OS-Genesis-7B-AC / trainer_state.json
QiushiSun's picture
Upload trainer_state.json with huggingface_hub
738137a verified
raw
history blame
67.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.996101364522417,
"eval_steps": 500,
"global_step": 384,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.005198180636777128,
"grad_norm": 46.50929641723633,
"learning_rate": 8.333333333333333e-08,
"loss": 1.4958,
"step": 1
},
{
"epoch": 0.010396361273554255,
"grad_norm": 36.618709564208984,
"learning_rate": 1.6666666666666665e-07,
"loss": 1.2222,
"step": 2
},
{
"epoch": 0.015594541910331383,
"grad_norm": 42.09670639038086,
"learning_rate": 2.5e-07,
"loss": 1.2627,
"step": 3
},
{
"epoch": 0.02079272254710851,
"grad_norm": 38.13116455078125,
"learning_rate": 3.333333333333333e-07,
"loss": 1.2579,
"step": 4
},
{
"epoch": 0.02599090318388564,
"grad_norm": 40.15380859375,
"learning_rate": 4.1666666666666667e-07,
"loss": 1.2421,
"step": 5
},
{
"epoch": 0.031189083820662766,
"grad_norm": 43.696563720703125,
"learning_rate": 5e-07,
"loss": 1.264,
"step": 6
},
{
"epoch": 0.036387264457439894,
"grad_norm": 44.71561813354492,
"learning_rate": 5.833333333333334e-07,
"loss": 1.3881,
"step": 7
},
{
"epoch": 0.04158544509421702,
"grad_norm": 41.07535171508789,
"learning_rate": 6.666666666666666e-07,
"loss": 1.1705,
"step": 8
},
{
"epoch": 0.04678362573099415,
"grad_norm": 37.13037109375,
"learning_rate": 7.5e-07,
"loss": 1.1108,
"step": 9
},
{
"epoch": 0.05198180636777128,
"grad_norm": 39.47488021850586,
"learning_rate": 8.333333333333333e-07,
"loss": 1.1024,
"step": 10
},
{
"epoch": 0.057179987004548405,
"grad_norm": 35.2398681640625,
"learning_rate": 9.166666666666665e-07,
"loss": 1.0848,
"step": 11
},
{
"epoch": 0.06237816764132553,
"grad_norm": 26.39617347717285,
"learning_rate": 1e-06,
"loss": 0.7678,
"step": 12
},
{
"epoch": 0.06757634827810266,
"grad_norm": 23.569713592529297,
"learning_rate": 9.999821700020548e-07,
"loss": 0.7753,
"step": 13
},
{
"epoch": 0.07277452891487979,
"grad_norm": 20.85965919494629,
"learning_rate": 9.99928681279855e-07,
"loss": 0.6663,
"step": 14
},
{
"epoch": 0.07797270955165692,
"grad_norm": 19.964326858520508,
"learning_rate": 9.998395376482152e-07,
"loss": 0.5468,
"step": 15
},
{
"epoch": 0.08317089018843404,
"grad_norm": 10.887548446655273,
"learning_rate": 9.997147454648588e-07,
"loss": 0.4754,
"step": 16
},
{
"epoch": 0.08836907082521117,
"grad_norm": 10.726633071899414,
"learning_rate": 9.995543136299635e-07,
"loss": 0.4547,
"step": 17
},
{
"epoch": 0.0935672514619883,
"grad_norm": 9.673012733459473,
"learning_rate": 9.993582535855263e-07,
"loss": 0.4634,
"step": 18
},
{
"epoch": 0.09876543209876543,
"grad_norm": 7.258286476135254,
"learning_rate": 9.991265793145479e-07,
"loss": 0.3635,
"step": 19
},
{
"epoch": 0.10396361273554255,
"grad_norm": 5.402667999267578,
"learning_rate": 9.988593073400354e-07,
"loss": 0.356,
"step": 20
},
{
"epoch": 0.10916179337231968,
"grad_norm": 4.859364032745361,
"learning_rate": 9.985564567238236e-07,
"loss": 0.3692,
"step": 21
},
{
"epoch": 0.11435997400909681,
"grad_norm": 3.8276686668395996,
"learning_rate": 9.982180490652164e-07,
"loss": 0.2976,
"step": 22
},
{
"epoch": 0.11955815464587394,
"grad_norm": 3.0185964107513428,
"learning_rate": 9.97844108499445e-07,
"loss": 0.2635,
"step": 23
},
{
"epoch": 0.12475633528265107,
"grad_norm": 2.6632726192474365,
"learning_rate": 9.974346616959475e-07,
"loss": 0.3086,
"step": 24
},
{
"epoch": 0.1299545159194282,
"grad_norm": 2.5440852642059326,
"learning_rate": 9.969897378564667e-07,
"loss": 0.2746,
"step": 25
},
{
"epoch": 0.13515269655620532,
"grad_norm": 2.642413377761841,
"learning_rate": 9.965093687129667e-07,
"loss": 0.2889,
"step": 26
},
{
"epoch": 0.14035087719298245,
"grad_norm": 2.513338565826416,
"learning_rate": 9.959935885253715e-07,
"loss": 0.2778,
"step": 27
},
{
"epoch": 0.14554905782975958,
"grad_norm": 2.2585127353668213,
"learning_rate": 9.954424340791195e-07,
"loss": 0.2311,
"step": 28
},
{
"epoch": 0.1507472384665367,
"grad_norm": 2.021958351135254,
"learning_rate": 9.948559446825411e-07,
"loss": 0.2403,
"step": 29
},
{
"epoch": 0.15594541910331383,
"grad_norm": 2.938659429550171,
"learning_rate": 9.942341621640557e-07,
"loss": 0.2984,
"step": 30
},
{
"epoch": 0.16114359974009096,
"grad_norm": 1.9811211824417114,
"learning_rate": 9.93577130869187e-07,
"loss": 0.2607,
"step": 31
},
{
"epoch": 0.1663417803768681,
"grad_norm": 1.8804433345794678,
"learning_rate": 9.928848976574018e-07,
"loss": 0.2236,
"step": 32
},
{
"epoch": 0.17153996101364521,
"grad_norm": 2.2095425128936768,
"learning_rate": 9.921575118987671e-07,
"loss": 0.247,
"step": 33
},
{
"epoch": 0.17673814165042234,
"grad_norm": 2.0361135005950928,
"learning_rate": 9.91395025470429e-07,
"loss": 0.2519,
"step": 34
},
{
"epoch": 0.18193632228719947,
"grad_norm": 1.9882704019546509,
"learning_rate": 9.905974927529133e-07,
"loss": 0.2387,
"step": 35
},
{
"epoch": 0.1871345029239766,
"grad_norm": 2.1970348358154297,
"learning_rate": 9.897649706262473e-07,
"loss": 0.2506,
"step": 36
},
{
"epoch": 0.19233268356075373,
"grad_norm": 1.9535129070281982,
"learning_rate": 9.888975184659016e-07,
"loss": 0.2491,
"step": 37
},
{
"epoch": 0.19753086419753085,
"grad_norm": 1.7368297576904297,
"learning_rate": 9.879951981385577e-07,
"loss": 0.2002,
"step": 38
},
{
"epoch": 0.20272904483430798,
"grad_norm": 1.8394443988800049,
"learning_rate": 9.870580739976935e-07,
"loss": 0.2107,
"step": 39
},
{
"epoch": 0.2079272254710851,
"grad_norm": 2.0104002952575684,
"learning_rate": 9.860862128789952e-07,
"loss": 0.2373,
"step": 40
},
{
"epoch": 0.21312540610786224,
"grad_norm": 1.7731590270996094,
"learning_rate": 9.850796840955899e-07,
"loss": 0.1881,
"step": 41
},
{
"epoch": 0.21832358674463936,
"grad_norm": 1.937873363494873,
"learning_rate": 9.840385594331022e-07,
"loss": 0.2238,
"step": 42
},
{
"epoch": 0.2235217673814165,
"grad_norm": 1.9435638189315796,
"learning_rate": 9.82962913144534e-07,
"loss": 0.2237,
"step": 43
},
{
"epoch": 0.22871994801819362,
"grad_norm": 1.786537766456604,
"learning_rate": 9.818528219449704e-07,
"loss": 0.1951,
"step": 44
},
{
"epoch": 0.23391812865497075,
"grad_norm": 1.7915631532669067,
"learning_rate": 9.807083650061062e-07,
"loss": 0.2257,
"step": 45
},
{
"epoch": 0.23911630929174787,
"grad_norm": 1.798353910446167,
"learning_rate": 9.79529623950601e-07,
"loss": 0.236,
"step": 46
},
{
"epoch": 0.244314489928525,
"grad_norm": 1.872049331665039,
"learning_rate": 9.783166828462572e-07,
"loss": 0.2354,
"step": 47
},
{
"epoch": 0.24951267056530213,
"grad_norm": 1.879210114479065,
"learning_rate": 9.770696282000244e-07,
"loss": 0.2229,
"step": 48
},
{
"epoch": 0.25471085120207926,
"grad_norm": 1.9663130044937134,
"learning_rate": 9.757885489518296e-07,
"loss": 0.2461,
"step": 49
},
{
"epoch": 0.2599090318388564,
"grad_norm": 1.6957286596298218,
"learning_rate": 9.744735364682344e-07,
"loss": 0.2065,
"step": 50
},
{
"epoch": 0.2651072124756335,
"grad_norm": 1.7848544120788574,
"learning_rate": 9.731246845359184e-07,
"loss": 0.1949,
"step": 51
},
{
"epoch": 0.27030539311241064,
"grad_norm": 1.8262349367141724,
"learning_rate": 9.7174208935499e-07,
"loss": 0.2135,
"step": 52
},
{
"epoch": 0.27550357374918777,
"grad_norm": 1.6100451946258545,
"learning_rate": 9.703258495321265e-07,
"loss": 0.1643,
"step": 53
},
{
"epoch": 0.2807017543859649,
"grad_norm": 1.6476277112960815,
"learning_rate": 9.688760660735402e-07,
"loss": 0.1796,
"step": 54
},
{
"epoch": 0.285899935022742,
"grad_norm": 1.6926974058151245,
"learning_rate": 9.673928423777756e-07,
"loss": 0.2048,
"step": 55
},
{
"epoch": 0.29109811565951915,
"grad_norm": 1.7797563076019287,
"learning_rate": 9.658762842283341e-07,
"loss": 0.1953,
"step": 56
},
{
"epoch": 0.2962962962962963,
"grad_norm": 1.7844983339309692,
"learning_rate": 9.643264997861312e-07,
"loss": 0.2103,
"step": 57
},
{
"epoch": 0.3014944769330734,
"grad_norm": 1.8552502393722534,
"learning_rate": 9.627435995817797e-07,
"loss": 0.1854,
"step": 58
},
{
"epoch": 0.30669265756985054,
"grad_norm": 1.8308689594268799,
"learning_rate": 9.611276965077097e-07,
"loss": 0.1892,
"step": 59
},
{
"epoch": 0.31189083820662766,
"grad_norm": 1.8636093139648438,
"learning_rate": 9.594789058101153e-07,
"loss": 0.216,
"step": 60
},
{
"epoch": 0.3170890188434048,
"grad_norm": 1.5857099294662476,
"learning_rate": 9.577973450807351e-07,
"loss": 0.1924,
"step": 61
},
{
"epoch": 0.3222871994801819,
"grad_norm": 1.670000433921814,
"learning_rate": 9.560831342484666e-07,
"loss": 0.2088,
"step": 62
},
{
"epoch": 0.32748538011695905,
"grad_norm": 1.8758388757705688,
"learning_rate": 9.543363955708124e-07,
"loss": 0.1697,
"step": 63
},
{
"epoch": 0.3326835607537362,
"grad_norm": 2.020310401916504,
"learning_rate": 9.525572536251605e-07,
"loss": 0.2249,
"step": 64
},
{
"epoch": 0.3378817413905133,
"grad_norm": 1.8294882774353027,
"learning_rate": 9.507458352999001e-07,
"loss": 0.1884,
"step": 65
},
{
"epoch": 0.34307992202729043,
"grad_norm": 1.606002926826477,
"learning_rate": 9.489022697853708e-07,
"loss": 0.1761,
"step": 66
},
{
"epoch": 0.34827810266406756,
"grad_norm": 1.6073530912399292,
"learning_rate": 9.470266885646503e-07,
"loss": 0.1871,
"step": 67
},
{
"epoch": 0.3534762833008447,
"grad_norm": 1.7087726593017578,
"learning_rate": 9.451192254041758e-07,
"loss": 0.173,
"step": 68
},
{
"epoch": 0.3586744639376218,
"grad_norm": 1.7764538526535034,
"learning_rate": 9.431800163442041e-07,
"loss": 0.1957,
"step": 69
},
{
"epoch": 0.36387264457439894,
"grad_norm": 1.8759775161743164,
"learning_rate": 9.412091996891095e-07,
"loss": 0.2154,
"step": 70
},
{
"epoch": 0.36907082521117607,
"grad_norm": 1.8281443119049072,
"learning_rate": 9.392069159975198e-07,
"loss": 0.1679,
"step": 71
},
{
"epoch": 0.3742690058479532,
"grad_norm": 1.7894129753112793,
"learning_rate": 9.37173308072291e-07,
"loss": 0.1679,
"step": 72
},
{
"epoch": 0.3794671864847303,
"grad_norm": 1.6492183208465576,
"learning_rate": 9.35108520950324e-07,
"loss": 0.1833,
"step": 73
},
{
"epoch": 0.38466536712150745,
"grad_norm": 1.6076239347457886,
"learning_rate": 9.330127018922193e-07,
"loss": 0.1507,
"step": 74
},
{
"epoch": 0.3898635477582846,
"grad_norm": 1.8182544708251953,
"learning_rate": 9.308860003717748e-07,
"loss": 0.1759,
"step": 75
},
{
"epoch": 0.3950617283950617,
"grad_norm": 2.183497667312622,
"learning_rate": 9.287285680653254e-07,
"loss": 0.2069,
"step": 76
},
{
"epoch": 0.40025990903183883,
"grad_norm": 1.9281930923461914,
"learning_rate": 9.265405588409256e-07,
"loss": 0.1813,
"step": 77
},
{
"epoch": 0.40545808966861596,
"grad_norm": 1.7534650564193726,
"learning_rate": 9.243221287473755e-07,
"loss": 0.1764,
"step": 78
},
{
"epoch": 0.4106562703053931,
"grad_norm": 1.7174078226089478,
"learning_rate": 9.220734360030906e-07,
"loss": 0.1863,
"step": 79
},
{
"epoch": 0.4158544509421702,
"grad_norm": 1.7550305128097534,
"learning_rate": 9.197946409848194e-07,
"loss": 0.1718,
"step": 80
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.4776816368103027,
"learning_rate": 9.174859062162037e-07,
"loss": 0.156,
"step": 81
},
{
"epoch": 0.4262508122157245,
"grad_norm": 1.7932229042053223,
"learning_rate": 9.151473963561882e-07,
"loss": 0.1821,
"step": 82
},
{
"epoch": 0.4314489928525016,
"grad_norm": 1.6103583574295044,
"learning_rate": 9.127792781872768e-07,
"loss": 0.1749,
"step": 83
},
{
"epoch": 0.43664717348927873,
"grad_norm": 1.8216729164123535,
"learning_rate": 9.103817206036382e-07,
"loss": 0.177,
"step": 84
},
{
"epoch": 0.44184535412605586,
"grad_norm": 1.7169886827468872,
"learning_rate": 9.079548945990592e-07,
"loss": 0.1845,
"step": 85
},
{
"epoch": 0.447043534762833,
"grad_norm": 1.4935150146484375,
"learning_rate": 9.054989732547506e-07,
"loss": 0.1518,
"step": 86
},
{
"epoch": 0.4522417153996101,
"grad_norm": 1.7215607166290283,
"learning_rate": 9.030141317270025e-07,
"loss": 0.1651,
"step": 87
},
{
"epoch": 0.45743989603638724,
"grad_norm": 1.885299801826477,
"learning_rate": 9.005005472346923e-07,
"loss": 0.1862,
"step": 88
},
{
"epoch": 0.46263807667316437,
"grad_norm": 1.6924781799316406,
"learning_rate": 8.979583990466452e-07,
"loss": 0.1834,
"step": 89
},
{
"epoch": 0.4678362573099415,
"grad_norm": 1.6620601415634155,
"learning_rate": 8.953878684688492e-07,
"loss": 0.1736,
"step": 90
},
{
"epoch": 0.4730344379467186,
"grad_norm": 1.7256325483322144,
"learning_rate": 8.92789138831524e-07,
"loss": 0.1792,
"step": 91
},
{
"epoch": 0.47823261858349575,
"grad_norm": 1.6039340496063232,
"learning_rate": 8.901623954760459e-07,
"loss": 0.1704,
"step": 92
},
{
"epoch": 0.4834307992202729,
"grad_norm": 1.6422524452209473,
"learning_rate": 8.875078257417294e-07,
"loss": 0.1621,
"step": 93
},
{
"epoch": 0.48862897985705,
"grad_norm": 1.6837060451507568,
"learning_rate": 8.84825618952466e-07,
"loss": 0.183,
"step": 94
},
{
"epoch": 0.49382716049382713,
"grad_norm": 1.750653862953186,
"learning_rate": 8.821159664032223e-07,
"loss": 0.1689,
"step": 95
},
{
"epoch": 0.49902534113060426,
"grad_norm": 1.6462229490280151,
"learning_rate": 8.793790613463954e-07,
"loss": 0.1394,
"step": 96
},
{
"epoch": 0.5042235217673814,
"grad_norm": 1.7336857318878174,
"learning_rate": 8.766150989780317e-07,
"loss": 0.1581,
"step": 97
},
{
"epoch": 0.5094217024041585,
"grad_norm": 1.8384933471679688,
"learning_rate": 8.738242764239046e-07,
"loss": 0.1918,
"step": 98
},
{
"epoch": 0.5146198830409356,
"grad_norm": 1.723486065864563,
"learning_rate": 8.710067927254554e-07,
"loss": 0.1737,
"step": 99
},
{
"epoch": 0.5198180636777128,
"grad_norm": 1.9092669486999512,
"learning_rate": 8.681628488255986e-07,
"loss": 0.1728,
"step": 100
},
{
"epoch": 0.5250162443144899,
"grad_norm": 1.729762315750122,
"learning_rate": 8.652926475543898e-07,
"loss": 0.162,
"step": 101
},
{
"epoch": 0.530214424951267,
"grad_norm": 1.7867392301559448,
"learning_rate": 8.623963936145599e-07,
"loss": 0.1658,
"step": 102
},
{
"epoch": 0.5354126055880442,
"grad_norm": 2.0217678546905518,
"learning_rate": 8.594742935669164e-07,
"loss": 0.1865,
"step": 103
},
{
"epoch": 0.5406107862248213,
"grad_norm": 1.7473349571228027,
"learning_rate": 8.565265558156101e-07,
"loss": 0.1535,
"step": 104
},
{
"epoch": 0.5458089668615984,
"grad_norm": 1.5292036533355713,
"learning_rate": 8.535533905932737e-07,
"loss": 0.1559,
"step": 105
},
{
"epoch": 0.5510071474983755,
"grad_norm": 1.5472049713134766,
"learning_rate": 8.505550099460263e-07,
"loss": 0.1423,
"step": 106
},
{
"epoch": 0.5562053281351527,
"grad_norm": 1.636443853378296,
"learning_rate": 8.475316277183508e-07,
"loss": 0.1747,
"step": 107
},
{
"epoch": 0.5614035087719298,
"grad_norm": 1.5992189645767212,
"learning_rate": 8.444834595378433e-07,
"loss": 0.1766,
"step": 108
},
{
"epoch": 0.5666016894087069,
"grad_norm": 1.6766347885131836,
"learning_rate": 8.414107227998328e-07,
"loss": 0.1421,
"step": 109
},
{
"epoch": 0.571799870045484,
"grad_norm": 1.7345399856567383,
"learning_rate": 8.383136366518787e-07,
"loss": 0.1752,
"step": 110
},
{
"epoch": 0.5769980506822612,
"grad_norm": 1.669264793395996,
"learning_rate": 8.351924219781392e-07,
"loss": 0.1661,
"step": 111
},
{
"epoch": 0.5821962313190383,
"grad_norm": 1.7636350393295288,
"learning_rate": 8.320473013836195e-07,
"loss": 0.1892,
"step": 112
},
{
"epoch": 0.5873944119558154,
"grad_norm": 1.8429635763168335,
"learning_rate": 8.288784991782945e-07,
"loss": 0.1883,
"step": 113
},
{
"epoch": 0.5925925925925926,
"grad_norm": 1.5329152345657349,
"learning_rate": 8.256862413611112e-07,
"loss": 0.1472,
"step": 114
},
{
"epoch": 0.5977907732293697,
"grad_norm": 1.9208284616470337,
"learning_rate": 8.22470755603871e-07,
"loss": 0.1714,
"step": 115
},
{
"epoch": 0.6029889538661468,
"grad_norm": 1.6381752490997314,
"learning_rate": 8.192322712349917e-07,
"loss": 0.1806,
"step": 116
},
{
"epoch": 0.6081871345029239,
"grad_norm": 1.5502922534942627,
"learning_rate": 8.159710192231519e-07,
"loss": 0.1653,
"step": 117
},
{
"epoch": 0.6133853151397011,
"grad_norm": 1.604650616645813,
"learning_rate": 8.126872321608183e-07,
"loss": 0.1478,
"step": 118
},
{
"epoch": 0.6185834957764782,
"grad_norm": 1.6860443353652954,
"learning_rate": 8.093811442476572e-07,
"loss": 0.1639,
"step": 119
},
{
"epoch": 0.6237816764132553,
"grad_norm": 1.5915076732635498,
"learning_rate": 8.060529912738314e-07,
"loss": 0.1511,
"step": 120
},
{
"epoch": 0.6289798570500325,
"grad_norm": 1.7241225242614746,
"learning_rate": 8.027030106031835e-07,
"loss": 0.1848,
"step": 121
},
{
"epoch": 0.6341780376868096,
"grad_norm": 1.7747095823287964,
"learning_rate": 7.993314411563075e-07,
"loss": 0.1816,
"step": 122
},
{
"epoch": 0.6393762183235867,
"grad_norm": 1.6497771739959717,
"learning_rate": 7.959385233935085e-07,
"loss": 0.1696,
"step": 123
},
{
"epoch": 0.6445743989603638,
"grad_norm": 1.4712307453155518,
"learning_rate": 7.925244992976537e-07,
"loss": 0.1297,
"step": 124
},
{
"epoch": 0.649772579597141,
"grad_norm": 1.618713140487671,
"learning_rate": 7.890896123569135e-07,
"loss": 0.1708,
"step": 125
},
{
"epoch": 0.6549707602339181,
"grad_norm": 1.8550593852996826,
"learning_rate": 7.856341075473961e-07,
"loss": 0.1646,
"step": 126
},
{
"epoch": 0.6601689408706952,
"grad_norm": 1.7929205894470215,
"learning_rate": 7.821582313156763e-07,
"loss": 0.1555,
"step": 127
},
{
"epoch": 0.6653671215074723,
"grad_norm": 1.8011633157730103,
"learning_rate": 7.786622315612181e-07,
"loss": 0.1882,
"step": 128
},
{
"epoch": 0.6705653021442495,
"grad_norm": 1.642986536026001,
"learning_rate": 7.751463576186957e-07,
"loss": 0.1659,
"step": 129
},
{
"epoch": 0.6757634827810266,
"grad_norm": 1.547602653503418,
"learning_rate": 7.716108602402094e-07,
"loss": 0.1479,
"step": 130
},
{
"epoch": 0.6809616634178037,
"grad_norm": 1.6602659225463867,
"learning_rate": 7.680559915774033e-07,
"loss": 0.1627,
"step": 131
},
{
"epoch": 0.6861598440545809,
"grad_norm": 1.8091386556625366,
"learning_rate": 7.644820051634812e-07,
"loss": 0.1637,
"step": 132
},
{
"epoch": 0.691358024691358,
"grad_norm": 1.669487476348877,
"learning_rate": 7.608891558951248e-07,
"loss": 0.1599,
"step": 133
},
{
"epoch": 0.6965562053281351,
"grad_norm": 1.9016000032424927,
"learning_rate": 7.572777000143145e-07,
"loss": 0.1654,
"step": 134
},
{
"epoch": 0.7017543859649122,
"grad_norm": 1.4672502279281616,
"learning_rate": 7.536478950900536e-07,
"loss": 0.1482,
"step": 135
},
{
"epoch": 0.7069525666016894,
"grad_norm": 1.4602234363555908,
"learning_rate": 7.5e-07,
"loss": 0.1214,
"step": 136
},
{
"epoch": 0.7121507472384665,
"grad_norm": 1.725661277770996,
"learning_rate": 7.463342749120013e-07,
"loss": 0.1406,
"step": 137
},
{
"epoch": 0.7173489278752436,
"grad_norm": 1.6164398193359375,
"learning_rate": 7.426509812655405e-07,
"loss": 0.1492,
"step": 138
},
{
"epoch": 0.7225471085120208,
"grad_norm": 1.609312891960144,
"learning_rate": 7.389503817530905e-07,
"loss": 0.1669,
"step": 139
},
{
"epoch": 0.7277452891487979,
"grad_norm": 1.512629508972168,
"learning_rate": 7.352327403013779e-07,
"loss": 0.1318,
"step": 140
},
{
"epoch": 0.732943469785575,
"grad_norm": 1.7129087448120117,
"learning_rate": 7.314983220525604e-07,
"loss": 0.1762,
"step": 141
},
{
"epoch": 0.7381416504223521,
"grad_norm": 1.6480506658554077,
"learning_rate": 7.277473933453169e-07,
"loss": 0.1738,
"step": 142
},
{
"epoch": 0.7433398310591293,
"grad_norm": 1.5904552936553955,
"learning_rate": 7.239802216958522e-07,
"loss": 0.1558,
"step": 143
},
{
"epoch": 0.7485380116959064,
"grad_norm": 1.6988767385482788,
"learning_rate": 7.201970757788171e-07,
"loss": 0.1661,
"step": 144
},
{
"epoch": 0.7537361923326835,
"grad_norm": 1.5458639860153198,
"learning_rate": 7.163982254081474e-07,
"loss": 0.1338,
"step": 145
},
{
"epoch": 0.7589343729694606,
"grad_norm": 1.5240118503570557,
"learning_rate": 7.125839415178203e-07,
"loss": 0.1405,
"step": 146
},
{
"epoch": 0.7641325536062378,
"grad_norm": 1.7464412450790405,
"learning_rate": 7.087544961425316e-07,
"loss": 0.1682,
"step": 147
},
{
"epoch": 0.7693307342430149,
"grad_norm": 1.7425211668014526,
"learning_rate": 7.049101623982937e-07,
"loss": 0.1839,
"step": 148
},
{
"epoch": 0.774528914879792,
"grad_norm": 1.4918522834777832,
"learning_rate": 7.010512144629579e-07,
"loss": 0.1124,
"step": 149
},
{
"epoch": 0.7797270955165692,
"grad_norm": 1.6756539344787598,
"learning_rate": 6.971779275566593e-07,
"loss": 0.1546,
"step": 150
},
{
"epoch": 0.7849252761533463,
"grad_norm": 1.5222876071929932,
"learning_rate": 6.93290577922188e-07,
"loss": 0.1313,
"step": 151
},
{
"epoch": 0.7901234567901234,
"grad_norm": 1.548453688621521,
"learning_rate": 6.89389442805288e-07,
"loss": 0.1349,
"step": 152
},
{
"epoch": 0.7953216374269005,
"grad_norm": 1.6898419857025146,
"learning_rate": 6.85474800434884e-07,
"loss": 0.1568,
"step": 153
},
{
"epoch": 0.8005198180636777,
"grad_norm": 1.8794304132461548,
"learning_rate": 6.815469300032373e-07,
"loss": 0.161,
"step": 154
},
{
"epoch": 0.8057179987004548,
"grad_norm": 1.6816418170928955,
"learning_rate": 6.776061116460352e-07,
"loss": 0.1615,
"step": 155
},
{
"epoch": 0.8109161793372319,
"grad_norm": 1.960444688796997,
"learning_rate": 6.7365262642241e-07,
"loss": 0.1948,
"step": 156
},
{
"epoch": 0.816114359974009,
"grad_norm": 1.6450730562210083,
"learning_rate": 6.696867562948962e-07,
"loss": 0.161,
"step": 157
},
{
"epoch": 0.8213125406107862,
"grad_norm": 1.4993230104446411,
"learning_rate": 6.657087841093179e-07,
"loss": 0.1476,
"step": 158
},
{
"epoch": 0.8265107212475633,
"grad_norm": 1.856066346168518,
"learning_rate": 6.61718993574619e-07,
"loss": 0.1599,
"step": 159
},
{
"epoch": 0.8317089018843404,
"grad_norm": 1.6243445873260498,
"learning_rate": 6.577176692426278e-07,
"loss": 0.1548,
"step": 160
},
{
"epoch": 0.8369070825211176,
"grad_norm": 1.538219928741455,
"learning_rate": 6.537050964877625e-07,
"loss": 0.1428,
"step": 161
},
{
"epoch": 0.8421052631578947,
"grad_norm": 1.429417371749878,
"learning_rate": 6.496815614866791e-07,
"loss": 0.1205,
"step": 162
},
{
"epoch": 0.8473034437946718,
"grad_norm": 1.7732073068618774,
"learning_rate": 6.456473511978606e-07,
"loss": 0.1903,
"step": 163
},
{
"epoch": 0.852501624431449,
"grad_norm": 1.575061321258545,
"learning_rate": 6.416027533411519e-07,
"loss": 0.1571,
"step": 164
},
{
"epoch": 0.8576998050682261,
"grad_norm": 1.6352499723434448,
"learning_rate": 6.375480563772389e-07,
"loss": 0.1484,
"step": 165
},
{
"epoch": 0.8628979857050032,
"grad_norm": 1.7170888185501099,
"learning_rate": 6.334835494870758e-07,
"loss": 0.1735,
"step": 166
},
{
"epoch": 0.8680961663417803,
"grad_norm": 1.5450496673583984,
"learning_rate": 6.294095225512604e-07,
"loss": 0.1339,
"step": 167
},
{
"epoch": 0.8732943469785575,
"grad_norm": 1.5989458560943604,
"learning_rate": 6.253262661293602e-07,
"loss": 0.1393,
"step": 168
},
{
"epoch": 0.8784925276153346,
"grad_norm": 1.464534878730774,
"learning_rate": 6.2123407143919e-07,
"loss": 0.1421,
"step": 169
},
{
"epoch": 0.8836907082521117,
"grad_norm": 1.6165345907211304,
"learning_rate": 6.17133230336041e-07,
"loss": 0.154,
"step": 170
},
{
"epoch": 0.8888888888888888,
"grad_norm": 1.5105384588241577,
"learning_rate": 6.130240352918674e-07,
"loss": 0.1614,
"step": 171
},
{
"epoch": 0.894087069525666,
"grad_norm": 1.6538264751434326,
"learning_rate": 6.089067793744257e-07,
"loss": 0.1213,
"step": 172
},
{
"epoch": 0.8992852501624431,
"grad_norm": 1.5659717321395874,
"learning_rate": 6.047817562263743e-07,
"loss": 0.1349,
"step": 173
},
{
"epoch": 0.9044834307992202,
"grad_norm": 1.6108099222183228,
"learning_rate": 6.0064926004433e-07,
"loss": 0.1572,
"step": 174
},
{
"epoch": 0.9096816114359974,
"grad_norm": 1.7230148315429688,
"learning_rate": 5.965095855578868e-07,
"loss": 0.1376,
"step": 175
},
{
"epoch": 0.9148797920727745,
"grad_norm": 1.7344483137130737,
"learning_rate": 5.923630280085947e-07,
"loss": 0.1572,
"step": 176
},
{
"epoch": 0.9200779727095516,
"grad_norm": 1.6481879949569702,
"learning_rate": 5.882098831289043e-07,
"loss": 0.1626,
"step": 177
},
{
"epoch": 0.9252761533463287,
"grad_norm": 1.7318065166473389,
"learning_rate": 5.840504471210741e-07,
"loss": 0.1756,
"step": 178
},
{
"epoch": 0.9304743339831059,
"grad_norm": 1.676165223121643,
"learning_rate": 5.79885016636046e-07,
"loss": 0.147,
"step": 179
},
{
"epoch": 0.935672514619883,
"grad_norm": 1.4620646238327026,
"learning_rate": 5.757138887522883e-07,
"loss": 0.1249,
"step": 180
},
{
"epoch": 0.9408706952566601,
"grad_norm": 1.65927255153656,
"learning_rate": 5.71537360954607e-07,
"loss": 0.163,
"step": 181
},
{
"epoch": 0.9460688758934372,
"grad_norm": 1.5536587238311768,
"learning_rate": 5.673557311129306e-07,
"loss": 0.1351,
"step": 182
},
{
"epoch": 0.9512670565302144,
"grad_norm": 1.7076836824417114,
"learning_rate": 5.631692974610647e-07,
"loss": 0.1771,
"step": 183
},
{
"epoch": 0.9564652371669915,
"grad_norm": 1.4979828596115112,
"learning_rate": 5.589783585754231e-07,
"loss": 0.121,
"step": 184
},
{
"epoch": 0.9616634178037686,
"grad_norm": 1.5839756727218628,
"learning_rate": 5.547832133537327e-07,
"loss": 0.1458,
"step": 185
},
{
"epoch": 0.9668615984405458,
"grad_norm": 1.7546137571334839,
"learning_rate": 5.505841609937161e-07,
"loss": 0.1671,
"step": 186
},
{
"epoch": 0.9720597790773229,
"grad_norm": 1.7105190753936768,
"learning_rate": 5.463815009717532e-07,
"loss": 0.1314,
"step": 187
},
{
"epoch": 0.9772579597141,
"grad_norm": 1.8557852506637573,
"learning_rate": 5.421755330215223e-07,
"loss": 0.1794,
"step": 188
},
{
"epoch": 0.9824561403508771,
"grad_norm": 1.569214105606079,
"learning_rate": 5.379665571126231e-07,
"loss": 0.1307,
"step": 189
},
{
"epoch": 0.9876543209876543,
"grad_norm": 1.6137492656707764,
"learning_rate": 5.337548734291826e-07,
"loss": 0.1412,
"step": 190
},
{
"epoch": 0.9928525016244314,
"grad_norm": 1.6707996129989624,
"learning_rate": 5.295407823484467e-07,
"loss": 0.1627,
"step": 191
},
{
"epoch": 0.9980506822612085,
"grad_norm": 2.3229496479034424,
"learning_rate": 5.253245844193564e-07,
"loss": 0.1872,
"step": 192
},
{
"epoch": 1.0032488628979856,
"grad_norm": 1.4374881982803345,
"learning_rate": 5.211065803411134e-07,
"loss": 0.1118,
"step": 193
},
{
"epoch": 1.0084470435347628,
"grad_norm": 1.7097264528274536,
"learning_rate": 5.168870709417341e-07,
"loss": 0.1603,
"step": 194
},
{
"epoch": 1.01364522417154,
"grad_norm": 1.5904935598373413,
"learning_rate": 5.126663571565939e-07,
"loss": 0.128,
"step": 195
},
{
"epoch": 1.018843404808317,
"grad_norm": 1.5433835983276367,
"learning_rate": 5.084447400069654e-07,
"loss": 0.1192,
"step": 196
},
{
"epoch": 1.0240415854450942,
"grad_norm": 1.40073561668396,
"learning_rate": 5.042225205785492e-07,
"loss": 0.1188,
"step": 197
},
{
"epoch": 1.0292397660818713,
"grad_norm": 1.6374619007110596,
"learning_rate": 5e-07,
"loss": 0.1486,
"step": 198
},
{
"epoch": 1.0344379467186484,
"grad_norm": 1.4800790548324585,
"learning_rate": 4.957774794214508e-07,
"loss": 0.1297,
"step": 199
},
{
"epoch": 1.0396361273554255,
"grad_norm": 1.5941686630249023,
"learning_rate": 4.915552599930345e-07,
"loss": 0.1466,
"step": 200
},
{
"epoch": 1.0448343079922027,
"grad_norm": 1.5258111953735352,
"learning_rate": 4.873336428434061e-07,
"loss": 0.1264,
"step": 201
},
{
"epoch": 1.0500324886289798,
"grad_norm": 1.725998878479004,
"learning_rate": 4.831129290582659e-07,
"loss": 0.1334,
"step": 202
},
{
"epoch": 1.055230669265757,
"grad_norm": 1.8626656532287598,
"learning_rate": 4.788934196588865e-07,
"loss": 0.1503,
"step": 203
},
{
"epoch": 1.060428849902534,
"grad_norm": 1.4593480825424194,
"learning_rate": 4.746754155806437e-07,
"loss": 0.1225,
"step": 204
},
{
"epoch": 1.0656270305393112,
"grad_norm": 1.6284171342849731,
"learning_rate": 4.7045921765155337e-07,
"loss": 0.1397,
"step": 205
},
{
"epoch": 1.0708252111760883,
"grad_norm": 1.4469428062438965,
"learning_rate": 4.662451265708174e-07,
"loss": 0.1082,
"step": 206
},
{
"epoch": 1.0760233918128654,
"grad_norm": 1.3986436128616333,
"learning_rate": 4.620334428873769e-07,
"loss": 0.1025,
"step": 207
},
{
"epoch": 1.0812215724496426,
"grad_norm": 1.6322413682937622,
"learning_rate": 4.5782446697847764e-07,
"loss": 0.126,
"step": 208
},
{
"epoch": 1.0864197530864197,
"grad_norm": 1.4022878408432007,
"learning_rate": 4.536184990282467e-07,
"loss": 0.0932,
"step": 209
},
{
"epoch": 1.0916179337231968,
"grad_norm": 1.8021215200424194,
"learning_rate": 4.4941583900628393e-07,
"loss": 0.1662,
"step": 210
},
{
"epoch": 1.096816114359974,
"grad_norm": 1.806060552597046,
"learning_rate": 4.4521678664626745e-07,
"loss": 0.1574,
"step": 211
},
{
"epoch": 1.102014294996751,
"grad_norm": 1.7470866441726685,
"learning_rate": 4.4102164142457705e-07,
"loss": 0.1467,
"step": 212
},
{
"epoch": 1.1072124756335282,
"grad_norm": 1.5723119974136353,
"learning_rate": 4.368307025389355e-07,
"loss": 0.1084,
"step": 213
},
{
"epoch": 1.1124106562703053,
"grad_norm": 1.5377353429794312,
"learning_rate": 4.326442688870696e-07,
"loss": 0.116,
"step": 214
},
{
"epoch": 1.1176088369070825,
"grad_norm": 1.7077683210372925,
"learning_rate": 4.2846263904539303e-07,
"loss": 0.1483,
"step": 215
},
{
"epoch": 1.1228070175438596,
"grad_norm": 1.8333083391189575,
"learning_rate": 4.242861112477118e-07,
"loss": 0.1527,
"step": 216
},
{
"epoch": 1.1280051981806367,
"grad_norm": 1.5769239664077759,
"learning_rate": 4.201149833639539e-07,
"loss": 0.121,
"step": 217
},
{
"epoch": 1.1332033788174138,
"grad_norm": 1.919921636581421,
"learning_rate": 4.15949552878926e-07,
"loss": 0.173,
"step": 218
},
{
"epoch": 1.138401559454191,
"grad_norm": 1.5571967363357544,
"learning_rate": 4.117901168710959e-07,
"loss": 0.1227,
"step": 219
},
{
"epoch": 1.143599740090968,
"grad_norm": 1.6683669090270996,
"learning_rate": 4.0763697199140546e-07,
"loss": 0.1422,
"step": 220
},
{
"epoch": 1.1487979207277452,
"grad_norm": 1.6401057243347168,
"learning_rate": 4.034904144421134e-07,
"loss": 0.1256,
"step": 221
},
{
"epoch": 1.1539961013645224,
"grad_norm": 1.5419056415557861,
"learning_rate": 3.9935073995566987e-07,
"loss": 0.1302,
"step": 222
},
{
"epoch": 1.1591942820012995,
"grad_norm": 1.650795817375183,
"learning_rate": 3.952182437736256e-07,
"loss": 0.1471,
"step": 223
},
{
"epoch": 1.1643924626380766,
"grad_norm": 1.672743797302246,
"learning_rate": 3.910932206255742e-07,
"loss": 0.1298,
"step": 224
},
{
"epoch": 1.1695906432748537,
"grad_norm": 1.6902425289154053,
"learning_rate": 3.869759647081325e-07,
"loss": 0.1414,
"step": 225
},
{
"epoch": 1.1747888239116309,
"grad_norm": 1.607485055923462,
"learning_rate": 3.828667696639589e-07,
"loss": 0.1032,
"step": 226
},
{
"epoch": 1.179987004548408,
"grad_norm": 1.5336533784866333,
"learning_rate": 3.7876592856081e-07,
"loss": 0.1116,
"step": 227
},
{
"epoch": 1.1851851851851851,
"grad_norm": 1.5528396368026733,
"learning_rate": 3.7467373387063964e-07,
"loss": 0.1243,
"step": 228
},
{
"epoch": 1.1903833658219622,
"grad_norm": 1.743318796157837,
"learning_rate": 3.7059047744873955e-07,
"loss": 0.1437,
"step": 229
},
{
"epoch": 1.1955815464587394,
"grad_norm": 1.6484525203704834,
"learning_rate": 3.665164505129241e-07,
"loss": 0.131,
"step": 230
},
{
"epoch": 1.2007797270955165,
"grad_norm": 1.6531956195831299,
"learning_rate": 3.6245194362276094e-07,
"loss": 0.1268,
"step": 231
},
{
"epoch": 1.2059779077322936,
"grad_norm": 1.491297960281372,
"learning_rate": 3.5839724665884795e-07,
"loss": 0.1261,
"step": 232
},
{
"epoch": 1.2111760883690708,
"grad_norm": 1.5535728931427002,
"learning_rate": 3.5435264880213937e-07,
"loss": 0.1233,
"step": 233
},
{
"epoch": 1.2163742690058479,
"grad_norm": 1.6621695756912231,
"learning_rate": 3.50318438513321e-07,
"loss": 0.1331,
"step": 234
},
{
"epoch": 1.221572449642625,
"grad_norm": 1.5829371213912964,
"learning_rate": 3.462949035122376e-07,
"loss": 0.1229,
"step": 235
},
{
"epoch": 1.2267706302794021,
"grad_norm": 1.7693301439285278,
"learning_rate": 3.4228233075737223e-07,
"loss": 0.1434,
"step": 236
},
{
"epoch": 1.2319688109161793,
"grad_norm": 1.6113789081573486,
"learning_rate": 3.3828100642538093e-07,
"loss": 0.1213,
"step": 237
},
{
"epoch": 1.2371669915529564,
"grad_norm": 1.5799354314804077,
"learning_rate": 3.342912158906821e-07,
"loss": 0.1191,
"step": 238
},
{
"epoch": 1.2423651721897335,
"grad_norm": 1.5467253923416138,
"learning_rate": 3.3031324370510396e-07,
"loss": 0.1133,
"step": 239
},
{
"epoch": 1.2475633528265107,
"grad_norm": 1.8147982358932495,
"learning_rate": 3.263473735775899e-07,
"loss": 0.1391,
"step": 240
},
{
"epoch": 1.2527615334632878,
"grad_norm": 1.702359676361084,
"learning_rate": 3.2239388835396484e-07,
"loss": 0.1339,
"step": 241
},
{
"epoch": 1.257959714100065,
"grad_norm": 1.7504138946533203,
"learning_rate": 3.184530699967627e-07,
"loss": 0.1565,
"step": 242
},
{
"epoch": 1.263157894736842,
"grad_norm": 1.7226463556289673,
"learning_rate": 3.1452519956511614e-07,
"loss": 0.1266,
"step": 243
},
{
"epoch": 1.2683560753736192,
"grad_norm": 1.8186461925506592,
"learning_rate": 3.1061055719471197e-07,
"loss": 0.1347,
"step": 244
},
{
"epoch": 1.2735542560103963,
"grad_norm": 1.6384614706039429,
"learning_rate": 3.0670942207781204e-07,
"loss": 0.1115,
"step": 245
},
{
"epoch": 1.2787524366471734,
"grad_norm": 1.7369823455810547,
"learning_rate": 3.028220724433408e-07,
"loss": 0.129,
"step": 246
},
{
"epoch": 1.2839506172839505,
"grad_norm": 1.6780822277069092,
"learning_rate": 2.989487855370421e-07,
"loss": 0.1385,
"step": 247
},
{
"epoch": 1.2891487979207277,
"grad_norm": 1.87428879737854,
"learning_rate": 2.9508983760170634e-07,
"loss": 0.1435,
"step": 248
},
{
"epoch": 1.2943469785575048,
"grad_norm": 1.7940468788146973,
"learning_rate": 2.9124550385746856e-07,
"loss": 0.1491,
"step": 249
},
{
"epoch": 1.299545159194282,
"grad_norm": 1.712099552154541,
"learning_rate": 2.8741605848217976e-07,
"loss": 0.131,
"step": 250
},
{
"epoch": 1.304743339831059,
"grad_norm": 1.6154824495315552,
"learning_rate": 2.8360177459185263e-07,
"loss": 0.1145,
"step": 251
},
{
"epoch": 1.3099415204678362,
"grad_norm": 1.6371185779571533,
"learning_rate": 2.7980292422118277e-07,
"loss": 0.1232,
"step": 252
},
{
"epoch": 1.3151397011046133,
"grad_norm": 1.8156847953796387,
"learning_rate": 2.7601977830414766e-07,
"loss": 0.1274,
"step": 253
},
{
"epoch": 1.3203378817413904,
"grad_norm": 1.6596229076385498,
"learning_rate": 2.72252606654683e-07,
"loss": 0.1168,
"step": 254
},
{
"epoch": 1.3255360623781676,
"grad_norm": 1.6106423139572144,
"learning_rate": 2.685016779474396e-07,
"loss": 0.1139,
"step": 255
},
{
"epoch": 1.3307342430149447,
"grad_norm": 1.6363728046417236,
"learning_rate": 2.6476725969862226e-07,
"loss": 0.1297,
"step": 256
},
{
"epoch": 1.3359324236517218,
"grad_norm": 1.4978957176208496,
"learning_rate": 2.6104961824690964e-07,
"loss": 0.1191,
"step": 257
},
{
"epoch": 1.341130604288499,
"grad_norm": 1.5889379978179932,
"learning_rate": 2.5734901873445956e-07,
"loss": 0.1236,
"step": 258
},
{
"epoch": 1.346328784925276,
"grad_norm": 1.534178376197815,
"learning_rate": 2.536657250879988e-07,
"loss": 0.1053,
"step": 259
},
{
"epoch": 1.3515269655620532,
"grad_norm": 1.8409833908081055,
"learning_rate": 2.500000000000001e-07,
"loss": 0.1427,
"step": 260
},
{
"epoch": 1.3567251461988303,
"grad_norm": 1.7446589469909668,
"learning_rate": 2.4635210490994647e-07,
"loss": 0.1194,
"step": 261
},
{
"epoch": 1.3619233268356075,
"grad_norm": 1.7886688709259033,
"learning_rate": 2.427222999856857e-07,
"loss": 0.1351,
"step": 262
},
{
"epoch": 1.3671215074723846,
"grad_norm": 1.6462031602859497,
"learning_rate": 2.391108441048753e-07,
"loss": 0.1249,
"step": 263
},
{
"epoch": 1.3723196881091617,
"grad_norm": 1.8700019121170044,
"learning_rate": 2.355179948365189e-07,
"loss": 0.1482,
"step": 264
},
{
"epoch": 1.3775178687459388,
"grad_norm": 1.8244132995605469,
"learning_rate": 2.3194400842259687e-07,
"loss": 0.134,
"step": 265
},
{
"epoch": 1.382716049382716,
"grad_norm": 1.8189918994903564,
"learning_rate": 2.283891397597908e-07,
"loss": 0.1258,
"step": 266
},
{
"epoch": 1.387914230019493,
"grad_norm": 1.5552480220794678,
"learning_rate": 2.2485364238130433e-07,
"loss": 0.1131,
"step": 267
},
{
"epoch": 1.3931124106562702,
"grad_norm": 1.659328579902649,
"learning_rate": 2.2133776843878183e-07,
"loss": 0.1119,
"step": 268
},
{
"epoch": 1.3983105912930474,
"grad_norm": 1.3867595195770264,
"learning_rate": 2.1784176868432375e-07,
"loss": 0.0851,
"step": 269
},
{
"epoch": 1.4035087719298245,
"grad_norm": 1.7647099494934082,
"learning_rate": 2.1436589245260372e-07,
"loss": 0.1158,
"step": 270
},
{
"epoch": 1.4087069525666016,
"grad_norm": 1.8926544189453125,
"learning_rate": 2.109103876430864e-07,
"loss": 0.1527,
"step": 271
},
{
"epoch": 1.4139051332033787,
"grad_norm": 1.839390516281128,
"learning_rate": 2.074755007023461e-07,
"loss": 0.1108,
"step": 272
},
{
"epoch": 1.4191033138401559,
"grad_norm": 1.797500729560852,
"learning_rate": 2.040614766064913e-07,
"loss": 0.1508,
"step": 273
},
{
"epoch": 1.424301494476933,
"grad_norm": 1.7864497900009155,
"learning_rate": 2.0066855884369243e-07,
"loss": 0.1242,
"step": 274
},
{
"epoch": 1.4294996751137101,
"grad_norm": 1.853615641593933,
"learning_rate": 1.9729698939681644e-07,
"loss": 0.122,
"step": 275
},
{
"epoch": 1.4346978557504872,
"grad_norm": 1.6054631471633911,
"learning_rate": 1.9394700872616853e-07,
"loss": 0.1212,
"step": 276
},
{
"epoch": 1.4398960363872644,
"grad_norm": 1.632055640220642,
"learning_rate": 1.906188557523427e-07,
"loss": 0.1101,
"step": 277
},
{
"epoch": 1.4450942170240415,
"grad_norm": 1.6560664176940918,
"learning_rate": 1.873127678391816e-07,
"loss": 0.1217,
"step": 278
},
{
"epoch": 1.4502923976608186,
"grad_norm": 1.4159197807312012,
"learning_rate": 1.8402898077684803e-07,
"loss": 0.0988,
"step": 279
},
{
"epoch": 1.4554905782975958,
"grad_norm": 1.6314151287078857,
"learning_rate": 1.8076772876500828e-07,
"loss": 0.1293,
"step": 280
},
{
"epoch": 1.4606887589343729,
"grad_norm": 1.7430942058563232,
"learning_rate": 1.775292443961291e-07,
"loss": 0.1401,
"step": 281
},
{
"epoch": 1.46588693957115,
"grad_norm": 1.7857812643051147,
"learning_rate": 1.7431375863888898e-07,
"loss": 0.1275,
"step": 282
},
{
"epoch": 1.4710851202079271,
"grad_norm": 1.5308443307876587,
"learning_rate": 1.7112150082170568e-07,
"loss": 0.1061,
"step": 283
},
{
"epoch": 1.4762833008447043,
"grad_norm": 1.6288737058639526,
"learning_rate": 1.679526986163804e-07,
"loss": 0.1119,
"step": 284
},
{
"epoch": 1.4814814814814814,
"grad_norm": 1.704690933227539,
"learning_rate": 1.6480757802186068e-07,
"loss": 0.1166,
"step": 285
},
{
"epoch": 1.4866796621182585,
"grad_norm": 1.5033763647079468,
"learning_rate": 1.6168636334812125e-07,
"loss": 0.1045,
"step": 286
},
{
"epoch": 1.4918778427550357,
"grad_norm": 1.4401872158050537,
"learning_rate": 1.5858927720016706e-07,
"loss": 0.0959,
"step": 287
},
{
"epoch": 1.4970760233918128,
"grad_norm": 1.6706205606460571,
"learning_rate": 1.555165404621567e-07,
"loss": 0.1124,
"step": 288
},
{
"epoch": 1.50227420402859,
"grad_norm": 1.8483508825302124,
"learning_rate": 1.5246837228164905e-07,
"loss": 0.1146,
"step": 289
},
{
"epoch": 1.507472384665367,
"grad_norm": 1.8255398273468018,
"learning_rate": 1.494449900539737e-07,
"loss": 0.1413,
"step": 290
},
{
"epoch": 1.5126705653021442,
"grad_norm": 1.8132373094558716,
"learning_rate": 1.4644660940672627e-07,
"loss": 0.1316,
"step": 291
},
{
"epoch": 1.5178687459389213,
"grad_norm": 1.7580801248550415,
"learning_rate": 1.434734441843899e-07,
"loss": 0.1252,
"step": 292
},
{
"epoch": 1.5230669265756984,
"grad_norm": 1.4802451133728027,
"learning_rate": 1.4052570643308375e-07,
"loss": 0.1087,
"step": 293
},
{
"epoch": 1.5282651072124755,
"grad_norm": 1.595434308052063,
"learning_rate": 1.376036063854401e-07,
"loss": 0.1063,
"step": 294
},
{
"epoch": 1.5334632878492527,
"grad_norm": 1.5052251815795898,
"learning_rate": 1.3470735244561027e-07,
"loss": 0.1071,
"step": 295
},
{
"epoch": 1.5386614684860298,
"grad_norm": 1.4966932535171509,
"learning_rate": 1.3183715117440142e-07,
"loss": 0.1003,
"step": 296
},
{
"epoch": 1.543859649122807,
"grad_norm": 1.5577093362808228,
"learning_rate": 1.2899320727454472e-07,
"loss": 0.1147,
"step": 297
},
{
"epoch": 1.549057829759584,
"grad_norm": 2.081566572189331,
"learning_rate": 1.2617572357609562e-07,
"loss": 0.1479,
"step": 298
},
{
"epoch": 1.5542560103963612,
"grad_norm": 1.5348504781723022,
"learning_rate": 1.2338490102196825e-07,
"loss": 0.1061,
"step": 299
},
{
"epoch": 1.5594541910331383,
"grad_norm": 1.7641793489456177,
"learning_rate": 1.2062093865360457e-07,
"loss": 0.1359,
"step": 300
},
{
"epoch": 1.5646523716699154,
"grad_norm": 1.5747112035751343,
"learning_rate": 1.1788403359677767e-07,
"loss": 0.1069,
"step": 301
},
{
"epoch": 1.5698505523066926,
"grad_norm": 1.6333017349243164,
"learning_rate": 1.1517438104753385e-07,
"loss": 0.1077,
"step": 302
},
{
"epoch": 1.5750487329434697,
"grad_norm": 1.5666186809539795,
"learning_rate": 1.1249217425827062e-07,
"loss": 0.1118,
"step": 303
},
{
"epoch": 1.5802469135802468,
"grad_norm": 1.5051664113998413,
"learning_rate": 1.0983760452395413e-07,
"loss": 0.1043,
"step": 304
},
{
"epoch": 1.585445094217024,
"grad_norm": 1.7494423389434814,
"learning_rate": 1.07210861168476e-07,
"loss": 0.1327,
"step": 305
},
{
"epoch": 1.590643274853801,
"grad_norm": 1.5114164352416992,
"learning_rate": 1.0461213153115079e-07,
"loss": 0.0938,
"step": 306
},
{
"epoch": 1.5958414554905782,
"grad_norm": 1.562703251838684,
"learning_rate": 1.0204160095335479e-07,
"loss": 0.1056,
"step": 307
},
{
"epoch": 1.6010396361273553,
"grad_norm": 1.6739459037780762,
"learning_rate": 9.94994527653078e-08,
"loss": 0.1223,
"step": 308
},
{
"epoch": 1.6062378167641325,
"grad_norm": 1.6659190654754639,
"learning_rate": 9.69858682729976e-08,
"loss": 0.1201,
"step": 309
},
{
"epoch": 1.6114359974009096,
"grad_norm": 1.6198205947875977,
"learning_rate": 9.45010267452495e-08,
"loss": 0.1041,
"step": 310
},
{
"epoch": 1.6166341780376867,
"grad_norm": 1.6569786071777344,
"learning_rate": 9.204510540094095e-08,
"loss": 0.1153,
"step": 311
},
{
"epoch": 1.6218323586744638,
"grad_norm": 1.6825376749038696,
"learning_rate": 8.961827939636196e-08,
"loss": 0.1195,
"step": 312
},
{
"epoch": 1.627030539311241,
"grad_norm": 1.6692954301834106,
"learning_rate": 8.722072181272311e-08,
"loss": 0.14,
"step": 313
},
{
"epoch": 1.632228719948018,
"grad_norm": 1.86336350440979,
"learning_rate": 8.485260364381186e-08,
"loss": 0.1154,
"step": 314
},
{
"epoch": 1.6374269005847952,
"grad_norm": 1.6375104188919067,
"learning_rate": 8.251409378379637e-08,
"loss": 0.1087,
"step": 315
},
{
"epoch": 1.6426250812215724,
"grad_norm": 1.6260446310043335,
"learning_rate": 8.02053590151805e-08,
"loss": 0.1099,
"step": 316
},
{
"epoch": 1.6478232618583495,
"grad_norm": 1.7146011590957642,
"learning_rate": 7.792656399690922e-08,
"loss": 0.1167,
"step": 317
},
{
"epoch": 1.6530214424951266,
"grad_norm": 1.6423556804656982,
"learning_rate": 7.567787125262449e-08,
"loss": 0.1171,
"step": 318
},
{
"epoch": 1.6582196231319037,
"grad_norm": 1.5510940551757812,
"learning_rate": 7.345944115907421e-08,
"loss": 0.1013,
"step": 319
},
{
"epoch": 1.6634178037686809,
"grad_norm": 1.863503336906433,
"learning_rate": 7.127143193467445e-08,
"loss": 0.1423,
"step": 320
},
{
"epoch": 1.668615984405458,
"grad_norm": 1.7284936904907227,
"learning_rate": 6.911399962822518e-08,
"loss": 0.1112,
"step": 321
},
{
"epoch": 1.6738141650422351,
"grad_norm": 1.7205549478530884,
"learning_rate": 6.698729810778064e-08,
"loss": 0.136,
"step": 322
},
{
"epoch": 1.6790123456790123,
"grad_norm": 1.4762628078460693,
"learning_rate": 6.48914790496759e-08,
"loss": 0.1144,
"step": 323
},
{
"epoch": 1.6842105263157894,
"grad_norm": 1.8362925052642822,
"learning_rate": 6.282669192770895e-08,
"loss": 0.1328,
"step": 324
},
{
"epoch": 1.6894087069525665,
"grad_norm": 1.4527249336242676,
"learning_rate": 6.079308400248029e-08,
"loss": 0.1055,
"step": 325
},
{
"epoch": 1.6946068875893436,
"grad_norm": 1.7803164720535278,
"learning_rate": 5.8790800310890456e-08,
"loss": 0.1451,
"step": 326
},
{
"epoch": 1.6998050682261208,
"grad_norm": 2.059589147567749,
"learning_rate": 5.6819983655795936e-08,
"loss": 0.1458,
"step": 327
},
{
"epoch": 1.705003248862898,
"grad_norm": 1.7053784132003784,
"learning_rate": 5.4880774595824245e-08,
"loss": 0.1257,
"step": 328
},
{
"epoch": 1.710201429499675,
"grad_norm": 1.97287118434906,
"learning_rate": 5.297331143534972e-08,
"loss": 0.1381,
"step": 329
},
{
"epoch": 1.7153996101364521,
"grad_norm": 1.6465667486190796,
"learning_rate": 5.109773021462921e-08,
"loss": 0.1155,
"step": 330
},
{
"epoch": 1.7205977907732293,
"grad_norm": 1.6381739377975464,
"learning_rate": 4.925416470009991e-08,
"loss": 0.1224,
"step": 331
},
{
"epoch": 1.7257959714100064,
"grad_norm": 1.5278061628341675,
"learning_rate": 4.744274637483936e-08,
"loss": 0.1203,
"step": 332
},
{
"epoch": 1.7309941520467835,
"grad_norm": 1.7712794542312622,
"learning_rate": 4.566360442918754e-08,
"loss": 0.1334,
"step": 333
},
{
"epoch": 1.7361923326835607,
"grad_norm": 1.912316083908081,
"learning_rate": 4.3916865751533306e-08,
"loss": 0.1475,
"step": 334
},
{
"epoch": 1.7413905133203378,
"grad_norm": 1.7500931024551392,
"learning_rate": 4.220265491926489e-08,
"loss": 0.1165,
"step": 335
},
{
"epoch": 1.746588693957115,
"grad_norm": 1.7582532167434692,
"learning_rate": 4.0521094189884696e-08,
"loss": 0.119,
"step": 336
},
{
"epoch": 1.751786874593892,
"grad_norm": 1.6794097423553467,
"learning_rate": 3.887230349229015e-08,
"loss": 0.1093,
"step": 337
},
{
"epoch": 1.7569850552306692,
"grad_norm": 1.797913908958435,
"learning_rate": 3.7256400418220256e-08,
"loss": 0.1235,
"step": 338
},
{
"epoch": 1.7621832358674463,
"grad_norm": 1.6042097806930542,
"learning_rate": 3.567350021386895e-08,
"loss": 0.096,
"step": 339
},
{
"epoch": 1.7673814165042234,
"grad_norm": 1.7623355388641357,
"learning_rate": 3.412371577166578e-08,
"loss": 0.1153,
"step": 340
},
{
"epoch": 1.7725795971410006,
"grad_norm": 1.737052321434021,
"learning_rate": 3.260715762222449e-08,
"loss": 0.1327,
"step": 341
},
{
"epoch": 1.7777777777777777,
"grad_norm": 1.61103355884552,
"learning_rate": 3.1123933926459845e-08,
"loss": 0.1213,
"step": 342
},
{
"epoch": 1.7829759584145548,
"grad_norm": 1.78020441532135,
"learning_rate": 2.9674150467873527e-08,
"loss": 0.1262,
"step": 343
},
{
"epoch": 1.788174139051332,
"grad_norm": 1.7625352144241333,
"learning_rate": 2.825791064500993e-08,
"loss": 0.1465,
"step": 344
},
{
"epoch": 1.793372319688109,
"grad_norm": 1.7474111318588257,
"learning_rate": 2.6875315464081562e-08,
"loss": 0.1289,
"step": 345
},
{
"epoch": 1.7985705003248862,
"grad_norm": 1.7776579856872559,
"learning_rate": 2.5526463531765463e-08,
"loss": 0.1219,
"step": 346
},
{
"epoch": 1.8037686809616633,
"grad_norm": 1.7461159229278564,
"learning_rate": 2.4211451048170296e-08,
"loss": 0.1361,
"step": 347
},
{
"epoch": 1.8089668615984404,
"grad_norm": 1.701060175895691,
"learning_rate": 2.293037179997559e-08,
"loss": 0.1463,
"step": 348
},
{
"epoch": 1.8141650422352176,
"grad_norm": 1.7974237203598022,
"learning_rate": 2.1683317153742775e-08,
"loss": 0.14,
"step": 349
},
{
"epoch": 1.8193632228719947,
"grad_norm": 1.5575218200683594,
"learning_rate": 2.047037604939894e-08,
"loss": 0.1234,
"step": 350
},
{
"epoch": 1.8245614035087718,
"grad_norm": 1.6900442838668823,
"learning_rate": 1.92916349938938e-08,
"loss": 0.1119,
"step": 351
},
{
"epoch": 1.829759584145549,
"grad_norm": 1.7725763320922852,
"learning_rate": 1.8147178055029577e-08,
"loss": 0.1456,
"step": 352
},
{
"epoch": 1.834957764782326,
"grad_norm": 1.677262544631958,
"learning_rate": 1.7037086855465898e-08,
"loss": 0.1041,
"step": 353
},
{
"epoch": 1.8401559454191032,
"grad_norm": 1.5990703105926514,
"learning_rate": 1.596144056689791e-08,
"loss": 0.1149,
"step": 354
},
{
"epoch": 1.8453541260558803,
"grad_norm": 2.083341598510742,
"learning_rate": 1.4920315904410064e-08,
"loss": 0.1375,
"step": 355
},
{
"epoch": 1.8505523066926575,
"grad_norm": 1.7379205226898193,
"learning_rate": 1.3913787121004716e-08,
"loss": 0.1365,
"step": 356
},
{
"epoch": 1.8557504873294346,
"grad_norm": 1.6175756454467773,
"learning_rate": 1.2941926002306536e-08,
"loss": 0.1348,
"step": 357
},
{
"epoch": 1.8609486679662117,
"grad_norm": 1.7680426836013794,
"learning_rate": 1.200480186144237e-08,
"loss": 0.1232,
"step": 358
},
{
"epoch": 1.8661468486029889,
"grad_norm": 1.3798372745513916,
"learning_rate": 1.1102481534098374e-08,
"loss": 0.0966,
"step": 359
},
{
"epoch": 1.871345029239766,
"grad_norm": 1.584373950958252,
"learning_rate": 1.0235029373752757e-08,
"loss": 0.1082,
"step": 360
},
{
"epoch": 1.876543209876543,
"grad_norm": 1.7235833406448364,
"learning_rate": 9.402507247086578e-09,
"loss": 0.1348,
"step": 361
},
{
"epoch": 1.8817413905133202,
"grad_norm": 1.6077182292938232,
"learning_rate": 8.60497452957104e-09,
"loss": 0.1234,
"step": 362
},
{
"epoch": 1.8869395711500974,
"grad_norm": 1.5776941776275635,
"learning_rate": 7.842488101232891e-09,
"loss": 0.1177,
"step": 363
},
{
"epoch": 1.8921377517868745,
"grad_norm": 1.5909714698791504,
"learning_rate": 7.115102342598101e-09,
"loss": 0.1133,
"step": 364
},
{
"epoch": 1.8973359324236516,
"grad_norm": 1.7370619773864746,
"learning_rate": 6.422869130812913e-09,
"loss": 0.1269,
"step": 365
},
{
"epoch": 1.9025341130604287,
"grad_norm": 1.7405683994293213,
"learning_rate": 5.765837835944309e-09,
"loss": 0.1307,
"step": 366
},
{
"epoch": 1.9077322936972059,
"grad_norm": 1.9631272554397583,
"learning_rate": 5.144055317458817e-09,
"loss": 0.1546,
"step": 367
},
{
"epoch": 1.912930474333983,
"grad_norm": 1.6291249990463257,
"learning_rate": 4.55756592088058e-09,
"loss": 0.1249,
"step": 368
},
{
"epoch": 1.9181286549707601,
"grad_norm": 1.528748869895935,
"learning_rate": 4.0064114746284905e-09,
"loss": 0.1027,
"step": 369
},
{
"epoch": 1.9233268356075373,
"grad_norm": 1.8311703205108643,
"learning_rate": 3.4906312870331965e-09,
"loss": 0.1128,
"step": 370
},
{
"epoch": 1.9285250162443144,
"grad_norm": 1.8863190412521362,
"learning_rate": 3.010262143533393e-09,
"loss": 0.1441,
"step": 371
},
{
"epoch": 1.9337231968810915,
"grad_norm": 1.5785506963729858,
"learning_rate": 2.5653383040524224e-09,
"loss": 0.1198,
"step": 372
},
{
"epoch": 1.9389213775178686,
"grad_norm": 1.6531542539596558,
"learning_rate": 2.155891500554896e-09,
"loss": 0.1224,
"step": 373
},
{
"epoch": 1.9441195581546458,
"grad_norm": 1.5542500019073486,
"learning_rate": 1.7819509347835049e-09,
"loss": 0.1149,
"step": 374
},
{
"epoch": 1.949317738791423,
"grad_norm": 1.5082885026931763,
"learning_rate": 1.4435432761762955e-09,
"loss": 0.1061,
"step": 375
},
{
"epoch": 1.9545159194282,
"grad_norm": 1.5856765508651733,
"learning_rate": 1.1406926599646372e-09,
"loss": 0.1327,
"step": 376
},
{
"epoch": 1.9597141000649771,
"grad_norm": 1.6566141843795776,
"learning_rate": 8.73420685452042e-10,
"loss": 0.1105,
"step": 377
},
{
"epoch": 1.9649122807017543,
"grad_norm": 1.7047754526138306,
"learning_rate": 6.417464144736207e-10,
"loss": 0.1301,
"step": 378
},
{
"epoch": 1.9701104613385314,
"grad_norm": 1.7699244022369385,
"learning_rate": 4.4568637003633556e-10,
"loss": 0.1412,
"step": 379
},
{
"epoch": 1.9753086419753085,
"grad_norm": 1.8177380561828613,
"learning_rate": 2.852545351409996e-10,
"loss": 0.1571,
"step": 380
},
{
"epoch": 1.9805068226120857,
"grad_norm": 1.5334748029708862,
"learning_rate": 1.6046235178474033e-10,
"loss": 0.104,
"step": 381
},
{
"epoch": 1.9857050032488628,
"grad_norm": 1.7260781526565552,
"learning_rate": 7.13187201450971e-11,
"loss": 0.1263,
"step": 382
},
{
"epoch": 1.99090318388564,
"grad_norm": 1.6607623100280762,
"learning_rate": 1.7829997945084662e-11,
"loss": 0.1135,
"step": 383
},
{
"epoch": 1.996101364522417,
"grad_norm": 1.7004939317703247,
"learning_rate": 0.0,
"loss": 0.1315,
"step": 384
},
{
"epoch": 1.996101364522417,
"step": 384,
"total_flos": 3.0550374682743276e+18,
"train_loss": 0.19105235013800362,
"train_runtime": 6828.3197,
"train_samples_per_second": 7.211,
"train_steps_per_second": 0.056
}
],
"logging_steps": 1.0,
"max_steps": 384,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.0550374682743276e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}