Mistral-7B-v0.1-focus-sw / trainer_state.json
atsuki-yamaguchi's picture
Upload folder using huggingface_hub
087d2ea verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.8313847752663029,
"eval_steps": 500,
"global_step": 4000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 137.020263671875,
"learning_rate": 5e-06,
"loss": 29.3366,
"step": 5
},
{
"epoch": 0.0,
"grad_norm": 109.07556915283203,
"learning_rate": 1e-05,
"loss": 27.4529,
"step": 10
},
{
"epoch": 0.0,
"grad_norm": 170.4891815185547,
"learning_rate": 1.5e-05,
"loss": 23.5536,
"step": 15
},
{
"epoch": 0.0,
"grad_norm": 55.156646728515625,
"learning_rate": 2e-05,
"loss": 18.5847,
"step": 20
},
{
"epoch": 0.01,
"grad_norm": 27.853300094604492,
"learning_rate": 2.5e-05,
"loss": 13.4544,
"step": 25
},
{
"epoch": 0.01,
"grad_norm": 13.67708969116211,
"learning_rate": 3e-05,
"loss": 10.6938,
"step": 30
},
{
"epoch": 0.01,
"grad_norm": 11.794526100158691,
"learning_rate": 3.5e-05,
"loss": 9.4631,
"step": 35
},
{
"epoch": 0.01,
"grad_norm": 26.49574089050293,
"learning_rate": 4e-05,
"loss": 8.3764,
"step": 40
},
{
"epoch": 0.01,
"grad_norm": 6.533830165863037,
"learning_rate": 4.5e-05,
"loss": 8.0701,
"step": 45
},
{
"epoch": 0.01,
"grad_norm": 3.0623011589050293,
"learning_rate": 5e-05,
"loss": 7.8822,
"step": 50
},
{
"epoch": 0.01,
"grad_norm": 2.7227776050567627,
"learning_rate": 5.500000000000001e-05,
"loss": 7.8201,
"step": 55
},
{
"epoch": 0.01,
"grad_norm": 3.0419743061065674,
"learning_rate": 6e-05,
"loss": 7.6437,
"step": 60
},
{
"epoch": 0.01,
"grad_norm": 2.9724104404449463,
"learning_rate": 6.500000000000001e-05,
"loss": 7.4267,
"step": 65
},
{
"epoch": 0.01,
"grad_norm": 6.309749126434326,
"learning_rate": 7e-05,
"loss": 7.1598,
"step": 70
},
{
"epoch": 0.02,
"grad_norm": 3.2922019958496094,
"learning_rate": 7.500000000000001e-05,
"loss": 6.9908,
"step": 75
},
{
"epoch": 0.02,
"grad_norm": 7.097780704498291,
"learning_rate": 8e-05,
"loss": 6.7119,
"step": 80
},
{
"epoch": 0.02,
"grad_norm": 4.965648174285889,
"learning_rate": 8.5e-05,
"loss": 6.5169,
"step": 85
},
{
"epoch": 0.02,
"grad_norm": 5.251701831817627,
"learning_rate": 9e-05,
"loss": 6.364,
"step": 90
},
{
"epoch": 0.02,
"grad_norm": 4.818133354187012,
"learning_rate": 9.5e-05,
"loss": 6.1596,
"step": 95
},
{
"epoch": 0.02,
"grad_norm": 32.90806198120117,
"learning_rate": 0.0001,
"loss": 6.0288,
"step": 100
},
{
"epoch": 0.02,
"grad_norm": 4.648930072784424,
"learning_rate": 9.999972205865686e-05,
"loss": 5.9203,
"step": 105
},
{
"epoch": 0.02,
"grad_norm": 6.306034564971924,
"learning_rate": 9.999888823771751e-05,
"loss": 5.7512,
"step": 110
},
{
"epoch": 0.02,
"grad_norm": 5.2540459632873535,
"learning_rate": 9.999749854645204e-05,
"loss": 5.706,
"step": 115
},
{
"epoch": 0.02,
"grad_norm": 3.670748710632324,
"learning_rate": 9.99955530003106e-05,
"loss": 5.5649,
"step": 120
},
{
"epoch": 0.03,
"grad_norm": 3.2438762187957764,
"learning_rate": 9.99930516209231e-05,
"loss": 5.451,
"step": 125
},
{
"epoch": 0.03,
"grad_norm": 3.2759511470794678,
"learning_rate": 9.998999443609897e-05,
"loss": 5.3877,
"step": 130
},
{
"epoch": 0.03,
"grad_norm": 3.265584945678711,
"learning_rate": 9.998638147982696e-05,
"loss": 5.298,
"step": 135
},
{
"epoch": 0.03,
"grad_norm": 3.1032111644744873,
"learning_rate": 9.998221279227467e-05,
"loss": 5.2301,
"step": 140
},
{
"epoch": 0.03,
"grad_norm": 2.569101572036743,
"learning_rate": 9.997748841978812e-05,
"loss": 5.0846,
"step": 145
},
{
"epoch": 0.03,
"grad_norm": 4.325613498687744,
"learning_rate": 9.997220841489122e-05,
"loss": 5.0065,
"step": 150
},
{
"epoch": 0.03,
"grad_norm": 3.926607608795166,
"learning_rate": 9.996637283628528e-05,
"loss": 4.9818,
"step": 155
},
{
"epoch": 0.03,
"grad_norm": 2.6057651042938232,
"learning_rate": 9.995998174884821e-05,
"loss": 4.9515,
"step": 160
},
{
"epoch": 0.03,
"grad_norm": 2.8182785511016846,
"learning_rate": 9.995303522363394e-05,
"loss": 4.9315,
"step": 165
},
{
"epoch": 0.04,
"grad_norm": 2.7026009559631348,
"learning_rate": 9.99455333378715e-05,
"loss": 4.8364,
"step": 170
},
{
"epoch": 0.04,
"grad_norm": 2.4167022705078125,
"learning_rate": 9.993747617496428e-05,
"loss": 4.7815,
"step": 175
},
{
"epoch": 0.04,
"grad_norm": 2.470336675643921,
"learning_rate": 9.9928863824489e-05,
"loss": 4.7145,
"step": 180
},
{
"epoch": 0.04,
"grad_norm": 4.033043384552002,
"learning_rate": 9.99196963821948e-05,
"loss": 4.6346,
"step": 185
},
{
"epoch": 0.04,
"grad_norm": 3.112842559814453,
"learning_rate": 9.990997395000217e-05,
"loss": 4.6651,
"step": 190
},
{
"epoch": 0.04,
"grad_norm": 2.9468941688537598,
"learning_rate": 9.989969663600169e-05,
"loss": 4.6578,
"step": 195
},
{
"epoch": 0.04,
"grad_norm": 2.681847095489502,
"learning_rate": 9.9888864554453e-05,
"loss": 4.5354,
"step": 200
},
{
"epoch": 0.04,
"grad_norm": 2.2983462810516357,
"learning_rate": 9.987747782578342e-05,
"loss": 4.573,
"step": 205
},
{
"epoch": 0.04,
"grad_norm": 2.0980384349823,
"learning_rate": 9.986553657658668e-05,
"loss": 4.4353,
"step": 210
},
{
"epoch": 0.04,
"grad_norm": 2.3319320678710938,
"learning_rate": 9.985304093962145e-05,
"loss": 4.3826,
"step": 215
},
{
"epoch": 0.05,
"grad_norm": 2.326267719268799,
"learning_rate": 9.983999105380988e-05,
"loss": 4.4225,
"step": 220
},
{
"epoch": 0.05,
"grad_norm": 2.1649839878082275,
"learning_rate": 9.982638706423608e-05,
"loss": 4.3784,
"step": 225
},
{
"epoch": 0.05,
"grad_norm": 2.3909122943878174,
"learning_rate": 9.98122291221445e-05,
"loss": 4.3201,
"step": 230
},
{
"epoch": 0.05,
"grad_norm": 3.895136594772339,
"learning_rate": 9.979751738493826e-05,
"loss": 4.327,
"step": 235
},
{
"epoch": 0.05,
"grad_norm": 4.401449203491211,
"learning_rate": 9.978225201617732e-05,
"loss": 4.3528,
"step": 240
},
{
"epoch": 0.05,
"grad_norm": 2.517158031463623,
"learning_rate": 9.976643318557678e-05,
"loss": 4.3535,
"step": 245
},
{
"epoch": 0.05,
"grad_norm": 2.185535430908203,
"learning_rate": 9.975006106900495e-05,
"loss": 4.172,
"step": 250
},
{
"epoch": 0.05,
"grad_norm": 2.6196024417877197,
"learning_rate": 9.973313584848132e-05,
"loss": 4.285,
"step": 255
},
{
"epoch": 0.05,
"grad_norm": 2.339275360107422,
"learning_rate": 9.971565771217464e-05,
"loss": 4.1991,
"step": 260
},
{
"epoch": 0.06,
"grad_norm": 2.7671592235565186,
"learning_rate": 9.969762685440076e-05,
"loss": 4.2154,
"step": 265
},
{
"epoch": 0.06,
"grad_norm": 2.326404094696045,
"learning_rate": 9.967904347562054e-05,
"loss": 4.1889,
"step": 270
},
{
"epoch": 0.06,
"grad_norm": 2.097139358520508,
"learning_rate": 9.965990778243755e-05,
"loss": 4.0934,
"step": 275
},
{
"epoch": 0.06,
"grad_norm": 3.074829578399658,
"learning_rate": 9.964021998759577e-05,
"loss": 4.1701,
"step": 280
},
{
"epoch": 0.06,
"grad_norm": 2.592827320098877,
"learning_rate": 9.961998030997733e-05,
"loss": 4.2065,
"step": 285
},
{
"epoch": 0.06,
"grad_norm": 2.123453140258789,
"learning_rate": 9.95991889745999e-05,
"loss": 4.0641,
"step": 290
},
{
"epoch": 0.06,
"grad_norm": 1.8817672729492188,
"learning_rate": 9.957784621261441e-05,
"loss": 4.0941,
"step": 295
},
{
"epoch": 0.06,
"grad_norm": 2.40132212638855,
"learning_rate": 9.955595226130226e-05,
"loss": 4.0269,
"step": 300
},
{
"epoch": 0.06,
"grad_norm": 2.1056606769561768,
"learning_rate": 9.953350736407282e-05,
"loss": 4.0569,
"step": 305
},
{
"epoch": 0.06,
"grad_norm": 2.0326855182647705,
"learning_rate": 9.951051177046069e-05,
"loss": 3.9838,
"step": 310
},
{
"epoch": 0.07,
"grad_norm": 2.248284339904785,
"learning_rate": 9.948696573612292e-05,
"loss": 4.0631,
"step": 315
},
{
"epoch": 0.07,
"grad_norm": 1.9347286224365234,
"learning_rate": 9.946286952283618e-05,
"loss": 3.9923,
"step": 320
},
{
"epoch": 0.07,
"grad_norm": 2.2586050033569336,
"learning_rate": 9.943822339849381e-05,
"loss": 3.9914,
"step": 325
},
{
"epoch": 0.07,
"grad_norm": 1.936707615852356,
"learning_rate": 9.941302763710288e-05,
"loss": 3.9033,
"step": 330
},
{
"epoch": 0.07,
"grad_norm": 2.02547287940979,
"learning_rate": 9.938728251878116e-05,
"loss": 3.9657,
"step": 335
},
{
"epoch": 0.07,
"grad_norm": 1.957825779914856,
"learning_rate": 9.936098832975393e-05,
"loss": 3.9216,
"step": 340
},
{
"epoch": 0.07,
"grad_norm": 1.727043867111206,
"learning_rate": 9.933414536235091e-05,
"loss": 3.8126,
"step": 345
},
{
"epoch": 0.07,
"grad_norm": 1.8693815469741821,
"learning_rate": 9.93067539150029e-05,
"loss": 3.8927,
"step": 350
},
{
"epoch": 0.07,
"grad_norm": 1.79536771774292,
"learning_rate": 9.927881429223853e-05,
"loss": 3.828,
"step": 355
},
{
"epoch": 0.07,
"grad_norm": 1.853402853012085,
"learning_rate": 9.925032680468085e-05,
"loss": 3.9035,
"step": 360
},
{
"epoch": 0.08,
"grad_norm": 1.7830322980880737,
"learning_rate": 9.922129176904388e-05,
"loss": 3.7961,
"step": 365
},
{
"epoch": 0.08,
"grad_norm": 1.9644182920455933,
"learning_rate": 9.919170950812911e-05,
"loss": 3.7411,
"step": 370
},
{
"epoch": 0.08,
"grad_norm": 1.8130580186843872,
"learning_rate": 9.916158035082184e-05,
"loss": 3.6928,
"step": 375
},
{
"epoch": 0.08,
"grad_norm": 1.8783149719238281,
"learning_rate": 9.913090463208763e-05,
"loss": 3.7242,
"step": 380
},
{
"epoch": 0.08,
"grad_norm": 1.7080426216125488,
"learning_rate": 9.90996826929685e-05,
"loss": 3.7248,
"step": 385
},
{
"epoch": 0.08,
"grad_norm": 1.8799563646316528,
"learning_rate": 9.906791488057916e-05,
"loss": 3.6945,
"step": 390
},
{
"epoch": 0.08,
"grad_norm": 1.9748135805130005,
"learning_rate": 9.903560154810313e-05,
"loss": 3.7439,
"step": 395
},
{
"epoch": 0.08,
"grad_norm": 1.94729483127594,
"learning_rate": 9.900274305478887e-05,
"loss": 3.7472,
"step": 400
},
{
"epoch": 0.08,
"grad_norm": 1.991847276687622,
"learning_rate": 9.896933976594572e-05,
"loss": 3.7009,
"step": 405
},
{
"epoch": 0.09,
"grad_norm": 1.906658411026001,
"learning_rate": 9.893539205293989e-05,
"loss": 3.6658,
"step": 410
},
{
"epoch": 0.09,
"grad_norm": 1.7650723457336426,
"learning_rate": 9.890090029319028e-05,
"loss": 3.6477,
"step": 415
},
{
"epoch": 0.09,
"grad_norm": 1.9193689823150635,
"learning_rate": 9.886586487016433e-05,
"loss": 3.6337,
"step": 420
},
{
"epoch": 0.09,
"grad_norm": 1.7722522020339966,
"learning_rate": 9.883028617337378e-05,
"loss": 3.7095,
"step": 425
},
{
"epoch": 0.09,
"grad_norm": 2.179424285888672,
"learning_rate": 9.879416459837022e-05,
"loss": 3.6381,
"step": 430
},
{
"epoch": 0.09,
"grad_norm": 1.8186856508255005,
"learning_rate": 9.875750054674082e-05,
"loss": 3.6123,
"step": 435
},
{
"epoch": 0.09,
"grad_norm": 1.7127214670181274,
"learning_rate": 9.872029442610382e-05,
"loss": 3.6322,
"step": 440
},
{
"epoch": 0.09,
"grad_norm": 1.611188530921936,
"learning_rate": 9.8682546650104e-05,
"loss": 3.551,
"step": 445
},
{
"epoch": 0.09,
"grad_norm": 1.6156830787658691,
"learning_rate": 9.864425763840802e-05,
"loss": 3.5886,
"step": 450
},
{
"epoch": 0.09,
"grad_norm": 2.6605937480926514,
"learning_rate": 9.860542781669988e-05,
"loss": 3.5478,
"step": 455
},
{
"epoch": 0.1,
"grad_norm": 1.6025112867355347,
"learning_rate": 9.85660576166761e-05,
"loss": 3.515,
"step": 460
},
{
"epoch": 0.1,
"grad_norm": 1.6478216648101807,
"learning_rate": 9.852614747604093e-05,
"loss": 3.4571,
"step": 465
},
{
"epoch": 0.1,
"grad_norm": 1.4119625091552734,
"learning_rate": 9.848569783850145e-05,
"loss": 3.4977,
"step": 470
},
{
"epoch": 0.1,
"grad_norm": 1.4904718399047852,
"learning_rate": 9.844470915376278e-05,
"loss": 3.5384,
"step": 475
},
{
"epoch": 0.1,
"grad_norm": 1.438099980354309,
"learning_rate": 9.840318187752292e-05,
"loss": 3.5476,
"step": 480
},
{
"epoch": 0.1,
"grad_norm": 1.456661343574524,
"learning_rate": 9.836111647146771e-05,
"loss": 3.3344,
"step": 485
},
{
"epoch": 0.1,
"grad_norm": 1.5118834972381592,
"learning_rate": 9.831851340326577e-05,
"loss": 3.4097,
"step": 490
},
{
"epoch": 0.1,
"grad_norm": 1.5433509349822998,
"learning_rate": 9.82753731465633e-05,
"loss": 3.4497,
"step": 495
},
{
"epoch": 0.1,
"grad_norm": 1.6322232484817505,
"learning_rate": 9.823169618097871e-05,
"loss": 3.4612,
"step": 500
},
{
"epoch": 0.1,
"grad_norm": 1.4391494989395142,
"learning_rate": 9.81874829920974e-05,
"loss": 3.3981,
"step": 505
},
{
"epoch": 0.11,
"grad_norm": 1.780336618423462,
"learning_rate": 9.814273407146623e-05,
"loss": 3.4524,
"step": 510
},
{
"epoch": 0.11,
"grad_norm": 1.6080727577209473,
"learning_rate": 9.809744991658829e-05,
"loss": 3.4337,
"step": 515
},
{
"epoch": 0.11,
"grad_norm": 1.6014811992645264,
"learning_rate": 9.805163103091708e-05,
"loss": 3.4079,
"step": 520
},
{
"epoch": 0.11,
"grad_norm": 1.3736337423324585,
"learning_rate": 9.800527792385112e-05,
"loss": 3.3513,
"step": 525
},
{
"epoch": 0.11,
"grad_norm": 1.451069712638855,
"learning_rate": 9.79583911107282e-05,
"loss": 3.365,
"step": 530
},
{
"epoch": 0.11,
"grad_norm": 1.4618583917617798,
"learning_rate": 9.791097111281968e-05,
"loss": 3.3863,
"step": 535
},
{
"epoch": 0.11,
"grad_norm": 1.4862762689590454,
"learning_rate": 9.786301845732467e-05,
"loss": 3.295,
"step": 540
},
{
"epoch": 0.11,
"grad_norm": 1.5561633110046387,
"learning_rate": 9.781453367736418e-05,
"loss": 3.2993,
"step": 545
},
{
"epoch": 0.11,
"grad_norm": 1.4097754955291748,
"learning_rate": 9.776551731197524e-05,
"loss": 3.2775,
"step": 550
},
{
"epoch": 0.12,
"grad_norm": 1.3923735618591309,
"learning_rate": 9.771596990610478e-05,
"loss": 3.2167,
"step": 555
},
{
"epoch": 0.12,
"grad_norm": 1.4300577640533447,
"learning_rate": 9.766589201060372e-05,
"loss": 3.3663,
"step": 560
},
{
"epoch": 0.12,
"grad_norm": 1.46330988407135,
"learning_rate": 9.761528418222077e-05,
"loss": 3.2499,
"step": 565
},
{
"epoch": 0.12,
"grad_norm": 1.569105625152588,
"learning_rate": 9.756414698359624e-05,
"loss": 3.2469,
"step": 570
},
{
"epoch": 0.12,
"grad_norm": 1.5255167484283447,
"learning_rate": 9.75124809832558e-05,
"loss": 3.2909,
"step": 575
},
{
"epoch": 0.12,
"grad_norm": 1.4769233465194702,
"learning_rate": 9.746028675560413e-05,
"loss": 3.2518,
"step": 580
},
{
"epoch": 0.12,
"grad_norm": 1.4233957529067993,
"learning_rate": 9.740756488091861e-05,
"loss": 3.2779,
"step": 585
},
{
"epoch": 0.12,
"grad_norm": 1.4301520586013794,
"learning_rate": 9.735431594534277e-05,
"loss": 3.2006,
"step": 590
},
{
"epoch": 0.12,
"grad_norm": 1.4157580137252808,
"learning_rate": 9.730054054087983e-05,
"loss": 3.2719,
"step": 595
},
{
"epoch": 0.12,
"grad_norm": 1.5442814826965332,
"learning_rate": 9.724623926538612e-05,
"loss": 3.2103,
"step": 600
},
{
"epoch": 0.13,
"grad_norm": 1.3368921279907227,
"learning_rate": 9.719141272256443e-05,
"loss": 3.1553,
"step": 605
},
{
"epoch": 0.13,
"grad_norm": 1.2515050172805786,
"learning_rate": 9.713606152195726e-05,
"loss": 3.234,
"step": 610
},
{
"epoch": 0.13,
"grad_norm": 1.4154413938522339,
"learning_rate": 9.708018627894011e-05,
"loss": 3.2055,
"step": 615
},
{
"epoch": 0.13,
"grad_norm": 1.4324495792388916,
"learning_rate": 9.702378761471456e-05,
"loss": 3.2533,
"step": 620
},
{
"epoch": 0.13,
"grad_norm": 1.2748337984085083,
"learning_rate": 9.696686615630146e-05,
"loss": 3.1449,
"step": 625
},
{
"epoch": 0.13,
"grad_norm": 1.4128867387771606,
"learning_rate": 9.690942253653385e-05,
"loss": 3.1894,
"step": 630
},
{
"epoch": 0.13,
"grad_norm": 1.2542297840118408,
"learning_rate": 9.685145739405002e-05,
"loss": 3.1965,
"step": 635
},
{
"epoch": 0.13,
"grad_norm": 1.266688346862793,
"learning_rate": 9.679297137328634e-05,
"loss": 3.1809,
"step": 640
},
{
"epoch": 0.13,
"grad_norm": 1.442312479019165,
"learning_rate": 9.673396512447013e-05,
"loss": 3.153,
"step": 645
},
{
"epoch": 0.14,
"grad_norm": 1.3614144325256348,
"learning_rate": 9.667443930361247e-05,
"loss": 3.1016,
"step": 650
},
{
"epoch": 0.14,
"grad_norm": 1.343624472618103,
"learning_rate": 9.661439457250076e-05,
"loss": 3.1734,
"step": 655
},
{
"epoch": 0.14,
"grad_norm": 1.695099115371704,
"learning_rate": 9.655383159869158e-05,
"loss": 3.1199,
"step": 660
},
{
"epoch": 0.14,
"grad_norm": 1.5576951503753662,
"learning_rate": 9.649275105550309e-05,
"loss": 3.0744,
"step": 665
},
{
"epoch": 0.14,
"grad_norm": 1.4647527933120728,
"learning_rate": 9.643115362200762e-05,
"loss": 3.0753,
"step": 670
},
{
"epoch": 0.14,
"grad_norm": 1.3130619525909424,
"learning_rate": 9.636903998302409e-05,
"loss": 3.1342,
"step": 675
},
{
"epoch": 0.14,
"grad_norm": 1.2923471927642822,
"learning_rate": 9.630641082911045e-05,
"loss": 3.0888,
"step": 680
},
{
"epoch": 0.14,
"grad_norm": 1.3009761571884155,
"learning_rate": 9.624326685655593e-05,
"loss": 3.1669,
"step": 685
},
{
"epoch": 0.14,
"grad_norm": 1.2862359285354614,
"learning_rate": 9.617960876737337e-05,
"loss": 3.0965,
"step": 690
},
{
"epoch": 0.14,
"grad_norm": 1.228393316268921,
"learning_rate": 9.611543726929134e-05,
"loss": 3.0653,
"step": 695
},
{
"epoch": 0.15,
"grad_norm": 1.1705549955368042,
"learning_rate": 9.605075307574635e-05,
"loss": 3.1241,
"step": 700
},
{
"epoch": 0.15,
"grad_norm": 1.1654901504516602,
"learning_rate": 9.598555690587487e-05,
"loss": 3.0948,
"step": 705
},
{
"epoch": 0.15,
"grad_norm": 1.1867812871932983,
"learning_rate": 9.591984948450532e-05,
"loss": 3.1185,
"step": 710
},
{
"epoch": 0.15,
"grad_norm": 1.3131000995635986,
"learning_rate": 9.585363154215008e-05,
"loss": 3.0882,
"step": 715
},
{
"epoch": 0.15,
"grad_norm": 1.2663651704788208,
"learning_rate": 9.578690381499728e-05,
"loss": 3.0816,
"step": 720
},
{
"epoch": 0.15,
"grad_norm": 1.226505994796753,
"learning_rate": 9.571966704490271e-05,
"loss": 3.0173,
"step": 725
},
{
"epoch": 0.15,
"grad_norm": 1.3454265594482422,
"learning_rate": 9.565192197938148e-05,
"loss": 3.0315,
"step": 730
},
{
"epoch": 0.15,
"grad_norm": 1.187574028968811,
"learning_rate": 9.558366937159977e-05,
"loss": 3.0901,
"step": 735
},
{
"epoch": 0.15,
"grad_norm": 1.2839643955230713,
"learning_rate": 9.551490998036646e-05,
"loss": 3.0159,
"step": 740
},
{
"epoch": 0.15,
"grad_norm": 1.2916148900985718,
"learning_rate": 9.544564457012463e-05,
"loss": 3.1099,
"step": 745
},
{
"epoch": 0.16,
"grad_norm": 1.3566361665725708,
"learning_rate": 9.537587391094314e-05,
"loss": 3.0296,
"step": 750
},
{
"epoch": 0.16,
"grad_norm": 1.237157940864563,
"learning_rate": 9.5305598778508e-05,
"loss": 3.0956,
"step": 755
},
{
"epoch": 0.16,
"grad_norm": 1.163589596748352,
"learning_rate": 9.52348199541138e-05,
"loss": 3.049,
"step": 760
},
{
"epoch": 0.16,
"grad_norm": 1.243045687675476,
"learning_rate": 9.516353822465504e-05,
"loss": 3.1106,
"step": 765
},
{
"epoch": 0.16,
"grad_norm": 1.2744767665863037,
"learning_rate": 9.509175438261726e-05,
"loss": 3.0186,
"step": 770
},
{
"epoch": 0.16,
"grad_norm": 1.3286916017532349,
"learning_rate": 9.501946922606838e-05,
"loss": 3.0464,
"step": 775
},
{
"epoch": 0.16,
"grad_norm": 1.2067335844039917,
"learning_rate": 9.494668355864973e-05,
"loss": 2.9649,
"step": 780
},
{
"epoch": 0.16,
"grad_norm": 1.151553988456726,
"learning_rate": 9.487339818956716e-05,
"loss": 2.9769,
"step": 785
},
{
"epoch": 0.16,
"grad_norm": 1.1733759641647339,
"learning_rate": 9.479961393358203e-05,
"loss": 3.0688,
"step": 790
},
{
"epoch": 0.17,
"grad_norm": 1.123901128768921,
"learning_rate": 9.472533161100215e-05,
"loss": 2.9925,
"step": 795
},
{
"epoch": 0.17,
"grad_norm": 1.169124722480774,
"learning_rate": 9.465055204767265e-05,
"loss": 2.9526,
"step": 800
},
{
"epoch": 0.17,
"grad_norm": 1.7418535947799683,
"learning_rate": 9.457527607496685e-05,
"loss": 2.9742,
"step": 805
},
{
"epoch": 0.17,
"grad_norm": 1.252622127532959,
"learning_rate": 9.44995045297769e-05,
"loss": 3.0665,
"step": 810
},
{
"epoch": 0.17,
"grad_norm": 1.1671171188354492,
"learning_rate": 9.442323825450464e-05,
"loss": 2.9878,
"step": 815
},
{
"epoch": 0.17,
"grad_norm": 1.1918081045150757,
"learning_rate": 9.43464780970521e-05,
"loss": 2.9364,
"step": 820
},
{
"epoch": 0.17,
"grad_norm": 1.134225845336914,
"learning_rate": 9.426922491081212e-05,
"loss": 2.9205,
"step": 825
},
{
"epoch": 0.17,
"grad_norm": 1.1393715143203735,
"learning_rate": 9.419147955465888e-05,
"loss": 3.0289,
"step": 830
},
{
"epoch": 0.17,
"grad_norm": 1.1497019529342651,
"learning_rate": 9.411324289293832e-05,
"loss": 2.9906,
"step": 835
},
{
"epoch": 0.17,
"grad_norm": 1.0984611511230469,
"learning_rate": 9.403451579545859e-05,
"loss": 2.964,
"step": 840
},
{
"epoch": 0.18,
"grad_norm": 1.2375845909118652,
"learning_rate": 9.395529913748025e-05,
"loss": 2.9787,
"step": 845
},
{
"epoch": 0.18,
"grad_norm": 1.1662236452102661,
"learning_rate": 9.387559379970672e-05,
"loss": 2.9684,
"step": 850
},
{
"epoch": 0.18,
"grad_norm": 1.1607048511505127,
"learning_rate": 9.379540066827431e-05,
"loss": 2.9925,
"step": 855
},
{
"epoch": 0.18,
"grad_norm": 1.17828369140625,
"learning_rate": 9.371472063474248e-05,
"loss": 2.9272,
"step": 860
},
{
"epoch": 0.18,
"grad_norm": 1.0758649110794067,
"learning_rate": 9.363355459608394e-05,
"loss": 2.9528,
"step": 865
},
{
"epoch": 0.18,
"grad_norm": 1.188213586807251,
"learning_rate": 9.355190345467457e-05,
"loss": 2.9892,
"step": 870
},
{
"epoch": 0.18,
"grad_norm": 1.0760455131530762,
"learning_rate": 9.346976811828352e-05,
"loss": 2.9936,
"step": 875
},
{
"epoch": 0.18,
"grad_norm": 1.0703479051589966,
"learning_rate": 9.338714950006297e-05,
"loss": 2.8893,
"step": 880
},
{
"epoch": 0.18,
"grad_norm": 1.0391613245010376,
"learning_rate": 9.330404851853817e-05,
"loss": 2.9639,
"step": 885
},
{
"epoch": 0.18,
"grad_norm": 1.0646852254867554,
"learning_rate": 9.3220466097597e-05,
"loss": 2.978,
"step": 890
},
{
"epoch": 0.19,
"grad_norm": 1.18621027469635,
"learning_rate": 9.313640316647991e-05,
"loss": 2.9308,
"step": 895
},
{
"epoch": 0.19,
"grad_norm": 1.1788042783737183,
"learning_rate": 9.305186065976945e-05,
"loss": 2.9228,
"step": 900
},
{
"epoch": 0.19,
"grad_norm": 1.0776159763336182,
"learning_rate": 9.296683951737993e-05,
"loss": 2.9048,
"step": 905
},
{
"epoch": 0.19,
"grad_norm": 1.0837948322296143,
"learning_rate": 9.288134068454697e-05,
"loss": 2.8496,
"step": 910
},
{
"epoch": 0.19,
"grad_norm": 1.1005736589431763,
"learning_rate": 9.2795365111817e-05,
"loss": 2.9368,
"step": 915
},
{
"epoch": 0.19,
"grad_norm": 1.1075985431671143,
"learning_rate": 9.270891375503665e-05,
"loss": 2.963,
"step": 920
},
{
"epoch": 0.19,
"grad_norm": 1.063977837562561,
"learning_rate": 9.262198757534218e-05,
"loss": 2.8108,
"step": 925
},
{
"epoch": 0.19,
"grad_norm": 1.288533329963684,
"learning_rate": 9.253458753914874e-05,
"loss": 2.8746,
"step": 930
},
{
"epoch": 0.19,
"grad_norm": 1.0941237211227417,
"learning_rate": 9.244671461813969e-05,
"loss": 2.9319,
"step": 935
},
{
"epoch": 0.2,
"grad_norm": 1.0808980464935303,
"learning_rate": 9.235836978925572e-05,
"loss": 2.9179,
"step": 940
},
{
"epoch": 0.2,
"grad_norm": 1.2530587911605835,
"learning_rate": 9.226955403468406e-05,
"loss": 2.9531,
"step": 945
},
{
"epoch": 0.2,
"grad_norm": 1.2044453620910645,
"learning_rate": 9.21802683418475e-05,
"loss": 2.8782,
"step": 950
},
{
"epoch": 0.2,
"grad_norm": 1.155251383781433,
"learning_rate": 9.209051370339347e-05,
"loss": 2.9454,
"step": 955
},
{
"epoch": 0.2,
"grad_norm": 1.301200270652771,
"learning_rate": 9.200029111718295e-05,
"loss": 2.817,
"step": 960
},
{
"epoch": 0.2,
"grad_norm": 1.1260250806808472,
"learning_rate": 9.190960158627941e-05,
"loss": 2.9268,
"step": 965
},
{
"epoch": 0.2,
"grad_norm": 1.0131049156188965,
"learning_rate": 9.181844611893766e-05,
"loss": 2.9798,
"step": 970
},
{
"epoch": 0.2,
"grad_norm": 1.2464625835418701,
"learning_rate": 9.172682572859261e-05,
"loss": 2.8194,
"step": 975
},
{
"epoch": 0.2,
"grad_norm": 1.1114532947540283,
"learning_rate": 9.163474143384806e-05,
"loss": 2.9421,
"step": 980
},
{
"epoch": 0.2,
"grad_norm": 1.0470279455184937,
"learning_rate": 9.154219425846528e-05,
"loss": 2.8578,
"step": 985
},
{
"epoch": 0.21,
"grad_norm": 3.8998446464538574,
"learning_rate": 9.144918523135175e-05,
"loss": 2.9439,
"step": 990
},
{
"epoch": 0.21,
"grad_norm": 1.0205687284469604,
"learning_rate": 9.13557153865496e-05,
"loss": 2.8819,
"step": 995
},
{
"epoch": 0.21,
"grad_norm": 2.913062810897827,
"learning_rate": 9.12617857632242e-05,
"loss": 2.8762,
"step": 1000
},
{
"epoch": 0.21,
"grad_norm": 1.112969160079956,
"learning_rate": 9.116739740565259e-05,
"loss": 2.9525,
"step": 1005
},
{
"epoch": 0.21,
"grad_norm": 1.0801798105239868,
"learning_rate": 9.107255136321184e-05,
"loss": 2.8277,
"step": 1010
},
{
"epoch": 0.21,
"grad_norm": 1.1021159887313843,
"learning_rate": 9.09772486903674e-05,
"loss": 2.9052,
"step": 1015
},
{
"epoch": 0.21,
"grad_norm": 1.0213983058929443,
"learning_rate": 9.08814904466614e-05,
"loss": 2.8455,
"step": 1020
},
{
"epoch": 0.21,
"grad_norm": 1.131356954574585,
"learning_rate": 9.078527769670085e-05,
"loss": 2.8025,
"step": 1025
},
{
"epoch": 0.21,
"grad_norm": 1.076568365097046,
"learning_rate": 9.068861151014575e-05,
"loss": 2.8286,
"step": 1030
},
{
"epoch": 0.22,
"grad_norm": 1.095739722251892,
"learning_rate": 9.05914929616973e-05,
"loss": 2.8479,
"step": 1035
},
{
"epoch": 0.22,
"grad_norm": 1.0352883338928223,
"learning_rate": 9.04939231310859e-05,
"loss": 2.8552,
"step": 1040
},
{
"epoch": 0.22,
"grad_norm": 1.1047638654708862,
"learning_rate": 9.039590310305914e-05,
"loss": 2.8125,
"step": 1045
},
{
"epoch": 0.22,
"grad_norm": 1.0611072778701782,
"learning_rate": 9.029743396736974e-05,
"loss": 2.8543,
"step": 1050
},
{
"epoch": 0.22,
"grad_norm": 1.0438469648361206,
"learning_rate": 9.019851681876348e-05,
"loss": 2.8831,
"step": 1055
},
{
"epoch": 0.22,
"grad_norm": 1.028063416481018,
"learning_rate": 9.009915275696693e-05,
"loss": 2.867,
"step": 1060
},
{
"epoch": 0.22,
"grad_norm": 1.1007819175720215,
"learning_rate": 8.999934288667534e-05,
"loss": 2.8624,
"step": 1065
},
{
"epoch": 0.22,
"grad_norm": 1.106092095375061,
"learning_rate": 8.989908831754028e-05,
"loss": 2.9078,
"step": 1070
},
{
"epoch": 0.22,
"grad_norm": 1.1761088371276855,
"learning_rate": 8.979839016415735e-05,
"loss": 2.8175,
"step": 1075
},
{
"epoch": 0.22,
"grad_norm": 1.0588090419769287,
"learning_rate": 8.969724954605373e-05,
"loss": 2.8673,
"step": 1080
},
{
"epoch": 0.23,
"grad_norm": 1.0640250444412231,
"learning_rate": 8.959566758767581e-05,
"loss": 2.8766,
"step": 1085
},
{
"epoch": 0.23,
"grad_norm": 1.0558903217315674,
"learning_rate": 8.949364541837661e-05,
"loss": 2.8431,
"step": 1090
},
{
"epoch": 0.23,
"grad_norm": 1.0601201057434082,
"learning_rate": 8.939118417240329e-05,
"loss": 2.848,
"step": 1095
},
{
"epoch": 0.23,
"grad_norm": 1.0075984001159668,
"learning_rate": 8.92882849888845e-05,
"loss": 2.8221,
"step": 1100
},
{
"epoch": 0.23,
"grad_norm": 1.0282323360443115,
"learning_rate": 8.918494901181773e-05,
"loss": 2.8628,
"step": 1105
},
{
"epoch": 0.23,
"grad_norm": 1.0520479679107666,
"learning_rate": 8.908117739005659e-05,
"loss": 2.8436,
"step": 1110
},
{
"epoch": 0.23,
"grad_norm": 1.0297996997833252,
"learning_rate": 8.897697127729805e-05,
"loss": 2.7828,
"step": 1115
},
{
"epoch": 0.23,
"grad_norm": 1.0245318412780762,
"learning_rate": 8.887233183206957e-05,
"loss": 2.8392,
"step": 1120
},
{
"epoch": 0.23,
"grad_norm": 0.9845367670059204,
"learning_rate": 8.876726021771627e-05,
"loss": 2.8087,
"step": 1125
},
{
"epoch": 0.23,
"grad_norm": 0.9897079467773438,
"learning_rate": 8.866175760238798e-05,
"loss": 2.8482,
"step": 1130
},
{
"epoch": 0.24,
"grad_norm": 1.0339479446411133,
"learning_rate": 8.855582515902625e-05,
"loss": 2.7912,
"step": 1135
},
{
"epoch": 0.24,
"grad_norm": 1.0386319160461426,
"learning_rate": 8.844946406535131e-05,
"loss": 2.8598,
"step": 1140
},
{
"epoch": 0.24,
"grad_norm": 1.035483956336975,
"learning_rate": 8.834267550384893e-05,
"loss": 2.7932,
"step": 1145
},
{
"epoch": 0.24,
"grad_norm": 1.0350147485733032,
"learning_rate": 8.823546066175741e-05,
"loss": 2.8413,
"step": 1150
},
{
"epoch": 0.24,
"grad_norm": 1.0898412466049194,
"learning_rate": 8.81278207310542e-05,
"loss": 2.7879,
"step": 1155
},
{
"epoch": 0.24,
"grad_norm": 1.0223238468170166,
"learning_rate": 8.801975690844278e-05,
"loss": 2.7522,
"step": 1160
},
{
"epoch": 0.24,
"grad_norm": 1.1877025365829468,
"learning_rate": 8.791127039533934e-05,
"loss": 2.6965,
"step": 1165
},
{
"epoch": 0.24,
"grad_norm": 0.9616990685462952,
"learning_rate": 8.780236239785935e-05,
"loss": 2.7613,
"step": 1170
},
{
"epoch": 0.24,
"grad_norm": 1.0233880281448364,
"learning_rate": 8.76930341268042e-05,
"loss": 2.7835,
"step": 1175
},
{
"epoch": 0.25,
"grad_norm": 1.0146315097808838,
"learning_rate": 8.758328679764776e-05,
"loss": 2.8325,
"step": 1180
},
{
"epoch": 0.25,
"grad_norm": 1.000049352645874,
"learning_rate": 8.747312163052284e-05,
"loss": 2.871,
"step": 1185
},
{
"epoch": 0.25,
"grad_norm": 1.0159934759140015,
"learning_rate": 8.736253985020761e-05,
"loss": 2.743,
"step": 1190
},
{
"epoch": 0.25,
"grad_norm": 1.0828293561935425,
"learning_rate": 8.725154268611203e-05,
"loss": 2.7691,
"step": 1195
},
{
"epoch": 0.25,
"grad_norm": 1.010151743888855,
"learning_rate": 8.714013137226411e-05,
"loss": 2.7705,
"step": 1200
},
{
"epoch": 0.25,
"grad_norm": 0.9873592853546143,
"learning_rate": 8.702830714729628e-05,
"loss": 2.8599,
"step": 1205
},
{
"epoch": 0.25,
"grad_norm": 1.074005365371704,
"learning_rate": 8.691607125443153e-05,
"loss": 2.7302,
"step": 1210
},
{
"epoch": 0.25,
"grad_norm": 0.9528563022613525,
"learning_rate": 8.680342494146967e-05,
"loss": 2.7571,
"step": 1215
},
{
"epoch": 0.25,
"grad_norm": 0.9532042145729065,
"learning_rate": 8.66903694607734e-05,
"loss": 2.78,
"step": 1220
},
{
"epoch": 0.25,
"grad_norm": 0.9630159735679626,
"learning_rate": 8.65769060692544e-05,
"loss": 2.7383,
"step": 1225
},
{
"epoch": 0.26,
"grad_norm": 0.9430820941925049,
"learning_rate": 8.646303602835936e-05,
"loss": 2.8103,
"step": 1230
},
{
"epoch": 0.26,
"grad_norm": 0.995627760887146,
"learning_rate": 8.634876060405597e-05,
"loss": 2.783,
"step": 1235
},
{
"epoch": 0.26,
"grad_norm": 0.9373460412025452,
"learning_rate": 8.623408106681884e-05,
"loss": 2.748,
"step": 1240
},
{
"epoch": 0.26,
"grad_norm": 0.9708099961280823,
"learning_rate": 8.611899869161535e-05,
"loss": 2.8299,
"step": 1245
},
{
"epoch": 0.26,
"grad_norm": 1.0436069965362549,
"learning_rate": 8.600351475789147e-05,
"loss": 2.7038,
"step": 1250
},
{
"epoch": 0.26,
"grad_norm": 0.9993197321891785,
"learning_rate": 8.588763054955764e-05,
"loss": 2.766,
"step": 1255
},
{
"epoch": 0.26,
"grad_norm": 0.9319871664047241,
"learning_rate": 8.57713473549743e-05,
"loss": 2.7069,
"step": 1260
},
{
"epoch": 0.26,
"grad_norm": 0.9643238186836243,
"learning_rate": 8.565466646693778e-05,
"loss": 2.7478,
"step": 1265
},
{
"epoch": 0.26,
"grad_norm": 0.9584718942642212,
"learning_rate": 8.553758918266578e-05,
"loss": 2.7739,
"step": 1270
},
{
"epoch": 0.27,
"grad_norm": 0.9439178109169006,
"learning_rate": 8.5420116803783e-05,
"loss": 2.7627,
"step": 1275
},
{
"epoch": 0.27,
"grad_norm": 0.9363665580749512,
"learning_rate": 8.530225063630668e-05,
"loss": 2.8139,
"step": 1280
},
{
"epoch": 0.27,
"grad_norm": 0.9967987537384033,
"learning_rate": 8.518399199063205e-05,
"loss": 2.733,
"step": 1285
},
{
"epoch": 0.27,
"grad_norm": 1.022873878479004,
"learning_rate": 8.50653421815178e-05,
"loss": 2.8,
"step": 1290
},
{
"epoch": 0.27,
"grad_norm": 1.0046098232269287,
"learning_rate": 8.494630252807138e-05,
"loss": 2.7536,
"step": 1295
},
{
"epoch": 0.27,
"grad_norm": 0.9903509020805359,
"learning_rate": 8.482687435373449e-05,
"loss": 2.7765,
"step": 1300
},
{
"epoch": 0.27,
"grad_norm": 1.0031121969223022,
"learning_rate": 8.470705898626817e-05,
"loss": 2.7756,
"step": 1305
},
{
"epoch": 0.27,
"grad_norm": 0.9923064112663269,
"learning_rate": 8.458685775773822e-05,
"loss": 2.807,
"step": 1310
},
{
"epoch": 0.27,
"grad_norm": 1.0844812393188477,
"learning_rate": 8.446627200450025e-05,
"loss": 2.7844,
"step": 1315
},
{
"epoch": 0.27,
"grad_norm": 0.9454586505889893,
"learning_rate": 8.434530306718493e-05,
"loss": 2.7602,
"step": 1320
},
{
"epoch": 0.28,
"grad_norm": 0.9430227279663086,
"learning_rate": 8.4223952290683e-05,
"loss": 2.7727,
"step": 1325
},
{
"epoch": 0.28,
"grad_norm": 0.9255975484848022,
"learning_rate": 8.41022210241304e-05,
"loss": 2.6946,
"step": 1330
},
{
"epoch": 0.28,
"grad_norm": 1.0021836757659912,
"learning_rate": 8.398011062089316e-05,
"loss": 2.6874,
"step": 1335
},
{
"epoch": 0.28,
"grad_norm": 0.9358190298080444,
"learning_rate": 8.385762243855249e-05,
"loss": 2.7553,
"step": 1340
},
{
"epoch": 0.28,
"grad_norm": 0.9901856184005737,
"learning_rate": 8.373475783888958e-05,
"loss": 2.7411,
"step": 1345
},
{
"epoch": 0.28,
"grad_norm": 0.959700882434845,
"learning_rate": 8.36115181878705e-05,
"loss": 2.6464,
"step": 1350
},
{
"epoch": 0.28,
"grad_norm": 1.0037727355957031,
"learning_rate": 8.348790485563101e-05,
"loss": 2.711,
"step": 1355
},
{
"epoch": 0.28,
"grad_norm": 1.0071569681167603,
"learning_rate": 8.336391921646134e-05,
"loss": 2.7203,
"step": 1360
},
{
"epoch": 0.28,
"grad_norm": 1.0054467916488647,
"learning_rate": 8.323956264879089e-05,
"loss": 2.6529,
"step": 1365
},
{
"epoch": 0.28,
"grad_norm": 0.9445329904556274,
"learning_rate": 8.311483653517294e-05,
"loss": 2.7707,
"step": 1370
},
{
"epoch": 0.29,
"grad_norm": 1.0042121410369873,
"learning_rate": 8.298974226226919e-05,
"loss": 2.6704,
"step": 1375
},
{
"epoch": 0.29,
"grad_norm": 0.9446637034416199,
"learning_rate": 8.28642812208345e-05,
"loss": 2.7243,
"step": 1380
},
{
"epoch": 0.29,
"grad_norm": 1.0029373168945312,
"learning_rate": 8.273845480570123e-05,
"loss": 2.8032,
"step": 1385
},
{
"epoch": 0.29,
"grad_norm": 1.0980321168899536,
"learning_rate": 8.26122644157639e-05,
"loss": 2.7051,
"step": 1390
},
{
"epoch": 0.29,
"grad_norm": 0.9893448948860168,
"learning_rate": 8.248571145396362e-05,
"loss": 2.7615,
"step": 1395
},
{
"epoch": 0.29,
"grad_norm": 1.026167392730713,
"learning_rate": 8.235879732727236e-05,
"loss": 2.7021,
"step": 1400
},
{
"epoch": 0.29,
"grad_norm": 1.019116997718811,
"learning_rate": 8.223152344667745e-05,
"loss": 2.7147,
"step": 1405
},
{
"epoch": 0.29,
"grad_norm": 0.9762393832206726,
"learning_rate": 8.21038912271658e-05,
"loss": 2.7423,
"step": 1410
},
{
"epoch": 0.29,
"grad_norm": 0.9256317615509033,
"learning_rate": 8.197590208770824e-05,
"loss": 2.6669,
"step": 1415
},
{
"epoch": 0.3,
"grad_norm": 0.9092603325843811,
"learning_rate": 8.184755745124371e-05,
"loss": 2.7058,
"step": 1420
},
{
"epoch": 0.3,
"grad_norm": 0.9133468270301819,
"learning_rate": 8.171885874466342e-05,
"loss": 2.7317,
"step": 1425
},
{
"epoch": 0.3,
"grad_norm": 0.9461574554443359,
"learning_rate": 8.158980739879507e-05,
"loss": 2.7046,
"step": 1430
},
{
"epoch": 0.3,
"grad_norm": 0.8915404081344604,
"learning_rate": 8.146040484838677e-05,
"loss": 2.7193,
"step": 1435
},
{
"epoch": 0.3,
"grad_norm": 0.9734158515930176,
"learning_rate": 8.133065253209132e-05,
"loss": 2.6976,
"step": 1440
},
{
"epoch": 0.3,
"grad_norm": 0.9868660569190979,
"learning_rate": 8.120055189245e-05,
"loss": 2.6721,
"step": 1445
},
{
"epoch": 0.3,
"grad_norm": 0.953726589679718,
"learning_rate": 8.10701043758767e-05,
"loss": 2.6958,
"step": 1450
},
{
"epoch": 0.3,
"grad_norm": 0.9402896761894226,
"learning_rate": 8.093931143264174e-05,
"loss": 2.7276,
"step": 1455
},
{
"epoch": 0.3,
"grad_norm": 1.0774366855621338,
"learning_rate": 8.080817451685576e-05,
"loss": 2.6766,
"step": 1460
},
{
"epoch": 0.3,
"grad_norm": 1.0757198333740234,
"learning_rate": 8.067669508645356e-05,
"loss": 2.6647,
"step": 1465
},
{
"epoch": 0.31,
"grad_norm": 0.9260072112083435,
"learning_rate": 8.054487460317797e-05,
"loss": 2.6653,
"step": 1470
},
{
"epoch": 0.31,
"grad_norm": 0.9869867563247681,
"learning_rate": 8.041271453256345e-05,
"loss": 2.7253,
"step": 1475
},
{
"epoch": 0.31,
"grad_norm": 0.9941818118095398,
"learning_rate": 8.02802163439199e-05,
"loss": 2.7085,
"step": 1480
},
{
"epoch": 0.31,
"grad_norm": 0.9644975066184998,
"learning_rate": 8.01473815103163e-05,
"loss": 2.7263,
"step": 1485
},
{
"epoch": 0.31,
"grad_norm": 0.9745631814002991,
"learning_rate": 8.001421150856434e-05,
"loss": 2.6906,
"step": 1490
},
{
"epoch": 0.31,
"grad_norm": 0.9564938545227051,
"learning_rate": 7.988070781920197e-05,
"loss": 2.6963,
"step": 1495
},
{
"epoch": 0.31,
"grad_norm": 1.0416967868804932,
"learning_rate": 7.9746871926477e-05,
"loss": 2.6825,
"step": 1500
},
{
"epoch": 0.31,
"grad_norm": 1.0030397176742554,
"learning_rate": 7.961270531833052e-05,
"loss": 2.6642,
"step": 1505
},
{
"epoch": 0.31,
"grad_norm": 0.9303964972496033,
"learning_rate": 7.947820948638045e-05,
"loss": 2.6724,
"step": 1510
},
{
"epoch": 0.31,
"grad_norm": 0.8955966234207153,
"learning_rate": 7.934338592590486e-05,
"loss": 2.7086,
"step": 1515
},
{
"epoch": 0.32,
"grad_norm": 0.9256662130355835,
"learning_rate": 7.92082361358254e-05,
"loss": 2.6202,
"step": 1520
},
{
"epoch": 0.32,
"grad_norm": 1.0176639556884766,
"learning_rate": 7.907276161869065e-05,
"loss": 2.6799,
"step": 1525
},
{
"epoch": 0.32,
"grad_norm": 0.9430251121520996,
"learning_rate": 7.893696388065936e-05,
"loss": 2.6526,
"step": 1530
},
{
"epoch": 0.32,
"grad_norm": 0.9827606678009033,
"learning_rate": 7.88008444314838e-05,
"loss": 2.722,
"step": 1535
},
{
"epoch": 0.32,
"grad_norm": 0.9310677647590637,
"learning_rate": 7.866440478449283e-05,
"loss": 2.7176,
"step": 1540
},
{
"epoch": 0.32,
"grad_norm": 0.9184085130691528,
"learning_rate": 7.852764645657522e-05,
"loss": 2.7291,
"step": 1545
},
{
"epoch": 0.32,
"grad_norm": 0.9483653903007507,
"learning_rate": 7.839057096816271e-05,
"loss": 2.6791,
"step": 1550
},
{
"epoch": 0.32,
"grad_norm": 0.924527645111084,
"learning_rate": 7.82531798432131e-05,
"loss": 2.7046,
"step": 1555
},
{
"epoch": 0.32,
"grad_norm": 0.8874161243438721,
"learning_rate": 7.811547460919333e-05,
"loss": 2.7334,
"step": 1560
},
{
"epoch": 0.33,
"grad_norm": 1.0274156332015991,
"learning_rate": 7.797745679706254e-05,
"loss": 2.7015,
"step": 1565
},
{
"epoch": 0.33,
"grad_norm": 0.9184699654579163,
"learning_rate": 7.783912794125496e-05,
"loss": 2.6723,
"step": 1570
},
{
"epoch": 0.33,
"grad_norm": 0.9533879160881042,
"learning_rate": 7.770048957966291e-05,
"loss": 2.6385,
"step": 1575
},
{
"epoch": 0.33,
"grad_norm": 0.9507087469100952,
"learning_rate": 7.756154325361967e-05,
"loss": 2.7001,
"step": 1580
},
{
"epoch": 0.33,
"grad_norm": 0.9311648011207581,
"learning_rate": 7.74222905078824e-05,
"loss": 2.688,
"step": 1585
},
{
"epoch": 0.33,
"grad_norm": 0.9226982593536377,
"learning_rate": 7.728273289061489e-05,
"loss": 2.7079,
"step": 1590
},
{
"epoch": 0.33,
"grad_norm": 0.9227792024612427,
"learning_rate": 7.714287195337044e-05,
"loss": 2.6676,
"step": 1595
},
{
"epoch": 0.33,
"grad_norm": 0.9729253053665161,
"learning_rate": 7.700270925107448e-05,
"loss": 2.7251,
"step": 1600
},
{
"epoch": 0.33,
"grad_norm": 0.9518150091171265,
"learning_rate": 7.686224634200742e-05,
"loss": 2.6562,
"step": 1605
},
{
"epoch": 0.33,
"grad_norm": 0.9133305549621582,
"learning_rate": 7.672148478778722e-05,
"loss": 2.599,
"step": 1610
},
{
"epoch": 0.34,
"grad_norm": 1.0146985054016113,
"learning_rate": 7.658042615335212e-05,
"loss": 2.7134,
"step": 1615
},
{
"epoch": 0.34,
"grad_norm": 0.8882867097854614,
"learning_rate": 7.643907200694318e-05,
"loss": 2.6559,
"step": 1620
},
{
"epoch": 0.34,
"grad_norm": 0.9086301326751709,
"learning_rate": 7.629742392008684e-05,
"loss": 2.6441,
"step": 1625
},
{
"epoch": 0.34,
"grad_norm": 0.8805155158042908,
"learning_rate": 7.615548346757749e-05,
"loss": 2.6748,
"step": 1630
},
{
"epoch": 0.34,
"grad_norm": 0.8923571109771729,
"learning_rate": 7.60132522274599e-05,
"loss": 2.7604,
"step": 1635
},
{
"epoch": 0.34,
"grad_norm": 0.8995354175567627,
"learning_rate": 7.587073178101178e-05,
"loss": 2.695,
"step": 1640
},
{
"epoch": 0.34,
"grad_norm": 0.9369312524795532,
"learning_rate": 7.572792371272609e-05,
"loss": 2.6368,
"step": 1645
},
{
"epoch": 0.34,
"grad_norm": 0.8683916926383972,
"learning_rate": 7.55848296102935e-05,
"loss": 2.7157,
"step": 1650
},
{
"epoch": 0.34,
"grad_norm": 0.8939131498336792,
"learning_rate": 7.544145106458465e-05,
"loss": 2.6221,
"step": 1655
},
{
"epoch": 0.35,
"grad_norm": 0.9551239609718323,
"learning_rate": 7.529778966963259e-05,
"loss": 2.6503,
"step": 1660
},
{
"epoch": 0.35,
"grad_norm": 0.9065306186676025,
"learning_rate": 7.515384702261496e-05,
"loss": 2.6772,
"step": 1665
},
{
"epoch": 0.35,
"grad_norm": 0.8881747722625732,
"learning_rate": 7.500962472383627e-05,
"loss": 2.706,
"step": 1670
},
{
"epoch": 0.35,
"grad_norm": 0.9851480722427368,
"learning_rate": 7.486512437671011e-05,
"loss": 2.7238,
"step": 1675
},
{
"epoch": 0.35,
"grad_norm": 0.8894233703613281,
"learning_rate": 7.472034758774128e-05,
"loss": 2.6108,
"step": 1680
},
{
"epoch": 0.35,
"grad_norm": 0.9191601872444153,
"learning_rate": 7.457529596650797e-05,
"loss": 2.661,
"step": 1685
},
{
"epoch": 0.35,
"grad_norm": 0.9564477801322937,
"learning_rate": 7.442997112564392e-05,
"loss": 2.6915,
"step": 1690
},
{
"epoch": 0.35,
"grad_norm": 0.864436149597168,
"learning_rate": 7.428437468082037e-05,
"loss": 2.5873,
"step": 1695
},
{
"epoch": 0.35,
"grad_norm": 0.9332630634307861,
"learning_rate": 7.413850825072817e-05,
"loss": 2.7157,
"step": 1700
},
{
"epoch": 0.35,
"grad_norm": 0.8967276215553284,
"learning_rate": 7.39923734570598e-05,
"loss": 2.6271,
"step": 1705
},
{
"epoch": 0.36,
"grad_norm": 0.8776575326919556,
"learning_rate": 7.384597192449126e-05,
"loss": 2.6399,
"step": 1710
},
{
"epoch": 0.36,
"grad_norm": 0.8810414671897888,
"learning_rate": 7.369930528066412e-05,
"loss": 2.6278,
"step": 1715
},
{
"epoch": 0.36,
"grad_norm": 0.9536029100418091,
"learning_rate": 7.355237515616732e-05,
"loss": 2.6533,
"step": 1720
},
{
"epoch": 0.36,
"grad_norm": 0.8947739601135254,
"learning_rate": 7.340518318451914e-05,
"loss": 2.7199,
"step": 1725
},
{
"epoch": 0.36,
"grad_norm": 0.8960339426994324,
"learning_rate": 7.325773100214893e-05,
"loss": 2.6842,
"step": 1730
},
{
"epoch": 0.36,
"grad_norm": 0.9321779608726501,
"learning_rate": 7.311002024837899e-05,
"loss": 2.6393,
"step": 1735
},
{
"epoch": 0.36,
"grad_norm": 0.9875207543373108,
"learning_rate": 7.296205256540633e-05,
"loss": 2.631,
"step": 1740
},
{
"epoch": 0.36,
"grad_norm": 0.9218025803565979,
"learning_rate": 7.281382959828443e-05,
"loss": 2.6646,
"step": 1745
},
{
"epoch": 0.36,
"grad_norm": 0.8971613645553589,
"learning_rate": 7.26653529949049e-05,
"loss": 2.6052,
"step": 1750
},
{
"epoch": 0.36,
"grad_norm": 0.9385172128677368,
"learning_rate": 7.25166244059792e-05,
"loss": 2.6346,
"step": 1755
},
{
"epoch": 0.37,
"grad_norm": 0.9440680146217346,
"learning_rate": 7.236764548502029e-05,
"loss": 2.6613,
"step": 1760
},
{
"epoch": 0.37,
"grad_norm": 0.8777347207069397,
"learning_rate": 7.221841788832421e-05,
"loss": 2.6559,
"step": 1765
},
{
"epoch": 0.37,
"grad_norm": 0.9107197523117065,
"learning_rate": 7.206894327495173e-05,
"loss": 2.5924,
"step": 1770
},
{
"epoch": 0.37,
"grad_norm": 0.9251880645751953,
"learning_rate": 7.191922330670982e-05,
"loss": 2.5643,
"step": 1775
},
{
"epoch": 0.37,
"grad_norm": 0.9331080913543701,
"learning_rate": 7.176925964813326e-05,
"loss": 2.59,
"step": 1780
},
{
"epoch": 0.37,
"grad_norm": 0.8959988951683044,
"learning_rate": 7.161905396646607e-05,
"loss": 2.5484,
"step": 1785
},
{
"epoch": 0.37,
"grad_norm": 0.9152870774269104,
"learning_rate": 7.146860793164299e-05,
"loss": 2.6532,
"step": 1790
},
{
"epoch": 0.37,
"grad_norm": 0.8978911638259888,
"learning_rate": 7.131792321627098e-05,
"loss": 2.6451,
"step": 1795
},
{
"epoch": 0.37,
"grad_norm": 0.8633766174316406,
"learning_rate": 7.116700149561048e-05,
"loss": 2.6948,
"step": 1800
},
{
"epoch": 0.38,
"grad_norm": 0.8857610821723938,
"learning_rate": 7.101584444755696e-05,
"loss": 2.713,
"step": 1805
},
{
"epoch": 0.38,
"grad_norm": 1.3501237630844116,
"learning_rate": 7.086445375262212e-05,
"loss": 2.6084,
"step": 1810
},
{
"epoch": 0.38,
"grad_norm": 0.91219562292099,
"learning_rate": 7.071283109391528e-05,
"loss": 2.6028,
"step": 1815
},
{
"epoch": 0.38,
"grad_norm": 0.8799169659614563,
"learning_rate": 7.056097815712466e-05,
"loss": 2.632,
"step": 1820
},
{
"epoch": 0.38,
"grad_norm": 0.9489891529083252,
"learning_rate": 7.040889663049862e-05,
"loss": 2.6587,
"step": 1825
},
{
"epoch": 0.38,
"grad_norm": 0.8914775252342224,
"learning_rate": 7.025658820482693e-05,
"loss": 2.691,
"step": 1830
},
{
"epoch": 0.38,
"grad_norm": 0.8673481941223145,
"learning_rate": 7.010405457342192e-05,
"loss": 2.5965,
"step": 1835
},
{
"epoch": 0.38,
"grad_norm": 0.9208853244781494,
"learning_rate": 6.995129743209967e-05,
"loss": 2.6357,
"step": 1840
},
{
"epoch": 0.38,
"grad_norm": 0.8941606879234314,
"learning_rate": 6.97983184791612e-05,
"loss": 2.6235,
"step": 1845
},
{
"epoch": 0.38,
"grad_norm": 0.8743651509284973,
"learning_rate": 6.964511941537355e-05,
"loss": 2.7112,
"step": 1850
},
{
"epoch": 0.39,
"grad_norm": 0.901387095451355,
"learning_rate": 6.949170194395083e-05,
"loss": 2.6553,
"step": 1855
},
{
"epoch": 0.39,
"grad_norm": 0.8705363869667053,
"learning_rate": 6.933806777053536e-05,
"loss": 2.6379,
"step": 1860
},
{
"epoch": 0.39,
"grad_norm": 0.927294135093689,
"learning_rate": 6.918421860317872e-05,
"loss": 2.6143,
"step": 1865
},
{
"epoch": 0.39,
"grad_norm": 0.9013956785202026,
"learning_rate": 6.903015615232263e-05,
"loss": 2.5986,
"step": 1870
},
{
"epoch": 0.39,
"grad_norm": 0.8557500243186951,
"learning_rate": 6.887588213078012e-05,
"loss": 2.6973,
"step": 1875
},
{
"epoch": 0.39,
"grad_norm": 0.9382877349853516,
"learning_rate": 6.87213982537163e-05,
"loss": 2.6378,
"step": 1880
},
{
"epoch": 0.39,
"grad_norm": 0.8524401187896729,
"learning_rate": 6.856670623862943e-05,
"loss": 2.5714,
"step": 1885
},
{
"epoch": 0.39,
"grad_norm": 0.9000129699707031,
"learning_rate": 6.841180780533179e-05,
"loss": 2.6309,
"step": 1890
},
{
"epoch": 0.39,
"grad_norm": 0.885753333568573,
"learning_rate": 6.82567046759305e-05,
"loss": 2.5606,
"step": 1895
},
{
"epoch": 0.39,
"grad_norm": 0.9170253872871399,
"learning_rate": 6.810139857480844e-05,
"loss": 2.5845,
"step": 1900
},
{
"epoch": 0.4,
"grad_norm": 0.9164355993270874,
"learning_rate": 6.794589122860509e-05,
"loss": 2.6819,
"step": 1905
},
{
"epoch": 0.4,
"grad_norm": 0.9106840491294861,
"learning_rate": 6.779018436619725e-05,
"loss": 2.649,
"step": 1910
},
{
"epoch": 0.4,
"grad_norm": 0.9228881001472473,
"learning_rate": 6.763427971867992e-05,
"loss": 2.6524,
"step": 1915
},
{
"epoch": 0.4,
"grad_norm": 0.9677906036376953,
"learning_rate": 6.747817901934699e-05,
"loss": 2.5003,
"step": 1920
},
{
"epoch": 0.4,
"grad_norm": 0.8591650724411011,
"learning_rate": 6.732188400367197e-05,
"loss": 2.5631,
"step": 1925
},
{
"epoch": 0.4,
"grad_norm": 0.8764705657958984,
"learning_rate": 6.716539640928871e-05,
"loss": 2.6895,
"step": 1930
},
{
"epoch": 0.4,
"grad_norm": 0.9018672704696655,
"learning_rate": 6.70087179759721e-05,
"loss": 2.6417,
"step": 1935
},
{
"epoch": 0.4,
"grad_norm": 0.8887489438056946,
"learning_rate": 6.685185044561874e-05,
"loss": 2.5926,
"step": 1940
},
{
"epoch": 0.4,
"grad_norm": 0.9418622851371765,
"learning_rate": 6.669479556222747e-05,
"loss": 2.6149,
"step": 1945
},
{
"epoch": 0.41,
"grad_norm": 0.8539125323295593,
"learning_rate": 6.653755507188013e-05,
"loss": 2.6164,
"step": 1950
},
{
"epoch": 0.41,
"grad_norm": 0.8717061281204224,
"learning_rate": 6.638013072272205e-05,
"loss": 2.6717,
"step": 1955
},
{
"epoch": 0.41,
"grad_norm": 0.8727285861968994,
"learning_rate": 6.622252426494259e-05,
"loss": 2.659,
"step": 1960
},
{
"epoch": 0.41,
"grad_norm": 0.9590850472450256,
"learning_rate": 6.606473745075581e-05,
"loss": 2.6064,
"step": 1965
},
{
"epoch": 0.41,
"grad_norm": 0.843034565448761,
"learning_rate": 6.590677203438084e-05,
"loss": 2.5781,
"step": 1970
},
{
"epoch": 0.41,
"grad_norm": 0.9349545836448669,
"learning_rate": 6.574862977202252e-05,
"loss": 2.6224,
"step": 1975
},
{
"epoch": 0.41,
"grad_norm": 0.9160761833190918,
"learning_rate": 6.559031242185174e-05,
"loss": 2.614,
"step": 1980
},
{
"epoch": 0.41,
"grad_norm": 0.8734063506126404,
"learning_rate": 6.543182174398597e-05,
"loss": 2.6367,
"step": 1985
},
{
"epoch": 0.41,
"grad_norm": 0.9030283689498901,
"learning_rate": 6.52731595004697e-05,
"loss": 2.6176,
"step": 1990
},
{
"epoch": 0.41,
"grad_norm": 0.8655874729156494,
"learning_rate": 6.51143274552548e-05,
"loss": 2.6403,
"step": 1995
},
{
"epoch": 0.42,
"grad_norm": 0.8767572045326233,
"learning_rate": 6.495532737418098e-05,
"loss": 2.5767,
"step": 2000
},
{
"epoch": 0.42,
"grad_norm": 0.8731923699378967,
"learning_rate": 6.479616102495605e-05,
"loss": 2.6223,
"step": 2005
},
{
"epoch": 0.42,
"grad_norm": 0.8789691925048828,
"learning_rate": 6.463683017713638e-05,
"loss": 2.6176,
"step": 2010
},
{
"epoch": 0.42,
"grad_norm": 0.8845729827880859,
"learning_rate": 6.447733660210715e-05,
"loss": 2.6908,
"step": 2015
},
{
"epoch": 0.42,
"grad_norm": 0.8749977946281433,
"learning_rate": 6.431768207306272e-05,
"loss": 2.5775,
"step": 2020
},
{
"epoch": 0.42,
"grad_norm": 0.9047712683677673,
"learning_rate": 6.415786836498684e-05,
"loss": 2.5204,
"step": 2025
},
{
"epoch": 0.42,
"grad_norm": 0.8837469220161438,
"learning_rate": 6.399789725463298e-05,
"loss": 2.6177,
"step": 2030
},
{
"epoch": 0.42,
"grad_norm": 0.8328371644020081,
"learning_rate": 6.383777052050458e-05,
"loss": 2.5712,
"step": 2035
},
{
"epoch": 0.42,
"grad_norm": 0.8723359107971191,
"learning_rate": 6.367748994283518e-05,
"loss": 2.6146,
"step": 2040
},
{
"epoch": 0.43,
"grad_norm": 0.8871758580207825,
"learning_rate": 6.351705730356877e-05,
"loss": 2.6646,
"step": 2045
},
{
"epoch": 0.43,
"grad_norm": 0.8672588467597961,
"learning_rate": 6.335647438633987e-05,
"loss": 2.5867,
"step": 2050
},
{
"epoch": 0.43,
"grad_norm": 0.9469790458679199,
"learning_rate": 6.319574297645374e-05,
"loss": 2.5729,
"step": 2055
},
{
"epoch": 0.43,
"grad_norm": 0.914804995059967,
"learning_rate": 6.303486486086654e-05,
"loss": 2.5805,
"step": 2060
},
{
"epoch": 0.43,
"grad_norm": 0.8987873792648315,
"learning_rate": 6.287384182816546e-05,
"loss": 2.5373,
"step": 2065
},
{
"epoch": 0.43,
"grad_norm": 0.854649007320404,
"learning_rate": 6.271267566854883e-05,
"loss": 2.5961,
"step": 2070
},
{
"epoch": 0.43,
"grad_norm": 0.8717460632324219,
"learning_rate": 6.255136817380618e-05,
"loss": 2.6651,
"step": 2075
},
{
"epoch": 0.43,
"grad_norm": 0.8450790047645569,
"learning_rate": 6.23899211372984e-05,
"loss": 2.5674,
"step": 2080
},
{
"epoch": 0.43,
"grad_norm": 0.8864030241966248,
"learning_rate": 6.222833635393772e-05,
"loss": 2.5799,
"step": 2085
},
{
"epoch": 0.43,
"grad_norm": 0.8747897148132324,
"learning_rate": 6.206661562016782e-05,
"loss": 2.6777,
"step": 2090
},
{
"epoch": 0.44,
"grad_norm": 0.9257798194885254,
"learning_rate": 6.190476073394382e-05,
"loss": 2.6753,
"step": 2095
},
{
"epoch": 0.44,
"grad_norm": 0.894865870475769,
"learning_rate": 6.17427734947123e-05,
"loss": 2.5385,
"step": 2100
},
{
"epoch": 0.44,
"grad_norm": 0.8680559992790222,
"learning_rate": 6.158065570339127e-05,
"loss": 2.623,
"step": 2105
},
{
"epoch": 0.44,
"grad_norm": 0.8899837136268616,
"learning_rate": 6.141840916235021e-05,
"loss": 2.5333,
"step": 2110
},
{
"epoch": 0.44,
"grad_norm": 0.87552410364151,
"learning_rate": 6.125603567539001e-05,
"loss": 2.6303,
"step": 2115
},
{
"epoch": 0.44,
"grad_norm": 0.8710044026374817,
"learning_rate": 6.109353704772284e-05,
"loss": 2.5336,
"step": 2120
},
{
"epoch": 0.44,
"grad_norm": 0.9061421751976013,
"learning_rate": 6.0930915085952164e-05,
"loss": 2.6329,
"step": 2125
},
{
"epoch": 0.44,
"grad_norm": 0.8576582074165344,
"learning_rate": 6.076817159805267e-05,
"loss": 2.6401,
"step": 2130
},
{
"epoch": 0.44,
"grad_norm": 0.9094964265823364,
"learning_rate": 6.06053083933501e-05,
"loss": 2.5809,
"step": 2135
},
{
"epoch": 0.44,
"grad_norm": 0.944002628326416,
"learning_rate": 6.044232728250116e-05,
"loss": 2.6436,
"step": 2140
},
{
"epoch": 0.45,
"grad_norm": 0.8766120672225952,
"learning_rate": 6.027923007747339e-05,
"loss": 2.6337,
"step": 2145
},
{
"epoch": 0.45,
"grad_norm": 0.8836204409599304,
"learning_rate": 6.011601859152506e-05,
"loss": 2.5756,
"step": 2150
},
{
"epoch": 0.45,
"grad_norm": 0.8940385580062866,
"learning_rate": 5.995269463918495e-05,
"loss": 2.5361,
"step": 2155
},
{
"epoch": 0.45,
"grad_norm": 0.8867673873901367,
"learning_rate": 5.97892600362322e-05,
"loss": 2.5409,
"step": 2160
},
{
"epoch": 0.45,
"grad_norm": 0.8697126507759094,
"learning_rate": 5.962571659967614e-05,
"loss": 2.5674,
"step": 2165
},
{
"epoch": 0.45,
"grad_norm": 0.8790608644485474,
"learning_rate": 5.946206614773606e-05,
"loss": 2.6999,
"step": 2170
},
{
"epoch": 0.45,
"grad_norm": 0.9058618545532227,
"learning_rate": 5.929831049982103e-05,
"loss": 2.5889,
"step": 2175
},
{
"epoch": 0.45,
"grad_norm": 0.8863576650619507,
"learning_rate": 5.9134451476509633e-05,
"loss": 2.633,
"step": 2180
},
{
"epoch": 0.45,
"grad_norm": 0.8856499791145325,
"learning_rate": 5.897049089952974e-05,
"loss": 2.5503,
"step": 2185
},
{
"epoch": 0.46,
"grad_norm": 0.8821172714233398,
"learning_rate": 5.880643059173826e-05,
"loss": 2.5299,
"step": 2190
},
{
"epoch": 0.46,
"grad_norm": 0.9050044417381287,
"learning_rate": 5.864227237710093e-05,
"loss": 2.5789,
"step": 2195
},
{
"epoch": 0.46,
"grad_norm": 0.839778482913971,
"learning_rate": 5.847801808067189e-05,
"loss": 2.5876,
"step": 2200
},
{
"epoch": 0.46,
"grad_norm": 0.8772847652435303,
"learning_rate": 5.831366952857357e-05,
"loss": 2.6012,
"step": 2205
},
{
"epoch": 0.46,
"grad_norm": 0.9180009365081787,
"learning_rate": 5.814922854797622e-05,
"loss": 2.4858,
"step": 2210
},
{
"epoch": 0.46,
"grad_norm": 0.8592243194580078,
"learning_rate": 5.798469696707775e-05,
"loss": 2.5631,
"step": 2215
},
{
"epoch": 0.46,
"grad_norm": 0.9094417095184326,
"learning_rate": 5.782007661508331e-05,
"loss": 2.5793,
"step": 2220
},
{
"epoch": 0.46,
"grad_norm": 0.9196504354476929,
"learning_rate": 5.765536932218495e-05,
"loss": 2.5644,
"step": 2225
},
{
"epoch": 0.46,
"grad_norm": 0.8280788064002991,
"learning_rate": 5.7490576919541315e-05,
"loss": 2.5765,
"step": 2230
},
{
"epoch": 0.46,
"grad_norm": 0.8863653540611267,
"learning_rate": 5.732570123925729e-05,
"loss": 2.5885,
"step": 2235
},
{
"epoch": 0.47,
"grad_norm": 1.1553986072540283,
"learning_rate": 5.7160744114363593e-05,
"loss": 2.5468,
"step": 2240
},
{
"epoch": 0.47,
"grad_norm": 0.867210865020752,
"learning_rate": 5.699570737879641e-05,
"loss": 2.5728,
"step": 2245
},
{
"epoch": 0.47,
"grad_norm": 0.8991091251373291,
"learning_rate": 5.683059286737702e-05,
"loss": 2.6365,
"step": 2250
},
{
"epoch": 0.47,
"grad_norm": 0.8775330185890198,
"learning_rate": 5.666540241579139e-05,
"loss": 2.596,
"step": 2255
},
{
"epoch": 0.47,
"grad_norm": 0.870933473110199,
"learning_rate": 5.6500137860569766e-05,
"loss": 2.6069,
"step": 2260
},
{
"epoch": 0.47,
"grad_norm": 0.8916445374488831,
"learning_rate": 5.633480103906624e-05,
"loss": 2.5785,
"step": 2265
},
{
"epoch": 0.47,
"grad_norm": 0.9029600620269775,
"learning_rate": 5.616939378943834e-05,
"loss": 2.5859,
"step": 2270
},
{
"epoch": 0.47,
"grad_norm": 0.8787057995796204,
"learning_rate": 5.6003917950626595e-05,
"loss": 2.5044,
"step": 2275
},
{
"epoch": 0.47,
"grad_norm": 0.865903913974762,
"learning_rate": 5.583837536233407e-05,
"loss": 2.6433,
"step": 2280
},
{
"epoch": 0.47,
"grad_norm": 0.9330430626869202,
"learning_rate": 5.567276786500596e-05,
"loss": 2.5532,
"step": 2285
},
{
"epoch": 0.48,
"grad_norm": 0.8771436810493469,
"learning_rate": 5.5507097299809054e-05,
"loss": 2.5918,
"step": 2290
},
{
"epoch": 0.48,
"grad_norm": 1.0670830011367798,
"learning_rate": 5.534136550861133e-05,
"loss": 2.5587,
"step": 2295
},
{
"epoch": 0.48,
"grad_norm": 0.8688689470291138,
"learning_rate": 5.5175574333961465e-05,
"loss": 2.5261,
"step": 2300
},
{
"epoch": 0.48,
"grad_norm": 0.8848040103912354,
"learning_rate": 5.500972561906832e-05,
"loss": 2.5148,
"step": 2305
},
{
"epoch": 0.48,
"grad_norm": 0.8618828654289246,
"learning_rate": 5.484382120778048e-05,
"loss": 2.5009,
"step": 2310
},
{
"epoch": 0.48,
"grad_norm": 0.8698475956916809,
"learning_rate": 5.467786294456575e-05,
"loss": 2.6039,
"step": 2315
},
{
"epoch": 0.48,
"grad_norm": 0.8511532545089722,
"learning_rate": 5.451185267449061e-05,
"loss": 2.534,
"step": 2320
},
{
"epoch": 0.48,
"grad_norm": 0.8999781012535095,
"learning_rate": 5.43457922431998e-05,
"loss": 2.5156,
"step": 2325
},
{
"epoch": 0.48,
"grad_norm": 0.8377700448036194,
"learning_rate": 5.417968349689566e-05,
"loss": 2.6344,
"step": 2330
},
{
"epoch": 0.49,
"grad_norm": 0.8604027032852173,
"learning_rate": 5.401352828231772e-05,
"loss": 2.4831,
"step": 2335
},
{
"epoch": 0.49,
"grad_norm": 0.9010083675384521,
"learning_rate": 5.384732844672211e-05,
"loss": 2.5416,
"step": 2340
},
{
"epoch": 0.49,
"grad_norm": 0.8889869451522827,
"learning_rate": 5.368108583786107e-05,
"loss": 2.5559,
"step": 2345
},
{
"epoch": 0.49,
"grad_norm": 0.8636956810951233,
"learning_rate": 5.3514802303962344e-05,
"loss": 2.5414,
"step": 2350
},
{
"epoch": 0.49,
"grad_norm": 0.8281151652336121,
"learning_rate": 5.334847969370868e-05,
"loss": 2.5195,
"step": 2355
},
{
"epoch": 0.49,
"grad_norm": 0.8393617868423462,
"learning_rate": 5.3182119856217284e-05,
"loss": 2.5848,
"step": 2360
},
{
"epoch": 0.49,
"grad_norm": 0.870062530040741,
"learning_rate": 5.3015724641019214e-05,
"loss": 2.5202,
"step": 2365
},
{
"epoch": 0.49,
"grad_norm": 0.8297975659370422,
"learning_rate": 5.284929589803884e-05,
"loss": 2.4958,
"step": 2370
},
{
"epoch": 0.49,
"grad_norm": 0.879835307598114,
"learning_rate": 5.2682835477573336e-05,
"loss": 2.5438,
"step": 2375
},
{
"epoch": 0.49,
"grad_norm": 0.8568702340126038,
"learning_rate": 5.2516345230271965e-05,
"loss": 2.5385,
"step": 2380
},
{
"epoch": 0.5,
"grad_norm": 0.8864716291427612,
"learning_rate": 5.234982700711569e-05,
"loss": 2.4734,
"step": 2385
},
{
"epoch": 0.5,
"grad_norm": 0.8865468502044678,
"learning_rate": 5.218328265939643e-05,
"loss": 2.5695,
"step": 2390
},
{
"epoch": 0.5,
"grad_norm": 0.8695160746574402,
"learning_rate": 5.201671403869657e-05,
"loss": 2.5463,
"step": 2395
},
{
"epoch": 0.5,
"grad_norm": 0.8666235208511353,
"learning_rate": 5.1850122996868366e-05,
"loss": 2.5062,
"step": 2400
},
{
"epoch": 0.5,
"grad_norm": 0.8991391658782959,
"learning_rate": 5.168351138601334e-05,
"loss": 2.5438,
"step": 2405
},
{
"epoch": 0.5,
"grad_norm": 0.85994553565979,
"learning_rate": 5.1516881058461675e-05,
"loss": 2.4915,
"step": 2410
},
{
"epoch": 0.5,
"grad_norm": 0.9064061045646667,
"learning_rate": 5.135023386675166e-05,
"loss": 2.591,
"step": 2415
},
{
"epoch": 0.5,
"grad_norm": 0.9430391192436218,
"learning_rate": 5.118357166360906e-05,
"loss": 2.5147,
"step": 2420
},
{
"epoch": 0.5,
"grad_norm": 0.8593295216560364,
"learning_rate": 5.101689630192655e-05,
"loss": 2.5574,
"step": 2425
},
{
"epoch": 0.51,
"grad_norm": 0.9219358563423157,
"learning_rate": 5.085020963474307e-05,
"loss": 2.5358,
"step": 2430
},
{
"epoch": 0.51,
"grad_norm": 0.8650044798851013,
"learning_rate": 5.068351351522329e-05,
"loss": 2.5807,
"step": 2435
},
{
"epoch": 0.51,
"grad_norm": 0.841683566570282,
"learning_rate": 5.0516809796636935e-05,
"loss": 2.5602,
"step": 2440
},
{
"epoch": 0.51,
"grad_norm": 0.8546695709228516,
"learning_rate": 5.035010033233821e-05,
"loss": 2.5168,
"step": 2445
},
{
"epoch": 0.51,
"grad_norm": 0.8721759915351868,
"learning_rate": 5.018338697574523e-05,
"loss": 2.5257,
"step": 2450
},
{
"epoch": 0.51,
"grad_norm": 0.8715779185295105,
"learning_rate": 5.0016671580319354e-05,
"loss": 2.5634,
"step": 2455
},
{
"epoch": 0.51,
"grad_norm": 0.8524410128593445,
"learning_rate": 4.984995599954461e-05,
"loss": 2.5336,
"step": 2460
},
{
"epoch": 0.51,
"grad_norm": 0.8427689671516418,
"learning_rate": 4.968324208690712e-05,
"loss": 2.5532,
"step": 2465
},
{
"epoch": 0.51,
"grad_norm": 0.8556156158447266,
"learning_rate": 4.951653169587441e-05,
"loss": 2.5122,
"step": 2470
},
{
"epoch": 0.51,
"grad_norm": 0.8254562616348267,
"learning_rate": 4.93498266798749e-05,
"loss": 2.553,
"step": 2475
},
{
"epoch": 0.52,
"grad_norm": 0.8645380735397339,
"learning_rate": 4.918312889227722e-05,
"loss": 2.5821,
"step": 2480
},
{
"epoch": 0.52,
"grad_norm": 0.8345059156417847,
"learning_rate": 4.901644018636966e-05,
"loss": 2.4992,
"step": 2485
},
{
"epoch": 0.52,
"grad_norm": 0.874608039855957,
"learning_rate": 4.8849762415339526e-05,
"loss": 2.5255,
"step": 2490
},
{
"epoch": 0.52,
"grad_norm": 0.8931384086608887,
"learning_rate": 4.868309743225256e-05,
"loss": 2.5809,
"step": 2495
},
{
"epoch": 0.52,
"grad_norm": 0.8413265943527222,
"learning_rate": 4.851644709003233e-05,
"loss": 2.5034,
"step": 2500
},
{
"epoch": 0.52,
"grad_norm": 0.8653162121772766,
"learning_rate": 4.834981324143964e-05,
"loss": 2.613,
"step": 2505
},
{
"epoch": 0.52,
"grad_norm": 0.8855701684951782,
"learning_rate": 4.818319773905191e-05,
"loss": 2.527,
"step": 2510
},
{
"epoch": 0.52,
"grad_norm": 0.8463268280029297,
"learning_rate": 4.801660243524261e-05,
"loss": 2.4958,
"step": 2515
},
{
"epoch": 0.52,
"grad_norm": 0.8783470392227173,
"learning_rate": 4.7850029182160626e-05,
"loss": 2.5776,
"step": 2520
},
{
"epoch": 0.52,
"grad_norm": 0.8292690515518188,
"learning_rate": 4.768347983170973e-05,
"loss": 2.5522,
"step": 2525
},
{
"epoch": 0.53,
"grad_norm": 0.8445718288421631,
"learning_rate": 4.7516956235527884e-05,
"loss": 2.5385,
"step": 2530
},
{
"epoch": 0.53,
"grad_norm": 0.8653412461280823,
"learning_rate": 4.735046024496682e-05,
"loss": 2.458,
"step": 2535
},
{
"epoch": 0.53,
"grad_norm": 0.8523708581924438,
"learning_rate": 4.7183993711071286e-05,
"loss": 2.5166,
"step": 2540
},
{
"epoch": 0.53,
"grad_norm": 0.895020604133606,
"learning_rate": 4.7017558484558554e-05,
"loss": 2.563,
"step": 2545
},
{
"epoch": 0.53,
"grad_norm": 0.8685027956962585,
"learning_rate": 4.6851156415797844e-05,
"loss": 2.6202,
"step": 2550
},
{
"epoch": 0.53,
"grad_norm": 0.8594679832458496,
"learning_rate": 4.6684789354789746e-05,
"loss": 2.5202,
"step": 2555
},
{
"epoch": 0.53,
"grad_norm": 0.8136383295059204,
"learning_rate": 4.651845915114563e-05,
"loss": 2.5128,
"step": 2560
},
{
"epoch": 0.53,
"grad_norm": 0.8476276397705078,
"learning_rate": 4.6352167654067095e-05,
"loss": 2.5663,
"step": 2565
},
{
"epoch": 0.53,
"grad_norm": 0.9013127088546753,
"learning_rate": 4.618591671232544e-05,
"loss": 2.5858,
"step": 2570
},
{
"epoch": 0.54,
"grad_norm": 0.8896850943565369,
"learning_rate": 4.601970817424106e-05,
"loss": 2.5242,
"step": 2575
},
{
"epoch": 0.54,
"grad_norm": 0.8470696210861206,
"learning_rate": 4.585354388766292e-05,
"loss": 2.5343,
"step": 2580
},
{
"epoch": 0.54,
"grad_norm": 0.9307103157043457,
"learning_rate": 4.568742569994802e-05,
"loss": 2.5792,
"step": 2585
},
{
"epoch": 0.54,
"grad_norm": 0.8707754015922546,
"learning_rate": 4.552135545794086e-05,
"loss": 2.501,
"step": 2590
},
{
"epoch": 0.54,
"grad_norm": 0.87799471616745,
"learning_rate": 4.535533500795288e-05,
"loss": 2.5384,
"step": 2595
},
{
"epoch": 0.54,
"grad_norm": 0.8485037684440613,
"learning_rate": 4.5189366195741953e-05,
"loss": 2.4334,
"step": 2600
},
{
"epoch": 0.54,
"grad_norm": 0.8210888504981995,
"learning_rate": 4.502345086649186e-05,
"loss": 2.4754,
"step": 2605
},
{
"epoch": 0.54,
"grad_norm": 0.876570999622345,
"learning_rate": 4.485759086479179e-05,
"loss": 2.5798,
"step": 2610
},
{
"epoch": 0.54,
"grad_norm": 0.8352026343345642,
"learning_rate": 4.469178803461579e-05,
"loss": 2.5406,
"step": 2615
},
{
"epoch": 0.54,
"grad_norm": 0.8169410824775696,
"learning_rate": 4.4526044219302326e-05,
"loss": 2.5048,
"step": 2620
},
{
"epoch": 0.55,
"grad_norm": 0.883799135684967,
"learning_rate": 4.4360361261533745e-05,
"loss": 2.5406,
"step": 2625
},
{
"epoch": 0.55,
"grad_norm": 0.8556325435638428,
"learning_rate": 4.419474100331579e-05,
"loss": 2.5069,
"step": 2630
},
{
"epoch": 0.55,
"grad_norm": 0.8517823815345764,
"learning_rate": 4.402918528595715e-05,
"loss": 2.5081,
"step": 2635
},
{
"epoch": 0.55,
"grad_norm": 0.8794567584991455,
"learning_rate": 4.386369595004896e-05,
"loss": 2.5115,
"step": 2640
},
{
"epoch": 0.55,
"grad_norm": 0.8702395558357239,
"learning_rate": 4.3698274835444354e-05,
"loss": 2.5159,
"step": 2645
},
{
"epoch": 0.55,
"grad_norm": 0.8563154339790344,
"learning_rate": 4.3532923781238e-05,
"loss": 2.5139,
"step": 2650
},
{
"epoch": 0.55,
"grad_norm": 0.8651615381240845,
"learning_rate": 4.336764462574566e-05,
"loss": 2.5394,
"step": 2655
},
{
"epoch": 0.55,
"grad_norm": 0.8769192099571228,
"learning_rate": 4.320243920648376e-05,
"loss": 2.4724,
"step": 2660
},
{
"epoch": 0.55,
"grad_norm": 0.8263276219367981,
"learning_rate": 4.303730936014894e-05,
"loss": 2.5816,
"step": 2665
},
{
"epoch": 0.55,
"grad_norm": 0.8607973456382751,
"learning_rate": 4.287225692259765e-05,
"loss": 2.4292,
"step": 2670
},
{
"epoch": 0.56,
"grad_norm": 0.8444298505783081,
"learning_rate": 4.270728372882575e-05,
"loss": 2.5093,
"step": 2675
},
{
"epoch": 0.56,
"grad_norm": 0.8862450122833252,
"learning_rate": 4.254239161294804e-05,
"loss": 2.515,
"step": 2680
},
{
"epoch": 0.56,
"grad_norm": 0.8475955128669739,
"learning_rate": 4.237758240817802e-05,
"loss": 2.4614,
"step": 2685
},
{
"epoch": 0.56,
"grad_norm": 0.8355785012245178,
"learning_rate": 4.2212857946807336e-05,
"loss": 2.4804,
"step": 2690
},
{
"epoch": 0.56,
"grad_norm": 0.8851509690284729,
"learning_rate": 4.2048220060185516e-05,
"loss": 2.5098,
"step": 2695
},
{
"epoch": 0.56,
"grad_norm": 0.8373569846153259,
"learning_rate": 4.188367057869957e-05,
"loss": 2.4776,
"step": 2700
},
{
"epoch": 0.56,
"grad_norm": 0.8498626947402954,
"learning_rate": 4.171921133175365e-05,
"loss": 2.52,
"step": 2705
},
{
"epoch": 0.56,
"grad_norm": 0.8307340741157532,
"learning_rate": 4.155484414774872e-05,
"loss": 2.5165,
"step": 2710
},
{
"epoch": 0.56,
"grad_norm": 0.8285654783248901,
"learning_rate": 4.139057085406221e-05,
"loss": 2.4582,
"step": 2715
},
{
"epoch": 0.57,
"grad_norm": 0.8493524789810181,
"learning_rate": 4.1226393277027726e-05,
"loss": 2.4845,
"step": 2720
},
{
"epoch": 0.57,
"grad_norm": 0.856451153755188,
"learning_rate": 4.106231324191471e-05,
"loss": 2.5474,
"step": 2725
},
{
"epoch": 0.57,
"grad_norm": 0.8426300287246704,
"learning_rate": 4.089833257290817e-05,
"loss": 2.4962,
"step": 2730
},
{
"epoch": 0.57,
"grad_norm": 0.8221860527992249,
"learning_rate": 4.073445309308842e-05,
"loss": 2.4739,
"step": 2735
},
{
"epoch": 0.57,
"grad_norm": 0.8343910574913025,
"learning_rate": 4.0570676624410756e-05,
"loss": 2.4457,
"step": 2740
},
{
"epoch": 0.57,
"grad_norm": 0.8438074588775635,
"learning_rate": 4.040700498768525e-05,
"loss": 2.4766,
"step": 2745
},
{
"epoch": 0.57,
"grad_norm": 0.8355539441108704,
"learning_rate": 4.024344000255648e-05,
"loss": 2.4595,
"step": 2750
},
{
"epoch": 0.57,
"grad_norm": 0.8517106175422668,
"learning_rate": 4.0079983487483313e-05,
"loss": 2.4489,
"step": 2755
},
{
"epoch": 0.57,
"grad_norm": 0.8524771928787231,
"learning_rate": 3.9916637259718683e-05,
"loss": 2.5292,
"step": 2760
},
{
"epoch": 0.57,
"grad_norm": 0.8597675561904907,
"learning_rate": 3.9753403135289396e-05,
"loss": 2.4947,
"step": 2765
},
{
"epoch": 0.58,
"grad_norm": 0.8470887541770935,
"learning_rate": 3.9590282928975914e-05,
"loss": 2.5049,
"step": 2770
},
{
"epoch": 0.58,
"grad_norm": 0.856467604637146,
"learning_rate": 3.942727845429221e-05,
"loss": 2.4955,
"step": 2775
},
{
"epoch": 0.58,
"grad_norm": 0.8237487077713013,
"learning_rate": 3.926439152346558e-05,
"loss": 2.401,
"step": 2780
},
{
"epoch": 0.58,
"grad_norm": 0.833526611328125,
"learning_rate": 3.910162394741653e-05,
"loss": 2.4341,
"step": 2785
},
{
"epoch": 0.58,
"grad_norm": 0.8656613826751709,
"learning_rate": 3.893897753573861e-05,
"loss": 2.5219,
"step": 2790
},
{
"epoch": 0.58,
"grad_norm": 0.8569719791412354,
"learning_rate": 3.877645409667829e-05,
"loss": 2.5026,
"step": 2795
},
{
"epoch": 0.58,
"grad_norm": 0.8303122520446777,
"learning_rate": 3.861405543711491e-05,
"loss": 2.4507,
"step": 2800
},
{
"epoch": 0.58,
"grad_norm": 0.8535102605819702,
"learning_rate": 3.8451783362540507e-05,
"loss": 2.5029,
"step": 2805
},
{
"epoch": 0.58,
"grad_norm": 0.839522123336792,
"learning_rate": 3.828963967703983e-05,
"loss": 2.4875,
"step": 2810
},
{
"epoch": 0.59,
"grad_norm": 0.8629278540611267,
"learning_rate": 3.8127626183270223e-05,
"loss": 2.4474,
"step": 2815
},
{
"epoch": 0.59,
"grad_norm": 0.8673598766326904,
"learning_rate": 3.796574468244161e-05,
"loss": 2.4495,
"step": 2820
},
{
"epoch": 0.59,
"grad_norm": 0.8475494980812073,
"learning_rate": 3.7803996974296444e-05,
"loss": 2.4441,
"step": 2825
},
{
"epoch": 0.59,
"grad_norm": 0.8437146544456482,
"learning_rate": 3.7642384857089776e-05,
"loss": 2.5006,
"step": 2830
},
{
"epoch": 0.59,
"grad_norm": 0.876169741153717,
"learning_rate": 3.748091012756915e-05,
"loss": 2.4881,
"step": 2835
},
{
"epoch": 0.59,
"grad_norm": 0.8096486926078796,
"learning_rate": 3.731957458095467e-05,
"loss": 2.512,
"step": 2840
},
{
"epoch": 0.59,
"grad_norm": 0.8454575538635254,
"learning_rate": 3.71583800109191e-05,
"loss": 2.5171,
"step": 2845
},
{
"epoch": 0.59,
"grad_norm": 0.8605650663375854,
"learning_rate": 3.699732820956784e-05,
"loss": 2.5342,
"step": 2850
},
{
"epoch": 0.59,
"grad_norm": 0.8938848972320557,
"learning_rate": 3.6836420967419057e-05,
"loss": 2.4644,
"step": 2855
},
{
"epoch": 0.59,
"grad_norm": 0.8539745807647705,
"learning_rate": 3.6675660073383745e-05,
"loss": 2.4343,
"step": 2860
},
{
"epoch": 0.6,
"grad_norm": 0.8652215003967285,
"learning_rate": 3.6515047314745856e-05,
"loss": 2.5405,
"step": 2865
},
{
"epoch": 0.6,
"grad_norm": 0.8296924829483032,
"learning_rate": 3.6354584477142437e-05,
"loss": 2.4473,
"step": 2870
},
{
"epoch": 0.6,
"grad_norm": 0.8303621411323547,
"learning_rate": 3.6194273344543736e-05,
"loss": 2.4972,
"step": 2875
},
{
"epoch": 0.6,
"grad_norm": 0.8642436265945435,
"learning_rate": 3.6034115699233425e-05,
"loss": 2.4831,
"step": 2880
},
{
"epoch": 0.6,
"grad_norm": 0.8456935882568359,
"learning_rate": 3.5874113321788736e-05,
"loss": 2.4794,
"step": 2885
},
{
"epoch": 0.6,
"grad_norm": 0.8381088376045227,
"learning_rate": 3.571426799106071e-05,
"loss": 2.5138,
"step": 2890
},
{
"epoch": 0.6,
"grad_norm": 0.853823184967041,
"learning_rate": 3.555458148415437e-05,
"loss": 2.4311,
"step": 2895
},
{
"epoch": 0.6,
"grad_norm": 0.8326506614685059,
"learning_rate": 3.539505557640901e-05,
"loss": 2.4966,
"step": 2900
},
{
"epoch": 0.6,
"grad_norm": 0.8568333387374878,
"learning_rate": 3.523569204137843e-05,
"loss": 2.5046,
"step": 2905
},
{
"epoch": 0.6,
"grad_norm": 0.8463303446769714,
"learning_rate": 3.5076492650811246e-05,
"loss": 2.5079,
"step": 2910
},
{
"epoch": 0.61,
"grad_norm": 0.8285181522369385,
"learning_rate": 3.491745917463113e-05,
"loss": 2.4768,
"step": 2915
},
{
"epoch": 0.61,
"grad_norm": 0.8650281429290771,
"learning_rate": 3.475859338091721e-05,
"loss": 2.4288,
"step": 2920
},
{
"epoch": 0.61,
"grad_norm": 0.8329194784164429,
"learning_rate": 3.4599897035884374e-05,
"loss": 2.4516,
"step": 2925
},
{
"epoch": 0.61,
"grad_norm": 0.8466298580169678,
"learning_rate": 3.444137190386363e-05,
"loss": 2.5033,
"step": 2930
},
{
"epoch": 0.61,
"grad_norm": 0.8329981565475464,
"learning_rate": 3.4283019747282514e-05,
"loss": 2.4784,
"step": 2935
},
{
"epoch": 0.61,
"grad_norm": 0.8323549032211304,
"learning_rate": 3.412484232664545e-05,
"loss": 2.4912,
"step": 2940
},
{
"epoch": 0.61,
"grad_norm": 0.860688328742981,
"learning_rate": 3.396684140051424e-05,
"loss": 2.4394,
"step": 2945
},
{
"epoch": 0.61,
"grad_norm": 0.816442608833313,
"learning_rate": 3.3809018725488466e-05,
"loss": 2.4595,
"step": 2950
},
{
"epoch": 0.61,
"grad_norm": 0.8673380613327026,
"learning_rate": 3.365137605618598e-05,
"loss": 2.4981,
"step": 2955
},
{
"epoch": 0.62,
"grad_norm": 0.8419741988182068,
"learning_rate": 3.3493915145223395e-05,
"loss": 2.4396,
"step": 2960
},
{
"epoch": 0.62,
"grad_norm": 0.860354483127594,
"learning_rate": 3.3336637743196584e-05,
"loss": 2.4571,
"step": 2965
},
{
"epoch": 0.62,
"grad_norm": 0.8575332760810852,
"learning_rate": 3.317954559866126e-05,
"loss": 2.4907,
"step": 2970
},
{
"epoch": 0.62,
"grad_norm": 0.8085770606994629,
"learning_rate": 3.302264045811344e-05,
"loss": 2.4579,
"step": 2975
},
{
"epoch": 0.62,
"grad_norm": 0.8606233596801758,
"learning_rate": 3.286592406597021e-05,
"loss": 2.5333,
"step": 2980
},
{
"epoch": 0.62,
"grad_norm": 0.8195111155509949,
"learning_rate": 3.270939816455012e-05,
"loss": 2.4749,
"step": 2985
},
{
"epoch": 0.62,
"grad_norm": 0.8383525013923645,
"learning_rate": 3.255306449405395e-05,
"loss": 2.4935,
"step": 2990
},
{
"epoch": 0.62,
"grad_norm": 0.8518692851066589,
"learning_rate": 3.2396924792545304e-05,
"loss": 2.4706,
"step": 2995
},
{
"epoch": 0.62,
"grad_norm": 0.8205353617668152,
"learning_rate": 3.224098079593132e-05,
"loss": 2.4671,
"step": 3000
},
{
"epoch": 0.62,
"grad_norm": 0.8170686364173889,
"learning_rate": 3.2085234237943354e-05,
"loss": 2.4938,
"step": 3005
},
{
"epoch": 0.63,
"grad_norm": 0.8775966167449951,
"learning_rate": 3.19296868501177e-05,
"loss": 2.51,
"step": 3010
},
{
"epoch": 0.63,
"grad_norm": 0.8382036685943604,
"learning_rate": 3.177434036177636e-05,
"loss": 2.4934,
"step": 3015
},
{
"epoch": 0.63,
"grad_norm": 0.8418590426445007,
"learning_rate": 3.1619196500007804e-05,
"loss": 2.4608,
"step": 3020
},
{
"epoch": 0.63,
"grad_norm": 0.8652852177619934,
"learning_rate": 3.146425698964776e-05,
"loss": 2.4699,
"step": 3025
},
{
"epoch": 0.63,
"grad_norm": 0.8524035811424255,
"learning_rate": 3.1309523553260046e-05,
"loss": 2.5358,
"step": 3030
},
{
"epoch": 0.63,
"grad_norm": 0.8546153903007507,
"learning_rate": 3.115499791111743e-05,
"loss": 2.5292,
"step": 3035
},
{
"epoch": 0.63,
"grad_norm": 0.8227338194847107,
"learning_rate": 3.10006817811825e-05,
"loss": 2.4798,
"step": 3040
},
{
"epoch": 0.63,
"grad_norm": 0.8461725115776062,
"learning_rate": 3.084657687908855e-05,
"loss": 2.5442,
"step": 3045
},
{
"epoch": 0.63,
"grad_norm": 0.8724225163459778,
"learning_rate": 3.069268491812052e-05,
"loss": 2.52,
"step": 3050
},
{
"epoch": 0.63,
"grad_norm": 0.8465836644172668,
"learning_rate": 3.0539007609195934e-05,
"loss": 2.5311,
"step": 3055
},
{
"epoch": 0.64,
"grad_norm": 0.8338139057159424,
"learning_rate": 3.0385546660845908e-05,
"loss": 2.5137,
"step": 3060
},
{
"epoch": 0.64,
"grad_norm": 0.8537865877151489,
"learning_rate": 3.0232303779196132e-05,
"loss": 2.4483,
"step": 3065
},
{
"epoch": 0.64,
"grad_norm": 0.8705686926841736,
"learning_rate": 3.0079280667947885e-05,
"loss": 2.4951,
"step": 3070
},
{
"epoch": 0.64,
"grad_norm": 0.8283040523529053,
"learning_rate": 2.9926479028359132e-05,
"loss": 2.4763,
"step": 3075
},
{
"epoch": 0.64,
"grad_norm": 0.8348056674003601,
"learning_rate": 2.97739005592256e-05,
"loss": 2.4961,
"step": 3080
},
{
"epoch": 0.64,
"grad_norm": 0.8377600312232971,
"learning_rate": 2.962154695686187e-05,
"loss": 2.4741,
"step": 3085
},
{
"epoch": 0.64,
"grad_norm": 0.8565430641174316,
"learning_rate": 2.9469419915082536e-05,
"loss": 2.5142,
"step": 3090
},
{
"epoch": 0.64,
"grad_norm": 0.8483173251152039,
"learning_rate": 2.9317521125183368e-05,
"loss": 2.4468,
"step": 3095
},
{
"epoch": 0.64,
"grad_norm": 0.8331131935119629,
"learning_rate": 2.9165852275922524e-05,
"loss": 2.4343,
"step": 3100
},
{
"epoch": 0.65,
"grad_norm": 0.854733943939209,
"learning_rate": 2.901441505350174e-05,
"loss": 2.4732,
"step": 3105
},
{
"epoch": 0.65,
"grad_norm": 0.8639799952507019,
"learning_rate": 2.886321114154762e-05,
"loss": 2.5379,
"step": 3110
},
{
"epoch": 0.65,
"grad_norm": 0.8178684115409851,
"learning_rate": 2.87122422210929e-05,
"loss": 2.4946,
"step": 3115
},
{
"epoch": 0.65,
"grad_norm": 0.8145985007286072,
"learning_rate": 2.8561509970557736e-05,
"loss": 2.4871,
"step": 3120
},
{
"epoch": 0.65,
"grad_norm": 0.8069058656692505,
"learning_rate": 2.8411016065731146e-05,
"loss": 2.4385,
"step": 3125
},
{
"epoch": 0.65,
"grad_norm": 0.8580114245414734,
"learning_rate": 2.826076217975222e-05,
"loss": 2.4858,
"step": 3130
},
{
"epoch": 0.65,
"grad_norm": 0.8153144717216492,
"learning_rate": 2.8110749983091632e-05,
"loss": 2.4465,
"step": 3135
},
{
"epoch": 0.65,
"grad_norm": 0.8372932076454163,
"learning_rate": 2.7960981143533053e-05,
"loss": 2.4108,
"step": 3140
},
{
"epoch": 0.65,
"grad_norm": 0.8401663303375244,
"learning_rate": 2.781145732615457e-05,
"loss": 2.4682,
"step": 3145
},
{
"epoch": 0.65,
"grad_norm": 0.8417117595672607,
"learning_rate": 2.7662180193310218e-05,
"loss": 2.4767,
"step": 3150
},
{
"epoch": 0.66,
"grad_norm": 0.997302234172821,
"learning_rate": 2.751315140461145e-05,
"loss": 2.5469,
"step": 3155
},
{
"epoch": 0.66,
"grad_norm": 0.8078846335411072,
"learning_rate": 2.7364372616908744e-05,
"loss": 2.466,
"step": 3160
},
{
"epoch": 0.66,
"grad_norm": 0.8174634575843811,
"learning_rate": 2.7215845484273152e-05,
"loss": 2.4345,
"step": 3165
},
{
"epoch": 0.66,
"grad_norm": 0.8275050520896912,
"learning_rate": 2.7067571657977893e-05,
"loss": 2.4417,
"step": 3170
},
{
"epoch": 0.66,
"grad_norm": 0.8436856865882874,
"learning_rate": 2.691955278648003e-05,
"loss": 2.4209,
"step": 3175
},
{
"epoch": 0.66,
"grad_norm": 0.8604226112365723,
"learning_rate": 2.6771790515402112e-05,
"loss": 2.4584,
"step": 3180
},
{
"epoch": 0.66,
"grad_norm": 0.8629764914512634,
"learning_rate": 2.6624286487513916e-05,
"loss": 2.4699,
"step": 3185
},
{
"epoch": 0.66,
"grad_norm": 0.8471906781196594,
"learning_rate": 2.6477042342714137e-05,
"loss": 2.4888,
"step": 3190
},
{
"epoch": 0.66,
"grad_norm": 0.8521903157234192,
"learning_rate": 2.633005971801219e-05,
"loss": 2.5484,
"step": 3195
},
{
"epoch": 0.67,
"grad_norm": 0.8371425867080688,
"learning_rate": 2.6183340247510013e-05,
"loss": 2.5013,
"step": 3200
},
{
"epoch": 0.67,
"grad_norm": 0.837877094745636,
"learning_rate": 2.6036885562383856e-05,
"loss": 2.4266,
"step": 3205
},
{
"epoch": 0.67,
"grad_norm": 0.8451969623565674,
"learning_rate": 2.5890697290866206e-05,
"loss": 2.4487,
"step": 3210
},
{
"epoch": 0.67,
"grad_norm": 0.8365843296051025,
"learning_rate": 2.5744777058227642e-05,
"loss": 2.4694,
"step": 3215
},
{
"epoch": 0.67,
"grad_norm": 0.85359787940979,
"learning_rate": 2.5599126486758777e-05,
"loss": 2.4277,
"step": 3220
},
{
"epoch": 0.67,
"grad_norm": 0.8310922980308533,
"learning_rate": 2.5453747195752243e-05,
"loss": 2.3899,
"step": 3225
},
{
"epoch": 0.67,
"grad_norm": 0.837394118309021,
"learning_rate": 2.530864080148464e-05,
"loss": 2.5018,
"step": 3230
},
{
"epoch": 0.67,
"grad_norm": 0.8306896686553955,
"learning_rate": 2.5163808917198615e-05,
"loss": 2.4753,
"step": 3235
},
{
"epoch": 0.67,
"grad_norm": 0.8381210565567017,
"learning_rate": 2.501925315308492e-05,
"loss": 2.5005,
"step": 3240
},
{
"epoch": 0.67,
"grad_norm": 0.8402901887893677,
"learning_rate": 2.4874975116264477e-05,
"loss": 2.4355,
"step": 3245
},
{
"epoch": 0.68,
"grad_norm": 0.8797227740287781,
"learning_rate": 2.4730976410770534e-05,
"loss": 2.496,
"step": 3250
},
{
"epoch": 0.68,
"grad_norm": 0.8176755309104919,
"learning_rate": 2.458725863753084e-05,
"loss": 2.4882,
"step": 3255
},
{
"epoch": 0.68,
"grad_norm": 0.8236318826675415,
"learning_rate": 2.4443823394349834e-05,
"loss": 2.4739,
"step": 3260
},
{
"epoch": 0.68,
"grad_norm": 0.8401728868484497,
"learning_rate": 2.430067227589088e-05,
"loss": 2.3903,
"step": 3265
},
{
"epoch": 0.68,
"grad_norm": 0.8464370369911194,
"learning_rate": 2.4157806873658517e-05,
"loss": 2.5006,
"step": 3270
},
{
"epoch": 0.68,
"grad_norm": 0.8629506826400757,
"learning_rate": 2.401522877598087e-05,
"loss": 2.4274,
"step": 3275
},
{
"epoch": 0.68,
"grad_norm": 0.8365341424942017,
"learning_rate": 2.3872939567991827e-05,
"loss": 2.454,
"step": 3280
},
{
"epoch": 0.68,
"grad_norm": 0.8436932563781738,
"learning_rate": 2.373094083161353e-05,
"loss": 2.4396,
"step": 3285
},
{
"epoch": 0.68,
"grad_norm": 0.8262729644775391,
"learning_rate": 2.358923414553877e-05,
"loss": 2.4379,
"step": 3290
},
{
"epoch": 0.68,
"grad_norm": 0.8110555410385132,
"learning_rate": 2.3447821085213405e-05,
"loss": 2.3887,
"step": 3295
},
{
"epoch": 0.69,
"grad_norm": 0.8610245585441589,
"learning_rate": 2.3306703222818878e-05,
"loss": 2.497,
"step": 3300
},
{
"epoch": 0.69,
"grad_norm": 0.8528003692626953,
"learning_rate": 2.3165882127254705e-05,
"loss": 2.4406,
"step": 3305
},
{
"epoch": 0.69,
"grad_norm": 0.8422110080718994,
"learning_rate": 2.302535936412108e-05,
"loss": 2.531,
"step": 3310
},
{
"epoch": 0.69,
"grad_norm": 0.83036208152771,
"learning_rate": 2.2885136495701415e-05,
"loss": 2.4128,
"step": 3315
},
{
"epoch": 0.69,
"grad_norm": 0.8456906080245972,
"learning_rate": 2.274521508094501e-05,
"loss": 2.4396,
"step": 3320
},
{
"epoch": 0.69,
"grad_norm": 0.8553588390350342,
"learning_rate": 2.2605596675449698e-05,
"loss": 2.4208,
"step": 3325
},
{
"epoch": 0.69,
"grad_norm": 0.848984956741333,
"learning_rate": 2.246628283144457e-05,
"loss": 2.4647,
"step": 3330
},
{
"epoch": 0.69,
"grad_norm": 0.8481738567352295,
"learning_rate": 2.232727509777269e-05,
"loss": 2.4427,
"step": 3335
},
{
"epoch": 0.69,
"grad_norm": 0.8750231862068176,
"learning_rate": 2.2188575019873932e-05,
"loss": 2.5128,
"step": 3340
},
{
"epoch": 0.7,
"grad_norm": 1.141512155532837,
"learning_rate": 2.2050184139767704e-05,
"loss": 2.429,
"step": 3345
},
{
"epoch": 0.7,
"grad_norm": 0.9442705512046814,
"learning_rate": 2.191210399603591e-05,
"loss": 2.4569,
"step": 3350
},
{
"epoch": 0.7,
"grad_norm": 0.832524836063385,
"learning_rate": 2.1774336123805772e-05,
"loss": 2.4379,
"step": 3355
},
{
"epoch": 0.7,
"grad_norm": 0.879459023475647,
"learning_rate": 2.1636882054732776e-05,
"loss": 2.4453,
"step": 3360
},
{
"epoch": 0.7,
"grad_norm": 0.8389520049095154,
"learning_rate": 2.1499743316983684e-05,
"loss": 2.4836,
"step": 3365
},
{
"epoch": 0.7,
"grad_norm": 0.8538713455200195,
"learning_rate": 2.1362921435219473e-05,
"loss": 2.4503,
"step": 3370
},
{
"epoch": 0.7,
"grad_norm": 0.845042884349823,
"learning_rate": 2.1226417930578464e-05,
"loss": 2.4747,
"step": 3375
},
{
"epoch": 0.7,
"grad_norm": 0.8627697825431824,
"learning_rate": 2.109023432065935e-05,
"loss": 2.4204,
"step": 3380
},
{
"epoch": 0.7,
"grad_norm": 0.8196375370025635,
"learning_rate": 2.095437211950434e-05,
"loss": 2.45,
"step": 3385
},
{
"epoch": 0.7,
"grad_norm": 0.8441497683525085,
"learning_rate": 2.0818832837582352e-05,
"loss": 2.4576,
"step": 3390
},
{
"epoch": 0.71,
"grad_norm": 0.824752151966095,
"learning_rate": 2.068361798177218e-05,
"loss": 2.3914,
"step": 3395
},
{
"epoch": 0.71,
"grad_norm": 0.8222182393074036,
"learning_rate": 2.0548729055345778e-05,
"loss": 2.4152,
"step": 3400
},
{
"epoch": 0.71,
"grad_norm": 0.8592625260353088,
"learning_rate": 2.0414167557951514e-05,
"loss": 2.4308,
"step": 3405
},
{
"epoch": 0.71,
"grad_norm": 0.8408738970756531,
"learning_rate": 2.0279934985597527e-05,
"loss": 2.3937,
"step": 3410
},
{
"epoch": 0.71,
"grad_norm": 0.8241609334945679,
"learning_rate": 2.0146032830635054e-05,
"loss": 2.4739,
"step": 3415
},
{
"epoch": 0.71,
"grad_norm": 0.8782637119293213,
"learning_rate": 2.001246258174192e-05,
"loss": 2.4966,
"step": 3420
},
{
"epoch": 0.71,
"grad_norm": 0.8682647943496704,
"learning_rate": 1.9879225723905886e-05,
"loss": 2.3827,
"step": 3425
},
{
"epoch": 0.71,
"grad_norm": 0.8345317244529724,
"learning_rate": 1.9746323738408203e-05,
"loss": 2.5048,
"step": 3430
},
{
"epoch": 0.71,
"grad_norm": 0.8492810130119324,
"learning_rate": 1.9613758102807117e-05,
"loss": 2.4702,
"step": 3435
},
{
"epoch": 0.71,
"grad_norm": 0.842808187007904,
"learning_rate": 1.9481530290921474e-05,
"loss": 2.4513,
"step": 3440
},
{
"epoch": 0.72,
"grad_norm": 0.8469889760017395,
"learning_rate": 1.934964177281428e-05,
"loss": 2.4091,
"step": 3445
},
{
"epoch": 0.72,
"grad_norm": 0.8806920647621155,
"learning_rate": 1.9218094014776434e-05,
"loss": 2.4841,
"step": 3450
},
{
"epoch": 0.72,
"grad_norm": 0.8436346054077148,
"learning_rate": 1.9086888479310333e-05,
"loss": 2.483,
"step": 3455
},
{
"epoch": 0.72,
"grad_norm": 0.8514381051063538,
"learning_rate": 1.895602662511371e-05,
"loss": 2.4133,
"step": 3460
},
{
"epoch": 0.72,
"grad_norm": 0.8585630655288696,
"learning_rate": 1.8825509907063327e-05,
"loss": 2.5086,
"step": 3465
},
{
"epoch": 0.72,
"grad_norm": 0.8401122689247131,
"learning_rate": 1.8695339776198872e-05,
"loss": 2.4549,
"step": 3470
},
{
"epoch": 0.72,
"grad_norm": 0.8682958483695984,
"learning_rate": 1.8565517679706783e-05,
"loss": 2.4149,
"step": 3475
},
{
"epoch": 0.72,
"grad_norm": 0.8392828106880188,
"learning_rate": 1.8436045060904174e-05,
"loss": 2.418,
"step": 3480
},
{
"epoch": 0.72,
"grad_norm": 0.8359572291374207,
"learning_rate": 1.830692335922279e-05,
"loss": 2.4089,
"step": 3485
},
{
"epoch": 0.73,
"grad_norm": 0.8385863900184631,
"learning_rate": 1.8178154010192994e-05,
"loss": 2.5022,
"step": 3490
},
{
"epoch": 0.73,
"grad_norm": 0.8350293636322021,
"learning_rate": 1.8049738445427822e-05,
"loss": 2.4515,
"step": 3495
},
{
"epoch": 0.73,
"grad_norm": 0.8670375347137451,
"learning_rate": 1.7921678092607052e-05,
"loss": 2.4437,
"step": 3500
},
{
"epoch": 0.73,
"grad_norm": 0.8225032687187195,
"learning_rate": 1.7793974375461352e-05,
"loss": 2.3894,
"step": 3505
},
{
"epoch": 0.73,
"grad_norm": 0.8368529677391052,
"learning_rate": 1.7666628713756417e-05,
"loss": 2.3898,
"step": 3510
},
{
"epoch": 0.73,
"grad_norm": 0.8342825174331665,
"learning_rate": 1.7539642523277228e-05,
"loss": 2.4032,
"step": 3515
},
{
"epoch": 0.73,
"grad_norm": 0.838580846786499,
"learning_rate": 1.7413017215812273e-05,
"loss": 2.4454,
"step": 3520
},
{
"epoch": 0.73,
"grad_norm": 0.8412612080574036,
"learning_rate": 1.728675419913788e-05,
"loss": 2.4327,
"step": 3525
},
{
"epoch": 0.73,
"grad_norm": 0.8314975500106812,
"learning_rate": 1.716085487700253e-05,
"loss": 2.4657,
"step": 3530
},
{
"epoch": 0.73,
"grad_norm": 0.8533310890197754,
"learning_rate": 1.703532064911131e-05,
"loss": 2.3868,
"step": 3535
},
{
"epoch": 0.74,
"grad_norm": 0.8259750604629517,
"learning_rate": 1.6910152911110283e-05,
"loss": 2.4788,
"step": 3540
},
{
"epoch": 0.74,
"grad_norm": 0.8331064581871033,
"learning_rate": 1.6785353054571024e-05,
"loss": 2.4237,
"step": 3545
},
{
"epoch": 0.74,
"grad_norm": 0.8288483023643494,
"learning_rate": 1.666092246697512e-05,
"loss": 2.4061,
"step": 3550
},
{
"epoch": 0.74,
"grad_norm": 0.8371803164482117,
"learning_rate": 1.6536862531698766e-05,
"loss": 2.4104,
"step": 3555
},
{
"epoch": 0.74,
"grad_norm": 0.859487771987915,
"learning_rate": 1.6413174627997328e-05,
"loss": 2.4433,
"step": 3560
},
{
"epoch": 0.74,
"grad_norm": 0.8278236389160156,
"learning_rate": 1.6289860130990147e-05,
"loss": 2.3789,
"step": 3565
},
{
"epoch": 0.74,
"grad_norm": 0.8158142566680908,
"learning_rate": 1.6166920411645064e-05,
"loss": 2.4658,
"step": 3570
},
{
"epoch": 0.74,
"grad_norm": 0.8452654480934143,
"learning_rate": 1.6044356836763315e-05,
"loss": 2.4245,
"step": 3575
},
{
"epoch": 0.74,
"grad_norm": 0.8108316659927368,
"learning_rate": 1.5922170768964285e-05,
"loss": 2.3574,
"step": 3580
},
{
"epoch": 0.75,
"grad_norm": 0.8247130513191223,
"learning_rate": 1.5800363566670362e-05,
"loss": 2.427,
"step": 3585
},
{
"epoch": 0.75,
"grad_norm": 0.8374765515327454,
"learning_rate": 1.5678936584091852e-05,
"loss": 2.441,
"step": 3590
},
{
"epoch": 0.75,
"grad_norm": 0.8417315483093262,
"learning_rate": 1.5557891171211892e-05,
"loss": 2.4637,
"step": 3595
},
{
"epoch": 0.75,
"grad_norm": 0.8394094109535217,
"learning_rate": 1.5437228673771465e-05,
"loss": 2.3891,
"step": 3600
},
{
"epoch": 0.75,
"grad_norm": 0.8402270674705505,
"learning_rate": 1.5316950433254445e-05,
"loss": 2.4188,
"step": 3605
},
{
"epoch": 0.75,
"grad_norm": 0.8152482509613037,
"learning_rate": 1.5197057786872649e-05,
"loss": 2.3605,
"step": 3610
},
{
"epoch": 0.75,
"grad_norm": 0.8334120512008667,
"learning_rate": 1.5077552067551015e-05,
"loss": 2.4525,
"step": 3615
},
{
"epoch": 0.75,
"grad_norm": 0.8562144041061401,
"learning_rate": 1.4958434603912747e-05,
"loss": 2.4727,
"step": 3620
},
{
"epoch": 0.75,
"grad_norm": 0.8257231116294861,
"learning_rate": 1.4839706720264546e-05,
"loss": 2.4352,
"step": 3625
},
{
"epoch": 0.75,
"grad_norm": 0.8590261936187744,
"learning_rate": 1.4721369736581924e-05,
"loss": 2.3921,
"step": 3630
},
{
"epoch": 0.76,
"grad_norm": 0.883569061756134,
"learning_rate": 1.4603424968494484e-05,
"loss": 2.3616,
"step": 3635
},
{
"epoch": 0.76,
"grad_norm": 0.8439374566078186,
"learning_rate": 1.448587372727132e-05,
"loss": 2.4843,
"step": 3640
},
{
"epoch": 0.76,
"grad_norm": 0.8474141359329224,
"learning_rate": 1.4368717319806419e-05,
"loss": 2.4002,
"step": 3645
},
{
"epoch": 0.76,
"grad_norm": 0.8499286770820618,
"learning_rate": 1.4251957048604152e-05,
"loss": 2.3967,
"step": 3650
},
{
"epoch": 0.76,
"grad_norm": 0.8370791077613831,
"learning_rate": 1.413559421176479e-05,
"loss": 2.3787,
"step": 3655
},
{
"epoch": 0.76,
"grad_norm": 0.8443555235862732,
"learning_rate": 1.4019630102970056e-05,
"loss": 2.4941,
"step": 3660
},
{
"epoch": 0.76,
"grad_norm": 0.8790743947029114,
"learning_rate": 1.3904066011468753e-05,
"loss": 2.4577,
"step": 3665
},
{
"epoch": 0.76,
"grad_norm": 0.8634971976280212,
"learning_rate": 1.3788903222062433e-05,
"loss": 2.4719,
"step": 3670
},
{
"epoch": 0.76,
"grad_norm": 0.8475700616836548,
"learning_rate": 1.3674143015091118e-05,
"loss": 2.4178,
"step": 3675
},
{
"epoch": 0.76,
"grad_norm": 0.8289123177528381,
"learning_rate": 1.355978666641905e-05,
"loss": 2.3726,
"step": 3680
},
{
"epoch": 0.77,
"grad_norm": 0.8187718987464905,
"learning_rate": 1.3445835447420507e-05,
"loss": 2.3948,
"step": 3685
},
{
"epoch": 0.77,
"grad_norm": 0.8293814659118652,
"learning_rate": 1.3332290624965688e-05,
"loss": 2.4331,
"step": 3690
},
{
"epoch": 0.77,
"grad_norm": 0.8501749634742737,
"learning_rate": 1.3219153461406609e-05,
"loss": 2.4301,
"step": 3695
},
{
"epoch": 0.77,
"grad_norm": 0.8464502096176147,
"learning_rate": 1.3106425214563078e-05,
"loss": 2.4247,
"step": 3700
},
{
"epoch": 0.77,
"grad_norm": 0.8278773427009583,
"learning_rate": 1.2994107137708716e-05,
"loss": 2.4813,
"step": 3705
},
{
"epoch": 0.77,
"grad_norm": 0.8624427318572998,
"learning_rate": 1.2882200479556988e-05,
"loss": 2.4825,
"step": 3710
},
{
"epoch": 0.77,
"grad_norm": 0.8700522184371948,
"learning_rate": 1.2770706484247397e-05,
"loss": 2.3882,
"step": 3715
},
{
"epoch": 0.77,
"grad_norm": 0.8325879573822021,
"learning_rate": 1.2659626391331564e-05,
"loss": 2.3853,
"step": 3720
},
{
"epoch": 0.77,
"grad_norm": 0.8339242935180664,
"learning_rate": 1.2548961435759493e-05,
"loss": 2.4383,
"step": 3725
},
{
"epoch": 0.78,
"grad_norm": 0.8548488616943359,
"learning_rate": 1.2438712847865846e-05,
"loss": 2.482,
"step": 3730
},
{
"epoch": 0.78,
"grad_norm": 0.8385611176490784,
"learning_rate": 1.2328881853356244e-05,
"loss": 2.3893,
"step": 3735
},
{
"epoch": 0.78,
"grad_norm": 0.8631219863891602,
"learning_rate": 1.221946967329365e-05,
"loss": 2.3657,
"step": 3740
},
{
"epoch": 0.78,
"grad_norm": 0.8660287261009216,
"learning_rate": 1.2110477524084796e-05,
"loss": 2.4005,
"step": 3745
},
{
"epoch": 0.78,
"grad_norm": 0.8280443549156189,
"learning_rate": 1.2001906617466657e-05,
"loss": 2.4671,
"step": 3750
},
{
"epoch": 0.78,
"grad_norm": 0.8248807787895203,
"learning_rate": 1.1893758160492978e-05,
"loss": 2.4249,
"step": 3755
},
{
"epoch": 0.78,
"grad_norm": 0.8332074284553528,
"learning_rate": 1.1786033355520859e-05,
"loss": 2.4044,
"step": 3760
},
{
"epoch": 0.78,
"grad_norm": 0.8442416787147522,
"learning_rate": 1.1678733400197373e-05,
"loss": 2.3946,
"step": 3765
},
{
"epoch": 0.78,
"grad_norm": 0.8986250758171082,
"learning_rate": 1.1571859487446263e-05,
"loss": 2.3938,
"step": 3770
},
{
"epoch": 0.78,
"grad_norm": 0.8431919813156128,
"learning_rate": 1.1465412805454695e-05,
"loss": 2.4325,
"step": 3775
},
{
"epoch": 0.79,
"grad_norm": 0.8419010043144226,
"learning_rate": 1.1359394537660011e-05,
"loss": 2.3799,
"step": 3780
},
{
"epoch": 0.79,
"grad_norm": 0.8507583141326904,
"learning_rate": 1.125380586273661e-05,
"loss": 2.4064,
"step": 3785
},
{
"epoch": 0.79,
"grad_norm": 0.8361155390739441,
"learning_rate": 1.1148647954582808e-05,
"loss": 2.4073,
"step": 3790
},
{
"epoch": 0.79,
"grad_norm": 0.8001879453659058,
"learning_rate": 1.1043921982307819e-05,
"loss": 2.4172,
"step": 3795
},
{
"epoch": 0.79,
"grad_norm": 0.8469948172569275,
"learning_rate": 1.0939629110218735e-05,
"loss": 2.4288,
"step": 3800
},
{
"epoch": 0.79,
"grad_norm": 0.825676679611206,
"learning_rate": 1.0835770497807596e-05,
"loss": 2.3832,
"step": 3805
},
{
"epoch": 0.79,
"grad_norm": 0.8198089599609375,
"learning_rate": 1.0732347299738493e-05,
"loss": 2.3916,
"step": 3810
},
{
"epoch": 0.79,
"grad_norm": 0.8248541951179504,
"learning_rate": 1.0629360665834732e-05,
"loss": 2.4436,
"step": 3815
},
{
"epoch": 0.79,
"grad_norm": 0.8490849137306213,
"learning_rate": 1.052681174106604e-05,
"loss": 2.4666,
"step": 3820
},
{
"epoch": 0.8,
"grad_norm": 0.8369426131248474,
"learning_rate": 1.0424701665535852e-05,
"loss": 2.4508,
"step": 3825
},
{
"epoch": 0.8,
"grad_norm": 0.8336918950080872,
"learning_rate": 1.0323031574468638e-05,
"loss": 2.4203,
"step": 3830
},
{
"epoch": 0.8,
"grad_norm": 0.8324209451675415,
"learning_rate": 1.0221802598197261e-05,
"loss": 2.4469,
"step": 3835
},
{
"epoch": 0.8,
"grad_norm": 0.8504571318626404,
"learning_rate": 1.0121015862150423e-05,
"loss": 2.4261,
"step": 3840
},
{
"epoch": 0.8,
"grad_norm": 0.8121498823165894,
"learning_rate": 1.0020672486840154e-05,
"loss": 2.4624,
"step": 3845
},
{
"epoch": 0.8,
"grad_norm": 0.8450626730918884,
"learning_rate": 9.920773587849364e-06,
"loss": 2.4759,
"step": 3850
},
{
"epoch": 0.8,
"grad_norm": 0.8238378167152405,
"learning_rate": 9.821320275819401e-06,
"loss": 2.4382,
"step": 3855
},
{
"epoch": 0.8,
"grad_norm": 0.8515500426292419,
"learning_rate": 9.72231365643777e-06,
"loss": 2.3913,
"step": 3860
},
{
"epoch": 0.8,
"grad_norm": 0.8674489855766296,
"learning_rate": 9.623754830425779e-06,
"loss": 2.4097,
"step": 3865
},
{
"epoch": 0.8,
"grad_norm": 0.8425604701042175,
"learning_rate": 9.52564489352632e-06,
"loss": 2.4279,
"step": 3870
},
{
"epoch": 0.81,
"grad_norm": 0.8549066185951233,
"learning_rate": 9.427984936491702e-06,
"loss": 2.387,
"step": 3875
},
{
"epoch": 0.81,
"grad_norm": 0.8391668796539307,
"learning_rate": 9.330776045071509e-06,
"loss": 2.3812,
"step": 3880
},
{
"epoch": 0.81,
"grad_norm": 0.8398712277412415,
"learning_rate": 9.23401930000054e-06,
"loss": 2.4282,
"step": 3885
},
{
"epoch": 0.81,
"grad_norm": 0.8507668972015381,
"learning_rate": 9.137715776986772e-06,
"loss": 2.3517,
"step": 3890
},
{
"epoch": 0.81,
"grad_norm": 0.8810450434684753,
"learning_rate": 9.041866546699434e-06,
"loss": 2.4328,
"step": 3895
},
{
"epoch": 0.81,
"grad_norm": 0.838951587677002,
"learning_rate": 8.946472674757078e-06,
"loss": 2.3611,
"step": 3900
},
{
"epoch": 0.81,
"grad_norm": 0.8396971225738525,
"learning_rate": 8.851535221715735e-06,
"loss": 2.4292,
"step": 3905
},
{
"epoch": 0.81,
"grad_norm": 0.8501064777374268,
"learning_rate": 8.757055243057132e-06,
"loss": 2.4011,
"step": 3910
},
{
"epoch": 0.81,
"grad_norm": 0.8223648071289062,
"learning_rate": 8.663033789176967e-06,
"loss": 2.4684,
"step": 3915
},
{
"epoch": 0.81,
"grad_norm": 0.8593500852584839,
"learning_rate": 8.5694719053732e-06,
"loss": 2.403,
"step": 3920
},
{
"epoch": 0.82,
"grad_norm": 0.8248060941696167,
"learning_rate": 8.476370631834458e-06,
"loss": 2.4386,
"step": 3925
},
{
"epoch": 0.82,
"grad_norm": 0.8380996584892273,
"learning_rate": 8.383731003628452e-06,
"loss": 2.4916,
"step": 3930
},
{
"epoch": 0.82,
"grad_norm": 0.8650656938552856,
"learning_rate": 8.291554050690508e-06,
"loss": 2.4159,
"step": 3935
},
{
"epoch": 0.82,
"grad_norm": 0.8511679768562317,
"learning_rate": 8.199840797812058e-06,
"loss": 2.3887,
"step": 3940
},
{
"epoch": 0.82,
"grad_norm": 0.8332497477531433,
"learning_rate": 8.108592264629295e-06,
"loss": 2.3583,
"step": 3945
},
{
"epoch": 0.82,
"grad_norm": 0.8381503820419312,
"learning_rate": 8.017809465611803e-06,
"loss": 2.4219,
"step": 3950
},
{
"epoch": 0.82,
"grad_norm": 0.8456980586051941,
"learning_rate": 7.927493410051324e-06,
"loss": 2.3751,
"step": 3955
},
{
"epoch": 0.82,
"grad_norm": 0.8301672339439392,
"learning_rate": 7.837645102050473e-06,
"loss": 2.4167,
"step": 3960
},
{
"epoch": 0.82,
"grad_norm": 0.8242828845977783,
"learning_rate": 7.748265540511635e-06,
"loss": 2.3247,
"step": 3965
},
{
"epoch": 0.83,
"grad_norm": 0.8471345901489258,
"learning_rate": 7.65935571912582e-06,
"loss": 2.3874,
"step": 3970
},
{
"epoch": 0.83,
"grad_norm": 0.8570038080215454,
"learning_rate": 7.5709166263616405e-06,
"loss": 2.4379,
"step": 3975
},
{
"epoch": 0.83,
"grad_norm": 0.843518853187561,
"learning_rate": 7.482949245454302e-06,
"loss": 2.3743,
"step": 3980
},
{
"epoch": 0.83,
"grad_norm": 0.8229020237922668,
"learning_rate": 7.3954545543946876e-06,
"loss": 2.4066,
"step": 3985
},
{
"epoch": 0.83,
"grad_norm": 0.8295770287513733,
"learning_rate": 7.308433525918468e-06,
"loss": 2.4312,
"step": 3990
},
{
"epoch": 0.83,
"grad_norm": 0.8365007638931274,
"learning_rate": 7.221887127495313e-06,
"loss": 2.4288,
"step": 3995
},
{
"epoch": 0.83,
"grad_norm": 0.8317162394523621,
"learning_rate": 7.1358163213181114e-06,
"loss": 2.4077,
"step": 4000
}
],
"logging_steps": 5,
"max_steps": 4811,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"total_flos": 1.1658574908358656e+19,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}