llama3-8b-math-sft-full-test / trainer_state.json
Dynosaur's picture
Model save
78b5217 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9977298524404086,
"eval_steps": 500,
"global_step": 660,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003026863412788498,
"grad_norm": 2.653642309001846,
"learning_rate": 3.0303030303030305e-07,
"loss": 0.7607,
"step": 1
},
{
"epoch": 0.006053726825576996,
"grad_norm": 2.5922112333202088,
"learning_rate": 6.060606060606061e-07,
"loss": 0.7518,
"step": 2
},
{
"epoch": 0.009080590238365494,
"grad_norm": 2.554194759996629,
"learning_rate": 9.090909090909091e-07,
"loss": 0.7257,
"step": 3
},
{
"epoch": 0.012107453651153992,
"grad_norm": 2.63438925636528,
"learning_rate": 1.2121212121212122e-06,
"loss": 0.7808,
"step": 4
},
{
"epoch": 0.01513431706394249,
"grad_norm": 2.462685042118507,
"learning_rate": 1.5151515151515152e-06,
"loss": 0.7046,
"step": 5
},
{
"epoch": 0.018161180476730987,
"grad_norm": 2.352286315669494,
"learning_rate": 1.8181818181818183e-06,
"loss": 0.7208,
"step": 6
},
{
"epoch": 0.021188043889519486,
"grad_norm": 2.063556901218873,
"learning_rate": 2.1212121212121216e-06,
"loss": 0.7149,
"step": 7
},
{
"epoch": 0.024214907302307985,
"grad_norm": 1.8154847790671955,
"learning_rate": 2.4242424242424244e-06,
"loss": 0.6994,
"step": 8
},
{
"epoch": 0.02724177071509648,
"grad_norm": 1.2076222176550426,
"learning_rate": 2.7272727272727272e-06,
"loss": 0.6912,
"step": 9
},
{
"epoch": 0.03026863412788498,
"grad_norm": 1.214170698873997,
"learning_rate": 3.0303030303030305e-06,
"loss": 0.6898,
"step": 10
},
{
"epoch": 0.03329549754067348,
"grad_norm": 1.249393928614815,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.659,
"step": 11
},
{
"epoch": 0.036322360953461974,
"grad_norm": 1.9156334051480104,
"learning_rate": 3.6363636363636366e-06,
"loss": 0.6679,
"step": 12
},
{
"epoch": 0.03934922436625047,
"grad_norm": 3.0417637453190634,
"learning_rate": 3.93939393939394e-06,
"loss": 0.6902,
"step": 13
},
{
"epoch": 0.04237608777903897,
"grad_norm": 2.459496434867318,
"learning_rate": 4.242424242424243e-06,
"loss": 0.6584,
"step": 14
},
{
"epoch": 0.04540295119182747,
"grad_norm": 2.0017984414461925,
"learning_rate": 4.5454545454545455e-06,
"loss": 0.662,
"step": 15
},
{
"epoch": 0.04842981460461597,
"grad_norm": 1.3619786408678614,
"learning_rate": 4.848484848484849e-06,
"loss": 0.5831,
"step": 16
},
{
"epoch": 0.051456678017404466,
"grad_norm": 1.1123523922758016,
"learning_rate": 5.151515151515152e-06,
"loss": 0.6011,
"step": 17
},
{
"epoch": 0.05448354143019296,
"grad_norm": 1.2313268742997454,
"learning_rate": 5.4545454545454545e-06,
"loss": 0.6159,
"step": 18
},
{
"epoch": 0.057510404842981463,
"grad_norm": 1.1775618980909193,
"learning_rate": 5.7575757575757586e-06,
"loss": 0.6232,
"step": 19
},
{
"epoch": 0.06053726825576996,
"grad_norm": 1.0161247742343753,
"learning_rate": 6.060606060606061e-06,
"loss": 0.5972,
"step": 20
},
{
"epoch": 0.06356413166855845,
"grad_norm": 0.8477184006555634,
"learning_rate": 6.363636363636364e-06,
"loss": 0.589,
"step": 21
},
{
"epoch": 0.06659099508134696,
"grad_norm": 0.7581805666749508,
"learning_rate": 6.666666666666667e-06,
"loss": 0.5371,
"step": 22
},
{
"epoch": 0.06961785849413545,
"grad_norm": 0.8705917577466252,
"learning_rate": 6.969696969696971e-06,
"loss": 0.5591,
"step": 23
},
{
"epoch": 0.07264472190692395,
"grad_norm": 0.8522831911904432,
"learning_rate": 7.272727272727273e-06,
"loss": 0.6045,
"step": 24
},
{
"epoch": 0.07567158531971245,
"grad_norm": 0.7352647897854065,
"learning_rate": 7.5757575757575764e-06,
"loss": 0.5433,
"step": 25
},
{
"epoch": 0.07869844873250094,
"grad_norm": 0.7194423582606616,
"learning_rate": 7.87878787878788e-06,
"loss": 0.5745,
"step": 26
},
{
"epoch": 0.08172531214528944,
"grad_norm": 0.7178198575109902,
"learning_rate": 8.181818181818183e-06,
"loss": 0.5365,
"step": 27
},
{
"epoch": 0.08475217555807794,
"grad_norm": 0.7287566186498791,
"learning_rate": 8.484848484848486e-06,
"loss": 0.5047,
"step": 28
},
{
"epoch": 0.08777903897086645,
"grad_norm": 0.795016826843346,
"learning_rate": 8.787878787878788e-06,
"loss": 0.5747,
"step": 29
},
{
"epoch": 0.09080590238365494,
"grad_norm": 0.7070438081150247,
"learning_rate": 9.090909090909091e-06,
"loss": 0.5558,
"step": 30
},
{
"epoch": 0.09383276579644344,
"grad_norm": 0.7189224373814807,
"learning_rate": 9.393939393939396e-06,
"loss": 0.5253,
"step": 31
},
{
"epoch": 0.09685962920923194,
"grad_norm": 0.708188733507097,
"learning_rate": 9.696969696969698e-06,
"loss": 0.5357,
"step": 32
},
{
"epoch": 0.09988649262202043,
"grad_norm": 0.709892169548937,
"learning_rate": 1e-05,
"loss": 0.5542,
"step": 33
},
{
"epoch": 0.10291335603480893,
"grad_norm": 0.6689205883594461,
"learning_rate": 1.0303030303030304e-05,
"loss": 0.5251,
"step": 34
},
{
"epoch": 0.10594021944759743,
"grad_norm": 0.7708338817589518,
"learning_rate": 1.0606060606060606e-05,
"loss": 0.5624,
"step": 35
},
{
"epoch": 0.10896708286038592,
"grad_norm": 0.6607276883497549,
"learning_rate": 1.0909090909090909e-05,
"loss": 0.5277,
"step": 36
},
{
"epoch": 0.11199394627317442,
"grad_norm": 0.6782181867807554,
"learning_rate": 1.1212121212121212e-05,
"loss": 0.5471,
"step": 37
},
{
"epoch": 0.11502080968596293,
"grad_norm": 0.6633295184018997,
"learning_rate": 1.1515151515151517e-05,
"loss": 0.5076,
"step": 38
},
{
"epoch": 0.11804767309875142,
"grad_norm": 0.6831770809175276,
"learning_rate": 1.181818181818182e-05,
"loss": 0.5085,
"step": 39
},
{
"epoch": 0.12107453651153992,
"grad_norm": 0.7163772729536908,
"learning_rate": 1.2121212121212122e-05,
"loss": 0.5025,
"step": 40
},
{
"epoch": 0.12410139992432842,
"grad_norm": 0.673621628700286,
"learning_rate": 1.2424242424242425e-05,
"loss": 0.5404,
"step": 41
},
{
"epoch": 0.1271282633371169,
"grad_norm": 0.6978798166145279,
"learning_rate": 1.2727272727272728e-05,
"loss": 0.5206,
"step": 42
},
{
"epoch": 0.1301551267499054,
"grad_norm": 0.6758251776071441,
"learning_rate": 1.3030303030303032e-05,
"loss": 0.5151,
"step": 43
},
{
"epoch": 0.13318199016269391,
"grad_norm": 0.6320608089101,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.5084,
"step": 44
},
{
"epoch": 0.1362088535754824,
"grad_norm": 0.6920400003814712,
"learning_rate": 1.3636363636363637e-05,
"loss": 0.5215,
"step": 45
},
{
"epoch": 0.1392357169882709,
"grad_norm": 0.6426589394479636,
"learning_rate": 1.3939393939393942e-05,
"loss": 0.5478,
"step": 46
},
{
"epoch": 0.1422625804010594,
"grad_norm": 0.6592544978872925,
"learning_rate": 1.4242424242424245e-05,
"loss": 0.5363,
"step": 47
},
{
"epoch": 0.1452894438138479,
"grad_norm": 0.6303886033476793,
"learning_rate": 1.4545454545454546e-05,
"loss": 0.4778,
"step": 48
},
{
"epoch": 0.14831630722663638,
"grad_norm": 0.6825469178984898,
"learning_rate": 1.484848484848485e-05,
"loss": 0.5313,
"step": 49
},
{
"epoch": 0.1513431706394249,
"grad_norm": 0.6511075065453922,
"learning_rate": 1.5151515151515153e-05,
"loss": 0.5037,
"step": 50
},
{
"epoch": 0.1543700340522134,
"grad_norm": 0.6783685695103729,
"learning_rate": 1.5454545454545454e-05,
"loss": 0.4632,
"step": 51
},
{
"epoch": 0.15739689746500188,
"grad_norm": 0.6778524173791399,
"learning_rate": 1.575757575757576e-05,
"loss": 0.5347,
"step": 52
},
{
"epoch": 0.1604237608777904,
"grad_norm": 0.6713616058177643,
"learning_rate": 1.606060606060606e-05,
"loss": 0.4869,
"step": 53
},
{
"epoch": 0.16345062429057888,
"grad_norm": 0.7146450729124729,
"learning_rate": 1.6363636363636366e-05,
"loss": 0.4758,
"step": 54
},
{
"epoch": 0.16647748770336737,
"grad_norm": 0.6877829806421919,
"learning_rate": 1.6666666666666667e-05,
"loss": 0.5339,
"step": 55
},
{
"epoch": 0.1695043511161559,
"grad_norm": 0.638685196422612,
"learning_rate": 1.6969696969696972e-05,
"loss": 0.4868,
"step": 56
},
{
"epoch": 0.17253121452894438,
"grad_norm": 0.7094289181124012,
"learning_rate": 1.7272727272727274e-05,
"loss": 0.5111,
"step": 57
},
{
"epoch": 0.1755580779417329,
"grad_norm": 0.6714538266612285,
"learning_rate": 1.7575757575757576e-05,
"loss": 0.5121,
"step": 58
},
{
"epoch": 0.17858494135452138,
"grad_norm": 0.6969505485711796,
"learning_rate": 1.787878787878788e-05,
"loss": 0.4897,
"step": 59
},
{
"epoch": 0.18161180476730987,
"grad_norm": 0.7349628866367,
"learning_rate": 1.8181818181818182e-05,
"loss": 0.553,
"step": 60
},
{
"epoch": 0.1846386681800984,
"grad_norm": 0.6696445466750143,
"learning_rate": 1.8484848484848487e-05,
"loss": 0.4951,
"step": 61
},
{
"epoch": 0.18766553159288688,
"grad_norm": 0.6617587898318434,
"learning_rate": 1.8787878787878792e-05,
"loss": 0.5284,
"step": 62
},
{
"epoch": 0.19069239500567536,
"grad_norm": 0.706221486781084,
"learning_rate": 1.9090909090909094e-05,
"loss": 0.4885,
"step": 63
},
{
"epoch": 0.19371925841846388,
"grad_norm": 0.6614213463731939,
"learning_rate": 1.9393939393939395e-05,
"loss": 0.5159,
"step": 64
},
{
"epoch": 0.19674612183125237,
"grad_norm": 0.726559409699496,
"learning_rate": 1.96969696969697e-05,
"loss": 0.5148,
"step": 65
},
{
"epoch": 0.19977298524404086,
"grad_norm": 0.703206497838019,
"learning_rate": 2e-05,
"loss": 0.5015,
"step": 66
},
{
"epoch": 0.20279984865682937,
"grad_norm": 0.8434838159306466,
"learning_rate": 1.9999860139251737e-05,
"loss": 0.527,
"step": 67
},
{
"epoch": 0.20582671206961786,
"grad_norm": 0.6132658265405436,
"learning_rate": 1.9999440560919153e-05,
"loss": 0.4921,
"step": 68
},
{
"epoch": 0.20885357548240635,
"grad_norm": 0.7505502804477754,
"learning_rate": 1.9998741276738753e-05,
"loss": 0.469,
"step": 69
},
{
"epoch": 0.21188043889519487,
"grad_norm": 0.7115606662498123,
"learning_rate": 1.999776230627102e-05,
"loss": 0.5244,
"step": 70
},
{
"epoch": 0.21490730230798336,
"grad_norm": 0.7570799848249147,
"learning_rate": 1.9996503676899863e-05,
"loss": 0.4871,
"step": 71
},
{
"epoch": 0.21793416572077184,
"grad_norm": 0.6745168914567232,
"learning_rate": 1.9994965423831853e-05,
"loss": 0.5216,
"step": 72
},
{
"epoch": 0.22096102913356036,
"grad_norm": 0.8026032876428376,
"learning_rate": 1.9993147590095232e-05,
"loss": 0.4818,
"step": 73
},
{
"epoch": 0.22398789254634885,
"grad_norm": 0.7703357098063327,
"learning_rate": 1.999105022653872e-05,
"loss": 0.5289,
"step": 74
},
{
"epoch": 0.22701475595913734,
"grad_norm": 0.6884525063015393,
"learning_rate": 1.9988673391830082e-05,
"loss": 0.4997,
"step": 75
},
{
"epoch": 0.23004161937192585,
"grad_norm": 0.7908440606495958,
"learning_rate": 1.9986017152454497e-05,
"loss": 0.5035,
"step": 76
},
{
"epoch": 0.23306848278471434,
"grad_norm": 0.6217642015366157,
"learning_rate": 1.9983081582712684e-05,
"loss": 0.4673,
"step": 77
},
{
"epoch": 0.23609534619750283,
"grad_norm": 0.8086567400881232,
"learning_rate": 1.9979866764718846e-05,
"loss": 0.4936,
"step": 78
},
{
"epoch": 0.23912220961029135,
"grad_norm": 0.6748553548070318,
"learning_rate": 1.997637278839835e-05,
"loss": 0.5007,
"step": 79
},
{
"epoch": 0.24214907302307984,
"grad_norm": 0.7909118323454936,
"learning_rate": 1.9972599751485225e-05,
"loss": 0.4929,
"step": 80
},
{
"epoch": 0.24517593643586832,
"grad_norm": 0.6761702566000428,
"learning_rate": 1.9968547759519426e-05,
"loss": 0.5018,
"step": 81
},
{
"epoch": 0.24820279984865684,
"grad_norm": 0.7713095553128116,
"learning_rate": 1.9964216925843876e-05,
"loss": 0.4794,
"step": 82
},
{
"epoch": 0.2512296632614453,
"grad_norm": 0.6765563475808045,
"learning_rate": 1.9959607371601303e-05,
"loss": 0.4839,
"step": 83
},
{
"epoch": 0.2542565266742338,
"grad_norm": 0.7848989238135033,
"learning_rate": 1.9954719225730847e-05,
"loss": 0.503,
"step": 84
},
{
"epoch": 0.25728339008702233,
"grad_norm": 0.6752303325964191,
"learning_rate": 1.994955262496446e-05,
"loss": 0.4929,
"step": 85
},
{
"epoch": 0.2603102534998108,
"grad_norm": 0.8090157427997055,
"learning_rate": 1.9944107713823068e-05,
"loss": 0.5074,
"step": 86
},
{
"epoch": 0.2633371169125993,
"grad_norm": 0.6148199872593046,
"learning_rate": 1.9938384644612542e-05,
"loss": 0.4761,
"step": 87
},
{
"epoch": 0.26636398032538783,
"grad_norm": 0.6818622789453987,
"learning_rate": 1.9932383577419432e-05,
"loss": 0.4856,
"step": 88
},
{
"epoch": 0.2693908437381763,
"grad_norm": 0.6706670307612664,
"learning_rate": 1.9926104680106484e-05,
"loss": 0.5031,
"step": 89
},
{
"epoch": 0.2724177071509648,
"grad_norm": 0.6707470958888189,
"learning_rate": 1.9919548128307954e-05,
"loss": 0.4713,
"step": 90
},
{
"epoch": 0.2754445705637533,
"grad_norm": 0.7024692529494156,
"learning_rate": 1.9912714105424694e-05,
"loss": 0.4836,
"step": 91
},
{
"epoch": 0.2784714339765418,
"grad_norm": 0.6826575867532084,
"learning_rate": 1.990560280261901e-05,
"loss": 0.4818,
"step": 92
},
{
"epoch": 0.2814982973893303,
"grad_norm": 0.6875484519608218,
"learning_rate": 1.989821441880933e-05,
"loss": 0.4707,
"step": 93
},
{
"epoch": 0.2845251608021188,
"grad_norm": 0.6989815243455156,
"learning_rate": 1.9890549160664633e-05,
"loss": 0.4555,
"step": 94
},
{
"epoch": 0.2875520242149073,
"grad_norm": 0.6881951073889534,
"learning_rate": 1.9882607242598663e-05,
"loss": 0.5173,
"step": 95
},
{
"epoch": 0.2905788876276958,
"grad_norm": 0.7746163150218435,
"learning_rate": 1.9874388886763944e-05,
"loss": 0.4493,
"step": 96
},
{
"epoch": 0.2936057510404843,
"grad_norm": 0.7874564596528957,
"learning_rate": 1.9865894323045558e-05,
"loss": 0.5304,
"step": 97
},
{
"epoch": 0.29663261445327277,
"grad_norm": 0.701524225555686,
"learning_rate": 1.9857123789054707e-05,
"loss": 0.5377,
"step": 98
},
{
"epoch": 0.2996594778660613,
"grad_norm": 0.8145646846456013,
"learning_rate": 1.9848077530122083e-05,
"loss": 0.502,
"step": 99
},
{
"epoch": 0.3026863412788498,
"grad_norm": 0.6768328450225374,
"learning_rate": 1.9838755799290993e-05,
"loss": 0.4821,
"step": 100
},
{
"epoch": 0.30571320469163826,
"grad_norm": 0.8263833925233237,
"learning_rate": 1.9829158857310288e-05,
"loss": 0.513,
"step": 101
},
{
"epoch": 0.3087400681044268,
"grad_norm": 0.705647528914653,
"learning_rate": 1.9819286972627066e-05,
"loss": 0.4954,
"step": 102
},
{
"epoch": 0.3117669315172153,
"grad_norm": 1.112583589601411,
"learning_rate": 1.9809140421379168e-05,
"loss": 0.4735,
"step": 103
},
{
"epoch": 0.31479379493000376,
"grad_norm": 1.0573008413710339,
"learning_rate": 1.979871948738743e-05,
"loss": 0.4988,
"step": 104
},
{
"epoch": 0.3178206583427923,
"grad_norm": 4.563156628451907,
"learning_rate": 1.978802446214779e-05,
"loss": 0.5498,
"step": 105
},
{
"epoch": 0.3208475217555808,
"grad_norm": 1.5419963879262906,
"learning_rate": 1.9777055644823087e-05,
"loss": 0.518,
"step": 106
},
{
"epoch": 0.32387438516836925,
"grad_norm": 0.9957849878521595,
"learning_rate": 1.9765813342234726e-05,
"loss": 0.5223,
"step": 107
},
{
"epoch": 0.32690124858115777,
"grad_norm": 0.9918631792361824,
"learning_rate": 1.9754297868854075e-05,
"loss": 0.491,
"step": 108
},
{
"epoch": 0.3299281119939463,
"grad_norm": 0.7501498300964237,
"learning_rate": 1.9742509546793673e-05,
"loss": 0.527,
"step": 109
},
{
"epoch": 0.33295497540673474,
"grad_norm": 0.8475448493105577,
"learning_rate": 1.973044870579824e-05,
"loss": 0.5106,
"step": 110
},
{
"epoch": 0.33598183881952326,
"grad_norm": 0.6097368728455299,
"learning_rate": 1.9718115683235418e-05,
"loss": 0.4682,
"step": 111
},
{
"epoch": 0.3390087022323118,
"grad_norm": 0.7885205586754842,
"learning_rate": 1.970551082408636e-05,
"loss": 0.4548,
"step": 112
},
{
"epoch": 0.3420355656451003,
"grad_norm": 0.6172104545055264,
"learning_rate": 1.969263448093608e-05,
"loss": 0.4661,
"step": 113
},
{
"epoch": 0.34506242905788875,
"grad_norm": 0.719254702740748,
"learning_rate": 1.9679487013963566e-05,
"loss": 0.5002,
"step": 114
},
{
"epoch": 0.34808929247067727,
"grad_norm": 0.6117151881946589,
"learning_rate": 1.9666068790931733e-05,
"loss": 0.4877,
"step": 115
},
{
"epoch": 0.3511161558834658,
"grad_norm": 0.7027132398898194,
"learning_rate": 1.9652380187177128e-05,
"loss": 0.4677,
"step": 116
},
{
"epoch": 0.35414301929625425,
"grad_norm": 0.6054380731971206,
"learning_rate": 1.9638421585599422e-05,
"loss": 0.4619,
"step": 117
},
{
"epoch": 0.35716988270904276,
"grad_norm": 0.6697458443391867,
"learning_rate": 1.9624193376650708e-05,
"loss": 0.4725,
"step": 118
},
{
"epoch": 0.3601967461218313,
"grad_norm": 0.6320885158836743,
"learning_rate": 1.960969595832457e-05,
"loss": 0.4735,
"step": 119
},
{
"epoch": 0.36322360953461974,
"grad_norm": 0.7297143353827921,
"learning_rate": 1.9594929736144978e-05,
"loss": 0.4995,
"step": 120
},
{
"epoch": 0.36625047294740826,
"grad_norm": 0.6882788090108111,
"learning_rate": 1.957989512315489e-05,
"loss": 0.4802,
"step": 121
},
{
"epoch": 0.3692773363601968,
"grad_norm": 0.7103992968091178,
"learning_rate": 1.956459253990476e-05,
"loss": 0.4671,
"step": 122
},
{
"epoch": 0.37230419977298523,
"grad_norm": 0.6596711674879402,
"learning_rate": 1.9549022414440738e-05,
"loss": 0.4845,
"step": 123
},
{
"epoch": 0.37533106318577375,
"grad_norm": 0.7056786527520791,
"learning_rate": 1.9533185182292705e-05,
"loss": 0.5066,
"step": 124
},
{
"epoch": 0.37835792659856227,
"grad_norm": 0.6463010214185686,
"learning_rate": 1.9517081286462082e-05,
"loss": 0.4658,
"step": 125
},
{
"epoch": 0.3813847900113507,
"grad_norm": 0.6449317177074809,
"learning_rate": 1.9500711177409456e-05,
"loss": 0.4439,
"step": 126
},
{
"epoch": 0.38441165342413924,
"grad_norm": 0.6711182804100264,
"learning_rate": 1.9484075313041968e-05,
"loss": 0.4639,
"step": 127
},
{
"epoch": 0.38743851683692776,
"grad_norm": 0.6216750305908364,
"learning_rate": 1.9467174158700507e-05,
"loss": 0.4639,
"step": 128
},
{
"epoch": 0.3904653802497162,
"grad_norm": 0.6715332804368,
"learning_rate": 1.9450008187146685e-05,
"loss": 0.5023,
"step": 129
},
{
"epoch": 0.39349224366250474,
"grad_norm": 0.7147058467691788,
"learning_rate": 1.9432577878549635e-05,
"loss": 0.468,
"step": 130
},
{
"epoch": 0.39651910707529325,
"grad_norm": 0.6787351073466207,
"learning_rate": 1.9414883720472557e-05,
"loss": 0.4886,
"step": 131
},
{
"epoch": 0.3995459704880817,
"grad_norm": 0.6542651634964172,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.4928,
"step": 132
},
{
"epoch": 0.40257283390087023,
"grad_norm": 0.6730854186948818,
"learning_rate": 1.937870584301945e-05,
"loss": 0.4568,
"step": 133
},
{
"epoch": 0.40559969731365875,
"grad_norm": 0.6525479754112047,
"learning_rate": 1.9360223135616423e-05,
"loss": 0.4742,
"step": 134
},
{
"epoch": 0.4086265607264472,
"grad_norm": 0.6176641561031895,
"learning_rate": 1.9341478602651068e-05,
"loss": 0.4674,
"step": 135
},
{
"epoch": 0.4116534241392357,
"grad_norm": 0.6178984085617434,
"learning_rate": 1.932247276844826e-05,
"loss": 0.4748,
"step": 136
},
{
"epoch": 0.41468028755202424,
"grad_norm": 0.6321251612378418,
"learning_rate": 1.9303206164642037e-05,
"loss": 0.4763,
"step": 137
},
{
"epoch": 0.4177071509648127,
"grad_norm": 0.6539134673370697,
"learning_rate": 1.9283679330160726e-05,
"loss": 0.5008,
"step": 138
},
{
"epoch": 0.4207340143776012,
"grad_norm": 0.629247203087845,
"learning_rate": 1.9263892811211865e-05,
"loss": 0.4959,
"step": 139
},
{
"epoch": 0.42376087779038973,
"grad_norm": 0.6615716325218832,
"learning_rate": 1.9243847161266924e-05,
"loss": 0.4781,
"step": 140
},
{
"epoch": 0.4267877412031782,
"grad_norm": 0.6807124736623016,
"learning_rate": 1.9223542941045817e-05,
"loss": 0.4456,
"step": 141
},
{
"epoch": 0.4298146046159667,
"grad_norm": 0.5781124762880367,
"learning_rate": 1.920298071850123e-05,
"loss": 0.4461,
"step": 142
},
{
"epoch": 0.4328414680287552,
"grad_norm": 0.6966598711355443,
"learning_rate": 1.9182161068802742e-05,
"loss": 0.446,
"step": 143
},
{
"epoch": 0.4358683314415437,
"grad_norm": 0.6567319123851109,
"learning_rate": 1.9161084574320696e-05,
"loss": 0.4751,
"step": 144
},
{
"epoch": 0.4388951948543322,
"grad_norm": 0.7402740815269796,
"learning_rate": 1.913975182460996e-05,
"loss": 0.4812,
"step": 145
},
{
"epoch": 0.4419220582671207,
"grad_norm": 0.611532484278698,
"learning_rate": 1.9118163416393392e-05,
"loss": 0.4636,
"step": 146
},
{
"epoch": 0.4449489216799092,
"grad_norm": 0.7063333694418584,
"learning_rate": 1.9096319953545186e-05,
"loss": 0.492,
"step": 147
},
{
"epoch": 0.4479757850926977,
"grad_norm": 0.6347556617125097,
"learning_rate": 1.9074222047073945e-05,
"loss": 0.4441,
"step": 148
},
{
"epoch": 0.4510026485054862,
"grad_norm": 0.7009661095563158,
"learning_rate": 1.9051870315105626e-05,
"loss": 0.4595,
"step": 149
},
{
"epoch": 0.4540295119182747,
"grad_norm": 0.6155698353426865,
"learning_rate": 1.9029265382866216e-05,
"loss": 0.5157,
"step": 150
},
{
"epoch": 0.4570563753310632,
"grad_norm": 0.7565512880981403,
"learning_rate": 1.9006407882664256e-05,
"loss": 0.4718,
"step": 151
},
{
"epoch": 0.4600832387438517,
"grad_norm": 0.6377798494614118,
"learning_rate": 1.8983298453873172e-05,
"loss": 0.4734,
"step": 152
},
{
"epoch": 0.46311010215664017,
"grad_norm": 0.7115285069601067,
"learning_rate": 1.895993774291336e-05,
"loss": 0.4446,
"step": 153
},
{
"epoch": 0.4661369655694287,
"grad_norm": 0.626546307492942,
"learning_rate": 1.8936326403234125e-05,
"loss": 0.4235,
"step": 154
},
{
"epoch": 0.4691638289822172,
"grad_norm": 0.6989309565590944,
"learning_rate": 1.891246509529539e-05,
"loss": 0.4648,
"step": 155
},
{
"epoch": 0.47219069239500566,
"grad_norm": 0.6264462766886602,
"learning_rate": 1.8888354486549238e-05,
"loss": 0.4734,
"step": 156
},
{
"epoch": 0.4752175558077942,
"grad_norm": 0.6586007898441965,
"learning_rate": 1.886399525142122e-05,
"loss": 0.4606,
"step": 157
},
{
"epoch": 0.4782444192205827,
"grad_norm": 0.6762512929813792,
"learning_rate": 1.8839388071291506e-05,
"loss": 0.4673,
"step": 158
},
{
"epoch": 0.48127128263337116,
"grad_norm": 0.6520802178788475,
"learning_rate": 1.881453363447582e-05,
"loss": 0.4399,
"step": 159
},
{
"epoch": 0.48429814604615967,
"grad_norm": 0.6315321840707008,
"learning_rate": 1.8789432636206197e-05,
"loss": 0.4565,
"step": 160
},
{
"epoch": 0.4873250094589482,
"grad_norm": 0.6173557754039471,
"learning_rate": 1.8764085778611507e-05,
"loss": 0.4882,
"step": 161
},
{
"epoch": 0.49035187287173665,
"grad_norm": 0.6632428987221951,
"learning_rate": 1.873849377069785e-05,
"loss": 0.4563,
"step": 162
},
{
"epoch": 0.49337873628452517,
"grad_norm": 0.6173524446577745,
"learning_rate": 1.87126573283287e-05,
"loss": 0.4737,
"step": 163
},
{
"epoch": 0.4964055996973137,
"grad_norm": 0.7448052847307424,
"learning_rate": 1.8686577174204887e-05,
"loss": 0.4581,
"step": 164
},
{
"epoch": 0.49943246311010214,
"grad_norm": 0.6162007595348211,
"learning_rate": 1.866025403784439e-05,
"loss": 0.4597,
"step": 165
},
{
"epoch": 0.5024593265228906,
"grad_norm": 0.644579592262543,
"learning_rate": 1.863368865556191e-05,
"loss": 0.4656,
"step": 166
},
{
"epoch": 0.5054861899356792,
"grad_norm": 0.6081483512075876,
"learning_rate": 1.8606881770448305e-05,
"loss": 0.446,
"step": 167
},
{
"epoch": 0.5085130533484676,
"grad_norm": 0.6164087468455623,
"learning_rate": 1.8579834132349773e-05,
"loss": 0.4597,
"step": 168
},
{
"epoch": 0.5115399167612561,
"grad_norm": 0.5778923434429061,
"learning_rate": 1.8552546497846893e-05,
"loss": 0.475,
"step": 169
},
{
"epoch": 0.5145667801740447,
"grad_norm": 0.6904024690925634,
"learning_rate": 1.8525019630233463e-05,
"loss": 0.4591,
"step": 170
},
{
"epoch": 0.5175936435868331,
"grad_norm": 0.5733852004940836,
"learning_rate": 1.8497254299495147e-05,
"loss": 0.454,
"step": 171
},
{
"epoch": 0.5206205069996216,
"grad_norm": 0.7226227274606795,
"learning_rate": 1.8469251282287925e-05,
"loss": 0.4801,
"step": 172
},
{
"epoch": 0.5236473704124102,
"grad_norm": 0.6999327344798008,
"learning_rate": 1.8441011361916387e-05,
"loss": 0.4891,
"step": 173
},
{
"epoch": 0.5266742338251986,
"grad_norm": 0.6731236040321812,
"learning_rate": 1.8412535328311813e-05,
"loss": 0.4923,
"step": 174
},
{
"epoch": 0.5297010972379871,
"grad_norm": 0.7284066217196885,
"learning_rate": 1.8383823978010077e-05,
"loss": 0.4535,
"step": 175
},
{
"epoch": 0.5327279606507757,
"grad_norm": 0.6675316294269169,
"learning_rate": 1.8354878114129368e-05,
"loss": 0.4691,
"step": 176
},
{
"epoch": 0.5357548240635641,
"grad_norm": 0.6430468297631092,
"learning_rate": 1.8325698546347714e-05,
"loss": 0.438,
"step": 177
},
{
"epoch": 0.5387816874763526,
"grad_norm": 0.6236167988296556,
"learning_rate": 1.8296286090880362e-05,
"loss": 0.4764,
"step": 178
},
{
"epoch": 0.5418085508891411,
"grad_norm": 0.6833528437786831,
"learning_rate": 1.8266641570456915e-05,
"loss": 0.4594,
"step": 179
},
{
"epoch": 0.5448354143019296,
"grad_norm": 0.6191943776744446,
"learning_rate": 1.8236765814298328e-05,
"loss": 0.4311,
"step": 180
},
{
"epoch": 0.5478622777147181,
"grad_norm": 0.7094813750809941,
"learning_rate": 1.820665965809373e-05,
"loss": 0.4569,
"step": 181
},
{
"epoch": 0.5508891411275066,
"grad_norm": 0.6172794257535029,
"learning_rate": 1.8176323943977034e-05,
"loss": 0.4462,
"step": 182
},
{
"epoch": 0.5539160045402951,
"grad_norm": 0.6970614533926457,
"learning_rate": 1.814575952050336e-05,
"loss": 0.5002,
"step": 183
},
{
"epoch": 0.5569428679530836,
"grad_norm": 0.6134218686673267,
"learning_rate": 1.8114967242625342e-05,
"loss": 0.4541,
"step": 184
},
{
"epoch": 0.5599697313658721,
"grad_norm": 0.6279799927111867,
"learning_rate": 1.808394797166919e-05,
"loss": 0.4619,
"step": 185
},
{
"epoch": 0.5629965947786606,
"grad_norm": 0.6062609988486924,
"learning_rate": 1.8052702575310588e-05,
"loss": 0.4572,
"step": 186
},
{
"epoch": 0.5660234581914491,
"grad_norm": 0.6765594721594622,
"learning_rate": 1.802123192755044e-05,
"loss": 0.4923,
"step": 187
},
{
"epoch": 0.5690503216042376,
"grad_norm": 0.5989714922185403,
"learning_rate": 1.7989536908690413e-05,
"loss": 0.4566,
"step": 188
},
{
"epoch": 0.5720771850170261,
"grad_norm": 0.5654098894953311,
"learning_rate": 1.7957618405308323e-05,
"loss": 0.4383,
"step": 189
},
{
"epoch": 0.5751040484298146,
"grad_norm": 0.6094404762659081,
"learning_rate": 1.792547731023332e-05,
"loss": 0.4665,
"step": 190
},
{
"epoch": 0.5781309118426031,
"grad_norm": 0.578404340650778,
"learning_rate": 1.789311452252092e-05,
"loss": 0.4566,
"step": 191
},
{
"epoch": 0.5811577752553916,
"grad_norm": 0.5887549251095974,
"learning_rate": 1.7860530947427878e-05,
"loss": 0.4377,
"step": 192
},
{
"epoch": 0.58418463866818,
"grad_norm": 0.5878909756117219,
"learning_rate": 1.782772749638682e-05,
"loss": 0.4443,
"step": 193
},
{
"epoch": 0.5872115020809686,
"grad_norm": 0.6230290972876099,
"learning_rate": 1.779470508698079e-05,
"loss": 0.4407,
"step": 194
},
{
"epoch": 0.5902383654937571,
"grad_norm": 0.5883712349316433,
"learning_rate": 1.776146464291757e-05,
"loss": 0.4586,
"step": 195
},
{
"epoch": 0.5932652289065455,
"grad_norm": 0.5818914639555242,
"learning_rate": 1.772800709400383e-05,
"loss": 0.4422,
"step": 196
},
{
"epoch": 0.5962920923193341,
"grad_norm": 0.5925425243464446,
"learning_rate": 1.7694333376119144e-05,
"loss": 0.4741,
"step": 197
},
{
"epoch": 0.5993189557321226,
"grad_norm": 0.5904170861473932,
"learning_rate": 1.766044443118978e-05,
"loss": 0.4431,
"step": 198
},
{
"epoch": 0.602345819144911,
"grad_norm": 0.6806471413231173,
"learning_rate": 1.762634120716238e-05,
"loss": 0.5139,
"step": 199
},
{
"epoch": 0.6053726825576996,
"grad_norm": 0.6286910641028791,
"learning_rate": 1.7592024657977432e-05,
"loss": 0.4989,
"step": 200
},
{
"epoch": 0.6083995459704881,
"grad_norm": 0.6245618264528192,
"learning_rate": 1.7557495743542586e-05,
"loss": 0.4511,
"step": 201
},
{
"epoch": 0.6114264093832765,
"grad_norm": 0.6132991964860224,
"learning_rate": 1.75227554297058e-05,
"loss": 0.4676,
"step": 202
},
{
"epoch": 0.6144532727960651,
"grad_norm": 0.6154910007964906,
"learning_rate": 1.7487804688228327e-05,
"loss": 0.5012,
"step": 203
},
{
"epoch": 0.6174801362088536,
"grad_norm": 0.6020062397729947,
"learning_rate": 1.745264449675755e-05,
"loss": 0.4303,
"step": 204
},
{
"epoch": 0.620506999621642,
"grad_norm": 0.6210587387537155,
"learning_rate": 1.7417275838799596e-05,
"loss": 0.4502,
"step": 205
},
{
"epoch": 0.6235338630344306,
"grad_norm": 0.5997066355321784,
"learning_rate": 1.7381699703691866e-05,
"loss": 0.4353,
"step": 206
},
{
"epoch": 0.626560726447219,
"grad_norm": 0.5985750220470296,
"learning_rate": 1.734591708657533e-05,
"loss": 0.4601,
"step": 207
},
{
"epoch": 0.6295875898600075,
"grad_norm": 0.6082984419772222,
"learning_rate": 1.730992898836672e-05,
"loss": 0.458,
"step": 208
},
{
"epoch": 0.6326144532727961,
"grad_norm": 0.6288608594750572,
"learning_rate": 1.7273736415730488e-05,
"loss": 0.4866,
"step": 209
},
{
"epoch": 0.6356413166855845,
"grad_norm": 0.624761341133163,
"learning_rate": 1.72373403810507e-05,
"loss": 0.5011,
"step": 210
},
{
"epoch": 0.638668180098373,
"grad_norm": 0.5636976950324082,
"learning_rate": 1.720074190240269e-05,
"loss": 0.455,
"step": 211
},
{
"epoch": 0.6416950435111616,
"grad_norm": 0.6045340371994375,
"learning_rate": 1.7163942003524574e-05,
"loss": 0.4496,
"step": 212
},
{
"epoch": 0.64472190692395,
"grad_norm": 0.6246910650251282,
"learning_rate": 1.7126941713788633e-05,
"loss": 0.4687,
"step": 213
},
{
"epoch": 0.6477487703367385,
"grad_norm": 0.6048947001058114,
"learning_rate": 1.70897420681725e-05,
"loss": 0.4829,
"step": 214
},
{
"epoch": 0.6507756337495271,
"grad_norm": 0.6363021906243671,
"learning_rate": 1.7052344107230244e-05,
"loss": 0.4543,
"step": 215
},
{
"epoch": 0.6538024971623155,
"grad_norm": 0.5683959474931025,
"learning_rate": 1.7014748877063212e-05,
"loss": 0.4512,
"step": 216
},
{
"epoch": 0.656829360575104,
"grad_norm": 0.590705297248844,
"learning_rate": 1.697695742929082e-05,
"loss": 0.4765,
"step": 217
},
{
"epoch": 0.6598562239878926,
"grad_norm": 0.6162427119097076,
"learning_rate": 1.693897082102109e-05,
"loss": 0.4874,
"step": 218
},
{
"epoch": 0.662883087400681,
"grad_norm": 0.5762169004257623,
"learning_rate": 1.6900790114821122e-05,
"loss": 0.4632,
"step": 219
},
{
"epoch": 0.6659099508134695,
"grad_norm": 0.6012524947073945,
"learning_rate": 1.686241637868734e-05,
"loss": 0.4125,
"step": 220
},
{
"epoch": 0.6689368142262581,
"grad_norm": 0.5213993348344866,
"learning_rate": 1.682385068601563e-05,
"loss": 0.4436,
"step": 221
},
{
"epoch": 0.6719636776390465,
"grad_norm": 0.6391231575011538,
"learning_rate": 1.6785094115571323e-05,
"loss": 0.4496,
"step": 222
},
{
"epoch": 0.6749905410518351,
"grad_norm": 0.5860781493940427,
"learning_rate": 1.674614775145901e-05,
"loss": 0.4325,
"step": 223
},
{
"epoch": 0.6780174044646236,
"grad_norm": 0.6177729967982134,
"learning_rate": 1.670701268309221e-05,
"loss": 0.4535,
"step": 224
},
{
"epoch": 0.681044267877412,
"grad_norm": 0.592088195171676,
"learning_rate": 1.666769000516292e-05,
"loss": 0.4445,
"step": 225
},
{
"epoch": 0.6840711312902006,
"grad_norm": 0.6032920039386354,
"learning_rate": 1.6628180817610963e-05,
"loss": 0.4652,
"step": 226
},
{
"epoch": 0.687097994702989,
"grad_norm": 0.584142199464464,
"learning_rate": 1.658848622559325e-05,
"loss": 0.4672,
"step": 227
},
{
"epoch": 0.6901248581157775,
"grad_norm": 0.5141795455869701,
"learning_rate": 1.6548607339452853e-05,
"loss": 0.4346,
"step": 228
},
{
"epoch": 0.6931517215285661,
"grad_norm": 0.620751280657942,
"learning_rate": 1.6508545274687936e-05,
"loss": 0.4902,
"step": 229
},
{
"epoch": 0.6961785849413545,
"grad_norm": 0.5307868762464378,
"learning_rate": 1.6468301151920576e-05,
"loss": 0.4298,
"step": 230
},
{
"epoch": 0.699205448354143,
"grad_norm": 0.5693977942565753,
"learning_rate": 1.6427876096865394e-05,
"loss": 0.476,
"step": 231
},
{
"epoch": 0.7022323117669316,
"grad_norm": 0.5976434354939867,
"learning_rate": 1.6387271240298082e-05,
"loss": 0.4648,
"step": 232
},
{
"epoch": 0.70525917517972,
"grad_norm": 0.5338202178939262,
"learning_rate": 1.6346487718023762e-05,
"loss": 0.4324,
"step": 233
},
{
"epoch": 0.7082860385925085,
"grad_norm": 0.5911800531857866,
"learning_rate": 1.6305526670845225e-05,
"loss": 0.4827,
"step": 234
},
{
"epoch": 0.7113129020052971,
"grad_norm": 0.5396182502659621,
"learning_rate": 1.6264389244531015e-05,
"loss": 0.4478,
"step": 235
},
{
"epoch": 0.7143397654180855,
"grad_norm": 0.5323330525505672,
"learning_rate": 1.6223076589783368e-05,
"loss": 0.4096,
"step": 236
},
{
"epoch": 0.717366628830874,
"grad_norm": 0.5701117874434086,
"learning_rate": 1.6181589862206053e-05,
"loss": 0.4736,
"step": 237
},
{
"epoch": 0.7203934922436626,
"grad_norm": 0.5840093135240358,
"learning_rate": 1.613993022227202e-05,
"loss": 0.4528,
"step": 238
},
{
"epoch": 0.723420355656451,
"grad_norm": 0.6024058976210077,
"learning_rate": 1.6098098835290955e-05,
"loss": 0.4756,
"step": 239
},
{
"epoch": 0.7264472190692395,
"grad_norm": 0.6168880034198778,
"learning_rate": 1.6056096871376667e-05,
"loss": 0.4184,
"step": 240
},
{
"epoch": 0.729474082482028,
"grad_norm": 0.550641358597231,
"learning_rate": 1.6013925505414386e-05,
"loss": 0.4439,
"step": 241
},
{
"epoch": 0.7325009458948165,
"grad_norm": 0.6370378803561358,
"learning_rate": 1.5971585917027864e-05,
"loss": 0.4515,
"step": 242
},
{
"epoch": 0.735527809307605,
"grad_norm": 0.5204395799064176,
"learning_rate": 1.5929079290546408e-05,
"loss": 0.4355,
"step": 243
},
{
"epoch": 0.7385546727203935,
"grad_norm": 0.6046579886324439,
"learning_rate": 1.5886406814971728e-05,
"loss": 0.461,
"step": 244
},
{
"epoch": 0.741581536133182,
"grad_norm": 0.5580701920683901,
"learning_rate": 1.584356968394471e-05,
"loss": 0.4376,
"step": 245
},
{
"epoch": 0.7446083995459705,
"grad_norm": 0.5613850747261606,
"learning_rate": 1.5800569095711983e-05,
"loss": 0.4807,
"step": 246
},
{
"epoch": 0.747635262958759,
"grad_norm": 0.5525638182055364,
"learning_rate": 1.575740625309244e-05,
"loss": 0.4687,
"step": 247
},
{
"epoch": 0.7506621263715475,
"grad_norm": 0.581489547733923,
"learning_rate": 1.5714082363443576e-05,
"loss": 0.4757,
"step": 248
},
{
"epoch": 0.753688989784336,
"grad_norm": 0.5665898871787093,
"learning_rate": 1.5670598638627707e-05,
"loss": 0.4571,
"step": 249
},
{
"epoch": 0.7567158531971245,
"grad_norm": 0.5719018697005243,
"learning_rate": 1.5626956294978103e-05,
"loss": 0.4319,
"step": 250
},
{
"epoch": 0.759742716609913,
"grad_norm": 0.5955991048110572,
"learning_rate": 1.5583156553264923e-05,
"loss": 0.4292,
"step": 251
},
{
"epoch": 0.7627695800227015,
"grad_norm": 0.6602276921957603,
"learning_rate": 1.5539200638661106e-05,
"loss": 0.4525,
"step": 252
},
{
"epoch": 0.76579644343549,
"grad_norm": 0.5647884583743515,
"learning_rate": 1.5495089780708062e-05,
"loss": 0.4626,
"step": 253
},
{
"epoch": 0.7688233068482785,
"grad_norm": 0.7054424845478967,
"learning_rate": 1.5450825213281317e-05,
"loss": 0.4577,
"step": 254
},
{
"epoch": 0.771850170261067,
"grad_norm": 0.6207669242248656,
"learning_rate": 1.5406408174555978e-05,
"loss": 0.4207,
"step": 255
},
{
"epoch": 0.7748770336738555,
"grad_norm": 0.5521778957471621,
"learning_rate": 1.5361839906972095e-05,
"loss": 0.4103,
"step": 256
},
{
"epoch": 0.777903897086644,
"grad_norm": 0.7317913688406207,
"learning_rate": 1.531712165719992e-05,
"loss": 0.437,
"step": 257
},
{
"epoch": 0.7809307604994324,
"grad_norm": 0.5188816297108134,
"learning_rate": 1.5272254676105026e-05,
"loss": 0.4605,
"step": 258
},
{
"epoch": 0.783957623912221,
"grad_norm": 0.6252471165524401,
"learning_rate": 1.5227240218713326e-05,
"loss": 0.4353,
"step": 259
},
{
"epoch": 0.7869844873250095,
"grad_norm": 0.6034710012703107,
"learning_rate": 1.5182079544175957e-05,
"loss": 0.4715,
"step": 260
},
{
"epoch": 0.7900113507377979,
"grad_norm": 0.5758162063303118,
"learning_rate": 1.5136773915734067e-05,
"loss": 0.4489,
"step": 261
},
{
"epoch": 0.7930382141505865,
"grad_norm": 0.6351383122797396,
"learning_rate": 1.5091324600683472e-05,
"loss": 0.4473,
"step": 262
},
{
"epoch": 0.796065077563375,
"grad_norm": 0.5791080706096535,
"learning_rate": 1.5045732870339213e-05,
"loss": 0.4865,
"step": 263
},
{
"epoch": 0.7990919409761634,
"grad_norm": 0.5833343821256102,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.4451,
"step": 264
},
{
"epoch": 0.802118804388952,
"grad_norm": 0.6667205638269371,
"learning_rate": 1.4954127268912525e-05,
"loss": 0.4564,
"step": 265
},
{
"epoch": 0.8051456678017405,
"grad_norm": 0.5507926227382565,
"learning_rate": 1.4908115960235683e-05,
"loss": 0.4678,
"step": 266
},
{
"epoch": 0.8081725312145289,
"grad_norm": 0.6280136813805461,
"learning_rate": 1.4861967361004687e-05,
"loss": 0.4628,
"step": 267
},
{
"epoch": 0.8111993946273175,
"grad_norm": 0.5548303535061571,
"learning_rate": 1.4815682762095065e-05,
"loss": 0.4328,
"step": 268
},
{
"epoch": 0.814226258040106,
"grad_norm": 0.6167752775833147,
"learning_rate": 1.476926345818654e-05,
"loss": 0.4505,
"step": 269
},
{
"epoch": 0.8172531214528944,
"grad_norm": 0.5453107837103142,
"learning_rate": 1.472271074772683e-05,
"loss": 0.4374,
"step": 270
},
{
"epoch": 0.820279984865683,
"grad_norm": 0.5661867840834097,
"learning_rate": 1.4676025932895315e-05,
"loss": 0.4416,
"step": 271
},
{
"epoch": 0.8233068482784714,
"grad_norm": 0.5744750357185142,
"learning_rate": 1.4629210319566626e-05,
"loss": 0.4414,
"step": 272
},
{
"epoch": 0.8263337116912599,
"grad_norm": 0.557043473985475,
"learning_rate": 1.4582265217274105e-05,
"loss": 0.4482,
"step": 273
},
{
"epoch": 0.8293605751040485,
"grad_norm": 0.5129476555713144,
"learning_rate": 1.4535191939173179e-05,
"loss": 0.4402,
"step": 274
},
{
"epoch": 0.8323874385168369,
"grad_norm": 0.5663442978416631,
"learning_rate": 1.4487991802004625e-05,
"loss": 0.4488,
"step": 275
},
{
"epoch": 0.8354143019296254,
"grad_norm": 0.5477943321049938,
"learning_rate": 1.4440666126057743e-05,
"loss": 0.4286,
"step": 276
},
{
"epoch": 0.838441165342414,
"grad_norm": 0.5941854410561181,
"learning_rate": 1.4393216235133427e-05,
"loss": 0.4554,
"step": 277
},
{
"epoch": 0.8414680287552024,
"grad_norm": 0.5235597813977861,
"learning_rate": 1.4345643456507126e-05,
"loss": 0.4345,
"step": 278
},
{
"epoch": 0.8444948921679909,
"grad_norm": 0.5950903975671218,
"learning_rate": 1.4297949120891718e-05,
"loss": 0.4396,
"step": 279
},
{
"epoch": 0.8475217555807795,
"grad_norm": 0.5710408902833406,
"learning_rate": 1.4250134562400301e-05,
"loss": 0.4877,
"step": 280
},
{
"epoch": 0.8505486189935679,
"grad_norm": 0.5638725014761413,
"learning_rate": 1.4202201118508863e-05,
"loss": 0.4267,
"step": 281
},
{
"epoch": 0.8535754824063564,
"grad_norm": 0.6225407085931399,
"learning_rate": 1.4154150130018867e-05,
"loss": 0.455,
"step": 282
},
{
"epoch": 0.856602345819145,
"grad_norm": 0.603119894922715,
"learning_rate": 1.4105982941019751e-05,
"loss": 0.4414,
"step": 283
},
{
"epoch": 0.8596292092319334,
"grad_norm": 0.6460924211895752,
"learning_rate": 1.405770089885134e-05,
"loss": 0.5052,
"step": 284
},
{
"epoch": 0.8626560726447219,
"grad_norm": 0.5549576920365494,
"learning_rate": 1.4009305354066138e-05,
"loss": 0.4415,
"step": 285
},
{
"epoch": 0.8656829360575105,
"grad_norm": 0.6062500402242622,
"learning_rate": 1.396079766039157e-05,
"loss": 0.4162,
"step": 286
},
{
"epoch": 0.8687097994702989,
"grad_norm": 0.598862738611488,
"learning_rate": 1.39121791746921e-05,
"loss": 0.4469,
"step": 287
},
{
"epoch": 0.8717366628830874,
"grad_norm": 0.5987421250819631,
"learning_rate": 1.3863451256931286e-05,
"loss": 0.438,
"step": 288
},
{
"epoch": 0.874763526295876,
"grad_norm": 0.6379362856437264,
"learning_rate": 1.381461527013374e-05,
"loss": 0.4386,
"step": 289
},
{
"epoch": 0.8777903897086644,
"grad_norm": 0.5743854647009672,
"learning_rate": 1.3765672580346986e-05,
"loss": 0.4506,
"step": 290
},
{
"epoch": 0.8808172531214529,
"grad_norm": 0.6954210871152465,
"learning_rate": 1.3716624556603275e-05,
"loss": 0.437,
"step": 291
},
{
"epoch": 0.8838441165342414,
"grad_norm": 0.6224537750665278,
"learning_rate": 1.3667472570881264e-05,
"loss": 0.4519,
"step": 292
},
{
"epoch": 0.8868709799470299,
"grad_norm": 0.569215588097282,
"learning_rate": 1.361821799806765e-05,
"loss": 0.4227,
"step": 293
},
{
"epoch": 0.8898978433598184,
"grad_norm": 0.7020677259891825,
"learning_rate": 1.356886221591872e-05,
"loss": 0.4238,
"step": 294
},
{
"epoch": 0.8929247067726069,
"grad_norm": 0.6613381233174982,
"learning_rate": 1.3519406605021797e-05,
"loss": 0.4671,
"step": 295
},
{
"epoch": 0.8959515701853954,
"grad_norm": 0.6658023274346871,
"learning_rate": 1.3469852548756626e-05,
"loss": 0.4282,
"step": 296
},
{
"epoch": 0.8989784335981839,
"grad_norm": 0.6039718483326141,
"learning_rate": 1.342020143325669e-05,
"loss": 0.4487,
"step": 297
},
{
"epoch": 0.9020052970109724,
"grad_norm": 0.5928115354710685,
"learning_rate": 1.3370454647370418e-05,
"loss": 0.4306,
"step": 298
},
{
"epoch": 0.9050321604237609,
"grad_norm": 0.6297118136475328,
"learning_rate": 1.3320613582622354e-05,
"loss": 0.408,
"step": 299
},
{
"epoch": 0.9080590238365494,
"grad_norm": 0.5631704790574225,
"learning_rate": 1.3270679633174219e-05,
"loss": 0.4348,
"step": 300
},
{
"epoch": 0.9110858872493379,
"grad_norm": 0.5601372412079054,
"learning_rate": 1.3220654195785917e-05,
"loss": 0.4384,
"step": 301
},
{
"epoch": 0.9141127506621264,
"grad_norm": 0.6116931393291521,
"learning_rate": 1.3170538669776469e-05,
"loss": 0.4372,
"step": 302
},
{
"epoch": 0.9171396140749148,
"grad_norm": 0.5785602434152506,
"learning_rate": 1.3120334456984871e-05,
"loss": 0.4298,
"step": 303
},
{
"epoch": 0.9201664774877034,
"grad_norm": 0.5864172787075979,
"learning_rate": 1.3070042961730878e-05,
"loss": 0.4419,
"step": 304
},
{
"epoch": 0.9231933409004919,
"grad_norm": 0.5777144316756396,
"learning_rate": 1.3019665590775717e-05,
"loss": 0.4395,
"step": 305
},
{
"epoch": 0.9262202043132803,
"grad_norm": 0.6290443229794741,
"learning_rate": 1.296920375328275e-05,
"loss": 0.415,
"step": 306
},
{
"epoch": 0.9292470677260689,
"grad_norm": 0.604212434657053,
"learning_rate": 1.2918658860778046e-05,
"loss": 0.4586,
"step": 307
},
{
"epoch": 0.9322739311388574,
"grad_norm": 0.5768569815666824,
"learning_rate": 1.2868032327110904e-05,
"loss": 0.4267,
"step": 308
},
{
"epoch": 0.9353007945516458,
"grad_norm": 0.6080643887747446,
"learning_rate": 1.2817325568414299e-05,
"loss": 0.4304,
"step": 309
},
{
"epoch": 0.9383276579644344,
"grad_norm": 0.5949587703622018,
"learning_rate": 1.2766540003065272e-05,
"loss": 0.438,
"step": 310
},
{
"epoch": 0.9413545213772229,
"grad_norm": 0.6020334120966452,
"learning_rate": 1.2715677051645259e-05,
"loss": 0.4111,
"step": 311
},
{
"epoch": 0.9443813847900113,
"grad_norm": 0.6333161140126744,
"learning_rate": 1.266473813690035e-05,
"loss": 0.4273,
"step": 312
},
{
"epoch": 0.9474082482027999,
"grad_norm": 0.5351338292144018,
"learning_rate": 1.2613724683701491e-05,
"loss": 0.415,
"step": 313
},
{
"epoch": 0.9504351116155884,
"grad_norm": 0.5748297206760773,
"learning_rate": 1.2562638119004627e-05,
"loss": 0.4389,
"step": 314
},
{
"epoch": 0.9534619750283768,
"grad_norm": 0.6118311897287021,
"learning_rate": 1.2511479871810792e-05,
"loss": 0.4258,
"step": 315
},
{
"epoch": 0.9564888384411654,
"grad_norm": 0.5724280253975008,
"learning_rate": 1.2460251373126136e-05,
"loss": 0.4368,
"step": 316
},
{
"epoch": 0.9595157018539539,
"grad_norm": 0.5419387606143478,
"learning_rate": 1.2408954055921884e-05,
"loss": 0.454,
"step": 317
},
{
"epoch": 0.9625425652667423,
"grad_norm": 0.6271287805897796,
"learning_rate": 1.2357589355094275e-05,
"loss": 0.4519,
"step": 318
},
{
"epoch": 0.9655694286795309,
"grad_norm": 0.5450391400283195,
"learning_rate": 1.2306158707424402e-05,
"loss": 0.4393,
"step": 319
},
{
"epoch": 0.9685962920923193,
"grad_norm": 0.5938246198781107,
"learning_rate": 1.2254663551538047e-05,
"loss": 0.4359,
"step": 320
},
{
"epoch": 0.9716231555051078,
"grad_norm": 0.6458508616085348,
"learning_rate": 1.2203105327865407e-05,
"loss": 0.4555,
"step": 321
},
{
"epoch": 0.9746500189178964,
"grad_norm": 0.5392315001442343,
"learning_rate": 1.215148547860084e-05,
"loss": 0.4356,
"step": 322
},
{
"epoch": 0.9776768823306848,
"grad_norm": 0.555583837981962,
"learning_rate": 1.2099805447662485e-05,
"loss": 0.4331,
"step": 323
},
{
"epoch": 0.9807037457434733,
"grad_norm": 0.6233182145710846,
"learning_rate": 1.2048066680651908e-05,
"loss": 0.4457,
"step": 324
},
{
"epoch": 0.9837306091562619,
"grad_norm": 0.5706575010137057,
"learning_rate": 1.1996270624813642e-05,
"loss": 0.4809,
"step": 325
},
{
"epoch": 0.9867574725690503,
"grad_norm": 0.5607058562617756,
"learning_rate": 1.194441872899471e-05,
"loss": 0.433,
"step": 326
},
{
"epoch": 0.9897843359818388,
"grad_norm": 0.6337675952394343,
"learning_rate": 1.1892512443604103e-05,
"loss": 0.4092,
"step": 327
},
{
"epoch": 0.9928111993946274,
"grad_norm": 0.5601593624177207,
"learning_rate": 1.1840553220572204e-05,
"loss": 0.4389,
"step": 328
},
{
"epoch": 0.9958380628074158,
"grad_norm": 0.5648069788997735,
"learning_rate": 1.1788542513310178e-05,
"loss": 0.4145,
"step": 329
},
{
"epoch": 0.9988649262202043,
"grad_norm": 0.6295747555116963,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.4647,
"step": 330
},
{
"epoch": 1.0018917896329929,
"grad_norm": 0.6763440004521155,
"learning_rate": 1.1684372466900306e-05,
"loss": 0.3761,
"step": 331
},
{
"epoch": 1.0049186530457812,
"grad_norm": 0.7458101987680528,
"learning_rate": 1.1632216041612595e-05,
"loss": 0.387,
"step": 332
},
{
"epoch": 1.0079455164585698,
"grad_norm": 0.7357311079701176,
"learning_rate": 1.15800139597335e-05,
"loss": 0.3764,
"step": 333
},
{
"epoch": 1.0109723798713584,
"grad_norm": 0.7717686853169455,
"learning_rate": 1.1527767681467472e-05,
"loss": 0.3635,
"step": 334
},
{
"epoch": 1.0139992432841467,
"grad_norm": 0.7065795662400062,
"learning_rate": 1.1475478668255223e-05,
"loss": 0.383,
"step": 335
},
{
"epoch": 1.0170261066969353,
"grad_norm": 0.9177552875156756,
"learning_rate": 1.1423148382732854e-05,
"loss": 0.3348,
"step": 336
},
{
"epoch": 1.0200529701097238,
"grad_norm": 0.6321450925810481,
"learning_rate": 1.1370778288690947e-05,
"loss": 0.3591,
"step": 337
},
{
"epoch": 1.0230798335225122,
"grad_norm": 0.6289265068415445,
"learning_rate": 1.1318369851033604e-05,
"loss": 0.3671,
"step": 338
},
{
"epoch": 1.0261066969353008,
"grad_norm": 0.6840203382918636,
"learning_rate": 1.1265924535737494e-05,
"loss": 0.3454,
"step": 339
},
{
"epoch": 1.0291335603480893,
"grad_norm": 0.6241092308430238,
"learning_rate": 1.121344380981082e-05,
"loss": 0.3334,
"step": 340
},
{
"epoch": 1.0321604237608777,
"grad_norm": 0.6165040083323652,
"learning_rate": 1.1160929141252303e-05,
"loss": 0.3286,
"step": 341
},
{
"epoch": 1.0351872871736663,
"grad_norm": 0.6540210711282685,
"learning_rate": 1.1108381999010111e-05,
"loss": 0.3532,
"step": 342
},
{
"epoch": 1.0382141505864548,
"grad_norm": 0.6361004350122359,
"learning_rate": 1.1055803852940772e-05,
"loss": 0.3448,
"step": 343
},
{
"epoch": 1.0412410139992432,
"grad_norm": 0.5995394272088068,
"learning_rate": 1.1003196173768051e-05,
"loss": 0.3395,
"step": 344
},
{
"epoch": 1.0442678774120318,
"grad_norm": 0.6575253459959216,
"learning_rate": 1.0950560433041825e-05,
"loss": 0.3384,
"step": 345
},
{
"epoch": 1.0472947408248203,
"grad_norm": 0.5986410573344175,
"learning_rate": 1.0897898103096917e-05,
"loss": 0.3403,
"step": 346
},
{
"epoch": 1.0503216042376087,
"grad_norm": 0.6059712807992255,
"learning_rate": 1.0845210657011893e-05,
"loss": 0.3469,
"step": 347
},
{
"epoch": 1.0533484676503972,
"grad_norm": 0.5856765463404571,
"learning_rate": 1.0792499568567885e-05,
"loss": 0.3432,
"step": 348
},
{
"epoch": 1.0563753310631858,
"grad_norm": 0.6060734395289886,
"learning_rate": 1.0739766312207344e-05,
"loss": 0.346,
"step": 349
},
{
"epoch": 1.0594021944759742,
"grad_norm": 0.594419760076603,
"learning_rate": 1.068701236299281e-05,
"loss": 0.3376,
"step": 350
},
{
"epoch": 1.0624290578887627,
"grad_norm": 0.6045655187641306,
"learning_rate": 1.0634239196565646e-05,
"loss": 0.3369,
"step": 351
},
{
"epoch": 1.0654559213015513,
"grad_norm": 0.6392684858621426,
"learning_rate": 1.0581448289104759e-05,
"loss": 0.3431,
"step": 352
},
{
"epoch": 1.0684827847143397,
"grad_norm": 0.5855656032638442,
"learning_rate": 1.0528641117285315e-05,
"loss": 0.3544,
"step": 353
},
{
"epoch": 1.0715096481271282,
"grad_norm": 0.5920665399355434,
"learning_rate": 1.0475819158237426e-05,
"loss": 0.3562,
"step": 354
},
{
"epoch": 1.0745365115399168,
"grad_norm": 0.5693121413786889,
"learning_rate": 1.0422983889504831e-05,
"loss": 0.3339,
"step": 355
},
{
"epoch": 1.0775633749527052,
"grad_norm": 0.6223813639455659,
"learning_rate": 1.0370136789003582e-05,
"loss": 0.3674,
"step": 356
},
{
"epoch": 1.0805902383654937,
"grad_norm": 0.5983549752495387,
"learning_rate": 1.031727933498068e-05,
"loss": 0.3379,
"step": 357
},
{
"epoch": 1.0836171017782823,
"grad_norm": 0.5698102905876657,
"learning_rate": 1.0264413005972736e-05,
"loss": 0.3404,
"step": 358
},
{
"epoch": 1.0866439651910706,
"grad_norm": 0.5628619535872136,
"learning_rate": 1.0211539280764617e-05,
"loss": 0.3228,
"step": 359
},
{
"epoch": 1.0896708286038592,
"grad_norm": 0.5731047846086312,
"learning_rate": 1.015865963834808e-05,
"loss": 0.3421,
"step": 360
},
{
"epoch": 1.0926976920166478,
"grad_norm": 0.6109810528662306,
"learning_rate": 1.0105775557880398e-05,
"loss": 0.3647,
"step": 361
},
{
"epoch": 1.0957245554294364,
"grad_norm": 0.5621973962374506,
"learning_rate": 1.0052888518642978e-05,
"loss": 0.3575,
"step": 362
},
{
"epoch": 1.0987514188422247,
"grad_norm": 0.5870593645318578,
"learning_rate": 1e-05,
"loss": 0.3639,
"step": 363
},
{
"epoch": 1.1017782822550133,
"grad_norm": 0.6192841510938353,
"learning_rate": 9.947111481357023e-06,
"loss": 0.3481,
"step": 364
},
{
"epoch": 1.1048051456678016,
"grad_norm": 0.5708840092102739,
"learning_rate": 9.894224442119606e-06,
"loss": 0.3674,
"step": 365
},
{
"epoch": 1.1078320090805902,
"grad_norm": 0.6270214395862005,
"learning_rate": 9.841340361651921e-06,
"loss": 0.3454,
"step": 366
},
{
"epoch": 1.1108588724933788,
"grad_norm": 0.5751327114907606,
"learning_rate": 9.788460719235386e-06,
"loss": 0.3644,
"step": 367
},
{
"epoch": 1.1138857359061674,
"grad_norm": 0.6593656572115276,
"learning_rate": 9.735586994027267e-06,
"loss": 0.3189,
"step": 368
},
{
"epoch": 1.1169125993189557,
"grad_norm": 0.5697961927811549,
"learning_rate": 9.682720665019325e-06,
"loss": 0.3351,
"step": 369
},
{
"epoch": 1.1199394627317443,
"grad_norm": 0.5783005136573245,
"learning_rate": 9.62986321099642e-06,
"loss": 0.3172,
"step": 370
},
{
"epoch": 1.1229663261445326,
"grad_norm": 0.6602814212616022,
"learning_rate": 9.57701611049517e-06,
"loss": 0.3477,
"step": 371
},
{
"epoch": 1.1259931895573212,
"grad_norm": 0.5466495486897278,
"learning_rate": 9.524180841762577e-06,
"loss": 0.3449,
"step": 372
},
{
"epoch": 1.1290200529701098,
"grad_norm": 0.5898747145363605,
"learning_rate": 9.471358882714687e-06,
"loss": 0.3538,
"step": 373
},
{
"epoch": 1.1320469163828983,
"grad_norm": 0.6055038084990242,
"learning_rate": 9.418551710895243e-06,
"loss": 0.3356,
"step": 374
},
{
"epoch": 1.1350737797956867,
"grad_norm": 0.5711419608727597,
"learning_rate": 9.365760803434356e-06,
"loss": 0.3374,
"step": 375
},
{
"epoch": 1.1381006432084753,
"grad_norm": 0.5582308120665458,
"learning_rate": 9.312987637007191e-06,
"loss": 0.3457,
"step": 376
},
{
"epoch": 1.1411275066212636,
"grad_norm": 0.6228675184029376,
"learning_rate": 9.260233687792657e-06,
"loss": 0.3437,
"step": 377
},
{
"epoch": 1.1441543700340522,
"grad_norm": 0.571541793087105,
"learning_rate": 9.207500431432115e-06,
"loss": 0.3417,
"step": 378
},
{
"epoch": 1.1471812334468408,
"grad_norm": 0.5827669475909949,
"learning_rate": 9.154789342988108e-06,
"loss": 0.3511,
"step": 379
},
{
"epoch": 1.1502080968596293,
"grad_norm": 0.5997040669168835,
"learning_rate": 9.102101896903084e-06,
"loss": 0.3495,
"step": 380
},
{
"epoch": 1.1532349602724177,
"grad_norm": 0.5846826538547297,
"learning_rate": 9.049439566958176e-06,
"loss": 0.3437,
"step": 381
},
{
"epoch": 1.1562618236852062,
"grad_norm": 0.6170682848154051,
"learning_rate": 8.99680382623195e-06,
"loss": 0.35,
"step": 382
},
{
"epoch": 1.1592886870979946,
"grad_norm": 0.5409605082135953,
"learning_rate": 8.944196147059233e-06,
"loss": 0.3397,
"step": 383
},
{
"epoch": 1.1623155505107832,
"grad_norm": 0.5422397824189615,
"learning_rate": 8.89161800098989e-06,
"loss": 0.3198,
"step": 384
},
{
"epoch": 1.1653424139235717,
"grad_norm": 0.6709624091575735,
"learning_rate": 8.839070858747697e-06,
"loss": 0.3709,
"step": 385
},
{
"epoch": 1.1683692773363603,
"grad_norm": 0.5991988690501839,
"learning_rate": 8.786556190189183e-06,
"loss": 0.3378,
"step": 386
},
{
"epoch": 1.1713961407491487,
"grad_norm": 0.5844740298871881,
"learning_rate": 8.734075464262507e-06,
"loss": 0.3385,
"step": 387
},
{
"epoch": 1.1744230041619372,
"grad_norm": 0.5787295522146091,
"learning_rate": 8.681630148966397e-06,
"loss": 0.3504,
"step": 388
},
{
"epoch": 1.1774498675747256,
"grad_norm": 0.5705869702784712,
"learning_rate": 8.629221711309056e-06,
"loss": 0.3472,
"step": 389
},
{
"epoch": 1.1804767309875142,
"grad_norm": 0.5469340675580189,
"learning_rate": 8.576851617267151e-06,
"loss": 0.3262,
"step": 390
},
{
"epoch": 1.1835035944003027,
"grad_norm": 0.6390614574325132,
"learning_rate": 8.52452133174478e-06,
"loss": 0.3338,
"step": 391
},
{
"epoch": 1.1865304578130913,
"grad_norm": 0.5626809617022358,
"learning_rate": 8.472232318532531e-06,
"loss": 0.3522,
"step": 392
},
{
"epoch": 1.1895573212258796,
"grad_norm": 0.6074343592856284,
"learning_rate": 8.419986040266502e-06,
"loss": 0.3335,
"step": 393
},
{
"epoch": 1.1925841846386682,
"grad_norm": 0.6108834302902758,
"learning_rate": 8.367783958387407e-06,
"loss": 0.3634,
"step": 394
},
{
"epoch": 1.1956110480514566,
"grad_norm": 0.5924041613120298,
"learning_rate": 8.315627533099697e-06,
"loss": 0.3352,
"step": 395
},
{
"epoch": 1.1986379114642451,
"grad_norm": 0.5537237413168906,
"learning_rate": 8.263518223330698e-06,
"loss": 0.3397,
"step": 396
},
{
"epoch": 1.2016647748770337,
"grad_norm": 0.5995259959826205,
"learning_rate": 8.211457486689829e-06,
"loss": 0.3288,
"step": 397
},
{
"epoch": 1.2046916382898223,
"grad_norm": 0.586227691301758,
"learning_rate": 8.159446779427798e-06,
"loss": 0.3469,
"step": 398
},
{
"epoch": 1.2077185017026106,
"grad_norm": 0.5721294421911566,
"learning_rate": 8.107487556395902e-06,
"loss": 0.349,
"step": 399
},
{
"epoch": 1.2107453651153992,
"grad_norm": 0.5335405620365062,
"learning_rate": 8.055581271005292e-06,
"loss": 0.3185,
"step": 400
},
{
"epoch": 1.2137722285281876,
"grad_norm": 0.6405685814283789,
"learning_rate": 8.00372937518636e-06,
"loss": 0.3947,
"step": 401
},
{
"epoch": 1.2167990919409761,
"grad_norm": 0.5643907919650019,
"learning_rate": 7.951933319348095e-06,
"loss": 0.335,
"step": 402
},
{
"epoch": 1.2198259553537647,
"grad_norm": 0.5467062061450758,
"learning_rate": 7.900194552337516e-06,
"loss": 0.3385,
"step": 403
},
{
"epoch": 1.2228528187665533,
"grad_norm": 0.618770713655019,
"learning_rate": 7.848514521399167e-06,
"loss": 0.361,
"step": 404
},
{
"epoch": 1.2258796821793416,
"grad_norm": 0.5842007696391234,
"learning_rate": 7.796894672134594e-06,
"loss": 0.3379,
"step": 405
},
{
"epoch": 1.2289065455921302,
"grad_norm": 0.5266206356487769,
"learning_rate": 7.745336448461958e-06,
"loss": 0.3362,
"step": 406
},
{
"epoch": 1.2319334090049185,
"grad_norm": 0.5513791529472379,
"learning_rate": 7.6938412925756e-06,
"loss": 0.3617,
"step": 407
},
{
"epoch": 1.2349602724177071,
"grad_norm": 0.5899581433906347,
"learning_rate": 7.642410644905726e-06,
"loss": 0.347,
"step": 408
},
{
"epoch": 1.2379871358304957,
"grad_norm": 0.5632085415653629,
"learning_rate": 7.591045944078119e-06,
"loss": 0.3227,
"step": 409
},
{
"epoch": 1.2410139992432843,
"grad_norm": 0.646688847920783,
"learning_rate": 7.539748626873866e-06,
"loss": 0.3835,
"step": 410
},
{
"epoch": 1.2440408626560726,
"grad_norm": 0.565815570373247,
"learning_rate": 7.488520128189209e-06,
"loss": 0.3381,
"step": 411
},
{
"epoch": 1.2470677260688612,
"grad_norm": 0.5508218814977722,
"learning_rate": 7.4373618809953755e-06,
"loss": 0.3254,
"step": 412
},
{
"epoch": 1.2500945894816495,
"grad_norm": 0.5775285820935306,
"learning_rate": 7.386275316298513e-06,
"loss": 0.333,
"step": 413
},
{
"epoch": 1.253121452894438,
"grad_norm": 0.5419978291783134,
"learning_rate": 7.335261863099652e-06,
"loss": 0.3254,
"step": 414
},
{
"epoch": 1.2561483163072267,
"grad_norm": 0.5484080196401268,
"learning_rate": 7.2843229483547405e-06,
"loss": 0.3291,
"step": 415
},
{
"epoch": 1.2591751797200152,
"grad_norm": 0.5923578982303893,
"learning_rate": 7.233459996934731e-06,
"loss": 0.3165,
"step": 416
},
{
"epoch": 1.2622020431328036,
"grad_norm": 0.5323464300979155,
"learning_rate": 7.182674431585703e-06,
"loss": 0.3225,
"step": 417
},
{
"epoch": 1.2652289065455922,
"grad_norm": 0.597156918147518,
"learning_rate": 7.131967672889101e-06,
"loss": 0.3452,
"step": 418
},
{
"epoch": 1.2682557699583805,
"grad_norm": 0.5903084563181314,
"learning_rate": 7.081341139221955e-06,
"loss": 0.3604,
"step": 419
},
{
"epoch": 1.271282633371169,
"grad_norm": 0.593569173911432,
"learning_rate": 7.0307962467172555e-06,
"loss": 0.3192,
"step": 420
},
{
"epoch": 1.2743094967839577,
"grad_norm": 0.553599303532681,
"learning_rate": 6.9803344092242855e-06,
"loss": 0.3233,
"step": 421
},
{
"epoch": 1.2773363601967462,
"grad_norm": 0.5722066857943819,
"learning_rate": 6.929957038269123e-06,
"loss": 0.3656,
"step": 422
},
{
"epoch": 1.2803632236095346,
"grad_norm": 0.6229815969094524,
"learning_rate": 6.87966554301513e-06,
"loss": 0.3466,
"step": 423
},
{
"epoch": 1.2833900870223232,
"grad_norm": 0.5726929444576085,
"learning_rate": 6.8294613302235325e-06,
"loss": 0.3389,
"step": 424
},
{
"epoch": 1.2864169504351115,
"grad_norm": 0.5555540574035858,
"learning_rate": 6.779345804214088e-06,
"loss": 0.3345,
"step": 425
},
{
"epoch": 1.2894438138479,
"grad_norm": 0.5553338955130577,
"learning_rate": 6.729320366825785e-06,
"loss": 0.3514,
"step": 426
},
{
"epoch": 1.2924706772606886,
"grad_norm": 0.6243720745316447,
"learning_rate": 6.679386417377649e-06,
"loss": 0.3635,
"step": 427
},
{
"epoch": 1.2954975406734772,
"grad_norm": 0.5531397142233633,
"learning_rate": 6.629545352629583e-06,
"loss": 0.3638,
"step": 428
},
{
"epoch": 1.2985244040862656,
"grad_norm": 0.5613951944973024,
"learning_rate": 6.579798566743314e-06,
"loss": 0.3458,
"step": 429
},
{
"epoch": 1.3015512674990541,
"grad_norm": 0.5100453016571512,
"learning_rate": 6.530147451243377e-06,
"loss": 0.3144,
"step": 430
},
{
"epoch": 1.3045781309118425,
"grad_norm": 0.5686461287674612,
"learning_rate": 6.480593394978208e-06,
"loss": 0.3516,
"step": 431
},
{
"epoch": 1.307604994324631,
"grad_norm": 0.5724419417242387,
"learning_rate": 6.431137784081283e-06,
"loss": 0.3434,
"step": 432
},
{
"epoch": 1.3106318577374196,
"grad_norm": 0.5547824812824002,
"learning_rate": 6.381782001932352e-06,
"loss": 0.3521,
"step": 433
},
{
"epoch": 1.3136587211502082,
"grad_norm": 0.5399423116097835,
"learning_rate": 6.33252742911874e-06,
"loss": 0.333,
"step": 434
},
{
"epoch": 1.3166855845629966,
"grad_norm": 0.5874679725515185,
"learning_rate": 6.283375443396726e-06,
"loss": 0.3382,
"step": 435
},
{
"epoch": 1.3197124479757851,
"grad_norm": 0.5737078845333226,
"learning_rate": 6.234327419653013e-06,
"loss": 0.3577,
"step": 436
},
{
"epoch": 1.3227393113885735,
"grad_norm": 0.5601875807415152,
"learning_rate": 6.185384729866264e-06,
"loss": 0.3342,
"step": 437
},
{
"epoch": 1.325766174801362,
"grad_norm": 0.538551682042874,
"learning_rate": 6.136548743068713e-06,
"loss": 0.324,
"step": 438
},
{
"epoch": 1.3287930382141506,
"grad_norm": 0.5579579164805096,
"learning_rate": 6.087820825307904e-06,
"loss": 0.3477,
"step": 439
},
{
"epoch": 1.3318199016269392,
"grad_norm": 0.5572643172617712,
"learning_rate": 6.039202339608432e-06,
"loss": 0.3391,
"step": 440
},
{
"epoch": 1.3348467650397275,
"grad_norm": 0.5681301827042138,
"learning_rate": 5.990694645933866e-06,
"loss": 0.3646,
"step": 441
},
{
"epoch": 1.3378736284525161,
"grad_norm": 0.5367557097686797,
"learning_rate": 5.9422991011486635e-06,
"loss": 0.3385,
"step": 442
},
{
"epoch": 1.3409004918653045,
"grad_norm": 0.5426137109937247,
"learning_rate": 5.894017058980249e-06,
"loss": 0.3176,
"step": 443
},
{
"epoch": 1.343927355278093,
"grad_norm": 0.5477836510987404,
"learning_rate": 5.845849869981137e-06,
"loss": 0.3138,
"step": 444
},
{
"epoch": 1.3469542186908816,
"grad_norm": 0.5456791910076206,
"learning_rate": 5.797798881491138e-06,
"loss": 0.344,
"step": 445
},
{
"epoch": 1.3499810821036702,
"grad_norm": 0.5520282720983019,
"learning_rate": 5.749865437599703e-06,
"loss": 0.3485,
"step": 446
},
{
"epoch": 1.3530079455164585,
"grad_norm": 0.5521474031345254,
"learning_rate": 5.702050879108284e-06,
"loss": 0.3408,
"step": 447
},
{
"epoch": 1.356034808929247,
"grad_norm": 0.548564658972675,
"learning_rate": 5.654356543492883e-06,
"loss": 0.3433,
"step": 448
},
{
"epoch": 1.3590616723420355,
"grad_norm": 0.5438952579132498,
"learning_rate": 5.606783764866576e-06,
"loss": 0.3551,
"step": 449
},
{
"epoch": 1.362088535754824,
"grad_norm": 0.5318392433791388,
"learning_rate": 5.559333873942259e-06,
"loss": 0.348,
"step": 450
},
{
"epoch": 1.3651153991676126,
"grad_norm": 0.5061705585504341,
"learning_rate": 5.512008197995379e-06,
"loss": 0.3199,
"step": 451
},
{
"epoch": 1.3681422625804012,
"grad_norm": 0.5393172856391282,
"learning_rate": 5.464808060826825e-06,
"loss": 0.3497,
"step": 452
},
{
"epoch": 1.3711691259931895,
"grad_norm": 0.5668109293418759,
"learning_rate": 5.417734782725896e-06,
"loss": 0.3482,
"step": 453
},
{
"epoch": 1.374195989405978,
"grad_norm": 0.544311693084174,
"learning_rate": 5.370789680433376e-06,
"loss": 0.3421,
"step": 454
},
{
"epoch": 1.3772228528187664,
"grad_norm": 0.5157258721726508,
"learning_rate": 5.323974067104687e-06,
"loss": 0.3149,
"step": 455
},
{
"epoch": 1.380249716231555,
"grad_norm": 0.5551022028976371,
"learning_rate": 5.277289252273175e-06,
"loss": 0.3421,
"step": 456
},
{
"epoch": 1.3832765796443436,
"grad_norm": 0.5795019652706563,
"learning_rate": 5.230736541813463e-06,
"loss": 0.3237,
"step": 457
},
{
"epoch": 1.3863034430571322,
"grad_norm": 0.5555972127211857,
"learning_rate": 5.184317237904939e-06,
"loss": 0.3614,
"step": 458
},
{
"epoch": 1.3893303064699205,
"grad_norm": 0.5429781404014025,
"learning_rate": 5.138032638995315e-06,
"loss": 0.3364,
"step": 459
},
{
"epoch": 1.392357169882709,
"grad_norm": 0.5428787651268926,
"learning_rate": 5.091884039764321e-06,
"loss": 0.3308,
"step": 460
},
{
"epoch": 1.3953840332954974,
"grad_norm": 0.546411228561912,
"learning_rate": 5.045872731087479e-06,
"loss": 0.3303,
"step": 461
},
{
"epoch": 1.398410896708286,
"grad_norm": 0.5776370787405458,
"learning_rate": 5.000000000000003e-06,
"loss": 0.3604,
"step": 462
},
{
"epoch": 1.4014377601210746,
"grad_norm": 0.5412848296262482,
"learning_rate": 4.954267129660789e-06,
"loss": 0.3378,
"step": 463
},
{
"epoch": 1.4044646235338631,
"grad_norm": 0.5193209876471725,
"learning_rate": 4.908675399316534e-06,
"loss": 0.3462,
"step": 464
},
{
"epoch": 1.4074914869466515,
"grad_norm": 0.5585630158275061,
"learning_rate": 4.863226084265939e-06,
"loss": 0.3587,
"step": 465
},
{
"epoch": 1.41051835035944,
"grad_norm": 0.534922999597849,
"learning_rate": 4.817920455824045e-06,
"loss": 0.3539,
"step": 466
},
{
"epoch": 1.4135452137722284,
"grad_norm": 0.5603280075048959,
"learning_rate": 4.772759781286679e-06,
"loss": 0.3188,
"step": 467
},
{
"epoch": 1.416572077185017,
"grad_norm": 0.5547404479145099,
"learning_rate": 4.727745323894976e-06,
"loss": 0.3536,
"step": 468
},
{
"epoch": 1.4195989405978056,
"grad_norm": 0.5360397333868736,
"learning_rate": 4.682878342800087e-06,
"loss": 0.3295,
"step": 469
},
{
"epoch": 1.4226258040105941,
"grad_norm": 0.5594639544278357,
"learning_rate": 4.638160093027908e-06,
"loss": 0.3365,
"step": 470
},
{
"epoch": 1.4256526674233825,
"grad_norm": 0.5569096723075034,
"learning_rate": 4.593591825444028e-06,
"loss": 0.3454,
"step": 471
},
{
"epoch": 1.428679530836171,
"grad_norm": 0.5599723522714323,
"learning_rate": 4.549174786718684e-06,
"loss": 0.3337,
"step": 472
},
{
"epoch": 1.4317063942489594,
"grad_norm": 0.5747715584978174,
"learning_rate": 4.504910219291941e-06,
"loss": 0.3443,
"step": 473
},
{
"epoch": 1.434733257661748,
"grad_norm": 0.5693242747624353,
"learning_rate": 4.460799361338898e-06,
"loss": 0.3408,
"step": 474
},
{
"epoch": 1.4377601210745365,
"grad_norm": 0.5951141720172414,
"learning_rate": 4.416843446735077e-06,
"loss": 0.3485,
"step": 475
},
{
"epoch": 1.4407869844873251,
"grad_norm": 0.5369277042296827,
"learning_rate": 4.373043705021899e-06,
"loss": 0.3361,
"step": 476
},
{
"epoch": 1.4438138479001135,
"grad_norm": 0.5363526804804581,
"learning_rate": 4.3294013613722944e-06,
"loss": 0.3615,
"step": 477
},
{
"epoch": 1.446840711312902,
"grad_norm": 0.5620570177865998,
"learning_rate": 4.2859176365564294e-06,
"loss": 0.3358,
"step": 478
},
{
"epoch": 1.4498675747256904,
"grad_norm": 0.5518198256530519,
"learning_rate": 4.2425937469075626e-06,
"loss": 0.3328,
"step": 479
},
{
"epoch": 1.452894438138479,
"grad_norm": 0.5556520541778592,
"learning_rate": 4.19943090428802e-06,
"loss": 0.3376,
"step": 480
},
{
"epoch": 1.4559213015512675,
"grad_norm": 0.5303514005471629,
"learning_rate": 4.1564303160552935e-06,
"loss": 0.3305,
"step": 481
},
{
"epoch": 1.458948164964056,
"grad_norm": 0.533081192300205,
"learning_rate": 4.113593185028273e-06,
"loss": 0.3296,
"step": 482
},
{
"epoch": 1.4619750283768445,
"grad_norm": 0.5219490519033269,
"learning_rate": 4.070920709453597e-06,
"loss": 0.3104,
"step": 483
},
{
"epoch": 1.465001891789633,
"grad_norm": 0.5426105571932429,
"learning_rate": 4.028414082972141e-06,
"loss": 0.3219,
"step": 484
},
{
"epoch": 1.4680287552024214,
"grad_norm": 0.5497656570570758,
"learning_rate": 3.986074494585619e-06,
"loss": 0.3184,
"step": 485
},
{
"epoch": 1.47105561861521,
"grad_norm": 0.5239164687528085,
"learning_rate": 3.943903128623336e-06,
"loss": 0.3188,
"step": 486
},
{
"epoch": 1.4740824820279985,
"grad_norm": 0.516373792095981,
"learning_rate": 3.9019011647090465e-06,
"loss": 0.3219,
"step": 487
},
{
"epoch": 1.477109345440787,
"grad_norm": 0.5367408204940652,
"learning_rate": 3.860069777727983e-06,
"loss": 0.3231,
"step": 488
},
{
"epoch": 1.4801362088535754,
"grad_norm": 0.5496329286629653,
"learning_rate": 3.818410137793947e-06,
"loss": 0.3301,
"step": 489
},
{
"epoch": 1.483163072266364,
"grad_norm": 0.5334420863096035,
"learning_rate": 3.7769234102166365e-06,
"loss": 0.3245,
"step": 490
},
{
"epoch": 1.4861899356791524,
"grad_norm": 0.5323076960619461,
"learning_rate": 3.735610755468988e-06,
"loss": 0.3405,
"step": 491
},
{
"epoch": 1.489216799091941,
"grad_norm": 0.5380616673054802,
"learning_rate": 3.6944733291547784e-06,
"loss": 0.3523,
"step": 492
},
{
"epoch": 1.4922436625047295,
"grad_norm": 0.5817667007710551,
"learning_rate": 3.653512281976238e-06,
"loss": 0.3433,
"step": 493
},
{
"epoch": 1.495270525917518,
"grad_norm": 0.5452858852329028,
"learning_rate": 3.612728759701919e-06,
"loss": 0.3202,
"step": 494
},
{
"epoch": 1.4982973893303064,
"grad_norm": 0.5499290913033745,
"learning_rate": 3.5721239031346067e-06,
"loss": 0.2938,
"step": 495
},
{
"epoch": 1.501324252743095,
"grad_norm": 0.5732273600347847,
"learning_rate": 3.5316988480794255e-06,
"loss": 0.3386,
"step": 496
},
{
"epoch": 1.5043511161558833,
"grad_norm": 0.5669908963512301,
"learning_rate": 3.4914547253120655e-06,
"loss": 0.3552,
"step": 497
},
{
"epoch": 1.507377979568672,
"grad_norm": 0.5344679664912712,
"learning_rate": 3.4513926605471504e-06,
"loss": 0.3259,
"step": 498
},
{
"epoch": 1.5104048429814605,
"grad_norm": 0.5447672402613514,
"learning_rate": 3.4115137744067516e-06,
"loss": 0.3481,
"step": 499
},
{
"epoch": 1.513431706394249,
"grad_norm": 0.5625756836572919,
"learning_rate": 3.37181918238904e-06,
"loss": 0.3693,
"step": 500
},
{
"epoch": 1.5164585698070374,
"grad_norm": 0.5544132310528578,
"learning_rate": 3.3323099948370853e-06,
"loss": 0.3294,
"step": 501
},
{
"epoch": 1.519485433219826,
"grad_norm": 0.5510107028882012,
"learning_rate": 3.292987316907792e-06,
"loss": 0.346,
"step": 502
},
{
"epoch": 1.5225122966326143,
"grad_norm": 0.5397126036092406,
"learning_rate": 3.253852248540994e-06,
"loss": 0.3445,
"step": 503
},
{
"epoch": 1.525539160045403,
"grad_norm": 0.5439152626942458,
"learning_rate": 3.2149058844286796e-06,
"loss": 0.3422,
"step": 504
},
{
"epoch": 1.5285660234581915,
"grad_norm": 0.5424825124147472,
"learning_rate": 3.1761493139843734e-06,
"loss": 0.347,
"step": 505
},
{
"epoch": 1.53159288687098,
"grad_norm": 0.5153711897693964,
"learning_rate": 3.1375836213126653e-06,
"loss": 0.3096,
"step": 506
},
{
"epoch": 1.5346197502837684,
"grad_norm": 0.5469298988447148,
"learning_rate": 3.099209885178882e-06,
"loss": 0.3231,
"step": 507
},
{
"epoch": 1.537646613696557,
"grad_norm": 0.5343635050441716,
"learning_rate": 3.0610291789789094e-06,
"loss": 0.339,
"step": 508
},
{
"epoch": 1.5406734771093453,
"grad_norm": 0.5412668739465106,
"learning_rate": 3.023042570709185e-06,
"loss": 0.3427,
"step": 509
},
{
"epoch": 1.543700340522134,
"grad_norm": 0.5545585816484584,
"learning_rate": 2.9852511229367862e-06,
"loss": 0.3532,
"step": 510
},
{
"epoch": 1.5467272039349225,
"grad_norm": 0.5309574182255647,
"learning_rate": 2.9476558927697605e-06,
"loss": 0.3346,
"step": 511
},
{
"epoch": 1.549754067347711,
"grad_norm": 0.5337453389866483,
"learning_rate": 2.9102579318274994e-06,
"loss": 0.3209,
"step": 512
},
{
"epoch": 1.5527809307604994,
"grad_norm": 0.5909095579929825,
"learning_rate": 2.8730582862113743e-06,
"loss": 0.3577,
"step": 513
},
{
"epoch": 1.555807794173288,
"grad_norm": 0.5313817380647149,
"learning_rate": 2.8360579964754277e-06,
"loss": 0.3166,
"step": 514
},
{
"epoch": 1.5588346575860763,
"grad_norm": 0.5281743372920349,
"learning_rate": 2.7992580975973136e-06,
"loss": 0.3304,
"step": 515
},
{
"epoch": 1.5618615209988649,
"grad_norm": 0.5494324841890594,
"learning_rate": 2.7626596189492983e-06,
"loss": 0.3332,
"step": 516
},
{
"epoch": 1.5648883844116535,
"grad_norm": 0.5344538032398312,
"learning_rate": 2.726263584269513e-06,
"loss": 0.3546,
"step": 517
},
{
"epoch": 1.567915247824442,
"grad_norm": 0.574286734188576,
"learning_rate": 2.690071011633284e-06,
"loss": 0.3393,
"step": 518
},
{
"epoch": 1.5709421112372304,
"grad_norm": 0.5768349285826244,
"learning_rate": 2.6540829134246683e-06,
"loss": 0.3315,
"step": 519
},
{
"epoch": 1.573968974650019,
"grad_norm": 0.5625024418230359,
"learning_rate": 2.618300296308135e-06,
"loss": 0.3509,
"step": 520
},
{
"epoch": 1.5769958380628073,
"grad_norm": 0.5327769230849426,
"learning_rate": 2.582724161200405e-06,
"loss": 0.3344,
"step": 521
},
{
"epoch": 1.5800227014755959,
"grad_norm": 0.5421050697925652,
"learning_rate": 2.5473555032424534e-06,
"loss": 0.3229,
"step": 522
},
{
"epoch": 1.5830495648883844,
"grad_norm": 0.5387385790080916,
"learning_rate": 2.5121953117716744e-06,
"loss": 0.3445,
"step": 523
},
{
"epoch": 1.586076428301173,
"grad_norm": 0.5191243783724272,
"learning_rate": 2.477244570294206e-06,
"loss": 0.3227,
"step": 524
},
{
"epoch": 1.5891032917139614,
"grad_norm": 0.545568450389557,
"learning_rate": 2.4425042564574186e-06,
"loss": 0.3357,
"step": 525
},
{
"epoch": 1.59213015512675,
"grad_norm": 0.5473005411978172,
"learning_rate": 2.4079753420225694e-06,
"loss": 0.3341,
"step": 526
},
{
"epoch": 1.5951570185395383,
"grad_norm": 0.5214560020541606,
"learning_rate": 2.3736587928376197e-06,
"loss": 0.3311,
"step": 527
},
{
"epoch": 1.5981838819523269,
"grad_norm": 0.5121042197817159,
"learning_rate": 2.339555568810221e-06,
"loss": 0.3145,
"step": 528
},
{
"epoch": 1.6012107453651154,
"grad_norm": 0.5354363863049165,
"learning_rate": 2.305666623880858e-06,
"loss": 0.3319,
"step": 529
},
{
"epoch": 1.604237608777904,
"grad_norm": 0.5531462253765186,
"learning_rate": 2.27199290599617e-06,
"loss": 0.3435,
"step": 530
},
{
"epoch": 1.6072644721906924,
"grad_norm": 0.5433550687181965,
"learning_rate": 2.2385353570824308e-06,
"loss": 0.337,
"step": 531
},
{
"epoch": 1.610291335603481,
"grad_norm": 0.5307937589601727,
"learning_rate": 2.2052949130192136e-06,
"loss": 0.3361,
"step": 532
},
{
"epoch": 1.6133181990162693,
"grad_norm": 0.5349935044913042,
"learning_rate": 2.172272503613183e-06,
"loss": 0.3347,
"step": 533
},
{
"epoch": 1.6163450624290578,
"grad_norm": 0.5196297934178653,
"learning_rate": 2.1394690525721275e-06,
"loss": 0.3399,
"step": 534
},
{
"epoch": 1.6193719258418464,
"grad_norm": 0.5297545299951213,
"learning_rate": 2.1068854774790783e-06,
"loss": 0.3423,
"step": 535
},
{
"epoch": 1.622398789254635,
"grad_norm": 0.5514786858614504,
"learning_rate": 2.0745226897666858e-06,
"loss": 0.3441,
"step": 536
},
{
"epoch": 1.6254256526674233,
"grad_norm": 0.5369609797548685,
"learning_rate": 2.0423815946916783e-06,
"loss": 0.3412,
"step": 537
},
{
"epoch": 1.628452516080212,
"grad_norm": 0.5626768055990692,
"learning_rate": 2.010463091309587e-06,
"loss": 0.3312,
"step": 538
},
{
"epoch": 1.6314793794930003,
"grad_norm": 0.5409337306023044,
"learning_rate": 1.9787680724495617e-06,
"loss": 0.3626,
"step": 539
},
{
"epoch": 1.6345062429057888,
"grad_norm": 0.5694383205006677,
"learning_rate": 1.947297424689414e-06,
"loss": 0.3339,
"step": 540
},
{
"epoch": 1.6375331063185774,
"grad_norm": 0.5350701082959706,
"learning_rate": 1.9160520283308115e-06,
"loss": 0.3256,
"step": 541
},
{
"epoch": 1.640559969731366,
"grad_norm": 0.5271503018364957,
"learning_rate": 1.8850327573746584e-06,
"loss": 0.3321,
"step": 542
},
{
"epoch": 1.6435868331441543,
"grad_norm": 0.5326263939812926,
"learning_rate": 1.854240479496643e-06,
"loss": 0.3256,
"step": 543
},
{
"epoch": 1.646613696556943,
"grad_norm": 0.5895426836301729,
"learning_rate": 1.8236760560229715e-06,
"loss": 0.338,
"step": 544
},
{
"epoch": 1.6496405599697312,
"grad_norm": 0.5484900166169228,
"learning_rate": 1.7933403419062689e-06,
"loss": 0.3459,
"step": 545
},
{
"epoch": 1.6526674233825198,
"grad_norm": 0.5531195417760699,
"learning_rate": 1.7632341857016733e-06,
"loss": 0.3348,
"step": 546
},
{
"epoch": 1.6556942867953084,
"grad_norm": 0.5226837919983212,
"learning_rate": 1.7333584295430894e-06,
"loss": 0.315,
"step": 547
},
{
"epoch": 1.658721150208097,
"grad_norm": 0.545918979119333,
"learning_rate": 1.7037139091196396e-06,
"loss": 0.334,
"step": 548
},
{
"epoch": 1.6617480136208853,
"grad_norm": 0.5111664636037931,
"learning_rate": 1.6743014536522872e-06,
"loss": 0.3535,
"step": 549
},
{
"epoch": 1.6647748770336739,
"grad_norm": 0.5368881403465131,
"learning_rate": 1.6451218858706374e-06,
"loss": 0.3339,
"step": 550
},
{
"epoch": 1.6678017404464622,
"grad_norm": 0.5576209299099644,
"learning_rate": 1.616176021989926e-06,
"loss": 0.3376,
"step": 551
},
{
"epoch": 1.6708286038592508,
"grad_norm": 0.5089944495209637,
"learning_rate": 1.587464671688187e-06,
"loss": 0.3377,
"step": 552
},
{
"epoch": 1.6738554672720394,
"grad_norm": 0.5714316099941706,
"learning_rate": 1.558988638083616e-06,
"loss": 0.3445,
"step": 553
},
{
"epoch": 1.676882330684828,
"grad_norm": 0.5135280939181603,
"learning_rate": 1.5307487177120773e-06,
"loss": 0.3438,
"step": 554
},
{
"epoch": 1.6799091940976163,
"grad_norm": 0.5642811750118364,
"learning_rate": 1.5027457005048573e-06,
"loss": 0.3567,
"step": 555
},
{
"epoch": 1.6829360575104049,
"grad_norm": 0.5199707102235239,
"learning_rate": 1.4749803697665366e-06,
"loss": 0.3441,
"step": 556
},
{
"epoch": 1.6859629209231932,
"grad_norm": 0.5205676338074903,
"learning_rate": 1.4474535021531099e-06,
"loss": 0.3358,
"step": 557
},
{
"epoch": 1.6889897843359818,
"grad_norm": 0.5362996155339582,
"learning_rate": 1.4201658676502294e-06,
"loss": 0.3362,
"step": 558
},
{
"epoch": 1.6920166477487704,
"grad_norm": 0.548955323984696,
"learning_rate": 1.3931182295516965e-06,
"loss": 0.3464,
"step": 559
},
{
"epoch": 1.695043511161559,
"grad_norm": 0.5289697187345612,
"learning_rate": 1.3663113444380905e-06,
"loss": 0.3184,
"step": 560
},
{
"epoch": 1.6980703745743473,
"grad_norm": 0.5298888577780704,
"learning_rate": 1.339745962155613e-06,
"loss": 0.3235,
"step": 561
},
{
"epoch": 1.7010972379871359,
"grad_norm": 0.541301268416102,
"learning_rate": 1.3134228257951142e-06,
"loss": 0.3415,
"step": 562
},
{
"epoch": 1.7041241013999242,
"grad_norm": 0.531286474847516,
"learning_rate": 1.2873426716713012e-06,
"loss": 0.3332,
"step": 563
},
{
"epoch": 1.7071509648127128,
"grad_norm": 0.5523472234916659,
"learning_rate": 1.2615062293021508e-06,
"loss": 0.3318,
"step": 564
},
{
"epoch": 1.7101778282255014,
"grad_norm": 0.5095391750771804,
"learning_rate": 1.2359142213884933e-06,
"loss": 0.3148,
"step": 565
},
{
"epoch": 1.71320469163829,
"grad_norm": 0.5066186192515475,
"learning_rate": 1.2105673637938054e-06,
"loss": 0.3304,
"step": 566
},
{
"epoch": 1.7162315550510783,
"grad_norm": 0.516379881184284,
"learning_rate": 1.1854663655241804e-06,
"loss": 0.3389,
"step": 567
},
{
"epoch": 1.7192584184638668,
"grad_norm": 0.5194691989074978,
"learning_rate": 1.1606119287084982e-06,
"loss": 0.3227,
"step": 568
},
{
"epoch": 1.7222852818766552,
"grad_norm": 0.5234492179327196,
"learning_rate": 1.136004748578785e-06,
"loss": 0.3329,
"step": 569
},
{
"epoch": 1.7253121452894438,
"grad_norm": 0.5419676642514104,
"learning_rate": 1.1116455134507665e-06,
"loss": 0.326,
"step": 570
},
{
"epoch": 1.7283390087022323,
"grad_norm": 0.5034568150745946,
"learning_rate": 1.0875349047046113e-06,
"loss": 0.323,
"step": 571
},
{
"epoch": 1.731365872115021,
"grad_norm": 0.5255454997274771,
"learning_rate": 1.0636735967658785e-06,
"loss": 0.3419,
"step": 572
},
{
"epoch": 1.7343927355278093,
"grad_norm": 0.5240983345143586,
"learning_rate": 1.0400622570866426e-06,
"loss": 0.3359,
"step": 573
},
{
"epoch": 1.7374195989405978,
"grad_norm": 0.5211210443855394,
"learning_rate": 1.0167015461268303e-06,
"loss": 0.3489,
"step": 574
},
{
"epoch": 1.7404464623533862,
"grad_norm": 0.5025003403433331,
"learning_rate": 9.935921173357444e-07,
"loss": 0.331,
"step": 575
},
{
"epoch": 1.7434733257661748,
"grad_norm": 0.535412370103588,
"learning_rate": 9.707346171337895e-07,
"loss": 0.3455,
"step": 576
},
{
"epoch": 1.7465001891789633,
"grad_norm": 0.5405804985692432,
"learning_rate": 9.481296848943744e-07,
"loss": 0.3613,
"step": 577
},
{
"epoch": 1.749527052591752,
"grad_norm": 0.5171554201549999,
"learning_rate": 9.257779529260558e-07,
"loss": 0.3374,
"step": 578
},
{
"epoch": 1.7525539160045402,
"grad_norm": 0.541760899431086,
"learning_rate": 9.036800464548157e-07,
"loss": 0.3363,
"step": 579
},
{
"epoch": 1.7555807794173288,
"grad_norm": 0.5284347732678949,
"learning_rate": 8.818365836066101e-07,
"loss": 0.3371,
"step": 580
},
{
"epoch": 1.7586076428301172,
"grad_norm": 0.515647569313736,
"learning_rate": 8.602481753900427e-07,
"loss": 0.3343,
"step": 581
},
{
"epoch": 1.7616345062429057,
"grad_norm": 0.5493374263487832,
"learning_rate": 8.389154256793042e-07,
"loss": 0.3372,
"step": 582
},
{
"epoch": 1.7646613696556943,
"grad_norm": 0.520056067438565,
"learning_rate": 8.178389311972612e-07,
"loss": 0.3076,
"step": 583
},
{
"epoch": 1.7676882330684829,
"grad_norm": 0.515783537561946,
"learning_rate": 7.970192814987676e-07,
"loss": 0.3013,
"step": 584
},
{
"epoch": 1.7707150964812712,
"grad_norm": 0.511263622854028,
"learning_rate": 7.764570589541876e-07,
"loss": 0.3291,
"step": 585
},
{
"epoch": 1.7737419598940598,
"grad_norm": 0.5022716662793987,
"learning_rate": 7.561528387330797e-07,
"loss": 0.3165,
"step": 586
},
{
"epoch": 1.7767688233068482,
"grad_norm": 0.49706334384218254,
"learning_rate": 7.361071887881376e-07,
"loss": 0.3069,
"step": 587
},
{
"epoch": 1.7797956867196367,
"grad_norm": 0.5147167557332147,
"learning_rate": 7.163206698392744e-07,
"loss": 0.323,
"step": 588
},
{
"epoch": 1.7828225501324253,
"grad_norm": 0.5694318054287043,
"learning_rate": 6.96793835357964e-07,
"loss": 0.3546,
"step": 589
},
{
"epoch": 1.7858494135452139,
"grad_norm": 0.5114207026118898,
"learning_rate": 6.775272315517423e-07,
"loss": 0.319,
"step": 590
},
{
"epoch": 1.7888762769580022,
"grad_norm": 0.5176754574147087,
"learning_rate": 6.585213973489335e-07,
"loss": 0.3323,
"step": 591
},
{
"epoch": 1.7919031403707908,
"grad_norm": 0.5161157484947948,
"learning_rate": 6.397768643835755e-07,
"loss": 0.3291,
"step": 592
},
{
"epoch": 1.7949300037835791,
"grad_norm": 0.5447678048296611,
"learning_rate": 6.212941569805508e-07,
"loss": 0.3609,
"step": 593
},
{
"epoch": 1.7979568671963677,
"grad_norm": 0.4946458588220733,
"learning_rate": 6.030737921409169e-07,
"loss": 0.3171,
"step": 594
},
{
"epoch": 1.8009837306091563,
"grad_norm": 0.5234463435365085,
"learning_rate": 5.851162795274445e-07,
"loss": 0.3178,
"step": 595
},
{
"epoch": 1.8040105940219449,
"grad_norm": 0.5293792192577068,
"learning_rate": 5.674221214503639e-07,
"loss": 0.3126,
"step": 596
},
{
"epoch": 1.8070374574347332,
"grad_norm": 0.5277044034885998,
"learning_rate": 5.499918128533155e-07,
"loss": 0.3313,
"step": 597
},
{
"epoch": 1.8100643208475218,
"grad_norm": 0.5092427859911447,
"learning_rate": 5.328258412994958e-07,
"loss": 0.3239,
"step": 598
},
{
"epoch": 1.8130911842603101,
"grad_norm": 0.5149352616221561,
"learning_rate": 5.159246869580348e-07,
"loss": 0.3279,
"step": 599
},
{
"epoch": 1.8161180476730987,
"grad_norm": 0.5419286189342575,
"learning_rate": 4.992888225905467e-07,
"loss": 0.3437,
"step": 600
},
{
"epoch": 1.8191449110858873,
"grad_norm": 0.5324093053643671,
"learning_rate": 4.829187135379221e-07,
"loss": 0.359,
"step": 601
},
{
"epoch": 1.8221717744986758,
"grad_norm": 0.5201206845094479,
"learning_rate": 4.6681481770729844e-07,
"loss": 0.3346,
"step": 602
},
{
"epoch": 1.8251986379114642,
"grad_norm": 0.5521218832839953,
"learning_rate": 4.509775855592613e-07,
"loss": 0.3448,
"step": 603
},
{
"epoch": 1.8282255013242528,
"grad_norm": 0.5015868720849281,
"learning_rate": 4.354074600952407e-07,
"loss": 0.3078,
"step": 604
},
{
"epoch": 1.8312523647370411,
"grad_norm": 0.5243157913445379,
"learning_rate": 4.2010487684511105e-07,
"loss": 0.338,
"step": 605
},
{
"epoch": 1.8342792281498297,
"grad_norm": 0.49840216223213707,
"learning_rate": 4.0507026385502747e-07,
"loss": 0.3216,
"step": 606
},
{
"epoch": 1.8373060915626183,
"grad_norm": 0.5213845291767036,
"learning_rate": 3.9030404167542777e-07,
"loss": 0.3222,
"step": 607
},
{
"epoch": 1.8403329549754068,
"grad_norm": 0.5403603786906672,
"learning_rate": 3.7580662334929517e-07,
"loss": 0.3265,
"step": 608
},
{
"epoch": 1.8433598183881952,
"grad_norm": 0.49776468943532864,
"learning_rate": 3.615784144005796e-07,
"loss": 0.3171,
"step": 609
},
{
"epoch": 1.8463866818009838,
"grad_norm": 0.5257081748863395,
"learning_rate": 3.476198128228736e-07,
"loss": 0.3397,
"step": 610
},
{
"epoch": 1.849413545213772,
"grad_norm": 0.5539772514631744,
"learning_rate": 3.339312090682689e-07,
"loss": 0.3204,
"step": 611
},
{
"epoch": 1.8524404086265607,
"grad_norm": 0.4977270342121302,
"learning_rate": 3.2051298603643754e-07,
"loss": 0.3242,
"step": 612
},
{
"epoch": 1.8554672720393492,
"grad_norm": 0.5067179030194061,
"learning_rate": 3.0736551906392354e-07,
"loss": 0.3181,
"step": 613
},
{
"epoch": 1.8584941354521378,
"grad_norm": 0.5212549687159126,
"learning_rate": 2.9448917591363923e-07,
"loss": 0.3353,
"step": 614
},
{
"epoch": 1.8615209988649262,
"grad_norm": 0.4955350740015368,
"learning_rate": 2.818843167645835e-07,
"loss": 0.3247,
"step": 615
},
{
"epoch": 1.8645478622777147,
"grad_norm": 0.5062996754742579,
"learning_rate": 2.6955129420176193e-07,
"loss": 0.3144,
"step": 616
},
{
"epoch": 1.867574725690503,
"grad_norm": 0.5125940645385597,
"learning_rate": 2.5749045320632824e-07,
"loss": 0.3163,
"step": 617
},
{
"epoch": 1.8706015891032917,
"grad_norm": 0.5141111700527322,
"learning_rate": 2.4570213114592957e-07,
"loss": 0.3114,
"step": 618
},
{
"epoch": 1.8736284525160802,
"grad_norm": 0.5121904268534225,
"learning_rate": 2.3418665776527738e-07,
"loss": 0.3541,
"step": 619
},
{
"epoch": 1.8766553159288688,
"grad_norm": 0.5231636218499177,
"learning_rate": 2.2294435517691504e-07,
"loss": 0.3407,
"step": 620
},
{
"epoch": 1.8796821793416572,
"grad_norm": 0.5046178923309044,
"learning_rate": 2.119755378522137e-07,
"loss": 0.3301,
"step": 621
},
{
"epoch": 1.8827090427544457,
"grad_norm": 0.5401514722744142,
"learning_rate": 2.0128051261257165e-07,
"loss": 0.3626,
"step": 622
},
{
"epoch": 1.885735906167234,
"grad_norm": 0.48001410503161174,
"learning_rate": 1.908595786208367e-07,
"loss": 0.2978,
"step": 623
},
{
"epoch": 1.8887627695800226,
"grad_norm": 0.5086952825902071,
"learning_rate": 1.8071302737293294e-07,
"loss": 0.3226,
"step": 624
},
{
"epoch": 1.8917896329928112,
"grad_norm": 0.51829699327428,
"learning_rate": 1.7084114268971275e-07,
"loss": 0.3333,
"step": 625
},
{
"epoch": 1.8948164964055998,
"grad_norm": 0.5148122297877614,
"learning_rate": 1.612442007090076e-07,
"loss": 0.3305,
"step": 626
},
{
"epoch": 1.8978433598183881,
"grad_norm": 0.5051932238631632,
"learning_rate": 1.519224698779198e-07,
"loss": 0.3167,
"step": 627
},
{
"epoch": 1.9008702232311767,
"grad_norm": 0.5213823161183292,
"learning_rate": 1.4287621094529524e-07,
"loss": 0.3319,
"step": 628
},
{
"epoch": 1.903897086643965,
"grad_norm": 0.5407521383797909,
"learning_rate": 1.3410567695444576e-07,
"loss": 0.3408,
"step": 629
},
{
"epoch": 1.9069239500567536,
"grad_norm": 0.5013542586668778,
"learning_rate": 1.2561111323605714e-07,
"loss": 0.3196,
"step": 630
},
{
"epoch": 1.9099508134695422,
"grad_norm": 0.5235081553246185,
"learning_rate": 1.1739275740134004e-07,
"loss": 0.3381,
"step": 631
},
{
"epoch": 1.9129776768823308,
"grad_norm": 0.5476725963638974,
"learning_rate": 1.0945083933537104e-07,
"loss": 0.3532,
"step": 632
},
{
"epoch": 1.9160045402951191,
"grad_norm": 0.5373293708500385,
"learning_rate": 1.0178558119067316e-07,
"loss": 0.3428,
"step": 633
},
{
"epoch": 1.9190314037079077,
"grad_norm": 0.5173819725119972,
"learning_rate": 9.439719738099318e-08,
"loss": 0.3557,
"step": 634
},
{
"epoch": 1.922058267120696,
"grad_norm": 0.4929227055067737,
"learning_rate": 8.728589457530857e-08,
"loss": 0.3093,
"step": 635
},
{
"epoch": 1.9250851305334846,
"grad_norm": 0.5057133944809219,
"learning_rate": 8.04518716920466e-08,
"loss": 0.3304,
"step": 636
},
{
"epoch": 1.9281119939462732,
"grad_norm": 0.5236307503553929,
"learning_rate": 7.389531989351773e-08,
"loss": 0.3395,
"step": 637
},
{
"epoch": 1.9311388573590618,
"grad_norm": 0.5118149547852485,
"learning_rate": 6.761642258056977e-08,
"loss": 0.3097,
"step": 638
},
{
"epoch": 1.9341657207718501,
"grad_norm": 0.5293203250244016,
"learning_rate": 6.161535538745877e-08,
"loss": 0.3147,
"step": 639
},
{
"epoch": 1.9371925841846387,
"grad_norm": 0.5243005861352074,
"learning_rate": 5.5892286176932875e-08,
"loss": 0.311,
"step": 640
},
{
"epoch": 1.940219447597427,
"grad_norm": 0.5025317767629421,
"learning_rate": 5.044737503554165e-08,
"loss": 0.3321,
"step": 641
},
{
"epoch": 1.9432463110102156,
"grad_norm": 0.510055518279952,
"learning_rate": 4.528077426915412e-08,
"loss": 0.3429,
"step": 642
},
{
"epoch": 1.9462731744230042,
"grad_norm": 0.4983351520579858,
"learning_rate": 4.0392628398699954e-08,
"loss": 0.316,
"step": 643
},
{
"epoch": 1.9493000378357928,
"grad_norm": 0.51961463710477,
"learning_rate": 3.578307415612714e-08,
"loss": 0.3306,
"step": 644
},
{
"epoch": 1.952326901248581,
"grad_norm": 0.5241069831320792,
"learning_rate": 3.1452240480577265e-08,
"loss": 0.3366,
"step": 645
},
{
"epoch": 1.9553537646613697,
"grad_norm": 0.5076726777197932,
"learning_rate": 2.7400248514776184e-08,
"loss": 0.311,
"step": 646
},
{
"epoch": 1.958380628074158,
"grad_norm": 0.5270372365001028,
"learning_rate": 2.3627211601651157e-08,
"loss": 0.327,
"step": 647
},
{
"epoch": 1.9614074914869466,
"grad_norm": 0.5345706651089321,
"learning_rate": 2.013323528115674e-08,
"loss": 0.3305,
"step": 648
},
{
"epoch": 1.9644343548997352,
"grad_norm": 0.5701176651055055,
"learning_rate": 1.6918417287318245e-08,
"loss": 0.3597,
"step": 649
},
{
"epoch": 1.9674612183125237,
"grad_norm": 0.5096275397795988,
"learning_rate": 1.3982847545507271e-08,
"loss": 0.3191,
"step": 650
},
{
"epoch": 1.970488081725312,
"grad_norm": 0.5409876712500923,
"learning_rate": 1.1326608169920373e-08,
"loss": 0.3404,
"step": 651
},
{
"epoch": 1.9735149451381007,
"grad_norm": 0.5203072840312353,
"learning_rate": 8.949773461282008e-09,
"loss": 0.3225,
"step": 652
},
{
"epoch": 1.976541808550889,
"grad_norm": 0.5320993369980339,
"learning_rate": 6.8524099047695415e-09,
"loss": 0.348,
"step": 653
},
{
"epoch": 1.9795686719636776,
"grad_norm": 0.5094907409675443,
"learning_rate": 5.034576168149175e-09,
"loss": 0.3287,
"step": 654
},
{
"epoch": 1.9825955353764662,
"grad_norm": 0.507089285164817,
"learning_rate": 3.4963231001383657e-09,
"loss": 0.336,
"step": 655
},
{
"epoch": 1.9856223987892547,
"grad_norm": 0.5138496310370845,
"learning_rate": 2.237693728981416e-09,
"loss": 0.3508,
"step": 656
},
{
"epoch": 1.988649262202043,
"grad_norm": 0.5325962366267458,
"learning_rate": 1.2587232612493172e-09,
"loss": 0.338,
"step": 657
},
{
"epoch": 1.9916761256148316,
"grad_norm": 0.5281655407968793,
"learning_rate": 5.594390808494332e-10,
"loss": 0.3279,
"step": 658
},
{
"epoch": 1.99470298902762,
"grad_norm": 0.5236655600734001,
"learning_rate": 1.3986074826388697e-10,
"loss": 0.3421,
"step": 659
},
{
"epoch": 1.9977298524404086,
"grad_norm": 0.49841312880749544,
"learning_rate": 0.0,
"loss": 0.3118,
"step": 660
},
{
"epoch": 1.9977298524404086,
"step": 660,
"total_flos": 2.708198372933632e+17,
"train_loss": 0.41113431255022687,
"train_runtime": 20533.8574,
"train_samples_per_second": 4.117,
"train_steps_per_second": 0.032
}
],
"logging_steps": 1,
"max_steps": 660,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.708198372933632e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}