c1_science_10d_1s / trainer_state.json
neginr's picture
End of training
14694d1 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.989367088607595,
"eval_steps": 500,
"global_step": 1230,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0040547389761784085,
"grad_norm": 6.72721516957798,
"learning_rate": 3.2520325203252037e-07,
"loss": 1.0577,
"step": 1
},
{
"epoch": 0.008109477952356817,
"grad_norm": 6.46823017577869,
"learning_rate": 6.504065040650407e-07,
"loss": 1.0413,
"step": 2
},
{
"epoch": 0.012164216928535226,
"grad_norm": 6.542268116316752,
"learning_rate": 9.75609756097561e-07,
"loss": 1.0997,
"step": 3
},
{
"epoch": 0.016218955904713634,
"grad_norm": 6.510945638929656,
"learning_rate": 1.3008130081300815e-06,
"loss": 1.0543,
"step": 4
},
{
"epoch": 0.02027369488089204,
"grad_norm": 6.170439325045293,
"learning_rate": 1.6260162601626018e-06,
"loss": 1.0698,
"step": 5
},
{
"epoch": 0.024328433857070453,
"grad_norm": 5.957792791790743,
"learning_rate": 1.951219512195122e-06,
"loss": 1.0583,
"step": 6
},
{
"epoch": 0.02838317283324886,
"grad_norm": 4.818340065759815,
"learning_rate": 2.2764227642276426e-06,
"loss": 1.0211,
"step": 7
},
{
"epoch": 0.03243791180942727,
"grad_norm": 4.500911870744693,
"learning_rate": 2.601626016260163e-06,
"loss": 1.008,
"step": 8
},
{
"epoch": 0.03649265078560568,
"grad_norm": 2.569671445653381,
"learning_rate": 2.926829268292683e-06,
"loss": 0.9531,
"step": 9
},
{
"epoch": 0.04054738976178408,
"grad_norm": 2.2777855155348594,
"learning_rate": 3.2520325203252037e-06,
"loss": 0.96,
"step": 10
},
{
"epoch": 0.044602128737962494,
"grad_norm": 2.1423184761928775,
"learning_rate": 3.577235772357724e-06,
"loss": 0.9495,
"step": 11
},
{
"epoch": 0.048656867714140906,
"grad_norm": 4.503299842746655,
"learning_rate": 3.902439024390244e-06,
"loss": 0.9885,
"step": 12
},
{
"epoch": 0.05271160669031931,
"grad_norm": 4.449491752371554,
"learning_rate": 4.227642276422765e-06,
"loss": 0.9298,
"step": 13
},
{
"epoch": 0.05676634566649772,
"grad_norm": 4.687406736300778,
"learning_rate": 4.552845528455285e-06,
"loss": 0.9581,
"step": 14
},
{
"epoch": 0.060821084642676125,
"grad_norm": 4.179566361341069,
"learning_rate": 4.8780487804878055e-06,
"loss": 0.9229,
"step": 15
},
{
"epoch": 0.06487582361885454,
"grad_norm": 3.3430878107594646,
"learning_rate": 5.203252032520326e-06,
"loss": 0.9136,
"step": 16
},
{
"epoch": 0.06893056259503294,
"grad_norm": 3.1151188640810967,
"learning_rate": 5.528455284552846e-06,
"loss": 0.8693,
"step": 17
},
{
"epoch": 0.07298530157121136,
"grad_norm": 2.8364276261767274,
"learning_rate": 5.853658536585366e-06,
"loss": 0.8641,
"step": 18
},
{
"epoch": 0.07704004054738976,
"grad_norm": 2.3652276275629927,
"learning_rate": 6.178861788617887e-06,
"loss": 0.8179,
"step": 19
},
{
"epoch": 0.08109477952356817,
"grad_norm": 2.058477206053389,
"learning_rate": 6.504065040650407e-06,
"loss": 0.8681,
"step": 20
},
{
"epoch": 0.08514951849974658,
"grad_norm": 1.8143531779298494,
"learning_rate": 6.829268292682928e-06,
"loss": 0.824,
"step": 21
},
{
"epoch": 0.08920425747592499,
"grad_norm": 1.8171647874407701,
"learning_rate": 7.154471544715448e-06,
"loss": 0.8199,
"step": 22
},
{
"epoch": 0.0932589964521034,
"grad_norm": 1.867787038405791,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.8015,
"step": 23
},
{
"epoch": 0.09731373542828181,
"grad_norm": 1.8661866727755845,
"learning_rate": 7.804878048780489e-06,
"loss": 0.8087,
"step": 24
},
{
"epoch": 0.10136847440446022,
"grad_norm": 1.5187077254983465,
"learning_rate": 8.130081300813009e-06,
"loss": 0.7823,
"step": 25
},
{
"epoch": 0.10542321338063862,
"grad_norm": 1.320351830303469,
"learning_rate": 8.45528455284553e-06,
"loss": 0.8034,
"step": 26
},
{
"epoch": 0.10947795235681702,
"grad_norm": 1.3710915571363496,
"learning_rate": 8.78048780487805e-06,
"loss": 0.7643,
"step": 27
},
{
"epoch": 0.11353269133299544,
"grad_norm": 1.4574618640892707,
"learning_rate": 9.10569105691057e-06,
"loss": 0.7792,
"step": 28
},
{
"epoch": 0.11758743030917385,
"grad_norm": 1.161070280425468,
"learning_rate": 9.43089430894309e-06,
"loss": 0.7906,
"step": 29
},
{
"epoch": 0.12164216928535225,
"grad_norm": 1.0272009235272497,
"learning_rate": 9.756097560975611e-06,
"loss": 0.7727,
"step": 30
},
{
"epoch": 0.12569690826153065,
"grad_norm": 1.05818562644255,
"learning_rate": 1.008130081300813e-05,
"loss": 0.7917,
"step": 31
},
{
"epoch": 0.12975164723770907,
"grad_norm": 0.9849077717276795,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.7549,
"step": 32
},
{
"epoch": 0.1338063862138875,
"grad_norm": 0.9190339900565674,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.764,
"step": 33
},
{
"epoch": 0.13786112519006588,
"grad_norm": 0.7710236672465931,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.7592,
"step": 34
},
{
"epoch": 0.1419158641662443,
"grad_norm": 0.8569230882968184,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.7547,
"step": 35
},
{
"epoch": 0.14597060314242272,
"grad_norm": 0.8436470681793244,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.7762,
"step": 36
},
{
"epoch": 0.1500253421186011,
"grad_norm": 0.6634761360025331,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.7378,
"step": 37
},
{
"epoch": 0.15408008109477953,
"grad_norm": 0.7360792589692458,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.7425,
"step": 38
},
{
"epoch": 0.15813482007095794,
"grad_norm": 1.0972974069836425,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.7662,
"step": 39
},
{
"epoch": 0.16218955904713633,
"grad_norm": 0.7497035326618606,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.7316,
"step": 40
},
{
"epoch": 0.16624429802331475,
"grad_norm": 0.7873091636221122,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.718,
"step": 41
},
{
"epoch": 0.17029903699949317,
"grad_norm": 0.9796648682801602,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.7328,
"step": 42
},
{
"epoch": 0.17435377597567156,
"grad_norm": 0.6688871725941729,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.7769,
"step": 43
},
{
"epoch": 0.17840851495184998,
"grad_norm": 0.7403041425526135,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.7145,
"step": 44
},
{
"epoch": 0.1824632539280284,
"grad_norm": 0.7090195830408474,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.7457,
"step": 45
},
{
"epoch": 0.1865179929042068,
"grad_norm": 0.6589003304523197,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.7396,
"step": 46
},
{
"epoch": 0.1905727318803852,
"grad_norm": 0.7300942368681973,
"learning_rate": 1.528455284552846e-05,
"loss": 0.7171,
"step": 47
},
{
"epoch": 0.19462747085656362,
"grad_norm": 0.7424528805957988,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.7282,
"step": 48
},
{
"epoch": 0.198682209832742,
"grad_norm": 0.6969522479998541,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.7232,
"step": 49
},
{
"epoch": 0.20273694880892043,
"grad_norm": 0.7384738914554233,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.7286,
"step": 50
},
{
"epoch": 0.20679168778509882,
"grad_norm": 0.7142741837045874,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.7211,
"step": 51
},
{
"epoch": 0.21084642676127724,
"grad_norm": 0.7702480778247397,
"learning_rate": 1.691056910569106e-05,
"loss": 0.7412,
"step": 52
},
{
"epoch": 0.21490116573745566,
"grad_norm": 0.7672053437668961,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.7171,
"step": 53
},
{
"epoch": 0.21895590471363405,
"grad_norm": 0.6837919266541023,
"learning_rate": 1.75609756097561e-05,
"loss": 0.7211,
"step": 54
},
{
"epoch": 0.22301064368981247,
"grad_norm": 0.7145738436200555,
"learning_rate": 1.788617886178862e-05,
"loss": 0.7116,
"step": 55
},
{
"epoch": 0.22706538266599088,
"grad_norm": 0.6832268719838818,
"learning_rate": 1.821138211382114e-05,
"loss": 0.6856,
"step": 56
},
{
"epoch": 0.23112012164216927,
"grad_norm": 0.5845620879910963,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.7133,
"step": 57
},
{
"epoch": 0.2351748606183477,
"grad_norm": 0.6355335848331151,
"learning_rate": 1.886178861788618e-05,
"loss": 0.7093,
"step": 58
},
{
"epoch": 0.2392295995945261,
"grad_norm": 0.6948749141582691,
"learning_rate": 1.91869918699187e-05,
"loss": 0.6769,
"step": 59
},
{
"epoch": 0.2432843385707045,
"grad_norm": 0.6458388030223168,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.7094,
"step": 60
},
{
"epoch": 0.24733907754688292,
"grad_norm": 0.7694557577089198,
"learning_rate": 1.983739837398374e-05,
"loss": 0.702,
"step": 61
},
{
"epoch": 0.2513938165230613,
"grad_norm": 0.7458684625864561,
"learning_rate": 2.016260162601626e-05,
"loss": 0.689,
"step": 62
},
{
"epoch": 0.2554485554992397,
"grad_norm": 0.779798108616507,
"learning_rate": 2.048780487804878e-05,
"loss": 0.7211,
"step": 63
},
{
"epoch": 0.25950329447541814,
"grad_norm": 0.7377844562285283,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.7285,
"step": 64
},
{
"epoch": 0.26355803345159656,
"grad_norm": 0.6742920634485373,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.6741,
"step": 65
},
{
"epoch": 0.267612772427775,
"grad_norm": 0.6913792974161947,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.7003,
"step": 66
},
{
"epoch": 0.27166751140395334,
"grad_norm": 0.7630696019471577,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.6997,
"step": 67
},
{
"epoch": 0.27572225038013176,
"grad_norm": 0.7890619061259875,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.6841,
"step": 68
},
{
"epoch": 0.2797769893563102,
"grad_norm": 0.763429565427008,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.6909,
"step": 69
},
{
"epoch": 0.2838317283324886,
"grad_norm": 0.7286382276397224,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.7088,
"step": 70
},
{
"epoch": 0.287886467308667,
"grad_norm": 0.7048885671940769,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.6983,
"step": 71
},
{
"epoch": 0.29194120628484543,
"grad_norm": 0.9644537757352696,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.6856,
"step": 72
},
{
"epoch": 0.2959959452610238,
"grad_norm": 0.7278462733897373,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.7066,
"step": 73
},
{
"epoch": 0.3000506842372022,
"grad_norm": 0.8379959872080561,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.7168,
"step": 74
},
{
"epoch": 0.30410542321338063,
"grad_norm": 1.0490128911944459,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.726,
"step": 75
},
{
"epoch": 0.30816016218955905,
"grad_norm": 0.6734185042288604,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.6918,
"step": 76
},
{
"epoch": 0.31221490116573747,
"grad_norm": 1.1451572755516948,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.709,
"step": 77
},
{
"epoch": 0.3162696401419159,
"grad_norm": 0.8185756716696622,
"learning_rate": 2.536585365853659e-05,
"loss": 0.6992,
"step": 78
},
{
"epoch": 0.32032437911809425,
"grad_norm": 1.0328469721562996,
"learning_rate": 2.569105691056911e-05,
"loss": 0.7066,
"step": 79
},
{
"epoch": 0.32437911809427267,
"grad_norm": 1.3444900383346927,
"learning_rate": 2.601626016260163e-05,
"loss": 0.7102,
"step": 80
},
{
"epoch": 0.3284338570704511,
"grad_norm": 0.7389552280306783,
"learning_rate": 2.634146341463415e-05,
"loss": 0.7014,
"step": 81
},
{
"epoch": 0.3324885960466295,
"grad_norm": 1.0803164134362695,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.7047,
"step": 82
},
{
"epoch": 0.3365433350228079,
"grad_norm": 0.7485845858411434,
"learning_rate": 2.699186991869919e-05,
"loss": 0.6837,
"step": 83
},
{
"epoch": 0.34059807399898634,
"grad_norm": 0.9760047392445128,
"learning_rate": 2.731707317073171e-05,
"loss": 0.679,
"step": 84
},
{
"epoch": 0.3446528129751647,
"grad_norm": 0.7632569083394044,
"learning_rate": 2.764227642276423e-05,
"loss": 0.698,
"step": 85
},
{
"epoch": 0.3487075519513431,
"grad_norm": 0.8293891276518387,
"learning_rate": 2.796747967479675e-05,
"loss": 0.69,
"step": 86
},
{
"epoch": 0.35276229092752154,
"grad_norm": 0.9739090243740767,
"learning_rate": 2.829268292682927e-05,
"loss": 0.7075,
"step": 87
},
{
"epoch": 0.35681702990369996,
"grad_norm": 0.6430347870691953,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.6846,
"step": 88
},
{
"epoch": 0.3608717688798784,
"grad_norm": 0.9133982026270139,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.6879,
"step": 89
},
{
"epoch": 0.3649265078560568,
"grad_norm": 0.7595798983853993,
"learning_rate": 2.926829268292683e-05,
"loss": 0.6873,
"step": 90
},
{
"epoch": 0.36898124683223515,
"grad_norm": 0.8257891758311616,
"learning_rate": 2.959349593495935e-05,
"loss": 0.6923,
"step": 91
},
{
"epoch": 0.3730359858084136,
"grad_norm": 0.9024576487657057,
"learning_rate": 2.991869918699187e-05,
"loss": 0.692,
"step": 92
},
{
"epoch": 0.377090724784592,
"grad_norm": 0.6466942341342199,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.6887,
"step": 93
},
{
"epoch": 0.3811454637607704,
"grad_norm": 0.9232596534235612,
"learning_rate": 3.056910569105692e-05,
"loss": 0.6741,
"step": 94
},
{
"epoch": 0.3852002027369488,
"grad_norm": 0.8080656528815979,
"learning_rate": 3.089430894308943e-05,
"loss": 0.691,
"step": 95
},
{
"epoch": 0.38925494171312724,
"grad_norm": 1.0058008730398007,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.6852,
"step": 96
},
{
"epoch": 0.3933096806893056,
"grad_norm": 1.0280233982249636,
"learning_rate": 3.154471544715447e-05,
"loss": 0.6911,
"step": 97
},
{
"epoch": 0.397364419665484,
"grad_norm": 0.7517810818794654,
"learning_rate": 3.186991869918699e-05,
"loss": 0.6863,
"step": 98
},
{
"epoch": 0.40141915864166244,
"grad_norm": 1.0758098938437253,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.683,
"step": 99
},
{
"epoch": 0.40547389761784086,
"grad_norm": 0.7342481230908802,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.711,
"step": 100
},
{
"epoch": 0.4095286365940193,
"grad_norm": 0.9443883063108075,
"learning_rate": 3.284552845528456e-05,
"loss": 0.6931,
"step": 101
},
{
"epoch": 0.41358337557019764,
"grad_norm": 0.7967871031804306,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.6787,
"step": 102
},
{
"epoch": 0.41763811454637606,
"grad_norm": 0.8211351724614182,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.699,
"step": 103
},
{
"epoch": 0.4216928535225545,
"grad_norm": 1.0724463058993667,
"learning_rate": 3.382113821138212e-05,
"loss": 0.6809,
"step": 104
},
{
"epoch": 0.4257475924987329,
"grad_norm": 1.1780972529280092,
"learning_rate": 3.414634146341463e-05,
"loss": 0.6875,
"step": 105
},
{
"epoch": 0.4298023314749113,
"grad_norm": 0.7402523042919582,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.6664,
"step": 106
},
{
"epoch": 0.43385707045108973,
"grad_norm": 1.3308853636327231,
"learning_rate": 3.479674796747968e-05,
"loss": 0.7151,
"step": 107
},
{
"epoch": 0.4379118094272681,
"grad_norm": 0.8188608634411022,
"learning_rate": 3.51219512195122e-05,
"loss": 0.6912,
"step": 108
},
{
"epoch": 0.4419665484034465,
"grad_norm": 1.315123879380671,
"learning_rate": 3.544715447154472e-05,
"loss": 0.6941,
"step": 109
},
{
"epoch": 0.44602128737962493,
"grad_norm": 0.9525459481311808,
"learning_rate": 3.577235772357724e-05,
"loss": 0.678,
"step": 110
},
{
"epoch": 0.45007602635580335,
"grad_norm": 2.3901357503401868,
"learning_rate": 3.609756097560976e-05,
"loss": 0.7146,
"step": 111
},
{
"epoch": 0.45413076533198177,
"grad_norm": 1.4990202315034267,
"learning_rate": 3.642276422764228e-05,
"loss": 0.6937,
"step": 112
},
{
"epoch": 0.4581855043081602,
"grad_norm": 0.8426609396121021,
"learning_rate": 3.67479674796748e-05,
"loss": 0.6987,
"step": 113
},
{
"epoch": 0.46224024328433855,
"grad_norm": 1.6349726015218022,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.6866,
"step": 114
},
{
"epoch": 0.46629498226051697,
"grad_norm": 1.196905079918351,
"learning_rate": 3.739837398373984e-05,
"loss": 0.6705,
"step": 115
},
{
"epoch": 0.4703497212366954,
"grad_norm": 0.9259170416416036,
"learning_rate": 3.772357723577236e-05,
"loss": 0.7079,
"step": 116
},
{
"epoch": 0.4744044602128738,
"grad_norm": 1.0180263895462414,
"learning_rate": 3.804878048780488e-05,
"loss": 0.7203,
"step": 117
},
{
"epoch": 0.4784591991890522,
"grad_norm": 1.0078834521339362,
"learning_rate": 3.83739837398374e-05,
"loss": 0.6731,
"step": 118
},
{
"epoch": 0.48251393816523064,
"grad_norm": 1.5309138577841397,
"learning_rate": 3.869918699186992e-05,
"loss": 0.6875,
"step": 119
},
{
"epoch": 0.486568677141409,
"grad_norm": 0.821718488295634,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.6718,
"step": 120
},
{
"epoch": 0.4906234161175874,
"grad_norm": 1.7320738012628334,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.6967,
"step": 121
},
{
"epoch": 0.49467815509376584,
"grad_norm": 1.28806423710955,
"learning_rate": 3.967479674796748e-05,
"loss": 0.6951,
"step": 122
},
{
"epoch": 0.49873289406994425,
"grad_norm": 1.9807858710872799,
"learning_rate": 4e-05,
"loss": 0.743,
"step": 123
},
{
"epoch": 0.5027876330461226,
"grad_norm": 1.835357066309503,
"learning_rate": 3.999991946137476e-05,
"loss": 0.6937,
"step": 124
},
{
"epoch": 0.5068423720223011,
"grad_norm": 1.2070749708501356,
"learning_rate": 3.999967784614766e-05,
"loss": 0.7031,
"step": 125
},
{
"epoch": 0.5108971109984795,
"grad_norm": 4.614800436607694,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.7107,
"step": 126
},
{
"epoch": 0.5149518499746579,
"grad_norm": 1.6327098536948497,
"learning_rate": 3.999871139496895e-05,
"loss": 0.694,
"step": 127
},
{
"epoch": 0.5190065889508363,
"grad_norm": 1.3013993777388393,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.6714,
"step": 128
},
{
"epoch": 0.5230613279270147,
"grad_norm": 1.9289443360186063,
"learning_rate": 3.999710067759846e-05,
"loss": 0.6753,
"step": 129
},
{
"epoch": 0.5271160669031931,
"grad_norm": 1.7090956719629866,
"learning_rate": 3.999605373449617e-05,
"loss": 0.7147,
"step": 130
},
{
"epoch": 0.5311708058793715,
"grad_norm": 1.0722214564616033,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.667,
"step": 131
},
{
"epoch": 0.53522554485555,
"grad_norm": 1.1286918966506552,
"learning_rate": 3.999347672161713e-05,
"loss": 0.7076,
"step": 132
},
{
"epoch": 0.5392802838317283,
"grad_norm": 1.8596403138033017,
"learning_rate": 3.999194667259528e-05,
"loss": 0.7029,
"step": 133
},
{
"epoch": 0.5433350228079067,
"grad_norm": 0.9980098596120747,
"learning_rate": 3.999025561118334e-05,
"loss": 0.7003,
"step": 134
},
{
"epoch": 0.5473897617840852,
"grad_norm": 1.0646233674967533,
"learning_rate": 3.998840355100086e-05,
"loss": 0.6803,
"step": 135
},
{
"epoch": 0.5514445007602635,
"grad_norm": 1.1747189909613964,
"learning_rate": 3.998639050696409e-05,
"loss": 0.6981,
"step": 136
},
{
"epoch": 0.555499239736442,
"grad_norm": 0.8890022070734527,
"learning_rate": 3.998421649528582e-05,
"loss": 0.6801,
"step": 137
},
{
"epoch": 0.5595539787126204,
"grad_norm": 0.9237520878918816,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.6928,
"step": 138
},
{
"epoch": 0.5636087176887988,
"grad_norm": 0.689021448921926,
"learning_rate": 3.997938564033779e-05,
"loss": 0.671,
"step": 139
},
{
"epoch": 0.5676634566649772,
"grad_norm": 0.9582186141418718,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.6598,
"step": 140
},
{
"epoch": 0.5717181956411556,
"grad_norm": 0.7340409097480117,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.6803,
"step": 141
},
{
"epoch": 0.575772934617334,
"grad_norm": 0.705798465159733,
"learning_rate": 3.997093258045973e-05,
"loss": 0.6933,
"step": 142
},
{
"epoch": 0.5798276735935124,
"grad_norm": 0.7765603061788923,
"learning_rate": 3.996779317598936e-05,
"loss": 0.6783,
"step": 143
},
{
"epoch": 0.5838824125696909,
"grad_norm": 0.6238745530139232,
"learning_rate": 3.996449295365782e-05,
"loss": 0.6601,
"step": 144
},
{
"epoch": 0.5879371515458692,
"grad_norm": 0.9929072620924326,
"learning_rate": 3.996103194004467e-05,
"loss": 0.6777,
"step": 145
},
{
"epoch": 0.5919918905220476,
"grad_norm": 0.7143742837978996,
"learning_rate": 3.995741016302441e-05,
"loss": 0.6761,
"step": 146
},
{
"epoch": 0.5960466294982261,
"grad_norm": 0.8774888500952484,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.6701,
"step": 147
},
{
"epoch": 0.6001013684744044,
"grad_norm": 0.8343025750986708,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.706,
"step": 148
},
{
"epoch": 0.6041561074505829,
"grad_norm": 0.8102723088966813,
"learning_rate": 3.994558054968643e-05,
"loss": 0.6822,
"step": 149
},
{
"epoch": 0.6082108464267613,
"grad_norm": 0.9568927730397531,
"learning_rate": 3.994131602367481e-05,
"loss": 0.7287,
"step": 150
},
{
"epoch": 0.6122655854029397,
"grad_norm": 0.6015408850621344,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.6822,
"step": 151
},
{
"epoch": 0.6163203243791181,
"grad_norm": 1.2001824966184542,
"learning_rate": 3.993230519343752e-05,
"loss": 0.6833,
"step": 152
},
{
"epoch": 0.6203750633552965,
"grad_norm": 0.7698179885632005,
"learning_rate": 3.992755896178383e-05,
"loss": 0.6715,
"step": 153
},
{
"epoch": 0.6244298023314749,
"grad_norm": 1.0977983341598858,
"learning_rate": 3.992265223630981e-05,
"loss": 0.6841,
"step": 154
},
{
"epoch": 0.6284845413076533,
"grad_norm": 0.8188334478534633,
"learning_rate": 3.991758505653355e-05,
"loss": 0.6792,
"step": 155
},
{
"epoch": 0.6325392802838318,
"grad_norm": 0.7417262938692606,
"learning_rate": 3.991235746326543e-05,
"loss": 0.6498,
"step": 156
},
{
"epoch": 0.6365940192600101,
"grad_norm": 0.7619102560385524,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.6456,
"step": 157
},
{
"epoch": 0.6406487582361885,
"grad_norm": 0.6784465965155523,
"learning_rate": 3.990142120595444e-05,
"loss": 0.6629,
"step": 158
},
{
"epoch": 0.644703497212367,
"grad_norm": 0.722121613112185,
"learning_rate": 3.98957126299907e-05,
"loss": 0.6637,
"step": 159
},
{
"epoch": 0.6487582361885453,
"grad_norm": 0.6447945770108525,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.6549,
"step": 160
},
{
"epoch": 0.6528129751647238,
"grad_norm": 0.7035653078450818,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.6813,
"step": 161
},
{
"epoch": 0.6568677141409022,
"grad_norm": 0.6624902074090185,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.6602,
"step": 162
},
{
"epoch": 0.6609224531170805,
"grad_norm": 0.575113864148145,
"learning_rate": 3.98712764319087e-05,
"loss": 0.6692,
"step": 163
},
{
"epoch": 0.664977192093259,
"grad_norm": 0.6722069364099436,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.668,
"step": 164
},
{
"epoch": 0.6690319310694374,
"grad_norm": 0.6151763300858886,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.6727,
"step": 165
},
{
"epoch": 0.6730866700456158,
"grad_norm": 0.5765462457314585,
"learning_rate": 3.985126869010129e-05,
"loss": 0.6843,
"step": 166
},
{
"epoch": 0.6771414090217942,
"grad_norm": 0.8147294856093195,
"learning_rate": 3.984427961114833e-05,
"loss": 0.6813,
"step": 167
},
{
"epoch": 0.6811961479979727,
"grad_norm": 0.6653934331750595,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.6906,
"step": 168
},
{
"epoch": 0.685250886974151,
"grad_norm": 0.6367427096971082,
"learning_rate": 3.982982204151901e-05,
"loss": 0.6988,
"step": 169
},
{
"epoch": 0.6893056259503294,
"grad_norm": 0.599698466328495,
"learning_rate": 3.982235366728193e-05,
"loss": 0.6692,
"step": 170
},
{
"epoch": 0.6933603649265079,
"grad_norm": 0.6849336630476502,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.6536,
"step": 171
},
{
"epoch": 0.6974151039026862,
"grad_norm": 0.5733465390017347,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.6961,
"step": 172
},
{
"epoch": 0.7014698428788647,
"grad_norm": 0.6818706973421986,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.6856,
"step": 173
},
{
"epoch": 0.7055245818550431,
"grad_norm": 0.6132588090722377,
"learning_rate": 3.979088432599627e-05,
"loss": 0.6828,
"step": 174
},
{
"epoch": 0.7095793208312214,
"grad_norm": 0.5894185438945142,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.6637,
"step": 175
},
{
"epoch": 0.7136340598073999,
"grad_norm": 0.6342968095007281,
"learning_rate": 3.977419304032111e-05,
"loss": 0.6862,
"step": 176
},
{
"epoch": 0.7176887987835783,
"grad_norm": 0.6172904989559048,
"learning_rate": 3.976560847560697e-05,
"loss": 0.6754,
"step": 177
},
{
"epoch": 0.7217435377597567,
"grad_norm": 0.6513584061429492,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.658,
"step": 178
},
{
"epoch": 0.7257982767359351,
"grad_norm": 0.564461934281462,
"learning_rate": 3.974796184811956e-05,
"loss": 0.6626,
"step": 179
},
{
"epoch": 0.7298530157121136,
"grad_norm": 0.6090058825632219,
"learning_rate": 3.973889992746979e-05,
"loss": 0.6757,
"step": 180
},
{
"epoch": 0.733907754688292,
"grad_norm": 0.5585227586359442,
"learning_rate": 3.972967903243361e-05,
"loss": 0.6717,
"step": 181
},
{
"epoch": 0.7379624936644703,
"grad_norm": 0.4707154002171439,
"learning_rate": 3.972029923727486e-05,
"loss": 0.6582,
"step": 182
},
{
"epoch": 0.7420172326406488,
"grad_norm": 0.544674928057688,
"learning_rate": 3.971076061753709e-05,
"loss": 0.6474,
"step": 183
},
{
"epoch": 0.7460719716168271,
"grad_norm": 0.5365234891982672,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.6707,
"step": 184
},
{
"epoch": 0.7501267105930056,
"grad_norm": 0.4955924399686322,
"learning_rate": 3.969120721289402e-05,
"loss": 0.6644,
"step": 185
},
{
"epoch": 0.754181449569184,
"grad_norm": 0.5444771343256386,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.6626,
"step": 186
},
{
"epoch": 0.7582361885453623,
"grad_norm": 0.5711354858382636,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.6952,
"step": 187
},
{
"epoch": 0.7622909275215408,
"grad_norm": 0.49033415499089034,
"learning_rate": 3.966068788369422e-05,
"loss": 0.6643,
"step": 188
},
{
"epoch": 0.7663456664977192,
"grad_norm": 0.7977718416858243,
"learning_rate": 3.965019797448622e-05,
"loss": 0.6915,
"step": 189
},
{
"epoch": 0.7704004054738977,
"grad_norm": 0.508539756624699,
"learning_rate": 3.963954980528515e-05,
"loss": 0.6494,
"step": 190
},
{
"epoch": 0.774455144450076,
"grad_norm": 0.5835751436218239,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.6596,
"step": 191
},
{
"epoch": 0.7785098834262545,
"grad_norm": 0.49457305443770033,
"learning_rate": 3.961777903121329e-05,
"loss": 0.6445,
"step": 192
},
{
"epoch": 0.7825646224024329,
"grad_norm": 0.5395641244777112,
"learning_rate": 3.960665660168131e-05,
"loss": 0.6867,
"step": 193
},
{
"epoch": 0.7866193613786112,
"grad_norm": 0.5514382799732869,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.6863,
"step": 194
},
{
"epoch": 0.7906741003547897,
"grad_norm": 0.5663218656289281,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.6692,
"step": 195
},
{
"epoch": 0.794728839330968,
"grad_norm": 0.5720308948620724,
"learning_rate": 3.957234222185657e-05,
"loss": 0.667,
"step": 196
},
{
"epoch": 0.7987835783071465,
"grad_norm": 0.6308084088213408,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.6626,
"step": 197
},
{
"epoch": 0.8028383172833249,
"grad_norm": 0.7116301575492819,
"learning_rate": 3.954867765033605e-05,
"loss": 0.6598,
"step": 198
},
{
"epoch": 0.8068930562595032,
"grad_norm": 0.6410050012636682,
"learning_rate": 3.953660915306728e-05,
"loss": 0.6564,
"step": 199
},
{
"epoch": 0.8109477952356817,
"grad_norm": 0.603593734567084,
"learning_rate": 3.952438331063419e-05,
"loss": 0.6631,
"step": 200
},
{
"epoch": 0.8150025342118601,
"grad_norm": 0.6479894849496406,
"learning_rate": 3.951200022150205e-05,
"loss": 0.6794,
"step": 201
},
{
"epoch": 0.8190572731880386,
"grad_norm": 0.5743847940643024,
"learning_rate": 3.949945998540253e-05,
"loss": 0.6757,
"step": 202
},
{
"epoch": 0.8231120121642169,
"grad_norm": 0.5813175104068687,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.6915,
"step": 203
},
{
"epoch": 0.8271667511403953,
"grad_norm": 0.5622095934511784,
"learning_rate": 3.947390847755559e-05,
"loss": 0.6586,
"step": 204
},
{
"epoch": 0.8312214901165738,
"grad_norm": 0.5268665217538123,
"learning_rate": 3.946089741159648e-05,
"loss": 0.6542,
"step": 205
},
{
"epoch": 0.8352762290927521,
"grad_norm": 0.5409841799868024,
"learning_rate": 3.944772961024501e-05,
"loss": 0.6562,
"step": 206
},
{
"epoch": 0.8393309680689306,
"grad_norm": 0.635540461494687,
"learning_rate": 3.943440517955285e-05,
"loss": 0.669,
"step": 207
},
{
"epoch": 0.843385707045109,
"grad_norm": 0.6637248171877679,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.6677,
"step": 208
},
{
"epoch": 0.8474404460212874,
"grad_norm": 0.6656738896112339,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.662,
"step": 209
},
{
"epoch": 0.8514951849974658,
"grad_norm": 0.6871767253499165,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.6627,
"step": 210
},
{
"epoch": 0.8555499239736442,
"grad_norm": 0.7622749720064445,
"learning_rate": 3.937954332854371e-05,
"loss": 0.6558,
"step": 211
},
{
"epoch": 0.8596046629498226,
"grad_norm": 0.6414294057685311,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.6754,
"step": 212
},
{
"epoch": 0.863659401926001,
"grad_norm": 0.7953161065568234,
"learning_rate": 3.935117547697387e-05,
"loss": 0.6677,
"step": 213
},
{
"epoch": 0.8677141409021795,
"grad_norm": 0.594375165813602,
"learning_rate": 3.933675771619675e-05,
"loss": 0.6727,
"step": 214
},
{
"epoch": 0.8717688798783578,
"grad_norm": 0.7344808009390653,
"learning_rate": 3.932218421983131e-05,
"loss": 0.681,
"step": 215
},
{
"epoch": 0.8758236188545362,
"grad_norm": 0.7167765957206945,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.6759,
"step": 216
},
{
"epoch": 0.8798783578307147,
"grad_norm": 0.603462550356937,
"learning_rate": 3.929257049108054e-05,
"loss": 0.6387,
"step": 217
},
{
"epoch": 0.883933096806893,
"grad_norm": 0.571884635497545,
"learning_rate": 3.927753049720011e-05,
"loss": 0.6519,
"step": 218
},
{
"epoch": 0.8879878357830715,
"grad_norm": 0.6559833796555712,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.6928,
"step": 219
},
{
"epoch": 0.8920425747592499,
"grad_norm": 0.53928088476035,
"learning_rate": 3.92469848560784e-05,
"loss": 0.6468,
"step": 220
},
{
"epoch": 0.8960973137354283,
"grad_norm": 0.6372413119697461,
"learning_rate": 3.923147945484751e-05,
"loss": 0.6565,
"step": 221
},
{
"epoch": 0.9001520527116067,
"grad_norm": 0.5878945612482709,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.6627,
"step": 222
},
{
"epoch": 0.9042067916877851,
"grad_norm": 0.7502268667930518,
"learning_rate": 3.920000411543654e-05,
"loss": 0.6728,
"step": 223
},
{
"epoch": 0.9082615306639635,
"grad_norm": 0.5258230439693216,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.6403,
"step": 224
},
{
"epoch": 0.9123162696401419,
"grad_norm": 0.6898307257490985,
"learning_rate": 3.916791024049648e-05,
"loss": 0.6902,
"step": 225
},
{
"epoch": 0.9163710086163204,
"grad_norm": 0.550891947167564,
"learning_rate": 3.91516316745245e-05,
"loss": 0.6747,
"step": 226
},
{
"epoch": 0.9204257475924987,
"grad_norm": 0.6560712993548744,
"learning_rate": 3.913519886394389e-05,
"loss": 0.6422,
"step": 227
},
{
"epoch": 0.9244804865686771,
"grad_norm": 0.5600888821358391,
"learning_rate": 3.911861194110225e-05,
"loss": 0.6558,
"step": 228
},
{
"epoch": 0.9285352255448556,
"grad_norm": 0.6302155881324434,
"learning_rate": 3.910187103958837e-05,
"loss": 0.7023,
"step": 229
},
{
"epoch": 0.9325899645210339,
"grad_norm": 0.4869967136247146,
"learning_rate": 3.908497629423117e-05,
"loss": 0.6598,
"step": 230
},
{
"epoch": 0.9366447034972124,
"grad_norm": 0.6569008883330597,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.6672,
"step": 231
},
{
"epoch": 0.9406994424733908,
"grad_norm": 0.5552260012456819,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.6554,
"step": 232
},
{
"epoch": 0.9447541814495691,
"grad_norm": 0.5841537947243322,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.6797,
"step": 233
},
{
"epoch": 0.9488089204257476,
"grad_norm": 0.5805402020053461,
"learning_rate": 3.901586161429081e-05,
"loss": 0.6525,
"step": 234
},
{
"epoch": 0.952863659401926,
"grad_norm": 0.5175954006831186,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.6587,
"step": 235
},
{
"epoch": 0.9569183983781044,
"grad_norm": 0.5235935548422253,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.6731,
"step": 236
},
{
"epoch": 0.9609731373542828,
"grad_norm": 0.5240565344545104,
"learning_rate": 3.896241703466667e-05,
"loss": 0.668,
"step": 237
},
{
"epoch": 0.9650278763304613,
"grad_norm": 0.5566347217339349,
"learning_rate": 3.894429654085585e-05,
"loss": 0.6512,
"step": 238
},
{
"epoch": 0.9690826153066396,
"grad_norm": 0.5467691540646846,
"learning_rate": 3.892602347228505e-05,
"loss": 0.6656,
"step": 239
},
{
"epoch": 0.973137354282818,
"grad_norm": 0.5643909729738225,
"learning_rate": 3.890759797612307e-05,
"loss": 0.6786,
"step": 240
},
{
"epoch": 0.9771920932589965,
"grad_norm": 0.5488436036709752,
"learning_rate": 3.888902020076632e-05,
"loss": 0.6521,
"step": 241
},
{
"epoch": 0.9812468322351748,
"grad_norm": 0.5759615663503065,
"learning_rate": 3.887029029583764e-05,
"loss": 0.6472,
"step": 242
},
{
"epoch": 0.9853015712113533,
"grad_norm": 0.5164525825813331,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.6488,
"step": 243
},
{
"epoch": 0.9893563101875317,
"grad_norm": 0.6821928673060211,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.6657,
"step": 244
},
{
"epoch": 0.99341104916371,
"grad_norm": 0.4928621635756573,
"learning_rate": 3.881318931821972e-05,
"loss": 0.648,
"step": 245
},
{
"epoch": 0.9974657881398885,
"grad_norm": 0.6463592667023097,
"learning_rate": 3.879385241571817e-05,
"loss": 0.6684,
"step": 246
},
{
"epoch": 1.0030410542321337,
"grad_norm": 0.5878757829771379,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.5734,
"step": 247
},
{
"epoch": 1.0070957932083122,
"grad_norm": 0.5758660605379867,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.5413,
"step": 248
},
{
"epoch": 1.0111505321844907,
"grad_norm": 0.5687600864259809,
"learning_rate": 3.873493415863256e-05,
"loss": 0.5414,
"step": 249
},
{
"epoch": 1.015205271160669,
"grad_norm": 0.5857842299726084,
"learning_rate": 3.871499275032111e-05,
"loss": 0.5532,
"step": 250
},
{
"epoch": 1.0192600101368474,
"grad_norm": 0.6289664075593226,
"learning_rate": 3.869490061403091e-05,
"loss": 0.5583,
"step": 251
},
{
"epoch": 1.023314749113026,
"grad_norm": 0.6235876622520403,
"learning_rate": 3.867465791158124e-05,
"loss": 0.5648,
"step": 252
},
{
"epoch": 1.0273694880892044,
"grad_norm": 0.5381281528998375,
"learning_rate": 3.865426480600407e-05,
"loss": 0.5428,
"step": 253
},
{
"epoch": 1.0314242270653826,
"grad_norm": 0.5254876407102241,
"learning_rate": 3.863372146154264e-05,
"loss": 0.5346,
"step": 254
},
{
"epoch": 1.035478966041561,
"grad_norm": 0.5927676727148181,
"learning_rate": 3.861302804365024e-05,
"loss": 0.5399,
"step": 255
},
{
"epoch": 1.0395337050177396,
"grad_norm": 0.49794763930707314,
"learning_rate": 3.85921847189888e-05,
"loss": 0.5386,
"step": 256
},
{
"epoch": 1.0435884439939178,
"grad_norm": 0.5549831303164778,
"learning_rate": 3.85711916554276e-05,
"loss": 0.5547,
"step": 257
},
{
"epoch": 1.0476431829700963,
"grad_norm": 0.525078912106115,
"learning_rate": 3.85500490220419e-05,
"loss": 0.5363,
"step": 258
},
{
"epoch": 1.0516979219462748,
"grad_norm": 0.5418998537277435,
"learning_rate": 3.852875698911154e-05,
"loss": 0.4941,
"step": 259
},
{
"epoch": 1.055752660922453,
"grad_norm": 0.5154304920568932,
"learning_rate": 3.850731572811963e-05,
"loss": 0.4957,
"step": 260
},
{
"epoch": 1.0598073998986315,
"grad_norm": 0.5810323798109459,
"learning_rate": 3.848572541175116e-05,
"loss": 0.5211,
"step": 261
},
{
"epoch": 1.06386213887481,
"grad_norm": 0.5744998068855989,
"learning_rate": 3.846398621389154e-05,
"loss": 0.5099,
"step": 262
},
{
"epoch": 1.0679168778509884,
"grad_norm": 0.648411930307603,
"learning_rate": 3.84420983096253e-05,
"loss": 0.558,
"step": 263
},
{
"epoch": 1.0719716168271667,
"grad_norm": 0.6270281388129116,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.5211,
"step": 264
},
{
"epoch": 1.0760263558033452,
"grad_norm": 0.5064640571790376,
"learning_rate": 3.839787708819787e-05,
"loss": 0.5245,
"step": 265
},
{
"epoch": 1.0800810947795236,
"grad_norm": 0.6718727630974323,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.5385,
"step": 266
},
{
"epoch": 1.0841358337557019,
"grad_norm": 0.5479619482607734,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.5481,
"step": 267
},
{
"epoch": 1.0881905727318804,
"grad_norm": 0.5824318189656096,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.541,
"step": 268
},
{
"epoch": 1.0922453117080588,
"grad_norm": 0.594580577223674,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.5386,
"step": 269
},
{
"epoch": 1.0963000506842373,
"grad_norm": 0.5971130485367072,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.5503,
"step": 270
},
{
"epoch": 1.1003547896604156,
"grad_norm": 0.7838587020785184,
"learning_rate": 3.826166304956594e-05,
"loss": 0.5187,
"step": 271
},
{
"epoch": 1.104409528636594,
"grad_norm": 0.5669595496349817,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.5361,
"step": 272
},
{
"epoch": 1.1084642676127725,
"grad_norm": 0.7102136463147705,
"learning_rate": 3.821507978767479e-05,
"loss": 0.5479,
"step": 273
},
{
"epoch": 1.1125190065889508,
"grad_norm": 0.5871403310820278,
"learning_rate": 3.819156801001693e-05,
"loss": 0.5419,
"step": 274
},
{
"epoch": 1.1165737455651292,
"grad_norm": 0.7350609226387498,
"learning_rate": 3.816790971997121e-05,
"loss": 0.5435,
"step": 275
},
{
"epoch": 1.1206284845413077,
"grad_norm": 0.5332760748470078,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.5295,
"step": 276
},
{
"epoch": 1.124683223517486,
"grad_norm": 0.6980458593997768,
"learning_rate": 3.81201543660571e-05,
"loss": 0.5415,
"step": 277
},
{
"epoch": 1.1287379624936644,
"grad_norm": 0.5822645044689777,
"learning_rate": 3.809605768680377e-05,
"loss": 0.5188,
"step": 278
},
{
"epoch": 1.132792701469843,
"grad_norm": 0.6945976103783668,
"learning_rate": 3.807181526438958e-05,
"loss": 0.5256,
"step": 279
},
{
"epoch": 1.1368474404460214,
"grad_norm": 0.5708085225855689,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.5484,
"step": 280
},
{
"epoch": 1.1409021794221996,
"grad_norm": 0.705893306854577,
"learning_rate": 3.802289397223145e-05,
"loss": 0.5115,
"step": 281
},
{
"epoch": 1.144956918398378,
"grad_norm": 0.6254079685042802,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.5297,
"step": 282
},
{
"epoch": 1.1490116573745566,
"grad_norm": 0.5534058307056284,
"learning_rate": 3.797339206560096e-05,
"loss": 0.507,
"step": 283
},
{
"epoch": 1.1530663963507348,
"grad_norm": 0.710682756301194,
"learning_rate": 3.794842387948027e-05,
"loss": 0.5236,
"step": 284
},
{
"epoch": 1.1571211353269133,
"grad_norm": 0.5741554664737155,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.5348,
"step": 285
},
{
"epoch": 1.1611758743030918,
"grad_norm": 0.5615514090579279,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.514,
"step": 286
},
{
"epoch": 1.1652306132792702,
"grad_norm": 0.599102775121823,
"learning_rate": 3.787265280646825e-05,
"loss": 0.5119,
"step": 287
},
{
"epoch": 1.1692853522554485,
"grad_norm": 0.5160057202823605,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.563,
"step": 288
},
{
"epoch": 1.173340091231627,
"grad_norm": 0.5128746063076736,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.5173,
"step": 289
},
{
"epoch": 1.1773948302078054,
"grad_norm": 0.5469712587296043,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.5493,
"step": 290
},
{
"epoch": 1.1814495691839837,
"grad_norm": 0.519459832649935,
"learning_rate": 3.776961046830306e-05,
"loss": 0.5217,
"step": 291
},
{
"epoch": 1.1855043081601622,
"grad_norm": 0.48836631195223607,
"learning_rate": 3.774349157719215e-05,
"loss": 0.5253,
"step": 292
},
{
"epoch": 1.1895590471363406,
"grad_norm": 0.5250048240890355,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.5176,
"step": 293
},
{
"epoch": 1.1936137861125191,
"grad_norm": 0.5205090925529059,
"learning_rate": 3.769082529555359e-05,
"loss": 0.5271,
"step": 294
},
{
"epoch": 1.1976685250886974,
"grad_norm": 0.6366825317299508,
"learning_rate": 3.766427832919294e-05,
"loss": 0.5549,
"step": 295
},
{
"epoch": 1.2017232640648758,
"grad_norm": 0.6167040022733965,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.5355,
"step": 296
},
{
"epoch": 1.2057780030410543,
"grad_norm": 0.6366891094836108,
"learning_rate": 3.761075781441526e-05,
"loss": 0.5227,
"step": 297
},
{
"epoch": 1.2098327420172326,
"grad_norm": 0.8359561651386813,
"learning_rate": 3.75837846970451e-05,
"loss": 0.5554,
"step": 298
},
{
"epoch": 1.213887480993411,
"grad_norm": 0.5294761511513448,
"learning_rate": 3.755666996229032e-05,
"loss": 0.5136,
"step": 299
},
{
"epoch": 1.2179422199695895,
"grad_norm": 0.6624181817991323,
"learning_rate": 3.752941382852927e-05,
"loss": 0.5645,
"step": 300
},
{
"epoch": 1.221996958945768,
"grad_norm": 0.5877995515862958,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.5354,
"step": 301
},
{
"epoch": 1.2260516979219462,
"grad_norm": 0.5875575199591586,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.5308,
"step": 302
},
{
"epoch": 1.2301064368981247,
"grad_norm": 0.5614185046630379,
"learning_rate": 3.744679923406351e-05,
"loss": 0.5529,
"step": 303
},
{
"epoch": 1.2341611758743032,
"grad_norm": 1.0664359102077074,
"learning_rate": 3.741897971081043e-05,
"loss": 0.5508,
"step": 304
},
{
"epoch": 1.2382159148504814,
"grad_norm": 0.6046973911120708,
"learning_rate": 3.739101989748946e-05,
"loss": 0.5397,
"step": 305
},
{
"epoch": 1.24227065382666,
"grad_norm": 1.2673124349483489,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.5705,
"step": 306
},
{
"epoch": 1.2463253928028384,
"grad_norm": 0.598233704017945,
"learning_rate": 3.73346803025098e-05,
"loss": 0.5146,
"step": 307
},
{
"epoch": 1.2503801317790169,
"grad_norm": 0.5651528299876605,
"learning_rate": 3.730630097460247e-05,
"loss": 0.5261,
"step": 308
},
{
"epoch": 1.254434870755195,
"grad_norm": 0.5878193006363205,
"learning_rate": 3.727778226412628e-05,
"loss": 0.5459,
"step": 309
},
{
"epoch": 1.2584896097313736,
"grad_norm": 0.6599534372703165,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.5515,
"step": 310
},
{
"epoch": 1.2625443487075518,
"grad_norm": 0.5382395668177711,
"learning_rate": 3.722032761533114e-05,
"loss": 0.5447,
"step": 311
},
{
"epoch": 1.2665990876837303,
"grad_norm": 0.5821648363520998,
"learning_rate": 3.719139213974403e-05,
"loss": 0.5666,
"step": 312
},
{
"epoch": 1.2706538266599088,
"grad_norm": 0.5225471012481449,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.5115,
"step": 313
},
{
"epoch": 1.2747085656360873,
"grad_norm": 0.5312525144558344,
"learning_rate": 3.713310605140055e-05,
"loss": 0.5358,
"step": 314
},
{
"epoch": 1.2787633046122655,
"grad_norm": 0.5694871694762195,
"learning_rate": 3.710375590807233e-05,
"loss": 0.5293,
"step": 315
},
{
"epoch": 1.282818043588444,
"grad_norm": 0.5824208302644457,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.5593,
"step": 316
},
{
"epoch": 1.2868727825646225,
"grad_norm": 0.5578603594395953,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.5195,
"step": 317
},
{
"epoch": 1.2909275215408007,
"grad_norm": 0.5206883755354508,
"learning_rate": 3.701487992136854e-05,
"loss": 0.5463,
"step": 318
},
{
"epoch": 1.2949822605169792,
"grad_norm": 0.6322592912350286,
"learning_rate": 3.69849802022222e-05,
"loss": 0.5381,
"step": 319
},
{
"epoch": 1.2990369994931577,
"grad_norm": 0.50976856555698,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.5337,
"step": 320
},
{
"epoch": 1.3030917384693361,
"grad_norm": 0.5689241990448529,
"learning_rate": 3.692477062175289e-05,
"loss": 0.5371,
"step": 321
},
{
"epoch": 1.3071464774455144,
"grad_norm": 0.5858422733430532,
"learning_rate": 3.689446124534958e-05,
"loss": 0.5185,
"step": 322
},
{
"epoch": 1.3112012164216929,
"grad_norm": 0.5150980661542439,
"learning_rate": 3.686401580327799e-05,
"loss": 0.5723,
"step": 323
},
{
"epoch": 1.3152559553978713,
"grad_norm": 0.6723271773916295,
"learning_rate": 3.683343454074149e-05,
"loss": 0.5165,
"step": 324
},
{
"epoch": 1.3193106943740496,
"grad_norm": 0.48189240381178344,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.541,
"step": 325
},
{
"epoch": 1.323365433350228,
"grad_norm": 0.5375564443419653,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.5275,
"step": 326
},
{
"epoch": 1.3274201723264065,
"grad_norm": 0.5243751094815126,
"learning_rate": 3.674087829877297e-05,
"loss": 0.5396,
"step": 327
},
{
"epoch": 1.331474911302585,
"grad_norm": 0.5009438956429274,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.5393,
"step": 328
},
{
"epoch": 1.3355296502787632,
"grad_norm": 0.526887141689783,
"learning_rate": 3.667849957966501e-05,
"loss": 0.5305,
"step": 329
},
{
"epoch": 1.3395843892549417,
"grad_norm": 0.5728732776288902,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.5186,
"step": 330
},
{
"epoch": 1.3436391282311202,
"grad_norm": 0.4878002753317562,
"learning_rate": 3.661558355626795e-05,
"loss": 0.5384,
"step": 331
},
{
"epoch": 1.3476938672072984,
"grad_norm": 0.5984357470604207,
"learning_rate": 3.658392468818163e-05,
"loss": 0.5389,
"step": 332
},
{
"epoch": 1.351748606183477,
"grad_norm": 0.45746468176179467,
"learning_rate": 3.655213225544574e-05,
"loss": 0.5452,
"step": 333
},
{
"epoch": 1.3558033451596554,
"grad_norm": 0.5288815754613176,
"learning_rate": 3.652020651411218e-05,
"loss": 0.5303,
"step": 334
},
{
"epoch": 1.3598580841358339,
"grad_norm": 0.5155283035957694,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.5631,
"step": 335
},
{
"epoch": 1.3639128231120121,
"grad_norm": 0.5056019808696458,
"learning_rate": 3.645595613522574e-05,
"loss": 0.5524,
"step": 336
},
{
"epoch": 1.3679675620881906,
"grad_norm": 0.5278228931185853,
"learning_rate": 3.642363201513657e-05,
"loss": 0.5084,
"step": 337
},
{
"epoch": 1.3720223010643688,
"grad_norm": 0.6461715292614261,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.5464,
"step": 338
},
{
"epoch": 1.3760770400405473,
"grad_norm": 0.5661481738074413,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.5185,
"step": 339
},
{
"epoch": 1.3801317790167258,
"grad_norm": 0.5942531936530784,
"learning_rate": 3.632586705948318e-05,
"loss": 0.5403,
"step": 340
},
{
"epoch": 1.3841865179929043,
"grad_norm": 0.5850682980282983,
"learning_rate": 3.629301541734311e-05,
"loss": 0.5458,
"step": 341
},
{
"epoch": 1.3882412569690827,
"grad_norm": 0.5333424760688134,
"learning_rate": 3.626003255349676e-05,
"loss": 0.5263,
"step": 342
},
{
"epoch": 1.392295995945261,
"grad_norm": 0.5429442872554063,
"learning_rate": 3.622691873358357e-05,
"loss": 0.5292,
"step": 343
},
{
"epoch": 1.3963507349214395,
"grad_norm": 0.47037294092694815,
"learning_rate": 3.61936742242977e-05,
"loss": 0.5229,
"step": 344
},
{
"epoch": 1.4004054738976177,
"grad_norm": 0.5942873647263273,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.5458,
"step": 345
},
{
"epoch": 1.4044602128737962,
"grad_norm": 0.5273951524953545,
"learning_rate": 3.612679420964516e-05,
"loss": 0.5393,
"step": 346
},
{
"epoch": 1.4085149518499747,
"grad_norm": 0.5195904407948269,
"learning_rate": 3.609315924292092e-05,
"loss": 0.5323,
"step": 347
},
{
"epoch": 1.4125696908261531,
"grad_norm": 0.4463577929614556,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.5342,
"step": 348
},
{
"epoch": 1.4166244298023316,
"grad_norm": 0.4761680337625318,
"learning_rate": 3.602550074513133e-05,
"loss": 0.5424,
"step": 349
},
{
"epoch": 1.4206791687785099,
"grad_norm": 0.5127915314976486,
"learning_rate": 3.599147775897822e-05,
"loss": 0.5422,
"step": 350
},
{
"epoch": 1.4247339077546883,
"grad_norm": 0.4814083880035532,
"learning_rate": 3.595732597966167e-05,
"loss": 0.5411,
"step": 351
},
{
"epoch": 1.4287886467308666,
"grad_norm": 0.5511609873924513,
"learning_rate": 3.592304568223542e-05,
"loss": 0.5067,
"step": 352
},
{
"epoch": 1.432843385707045,
"grad_norm": 0.5328538616416393,
"learning_rate": 3.588863714278826e-05,
"loss": 0.5127,
"step": 353
},
{
"epoch": 1.4368981246832235,
"grad_norm": 0.5488266333980802,
"learning_rate": 3.585410063844186e-05,
"loss": 0.5441,
"step": 354
},
{
"epoch": 1.440952863659402,
"grad_norm": 0.504600725213016,
"learning_rate": 3.581943644734846e-05,
"loss": 0.5341,
"step": 355
},
{
"epoch": 1.4450076026355803,
"grad_norm": 0.5029063722907915,
"learning_rate": 3.578464484868869e-05,
"loss": 0.5359,
"step": 356
},
{
"epoch": 1.4490623416117587,
"grad_norm": 0.4956156311289888,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.5371,
"step": 357
},
{
"epoch": 1.4531170805879372,
"grad_norm": 0.4906625786417074,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.5212,
"step": 358
},
{
"epoch": 1.4571718195641155,
"grad_norm": 0.5964700062030917,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.5648,
"step": 359
},
{
"epoch": 1.461226558540294,
"grad_norm": 0.4516001940398592,
"learning_rate": 3.564420999786543e-05,
"loss": 0.5319,
"step": 360
},
{
"epoch": 1.4652812975164724,
"grad_norm": 0.6184123814858841,
"learning_rate": 3.560878558491842e-05,
"loss": 0.5228,
"step": 361
},
{
"epoch": 1.4693360364926509,
"grad_norm": 0.4956415544231763,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.5413,
"step": 362
},
{
"epoch": 1.4733907754688291,
"grad_norm": 0.5060737178518595,
"learning_rate": 3.553755991230039e-05,
"loss": 0.5374,
"step": 363
},
{
"epoch": 1.4774455144450076,
"grad_norm": 0.5207016195924674,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.5308,
"step": 364
},
{
"epoch": 1.481500253421186,
"grad_norm": 0.4775863315679151,
"learning_rate": 3.546583369120419e-05,
"loss": 0.5224,
"step": 365
},
{
"epoch": 1.4855549923973643,
"grad_norm": 0.46080617734858675,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.5288,
"step": 366
},
{
"epoch": 1.4896097313735428,
"grad_norm": 0.66877228131714,
"learning_rate": 3.539360923231766e-05,
"loss": 0.5241,
"step": 367
},
{
"epoch": 1.4936644703497213,
"grad_norm": 0.48274572205584454,
"learning_rate": 3.535731089018394e-05,
"loss": 0.5278,
"step": 368
},
{
"epoch": 1.4977192093258997,
"grad_norm": 0.5116727879741129,
"learning_rate": 3.532088886237956e-05,
"loss": 0.5412,
"step": 369
},
{
"epoch": 1.501773948302078,
"grad_norm": 0.44640795006887096,
"learning_rate": 3.528434344224253e-05,
"loss": 0.553,
"step": 370
},
{
"epoch": 1.5058286872782565,
"grad_norm": 0.45238109884301936,
"learning_rate": 3.524767492410464e-05,
"loss": 0.512,
"step": 371
},
{
"epoch": 1.5098834262544347,
"grad_norm": 0.4870707312978545,
"learning_rate": 3.521088360328908e-05,
"loss": 0.5421,
"step": 372
},
{
"epoch": 1.5139381652306132,
"grad_norm": 0.48851249073331915,
"learning_rate": 3.517396977610811e-05,
"loss": 0.553,
"step": 373
},
{
"epoch": 1.5179929042067917,
"grad_norm": 0.4956952448579774,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.5423,
"step": 374
},
{
"epoch": 1.5220476431829701,
"grad_norm": 0.5782921657739899,
"learning_rate": 3.509977579282971e-05,
"loss": 0.5271,
"step": 375
},
{
"epoch": 1.5261023821591486,
"grad_norm": 0.5249983066236462,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.5403,
"step": 376
},
{
"epoch": 1.5301571211353269,
"grad_norm": 0.5478128513689535,
"learning_rate": 3.502509536445719e-05,
"loss": 0.5252,
"step": 377
},
{
"epoch": 1.5342118601115053,
"grad_norm": 0.6173517839190047,
"learning_rate": 3.498757348458147e-05,
"loss": 0.5515,
"step": 378
},
{
"epoch": 1.5382665990876836,
"grad_norm": 0.4641534498295412,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.5149,
"step": 379
},
{
"epoch": 1.542321338063862,
"grad_norm": 0.5500271294123625,
"learning_rate": 3.491216790442899e-05,
"loss": 0.5441,
"step": 380
},
{
"epoch": 1.5463760770400405,
"grad_norm": 0.5498120190656972,
"learning_rate": 3.487428481145839e-05,
"loss": 0.5243,
"step": 381
},
{
"epoch": 1.550430816016219,
"grad_norm": 0.5885044583121567,
"learning_rate": 3.483628192304278e-05,
"loss": 0.5313,
"step": 382
},
{
"epoch": 1.5544855549923975,
"grad_norm": 0.5766803604146636,
"learning_rate": 3.479815954525219e-05,
"loss": 0.5281,
"step": 383
},
{
"epoch": 1.5585402939685757,
"grad_norm": 0.5243422140746384,
"learning_rate": 3.475991798511899e-05,
"loss": 0.5459,
"step": 384
},
{
"epoch": 1.5625950329447542,
"grad_norm": 0.4944265458657894,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.5276,
"step": 385
},
{
"epoch": 1.5666497719209325,
"grad_norm": 0.5326287391447962,
"learning_rate": 3.468307855075128e-05,
"loss": 0.5563,
"step": 386
},
{
"epoch": 1.570704510897111,
"grad_norm": 2.6035134395372577,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.5476,
"step": 387
},
{
"epoch": 1.5747592498732894,
"grad_norm": 0.6258440108575806,
"learning_rate": 3.460576609535163e-05,
"loss": 0.5322,
"step": 388
},
{
"epoch": 1.5788139888494679,
"grad_norm": 0.4810662385963842,
"learning_rate": 3.456693326250006e-05,
"loss": 0.5428,
"step": 389
},
{
"epoch": 1.5828687278256464,
"grad_norm": 0.5623407303060836,
"learning_rate": 3.452798310957058e-05,
"loss": 0.5571,
"step": 390
},
{
"epoch": 1.5869234668018246,
"grad_norm": 0.44396895174513346,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.5296,
"step": 391
},
{
"epoch": 1.590978205778003,
"grad_norm": 0.4958270417824638,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.5147,
"step": 392
},
{
"epoch": 1.5950329447541813,
"grad_norm": 0.46159842793845884,
"learning_rate": 3.441043187201574e-05,
"loss": 0.5363,
"step": 393
},
{
"epoch": 1.5990876837303598,
"grad_norm": 0.5161643738943553,
"learning_rate": 3.437101558517728e-05,
"loss": 0.5268,
"step": 394
},
{
"epoch": 1.6031424227065383,
"grad_norm": 0.4820851667279659,
"learning_rate": 3.433148355615496e-05,
"loss": 0.5133,
"step": 395
},
{
"epoch": 1.6071971616827168,
"grad_norm": 0.48513810682155595,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.5373,
"step": 396
},
{
"epoch": 1.6112519006588952,
"grad_norm": 0.5181766436927846,
"learning_rate": 3.425207354603043e-05,
"loss": 0.5363,
"step": 397
},
{
"epoch": 1.6153066396350735,
"grad_norm": 0.5059590753444507,
"learning_rate": 3.421219620448553e-05,
"loss": 0.554,
"step": 398
},
{
"epoch": 1.619361378611252,
"grad_norm": 0.4940994689949252,
"learning_rate": 3.417220439986623e-05,
"loss": 0.5288,
"step": 399
},
{
"epoch": 1.6234161175874302,
"grad_norm": 0.5179628733409677,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.5396,
"step": 400
},
{
"epoch": 1.6274708565636087,
"grad_norm": 0.4851519221329909,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.5385,
"step": 401
},
{
"epoch": 1.6315255955397872,
"grad_norm": 0.5941817086304318,
"learning_rate": 3.405154543304065e-05,
"loss": 0.5135,
"step": 402
},
{
"epoch": 1.6355803345159656,
"grad_norm": 0.561310083920348,
"learning_rate": 3.401109900618843e-05,
"loss": 0.5467,
"step": 403
},
{
"epoch": 1.639635073492144,
"grad_norm": 0.5420360500350442,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.5539,
"step": 404
},
{
"epoch": 1.6436898124683224,
"grad_norm": 0.5939028717263156,
"learning_rate": 3.392986794874714e-05,
"loss": 0.5272,
"step": 405
},
{
"epoch": 1.6477445514445006,
"grad_norm": 0.5090038493094223,
"learning_rate": 3.388908397238184e-05,
"loss": 0.556,
"step": 406
},
{
"epoch": 1.651799290420679,
"grad_norm": 0.5590088456106946,
"learning_rate": 3.384818813524362e-05,
"loss": 0.5307,
"step": 407
},
{
"epoch": 1.6558540293968576,
"grad_norm": 0.4972841137206524,
"learning_rate": 3.380718076670195e-05,
"loss": 0.5131,
"step": 408
},
{
"epoch": 1.659908768373036,
"grad_norm": 0.6111036878491439,
"learning_rate": 3.376606219702454e-05,
"loss": 0.5235,
"step": 409
},
{
"epoch": 1.6639635073492145,
"grad_norm": 0.5252236211887534,
"learning_rate": 3.372483275737468e-05,
"loss": 0.5489,
"step": 410
},
{
"epoch": 1.6680182463253928,
"grad_norm": 0.5599101064971196,
"learning_rate": 3.368349277980861e-05,
"loss": 0.5441,
"step": 411
},
{
"epoch": 1.6720729853015712,
"grad_norm": 0.5522006831385899,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.5462,
"step": 412
},
{
"epoch": 1.6761277242777495,
"grad_norm": 0.5580887951556128,
"learning_rate": 3.360048254360144e-05,
"loss": 0.5558,
"step": 413
},
{
"epoch": 1.680182463253928,
"grad_norm": 0.520869625593591,
"learning_rate": 3.355881295351336e-05,
"loss": 0.5222,
"step": 414
},
{
"epoch": 1.6842372022301064,
"grad_norm": 0.5003440341412961,
"learning_rate": 3.351703416260975e-05,
"loss": 0.5646,
"step": 415
},
{
"epoch": 1.688291941206285,
"grad_norm": 0.5364881366475122,
"learning_rate": 3.347514650737126e-05,
"loss": 0.5386,
"step": 416
},
{
"epoch": 1.6923466801824634,
"grad_norm": 0.4914345427547982,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.535,
"step": 417
},
{
"epoch": 1.6964014191586416,
"grad_norm": 0.5432848620089233,
"learning_rate": 3.339104595419334e-05,
"loss": 0.5453,
"step": 418
},
{
"epoch": 1.70045615813482,
"grad_norm": 0.5458384243945,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.5255,
"step": 419
},
{
"epoch": 1.7045108971109983,
"grad_norm": 0.5048768442341492,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.5138,
"step": 420
},
{
"epoch": 1.7085656360871768,
"grad_norm": 0.5830659104068939,
"learning_rate": 3.326408710419996e-05,
"loss": 0.5322,
"step": 421
},
{
"epoch": 1.7126203750633553,
"grad_norm": 0.4421170526539174,
"learning_rate": 3.322155337795454e-05,
"loss": 0.5317,
"step": 422
},
{
"epoch": 1.7166751140395338,
"grad_norm": 0.46738927578621453,
"learning_rate": 3.317891316713587e-05,
"loss": 0.5268,
"step": 423
},
{
"epoch": 1.7207298530157122,
"grad_norm": 0.43954399294717533,
"learning_rate": 3.313616681516231e-05,
"loss": 0.5033,
"step": 424
},
{
"epoch": 1.7247845919918905,
"grad_norm": 0.5301730762424381,
"learning_rate": 3.309331466630713e-05,
"loss": 0.5359,
"step": 425
},
{
"epoch": 1.728839330968069,
"grad_norm": 0.4974641831720144,
"learning_rate": 3.305035706569563e-05,
"loss": 0.5369,
"step": 426
},
{
"epoch": 1.7328940699442472,
"grad_norm": 0.5085663630067726,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.5121,
"step": 427
},
{
"epoch": 1.7369488089204257,
"grad_norm": 0.4781764748466198,
"learning_rate": 3.296412689394864e-05,
"loss": 0.5424,
"step": 428
},
{
"epoch": 1.7410035478966042,
"grad_norm": 0.5022048925017011,
"learning_rate": 3.292085501729909e-05,
"loss": 0.5283,
"step": 429
},
{
"epoch": 1.7450582868727826,
"grad_norm": 0.49529966343334775,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.5245,
"step": 430
},
{
"epoch": 1.7491130258489611,
"grad_norm": 0.47802835185709214,
"learning_rate": 3.283399942497381e-05,
"loss": 0.5233,
"step": 431
},
{
"epoch": 1.7531677648251394,
"grad_norm": 0.5510415336443616,
"learning_rate": 3.279041640882108e-05,
"loss": 0.5509,
"step": 432
},
{
"epoch": 1.7572225038013178,
"grad_norm": 0.48110439821751527,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.5458,
"step": 433
},
{
"epoch": 1.761277242777496,
"grad_norm": 0.5651952979493425,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.538,
"step": 434
},
{
"epoch": 1.7653319817536746,
"grad_norm": 0.4415340649904317,
"learning_rate": 3.265905069502244e-05,
"loss": 0.5248,
"step": 435
},
{
"epoch": 1.769386720729853,
"grad_norm": 0.5882088681415918,
"learning_rate": 3.261505774420016e-05,
"loss": 0.5433,
"step": 436
},
{
"epoch": 1.7734414597060315,
"grad_norm": 0.8980241511942082,
"learning_rate": 3.257096319343707e-05,
"loss": 0.5716,
"step": 437
},
{
"epoch": 1.77749619868221,
"grad_norm": 0.509641809505027,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.5305,
"step": 438
},
{
"epoch": 1.7815509376583882,
"grad_norm": 0.4808299965743913,
"learning_rate": 3.248247071342966e-05,
"loss": 0.5261,
"step": 439
},
{
"epoch": 1.7856056766345665,
"grad_norm": 0.5334613562199398,
"learning_rate": 3.243807349689161e-05,
"loss": 0.5338,
"step": 440
},
{
"epoch": 1.789660415610745,
"grad_norm": 0.49099683256519816,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.5019,
"step": 441
},
{
"epoch": 1.7937151545869234,
"grad_norm": 0.5517096871871732,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.5405,
"step": 442
},
{
"epoch": 1.797769893563102,
"grad_norm": 0.4783643584397025,
"learning_rate": 3.230428223438075e-05,
"loss": 0.525,
"step": 443
},
{
"epoch": 1.8018246325392804,
"grad_norm": 0.46894895880465415,
"learning_rate": 3.225948647317459e-05,
"loss": 0.52,
"step": 444
},
{
"epoch": 1.8058793715154589,
"grad_norm": 0.45510230722092476,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.5417,
"step": 445
},
{
"epoch": 1.809934110491637,
"grad_norm": 0.4816100222096817,
"learning_rate": 3.216959910368034e-05,
"loss": 0.5475,
"step": 446
},
{
"epoch": 1.8139888494678154,
"grad_norm": 0.45400023942209033,
"learning_rate": 3.212450821933277e-05,
"loss": 0.5401,
"step": 447
},
{
"epoch": 1.8180435884439938,
"grad_norm": 0.4615435336089353,
"learning_rate": 3.207931968586281e-05,
"loss": 0.508,
"step": 448
},
{
"epoch": 1.8220983274201723,
"grad_norm": 0.49089976896884613,
"learning_rate": 3.203403386721272e-05,
"loss": 0.5577,
"step": 449
},
{
"epoch": 1.8261530663963508,
"grad_norm": 0.5509226939141177,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.5473,
"step": 450
},
{
"epoch": 1.8302078053725293,
"grad_norm": 0.44818948714898676,
"learning_rate": 3.194317183405573e-05,
"loss": 0.5395,
"step": 451
},
{
"epoch": 1.8342625443487075,
"grad_norm": 0.5055755261923557,
"learning_rate": 3.189759635133914e-05,
"loss": 0.5512,
"step": 452
},
{
"epoch": 1.838317283324886,
"grad_norm": 0.4109031998193964,
"learning_rate": 3.185192504701718e-05,
"loss": 0.5475,
"step": 453
},
{
"epoch": 1.8423720223010642,
"grad_norm": 0.5786829193210274,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.525,
"step": 454
},
{
"epoch": 1.8464267612772427,
"grad_norm": 0.4486101306631423,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.5645,
"step": 455
},
{
"epoch": 1.8504815002534212,
"grad_norm": 0.5297784092734797,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.5422,
"step": 456
},
{
"epoch": 1.8545362392295996,
"grad_norm": 0.49083680759995885,
"learning_rate": 3.166828898179731e-05,
"loss": 0.5123,
"step": 457
},
{
"epoch": 1.8585909782057781,
"grad_norm": 0.4864951857743798,
"learning_rate": 3.162214410223536e-05,
"loss": 0.5214,
"step": 458
},
{
"epoch": 1.8626457171819564,
"grad_norm": 0.460905485383442,
"learning_rate": 3.157590561952257e-05,
"loss": 0.5104,
"step": 459
},
{
"epoch": 1.8667004561581348,
"grad_norm": 0.5165462744307608,
"learning_rate": 3.152957390605732e-05,
"loss": 0.5546,
"step": 460
},
{
"epoch": 1.870755195134313,
"grad_norm": 0.43255359476263083,
"learning_rate": 3.148314933498886e-05,
"loss": 0.5366,
"step": 461
},
{
"epoch": 1.8748099341104916,
"grad_norm": 0.48036008287513215,
"learning_rate": 3.143663228021431e-05,
"loss": 0.5402,
"step": 462
},
{
"epoch": 1.87886467308667,
"grad_norm": 0.47314553250052255,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.5319,
"step": 463
},
{
"epoch": 1.8829194120628485,
"grad_norm": 0.514252925655343,
"learning_rate": 3.134332221885661e-05,
"loss": 0.5347,
"step": 464
},
{
"epoch": 1.886974151039027,
"grad_norm": 0.5142539640243031,
"learning_rate": 3.129652996377987e-05,
"loss": 0.5432,
"step": 465
},
{
"epoch": 1.8910288900152052,
"grad_norm": 0.5315747901151392,
"learning_rate": 3.12496467280038e-05,
"loss": 0.5311,
"step": 466
},
{
"epoch": 1.8950836289913837,
"grad_norm": 0.44734745454945035,
"learning_rate": 3.120267288911952e-05,
"loss": 0.5279,
"step": 467
},
{
"epoch": 1.899138367967562,
"grad_norm": 0.49504605802902835,
"learning_rate": 3.11556088254479e-05,
"loss": 0.5387,
"step": 468
},
{
"epoch": 1.9031931069437404,
"grad_norm": 0.5009752421254353,
"learning_rate": 3.11084549160364e-05,
"loss": 0.5406,
"step": 469
},
{
"epoch": 1.907247845919919,
"grad_norm": 0.4655278319823632,
"learning_rate": 3.106121154065615e-05,
"loss": 0.5385,
"step": 470
},
{
"epoch": 1.9113025848960974,
"grad_norm": 0.49026279539403034,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.5142,
"step": 471
},
{
"epoch": 1.9153573238722759,
"grad_norm": 0.477240083669834,
"learning_rate": 3.096645791467348e-05,
"loss": 0.5172,
"step": 472
},
{
"epoch": 1.9194120628484541,
"grad_norm": 0.5549564840556759,
"learning_rate": 3.091894842720373e-05,
"loss": 0.5375,
"step": 473
},
{
"epoch": 1.9234668018246326,
"grad_norm": 0.5179330320251824,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.5543,
"step": 474
},
{
"epoch": 1.9275215408008108,
"grad_norm": 0.49761554362326116,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.5321,
"step": 475
},
{
"epoch": 1.9315762797769893,
"grad_norm": 0.5209631079429944,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.5348,
"step": 476
},
{
"epoch": 1.9356310187531678,
"grad_norm": 0.5078173890091245,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.5303,
"step": 477
},
{
"epoch": 1.9396857577293463,
"grad_norm": 0.48133933547201196,
"learning_rate": 3.06800895716334e-05,
"loss": 0.5388,
"step": 478
},
{
"epoch": 1.9437404967055247,
"grad_norm": 0.5233547338398804,
"learning_rate": 3.063205821011029e-05,
"loss": 0.5278,
"step": 479
},
{
"epoch": 1.947795235681703,
"grad_norm": 0.49342115532441005,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.5438,
"step": 480
},
{
"epoch": 1.9518499746578812,
"grad_norm": 0.4244972146497058,
"learning_rate": 3.053573898718618e-05,
"loss": 0.5315,
"step": 481
},
{
"epoch": 1.9559047136340597,
"grad_norm": 0.5229600740799853,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.5305,
"step": 482
},
{
"epoch": 1.9599594526102382,
"grad_norm": 0.46883505549131543,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.536,
"step": 483
},
{
"epoch": 1.9640141915864167,
"grad_norm": 0.5634634624180843,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.5277,
"step": 484
},
{
"epoch": 1.9680689305625951,
"grad_norm": 0.45877238281158705,
"learning_rate": 3.034208541656255e-05,
"loss": 0.548,
"step": 485
},
{
"epoch": 1.9721236695387736,
"grad_norm": 0.507123641273262,
"learning_rate": 3.029346281309218e-05,
"loss": 0.5317,
"step": 486
},
{
"epoch": 1.9761784085149519,
"grad_norm": 0.45884827938264094,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.5562,
"step": 487
},
{
"epoch": 1.98023314749113,
"grad_norm": 0.4877600804194238,
"learning_rate": 3.019596929201569e-05,
"loss": 0.5574,
"step": 488
},
{
"epoch": 1.9842878864673086,
"grad_norm": 0.45635131041346894,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.5204,
"step": 489
},
{
"epoch": 1.988342625443487,
"grad_norm": 0.4974258569485135,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.5356,
"step": 490
},
{
"epoch": 1.9923973644196655,
"grad_norm": 0.47201410049283277,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.5333,
"step": 491
},
{
"epoch": 1.996452103395844,
"grad_norm": 0.4182359857009549,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.5179,
"step": 492
},
{
"epoch": 2.0040506329113925,
"grad_norm": 0.8514659742359979,
"learning_rate": 2.995080534235264e-05,
"loss": 0.4399,
"step": 493
},
{
"epoch": 2.008101265822785,
"grad_norm": 0.6349238377917809,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.389,
"step": 494
},
{
"epoch": 2.012151898734177,
"grad_norm": 0.9061423615945333,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.3673,
"step": 495
},
{
"epoch": 2.0162025316455696,
"grad_norm": 0.6522084160793656,
"learning_rate": 2.980274210295326e-05,
"loss": 0.3749,
"step": 496
},
{
"epoch": 2.020253164556962,
"grad_norm": 0.6077870581138277,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.3551,
"step": 497
},
{
"epoch": 2.0243037974683546,
"grad_norm": 0.553831455743028,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.3934,
"step": 498
},
{
"epoch": 2.0283544303797467,
"grad_norm": 0.5464039098966776,
"learning_rate": 2.965396831793362e-05,
"loss": 0.3804,
"step": 499
},
{
"epoch": 2.0324050632911392,
"grad_norm": 0.5718192802534139,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.3831,
"step": 500
},
{
"epoch": 2.0364556962025318,
"grad_norm": 0.6045329581908979,
"learning_rate": 2.955439637024526e-05,
"loss": 0.3872,
"step": 501
},
{
"epoch": 2.0405063291139243,
"grad_norm": 0.5196942241125314,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.3935,
"step": 502
},
{
"epoch": 2.0445569620253163,
"grad_norm": 0.5754960964732723,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.3941,
"step": 503
},
{
"epoch": 2.048607594936709,
"grad_norm": 0.5752782606552806,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.3729,
"step": 504
},
{
"epoch": 2.0526582278481014,
"grad_norm": 0.5621345611474224,
"learning_rate": 2.93543322968538e-05,
"loss": 0.3661,
"step": 505
},
{
"epoch": 2.056708860759494,
"grad_norm": 0.634996690632848,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.3799,
"step": 506
},
{
"epoch": 2.060759493670886,
"grad_norm": 0.5718480588242677,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.397,
"step": 507
},
{
"epoch": 2.0648101265822785,
"grad_norm": 0.496144593015699,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.3691,
"step": 508
},
{
"epoch": 2.068860759493671,
"grad_norm": 0.6145546482098673,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.3838,
"step": 509
},
{
"epoch": 2.0729113924050635,
"grad_norm": 0.46273333127496863,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.3921,
"step": 510
},
{
"epoch": 2.0769620253164556,
"grad_norm": 0.4879514365498469,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.3722,
"step": 511
},
{
"epoch": 2.081012658227848,
"grad_norm": 0.4688509947375713,
"learning_rate": 2.900133526131588e-05,
"loss": 0.3709,
"step": 512
},
{
"epoch": 2.0850632911392406,
"grad_norm": 0.4805031988742675,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.367,
"step": 513
},
{
"epoch": 2.089113924050633,
"grad_norm": 0.5194697326094561,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.3781,
"step": 514
},
{
"epoch": 2.093164556962025,
"grad_norm": 0.5074432805079186,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.3928,
"step": 515
},
{
"epoch": 2.0972151898734177,
"grad_norm": 0.4957122474905427,
"learning_rate": 2.879801882631847e-05,
"loss": 0.3808,
"step": 516
},
{
"epoch": 2.1012658227848102,
"grad_norm": 0.4615971405441044,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.3744,
"step": 517
},
{
"epoch": 2.1053164556962027,
"grad_norm": 0.4395519719873443,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.3628,
"step": 518
},
{
"epoch": 2.109367088607595,
"grad_norm": 0.4646395339909091,
"learning_rate": 2.864478605989494e-05,
"loss": 0.3815,
"step": 519
},
{
"epoch": 2.1134177215189873,
"grad_norm": 0.4436803364537414,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.3877,
"step": 520
},
{
"epoch": 2.11746835443038,
"grad_norm": 0.44069091066092003,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.3986,
"step": 521
},
{
"epoch": 2.1215189873417724,
"grad_norm": 0.4292878422341503,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.3846,
"step": 522
},
{
"epoch": 2.1255696202531644,
"grad_norm": 0.4180214731459623,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.3721,
"step": 523
},
{
"epoch": 2.129620253164557,
"grad_norm": 0.45778394470256106,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.3741,
"step": 524
},
{
"epoch": 2.1336708860759495,
"grad_norm": 0.4414769691721126,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.3708,
"step": 525
},
{
"epoch": 2.137721518987342,
"grad_norm": 0.44197762460745843,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.3899,
"step": 526
},
{
"epoch": 2.141772151898734,
"grad_norm": 0.46521853549160674,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.396,
"step": 527
},
{
"epoch": 2.1458227848101266,
"grad_norm": 0.4415683747545979,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.382,
"step": 528
},
{
"epoch": 2.149873417721519,
"grad_norm": 0.5139737887003896,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.3901,
"step": 529
},
{
"epoch": 2.1539240506329116,
"grad_norm": 0.45221784610277393,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.371,
"step": 530
},
{
"epoch": 2.1579746835443037,
"grad_norm": 0.4655033671224608,
"learning_rate": 2.802570062143278e-05,
"loss": 0.3961,
"step": 531
},
{
"epoch": 2.162025316455696,
"grad_norm": 0.4667442846665353,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.3775,
"step": 532
},
{
"epoch": 2.1660759493670887,
"grad_norm": 0.4394568000032489,
"learning_rate": 2.792159532078314e-05,
"loss": 0.367,
"step": 533
},
{
"epoch": 2.170126582278481,
"grad_norm": 0.40914103030178883,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.3725,
"step": 534
},
{
"epoch": 2.1741772151898733,
"grad_norm": 0.4151878532671488,
"learning_rate": 2.781723482288857e-05,
"loss": 0.3762,
"step": 535
},
{
"epoch": 2.178227848101266,
"grad_norm": 0.4403755362145964,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.38,
"step": 536
},
{
"epoch": 2.1822784810126583,
"grad_norm": 0.41790693126486,
"learning_rate": 2.771262248976272e-05,
"loss": 0.3392,
"step": 537
},
{
"epoch": 2.186329113924051,
"grad_norm": 0.458924104647922,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.3753,
"step": 538
},
{
"epoch": 2.190379746835443,
"grad_norm": 0.43690722027568707,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.3771,
"step": 539
},
{
"epoch": 2.1944303797468354,
"grad_norm": 0.44516561175180225,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.3522,
"step": 540
},
{
"epoch": 2.198481012658228,
"grad_norm": 0.44725695226517453,
"learning_rate": 2.7502655806328e-05,
"loss": 0.3914,
"step": 541
},
{
"epoch": 2.2025316455696204,
"grad_norm": 0.4025650578072138,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.3767,
"step": 542
},
{
"epoch": 2.2065822784810125,
"grad_norm": 0.4768323396497322,
"learning_rate": 2.739730822017673e-05,
"loss": 0.386,
"step": 543
},
{
"epoch": 2.210632911392405,
"grad_norm": 0.43000345511115917,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.3899,
"step": 544
},
{
"epoch": 2.2146835443037975,
"grad_norm": 0.4134392985344769,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.3814,
"step": 545
},
{
"epoch": 2.21873417721519,
"grad_norm": 0.4379128240932747,
"learning_rate": 2.723884107774236e-05,
"loss": 0.3812,
"step": 546
},
{
"epoch": 2.222784810126582,
"grad_norm": 0.4263783441745549,
"learning_rate": 2.718590152796239e-05,
"loss": 0.3707,
"step": 547
},
{
"epoch": 2.2268354430379746,
"grad_norm": 0.41135737054801624,
"learning_rate": 2.71329041039194e-05,
"loss": 0.3719,
"step": 548
},
{
"epoch": 2.230886075949367,
"grad_norm": 0.4041013134577113,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.3827,
"step": 549
},
{
"epoch": 2.2349367088607597,
"grad_norm": 0.4189157852552827,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.3756,
"step": 550
},
{
"epoch": 2.2389873417721518,
"grad_norm": 0.416942921606829,
"learning_rate": 2.697356885686189e-05,
"loss": 0.3697,
"step": 551
},
{
"epoch": 2.2430379746835443,
"grad_norm": 0.41674426895198113,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.3605,
"step": 552
},
{
"epoch": 2.247088607594937,
"grad_norm": 0.406372058924128,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.3925,
"step": 553
},
{
"epoch": 2.2511392405063293,
"grad_norm": 0.4382254459998848,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.3749,
"step": 554
},
{
"epoch": 2.2551898734177214,
"grad_norm": 0.4815894076427739,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.3908,
"step": 555
},
{
"epoch": 2.259240506329114,
"grad_norm": 0.45091781178998813,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.3737,
"step": 556
},
{
"epoch": 2.2632911392405064,
"grad_norm": 0.49430336208261255,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.3731,
"step": 557
},
{
"epoch": 2.267341772151899,
"grad_norm": 0.40777031627160537,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.3851,
"step": 558
},
{
"epoch": 2.271392405063291,
"grad_norm": 0.47250510727840067,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.3942,
"step": 559
},
{
"epoch": 2.2754430379746835,
"grad_norm": 0.42719932978501135,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.378,
"step": 560
},
{
"epoch": 2.279493670886076,
"grad_norm": 0.4403029109216816,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.3743,
"step": 561
},
{
"epoch": 2.2835443037974685,
"grad_norm": 0.4599031556625358,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.3831,
"step": 562
},
{
"epoch": 2.2875949367088606,
"grad_norm": 0.47139164013460116,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.3779,
"step": 563
},
{
"epoch": 2.291645569620253,
"grad_norm": 0.4362106117279933,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.3801,
"step": 564
},
{
"epoch": 2.2956962025316456,
"grad_norm": 0.4602285139217017,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.3884,
"step": 565
},
{
"epoch": 2.299746835443038,
"grad_norm": 0.43036654970037863,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.3922,
"step": 566
},
{
"epoch": 2.3037974683544302,
"grad_norm": 0.4034970006165686,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.3822,
"step": 567
},
{
"epoch": 2.3078481012658227,
"grad_norm": 0.4409206494805822,
"learning_rate": 2.606146264569603e-05,
"loss": 0.387,
"step": 568
},
{
"epoch": 2.3118987341772153,
"grad_norm": 0.4189460471882211,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.3434,
"step": 569
},
{
"epoch": 2.3159493670886078,
"grad_norm": 0.4272460465614016,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.3991,
"step": 570
},
{
"epoch": 2.32,
"grad_norm": 0.41491420257245454,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.3749,
"step": 571
},
{
"epoch": 2.3240506329113924,
"grad_norm": 0.47338763427153185,
"learning_rate": 2.584472010606015e-05,
"loss": 0.3862,
"step": 572
},
{
"epoch": 2.328101265822785,
"grad_norm": 0.4163074684373198,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.3867,
"step": 573
},
{
"epoch": 2.3321518987341774,
"grad_norm": 0.471018531067937,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.3544,
"step": 574
},
{
"epoch": 2.3362025316455695,
"grad_norm": 0.43442359660927177,
"learning_rate": 2.568166741322048e-05,
"loss": 0.3727,
"step": 575
},
{
"epoch": 2.340253164556962,
"grad_norm": 0.5302753589909951,
"learning_rate": 2.56272244128509e-05,
"loss": 0.3919,
"step": 576
},
{
"epoch": 2.3443037974683545,
"grad_norm": 0.4112882784464308,
"learning_rate": 2.55727360915895e-05,
"loss": 0.3657,
"step": 577
},
{
"epoch": 2.348354430379747,
"grad_norm": 0.5008569811448608,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.4041,
"step": 578
},
{
"epoch": 2.352405063291139,
"grad_norm": 0.4233472626503316,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.3763,
"step": 579
},
{
"epoch": 2.3564556962025316,
"grad_norm": 0.48799912544884766,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.3901,
"step": 580
},
{
"epoch": 2.360506329113924,
"grad_norm": 0.4169614626129584,
"learning_rate": 2.535433837985559e-05,
"loss": 0.3729,
"step": 581
},
{
"epoch": 2.3645569620253166,
"grad_norm": 0.45754448910417467,
"learning_rate": 2.529963004393324e-05,
"loss": 0.381,
"step": 582
},
{
"epoch": 2.3686075949367087,
"grad_norm": 0.43770243940697023,
"learning_rate": 2.524487902551908e-05,
"loss": 0.3585,
"step": 583
},
{
"epoch": 2.372658227848101,
"grad_norm": 0.4258747818832008,
"learning_rate": 2.519008576557029e-05,
"loss": 0.3968,
"step": 584
},
{
"epoch": 2.3767088607594937,
"grad_norm": 0.4398642474992483,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.3897,
"step": 585
},
{
"epoch": 2.3807594936708862,
"grad_norm": 0.41520773835991953,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.363,
"step": 586
},
{
"epoch": 2.3848101265822783,
"grad_norm": 0.4451833148839736,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.378,
"step": 587
},
{
"epoch": 2.388860759493671,
"grad_norm": 0.4311479119781443,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.3893,
"step": 588
},
{
"epoch": 2.3929113924050633,
"grad_norm": 0.43703915111345937,
"learning_rate": 2.491550129992345e-05,
"loss": 0.3788,
"step": 589
},
{
"epoch": 2.396962025316456,
"grad_norm": 0.4176600456529176,
"learning_rate": 2.486046386967024e-05,
"loss": 0.3667,
"step": 590
},
{
"epoch": 2.401012658227848,
"grad_norm": 0.41982478162628195,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.388,
"step": 591
},
{
"epoch": 2.4050632911392404,
"grad_norm": 0.39977671557861466,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.3817,
"step": 592
},
{
"epoch": 2.409113924050633,
"grad_norm": 0.4345144432237986,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.3804,
"step": 593
},
{
"epoch": 2.4131645569620255,
"grad_norm": 0.40733954810248213,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.369,
"step": 594
},
{
"epoch": 2.4172151898734175,
"grad_norm": 0.42889376853620476,
"learning_rate": 2.458469841202444e-05,
"loss": 0.3875,
"step": 595
},
{
"epoch": 2.42126582278481,
"grad_norm": 0.4106525470388025,
"learning_rate": 2.452943276859226e-05,
"loss": 0.3803,
"step": 596
},
{
"epoch": 2.4253164556962026,
"grad_norm": 0.4177030261720476,
"learning_rate": 2.447413064573125e-05,
"loss": 0.3857,
"step": 597
},
{
"epoch": 2.429367088607595,
"grad_norm": 0.39587497567752467,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.3875,
"step": 598
},
{
"epoch": 2.433417721518987,
"grad_norm": 0.46353358609537676,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.4015,
"step": 599
},
{
"epoch": 2.4374683544303797,
"grad_norm": 0.41487499334008693,
"learning_rate": 2.430800985597963e-05,
"loss": 0.3821,
"step": 600
},
{
"epoch": 2.441518987341772,
"grad_norm": 0.44465171632830636,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.3929,
"step": 601
},
{
"epoch": 2.4455696202531647,
"grad_norm": 0.44092919368685135,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.3648,
"step": 602
},
{
"epoch": 2.449620253164557,
"grad_norm": 0.44062554747559957,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.3842,
"step": 603
},
{
"epoch": 2.4536708860759493,
"grad_norm": 0.4698188980278592,
"learning_rate": 2.408603181104957e-05,
"loss": 0.3902,
"step": 604
},
{
"epoch": 2.457721518987342,
"grad_norm": 0.40831461197425645,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.3863,
"step": 605
},
{
"epoch": 2.4617721518987343,
"grad_norm": 0.4869423238423921,
"learning_rate": 2.397484355008521e-05,
"loss": 0.3924,
"step": 606
},
{
"epoch": 2.4658227848101264,
"grad_norm": 0.41456226962415793,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.3721,
"step": 607
},
{
"epoch": 2.469873417721519,
"grad_norm": 0.4408498532878629,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.3812,
"step": 608
},
{
"epoch": 2.4739240506329114,
"grad_norm": 0.4406169288650729,
"learning_rate": 2.380782218329337e-05,
"loss": 0.3747,
"step": 609
},
{
"epoch": 2.477974683544304,
"grad_norm": 0.4293197488150918,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.373,
"step": 610
},
{
"epoch": 2.482025316455696,
"grad_norm": 0.4386511917618956,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.3631,
"step": 611
},
{
"epoch": 2.4860759493670885,
"grad_norm": 0.4389837360068803,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.3786,
"step": 612
},
{
"epoch": 2.490126582278481,
"grad_norm": 0.40594255747539304,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.3743,
"step": 613
},
{
"epoch": 2.4941772151898736,
"grad_norm": 0.4330734763237425,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.3874,
"step": 614
},
{
"epoch": 2.4982278481012656,
"grad_norm": 0.4394737090135893,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.4031,
"step": 615
},
{
"epoch": 2.502278481012658,
"grad_norm": 0.45205534624227534,
"learning_rate": 2.341705325992901e-05,
"loss": 0.4057,
"step": 616
},
{
"epoch": 2.5063291139240507,
"grad_norm": 0.41409128599825973,
"learning_rate": 2.336111544604222e-05,
"loss": 0.3886,
"step": 617
},
{
"epoch": 2.510379746835443,
"grad_norm": 0.47581991481530356,
"learning_rate": 2.33051505621937e-05,
"loss": 0.3717,
"step": 618
},
{
"epoch": 2.5144303797468357,
"grad_norm": 0.4423291440906091,
"learning_rate": 2.324915905911693e-05,
"loss": 0.3756,
"step": 619
},
{
"epoch": 2.5184810126582278,
"grad_norm": 0.47092930287137563,
"learning_rate": 2.319314138775977e-05,
"loss": 0.3763,
"step": 620
},
{
"epoch": 2.5225316455696203,
"grad_norm": 0.4793937399640029,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.3674,
"step": 621
},
{
"epoch": 2.526582278481013,
"grad_norm": 0.4043774353406513,
"learning_rate": 2.308102934504593e-05,
"loss": 0.3713,
"step": 622
},
{
"epoch": 2.530632911392405,
"grad_norm": 0.4606441949085103,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.3701,
"step": 623
},
{
"epoch": 2.5346835443037974,
"grad_norm": 0.46009816105668844,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.3834,
"step": 624
},
{
"epoch": 2.53873417721519,
"grad_norm": 0.4119991386896808,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.3779,
"step": 625
},
{
"epoch": 2.5427848101265824,
"grad_norm": 0.4669020052794965,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.3636,
"step": 626
},
{
"epoch": 2.546835443037975,
"grad_norm": 0.46354206008873916,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.3742,
"step": 627
},
{
"epoch": 2.550886075949367,
"grad_norm": 0.4252775526478063,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.3962,
"step": 628
},
{
"epoch": 2.5549367088607595,
"grad_norm": 0.47465774495281027,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.3824,
"step": 629
},
{
"epoch": 2.558987341772152,
"grad_norm": 0.421692256516435,
"learning_rate": 2.26316247733093e-05,
"loss": 0.3659,
"step": 630
},
{
"epoch": 2.563037974683544,
"grad_norm": 0.4580246597726559,
"learning_rate": 2.257534907110328e-05,
"loss": 0.3717,
"step": 631
},
{
"epoch": 2.5670886075949366,
"grad_norm": 0.4621492621448863,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.3786,
"step": 632
},
{
"epoch": 2.571139240506329,
"grad_norm": 0.4752863774681814,
"learning_rate": 2.246273589557294e-05,
"loss": 0.3602,
"step": 633
},
{
"epoch": 2.5751898734177217,
"grad_norm": 0.42824990418849496,
"learning_rate": 2.240639932921966e-05,
"loss": 0.385,
"step": 634
},
{
"epoch": 2.579240506329114,
"grad_norm": 0.4098358299368642,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.3792,
"step": 635
},
{
"epoch": 2.5832911392405062,
"grad_norm": 0.43435825711162906,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.3784,
"step": 636
},
{
"epoch": 2.5873417721518988,
"grad_norm": 0.41667862858882404,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.365,
"step": 637
},
{
"epoch": 2.5913924050632913,
"grad_norm": 0.4505303321596465,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.405,
"step": 638
},
{
"epoch": 2.5954430379746833,
"grad_norm": 0.4070092134738687,
"learning_rate": 2.212443486523819e-05,
"loss": 0.3821,
"step": 639
},
{
"epoch": 2.599493670886076,
"grad_norm": 0.4543491482895516,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.3599,
"step": 640
},
{
"epoch": 2.6035443037974684,
"grad_norm": 0.42360153355796726,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.3969,
"step": 641
},
{
"epoch": 2.607594936708861,
"grad_norm": 0.3920020771867113,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.3807,
"step": 642
},
{
"epoch": 2.6116455696202534,
"grad_norm": 0.44887307943731053,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.3706,
"step": 643
},
{
"epoch": 2.6156962025316455,
"grad_norm": 0.3992780268593902,
"learning_rate": 2.184204266048803e-05,
"loss": 0.3676,
"step": 644
},
{
"epoch": 2.619746835443038,
"grad_norm": 0.415901225856746,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.3949,
"step": 645
},
{
"epoch": 2.6237974683544305,
"grad_norm": 0.4090818765960425,
"learning_rate": 2.17289787446841e-05,
"loss": 0.3804,
"step": 646
},
{
"epoch": 2.6278481012658226,
"grad_norm": 0.4131800708541863,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.3958,
"step": 647
},
{
"epoch": 2.631898734177215,
"grad_norm": 0.4263157909775713,
"learning_rate": 2.161585912916385e-05,
"loss": 0.3798,
"step": 648
},
{
"epoch": 2.6359493670886076,
"grad_norm": 0.39291049701338837,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.3945,
"step": 649
},
{
"epoch": 2.64,
"grad_norm": 0.3878734683088336,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.3522,
"step": 650
},
{
"epoch": 2.6440506329113926,
"grad_norm": 0.45832450848251055,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.372,
"step": 651
},
{
"epoch": 2.6481012658227847,
"grad_norm": 0.39461053125464596,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.3746,
"step": 652
},
{
"epoch": 2.6521518987341772,
"grad_norm": 0.3856506637445099,
"learning_rate": 2.133284032321232e-05,
"loss": 0.3887,
"step": 653
},
{
"epoch": 2.6562025316455697,
"grad_norm": 0.3960546128671372,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.3653,
"step": 654
},
{
"epoch": 2.660253164556962,
"grad_norm": 0.38557165372639396,
"learning_rate": 2.121955446741306e-05,
"loss": 0.3881,
"step": 655
},
{
"epoch": 2.6643037974683543,
"grad_norm": 0.41218011504196184,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.3815,
"step": 656
},
{
"epoch": 2.668354430379747,
"grad_norm": 0.42583623566175693,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.4014,
"step": 657
},
{
"epoch": 2.6724050632911394,
"grad_norm": 0.3809378393949005,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.3885,
"step": 658
},
{
"epoch": 2.676455696202532,
"grad_norm": 0.3906216862453747,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.3659,
"step": 659
},
{
"epoch": 2.680506329113924,
"grad_norm": 0.40799850112247493,
"learning_rate": 2.093617592756158e-05,
"loss": 0.4145,
"step": 660
},
{
"epoch": 2.6845569620253165,
"grad_norm": 0.39334696177520273,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.3763,
"step": 661
},
{
"epoch": 2.688607594936709,
"grad_norm": 0.4510106953320501,
"learning_rate": 2.082276853708978e-05,
"loss": 0.3749,
"step": 662
},
{
"epoch": 2.692658227848101,
"grad_norm": 0.4202402397472803,
"learning_rate": 2.076605467380071e-05,
"loss": 0.3754,
"step": 663
},
{
"epoch": 2.6967088607594936,
"grad_norm": 0.439002952818395,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.3868,
"step": 664
},
{
"epoch": 2.700759493670886,
"grad_norm": 0.42195715110805065,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.3862,
"step": 665
},
{
"epoch": 2.7048101265822786,
"grad_norm": 0.4124142472125217,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.3726,
"step": 666
},
{
"epoch": 2.708860759493671,
"grad_norm": 0.41449142844075326,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.3906,
"step": 667
},
{
"epoch": 2.712911392405063,
"grad_norm": 0.41708339342414497,
"learning_rate": 2.048240194884496e-05,
"loss": 0.3925,
"step": 668
},
{
"epoch": 2.7169620253164557,
"grad_norm": 0.42672534102130943,
"learning_rate": 2.042565792045695e-05,
"loss": 0.3809,
"step": 669
},
{
"epoch": 2.721012658227848,
"grad_norm": 0.3681459036950113,
"learning_rate": 2.036891046387857e-05,
"loss": 0.3856,
"step": 670
},
{
"epoch": 2.7250632911392403,
"grad_norm": 0.44252339116550843,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.3823,
"step": 671
},
{
"epoch": 2.729113924050633,
"grad_norm": 0.4079305528932944,
"learning_rate": 2.025540709431948e-05,
"loss": 0.3906,
"step": 672
},
{
"epoch": 2.7331645569620253,
"grad_norm": 0.40164959837109615,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.3807,
"step": 673
},
{
"epoch": 2.737215189873418,
"grad_norm": 0.38609473494953733,
"learning_rate": 2.014189549672245e-05,
"loss": 0.3746,
"step": 674
},
{
"epoch": 2.7412658227848103,
"grad_norm": 1.1350494039209302,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.3928,
"step": 675
},
{
"epoch": 2.7453164556962024,
"grad_norm": 0.4963194170270538,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.3902,
"step": 676
},
{
"epoch": 2.749367088607595,
"grad_norm": 0.427327526614533,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.3615,
"step": 677
},
{
"epoch": 2.7534177215189874,
"grad_norm": 0.47922896065252446,
"learning_rate": 1.991486224484123e-05,
"loss": 0.4028,
"step": 678
},
{
"epoch": 2.7574683544303795,
"grad_norm": 0.441189769787658,
"learning_rate": 1.985810450327756e-05,
"loss": 0.3878,
"step": 679
},
{
"epoch": 2.761518987341772,
"grad_norm": 0.41250057929751544,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.3624,
"step": 680
},
{
"epoch": 2.7655696202531646,
"grad_norm": 0.4695529407736641,
"learning_rate": 1.974459290568053e-05,
"loss": 0.3825,
"step": 681
},
{
"epoch": 2.769620253164557,
"grad_norm": 0.3966265027802706,
"learning_rate": 1.968783996385397e-05,
"loss": 0.362,
"step": 682
},
{
"epoch": 2.7736708860759496,
"grad_norm": 0.4024416401868956,
"learning_rate": 1.963108953612143e-05,
"loss": 0.378,
"step": 683
},
{
"epoch": 2.7777215189873417,
"grad_norm": 0.4170506669029113,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.3718,
"step": 684
},
{
"epoch": 2.781772151898734,
"grad_norm": 0.4332398218236451,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.3955,
"step": 685
},
{
"epoch": 2.7858227848101267,
"grad_norm": 0.4237296085059462,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.3637,
"step": 686
},
{
"epoch": 2.7898734177215188,
"grad_norm": 0.41504591325602147,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.3636,
"step": 687
},
{
"epoch": 2.7939240506329113,
"grad_norm": 0.5403379193757628,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.3945,
"step": 688
},
{
"epoch": 2.797974683544304,
"grad_norm": 0.4101441159839735,
"learning_rate": 1.92906653591874e-05,
"loss": 0.3827,
"step": 689
},
{
"epoch": 2.8020253164556963,
"grad_norm": 0.4234197936780648,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.3752,
"step": 690
},
{
"epoch": 2.806075949367089,
"grad_norm": 0.39637366299076726,
"learning_rate": 1.917723146291022e-05,
"loss": 0.387,
"step": 691
},
{
"epoch": 2.810126582278481,
"grad_norm": 0.3931711652317323,
"learning_rate": 1.912052422608584e-05,
"loss": 0.3785,
"step": 692
},
{
"epoch": 2.8141772151898734,
"grad_norm": 0.36437457992488465,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.4073,
"step": 693
},
{
"epoch": 2.818227848101266,
"grad_norm": 0.3870292079949069,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.3517,
"step": 694
},
{
"epoch": 2.822278481012658,
"grad_norm": 0.4138726301877295,
"learning_rate": 1.895044684123479e-05,
"loss": 0.3842,
"step": 695
},
{
"epoch": 2.8263291139240505,
"grad_norm": 0.42067235889024684,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.3746,
"step": 696
},
{
"epoch": 2.830379746835443,
"grad_norm": 0.393923361503678,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.3587,
"step": 697
},
{
"epoch": 2.8344303797468355,
"grad_norm": 0.43159474269177983,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.3673,
"step": 698
},
{
"epoch": 2.838481012658228,
"grad_norm": 0.40439460292257445,
"learning_rate": 1.872379746550743e-05,
"loss": 0.3794,
"step": 699
},
{
"epoch": 2.84253164556962,
"grad_norm": 0.40099408509512946,
"learning_rate": 1.866715967678769e-05,
"loss": 0.3983,
"step": 700
},
{
"epoch": 2.8465822784810126,
"grad_norm": 0.4376332719050142,
"learning_rate": 1.861053262258067e-05,
"loss": 0.3653,
"step": 701
},
{
"epoch": 2.850632911392405,
"grad_norm": 0.4132331093978461,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.3559,
"step": 702
},
{
"epoch": 2.8546835443037972,
"grad_norm": 1.7227765412204687,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.39,
"step": 703
},
{
"epoch": 2.8587341772151897,
"grad_norm": 0.4258890416578788,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.3465,
"step": 704
},
{
"epoch": 2.8627848101265823,
"grad_norm": 0.4332376304143481,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.3752,
"step": 705
},
{
"epoch": 2.8668354430379748,
"grad_norm": 0.4222924462581683,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.3662,
"step": 706
},
{
"epoch": 2.8708860759493673,
"grad_norm": 0.41837856304984805,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.3769,
"step": 707
},
{
"epoch": 2.8749367088607594,
"grad_norm": 0.43141509558522906,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.3791,
"step": 708
},
{
"epoch": 2.878987341772152,
"grad_norm": 0.417321750318619,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.3874,
"step": 709
},
{
"epoch": 2.8830379746835444,
"grad_norm": 0.4078955875821102,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.3747,
"step": 710
},
{
"epoch": 2.8870886075949365,
"grad_norm": 0.4128869979881167,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.3751,
"step": 711
},
{
"epoch": 2.891139240506329,
"grad_norm": 0.38726665699784746,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.3993,
"step": 712
},
{
"epoch": 2.8951898734177215,
"grad_norm": 0.3839015960360953,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.3855,
"step": 713
},
{
"epoch": 2.899240506329114,
"grad_norm": 0.4002774380570917,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.3599,
"step": 714
},
{
"epoch": 2.9032911392405065,
"grad_norm": 0.36345814099517143,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.368,
"step": 715
},
{
"epoch": 2.9073417721518986,
"grad_norm": 0.3911740568045173,
"learning_rate": 1.776272483901182e-05,
"loss": 0.3582,
"step": 716
},
{
"epoch": 2.911392405063291,
"grad_norm": 0.39753827282648824,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.343,
"step": 717
},
{
"epoch": 2.9154430379746836,
"grad_norm": 0.40221455277216717,
"learning_rate": 1.764995661794301e-05,
"loss": 0.4067,
"step": 718
},
{
"epoch": 2.9194936708860757,
"grad_norm": 0.4363619105989375,
"learning_rate": 1.759360067078035e-05,
"loss": 0.393,
"step": 719
},
{
"epoch": 2.923544303797468,
"grad_norm": 0.3876145888345104,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.3943,
"step": 720
},
{
"epoch": 2.9275949367088607,
"grad_norm": 0.3913732310462001,
"learning_rate": 1.748094737261012e-05,
"loss": 0.3808,
"step": 721
},
{
"epoch": 2.9316455696202532,
"grad_norm": 0.39921756291313915,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.3883,
"step": 722
},
{
"epoch": 2.9356962025316458,
"grad_norm": 0.3791669443725468,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.3841,
"step": 723
},
{
"epoch": 2.939746835443038,
"grad_norm": 0.3849742265295548,
"learning_rate": 1.731212071922883e-05,
"loss": 0.3652,
"step": 724
},
{
"epoch": 2.9437974683544303,
"grad_norm": 0.40887517963441394,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.3591,
"step": 725
},
{
"epoch": 2.947848101265823,
"grad_norm": 0.39220754084126147,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.3516,
"step": 726
},
{
"epoch": 2.951898734177215,
"grad_norm": 0.41217843010157806,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.3657,
"step": 727
},
{
"epoch": 2.9559493670886074,
"grad_norm": 0.3772143885894634,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.4001,
"step": 728
},
{
"epoch": 2.96,
"grad_norm": 0.3831326664846604,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.3876,
"step": 729
},
{
"epoch": 2.9640506329113925,
"grad_norm": 0.4285310955075387,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.3724,
"step": 730
},
{
"epoch": 2.968101265822785,
"grad_norm": 0.38721629331957647,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.3781,
"step": 731
},
{
"epoch": 2.972151898734177,
"grad_norm": 0.3864928134472797,
"learning_rate": 1.686290200071915e-05,
"loss": 0.3777,
"step": 732
},
{
"epoch": 2.9762025316455696,
"grad_norm": 0.38979502240280545,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.3842,
"step": 733
},
{
"epoch": 2.980253164556962,
"grad_norm": 0.39979176511570574,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.3711,
"step": 734
},
{
"epoch": 2.984303797468354,
"grad_norm": 0.3730073375706051,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.3679,
"step": 735
},
{
"epoch": 2.9883544303797467,
"grad_norm": 0.38705054036823866,
"learning_rate": 1.663888455395778e-05,
"loss": 0.3704,
"step": 736
},
{
"epoch": 2.992405063291139,
"grad_norm": 0.38611799081772286,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.3671,
"step": 737
},
{
"epoch": 2.9964556962025317,
"grad_norm": 0.38796509798672774,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.3718,
"step": 738
},
{
"epoch": 3.0005063291139242,
"grad_norm": 0.37948874827646206,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.3772,
"step": 739
},
{
"epoch": 3.0045569620253163,
"grad_norm": 0.596791084204251,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.2749,
"step": 740
},
{
"epoch": 3.008607594936709,
"grad_norm": 0.3743619798467806,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.275,
"step": 741
},
{
"epoch": 3.0126582278481013,
"grad_norm": 0.5669373737728506,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.2584,
"step": 742
},
{
"epoch": 3.016708860759494,
"grad_norm": 0.6758891561553594,
"learning_rate": 1.624791353909428e-05,
"loss": 0.2682,
"step": 743
},
{
"epoch": 3.020759493670886,
"grad_norm": 0.4112812875826829,
"learning_rate": 1.619217781670663e-05,
"loss": 0.2685,
"step": 744
},
{
"epoch": 3.0248101265822784,
"grad_norm": 0.5582814357935626,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.2529,
"step": 745
},
{
"epoch": 3.028860759493671,
"grad_norm": 0.5839576198760028,
"learning_rate": 1.608079882360134e-05,
"loss": 0.2782,
"step": 746
},
{
"epoch": 3.0329113924050635,
"grad_norm": 0.4113209083584426,
"learning_rate": 1.60251564499148e-05,
"loss": 0.2682,
"step": 747
},
{
"epoch": 3.0369620253164555,
"grad_norm": 0.4746085930303214,
"learning_rate": 1.596954608907176e-05,
"loss": 0.2738,
"step": 748
},
{
"epoch": 3.041012658227848,
"grad_norm": 0.5441305964014203,
"learning_rate": 1.591396818895043e-05,
"loss": 0.2675,
"step": 749
},
{
"epoch": 3.0450632911392406,
"grad_norm": 0.4139902728823283,
"learning_rate": 1.585842319716759e-05,
"loss": 0.2831,
"step": 750
},
{
"epoch": 3.049113924050633,
"grad_norm": 0.46714488110479824,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.2559,
"step": 751
},
{
"epoch": 3.053164556962025,
"grad_norm": 0.4610049016148163,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.2686,
"step": 752
},
{
"epoch": 3.0572151898734177,
"grad_norm": 0.41698704893081256,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.2754,
"step": 753
},
{
"epoch": 3.06126582278481,
"grad_norm": 0.40250335198182124,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.2741,
"step": 754
},
{
"epoch": 3.0653164556962027,
"grad_norm": 0.4297700299619123,
"learning_rate": 1.558120751116291e-05,
"loss": 0.2636,
"step": 755
},
{
"epoch": 3.0693670886075948,
"grad_norm": 0.3790307629643322,
"learning_rate": 1.552586935426876e-05,
"loss": 0.2629,
"step": 756
},
{
"epoch": 3.0734177215189873,
"grad_norm": 0.4007262827931162,
"learning_rate": 1.547056723140774e-05,
"loss": 0.2488,
"step": 757
},
{
"epoch": 3.07746835443038,
"grad_norm": 0.3724225042477577,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.2667,
"step": 758
},
{
"epoch": 3.0815189873417723,
"grad_norm": 0.3576009291662094,
"learning_rate": 1.536007286907411e-05,
"loss": 0.2429,
"step": 759
},
{
"epoch": 3.0855696202531644,
"grad_norm": 1.3270443313498466,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.2908,
"step": 760
},
{
"epoch": 3.089620253164557,
"grad_norm": 0.3973664928458626,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.2667,
"step": 761
},
{
"epoch": 3.0936708860759494,
"grad_norm": 0.36113786435375406,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.243,
"step": 762
},
{
"epoch": 3.097721518987342,
"grad_norm": 0.35293452717245266,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.2569,
"step": 763
},
{
"epoch": 3.101772151898734,
"grad_norm": 0.38725685685440875,
"learning_rate": 1.508449870007656e-05,
"loss": 0.2606,
"step": 764
},
{
"epoch": 3.1058227848101265,
"grad_norm": 0.37101523545072473,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.2398,
"step": 765
},
{
"epoch": 3.109873417721519,
"grad_norm": 0.36339484109530806,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.243,
"step": 766
},
{
"epoch": 3.1139240506329116,
"grad_norm": 0.40935128420094374,
"learning_rate": 1.4919625713405e-05,
"loss": 0.2768,
"step": 767
},
{
"epoch": 3.1179746835443036,
"grad_norm": 0.34270105050856187,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.2924,
"step": 768
},
{
"epoch": 3.122025316455696,
"grad_norm": 0.37208155715100555,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.2803,
"step": 769
},
{
"epoch": 3.1260759493670887,
"grad_norm": 0.3742273536543627,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.2486,
"step": 770
},
{
"epoch": 3.130126582278481,
"grad_norm": 0.34419303508524507,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.2661,
"step": 771
},
{
"epoch": 3.1341772151898732,
"grad_norm": 0.35441114620378494,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.2514,
"step": 772
},
{
"epoch": 3.1382278481012658,
"grad_norm": 0.3719625090077885,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.264,
"step": 773
},
{
"epoch": 3.1422784810126583,
"grad_norm": 0.33440958360879275,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.268,
"step": 774
},
{
"epoch": 3.146329113924051,
"grad_norm": 0.36378541262525144,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.2396,
"step": 775
},
{
"epoch": 3.150379746835443,
"grad_norm": 0.34303575056612884,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.2618,
"step": 776
},
{
"epoch": 3.1544303797468354,
"grad_norm": 0.33745475019203186,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.2534,
"step": 777
},
{
"epoch": 3.158481012658228,
"grad_norm": 0.3360773666946644,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.271,
"step": 778
},
{
"epoch": 3.1625316455696204,
"grad_norm": 0.35439914138463163,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.2699,
"step": 779
},
{
"epoch": 3.1665822784810125,
"grad_norm": 0.3394071725567932,
"learning_rate": 1.420958430225303e-05,
"loss": 0.2474,
"step": 780
},
{
"epoch": 3.170632911392405,
"grad_norm": 0.35032742427151126,
"learning_rate": 1.415527989393985e-05,
"loss": 0.2655,
"step": 781
},
{
"epoch": 3.1746835443037975,
"grad_norm": 0.32854896275880807,
"learning_rate": 1.410102255819891e-05,
"loss": 0.2795,
"step": 782
},
{
"epoch": 3.17873417721519,
"grad_norm": 0.33987120407592275,
"learning_rate": 1.404681273201131e-05,
"loss": 0.2676,
"step": 783
},
{
"epoch": 3.182784810126582,
"grad_norm": 0.347708890164154,
"learning_rate": 1.399265085197556e-05,
"loss": 0.2635,
"step": 784
},
{
"epoch": 3.1868354430379746,
"grad_norm": 0.3428945202025847,
"learning_rate": 1.393853735430398e-05,
"loss": 0.2585,
"step": 785
},
{
"epoch": 3.190886075949367,
"grad_norm": 0.33905493766893297,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.267,
"step": 786
},
{
"epoch": 3.1949367088607596,
"grad_norm": 0.4598968270116413,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.2739,
"step": 787
},
{
"epoch": 3.1989873417721517,
"grad_norm": 0.35877644821741234,
"learning_rate": 1.377649151173163e-05,
"loss": 0.2553,
"step": 788
},
{
"epoch": 3.2030379746835442,
"grad_norm": 0.3320817894243063,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.251,
"step": 789
},
{
"epoch": 3.2070886075949367,
"grad_norm": 0.35152550235910696,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.2641,
"step": 790
},
{
"epoch": 3.2111392405063293,
"grad_norm": 0.3521510610980603,
"learning_rate": 1.361489677627324e-05,
"loss": 0.2902,
"step": 791
},
{
"epoch": 3.2151898734177213,
"grad_norm": 0.31889014123417664,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.277,
"step": 792
},
{
"epoch": 3.219240506329114,
"grad_norm": 0.35637823460930973,
"learning_rate": 1.350742335331241e-05,
"loss": 0.2725,
"step": 793
},
{
"epoch": 3.2232911392405064,
"grad_norm": 0.35794649484589247,
"learning_rate": 1.345376486102198e-05,
"loss": 0.259,
"step": 794
},
{
"epoch": 3.227341772151899,
"grad_norm": 0.344872623380073,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.2506,
"step": 795
},
{
"epoch": 3.231392405063291,
"grad_norm": 0.3396139615584749,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.2511,
"step": 796
},
{
"epoch": 3.2354430379746835,
"grad_norm": 0.3455099205873177,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.2521,
"step": 797
},
{
"epoch": 3.239493670886076,
"grad_norm": 0.3354503393861874,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.2624,
"step": 798
},
{
"epoch": 3.2435443037974685,
"grad_norm": 0.3485638615335572,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.2509,
"step": 799
},
{
"epoch": 3.2475949367088606,
"grad_norm": 0.335781505732869,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.2553,
"step": 800
},
{
"epoch": 3.251645569620253,
"grad_norm": 0.3338145595130177,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.2657,
"step": 801
},
{
"epoch": 3.2556962025316456,
"grad_norm": 0.34741127195352417,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.2451,
"step": 802
},
{
"epoch": 3.259746835443038,
"grad_norm": 0.3285673822019449,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.2752,
"step": 803
},
{
"epoch": 3.26379746835443,
"grad_norm": 0.33388344256167163,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.2381,
"step": 804
},
{
"epoch": 3.2678481012658227,
"grad_norm": 0.33142962343022947,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.2349,
"step": 805
},
{
"epoch": 3.271898734177215,
"grad_norm": 0.3190754524680187,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.2821,
"step": 806
},
{
"epoch": 3.2759493670886077,
"grad_norm": 0.32096090705479924,
"learning_rate": 1.276115892225764e-05,
"loss": 0.2785,
"step": 807
},
{
"epoch": 3.2800000000000002,
"grad_norm": 0.3280607006667989,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.2513,
"step": 808
},
{
"epoch": 3.2840506329113923,
"grad_norm": 0.35596552114047153,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.277,
"step": 809
},
{
"epoch": 3.288101265822785,
"grad_norm": 0.9192821929072098,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.2694,
"step": 810
},
{
"epoch": 3.2921518987341774,
"grad_norm": 0.36768239311450734,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.2695,
"step": 811
},
{
"epoch": 3.2962025316455694,
"grad_norm": 0.3247528593154617,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.2707,
"step": 812
},
{
"epoch": 3.300253164556962,
"grad_norm": 0.35341545440612115,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.2349,
"step": 813
},
{
"epoch": 3.3043037974683545,
"grad_norm": 0.3324977073379,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.2457,
"step": 814
},
{
"epoch": 3.308354430379747,
"grad_norm": 0.3444773402103227,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.2381,
"step": 815
},
{
"epoch": 3.3124050632911395,
"grad_norm": 0.3265785857952916,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.2604,
"step": 816
},
{
"epoch": 3.3164556962025316,
"grad_norm": 0.3520298512170093,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.2551,
"step": 817
},
{
"epoch": 3.320506329113924,
"grad_norm": 0.33967816740067275,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.2635,
"step": 818
},
{
"epoch": 3.3245569620253166,
"grad_norm": 0.3469211593077021,
"learning_rate": 1.213055323844297e-05,
"loss": 0.2406,
"step": 819
},
{
"epoch": 3.3286075949367087,
"grad_norm": 0.36409982342407027,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.2965,
"step": 820
},
{
"epoch": 3.332658227848101,
"grad_norm": 0.35445620973485914,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.2779,
"step": 821
},
{
"epoch": 3.3367088607594937,
"grad_norm": 0.3467142083036966,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.2658,
"step": 822
},
{
"epoch": 3.340759493670886,
"grad_norm": 0.3225049438747772,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.2956,
"step": 823
},
{
"epoch": 3.3448101265822787,
"grad_norm": 0.33111998498364537,
"learning_rate": 1.187045262895488e-05,
"loss": 0.2789,
"step": 824
},
{
"epoch": 3.348860759493671,
"grad_norm": 0.3422646532384314,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.2603,
"step": 825
},
{
"epoch": 3.3529113924050633,
"grad_norm": 0.35182123732980874,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.2648,
"step": 826
},
{
"epoch": 3.356962025316456,
"grad_norm": 0.3404137775972387,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.2441,
"step": 827
},
{
"epoch": 3.361012658227848,
"grad_norm": 0.322826275731296,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.2797,
"step": 828
},
{
"epoch": 3.3650632911392404,
"grad_norm": 0.32936390271439575,
"learning_rate": 1.161198884952377e-05,
"loss": 0.2516,
"step": 829
},
{
"epoch": 3.369113924050633,
"grad_norm": 0.3203961294386798,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.2557,
"step": 830
},
{
"epoch": 3.3731645569620254,
"grad_norm": 0.3418066517199913,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.2642,
"step": 831
},
{
"epoch": 3.377215189873418,
"grad_norm": 0.3429738693246749,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.2606,
"step": 832
},
{
"epoch": 3.38126582278481,
"grad_norm": 0.3353771093082826,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.2814,
"step": 833
},
{
"epoch": 3.3853164556962025,
"grad_norm": 0.33381460248100353,
"learning_rate": 1.135521394010506e-05,
"loss": 0.2544,
"step": 834
},
{
"epoch": 3.389367088607595,
"grad_norm": 0.33260110677968546,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.255,
"step": 835
},
{
"epoch": 3.393417721518987,
"grad_norm": 0.3440022331910399,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.2536,
"step": 836
},
{
"epoch": 3.3974683544303796,
"grad_norm": 0.36983894279927965,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.2749,
"step": 837
},
{
"epoch": 3.401518987341772,
"grad_norm": 0.32157201310196054,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.2822,
"step": 838
},
{
"epoch": 3.4055696202531647,
"grad_norm": 0.3609607935290914,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.2402,
"step": 839
},
{
"epoch": 3.409620253164557,
"grad_norm": 0.33477461489221555,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.2589,
"step": 840
},
{
"epoch": 3.4136708860759493,
"grad_norm": 0.34080248593658424,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.2486,
"step": 841
},
{
"epoch": 3.4177215189873418,
"grad_norm": 0.3309446565199687,
"learning_rate": 1.0948015846744e-05,
"loss": 0.2688,
"step": 842
},
{
"epoch": 3.4217721518987343,
"grad_norm": 0.3143740414721356,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.2744,
"step": 843
},
{
"epoch": 3.4258227848101264,
"grad_norm": 0.3314178034728436,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.2546,
"step": 844
},
{
"epoch": 3.429873417721519,
"grad_norm": 0.3287977614860553,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.2458,
"step": 845
},
{
"epoch": 3.4339240506329114,
"grad_norm": 0.3355423312588682,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.2739,
"step": 846
},
{
"epoch": 3.437974683544304,
"grad_norm": 0.33354755684911136,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.2607,
"step": 847
},
{
"epoch": 3.4420253164556964,
"grad_norm": 0.33889693383208186,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.241,
"step": 848
},
{
"epoch": 3.4460759493670885,
"grad_norm": 0.33233405465158594,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.243,
"step": 849
},
{
"epoch": 3.450126582278481,
"grad_norm": 0.3313170908696168,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.2615,
"step": 850
},
{
"epoch": 3.4541772151898735,
"grad_norm": 0.3505304846832336,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.2358,
"step": 851
},
{
"epoch": 3.4582278481012656,
"grad_norm": 0.31732070786780747,
"learning_rate": 1.044560362975474e-05,
"loss": 0.2675,
"step": 852
},
{
"epoch": 3.462278481012658,
"grad_norm": 0.3310540127759703,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.2547,
"step": 853
},
{
"epoch": 3.4663291139240506,
"grad_norm": 0.3639672971964621,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.2873,
"step": 854
},
{
"epoch": 3.470379746835443,
"grad_norm": 0.3362387765306904,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.2638,
"step": 855
},
{
"epoch": 3.4744303797468357,
"grad_norm": 0.3207178162833066,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.2733,
"step": 856
},
{
"epoch": 3.4784810126582277,
"grad_norm": 0.3376652406664714,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.2526,
"step": 857
},
{
"epoch": 3.4825316455696202,
"grad_norm": 0.35513470677101816,
"learning_rate": 1.014782400334433e-05,
"loss": 0.2546,
"step": 858
},
{
"epoch": 3.4865822784810128,
"grad_norm": 0.3425472724515098,
"learning_rate": 1.009846945771296e-05,
"loss": 0.264,
"step": 859
},
{
"epoch": 3.490632911392405,
"grad_norm": 0.32566631763620035,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.2339,
"step": 860
},
{
"epoch": 3.4946835443037974,
"grad_norm": 0.3336027231848593,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.2369,
"step": 861
},
{
"epoch": 3.49873417721519,
"grad_norm": 0.3415437067338059,
"learning_rate": 9.950885880977891e-06,
"loss": 0.2472,
"step": 862
},
{
"epoch": 3.5027848101265824,
"grad_norm": 0.3415662529897534,
"learning_rate": 9.901852696139382e-06,
"loss": 0.2847,
"step": 863
},
{
"epoch": 3.506835443037975,
"grad_norm": 0.3400580440728001,
"learning_rate": 9.852900840391027e-06,
"loss": 0.2897,
"step": 864
},
{
"epoch": 3.510886075949367,
"grad_norm": 0.3211091122594664,
"learning_rate": 9.804030707984313e-06,
"loss": 0.2555,
"step": 865
},
{
"epoch": 3.5149367088607595,
"grad_norm": 0.32669894847812564,
"learning_rate": 9.755242692512599e-06,
"loss": 0.2479,
"step": 866
},
{
"epoch": 3.518987341772152,
"grad_norm": 0.32826980976882586,
"learning_rate": 9.70653718690782e-06,
"loss": 0.2808,
"step": 867
},
{
"epoch": 3.523037974683544,
"grad_norm": 0.32101754282768347,
"learning_rate": 9.657914583437454e-06,
"loss": 0.2276,
"step": 868
},
{
"epoch": 3.5270886075949366,
"grad_norm": 0.3222837954142063,
"learning_rate": 9.609375273701246e-06,
"loss": 0.2601,
"step": 869
},
{
"epoch": 3.531139240506329,
"grad_norm": 0.31807716383970397,
"learning_rate": 9.560919648628133e-06,
"loss": 0.2401,
"step": 870
},
{
"epoch": 3.5351898734177216,
"grad_norm": 0.3463343759520837,
"learning_rate": 9.512548098473047e-06,
"loss": 0.2466,
"step": 871
},
{
"epoch": 3.539240506329114,
"grad_norm": 0.3263269928201352,
"learning_rate": 9.464261012813825e-06,
"loss": 0.2763,
"step": 872
},
{
"epoch": 3.543291139240506,
"grad_norm": 7.034552630776023,
"learning_rate": 9.416058780547987e-06,
"loss": 0.405,
"step": 873
},
{
"epoch": 3.5473417721518987,
"grad_norm": 0.3295153865359594,
"learning_rate": 9.367941789889714e-06,
"loss": 0.2442,
"step": 874
},
{
"epoch": 3.5513924050632912,
"grad_norm": 0.3418933618238107,
"learning_rate": 9.319910428366607e-06,
"loss": 0.265,
"step": 875
},
{
"epoch": 3.5554430379746833,
"grad_norm": 0.3328151183992317,
"learning_rate": 9.271965082816667e-06,
"loss": 0.248,
"step": 876
},
{
"epoch": 3.559493670886076,
"grad_norm": 0.3306094626190365,
"learning_rate": 9.224106139385111e-06,
"loss": 0.2418,
"step": 877
},
{
"epoch": 3.5635443037974683,
"grad_norm": 0.32806962845348664,
"learning_rate": 9.176333983521291e-06,
"loss": 0.2603,
"step": 878
},
{
"epoch": 3.567594936708861,
"grad_norm": 0.3301275003671268,
"learning_rate": 9.12864899997558e-06,
"loss": 0.2409,
"step": 879
},
{
"epoch": 3.5716455696202534,
"grad_norm": 0.329848726269475,
"learning_rate": 9.08105157279628e-06,
"loss": 0.256,
"step": 880
},
{
"epoch": 3.5756962025316454,
"grad_norm": 0.3395409258655622,
"learning_rate": 9.03354208532653e-06,
"loss": 0.2775,
"step": 881
},
{
"epoch": 3.579746835443038,
"grad_norm": 0.33217973977294996,
"learning_rate": 8.986120920201205e-06,
"loss": 0.2466,
"step": 882
},
{
"epoch": 3.5837974683544305,
"grad_norm": 0.32309465516487984,
"learning_rate": 8.938788459343852e-06,
"loss": 0.2562,
"step": 883
},
{
"epoch": 3.5878481012658225,
"grad_norm": 0.3036419039102078,
"learning_rate": 8.8915450839636e-06,
"loss": 0.2469,
"step": 884
},
{
"epoch": 3.591898734177215,
"grad_norm": 0.31160634447617674,
"learning_rate": 8.844391174552116e-06,
"loss": 0.2474,
"step": 885
},
{
"epoch": 3.5959493670886076,
"grad_norm": 0.33212298440016336,
"learning_rate": 8.797327110880479e-06,
"loss": 0.2795,
"step": 886
},
{
"epoch": 3.6,
"grad_norm": 0.3221819310821201,
"learning_rate": 8.750353271996206e-06,
"loss": 0.2505,
"step": 887
},
{
"epoch": 3.6040506329113926,
"grad_norm": 0.34609774776592417,
"learning_rate": 8.703470036220132e-06,
"loss": 0.2384,
"step": 888
},
{
"epoch": 3.6081012658227847,
"grad_norm": 0.3266750966100086,
"learning_rate": 8.656677781143394e-06,
"loss": 0.2479,
"step": 889
},
{
"epoch": 3.612151898734177,
"grad_norm": 0.32239102147594784,
"learning_rate": 8.609976883624377e-06,
"loss": 0.2444,
"step": 890
},
{
"epoch": 3.6162025316455697,
"grad_norm": 0.31854634569233964,
"learning_rate": 8.563367719785698e-06,
"loss": 0.2577,
"step": 891
},
{
"epoch": 3.620253164556962,
"grad_norm": 0.3242615893314945,
"learning_rate": 8.516850665011138e-06,
"loss": 0.2468,
"step": 892
},
{
"epoch": 3.6243037974683543,
"grad_norm": 0.3322918424139079,
"learning_rate": 8.47042609394269e-06,
"loss": 0.2737,
"step": 893
},
{
"epoch": 3.628354430379747,
"grad_norm": 0.3166888744758024,
"learning_rate": 8.424094380477432e-06,
"loss": 0.2607,
"step": 894
},
{
"epoch": 3.6324050632911393,
"grad_norm": 0.31179191146796253,
"learning_rate": 8.37785589776465e-06,
"loss": 0.2693,
"step": 895
},
{
"epoch": 3.636455696202532,
"grad_norm": 0.3291715961451965,
"learning_rate": 8.331711018202694e-06,
"loss": 0.2604,
"step": 896
},
{
"epoch": 3.640506329113924,
"grad_norm": 0.3346112267939724,
"learning_rate": 8.285660113436104e-06,
"loss": 0.2544,
"step": 897
},
{
"epoch": 3.6445569620253164,
"grad_norm": 0.33824852742384787,
"learning_rate": 8.239703554352527e-06,
"loss": 0.2615,
"step": 898
},
{
"epoch": 3.648607594936709,
"grad_norm": 0.3290725811707173,
"learning_rate": 8.193841711079775e-06,
"loss": 0.2687,
"step": 899
},
{
"epoch": 3.652658227848101,
"grad_norm": 0.33961955653639647,
"learning_rate": 8.148074952982828e-06,
"loss": 0.2717,
"step": 900
},
{
"epoch": 3.6567088607594935,
"grad_norm": 0.3115553425491023,
"learning_rate": 8.102403648660859e-06,
"loss": 0.2724,
"step": 901
},
{
"epoch": 3.660759493670886,
"grad_norm": 0.3036572787445494,
"learning_rate": 8.056828165944282e-06,
"loss": 0.2709,
"step": 902
},
{
"epoch": 3.6648101265822786,
"grad_norm": 0.328819526216983,
"learning_rate": 8.011348871891762e-06,
"loss": 0.2665,
"step": 903
},
{
"epoch": 3.668860759493671,
"grad_norm": 0.32333906708281607,
"learning_rate": 7.965966132787287e-06,
"loss": 0.2449,
"step": 904
},
{
"epoch": 3.672911392405063,
"grad_norm": 0.3352837269176696,
"learning_rate": 7.920680314137189e-06,
"loss": 0.2705,
"step": 905
},
{
"epoch": 3.6769620253164557,
"grad_norm": 0.31112564557052813,
"learning_rate": 7.875491780667246e-06,
"loss": 0.2593,
"step": 906
},
{
"epoch": 3.681012658227848,
"grad_norm": 0.3164141705824694,
"learning_rate": 7.830400896319667e-06,
"loss": 0.2744,
"step": 907
},
{
"epoch": 3.6850632911392402,
"grad_norm": 0.3191144278439063,
"learning_rate": 7.785408024250259e-06,
"loss": 0.2722,
"step": 908
},
{
"epoch": 3.6891139240506328,
"grad_norm": 0.32986802592082554,
"learning_rate": 7.74051352682542e-06,
"loss": 0.2406,
"step": 909
},
{
"epoch": 3.6931645569620253,
"grad_norm": 0.3184330066566083,
"learning_rate": 7.695717765619257e-06,
"loss": 0.2715,
"step": 910
},
{
"epoch": 3.697215189873418,
"grad_norm": 0.30090760551445567,
"learning_rate": 7.651021101410673e-06,
"loss": 0.28,
"step": 911
},
{
"epoch": 3.7012658227848103,
"grad_norm": 0.30533829699862847,
"learning_rate": 7.606423894180464e-06,
"loss": 0.2798,
"step": 912
},
{
"epoch": 3.7053164556962024,
"grad_norm": 0.31699711436663564,
"learning_rate": 7.56192650310839e-06,
"loss": 0.2754,
"step": 913
},
{
"epoch": 3.709367088607595,
"grad_norm": 0.32482823891923296,
"learning_rate": 7.517529286570349e-06,
"loss": 0.254,
"step": 914
},
{
"epoch": 3.7134177215189874,
"grad_norm": 0.31427595748544096,
"learning_rate": 7.473232602135387e-06,
"loss": 0.253,
"step": 915
},
{
"epoch": 3.7174683544303795,
"grad_norm": 0.32045818822986416,
"learning_rate": 7.429036806562935e-06,
"loss": 0.2606,
"step": 916
},
{
"epoch": 3.721518987341772,
"grad_norm": 0.3303491268711638,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.2563,
"step": 917
},
{
"epoch": 3.7255696202531645,
"grad_norm": 0.33649458908875574,
"learning_rate": 7.340949304977567e-06,
"loss": 0.2714,
"step": 918
},
{
"epoch": 3.729620253164557,
"grad_norm": 0.3142733069089562,
"learning_rate": 7.297058308409282e-06,
"loss": 0.2639,
"step": 919
},
{
"epoch": 3.7336708860759495,
"grad_norm": 0.33326018461813206,
"learning_rate": 7.25326961958704e-06,
"loss": 0.2662,
"step": 920
},
{
"epoch": 3.7377215189873416,
"grad_norm": 0.30958204640505665,
"learning_rate": 7.209583591178921e-06,
"loss": 0.2721,
"step": 921
},
{
"epoch": 3.741772151898734,
"grad_norm": 0.3224744714892965,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.2346,
"step": 922
},
{
"epoch": 3.7458227848101266,
"grad_norm": 0.32676212409422484,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.2533,
"step": 923
},
{
"epoch": 3.7498734177215187,
"grad_norm": 0.32665020044605836,
"learning_rate": 7.079144982700909e-06,
"loss": 0.2515,
"step": 924
},
{
"epoch": 3.7539240506329112,
"grad_norm": 0.3128878165016944,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.2828,
"step": 925
},
{
"epoch": 3.7579746835443038,
"grad_norm": 0.312742438714276,
"learning_rate": 6.99270564069757e-06,
"loss": 0.2415,
"step": 926
},
{
"epoch": 3.7620253164556963,
"grad_norm": 0.33888126120169765,
"learning_rate": 6.949642934304375e-06,
"loss": 0.2622,
"step": 927
},
{
"epoch": 3.766075949367089,
"grad_norm": 0.3341788676777151,
"learning_rate": 6.906685333692871e-06,
"loss": 0.2522,
"step": 928
},
{
"epoch": 3.770126582278481,
"grad_norm": 0.32306480628745643,
"learning_rate": 6.86383318483769e-06,
"loss": 0.2538,
"step": 929
},
{
"epoch": 3.7741772151898734,
"grad_norm": 0.35206994963337995,
"learning_rate": 6.821086832864139e-06,
"loss": 0.2522,
"step": 930
},
{
"epoch": 3.778227848101266,
"grad_norm": 0.3325883610549248,
"learning_rate": 6.77844662204546e-06,
"loss": 0.2678,
"step": 931
},
{
"epoch": 3.782278481012658,
"grad_norm": 0.3247730800981493,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.2559,
"step": 932
},
{
"epoch": 3.7863291139240505,
"grad_norm": 0.31778710489835665,
"learning_rate": 6.693485996688695e-06,
"loss": 0.2642,
"step": 933
},
{
"epoch": 3.790379746835443,
"grad_norm": 0.31700293089972714,
"learning_rate": 6.651166266411801e-06,
"loss": 0.2557,
"step": 934
},
{
"epoch": 3.7944303797468355,
"grad_norm": 0.335717004806811,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.2755,
"step": 935
},
{
"epoch": 3.798481012658228,
"grad_norm": 0.3296377953470558,
"learning_rate": 6.566849674844711e-06,
"loss": 0.2405,
"step": 936
},
{
"epoch": 3.80253164556962,
"grad_norm": 0.3165897833899535,
"learning_rate": 6.524853492628747e-06,
"loss": 0.2364,
"step": 937
},
{
"epoch": 3.8065822784810126,
"grad_norm": 0.324184524462777,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.2644,
"step": 938
},
{
"epoch": 3.810632911392405,
"grad_norm": 0.3215291068279111,
"learning_rate": 6.441187046486648e-06,
"loss": 0.244,
"step": 939
},
{
"epoch": 3.814683544303797,
"grad_norm": 0.3196898873599333,
"learning_rate": 6.399517456398567e-06,
"loss": 0.2422,
"step": 940
},
{
"epoch": 3.81873417721519,
"grad_norm": 0.3248229612444475,
"learning_rate": 6.357957402727164e-06,
"loss": 0.2479,
"step": 941
},
{
"epoch": 3.8227848101265822,
"grad_norm": 0.3179200740314193,
"learning_rate": 6.316507220191395e-06,
"loss": 0.2592,
"step": 942
},
{
"epoch": 3.8268354430379747,
"grad_norm": 0.31487350547600124,
"learning_rate": 6.275167242625331e-06,
"loss": 0.2515,
"step": 943
},
{
"epoch": 3.8308860759493673,
"grad_norm": 0.31793330076884496,
"learning_rate": 6.233937802975471e-06,
"loss": 0.2594,
"step": 944
},
{
"epoch": 3.8349367088607593,
"grad_norm": 0.3139751020423621,
"learning_rate": 6.192819233298046e-06,
"loss": 0.2705,
"step": 945
},
{
"epoch": 3.838987341772152,
"grad_norm": 0.31683723918513695,
"learning_rate": 6.151811864756383e-06,
"loss": 0.2649,
"step": 946
},
{
"epoch": 3.8430379746835444,
"grad_norm": 0.309998678656485,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.2342,
"step": 947
},
{
"epoch": 3.847088607594937,
"grad_norm": 0.3241766729835554,
"learning_rate": 6.070132051252868e-06,
"loss": 0.265,
"step": 948
},
{
"epoch": 3.8511392405063294,
"grad_norm": 0.3185372798053361,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.2619,
"step": 949
},
{
"epoch": 3.8551898734177215,
"grad_norm": 0.30812637821772376,
"learning_rate": 5.988900993811575e-06,
"loss": 0.2808,
"step": 950
},
{
"epoch": 3.859240506329114,
"grad_norm": 0.3158308701665354,
"learning_rate": 5.948454566959363e-06,
"loss": 0.2531,
"step": 951
},
{
"epoch": 3.8632911392405065,
"grad_norm": 0.318019991207333,
"learning_rate": 5.908121309322328e-06,
"loss": 0.2519,
"step": 952
},
{
"epoch": 3.8673417721518986,
"grad_norm": 0.31150605253875063,
"learning_rate": 5.867901545738976e-06,
"loss": 0.2517,
"step": 953
},
{
"epoch": 3.871392405063291,
"grad_norm": 0.3261892215969753,
"learning_rate": 5.827795600133774e-06,
"loss": 0.2687,
"step": 954
},
{
"epoch": 3.8754430379746836,
"grad_norm": 0.31779357618329623,
"learning_rate": 5.787803795514466e-06,
"loss": 0.261,
"step": 955
},
{
"epoch": 3.879493670886076,
"grad_norm": 0.3231061100526691,
"learning_rate": 5.747926453969576e-06,
"loss": 0.2542,
"step": 956
},
{
"epoch": 3.8835443037974686,
"grad_norm": 0.3229436760909751,
"learning_rate": 5.708163896665708e-06,
"loss": 0.2648,
"step": 957
},
{
"epoch": 3.8875949367088607,
"grad_norm": 0.31552205593922894,
"learning_rate": 5.668516443845047e-06,
"loss": 0.2771,
"step": 958
},
{
"epoch": 3.891645569620253,
"grad_norm": 0.31422354392116614,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.2554,
"step": 959
},
{
"epoch": 3.8956962025316457,
"grad_norm": 0.31696782828662334,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.2726,
"step": 960
},
{
"epoch": 3.899746835443038,
"grad_norm": 0.31732079090620985,
"learning_rate": 5.550267900783019e-06,
"loss": 0.255,
"step": 961
},
{
"epoch": 3.9037974683544303,
"grad_norm": 0.3136492167991513,
"learning_rate": 5.511084049737623e-06,
"loss": 0.2592,
"step": 962
},
{
"epoch": 3.907848101265823,
"grad_norm": 0.30778638520712104,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.2348,
"step": 963
},
{
"epoch": 3.9118987341772153,
"grad_norm": 0.3379906949877354,
"learning_rate": 5.433066737499948e-06,
"loss": 0.2481,
"step": 964
},
{
"epoch": 3.915949367088608,
"grad_norm": 0.32021485322534854,
"learning_rate": 5.394233904648376e-06,
"loss": 0.2492,
"step": 965
},
{
"epoch": 3.92,
"grad_norm": 0.30948397177020814,
"learning_rate": 5.355518704628997e-06,
"loss": 0.256,
"step": 966
},
{
"epoch": 3.9240506329113924,
"grad_norm": 0.3247904016291046,
"learning_rate": 5.316921449248731e-06,
"loss": 0.2667,
"step": 967
},
{
"epoch": 3.928101265822785,
"grad_norm": 0.3210654392478939,
"learning_rate": 5.278442449364538e-06,
"loss": 0.2571,
"step": 968
},
{
"epoch": 3.932151898734177,
"grad_norm": 0.31931131942559504,
"learning_rate": 5.240082014881016e-06,
"loss": 0.2349,
"step": 969
},
{
"epoch": 3.9362025316455695,
"grad_norm": 0.3155224256183819,
"learning_rate": 5.201840454747822e-06,
"loss": 0.2798,
"step": 970
},
{
"epoch": 3.940253164556962,
"grad_norm": 0.31924396383671494,
"learning_rate": 5.163718076957223e-06,
"loss": 0.2438,
"step": 971
},
{
"epoch": 3.9443037974683546,
"grad_norm": 0.3295995565140276,
"learning_rate": 5.125715188541609e-06,
"loss": 0.2418,
"step": 972
},
{
"epoch": 3.948354430379747,
"grad_norm": 0.3212526545070598,
"learning_rate": 5.087832095571021e-06,
"loss": 0.2817,
"step": 973
},
{
"epoch": 3.952405063291139,
"grad_norm": 0.31368977543121995,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.2433,
"step": 974
},
{
"epoch": 3.9564556962025317,
"grad_norm": 0.32020514956326707,
"learning_rate": 5.01242651541854e-06,
"loss": 0.2624,
"step": 975
},
{
"epoch": 3.960506329113924,
"grad_norm": 0.32527273560284603,
"learning_rate": 4.974904635542815e-06,
"loss": 0.266,
"step": 976
},
{
"epoch": 3.9645569620253163,
"grad_norm": 0.31623854751394964,
"learning_rate": 4.937503765719582e-06,
"loss": 0.2485,
"step": 977
},
{
"epoch": 3.968607594936709,
"grad_norm": 0.32830916613393374,
"learning_rate": 4.900224207170299e-06,
"loss": 0.2518,
"step": 978
},
{
"epoch": 3.9726582278481013,
"grad_norm": 0.33429318038902406,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.2708,
"step": 979
},
{
"epoch": 3.976708860759494,
"grad_norm": 0.3352124050928428,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.2436,
"step": 980
},
{
"epoch": 3.9807594936708863,
"grad_norm": 0.30952264080535413,
"learning_rate": 4.789116396710924e-06,
"loss": 0.2598,
"step": 981
},
{
"epoch": 3.9848101265822784,
"grad_norm": 0.31679038779075197,
"learning_rate": 4.752325075895368e-06,
"loss": 0.2472,
"step": 982
},
{
"epoch": 3.988860759493671,
"grad_norm": 0.3258093025382478,
"learning_rate": 4.715656557757473e-06,
"loss": 0.2535,
"step": 983
},
{
"epoch": 3.9929113924050634,
"grad_norm": 0.31246740241386456,
"learning_rate": 4.679111137620442e-06,
"loss": 0.2712,
"step": 984
},
{
"epoch": 3.9969620253164555,
"grad_norm": 0.31391694986226065,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.2712,
"step": 985
},
{
"epoch": 4.0010126582278485,
"grad_norm": 0.3047249200476688,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.2612,
"step": 986
},
{
"epoch": 4.0050632911392405,
"grad_norm": 0.44241131725676663,
"learning_rate": 4.570216403561141e-06,
"loss": 0.197,
"step": 987
},
{
"epoch": 4.009113924050633,
"grad_norm": 0.3265966350632202,
"learning_rate": 4.534166308795815e-06,
"loss": 0.2262,
"step": 988
},
{
"epoch": 4.013164556962026,
"grad_norm": 0.2738409540777884,
"learning_rate": 4.498240773728859e-06,
"loss": 0.1877,
"step": 989
},
{
"epoch": 4.017215189873418,
"grad_norm": 0.3192965982513776,
"learning_rate": 4.462440087699609e-06,
"loss": 0.2029,
"step": 990
},
{
"epoch": 4.02126582278481,
"grad_norm": 0.3467720860959134,
"learning_rate": 4.426764539041861e-06,
"loss": 0.195,
"step": 991
},
{
"epoch": 4.025316455696203,
"grad_norm": 0.38845727178963896,
"learning_rate": 4.391214415081582e-06,
"loss": 0.1864,
"step": 992
},
{
"epoch": 4.029367088607595,
"grad_norm": 0.40546089067312313,
"learning_rate": 4.355790002134579e-06,
"loss": 0.2231,
"step": 993
},
{
"epoch": 4.033417721518988,
"grad_norm": 0.3327439149250507,
"learning_rate": 4.320491585504207e-06,
"loss": 0.2039,
"step": 994
},
{
"epoch": 4.03746835443038,
"grad_norm": 0.3238421982711312,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.1985,
"step": 995
},
{
"epoch": 4.041518987341772,
"grad_norm": 0.3401470372748014,
"learning_rate": 4.250273877330691e-06,
"loss": 0.2108,
"step": 996
},
{
"epoch": 4.045569620253165,
"grad_norm": 0.3235429429410752,
"learning_rate": 4.215355151311313e-06,
"loss": 0.2019,
"step": 997
},
{
"epoch": 4.049620253164557,
"grad_norm": 0.32656106010699437,
"learning_rate": 4.180563552651542e-06,
"loss": 0.1935,
"step": 998
},
{
"epoch": 4.053670886075949,
"grad_norm": 0.29409029662095665,
"learning_rate": 4.145899361558147e-06,
"loss": 0.207,
"step": 999
},
{
"epoch": 4.057721518987342,
"grad_norm": 0.26806007558622735,
"learning_rate": 4.111362857211738e-06,
"loss": 0.1902,
"step": 1000
},
{
"epoch": 4.061772151898734,
"grad_norm": 0.30354614308193606,
"learning_rate": 4.076954317764592e-06,
"loss": 0.196,
"step": 1001
},
{
"epoch": 4.065822784810127,
"grad_norm": 0.32670571327561365,
"learning_rate": 4.042674020338335e-06,
"loss": 0.1966,
"step": 1002
},
{
"epoch": 4.069873417721519,
"grad_norm": 0.32539979056424606,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.1959,
"step": 1003
},
{
"epoch": 4.073924050632911,
"grad_norm": 0.2958411649222161,
"learning_rate": 3.974499254868674e-06,
"loss": 0.1975,
"step": 1004
},
{
"epoch": 4.077974683544304,
"grad_norm": 0.273500123498002,
"learning_rate": 3.940605335895451e-06,
"loss": 0.1865,
"step": 1005
},
{
"epoch": 4.082025316455696,
"grad_norm": 0.2736251988723321,
"learning_rate": 3.90684075707908e-06,
"loss": 0.197,
"step": 1006
},
{
"epoch": 4.086075949367088,
"grad_norm": 0.2598125408563322,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.1662,
"step": 1007
},
{
"epoch": 4.090126582278481,
"grad_norm": 0.29269859812924154,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.1888,
"step": 1008
},
{
"epoch": 4.094177215189873,
"grad_norm": 0.2858542297591008,
"learning_rate": 3.806325775702304e-06,
"loss": 0.2056,
"step": 1009
},
{
"epoch": 4.098227848101266,
"grad_norm": 0.27614885822567387,
"learning_rate": 3.773081266416434e-06,
"loss": 0.1871,
"step": 1010
},
{
"epoch": 4.102278481012658,
"grad_norm": 0.26451519490895925,
"learning_rate": 3.739967446503245e-06,
"loss": 0.17,
"step": 1011
},
{
"epoch": 4.10632911392405,
"grad_norm": 0.2752467808155445,
"learning_rate": 3.706984582656894e-06,
"loss": 0.203,
"step": 1012
},
{
"epoch": 4.110379746835443,
"grad_norm": 0.24611645460669995,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.2059,
"step": 1013
},
{
"epoch": 4.114430379746835,
"grad_norm": 0.2497337452771658,
"learning_rate": 3.641412784665648e-06,
"loss": 0.1833,
"step": 1014
},
{
"epoch": 4.118481012658227,
"grad_norm": 0.26198152584102985,
"learning_rate": 3.608824378627005e-06,
"loss": 0.1944,
"step": 1015
},
{
"epoch": 4.12253164556962,
"grad_norm": 0.2757730992347448,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.2076,
"step": 1016
},
{
"epoch": 4.1265822784810124,
"grad_norm": 0.269706650788486,
"learning_rate": 3.544043864774269e-06,
"loss": 0.2121,
"step": 1017
},
{
"epoch": 4.130632911392405,
"grad_norm": 0.25455827836776374,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.18,
"step": 1018
},
{
"epoch": 4.1346835443037975,
"grad_norm": 0.2596168327415316,
"learning_rate": 3.479793485887819e-06,
"loss": 0.1888,
"step": 1019
},
{
"epoch": 4.1387341772151895,
"grad_norm": 0.25123397163434336,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.2149,
"step": 1020
},
{
"epoch": 4.1427848101265825,
"grad_norm": 0.25535819854084696,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.2118,
"step": 1021
},
{
"epoch": 4.146835443037975,
"grad_norm": 0.2662978614786831,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.1956,
"step": 1022
},
{
"epoch": 4.1508860759493675,
"grad_norm": 0.7788986744259833,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.2174,
"step": 1023
},
{
"epoch": 4.15493670886076,
"grad_norm": 0.2621996509324186,
"learning_rate": 3.321500420335e-06,
"loss": 0.2098,
"step": 1024
},
{
"epoch": 4.158987341772152,
"grad_norm": 0.2654694698925293,
"learning_rate": 3.290243771741275e-06,
"loss": 0.2208,
"step": 1025
},
{
"epoch": 4.163037974683545,
"grad_norm": 0.2809389323094826,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.2015,
"step": 1026
},
{
"epoch": 4.167088607594937,
"grad_norm": 0.27030541776004285,
"learning_rate": 3.228134459445149e-06,
"loss": 0.2074,
"step": 1027
},
{
"epoch": 4.171139240506329,
"grad_norm": 0.2705151656557634,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.2044,
"step": 1028
},
{
"epoch": 4.175189873417722,
"grad_norm": 0.3522771785692766,
"learning_rate": 3.166565459258513e-06,
"loss": 0.2073,
"step": 1029
},
{
"epoch": 4.179240506329114,
"grad_norm": 0.38156743222081335,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.1956,
"step": 1030
},
{
"epoch": 4.183291139240506,
"grad_norm": 0.2631111978718513,
"learning_rate": 3.105538754650419e-06,
"loss": 0.2069,
"step": 1031
},
{
"epoch": 4.187341772151899,
"grad_norm": 0.2659998398333357,
"learning_rate": 3.07522937824712e-06,
"loss": 0.2114,
"step": 1032
},
{
"epoch": 4.191392405063291,
"grad_norm": 0.2522949585873704,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.2041,
"step": 1033
},
{
"epoch": 4.195443037974684,
"grad_norm": 0.2570931118633325,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.2032,
"step": 1034
},
{
"epoch": 4.199493670886076,
"grad_norm": 0.24688101620439035,
"learning_rate": 2.985120078631465e-06,
"loss": 0.1921,
"step": 1035
},
{
"epoch": 4.203544303797468,
"grad_norm": 0.2527030808501946,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.2032,
"step": 1036
},
{
"epoch": 4.207594936708861,
"grad_norm": 0.25449864893348595,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.1647,
"step": 1037
},
{
"epoch": 4.211645569620253,
"grad_norm": 0.2528289474002866,
"learning_rate": 2.896244091927678e-06,
"loss": 0.2037,
"step": 1038
},
{
"epoch": 4.215696202531646,
"grad_norm": 0.26499414921145725,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.2131,
"step": 1039
},
{
"epoch": 4.219746835443038,
"grad_norm": 0.2601209678660675,
"learning_rate": 2.837681792951994e-06,
"loss": 0.224,
"step": 1040
},
{
"epoch": 4.22379746835443,
"grad_norm": 0.2675937211703339,
"learning_rate": 2.808607860255981e-06,
"loss": 0.2089,
"step": 1041
},
{
"epoch": 4.227848101265823,
"grad_norm": 0.26779146765265943,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.1984,
"step": 1042
},
{
"epoch": 4.231898734177215,
"grad_norm": 0.2594427451798492,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.2199,
"step": 1043
},
{
"epoch": 4.235949367088607,
"grad_norm": 0.25616734100604494,
"learning_rate": 2.722217735873718e-06,
"loss": 0.1929,
"step": 1044
},
{
"epoch": 4.24,
"grad_norm": 0.24737601009852525,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.2041,
"step": 1045
},
{
"epoch": 4.244050632911392,
"grad_norm": 0.2643382283989595,
"learning_rate": 2.665319697490205e-06,
"loss": 0.198,
"step": 1046
},
{
"epoch": 4.248101265822784,
"grad_norm": 0.26358066831277943,
"learning_rate": 2.637079980714945e-06,
"loss": 0.1789,
"step": 1047
},
{
"epoch": 4.252151898734177,
"grad_norm": 0.24440559702895612,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.1651,
"step": 1048
},
{
"epoch": 4.256202531645569,
"grad_norm": 0.2393972856157291,
"learning_rate": 2.581020289189571e-06,
"loss": 0.211,
"step": 1049
},
{
"epoch": 4.260253164556962,
"grad_norm": 0.259776256729944,
"learning_rate": 2.553200765936501e-06,
"loss": 0.1999,
"step": 1050
},
{
"epoch": 4.264303797468354,
"grad_norm": 0.2527333101787348,
"learning_rate": 2.525521756805962e-06,
"loss": 0.1857,
"step": 1051
},
{
"epoch": 4.2683544303797465,
"grad_norm": 0.2621938410536013,
"learning_rate": 2.497983484720885e-06,
"loss": 0.2093,
"step": 1052
},
{
"epoch": 4.2724050632911394,
"grad_norm": 0.26870903886202435,
"learning_rate": 2.470586171470728e-06,
"loss": 0.1901,
"step": 1053
},
{
"epoch": 4.2764556962025315,
"grad_norm": 0.25477152199533687,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.1942,
"step": 1054
},
{
"epoch": 4.2805063291139245,
"grad_norm": 0.2582249271042767,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.2061,
"step": 1055
},
{
"epoch": 4.2845569620253166,
"grad_norm": 0.2743432462506923,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.2018,
"step": 1056
},
{
"epoch": 4.288607594936709,
"grad_norm": 0.2531837157003946,
"learning_rate": 2.362410902836978e-06,
"loss": 0.2012,
"step": 1057
},
{
"epoch": 4.292658227848102,
"grad_norm": 0.2556271962399679,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.1946,
"step": 1058
},
{
"epoch": 4.296708860759494,
"grad_norm": 0.2612984899381395,
"learning_rate": 2.309174704446411e-06,
"loss": 0.2088,
"step": 1059
},
{
"epoch": 4.300759493670886,
"grad_norm": 0.25992092434618197,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.211,
"step": 1060
},
{
"epoch": 4.304810126582279,
"grad_norm": 0.25312328290487046,
"learning_rate": 2.256508422807855e-06,
"loss": 0.1951,
"step": 1061
},
{
"epoch": 4.308860759493671,
"grad_norm": 0.2553521105102254,
"learning_rate": 2.230389531696946e-06,
"loss": 0.1975,
"step": 1062
},
{
"epoch": 4.312911392405063,
"grad_norm": 0.2587245757264993,
"learning_rate": 2.204413754585857e-06,
"loss": 0.1904,
"step": 1063
},
{
"epoch": 4.316962025316456,
"grad_norm": 0.2622452162251132,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.1832,
"step": 1064
},
{
"epoch": 4.321012658227848,
"grad_norm": 0.2606499579107849,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.2142,
"step": 1065
},
{
"epoch": 4.325063291139241,
"grad_norm": 0.24828440558137282,
"learning_rate": 2.127347193531757e-06,
"loss": 0.2134,
"step": 1066
},
{
"epoch": 4.329113924050633,
"grad_norm": 0.25789451689875287,
"learning_rate": 2.101945952921942e-06,
"loss": 0.1849,
"step": 1067
},
{
"epoch": 4.333164556962025,
"grad_norm": 0.2621031394073994,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.2112,
"step": 1068
},
{
"epoch": 4.337215189873418,
"grad_norm": 0.26534879809308376,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.218,
"step": 1069
},
{
"epoch": 4.34126582278481,
"grad_norm": 0.2615573838473744,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.1909,
"step": 1070
},
{
"epoch": 4.345316455696203,
"grad_norm": 0.26729624004152985,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.2187,
"step": 1071
},
{
"epoch": 4.349367088607595,
"grad_norm": 0.25982169370838243,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.241,
"step": 1072
},
{
"epoch": 4.353417721518987,
"grad_norm": 0.2510727580117948,
"learning_rate": 1.95257270594031e-06,
"loss": 0.235,
"step": 1073
},
{
"epoch": 4.35746835443038,
"grad_norm": 0.25928975119566805,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.1924,
"step": 1074
},
{
"epoch": 4.361518987341772,
"grad_norm": 0.26253850141539453,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.2038,
"step": 1075
},
{
"epoch": 4.365569620253164,
"grad_norm": 0.2633374084629919,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.2199,
"step": 1076
},
{
"epoch": 4.369620253164557,
"grad_norm": 0.268085867203902,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.1946,
"step": 1077
},
{
"epoch": 4.373670886075949,
"grad_norm": 0.253733320758435,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.2097,
"step": 1078
},
{
"epoch": 4.377721518987342,
"grad_norm": 0.2669154712162767,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.1943,
"step": 1079
},
{
"epoch": 4.381772151898734,
"grad_norm": 0.25382443279494127,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.214,
"step": 1080
},
{
"epoch": 4.385822784810126,
"grad_norm": 0.259573618715521,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.1921,
"step": 1081
},
{
"epoch": 4.389873417721519,
"grad_norm": 0.2602820198263081,
"learning_rate": 1.738336950434061e-06,
"loss": 0.181,
"step": 1082
},
{
"epoch": 4.393924050632911,
"grad_norm": 0.2689217864221219,
"learning_rate": 1.715265841375957e-06,
"loss": 0.196,
"step": 1083
},
{
"epoch": 4.397974683544303,
"grad_norm": 0.24901160695872324,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.2044,
"step": 1084
},
{
"epoch": 4.402025316455696,
"grad_norm": 0.25002610658720037,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.1902,
"step": 1085
},
{
"epoch": 4.4060759493670885,
"grad_norm": 0.25754384170720096,
"learning_rate": 1.646936827927441e-06,
"loss": 0.2025,
"step": 1086
},
{
"epoch": 4.410126582278481,
"grad_norm": 0.25340960916236643,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.2047,
"step": 1087
},
{
"epoch": 4.4141772151898735,
"grad_norm": 0.25521768141931783,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.2165,
"step": 1088
},
{
"epoch": 4.418227848101266,
"grad_norm": 0.2501062855806572,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.1865,
"step": 1089
},
{
"epoch": 4.4222784810126585,
"grad_norm": 0.2607938935293541,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.1904,
"step": 1090
},
{
"epoch": 4.426329113924051,
"grad_norm": 0.25704591235078983,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.1986,
"step": 1091
},
{
"epoch": 4.430379746835443,
"grad_norm": 0.24648704314743006,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.189,
"step": 1092
},
{
"epoch": 4.434430379746836,
"grad_norm": 0.26206852461848434,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.2169,
"step": 1093
},
{
"epoch": 4.438481012658228,
"grad_norm": 0.24055572295585215,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.197,
"step": 1094
},
{
"epoch": 4.442531645569621,
"grad_norm": 0.2497119492424784,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.1936,
"step": 1095
},
{
"epoch": 4.446582278481013,
"grad_norm": 0.25159642453577336,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.1842,
"step": 1096
},
{
"epoch": 4.450632911392405,
"grad_norm": 0.2595686473383348,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.1854,
"step": 1097
},
{
"epoch": 4.454683544303798,
"grad_norm": 0.2593917299568197,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.1934,
"step": 1098
},
{
"epoch": 4.45873417721519,
"grad_norm": 0.26258270721543425,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.1868,
"step": 1099
},
{
"epoch": 4.462784810126582,
"grad_norm": 0.2618893741218843,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.1969,
"step": 1100
},
{
"epoch": 4.466835443037975,
"grad_norm": 0.26113182344245556,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.2034,
"step": 1101
},
{
"epoch": 4.470886075949367,
"grad_norm": 0.2493228255722369,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.2288,
"step": 1102
},
{
"epoch": 4.47493670886076,
"grad_norm": 0.25307855044586813,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.1841,
"step": 1103
},
{
"epoch": 4.478987341772152,
"grad_norm": 0.250273955747193,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.1811,
"step": 1104
},
{
"epoch": 4.483037974683544,
"grad_norm": 0.2530546259876481,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.2032,
"step": 1105
},
{
"epoch": 4.487088607594937,
"grad_norm": 0.27182433189447064,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.1965,
"step": 1106
},
{
"epoch": 4.491139240506329,
"grad_norm": 0.24566165376198426,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.1669,
"step": 1107
},
{
"epoch": 4.495189873417721,
"grad_norm": 0.2540728448571459,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.2028,
"step": 1108
},
{
"epoch": 4.499240506329114,
"grad_norm": 0.2524915126414455,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.1979,
"step": 1109
},
{
"epoch": 4.503291139240506,
"grad_norm": 0.26544147011520597,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.2312,
"step": 1110
},
{
"epoch": 4.507341772151898,
"grad_norm": 0.2487871584162263,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.2053,
"step": 1111
},
{
"epoch": 4.511392405063291,
"grad_norm": 0.2519491475535857,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.2275,
"step": 1112
},
{
"epoch": 4.515443037974683,
"grad_norm": 0.2581338019232022,
"learning_rate": 1.092402023876933e-06,
"loss": 0.2044,
"step": 1113
},
{
"epoch": 4.519493670886076,
"grad_norm": 0.26641060531169786,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.2092,
"step": 1114
},
{
"epoch": 4.523544303797468,
"grad_norm": 0.26353998030588965,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.1921,
"step": 1115
},
{
"epoch": 4.52759493670886,
"grad_norm": 0.2510265050845937,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.2067,
"step": 1116
},
{
"epoch": 4.531645569620253,
"grad_norm": 0.25943592805793003,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.1768,
"step": 1117
},
{
"epoch": 4.535696202531645,
"grad_norm": 0.2630884108152766,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.1807,
"step": 1118
},
{
"epoch": 4.539746835443038,
"grad_norm": 0.26324800027858136,
"learning_rate": 9.841383857091947e-07,
"loss": 0.1754,
"step": 1119
},
{
"epoch": 4.54379746835443,
"grad_norm": 0.25600705339287205,
"learning_rate": 9.666296380321616e-07,
"loss": 0.2092,
"step": 1120
},
{
"epoch": 4.5478481012658225,
"grad_norm": 0.2624629292750856,
"learning_rate": 9.492741825034124e-07,
"loss": 0.1975,
"step": 1121
},
{
"epoch": 4.5518987341772155,
"grad_norm": 0.24907389514885084,
"learning_rate": 9.320721589013892e-07,
"loss": 0.192,
"step": 1122
},
{
"epoch": 4.5559493670886075,
"grad_norm": 0.25703402155812216,
"learning_rate": 9.150237057688339e-07,
"loss": 0.2098,
"step": 1123
},
{
"epoch": 4.5600000000000005,
"grad_norm": 0.25515075367773454,
"learning_rate": 8.981289604116328e-07,
"loss": 0.2117,
"step": 1124
},
{
"epoch": 4.564050632911393,
"grad_norm": 0.2678256814668536,
"learning_rate": 8.813880588977542e-07,
"loss": 0.195,
"step": 1125
},
{
"epoch": 4.568101265822785,
"grad_norm": 0.25262397656361074,
"learning_rate": 8.648011360561126e-07,
"loss": 0.2008,
"step": 1126
},
{
"epoch": 4.572151898734178,
"grad_norm": 0.2541373039734592,
"learning_rate": 8.483683254755037e-07,
"loss": 0.2145,
"step": 1127
},
{
"epoch": 4.57620253164557,
"grad_norm": 0.24387272303640745,
"learning_rate": 8.320897595035227e-07,
"loss": 0.1843,
"step": 1128
},
{
"epoch": 4.580253164556962,
"grad_norm": 0.24887737193811613,
"learning_rate": 8.159655692455093e-07,
"loss": 0.1959,
"step": 1129
},
{
"epoch": 4.584303797468355,
"grad_norm": 0.27308648375147004,
"learning_rate": 7.999958845634648e-07,
"loss": 0.1847,
"step": 1130
},
{
"epoch": 4.588354430379747,
"grad_norm": 0.24398006413986845,
"learning_rate": 7.841808340750478e-07,
"loss": 0.2,
"step": 1131
},
{
"epoch": 4.592405063291139,
"grad_norm": 0.24717239191212276,
"learning_rate": 7.685205451524869e-07,
"loss": 0.1773,
"step": 1132
},
{
"epoch": 4.596455696202532,
"grad_norm": 0.25164180216931303,
"learning_rate": 7.530151439216027e-07,
"loss": 0.2055,
"step": 1133
},
{
"epoch": 4.600506329113924,
"grad_norm": 0.264802247610348,
"learning_rate": 7.376647552607675e-07,
"loss": 0.2,
"step": 1134
},
{
"epoch": 4.604556962025317,
"grad_norm": 0.2353577136263203,
"learning_rate": 7.224695027998963e-07,
"loss": 0.1785,
"step": 1135
},
{
"epoch": 4.608607594936709,
"grad_norm": 0.25532222542925975,
"learning_rate": 7.07429508919466e-07,
"loss": 0.1957,
"step": 1136
},
{
"epoch": 4.612658227848101,
"grad_norm": 0.26145666378599247,
"learning_rate": 6.925448947495206e-07,
"loss": 0.2017,
"step": 1137
},
{
"epoch": 4.616708860759494,
"grad_norm": 0.24196088866113447,
"learning_rate": 6.778157801686936e-07,
"loss": 0.1751,
"step": 1138
},
{
"epoch": 4.620759493670886,
"grad_norm": 0.25880424892800513,
"learning_rate": 6.632422838032515e-07,
"loss": 0.2096,
"step": 1139
},
{
"epoch": 4.624810126582279,
"grad_norm": 0.24567283440680557,
"learning_rate": 6.488245230261281e-07,
"loss": 0.1802,
"step": 1140
},
{
"epoch": 4.628860759493671,
"grad_norm": 0.2477971838152083,
"learning_rate": 6.345626139559868e-07,
"loss": 0.1774,
"step": 1141
},
{
"epoch": 4.632911392405063,
"grad_norm": 0.2497714560754988,
"learning_rate": 6.204566714562866e-07,
"loss": 0.2175,
"step": 1142
},
{
"epoch": 4.636962025316456,
"grad_norm": 0.2523938406424071,
"learning_rate": 6.06506809134344e-07,
"loss": 0.2038,
"step": 1143
},
{
"epoch": 4.641012658227848,
"grad_norm": 0.26475387315292453,
"learning_rate": 5.927131393404373e-07,
"loss": 0.2144,
"step": 1144
},
{
"epoch": 4.64506329113924,
"grad_norm": 0.25058119184284355,
"learning_rate": 5.790757731668817e-07,
"loss": 0.2115,
"step": 1145
},
{
"epoch": 4.649113924050633,
"grad_norm": 2.108012870931347,
"learning_rate": 5.655948204471507e-07,
"loss": 0.245,
"step": 1146
},
{
"epoch": 4.653164556962025,
"grad_norm": 0.2536618603065724,
"learning_rate": 5.522703897549875e-07,
"loss": 0.2089,
"step": 1147
},
{
"epoch": 4.657215189873417,
"grad_norm": 0.2566944988894761,
"learning_rate": 5.391025884035239e-07,
"loss": 0.1958,
"step": 1148
},
{
"epoch": 4.66126582278481,
"grad_norm": 0.24443761724587734,
"learning_rate": 5.260915224444207e-07,
"loss": 0.1844,
"step": 1149
},
{
"epoch": 4.665316455696202,
"grad_norm": 0.2564917809340935,
"learning_rate": 5.132372966670129e-07,
"loss": 0.2015,
"step": 1150
},
{
"epoch": 4.669367088607595,
"grad_norm": 0.24772010095010105,
"learning_rate": 5.005400145974704e-07,
"loss": 0.1815,
"step": 1151
},
{
"epoch": 4.673417721518987,
"grad_norm": 0.25911716775552734,
"learning_rate": 4.879997784979562e-07,
"loss": 0.2213,
"step": 1152
},
{
"epoch": 4.6774683544303794,
"grad_norm": 0.2524699359966587,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.1924,
"step": 1153
},
{
"epoch": 4.681518987341772,
"grad_norm": 0.24929623373696363,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.214,
"step": 1154
},
{
"epoch": 4.6855696202531645,
"grad_norm": 0.2683524558300083,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.2059,
"step": 1155
},
{
"epoch": 4.689620253164557,
"grad_norm": 0.24047561905779008,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.1928,
"step": 1156
},
{
"epoch": 4.6936708860759495,
"grad_norm": 0.2627637410715735,
"learning_rate": 4.27657778143431e-07,
"loss": 0.2049,
"step": 1157
},
{
"epoch": 4.697721518987342,
"grad_norm": 0.2663025997674183,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.2097,
"step": 1158
},
{
"epoch": 4.7017721518987345,
"grad_norm": 0.2424310718577902,
"learning_rate": 4.046237371675177e-07,
"loss": 0.2157,
"step": 1159
},
{
"epoch": 4.705822784810127,
"grad_norm": 0.24475612559285465,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.2077,
"step": 1160
},
{
"epoch": 4.709873417721519,
"grad_norm": 0.2458985882504158,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.1873,
"step": 1161
},
{
"epoch": 4.713924050632912,
"grad_norm": 0.2505212663682799,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.2151,
"step": 1162
},
{
"epoch": 4.717974683544304,
"grad_norm": 0.24842055840912042,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.1751,
"step": 1163
},
{
"epoch": 4.722025316455696,
"grad_norm": 0.25703797192125316,
"learning_rate": 3.498020255137813e-07,
"loss": 0.2095,
"step": 1164
},
{
"epoch": 4.726075949367089,
"grad_norm": 0.23809600519830923,
"learning_rate": 3.393121163057811e-07,
"loss": 0.1659,
"step": 1165
},
{
"epoch": 4.730126582278481,
"grad_norm": 0.24732234230931982,
"learning_rate": 3.289805515751399e-07,
"loss": 0.1677,
"step": 1166
},
{
"epoch": 4.734177215189874,
"grad_norm": 0.25506541907425545,
"learning_rate": 3.188074145308573e-07,
"loss": 0.2036,
"step": 1167
},
{
"epoch": 4.738227848101266,
"grad_norm": 0.24424433479851657,
"learning_rate": 3.087927871059804e-07,
"loss": 0.2253,
"step": 1168
},
{
"epoch": 4.742278481012658,
"grad_norm": 0.2556828833208802,
"learning_rate": 2.989367499569418e-07,
"loss": 0.2024,
"step": 1169
},
{
"epoch": 4.746329113924051,
"grad_norm": 0.24532063389595316,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.2023,
"step": 1170
},
{
"epoch": 4.750379746835443,
"grad_norm": 0.24175238452529577,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.178,
"step": 1171
},
{
"epoch": 4.754430379746836,
"grad_norm": 0.2456035522057982,
"learning_rate": 2.703209675663887e-07,
"loss": 0.1892,
"step": 1172
},
{
"epoch": 4.758481012658228,
"grad_norm": 0.25366356707654963,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.1958,
"step": 1173
},
{
"epoch": 4.76253164556962,
"grad_norm": 0.24364209726860575,
"learning_rate": 2.520381518804471e-07,
"loss": 0.2062,
"step": 1174
},
{
"epoch": 4.766582278481013,
"grad_norm": 0.2532719586223214,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.2087,
"step": 1175
},
{
"epoch": 4.770632911392405,
"grad_norm": 0.25072301879287917,
"learning_rate": 2.343915243930317e-07,
"loss": 0.1903,
"step": 1176
},
{
"epoch": 4.774683544303797,
"grad_norm": 0.2554022921665841,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.187,
"step": 1177
},
{
"epoch": 4.77873417721519,
"grad_norm": 0.25112185507988183,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.1971,
"step": 1178
},
{
"epoch": 4.782784810126582,
"grad_norm": 0.26766889986065084,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.1995,
"step": 1179
},
{
"epoch": 4.786835443037974,
"grad_norm": 0.2460592150688571,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.1545,
"step": 1180
},
{
"epoch": 4.790886075949367,
"grad_norm": 0.24314463310606263,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.1779,
"step": 1181
},
{
"epoch": 4.794936708860759,
"grad_norm": 0.2566677738249757,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.2106,
"step": 1182
},
{
"epoch": 4.798987341772152,
"grad_norm": 0.25033412557820445,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.1826,
"step": 1183
},
{
"epoch": 4.803037974683544,
"grad_norm": 0.24026372312028166,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.1737,
"step": 1184
},
{
"epoch": 4.807088607594936,
"grad_norm": 0.2612560648344739,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.2046,
"step": 1185
},
{
"epoch": 4.811139240506329,
"grad_norm": 0.24371697897247008,
"learning_rate": 1.557203888516745e-07,
"loss": 0.2067,
"step": 1186
},
{
"epoch": 4.815189873417721,
"grad_norm": 0.25740540458714745,
"learning_rate": 1.487313098987131e-07,
"loss": 0.2182,
"step": 1187
},
{
"epoch": 4.819240506329114,
"grad_norm": 0.2571647592095984,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.1979,
"step": 1188
},
{
"epoch": 4.8232911392405065,
"grad_norm": 0.2383786360630127,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.1934,
"step": 1189
},
{
"epoch": 4.8273417721518985,
"grad_norm": 0.2554511057628668,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.1884,
"step": 1190
},
{
"epoch": 4.8313924050632915,
"grad_norm": 0.25637506367302954,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.1814,
"step": 1191
},
{
"epoch": 4.8354430379746836,
"grad_norm": 0.24815815988146436,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.1785,
"step": 1192
},
{
"epoch": 4.839493670886076,
"grad_norm": 0.23917163802828875,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.1957,
"step": 1193
},
{
"epoch": 4.843544303797469,
"grad_norm": 0.24775464265333164,
"learning_rate": 1.042873700093061e-07,
"loss": 0.1798,
"step": 1194
},
{
"epoch": 4.847594936708861,
"grad_norm": 0.2655339450441836,
"learning_rate": 9.857879404556291e-08,
"loss": 0.2255,
"step": 1195
},
{
"epoch": 4.851645569620253,
"grad_norm": 0.2500842331193114,
"learning_rate": 9.303050139225722e-08,
"loss": 0.181,
"step": 1196
},
{
"epoch": 4.855696202531646,
"grad_norm": 0.24828085944143033,
"learning_rate": 8.76425367345779e-08,
"loss": 0.1971,
"step": 1197
},
{
"epoch": 4.859746835443038,
"grad_norm": 0.25149439671060814,
"learning_rate": 8.241494346644897e-08,
"loss": 0.1873,
"step": 1198
},
{
"epoch": 4.863797468354431,
"grad_norm": 0.2504313440873231,
"learning_rate": 7.734776369019204e-08,
"loss": 0.1872,
"step": 1199
},
{
"epoch": 4.867848101265823,
"grad_norm": 0.24756551723862585,
"learning_rate": 7.244103821617332e-08,
"loss": 0.203,
"step": 1200
},
{
"epoch": 4.871898734177215,
"grad_norm": 0.25459428437084064,
"learning_rate": 6.769480656248606e-08,
"loss": 0.204,
"step": 1201
},
{
"epoch": 4.875949367088608,
"grad_norm": 0.23784856196722465,
"learning_rate": 6.310910695462635e-08,
"loss": 0.1866,
"step": 1202
},
{
"epoch": 4.88,
"grad_norm": 0.2528393601485236,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.2102,
"step": 1203
},
{
"epoch": 4.884050632911393,
"grad_norm": 0.24753672042114178,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.2046,
"step": 1204
},
{
"epoch": 4.888101265822785,
"grad_norm": 0.2617709211633615,
"learning_rate": 5.031556326567488e-08,
"loss": 0.1912,
"step": 1205
},
{
"epoch": 4.892151898734177,
"grad_norm": 0.24917855199019504,
"learning_rate": 4.637234823364312e-08,
"loss": 0.2113,
"step": 1206
},
{
"epoch": 4.89620253164557,
"grad_norm": 0.2585955363410735,
"learning_rate": 4.258983697558838e-08,
"loss": 0.1992,
"step": 1207
},
{
"epoch": 4.900253164556962,
"grad_norm": 0.248815513161566,
"learning_rate": 3.896805995533548e-08,
"loss": 0.2291,
"step": 1208
},
{
"epoch": 4.904303797468354,
"grad_norm": 0.2492963591007819,
"learning_rate": 3.550704634218028e-08,
"loss": 0.1823,
"step": 1209
},
{
"epoch": 4.908354430379747,
"grad_norm": 0.2588460081211886,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.1981,
"step": 1210
},
{
"epoch": 4.912405063291139,
"grad_norm": 0.24042186296805768,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.1971,
"step": 1211
},
{
"epoch": 4.916455696202531,
"grad_norm": 0.25605520119864716,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.1835,
"step": 1212
},
{
"epoch": 4.920506329113924,
"grad_norm": 0.26764246034256634,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.2126,
"step": 1213
},
{
"epoch": 4.924556962025316,
"grad_norm": 0.2541211851690507,
"learning_rate": 2.061435966221881e-08,
"loss": 0.1896,
"step": 1214
},
{
"epoch": 4.928607594936709,
"grad_norm": 0.24358458062765112,
"learning_rate": 1.811846652477245e-08,
"loss": 0.1832,
"step": 1215
},
{
"epoch": 4.932658227848101,
"grad_norm": 0.25294044643795005,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.1907,
"step": 1216
},
{
"epoch": 4.936708860759493,
"grad_norm": 0.24353032529399743,
"learning_rate": 1.360949303591097e-08,
"loss": 0.2112,
"step": 1217
},
{
"epoch": 4.940759493670886,
"grad_norm": 0.2466978094075244,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.1974,
"step": 1218
},
{
"epoch": 4.944810126582278,
"grad_norm": 0.24367181497075122,
"learning_rate": 9.744388816668172e-09,
"loss": 0.171,
"step": 1219
},
{
"epoch": 4.948860759493671,
"grad_norm": 0.2625426681324267,
"learning_rate": 8.05332740472009e-09,
"loss": 0.1857,
"step": 1220
},
{
"epoch": 4.952911392405063,
"grad_norm": 0.24955477379731322,
"learning_rate": 6.523278382872811e-09,
"loss": 0.1712,
"step": 1221
},
{
"epoch": 4.9569620253164555,
"grad_norm": 0.25297319357389847,
"learning_rate": 5.15425407393133e-09,
"loss": 0.1959,
"step": 1222
},
{
"epoch": 4.961012658227848,
"grad_norm": 0.2527721545261932,
"learning_rate": 3.94626550383137e-09,
"loss": 0.203,
"step": 1223
},
{
"epoch": 4.9650632911392405,
"grad_norm": 0.2424206745400902,
"learning_rate": 2.899322401546112e-09,
"loss": 0.1804,
"step": 1224
},
{
"epoch": 4.969113924050633,
"grad_norm": 0.25698019306593506,
"learning_rate": 2.013433199010706e-09,
"loss": 0.1995,
"step": 1225
},
{
"epoch": 4.9731645569620255,
"grad_norm": 0.25422585350548893,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.2084,
"step": 1226
},
{
"epoch": 4.977215189873418,
"grad_norm": 0.2578738457889016,
"learning_rate": 7.248437353468695e-10,
"loss": 0.1996,
"step": 1227
},
{
"epoch": 4.98126582278481,
"grad_norm": 0.23770006713593,
"learning_rate": 3.221538523412449e-10,
"loss": 0.1895,
"step": 1228
},
{
"epoch": 4.985316455696203,
"grad_norm": 0.26211143695661016,
"learning_rate": 8.053862524670663e-11,
"loss": 0.1732,
"step": 1229
},
{
"epoch": 4.989367088607595,
"grad_norm": 0.24068550557756116,
"learning_rate": 0.0,
"loss": 0.1817,
"step": 1230
},
{
"epoch": 4.989367088607595,
"step": 1230,
"total_flos": 2.53664046736153e+18,
"train_loss": 0.16788045876636737,
"train_runtime": 24163.9435,
"train_samples_per_second": 6.539,
"train_steps_per_second": 0.051
}
],
"logging_steps": 1,
"max_steps": 1230,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"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": 2.53664046736153e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}