Agron / report.json
teddyoas's picture
Add report.json for 1-epoch run
0728f64 verified
Raw History Blame Contribute Delete
50.7 kB
{
"status": "completed",
"model": "teddyoas/Agron (continued pretraining)",
"data_dir": "/content/data/agron-domain",
"corpus": {
"created": "2026-10-26T11:57:09Z",
"tokenizer": "teddyoas/Agron tokenizer.json",
"seed": 1234,
"sources": {
"code": {
"dataset": "codeparrot/github-code",
"shards": 8
},
"cyber": {
"dataset": "Trendyol/Trendyol-Cybersecurity-Instruction-Tuning-Dataset",
"file": "CyberSec-Dataset_escaped.jsonl"
},
"replay": {
"dataset": "Salesforce/wikitext",
"config": "wikitext-2-raw-v1"
}
},
"caps_chars": {
"code_total": 150000000,
"cyber": 120000000,
"replay": 15000000
},
"val": {
"code": {
"tokens": 455373,
"docs": 150
},
"cyber": {
"tokens": 137298,
"docs": 150
},
"replay": {
"tokens": 22625,
"docs": 150
}
},
"train": {
"tokens": 118720966,
"docs": 74018,
"tokens_per_source": {
"C": 7064900,
"cyber": 30769293,
"JavaScript": 14200205,
"replay": 2342084,
"C++": 9751154,
"Java": 8843792,
"Rust": 5327110,
"TypeScript": 9911774,
"Python": 20339178,
"Go": 5748972,
"C#": 4422504
}
},
"disk_bytes": 238672524
},
"hyperparameters": {
"block_size": 512,
"batch_size": 16,
"grad_accum": 2,
"epochs": 1.0,
"steps": 7246,
"lr": 0.0003,
"warmup": 100,
"weight_decay": 0.01,
"grad_clip": 1.0,
"seed": 42
},
"tokens_per_step": 16384,
"total_tokens": 118718464,
"train_wall_seconds": 3659.7,
"tokens_per_second": 32439.5,
"eval_before": {
"code": {
"loss": 1.4532,
"perplexity": 4.28,
"blocks": 16
},
"cyber": {
"loss": 1.358,
"perplexity": 3.89,
"blocks": 16
},
"replay": {
"loss": 5.626,
"perplexity": 277.55,
"blocks": 5
}
},
"eval_after": {
"code": {
"loss": 1.2451,
"perplexity": 3.47,
"blocks": 16
},
"cyber": {
"loss": 1.0636,
"perplexity": 2.9,
"blocks": 16
},
"replay": {
"loss": 5.0238,
"perplexity": 151.98,
"blocks": 5
}
},
"eval_history": [
{
"step": 500,
"code": {
"loss": 1.5055,
"perplexity": 4.51,
"blocks": 16
},
"cyber": {
"loss": 1.4013,
"perplexity": 4.06,
"blocks": 16
},
"replay": {
"loss": 5.793,
"perplexity": 327.99,
"blocks": 5
}
},
{
"step": 1000,
"code": {
"loss": 1.4807,
"perplexity": 4.4,
"blocks": 16
},
"cyber": {
"loss": 1.3501,
"perplexity": 3.86,
"blocks": 16
},
"replay": {
"loss": 5.6901,
"perplexity": 295.92,
"blocks": 5
}
},
{
"step": 1500,
"code": {
"loss": 1.4339,
"perplexity": 4.19,
"blocks": 16
},
"cyber": {
"loss": 1.3048,
"perplexity": 3.69,
"blocks": 16
},
"replay": {
"loss": 5.5986,
"perplexity": 270.06,
"blocks": 5
}
},
{
"step": 2000,
"code": {
"loss": 1.4063,
"perplexity": 4.08,
"blocks": 16
},
"cyber": {
"loss": 1.2648,
"perplexity": 3.54,
"blocks": 16
},
"replay": {
"loss": 5.5139,
"perplexity": 248.12,
"blocks": 5
}
},
{
"step": 2500,
"code": {
"loss": 1.3737,
"perplexity": 3.95,
"blocks": 16
},
"cyber": {
"loss": 1.2266,
"perplexity": 3.41,
"blocks": 16
},
"replay": {
"loss": 5.4098,
"perplexity": 223.59,
"blocks": 5
}
},
{
"step": 3000,
"code": {
"loss": 1.3446,
"perplexity": 3.84,
"blocks": 16
},
"cyber": {
"loss": 1.1879,
"perplexity": 3.28,
"blocks": 16
},
"replay": {
"loss": 5.3373,
"perplexity": 207.95,
"blocks": 5
}
},
{
"step": 3500,
"code": {
"loss": 1.3275,
"perplexity": 3.77,
"blocks": 16
},
"cyber": {
"loss": 1.1612,
"perplexity": 3.19,
"blocks": 16
},
"replay": {
"loss": 5.2525,
"perplexity": 191.05,
"blocks": 5
}
},
{
"step": 4000,
"code": {
"loss": 1.3077,
"perplexity": 3.7,
"blocks": 16
},
"cyber": {
"loss": 1.1404,
"perplexity": 3.13,
"blocks": 16
},
"replay": {
"loss": 5.1941,
"perplexity": 180.21,
"blocks": 5
}
},
{
"step": 4500,
"code": {
"loss": 1.2838,
"perplexity": 3.61,
"blocks": 16
},
"cyber": {
"loss": 1.1155,
"perplexity": 3.05,
"blocks": 16
},
"replay": {
"loss": 5.1429,
"perplexity": 171.22,
"blocks": 5
}
},
{
"step": 5000,
"code": {
"loss": 1.2708,
"perplexity": 3.56,
"blocks": 16
},
"cyber": {
"loss": 1.0989,
"perplexity": 3.0,
"blocks": 16
},
"replay": {
"loss": 5.1017,
"perplexity": 164.31,
"blocks": 5
}
},
{
"step": 5500,
"code": {
"loss": 1.261,
"perplexity": 3.53,
"blocks": 16
},
"cyber": {
"loss": 1.0821,
"perplexity": 2.95,
"blocks": 16
},
"replay": {
"loss": 5.0702,
"perplexity": 159.21,
"blocks": 5
}
},
{
"step": 6000,
"code": {
"loss": 1.2511,
"perplexity": 3.49,
"blocks": 16
},
"cyber": {
"loss": 1.0722,
"perplexity": 2.92,
"blocks": 16
},
"replay": {
"loss": 5.0406,
"perplexity": 154.56,
"blocks": 5
}
},
{
"step": 6500,
"code": {
"loss": 1.2471,
"perplexity": 3.48,
"blocks": 16
},
"cyber": {
"loss": 1.0659,
"perplexity": 2.9,
"blocks": 16
},
"replay": {
"loss": 5.0278,
"perplexity": 152.59,
"blocks": 5
}
},
{
"step": 7000,
"code": {
"loss": 1.2451,
"perplexity": 3.47,
"blocks": 16
},
"cyber": {
"loss": 1.0637,
"perplexity": 2.9,
"blocks": 16
},
"replay": {
"loss": 5.0241,
"perplexity": 152.03,
"blocks": 5
}
}
],
"train_loss_history": [
{
"step": 25,
"loss": 1.465992431640625,
"learning_rate": 7.199999999999999e-05,
"grad_norm": 0.503962516784668
},
{
"step": 50,
"loss": 1.4964903259277345,
"learning_rate": 0.000147,
"grad_norm": 0.5079663991928101
},
{
"step": 75,
"loss": 1.4984529113769531,
"learning_rate": 0.00022199999999999998,
"grad_norm": 0.5481196641921997
},
{
"step": 100,
"loss": 1.53364013671875,
"learning_rate": 0.00029699999999999996,
"grad_norm": 0.5628758072853088
},
{
"step": 125,
"loss": 1.503697509765625,
"learning_rate": 0.00029999165063514596,
"grad_norm": 0.6324927806854248
},
{
"step": 150,
"loss": 1.5965298461914061,
"learning_rate": 0.00029996519750734643,
"grad_norm": 0.5400546193122864
},
{
"step": 175,
"loss": 1.5519804382324218,
"learning_rate": 0.0002999206293164506,
"grad_norm": 0.5300002098083496
},
{
"step": 200,
"loss": 1.5133120727539062,
"learning_rate": 0.0002998579514460781,
"grad_norm": 0.5625991821289062
},
{
"step": 225,
"loss": 1.560133819580078,
"learning_rate": 0.00029977717146740954,
"grad_norm": 0.5264961123466492
},
{
"step": 250,
"loss": 1.5489260864257812,
"learning_rate": 0.0002996782991382715,
"grad_norm": 0.5348047018051147
},
{
"step": 275,
"loss": 1.4816361999511718,
"learning_rate": 0.0002995613464019583,
"grad_norm": 0.5102686285972595
},
{
"step": 300,
"loss": 1.5427243041992187,
"learning_rate": 0.00029942632738578904,
"grad_norm": 0.9355897307395935
},
{
"step": 325,
"loss": 1.4962777709960937,
"learning_rate": 0.00029927325839940107,
"grad_norm": 0.515880286693573
},
{
"step": 350,
"loss": 1.5118930053710937,
"learning_rate": 0.0002991021579327801,
"grad_norm": 0.5115503072738647
},
{
"step": 375,
"loss": 1.489414520263672,
"learning_rate": 0.0002989130466540265,
"grad_norm": 0.5668929815292358
},
{
"step": 400,
"loss": 1.531691436767578,
"learning_rate": 0.00029870594740685855,
"grad_norm": 0.5156782269477844
},
{
"step": 425,
"loss": 1.5007366943359375,
"learning_rate": 0.00029848088520785347,
"grad_norm": 0.4919520318508148
},
{
"step": 450,
"loss": 1.5251170349121095,
"learning_rate": 0.0002982378872434252,
"grad_norm": 0.5143119096755981
},
{
"step": 475,
"loss": 1.5015919494628907,
"learning_rate": 0.0002979769828665405,
"grad_norm": 0.5741536021232605
},
{
"step": 500,
"loss": 1.4777200317382813,
"learning_rate": 0.0002976982035931733,
"grad_norm": 0.5316178202629089
},
{
"step": 525,
"loss": 1.5096803283691407,
"learning_rate": 0.0002974015830984975,
"grad_norm": 0.51364666223526
},
{
"step": 550,
"loss": 1.4988615417480469,
"learning_rate": 0.0002970871572128197,
"grad_norm": 0.4849585294723511
},
{
"step": 575,
"loss": 1.5281857299804686,
"learning_rate": 0.00029675496391725037,
"grad_norm": 0.4936002492904663
},
{
"step": 600,
"loss": 1.4581077575683594,
"learning_rate": 0.00029640504333911655,
"grad_norm": 0.4904490113258362
},
{
"step": 625,
"loss": 1.4236958312988282,
"learning_rate": 0.0002960374377471143,
"grad_norm": 0.49264708161354065
},
{
"step": 650,
"loss": 1.463341522216797,
"learning_rate": 0.0002956521915462028,
"grad_norm": 0.5007567405700684
},
{
"step": 675,
"loss": 1.4722225952148438,
"learning_rate": 0.0002952493512722409,
"grad_norm": 0.48851534724235535
},
{
"step": 700,
"loss": 1.4606297302246094,
"learning_rate": 0.00029482896558636517,
"grad_norm": 0.5092523694038391
},
{
"step": 725,
"loss": 1.4079983520507813,
"learning_rate": 0.00029439108526911243,
"grad_norm": 0.5120623707771301
},
{
"step": 750,
"loss": 1.472628631591797,
"learning_rate": 0.0002939357632142855,
"grad_norm": 0.5220867991447449
},
{
"step": 775,
"loss": 1.4448049926757813,
"learning_rate": 0.0002934630544225637,
"grad_norm": 0.510402262210846
},
{
"step": 800,
"loss": 1.500727996826172,
"learning_rate": 0.0002929730159948594,
"grad_norm": 0.46804383397102356
},
{
"step": 825,
"loss": 1.4285128784179688,
"learning_rate": 0.0002924657071254201,
"grad_norm": 0.511221170425415
},
{
"step": 850,
"loss": 1.4769186401367187,
"learning_rate": 0.0002919411890946786,
"grad_norm": 0.5099458694458008
},
{
"step": 875,
"loss": 1.435450897216797,
"learning_rate": 0.0002913995252618499,
"grad_norm": 0.513334333896637
},
{
"step": 900,
"loss": 1.4503627014160156,
"learning_rate": 0.0002908407810572784,
"grad_norm": 0.48919397592544556
},
{
"step": 925,
"loss": 1.4501026916503905,
"learning_rate": 0.00029026502397453395,
"grad_norm": 0.46839049458503723
},
{
"step": 950,
"loss": 1.4088829040527344,
"learning_rate": 0.0002896723235622588,
"grad_norm": 0.5264255404472351
},
{
"step": 975,
"loss": 1.4610905456542969,
"learning_rate": 0.0002890627514157666,
"grad_norm": 0.4739971458911896
},
{
"step": 1000,
"loss": 1.439478759765625,
"learning_rate": 0.0002884363811683943,
"grad_norm": 0.5118926167488098
},
{
"step": 1025,
"loss": 1.3943740844726562,
"learning_rate": 0.00028779328848260716,
"grad_norm": 0.49222490191459656
},
{
"step": 1050,
"loss": 1.4404747009277343,
"learning_rate": 0.0002871335510408592,
"grad_norm": 0.48402321338653564
},
{
"step": 1075,
"loss": 1.3960848999023439,
"learning_rate": 0.00028645724853620995,
"grad_norm": 0.48026469349861145
},
{
"step": 1100,
"loss": 1.4121771240234375,
"learning_rate": 0.0002857644626626973,
"grad_norm": 0.5774903297424316
},
{
"step": 1125,
"loss": 1.3999868774414062,
"learning_rate": 0.00028505527710546974,
"grad_norm": 0.5563474297523499
},
{
"step": 1150,
"loss": 1.4009886169433594,
"learning_rate": 0.00028432977753067726,
"grad_norm": 0.5073521733283997
},
{
"step": 1175,
"loss": 1.375696258544922,
"learning_rate": 0.0002835880515751236,
"grad_norm": 0.5103248357772827
},
{
"step": 1200,
"loss": 1.4315164184570313,
"learning_rate": 0.00028283018883568,
"grad_norm": 0.4903603494167328
},
{
"step": 1225,
"loss": 1.4241004943847657,
"learning_rate": 0.0002820562808584622,
"grad_norm": 0.506152868270874
},
{
"step": 1250,
"loss": 1.4256643676757812,
"learning_rate": 0.0002812664211277724,
"grad_norm": 0.4851268529891968
},
{
"step": 1275,
"loss": 1.4002693176269532,
"learning_rate": 0.00028046070505480665,
"grad_norm": 0.5567605495452881
},
{
"step": 1300,
"loss": 1.4149452209472657,
"learning_rate": 0.0002796392299661297,
"grad_norm": 0.49636349081993103
},
{
"step": 1325,
"loss": 1.3971769714355469,
"learning_rate": 0.0002788020950919183,
"grad_norm": 0.4663887917995453
},
{
"step": 1350,
"loss": 1.4440313720703124,
"learning_rate": 0.0002779494015539749,
"grad_norm": 0.5311484932899475
},
{
"step": 1375,
"loss": 1.4248406982421875,
"learning_rate": 0.00027708125235351237,
"grad_norm": 0.4884320795536041
},
{
"step": 1400,
"loss": 1.3670068359375,
"learning_rate": 0.00027619775235871217,
"grad_norm": 0.5010215640068054
},
{
"step": 1425,
"loss": 1.3905780029296875,
"learning_rate": 0.0002752990082920567,
"grad_norm": 0.5153194069862366
},
{
"step": 1450,
"loss": 1.4214459228515626,
"learning_rate": 0.0002743851287174378,
"grad_norm": 0.4822602868080139
},
{
"step": 1475,
"loss": 1.4260044860839844,
"learning_rate": 0.00027345622402704257,
"grad_norm": 0.503484845161438
},
{
"step": 1500,
"loss": 1.387130126953125,
"learning_rate": 0.00027251240642801893,
"grad_norm": 0.5117263793945312
},
{
"step": 1525,
"loss": 1.4189169311523437,
"learning_rate": 0.000271553789928921,
"grad_norm": 0.5162776112556458
},
{
"step": 1550,
"loss": 1.3924339294433594,
"learning_rate": 0.0002705804903259379,
"grad_norm": 0.47961705923080444
},
{
"step": 1575,
"loss": 1.43085693359375,
"learning_rate": 0.000269592625188906,
"grad_norm": 0.4925697147846222
},
{
"step": 1600,
"loss": 1.3906794738769532,
"learning_rate": 0.00026859031384710686,
"grad_norm": 0.5015848278999329
},
{
"step": 1625,
"loss": 1.4191299438476563,
"learning_rate": 0.000267573677374853,
"grad_norm": 0.47783127427101135
},
{
"step": 1650,
"loss": 1.4013314819335938,
"learning_rate": 0.0002665428385768629,
"grad_norm": 0.44291064143180847
},
{
"step": 1675,
"loss": 1.4088816833496094,
"learning_rate": 0.00026549792197342627,
"grad_norm": 0.4942842721939087
},
{
"step": 1700,
"loss": 1.4031414794921875,
"learning_rate": 0.0002644390537853633,
"grad_norm": 0.4807228744029999
},
{
"step": 1725,
"loss": 1.4029275512695312,
"learning_rate": 0.0002633663619187772,
"grad_norm": 0.46695584058761597
},
{
"step": 1750,
"loss": 1.3980467224121094,
"learning_rate": 0.000262279975949604,
"grad_norm": 0.5140073299407959
},
{
"step": 1775,
"loss": 1.3929660034179687,
"learning_rate": 0.0002611800271079604,
"grad_norm": 0.4695814251899719
},
{
"step": 1800,
"loss": 1.3747761535644532,
"learning_rate": 0.00026006664826229203,
"grad_norm": 0.45960962772369385
},
{
"step": 1825,
"loss": 1.373397979736328,
"learning_rate": 0.00025893997390332314,
"grad_norm": 0.4975016415119171
},
{
"step": 1850,
"loss": 1.3793846130371095,
"learning_rate": 0.00025780014012781113,
"grad_norm": 0.48639926314353943
},
{
"step": 1875,
"loss": 1.4201519775390625,
"learning_rate": 0.0002566472846221067,
"grad_norm": 0.48666369915008545
},
{
"step": 1900,
"loss": 1.4070872497558593,
"learning_rate": 0.00025548154664552166,
"grad_norm": 0.4822351634502411
},
{
"step": 1925,
"loss": 1.4160031127929686,
"learning_rate": 0.00025430306701350764,
"grad_norm": 0.49669837951660156
},
{
"step": 1950,
"loss": 1.4190867614746094,
"learning_rate": 0.0002531119880806459,
"grad_norm": 0.4919108748435974
},
{
"step": 1975,
"loss": 1.4013197326660156,
"learning_rate": 0.0002519084537234516,
"grad_norm": 0.48018497228622437
},
{
"step": 2000,
"loss": 1.3750941467285156,
"learning_rate": 0.0002506926093229944,
"grad_norm": 0.5125364065170288
},
{
"step": 2025,
"loss": 1.3990634155273438,
"learning_rate": 0.000249464601747337,
"grad_norm": 0.4900771677494049
},
{
"step": 2050,
"loss": 1.3854647827148439,
"learning_rate": 0.0002482245793337941,
"grad_norm": 0.4588489532470703
},
{
"step": 2075,
"loss": 1.3835163879394532,
"learning_rate": 0.00024697269187101445,
"grad_norm": 0.46798691153526306
},
{
"step": 2100,
"loss": 1.4211228942871095,
"learning_rate": 0.0002457090905808864,
"grad_norm": 0.4728078842163086
},
{
"step": 2125,
"loss": 1.4086012268066406,
"learning_rate": 0.00024443392810027167,
"grad_norm": 0.4489896893501282
},
{
"step": 2150,
"loss": 1.3630841064453125,
"learning_rate": 0.0002431473584625671,
"grad_norm": 0.45225420594215393
},
{
"step": 2175,
"loss": 1.3702024841308593,
"learning_rate": 0.00024184953707909852,
"grad_norm": 0.4957192540168762
},
{
"step": 2200,
"loss": 1.397852783203125,
"learning_rate": 0.0002405406207203476,
"grad_norm": 0.45574721693992615
},
{
"step": 2225,
"loss": 1.4111256408691406,
"learning_rate": 0.00023922076749701487,
"grad_norm": 0.45370936393737793
},
{
"step": 2250,
"loss": 1.36464599609375,
"learning_rate": 0.00023789013684092053,
"grad_norm": 0.45561283826828003
},
{
"step": 2275,
"loss": 1.420698699951172,
"learning_rate": 0.00023654888948574615,
"grad_norm": 0.48367705941200256
},
{
"step": 2300,
"loss": 1.3629957580566405,
"learning_rate": 0.0002351971874476186,
"grad_norm": 0.45925864577293396
},
{
"step": 2325,
"loss": 1.3583328247070312,
"learning_rate": 0.00023383519400553934,
"grad_norm": 0.4538702964782715
},
{
"step": 2350,
"loss": 1.3273094177246094,
"learning_rate": 0.00023246307368166122,
"grad_norm": 0.4757903516292572
},
{
"step": 2375,
"loss": 1.3557374572753906,
"learning_rate": 0.00023108099222141494,
"grad_norm": 0.44728630781173706
},
{
"step": 2400,
"loss": 1.3738319396972656,
"learning_rate": 0.0002296891165734876,
"grad_norm": 0.5025182366371155
},
{
"step": 2425,
"loss": 1.3842662048339844,
"learning_rate": 0.00022828761486965658,
"grad_norm": 0.46316027641296387
},
{
"step": 2450,
"loss": 1.4199676513671875,
"learning_rate": 0.0002268766564044796,
"grad_norm": 0.4852520227432251
},
{
"step": 2475,
"loss": 1.3864126586914063,
"learning_rate": 0.0002254564116148451,
"grad_norm": 0.47896522283554077
},
{
"step": 2500,
"loss": 1.3791714477539063,
"learning_rate": 0.0002240270520593839,
"grad_norm": 0.7034990787506104
},
{
"step": 2525,
"loss": 1.4165682983398438,
"learning_rate": 0.0002225887503977462,
"grad_norm": 0.4801398515701294
},
{
"step": 2550,
"loss": 1.376103515625,
"learning_rate": 0.0002211416803697449,
"grad_norm": 0.4766090214252472
},
{
"step": 2575,
"loss": 1.3946040344238282,
"learning_rate": 0.00021968601677436867,
"grad_norm": 0.46213752031326294
},
{
"step": 2600,
"loss": 1.3469049072265624,
"learning_rate": 0.00021822193544866712,
"grad_norm": 0.48614174127578735
},
{
"step": 2625,
"loss": 1.4027534484863282,
"learning_rate": 0.00021674961324651061,
"grad_norm": 0.46639734506607056
},
{
"step": 2650,
"loss": 1.3664921569824218,
"learning_rate": 0.000215269228017227,
"grad_norm": 0.4851883351802826
},
{
"step": 2675,
"loss": 1.3809815979003905,
"learning_rate": 0.00021378095858411863,
"grad_norm": 0.45697319507598877
},
{
"step": 2700,
"loss": 1.3494392395019532,
"learning_rate": 0.0002122849847228609,
"grad_norm": 0.47020864486694336
},
{
"step": 2725,
"loss": 1.338717041015625,
"learning_rate": 0.00021078148713978665,
"grad_norm": 0.45249685645103455
},
{
"step": 2750,
"loss": 1.3287506103515625,
"learning_rate": 0.00020927064745005744,
"grad_norm": 0.44328123331069946
},
{
"step": 2775,
"loss": 1.3808297729492187,
"learning_rate": 0.00020775264815572558,
"grad_norm": 0.44321563839912415
},
{
"step": 2800,
"loss": 1.3122210693359375,
"learning_rate": 0.00020622767262368853,
"grad_norm": 0.4848768711090088
},
{
"step": 2825,
"loss": 1.3498515319824218,
"learning_rate": 0.00020469590506353916,
"grad_norm": 0.47914770245552063
},
{
"step": 2850,
"loss": 1.3182548522949218,
"learning_rate": 0.0002031575305053142,
"grad_norm": 0.4577605426311493
},
{
"step": 2875,
"loss": 1.3279336547851563,
"learning_rate": 0.00020161273477714345,
"grad_norm": 0.4621855318546295
},
{
"step": 2900,
"loss": 1.336206817626953,
"learning_rate": 0.0002000617044828027,
"grad_norm": 0.4753350019454956
},
{
"step": 2925,
"loss": 1.3457479858398438,
"learning_rate": 0.00019850462697917277,
"grad_norm": 0.45626455545425415
},
{
"step": 2950,
"loss": 1.3458819580078125,
"learning_rate": 0.00019694169035360792,
"grad_norm": 0.4753391742706299
},
{
"step": 2975,
"loss": 1.3453900146484374,
"learning_rate": 0.0001953730834012157,
"grad_norm": 0.4793599545955658
},
{
"step": 3000,
"loss": 1.3578317260742188,
"learning_rate": 0.00019379899560205134,
"grad_norm": 0.4540102779865265
},
{
"step": 3025,
"loss": 1.3426896667480468,
"learning_rate": 0.00019221961709822968,
"grad_norm": 0.4781694710254669
},
{
"step": 3050,
"loss": 1.3274795532226562,
"learning_rate": 0.00019063513867095678,
"grad_norm": 0.4815289080142975
},
{
"step": 3075,
"loss": 1.3402708435058595,
"learning_rate": 0.00018904575171748458,
"grad_norm": 0.44584402441978455
},
{
"step": 3100,
"loss": 1.3204678344726561,
"learning_rate": 0.00018745164822799094,
"grad_norm": 0.4505169987678528
},
{
"step": 3125,
"loss": 1.3190098571777344,
"learning_rate": 0.00018585302076238827,
"grad_norm": 0.47513049840927124
},
{
"step": 3150,
"loss": 1.318228759765625,
"learning_rate": 0.00018425006242706312,
"grad_norm": 0.46475449204444885
},
{
"step": 3175,
"loss": 1.3169645690917968,
"learning_rate": 0.00018264296685155002,
"grad_norm": 0.49327173829078674
},
{
"step": 3200,
"loss": 1.3219915771484374,
"learning_rate": 0.00018103192816514195,
"grad_norm": 0.4188969135284424
},
{
"step": 3225,
"loss": 1.3042182922363281,
"learning_rate": 0.00017941714097344027,
"grad_norm": 0.4400496482849121
},
{
"step": 3250,
"loss": 1.388089141845703,
"learning_rate": 0.00017779880033484753,
"grad_norm": 0.4402545690536499
},
{
"step": 3275,
"loss": 1.2967855834960937,
"learning_rate": 0.00017617710173700537,
"grad_norm": 0.4542015492916107
},
{
"step": 3300,
"loss": 1.3468402099609376,
"learning_rate": 0.00017455224107318037,
"grad_norm": 0.44114288687705994
},
{
"step": 3325,
"loss": 1.314374237060547,
"learning_rate": 0.0001729244146186013,
"grad_norm": 0.448632150888443
},
{
"step": 3350,
"loss": 1.3190969848632812,
"learning_rate": 0.00017129381900674997,
"grad_norm": 0.49022868275642395
},
{
"step": 3375,
"loss": 1.3036573791503907,
"learning_rate": 0.00016966065120560889,
"grad_norm": 0.465278685092926
},
{
"step": 3400,
"loss": 1.3122999572753906,
"learning_rate": 0.00016802510849386838,
"grad_norm": 0.46860554814338684
},
{
"step": 3425,
"loss": 1.3047955322265625,
"learning_rate": 0.00016638738843709644,
"grad_norm": 0.4665217399597168
},
{
"step": 3450,
"loss": 1.2953912353515624,
"learning_rate": 0.0001647476888638736,
"grad_norm": 0.4496661424636841
},
{
"step": 3475,
"loss": 1.2946479797363282,
"learning_rate": 0.00016310620784189646,
"grad_norm": 0.4517843425273895
},
{
"step": 3500,
"loss": 1.2945361328125,
"learning_rate": 0.00016146314365405177,
"grad_norm": 0.4854501783847809
},
{
"step": 3525,
"loss": 1.299219970703125,
"learning_rate": 0.00015981869477446509,
"grad_norm": 0.4721055030822754
},
{
"step": 3550,
"loss": 1.2779254913330078,
"learning_rate": 0.00015817305984452593,
"grad_norm": 0.4885696768760681
},
{
"step": 3575,
"loss": 1.3132022094726563,
"learning_rate": 0.00015652643764889294,
"grad_norm": 0.46799078583717346
},
{
"step": 3600,
"loss": 1.3397958374023438,
"learning_rate": 0.00015487902709148153,
"grad_norm": 0.4386662244796753
},
{
"step": 3625,
"loss": 1.298585205078125,
"learning_rate": 0.00015323102717143747,
"grad_norm": 0.439731627702713
},
{
"step": 3650,
"loss": 1.2728870391845704,
"learning_rate": 0.00015158263695909857,
"grad_norm": 0.47465574741363525
},
{
"step": 3675,
"loss": 1.3068107604980468,
"learning_rate": 0.00014993405557194814,
"grad_norm": 0.4406924247741699
},
{
"step": 3700,
"loss": 1.3211735534667968,
"learning_rate": 0.00014828548215056232,
"grad_norm": 0.4591175317764282
},
{
"step": 3725,
"loss": 1.2942324829101564,
"learning_rate": 0.0001466371158345553,
"grad_norm": 0.4335545301437378
},
{
"step": 3750,
"loss": 1.3200546264648438,
"learning_rate": 0.00014498915573852368,
"grad_norm": 0.4724709093570709
},
{
"step": 3775,
"loss": 1.2709454345703124,
"learning_rate": 0.00014334180092799478,
"grad_norm": 0.4954122304916382
},
{
"step": 3800,
"loss": 1.307467498779297,
"learning_rate": 0.00014169525039538032,
"grad_norm": 0.4750748574733734
},
{
"step": 3825,
"loss": 1.2666266632080079,
"learning_rate": 0.0001400497030359393,
"grad_norm": 0.4433806538581848
},
{
"step": 3850,
"loss": 1.3017340087890625,
"learning_rate": 0.00013840535762375225,
"grad_norm": 0.44521960616111755
},
{
"step": 3875,
"loss": 1.3061532592773437,
"learning_rate": 0.0001367624127877103,
"grad_norm": 0.44376805424690247
},
{
"step": 3900,
"loss": 1.3032217407226563,
"learning_rate": 0.0001351210669875218,
"grad_norm": 0.4341404438018799
},
{
"step": 3925,
"loss": 1.272267074584961,
"learning_rate": 0.00013348151848973955,
"grad_norm": 0.45423921942710876
},
{
"step": 3950,
"loss": 1.339799346923828,
"learning_rate": 0.00013184396534381086,
"grad_norm": 0.4589066803455353
},
{
"step": 3975,
"loss": 1.2691211700439453,
"learning_rate": 0.00013020860535815427,
"grad_norm": 0.44028106331825256
},
{
"step": 4000,
"loss": 1.2714073181152343,
"learning_rate": 0.0001285756360762654,
"grad_norm": 0.4480900466442108
},
{
"step": 4025,
"loss": 1.3150112915039063,
"learning_rate": 0.00012694525475285442,
"grad_norm": 0.471388578414917
},
{
"step": 4050,
"loss": 1.2618749237060547,
"learning_rate": 0.00012531765833001876,
"grad_norm": 0.45005708932876587
},
{
"step": 4075,
"loss": 1.2932235717773437,
"learning_rate": 0.00012369304341345355,
"grad_norm": 0.45370563864707947
},
{
"step": 4100,
"loss": 1.214958953857422,
"learning_rate": 0.00012207160624870243,
"grad_norm": 0.45486775040626526
},
{
"step": 4125,
"loss": 1.2909408569335938,
"learning_rate": 0.00012045354269745233,
"grad_norm": 0.452974796295166
},
{
"step": 4150,
"loss": 1.283765106201172,
"learning_rate": 0.00011883904821387391,
"grad_norm": 0.46668413281440735
},
{
"step": 4175,
"loss": 1.238913803100586,
"learning_rate": 0.0001172283178210121,
"grad_norm": 0.4666079580783844
},
{
"step": 4200,
"loss": 1.303359375,
"learning_rate": 0.00011562154608722795,
"grad_norm": 0.4673307538032532
},
{
"step": 4225,
"loss": 1.2495006561279296,
"learning_rate": 0.00011401892710269586,
"grad_norm": 0.465811550617218
},
{
"step": 4250,
"loss": 1.263448257446289,
"learning_rate": 0.00011242065445595839,
"grad_norm": 0.43971243500709534
},
{
"step": 4275,
"loss": 1.2942657470703125,
"learning_rate": 0.00011082692121054178,
"grad_norm": 0.5017976760864258
},
{
"step": 4300,
"loss": 1.2541777038574218,
"learning_rate": 0.00010923791988163473,
"grad_norm": 0.44632741808891296
},
{
"step": 4325,
"loss": 1.2856007385253907,
"learning_rate": 0.0001076538424128336,
"grad_norm": 0.47356122732162476
},
{
"step": 4350,
"loss": 1.24027099609375,
"learning_rate": 0.00010607488015295649,
"grad_norm": 0.47505736351013184
},
{
"step": 4375,
"loss": 1.2527085113525391,
"learning_rate": 0.00010450122383292921,
"grad_norm": 0.4675339460372925
},
{
"step": 4400,
"loss": 1.2761446380615233,
"learning_rate": 0.000102933063542746,
"grad_norm": 0.4455282390117645
},
{
"step": 4425,
"loss": 1.2683029174804688,
"learning_rate": 0.00010137058870850752,
"grad_norm": 0.4367285668849945
},
{
"step": 4450,
"loss": 1.2697316741943359,
"learning_rate": 9.981398806953917e-05,
"grad_norm": 0.4658321738243103
},
{
"step": 4475,
"loss": 1.290264892578125,
"learning_rate": 9.826344965559224e-05,
"grad_norm": 0.48173364996910095
},
{
"step": 4500,
"loss": 1.2587723541259765,
"learning_rate": 9.671916076413083e-05,
"grad_norm": 0.5190169215202332
},
{
"step": 4525,
"loss": 1.2330872344970703,
"learning_rate": 9.51813079377074e-05,
"grad_norm": 0.4717499315738678
},
{
"step": 4550,
"loss": 1.2380457305908203,
"learning_rate": 9.365007694142903e-05,
"grad_norm": 0.42915889620780945
},
{
"step": 4575,
"loss": 1.264954833984375,
"learning_rate": 9.212565274051817e-05,
"grad_norm": 0.4541546106338501
},
{
"step": 4600,
"loss": 1.2391783142089843,
"learning_rate": 9.060821947796955e-05,
"grad_norm": 0.4281995892524719
},
{
"step": 4625,
"loss": 1.2423888397216798,
"learning_rate": 8.909796045230688e-05,
"grad_norm": 0.445817768573761
},
{
"step": 4650,
"loss": 1.2846461486816407,
"learning_rate": 8.759505809544096e-05,
"grad_norm": 0.4558369815349579
},
{
"step": 4675,
"loss": 1.2639895629882814,
"learning_rate": 8.609969395063299e-05,
"grad_norm": 0.4140309989452362
},
{
"step": 4700,
"loss": 1.1922022247314452,
"learning_rate": 8.461204865056494e-05,
"grad_norm": 0.42347386479377747
},
{
"step": 4725,
"loss": 1.2776814270019532,
"learning_rate": 8.313230189552016e-05,
"grad_norm": 0.4576684236526489
},
{
"step": 4750,
"loss": 1.2211253356933593,
"learning_rate": 8.166063243167627e-05,
"grad_norm": 0.5280000567436218
},
{
"step": 4775,
"loss": 1.2126011657714844,
"learning_rate": 8.019721802951368e-05,
"grad_norm": 0.45232927799224854
},
{
"step": 4800,
"loss": 1.237452163696289,
"learning_rate": 7.874223546234149e-05,
"grad_norm": 0.4481607973575592
},
{
"step": 4825,
"loss": 1.198594436645508,
"learning_rate": 7.72958604849445e-05,
"grad_norm": 0.45539891719818115
},
{
"step": 4850,
"loss": 1.212102279663086,
"learning_rate": 7.585826781235274e-05,
"grad_norm": 0.45331281423568726
},
{
"step": 4875,
"loss": 1.2540426635742188,
"learning_rate": 7.442963109873657e-05,
"grad_norm": 0.4342852830886841
},
{
"step": 4900,
"loss": 1.2529083251953126,
"learning_rate": 7.301012291643042e-05,
"grad_norm": 0.4284408688545227
},
{
"step": 4925,
"loss": 1.2232284545898438,
"learning_rate": 7.159991473508666e-05,
"grad_norm": 0.4916747808456421
},
{
"step": 4950,
"loss": 1.2357479858398437,
"learning_rate": 7.019917690096307e-05,
"grad_norm": 0.44929102063179016
},
{
"step": 4975,
"loss": 1.2480144500732422,
"learning_rate": 6.880807861634563e-05,
"grad_norm": 0.4392186403274536
},
{
"step": 5000,
"loss": 1.2522673797607422,
"learning_rate": 6.742678791910994e-05,
"grad_norm": 0.4787065386772156
},
{
"step": 5025,
"loss": 1.253096923828125,
"learning_rate": 6.6055471662423e-05,
"grad_norm": 0.4620949923992157
},
{
"step": 5050,
"loss": 1.2833746337890626,
"learning_rate": 6.469429549458793e-05,
"grad_norm": 0.4496528208255768
},
{
"step": 5075,
"loss": 1.2286983489990235,
"learning_rate": 6.334342383903497e-05,
"grad_norm": 0.4651983380317688
},
{
"step": 5100,
"loss": 1.2594075775146485,
"learning_rate": 6.200301987445924e-05,
"grad_norm": 0.49068841338157654
},
{
"step": 5125,
"loss": 1.2662586975097656,
"learning_rate": 6.067324551511042e-05,
"grad_norm": 0.4658633768558502
},
{
"step": 5150,
"loss": 1.2304222106933593,
"learning_rate": 5.935426139123334e-05,
"grad_norm": 0.44955214858055115
},
{
"step": 5175,
"loss": 1.229621353149414,
"learning_rate": 5.80462268296654e-05,
"grad_norm": 0.4408133029937744
},
{
"step": 5200,
"loss": 1.2133523559570312,
"learning_rate": 5.674929983459037e-05,
"grad_norm": 0.44262009859085083
},
{
"step": 5225,
"loss": 1.2221240234375,
"learning_rate": 5.546363706845204e-05,
"grad_norm": 0.43076950311660767
},
{
"step": 5250,
"loss": 1.2376617431640624,
"learning_rate": 5.418939383303067e-05,
"grad_norm": 0.45389309525489807
},
{
"step": 5275,
"loss": 1.2449069976806642,
"learning_rate": 5.2926724050682747e-05,
"grad_norm": 0.48846372961997986
},
{
"step": 5300,
"loss": 1.2519512939453126,
"learning_rate": 5.167578024574836e-05,
"grad_norm": 0.4938410520553589
},
{
"step": 5325,
"loss": 1.1897734832763671,
"learning_rate": 5.0436713526126756e-05,
"grad_norm": 0.4553002417087555
},
{
"step": 5350,
"loss": 1.232283477783203,
"learning_rate": 4.920967356502332e-05,
"grad_norm": 0.49487122893333435
},
{
"step": 5375,
"loss": 1.2140327453613282,
"learning_rate": 4.799480858286981e-05,
"grad_norm": 0.4383367896080017
},
{
"step": 5400,
"loss": 1.2149247741699218,
"learning_rate": 4.679226532941989e-05,
"grad_norm": 0.4817405939102173
},
{
"step": 5425,
"loss": 1.2574878692626954,
"learning_rate": 4.560218906602277e-05,
"grad_norm": 0.4749167859554291
},
{
"step": 5450,
"loss": 1.2151976013183594,
"learning_rate": 4.4424723548076014e-05,
"grad_norm": 0.49512583017349243
},
{
"step": 5475,
"loss": 1.226508560180664,
"learning_rate": 4.32600110076609e-05,
"grad_norm": 0.46990203857421875
},
{
"step": 5500,
"loss": 1.218799819946289,
"learning_rate": 4.210819213636126e-05,
"grad_norm": 0.4672454595565796
},
{
"step": 5525,
"loss": 1.2233743286132812,
"learning_rate": 4.096940606826865e-05,
"grad_norm": 0.45172497630119324
},
{
"step": 5550,
"loss": 1.194942626953125,
"learning_rate": 3.9843790363175865e-05,
"grad_norm": 0.4389333426952362
},
{
"step": 5575,
"loss": 1.2313622283935546,
"learning_rate": 3.8731480989959944e-05,
"grad_norm": 0.46159827709198
},
{
"step": 5600,
"loss": 1.1869064331054688,
"learning_rate": 3.763261231015831e-05,
"grad_norm": 0.4371712803840637
},
{
"step": 5625,
"loss": 1.2109555053710936,
"learning_rate": 3.6547317061738065e-05,
"grad_norm": 0.4667040705680847
},
{
"step": 5650,
"loss": 1.1982355499267578,
"learning_rate": 3.547572634306225e-05,
"grad_norm": 0.4180281460285187
},
{
"step": 5675,
"loss": 1.2502777099609375,
"learning_rate": 3.441796959705373e-05,
"grad_norm": 0.4566985070705414
},
{
"step": 5700,
"loss": 1.2031982421875,
"learning_rate": 3.3374174595558906e-05,
"grad_norm": 0.48893609642982483
},
{
"step": 5725,
"loss": 1.2035820007324218,
"learning_rate": 3.2344467423913764e-05,
"grad_norm": 0.47163310647010803
},
{
"step": 5750,
"loss": 1.2236875152587892,
"learning_rate": 3.1328972465713305e-05,
"grad_norm": 0.4129290282726288
},
{
"step": 5775,
"loss": 1.209072036743164,
"learning_rate": 3.032781238778656e-05,
"grad_norm": 0.5045214295387268
},
{
"step": 5800,
"loss": 1.2049720001220703,
"learning_rate": 2.9341108125378987e-05,
"grad_norm": 1.177187442779541
},
{
"step": 5825,
"loss": 1.2237082672119142,
"learning_rate": 2.836897886754428e-05,
"grad_norm": 0.44473758339881897
},
{
"step": 5850,
"loss": 1.211065444946289,
"learning_rate": 2.741154204274677e-05,
"grad_norm": 0.4674845039844513
},
{
"step": 5875,
"loss": 1.1834144592285156,
"learning_rate": 2.6468913304676588e-05,
"grad_norm": 0.4646168649196625
},
{
"step": 5900,
"loss": 1.2036126708984376,
"learning_rate": 2.5541206518279465e-05,
"grad_norm": 0.4711571931838989
},
{
"step": 5925,
"loss": 1.2229229736328124,
"learning_rate": 2.4628533746002134e-05,
"grad_norm": 0.47303342819213867
},
{
"step": 5950,
"loss": 1.2234363555908203,
"learning_rate": 2.37310052342561e-05,
"grad_norm": 0.429867684841156
},
{
"step": 5975,
"loss": 1.2309221649169921,
"learning_rate": 2.2848729400099936e-05,
"grad_norm": 0.5267869830131531
},
{
"step": 6000,
"loss": 1.2363741302490234,
"learning_rate": 2.198181281814349e-05,
"grad_norm": 0.4759523570537567
},
{
"step": 6025,
"loss": 1.220240478515625,
"learning_rate": 2.1130360207673923e-05,
"grad_norm": 0.4383280575275421
},
{
"step": 6050,
"loss": 1.1791091918945313,
"learning_rate": 2.0294474420006134e-05,
"grad_norm": 0.4518462121486664
},
{
"step": 6075,
"loss": 1.205836639404297,
"learning_rate": 1.9474256426059033e-05,
"grad_norm": 0.4878427982330322
},
{
"step": 6100,
"loss": 1.2177058410644532,
"learning_rate": 1.866980530415844e-05,
"grad_norm": 0.45281848311424255
},
{
"step": 6125,
"loss": 1.2058243560791015,
"learning_rate": 1.7881218228069215e-05,
"grad_norm": 0.467458039522171
},
{
"step": 6150,
"loss": 1.2168304443359375,
"learning_rate": 1.710859045525692e-05,
"grad_norm": 0.4621610939502716
},
{
"step": 6175,
"loss": 1.239758529663086,
"learning_rate": 1.6352015315381306e-05,
"grad_norm": 0.44325652718544006
},
{
"step": 6200,
"loss": 1.2028849029541016,
"learning_rate": 1.5611584199022474e-05,
"grad_norm": 0.44522371888160706
},
{
"step": 6225,
"loss": 1.2188538360595702,
"learning_rate": 1.488738654664135e-05,
"grad_norm": 0.4577750861644745
},
{
"step": 6250,
"loss": 1.2289091491699218,
"learning_rate": 1.4179509837775794e-05,
"grad_norm": 0.44208669662475586
},
{
"step": 6275,
"loss": 1.233956527709961,
"learning_rate": 1.348803958047333e-05,
"grad_norm": 0.4531864821910858
},
{
"step": 6300,
"loss": 1.2187274169921876,
"learning_rate": 1.281305930096243e-05,
"grad_norm": 0.4565826654434204
},
{
"step": 6325,
"loss": 1.2017286682128907,
"learning_rate": 1.215465053356272e-05,
"grad_norm": 0.47588875889778137
},
{
"step": 6350,
"loss": 1.219265899658203,
"learning_rate": 1.1512892810836188e-05,
"grad_norm": 0.4373761713504791
},
{
"step": 6375,
"loss": 1.2092100524902343,
"learning_rate": 1.0887863653980012e-05,
"grad_norm": 0.4859849214553833
},
{
"step": 6400,
"loss": 1.214217987060547,
"learning_rate": 1.0279638563462234e-05,
"grad_norm": 0.4610133469104767
},
{
"step": 6425,
"loss": 1.1977297973632812,
"learning_rate": 9.68829100990191e-06,
"grad_norm": 0.46569377183914185
},
{
"step": 6450,
"loss": 1.2287699127197265,
"learning_rate": 9.113892425194091e-06,
"grad_norm": 0.45601311326026917
},
{
"step": 6475,
"loss": 1.1981573486328125,
"learning_rate": 8.556512193881105e-06,
"grad_norm": 0.46404409408569336
},
{
"step": 6500,
"loss": 1.2465553283691406,
"learning_rate": 8.01621764477146e-06,
"grad_norm": 0.47190892696380615
},
{
"step": 6525,
"loss": 1.197098846435547,
"learning_rate": 7.4930740428066294e-06,
"grad_norm": 0.45320889353752136
},
{
"step": 6550,
"loss": 1.1829496002197266,
"learning_rate": 6.987144581177639e-06,
"grad_norm": 0.44584017992019653
},
{
"step": 6575,
"loss": 1.2278943634033204,
"learning_rate": 6.498490373691289e-06,
"grad_norm": 0.4520955979824066
},
{
"step": 6600,
"loss": 1.2047061157226562,
"learning_rate": 6.027170447388219e-06,
"grad_norm": 0.4502451717853546
},
{
"step": 6625,
"loss": 1.2211988830566407,
"learning_rate": 5.573241735412543e-06,
"grad_norm": 0.4682001769542694
},
{
"step": 6650,
"loss": 1.2268491363525391,
"learning_rate": 5.136759070134638e-06,
"grad_norm": 0.44889962673187256
},
{
"step": 6675,
"loss": 1.2038793182373047,
"learning_rate": 4.7177751765277775e-06,
"grad_norm": 0.44545120000839233
},
{
"step": 6700,
"loss": 1.23889892578125,
"learning_rate": 4.316340665799017e-06,
"grad_norm": 0.42067432403564453
},
{
"step": 6725,
"loss": 1.1800360107421874,
"learning_rate": 3.932504029275779e-06,
"grad_norm": 0.44353973865509033
},
{
"step": 6750,
"loss": 1.202306671142578,
"learning_rate": 3.5663116325482466e-06,
"grad_norm": 0.443708211183548
},
{
"step": 6775,
"loss": 1.1959423828125,
"learning_rate": 3.2178077098686304e-06,
"grad_norm": 0.4556666612625122
},
{
"step": 6800,
"loss": 1.1825554656982422,
"learning_rate": 2.88703435880796e-06,
"grad_norm": 0.45425671339035034
},
{
"step": 6825,
"loss": 1.1897866058349609,
"learning_rate": 2.5740315351707843e-06,
"grad_norm": 0.4512813091278076
},
{
"step": 6850,
"loss": 1.1878171539306641,
"learning_rate": 2.2788370481687965e-06,
"grad_norm": 0.43827131390571594
},
{
"step": 6875,
"loss": 1.2006049346923828,
"learning_rate": 2.0014865558536596e-06,
"grad_norm": 0.4461023211479187
},
{
"step": 6900,
"loss": 1.179116439819336,
"learning_rate": 1.7420135608096297e-06,
"grad_norm": 0.43480661511421204
},
{
"step": 6925,
"loss": 1.2001779937744141,
"learning_rate": 1.5004494061066797e-06,
"grad_norm": 0.4782886207103729
},
{
"step": 6950,
"loss": 1.2102194976806642,
"learning_rate": 1.27682327151436e-06,
"grad_norm": 0.4564678370952606
},
{
"step": 6975,
"loss": 1.174234848022461,
"learning_rate": 1.0711621699770522e-06,
"grad_norm": 0.4488446116447449
},
{
"step": 7000,
"loss": 1.1999745178222656,
"learning_rate": 8.834909443508953e-07,
"grad_norm": 0.458385169506073
},
{
"step": 7025,
"loss": 1.1988882446289062,
"learning_rate": 7.138322644029692e-07,
"grad_norm": 0.44504058361053467
},
{
"step": 7050,
"loss": 1.2174333953857421,
"learning_rate": 5.622066240728029e-07,
"grad_norm": 0.430634081363678
},
{
"step": 7075,
"loss": 1.2466523742675781,
"learning_rate": 4.286323389968871e-07,
"grad_norm": 0.4718645513057709
},
{
"step": 7100,
"loss": 1.216027069091797,
"learning_rate": 3.1312554429617776e-07,
"grad_norm": 0.466147243976593
},
{
"step": 7125,
"loss": 1.2029841613769532,
"learning_rate": 2.1570019262707185e-07,
"grad_norm": 0.45927485823631287
},
{
"step": 7150,
"loss": 1.1904071044921876,
"learning_rate": 1.3636805249603866e-07,
"grad_norm": 0.4847685992717743
},
{
"step": 7175,
"loss": 1.2437645721435546,
"learning_rate": 7.513870683794054e-08,
"grad_norm": 0.46799784898757935
},
{
"step": 7200,
"loss": 1.1522864532470702,
"learning_rate": 3.20195518585753e-08,
"grad_norm": 0.4408682882785797
},
{
"step": 7225,
"loss": 1.1922702026367187,
"learning_rate": 7.015796141157659e-09,
"grad_norm": 0.4327032268047333
}
],
"final_checkpoint": "/content/agron-epoch1/final"
}