0377, "step": 579 }, { "epoch": 0.13196814562002276, "grad_norm": 0.8431904554396206, "learning_rate": 1.2478600622447001e-06, "loss": 0.0404, "step": 580 }, { "epoch": 0.13219567690557452, "grad_norm": 1.1168682097019669, "learning_rate": 1.2478526682604512e-06, "loss": 0.0324, "step": 581 }, { "epoch": 0.13242320819112627, "grad_norm": 1.3377796205096442, "learning_rate": 1.2478452615462345e-06, "loss": 0.0663, "step": 582 }, { "epoch": 0.13265073947667805, "grad_norm": 3.2228218888338667, "learning_rate": 1.247837842102201e-06, "loss": 0.1264, "step": 583 }, { "epoch": 0.1328782707622298, "grad_norm": 1.5298131963208237, "learning_rate": 1.2478304099285031e-06, "loss": 0.0525, "step": 584 }, { "epoch": 0.13310580204778158, "grad_norm": 1.1458963041110781, "learning_rate": 1.2478229650252921e-06, "loss": 0.0464, "step": 585 }, { "epoch": 0.13333333333333333, "grad_norm": 1.7278138919558592, "learning_rate": 1.2478155073927204e-06, "loss": 0.0556, "step": 586 }, { "epoch": 0.13356086461888508, "grad_norm": 1.2475569782092948, "learning_rate": 1.2478080370309404e-06, "loss": 0.0428, "step": 587 }, { "epoch": 0.13378839590443686, "grad_norm": 1.088342211454506, "learning_rate": 1.2478005539401046e-06, "loss": 0.0295, "step": 588 }, { "epoch": 0.13401592718998862, "grad_norm": 1.7587707539606798, "learning_rate": 1.2477930581203663e-06, "loss": 0.0442, "step": 589 }, { "epoch": 0.1342434584755404, "grad_norm": 1.1629708095404674, "learning_rate": 1.2477855495718782e-06, "loss": 0.0548, "step": 590 }, { "epoch": 0.13447098976109215, "grad_norm": 1.551063066789381, "learning_rate": 1.2477780282947942e-06, "loss": 0.037, "step": 591 }, { "epoch": 0.13469852104664393, "grad_norm": 1.0301289542876222, "learning_rate": 1.2477704942892677e-06, "loss": 0.0443, "step": 592 }, { "epoch": 0.13492605233219568, "grad_norm": 3.67842928304707, "learning_rate": 1.2477629475554532e-06, "loss": 0.0909, "step": 593 }, { "epoch": 0.13515358361774743, "grad_norm": 1.6590539267977245, "learning_rate": 1.2477553880935043e-06, "loss": 0.0523, "step": 594 }, { "epoch": 0.1353811149032992, "grad_norm": 1.9972940081293025, "learning_rate": 1.2477478159035758e-06, "loss": 0.0745, "step": 595 }, { "epoch": 0.13560864618885096, "grad_norm": 2.5034292233621693, "learning_rate": 1.2477402309858226e-06, "loss": 0.0731, "step": 596 }, { "epoch": 0.13583617747440274, "grad_norm": 1.7292171438267574, "learning_rate": 1.2477326333403995e-06, "loss": 0.0947, "step": 597 }, { "epoch": 0.1360637087599545, "grad_norm": 1.0893865478354072, "learning_rate": 1.2477250229674618e-06, "loss": 0.036, "step": 598 }, { "epoch": 0.13629124004550625, "grad_norm": 2.0220916101870854, "learning_rate": 1.2477173998671653e-06, "loss": 0.0695, "step": 599 }, { "epoch": 0.13651877133105803, "grad_norm": 1.6663966081636725, "learning_rate": 1.2477097640396655e-06, "loss": 0.0239, "step": 600 }, { "epoch": 0.13674630261660978, "grad_norm": 2.160820824239015, "learning_rate": 1.2477021154851185e-06, "loss": 0.041, "step": 601 }, { "epoch": 0.13697383390216156, "grad_norm": 1.7560963022061893, "learning_rate": 1.2476944542036806e-06, "loss": 0.0646, "step": 602 }, { "epoch": 0.1372013651877133, "grad_norm": 1.0045761668266382, "learning_rate": 1.2476867801955086e-06, "loss": 0.0319, "step": 603 }, { "epoch": 0.13742889647326506, "grad_norm": 1.1722883805586992, "learning_rate": 1.247679093460759e-06, "loss": 0.0421, "step": 604 }, { "epoch": 0.13765642775881684, "grad_norm": 1.5520795851774112, "learning_rate": 1.2476713939995895e-06, "loss": 0.0545, "step": 605 }, { "epoch": 0.1378839590443686, "grad_norm": 1.5341083321869782, "learning_rate": 1.2476636818121568e-06, "loss": 0.0587, "step": 606 }, { "epoch": 0.13811149032992037, "grad_norm": 1.427789001931161, "learning_rate": 1.247655956898619e-06, "loss": 0.0783, "step": 607 }, { "epoch": 0.13833902161547212, "grad_norm": 1.6612253896109521, "learning_rate": 1.2476482192591335e-06, "loss": 0.0454, "step": 608 }, { "epoch": 0.1385665529010239, "grad_norm": 4.06982296073183, "learning_rate": 1.247640468893859e-06, "loss": 0.0518, "step": 609 }, { "epoch": 0.13879408418657566, "grad_norm": 0.9984415328300282, "learning_rate": 1.2476327058029534e-06, "loss": 0.0286, "step": 610 }, { "epoch": 0.1390216154721274, "grad_norm": 1.5505955919902168, "learning_rate": 1.2476249299865757e-06, "loss": 0.0679, "step": 611 }, { "epoch": 0.1392491467576792, "grad_norm": 1.1777984076454868, "learning_rate": 1.2476171414448847e-06, "loss": 0.0333, "step": 612 }, { "epoch": 0.13947667804323094, "grad_norm": 1.5134654111327426, "learning_rate": 1.2476093401780397e-06, "loss": 0.0464, "step": 613 }, { "epoch": 0.13970420932878272, "grad_norm": 8.868555700718426, "learning_rate": 1.2476015261861998e-06, "loss": 0.0588, "step": 614 }, { "epoch": 0.13993174061433447, "grad_norm": 1.07185240181413, "learning_rate": 1.247593699469525e-06, "loss": 0.0508, "step": 615 }, { "epoch": 0.14015927189988622, "grad_norm": 1.0467231615661654, "learning_rate": 1.2475858600281754e-06, "loss": 0.0375, "step": 616 }, { "epoch": 0.140386803185438, "grad_norm": 1.6914112039280367, "learning_rate": 1.247578007862311e-06, "loss": 0.0385, "step": 617 }, { "epoch": 0.14061433447098975, "grad_norm": 1.6162335462442214, "learning_rate": 1.2475701429720923e-06, "loss": 0.0617, "step": 618 }, { "epoch": 0.14084186575654153, "grad_norm": 0.8727030068232594, "learning_rate": 1.24756226535768e-06, "loss": 0.0388, "step": 619 }, { "epoch": 0.1410693970420933, "grad_norm": 1.5566079211643449, "learning_rate": 1.2475543750192352e-06, "loss": 0.094, "step": 620 }, { "epoch": 0.14129692832764504, "grad_norm": 1.1473885930159387, "learning_rate": 1.2475464719569192e-06, "loss": 0.0563, "step": 621 }, { "epoch": 0.14152445961319682, "grad_norm": 1.111358042304426, "learning_rate": 1.2475385561708934e-06, "loss": 0.058, "step": 622 }, { "epoch": 0.14175199089874857, "grad_norm": 0.8961189135726134, "learning_rate": 1.2475306276613194e-06, "loss": 0.0393, "step": 623 }, { "epoch": 0.14197952218430035, "grad_norm": 3.442046327889576, "learning_rate": 1.2475226864283596e-06, "loss": 0.0713, "step": 624 }, { "epoch": 0.1422070534698521, "grad_norm": 4.150666880510046, "learning_rate": 1.2475147324721764e-06, "loss": 0.0442, "step": 625 }, { "epoch": 0.14243458475540388, "grad_norm": 0.9296406182877743, "learning_rate": 1.2475067657929319e-06, "loss": 0.0414, "step": 626 }, { "epoch": 0.14266211604095563, "grad_norm": 1.0829545070929647, "learning_rate": 1.2474987863907894e-06, "loss": 0.0451, "step": 627 }, { "epoch": 0.14288964732650739, "grad_norm": 1.8334147294224943, "learning_rate": 1.2474907942659116e-06, "loss": 0.0771, "step": 628 }, { "epoch": 0.14311717861205916, "grad_norm": 1.7815672754482244, "learning_rate": 1.247482789418462e-06, "loss": 0.0897, "step": 629 }, { "epoch": 0.14334470989761092, "grad_norm": 1.3733903816478978, "learning_rate": 1.2474747718486044e-06, "loss": 0.0358, "step": 630 }, { "epoch": 0.1435722411831627, "grad_norm": 2.28709865860884, "learning_rate": 1.2474667415565022e-06, "loss": 0.0652, "step": 631 }, { "epoch": 0.14379977246871445, "grad_norm": 1.231780324099585, "learning_rate": 1.24745869854232e-06, "loss": 0.0425, "step": 632 }, { "epoch": 0.1440273037542662, "grad_norm": 1.335370738840473, "learning_rate": 1.2474506428062219e-06, "loss": 0.0422, "step": 633 }, { "epoch": 0.14425483503981798, "grad_norm": 2.966338619611093, "learning_rate": 1.2474425743483726e-06, "loss": 0.1214, "step": 634 }, { "epoch": 0.14448236632536973, "grad_norm": 0.8989763932220884, "learning_rate": 1.2474344931689371e-06, "loss": 0.0341, "step": 635 }, { "epoch": 0.1447098976109215, "grad_norm": 0.937533756927628, "learning_rate": 1.2474263992680805e-06, "loss": 0.0394, "step": 636 }, { "epoch": 0.14493742889647326, "grad_norm": 1.8912948524392619, "learning_rate": 1.247418292645968e-06, "loss": 0.058, "step": 637 }, { "epoch": 0.14516496018202502, "grad_norm": 2.6567132856925877, "learning_rate": 1.2474101733027659e-06, "loss": 0.1061, "step": 638 }, { "epoch": 0.1453924914675768, "grad_norm": 1.0544164694645533, "learning_rate": 1.2474020412386395e-06, "loss": 0.0327, "step": 639 }, { "epoch": 0.14562002275312855, "grad_norm": 0.8862493084977525, "learning_rate": 1.2473938964537551e-06, "loss": 0.0337, "step": 640 }, { "epoch": 0.14584755403868033, "grad_norm": 1.1660786205758522, "learning_rate": 1.2473857389482797e-06, "loss": 0.0379, "step": 641 }, { "epoch": 0.14607508532423208, "grad_norm": 1.275063352619887, "learning_rate": 1.2473775687223794e-06, "loss": 0.0677, "step": 642 }, { "epoch": 0.14630261660978386, "grad_norm": 1.033368353907448, "learning_rate": 1.2473693857762215e-06, "loss": 0.0499, "step": 643 }, { "epoch": 0.1465301478953356, "grad_norm": 0.8176955754401027, "learning_rate": 1.247361190109973e-06, "loss": 0.028, "step": 644 }, { "epoch": 0.14675767918088736, "grad_norm": 1.5868301680057004, "learning_rate": 1.2473529817238016e-06, "loss": 0.0758, "step": 645 }, { "epoch": 0.14698521046643914, "grad_norm": 1.1116607935229232, "learning_rate": 1.2473447606178754e-06, "loss": 0.0464, "step": 646 }, { "epoch": 0.1472127417519909, "grad_norm": 1.3737040045295303, "learning_rate": 1.2473365267923617e-06, "loss": 0.038, "step": 647 }, { "epoch": 0.14744027303754267, "grad_norm": 1.6851756106039961, "learning_rate": 1.2473282802474293e-06, "loss": 0.0521, "step": 648 }, { "epoch": 0.14766780432309443, "grad_norm": 1.067298154630741, "learning_rate": 1.2473200209832465e-06, "loss": 0.0339, "step": 649 }, { "epoch": 0.14789533560864618, "grad_norm": 3.389712283735002, "learning_rate": 1.2473117489999823e-06, "loss": 0.1032, "step": 650 }, { "epoch": 0.14812286689419796, "grad_norm": 1.6568457682006943, "learning_rate": 1.2473034642978057e-06, "loss": 0.0619, "step": 651 }, { "epoch": 0.1483503981797497, "grad_norm": 1.490856549646787, "learning_rate": 1.247295166876886e-06, "loss": 0.0516, "step": 652 }, { "epoch": 0.1485779294653015, "grad_norm": 1.1851614081448008, "learning_rate": 1.2472868567373924e-06, "loss": 0.0431, "step": 653 }, { "epoch": 0.14880546075085324, "grad_norm": 1.644166757760147, "learning_rate": 1.2472785338794953e-06, "loss": 0.082, "step": 654 }, { "epoch": 0.149032992036405, "grad_norm": 1.556466054889254, "learning_rate": 1.247270198303365e-06, "loss": 0.0573, "step": 655 }, { "epoch": 0.14926052332195677, "grad_norm": 2.233896039541245, "learning_rate": 1.247261850009171e-06, "loss": 0.1073, "step": 656 }, { "epoch": 0.14948805460750852, "grad_norm": 1.5970780316095814, "learning_rate": 1.2472534889970848e-06, "loss": 0.0741, "step": 657 }, { "epoch": 0.1497155858930603, "grad_norm": 1.2400037573314944, "learning_rate": 1.2472451152672766e-06, "loss": 0.0483, "step": 658 }, { "epoch": 0.14994311717861206, "grad_norm": 0.7241628580276863, "learning_rate": 1.2472367288199177e-06, "loss": 0.0294, "step": 659 }, { "epoch": 0.15017064846416384, "grad_norm": 1.7234352219783362, "learning_rate": 1.2472283296551798e-06, "loss": 0.0462, "step": 660 }, { "epoch": 0.1503981797497156, "grad_norm": 1.4734355231360154, "learning_rate": 1.2472199177732346e-06, "loss": 0.0582, "step": 661 }, { "epoch": 0.15062571103526734, "grad_norm": 2.1378423903625317, "learning_rate": 1.2472114931742537e-06, "loss": 0.0584, "step": 662 }, { "epoch": 0.15085324232081912, "grad_norm": 10.24253063417928, "learning_rate": 1.2472030558584093e-06, "loss": 0.0477, "step": 663 }, { "epoch": 0.15108077360637087, "grad_norm": 1.8691143785999473, "learning_rate": 1.2471946058258742e-06, "loss": 0.0739, "step": 664 }, { "epoch": 0.15130830489192265, "grad_norm": 2.241077277544209, "learning_rate": 1.2471861430768205e-06, "loss": 0.0672, "step": 665 }, { "epoch": 0.1515358361774744, "grad_norm": 2.2720454211658385, "learning_rate": 1.2471776676114217e-06, "loss": 0.0744, "step": 666 }, { "epoch": 0.15176336746302616, "grad_norm": 1.5599655593025843, "learning_rate": 1.2471691794298508e-06, "loss": 0.0708, "step": 667 }, { "epoch": 0.15199089874857794, "grad_norm": 1.2105729421418516, "learning_rate": 1.2471606785322814e-06, "loss": 0.0408, "step": 668 }, { "epoch": 0.1522184300341297, "grad_norm": 1.2672422777930856, "learning_rate": 1.247152164918887e-06, "loss": 0.0406, "step": 669 }, { "epoch": 0.15244596131968147, "grad_norm": 1.4429003913192175, "learning_rate": 1.247143638589842e-06, "loss": 0.0537, "step": 670 }, { "epoch": 0.15267349260523322, "grad_norm": 2.4501715529962667, "learning_rate": 1.2471350995453203e-06, "loss": 0.0658, "step": 671 }, { "epoch": 0.15290102389078497, "grad_norm": 1.48398143402129, "learning_rate": 1.2471265477854966e-06, "loss": 0.0556, "step": 672 }, { "epoch": 0.15312855517633675, "grad_norm": 2.1744278885096278, "learning_rate": 1.2471179833105454e-06, "loss": 0.0404, "step": 673 }, { "epoch": 0.1533560864618885, "grad_norm": 6.901371774679421, "learning_rate": 1.2471094061206422e-06, "loss": 0.0869, "step": 674 }, { "epoch": 0.15358361774744028, "grad_norm": 1.1601432497447364, "learning_rate": 1.247100816215962e-06, "loss": 0.0379, "step": 675 }, { "epoch": 0.15381114903299203, "grad_norm": 1.2820824697846318, "learning_rate": 1.2470922135966806e-06, "loss": 0.032, "step": 676 }, { "epoch": 0.1540386803185438, "grad_norm": 1.2705594517037477, "learning_rate": 1.2470835982629736e-06, "loss": 0.0494, "step": 677 }, { "epoch": 0.15426621160409557, "grad_norm": 1.5982674187017603, "learning_rate": 1.247074970215017e-06, "loss": 0.0495, "step": 678 }, { "epoch": 0.15449374288964732, "grad_norm": 1.7109186334283122, "learning_rate": 1.2470663294529873e-06, "loss": 0.0414, "step": 679 }, { "epoch": 0.1547212741751991, "grad_norm": 1.0715248913545208, "learning_rate": 1.2470576759770612e-06, "loss": 0.0325, "step": 680 }, { "epoch": 0.15494880546075085, "grad_norm": 1.824156765255701, "learning_rate": 1.2470490097874155e-06, "loss": 0.0466, "step": 681 }, { "epoch": 0.15517633674630263, "grad_norm": 2.4373458213618897, "learning_rate": 1.247040330884227e-06, "loss": 0.0539, "step": 682 }, { "epoch": 0.15540386803185438, "grad_norm": 1.0065301428430162, "learning_rate": 1.2470316392676738e-06, "loss": 0.0525, "step": 683 }, { "epoch": 0.15563139931740613, "grad_norm": 1.5815427583099921, "learning_rate": 1.2470229349379326e-06, "loss": 0.0519, "step": 684 }, { "epoch": 0.1558589306029579, "grad_norm": 1.9652039642446397, "learning_rate": 1.2470142178951822e-06, "loss": 0.0768, "step": 685 }, { "epoch": 0.15608646188850966, "grad_norm": 1.6149386041596687, "learning_rate": 1.2470054881396002e-06, "loss": 0.0832, "step": 686 }, { "epoch": 0.15631399317406144, "grad_norm": 1.1987854832655134, "learning_rate": 1.246996745671365e-06, "loss": 0.0659, "step": 687 }, { "epoch": 0.1565415244596132, "grad_norm": 1.3921581311290316, "learning_rate": 1.2469879904906556e-06, "loss": 0.0477, "step": 688 }, { "epoch": 0.15676905574516495, "grad_norm": 1.5291818763479217, "learning_rate": 1.2469792225976507e-06, "loss": 0.0597, "step": 689 }, { "epoch": 0.15699658703071673, "grad_norm": 1.1978579028524097, "learning_rate": 1.2469704419925296e-06, "loss": 0.0551, "step": 690 }, { "epoch": 0.15722411831626848, "grad_norm": 2.2562422781402827, "learning_rate": 1.246961648675472e-06, "loss": 0.0709, "step": 691 }, { "epoch": 0.15745164960182026, "grad_norm": 1.947406282057691, "learning_rate": 1.246952842646657e-06, "loss": 0.115, "step": 692 }, { "epoch": 0.157679180887372, "grad_norm": 1.2450058586298152, "learning_rate": 1.2469440239062653e-06, "loss": 0.0499, "step": 693 }, { "epoch": 0.15790671217292376, "grad_norm": 1.8197388877884058, "learning_rate": 1.2469351924544766e-06, "loss": 0.0927, "step": 694 }, { "epoch": 0.15813424345847554, "grad_norm": 1.904635209687238, "learning_rate": 1.2469263482914716e-06, "loss": 0.0814, "step": 695 }, { "epoch": 0.1583617747440273, "grad_norm": 1.4011214554570595, "learning_rate": 1.246917491417431e-06, "loss": 0.0585, "step": 696 }, { "epoch": 0.15858930602957907, "grad_norm": 1.9445852493644915, "learning_rate": 1.246908621832536e-06, "loss": 0.1005, "step": 697 }, { "epoch": 0.15881683731513083, "grad_norm": 1.363848979307397, "learning_rate": 1.2468997395369677e-06, "loss": 0.0717, "step": 698 }, { "epoch": 0.1590443686006826, "grad_norm": 1.3903385726286603, "learning_rate": 1.2468908445309077e-06, "loss": 0.0557, "step": 699 }, { "epoch": 0.15927189988623436, "grad_norm": 0.8725919788890233, "learning_rate": 1.2468819368145376e-06, "loss": 0.0365, "step": 700 }, { "epoch": 0.1594994311717861, "grad_norm": 1.33354116221911, "learning_rate": 1.2468730163880398e-06, "loss": 0.0486, "step": 701 }, { "epoch": 0.1597269624573379, "grad_norm": 4.544549924474667, "learning_rate": 1.2468640832515962e-06, "loss": 0.0302, "step": 702 }, { "epoch": 0.15995449374288964, "grad_norm": 2.9489517825532308, "learning_rate": 1.24685513740539e-06, "loss": 0.0904, "step": 703 }, { "epoch": 0.16018202502844142, "grad_norm": 1.3714756073146785, "learning_rate": 1.2468461788496036e-06, "loss": 0.056, "step": 704 }, { "epoch": 0.16040955631399317, "grad_norm": 1.9329458786445137, "learning_rate": 1.24683720758442e-06, "loss": 0.0848, "step": 705 }, { "epoch": 0.16063708759954493, "grad_norm": 1.0521231330897036, "learning_rate": 1.2468282236100226e-06, "loss": 0.0438, "step": 706 }, { "epoch": 0.1608646188850967, "grad_norm": 1.7554793724577804, "learning_rate": 1.2468192269265955e-06, "loss": 0.1102, "step": 707 }, { "epoch": 0.16109215017064846, "grad_norm": 1.0071708604696206, "learning_rate": 1.246810217534322e-06, "loss": 0.0266, "step": 708 }, { "epoch": 0.16131968145620024, "grad_norm": 1.3334260212411804, "learning_rate": 1.2468011954333864e-06, "loss": 0.0609, "step": 709 }, { "epoch": 0.161547212741752, "grad_norm": 1.1880594144178185, "learning_rate": 1.2467921606239734e-06, "loss": 0.0465, "step": 710 }, { "epoch": 0.16177474402730374, "grad_norm": 2.7036981936596542, "learning_rate": 1.2467831131062672e-06, "loss": 0.0516, "step": 711 }, { "epoch": 0.16200227531285552, "grad_norm": 1.416736687413605, "learning_rate": 1.2467740528804528e-06, "loss": 0.0552, "step": 712 }, { "epoch": 0.16222980659840727, "grad_norm": 1.3251728729578673, "learning_rate": 1.2467649799467156e-06, "loss": 0.0615, "step": 713 }, { "epoch": 0.16245733788395905, "grad_norm": 0.9167577669577527, "learning_rate": 1.246755894305241e-06, "loss": 0.0312, "step": 714 }, { "epoch": 0.1626848691695108, "grad_norm": 1.4735705097484915, "learning_rate": 1.2467467959562143e-06, "loss": 0.0676, "step": 715 }, { "epoch": 0.16291240045506258, "grad_norm": 1.6890201461846701, "learning_rate": 1.2467376848998221e-06, "loss": 0.0597, "step": 716 }, { "epoch": 0.16313993174061434, "grad_norm": 1.4659558773380117, "learning_rate": 1.2467285611362501e-06, "loss": 0.052, "step": 717 }, { "epoch": 0.1633674630261661, "grad_norm": 1.7717923581360278, "learning_rate": 1.2467194246656851e-06, "loss": 0.0563, "step": 718 }, { "epoch": 0.16359499431171787, "grad_norm": 1.0756915461750918, "learning_rate": 1.2467102754883136e-06, "loss": 0.0411, "step": 719 }, { "epoch": 0.16382252559726962, "grad_norm": 1.2683417009383617, "learning_rate": 1.2467011136043228e-06, "loss": 0.0454, "step": 720 }, { "epoch": 0.1640500568828214, "grad_norm": 1.5509079166930662, "learning_rate": 1.2466919390138995e-06, "loss": 0.0745, "step": 721 }, { "epoch": 0.16427758816837315, "grad_norm": 1.799801190377342, "learning_rate": 1.246682751717232e-06, "loss": 0.0531, "step": 722 }, { "epoch": 0.1645051194539249, "grad_norm": 1.734480174185366, "learning_rate": 1.2466735517145074e-06, "loss": 0.0792, "step": 723 }, { "epoch": 0.16473265073947668, "grad_norm": 1.961760175196997, "learning_rate": 1.2466643390059138e-06, "loss": 0.0685, "step": 724 }, { "epoch": 0.16496018202502843, "grad_norm": 0.8466307246278059, "learning_rate": 1.2466551135916398e-06, "loss": 0.0229, "step": 725 }, { "epoch": 0.16518771331058021, "grad_norm": 1.2173112499342529, "learning_rate": 1.2466458754718737e-06, "loss": 0.0632, "step": 726 }, { "epoch": 0.16541524459613197, "grad_norm": 2.194872247572006, "learning_rate": 1.2466366246468045e-06, "loss": 0.0432, "step": 727 }, { "epoch": 0.16564277588168372, "grad_norm": 1.4416201742124337, "learning_rate": 1.246627361116621e-06, "loss": 0.0619, "step": 728 }, { "epoch": 0.1658703071672355, "grad_norm": 1.8390552268184857, "learning_rate": 1.246618084881513e-06, "loss": 0.1148, "step": 729 }, { "epoch": 0.16609783845278725, "grad_norm": 1.7117932742809556, "learning_rate": 1.2466087959416695e-06, "loss": 0.0354, "step": 730 }, { "epoch": 0.16632536973833903, "grad_norm": 2.1473525159969644, "learning_rate": 1.2465994942972805e-06, "loss": 0.0479, "step": 731 }, { "epoch": 0.16655290102389078, "grad_norm": 2.0228728542376846, "learning_rate": 1.2465901799485366e-06, "loss": 0.0819, "step": 732 }, { "epoch": 0.16678043230944256, "grad_norm": 0.7672679660904389, "learning_rate": 1.2465808528956277e-06, "loss": 0.0251, "step": 733 }, { "epoch": 0.1670079635949943, "grad_norm": 6.58955540161242, "learning_rate": 1.2465715131387446e-06, "loss": 0.0608, "step": 734 }, { "epoch": 0.16723549488054607, "grad_norm": 2.0572059988722873, "learning_rate": 1.2465621606780778e-06, "loss": 0.0841, "step": 735 }, { "epoch": 0.16746302616609784, "grad_norm": 1.628470993180389, "learning_rate": 1.2465527955138191e-06, "loss": 0.059, "step": 736 }, { "epoch": 0.1676905574516496, "grad_norm": 1.0886679327758462, "learning_rate": 1.2465434176461596e-06, "loss": 0.0419, "step": 737 }, { "epoch": 0.16791808873720138, "grad_norm": 1.4286021962604534, "learning_rate": 1.2465340270752908e-06, "loss": 0.0481, "step": 738 }, { "epoch": 0.16814562002275313, "grad_norm": 0.8663035560041373, "learning_rate": 1.2465246238014047e-06, "loss": 0.0295, "step": 739 }, { "epoch": 0.16837315130830488, "grad_norm": 1.8501418707953075, "learning_rate": 1.2465152078246936e-06, "loss": 0.0829, "step": 740 }, { "epoch": 0.16860068259385666, "grad_norm": 1.311824892930283, "learning_rate": 1.24650577914535e-06, "loss": 0.0418, "step": 741 }, { "epoch": 0.1688282138794084, "grad_norm": 1.249901682454073, "learning_rate": 1.2464963377635667e-06, "loss": 0.0479, "step": 742 }, { "epoch": 0.1690557451649602, "grad_norm": 1.9982982994055205, "learning_rate": 1.246486883679536e-06, "loss": 0.112, "step": 743 }, { "epoch": 0.16928327645051194, "grad_norm": 1.8891194873050854, "learning_rate": 1.246477416893452e-06, "loss": 0.0489, "step": 744 }, { "epoch": 0.1695108077360637, "grad_norm": 1.3265078136556134, "learning_rate": 1.2464679374055074e-06, "loss": 0.0492, "step": 745 }, { "epoch": 0.16973833902161548, "grad_norm": 1.1620433482725554, "learning_rate": 1.2464584452158968e-06, "loss": 0.0217, "step": 746 }, { "epoch": 0.16996587030716723, "grad_norm": 1.8320124858819533, "learning_rate": 1.2464489403248133e-06, "loss": 0.057, "step": 747 }, { "epoch": 0.170193401592719, "grad_norm": 0.8996339413245623, "learning_rate": 1.246439422732452e-06, "loss": 0.0308, "step": 748 }, { "epoch": 0.17042093287827076, "grad_norm": 1.1199200581464024, "learning_rate": 1.2464298924390066e-06, "loss": 0.0449, "step": 749 }, { "epoch": 0.17064846416382254, "grad_norm": 1.6610024470498899, "learning_rate": 1.2464203494446725e-06, "loss": 0.0667, "step": 750 }, { "epoch": 0.1708759954493743, "grad_norm": 1.1007524784666796, "learning_rate": 1.2464107937496444e-06, "loss": 0.0341, "step": 751 }, { "epoch": 0.17110352673492604, "grad_norm": 1.0836506765705751, "learning_rate": 1.246401225354118e-06, "loss": 0.0397, "step": 752 }, { "epoch": 0.17133105802047782, "grad_norm": 1.320442517854178, "learning_rate": 1.2463916442582883e-06, "loss": 0.0694, "step": 753 }, { "epoch": 0.17155858930602957, "grad_norm": 1.0431861088552385, "learning_rate": 1.2463820504623516e-06, "loss": 0.0285, "step": 754 }, { "epoch": 0.17178612059158135, "grad_norm": 0.9128127625945931, "learning_rate": 1.246372443966504e-06, "loss": 0.0394, "step": 755 }, { "epoch": 0.1720136518771331, "grad_norm": 1.3898389984950463, "learning_rate": 1.246362824770941e-06, "loss": 0.0451, "step": 756 }, { "epoch": 0.17224118316268486, "grad_norm": 0.9752697300280603, "learning_rate": 1.2463531928758605e-06, "loss": 0.0398, "step": 757 }, { "epoch": 0.17246871444823664, "grad_norm": 1.8211753063899252, "learning_rate": 1.2463435482814585e-06, "loss": 0.0947, "step": 758 }, { "epoch": 0.1726962457337884, "grad_norm": 1.5787941280691695, "learning_rate": 1.246333890987932e-06, "loss": 0.1047, "step": 759 }, { "epoch": 0.17292377701934017, "grad_norm": 1.8356072487697, "learning_rate": 1.246324220995479e-06, "loss": 0.0729, "step": 760 }, { "epoch": 0.17315130830489192, "grad_norm": 4.476837441740864, "learning_rate": 1.2463145383042966e-06, "loss": 0.0686, "step": 761 }, { "epoch": 0.17337883959044367, "grad_norm": 252.52198381660855, "learning_rate": 1.2463048429145832e-06, "loss": 1.0972, "step": 762 }, { "epoch": 0.17360637087599545, "grad_norm": 1.6008327751394016, "learning_rate": 1.2462951348265364e-06, "loss": 0.067, "step": 763 }, { "epoch": 0.1738339021615472, "grad_norm": 1.087447761528399, "learning_rate": 1.2462854140403553e-06, "loss": 0.0433, "step": 764 }, { "epoch": 0.17406143344709898, "grad_norm": 2.5989190208674384, "learning_rate": 1.2462756805562378e-06, "loss": 0.1183, "step": 765 }, { "epoch": 0.17428896473265074, "grad_norm": 1.0290974538586322, "learning_rate": 1.2462659343743832e-06, "loss": 0.048, "step": 766 }, { "epoch": 0.17451649601820252, "grad_norm": 1.2617664816926566, "learning_rate": 1.2462561754949908e-06, "loss": 0.0498, "step": 767 }, { "epoch": 0.17474402730375427, "grad_norm": 1.6415043644592435, "learning_rate": 1.2462464039182598e-06, "loss": 0.062, "step": 768 }, { "epoch": 0.17497155858930602, "grad_norm": 1.6673873794852758, "learning_rate": 1.2462366196443903e-06, "loss": 0.0656, "step": 769 }, { "epoch": 0.1751990898748578, "grad_norm": 3.3205898186422718, "learning_rate": 1.246226822673582e-06, "loss": 0.1071, "step": 770 }, { "epoch": 0.17542662116040955, "grad_norm": 1.280610915532866, "learning_rate": 1.2462170130060351e-06, "loss": 0.0318, "step": 771 }, { "epoch": 0.17565415244596133, "grad_norm": 1.9843858659809077, "learning_rate": 1.24620719064195e-06, "loss": 0.0829, "step": 772 }, { "epoch": 0.17588168373151308, "grad_norm": 2.184973202731946, "learning_rate": 1.246197355581528e-06, "loss": 0.0998, "step": 773 }, { "epoch": 0.17610921501706484, "grad_norm": 1.361748313569626, "learning_rate": 1.2461875078249694e-06, "loss": 0.0459, "step": 774 }, { "epoch": 0.17633674630261661, "grad_norm": 1.5219919911848705, "learning_rate": 1.246177647372476e-06, "loss": 0.0539, "step": 775 }, { "epoch": 0.17656427758816837, "grad_norm": 49.0672993271897, "learning_rate": 1.246167774224249e-06, "loss": 0.5261, "step": 776 }, { "epoch": 0.17679180887372015, "grad_norm": 1.5576377429332795, "learning_rate": 1.2461578883804903e-06, "loss": 0.036, "step": 777 }, { "epoch": 0.1770193401592719, "grad_norm": 0.869488380021361, "learning_rate": 1.246147989841402e-06, "loss": 0.0252, "step": 778 }, { "epoch": 0.17724687144482365, "grad_norm": 1.2894143452262044, "learning_rate": 1.2461380786071863e-06, "loss": 0.0401, "step": 779 }, { "epoch": 0.17747440273037543, "grad_norm": 1.4297168158807256, "learning_rate": 1.246128154678046e-06, "loss": 0.0586, "step": 780 }, { "epoch": 0.17770193401592718, "grad_norm": 4.251777831616221, "learning_rate": 1.2461182180541835e-06, "loss": 0.0484, "step": 781 }, { "epoch": 0.17792946530147896, "grad_norm": 1.1899715722609847, "learning_rate": 1.2461082687358022e-06, "loss": 0.0391, "step": 782 }, { "epoch": 0.1781569965870307, "grad_norm": 1.0511863654570817, "learning_rate": 1.2460983067231055e-06, "loss": 0.0424, "step": 783 }, { "epoch": 0.1783845278725825, "grad_norm": 1.6108410955721733, "learning_rate": 1.246088332016297e-06, "loss": 0.0453, "step": 784 }, { "epoch": 0.17861205915813425, "grad_norm": 1.014732446092101, "learning_rate": 1.2460783446155802e-06, "loss": 0.0398, "step": 785 }, { "epoch": 0.178839590443686, "grad_norm": 3.365599666409623, "learning_rate": 1.2460683445211596e-06, "loss": 0.1579, "step": 786 }, { "epoch": 0.17906712172923778, "grad_norm": 2.0790980638653984, "learning_rate": 1.2460583317332395e-06, "loss": 0.1013, "step": 787 }, { "epoch": 0.17929465301478953, "grad_norm": 1.0292372040908986, "learning_rate": 1.2460483062520246e-06, "loss": 0.0399, "step": 788 }, { "epoch": 0.1795221843003413, "grad_norm": 1.056684060389556, "learning_rate": 1.2460382680777196e-06, "loss": 0.0342, "step": 789 }, { "epoch": 0.17974971558589306, "grad_norm": 1.3734567420277208, "learning_rate": 1.2460282172105298e-06, "loss": 0.0673, "step": 790 }, { "epoch": 0.1799772468714448, "grad_norm": 1.519969265957327, "learning_rate": 1.2460181536506608e-06, "loss": 0.0775, "step": 791 }, { "epoch": 0.1802047781569966, "grad_norm": 1.514321767455213, "learning_rate": 1.2460080773983177e-06, "loss": 0.0659, "step": 792 }, { "epoch": 0.18043230944254834, "grad_norm": 1.0361490916169322, "learning_rate": 1.2459979884537072e-06, "loss": 0.0424, "step": 793 }, { "epoch": 0.18065984072810012, "grad_norm": 1.3102760225972407, "learning_rate": 1.2459878868170348e-06, "loss": 0.0476, "step": 794 }, { "epoch": 0.18088737201365188, "grad_norm": 1.4645693998145335, "learning_rate": 1.2459777724885075e-06, "loss": 0.0522, "step": 795 }, { "epoch": 0.18111490329920363, "grad_norm": 1.6043429433204082, "learning_rate": 1.2459676454683318e-06, "loss": 0.0415, "step": 796 }, { "epoch": 0.1813424345847554, "grad_norm": 2.3566952879735963, "learning_rate": 1.2459575057567144e-06, "loss": 0.0916, "step": 797 }, { "epoch": 0.18156996587030716, "grad_norm": 1.5769029444288707, "learning_rate": 1.245947353353863e-06, "loss": 0.0596, "step": 798 }, { "epoch": 0.18179749715585894, "grad_norm": 2.293268133767521, "learning_rate": 1.245937188259985e-06, "loss": 0.1087, "step": 799 }, { "epoch": 0.1820250284414107, "grad_norm": 2.0227046414956096, "learning_rate": 1.245927010475288e-06, "loss": 0.0986, "step": 800 }, { "epoch": 0.18225255972696247, "grad_norm": 1.4895840436039498, "learning_rate": 1.24591681999998e-06, "loss": 0.0633, "step": 801 }, { "epoch": 0.18248009101251422, "grad_norm": 1.9671715135309942, "learning_rate": 1.2459066168342693e-06, "loss": 0.0551, "step": 802 }, { "epoch": 0.18270762229806597, "grad_norm": 2.1525267147108322, "learning_rate": 1.2458964009783646e-06, "loss": 0.0809, "step": 803 }, { "epoch": 0.18293515358361775, "grad_norm": 1.2864874343026942, "learning_rate": 1.2458861724324745e-06, "loss": 0.064, "step": 804 }, { "epoch": 0.1831626848691695, "grad_norm": 3.1883103987924444, "learning_rate": 1.2458759311968084e-06, "loss": 0.1303, "step": 805 }, { "epoch": 0.1833902161547213, "grad_norm": 1.0904025457159896, "learning_rate": 1.245865677271575e-06, "loss": 0.0419, "step": 806 }, { "epoch": 0.18361774744027304, "grad_norm": 1.4716173202867617, "learning_rate": 1.2458554106569844e-06, "loss": 0.072, "step": 807 }, { "epoch": 0.1838452787258248, "grad_norm": 0.9197168745315846, "learning_rate": 1.2458451313532463e-06, "loss": 0.04, "step": 808 }, { "epoch": 0.18407281001137657, "grad_norm": 1.2386933488526581, "learning_rate": 1.2458348393605708e-06, "loss": 0.0383, "step": 809 }, { "epoch": 0.18430034129692832, "grad_norm": 1.2129381734644349, "learning_rate": 1.2458245346791678e-06, "loss": 0.0486, "step": 810 }, { "epoch": 0.1845278725824801, "grad_norm": 1.0772770126401692, "learning_rate": 1.2458142173092486e-06, "loss": 0.0654, "step": 811 }, { "epoch": 0.18475540386803185, "grad_norm": 5.609205856069292, "learning_rate": 1.2458038872510237e-06, "loss": 0.0565, "step": 812 }, { "epoch": 0.1849829351535836, "grad_norm": 1.1271779547323408, "learning_rate": 1.2457935445047042e-06, "loss": 0.0414, "step": 813 }, { "epoch": 0.18521046643913538, "grad_norm": 1.1585514417402933, "learning_rate": 1.2457831890705018e-06, "loss": 0.0646, "step": 814 }, { "epoch": 0.18543799772468714, "grad_norm": 1.6262463401225817, "learning_rate": 1.2457728209486279e-06, "loss": 0.0555, "step": 815 }, { "epoch": 0.18566552901023892, "grad_norm": 1.6971282783367077, "learning_rate": 1.2457624401392943e-06, "loss": 0.0803, "step": 816 }, { "epoch": 0.18589306029579067, "grad_norm": 1.5890865721783058, "learning_rate": 1.2457520466427135e-06, "loss": 0.0637, "step": 817 }, { "epoch": 0.18612059158134245, "grad_norm": 1.663587359707044, "learning_rate": 1.2457416404590974e-06, "loss": 0.063, "step": 818 }, { "epoch": 0.1863481228668942, "grad_norm": 1.4589989522369327, "learning_rate": 1.2457312215886592e-06, "loss": 0.0665, "step": 819 }, { "epoch": 0.18657565415244595, "grad_norm": 1.2689131726245948, "learning_rate": 1.2457207900316115e-06, "loss": 0.0653, "step": 820 }, { "epoch": 0.18680318543799773, "grad_norm": 1.2763445646366045, "learning_rate": 1.245710345788168e-06, "loss": 0.0691, "step": 821 }, { "epoch": 0.18703071672354948, "grad_norm": 2.1044109818141523, "learning_rate": 1.2456998888585414e-06, "loss": 0.0334, "step": 822 }, { "epoch": 0.18725824800910126, "grad_norm": 1.1582960528680695, "learning_rate": 1.245689419242946e-06, "loss": 0.0513, "step": 823 }, { "epoch": 0.18748577929465302, "grad_norm": 1.4301422945996285, "learning_rate": 1.2456789369415955e-06, "loss": 0.0632, "step": 824 }, { "epoch": 0.18771331058020477, "grad_norm": 1.5961327622673143, "learning_rate": 1.2456684419547044e-06, "loss": 0.0915, "step": 825 }, { "epoch": 0.18794084186575655, "grad_norm": 1.156039499623056, "learning_rate": 1.245657934282487e-06, "loss": 0.0406, "step": 826 }, { "epoch": 0.1881683731513083, "grad_norm": 0.7964273734238357, "learning_rate": 1.245647413925158e-06, "loss": 0.0339, "step": 827 }, { "epoch": 0.18839590443686008, "grad_norm": 1.3110827687978466, "learning_rate": 1.2456368808829327e-06, "loss": 0.0315, "step": 828 }, { "epoch": 0.18862343572241183, "grad_norm": 2.167581797519473, "learning_rate": 1.2456263351560261e-06, "loss": 0.0944, "step": 829 }, { "epoch": 0.18885096700796358, "grad_norm": 1.7008701552637757, "learning_rate": 1.2456157767446538e-06, "loss": 0.0675, "step": 830 }, { "epoch": 0.18907849829351536, "grad_norm": 0.9862167570499358, "learning_rate": 1.245605205649032e-06, "loss": 0.0283, "step": 831 }, { "epoch": 0.18930602957906711, "grad_norm": 1.4504322451994889, "learning_rate": 1.245594621869376e-06, "loss": 0.0855, "step": 832 }, { "epoch": 0.1895335608646189, "grad_norm": 0.8637440563578112, "learning_rate": 1.2455840254059026e-06, "loss": 0.0457, "step": 833 }, { "epoch": 0.18976109215017065, "grad_norm": 1.365450351316701, "learning_rate": 1.2455734162588282e-06, "loss": 0.0523, "step": 834 }, { "epoch": 0.1899886234357224, "grad_norm": 1.9203688742265796, "learning_rate": 1.2455627944283697e-06, "loss": 0.0421, "step": 835 }, { "epoch": 0.19021615472127418, "grad_norm": 1.947700834519252, "learning_rate": 1.245552159914744e-06, "loss": 0.0631, "step": 836 }, { "epoch": 0.19044368600682593, "grad_norm": 1.274745482531776, "learning_rate": 1.245541512718169e-06, "loss": 0.0408, "step": 837 }, { "epoch": 0.1906712172923777, "grad_norm": 1.2840926844609297, "learning_rate": 1.245530852838862e-06, "loss": 0.0435, "step": 838 }, { "epoch": 0.19089874857792946, "grad_norm": 1.5640003481699476, "learning_rate": 1.2455201802770405e-06, "loss": 0.0616, "step": 839 }, { "epoch": 0.19112627986348124, "grad_norm": 1.2705630922274525, "learning_rate": 1.245509495032923e-06, "loss": 0.0604, "step": 840 }, { "epoch": 0.191353811149033, "grad_norm": 1.2721765422307298, "learning_rate": 1.2454987971067278e-06, "loss": 0.0431, "step": 841 }, { "epoch": 0.19158134243458474, "grad_norm": 1.7662664353990207, "learning_rate": 1.2454880864986737e-06, "loss": 0.0881, "step": 842 }, { "epoch": 0.19180887372013652, "grad_norm": 0.8548688062918027, "learning_rate": 1.2454773632089795e-06, "loss": 0.0219, "step": 843 }, { "epoch": 0.19203640500568828, "grad_norm": 1.1833145714689994, "learning_rate": 1.2454666272378644e-06, "loss": 0.0502, "step": 844 }, { "epoch": 0.19226393629124006, "grad_norm": 2.15061289434072, "learning_rate": 1.2454558785855475e-06, "loss": 0.0687, "step": 845 }, { "epoch": 0.1924914675767918, "grad_norm": 1.0201954402445583, "learning_rate": 1.245445117252249e-06, "loss": 0.0377, "step": 846 }, { "epoch": 0.19271899886234356, "grad_norm": 2.256502679796923, "learning_rate": 1.2454343432381886e-06, "loss": 0.1344, "step": 847 }, { "epoch": 0.19294653014789534, "grad_norm": 0.9607197396814332, "learning_rate": 1.2454235565435862e-06, "loss": 0.0293, "step": 848 }, { "epoch": 0.1931740614334471, "grad_norm": 1.1041321579517591, "learning_rate": 1.2454127571686629e-06, "loss": 0.0588, "step": 849 }, { "epoch": 0.19340159271899887, "grad_norm": 1.551888854493328, "learning_rate": 1.245401945113639e-06, "loss": 0.0884, "step": 850 }, { "epoch": 0.19362912400455062, "grad_norm": 1.2207595212151812, "learning_rate": 1.2453911203787355e-06, "loss": 0.0495, "step": 851 }, { "epoch": 0.19385665529010238, "grad_norm": 3.151128795398426, "learning_rate": 1.2453802829641736e-06, "loss": 0.0959, "step": 852 }, { "epoch": 0.19408418657565416, "grad_norm": 1.0306148148350838, "learning_rate": 1.2453694328701752e-06, "loss": 0.0362, "step": 853 }, { "epoch": 0.1943117178612059, "grad_norm": 1.4686426621438413, "learning_rate": 1.2453585700969614e-06, "loss": 0.0519, "step": 854 }, { "epoch": 0.1945392491467577, "grad_norm": 4.418162916580233, "learning_rate": 1.2453476946447547e-06, "loss": 0.0718, "step": 855 }, { "epoch": 0.19476678043230944, "grad_norm": 1.2510611267805927, "learning_rate": 1.2453368065137772e-06, "loss": 0.0471, "step": 856 }, { "epoch": 0.19499431171786122, "grad_norm": 3.8279701209646566, "learning_rate": 1.2453259057042514e-06, "loss": 0.0962, "step": 857 }, { "epoch": 0.19522184300341297, "grad_norm": 0.9857769216686043, "learning_rate": 1.2453149922164003e-06, "loss": 0.0361, "step": 858 }, { "epoch": 0.19544937428896472, "grad_norm": 1.2075815978254054, "learning_rate": 1.2453040660504468e-06, "loss": 0.0361, "step": 859 }, { "epoch": 0.1956769055745165, "grad_norm": 1.5382334773010704, "learning_rate": 1.2452931272066141e-06, "loss": 0.0489, "step": 860 }, { "epoch": 0.19590443686006825, "grad_norm": 1.2911029078532041, "learning_rate": 1.245282175685126e-06, "loss": 0.0743, "step": 861 }, { "epoch": 0.19613196814562003, "grad_norm": 2.432722962095318, "learning_rate": 1.2452712114862063e-06, "loss": 0.106, "step": 862 }, { "epoch": 0.19635949943117179, "grad_norm": 1.3235947439225328, "learning_rate": 1.245260234610079e-06, "loss": 0.045, "step": 863 }, { "epoch": 0.19658703071672354, "grad_norm": 1.1324676509010458, "learning_rate": 1.2452492450569682e-06, "loss": 0.0554, "step": 864 }, { "epoch": 0.19681456200227532, "grad_norm": 2.2307637995876863, "learning_rate": 1.245238242827099e-06, "loss": 0.0832, "step": 865 }, { "epoch": 0.19704209328782707, "grad_norm": 1.4079128192407386, "learning_rate": 1.245227227920696e-06, "loss": 0.079, "step": 866 }, { "epoch": 0.19726962457337885, "grad_norm": 1.9175023199810017, "learning_rate": 1.2452162003379842e-06, "loss": 0.0917, "step": 867 }, { "epoch": 0.1974971558589306, "grad_norm": 1.731489599286668, "learning_rate": 1.2452051600791891e-06, "loss": 0.0571, "step": 868 }, { "epoch": 0.19772468714448235, "grad_norm": 1.3722463786220256, "learning_rate": 1.2451941071445367e-06, "loss": 0.039, "step": 869 }, { "epoch": 0.19795221843003413, "grad_norm": 1.287569176842529, "learning_rate": 1.2451830415342524e-06, "loss": 0.0427, "step": 870 }, { "epoch": 0.19817974971558588, "grad_norm": 2.2505234649573795, "learning_rate": 1.2451719632485627e-06, "loss": 0.0606, "step": 871 }, { "epoch": 0.19840728100113766, "grad_norm": 1.3085408197305148, "learning_rate": 1.2451608722876938e-06, "loss": 0.0659, "step": 872 }, { "epoch": 0.19863481228668942, "grad_norm": 2.3075742049076085, "learning_rate": 1.2451497686518722e-06, "loss": 0.0762, "step": 873 }, { "epoch": 0.1988623435722412, "grad_norm": 1.1650796320462304, "learning_rate": 1.2451386523413252e-06, "loss": 0.0559, "step": 874 }, { "epoch": 0.19908987485779295, "grad_norm": 1.199672804247241, "learning_rate": 1.24512752335628e-06, "loss": 0.0417, "step": 875 }, { "epoch": 0.1993174061433447, "grad_norm": 1.4954309542074338, "learning_rate": 1.2451163816969639e-06, "loss": 0.0841, "step": 876 }, { "epoch": 0.19954493742889648, "grad_norm": 1.4893257668939828, "learning_rate": 1.2451052273636045e-06, "loss": 0.0639, "step": 877 }, { "epoch": 0.19977246871444823, "grad_norm": 1.1681659159861986, "learning_rate": 1.24509406035643e-06, "loss": 0.0487, "step": 878 }, { "epoch": 0.2, "grad_norm": 1.337002748039853, "learning_rate": 1.2450828806756685e-06, "loss": 0.0408, "step": 879 }, { "epoch": 0.20022753128555176, "grad_norm": 1.238199679938109, "learning_rate": 1.245071688321549e-06, "loss": 0.0452, "step": 880 }, { "epoch": 0.20045506257110352, "grad_norm": 1.4016981334625263, "learning_rate": 1.2450604832942991e-06, "loss": 0.0462, "step": 881 }, { "epoch": 0.2006825938566553, "grad_norm": 1.4932886959066234, "learning_rate": 1.245049265594149e-06, "loss": 0.0735, "step": 882 }, { "epoch": 0.20091012514220705, "grad_norm": 1.2357598702477623, "learning_rate": 1.2450380352213271e-06, "loss": 0.0504, "step": 883 }, { "epoch": 0.20113765642775883, "grad_norm": 2.015377601632808, "learning_rate": 1.2450267921760636e-06, "loss": 0.0523, "step": 884 }, { "epoch": 0.20136518771331058, "grad_norm": 1.8296148009850803, "learning_rate": 1.2450155364585878e-06, "loss": 0.0554, "step": 885 }, { "epoch": 0.20159271899886233, "grad_norm": 1.9872439975590221, "learning_rate": 1.2450042680691301e-06, "loss": 0.0737, "step": 886 }, { "epoch": 0.2018202502844141, "grad_norm": 0.832951332510299, "learning_rate": 1.2449929870079206e-06, "loss": 0.0457, "step": 887 }, { "epoch": 0.20204778156996586, "grad_norm": 0.7896517265085776, "learning_rate": 1.24498169327519e-06, "loss": 0.0261, "step": 888 }, { "epoch": 0.20227531285551764, "grad_norm": 1.3980548207185421, "learning_rate": 1.2449703868711688e-06, "loss": 0.061, "step": 889 }, { "epoch": 0.2025028441410694, "grad_norm": 1.3109982570334282, "learning_rate": 1.2449590677960886e-06, "loss": 0.0525, "step": 890 }, { "epoch": 0.20273037542662117, "grad_norm": 0.8611577720266457, "learning_rate": 1.2449477360501802e-06, "loss": 0.0297, "step": 891 }, { "epoch": 0.20295790671217293, "grad_norm": 2.2537154917621267, "learning_rate": 1.2449363916336756e-06, "loss": 0.0658, "step": 892 }, { "epoch": 0.20318543799772468, "grad_norm": 2.7860866459618823, "learning_rate": 1.2449250345468065e-06, "loss": 0.0853, "step": 893 }, { "epoch": 0.20341296928327646, "grad_norm": 2.063772825536578, "learning_rate": 1.244913664789805e-06, "loss": 0.0413, "step": 894 }, { "epoch": 0.2036405005688282, "grad_norm": 1.3514226592309118, "learning_rate": 1.2449022823629036e-06, "loss": 0.0445, "step": 895 }, { "epoch": 0.20386803185438, "grad_norm": 1.2525160508368907, "learning_rate": 1.2448908872663347e-06, "loss": 0.0337, "step": 896 }, { "epoch": 0.20409556313993174, "grad_norm": 0.9460587754259561, "learning_rate": 1.2448794795003313e-06, "loss": 0.0391, "step": 897 }, { "epoch": 0.2043230944254835, "grad_norm": 1.146769061895453, "learning_rate": 1.2448680590651269e-06, "loss": 0.0618, "step": 898 }, { "epoch": 0.20455062571103527, "grad_norm": 1.277169361575558, "learning_rate": 1.2448566259609543e-06, "loss": 0.0479, "step": 899 }, { "epoch": 0.20477815699658702, "grad_norm": 1.2913475907417995, "learning_rate": 1.2448451801880476e-06, "loss": 0.0352, "step": 900 }, { "epoch": 0.2050056882821388, "grad_norm": 1.614343484279519, "learning_rate": 1.2448337217466404e-06, "loss": 0.0672, "step": 901 }, { "epoch": 0.20523321956769056, "grad_norm": 1.1122189034817742, "learning_rate": 1.2448222506369675e-06, "loss": 0.0385, "step": 902 }, { "epoch": 0.2054607508532423, "grad_norm": 1.5480813920575183, "learning_rate": 1.2448107668592626e-06, "loss": 0.0696, "step": 903 }, { "epoch": 0.2056882821387941, "grad_norm": 2.133323122403055, "learning_rate": 1.244799270413761e-06, "loss": 0.0726, "step": 904 }, { "epoch": 0.20591581342434584, "grad_norm": 1.9375085657983182, "learning_rate": 1.2447877613006972e-06, "loss": 0.0762, "step": 905 }, { "epoch": 0.20614334470989762, "grad_norm": 2.138121754084785, "learning_rate": 1.244776239520307e-06, "loss": 0.1251, "step": 906 }, { "epoch": 0.20637087599544937, "grad_norm": 69.30343641276347, "learning_rate": 1.244764705072825e-06, "loss": 1.5633, "step": 907 }, { "epoch": 0.20659840728100115, "grad_norm": 0.9136079997375776, "learning_rate": 1.2447531579584878e-06, "loss": 0.0359, "step": 908 }, { "epoch": 0.2068259385665529, "grad_norm": 2.258737842455169, "learning_rate": 1.2447415981775312e-06, "loss": 0.0815, "step": 909 }, { "epoch": 0.20705346985210465, "grad_norm": 1.1302894088592583, "learning_rate": 1.2447300257301912e-06, "loss": 0.0377, "step": 910 }, { "epoch": 0.20728100113765643, "grad_norm": 1.4921976481695052, "learning_rate": 1.2447184406167045e-06, "loss": 0.0867, "step": 911 }, { "epoch": 0.2075085324232082, "grad_norm": 2.2983270765657546, "learning_rate": 1.2447068428373077e-06, "loss": 0.0726, "step": 912 }, { "epoch": 0.20773606370875997, "grad_norm": 1.9446016021679782, "learning_rate": 1.244695232392238e-06, "loss": 0.064, "step": 913 }, { "epoch": 0.20796359499431172, "grad_norm": 0.9872094991366439, "learning_rate": 1.2446836092817328e-06, "loss": 0.0432, "step": 914 }, { "epoch": 0.20819112627986347, "grad_norm": 6.934093481791269, "learning_rate": 1.2446719735060293e-06, "loss": 0.0945, "step": 915 }, { "epoch": 0.20841865756541525, "grad_norm": 1.360488111746127, "learning_rate": 1.2446603250653658e-06, "loss": 0.0603, "step": 916 }, { "epoch": 0.208646188850967, "grad_norm": 1.3310461950551042, "learning_rate": 1.24464866395998e-06, "loss": 0.0365, "step": 917 }, { "epoch": 0.20887372013651878, "grad_norm": 0.9182393241316875, "learning_rate": 1.2446369901901102e-06, "loss": 0.0559, "step": 918 }, { "epoch": 0.20910125142207053, "grad_norm": 1.4023619755477905, "learning_rate": 1.2446253037559952e-06, "loss": 0.0457, "step": 919 }, { "epoch": 0.20932878270762229, "grad_norm": 1.1256168684655514, "learning_rate": 1.2446136046578739e-06, "loss": 0.0455, "step": 920 }, { "epoch": 0.20955631399317406, "grad_norm": 1.9433578174394295, "learning_rate": 1.2446018928959853e-06, "loss": 0.0962, "step": 921 }, { "epoch": 0.20978384527872582, "grad_norm": 1.5957841959716605, "learning_rate": 1.2445901684705685e-06, "loss": 0.0408, "step": 922 }, { "epoch": 0.2100113765642776, "grad_norm": 1.075018126077203, "learning_rate": 1.2445784313818638e-06, "loss": 0.0378, "step": 923 }, { "epoch": 0.21023890784982935, "grad_norm": 1.5370330862679797, "learning_rate": 1.2445666816301102e-06, "loss": 0.0578, "step": 924 }, { "epoch": 0.21046643913538113, "grad_norm": 1.4943129389397118, "learning_rate": 1.2445549192155487e-06, "loss": 0.0875, "step": 925 }, { "epoch": 0.21069397042093288, "grad_norm": 1.1898394401405206, "learning_rate": 1.244543144138419e-06, "loss": 0.0486, "step": 926 }, { "epoch": 0.21092150170648463, "grad_norm": 1.6954763665720234, "learning_rate": 1.2445313563989624e-06, "loss": 0.0641, "step": 927 }, { "epoch": 0.2111490329920364, "grad_norm": 1.0426224473753456, "learning_rate": 1.2445195559974194e-06, "loss": 0.0404, "step": 928 }, { "epoch": 0.21137656427758816, "grad_norm": 1.6481528761515403, "learning_rate": 1.244507742934031e-06, "loss": 0.0839, "step": 929 }, { "epoch": 0.21160409556313994, "grad_norm": 0.9619424275618367, "learning_rate": 1.2444959172090393e-06, "loss": 0.0273, "step": 930 }, { "epoch": 0.2118316268486917, "grad_norm": 3.69544520271876, "learning_rate": 1.2444840788226854e-06, "loss": 0.1652, "step": 931 }, { "epoch": 0.21205915813424345, "grad_norm": 1.804306156177365, "learning_rate": 1.2444722277752114e-06, "loss": 0.1027, "step": 932 }, { "epoch": 0.21228668941979523, "grad_norm": 1.768050037917839, "learning_rate": 1.2444603640668596e-06, "loss": 0.095, "step": 933 }, { "epoch": 0.21251422070534698, "grad_norm": 1.1285321294388682, "learning_rate": 1.2444484876978725e-06, "loss": 0.0465, "step": 934 }, { "epoch": 0.21274175199089876, "grad_norm": 1.7249462106069577, "learning_rate": 1.2444365986684929e-06, "loss": 0.0842, "step": 935 }, { "epoch": 0.2129692832764505, "grad_norm": 1.2306924671997328, "learning_rate": 1.2444246969789633e-06, "loss": 0.0447, "step": 936 }, { "epoch": 0.21319681456200226, "grad_norm": 1.1194256262692541, "learning_rate": 1.2444127826295277e-06, "loss": 0.0387, "step": 937 }, { "epoch": 0.21342434584755404, "grad_norm": 0.9357846968568098, "learning_rate": 1.244400855620429e-06, "loss": 0.0311, "step": 938 }, { "epoch": 0.2136518771331058, "grad_norm": 0.9399277842933605, "learning_rate": 1.2443889159519113e-06, "loss": 0.0408, "step": 939 }, { "epoch": 0.21387940841865757, "grad_norm": 1.038736805523899, "learning_rate": 1.2443769636242185e-06, "loss": 0.0573, "step": 940 }, { "epoch": 0.21410693970420933, "grad_norm": 1.179286632772412, "learning_rate": 1.244364998637595e-06, "loss": 0.0578, "step": 941 }, { "epoch": 0.2143344709897611, "grad_norm": 3.1404649465639594, "learning_rate": 1.2443530209922848e-06, "loss": 0.0549, "step": 942 }, { "epoch": 0.21456200227531286, "grad_norm": 1.4304654332903906, "learning_rate": 1.2443410306885337e-06, "loss": 0.0408, "step": 943 }, { "epoch": 0.2147895335608646, "grad_norm": 1.3926671923537388, "learning_rate": 1.244329027726586e-06, "loss": 0.0675, "step": 944 }, { "epoch": 0.2150170648464164, "grad_norm": 0.8473568816417418, "learning_rate": 1.2443170121066872e-06, "loss": 0.0388, "step": 945 }, { "epoch": 0.21524459613196814, "grad_norm": 1.7456532933762987, "learning_rate": 1.2443049838290827e-06, "loss": 0.0655, "step": 946 }, { "epoch": 0.21547212741751992, "grad_norm": 5.065824955296096, "learning_rate": 1.2442929428940186e-06, "loss": 0.196, "step": 947 }, { "epoch": 0.21569965870307167, "grad_norm": 0.9216519176270381, "learning_rate": 1.2442808893017414e-06, "loss": 0.0376, "step": 948 }, { "epoch": 0.21592718998862342, "grad_norm": 1.111383779669666, "learning_rate": 1.2442688230524965e-06, "loss": 0.0403, "step": 949 }, { "epoch": 0.2161547212741752, "grad_norm": 1.557530476173142, "learning_rate": 1.244256744146531e-06, "loss": 0.0365, "step": 950 }, { "epoch": 0.21638225255972696, "grad_norm": 1.1787378400008193, "learning_rate": 1.244244652584092e-06, "loss": 0.0507, "step": 951 }, { "epoch": 0.21660978384527874, "grad_norm": 0.9890913425273592, "learning_rate": 1.2442325483654263e-06, "loss": 0.0533, "step": 952 }, { "epoch": 0.2168373151308305, "grad_norm": 2.1206498633906294, "learning_rate": 1.2442204314907812e-06, "loss": 0.0794, "step": 953 }, { "epoch": 0.21706484641638224, "grad_norm": 1.3153818338446894, "learning_rate": 1.2442083019604047e-06, "loss": 0.0706, "step": 954 }, { "epoch": 0.21729237770193402, "grad_norm": 1.2497584639698338, "learning_rate": 1.2441961597745447e-06, "loss": 0.0474, "step": 955 }, { "epoch": 0.21751990898748577, "grad_norm": 1.0089839713441826, "learning_rate": 1.244184004933449e-06, "loss": 0.0354, "step": 956 }, { "epoch": 0.21774744027303755, "grad_norm": 1.967078988299315, "learning_rate": 1.2441718374373662e-06, "loss": 0.0371, "step": 957 }, { "epoch": 0.2179749715585893, "grad_norm": 1.1714442804547354, "learning_rate": 1.244159657286545e-06, "loss": 0.0452, "step": 958 }, { "epoch": 0.21820250284414108, "grad_norm": 0.940255872741034, "learning_rate": 1.2441474644812345e-06, "loss": 0.0363, "step": 959 }, { "epoch": 0.21843003412969283, "grad_norm": 1.100910872233546, "learning_rate": 1.2441352590216836e-06, "loss": 0.0357, "step": 960 }, { "epoch": 0.2186575654152446, "grad_norm": 1.9969980893987056, "learning_rate": 1.244123040908142e-06, "loss": 0.1572, "step": 961 }, { "epoch": 0.21888509670079637, "grad_norm": 1.5961804013592509, "learning_rate": 1.2441108101408592e-06, "loss": 0.0425, "step": 962 }, { "epoch": 0.21911262798634812, "grad_norm": 2.1162445389078024, "learning_rate": 1.2440985667200853e-06, "loss": 0.0517, "step": 963 }, { "epoch": 0.2193401592718999, "grad_norm": 1.064967765846801, "learning_rate": 1.2440863106460705e-06, "loss": 0.023, "step": 964 }, { "epoch": 0.21956769055745165, "grad_norm": 2.08098035050502, "learning_rate": 1.2440740419190655e-06, "loss": 0.0796, "step": 965 }, { "epoch": 0.2197952218430034, "grad_norm": 2.4340967444170136, "learning_rate": 1.2440617605393208e-06, "loss": 0.0673, "step": 966 }, { "epoch": 0.22002275312855518, "grad_norm": 2.139025801414634, "learning_rate": 1.2440494665070874e-06, "loss": 0.102, "step": 967 }, { "epoch": 0.22025028441410693, "grad_norm": 1.3877680351054698, "learning_rate": 1.2440371598226165e-06, "loss": 0.0588, "step": 968 }, { "epoch": 0.2204778156996587, "grad_norm": 1.3210352246509975, "learning_rate": 1.2440248404861598e-06, "loss": 0.0528, "step": 969 }, { "epoch": 0.22070534698521047, "grad_norm": 1.4713104262343806, "learning_rate": 1.2440125084979693e-06, "loss": 0.0468, "step": 970 }, { "epoch": 0.22093287827076222, "grad_norm": 1.4854620779239043, "learning_rate": 1.2440001638582965e-06, "loss": 0.0494, "step": 971 }, { "epoch": 0.221160409556314, "grad_norm": 1.1733972765082468, "learning_rate": 1.2439878065673944e-06, "loss": 0.0517, "step": 972 }, { "epoch": 0.22138794084186575, "grad_norm": 2.818817007281193, "learning_rate": 1.2439754366255149e-06, "loss": 0.0332, "step": 973 }, { "epoch": 0.22161547212741753, "grad_norm": 1.3519809642179033, "learning_rate": 1.2439630540329111e-06, "loss": 0.0714, "step": 974 }, { "epoch": 0.22184300341296928, "grad_norm": 1.2366129488699829, "learning_rate": 1.2439506587898358e-06, "loss": 0.0307, "step": 975 }, { "epoch": 0.22207053469852106, "grad_norm": 1.2264022197392555, "learning_rate": 1.243938250896543e-06, "loss": 0.0505, "step": 976 }, { "epoch": 0.2222980659840728, "grad_norm": 1.7725437520836973, "learning_rate": 1.2439258303532858e-06, "loss": 0.0551, "step": 977 }, { "epoch": 0.22252559726962456, "grad_norm": 0.9663835721436027, "learning_rate": 1.243913397160318e-06, "loss": 0.0375, "step": 978 }, { "epoch": 0.22275312855517634, "grad_norm": 0.8743576163827176, "learning_rate": 1.2439009513178938e-06, "loss": 0.0245, "step": 979 }, { "epoch": 0.2229806598407281, "grad_norm": 1.054127962193629, "learning_rate": 1.2438884928262678e-06, "loss": 0.0323, "step": 980 }, { "epoch": 0.22320819112627988, "grad_norm": 0.9874001278398739, "learning_rate": 1.2438760216856944e-06, "loss": 0.0435, "step": 981 }, { "epoch": 0.22343572241183163, "grad_norm": 1.3544331214991197, "learning_rate": 1.2438635378964284e-06, "loss": 0.0707, "step": 982 }, { "epoch": 0.22366325369738338, "grad_norm": 1.3272223315121154, "learning_rate": 1.2438510414587251e-06, "loss": 0.0553, "step": 983 }, { "epoch": 0.22389078498293516, "grad_norm": 1.4246609306991453, "learning_rate": 1.24383853237284e-06, "loss": 0.0616, "step": 984 }, { "epoch": 0.2241183162684869, "grad_norm": 1.3509367739810239, "learning_rate": 1.2438260106390285e-06, "loss": 0.054, "step": 985 }, { "epoch": 0.2243458475540387, "grad_norm": 2.743016610014617, "learning_rate": 1.2438134762575467e-06, "loss": 0.1654, "step": 986 }, { "epoch": 0.22457337883959044, "grad_norm": 1.1104234557718136, "learning_rate": 1.243800929228651e-06, "loss": 0.0438, "step": 987 }, { "epoch": 0.2248009101251422, "grad_norm": 1.5813628573176925, "learning_rate": 1.2437883695525974e-06, "loss": 0.0577, "step": 988 }, { "epoch": 0.22502844141069397, "grad_norm": 0.9637284966057095, "learning_rate": 1.2437757972296427e-06, "loss": 0.0441, "step": 989 }, { "epoch": 0.22525597269624573, "grad_norm": 1.2736418998037702, "learning_rate": 1.2437632122600442e-06, "loss": 0.0627, "step": 990 }, { "epoch": 0.2254835039817975, "grad_norm": 1.704338839028225, "learning_rate": 1.2437506146440587e-06, "loss": 0.0794, "step": 991 }, { "epoch": 0.22571103526734926, "grad_norm": 1.3083723593285776, "learning_rate": 1.243738004381944e-06, "loss": 0.0367, "step": 992 }, { "epoch": 0.225938566552901, "grad_norm": 1.2372256320814667, "learning_rate": 1.2437253814739572e-06, "loss": 0.0478, "step": 993 }, { "epoch": 0.2261660978384528, "grad_norm": 2.064272199741316, "learning_rate": 1.2437127459203572e-06, "loss": 0.0673, "step": 994 }, { "epoch": 0.22639362912400454, "grad_norm": 0.9919600393162777, "learning_rate": 1.2437000977214015e-06, "loss": 0.031, "step": 995 }, { "epoch": 0.22662116040955632, "grad_norm": 1.6080832002027023, "learning_rate": 1.243687436877349e-06, "loss": 0.0841, "step": 996 }, { "epoch": 0.22684869169510807, "grad_norm": 1.3930180180919087, "learning_rate": 1.2436747633884583e-06, "loss": 0.0616, "step": 997 }, { "epoch": 0.22707622298065985, "grad_norm": 4.70664655158806, "learning_rate": 1.2436620772549885e-06, "loss": 0.1541, "step": 998 }, { "epoch": 0.2273037542662116, "grad_norm": 2.3548939922944556, "learning_rate": 1.243649378477199e-06, "loss": 0.1161, "step": 999 }, { "epoch": 0.22753128555176336, "grad_norm": 1.9186203633212395, "learning_rate": 1.2436366670553491e-06, "loss": 0.0725, "step": 1000 }, { "epoch": 0.22775881683731514, "grad_norm": 1.726012700048525, "learning_rate": 1.2436239429896988e-06, "loss": 0.1038, "step": 1001 }, { "epoch": 0.2279863481228669, "grad_norm": 1.7981785549763478, "learning_rate": 1.2436112062805081e-06, "loss": 0.0485, "step": 1002 }, { "epoch": 0.22821387940841867, "grad_norm": 1.935791028636196, "learning_rate": 1.2435984569280372e-06, "loss": 0.0773, "step": 1003 }, { "epoch": 0.22844141069397042, "grad_norm": 1.4700273599627696, "learning_rate": 1.2435856949325467e-06, "loss": 0.0584, "step": 1004 }, { "epoch": 0.22866894197952217, "grad_norm": 4.712742088498197, "learning_rate": 1.2435729202942972e-06, "loss": 0.073, "step": 1005 }, { "epoch": 0.22889647326507395, "grad_norm": 1.3146906101281273, "learning_rate": 1.2435601330135506e-06, "loss": 0.0357, "step": 1006 }, { "epoch": 0.2291240045506257, "grad_norm": 2.3674281228532066, "learning_rate": 1.2435473330905674e-06, "loss": 0.0701, "step": 1007 }, { "epoch": 0.22935153583617748, "grad_norm": 1.2459812264205032, "learning_rate": 1.2435345205256097e-06, "loss": 0.0375, "step": 1008 }, { "epoch": 0.22957906712172924, "grad_norm": 1.4505448834484096, "learning_rate": 1.243521695318939e-06, "loss": 0.057, "step": 1009 }, { "epoch": 0.229806598407281, "grad_norm": 1.014039827241364, "learning_rate": 1.2435088574708178e-06, "loss": 0.027, "step": 1010 }, { "epoch": 0.23003412969283277, "grad_norm": 2.0117138797089766, "learning_rate": 1.2434960069815083e-06, "loss": 0.0583, "step": 1011 }, { "epoch": 0.23026166097838452, "grad_norm": 2.319983736807929, "learning_rate": 1.243483143851273e-06, "loss": 0.0687, "step": 1012 }, { "epoch": 0.2304891922639363, "grad_norm": 1.5313544208925962, "learning_rate": 1.2434702680803751e-06, "loss": 0.0531, "step": 1013 }, { "epoch": 0.23071672354948805, "grad_norm": 2.4005097874561616, "learning_rate": 1.2434573796690774e-06, "loss": 0.0929, "step": 1014 }, { "epoch": 0.23094425483503983, "grad_norm": 1.4786754339201256, "learning_rate": 1.2434444786176435e-06, "loss": 0.072, "step": 1015 }, { "epoch": 0.23117178612059158, "grad_norm": 1.0190695534603456, "learning_rate": 1.2434315649263372e-06, "loss": 0.0336, "step": 1016 }, { "epoch": 0.23139931740614333, "grad_norm": 0.85632474720689, "learning_rate": 1.2434186385954225e-06, "loss": 0.035, "step": 1017 }, { "epoch": 0.23162684869169511, "grad_norm": 1.2607686804487748, "learning_rate": 1.243405699625163e-06, "loss": 0.0442, "step": 1018 }, { "epoch": 0.23185437997724687, "grad_norm": 1.2717127046145194, "learning_rate": 1.243392748015824e-06, "loss": 0.0377, "step": 1019 }, { "epoch": 0.23208191126279865, "grad_norm": 0.9597540032341592, "learning_rate": 1.2433797837676694e-06, "loss": 0.0434, "step": 1020 }, { "epoch": 0.2323094425483504, "grad_norm": 0.9627749018367474, "learning_rate": 1.2433668068809648e-06, "loss": 0.0303, "step": 1021 }, { "epoch": 0.23253697383390215, "grad_norm": 1.2513646734831103, "learning_rate": 1.243353817355975e-06, "loss": 0.039, "step": 1022 }, { "epoch": 0.23276450511945393, "grad_norm": 1.2104231875454026, "learning_rate": 1.2433408151929655e-06, "loss": 0.0361, "step": 1023 }, { "epoch": 0.23299203640500568, "grad_norm": 1.0202244699658645, "learning_rate": 1.2433278003922026e-06, "loss": 0.0351, "step": 1024 }, { "epoch": 0.23321956769055746, "grad_norm": 1.4263990244044498, "learning_rate": 1.2433147729539514e-06, "loss": 0.0815, "step": 1025 }, { "epoch": 0.2334470989761092, "grad_norm": 1.5336205920808743, "learning_rate": 1.2433017328784788e-06, "loss": 0.0637, "step": 1026 }, { "epoch": 0.23367463026166096, "grad_norm": 1.4842765531420945, "learning_rate": 1.2432886801660513e-06, "loss": 0.0619, "step": 1027 }, { "epoch": 0.23390216154721274, "grad_norm": 0.9615217436557268, "learning_rate": 1.2432756148169354e-06, "loss": 0.0424, "step": 1028 }, { "epoch": 0.2341296928327645, "grad_norm": 2.5880116161383273, "learning_rate": 1.2432625368313983e-06, "loss": 0.0925, "step": 1029 }, { "epoch": 0.23435722411831628, "grad_norm": 1.1806940004387316, "learning_rate": 1.2432494462097072e-06, "loss": 0.0547, "step": 1030 }, { "epoch": 0.23458475540386803, "grad_norm": 0.9929303526421415, "learning_rate": 1.2432363429521295e-06, "loss": 0.0529, "step": 1031 }, { "epoch": 0.2348122866894198, "grad_norm": 0.9571908888077277, "learning_rate": 1.2432232270589335e-06, "loss": 0.0334, "step": 1032 }, { "epoch": 0.23503981797497156, "grad_norm": 2.067393897730002, "learning_rate": 1.2432100985303868e-06, "loss": 0.106, "step": 1033 }, { "epoch": 0.2352673492605233, "grad_norm": 0.8691321322200917, "learning_rate": 1.243196957366758e-06, "loss": 0.0346, "step": 1034 }, { "epoch": 0.2354948805460751, "grad_norm": 1.5305165337075897, "learning_rate": 1.2431838035683155e-06, "loss": 0.0848, "step": 1035 }, { "epoch": 0.23572241183162684, "grad_norm": 1.3223616364406545, "learning_rate": 1.2431706371353282e-06, "loss": 0.0687, "step": 1036 }, { "epoch": 0.23594994311717862, "grad_norm": 1.928604930493819, "learning_rate": 1.2431574580680653e-06, "loss": 0.0649, "step": 1037 }, { "epoch": 0.23617747440273038, "grad_norm": 1.2898613078943457, "learning_rate": 1.2431442663667958e-06, "loss": 0.0542, "step": 1038 }, { "epoch": 0.23640500568828213, "grad_norm": 1.0765527225532854, "learning_rate": 1.2431310620317898e-06, "loss": 0.0601, "step": 1039 }, { "epoch": 0.2366325369738339, "grad_norm": 0.7732520151584088, "learning_rate": 1.2431178450633168e-06, "loss": 0.0285, "step": 1040 }, { "epoch": 0.23686006825938566, "grad_norm": 1.1685154356326117, "learning_rate": 1.2431046154616473e-06, "loss": 0.0356, "step": 1041 }, { "epoch": 0.23708759954493744, "grad_norm": 1.3543617648083628, "learning_rate": 1.2430913732270512e-06, "loss": 0.0446, "step": 1042 }, { "epoch": 0.2373151308304892, "grad_norm": 1.707962300050113, "learning_rate": 1.2430781183597995e-06, "loss": 0.0555, "step": 1043 }, { "epoch": 0.23754266211604094, "grad_norm": 1.2841428640766417, "learning_rate": 1.243064850860163e-06, "loss": 0.0447, "step": 1044 }, { "epoch": 0.23777019340159272, "grad_norm": 6.261159556049287, "learning_rate": 1.243051570728413e-06, "loss": 0.0749, "step": 1045 }, { "epoch": 0.23799772468714447, "grad_norm": 1.1364071139867686, "learning_rate": 1.2430382779648208e-06, "loss": 0.0326, "step": 1046 }, { "epoch": 0.23822525597269625, "grad_norm": 1.484581864381779, "learning_rate": 1.243024972569658e-06, "loss": 0.0645, "step": 1047 }, { "epoch": 0.238452787258248, "grad_norm": 1.1545473703015605, "learning_rate": 1.2430116545431966e-06, "loss": 0.0481, "step": 1048 }, { "epoch": 0.23868031854379979, "grad_norm": 0.8983805133665866, "learning_rate": 1.2429983238857088e-06, "loss": 0.0369, "step": 1049 }, { "epoch": 0.23890784982935154, "grad_norm": 1.068742935236811, "learning_rate": 1.2429849805974673e-06, "loss": 0.039, "step": 1050 }, { "epoch": 0.2391353811149033, "grad_norm": 1.4873654976644912, "learning_rate": 1.2429716246787444e-06, "loss": 0.0312, "step": 1051 }, { "epoch": 0.23936291240045507, "grad_norm": 0.9449501930869969, "learning_rate": 1.242958256129813e-06, "loss": 0.0567, "step": 1052 }, { "epoch": 0.23959044368600682, "grad_norm": 1.6243209526441913, "learning_rate": 1.242944874950947e-06, "loss": 0.0647, "step": 1053 }, { "epoch": 0.2398179749715586, "grad_norm": 1.4569612537340462, "learning_rate": 1.2429314811424192e-06, "loss": 0.0758, "step": 1054 }, { "epoch": 0.24004550625711035, "grad_norm": 1.5825027859681509, "learning_rate": 1.242918074704504e-06, "loss": 0.0621, "step": 1055 }, { "epoch": 0.2402730375426621, "grad_norm": 1.0511003317201362, "learning_rate": 1.2429046556374747e-06, "loss": 0.037, "step": 1056 }, { "epoch": 0.24050056882821388, "grad_norm": 1.5928024296302492, "learning_rate": 1.2428912239416057e-06, "loss": 0.0453, "step": 1057 }, { "epoch": 0.24072810011376564, "grad_norm": 1.1044477930026537, "learning_rate": 1.242877779617172e-06, "loss": 0.0419, "step": 1058 }, { "epoch": 0.24095563139931742, "grad_norm": 1.3035104956572268, "learning_rate": 1.242864322664448e-06, "loss": 0.0531, "step": 1059 }, { "epoch": 0.24118316268486917, "grad_norm": 1.4034057924806689, "learning_rate": 1.2428508530837088e-06, "loss": 0.0753, "step": 1060 }, { "epoch": 0.24141069397042092, "grad_norm": 45.147488919338755, "learning_rate": 1.2428373708752298e-06, "loss": 0.4166, "step": 1061 }, { "epoch": 0.2416382252559727, "grad_norm": 1.804049071201822, "learning_rate": 1.2428238760392862e-06, "loss": 0.0881, "step": 1062 }, { "epoch": 0.24186575654152445, "grad_norm": 1.2662139418382417, "learning_rate": 1.2428103685761543e-06, "loss": 0.0592, "step": 1063 }, { "epoch": 0.24209328782707623, "grad_norm": 2.633674196903193, "learning_rate": 1.2427968484861097e-06, "loss": 0.1104, "step": 1064 }, { "epoch": 0.24232081911262798, "grad_norm": 1.437345835314508, "learning_rate": 1.2427833157694292e-06, "loss": 0.059, "step": 1065 }, { "epoch": 0.24254835039817976, "grad_norm": 1.4759939750585893, "learning_rate": 1.2427697704263892e-06, "loss": 0.0564, "step": 1066 }, { "epoch": 0.24277588168373151, "grad_norm": 1.4003983063384298, "learning_rate": 1.2427562124572663e-06, "loss": 0.0756, "step": 1067 }, { "epoch": 0.24300341296928327, "grad_norm": 1.2893015359567934, "learning_rate": 1.2427426418623377e-06, "loss": 0.0581, "step": 1068 }, { "epoch": 0.24323094425483505, "grad_norm": 0.9083897855186556, "learning_rate": 1.242729058641881e-06, "loss": 0.0353, "step": 1069 }, { "epoch": 0.2434584755403868, "grad_norm": 2.339042375493057, "learning_rate": 1.2427154627961737e-06, "loss": 0.1366, "step": 1070 }, { "epoch": 0.24368600682593858, "grad_norm": 2.1438385712601593, "learning_rate": 1.2427018543254935e-06, "loss": 0.082, "step": 1071 }, { "epoch": 0.24391353811149033, "grad_norm": 1.627313372349946, "learning_rate": 1.2426882332301187e-06, "loss": 0.076, "step": 1072 }, { "epoch": 0.24414106939704208, "grad_norm": 0.7791355914048339, "learning_rate": 1.2426745995103277e-06, "loss": 0.0309, "step": 1073 }, { "epoch": 0.24436860068259386, "grad_norm": 1.1486991259203154, "learning_rate": 1.242660953166399e-06, "loss": 0.0509, "step": 1074 }, { "epoch": 0.2445961319681456, "grad_norm": 1.7757375833942548, "learning_rate": 1.2426472941986117e-06, "loss": 0.0731, "step": 1075 }, { "epoch": 0.2448236632536974, "grad_norm": 1.4454374080664514, "learning_rate": 1.2426336226072449e-06, "loss": 0.0868, "step": 1076 }, { "epoch": 0.24505119453924915, "grad_norm": 1.0698078663784565, "learning_rate": 1.242619938392578e-06, "loss": 0.037, "step": 1077 }, { "epoch": 0.2452787258248009, "grad_norm": 1.4863388002827098, "learning_rate": 1.2426062415548907e-06, "loss": 0.0677, "step": 1078 }, { "epoch": 0.24550625711035268, "grad_norm": 0.8732937705027402, "learning_rate": 1.2425925320944628e-06, "loss": 0.0293, "step": 1079 }, { "epoch": 0.24573378839590443, "grad_norm": 1.2595288446902135, "learning_rate": 1.2425788100115747e-06, "loss": 0.056, "step": 1080 }, { "epoch": 0.2459613196814562, "grad_norm": 1.5998408970282165, "learning_rate": 1.2425650753065065e-06, "loss": 0.0764, "step": 1081 }, { "epoch": 0.24618885096700796, "grad_norm": 1.8027463727170587, "learning_rate": 1.2425513279795395e-06, "loss": 0.0615, "step": 1082 }, { "epoch": 0.24641638225255974, "grad_norm": 1.5654018901455387, "learning_rate": 1.2425375680309543e-06, "loss": 0.0661, "step": 1083 }, { "epoch": 0.2466439135381115, "grad_norm": 1.0676862677438599, "learning_rate": 1.2425237954610322e-06, "loss": 0.0469, "step": 1084 }, { "epoch": 0.24687144482366324, "grad_norm": 1.097337057406586, "learning_rate": 1.2425100102700547e-06, "loss": 0.0457, "step": 1085 }, { "epoch": 0.24709897610921502, "grad_norm": 1.7432427011380645, "learning_rate": 1.2424962124583033e-06, "loss": 0.0566, "step": 1086 }, { "epoch": 0.24732650739476678, "grad_norm": 1.2401111182610864, "learning_rate": 1.2424824020260603e-06, "loss": 0.0471, "step": 1087 }, { "epoch": 0.24755403868031856, "grad_norm": 1.6086762249686064, "learning_rate": 1.2424685789736077e-06, "loss": 0.0712, "step": 1088 }, { "epoch": 0.2477815699658703, "grad_norm": 2.223604715087674, "learning_rate": 1.2424547433012284e-06, "loss": 0.059, "step": 1089 }, { "epoch": 0.24800910125142206, "grad_norm": 1.763749275351197, "learning_rate": 1.2424408950092049e-06, "loss": 0.0768, "step": 1090 }, { "epoch": 0.24823663253697384, "grad_norm": 1.7828992657348108, "learning_rate": 1.2424270340978204e-06, "loss": 0.0794, "step": 1091 }, { "epoch": 0.2484641638225256, "grad_norm": 1.5001450767436246, "learning_rate": 1.2424131605673582e-06, "loss": 0.0693, "step": 1092 }, { "epoch": 0.24869169510807737, "grad_norm": 1.2301421279887537, "learning_rate": 1.2423992744181015e-06, "loss": 0.0512, "step": 1093 }, { "epoch": 0.24891922639362912, "grad_norm": 1.2844864578504145, "learning_rate": 1.2423853756503343e-06, "loss": 0.0447, "step": 1094 }, { "epoch": 0.24914675767918087, "grad_norm": 1.7196177827780137, "learning_rate": 1.2423714642643408e-06, "loss": 0.1182, "step": 1095 }, { "epoch": 0.24937428896473265, "grad_norm": 1.46657616590301, "learning_rate": 1.2423575402604051e-06, "loss": 0.0656, "step": 1096 }, { "epoch": 0.2496018202502844, "grad_norm": 1.326342830248352, "learning_rate": 1.2423436036388122e-06, "loss": 0.0544, "step": 1097 }, { "epoch": 0.24982935153583619, "grad_norm": 1.952008099521945, "learning_rate": 1.2423296543998465e-06, "loss": 0.078, "step": 1098 }, { "epoch": 0.25005688282138794, "grad_norm": 2.219926087555926, "learning_rate": 1.2423156925437932e-06, "loss": 0.0976, "step": 1099 }, { "epoch": 0.2502844141069397, "grad_norm": 1.4678969382750013, "learning_rate": 1.2423017180709376e-06, "loss": 0.0727, "step": 1100 }, { "epoch": 0.25051194539249144, "grad_norm": 1.9083840830954104, "learning_rate": 1.2422877309815656e-06, "loss": 0.0647, "step": 1101 }, { "epoch": 0.25073947667804325, "grad_norm": 1.3418972776139215, "learning_rate": 1.242273731275963e-06, "loss": 0.0616, "step": 1102 }, { "epoch": 0.250967007963595, "grad_norm": 1.0799686257073313, "learning_rate": 1.2422597189544155e-06, "loss": 0.0306, "step": 1103 }, { "epoch": 0.25119453924914675, "grad_norm": 1.4786340245230432, "learning_rate": 1.2422456940172101e-06, "loss": 0.0752, "step": 1104 }, { "epoch": 0.2514220705346985, "grad_norm": 1.3574468048142418, "learning_rate": 1.2422316564646331e-06, "loss": 0.0469, "step": 1105 }, { "epoch": 0.25164960182025026, "grad_norm": 1.0996737224918256, "learning_rate": 1.2422176062969713e-06, "loss": 0.06, "step": 1106 }, { "epoch": 0.25187713310580206, "grad_norm": 1.2079586241887055, "learning_rate": 1.2422035435145121e-06, "loss": 0.0417, "step": 1107 }, { "epoch": 0.2521046643913538, "grad_norm": 1.8484139155779635, "learning_rate": 1.2421894681175428e-06, "loss": 0.1201, "step": 1108 }, { "epoch": 0.25233219567690557, "grad_norm": 2.2964405416217977, "learning_rate": 1.2421753801063511e-06, "loss": 0.0661, "step": 1109 }, { "epoch": 0.2525597269624573, "grad_norm": 1.233798632970271, "learning_rate": 1.2421612794812248e-06, "loss": 0.0533, "step": 1110 }, { "epoch": 0.25278725824800913, "grad_norm": 0.8794754874897922, "learning_rate": 1.2421471662424525e-06, "loss": 0.0374, "step": 1111 }, { "epoch": 0.2530147895335609, "grad_norm": 1.1655828487104456, "learning_rate": 1.2421330403903222e-06, "loss": 0.0417, "step": 1112 }, { "epoch": 0.25324232081911263, "grad_norm": 1.978604856786128, "learning_rate": 1.2421189019251228e-06, "loss": 0.0613, "step": 1113 }, { "epoch": 0.2534698521046644, "grad_norm": 2.776603540550928, "learning_rate": 1.2421047508471433e-06, "loss": 0.0945, "step": 1114 }, { "epoch": 0.25369738339021614, "grad_norm": 1.4281147024591225, "learning_rate": 1.242090587156673e-06, "loss": 0.0634, "step": 1115 }, { "epoch": 0.25392491467576794, "grad_norm": 1.9687323288611094, "learning_rate": 1.242076410854001e-06, "loss": 0.0637, "step": 1116 }, { "epoch": 0.2541524459613197, "grad_norm": 0.9824968425418638, "learning_rate": 1.2420622219394174e-06, "loss": 0.0361, "step": 1117 }, { "epoch": 0.25437997724687145, "grad_norm": 1.1748756467082628, "learning_rate": 1.2420480204132117e-06, "loss": 0.0438, "step": 1118 }, { "epoch": 0.2546075085324232, "grad_norm": 1.1687970818584577, "learning_rate": 1.242033806275675e-06, "loss": 0.0517, "step": 1119 }, { "epoch": 0.25483503981797495, "grad_norm": 0.7243821856689469, "learning_rate": 1.2420195795270973e-06, "loss": 0.0341, "step": 1120 }, { "epoch": 0.25506257110352676, "grad_norm": 1.5334379290667297, "learning_rate": 1.2420053401677693e-06, "loss": 0.0834, "step": 1121 }, { "epoch": 0.2552901023890785, "grad_norm": 1.1475477659525273, "learning_rate": 1.241991088197982e-06, "loss": 0.0409, "step": 1122 }, { "epoch": 0.25551763367463026, "grad_norm": 1.0609745683681193, "learning_rate": 1.241976823618027e-06, "loss": 0.0521, "step": 1123 }, { "epoch": 0.255745164960182, "grad_norm": 2.0039923086463602, "learning_rate": 1.241962546428196e-06, "loss": 0.0912, "step": 1124 }, { "epoch": 0.25597269624573377, "grad_norm": 2.1894810646011846, "learning_rate": 1.24194825662878e-06, "loss": 0.0785, "step": 1125 }, { "epoch": 0.2562002275312856, "grad_norm": 1.4851813238564915, "learning_rate": 1.2419339542200715e-06, "loss": 0.0688, "step": 1126 }, { "epoch": 0.2564277588168373, "grad_norm": 1.2728572943407013, "learning_rate": 1.241919639202363e-06, "loss": 0.0794, "step": 1127 }, { "epoch": 0.2566552901023891, "grad_norm": 1.8082441837013492, "learning_rate": 1.2419053115759468e-06, "loss": 0.0677, "step": 1128 }, { "epoch": 0.25688282138794083, "grad_norm": 2.0609613364729267, "learning_rate": 1.2418909713411161e-06, "loss": 0.08, "step": 1129 }, { "epoch": 0.2571103526734926, "grad_norm": 1.533485951338072, "learning_rate": 1.2418766184981634e-06, "loss": 0.0606, "step": 1130 }, { "epoch": 0.2573378839590444, "grad_norm": 1.4347136829017464, "learning_rate": 1.2418622530473825e-06, "loss": 0.0825, "step": 1131 }, { "epoch": 0.25756541524459614, "grad_norm": 0.8595888516934124, "learning_rate": 1.2418478749890672e-06, "loss": 0.0351, "step": 1132 }, { "epoch": 0.2577929465301479, "grad_norm": 0.8988536535150867, "learning_rate": 1.2418334843235105e-06, "loss": 0.0465, "step": 1133 }, { "epoch": 0.25802047781569964, "grad_norm": 0.8536639563076862, "learning_rate": 1.2418190810510075e-06, "loss": 0.0258, "step": 1134 }, { "epoch": 0.2582480091012514, "grad_norm": 1.8139538056092865, "learning_rate": 1.2418046651718518e-06, "loss": 0.0545, "step": 1135 }, { "epoch": 0.2584755403868032, "grad_norm": 1.0709084305362149, "learning_rate": 1.2417902366863386e-06, "loss": 0.0488, "step": 1136 }, { "epoch": 0.25870307167235496, "grad_norm": 1.5965977494143468, "learning_rate": 1.2417757955947623e-06, "loss": 0.0549, "step": 1137 }, { "epoch": 0.2589306029579067, "grad_norm": 1.258938457199683, "learning_rate": 1.2417613418974187e-06, "loss": 0.0422, "step": 1138 }, { "epoch": 0.25915813424345846, "grad_norm": 0.9954800667569594, "learning_rate": 1.2417468755946025e-06, "loss": 0.0297, "step": 1139 }, { "epoch": 0.2593856655290102, "grad_norm": 1.410994342027921, "learning_rate": 1.2417323966866097e-06, "loss": 0.054, "step": 1140 }, { "epoch": 0.259613196814562, "grad_norm": 1.5107757792275114, "learning_rate": 1.2417179051737364e-06, "loss": 0.0589, "step": 1141 }, { "epoch": 0.25984072810011377, "grad_norm": 1.6398671824805653, "learning_rate": 1.2417034010562784e-06, "loss": 0.0666, "step": 1142 }, { "epoch": 0.2600682593856655, "grad_norm": 1.3520000022781709, "learning_rate": 1.2416888843345323e-06, "loss": 0.0416, "step": 1143 }, { "epoch": 0.2602957906712173, "grad_norm": 2.915495796493339, "learning_rate": 1.2416743550087951e-06, "loss": 0.1129, "step": 1144 }, { "epoch": 0.2605233219567691, "grad_norm": 1.2846258547817568, "learning_rate": 1.241659813079363e-06, "loss": 0.0579, "step": 1145 }, { "epoch": 0.26075085324232083, "grad_norm": 2.7427597460101936, "learning_rate": 1.2416452585465342e-06, "loss": 0.0702, "step": 1146 }, { "epoch": 0.2609783845278726, "grad_norm": 0.5884709313504546, "learning_rate": 1.2416306914106053e-06, "loss": 0.0237, "step": 1147 }, { "epoch": 0.26120591581342434, "grad_norm": 1.2939407540414039, "learning_rate": 1.2416161116718744e-06, "loss": 0.0583, "step": 1148 }, { "epoch": 0.2614334470989761, "grad_norm": 1.96584405463017, "learning_rate": 1.2416015193306397e-06, "loss": 0.0597, "step": 1149 }, { "epoch": 0.2616609783845279, "grad_norm": 1.367578620212531, "learning_rate": 1.241586914387199e-06, "loss": 0.0525, "step": 1150 }, { "epoch": 0.26188850967007965, "grad_norm": 1.5579301663665894, "learning_rate": 1.2415722968418508e-06, "loss": 0.0701, "step": 1151 }, { "epoch": 0.2621160409556314, "grad_norm": 2.113746394897606, "learning_rate": 1.2415576666948945e-06, "loss": 0.1158, "step": 1152 }, { "epoch": 0.26234357224118315, "grad_norm": 1.450032906329583, "learning_rate": 1.2415430239466283e-06, "loss": 0.0602, "step": 1153 }, { "epoch": 0.2625711035267349, "grad_norm": 1.3272413443266664, "learning_rate": 1.241528368597352e-06, "loss": 0.0523, "step": 1154 }, { "epoch": 0.2627986348122867, "grad_norm": 1.179039651442814, "learning_rate": 1.2415137006473649e-06, "loss": 0.0647, "step": 1155 }, { "epoch": 0.26302616609783847, "grad_norm": 1.145520426253407, "learning_rate": 1.241499020096967e-06, "loss": 0.0477, "step": 1156 }, { "epoch": 0.2632536973833902, "grad_norm": 1.328947872893636, "learning_rate": 1.2414843269464579e-06, "loss": 0.0452, "step": 1157 }, { "epoch": 0.26348122866894197, "grad_norm": 1.8843518385578133, "learning_rate": 1.2414696211961386e-06, "loss": 0.0603, "step": 1158 }, { "epoch": 0.2637087599544937, "grad_norm": 2.1467164256006943, "learning_rate": 1.2414549028463087e-06, "loss": 0.0908, "step": 1159 }, { "epoch": 0.26393629124004553, "grad_norm": 1.1918005997211691, "learning_rate": 1.2414401718972703e-06, "loss": 0.0576, "step": 1160 }, { "epoch": 0.2641638225255973, "grad_norm": 1.2786644068555166, "learning_rate": 1.2414254283493232e-06, "loss": 0.0416, "step": 1161 }, { "epoch": 0.26439135381114903, "grad_norm": 0.9452023930772737, "learning_rate": 1.2414106722027694e-06, "loss": 0.0537, "step": 1162 }, { "epoch": 0.2646188850967008, "grad_norm": 1.1756293508640332, "learning_rate": 1.2413959034579104e-06, "loss": 0.0511, "step": 1163 }, { "epoch": 0.26484641638225254, "grad_norm": 1.0472201568449635, "learning_rate": 1.2413811221150478e-06, "loss": 0.0442, "step": 1164 }, { "epoch": 0.26507394766780434, "grad_norm": 1.916314085543847, "learning_rate": 1.2413663281744843e-06, "loss": 0.0768, "step": 1165 }, { "epoch": 0.2653014789533561, "grad_norm": 1.0894771604167501, "learning_rate": 1.2413515216365216e-06, "loss": 0.0602, "step": 1166 }, { "epoch": 0.26552901023890785, "grad_norm": 1.00975683219884, "learning_rate": 1.2413367025014628e-06, "loss": 0.0432, "step": 1167 }, { "epoch": 0.2657565415244596, "grad_norm": 2.1765273948785913, "learning_rate": 1.2413218707696103e-06, "loss": 0.1186, "step": 1168 }, { "epoch": 0.26598407281001135, "grad_norm": 1.5268905455044306, "learning_rate": 1.2413070264412677e-06, "loss": 0.0671, "step": 1169 }, { "epoch": 0.26621160409556316, "grad_norm": 0.8198195830517153, "learning_rate": 1.2412921695167381e-06, "loss": 0.0303, "step": 1170 }, { "epoch": 0.2664391353811149, "grad_norm": 0.916743439302086, "learning_rate": 1.2412772999963253e-06, "loss": 0.0519, "step": 1171 }, { "epoch": 0.26666666666666666, "grad_norm": 0.9592599841051566, "learning_rate": 1.2412624178803332e-06, "loss": 0.0465, "step": 1172 }, { "epoch": 0.2668941979522184, "grad_norm": 1.2610976325050403, "learning_rate": 1.2412475231690656e-06, "loss": 0.0345, "step": 1173 }, { "epoch": 0.26712172923777017, "grad_norm": 1.5016024974886808, "learning_rate": 1.2412326158628275e-06, "loss": 0.0714, "step": 1174 }, { "epoch": 0.267349260523322, "grad_norm": 1.4314762978749895, "learning_rate": 1.2412176959619232e-06, "loss": 0.0591, "step": 1175 }, { "epoch": 0.2675767918088737, "grad_norm": 2.2495091900368562, "learning_rate": 1.2412027634666578e-06, "loss": 0.1059, "step": 1176 }, { "epoch": 0.2678043230944255, "grad_norm": 1.706282565066305, "learning_rate": 1.2411878183773366e-06, "loss": 0.0665, "step": 1177 }, { "epoch": 0.26803185437997723, "grad_norm": 0.7041103017092823, "learning_rate": 1.2411728606942647e-06, "loss": 0.0218, "step": 1178 }, { "epoch": 0.26825938566552904, "grad_norm": 0.9725397335622218, "learning_rate": 1.241157890417748e-06, "loss": 0.0452, "step": 1179 }, { "epoch": 0.2684869169510808, "grad_norm": 1.9030708757385315, "learning_rate": 1.2411429075480923e-06, "loss": 0.0827, "step": 1180 }, { "epoch": 0.26871444823663254, "grad_norm": 1.1973351649414545, "learning_rate": 1.2411279120856042e-06, "loss": 0.0476, "step": 1181 }, { "epoch": 0.2689419795221843, "grad_norm": 1.1061135820635588, "learning_rate": 1.24111290403059e-06, "loss": 0.0387, "step": 1182 }, { "epoch": 0.26916951080773605, "grad_norm": 1.1523226678516665, "learning_rate": 1.2410978833833564e-06, "loss": 0.0614, "step": 1183 }, { "epoch": 0.26939704209328785, "grad_norm": 1.0514874828378626, "learning_rate": 1.2410828501442104e-06, "loss": 0.0439, "step": 1184 }, { "epoch": 0.2696245733788396, "grad_norm": 1.5362061949507542, "learning_rate": 1.2410678043134591e-06, "loss": 0.0804, "step": 1185 }, { "epoch": 0.26985210466439136, "grad_norm": 1.8173832515444213, "learning_rate": 1.2410527458914103e-06, "loss": 0.063, "step": 1186 }, { "epoch": 0.2700796359499431, "grad_norm": 1.556320593990832, "learning_rate": 1.2410376748783714e-06, "loss": 0.0795, "step": 1187 }, { "epoch": 0.27030716723549486, "grad_norm": 1.321348596107485, "learning_rate": 1.241022591274651e-06, "loss": 0.0802, "step": 1188 }, { "epoch": 0.27053469852104667, "grad_norm": 1.0652846616165823, "learning_rate": 1.2410074950805567e-06, "loss": 0.036, "step": 1189 }, { "epoch": 0.2707622298065984, "grad_norm": 1.2080698971206665, "learning_rate": 1.2409923862963973e-06, "loss": 0.0488, "step": 1190 }, { "epoch": 0.27098976109215017, "grad_norm": 1.0230330657283313, "learning_rate": 1.240977264922482e-06, "loss": 0.0408, "step": 1191 }, { "epoch": 0.2712172923777019, "grad_norm": 1.243454259162499, "learning_rate": 1.2409621309591195e-06, "loss": 0.0359, "step": 1192 }, { "epoch": 0.2714448236632537, "grad_norm": 1.3001395180964366, "learning_rate": 1.2409469844066188e-06, "loss": 0.0454, "step": 1193 }, { "epoch": 0.2716723549488055, "grad_norm": 1.4147947246201094, "learning_rate": 1.2409318252652899e-06, "loss": 0.0673, "step": 1194 }, { "epoch": 0.27189988623435724, "grad_norm": 2.0607810868090977, "learning_rate": 1.2409166535354428e-06, "loss": 0.1099, "step": 1195 }, { "epoch": 0.272127417519909, "grad_norm": 0.8856660147431201, "learning_rate": 1.2409014692173872e-06, "loss": 0.037, "step": 1196 }, { "epoch": 0.27235494880546074, "grad_norm": 0.8588236308520142, "learning_rate": 1.240886272311433e-06, "loss": 0.0317, "step": 1197 }, { "epoch": 0.2725824800910125, "grad_norm": 1.2966972282013072, "learning_rate": 1.240871062817892e-06, "loss": 0.0519, "step": 1198 }, { "epoch": 0.2728100113765643, "grad_norm": 1.0141950682821237, "learning_rate": 1.240855840737074e-06, "loss": 0.0362, "step": 1199 }, { "epoch": 0.27303754266211605, "grad_norm": 1.1327532340524933, "learning_rate": 1.2408406060692909e-06, "loss": 0.0411, "step": 1200 }, { "epoch": 0.2732650739476678, "grad_norm": 38.94813584323084, "learning_rate": 1.2408253588148532e-06, "loss": 0.4954, "step": 1201 }, { "epoch": 0.27349260523321955, "grad_norm": 1.3419172753176944, "learning_rate": 1.2408100989740735e-06, "loss": 0.0629, "step": 1202 }, { "epoch": 0.2737201365187713, "grad_norm": 1.164418605632623, "learning_rate": 1.2407948265472628e-06, "loss": 0.0327, "step": 1203 }, { "epoch": 0.2739476678043231, "grad_norm": 1.2988536193125668, "learning_rate": 1.2407795415347336e-06, "loss": 0.0567, "step": 1204 }, { "epoch": 0.27417519908987487, "grad_norm": 1.361893473954927, "learning_rate": 1.2407642439367986e-06, "loss": 0.0438, "step": 1205 }, { "epoch": 0.2744027303754266, "grad_norm": 1.0391897172547626, "learning_rate": 1.24074893375377e-06, "loss": 0.0557, "step": 1206 }, { "epoch": 0.27463026166097837, "grad_norm": 0.9962942412430917, "learning_rate": 1.2407336109859607e-06, "loss": 0.0371, "step": 1207 }, { "epoch": 0.2748577929465301, "grad_norm": 1.215727870574286, "learning_rate": 1.2407182756336844e-06, "loss": 0.0437, "step": 1208 }, { "epoch": 0.27508532423208193, "grad_norm": 1.8508042963184421, "learning_rate": 1.240702927697254e-06, "loss": 0.0626, "step": 1209 }, { "epoch": 0.2753128555176337, "grad_norm": 1.8022236633570414, "learning_rate": 1.2406875671769837e-06, "loss": 0.0584, "step": 1210 }, { "epoch": 0.27554038680318543, "grad_norm": 1.606815955580643, "learning_rate": 1.2406721940731866e-06, "loss": 0.0599, "step": 1211 }, { "epoch": 0.2757679180887372, "grad_norm": 1.0652074757978982, "learning_rate": 1.2406568083861776e-06, "loss": 0.0531, "step": 1212 }, { "epoch": 0.27599544937428894, "grad_norm": 1.4187270182072722, "learning_rate": 1.2406414101162708e-06, "loss": 0.0528, "step": 1213 }, { "epoch": 0.27622298065984074, "grad_norm": 1.0888371326459914, "learning_rate": 1.2406259992637815e-06, "loss": 0.0552, "step": 1214 }, { "epoch": 0.2764505119453925, "grad_norm": 2.1125996368288553, "learning_rate": 1.240610575829024e-06, "loss": 0.0735, "step": 1215 }, { "epoch": 0.27667804323094425, "grad_norm": 1.288876395538611, "learning_rate": 1.2405951398123136e-06, "loss": 0.0553, "step": 1216 }, { "epoch": 0.276905574516496, "grad_norm": 1.711303610363005, "learning_rate": 1.2405796912139662e-06, "loss": 0.0673, "step": 1217 }, { "epoch": 0.2771331058020478, "grad_norm": 1.3330966398511392, "learning_rate": 1.240564230034297e-06, "loss": 0.0347, "step": 1218 }, { "epoch": 0.27736063708759956, "grad_norm": 0.9030680288819154, "learning_rate": 1.2405487562736226e-06, "loss": 0.0365, "step": 1219 }, { "epoch": 0.2775881683731513, "grad_norm": 1.20973479021967, "learning_rate": 1.240533269932259e-06, "loss": 0.0647, "step": 1220 }, { "epoch": 0.27781569965870306, "grad_norm": 1.2330537480211692, "learning_rate": 1.2405177710105223e-06, "loss": 0.0415, "step": 1221 }, { "epoch": 0.2780432309442548, "grad_norm": 1.4447131783576561, "learning_rate": 1.24050225950873e-06, "loss": 0.04, "step": 1222 }, { "epoch": 0.2782707622298066, "grad_norm": 1.5348321099487485, "learning_rate": 1.2404867354271984e-06, "loss": 0.0699, "step": 1223 }, { "epoch": 0.2784982935153584, "grad_norm": 1.103797642514096, "learning_rate": 1.2404711987662452e-06, "loss": 0.0501, "step": 1224 }, { "epoch": 0.2787258248009101, "grad_norm": 1.8543277255058668, "learning_rate": 1.240455649526188e-06, "loss": 0.0779, "step": 1225 }, { "epoch": 0.2789533560864619, "grad_norm": 1.461515751555683, "learning_rate": 1.2404400877073446e-06, "loss": 0.0589, "step": 1226 }, { "epoch": 0.27918088737201363, "grad_norm": 1.7282210636065953, "learning_rate": 1.2404245133100328e-06, "loss": 0.0616, "step": 1227 }, { "epoch": 0.27940841865756544, "grad_norm": 0.9274643279367238, "learning_rate": 1.240408926334571e-06, "loss": 0.0259, "step": 1228 }, { "epoch": 0.2796359499431172, "grad_norm": 1.033501452595113, "learning_rate": 1.240393326781278e-06, "loss": 0.0388, "step": 1229 }, { "epoch": 0.27986348122866894, "grad_norm": 1.2277681625095047, "learning_rate": 1.2403777146504722e-06, "loss": 0.0417, "step": 1230 }, { "epoch": 0.2800910125142207, "grad_norm": 1.6056103902463323, "learning_rate": 1.240362089942473e-06, "loss": 0.0869, "step": 1231 }, { "epoch": 0.28031854379977245, "grad_norm": 0.9379340266438139, "learning_rate": 1.2403464526575997e-06, "loss": 0.0288, "step": 1232 }, { "epoch": 0.28054607508532425, "grad_norm": 1.791400262370738, "learning_rate": 1.240330802796172e-06, "loss": 0.0494, "step": 1233 }, { "epoch": 0.280773606370876, "grad_norm": 1.5180897393064074, "learning_rate": 1.2403151403585093e-06, "loss": 0.0555, "step": 1234 }, { "epoch": 0.28100113765642776, "grad_norm": 1.3796072286824306, "learning_rate": 1.240299465344932e-06, "loss": 0.0405, "step": 1235 }, { "epoch": 0.2812286689419795, "grad_norm": 0.968124324669557, "learning_rate": 1.2402837777557608e-06, "loss": 0.0462, "step": 1236 }, { "epoch": 0.28145620022753126, "grad_norm": 2.8732703815868157, "learning_rate": 1.240268077591316e-06, "loss": 0.1432, "step": 1237 }, { "epoch": 0.28168373151308307, "grad_norm": 1.7096825525845536, "learning_rate": 1.2402523648519184e-06, "loss": 0.0709, "step": 1238 }, { "epoch": 0.2819112627986348, "grad_norm": 1.4537193421874715, "learning_rate": 1.2402366395378892e-06, "loss": 0.0477, "step": 1239 }, { "epoch": 0.2821387940841866, "grad_norm": 1.0525576505096776, "learning_rate": 1.24022090164955e-06, "loss": 0.0381, "step": 1240 }, { "epoch": 0.2823663253697383, "grad_norm": 2.0691434586350135, "learning_rate": 1.240205151187222e-06, "loss": 0.0749, "step": 1241 }, { "epoch": 0.2825938566552901, "grad_norm": 1.4855492490109166, "learning_rate": 1.2401893881512278e-06, "loss": 0.0486, "step": 1242 }, { "epoch": 0.2828213879408419, "grad_norm": 2.3507672202858427, "learning_rate": 1.240173612541889e-06, "loss": 0.1187, "step": 1243 }, { "epoch": 0.28304891922639364, "grad_norm": 1.4271939145537942, "learning_rate": 1.2401578243595281e-06, "loss": 0.0459, "step": 1244 }, { "epoch": 0.2832764505119454, "grad_norm": 0.7153929012800043, "learning_rate": 1.2401420236044678e-06, "loss": 0.0203, "step": 1245 }, { "epoch": 0.28350398179749714, "grad_norm": 1.8926578555618896, "learning_rate": 1.2401262102770314e-06, "loss": 0.0939, "step": 1246 }, { "epoch": 0.2837315130830489, "grad_norm": 1.5820745105272767, "learning_rate": 1.2401103843775416e-06, "loss": 0.0885, "step": 1247 }, { "epoch": 0.2839590443686007, "grad_norm": 1.3120392754023362, "learning_rate": 1.240094545906322e-06, "loss": 0.0629, "step": 1248 }, { "epoch": 0.28418657565415245, "grad_norm": 1.2780609773747653, "learning_rate": 1.2400786948636966e-06, "loss": 0.0669, "step": 1249 }, { "epoch": 0.2844141069397042, "grad_norm": 1.4394753485796945, "learning_rate": 1.2400628312499892e-06, "loss": 0.0457, "step": 1250 }, { "epoch": 0.28464163822525596, "grad_norm": 1.4902299382519306, "learning_rate": 1.2400469550655239e-06, "loss": 0.0788, "step": 1251 }, { "epoch": 0.28486916951080776, "grad_norm": 2.1373698726270614, "learning_rate": 1.240031066310625e-06, "loss": 0.0709, "step": 1252 }, { "epoch": 0.2850967007963595, "grad_norm": 2.6364110769644467, "learning_rate": 1.2400151649856176e-06, "loss": 0.1474, "step": 1253 }, { "epoch": 0.28532423208191127, "grad_norm": 1.54489334981108, "learning_rate": 1.2399992510908266e-06, "loss": 0.0681, "step": 1254 }, { "epoch": 0.285551763367463, "grad_norm": 1.2035560270921895, "learning_rate": 1.2399833246265772e-06, "loss": 0.0624, "step": 1255 }, { "epoch": 0.28577929465301477, "grad_norm": 2.3200664566107427, "learning_rate": 1.2399673855931951e-06, "loss": 0.0948, "step": 1256 }, { "epoch": 0.2860068259385666, "grad_norm": 0.911546303051303, "learning_rate": 1.2399514339910058e-06, "loss": 0.0371, "step": 1257 }, { "epoch": 0.28623435722411833, "grad_norm": 1.1459525470306917, "learning_rate": 1.2399354698203353e-06, "loss": 0.063, "step": 1258 }, { "epoch": 0.2864618885096701, "grad_norm": 5.90970905115882, "learning_rate": 1.2399194930815103e-06, "loss": 0.2573, "step": 1259 }, { "epoch": 0.28668941979522183, "grad_norm": 1.6075560430361318, "learning_rate": 1.2399035037748567e-06, "loss": 0.0712, "step": 1260 }, { "epoch": 0.2869169510807736, "grad_norm": 1.059504335911741, "learning_rate": 1.2398875019007017e-06, "loss": 0.0336, "step": 1261 }, { "epoch": 0.2871444823663254, "grad_norm": 1.0386465638630533, "learning_rate": 1.2398714874593722e-06, "loss": 0.028, "step": 1262 }, { "epoch": 0.28737201365187715, "grad_norm": 1.2034325585223464, "learning_rate": 1.2398554604511958e-06, "loss": 0.0426, "step": 1263 }, { "epoch": 0.2875995449374289, "grad_norm": 1.130614264631673, "learning_rate": 1.2398394208764998e-06, "loss": 0.0449, "step": 1264 }, { "epoch": 0.28782707622298065, "grad_norm": 1.8759349581624183, "learning_rate": 1.239823368735612e-06, "loss": 0.0738, "step": 1265 }, { "epoch": 0.2880546075085324, "grad_norm": 1.534877012833876, "learning_rate": 1.2398073040288605e-06, "loss": 0.0565, "step": 1266 }, { "epoch": 0.2882821387940842, "grad_norm": 1.1199756497261975, "learning_rate": 1.2397912267565738e-06, "loss": 0.0434, "step": 1267 }, { "epoch": 0.28850967007963596, "grad_norm": 1.2767125637243415, "learning_rate": 1.2397751369190803e-06, "loss": 0.053, "step": 1268 }, { "epoch": 0.2887372013651877, "grad_norm": 1.6866930161803586, "learning_rate": 1.2397590345167088e-06, "loss": 0.0977, "step": 1269 }, { "epoch": 0.28896473265073946, "grad_norm": 1.2795958593835728, "learning_rate": 1.2397429195497887e-06, "loss": 0.0435, "step": 1270 }, { "epoch": 0.2891922639362912, "grad_norm": 1.2653384358613777, "learning_rate": 1.239726792018649e-06, "loss": 0.0366, "step": 1271 }, { "epoch": 0.289419795221843, "grad_norm": 1.175013233181025, "learning_rate": 1.2397106519236197e-06, "loss": 0.0589, "step": 1272 }, { "epoch": 0.2896473265073948, "grad_norm": 0.8118499461560013, "learning_rate": 1.2396944992650303e-06, "loss": 0.0341, "step": 1273 }, { "epoch": 0.2898748577929465, "grad_norm": 1.7245088180198505, "learning_rate": 1.2396783340432111e-06, "loss": 0.0712, "step": 1274 }, { "epoch": 0.2901023890784983, "grad_norm": 1.453392643848313, "learning_rate": 1.2396621562584925e-06, "loss": 0.0552, "step": 1275 }, { "epoch": 0.29032992036405003, "grad_norm": 0.9686308053813132, "learning_rate": 1.2396459659112052e-06, "loss": 0.0298, "step": 1276 }, { "epoch": 0.29055745164960184, "grad_norm": 1.2190161105816313, "learning_rate": 1.23962976300168e-06, "loss": 0.0657, "step": 1277 }, { "epoch": 0.2907849829351536, "grad_norm": 1.4400910874681545, "learning_rate": 1.2396135475302478e-06, "loss": 0.0689, "step": 1278 }, { "epoch": 0.29101251422070534, "grad_norm": 1.6072456813020977, "learning_rate": 1.2395973194972406e-06, "loss": 0.0519, "step": 1279 }, { "epoch": 0.2912400455062571, "grad_norm": 0.9134133498246579, "learning_rate": 1.2395810789029898e-06, "loss": 0.0305, "step": 1280 }, { "epoch": 0.29146757679180885, "grad_norm": 1.4368452567354464, "learning_rate": 1.2395648257478271e-06, "loss": 0.0842, "step": 1281 }, { "epoch": 0.29169510807736065, "grad_norm": 1.2627048123993603, "learning_rate": 1.239548560032085e-06, "loss": 0.0556, "step": 1282 }, { "epoch": 0.2919226393629124, "grad_norm": 1.8157780106000843, "learning_rate": 1.239532281756096e-06, "loss": 0.0812, "step": 1283 }, { "epoch": 0.29215017064846416, "grad_norm": 1.7031472200076767, "learning_rate": 1.2395159909201924e-06, "loss": 0.1025, "step": 1284 }, { "epoch": 0.2923777019340159, "grad_norm": 1.4512078193177647, "learning_rate": 1.2394996875247075e-06, "loss": 0.0878, "step": 1285 }, { "epoch": 0.2926052332195677, "grad_norm": 1.1558527774450407, "learning_rate": 1.2394833715699743e-06, "loss": 0.0517, "step": 1286 }, { "epoch": 0.29283276450511947, "grad_norm": 1.1961425724018928, "learning_rate": 1.2394670430563264e-06, "loss": 0.0822, "step": 1287 }, { "epoch": 0.2930602957906712, "grad_norm": 1.1257477613693543, "learning_rate": 1.2394507019840974e-06, "loss": 0.0433, "step": 1288 }, { "epoch": 0.293287827076223, "grad_norm": 1.7490981885592791, "learning_rate": 1.2394343483536215e-06, "loss": 0.0731, "step": 1289 }, { "epoch": 0.2935153583617747, "grad_norm": 0.7736925600695324, "learning_rate": 1.2394179821652326e-06, "loss": 0.048, "step": 1290 }, { "epoch": 0.29374288964732653, "grad_norm": 1.2742121789528693, "learning_rate": 1.2394016034192654e-06, "loss": 0.0523, "step": 1291 }, { "epoch": 0.2939704209328783, "grad_norm": 1.4246615440711465, "learning_rate": 1.2393852121160548e-06, "loss": 0.0813, "step": 1292 }, { "epoch": 0.29419795221843004, "grad_norm": 1.656527882332843, "learning_rate": 1.2393688082559357e-06, "loss": 0.0699, "step": 1293 }, { "epoch": 0.2944254835039818, "grad_norm": 1.2504233198660668, "learning_rate": 1.2393523918392433e-06, "loss": 0.0457, "step": 1294 }, { "epoch": 0.29465301478953354, "grad_norm": 2.1835557352018458, "learning_rate": 1.2393359628663133e-06, "loss": 0.1184, "step": 1295 }, { "epoch": 0.29488054607508535, "grad_norm": 1.1347886652336971, "learning_rate": 1.239319521337481e-06, "loss": 0.0655, "step": 1296 }, { "epoch": 0.2951080773606371, "grad_norm": 1.1763338033692377, "learning_rate": 1.2393030672530828e-06, "loss": 0.0568, "step": 1297 }, { "epoch": 0.29533560864618885, "grad_norm": 1.3990303438853744, "learning_rate": 1.239286600613455e-06, "loss": 0.041, "step": 1298 }, { "epoch": 0.2955631399317406, "grad_norm": 0.9401022615309382, "learning_rate": 1.2392701214189343e-06, "loss": 0.0349, "step": 1299 }, { "epoch": 0.29579067121729236, "grad_norm": 2.3948134916737516, "learning_rate": 1.2392536296698571e-06, "loss": 0.1472, "step": 1300 }, { "epoch": 0.29601820250284416, "grad_norm": 1.320538326644519, "learning_rate": 1.2392371253665605e-06, "loss": 0.0767, "step": 1301 }, { "epoch": 0.2962457337883959, "grad_norm": 1.354776649937033, "learning_rate": 1.2392206085093823e-06, "loss": 0.0584, "step": 1302 }, { "epoch": 0.29647326507394767, "grad_norm": 1.4302761520764427, "learning_rate": 1.2392040790986594e-06, "loss": 0.0823, "step": 1303 }, { "epoch": 0.2967007963594994, "grad_norm": 1.4047156580871927, "learning_rate": 1.2391875371347303e-06, "loss": 0.0549, "step": 1304 }, { "epoch": 0.29692832764505117, "grad_norm": 0.7970562047989252, "learning_rate": 1.2391709826179327e-06, "loss": 0.0336, "step": 1305 }, { "epoch": 0.297155858930603, "grad_norm": 1.3727411496357969, "learning_rate": 1.239154415548605e-06, "loss": 0.0392, "step": 1306 }, { "epoch": 0.29738339021615473, "grad_norm": 0.9054264955476324, "learning_rate": 1.2391378359270859e-06, "loss": 0.0389, "step": 1307 }, { "epoch": 0.2976109215017065, "grad_norm": 1.814490926829392, "learning_rate": 1.2391212437537138e-06, "loss": 0.072, "step": 1308 }, { "epoch": 0.29783845278725823, "grad_norm": 1.7618287946971354, "learning_rate": 1.2391046390288287e-06, "loss": 0.058, "step": 1309 }, { "epoch": 0.29806598407281, "grad_norm": 1.5269715641275399, "learning_rate": 1.2390880217527692e-06, "loss": 0.0656, "step": 1310 }, { "epoch": 0.2982935153583618, "grad_norm": 1.262642874805734, "learning_rate": 1.2390713919258752e-06, "loss": 0.0532, "step": 1311 }, { "epoch": 0.29852104664391355, "grad_norm": 1.9617860637916495, "learning_rate": 1.2390547495484866e-06, "loss": 0.1163, "step": 1312 }, { "epoch": 0.2987485779294653, "grad_norm": 1.288057411790914, "learning_rate": 1.2390380946209436e-06, "loss": 0.0508, "step": 1313 }, { "epoch": 0.29897610921501705, "grad_norm": 1.1943768298905832, "learning_rate": 1.2390214271435863e-06, "loss": 0.0388, "step": 1314 }, { "epoch": 0.2992036405005688, "grad_norm": 1.291453023419144, "learning_rate": 1.2390047471167557e-06, "loss": 0.0459, "step": 1315 }, { "epoch": 0.2994311717861206, "grad_norm": 1.014250520116699, "learning_rate": 1.2389880545407926e-06, "loss": 0.0415, "step": 1316 }, { "epoch": 0.29965870307167236, "grad_norm": 0.8604020785970597, "learning_rate": 1.2389713494160379e-06, "loss": 0.0256, "step": 1317 }, { "epoch": 0.2998862343572241, "grad_norm": 1.5717854535146285, "learning_rate": 1.2389546317428335e-06, "loss": 0.0659, "step": 1318 }, { "epoch": 0.30011376564277586, "grad_norm": 1.1712963748867264, "learning_rate": 1.2389379015215208e-06, "loss": 0.0576, "step": 1319 }, { "epoch": 0.3003412969283277, "grad_norm": 1.9673409422404795, "learning_rate": 1.2389211587524416e-06, "loss": 0.0724, "step": 1320 }, { "epoch": 0.3005688282138794, "grad_norm": 1.7254424987946275, "learning_rate": 1.2389044034359383e-06, "loss": 0.0667, "step": 1321 }, { "epoch": 0.3007963594994312, "grad_norm": 1.7304834437682375, "learning_rate": 1.2388876355723533e-06, "loss": 0.0669, "step": 1322 }, { "epoch": 0.30102389078498293, "grad_norm": 1.5970096808225174, "learning_rate": 1.2388708551620295e-06, "loss": 0.0606, "step": 1323 }, { "epoch": 0.3012514220705347, "grad_norm": 1.4122576326239549, "learning_rate": 1.2388540622053095e-06, "loss": 0.0521, "step": 1324 }, { "epoch": 0.3014789533560865, "grad_norm": 0.955203381580325, "learning_rate": 1.2388372567025367e-06, "loss": 0.0338, "step": 1325 }, { "epoch": 0.30170648464163824, "grad_norm": 1.8026618764076134, "learning_rate": 1.2388204386540546e-06, "loss": 0.0667, "step": 1326 }, { "epoch": 0.30193401592719, "grad_norm": 1.4712519094439878, "learning_rate": 1.238803608060207e-06, "loss": 0.0573, "step": 1327 }, { "epoch": 0.30216154721274174, "grad_norm": 1.7736677880474818, "learning_rate": 1.2387867649213376e-06, "loss": 0.066, "step": 1328 }, { "epoch": 0.3023890784982935, "grad_norm": 2.3477972521623465, "learning_rate": 1.2387699092377908e-06, "loss": 0.1076, "step": 1329 }, { "epoch": 0.3026166097838453, "grad_norm": 1.1616522138628205, "learning_rate": 1.2387530410099113e-06, "loss": 0.0439, "step": 1330 }, { "epoch": 0.30284414106939705, "grad_norm": 1.4363882653748854, "learning_rate": 1.2387361602380436e-06, "loss": 0.0571, "step": 1331 }, { "epoch": 0.3030716723549488, "grad_norm": 0.7445259748127024, "learning_rate": 1.238719266922533e-06, "loss": 0.0254, "step": 1332 }, { "epoch": 0.30329920364050056, "grad_norm": 1.5848968542889414, "learning_rate": 1.238702361063724e-06, "loss": 0.0656, "step": 1333 }, { "epoch": 0.3035267349260523, "grad_norm": 6.520543814327786, "learning_rate": 1.2386854426619633e-06, "loss": 0.168, "step": 1334 }, { "epoch": 0.3037542662116041, "grad_norm": 1.1879669128675703, "learning_rate": 1.2386685117175956e-06, "loss": 0.0414, "step": 1335 }, { "epoch": 0.30398179749715587, "grad_norm": 1.627644346798232, "learning_rate": 1.2386515682309676e-06, "loss": 0.0524, "step": 1336 }, { "epoch": 0.3042093287827076, "grad_norm": 1.5535932914971018, "learning_rate": 1.2386346122024253e-06, "loss": 0.0783, "step": 1337 }, { "epoch": 0.3044368600682594, "grad_norm": 1.9629237703618092, "learning_rate": 1.2386176436323154e-06, "loss": 0.0682, "step": 1338 }, { "epoch": 0.3046643913538111, "grad_norm": 1.1645585598673585, "learning_rate": 1.2386006625209847e-06, "loss": 0.0459, "step": 1339 }, { "epoch": 0.30489192263936293, "grad_norm": 1.7630197143088098, "learning_rate": 1.2385836688687802e-06, "loss": 0.0934, "step": 1340 }, { "epoch": 0.3051194539249147, "grad_norm": 1.6191880044317468, "learning_rate": 1.2385666626760493e-06, "loss": 0.0633, "step": 1341 }, { "epoch": 0.30534698521046644, "grad_norm": 1.8913527150034817, "learning_rate": 1.2385496439431395e-06, "loss": 0.0914, "step": 1342 }, { "epoch": 0.3055745164960182, "grad_norm": 1.445418394402971, "learning_rate": 1.2385326126703986e-06, "loss": 0.0687, "step": 1343 }, { "epoch": 0.30580204778156994, "grad_norm": 1.8288680303956055, "learning_rate": 1.2385155688581746e-06, "loss": 0.0739, "step": 1344 }, { "epoch": 0.30602957906712175, "grad_norm": 1.0257171242110983, "learning_rate": 1.238498512506816e-06, "loss": 0.0384, "step": 1345 }, { "epoch": 0.3062571103526735, "grad_norm": 1.8747685549411612, "learning_rate": 1.2384814436166715e-06, "loss": 0.0818, "step": 1346 }, { "epoch": 0.30648464163822525, "grad_norm": 1.2409585891810266, "learning_rate": 1.2384643621880898e-06, "loss": 0.0559, "step": 1347 }, { "epoch": 0.306712172923777, "grad_norm": 1.5549364252372342, "learning_rate": 1.2384472682214201e-06, "loss": 0.0633, "step": 1348 }, { "epoch": 0.30693970420932876, "grad_norm": 2.521195686268785, "learning_rate": 1.2384301617170116e-06, "loss": 0.076, "step": 1349 }, { "epoch": 0.30716723549488056, "grad_norm": 1.5977696260072916, "learning_rate": 1.2384130426752142e-06, "loss": 0.0849, "step": 1350 }, { "epoch": 0.3073947667804323, "grad_norm": 1.8196313465168323, "learning_rate": 1.2383959110963775e-06, "loss": 0.0585, "step": 1351 }, { "epoch": 0.30762229806598407, "grad_norm": 1.349543502368748, "learning_rate": 1.2383787669808518e-06, "loss": 0.0627, "step": 1352 }, { "epoch": 0.3078498293515358, "grad_norm": 1.6887455424782774, "learning_rate": 1.2383616103289871e-06, "loss": 0.0497, "step": 1353 }, { "epoch": 0.3080773606370876, "grad_norm": 0.6215713527290292, "learning_rate": 1.238344441141135e-06, "loss": 0.0178, "step": 1354 }, { "epoch": 0.3083048919226394, "grad_norm": 1.640454410435492, "learning_rate": 1.2383272594176454e-06, "loss": 0.0638, "step": 1355 }, { "epoch": 0.30853242320819113, "grad_norm": 2.278012967969371, "learning_rate": 1.23831006515887e-06, "loss": 0.0815, "step": 1356 }, { "epoch": 0.3087599544937429, "grad_norm": 1.272401712072303, "learning_rate": 1.2382928583651601e-06, "loss": 0.0654, "step": 1357 }, { "epoch": 0.30898748577929463, "grad_norm": 2.586692416689192, "learning_rate": 1.2382756390368674e-06, "loss": 0.0821, "step": 1358 }, { "epoch": 0.30921501706484644, "grad_norm": 1.0904063045445367, "learning_rate": 1.2382584071743438e-06, "loss": 0.0461, "step": 1359 }, { "epoch": 0.3094425483503982, "grad_norm": 1.8779937542492704, "learning_rate": 1.2382411627779414e-06, "loss": 0.0883, "step": 1360 }, { "epoch": 0.30967007963594995, "grad_norm": 1.9814293020102243, "learning_rate": 1.2382239058480128e-06, "loss": 0.0778, "step": 1361 }, { "epoch": 0.3098976109215017, "grad_norm": 1.4818784709766222, "learning_rate": 1.2382066363849106e-06, "loss": 0.0476, "step": 1362 }, { "epoch": 0.31012514220705345, "grad_norm": 1.5363329257464977, "learning_rate": 1.2381893543889878e-06, "loss": 0.0539, "step": 1363 }, { "epoch": 0.31035267349260526, "grad_norm": 1.58062550907191, "learning_rate": 1.2381720598605976e-06, "loss": 0.0761, "step": 1364 }, { "epoch": 0.310580204778157, "grad_norm": 1.358541490432856, "learning_rate": 1.2381547528000934e-06, "loss": 0.0923, "step": 1365 }, { "epoch": 0.31080773606370876, "grad_norm": 1.0458300282796855, "learning_rate": 1.238137433207829e-06, "loss": 0.0377, "step": 1366 }, { "epoch": 0.3110352673492605, "grad_norm": 1.183742190968498, "learning_rate": 1.2381201010841585e-06, "loss": 0.0402, "step": 1367 }, { "epoch": 0.31126279863481227, "grad_norm": 1.2571454436583878, "learning_rate": 1.2381027564294359e-06, "loss": 0.062, "step": 1368 }, { "epoch": 0.3114903299203641, "grad_norm": 1.503878059552754, "learning_rate": 1.238085399244016e-06, "loss": 0.0751, "step": 1369 }, { "epoch": 0.3117178612059158, "grad_norm": 2.1756129961580317, "learning_rate": 1.2380680295282532e-06, "loss": 0.099, "step": 1370 }, { "epoch": 0.3119453924914676, "grad_norm": 1.478282276295607, "learning_rate": 1.2380506472825025e-06, "loss": 0.0451, "step": 1371 }, { "epoch": 0.31217292377701933, "grad_norm": 1.210809744948102, "learning_rate": 1.2380332525071194e-06, "loss": 0.0371, "step": 1372 }, { "epoch": 0.3124004550625711, "grad_norm": 1.4368943652552293, "learning_rate": 1.2380158452024595e-06, "loss": 0.0407, "step": 1373 }, { "epoch": 0.3126279863481229, "grad_norm": 0.7640327894504174, "learning_rate": 1.2379984253688783e-06, "loss": 0.0337, "step": 1374 }, { "epoch": 0.31285551763367464, "grad_norm": 1.8295475899815037, "learning_rate": 1.237980993006732e-06, "loss": 0.1147, "step": 1375 }, { "epoch": 0.3130830489192264, "grad_norm": 1.4807139063710786, "learning_rate": 1.2379635481163768e-06, "loss": 0.0743, "step": 1376 }, { "epoch": 0.31331058020477814, "grad_norm": 2.448286916518216, "learning_rate": 1.2379460906981692e-06, "loss": 0.0534, "step": 1377 }, { "epoch": 0.3135381114903299, "grad_norm": 0.9023973811346913, "learning_rate": 1.237928620752466e-06, "loss": 0.052, "step": 1378 }, { "epoch": 0.3137656427758817, "grad_norm": 2.013947590986497, "learning_rate": 1.2379111382796246e-06, "loss": 0.0575, "step": 1379 }, { "epoch": 0.31399317406143346, "grad_norm": 1.6032640725703056, "learning_rate": 1.2378936432800017e-06, "loss": 0.0706, "step": 1380 }, { "epoch": 0.3142207053469852, "grad_norm": 1.239992146480516, "learning_rate": 1.2378761357539554e-06, "loss": 0.0569, "step": 1381 }, { "epoch": 0.31444823663253696, "grad_norm": 1.0642429061701875, "learning_rate": 1.2378586157018434e-06, "loss": 0.0381, "step": 1382 }, { "epoch": 0.3146757679180887, "grad_norm": 1.1087768689334148, "learning_rate": 1.2378410831240235e-06, "loss": 0.0558, "step": 1383 }, { "epoch": 0.3149032992036405, "grad_norm": 1.9572648591918616, "learning_rate": 1.2378235380208542e-06, "loss": 0.0738, "step": 1384 }, { "epoch": 0.31513083048919227, "grad_norm": 0.9929386411259282, "learning_rate": 1.2378059803926941e-06, "loss": 0.0394, "step": 1385 }, { "epoch": 0.315358361774744, "grad_norm": 1.202982993512815, "learning_rate": 1.2377884102399023e-06, "loss": 0.0461, "step": 1386 }, { "epoch": 0.3155858930602958, "grad_norm": 1.1752667673190271, "learning_rate": 1.2377708275628375e-06, "loss": 0.0447, "step": 1387 }, { "epoch": 0.3158134243458475, "grad_norm": 1.5891864361864994, "learning_rate": 1.2377532323618593e-06, "loss": 0.0456, "step": 1388 }, { "epoch": 0.31604095563139933, "grad_norm": 1.6341111744531334, "learning_rate": 1.2377356246373271e-06, "loss": 0.069, "step": 1389 }, { "epoch": 0.3162684869169511, "grad_norm": 1.440731938439493, "learning_rate": 1.2377180043896012e-06, "loss": 0.0595, "step": 1390 }, { "epoch": 0.31649601820250284, "grad_norm": 0.9756352806439903, "learning_rate": 1.2377003716190411e-06, "loss": 0.0469, "step": 1391 }, { "epoch": 0.3167235494880546, "grad_norm": 1.4676165718616923, "learning_rate": 1.2376827263260078e-06, "loss": 0.0516, "step": 1392 }, { "epoch": 0.3169510807736064, "grad_norm": 1.5939577885228815, "learning_rate": 1.2376650685108612e-06, "loss": 0.0503, "step": 1393 }, { "epoch": 0.31717861205915815, "grad_norm": 2.1816410059989098, "learning_rate": 1.2376473981739632e-06, "loss": 0.0929, "step": 1394 }, { "epoch": 0.3174061433447099, "grad_norm": 2.718683093963833, "learning_rate": 1.237629715315674e-06, "loss": 0.1129, "step": 1395 }, { "epoch": 0.31763367463026165, "grad_norm": 0.9979838242263193, "learning_rate": 1.2376120199363554e-06, "loss": 0.0536, "step": 1396 }, { "epoch": 0.3178612059158134, "grad_norm": 1.9720298867063506, "learning_rate": 1.2375943120363692e-06, "loss": 0.0969, "step": 1397 }, { "epoch": 0.3180887372013652, "grad_norm": 0.6502262811463818, "learning_rate": 1.2375765916160773e-06, "loss": 0.0287, "step": 1398 }, { "epoch": 0.31831626848691696, "grad_norm": 1.3837265125988647, "learning_rate": 1.2375588586758415e-06, "loss": 0.0441, "step": 1399 }, { "epoch": 0.3185437997724687, "grad_norm": 1.593794924686227, "learning_rate": 1.2375411132160245e-06, "loss": 0.0595, "step": 1400 }, { "epoch": 0.31877133105802047, "grad_norm": 1.776611213440541, "learning_rate": 1.2375233552369892e-06, "loss": 0.0694, "step": 1401 }, { "epoch": 0.3189988623435722, "grad_norm": 1.2580919074923334, "learning_rate": 1.237505584739098e-06, "loss": 0.0491, "step": 1402 }, { "epoch": 0.31922639362912403, "grad_norm": 0.994481303076797, "learning_rate": 1.2374878017227147e-06, "loss": 0.0291, "step": 1403 }, { "epoch": 0.3194539249146758, "grad_norm": 0.7265755658718445, "learning_rate": 1.237470006188202e-06, "loss": 0.027, "step": 1404 }, { "epoch": 0.31968145620022753, "grad_norm": 0.9576296817818531, "learning_rate": 1.2374521981359245e-06, "loss": 0.0423, "step": 1405 }, { "epoch": 0.3199089874857793, "grad_norm": 1.7776464530772098, "learning_rate": 1.2374343775662456e-06, "loss": 0.0764, "step": 1406 }, { "epoch": 0.32013651877133104, "grad_norm": 1.2848189857025718, "learning_rate": 1.2374165444795296e-06, "loss": 0.0398, "step": 1407 }, { "epoch": 0.32036405005688284, "grad_norm": 2.0114145603579625, "learning_rate": 1.237398698876141e-06, "loss": 0.0934, "step": 1408 }, { "epoch": 0.3205915813424346, "grad_norm": 1.3984905897111508, "learning_rate": 1.2373808407564446e-06, "loss": 0.0513, "step": 1409 }, { "epoch": 0.32081911262798635, "grad_norm": 0.9464490409610342, "learning_rate": 1.2373629701208053e-06, "loss": 0.0369, "step": 1410 }, { "epoch": 0.3210466439135381, "grad_norm": 1.4072284744296988, "learning_rate": 1.2373450869695883e-06, "loss": 0.0479, "step": 1411 }, { "epoch": 0.32127417519908985, "grad_norm": 1.5797071346301892, "learning_rate": 1.2373271913031593e-06, "loss": 0.0581, "step": 1412 }, { "epoch": 0.32150170648464166, "grad_norm": 1.5741999856445812, "learning_rate": 1.237309283121884e-06, "loss": 0.0695, "step": 1413 }, { "epoch": 0.3217292377701934, "grad_norm": 1.004256336444922, "learning_rate": 1.237291362426128e-06, "loss": 0.0584, "step": 1414 }, { "epoch": 0.32195676905574516, "grad_norm": 1.2610248823720653, "learning_rate": 1.2372734292162584e-06, "loss": 0.0508, "step": 1415 }, { "epoch": 0.3221843003412969, "grad_norm": 1.138966176373832, "learning_rate": 1.237255483492641e-06, "loss": 0.0346, "step": 1416 }, { "epoch": 0.32241183162684867, "grad_norm": 1.279726266655994, "learning_rate": 1.2372375252556429e-06, "loss": 0.0526, "step": 1417 }, { "epoch": 0.3226393629124005, "grad_norm": 2.259244357087649, "learning_rate": 1.2372195545056308e-06, "loss": 0.0737, "step": 1418 }, { "epoch": 0.3228668941979522, "grad_norm": 1.260233063486729, "learning_rate": 1.2372015712429725e-06, "loss": 0.0439, "step": 1419 }, { "epoch": 0.323094425483504, "grad_norm": 1.180315594146752, "learning_rate": 1.237183575468035e-06, "loss": 0.0506, "step": 1420 }, { "epoch": 0.32332195676905573, "grad_norm": 0.9777167019905004, "learning_rate": 1.2371655671811866e-06, "loss": 0.0308, "step": 1421 }, { "epoch": 0.3235494880546075, "grad_norm": 0.8734921365662196, "learning_rate": 1.237147546382795e-06, "loss": 0.0281, "step": 1422 }, { "epoch": 0.3237770193401593, "grad_norm": 1.1112810388031804, "learning_rate": 1.237129513073229e-06, "loss": 0.04, "step": 1423 }, { "epoch": 0.32400455062571104, "grad_norm": 1.156890481720808, "learning_rate": 1.2371114672528565e-06, "loss": 0.0593, "step": 1424 }, { "epoch": 0.3242320819112628, "grad_norm": 0.8688160126362955, "learning_rate": 1.2370934089220466e-06, "loss": 0.0239, "step": 1425 }, { "epoch": 0.32445961319681454, "grad_norm": 1.1430764409107144, "learning_rate": 1.2370753380811685e-06, "loss": 0.0343, "step": 1426 }, { "epoch": 0.32468714448236635, "grad_norm": 0.94277640364479, "learning_rate": 1.2370572547305915e-06, "loss": 0.0254, "step": 1427 }, { "epoch": 0.3249146757679181, "grad_norm": 0.8543512868236353, "learning_rate": 1.237039158870685e-06, "loss": 0.0359, "step": 1428 }, { "epoch": 0.32514220705346986, "grad_norm": 0.9268593896369735, "learning_rate": 1.237021050501819e-06, "loss": 0.0462, "step": 1429 }, { "epoch": 0.3253697383390216, "grad_norm": 1.0492131527271786, "learning_rate": 1.2370029296243638e-06, "loss": 0.0483, "step": 1430 }, { "epoch": 0.32559726962457336, "grad_norm": 1.2265145141334126, "learning_rate": 1.2369847962386893e-06, "loss": 0.0605, "step": 1431 }, { "epoch": 0.32582480091012517, "grad_norm": 1.4614421557097579, "learning_rate": 1.2369666503451665e-06, "loss": 0.0792, "step": 1432 }, { "epoch": 0.3260523321956769, "grad_norm": 1.3551154178225278, "learning_rate": 1.2369484919441662e-06, "loss": 0.0504, "step": 1433 }, { "epoch": 0.32627986348122867, "grad_norm": 2.070524215459429, "learning_rate": 1.2369303210360592e-06, "loss": 0.0813, "step": 1434 }, { "epoch": 0.3265073947667804, "grad_norm": 2.674887998532007, "learning_rate": 1.2369121376212174e-06, "loss": 0.1356, "step": 1435 }, { "epoch": 0.3267349260523322, "grad_norm": 2.0059123123271543, "learning_rate": 1.236893941700012e-06, "loss": 0.0815, "step": 1436 }, { "epoch": 0.326962457337884, "grad_norm": 1.614422308046532, "learning_rate": 1.236875733272815e-06, "loss": 0.0632, "step": 1437 }, { "epoch": 0.32718998862343573, "grad_norm": 1.0273415965735717, "learning_rate": 1.236857512339999e-06, "loss": 0.0636, "step": 1438 }, { "epoch": 0.3274175199089875, "grad_norm": 2.063298633572998, "learning_rate": 1.2368392789019356e-06, "loss": 0.1028, "step": 1439 }, { "epoch": 0.32764505119453924, "grad_norm": 1.0762629811135433, "learning_rate": 1.2368210329589982e-06, "loss": 0.0287, "step": 1440 }, { "epoch": 0.327872582480091, "grad_norm": 1.5998871926979066, "learning_rate": 1.236802774511559e-06, "loss": 0.0862, "step": 1441 }, { "epoch": 0.3281001137656428, "grad_norm": 1.4906993606341739, "learning_rate": 1.2367845035599919e-06, "loss": 0.094, "step": 1442 }, { "epoch": 0.32832764505119455, "grad_norm": 0.8173626977187779, "learning_rate": 1.2367662201046698e-06, "loss": 0.0348, "step": 1443 }, { "epoch": 0.3285551763367463, "grad_norm": 1.260244704516005, "learning_rate": 1.2367479241459666e-06, "loss": 0.0388, "step": 1444 }, { "epoch": 0.32878270762229805, "grad_norm": 1.2263377322773583, "learning_rate": 1.2367296156842562e-06, "loss": 0.0514, "step": 1445 }, { "epoch": 0.3290102389078498, "grad_norm": 1.1202578543878008, "learning_rate": 1.2367112947199128e-06, "loss": 0.0332, "step": 1446 }, { "epoch": 0.3292377701934016, "grad_norm": 1.3736971576588957, "learning_rate": 1.2366929612533109e-06, "loss": 0.0539, "step": 1447 }, { "epoch": 0.32946530147895337, "grad_norm": 1.2893217491287035, "learning_rate": 1.2366746152848249e-06, "loss": 0.058, "step": 1448 }, { "epoch": 0.3296928327645051, "grad_norm": 1.4381578160793065, "learning_rate": 1.2366562568148301e-06, "loss": 0.0521, "step": 1449 }, { "epoch": 0.32992036405005687, "grad_norm": 1.6844461549267769, "learning_rate": 1.2366378858437016e-06, "loss": 0.1036, "step": 1450 }, { "epoch": 0.3301478953356086, "grad_norm": 1.397177590337732, "learning_rate": 1.2366195023718152e-06, "loss": 0.0534, "step": 1451 }, { "epoch": 0.33037542662116043, "grad_norm": 1.0456464972065773, "learning_rate": 1.2366011063995458e-06, "loss": 0.0558, "step": 1452 }, { "epoch": 0.3306029579067122, "grad_norm": 1.9835006920707845, "learning_rate": 1.2365826979272702e-06, "loss": 0.1249, "step": 1453 }, { "epoch": 0.33083048919226393, "grad_norm": 2.6314184120557376, "learning_rate": 1.2365642769553644e-06, "loss": 0.1484, "step": 1454 }, { "epoch": 0.3310580204778157, "grad_norm": 1.4200073439085166, "learning_rate": 1.2365458434842046e-06, "loss": 0.0483, "step": 1455 }, { "epoch": 0.33128555176336744, "grad_norm": 1.1143001688120713, "learning_rate": 1.2365273975141675e-06, "loss": 0.0369, "step": 1456 }, { "epoch": 0.33151308304891924, "grad_norm": 1.0377938140431064, "learning_rate": 1.2365089390456308e-06, "loss": 0.0518, "step": 1457 }, { "epoch": 0.331740614334471, "grad_norm": 1.1599848352897109, "learning_rate": 1.236490468078971e-06, "loss": 0.0432, "step": 1458 }, { "epoch": 0.33196814562002275, "grad_norm": 1.5939156162998587, "learning_rate": 1.2364719846145662e-06, "loss": 0.0479, "step": 1459 }, { "epoch": 0.3321956769055745, "grad_norm": 1.0154765270633663, "learning_rate": 1.2364534886527937e-06, "loss": 0.0431, "step": 1460 }, { "epoch": 0.3324232081911263, "grad_norm": 1.8584242211532112, "learning_rate": 1.2364349801940317e-06, "loss": 0.1002, "step": 1461 }, { "epoch": 0.33265073947667806, "grad_norm": 1.3694201211070989, "learning_rate": 1.2364164592386588e-06, "loss": 0.0733, "step": 1462 }, { "epoch": 0.3328782707622298, "grad_norm": 1.9119781832863982, "learning_rate": 1.2363979257870528e-06, "loss": 0.0696, "step": 1463 }, { "epoch": 0.33310580204778156, "grad_norm": 1.9001104021651412, "learning_rate": 1.2363793798395932e-06, "loss": 0.0572, "step": 1464 }, { "epoch": 0.3333333333333333, "grad_norm": 2.498897834454089, "learning_rate": 1.2363608213966588e-06, "loss": 0.0848, "step": 1465 }, { "epoch": 0.3335608646188851, "grad_norm": 1.1620781742745308, "learning_rate": 1.2363422504586286e-06, "loss": 0.0504, "step": 1466 }, { "epoch": 0.3337883959044369, "grad_norm": 1.2505263772670705, "learning_rate": 1.2363236670258827e-06, "loss": 0.05, "step": 1467 }, { "epoch": 0.3340159271899886, "grad_norm": 1.4633039559699816, "learning_rate": 1.2363050710988003e-06, "loss": 0.079, "step": 1468 }, { "epoch": 0.3342434584755404, "grad_norm": 1.2463657051053827, "learning_rate": 1.236286462677762e-06, "loss": 0.0418, "step": 1469 }, { "epoch": 0.33447098976109213, "grad_norm": 1.683583709704665, "learning_rate": 1.236267841763148e-06, "loss": 0.0812, "step": 1470 }, { "epoch": 0.33469852104664394, "grad_norm": 0.7484223072827635, "learning_rate": 1.2362492083553387e-06, "loss": 0.0356, "step": 1471 }, { "epoch": 0.3349260523321957, "grad_norm": 1.639323852335928, "learning_rate": 1.236230562454715e-06, "loss": 0.0401, "step": 1472 }, { "epoch": 0.33515358361774744, "grad_norm": 1.2064100614992055, "learning_rate": 1.2362119040616582e-06, "loss": 0.0677, "step": 1473 }, { "epoch": 0.3353811149032992, "grad_norm": 1.107381501336236, "learning_rate": 1.2361932331765492e-06, "loss": 0.0404, "step": 1474 }, { "epoch": 0.33560864618885095, "grad_norm": 1.4911394528443533, "learning_rate": 1.2361745497997702e-06, "loss": 0.0623, "step": 1475 }, { "epoch": 0.33583617747440275, "grad_norm": 1.2262949506307204, "learning_rate": 1.2361558539317023e-06, "loss": 0.0651, "step": 1476 }, { "epoch": 0.3360637087599545, "grad_norm": 0.9519435590816607, "learning_rate": 1.2361371455727284e-06, "loss": 0.0404, "step": 1477 }, { "epoch": 0.33629124004550626, "grad_norm": 1.5689690961881004, "learning_rate": 1.2361184247232302e-06, "loss": 0.087, "step": 1478 }, { "epoch": 0.336518771331058, "grad_norm": 5.906098295875717, "learning_rate": 1.2360996913835907e-06, "loss": 0.236, "step": 1479 }, { "epoch": 0.33674630261660976, "grad_norm": 0.8654279575626805, "learning_rate": 1.2360809455541928e-06, "loss": 0.0549, "step": 1480 }, { "epoch": 0.33697383390216157, "grad_norm": 1.776832317476583, "learning_rate": 1.2360621872354195e-06, "loss": 0.0453, "step": 1481 }, { "epoch": 0.3372013651877133, "grad_norm": 1.347252653711477, "learning_rate": 1.236043416427654e-06, "loss": 0.0697, "step": 1482 }, { "epoch": 0.33742889647326507, "grad_norm": 1.2864149376371685, "learning_rate": 1.2360246331312804e-06, "loss": 0.0381, "step": 1483 }, { "epoch": 0.3376564277588168, "grad_norm": 1.6142130648155326, "learning_rate": 1.2360058373466821e-06, "loss": 0.0477, "step": 1484 }, { "epoch": 0.3378839590443686, "grad_norm": 2.106670043543402, "learning_rate": 1.2359870290742437e-06, "loss": 0.097, "step": 1485 }, { "epoch": 0.3381114903299204, "grad_norm": 1.2681226361832132, "learning_rate": 1.2359682083143494e-06, "loss": 0.0682, "step": 1486 }, { "epoch": 0.33833902161547214, "grad_norm": 0.8649596148543436, "learning_rate": 1.2359493750673835e-06, "loss": 0.0302, "step": 1487 }, { "epoch": 0.3385665529010239, "grad_norm": 1.3533089587285108, "learning_rate": 1.2359305293337316e-06, "loss": 0.0526, "step": 1488 }, { "epoch": 0.33879408418657564, "grad_norm": 0.7446203717116848, "learning_rate": 1.2359116711137785e-06, "loss": 0.0247, "step": 1489 }, { "epoch": 0.3390216154721274, "grad_norm": 1.1703255026021544, "learning_rate": 1.2358928004079095e-06, "loss": 0.0469, "step": 1490 }, { "epoch": 0.3392491467576792, "grad_norm": 2.2956044408157554, "learning_rate": 1.2358739172165108e-06, "loss": 0.1262, "step": 1491 }, { "epoch": 0.33947667804323095, "grad_norm": 1.425210539470094, "learning_rate": 1.235855021539968e-06, "loss": 0.0539, "step": 1492 }, { "epoch": 0.3397042093287827, "grad_norm": 1.2958224989383627, "learning_rate": 1.2358361133786668e-06, "loss": 0.0417, "step": 1493 }, { "epoch": 0.33993174061433445, "grad_norm": 0.8383879482270241, "learning_rate": 1.2358171927329946e-06, "loss": 0.0411, "step": 1494 }, { "epoch": 0.34015927189988626, "grad_norm": 1.0816879292496617, "learning_rate": 1.2357982596033374e-06, "loss": 0.0321, "step": 1495 }, { "epoch": 0.340386803185438, "grad_norm": 1.2228392090454494, "learning_rate": 1.2357793139900823e-06, "loss": 0.076, "step": 1496 }, { "epoch": 0.34061433447098977, "grad_norm": 0.9376305546071186, "learning_rate": 1.2357603558936168e-06, "loss": 0.0294, "step": 1497 }, { "epoch": 0.3408418657565415, "grad_norm": 2.032138065689827, "learning_rate": 1.2357413853143282e-06, "loss": 0.1073, "step": 1498 }, { "epoch": 0.34106939704209327, "grad_norm": 1.203066065214765, "learning_rate": 1.2357224022526041e-06, "loss": 0.0585, "step": 1499 }, { "epoch": 0.3412969283276451, "grad_norm": 1.5240232506976108, "learning_rate": 1.2357034067088327e-06, "loss": 0.0566, "step": 1500 }, { "epoch": 0.34152445961319683, "grad_norm": 1.4069261769751387, "learning_rate": 1.235684398683402e-06, "loss": 0.0911, "step": 1501 }, { "epoch": 0.3417519908987486, "grad_norm": 1.2794137407821935, "learning_rate": 1.2356653781767009e-06, "loss": 0.0387, "step": 1502 }, { "epoch": 0.34197952218430033, "grad_norm": 1.7747949717550477, "learning_rate": 1.2356463451891174e-06, "loss": 0.0928, "step": 1503 }, { "epoch": 0.3422070534698521, "grad_norm": 1.0081856733209773, "learning_rate": 1.2356272997210414e-06, "loss": 0.0289, "step": 1504 }, { "epoch": 0.3424345847554039, "grad_norm": 1.8655840562261654, "learning_rate": 1.2356082417728612e-06, "loss": 0.0837, "step": 1505 }, { "epoch": 0.34266211604095564, "grad_norm": 0.9436733972164556, "learning_rate": 1.2355891713449672e-06, "loss": 0.0406, "step": 1506 }, { "epoch": 0.3428896473265074, "grad_norm": 1.2922255822019764, "learning_rate": 1.2355700884377485e-06, "loss": 0.0363, "step": 1507 }, { "epoch": 0.34311717861205915, "grad_norm": 1.3543650706739145, "learning_rate": 1.2355509930515958e-06, "loss": 0.0697, "step": 1508 }, { "epoch": 0.3433447098976109, "grad_norm": 1.3352155504337506, "learning_rate": 1.2355318851868987e-06, "loss": 0.0723, "step": 1509 }, { "epoch": 0.3435722411831627, "grad_norm": 0.8330395399945464, "learning_rate": 1.235512764844048e-06, "loss": 0.0427, "step": 1510 }, { "epoch": 0.34379977246871446, "grad_norm": 0.7599055129001592, "learning_rate": 1.2354936320234345e-06, "loss": 0.0275, "step": 1511 }, { "epoch": 0.3440273037542662, "grad_norm": 1.107814773629098, "learning_rate": 1.2354744867254493e-06, "loss": 0.0519, "step": 1512 }, { "epoch": 0.34425483503981796, "grad_norm": 0.9874953818051542, "learning_rate": 1.2354553289504836e-06, "loss": 0.0276, "step": 1513 }, { "epoch": 0.3444823663253697, "grad_norm": 1.4195908498124084, "learning_rate": 1.2354361586989287e-06, "loss": 0.0658, "step": 1514 }, { "epoch": 0.3447098976109215, "grad_norm": 1.1536105740853742, "learning_rate": 1.235416975971177e-06, "loss": 0.0693, "step": 1515 }, { "epoch": 0.3449374288964733, "grad_norm": 1.1208196458039765, "learning_rate": 1.2353977807676205e-06, "loss": 0.0451, "step": 1516 }, { "epoch": 0.345164960182025, "grad_norm": 1.6664132616304859, "learning_rate": 1.2353785730886506e-06, "loss": 0.0659, "step": 1517 }, { "epoch": 0.3453924914675768, "grad_norm": 1.6331880132019159, "learning_rate": 1.235359352934661e-06, "loss": 0.0518, "step": 1518 }, { "epoch": 0.34562002275312853, "grad_norm": 1.9450367879491788, "learning_rate": 1.235340120306044e-06, "loss": 0.108, "step": 1519 }, { "epoch": 0.34584755403868034, "grad_norm": 1.4426224264383893, "learning_rate": 1.2353208752031925e-06, "loss": 0.0598, "step": 1520 }, { "epoch": 0.3460750853242321, "grad_norm": 1.2111901419847948, "learning_rate": 1.2353016176265002e-06, "loss": 0.0304, "step": 1521 }, { "epoch": 0.34630261660978384, "grad_norm": 0.841669446586875, "learning_rate": 1.2352823475763603e-06, "loss": 0.029, "step": 1522 }, { "epoch": 0.3465301478953356, "grad_norm": 1.6324873257088213, "learning_rate": 1.2352630650531672e-06, "loss": 0.0737, "step": 1523 }, { "epoch": 0.34675767918088735, "grad_norm": 1.5631645650045773, "learning_rate": 1.2352437700573147e-06, "loss": 0.0555, "step": 1524 }, { "epoch": 0.34698521046643915, "grad_norm": 1.3689087939433322, "learning_rate": 1.235224462589197e-06, "loss": 0.0699, "step": 1525 }, { "epoch": 0.3472127417519909, "grad_norm": 1.729486346174357, "learning_rate": 1.2352051426492089e-06, "loss": 0.0795, "step": 1526 }, { "epoch": 0.34744027303754266, "grad_norm": 1.2401649575621747, "learning_rate": 1.2351858102377455e-06, "loss": 0.0459, "step": 1527 }, { "epoch": 0.3476678043230944, "grad_norm": 1.1736397484657297, "learning_rate": 1.2351664653552012e-06, "loss": 0.0374, "step": 1528 }, { "epoch": 0.34789533560864616, "grad_norm": 0.8775289996354035, "learning_rate": 1.235147108001972e-06, "loss": 0.0415, "step": 1529 }, { "epoch": 0.34812286689419797, "grad_norm": 1.5150051901632806, "learning_rate": 1.2351277381784532e-06, "loss": 0.0499, "step": 1530 }, { "epoch": 0.3483503981797497, "grad_norm": 0.7367153174028898, "learning_rate": 1.235108355885041e-06, "loss": 0.0405, "step": 1531 }, { "epoch": 0.3485779294653015, "grad_norm": 1.0973156282921959, "learning_rate": 1.2350889611221315e-06, "loss": 0.0593, "step": 1532 }, { "epoch": 0.3488054607508532, "grad_norm": 1.2029539420964899, "learning_rate": 1.2350695538901207e-06, "loss": 0.052, "step": 1533 }, { "epoch": 0.34903299203640503, "grad_norm": 1.2254789545948275, "learning_rate": 1.2350501341894055e-06, "loss": 0.0442, "step": 1534 }, { "epoch": 0.3492605233219568, "grad_norm": 1.173455391781939, "learning_rate": 1.235030702020383e-06, "loss": 0.0627, "step": 1535 }, { "epoch": 0.34948805460750854, "grad_norm": 1.1576335152673878, "learning_rate": 1.23501125738345e-06, "loss": 0.0467, "step": 1536 }, { "epoch": 0.3497155858930603, "grad_norm": 1.0478326366866813, "learning_rate": 1.2349918002790043e-06, "loss": 0.0409, "step": 1537 }, { "epoch": 0.34994311717861204, "grad_norm": 0.7625837183082101, "learning_rate": 1.2349723307074432e-06, "loss": 0.037, "step": 1538 }, { "epoch": 0.35017064846416385, "grad_norm": 1.2469855224758288, "learning_rate": 1.2349528486691648e-06, "loss": 0.0668, "step": 1539 }, { "epoch": 0.3503981797497156, "grad_norm": 1.0826914171234396, "learning_rate": 1.2349333541645672e-06, "loss": 0.0443, "step": 1540 }, { "epoch": 0.35062571103526735, "grad_norm": 1.195949344430507, "learning_rate": 1.2349138471940489e-06, "loss": 0.0362, "step": 1541 }, { "epoch": 0.3508532423208191, "grad_norm": 1.2921680529996942, "learning_rate": 1.2348943277580086e-06, "loss": 0.0534, "step": 1542 }, { "epoch": 0.35108077360637086, "grad_norm": 0.713008372789458, "learning_rate": 1.2348747958568452e-06, "loss": 0.0211, "step": 1543 }, { "epoch": 0.35130830489192266, "grad_norm": 1.5116593555386681, "learning_rate": 1.2348552514909579e-06, "loss": 0.0712, "step": 1544 }, { "epoch": 0.3515358361774744, "grad_norm": 1.0248863840699733, "learning_rate": 1.2348356946607462e-06, "loss": 0.0305, "step": 1545 }, { "epoch": 0.35176336746302617, "grad_norm": 1.1409421918168505, "learning_rate": 1.2348161253666096e-06, "loss": 0.0529, "step": 1546 }, { "epoch": 0.3519908987485779, "grad_norm": 1.768834998548112, "learning_rate": 1.2347965436089484e-06, "loss": 0.0888, "step": 1547 }, { "epoch": 0.35221843003412967, "grad_norm": 1.4622739241779608, "learning_rate": 1.2347769493881625e-06, "loss": 0.0466, "step": 1548 }, { "epoch": 0.3524459613196815, "grad_norm": 0.6584121088377994, "learning_rate": 1.2347573427046527e-06, "loss": 0.0348, "step": 1549 }, { "epoch": 0.35267349260523323, "grad_norm": 2.2141006455976604, "learning_rate": 1.2347377235588193e-06, "loss": 0.1474, "step": 1550 }, { "epoch": 0.352901023890785, "grad_norm": 1.19254071976006, "learning_rate": 1.2347180919510637e-06, "loss": 0.0511, "step": 1551 }, { "epoch": 0.35312855517633673, "grad_norm": 1.4469931390196393, "learning_rate": 1.234698447881787e-06, "loss": 0.0457, "step": 1552 }, { "epoch": 0.3533560864618885, "grad_norm": 1.0953678059195961, "learning_rate": 1.2346787913513904e-06, "loss": 0.0351, "step": 1553 }, { "epoch": 0.3535836177474403, "grad_norm": 1.2220799955288675, "learning_rate": 1.234659122360276e-06, "loss": 0.0605, "step": 1554 }, { "epoch": 0.35381114903299204, "grad_norm": 0.7125746349183502, "learning_rate": 1.2346394409088457e-06, "loss": 0.029, "step": 1555 }, { "epoch": 0.3540386803185438, "grad_norm": 0.7700035617297257, "learning_rate": 1.2346197469975016e-06, "loss": 0.0399, "step": 1556 }, { "epoch": 0.35426621160409555, "grad_norm": 1.4287678945741027, "learning_rate": 1.2346000406266466e-06, "loss": 0.0849, "step": 1557 }, { "epoch": 0.3544937428896473, "grad_norm": 1.4371209706481878, "learning_rate": 1.2345803217966829e-06, "loss": 0.0497, "step": 1558 }, { "epoch": 0.3547212741751991, "grad_norm": 1.5348631747891068, "learning_rate": 1.2345605905080141e-06, "loss": 0.0625, "step": 1559 }, { "epoch": 0.35494880546075086, "grad_norm": 1.8970470757470723, "learning_rate": 1.234540846761043e-06, "loss": 0.0613, "step": 1560 }, { "epoch": 0.3551763367463026, "grad_norm": 2.2382508192831416, "learning_rate": 1.2345210905561733e-06, "loss": 0.0889, "step": 1561 }, { "epoch": 0.35540386803185436, "grad_norm": 0.9465351044370347, "learning_rate": 1.2345013218938089e-06, "loss": 0.0478, "step": 1562 }, { "epoch": 0.3556313993174061, "grad_norm": 0.7802531004508686, "learning_rate": 1.2344815407743537e-06, "loss": 0.0265, "step": 1563 }, { "epoch": 0.3558589306029579, "grad_norm": 1.3777470813147898, "learning_rate": 1.2344617471982119e-06, "loss": 0.0595, "step": 1564 }, { "epoch": 0.3560864618885097, "grad_norm": 1.9091835939621151, "learning_rate": 1.2344419411657883e-06, "loss": 0.0652, "step": 1565 }, { "epoch": 0.3563139931740614, "grad_norm": 1.8891000138650136, "learning_rate": 1.2344221226774874e-06, "loss": 0.0894, "step": 1566 }, { "epoch": 0.3565415244596132, "grad_norm": 2.3333751351582968, "learning_rate": 1.2344022917337147e-06, "loss": 0.1068, "step": 1567 }, { "epoch": 0.356769055745165, "grad_norm": 1.5543425365273886, "learning_rate": 1.234382448334875e-06, "loss": 0.0549, "step": 1568 }, { "epoch": 0.35699658703071674, "grad_norm": 0.7049933258516752, "learning_rate": 1.2343625924813741e-06, "loss": 0.024, "step": 1569 }, { "epoch": 0.3572241183162685, "grad_norm": 1.7675084632445353, "learning_rate": 1.234342724173618e-06, "loss": 0.0485, "step": 1570 }, { "epoch": 0.35745164960182024, "grad_norm": 1.3808807084264496, "learning_rate": 1.2343228434120124e-06, "loss": 0.067, "step": 1571 }, { "epoch": 0.357679180887372, "grad_norm": 1.2453962587688918, "learning_rate": 1.2343029501969638e-06, "loss": 0.0524, "step": 1572 }, { "epoch": 0.3579067121729238, "grad_norm": 1.343656065138551, "learning_rate": 1.2342830445288788e-06, "loss": 0.0699, "step": 1573 }, { "epoch": 0.35813424345847555, "grad_norm": 1.0013673586639926, "learning_rate": 1.2342631264081643e-06, "loss": 0.051, "step": 1574 }, { "epoch": 0.3583617747440273, "grad_norm": 1.1273372730897988, "learning_rate": 1.234243195835227e-06, "loss": 0.0564, "step": 1575 }, { "epoch": 0.35858930602957906, "grad_norm": 1.3457068038564708, "learning_rate": 1.234223252810475e-06, "loss": 0.0445, "step": 1576 }, { "epoch": 0.3588168373151308, "grad_norm": 1.1036016217002649, "learning_rate": 1.2342032973343152e-06, "loss": 0.0494, "step": 1577 }, { "epoch": 0.3590443686006826, "grad_norm": 1.1088066035106612, "learning_rate": 1.2341833294071558e-06, "loss": 0.0483, "step": 1578 }, { "epoch": 0.35927189988623437, "grad_norm": 1.81529439176539, "learning_rate": 1.2341633490294046e-06, "loss": 0.1067, "step": 1579 }, { "epoch": 0.3594994311717861, "grad_norm": 1.7525830328724732, "learning_rate": 1.2341433562014705e-06, "loss": 0.0977, "step": 1580 }, { "epoch": 0.3597269624573379, "grad_norm": 1.1410352301000704, "learning_rate": 1.2341233509237616e-06, "loss": 0.0383, "step": 1581 }, { "epoch": 0.3599544937428896, "grad_norm": 1.3291484690769528, "learning_rate": 1.234103333196687e-06, "loss": 0.0414, "step": 1582 }, { "epoch": 0.36018202502844143, "grad_norm": 1.8297830185646766, "learning_rate": 1.2340833030206558e-06, "loss": 0.0578, "step": 1583 }, { "epoch": 0.3604095563139932, "grad_norm": 1.4265200616837415, "learning_rate": 1.2340632603960774e-06, "loss": 0.0533, "step": 1584 }, { "epoch": 0.36063708759954494, "grad_norm": 1.5565200581725265, "learning_rate": 1.2340432053233615e-06, "loss": 0.0558, "step": 1585 }, { "epoch": 0.3608646188850967, "grad_norm": 0.8252310606061587, "learning_rate": 1.2340231378029177e-06, "loss": 0.0415, "step": 1586 }, { "epoch": 0.36109215017064844, "grad_norm": 1.1625630250416708, "learning_rate": 1.2340030578351564e-06, "loss": 0.0536, "step": 1587 }, { "epoch": 0.36131968145620025, "grad_norm": 1.4833717042031473, "learning_rate": 1.2339829654204878e-06, "loss": 0.054, "step": 1588 }, { "epoch": 0.361547212741752, "grad_norm": 1.2867709045464473, "learning_rate": 1.2339628605593229e-06, "loss": 0.0486, "step": 1589 }, { "epoch": 0.36177474402730375, "grad_norm": 1.7822365018575843, "learning_rate": 1.2339427432520722e-06, "loss": 0.0538, "step": 1590 }, { "epoch": 0.3620022753128555, "grad_norm": 1.7621725916894804, "learning_rate": 1.2339226134991471e-06, "loss": 0.0742, "step": 1591 }, { "epoch": 0.36222980659840726, "grad_norm": 2.2648964065061223, "learning_rate": 1.2339024713009592e-06, "loss": 0.0702, "step": 1592 }, { "epoch": 0.36245733788395906, "grad_norm": 1.3626520454398885, "learning_rate": 1.2338823166579197e-06, "loss": 0.0896, "step": 1593 }, { "epoch": 0.3626848691695108, "grad_norm": 1.3432892967430223, "learning_rate": 1.2338621495704409e-06, "loss": 0.0442, "step": 1594 }, { "epoch": 0.36291240045506257, "grad_norm": 1.920266034051075, "learning_rate": 1.2338419700389349e-06, "loss": 0.0949, "step": 1595 }, { "epoch": 0.3631399317406143, "grad_norm": 1.556412496157174, "learning_rate": 1.2338217780638137e-06, "loss": 0.0496, "step": 1596 }, { "epoch": 0.36336746302616607, "grad_norm": 2.160817778943729, "learning_rate": 1.2338015736454908e-06, "loss": 0.1306, "step": 1597 }, { "epoch": 0.3635949943117179, "grad_norm": 1.3086877468472204, "learning_rate": 1.2337813567843784e-06, "loss": 0.0441, "step": 1598 }, { "epoch": 0.36382252559726963, "grad_norm": 1.1991943484955294, "learning_rate": 1.2337611274808901e-06, "loss": 0.0684, "step": 1599 }, { "epoch": 0.3640500568828214, "grad_norm": 1.223483082152576, "learning_rate": 1.2337408857354394e-06, "loss": 0.0464, "step": 1600 }, { "epoch": 0.36427758816837313, "grad_norm": 0.8679079039579355, "learning_rate": 1.2337206315484396e-06, "loss": 0.036, "step": 1601 }, { "epoch": 0.36450511945392494, "grad_norm": 1.1879547293476054, "learning_rate": 1.2337003649203049e-06, "loss": 0.0481, "step": 1602 }, { "epoch": 0.3647326507394767, "grad_norm": 1.3048080561875424, "learning_rate": 1.2336800858514498e-06, "loss": 0.057, "step": 1603 }, { "epoch": 0.36496018202502845, "grad_norm": 1.3761316306984896, "learning_rate": 1.2336597943422883e-06, "loss": 0.0806, "step": 1604 }, { "epoch": 0.3651877133105802, "grad_norm": 1.3617767409383836, "learning_rate": 1.2336394903932353e-06, "loss": 0.0636, "step": 1605 }, { "epoch": 0.36541524459613195, "grad_norm": 0.8022215628059131, "learning_rate": 1.233619174004706e-06, "loss": 0.0419, "step": 1606 }, { "epoch": 0.36564277588168376, "grad_norm": 1.0273812494717272, "learning_rate": 1.233598845177115e-06, "loss": 0.052, "step": 1607 }, { "epoch": 0.3658703071672355, "grad_norm": 1.262245877196861, "learning_rate": 1.2335785039108787e-06, "loss": 0.0494, "step": 1608 }, { "epoch": 0.36609783845278726, "grad_norm": 1.32897975384147, "learning_rate": 1.233558150206412e-06, "loss": 0.069, "step": 1609 }, { "epoch": 0.366325369738339, "grad_norm": 1.2103326320958596, "learning_rate": 1.2335377840641314e-06, "loss": 0.0388, "step": 1610 }, { "epoch": 0.36655290102389076, "grad_norm": 1.6475678381055654, "learning_rate": 1.233517405484453e-06, "loss": 0.0584, "step": 1611 }, { "epoch": 0.3667804323094426, "grad_norm": 0.8570759088496013, "learning_rate": 1.2334970144677929e-06, "loss": 0.0268, "step": 1612 }, { "epoch": 0.3670079635949943, "grad_norm": 1.2115212249577887, "learning_rate": 1.2334766110145684e-06, "loss": 0.0492, "step": 1613 }, { "epoch": 0.3672354948805461, "grad_norm": 1.6505586244119963, "learning_rate": 1.2334561951251967e-06, "loss": 0.0696, "step": 1614 }, { "epoch": 0.36746302616609783, "grad_norm": 1.0033818903908105, "learning_rate": 1.2334357668000943e-06, "loss": 0.0485, "step": 1615 }, { "epoch": 0.3676905574516496, "grad_norm": 0.8359701120507522, "learning_rate": 1.2334153260396795e-06, "loss": 0.0428, "step": 1616 }, { "epoch": 0.3679180887372014, "grad_norm": 1.3307257141956392, "learning_rate": 1.2333948728443692e-06, "loss": 0.0609, "step": 1617 }, { "epoch": 0.36814562002275314, "grad_norm": 0.9161807224363929, "learning_rate": 1.2333744072145824e-06, "loss": 0.0357, "step": 1618 }, { "epoch": 0.3683731513083049, "grad_norm": 1.5055305786265103, "learning_rate": 1.2333539291507365e-06, "loss": 0.1025, "step": 1619 }, { "epoch": 0.36860068259385664, "grad_norm": 1.936271932194028, "learning_rate": 1.2333334386532507e-06, "loss": 0.073, "step": 1620 }, { "epoch": 0.3688282138794084, "grad_norm": 0.9636982724990819, "learning_rate": 1.2333129357225434e-06, "loss": 0.0366, "step": 1621 }, { "epoch": 0.3690557451649602, "grad_norm": 1.4864238412613773, "learning_rate": 1.2332924203590341e-06, "loss": 0.0579, "step": 1622 }, { "epoch": 0.36928327645051195, "grad_norm": 1.1141779480557945, "learning_rate": 1.2332718925631414e-06, "loss": 0.0626, "step": 1623 }, { "epoch": 0.3695108077360637, "grad_norm": 1.4180072004359519, "learning_rate": 1.2332513523352853e-06, "loss": 0.0583, "step": 1624 }, { "epoch": 0.36973833902161546, "grad_norm": 1.3854251776786142, "learning_rate": 1.2332307996758854e-06, "loss": 0.0499, "step": 1625 }, { "epoch": 0.3699658703071672, "grad_norm": 2.0139670039119766, "learning_rate": 1.233210234585362e-06, "loss": 0.083, "step": 1626 }, { "epoch": 0.370193401592719, "grad_norm": 1.951694104084049, "learning_rate": 1.2331896570641354e-06, "loss": 0.08, "step": 1627 }, { "epoch": 0.37042093287827077, "grad_norm": 1.2223571341809563, "learning_rate": 1.233169067112626e-06, "loss": 0.0515, "step": 1628 }, { "epoch": 0.3706484641638225, "grad_norm": 1.845536817940592, "learning_rate": 1.2331484647312545e-06, "loss": 0.1018, "step": 1629 }, { "epoch": 0.3708759954493743, "grad_norm": 1.2829067703722639, "learning_rate": 1.2331278499204423e-06, "loss": 0.0577, "step": 1630 }, { "epoch": 0.371103526734926, "grad_norm": 1.5585929030011767, "learning_rate": 1.2331072226806107e-06, "loss": 0.0466, "step": 1631 }, { "epoch": 0.37133105802047783, "grad_norm": 1.6427062066371947, "learning_rate": 1.233086583012181e-06, "loss": 0.0855, "step": 1632 }, { "epoch": 0.3715585893060296, "grad_norm": 1.1255582962725759, "learning_rate": 1.233065930915575e-06, "loss": 0.0514, "step": 1633 }, { "epoch": 0.37178612059158134, "grad_norm": 1.1524468415244595, "learning_rate": 1.2330452663912155e-06, "loss": 0.0489, "step": 1634 }, { "epoch": 0.3720136518771331, "grad_norm": 3.45030750590545, "learning_rate": 1.233024589439524e-06, "loss": 0.0984, "step": 1635 }, { "epoch": 0.3722411831626849, "grad_norm": 1.3099017670753825, "learning_rate": 1.2330039000609233e-06, "loss": 0.0402, "step": 1636 }, { "epoch": 0.37246871444823665, "grad_norm": 1.9185688290740377, "learning_rate": 1.2329831982558365e-06, "loss": 0.0612, "step": 1637 }, { "epoch": 0.3726962457337884, "grad_norm": 1.0515979623712708, "learning_rate": 1.2329624840246867e-06, "loss": 0.0341, "step": 1638 }, { "epoch": 0.37292377701934015, "grad_norm": 1.0104966308830603, "learning_rate": 1.2329417573678974e-06, "loss": 0.0363, "step": 1639 }, { "epoch": 0.3731513083048919, "grad_norm": 0.8767867018961214, "learning_rate": 1.2329210182858915e-06, "loss": 0.0358, "step": 1640 }, { "epoch": 0.3733788395904437, "grad_norm": 1.2786801769613858, "learning_rate": 1.2329002667790937e-06, "loss": 0.0727, "step": 1641 }, { "epoch": 0.37360637087599546, "grad_norm": 1.0430455326740677, "learning_rate": 1.2328795028479275e-06, "loss": 0.0293, "step": 1642 }, { "epoch": 0.3738339021615472, "grad_norm": 1.3214304805194543, "learning_rate": 1.2328587264928176e-06, "loss": 0.0487, "step": 1643 }, { "epoch": 0.37406143344709897, "grad_norm": 2.17352926519868, "learning_rate": 1.2328379377141885e-06, "loss": 0.1179, "step": 1644 }, { "epoch": 0.3742889647326507, "grad_norm": 1.6384210150692329, "learning_rate": 1.2328171365124655e-06, "loss": 0.0356, "step": 1645 }, { "epoch": 0.3745164960182025, "grad_norm": 1.0746654458597755, "learning_rate": 1.2327963228880733e-06, "loss": 0.0348, "step": 1646 }, { "epoch": 0.3747440273037543, "grad_norm": 1.1554129673310896, "learning_rate": 1.2327754968414372e-06, "loss": 0.0559, "step": 1647 }, { "epoch": 0.37497155858930603, "grad_norm": 1.381713868304537, "learning_rate": 1.232754658372983e-06, "loss": 0.0429, "step": 1648 }, { "epoch": 0.3751990898748578, "grad_norm": 1.0461255300086107, "learning_rate": 1.2327338074831366e-06, "loss": 0.0498, "step": 1649 }, { "epoch": 0.37542662116040953, "grad_norm": 0.9696935937818977, "learning_rate": 1.2327129441723242e-06, "loss": 0.0282, "step": 1650 }, { "epoch": 0.37565415244596134, "grad_norm": 2.070008088112287, "learning_rate": 1.2326920684409724e-06, "loss": 0.0936, "step": 1651 }, { "epoch": 0.3758816837315131, "grad_norm": 1.9843357113833955, "learning_rate": 1.2326711802895077e-06, "loss": 0.068, "step": 1652 }, { "epoch": 0.37610921501706485, "grad_norm": 1.3824487035448465, "learning_rate": 1.2326502797183568e-06, "loss": 0.0609, "step": 1653 }, { "epoch": 0.3763367463026166, "grad_norm": 1.4651223595589276, "learning_rate": 1.2326293667279472e-06, "loss": 0.0448, "step": 1654 }, { "epoch": 0.37656427758816835, "grad_norm": 0.9774603013378036, "learning_rate": 1.232608441318706e-06, "loss": 0.0282, "step": 1655 }, { "epoch": 0.37679180887372016, "grad_norm": 1.1003503115295237, "learning_rate": 1.232587503491061e-06, "loss": 0.0447, "step": 1656 }, { "epoch": 0.3770193401592719, "grad_norm": 1.8648740699841149, "learning_rate": 1.2325665532454403e-06, "loss": 0.0879, "step": 1657 }, { "epoch": 0.37724687144482366, "grad_norm": 1.4508296738594828, "learning_rate": 1.2325455905822719e-06, "loss": 0.048, "step": 1658 }, { "epoch": 0.3774744027303754, "grad_norm": 1.5446190575894356, "learning_rate": 1.2325246155019844e-06, "loss": 0.0713, "step": 1659 }, { "epoch": 0.37770193401592717, "grad_norm": 0.657557765969774, "learning_rate": 1.2325036280050063e-06, "loss": 0.0252, "step": 1660 }, { "epoch": 0.377929465301479, "grad_norm": 0.7023883517908608, "learning_rate": 1.2324826280917664e-06, "loss": 0.0222, "step": 1661 }, { "epoch": 0.3781569965870307, "grad_norm": 1.2534079220919763, "learning_rate": 1.2324616157626943e-06, "loss": 0.0467, "step": 1662 }, { "epoch": 0.3783845278725825, "grad_norm": 1.4009224674260483, "learning_rate": 1.2324405910182195e-06, "loss": 0.0796, "step": 1663 }, { "epoch": 0.37861205915813423, "grad_norm": 2.444160860575137, "learning_rate": 1.2324195538587713e-06, "loss": 0.0772, "step": 1664 }, { "epoch": 0.378839590443686, "grad_norm": 1.1670387195512752, "learning_rate": 1.23239850428478e-06, "loss": 0.0648, "step": 1665 }, { "epoch": 0.3790671217292378, "grad_norm": 1.1973819971714186, "learning_rate": 1.2323774422966756e-06, "loss": 0.0466, "step": 1666 }, { "epoch": 0.37929465301478954, "grad_norm": 2.3889623346599484, "learning_rate": 1.2323563678948885e-06, "loss": 0.0995, "step": 1667 }, { "epoch": 0.3795221843003413, "grad_norm": 1.0719186425987624, "learning_rate": 1.2323352810798498e-06, "loss": 0.0434, "step": 1668 }, { "epoch": 0.37974971558589304, "grad_norm": 1.5135458293019917, "learning_rate": 1.23231418185199e-06, "loss": 0.1, "step": 1669 }, { "epoch": 0.3799772468714448, "grad_norm": 1.6715828908867492, "learning_rate": 1.2322930702117406e-06, "loss": 0.0654, "step": 1670 }, { "epoch": 0.3802047781569966, "grad_norm": 1.1117468704392939, "learning_rate": 1.232271946159533e-06, "loss": 0.0391, "step": 1671 }, { "epoch": 0.38043230944254836, "grad_norm": 1.7072937667796886, "learning_rate": 1.2322508096957992e-06, "loss": 0.0597, "step": 1672 }, { "epoch": 0.3806598407281001, "grad_norm": 2.695721182769726, "learning_rate": 1.2322296608209709e-06, "loss": 0.099, "step": 1673 }, { "epoch": 0.38088737201365186, "grad_norm": 2.29172385709693, "learning_rate": 1.2322084995354805e-06, "loss": 0.0939, "step": 1674 }, { "epoch": 0.38111490329920367, "grad_norm": 1.1798892776179337, "learning_rate": 1.2321873258397602e-06, "loss": 0.0544, "step": 1675 }, { "epoch": 0.3813424345847554, "grad_norm": 1.4030630144768808, "learning_rate": 1.232166139734243e-06, "loss": 0.0533, "step": 1676 }, { "epoch": 0.38156996587030717, "grad_norm": 1.2796792180280614, "learning_rate": 1.2321449412193622e-06, "loss": 0.0607, "step": 1677 }, { "epoch": 0.3817974971558589, "grad_norm": 1.3330548972713963, "learning_rate": 1.2321237302955505e-06, "loss": 0.0474, "step": 1678 }, { "epoch": 0.3820250284414107, "grad_norm": 1.0575562905738431, "learning_rate": 1.2321025069632416e-06, "loss": 0.0495, "step": 1679 }, { "epoch": 0.3822525597269625, "grad_norm": 1.1768156879508973, "learning_rate": 1.2320812712228694e-06, "loss": 0.0451, "step": 1680 }, { "epoch": 0.38248009101251423, "grad_norm": 0.7767797500459915, "learning_rate": 1.2320600230748677e-06, "loss": 0.0347, "step": 1681 }, { "epoch": 0.382707622298066, "grad_norm": 1.0249897852947059, "learning_rate": 1.232038762519671e-06, "loss": 0.0337, "step": 1682 }, { "epoch": 0.38293515358361774, "grad_norm": 1.6718567212147661, "learning_rate": 1.2320174895577138e-06, "loss": 0.1132, "step": 1683 }, { "epoch": 0.3831626848691695, "grad_norm": 0.994384807721514, "learning_rate": 1.2319962041894307e-06, "loss": 0.0362, "step": 1684 }, { "epoch": 0.3833902161547213, "grad_norm": 2.303334030981133, "learning_rate": 1.2319749064152569e-06, "loss": 0.1277, "step": 1685 }, { "epoch": 0.38361774744027305, "grad_norm": 1.609337504264491, "learning_rate": 1.2319535962356277e-06, "loss": 0.0575, "step": 1686 }, { "epoch": 0.3838452787258248, "grad_norm": 1.2011777956026055, "learning_rate": 1.2319322736509784e-06, "loss": 0.0552, "step": 1687 }, { "epoch": 0.38407281001137655, "grad_norm": 1.7322047315095004, "learning_rate": 1.2319109386617452e-06, "loss": 0.0753, "step": 1688 }, { "epoch": 0.3843003412969283, "grad_norm": 1.4045767611712419, "learning_rate": 1.2318895912683638e-06, "loss": 0.0726, "step": 1689 }, { "epoch": 0.3845278725824801, "grad_norm": 0.8734396508754521, "learning_rate": 1.2318682314712706e-06, "loss": 0.029, "step": 1690 }, { "epoch": 0.38475540386803186, "grad_norm": 3.125984298027021, "learning_rate": 1.2318468592709022e-06, "loss": 0.0794, "step": 1691 }, { "epoch": 0.3849829351535836, "grad_norm": 0.9959306519128199, "learning_rate": 1.2318254746676954e-06, "loss": 0.0504, "step": 1692 }, { "epoch": 0.38521046643913537, "grad_norm": 1.0411287511380087, "learning_rate": 1.2318040776620872e-06, "loss": 0.0423, "step": 1693 }, { "epoch": 0.3854379977246871, "grad_norm": 1.299794772066647, "learning_rate": 1.231782668254515e-06, "loss": 0.0633, "step": 1694 }, { "epoch": 0.3856655290102389, "grad_norm": 1.5302626083298043, "learning_rate": 1.2317612464454161e-06, "loss": 0.0581, "step": 1695 }, { "epoch": 0.3858930602957907, "grad_norm": 1.3051944839085843, "learning_rate": 1.2317398122352289e-06, "loss": 0.0524, "step": 1696 }, { "epoch": 0.38612059158134243, "grad_norm": 2.5811565293035166, "learning_rate": 1.2317183656243912e-06, "loss": 0.1304, "step": 1697 }, { "epoch": 0.3863481228668942, "grad_norm": 0.7679176052329142, "learning_rate": 1.231696906613341e-06, "loss": 0.0311, "step": 1698 }, { "epoch": 0.38657565415244594, "grad_norm": 1.2458183734125623, "learning_rate": 1.2316754352025173e-06, "loss": 0.0601, "step": 1699 }, { "epoch": 0.38680318543799774, "grad_norm": 1.4002525637777081, "learning_rate": 1.2316539513923585e-06, "loss": 0.052, "step": 1700 }, { "epoch": 0.3870307167235495, "grad_norm": 1.40244177955916, "learning_rate": 1.2316324551833042e-06, "loss": 0.0798, "step": 1701 }, { "epoch": 0.38725824800910125, "grad_norm": 0.636865674778868, "learning_rate": 1.2316109465757934e-06, "loss": 0.0236, "step": 1702 }, { "epoch": 0.387485779294653, "grad_norm": 0.911548468016125, "learning_rate": 1.231589425570266e-06, "loss": 0.048, "step": 1703 }, { "epoch": 0.38771331058020475, "grad_norm": 0.916105640198358, "learning_rate": 1.2315678921671615e-06, "loss": 0.0559, "step": 1704 }, { "epoch": 0.38794084186575656, "grad_norm": 1.1465421388050945, "learning_rate": 1.2315463463669202e-06, "loss": 0.0561, "step": 1705 }, { "epoch": 0.3881683731513083, "grad_norm": 1.441271671684643, "learning_rate": 1.2315247881699825e-06, "loss": 0.0507, "step": 1706 }, { "epoch": 0.38839590443686006, "grad_norm": 1.5270606085562182, "learning_rate": 1.2315032175767887e-06, "loss": 0.049, "step": 1707 }, { "epoch": 0.3886234357224118, "grad_norm": 1.5250540029402313, "learning_rate": 1.23148163458778e-06, "loss": 0.0597, "step": 1708 }, { "epoch": 0.3888509670079636, "grad_norm": 1.4207097878273431, "learning_rate": 1.2314600392033974e-06, "loss": 0.0676, "step": 1709 }, { "epoch": 0.3890784982935154, "grad_norm": 1.5896950089365356, "learning_rate": 1.2314384314240824e-06, "loss": 0.0617, "step": 1710 }, { "epoch": 0.3893060295790671, "grad_norm": 2.036043003063475, "learning_rate": 1.2314168112502765e-06, "loss": 0.1324, "step": 1711 }, { "epoch": 0.3895335608646189, "grad_norm": 1.1640099693184898, "learning_rate": 1.2313951786824213e-06, "loss": 0.0634, "step": 1712 }, { "epoch": 0.38976109215017063, "grad_norm": 1.4865385157328637, "learning_rate": 1.2313735337209593e-06, "loss": 0.0584, "step": 1713 }, { "epoch": 0.38998862343572244, "grad_norm": 1.252450965977985, "learning_rate": 1.231351876366333e-06, "loss": 0.0368, "step": 1714 }, { "epoch": 0.3902161547212742, "grad_norm": 1.4371658281059463, "learning_rate": 1.2313302066189846e-06, "loss": 0.0557, "step": 1715 }, { "epoch": 0.39044368600682594, "grad_norm": 1.579902726372128, "learning_rate": 1.2313085244793573e-06, "loss": 0.063, "step": 1716 }, { "epoch": 0.3906712172923777, "grad_norm": 2.3512912102608277, "learning_rate": 1.2312868299478944e-06, "loss": 0.0774, "step": 1717 }, { "epoch": 0.39089874857792944, "grad_norm": 1.2308435632473214, "learning_rate": 1.2312651230250387e-06, "loss": 0.0585, "step": 1718 }, { "epoch": 0.39112627986348125, "grad_norm": 1.0173462925036059, "learning_rate": 1.2312434037112345e-06, "loss": 0.0306, "step": 1719 }, { "epoch": 0.391353811149033, "grad_norm": 1.3611629215931131, "learning_rate": 1.2312216720069251e-06, "loss": 0.0822, "step": 1720 }, { "epoch": 0.39158134243458476, "grad_norm": 2.0854012174946135, "learning_rate": 1.2311999279125552e-06, "loss": 0.0877, "step": 1721 }, { "epoch": 0.3918088737201365, "grad_norm": 0.9865613588959612, "learning_rate": 1.2311781714285689e-06, "loss": 0.0557, "step": 1722 }, { "epoch": 0.39203640500568826, "grad_norm": 1.4304461856241661, "learning_rate": 1.231156402555411e-06, "loss": 0.0586, "step": 1723 }, { "epoch": 0.39226393629124007, "grad_norm": 1.0337361927182938, "learning_rate": 1.2311346212935262e-06, "loss": 0.0433, "step": 1724 }, { "epoch": 0.3924914675767918, "grad_norm": 0.7476757498111144, "learning_rate": 1.23111282764336e-06, "loss": 0.0303, "step": 1725 }, { "epoch": 0.39271899886234357, "grad_norm": 1.1404795389862592, "learning_rate": 1.2310910216053576e-06, "loss": 0.0518, "step": 1726 }, { "epoch": 0.3929465301478953, "grad_norm": 1.6483693490580467, "learning_rate": 1.2310692031799646e-06, "loss": 0.0754, "step": 1727 }, { "epoch": 0.3931740614334471, "grad_norm": 1.0959398843175312, "learning_rate": 1.2310473723676272e-06, "loss": 0.0444, "step": 1728 }, { "epoch": 0.3934015927189989, "grad_norm": 4.125333379715717, "learning_rate": 1.2310255291687913e-06, "loss": 0.0891, "step": 1729 }, { "epoch": 0.39362912400455063, "grad_norm": 1.4792691340855277, "learning_rate": 1.2310036735839037e-06, "loss": 0.0414, "step": 1730 }, { "epoch": 0.3938566552901024, "grad_norm": 1.4532148859913319, "learning_rate": 1.2309818056134108e-06, "loss": 0.0631, "step": 1731 }, { "epoch": 0.39408418657565414, "grad_norm": 0.7598305700170358, "learning_rate": 1.2309599252577593e-06, "loss": 0.026, "step": 1732 }, { "epoch": 0.3943117178612059, "grad_norm": 1.3606922738087333, "learning_rate": 1.230938032517397e-06, "loss": 0.0845, "step": 1733 }, { "epoch": 0.3945392491467577, "grad_norm": 0.9458332790402086, "learning_rate": 1.2309161273927708e-06, "loss": 0.0405, "step": 1734 }, { "epoch": 0.39476678043230945, "grad_norm": 1.338928802144235, "learning_rate": 1.2308942098843289e-06, "loss": 0.0506, "step": 1735 }, { "epoch": 0.3949943117178612, "grad_norm": 0.9875822099203543, "learning_rate": 1.2308722799925188e-06, "loss": 0.0474, "step": 1736 }, { "epoch": 0.39522184300341295, "grad_norm": 1.144643764479203, "learning_rate": 1.2308503377177887e-06, "loss": 0.0482, "step": 1737 }, { "epoch": 0.3954493742889647, "grad_norm": 1.2433153307693834, "learning_rate": 1.2308283830605877e-06, "loss": 0.0382, "step": 1738 }, { "epoch": 0.3956769055745165, "grad_norm": 0.8868244265843614, "learning_rate": 1.230806416021364e-06, "loss": 0.0312, "step": 1739 }, { "epoch": 0.39590443686006827, "grad_norm": 1.2733746862834583, "learning_rate": 1.2307844366005665e-06, "loss": 0.0558, "step": 1740 }, { "epoch": 0.39613196814562, "grad_norm": 1.5929660791826068, "learning_rate": 1.2307624447986446e-06, "loss": 0.0939, "step": 1741 }, { "epoch": 0.39635949943117177, "grad_norm": 1.6706588043136787, "learning_rate": 1.2307404406160476e-06, "loss": 0.0856, "step": 1742 }, { "epoch": 0.3965870307167236, "grad_norm": 1.4925539377680157, "learning_rate": 1.2307184240532255e-06, "loss": 0.051, "step": 1743 }, { "epoch": 0.39681456200227533, "grad_norm": 0.8314484619002069, "learning_rate": 1.2306963951106283e-06, "loss": 0.0398, "step": 1744 }, { "epoch": 0.3970420932878271, "grad_norm": 1.8438134884760746, "learning_rate": 1.2306743537887058e-06, "loss": 0.1539, "step": 1745 }, { "epoch": 0.39726962457337883, "grad_norm": 1.1012953571620363, "learning_rate": 1.2306523000879086e-06, "loss": 0.0395, "step": 1746 }, { "epoch": 0.3974971558589306, "grad_norm": 1.2928925052093003, "learning_rate": 1.230630234008688e-06, "loss": 0.0807, "step": 1747 }, { "epoch": 0.3977246871444824, "grad_norm": 0.9015308252174471, "learning_rate": 1.2306081555514942e-06, "loss": 0.0286, "step": 1748 }, { "epoch": 0.39795221843003414, "grad_norm": 1.4467952739416505, "learning_rate": 1.2305860647167792e-06, "loss": 0.0557, "step": 1749 }, { "epoch": 0.3981797497155859, "grad_norm": 0.8808019982734657, "learning_rate": 1.2305639615049938e-06, "loss": 0.033, "step": 1750 }, { "epoch": 0.39840728100113765, "grad_norm": 1.3896950910533592, "learning_rate": 1.2305418459165902e-06, "loss": 0.0425, "step": 1751 }, { "epoch": 0.3986348122866894, "grad_norm": 1.0876862572580541, "learning_rate": 1.2305197179520203e-06, "loss": 0.0529, "step": 1752 }, { "epoch": 0.3988623435722412, "grad_norm": 1.3270997207033957, "learning_rate": 1.2304975776117362e-06, "loss": 0.0539, "step": 1753 }, { "epoch": 0.39908987485779296, "grad_norm": 0.7035727027322921, "learning_rate": 1.2304754248961906e-06, "loss": 0.0319, "step": 1754 }, { "epoch": 0.3993174061433447, "grad_norm": 1.2785559710612535, "learning_rate": 1.2304532598058363e-06, "loss": 0.0481, "step": 1755 }, { "epoch": 0.39954493742889646, "grad_norm": 1.0072143636243023, "learning_rate": 1.230431082341126e-06, "loss": 0.0481, "step": 1756 }, { "epoch": 0.3997724687144482, "grad_norm": 1.063335290490128, "learning_rate": 1.2304088925025133e-06, "loss": 0.0405, "step": 1757 }, { "epoch": 0.4, "grad_norm": 1.3383471149297288, "learning_rate": 1.2303866902904515e-06, "loss": 0.0551, "step": 1758 }, { "epoch": 0.4002275312855518, "grad_norm": 1.3624741606263786, "learning_rate": 1.2303644757053945e-06, "loss": 0.0608, "step": 1759 }, { "epoch": 0.4004550625711035, "grad_norm": 1.5467274319242612, "learning_rate": 1.2303422487477965e-06, "loss": 0.0824, "step": 1760 }, { "epoch": 0.4006825938566553, "grad_norm": 0.9214849308730754, "learning_rate": 1.230320009418111e-06, "loss": 0.0367, "step": 1761 }, { "epoch": 0.40091012514220703, "grad_norm": 1.1758563975131424, "learning_rate": 1.2302977577167937e-06, "loss": 0.0549, "step": 1762 }, { "epoch": 0.40113765642775884, "grad_norm": 1.2826994675732137, "learning_rate": 1.2302754936442986e-06, "loss": 0.0591, "step": 1763 }, { "epoch": 0.4013651877133106, "grad_norm": 1.1321972166341843, "learning_rate": 1.2302532172010809e-06, "loss": 0.0309, "step": 1764 }, { "epoch": 0.40159271899886234, "grad_norm": 0.8928863966424516, "learning_rate": 1.2302309283875958e-06, "loss": 0.0443, "step": 1765 }, { "epoch": 0.4018202502844141, "grad_norm": 0.9140748312352776, "learning_rate": 1.230208627204299e-06, "loss": 0.0462, "step": 1766 }, { "epoch": 0.40204778156996585, "grad_norm": 1.365652466902771, "learning_rate": 1.2301863136516463e-06, "loss": 0.0575, "step": 1767 }, { "epoch": 0.40227531285551765, "grad_norm": 1.8435555715075644, "learning_rate": 1.2301639877300937e-06, "loss": 0.0835, "step": 1768 }, { "epoch": 0.4025028441410694, "grad_norm": 1.0698988090252757, "learning_rate": 1.2301416494400974e-06, "loss": 0.039, "step": 1769 }, { "epoch": 0.40273037542662116, "grad_norm": 1.5392783242527026, "learning_rate": 1.2301192987821142e-06, "loss": 0.0635, "step": 1770 }, { "epoch": 0.4029579067121729, "grad_norm": 1.992163233200094, "learning_rate": 1.2300969357566008e-06, "loss": 0.0687, "step": 1771 }, { "epoch": 0.40318543799772466, "grad_norm": 1.1143248581214018, "learning_rate": 1.230074560364014e-06, "loss": 0.0431, "step": 1772 }, { "epoch": 0.40341296928327647, "grad_norm": 0.9736384244561508, "learning_rate": 1.2300521726048114e-06, "loss": 0.0502, "step": 1773 }, { "epoch": 0.4036405005688282, "grad_norm": 1.6202139485972007, "learning_rate": 1.2300297724794506e-06, "loss": 0.0501, "step": 1774 }, { "epoch": 0.40386803185437997, "grad_norm": 2.6531740859909854, "learning_rate": 1.2300073599883892e-06, "loss": 0.0674, "step": 1775 }, { "epoch": 0.4040955631399317, "grad_norm": 1.4336293766578745, "learning_rate": 1.2299849351320854e-06, "loss": 0.0513, "step": 1776 }, { "epoch": 0.40432309442548353, "grad_norm": 0.9141317140330615, "learning_rate": 1.2299624979109976e-06, "loss": 0.0407, "step": 1777 }, { "epoch": 0.4045506257110353, "grad_norm": 1.1533277774875146, "learning_rate": 1.229940048325584e-06, "loss": 0.0561, "step": 1778 }, { "epoch": 0.40477815699658704, "grad_norm": 1.502695522459793, "learning_rate": 1.229917586376304e-06, "loss": 0.045, "step": 1779 }, { "epoch": 0.4050056882821388, "grad_norm": 1.8452181062352009, "learning_rate": 1.2298951120636163e-06, "loss": 0.0805, "step": 1780 }, { "epoch": 0.40523321956769054, "grad_norm": 1.5002423894959154, "learning_rate": 1.2298726253879802e-06, "loss": 0.0599, "step": 1781 }, { "epoch": 0.40546075085324235, "grad_norm": 1.988959282415028, "learning_rate": 1.2298501263498557e-06, "loss": 0.0951, "step": 1782 }, { "epoch": 0.4056882821387941, "grad_norm": 1.0188710069870275, "learning_rate": 1.229827614949702e-06, "loss": 0.0451, "step": 1783 }, { "epoch": 0.40591581342434585, "grad_norm": 1.44475701906824, "learning_rate": 1.22980509118798e-06, "loss": 0.0707, "step": 1784 }, { "epoch": 0.4061433447098976, "grad_norm": 1.5656238194585963, "learning_rate": 1.2297825550651491e-06, "loss": 0.0658, "step": 1785 }, { "epoch": 0.40637087599544935, "grad_norm": 1.1011276149008262, "learning_rate": 1.2297600065816707e-06, "loss": 0.0786, "step": 1786 }, { "epoch": 0.40659840728100116, "grad_norm": 1.5148687422019658, "learning_rate": 1.229737445738005e-06, "loss": 0.0519, "step": 1787 }, { "epoch": 0.4068259385665529, "grad_norm": 2.402810934984467, "learning_rate": 1.2297148725346137e-06, "loss": 0.0844, "step": 1788 }, { "epoch": 0.40705346985210467, "grad_norm": 0.9099961278233849, "learning_rate": 1.229692286971958e-06, "loss": 0.0249, "step": 1789 }, { "epoch": 0.4072810011376564, "grad_norm": 1.4154667950181927, "learning_rate": 1.2296696890504992e-06, "loss": 0.0473, "step": 1790 }, { "epoch": 0.40750853242320817, "grad_norm": 0.6978007467353329, "learning_rate": 1.2296470787706993e-06, "loss": 0.0294, "step": 1791 }, { "epoch": 0.40773606370876, "grad_norm": 0.7284846144551249, "learning_rate": 1.2296244561330206e-06, "loss": 0.0294, "step": 1792 }, { "epoch": 0.40796359499431173, "grad_norm": 1.3976750286698911, "learning_rate": 1.2296018211379253e-06, "loss": 0.0719, "step": 1793 }, { "epoch": 0.4081911262798635, "grad_norm": 1.7561963591958132, "learning_rate": 1.229579173785876e-06, "loss": 0.0929, "step": 1794 }, { "epoch": 0.40841865756541523, "grad_norm": 1.2130144152724294, "learning_rate": 1.2295565140773357e-06, "loss": 0.055, "step": 1795 }, { "epoch": 0.408646188850967, "grad_norm": 1.5931552656845787, "learning_rate": 1.2295338420127673e-06, "loss": 0.0996, "step": 1796 }, { "epoch": 0.4088737201365188, "grad_norm": 2.213301169642664, "learning_rate": 1.2295111575926344e-06, "loss": 0.0757, "step": 1797 }, { "epoch": 0.40910125142207054, "grad_norm": 0.7230584272794064, "learning_rate": 1.2294884608174007e-06, "loss": 0.0268, "step": 1798 }, { "epoch": 0.4093287827076223, "grad_norm": 1.8735567431020124, "learning_rate": 1.2294657516875297e-06, "loss": 0.0777, "step": 1799 }, { "epoch": 0.40955631399317405, "grad_norm": 1.4508553485648286, "learning_rate": 1.229443030203486e-06, "loss": 0.0592, "step": 1800 }, { "epoch": 0.4097838452787258, "grad_norm": 1.649254441958079, "learning_rate": 1.2294202963657335e-06, "loss": 0.0663, "step": 1801 }, { "epoch": 0.4100113765642776, "grad_norm": 1.2184400652023568, "learning_rate": 1.2293975501747372e-06, "loss": 0.058, "step": 1802 }, { "epoch": 0.41023890784982936, "grad_norm": 1.3977455445674583, "learning_rate": 1.229374791630962e-06, "loss": 0.0513, "step": 1803 }, { "epoch": 0.4104664391353811, "grad_norm": 1.7032814483710292, "learning_rate": 1.2293520207348727e-06, "loss": 0.0526, "step": 1804 }, { "epoch": 0.41069397042093286, "grad_norm": 1.506685007842939, "learning_rate": 1.229329237486935e-06, "loss": 0.0584, "step": 1805 }, { "epoch": 0.4109215017064846, "grad_norm": 1.1882085263706217, "learning_rate": 1.2293064418876145e-06, "loss": 0.0379, "step": 1806 }, { "epoch": 0.4111490329920364, "grad_norm": 0.8676322681005716, "learning_rate": 1.2292836339373771e-06, "loss": 0.0423, "step": 1807 }, { "epoch": 0.4113765642775882, "grad_norm": 1.0357464791970643, "learning_rate": 1.2292608136366887e-06, "loss": 0.0384, "step": 1808 }, { "epoch": 0.4116040955631399, "grad_norm": 1.2686531396882506, "learning_rate": 1.2292379809860162e-06, "loss": 0.048, "step": 1809 }, { "epoch": 0.4118316268486917, "grad_norm": 0.6203768192444805, "learning_rate": 1.2292151359858258e-06, "loss": 0.028, "step": 1810 }, { "epoch": 0.4120591581342435, "grad_norm": 0.9010901499957009, "learning_rate": 1.2291922786365846e-06, "loss": 0.0341, "step": 1811 }, { "epoch": 0.41228668941979524, "grad_norm": 1.1445391336093391, "learning_rate": 1.2291694089387599e-06, "loss": 0.0479, "step": 1812 }, { "epoch": 0.412514220705347, "grad_norm": 1.6521327765404437, "learning_rate": 1.2291465268928187e-06, "loss": 0.1017, "step": 1813 }, { "epoch": 0.41274175199089874, "grad_norm": 1.79397743679785, "learning_rate": 1.2291236324992291e-06, "loss": 0.1033, "step": 1814 }, { "epoch": 0.4129692832764505, "grad_norm": 1.5152525260856162, "learning_rate": 1.229100725758459e-06, "loss": 0.0685, "step": 1815 }, { "epoch": 0.4131968145620023, "grad_norm": 1.483120669394339, "learning_rate": 1.2290778066709763e-06, "loss": 0.0487, "step": 1816 }, { "epoch": 0.41342434584755405, "grad_norm": 0.9540882649733521, "learning_rate": 1.2290548752372494e-06, "loss": 0.0311, "step": 1817 }, { "epoch": 0.4136518771331058, "grad_norm": 0.7907197059168389, "learning_rate": 1.2290319314577473e-06, "loss": 0.0319, "step": 1818 }, { "epoch": 0.41387940841865756, "grad_norm": 0.9184118117485006, "learning_rate": 1.2290089753329386e-06, "loss": 0.0364, "step": 1819 }, { "epoch": 0.4141069397042093, "grad_norm": 0.7846265232624993, "learning_rate": 1.2289860068632929e-06, "loss": 0.0294, "step": 1820 }, { "epoch": 0.4143344709897611, "grad_norm": 1.507108248570942, "learning_rate": 1.228963026049279e-06, "loss": 0.0664, "step": 1821 }, { "epoch": 0.41456200227531287, "grad_norm": 1.9973321813531546, "learning_rate": 1.228940032891367e-06, "loss": 0.0638, "step": 1822 }, { "epoch": 0.4147895335608646, "grad_norm": 1.5615035868464353, "learning_rate": 1.2289170273900272e-06, "loss": 0.0643, "step": 1823 }, { "epoch": 0.4150170648464164, "grad_norm": 0.9932945336330554, "learning_rate": 1.228894009545729e-06, "loss": 0.0645, "step": 1824 }, { "epoch": 0.4152445961319681, "grad_norm": 1.5587019691492865, "learning_rate": 1.2288709793589434e-06, "loss": 0.1044, "step": 1825 }, { "epoch": 0.41547212741751993, "grad_norm": 1.786679686535424, "learning_rate": 1.2288479368301408e-06, "loss": 0.0724, "step": 1826 }, { "epoch": 0.4156996587030717, "grad_norm": 1.690300262458873, "learning_rate": 1.2288248819597922e-06, "loss": 0.091, "step": 1827 }, { "epoch": 0.41592718998862344, "grad_norm": 0.7331004990082142, "learning_rate": 1.228801814748369e-06, "loss": 0.0219, "step": 1828 }, { "epoch": 0.4161547212741752, "grad_norm": 1.3161127926479836, "learning_rate": 1.2287787351963427e-06, "loss": 0.0549, "step": 1829 }, { "epoch": 0.41638225255972694, "grad_norm": 2.165472731866584, "learning_rate": 1.2287556433041845e-06, "loss": 0.0976, "step": 1830 }, { "epoch": 0.41660978384527875, "grad_norm": 1.5021088638901345, "learning_rate": 1.2287325390723669e-06, "loss": 0.0676, "step": 1831 }, { "epoch": 0.4168373151308305, "grad_norm": 1.776482868261773, "learning_rate": 1.2287094225013618e-06, "loss": 0.0955, "step": 1832 }, { "epoch": 0.41706484641638225, "grad_norm": 1.4781242043850178, "learning_rate": 1.2286862935916416e-06, "loss": 0.087, "step": 1833 }, { "epoch": 0.417292377701934, "grad_norm": 0.5015421036696045, "learning_rate": 1.2286631523436793e-06, "loss": 0.0127, "step": 1834 }, { "epoch": 0.41751990898748575, "grad_norm": 1.0956161872763561, "learning_rate": 1.2286399987579478e-06, "loss": 0.0569, "step": 1835 }, { "epoch": 0.41774744027303756, "grad_norm": 0.9950125745841572, "learning_rate": 1.2286168328349202e-06, "loss": 0.0464, "step": 1836 }, { "epoch": 0.4179749715585893, "grad_norm": 0.9975401241840096, "learning_rate": 1.2285936545750698e-06, "loss": 0.0484, "step": 1837 }, { "epoch": 0.41820250284414107, "grad_norm": 1.1811400302907533, "learning_rate": 1.2285704639788707e-06, "loss": 0.0543, "step": 1838 }, { "epoch": 0.4184300341296928, "grad_norm": 1.3042909512695973, "learning_rate": 1.2285472610467969e-06, "loss": 0.0426, "step": 1839 }, { "epoch": 0.41865756541524457, "grad_norm": 0.9151776818093462, "learning_rate": 1.228524045779322e-06, "loss": 0.0403, "step": 1840 }, { "epoch": 0.4188850967007964, "grad_norm": 1.6251189485123572, "learning_rate": 1.2285008181769212e-06, "loss": 0.0503, "step": 1841 }, { "epoch": 0.41911262798634813, "grad_norm": 1.3392895278247294, "learning_rate": 1.228477578240069e-06, "loss": 0.0489, "step": 1842 }, { "epoch": 0.4193401592718999, "grad_norm": 1.2464077367956892, "learning_rate": 1.22845432596924e-06, "loss": 0.0405, "step": 1843 }, { "epoch": 0.41956769055745163, "grad_norm": 1.1623606536272444, "learning_rate": 1.22843106136491e-06, "loss": 0.0358, "step": 1844 }, { "epoch": 0.4197952218430034, "grad_norm": 1.0651018669316255, "learning_rate": 1.2284077844275543e-06, "loss": 0.0416, "step": 1845 }, { "epoch": 0.4200227531285552, "grad_norm": 1.84496334967716, "learning_rate": 1.2283844951576484e-06, "loss": 0.0468, "step": 1846 }, { "epoch": 0.42025028441410694, "grad_norm": 1.0014097401871376, "learning_rate": 1.2283611935556686e-06, "loss": 0.0334, "step": 1847 }, { "epoch": 0.4204778156996587, "grad_norm": 1.2875305555648238, "learning_rate": 1.2283378796220909e-06, "loss": 0.0568, "step": 1848 }, { "epoch": 0.42070534698521045, "grad_norm": 2.1705410042003095, "learning_rate": 1.228314553357392e-06, "loss": 0.0814, "step": 1849 }, { "epoch": 0.42093287827076226, "grad_norm": 0.9864984577581978, "learning_rate": 1.2282912147620483e-06, "loss": 0.035, "step": 1850 }, { "epoch": 0.421160409556314, "grad_norm": 1.0361264682055424, "learning_rate": 1.2282678638365373e-06, "loss": 0.0555, "step": 1851 }, { "epoch": 0.42138794084186576, "grad_norm": 1.2255325101947971, "learning_rate": 1.2282445005813359e-06, "loss": 0.0487, "step": 1852 }, { "epoch": 0.4216154721274175, "grad_norm": 1.21533819095114, "learning_rate": 1.2282211249969217e-06, "loss": 0.0401, "step": 1853 }, { "epoch": 0.42184300341296926, "grad_norm": 1.3664419425404826, "learning_rate": 1.2281977370837725e-06, "loss": 0.0615, "step": 1854 }, { "epoch": 0.42207053469852107, "grad_norm": 1.0812048471076985, "learning_rate": 1.2281743368423662e-06, "loss": 0.0346, "step": 1855 }, { "epoch": 0.4222980659840728, "grad_norm": 1.60512477075616, "learning_rate": 1.2281509242731813e-06, "loss": 0.0945, "step": 1856 }, { "epoch": 0.4225255972696246, "grad_norm": 1.6388845900863238, "learning_rate": 1.228127499376696e-06, "loss": 0.0818, "step": 1857 }, { "epoch": 0.4227531285551763, "grad_norm": 0.9782543566946358, "learning_rate": 1.228104062153389e-06, "loss": 0.0417, "step": 1858 }, { "epoch": 0.4229806598407281, "grad_norm": 0.9911432963854596, "learning_rate": 1.2280806126037396e-06, "loss": 0.0652, "step": 1859 }, { "epoch": 0.4232081911262799, "grad_norm": 1.6481089463408192, "learning_rate": 1.2280571507282272e-06, "loss": 0.104, "step": 1860 }, { "epoch": 0.42343572241183164, "grad_norm": 1.1739506923046776, "learning_rate": 1.2280336765273309e-06, "loss": 0.0499, "step": 1861 }, { "epoch": 0.4236632536973834, "grad_norm": 1.0261276957174543, "learning_rate": 1.2280101900015306e-06, "loss": 0.0467, "step": 1862 }, { "epoch": 0.42389078498293514, "grad_norm": 1.4414669057942235, "learning_rate": 1.2279866911513064e-06, "loss": 0.0559, "step": 1863 }, { "epoch": 0.4241183162684869, "grad_norm": 1.4975361597068904, "learning_rate": 1.2279631799771386e-06, "loss": 0.0816, "step": 1864 }, { "epoch": 0.4243458475540387, "grad_norm": 1.9040003169842432, "learning_rate": 1.2279396564795077e-06, "loss": 0.0914, "step": 1865 }, { "epoch": 0.42457337883959045, "grad_norm": 1.3040862552521917, "learning_rate": 1.2279161206588944e-06, "loss": 0.0402, "step": 1866 }, { "epoch": 0.4248009101251422, "grad_norm": 1.4037836584117365, "learning_rate": 1.2278925725157798e-06, "loss": 0.0583, "step": 1867 }, { "epoch": 0.42502844141069396, "grad_norm": 0.9243184821654356, "learning_rate": 1.2278690120506451e-06, "loss": 0.0461, "step": 1868 }, { "epoch": 0.4252559726962457, "grad_norm": 0.9991936622504879, "learning_rate": 1.2278454392639722e-06, "loss": 0.0342, "step": 1869 }, { "epoch": 0.4254835039817975, "grad_norm": 0.8487853121482433, "learning_rate": 1.2278218541562422e-06, "loss": 0.0301, "step": 1870 }, { "epoch": 0.42571103526734927, "grad_norm": 1.6907022090274095, "learning_rate": 1.2277982567279377e-06, "loss": 0.0786, "step": 1871 }, { "epoch": 0.425938566552901, "grad_norm": 1.7052726832363685, "learning_rate": 1.2277746469795407e-06, "loss": 0.0737, "step": 1872 }, { "epoch": 0.4261660978384528, "grad_norm": 1.3329829339197778, "learning_rate": 1.227751024911534e-06, "loss": 0.055, "step": 1873 }, { "epoch": 0.4263936291240045, "grad_norm": 1.3273033221545125, "learning_rate": 1.2277273905244002e-06, "loss": 0.0484, "step": 1874 }, { "epoch": 0.42662116040955633, "grad_norm": 1.1982373731004552, "learning_rate": 1.2277037438186224e-06, "loss": 0.0466, "step": 1875 }, { "epoch": 0.4268486916951081, "grad_norm": 1.0093718042807764, "learning_rate": 1.2276800847946839e-06, "loss": 0.0363, "step": 1876 }, { "epoch": 0.42707622298065984, "grad_norm": 0.9811547399478636, "learning_rate": 1.227656413453068e-06, "loss": 0.0425, "step": 1877 }, { "epoch": 0.4273037542662116, "grad_norm": 1.146528135934977, "learning_rate": 1.227632729794259e-06, "loss": 0.0485, "step": 1878 }, { "epoch": 0.42753128555176334, "grad_norm": 1.5824604547576673, "learning_rate": 1.2276090338187403e-06, "loss": 0.0487, "step": 1879 }, { "epoch": 0.42775881683731515, "grad_norm": 1.3160117730674203, "learning_rate": 1.2275853255269967e-06, "loss": 0.0446, "step": 1880 }, { "epoch": 0.4279863481228669, "grad_norm": 1.3027261584989582, "learning_rate": 1.2275616049195129e-06, "loss": 0.0395, "step": 1881 }, { "epoch": 0.42821387940841865, "grad_norm": 1.0814544003095643, "learning_rate": 1.2275378719967733e-06, "loss": 0.0292, "step": 1882 }, { "epoch": 0.4284414106939704, "grad_norm": 0.7587179405824312, "learning_rate": 1.227514126759263e-06, "loss": 0.0256, "step": 1883 }, { "epoch": 0.4286689419795222, "grad_norm": 1.4079924288332035, "learning_rate": 1.2274903692074674e-06, "loss": 0.0636, "step": 1884 }, { "epoch": 0.42889647326507396, "grad_norm": 1.024771832790498, "learning_rate": 1.2274665993418722e-06, "loss": 0.0399, "step": 1885 }, { "epoch": 0.4291240045506257, "grad_norm": 1.1395572191018248, "learning_rate": 1.227442817162963e-06, "loss": 0.0427, "step": 1886 }, { "epoch": 0.42935153583617747, "grad_norm": 1.9966203221580272, "learning_rate": 1.227419022671226e-06, "loss": 0.1059, "step": 1887 }, { "epoch": 0.4295790671217292, "grad_norm": 1.6299484076057391, "learning_rate": 1.2273952158671472e-06, "loss": 0.0633, "step": 1888 }, { "epoch": 0.429806598407281, "grad_norm": 1.2903207410298125, "learning_rate": 1.2273713967512137e-06, "loss": 0.0618, "step": 1889 }, { "epoch": 0.4300341296928328, "grad_norm": 1.3026692339522694, "learning_rate": 1.227347565323912e-06, "loss": 0.0795, "step": 1890 }, { "epoch": 0.43026166097838453, "grad_norm": 1.175449820116291, "learning_rate": 1.2273237215857293e-06, "loss": 0.0401, "step": 1891 }, { "epoch": 0.4304891922639363, "grad_norm": 1.6959835304167243, "learning_rate": 1.2272998655371526e-06, "loss": 0.0825, "step": 1892 }, { "epoch": 0.43071672354948803, "grad_norm": 1.4197022325382747, "learning_rate": 1.22727599717867e-06, "loss": 0.0576, "step": 1893 }, { "epoch": 0.43094425483503984, "grad_norm": 1.2471339436500026, "learning_rate": 1.2272521165107687e-06, "loss": 0.0722, "step": 1894 }, { "epoch": 0.4311717861205916, "grad_norm": 1.3643717577874817, "learning_rate": 1.2272282235339372e-06, "loss": 0.0449, "step": 1895 }, { "epoch": 0.43139931740614335, "grad_norm": 1.1744612621898987, "learning_rate": 1.2272043182486638e-06, "loss": 0.0416, "step": 1896 }, { "epoch": 0.4316268486916951, "grad_norm": 1.0384312678976468, "learning_rate": 1.227180400655437e-06, "loss": 0.0399, "step": 1897 }, { "epoch": 0.43185437997724685, "grad_norm": 1.7243528985527803, "learning_rate": 1.2271564707547457e-06, "loss": 0.1495, "step": 1898 }, { "epoch": 0.43208191126279866, "grad_norm": 0.9548507541370914, "learning_rate": 1.227132528547079e-06, "loss": 0.0415, "step": 1899 }, { "epoch": 0.4323094425483504, "grad_norm": 1.3617841660517656, "learning_rate": 1.2271085740329261e-06, "loss": 0.0607, "step": 1900 }, { "epoch": 0.43253697383390216, "grad_norm": 1.6402508487098346, "learning_rate": 1.2270846072127764e-06, "loss": 0.064, "step": 1901 }, { "epoch": 0.4327645051194539, "grad_norm": 1.131250294762535, "learning_rate": 1.2270606280871205e-06, "loss": 0.034, "step": 1902 }, { "epoch": 0.43299203640500566, "grad_norm": 1.2541444770869459, "learning_rate": 1.2270366366564476e-06, "loss": 0.0354, "step": 1903 }, { "epoch": 0.43321956769055747, "grad_norm": 1.1469941807844757, "learning_rate": 1.2270126329212486e-06, "loss": 0.0568, "step": 1904 }, { "epoch": 0.4334470989761092, "grad_norm": 0.9467488862348428, "learning_rate": 1.2269886168820138e-06, "loss": 0.0433, "step": 1905 }, { "epoch": 0.433674630261661, "grad_norm": 2.421156992660593, "learning_rate": 1.2269645885392342e-06, "loss": 0.1575, "step": 1906 }, { "epoch": 0.43390216154721273, "grad_norm": 1.2853239102420944, "learning_rate": 1.226940547893401e-06, "loss": 0.0523, "step": 1907 }, { "epoch": 0.4341296928327645, "grad_norm": 0.7677705648665133, "learning_rate": 1.2269164949450052e-06, "loss": 0.0367, "step": 1908 }, { "epoch": 0.4343572241183163, "grad_norm": 1.7484475434950735, "learning_rate": 1.2268924296945387e-06, "loss": 0.0551, "step": 1909 }, { "epoch": 0.43458475540386804, "grad_norm": 1.8872851215095336, "learning_rate": 1.2268683521424932e-06, "loss": 0.0713, "step": 1910 }, { "epoch": 0.4348122866894198, "grad_norm": 1.2082493686491087, "learning_rate": 1.226844262289361e-06, "loss": 0.0411, "step": 1911 }, { "epoch": 0.43503981797497154, "grad_norm": 1.7963587387869338, "learning_rate": 1.2268201601356342e-06, "loss": 0.0829, "step": 1912 }, { "epoch": 0.4352673492605233, "grad_norm": 1.2079928907070248, "learning_rate": 1.2267960456818054e-06, "loss": 0.0399, "step": 1913 }, { "epoch": 0.4354948805460751, "grad_norm": 1.3116442013593013, "learning_rate": 1.2267719189283676e-06, "loss": 0.0483, "step": 1914 }, { "epoch": 0.43572241183162685, "grad_norm": 1.3635629491104682, "learning_rate": 1.2267477798758141e-06, "loss": 0.0415, "step": 1915 }, { "epoch": 0.4359499431171786, "grad_norm": 1.1777595750610328, "learning_rate": 1.2267236285246376e-06, "loss": 0.0635, "step": 1916 }, { "epoch": 0.43617747440273036, "grad_norm": 1.54425482806484, "learning_rate": 1.2266994648753325e-06, "loss": 0.0713, "step": 1917 }, { "epoch": 0.43640500568828217, "grad_norm": 1.3128341631768956, "learning_rate": 1.2266752889283923e-06, "loss": 0.064, "step": 1918 }, { "epoch": 0.4366325369738339, "grad_norm": 1.3460506007652075, "learning_rate": 1.226651100684311e-06, "loss": 0.0489, "step": 1919 }, { "epoch": 0.43686006825938567, "grad_norm": 1.3320121603668689, "learning_rate": 1.2266269001435829e-06, "loss": 0.0756, "step": 1920 }, { "epoch": 0.4370875995449374, "grad_norm": 1.1661019863589135, "learning_rate": 1.226602687306703e-06, "loss": 0.0313, "step": 1921 }, { "epoch": 0.4373151308304892, "grad_norm": 1.320896853626411, "learning_rate": 1.226578462174166e-06, "loss": 0.0421, "step": 1922 }, { "epoch": 0.437542662116041, "grad_norm": 2.159753678644095, "learning_rate": 1.2265542247464668e-06, "loss": 0.1307, "step": 1923 }, { "epoch": 0.43777019340159273, "grad_norm": 1.5563003879211055, "learning_rate": 1.226529975024101e-06, "loss": 0.0593, "step": 1924 }, { "epoch": 0.4379977246871445, "grad_norm": 0.9689169184561489, "learning_rate": 1.2265057130075641e-06, "loss": 0.0552, "step": 1925 }, { "epoch": 0.43822525597269624, "grad_norm": 1.4053723286715865, "learning_rate": 1.2264814386973523e-06, "loss": 0.0761, "step": 1926 }, { "epoch": 0.438452787258248, "grad_norm": 1.373860475632686, "learning_rate": 1.2264571520939612e-06, "loss": 0.061, "step": 1927 }, { "epoch": 0.4386803185437998, "grad_norm": 1.5229851486555135, "learning_rate": 1.2264328531978875e-06, "loss": 0.0871, "step": 1928 }, { "epoch": 0.43890784982935155, "grad_norm": 0.9510671111937553, "learning_rate": 1.2264085420096277e-06, "loss": 0.0372, "step": 1929 }, { "epoch": 0.4391353811149033, "grad_norm": 1.411692319953813, "learning_rate": 1.226384218529679e-06, "loss": 0.0744, "step": 1930 }, { "epoch": 0.43936291240045505, "grad_norm": 0.9987666361393835, "learning_rate": 1.2263598827585379e-06, "loss": 0.0392, "step": 1931 }, { "epoch": 0.4395904436860068, "grad_norm": 0.8661437238334854, "learning_rate": 1.2263355346967023e-06, "loss": 0.0418, "step": 1932 }, { "epoch": 0.4398179749715586, "grad_norm": 1.2249390316316808, "learning_rate": 1.2263111743446697e-06, "loss": 0.0375, "step": 1933 }, { "epoch": 0.44004550625711036, "grad_norm": 1.5944859887412242, "learning_rate": 1.2262868017029377e-06, "loss": 0.062, "step": 1934 }, { "epoch": 0.4402730375426621, "grad_norm": 1.3941560967832323, "learning_rate": 1.226262416772005e-06, "loss": 0.065, "step": 1935 }, { "epoch": 0.44050056882821387, "grad_norm": 0.9120544787723815, "learning_rate": 1.2262380195523696e-06, "loss": 0.0392, "step": 1936 }, { "epoch": 0.4407281001137656, "grad_norm": 1.764018460971706, "learning_rate": 1.2262136100445303e-06, "loss": 0.0805, "step": 1937 }, { "epoch": 0.4409556313993174, "grad_norm": 1.1144365245430385, "learning_rate": 1.2261891882489855e-06, "loss": 0.0394, "step": 1938 }, { "epoch": 0.4411831626848692, "grad_norm": 1.1821432018479405, "learning_rate": 1.226164754166235e-06, "loss": 0.0411, "step": 1939 }, { "epoch": 0.44141069397042093, "grad_norm": 1.2454202923866486, "learning_rate": 1.2261403077967778e-06, "loss": 0.0441, "step": 1940 }, { "epoch": 0.4416382252559727, "grad_norm": 1.271626432566145, "learning_rate": 1.2261158491411136e-06, "loss": 0.0452, "step": 1941 }, { "epoch": 0.44186575654152443, "grad_norm": 1.0596255221554163, "learning_rate": 1.2260913781997425e-06, "loss": 0.0394, "step": 1942 }, { "epoch": 0.44209328782707624, "grad_norm": 1.4836582697063838, "learning_rate": 1.2260668949731644e-06, "loss": 0.0579, "step": 1943 }, { "epoch": 0.442320819112628, "grad_norm": 1.687109016999551, "learning_rate": 1.2260423994618798e-06, "loss": 0.0544, "step": 1944 }, { "epoch": 0.44254835039817975, "grad_norm": 1.2086563543761475, "learning_rate": 1.2260178916663892e-06, "loss": 0.0351, "step": 1945 }, { "epoch": 0.4427758816837315, "grad_norm": 0.9870058509060623, "learning_rate": 1.2259933715871935e-06, "loss": 0.0458, "step": 1946 }, { "epoch": 0.44300341296928325, "grad_norm": 1.2117861018085792, "learning_rate": 1.2259688392247942e-06, "loss": 0.0466, "step": 1947 }, { "epoch": 0.44323094425483506, "grad_norm": 1.4026535262541224, "learning_rate": 1.2259442945796926e-06, "loss": 0.0695, "step": 1948 }, { "epoch": 0.4434584755403868, "grad_norm": 0.7731288867011977, "learning_rate": 1.2259197376523898e-06, "loss": 0.033, "step": 1949 }, { "epoch": 0.44368600682593856, "grad_norm": 0.7657804826696037, "learning_rate": 1.2258951684433883e-06, "loss": 0.041, "step": 1950 }, { "epoch": 0.4439135381114903, "grad_norm": 1.8552107522089532, "learning_rate": 1.2258705869531901e-06, "loss": 0.0944, "step": 1951 }, { "epoch": 0.4441410693970421, "grad_norm": 1.111099312238907, "learning_rate": 1.2258459931822974e-06, "loss": 0.0651, "step": 1952 }, { "epoch": 0.4443686006825939, "grad_norm": 1.6141386501644925, "learning_rate": 1.225821387131213e-06, "loss": 0.0574, "step": 1953 }, { "epoch": 0.4445961319681456, "grad_norm": 0.9741551342534741, "learning_rate": 1.22579676880044e-06, "loss": 0.0398, "step": 1954 }, { "epoch": 0.4448236632536974, "grad_norm": 1.0138692072583615, "learning_rate": 1.2257721381904811e-06, "loss": 0.0547, "step": 1955 }, { "epoch": 0.44505119453924913, "grad_norm": 1.4460053506809138, "learning_rate": 1.22574749530184e-06, "loss": 0.0624, "step": 1956 }, { "epoch": 0.44527872582480094, "grad_norm": 1.9350781623487736, "learning_rate": 1.2257228401350205e-06, "loss": 0.0694, "step": 1957 }, { "epoch": 0.4455062571103527, "grad_norm": 0.8261029534789511, "learning_rate": 1.2256981726905262e-06, "loss": 0.0295, "step": 1958 }, { "epoch": 0.44573378839590444, "grad_norm": 1.6085761442666993, "learning_rate": 1.2256734929688612e-06, "loss": 0.0699, "step": 1959 }, { "epoch": 0.4459613196814562, "grad_norm": 1.2323407982068377, "learning_rate": 1.2256488009705303e-06, "loss": 0.0409, "step": 1960 }, { "epoch": 0.44618885096700794, "grad_norm": 1.6360239341263807, "learning_rate": 1.225624096696038e-06, "loss": 0.0645, "step": 1961 }, { "epoch": 0.44641638225255975, "grad_norm": 0.9576742303137104, "learning_rate": 1.225599380145889e-06, "loss": 0.0531, "step": 1962 }, { "epoch": 0.4466439135381115, "grad_norm": 1.545322326325835, "learning_rate": 1.2255746513205889e-06, "loss": 0.058, "step": 1963 }, { "epoch": 0.44687144482366326, "grad_norm": 0.9718630260311564, "learning_rate": 1.2255499102206423e-06, "loss": 0.0303, "step": 1964 }, { "epoch": 0.447098976109215, "grad_norm": 1.992196356794709, "learning_rate": 1.2255251568465558e-06, "loss": 0.0711, "step": 1965 }, { "epoch": 0.44732650739476676, "grad_norm": 1.758374736199823, "learning_rate": 1.2255003911988348e-06, "loss": 0.097, "step": 1966 }, { "epoch": 0.44755403868031857, "grad_norm": 1.1999149710277988, "learning_rate": 1.2254756132779855e-06, "loss": 0.0762, "step": 1967 }, { "epoch": 0.4477815699658703, "grad_norm": 0.750734148431945, "learning_rate": 1.2254508230845144e-06, "loss": 0.0366, "step": 1968 }, { "epoch": 0.44800910125142207, "grad_norm": 0.9217944145406265, "learning_rate": 1.2254260206189283e-06, "loss": 0.0388, "step": 1969 }, { "epoch": 0.4482366325369738, "grad_norm": 1.0142165579258442, "learning_rate": 1.2254012058817337e-06, "loss": 0.0582, "step": 1970 }, { "epoch": 0.4484641638225256, "grad_norm": 1.7763236230644281, "learning_rate": 1.2253763788734384e-06, "loss": 0.049, "step": 1971 }, { "epoch": 0.4486916951080774, "grad_norm": 1.1246346527787483, "learning_rate": 1.2253515395945492e-06, "loss": 0.0605, "step": 1972 }, { "epoch": 0.44891922639362913, "grad_norm": 1.0749214639270723, "learning_rate": 1.2253266880455742e-06, "loss": 0.0617, "step": 1973 }, { "epoch": 0.4491467576791809, "grad_norm": 0.7801246159886271, "learning_rate": 1.225301824227021e-06, "loss": 0.028, "step": 1974 }, { "epoch": 0.44937428896473264, "grad_norm": 1.3195521199092928, "learning_rate": 1.2252769481393979e-06, "loss": 0.0471, "step": 1975 }, { "epoch": 0.4496018202502844, "grad_norm": 1.5596892760854768, "learning_rate": 1.2252520597832132e-06, "loss": 0.0337, "step": 1976 }, { "epoch": 0.4498293515358362, "grad_norm": 0.8505107972505791, "learning_rate": 1.2252271591589759e-06, "loss": 0.0398, "step": 1977 }, { "epoch": 0.45005688282138795, "grad_norm": 1.1325746921296718, "learning_rate": 1.2252022462671947e-06, "loss": 0.0486, "step": 1978 }, { "epoch": 0.4502844141069397, "grad_norm": 1.813164319739497, "learning_rate": 1.2251773211083789e-06, "loss": 0.0991, "step": 1979 }, { "epoch": 0.45051194539249145, "grad_norm": 2.0540653923112164, "learning_rate": 1.225152383683038e-06, "loss": 0.0985, "step": 1980 }, { "epoch": 0.4507394766780432, "grad_norm": 1.3064957694509771, "learning_rate": 1.225127433991681e-06, "loss": 0.0408, "step": 1981 }, { "epoch": 0.450967007963595, "grad_norm": 1.055797400460298, "learning_rate": 1.2251024720348186e-06, "loss": 0.0427, "step": 1982 }, { "epoch": 0.45119453924914676, "grad_norm": 1.1174321415520143, "learning_rate": 1.2250774978129606e-06, "loss": 0.0436, "step": 1983 }, { "epoch": 0.4514220705346985, "grad_norm": 1.7220669335096757, "learning_rate": 1.2250525113266175e-06, "loss": 0.0855, "step": 1984 }, { "epoch": 0.45164960182025027, "grad_norm": 1.061543218680689, "learning_rate": 1.2250275125763002e-06, "loss": 0.0386, "step": 1985 }, { "epoch": 0.451877133105802, "grad_norm": 1.4190135452138255, "learning_rate": 1.2250025015625194e-06, "loss": 0.0616, "step": 1986 }, { "epoch": 0.4521046643913538, "grad_norm": 1.8336951622768547, "learning_rate": 1.2249774782857863e-06, "loss": 0.0527, "step": 1987 }, { "epoch": 0.4523321956769056, "grad_norm": 1.3808921921483752, "learning_rate": 1.2249524427466123e-06, "loss": 0.0449, "step": 1988 }, { "epoch": 0.45255972696245733, "grad_norm": 1.0413790190683898, "learning_rate": 1.2249273949455092e-06, "loss": 0.0571, "step": 1989 }, { "epoch": 0.4527872582480091, "grad_norm": 1.1506775133869898, "learning_rate": 1.224902334882989e-06, "loss": 0.0362, "step": 1990 }, { "epoch": 0.4530147895335609, "grad_norm": 0.5659500114505296, "learning_rate": 1.2248772625595633e-06, "loss": 0.0243, "step": 1991 }, { "epoch": 0.45324232081911264, "grad_norm": 1.8858126631285193, "learning_rate": 1.2248521779757455e-06, "loss": 0.0921, "step": 1992 }, { "epoch": 0.4534698521046644, "grad_norm": 1.4534701341984988, "learning_rate": 1.2248270811320473e-06, "loss": 0.0626, "step": 1993 }, { "epoch": 0.45369738339021615, "grad_norm": 0.957481343542944, "learning_rate": 1.2248019720289826e-06, "loss": 0.0273, "step": 1994 }, { "epoch": 0.4539249146757679, "grad_norm": 0.8791382523836305, "learning_rate": 1.2247768506670639e-06, "loss": 0.0589, "step": 1995 }, { "epoch": 0.4541524459613197, "grad_norm": 2.2598713011024794, "learning_rate": 1.2247517170468045e-06, "loss": 0.0993, "step": 1996 }, { "epoch": 0.45437997724687146, "grad_norm": 0.9870730424185967, "learning_rate": 1.2247265711687186e-06, "loss": 0.0355, "step": 1997 }, { "epoch": 0.4546075085324232, "grad_norm": 1.1900169451289013, "learning_rate": 1.2247014130333198e-06, "loss": 0.0525, "step": 1998 }, { "epoch": 0.45483503981797496, "grad_norm": 1.384236012560106, "learning_rate": 1.2246762426411227e-06, "loss": 0.0653, "step": 1999 }, { "epoch": 0.4550625711035267, "grad_norm": 1.3459272018496387, "learning_rate": 1.2246510599926412e-06, "loss": 0.0587, "step": 2000 }, { "epoch": 0.4552901023890785, "grad_norm": 1.4831181130397366, "learning_rate": 1.2246258650883904e-06, "loss": 0.0632, "step": 2001 }, { "epoch": 0.4555176336746303, "grad_norm": 1.4710214394162384, "learning_rate": 1.224600657928885e-06, "loss": 0.0707, "step": 2002 }, { "epoch": 0.455745164960182, "grad_norm": 1.6318800586995355, "learning_rate": 1.2245754385146402e-06, "loss": 0.0737, "step": 2003 }, { "epoch": 0.4559726962457338, "grad_norm": 1.3377789919521947, "learning_rate": 1.2245502068461718e-06, "loss": 0.0549, "step": 2004 }, { "epoch": 0.45620022753128553, "grad_norm": 1.5464491108432568, "learning_rate": 1.224524962923995e-06, "loss": 0.0414, "step": 2005 }, { "epoch": 0.45642775881683734, "grad_norm": 1.7217302295238814, "learning_rate": 1.2244997067486258e-06, "loss": 0.1082, "step": 2006 }, { "epoch": 0.4566552901023891, "grad_norm": 1.3057544355327637, "learning_rate": 1.2244744383205807e-06, "loss": 0.0553, "step": 2007 }, { "epoch": 0.45688282138794084, "grad_norm": 1.4993635003547536, "learning_rate": 1.224449157640376e-06, "loss": 0.0694, "step": 2008 }, { "epoch": 0.4571103526734926, "grad_norm": 1.4893330797563609, "learning_rate": 1.2244238647085283e-06, "loss": 0.0715, "step": 2009 }, { "epoch": 0.45733788395904434, "grad_norm": 1.8914472286199435, "learning_rate": 1.2243985595255546e-06, "loss": 0.0577, "step": 2010 }, { "epoch": 0.45756541524459615, "grad_norm": 1.8242839299021114, "learning_rate": 1.224373242091972e-06, "loss": 0.0565, "step": 2011 }, { "epoch": 0.4577929465301479, "grad_norm": 1.0856942810712096, "learning_rate": 1.224347912408298e-06, "loss": 0.0357, "step": 2012 }, { "epoch": 0.45802047781569966, "grad_norm": 1.3194141628678389, "learning_rate": 1.2243225704750506e-06, "loss": 0.0449, "step": 2013 }, { "epoch": 0.4582480091012514, "grad_norm": 1.860571028270787, "learning_rate": 1.2242972162927472e-06, "loss": 0.0855, "step": 2014 }, { "epoch": 0.45847554038680316, "grad_norm": 1.0986966810344647, "learning_rate": 1.2242718498619063e-06, "loss": 0.048, "step": 2015 }, { "epoch": 0.45870307167235497, "grad_norm": 1.286949991423601, "learning_rate": 1.2242464711830463e-06, "loss": 0.0648, "step": 2016 }, { "epoch": 0.4589306029579067, "grad_norm": 1.7867407908291044, "learning_rate": 1.224221080256686e-06, "loss": 0.0656, "step": 2017 }, { "epoch": 0.45915813424345847, "grad_norm": 0.9964611889527885, "learning_rate": 1.2241956770833444e-06, "loss": 0.0287, "step": 2018 }, { "epoch": 0.4593856655290102, "grad_norm": 0.758262966491813, "learning_rate": 1.2241702616635403e-06, "loss": 0.0266, "step": 2019 }, { "epoch": 0.459613196814562, "grad_norm": 2.5974139992788707, "learning_rate": 1.2241448339977931e-06, "loss": 0.111, "step": 2020 }, { "epoch": 0.4598407281001138, "grad_norm": 0.771625496801363, "learning_rate": 1.2241193940866234e-06, "loss": 0.0291, "step": 2021 }, { "epoch": 0.46006825938566553, "grad_norm": 1.04981191475765, "learning_rate": 1.22409394193055e-06, "loss": 0.0445, "step": 2022 }, { "epoch": 0.4602957906712173, "grad_norm": 1.1765006507267777, "learning_rate": 1.224068477530094e-06, "loss": 0.0534, "step": 2023 }, { "epoch": 0.46052332195676904, "grad_norm": 1.7004930482631033, "learning_rate": 1.2240430008857751e-06, "loss": 0.1093, "step": 2024 }, { "epoch": 0.46075085324232085, "grad_norm": 0.9880326133501596, "learning_rate": 1.2240175119981146e-06, "loss": 0.0596, "step": 2025 }, { "epoch": 0.4609783845278726, "grad_norm": 0.9331212134306713, "learning_rate": 1.223992010867633e-06, "loss": 0.0364, "step": 2026 }, { "epoch": 0.46120591581342435, "grad_norm": 0.592705357456653, "learning_rate": 1.223966497494852e-06, "loss": 0.0267, "step": 2027 }, { "epoch": 0.4614334470989761, "grad_norm": 2.103103329890297, "learning_rate": 1.2239409718802927e-06, "loss": 0.0789, "step": 2028 }, { "epoch": 0.46166097838452785, "grad_norm": 1.2774828791183843, "learning_rate": 1.2239154340244766e-06, "loss": 0.0617, "step": 2029 }, { "epoch": 0.46188850967007966, "grad_norm": 1.9492863350229845, "learning_rate": 1.2238898839279262e-06, "loss": 0.0715, "step": 2030 }, { "epoch": 0.4621160409556314, "grad_norm": 1.7906002686883862, "learning_rate": 1.223864321591163e-06, "loss": 0.0795, "step": 2031 }, { "epoch": 0.46234357224118316, "grad_norm": 1.0120261476046106, "learning_rate": 1.2238387470147103e-06, "loss": 0.0397, "step": 2032 }, { "epoch": 0.4625711035267349, "grad_norm": 1.4202832673434427, "learning_rate": 1.22381316019909e-06, "loss": 0.0551, "step": 2033 }, { "epoch": 0.46279863481228667, "grad_norm": 0.9745272894465992, "learning_rate": 1.2237875611448254e-06, "loss": 0.0634, "step": 2034 }, { "epoch": 0.4630261660978385, "grad_norm": 0.7432221539823165, "learning_rate": 1.2237619498524397e-06, "loss": 0.0218, "step": 2035 }, { "epoch": 0.46325369738339023, "grad_norm": 1.9819427368647708, "learning_rate": 1.2237363263224564e-06, "loss": 0.1136, "step": 2036 }, { "epoch": 0.463481228668942, "grad_norm": 1.538163547331391, "learning_rate": 1.2237106905553991e-06, "loss": 0.0581, "step": 2037 }, { "epoch": 0.46370875995449373, "grad_norm": 1.1781797084882926, "learning_rate": 1.2236850425517918e-06, "loss": 0.0593, "step": 2038 }, { "epoch": 0.4639362912400455, "grad_norm": 1.7417190267667673, "learning_rate": 1.2236593823121586e-06, "loss": 0.092, "step": 2039 }, { "epoch": 0.4641638225255973, "grad_norm": 1.2090125054365826, "learning_rate": 1.223633709837024e-06, "loss": 0.0238, "step": 2040 }, { "epoch": 0.46439135381114904, "grad_norm": 0.8736586233913058, "learning_rate": 1.2236080251269129e-06, "loss": 0.0436, "step": 2041 }, { "epoch": 0.4646188850967008, "grad_norm": 1.4877382030386759, "learning_rate": 1.2235823281823498e-06, "loss": 0.0673, "step": 2042 }, { "epoch": 0.46484641638225255, "grad_norm": 1.8784786331322982, "learning_rate": 1.2235566190038602e-06, "loss": 0.0872, "step": 2043 }, { "epoch": 0.4650739476678043, "grad_norm": 0.9749064913400318, "learning_rate": 1.2235308975919697e-06, "loss": 0.0408, "step": 2044 }, { "epoch": 0.4653014789533561, "grad_norm": 1.6562176997205134, "learning_rate": 1.2235051639472037e-06, "loss": 0.0593, "step": 2045 }, { "epoch": 0.46552901023890786, "grad_norm": 7.178097985744176, "learning_rate": 1.2234794180700883e-06, "loss": 0.2677, "step": 2046 }, { "epoch": 0.4657565415244596, "grad_norm": 1.2664824384389577, "learning_rate": 1.2234536599611496e-06, "loss": 0.0578, "step": 2047 }, { "epoch": 0.46598407281001136, "grad_norm": 1.0979349025374328, "learning_rate": 1.2234278896209142e-06, "loss": 0.053, "step": 2048 }, { "epoch": 0.4662116040955631, "grad_norm": 0.6490727646887132, "learning_rate": 1.2234021070499086e-06, "loss": 0.0262, "step": 2049 }, { "epoch": 0.4664391353811149, "grad_norm": 0.9140641916499196, "learning_rate": 1.22337631224866e-06, "loss": 0.0325, "step": 2050 }, { "epoch": 0.4666666666666667, "grad_norm": 1.0654195689636294, "learning_rate": 1.2233505052176952e-06, "loss": 0.0539, "step": 2051 }, { "epoch": 0.4668941979522184, "grad_norm": 1.2246413980257993, "learning_rate": 1.223324685957542e-06, "loss": 0.0539, "step": 2052 }, { "epoch": 0.4671217292377702, "grad_norm": 0.9496001567387945, "learning_rate": 1.2232988544687282e-06, "loss": 0.0523, "step": 2053 }, { "epoch": 0.46734926052332193, "grad_norm": 1.1280442595857205, "learning_rate": 1.2232730107517813e-06, "loss": 0.0483, "step": 2054 }, { "epoch": 0.46757679180887374, "grad_norm": 0.939380252104929, "learning_rate": 1.22324715480723e-06, "loss": 0.032, "step": 2055 }, { "epoch": 0.4678043230944255, "grad_norm": 1.56220155639698, "learning_rate": 1.2232212866356022e-06, "loss": 0.0744, "step": 2056 }, { "epoch": 0.46803185437997724, "grad_norm": 0.8923779970732735, "learning_rate": 1.2231954062374272e-06, "loss": 0.0278, "step": 2057 }, { "epoch": 0.468259385665529, "grad_norm": 1.366515353592018, "learning_rate": 1.2231695136132333e-06, "loss": 0.0699, "step": 2058 }, { "epoch": 0.4684869169510808, "grad_norm": 2.0148188928409616, "learning_rate": 1.2231436087635504e-06, "loss": 0.0599, "step": 2059 }, { "epoch": 0.46871444823663255, "grad_norm": 0.8270174628565943, "learning_rate": 1.2231176916889072e-06, "loss": 0.027, "step": 2060 }, { "epoch": 0.4689419795221843, "grad_norm": 1.2532661876313687, "learning_rate": 1.223091762389834e-06, "loss": 0.0684, "step": 2061 }, { "epoch": 0.46916951080773606, "grad_norm": 0.961835609964009, "learning_rate": 1.2230658208668606e-06, "loss": 0.0419, "step": 2062 }, { "epoch": 0.4693970420932878, "grad_norm": 1.2485009940950207, "learning_rate": 1.223039867120517e-06, "loss": 0.06, "step": 2063 }, { "epoch": 0.4696245733788396, "grad_norm": 1.0550492918928736, "learning_rate": 1.2230139011513338e-06, "loss": 0.0325, "step": 2064 }, { "epoch": 0.46985210466439137, "grad_norm": 1.0002830612932734, "learning_rate": 1.2229879229598416e-06, "loss": 0.0309, "step": 2065 }, { "epoch": 0.4700796359499431, "grad_norm": 1.0559040750849908, "learning_rate": 1.2229619325465717e-06, "loss": 0.044, "step": 2066 }, { "epoch": 0.47030716723549487, "grad_norm": 1.8211747050798512, "learning_rate": 1.2229359299120548e-06, "loss": 0.0544, "step": 2067 }, { "epoch": 0.4705346985210466, "grad_norm": 1.1244511125230865, "learning_rate": 1.2229099150568224e-06, "loss": 0.0458, "step": 2068 }, { "epoch": 0.47076222980659843, "grad_norm": 1.2446047468564831, "learning_rate": 1.2228838879814065e-06, "loss": 0.0454, "step": 2069 }, { "epoch": 0.4709897610921502, "grad_norm": 2.7103883547474132, "learning_rate": 1.222857848686339e-06, "loss": 0.1164, "step": 2070 }, { "epoch": 0.47121729237770194, "grad_norm": 1.0755709452712496, "learning_rate": 1.2228317971721518e-06, "loss": 0.0497, "step": 2071 }, { "epoch": 0.4714448236632537, "grad_norm": 1.3822921197656612, "learning_rate": 1.2228057334393777e-06, "loss": 0.0606, "step": 2072 }, { "epoch": 0.47167235494880544, "grad_norm": 1.436834207352602, "learning_rate": 1.2227796574885493e-06, "loss": 0.0759, "step": 2073 }, { "epoch": 0.47189988623435725, "grad_norm": 0.8895801225433585, "learning_rate": 1.2227535693201994e-06, "loss": 0.0278, "step": 2074 }, { "epoch": 0.472127417519909, "grad_norm": 0.9871504029067106, "learning_rate": 1.2227274689348612e-06, "loss": 0.0339, "step": 2075 }, { "epoch": 0.47235494880546075, "grad_norm": 2.4304810090092217, "learning_rate": 1.2227013563330684e-06, "loss": 0.1278, "step": 2076 }, { "epoch": 0.4725824800910125, "grad_norm": 1.3611006420880847, "learning_rate": 1.2226752315153543e-06, "loss": 0.056, "step": 2077 }, { "epoch": 0.47281001137656425, "grad_norm": 1.2835019677874184, "learning_rate": 1.2226490944822533e-06, "loss": 0.0495, "step": 2078 }, { "epoch": 0.47303754266211606, "grad_norm": 0.9616764487213456, "learning_rate": 1.2226229452342991e-06, "loss": 0.0346, "step": 2079 }, { "epoch": 0.4732650739476678, "grad_norm": 1.0696378165797904, "learning_rate": 1.2225967837720265e-06, "loss": 0.0408, "step": 2080 }, { "epoch": 0.47349260523321957, "grad_norm": 1.40355131716461, "learning_rate": 1.2225706100959702e-06, "loss": 0.0567, "step": 2081 }, { "epoch": 0.4737201365187713, "grad_norm": 1.228435046083943, "learning_rate": 1.2225444242066648e-06, "loss": 0.0483, "step": 2082 }, { "epoch": 0.47394766780432307, "grad_norm": 2.414708407757396, "learning_rate": 1.2225182261046458e-06, "loss": 0.0817, "step": 2083 }, { "epoch": 0.4741751990898749, "grad_norm": 1.3036262799520595, "learning_rate": 1.2224920157904487e-06, "loss": 0.0552, "step": 2084 }, { "epoch": 0.47440273037542663, "grad_norm": 1.4961223747917856, "learning_rate": 1.222465793264609e-06, "loss": 0.0653, "step": 2085 }, { "epoch": 0.4746302616609784, "grad_norm": 0.9962870632246056, "learning_rate": 1.2224395585276626e-06, "loss": 0.059, "step": 2086 }, { "epoch": 0.47485779294653013, "grad_norm": 1.1085210761396838, "learning_rate": 1.2224133115801458e-06, "loss": 0.0415, "step": 2087 }, { "epoch": 0.4750853242320819, "grad_norm": 2.147238444020511, "learning_rate": 1.222387052422595e-06, "loss": 0.0832, "step": 2088 }, { "epoch": 0.4753128555176337, "grad_norm": 1.1789213043757707, "learning_rate": 1.222360781055547e-06, "loss": 0.0439, "step": 2089 }, { "epoch": 0.47554038680318544, "grad_norm": 1.0136823057916216, "learning_rate": 1.2223344974795386e-06, "loss": 0.0491, "step": 2090 }, { "epoch": 0.4757679180887372, "grad_norm": 1.0762034069369268, "learning_rate": 1.2223082016951071e-06, "loss": 0.0414, "step": 2091 }, { "epoch": 0.47599544937428895, "grad_norm": 0.7861883267498344, "learning_rate": 1.2222818937027898e-06, "loss": 0.0225, "step": 2092 }, { "epoch": 0.47622298065984076, "grad_norm": 1.4307219009209873, "learning_rate": 1.2222555735031245e-06, "loss": 0.0518, "step": 2093 }, { "epoch": 0.4764505119453925, "grad_norm": 1.7521480356699026, "learning_rate": 1.222229241096649e-06, "loss": 0.0686, "step": 2094 }, { "epoch": 0.47667804323094426, "grad_norm": 1.084257413613486, "learning_rate": 1.2222028964839017e-06, "loss": 0.0623, "step": 2095 }, { "epoch": 0.476905574516496, "grad_norm": 0.9853812502387866, "learning_rate": 1.2221765396654209e-06, "loss": 0.0322, "step": 2096 }, { "epoch": 0.47713310580204776, "grad_norm": 1.593782379839956, "learning_rate": 1.2221501706417454e-06, "loss": 0.0776, "step": 2097 }, { "epoch": 0.47736063708759957, "grad_norm": 1.2316203058040796, "learning_rate": 1.2221237894134138e-06, "loss": 0.0415, "step": 2098 }, { "epoch": 0.4775881683731513, "grad_norm": 1.7492243550591817, "learning_rate": 1.222097395980966e-06, "loss": 0.0863, "step": 2099 }, { "epoch": 0.4778156996587031, "grad_norm": 1.5123822677786471, "learning_rate": 1.2220709903449403e-06, "loss": 0.0839, "step": 2100 }, { "epoch": 0.4780432309442548, "grad_norm": 0.6925370489166974, "learning_rate": 1.2220445725058775e-06, "loss": 0.036, "step": 2101 }, { "epoch": 0.4782707622298066, "grad_norm": 1.416495751104088, "learning_rate": 1.222018142464317e-06, "loss": 0.0745, "step": 2102 }, { "epoch": 0.4784982935153584, "grad_norm": 1.3844263789079834, "learning_rate": 1.221991700220799e-06, "loss": 0.0635, "step": 2103 }, { "epoch": 0.47872582480091014, "grad_norm": 1.293007537973823, "learning_rate": 1.221965245775864e-06, "loss": 0.0489, "step": 2104 }, { "epoch": 0.4789533560864619, "grad_norm": 1.4567681046992622, "learning_rate": 1.2219387791300527e-06, "loss": 0.0645, "step": 2105 }, { "epoch": 0.47918088737201364, "grad_norm": 0.9331875777291676, "learning_rate": 1.2219123002839058e-06, "loss": 0.0374, "step": 2106 }, { "epoch": 0.4794084186575654, "grad_norm": 1.0880231881224058, "learning_rate": 1.2218858092379648e-06, "loss": 0.0495, "step": 2107 }, { "epoch": 0.4796359499431172, "grad_norm": 2.087839894098663, "learning_rate": 1.221859305992771e-06, "loss": 0.0775, "step": 2108 }, { "epoch": 0.47986348122866895, "grad_norm": 2.0258749667718408, "learning_rate": 1.2218327905488662e-06, "loss": 0.0928, "step": 2109 }, { "epoch": 0.4800910125142207, "grad_norm": 1.077278940685259, "learning_rate": 1.221806262906792e-06, "loss": 0.0395, "step": 2110 }, { "epoch": 0.48031854379977246, "grad_norm": 1.3800617021949446, "learning_rate": 1.221779723067091e-06, "loss": 0.0547, "step": 2111 }, { "epoch": 0.4805460750853242, "grad_norm": 1.1882918051107954, "learning_rate": 1.2217531710303053e-06, "loss": 0.0646, "step": 2112 }, { "epoch": 0.480773606370876, "grad_norm": 1.648037642948223, "learning_rate": 1.2217266067969778e-06, "loss": 0.0883, "step": 2113 }, { "epoch": 0.48100113765642777, "grad_norm": 1.1114503968376748, "learning_rate": 1.221700030367651e-06, "loss": 0.0532, "step": 2114 }, { "epoch": 0.4812286689419795, "grad_norm": 2.318260685332167, "learning_rate": 1.2216734417428686e-06, "loss": 0.1156, "step": 2115 }, { "epoch": 0.4814562002275313, "grad_norm": 0.943158220881948, "learning_rate": 1.2216468409231738e-06, "loss": 0.0478, "step": 2116 }, { "epoch": 0.481683731513083, "grad_norm": 1.3059340047526014, "learning_rate": 1.2216202279091104e-06, "loss": 0.0474, "step": 2117 }, { "epoch": 0.48191126279863483, "grad_norm": 1.5108825640696393, "learning_rate": 1.2215936027012221e-06, "loss": 0.0528, "step": 2118 }, { "epoch": 0.4821387940841866, "grad_norm": 1.6389721165810114, "learning_rate": 1.2215669653000532e-06, "loss": 0.0636, "step": 2119 }, { "epoch": 0.48236632536973834, "grad_norm": 0.8263598171620422, "learning_rate": 1.2215403157061478e-06, "loss": 0.0288, "step": 2120 }, { "epoch": 0.4825938566552901, "grad_norm": 1.9615999942786497, "learning_rate": 1.2215136539200512e-06, "loss": 0.0735, "step": 2121 }, { "epoch": 0.48282138794084184, "grad_norm": 1.052892038422689, "learning_rate": 1.2214869799423078e-06, "loss": 0.0375, "step": 2122 }, { "epoch": 0.48304891922639365, "grad_norm": 1.0184495668702673, "learning_rate": 1.2214602937734632e-06, "loss": 0.0457, "step": 2123 }, { "epoch": 0.4832764505119454, "grad_norm": 2.1042829626272868, "learning_rate": 1.2214335954140624e-06, "loss": 0.1008, "step": 2124 }, { "epoch": 0.48350398179749715, "grad_norm": 1.2825506743830564, "learning_rate": 1.221406884864651e-06, "loss": 0.0542, "step": 2125 }, { "epoch": 0.4837315130830489, "grad_norm": 1.2404980369925334, "learning_rate": 1.2213801621257754e-06, "loss": 0.076, "step": 2126 }, { "epoch": 0.48395904436860065, "grad_norm": 1.6535875011294203, "learning_rate": 1.2213534271979815e-06, "loss": 0.0826, "step": 2127 }, { "epoch": 0.48418657565415246, "grad_norm": 1.5742759100449686, "learning_rate": 1.2213266800818158e-06, "loss": 0.0722, "step": 2128 }, { "epoch": 0.4844141069397042, "grad_norm": 1.4590518459696438, "learning_rate": 1.2212999207778246e-06, "loss": 0.0626, "step": 2129 }, { "epoch": 0.48464163822525597, "grad_norm": 1.032592661119547, "learning_rate": 1.2212731492865552e-06, "loss": 0.0352, "step": 2130 }, { "epoch": 0.4848691695108077, "grad_norm": 1.943068406348077, "learning_rate": 1.2212463656085548e-06, "loss": 0.0878, "step": 2131 }, { "epoch": 0.4850967007963595, "grad_norm": 1.4222918495364802, "learning_rate": 1.2212195697443704e-06, "loss": 0.0662, "step": 2132 }, { "epoch": 0.4853242320819113, "grad_norm": 1.0835698936451348, "learning_rate": 1.2211927616945502e-06, "loss": 0.0315, "step": 2133 }, { "epoch": 0.48555176336746303, "grad_norm": 1.4804001036145082, "learning_rate": 1.2211659414596417e-06, "loss": 0.057, "step": 2134 }, { "epoch": 0.4857792946530148, "grad_norm": 1.0372510496516572, "learning_rate": 1.2211391090401931e-06, "loss": 0.0347, "step": 2135 }, { "epoch": 0.48600682593856653, "grad_norm": 1.4539414447995862, "learning_rate": 1.2211122644367531e-06, "loss": 0.0546, "step": 2136 }, { "epoch": 0.48623435722411834, "grad_norm": 0.8272464584595134, "learning_rate": 1.2210854076498699e-06, "loss": 0.0291, "step": 2137 }, { "epoch": 0.4864618885096701, "grad_norm": 1.082723119747633, "learning_rate": 1.2210585386800927e-06, "loss": 0.0631, "step": 2138 }, { "epoch": 0.48668941979522184, "grad_norm": 1.7613862158027351, "learning_rate": 1.2210316575279707e-06, "loss": 0.0516, "step": 2139 }, { "epoch": 0.4869169510807736, "grad_norm": 1.05018126679848, "learning_rate": 1.221004764194053e-06, "loss": 0.0503, "step": 2140 }, { "epoch": 0.48714448236632535, "grad_norm": 1.4765291215619578, "learning_rate": 1.22097785867889e-06, "loss": 0.0462, "step": 2141 }, { "epoch": 0.48737201365187716, "grad_norm": 1.6033668171289224, "learning_rate": 1.2209509409830304e-06, "loss": 0.0656, "step": 2142 }, { "epoch": 0.4875995449374289, "grad_norm": 1.2469591088986771, "learning_rate": 1.2209240111070254e-06, "loss": 0.0644, "step": 2143 }, { "epoch": 0.48782707622298066, "grad_norm": 1.676030146319255, "learning_rate": 1.2208970690514247e-06, "loss": 0.0502, "step": 2144 }, { "epoch": 0.4880546075085324, "grad_norm": 0.89826172703067, "learning_rate": 1.2208701148167795e-06, "loss": 0.0378, "step": 2145 }, { "epoch": 0.48828213879408416, "grad_norm": 1.2953659435102682, "learning_rate": 1.2208431484036405e-06, "loss": 0.0532, "step": 2146 }, { "epoch": 0.48850967007963597, "grad_norm": 1.6937135620824826, "learning_rate": 1.2208161698125584e-06, "loss": 0.0777, "step": 2147 }, { "epoch": 0.4887372013651877, "grad_norm": 1.4516409315976513, "learning_rate": 1.2207891790440852e-06, "loss": 0.0532, "step": 2148 }, { "epoch": 0.4889647326507395, "grad_norm": 1.466966804713568, "learning_rate": 1.2207621760987723e-06, "loss": 0.0542, "step": 2149 }, { "epoch": 0.4891922639362912, "grad_norm": 1.5403030963325346, "learning_rate": 1.2207351609771718e-06, "loss": 0.0699, "step": 2150 }, { "epoch": 0.489419795221843, "grad_norm": 1.1658099032169384, "learning_rate": 1.2207081336798352e-06, "loss": 0.0747, "step": 2151 }, { "epoch": 0.4896473265073948, "grad_norm": 1.272828829650441, "learning_rate": 1.2206810942073158e-06, "loss": 0.0339, "step": 2152 }, { "epoch": 0.48987485779294654, "grad_norm": 0.8638651055871553, "learning_rate": 1.2206540425601653e-06, "loss": 0.0258, "step": 2153 }, { "epoch": 0.4901023890784983, "grad_norm": 1.1406798846203596, "learning_rate": 1.2206269787389374e-06, "loss": 0.0434, "step": 2154 }, { "epoch": 0.49032992036405004, "grad_norm": 1.3546994084602113, "learning_rate": 1.220599902744185e-06, "loss": 0.0505, "step": 2155 }, { "epoch": 0.4905574516496018, "grad_norm": 0.8028716165840231, "learning_rate": 1.220572814576461e-06, "loss": 0.0252, "step": 2156 }, { "epoch": 0.4907849829351536, "grad_norm": 1.5525391069467074, "learning_rate": 1.2205457142363197e-06, "loss": 0.0671, "step": 2157 }, { "epoch": 0.49101251422070535, "grad_norm": 1.5565330239281656, "learning_rate": 1.2205186017243146e-06, "loss": 0.0485, "step": 2158 }, { "epoch": 0.4912400455062571, "grad_norm": 0.9990569085398061, "learning_rate": 1.2204914770409999e-06, "loss": 0.0379, "step": 2159 }, { "epoch": 0.49146757679180886, "grad_norm": 1.3885909890564816, "learning_rate": 1.22046434018693e-06, "loss": 0.0815, "step": 2160 }, { "epoch": 0.4916951080773606, "grad_norm": 1.0037322866680125, "learning_rate": 1.2204371911626597e-06, "loss": 0.0409, "step": 2161 }, { "epoch": 0.4919226393629124, "grad_norm": 2.005545489895357, "learning_rate": 1.2204100299687436e-06, "loss": 0.1019, "step": 2162 }, { "epoch": 0.49215017064846417, "grad_norm": 1.5901419467739697, "learning_rate": 1.2203828566057368e-06, "loss": 0.0497, "step": 2163 }, { "epoch": 0.4923777019340159, "grad_norm": 1.0024861467422042, "learning_rate": 1.2203556710741948e-06, "loss": 0.0537, "step": 2164 }, { "epoch": 0.4926052332195677, "grad_norm": 1.1764359755022478, "learning_rate": 1.2203284733746734e-06, "loss": 0.0609, "step": 2165 }, { "epoch": 0.4928327645051195, "grad_norm": 1.4780687059085984, "learning_rate": 1.2203012635077283e-06, "loss": 0.0467, "step": 2166 }, { "epoch": 0.49306029579067123, "grad_norm": 1.0636152584968832, "learning_rate": 1.2202740414739153e-06, "loss": 0.0334, "step": 2167 }, { "epoch": 0.493287827076223, "grad_norm": 0.9777542514618388, "learning_rate": 1.2202468072737914e-06, "loss": 0.0299, "step": 2168 }, { "epoch": 0.49351535836177474, "grad_norm": 1.099998345093073, "learning_rate": 1.2202195609079128e-06, "loss": 0.0349, "step": 2169 }, { "epoch": 0.4937428896473265, "grad_norm": 0.9946626000618751, "learning_rate": 1.2201923023768365e-06, "loss": 0.0439, "step": 2170 }, { "epoch": 0.4939704209328783, "grad_norm": 2.1311963123579707, "learning_rate": 1.2201650316811194e-06, "loss": 0.0871, "step": 2171 }, { "epoch": 0.49419795221843005, "grad_norm": 1.492842268786691, "learning_rate": 1.2201377488213195e-06, "loss": 0.0851, "step": 2172 }, { "epoch": 0.4944254835039818, "grad_norm": 2.5237867181473876, "learning_rate": 1.2201104537979934e-06, "loss": 0.0637, "step": 2173 }, { "epoch": 0.49465301478953355, "grad_norm": 2.556432817981331, "learning_rate": 1.2200831466117e-06, "loss": 0.1105, "step": 2174 }, { "epoch": 0.4948805460750853, "grad_norm": 0.8981926808894833, "learning_rate": 1.2200558272629967e-06, "loss": 0.03, "step": 2175 }, { "epoch": 0.4951080773606371, "grad_norm": 1.6324971068278198, "learning_rate": 1.2200284957524421e-06, "loss": 0.063, "step": 2176 }, { "epoch": 0.49533560864618886, "grad_norm": 1.0341597824022535, "learning_rate": 1.2200011520805947e-06, "loss": 0.0533, "step": 2177 }, { "epoch": 0.4955631399317406, "grad_norm": 0.8107094209291485, "learning_rate": 1.2199737962480136e-06, "loss": 0.0427, "step": 2178 }, { "epoch": 0.49579067121729237, "grad_norm": 1.106894286453914, "learning_rate": 1.2199464282552576e-06, "loss": 0.0594, "step": 2179 }, { "epoch": 0.4960182025028441, "grad_norm": 1.7291263581636043, "learning_rate": 1.2199190481028864e-06, "loss": 0.0675, "step": 2180 }, { "epoch": 0.4962457337883959, "grad_norm": 0.9412668127265857, "learning_rate": 1.2198916557914592e-06, "loss": 0.0364, "step": 2181 }, { "epoch": 0.4964732650739477, "grad_norm": 1.4114602905566138, "learning_rate": 1.2198642513215362e-06, "loss": 0.0652, "step": 2182 }, { "epoch": 0.49670079635949943, "grad_norm": 0.609976074205654, "learning_rate": 1.2198368346936774e-06, "loss": 0.03, "step": 2183 }, { "epoch": 0.4969283276450512, "grad_norm": 1.2570918433725233, "learning_rate": 1.219809405908443e-06, "loss": 0.0479, "step": 2184 }, { "epoch": 0.49715585893060293, "grad_norm": 1.0852743744371969, "learning_rate": 1.2197819649663934e-06, "loss": 0.0383, "step": 2185 }, { "epoch": 0.49738339021615474, "grad_norm": 0.8202048600920889, "learning_rate": 1.21975451186809e-06, "loss": 0.0242, "step": 2186 }, { "epoch": 0.4976109215017065, "grad_norm": 1.180630635637642, "learning_rate": 1.2197270466140936e-06, "loss": 0.0597, "step": 2187 }, { "epoch": 0.49783845278725825, "grad_norm": 0.5499273259791628, "learning_rate": 1.2196995692049656e-06, "loss": 0.0194, "step": 2188 }, { "epoch": 0.49806598407281, "grad_norm": 1.3699181548074078, "learning_rate": 1.2196720796412675e-06, "loss": 0.0729, "step": 2189 }, { "epoch": 0.49829351535836175, "grad_norm": 1.6813076102860163, "learning_rate": 1.219644577923561e-06, "loss": 0.0576, "step": 2190 }, { "epoch": 0.49852104664391356, "grad_norm": 1.3365991781516084, "learning_rate": 1.2196170640524087e-06, "loss": 0.0682, "step": 2191 }, { "epoch": 0.4987485779294653, "grad_norm": 1.5433585155887166, "learning_rate": 1.2195895380283725e-06, "loss": 0.0712, "step": 2192 }, { "epoch": 0.49897610921501706, "grad_norm": 0.7540413090381725, "learning_rate": 1.219561999852015e-06, "loss": 0.0358, "step": 2193 }, { "epoch": 0.4992036405005688, "grad_norm": 1.0443802296277052, "learning_rate": 1.2195344495238993e-06, "loss": 0.0443, "step": 2194 }, { "epoch": 0.49943117178612056, "grad_norm": 1.0620744241395403, "learning_rate": 1.2195068870445878e-06, "loss": 0.0317, "step": 2195 }, { "epoch": 0.49965870307167237, "grad_norm": 0.8699498301198716, "learning_rate": 1.2194793124146447e-06, "loss": 0.0291, "step": 2196 }, { "epoch": 0.4998862343572241, "grad_norm": 1.4610167793964797, "learning_rate": 1.2194517256346333e-06, "loss": 0.0701, "step": 2197 }, { "epoch": 0.5001137656427759, "grad_norm": 1.2434868855241086, "learning_rate": 1.2194241267051172e-06, "loss": 0.0454, "step": 2198 }, { "epoch": 0.5003412969283276, "grad_norm": 1.4551693858285264, "learning_rate": 1.2193965156266605e-06, "loss": 0.0547, "step": 2199 }, { "epoch": 0.5005688282138794, "grad_norm": 1.9356847413883649, "learning_rate": 1.2193688923998277e-06, "loss": 0.0644, "step": 2200 }, { "epoch": 0.5007963594994311, "grad_norm": 1.1503513652252215, "learning_rate": 1.219341257025183e-06, "loss": 0.0474, "step": 2201 }, { "epoch": 0.5010238907849829, "grad_norm": 1.5368491876593886, "learning_rate": 1.2193136095032918e-06, "loss": 0.0573, "step": 2202 }, { "epoch": 0.5012514220705347, "grad_norm": 1.7879629590621755, "learning_rate": 1.2192859498347189e-06, "loss": 0.0917, "step": 2203 }, { "epoch": 0.5014789533560865, "grad_norm": 1.5178773819159281, "learning_rate": 1.2192582780200293e-06, "loss": 0.0622, "step": 2204 }, { "epoch": 0.5017064846416383, "grad_norm": 1.3331971965539129, "learning_rate": 1.219230594059789e-06, "loss": 0.0649, "step": 2205 }, { "epoch": 0.50193401592719, "grad_norm": 2.165221254848427, "learning_rate": 1.2192028979545637e-06, "loss": 0.1342, "step": 2206 }, { "epoch": 0.5021615472127418, "grad_norm": 1.743824072315693, "learning_rate": 1.2191751897049192e-06, "loss": 0.0792, "step": 2207 }, { "epoch": 0.5023890784982935, "grad_norm": 1.3495321864370744, "learning_rate": 1.2191474693114223e-06, "loss": 0.0553, "step": 2208 }, { "epoch": 0.5026166097838453, "grad_norm": 1.2047719291625048, "learning_rate": 1.2191197367746389e-06, "loss": 0.0493, "step": 2209 }, { "epoch": 0.502844141069397, "grad_norm": 0.7568498406042676, "learning_rate": 1.2190919920951363e-06, "loss": 0.0425, "step": 2210 }, { "epoch": 0.5030716723549488, "grad_norm": 1.0905266622637186, "learning_rate": 1.2190642352734814e-06, "loss": 0.0392, "step": 2211 }, { "epoch": 0.5032992036405005, "grad_norm": 1.0772245840535826, "learning_rate": 1.2190364663102417e-06, "loss": 0.0602, "step": 2212 }, { "epoch": 0.5035267349260524, "grad_norm": 1.2112393942159567, "learning_rate": 1.2190086852059844e-06, "loss": 0.0561, "step": 2213 }, { "epoch": 0.5037542662116041, "grad_norm": 1.164381821277193, "learning_rate": 1.2189808919612773e-06, "loss": 0.0567, "step": 2214 }, { "epoch": 0.5039817974971559, "grad_norm": 1.980841011359919, "learning_rate": 1.2189530865766888e-06, "loss": 0.1002, "step": 2215 }, { "epoch": 0.5042093287827076, "grad_norm": 1.0195390074518678, "learning_rate": 1.218925269052787e-06, "loss": 0.0378, "step": 2216 }, { "epoch": 0.5044368600682594, "grad_norm": 1.58039989452312, "learning_rate": 1.2188974393901404e-06, "loss": 0.079, "step": 2217 }, { "epoch": 0.5046643913538111, "grad_norm": 1.7632794902075288, "learning_rate": 1.2188695975893177e-06, "loss": 0.0704, "step": 2218 }, { "epoch": 0.5048919226393629, "grad_norm": 1.8641645104616156, "learning_rate": 1.218841743650888e-06, "loss": 0.0767, "step": 2219 }, { "epoch": 0.5051194539249146, "grad_norm": 0.7736340615610193, "learning_rate": 1.2188138775754207e-06, "loss": 0.0255, "step": 2220 }, { "epoch": 0.5053469852104664, "grad_norm": 1.706014884062171, "learning_rate": 1.2187859993634854e-06, "loss": 0.0842, "step": 2221 }, { "epoch": 0.5055745164960183, "grad_norm": 1.0275306984300476, "learning_rate": 1.2187581090156518e-06, "loss": 0.0372, "step": 2222 }, { "epoch": 0.50580204778157, "grad_norm": 1.560530367319729, "learning_rate": 1.2187302065324896e-06, "loss": 0.0957, "step": 2223 }, { "epoch": 0.5060295790671218, "grad_norm": 1.1142447765551218, "learning_rate": 1.2187022919145695e-06, "loss": 0.0372, "step": 2224 }, { "epoch": 0.5062571103526735, "grad_norm": 2.0014178600777157, "learning_rate": 1.2186743651624617e-06, "loss": 0.1222, "step": 2225 }, { "epoch": 0.5064846416382253, "grad_norm": 0.6371915222683989, "learning_rate": 1.2186464262767372e-06, "loss": 0.0303, "step": 2226 }, { "epoch": 0.506712172923777, "grad_norm": 1.1677216592762003, "learning_rate": 1.2186184752579671e-06, "loss": 0.0375, "step": 2227 }, { "epoch": 0.5069397042093288, "grad_norm": 1.0363110917575087, "learning_rate": 1.2185905121067223e-06, "loss": 0.0465, "step": 2228 }, { "epoch": 0.5071672354948805, "grad_norm": 0.7880205913361013, "learning_rate": 1.218562536823575e-06, "loss": 0.0461, "step": 2229 }, { "epoch": 0.5073947667804323, "grad_norm": 1.5494942731094494, "learning_rate": 1.2185345494090959e-06, "loss": 0.0633, "step": 2230 }, { "epoch": 0.507622298065984, "grad_norm": 1.24395335626353, "learning_rate": 1.218506549863858e-06, "loss": 0.037, "step": 2231 }, { "epoch": 0.5078498293515359, "grad_norm": 0.8497919211723142, "learning_rate": 1.2184785381884332e-06, "loss": 0.0281, "step": 2232 }, { "epoch": 0.5080773606370876, "grad_norm": 1.6659795749856976, "learning_rate": 1.218450514383394e-06, "loss": 0.0684, "step": 2233 }, { "epoch": 0.5083048919226394, "grad_norm": 1.2993624404216275, "learning_rate": 1.2184224784493127e-06, "loss": 0.0737, "step": 2234 }, { "epoch": 0.5085324232081911, "grad_norm": 2.2363218045516287, "learning_rate": 1.218394430386763e-06, "loss": 0.1023, "step": 2235 }, { "epoch": 0.5087599544937429, "grad_norm": 1.9367152712705378, "learning_rate": 1.2183663701963181e-06, "loss": 0.0801, "step": 2236 }, { "epoch": 0.5089874857792946, "grad_norm": 1.683178965938102, "learning_rate": 1.218338297878551e-06, "loss": 0.0511, "step": 2237 }, { "epoch": 0.5092150170648464, "grad_norm": 1.2570305483870696, "learning_rate": 1.2183102134340361e-06, "loss": 0.0531, "step": 2238 }, { "epoch": 0.5094425483503981, "grad_norm": 1.3387251635869548, "learning_rate": 1.2182821168633466e-06, "loss": 0.0488, "step": 2239 }, { "epoch": 0.5096700796359499, "grad_norm": 0.9109698538663267, "learning_rate": 1.2182540081670574e-06, "loss": 0.0297, "step": 2240 }, { "epoch": 0.5098976109215017, "grad_norm": 2.0458637944350917, "learning_rate": 1.2182258873457428e-06, "loss": 0.0932, "step": 2241 }, { "epoch": 0.5101251422070535, "grad_norm": 1.546517047083001, "learning_rate": 1.2181977543999776e-06, "loss": 0.0831, "step": 2242 }, { "epoch": 0.5103526734926053, "grad_norm": 0.5799180817857912, "learning_rate": 1.2181696093303363e-06, "loss": 0.0246, "step": 2243 }, { "epoch": 0.510580204778157, "grad_norm": 1.5842605922722366, "learning_rate": 1.218141452137395e-06, "loss": 0.1109, "step": 2244 }, { "epoch": 0.5108077360637088, "grad_norm": 1.1518936456507234, "learning_rate": 1.2181132828217284e-06, "loss": 0.0392, "step": 2245 }, { "epoch": 0.5110352673492605, "grad_norm": 1.5601510600984896, "learning_rate": 1.2180851013839127e-06, "loss": 0.0516, "step": 2246 }, { "epoch": 0.5112627986348123, "grad_norm": 1.1933267652037376, "learning_rate": 1.2180569078245236e-06, "loss": 0.0768, "step": 2247 }, { "epoch": 0.511490329920364, "grad_norm": 1.0276357760825077, "learning_rate": 1.2180287021441372e-06, "loss": 0.0392, "step": 2248 }, { "epoch": 0.5117178612059158, "grad_norm": 1.4568173829581255, "learning_rate": 1.2180004843433305e-06, "loss": 0.0711, "step": 2249 }, { "epoch": 0.5119453924914675, "grad_norm": 1.6296361312434235, "learning_rate": 1.21797225442268e-06, "loss": 0.061, "step": 2250 }, { "epoch": 0.5121729237770194, "grad_norm": 1.4099321985545326, "learning_rate": 1.2179440123827625e-06, "loss": 0.0562, "step": 2251 }, { "epoch": 0.5124004550625711, "grad_norm": 1.0111117892815282, "learning_rate": 1.2179157582241554e-06, "loss": 0.0425, "step": 2252 }, { "epoch": 0.5126279863481229, "grad_norm": 1.100003009354464, "learning_rate": 1.2178874919474359e-06, "loss": 0.059, "step": 2253 }, { "epoch": 0.5128555176336747, "grad_norm": 1.325363901130602, "learning_rate": 1.217859213553182e-06, "loss": 0.0711, "step": 2254 }, { "epoch": 0.5130830489192264, "grad_norm": 0.846300456406513, "learning_rate": 1.2178309230419714e-06, "loss": 0.0271, "step": 2255 }, { "epoch": 0.5133105802047782, "grad_norm": 1.2410062157733406, "learning_rate": 1.2178026204143827e-06, "loss": 0.0556, "step": 2256 }, { "epoch": 0.5135381114903299, "grad_norm": 1.6420974990981174, "learning_rate": 1.217774305670994e-06, "loss": 0.0712, "step": 2257 }, { "epoch": 0.5137656427758817, "grad_norm": 1.2130253789178054, "learning_rate": 1.2177459788123841e-06, "loss": 0.045, "step": 2258 }, { "epoch": 0.5139931740614334, "grad_norm": 1.5302885687128625, "learning_rate": 1.217717639839132e-06, "loss": 0.0788, "step": 2259 }, { "epoch": 0.5142207053469852, "grad_norm": 0.9160947105126066, "learning_rate": 1.2176892887518166e-06, "loss": 0.0381, "step": 2260 }, { "epoch": 0.514448236632537, "grad_norm": 0.9460716271280594, "learning_rate": 1.2176609255510176e-06, "loss": 0.0453, "step": 2261 }, { "epoch": 0.5146757679180888, "grad_norm": 0.6452347430540285, "learning_rate": 1.217632550237315e-06, "loss": 0.0225, "step": 2262 }, { "epoch": 0.5149032992036405, "grad_norm": 1.1495647599088414, "learning_rate": 1.2176041628112884e-06, "loss": 0.0506, "step": 2263 }, { "epoch": 0.5151308304891923, "grad_norm": 1.4009996525001283, "learning_rate": 1.217575763273518e-06, "loss": 0.0667, "step": 2264 }, { "epoch": 0.515358361774744, "grad_norm": 0.9674758245027556, "learning_rate": 1.217547351624584e-06, "loss": 0.0383, "step": 2265 }, { "epoch": 0.5155858930602958, "grad_norm": 1.6049309598180266, "learning_rate": 1.2175189278650677e-06, "loss": 0.0685, "step": 2266 }, { "epoch": 0.5158134243458475, "grad_norm": 1.047951830927466, "learning_rate": 1.2174904919955493e-06, "loss": 0.031, "step": 2267 }, { "epoch": 0.5160409556313993, "grad_norm": 1.0060223411128129, "learning_rate": 1.2174620440166105e-06, "loss": 0.0442, "step": 2268 }, { "epoch": 0.516268486916951, "grad_norm": 1.2945214325099708, "learning_rate": 1.2174335839288326e-06, "loss": 0.0558, "step": 2269 }, { "epoch": 0.5164960182025028, "grad_norm": 1.2199251228288925, "learning_rate": 1.2174051117327972e-06, "loss": 0.0398, "step": 2270 }, { "epoch": 0.5167235494880547, "grad_norm": 1.8177803789021154, "learning_rate": 1.2173766274290864e-06, "loss": 0.0618, "step": 2271 }, { "epoch": 0.5169510807736064, "grad_norm": 2.1534503937526557, "learning_rate": 1.217348131018282e-06, "loss": 0.0946, "step": 2272 }, { "epoch": 0.5171786120591582, "grad_norm": 1.335439582615335, "learning_rate": 1.2173196225009666e-06, "loss": 0.0467, "step": 2273 }, { "epoch": 0.5174061433447099, "grad_norm": 3.39634994855923, "learning_rate": 1.217291101877723e-06, "loss": 0.1597, "step": 2274 }, { "epoch": 0.5176336746302617, "grad_norm": 13.151087600523141, "learning_rate": 1.2172625691491342e-06, "loss": 0.061, "step": 2275 }, { "epoch": 0.5178612059158134, "grad_norm": 0.9388770540310672, "learning_rate": 1.217234024315783e-06, "loss": 0.0469, "step": 2276 }, { "epoch": 0.5180887372013652, "grad_norm": 1.237623086215354, "learning_rate": 1.2172054673782527e-06, "loss": 0.0494, "step": 2277 }, { "epoch": 0.5183162684869169, "grad_norm": 1.3918576506801177, "learning_rate": 1.2171768983371276e-06, "loss": 0.0688, "step": 2278 }, { "epoch": 0.5185437997724687, "grad_norm": 1.2775225468090683, "learning_rate": 1.217148317192991e-06, "loss": 0.034, "step": 2279 }, { "epoch": 0.5187713310580204, "grad_norm": 1.2629250238471204, "learning_rate": 1.2171197239464274e-06, "loss": 0.0477, "step": 2280 }, { "epoch": 0.5189988623435723, "grad_norm": 0.9222994348691986, "learning_rate": 1.217091118598021e-06, "loss": 0.0518, "step": 2281 }, { "epoch": 0.519226393629124, "grad_norm": 1.1733465242259884, "learning_rate": 1.2170625011483565e-06, "loss": 0.0362, "step": 2282 }, { "epoch": 0.5194539249146758, "grad_norm": 1.0540409111201516, "learning_rate": 1.2170338715980187e-06, "loss": 0.0333, "step": 2283 }, { "epoch": 0.5196814562002275, "grad_norm": 1.0390405907126188, "learning_rate": 1.217005229947593e-06, "loss": 0.0307, "step": 2284 }, { "epoch": 0.5199089874857793, "grad_norm": 0.7575429344893744, "learning_rate": 1.2169765761976646e-06, "loss": 0.0416, "step": 2285 }, { "epoch": 0.520136518771331, "grad_norm": 1.3326371343495533, "learning_rate": 1.216947910348819e-06, "loss": 0.054, "step": 2286 }, { "epoch": 0.5203640500568828, "grad_norm": 1.386966435546428, "learning_rate": 1.2169192324016423e-06, "loss": 0.0552, "step": 2287 }, { "epoch": 0.5205915813424346, "grad_norm": 1.957649195968788, "learning_rate": 1.2168905423567205e-06, "loss": 0.0941, "step": 2288 }, { "epoch": 0.5208191126279863, "grad_norm": 2.448802297286155, "learning_rate": 1.21686184021464e-06, "loss": 0.1116, "step": 2289 }, { "epoch": 0.5210466439135382, "grad_norm": 1.1844985987836856, "learning_rate": 1.2168331259759875e-06, "loss": 0.0449, "step": 2290 }, { "epoch": 0.5212741751990899, "grad_norm": 1.8104263725287626, "learning_rate": 1.2168043996413497e-06, "loss": 0.0575, "step": 2291 }, { "epoch": 0.5215017064846417, "grad_norm": 1.1147839216238626, "learning_rate": 1.216775661211314e-06, "loss": 0.05, "step": 2292 }, { "epoch": 0.5217292377701934, "grad_norm": 1.543703515430326, "learning_rate": 1.2167469106864673e-06, "loss": 0.0813, "step": 2293 }, { "epoch": 0.5219567690557452, "grad_norm": 1.5787563749864177, "learning_rate": 1.2167181480673977e-06, "loss": 0.0843, "step": 2294 }, { "epoch": 0.5221843003412969, "grad_norm": 1.3389543319065593, "learning_rate": 1.2166893733546927e-06, "loss": 0.0854, "step": 2295 }, { "epoch": 0.5224118316268487, "grad_norm": 1.79843376105108, "learning_rate": 1.2166605865489406e-06, "loss": 0.111, "step": 2296 }, { "epoch": 0.5226393629124004, "grad_norm": 1.2843560350462406, "learning_rate": 1.2166317876507296e-06, "loss": 0.0617, "step": 2297 }, { "epoch": 0.5228668941979522, "grad_norm": 1.2028371451089743, "learning_rate": 1.2166029766606486e-06, "loss": 0.0721, "step": 2298 }, { "epoch": 0.5230944254835039, "grad_norm": 2.882673882589606, "learning_rate": 1.2165741535792861e-06, "loss": 0.1223, "step": 2299 }, { "epoch": 0.5233219567690558, "grad_norm": 1.3155493992738794, "learning_rate": 1.2165453184072312e-06, "loss": 0.0591, "step": 2300 }, { "epoch": 0.5235494880546075, "grad_norm": 1.6629360145623788, "learning_rate": 1.2165164711450735e-06, "loss": 0.0775, "step": 2301 }, { "epoch": 0.5237770193401593, "grad_norm": 1.3013507571586798, "learning_rate": 1.2164876117934024e-06, "loss": 0.0638, "step": 2302 }, { "epoch": 0.524004550625711, "grad_norm": 2.648828931200833, "learning_rate": 1.2164587403528078e-06, "loss": 0.0864, "step": 2303 }, { "epoch": 0.5242320819112628, "grad_norm": 1.4929308932558032, "learning_rate": 1.2164298568238797e-06, "loss": 0.0679, "step": 2304 }, { "epoch": 0.5244596131968146, "grad_norm": 0.8718556925311854, "learning_rate": 1.2164009612072085e-06, "loss": 0.0359, "step": 2305 }, { "epoch": 0.5246871444823663, "grad_norm": 0.9783095053030553, "learning_rate": 1.2163720535033845e-06, "loss": 0.0442, "step": 2306 }, { "epoch": 0.5249146757679181, "grad_norm": 1.5130471240864227, "learning_rate": 1.2163431337129993e-06, "loss": 0.0512, "step": 2307 }, { "epoch": 0.5251422070534698, "grad_norm": 1.5432509321722079, "learning_rate": 1.216314201836643e-06, "loss": 0.0704, "step": 2308 }, { "epoch": 0.5253697383390216, "grad_norm": 1.069477672250475, "learning_rate": 1.2162852578749076e-06, "loss": 0.0445, "step": 2309 }, { "epoch": 0.5255972696245734, "grad_norm": 1.831811758429303, "learning_rate": 1.2162563018283843e-06, "loss": 0.086, "step": 2310 }, { "epoch": 0.5258248009101252, "grad_norm": 0.9600707211513121, "learning_rate": 1.216227333697665e-06, "loss": 0.0387, "step": 2311 }, { "epoch": 0.5260523321956769, "grad_norm": 1.0398262718871747, "learning_rate": 1.216198353483342e-06, "loss": 0.0336, "step": 2312 }, { "epoch": 0.5262798634812287, "grad_norm": 0.7659268750272631, "learning_rate": 1.2161693611860072e-06, "loss": 0.0292, "step": 2313 }, { "epoch": 0.5265073947667804, "grad_norm": 1.003194143526719, "learning_rate": 1.2161403568062533e-06, "loss": 0.0446, "step": 2314 }, { "epoch": 0.5267349260523322, "grad_norm": 0.8822650643996454, "learning_rate": 1.2161113403446733e-06, "loss": 0.036, "step": 2315 }, { "epoch": 0.5269624573378839, "grad_norm": 1.2445369505027863, "learning_rate": 1.21608231180186e-06, "loss": 0.056, "step": 2316 }, { "epoch": 0.5271899886234357, "grad_norm": 1.9110079675084388, "learning_rate": 1.2160532711784066e-06, "loss": 0.085, "step": 2317 }, { "epoch": 0.5274175199089874, "grad_norm": 1.3201545262151622, "learning_rate": 1.2160242184749069e-06, "loss": 0.0596, "step": 2318 }, { "epoch": 0.5276450511945392, "grad_norm": 1.3929491912851009, "learning_rate": 1.2159951536919547e-06, "loss": 0.0575, "step": 2319 }, { "epoch": 0.5278725824800911, "grad_norm": 1.4551403932656901, "learning_rate": 1.2159660768301438e-06, "loss": 0.0492, "step": 2320 }, { "epoch": 0.5281001137656428, "grad_norm": 1.4130548917935428, "learning_rate": 1.2159369878900687e-06, "loss": 0.047, "step": 2321 }, { "epoch": 0.5283276450511946, "grad_norm": 1.4452980530222412, "learning_rate": 1.2159078868723238e-06, "loss": 0.0525, "step": 2322 }, { "epoch": 0.5285551763367463, "grad_norm": 1.6104531531662376, "learning_rate": 1.2158787737775037e-06, "loss": 0.0755, "step": 2323 }, { "epoch": 0.5287827076222981, "grad_norm": 0.8559219056147712, "learning_rate": 1.2158496486062039e-06, "loss": 0.0415, "step": 2324 }, { "epoch": 0.5290102389078498, "grad_norm": 1.789699350536378, "learning_rate": 1.215820511359019e-06, "loss": 0.081, "step": 2325 }, { "epoch": 0.5292377701934016, "grad_norm": 1.3025849756724603, "learning_rate": 1.215791362036545e-06, "loss": 0.0485, "step": 2326 }, { "epoch": 0.5294653014789533, "grad_norm": 1.8682155789508528, "learning_rate": 1.2157622006393777e-06, "loss": 0.0634, "step": 2327 }, { "epoch": 0.5296928327645051, "grad_norm": 1.176130717608952, "learning_rate": 1.2157330271681129e-06, "loss": 0.039, "step": 2328 }, { "epoch": 0.5299203640500569, "grad_norm": 1.5070209136147386, "learning_rate": 1.215703841623347e-06, "loss": 0.0515, "step": 2329 }, { "epoch": 0.5301478953356087, "grad_norm": 1.2847130041310237, "learning_rate": 1.2156746440056762e-06, "loss": 0.0515, "step": 2330 }, { "epoch": 0.5303754266211604, "grad_norm": 2.141191883255034, "learning_rate": 1.2156454343156976e-06, "loss": 0.0828, "step": 2331 }, { "epoch": 0.5306029579067122, "grad_norm": 0.9805091402589389, "learning_rate": 1.2156162125540078e-06, "loss": 0.0315, "step": 2332 }, { "epoch": 0.5308304891922639, "grad_norm": 0.7809668560849007, "learning_rate": 1.2155869787212046e-06, "loss": 0.0365, "step": 2333 }, { "epoch": 0.5310580204778157, "grad_norm": 1.0616410014514985, "learning_rate": 1.215557732817885e-06, "loss": 0.0754, "step": 2334 }, { "epoch": 0.5312855517633674, "grad_norm": 1.74228651169713, "learning_rate": 1.2155284748446469e-06, "loss": 0.0628, "step": 2335 }, { "epoch": 0.5315130830489192, "grad_norm": 1.8258581718716098, "learning_rate": 1.2154992048020882e-06, "loss": 0.0957, "step": 2336 }, { "epoch": 0.531740614334471, "grad_norm": 1.5209500494569268, "learning_rate": 1.2154699226908072e-06, "loss": 0.0719, "step": 2337 }, { "epoch": 0.5319681456200227, "grad_norm": 0.9184689674787639, "learning_rate": 1.2154406285114025e-06, "loss": 0.0512, "step": 2338 }, { "epoch": 0.5321956769055746, "grad_norm": 0.92211428137911, "learning_rate": 1.2154113222644727e-06, "loss": 0.048, "step": 2339 }, { "epoch": 0.5324232081911263, "grad_norm": 1.2486653702619341, "learning_rate": 1.2153820039506167e-06, "loss": 0.036, "step": 2340 }, { "epoch": 0.5326507394766781, "grad_norm": 1.4084664381645906, "learning_rate": 1.2153526735704337e-06, "loss": 0.0641, "step": 2341 }, { "epoch": 0.5328782707622298, "grad_norm": 1.006089995573004, "learning_rate": 1.2153233311245234e-06, "loss": 0.0471, "step": 2342 }, { "epoch": 0.5331058020477816, "grad_norm": 1.2533267107040431, "learning_rate": 1.2152939766134852e-06, "loss": 0.044, "step": 2343 }, { "epoch": 0.5333333333333333, "grad_norm": 1.7169848907323204, "learning_rate": 1.2152646100379193e-06, "loss": 0.0979, "step": 2344 }, { "epoch": 0.5335608646188851, "grad_norm": 1.1457088139957912, "learning_rate": 1.2152352313984257e-06, "loss": 0.0539, "step": 2345 }, { "epoch": 0.5337883959044368, "grad_norm": 0.9527962003003786, "learning_rate": 1.2152058406956049e-06, "loss": 0.0389, "step": 2346 }, { "epoch": 0.5340159271899886, "grad_norm": 1.5909759190824926, "learning_rate": 1.2151764379300578e-06, "loss": 0.058, "step": 2347 }, { "epoch": 0.5342434584755403, "grad_norm": 1.065418317407388, "learning_rate": 1.2151470231023851e-06, "loss": 0.0304, "step": 2348 }, { "epoch": 0.5344709897610922, "grad_norm": 1.007958090684921, "learning_rate": 1.2151175962131881e-06, "loss": 0.049, "step": 2349 }, { "epoch": 0.534698521046644, "grad_norm": 1.7395647605444997, "learning_rate": 1.215088157263068e-06, "loss": 0.06, "step": 2350 }, { "epoch": 0.5349260523321957, "grad_norm": 1.0555560500880805, "learning_rate": 1.2150587062526267e-06, "loss": 0.0466, "step": 2351 }, { "epoch": 0.5351535836177475, "grad_norm": 1.4330290353052704, "learning_rate": 1.2150292431824663e-06, "loss": 0.1029, "step": 2352 }, { "epoch": 0.5353811149032992, "grad_norm": 1.0120120325632629, "learning_rate": 1.2149997680531886e-06, "loss": 0.0372, "step": 2353 }, { "epoch": 0.535608646188851, "grad_norm": 1.6445264766335106, "learning_rate": 1.214970280865396e-06, "loss": 0.0601, "step": 2354 }, { "epoch": 0.5358361774744027, "grad_norm": 1.479298949629959, "learning_rate": 1.2149407816196917e-06, "loss": 0.0713, "step": 2355 }, { "epoch": 0.5360637087599545, "grad_norm": 1.0674415697842976, "learning_rate": 1.214911270316678e-06, "loss": 0.0529, "step": 2356 }, { "epoch": 0.5362912400455062, "grad_norm": 1.1601322180484739, "learning_rate": 1.2148817469569584e-06, "loss": 0.0762, "step": 2357 }, { "epoch": 0.5365187713310581, "grad_norm": 0.6289268827845146, "learning_rate": 1.214852211541136e-06, "loss": 0.0271, "step": 2358 }, { "epoch": 0.5367463026166098, "grad_norm": 1.127391002234741, "learning_rate": 1.2148226640698148e-06, "loss": 0.0374, "step": 2359 }, { "epoch": 0.5369738339021616, "grad_norm": 1.4660398732133397, "learning_rate": 1.2147931045435988e-06, "loss": 0.0851, "step": 2360 }, { "epoch": 0.5372013651877133, "grad_norm": 1.4378197818055753, "learning_rate": 1.2147635329630916e-06, "loss": 0.0887, "step": 2361 }, { "epoch": 0.5374288964732651, "grad_norm": 1.0407552632451906, "learning_rate": 1.214733949328898e-06, "loss": 0.0274, "step": 2362 }, { "epoch": 0.5376564277588168, "grad_norm": 1.2421238364546752, "learning_rate": 1.2147043536416226e-06, "loss": 0.0547, "step": 2363 }, { "epoch": 0.5378839590443686, "grad_norm": 1.9182602099931712, "learning_rate": 1.21467474590187e-06, "loss": 0.1061, "step": 2364 }, { "epoch": 0.5381114903299203, "grad_norm": 1.7756462089099831, "learning_rate": 1.2146451261102458e-06, "loss": 0.0633, "step": 2365 }, { "epoch": 0.5383390216154721, "grad_norm": 0.9945218699288574, "learning_rate": 1.2146154942673548e-06, "loss": 0.0458, "step": 2366 }, { "epoch": 0.5385665529010238, "grad_norm": 0.8213845393107763, "learning_rate": 1.2145858503738032e-06, "loss": 0.0425, "step": 2367 }, { "epoch": 0.5387940841865757, "grad_norm": 1.8457471393460891, "learning_rate": 1.2145561944301963e-06, "loss": 0.0922, "step": 2368 }, { "epoch": 0.5390216154721275, "grad_norm": 1.264742540597221, "learning_rate": 1.2145265264371406e-06, "loss": 0.0668, "step": 2369 }, { "epoch": 0.5392491467576792, "grad_norm": 0.8301569695908675, "learning_rate": 1.2144968463952425e-06, "loss": 0.0312, "step": 2370 }, { "epoch": 0.539476678043231, "grad_norm": 1.4005462540595164, "learning_rate": 1.2144671543051085e-06, "loss": 0.0385, "step": 2371 }, { "epoch": 0.5397042093287827, "grad_norm": 1.0427937857498775, "learning_rate": 1.2144374501673454e-06, "loss": 0.0405, "step": 2372 }, { "epoch": 0.5399317406143345, "grad_norm": 2.509030531929466, "learning_rate": 1.2144077339825603e-06, "loss": 0.129, "step": 2373 }, { "epoch": 0.5401592718998862, "grad_norm": 1.4490653559115954, "learning_rate": 1.2143780057513605e-06, "loss": 0.076, "step": 2374 }, { "epoch": 0.540386803185438, "grad_norm": 1.1167399295770026, "learning_rate": 1.2143482654743535e-06, "loss": 0.0433, "step": 2375 }, { "epoch": 0.5406143344709897, "grad_norm": 1.1732119937209518, "learning_rate": 1.2143185131521475e-06, "loss": 0.0639, "step": 2376 }, { "epoch": 0.5408418657565415, "grad_norm": 1.083852720432599, "learning_rate": 1.2142887487853503e-06, "loss": 0.0534, "step": 2377 }, { "epoch": 0.5410693970420933, "grad_norm": 1.2045837567267914, "learning_rate": 1.2142589723745705e-06, "loss": 0.047, "step": 2378 }, { "epoch": 0.5412969283276451, "grad_norm": 1.149610953840798, "learning_rate": 1.2142291839204163e-06, "loss": 0.0451, "step": 2379 }, { "epoch": 0.5415244596131968, "grad_norm": 1.17145770612206, "learning_rate": 1.2141993834234967e-06, "loss": 0.0594, "step": 2380 }, { "epoch": 0.5417519908987486, "grad_norm": 0.8279362230767302, "learning_rate": 1.2141695708844209e-06, "loss": 0.0331, "step": 2381 }, { "epoch": 0.5419795221843003, "grad_norm": 1.1783984776248138, "learning_rate": 1.214139746303798e-06, "loss": 0.0479, "step": 2382 }, { "epoch": 0.5422070534698521, "grad_norm": 1.6263426873577607, "learning_rate": 1.2141099096822376e-06, "loss": 0.0788, "step": 2383 }, { "epoch": 0.5424345847554038, "grad_norm": 1.339603186745851, "learning_rate": 1.2140800610203497e-06, "loss": 0.0819, "step": 2384 }, { "epoch": 0.5426621160409556, "grad_norm": 1.1850152950049393, "learning_rate": 1.214050200318744e-06, "loss": 0.0515, "step": 2385 }, { "epoch": 0.5428896473265074, "grad_norm": 1.4608498983467832, "learning_rate": 1.214020327578031e-06, "loss": 0.0528, "step": 2386 }, { "epoch": 0.5431171786120591, "grad_norm": 1.1158219646844105, "learning_rate": 1.2139904427988213e-06, "loss": 0.0328, "step": 2387 }, { "epoch": 0.543344709897611, "grad_norm": 0.8262393644685219, "learning_rate": 1.2139605459817259e-06, "loss": 0.0415, "step": 2388 }, { "epoch": 0.5435722411831627, "grad_norm": 1.096976415569224, "learning_rate": 1.2139306371273552e-06, "loss": 0.0619, "step": 2389 }, { "epoch": 0.5437997724687145, "grad_norm": 0.8704738503788053, "learning_rate": 1.213900716236321e-06, "loss": 0.0522, "step": 2390 }, { "epoch": 0.5440273037542662, "grad_norm": 0.9955866102135448, "learning_rate": 1.2138707833092345e-06, "loss": 0.0606, "step": 2391 }, { "epoch": 0.544254835039818, "grad_norm": 1.488547151459133, "learning_rate": 1.213840838346708e-06, "loss": 0.0871, "step": 2392 }, { "epoch": 0.5444823663253697, "grad_norm": 1.3949089434078865, "learning_rate": 1.213810881349353e-06, "loss": 0.0833, "step": 2393 }, { "epoch": 0.5447098976109215, "grad_norm": 1.5448754824040467, "learning_rate": 1.213780912317782e-06, "loss": 0.0567, "step": 2394 }, { "epoch": 0.5449374288964732, "grad_norm": 1.2930562589697765, "learning_rate": 1.2137509312526074e-06, "loss": 0.0486, "step": 2395 }, { "epoch": 0.545164960182025, "grad_norm": 0.8149733951933631, "learning_rate": 1.213720938154442e-06, "loss": 0.0447, "step": 2396 }, { "epoch": 0.5453924914675768, "grad_norm": 1.5207149596968195, "learning_rate": 1.2136909330238988e-06, "loss": 0.0712, "step": 2397 }, { "epoch": 0.5456200227531286, "grad_norm": 1.6094750936001516, "learning_rate": 1.213660915861591e-06, "loss": 0.0717, "step": 2398 }, { "epoch": 0.5458475540386803, "grad_norm": 1.0354263989549572, "learning_rate": 1.2136308866681323e-06, "loss": 0.0543, "step": 2399 }, { "epoch": 0.5460750853242321, "grad_norm": 1.312237280962152, "learning_rate": 1.2136008454441363e-06, "loss": 0.0651, "step": 2400 }, { "epoch": 0.5463026166097839, "grad_norm": 0.9265942352701293, "learning_rate": 1.213570792190217e-06, "loss": 0.0399, "step": 2401 }, { "epoch": 0.5465301478953356, "grad_norm": 1.3165932395587605, "learning_rate": 1.2135407269069885e-06, "loss": 0.0711, "step": 2402 }, { "epoch": 0.5467576791808874, "grad_norm": 1.3250744955676776, "learning_rate": 1.2135106495950655e-06, "loss": 0.055, "step": 2403 }, { "epoch": 0.5469852104664391, "grad_norm": 1.1269361373493476, "learning_rate": 1.2134805602550625e-06, "loss": 0.0391, "step": 2404 }, { "epoch": 0.5472127417519909, "grad_norm": 1.6294590878339321, "learning_rate": 1.2134504588875948e-06, "loss": 0.0752, "step": 2405 }, { "epoch": 0.5474402730375426, "grad_norm": 1.942161162101733, "learning_rate": 1.2134203454932772e-06, "loss": 0.111, "step": 2406 }, { "epoch": 0.5476678043230945, "grad_norm": 0.9091105494763078, "learning_rate": 1.2133902200727256e-06, "loss": 0.0366, "step": 2407 }, { "epoch": 0.5478953356086462, "grad_norm": 0.8734404453486602, "learning_rate": 1.2133600826265555e-06, "loss": 0.0339, "step": 2408 }, { "epoch": 0.548122866894198, "grad_norm": 1.4745463801324197, "learning_rate": 1.2133299331553826e-06, "loss": 0.0446, "step": 2409 }, { "epoch": 0.5483503981797497, "grad_norm": 0.9275974290923358, "learning_rate": 1.2132997716598236e-06, "loss": 0.0395, "step": 2410 }, { "epoch": 0.5485779294653015, "grad_norm": 0.8344512585289461, "learning_rate": 1.2132695981404943e-06, "loss": 0.0315, "step": 2411 }, { "epoch": 0.5488054607508532, "grad_norm": 1.6903517654405738, "learning_rate": 1.2132394125980122e-06, "loss": 0.0618, "step": 2412 }, { "epoch": 0.549032992036405, "grad_norm": 1.5596571212256216, "learning_rate": 1.2132092150329936e-06, "loss": 0.0623, "step": 2413 }, { "epoch": 0.5492605233219567, "grad_norm": 1.5970478431414654, "learning_rate": 1.213179005446056e-06, "loss": 0.0676, "step": 2414 }, { "epoch": 0.5494880546075085, "grad_norm": 1.1358043470989139, "learning_rate": 1.2131487838378167e-06, "loss": 0.0444, "step": 2415 }, { "epoch": 0.5497155858930602, "grad_norm": 0.7736657460795575, "learning_rate": 1.2131185502088932e-06, "loss": 0.0234, "step": 2416 }, { "epoch": 0.5499431171786121, "grad_norm": 1.0171839684581578, "learning_rate": 1.213088304559904e-06, "loss": 0.0354, "step": 2417 }, { "epoch": 0.5501706484641639, "grad_norm": 1.5638848346839926, "learning_rate": 1.2130580468914665e-06, "loss": 0.0907, "step": 2418 }, { "epoch": 0.5503981797497156, "grad_norm": 1.195018951013028, "learning_rate": 1.2130277772041999e-06, "loss": 0.0492, "step": 2419 }, { "epoch": 0.5506257110352674, "grad_norm": 1.0599309139929776, "learning_rate": 1.2129974954987222e-06, "loss": 0.0468, "step": 2420 }, { "epoch": 0.5508532423208191, "grad_norm": 0.9174363109363153, "learning_rate": 1.2129672017756524e-06, "loss": 0.0396, "step": 2421 }, { "epoch": 0.5510807736063709, "grad_norm": 0.996062027313234, "learning_rate": 1.2129368960356102e-06, "loss": 0.0461, "step": 2422 }, { "epoch": 0.5513083048919226, "grad_norm": 1.1748800644809907, "learning_rate": 1.2129065782792142e-06, "loss": 0.0468, "step": 2423 }, { "epoch": 0.5515358361774744, "grad_norm": 1.0572109261778357, "learning_rate": 1.2128762485070848e-06, "loss": 0.0337, "step": 2424 }, { "epoch": 0.5517633674630261, "grad_norm": 0.7078117825584842, "learning_rate": 1.2128459067198414e-06, "loss": 0.0279, "step": 2425 }, { "epoch": 0.5519908987485779, "grad_norm": 1.2504565872577411, "learning_rate": 1.2128155529181042e-06, "loss": 0.0432, "step": 2426 }, { "epoch": 0.5522184300341297, "grad_norm": 2.0545311286439656, "learning_rate": 1.2127851871024937e-06, "loss": 0.13, "step": 2427 }, { "epoch": 0.5524459613196815, "grad_norm": 1.6325039766646547, "learning_rate": 1.2127548092736305e-06, "loss": 0.0712, "step": 2428 }, { "epoch": 0.5526734926052332, "grad_norm": 1.4302972588358862, "learning_rate": 1.2127244194321353e-06, "loss": 0.0928, "step": 2429 }, { "epoch": 0.552901023890785, "grad_norm": 0.9939753923200273, "learning_rate": 1.2126940175786294e-06, "loss": 0.0537, "step": 2430 }, { "epoch": 0.5531285551763367, "grad_norm": 1.1837217608085497, "learning_rate": 1.212663603713734e-06, "loss": 0.0411, "step": 2431 }, { "epoch": 0.5533560864618885, "grad_norm": 1.0395121186554928, "learning_rate": 1.212633177838071e-06, "loss": 0.0499, "step": 2432 }, { "epoch": 0.5535836177474402, "grad_norm": 0.9514659467888016, "learning_rate": 1.2126027399522617e-06, "loss": 0.0453, "step": 2433 }, { "epoch": 0.553811149032992, "grad_norm": 1.6880603772633955, "learning_rate": 1.2125722900569288e-06, "loss": 0.0745, "step": 2434 }, { "epoch": 0.5540386803185438, "grad_norm": 1.8855045585505437, "learning_rate": 1.2125418281526944e-06, "loss": 0.1082, "step": 2435 }, { "epoch": 0.5542662116040956, "grad_norm": 1.2963290000539638, "learning_rate": 1.212511354240181e-06, "loss": 0.0571, "step": 2436 }, { "epoch": 0.5544937428896474, "grad_norm": 1.075287013825043, "learning_rate": 1.2124808683200113e-06, "loss": 0.0522, "step": 2437 }, { "epoch": 0.5547212741751991, "grad_norm": 1.5923929041091236, "learning_rate": 1.2124503703928088e-06, "loss": 0.0634, "step": 2438 }, { "epoch": 0.5549488054607509, "grad_norm": 1.832683570099282, "learning_rate": 1.2124198604591965e-06, "loss": 0.0659, "step": 2439 }, { "epoch": 0.5551763367463026, "grad_norm": 1.4530576893525877, "learning_rate": 1.2123893385197982e-06, "loss": 0.0508, "step": 2440 }, { "epoch": 0.5554038680318544, "grad_norm": 2.2165292342455967, "learning_rate": 1.2123588045752373e-06, "loss": 0.0788, "step": 2441 }, { "epoch": 0.5556313993174061, "grad_norm": 0.7214271689112302, "learning_rate": 1.2123282586261384e-06, "loss": 0.0373, "step": 2442 }, { "epoch": 0.5558589306029579, "grad_norm": 1.733249440041101, "learning_rate": 1.2122977006731256e-06, "loss": 0.0849, "step": 2443 }, { "epoch": 0.5560864618885096, "grad_norm": 0.9877695265064453, "learning_rate": 1.2122671307168232e-06, "loss": 0.0293, "step": 2444 }, { "epoch": 0.5563139931740614, "grad_norm": 2.369119537407654, "learning_rate": 1.212236548757856e-06, "loss": 0.1193, "step": 2445 }, { "epoch": 0.5565415244596132, "grad_norm": 0.9124531328873139, "learning_rate": 1.2122059547968496e-06, "loss": 0.0304, "step": 2446 }, { "epoch": 0.556769055745165, "grad_norm": 2.08793896339006, "learning_rate": 1.2121753488344286e-06, "loss": 0.102, "step": 2447 }, { "epoch": 0.5569965870307167, "grad_norm": 1.128455134895707, "learning_rate": 1.212144730871219e-06, "loss": 0.0446, "step": 2448 }, { "epoch": 0.5572241183162685, "grad_norm": 1.6401919927448718, "learning_rate": 1.2121141009078462e-06, "loss": 0.0557, "step": 2449 }, { "epoch": 0.5574516496018203, "grad_norm": 1.058585660351393, "learning_rate": 1.2120834589449368e-06, "loss": 0.0397, "step": 2450 }, { "epoch": 0.557679180887372, "grad_norm": 1.850497858074919, "learning_rate": 1.2120528049831165e-06, "loss": 0.0898, "step": 2451 }, { "epoch": 0.5579067121729238, "grad_norm": 2.1238500476076614, "learning_rate": 1.212022139023012e-06, "loss": 0.0775, "step": 2452 }, { "epoch": 0.5581342434584755, "grad_norm": 1.1451793439610103, "learning_rate": 1.21199146106525e-06, "loss": 0.0395, "step": 2453 }, { "epoch": 0.5583617747440273, "grad_norm": 0.8655064333461899, "learning_rate": 1.2119607711104574e-06, "loss": 0.0414, "step": 2454 }, { "epoch": 0.558589306029579, "grad_norm": 0.7739545180233858, "learning_rate": 1.211930069159262e-06, "loss": 0.0376, "step": 2455 }, { "epoch": 0.5588168373151309, "grad_norm": 1.1255668925149045, "learning_rate": 1.2118993552122907e-06, "loss": 0.0615, "step": 2456 }, { "epoch": 0.5590443686006826, "grad_norm": 0.8837865527117891, "learning_rate": 1.2118686292701715e-06, "loss": 0.033, "step": 2457 }, { "epoch": 0.5592718998862344, "grad_norm": 1.4857021615037345, "learning_rate": 1.211837891333532e-06, "loss": 0.0515, "step": 2458 }, { "epoch": 0.5594994311717861, "grad_norm": 1.687364075587055, "learning_rate": 1.211807141403001e-06, "loss": 0.0788, "step": 2459 }, { "epoch": 0.5597269624573379, "grad_norm": 1.1101722611046627, "learning_rate": 1.211776379479207e-06, "loss": 0.0361, "step": 2460 }, { "epoch": 0.5599544937428896, "grad_norm": 1.4952697824972245, "learning_rate": 1.211745605562778e-06, "loss": 0.0557, "step": 2461 }, { "epoch": 0.5601820250284414, "grad_norm": 0.8923044749255172, "learning_rate": 1.2117148196543436e-06, "loss": 0.0577, "step": 2462 }, { "epoch": 0.5604095563139931, "grad_norm": 2.562794461200439, "learning_rate": 1.2116840217545329e-06, "loss": 0.1001, "step": 2463 }, { "epoch": 0.5606370875995449, "grad_norm": 0.9586201769696352, "learning_rate": 1.211653211863975e-06, "loss": 0.0502, "step": 2464 }, { "epoch": 0.5608646188850968, "grad_norm": 1.4257435635791937, "learning_rate": 1.2116223899833e-06, "loss": 0.0544, "step": 2465 }, { "epoch": 0.5610921501706485, "grad_norm": 1.1082780136920223, "learning_rate": 1.2115915561131376e-06, "loss": 0.0368, "step": 2466 }, { "epoch": 0.5613196814562003, "grad_norm": 1.1191301112997847, "learning_rate": 1.211560710254118e-06, "loss": 0.0346, "step": 2467 }, { "epoch": 0.561547212741752, "grad_norm": 1.5912996699076145, "learning_rate": 1.211529852406872e-06, "loss": 0.0596, "step": 2468 }, { "epoch": 0.5617747440273038, "grad_norm": 1.3053443148766963, "learning_rate": 1.2114989825720298e-06, "loss": 0.0617, "step": 2469 }, { "epoch": 0.5620022753128555, "grad_norm": 1.935737911886316, "learning_rate": 1.2114681007502227e-06, "loss": 0.0713, "step": 2470 }, { "epoch": 0.5622298065984073, "grad_norm": 1.6784690811076477, "learning_rate": 1.2114372069420815e-06, "loss": 0.0898, "step": 2471 }, { "epoch": 0.562457337883959, "grad_norm": 1.6677773674255931, "learning_rate": 1.2114063011482378e-06, "loss": 0.0515, "step": 2472 }, { "epoch": 0.5626848691695108, "grad_norm": 0.7751790351928534, "learning_rate": 1.2113753833693234e-06, "loss": 0.0276, "step": 2473 }, { "epoch": 0.5629124004550625, "grad_norm": 1.5172528416500193, "learning_rate": 1.2113444536059699e-06, "loss": 0.078, "step": 2474 }, { "epoch": 0.5631399317406144, "grad_norm": 1.0306067244802375, "learning_rate": 1.2113135118588096e-06, "loss": 0.0601, "step": 2475 }, { "epoch": 0.5633674630261661, "grad_norm": 0.9276551177060361, "learning_rate": 1.2112825581284752e-06, "loss": 0.055, "step": 2476 }, { "epoch": 0.5635949943117179, "grad_norm": 1.2647808751954976, "learning_rate": 1.2112515924155987e-06, "loss": 0.048, "step": 2477 }, { "epoch": 0.5638225255972696, "grad_norm": 1.1803809621410315, "learning_rate": 1.2112206147208134e-06, "loss": 0.0513, "step": 2478 }, { "epoch": 0.5640500568828214, "grad_norm": 1.8816100409430934, "learning_rate": 1.2111896250447525e-06, "loss": 0.0972, "step": 2479 }, { "epoch": 0.5642775881683731, "grad_norm": 1.2512746487710684, "learning_rate": 1.211158623388049e-06, "loss": 0.0416, "step": 2480 }, { "epoch": 0.5645051194539249, "grad_norm": 1.1080723770280754, "learning_rate": 1.2111276097513369e-06, "loss": 0.044, "step": 2481 }, { "epoch": 0.5647326507394766, "grad_norm": 1.060740374175569, "learning_rate": 1.2110965841352498e-06, "loss": 0.0633, "step": 2482 }, { "epoch": 0.5649601820250284, "grad_norm": 1.0075017492283267, "learning_rate": 1.211065546540422e-06, "loss": 0.0341, "step": 2483 }, { "epoch": 0.5651877133105802, "grad_norm": 0.7064865281725817, "learning_rate": 1.2110344969674877e-06, "loss": 0.0286, "step": 2484 }, { "epoch": 0.565415244596132, "grad_norm": 1.1946068302141277, "learning_rate": 1.2110034354170816e-06, "loss": 0.0659, "step": 2485 }, { "epoch": 0.5656427758816838, "grad_norm": 0.9370117142292799, "learning_rate": 1.2109723618898383e-06, "loss": 0.0288, "step": 2486 }, { "epoch": 0.5658703071672355, "grad_norm": 2.2018132198270046, "learning_rate": 1.2109412763863933e-06, "loss": 0.1172, "step": 2487 }, { "epoch": 0.5660978384527873, "grad_norm": 1.1995855232529598, "learning_rate": 1.2109101789073815e-06, "loss": 0.0511, "step": 2488 }, { "epoch": 0.566325369738339, "grad_norm": 1.1031248857035114, "learning_rate": 1.2108790694534389e-06, "loss": 0.064, "step": 2489 }, { "epoch": 0.5665529010238908, "grad_norm": 1.738059897032761, "learning_rate": 1.2108479480252011e-06, "loss": 0.0756, "step": 2490 }, { "epoch": 0.5667804323094425, "grad_norm": 1.8262819909910217, "learning_rate": 1.210816814623304e-06, "loss": 0.1204, "step": 2491 }, { "epoch": 0.5670079635949943, "grad_norm": 0.9990130696869628, "learning_rate": 1.2107856692483843e-06, "loss": 0.0357, "step": 2492 }, { "epoch": 0.567235494880546, "grad_norm": 1.3613754270975984, "learning_rate": 1.2107545119010783e-06, "loss": 0.0638, "step": 2493 }, { "epoch": 0.5674630261660978, "grad_norm": 1.2474975066272391, "learning_rate": 1.2107233425820229e-06, "loss": 0.0599, "step": 2494 }, { "epoch": 0.5676905574516496, "grad_norm": 1.0265382923896895, "learning_rate": 1.2106921612918549e-06, "loss": 0.0385, "step": 2495 }, { "epoch": 0.5679180887372014, "grad_norm": 1.9111381375185976, "learning_rate": 1.2106609680312117e-06, "loss": 0.0686, "step": 2496 }, { "epoch": 0.5681456200227532, "grad_norm": 1.2560597740440176, "learning_rate": 1.210629762800731e-06, "loss": 0.0642, "step": 2497 }, { "epoch": 0.5683731513083049, "grad_norm": 1.0400297969186523, "learning_rate": 1.2105985456010506e-06, "loss": 0.0419, "step": 2498 }, { "epoch": 0.5686006825938567, "grad_norm": 1.5436476934145886, "learning_rate": 1.2105673164328081e-06, "loss": 0.0771, "step": 2499 }, { "epoch": 0.5688282138794084, "grad_norm": 1.2147723738279594, "learning_rate": 1.2105360752966424e-06, "loss": 0.044, "step": 2500 }, { "epoch": 0.5690557451649602, "grad_norm": 1.2940548926007058, "learning_rate": 1.2105048221931915e-06, "loss": 0.0556, "step": 2501 }, { "epoch": 0.5692832764505119, "grad_norm": 1.1343959115342286, "learning_rate": 1.2104735571230944e-06, "loss": 0.0325, "step": 2502 }, { "epoch": 0.5695108077360637, "grad_norm": 0.7696868285921366, "learning_rate": 1.21044228008699e-06, "loss": 0.0321, "step": 2503 }, { "epoch": 0.5697383390216155, "grad_norm": 1.277277873025092, "learning_rate": 1.2104109910855176e-06, "loss": 0.0597, "step": 2504 }, { "epoch": 0.5699658703071673, "grad_norm": 1.433825874950651, "learning_rate": 1.2103796901193166e-06, "loss": 0.0763, "step": 2505 }, { "epoch": 0.570193401592719, "grad_norm": 0.8577837659660162, "learning_rate": 1.210348377189027e-06, "loss": 0.0284, "step": 2506 }, { "epoch": 0.5704209328782708, "grad_norm": 1.1944391480100813, "learning_rate": 1.2103170522952885e-06, "loss": 0.0562, "step": 2507 }, { "epoch": 0.5706484641638225, "grad_norm": 1.283294042740657, "learning_rate": 1.2102857154387413e-06, "loss": 0.0468, "step": 2508 }, { "epoch": 0.5708759954493743, "grad_norm": 1.7437964279202711, "learning_rate": 1.210254366620026e-06, "loss": 0.097, "step": 2509 }, { "epoch": 0.571103526734926, "grad_norm": 1.224329122924591, "learning_rate": 1.2102230058397832e-06, "loss": 0.0595, "step": 2510 }, { "epoch": 0.5713310580204778, "grad_norm": 1.3007053728417897, "learning_rate": 1.210191633098654e-06, "loss": 0.0685, "step": 2511 }, { "epoch": 0.5715585893060295, "grad_norm": 1.1429406193383025, "learning_rate": 1.2101602483972797e-06, "loss": 0.048, "step": 2512 }, { "epoch": 0.5717861205915813, "grad_norm": 0.7595899866293734, "learning_rate": 1.2101288517363016e-06, "loss": 0.0194, "step": 2513 }, { "epoch": 0.5720136518771332, "grad_norm": 1.025256317864455, "learning_rate": 1.2100974431163614e-06, "loss": 0.0423, "step": 2514 }, { "epoch": 0.5722411831626849, "grad_norm": 1.053909591806734, "learning_rate": 1.2100660225381008e-06, "loss": 0.0284, "step": 2515 }, { "epoch": 0.5724687144482367, "grad_norm": 1.384456570911114, "learning_rate": 1.2100345900021624e-06, "loss": 0.0659, "step": 2516 }, { "epoch": 0.5726962457337884, "grad_norm": 1.3325135654239626, "learning_rate": 1.2100031455091883e-06, "loss": 0.0675, "step": 2517 }, { "epoch": 0.5729237770193402, "grad_norm": 1.179139361018548, "learning_rate": 1.2099716890598212e-06, "loss": 0.0475, "step": 2518 }, { "epoch": 0.5731513083048919, "grad_norm": 1.1347488592526416, "learning_rate": 1.2099402206547042e-06, "loss": 0.0563, "step": 2519 }, { "epoch": 0.5733788395904437, "grad_norm": 1.2845284052496178, "learning_rate": 1.2099087402944805e-06, "loss": 0.053, "step": 2520 }, { "epoch": 0.5736063708759954, "grad_norm": 0.9939680061445038, "learning_rate": 1.2098772479797933e-06, "loss": 0.0388, "step": 2521 }, { "epoch": 0.5738339021615472, "grad_norm": 1.2948668621318795, "learning_rate": 1.2098457437112862e-06, "loss": 0.0451, "step": 2522 }, { "epoch": 0.5740614334470989, "grad_norm": 1.6486523140737381, "learning_rate": 1.2098142274896033e-06, "loss": 0.0722, "step": 2523 }, { "epoch": 0.5742889647326508, "grad_norm": 1.1690776697359815, "learning_rate": 1.2097826993153886e-06, "loss": 0.0504, "step": 2524 }, { "epoch": 0.5745164960182025, "grad_norm": 0.8812949923396778, "learning_rate": 1.2097511591892863e-06, "loss": 0.0232, "step": 2525 }, { "epoch": 0.5747440273037543, "grad_norm": 1.6847623228179178, "learning_rate": 1.2097196071119415e-06, "loss": 0.0643, "step": 2526 }, { "epoch": 0.574971558589306, "grad_norm": 1.5124757371829332, "learning_rate": 1.2096880430839985e-06, "loss": 0.0859, "step": 2527 }, { "epoch": 0.5751990898748578, "grad_norm": 1.6240299660391282, "learning_rate": 1.2096564671061031e-06, "loss": 0.0637, "step": 2528 }, { "epoch": 0.5754266211604095, "grad_norm": 1.5970094647815496, "learning_rate": 1.2096248791789e-06, "loss": 0.0654, "step": 2529 }, { "epoch": 0.5756541524459613, "grad_norm": 1.381679821195058, "learning_rate": 1.2095932793030353e-06, "loss": 0.0513, "step": 2530 }, { "epoch": 0.575881683731513, "grad_norm": 1.468449586854502, "learning_rate": 1.2095616674791543e-06, "loss": 0.064, "step": 2531 }, { "epoch": 0.5761092150170648, "grad_norm": 1.3812350043911665, "learning_rate": 1.2095300437079034e-06, "loss": 0.0532, "step": 2532 }, { "epoch": 0.5763367463026167, "grad_norm": 0.9730226451517775, "learning_rate": 1.2094984079899292e-06, "loss": 0.0355, "step": 2533 }, { "epoch": 0.5765642775881684, "grad_norm": 1.0694961505250529, "learning_rate": 1.2094667603258779e-06, "loss": 0.0404, "step": 2534 }, { "epoch": 0.5767918088737202, "grad_norm": 1.2344551784501863, "learning_rate": 1.2094351007163962e-06, "loss": 0.0472, "step": 2535 }, { "epoch": 0.5770193401592719, "grad_norm": 1.256788629460907, "learning_rate": 1.2094034291621315e-06, "loss": 0.0744, "step": 2536 }, { "epoch": 0.5772468714448237, "grad_norm": 1.1106837372555467, "learning_rate": 1.2093717456637311e-06, "loss": 0.0338, "step": 2537 }, { "epoch": 0.5774744027303754, "grad_norm": 1.0354356057027125, "learning_rate": 1.2093400502218422e-06, "loss": 0.0525, "step": 2538 }, { "epoch": 0.5777019340159272, "grad_norm": 1.3796812821808617, "learning_rate": 1.209308342837113e-06, "loss": 0.0582, "step": 2539 }, { "epoch": 0.5779294653014789, "grad_norm": 1.0890955420481698, "learning_rate": 1.2092766235101917e-06, "loss": 0.0597, "step": 2540 }, { "epoch": 0.5781569965870307, "grad_norm": 1.0327303160169794, "learning_rate": 1.2092448922417258e-06, "loss": 0.0549, "step": 2541 }, { "epoch": 0.5783845278725824, "grad_norm": 1.0446209152101336, "learning_rate": 1.2092131490323644e-06, "loss": 0.0512, "step": 2542 }, { "epoch": 0.5786120591581343, "grad_norm": 1.1872918690030627, "learning_rate": 1.2091813938827563e-06, "loss": 0.0467, "step": 2543 }, { "epoch": 0.578839590443686, "grad_norm": 1.6213845013127128, "learning_rate": 1.2091496267935502e-06, "loss": 0.0913, "step": 2544 }, { "epoch": 0.5790671217292378, "grad_norm": 0.25107695804250185, "learning_rate": 1.2091178477653957e-06, "loss": 0.017, "step": 2545 }, { "epoch": 0.5792946530147896, "grad_norm": 0.8996867474301544, "learning_rate": 1.209086056798942e-06, "loss": 0.0386, "step": 2546 }, { "epoch": 0.5795221843003413, "grad_norm": 2.2594305507095944, "learning_rate": 1.2090542538948392e-06, "loss": 0.1133, "step": 2547 }, { "epoch": 0.579749715585893, "grad_norm": 1.1393658831306852, "learning_rate": 1.209022439053737e-06, "loss": 0.0445, "step": 2548 }, { "epoch": 0.5799772468714448, "grad_norm": 1.2977706300782228, "learning_rate": 1.2089906122762859e-06, "loss": 0.0754, "step": 2549 }, { "epoch": 0.5802047781569966, "grad_norm": 1.8878513858209192, "learning_rate": 1.208958773563136e-06, "loss": 0.0896, "step": 2550 }, { "epoch": 0.5804323094425483, "grad_norm": 1.8406180655967013, "learning_rate": 1.2089269229149383e-06, "loss": 0.0884, "step": 2551 }, { "epoch": 0.5806598407281001, "grad_norm": 1.0239444959470219, "learning_rate": 1.2088950603323438e-06, "loss": 0.0461, "step": 2552 }, { "epoch": 0.5808873720136519, "grad_norm": 1.0773801965795455, "learning_rate": 1.2088631858160033e-06, "loss": 0.0415, "step": 2553 }, { "epoch": 0.5811149032992037, "grad_norm": 0.9310673305069316, "learning_rate": 1.2088312993665689e-06, "loss": 0.042, "step": 2554 }, { "epoch": 0.5813424345847554, "grad_norm": 1.3375692899110723, "learning_rate": 1.208799400984692e-06, "loss": 0.0727, "step": 2555 }, { "epoch": 0.5815699658703072, "grad_norm": 1.4881435052347582, "learning_rate": 1.2087674906710242e-06, "loss": 0.0482, "step": 2556 }, { "epoch": 0.5817974971558589, "grad_norm": 0.9553759696828109, "learning_rate": 1.2087355684262183e-06, "loss": 0.0631, "step": 2557 }, { "epoch": 0.5820250284414107, "grad_norm": 0.6274163364554146, "learning_rate": 1.2087036342509265e-06, "loss": 0.0219, "step": 2558 }, { "epoch": 0.5822525597269624, "grad_norm": 1.3757750182752106, "learning_rate": 1.208671688145801e-06, "loss": 0.0626, "step": 2559 }, { "epoch": 0.5824800910125142, "grad_norm": 2.481636121235935, "learning_rate": 1.2086397301114955e-06, "loss": 0.1418, "step": 2560 }, { "epoch": 0.5827076222980659, "grad_norm": 1.0378265608573296, "learning_rate": 1.2086077601486627e-06, "loss": 0.0576, "step": 2561 }, { "epoch": 0.5829351535836177, "grad_norm": 1.4924581569226714, "learning_rate": 1.2085757782579562e-06, "loss": 0.0816, "step": 2562 }, { "epoch": 0.5831626848691696, "grad_norm": 0.8886308787621965, "learning_rate": 1.2085437844400293e-06, "loss": 0.0278, "step": 2563 }, { "epoch": 0.5833902161547213, "grad_norm": 1.9559855395157613, "learning_rate": 1.2085117786955366e-06, "loss": 0.107, "step": 2564 }, { "epoch": 0.5836177474402731, "grad_norm": 1.3285689681822581, "learning_rate": 1.2084797610251314e-06, "loss": 0.0593, "step": 2565 }, { "epoch": 0.5838452787258248, "grad_norm": 1.042908390638745, "learning_rate": 1.2084477314294688e-06, "loss": 0.0299, "step": 2566 }, { "epoch": 0.5840728100113766, "grad_norm": 0.9604010784829523, "learning_rate": 1.2084156899092028e-06, "loss": 0.0336, "step": 2567 }, { "epoch": 0.5843003412969283, "grad_norm": 0.9621929837602977, "learning_rate": 1.2083836364649888e-06, "loss": 0.0431, "step": 2568 }, { "epoch": 0.5845278725824801, "grad_norm": 1.2097619610555346, "learning_rate": 1.2083515710974813e-06, "loss": 0.0446, "step": 2569 }, { "epoch": 0.5847554038680318, "grad_norm": 1.3118996444529896, "learning_rate": 1.2083194938073363e-06, "loss": 0.0491, "step": 2570 }, { "epoch": 0.5849829351535836, "grad_norm": 2.2788662451246915, "learning_rate": 1.2082874045952094e-06, "loss": 0.0905, "step": 2571 }, { "epoch": 0.5852104664391354, "grad_norm": 1.7999056722845332, "learning_rate": 1.2082553034617559e-06, "loss": 0.0798, "step": 2572 }, { "epoch": 0.5854379977246872, "grad_norm": 1.1859950837264095, "learning_rate": 1.2082231904076323e-06, "loss": 0.0361, "step": 2573 }, { "epoch": 0.5856655290102389, "grad_norm": 2.3662421804783618, "learning_rate": 1.2081910654334948e-06, "loss": 0.0962, "step": 2574 }, { "epoch": 0.5858930602957907, "grad_norm": 1.248609581967699, "learning_rate": 1.2081589285399998e-06, "loss": 0.0509, "step": 2575 }, { "epoch": 0.5861205915813424, "grad_norm": 1.1728856094231628, "learning_rate": 1.2081267797278043e-06, "loss": 0.0474, "step": 2576 }, { "epoch": 0.5863481228668942, "grad_norm": 1.3290932236966655, "learning_rate": 1.2080946189975656e-06, "loss": 0.0608, "step": 2577 }, { "epoch": 0.586575654152446, "grad_norm": 1.3099426521521451, "learning_rate": 1.2080624463499407e-06, "loss": 0.0697, "step": 2578 }, { "epoch": 0.5868031854379977, "grad_norm": 1.4265140149288442, "learning_rate": 1.2080302617855874e-06, "loss": 0.0802, "step": 2579 }, { "epoch": 0.5870307167235495, "grad_norm": 0.6748962342724948, "learning_rate": 1.2079980653051629e-06, "loss": 0.0194, "step": 2580 }, { "epoch": 0.5872582480091012, "grad_norm": 1.1857133228322694, "learning_rate": 1.207965856909326e-06, "loss": 0.0603, "step": 2581 }, { "epoch": 0.5874857792946531, "grad_norm": 1.4269876330750941, "learning_rate": 1.2079336365987345e-06, "loss": 0.0674, "step": 2582 }, { "epoch": 0.5877133105802048, "grad_norm": 1.1171535242191324, "learning_rate": 1.2079014043740471e-06, "loss": 0.0474, "step": 2583 }, { "epoch": 0.5879408418657566, "grad_norm": 1.049798577134009, "learning_rate": 1.2078691602359224e-06, "loss": 0.0515, "step": 2584 }, { "epoch": 0.5881683731513083, "grad_norm": 1.4079703732165962, "learning_rate": 1.2078369041850197e-06, "loss": 0.0614, "step": 2585 }, { "epoch": 0.5883959044368601, "grad_norm": 0.6740789590290865, "learning_rate": 1.207804636221998e-06, "loss": 0.0259, "step": 2586 }, { "epoch": 0.5886234357224118, "grad_norm": 1.189214147562108, "learning_rate": 1.207772356347517e-06, "loss": 0.0527, "step": 2587 }, { "epoch": 0.5888509670079636, "grad_norm": 1.1360915553171849, "learning_rate": 1.2077400645622363e-06, "loss": 0.0458, "step": 2588 }, { "epoch": 0.5890784982935153, "grad_norm": 0.9198535003511792, "learning_rate": 1.207707760866816e-06, "loss": 0.0402, "step": 2589 }, { "epoch": 0.5893060295790671, "grad_norm": 1.174521842214451, "learning_rate": 1.207675445261916e-06, "loss": 0.0421, "step": 2590 }, { "epoch": 0.5895335608646188, "grad_norm": 2.8929933375186287, "learning_rate": 1.207643117748197e-06, "loss": 0.1536, "step": 2591 }, { "epoch": 0.5897610921501707, "grad_norm": 1.5474448385855322, "learning_rate": 1.2076107783263199e-06, "loss": 0.0567, "step": 2592 }, { "epoch": 0.5899886234357224, "grad_norm": 0.8451598056278512, "learning_rate": 1.2075784269969457e-06, "loss": 0.0362, "step": 2593 }, { "epoch": 0.5902161547212742, "grad_norm": 0.7031869485161736, "learning_rate": 1.2075460637607351e-06, "loss": 0.0312, "step": 2594 }, { "epoch": 0.590443686006826, "grad_norm": 1.449466507606748, "learning_rate": 1.20751368861835e-06, "loss": 0.0573, "step": 2595 }, { "epoch": 0.5906712172923777, "grad_norm": 1.1927092104354498, "learning_rate": 1.2074813015704518e-06, "loss": 0.0402, "step": 2596 }, { "epoch": 0.5908987485779295, "grad_norm": 1.636503401466634, "learning_rate": 1.2074489026177024e-06, "loss": 0.0657, "step": 2597 }, { "epoch": 0.5911262798634812, "grad_norm": 1.4909987612945546, "learning_rate": 1.2074164917607644e-06, "loss": 0.0658, "step": 2598 }, { "epoch": 0.591353811149033, "grad_norm": 1.1952099163133536, "learning_rate": 1.2073840690003e-06, "loss": 0.0603, "step": 2599 }, { "epoch": 0.5915813424345847, "grad_norm": 1.0622134341599558, "learning_rate": 1.2073516343369717e-06, "loss": 0.0459, "step": 2600 }, { "epoch": 0.5918088737201365, "grad_norm": 0.8032008417926766, "learning_rate": 1.2073191877714424e-06, "loss": 0.0368, "step": 2601 }, { "epoch": 0.5920364050056883, "grad_norm": 1.1983356217235557, "learning_rate": 1.2072867293043755e-06, "loss": 0.045, "step": 2602 }, { "epoch": 0.5922639362912401, "grad_norm": 1.1811729971845715, "learning_rate": 1.2072542589364343e-06, "loss": 0.0413, "step": 2603 }, { "epoch": 0.5924914675767918, "grad_norm": 0.9199950791965811, "learning_rate": 1.2072217766682822e-06, "loss": 0.0471, "step": 2604 }, { "epoch": 0.5927189988623436, "grad_norm": 1.5522989255979738, "learning_rate": 1.2071892825005835e-06, "loss": 0.0531, "step": 2605 }, { "epoch": 0.5929465301478953, "grad_norm": 2.34185081988946, "learning_rate": 1.207156776434002e-06, "loss": 0.1052, "step": 2606 }, { "epoch": 0.5931740614334471, "grad_norm": 1.6203202208735543, "learning_rate": 1.2071242584692022e-06, "loss": 0.0824, "step": 2607 }, { "epoch": 0.5934015927189988, "grad_norm": 1.8225272194053617, "learning_rate": 1.2070917286068486e-06, "loss": 0.0816, "step": 2608 }, { "epoch": 0.5936291240045506, "grad_norm": 1.0482127292144585, "learning_rate": 1.2070591868476062e-06, "loss": 0.0336, "step": 2609 }, { "epoch": 0.5938566552901023, "grad_norm": 1.2383829160698303, "learning_rate": 1.20702663319214e-06, "loss": 0.0558, "step": 2610 }, { "epoch": 0.5940841865756542, "grad_norm": 1.0803065071888958, "learning_rate": 1.2069940676411154e-06, "loss": 0.0663, "step": 2611 }, { "epoch": 0.594311717861206, "grad_norm": 1.5171982637367767, "learning_rate": 1.2069614901951978e-06, "loss": 0.0601, "step": 2612 }, { "epoch": 0.5945392491467577, "grad_norm": 1.3847567182951876, "learning_rate": 1.2069289008550533e-06, "loss": 0.0611, "step": 2613 }, { "epoch": 0.5947667804323095, "grad_norm": 0.9955435148692463, "learning_rate": 1.2068962996213476e-06, "loss": 0.0403, "step": 2614 }, { "epoch": 0.5949943117178612, "grad_norm": 1.13405892693637, "learning_rate": 1.2068636864947475e-06, "loss": 0.0519, "step": 2615 }, { "epoch": 0.595221843003413, "grad_norm": 0.7900528990598964, "learning_rate": 1.206831061475919e-06, "loss": 0.027, "step": 2616 }, { "epoch": 0.5954493742889647, "grad_norm": 0.6257774504076663, "learning_rate": 1.2067984245655292e-06, "loss": 0.0174, "step": 2617 }, { "epoch": 0.5956769055745165, "grad_norm": 0.7037772555599539, "learning_rate": 1.2067657757642453e-06, "loss": 0.0168, "step": 2618 }, { "epoch": 0.5959044368600682, "grad_norm": 0.8538766117483473, "learning_rate": 1.2067331150727343e-06, "loss": 0.0382, "step": 2619 }, { "epoch": 0.59613196814562, "grad_norm": 0.9862142615994054, "learning_rate": 1.206700442491664e-06, "loss": 0.0359, "step": 2620 }, { "epoch": 0.5963594994311718, "grad_norm": 1.5267273993929875, "learning_rate": 1.2066677580217018e-06, "loss": 0.0625, "step": 2621 }, { "epoch": 0.5965870307167236, "grad_norm": 1.5028260871816748, "learning_rate": 1.2066350616635159e-06, "loss": 0.0545, "step": 2622 }, { "epoch": 0.5968145620022753, "grad_norm": 2.64736254484509, "learning_rate": 1.2066023534177746e-06, "loss": 0.1266, "step": 2623 }, { "epoch": 0.5970420932878271, "grad_norm": 1.4120526861644525, "learning_rate": 1.2065696332851463e-06, "loss": 0.0591, "step": 2624 }, { "epoch": 0.5972696245733788, "grad_norm": 0.9067523643094818, "learning_rate": 1.2065369012662999e-06, "loss": 0.0331, "step": 2625 }, { "epoch": 0.5974971558589306, "grad_norm": 1.0988059318675343, "learning_rate": 1.206504157361904e-06, "loss": 0.0605, "step": 2626 }, { "epoch": 0.5977246871444823, "grad_norm": 1.2270739952449312, "learning_rate": 1.2064714015726283e-06, "loss": 0.056, "step": 2627 }, { "epoch": 0.5979522184300341, "grad_norm": 0.9257304826598889, "learning_rate": 1.2064386338991423e-06, "loss": 0.0433, "step": 2628 }, { "epoch": 0.5981797497155859, "grad_norm": 1.8761997788068991, "learning_rate": 1.2064058543421152e-06, "loss": 0.084, "step": 2629 }, { "epoch": 0.5984072810011376, "grad_norm": 1.5801176338760017, "learning_rate": 1.2063730629022173e-06, "loss": 0.0419, "step": 2630 }, { "epoch": 0.5986348122866895, "grad_norm": 1.1971080645956096, "learning_rate": 1.2063402595801187e-06, "loss": 0.0438, "step": 2631 }, { "epoch": 0.5988623435722412, "grad_norm": 1.405844210266892, "learning_rate": 1.2063074443764897e-06, "loss": 0.0624, "step": 2632 }, { "epoch": 0.599089874857793, "grad_norm": 1.5341316627448498, "learning_rate": 1.2062746172920014e-06, "loss": 0.088, "step": 2633 }, { "epoch": 0.5993174061433447, "grad_norm": 1.0531520309166145, "learning_rate": 1.2062417783273246e-06, "loss": 0.0474, "step": 2634 }, { "epoch": 0.5995449374288965, "grad_norm": 1.36821624035935, "learning_rate": 1.2062089274831299e-06, "loss": 0.0677, "step": 2635 }, { "epoch": 0.5997724687144482, "grad_norm": 0.7040976905373539, "learning_rate": 1.2061760647600894e-06, "loss": 0.0308, "step": 2636 }, { "epoch": 0.6, "grad_norm": 0.9178208746030599, "learning_rate": 1.2061431901588747e-06, "loss": 0.0205, "step": 2637 }, { "epoch": 0.6002275312855517, "grad_norm": 0.9105858796703735, "learning_rate": 1.2061103036801573e-06, "loss": 0.0442, "step": 2638 }, { "epoch": 0.6004550625711035, "grad_norm": 1.40951436107761, "learning_rate": 1.2060774053246096e-06, "loss": 0.0644, "step": 2639 }, { "epoch": 0.6006825938566553, "grad_norm": 1.1409764583060835, "learning_rate": 1.2060444950929038e-06, "loss": 0.0705, "step": 2640 }, { "epoch": 0.6009101251422071, "grad_norm": 1.1909546375772935, "learning_rate": 1.2060115729857128e-06, "loss": 0.0515, "step": 2641 }, { "epoch": 0.6011376564277588, "grad_norm": 1.2179103001644596, "learning_rate": 1.2059786390037093e-06, "loss": 0.0387, "step": 2642 }, { "epoch": 0.6013651877133106, "grad_norm": 1.135969807370913, "learning_rate": 1.2059456931475663e-06, "loss": 0.0224, "step": 2643 }, { "epoch": 0.6015927189988624, "grad_norm": 1.6455009465660149, "learning_rate": 1.2059127354179573e-06, "loss": 0.1139, "step": 2644 }, { "epoch": 0.6018202502844141, "grad_norm": 1.0904940664660454, "learning_rate": 1.205879765815556e-06, "loss": 0.0409, "step": 2645 }, { "epoch": 0.6020477815699659, "grad_norm": 1.8187560107849106, "learning_rate": 1.205846784341036e-06, "loss": 0.0746, "step": 2646 }, { "epoch": 0.6022753128555176, "grad_norm": 1.2846083666249735, "learning_rate": 1.2058137909950719e-06, "loss": 0.0397, "step": 2647 }, { "epoch": 0.6025028441410694, "grad_norm": 1.3139384696968406, "learning_rate": 1.2057807857783371e-06, "loss": 0.0513, "step": 2648 }, { "epoch": 0.6027303754266211, "grad_norm": 1.5334002395835504, "learning_rate": 1.205747768691507e-06, "loss": 0.0451, "step": 2649 }, { "epoch": 0.602957906712173, "grad_norm": 0.9668095767338745, "learning_rate": 1.2057147397352559e-06, "loss": 0.0519, "step": 2650 }, { "epoch": 0.6031854379977247, "grad_norm": 1.1043891282747957, "learning_rate": 1.2056816989102591e-06, "loss": 0.0421, "step": 2651 }, { "epoch": 0.6034129692832765, "grad_norm": 1.63627667644995, "learning_rate": 1.2056486462171918e-06, "loss": 0.0715, "step": 2652 }, { "epoch": 0.6036405005688282, "grad_norm": 0.986630375032043, "learning_rate": 1.2056155816567297e-06, "loss": 0.0497, "step": 2653 }, { "epoch": 0.60386803185438, "grad_norm": 1.198264842052037, "learning_rate": 1.2055825052295486e-06, "loss": 0.0565, "step": 2654 }, { "epoch": 0.6040955631399317, "grad_norm": 0.7380211191011931, "learning_rate": 1.205549416936324e-06, "loss": 0.0423, "step": 2655 }, { "epoch": 0.6043230944254835, "grad_norm": 0.8003290473544856, "learning_rate": 1.2055163167777328e-06, "loss": 0.057, "step": 2656 }, { "epoch": 0.6045506257110352, "grad_norm": 1.2699158922395328, "learning_rate": 1.2054832047544512e-06, "loss": 0.0588, "step": 2657 }, { "epoch": 0.604778156996587, "grad_norm": 0.7610615046539442, "learning_rate": 1.205450080867156e-06, "loss": 0.0343, "step": 2658 }, { "epoch": 0.6050056882821387, "grad_norm": 1.3103630700346172, "learning_rate": 1.205416945116524e-06, "loss": 0.0406, "step": 2659 }, { "epoch": 0.6052332195676906, "grad_norm": 1.0165185597393296, "learning_rate": 1.2053837975032328e-06, "loss": 0.0337, "step": 2660 }, { "epoch": 0.6054607508532424, "grad_norm": 1.387250593088855, "learning_rate": 1.2053506380279597e-06, "loss": 0.0603, "step": 2661 }, { "epoch": 0.6056882821387941, "grad_norm": 1.335719349119118, "learning_rate": 1.2053174666913826e-06, "loss": 0.0729, "step": 2662 }, { "epoch": 0.6059158134243459, "grad_norm": 0.7777911730162685, "learning_rate": 1.2052842834941791e-06, "loss": 0.0351, "step": 2663 }, { "epoch": 0.6061433447098976, "grad_norm": 0.9791113605180757, "learning_rate": 1.2052510884370274e-06, "loss": 0.0447, "step": 2664 }, { "epoch": 0.6063708759954494, "grad_norm": 0.967866940301246, "learning_rate": 1.2052178815206064e-06, "loss": 0.0371, "step": 2665 }, { "epoch": 0.6065984072810011, "grad_norm": 1.534725370801602, "learning_rate": 1.2051846627455946e-06, "loss": 0.0826, "step": 2666 }, { "epoch": 0.6068259385665529, "grad_norm": 0.9691101488135158, "learning_rate": 1.2051514321126705e-06, "loss": 0.0486, "step": 2667 }, { "epoch": 0.6070534698521046, "grad_norm": 1.352223067560777, "learning_rate": 1.2051181896225139e-06, "loss": 0.0496, "step": 2668 }, { "epoch": 0.6072810011376564, "grad_norm": 0.9959104730289088, "learning_rate": 1.205084935275804e-06, "loss": 0.0458, "step": 2669 }, { "epoch": 0.6075085324232082, "grad_norm": 1.8958304200279301, "learning_rate": 1.20505166907322e-06, "loss": 0.0702, "step": 2670 }, { "epoch": 0.60773606370876, "grad_norm": 1.1373672856837178, "learning_rate": 1.2050183910154425e-06, "loss": 0.0642, "step": 2671 }, { "epoch": 0.6079635949943117, "grad_norm": 1.5146568276899794, "learning_rate": 1.2049851011031514e-06, "loss": 0.0615, "step": 2672 }, { "epoch": 0.6081911262798635, "grad_norm": 1.3486666030528254, "learning_rate": 1.2049517993370269e-06, "loss": 0.0407, "step": 2673 }, { "epoch": 0.6084186575654152, "grad_norm": 1.7553759733095198, "learning_rate": 1.2049184857177498e-06, "loss": 0.089, "step": 2674 }, { "epoch": 0.608646188850967, "grad_norm": 1.6341921000073825, "learning_rate": 1.204885160246001e-06, "loss": 0.1034, "step": 2675 }, { "epoch": 0.6088737201365187, "grad_norm": 1.2886908765685168, "learning_rate": 1.2048518229224613e-06, "loss": 0.0371, "step": 2676 }, { "epoch": 0.6091012514220705, "grad_norm": 1.147794266112336, "learning_rate": 1.2048184737478124e-06, "loss": 0.0513, "step": 2677 }, { "epoch": 0.6093287827076223, "grad_norm": 1.108848937036823, "learning_rate": 1.2047851127227358e-06, "loss": 0.0372, "step": 2678 }, { "epoch": 0.6095563139931741, "grad_norm": 0.9494871582849728, "learning_rate": 1.2047517398479135e-06, "loss": 0.0313, "step": 2679 }, { "epoch": 0.6097838452787259, "grad_norm": 1.4403996402965795, "learning_rate": 1.204718355124027e-06, "loss": 0.0648, "step": 2680 }, { "epoch": 0.6100113765642776, "grad_norm": 1.1467814441502786, "learning_rate": 1.2046849585517595e-06, "loss": 0.0363, "step": 2681 }, { "epoch": 0.6102389078498294, "grad_norm": 0.8769285585937754, "learning_rate": 1.2046515501317927e-06, "loss": 0.0439, "step": 2682 }, { "epoch": 0.6104664391353811, "grad_norm": 1.0907507673380405, "learning_rate": 1.2046181298648101e-06, "loss": 0.0362, "step": 2683 }, { "epoch": 0.6106939704209329, "grad_norm": 1.4400540200529313, "learning_rate": 1.2045846977514943e-06, "loss": 0.0695, "step": 2684 }, { "epoch": 0.6109215017064846, "grad_norm": 0.926017693420237, "learning_rate": 1.2045512537925287e-06, "loss": 0.0319, "step": 2685 }, { "epoch": 0.6111490329920364, "grad_norm": 1.9422364837523665, "learning_rate": 1.2045177979885969e-06, "loss": 0.0678, "step": 2686 }, { "epoch": 0.6113765642775881, "grad_norm": 1.0780377835715493, "learning_rate": 1.2044843303403827e-06, "loss": 0.051, "step": 2687 }, { "epoch": 0.6116040955631399, "grad_norm": 1.522022597658668, "learning_rate": 1.20445085084857e-06, "loss": 0.0917, "step": 2688 }, { "epoch": 0.6118316268486917, "grad_norm": 1.0239752154710169, "learning_rate": 1.204417359513843e-06, "loss": 0.0286, "step": 2689 }, { "epoch": 0.6120591581342435, "grad_norm": 1.150678969827508, "learning_rate": 1.2043838563368865e-06, "loss": 0.0554, "step": 2690 }, { "epoch": 0.6122866894197952, "grad_norm": 1.2280896133913035, "learning_rate": 1.204350341318385e-06, "loss": 0.0562, "step": 2691 }, { "epoch": 0.612514220705347, "grad_norm": 1.0374592547169612, "learning_rate": 1.2043168144590237e-06, "loss": 0.0483, "step": 2692 }, { "epoch": 0.6127417519908988, "grad_norm": 1.8353693399206994, "learning_rate": 1.2042832757594875e-06, "loss": 0.0901, "step": 2693 }, { "epoch": 0.6129692832764505, "grad_norm": 1.6915158293911816, "learning_rate": 1.204249725220462e-06, "loss": 0.0741, "step": 2694 }, { "epoch": 0.6131968145620023, "grad_norm": 1.18563782209136, "learning_rate": 1.204216162842633e-06, "loss": 0.0406, "step": 2695 }, { "epoch": 0.613424345847554, "grad_norm": 1.2961769715929403, "learning_rate": 1.2041825886266866e-06, "loss": 0.0449, "step": 2696 }, { "epoch": 0.6136518771331058, "grad_norm": 1.5168600612694154, "learning_rate": 1.2041490025733089e-06, "loss": 0.0853, "step": 2697 }, { "epoch": 0.6138794084186575, "grad_norm": 1.0552035992692197, "learning_rate": 1.2041154046831859e-06, "loss": 0.0459, "step": 2698 }, { "epoch": 0.6141069397042094, "grad_norm": 1.0052589171106947, "learning_rate": 1.2040817949570046e-06, "loss": 0.0519, "step": 2699 }, { "epoch": 0.6143344709897611, "grad_norm": 1.5847978638614886, "learning_rate": 1.2040481733954523e-06, "loss": 0.0489, "step": 2700 }, { "epoch": 0.6145620022753129, "grad_norm": 1.5433588402630525, "learning_rate": 1.2040145399992157e-06, "loss": 0.0554, "step": 2701 }, { "epoch": 0.6147895335608646, "grad_norm": 0.9199676231545262, "learning_rate": 1.203980894768982e-06, "loss": 0.051, "step": 2702 }, { "epoch": 0.6150170648464164, "grad_norm": 1.6984060525727276, "learning_rate": 1.2039472377054395e-06, "loss": 0.0951, "step": 2703 }, { "epoch": 0.6152445961319681, "grad_norm": 1.9254054628589912, "learning_rate": 1.2039135688092757e-06, "loss": 0.0614, "step": 2704 }, { "epoch": 0.6154721274175199, "grad_norm": 1.3832191004617072, "learning_rate": 1.203879888081179e-06, "loss": 0.0666, "step": 2705 }, { "epoch": 0.6156996587030716, "grad_norm": 1.915733750626152, "learning_rate": 1.203846195521837e-06, "loss": 0.0622, "step": 2706 }, { "epoch": 0.6159271899886234, "grad_norm": 1.1800428967306633, "learning_rate": 1.2038124911319393e-06, "loss": 0.0428, "step": 2707 }, { "epoch": 0.6161547212741753, "grad_norm": 1.464495992586222, "learning_rate": 1.2037787749121741e-06, "loss": 0.069, "step": 2708 }, { "epoch": 0.616382252559727, "grad_norm": 1.1760191210505098, "learning_rate": 1.203745046863231e-06, "loss": 0.0503, "step": 2709 }, { "epoch": 0.6166097838452788, "grad_norm": 1.2442850515401647, "learning_rate": 1.203711306985799e-06, "loss": 0.04, "step": 2710 }, { "epoch": 0.6168373151308305, "grad_norm": 0.865019588027552, "learning_rate": 1.2036775552805674e-06, "loss": 0.0294, "step": 2711 }, { "epoch": 0.6170648464163823, "grad_norm": 2.6756015679269978, "learning_rate": 1.2036437917482267e-06, "loss": 0.0814, "step": 2712 }, { "epoch": 0.617292377701934, "grad_norm": 0.8893160007809519, "learning_rate": 1.2036100163894665e-06, "loss": 0.0434, "step": 2713 }, { "epoch": 0.6175199089874858, "grad_norm": 1.5121328039736281, "learning_rate": 1.2035762292049772e-06, "loss": 0.0848, "step": 2714 }, { "epoch": 0.6177474402730375, "grad_norm": 1.1556104934545364, "learning_rate": 1.2035424301954496e-06, "loss": 0.0564, "step": 2715 }, { "epoch": 0.6179749715585893, "grad_norm": 1.344567899518094, "learning_rate": 1.2035086193615743e-06, "loss": 0.0449, "step": 2716 }, { "epoch": 0.618202502844141, "grad_norm": 1.6415013868308683, "learning_rate": 1.203474796704042e-06, "loss": 0.0598, "step": 2717 }, { "epoch": 0.6184300341296929, "grad_norm": 0.9852915602376644, "learning_rate": 1.2034409622235444e-06, "loss": 0.0285, "step": 2718 }, { "epoch": 0.6186575654152446, "grad_norm": 1.1301434136989037, "learning_rate": 1.203407115920773e-06, "loss": 0.0473, "step": 2719 }, { "epoch": 0.6188850967007964, "grad_norm": 1.3150440764406408, "learning_rate": 1.2033732577964194e-06, "loss": 0.0717, "step": 2720 }, { "epoch": 0.6191126279863481, "grad_norm": 1.4850057002676174, "learning_rate": 1.2033393878511756e-06, "loss": 0.0589, "step": 2721 }, { "epoch": 0.6193401592718999, "grad_norm": 1.5995015548370441, "learning_rate": 1.203305506085734e-06, "loss": 0.0705, "step": 2722 }, { "epoch": 0.6195676905574516, "grad_norm": 0.8404790504075466, "learning_rate": 1.2032716125007868e-06, "loss": 0.0437, "step": 2723 }, { "epoch": 0.6197952218430034, "grad_norm": 1.3772954832300879, "learning_rate": 1.2032377070970268e-06, "loss": 0.0318, "step": 2724 }, { "epoch": 0.6200227531285551, "grad_norm": 1.4515256482801804, "learning_rate": 1.2032037898751475e-06, "loss": 0.0473, "step": 2725 }, { "epoch": 0.6202502844141069, "grad_norm": 0.7545612524043191, "learning_rate": 1.2031698608358414e-06, "loss": 0.0269, "step": 2726 }, { "epoch": 0.6204778156996587, "grad_norm": 0.743007367054589, "learning_rate": 1.2031359199798021e-06, "loss": 0.0283, "step": 2727 }, { "epoch": 0.6207053469852105, "grad_norm": 1.2718184804954529, "learning_rate": 1.2031019673077237e-06, "loss": 0.0634, "step": 2728 }, { "epoch": 0.6209328782707623, "grad_norm": 1.775193564914863, "learning_rate": 1.2030680028202995e-06, "loss": 0.1335, "step": 2729 }, { "epoch": 0.621160409556314, "grad_norm": 1.1547031507192722, "learning_rate": 1.2030340265182242e-06, "loss": 0.026, "step": 2730 }, { "epoch": 0.6213879408418658, "grad_norm": 1.81101747698367, "learning_rate": 1.2030000384021919e-06, "loss": 0.1018, "step": 2731 }, { "epoch": 0.6216154721274175, "grad_norm": 1.5054245916089843, "learning_rate": 1.2029660384728974e-06, "loss": 0.0495, "step": 2732 }, { "epoch": 0.6218430034129693, "grad_norm": 1.6121127283946548, "learning_rate": 1.2029320267310359e-06, "loss": 0.0782, "step": 2733 }, { "epoch": 0.622070534698521, "grad_norm": 1.6042883442076186, "learning_rate": 1.2028980031773018e-06, "loss": 0.0589, "step": 2734 }, { "epoch": 0.6222980659840728, "grad_norm": 1.2540922996339912, "learning_rate": 1.202863967812391e-06, "loss": 0.0774, "step": 2735 }, { "epoch": 0.6225255972696245, "grad_norm": 1.518635659663246, "learning_rate": 1.2028299206369991e-06, "loss": 0.1072, "step": 2736 }, { "epoch": 0.6227531285551763, "grad_norm": 1.3308844577317762, "learning_rate": 1.2027958616518218e-06, "loss": 0.0445, "step": 2737 }, { "epoch": 0.6229806598407281, "grad_norm": 2.209735756502863, "learning_rate": 1.2027617908575553e-06, "loss": 0.1124, "step": 2738 }, { "epoch": 0.6232081911262799, "grad_norm": 1.3929457690467317, "learning_rate": 1.2027277082548958e-06, "loss": 0.0545, "step": 2739 }, { "epoch": 0.6234357224118316, "grad_norm": 1.051968724737917, "learning_rate": 1.20269361384454e-06, "loss": 0.033, "step": 2740 }, { "epoch": 0.6236632536973834, "grad_norm": 1.6764017429869924, "learning_rate": 1.2026595076271848e-06, "loss": 0.0834, "step": 2741 }, { "epoch": 0.6238907849829352, "grad_norm": 1.6397713261851863, "learning_rate": 1.2026253896035273e-06, "loss": 0.0693, "step": 2742 }, { "epoch": 0.6241183162684869, "grad_norm": 1.591510561765584, "learning_rate": 1.2025912597742646e-06, "loss": 0.0768, "step": 2743 }, { "epoch": 0.6243458475540387, "grad_norm": 1.5008432457963459, "learning_rate": 1.2025571181400944e-06, "loss": 0.0771, "step": 2744 }, { "epoch": 0.6245733788395904, "grad_norm": 1.362876261935557, "learning_rate": 1.2025229647017145e-06, "loss": 0.0523, "step": 2745 }, { "epoch": 0.6248009101251422, "grad_norm": 2.2117120292707977, "learning_rate": 1.2024887994598227e-06, "loss": 0.0941, "step": 2746 }, { "epoch": 0.625028441410694, "grad_norm": 0.962824420377618, "learning_rate": 1.2024546224151176e-06, "loss": 0.0339, "step": 2747 }, { "epoch": 0.6252559726962458, "grad_norm": 0.8088610019406368, "learning_rate": 1.2024204335682977e-06, "loss": 0.0269, "step": 2748 }, { "epoch": 0.6254835039817975, "grad_norm": 2.415791185781217, "learning_rate": 1.2023862329200613e-06, "loss": 0.0925, "step": 2749 }, { "epoch": 0.6257110352673493, "grad_norm": 1.0450350621113902, "learning_rate": 1.2023520204711078e-06, "loss": 0.0539, "step": 2750 }, { "epoch": 0.625938566552901, "grad_norm": 1.3226699575436216, "learning_rate": 1.2023177962221366e-06, "loss": 0.098, "step": 2751 }, { "epoch": 0.6261660978384528, "grad_norm": 1.6169539390839551, "learning_rate": 1.2022835601738467e-06, "loss": 0.0674, "step": 2752 }, { "epoch": 0.6263936291240045, "grad_norm": 0.926101104299797, "learning_rate": 1.2022493123269383e-06, "loss": 0.0326, "step": 2753 }, { "epoch": 0.6266211604095563, "grad_norm": 1.7687331075293982, "learning_rate": 1.202215052682111e-06, "loss": 0.1007, "step": 2754 }, { "epoch": 0.626848691695108, "grad_norm": 1.2611712167362608, "learning_rate": 1.2021807812400652e-06, "loss": 0.0739, "step": 2755 }, { "epoch": 0.6270762229806598, "grad_norm": 1.5800647590282382, "learning_rate": 1.2021464980015014e-06, "loss": 0.078, "step": 2756 }, { "epoch": 0.6273037542662117, "grad_norm": 1.2594753097389422, "learning_rate": 1.20211220296712e-06, "loss": 0.0642, "step": 2757 }, { "epoch": 0.6275312855517634, "grad_norm": 1.121495104083563, "learning_rate": 1.2020778961376223e-06, "loss": 0.0328, "step": 2758 }, { "epoch": 0.6277588168373152, "grad_norm": 1.4683080872027747, "learning_rate": 1.202043577513709e-06, "loss": 0.0902, "step": 2759 }, { "epoch": 0.6279863481228669, "grad_norm": 1.8245567164154206, "learning_rate": 1.202009247096082e-06, "loss": 0.0761, "step": 2760 }, { "epoch": 0.6282138794084187, "grad_norm": 1.1125392914595493, "learning_rate": 1.2019749048854426e-06, "loss": 0.0646, "step": 2761 }, { "epoch": 0.6284414106939704, "grad_norm": 0.8642533280519576, "learning_rate": 1.2019405508824927e-06, "loss": 0.0324, "step": 2762 }, { "epoch": 0.6286689419795222, "grad_norm": 1.7625633245035697, "learning_rate": 1.201906185087935e-06, "loss": 0.0776, "step": 2763 }, { "epoch": 0.6288964732650739, "grad_norm": 1.9920964212571328, "learning_rate": 1.201871807502471e-06, "loss": 0.0628, "step": 2764 }, { "epoch": 0.6291240045506257, "grad_norm": 1.371538527134396, "learning_rate": 1.2018374181268039e-06, "loss": 0.054, "step": 2765 }, { "epoch": 0.6293515358361774, "grad_norm": 1.603991934514247, "learning_rate": 1.2018030169616363e-06, "loss": 0.0675, "step": 2766 }, { "epoch": 0.6295790671217293, "grad_norm": 1.3082298269977088, "learning_rate": 1.2017686040076715e-06, "loss": 0.059, "step": 2767 }, { "epoch": 0.629806598407281, "grad_norm": 1.0629082940323715, "learning_rate": 1.2017341792656129e-06, "loss": 0.0526, "step": 2768 }, { "epoch": 0.6300341296928328, "grad_norm": 1.5246818248480305, "learning_rate": 1.2016997427361635e-06, "loss": 0.0489, "step": 2769 }, { "epoch": 0.6302616609783845, "grad_norm": 1.4969549226937944, "learning_rate": 1.201665294420028e-06, "loss": 0.079, "step": 2770 }, { "epoch": 0.6304891922639363, "grad_norm": 1.195617244868444, "learning_rate": 1.2016308343179098e-06, "loss": 0.038, "step": 2771 }, { "epoch": 0.630716723549488, "grad_norm": 1.0734726876278198, "learning_rate": 1.2015963624305132e-06, "loss": 0.0528, "step": 2772 }, { "epoch": 0.6309442548350398, "grad_norm": 0.9060818686627535, "learning_rate": 1.201561878758543e-06, "loss": 0.0321, "step": 2773 }, { "epoch": 0.6311717861205915, "grad_norm": 1.9200212018830873, "learning_rate": 1.2015273833027041e-06, "loss": 0.0827, "step": 2774 }, { "epoch": 0.6313993174061433, "grad_norm": 1.1822527617116831, "learning_rate": 1.2014928760637014e-06, "loss": 0.0537, "step": 2775 }, { "epoch": 0.631626848691695, "grad_norm": 1.618374665019338, "learning_rate": 1.20145835704224e-06, "loss": 0.0563, "step": 2776 }, { "epoch": 0.6318543799772469, "grad_norm": 2.134594638963389, "learning_rate": 1.2014238262390254e-06, "loss": 0.0709, "step": 2777 }, { "epoch": 0.6320819112627987, "grad_norm": 2.1175666116051968, "learning_rate": 1.2013892836547635e-06, "loss": 0.0924, "step": 2778 }, { "epoch": 0.6323094425483504, "grad_norm": 1.5175276686793735, "learning_rate": 1.2013547292901605e-06, "loss": 0.0818, "step": 2779 }, { "epoch": 0.6325369738339022, "grad_norm": 1.1839298832479785, "learning_rate": 1.2013201631459222e-06, "loss": 0.0709, "step": 2780 }, { "epoch": 0.6327645051194539, "grad_norm": 2.288412160525939, "learning_rate": 1.201285585222755e-06, "loss": 0.105, "step": 2781 }, { "epoch": 0.6329920364050057, "grad_norm": 1.569225058207049, "learning_rate": 1.2012509955213664e-06, "loss": 0.0477, "step": 2782 }, { "epoch": 0.6332195676905574, "grad_norm": 0.7422297833198885, "learning_rate": 1.2012163940424624e-06, "loss": 0.0359, "step": 2783 }, { "epoch": 0.6334470989761092, "grad_norm": 1.009892072927878, "learning_rate": 1.2011817807867507e-06, "loss": 0.0472, "step": 2784 }, { "epoch": 0.6336746302616609, "grad_norm": 1.8711194663634534, "learning_rate": 1.2011471557549387e-06, "loss": 0.0621, "step": 2785 }, { "epoch": 0.6339021615472128, "grad_norm": 1.0981859377257484, "learning_rate": 1.2011125189477339e-06, "loss": 0.0407, "step": 2786 }, { "epoch": 0.6341296928327645, "grad_norm": 2.3667510812808903, "learning_rate": 1.2010778703658441e-06, "loss": 0.103, "step": 2787 }, { "epoch": 0.6343572241183163, "grad_norm": 1.4344646107903631, "learning_rate": 1.2010432100099781e-06, "loss": 0.0796, "step": 2788 }, { "epoch": 0.634584755403868, "grad_norm": 1.6112614680461406, "learning_rate": 1.2010085378808437e-06, "loss": 0.0635, "step": 2789 }, { "epoch": 0.6348122866894198, "grad_norm": 1.8958308802132453, "learning_rate": 1.2009738539791497e-06, "loss": 0.1027, "step": 2790 }, { "epoch": 0.6350398179749716, "grad_norm": 1.4317662705377017, "learning_rate": 1.2009391583056048e-06, "loss": 0.0689, "step": 2791 }, { "epoch": 0.6352673492605233, "grad_norm": 1.2007483338261875, "learning_rate": 1.2009044508609182e-06, "loss": 0.0584, "step": 2792 }, { "epoch": 0.6354948805460751, "grad_norm": 1.9010319599598573, "learning_rate": 1.2008697316457997e-06, "loss": 0.0966, "step": 2793 }, { "epoch": 0.6357224118316268, "grad_norm": 0.9068372322632697, "learning_rate": 1.2008350006609584e-06, "loss": 0.0319, "step": 2794 }, { "epoch": 0.6359499431171786, "grad_norm": 1.9186704917037192, "learning_rate": 1.2008002579071043e-06, "loss": 0.0834, "step": 2795 }, { "epoch": 0.6361774744027304, "grad_norm": 1.5719062287923162, "learning_rate": 1.2007655033849474e-06, "loss": 0.0719, "step": 2796 }, { "epoch": 0.6364050056882822, "grad_norm": 0.8675484564571867, "learning_rate": 1.2007307370951983e-06, "loss": 0.0382, "step": 2797 }, { "epoch": 0.6366325369738339, "grad_norm": 1.2487471267893597, "learning_rate": 1.200695959038567e-06, "loss": 0.0541, "step": 2798 }, { "epoch": 0.6368600682593857, "grad_norm": 1.340853119074798, "learning_rate": 1.2006611692157648e-06, "loss": 0.0625, "step": 2799 }, { "epoch": 0.6370875995449374, "grad_norm": 1.0090275039380872, "learning_rate": 1.2006263676275026e-06, "loss": 0.0386, "step": 2800 }, { "epoch": 0.6373151308304892, "grad_norm": 1.3604638479008277, "learning_rate": 1.2005915542744915e-06, "loss": 0.0579, "step": 2801 }, { "epoch": 0.6375426621160409, "grad_norm": 1.3626136251906178, "learning_rate": 1.2005567291574434e-06, "loss": 0.0506, "step": 2802 }, { "epoch": 0.6377701934015927, "grad_norm": 2.24330408064307, "learning_rate": 1.2005218922770695e-06, "loss": 0.0934, "step": 2803 }, { "epoch": 0.6379977246871444, "grad_norm": 1.652449071090047, "learning_rate": 1.2004870436340824e-06, "loss": 0.0916, "step": 2804 }, { "epoch": 0.6382252559726962, "grad_norm": 1.335757404355839, "learning_rate": 1.2004521832291943e-06, "loss": 0.0319, "step": 2805 }, { "epoch": 0.6384527872582481, "grad_norm": 1.2669579033763387, "learning_rate": 1.200417311063117e-06, "loss": 0.0452, "step": 2806 }, { "epoch": 0.6386803185437998, "grad_norm": 1.3480996457776724, "learning_rate": 1.200382427136564e-06, "loss": 0.0664, "step": 2807 }, { "epoch": 0.6389078498293516, "grad_norm": 1.6036199545645866, "learning_rate": 1.2003475314502477e-06, "loss": 0.0715, "step": 2808 }, { "epoch": 0.6391353811149033, "grad_norm": 1.0938499350027229, "learning_rate": 1.200312624004882e-06, "loss": 0.0467, "step": 2809 }, { "epoch": 0.6393629124004551, "grad_norm": 1.301558947668022, "learning_rate": 1.2002777048011794e-06, "loss": 0.0611, "step": 2810 }, { "epoch": 0.6395904436860068, "grad_norm": 1.2950309349789517, "learning_rate": 1.2002427738398543e-06, "loss": 0.0501, "step": 2811 }, { "epoch": 0.6398179749715586, "grad_norm": 1.074515099435565, "learning_rate": 1.2002078311216205e-06, "loss": 0.0273, "step": 2812 }, { "epoch": 0.6400455062571103, "grad_norm": 0.9534944533034787, "learning_rate": 1.2001728766471919e-06, "loss": 0.0517, "step": 2813 }, { "epoch": 0.6402730375426621, "grad_norm": 1.5158088143964041, "learning_rate": 1.2001379104172832e-06, "loss": 0.0638, "step": 2814 }, { "epoch": 0.6405005688282139, "grad_norm": 1.247966598668176, "learning_rate": 1.2001029324326087e-06, "loss": 0.054, "step": 2815 }, { "epoch": 0.6407281001137657, "grad_norm": 0.9514285262896064, "learning_rate": 1.2000679426938838e-06, "loss": 0.0324, "step": 2816 }, { "epoch": 0.6409556313993174, "grad_norm": 1.5357576363213208, "learning_rate": 1.2000329412018233e-06, "loss": 0.0504, "step": 2817 }, { "epoch": 0.6411831626848692, "grad_norm": 1.3766532946507657, "learning_rate": 1.1999979279571425e-06, "loss": 0.0785, "step": 2818 }, { "epoch": 0.6414106939704209, "grad_norm": 1.1854802730452816, "learning_rate": 1.1999629029605572e-06, "loss": 0.0471, "step": 2819 }, { "epoch": 0.6416382252559727, "grad_norm": 0.9120829198047485, "learning_rate": 1.1999278662127832e-06, "loss": 0.0465, "step": 2820 }, { "epoch": 0.6418657565415244, "grad_norm": 1.215725225022476, "learning_rate": 1.1998928177145363e-06, "loss": 0.0539, "step": 2821 }, { "epoch": 0.6420932878270762, "grad_norm": 1.6055605292211583, "learning_rate": 1.1998577574665334e-06, "loss": 0.0794, "step": 2822 }, { "epoch": 0.642320819112628, "grad_norm": 1.5194681984361975, "learning_rate": 1.1998226854694906e-06, "loss": 0.0775, "step": 2823 }, { "epoch": 0.6425483503981797, "grad_norm": 1.217305433486093, "learning_rate": 1.1997876017241248e-06, "loss": 0.0363, "step": 2824 }, { "epoch": 0.6427758816837316, "grad_norm": 1.2554857295036752, "learning_rate": 1.199752506231153e-06, "loss": 0.0451, "step": 2825 }, { "epoch": 0.6430034129692833, "grad_norm": 1.4903034342688748, "learning_rate": 1.1997173989912928e-06, "loss": 0.0651, "step": 2826 }, { "epoch": 0.6432309442548351, "grad_norm": 0.9479490152328633, "learning_rate": 1.1996822800052614e-06, "loss": 0.0416, "step": 2827 }, { "epoch": 0.6434584755403868, "grad_norm": 1.3630232299101448, "learning_rate": 1.1996471492737767e-06, "loss": 0.053, "step": 2828 }, { "epoch": 0.6436860068259386, "grad_norm": 0.8310504363214145, "learning_rate": 1.1996120067975568e-06, "loss": 0.0401, "step": 2829 }, { "epoch": 0.6439135381114903, "grad_norm": 0.7817607162647777, "learning_rate": 1.1995768525773195e-06, "loss": 0.0381, "step": 2830 }, { "epoch": 0.6441410693970421, "grad_norm": 1.19151900025748, "learning_rate": 1.1995416866137837e-06, "loss": 0.0401, "step": 2831 }, { "epoch": 0.6443686006825938, "grad_norm": 0.8480287553862164, "learning_rate": 1.1995065089076682e-06, "loss": 0.0427, "step": 2832 }, { "epoch": 0.6445961319681456, "grad_norm": 1.3823956321024167, "learning_rate": 1.199471319459692e-06, "loss": 0.0678, "step": 2833 }, { "epoch": 0.6448236632536973, "grad_norm": 1.0881726943732215, "learning_rate": 1.199436118270574e-06, "loss": 0.0463, "step": 2834 }, { "epoch": 0.6450511945392492, "grad_norm": 0.9603443906924007, "learning_rate": 1.1994009053410336e-06, "loss": 0.0334, "step": 2835 }, { "epoch": 0.645278725824801, "grad_norm": 0.7257173068914297, "learning_rate": 1.1993656806717906e-06, "loss": 0.022, "step": 2836 }, { "epoch": 0.6455062571103527, "grad_norm": 1.3460255228704814, "learning_rate": 1.199330444263565e-06, "loss": 0.0589, "step": 2837 }, { "epoch": 0.6457337883959045, "grad_norm": 1.8598440039360982, "learning_rate": 1.1992951961170774e-06, "loss": 0.0884, "step": 2838 }, { "epoch": 0.6459613196814562, "grad_norm": 1.7631231436296406, "learning_rate": 1.1992599362330474e-06, "loss": 0.0691, "step": 2839 }, { "epoch": 0.646188850967008, "grad_norm": 1.0081908118372565, "learning_rate": 1.199224664612196e-06, "loss": 0.046, "step": 2840 }, { "epoch": 0.6464163822525597, "grad_norm": 1.5846986206428932, "learning_rate": 1.199189381255244e-06, "loss": 0.0819, "step": 2841 }, { "epoch": 0.6466439135381115, "grad_norm": 1.3965515241420483, "learning_rate": 1.199154086162913e-06, "loss": 0.0675, "step": 2842 }, { "epoch": 0.6468714448236632, "grad_norm": 1.7830102963673449, "learning_rate": 1.1991187793359239e-06, "loss": 0.1173, "step": 2843 }, { "epoch": 0.647098976109215, "grad_norm": 0.7979390598594543, "learning_rate": 1.1990834607749981e-06, "loss": 0.0288, "step": 2844 }, { "epoch": 0.6473265073947668, "grad_norm": 1.2345551184177335, "learning_rate": 1.199048130480858e-06, "loss": 0.0823, "step": 2845 }, { "epoch": 0.6475540386803186, "grad_norm": 1.3892698568903075, "learning_rate": 1.1990127884542251e-06, "loss": 0.0739, "step": 2846 }, { "epoch": 0.6477815699658703, "grad_norm": 1.5536465235111332, "learning_rate": 1.1989774346958225e-06, "loss": 0.0725, "step": 2847 }, { "epoch": 0.6480091012514221, "grad_norm": 1.482520572829848, "learning_rate": 1.1989420692063723e-06, "loss": 0.0902, "step": 2848 }, { "epoch": 0.6482366325369738, "grad_norm": 1.0973672503057945, "learning_rate": 1.198906691986597e-06, "loss": 0.0567, "step": 2849 }, { "epoch": 0.6484641638225256, "grad_norm": 1.974255381506691, "learning_rate": 1.1988713030372202e-06, "loss": 0.1019, "step": 2850 }, { "epoch": 0.6486916951080773, "grad_norm": 0.9822063316024513, "learning_rate": 1.198835902358965e-06, "loss": 0.0556, "step": 2851 }, { "epoch": 0.6489192263936291, "grad_norm": 0.9883294090430303, "learning_rate": 1.1988004899525547e-06, "loss": 0.0511, "step": 2852 }, { "epoch": 0.6491467576791808, "grad_norm": 1.7242005203993112, "learning_rate": 1.1987650658187133e-06, "loss": 0.0684, "step": 2853 }, { "epoch": 0.6493742889647327, "grad_norm": 1.163292058133356, "learning_rate": 1.1987296299581648e-06, "loss": 0.055, "step": 2854 }, { "epoch": 0.6496018202502845, "grad_norm": 0.842300844455492, "learning_rate": 1.1986941823716333e-06, "loss": 0.0361, "step": 2855 }, { "epoch": 0.6498293515358362, "grad_norm": 0.7877836086747282, "learning_rate": 1.1986587230598437e-06, "loss": 0.0269, "step": 2856 }, { "epoch": 0.650056882821388, "grad_norm": 1.1854138002017856, "learning_rate": 1.19862325202352e-06, "loss": 0.0789, "step": 2857 }, { "epoch": 0.6502844141069397, "grad_norm": 1.5193193171125443, "learning_rate": 1.198587769263388e-06, "loss": 0.0647, "step": 2858 }, { "epoch": 0.6505119453924915, "grad_norm": 0.9471414553218402, "learning_rate": 1.198552274780172e-06, "loss": 0.0394, "step": 2859 }, { "epoch": 0.6507394766780432, "grad_norm": 1.142561403387133, "learning_rate": 1.1985167685745982e-06, "loss": 0.0427, "step": 2860 }, { "epoch": 0.650967007963595, "grad_norm": 1.163361004688848, "learning_rate": 1.198481250647392e-06, "loss": 0.0583, "step": 2861 }, { "epoch": 0.6511945392491467, "grad_norm": 1.2956201317841325, "learning_rate": 1.1984457209992792e-06, "loss": 0.0497, "step": 2862 }, { "epoch": 0.6514220705346985, "grad_norm": 1.0871624809389135, "learning_rate": 1.1984101796309862e-06, "loss": 0.0719, "step": 2863 }, { "epoch": 0.6516496018202503, "grad_norm": 1.3146485801211771, "learning_rate": 1.1983746265432392e-06, "loss": 0.0589, "step": 2864 }, { "epoch": 0.6518771331058021, "grad_norm": 1.5681925610927387, "learning_rate": 1.1983390617367649e-06, "loss": 0.0701, "step": 2865 }, { "epoch": 0.6521046643913538, "grad_norm": 2.2916879452331926, "learning_rate": 1.1983034852122902e-06, "loss": 0.1033, "step": 2866 }, { "epoch": 0.6523321956769056, "grad_norm": 1.0785758663467682, "learning_rate": 1.1982678969705425e-06, "loss": 0.0264, "step": 2867 }, { "epoch": 0.6525597269624573, "grad_norm": 1.8613288950042697, "learning_rate": 1.1982322970122485e-06, "loss": 0.066, "step": 2868 }, { "epoch": 0.6527872582480091, "grad_norm": 1.1933202650904684, "learning_rate": 1.1981966853381364e-06, "loss": 0.0553, "step": 2869 }, { "epoch": 0.6530147895335608, "grad_norm": 0.9553057216020039, "learning_rate": 1.1981610619489337e-06, "loss": 0.0391, "step": 2870 }, { "epoch": 0.6532423208191126, "grad_norm": 1.1836315703365476, "learning_rate": 1.1981254268453684e-06, "loss": 0.054, "step": 2871 }, { "epoch": 0.6534698521046644, "grad_norm": 1.276765336289362, "learning_rate": 1.1980897800281694e-06, "loss": 0.0585, "step": 2872 }, { "epoch": 0.6536973833902161, "grad_norm": 1.5829433588482489, "learning_rate": 1.1980541214980646e-06, "loss": 0.0811, "step": 2873 }, { "epoch": 0.653924914675768, "grad_norm": 2.2299790037798854, "learning_rate": 1.1980184512557833e-06, "loss": 0.0838, "step": 2874 }, { "epoch": 0.6541524459613197, "grad_norm": 1.8246918071788611, "learning_rate": 1.1979827693020541e-06, "loss": 0.0778, "step": 2875 }, { "epoch": 0.6543799772468715, "grad_norm": 0.8023113638584155, "learning_rate": 1.1979470756376064e-06, "loss": 0.0374, "step": 2876 }, { "epoch": 0.6546075085324232, "grad_norm": 1.0665996787354195, "learning_rate": 1.1979113702631697e-06, "loss": 0.0642, "step": 2877 }, { "epoch": 0.654835039817975, "grad_norm": 1.0607142907030562, "learning_rate": 1.197875653179474e-06, "loss": 0.0379, "step": 2878 }, { "epoch": 0.6550625711035267, "grad_norm": 1.509330774335647, "learning_rate": 1.1978399243872492e-06, "loss": 0.0429, "step": 2879 }, { "epoch": 0.6552901023890785, "grad_norm": 1.1080169650246614, "learning_rate": 1.1978041838872253e-06, "loss": 0.041, "step": 2880 }, { "epoch": 0.6555176336746302, "grad_norm": 0.8616722922724181, "learning_rate": 1.197768431680133e-06, "loss": 0.0308, "step": 2881 }, { "epoch": 0.655745164960182, "grad_norm": 1.625281473798501, "learning_rate": 1.197732667766703e-06, "loss": 0.0732, "step": 2882 }, { "epoch": 0.6559726962457337, "grad_norm": 2.0203110844795056, "learning_rate": 1.1976968921476662e-06, "loss": 0.0885, "step": 2883 }, { "epoch": 0.6562002275312856, "grad_norm": 1.325667144252201, "learning_rate": 1.1976611048237534e-06, "loss": 0.0471, "step": 2884 }, { "epoch": 0.6564277588168373, "grad_norm": 1.108094823287875, "learning_rate": 1.1976253057956968e-06, "loss": 0.0486, "step": 2885 }, { "epoch": 0.6566552901023891, "grad_norm": 1.5070679596895156, "learning_rate": 1.1975894950642276e-06, "loss": 0.0609, "step": 2886 }, { "epoch": 0.6568828213879409, "grad_norm": 1.8747406123784733, "learning_rate": 1.1975536726300776e-06, "loss": 0.084, "step": 2887 }, { "epoch": 0.6571103526734926, "grad_norm": 1.0968771480268504, "learning_rate": 1.1975178384939793e-06, "loss": 0.0518, "step": 2888 }, { "epoch": 0.6573378839590444, "grad_norm": 1.2149776100702219, "learning_rate": 1.197481992656665e-06, "loss": 0.0663, "step": 2889 }, { "epoch": 0.6575654152445961, "grad_norm": 1.0219972948941853, "learning_rate": 1.1974461351188668e-06, "loss": 0.0421, "step": 2890 }, { "epoch": 0.6577929465301479, "grad_norm": 1.0668851298436286, "learning_rate": 1.1974102658813183e-06, "loss": 0.0372, "step": 2891 }, { "epoch": 0.6580204778156996, "grad_norm": 0.9280161130235592, "learning_rate": 1.1973743849447522e-06, "loss": 0.041, "step": 2892 }, { "epoch": 0.6582480091012515, "grad_norm": 1.5811942617888033, "learning_rate": 1.197338492309902e-06, "loss": 0.0644, "step": 2893 }, { "epoch": 0.6584755403868032, "grad_norm": 0.9301441187543878, "learning_rate": 1.1973025879775011e-06, "loss": 0.0452, "step": 2894 }, { "epoch": 0.658703071672355, "grad_norm": 1.4309181679010228, "learning_rate": 1.1972666719482833e-06, "loss": 0.0594, "step": 2895 }, { "epoch": 0.6589306029579067, "grad_norm": 1.2896695199681876, "learning_rate": 1.197230744222983e-06, "loss": 0.0462, "step": 2896 }, { "epoch": 0.6591581342434585, "grad_norm": 1.3825152304918809, "learning_rate": 1.1971948048023343e-06, "loss": 0.0645, "step": 2897 }, { "epoch": 0.6593856655290102, "grad_norm": 1.5004332728601224, "learning_rate": 1.1971588536870717e-06, "loss": 0.0709, "step": 2898 }, { "epoch": 0.659613196814562, "grad_norm": 1.4884045373736343, "learning_rate": 1.19712289087793e-06, "loss": 0.0779, "step": 2899 }, { "epoch": 0.6598407281001137, "grad_norm": 1.8129968316744782, "learning_rate": 1.197086916375644e-06, "loss": 0.0848, "step": 2900 }, { "epoch": 0.6600682593856655, "grad_norm": 0.6798855950472126, "learning_rate": 1.1970509301809493e-06, "loss": 0.0182, "step": 2901 }, { "epoch": 0.6602957906712172, "grad_norm": 1.3632444561301862, "learning_rate": 1.1970149322945812e-06, "loss": 0.0885, "step": 2902 }, { "epoch": 0.6605233219567691, "grad_norm": 1.1160426689933205, "learning_rate": 1.1969789227172755e-06, "loss": 0.0467, "step": 2903 }, { "epoch": 0.6607508532423209, "grad_norm": 0.9677303090234676, "learning_rate": 1.1969429014497684e-06, "loss": 0.0257, "step": 2904 }, { "epoch": 0.6609783845278726, "grad_norm": 17.353739180619908, "learning_rate": 1.1969068684927954e-06, "loss": 0.0818, "step": 2905 }, { "epoch": 0.6612059158134244, "grad_norm": 1.2792498241135908, "learning_rate": 1.1968708238470936e-06, "loss": 0.058, "step": 2906 }, { "epoch": 0.6614334470989761, "grad_norm": 1.4555868612084297, "learning_rate": 1.1968347675133995e-06, "loss": 0.0562, "step": 2907 }, { "epoch": 0.6616609783845279, "grad_norm": 1.5729405842398636, "learning_rate": 1.19679869949245e-06, "loss": 0.0901, "step": 2908 }, { "epoch": 0.6618885096700796, "grad_norm": 2.205720994640489, "learning_rate": 1.1967626197849824e-06, "loss": 0.0712, "step": 2909 }, { "epoch": 0.6621160409556314, "grad_norm": 2.2706627078782478, "learning_rate": 1.1967265283917339e-06, "loss": 0.1286, "step": 2910 }, { "epoch": 0.6623435722411831, "grad_norm": 1.3583089823132908, "learning_rate": 1.1966904253134422e-06, "loss": 0.051, "step": 2911 }, { "epoch": 0.6625711035267349, "grad_norm": 1.2040227987956196, "learning_rate": 1.1966543105508454e-06, "loss": 0.0303, "step": 2912 }, { "epoch": 0.6627986348122867, "grad_norm": 1.8898241749294495, "learning_rate": 1.1966181841046812e-06, "loss": 0.1013, "step": 2913 }, { "epoch": 0.6630261660978385, "grad_norm": 1.1212373105622575, "learning_rate": 1.1965820459756882e-06, "loss": 0.0423, "step": 2914 }, { "epoch": 0.6632536973833902, "grad_norm": 1.4942787777207431, "learning_rate": 1.1965458961646051e-06, "loss": 0.074, "step": 2915 }, { "epoch": 0.663481228668942, "grad_norm": 1.2554067286772215, "learning_rate": 1.1965097346721707e-06, "loss": 0.0569, "step": 2916 }, { "epoch": 0.6637087599544937, "grad_norm": 2.064257552726771, "learning_rate": 1.1964735614991237e-06, "loss": 0.1222, "step": 2917 }, { "epoch": 0.6639362912400455, "grad_norm": 1.7127511151118642, "learning_rate": 1.196437376646204e-06, "loss": 0.059, "step": 2918 }, { "epoch": 0.6641638225255972, "grad_norm": 1.747734548367446, "learning_rate": 1.1964011801141505e-06, "loss": 0.0699, "step": 2919 }, { "epoch": 0.664391353811149, "grad_norm": 1.7260170721527037, "learning_rate": 1.1963649719037037e-06, "loss": 0.0682, "step": 2920 }, { "epoch": 0.6646188850967008, "grad_norm": 1.0412792846070889, "learning_rate": 1.196328752015603e-06, "loss": 0.0469, "step": 2921 }, { "epoch": 0.6648464163822526, "grad_norm": 1.0652258289806211, "learning_rate": 1.1962925204505894e-06, "loss": 0.0292, "step": 2922 }, { "epoch": 0.6650739476678044, "grad_norm": 1.05165136662707, "learning_rate": 1.1962562772094024e-06, "loss": 0.0707, "step": 2923 }, { "epoch": 0.6653014789533561, "grad_norm": 0.877541367497127, "learning_rate": 1.1962200222927836e-06, "loss": 0.0382, "step": 2924 }, { "epoch": 0.6655290102389079, "grad_norm": 1.0400570459306038, "learning_rate": 1.1961837557014736e-06, "loss": 0.0334, "step": 2925 }, { "epoch": 0.6657565415244596, "grad_norm": 1.6730636365537126, "learning_rate": 1.196147477436214e-06, "loss": 0.1133, "step": 2926 }, { "epoch": 0.6659840728100114, "grad_norm": 0.9388002444280849, "learning_rate": 1.1961111874977455e-06, "loss": 0.0425, "step": 2927 }, { "epoch": 0.6662116040955631, "grad_norm": 1.2971959782913498, "learning_rate": 1.1960748858868104e-06, "loss": 0.0598, "step": 2928 }, { "epoch": 0.6664391353811149, "grad_norm": 1.6257500523277735, "learning_rate": 1.1960385726041507e-06, "loss": 0.0773, "step": 2929 }, { "epoch": 0.6666666666666666, "grad_norm": 1.113728651042982, "learning_rate": 1.1960022476505082e-06, "loss": 0.0406, "step": 2930 }, { "epoch": 0.6668941979522184, "grad_norm": 0.8908601808254871, "learning_rate": 1.1959659110266256e-06, "loss": 0.0392, "step": 2931 }, { "epoch": 0.6671217292377702, "grad_norm": 1.6390262789672483, "learning_rate": 1.1959295627332454e-06, "loss": 0.0888, "step": 2932 }, { "epoch": 0.667349260523322, "grad_norm": 1.4346808601596435, "learning_rate": 1.1958932027711106e-06, "loss": 0.0534, "step": 2933 }, { "epoch": 0.6675767918088737, "grad_norm": 0.9348135708089949, "learning_rate": 1.1958568311409643e-06, "loss": 0.0472, "step": 2934 }, { "epoch": 0.6678043230944255, "grad_norm": 1.159822973816305, "learning_rate": 1.1958204478435497e-06, "loss": 0.0602, "step": 2935 }, { "epoch": 0.6680318543799773, "grad_norm": 1.0013241030005304, "learning_rate": 1.195784052879611e-06, "loss": 0.0347, "step": 2936 }, { "epoch": 0.668259385665529, "grad_norm": 0.7712539814391252, "learning_rate": 1.195747646249891e-06, "loss": 0.0238, "step": 2937 }, { "epoch": 0.6684869169510808, "grad_norm": 1.8913673239815798, "learning_rate": 1.1957112279551347e-06, "loss": 0.0967, "step": 2938 }, { "epoch": 0.6687144482366325, "grad_norm": 1.3385349237767592, "learning_rate": 1.195674797996086e-06, "loss": 0.0452, "step": 2939 }, { "epoch": 0.6689419795221843, "grad_norm": 1.8314127610766537, "learning_rate": 1.1956383563734897e-06, "loss": 0.0888, "step": 2940 }, { "epoch": 0.669169510807736, "grad_norm": 1.4820883842924408, "learning_rate": 1.1956019030880902e-06, "loss": 0.0629, "step": 2941 }, { "epoch": 0.6693970420932879, "grad_norm": 0.9377075991001586, "learning_rate": 1.1955654381406331e-06, "loss": 0.0541, "step": 2942 }, { "epoch": 0.6696245733788396, "grad_norm": 1.1625402940924978, "learning_rate": 1.1955289615318632e-06, "loss": 0.0583, "step": 2943 }, { "epoch": 0.6698521046643914, "grad_norm": 0.9043356889096931, "learning_rate": 1.1954924732625264e-06, "loss": 0.0386, "step": 2944 }, { "epoch": 0.6700796359499431, "grad_norm": 1.0364482997261704, "learning_rate": 1.1954559733333681e-06, "loss": 0.0353, "step": 2945 }, { "epoch": 0.6703071672354949, "grad_norm": 1.6526326184491673, "learning_rate": 1.1954194617451345e-06, "loss": 0.0531, "step": 2946 }, { "epoch": 0.6705346985210466, "grad_norm": 0.9766619653449208, "learning_rate": 1.1953829384985716e-06, "loss": 0.0391, "step": 2947 }, { "epoch": 0.6707622298065984, "grad_norm": 2.163216907401221, "learning_rate": 1.1953464035944262e-06, "loss": 0.0835, "step": 2948 }, { "epoch": 0.6709897610921501, "grad_norm": 1.4844803459994427, "learning_rate": 1.195309857033445e-06, "loss": 0.0842, "step": 2949 }, { "epoch": 0.6712172923777019, "grad_norm": 0.6644024917990391, "learning_rate": 1.1952732988163745e-06, "loss": 0.0347, "step": 2950 }, { "epoch": 0.6714448236632536, "grad_norm": 1.3392424005542798, "learning_rate": 1.1952367289439624e-06, "loss": 0.0593, "step": 2951 }, { "epoch": 0.6716723549488055, "grad_norm": 1.5693502580306737, "learning_rate": 1.1952001474169558e-06, "loss": 0.0723, "step": 2952 }, { "epoch": 0.6718998862343573, "grad_norm": 1.3882537221447417, "learning_rate": 1.1951635542361025e-06, "loss": 0.0482, "step": 2953 }, { "epoch": 0.672127417519909, "grad_norm": 1.6367400788566016, "learning_rate": 1.1951269494021503e-06, "loss": 0.0541, "step": 2954 }, { "epoch": 0.6723549488054608, "grad_norm": 1.2264958404782558, "learning_rate": 1.1950903329158475e-06, "loss": 0.0543, "step": 2955 }, { "epoch": 0.6725824800910125, "grad_norm": 0.9638518358259208, "learning_rate": 1.1950537047779424e-06, "loss": 0.0425, "step": 2956 }, { "epoch": 0.6728100113765643, "grad_norm": 0.7797808668666366, "learning_rate": 1.1950170649891836e-06, "loss": 0.0266, "step": 2957 }, { "epoch": 0.673037542662116, "grad_norm": 1.256662868306741, "learning_rate": 1.1949804135503196e-06, "loss": 0.0762, "step": 2958 }, { "epoch": 0.6732650739476678, "grad_norm": 1.0816193366991107, "learning_rate": 1.1949437504621e-06, "loss": 0.047, "step": 2959 }, { "epoch": 0.6734926052332195, "grad_norm": 1.5164857137673318, "learning_rate": 1.194907075725274e-06, "loss": 0.0995, "step": 2960 }, { "epoch": 0.6737201365187714, "grad_norm": 1.1360766280459287, "learning_rate": 1.1948703893405911e-06, "loss": 0.0427, "step": 2961 }, { "epoch": 0.6739476678043231, "grad_norm": 1.342451075159084, "learning_rate": 1.194833691308801e-06, "loss": 0.0422, "step": 2962 }, { "epoch": 0.6741751990898749, "grad_norm": 1.7047601826121286, "learning_rate": 1.194796981630654e-06, "loss": 0.0532, "step": 2963 }, { "epoch": 0.6744027303754266, "grad_norm": 1.3289514422481006, "learning_rate": 1.1947602603069002e-06, "loss": 0.0648, "step": 2964 }, { "epoch": 0.6746302616609784, "grad_norm": 1.8203278250275807, "learning_rate": 1.19472352733829e-06, "loss": 0.0591, "step": 2965 }, { "epoch": 0.6748577929465301, "grad_norm": 0.9130861662390535, "learning_rate": 1.1946867827255744e-06, "loss": 0.0392, "step": 2966 }, { "epoch": 0.6750853242320819, "grad_norm": 2.1792488178961866, "learning_rate": 1.1946500264695044e-06, "loss": 0.0852, "step": 2967 }, { "epoch": 0.6753128555176336, "grad_norm": 1.3430961956554572, "learning_rate": 1.194613258570831e-06, "loss": 0.054, "step": 2968 }, { "epoch": 0.6755403868031854, "grad_norm": 1.0194337724675724, "learning_rate": 1.1945764790303059e-06, "loss": 0.0577, "step": 2969 }, { "epoch": 0.6757679180887372, "grad_norm": 1.414677659719635, "learning_rate": 1.1945396878486805e-06, "loss": 0.0575, "step": 2970 }, { "epoch": 0.675995449374289, "grad_norm": 1.3881918329356409, "learning_rate": 1.194502885026707e-06, "loss": 0.07, "step": 2971 }, { "epoch": 0.6762229806598408, "grad_norm": 0.8742447812198808, "learning_rate": 1.1944660705651375e-06, "loss": 0.0545, "step": 2972 }, { "epoch": 0.6764505119453925, "grad_norm": 0.9000949345293963, "learning_rate": 1.1944292444647248e-06, "loss": 0.0465, "step": 2973 }, { "epoch": 0.6766780432309443, "grad_norm": 1.0764234435889581, "learning_rate": 1.1943924067262208e-06, "loss": 0.0432, "step": 2974 }, { "epoch": 0.676905574516496, "grad_norm": 1.249115475491158, "learning_rate": 1.194355557350379e-06, "loss": 0.0548, "step": 2975 }, { "epoch": 0.6771331058020478, "grad_norm": 1.067373727393787, "learning_rate": 1.1943186963379522e-06, "loss": 0.0528, "step": 2976 }, { "epoch": 0.6773606370875995, "grad_norm": 1.5676460836455215, "learning_rate": 1.194281823689694e-06, "loss": 0.0672, "step": 2977 }, { "epoch": 0.6775881683731513, "grad_norm": 1.334005489011124, "learning_rate": 1.194244939406358e-06, "loss": 0.0565, "step": 2978 }, { "epoch": 0.677815699658703, "grad_norm": 1.324440667565256, "learning_rate": 1.1942080434886978e-06, "loss": 0.0488, "step": 2979 }, { "epoch": 0.6780432309442548, "grad_norm": 1.155754728183158, "learning_rate": 1.1941711359374678e-06, "loss": 0.0622, "step": 2980 }, { "epoch": 0.6782707622298066, "grad_norm": 0.7220306306891887, "learning_rate": 1.194134216753422e-06, "loss": 0.0363, "step": 2981 }, { "epoch": 0.6784982935153584, "grad_norm": 2.259664003454956, "learning_rate": 1.1940972859373151e-06, "loss": 0.084, "step": 2982 }, { "epoch": 0.6787258248009101, "grad_norm": 0.9209026025756932, "learning_rate": 1.194060343489902e-06, "loss": 0.0247, "step": 2983 }, { "epoch": 0.6789533560864619, "grad_norm": 2.237931268641456, "learning_rate": 1.1940233894119377e-06, "loss": 0.1003, "step": 2984 }, { "epoch": 0.6791808873720137, "grad_norm": 1.45745994111441, "learning_rate": 1.1939864237041774e-06, "loss": 0.0656, "step": 2985 }, { "epoch": 0.6794084186575654, "grad_norm": 0.9550126427103799, "learning_rate": 1.1939494463673767e-06, "loss": 0.0442, "step": 2986 }, { "epoch": 0.6796359499431172, "grad_norm": 1.9833777656739968, "learning_rate": 1.1939124574022914e-06, "loss": 0.0904, "step": 2987 }, { "epoch": 0.6798634812286689, "grad_norm": 1.3661977813438515, "learning_rate": 1.1938754568096771e-06, "loss": 0.0644, "step": 2988 }, { "epoch": 0.6800910125142207, "grad_norm": 0.8616871036933312, "learning_rate": 1.1938384445902905e-06, "loss": 0.0393, "step": 2989 }, { "epoch": 0.6803185437997725, "grad_norm": 1.1588349083327987, "learning_rate": 1.1938014207448877e-06, "loss": 0.047, "step": 2990 }, { "epoch": 0.6805460750853243, "grad_norm": 1.0841199268841493, "learning_rate": 1.1937643852742258e-06, "loss": 0.0315, "step": 2991 }, { "epoch": 0.680773606370876, "grad_norm": 4.593658878524584, "learning_rate": 1.1937273381790615e-06, "loss": 0.1461, "step": 2992 }, { "epoch": 0.6810011376564278, "grad_norm": 1.1490732435329922, "learning_rate": 1.1936902794601518e-06, "loss": 0.0479, "step": 2993 }, { "epoch": 0.6812286689419795, "grad_norm": 1.450896272579947, "learning_rate": 1.1936532091182544e-06, "loss": 0.0613, "step": 2994 }, { "epoch": 0.6814562002275313, "grad_norm": 1.2205030830111818, "learning_rate": 1.1936161271541268e-06, "loss": 0.0626, "step": 2995 }, { "epoch": 0.681683731513083, "grad_norm": 1.1653942327121265, "learning_rate": 1.1935790335685272e-06, "loss": 0.0464, "step": 2996 }, { "epoch": 0.6819112627986348, "grad_norm": 1.9153169444858846, "learning_rate": 1.193541928362213e-06, "loss": 0.0732, "step": 2997 }, { "epoch": 0.6821387940841865, "grad_norm": 1.4815686264003924, "learning_rate": 1.1935048115359432e-06, "loss": 0.0638, "step": 2998 }, { "epoch": 0.6823663253697383, "grad_norm": 1.6915818417968453, "learning_rate": 1.1934676830904763e-06, "loss": 0.0773, "step": 2999 }, { "epoch": 0.6825938566552902, "grad_norm": 1.3182256509758368, "learning_rate": 1.193430543026571e-06, "loss": 0.0627, "step": 3000 }, { "epoch": 0.6828213879408419, "grad_norm": 1.6374773596627972, "learning_rate": 1.1933933913449867e-06, "loss": 0.1202, "step": 3001 }, { "epoch": 0.6830489192263937, "grad_norm": 1.6225368015279966, "learning_rate": 1.193356228046482e-06, "loss": 0.1016, "step": 3002 }, { "epoch": 0.6832764505119454, "grad_norm": 1.158897733977453, "learning_rate": 1.1933190531318172e-06, "loss": 0.0455, "step": 3003 }, { "epoch": 0.6835039817974972, "grad_norm": 1.2180725170363247, "learning_rate": 1.1932818666017516e-06, "loss": 0.0335, "step": 3004 }, { "epoch": 0.6837315130830489, "grad_norm": 1.5088158398690277, "learning_rate": 1.1932446684570455e-06, "loss": 0.0478, "step": 3005 }, { "epoch": 0.6839590443686007, "grad_norm": 1.0577989164533361, "learning_rate": 1.1932074586984592e-06, "loss": 0.0411, "step": 3006 }, { "epoch": 0.6841865756541524, "grad_norm": 1.265620028127131, "learning_rate": 1.1931702373267527e-06, "loss": 0.0373, "step": 3007 }, { "epoch": 0.6844141069397042, "grad_norm": 1.015589756317499, "learning_rate": 1.1931330043426872e-06, "loss": 0.0701, "step": 3008 }, { "epoch": 0.6846416382252559, "grad_norm": 0.9468195016052887, "learning_rate": 1.1930957597470238e-06, "loss": 0.0502, "step": 3009 }, { "epoch": 0.6848691695108078, "grad_norm": 1.4788516581774798, "learning_rate": 1.1930585035405235e-06, "loss": 0.0579, "step": 3010 }, { "epoch": 0.6850967007963595, "grad_norm": 1.5755558972332642, "learning_rate": 1.1930212357239475e-06, "loss": 0.0525, "step": 3011 }, { "epoch": 0.6853242320819113, "grad_norm": 1.2998868209788172, "learning_rate": 1.1929839562980579e-06, "loss": 0.0585, "step": 3012 }, { "epoch": 0.685551763367463, "grad_norm": 1.5356635666634295, "learning_rate": 1.1929466652636164e-06, "loss": 0.0604, "step": 3013 }, { "epoch": 0.6857792946530148, "grad_norm": 2.0893811231577524, "learning_rate": 1.1929093626213852e-06, "loss": 0.1015, "step": 3014 }, { "epoch": 0.6860068259385665, "grad_norm": 0.5321529547637093, "learning_rate": 1.1928720483721269e-06, "loss": 0.0139, "step": 3015 }, { "epoch": 0.6862343572241183, "grad_norm": 1.0580052416688368, "learning_rate": 1.1928347225166035e-06, "loss": 0.0499, "step": 3016 }, { "epoch": 0.68646188850967, "grad_norm": 1.3320645495655943, "learning_rate": 1.1927973850555785e-06, "loss": 0.0465, "step": 3017 }, { "epoch": 0.6866894197952218, "grad_norm": 1.3194332531036985, "learning_rate": 1.192760035989815e-06, "loss": 0.0489, "step": 3018 }, { "epoch": 0.6869169510807736, "grad_norm": 1.7873894152489311, "learning_rate": 1.192722675320076e-06, "loss": 0.0683, "step": 3019 }, { "epoch": 0.6871444823663254, "grad_norm": 1.533578323181438, "learning_rate": 1.1926853030471253e-06, "loss": 0.0536, "step": 3020 }, { "epoch": 0.6873720136518772, "grad_norm": 0.9261434578357977, "learning_rate": 1.1926479191717267e-06, "loss": 0.0437, "step": 3021 }, { "epoch": 0.6875995449374289, "grad_norm": 1.1444794832316916, "learning_rate": 1.1926105236946443e-06, "loss": 0.0577, "step": 3022 }, { "epoch": 0.6878270762229807, "grad_norm": 1.335750162762178, "learning_rate": 1.192573116616642e-06, "loss": 0.0529, "step": 3023 }, { "epoch": 0.6880546075085324, "grad_norm": 1.434078699306857, "learning_rate": 1.192535697938485e-06, "loss": 0.0529, "step": 3024 }, { "epoch": 0.6882821387940842, "grad_norm": 1.3416490998400092, "learning_rate": 1.1924982676609377e-06, "loss": 0.074, "step": 3025 }, { "epoch": 0.6885096700796359, "grad_norm": 1.3232433001494412, "learning_rate": 1.1924608257847651e-06, "loss": 0.0451, "step": 3026 }, { "epoch": 0.6887372013651877, "grad_norm": 0.8764669981180951, "learning_rate": 1.1924233723107322e-06, "loss": 0.0484, "step": 3027 }, { "epoch": 0.6889647326507394, "grad_norm": 1.1655688024633921, "learning_rate": 1.1923859072396051e-06, "loss": 0.0606, "step": 3028 }, { "epoch": 0.6891922639362913, "grad_norm": 1.0179281217422282, "learning_rate": 1.1923484305721489e-06, "loss": 0.0466, "step": 3029 }, { "epoch": 0.689419795221843, "grad_norm": 0.8659456244009691, "learning_rate": 1.19231094230913e-06, "loss": 0.0404, "step": 3030 }, { "epoch": 0.6896473265073948, "grad_norm": 2.0636576063861565, "learning_rate": 1.1922734424513144e-06, "loss": 0.0784, "step": 3031 }, { "epoch": 0.6898748577929465, "grad_norm": 0.8695263646618282, "learning_rate": 1.1922359309994685e-06, "loss": 0.0277, "step": 3032 }, { "epoch": 0.6901023890784983, "grad_norm": 1.4190056652396834, "learning_rate": 1.192198407954359e-06, "loss": 0.0616, "step": 3033 }, { "epoch": 0.69032992036405, "grad_norm": 1.4753814874094142, "learning_rate": 1.192160873316753e-06, "loss": 0.0517, "step": 3034 }, { "epoch": 0.6905574516496018, "grad_norm": 1.5554050029090873, "learning_rate": 1.1921233270874174e-06, "loss": 0.0563, "step": 3035 }, { "epoch": 0.6907849829351536, "grad_norm": 0.7022912796986149, "learning_rate": 1.1920857692671196e-06, "loss": 0.0312, "step": 3036 }, { "epoch": 0.6910125142207053, "grad_norm": 2.0009143095316646, "learning_rate": 1.192048199856627e-06, "loss": 0.0994, "step": 3037 }, { "epoch": 0.6912400455062571, "grad_norm": 2.134476815905357, "learning_rate": 1.192010618856708e-06, "loss": 0.1024, "step": 3038 }, { "epoch": 0.6914675767918089, "grad_norm": 1.1849138073132004, "learning_rate": 1.1919730262681304e-06, "loss": 0.055, "step": 3039 }, { "epoch": 0.6916951080773607, "grad_norm": 1.845998643453397, "learning_rate": 1.1919354220916624e-06, "loss": 0.0983, "step": 3040 }, { "epoch": 0.6919226393629124, "grad_norm": 2.449072929938712, "learning_rate": 1.1918978063280726e-06, "loss": 0.0849, "step": 3041 }, { "epoch": 0.6921501706484642, "grad_norm": 0.8729079689297836, "learning_rate": 1.1918601789781299e-06, "loss": 0.0306, "step": 3042 }, { "epoch": 0.6923777019340159, "grad_norm": 1.780329225544261, "learning_rate": 1.1918225400426032e-06, "loss": 0.0689, "step": 3043 }, { "epoch": 0.6926052332195677, "grad_norm": 1.063152792810893, "learning_rate": 1.191784889522262e-06, "loss": 0.0408, "step": 3044 }, { "epoch": 0.6928327645051194, "grad_norm": 1.6901254212789465, "learning_rate": 1.1917472274178757e-06, "loss": 0.0846, "step": 3045 }, { "epoch": 0.6930602957906712, "grad_norm": 1.4002132139198793, "learning_rate": 1.191709553730214e-06, "loss": 0.0772, "step": 3046 }, { "epoch": 0.6932878270762229, "grad_norm": 1.5076734596253167, "learning_rate": 1.1916718684600469e-06, "loss": 0.0819, "step": 3047 }, { "epoch": 0.6935153583617747, "grad_norm": 1.0960152159965912, "learning_rate": 1.1916341716081446e-06, "loss": 0.049, "step": 3048 }, { "epoch": 0.6937428896473266, "grad_norm": 2.557391206512559, "learning_rate": 1.1915964631752775e-06, "loss": 0.0876, "step": 3049 }, { "epoch": 0.6939704209328783, "grad_norm": 0.920645510461067, "learning_rate": 1.1915587431622164e-06, "loss": 0.0307, "step": 3050 }, { "epoch": 0.6941979522184301, "grad_norm": 1.04173789253993, "learning_rate": 1.1915210115697324e-06, "loss": 0.0539, "step": 3051 }, { "epoch": 0.6944254835039818, "grad_norm": 1.6810913949726196, "learning_rate": 1.1914832683985962e-06, "loss": 0.0544, "step": 3052 }, { "epoch": 0.6946530147895336, "grad_norm": 1.850644903401837, "learning_rate": 1.1914455136495796e-06, "loss": 0.0899, "step": 3053 }, { "epoch": 0.6948805460750853, "grad_norm": 0.7506486967154027, "learning_rate": 1.191407747323454e-06, "loss": 0.0265, "step": 3054 }, { "epoch": 0.6951080773606371, "grad_norm": 2.1012986938124696, "learning_rate": 1.1913699694209914e-06, "loss": 0.1144, "step": 3055 }, { "epoch": 0.6953356086461888, "grad_norm": 1.2089722077198737, "learning_rate": 1.191332179942964e-06, "loss": 0.0609, "step": 3056 }, { "epoch": 0.6955631399317406, "grad_norm": 1.5324615273992814, "learning_rate": 1.1912943788901438e-06, "loss": 0.067, "step": 3057 }, { "epoch": 0.6957906712172923, "grad_norm": 1.4377403085013738, "learning_rate": 1.191256566263304e-06, "loss": 0.0841, "step": 3058 }, { "epoch": 0.6960182025028442, "grad_norm": 1.7028519956627337, "learning_rate": 1.1912187420632165e-06, "loss": 0.0796, "step": 3059 }, { "epoch": 0.6962457337883959, "grad_norm": 1.535145639508393, "learning_rate": 1.1911809062906552e-06, "loss": 0.0581, "step": 3060 }, { "epoch": 0.6964732650739477, "grad_norm": 1.3876603600761275, "learning_rate": 1.1911430589463931e-06, "loss": 0.0548, "step": 3061 }, { "epoch": 0.6967007963594994, "grad_norm": 2.3585378108380057, "learning_rate": 1.1911052000312038e-06, "loss": 0.1131, "step": 3062 }, { "epoch": 0.6969283276450512, "grad_norm": 1.224179370923149, "learning_rate": 1.1910673295458607e-06, "loss": 0.0733, "step": 3063 }, { "epoch": 0.697155858930603, "grad_norm": 1.080344689666754, "learning_rate": 1.1910294474911382e-06, "loss": 0.0339, "step": 3064 }, { "epoch": 0.6973833902161547, "grad_norm": 1.132793254903834, "learning_rate": 1.1909915538678105e-06, "loss": 0.0668, "step": 3065 }, { "epoch": 0.6976109215017064, "grad_norm": 0.9670480449446429, "learning_rate": 1.1909536486766522e-06, "loss": 0.0613, "step": 3066 }, { "epoch": 0.6978384527872582, "grad_norm": 0.8333929861107748, "learning_rate": 1.1909157319184373e-06, "loss": 0.0313, "step": 3067 }, { "epoch": 0.6980659840728101, "grad_norm": 1.5088053682741849, "learning_rate": 1.1908778035939416e-06, "loss": 0.0723, "step": 3068 }, { "epoch": 0.6982935153583618, "grad_norm": 2.496852043007627, "learning_rate": 1.1908398637039398e-06, "loss": 0.1006, "step": 3069 }, { "epoch": 0.6985210466439136, "grad_norm": 1.340716909671611, "learning_rate": 1.1908019122492077e-06, "loss": 0.0447, "step": 3070 }, { "epoch": 0.6987485779294653, "grad_norm": 2.0378089803817887, "learning_rate": 1.1907639492305205e-06, "loss": 0.0808, "step": 3071 }, { "epoch": 0.6989761092150171, "grad_norm": 0.7800931599177211, "learning_rate": 1.1907259746486547e-06, "loss": 0.0358, "step": 3072 }, { "epoch": 0.6992036405005688, "grad_norm": 0.9202110065520366, "learning_rate": 1.1906879885043856e-06, "loss": 0.0571, "step": 3073 }, { "epoch": 0.6994311717861206, "grad_norm": 0.7009452078498103, "learning_rate": 1.1906499907984903e-06, "loss": 0.0318, "step": 3074 }, { "epoch": 0.6996587030716723, "grad_norm": 1.1314308444671177, "learning_rate": 1.190611981531745e-06, "loss": 0.0446, "step": 3075 }, { "epoch": 0.6998862343572241, "grad_norm": 1.332217799137746, "learning_rate": 1.1905739607049267e-06, "loss": 0.0744, "step": 3076 }, { "epoch": 0.7001137656427758, "grad_norm": 0.9698337291959647, "learning_rate": 1.1905359283188126e-06, "loss": 0.0368, "step": 3077 }, { "epoch": 0.7003412969283277, "grad_norm": 0.5221877787168306, "learning_rate": 1.1904978843741796e-06, "loss": 0.0167, "step": 3078 }, { "epoch": 0.7005688282138794, "grad_norm": 1.4498948354718102, "learning_rate": 1.1904598288718055e-06, "loss": 0.0727, "step": 3079 }, { "epoch": 0.7007963594994312, "grad_norm": 1.5160712367792761, "learning_rate": 1.1904217618124684e-06, "loss": 0.0787, "step": 3080 }, { "epoch": 0.701023890784983, "grad_norm": 1.0604730462463068, "learning_rate": 1.1903836831969458e-06, "loss": 0.0639, "step": 3081 }, { "epoch": 0.7012514220705347, "grad_norm": 0.7275851025633676, "learning_rate": 1.190345593026016e-06, "loss": 0.0286, "step": 3082 }, { "epoch": 0.7014789533560865, "grad_norm": 1.4202168912069701, "learning_rate": 1.1903074913004579e-06, "loss": 0.0717, "step": 3083 }, { "epoch": 0.7017064846416382, "grad_norm": 1.6514017164531183, "learning_rate": 1.19026937802105e-06, "loss": 0.0787, "step": 3084 }, { "epoch": 0.70193401592719, "grad_norm": 0.6956539774170061, "learning_rate": 1.1902312531885712e-06, "loss": 0.0252, "step": 3085 }, { "epoch": 0.7021615472127417, "grad_norm": 1.5569684152870062, "learning_rate": 1.1901931168038007e-06, "loss": 0.0685, "step": 3086 }, { "epoch": 0.7023890784982935, "grad_norm": 1.2406976952968178, "learning_rate": 1.190154968867518e-06, "loss": 0.0454, "step": 3087 }, { "epoch": 0.7026166097838453, "grad_norm": 0.813921596018141, "learning_rate": 1.190116809380503e-06, "loss": 0.0253, "step": 3088 }, { "epoch": 0.7028441410693971, "grad_norm": 1.0275424113403453, "learning_rate": 1.190078638343535e-06, "loss": 0.0514, "step": 3089 }, { "epoch": 0.7030716723549488, "grad_norm": 1.0727110658094747, "learning_rate": 1.1900404557573948e-06, "loss": 0.0453, "step": 3090 }, { "epoch": 0.7032992036405006, "grad_norm": 1.6426665504516007, "learning_rate": 1.1900022616228624e-06, "loss": 0.0832, "step": 3091 }, { "epoch": 0.7035267349260523, "grad_norm": 1.9180762135067262, "learning_rate": 1.1899640559407186e-06, "loss": 0.0934, "step": 3092 }, { "epoch": 0.7037542662116041, "grad_norm": 1.7251637250688259, "learning_rate": 1.189925838711744e-06, "loss": 0.0744, "step": 3093 }, { "epoch": 0.7039817974971558, "grad_norm": 2.301009331337415, "learning_rate": 1.18988760993672e-06, "loss": 0.1061, "step": 3094 }, { "epoch": 0.7042093287827076, "grad_norm": 0.8730951635346749, "learning_rate": 1.1898493696164279e-06, "loss": 0.0251, "step": 3095 }, { "epoch": 0.7044368600682593, "grad_norm": 0.6226893588047131, "learning_rate": 1.1898111177516488e-06, "loss": 0.0396, "step": 3096 }, { "epoch": 0.7046643913538112, "grad_norm": 2.39998461661108, "learning_rate": 1.1897728543431653e-06, "loss": 0.1219, "step": 3097 }, { "epoch": 0.704891922639363, "grad_norm": 1.834814384170055, "learning_rate": 1.1897345793917589e-06, "loss": 0.0615, "step": 3098 }, { "epoch": 0.7051194539249147, "grad_norm": 1.0235189868055947, "learning_rate": 1.1896962928982116e-06, "loss": 0.0256, "step": 3099 }, { "epoch": 0.7053469852104665, "grad_norm": 1.084986587012269, "learning_rate": 1.1896579948633067e-06, "loss": 0.0839, "step": 3100 }, { "epoch": 0.7055745164960182, "grad_norm": 1.0758159672428005, "learning_rate": 1.1896196852878262e-06, "loss": 0.0554, "step": 3101 }, { "epoch": 0.70580204778157, "grad_norm": 1.653037063661443, "learning_rate": 1.1895813641725535e-06, "loss": 0.0592, "step": 3102 }, { "epoch": 0.7060295790671217, "grad_norm": 0.888432730316198, "learning_rate": 1.1895430315182719e-06, "loss": 0.0346, "step": 3103 }, { "epoch": 0.7062571103526735, "grad_norm": 0.7144453665457149, "learning_rate": 1.1895046873257644e-06, "loss": 0.0339, "step": 3104 }, { "epoch": 0.7064846416382252, "grad_norm": 1.1974179578050035, "learning_rate": 1.189466331595815e-06, "loss": 0.0584, "step": 3105 }, { "epoch": 0.706712172923777, "grad_norm": 2.0074894159842804, "learning_rate": 1.1894279643292074e-06, "loss": 0.0702, "step": 3106 }, { "epoch": 0.7069397042093288, "grad_norm": 1.1258269572183623, "learning_rate": 1.1893895855267262e-06, "loss": 0.0483, "step": 3107 }, { "epoch": 0.7071672354948806, "grad_norm": 1.8909664822572771, "learning_rate": 1.1893511951891553e-06, "loss": 0.0426, "step": 3108 }, { "epoch": 0.7073947667804323, "grad_norm": 0.9858584509738625, "learning_rate": 1.1893127933172794e-06, "loss": 0.0358, "step": 3109 }, { "epoch": 0.7076222980659841, "grad_norm": 1.6798238240138743, "learning_rate": 1.1892743799118838e-06, "loss": 0.0926, "step": 3110 }, { "epoch": 0.7078498293515358, "grad_norm": 1.8662422210150158, "learning_rate": 1.189235954973753e-06, "loss": 0.0913, "step": 3111 }, { "epoch": 0.7080773606370876, "grad_norm": 1.2592951116026574, "learning_rate": 1.189197518503673e-06, "loss": 0.0834, "step": 3112 }, { "epoch": 0.7083048919226393, "grad_norm": 2.2263635127158707, "learning_rate": 1.1891590705024288e-06, "loss": 0.1094, "step": 3113 }, { "epoch": 0.7085324232081911, "grad_norm": 1.2398530882425465, "learning_rate": 1.1891206109708065e-06, "loss": 0.0634, "step": 3114 }, { "epoch": 0.7087599544937428, "grad_norm": 0.8888283163380168, "learning_rate": 1.1890821399095917e-06, "loss": 0.0518, "step": 3115 }, { "epoch": 0.7089874857792946, "grad_norm": 1.7877536696152463, "learning_rate": 1.1890436573195714e-06, "loss": 0.0799, "step": 3116 }, { "epoch": 0.7092150170648465, "grad_norm": 0.6507653429669747, "learning_rate": 1.1890051632015315e-06, "loss": 0.0177, "step": 3117 }, { "epoch": 0.7094425483503982, "grad_norm": 0.810817143611541, "learning_rate": 1.1889666575562593e-06, "loss": 0.0294, "step": 3118 }, { "epoch": 0.70967007963595, "grad_norm": 1.351260714026528, "learning_rate": 1.1889281403845413e-06, "loss": 0.0764, "step": 3119 }, { "epoch": 0.7098976109215017, "grad_norm": 0.9662989115635772, "learning_rate": 1.1888896116871649e-06, "loss": 0.0592, "step": 3120 }, { "epoch": 0.7101251422070535, "grad_norm": 1.3131475158616286, "learning_rate": 1.1888510714649176e-06, "loss": 0.0485, "step": 3121 }, { "epoch": 0.7103526734926052, "grad_norm": 1.4877436902230308, "learning_rate": 1.1888125197185867e-06, "loss": 0.0786, "step": 3122 }, { "epoch": 0.710580204778157, "grad_norm": 0.9404627141658739, "learning_rate": 1.1887739564489608e-06, "loss": 0.0363, "step": 3123 }, { "epoch": 0.7108077360637087, "grad_norm": 1.664364740849866, "learning_rate": 1.1887353816568277e-06, "loss": 0.0836, "step": 3124 }, { "epoch": 0.7110352673492605, "grad_norm": 1.1581179507287511, "learning_rate": 1.188696795342976e-06, "loss": 0.0447, "step": 3125 }, { "epoch": 0.7112627986348122, "grad_norm": 1.4278480990598597, "learning_rate": 1.188658197508194e-06, "loss": 0.0702, "step": 3126 }, { "epoch": 0.7114903299203641, "grad_norm": 1.8166789381335957, "learning_rate": 1.1886195881532705e-06, "loss": 0.0756, "step": 3127 }, { "epoch": 0.7117178612059158, "grad_norm": 1.0532098621527515, "learning_rate": 1.1885809672789953e-06, "loss": 0.0427, "step": 3128 }, { "epoch": 0.7119453924914676, "grad_norm": 1.1496591085120342, "learning_rate": 1.188542334886157e-06, "loss": 0.0345, "step": 3129 }, { "epoch": 0.7121729237770194, "grad_norm": 1.3871333419583287, "learning_rate": 1.1885036909755454e-06, "loss": 0.0591, "step": 3130 }, { "epoch": 0.7124004550625711, "grad_norm": 1.4115338701891487, "learning_rate": 1.1884650355479505e-06, "loss": 0.0824, "step": 3131 }, { "epoch": 0.7126279863481229, "grad_norm": 1.4867197919013324, "learning_rate": 1.1884263686041622e-06, "loss": 0.0793, "step": 3132 }, { "epoch": 0.7128555176336746, "grad_norm": 1.2824607789793006, "learning_rate": 1.1883876901449707e-06, "loss": 0.0451, "step": 3133 }, { "epoch": 0.7130830489192264, "grad_norm": 0.6708570443551615, "learning_rate": 1.1883490001711667e-06, "loss": 0.0219, "step": 3134 }, { "epoch": 0.7133105802047781, "grad_norm": 2.015395007692978, "learning_rate": 1.1883102986835408e-06, "loss": 0.0757, "step": 3135 }, { "epoch": 0.71353811149033, "grad_norm": 1.6702641230416622, "learning_rate": 1.1882715856828842e-06, "loss": 0.075, "step": 3136 }, { "epoch": 0.7137656427758817, "grad_norm": 0.912883335407981, "learning_rate": 1.1882328611699879e-06, "loss": 0.0388, "step": 3137 }, { "epoch": 0.7139931740614335, "grad_norm": 1.9915888368964108, "learning_rate": 1.1881941251456434e-06, "loss": 0.0736, "step": 3138 }, { "epoch": 0.7142207053469852, "grad_norm": 1.432753078445199, "learning_rate": 1.1881553776106423e-06, "loss": 0.0686, "step": 3139 }, { "epoch": 0.714448236632537, "grad_norm": 1.8572484999557712, "learning_rate": 1.1881166185657765e-06, "loss": 0.0639, "step": 3140 }, { "epoch": 0.7146757679180887, "grad_norm": 1.1698682146531856, "learning_rate": 1.1880778480118388e-06, "loss": 0.0506, "step": 3141 }, { "epoch": 0.7149032992036405, "grad_norm": 1.4247011889264531, "learning_rate": 1.1880390659496207e-06, "loss": 0.0688, "step": 3142 }, { "epoch": 0.7151308304891922, "grad_norm": 1.1897422778029059, "learning_rate": 1.1880002723799155e-06, "loss": 0.0488, "step": 3143 }, { "epoch": 0.715358361774744, "grad_norm": 0.7067410656877349, "learning_rate": 1.1879614673035158e-06, "loss": 0.0316, "step": 3144 }, { "epoch": 0.7155858930602957, "grad_norm": 1.1073184850973958, "learning_rate": 1.1879226507212146e-06, "loss": 0.04, "step": 3145 }, { "epoch": 0.7158134243458476, "grad_norm": 1.551329079095373, "learning_rate": 1.1878838226338054e-06, "loss": 0.0449, "step": 3146 }, { "epoch": 0.7160409556313994, "grad_norm": 1.2160812373705194, "learning_rate": 1.187844983042082e-06, "loss": 0.0547, "step": 3147 }, { "epoch": 0.7162684869169511, "grad_norm": 1.463195276621994, "learning_rate": 1.1878061319468376e-06, "loss": 0.0604, "step": 3148 }, { "epoch": 0.7164960182025029, "grad_norm": 0.7715205659161936, "learning_rate": 1.1877672693488669e-06, "loss": 0.0395, "step": 3149 }, { "epoch": 0.7167235494880546, "grad_norm": 0.9856101752123785, "learning_rate": 1.1877283952489636e-06, "loss": 0.0419, "step": 3150 }, { "epoch": 0.7169510807736064, "grad_norm": 1.1172173173435798, "learning_rate": 1.1876895096479226e-06, "loss": 0.044, "step": 3151 }, { "epoch": 0.7171786120591581, "grad_norm": 1.7824313127710856, "learning_rate": 1.1876506125465386e-06, "loss": 0.0882, "step": 3152 }, { "epoch": 0.7174061433447099, "grad_norm": 0.6797130682621123, "learning_rate": 1.1876117039456065e-06, "loss": 0.0241, "step": 3153 }, { "epoch": 0.7176336746302616, "grad_norm": 1.8479763283658723, "learning_rate": 1.1875727838459213e-06, "loss": 0.0718, "step": 3154 }, { "epoch": 0.7178612059158134, "grad_norm": 0.8348967152826982, "learning_rate": 1.187533852248279e-06, "loss": 0.0293, "step": 3155 }, { "epoch": 0.7180887372013652, "grad_norm": 1.0039148482185, "learning_rate": 1.1874949091534749e-06, "loss": 0.0465, "step": 3156 }, { "epoch": 0.718316268486917, "grad_norm": 1.6124637787966882, "learning_rate": 1.1874559545623049e-06, "loss": 0.093, "step": 3157 }, { "epoch": 0.7185437997724687, "grad_norm": 1.3428421165694695, "learning_rate": 1.1874169884755654e-06, "loss": 0.068, "step": 3158 }, { "epoch": 0.7187713310580205, "grad_norm": 0.9564008146931468, "learning_rate": 1.1873780108940527e-06, "loss": 0.0424, "step": 3159 }, { "epoch": 0.7189988623435722, "grad_norm": 0.8282933582416673, "learning_rate": 1.1873390218185636e-06, "loss": 0.0237, "step": 3160 }, { "epoch": 0.719226393629124, "grad_norm": 0.8573276208421515, "learning_rate": 1.1873000212498942e-06, "loss": 0.0431, "step": 3161 }, { "epoch": 0.7194539249146757, "grad_norm": 0.9698959738736916, "learning_rate": 1.1872610091888426e-06, "loss": 0.0444, "step": 3162 }, { "epoch": 0.7196814562002275, "grad_norm": 1.8230237032369256, "learning_rate": 1.1872219856362057e-06, "loss": 0.0878, "step": 3163 }, { "epoch": 0.7199089874857793, "grad_norm": 1.1950764632445416, "learning_rate": 1.187182950592781e-06, "loss": 0.0559, "step": 3164 }, { "epoch": 0.7201365187713311, "grad_norm": 1.1584194920218946, "learning_rate": 1.1871439040593663e-06, "loss": 0.0659, "step": 3165 }, { "epoch": 0.7203640500568829, "grad_norm": 1.149871056931848, "learning_rate": 1.1871048460367598e-06, "loss": 0.0385, "step": 3166 }, { "epoch": 0.7205915813424346, "grad_norm": 1.1250408047699638, "learning_rate": 1.1870657765257595e-06, "loss": 0.041, "step": 3167 }, { "epoch": 0.7208191126279864, "grad_norm": 0.9750498880779663, "learning_rate": 1.1870266955271645e-06, "loss": 0.0286, "step": 3168 }, { "epoch": 0.7210466439135381, "grad_norm": 1.93920042584572, "learning_rate": 1.186987603041773e-06, "loss": 0.1038, "step": 3169 }, { "epoch": 0.7212741751990899, "grad_norm": 1.1716818159995082, "learning_rate": 1.1869484990703839e-06, "loss": 0.0567, "step": 3170 }, { "epoch": 0.7215017064846416, "grad_norm": 1.39986755402065, "learning_rate": 1.1869093836137968e-06, "loss": 0.0763, "step": 3171 }, { "epoch": 0.7217292377701934, "grad_norm": 1.3078245254309557, "learning_rate": 1.186870256672811e-06, "loss": 0.061, "step": 3172 }, { "epoch": 0.7219567690557451, "grad_norm": 1.3174384504374115, "learning_rate": 1.1868311182482262e-06, "loss": 0.0416, "step": 3173 }, { "epoch": 0.7221843003412969, "grad_norm": 1.404891211071726, "learning_rate": 1.1867919683408421e-06, "loss": 0.0272, "step": 3174 }, { "epoch": 0.7224118316268487, "grad_norm": 1.4160327712086356, "learning_rate": 1.1867528069514591e-06, "loss": 0.0863, "step": 3175 }, { "epoch": 0.7226393629124005, "grad_norm": 1.400708180969138, "learning_rate": 1.1867136340808778e-06, "loss": 0.0502, "step": 3176 }, { "epoch": 0.7228668941979522, "grad_norm": 1.6148464588484233, "learning_rate": 1.1866744497298982e-06, "loss": 0.0693, "step": 3177 }, { "epoch": 0.723094425483504, "grad_norm": 1.0345610557360225, "learning_rate": 1.1866352538993216e-06, "loss": 0.0441, "step": 3178 }, { "epoch": 0.7233219567690558, "grad_norm": 1.2686233759159034, "learning_rate": 1.1865960465899492e-06, "loss": 0.0658, "step": 3179 }, { "epoch": 0.7235494880546075, "grad_norm": 1.0561518957357936, "learning_rate": 1.186556827802582e-06, "loss": 0.0316, "step": 3180 }, { "epoch": 0.7237770193401593, "grad_norm": 1.2859639053927798, "learning_rate": 1.1865175975380218e-06, "loss": 0.0748, "step": 3181 }, { "epoch": 0.724004550625711, "grad_norm": 1.0612859349624024, "learning_rate": 1.18647835579707e-06, "loss": 0.0533, "step": 3182 }, { "epoch": 0.7242320819112628, "grad_norm": 1.252216188797781, "learning_rate": 1.186439102580529e-06, "loss": 0.0582, "step": 3183 }, { "epoch": 0.7244596131968145, "grad_norm": 1.4352458095090412, "learning_rate": 1.1863998378892011e-06, "loss": 0.0498, "step": 3184 }, { "epoch": 0.7246871444823664, "grad_norm": 2.834373155221246, "learning_rate": 1.1863605617238885e-06, "loss": 0.1159, "step": 3185 }, { "epoch": 0.7249146757679181, "grad_norm": 1.1067917536466692, "learning_rate": 1.1863212740853941e-06, "loss": 0.0721, "step": 3186 }, { "epoch": 0.7251422070534699, "grad_norm": 1.2158906431759549, "learning_rate": 1.1862819749745212e-06, "loss": 0.0272, "step": 3187 }, { "epoch": 0.7253697383390216, "grad_norm": 1.6050627766428684, "learning_rate": 1.1862426643920722e-06, "loss": 0.0567, "step": 3188 }, { "epoch": 0.7255972696245734, "grad_norm": 1.0752979084787344, "learning_rate": 1.1862033423388513e-06, "loss": 0.0442, "step": 3189 }, { "epoch": 0.7258248009101251, "grad_norm": 0.9789837014009594, "learning_rate": 1.1861640088156617e-06, "loss": 0.0524, "step": 3190 }, { "epoch": 0.7260523321956769, "grad_norm": 1.085883546609774, "learning_rate": 1.1861246638233077e-06, "loss": 0.043, "step": 3191 }, { "epoch": 0.7262798634812286, "grad_norm": 2.3885050844926305, "learning_rate": 1.1860853073625931e-06, "loss": 0.052, "step": 3192 }, { "epoch": 0.7265073947667804, "grad_norm": 1.8498560328174254, "learning_rate": 1.1860459394343223e-06, "loss": 0.1156, "step": 3193 }, { "epoch": 0.7267349260523321, "grad_norm": 1.1970597608121667, "learning_rate": 1.1860065600393002e-06, "loss": 0.0409, "step": 3194 }, { "epoch": 0.726962457337884, "grad_norm": 1.5076572874006593, "learning_rate": 1.1859671691783315e-06, "loss": 0.0561, "step": 3195 }, { "epoch": 0.7271899886234358, "grad_norm": 0.7392429484644525, "learning_rate": 1.1859277668522209e-06, "loss": 0.0228, "step": 3196 }, { "epoch": 0.7274175199089875, "grad_norm": 2.2057392375122973, "learning_rate": 1.1858883530617743e-06, "loss": 0.1463, "step": 3197 }, { "epoch": 0.7276450511945393, "grad_norm": 2.23022548380175, "learning_rate": 1.185848927807797e-06, "loss": 0.0798, "step": 3198 }, { "epoch": 0.727872582480091, "grad_norm": 1.7101734591640143, "learning_rate": 1.1858094910910945e-06, "loss": 0.0721, "step": 3199 }, { "epoch": 0.7281001137656428, "grad_norm": 1.013052127850482, "learning_rate": 1.1857700429124733e-06, "loss": 0.044, "step": 3200 }, { "epoch": 0.7283276450511945, "grad_norm": 0.9338207707410625, "learning_rate": 1.1857305832727395e-06, "loss": 0.0356, "step": 3201 }, { "epoch": 0.7285551763367463, "grad_norm": 1.151011668858322, "learning_rate": 1.1856911121726993e-06, "loss": 0.055, "step": 3202 }, { "epoch": 0.728782707622298, "grad_norm": 0.882191391961338, "learning_rate": 1.1856516296131596e-06, "loss": 0.0293, "step": 3203 }, { "epoch": 0.7290102389078499, "grad_norm": 1.701498982454748, "learning_rate": 1.1856121355949276e-06, "loss": 0.0632, "step": 3204 }, { "epoch": 0.7292377701934016, "grad_norm": 0.9653458303805711, "learning_rate": 1.18557263011881e-06, "loss": 0.0357, "step": 3205 }, { "epoch": 0.7294653014789534, "grad_norm": 1.2647182576075804, "learning_rate": 1.1855331131856146e-06, "loss": 0.0988, "step": 3206 }, { "epoch": 0.7296928327645051, "grad_norm": 1.2708027008753917, "learning_rate": 1.185493584796149e-06, "loss": 0.0659, "step": 3207 }, { "epoch": 0.7299203640500569, "grad_norm": 1.7886995517155944, "learning_rate": 1.185454044951221e-06, "loss": 0.0779, "step": 3208 }, { "epoch": 0.7301478953356086, "grad_norm": 0.6845065442131042, "learning_rate": 1.1854144936516388e-06, "loss": 0.0325, "step": 3209 }, { "epoch": 0.7303754266211604, "grad_norm": 0.9425373691725761, "learning_rate": 1.1853749308982107e-06, "loss": 0.0353, "step": 3210 }, { "epoch": 0.7306029579067121, "grad_norm": 1.6009640208057585, "learning_rate": 1.1853353566917452e-06, "loss": 0.0689, "step": 3211 }, { "epoch": 0.7308304891922639, "grad_norm": 0.9565911200831438, "learning_rate": 1.1852957710330511e-06, "loss": 0.0438, "step": 3212 }, { "epoch": 0.7310580204778157, "grad_norm": 1.6166034190859908, "learning_rate": 1.1852561739229377e-06, "loss": 0.0521, "step": 3213 }, { "epoch": 0.7312855517633675, "grad_norm": 2.3048115272176983, "learning_rate": 1.1852165653622141e-06, "loss": 0.1036, "step": 3214 }, { "epoch": 0.7315130830489193, "grad_norm": 0.6867861750300609, "learning_rate": 1.18517694535169e-06, "loss": 0.0285, "step": 3215 }, { "epoch": 0.731740614334471, "grad_norm": 0.7222977767557096, "learning_rate": 1.1851373138921752e-06, "loss": 0.0332, "step": 3216 }, { "epoch": 0.7319681456200228, "grad_norm": 1.0427932498841996, "learning_rate": 1.1850976709844792e-06, "loss": 0.0654, "step": 3217 }, { "epoch": 0.7321956769055745, "grad_norm": 0.7626880126713813, "learning_rate": 1.1850580166294127e-06, "loss": 0.0257, "step": 3218 }, { "epoch": 0.7324232081911263, "grad_norm": 1.3204712038737856, "learning_rate": 1.1850183508277862e-06, "loss": 0.0821, "step": 3219 }, { "epoch": 0.732650739476678, "grad_norm": 1.920439901931252, "learning_rate": 1.18497867358041e-06, "loss": 0.0843, "step": 3220 }, { "epoch": 0.7328782707622298, "grad_norm": 1.4946908757395634, "learning_rate": 1.1849389848880955e-06, "loss": 0.075, "step": 3221 }, { "epoch": 0.7331058020477815, "grad_norm": 0.5605078244225655, "learning_rate": 1.1848992847516535e-06, "loss": 0.0188, "step": 3222 }, { "epoch": 0.7333333333333333, "grad_norm": 0.7658587959864014, "learning_rate": 1.1848595731718955e-06, "loss": 0.0246, "step": 3223 }, { "epoch": 0.7335608646188851, "grad_norm": 1.6693233826267546, "learning_rate": 1.1848198501496331e-06, "loss": 0.0709, "step": 3224 }, { "epoch": 0.7337883959044369, "grad_norm": 1.6483972044168969, "learning_rate": 1.1847801156856783e-06, "loss": 0.0735, "step": 3225 }, { "epoch": 0.7340159271899886, "grad_norm": 1.4807700617572261, "learning_rate": 1.1847403697808433e-06, "loss": 0.0838, "step": 3226 }, { "epoch": 0.7342434584755404, "grad_norm": 1.485456481779637, "learning_rate": 1.18470061243594e-06, "loss": 0.0687, "step": 3227 }, { "epoch": 0.7344709897610922, "grad_norm": 1.4565020245267035, "learning_rate": 1.1846608436517813e-06, "loss": 0.0593, "step": 3228 }, { "epoch": 0.7346985210466439, "grad_norm": 1.3974828149244531, "learning_rate": 1.1846210634291799e-06, "loss": 0.0683, "step": 3229 }, { "epoch": 0.7349260523321957, "grad_norm": 0.9824198436565561, "learning_rate": 1.184581271768949e-06, "loss": 0.0474, "step": 3230 }, { "epoch": 0.7351535836177474, "grad_norm": 1.4337603662453289, "learning_rate": 1.1845414686719014e-06, "loss": 0.0621, "step": 3231 }, { "epoch": 0.7353811149032992, "grad_norm": 0.8374788974669934, "learning_rate": 1.1845016541388513e-06, "loss": 0.0388, "step": 3232 }, { "epoch": 0.7356086461888509, "grad_norm": 1.6437938029709063, "learning_rate": 1.184461828170612e-06, "loss": 0.0822, "step": 3233 }, { "epoch": 0.7358361774744028, "grad_norm": 0.9054278416004449, "learning_rate": 1.1844219907679973e-06, "loss": 0.0364, "step": 3234 }, { "epoch": 0.7360637087599545, "grad_norm": 1.2269654818634814, "learning_rate": 1.184382141931822e-06, "loss": 0.0474, "step": 3235 }, { "epoch": 0.7362912400455063, "grad_norm": 1.2170865361290026, "learning_rate": 1.1843422816628998e-06, "loss": 0.0463, "step": 3236 }, { "epoch": 0.736518771331058, "grad_norm": 1.224432806747687, "learning_rate": 1.184302409962046e-06, "loss": 0.047, "step": 3237 }, { "epoch": 0.7367463026166098, "grad_norm": 1.0965112824005279, "learning_rate": 1.1842625268300754e-06, "loss": 0.0359, "step": 3238 }, { "epoch": 0.7369738339021615, "grad_norm": 1.0155537003395356, "learning_rate": 1.1842226322678028e-06, "loss": 0.0325, "step": 3239 }, { "epoch": 0.7372013651877133, "grad_norm": 1.6425941417778662, "learning_rate": 1.1841827262760436e-06, "loss": 0.0666, "step": 3240 }, { "epoch": 0.737428896473265, "grad_norm": 1.0729010350000274, "learning_rate": 1.1841428088556137e-06, "loss": 0.0446, "step": 3241 }, { "epoch": 0.7376564277588168, "grad_norm": 1.5054940909406098, "learning_rate": 1.184102880007329e-06, "loss": 0.0734, "step": 3242 }, { "epoch": 0.7378839590443687, "grad_norm": 1.4685805990987935, "learning_rate": 1.1840629397320052e-06, "loss": 0.0657, "step": 3243 }, { "epoch": 0.7381114903299204, "grad_norm": 1.0491905607582221, "learning_rate": 1.1840229880304589e-06, "loss": 0.0553, "step": 3244 }, { "epoch": 0.7383390216154722, "grad_norm": 1.6183523557552497, "learning_rate": 1.1839830249035062e-06, "loss": 0.0401, "step": 3245 }, { "epoch": 0.7385665529010239, "grad_norm": 2.0313633648707885, "learning_rate": 1.1839430503519645e-06, "loss": 0.0871, "step": 3246 }, { "epoch": 0.7387940841865757, "grad_norm": 2.2817905288401485, "learning_rate": 1.1839030643766505e-06, "loss": 0.0876, "step": 3247 }, { "epoch": 0.7390216154721274, "grad_norm": 1.2662039437671349, "learning_rate": 1.1838630669783814e-06, "loss": 0.0597, "step": 3248 }, { "epoch": 0.7392491467576792, "grad_norm": 2.4757641907590116, "learning_rate": 1.1838230581579746e-06, "loss": 0.0364, "step": 3249 }, { "epoch": 0.7394766780432309, "grad_norm": 1.509162194685489, "learning_rate": 1.183783037916248e-06, "loss": 0.0728, "step": 3250 }, { "epoch": 0.7397042093287827, "grad_norm": 1.8117671336209638, "learning_rate": 1.1837430062540196e-06, "loss": 0.074, "step": 3251 }, { "epoch": 0.7399317406143344, "grad_norm": 1.426185256628181, "learning_rate": 1.1837029631721072e-06, "loss": 0.0663, "step": 3252 }, { "epoch": 0.7401592718998863, "grad_norm": 1.930245275398741, "learning_rate": 1.1836629086713296e-06, "loss": 0.0579, "step": 3253 }, { "epoch": 0.740386803185438, "grad_norm": 2.323978623936963, "learning_rate": 1.1836228427525054e-06, "loss": 0.0314, "step": 3254 }, { "epoch": 0.7406143344709898, "grad_norm": 1.2652157180134698, "learning_rate": 1.183582765416453e-06, "loss": 0.0451, "step": 3255 }, { "epoch": 0.7408418657565415, "grad_norm": 1.1001275189243191, "learning_rate": 1.1835426766639923e-06, "loss": 0.034, "step": 3256 }, { "epoch": 0.7410693970420933, "grad_norm": 1.2716136324777447, "learning_rate": 1.183502576495942e-06, "loss": 0.0434, "step": 3257 }, { "epoch": 0.741296928327645, "grad_norm": 1.8985708463935187, "learning_rate": 1.1834624649131218e-06, "loss": 0.0825, "step": 3258 }, { "epoch": 0.7415244596131968, "grad_norm": 0.9934071400912606, "learning_rate": 1.1834223419163518e-06, "loss": 0.0332, "step": 3259 }, { "epoch": 0.7417519908987485, "grad_norm": 1.9148887694046421, "learning_rate": 1.1833822075064517e-06, "loss": 0.0288, "step": 3260 }, { "epoch": 0.7419795221843003, "grad_norm": 1.1004472895017543, "learning_rate": 1.183342061684242e-06, "loss": 0.0487, "step": 3261 }, { "epoch": 0.742207053469852, "grad_norm": 1.203746668995667, "learning_rate": 1.183301904450543e-06, "loss": 0.0423, "step": 3262 }, { "epoch": 0.7424345847554039, "grad_norm": 2.131599554211613, "learning_rate": 1.1832617358061756e-06, "loss": 0.078, "step": 3263 }, { "epoch": 0.7426621160409557, "grad_norm": 1.3232000144442189, "learning_rate": 1.1832215557519608e-06, "loss": 0.0542, "step": 3264 }, { "epoch": 0.7428896473265074, "grad_norm": 1.3347834617776861, "learning_rate": 1.1831813642887196e-06, "loss": 0.0571, "step": 3265 }, { "epoch": 0.7431171786120592, "grad_norm": 1.491233272879141, "learning_rate": 1.1831411614172735e-06, "loss": 0.0706, "step": 3266 }, { "epoch": 0.7433447098976109, "grad_norm": 2.338157694562792, "learning_rate": 1.1831009471384445e-06, "loss": 0.0591, "step": 3267 }, { "epoch": 0.7435722411831627, "grad_norm": 1.1906122468487437, "learning_rate": 1.1830607214530543e-06, "loss": 0.0693, "step": 3268 }, { "epoch": 0.7437997724687144, "grad_norm": 1.2940810772686644, "learning_rate": 1.1830204843619248e-06, "loss": 0.0517, "step": 3269 }, { "epoch": 0.7440273037542662, "grad_norm": 2.214601640279187, "learning_rate": 1.1829802358658785e-06, "loss": 0.1046, "step": 3270 }, { "epoch": 0.7442548350398179, "grad_norm": 1.039009352489729, "learning_rate": 1.1829399759657383e-06, "loss": 0.0592, "step": 3271 }, { "epoch": 0.7444823663253698, "grad_norm": 1.8684769750059191, "learning_rate": 1.1828997046623267e-06, "loss": 0.093, "step": 3272 }, { "epoch": 0.7447098976109215, "grad_norm": 1.823521554100394, "learning_rate": 1.1828594219564669e-06, "loss": 0.049, "step": 3273 }, { "epoch": 0.7449374288964733, "grad_norm": 1.8271992779263462, "learning_rate": 1.182819127848982e-06, "loss": 0.0303, "step": 3274 }, { "epoch": 0.745164960182025, "grad_norm": 3.0795350516203928, "learning_rate": 1.1827788223406959e-06, "loss": 0.1191, "step": 3275 }, { "epoch": 0.7453924914675768, "grad_norm": 1.7589178709629525, "learning_rate": 1.1827385054324323e-06, "loss": 0.0616, "step": 3276 }, { "epoch": 0.7456200227531286, "grad_norm": 2.4742706309940976, "learning_rate": 1.1826981771250148e-06, "loss": 0.077, "step": 3277 }, { "epoch": 0.7458475540386803, "grad_norm": 1.9750018572038446, "learning_rate": 1.1826578374192681e-06, "loss": 0.0795, "step": 3278 }, { "epoch": 0.7460750853242321, "grad_norm": 1.5314832082240002, "learning_rate": 1.1826174863160168e-06, "loss": 0.0694, "step": 3279 }, { "epoch": 0.7463026166097838, "grad_norm": 0.9199631256267508, "learning_rate": 1.182577123816085e-06, "loss": 0.0407, "step": 3280 }, { "epoch": 0.7465301478953356, "grad_norm": 1.6996895039706184, "learning_rate": 1.1825367499202978e-06, "loss": 0.0648, "step": 3281 }, { "epoch": 0.7467576791808874, "grad_norm": 1.538865347545186, "learning_rate": 1.1824963646294806e-06, "loss": 0.0483, "step": 3282 }, { "epoch": 0.7469852104664392, "grad_norm": 1.5246278126883535, "learning_rate": 1.1824559679444588e-06, "loss": 0.0652, "step": 3283 }, { "epoch": 0.7472127417519909, "grad_norm": 0.7990923695187152, "learning_rate": 1.182415559866058e-06, "loss": 0.0416, "step": 3284 }, { "epoch": 0.7474402730375427, "grad_norm": 1.0564862923399663, "learning_rate": 1.182375140395104e-06, "loss": 0.0414, "step": 3285 }, { "epoch": 0.7476678043230944, "grad_norm": 1.1846269731523205, "learning_rate": 1.1823347095324228e-06, "loss": 0.0621, "step": 3286 }, { "epoch": 0.7478953356086462, "grad_norm": 0.8569640753381744, "learning_rate": 1.1822942672788409e-06, "loss": 0.0278, "step": 3287 }, { "epoch": 0.7481228668941979, "grad_norm": 1.571726141718777, "learning_rate": 1.1822538136351849e-06, "loss": 0.0613, "step": 3288 }, { "epoch": 0.7483503981797497, "grad_norm": 1.1282851528775475, "learning_rate": 1.1822133486022815e-06, "loss": 0.0392, "step": 3289 }, { "epoch": 0.7485779294653014, "grad_norm": 1.0669556173185228, "learning_rate": 1.1821728721809577e-06, "loss": 0.0294, "step": 3290 }, { "epoch": 0.7488054607508532, "grad_norm": 2.466980162201729, "learning_rate": 1.1821323843720408e-06, "loss": 0.0745, "step": 3291 }, { "epoch": 0.749032992036405, "grad_norm": 1.8125453554610331, "learning_rate": 1.1820918851763582e-06, "loss": 0.0798, "step": 3292 }, { "epoch": 0.7492605233219568, "grad_norm": 1.6160104087285256, "learning_rate": 1.182051374594738e-06, "loss": 0.0769, "step": 3293 }, { "epoch": 0.7494880546075086, "grad_norm": 1.5687919942801445, "learning_rate": 1.1820108526280076e-06, "loss": 0.0626, "step": 3294 }, { "epoch": 0.7497155858930603, "grad_norm": 0.9398572843854899, "learning_rate": 1.1819703192769955e-06, "loss": 0.0482, "step": 3295 }, { "epoch": 0.7499431171786121, "grad_norm": 12.232356704164316, "learning_rate": 1.1819297745425304e-06, "loss": 0.0627, "step": 3296 }, { "epoch": 0.7501706484641638, "grad_norm": 1.4183355936794655, "learning_rate": 1.1818892184254404e-06, "loss": 0.0731, "step": 3297 }, { "epoch": 0.7503981797497156, "grad_norm": 0.7929826048625227, "learning_rate": 1.1818486509265547e-06, "loss": 0.028, "step": 3298 }, { "epoch": 0.7506257110352673, "grad_norm": 0.9321132746109633, "learning_rate": 1.1818080720467026e-06, "loss": 0.0406, "step": 3299 }, { "epoch": 0.7508532423208191, "grad_norm": 0.7184037331999952, "learning_rate": 1.1817674817867131e-06, "loss": 0.0292, "step": 3300 }, { "epoch": 0.7510807736063708, "grad_norm": 1.7585855147942302, "learning_rate": 1.181726880147416e-06, "loss": 0.074, "step": 3301 }, { "epoch": 0.7513083048919227, "grad_norm": 0.9632647926307707, "learning_rate": 1.181686267129641e-06, "loss": 0.0318, "step": 3302 }, { "epoch": 0.7515358361774744, "grad_norm": 1.2947286461920264, "learning_rate": 1.1816456427342181e-06, "loss": 0.0649, "step": 3303 }, { "epoch": 0.7517633674630262, "grad_norm": 0.9051783521749419, "learning_rate": 1.181605006961978e-06, "loss": 0.0318, "step": 3304 }, { "epoch": 0.7519908987485779, "grad_norm": 1.9818984386721539, "learning_rate": 1.1815643598137507e-06, "loss": 0.0909, "step": 3305 }, { "epoch": 0.7522184300341297, "grad_norm": 1.1059310163256162, "learning_rate": 1.1815237012903675e-06, "loss": 0.0422, "step": 3306 }, { "epoch": 0.7524459613196814, "grad_norm": 2.4093572946300865, "learning_rate": 1.1814830313926589e-06, "loss": 0.0944, "step": 3307 }, { "epoch": 0.7526734926052332, "grad_norm": 1.2458152163270595, "learning_rate": 1.1814423501214562e-06, "loss": 0.0422, "step": 3308 }, { "epoch": 0.752901023890785, "grad_norm": 1.1846230368435793, "learning_rate": 1.1814016574775909e-06, "loss": 0.0459, "step": 3309 }, { "epoch": 0.7531285551763367, "grad_norm": 0.7274927229665066, "learning_rate": 1.1813609534618948e-06, "loss": 0.031, "step": 3310 }, { "epoch": 0.7533560864618886, "grad_norm": 1.0084301910129103, "learning_rate": 1.1813202380751998e-06, "loss": 0.022, "step": 3311 }, { "epoch": 0.7535836177474403, "grad_norm": 1.2248641094215589, "learning_rate": 1.1812795113183378e-06, "loss": 0.0519, "step": 3312 }, { "epoch": 0.7538111490329921, "grad_norm": 1.2989392125838315, "learning_rate": 1.1812387731921415e-06, "loss": 0.0418, "step": 3313 }, { "epoch": 0.7540386803185438, "grad_norm": 1.8035598035830265, "learning_rate": 1.1811980236974435e-06, "loss": 0.047, "step": 3314 }, { "epoch": 0.7542662116040956, "grad_norm": 0.9445160919482642, "learning_rate": 1.1811572628350764e-06, "loss": 0.053, "step": 3315 }, { "epoch": 0.7544937428896473, "grad_norm": 1.820531092229777, "learning_rate": 1.1811164906058735e-06, "loss": 0.0898, "step": 3316 }, { "epoch": 0.7547212741751991, "grad_norm": 1.7017364368903272, "learning_rate": 1.1810757070106678e-06, "loss": 0.064, "step": 3317 }, { "epoch": 0.7549488054607508, "grad_norm": 1.589156926554848, "learning_rate": 1.1810349120502932e-06, "loss": 0.069, "step": 3318 }, { "epoch": 0.7551763367463026, "grad_norm": 0.9315330165580284, "learning_rate": 1.1809941057255834e-06, "loss": 0.043, "step": 3319 }, { "epoch": 0.7554038680318543, "grad_norm": 1.7002202165396425, "learning_rate": 1.1809532880373721e-06, "loss": 0.0662, "step": 3320 }, { "epoch": 0.7556313993174062, "grad_norm": 1.4641161178529223, "learning_rate": 1.180912458986494e-06, "loss": 0.0696, "step": 3321 }, { "epoch": 0.755858930602958, "grad_norm": 1.5316937073382941, "learning_rate": 1.180871618573783e-06, "loss": 0.0532, "step": 3322 }, { "epoch": 0.7560864618885097, "grad_norm": 1.5604240478591438, "learning_rate": 1.1808307668000745e-06, "loss": 0.0511, "step": 3323 }, { "epoch": 0.7563139931740614, "grad_norm": 1.7255746868374173, "learning_rate": 1.180789903666203e-06, "loss": 0.0646, "step": 3324 }, { "epoch": 0.7565415244596132, "grad_norm": 1.3143153209229907, "learning_rate": 1.1807490291730036e-06, "loss": 0.0792, "step": 3325 }, { "epoch": 0.756769055745165, "grad_norm": 0.8135022202076733, "learning_rate": 1.1807081433213122e-06, "loss": 0.0432, "step": 3326 }, { "epoch": 0.7569965870307167, "grad_norm": 1.4086605703937913, "learning_rate": 1.1806672461119637e-06, "loss": 0.0774, "step": 3327 }, { "epoch": 0.7572241183162685, "grad_norm": 1.7398967165213481, "learning_rate": 1.1806263375457947e-06, "loss": 0.0598, "step": 3328 }, { "epoch": 0.7574516496018202, "grad_norm": 1.4489241741578003, "learning_rate": 1.1805854176236406e-06, "loss": 0.0655, "step": 3329 }, { "epoch": 0.757679180887372, "grad_norm": 2.000585789486283, "learning_rate": 1.1805444863463384e-06, "loss": 0.1195, "step": 3330 }, { "epoch": 0.7579067121729238, "grad_norm": 1.1274809608332625, "learning_rate": 1.180503543714724e-06, "loss": 0.0355, "step": 3331 }, { "epoch": 0.7581342434584756, "grad_norm": 0.8018211568884192, "learning_rate": 1.1804625897296345e-06, "loss": 0.0354, "step": 3332 }, { "epoch": 0.7583617747440273, "grad_norm": 1.8306901932048323, "learning_rate": 1.1804216243919074e-06, "loss": 0.0604, "step": 3333 }, { "epoch": 0.7585893060295791, "grad_norm": 1.0198114562975762, "learning_rate": 1.1803806477023792e-06, "loss": 0.034, "step": 3334 }, { "epoch": 0.7588168373151308, "grad_norm": 1.0883117507836249, "learning_rate": 1.1803396596618878e-06, "loss": 0.0476, "step": 3335 }, { "epoch": 0.7590443686006826, "grad_norm": 1.2568287515264616, "learning_rate": 1.1802986602712705e-06, "loss": 0.0415, "step": 3336 }, { "epoch": 0.7592718998862343, "grad_norm": 1.5230920956011404, "learning_rate": 1.1802576495313657e-06, "loss": 0.0516, "step": 3337 }, { "epoch": 0.7594994311717861, "grad_norm": 3.0018010476610044, "learning_rate": 1.1802166274430116e-06, "loss": 0.0773, "step": 3338 }, { "epoch": 0.7597269624573378, "grad_norm": 0.8881506140039246, "learning_rate": 1.1801755940070464e-06, "loss": 0.0332, "step": 3339 }, { "epoch": 0.7599544937428896, "grad_norm": 1.5510159044642893, "learning_rate": 1.1801345492243087e-06, "loss": 0.0653, "step": 3340 }, { "epoch": 0.7601820250284415, "grad_norm": 1.732180484199383, "learning_rate": 1.1800934930956378e-06, "loss": 0.0681, "step": 3341 }, { "epoch": 0.7604095563139932, "grad_norm": 1.0655757603385652, "learning_rate": 1.1800524256218724e-06, "loss": 0.0635, "step": 3342 }, { "epoch": 0.760637087599545, "grad_norm": 1.3481251487359558, "learning_rate": 1.1800113468038518e-06, "loss": 0.0685, "step": 3343 }, { "epoch": 0.7608646188850967, "grad_norm": 2.2560803054292418, "learning_rate": 1.1799702566424159e-06, "loss": 0.1302, "step": 3344 }, { "epoch": 0.7610921501706485, "grad_norm": 1.6837334743270123, "learning_rate": 1.1799291551384042e-06, "loss": 0.0842, "step": 3345 }, { "epoch": 0.7613196814562002, "grad_norm": 1.3300039883273982, "learning_rate": 1.1798880422926568e-06, "loss": 0.0518, "step": 3346 }, { "epoch": 0.761547212741752, "grad_norm": 1.8839734220786841, "learning_rate": 1.1798469181060143e-06, "loss": 0.0994, "step": 3347 }, { "epoch": 0.7617747440273037, "grad_norm": 1.600150172121613, "learning_rate": 1.1798057825793167e-06, "loss": 0.0651, "step": 3348 }, { "epoch": 0.7620022753128555, "grad_norm": 1.3818697167200884, "learning_rate": 1.179764635713405e-06, "loss": 0.0473, "step": 3349 }, { "epoch": 0.7622298065984073, "grad_norm": 0.8317911257581198, "learning_rate": 1.1797234775091204e-06, "loss": 0.0337, "step": 3350 }, { "epoch": 0.7624573378839591, "grad_norm": 1.1808932366446725, "learning_rate": 1.1796823079673036e-06, "loss": 0.0376, "step": 3351 }, { "epoch": 0.7626848691695108, "grad_norm": 1.6710801191008726, "learning_rate": 1.1796411270887965e-06, "loss": 0.0592, "step": 3352 }, { "epoch": 0.7629124004550626, "grad_norm": 1.0835763404496392, "learning_rate": 1.1795999348744403e-06, "loss": 0.0525, "step": 3353 }, { "epoch": 0.7631399317406143, "grad_norm": 1.2436113476656212, "learning_rate": 1.1795587313250773e-06, "loss": 0.0451, "step": 3354 }, { "epoch": 0.7633674630261661, "grad_norm": 1.1914849618578496, "learning_rate": 1.1795175164415493e-06, "loss": 0.0578, "step": 3355 }, { "epoch": 0.7635949943117178, "grad_norm": 2.3441953010272867, "learning_rate": 1.1794762902246988e-06, "loss": 0.1105, "step": 3356 }, { "epoch": 0.7638225255972696, "grad_norm": 2.1756802775613364, "learning_rate": 1.1794350526753688e-06, "loss": 0.1026, "step": 3357 }, { "epoch": 0.7640500568828213, "grad_norm": 2.3594222993387652, "learning_rate": 1.1793938037944014e-06, "loss": 0.1163, "step": 3358 }, { "epoch": 0.7642775881683731, "grad_norm": 0.9175490163523685, "learning_rate": 1.1793525435826399e-06, "loss": 0.0287, "step": 3359 }, { "epoch": 0.764505119453925, "grad_norm": 1.9177635612847674, "learning_rate": 1.1793112720409277e-06, "loss": 0.0945, "step": 3360 }, { "epoch": 0.7647326507394767, "grad_norm": 1.0136874739515975, "learning_rate": 1.1792699891701085e-06, "loss": 0.0389, "step": 3361 }, { "epoch": 0.7649601820250285, "grad_norm": 0.9880415043802684, "learning_rate": 1.1792286949710254e-06, "loss": 0.052, "step": 3362 }, { "epoch": 0.7651877133105802, "grad_norm": 1.5402287968172736, "learning_rate": 1.1791873894445233e-06, "loss": 0.0698, "step": 3363 }, { "epoch": 0.765415244596132, "grad_norm": 1.3968644273802429, "learning_rate": 1.1791460725914455e-06, "loss": 0.065, "step": 3364 }, { "epoch": 0.7656427758816837, "grad_norm": 1.8023364682651095, "learning_rate": 1.179104744412637e-06, "loss": 0.0653, "step": 3365 }, { "epoch": 0.7658703071672355, "grad_norm": 1.3938733296222532, "learning_rate": 1.1790634049089425e-06, "loss": 0.0604, "step": 3366 }, { "epoch": 0.7660978384527872, "grad_norm": 1.6406363966121917, "learning_rate": 1.1790220540812063e-06, "loss": 0.0781, "step": 3367 }, { "epoch": 0.766325369738339, "grad_norm": 1.2609639774979144, "learning_rate": 1.1789806919302743e-06, "loss": 0.0429, "step": 3368 }, { "epoch": 0.7665529010238907, "grad_norm": 1.4055922666036842, "learning_rate": 1.1789393184569914e-06, "loss": 0.0447, "step": 3369 }, { "epoch": 0.7667804323094426, "grad_norm": 1.2844015411070886, "learning_rate": 1.1788979336622034e-06, "loss": 0.0686, "step": 3370 }, { "epoch": 0.7670079635949943, "grad_norm": 1.2894781437661027, "learning_rate": 1.178856537546756e-06, "loss": 0.0676, "step": 3371 }, { "epoch": 0.7672354948805461, "grad_norm": 1.0960231571013044, "learning_rate": 1.1788151301114952e-06, "loss": 0.0472, "step": 3372 }, { "epoch": 0.7674630261660979, "grad_norm": 1.504396802843696, "learning_rate": 1.1787737113572678e-06, "loss": 0.0588, "step": 3373 }, { "epoch": 0.7676905574516496, "grad_norm": 1.6683614299229446, "learning_rate": 1.1787322812849196e-06, "loss": 0.0711, "step": 3374 }, { "epoch": 0.7679180887372014, "grad_norm": 2.581634805020921, "learning_rate": 1.1786908398952977e-06, "loss": 0.1279, "step": 3375 }, { "epoch": 0.7681456200227531, "grad_norm": 1.1979082190366048, "learning_rate": 1.1786493871892491e-06, "loss": 0.0387, "step": 3376 }, { "epoch": 0.7683731513083049, "grad_norm": 0.7863088959802915, "learning_rate": 1.178607923167621e-06, "loss": 0.032, "step": 3377 }, { "epoch": 0.7686006825938566, "grad_norm": 1.395956623470284, "learning_rate": 1.1785664478312607e-06, "loss": 0.0557, "step": 3378 }, { "epoch": 0.7688282138794085, "grad_norm": 1.3990313414663118, "learning_rate": 1.1785249611810163e-06, "loss": 0.0675, "step": 3379 }, { "epoch": 0.7690557451649602, "grad_norm": 1.5727682457288055, "learning_rate": 1.1784834632177352e-06, "loss": 0.0488, "step": 3380 }, { "epoch": 0.769283276450512, "grad_norm": 1.2997891762689524, "learning_rate": 1.178441953942266e-06, "loss": 0.0497, "step": 3381 }, { "epoch": 0.7695108077360637, "grad_norm": 1.1951369765878084, "learning_rate": 1.1784004333554565e-06, "loss": 0.043, "step": 3382 }, { "epoch": 0.7697383390216155, "grad_norm": 1.3760280462231655, "learning_rate": 1.178358901458156e-06, "loss": 0.0637, "step": 3383 }, { "epoch": 0.7699658703071672, "grad_norm": 1.167678907961144, "learning_rate": 1.1783173582512127e-06, "loss": 0.0504, "step": 3384 }, { "epoch": 0.770193401592719, "grad_norm": 1.1188494063998051, "learning_rate": 1.178275803735476e-06, "loss": 0.0396, "step": 3385 }, { "epoch": 0.7704209328782707, "grad_norm": 1.6858173282426545, "learning_rate": 1.1782342379117954e-06, "loss": 0.0749, "step": 3386 }, { "epoch": 0.7706484641638225, "grad_norm": 1.6532847814641132, "learning_rate": 1.17819266078102e-06, "loss": 0.0531, "step": 3387 }, { "epoch": 0.7708759954493742, "grad_norm": 1.309511727163334, "learning_rate": 1.1781510723439995e-06, "loss": 0.0619, "step": 3388 }, { "epoch": 0.7711035267349261, "grad_norm": 1.8164710685094503, "learning_rate": 1.1781094726015842e-06, "loss": 0.0849, "step": 3389 }, { "epoch": 0.7713310580204779, "grad_norm": 1.5604092892535402, "learning_rate": 1.1780678615546245e-06, "loss": 0.0718, "step": 3390 }, { "epoch": 0.7715585893060296, "grad_norm": 1.3284971139146458, "learning_rate": 1.1780262392039706e-06, "loss": 0.0532, "step": 3391 }, { "epoch": 0.7717861205915814, "grad_norm": 0.8981286940096231, "learning_rate": 1.177984605550473e-06, "loss": 0.038, "step": 3392 }, { "epoch": 0.7720136518771331, "grad_norm": 1.2803703590807574, "learning_rate": 1.177942960594983e-06, "loss": 0.0381, "step": 3393 }, { "epoch": 0.7722411831626849, "grad_norm": 1.0532494793515876, "learning_rate": 1.1779013043383516e-06, "loss": 0.0274, "step": 3394 }, { "epoch": 0.7724687144482366, "grad_norm": 1.0370979550305721, "learning_rate": 1.17785963678143e-06, "loss": 0.0527, "step": 3395 }, { "epoch": 0.7726962457337884, "grad_norm": 0.8794501213484224, "learning_rate": 1.1778179579250699e-06, "loss": 0.0252, "step": 3396 }, { "epoch": 0.7729237770193401, "grad_norm": 1.2333346024054503, "learning_rate": 1.1777762677701232e-06, "loss": 0.0443, "step": 3397 }, { "epoch": 0.7731513083048919, "grad_norm": 1.0027591833023217, "learning_rate": 1.1777345663174419e-06, "loss": 0.0689, "step": 3398 }, { "epoch": 0.7733788395904437, "grad_norm": 1.1808711247974837, "learning_rate": 1.1776928535678784e-06, "loss": 0.0575, "step": 3399 }, { "epoch": 0.7736063708759955, "grad_norm": 1.3093615145015158, "learning_rate": 1.1776511295222852e-06, "loss": 0.0602, "step": 3400 }, { "epoch": 0.7738339021615472, "grad_norm": 1.488841199985786, "learning_rate": 1.177609394181515e-06, "loss": 0.0713, "step": 3401 }, { "epoch": 0.774061433447099, "grad_norm": 2.280972914382851, "learning_rate": 1.177567647546421e-06, "loss": 0.1191, "step": 3402 }, { "epoch": 0.7742889647326507, "grad_norm": 1.6714940535505132, "learning_rate": 1.177525889617856e-06, "loss": 0.0872, "step": 3403 }, { "epoch": 0.7745164960182025, "grad_norm": 0.7472197048359, "learning_rate": 1.177484120396674e-06, "loss": 0.0307, "step": 3404 }, { "epoch": 0.7747440273037542, "grad_norm": 0.9070022095243302, "learning_rate": 1.1774423398837282e-06, "loss": 0.0269, "step": 3405 }, { "epoch": 0.774971558589306, "grad_norm": 0.8847029817903657, "learning_rate": 1.177400548079873e-06, "loss": 0.038, "step": 3406 }, { "epoch": 0.7751990898748577, "grad_norm": 1.0038791040094233, "learning_rate": 1.177358744985962e-06, "loss": 0.0514, "step": 3407 }, { "epoch": 0.7754266211604095, "grad_norm": 0.9653957279075387, "learning_rate": 1.1773169306028498e-06, "loss": 0.0376, "step": 3408 }, { "epoch": 0.7756541524459614, "grad_norm": 1.025381489223499, "learning_rate": 1.1772751049313911e-06, "loss": 0.0404, "step": 3409 }, { "epoch": 0.7758816837315131, "grad_norm": 4.749905802391231, "learning_rate": 1.1772332679724408e-06, "loss": 0.0609, "step": 3410 }, { "epoch": 0.7761092150170649, "grad_norm": 0.9550524873306326, "learning_rate": 1.1771914197268538e-06, "loss": 0.0303, "step": 3411 }, { "epoch": 0.7763367463026166, "grad_norm": 0.7920355825602112, "learning_rate": 1.1771495601954856e-06, "loss": 0.0357, "step": 3412 }, { "epoch": 0.7765642775881684, "grad_norm": 1.0204752879838752, "learning_rate": 1.1771076893791914e-06, "loss": 0.0296, "step": 3413 }, { "epoch": 0.7767918088737201, "grad_norm": 1.1240956814281524, "learning_rate": 1.1770658072788272e-06, "loss": 0.0326, "step": 3414 }, { "epoch": 0.7770193401592719, "grad_norm": 2.188614006253551, "learning_rate": 1.1770239138952492e-06, "loss": 0.0814, "step": 3415 }, { "epoch": 0.7772468714448236, "grad_norm": 2.1028001694138916, "learning_rate": 1.1769820092293132e-06, "loss": 0.1016, "step": 3416 }, { "epoch": 0.7774744027303754, "grad_norm": 1.3041493538358495, "learning_rate": 1.1769400932818758e-06, "loss": 0.0689, "step": 3417 }, { "epoch": 0.7777019340159272, "grad_norm": 0.8082774287076608, "learning_rate": 1.1768981660537938e-06, "loss": 0.0465, "step": 3418 }, { "epoch": 0.777929465301479, "grad_norm": 0.9743628621052991, "learning_rate": 1.1768562275459242e-06, "loss": 0.0515, "step": 3419 }, { "epoch": 0.7781569965870307, "grad_norm": 1.3767418297539598, "learning_rate": 1.1768142777591235e-06, "loss": 0.0765, "step": 3420 }, { "epoch": 0.7783845278725825, "grad_norm": 3.29832363084199, "learning_rate": 1.17677231669425e-06, "loss": 0.1072, "step": 3421 }, { "epoch": 0.7786120591581343, "grad_norm": 1.3112653588991108, "learning_rate": 1.1767303443521608e-06, "loss": 0.0412, "step": 3422 }, { "epoch": 0.778839590443686, "grad_norm": 1.3040229403784145, "learning_rate": 1.1766883607337137e-06, "loss": 0.0748, "step": 3423 }, { "epoch": 0.7790671217292378, "grad_norm": 1.5632822983270742, "learning_rate": 1.176646365839767e-06, "loss": 0.0565, "step": 3424 }, { "epoch": 0.7792946530147895, "grad_norm": 0.8997709014511377, "learning_rate": 1.1766043596711787e-06, "loss": 0.0612, "step": 3425 }, { "epoch": 0.7795221843003413, "grad_norm": 1.0705681786627015, "learning_rate": 1.1765623422288078e-06, "loss": 0.0353, "step": 3426 }, { "epoch": 0.779749715585893, "grad_norm": 0.9644671179029716, "learning_rate": 1.1765203135135126e-06, "loss": 0.0371, "step": 3427 }, { "epoch": 0.7799772468714449, "grad_norm": 1.1638233882315414, "learning_rate": 1.176478273526152e-06, "loss": 0.0519, "step": 3428 }, { "epoch": 0.7802047781569966, "grad_norm": 1.6611610210478276, "learning_rate": 1.1764362222675857e-06, "loss": 0.0733, "step": 3429 }, { "epoch": 0.7804323094425484, "grad_norm": 1.3309978458504417, "learning_rate": 1.176394159738673e-06, "loss": 0.0658, "step": 3430 }, { "epoch": 0.7806598407281001, "grad_norm": 1.379255981946512, "learning_rate": 1.1763520859402735e-06, "loss": 0.0803, "step": 3431 }, { "epoch": 0.7808873720136519, "grad_norm": 1.2786437648131908, "learning_rate": 1.176310000873247e-06, "loss": 0.0523, "step": 3432 }, { "epoch": 0.7811149032992036, "grad_norm": 1.382541657289892, "learning_rate": 1.1762679045384537e-06, "loss": 0.0593, "step": 3433 }, { "epoch": 0.7813424345847554, "grad_norm": 1.141255046714792, "learning_rate": 1.1762257969367543e-06, "loss": 0.0404, "step": 3434 }, { "epoch": 0.7815699658703071, "grad_norm": 1.6145161686719605, "learning_rate": 1.176183678069009e-06, "loss": 0.0636, "step": 3435 }, { "epoch": 0.7817974971558589, "grad_norm": 2.0980684433512, "learning_rate": 1.1761415479360784e-06, "loss": 0.0895, "step": 3436 }, { "epoch": 0.7820250284414106, "grad_norm": 1.1114214017016597, "learning_rate": 1.1760994065388246e-06, "loss": 0.0443, "step": 3437 }, { "epoch": 0.7822525597269625, "grad_norm": 1.8826452294482943, "learning_rate": 1.1760572538781077e-06, "loss": 0.0543, "step": 3438 }, { "epoch": 0.7824800910125143, "grad_norm": 0.8873377874884448, "learning_rate": 1.17601508995479e-06, "loss": 0.0376, "step": 3439 }, { "epoch": 0.782707622298066, "grad_norm": 1.113759690507731, "learning_rate": 1.175972914769733e-06, "loss": 0.0387, "step": 3440 }, { "epoch": 0.7829351535836178, "grad_norm": 0.880447245871124, "learning_rate": 1.1759307283237986e-06, "loss": 0.0364, "step": 3441 }, { "epoch": 0.7831626848691695, "grad_norm": 0.687251787996786, "learning_rate": 1.175888530617849e-06, "loss": 0.0272, "step": 3442 }, { "epoch": 0.7833902161547213, "grad_norm": 1.9024845236880248, "learning_rate": 1.175846321652747e-06, "loss": 0.0815, "step": 3443 }, { "epoch": 0.783617747440273, "grad_norm": 0.9590041079160219, "learning_rate": 1.1758041014293548e-06, "loss": 0.0503, "step": 3444 }, { "epoch": 0.7838452787258248, "grad_norm": 0.8857462180203992, "learning_rate": 1.1757618699485353e-06, "loss": 0.0472, "step": 3445 }, { "epoch": 0.7840728100113765, "grad_norm": 1.299303333784786, "learning_rate": 1.1757196272111524e-06, "loss": 0.0555, "step": 3446 }, { "epoch": 0.7843003412969284, "grad_norm": 0.845660192083495, "learning_rate": 1.1756773732180684e-06, "loss": 0.025, "step": 3447 }, { "epoch": 0.7845278725824801, "grad_norm": 2.1486155155967737, "learning_rate": 1.1756351079701477e-06, "loss": 0.086, "step": 3448 }, { "epoch": 0.7847554038680319, "grad_norm": 1.373331795777772, "learning_rate": 1.1755928314682537e-06, "loss": 0.0432, "step": 3449 }, { "epoch": 0.7849829351535836, "grad_norm": 0.8852881088520247, "learning_rate": 1.1755505437132507e-06, "loss": 0.0291, "step": 3450 }, { "epoch": 0.7852104664391354, "grad_norm": 2.454713357414872, "learning_rate": 1.1755082447060029e-06, "loss": 0.1266, "step": 3451 }, { "epoch": 0.7854379977246871, "grad_norm": 1.4196724985388591, "learning_rate": 1.1754659344473747e-06, "loss": 0.0548, "step": 3452 }, { "epoch": 0.7856655290102389, "grad_norm": 1.0229434892505915, "learning_rate": 1.175423612938231e-06, "loss": 0.0558, "step": 3453 }, { "epoch": 0.7858930602957906, "grad_norm": 1.3800295427652078, "learning_rate": 1.1753812801794368e-06, "loss": 0.0617, "step": 3454 }, { "epoch": 0.7861205915813424, "grad_norm": 1.5307677991589903, "learning_rate": 1.175338936171857e-06, "loss": 0.0551, "step": 3455 }, { "epoch": 0.7863481228668942, "grad_norm": 1.1267953232901513, "learning_rate": 1.1752965809163574e-06, "loss": 0.0427, "step": 3456 }, { "epoch": 0.786575654152446, "grad_norm": 1.227541921588263, "learning_rate": 1.1752542144138033e-06, "loss": 0.0486, "step": 3457 }, { "epoch": 0.7868031854379978, "grad_norm": 1.679516124458468, "learning_rate": 1.1752118366650608e-06, "loss": 0.0589, "step": 3458 }, { "epoch": 0.7870307167235495, "grad_norm": 1.0236231221195786, "learning_rate": 1.1751694476709962e-06, "loss": 0.0306, "step": 3459 }, { "epoch": 0.7872582480091013, "grad_norm": 1.3331183973827108, "learning_rate": 1.1751270474324757e-06, "loss": 0.0651, "step": 3460 }, { "epoch": 0.787485779294653, "grad_norm": 1.3868556370726, "learning_rate": 1.1750846359503657e-06, "loss": 0.0753, "step": 3461 }, { "epoch": 0.7877133105802048, "grad_norm": 1.2017443079409735, "learning_rate": 1.1750422132255334e-06, "loss": 0.0626, "step": 3462 }, { "epoch": 0.7879408418657565, "grad_norm": 1.4474690077464614, "learning_rate": 1.1749997792588453e-06, "loss": 0.0449, "step": 3463 }, { "epoch": 0.7881683731513083, "grad_norm": 1.2518797374298931, "learning_rate": 1.1749573340511693e-06, "loss": 0.0621, "step": 3464 }, { "epoch": 0.78839590443686, "grad_norm": 1.639438514025765, "learning_rate": 1.1749148776033723e-06, "loss": 0.063, "step": 3465 }, { "epoch": 0.7886234357224118, "grad_norm": 0.8717910913192211, "learning_rate": 1.1748724099163225e-06, "loss": 0.042, "step": 3466 }, { "epoch": 0.7888509670079636, "grad_norm": 1.144252834465786, "learning_rate": 1.1748299309908878e-06, "loss": 0.0396, "step": 3467 }, { "epoch": 0.7890784982935154, "grad_norm": 1.2552868730516495, "learning_rate": 1.174787440827936e-06, "loss": 0.0561, "step": 3468 }, { "epoch": 0.7893060295790671, "grad_norm": 0.9915360367623126, "learning_rate": 1.1747449394283361e-06, "loss": 0.0417, "step": 3469 }, { "epoch": 0.7895335608646189, "grad_norm": 2.2809743574627612, "learning_rate": 1.174702426792956e-06, "loss": 0.122, "step": 3470 }, { "epoch": 0.7897610921501707, "grad_norm": 0.9319050863528758, "learning_rate": 1.1746599029226656e-06, "loss": 0.024, "step": 3471 }, { "epoch": 0.7899886234357224, "grad_norm": 1.0243621293291936, "learning_rate": 1.174617367818333e-06, "loss": 0.0334, "step": 3472 }, { "epoch": 0.7902161547212742, "grad_norm": 0.5162499932362775, "learning_rate": 1.174574821480828e-06, "loss": 0.028, "step": 3473 }, { "epoch": 0.7904436860068259, "grad_norm": 1.1251631282315664, "learning_rate": 1.1745322639110203e-06, "loss": 0.0314, "step": 3474 }, { "epoch": 0.7906712172923777, "grad_norm": 0.881794120596432, "learning_rate": 1.1744896951097798e-06, "loss": 0.0453, "step": 3475 }, { "epoch": 0.7908987485779294, "grad_norm": 0.7316213076139558, "learning_rate": 1.1744471150779758e-06, "loss": 0.0183, "step": 3476 }, { "epoch": 0.7911262798634813, "grad_norm": 1.6604024020521415, "learning_rate": 1.1744045238164793e-06, "loss": 0.0587, "step": 3477 }, { "epoch": 0.791353811149033, "grad_norm": 1.3826539577284072, "learning_rate": 1.1743619213261604e-06, "loss": 0.0626, "step": 3478 }, { "epoch": 0.7915813424345848, "grad_norm": 0.9873190669247738, "learning_rate": 1.1743193076078901e-06, "loss": 0.0407, "step": 3479 }, { "epoch": 0.7918088737201365, "grad_norm": 2.2069878657118243, "learning_rate": 1.174276682662539e-06, "loss": 0.0907, "step": 3480 }, { "epoch": 0.7920364050056883, "grad_norm": 0.9695329009047923, "learning_rate": 1.1742340464909786e-06, "loss": 0.0444, "step": 3481 }, { "epoch": 0.79226393629124, "grad_norm": 1.18743378027036, "learning_rate": 1.1741913990940801e-06, "loss": 0.0417, "step": 3482 }, { "epoch": 0.7924914675767918, "grad_norm": 1.2335459566427895, "learning_rate": 1.174148740472715e-06, "loss": 0.0357, "step": 3483 }, { "epoch": 0.7927189988623435, "grad_norm": 1.7123789864089465, "learning_rate": 1.1741060706277557e-06, "loss": 0.0549, "step": 3484 }, { "epoch": 0.7929465301478953, "grad_norm": 1.0088683967525243, "learning_rate": 1.1740633895600738e-06, "loss": 0.0487, "step": 3485 }, { "epoch": 0.7931740614334472, "grad_norm": 0.9729774800349912, "learning_rate": 1.1740206972705418e-06, "loss": 0.0487, "step": 3486 }, { "epoch": 0.7934015927189989, "grad_norm": 1.357974224485339, "learning_rate": 1.1739779937600323e-06, "loss": 0.0501, "step": 3487 }, { "epoch": 0.7936291240045507, "grad_norm": 1.2508424825549256, "learning_rate": 1.1739352790294177e-06, "loss": 0.0373, "step": 3488 }, { "epoch": 0.7938566552901024, "grad_norm": 1.3686653687913515, "learning_rate": 1.1738925530795716e-06, "loss": 0.0517, "step": 3489 }, { "epoch": 0.7940841865756542, "grad_norm": 1.3919271731699832, "learning_rate": 1.173849815911367e-06, "loss": 0.0513, "step": 3490 }, { "epoch": 0.7943117178612059, "grad_norm": 1.0732983866451746, "learning_rate": 1.1738070675256773e-06, "loss": 0.0485, "step": 3491 }, { "epoch": 0.7945392491467577, "grad_norm": 0.8920921807243749, "learning_rate": 1.1737643079233763e-06, "loss": 0.0423, "step": 3492 }, { "epoch": 0.7947667804323094, "grad_norm": 0.8985878758463217, "learning_rate": 1.1737215371053376e-06, "loss": 0.0315, "step": 3493 }, { "epoch": 0.7949943117178612, "grad_norm": 1.4217871233611445, "learning_rate": 1.1736787550724357e-06, "loss": 0.0634, "step": 3494 }, { "epoch": 0.7952218430034129, "grad_norm": 1.6601308191313218, "learning_rate": 1.1736359618255452e-06, "loss": 0.072, "step": 3495 }, { "epoch": 0.7954493742889648, "grad_norm": 1.9516506015444255, "learning_rate": 1.1735931573655402e-06, "loss": 0.0845, "step": 3496 }, { "epoch": 0.7956769055745165, "grad_norm": 13.731764530376946, "learning_rate": 1.1735503416932957e-06, "loss": 0.3361, "step": 3497 }, { "epoch": 0.7959044368600683, "grad_norm": 0.8458278086989922, "learning_rate": 1.1735075148096869e-06, "loss": 0.0459, "step": 3498 }, { "epoch": 0.79613196814562, "grad_norm": 1.0325577020668653, "learning_rate": 1.173464676715589e-06, "loss": 0.0491, "step": 3499 }, { "epoch": 0.7963594994311718, "grad_norm": 1.0941809224515442, "learning_rate": 1.1734218274118775e-06, "loss": 0.0509, "step": 3500 }, { "epoch": 0.7965870307167235, "grad_norm": 1.4998412599053896, "learning_rate": 1.1733789668994285e-06, "loss": 0.0674, "step": 3501 }, { "epoch": 0.7968145620022753, "grad_norm": 1.4510084408943071, "learning_rate": 1.1733360951791176e-06, "loss": 0.078, "step": 3502 }, { "epoch": 0.797042093287827, "grad_norm": 0.8199454039615341, "learning_rate": 1.173293212251821e-06, "loss": 0.0277, "step": 3503 }, { "epoch": 0.7972696245733788, "grad_norm": 1.4797806837589549, "learning_rate": 1.1732503181184155e-06, "loss": 0.0457, "step": 3504 }, { "epoch": 0.7974971558589306, "grad_norm": 1.6954088656786952, "learning_rate": 1.1732074127797773e-06, "loss": 0.0431, "step": 3505 }, { "epoch": 0.7977246871444824, "grad_norm": 0.78592343457484, "learning_rate": 1.1731644962367838e-06, "loss": 0.0289, "step": 3506 }, { "epoch": 0.7979522184300342, "grad_norm": 1.0800031446768374, "learning_rate": 1.173121568490312e-06, "loss": 0.068, "step": 3507 }, { "epoch": 0.7981797497155859, "grad_norm": 1.4924469013770905, "learning_rate": 1.173078629541239e-06, "loss": 0.0986, "step": 3508 }, { "epoch": 0.7984072810011377, "grad_norm": 1.0339163270277598, "learning_rate": 1.1730356793904426e-06, "loss": 0.0393, "step": 3509 }, { "epoch": 0.7986348122866894, "grad_norm": 2.1650674647440002, "learning_rate": 1.1729927180388008e-06, "loss": 0.0748, "step": 3510 }, { "epoch": 0.7988623435722412, "grad_norm": 1.3477613309939473, "learning_rate": 1.172949745487191e-06, "loss": 0.0511, "step": 3511 }, { "epoch": 0.7990898748577929, "grad_norm": 1.941320037871878, "learning_rate": 1.1729067617364923e-06, "loss": 0.0653, "step": 3512 }, { "epoch": 0.7993174061433447, "grad_norm": 1.3711705573827695, "learning_rate": 1.1728637667875825e-06, "loss": 0.0583, "step": 3513 }, { "epoch": 0.7995449374288964, "grad_norm": 1.4216556543524117, "learning_rate": 1.172820760641341e-06, "loss": 0.0767, "step": 3514 }, { "epoch": 0.7997724687144482, "grad_norm": 1.0065561289408549, "learning_rate": 1.1727777432986465e-06, "loss": 0.0469, "step": 3515 }, { "epoch": 0.8, "grad_norm": 1.3114054375469062, "learning_rate": 1.1727347147603778e-06, "loss": 0.051, "step": 3516 }, { "epoch": 0.8002275312855518, "grad_norm": 1.1635377251522268, "learning_rate": 1.1726916750274148e-06, "loss": 0.0425, "step": 3517 }, { "epoch": 0.8004550625711035, "grad_norm": 0.7109505941253476, "learning_rate": 1.172648624100637e-06, "loss": 0.0335, "step": 3518 }, { "epoch": 0.8006825938566553, "grad_norm": 2.140244010424768, "learning_rate": 1.1726055619809245e-06, "loss": 0.0764, "step": 3519 }, { "epoch": 0.800910125142207, "grad_norm": 1.444555156138925, "learning_rate": 1.172562488669157e-06, "loss": 0.0618, "step": 3520 }, { "epoch": 0.8011376564277588, "grad_norm": 1.7685121918701119, "learning_rate": 1.172519404166215e-06, "loss": 0.0857, "step": 3521 }, { "epoch": 0.8013651877133106, "grad_norm": 1.713827324352964, "learning_rate": 1.1724763084729792e-06, "loss": 0.0633, "step": 3522 }, { "epoch": 0.8015927189988623, "grad_norm": 1.1050834097261653, "learning_rate": 1.1724332015903303e-06, "loss": 0.0519, "step": 3523 }, { "epoch": 0.8018202502844141, "grad_norm": 1.6510556663604237, "learning_rate": 1.1723900835191494e-06, "loss": 0.0789, "step": 3524 }, { "epoch": 0.8020477815699659, "grad_norm": 0.9667126871814873, "learning_rate": 1.1723469542603174e-06, "loss": 0.042, "step": 3525 }, { "epoch": 0.8022753128555177, "grad_norm": 0.8657263063272098, "learning_rate": 1.1723038138147165e-06, "loss": 0.0313, "step": 3526 }, { "epoch": 0.8025028441410694, "grad_norm": 1.856452808149009, "learning_rate": 1.1722606621832278e-06, "loss": 0.0896, "step": 3527 }, { "epoch": 0.8027303754266212, "grad_norm": 0.6848148476167641, "learning_rate": 1.1722174993667335e-06, "loss": 0.0183, "step": 3528 }, { "epoch": 0.8029579067121729, "grad_norm": 1.019323438363341, "learning_rate": 1.1721743253661155e-06, "loss": 0.051, "step": 3529 }, { "epoch": 0.8031854379977247, "grad_norm": 0.6693914231036064, "learning_rate": 1.1721311401822566e-06, "loss": 0.0232, "step": 3530 }, { "epoch": 0.8034129692832764, "grad_norm": 0.920308592294961, "learning_rate": 1.1720879438160391e-06, "loss": 0.0305, "step": 3531 }, { "epoch": 0.8036405005688282, "grad_norm": 1.3695873071245632, "learning_rate": 1.1720447362683458e-06, "loss": 0.045, "step": 3532 }, { "epoch": 0.8038680318543799, "grad_norm": 1.7871009286158004, "learning_rate": 1.1720015175400602e-06, "loss": 0.0834, "step": 3533 }, { "epoch": 0.8040955631399317, "grad_norm": 0.6832287827660394, "learning_rate": 1.1719582876320655e-06, "loss": 0.0263, "step": 3534 }, { "epoch": 0.8043230944254836, "grad_norm": 0.7627184412086807, "learning_rate": 1.1719150465452447e-06, "loss": 0.0262, "step": 3535 }, { "epoch": 0.8045506257110353, "grad_norm": 1.716803621327735, "learning_rate": 1.1718717942804822e-06, "loss": 0.0734, "step": 3536 }, { "epoch": 0.8047781569965871, "grad_norm": 2.0486680528761854, "learning_rate": 1.1718285308386618e-06, "loss": 0.1516, "step": 3537 }, { "epoch": 0.8050056882821388, "grad_norm": 1.3104153747023377, "learning_rate": 1.1717852562206678e-06, "loss": 0.0414, "step": 3538 }, { "epoch": 0.8052332195676906, "grad_norm": 1.5524472443736617, "learning_rate": 1.1717419704273844e-06, "loss": 0.0532, "step": 3539 }, { "epoch": 0.8054607508532423, "grad_norm": 1.8181142023265993, "learning_rate": 1.1716986734596963e-06, "loss": 0.074, "step": 3540 }, { "epoch": 0.8056882821387941, "grad_norm": 1.5770991987541603, "learning_rate": 1.1716553653184887e-06, "loss": 0.0572, "step": 3541 }, { "epoch": 0.8059158134243458, "grad_norm": 1.603046856744777, "learning_rate": 1.1716120460046464e-06, "loss": 0.068, "step": 3542 }, { "epoch": 0.8061433447098976, "grad_norm": 0.9490597103722859, "learning_rate": 1.1715687155190553e-06, "loss": 0.044, "step": 3543 }, { "epoch": 0.8063708759954493, "grad_norm": 1.4475288739954406, "learning_rate": 1.1715253738626005e-06, "loss": 0.0859, "step": 3544 }, { "epoch": 0.8065984072810012, "grad_norm": 2.4651858821386132, "learning_rate": 1.171482021036168e-06, "loss": 0.1099, "step": 3545 }, { "epoch": 0.8068259385665529, "grad_norm": 0.6592130419111449, "learning_rate": 1.1714386570406439e-06, "loss": 0.0271, "step": 3546 }, { "epoch": 0.8070534698521047, "grad_norm": 1.7063955755708686, "learning_rate": 1.1713952818769142e-06, "loss": 0.111, "step": 3547 }, { "epoch": 0.8072810011376564, "grad_norm": 2.169506097036596, "learning_rate": 1.1713518955458657e-06, "loss": 0.0733, "step": 3548 }, { "epoch": 0.8075085324232082, "grad_norm": 1.2288435774277289, "learning_rate": 1.1713084980483849e-06, "loss": 0.042, "step": 3549 }, { "epoch": 0.8077360637087599, "grad_norm": 1.577397877205651, "learning_rate": 1.1712650893853591e-06, "loss": 0.0831, "step": 3550 }, { "epoch": 0.8079635949943117, "grad_norm": 1.1381216086876993, "learning_rate": 1.1712216695576753e-06, "loss": 0.0531, "step": 3551 }, { "epoch": 0.8081911262798634, "grad_norm": 1.1404596418271893, "learning_rate": 1.171178238566221e-06, "loss": 0.043, "step": 3552 }, { "epoch": 0.8084186575654152, "grad_norm": 1.317622472587367, "learning_rate": 1.1711347964118837e-06, "loss": 0.0529, "step": 3553 }, { "epoch": 0.8086461888509671, "grad_norm": 1.065643129987413, "learning_rate": 1.1710913430955514e-06, "loss": 0.0375, "step": 3554 }, { "epoch": 0.8088737201365188, "grad_norm": 1.3409199390597606, "learning_rate": 1.171047878618112e-06, "loss": 0.0424, "step": 3555 }, { "epoch": 0.8091012514220706, "grad_norm": 1.795521953652107, "learning_rate": 1.1710044029804542e-06, "loss": 0.0649, "step": 3556 }, { "epoch": 0.8093287827076223, "grad_norm": 1.4702215735242095, "learning_rate": 1.1709609161834663e-06, "loss": 0.0714, "step": 3557 }, { "epoch": 0.8095563139931741, "grad_norm": 1.4542304303142697, "learning_rate": 1.1709174182280371e-06, "loss": 0.074, "step": 3558 }, { "epoch": 0.8097838452787258, "grad_norm": 0.9222037242585802, "learning_rate": 1.1708739091150557e-06, "loss": 0.0479, "step": 3559 }, { "epoch": 0.8100113765642776, "grad_norm": 1.1706470717553774, "learning_rate": 1.1708303888454113e-06, "loss": 0.0499, "step": 3560 }, { "epoch": 0.8102389078498293, "grad_norm": 1.4734537784856048, "learning_rate": 1.1707868574199934e-06, "loss": 0.0684, "step": 3561 }, { "epoch": 0.8104664391353811, "grad_norm": 2.40284003722661, "learning_rate": 1.1707433148396918e-06, "loss": 0.1156, "step": 3562 }, { "epoch": 0.8106939704209328, "grad_norm": 1.283544001327138, "learning_rate": 1.1706997611053963e-06, "loss": 0.041, "step": 3563 }, { "epoch": 0.8109215017064847, "grad_norm": 1.2655368943151744, "learning_rate": 1.1706561962179968e-06, "loss": 0.0517, "step": 3564 }, { "epoch": 0.8111490329920364, "grad_norm": 1.2742964978863571, "learning_rate": 1.1706126201783844e-06, "loss": 0.0601, "step": 3565 }, { "epoch": 0.8113765642775882, "grad_norm": 1.1638624014319168, "learning_rate": 1.170569032987449e-06, "loss": 0.0408, "step": 3566 }, { "epoch": 0.81160409556314, "grad_norm": 1.3521980505141613, "learning_rate": 1.1705254346460818e-06, "loss": 0.0494, "step": 3567 }, { "epoch": 0.8118316268486917, "grad_norm": 1.002566765933997, "learning_rate": 1.170481825155174e-06, "loss": 0.0608, "step": 3568 }, { "epoch": 0.8120591581342435, "grad_norm": 1.0785551144421794, "learning_rate": 1.1704382045156162e-06, "loss": 0.0443, "step": 3569 }, { "epoch": 0.8122866894197952, "grad_norm": 0.9426048538607938, "learning_rate": 1.1703945727283008e-06, "loss": 0.0293, "step": 3570 }, { "epoch": 0.812514220705347, "grad_norm": 0.9121003153189483, "learning_rate": 1.1703509297941193e-06, "loss": 0.0469, "step": 3571 }, { "epoch": 0.8127417519908987, "grad_norm": 1.1024950352965313, "learning_rate": 1.1703072757139633e-06, "loss": 0.0432, "step": 3572 }, { "epoch": 0.8129692832764505, "grad_norm": 1.5884989646335343, "learning_rate": 1.1702636104887253e-06, "loss": 0.0567, "step": 3573 }, { "epoch": 0.8131968145620023, "grad_norm": 1.3193688692811854, "learning_rate": 1.1702199341192977e-06, "loss": 0.0492, "step": 3574 }, { "epoch": 0.8134243458475541, "grad_norm": 1.278989450905917, "learning_rate": 1.1701762466065733e-06, "loss": 0.0673, "step": 3575 }, { "epoch": 0.8136518771331058, "grad_norm": 1.2742872087629682, "learning_rate": 1.1701325479514446e-06, "loss": 0.0515, "step": 3576 }, { "epoch": 0.8138794084186576, "grad_norm": 1.0146683924087292, "learning_rate": 1.1700888381548052e-06, "loss": 0.0313, "step": 3577 }, { "epoch": 0.8141069397042093, "grad_norm": 1.2456789380747535, "learning_rate": 1.1700451172175482e-06, "loss": 0.0795, "step": 3578 }, { "epoch": 0.8143344709897611, "grad_norm": 1.0481095629438555, "learning_rate": 1.1700013851405673e-06, "loss": 0.0388, "step": 3579 }, { "epoch": 0.8145620022753128, "grad_norm": 2.2650366394721293, "learning_rate": 1.169957641924756e-06, "loss": 0.0791, "step": 3580 }, { "epoch": 0.8147895335608646, "grad_norm": 1.0382772002032885, "learning_rate": 1.1699138875710087e-06, "loss": 0.0378, "step": 3581 }, { "epoch": 0.8150170648464163, "grad_norm": 1.1375187926566632, "learning_rate": 1.1698701220802194e-06, "loss": 0.046, "step": 3582 }, { "epoch": 0.8152445961319681, "grad_norm": 1.1605133417672624, "learning_rate": 1.1698263454532827e-06, "loss": 0.0563, "step": 3583 }, { "epoch": 0.81547212741752, "grad_norm": 1.382705077452947, "learning_rate": 1.1697825576910933e-06, "loss": 0.0388, "step": 3584 }, { "epoch": 0.8156996587030717, "grad_norm": 24.85184337955777, "learning_rate": 1.1697387587945461e-06, "loss": 0.4189, "step": 3585 }, { "epoch": 0.8159271899886235, "grad_norm": 1.375750049796865, "learning_rate": 1.1696949487645365e-06, "loss": 0.0746, "step": 3586 }, { "epoch": 0.8161547212741752, "grad_norm": 1.9135630568236137, "learning_rate": 1.1696511276019595e-06, "loss": 0.0963, "step": 3587 }, { "epoch": 0.816382252559727, "grad_norm": 1.4697481500847787, "learning_rate": 1.169607295307711e-06, "loss": 0.0468, "step": 3588 }, { "epoch": 0.8166097838452787, "grad_norm": 1.8024021280856608, "learning_rate": 1.169563451882687e-06, "loss": 0.0734, "step": 3589 }, { "epoch": 0.8168373151308305, "grad_norm": 1.0001763795597935, "learning_rate": 1.1695195973277831e-06, "loss": 0.0377, "step": 3590 }, { "epoch": 0.8170648464163822, "grad_norm": 1.6468051827674142, "learning_rate": 1.169475731643896e-06, "loss": 0.0757, "step": 3591 }, { "epoch": 0.817292377701934, "grad_norm": 1.1181893907591056, "learning_rate": 1.169431854831922e-06, "loss": 0.0423, "step": 3592 }, { "epoch": 0.8175199089874858, "grad_norm": 1.153273064949388, "learning_rate": 1.169387966892758e-06, "loss": 0.0697, "step": 3593 }, { "epoch": 0.8177474402730376, "grad_norm": 1.694259041690198, "learning_rate": 1.169344067827301e-06, "loss": 0.0942, "step": 3594 }, { "epoch": 0.8179749715585893, "grad_norm": 1.1843299797048426, "learning_rate": 1.1693001576364483e-06, "loss": 0.0649, "step": 3595 }, { "epoch": 0.8182025028441411, "grad_norm": 1.5290468150322878, "learning_rate": 1.1692562363210969e-06, "loss": 0.0586, "step": 3596 }, { "epoch": 0.8184300341296928, "grad_norm": 1.020107068762973, "learning_rate": 1.169212303882145e-06, "loss": 0.045, "step": 3597 }, { "epoch": 0.8186575654152446, "grad_norm": 0.8683619857454433, "learning_rate": 1.1691683603204901e-06, "loss": 0.0628, "step": 3598 }, { "epoch": 0.8188850967007963, "grad_norm": 1.2042022871006006, "learning_rate": 1.1691244056370307e-06, "loss": 0.0461, "step": 3599 }, { "epoch": 0.8191126279863481, "grad_norm": 0.9629993703014715, "learning_rate": 1.169080439832665e-06, "loss": 0.0429, "step": 3600 }, { "epoch": 0.8193401592718998, "grad_norm": 1.3152436305816528, "learning_rate": 1.1690364629082915e-06, "loss": 0.0414, "step": 3601 }, { "epoch": 0.8195676905574516, "grad_norm": 1.2407550380695793, "learning_rate": 1.168992474864809e-06, "loss": 0.0426, "step": 3602 }, { "epoch": 0.8197952218430035, "grad_norm": 1.7099317456821084, "learning_rate": 1.1689484757031167e-06, "loss": 0.0661, "step": 3603 }, { "epoch": 0.8200227531285552, "grad_norm": 2.4921680206760173, "learning_rate": 1.1689044654241136e-06, "loss": 0.0975, "step": 3604 }, { "epoch": 0.820250284414107, "grad_norm": 0.9191594230152784, "learning_rate": 1.1688604440286994e-06, "loss": 0.0427, "step": 3605 }, { "epoch": 0.8204778156996587, "grad_norm": 1.5850443718709968, "learning_rate": 1.1688164115177737e-06, "loss": 0.064, "step": 3606 }, { "epoch": 0.8207053469852105, "grad_norm": 1.4447088659158112, "learning_rate": 1.1687723678922367e-06, "loss": 0.0644, "step": 3607 }, { "epoch": 0.8209328782707622, "grad_norm": 1.1876210795397029, "learning_rate": 1.1687283131529882e-06, "loss": 0.05, "step": 3608 }, { "epoch": 0.821160409556314, "grad_norm": 1.1124560475691172, "learning_rate": 1.1686842473009286e-06, "loss": 0.052, "step": 3609 }, { "epoch": 0.8213879408418657, "grad_norm": 1.1061762100158978, "learning_rate": 1.168640170336959e-06, "loss": 0.0386, "step": 3610 }, { "epoch": 0.8216154721274175, "grad_norm": 1.44352318845533, "learning_rate": 1.16859608226198e-06, "loss": 0.063, "step": 3611 }, { "epoch": 0.8218430034129692, "grad_norm": 1.6539718566592532, "learning_rate": 1.1685519830768923e-06, "loss": 0.0874, "step": 3612 }, { "epoch": 0.8220705346985211, "grad_norm": 1.6565633171924414, "learning_rate": 1.1685078727825978e-06, "loss": 0.0589, "step": 3613 }, { "epoch": 0.8222980659840728, "grad_norm": 1.461206103355595, "learning_rate": 1.1684637513799976e-06, "loss": 0.0627, "step": 3614 }, { "epoch": 0.8225255972696246, "grad_norm": 0.9887213175076792, "learning_rate": 1.1684196188699936e-06, "loss": 0.0459, "step": 3615 }, { "epoch": 0.8227531285551763, "grad_norm": 1.3363304037819075, "learning_rate": 1.168375475253488e-06, "loss": 0.0467, "step": 3616 }, { "epoch": 0.8229806598407281, "grad_norm": 1.178055634767022, "learning_rate": 1.168331320531383e-06, "loss": 0.0471, "step": 3617 }, { "epoch": 0.8232081911262799, "grad_norm": 1.586381066611804, "learning_rate": 1.1682871547045805e-06, "loss": 0.0708, "step": 3618 }, { "epoch": 0.8234357224118316, "grad_norm": 1.209859608299182, "learning_rate": 1.1682429777739836e-06, "loss": 0.0513, "step": 3619 }, { "epoch": 0.8236632536973834, "grad_norm": 1.7764116713616094, "learning_rate": 1.168198789740495e-06, "loss": 0.0638, "step": 3620 }, { "epoch": 0.8238907849829351, "grad_norm": 0.7763254752373752, "learning_rate": 1.1681545906050184e-06, "loss": 0.0343, "step": 3621 }, { "epoch": 0.824118316268487, "grad_norm": 1.1001948497692553, "learning_rate": 1.1681103803684564e-06, "loss": 0.056, "step": 3622 }, { "epoch": 0.8243458475540387, "grad_norm": 1.273383386134855, "learning_rate": 1.168066159031713e-06, "loss": 0.0468, "step": 3623 }, { "epoch": 0.8245733788395905, "grad_norm": 1.5388101293640168, "learning_rate": 1.1680219265956918e-06, "loss": 0.0558, "step": 3624 }, { "epoch": 0.8248009101251422, "grad_norm": 1.1316172883077475, "learning_rate": 1.167977683061297e-06, "loss": 0.0434, "step": 3625 }, { "epoch": 0.825028441410694, "grad_norm": 1.3402414111306105, "learning_rate": 1.1679334284294328e-06, "loss": 0.0601, "step": 3626 }, { "epoch": 0.8252559726962457, "grad_norm": 0.8814362527241864, "learning_rate": 1.1678891627010036e-06, "loss": 0.0498, "step": 3627 }, { "epoch": 0.8254835039817975, "grad_norm": 1.1662546055909706, "learning_rate": 1.1678448858769139e-06, "loss": 0.0493, "step": 3628 }, { "epoch": 0.8257110352673492, "grad_norm": 1.8562897841170296, "learning_rate": 1.1678005979580694e-06, "loss": 0.0834, "step": 3629 }, { "epoch": 0.825938566552901, "grad_norm": 0.8447551801193321, "learning_rate": 1.1677562989453745e-06, "loss": 0.0283, "step": 3630 }, { "epoch": 0.8261660978384527, "grad_norm": 1.0072638573362171, "learning_rate": 1.1677119888397348e-06, "loss": 0.049, "step": 3631 }, { "epoch": 0.8263936291240046, "grad_norm": 1.433958970376907, "learning_rate": 1.1676676676420562e-06, "loss": 0.0601, "step": 3632 }, { "epoch": 0.8266211604095564, "grad_norm": 1.123241025021127, "learning_rate": 1.1676233353532442e-06, "loss": 0.046, "step": 3633 }, { "epoch": 0.8268486916951081, "grad_norm": 0.7445525822110918, "learning_rate": 1.167578991974205e-06, "loss": 0.0181, "step": 3634 }, { "epoch": 0.8270762229806599, "grad_norm": 0.7798344340235215, "learning_rate": 1.167534637505845e-06, "loss": 0.0272, "step": 3635 }, { "epoch": 0.8273037542662116, "grad_norm": 1.0329542174985942, "learning_rate": 1.1674902719490704e-06, "loss": 0.0283, "step": 3636 }, { "epoch": 0.8275312855517634, "grad_norm": 1.3105774584535312, "learning_rate": 1.1674458953047883e-06, "loss": 0.0768, "step": 3637 }, { "epoch": 0.8277588168373151, "grad_norm": 1.403061308791836, "learning_rate": 1.1674015075739054e-06, "loss": 0.0524, "step": 3638 }, { "epoch": 0.8279863481228669, "grad_norm": 1.525826516533729, "learning_rate": 1.1673571087573293e-06, "loss": 0.035, "step": 3639 }, { "epoch": 0.8282138794084186, "grad_norm": 1.578985773093579, "learning_rate": 1.167312698855967e-06, "loss": 0.0686, "step": 3640 }, { "epoch": 0.8284414106939704, "grad_norm": 0.8598678068916811, "learning_rate": 1.1672682778707262e-06, "loss": 0.0434, "step": 3641 }, { "epoch": 0.8286689419795222, "grad_norm": 1.1746515581399284, "learning_rate": 1.167223845802515e-06, "loss": 0.0683, "step": 3642 }, { "epoch": 0.828896473265074, "grad_norm": 1.101963672107574, "learning_rate": 1.1671794026522417e-06, "loss": 0.0446, "step": 3643 }, { "epoch": 0.8291240045506257, "grad_norm": 1.29174643335613, "learning_rate": 1.1671349484208142e-06, "loss": 0.0631, "step": 3644 }, { "epoch": 0.8293515358361775, "grad_norm": 14.688190742848866, "learning_rate": 1.1670904831091412e-06, "loss": 0.3068, "step": 3645 }, { "epoch": 0.8295790671217292, "grad_norm": 1.997677369000564, "learning_rate": 1.1670460067181313e-06, "loss": 0.097, "step": 3646 }, { "epoch": 0.829806598407281, "grad_norm": 2.0170155054153844, "learning_rate": 1.167001519248694e-06, "loss": 0.091, "step": 3647 }, { "epoch": 0.8300341296928327, "grad_norm": 1.445412257420854, "learning_rate": 1.1669570207017384e-06, "loss": 0.0451, "step": 3648 }, { "epoch": 0.8302616609783845, "grad_norm": 1.5233401207808903, "learning_rate": 1.1669125110781737e-06, "loss": 0.0606, "step": 3649 }, { "epoch": 0.8304891922639362, "grad_norm": 0.7835188426142936, "learning_rate": 1.1668679903789095e-06, "loss": 0.0332, "step": 3650 }, { "epoch": 0.830716723549488, "grad_norm": 1.4515524685426144, "learning_rate": 1.1668234586048562e-06, "loss": 0.0596, "step": 3651 }, { "epoch": 0.8309442548350399, "grad_norm": 0.9655835502300143, "learning_rate": 1.1667789157569237e-06, "loss": 0.0409, "step": 3652 }, { "epoch": 0.8311717861205916, "grad_norm": 1.393653817047434, "learning_rate": 1.1667343618360224e-06, "loss": 0.0469, "step": 3653 }, { "epoch": 0.8313993174061434, "grad_norm": 1.483284458995605, "learning_rate": 1.166689796843063e-06, "loss": 0.071, "step": 3654 }, { "epoch": 0.8316268486916951, "grad_norm": 2.0171503740210004, "learning_rate": 1.166645220778956e-06, "loss": 0.113, "step": 3655 }, { "epoch": 0.8318543799772469, "grad_norm": 1.1504742515281876, "learning_rate": 1.1666006336446127e-06, "loss": 0.0524, "step": 3656 }, { "epoch": 0.8320819112627986, "grad_norm": 45.01747083144473, "learning_rate": 1.1665560354409444e-06, "loss": 0.4523, "step": 3657 }, { "epoch": 0.8323094425483504, "grad_norm": 1.112433200629069, "learning_rate": 1.1665114261688625e-06, "loss": 0.0668, "step": 3658 }, { "epoch": 0.8325369738339021, "grad_norm": 0.8757946163066259, "learning_rate": 1.1664668058292789e-06, "loss": 0.031, "step": 3659 }, { "epoch": 0.8327645051194539, "grad_norm": 1.4231045164348157, "learning_rate": 1.1664221744231052e-06, "loss": 0.0557, "step": 3660 }, { "epoch": 0.8329920364050057, "grad_norm": 0.817775869454815, "learning_rate": 1.1663775319512539e-06, "loss": 0.033, "step": 3661 }, { "epoch": 0.8332195676905575, "grad_norm": 1.1781931600979176, "learning_rate": 1.1663328784146375e-06, "loss": 0.0642, "step": 3662 }, { "epoch": 0.8334470989761092, "grad_norm": 0.9259664097181735, "learning_rate": 1.1662882138141684e-06, "loss": 0.04, "step": 3663 }, { "epoch": 0.833674630261661, "grad_norm": 0.9223727967515527, "learning_rate": 1.1662435381507595e-06, "loss": 0.0348, "step": 3664 }, { "epoch": 0.8339021615472128, "grad_norm": 2.505531351215165, "learning_rate": 1.1661988514253237e-06, "loss": 0.0541, "step": 3665 }, { "epoch": 0.8341296928327645, "grad_norm": 0.8185437890691741, "learning_rate": 1.1661541536387748e-06, "loss": 0.0304, "step": 3666 }, { "epoch": 0.8343572241183163, "grad_norm": 1.4951738593038986, "learning_rate": 1.166109444792026e-06, "loss": 0.0621, "step": 3667 }, { "epoch": 0.834584755403868, "grad_norm": 1.3025368287248427, "learning_rate": 1.1660647248859913e-06, "loss": 0.0516, "step": 3668 }, { "epoch": 0.8348122866894198, "grad_norm": 1.2218241621957284, "learning_rate": 1.1660199939215845e-06, "loss": 0.0425, "step": 3669 }, { "epoch": 0.8350398179749715, "grad_norm": 1.3776947069628735, "learning_rate": 1.1659752518997197e-06, "loss": 0.0398, "step": 3670 }, { "epoch": 0.8352673492605234, "grad_norm": 1.1397273421298342, "learning_rate": 1.1659304988213115e-06, "loss": 0.065, "step": 3671 }, { "epoch": 0.8354948805460751, "grad_norm": 1.0434688435298627, "learning_rate": 1.1658857346872745e-06, "loss": 0.0566, "step": 3672 }, { "epoch": 0.8357224118316269, "grad_norm": 1.5440568255758365, "learning_rate": 1.165840959498524e-06, "loss": 0.0688, "step": 3673 }, { "epoch": 0.8359499431171786, "grad_norm": 1.115123935337787, "learning_rate": 1.1657961732559745e-06, "loss": 0.032, "step": 3674 }, { "epoch": 0.8361774744027304, "grad_norm": 2.009941649682246, "learning_rate": 1.1657513759605417e-06, "loss": 0.1067, "step": 3675 }, { "epoch": 0.8364050056882821, "grad_norm": 1.0878721650862986, "learning_rate": 1.1657065676131412e-06, "loss": 0.039, "step": 3676 }, { "epoch": 0.8366325369738339, "grad_norm": 1.2823447557648144, "learning_rate": 1.1656617482146886e-06, "loss": 0.0565, "step": 3677 }, { "epoch": 0.8368600682593856, "grad_norm": 0.8640896631578039, "learning_rate": 1.1656169177661e-06, "loss": 0.0286, "step": 3678 }, { "epoch": 0.8370875995449374, "grad_norm": 1.5431241138574887, "learning_rate": 1.165572076268292e-06, "loss": 0.0777, "step": 3679 }, { "epoch": 0.8373151308304891, "grad_norm": 2.182613338071724, "learning_rate": 1.1655272237221804e-06, "loss": 0.0811, "step": 3680 }, { "epoch": 0.837542662116041, "grad_norm": 1.6601496114451646, "learning_rate": 1.1654823601286826e-06, "loss": 0.0705, "step": 3681 }, { "epoch": 0.8377701934015928, "grad_norm": 0.9845739508563496, "learning_rate": 1.165437485488715e-06, "loss": 0.0352, "step": 3682 }, { "epoch": 0.8379977246871445, "grad_norm": 1.7268236108563282, "learning_rate": 1.165392599803195e-06, "loss": 0.0813, "step": 3683 }, { "epoch": 0.8382252559726963, "grad_norm": 1.7016772440974146, "learning_rate": 1.16534770307304e-06, "loss": 0.0686, "step": 3684 }, { "epoch": 0.838452787258248, "grad_norm": 1.1115991957924602, "learning_rate": 1.1653027952991676e-06, "loss": 0.0453, "step": 3685 }, { "epoch": 0.8386803185437998, "grad_norm": 1.1625036802126243, "learning_rate": 1.1652578764824953e-06, "loss": 0.0547, "step": 3686 }, { "epoch": 0.8389078498293515, "grad_norm": 1.2752562383666506, "learning_rate": 1.1652129466239417e-06, "loss": 0.0557, "step": 3687 }, { "epoch": 0.8391353811149033, "grad_norm": 1.1708119141481568, "learning_rate": 1.1651680057244247e-06, "loss": 0.0543, "step": 3688 }, { "epoch": 0.839362912400455, "grad_norm": 1.5686634859706143, "learning_rate": 1.165123053784863e-06, "loss": 0.0631, "step": 3689 }, { "epoch": 0.8395904436860068, "grad_norm": 0.7862204435509706, "learning_rate": 1.1650780908061753e-06, "loss": 0.0268, "step": 3690 }, { "epoch": 0.8398179749715586, "grad_norm": 0.8894683444434565, "learning_rate": 1.1650331167892805e-06, "loss": 0.0447, "step": 3691 }, { "epoch": 0.8400455062571104, "grad_norm": 1.1634157087855133, "learning_rate": 1.1649881317350979e-06, "loss": 0.0563, "step": 3692 }, { "epoch": 0.8402730375426621, "grad_norm": 0.8154942879262764, "learning_rate": 1.1649431356445467e-06, "loss": 0.0286, "step": 3693 }, { "epoch": 0.8405005688282139, "grad_norm": 1.377478587198256, "learning_rate": 1.1648981285185464e-06, "loss": 0.0616, "step": 3694 }, { "epoch": 0.8407281001137656, "grad_norm": 1.4564713307212853, "learning_rate": 1.1648531103580178e-06, "loss": 0.0537, "step": 3695 }, { "epoch": 0.8409556313993174, "grad_norm": 1.1442584647082312, "learning_rate": 1.1648080811638797e-06, "loss": 0.0514, "step": 3696 }, { "epoch": 0.8411831626848691, "grad_norm": 0.6059546954643502, "learning_rate": 1.1647630409370533e-06, "loss": 0.024, "step": 3697 }, { "epoch": 0.8414106939704209, "grad_norm": 0.845020155561466, "learning_rate": 1.1647179896784588e-06, "loss": 0.0297, "step": 3698 }, { "epoch": 0.8416382252559726, "grad_norm": 1.1265316983407672, "learning_rate": 1.1646729273890172e-06, "loss": 0.0385, "step": 3699 }, { "epoch": 0.8418657565415245, "grad_norm": 1.5081431023506024, "learning_rate": 1.1646278540696493e-06, "loss": 0.0791, "step": 3700 }, { "epoch": 0.8420932878270763, "grad_norm": 1.0623449631960677, "learning_rate": 1.1645827697212762e-06, "loss": 0.0366, "step": 3701 }, { "epoch": 0.842320819112628, "grad_norm": 1.0697058302573508, "learning_rate": 1.1645376743448194e-06, "loss": 0.0365, "step": 3702 }, { "epoch": 0.8425483503981798, "grad_norm": 1.6166494157045328, "learning_rate": 1.1644925679412008e-06, "loss": 0.0917, "step": 3703 }, { "epoch": 0.8427758816837315, "grad_norm": 0.8909409395783954, "learning_rate": 1.1644474505113423e-06, "loss": 0.0483, "step": 3704 }, { "epoch": 0.8430034129692833, "grad_norm": 1.2947960727076373, "learning_rate": 1.1644023220561657e-06, "loss": 0.0841, "step": 3705 }, { "epoch": 0.843230944254835, "grad_norm": 2.338681558244469, "learning_rate": 1.1643571825765935e-06, "loss": 0.0884, "step": 3706 }, { "epoch": 0.8434584755403868, "grad_norm": 1.0808444836814672, "learning_rate": 1.1643120320735481e-06, "loss": 0.0468, "step": 3707 }, { "epoch": 0.8436860068259385, "grad_norm": 3.1483674383789007, "learning_rate": 1.164266870547953e-06, "loss": 0.1461, "step": 3708 }, { "epoch": 0.8439135381114903, "grad_norm": 1.1429682418848544, "learning_rate": 1.16422169800073e-06, "loss": 0.0575, "step": 3709 }, { "epoch": 0.8441410693970421, "grad_norm": 1.4092804732703959, "learning_rate": 1.1641765144328035e-06, "loss": 0.0824, "step": 3710 }, { "epoch": 0.8443686006825939, "grad_norm": 0.9907223592123003, "learning_rate": 1.1641313198450966e-06, "loss": 0.0408, "step": 3711 }, { "epoch": 0.8445961319681456, "grad_norm": 1.1953791862365195, "learning_rate": 1.1640861142385326e-06, "loss": 0.0417, "step": 3712 }, { "epoch": 0.8448236632536974, "grad_norm": 1.0300444272783411, "learning_rate": 1.1640408976140358e-06, "loss": 0.0292, "step": 3713 }, { "epoch": 0.8450511945392492, "grad_norm": 0.9396034384804057, "learning_rate": 1.1639956699725303e-06, "loss": 0.0251, "step": 3714 }, { "epoch": 0.8452787258248009, "grad_norm": 1.0450241907142919, "learning_rate": 1.1639504313149403e-06, "loss": 0.0427, "step": 3715 }, { "epoch": 0.8455062571103527, "grad_norm": 1.2352022092638624, "learning_rate": 1.1639051816421906e-06, "loss": 0.0622, "step": 3716 }, { "epoch": 0.8457337883959044, "grad_norm": 1.7465197987251446, "learning_rate": 1.163859920955206e-06, "loss": 0.0646, "step": 3717 }, { "epoch": 0.8459613196814562, "grad_norm": 1.1198862623805916, "learning_rate": 1.1638146492549112e-06, "loss": 0.057, "step": 3718 }, { "epoch": 0.8461888509670079, "grad_norm": 0.9891506478501289, "learning_rate": 1.163769366542232e-06, "loss": 0.043, "step": 3719 }, { "epoch": 0.8464163822525598, "grad_norm": 0.8307264131023004, "learning_rate": 1.1637240728180937e-06, "loss": 0.0374, "step": 3720 }, { "epoch": 0.8466439135381115, "grad_norm": 1.2015516317579826, "learning_rate": 1.163678768083422e-06, "loss": 0.0542, "step": 3721 }, { "epoch": 0.8468714448236633, "grad_norm": 0.8651705828559965, "learning_rate": 1.1636334523391426e-06, "loss": 0.0333, "step": 3722 }, { "epoch": 0.847098976109215, "grad_norm": 1.8721051492604341, "learning_rate": 1.1635881255861821e-06, "loss": 0.1021, "step": 3723 }, { "epoch": 0.8473265073947668, "grad_norm": 1.7947647079315026, "learning_rate": 1.1635427878254667e-06, "loss": 0.0741, "step": 3724 }, { "epoch": 0.8475540386803185, "grad_norm": 2.1673257995211506, "learning_rate": 1.163497439057923e-06, "loss": 0.084, "step": 3725 }, { "epoch": 0.8477815699658703, "grad_norm": 1.1269894105236045, "learning_rate": 1.1634520792844778e-06, "loss": 0.0675, "step": 3726 }, { "epoch": 0.848009101251422, "grad_norm": 1.205510809010935, "learning_rate": 1.1634067085060583e-06, "loss": 0.0495, "step": 3727 }, { "epoch": 0.8482366325369738, "grad_norm": 0.691849376485242, "learning_rate": 1.1633613267235915e-06, "loss": 0.0466, "step": 3728 }, { "epoch": 0.8484641638225257, "grad_norm": 1.7367118780028674, "learning_rate": 1.1633159339380054e-06, "loss": 0.0796, "step": 3729 }, { "epoch": 0.8486916951080774, "grad_norm": 1.0717660762894743, "learning_rate": 1.1632705301502272e-06, "loss": 0.04, "step": 3730 }, { "epoch": 0.8489192263936292, "grad_norm": 1.5860944536440218, "learning_rate": 1.1632251153611855e-06, "loss": 0.0494, "step": 3731 }, { "epoch": 0.8491467576791809, "grad_norm": 1.8503643346395675, "learning_rate": 1.163179689571808e-06, "loss": 0.0747, "step": 3732 }, { "epoch": 0.8493742889647327, "grad_norm": 1.51685414443071, "learning_rate": 1.1631342527830234e-06, "loss": 0.0605, "step": 3733 }, { "epoch": 0.8496018202502844, "grad_norm": 5.054888475581482, "learning_rate": 1.16308880499576e-06, "loss": 0.1453, "step": 3734 }, { "epoch": 0.8498293515358362, "grad_norm": 0.9938442986727953, "learning_rate": 1.163043346210947e-06, "loss": 0.0402, "step": 3735 }, { "epoch": 0.8500568828213879, "grad_norm": 1.5560463292525197, "learning_rate": 1.1629978764295133e-06, "loss": 0.0735, "step": 3736 }, { "epoch": 0.8502844141069397, "grad_norm": 1.6867577558544848, "learning_rate": 1.1629523956523883e-06, "loss": 0.0488, "step": 3737 }, { "epoch": 0.8505119453924914, "grad_norm": 1.0404339887342127, "learning_rate": 1.1629069038805018e-06, "loss": 0.0328, "step": 3738 }, { "epoch": 0.8507394766780433, "grad_norm": 1.2922834659902496, "learning_rate": 1.162861401114783e-06, "loss": 0.0643, "step": 3739 }, { "epoch": 0.850967007963595, "grad_norm": 1.0645197047514645, "learning_rate": 1.1628158873561624e-06, "loss": 0.0527, "step": 3740 }, { "epoch": 0.8511945392491468, "grad_norm": 1.2986459720013372, "learning_rate": 1.16277036260557e-06, "loss": 0.0562, "step": 3741 }, { "epoch": 0.8514220705346985, "grad_norm": 0.9856861581794601, "learning_rate": 1.1627248268639363e-06, "loss": 0.0395, "step": 3742 }, { "epoch": 0.8516496018202503, "grad_norm": 1.0675074473205326, "learning_rate": 1.1626792801321917e-06, "loss": 0.0675, "step": 3743 }, { "epoch": 0.851877133105802, "grad_norm": 2.38754548600943, "learning_rate": 1.1626337224112676e-06, "loss": 0.0971, "step": 3744 }, { "epoch": 0.8521046643913538, "grad_norm": 1.7266469693034243, "learning_rate": 1.1625881537020948e-06, "loss": 0.0671, "step": 3745 }, { "epoch": 0.8523321956769055, "grad_norm": 1.280551624533034, "learning_rate": 1.1625425740056046e-06, "loss": 0.0377, "step": 3746 }, { "epoch": 0.8525597269624573, "grad_norm": 1.237081721406354, "learning_rate": 1.1624969833227287e-06, "loss": 0.059, "step": 3747 }, { "epoch": 0.852787258248009, "grad_norm": 0.9836714855093116, "learning_rate": 1.1624513816543988e-06, "loss": 0.0362, "step": 3748 }, { "epoch": 0.8530147895335609, "grad_norm": 1.9867742968928817, "learning_rate": 1.162405769001547e-06, "loss": 0.0645, "step": 3749 }, { "epoch": 0.8532423208191127, "grad_norm": 0.8182858794812171, "learning_rate": 1.1623601453651053e-06, "loss": 0.0355, "step": 3750 }, { "epoch": 0.8534698521046644, "grad_norm": 1.4917273253143806, "learning_rate": 1.1623145107460065e-06, "loss": 0.0503, "step": 3751 }, { "epoch": 0.8536973833902162, "grad_norm": 0.7287193210043196, "learning_rate": 1.1622688651451833e-06, "loss": 0.0307, "step": 3752 }, { "epoch": 0.8539249146757679, "grad_norm": 1.4100424416120585, "learning_rate": 1.1622232085635683e-06, "loss": 0.0826, "step": 3753 }, { "epoch": 0.8541524459613197, "grad_norm": 1.2920863855722, "learning_rate": 1.162177541002095e-06, "loss": 0.057, "step": 3754 }, { "epoch": 0.8543799772468714, "grad_norm": 0.9760542142070106, "learning_rate": 1.162131862461696e-06, "loss": 0.0424, "step": 3755 }, { "epoch": 0.8546075085324232, "grad_norm": 1.1155922665776505, "learning_rate": 1.1620861729433062e-06, "loss": 0.0548, "step": 3756 }, { "epoch": 0.8548350398179749, "grad_norm": 2.0265637739770646, "learning_rate": 1.1620404724478582e-06, "loss": 0.1044, "step": 3757 }, { "epoch": 0.8550625711035267, "grad_norm": 1.2011014044384718, "learning_rate": 1.1619947609762867e-06, "loss": 0.0564, "step": 3758 }, { "epoch": 0.8552901023890785, "grad_norm": 6.130934132803181, "learning_rate": 1.1619490385295255e-06, "loss": 0.2002, "step": 3759 }, { "epoch": 0.8555176336746303, "grad_norm": 1.113651539392019, "learning_rate": 1.1619033051085094e-06, "loss": 0.0652, "step": 3760 }, { "epoch": 0.855745164960182, "grad_norm": 1.286325265245382, "learning_rate": 1.161857560714173e-06, "loss": 0.0588, "step": 3761 }, { "epoch": 0.8559726962457338, "grad_norm": 1.7600474072446048, "learning_rate": 1.1618118053474514e-06, "loss": 0.0893, "step": 3762 }, { "epoch": 0.8562002275312856, "grad_norm": 1.6589882412706265, "learning_rate": 1.1617660390092794e-06, "loss": 0.0745, "step": 3763 }, { "epoch": 0.8564277588168373, "grad_norm": 1.2790558914988392, "learning_rate": 1.161720261700593e-06, "loss": 0.057, "step": 3764 }, { "epoch": 0.856655290102389, "grad_norm": 1.0102177907448584, "learning_rate": 1.161674473422327e-06, "loss": 0.0413, "step": 3765 }, { "epoch": 0.8568828213879408, "grad_norm": 1.5326235862910704, "learning_rate": 1.1616286741754178e-06, "loss": 0.0605, "step": 3766 }, { "epoch": 0.8571103526734926, "grad_norm": 0.8465015567779428, "learning_rate": 1.1615828639608013e-06, "loss": 0.0312, "step": 3767 }, { "epoch": 0.8573378839590444, "grad_norm": 1.5793350610162067, "learning_rate": 1.1615370427794138e-06, "loss": 0.0945, "step": 3768 }, { "epoch": 0.8575654152445962, "grad_norm": 1.2162812251052018, "learning_rate": 1.1614912106321916e-06, "loss": 0.0752, "step": 3769 }, { "epoch": 0.8577929465301479, "grad_norm": 1.8496686418444572, "learning_rate": 1.1614453675200718e-06, "loss": 0.0722, "step": 3770 }, { "epoch": 0.8580204778156997, "grad_norm": 1.6913533101626717, "learning_rate": 1.1613995134439912e-06, "loss": 0.0725, "step": 3771 }, { "epoch": 0.8582480091012514, "grad_norm": 3.096645898390307, "learning_rate": 1.1613536484048866e-06, "loss": 0.1055, "step": 3772 }, { "epoch": 0.8584755403868032, "grad_norm": 1.6446149120178801, "learning_rate": 1.161307772403696e-06, "loss": 0.0597, "step": 3773 }, { "epoch": 0.8587030716723549, "grad_norm": 1.5393305823965842, "learning_rate": 1.1612618854413566e-06, "loss": 0.0732, "step": 3774 }, { "epoch": 0.8589306029579067, "grad_norm": 1.9421552759578808, "learning_rate": 1.1612159875188065e-06, "loss": 0.0693, "step": 3775 }, { "epoch": 0.8591581342434584, "grad_norm": 1.3707910624756399, "learning_rate": 1.1611700786369835e-06, "loss": 0.0585, "step": 3776 }, { "epoch": 0.8593856655290102, "grad_norm": 1.5332539464674453, "learning_rate": 1.161124158796826e-06, "loss": 0.0679, "step": 3777 }, { "epoch": 0.859613196814562, "grad_norm": 0.8257875510353138, "learning_rate": 1.1610782279992728e-06, "loss": 0.0405, "step": 3778 }, { "epoch": 0.8598407281001138, "grad_norm": 0.8774289661633246, "learning_rate": 1.161032286245262e-06, "loss": 0.0303, "step": 3779 }, { "epoch": 0.8600682593856656, "grad_norm": 2.167218867354055, "learning_rate": 1.1609863335357332e-06, "loss": 0.077, "step": 3780 }, { "epoch": 0.8602957906712173, "grad_norm": 1.8941215544947247, "learning_rate": 1.1609403698716255e-06, "loss": 0.0786, "step": 3781 }, { "epoch": 0.8605233219567691, "grad_norm": 1.1276115513513825, "learning_rate": 1.160894395253878e-06, "loss": 0.0537, "step": 3782 }, { "epoch": 0.8607508532423208, "grad_norm": 1.7706734723923294, "learning_rate": 1.1608484096834306e-06, "loss": 0.0922, "step": 3783 }, { "epoch": 0.8609783845278726, "grad_norm": 1.2558738456157548, "learning_rate": 1.1608024131612231e-06, "loss": 0.0652, "step": 3784 }, { "epoch": 0.8612059158134243, "grad_norm": 0.9032710072829923, "learning_rate": 1.1607564056881953e-06, "loss": 0.0379, "step": 3785 }, { "epoch": 0.8614334470989761, "grad_norm": 1.190052750152249, "learning_rate": 1.160710387265288e-06, "loss": 0.0444, "step": 3786 }, { "epoch": 0.8616609783845278, "grad_norm": 0.9786079939895075, "learning_rate": 1.1606643578934414e-06, "loss": 0.0418, "step": 3787 }, { "epoch": 0.8618885096700797, "grad_norm": 2.273287744213828, "learning_rate": 1.1606183175735963e-06, "loss": 0.0738, "step": 3788 }, { "epoch": 0.8621160409556314, "grad_norm": 1.0946877128994326, "learning_rate": 1.1605722663066938e-06, "loss": 0.0473, "step": 3789 }, { "epoch": 0.8623435722411832, "grad_norm": 1.3819454647043339, "learning_rate": 1.1605262040936752e-06, "loss": 0.0735, "step": 3790 }, { "epoch": 0.8625711035267349, "grad_norm": 1.263046575159459, "learning_rate": 1.1604801309354815e-06, "loss": 0.0595, "step": 3791 }, { "epoch": 0.8627986348122867, "grad_norm": 1.0133652978063783, "learning_rate": 1.1604340468330546e-06, "loss": 0.0307, "step": 3792 }, { "epoch": 0.8630261660978384, "grad_norm": 1.2379613127015137, "learning_rate": 1.1603879517873366e-06, "loss": 0.0517, "step": 3793 }, { "epoch": 0.8632536973833902, "grad_norm": 1.1750697280707245, "learning_rate": 1.160341845799269e-06, "loss": 0.0482, "step": 3794 }, { "epoch": 0.863481228668942, "grad_norm": 1.6791780297526506, "learning_rate": 1.160295728869795e-06, "loss": 0.1042, "step": 3795 }, { "epoch": 0.8637087599544937, "grad_norm": 1.1451895610966802, "learning_rate": 1.1602496009998562e-06, "loss": 0.0483, "step": 3796 }, { "epoch": 0.8639362912400455, "grad_norm": 0.737910160878505, "learning_rate": 1.160203462190396e-06, "loss": 0.0264, "step": 3797 }, { "epoch": 0.8641638225255973, "grad_norm": 1.0917499952318759, "learning_rate": 1.1601573124423573e-06, "loss": 0.0553, "step": 3798 }, { "epoch": 0.8643913538111491, "grad_norm": 1.4396023675409693, "learning_rate": 1.1601111517566831e-06, "loss": 0.0593, "step": 3799 }, { "epoch": 0.8646188850967008, "grad_norm": 1.302598540309529, "learning_rate": 1.1600649801343173e-06, "loss": 0.0761, "step": 3800 }, { "epoch": 0.8648464163822526, "grad_norm": 0.9487038984159104, "learning_rate": 1.1600187975762029e-06, "loss": 0.0533, "step": 3801 }, { "epoch": 0.8650739476678043, "grad_norm": 0.8318957052334727, "learning_rate": 1.159972604083284e-06, "loss": 0.0281, "step": 3802 }, { "epoch": 0.8653014789533561, "grad_norm": 1.2107437712653222, "learning_rate": 1.159926399656505e-06, "loss": 0.0465, "step": 3803 }, { "epoch": 0.8655290102389078, "grad_norm": 1.4023241592897753, "learning_rate": 1.1598801842968103e-06, "loss": 0.0395, "step": 3804 }, { "epoch": 0.8657565415244596, "grad_norm": 1.021600867386226, "learning_rate": 1.1598339580051439e-06, "loss": 0.042, "step": 3805 }, { "epoch": 0.8659840728100113, "grad_norm": 1.0092710955033857, "learning_rate": 1.159787720782451e-06, "loss": 0.0513, "step": 3806 }, { "epoch": 0.8662116040955632, "grad_norm": 1.517102557965297, "learning_rate": 1.1597414726296764e-06, "loss": 0.0823, "step": 3807 }, { "epoch": 0.8664391353811149, "grad_norm": 1.354754585750691, "learning_rate": 1.1596952135477656e-06, "loss": 0.0619, "step": 3808 }, { "epoch": 0.8666666666666667, "grad_norm": 2.107511992431661, "learning_rate": 1.1596489435376638e-06, "loss": 0.1053, "step": 3809 }, { "epoch": 0.8668941979522184, "grad_norm": 1.660224356687729, "learning_rate": 1.1596026626003168e-06, "loss": 0.1028, "step": 3810 }, { "epoch": 0.8671217292377702, "grad_norm": 1.1996277276445466, "learning_rate": 1.1595563707366705e-06, "loss": 0.0802, "step": 3811 }, { "epoch": 0.867349260523322, "grad_norm": 1.3676154758246724, "learning_rate": 1.1595100679476707e-06, "loss": 0.0662, "step": 3812 }, { "epoch": 0.8675767918088737, "grad_norm": 1.2961263261205425, "learning_rate": 1.1594637542342644e-06, "loss": 0.0569, "step": 3813 }, { "epoch": 0.8678043230944255, "grad_norm": 1.0638382967882807, "learning_rate": 1.1594174295973976e-06, "loss": 0.0527, "step": 3814 }, { "epoch": 0.8680318543799772, "grad_norm": 1.2853107355416882, "learning_rate": 1.1593710940380172e-06, "loss": 0.053, "step": 3815 }, { "epoch": 0.868259385665529, "grad_norm": 1.4112153541469374, "learning_rate": 1.1593247475570704e-06, "loss": 0.0652, "step": 3816 }, { "epoch": 0.8684869169510808, "grad_norm": 0.6600043862976881, "learning_rate": 1.1592783901555043e-06, "loss": 0.0235, "step": 3817 }, { "epoch": 0.8687144482366326, "grad_norm": 1.483562739056723, "learning_rate": 1.1592320218342665e-06, "loss": 0.0635, "step": 3818 }, { "epoch": 0.8689419795221843, "grad_norm": 1.0821712914528594, "learning_rate": 1.1591856425943044e-06, "loss": 0.0576, "step": 3819 }, { "epoch": 0.8691695108077361, "grad_norm": 1.2023944040312258, "learning_rate": 1.159139252436566e-06, "loss": 0.0448, "step": 3820 }, { "epoch": 0.8693970420932878, "grad_norm": 1.2261008981178048, "learning_rate": 1.1590928513619997e-06, "loss": 0.079, "step": 3821 }, { "epoch": 0.8696245733788396, "grad_norm": 0.9103768778718383, "learning_rate": 1.1590464393715536e-06, "loss": 0.0319, "step": 3822 }, { "epoch": 0.8698521046643913, "grad_norm": 1.0450339587824766, "learning_rate": 1.1590000164661763e-06, "loss": 0.0632, "step": 3823 }, { "epoch": 0.8700796359499431, "grad_norm": 1.0245424665524807, "learning_rate": 1.1589535826468168e-06, "loss": 0.0615, "step": 3824 }, { "epoch": 0.8703071672354948, "grad_norm": 1.624369945490237, "learning_rate": 1.158907137914424e-06, "loss": 0.0988, "step": 3825 }, { "epoch": 0.8705346985210466, "grad_norm": 1.1284177897194427, "learning_rate": 1.158860682269947e-06, "loss": 0.0488, "step": 3826 }, { "epoch": 0.8707622298065985, "grad_norm": 1.2276202548958175, "learning_rate": 1.1588142157143353e-06, "loss": 0.0608, "step": 3827 }, { "epoch": 0.8709897610921502, "grad_norm": 1.3822371198286005, "learning_rate": 1.1587677382485386e-06, "loss": 0.0468, "step": 3828 }, { "epoch": 0.871217292377702, "grad_norm": 1.106888167025077, "learning_rate": 1.158721249873507e-06, "loss": 0.0636, "step": 3829 }, { "epoch": 0.8714448236632537, "grad_norm": 0.9393892132586369, "learning_rate": 1.1586747505901904e-06, "loss": 0.0416, "step": 3830 }, { "epoch": 0.8716723549488055, "grad_norm": 1.1085193545572236, "learning_rate": 1.1586282403995395e-06, "loss": 0.0541, "step": 3831 }, { "epoch": 0.8718998862343572, "grad_norm": 1.345809504190513, "learning_rate": 1.1585817193025046e-06, "loss": 0.0513, "step": 3832 }, { "epoch": 0.872127417519909, "grad_norm": 0.8494118924798915, "learning_rate": 1.1585351873000365e-06, "loss": 0.03, "step": 3833 }, { "epoch": 0.8723549488054607, "grad_norm": 1.2837019051092946, "learning_rate": 1.1584886443930863e-06, "loss": 0.0529, "step": 3834 }, { "epoch": 0.8725824800910125, "grad_norm": 1.6558407244456526, "learning_rate": 1.1584420905826051e-06, "loss": 0.0833, "step": 3835 }, { "epoch": 0.8728100113765643, "grad_norm": 2.9844675595475887, "learning_rate": 1.1583955258695447e-06, "loss": 0.1321, "step": 3836 }, { "epoch": 0.8730375426621161, "grad_norm": 1.2652070426430007, "learning_rate": 1.1583489502548566e-06, "loss": 0.0795, "step": 3837 }, { "epoch": 0.8732650739476678, "grad_norm": 2.0137801774238344, "learning_rate": 1.1583023637394928e-06, "loss": 0.066, "step": 3838 }, { "epoch": 0.8734926052332196, "grad_norm": 1.2580173572147197, "learning_rate": 1.1582557663244052e-06, "loss": 0.0768, "step": 3839 }, { "epoch": 0.8737201365187713, "grad_norm": 1.300183703835354, "learning_rate": 1.1582091580105464e-06, "loss": 0.0609, "step": 3840 }, { "epoch": 0.8739476678043231, "grad_norm": 1.6278123129636566, "learning_rate": 1.158162538798869e-06, "loss": 0.0459, "step": 3841 }, { "epoch": 0.8741751990898748, "grad_norm": 1.4397538286299336, "learning_rate": 1.158115908690326e-06, "loss": 0.0895, "step": 3842 }, { "epoch": 0.8744027303754266, "grad_norm": 0.6209497855296535, "learning_rate": 1.1580692676858699e-06, "loss": 0.0383, "step": 3843 }, { "epoch": 0.8746302616609783, "grad_norm": 1.7840447917226478, "learning_rate": 1.1580226157864542e-06, "loss": 0.0761, "step": 3844 }, { "epoch": 0.8748577929465301, "grad_norm": 1.3094120438469554, "learning_rate": 1.1579759529930324e-06, "loss": 0.0403, "step": 3845 }, { "epoch": 0.875085324232082, "grad_norm": 2.141642949759066, "learning_rate": 1.1579292793065583e-06, "loss": 0.0849, "step": 3846 }, { "epoch": 0.8753128555176337, "grad_norm": 1.3320582954398776, "learning_rate": 1.157882594727986e-06, "loss": 0.0581, "step": 3847 }, { "epoch": 0.8755403868031855, "grad_norm": 2.7977842973683478, "learning_rate": 1.1578358992582689e-06, "loss": 0.0733, "step": 3848 }, { "epoch": 0.8757679180887372, "grad_norm": 1.273538584951798, "learning_rate": 1.1577891928983622e-06, "loss": 0.075, "step": 3849 }, { "epoch": 0.875995449374289, "grad_norm": 1.189871225174409, "learning_rate": 1.15774247564922e-06, "loss": 0.0533, "step": 3850 }, { "epoch": 0.8762229806598407, "grad_norm": 1.3416154016304795, "learning_rate": 1.1576957475117973e-06, "loss": 0.0687, "step": 3851 }, { "epoch": 0.8764505119453925, "grad_norm": 1.6440039914240385, "learning_rate": 1.1576490084870493e-06, "loss": 0.0573, "step": 3852 }, { "epoch": 0.8766780432309442, "grad_norm": 1.2895981463406705, "learning_rate": 1.1576022585759308e-06, "loss": 0.0626, "step": 3853 }, { "epoch": 0.876905574516496, "grad_norm": 1.0617961429852751, "learning_rate": 1.1575554977793975e-06, "loss": 0.0425, "step": 3854 }, { "epoch": 0.8771331058020477, "grad_norm": 0.7670265616309823, "learning_rate": 1.1575087260984056e-06, "loss": 0.0297, "step": 3855 }, { "epoch": 0.8773606370875996, "grad_norm": 0.9091034119562531, "learning_rate": 1.1574619435339101e-06, "loss": 0.0505, "step": 3856 }, { "epoch": 0.8775881683731513, "grad_norm": 0.721924678198625, "learning_rate": 1.157415150086868e-06, "loss": 0.0334, "step": 3857 }, { "epoch": 0.8778156996587031, "grad_norm": 1.028224217020184, "learning_rate": 1.1573683457582349e-06, "loss": 0.0504, "step": 3858 }, { "epoch": 0.8780432309442548, "grad_norm": 0.9265903973410824, "learning_rate": 1.157321530548968e-06, "loss": 0.0384, "step": 3859 }, { "epoch": 0.8782707622298066, "grad_norm": 1.0369597512628028, "learning_rate": 1.157274704460024e-06, "loss": 0.0421, "step": 3860 }, { "epoch": 0.8784982935153584, "grad_norm": 0.8116397369448615, "learning_rate": 1.1572278674923598e-06, "loss": 0.0382, "step": 3861 }, { "epoch": 0.8787258248009101, "grad_norm": 1.366642693958556, "learning_rate": 1.1571810196469326e-06, "loss": 0.0886, "step": 3862 }, { "epoch": 0.8789533560864619, "grad_norm": 0.9890955350681031, "learning_rate": 1.1571341609247003e-06, "loss": 0.0449, "step": 3863 }, { "epoch": 0.8791808873720136, "grad_norm": 1.4331436845492822, "learning_rate": 1.1570872913266202e-06, "loss": 0.0823, "step": 3864 }, { "epoch": 0.8794084186575654, "grad_norm": 0.9885779670626057, "learning_rate": 1.1570404108536501e-06, "loss": 0.0291, "step": 3865 }, { "epoch": 0.8796359499431172, "grad_norm": 1.0999651301917803, "learning_rate": 1.1569935195067487e-06, "loss": 0.041, "step": 3866 }, { "epoch": 0.879863481228669, "grad_norm": 1.1538168389141554, "learning_rate": 1.1569466172868737e-06, "loss": 0.0308, "step": 3867 }, { "epoch": 0.8800910125142207, "grad_norm": 0.737445292579344, "learning_rate": 1.1568997041949843e-06, "loss": 0.0304, "step": 3868 }, { "epoch": 0.8803185437997725, "grad_norm": 1.0078853967937846, "learning_rate": 1.156852780232039e-06, "loss": 0.0373, "step": 3869 }, { "epoch": 0.8805460750853242, "grad_norm": 1.1880834513401282, "learning_rate": 1.156805845398997e-06, "loss": 0.0483, "step": 3870 }, { "epoch": 0.880773606370876, "grad_norm": 1.0071545839978109, "learning_rate": 1.1567588996968173e-06, "loss": 0.0379, "step": 3871 }, { "epoch": 0.8810011376564277, "grad_norm": 1.0549498497637813, "learning_rate": 1.1567119431264598e-06, "loss": 0.058, "step": 3872 }, { "epoch": 0.8812286689419795, "grad_norm": 1.5102339027747615, "learning_rate": 1.156664975688884e-06, "loss": 0.0607, "step": 3873 }, { "epoch": 0.8814562002275312, "grad_norm": 1.8905446477817547, "learning_rate": 1.1566179973850496e-06, "loss": 0.11, "step": 3874 }, { "epoch": 0.8816837315130831, "grad_norm": 0.6411112129672515, "learning_rate": 1.156571008215917e-06, "loss": 0.0307, "step": 3875 }, { "epoch": 0.8819112627986349, "grad_norm": 0.748342490109174, "learning_rate": 1.1565240081824466e-06, "loss": 0.0284, "step": 3876 }, { "epoch": 0.8821387940841866, "grad_norm": 1.354485160913724, "learning_rate": 1.1564769972855987e-06, "loss": 0.0491, "step": 3877 }, { "epoch": 0.8823663253697384, "grad_norm": 1.0100612900654617, "learning_rate": 1.1564299755263345e-06, "loss": 0.061, "step": 3878 }, { "epoch": 0.8825938566552901, "grad_norm": 1.5275951944341757, "learning_rate": 1.1563829429056148e-06, "loss": 0.0997, "step": 3879 }, { "epoch": 0.8828213879408419, "grad_norm": 1.2004586776908042, "learning_rate": 1.156335899424401e-06, "loss": 0.0407, "step": 3880 }, { "epoch": 0.8830489192263936, "grad_norm": 1.191033638879632, "learning_rate": 1.1562888450836544e-06, "loss": 0.0705, "step": 3881 }, { "epoch": 0.8832764505119454, "grad_norm": 2.5519305254154117, "learning_rate": 1.156241779884337e-06, "loss": 0.0995, "step": 3882 }, { "epoch": 0.8835039817974971, "grad_norm": 1.1448529875713787, "learning_rate": 1.1561947038274104e-06, "loss": 0.0409, "step": 3883 }, { "epoch": 0.8837315130830489, "grad_norm": 1.3287084471615846, "learning_rate": 1.156147616913837e-06, "loss": 0.0658, "step": 3884 }, { "epoch": 0.8839590443686007, "grad_norm": 1.607480285345695, "learning_rate": 1.156100519144579e-06, "loss": 0.065, "step": 3885 }, { "epoch": 0.8841865756541525, "grad_norm": 0.8350311325241287, "learning_rate": 1.156053410520599e-06, "loss": 0.028, "step": 3886 }, { "epoch": 0.8844141069397042, "grad_norm": 1.2493913843113744, "learning_rate": 1.15600629104286e-06, "loss": 0.054, "step": 3887 }, { "epoch": 0.884641638225256, "grad_norm": 0.8917197871451082, "learning_rate": 1.1559591607123248e-06, "loss": 0.0529, "step": 3888 }, { "epoch": 0.8848691695108077, "grad_norm": 0.9543671098376685, "learning_rate": 1.1559120195299566e-06, "loss": 0.0482, "step": 3889 }, { "epoch": 0.8850967007963595, "grad_norm": 2.016256579364082, "learning_rate": 1.1558648674967191e-06, "loss": 0.0989, "step": 3890 }, { "epoch": 0.8853242320819112, "grad_norm": 1.1851720940142052, "learning_rate": 1.1558177046135761e-06, "loss": 0.0478, "step": 3891 }, { "epoch": 0.885551763367463, "grad_norm": 1.4662655950086683, "learning_rate": 1.1557705308814914e-06, "loss": 0.058, "step": 3892 }, { "epoch": 0.8857792946530147, "grad_norm": 0.7173696835484785, "learning_rate": 1.155723346301429e-06, "loss": 0.0316, "step": 3893 }, { "epoch": 0.8860068259385665, "grad_norm": 1.2706849402987643, "learning_rate": 1.1556761508743532e-06, "loss": 0.0822, "step": 3894 }, { "epoch": 0.8862343572241184, "grad_norm": 1.609473277966293, "learning_rate": 1.1556289446012292e-06, "loss": 0.0687, "step": 3895 }, { "epoch": 0.8864618885096701, "grad_norm": 1.388883872823661, "learning_rate": 1.155581727483021e-06, "loss": 0.0583, "step": 3896 }, { "epoch": 0.8866894197952219, "grad_norm": 1.2281956278668817, "learning_rate": 1.1555344995206941e-06, "loss": 0.0627, "step": 3897 }, { "epoch": 0.8869169510807736, "grad_norm": 0.814456870112058, "learning_rate": 1.1554872607152138e-06, "loss": 0.0294, "step": 3898 }, { "epoch": 0.8871444823663254, "grad_norm": 2.0519731112865855, "learning_rate": 1.1554400110675453e-06, "loss": 0.054, "step": 3899 }, { "epoch": 0.8873720136518771, "grad_norm": 0.61176670012776, "learning_rate": 1.1553927505786543e-06, "loss": 0.0195, "step": 3900 }, { "epoch": 0.8875995449374289, "grad_norm": 1.0712439727415621, "learning_rate": 1.1553454792495072e-06, "loss": 0.0397, "step": 3901 }, { "epoch": 0.8878270762229806, "grad_norm": 1.1593590248854935, "learning_rate": 1.1552981970810694e-06, "loss": 0.0551, "step": 3902 }, { "epoch": 0.8880546075085324, "grad_norm": 1.082392405172485, "learning_rate": 1.1552509040743078e-06, "loss": 0.0423, "step": 3903 }, { "epoch": 0.8882821387940842, "grad_norm": 1.2264806717691155, "learning_rate": 1.1552036002301891e-06, "loss": 0.0614, "step": 3904 }, { "epoch": 0.888509670079636, "grad_norm": 1.8769370812382, "learning_rate": 1.1551562855496796e-06, "loss": 0.1051, "step": 3905 }, { "epoch": 0.8887372013651877, "grad_norm": 0.8452780282160288, "learning_rate": 1.1551089600337465e-06, "loss": 0.0456, "step": 3906 }, { "epoch": 0.8889647326507395, "grad_norm": 1.0073082653206435, "learning_rate": 1.1550616236833574e-06, "loss": 0.0527, "step": 3907 }, { "epoch": 0.8891922639362912, "grad_norm": 1.4600634269759383, "learning_rate": 1.155014276499479e-06, "loss": 0.0589, "step": 3908 }, { "epoch": 0.889419795221843, "grad_norm": 0.9260386118996091, "learning_rate": 1.1549669184830796e-06, "loss": 0.0296, "step": 3909 }, { "epoch": 0.8896473265073948, "grad_norm": 1.4635888779797275, "learning_rate": 1.1549195496351271e-06, "loss": 0.092, "step": 3910 }, { "epoch": 0.8898748577929465, "grad_norm": 1.0776139556085673, "learning_rate": 1.1548721699565896e-06, "loss": 0.0517, "step": 3911 }, { "epoch": 0.8901023890784983, "grad_norm": 1.9827438270104123, "learning_rate": 1.1548247794484353e-06, "loss": 0.0953, "step": 3912 }, { "epoch": 0.89032992036405, "grad_norm": 1.0572692038363223, "learning_rate": 1.1547773781116326e-06, "loss": 0.0493, "step": 3913 }, { "epoch": 0.8905574516496019, "grad_norm": 0.9208219945202761, "learning_rate": 1.1547299659471509e-06, "loss": 0.0347, "step": 3914 }, { "epoch": 0.8907849829351536, "grad_norm": 0.7636650309650683, "learning_rate": 1.1546825429559585e-06, "loss": 0.0394, "step": 3915 }, { "epoch": 0.8910125142207054, "grad_norm": 1.8570270747724098, "learning_rate": 1.1546351091390253e-06, "loss": 0.0756, "step": 3916 }, { "epoch": 0.8912400455062571, "grad_norm": 2.2137237056291625, "learning_rate": 1.1545876644973202e-06, "loss": 0.0552, "step": 3917 }, { "epoch": 0.8914675767918089, "grad_norm": 1.314419860717642, "learning_rate": 1.1545402090318133e-06, "loss": 0.0746, "step": 3918 }, { "epoch": 0.8916951080773606, "grad_norm": 1.0660978255955464, "learning_rate": 1.1544927427434743e-06, "loss": 0.0544, "step": 3919 }, { "epoch": 0.8919226393629124, "grad_norm": 1.2951418524657141, "learning_rate": 1.1544452656332733e-06, "loss": 0.0585, "step": 3920 }, { "epoch": 0.8921501706484641, "grad_norm": 1.266569155658392, "learning_rate": 1.1543977777021808e-06, "loss": 0.0593, "step": 3921 }, { "epoch": 0.8923777019340159, "grad_norm": 1.0985430730465215, "learning_rate": 1.1543502789511671e-06, "loss": 0.057, "step": 3922 }, { "epoch": 0.8926052332195676, "grad_norm": 1.0051340610905692, "learning_rate": 1.1543027693812033e-06, "loss": 0.044, "step": 3923 }, { "epoch": 0.8928327645051195, "grad_norm": 0.8979646754734382, "learning_rate": 1.15425524899326e-06, "loss": 0.0381, "step": 3924 }, { "epoch": 0.8930602957906713, "grad_norm": 1.8955096750982607, "learning_rate": 1.154207717788309e-06, "loss": 0.0698, "step": 3925 }, { "epoch": 0.893287827076223, "grad_norm": 0.7837134343494108, "learning_rate": 1.1541601757673216e-06, "loss": 0.021, "step": 3926 }, { "epoch": 0.8935153583617748, "grad_norm": 1.406918667470208, "learning_rate": 1.154112622931269e-06, "loss": 0.0633, "step": 3927 }, { "epoch": 0.8937428896473265, "grad_norm": 1.6626757430888262, "learning_rate": 1.1540650592811233e-06, "loss": 0.0802, "step": 3928 }, { "epoch": 0.8939704209328783, "grad_norm": 2.7246018561325025, "learning_rate": 1.154017484817857e-06, "loss": 0.1196, "step": 3929 }, { "epoch": 0.89419795221843, "grad_norm": 1.3964197751853908, "learning_rate": 1.1539698995424423e-06, "loss": 0.0619, "step": 3930 }, { "epoch": 0.8944254835039818, "grad_norm": 1.2631409164792344, "learning_rate": 1.1539223034558513e-06, "loss": 0.0513, "step": 3931 }, { "epoch": 0.8946530147895335, "grad_norm": 1.0255766384543599, "learning_rate": 1.1538746965590572e-06, "loss": 0.0428, "step": 3932 }, { "epoch": 0.8948805460750853, "grad_norm": 0.4848782114580742, "learning_rate": 1.153827078853033e-06, "loss": 0.0267, "step": 3933 }, { "epoch": 0.8951080773606371, "grad_norm": 1.0020187672794154, "learning_rate": 1.1537794503387516e-06, "loss": 0.0494, "step": 3934 }, { "epoch": 0.8953356086461889, "grad_norm": 1.83637730371515, "learning_rate": 1.1537318110171867e-06, "loss": 0.0759, "step": 3935 }, { "epoch": 0.8955631399317406, "grad_norm": 1.9684228773442944, "learning_rate": 1.153684160889312e-06, "loss": 0.1322, "step": 3936 }, { "epoch": 0.8957906712172924, "grad_norm": 2.0616350638849172, "learning_rate": 1.1536364999561011e-06, "loss": 0.0855, "step": 3937 }, { "epoch": 0.8960182025028441, "grad_norm": 4.2442677487614, "learning_rate": 1.1535888282185283e-06, "loss": 0.1942, "step": 3938 }, { "epoch": 0.8962457337883959, "grad_norm": 1.5621794154117725, "learning_rate": 1.1535411456775682e-06, "loss": 0.0611, "step": 3939 }, { "epoch": 0.8964732650739476, "grad_norm": 1.0048473835575604, "learning_rate": 1.1534934523341952e-06, "loss": 0.0484, "step": 3940 }, { "epoch": 0.8967007963594994, "grad_norm": 0.9747410927457769, "learning_rate": 1.1534457481893834e-06, "loss": 0.0321, "step": 3941 }, { "epoch": 0.8969283276450511, "grad_norm": 0.8739461699351664, "learning_rate": 1.1533980332441085e-06, "loss": 0.0363, "step": 3942 }, { "epoch": 0.897155858930603, "grad_norm": 1.1216851883585208, "learning_rate": 1.1533503074993455e-06, "loss": 0.0423, "step": 3943 }, { "epoch": 0.8973833902161548, "grad_norm": 0.9832239028189882, "learning_rate": 1.15330257095607e-06, "loss": 0.0458, "step": 3944 }, { "epoch": 0.8976109215017065, "grad_norm": 0.8256408828846264, "learning_rate": 1.1532548236152574e-06, "loss": 0.0376, "step": 3945 }, { "epoch": 0.8978384527872583, "grad_norm": 2.409677737424434, "learning_rate": 1.1532070654778838e-06, "loss": 0.1075, "step": 3946 }, { "epoch": 0.89806598407281, "grad_norm": 1.6618831194214136, "learning_rate": 1.1531592965449249e-06, "loss": 0.0724, "step": 3947 }, { "epoch": 0.8982935153583618, "grad_norm": 1.1965129320838301, "learning_rate": 1.1531115168173574e-06, "loss": 0.0563, "step": 3948 }, { "epoch": 0.8985210466439135, "grad_norm": 1.6580223852928602, "learning_rate": 1.1530637262961574e-06, "loss": 0.0901, "step": 3949 }, { "epoch": 0.8987485779294653, "grad_norm": 2.0705952551342124, "learning_rate": 1.1530159249823022e-06, "loss": 0.1004, "step": 3950 }, { "epoch": 0.898976109215017, "grad_norm": 1.0142357819912748, "learning_rate": 1.1529681128767686e-06, "loss": 0.0412, "step": 3951 }, { "epoch": 0.8992036405005688, "grad_norm": 1.1350339928761526, "learning_rate": 1.1529202899805336e-06, "loss": 0.0538, "step": 3952 }, { "epoch": 0.8994311717861206, "grad_norm": 1.5799378264286232, "learning_rate": 1.1528724562945748e-06, "loss": 0.0833, "step": 3953 }, { "epoch": 0.8996587030716724, "grad_norm": 1.1208895769557774, "learning_rate": 1.1528246118198697e-06, "loss": 0.039, "step": 3954 }, { "epoch": 0.8998862343572241, "grad_norm": 1.160212672909766, "learning_rate": 1.1527767565573961e-06, "loss": 0.0271, "step": 3955 }, { "epoch": 0.9001137656427759, "grad_norm": 1.689340759598001, "learning_rate": 1.152728890508132e-06, "loss": 0.085, "step": 3956 }, { "epoch": 0.9003412969283277, "grad_norm": 6.049824984821008, "learning_rate": 1.1526810136730562e-06, "loss": 0.0767, "step": 3957 }, { "epoch": 0.9005688282138794, "grad_norm": 1.5293674377800182, "learning_rate": 1.1526331260531467e-06, "loss": 0.0652, "step": 3958 }, { "epoch": 0.9007963594994312, "grad_norm": 0.9268964770876272, "learning_rate": 1.1525852276493825e-06, "loss": 0.0373, "step": 3959 }, { "epoch": 0.9010238907849829, "grad_norm": 1.8729631573608785, "learning_rate": 1.1525373184627426e-06, "loss": 0.0883, "step": 3960 }, { "epoch": 0.9012514220705347, "grad_norm": 1.3972976693669097, "learning_rate": 1.1524893984942059e-06, "loss": 0.056, "step": 3961 }, { "epoch": 0.9014789533560864, "grad_norm": 1.0645332178479103, "learning_rate": 1.152441467744752e-06, "loss": 0.0482, "step": 3962 }, { "epoch": 0.9017064846416383, "grad_norm": 0.8491387015963779, "learning_rate": 1.1523935262153604e-06, "loss": 0.037, "step": 3963 }, { "epoch": 0.90193401592719, "grad_norm": 1.9173276752149597, "learning_rate": 1.152345573907011e-06, "loss": 0.0756, "step": 3964 }, { "epoch": 0.9021615472127418, "grad_norm": 1.0482343731205466, "learning_rate": 1.1522976108206838e-06, "loss": 0.0426, "step": 3965 }, { "epoch": 0.9023890784982935, "grad_norm": 1.5912063871488282, "learning_rate": 1.1522496369573592e-06, "loss": 0.0635, "step": 3966 }, { "epoch": 0.9026166097838453, "grad_norm": 1.0885136963654516, "learning_rate": 1.1522016523180177e-06, "loss": 0.0467, "step": 3967 }, { "epoch": 0.902844141069397, "grad_norm": 1.0253021604975354, "learning_rate": 1.15215365690364e-06, "loss": 0.0402, "step": 3968 }, { "epoch": 0.9030716723549488, "grad_norm": 1.6954250726492197, "learning_rate": 1.1521056507152068e-06, "loss": 0.0775, "step": 3969 }, { "epoch": 0.9032992036405005, "grad_norm": 2.193731037581251, "learning_rate": 1.1520576337536995e-06, "loss": 0.1161, "step": 3970 }, { "epoch": 0.9035267349260523, "grad_norm": 1.7077975253874977, "learning_rate": 1.1520096060200995e-06, "loss": 0.0947, "step": 3971 }, { "epoch": 0.903754266211604, "grad_norm": 1.5399563597984878, "learning_rate": 1.1519615675153884e-06, "loss": 0.1044, "step": 3972 }, { "epoch": 0.9039817974971559, "grad_norm": 1.7617983578353158, "learning_rate": 1.1519135182405477e-06, "loss": 0.0684, "step": 3973 }, { "epoch": 0.9042093287827077, "grad_norm": 1.3968116012863456, "learning_rate": 1.1518654581965597e-06, "loss": 0.0688, "step": 3974 }, { "epoch": 0.9044368600682594, "grad_norm": 1.1385754513385686, "learning_rate": 1.1518173873844068e-06, "loss": 0.0349, "step": 3975 }, { "epoch": 0.9046643913538112, "grad_norm": 0.8893866615862518, "learning_rate": 1.1517693058050714e-06, "loss": 0.0309, "step": 3976 }, { "epoch": 0.9048919226393629, "grad_norm": 1.539477830489965, "learning_rate": 1.151721213459536e-06, "loss": 0.0689, "step": 3977 }, { "epoch": 0.9051194539249147, "grad_norm": 1.1195140916495427, "learning_rate": 1.1516731103487836e-06, "loss": 0.051, "step": 3978 }, { "epoch": 0.9053469852104664, "grad_norm": 0.7220926512528832, "learning_rate": 1.1516249964737974e-06, "loss": 0.019, "step": 3979 }, { "epoch": 0.9055745164960182, "grad_norm": 1.0047441878180272, "learning_rate": 1.1515768718355607e-06, "loss": 0.0536, "step": 3980 }, { "epoch": 0.9058020477815699, "grad_norm": 1.1154458357876431, "learning_rate": 1.1515287364350573e-06, "loss": 0.0436, "step": 3981 }, { "epoch": 0.9060295790671218, "grad_norm": 1.083449314088406, "learning_rate": 1.1514805902732706e-06, "loss": 0.0403, "step": 3982 }, { "epoch": 0.9062571103526735, "grad_norm": 1.0839313718633037, "learning_rate": 1.151432433351185e-06, "loss": 0.0494, "step": 3983 }, { "epoch": 0.9064846416382253, "grad_norm": 1.6843996604661877, "learning_rate": 1.1513842656697844e-06, "loss": 0.0731, "step": 3984 }, { "epoch": 0.906712172923777, "grad_norm": 0.8893497549287934, "learning_rate": 1.1513360872300535e-06, "loss": 0.024, "step": 3985 }, { "epoch": 0.9069397042093288, "grad_norm": 1.6031300617012416, "learning_rate": 1.1512878980329771e-06, "loss": 0.0705, "step": 3986 }, { "epoch": 0.9071672354948805, "grad_norm": 1.129106422024724, "learning_rate": 1.1512396980795399e-06, "loss": 0.0568, "step": 3987 }, { "epoch": 0.9073947667804323, "grad_norm": 1.228932149349224, "learning_rate": 1.1511914873707269e-06, "loss": 0.0462, "step": 3988 }, { "epoch": 0.907622298065984, "grad_norm": 1.636371433136822, "learning_rate": 1.1511432659075234e-06, "loss": 0.0733, "step": 3989 }, { "epoch": 0.9078498293515358, "grad_norm": 1.1425256617133182, "learning_rate": 1.1510950336909154e-06, "loss": 0.0628, "step": 3990 }, { "epoch": 0.9080773606370875, "grad_norm": 1.01819882213196, "learning_rate": 1.1510467907218883e-06, "loss": 0.0442, "step": 3991 }, { "epoch": 0.9083048919226394, "grad_norm": 1.245888941943274, "learning_rate": 1.1509985370014283e-06, "loss": 0.0551, "step": 3992 }, { "epoch": 0.9085324232081912, "grad_norm": 1.3036968814546137, "learning_rate": 1.1509502725305214e-06, "loss": 0.0531, "step": 3993 }, { "epoch": 0.9087599544937429, "grad_norm": 0.9765098064954344, "learning_rate": 1.1509019973101542e-06, "loss": 0.0415, "step": 3994 }, { "epoch": 0.9089874857792947, "grad_norm": 0.8156281011799231, "learning_rate": 1.1508537113413134e-06, "loss": 0.041, "step": 3995 }, { "epoch": 0.9092150170648464, "grad_norm": 1.3936521974976825, "learning_rate": 1.1508054146249858e-06, "loss": 0.0397, "step": 3996 }, { "epoch": 0.9094425483503982, "grad_norm": 1.0646687699415507, "learning_rate": 1.1507571071621585e-06, "loss": 0.0357, "step": 3997 }, { "epoch": 0.9096700796359499, "grad_norm": 1.1522419098336925, "learning_rate": 1.1507087889538186e-06, "loss": 0.048, "step": 3998 }, { "epoch": 0.9098976109215017, "grad_norm": 1.2076402547928258, "learning_rate": 1.1506604600009542e-06, "loss": 0.0586, "step": 3999 }, { "epoch": 0.9101251422070534, "grad_norm": 1.0133219849273498, "learning_rate": 1.1506121203045524e-06, "loss": 0.0619, "step": 4000 }, { "epoch": 0.9103526734926052, "grad_norm": 0.8291805370083495, "learning_rate": 1.1505637698656014e-06, "loss": 0.0474, "step": 4001 }, { "epoch": 0.910580204778157, "grad_norm": 1.5524783948323588, "learning_rate": 1.1505154086850898e-06, "loss": 0.0696, "step": 4002 }, { "epoch": 0.9108077360637088, "grad_norm": 1.0896886188901207, "learning_rate": 1.1504670367640054e-06, "loss": 0.0349, "step": 4003 }, { "epoch": 0.9110352673492605, "grad_norm": 1.4368712196848394, "learning_rate": 1.1504186541033371e-06, "loss": 0.0972, "step": 4004 }, { "epoch": 0.9112627986348123, "grad_norm": 1.1609052845323367, "learning_rate": 1.150370260704074e-06, "loss": 0.0344, "step": 4005 }, { "epoch": 0.911490329920364, "grad_norm": 0.7261107332342037, "learning_rate": 1.1503218565672047e-06, "loss": 0.0378, "step": 4006 }, { "epoch": 0.9117178612059158, "grad_norm": 1.3224085375415395, "learning_rate": 1.1502734416937188e-06, "loss": 0.0806, "step": 4007 }, { "epoch": 0.9119453924914676, "grad_norm": 1.3655919422766778, "learning_rate": 1.150225016084606e-06, "loss": 0.0822, "step": 4008 }, { "epoch": 0.9121729237770193, "grad_norm": 0.8123755839338497, "learning_rate": 1.1501765797408558e-06, "loss": 0.0337, "step": 4009 }, { "epoch": 0.9124004550625711, "grad_norm": 1.8652137193155605, "learning_rate": 1.1501281326634578e-06, "loss": 0.0612, "step": 4010 }, { "epoch": 0.9126279863481229, "grad_norm": 0.9706539819037835, "learning_rate": 1.1500796748534026e-06, "loss": 0.0559, "step": 4011 }, { "epoch": 0.9128555176336747, "grad_norm": 0.6605470896041641, "learning_rate": 1.1500312063116803e-06, "loss": 0.0277, "step": 4012 }, { "epoch": 0.9130830489192264, "grad_norm": 2.0635463828713445, "learning_rate": 1.149982727039282e-06, "loss": 0.1237, "step": 4013 }, { "epoch": 0.9133105802047782, "grad_norm": 1.2700127801482208, "learning_rate": 1.149934237037198e-06, "loss": 0.0519, "step": 4014 }, { "epoch": 0.9135381114903299, "grad_norm": 1.0543579404973509, "learning_rate": 1.1498857363064198e-06, "loss": 0.0318, "step": 4015 }, { "epoch": 0.9137656427758817, "grad_norm": 1.3887452700800633, "learning_rate": 1.1498372248479383e-06, "loss": 0.0776, "step": 4016 }, { "epoch": 0.9139931740614334, "grad_norm": 1.1548765049955128, "learning_rate": 1.1497887026627451e-06, "loss": 0.0631, "step": 4017 }, { "epoch": 0.9142207053469852, "grad_norm": 1.4155291898606746, "learning_rate": 1.1497401697518318e-06, "loss": 0.0757, "step": 4018 }, { "epoch": 0.9144482366325369, "grad_norm": 1.5411778856885947, "learning_rate": 1.1496916261161908e-06, "loss": 0.0845, "step": 4019 }, { "epoch": 0.9146757679180887, "grad_norm": 0.8243491395066448, "learning_rate": 1.1496430717568136e-06, "loss": 0.0304, "step": 4020 }, { "epoch": 0.9149032992036406, "grad_norm": 1.0014800679630642, "learning_rate": 1.149594506674693e-06, "loss": 0.0443, "step": 4021 }, { "epoch": 0.9151308304891923, "grad_norm": 0.9825886288199939, "learning_rate": 1.1495459308708212e-06, "loss": 0.0285, "step": 4022 }, { "epoch": 0.9153583617747441, "grad_norm": 0.6708245548677098, "learning_rate": 1.1494973443461915e-06, "loss": 0.0258, "step": 4023 }, { "epoch": 0.9155858930602958, "grad_norm": 1.0354775698510426, "learning_rate": 1.1494487471017965e-06, "loss": 0.0516, "step": 4024 }, { "epoch": 0.9158134243458476, "grad_norm": 1.4705781324392022, "learning_rate": 1.1494001391386298e-06, "loss": 0.0588, "step": 4025 }, { "epoch": 0.9160409556313993, "grad_norm": 1.4496056117656928, "learning_rate": 1.1493515204576844e-06, "loss": 0.0453, "step": 4026 }, { "epoch": 0.9162684869169511, "grad_norm": 1.2256180174274225, "learning_rate": 1.1493028910599544e-06, "loss": 0.0474, "step": 4027 }, { "epoch": 0.9164960182025028, "grad_norm": 1.241479459058262, "learning_rate": 1.1492542509464333e-06, "loss": 0.0499, "step": 4028 }, { "epoch": 0.9167235494880546, "grad_norm": 1.0655237448875803, "learning_rate": 1.1492056001181157e-06, "loss": 0.04, "step": 4029 }, { "epoch": 0.9169510807736063, "grad_norm": 0.9311299042715484, "learning_rate": 1.1491569385759953e-06, "loss": 0.0353, "step": 4030 }, { "epoch": 0.9171786120591582, "grad_norm": 1.9476966702422809, "learning_rate": 1.1491082663210675e-06, "loss": 0.0822, "step": 4031 }, { "epoch": 0.9174061433447099, "grad_norm": 1.7080369705282124, "learning_rate": 1.1490595833543263e-06, "loss": 0.0921, "step": 4032 }, { "epoch": 0.9176336746302617, "grad_norm": 0.9700416110415793, "learning_rate": 1.1490108896767672e-06, "loss": 0.0376, "step": 4033 }, { "epoch": 0.9178612059158134, "grad_norm": 1.79218464701242, "learning_rate": 1.1489621852893849e-06, "loss": 0.0793, "step": 4034 }, { "epoch": 0.9180887372013652, "grad_norm": 1.5337857838361597, "learning_rate": 1.1489134701931753e-06, "loss": 0.0734, "step": 4035 }, { "epoch": 0.9183162684869169, "grad_norm": 2.2908857950335393, "learning_rate": 1.1488647443891339e-06, "loss": 0.1363, "step": 4036 }, { "epoch": 0.9185437997724687, "grad_norm": 0.9807386709388579, "learning_rate": 1.1488160078782565e-06, "loss": 0.0553, "step": 4037 }, { "epoch": 0.9187713310580204, "grad_norm": 1.610140836174014, "learning_rate": 1.148767260661539e-06, "loss": 0.0535, "step": 4038 }, { "epoch": 0.9189988623435722, "grad_norm": 1.8010870128097782, "learning_rate": 1.1487185027399783e-06, "loss": 0.0809, "step": 4039 }, { "epoch": 0.919226393629124, "grad_norm": 0.9946177354686938, "learning_rate": 1.1486697341145703e-06, "loss": 0.0308, "step": 4040 }, { "epoch": 0.9194539249146758, "grad_norm": 1.1243952505434611, "learning_rate": 1.148620954786312e-06, "loss": 0.0403, "step": 4041 }, { "epoch": 0.9196814562002276, "grad_norm": 1.1676663366717939, "learning_rate": 1.1485721647562005e-06, "loss": 0.0642, "step": 4042 }, { "epoch": 0.9199089874857793, "grad_norm": 1.2805472497836035, "learning_rate": 1.1485233640252328e-06, "loss": 0.0495, "step": 4043 }, { "epoch": 0.9201365187713311, "grad_norm": 1.176529756201314, "learning_rate": 1.1484745525944063e-06, "loss": 0.0425, "step": 4044 }, { "epoch": 0.9203640500568828, "grad_norm": 1.4292232195143082, "learning_rate": 1.1484257304647187e-06, "loss": 0.0792, "step": 4045 }, { "epoch": 0.9205915813424346, "grad_norm": 1.7932020137107854, "learning_rate": 1.1483768976371677e-06, "loss": 0.0727, "step": 4046 }, { "epoch": 0.9208191126279863, "grad_norm": 1.6644682540806308, "learning_rate": 1.1483280541127513e-06, "loss": 0.062, "step": 4047 }, { "epoch": 0.9210466439135381, "grad_norm": 1.1510951655579869, "learning_rate": 1.1482791998924681e-06, "loss": 0.0434, "step": 4048 }, { "epoch": 0.9212741751990898, "grad_norm": 1.19496202376025, "learning_rate": 1.1482303349773164e-06, "loss": 0.0679, "step": 4049 }, { "epoch": 0.9215017064846417, "grad_norm": 1.154368424237523, "learning_rate": 1.1481814593682946e-06, "loss": 0.0396, "step": 4050 }, { "epoch": 0.9217292377701934, "grad_norm": 1.3809033535643431, "learning_rate": 1.1481325730664023e-06, "loss": 0.0772, "step": 4051 }, { "epoch": 0.9219567690557452, "grad_norm": 0.9785189844102623, "learning_rate": 1.148083676072638e-06, "loss": 0.039, "step": 4052 }, { "epoch": 0.922184300341297, "grad_norm": 1.4126999693949231, "learning_rate": 1.1480347683880016e-06, "loss": 0.0704, "step": 4053 }, { "epoch": 0.9224118316268487, "grad_norm": 0.8860494719620977, "learning_rate": 1.1479858500134924e-06, "loss": 0.0269, "step": 4054 }, { "epoch": 0.9226393629124005, "grad_norm": 1.4070564902702778, "learning_rate": 1.14793692095011e-06, "loss": 0.0775, "step": 4055 }, { "epoch": 0.9228668941979522, "grad_norm": 1.1117624571631268, "learning_rate": 1.147887981198855e-06, "loss": 0.0474, "step": 4056 }, { "epoch": 0.923094425483504, "grad_norm": 1.4621020662224937, "learning_rate": 1.147839030760727e-06, "loss": 0.069, "step": 4057 }, { "epoch": 0.9233219567690557, "grad_norm": 1.4729100301695202, "learning_rate": 1.1477900696367269e-06, "loss": 0.0696, "step": 4058 }, { "epoch": 0.9235494880546075, "grad_norm": 0.9208872591470431, "learning_rate": 1.147741097827855e-06, "loss": 0.0365, "step": 4059 }, { "epoch": 0.9237770193401593, "grad_norm": 0.9240892354847577, "learning_rate": 1.1476921153351126e-06, "loss": 0.0405, "step": 4060 }, { "epoch": 0.9240045506257111, "grad_norm": 1.4959558432174869, "learning_rate": 1.1476431221595005e-06, "loss": 0.0463, "step": 4061 }, { "epoch": 0.9242320819112628, "grad_norm": 1.332519816167777, "learning_rate": 1.1475941183020203e-06, "loss": 0.0658, "step": 4062 }, { "epoch": 0.9244596131968146, "grad_norm": 2.187146405056203, "learning_rate": 1.1475451037636733e-06, "loss": 0.0943, "step": 4063 }, { "epoch": 0.9246871444823663, "grad_norm": 1.3435930034506374, "learning_rate": 1.1474960785454615e-06, "loss": 0.0726, "step": 4064 }, { "epoch": 0.9249146757679181, "grad_norm": 1.3012028567242737, "learning_rate": 1.1474470426483868e-06, "loss": 0.0528, "step": 4065 }, { "epoch": 0.9251422070534698, "grad_norm": 1.0819099650921233, "learning_rate": 1.1473979960734513e-06, "loss": 0.0426, "step": 4066 }, { "epoch": 0.9253697383390216, "grad_norm": 0.781744966406713, "learning_rate": 1.1473489388216574e-06, "loss": 0.0367, "step": 4067 }, { "epoch": 0.9255972696245733, "grad_norm": 1.151095346965489, "learning_rate": 1.1472998708940079e-06, "loss": 0.0619, "step": 4068 }, { "epoch": 0.9258248009101251, "grad_norm": 1.3142684365826751, "learning_rate": 1.1472507922915056e-06, "loss": 0.0446, "step": 4069 }, { "epoch": 0.926052332195677, "grad_norm": 1.3392798642883632, "learning_rate": 1.1472017030151536e-06, "loss": 0.0675, "step": 4070 }, { "epoch": 0.9262798634812287, "grad_norm": 0.7396122034506598, "learning_rate": 1.147152603065955e-06, "loss": 0.0436, "step": 4071 }, { "epoch": 0.9265073947667805, "grad_norm": 1.3292817664212946, "learning_rate": 1.1471034924449133e-06, "loss": 0.0562, "step": 4072 }, { "epoch": 0.9267349260523322, "grad_norm": 1.9912088396901564, "learning_rate": 1.1470543711530328e-06, "loss": 0.0629, "step": 4073 }, { "epoch": 0.926962457337884, "grad_norm": 1.3570840904239112, "learning_rate": 1.147005239191317e-06, "loss": 0.0667, "step": 4074 }, { "epoch": 0.9271899886234357, "grad_norm": 1.2559298173099012, "learning_rate": 1.1469560965607699e-06, "loss": 0.0518, "step": 4075 }, { "epoch": 0.9274175199089875, "grad_norm": 1.0348081180884323, "learning_rate": 1.1469069432623965e-06, "loss": 0.0517, "step": 4076 }, { "epoch": 0.9276450511945392, "grad_norm": 1.0484875570709722, "learning_rate": 1.1468577792972004e-06, "loss": 0.0465, "step": 4077 }, { "epoch": 0.927872582480091, "grad_norm": 1.09203642071682, "learning_rate": 1.1468086046661874e-06, "loss": 0.0548, "step": 4078 }, { "epoch": 0.9281001137656428, "grad_norm": 1.7712907658090584, "learning_rate": 1.146759419370362e-06, "loss": 0.1043, "step": 4079 }, { "epoch": 0.9283276450511946, "grad_norm": 0.8782441484375744, "learning_rate": 1.14671022341073e-06, "loss": 0.0241, "step": 4080 }, { "epoch": 0.9285551763367463, "grad_norm": 1.3964199439198308, "learning_rate": 1.1466610167882963e-06, "loss": 0.0789, "step": 4081 }, { "epoch": 0.9287827076222981, "grad_norm": 1.3164169043622294, "learning_rate": 1.1466117995040666e-06, "loss": 0.0635, "step": 4082 }, { "epoch": 0.9290102389078498, "grad_norm": 1.2381160175958454, "learning_rate": 1.1465625715590473e-06, "loss": 0.0427, "step": 4083 }, { "epoch": 0.9292377701934016, "grad_norm": 1.0698402996548797, "learning_rate": 1.146513332954244e-06, "loss": 0.0626, "step": 4084 }, { "epoch": 0.9294653014789533, "grad_norm": 1.534739951538693, "learning_rate": 1.1464640836906635e-06, "loss": 0.0587, "step": 4085 }, { "epoch": 0.9296928327645051, "grad_norm": 1.0552598639917363, "learning_rate": 1.146414823769312e-06, "loss": 0.046, "step": 4086 }, { "epoch": 0.9299203640500568, "grad_norm": 0.7826783824026263, "learning_rate": 1.1463655531911963e-06, "loss": 0.0308, "step": 4087 }, { "epoch": 0.9301478953356086, "grad_norm": 0.8273022304812389, "learning_rate": 1.1463162719573236e-06, "loss": 0.031, "step": 4088 }, { "epoch": 0.9303754266211605, "grad_norm": 1.3288942117067748, "learning_rate": 1.1462669800687012e-06, "loss": 0.0564, "step": 4089 }, { "epoch": 0.9306029579067122, "grad_norm": 0.8195424504406558, "learning_rate": 1.1462176775263365e-06, "loss": 0.0349, "step": 4090 }, { "epoch": 0.930830489192264, "grad_norm": 1.69823108141483, "learning_rate": 1.1461683643312366e-06, "loss": 0.0698, "step": 4091 }, { "epoch": 0.9310580204778157, "grad_norm": 1.5933893580265006, "learning_rate": 1.1461190404844103e-06, "loss": 0.0748, "step": 4092 }, { "epoch": 0.9312855517633675, "grad_norm": 1.4412482142890468, "learning_rate": 1.1460697059868648e-06, "loss": 0.086, "step": 4093 }, { "epoch": 0.9315130830489192, "grad_norm": 1.1113383236310055, "learning_rate": 1.146020360839609e-06, "loss": 0.0379, "step": 4094 }, { "epoch": 0.931740614334471, "grad_norm": 0.7200940731856813, "learning_rate": 1.1459710050436513e-06, "loss": 0.0356, "step": 4095 }, { "epoch": 0.9319681456200227, "grad_norm": 1.4853749484298682, "learning_rate": 1.1459216385999999e-06, "loss": 0.0559, "step": 4096 }, { "epoch": 0.9321956769055745, "grad_norm": 0.9574272073778142, "learning_rate": 1.1458722615096648e-06, "loss": 0.0276, "step": 4097 }, { "epoch": 0.9324232081911262, "grad_norm": 0.8174617084003479, "learning_rate": 1.1458228737736542e-06, "loss": 0.0327, "step": 4098 }, { "epoch": 0.9326507394766781, "grad_norm": 1.3250207983841498, "learning_rate": 1.145773475392978e-06, "loss": 0.0516, "step": 4099 }, { "epoch": 0.9328782707622298, "grad_norm": 1.22249857755224, "learning_rate": 1.145724066368646e-06, "loss": 0.0637, "step": 4100 }, { "epoch": 0.9331058020477816, "grad_norm": 1.4663762999847116, "learning_rate": 1.1456746467016675e-06, "loss": 0.0567, "step": 4101 }, { "epoch": 0.9333333333333333, "grad_norm": 1.4030554644294082, "learning_rate": 1.1456252163930528e-06, "loss": 0.0584, "step": 4102 }, { "epoch": 0.9335608646188851, "grad_norm": 1.5109418446571456, "learning_rate": 1.1455757754438122e-06, "loss": 0.0497, "step": 4103 }, { "epoch": 0.9337883959044369, "grad_norm": 1.0191531582003377, "learning_rate": 1.1455263238549563e-06, "loss": 0.0562, "step": 4104 }, { "epoch": 0.9340159271899886, "grad_norm": 0.9931115670125917, "learning_rate": 1.1454768616274955e-06, "loss": 0.0494, "step": 4105 }, { "epoch": 0.9342434584755404, "grad_norm": 1.0861122916963688, "learning_rate": 1.1454273887624407e-06, "loss": 0.0548, "step": 4106 }, { "epoch": 0.9344709897610921, "grad_norm": 0.9828952905986, "learning_rate": 1.1453779052608032e-06, "loss": 0.0374, "step": 4107 }, { "epoch": 0.9346985210466439, "grad_norm": 1.5222459429366253, "learning_rate": 1.1453284111235947e-06, "loss": 0.0818, "step": 4108 }, { "epoch": 0.9349260523321957, "grad_norm": 1.1153500509733376, "learning_rate": 1.145278906351826e-06, "loss": 0.0409, "step": 4109 }, { "epoch": 0.9351535836177475, "grad_norm": 1.0587413994265493, "learning_rate": 1.1452293909465095e-06, "loss": 0.0669, "step": 4110 }, { "epoch": 0.9353811149032992, "grad_norm": 0.809823129379949, "learning_rate": 1.145179864908657e-06, "loss": 0.0304, "step": 4111 }, { "epoch": 0.935608646188851, "grad_norm": 1.173627001973459, "learning_rate": 1.1451303282392808e-06, "loss": 0.046, "step": 4112 }, { "epoch": 0.9358361774744027, "grad_norm": 1.301208608992828, "learning_rate": 1.145080780939393e-06, "loss": 0.0457, "step": 4113 }, { "epoch": 0.9360637087599545, "grad_norm": 1.1311235005804838, "learning_rate": 1.145031223010007e-06, "loss": 0.0293, "step": 4114 }, { "epoch": 0.9362912400455062, "grad_norm": 1.754429989040699, "learning_rate": 1.1449816544521347e-06, "loss": 0.0962, "step": 4115 }, { "epoch": 0.936518771331058, "grad_norm": 1.1523409260488855, "learning_rate": 1.1449320752667898e-06, "loss": 0.0622, "step": 4116 }, { "epoch": 0.9367463026166097, "grad_norm": 0.9453806040405214, "learning_rate": 1.1448824854549856e-06, "loss": 0.0303, "step": 4117 }, { "epoch": 0.9369738339021616, "grad_norm": 1.270090887527193, "learning_rate": 1.1448328850177356e-06, "loss": 0.057, "step": 4118 }, { "epoch": 0.9372013651877134, "grad_norm": 1.1742323486339012, "learning_rate": 1.1447832739560533e-06, "loss": 0.0689, "step": 4119 }, { "epoch": 0.9374288964732651, "grad_norm": 1.0965078798779573, "learning_rate": 1.1447336522709528e-06, "loss": 0.0616, "step": 4120 }, { "epoch": 0.9376564277588169, "grad_norm": 1.1403738547142621, "learning_rate": 1.1446840199634483e-06, "loss": 0.0681, "step": 4121 }, { "epoch": 0.9378839590443686, "grad_norm": 1.176539532491646, "learning_rate": 1.1446343770345544e-06, "loss": 0.0369, "step": 4122 }, { "epoch": 0.9381114903299204, "grad_norm": 0.8950308841444714, "learning_rate": 1.1445847234852853e-06, "loss": 0.0508, "step": 4123 }, { "epoch": 0.9383390216154721, "grad_norm": 2.6033063822464553, "learning_rate": 1.1445350593166559e-06, "loss": 0.0865, "step": 4124 }, { "epoch": 0.9385665529010239, "grad_norm": 1.4419905128708395, "learning_rate": 1.1444853845296816e-06, "loss": 0.0922, "step": 4125 }, { "epoch": 0.9387940841865756, "grad_norm": 0.810209427625322, "learning_rate": 1.1444356991253774e-06, "loss": 0.0302, "step": 4126 }, { "epoch": 0.9390216154721274, "grad_norm": 1.4059215847764233, "learning_rate": 1.1443860031047589e-06, "loss": 0.0584, "step": 4127 }, { "epoch": 0.9392491467576792, "grad_norm": 1.3946503335439686, "learning_rate": 1.1443362964688416e-06, "loss": 0.0665, "step": 4128 }, { "epoch": 0.939476678043231, "grad_norm": 2.123996937731276, "learning_rate": 1.1442865792186413e-06, "loss": 0.0869, "step": 4129 }, { "epoch": 0.9397042093287827, "grad_norm": 1.0634024569590812, "learning_rate": 1.1442368513551746e-06, "loss": 0.0418, "step": 4130 }, { "epoch": 0.9399317406143345, "grad_norm": 1.063642439356403, "learning_rate": 1.1441871128794576e-06, "loss": 0.0341, "step": 4131 }, { "epoch": 0.9401592718998862, "grad_norm": 0.9460769468351368, "learning_rate": 1.1441373637925068e-06, "loss": 0.0533, "step": 4132 }, { "epoch": 0.940386803185438, "grad_norm": 1.764351905359598, "learning_rate": 1.1440876040953392e-06, "loss": 0.1242, "step": 4133 }, { "epoch": 0.9406143344709897, "grad_norm": 1.4558757328006684, "learning_rate": 1.1440378337889713e-06, "loss": 0.0763, "step": 4134 }, { "epoch": 0.9408418657565415, "grad_norm": 0.9021601245337508, "learning_rate": 1.143988052874421e-06, "loss": 0.0409, "step": 4135 }, { "epoch": 0.9410693970420932, "grad_norm": 1.7188429838557109, "learning_rate": 1.143938261352705e-06, "loss": 0.069, "step": 4136 }, { "epoch": 0.941296928327645, "grad_norm": 1.2572237568519273, "learning_rate": 1.1438884592248416e-06, "loss": 0.0368, "step": 4137 }, { "epoch": 0.9415244596131969, "grad_norm": 1.2286297911170685, "learning_rate": 1.1438386464918483e-06, "loss": 0.0371, "step": 4138 }, { "epoch": 0.9417519908987486, "grad_norm": 1.664747740752945, "learning_rate": 1.1437888231547434e-06, "loss": 0.0793, "step": 4139 }, { "epoch": 0.9419795221843004, "grad_norm": 0.9027458023794878, "learning_rate": 1.143738989214545e-06, "loss": 0.0359, "step": 4140 }, { "epoch": 0.9422070534698521, "grad_norm": 0.9844227932686808, "learning_rate": 1.1436891446722718e-06, "loss": 0.0398, "step": 4141 }, { "epoch": 0.9424345847554039, "grad_norm": 2.473079458381896, "learning_rate": 1.1436392895289423e-06, "loss": 0.1426, "step": 4142 }, { "epoch": 0.9426621160409556, "grad_norm": 1.6695549407826555, "learning_rate": 1.1435894237855754e-06, "loss": 0.0569, "step": 4143 }, { "epoch": 0.9428896473265074, "grad_norm": 0.9951741584954418, "learning_rate": 1.1435395474431906e-06, "loss": 0.0354, "step": 4144 }, { "epoch": 0.9431171786120591, "grad_norm": 1.3165779040583256, "learning_rate": 1.143489660502807e-06, "loss": 0.062, "step": 4145 }, { "epoch": 0.9433447098976109, "grad_norm": 1.1419121851198617, "learning_rate": 1.1434397629654445e-06, "loss": 0.0628, "step": 4146 }, { "epoch": 0.9435722411831626, "grad_norm": 1.1143992408718018, "learning_rate": 1.1433898548321226e-06, "loss": 0.0461, "step": 4147 }, { "epoch": 0.9437997724687145, "grad_norm": 1.4419058001216918, "learning_rate": 1.1433399361038614e-06, "loss": 0.0622, "step": 4148 }, { "epoch": 0.9440273037542662, "grad_norm": 1.0878632797009244, "learning_rate": 1.1432900067816813e-06, "loss": 0.051, "step": 4149 }, { "epoch": 0.944254835039818, "grad_norm": 0.8678017410178684, "learning_rate": 1.1432400668666028e-06, "loss": 0.036, "step": 4150 }, { "epoch": 0.9444823663253697, "grad_norm": 2.5565159682952325, "learning_rate": 1.1431901163596462e-06, "loss": 0.1481, "step": 4151 }, { "epoch": 0.9447098976109215, "grad_norm": 1.1719746596557403, "learning_rate": 1.1431401552618327e-06, "loss": 0.0654, "step": 4152 }, { "epoch": 0.9449374288964733, "grad_norm": 0.7399300932811994, "learning_rate": 1.1430901835741833e-06, "loss": 0.0426, "step": 4153 }, { "epoch": 0.945164960182025, "grad_norm": 1.5392915168318702, "learning_rate": 1.1430402012977195e-06, "loss": 0.064, "step": 4154 }, { "epoch": 0.9453924914675768, "grad_norm": 0.8642603027613659, "learning_rate": 1.1429902084334627e-06, "loss": 0.0391, "step": 4155 }, { "epoch": 0.9456200227531285, "grad_norm": 1.5792578097649426, "learning_rate": 1.1429402049824348e-06, "loss": 0.0864, "step": 4156 }, { "epoch": 0.9458475540386804, "grad_norm": 1.0656243196471682, "learning_rate": 1.1428901909456575e-06, "loss": 0.0643, "step": 4157 }, { "epoch": 0.9460750853242321, "grad_norm": 0.7834798421355951, "learning_rate": 1.1428401663241533e-06, "loss": 0.0255, "step": 4158 }, { "epoch": 0.9463026166097839, "grad_norm": 0.7290481198035724, "learning_rate": 1.1427901311189444e-06, "loss": 0.0347, "step": 4159 }, { "epoch": 0.9465301478953356, "grad_norm": 1.0906839196972904, "learning_rate": 1.1427400853310536e-06, "loss": 0.0416, "step": 4160 }, { "epoch": 0.9467576791808874, "grad_norm": 0.7797257513708735, "learning_rate": 1.1426900289615034e-06, "loss": 0.0332, "step": 4161 }, { "epoch": 0.9469852104664391, "grad_norm": 0.9828933034646482, "learning_rate": 1.1426399620113174e-06, "loss": 0.0426, "step": 4162 }, { "epoch": 0.9472127417519909, "grad_norm": 1.0815564353692642, "learning_rate": 1.1425898844815183e-06, "loss": 0.03, "step": 4163 }, { "epoch": 0.9474402730375426, "grad_norm": 1.3070693062923497, "learning_rate": 1.1425397963731303e-06, "loss": 0.0761, "step": 4164 }, { "epoch": 0.9476678043230944, "grad_norm": 1.9681632277415542, "learning_rate": 1.1424896976871763e-06, "loss": 0.0751, "step": 4165 }, { "epoch": 0.9478953356086461, "grad_norm": 1.3403750162598331, "learning_rate": 1.1424395884246808e-06, "loss": 0.0645, "step": 4166 }, { "epoch": 0.948122866894198, "grad_norm": 1.4856178047017083, "learning_rate": 1.1423894685866677e-06, "loss": 0.0524, "step": 4167 }, { "epoch": 0.9483503981797498, "grad_norm": 2.390346331051792, "learning_rate": 1.1423393381741614e-06, "loss": 0.1007, "step": 4168 }, { "epoch": 0.9485779294653015, "grad_norm": 1.9335775336716192, "learning_rate": 1.1422891971881867e-06, "loss": 0.0909, "step": 4169 }, { "epoch": 0.9488054607508533, "grad_norm": 1.4846712557080712, "learning_rate": 1.142239045629768e-06, "loss": 0.0706, "step": 4170 }, { "epoch": 0.949032992036405, "grad_norm": 1.2534969921049077, "learning_rate": 1.1421888834999306e-06, "loss": 0.0637, "step": 4171 }, { "epoch": 0.9492605233219568, "grad_norm": 1.5488362147727934, "learning_rate": 1.1421387107996993e-06, "loss": 0.065, "step": 4172 }, { "epoch": 0.9494880546075085, "grad_norm": 1.8544510157017915, "learning_rate": 1.1420885275301001e-06, "loss": 0.0783, "step": 4173 }, { "epoch": 0.9497155858930603, "grad_norm": 1.2775832861301255, "learning_rate": 1.1420383336921583e-06, "loss": 0.0628, "step": 4174 }, { "epoch": 0.949943117178612, "grad_norm": 1.1882749927515044, "learning_rate": 1.1419881292869e-06, "loss": 0.053, "step": 4175 }, { "epoch": 0.9501706484641638, "grad_norm": 1.1112594425917803, "learning_rate": 1.1419379143153511e-06, "loss": 0.0391, "step": 4176 }, { "epoch": 0.9503981797497156, "grad_norm": 1.4092744149498044, "learning_rate": 1.1418876887785379e-06, "loss": 0.0588, "step": 4177 }, { "epoch": 0.9506257110352674, "grad_norm": 1.2859140834039406, "learning_rate": 1.1418374526774872e-06, "loss": 0.0449, "step": 4178 }, { "epoch": 0.9508532423208191, "grad_norm": 1.7602664958939818, "learning_rate": 1.1417872060132251e-06, "loss": 0.066, "step": 4179 }, { "epoch": 0.9510807736063709, "grad_norm": 1.7687663881511184, "learning_rate": 1.1417369487867793e-06, "loss": 0.0737, "step": 4180 }, { "epoch": 0.9513083048919226, "grad_norm": 1.0504917072097133, "learning_rate": 1.1416866809991763e-06, "loss": 0.0571, "step": 4181 }, { "epoch": 0.9515358361774744, "grad_norm": 0.9175209845403075, "learning_rate": 1.1416364026514443e-06, "loss": 0.0473, "step": 4182 }, { "epoch": 0.9517633674630261, "grad_norm": 1.1584022734622255, "learning_rate": 1.1415861137446099e-06, "loss": 0.0615, "step": 4183 }, { "epoch": 0.9519908987485779, "grad_norm": 0.9605566486118688, "learning_rate": 1.1415358142797018e-06, "loss": 0.0468, "step": 4184 }, { "epoch": 0.9522184300341296, "grad_norm": 1.0547428180217657, "learning_rate": 1.1414855042577474e-06, "loss": 0.0578, "step": 4185 }, { "epoch": 0.9524459613196815, "grad_norm": 1.664335132859367, "learning_rate": 1.1414351836797755e-06, "loss": 0.0929, "step": 4186 }, { "epoch": 0.9526734926052333, "grad_norm": 1.326192413657784, "learning_rate": 1.1413848525468139e-06, "loss": 0.0583, "step": 4187 }, { "epoch": 0.952901023890785, "grad_norm": 1.475570664457709, "learning_rate": 1.1413345108598916e-06, "loss": 0.0626, "step": 4188 }, { "epoch": 0.9531285551763368, "grad_norm": 1.7815203562597235, "learning_rate": 1.1412841586200378e-06, "loss": 0.0717, "step": 4189 }, { "epoch": 0.9533560864618885, "grad_norm": 1.7331207453848758, "learning_rate": 1.1412337958282812e-06, "loss": 0.0621, "step": 4190 }, { "epoch": 0.9535836177474403, "grad_norm": 0.5286320379408398, "learning_rate": 1.1411834224856514e-06, "loss": 0.0279, "step": 4191 }, { "epoch": 0.953811149032992, "grad_norm": 0.8095894659074508, "learning_rate": 1.1411330385931776e-06, "loss": 0.0342, "step": 4192 }, { "epoch": 0.9540386803185438, "grad_norm": 1.1720468896203664, "learning_rate": 1.1410826441518898e-06, "loss": 0.0769, "step": 4193 }, { "epoch": 0.9542662116040955, "grad_norm": 1.6215729451396426, "learning_rate": 1.1410322391628179e-06, "loss": 0.064, "step": 4194 }, { "epoch": 0.9544937428896473, "grad_norm": 0.9975722592735233, "learning_rate": 1.140981823626992e-06, "loss": 0.033, "step": 4195 }, { "epoch": 0.9547212741751991, "grad_norm": 0.961294779276341, "learning_rate": 1.1409313975454429e-06, "loss": 0.0556, "step": 4196 }, { "epoch": 0.9549488054607509, "grad_norm": 1.195259518530016, "learning_rate": 1.1408809609192007e-06, "loss": 0.0685, "step": 4197 }, { "epoch": 0.9551763367463026, "grad_norm": 1.2021438816670518, "learning_rate": 1.1408305137492963e-06, "loss": 0.0567, "step": 4198 }, { "epoch": 0.9554038680318544, "grad_norm": 1.0136210761213766, "learning_rate": 1.1407800560367612e-06, "loss": 0.0304, "step": 4199 }, { "epoch": 0.9556313993174061, "grad_norm": 0.8478875067932324, "learning_rate": 1.140729587782626e-06, "loss": 0.0439, "step": 4200 }, { "epoch": 0.9558589306029579, "grad_norm": 1.660804148976066, "learning_rate": 1.1406791089879229e-06, "loss": 0.0731, "step": 4201 }, { "epoch": 0.9560864618885097, "grad_norm": 0.8895932107889849, "learning_rate": 1.1406286196536832e-06, "loss": 0.0363, "step": 4202 }, { "epoch": 0.9563139931740614, "grad_norm": 2.248321391825998, "learning_rate": 1.1405781197809388e-06, "loss": 0.0736, "step": 4203 }, { "epoch": 0.9565415244596132, "grad_norm": 0.5970206546121392, "learning_rate": 1.1405276093707218e-06, "loss": 0.02, "step": 4204 }, { "epoch": 0.9567690557451649, "grad_norm": 1.263024166664947, "learning_rate": 1.1404770884240645e-06, "loss": 0.0523, "step": 4205 }, { "epoch": 0.9569965870307168, "grad_norm": 1.3072039796205928, "learning_rate": 1.1404265569419998e-06, "loss": 0.0606, "step": 4206 }, { "epoch": 0.9572241183162685, "grad_norm": 1.2388621360387058, "learning_rate": 1.14037601492556e-06, "loss": 0.065, "step": 4207 }, { "epoch": 0.9574516496018203, "grad_norm": 0.6867586774624046, "learning_rate": 1.1403254623757785e-06, "loss": 0.0258, "step": 4208 }, { "epoch": 0.957679180887372, "grad_norm": 1.296038153739218, "learning_rate": 1.1402748992936881e-06, "loss": 0.0563, "step": 4209 }, { "epoch": 0.9579067121729238, "grad_norm": 1.3170165094768458, "learning_rate": 1.1402243256803228e-06, "loss": 0.0537, "step": 4210 }, { "epoch": 0.9581342434584755, "grad_norm": 2.115381990971279, "learning_rate": 1.1401737415367157e-06, "loss": 0.1093, "step": 4211 }, { "epoch": 0.9583617747440273, "grad_norm": 1.1322388794198333, "learning_rate": 1.1401231468639008e-06, "loss": 0.0477, "step": 4212 }, { "epoch": 0.958589306029579, "grad_norm": 1.7074659597722202, "learning_rate": 1.140072541662912e-06, "loss": 0.0883, "step": 4213 }, { "epoch": 0.9588168373151308, "grad_norm": 1.6179437373612184, "learning_rate": 1.1400219259347842e-06, "loss": 0.0583, "step": 4214 }, { "epoch": 0.9590443686006825, "grad_norm": 1.0176578115657482, "learning_rate": 1.139971299680551e-06, "loss": 0.0551, "step": 4215 }, { "epoch": 0.9592718998862344, "grad_norm": 1.07590619993706, "learning_rate": 1.1399206629012478e-06, "loss": 0.0367, "step": 4216 }, { "epoch": 0.9594994311717862, "grad_norm": 0.9678873702253064, "learning_rate": 1.1398700155979092e-06, "loss": 0.0394, "step": 4217 }, { "epoch": 0.9597269624573379, "grad_norm": 1.1238559057706377, "learning_rate": 1.1398193577715705e-06, "loss": 0.0623, "step": 4218 }, { "epoch": 0.9599544937428897, "grad_norm": 1.2051383992211526, "learning_rate": 1.1397686894232671e-06, "loss": 0.048, "step": 4219 }, { "epoch": 0.9601820250284414, "grad_norm": 1.5391029437375636, "learning_rate": 1.1397180105540343e-06, "loss": 0.0575, "step": 4220 }, { "epoch": 0.9604095563139932, "grad_norm": 1.620882109175618, "learning_rate": 1.1396673211649078e-06, "loss": 0.0401, "step": 4221 }, { "epoch": 0.9606370875995449, "grad_norm": 1.4900621743746256, "learning_rate": 1.139616621256924e-06, "loss": 0.0488, "step": 4222 }, { "epoch": 0.9608646188850967, "grad_norm": 1.2199319551548797, "learning_rate": 1.1395659108311192e-06, "loss": 0.0568, "step": 4223 }, { "epoch": 0.9610921501706484, "grad_norm": 0.7265505005040674, "learning_rate": 1.1395151898885293e-06, "loss": 0.03, "step": 4224 }, { "epoch": 0.9613196814562003, "grad_norm": 1.1651056301933327, "learning_rate": 1.1394644584301912e-06, "loss": 0.0496, "step": 4225 }, { "epoch": 0.961547212741752, "grad_norm": 1.3085058571423405, "learning_rate": 1.1394137164571418e-06, "loss": 0.0488, "step": 4226 }, { "epoch": 0.9617747440273038, "grad_norm": 1.3459918801148398, "learning_rate": 1.1393629639704182e-06, "loss": 0.0525, "step": 4227 }, { "epoch": 0.9620022753128555, "grad_norm": 1.8692658121925165, "learning_rate": 1.1393122009710575e-06, "loss": 0.0859, "step": 4228 }, { "epoch": 0.9622298065984073, "grad_norm": 1.1703507912941524, "learning_rate": 1.1392614274600975e-06, "loss": 0.048, "step": 4229 }, { "epoch": 0.962457337883959, "grad_norm": 1.0308219099107716, "learning_rate": 1.1392106434385754e-06, "loss": 0.0502, "step": 4230 }, { "epoch": 0.9626848691695108, "grad_norm": 1.67972447648695, "learning_rate": 1.1391598489075298e-06, "loss": 0.0898, "step": 4231 }, { "epoch": 0.9629124004550625, "grad_norm": 0.9225444046119701, "learning_rate": 1.1391090438679986e-06, "loss": 0.0589, "step": 4232 }, { "epoch": 0.9631399317406143, "grad_norm": 0.8217198859736241, "learning_rate": 1.1390582283210199e-06, "loss": 0.0515, "step": 4233 }, { "epoch": 0.963367463026166, "grad_norm": 1.4166504015117989, "learning_rate": 1.1390074022676325e-06, "loss": 0.0434, "step": 4234 }, { "epoch": 0.9635949943117179, "grad_norm": 1.8246388670882077, "learning_rate": 1.1389565657088752e-06, "loss": 0.0696, "step": 4235 }, { "epoch": 0.9638225255972697, "grad_norm": 1.984425508646575, "learning_rate": 1.1389057186457868e-06, "loss": 0.139, "step": 4236 }, { "epoch": 0.9640500568828214, "grad_norm": 2.480312140442047, "learning_rate": 1.1388548610794069e-06, "loss": 0.1062, "step": 4237 }, { "epoch": 0.9642775881683732, "grad_norm": 1.0531496575917552, "learning_rate": 1.1388039930107747e-06, "loss": 0.0416, "step": 4238 }, { "epoch": 0.9645051194539249, "grad_norm": 1.4873297621523849, "learning_rate": 1.1387531144409297e-06, "loss": 0.0418, "step": 4239 }, { "epoch": 0.9647326507394767, "grad_norm": 0.9204083159334068, "learning_rate": 1.138702225370912e-06, "loss": 0.0365, "step": 4240 }, { "epoch": 0.9649601820250284, "grad_norm": 1.193250713664879, "learning_rate": 1.1386513258017617e-06, "loss": 0.057, "step": 4241 }, { "epoch": 0.9651877133105802, "grad_norm": 1.1349796974488853, "learning_rate": 1.138600415734519e-06, "loss": 0.0321, "step": 4242 }, { "epoch": 0.9654152445961319, "grad_norm": 1.0770052024267605, "learning_rate": 1.1385494951702245e-06, "loss": 0.0683, "step": 4243 }, { "epoch": 0.9656427758816837, "grad_norm": 1.4502511589216216, "learning_rate": 1.1384985641099187e-06, "loss": 0.0729, "step": 4244 }, { "epoch": 0.9658703071672355, "grad_norm": 1.1356940815876257, "learning_rate": 1.1384476225546426e-06, "loss": 0.0366, "step": 4245 }, { "epoch": 0.9660978384527873, "grad_norm": 0.9898762934279589, "learning_rate": 1.1383966705054377e-06, "loss": 0.0408, "step": 4246 }, { "epoch": 0.966325369738339, "grad_norm": 1.7320554731048066, "learning_rate": 1.1383457079633448e-06, "loss": 0.0668, "step": 4247 }, { "epoch": 0.9665529010238908, "grad_norm": 1.1627771449836428, "learning_rate": 1.138294734929406e-06, "loss": 0.0565, "step": 4248 }, { "epoch": 0.9667804323094426, "grad_norm": 1.4667218418425674, "learning_rate": 1.1382437514046627e-06, "loss": 0.0572, "step": 4249 }, { "epoch": 0.9670079635949943, "grad_norm": 1.3102476422588094, "learning_rate": 1.1381927573901572e-06, "loss": 0.045, "step": 4250 }, { "epoch": 0.967235494880546, "grad_norm": 0.9659550904946845, "learning_rate": 1.1381417528869316e-06, "loss": 0.0562, "step": 4251 }, { "epoch": 0.9674630261660978, "grad_norm": 1.208698720888652, "learning_rate": 1.1380907378960282e-06, "loss": 0.0601, "step": 4252 }, { "epoch": 0.9676905574516496, "grad_norm": 1.3703900042435244, "learning_rate": 1.13803971241849e-06, "loss": 0.0498, "step": 4253 }, { "epoch": 0.9679180887372013, "grad_norm": 1.3965879380443744, "learning_rate": 1.1379886764553596e-06, "loss": 0.0457, "step": 4254 }, { "epoch": 0.9681456200227532, "grad_norm": 1.4360003219255868, "learning_rate": 1.1379376300076803e-06, "loss": 0.0689, "step": 4255 }, { "epoch": 0.9683731513083049, "grad_norm": 1.1977161122633277, "learning_rate": 1.137886573076495e-06, "loss": 0.0455, "step": 4256 }, { "epoch": 0.9686006825938567, "grad_norm": 1.5336831401293278, "learning_rate": 1.1378355056628474e-06, "loss": 0.0749, "step": 4257 }, { "epoch": 0.9688282138794084, "grad_norm": 1.842256713795995, "learning_rate": 1.1377844277677815e-06, "loss": 0.0702, "step": 4258 }, { "epoch": 0.9690557451649602, "grad_norm": 1.3257914742909727, "learning_rate": 1.1377333393923408e-06, "loss": 0.0427, "step": 4259 }, { "epoch": 0.9692832764505119, "grad_norm": 0.9651098248872604, "learning_rate": 1.1376822405375698e-06, "loss": 0.0309, "step": 4260 }, { "epoch": 0.9695108077360637, "grad_norm": 1.406730207103498, "learning_rate": 1.1376311312045128e-06, "loss": 0.0467, "step": 4261 }, { "epoch": 0.9697383390216154, "grad_norm": 1.1340714191157526, "learning_rate": 1.1375800113942144e-06, "loss": 0.0414, "step": 4262 }, { "epoch": 0.9699658703071672, "grad_norm": 1.0587038466584116, "learning_rate": 1.137528881107719e-06, "loss": 0.0411, "step": 4263 }, { "epoch": 0.970193401592719, "grad_norm": 1.375758096041912, "learning_rate": 1.137477740346072e-06, "loss": 0.0532, "step": 4264 }, { "epoch": 0.9704209328782708, "grad_norm": 1.3742948141863867, "learning_rate": 1.1374265891103187e-06, "loss": 0.0581, "step": 4265 }, { "epoch": 0.9706484641638226, "grad_norm": 1.8817386199561381, "learning_rate": 1.1373754274015044e-06, "loss": 0.0567, "step": 4266 }, { "epoch": 0.9708759954493743, "grad_norm": 1.8821493407596268, "learning_rate": 1.1373242552206744e-06, "loss": 0.0881, "step": 4267 }, { "epoch": 0.9711035267349261, "grad_norm": 1.2074529205184104, "learning_rate": 1.1372730725688754e-06, "loss": 0.0634, "step": 4268 }, { "epoch": 0.9713310580204778, "grad_norm": 1.193289676669864, "learning_rate": 1.1372218794471527e-06, "loss": 0.0435, "step": 4269 }, { "epoch": 0.9715585893060296, "grad_norm": 1.119752010069939, "learning_rate": 1.1371706758565529e-06, "loss": 0.0391, "step": 4270 }, { "epoch": 0.9717861205915813, "grad_norm": 1.4265721705104009, "learning_rate": 1.1371194617981224e-06, "loss": 0.0483, "step": 4271 }, { "epoch": 0.9720136518771331, "grad_norm": 2.0794425956744775, "learning_rate": 1.137068237272908e-06, "loss": 0.0806, "step": 4272 }, { "epoch": 0.9722411831626848, "grad_norm": 0.8292125442794063, "learning_rate": 1.1370170022819569e-06, "loss": 0.0266, "step": 4273 }, { "epoch": 0.9724687144482367, "grad_norm": 1.5988253258989058, "learning_rate": 1.1369657568263157e-06, "loss": 0.064, "step": 4274 }, { "epoch": 0.9726962457337884, "grad_norm": 1.4884666998450276, "learning_rate": 1.1369145009070323e-06, "loss": 0.0595, "step": 4275 }, { "epoch": 0.9729237770193402, "grad_norm": 1.0787898499294646, "learning_rate": 1.1368632345251538e-06, "loss": 0.0449, "step": 4276 }, { "epoch": 0.9731513083048919, "grad_norm": 1.2761275024398588, "learning_rate": 1.1368119576817283e-06, "loss": 0.0473, "step": 4277 }, { "epoch": 0.9733788395904437, "grad_norm": 0.9499141582616558, "learning_rate": 1.1367606703778037e-06, "loss": 0.0376, "step": 4278 }, { "epoch": 0.9736063708759954, "grad_norm": 1.0046565168504633, "learning_rate": 1.1367093726144283e-06, "loss": 0.0432, "step": 4279 }, { "epoch": 0.9738339021615472, "grad_norm": 1.0659038125870088, "learning_rate": 1.1366580643926506e-06, "loss": 0.0404, "step": 4280 }, { "epoch": 0.974061433447099, "grad_norm": 1.595011229834194, "learning_rate": 1.1366067457135188e-06, "loss": 0.0794, "step": 4281 }, { "epoch": 0.9742889647326507, "grad_norm": 1.968290906084048, "learning_rate": 1.1365554165780823e-06, "loss": 0.0673, "step": 4282 }, { "epoch": 0.9745164960182024, "grad_norm": 1.511125938151934, "learning_rate": 1.13650407698739e-06, "loss": 0.0347, "step": 4283 }, { "epoch": 0.9747440273037543, "grad_norm": 1.6319247438179008, "learning_rate": 1.136452726942491e-06, "loss": 0.0582, "step": 4284 }, { "epoch": 0.9749715585893061, "grad_norm": 1.3918200009388155, "learning_rate": 1.1364013664444351e-06, "loss": 0.053, "step": 4285 }, { "epoch": 0.9751990898748578, "grad_norm": 1.5863047987952235, "learning_rate": 1.1363499954942717e-06, "loss": 0.0704, "step": 4286 }, { "epoch": 0.9754266211604096, "grad_norm": 0.9811314326155632, "learning_rate": 1.1362986140930509e-06, "loss": 0.0341, "step": 4287 }, { "epoch": 0.9756541524459613, "grad_norm": 1.7299011978403993, "learning_rate": 1.1362472222418228e-06, "loss": 0.0831, "step": 4288 }, { "epoch": 0.9758816837315131, "grad_norm": 1.4033619817926257, "learning_rate": 1.1361958199416378e-06, "loss": 0.0559, "step": 4289 }, { "epoch": 0.9761092150170648, "grad_norm": 1.6430727260435547, "learning_rate": 1.1361444071935467e-06, "loss": 0.0666, "step": 4290 }, { "epoch": 0.9763367463026166, "grad_norm": 1.5397750371074748, "learning_rate": 1.1360929839985998e-06, "loss": 0.0954, "step": 4291 }, { "epoch": 0.9765642775881683, "grad_norm": 1.1112858103130956, "learning_rate": 1.1360415503578485e-06, "loss": 0.0524, "step": 4292 }, { "epoch": 0.9767918088737202, "grad_norm": 2.049966956578, "learning_rate": 1.1359901062723437e-06, "loss": 0.1009, "step": 4293 }, { "epoch": 0.9770193401592719, "grad_norm": 2.123284701337966, "learning_rate": 1.1359386517431366e-06, "loss": 0.0741, "step": 4294 }, { "epoch": 0.9772468714448237, "grad_norm": 1.8057163155909235, "learning_rate": 1.1358871867712797e-06, "loss": 0.1112, "step": 4295 }, { "epoch": 0.9774744027303754, "grad_norm": 0.6815528807119272, "learning_rate": 1.1358357113578242e-06, "loss": 0.0279, "step": 4296 }, { "epoch": 0.9777019340159272, "grad_norm": 1.9569360573215808, "learning_rate": 1.1357842255038222e-06, "loss": 0.0712, "step": 4297 }, { "epoch": 0.977929465301479, "grad_norm": 1.2238824938808284, "learning_rate": 1.1357327292103266e-06, "loss": 0.068, "step": 4298 }, { "epoch": 0.9781569965870307, "grad_norm": 1.3361935047633542, "learning_rate": 1.135681222478389e-06, "loss": 0.0593, "step": 4299 }, { "epoch": 0.9783845278725825, "grad_norm": 0.9204735606624523, "learning_rate": 1.1356297053090623e-06, "loss": 0.0433, "step": 4300 }, { "epoch": 0.9786120591581342, "grad_norm": 1.6545159240714982, "learning_rate": 1.1355781777033998e-06, "loss": 0.0716, "step": 4301 }, { "epoch": 0.978839590443686, "grad_norm": 1.3389750327964336, "learning_rate": 1.1355266396624545e-06, "loss": 0.0788, "step": 4302 }, { "epoch": 0.9790671217292378, "grad_norm": 1.2850002640158098, "learning_rate": 1.1354750911872795e-06, "loss": 0.0563, "step": 4303 }, { "epoch": 0.9792946530147896, "grad_norm": 2.1844096287203585, "learning_rate": 1.1354235322789286e-06, "loss": 0.1006, "step": 4304 }, { "epoch": 0.9795221843003413, "grad_norm": 1.2563539759355846, "learning_rate": 1.1353719629384554e-06, "loss": 0.0732, "step": 4305 }, { "epoch": 0.9797497155858931, "grad_norm": 1.4434315214333129, "learning_rate": 1.135320383166914e-06, "loss": 0.0704, "step": 4306 }, { "epoch": 0.9799772468714448, "grad_norm": 1.5960905043054407, "learning_rate": 1.1352687929653586e-06, "loss": 0.0837, "step": 4307 }, { "epoch": 0.9802047781569966, "grad_norm": 1.4152051878506084, "learning_rate": 1.1352171923348438e-06, "loss": 0.06, "step": 4308 }, { "epoch": 0.9804323094425483, "grad_norm": 1.3288142940732306, "learning_rate": 1.1351655812764236e-06, "loss": 0.0604, "step": 4309 }, { "epoch": 0.9806598407281001, "grad_norm": 0.9762848070268094, "learning_rate": 1.1351139597911536e-06, "loss": 0.0392, "step": 4310 }, { "epoch": 0.9808873720136518, "grad_norm": 2.015722359424784, "learning_rate": 1.1350623278800884e-06, "loss": 0.0572, "step": 4311 }, { "epoch": 0.9811149032992036, "grad_norm": 0.7081344517803854, "learning_rate": 1.1350106855442833e-06, "loss": 0.0276, "step": 4312 }, { "epoch": 0.9813424345847555, "grad_norm": 0.846185742491215, "learning_rate": 1.134959032784794e-06, "loss": 0.026, "step": 4313 }, { "epoch": 0.9815699658703072, "grad_norm": 1.3352329035862636, "learning_rate": 1.1349073696026759e-06, "loss": 0.057, "step": 4314 }, { "epoch": 0.981797497155859, "grad_norm": 0.8695753542918584, "learning_rate": 1.1348556959989848e-06, "loss": 0.0229, "step": 4315 }, { "epoch": 0.9820250284414107, "grad_norm": 1.3890644514305241, "learning_rate": 1.1348040119747771e-06, "loss": 0.052, "step": 4316 }, { "epoch": 0.9822525597269625, "grad_norm": 1.870756610162731, "learning_rate": 1.1347523175311092e-06, "loss": 0.0653, "step": 4317 }, { "epoch": 0.9824800910125142, "grad_norm": 1.4997312982720814, "learning_rate": 1.1347006126690377e-06, "loss": 0.1015, "step": 4318 }, { "epoch": 0.982707622298066, "grad_norm": 1.0165185093361495, "learning_rate": 1.1346488973896188e-06, "loss": 0.0291, "step": 4319 }, { "epoch": 0.9829351535836177, "grad_norm": 1.380478111887299, "learning_rate": 1.13459717169391e-06, "loss": 0.0632, "step": 4320 }, { "epoch": 0.9831626848691695, "grad_norm": 1.641625597159548, "learning_rate": 1.1345454355829682e-06, "loss": 0.0921, "step": 4321 }, { "epoch": 0.9833902161547212, "grad_norm": 1.0317618560602355, "learning_rate": 1.1344936890578508e-06, "loss": 0.0407, "step": 4322 }, { "epoch": 0.9836177474402731, "grad_norm": 1.310711884960193, "learning_rate": 1.1344419321196156e-06, "loss": 0.0685, "step": 4323 }, { "epoch": 0.9838452787258248, "grad_norm": 1.1047706035095923, "learning_rate": 1.1343901647693204e-06, "loss": 0.0565, "step": 4324 }, { "epoch": 0.9840728100113766, "grad_norm": 0.7843402054936461, "learning_rate": 1.134338387008023e-06, "loss": 0.0304, "step": 4325 }, { "epoch": 0.9843003412969283, "grad_norm": 1.1819853377536353, "learning_rate": 1.134286598836782e-06, "loss": 0.0661, "step": 4326 }, { "epoch": 0.9845278725824801, "grad_norm": 1.0132047480735697, "learning_rate": 1.1342348002566553e-06, "loss": 0.0327, "step": 4327 }, { "epoch": 0.9847554038680318, "grad_norm": 0.8226016309372456, "learning_rate": 1.1341829912687023e-06, "loss": 0.0437, "step": 4328 }, { "epoch": 0.9849829351535836, "grad_norm": 1.254492363059247, "learning_rate": 1.134131171873981e-06, "loss": 0.048, "step": 4329 }, { "epoch": 0.9852104664391353, "grad_norm": 1.9410455746016961, "learning_rate": 1.1340793420735514e-06, "loss": 0.0688, "step": 4330 }, { "epoch": 0.9854379977246871, "grad_norm": 1.2149172788419882, "learning_rate": 1.1340275018684722e-06, "loss": 0.055, "step": 4331 }, { "epoch": 0.985665529010239, "grad_norm": 1.135352471755647, "learning_rate": 1.133975651259803e-06, "loss": 0.0367, "step": 4332 }, { "epoch": 0.9858930602957907, "grad_norm": 1.3243570338612756, "learning_rate": 1.1339237902486037e-06, "loss": 0.0546, "step": 4333 }, { "epoch": 0.9861205915813425, "grad_norm": 1.3460820001864684, "learning_rate": 1.1338719188359343e-06, "loss": 0.0588, "step": 4334 }, { "epoch": 0.9863481228668942, "grad_norm": 0.9640088217580686, "learning_rate": 1.1338200370228546e-06, "loss": 0.0266, "step": 4335 }, { "epoch": 0.986575654152446, "grad_norm": 1.0087770932683686, "learning_rate": 1.1337681448104254e-06, "loss": 0.0549, "step": 4336 }, { "epoch": 0.9868031854379977, "grad_norm": 1.0628308289235069, "learning_rate": 1.1337162421997072e-06, "loss": 0.06, "step": 4337 }, { "epoch": 0.9870307167235495, "grad_norm": 0.9968869621613908, "learning_rate": 1.1336643291917604e-06, "loss": 0.0406, "step": 4338 }, { "epoch": 0.9872582480091012, "grad_norm": 1.3414457698222066, "learning_rate": 1.1336124057876464e-06, "loss": 0.059, "step": 4339 }, { "epoch": 0.987485779294653, "grad_norm": 1.1669110242397096, "learning_rate": 1.1335604719884264e-06, "loss": 0.0534, "step": 4340 }, { "epoch": 0.9877133105802047, "grad_norm": 2.0594889831279968, "learning_rate": 1.1335085277951616e-06, "loss": 0.108, "step": 4341 }, { "epoch": 0.9879408418657566, "grad_norm": 1.4063745875081306, "learning_rate": 1.1334565732089138e-06, "loss": 0.085, "step": 4342 }, { "epoch": 0.9881683731513083, "grad_norm": 1.501936982721738, "learning_rate": 1.133404608230745e-06, "loss": 0.0857, "step": 4343 }, { "epoch": 0.9883959044368601, "grad_norm": 1.2396246613907491, "learning_rate": 1.1333526328617168e-06, "loss": 0.0783, "step": 4344 }, { "epoch": 0.9886234357224118, "grad_norm": 1.4534107446508395, "learning_rate": 1.133300647102892e-06, "loss": 0.0359, "step": 4345 }, { "epoch": 0.9888509670079636, "grad_norm": 1.5794373443169556, "learning_rate": 1.1332486509553328e-06, "loss": 0.0433, "step": 4346 }, { "epoch": 0.9890784982935154, "grad_norm": 1.2904275476428004, "learning_rate": 1.133196644420102e-06, "loss": 0.0554, "step": 4347 }, { "epoch": 0.9893060295790671, "grad_norm": 1.1570569489010505, "learning_rate": 1.1331446274982625e-06, "loss": 0.0472, "step": 4348 }, { "epoch": 0.9895335608646189, "grad_norm": 1.145673861183969, "learning_rate": 1.1330926001908777e-06, "loss": 0.0458, "step": 4349 }, { "epoch": 0.9897610921501706, "grad_norm": 0.896667580653844, "learning_rate": 1.1330405624990104e-06, "loss": 0.0234, "step": 4350 }, { "epoch": 0.9899886234357224, "grad_norm": 0.9213529689842777, "learning_rate": 1.1329885144237243e-06, "loss": 0.0381, "step": 4351 }, { "epoch": 0.9902161547212742, "grad_norm": 1.0487384771504538, "learning_rate": 1.1329364559660836e-06, "loss": 0.0364, "step": 4352 }, { "epoch": 0.990443686006826, "grad_norm": 1.4233320390454558, "learning_rate": 1.132884387127152e-06, "loss": 0.0527, "step": 4353 }, { "epoch": 0.9906712172923777, "grad_norm": 1.4228059163710602, "learning_rate": 1.1328323079079934e-06, "loss": 0.0554, "step": 4354 }, { "epoch": 0.9908987485779295, "grad_norm": 1.1059331642608046, "learning_rate": 1.1327802183096725e-06, "loss": 0.0341, "step": 4355 }, { "epoch": 0.9911262798634812, "grad_norm": 0.9827406192573771, "learning_rate": 1.1327281183332542e-06, "loss": 0.0367, "step": 4356 }, { "epoch": 0.991353811149033, "grad_norm": 1.1884213703636224, "learning_rate": 1.1326760079798027e-06, "loss": 0.0491, "step": 4357 }, { "epoch": 0.9915813424345847, "grad_norm": 1.656723752079525, "learning_rate": 1.1326238872503837e-06, "loss": 0.045, "step": 4358 }, { "epoch": 0.9918088737201365, "grad_norm": 1.0844028312793037, "learning_rate": 1.1325717561460617e-06, "loss": 0.0339, "step": 4359 }, { "epoch": 0.9920364050056882, "grad_norm": 1.0891260266689944, "learning_rate": 1.132519614667903e-06, "loss": 0.0466, "step": 4360 }, { "epoch": 0.9922639362912401, "grad_norm": 1.7100161526383912, "learning_rate": 1.1324674628169725e-06, "loss": 0.0889, "step": 4361 }, { "epoch": 0.9924914675767919, "grad_norm": 1.0880829131016376, "learning_rate": 1.1324153005943367e-06, "loss": 0.0431, "step": 4362 }, { "epoch": 0.9927189988623436, "grad_norm": 0.8884823165051863, "learning_rate": 1.1323631280010611e-06, "loss": 0.0447, "step": 4363 }, { "epoch": 0.9929465301478954, "grad_norm": 1.8533973089871183, "learning_rate": 1.1323109450382128e-06, "loss": 0.082, "step": 4364 }, { "epoch": 0.9931740614334471, "grad_norm": 0.9974325526425775, "learning_rate": 1.1322587517068576e-06, "loss": 0.0319, "step": 4365 }, { "epoch": 0.9934015927189989, "grad_norm": 0.9460101960780305, "learning_rate": 1.1322065480080625e-06, "loss": 0.0433, "step": 4366 }, { "epoch": 0.9936291240045506, "grad_norm": 2.537972684614883, "learning_rate": 1.1321543339428946e-06, "loss": 0.1302, "step": 4367 }, { "epoch": 0.9938566552901024, "grad_norm": 0.9361607430120513, "learning_rate": 1.132102109512421e-06, "loss": 0.0414, "step": 4368 }, { "epoch": 0.9940841865756541, "grad_norm": 1.3835872703988714, "learning_rate": 1.1320498747177088e-06, "loss": 0.0539, "step": 4369 }, { "epoch": 0.9943117178612059, "grad_norm": 1.4613596711487815, "learning_rate": 1.1319976295598258e-06, "loss": 0.0695, "step": 4370 }, { "epoch": 0.9945392491467577, "grad_norm": 1.3564131183677717, "learning_rate": 1.1319453740398397e-06, "loss": 0.0438, "step": 4371 }, { "epoch": 0.9947667804323095, "grad_norm": 1.312368604701528, "learning_rate": 1.1318931081588188e-06, "loss": 0.0539, "step": 4372 }, { "epoch": 0.9949943117178612, "grad_norm": 2.852729748963175, "learning_rate": 1.1318408319178308e-06, "loss": 0.1051, "step": 4373 }, { "epoch": 0.995221843003413, "grad_norm": 1.174771913378081, "learning_rate": 1.1317885453179448e-06, "loss": 0.0345, "step": 4374 }, { "epoch": 0.9954493742889647, "grad_norm": 1.156608166877398, "learning_rate": 1.131736248360229e-06, "loss": 0.0507, "step": 4375 }, { "epoch": 0.9956769055745165, "grad_norm": 1.7246435858174831, "learning_rate": 1.1316839410457523e-06, "loss": 0.0753, "step": 4376 }, { "epoch": 0.9959044368600682, "grad_norm": 1.190785748773095, "learning_rate": 1.1316316233755837e-06, "loss": 0.048, "step": 4377 }, { "epoch": 0.99613196814562, "grad_norm": 1.3884121672018674, "learning_rate": 1.1315792953507924e-06, "loss": 0.054, "step": 4378 }, { "epoch": 0.9963594994311717, "grad_norm": 1.1099500717065285, "learning_rate": 1.1315269569724483e-06, "loss": 0.0578, "step": 4379 }, { "epoch": 0.9965870307167235, "grad_norm": 1.1511233150222433, "learning_rate": 1.131474608241621e-06, "loss": 0.0709, "step": 4380 }, { "epoch": 0.9968145620022754, "grad_norm": 0.9955642314650807, "learning_rate": 1.1314222491593798e-06, "loss": 0.0358, "step": 4381 }, { "epoch": 0.9970420932878271, "grad_norm": 1.434440905613479, "learning_rate": 1.1313698797267958e-06, "loss": 0.0486, "step": 4382 }, { "epoch": 0.9972696245733789, "grad_norm": 2.8294165726638822, "learning_rate": 1.1313174999449384e-06, "loss": 0.0813, "step": 4383 }, { "epoch": 0.9974971558589306, "grad_norm": 1.2319487216248912, "learning_rate": 1.1312651098148788e-06, "loss": 0.0432, "step": 4384 }, { "epoch": 0.9977246871444824, "grad_norm": 1.2286306009661918, "learning_rate": 1.1312127093376876e-06, "loss": 0.0498, "step": 4385 }, { "epoch": 0.9979522184300341, "grad_norm": 1.4020060740030407, "learning_rate": 1.1311602985144358e-06, "loss": 0.0618, "step": 4386 }, { "epoch": 0.9981797497155859, "grad_norm": 1.2218921395546485, "learning_rate": 1.1311078773461942e-06, "loss": 0.0704, "step": 4387 }, { "epoch": 0.9984072810011376, "grad_norm": 1.2701236465901293, "learning_rate": 1.1310554458340345e-06, "loss": 0.0494, "step": 4388 }, { "epoch": 0.9986348122866894, "grad_norm": 1.2231395933687852, "learning_rate": 1.1310030039790285e-06, "loss": 0.0548, "step": 4389 }, { "epoch": 0.9988623435722411, "grad_norm": 1.228744395379064, "learning_rate": 1.1309505517822476e-06, "loss": 0.0456, "step": 4390 }, { "epoch": 0.999089874857793, "grad_norm": 1.0175121426182296, "learning_rate": 1.1308980892447641e-06, "loss": 0.049, "step": 4391 }, { "epoch": 0.9993174061433447, "grad_norm": 1.144658935981794, "learning_rate": 1.1308456163676501e-06, "loss": 0.0487, "step": 4392 }, { "epoch": 0.9995449374288965, "grad_norm": 1.1385156907253697, "learning_rate": 1.1307931331519783e-06, "loss": 0.051, "step": 4393 }, { "epoch": 0.9997724687144482, "grad_norm": 1.2941484330489015, "learning_rate": 1.1307406395988211e-06, "loss": 0.0587, "step": 4394 }, { "epoch": 1.0, "grad_norm": 0.7645843200332254, "learning_rate": 1.1306881357092513e-06, "loss": 0.0273, "step": 4395 }, { "epoch": 1.0002275312855518, "grad_norm": 0.5440562375961928, "learning_rate": 1.1306356214843423e-06, "loss": 0.0133, "step": 4396 }, { "epoch": 1.0004550625711035, "grad_norm": 1.048557888593317, "learning_rate": 1.1305830969251672e-06, "loss": 0.0234, "step": 4397 }, { "epoch": 1.0006825938566553, "grad_norm": 0.7566376417333784, "learning_rate": 1.1305305620327994e-06, "loss": 0.0208, "step": 4398 }, { "epoch": 1.000910125142207, "grad_norm": 0.6722933508685324, "learning_rate": 1.1304780168083128e-06, "loss": 0.0252, "step": 4399 }, { "epoch": 1.0011376564277588, "grad_norm": 0.9157174579220916, "learning_rate": 1.1304254612527815e-06, "loss": 0.0292, "step": 4400 }, { "epoch": 1.0013651877133105, "grad_norm": 0.7682748184714413, "learning_rate": 1.130372895367279e-06, "loss": 0.0305, "step": 4401 }, { "epoch": 1.0015927189988623, "grad_norm": 0.4692149259132462, "learning_rate": 1.1303203191528803e-06, "loss": 0.011, "step": 4402 }, { "epoch": 1.001820250284414, "grad_norm": 0.6034838800549486, "learning_rate": 1.1302677326106598e-06, "loss": 0.0158, "step": 4403 }, { "epoch": 1.0020477815699658, "grad_norm": 0.5139829099740163, "learning_rate": 1.130215135741692e-06, "loss": 0.0097, "step": 4404 }, { "epoch": 1.0022753128555177, "grad_norm": 0.6233625169929209, "learning_rate": 1.1301625285470522e-06, "loss": 0.0209, "step": 4405 }, { "epoch": 1.0025028441410695, "grad_norm": 0.8715203529693301, "learning_rate": 1.1301099110278156e-06, "loss": 0.0349, "step": 4406 }, { "epoch": 1.0027303754266212, "grad_norm": 0.5809712252863496, "learning_rate": 1.1300572831850574e-06, "loss": 0.0173, "step": 4407 }, { "epoch": 1.002957906712173, "grad_norm": 0.869776635220801, "learning_rate": 1.1300046450198532e-06, "loss": 0.0349, "step": 4408 }, { "epoch": 1.0031854379977247, "grad_norm": 0.6891595017115912, "learning_rate": 1.1299519965332791e-06, "loss": 0.0218, "step": 4409 }, { "epoch": 1.0034129692832765, "grad_norm": 2.771272397471203, "learning_rate": 1.1298993377264108e-06, "loss": 0.1173, "step": 4410 }, { "epoch": 1.0036405005688283, "grad_norm": 0.6496153724080771, "learning_rate": 1.129846668600325e-06, "loss": 0.0165, "step": 4411 }, { "epoch": 1.00386803185438, "grad_norm": 0.9105625202645007, "learning_rate": 1.1297939891560975e-06, "loss": 0.0356, "step": 4412 }, { "epoch": 1.0040955631399318, "grad_norm": 0.954002149635865, "learning_rate": 1.1297412993948054e-06, "loss": 0.0419, "step": 4413 }, { "epoch": 1.0043230944254835, "grad_norm": 0.8420422197036498, "learning_rate": 1.1296885993175255e-06, "loss": 0.0249, "step": 4414 }, { "epoch": 1.0045506257110353, "grad_norm": 1.1194729404929513, "learning_rate": 1.1296358889253351e-06, "loss": 0.0282, "step": 4415 }, { "epoch": 1.004778156996587, "grad_norm": 0.664863679519903, "learning_rate": 1.1295831682193115e-06, "loss": 0.0232, "step": 4416 }, { "epoch": 1.0050056882821388, "grad_norm": 0.7352288120040293, "learning_rate": 1.1295304372005316e-06, "loss": 0.0308, "step": 4417 }, { "epoch": 1.0052332195676905, "grad_norm": 0.9912305346960562, "learning_rate": 1.129477695870074e-06, "loss": 0.0454, "step": 4418 }, { "epoch": 1.0054607508532423, "grad_norm": 0.5967781019650417, "learning_rate": 1.129424944229016e-06, "loss": 0.0145, "step": 4419 }, { "epoch": 1.005688282138794, "grad_norm": 0.8293958411943283, "learning_rate": 1.1293721822784359e-06, "loss": 0.0189, "step": 4420 }, { "epoch": 1.0059158134243458, "grad_norm": 0.9305770629796571, "learning_rate": 1.1293194100194121e-06, "loss": 0.0258, "step": 4421 }, { "epoch": 1.0061433447098975, "grad_norm": 0.7187164975510966, "learning_rate": 1.1292666274530232e-06, "loss": 0.0165, "step": 4422 }, { "epoch": 1.0063708759954493, "grad_norm": 0.6821469033121376, "learning_rate": 1.129213834580348e-06, "loss": 0.0147, "step": 4423 }, { "epoch": 1.006598407281001, "grad_norm": 0.5603983030125896, "learning_rate": 1.1291610314024653e-06, "loss": 0.0135, "step": 4424 }, { "epoch": 1.006825938566553, "grad_norm": 0.7667715549060277, "learning_rate": 1.1291082179204548e-06, "loss": 0.0202, "step": 4425 }, { "epoch": 1.0070534698521048, "grad_norm": 0.5411620244180794, "learning_rate": 1.1290553941353954e-06, "loss": 0.0103, "step": 4426 }, { "epoch": 1.0072810011376565, "grad_norm": 0.6055971311258457, "learning_rate": 1.1290025600483667e-06, "loss": 0.0228, "step": 4427 }, { "epoch": 1.0075085324232083, "grad_norm": 1.2109757139019808, "learning_rate": 1.1289497156604487e-06, "loss": 0.0437, "step": 4428 }, { "epoch": 1.00773606370876, "grad_norm": 0.5370646714370468, "learning_rate": 1.1288968609727216e-06, "loss": 0.0085, "step": 4429 }, { "epoch": 1.0079635949943118, "grad_norm": 0.9137582688267473, "learning_rate": 1.1288439959862654e-06, "loss": 0.0349, "step": 4430 }, { "epoch": 1.0081911262798635, "grad_norm": 0.6202104523174599, "learning_rate": 1.128791120702161e-06, "loss": 0.0244, "step": 4431 }, { "epoch": 1.0084186575654153, "grad_norm": 1.264397001473755, "learning_rate": 1.1287382351214884e-06, "loss": 0.0385, "step": 4432 }, { "epoch": 1.008646188850967, "grad_norm": 0.4733349727968622, "learning_rate": 1.128685339245329e-06, "loss": 0.0157, "step": 4433 }, { "epoch": 1.0088737201365188, "grad_norm": 0.7488096549319501, "learning_rate": 1.1286324330747637e-06, "loss": 0.0237, "step": 4434 }, { "epoch": 1.0091012514220705, "grad_norm": 0.6215090350542927, "learning_rate": 1.1285795166108735e-06, "loss": 0.0165, "step": 4435 }, { "epoch": 1.0093287827076223, "grad_norm": 0.9977188909479956, "learning_rate": 1.1285265898547406e-06, "loss": 0.0269, "step": 4436 }, { "epoch": 1.009556313993174, "grad_norm": 1.3828220847485446, "learning_rate": 1.1284736528074464e-06, "loss": 0.0327, "step": 4437 }, { "epoch": 1.0097838452787258, "grad_norm": 0.8745376030427436, "learning_rate": 1.1284207054700727e-06, "loss": 0.0223, "step": 4438 }, { "epoch": 1.0100113765642775, "grad_norm": 0.710421250041984, "learning_rate": 1.1283677478437016e-06, "loss": 0.0206, "step": 4439 }, { "epoch": 1.0102389078498293, "grad_norm": 0.9294364308935749, "learning_rate": 1.1283147799294158e-06, "loss": 0.023, "step": 4440 }, { "epoch": 1.010466439135381, "grad_norm": 1.1336723842425673, "learning_rate": 1.1282618017282977e-06, "loss": 0.0543, "step": 4441 }, { "epoch": 1.0106939704209328, "grad_norm": 0.6236066585507253, "learning_rate": 1.1282088132414297e-06, "loss": 0.0177, "step": 4442 }, { "epoch": 1.0109215017064845, "grad_norm": 0.4538882994044952, "learning_rate": 1.1281558144698956e-06, "loss": 0.005, "step": 4443 }, { "epoch": 1.0111490329920365, "grad_norm": 0.9922582222331733, "learning_rate": 1.128102805414778e-06, "loss": 0.024, "step": 4444 }, { "epoch": 1.0113765642775883, "grad_norm": 0.939282752346539, "learning_rate": 1.1280497860771603e-06, "loss": 0.0373, "step": 4445 }, { "epoch": 1.01160409556314, "grad_norm": 0.6323505737761085, "learning_rate": 1.1279967564581264e-06, "loss": 0.0121, "step": 4446 }, { "epoch": 1.0118316268486918, "grad_norm": 1.0956673474878567, "learning_rate": 1.12794371655876e-06, "loss": 0.02, "step": 4447 }, { "epoch": 1.0120591581342435, "grad_norm": 0.9129627353008005, "learning_rate": 1.127890666380145e-06, "loss": 0.025, "step": 4448 }, { "epoch": 1.0122866894197953, "grad_norm": 0.46435154090423725, "learning_rate": 1.1278376059233658e-06, "loss": 0.0083, "step": 4449 }, { "epoch": 1.012514220705347, "grad_norm": 0.823583660478901, "learning_rate": 1.127784535189507e-06, "loss": 0.0242, "step": 4450 }, { "epoch": 1.0127417519908988, "grad_norm": 1.244611532738964, "learning_rate": 1.127731454179653e-06, "loss": 0.0477, "step": 4451 }, { "epoch": 1.0129692832764505, "grad_norm": 0.6536691853613382, "learning_rate": 1.1276783628948887e-06, "loss": 0.0203, "step": 4452 }, { "epoch": 1.0131968145620023, "grad_norm": 0.7200287149249455, "learning_rate": 1.1276252613362995e-06, "loss": 0.0238, "step": 4453 }, { "epoch": 1.013424345847554, "grad_norm": 0.8521168883159115, "learning_rate": 1.1275721495049702e-06, "loss": 0.0308, "step": 4454 }, { "epoch": 1.0136518771331058, "grad_norm": 0.6812814467705595, "learning_rate": 1.1275190274019867e-06, "loss": 0.0223, "step": 4455 }, { "epoch": 1.0138794084186575, "grad_norm": 0.8073573367282825, "learning_rate": 1.1274658950284347e-06, "loss": 0.0159, "step": 4456 }, { "epoch": 1.0141069397042093, "grad_norm": 0.9929398468232915, "learning_rate": 1.1274127523854e-06, "loss": 0.0174, "step": 4457 }, { "epoch": 1.014334470989761, "grad_norm": 0.8110259409039485, "learning_rate": 1.1273595994739688e-06, "loss": 0.0231, "step": 4458 }, { "epoch": 1.0145620022753128, "grad_norm": 0.8503939357738989, "learning_rate": 1.1273064362952272e-06, "loss": 0.0181, "step": 4459 }, { "epoch": 1.0147895335608645, "grad_norm": 0.952005853323446, "learning_rate": 1.1272532628502621e-06, "loss": 0.011, "step": 4460 }, { "epoch": 1.0150170648464163, "grad_norm": 0.979632492475044, "learning_rate": 1.1272000791401602e-06, "loss": 0.0289, "step": 4461 }, { "epoch": 1.015244596131968, "grad_norm": 0.9141108737856498, "learning_rate": 1.1271468851660084e-06, "loss": 0.0325, "step": 4462 }, { "epoch": 1.01547212741752, "grad_norm": 0.7032331801339051, "learning_rate": 1.127093680928894e-06, "loss": 0.0188, "step": 4463 }, { "epoch": 1.0156996587030718, "grad_norm": 0.9755139681497196, "learning_rate": 1.1270404664299042e-06, "loss": 0.0212, "step": 4464 }, { "epoch": 1.0159271899886235, "grad_norm": 0.623696536501648, "learning_rate": 1.1269872416701267e-06, "loss": 0.0143, "step": 4465 }, { "epoch": 1.0161547212741753, "grad_norm": 0.5413765005494656, "learning_rate": 1.1269340066506493e-06, "loss": 0.0101, "step": 4466 }, { "epoch": 1.016382252559727, "grad_norm": 0.6689689779774286, "learning_rate": 1.12688076137256e-06, "loss": 0.0114, "step": 4467 }, { "epoch": 1.0166097838452788, "grad_norm": 0.6453292072094783, "learning_rate": 1.1268275058369472e-06, "loss": 0.0107, "step": 4468 }, { "epoch": 1.0168373151308305, "grad_norm": 1.1332110733517493, "learning_rate": 1.1267742400448992e-06, "loss": 0.0217, "step": 4469 }, { "epoch": 1.0170648464163823, "grad_norm": 0.630093494299339, "learning_rate": 1.1267209639975046e-06, "loss": 0.0089, "step": 4470 }, { "epoch": 1.017292377701934, "grad_norm": 1.0118536500106963, "learning_rate": 1.1266676776958523e-06, "loss": 0.0208, "step": 4471 }, { "epoch": 1.0175199089874858, "grad_norm": 1.0518953123729393, "learning_rate": 1.1266143811410317e-06, "loss": 0.0201, "step": 4472 }, { "epoch": 1.0177474402730375, "grad_norm": 0.757424135828587, "learning_rate": 1.1265610743341316e-06, "loss": 0.0217, "step": 4473 }, { "epoch": 1.0179749715585893, "grad_norm": 0.77124606602971, "learning_rate": 1.1265077572762418e-06, "loss": 0.0262, "step": 4474 }, { "epoch": 1.018202502844141, "grad_norm": 1.7312334858326726, "learning_rate": 1.1264544299684518e-06, "loss": 0.0511, "step": 4475 }, { "epoch": 1.0184300341296928, "grad_norm": 0.8077840361458416, "learning_rate": 1.1264010924118518e-06, "loss": 0.0156, "step": 4476 }, { "epoch": 1.0186575654152445, "grad_norm": 0.9380910246867491, "learning_rate": 1.1263477446075315e-06, "loss": 0.0387, "step": 4477 }, { "epoch": 1.0188850967007963, "grad_norm": 0.7652818963146196, "learning_rate": 1.1262943865565818e-06, "loss": 0.0173, "step": 4478 }, { "epoch": 1.019112627986348, "grad_norm": 1.3337882989968501, "learning_rate": 1.1262410182600927e-06, "loss": 0.0396, "step": 4479 }, { "epoch": 1.0193401592718998, "grad_norm": 0.45812987319685416, "learning_rate": 1.1261876397191554e-06, "loss": 0.015, "step": 4480 }, { "epoch": 1.0195676905574516, "grad_norm": 0.7602276965168047, "learning_rate": 1.1261342509348604e-06, "loss": 0.0202, "step": 4481 }, { "epoch": 1.0197952218430033, "grad_norm": 0.9601272476880518, "learning_rate": 1.126080851908299e-06, "loss": 0.0245, "step": 4482 }, { "epoch": 1.0200227531285553, "grad_norm": 0.9302450496938718, "learning_rate": 1.1260274426405629e-06, "loss": 0.0251, "step": 4483 }, { "epoch": 1.020250284414107, "grad_norm": 1.064523523991496, "learning_rate": 1.1259740231327434e-06, "loss": 0.0295, "step": 4484 }, { "epoch": 1.0204778156996588, "grad_norm": 1.4809063729774259, "learning_rate": 1.1259205933859325e-06, "loss": 0.0268, "step": 4485 }, { "epoch": 1.0207053469852105, "grad_norm": 1.2332212440546455, "learning_rate": 1.1258671534012216e-06, "loss": 0.0432, "step": 4486 }, { "epoch": 1.0209328782707623, "grad_norm": 0.7799860888625535, "learning_rate": 1.1258137031797037e-06, "loss": 0.0124, "step": 4487 }, { "epoch": 1.021160409556314, "grad_norm": 0.759267946121651, "learning_rate": 1.125760242722471e-06, "loss": 0.0091, "step": 4488 }, { "epoch": 1.0213879408418658, "grad_norm": 0.891169288145688, "learning_rate": 1.1257067720306159e-06, "loss": 0.0286, "step": 4489 }, { "epoch": 1.0216154721274175, "grad_norm": 1.0432400031887874, "learning_rate": 1.1256532911052313e-06, "loss": 0.0334, "step": 4490 }, { "epoch": 1.0218430034129693, "grad_norm": 1.0537890506909147, "learning_rate": 1.1255997999474105e-06, "loss": 0.02, "step": 4491 }, { "epoch": 1.022070534698521, "grad_norm": 0.7817598770936856, "learning_rate": 1.1255462985582465e-06, "loss": 0.017, "step": 4492 }, { "epoch": 1.0222980659840728, "grad_norm": 0.9383161200612665, "learning_rate": 1.125492786938833e-06, "loss": 0.011, "step": 4493 }, { "epoch": 1.0225255972696246, "grad_norm": 1.7998282210253833, "learning_rate": 1.1254392650902633e-06, "loss": 0.0469, "step": 4494 }, { "epoch": 1.0227531285551763, "grad_norm": 0.9796354836057241, "learning_rate": 1.1253857330136316e-06, "loss": 0.0252, "step": 4495 }, { "epoch": 1.022980659840728, "grad_norm": 0.8615898866534331, "learning_rate": 1.125332190710032e-06, "loss": 0.0103, "step": 4496 }, { "epoch": 1.0232081911262798, "grad_norm": 1.066758301808706, "learning_rate": 1.125278638180559e-06, "loss": 0.0176, "step": 4497 }, { "epoch": 1.0234357224118316, "grad_norm": 0.7537124723757486, "learning_rate": 1.1252250754263064e-06, "loss": 0.0243, "step": 4498 }, { "epoch": 1.0236632536973833, "grad_norm": 0.7310817973699005, "learning_rate": 1.1251715024483695e-06, "loss": 0.0213, "step": 4499 }, { "epoch": 1.023890784982935, "grad_norm": 0.8578238852047406, "learning_rate": 1.125117919247843e-06, "loss": 0.0319, "step": 4500 }, { "epoch": 1.0241183162684868, "grad_norm": 0.9913368014981602, "learning_rate": 1.1250643258258225e-06, "loss": 0.0208, "step": 4501 }, { "epoch": 1.0243458475540388, "grad_norm": 1.2683092562362872, "learning_rate": 1.1250107221834027e-06, "loss": 0.0504, "step": 4502 }, { "epoch": 1.0245733788395905, "grad_norm": 0.7751769469902451, "learning_rate": 1.12495710832168e-06, "loss": 0.0278, "step": 4503 }, { "epoch": 1.0248009101251423, "grad_norm": 0.7428863858706001, "learning_rate": 1.1249034842417489e-06, "loss": 0.0141, "step": 4504 }, { "epoch": 1.025028441410694, "grad_norm": 0.7095783022904831, "learning_rate": 1.1248498499447065e-06, "loss": 0.019, "step": 4505 }, { "epoch": 1.0252559726962458, "grad_norm": 1.0630325639796825, "learning_rate": 1.1247962054316485e-06, "loss": 0.0439, "step": 4506 }, { "epoch": 1.0254835039817976, "grad_norm": 0.8040239643121114, "learning_rate": 1.1247425507036715e-06, "loss": 0.0189, "step": 4507 }, { "epoch": 1.0257110352673493, "grad_norm": 1.8105078367140066, "learning_rate": 1.1246888857618719e-06, "loss": 0.0633, "step": 4508 }, { "epoch": 1.025938566552901, "grad_norm": 1.3028583651888686, "learning_rate": 1.1246352106073466e-06, "loss": 0.0233, "step": 4509 }, { "epoch": 1.0261660978384528, "grad_norm": 1.3773748170809028, "learning_rate": 1.1245815252411928e-06, "loss": 0.0563, "step": 4510 }, { "epoch": 1.0263936291240046, "grad_norm": 1.362685911234451, "learning_rate": 1.1245278296645073e-06, "loss": 0.0257, "step": 4511 }, { "epoch": 1.0266211604095563, "grad_norm": 0.7561719123407112, "learning_rate": 1.124474123878388e-06, "loss": 0.0137, "step": 4512 }, { "epoch": 1.026848691695108, "grad_norm": 0.6802103464690515, "learning_rate": 1.1244204078839325e-06, "loss": 0.0148, "step": 4513 }, { "epoch": 1.0270762229806598, "grad_norm": 1.023770837391378, "learning_rate": 1.1243666816822382e-06, "loss": 0.0383, "step": 4514 }, { "epoch": 1.0273037542662116, "grad_norm": 1.105198417368013, "learning_rate": 1.1243129452744036e-06, "loss": 0.0169, "step": 4515 }, { "epoch": 1.0275312855517633, "grad_norm": 0.8424066208751216, "learning_rate": 1.1242591986615268e-06, "loss": 0.0308, "step": 4516 }, { "epoch": 1.027758816837315, "grad_norm": 0.666774886288571, "learning_rate": 1.1242054418447063e-06, "loss": 0.0093, "step": 4517 }, { "epoch": 1.0279863481228668, "grad_norm": 0.6569322003852172, "learning_rate": 1.1241516748250408e-06, "loss": 0.0118, "step": 4518 }, { "epoch": 1.0282138794084186, "grad_norm": 1.2076404281398103, "learning_rate": 1.1240978976036294e-06, "loss": 0.0489, "step": 4519 }, { "epoch": 1.0284414106939703, "grad_norm": 1.1551438155853788, "learning_rate": 1.124044110181571e-06, "loss": 0.0311, "step": 4520 }, { "epoch": 1.028668941979522, "grad_norm": 1.389595859599812, "learning_rate": 1.1239903125599648e-06, "loss": 0.0491, "step": 4521 }, { "epoch": 1.028896473265074, "grad_norm": 1.0333918284834915, "learning_rate": 1.1239365047399106e-06, "loss": 0.0404, "step": 4522 }, { "epoch": 1.0291240045506258, "grad_norm": 0.8599318194273396, "learning_rate": 1.1238826867225077e-06, "loss": 0.0197, "step": 4523 }, { "epoch": 1.0293515358361776, "grad_norm": 0.8304017967466861, "learning_rate": 1.1238288585088567e-06, "loss": 0.0171, "step": 4524 }, { "epoch": 1.0295790671217293, "grad_norm": 1.136424973741861, "learning_rate": 1.1237750201000574e-06, "loss": 0.0316, "step": 4525 }, { "epoch": 1.029806598407281, "grad_norm": 1.4555472105148153, "learning_rate": 1.1237211714972098e-06, "loss": 0.062, "step": 4526 }, { "epoch": 1.0300341296928328, "grad_norm": 1.3336128205554842, "learning_rate": 1.1236673127014152e-06, "loss": 0.0243, "step": 4527 }, { "epoch": 1.0302616609783846, "grad_norm": 0.9479976161551643, "learning_rate": 1.1236134437137738e-06, "loss": 0.022, "step": 4528 }, { "epoch": 1.0304891922639363, "grad_norm": 0.851610494023467, "learning_rate": 1.1235595645353869e-06, "loss": 0.0304, "step": 4529 }, { "epoch": 1.030716723549488, "grad_norm": 0.8155261530618487, "learning_rate": 1.1235056751673554e-06, "loss": 0.017, "step": 4530 }, { "epoch": 1.0309442548350398, "grad_norm": 1.1243416422484538, "learning_rate": 1.123451775610781e-06, "loss": 0.0248, "step": 4531 }, { "epoch": 1.0311717861205916, "grad_norm": 0.6551716981893974, "learning_rate": 1.1233978658667651e-06, "loss": 0.0251, "step": 4532 }, { "epoch": 1.0313993174061433, "grad_norm": 0.9437616699526395, "learning_rate": 1.1233439459364097e-06, "loss": 0.0228, "step": 4533 }, { "epoch": 1.031626848691695, "grad_norm": 1.245047538582983, "learning_rate": 1.1232900158208166e-06, "loss": 0.0148, "step": 4534 }, { "epoch": 1.0318543799772468, "grad_norm": 0.9658390993690719, "learning_rate": 1.1232360755210883e-06, "loss": 0.0166, "step": 4535 }, { "epoch": 1.0320819112627986, "grad_norm": 1.1047045638558348, "learning_rate": 1.123182125038327e-06, "loss": 0.0316, "step": 4536 }, { "epoch": 1.0323094425483503, "grad_norm": 1.3208237077699243, "learning_rate": 1.1231281643736353e-06, "loss": 0.0303, "step": 4537 }, { "epoch": 1.032536973833902, "grad_norm": 0.7809775193316096, "learning_rate": 1.1230741935281163e-06, "loss": 0.0196, "step": 4538 }, { "epoch": 1.0327645051194538, "grad_norm": 0.991878604252392, "learning_rate": 1.123020212502873e-06, "loss": 0.02, "step": 4539 }, { "epoch": 1.0329920364050056, "grad_norm": 1.3869865701838429, "learning_rate": 1.1229662212990088e-06, "loss": 0.0612, "step": 4540 }, { "epoch": 1.0332195676905576, "grad_norm": 0.619858198046472, "learning_rate": 1.1229122199176268e-06, "loss": 0.0072, "step": 4541 }, { "epoch": 1.0334470989761093, "grad_norm": 0.7040478929448408, "learning_rate": 1.1228582083598311e-06, "loss": 0.0257, "step": 4542 }, { "epoch": 1.033674630261661, "grad_norm": 1.182074276350949, "learning_rate": 1.122804186626725e-06, "loss": 0.0206, "step": 4543 }, { "epoch": 1.0339021615472128, "grad_norm": 0.42845477644473695, "learning_rate": 1.1227501547194133e-06, "loss": 0.0111, "step": 4544 }, { "epoch": 1.0341296928327646, "grad_norm": 1.5120577926077867, "learning_rate": 1.1226961126390001e-06, "loss": 0.0402, "step": 4545 }, { "epoch": 1.0343572241183163, "grad_norm": 1.0258672333609278, "learning_rate": 1.1226420603865898e-06, "loss": 0.0266, "step": 4546 }, { "epoch": 1.034584755403868, "grad_norm": 0.6935436273674827, "learning_rate": 1.122587997963287e-06, "loss": 0.021, "step": 4547 }, { "epoch": 1.0348122866894198, "grad_norm": 2.2382185926621383, "learning_rate": 1.122533925370197e-06, "loss": 0.0311, "step": 4548 }, { "epoch": 1.0350398179749716, "grad_norm": 0.8597189295835992, "learning_rate": 1.1224798426084246e-06, "loss": 0.0251, "step": 4549 }, { "epoch": 1.0352673492605233, "grad_norm": 1.2187642427431231, "learning_rate": 1.1224257496790756e-06, "loss": 0.0307, "step": 4550 }, { "epoch": 1.035494880546075, "grad_norm": 0.8235532448084206, "learning_rate": 1.122371646583255e-06, "loss": 0.016, "step": 4551 }, { "epoch": 1.0357224118316268, "grad_norm": 1.1637372096776504, "learning_rate": 1.1223175333220688e-06, "loss": 0.0249, "step": 4552 }, { "epoch": 1.0359499431171786, "grad_norm": 1.2013323108837044, "learning_rate": 1.122263409896623e-06, "loss": 0.0211, "step": 4553 }, { "epoch": 1.0361774744027303, "grad_norm": 1.5241558913138389, "learning_rate": 1.1222092763080242e-06, "loss": 0.0614, "step": 4554 }, { "epoch": 1.036405005688282, "grad_norm": 0.9609458503216893, "learning_rate": 1.1221551325573779e-06, "loss": 0.0345, "step": 4555 }, { "epoch": 1.0366325369738338, "grad_norm": 1.067326425845944, "learning_rate": 1.1221009786457914e-06, "loss": 0.0251, "step": 4556 }, { "epoch": 1.0368600682593856, "grad_norm": 0.6952960225267327, "learning_rate": 1.1220468145743713e-06, "loss": 0.0158, "step": 4557 }, { "epoch": 1.0370875995449373, "grad_norm": 1.5415742411379196, "learning_rate": 1.1219926403442247e-06, "loss": 0.0398, "step": 4558 }, { "epoch": 1.037315130830489, "grad_norm": 1.0675139745280051, "learning_rate": 1.1219384559564587e-06, "loss": 0.0277, "step": 4559 }, { "epoch": 1.0375426621160408, "grad_norm": 0.7777707864923777, "learning_rate": 1.1218842614121806e-06, "loss": 0.0247, "step": 4560 }, { "epoch": 1.0377701934015928, "grad_norm": 1.6672178552337855, "learning_rate": 1.1218300567124983e-06, "loss": 0.0342, "step": 4561 }, { "epoch": 1.0379977246871446, "grad_norm": 1.0152141504441292, "learning_rate": 1.1217758418585195e-06, "loss": 0.0426, "step": 4562 }, { "epoch": 1.0382252559726963, "grad_norm": 1.0296746623158612, "learning_rate": 1.1217216168513522e-06, "loss": 0.0225, "step": 4563 }, { "epoch": 1.038452787258248, "grad_norm": 1.970275220352162, "learning_rate": 1.1216673816921048e-06, "loss": 0.0876, "step": 4564 }, { "epoch": 1.0386803185437998, "grad_norm": 1.1493211366430227, "learning_rate": 1.1216131363818859e-06, "loss": 0.0191, "step": 4565 }, { "epoch": 1.0389078498293516, "grad_norm": 0.9951641567260908, "learning_rate": 1.1215588809218038e-06, "loss": 0.013, "step": 4566 }, { "epoch": 1.0391353811149033, "grad_norm": 0.8580754292791186, "learning_rate": 1.1215046153129678e-06, "loss": 0.0183, "step": 4567 }, { "epoch": 1.039362912400455, "grad_norm": 1.4552256935681152, "learning_rate": 1.1214503395564866e-06, "loss": 0.0245, "step": 4568 }, { "epoch": 1.0395904436860068, "grad_norm": 0.6094461237074116, "learning_rate": 1.1213960536534698e-06, "loss": 0.0088, "step": 4569 }, { "epoch": 1.0398179749715586, "grad_norm": 1.6872520162524116, "learning_rate": 1.1213417576050267e-06, "loss": 0.0465, "step": 4570 }, { "epoch": 1.0400455062571103, "grad_norm": 0.8147199481078186, "learning_rate": 1.1212874514122669e-06, "loss": 0.0188, "step": 4571 }, { "epoch": 1.040273037542662, "grad_norm": 1.6710841073159661, "learning_rate": 1.1212331350763007e-06, "loss": 0.0183, "step": 4572 }, { "epoch": 1.0405005688282138, "grad_norm": 0.7521582291067465, "learning_rate": 1.1211788085982381e-06, "loss": 0.0141, "step": 4573 }, { "epoch": 1.0407281001137656, "grad_norm": 1.3851917572634034, "learning_rate": 1.1211244719791892e-06, "loss": 0.0306, "step": 4574 }, { "epoch": 1.0409556313993173, "grad_norm": 0.5793785303293703, "learning_rate": 1.1210701252202647e-06, "loss": 0.0106, "step": 4575 }, { "epoch": 1.041183162684869, "grad_norm": 1.0136644099615377, "learning_rate": 1.1210157683225753e-06, "loss": 0.0165, "step": 4576 }, { "epoch": 1.0414106939704209, "grad_norm": 0.45591615471037517, "learning_rate": 1.1209614012872323e-06, "loss": 0.0084, "step": 4577 }, { "epoch": 1.0416382252559726, "grad_norm": 1.3441880570081561, "learning_rate": 1.1209070241153462e-06, "loss": 0.031, "step": 4578 }, { "epoch": 1.0418657565415244, "grad_norm": 1.1464751988060768, "learning_rate": 1.1208526368080288e-06, "loss": 0.0332, "step": 4579 }, { "epoch": 1.0420932878270763, "grad_norm": 0.7884324251000338, "learning_rate": 1.120798239366392e-06, "loss": 0.0168, "step": 4580 }, { "epoch": 1.042320819112628, "grad_norm": 2.046969798862839, "learning_rate": 1.1207438317915468e-06, "loss": 0.0229, "step": 4581 }, { "epoch": 1.0425483503981798, "grad_norm": 0.6622639696340208, "learning_rate": 1.1206894140846055e-06, "loss": 0.0087, "step": 4582 }, { "epoch": 1.0427758816837316, "grad_norm": 0.7403190930069853, "learning_rate": 1.1206349862466807e-06, "loss": 0.015, "step": 4583 }, { "epoch": 1.0430034129692833, "grad_norm": 2.3626414035054175, "learning_rate": 1.1205805482788846e-06, "loss": 0.0417, "step": 4584 }, { "epoch": 1.043230944254835, "grad_norm": 2.471483368904896, "learning_rate": 1.1205261001823293e-06, "loss": 0.044, "step": 4585 }, { "epoch": 1.0434584755403868, "grad_norm": 2.416437888927223, "learning_rate": 1.1204716419581281e-06, "loss": 0.0098, "step": 4586 }, { "epoch": 1.0436860068259386, "grad_norm": 1.467076361857702, "learning_rate": 1.1204171736073942e-06, "loss": 0.0134, "step": 4587 }, { "epoch": 1.0439135381114903, "grad_norm": 1.1268621078307177, "learning_rate": 1.1203626951312405e-06, "loss": 0.0349, "step": 4588 }, { "epoch": 1.044141069397042, "grad_norm": 0.7695063155900477, "learning_rate": 1.1203082065307805e-06, "loss": 0.012, "step": 4589 }, { "epoch": 1.0443686006825939, "grad_norm": 1.446038392903303, "learning_rate": 1.1202537078071277e-06, "loss": 0.0547, "step": 4590 }, { "epoch": 1.0445961319681456, "grad_norm": 0.8969845928910839, "learning_rate": 1.1201991989613963e-06, "loss": 0.0111, "step": 4591 }, { "epoch": 1.0448236632536974, "grad_norm": 0.8115056392147382, "learning_rate": 1.1201446799947003e-06, "loss": 0.0092, "step": 4592 }, { "epoch": 1.045051194539249, "grad_norm": 1.2473917829189507, "learning_rate": 1.1200901509081537e-06, "loss": 0.0137, "step": 4593 }, { "epoch": 1.0452787258248009, "grad_norm": 1.0896608769495246, "learning_rate": 1.120035611702871e-06, "loss": 0.0209, "step": 4594 }, { "epoch": 1.0455062571103526, "grad_norm": 0.7977339624737255, "learning_rate": 1.1199810623799673e-06, "loss": 0.0176, "step": 4595 }, { "epoch": 1.0457337883959044, "grad_norm": 1.1634422015699433, "learning_rate": 1.119926502940557e-06, "loss": 0.0188, "step": 4596 }, { "epoch": 1.0459613196814561, "grad_norm": 1.092858194939433, "learning_rate": 1.1198719333857555e-06, "loss": 0.0312, "step": 4597 }, { "epoch": 1.0461888509670079, "grad_norm": 0.836130183985909, "learning_rate": 1.119817353716678e-06, "loss": 0.012, "step": 4598 }, { "epoch": 1.0464163822525596, "grad_norm": 0.767678738883129, "learning_rate": 1.11976276393444e-06, "loss": 0.0114, "step": 4599 }, { "epoch": 1.0466439135381116, "grad_norm": 1.1278576378072667, "learning_rate": 1.1197081640401572e-06, "loss": 0.0245, "step": 4600 }, { "epoch": 1.0468714448236633, "grad_norm": 1.3358445644446304, "learning_rate": 1.1196535540349453e-06, "loss": 0.0414, "step": 4601 }, { "epoch": 1.047098976109215, "grad_norm": 2.748783161026122, "learning_rate": 1.119598933919921e-06, "loss": 0.0374, "step": 4602 }, { "epoch": 1.0473265073947668, "grad_norm": 0.7346665523199191, "learning_rate": 1.1195443036962002e-06, "loss": 0.0137, "step": 4603 }, { "epoch": 1.0475540386803186, "grad_norm": 0.7926661565300546, "learning_rate": 1.1194896633648996e-06, "loss": 0.0234, "step": 4604 }, { "epoch": 1.0477815699658704, "grad_norm": 1.2266234047001268, "learning_rate": 1.1194350129271358e-06, "loss": 0.0218, "step": 4605 }, { "epoch": 1.048009101251422, "grad_norm": 1.1781905579948284, "learning_rate": 1.119380352384026e-06, "loss": 0.0298, "step": 4606 }, { "epoch": 1.0482366325369739, "grad_norm": 0.8103390645469172, "learning_rate": 1.1193256817366871e-06, "loss": 0.014, "step": 4607 }, { "epoch": 1.0484641638225256, "grad_norm": 1.2189569921919905, "learning_rate": 1.1192710009862365e-06, "loss": 0.0252, "step": 4608 }, { "epoch": 1.0486916951080774, "grad_norm": 0.6581776967473145, "learning_rate": 1.1192163101337921e-06, "loss": 0.0129, "step": 4609 }, { "epoch": 1.0489192263936291, "grad_norm": 0.7853755589384823, "learning_rate": 1.1191616091804712e-06, "loss": 0.0346, "step": 4610 }, { "epoch": 1.0491467576791809, "grad_norm": 0.48263244857420745, "learning_rate": 1.1191068981273919e-06, "loss": 0.0113, "step": 4611 }, { "epoch": 1.0493742889647326, "grad_norm": 0.9545125321950874, "learning_rate": 1.1190521769756729e-06, "loss": 0.0254, "step": 4612 }, { "epoch": 1.0496018202502844, "grad_norm": 1.247134305539268, "learning_rate": 1.118997445726432e-06, "loss": 0.0413, "step": 4613 }, { "epoch": 1.0498293515358361, "grad_norm": 1.1502298928735868, "learning_rate": 1.118942704380788e-06, "loss": 0.0137, "step": 4614 }, { "epoch": 1.0500568828213879, "grad_norm": 1.6594682241469216, "learning_rate": 1.11888795293986e-06, "loss": 0.0326, "step": 4615 }, { "epoch": 1.0502844141069396, "grad_norm": 0.7549275144222661, "learning_rate": 1.1188331914047666e-06, "loss": 0.0138, "step": 4616 }, { "epoch": 1.0505119453924914, "grad_norm": 0.8321159003301328, "learning_rate": 1.1187784197766269e-06, "loss": 0.0184, "step": 4617 }, { "epoch": 1.0507394766780431, "grad_norm": 1.209949785555879, "learning_rate": 1.1187236380565608e-06, "loss": 0.0251, "step": 4618 }, { "epoch": 1.050967007963595, "grad_norm": 0.7955918295120653, "learning_rate": 1.1186688462456879e-06, "loss": 0.0135, "step": 4619 }, { "epoch": 1.0511945392491469, "grad_norm": 1.1615893854076145, "learning_rate": 1.118614044345128e-06, "loss": 0.0282, "step": 4620 }, { "epoch": 1.0514220705346986, "grad_norm": 1.273989641669821, "learning_rate": 1.1185592323560006e-06, "loss": 0.038, "step": 4621 }, { "epoch": 1.0516496018202504, "grad_norm": 1.0165071618076773, "learning_rate": 1.1185044102794267e-06, "loss": 0.0328, "step": 4622 }, { "epoch": 1.051877133105802, "grad_norm": 1.1328617777044239, "learning_rate": 1.1184495781165263e-06, "loss": 0.0274, "step": 4623 }, { "epoch": 1.0521046643913539, "grad_norm": 1.1825756024120404, "learning_rate": 1.1183947358684203e-06, "loss": 0.0276, "step": 4624 }, { "epoch": 1.0523321956769056, "grad_norm": 0.517945392477519, "learning_rate": 1.1183398835362298e-06, "loss": 0.0174, "step": 4625 }, { "epoch": 1.0525597269624574, "grad_norm": 0.5843868560216139, "learning_rate": 1.1182850211210752e-06, "loss": 0.0161, "step": 4626 }, { "epoch": 1.0527872582480091, "grad_norm": 1.2595121663614455, "learning_rate": 1.1182301486240782e-06, "loss": 0.0359, "step": 4627 }, { "epoch": 1.0530147895335609, "grad_norm": 1.0715185413956108, "learning_rate": 1.1181752660463604e-06, "loss": 0.0214, "step": 4628 }, { "epoch": 1.0532423208191126, "grad_norm": 0.7795449522978944, "learning_rate": 1.1181203733890433e-06, "loss": 0.0177, "step": 4629 }, { "epoch": 1.0534698521046644, "grad_norm": 0.9800972489749913, "learning_rate": 1.118065470653249e-06, "loss": 0.0347, "step": 4630 }, { "epoch": 1.0536973833902161, "grad_norm": 1.0970470234975969, "learning_rate": 1.1180105578400993e-06, "loss": 0.0433, "step": 4631 }, { "epoch": 1.0539249146757679, "grad_norm": 1.0800910071281984, "learning_rate": 1.117955634950717e-06, "loss": 0.0227, "step": 4632 }, { "epoch": 1.0541524459613196, "grad_norm": 1.3530664066926406, "learning_rate": 1.117900701986224e-06, "loss": 0.0343, "step": 4633 }, { "epoch": 1.0543799772468714, "grad_norm": 1.4133875078213476, "learning_rate": 1.1178457589477434e-06, "loss": 0.027, "step": 4634 }, { "epoch": 1.0546075085324231, "grad_norm": 0.8440744418991889, "learning_rate": 1.1177908058363984e-06, "loss": 0.0098, "step": 4635 }, { "epoch": 1.0548350398179749, "grad_norm": 1.092273870951362, "learning_rate": 1.1177358426533115e-06, "loss": 0.02, "step": 4636 }, { "epoch": 1.0550625711035266, "grad_norm": 0.9187729071178374, "learning_rate": 1.1176808693996067e-06, "loss": 0.0183, "step": 4637 }, { "epoch": 1.0552901023890784, "grad_norm": 1.1090047987424088, "learning_rate": 1.117625886076407e-06, "loss": 0.0193, "step": 4638 }, { "epoch": 1.0555176336746304, "grad_norm": 1.0008875695337944, "learning_rate": 1.1175708926848363e-06, "loss": 0.0189, "step": 4639 }, { "epoch": 1.0557451649601821, "grad_norm": 2.724829300569884, "learning_rate": 1.1175158892260187e-06, "loss": 0.0362, "step": 4640 }, { "epoch": 1.0559726962457339, "grad_norm": 1.0478197516778056, "learning_rate": 1.1174608757010785e-06, "loss": 0.0152, "step": 4641 }, { "epoch": 1.0562002275312856, "grad_norm": 1.590272661414187, "learning_rate": 1.1174058521111398e-06, "loss": 0.0436, "step": 4642 }, { "epoch": 1.0564277588168374, "grad_norm": 0.8472430737385992, "learning_rate": 1.1173508184573273e-06, "loss": 0.0138, "step": 4643 }, { "epoch": 1.0566552901023891, "grad_norm": 0.8371060858404447, "learning_rate": 1.1172957747407657e-06, "loss": 0.0157, "step": 4644 }, { "epoch": 1.0568828213879409, "grad_norm": 0.9592085999503343, "learning_rate": 1.11724072096258e-06, "loss": 0.0323, "step": 4645 }, { "epoch": 1.0571103526734926, "grad_norm": 1.0524749179147639, "learning_rate": 1.1171856571238958e-06, "loss": 0.0296, "step": 4646 }, { "epoch": 1.0573378839590444, "grad_norm": 0.8834933966797389, "learning_rate": 1.1171305832258378e-06, "loss": 0.0218, "step": 4647 }, { "epoch": 1.0575654152445961, "grad_norm": 0.7743496062659715, "learning_rate": 1.117075499269532e-06, "loss": 0.0183, "step": 4648 }, { "epoch": 1.0577929465301479, "grad_norm": 1.0385592346274115, "learning_rate": 1.1170204052561045e-06, "loss": 0.0205, "step": 4649 }, { "epoch": 1.0580204778156996, "grad_norm": 0.905517805098956, "learning_rate": 1.1169653011866806e-06, "loss": 0.0274, "step": 4650 }, { "epoch": 1.0582480091012514, "grad_norm": 0.7908819622644908, "learning_rate": 1.1169101870623872e-06, "loss": 0.0137, "step": 4651 }, { "epoch": 1.0584755403868031, "grad_norm": 1.061810097297376, "learning_rate": 1.1168550628843506e-06, "loss": 0.0206, "step": 4652 }, { "epoch": 1.058703071672355, "grad_norm": 1.6341823289625474, "learning_rate": 1.116799928653697e-06, "loss": 0.0258, "step": 4653 }, { "epoch": 1.0589306029579066, "grad_norm": 1.4294137058728704, "learning_rate": 1.1167447843715536e-06, "loss": 0.0415, "step": 4654 }, { "epoch": 1.0591581342434584, "grad_norm": 1.9352681748212788, "learning_rate": 1.1166896300390475e-06, "loss": 0.0398, "step": 4655 }, { "epoch": 1.0593856655290101, "grad_norm": 0.623531295073113, "learning_rate": 1.1166344656573058e-06, "loss": 0.0096, "step": 4656 }, { "epoch": 1.059613196814562, "grad_norm": 0.7942878422407874, "learning_rate": 1.116579291227456e-06, "loss": 0.0246, "step": 4657 }, { "epoch": 1.0598407281001139, "grad_norm": 1.2220975624229167, "learning_rate": 1.1165241067506258e-06, "loss": 0.0361, "step": 4658 }, { "epoch": 1.0600682593856656, "grad_norm": 1.1286970072887417, "learning_rate": 1.116468912227943e-06, "loss": 0.0263, "step": 4659 }, { "epoch": 1.0602957906712174, "grad_norm": 1.030142424103505, "learning_rate": 1.1164137076605359e-06, "loss": 0.0179, "step": 4660 }, { "epoch": 1.0605233219567691, "grad_norm": 1.6790275883871597, "learning_rate": 1.1163584930495323e-06, "loss": 0.0422, "step": 4661 }, { "epoch": 1.0607508532423209, "grad_norm": 1.3814052631606326, "learning_rate": 1.1163032683960612e-06, "loss": 0.0209, "step": 4662 }, { "epoch": 1.0609783845278726, "grad_norm": 1.225549316988825, "learning_rate": 1.116248033701251e-06, "loss": 0.0271, "step": 4663 }, { "epoch": 1.0612059158134244, "grad_norm": 0.7598712432018329, "learning_rate": 1.1161927889662307e-06, "loss": 0.009, "step": 4664 }, { "epoch": 1.0614334470989761, "grad_norm": 0.940084790434426, "learning_rate": 1.1161375341921293e-06, "loss": 0.0248, "step": 4665 }, { "epoch": 1.0616609783845279, "grad_norm": 0.5873146342311836, "learning_rate": 1.1160822693800761e-06, "loss": 0.0078, "step": 4666 }, { "epoch": 1.0618885096700796, "grad_norm": 1.4244470289207833, "learning_rate": 1.116026994531201e-06, "loss": 0.0248, "step": 4667 }, { "epoch": 1.0621160409556314, "grad_norm": 1.0520117462705862, "learning_rate": 1.1159717096466332e-06, "loss": 0.0302, "step": 4668 }, { "epoch": 1.0623435722411831, "grad_norm": 1.062118932659138, "learning_rate": 1.1159164147275026e-06, "loss": 0.022, "step": 4669 }, { "epoch": 1.062571103526735, "grad_norm": 1.1122338472373463, "learning_rate": 1.11586110977494e-06, "loss": 0.0164, "step": 4670 }, { "epoch": 1.0627986348122866, "grad_norm": 0.8000529898204216, "learning_rate": 1.1158057947900749e-06, "loss": 0.0217, "step": 4671 }, { "epoch": 1.0630261660978384, "grad_norm": 0.6219194713042089, "learning_rate": 1.1157504697740384e-06, "loss": 0.0138, "step": 4672 }, { "epoch": 1.0632536973833902, "grad_norm": 1.3392201881028478, "learning_rate": 1.1156951347279612e-06, "loss": 0.028, "step": 4673 }, { "epoch": 1.063481228668942, "grad_norm": 1.7333136164138454, "learning_rate": 1.1156397896529739e-06, "loss": 0.05, "step": 4674 }, { "epoch": 1.0637087599544937, "grad_norm": 1.5262332959624183, "learning_rate": 1.1155844345502079e-06, "loss": 0.0312, "step": 4675 }, { "epoch": 1.0639362912400454, "grad_norm": 0.9932229349703479, "learning_rate": 1.1155290694207946e-06, "loss": 0.0226, "step": 4676 }, { "epoch": 1.0641638225255972, "grad_norm": 1.2628573637308051, "learning_rate": 1.1154736942658655e-06, "loss": 0.0214, "step": 4677 }, { "epoch": 1.0643913538111491, "grad_norm": 1.4330647372008305, "learning_rate": 1.1154183090865523e-06, "loss": 0.0256, "step": 4678 }, { "epoch": 1.0646188850967009, "grad_norm": 1.042200851154514, "learning_rate": 1.1153629138839869e-06, "loss": 0.0293, "step": 4679 }, { "epoch": 1.0648464163822526, "grad_norm": 0.9462778481987262, "learning_rate": 1.115307508659302e-06, "loss": 0.0245, "step": 4680 }, { "epoch": 1.0650739476678044, "grad_norm": 1.6564513386390822, "learning_rate": 1.115252093413629e-06, "loss": 0.0496, "step": 4681 }, { "epoch": 1.0653014789533561, "grad_norm": 1.0868647502168836, "learning_rate": 1.1151966681481013e-06, "loss": 0.0113, "step": 4682 }, { "epoch": 1.065529010238908, "grad_norm": 1.115171642211064, "learning_rate": 1.1151412328638516e-06, "loss": 0.0224, "step": 4683 }, { "epoch": 1.0657565415244596, "grad_norm": 0.5398893154965461, "learning_rate": 1.1150857875620129e-06, "loss": 0.0139, "step": 4684 }, { "epoch": 1.0659840728100114, "grad_norm": 1.0871499987134439, "learning_rate": 1.1150303322437179e-06, "loss": 0.0134, "step": 4685 }, { "epoch": 1.0662116040955631, "grad_norm": 1.1070037508570774, "learning_rate": 1.1149748669101005e-06, "loss": 0.0221, "step": 4686 }, { "epoch": 1.066439135381115, "grad_norm": 0.9425411653474299, "learning_rate": 1.1149193915622942e-06, "loss": 0.0398, "step": 4687 }, { "epoch": 1.0666666666666667, "grad_norm": 1.1772890828333267, "learning_rate": 1.1148639062014325e-06, "loss": 0.0211, "step": 4688 }, { "epoch": 1.0668941979522184, "grad_norm": 1.2557238706849523, "learning_rate": 1.11480841082865e-06, "loss": 0.0301, "step": 4689 }, { "epoch": 1.0671217292377702, "grad_norm": 0.7994780640965474, "learning_rate": 1.1147529054450805e-06, "loss": 0.0201, "step": 4690 }, { "epoch": 1.067349260523322, "grad_norm": 1.1399734758992979, "learning_rate": 1.1146973900518587e-06, "loss": 0.0293, "step": 4691 }, { "epoch": 1.0675767918088737, "grad_norm": 1.3046065395944408, "learning_rate": 1.1146418646501189e-06, "loss": 0.0286, "step": 4692 }, { "epoch": 1.0678043230944254, "grad_norm": 0.8748609055579856, "learning_rate": 1.114586329240996e-06, "loss": 0.0211, "step": 4693 }, { "epoch": 1.0680318543799772, "grad_norm": 0.8739676230672317, "learning_rate": 1.1145307838256255e-06, "loss": 0.011, "step": 4694 }, { "epoch": 1.068259385665529, "grad_norm": 1.249091385352741, "learning_rate": 1.1144752284051422e-06, "loss": 0.0301, "step": 4695 }, { "epoch": 1.068486916951081, "grad_norm": 1.562897636481294, "learning_rate": 1.1144196629806817e-06, "loss": 0.0501, "step": 4696 }, { "epoch": 1.0687144482366326, "grad_norm": 0.9143311380884951, "learning_rate": 1.1143640875533795e-06, "loss": 0.0123, "step": 4697 }, { "epoch": 1.0689419795221844, "grad_norm": 1.6200825529791434, "learning_rate": 1.1143085021243717e-06, "loss": 0.0225, "step": 4698 }, { "epoch": 1.0691695108077361, "grad_norm": 1.177872950810007, "learning_rate": 1.1142529066947941e-06, "loss": 0.0346, "step": 4699 }, { "epoch": 1.069397042093288, "grad_norm": 1.108013166594474, "learning_rate": 1.1141973012657834e-06, "loss": 0.0267, "step": 4700 }, { "epoch": 1.0696245733788396, "grad_norm": 1.0112425933651041, "learning_rate": 1.1141416858384753e-06, "loss": 0.0311, "step": 4701 }, { "epoch": 1.0698521046643914, "grad_norm": 1.095144486124691, "learning_rate": 1.1140860604140076e-06, "loss": 0.0286, "step": 4702 }, { "epoch": 1.0700796359499432, "grad_norm": 0.806336687142019, "learning_rate": 1.114030424993516e-06, "loss": 0.0142, "step": 4703 }, { "epoch": 1.070307167235495, "grad_norm": 0.7834952051331864, "learning_rate": 1.1139747795781382e-06, "loss": 0.0101, "step": 4704 }, { "epoch": 1.0705346985210467, "grad_norm": 0.7472560755343798, "learning_rate": 1.1139191241690116e-06, "loss": 0.0139, "step": 4705 }, { "epoch": 1.0707622298065984, "grad_norm": 0.5618125746941464, "learning_rate": 1.1138634587672734e-06, "loss": 0.0075, "step": 4706 }, { "epoch": 1.0709897610921502, "grad_norm": 0.7202976835755431, "learning_rate": 1.1138077833740616e-06, "loss": 0.0207, "step": 4707 }, { "epoch": 1.071217292377702, "grad_norm": 0.7307887063589182, "learning_rate": 1.1137520979905138e-06, "loss": 0.0223, "step": 4708 }, { "epoch": 1.0714448236632537, "grad_norm": 1.1867538184942938, "learning_rate": 1.1136964026177683e-06, "loss": 0.0288, "step": 4709 }, { "epoch": 1.0716723549488054, "grad_norm": 1.2840239837956307, "learning_rate": 1.113640697256963e-06, "loss": 0.0184, "step": 4710 }, { "epoch": 1.0718998862343572, "grad_norm": 1.1345754398737768, "learning_rate": 1.113584981909237e-06, "loss": 0.0238, "step": 4711 }, { "epoch": 1.072127417519909, "grad_norm": 0.9685288187674116, "learning_rate": 1.1135292565757288e-06, "loss": 0.0227, "step": 4712 }, { "epoch": 1.0723549488054607, "grad_norm": 0.7081273094248537, "learning_rate": 1.1134735212575772e-06, "loss": 0.0102, "step": 4713 }, { "epoch": 1.0725824800910124, "grad_norm": 0.9192931039025396, "learning_rate": 1.1134177759559216e-06, "loss": 0.0244, "step": 4714 }, { "epoch": 1.0728100113765642, "grad_norm": 0.9284668402003036, "learning_rate": 1.1133620206719011e-06, "loss": 0.0243, "step": 4715 }, { "epoch": 1.073037542662116, "grad_norm": 0.9115758932784177, "learning_rate": 1.1133062554066551e-06, "loss": 0.0205, "step": 4716 }, { "epoch": 1.073265073947668, "grad_norm": 0.8734771500689966, "learning_rate": 1.1132504801613237e-06, "loss": 0.0213, "step": 4717 }, { "epoch": 1.0734926052332197, "grad_norm": 1.126626337283734, "learning_rate": 1.1131946949370467e-06, "loss": 0.0327, "step": 4718 }, { "epoch": 1.0737201365187714, "grad_norm": 0.65003673022336, "learning_rate": 1.113138899734964e-06, "loss": 0.0162, "step": 4719 }, { "epoch": 1.0739476678043232, "grad_norm": 1.6524784863925304, "learning_rate": 1.1130830945562165e-06, "loss": 0.046, "step": 4720 }, { "epoch": 1.074175199089875, "grad_norm": 0.9704754115890377, "learning_rate": 1.1130272794019442e-06, "loss": 0.0165, "step": 4721 }, { "epoch": 1.0744027303754267, "grad_norm": 0.841189725514391, "learning_rate": 1.1129714542732882e-06, "loss": 0.0209, "step": 4722 }, { "epoch": 1.0746302616609784, "grad_norm": 0.6874778292132357, "learning_rate": 1.1129156191713893e-06, "loss": 0.0121, "step": 4723 }, { "epoch": 1.0748577929465302, "grad_norm": 0.9632218181611477, "learning_rate": 1.1128597740973886e-06, "loss": 0.0388, "step": 4724 }, { "epoch": 1.075085324232082, "grad_norm": 0.7703489214733116, "learning_rate": 1.1128039190524278e-06, "loss": 0.0059, "step": 4725 }, { "epoch": 1.0753128555176337, "grad_norm": 1.9572066317541847, "learning_rate": 1.112748054037648e-06, "loss": 0.0262, "step": 4726 }, { "epoch": 1.0755403868031854, "grad_norm": 1.0824319203641397, "learning_rate": 1.1126921790541915e-06, "loss": 0.0149, "step": 4727 }, { "epoch": 1.0757679180887372, "grad_norm": 0.670440042251469, "learning_rate": 1.1126362941032e-06, "loss": 0.0063, "step": 4728 }, { "epoch": 1.075995449374289, "grad_norm": 0.7800421077003349, "learning_rate": 1.1125803991858156e-06, "loss": 0.0277, "step": 4729 }, { "epoch": 1.0762229806598407, "grad_norm": 1.106352016812619, "learning_rate": 1.1125244943031809e-06, "loss": 0.0123, "step": 4730 }, { "epoch": 1.0764505119453924, "grad_norm": 1.1161991172358912, "learning_rate": 1.1124685794564383e-06, "loss": 0.026, "step": 4731 }, { "epoch": 1.0766780432309442, "grad_norm": 1.7881963402264416, "learning_rate": 1.1124126546467307e-06, "loss": 0.0385, "step": 4732 }, { "epoch": 1.076905574516496, "grad_norm": 1.2868577345652183, "learning_rate": 1.1123567198752012e-06, "loss": 0.0273, "step": 4733 }, { "epoch": 1.0771331058020477, "grad_norm": 2.0351098504278085, "learning_rate": 1.1123007751429928e-06, "loss": 0.0568, "step": 4734 }, { "epoch": 1.0773606370875997, "grad_norm": 0.9002868646294088, "learning_rate": 1.1122448204512493e-06, "loss": 0.0253, "step": 4735 }, { "epoch": 1.0775881683731514, "grad_norm": 0.9063013425685674, "learning_rate": 1.1121888558011136e-06, "loss": 0.0117, "step": 4736 }, { "epoch": 1.0778156996587032, "grad_norm": 0.5859089097561659, "learning_rate": 1.11213288119373e-06, "loss": 0.0074, "step": 4737 }, { "epoch": 1.078043230944255, "grad_norm": 1.3479752927519844, "learning_rate": 1.1120768966302424e-06, "loss": 0.0249, "step": 4738 }, { "epoch": 1.0782707622298067, "grad_norm": 0.7352263154143484, "learning_rate": 1.1120209021117953e-06, "loss": 0.0173, "step": 4739 }, { "epoch": 1.0784982935153584, "grad_norm": 0.7374112860235839, "learning_rate": 1.111964897639533e-06, "loss": 0.0174, "step": 4740 }, { "epoch": 1.0787258248009102, "grad_norm": 1.287591128879902, "learning_rate": 1.1119088832145999e-06, "loss": 0.0326, "step": 4741 }, { "epoch": 1.078953356086462, "grad_norm": 1.0787165555211229, "learning_rate": 1.1118528588381408e-06, "loss": 0.0192, "step": 4742 }, { "epoch": 1.0791808873720137, "grad_norm": 1.0815099906932875, "learning_rate": 1.111796824511301e-06, "loss": 0.0216, "step": 4743 }, { "epoch": 1.0794084186575654, "grad_norm": 0.8664538220627443, "learning_rate": 1.1117407802352257e-06, "loss": 0.0127, "step": 4744 }, { "epoch": 1.0796359499431172, "grad_norm": 1.33972724465675, "learning_rate": 1.11168472601106e-06, "loss": 0.0335, "step": 4745 }, { "epoch": 1.079863481228669, "grad_norm": 1.4666132789919955, "learning_rate": 1.1116286618399502e-06, "loss": 0.0457, "step": 4746 }, { "epoch": 1.0800910125142207, "grad_norm": 0.6790281378613789, "learning_rate": 1.1115725877230416e-06, "loss": 0.0102, "step": 4747 }, { "epoch": 1.0803185437997724, "grad_norm": 0.9620611501361773, "learning_rate": 1.1115165036614803e-06, "loss": 0.0261, "step": 4748 }, { "epoch": 1.0805460750853242, "grad_norm": 1.1237272287467874, "learning_rate": 1.1114604096564128e-06, "loss": 0.0316, "step": 4749 }, { "epoch": 1.080773606370876, "grad_norm": 1.0088779610949172, "learning_rate": 1.1114043057089855e-06, "loss": 0.0225, "step": 4750 }, { "epoch": 1.0810011376564277, "grad_norm": 1.497609697625705, "learning_rate": 1.1113481918203447e-06, "loss": 0.0442, "step": 4751 }, { "epoch": 1.0812286689419794, "grad_norm": 0.8914252374944144, "learning_rate": 1.111292067991638e-06, "loss": 0.0165, "step": 4752 }, { "epoch": 1.0814562002275312, "grad_norm": 1.276532043770101, "learning_rate": 1.1112359342240118e-06, "loss": 0.0387, "step": 4753 }, { "epoch": 1.081683731513083, "grad_norm": 0.9077763034087386, "learning_rate": 1.1111797905186137e-06, "loss": 0.0141, "step": 4754 }, { "epoch": 1.0819112627986347, "grad_norm": 1.6466072809663417, "learning_rate": 1.111123636876591e-06, "loss": 0.0349, "step": 4755 }, { "epoch": 1.0821387940841867, "grad_norm": 0.9430919880459814, "learning_rate": 1.1110674732990915e-06, "loss": 0.016, "step": 4756 }, { "epoch": 1.0823663253697384, "grad_norm": 1.6404185778699842, "learning_rate": 1.1110112997872627e-06, "loss": 0.0309, "step": 4757 }, { "epoch": 1.0825938566552902, "grad_norm": 0.9479110834953683, "learning_rate": 1.1109551163422535e-06, "loss": 0.0189, "step": 4758 }, { "epoch": 1.082821387940842, "grad_norm": 0.9870118719493604, "learning_rate": 1.1108989229652115e-06, "loss": 0.0221, "step": 4759 }, { "epoch": 1.0830489192263937, "grad_norm": 1.7588111531610175, "learning_rate": 1.1108427196572854e-06, "loss": 0.0327, "step": 4760 }, { "epoch": 1.0832764505119454, "grad_norm": 0.9021346736655393, "learning_rate": 1.110786506419624e-06, "loss": 0.036, "step": 4761 }, { "epoch": 1.0835039817974972, "grad_norm": 0.8266880922651803, "learning_rate": 1.110730283253376e-06, "loss": 0.0279, "step": 4762 }, { "epoch": 1.083731513083049, "grad_norm": 0.8011724413374426, "learning_rate": 1.1106740501596904e-06, "loss": 0.0195, "step": 4763 }, { "epoch": 1.0839590443686007, "grad_norm": 0.9073688021992705, "learning_rate": 1.110617807139717e-06, "loss": 0.023, "step": 4764 }, { "epoch": 1.0841865756541524, "grad_norm": 1.022169451105952, "learning_rate": 1.1105615541946049e-06, "loss": 0.0209, "step": 4765 }, { "epoch": 1.0844141069397042, "grad_norm": 0.43405837742396564, "learning_rate": 1.1105052913255038e-06, "loss": 0.0044, "step": 4766 }, { "epoch": 1.084641638225256, "grad_norm": 1.383667776402452, "learning_rate": 1.1104490185335638e-06, "loss": 0.0287, "step": 4767 }, { "epoch": 1.0848691695108077, "grad_norm": 0.9054919782471123, "learning_rate": 1.1103927358199349e-06, "loss": 0.0191, "step": 4768 }, { "epoch": 1.0850967007963594, "grad_norm": 1.2358115777885788, "learning_rate": 1.1103364431857672e-06, "loss": 0.0311, "step": 4769 }, { "epoch": 1.0853242320819112, "grad_norm": 0.7048379529354732, "learning_rate": 1.1102801406322118e-06, "loss": 0.0104, "step": 4770 }, { "epoch": 1.085551763367463, "grad_norm": 1.1457330667513694, "learning_rate": 1.110223828160419e-06, "loss": 0.0212, "step": 4771 }, { "epoch": 1.0857792946530147, "grad_norm": 0.7058751162839924, "learning_rate": 1.1101675057715396e-06, "loss": 0.0092, "step": 4772 }, { "epoch": 1.0860068259385665, "grad_norm": 1.1868874800053821, "learning_rate": 1.110111173466725e-06, "loss": 0.0416, "step": 4773 }, { "epoch": 1.0862343572241184, "grad_norm": 1.1919649645556245, "learning_rate": 1.1100548312471266e-06, "loss": 0.0287, "step": 4774 }, { "epoch": 1.0864618885096702, "grad_norm": 1.0824756647963032, "learning_rate": 1.1099984791138957e-06, "loss": 0.0152, "step": 4775 }, { "epoch": 1.086689419795222, "grad_norm": 1.4691023279738935, "learning_rate": 1.109942117068184e-06, "loss": 0.0201, "step": 4776 }, { "epoch": 1.0869169510807737, "grad_norm": 0.8472040472970495, "learning_rate": 1.1098857451111437e-06, "loss": 0.0142, "step": 4777 }, { "epoch": 1.0871444823663254, "grad_norm": 0.8480336544082094, "learning_rate": 1.1098293632439267e-06, "loss": 0.009, "step": 4778 }, { "epoch": 1.0873720136518772, "grad_norm": 1.1145190027001157, "learning_rate": 1.1097729714676855e-06, "loss": 0.0143, "step": 4779 }, { "epoch": 1.087599544937429, "grad_norm": 1.04224113160413, "learning_rate": 1.1097165697835726e-06, "loss": 0.024, "step": 4780 }, { "epoch": 1.0878270762229807, "grad_norm": 1.0620838647799002, "learning_rate": 1.1096601581927407e-06, "loss": 0.0178, "step": 4781 }, { "epoch": 1.0880546075085324, "grad_norm": 4.011373250629747, "learning_rate": 1.109603736696343e-06, "loss": 0.1504, "step": 4782 }, { "epoch": 1.0882821387940842, "grad_norm": 1.5174512485770184, "learning_rate": 1.1095473052955322e-06, "loss": 0.0405, "step": 4783 }, { "epoch": 1.088509670079636, "grad_norm": 0.884113509373173, "learning_rate": 1.1094908639914617e-06, "loss": 0.0224, "step": 4784 }, { "epoch": 1.0887372013651877, "grad_norm": 1.3619635032370507, "learning_rate": 1.1094344127852855e-06, "loss": 0.037, "step": 4785 }, { "epoch": 1.0889647326507395, "grad_norm": 1.9047416105199788, "learning_rate": 1.1093779516781571e-06, "loss": 0.0361, "step": 4786 }, { "epoch": 1.0891922639362912, "grad_norm": 1.1728707520462738, "learning_rate": 1.1093214806712305e-06, "loss": 0.0389, "step": 4787 }, { "epoch": 1.089419795221843, "grad_norm": 1.3254160466529121, "learning_rate": 1.1092649997656597e-06, "loss": 0.0232, "step": 4788 }, { "epoch": 1.0896473265073947, "grad_norm": 1.0725037617108157, "learning_rate": 1.1092085089625992e-06, "loss": 0.0164, "step": 4789 }, { "epoch": 1.0898748577929465, "grad_norm": 0.9758736007788956, "learning_rate": 1.1091520082632037e-06, "loss": 0.0218, "step": 4790 }, { "epoch": 1.0901023890784982, "grad_norm": 0.8477305521502946, "learning_rate": 1.1090954976686277e-06, "loss": 0.0245, "step": 4791 }, { "epoch": 1.09032992036405, "grad_norm": 0.9962558192273867, "learning_rate": 1.1090389771800264e-06, "loss": 0.0287, "step": 4792 }, { "epoch": 1.0905574516496017, "grad_norm": 1.2473669308596411, "learning_rate": 1.1089824467985549e-06, "loss": 0.0229, "step": 4793 }, { "epoch": 1.0907849829351535, "grad_norm": 2.510749402051243, "learning_rate": 1.1089259065253684e-06, "loss": 0.0259, "step": 4794 }, { "epoch": 1.0910125142207054, "grad_norm": 1.149992705234752, "learning_rate": 1.1088693563616226e-06, "loss": 0.0247, "step": 4795 }, { "epoch": 1.0912400455062572, "grad_norm": 0.5940460865164179, "learning_rate": 1.1088127963084736e-06, "loss": 0.0171, "step": 4796 }, { "epoch": 1.091467576791809, "grad_norm": 0.9164786766368149, "learning_rate": 1.108756226367077e-06, "loss": 0.0317, "step": 4797 }, { "epoch": 1.0916951080773607, "grad_norm": 1.1244091740172437, "learning_rate": 1.108699646538589e-06, "loss": 0.0255, "step": 4798 }, { "epoch": 1.0919226393629125, "grad_norm": 1.437903004135215, "learning_rate": 1.108643056824166e-06, "loss": 0.0464, "step": 4799 }, { "epoch": 1.0921501706484642, "grad_norm": 0.9254577109068766, "learning_rate": 1.108586457224965e-06, "loss": 0.0194, "step": 4800 }, { "epoch": 1.092377701934016, "grad_norm": 1.1931575439362083, "learning_rate": 1.1085298477421421e-06, "loss": 0.015, "step": 4801 }, { "epoch": 1.0926052332195677, "grad_norm": 0.6083574023703965, "learning_rate": 1.1084732283768548e-06, "loss": 0.0057, "step": 4802 }, { "epoch": 1.0928327645051195, "grad_norm": 0.8597843041320873, "learning_rate": 1.1084165991302601e-06, "loss": 0.0117, "step": 4803 }, { "epoch": 1.0930602957906712, "grad_norm": 1.4390855321709966, "learning_rate": 1.1083599600035155e-06, "loss": 0.0712, "step": 4804 }, { "epoch": 1.093287827076223, "grad_norm": 0.9189557088411734, "learning_rate": 1.1083033109977787e-06, "loss": 0.0248, "step": 4805 }, { "epoch": 1.0935153583617747, "grad_norm": 0.7847663302953185, "learning_rate": 1.1082466521142072e-06, "loss": 0.0138, "step": 4806 }, { "epoch": 1.0937428896473265, "grad_norm": 1.0427794259157794, "learning_rate": 1.1081899833539592e-06, "loss": 0.022, "step": 4807 }, { "epoch": 1.0939704209328782, "grad_norm": 0.7963208995055476, "learning_rate": 1.1081333047181928e-06, "loss": 0.0148, "step": 4808 }, { "epoch": 1.09419795221843, "grad_norm": 1.2861710540407572, "learning_rate": 1.1080766162080664e-06, "loss": 0.044, "step": 4809 }, { "epoch": 1.0944254835039817, "grad_norm": 1.0606108464391806, "learning_rate": 1.1080199178247388e-06, "loss": 0.0227, "step": 4810 }, { "epoch": 1.0946530147895335, "grad_norm": 1.0178256141989168, "learning_rate": 1.1079632095693688e-06, "loss": 0.0217, "step": 4811 }, { "epoch": 1.0948805460750852, "grad_norm": 0.3733490330289167, "learning_rate": 1.107906491443115e-06, "loss": 0.0041, "step": 4812 }, { "epoch": 1.0951080773606372, "grad_norm": 0.970660447630168, "learning_rate": 1.1078497634471373e-06, "loss": 0.0191, "step": 4813 }, { "epoch": 1.095335608646189, "grad_norm": 0.9571317568580685, "learning_rate": 1.1077930255825944e-06, "loss": 0.0146, "step": 4814 }, { "epoch": 1.0955631399317407, "grad_norm": 1.4227698048199817, "learning_rate": 1.1077362778506464e-06, "loss": 0.0259, "step": 4815 }, { "epoch": 1.0957906712172925, "grad_norm": 0.7937432954740787, "learning_rate": 1.107679520252453e-06, "loss": 0.0215, "step": 4816 }, { "epoch": 1.0960182025028442, "grad_norm": 1.2433291885534135, "learning_rate": 1.107622752789174e-06, "loss": 0.0271, "step": 4817 }, { "epoch": 1.096245733788396, "grad_norm": 0.5775769553380794, "learning_rate": 1.10756597546197e-06, "loss": 0.0067, "step": 4818 }, { "epoch": 1.0964732650739477, "grad_norm": 0.43052018257915775, "learning_rate": 1.1075091882720012e-06, "loss": 0.0047, "step": 4819 }, { "epoch": 1.0967007963594995, "grad_norm": 0.6965681550864847, "learning_rate": 1.1074523912204282e-06, "loss": 0.0069, "step": 4820 }, { "epoch": 1.0969283276450512, "grad_norm": 1.950972202735377, "learning_rate": 1.107395584308412e-06, "loss": 0.0384, "step": 4821 }, { "epoch": 1.097155858930603, "grad_norm": 1.2755184486485458, "learning_rate": 1.1073387675371134e-06, "loss": 0.021, "step": 4822 }, { "epoch": 1.0973833902161547, "grad_norm": 0.9195102681671121, "learning_rate": 1.1072819409076937e-06, "loss": 0.0233, "step": 4823 }, { "epoch": 1.0976109215017065, "grad_norm": 1.4467096094976202, "learning_rate": 1.1072251044213146e-06, "loss": 0.063, "step": 4824 }, { "epoch": 1.0978384527872582, "grad_norm": 0.9276103854462384, "learning_rate": 1.1071682580791375e-06, "loss": 0.0186, "step": 4825 }, { "epoch": 1.09806598407281, "grad_norm": 0.6788301923178457, "learning_rate": 1.107111401882324e-06, "loss": 0.0178, "step": 4826 }, { "epoch": 1.0982935153583617, "grad_norm": 0.8922254579510575, "learning_rate": 1.1070545358320367e-06, "loss": 0.0205, "step": 4827 }, { "epoch": 1.0985210466439135, "grad_norm": 1.5932312378336655, "learning_rate": 1.1069976599294374e-06, "loss": 0.0296, "step": 4828 }, { "epoch": 1.0987485779294652, "grad_norm": 1.3474193896215136, "learning_rate": 1.1069407741756884e-06, "loss": 0.0108, "step": 4829 }, { "epoch": 1.098976109215017, "grad_norm": 1.286192731643991, "learning_rate": 1.106883878571953e-06, "loss": 0.0306, "step": 4830 }, { "epoch": 1.0992036405005687, "grad_norm": 1.7859019215489673, "learning_rate": 1.1068269731193936e-06, "loss": 0.0467, "step": 4831 }, { "epoch": 1.0994311717861205, "grad_norm": 1.0673033994327072, "learning_rate": 1.106770057819173e-06, "loss": 0.011, "step": 4832 }, { "epoch": 1.0996587030716722, "grad_norm": 1.1919269451119046, "learning_rate": 1.1067131326724551e-06, "loss": 0.0275, "step": 4833 }, { "epoch": 1.0998862343572242, "grad_norm": 1.0610839759729118, "learning_rate": 1.106656197680403e-06, "loss": 0.0175, "step": 4834 }, { "epoch": 1.100113765642776, "grad_norm": 0.9398630402901391, "learning_rate": 1.10659925284418e-06, "loss": 0.028, "step": 4835 }, { "epoch": 1.1003412969283277, "grad_norm": 1.0905561404271749, "learning_rate": 1.1065422981649506e-06, "loss": 0.0408, "step": 4836 }, { "epoch": 1.1005688282138795, "grad_norm": 1.1760138038912833, "learning_rate": 1.1064853336438782e-06, "loss": 0.0491, "step": 4837 }, { "epoch": 1.1007963594994312, "grad_norm": 0.8244510688827178, "learning_rate": 1.1064283592821276e-06, "loss": 0.0197, "step": 4838 }, { "epoch": 1.101023890784983, "grad_norm": 0.747177088030229, "learning_rate": 1.106371375080863e-06, "loss": 0.017, "step": 4839 }, { "epoch": 1.1012514220705347, "grad_norm": 1.1336438441896446, "learning_rate": 1.106314381041249e-06, "loss": 0.0253, "step": 4840 }, { "epoch": 1.1014789533560865, "grad_norm": 0.5114058401483557, "learning_rate": 1.1062573771644506e-06, "loss": 0.004, "step": 4841 }, { "epoch": 1.1017064846416382, "grad_norm": 0.5746769321004888, "learning_rate": 1.106200363451633e-06, "loss": 0.0094, "step": 4842 }, { "epoch": 1.10193401592719, "grad_norm": 0.7680543679306212, "learning_rate": 1.1061433399039608e-06, "loss": 0.029, "step": 4843 }, { "epoch": 1.1021615472127417, "grad_norm": 0.7493587776816238, "learning_rate": 1.1060863065226002e-06, "loss": 0.0091, "step": 4844 }, { "epoch": 1.1023890784982935, "grad_norm": 1.3476928607388319, "learning_rate": 1.1060292633087167e-06, "loss": 0.0436, "step": 4845 }, { "epoch": 1.1026166097838452, "grad_norm": 1.087918078217182, "learning_rate": 1.1059722102634756e-06, "loss": 0.0201, "step": 4846 }, { "epoch": 1.102844141069397, "grad_norm": 1.0929486099530539, "learning_rate": 1.1059151473880439e-06, "loss": 0.0328, "step": 4847 }, { "epoch": 1.1030716723549487, "grad_norm": 1.3189660985869427, "learning_rate": 1.105858074683587e-06, "loss": 0.0142, "step": 4848 }, { "epoch": 1.1032992036405005, "grad_norm": 1.052151916181327, "learning_rate": 1.1058009921512717e-06, "loss": 0.0195, "step": 4849 }, { "epoch": 1.1035267349260522, "grad_norm": 1.0009913456254373, "learning_rate": 1.1057438997922648e-06, "loss": 0.022, "step": 4850 }, { "epoch": 1.103754266211604, "grad_norm": 0.9615349865705245, "learning_rate": 1.105686797607733e-06, "loss": 0.0111, "step": 4851 }, { "epoch": 1.103981797497156, "grad_norm": 1.3130785181889084, "learning_rate": 1.1056296855988432e-06, "loss": 0.0235, "step": 4852 }, { "epoch": 1.1042093287827077, "grad_norm": 0.8672033139838674, "learning_rate": 1.105572563766763e-06, "loss": 0.0243, "step": 4853 }, { "epoch": 1.1044368600682595, "grad_norm": 1.0127688429987634, "learning_rate": 1.1055154321126597e-06, "loss": 0.0153, "step": 4854 }, { "epoch": 1.1046643913538112, "grad_norm": 1.2643260607838762, "learning_rate": 1.105458290637701e-06, "loss": 0.04, "step": 4855 }, { "epoch": 1.104891922639363, "grad_norm": 0.6193207760495852, "learning_rate": 1.105401139343055e-06, "loss": 0.0095, "step": 4856 }, { "epoch": 1.1051194539249147, "grad_norm": 0.9678581892125716, "learning_rate": 1.105343978229889e-06, "loss": 0.0137, "step": 4857 }, { "epoch": 1.1053469852104665, "grad_norm": 1.2738862816924261, "learning_rate": 1.1052868072993723e-06, "loss": 0.0324, "step": 4858 }, { "epoch": 1.1055745164960182, "grad_norm": 2.111284557158375, "learning_rate": 1.1052296265526726e-06, "loss": 0.0436, "step": 4859 }, { "epoch": 1.10580204778157, "grad_norm": 1.7291659207116172, "learning_rate": 1.105172435990959e-06, "loss": 0.0443, "step": 4860 }, { "epoch": 1.1060295790671217, "grad_norm": 1.4583333294669347, "learning_rate": 1.1051152356154e-06, "loss": 0.0316, "step": 4861 }, { "epoch": 1.1062571103526735, "grad_norm": 0.9218920816273283, "learning_rate": 1.105058025427165e-06, "loss": 0.017, "step": 4862 }, { "epoch": 1.1064846416382252, "grad_norm": 1.2813414812852448, "learning_rate": 1.105000805427423e-06, "loss": 0.0262, "step": 4863 }, { "epoch": 1.106712172923777, "grad_norm": 1.6993734853125684, "learning_rate": 1.1049435756173439e-06, "loss": 0.0219, "step": 4864 }, { "epoch": 1.1069397042093287, "grad_norm": 0.8219715997662334, "learning_rate": 1.104886335998097e-06, "loss": 0.0166, "step": 4865 }, { "epoch": 1.1071672354948805, "grad_norm": 1.4898463042222714, "learning_rate": 1.104829086570852e-06, "loss": 0.0244, "step": 4866 }, { "epoch": 1.1073947667804322, "grad_norm": 0.8450834919239356, "learning_rate": 1.1047718273367794e-06, "loss": 0.0281, "step": 4867 }, { "epoch": 1.107622298065984, "grad_norm": 0.6448103593188148, "learning_rate": 1.1047145582970494e-06, "loss": 0.0086, "step": 4868 }, { "epoch": 1.1078498293515358, "grad_norm": 1.1979257342162584, "learning_rate": 1.1046572794528324e-06, "loss": 0.0295, "step": 4869 }, { "epoch": 1.1080773606370875, "grad_norm": 0.704626187571558, "learning_rate": 1.104599990805299e-06, "loss": 0.0141, "step": 4870 }, { "epoch": 1.1083048919226393, "grad_norm": 1.2920126084899004, "learning_rate": 1.1045426923556198e-06, "loss": 0.0226, "step": 4871 }, { "epoch": 1.108532423208191, "grad_norm": 0.9357951370023873, "learning_rate": 1.1044853841049668e-06, "loss": 0.0146, "step": 4872 }, { "epoch": 1.108759954493743, "grad_norm": 0.8290368347528208, "learning_rate": 1.1044280660545103e-06, "loss": 0.0098, "step": 4873 }, { "epoch": 1.1089874857792947, "grad_norm": 1.067183885716191, "learning_rate": 1.1043707382054223e-06, "loss": 0.0206, "step": 4874 }, { "epoch": 1.1092150170648465, "grad_norm": 0.9196445315136, "learning_rate": 1.1043134005588743e-06, "loss": 0.0196, "step": 4875 }, { "epoch": 1.1094425483503982, "grad_norm": 1.5830957213473278, "learning_rate": 1.1042560531160381e-06, "loss": 0.0426, "step": 4876 }, { "epoch": 1.10967007963595, "grad_norm": 1.0495445338165712, "learning_rate": 1.104198695878086e-06, "loss": 0.0223, "step": 4877 }, { "epoch": 1.1098976109215017, "grad_norm": 0.5938411642446392, "learning_rate": 1.1041413288461903e-06, "loss": 0.0118, "step": 4878 }, { "epoch": 1.1101251422070535, "grad_norm": 1.286777509620217, "learning_rate": 1.1040839520215233e-06, "loss": 0.0333, "step": 4879 }, { "epoch": 1.1103526734926052, "grad_norm": 0.9877268530404503, "learning_rate": 1.1040265654052575e-06, "loss": 0.0223, "step": 4880 }, { "epoch": 1.110580204778157, "grad_norm": 0.6647674925503221, "learning_rate": 1.103969168998566e-06, "loss": 0.0332, "step": 4881 }, { "epoch": 1.1108077360637088, "grad_norm": 1.2643667840897899, "learning_rate": 1.1039117628026222e-06, "loss": 0.0458, "step": 4882 }, { "epoch": 1.1110352673492605, "grad_norm": 0.9162239690822223, "learning_rate": 1.1038543468185988e-06, "loss": 0.0169, "step": 4883 }, { "epoch": 1.1112627986348123, "grad_norm": 0.8043833710259668, "learning_rate": 1.1037969210476696e-06, "loss": 0.013, "step": 4884 }, { "epoch": 1.111490329920364, "grad_norm": 0.9360911016853952, "learning_rate": 1.1037394854910082e-06, "loss": 0.0299, "step": 4885 }, { "epoch": 1.1117178612059158, "grad_norm": 0.8478138590871992, "learning_rate": 1.1036820401497884e-06, "loss": 0.0104, "step": 4886 }, { "epoch": 1.1119453924914675, "grad_norm": 0.7885738829466076, "learning_rate": 1.1036245850251844e-06, "loss": 0.0119, "step": 4887 }, { "epoch": 1.1121729237770193, "grad_norm": 0.711711942639679, "learning_rate": 1.1035671201183706e-06, "loss": 0.0167, "step": 4888 }, { "epoch": 1.112400455062571, "grad_norm": 0.6664367950082724, "learning_rate": 1.1035096454305213e-06, "loss": 0.0146, "step": 4889 }, { "epoch": 1.1126279863481228, "grad_norm": 1.5180095449935846, "learning_rate": 1.103452160962811e-06, "loss": 0.0206, "step": 4890 }, { "epoch": 1.1128555176336747, "grad_norm": 0.9314557397364934, "learning_rate": 1.103394666716415e-06, "loss": 0.0399, "step": 4891 }, { "epoch": 1.1130830489192265, "grad_norm": 0.7240617420566345, "learning_rate": 1.1033371626925079e-06, "loss": 0.0104, "step": 4892 }, { "epoch": 1.1133105802047782, "grad_norm": 1.000892668316659, "learning_rate": 1.1032796488922653e-06, "loss": 0.0186, "step": 4893 }, { "epoch": 1.11353811149033, "grad_norm": 1.0694112940902762, "learning_rate": 1.103222125316863e-06, "loss": 0.0151, "step": 4894 }, { "epoch": 1.1137656427758817, "grad_norm": 0.9676066050033223, "learning_rate": 1.1031645919674758e-06, "loss": 0.0139, "step": 4895 }, { "epoch": 1.1139931740614335, "grad_norm": 1.0428454696146314, "learning_rate": 1.10310704884528e-06, "loss": 0.0306, "step": 4896 }, { "epoch": 1.1142207053469853, "grad_norm": 1.2326650915996091, "learning_rate": 1.1030494959514521e-06, "loss": 0.0296, "step": 4897 }, { "epoch": 1.114448236632537, "grad_norm": 1.23258544073363, "learning_rate": 1.1029919332871678e-06, "loss": 0.0365, "step": 4898 }, { "epoch": 1.1146757679180888, "grad_norm": 0.9035657907747793, "learning_rate": 1.102934360853604e-06, "loss": 0.0128, "step": 4899 }, { "epoch": 1.1149032992036405, "grad_norm": 1.1444479344404774, "learning_rate": 1.1028767786519368e-06, "loss": 0.0141, "step": 4900 }, { "epoch": 1.1151308304891923, "grad_norm": 1.2429676441421305, "learning_rate": 1.1028191866833438e-06, "loss": 0.0285, "step": 4901 }, { "epoch": 1.115358361774744, "grad_norm": 0.9590079950756828, "learning_rate": 1.1027615849490014e-06, "loss": 0.0112, "step": 4902 }, { "epoch": 1.1155858930602958, "grad_norm": 1.572926999023847, "learning_rate": 1.1027039734500872e-06, "loss": 0.0369, "step": 4903 }, { "epoch": 1.1158134243458475, "grad_norm": 1.1352227453212334, "learning_rate": 1.102646352187779e-06, "loss": 0.0359, "step": 4904 }, { "epoch": 1.1160409556313993, "grad_norm": 1.1716125851334267, "learning_rate": 1.1025887211632538e-06, "loss": 0.02, "step": 4905 }, { "epoch": 1.116268486916951, "grad_norm": 1.1832800935639822, "learning_rate": 1.1025310803776898e-06, "loss": 0.0415, "step": 4906 }, { "epoch": 1.1164960182025028, "grad_norm": 1.1136610636887923, "learning_rate": 1.1024734298322655e-06, "loss": 0.0184, "step": 4907 }, { "epoch": 1.1167235494880545, "grad_norm": 1.1211312468158372, "learning_rate": 1.1024157695281582e-06, "loss": 0.029, "step": 4908 }, { "epoch": 1.1169510807736063, "grad_norm": 0.9972384903553551, "learning_rate": 1.1023580994665472e-06, "loss": 0.0201, "step": 4909 }, { "epoch": 1.117178612059158, "grad_norm": 1.7707402126553615, "learning_rate": 1.1023004196486108e-06, "loss": 0.043, "step": 4910 }, { "epoch": 1.11740614334471, "grad_norm": 1.134530398414991, "learning_rate": 1.102242730075528e-06, "loss": 0.0225, "step": 4911 }, { "epoch": 1.1176336746302618, "grad_norm": 1.627488244438378, "learning_rate": 1.1021850307484776e-06, "loss": 0.02, "step": 4912 }, { "epoch": 1.1178612059158135, "grad_norm": 1.1040646049948346, "learning_rate": 1.1021273216686397e-06, "loss": 0.0173, "step": 4913 }, { "epoch": 1.1180887372013653, "grad_norm": 1.5838051197154315, "learning_rate": 1.1020696028371926e-06, "loss": 0.024, "step": 4914 }, { "epoch": 1.118316268486917, "grad_norm": 0.8992624282859318, "learning_rate": 1.1020118742553166e-06, "loss": 0.0226, "step": 4915 }, { "epoch": 1.1185437997724688, "grad_norm": 1.3853418643898612, "learning_rate": 1.1019541359241917e-06, "loss": 0.0277, "step": 4916 }, { "epoch": 1.1187713310580205, "grad_norm": 1.0102231219601017, "learning_rate": 1.1018963878449976e-06, "loss": 0.013, "step": 4917 }, { "epoch": 1.1189988623435723, "grad_norm": 1.1790884765143759, "learning_rate": 1.1018386300189148e-06, "loss": 0.0176, "step": 4918 }, { "epoch": 1.119226393629124, "grad_norm": 0.7433554447296712, "learning_rate": 1.1017808624471237e-06, "loss": 0.0196, "step": 4919 }, { "epoch": 1.1194539249146758, "grad_norm": 0.9334325084247519, "learning_rate": 1.101723085130805e-06, "loss": 0.0241, "step": 4920 }, { "epoch": 1.1196814562002275, "grad_norm": 0.7643706032076052, "learning_rate": 1.1016652980711392e-06, "loss": 0.0138, "step": 4921 }, { "epoch": 1.1199089874857793, "grad_norm": 0.7474429513186569, "learning_rate": 1.1016075012693082e-06, "loss": 0.0266, "step": 4922 }, { "epoch": 1.120136518771331, "grad_norm": 1.1711233382681683, "learning_rate": 1.1015496947264923e-06, "loss": 0.0324, "step": 4923 }, { "epoch": 1.1203640500568828, "grad_norm": 2.534130515487931, "learning_rate": 1.1014918784438736e-06, "loss": 0.0624, "step": 4924 }, { "epoch": 1.1205915813424345, "grad_norm": 1.4712142976504, "learning_rate": 1.1014340524226337e-06, "loss": 0.0425, "step": 4925 }, { "epoch": 1.1208191126279863, "grad_norm": 0.8897734161115947, "learning_rate": 1.101376216663954e-06, "loss": 0.0146, "step": 4926 }, { "epoch": 1.121046643913538, "grad_norm": 1.6323492107077655, "learning_rate": 1.101318371169017e-06, "loss": 0.0446, "step": 4927 }, { "epoch": 1.1212741751990898, "grad_norm": 0.9646427394649012, "learning_rate": 1.1012605159390048e-06, "loss": 0.0312, "step": 4928 }, { "epoch": 1.1215017064846415, "grad_norm": 0.7920927385811001, "learning_rate": 1.1012026509751e-06, "loss": 0.0194, "step": 4929 }, { "epoch": 1.1217292377701935, "grad_norm": 1.1574113409003755, "learning_rate": 1.1011447762784849e-06, "loss": 0.0237, "step": 4930 }, { "epoch": 1.1219567690557453, "grad_norm": 1.2517329969278503, "learning_rate": 1.1010868918503429e-06, "loss": 0.0207, "step": 4931 }, { "epoch": 1.122184300341297, "grad_norm": 0.8071040660470671, "learning_rate": 1.1010289976918565e-06, "loss": 0.0207, "step": 4932 }, { "epoch": 1.1224118316268488, "grad_norm": 0.9373790639087559, "learning_rate": 1.1009710938042093e-06, "loss": 0.0153, "step": 4933 }, { "epoch": 1.1226393629124005, "grad_norm": 1.2887447933260439, "learning_rate": 1.1009131801885848e-06, "loss": 0.0239, "step": 4934 }, { "epoch": 1.1228668941979523, "grad_norm": 1.40441656204075, "learning_rate": 1.100855256846166e-06, "loss": 0.0343, "step": 4935 }, { "epoch": 1.123094425483504, "grad_norm": 0.689175568925837, "learning_rate": 1.1007973237781377e-06, "loss": 0.0148, "step": 4936 }, { "epoch": 1.1233219567690558, "grad_norm": 1.1787781392551855, "learning_rate": 1.1007393809856834e-06, "loss": 0.0286, "step": 4937 }, { "epoch": 1.1235494880546075, "grad_norm": 1.0271974875036538, "learning_rate": 1.1006814284699873e-06, "loss": 0.0373, "step": 4938 }, { "epoch": 1.1237770193401593, "grad_norm": 0.8297711019866251, "learning_rate": 1.100623466232234e-06, "loss": 0.0261, "step": 4939 }, { "epoch": 1.124004550625711, "grad_norm": 1.0858267777295185, "learning_rate": 1.1005654942736082e-06, "loss": 0.0261, "step": 4940 }, { "epoch": 1.1242320819112628, "grad_norm": 0.6790255085030746, "learning_rate": 1.1005075125952946e-06, "loss": 0.0156, "step": 4941 }, { "epoch": 1.1244596131968145, "grad_norm": 1.0737443834894482, "learning_rate": 1.1004495211984783e-06, "loss": 0.0342, "step": 4942 }, { "epoch": 1.1246871444823663, "grad_norm": 1.1759005758316916, "learning_rate": 1.1003915200843446e-06, "loss": 0.0296, "step": 4943 }, { "epoch": 1.124914675767918, "grad_norm": 1.4502960478357807, "learning_rate": 1.1003335092540787e-06, "loss": 0.0338, "step": 4944 }, { "epoch": 1.1251422070534698, "grad_norm": 1.0232031902807623, "learning_rate": 1.1002754887088665e-06, "loss": 0.0275, "step": 4945 }, { "epoch": 1.1253697383390215, "grad_norm": 1.1276836760259192, "learning_rate": 1.1002174584498938e-06, "loss": 0.018, "step": 4946 }, { "epoch": 1.1255972696245733, "grad_norm": 1.121912066969083, "learning_rate": 1.1001594184783464e-06, "loss": 0.0224, "step": 4947 }, { "epoch": 1.125824800910125, "grad_norm": 1.1026078266089805, "learning_rate": 1.1001013687954109e-06, "loss": 0.0265, "step": 4948 }, { "epoch": 1.1260523321956768, "grad_norm": 0.6996142941788891, "learning_rate": 1.1000433094022735e-06, "loss": 0.0204, "step": 4949 }, { "epoch": 1.1262798634812285, "grad_norm": 1.022975678784142, "learning_rate": 1.0999852403001208e-06, "loss": 0.0189, "step": 4950 }, { "epoch": 1.1265073947667805, "grad_norm": 1.2687033993187262, "learning_rate": 1.0999271614901396e-06, "loss": 0.0141, "step": 4951 }, { "epoch": 1.1267349260523323, "grad_norm": 1.799956791253596, "learning_rate": 1.099869072973517e-06, "loss": 0.0146, "step": 4952 }, { "epoch": 1.126962457337884, "grad_norm": 1.5492585986994671, "learning_rate": 1.0998109747514404e-06, "loss": 0.0447, "step": 4953 }, { "epoch": 1.1271899886234358, "grad_norm": 0.7598908512501374, "learning_rate": 1.099752866825097e-06, "loss": 0.0166, "step": 4954 }, { "epoch": 1.1274175199089875, "grad_norm": 0.6362839552889372, "learning_rate": 1.0996947491956745e-06, "loss": 0.0113, "step": 4955 }, { "epoch": 1.1276450511945393, "grad_norm": 1.0336058491900288, "learning_rate": 1.0996366218643607e-06, "loss": 0.0156, "step": 4956 }, { "epoch": 1.127872582480091, "grad_norm": 0.7457309100980003, "learning_rate": 1.0995784848323434e-06, "loss": 0.0151, "step": 4957 }, { "epoch": 1.1281001137656428, "grad_norm": 1.649804543332714, "learning_rate": 1.0995203381008112e-06, "loss": 0.0347, "step": 4958 }, { "epoch": 1.1283276450511945, "grad_norm": 1.2533333259409183, "learning_rate": 1.099462181670952e-06, "loss": 0.0288, "step": 4959 }, { "epoch": 1.1285551763367463, "grad_norm": 1.1428064969183531, "learning_rate": 1.0994040155439553e-06, "loss": 0.0263, "step": 4960 }, { "epoch": 1.128782707622298, "grad_norm": 0.7012991923657113, "learning_rate": 1.0993458397210092e-06, "loss": 0.0108, "step": 4961 }, { "epoch": 1.1290102389078498, "grad_norm": 1.319347853084701, "learning_rate": 1.0992876542033026e-06, "loss": 0.0232, "step": 4962 }, { "epoch": 1.1292377701934015, "grad_norm": 0.8983600851477538, "learning_rate": 1.0992294589920252e-06, "loss": 0.015, "step": 4963 }, { "epoch": 1.1294653014789533, "grad_norm": 0.6582052508958808, "learning_rate": 1.099171254088366e-06, "loss": 0.0129, "step": 4964 }, { "epoch": 1.129692832764505, "grad_norm": 2.0776942729185306, "learning_rate": 1.0991130394935148e-06, "loss": 0.0128, "step": 4965 }, { "epoch": 1.1299203640500568, "grad_norm": 1.3980552891281348, "learning_rate": 1.0990548152086616e-06, "loss": 0.0158, "step": 4966 }, { "epoch": 1.1301478953356086, "grad_norm": 1.399217668963092, "learning_rate": 1.098996581234996e-06, "loss": 0.0214, "step": 4967 }, { "epoch": 1.1303754266211605, "grad_norm": 1.0723453956148017, "learning_rate": 1.0989383375737081e-06, "loss": 0.0124, "step": 4968 }, { "epoch": 1.1306029579067123, "grad_norm": 1.1545644366726804, "learning_rate": 1.098880084225989e-06, "loss": 0.0412, "step": 4969 }, { "epoch": 1.130830489192264, "grad_norm": 0.942557146311076, "learning_rate": 1.0988218211930285e-06, "loss": 0.0201, "step": 4970 }, { "epoch": 1.1310580204778158, "grad_norm": 1.0043398319888264, "learning_rate": 1.0987635484760178e-06, "loss": 0.0198, "step": 4971 }, { "epoch": 1.1312855517633675, "grad_norm": 1.5698009312279446, "learning_rate": 1.098705266076148e-06, "loss": 0.0258, "step": 4972 }, { "epoch": 1.1315130830489193, "grad_norm": 1.4423264895380257, "learning_rate": 1.0986469739946102e-06, "loss": 0.0325, "step": 4973 }, { "epoch": 1.131740614334471, "grad_norm": 1.2136974558872335, "learning_rate": 1.0985886722325954e-06, "loss": 0.0217, "step": 4974 }, { "epoch": 1.1319681456200228, "grad_norm": 1.0091129092645157, "learning_rate": 1.0985303607912956e-06, "loss": 0.0134, "step": 4975 }, { "epoch": 1.1321956769055745, "grad_norm": 0.9544088452985304, "learning_rate": 1.0984720396719024e-06, "loss": 0.0243, "step": 4976 }, { "epoch": 1.1324232081911263, "grad_norm": 1.03400426394427, "learning_rate": 1.098413708875608e-06, "loss": 0.0165, "step": 4977 }, { "epoch": 1.132650739476678, "grad_norm": 0.9584699961425377, "learning_rate": 1.098355368403604e-06, "loss": 0.0159, "step": 4978 }, { "epoch": 1.1328782707622298, "grad_norm": 1.1633397537719241, "learning_rate": 1.0982970182570837e-06, "loss": 0.027, "step": 4979 }, { "epoch": 1.1331058020477816, "grad_norm": 0.6150701132014431, "learning_rate": 1.098238658437239e-06, "loss": 0.0134, "step": 4980 }, { "epoch": 1.1333333333333333, "grad_norm": 0.9628332044868487, "learning_rate": 1.0981802889452627e-06, "loss": 0.0244, "step": 4981 }, { "epoch": 1.133560864618885, "grad_norm": 0.8522049310117307, "learning_rate": 1.0981219097823479e-06, "loss": 0.0106, "step": 4982 }, { "epoch": 1.1337883959044368, "grad_norm": 0.9473163604101656, "learning_rate": 1.0980635209496878e-06, "loss": 0.024, "step": 4983 }, { "epoch": 1.1340159271899886, "grad_norm": 0.8002876838010067, "learning_rate": 1.0980051224484756e-06, "loss": 0.0083, "step": 4984 }, { "epoch": 1.1342434584755403, "grad_norm": 1.275340108026526, "learning_rate": 1.0979467142799052e-06, "loss": 0.0425, "step": 4985 }, { "epoch": 1.134470989761092, "grad_norm": 1.1242771585253066, "learning_rate": 1.0978882964451698e-06, "loss": 0.0235, "step": 4986 }, { "epoch": 1.1346985210466438, "grad_norm": 0.8316998784778787, "learning_rate": 1.097829868945464e-06, "loss": 0.0138, "step": 4987 }, { "epoch": 1.1349260523321956, "grad_norm": 1.2826664766461733, "learning_rate": 1.0977714317819812e-06, "loss": 0.0148, "step": 4988 }, { "epoch": 1.1351535836177473, "grad_norm": 1.7211760416844804, "learning_rate": 1.0977129849559165e-06, "loss": 0.0336, "step": 4989 }, { "epoch": 1.1353811149032993, "grad_norm": 0.8143435710252414, "learning_rate": 1.0976545284684642e-06, "loss": 0.0066, "step": 4990 }, { "epoch": 1.135608646188851, "grad_norm": 1.0342548252924428, "learning_rate": 1.0975960623208188e-06, "loss": 0.0162, "step": 4991 }, { "epoch": 1.1358361774744028, "grad_norm": 1.7590039011675342, "learning_rate": 1.0975375865141753e-06, "loss": 0.0362, "step": 4992 }, { "epoch": 1.1360637087599545, "grad_norm": 1.6822978128433002, "learning_rate": 1.097479101049729e-06, "loss": 0.0185, "step": 4993 }, { "epoch": 1.1362912400455063, "grad_norm": 1.2344613037091736, "learning_rate": 1.0974206059286752e-06, "loss": 0.0374, "step": 4994 }, { "epoch": 1.136518771331058, "grad_norm": 0.5280046364865614, "learning_rate": 1.0973621011522096e-06, "loss": 0.0084, "step": 4995 }, { "epoch": 1.1367463026166098, "grad_norm": 1.9687955658616516, "learning_rate": 1.0973035867215276e-06, "loss": 0.0336, "step": 4996 }, { "epoch": 1.1369738339021616, "grad_norm": 1.022624248487459, "learning_rate": 1.0972450626378254e-06, "loss": 0.0225, "step": 4997 }, { "epoch": 1.1372013651877133, "grad_norm": 0.48108692573634865, "learning_rate": 1.0971865289022988e-06, "loss": 0.008, "step": 4998 }, { "epoch": 1.137428896473265, "grad_norm": 0.9799413189160909, "learning_rate": 1.0971279855161442e-06, "loss": 0.0394, "step": 4999 }, { "epoch": 1.1376564277588168, "grad_norm": 1.466751462689922, "learning_rate": 1.0970694324805586e-06, "loss": 0.0167, "step": 5000 }, { "epoch": 1.1378839590443686, "grad_norm": 1.5693112940714422, "learning_rate": 1.0970108697967382e-06, "loss": 0.0325, "step": 5001 }, { "epoch": 1.1381114903299203, "grad_norm": 0.9255284204543452, "learning_rate": 1.09695229746588e-06, "loss": 0.0362, "step": 5002 }, { "epoch": 1.138339021615472, "grad_norm": 1.7087407221932875, "learning_rate": 1.0968937154891812e-06, "loss": 0.052, "step": 5003 }, { "epoch": 1.1385665529010238, "grad_norm": 2.428840915012362, "learning_rate": 1.096835123867839e-06, "loss": 0.0366, "step": 5004 }, { "epoch": 1.1387940841865756, "grad_norm": 0.6595273191874941, "learning_rate": 1.0967765226030512e-06, "loss": 0.0215, "step": 5005 }, { "epoch": 1.1390216154721273, "grad_norm": 1.2073236201929451, "learning_rate": 1.0967179116960153e-06, "loss": 0.0431, "step": 5006 }, { "epoch": 1.1392491467576793, "grad_norm": 1.517298131220002, "learning_rate": 1.096659291147929e-06, "loss": 0.0493, "step": 5007 }, { "epoch": 1.139476678043231, "grad_norm": 0.8459583474402816, "learning_rate": 1.0966006609599908e-06, "loss": 0.0161, "step": 5008 }, { "epoch": 1.1397042093287828, "grad_norm": 0.9402719162645292, "learning_rate": 1.0965420211333984e-06, "loss": 0.0199, "step": 5009 }, { "epoch": 1.1399317406143346, "grad_norm": 0.8824336019611823, "learning_rate": 1.0964833716693512e-06, "loss": 0.0236, "step": 5010 }, { "epoch": 1.1401592718998863, "grad_norm": 0.9061765730464424, "learning_rate": 1.096424712569047e-06, "loss": 0.0144, "step": 5011 }, { "epoch": 1.140386803185438, "grad_norm": 1.1463154922222984, "learning_rate": 1.0963660438336851e-06, "loss": 0.0405, "step": 5012 }, { "epoch": 1.1406143344709898, "grad_norm": 0.6924366133128478, "learning_rate": 1.0963073654644645e-06, "loss": 0.0139, "step": 5013 }, { "epoch": 1.1408418657565416, "grad_norm": 0.9228757161234781, "learning_rate": 1.0962486774625847e-06, "loss": 0.0227, "step": 5014 }, { "epoch": 1.1410693970420933, "grad_norm": 0.9665478631542946, "learning_rate": 1.096189979829245e-06, "loss": 0.0273, "step": 5015 }, { "epoch": 1.141296928327645, "grad_norm": 0.9370551975836519, "learning_rate": 1.096131272565645e-06, "loss": 0.0201, "step": 5016 }, { "epoch": 1.1415244596131968, "grad_norm": 1.0359990365775988, "learning_rate": 1.0960725556729845e-06, "loss": 0.0188, "step": 5017 }, { "epoch": 1.1417519908987486, "grad_norm": 0.6081258142385076, "learning_rate": 1.0960138291524637e-06, "loss": 0.0123, "step": 5018 }, { "epoch": 1.1419795221843003, "grad_norm": 0.8151694196582907, "learning_rate": 1.095955093005283e-06, "loss": 0.0178, "step": 5019 }, { "epoch": 1.142207053469852, "grad_norm": 0.972811628969212, "learning_rate": 1.0958963472326426e-06, "loss": 0.0114, "step": 5020 }, { "epoch": 1.1424345847554038, "grad_norm": 0.7206858563005965, "learning_rate": 1.0958375918357433e-06, "loss": 0.011, "step": 5021 }, { "epoch": 1.1426621160409556, "grad_norm": 0.8388894087062686, "learning_rate": 1.095778826815786e-06, "loss": 0.0214, "step": 5022 }, { "epoch": 1.1428896473265073, "grad_norm": 1.4288807810887136, "learning_rate": 1.0957200521739715e-06, "loss": 0.0315, "step": 5023 }, { "epoch": 1.143117178612059, "grad_norm": 0.9634756984934681, "learning_rate": 1.0956612679115012e-06, "loss": 0.0338, "step": 5024 }, { "epoch": 1.1433447098976108, "grad_norm": 1.2246259764288705, "learning_rate": 1.0956024740295767e-06, "loss": 0.0339, "step": 5025 }, { "epoch": 1.1435722411831626, "grad_norm": 0.9727990066152786, "learning_rate": 1.0955436705293996e-06, "loss": 0.0186, "step": 5026 }, { "epoch": 1.1437997724687143, "grad_norm": 1.4271212460831064, "learning_rate": 1.0954848574121715e-06, "loss": 0.0407, "step": 5027 }, { "epoch": 1.144027303754266, "grad_norm": 1.1642512101338942, "learning_rate": 1.0954260346790944e-06, "loss": 0.0305, "step": 5028 }, { "epoch": 1.144254835039818, "grad_norm": 3.382254274133764, "learning_rate": 1.0953672023313709e-06, "loss": 0.0462, "step": 5029 }, { "epoch": 1.1444823663253698, "grad_norm": 1.3043659225114033, "learning_rate": 1.0953083603702031e-06, "loss": 0.0269, "step": 5030 }, { "epoch": 1.1447098976109216, "grad_norm": 0.8941373741234482, "learning_rate": 1.0952495087967939e-06, "loss": 0.01, "step": 5031 }, { "epoch": 1.1449374288964733, "grad_norm": 1.0345429364832142, "learning_rate": 1.095190647612346e-06, "loss": 0.0313, "step": 5032 }, { "epoch": 1.145164960182025, "grad_norm": 0.9469945046545952, "learning_rate": 1.0951317768180623e-06, "loss": 0.0185, "step": 5033 }, { "epoch": 1.1453924914675768, "grad_norm": 1.5417022710277841, "learning_rate": 1.0950728964151457e-06, "loss": 0.0477, "step": 5034 }, { "epoch": 1.1456200227531286, "grad_norm": 1.2090565974916883, "learning_rate": 1.0950140064048005e-06, "loss": 0.0304, "step": 5035 }, { "epoch": 1.1458475540386803, "grad_norm": 1.6222929147315774, "learning_rate": 1.0949551067882297e-06, "loss": 0.0449, "step": 5036 }, { "epoch": 1.146075085324232, "grad_norm": 0.8458047187793855, "learning_rate": 1.094896197566637e-06, "loss": 0.0177, "step": 5037 }, { "epoch": 1.1463026166097838, "grad_norm": 0.8642635965418713, "learning_rate": 1.0948372787412267e-06, "loss": 0.019, "step": 5038 }, { "epoch": 1.1465301478953356, "grad_norm": 1.6962052412561806, "learning_rate": 1.094778350313203e-06, "loss": 0.0388, "step": 5039 }, { "epoch": 1.1467576791808873, "grad_norm": 1.2082274113676141, "learning_rate": 1.09471941228377e-06, "loss": 0.0443, "step": 5040 }, { "epoch": 1.146985210466439, "grad_norm": 1.551649129101969, "learning_rate": 1.0946604646541327e-06, "loss": 0.0458, "step": 5041 }, { "epoch": 1.1472127417519908, "grad_norm": 0.6322582451330618, "learning_rate": 1.0946015074254957e-06, "loss": 0.0063, "step": 5042 }, { "epoch": 1.1474402730375426, "grad_norm": 1.2390695869307191, "learning_rate": 1.0945425405990636e-06, "loss": 0.0438, "step": 5043 }, { "epoch": 1.1476678043230943, "grad_norm": 1.171113561669916, "learning_rate": 1.094483564176042e-06, "loss": 0.0304, "step": 5044 }, { "epoch": 1.147895335608646, "grad_norm": 1.111705225220342, "learning_rate": 1.0944245781576363e-06, "loss": 0.0199, "step": 5045 }, { "epoch": 1.148122866894198, "grad_norm": 0.7788906673524727, "learning_rate": 1.0943655825450517e-06, "loss": 0.017, "step": 5046 }, { "epoch": 1.1483503981797498, "grad_norm": 1.37808766490078, "learning_rate": 1.0943065773394943e-06, "loss": 0.027, "step": 5047 }, { "epoch": 1.1485779294653016, "grad_norm": 1.0016121975434131, "learning_rate": 1.0942475625421701e-06, "loss": 0.0217, "step": 5048 }, { "epoch": 1.1488054607508533, "grad_norm": 1.5673100085633027, "learning_rate": 1.094188538154285e-06, "loss": 0.039, "step": 5049 }, { "epoch": 1.149032992036405, "grad_norm": 1.1060010064378178, "learning_rate": 1.0941295041770453e-06, "loss": 0.0158, "step": 5050 }, { "epoch": 1.1492605233219568, "grad_norm": 1.1961653202374867, "learning_rate": 1.0940704606116578e-06, "loss": 0.0197, "step": 5051 }, { "epoch": 1.1494880546075086, "grad_norm": 0.9641941727834898, "learning_rate": 1.0940114074593292e-06, "loss": 0.0141, "step": 5052 }, { "epoch": 1.1497155858930603, "grad_norm": 0.6461548194155666, "learning_rate": 1.0939523447212665e-06, "loss": 0.0074, "step": 5053 }, { "epoch": 1.149943117178612, "grad_norm": 1.125395408330258, "learning_rate": 1.0938932723986766e-06, "loss": 0.0316, "step": 5054 }, { "epoch": 1.1501706484641638, "grad_norm": 0.6999270740525377, "learning_rate": 1.0938341904927669e-06, "loss": 0.0115, "step": 5055 }, { "epoch": 1.1503981797497156, "grad_norm": 0.824514611548941, "learning_rate": 1.093775099004745e-06, "loss": 0.0146, "step": 5056 }, { "epoch": 1.1506257110352673, "grad_norm": 1.2958988293911842, "learning_rate": 1.0937159979358186e-06, "loss": 0.0285, "step": 5057 }, { "epoch": 1.150853242320819, "grad_norm": 0.5661057032643344, "learning_rate": 1.0936568872871958e-06, "loss": 0.0085, "step": 5058 }, { "epoch": 1.1510807736063708, "grad_norm": 0.8911202539112528, "learning_rate": 1.0935977670600843e-06, "loss": 0.0249, "step": 5059 }, { "epoch": 1.1513083048919226, "grad_norm": 0.6119358774484015, "learning_rate": 1.0935386372556928e-06, "loss": 0.0081, "step": 5060 }, { "epoch": 1.1515358361774743, "grad_norm": 1.1732745638400495, "learning_rate": 1.0934794978752295e-06, "loss": 0.0304, "step": 5061 }, { "epoch": 1.151763367463026, "grad_norm": 1.4610976896610823, "learning_rate": 1.0934203489199033e-06, "loss": 0.026, "step": 5062 }, { "epoch": 1.1519908987485779, "grad_norm": 0.8751176409958491, "learning_rate": 1.093361190390923e-06, "loss": 0.0129, "step": 5063 }, { "epoch": 1.1522184300341296, "grad_norm": 0.9095379255951723, "learning_rate": 1.0933020222894978e-06, "loss": 0.0189, "step": 5064 }, { "epoch": 1.1524459613196814, "grad_norm": 0.886509655802644, "learning_rate": 1.0932428446168369e-06, "loss": 0.0094, "step": 5065 }, { "epoch": 1.152673492605233, "grad_norm": 1.0165308626390206, "learning_rate": 1.0931836573741498e-06, "loss": 0.0194, "step": 5066 }, { "epoch": 1.1529010238907849, "grad_norm": 2.2862767542657307, "learning_rate": 1.093124460562646e-06, "loss": 0.03, "step": 5067 }, { "epoch": 1.1531285551763368, "grad_norm": 3.0841316980515, "learning_rate": 1.0930652541835357e-06, "loss": 0.1031, "step": 5068 }, { "epoch": 1.1533560864618886, "grad_norm": 2.057276950609037, "learning_rate": 1.093006038238029e-06, "loss": 0.0498, "step": 5069 }, { "epoch": 1.1535836177474403, "grad_norm": 0.8535534714067085, "learning_rate": 1.0929468127273357e-06, "loss": 0.0262, "step": 5070 }, { "epoch": 1.153811149032992, "grad_norm": 0.7487264216032792, "learning_rate": 1.0928875776526667e-06, "loss": 0.0187, "step": 5071 }, { "epoch": 1.1540386803185438, "grad_norm": 1.011442354116883, "learning_rate": 1.0928283330152325e-06, "loss": 0.0237, "step": 5072 }, { "epoch": 1.1542662116040956, "grad_norm": 1.2385558290850083, "learning_rate": 1.092769078816244e-06, "loss": 0.0576, "step": 5073 }, { "epoch": 1.1544937428896473, "grad_norm": 1.2917165551359526, "learning_rate": 1.092709815056912e-06, "loss": 0.0211, "step": 5074 }, { "epoch": 1.154721274175199, "grad_norm": 1.0590341552017861, "learning_rate": 1.0926505417384482e-06, "loss": 0.0303, "step": 5075 }, { "epoch": 1.1549488054607508, "grad_norm": 0.9189460734693743, "learning_rate": 1.0925912588620637e-06, "loss": 0.0147, "step": 5076 }, { "epoch": 1.1551763367463026, "grad_norm": 0.998558466002832, "learning_rate": 1.0925319664289703e-06, "loss": 0.0219, "step": 5077 }, { "epoch": 1.1554038680318544, "grad_norm": 0.7945299298414331, "learning_rate": 1.0924726644403797e-06, "loss": 0.0176, "step": 5078 }, { "epoch": 1.155631399317406, "grad_norm": 1.3566259925414048, "learning_rate": 1.0924133528975039e-06, "loss": 0.0295, "step": 5079 }, { "epoch": 1.1558589306029579, "grad_norm": 0.8074114789352509, "learning_rate": 1.0923540318015552e-06, "loss": 0.022, "step": 5080 }, { "epoch": 1.1560864618885096, "grad_norm": 1.3411051591390433, "learning_rate": 1.092294701153746e-06, "loss": 0.029, "step": 5081 }, { "epoch": 1.1563139931740614, "grad_norm": 1.423080211058415, "learning_rate": 1.092235360955289e-06, "loss": 0.0503, "step": 5082 }, { "epoch": 1.1565415244596131, "grad_norm": 0.8213852817117168, "learning_rate": 1.092176011207397e-06, "loss": 0.021, "step": 5083 }, { "epoch": 1.1567690557451649, "grad_norm": 1.2933428126076256, "learning_rate": 1.0921166519112828e-06, "loss": 0.0442, "step": 5084 }, { "epoch": 1.1569965870307168, "grad_norm": 0.5857327857396418, "learning_rate": 1.0920572830681597e-06, "loss": 0.0107, "step": 5085 }, { "epoch": 1.1572241183162686, "grad_norm": 1.0152274364913478, "learning_rate": 1.0919979046792411e-06, "loss": 0.021, "step": 5086 }, { "epoch": 1.1574516496018203, "grad_norm": 1.6197058508435338, "learning_rate": 1.0919385167457408e-06, "loss": 0.0317, "step": 5087 }, { "epoch": 1.157679180887372, "grad_norm": 0.8321181279990257, "learning_rate": 1.0918791192688722e-06, "loss": 0.0083, "step": 5088 }, { "epoch": 1.1579067121729238, "grad_norm": 1.1078885245620338, "learning_rate": 1.0918197122498495e-06, "loss": 0.0417, "step": 5089 }, { "epoch": 1.1581342434584756, "grad_norm": 0.8666450018016355, "learning_rate": 1.0917602956898867e-06, "loss": 0.0235, "step": 5090 }, { "epoch": 1.1583617747440274, "grad_norm": 0.9389403634525508, "learning_rate": 1.0917008695901985e-06, "loss": 0.0161, "step": 5091 }, { "epoch": 1.158589306029579, "grad_norm": 1.6343340981795507, "learning_rate": 1.091641433951999e-06, "loss": 0.0589, "step": 5092 }, { "epoch": 1.1588168373151309, "grad_norm": 1.0221675248406095, "learning_rate": 1.0915819887765034e-06, "loss": 0.0209, "step": 5093 }, { "epoch": 1.1590443686006826, "grad_norm": 1.2783954239375135, "learning_rate": 1.0915225340649264e-06, "loss": 0.0314, "step": 5094 }, { "epoch": 1.1592718998862344, "grad_norm": 1.0876136046711122, "learning_rate": 1.091463069818483e-06, "loss": 0.0183, "step": 5095 }, { "epoch": 1.159499431171786, "grad_norm": 0.92935249083099, "learning_rate": 1.091403596038389e-06, "loss": 0.0284, "step": 5096 }, { "epoch": 1.1597269624573379, "grad_norm": 1.532892418760994, "learning_rate": 1.0913441127258596e-06, "loss": 0.0229, "step": 5097 }, { "epoch": 1.1599544937428896, "grad_norm": 1.0034643002001107, "learning_rate": 1.0912846198821105e-06, "loss": 0.0176, "step": 5098 }, { "epoch": 1.1601820250284414, "grad_norm": 1.1190616568728453, "learning_rate": 1.091225117508358e-06, "loss": 0.0201, "step": 5099 }, { "epoch": 1.1604095563139931, "grad_norm": 0.7869200780957125, "learning_rate": 1.0911656056058175e-06, "loss": 0.0085, "step": 5100 }, { "epoch": 1.1606370875995449, "grad_norm": 0.7075180696463311, "learning_rate": 1.0911060841757063e-06, "loss": 0.0143, "step": 5101 }, { "epoch": 1.1608646188850966, "grad_norm": 0.6587063658622069, "learning_rate": 1.09104655321924e-06, "loss": 0.0174, "step": 5102 }, { "epoch": 1.1610921501706484, "grad_norm": 1.2851325080822802, "learning_rate": 1.0909870127376358e-06, "loss": 0.0347, "step": 5103 }, { "epoch": 1.1613196814562001, "grad_norm": 1.4645102454670038, "learning_rate": 1.0909274627321106e-06, "loss": 0.0412, "step": 5104 }, { "epoch": 1.1615472127417519, "grad_norm": 1.983767386413822, "learning_rate": 1.090867903203881e-06, "loss": 0.0224, "step": 5105 }, { "epoch": 1.1617747440273036, "grad_norm": 1.2382132521135707, "learning_rate": 1.090808334154165e-06, "loss": 0.0172, "step": 5106 }, { "epoch": 1.1620022753128556, "grad_norm": 0.8691675756314506, "learning_rate": 1.0907487555841797e-06, "loss": 0.0158, "step": 5107 }, { "epoch": 1.1622298065984074, "grad_norm": 2.1283896982776853, "learning_rate": 1.0906891674951426e-06, "loss": 0.0345, "step": 5108 }, { "epoch": 1.162457337883959, "grad_norm": 1.6543688130648284, "learning_rate": 1.090629569888272e-06, "loss": 0.0491, "step": 5109 }, { "epoch": 1.1626848691695109, "grad_norm": 0.6293933164268644, "learning_rate": 1.0905699627647857e-06, "loss": 0.0094, "step": 5110 }, { "epoch": 1.1629124004550626, "grad_norm": 1.6953242717121608, "learning_rate": 1.090510346125902e-06, "loss": 0.0238, "step": 5111 }, { "epoch": 1.1631399317406144, "grad_norm": 1.159765223167702, "learning_rate": 1.0904507199728392e-06, "loss": 0.0406, "step": 5112 }, { "epoch": 1.1633674630261661, "grad_norm": 1.3015729382676753, "learning_rate": 1.0903910843068163e-06, "loss": 0.0318, "step": 5113 }, { "epoch": 1.1635949943117179, "grad_norm": 0.8775703929826588, "learning_rate": 1.090331439129052e-06, "loss": 0.0087, "step": 5114 }, { "epoch": 1.1638225255972696, "grad_norm": 2.0908466705122963, "learning_rate": 1.0902717844407651e-06, "loss": 0.0187, "step": 5115 }, { "epoch": 1.1640500568828214, "grad_norm": 0.8211627405295623, "learning_rate": 1.0902121202431754e-06, "loss": 0.0214, "step": 5116 }, { "epoch": 1.1642775881683731, "grad_norm": 1.0607426301957161, "learning_rate": 1.0901524465375015e-06, "loss": 0.0305, "step": 5117 }, { "epoch": 1.1645051194539249, "grad_norm": 1.0553962833230413, "learning_rate": 1.0900927633249638e-06, "loss": 0.0217, "step": 5118 }, { "epoch": 1.1647326507394766, "grad_norm": 1.0244204786724485, "learning_rate": 1.0900330706067818e-06, "loss": 0.0264, "step": 5119 }, { "epoch": 1.1649601820250284, "grad_norm": 0.9707054088367129, "learning_rate": 1.0899733683841753e-06, "loss": 0.0162, "step": 5120 }, { "epoch": 1.1651877133105801, "grad_norm": 1.1311434603614883, "learning_rate": 1.0899136566583647e-06, "loss": 0.028, "step": 5121 }, { "epoch": 1.1654152445961319, "grad_norm": 0.6912191457949579, "learning_rate": 1.0898539354305706e-06, "loss": 0.0099, "step": 5122 }, { "epoch": 1.1656427758816836, "grad_norm": 0.7902212704972374, "learning_rate": 1.0897942047020131e-06, "loss": 0.0128, "step": 5123 }, { "epoch": 1.1658703071672356, "grad_norm": 0.5967448578657041, "learning_rate": 1.0897344644739139e-06, "loss": 0.0081, "step": 5124 }, { "epoch": 1.1660978384527874, "grad_norm": 0.804672182956438, "learning_rate": 1.089674714747493e-06, "loss": 0.0121, "step": 5125 }, { "epoch": 1.1663253697383391, "grad_norm": 1.289356948925363, "learning_rate": 1.0896149555239717e-06, "loss": 0.0322, "step": 5126 }, { "epoch": 1.1665529010238909, "grad_norm": 1.8000751184542518, "learning_rate": 1.0895551868045718e-06, "loss": 0.055, "step": 5127 }, { "epoch": 1.1667804323094426, "grad_norm": 0.7403604907569553, "learning_rate": 1.0894954085905147e-06, "loss": 0.0213, "step": 5128 }, { "epoch": 1.1670079635949944, "grad_norm": 0.7673485205599404, "learning_rate": 1.0894356208830223e-06, "loss": 0.0146, "step": 5129 }, { "epoch": 1.1672354948805461, "grad_norm": 0.8057347260016878, "learning_rate": 1.089375823683316e-06, "loss": 0.0114, "step": 5130 }, { "epoch": 1.1674630261660979, "grad_norm": 1.031201971745674, "learning_rate": 1.0893160169926186e-06, "loss": 0.0224, "step": 5131 }, { "epoch": 1.1676905574516496, "grad_norm": 1.1471763380869269, "learning_rate": 1.0892562008121522e-06, "loss": 0.0199, "step": 5132 }, { "epoch": 1.1679180887372014, "grad_norm": 1.1194487112088174, "learning_rate": 1.0891963751431392e-06, "loss": 0.0202, "step": 5133 }, { "epoch": 1.1681456200227531, "grad_norm": 0.9431094841741555, "learning_rate": 1.0891365399868022e-06, "loss": 0.0242, "step": 5134 }, { "epoch": 1.1683731513083049, "grad_norm": 0.8152241328521926, "learning_rate": 1.0890766953443646e-06, "loss": 0.0135, "step": 5135 }, { "epoch": 1.1686006825938566, "grad_norm": 1.111189670935272, "learning_rate": 1.0890168412170493e-06, "loss": 0.0313, "step": 5136 }, { "epoch": 1.1688282138794084, "grad_norm": 1.0125526520638173, "learning_rate": 1.0889569776060796e-06, "loss": 0.0219, "step": 5137 }, { "epoch": 1.1690557451649601, "grad_norm": 1.1702719514469648, "learning_rate": 1.088897104512679e-06, "loss": 0.0349, "step": 5138 }, { "epoch": 1.1692832764505119, "grad_norm": 0.8014688872464257, "learning_rate": 1.0888372219380709e-06, "loss": 0.0134, "step": 5139 }, { "epoch": 1.1695108077360636, "grad_norm": 1.2120196426423329, "learning_rate": 1.0887773298834798e-06, "loss": 0.0218, "step": 5140 }, { "epoch": 1.1697383390216154, "grad_norm": 0.5093518570185542, "learning_rate": 1.0887174283501293e-06, "loss": 0.0069, "step": 5141 }, { "epoch": 1.1699658703071671, "grad_norm": 1.0586184070871325, "learning_rate": 1.0886575173392435e-06, "loss": 0.0212, "step": 5142 }, { "epoch": 1.170193401592719, "grad_norm": 0.9121062929597813, "learning_rate": 1.0885975968520476e-06, "loss": 0.01, "step": 5143 }, { "epoch": 1.1704209328782706, "grad_norm": 1.1192151515949293, "learning_rate": 1.0885376668897656e-06, "loss": 0.0297, "step": 5144 }, { "epoch": 1.1706484641638226, "grad_norm": 1.1083415447915677, "learning_rate": 1.0884777274536228e-06, "loss": 0.0288, "step": 5145 }, { "epoch": 1.1708759954493744, "grad_norm": 1.1157775165463915, "learning_rate": 1.0884177785448441e-06, "loss": 0.0179, "step": 5146 }, { "epoch": 1.1711035267349261, "grad_norm": 0.6942923206091653, "learning_rate": 1.0883578201646546e-06, "loss": 0.0186, "step": 5147 }, { "epoch": 1.1713310580204779, "grad_norm": 1.0660205113669436, "learning_rate": 1.08829785231428e-06, "loss": 0.0154, "step": 5148 }, { "epoch": 1.1715585893060296, "grad_norm": 1.2226468081458537, "learning_rate": 1.0882378749949456e-06, "loss": 0.027, "step": 5149 }, { "epoch": 1.1717861205915814, "grad_norm": 0.6736525742954345, "learning_rate": 1.0881778882078774e-06, "loss": 0.0104, "step": 5150 }, { "epoch": 1.1720136518771331, "grad_norm": 1.0175690285786059, "learning_rate": 1.0881178919543016e-06, "loss": 0.0209, "step": 5151 }, { "epoch": 1.1722411831626849, "grad_norm": 1.0274399291271918, "learning_rate": 1.0880578862354444e-06, "loss": 0.0169, "step": 5152 }, { "epoch": 1.1724687144482366, "grad_norm": 1.174579618024413, "learning_rate": 1.087997871052532e-06, "loss": 0.023, "step": 5153 }, { "epoch": 1.1726962457337884, "grad_norm": 1.0784512457128186, "learning_rate": 1.0879378464067906e-06, "loss": 0.0505, "step": 5154 }, { "epoch": 1.1729237770193401, "grad_norm": 1.2684504792555313, "learning_rate": 1.0878778122994477e-06, "loss": 0.0199, "step": 5155 }, { "epoch": 1.173151308304892, "grad_norm": 1.451789483548959, "learning_rate": 1.0878177687317302e-06, "loss": 0.0433, "step": 5156 }, { "epoch": 1.1733788395904436, "grad_norm": 0.9973415681599368, "learning_rate": 1.0877577157048648e-06, "loss": 0.0205, "step": 5157 }, { "epoch": 1.1736063708759954, "grad_norm": 2.814855357951941, "learning_rate": 1.0876976532200797e-06, "loss": 0.0309, "step": 5158 }, { "epoch": 1.1738339021615471, "grad_norm": 0.8916291744362826, "learning_rate": 1.0876375812786017e-06, "loss": 0.0207, "step": 5159 }, { "epoch": 1.174061433447099, "grad_norm": 0.7454883697223361, "learning_rate": 1.0875774998816586e-06, "loss": 0.0181, "step": 5160 }, { "epoch": 1.1742889647326507, "grad_norm": 1.5976521141289088, "learning_rate": 1.087517409030479e-06, "loss": 0.0352, "step": 5161 }, { "epoch": 1.1745164960182026, "grad_norm": 0.9023925994855438, "learning_rate": 1.0874573087262902e-06, "loss": 0.0189, "step": 5162 }, { "epoch": 1.1747440273037544, "grad_norm": 1.021608393538654, "learning_rate": 1.087397198970321e-06, "loss": 0.0248, "step": 5163 }, { "epoch": 1.1749715585893061, "grad_norm": 1.5179996755409662, "learning_rate": 1.0873370797638002e-06, "loss": 0.0346, "step": 5164 }, { "epoch": 1.1751990898748579, "grad_norm": 1.388608360949784, "learning_rate": 1.0872769511079561e-06, "loss": 0.0371, "step": 5165 }, { "epoch": 1.1754266211604096, "grad_norm": 0.9944177229048421, "learning_rate": 1.0872168130040175e-06, "loss": 0.0255, "step": 5166 }, { "epoch": 1.1756541524459614, "grad_norm": 1.460644807474802, "learning_rate": 1.087156665453214e-06, "loss": 0.0267, "step": 5167 }, { "epoch": 1.1758816837315131, "grad_norm": 1.201538811332985, "learning_rate": 1.0870965084567748e-06, "loss": 0.0322, "step": 5168 }, { "epoch": 1.176109215017065, "grad_norm": 0.9737685044159669, "learning_rate": 1.087036342015929e-06, "loss": 0.0238, "step": 5169 }, { "epoch": 1.1763367463026166, "grad_norm": 1.1123560277419813, "learning_rate": 1.0869761661319064e-06, "loss": 0.0323, "step": 5170 }, { "epoch": 1.1765642775881684, "grad_norm": 1.0917701556058614, "learning_rate": 1.0869159808059373e-06, "loss": 0.0239, "step": 5171 }, { "epoch": 1.1767918088737201, "grad_norm": 0.9254801632517604, "learning_rate": 1.0868557860392516e-06, "loss": 0.0225, "step": 5172 }, { "epoch": 1.177019340159272, "grad_norm": 1.1647756504926867, "learning_rate": 1.0867955818330792e-06, "loss": 0.0202, "step": 5173 }, { "epoch": 1.1772468714448237, "grad_norm": 0.9089361069774722, "learning_rate": 1.086735368188651e-06, "loss": 0.0167, "step": 5174 }, { "epoch": 1.1774744027303754, "grad_norm": 1.5111973761987758, "learning_rate": 1.0866751451071974e-06, "loss": 0.0295, "step": 5175 }, { "epoch": 1.1777019340159272, "grad_norm": 0.6596228779983846, "learning_rate": 1.0866149125899495e-06, "loss": 0.019, "step": 5176 }, { "epoch": 1.177929465301479, "grad_norm": 1.6565664804110074, "learning_rate": 1.086554670638138e-06, "loss": 0.031, "step": 5177 }, { "epoch": 1.1781569965870307, "grad_norm": 0.7431382614926502, "learning_rate": 1.0864944192529946e-06, "loss": 0.0125, "step": 5178 }, { "epoch": 1.1783845278725824, "grad_norm": 1.2886691602870353, "learning_rate": 1.08643415843575e-06, "loss": 0.0132, "step": 5179 }, { "epoch": 1.1786120591581342, "grad_norm": 1.2945110536777915, "learning_rate": 1.0863738881876369e-06, "loss": 0.0453, "step": 5180 }, { "epoch": 1.178839590443686, "grad_norm": 1.6944088405196147, "learning_rate": 1.0863136085098862e-06, "loss": 0.028, "step": 5181 }, { "epoch": 1.1790671217292377, "grad_norm": 0.9678714964974754, "learning_rate": 1.08625331940373e-06, "loss": 0.0186, "step": 5182 }, { "epoch": 1.1792946530147894, "grad_norm": 0.8164594441461164, "learning_rate": 1.086193020870401e-06, "loss": 0.0197, "step": 5183 }, { "epoch": 1.1795221843003414, "grad_norm": 0.847422876421066, "learning_rate": 1.0861327129111313e-06, "loss": 0.014, "step": 5184 }, { "epoch": 1.1797497155858931, "grad_norm": 0.8408136915565151, "learning_rate": 1.0860723955271533e-06, "loss": 0.0141, "step": 5185 }, { "epoch": 1.179977246871445, "grad_norm": 1.4185159384093218, "learning_rate": 1.0860120687196998e-06, "loss": 0.0174, "step": 5186 }, { "epoch": 1.1802047781569966, "grad_norm": 1.3653351030007033, "learning_rate": 1.0859517324900042e-06, "loss": 0.0318, "step": 5187 }, { "epoch": 1.1804323094425484, "grad_norm": 0.9616143252233853, "learning_rate": 1.0858913868392993e-06, "loss": 0.0324, "step": 5188 }, { "epoch": 1.1806598407281002, "grad_norm": 0.8826104173449131, "learning_rate": 1.0858310317688184e-06, "loss": 0.0194, "step": 5189 }, { "epoch": 1.180887372013652, "grad_norm": 0.573005849494334, "learning_rate": 1.0857706672797954e-06, "loss": 0.007, "step": 5190 }, { "epoch": 1.1811149032992037, "grad_norm": 0.792841731677897, "learning_rate": 1.0857102933734636e-06, "loss": 0.0228, "step": 5191 }, { "epoch": 1.1813424345847554, "grad_norm": 1.1198194956030572, "learning_rate": 1.0856499100510575e-06, "loss": 0.0204, "step": 5192 }, { "epoch": 1.1815699658703072, "grad_norm": 1.4060671628835655, "learning_rate": 1.0855895173138107e-06, "loss": 0.0335, "step": 5193 }, { "epoch": 1.181797497155859, "grad_norm": 1.0332252844504048, "learning_rate": 1.0855291151629576e-06, "loss": 0.0248, "step": 5194 }, { "epoch": 1.1820250284414107, "grad_norm": 1.3939283564653595, "learning_rate": 1.0854687035997329e-06, "loss": 0.0503, "step": 5195 }, { "epoch": 1.1822525597269624, "grad_norm": 1.389414773323681, "learning_rate": 1.0854082826253712e-06, "loss": 0.0179, "step": 5196 }, { "epoch": 1.1824800910125142, "grad_norm": 1.3843557347025581, "learning_rate": 1.0853478522411072e-06, "loss": 0.0235, "step": 5197 }, { "epoch": 1.182707622298066, "grad_norm": 3.1968855275158368, "learning_rate": 1.0852874124481764e-06, "loss": 0.0411, "step": 5198 }, { "epoch": 1.1829351535836177, "grad_norm": 1.4480239507963797, "learning_rate": 1.0852269632478138e-06, "loss": 0.0363, "step": 5199 }, { "epoch": 1.1831626848691694, "grad_norm": 1.9126854541087823, "learning_rate": 1.085166504641255e-06, "loss": 0.0181, "step": 5200 }, { "epoch": 1.1833902161547214, "grad_norm": 1.1213805046183494, "learning_rate": 1.0851060366297356e-06, "loss": 0.0353, "step": 5201 }, { "epoch": 1.1836177474402731, "grad_norm": 2.0688040255631126, "learning_rate": 1.0850455592144916e-06, "loss": 0.0451, "step": 5202 }, { "epoch": 1.183845278725825, "grad_norm": 1.7510399141137207, "learning_rate": 1.0849850723967585e-06, "loss": 0.0109, "step": 5203 }, { "epoch": 1.1840728100113767, "grad_norm": 6.664020652459381, "learning_rate": 1.0849245761777733e-06, "loss": 0.0381, "step": 5204 }, { "epoch": 1.1843003412969284, "grad_norm": 2.621359696329597, "learning_rate": 1.0848640705587718e-06, "loss": 0.0216, "step": 5205 }, { "epoch": 1.1845278725824802, "grad_norm": 1.8030419214813977, "learning_rate": 1.0848035555409911e-06, "loss": 0.0303, "step": 5206 }, { "epoch": 1.184755403868032, "grad_norm": 1.186220362313326, "learning_rate": 1.0847430311256676e-06, "loss": 0.0197, "step": 5207 }, { "epoch": 1.1849829351535837, "grad_norm": 1.5048336995036211, "learning_rate": 1.0846824973140388e-06, "loss": 0.0185, "step": 5208 }, { "epoch": 1.1852104664391354, "grad_norm": 1.8320739469192913, "learning_rate": 1.0846219541073417e-06, "loss": 0.0217, "step": 5209 }, { "epoch": 1.1854379977246872, "grad_norm": 1.6975307868867775, "learning_rate": 1.084561401506813e-06, "loss": 0.0491, "step": 5210 }, { "epoch": 1.185665529010239, "grad_norm": 2.1632818389721153, "learning_rate": 1.0845008395136915e-06, "loss": 0.057, "step": 5211 }, { "epoch": 1.1858930602957907, "grad_norm": 1.271479589500172, "learning_rate": 1.0844402681292144e-06, "loss": 0.0369, "step": 5212 }, { "epoch": 1.1861205915813424, "grad_norm": 1.0411295086764045, "learning_rate": 1.084379687354619e-06, "loss": 0.0097, "step": 5213 }, { "epoch": 1.1863481228668942, "grad_norm": 2.6559141117447225, "learning_rate": 1.0843190971911447e-06, "loss": 0.0228, "step": 5214 }, { "epoch": 1.186575654152446, "grad_norm": 1.5202925528678082, "learning_rate": 1.0842584976400292e-06, "loss": 0.0338, "step": 5215 }, { "epoch": 1.1868031854379977, "grad_norm": 1.1093380023536548, "learning_rate": 1.084197888702511e-06, "loss": 0.0168, "step": 5216 }, { "epoch": 1.1870307167235494, "grad_norm": 1.5307787654530016, "learning_rate": 1.0841372703798289e-06, "loss": 0.0278, "step": 5217 }, { "epoch": 1.1872582480091012, "grad_norm": 1.3322857454820025, "learning_rate": 1.0840766426732219e-06, "loss": 0.0347, "step": 5218 }, { "epoch": 1.187485779294653, "grad_norm": 0.8608991873903781, "learning_rate": 1.0840160055839291e-06, "loss": 0.0238, "step": 5219 }, { "epoch": 1.1877133105802047, "grad_norm": 1.119129138655594, "learning_rate": 1.08395535911319e-06, "loss": 0.0131, "step": 5220 }, { "epoch": 1.1879408418657564, "grad_norm": 0.9520127980932701, "learning_rate": 1.0838947032622436e-06, "loss": 0.0187, "step": 5221 }, { "epoch": 1.1881683731513082, "grad_norm": 1.0013994149687084, "learning_rate": 1.0838340380323297e-06, "loss": 0.0217, "step": 5222 }, { "epoch": 1.1883959044368602, "grad_norm": 2.3034636462642166, "learning_rate": 1.083773363424689e-06, "loss": 0.0253, "step": 5223 }, { "epoch": 1.188623435722412, "grad_norm": 0.9604142125712335, "learning_rate": 1.0837126794405603e-06, "loss": 0.0304, "step": 5224 }, { "epoch": 1.1888509670079637, "grad_norm": 0.9680611296826206, "learning_rate": 1.083651986081185e-06, "loss": 0.0233, "step": 5225 }, { "epoch": 1.1890784982935154, "grad_norm": 1.1564232624147677, "learning_rate": 1.0835912833478029e-06, "loss": 0.0514, "step": 5226 }, { "epoch": 1.1893060295790672, "grad_norm": 1.0708217476839645, "learning_rate": 1.0835305712416546e-06, "loss": 0.0244, "step": 5227 }, { "epoch": 1.189533560864619, "grad_norm": 1.1356456632073215, "learning_rate": 1.0834698497639817e-06, "loss": 0.0453, "step": 5228 }, { "epoch": 1.1897610921501707, "grad_norm": 0.9643109850754426, "learning_rate": 1.0834091189160243e-06, "loss": 0.0148, "step": 5229 }, { "epoch": 1.1899886234357224, "grad_norm": 1.4647639680120725, "learning_rate": 1.0833483786990243e-06, "loss": 0.0398, "step": 5230 }, { "epoch": 1.1902161547212742, "grad_norm": 0.6967268185621851, "learning_rate": 1.0832876291142228e-06, "loss": 0.0153, "step": 5231 }, { "epoch": 1.190443686006826, "grad_norm": 1.5226277581599168, "learning_rate": 1.0832268701628613e-06, "loss": 0.0387, "step": 5232 }, { "epoch": 1.1906712172923777, "grad_norm": 1.5725234513516795, "learning_rate": 1.083166101846182e-06, "loss": 0.059, "step": 5233 }, { "epoch": 1.1908987485779294, "grad_norm": 1.0092630282850967, "learning_rate": 1.0831053241654265e-06, "loss": 0.019, "step": 5234 }, { "epoch": 1.1911262798634812, "grad_norm": 1.1102622210079196, "learning_rate": 1.0830445371218374e-06, "loss": 0.0234, "step": 5235 }, { "epoch": 1.191353811149033, "grad_norm": 1.2854110941773407, "learning_rate": 1.0829837407166565e-06, "loss": 0.0243, "step": 5236 }, { "epoch": 1.1915813424345847, "grad_norm": 0.38016640183110134, "learning_rate": 1.082922934951127e-06, "loss": 0.0134, "step": 5237 }, { "epoch": 1.1918088737201364, "grad_norm": 1.168241176946775, "learning_rate": 1.082862119826491e-06, "loss": 0.031, "step": 5238 }, { "epoch": 1.1920364050056882, "grad_norm": 1.1960841025032027, "learning_rate": 1.0828012953439921e-06, "loss": 0.052, "step": 5239 }, { "epoch": 1.1922639362912402, "grad_norm": 0.7158528800285552, "learning_rate": 1.082740461504873e-06, "loss": 0.011, "step": 5240 }, { "epoch": 1.192491467576792, "grad_norm": 1.1123900753579041, "learning_rate": 1.0826796183103774e-06, "loss": 0.0442, "step": 5241 }, { "epoch": 1.1927189988623437, "grad_norm": 1.3167322420939016, "learning_rate": 1.0826187657617484e-06, "loss": 0.0351, "step": 5242 }, { "epoch": 1.1929465301478954, "grad_norm": 1.8256977932069058, "learning_rate": 1.0825579038602298e-06, "loss": 0.0228, "step": 5243 }, { "epoch": 1.1931740614334472, "grad_norm": 0.8153271667842001, "learning_rate": 1.082497032607066e-06, "loss": 0.0218, "step": 5244 }, { "epoch": 1.193401592718999, "grad_norm": 1.13052228865416, "learning_rate": 1.0824361520035004e-06, "loss": 0.0269, "step": 5245 }, { "epoch": 1.1936291240045507, "grad_norm": 0.8769614800574006, "learning_rate": 1.0823752620507778e-06, "loss": 0.0196, "step": 5246 }, { "epoch": 1.1938566552901024, "grad_norm": 2.442105507702053, "learning_rate": 1.0823143627501423e-06, "loss": 0.0223, "step": 5247 }, { "epoch": 1.1940841865756542, "grad_norm": 0.9690120767439246, "learning_rate": 1.082253454102839e-06, "loss": 0.0293, "step": 5248 }, { "epoch": 1.194311717861206, "grad_norm": 0.9656302532397552, "learning_rate": 1.0821925361101124e-06, "loss": 0.0194, "step": 5249 }, { "epoch": 1.1945392491467577, "grad_norm": 1.2220527685129399, "learning_rate": 1.0821316087732075e-06, "loss": 0.0361, "step": 5250 }, { "epoch": 1.1947667804323094, "grad_norm": 1.1131169147903066, "learning_rate": 1.0820706720933698e-06, "loss": 0.0282, "step": 5251 }, { "epoch": 1.1949943117178612, "grad_norm": 1.1331333681292557, "learning_rate": 1.0820097260718448e-06, "loss": 0.0117, "step": 5252 }, { "epoch": 1.195221843003413, "grad_norm": 1.126019194986974, "learning_rate": 1.081948770709878e-06, "loss": 0.0142, "step": 5253 }, { "epoch": 1.1954493742889647, "grad_norm": 1.9300596010130329, "learning_rate": 1.0818878060087151e-06, "loss": 0.0305, "step": 5254 }, { "epoch": 1.1956769055745164, "grad_norm": 3.8484100820913096, "learning_rate": 1.081826831969602e-06, "loss": 0.0175, "step": 5255 }, { "epoch": 1.1959044368600682, "grad_norm": 2.6177072747321493, "learning_rate": 1.0817658485937855e-06, "loss": 0.0356, "step": 5256 }, { "epoch": 1.19613196814562, "grad_norm": 1.8647005779081272, "learning_rate": 1.0817048558825114e-06, "loss": 0.0416, "step": 5257 }, { "epoch": 1.1963594994311717, "grad_norm": 1.8201241368258048, "learning_rate": 1.0816438538370262e-06, "loss": 0.03, "step": 5258 }, { "epoch": 1.1965870307167235, "grad_norm": 1.309838446791344, "learning_rate": 1.0815828424585772e-06, "loss": 0.0221, "step": 5259 }, { "epoch": 1.1968145620022752, "grad_norm": 2.2126363422898736, "learning_rate": 1.081521821748411e-06, "loss": 0.01, "step": 5260 }, { "epoch": 1.197042093287827, "grad_norm": 1.5225649498436484, "learning_rate": 1.081460791707775e-06, "loss": 0.0106, "step": 5261 }, { "epoch": 1.197269624573379, "grad_norm": 2.926124794782369, "learning_rate": 1.0813997523379163e-06, "loss": 0.0545, "step": 5262 }, { "epoch": 1.1974971558589307, "grad_norm": 3.121781275137615, "learning_rate": 1.0813387036400825e-06, "loss": 0.039, "step": 5263 }, { "epoch": 1.1977246871444824, "grad_norm": 2.991132032943111, "learning_rate": 1.081277645615521e-06, "loss": 0.0217, "step": 5264 }, { "epoch": 1.1979522184300342, "grad_norm": 0.7510540265715727, "learning_rate": 1.0812165782654806e-06, "loss": 0.0113, "step": 5265 }, { "epoch": 1.198179749715586, "grad_norm": 0.9442454590864651, "learning_rate": 1.0811555015912086e-06, "loss": 0.0128, "step": 5266 }, { "epoch": 1.1984072810011377, "grad_norm": 1.0369867334237874, "learning_rate": 1.0810944155939536e-06, "loss": 0.0289, "step": 5267 }, { "epoch": 1.1986348122866894, "grad_norm": 1.2893237225891867, "learning_rate": 1.081033320274964e-06, "loss": 0.0225, "step": 5268 }, { "epoch": 1.1988623435722412, "grad_norm": 1.4348798625399937, "learning_rate": 1.0809722156354884e-06, "loss": 0.0196, "step": 5269 }, { "epoch": 1.199089874857793, "grad_norm": 0.7783433954066992, "learning_rate": 1.0809111016767762e-06, "loss": 0.0132, "step": 5270 }, { "epoch": 1.1993174061433447, "grad_norm": 2.29125273839501, "learning_rate": 1.0808499784000756e-06, "loss": 0.0155, "step": 5271 }, { "epoch": 1.1995449374288965, "grad_norm": 1.8766245713656982, "learning_rate": 1.0807888458066364e-06, "loss": 0.0405, "step": 5272 }, { "epoch": 1.1997724687144482, "grad_norm": 2.0497369482648997, "learning_rate": 1.0807277038977083e-06, "loss": 0.0483, "step": 5273 }, { "epoch": 1.2, "grad_norm": 1.3799009247077467, "learning_rate": 1.0806665526745403e-06, "loss": 0.024, "step": 5274 }, { "epoch": 1.2002275312855517, "grad_norm": 0.9219092351160844, "learning_rate": 1.0806053921383823e-06, "loss": 0.0206, "step": 5275 }, { "epoch": 1.2004550625711035, "grad_norm": 0.7062087076877579, "learning_rate": 1.0805442222904846e-06, "loss": 0.0197, "step": 5276 }, { "epoch": 1.2006825938566552, "grad_norm": 1.1815501920868743, "learning_rate": 1.0804830431320972e-06, "loss": 0.0311, "step": 5277 }, { "epoch": 1.200910125142207, "grad_norm": 1.4194267958205675, "learning_rate": 1.0804218546644708e-06, "loss": 0.0269, "step": 5278 }, { "epoch": 1.201137656427759, "grad_norm": 1.5374413742413802, "learning_rate": 1.0803606568888557e-06, "loss": 0.0371, "step": 5279 }, { "epoch": 1.2013651877133107, "grad_norm": 1.4162269983170825, "learning_rate": 1.0802994498065027e-06, "loss": 0.0433, "step": 5280 }, { "epoch": 1.2015927189988624, "grad_norm": 1.3982466009914463, "learning_rate": 1.0802382334186627e-06, "loss": 0.02, "step": 5281 }, { "epoch": 1.2018202502844142, "grad_norm": 1.0353374922029346, "learning_rate": 1.080177007726587e-06, "loss": 0.0314, "step": 5282 }, { "epoch": 1.202047781569966, "grad_norm": 1.4144577926145716, "learning_rate": 1.080115772731527e-06, "loss": 0.0237, "step": 5283 }, { "epoch": 1.2022753128555177, "grad_norm": 1.270246359088111, "learning_rate": 1.0800545284347338e-06, "loss": 0.0278, "step": 5284 }, { "epoch": 1.2025028441410694, "grad_norm": 1.7811867988957832, "learning_rate": 1.0799932748374598e-06, "loss": 0.0351, "step": 5285 }, { "epoch": 1.2027303754266212, "grad_norm": 0.811331257305328, "learning_rate": 1.0799320119409562e-06, "loss": 0.0154, "step": 5286 }, { "epoch": 1.202957906712173, "grad_norm": 1.326967549928065, "learning_rate": 1.0798707397464756e-06, "loss": 0.0251, "step": 5287 }, { "epoch": 1.2031854379977247, "grad_norm": 1.2668064811271582, "learning_rate": 1.0798094582552703e-06, "loss": 0.0214, "step": 5288 }, { "epoch": 1.2034129692832765, "grad_norm": 1.7316590095255162, "learning_rate": 1.0797481674685925e-06, "loss": 0.0595, "step": 5289 }, { "epoch": 1.2036405005688282, "grad_norm": 0.8626668121425645, "learning_rate": 1.0796868673876947e-06, "loss": 0.0138, "step": 5290 }, { "epoch": 1.20386803185438, "grad_norm": 1.688909270364021, "learning_rate": 1.0796255580138303e-06, "loss": 0.0376, "step": 5291 }, { "epoch": 1.2040955631399317, "grad_norm": 1.3043440282984107, "learning_rate": 1.0795642393482523e-06, "loss": 0.0511, "step": 5292 }, { "epoch": 1.2043230944254835, "grad_norm": 1.2730397806370202, "learning_rate": 1.0795029113922136e-06, "loss": 0.0232, "step": 5293 }, { "epoch": 1.2045506257110352, "grad_norm": 0.5816881949668542, "learning_rate": 1.0794415741469677e-06, "loss": 0.0086, "step": 5294 }, { "epoch": 1.204778156996587, "grad_norm": 1.0835756813832158, "learning_rate": 1.0793802276137683e-06, "loss": 0.0204, "step": 5295 }, { "epoch": 1.2050056882821387, "grad_norm": 1.8276533196679292, "learning_rate": 1.0793188717938693e-06, "loss": 0.0304, "step": 5296 }, { "epoch": 1.2052332195676905, "grad_norm": 2.167032474707277, "learning_rate": 1.0792575066885245e-06, "loss": 0.0331, "step": 5297 }, { "epoch": 1.2054607508532422, "grad_norm": 0.7493889811458411, "learning_rate": 1.0791961322989882e-06, "loss": 0.0147, "step": 5298 }, { "epoch": 1.205688282138794, "grad_norm": 1.329356172948289, "learning_rate": 1.0791347486265147e-06, "loss": 0.0353, "step": 5299 }, { "epoch": 1.2059158134243457, "grad_norm": 0.8797921570409813, "learning_rate": 1.0790733556723589e-06, "loss": 0.0242, "step": 5300 }, { "epoch": 1.2061433447098977, "grad_norm": 0.8299000303047913, "learning_rate": 1.079011953437775e-06, "loss": 0.0131, "step": 5301 }, { "epoch": 1.2063708759954495, "grad_norm": 0.5695113688949743, "learning_rate": 1.0789505419240185e-06, "loss": 0.0091, "step": 5302 }, { "epoch": 1.2065984072810012, "grad_norm": 1.007416588849143, "learning_rate": 1.0788891211323442e-06, "loss": 0.0211, "step": 5303 }, { "epoch": 1.206825938566553, "grad_norm": 1.0207685534429214, "learning_rate": 1.0788276910640074e-06, "loss": 0.0167, "step": 5304 }, { "epoch": 1.2070534698521047, "grad_norm": 0.9531066995071166, "learning_rate": 1.0787662517202641e-06, "loss": 0.0264, "step": 5305 }, { "epoch": 1.2072810011376565, "grad_norm": 0.8593609085890438, "learning_rate": 1.0787048031023693e-06, "loss": 0.0192, "step": 5306 }, { "epoch": 1.2075085324232082, "grad_norm": 0.7570873924698162, "learning_rate": 1.0786433452115794e-06, "loss": 0.0107, "step": 5307 }, { "epoch": 1.20773606370876, "grad_norm": 0.6192632652453488, "learning_rate": 1.0785818780491502e-06, "loss": 0.01, "step": 5308 }, { "epoch": 1.2079635949943117, "grad_norm": 1.1762597971964037, "learning_rate": 1.0785204016163384e-06, "loss": 0.017, "step": 5309 }, { "epoch": 1.2081911262798635, "grad_norm": 0.8264299611729337, "learning_rate": 1.0784589159143999e-06, "loss": 0.0137, "step": 5310 }, { "epoch": 1.2084186575654152, "grad_norm": 0.5022797081392271, "learning_rate": 1.0783974209445915e-06, "loss": 0.0069, "step": 5311 }, { "epoch": 1.208646188850967, "grad_norm": 0.6561114668173221, "learning_rate": 1.0783359167081705e-06, "loss": 0.0178, "step": 5312 }, { "epoch": 1.2088737201365187, "grad_norm": 1.860198594863872, "learning_rate": 1.0782744032063935e-06, "loss": 0.0395, "step": 5313 }, { "epoch": 1.2091012514220705, "grad_norm": 1.2557407515136827, "learning_rate": 1.078212880440518e-06, "loss": 0.0198, "step": 5314 }, { "epoch": 1.2093287827076222, "grad_norm": 0.9774237973356178, "learning_rate": 1.0781513484118008e-06, "loss": 0.033, "step": 5315 }, { "epoch": 1.209556313993174, "grad_norm": 1.1404558362365051, "learning_rate": 1.0780898071215004e-06, "loss": 0.0153, "step": 5316 }, { "epoch": 1.2097838452787257, "grad_norm": 1.2711927135952792, "learning_rate": 1.0780282565708738e-06, "loss": 0.0452, "step": 5317 }, { "epoch": 1.2100113765642777, "grad_norm": 1.572981203889357, "learning_rate": 1.0779666967611796e-06, "loss": 0.0366, "step": 5318 }, { "epoch": 1.2102389078498295, "grad_norm": 1.0432053268711592, "learning_rate": 1.0779051276936755e-06, "loss": 0.0193, "step": 5319 }, { "epoch": 1.2104664391353812, "grad_norm": 1.108281271730141, "learning_rate": 1.0778435493696202e-06, "loss": 0.0514, "step": 5320 }, { "epoch": 1.210693970420933, "grad_norm": 1.0307341066968732, "learning_rate": 1.0777819617902718e-06, "loss": 0.0232, "step": 5321 }, { "epoch": 1.2109215017064847, "grad_norm": 1.1269375808316393, "learning_rate": 1.0777203649568896e-06, "loss": 0.0111, "step": 5322 }, { "epoch": 1.2111490329920365, "grad_norm": 0.8681106412081948, "learning_rate": 1.077658758870732e-06, "loss": 0.0134, "step": 5323 }, { "epoch": 1.2113765642775882, "grad_norm": 1.6310011066128904, "learning_rate": 1.0775971435330588e-06, "loss": 0.0443, "step": 5324 }, { "epoch": 1.21160409556314, "grad_norm": 1.9800926111893868, "learning_rate": 1.0775355189451286e-06, "loss": 0.0491, "step": 5325 }, { "epoch": 1.2118316268486917, "grad_norm": 1.0735676656079027, "learning_rate": 1.0774738851082011e-06, "loss": 0.0137, "step": 5326 }, { "epoch": 1.2120591581342435, "grad_norm": 1.4112445459881595, "learning_rate": 1.0774122420235363e-06, "loss": 0.0267, "step": 5327 }, { "epoch": 1.2122866894197952, "grad_norm": 1.6264263710804632, "learning_rate": 1.0773505896923936e-06, "loss": 0.0601, "step": 5328 }, { "epoch": 1.212514220705347, "grad_norm": 1.4435922348969275, "learning_rate": 1.0772889281160335e-06, "loss": 0.0487, "step": 5329 }, { "epoch": 1.2127417519908987, "grad_norm": 1.6845252736050549, "learning_rate": 1.0772272572957158e-06, "loss": 0.0384, "step": 5330 }, { "epoch": 1.2129692832764505, "grad_norm": 0.620039095627728, "learning_rate": 1.0771655772327013e-06, "loss": 0.0087, "step": 5331 }, { "epoch": 1.2131968145620022, "grad_norm": 1.0155486269794136, "learning_rate": 1.0771038879282505e-06, "loss": 0.0316, "step": 5332 }, { "epoch": 1.213424345847554, "grad_norm": 1.0720369582435914, "learning_rate": 1.0770421893836243e-06, "loss": 0.0297, "step": 5333 }, { "epoch": 1.2136518771331057, "grad_norm": 1.0641000248930415, "learning_rate": 1.0769804816000835e-06, "loss": 0.0228, "step": 5334 }, { "epoch": 1.2138794084186575, "grad_norm": 1.3185310795364764, "learning_rate": 1.0769187645788895e-06, "loss": 0.0325, "step": 5335 }, { "epoch": 1.2141069397042092, "grad_norm": 0.8019919288175986, "learning_rate": 1.0768570383213035e-06, "loss": 0.0108, "step": 5336 }, { "epoch": 1.214334470989761, "grad_norm": 1.0171180200964656, "learning_rate": 1.0767953028285872e-06, "loss": 0.0162, "step": 5337 }, { "epoch": 1.2145620022753127, "grad_norm": 0.6898326225080136, "learning_rate": 1.0767335581020024e-06, "loss": 0.0105, "step": 5338 }, { "epoch": 1.2147895335608645, "grad_norm": 0.5713039580718985, "learning_rate": 1.076671804142811e-06, "loss": 0.0077, "step": 5339 }, { "epoch": 1.2150170648464165, "grad_norm": 1.2459220754175229, "learning_rate": 1.076610040952275e-06, "loss": 0.0332, "step": 5340 }, { "epoch": 1.2152445961319682, "grad_norm": 1.1187776697921044, "learning_rate": 1.076548268531657e-06, "loss": 0.0191, "step": 5341 }, { "epoch": 1.21547212741752, "grad_norm": 2.0317760107761704, "learning_rate": 1.0764864868822194e-06, "loss": 0.0516, "step": 5342 }, { "epoch": 1.2156996587030717, "grad_norm": 1.9233690667339316, "learning_rate": 1.0764246960052247e-06, "loss": 0.0134, "step": 5343 }, { "epoch": 1.2159271899886235, "grad_norm": 1.6064839097939867, "learning_rate": 1.0763628959019359e-06, "loss": 0.0209, "step": 5344 }, { "epoch": 1.2161547212741752, "grad_norm": 1.2879343782626302, "learning_rate": 1.076301086573616e-06, "loss": 0.016, "step": 5345 }, { "epoch": 1.216382252559727, "grad_norm": 1.1521103656454588, "learning_rate": 1.076239268021529e-06, "loss": 0.0294, "step": 5346 }, { "epoch": 1.2166097838452787, "grad_norm": 0.7645894367138184, "learning_rate": 1.0761774402469375e-06, "loss": 0.0146, "step": 5347 }, { "epoch": 1.2168373151308305, "grad_norm": 1.0900116364078585, "learning_rate": 1.0761156032511052e-06, "loss": 0.0148, "step": 5348 }, { "epoch": 1.2170648464163822, "grad_norm": 1.0426468341641788, "learning_rate": 1.0760537570352963e-06, "loss": 0.0205, "step": 5349 }, { "epoch": 1.217292377701934, "grad_norm": 0.968560865785141, "learning_rate": 1.0759919016007747e-06, "loss": 0.028, "step": 5350 }, { "epoch": 1.2175199089874857, "grad_norm": 1.3908329783187319, "learning_rate": 1.0759300369488046e-06, "loss": 0.0311, "step": 5351 }, { "epoch": 1.2177474402730375, "grad_norm": 0.9336674383856225, "learning_rate": 1.0758681630806502e-06, "loss": 0.0126, "step": 5352 }, { "epoch": 1.2179749715585892, "grad_norm": 0.9822161525178283, "learning_rate": 1.0758062799975765e-06, "loss": 0.0328, "step": 5353 }, { "epoch": 1.218202502844141, "grad_norm": 1.56822129701628, "learning_rate": 1.075744387700848e-06, "loss": 0.0543, "step": 5354 }, { "epoch": 1.2184300341296928, "grad_norm": 0.8508512043785544, "learning_rate": 1.0756824861917294e-06, "loss": 0.0186, "step": 5355 }, { "epoch": 1.2186575654152445, "grad_norm": 1.6471064921915508, "learning_rate": 1.0756205754714866e-06, "loss": 0.0419, "step": 5356 }, { "epoch": 1.2188850967007965, "grad_norm": 0.9710323188808565, "learning_rate": 1.0755586555413845e-06, "loss": 0.0165, "step": 5357 }, { "epoch": 1.2191126279863482, "grad_norm": 1.1740838398504168, "learning_rate": 1.0754967264026882e-06, "loss": 0.0171, "step": 5358 }, { "epoch": 1.2193401592719, "grad_norm": 0.821398345986984, "learning_rate": 1.0754347880566643e-06, "loss": 0.0144, "step": 5359 }, { "epoch": 1.2195676905574517, "grad_norm": 0.737155613985261, "learning_rate": 1.075372840504578e-06, "loss": 0.0215, "step": 5360 }, { "epoch": 1.2197952218430035, "grad_norm": 1.2248403884170755, "learning_rate": 1.0753108837476958e-06, "loss": 0.023, "step": 5361 }, { "epoch": 1.2200227531285552, "grad_norm": 1.5213992441300637, "learning_rate": 1.0752489177872839e-06, "loss": 0.0211, "step": 5362 }, { "epoch": 1.220250284414107, "grad_norm": 0.9276211966706847, "learning_rate": 1.0751869426246086e-06, "loss": 0.0248, "step": 5363 }, { "epoch": 1.2204778156996587, "grad_norm": 0.8339386449394393, "learning_rate": 1.0751249582609368e-06, "loss": 0.0121, "step": 5364 }, { "epoch": 1.2207053469852105, "grad_norm": 0.814490781430545, "learning_rate": 1.075062964697535e-06, "loss": 0.0169, "step": 5365 }, { "epoch": 1.2209328782707622, "grad_norm": 1.368386097389221, "learning_rate": 1.0750009619356706e-06, "loss": 0.0406, "step": 5366 }, { "epoch": 1.221160409556314, "grad_norm": 1.0996742222363038, "learning_rate": 1.0749389499766106e-06, "loss": 0.0259, "step": 5367 }, { "epoch": 1.2213879408418657, "grad_norm": 1.468334437179015, "learning_rate": 1.0748769288216226e-06, "loss": 0.0272, "step": 5368 }, { "epoch": 1.2216154721274175, "grad_norm": 1.3908865445482075, "learning_rate": 1.074814898471974e-06, "loss": 0.0461, "step": 5369 }, { "epoch": 1.2218430034129693, "grad_norm": 1.3737903263713396, "learning_rate": 1.0747528589289327e-06, "loss": 0.0109, "step": 5370 }, { "epoch": 1.222070534698521, "grad_norm": 0.9334616229837778, "learning_rate": 1.0746908101937666e-06, "loss": 0.0225, "step": 5371 }, { "epoch": 1.2222980659840728, "grad_norm": 3.4821828449919656, "learning_rate": 1.0746287522677439e-06, "loss": 0.1442, "step": 5372 }, { "epoch": 1.2225255972696245, "grad_norm": 1.1612571872192434, "learning_rate": 1.074566685152133e-06, "loss": 0.024, "step": 5373 }, { "epoch": 1.2227531285551763, "grad_norm": 0.6082887587191537, "learning_rate": 1.0745046088482025e-06, "loss": 0.0091, "step": 5374 }, { "epoch": 1.222980659840728, "grad_norm": 0.8947279579511637, "learning_rate": 1.074442523357221e-06, "loss": 0.018, "step": 5375 }, { "epoch": 1.2232081911262798, "grad_norm": 0.8578464895598875, "learning_rate": 1.0743804286804573e-06, "loss": 0.0083, "step": 5376 }, { "epoch": 1.2234357224118315, "grad_norm": 0.9832599316516126, "learning_rate": 1.0743183248191806e-06, "loss": 0.0112, "step": 5377 }, { "epoch": 1.2236632536973833, "grad_norm": 0.9942384197604868, "learning_rate": 1.0742562117746604e-06, "loss": 0.0266, "step": 5378 }, { "epoch": 1.2238907849829352, "grad_norm": 0.907525459976151, "learning_rate": 1.074194089548166e-06, "loss": 0.0182, "step": 5379 }, { "epoch": 1.224118316268487, "grad_norm": 0.6044100806377944, "learning_rate": 1.0741319581409667e-06, "loss": 0.0132, "step": 5380 }, { "epoch": 1.2243458475540387, "grad_norm": 0.7851876560740085, "learning_rate": 1.0740698175543332e-06, "loss": 0.008, "step": 5381 }, { "epoch": 1.2245733788395905, "grad_norm": 1.1184528122514146, "learning_rate": 1.0740076677895348e-06, "loss": 0.0295, "step": 5382 }, { "epoch": 1.2248009101251423, "grad_norm": 1.0110121442428808, "learning_rate": 1.0739455088478422e-06, "loss": 0.0246, "step": 5383 }, { "epoch": 1.225028441410694, "grad_norm": 1.3430003631618936, "learning_rate": 1.0738833407305254e-06, "loss": 0.0311, "step": 5384 }, { "epoch": 1.2252559726962458, "grad_norm": 0.714109028797083, "learning_rate": 1.0738211634388554e-06, "loss": 0.0137, "step": 5385 }, { "epoch": 1.2254835039817975, "grad_norm": 1.7977112803389215, "learning_rate": 1.0737589769741025e-06, "loss": 0.0311, "step": 5386 }, { "epoch": 1.2257110352673493, "grad_norm": 1.595010039888779, "learning_rate": 1.0736967813375382e-06, "loss": 0.0264, "step": 5387 }, { "epoch": 1.225938566552901, "grad_norm": 0.9599396412918755, "learning_rate": 1.0736345765304335e-06, "loss": 0.0183, "step": 5388 }, { "epoch": 1.2261660978384528, "grad_norm": 1.4002183993615684, "learning_rate": 1.0735723625540596e-06, "loss": 0.0584, "step": 5389 }, { "epoch": 1.2263936291240045, "grad_norm": 0.5289624966838866, "learning_rate": 1.073510139409688e-06, "loss": 0.0101, "step": 5390 }, { "epoch": 1.2266211604095563, "grad_norm": 1.057508406840753, "learning_rate": 1.0734479070985908e-06, "loss": 0.0142, "step": 5391 }, { "epoch": 1.226848691695108, "grad_norm": 0.7921993020441022, "learning_rate": 1.0733856656220396e-06, "loss": 0.0111, "step": 5392 }, { "epoch": 1.2270762229806598, "grad_norm": 0.7881195760466057, "learning_rate": 1.0733234149813065e-06, "loss": 0.0251, "step": 5393 }, { "epoch": 1.2273037542662115, "grad_norm": 0.7174133775125424, "learning_rate": 1.0732611551776639e-06, "loss": 0.0123, "step": 5394 }, { "epoch": 1.2275312855517633, "grad_norm": 1.2047358857096602, "learning_rate": 1.0731988862123841e-06, "loss": 0.012, "step": 5395 }, { "epoch": 1.2277588168373152, "grad_norm": 0.8067610993565308, "learning_rate": 1.07313660808674e-06, "loss": 0.0093, "step": 5396 }, { "epoch": 1.227986348122867, "grad_norm": 0.9414182769464757, "learning_rate": 1.0730743208020044e-06, "loss": 0.0138, "step": 5397 }, { "epoch": 1.2282138794084188, "grad_norm": 1.0848537565553817, "learning_rate": 1.0730120243594504e-06, "loss": 0.0235, "step": 5398 }, { "epoch": 1.2284414106939705, "grad_norm": 1.0189138352280012, "learning_rate": 1.0729497187603508e-06, "loss": 0.0264, "step": 5399 }, { "epoch": 1.2286689419795223, "grad_norm": 1.0924034497075088, "learning_rate": 1.0728874040059798e-06, "loss": 0.0115, "step": 5400 }, { "epoch": 1.228896473265074, "grad_norm": 0.6433664647544001, "learning_rate": 1.07282508009761e-06, "loss": 0.01, "step": 5401 }, { "epoch": 1.2291240045506258, "grad_norm": 0.7843372574416333, "learning_rate": 1.072762747036516e-06, "loss": 0.0215, "step": 5402 }, { "epoch": 1.2293515358361775, "grad_norm": 1.0890903190836254, "learning_rate": 1.0727004048239715e-06, "loss": 0.0272, "step": 5403 }, { "epoch": 1.2295790671217293, "grad_norm": 0.8996000171907133, "learning_rate": 1.0726380534612507e-06, "loss": 0.0171, "step": 5404 }, { "epoch": 1.229806598407281, "grad_norm": 0.8250699810825667, "learning_rate": 1.0725756929496277e-06, "loss": 0.007, "step": 5405 }, { "epoch": 1.2300341296928328, "grad_norm": 1.0264353557572934, "learning_rate": 1.0725133232903773e-06, "loss": 0.0345, "step": 5406 }, { "epoch": 1.2302616609783845, "grad_norm": 1.137616173296977, "learning_rate": 1.0724509444847741e-06, "loss": 0.036, "step": 5407 }, { "epoch": 1.2304891922639363, "grad_norm": 1.3664447826228183, "learning_rate": 1.0723885565340933e-06, "loss": 0.0285, "step": 5408 }, { "epoch": 1.230716723549488, "grad_norm": 1.1602041449374105, "learning_rate": 1.0723261594396095e-06, "loss": 0.0332, "step": 5409 }, { "epoch": 1.2309442548350398, "grad_norm": 1.090010681276093, "learning_rate": 1.0722637532025984e-06, "loss": 0.0218, "step": 5410 }, { "epoch": 1.2311717861205915, "grad_norm": 1.4413107732179307, "learning_rate": 1.0722013378243354e-06, "loss": 0.0321, "step": 5411 }, { "epoch": 1.2313993174061433, "grad_norm": 1.8979249234317928, "learning_rate": 1.0721389133060958e-06, "loss": 0.0266, "step": 5412 }, { "epoch": 1.231626848691695, "grad_norm": 1.7124518471345664, "learning_rate": 1.0720764796491559e-06, "loss": 0.0281, "step": 5413 }, { "epoch": 1.2318543799772468, "grad_norm": 1.466076682039334, "learning_rate": 1.0720140368547915e-06, "loss": 0.0445, "step": 5414 }, { "epoch": 1.2320819112627985, "grad_norm": 1.307580416194323, "learning_rate": 1.0719515849242787e-06, "loss": 0.0233, "step": 5415 }, { "epoch": 1.2323094425483503, "grad_norm": 0.8318083081384708, "learning_rate": 1.0718891238588943e-06, "loss": 0.022, "step": 5416 }, { "epoch": 1.232536973833902, "grad_norm": 1.3996049895553466, "learning_rate": 1.0718266536599145e-06, "loss": 0.046, "step": 5417 }, { "epoch": 1.232764505119454, "grad_norm": 1.4470303584385065, "learning_rate": 1.0717641743286163e-06, "loss": 0.0315, "step": 5418 }, { "epoch": 1.2329920364050058, "grad_norm": 1.251825573353438, "learning_rate": 1.0717016858662766e-06, "loss": 0.0187, "step": 5419 }, { "epoch": 1.2332195676905575, "grad_norm": 1.2025353214638472, "learning_rate": 1.0716391882741722e-06, "loss": 0.0512, "step": 5420 }, { "epoch": 1.2334470989761093, "grad_norm": 1.1125647041962026, "learning_rate": 1.0715766815535812e-06, "loss": 0.0191, "step": 5421 }, { "epoch": 1.233674630261661, "grad_norm": 1.0140086609338004, "learning_rate": 1.0715141657057805e-06, "loss": 0.0286, "step": 5422 }, { "epoch": 1.2339021615472128, "grad_norm": 1.6413125694512114, "learning_rate": 1.0714516407320482e-06, "loss": 0.0277, "step": 5423 }, { "epoch": 1.2341296928327645, "grad_norm": 1.113912252435204, "learning_rate": 1.0713891066336619e-06, "loss": 0.0459, "step": 5424 }, { "epoch": 1.2343572241183163, "grad_norm": 1.4211151899069139, "learning_rate": 1.0713265634118998e-06, "loss": 0.03, "step": 5425 }, { "epoch": 1.234584755403868, "grad_norm": 0.5439788203180792, "learning_rate": 1.0712640110680398e-06, "loss": 0.0051, "step": 5426 }, { "epoch": 1.2348122866894198, "grad_norm": 1.1066572512164194, "learning_rate": 1.071201449603361e-06, "loss": 0.031, "step": 5427 }, { "epoch": 1.2350398179749715, "grad_norm": 1.0734688953953, "learning_rate": 1.0711388790191418e-06, "loss": 0.0159, "step": 5428 }, { "epoch": 1.2352673492605233, "grad_norm": 0.9674963008176792, "learning_rate": 1.071076299316661e-06, "loss": 0.0292, "step": 5429 }, { "epoch": 1.235494880546075, "grad_norm": 0.8648315211359087, "learning_rate": 1.0710137104971973e-06, "loss": 0.015, "step": 5430 }, { "epoch": 1.2357224118316268, "grad_norm": 1.0177696413333797, "learning_rate": 1.0709511125620306e-06, "loss": 0.0187, "step": 5431 }, { "epoch": 1.2359499431171785, "grad_norm": 1.080479875370983, "learning_rate": 1.0708885055124396e-06, "loss": 0.0202, "step": 5432 }, { "epoch": 1.2361774744027303, "grad_norm": 1.2213741714117063, "learning_rate": 1.0708258893497043e-06, "loss": 0.0347, "step": 5433 }, { "epoch": 1.236405005688282, "grad_norm": 0.6102850699242401, "learning_rate": 1.0707632640751042e-06, "loss": 0.011, "step": 5434 }, { "epoch": 1.236632536973834, "grad_norm": 0.7162928033199788, "learning_rate": 1.0707006296899194e-06, "loss": 0.0104, "step": 5435 }, { "epoch": 1.2368600682593858, "grad_norm": 1.6643057799440983, "learning_rate": 1.0706379861954299e-06, "loss": 0.0474, "step": 5436 }, { "epoch": 1.2370875995449375, "grad_norm": 1.0636970768259353, "learning_rate": 1.0705753335929162e-06, "loss": 0.0441, "step": 5437 }, { "epoch": 1.2373151308304893, "grad_norm": 0.5899981023336623, "learning_rate": 1.070512671883659e-06, "loss": 0.0081, "step": 5438 }, { "epoch": 1.237542662116041, "grad_norm": 1.1358807681666672, "learning_rate": 1.0704500010689383e-06, "loss": 0.0231, "step": 5439 }, { "epoch": 1.2377701934015928, "grad_norm": 1.483924786030886, "learning_rate": 1.0703873211500356e-06, "loss": 0.0434, "step": 5440 }, { "epoch": 1.2379977246871445, "grad_norm": 0.5564175099477741, "learning_rate": 1.0703246321282316e-06, "loss": 0.0111, "step": 5441 }, { "epoch": 1.2382252559726963, "grad_norm": 0.6919558166199928, "learning_rate": 1.0702619340048077e-06, "loss": 0.0107, "step": 5442 }, { "epoch": 1.238452787258248, "grad_norm": 0.916701734430617, "learning_rate": 1.0701992267810454e-06, "loss": 0.0095, "step": 5443 }, { "epoch": 1.2386803185437998, "grad_norm": 0.6806190525241188, "learning_rate": 1.0701365104582262e-06, "loss": 0.0166, "step": 5444 }, { "epoch": 1.2389078498293515, "grad_norm": 0.9812678457663916, "learning_rate": 1.070073785037632e-06, "loss": 0.0179, "step": 5445 }, { "epoch": 1.2391353811149033, "grad_norm": 1.2184326624890787, "learning_rate": 1.0700110505205447e-06, "loss": 0.0285, "step": 5446 }, { "epoch": 1.239362912400455, "grad_norm": 0.9122394182634046, "learning_rate": 1.0699483069082468e-06, "loss": 0.0206, "step": 5447 }, { "epoch": 1.2395904436860068, "grad_norm": 1.776304008174226, "learning_rate": 1.0698855542020201e-06, "loss": 0.0406, "step": 5448 }, { "epoch": 1.2398179749715585, "grad_norm": 2.144224863401225, "learning_rate": 1.0698227924031474e-06, "loss": 0.0269, "step": 5449 }, { "epoch": 1.2400455062571103, "grad_norm": 0.8197781138940033, "learning_rate": 1.0697600215129113e-06, "loss": 0.0132, "step": 5450 }, { "epoch": 1.240273037542662, "grad_norm": 1.112356089568793, "learning_rate": 1.0696972415325954e-06, "loss": 0.0188, "step": 5451 }, { "epoch": 1.2405005688282138, "grad_norm": 1.5519612957572413, "learning_rate": 1.069634452463482e-06, "loss": 0.0649, "step": 5452 }, { "epoch": 1.2407281001137656, "grad_norm": 1.7090640826753272, "learning_rate": 1.0695716543068548e-06, "loss": 0.021, "step": 5453 }, { "epoch": 1.2409556313993173, "grad_norm": 0.9894882585365135, "learning_rate": 1.0695088470639973e-06, "loss": 0.031, "step": 5454 }, { "epoch": 1.241183162684869, "grad_norm": 1.2111142205989058, "learning_rate": 1.069446030736193e-06, "loss": 0.0328, "step": 5455 }, { "epoch": 1.2414106939704208, "grad_norm": 0.8409351330327227, "learning_rate": 1.0693832053247256e-06, "loss": 0.0147, "step": 5456 }, { "epoch": 1.2416382252559728, "grad_norm": 1.2297382097518612, "learning_rate": 1.0693203708308792e-06, "loss": 0.018, "step": 5457 }, { "epoch": 1.2418657565415245, "grad_norm": 0.7856520845586189, "learning_rate": 1.0692575272559385e-06, "loss": 0.0132, "step": 5458 }, { "epoch": 1.2420932878270763, "grad_norm": 2.9495226963416092, "learning_rate": 1.0691946746011874e-06, "loss": 0.0753, "step": 5459 }, { "epoch": 1.242320819112628, "grad_norm": 1.1311124384842985, "learning_rate": 1.0691318128679107e-06, "loss": 0.0282, "step": 5460 }, { "epoch": 1.2425483503981798, "grad_norm": 1.2641663291780139, "learning_rate": 1.0690689420573933e-06, "loss": 0.0359, "step": 5461 }, { "epoch": 1.2427758816837315, "grad_norm": 1.5841963404042003, "learning_rate": 1.0690060621709198e-06, "loss": 0.0334, "step": 5462 }, { "epoch": 1.2430034129692833, "grad_norm": 0.7353352974419007, "learning_rate": 1.0689431732097754e-06, "loss": 0.0272, "step": 5463 }, { "epoch": 1.243230944254835, "grad_norm": 1.0500642122401156, "learning_rate": 1.0688802751752458e-06, "loss": 0.0275, "step": 5464 }, { "epoch": 1.2434584755403868, "grad_norm": 0.7175451483028665, "learning_rate": 1.0688173680686164e-06, "loss": 0.0114, "step": 5465 }, { "epoch": 1.2436860068259386, "grad_norm": 0.5867289910981424, "learning_rate": 1.0687544518911726e-06, "loss": 0.0111, "step": 5466 }, { "epoch": 1.2439135381114903, "grad_norm": 0.6712196944383615, "learning_rate": 1.0686915266442005e-06, "loss": 0.0162, "step": 5467 }, { "epoch": 1.244141069397042, "grad_norm": 1.2463644416881146, "learning_rate": 1.0686285923289863e-06, "loss": 0.0442, "step": 5468 }, { "epoch": 1.2443686006825938, "grad_norm": 0.8847672705631197, "learning_rate": 1.0685656489468161e-06, "loss": 0.0137, "step": 5469 }, { "epoch": 1.2445961319681456, "grad_norm": 0.8604702504051126, "learning_rate": 1.0685026964989764e-06, "loss": 0.0168, "step": 5470 }, { "epoch": 1.2448236632536973, "grad_norm": 1.418689898486085, "learning_rate": 1.0684397349867537e-06, "loss": 0.0292, "step": 5471 }, { "epoch": 1.245051194539249, "grad_norm": 1.846189023146859, "learning_rate": 1.068376764411435e-06, "loss": 0.0448, "step": 5472 }, { "epoch": 1.2452787258248008, "grad_norm": 1.252366571384877, "learning_rate": 1.0683137847743076e-06, "loss": 0.0301, "step": 5473 }, { "epoch": 1.2455062571103528, "grad_norm": 1.2815205403700491, "learning_rate": 1.0682507960766578e-06, "loss": 0.0317, "step": 5474 }, { "epoch": 1.2457337883959045, "grad_norm": 1.5712499828336253, "learning_rate": 1.0681877983197738e-06, "loss": 0.0511, "step": 5475 }, { "epoch": 1.2459613196814563, "grad_norm": 1.3120502665157674, "learning_rate": 1.0681247915049428e-06, "loss": 0.0239, "step": 5476 }, { "epoch": 1.246188850967008, "grad_norm": 1.2372734878283551, "learning_rate": 1.0680617756334527e-06, "loss": 0.025, "step": 5477 }, { "epoch": 1.2464163822525598, "grad_norm": 1.1291827306006323, "learning_rate": 1.0679987507065912e-06, "loss": 0.0155, "step": 5478 }, { "epoch": 1.2466439135381115, "grad_norm": 1.1166618335200997, "learning_rate": 1.0679357167256465e-06, "loss": 0.015, "step": 5479 }, { "epoch": 1.2468714448236633, "grad_norm": 0.9525747638788947, "learning_rate": 1.067872673691907e-06, "loss": 0.0379, "step": 5480 }, { "epoch": 1.247098976109215, "grad_norm": 0.9015026474953439, "learning_rate": 1.0678096216066611e-06, "loss": 0.0171, "step": 5481 }, { "epoch": 1.2473265073947668, "grad_norm": 0.7251387961734896, "learning_rate": 1.0677465604711975e-06, "loss": 0.0183, "step": 5482 }, { "epoch": 1.2475540386803186, "grad_norm": 1.1041778984913362, "learning_rate": 1.0676834902868051e-06, "loss": 0.0449, "step": 5483 }, { "epoch": 1.2477815699658703, "grad_norm": 1.5486690407695638, "learning_rate": 1.067620411054773e-06, "loss": 0.0559, "step": 5484 }, { "epoch": 1.248009101251422, "grad_norm": 1.0936433230281484, "learning_rate": 1.0675573227763903e-06, "loss": 0.041, "step": 5485 }, { "epoch": 1.2482366325369738, "grad_norm": 0.7723592633898496, "learning_rate": 1.0674942254529463e-06, "loss": 0.0196, "step": 5486 }, { "epoch": 1.2484641638225256, "grad_norm": 1.3114855555869747, "learning_rate": 1.0674311190857308e-06, "loss": 0.0231, "step": 5487 }, { "epoch": 1.2486916951080773, "grad_norm": 1.781546684980227, "learning_rate": 1.0673680036760333e-06, "loss": 0.0336, "step": 5488 }, { "epoch": 1.248919226393629, "grad_norm": 1.4863914973995318, "learning_rate": 1.0673048792251443e-06, "loss": 0.0284, "step": 5489 }, { "epoch": 1.2491467576791808, "grad_norm": 0.7255040029884252, "learning_rate": 1.0672417457343535e-06, "loss": 0.0119, "step": 5490 }, { "epoch": 1.2493742889647326, "grad_norm": 1.1732515422757923, "learning_rate": 1.0671786032049512e-06, "loss": 0.0247, "step": 5491 }, { "epoch": 1.2496018202502843, "grad_norm": 0.8454940508394108, "learning_rate": 1.0671154516382283e-06, "loss": 0.0126, "step": 5492 }, { "epoch": 1.249829351535836, "grad_norm": 1.7717426170006738, "learning_rate": 1.067052291035475e-06, "loss": 0.0288, "step": 5493 }, { "epoch": 1.2500568828213878, "grad_norm": 0.9733708679345942, "learning_rate": 1.0669891213979826e-06, "loss": 0.0424, "step": 5494 }, { "epoch": 1.2502844141069396, "grad_norm": 0.8099462019509774, "learning_rate": 1.066925942727042e-06, "loss": 0.0135, "step": 5495 }, { "epoch": 1.2505119453924913, "grad_norm": 1.0927120951454266, "learning_rate": 1.0668627550239444e-06, "loss": 0.02, "step": 5496 }, { "epoch": 1.2507394766780433, "grad_norm": 0.9823491330955677, "learning_rate": 1.0667995582899815e-06, "loss": 0.0117, "step": 5497 }, { "epoch": 1.250967007963595, "grad_norm": 1.3444249454765562, "learning_rate": 1.0667363525264446e-06, "loss": 0.0272, "step": 5498 }, { "epoch": 1.2511945392491468, "grad_norm": 1.5794185821366846, "learning_rate": 1.0666731377346257e-06, "loss": 0.0429, "step": 5499 }, { "epoch": 1.2514220705346986, "grad_norm": 0.827330081267108, "learning_rate": 1.0666099139158168e-06, "loss": 0.0133, "step": 5500 }, { "epoch": 1.2516496018202503, "grad_norm": 1.5562286915608332, "learning_rate": 1.06654668107131e-06, "loss": 0.037, "step": 5501 }, { "epoch": 1.251877133105802, "grad_norm": 1.8342841703579758, "learning_rate": 1.0664834392023975e-06, "loss": 0.0369, "step": 5502 }, { "epoch": 1.2521046643913538, "grad_norm": 1.9754022342502022, "learning_rate": 1.0664201883103722e-06, "loss": 0.0576, "step": 5503 }, { "epoch": 1.2523321956769056, "grad_norm": 1.704003080347443, "learning_rate": 1.066356928396527e-06, "loss": 0.041, "step": 5504 }, { "epoch": 1.2525597269624573, "grad_norm": 0.8152446360977106, "learning_rate": 1.066293659462154e-06, "loss": 0.0092, "step": 5505 }, { "epoch": 1.252787258248009, "grad_norm": 1.187665186303167, "learning_rate": 1.066230381508547e-06, "loss": 0.03, "step": 5506 }, { "epoch": 1.2530147895335608, "grad_norm": 1.515571401940605, "learning_rate": 1.0661670945369991e-06, "loss": 0.0658, "step": 5507 }, { "epoch": 1.2532423208191126, "grad_norm": 0.7829970004154015, "learning_rate": 1.0661037985488037e-06, "loss": 0.0221, "step": 5508 }, { "epoch": 1.2534698521046643, "grad_norm": 0.9344601178136276, "learning_rate": 1.0660404935452545e-06, "loss": 0.0176, "step": 5509 }, { "epoch": 1.253697383390216, "grad_norm": 1.1741631037543796, "learning_rate": 1.0659771795276451e-06, "loss": 0.0383, "step": 5510 }, { "epoch": 1.253924914675768, "grad_norm": 0.6952087144812199, "learning_rate": 1.06591385649727e-06, "loss": 0.0166, "step": 5511 }, { "epoch": 1.2541524459613198, "grad_norm": 1.148195580887031, "learning_rate": 1.0658505244554233e-06, "loss": 0.0336, "step": 5512 }, { "epoch": 1.2543799772468716, "grad_norm": 0.9686864698151374, "learning_rate": 1.0657871834033992e-06, "loss": 0.017, "step": 5513 }, { "epoch": 1.2546075085324233, "grad_norm": 0.6541574267823076, "learning_rate": 1.0657238333424922e-06, "loss": 0.0149, "step": 5514 }, { "epoch": 1.254835039817975, "grad_norm": 0.682956550720555, "learning_rate": 1.0656604742739974e-06, "loss": 0.0106, "step": 5515 }, { "epoch": 1.2550625711035268, "grad_norm": 1.6100348965042812, "learning_rate": 1.0655971061992093e-06, "loss": 0.0445, "step": 5516 }, { "epoch": 1.2552901023890786, "grad_norm": 1.1979735909844647, "learning_rate": 1.0655337291194235e-06, "loss": 0.0369, "step": 5517 }, { "epoch": 1.2555176336746303, "grad_norm": 0.7815830769361026, "learning_rate": 1.0654703430359348e-06, "loss": 0.0231, "step": 5518 }, { "epoch": 1.255745164960182, "grad_norm": 0.6929286637595476, "learning_rate": 1.065406947950039e-06, "loss": 0.0106, "step": 5519 }, { "epoch": 1.2559726962457338, "grad_norm": 1.1236746085503946, "learning_rate": 1.065343543863032e-06, "loss": 0.0344, "step": 5520 }, { "epoch": 1.2562002275312856, "grad_norm": 0.570375538462011, "learning_rate": 1.0652801307762093e-06, "loss": 0.0154, "step": 5521 }, { "epoch": 1.2564277588168373, "grad_norm": 1.182265579469843, "learning_rate": 1.065216708690867e-06, "loss": 0.0251, "step": 5522 }, { "epoch": 1.256655290102389, "grad_norm": 1.0748877350333434, "learning_rate": 1.0651532776083014e-06, "loss": 0.0252, "step": 5523 }, { "epoch": 1.2568828213879408, "grad_norm": 0.6801381732217795, "learning_rate": 1.0650898375298088e-06, "loss": 0.0175, "step": 5524 }, { "epoch": 1.2571103526734926, "grad_norm": 0.8236115723758932, "learning_rate": 1.0650263884566863e-06, "loss": 0.0122, "step": 5525 }, { "epoch": 1.2573378839590443, "grad_norm": 1.2072865539422744, "learning_rate": 1.06496293039023e-06, "loss": 0.0307, "step": 5526 }, { "epoch": 1.257565415244596, "grad_norm": 0.8638078889384445, "learning_rate": 1.0648994633317373e-06, "loss": 0.0338, "step": 5527 }, { "epoch": 1.2577929465301478, "grad_norm": 1.3903600138449066, "learning_rate": 1.064835987282505e-06, "loss": 0.0427, "step": 5528 }, { "epoch": 1.2580204778156996, "grad_norm": 1.2546291636301692, "learning_rate": 1.0647725022438307e-06, "loss": 0.0159, "step": 5529 }, { "epoch": 1.2582480091012513, "grad_norm": 0.756663737422429, "learning_rate": 1.0647090082170118e-06, "loss": 0.013, "step": 5530 }, { "epoch": 1.258475540386803, "grad_norm": 0.8566876022289612, "learning_rate": 1.0646455052033463e-06, "loss": 0.0201, "step": 5531 }, { "epoch": 1.2587030716723548, "grad_norm": 1.1638253388866646, "learning_rate": 1.0645819932041317e-06, "loss": 0.0166, "step": 5532 }, { "epoch": 1.2589306029579066, "grad_norm": 1.1215067967587442, "learning_rate": 1.064518472220666e-06, "loss": 0.0138, "step": 5533 }, { "epoch": 1.2591581342434583, "grad_norm": 1.2121221098665633, "learning_rate": 1.064454942254248e-06, "loss": 0.0214, "step": 5534 }, { "epoch": 1.25938566552901, "grad_norm": 0.7454316468560616, "learning_rate": 1.0643914033061757e-06, "loss": 0.0129, "step": 5535 }, { "epoch": 1.259613196814562, "grad_norm": 1.7146758293381863, "learning_rate": 1.0643278553777477e-06, "loss": 0.0243, "step": 5536 }, { "epoch": 1.2598407281001138, "grad_norm": 32.10042100778171, "learning_rate": 1.0642642984702632e-06, "loss": 0.4248, "step": 5537 }, { "epoch": 1.2600682593856656, "grad_norm": 1.0484198163247316, "learning_rate": 1.0642007325850207e-06, "loss": 0.0287, "step": 5538 }, { "epoch": 1.2602957906712173, "grad_norm": 2.3948890958568487, "learning_rate": 1.0641371577233197e-06, "loss": 0.0669, "step": 5539 }, { "epoch": 1.260523321956769, "grad_norm": 0.9801146567578284, "learning_rate": 1.0640735738864594e-06, "loss": 0.0162, "step": 5540 }, { "epoch": 1.2607508532423208, "grad_norm": 0.9922579714313251, "learning_rate": 1.0640099810757394e-06, "loss": 0.0181, "step": 5541 }, { "epoch": 1.2609783845278726, "grad_norm": 0.8815528286158882, "learning_rate": 1.0639463792924592e-06, "loss": 0.0217, "step": 5542 }, { "epoch": 1.2612059158134243, "grad_norm": 0.8238976020134005, "learning_rate": 1.0638827685379191e-06, "loss": 0.0145, "step": 5543 }, { "epoch": 1.261433447098976, "grad_norm": 1.1437101469135258, "learning_rate": 1.063819148813419e-06, "loss": 0.0248, "step": 5544 }, { "epoch": 1.2616609783845278, "grad_norm": 2.5729371322737475, "learning_rate": 1.063755520120259e-06, "loss": 0.0719, "step": 5545 }, { "epoch": 1.2618885096700796, "grad_norm": 0.742178098402897, "learning_rate": 1.0636918824597397e-06, "loss": 0.011, "step": 5546 }, { "epoch": 1.2621160409556313, "grad_norm": 1.2993072879022185, "learning_rate": 1.063628235833162e-06, "loss": 0.0127, "step": 5547 }, { "epoch": 1.262343572241183, "grad_norm": 1.050947969366834, "learning_rate": 1.0635645802418263e-06, "loss": 0.0203, "step": 5548 }, { "epoch": 1.2625711035267349, "grad_norm": 1.0789000058560003, "learning_rate": 1.0635009156870338e-06, "loss": 0.0292, "step": 5549 }, { "epoch": 1.2627986348122868, "grad_norm": 0.7771290887161867, "learning_rate": 1.0634372421700858e-06, "loss": 0.0081, "step": 5550 } ], "logging_steps": 1, "max_steps": 21975, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 1110, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 18269774659584.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }