FMU-Agent-lora / trainer_state.json
wolfvoid's picture
upload lora ckpts
5da71ac verified
Raw
History Blame Contribute Delete
53 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 2997,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.010012515644555695,
"grad_norm": 6.284793376922607,
"learning_rate": 5e-06,
"loss": 3.2507,
"step": 10
},
{
"epoch": 0.02002503128911139,
"grad_norm": 4.173284530639648,
"learning_rate": 1.0555555555555555e-05,
"loss": 2.8994,
"step": 20
},
{
"epoch": 0.030037546933667083,
"grad_norm": 2.0444984436035156,
"learning_rate": 1.6111111111111115e-05,
"loss": 1.8864,
"step": 30
},
{
"epoch": 0.04005006257822278,
"grad_norm": 0.7631794214248657,
"learning_rate": 2.1666666666666667e-05,
"loss": 1.1103,
"step": 40
},
{
"epoch": 0.05006257822277847,
"grad_norm": 0.47767218947410583,
"learning_rate": 2.7222222222222223e-05,
"loss": 0.8778,
"step": 50
},
{
"epoch": 0.060075093867334166,
"grad_norm": 0.3519922196865082,
"learning_rate": 3.277777777777778e-05,
"loss": 0.8132,
"step": 60
},
{
"epoch": 0.07008760951188986,
"grad_norm": 0.39849111437797546,
"learning_rate": 3.8333333333333334e-05,
"loss": 0.7614,
"step": 70
},
{
"epoch": 0.08010012515644556,
"grad_norm": 0.304582417011261,
"learning_rate": 4.388888888888889e-05,
"loss": 0.7313,
"step": 80
},
{
"epoch": 0.09011264080100125,
"grad_norm": 0.5032485723495483,
"learning_rate": 4.9444444444444446e-05,
"loss": 0.7115,
"step": 90
},
{
"epoch": 0.10012515644555695,
"grad_norm": 0.38174888491630554,
"learning_rate": 4.999881749961716e-05,
"loss": 0.6927,
"step": 100
},
{
"epoch": 0.11013767209011265,
"grad_norm": 0.3855612874031067,
"learning_rate": 4.999472998758978e-05,
"loss": 0.6787,
"step": 110
},
{
"epoch": 0.12015018773466833,
"grad_norm": 0.5459920763969421,
"learning_rate": 4.9987723342414355e-05,
"loss": 0.6812,
"step": 120
},
{
"epoch": 0.13016270337922403,
"grad_norm": 0.40923094749450684,
"learning_rate": 4.997779838239658e-05,
"loss": 0.6728,
"step": 130
},
{
"epoch": 0.14017521902377972,
"grad_norm": 0.4264446794986725,
"learning_rate": 4.9964956266671986e-05,
"loss": 0.6652,
"step": 140
},
{
"epoch": 0.15018773466833543,
"grad_norm": 0.4102150499820709,
"learning_rate": 4.994919849507058e-05,
"loss": 0.6659,
"step": 150
},
{
"epoch": 0.16020025031289112,
"grad_norm": 1.234883189201355,
"learning_rate": 4.993052690794165e-05,
"loss": 0.6618,
"step": 160
},
{
"epoch": 0.1702127659574468,
"grad_norm": 0.7370510101318359,
"learning_rate": 4.9908943685938804e-05,
"loss": 0.6612,
"step": 170
},
{
"epoch": 0.1802252816020025,
"grad_norm": 0.41938677430152893,
"learning_rate": 4.9884451349765395e-05,
"loss": 0.6574,
"step": 180
},
{
"epoch": 0.1902377972465582,
"grad_norm": 0.425245076417923,
"learning_rate": 4.985705275988001e-05,
"loss": 0.6521,
"step": 190
},
{
"epoch": 0.2002503128911139,
"grad_norm": 0.4455544352531433,
"learning_rate": 4.982675111616246e-05,
"loss": 0.6454,
"step": 200
},
{
"epoch": 0.21026282853566958,
"grad_norm": 0.45250239968299866,
"learning_rate": 4.979354995754006e-05,
"loss": 0.6613,
"step": 210
},
{
"epoch": 0.2202753441802253,
"grad_norm": 0.39480283856391907,
"learning_rate": 4.975745316157431e-05,
"loss": 0.6358,
"step": 220
},
{
"epoch": 0.23028785982478098,
"grad_norm": 0.40793269872665405,
"learning_rate": 4.9718464944008045e-05,
"loss": 0.623,
"step": 230
},
{
"epoch": 0.24030037546933666,
"grad_norm": 0.437458872795105,
"learning_rate": 4.967658985827305e-05,
"loss": 0.644,
"step": 240
},
{
"epoch": 0.2503128911138924,
"grad_norm": 0.44151055812835693,
"learning_rate": 4.963183279495829e-05,
"loss": 0.6352,
"step": 250
},
{
"epoch": 0.26032540675844806,
"grad_norm": 0.4475908875465393,
"learning_rate": 4.9584198981238765e-05,
"loss": 0.6372,
"step": 260
},
{
"epoch": 0.27033792240300375,
"grad_norm": 0.47630321979522705,
"learning_rate": 4.953369398026496e-05,
"loss": 0.6312,
"step": 270
},
{
"epoch": 0.28035043804755944,
"grad_norm": 0.4477619528770447,
"learning_rate": 4.9480323690513175e-05,
"loss": 0.6228,
"step": 280
},
{
"epoch": 0.2903629536921151,
"grad_norm": 0.46181827783584595,
"learning_rate": 4.942409434509663e-05,
"loss": 0.6285,
"step": 290
},
{
"epoch": 0.30037546933667086,
"grad_norm": 0.5095700025558472,
"learning_rate": 4.9365012511037514e-05,
"loss": 0.6224,
"step": 300
},
{
"epoch": 0.31038798498122655,
"grad_norm": 0.524527907371521,
"learning_rate": 4.9303085088499965e-05,
"loss": 0.6127,
"step": 310
},
{
"epoch": 0.32040050062578224,
"grad_norm": 0.4272041618824005,
"learning_rate": 4.9238319309984315e-05,
"loss": 0.6048,
"step": 320
},
{
"epoch": 0.3304130162703379,
"grad_norm": 0.4468429982662201,
"learning_rate": 4.917072273948226e-05,
"loss": 0.6194,
"step": 330
},
{
"epoch": 0.3404255319148936,
"grad_norm": 0.4301135241985321,
"learning_rate": 4.910030327159359e-05,
"loss": 0.6204,
"step": 340
},
{
"epoch": 0.3504380475594493,
"grad_norm": 0.4218313694000244,
"learning_rate": 4.90270691306041e-05,
"loss": 0.6055,
"step": 350
},
{
"epoch": 0.360450563204005,
"grad_norm": 0.44785287976264954,
"learning_rate": 4.8951028869525104e-05,
"loss": 0.6188,
"step": 360
},
{
"epoch": 0.3704630788485607,
"grad_norm": 0.463697612285614,
"learning_rate": 4.887219136909456e-05,
"loss": 0.6103,
"step": 370
},
{
"epoch": 0.3804755944931164,
"grad_norm": 0.4042550325393677,
"learning_rate": 4.87905658367398e-05,
"loss": 0.6058,
"step": 380
},
{
"epoch": 0.3904881101376721,
"grad_norm": 0.40829843282699585,
"learning_rate": 4.8706161805502296e-05,
"loss": 0.6083,
"step": 390
},
{
"epoch": 0.4005006257822278,
"grad_norm": 0.4917544722557068,
"learning_rate": 4.861898913292424e-05,
"loss": 0.5929,
"step": 400
},
{
"epoch": 0.41051314142678347,
"grad_norm": 0.42198094725608826,
"learning_rate": 4.85290579998973e-05,
"loss": 0.5978,
"step": 410
},
{
"epoch": 0.42052565707133915,
"grad_norm": 0.4416494369506836,
"learning_rate": 4.8436378909473576e-05,
"loss": 0.6046,
"step": 420
},
{
"epoch": 0.4305381727158949,
"grad_norm": 0.43621471524238586,
"learning_rate": 4.8340962685638964e-05,
"loss": 0.6145,
"step": 430
},
{
"epoch": 0.4405506883604506,
"grad_norm": 0.4515362083911896,
"learning_rate": 4.8242820472049035e-05,
"loss": 0.6075,
"step": 440
},
{
"epoch": 0.45056320400500627,
"grad_norm": 0.44357553124427795,
"learning_rate": 4.814196373072754e-05,
"loss": 0.6098,
"step": 450
},
{
"epoch": 0.46057571964956195,
"grad_norm": 0.4832551181316376,
"learning_rate": 4.803840424072779e-05,
"loss": 0.6004,
"step": 460
},
{
"epoch": 0.47058823529411764,
"grad_norm": 0.44098058342933655,
"learning_rate": 4.793215409675694e-05,
"loss": 0.6031,
"step": 470
},
{
"epoch": 0.4806007509386733,
"grad_norm": 0.45025014877319336,
"learning_rate": 4.782322570776352e-05,
"loss": 0.6068,
"step": 480
},
{
"epoch": 0.490613266583229,
"grad_norm": 0.4705432653427124,
"learning_rate": 4.7711631795488096e-05,
"loss": 0.5877,
"step": 490
},
{
"epoch": 0.5006257822277848,
"grad_norm": 0.4505113661289215,
"learning_rate": 4.759738539297759e-05,
"loss": 0.5844,
"step": 500
},
{
"epoch": 0.5106382978723404,
"grad_norm": 0.46183329820632935,
"learning_rate": 4.74804998430631e-05,
"loss": 0.5902,
"step": 510
},
{
"epoch": 0.5206508135168961,
"grad_norm": 0.47073203325271606,
"learning_rate": 4.7360988796801606e-05,
"loss": 0.6057,
"step": 520
},
{
"epoch": 0.5306633291614519,
"grad_norm": 0.46323421597480774,
"learning_rate": 4.723886621188167e-05,
"loss": 0.593,
"step": 530
},
{
"epoch": 0.5406758448060075,
"grad_norm": 0.4409252107143402,
"learning_rate": 4.7114146350993305e-05,
"loss": 0.5887,
"step": 540
},
{
"epoch": 0.5506883604505632,
"grad_norm": 0.4844597280025482,
"learning_rate": 4.698684378016222e-05,
"loss": 0.6012,
"step": 550
},
{
"epoch": 0.5607008760951189,
"grad_norm": 0.45070862770080566,
"learning_rate": 4.685697336704874e-05,
"loss": 0.5959,
"step": 560
},
{
"epoch": 0.5707133917396746,
"grad_norm": 0.4574355483055115,
"learning_rate": 4.6724550279211265e-05,
"loss": 0.6002,
"step": 570
},
{
"epoch": 0.5807259073842302,
"grad_norm": 0.534294068813324,
"learning_rate": 4.658958998233501e-05,
"loss": 0.59,
"step": 580
},
{
"epoch": 0.590738423028786,
"grad_norm": 0.9012548327445984,
"learning_rate": 4.645210823842565e-05,
"loss": 0.5918,
"step": 590
},
{
"epoch": 0.6007509386733417,
"grad_norm": 0.4650445282459259,
"learning_rate": 4.6312121103968555e-05,
"loss": 0.5913,
"step": 600
},
{
"epoch": 0.6107634543178974,
"grad_norm": 0.4305174946784973,
"learning_rate": 4.616964492805347e-05,
"loss": 0.5889,
"step": 610
},
{
"epoch": 0.6207759699624531,
"grad_norm": 0.4870530366897583,
"learning_rate": 4.602469635046521e-05,
"loss": 0.5747,
"step": 620
},
{
"epoch": 0.6307884856070087,
"grad_norm": 0.43425261974334717,
"learning_rate": 4.587729229974021e-05,
"loss": 0.6031,
"step": 630
},
{
"epoch": 0.6408010012515645,
"grad_norm": 0.46597158908843994,
"learning_rate": 4.572744999118951e-05,
"loss": 0.5903,
"step": 640
},
{
"epoch": 0.6508135168961201,
"grad_norm": 0.4396601915359497,
"learning_rate": 4.5575186924888115e-05,
"loss": 0.5937,
"step": 650
},
{
"epoch": 0.6608260325406758,
"grad_norm": 0.41379213333129883,
"learning_rate": 4.54205208836312e-05,
"loss": 0.5817,
"step": 660
},
{
"epoch": 0.6708385481852316,
"grad_norm": 0.45009931921958923,
"learning_rate": 4.5263469930857286e-05,
"loss": 0.5814,
"step": 670
},
{
"epoch": 0.6808510638297872,
"grad_norm": 0.4549204111099243,
"learning_rate": 4.510405240853854e-05,
"loss": 0.576,
"step": 680
},
{
"epoch": 0.690863579474343,
"grad_norm": 0.44143491983413696,
"learning_rate": 4.494228693503869e-05,
"loss": 0.5813,
"step": 690
},
{
"epoch": 0.7008760951188986,
"grad_norm": 0.4684794247150421,
"learning_rate": 4.477819240293852e-05,
"loss": 0.5751,
"step": 700
},
{
"epoch": 0.7108886107634543,
"grad_norm": 0.431362122297287,
"learning_rate": 4.461178797682947e-05,
"loss": 0.573,
"step": 710
},
{
"epoch": 0.72090112640801,
"grad_norm": 0.43078383803367615,
"learning_rate": 4.444309309107535e-05,
"loss": 0.5736,
"step": 720
},
{
"epoch": 0.7309136420525657,
"grad_norm": 0.4367680251598358,
"learning_rate": 4.427212744754266e-05,
"loss": 0.5735,
"step": 730
},
{
"epoch": 0.7409261576971214,
"grad_norm": 0.49401211738586426,
"learning_rate": 4.409891101329953e-05,
"loss": 0.5789,
"step": 740
},
{
"epoch": 0.7509386733416771,
"grad_norm": 0.4479840099811554,
"learning_rate": 4.3923464018283866e-05,
"loss": 0.5732,
"step": 750
},
{
"epoch": 0.7609511889862328,
"grad_norm": 0.5013871192932129,
"learning_rate": 4.374580695294061e-05,
"loss": 0.5593,
"step": 760
},
{
"epoch": 0.7709637046307884,
"grad_norm": 0.45542237162590027,
"learning_rate": 4.356596056582869e-05,
"loss": 0.5868,
"step": 770
},
{
"epoch": 0.7809762202753442,
"grad_norm": 0.46875742077827454,
"learning_rate": 4.3383945861197845e-05,
"loss": 0.5846,
"step": 780
},
{
"epoch": 0.7909887359198998,
"grad_norm": 0.45340389013290405,
"learning_rate": 4.319978409653545e-05,
"loss": 0.5745,
"step": 790
},
{
"epoch": 0.8010012515644556,
"grad_norm": 0.49433958530426025,
"learning_rate": 4.301349678008395e-05,
"loss": 0.5661,
"step": 800
},
{
"epoch": 0.8110137672090113,
"grad_norm": 0.4294925332069397,
"learning_rate": 4.282510566832883e-05,
"loss": 0.5719,
"step": 810
},
{
"epoch": 0.8210262828535669,
"grad_norm": 0.43871548771858215,
"learning_rate": 4.263463276345772e-05,
"loss": 0.5745,
"step": 820
},
{
"epoch": 0.8310387984981227,
"grad_norm": 0.4298272728919983,
"learning_rate": 4.244210031079077e-05,
"loss": 0.5721,
"step": 830
},
{
"epoch": 0.8410513141426783,
"grad_norm": 0.4713990092277527,
"learning_rate": 4.22475307961826e-05,
"loss": 0.585,
"step": 840
},
{
"epoch": 0.851063829787234,
"grad_norm": 0.45330944657325745,
"learning_rate": 4.205094694339617e-05,
"loss": 0.5705,
"step": 850
},
{
"epoch": 0.8610763454317898,
"grad_norm": 0.44769591093063354,
"learning_rate": 4.18523717114489e-05,
"loss": 0.5634,
"step": 860
},
{
"epoch": 0.8710888610763454,
"grad_norm": 0.43921416997909546,
"learning_rate": 4.1651828291931264e-05,
"loss": 0.5708,
"step": 870
},
{
"epoch": 0.8811013767209012,
"grad_norm": 0.45106759667396545,
"learning_rate": 4.144934010629825e-05,
"loss": 0.568,
"step": 880
},
{
"epoch": 0.8911138923654568,
"grad_norm": 0.4386798143386841,
"learning_rate": 4.1244930803134e-05,
"loss": 0.5694,
"step": 890
},
{
"epoch": 0.9011264080100125,
"grad_norm": 0.5129058957099915,
"learning_rate": 4.103862425538982e-05,
"loss": 0.5733,
"step": 900
},
{
"epoch": 0.9111389236545682,
"grad_norm": 8.963372230529785,
"learning_rate": 4.083044455759617e-05,
"loss": 0.5679,
"step": 910
},
{
"epoch": 0.9211514392991239,
"grad_norm": 0.45170068740844727,
"learning_rate": 4.062041602304856e-05,
"loss": 0.5622,
"step": 920
},
{
"epoch": 0.9311639549436797,
"grad_norm": 0.4376923143863678,
"learning_rate": 4.040856318096804e-05,
"loss": 0.5705,
"step": 930
},
{
"epoch": 0.9411764705882353,
"grad_norm": 0.45813918113708496,
"learning_rate": 4.019491077363644e-05,
"loss": 0.5599,
"step": 940
},
{
"epoch": 0.951188986232791,
"grad_norm": 0.49614930152893066,
"learning_rate": 3.997948375350675e-05,
"loss": 0.5798,
"step": 950
},
{
"epoch": 0.9612015018773467,
"grad_norm": 0.4715845286846161,
"learning_rate": 3.9762307280288863e-05,
"loss": 0.583,
"step": 960
},
{
"epoch": 0.9712140175219024,
"grad_norm": 0.4904944598674774,
"learning_rate": 3.9543406718011196e-05,
"loss": 0.5738,
"step": 970
},
{
"epoch": 0.981226533166458,
"grad_norm": 0.4361754059791565,
"learning_rate": 3.9322807632058447e-05,
"loss": 0.5498,
"step": 980
},
{
"epoch": 0.9912390488110138,
"grad_norm": 0.46598365902900696,
"learning_rate": 3.910053578618579e-05,
"loss": 0.569,
"step": 990
},
{
"epoch": 1.0010012515644555,
"grad_norm": 0.44678816199302673,
"learning_rate": 3.8876617139509965e-05,
"loss": 0.5683,
"step": 1000
},
{
"epoch": 1.0110137672090114,
"grad_norm": 0.42059096693992615,
"learning_rate": 3.865107784347746e-05,
"loss": 0.5392,
"step": 1010
},
{
"epoch": 1.021026282853567,
"grad_norm": 0.4724879562854767,
"learning_rate": 3.842394423881031e-05,
"loss": 0.5519,
"step": 1020
},
{
"epoch": 1.0310387984981226,
"grad_norm": 0.47808629274368286,
"learning_rate": 3.81952428524298e-05,
"loss": 0.5294,
"step": 1030
},
{
"epoch": 1.0410513141426783,
"grad_norm": 0.4879685044288635,
"learning_rate": 3.796500039435832e-05,
"loss": 0.5258,
"step": 1040
},
{
"epoch": 1.0510638297872341,
"grad_norm": 0.4844951927661896,
"learning_rate": 3.773324375459996e-05,
"loss": 0.5237,
"step": 1050
},
{
"epoch": 1.0610763454317897,
"grad_norm": 0.5054337978363037,
"learning_rate": 3.7500000000000003e-05,
"loss": 0.5485,
"step": 1060
},
{
"epoch": 1.0710888610763454,
"grad_norm": 0.49947911500930786,
"learning_rate": 3.726529637108379e-05,
"loss": 0.5381,
"step": 1070
},
{
"epoch": 1.0811013767209012,
"grad_norm": 0.5068179965019226,
"learning_rate": 3.70291602788753e-05,
"loss": 0.5382,
"step": 1080
},
{
"epoch": 1.0911138923654569,
"grad_norm": 0.5623748898506165,
"learning_rate": 3.679161930169582e-05,
"loss": 0.5402,
"step": 1090
},
{
"epoch": 1.1011264080100125,
"grad_norm": 0.4831792712211609,
"learning_rate": 3.655270118194311e-05,
"loss": 0.5314,
"step": 1100
},
{
"epoch": 1.1111389236545681,
"grad_norm": 0.46407610177993774,
"learning_rate": 3.6312433822851266e-05,
"loss": 0.5335,
"step": 1110
},
{
"epoch": 1.121151439299124,
"grad_norm": 0.4935677647590637,
"learning_rate": 3.607084528523203e-05,
"loss": 0.5337,
"step": 1120
},
{
"epoch": 1.1311639549436796,
"grad_norm": 0.48637017607688904,
"learning_rate": 3.5827963784197524e-05,
"loss": 0.5365,
"step": 1130
},
{
"epoch": 1.1411764705882352,
"grad_norm": 0.5110608339309692,
"learning_rate": 3.558381768586493e-05,
"loss": 0.5252,
"step": 1140
},
{
"epoch": 1.1511889862327909,
"grad_norm": 0.4810512065887451,
"learning_rate": 3.533843550404373e-05,
"loss": 0.5433,
"step": 1150
},
{
"epoch": 1.1612015018773467,
"grad_norm": 0.5119523406028748,
"learning_rate": 3.509184589690553e-05,
"loss": 0.5373,
"step": 1160
},
{
"epoch": 1.1712140175219024,
"grad_norm": 0.4847480058670044,
"learning_rate": 3.484407766363704e-05,
"loss": 0.5502,
"step": 1170
},
{
"epoch": 1.181226533166458,
"grad_norm": 0.48766791820526123,
"learning_rate": 3.459515974107667e-05,
"loss": 0.5511,
"step": 1180
},
{
"epoch": 1.1912390488110138,
"grad_norm": 0.4738394618034363,
"learning_rate": 3.4345121200335e-05,
"loss": 0.5308,
"step": 1190
},
{
"epoch": 1.2012515644555695,
"grad_norm": 0.5227957367897034,
"learning_rate": 3.4093991243399516e-05,
"loss": 0.5339,
"step": 1200
},
{
"epoch": 1.211264080100125,
"grad_norm": 0.4802178740501404,
"learning_rate": 3.3841799199724144e-05,
"loss": 0.5322,
"step": 1210
},
{
"epoch": 1.2212765957446807,
"grad_norm": 0.4954160451889038,
"learning_rate": 3.358857452280388e-05,
"loss": 0.5381,
"step": 1220
},
{
"epoch": 1.2312891113892366,
"grad_norm": 0.4919920563697815,
"learning_rate": 3.333434678673489e-05,
"loss": 0.5315,
"step": 1230
},
{
"epoch": 1.2413016270337922,
"grad_norm": 0.47537413239479065,
"learning_rate": 3.307914568276056e-05,
"loss": 0.5392,
"step": 1240
},
{
"epoch": 1.251314142678348,
"grad_norm": 0.471064031124115,
"learning_rate": 3.282300101580386e-05,
"loss": 0.5349,
"step": 1250
},
{
"epoch": 1.2613266583229037,
"grad_norm": 0.5098879933357239,
"learning_rate": 3.256594270098644e-05,
"loss": 0.5367,
"step": 1260
},
{
"epoch": 1.2713391739674593,
"grad_norm": 0.5475518107414246,
"learning_rate": 3.2308000760134824e-05,
"loss": 0.5344,
"step": 1270
},
{
"epoch": 1.281351689612015,
"grad_norm": 0.50394207239151,
"learning_rate": 3.204920531827418e-05,
"loss": 0.545,
"step": 1280
},
{
"epoch": 1.2913642052565706,
"grad_norm": 0.5121902823448181,
"learning_rate": 3.1789586600109995e-05,
"loss": 0.5436,
"step": 1290
},
{
"epoch": 1.3013767209011264,
"grad_norm": 0.4841352701187134,
"learning_rate": 3.1529174926498175e-05,
"loss": 0.5518,
"step": 1300
},
{
"epoch": 1.311389236545682,
"grad_norm": 0.5672234892845154,
"learning_rate": 3.126800071090382e-05,
"loss": 0.5359,
"step": 1310
},
{
"epoch": 1.321401752190238,
"grad_norm": 1.4713329076766968,
"learning_rate": 3.1006094455849305e-05,
"loss": 0.5295,
"step": 1320
},
{
"epoch": 1.3314142678347936,
"grad_norm": 0.5116292238235474,
"learning_rate": 3.0743486749351817e-05,
"loss": 0.5357,
"step": 1330
},
{
"epoch": 1.3414267834793492,
"grad_norm": 0.5375059843063354,
"learning_rate": 3.048020826135104e-05,
"loss": 0.5344,
"step": 1340
},
{
"epoch": 1.3514392991239048,
"grad_norm": 0.5401167869567871,
"learning_rate": 3.0216289740127185e-05,
"loss": 0.5398,
"step": 1350
},
{
"epoch": 1.3614518147684604,
"grad_norm": 0.5380666255950928,
"learning_rate": 2.9951762008709904e-05,
"loss": 0.5264,
"step": 1360
},
{
"epoch": 1.3714643304130163,
"grad_norm": 0.49941694736480713,
"learning_rate": 2.968665596127848e-05,
"loss": 0.5303,
"step": 1370
},
{
"epoch": 1.381476846057572,
"grad_norm": 0.5503513216972351,
"learning_rate": 2.9421002559553706e-05,
"loss": 0.5315,
"step": 1380
},
{
"epoch": 1.3914893617021278,
"grad_norm": 0.5197814106941223,
"learning_rate": 2.915483282918182e-05,
"loss": 0.5257,
"step": 1390
},
{
"epoch": 1.4015018773466834,
"grad_norm": 0.576173722743988,
"learning_rate": 2.8888177856111083e-05,
"loss": 0.5344,
"step": 1400
},
{
"epoch": 1.411514392991239,
"grad_norm": 0.4857792854309082,
"learning_rate": 2.862106878296121e-05,
"loss": 0.538,
"step": 1410
},
{
"epoch": 1.4215269086357947,
"grad_norm": 0.5633933544158936,
"learning_rate": 2.8353536805386245e-05,
"loss": 0.5277,
"step": 1420
},
{
"epoch": 1.4315394242803503,
"grad_norm": 0.4864403307437897,
"learning_rate": 2.8085613168431217e-05,
"loss": 0.5274,
"step": 1430
},
{
"epoch": 1.4415519399249062,
"grad_norm": 0.544262170791626,
"learning_rate": 2.781732916288303e-05,
"loss": 0.529,
"step": 1440
},
{
"epoch": 1.4515644555694618,
"grad_norm": 0.5167874693870544,
"learning_rate": 2.7548716121616015e-05,
"loss": 0.5335,
"step": 1450
},
{
"epoch": 1.4615769712140176,
"grad_norm": 0.5315885543823242,
"learning_rate": 2.727980541593257e-05,
"loss": 0.5428,
"step": 1460
},
{
"epoch": 1.4715894868585733,
"grad_norm": 0.5589486360549927,
"learning_rate": 2.70106284518993e-05,
"loss": 0.5227,
"step": 1470
},
{
"epoch": 1.481602002503129,
"grad_norm": 0.5045379400253296,
"learning_rate": 2.6741216666679115e-05,
"loss": 0.5235,
"step": 1480
},
{
"epoch": 1.4916145181476845,
"grad_norm": 0.5237298011779785,
"learning_rate": 2.6471601524859675e-05,
"loss": 0.5335,
"step": 1490
},
{
"epoch": 1.5016270337922402,
"grad_norm": 0.5030074715614319,
"learning_rate": 2.6201814514778638e-05,
"loss": 0.5256,
"step": 1500
},
{
"epoch": 1.511639549436796,
"grad_norm": 0.5370875597000122,
"learning_rate": 2.5931887144846157e-05,
"loss": 0.5347,
"step": 1510
},
{
"epoch": 1.5216520650813516,
"grad_norm": 0.5221341252326965,
"learning_rate": 2.5661850939865006e-05,
"loss": 0.5299,
"step": 1520
},
{
"epoch": 1.5316645807259075,
"grad_norm": 0.5204113721847534,
"learning_rate": 2.539173743734879e-05,
"loss": 0.5219,
"step": 1530
},
{
"epoch": 1.5416770963704631,
"grad_norm": 0.5480059385299683,
"learning_rate": 2.512157818383869e-05,
"loss": 0.5356,
"step": 1540
},
{
"epoch": 1.5516896120150188,
"grad_norm": 0.5181515216827393,
"learning_rate": 2.4851404731219154e-05,
"loss": 0.5362,
"step": 1550
},
{
"epoch": 1.5617021276595744,
"grad_norm": 0.49918684363365173,
"learning_rate": 2.4581248633032935e-05,
"loss": 0.5205,
"step": 1560
},
{
"epoch": 1.57171464330413,
"grad_norm": 1.2904787063598633,
"learning_rate": 2.4311141440795953e-05,
"loss": 0.523,
"step": 1570
},
{
"epoch": 1.5817271589486859,
"grad_norm": 0.5487804412841797,
"learning_rate": 2.4041114700312425e-05,
"loss": 0.5237,
"step": 1580
},
{
"epoch": 1.5917396745932415,
"grad_norm": 0.5446999669075012,
"learning_rate": 2.3771199947990585e-05,
"loss": 0.5211,
"step": 1590
},
{
"epoch": 1.6017521902377974,
"grad_norm": 0.5772347450256348,
"learning_rate": 2.3501428707159617e-05,
"loss": 0.53,
"step": 1600
},
{
"epoch": 1.611764705882353,
"grad_norm": 0.545523464679718,
"learning_rate": 2.3231832484387965e-05,
"loss": 0.5252,
"step": 1610
},
{
"epoch": 1.6217772215269086,
"grad_norm": 0.5056798458099365,
"learning_rate": 2.296244276580375e-05,
"loss": 0.5274,
"step": 1620
},
{
"epoch": 1.6317897371714642,
"grad_norm": 0.5340022444725037,
"learning_rate": 2.2693291013417453e-05,
"loss": 0.5273,
"step": 1630
},
{
"epoch": 1.6418022528160199,
"grad_norm": 0.5341758728027344,
"learning_rate": 2.2424408661447534e-05,
"loss": 0.5179,
"step": 1640
},
{
"epoch": 1.6518147684605757,
"grad_norm": 0.520226240158081,
"learning_rate": 2.2155827112649147e-05,
"loss": 0.5307,
"step": 1650
},
{
"epoch": 1.6618272841051314,
"grad_norm": 0.6203330159187317,
"learning_rate": 2.1887577734646706e-05,
"loss": 0.5359,
"step": 1660
},
{
"epoch": 1.6718397997496872,
"grad_norm": 0.5315065979957581,
"learning_rate": 2.1619691856270395e-05,
"loss": 0.534,
"step": 1670
},
{
"epoch": 1.6818523153942428,
"grad_norm": 0.5280863642692566,
"learning_rate": 2.1352200763897282e-05,
"loss": 0.5303,
"step": 1680
},
{
"epoch": 1.6918648310387985,
"grad_norm": 0.5708233118057251,
"learning_rate": 2.1085135697797426e-05,
"loss": 0.527,
"step": 1690
},
{
"epoch": 1.701877346683354,
"grad_norm": 0.503731369972229,
"learning_rate": 2.0818527848485275e-05,
"loss": 0.5196,
"step": 1700
},
{
"epoch": 1.7118898623279097,
"grad_norm": 0.5757652521133423,
"learning_rate": 2.0552408353076992e-05,
"loss": 0.5434,
"step": 1710
},
{
"epoch": 1.7219023779724656,
"grad_norm": 0.532005250453949,
"learning_rate": 2.0286808291653853e-05,
"loss": 0.5217,
"step": 1720
},
{
"epoch": 1.7319148936170212,
"grad_norm": 0.5191890001296997,
"learning_rate": 2.0021758683632537e-05,
"loss": 0.5215,
"step": 1730
},
{
"epoch": 1.741927409261577,
"grad_norm": 0.5299850702285767,
"learning_rate": 1.9757290484142244e-05,
"loss": 0.5118,
"step": 1740
},
{
"epoch": 1.7519399249061327,
"grad_norm": 0.5394216775894165,
"learning_rate": 1.9493434580409546e-05,
"loss": 0.5129,
"step": 1750
},
{
"epoch": 1.7619524405506883,
"grad_norm": 0.5350205898284912,
"learning_rate": 1.9230221788150983e-05,
"loss": 0.526,
"step": 1760
},
{
"epoch": 1.771964956195244,
"grad_norm": 0.5062662959098816,
"learning_rate": 1.8967682847974193e-05,
"loss": 0.524,
"step": 1770
},
{
"epoch": 1.7819774718397996,
"grad_norm": 0.5257699489593506,
"learning_rate": 1.870584842178761e-05,
"loss": 0.5236,
"step": 1780
},
{
"epoch": 1.7919899874843555,
"grad_norm": 0.48296642303466797,
"learning_rate": 1.844474908921957e-05,
"loss": 0.5217,
"step": 1790
},
{
"epoch": 1.8020025031289113,
"grad_norm": 0.5666847229003906,
"learning_rate": 1.8184415344046825e-05,
"loss": 0.5414,
"step": 1800
},
{
"epoch": 1.812015018773467,
"grad_norm": 0.6386372447013855,
"learning_rate": 1.7924877590633237e-05,
"loss": 0.5286,
"step": 1810
},
{
"epoch": 1.8220275344180226,
"grad_norm": 0.5515421032905579,
"learning_rate": 1.7666166140378852e-05,
"loss": 0.5268,
"step": 1820
},
{
"epoch": 1.8320400500625782,
"grad_norm": 0.5558871626853943,
"learning_rate": 1.740831120817976e-05,
"loss": 0.5119,
"step": 1830
},
{
"epoch": 1.8420525657071338,
"grad_norm": 0.5632119178771973,
"learning_rate": 1.7151342908899415e-05,
"loss": 0.5163,
"step": 1840
},
{
"epoch": 1.8520650813516895,
"grad_norm": 0.5401710867881775,
"learning_rate": 1.6895291253851413e-05,
"loss": 0.5267,
"step": 1850
},
{
"epoch": 1.8620775969962453,
"grad_norm": 0.5261772274971008,
"learning_rate": 1.6640186147294563e-05,
"loss": 0.5216,
"step": 1860
},
{
"epoch": 1.8720901126408012,
"grad_norm": 0.5251931548118591,
"learning_rate": 1.6386057382940283e-05,
"loss": 0.5342,
"step": 1870
},
{
"epoch": 1.8821026282853568,
"grad_norm": 0.5322995185852051,
"learning_rate": 1.613293464047308e-05,
"loss": 0.5189,
"step": 1880
},
{
"epoch": 1.8921151439299124,
"grad_norm": 0.6049567461013794,
"learning_rate": 1.588084748208418e-05,
"loss": 0.5174,
"step": 1890
},
{
"epoch": 1.902127659574468,
"grad_norm": 0.5695785880088806,
"learning_rate": 1.5629825349019054e-05,
"loss": 0.5216,
"step": 1900
},
{
"epoch": 1.9121401752190237,
"grad_norm": 0.541617214679718,
"learning_rate": 1.5379897558138862e-05,
"loss": 0.5219,
"step": 1910
},
{
"epoch": 1.9221526908635793,
"grad_norm": 0.550284206867218,
"learning_rate": 1.5131093298496646e-05,
"loss": 0.5228,
"step": 1920
},
{
"epoch": 1.9321652065081352,
"grad_norm": 0.5386470556259155,
"learning_rate": 1.4883441627928274e-05,
"loss": 0.5273,
"step": 1930
},
{
"epoch": 1.942177722152691,
"grad_norm": 0.5703060030937195,
"learning_rate": 1.4636971469658812e-05,
"loss": 0.5242,
"step": 1940
},
{
"epoch": 1.9521902377972467,
"grad_norm": 0.5564872026443481,
"learning_rate": 1.4391711608924544e-05,
"loss": 0.521,
"step": 1950
},
{
"epoch": 1.9622027534418023,
"grad_norm": 0.570838987827301,
"learning_rate": 1.4147690689611214e-05,
"loss": 0.5214,
"step": 1960
},
{
"epoch": 1.972215269086358,
"grad_norm": 0.5769771337509155,
"learning_rate": 1.3904937210908631e-05,
"loss": 0.5263,
"step": 1970
},
{
"epoch": 1.9822277847309135,
"grad_norm": 0.542021632194519,
"learning_rate": 1.366347952398228e-05,
"loss": 0.5154,
"step": 1980
},
{
"epoch": 1.9922403003754692,
"grad_norm": 0.6000685095787048,
"learning_rate": 1.3423345828662236e-05,
"loss": 0.5224,
"step": 1990
},
{
"epoch": 2.002002503128911,
"grad_norm": 0.5418619513511658,
"learning_rate": 1.3184564170149616e-05,
"loss": 0.5213,
"step": 2000
},
{
"epoch": 2.0120150187734667,
"grad_norm": 0.5431219935417175,
"learning_rate": 1.2947162435741278e-05,
"loss": 0.4901,
"step": 2010
},
{
"epoch": 2.0220275344180227,
"grad_norm": 0.5284715294837952,
"learning_rate": 1.2711168351572789e-05,
"loss": 0.4873,
"step": 2020
},
{
"epoch": 2.0320400500625784,
"grad_norm": 0.5649401545524597,
"learning_rate": 1.2476609479380324e-05,
"loss": 0.4974,
"step": 2030
},
{
"epoch": 2.042052565707134,
"grad_norm": 0.5734306573867798,
"learning_rate": 1.2243513213281735e-05,
"loss": 0.4919,
"step": 2040
},
{
"epoch": 2.0520650813516896,
"grad_norm": 0.5586088299751282,
"learning_rate": 1.2011906776577203e-05,
"loss": 0.4929,
"step": 2050
},
{
"epoch": 2.0620775969962453,
"grad_norm": 0.5533231496810913,
"learning_rate": 1.1781817218569782e-05,
"loss": 0.4759,
"step": 2060
},
{
"epoch": 2.072090112640801,
"grad_norm": 0.5898957252502441,
"learning_rate": 1.1553271411406392e-05,
"loss": 0.4965,
"step": 2070
},
{
"epoch": 2.0821026282853565,
"grad_norm": 0.57174152135849,
"learning_rate": 1.1326296046939333e-05,
"loss": 0.4963,
"step": 2080
},
{
"epoch": 2.0921151439299126,
"grad_norm": 0.6051098704338074,
"learning_rate": 1.1100917633609009e-05,
"loss": 0.4977,
"step": 2090
},
{
"epoch": 2.1021276595744682,
"grad_norm": 0.5997815728187561,
"learning_rate": 1.0877162493347978e-05,
"loss": 0.4901,
"step": 2100
},
{
"epoch": 2.112140175219024,
"grad_norm": 0.5957483649253845,
"learning_rate": 1.0655056758506846e-05,
"loss": 0.4903,
"step": 2110
},
{
"epoch": 2.1221526908635795,
"grad_norm": 0.5864090919494629,
"learning_rate": 1.0434626368802222e-05,
"loss": 0.489,
"step": 2120
},
{
"epoch": 2.132165206508135,
"grad_norm": 0.6237981915473938,
"learning_rate": 1.0215897068287274e-05,
"loss": 0.4935,
"step": 2130
},
{
"epoch": 2.1421777221526908,
"grad_norm": 0.651959240436554,
"learning_rate": 9.998894402345043e-06,
"loss": 0.4827,
"step": 2140
},
{
"epoch": 2.1521902377972464,
"grad_norm": 0.6107705235481262,
"learning_rate": 9.783643714705001e-06,
"loss": 0.4972,
"step": 2150
},
{
"epoch": 2.1622027534418025,
"grad_norm": 0.5812814235687256,
"learning_rate": 9.5701701444832e-06,
"loss": 0.4951,
"step": 2160
},
{
"epoch": 2.172215269086358,
"grad_norm": 0.5957708358764648,
"learning_rate": 9.35849862324622e-06,
"loss": 0.4935,
"step": 2170
},
{
"epoch": 2.1822277847309137,
"grad_norm": 0.5565418601036072,
"learning_rate": 9.148653872099455e-06,
"loss": 0.4882,
"step": 2180
},
{
"epoch": 2.1922403003754694,
"grad_norm": 0.5700868964195251,
"learning_rate": 8.940660398799911e-06,
"loss": 0.4911,
"step": 2190
},
{
"epoch": 2.202252816020025,
"grad_norm": 0.5764639377593994,
"learning_rate": 8.734542494893955e-06,
"loss": 0.4981,
"step": 2200
},
{
"epoch": 2.2122653316645806,
"grad_norm": 0.5823454856872559,
"learning_rate": 8.530324232880299e-06,
"loss": 0.4923,
"step": 2210
},
{
"epoch": 2.2222778473091362,
"grad_norm": 0.5832889676094055,
"learning_rate": 8.328029463398615e-06,
"loss": 0.4913,
"step": 2220
},
{
"epoch": 2.2322903629536923,
"grad_norm": 0.5884621739387512,
"learning_rate": 8.127681812443947e-06,
"loss": 0.4935,
"step": 2230
},
{
"epoch": 2.242302878598248,
"grad_norm": 0.5512953996658325,
"learning_rate": 7.929304678607505e-06,
"loss": 0.488,
"step": 2240
},
{
"epoch": 2.2523153942428036,
"grad_norm": 0.6155216097831726,
"learning_rate": 7.732921230343892e-06,
"loss": 0.4964,
"step": 2250
},
{
"epoch": 2.262327909887359,
"grad_norm": 0.6091813445091248,
"learning_rate": 7.538554403265291e-06,
"loss": 0.4922,
"step": 2260
},
{
"epoch": 2.272340425531915,
"grad_norm": 0.6014239192008972,
"learning_rate": 7.346226897462793e-06,
"loss": 0.4918,
"step": 2270
},
{
"epoch": 2.2823529411764705,
"grad_norm": 0.600447416305542,
"learning_rate": 7.155961174855297e-06,
"loss": 0.4866,
"step": 2280
},
{
"epoch": 2.292365456821026,
"grad_norm": 0.592515230178833,
"learning_rate": 6.967779456566126e-06,
"loss": 0.4929,
"step": 2290
},
{
"epoch": 2.3023779724655817,
"grad_norm": 0.6076403856277466,
"learning_rate": 6.7817037203278786e-06,
"loss": 0.4858,
"step": 2300
},
{
"epoch": 2.312390488110138,
"grad_norm": 0.6241127848625183,
"learning_rate": 6.597755697915606e-06,
"loss": 0.4879,
"step": 2310
},
{
"epoch": 2.3224030037546934,
"grad_norm": 0.6401041150093079,
"learning_rate": 6.415956872608775e-06,
"loss": 0.491,
"step": 2320
},
{
"epoch": 2.332415519399249,
"grad_norm": 0.6090307235717773,
"learning_rate": 6.2363284766822625e-06,
"loss": 0.4846,
"step": 2330
},
{
"epoch": 2.3424280350438047,
"grad_norm": 0.621647298336029,
"learning_rate": 6.058891488926591e-06,
"loss": 0.4908,
"step": 2340
},
{
"epoch": 2.3524405506883603,
"grad_norm": 0.6762897372245789,
"learning_rate": 5.883666632197868e-06,
"loss": 0.4861,
"step": 2350
},
{
"epoch": 2.362453066332916,
"grad_norm": 0.6175228357315063,
"learning_rate": 5.7106743709975204e-06,
"loss": 0.4841,
"step": 2360
},
{
"epoch": 2.3724655819774716,
"grad_norm": 0.6411292552947998,
"learning_rate": 5.53993490908227e-06,
"loss": 0.4848,
"step": 2370
},
{
"epoch": 2.3824780976220277,
"grad_norm": 0.6363428831100464,
"learning_rate": 5.371468187104528e-06,
"loss": 0.4899,
"step": 2380
},
{
"epoch": 2.3924906132665833,
"grad_norm": 0.6230722665786743,
"learning_rate": 5.205293880283552e-06,
"loss": 0.4953,
"step": 2390
},
{
"epoch": 2.402503128911139,
"grad_norm": 0.6571158766746521,
"learning_rate": 5.041431396107532e-06,
"loss": 0.4953,
"step": 2400
},
{
"epoch": 2.4125156445556946,
"grad_norm": 0.616431474685669,
"learning_rate": 4.879899872067045e-06,
"loss": 0.494,
"step": 2410
},
{
"epoch": 2.42252816020025,
"grad_norm": 0.5991277694702148,
"learning_rate": 4.720718173419947e-06,
"loss": 0.4952,
"step": 2420
},
{
"epoch": 2.432540675844806,
"grad_norm": 0.5943247079849243,
"learning_rate": 4.563904890988107e-06,
"loss": 0.496,
"step": 2430
},
{
"epoch": 2.4425531914893615,
"grad_norm": 0.6258060336112976,
"learning_rate": 4.409478338986203e-06,
"loss": 0.4788,
"step": 2440
},
{
"epoch": 2.4525657071339175,
"grad_norm": 0.6271305680274963,
"learning_rate": 4.257456552882808e-06,
"loss": 0.4787,
"step": 2450
},
{
"epoch": 2.462578222778473,
"grad_norm": 0.6384322643280029,
"learning_rate": 4.10785728729399e-06,
"loss": 0.4979,
"step": 2460
},
{
"epoch": 2.472590738423029,
"grad_norm": 0.6410216689109802,
"learning_rate": 3.960698013909809e-06,
"loss": 0.4829,
"step": 2470
},
{
"epoch": 2.4826032540675844,
"grad_norm": 0.6721149682998657,
"learning_rate": 3.815995919453743e-06,
"loss": 0.4925,
"step": 2480
},
{
"epoch": 2.49261576971214,
"grad_norm": 0.5894445180892944,
"learning_rate": 3.67376790367549e-06,
"loss": 0.4795,
"step": 2490
},
{
"epoch": 2.502628285356696,
"grad_norm": 0.63447105884552,
"learning_rate": 3.5340305773772425e-06,
"loss": 0.4929,
"step": 2500
},
{
"epoch": 2.5126408010012513,
"grad_norm": 0.588717520236969,
"learning_rate": 3.3968002604736785e-06,
"loss": 0.4884,
"step": 2510
},
{
"epoch": 2.5226533166458074,
"grad_norm": 0.6290906667709351,
"learning_rate": 3.2620929800860205e-06,
"loss": 0.4736,
"step": 2520
},
{
"epoch": 2.532665832290363,
"grad_norm": 0.6323853731155396,
"learning_rate": 3.12992446867017e-06,
"loss": 0.4895,
"step": 2530
},
{
"epoch": 2.5426783479349186,
"grad_norm": 0.6200573444366455,
"learning_rate": 3.000310162179343e-06,
"loss": 0.4851,
"step": 2540
},
{
"epoch": 2.5526908635794743,
"grad_norm": 0.6537245512008667,
"learning_rate": 2.873265198261299e-06,
"loss": 0.4905,
"step": 2550
},
{
"epoch": 2.56270337922403,
"grad_norm": 0.6241253018379211,
"learning_rate": 2.7488044144904384e-06,
"loss": 0.4961,
"step": 2560
},
{
"epoch": 2.572715894868586,
"grad_norm": 0.683192789554596,
"learning_rate": 2.626942346634878e-06,
"loss": 0.4987,
"step": 2570
},
{
"epoch": 2.582728410513141,
"grad_norm": 0.6398680806159973,
"learning_rate": 2.507693226958871e-06,
"loss": 0.4812,
"step": 2580
},
{
"epoch": 2.5927409261576972,
"grad_norm": 0.6247938275337219,
"learning_rate": 2.391070982560564e-06,
"loss": 0.4886,
"step": 2590
},
{
"epoch": 2.602753441802253,
"grad_norm": 0.5934692025184631,
"learning_rate": 2.2770892337455135e-06,
"loss": 0.4861,
"step": 2600
},
{
"epoch": 2.6127659574468085,
"grad_norm": 0.6466832756996155,
"learning_rate": 2.165761292435922e-06,
"loss": 0.4943,
"step": 2610
},
{
"epoch": 2.622778473091364,
"grad_norm": 0.6184497475624084,
"learning_rate": 2.0571001606159686e-06,
"loss": 0.49,
"step": 2620
},
{
"epoch": 2.6327909887359198,
"grad_norm": 0.6564368009567261,
"learning_rate": 1.9511185288133004e-06,
"loss": 0.4822,
"step": 2630
},
{
"epoch": 2.642803504380476,
"grad_norm": 0.6337063312530518,
"learning_rate": 1.8478287746168887e-06,
"loss": 0.4899,
"step": 2640
},
{
"epoch": 2.652816020025031,
"grad_norm": 0.6827977895736694,
"learning_rate": 1.7472429612314823e-06,
"loss": 0.4883,
"step": 2650
},
{
"epoch": 2.662828535669587,
"grad_norm": 0.6701290011405945,
"learning_rate": 1.649372836068719e-06,
"loss": 0.4962,
"step": 2660
},
{
"epoch": 2.6728410513141427,
"grad_norm": 0.5762397050857544,
"learning_rate": 1.55422982937517e-06,
"loss": 0.4965,
"step": 2670
},
{
"epoch": 2.6828535669586984,
"grad_norm": 0.6099869608879089,
"learning_rate": 1.4618250528973755e-06,
"loss": 0.4943,
"step": 2680
},
{
"epoch": 2.692866082603254,
"grad_norm": 0.6205244064331055,
"learning_rate": 1.3721692985841361e-06,
"loss": 0.4927,
"step": 2690
},
{
"epoch": 2.7028785982478096,
"grad_norm": 0.65453040599823,
"learning_rate": 1.2852730373260845e-06,
"loss": 0.4913,
"step": 2700
},
{
"epoch": 2.7128911138923657,
"grad_norm": 0.6482648253440857,
"learning_rate": 1.2011464177328236e-06,
"loss": 0.4972,
"step": 2710
},
{
"epoch": 2.722903629536921,
"grad_norm": 0.5976265668869019,
"learning_rate": 1.1197992649476492e-06,
"loss": 0.4864,
"step": 2720
},
{
"epoch": 2.732916145181477,
"grad_norm": 0.6042509078979492,
"learning_rate": 1.0412410795000964e-06,
"loss": 0.4984,
"step": 2730
},
{
"epoch": 2.7429286608260326,
"grad_norm": 0.6241819858551025,
"learning_rate": 9.654810361963362e-07,
"loss": 0.4876,
"step": 2740
},
{
"epoch": 2.7529411764705882,
"grad_norm": 0.6548027396202087,
"learning_rate": 8.92527983047689e-07,
"loss": 0.4893,
"step": 2750
},
{
"epoch": 2.762953692115144,
"grad_norm": 0.5953883528709412,
"learning_rate": 8.223904402372334e-07,
"loss": 0.4782,
"step": 2760
},
{
"epoch": 2.7729662077596995,
"grad_norm": 0.6101512312889099,
"learning_rate": 7.550765991247654e-07,
"loss": 0.4899,
"step": 2770
},
{
"epoch": 2.7829787234042556,
"grad_norm": 0.6580602526664734,
"learning_rate": 6.905943212901e-07,
"loss": 0.5002,
"step": 2780
},
{
"epoch": 2.7929912390488107,
"grad_norm": 0.6054741144180298,
"learning_rate": 6.289511376149471e-07,
"loss": 0.4918,
"step": 2790
},
{
"epoch": 2.803003754693367,
"grad_norm": 0.6138226985931396,
"learning_rate": 5.701542474033511e-07,
"loss": 0.4945,
"step": 2800
},
{
"epoch": 2.8130162703379225,
"grad_norm": 0.6270805597305298,
"learning_rate": 5.142105175409001e-07,
"loss": 0.4939,
"step": 2810
},
{
"epoch": 2.823028785982478,
"grad_norm": 0.6723251938819885,
"learning_rate": 4.611264816927502e-07,
"loss": 0.4927,
"step": 2820
},
{
"epoch": 2.8330413016270337,
"grad_norm": 0.653648853302002,
"learning_rate": 4.109083395405416e-07,
"loss": 0.4981,
"step": 2830
},
{
"epoch": 2.8430538172715893,
"grad_norm": 0.5800461173057556,
"learning_rate": 3.635619560583503e-07,
"loss": 0.4813,
"step": 2840
},
{
"epoch": 2.8530663329161454,
"grad_norm": 0.6078097224235535,
"learning_rate": 3.190928608277161e-07,
"loss": 0.4836,
"step": 2850
},
{
"epoch": 2.8630788485607006,
"grad_norm": 0.6220079064369202,
"learning_rate": 2.775062473918344e-07,
"loss": 0.4966,
"step": 2860
},
{
"epoch": 2.8730913642052567,
"grad_norm": 0.6504079699516296,
"learning_rate": 2.388069726490139e-07,
"loss": 0.483,
"step": 2870
},
{
"epoch": 2.8831038798498123,
"grad_norm": 0.6452395915985107,
"learning_rate": 2.029995562854381e-07,
"loss": 0.4976,
"step": 2880
},
{
"epoch": 2.893116395494368,
"grad_norm": 0.6240577101707458,
"learning_rate": 1.70088180247302e-07,
"loss": 0.4873,
"step": 2890
},
{
"epoch": 2.9031289111389236,
"grad_norm": 0.6439421772956848,
"learning_rate": 1.4007668825241627e-07,
"loss": 0.4868,
"step": 2900
},
{
"epoch": 2.913141426783479,
"grad_norm": 0.6726459860801697,
"learning_rate": 1.1296858534128596e-07,
"loss": 0.4901,
"step": 2910
},
{
"epoch": 2.9231539424280353,
"grad_norm": 0.6149904727935791,
"learning_rate": 8.876703746776848e-08,
"loss": 0.4921,
"step": 2920
},
{
"epoch": 2.9331664580725905,
"grad_norm": 0.63145911693573,
"learning_rate": 6.747487112931661e-08,
"loss": 0.498,
"step": 2930
},
{
"epoch": 2.9431789737171465,
"grad_norm": 0.5948153138160706,
"learning_rate": 4.9094573036873035e-08,
"loss": 0.4948,
"step": 2940
},
{
"epoch": 2.953191489361702,
"grad_norm": 0.5933936834335327,
"learning_rate": 3.3628289824444436e-08,
"loss": 0.4957,
"step": 2950
},
{
"epoch": 2.963204005006258,
"grad_norm": 0.6102668642997742,
"learning_rate": 2.1077827798404726e-08,
"loss": 0.4868,
"step": 2960
},
{
"epoch": 2.9732165206508134,
"grad_norm": 0.6177405714988708,
"learning_rate": 1.1444652726522176e-08,
"loss": 0.4941,
"step": 2970
},
{
"epoch": 2.983229036295369,
"grad_norm": 0.6207489371299744,
"learning_rate": 4.729889666793552e-09,
"loss": 0.4758,
"step": 2980
},
{
"epoch": 2.993241551939925,
"grad_norm": 0.5783057808876038,
"learning_rate": 9.343228360214618e-10,
"loss": 0.4917,
"step": 2990
},
{
"epoch": 3.0,
"step": 2997,
"total_flos": 5.543078096363586e+18,
"train_loss": 0.5661044798575126,
"train_runtime": 20294.7561,
"train_samples_per_second": 9.447,
"train_steps_per_second": 0.148
}
],
"logging_steps": 10,
"max_steps": 2997,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"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": 5.543078096363586e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}