feedback_gemma_dirty / trainer_state.json
terry69's picture
Model save
3fda082 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1699,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0005885815185403178,
"grad_norm": 81.18565670358198,
"learning_rate": 5.882352941176471e-08,
"loss": 4.1356,
"step": 1
},
{
"epoch": 0.002942907592701589,
"grad_norm": 77.04540875822474,
"learning_rate": 2.9411764705882356e-07,
"loss": 4.1935,
"step": 5
},
{
"epoch": 0.005885815185403178,
"grad_norm": 35.44829095248985,
"learning_rate": 5.882352941176471e-07,
"loss": 3.9438,
"step": 10
},
{
"epoch": 0.008828722778104767,
"grad_norm": 22.5439620439223,
"learning_rate": 8.823529411764707e-07,
"loss": 3.4429,
"step": 15
},
{
"epoch": 0.011771630370806356,
"grad_norm": 13.2019843638693,
"learning_rate": 1.1764705882352942e-06,
"loss": 3.1438,
"step": 20
},
{
"epoch": 0.014714537963507945,
"grad_norm": 14.772036939700145,
"learning_rate": 1.4705882352941177e-06,
"loss": 2.9141,
"step": 25
},
{
"epoch": 0.017657445556209534,
"grad_norm": 20.5779186199068,
"learning_rate": 1.7647058823529414e-06,
"loss": 2.6074,
"step": 30
},
{
"epoch": 0.020600353148911125,
"grad_norm": 27.382526732783585,
"learning_rate": 2.058823529411765e-06,
"loss": 2.2919,
"step": 35
},
{
"epoch": 0.023543260741612712,
"grad_norm": 20.764560265022364,
"learning_rate": 2.3529411764705885e-06,
"loss": 1.7801,
"step": 40
},
{
"epoch": 0.026486168334314303,
"grad_norm": 11.788637809965795,
"learning_rate": 2.647058823529412e-06,
"loss": 1.3328,
"step": 45
},
{
"epoch": 0.02942907592701589,
"grad_norm": 22.53245322685116,
"learning_rate": 2.9411764705882355e-06,
"loss": 1.1855,
"step": 50
},
{
"epoch": 0.032371983519717484,
"grad_norm": 6.015914744919097,
"learning_rate": 3.2352941176470594e-06,
"loss": 1.1366,
"step": 55
},
{
"epoch": 0.03531489111241907,
"grad_norm": 5.217837662899846,
"learning_rate": 3.529411764705883e-06,
"loss": 1.1055,
"step": 60
},
{
"epoch": 0.03825779870512066,
"grad_norm": 3.357783317857939,
"learning_rate": 3.8235294117647055e-06,
"loss": 1.0953,
"step": 65
},
{
"epoch": 0.04120070629782225,
"grad_norm": 4.996001471413591,
"learning_rate": 4.11764705882353e-06,
"loss": 1.0496,
"step": 70
},
{
"epoch": 0.04414361389052384,
"grad_norm": 8.621809656222444,
"learning_rate": 4.411764705882353e-06,
"loss": 1.0292,
"step": 75
},
{
"epoch": 0.047086521483225424,
"grad_norm": 5.339001901811776,
"learning_rate": 4.705882352941177e-06,
"loss": 1.0423,
"step": 80
},
{
"epoch": 0.050029429075927015,
"grad_norm": 3.493074689207993,
"learning_rate": 5e-06,
"loss": 1.026,
"step": 85
},
{
"epoch": 0.052972336668628606,
"grad_norm": 9.685062179270181,
"learning_rate": 5.294117647058824e-06,
"loss": 1.005,
"step": 90
},
{
"epoch": 0.055915244261330196,
"grad_norm": 6.933487463420427,
"learning_rate": 5.588235294117647e-06,
"loss": 0.9885,
"step": 95
},
{
"epoch": 0.05885815185403178,
"grad_norm": 8.28144947566066,
"learning_rate": 5.882352941176471e-06,
"loss": 0.9873,
"step": 100
},
{
"epoch": 0.06180105944673337,
"grad_norm": 17.077053218905366,
"learning_rate": 6.176470588235295e-06,
"loss": 0.9694,
"step": 105
},
{
"epoch": 0.06474396703943497,
"grad_norm": 19.687199094462986,
"learning_rate": 6.470588235294119e-06,
"loss": 0.9622,
"step": 110
},
{
"epoch": 0.06768687463213655,
"grad_norm": 13.804537783692052,
"learning_rate": 6.764705882352942e-06,
"loss": 0.9563,
"step": 115
},
{
"epoch": 0.07062978222483814,
"grad_norm": 10.94725198202626,
"learning_rate": 7.058823529411766e-06,
"loss": 0.9414,
"step": 120
},
{
"epoch": 0.07357268981753973,
"grad_norm": 18.46306700568472,
"learning_rate": 7.352941176470589e-06,
"loss": 0.9289,
"step": 125
},
{
"epoch": 0.07651559741024132,
"grad_norm": 14.894213258364779,
"learning_rate": 7.647058823529411e-06,
"loss": 0.9238,
"step": 130
},
{
"epoch": 0.07945850500294291,
"grad_norm": 7.569487604501164,
"learning_rate": 7.941176470588236e-06,
"loss": 0.9183,
"step": 135
},
{
"epoch": 0.0824014125956445,
"grad_norm": 21.99231321792487,
"learning_rate": 8.23529411764706e-06,
"loss": 0.911,
"step": 140
},
{
"epoch": 0.08534432018834609,
"grad_norm": 18.115769647379153,
"learning_rate": 8.529411764705883e-06,
"loss": 0.9047,
"step": 145
},
{
"epoch": 0.08828722778104768,
"grad_norm": 10.705484744227235,
"learning_rate": 8.823529411764707e-06,
"loss": 0.8932,
"step": 150
},
{
"epoch": 0.09123013537374926,
"grad_norm": 16.313242301856516,
"learning_rate": 9.11764705882353e-06,
"loss": 0.8899,
"step": 155
},
{
"epoch": 0.09417304296645085,
"grad_norm": 12.978636094048102,
"learning_rate": 9.411764705882354e-06,
"loss": 0.8916,
"step": 160
},
{
"epoch": 0.09711595055915244,
"grad_norm": 11.506756600035605,
"learning_rate": 9.705882352941177e-06,
"loss": 0.8789,
"step": 165
},
{
"epoch": 0.10005885815185403,
"grad_norm": 24.989193177477567,
"learning_rate": 1e-05,
"loss": 0.8768,
"step": 170
},
{
"epoch": 0.10300176574455562,
"grad_norm": 13.397711202568765,
"learning_rate": 9.99973614764657e-06,
"loss": 0.8801,
"step": 175
},
{
"epoch": 0.10594467333725721,
"grad_norm": 12.037309022627504,
"learning_rate": 9.99894461843351e-06,
"loss": 0.8731,
"step": 180
},
{
"epoch": 0.1088875809299588,
"grad_norm": 22.40769728661026,
"learning_rate": 9.997625495899555e-06,
"loss": 0.8712,
"step": 185
},
{
"epoch": 0.11183048852266039,
"grad_norm": 14.584492756570954,
"learning_rate": 9.99577891926614e-06,
"loss": 0.8491,
"step": 190
},
{
"epoch": 0.11477339611536198,
"grad_norm": 9.426683321327127,
"learning_rate": 9.9934050834227e-06,
"loss": 0.8674,
"step": 195
},
{
"epoch": 0.11771630370806356,
"grad_norm": 25.968161707891955,
"learning_rate": 9.990504238906107e-06,
"loss": 0.8573,
"step": 200
},
{
"epoch": 0.12065921130076515,
"grad_norm": 13.072895533976467,
"learning_rate": 9.98707669187422e-06,
"loss": 0.8525,
"step": 205
},
{
"epoch": 0.12360211889346674,
"grad_norm": 11.095395155610879,
"learning_rate": 9.98312280407358e-06,
"loss": 0.8553,
"step": 210
},
{
"epoch": 0.12654502648616833,
"grad_norm": 16.418159232638153,
"learning_rate": 9.978642992801229e-06,
"loss": 0.8555,
"step": 215
},
{
"epoch": 0.12948793407886994,
"grad_norm": 4.450231202820506,
"learning_rate": 9.973637730860664e-06,
"loss": 0.8413,
"step": 220
},
{
"epoch": 0.13243084167157151,
"grad_norm": 4.111542494606179,
"learning_rate": 9.968107546511942e-06,
"loss": 0.8424,
"step": 225
},
{
"epoch": 0.1353737492642731,
"grad_norm": 13.693095667132473,
"learning_rate": 9.962053023415926e-06,
"loss": 0.8561,
"step": 230
},
{
"epoch": 0.1383166568569747,
"grad_norm": 5.646518368088086,
"learning_rate": 9.955474800572684e-06,
"loss": 0.8636,
"step": 235
},
{
"epoch": 0.14125956444967627,
"grad_norm": 5.108991758643153,
"learning_rate": 9.948373572254044e-06,
"loss": 0.8503,
"step": 240
},
{
"epoch": 0.14420247204237788,
"grad_norm": 9.81020252703176,
"learning_rate": 9.94075008793033e-06,
"loss": 0.873,
"step": 245
},
{
"epoch": 0.14714537963507945,
"grad_norm": 7.935285563014131,
"learning_rate": 9.932605152191255e-06,
"loss": 0.8433,
"step": 250
},
{
"epoch": 0.15008828722778106,
"grad_norm": 3.3393405756912484,
"learning_rate": 9.923939624661005e-06,
"loss": 0.8296,
"step": 255
},
{
"epoch": 0.15303119482048264,
"grad_norm": 3.830858966207542,
"learning_rate": 9.914754419907508e-06,
"loss": 0.8317,
"step": 260
},
{
"epoch": 0.1559741024131842,
"grad_norm": 16.404195583675744,
"learning_rate": 9.905050507345926e-06,
"loss": 0.8419,
"step": 265
},
{
"epoch": 0.15891701000588582,
"grad_norm": 9.685851257998209,
"learning_rate": 9.894828911136323e-06,
"loss": 0.8649,
"step": 270
},
{
"epoch": 0.1618599175985874,
"grad_norm": 5.518903601787907,
"learning_rate": 9.884090710075585e-06,
"loss": 0.8508,
"step": 275
},
{
"epoch": 0.164802825191289,
"grad_norm": 6.557040852323014,
"learning_rate": 9.87283703748356e-06,
"loss": 0.8537,
"step": 280
},
{
"epoch": 0.16774573278399058,
"grad_norm": 6.51479177962607,
"learning_rate": 9.861069081083452e-06,
"loss": 0.8454,
"step": 285
},
{
"epoch": 0.17068864037669218,
"grad_norm": 4.48759740362246,
"learning_rate": 9.84878808287645e-06,
"loss": 0.8494,
"step": 290
},
{
"epoch": 0.17363154796939376,
"grad_norm": 3.077914439009112,
"learning_rate": 9.835995339010673e-06,
"loss": 0.8371,
"step": 295
},
{
"epoch": 0.17657445556209536,
"grad_norm": 7.741990389634322,
"learning_rate": 9.822692199644346e-06,
"loss": 0.8489,
"step": 300
},
{
"epoch": 0.17951736315479694,
"grad_norm": 7.12110354743661,
"learning_rate": 9.808880068803323e-06,
"loss": 0.8386,
"step": 305
},
{
"epoch": 0.18246027074749852,
"grad_norm": 4.874821068849924,
"learning_rate": 9.794560404232898e-06,
"loss": 0.8399,
"step": 310
},
{
"epoch": 0.18540317834020012,
"grad_norm": 4.563419733002005,
"learning_rate": 9.779734717243946e-06,
"loss": 0.8271,
"step": 315
},
{
"epoch": 0.1883460859329017,
"grad_norm": 10.93234206126768,
"learning_rate": 9.764404572553427e-06,
"loss": 0.8123,
"step": 320
},
{
"epoch": 0.1912889935256033,
"grad_norm": 9.419206687007401,
"learning_rate": 9.748571588119247e-06,
"loss": 0.8188,
"step": 325
},
{
"epoch": 0.19423190111830488,
"grad_norm": 6.977872680215194,
"learning_rate": 9.732237434969487e-06,
"loss": 0.8255,
"step": 330
},
{
"epoch": 0.19717480871100648,
"grad_norm": 5.908304115359454,
"learning_rate": 9.715403837026046e-06,
"loss": 0.8265,
"step": 335
},
{
"epoch": 0.20011771630370806,
"grad_norm": 6.797809972888078,
"learning_rate": 9.698072570922698e-06,
"loss": 0.8118,
"step": 340
},
{
"epoch": 0.20306062389640966,
"grad_norm": 6.704121135386284,
"learning_rate": 9.680245465817583e-06,
"loss": 0.8096,
"step": 345
},
{
"epoch": 0.20600353148911124,
"grad_norm": 8.645898444887452,
"learning_rate": 9.661924403200155e-06,
"loss": 0.8038,
"step": 350
},
{
"epoch": 0.20894643908181282,
"grad_norm": 6.272395534856629,
"learning_rate": 9.64311131669261e-06,
"loss": 0.7848,
"step": 355
},
{
"epoch": 0.21188934667451442,
"grad_norm": 7.0264316123358155,
"learning_rate": 9.623808191845806e-06,
"loss": 0.7885,
"step": 360
},
{
"epoch": 0.214832254267216,
"grad_norm": 6.754946681714514,
"learning_rate": 9.604017065929715e-06,
"loss": 0.7937,
"step": 365
},
{
"epoch": 0.2177751618599176,
"grad_norm": 6.611299682141483,
"learning_rate": 9.583740027718396e-06,
"loss": 0.8036,
"step": 370
},
{
"epoch": 0.22071806945261918,
"grad_norm": 4.018544136474155,
"learning_rate": 9.562979217269547e-06,
"loss": 0.7969,
"step": 375
},
{
"epoch": 0.22366097704532079,
"grad_norm": 7.101373596601596,
"learning_rate": 9.54173682569865e-06,
"loss": 0.796,
"step": 380
},
{
"epoch": 0.22660388463802236,
"grad_norm": 6.399510185421723,
"learning_rate": 9.520015094947704e-06,
"loss": 0.8075,
"step": 385
},
{
"epoch": 0.22954679223072397,
"grad_norm": 9.296984628579494,
"learning_rate": 9.497816317548625e-06,
"loss": 0.7799,
"step": 390
},
{
"epoch": 0.23248969982342554,
"grad_norm": 4.786255899468576,
"learning_rate": 9.475142836381274e-06,
"loss": 0.7695,
"step": 395
},
{
"epoch": 0.23543260741612712,
"grad_norm": 4.559456259891477,
"learning_rate": 9.4519970444262e-06,
"loss": 0.7741,
"step": 400
},
{
"epoch": 0.23837551500882873,
"grad_norm": 7.466092853647954,
"learning_rate": 9.428381384512073e-06,
"loss": 0.7619,
"step": 405
},
{
"epoch": 0.2413184226015303,
"grad_norm": 6.623530673850783,
"learning_rate": 9.404298349057873e-06,
"loss": 0.7731,
"step": 410
},
{
"epoch": 0.2442613301942319,
"grad_norm": 6.068028651768666,
"learning_rate": 9.379750479809832e-06,
"loss": 0.7574,
"step": 415
},
{
"epoch": 0.24720423778693348,
"grad_norm": 8.419516596990583,
"learning_rate": 9.354740367573178e-06,
"loss": 0.7821,
"step": 420
},
{
"epoch": 0.25014714537963506,
"grad_norm": 4.198177608767893,
"learning_rate": 9.329270651938702e-06,
"loss": 0.7547,
"step": 425
},
{
"epoch": 0.25309005297233667,
"grad_norm": 4.566575750438902,
"learning_rate": 9.303344021004169e-06,
"loss": 0.7664,
"step": 430
},
{
"epoch": 0.25603296056503827,
"grad_norm": 8.680313490795726,
"learning_rate": 9.27696321109061e-06,
"loss": 0.7668,
"step": 435
},
{
"epoch": 0.2589758681577399,
"grad_norm": 9.12784954370401,
"learning_rate": 9.250131006453544e-06,
"loss": 0.7678,
"step": 440
},
{
"epoch": 0.2619187757504414,
"grad_norm": 10.144210993180703,
"learning_rate": 9.222850238989104e-06,
"loss": 0.765,
"step": 445
},
{
"epoch": 0.26486168334314303,
"grad_norm": 3.0429868552146964,
"learning_rate": 9.195123787935168e-06,
"loss": 0.7445,
"step": 450
},
{
"epoch": 0.26780459093584463,
"grad_norm": 4.691784082396774,
"learning_rate": 9.166954579567483e-06,
"loss": 0.7649,
"step": 455
},
{
"epoch": 0.2707474985285462,
"grad_norm": 4.7416354261595615,
"learning_rate": 9.138345586890818e-06,
"loss": 0.7496,
"step": 460
},
{
"epoch": 0.2736904061212478,
"grad_norm": 6.223772502898528,
"learning_rate": 9.109299829325191e-06,
"loss": 0.7439,
"step": 465
},
{
"epoch": 0.2766333137139494,
"grad_norm": 5.069169897647759,
"learning_rate": 9.0798203723872e-06,
"loss": 0.7552,
"step": 470
},
{
"epoch": 0.279576221306651,
"grad_norm": 4.067678548778218,
"learning_rate": 9.049910327366481e-06,
"loss": 0.7479,
"step": 475
},
{
"epoch": 0.28251912889935255,
"grad_norm": 7.216787820352084,
"learning_rate": 9.019572850997339e-06,
"loss": 0.7456,
"step": 480
},
{
"epoch": 0.28546203649205415,
"grad_norm": 6.118517874215981,
"learning_rate": 8.988811145125592e-06,
"loss": 0.7373,
"step": 485
},
{
"epoch": 0.28840494408475575,
"grad_norm": 6.974203549951666,
"learning_rate": 8.957628456370636e-06,
"loss": 0.7544,
"step": 490
},
{
"epoch": 0.2913478516774573,
"grad_norm": 6.659489098506463,
"learning_rate": 8.926028075782791e-06,
"loss": 0.7485,
"step": 495
},
{
"epoch": 0.2942907592701589,
"grad_norm": 6.963976313909436,
"learning_rate": 8.89401333849598e-06,
"loss": 0.7364,
"step": 500
},
{
"epoch": 0.2972336668628605,
"grad_norm": 5.060368338459147,
"learning_rate": 8.861587623375711e-06,
"loss": 0.7335,
"step": 505
},
{
"epoch": 0.3001765744555621,
"grad_norm": 11.693154707714637,
"learning_rate": 8.828754352662483e-06,
"loss": 0.7469,
"step": 510
},
{
"epoch": 0.30311948204826367,
"grad_norm": 8.708261644979073,
"learning_rate": 8.795516991610592e-06,
"loss": 0.7296,
"step": 515
},
{
"epoch": 0.30606238964096527,
"grad_norm": 5.926416635055216,
"learning_rate": 8.761879048122414e-06,
"loss": 0.7443,
"step": 520
},
{
"epoch": 0.3090052972336669,
"grad_norm": 12.904699986898022,
"learning_rate": 8.727844072378173e-06,
"loss": 0.7182,
"step": 525
},
{
"epoch": 0.3119482048263684,
"grad_norm": 6.368508385097698,
"learning_rate": 8.693415656461244e-06,
"loss": 0.734,
"step": 530
},
{
"epoch": 0.31489111241907003,
"grad_norm": 3.3588114427301607,
"learning_rate": 8.658597433979057e-06,
"loss": 0.7248,
"step": 535
},
{
"epoch": 0.31783402001177163,
"grad_norm": 12.331947606606729,
"learning_rate": 8.623393079679588e-06,
"loss": 0.7344,
"step": 540
},
{
"epoch": 0.32077692760447324,
"grad_norm": 12.166702155420914,
"learning_rate": 8.587806309063531e-06,
"loss": 0.726,
"step": 545
},
{
"epoch": 0.3237198351971748,
"grad_norm": 4.612832603972562,
"learning_rate": 8.551840877992155e-06,
"loss": 0.7283,
"step": 550
},
{
"epoch": 0.3266627427898764,
"grad_norm": 5.77424459107182,
"learning_rate": 8.515500582290914e-06,
"loss": 0.7155,
"step": 555
},
{
"epoch": 0.329605650382578,
"grad_norm": 3.6459134260142214,
"learning_rate": 8.478789257348827e-06,
"loss": 0.7116,
"step": 560
},
{
"epoch": 0.3325485579752796,
"grad_norm": 3.0011507986582733,
"learning_rate": 8.441710777713683e-06,
"loss": 0.7037,
"step": 565
},
{
"epoch": 0.33549146556798115,
"grad_norm": 5.360195180614321,
"learning_rate": 8.404269056683133e-06,
"loss": 0.7169,
"step": 570
},
{
"epoch": 0.33843437316068276,
"grad_norm": 5.707195645216086,
"learning_rate": 8.366468045891658e-06,
"loss": 0.7102,
"step": 575
},
{
"epoch": 0.34137728075338436,
"grad_norm": 3.5974238955778457,
"learning_rate": 8.32831173489352e-06,
"loss": 0.7115,
"step": 580
},
{
"epoch": 0.3443201883460859,
"grad_norm": 3.2599056325905367,
"learning_rate": 8.289804150741704e-06,
"loss": 0.7154,
"step": 585
},
{
"epoch": 0.3472630959387875,
"grad_norm": 4.229649245193178,
"learning_rate": 8.250949357562892e-06,
"loss": 0.7057,
"step": 590
},
{
"epoch": 0.3502060035314891,
"grad_norm": 5.194627560952757,
"learning_rate": 8.211751456128532e-06,
"loss": 0.7251,
"step": 595
},
{
"epoch": 0.3531489111241907,
"grad_norm": 3.0868739009823583,
"learning_rate": 8.172214583422038e-06,
"loss": 0.7067,
"step": 600
},
{
"epoch": 0.35609181871689227,
"grad_norm": 4.2974502097872405,
"learning_rate": 8.132342912202178e-06,
"loss": 0.7144,
"step": 605
},
{
"epoch": 0.3590347263095939,
"grad_norm": 3.271991629082987,
"learning_rate": 8.092140650562665e-06,
"loss": 0.7103,
"step": 610
},
{
"epoch": 0.3619776339022955,
"grad_norm": 4.2083604726976915,
"learning_rate": 8.051612041488038e-06,
"loss": 0.7211,
"step": 615
},
{
"epoch": 0.36492054149499703,
"grad_norm": 3.6340200304717114,
"learning_rate": 8.010761362405851e-06,
"loss": 0.7272,
"step": 620
},
{
"epoch": 0.36786344908769864,
"grad_norm": 5.783094842613372,
"learning_rate": 7.969592924735234e-06,
"loss": 0.7054,
"step": 625
},
{
"epoch": 0.37080635668040024,
"grad_norm": 5.0424325252776745,
"learning_rate": 7.928111073431847e-06,
"loss": 0.7122,
"step": 630
},
{
"epoch": 0.37374926427310184,
"grad_norm": 5.196701737520978,
"learning_rate": 7.886320186529327e-06,
"loss": 0.6953,
"step": 635
},
{
"epoch": 0.3766921718658034,
"grad_norm": 3.608090265928746,
"learning_rate": 7.844224674677225e-06,
"loss": 0.6978,
"step": 640
},
{
"epoch": 0.379635079458505,
"grad_norm": 4.731749503563869,
"learning_rate": 7.801828980675485e-06,
"loss": 0.6953,
"step": 645
},
{
"epoch": 0.3825779870512066,
"grad_norm": 3.0237351828264547,
"learning_rate": 7.75913757900556e-06,
"loss": 0.6738,
"step": 650
},
{
"epoch": 0.3855208946439082,
"grad_norm": 3.0717508997327365,
"learning_rate": 7.716154975358174e-06,
"loss": 0.7046,
"step": 655
},
{
"epoch": 0.38846380223660976,
"grad_norm": 3.4394705043914477,
"learning_rate": 7.672885706157778e-06,
"loss": 0.6982,
"step": 660
},
{
"epoch": 0.39140670982931136,
"grad_norm": 3.354889681681399,
"learning_rate": 7.629334338083774e-06,
"loss": 0.665,
"step": 665
},
{
"epoch": 0.39434961742201297,
"grad_norm": 4.968268480091777,
"learning_rate": 7.585505467588547e-06,
"loss": 0.6836,
"step": 670
},
{
"epoch": 0.3972925250147145,
"grad_norm": 3.4164646148678917,
"learning_rate": 7.541403720412353e-06,
"loss": 0.6951,
"step": 675
},
{
"epoch": 0.4002354326074161,
"grad_norm": 5.683831574907592,
"learning_rate": 7.497033751095099e-06,
"loss": 0.6928,
"step": 680
},
{
"epoch": 0.4031783402001177,
"grad_norm": 4.472151489983979,
"learning_rate": 7.4524002424851185e-06,
"loss": 0.6797,
"step": 685
},
{
"epoch": 0.40612124779281933,
"grad_norm": 6.186377434937082,
"learning_rate": 7.407507905244924e-06,
"loss": 0.672,
"step": 690
},
{
"epoch": 0.4090641553855209,
"grad_norm": 2.9651290505497045,
"learning_rate": 7.362361477354052e-06,
"loss": 0.6791,
"step": 695
},
{
"epoch": 0.4120070629782225,
"grad_norm": 5.390553555502008,
"learning_rate": 7.3169657236090004e-06,
"loss": 0.6689,
"step": 700
},
{
"epoch": 0.4149499705709241,
"grad_norm": 4.10914126631982,
"learning_rate": 7.271325435120353e-06,
"loss": 0.6845,
"step": 705
},
{
"epoch": 0.41789287816362564,
"grad_norm": 8.644264069486146,
"learning_rate": 7.225445428807121e-06,
"loss": 0.6791,
"step": 710
},
{
"epoch": 0.42083578575632724,
"grad_norm": 4.677705564927546,
"learning_rate": 7.179330546888364e-06,
"loss": 0.6713,
"step": 715
},
{
"epoch": 0.42377869334902885,
"grad_norm": 6.154537296560613,
"learning_rate": 7.132985656372126e-06,
"loss": 0.6579,
"step": 720
},
{
"epoch": 0.42672160094173045,
"grad_norm": 8.540942755538525,
"learning_rate": 7.086415648541782e-06,
"loss": 0.6576,
"step": 725
},
{
"epoch": 0.429664508534432,
"grad_norm": 5.046772189845541,
"learning_rate": 7.0396254384398e-06,
"loss": 0.6435,
"step": 730
},
{
"epoch": 0.4326074161271336,
"grad_norm": 3.4905163179341967,
"learning_rate": 6.9926199643489985e-06,
"loss": 0.6405,
"step": 735
},
{
"epoch": 0.4355503237198352,
"grad_norm": 3.005636207711306,
"learning_rate": 6.945404187271365e-06,
"loss": 0.6461,
"step": 740
},
{
"epoch": 0.4384932313125368,
"grad_norm": 3.4654929507322803,
"learning_rate": 6.897983090404457e-06,
"loss": 0.6239,
"step": 745
},
{
"epoch": 0.44143613890523836,
"grad_norm": 3.6846119188733666,
"learning_rate": 6.850361678615481e-06,
"loss": 0.6548,
"step": 750
},
{
"epoch": 0.44437904649793997,
"grad_norm": 3.2356789677548106,
"learning_rate": 6.802544977913067e-06,
"loss": 0.6741,
"step": 755
},
{
"epoch": 0.44732195409064157,
"grad_norm": 4.998250903563615,
"learning_rate": 6.754538034916819e-06,
"loss": 0.6631,
"step": 760
},
{
"epoch": 0.4502648616833431,
"grad_norm": 5.140876271152121,
"learning_rate": 6.706345916324692e-06,
"loss": 0.6458,
"step": 765
},
{
"epoch": 0.4532077692760447,
"grad_norm": 3.486582775324299,
"learning_rate": 6.657973708378251e-06,
"loss": 0.6468,
"step": 770
},
{
"epoch": 0.45615067686874633,
"grad_norm": 3.2392766779187543,
"learning_rate": 6.609426516325859e-06,
"loss": 0.6453,
"step": 775
},
{
"epoch": 0.45909358446144793,
"grad_norm": 8.066053393796732,
"learning_rate": 6.560709463883865e-06,
"loss": 0.629,
"step": 780
},
{
"epoch": 0.4620364920541495,
"grad_norm": 4.508373615955478,
"learning_rate": 6.511827692695847e-06,
"loss": 0.6364,
"step": 785
},
{
"epoch": 0.4649793996468511,
"grad_norm": 3.7473795272990755,
"learning_rate": 6.462786361789948e-06,
"loss": 0.6499,
"step": 790
},
{
"epoch": 0.4679223072395527,
"grad_norm": 2.9615724505608543,
"learning_rate": 6.413590647034401e-06,
"loss": 0.6346,
"step": 795
},
{
"epoch": 0.47086521483225424,
"grad_norm": 3.04892457656011,
"learning_rate": 6.364245740591251e-06,
"loss": 0.6322,
"step": 800
},
{
"epoch": 0.47380812242495585,
"grad_norm": 2.918336925557892,
"learning_rate": 6.314756850368377e-06,
"loss": 0.648,
"step": 805
},
{
"epoch": 0.47675103001765745,
"grad_norm": 3.7501594414028516,
"learning_rate": 6.2651291994698425e-06,
"loss": 0.6372,
"step": 810
},
{
"epoch": 0.47969393761035906,
"grad_norm": 2.6175973358406597,
"learning_rate": 6.215368025644637e-06,
"loss": 0.6233,
"step": 815
},
{
"epoch": 0.4826368452030606,
"grad_norm": 3.216759692623588,
"learning_rate": 6.1654785807338925e-06,
"loss": 0.6261,
"step": 820
},
{
"epoch": 0.4855797527957622,
"grad_norm": 4.2663028236726,
"learning_rate": 6.11546613011659e-06,
"loss": 0.6402,
"step": 825
},
{
"epoch": 0.4885226603884638,
"grad_norm": 3.0098989981042683,
"learning_rate": 6.065335952153846e-06,
"loss": 0.6283,
"step": 830
},
{
"epoch": 0.4914655679811654,
"grad_norm": 2.9805289997816997,
"learning_rate": 6.015093337631833e-06,
"loss": 0.625,
"step": 835
},
{
"epoch": 0.49440847557386697,
"grad_norm": 2.828079466519479,
"learning_rate": 5.964743589203388e-06,
"loss": 0.6271,
"step": 840
},
{
"epoch": 0.4973513831665686,
"grad_norm": 3.056832170036487,
"learning_rate": 5.9142920208283564e-06,
"loss": 0.6163,
"step": 845
},
{
"epoch": 0.5002942907592701,
"grad_norm": 4.967318743335985,
"learning_rate": 5.863743957212759e-06,
"loss": 0.6231,
"step": 850
},
{
"epoch": 0.5032371983519718,
"grad_norm": 7.691396927048123,
"learning_rate": 5.8131047332468106e-06,
"loss": 0.6279,
"step": 855
},
{
"epoch": 0.5061801059446733,
"grad_norm": 10.485146930542516,
"learning_rate": 5.762379693441883e-06,
"loss": 0.617,
"step": 860
},
{
"epoch": 0.5091230135373749,
"grad_norm": 5.793807677023011,
"learning_rate": 5.711574191366427e-06,
"loss": 0.6117,
"step": 865
},
{
"epoch": 0.5120659211300765,
"grad_norm": 3.75783456933447,
"learning_rate": 5.660693589080958e-06,
"loss": 0.6083,
"step": 870
},
{
"epoch": 0.5150088287227781,
"grad_norm": 4.787442679225073,
"learning_rate": 5.60974325657214e-06,
"loss": 0.6062,
"step": 875
},
{
"epoch": 0.5179517363154797,
"grad_norm": 4.424520409909319,
"learning_rate": 5.558728571186027e-06,
"loss": 0.6087,
"step": 880
},
{
"epoch": 0.5208946439081813,
"grad_norm": 4.60877448542516,
"learning_rate": 5.507654917060541e-06,
"loss": 0.6144,
"step": 885
},
{
"epoch": 0.5238375515008828,
"grad_norm": 5.463543352263803,
"learning_rate": 5.456527684557217e-06,
"loss": 0.6234,
"step": 890
},
{
"epoch": 0.5267804590935845,
"grad_norm": 4.713832072636091,
"learning_rate": 5.405352269692301e-06,
"loss": 0.6036,
"step": 895
},
{
"epoch": 0.5297233666862861,
"grad_norm": 3.074792807909731,
"learning_rate": 5.354134073567255e-06,
"loss": 0.5953,
"step": 900
},
{
"epoch": 0.5326662742789876,
"grad_norm": 2.624507780176144,
"learning_rate": 5.302878501798714e-06,
"loss": 0.6024,
"step": 905
},
{
"epoch": 0.5356091818716893,
"grad_norm": 2.4586044761464505,
"learning_rate": 5.2515909639479715e-06,
"loss": 0.583,
"step": 910
},
{
"epoch": 0.5385520894643908,
"grad_norm": 3.1643013838767318,
"learning_rate": 5.200276872950052e-06,
"loss": 0.5848,
"step": 915
},
{
"epoch": 0.5414949970570924,
"grad_norm": 5.524616335650943,
"learning_rate": 5.148941644542428e-06,
"loss": 0.6066,
"step": 920
},
{
"epoch": 0.544437904649794,
"grad_norm": 2.9626617937595383,
"learning_rate": 5.09759069669343e-06,
"loss": 0.5868,
"step": 925
},
{
"epoch": 0.5473808122424956,
"grad_norm": 4.179026591658555,
"learning_rate": 5.046229449030432e-06,
"loss": 0.5938,
"step": 930
},
{
"epoch": 0.5503237198351971,
"grad_norm": 2.9736230937939436,
"learning_rate": 4.994863322267865e-06,
"loss": 0.5854,
"step": 935
},
{
"epoch": 0.5532666274278988,
"grad_norm": 2.6674395025105384,
"learning_rate": 4.943497737635103e-06,
"loss": 0.5773,
"step": 940
},
{
"epoch": 0.5562095350206003,
"grad_norm": 3.3999264932445965,
"learning_rate": 4.892138116304297e-06,
"loss": 0.5899,
"step": 945
},
{
"epoch": 0.559152442613302,
"grad_norm": 3.7946695534722954,
"learning_rate": 4.840789878818238e-06,
"loss": 0.5909,
"step": 950
},
{
"epoch": 0.5620953502060035,
"grad_norm": 2.9170374244463777,
"learning_rate": 4.789458444518243e-06,
"loss": 0.585,
"step": 955
},
{
"epoch": 0.5650382577987051,
"grad_norm": 3.804085500882333,
"learning_rate": 4.73814923097221e-06,
"loss": 0.5811,
"step": 960
},
{
"epoch": 0.5679811653914068,
"grad_norm": 2.627248184612479,
"learning_rate": 4.6868676534028415e-06,
"loss": 0.5741,
"step": 965
},
{
"epoch": 0.5709240729841083,
"grad_norm": 2.758710046147963,
"learning_rate": 4.635619124116105e-06,
"loss": 0.5751,
"step": 970
},
{
"epoch": 0.5738669805768098,
"grad_norm": 2.533940367785718,
"learning_rate": 4.584409051930027e-06,
"loss": 0.5727,
"step": 975
},
{
"epoch": 0.5768098881695115,
"grad_norm": 4.7034871762477755,
"learning_rate": 4.533242841603836e-06,
"loss": 0.5958,
"step": 980
},
{
"epoch": 0.5797527957622131,
"grad_norm": 3.9130158274657205,
"learning_rate": 4.4821258932675324e-06,
"loss": 0.5752,
"step": 985
},
{
"epoch": 0.5826957033549146,
"grad_norm": 2.923581564704699,
"learning_rate": 4.431063601851965e-06,
"loss": 0.5855,
"step": 990
},
{
"epoch": 0.5856386109476163,
"grad_norm": 2.68100418754764,
"learning_rate": 4.38006135651944e-06,
"loss": 0.5592,
"step": 995
},
{
"epoch": 0.5885815185403178,
"grad_norm": 2.5427064423534054,
"learning_rate": 4.329124540094939e-06,
"loss": 0.5738,
"step": 1000
},
{
"epoch": 0.5915244261330195,
"grad_norm": 2.56361328953547,
"learning_rate": 4.278258528498023e-06,
"loss": 0.5518,
"step": 1005
},
{
"epoch": 0.594467333725721,
"grad_norm": 3.570826652815698,
"learning_rate": 4.227468690175433e-06,
"loss": 0.5565,
"step": 1010
},
{
"epoch": 0.5974102413184226,
"grad_norm": 3.144931194434484,
"learning_rate": 4.17676038553452e-06,
"loss": 0.5542,
"step": 1015
},
{
"epoch": 0.6003531489111242,
"grad_norm": 3.0428538279770487,
"learning_rate": 4.126138966377495e-06,
"loss": 0.5724,
"step": 1020
},
{
"epoch": 0.6032960565038258,
"grad_norm": 2.9781764362393863,
"learning_rate": 4.075609775336585e-06,
"loss": 0.5651,
"step": 1025
},
{
"epoch": 0.6062389640965273,
"grad_norm": 2.8792302508241896,
"learning_rate": 4.02517814531018e-06,
"loss": 0.5538,
"step": 1030
},
{
"epoch": 0.609181871689229,
"grad_norm": 2.852533794505324,
"learning_rate": 3.974849398899991e-06,
"loss": 0.5488,
"step": 1035
},
{
"epoch": 0.6121247792819305,
"grad_norm": 3.645969861195433,
"learning_rate": 3.924628847849287e-06,
"loss": 0.5534,
"step": 1040
},
{
"epoch": 0.6150676868746321,
"grad_norm": 2.4948233282423002,
"learning_rate": 3.874521792482305e-06,
"loss": 0.5661,
"step": 1045
},
{
"epoch": 0.6180105944673338,
"grad_norm": 2.6660019183479577,
"learning_rate": 3.8245335211448404e-06,
"loss": 0.5601,
"step": 1050
},
{
"epoch": 0.6209535020600353,
"grad_norm": 3.7571638678391346,
"learning_rate": 3.7746693096461027e-06,
"loss": 0.5522,
"step": 1055
},
{
"epoch": 0.6238964096527368,
"grad_norm": 2.891024598849757,
"learning_rate": 3.7249344207019163e-06,
"loss": 0.5384,
"step": 1060
},
{
"epoch": 0.6268393172454385,
"grad_norm": 3.117981111832482,
"learning_rate": 3.675334103379282e-06,
"loss": 0.5297,
"step": 1065
},
{
"epoch": 0.6297822248381401,
"grad_norm": 2.471751054844093,
"learning_rate": 3.625873592542377e-06,
"loss": 0.5488,
"step": 1070
},
{
"epoch": 0.6327251324308417,
"grad_norm": 2.9406065713758545,
"learning_rate": 3.576558108300081e-06,
"loss": 0.5406,
"step": 1075
},
{
"epoch": 0.6356680400235433,
"grad_norm": 2.5304758647440466,
"learning_rate": 3.5273928554550218e-06,
"loss": 0.5377,
"step": 1080
},
{
"epoch": 0.6386109476162448,
"grad_norm": 4.358193416602383,
"learning_rate": 3.478383022954267e-06,
"loss": 0.5262,
"step": 1085
},
{
"epoch": 0.6415538552089465,
"grad_norm": 3.751043254388853,
"learning_rate": 3.4295337833416797e-06,
"loss": 0.5347,
"step": 1090
},
{
"epoch": 0.644496762801648,
"grad_norm": 2.8478083940715058,
"learning_rate": 3.3808502922119884e-06,
"loss": 0.5517,
"step": 1095
},
{
"epoch": 0.6474396703943496,
"grad_norm": 3.592053051641115,
"learning_rate": 3.3323376876666787e-06,
"loss": 0.5185,
"step": 1100
},
{
"epoch": 0.6503825779870512,
"grad_norm": 3.2832783558616523,
"learning_rate": 3.2840010897717045e-06,
"loss": 0.5283,
"step": 1105
},
{
"epoch": 0.6533254855797528,
"grad_norm": 3.7958128452191313,
"learning_rate": 3.235845600017107e-06,
"loss": 0.5304,
"step": 1110
},
{
"epoch": 0.6562683931724543,
"grad_norm": 2.458980007684118,
"learning_rate": 3.1878763007786075e-06,
"loss": 0.5119,
"step": 1115
},
{
"epoch": 0.659211300765156,
"grad_norm": 5.356928896927627,
"learning_rate": 3.1400982547812088e-06,
"loss": 0.5356,
"step": 1120
},
{
"epoch": 0.6621542083578575,
"grad_norm": 4.006321789010636,
"learning_rate": 3.0925165045648564e-06,
"loss": 0.5144,
"step": 1125
},
{
"epoch": 0.6650971159505592,
"grad_norm": 3.1366794171510164,
"learning_rate": 3.0451360719522653e-06,
"loss": 0.5259,
"step": 1130
},
{
"epoch": 0.6680400235432608,
"grad_norm": 3.075050005395039,
"learning_rate": 2.997961957518892e-06,
"loss": 0.5126,
"step": 1135
},
{
"epoch": 0.6709829311359623,
"grad_norm": 2.614210901095361,
"learning_rate": 2.950999140065184e-06,
"loss": 0.526,
"step": 1140
},
{
"epoch": 0.673925838728664,
"grad_norm": 2.668099567229874,
"learning_rate": 2.904252576091105e-06,
"loss": 0.5207,
"step": 1145
},
{
"epoch": 0.6768687463213655,
"grad_norm": 2.822258046228147,
"learning_rate": 2.8577271992730223e-06,
"loss": 0.5179,
"step": 1150
},
{
"epoch": 0.6798116539140671,
"grad_norm": 2.7498849578418842,
"learning_rate": 2.8114279199430027e-06,
"loss": 0.5208,
"step": 1155
},
{
"epoch": 0.6827545615067687,
"grad_norm": 3.7794257172644086,
"learning_rate": 2.765359624570574e-06,
"loss": 0.5156,
"step": 1160
},
{
"epoch": 0.6856974690994703,
"grad_norm": 2.9809135197781527,
"learning_rate": 2.719527175246993e-06,
"loss": 0.5233,
"step": 1165
},
{
"epoch": 0.6886403766921718,
"grad_norm": 2.600094892048898,
"learning_rate": 2.6739354091721094e-06,
"loss": 0.5223,
"step": 1170
},
{
"epoch": 0.6915832842848735,
"grad_norm": 3.392788220792135,
"learning_rate": 2.6285891381438346e-06,
"loss": 0.5343,
"step": 1175
},
{
"epoch": 0.694526191877575,
"grad_norm": 2.620537806287387,
"learning_rate": 2.583493148050297e-06,
"loss": 0.5113,
"step": 1180
},
{
"epoch": 0.6974690994702767,
"grad_norm": 2.4542848921872817,
"learning_rate": 2.5386521983647468e-06,
"loss": 0.511,
"step": 1185
},
{
"epoch": 0.7004120070629782,
"grad_norm": 2.5882709829545556,
"learning_rate": 2.494071021643222e-06,
"loss": 0.5061,
"step": 1190
},
{
"epoch": 0.7033549146556798,
"grad_norm": 2.2977723019021306,
"learning_rate": 2.4497543230250843e-06,
"loss": 0.5041,
"step": 1195
},
{
"epoch": 0.7062978222483814,
"grad_norm": 2.706988899429715,
"learning_rate": 2.4057067797364247e-06,
"loss": 0.5034,
"step": 1200
},
{
"epoch": 0.709240729841083,
"grad_norm": 2.8986602480661716,
"learning_rate": 2.361933040596424e-06,
"loss": 0.5081,
"step": 1205
},
{
"epoch": 0.7121836374337845,
"grad_norm": 3.069283780524911,
"learning_rate": 2.31843772552672e-06,
"loss": 0.4812,
"step": 1210
},
{
"epoch": 0.7151265450264862,
"grad_norm": 2.942823198350691,
"learning_rate": 2.275225425063813e-06,
"loss": 0.4862,
"step": 1215
},
{
"epoch": 0.7180694526191878,
"grad_norm": 4.0835750894481,
"learning_rate": 2.2323006998745673e-06,
"loss": 0.5002,
"step": 1220
},
{
"epoch": 0.7210123602118893,
"grad_norm": 2.6814535983918764,
"learning_rate": 2.189668080274891e-06,
"loss": 0.4844,
"step": 1225
},
{
"epoch": 0.723955267804591,
"grad_norm": 2.476076418297489,
"learning_rate": 2.1473320657515913e-06,
"loss": 0.4882,
"step": 1230
},
{
"epoch": 0.7268981753972925,
"grad_norm": 2.9506625115965073,
"learning_rate": 2.1052971244874905e-06,
"loss": 0.496,
"step": 1235
},
{
"epoch": 0.7298410829899941,
"grad_norm": 2.4718878265701214,
"learning_rate": 2.063567692889864e-06,
"loss": 0.496,
"step": 1240
},
{
"epoch": 0.7327839905826957,
"grad_norm": 2.5394148181117724,
"learning_rate": 2.0221481751222027e-06,
"loss": 0.4918,
"step": 1245
},
{
"epoch": 0.7357268981753973,
"grad_norm": 2.4239900265691436,
"learning_rate": 1.981042942639404e-06,
"loss": 0.4825,
"step": 1250
},
{
"epoch": 0.7386698057680989,
"grad_norm": 2.582847306611647,
"learning_rate": 1.9402563337264012e-06,
"loss": 0.479,
"step": 1255
},
{
"epoch": 0.7416127133608005,
"grad_norm": 2.8447951582688344,
"learning_rate": 1.899792653040291e-06,
"loss": 0.4856,
"step": 1260
},
{
"epoch": 0.744555620953502,
"grad_norm": 2.7091883817883504,
"learning_rate": 1.8596561711560267e-06,
"loss": 0.4785,
"step": 1265
},
{
"epoch": 0.7474985285462037,
"grad_norm": 2.5437072619561443,
"learning_rate": 1.8198511241156902e-06,
"loss": 0.4846,
"step": 1270
},
{
"epoch": 0.7504414361389052,
"grad_norm": 3.156459227921938,
"learning_rate": 1.780381712981415e-06,
"loss": 0.4765,
"step": 1275
},
{
"epoch": 0.7533843437316068,
"grad_norm": 2.606788537147239,
"learning_rate": 1.7412521033920087e-06,
"loss": 0.4945,
"step": 1280
},
{
"epoch": 0.7563272513243084,
"grad_norm": 2.7549835249891057,
"learning_rate": 1.7024664251233052e-06,
"loss": 0.4795,
"step": 1285
},
{
"epoch": 0.75927015891701,
"grad_norm": 2.499144800895581,
"learning_rate": 1.6640287716522975e-06,
"loss": 0.4691,
"step": 1290
},
{
"epoch": 0.7622130665097115,
"grad_norm": 2.4799314569355677,
"learning_rate": 1.6259431997251191e-06,
"loss": 0.4972,
"step": 1295
},
{
"epoch": 0.7651559741024132,
"grad_norm": 2.5207515424299034,
"learning_rate": 1.5882137289288812e-06,
"loss": 0.4734,
"step": 1300
},
{
"epoch": 0.7680988816951148,
"grad_norm": 2.383436407772927,
"learning_rate": 1.5508443412674518e-06,
"loss": 0.4944,
"step": 1305
},
{
"epoch": 0.7710417892878164,
"grad_norm": 2.7793788321817656,
"learning_rate": 1.513838980741183e-06,
"loss": 0.4749,
"step": 1310
},
{
"epoch": 0.773984696880518,
"grad_norm": 2.301895560667882,
"learning_rate": 1.4772015529306587e-06,
"loss": 0.4619,
"step": 1315
},
{
"epoch": 0.7769276044732195,
"grad_norm": 2.446229173062197,
"learning_rate": 1.4409359245845005e-06,
"loss": 0.4671,
"step": 1320
},
{
"epoch": 0.7798705120659212,
"grad_norm": 2.998324656500335,
"learning_rate": 1.4050459232112652e-06,
"loss": 0.4708,
"step": 1325
},
{
"epoch": 0.7828134196586227,
"grad_norm": 2.287094207368545,
"learning_rate": 1.3695353366754798e-06,
"loss": 0.4666,
"step": 1330
},
{
"epoch": 0.7857563272513243,
"grad_norm": 2.652209930668726,
"learning_rate": 1.3344079127978788e-06,
"loss": 0.458,
"step": 1335
},
{
"epoch": 0.7886992348440259,
"grad_norm": 2.6373384490448872,
"learning_rate": 1.299667358959847e-06,
"loss": 0.4802,
"step": 1340
},
{
"epoch": 0.7916421424367275,
"grad_norm": 2.6241240764044034,
"learning_rate": 1.2653173417121367e-06,
"loss": 0.458,
"step": 1345
},
{
"epoch": 0.794585050029429,
"grad_norm": 2.596475069136379,
"learning_rate": 1.2313614863879074e-06,
"loss": 0.4589,
"step": 1350
},
{
"epoch": 0.7975279576221307,
"grad_norm": 2.6042310296871856,
"learning_rate": 1.1978033767200936e-06,
"loss": 0.4613,
"step": 1355
},
{
"epoch": 0.8004708652148322,
"grad_norm": 2.4907632128373614,
"learning_rate": 1.1646465544631802e-06,
"loss": 0.4672,
"step": 1360
},
{
"epoch": 0.8034137728075338,
"grad_norm": 2.8404270982027624,
"learning_rate": 1.131894519019403e-06,
"loss": 0.4647,
"step": 1365
},
{
"epoch": 0.8063566804002354,
"grad_norm": 2.511899627431363,
"learning_rate": 1.0995507270694121e-06,
"loss": 0.4568,
"step": 1370
},
{
"epoch": 0.809299587992937,
"grad_norm": 2.49013220750732,
"learning_rate": 1.0676185922074583e-06,
"loss": 0.4657,
"step": 1375
},
{
"epoch": 0.8122424955856387,
"grad_norm": 2.4156192880535747,
"learning_rate": 1.036101484581117e-06,
"loss": 0.4616,
"step": 1380
},
{
"epoch": 0.8151854031783402,
"grad_norm": 2.4647230171771364,
"learning_rate": 1.0050027305355926e-06,
"loss": 0.4651,
"step": 1385
},
{
"epoch": 0.8181283107710418,
"grad_norm": 2.90805484751002,
"learning_rate": 9.743256122626644e-07,
"loss": 0.4582,
"step": 1390
},
{
"epoch": 0.8210712183637434,
"grad_norm": 2.770355546856387,
"learning_rate": 9.440733674542751e-07,
"loss": 0.47,
"step": 1395
},
{
"epoch": 0.824014125956445,
"grad_norm": 2.894362207485248,
"learning_rate": 9.142491889608179e-07,
"loss": 0.4525,
"step": 1400
},
{
"epoch": 0.8269570335491465,
"grad_norm": 2.404589039415564,
"learning_rate": 8.848562244541675e-07,
"loss": 0.4682,
"step": 1405
},
{
"epoch": 0.8298999411418482,
"grad_norm": 2.492758443409389,
"learning_rate": 8.558975760954685e-07,
"loss": 0.4595,
"step": 1410
},
{
"epoch": 0.8328428487345497,
"grad_norm": 3.0988070143588202,
"learning_rate": 8.273763002077284e-07,
"loss": 0.4568,
"step": 1415
},
{
"epoch": 0.8357857563272513,
"grad_norm": 3.440047173962592,
"learning_rate": 7.992954069532543e-07,
"loss": 0.461,
"step": 1420
},
{
"epoch": 0.8387286639199529,
"grad_norm": 2.657598459534566,
"learning_rate": 7.716578600159535e-07,
"loss": 0.4503,
"step": 1425
},
{
"epoch": 0.8416715715126545,
"grad_norm": 2.572563487365978,
"learning_rate": 7.444665762885484e-07,
"loss": 0.4615,
"step": 1430
},
{
"epoch": 0.8446144791053561,
"grad_norm": 2.439187441026128,
"learning_rate": 7.177244255647209e-07,
"loss": 0.4446,
"step": 1435
},
{
"epoch": 0.8475573866980577,
"grad_norm": 2.233969483468557,
"learning_rate": 6.914342302362309e-07,
"loss": 0.4511,
"step": 1440
},
{
"epoch": 0.8505002942907592,
"grad_norm": 2.471799044846796,
"learning_rate": 6.65598764995043e-07,
"loss": 0.4497,
"step": 1445
},
{
"epoch": 0.8534432018834609,
"grad_norm": 2.755442497972038,
"learning_rate": 6.402207565404811e-07,
"loss": 0.4487,
"step": 1450
},
{
"epoch": 0.8563861094761624,
"grad_norm": 2.4539325525236326,
"learning_rate": 6.153028832914453e-07,
"loss": 0.4541,
"step": 1455
},
{
"epoch": 0.859329017068864,
"grad_norm": 2.2940758992178014,
"learning_rate": 5.908477751037361e-07,
"loss": 0.4425,
"step": 1460
},
{
"epoch": 0.8622719246615657,
"grad_norm": 2.5143891555656435,
"learning_rate": 5.668580129924949e-07,
"loss": 0.4475,
"step": 1465
},
{
"epoch": 0.8652148322542672,
"grad_norm": 2.643528212719313,
"learning_rate": 5.433361288597949e-07,
"loss": 0.4374,
"step": 1470
},
{
"epoch": 0.8681577398469688,
"grad_norm": 2.3103830994227774,
"learning_rate": 5.20284605227433e-07,
"loss": 0.4434,
"step": 1475
},
{
"epoch": 0.8711006474396704,
"grad_norm": 2.1890313025355668,
"learning_rate": 4.977058749749115e-07,
"loss": 0.4464,
"step": 1480
},
{
"epoch": 0.874043555032372,
"grad_norm": 2.311388455333076,
"learning_rate": 4.7560232108267714e-07,
"loss": 0.4501,
"step": 1485
},
{
"epoch": 0.8769864626250736,
"grad_norm": 2.2768695736500817,
"learning_rate": 4.5397627638061604e-07,
"loss": 0.4519,
"step": 1490
},
{
"epoch": 0.8799293702177752,
"grad_norm": 2.4901928210573305,
"learning_rate": 4.3283002330184233e-07,
"loss": 0.4426,
"step": 1495
},
{
"epoch": 0.8828722778104767,
"grad_norm": 2.2512157372682213,
"learning_rate": 4.1216579364181333e-07,
"loss": 0.4465,
"step": 1500
},
{
"epoch": 0.8858151854031784,
"grad_norm": 2.2396613255090947,
"learning_rate": 3.919857683227818e-07,
"loss": 0.447,
"step": 1505
},
{
"epoch": 0.8887580929958799,
"grad_norm": 2.4441734845394096,
"learning_rate": 3.72292077163614e-07,
"loss": 0.4455,
"step": 1510
},
{
"epoch": 0.8917010005885815,
"grad_norm": 2.3390391914462128,
"learning_rate": 3.5308679865501516e-07,
"loss": 0.4329,
"step": 1515
},
{
"epoch": 0.8946439081812831,
"grad_norm": 2.467951011914622,
"learning_rate": 3.343719597401596e-07,
"loss": 0.4373,
"step": 1520
},
{
"epoch": 0.8975868157739847,
"grad_norm": 2.506469216208831,
"learning_rate": 3.1614953560076213e-07,
"loss": 0.4413,
"step": 1525
},
{
"epoch": 0.9005297233666862,
"grad_norm": 2.3383168243956174,
"learning_rate": 2.984214494486215e-07,
"loss": 0.4548,
"step": 1530
},
{
"epoch": 0.9034726309593879,
"grad_norm": 2.630425109995007,
"learning_rate": 2.811895723226382e-07,
"loss": 0.4524,
"step": 1535
},
{
"epoch": 0.9064155385520895,
"grad_norm": 2.4286937188874775,
"learning_rate": 2.6445572289134634e-07,
"loss": 0.4376,
"step": 1540
},
{
"epoch": 0.909358446144791,
"grad_norm": 2.216575869780935,
"learning_rate": 2.482216672609677e-07,
"loss": 0.4351,
"step": 1545
},
{
"epoch": 0.9123013537374927,
"grad_norm": 2.3654422420764973,
"learning_rate": 2.3248911878901547e-07,
"loss": 0.4355,
"step": 1550
},
{
"epoch": 0.9152442613301942,
"grad_norm": 2.3549052063334495,
"learning_rate": 2.1725973790346522e-07,
"loss": 0.4346,
"step": 1555
},
{
"epoch": 0.9181871689228959,
"grad_norm": 2.3113103859994535,
"learning_rate": 2.0253513192751374e-07,
"loss": 0.4367,
"step": 1560
},
{
"epoch": 0.9211300765155974,
"grad_norm": 2.7576756319444553,
"learning_rate": 1.8831685490993457e-07,
"loss": 0.4381,
"step": 1565
},
{
"epoch": 0.924072984108299,
"grad_norm": 2.369997826369225,
"learning_rate": 1.7460640746107094e-07,
"loss": 0.4454,
"step": 1570
},
{
"epoch": 0.9270158917010006,
"grad_norm": 2.6393554236756,
"learning_rate": 1.614052365944524e-07,
"loss": 0.4481,
"step": 1575
},
{
"epoch": 0.9299587992937022,
"grad_norm": 2.4054005953759128,
"learning_rate": 1.4871473557407924e-07,
"loss": 0.4412,
"step": 1580
},
{
"epoch": 0.9329017068864037,
"grad_norm": 2.491831338754737,
"learning_rate": 1.3653624376737562e-07,
"loss": 0.4451,
"step": 1585
},
{
"epoch": 0.9358446144791054,
"grad_norm": 3.0815694783925305,
"learning_rate": 1.2487104650383165e-07,
"loss": 0.4283,
"step": 1590
},
{
"epoch": 0.9387875220718069,
"grad_norm": 2.2641401566584545,
"learning_rate": 1.1372037493934785e-07,
"loss": 0.4237,
"step": 1595
},
{
"epoch": 0.9417304296645085,
"grad_norm": 2.7762443739962057,
"learning_rate": 1.0308540592629756e-07,
"loss": 0.4404,
"step": 1600
},
{
"epoch": 0.9446733372572101,
"grad_norm": 2.6178876501865145,
"learning_rate": 9.296726188932293e-08,
"loss": 0.4306,
"step": 1605
},
{
"epoch": 0.9476162448499117,
"grad_norm": 2.6239109904141604,
"learning_rate": 8.336701070686904e-08,
"loss": 0.4461,
"step": 1610
},
{
"epoch": 0.9505591524426134,
"grad_norm": 2.2437383637665715,
"learning_rate": 7.428566559848416e-08,
"loss": 0.4345,
"step": 1615
},
{
"epoch": 0.9535020600353149,
"grad_norm": 2.1833908010974605,
"learning_rate": 6.572418501788025e-08,
"loss": 0.4359,
"step": 1620
},
{
"epoch": 0.9564449676280165,
"grad_norm": 2.3432337307633215,
"learning_rate": 5.768347255177609e-08,
"loss": 0.4169,
"step": 1625
},
{
"epoch": 0.9593878752207181,
"grad_norm": 2.809678686382748,
"learning_rate": 5.0164376824535235e-08,
"loss": 0.4353,
"step": 1630
},
{
"epoch": 0.9623307828134197,
"grad_norm": 2.36827229347058,
"learning_rate": 4.3167691408599334e-08,
"loss": 0.4284,
"step": 1635
},
{
"epoch": 0.9652736904061212,
"grad_norm": 2.29034432840639,
"learning_rate": 3.6694154740733456e-08,
"loss": 0.4443,
"step": 1640
},
{
"epoch": 0.9682165979988229,
"grad_norm": 2.313570657915266,
"learning_rate": 3.074445004409121e-08,
"loss": 0.4382,
"step": 1645
},
{
"epoch": 0.9711595055915244,
"grad_norm": 2.3743788836695776,
"learning_rate": 2.5319205256107425e-08,
"loss": 0.418,
"step": 1650
},
{
"epoch": 0.974102413184226,
"grad_norm": 2.3747057502664854,
"learning_rate": 2.0418992962224495e-08,
"loss": 0.4463,
"step": 1655
},
{
"epoch": 0.9770453207769276,
"grad_norm": 2.2646005029150618,
"learning_rate": 1.6044330335460712e-08,
"loss": 0.4517,
"step": 1660
},
{
"epoch": 0.9799882283696292,
"grad_norm": 2.3395141439258302,
"learning_rate": 1.2195679081827283e-08,
"loss": 0.447,
"step": 1665
},
{
"epoch": 0.9829311359623308,
"grad_norm": 2.7310828433799377,
"learning_rate": 8.873445391601176e-09,
"loss": 0.4445,
"step": 1670
},
{
"epoch": 0.9858740435550324,
"grad_norm": 2.3333374487486203,
"learning_rate": 6.077979896453312e-09,
"loss": 0.4354,
"step": 1675
},
{
"epoch": 0.9888169511477339,
"grad_norm": 2.3497335141047513,
"learning_rate": 3.809577632443717e-09,
"loss": 0.4321,
"step": 1680
},
{
"epoch": 0.9917598587404356,
"grad_norm": 2.3914591270933983,
"learning_rate": 2.0684780088825463e-09,
"loss": 0.4372,
"step": 1685
},
{
"epoch": 0.9947027663331371,
"grad_norm": 2.3017358226327653,
"learning_rate": 8.548647830630696e-10,
"loss": 0.4339,
"step": 1690
},
{
"epoch": 0.9976456739258387,
"grad_norm": 2.621035259676665,
"learning_rate": 1.6886604086774337e-10,
"loss": 0.4319,
"step": 1695
},
{
"epoch": 1.0,
"eval_runtime": 4.2193,
"eval_samples_per_second": 2.37,
"eval_steps_per_second": 0.711,
"step": 1699
},
{
"epoch": 1.0,
"step": 1699,
"total_flos": 116982426894336.0,
"train_loss": 0.6940072515278861,
"train_runtime": 20640.2326,
"train_samples_per_second": 1.317,
"train_steps_per_second": 0.082
}
],
"logging_steps": 5,
"max_steps": 1699,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 116982426894336.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}