u5 / trainer_state.json
Kaipengm2's picture
Upload folder using huggingface_hub
692270a verified
Raw
History Blame Contribute Delete
88.5 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2359,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00211954217888936,
"grad_norm": 1.8612931966781616,
"learning_rate": 4.507042253521127e-06,
"loss": 2.944260025024414,
"step": 5
},
{
"epoch": 0.00423908435777872,
"grad_norm": 1.1160953044891357,
"learning_rate": 1.0140845070422535e-05,
"loss": 2.837664031982422,
"step": 10
},
{
"epoch": 0.00635862653666808,
"grad_norm": 0.7302947640419006,
"learning_rate": 1.5774647887323945e-05,
"loss": 2.802225685119629,
"step": 15
},
{
"epoch": 0.00847816871555744,
"grad_norm": 0.40593045949935913,
"learning_rate": 2.140845070422535e-05,
"loss": 2.6366830825805665,
"step": 20
},
{
"epoch": 0.0105977108944468,
"grad_norm": 0.31648972630500793,
"learning_rate": 2.7042253521126763e-05,
"loss": 2.4947330474853517,
"step": 25
},
{
"epoch": 0.01271725307333616,
"grad_norm": 0.3702400326728821,
"learning_rate": 3.267605633802817e-05,
"loss": 2.512411117553711,
"step": 30
},
{
"epoch": 0.01483679525222552,
"grad_norm": 0.27157336473464966,
"learning_rate": 3.8309859154929575e-05,
"loss": 2.375998687744141,
"step": 35
},
{
"epoch": 0.01695633743111488,
"grad_norm": 0.21749232709407806,
"learning_rate": 4.394366197183099e-05,
"loss": 2.2622447967529298,
"step": 40
},
{
"epoch": 0.01907587961000424,
"grad_norm": 0.23159104585647583,
"learning_rate": 4.9577464788732394e-05,
"loss": 2.2170639038085938,
"step": 45
},
{
"epoch": 0.0211954217888936,
"grad_norm": 0.19265541434288025,
"learning_rate": 5.52112676056338e-05,
"loss": 2.174176025390625,
"step": 50
},
{
"epoch": 0.02331496396778296,
"grad_norm": 0.2098459154367447,
"learning_rate": 6.084507042253521e-05,
"loss": 2.1516395568847657,
"step": 55
},
{
"epoch": 0.02543450614667232,
"grad_norm": 0.22788824141025543,
"learning_rate": 6.647887323943662e-05,
"loss": 2.0522294998168946,
"step": 60
},
{
"epoch": 0.02755404832556168,
"grad_norm": 0.2034478485584259,
"learning_rate": 7.211267605633803e-05,
"loss": 2.0236040115356446,
"step": 65
},
{
"epoch": 0.02967359050445104,
"grad_norm": 0.27458181977272034,
"learning_rate": 7.774647887323944e-05,
"loss": 1.910360336303711,
"step": 70
},
{
"epoch": 0.0317931326833404,
"grad_norm": 0.19724486768245697,
"learning_rate": 7.99996606408244e-05,
"loss": 1.9117116928100586,
"step": 75
},
{
"epoch": 0.03391267486222976,
"grad_norm": 0.23376719653606415,
"learning_rate": 7.999758680004851e-05,
"loss": 1.8922744750976563,
"step": 80
},
{
"epoch": 0.03603221704111912,
"grad_norm": 0.21566373109817505,
"learning_rate": 7.999362774900014e-05,
"loss": 1.893143081665039,
"step": 85
},
{
"epoch": 0.03815175922000848,
"grad_norm": 0.22187377512454987,
"learning_rate": 7.998778367428163e-05,
"loss": 1.8458093643188476,
"step": 90
},
{
"epoch": 0.04027130139889784,
"grad_norm": 0.2085333913564682,
"learning_rate": 7.998005485134234e-05,
"loss": 1.7947025299072266,
"step": 95
},
{
"epoch": 0.0423908435777872,
"grad_norm": 0.19334450364112854,
"learning_rate": 7.997044164446566e-05,
"loss": 1.8084171295166016,
"step": 100
},
{
"epoch": 0.04451038575667656,
"grad_norm": 0.22427749633789062,
"learning_rate": 7.995894450675184e-05,
"loss": 1.763449478149414,
"step": 105
},
{
"epoch": 0.04662992793556592,
"grad_norm": 0.21727879345417023,
"learning_rate": 7.994556398009663e-05,
"loss": 1.7564985275268554,
"step": 110
},
{
"epoch": 0.048749470114455275,
"grad_norm": 0.2635990083217621,
"learning_rate": 7.993030069516571e-05,
"loss": 1.7502134323120118,
"step": 115
},
{
"epoch": 0.05086901229334464,
"grad_norm": 0.24232925474643707,
"learning_rate": 7.991315537136504e-05,
"loss": 1.6973617553710938,
"step": 120
},
{
"epoch": 0.052988554472233995,
"grad_norm": 0.21670188009738922,
"learning_rate": 7.989412881680689e-05,
"loss": 1.724782943725586,
"step": 125
},
{
"epoch": 0.05510809665112336,
"grad_norm": 0.2662172019481659,
"learning_rate": 7.987322192827177e-05,
"loss": 1.7772514343261718,
"step": 130
},
{
"epoch": 0.057227638830012716,
"grad_norm": 0.26820454001426697,
"learning_rate": 7.985043569116612e-05,
"loss": 1.723222541809082,
"step": 135
},
{
"epoch": 0.05934718100890208,
"grad_norm": 0.2372129261493683,
"learning_rate": 7.982577117947597e-05,
"loss": 1.7164798736572267,
"step": 140
},
{
"epoch": 0.061466723187791436,
"grad_norm": 0.23795653879642487,
"learning_rate": 7.979922955571623e-05,
"loss": 1.6659355163574219,
"step": 145
},
{
"epoch": 0.0635862653666808,
"grad_norm": 0.22253049910068512,
"learning_rate": 7.977081207087591e-05,
"loss": 1.7030624389648437,
"step": 150
},
{
"epoch": 0.06570580754557016,
"grad_norm": 0.24633541703224182,
"learning_rate": 7.974052006435915e-05,
"loss": 1.651565933227539,
"step": 155
},
{
"epoch": 0.06782534972445951,
"grad_norm": 0.2382444590330124,
"learning_rate": 7.970835496392216e-05,
"loss": 1.6775833129882813,
"step": 160
},
{
"epoch": 0.06994489190334888,
"grad_norm": 0.2766229510307312,
"learning_rate": 7.967431828560586e-05,
"loss": 1.6641559600830078,
"step": 165
},
{
"epoch": 0.07206443408223824,
"grad_norm": 0.27501192688941956,
"learning_rate": 7.963841163366438e-05,
"loss": 1.640512466430664,
"step": 170
},
{
"epoch": 0.07418397626112759,
"grad_norm": 0.23321472108364105,
"learning_rate": 7.96006367004896e-05,
"loss": 1.639167022705078,
"step": 175
},
{
"epoch": 0.07630351844001695,
"grad_norm": 0.21392957866191864,
"learning_rate": 7.956099526653118e-05,
"loss": 1.6127048492431642,
"step": 180
},
{
"epoch": 0.07842306061890632,
"grad_norm": 0.27146685123443604,
"learning_rate": 7.951948920021286e-05,
"loss": 1.6198917388916017,
"step": 185
},
{
"epoch": 0.08054260279779568,
"grad_norm": 0.2585868239402771,
"learning_rate": 7.947612045784418e-05,
"loss": 1.618967056274414,
"step": 190
},
{
"epoch": 0.08266214497668503,
"grad_norm": 0.2556251883506775,
"learning_rate": 7.943089108352847e-05,
"loss": 1.6223142623901368,
"step": 195
},
{
"epoch": 0.0847816871555744,
"grad_norm": 0.24963316321372986,
"learning_rate": 7.938380320906634e-05,
"loss": 1.6074459075927734,
"step": 200
},
{
"epoch": 0.08690122933446376,
"grad_norm": 0.24312998354434967,
"learning_rate": 7.933485905385532e-05,
"loss": 1.6534408569335937,
"step": 205
},
{
"epoch": 0.08902077151335312,
"grad_norm": 0.24531760811805725,
"learning_rate": 7.928406092478516e-05,
"loss": 1.606509017944336,
"step": 210
},
{
"epoch": 0.09114031369224247,
"grad_norm": 0.30584201216697693,
"learning_rate": 7.923141121612922e-05,
"loss": 1.6016658782958983,
"step": 215
},
{
"epoch": 0.09325985587113184,
"grad_norm": 0.26060405373573303,
"learning_rate": 7.917691240943147e-05,
"loss": 1.5423978805541991,
"step": 220
},
{
"epoch": 0.0953793980500212,
"grad_norm": 0.24255937337875366,
"learning_rate": 7.912056707338964e-05,
"loss": 1.6322971343994142,
"step": 225
},
{
"epoch": 0.09749894022891055,
"grad_norm": 0.28530821204185486,
"learning_rate": 7.906237786373411e-05,
"loss": 1.5935253143310546,
"step": 230
},
{
"epoch": 0.09961848240779991,
"grad_norm": 0.28408628702163696,
"learning_rate": 7.900234752310274e-05,
"loss": 1.6382963180541992,
"step": 235
},
{
"epoch": 0.10173802458668928,
"grad_norm": 0.24532373249530792,
"learning_rate": 7.89404788809116e-05,
"loss": 1.6328113555908204,
"step": 240
},
{
"epoch": 0.10385756676557864,
"grad_norm": 0.2462259978055954,
"learning_rate": 7.88767748532216e-05,
"loss": 1.5996957778930665,
"step": 245
},
{
"epoch": 0.10597710894446799,
"grad_norm": 0.2827068865299225,
"learning_rate": 7.881123844260109e-05,
"loss": 1.5423406600952148,
"step": 250
},
{
"epoch": 0.10809665112335735,
"grad_norm": 0.27817490696907043,
"learning_rate": 7.874387273798427e-05,
"loss": 1.5872836112976074,
"step": 255
},
{
"epoch": 0.11021619330224672,
"grad_norm": 0.30322763323783875,
"learning_rate": 7.867468091452566e-05,
"loss": 1.555913543701172,
"step": 260
},
{
"epoch": 0.11233573548113608,
"grad_norm": 0.2901754677295685,
"learning_rate": 7.860366623345043e-05,
"loss": 1.5874847412109374,
"step": 265
},
{
"epoch": 0.11445527766002543,
"grad_norm": 0.25083526968955994,
"learning_rate": 7.853083204190066e-05,
"loss": 1.5477298736572265,
"step": 270
},
{
"epoch": 0.1165748198389148,
"grad_norm": 0.3856332302093506,
"learning_rate": 7.845618177277761e-05,
"loss": 1.5994754791259767,
"step": 275
},
{
"epoch": 0.11869436201780416,
"grad_norm": 0.3508598804473877,
"learning_rate": 7.837971894457991e-05,
"loss": 1.5365802764892578,
"step": 280
},
{
"epoch": 0.12081390419669351,
"grad_norm": 0.3150509297847748,
"learning_rate": 7.830144716123769e-05,
"loss": 1.5412468910217285,
"step": 285
},
{
"epoch": 0.12293344637558287,
"grad_norm": 0.34235191345214844,
"learning_rate": 7.822137011194281e-05,
"loss": 1.580563735961914,
"step": 290
},
{
"epoch": 0.12505298855447222,
"grad_norm": 0.2897155284881592,
"learning_rate": 7.813949157097485e-05,
"loss": 1.5645181655883789,
"step": 295
},
{
"epoch": 0.1271725307333616,
"grad_norm": 0.2858702838420868,
"learning_rate": 7.80558153975233e-05,
"loss": 1.5653106689453125,
"step": 300
},
{
"epoch": 0.12929207291225095,
"grad_norm": 0.34633851051330566,
"learning_rate": 7.797034553550563e-05,
"loss": 1.5471956253051757,
"step": 305
},
{
"epoch": 0.13141161509114033,
"grad_norm": 0.27321475744247437,
"learning_rate": 7.788308601338143e-05,
"loss": 1.5335832595825196,
"step": 310
},
{
"epoch": 0.13353115727002968,
"grad_norm": 0.26764094829559326,
"learning_rate": 7.779404094396253e-05,
"loss": 1.5111331939697266,
"step": 315
},
{
"epoch": 0.13565069944891903,
"grad_norm": 0.28431782126426697,
"learning_rate": 7.770321452421909e-05,
"loss": 1.5610011100769043,
"step": 320
},
{
"epoch": 0.1377702416278084,
"grad_norm": 0.251496285200119,
"learning_rate": 7.76106110350819e-05,
"loss": 1.6016853332519532,
"step": 325
},
{
"epoch": 0.13988978380669775,
"grad_norm": 0.2996198832988739,
"learning_rate": 7.75162348412405e-05,
"loss": 1.587802028656006,
"step": 330
},
{
"epoch": 0.1420093259855871,
"grad_norm": 0.28139063715934753,
"learning_rate": 7.742009039093747e-05,
"loss": 1.5232866287231446,
"step": 335
},
{
"epoch": 0.14412886816447648,
"grad_norm": 0.27496451139450073,
"learning_rate": 7.732218221575888e-05,
"loss": 1.508912944793701,
"step": 340
},
{
"epoch": 0.14624841034336583,
"grad_norm": 0.2885613441467285,
"learning_rate": 7.722251493042056e-05,
"loss": 1.4782177925109863,
"step": 345
},
{
"epoch": 0.14836795252225518,
"grad_norm": 0.2669691741466522,
"learning_rate": 7.712109323255065e-05,
"loss": 1.4798032760620117,
"step": 350
},
{
"epoch": 0.15048749470114456,
"grad_norm": 0.2620970606803894,
"learning_rate": 7.701792190246825e-05,
"loss": 1.54260311126709,
"step": 355
},
{
"epoch": 0.1526070368800339,
"grad_norm": 0.24850228428840637,
"learning_rate": 7.691300580295798e-05,
"loss": 1.4747392654418945,
"step": 360
},
{
"epoch": 0.15472657905892329,
"grad_norm": 0.2908382713794708,
"learning_rate": 7.680634987904091e-05,
"loss": 1.5131176948547362,
"step": 365
},
{
"epoch": 0.15684612123781264,
"grad_norm": 0.29605597257614136,
"learning_rate": 7.669795915774139e-05,
"loss": 1.5169933319091797,
"step": 370
},
{
"epoch": 0.15896566341670199,
"grad_norm": 0.2789711356163025,
"learning_rate": 7.658783874785018e-05,
"loss": 1.538616943359375,
"step": 375
},
{
"epoch": 0.16108520559559136,
"grad_norm": 0.24284884333610535,
"learning_rate": 7.64759938396836e-05,
"loss": 1.5400668144226075,
"step": 380
},
{
"epoch": 0.1632047477744807,
"grad_norm": 0.27422091364860535,
"learning_rate": 7.636242970483894e-05,
"loss": 1.5287532806396484,
"step": 385
},
{
"epoch": 0.16532428995337006,
"grad_norm": 0.2600255310535431,
"learning_rate": 7.624715169594596e-05,
"loss": 1.5154428482055664,
"step": 390
},
{
"epoch": 0.16744383213225944,
"grad_norm": 0.30100488662719727,
"learning_rate": 7.613016524641462e-05,
"loss": 1.4649045944213868,
"step": 395
},
{
"epoch": 0.1695633743111488,
"grad_norm": 0.29589682817459106,
"learning_rate": 7.601147587017897e-05,
"loss": 1.4927549362182617,
"step": 400
},
{
"epoch": 0.17168291649003814,
"grad_norm": 0.28111886978149414,
"learning_rate": 7.58910891614373e-05,
"loss": 1.5342559814453125,
"step": 405
},
{
"epoch": 0.17380245866892752,
"grad_norm": 0.26966726779937744,
"learning_rate": 7.576901079438845e-05,
"loss": 1.5068151473999023,
"step": 410
},
{
"epoch": 0.17592200084781687,
"grad_norm": 0.3016875684261322,
"learning_rate": 7.564524652296434e-05,
"loss": 1.5353782653808594,
"step": 415
},
{
"epoch": 0.17804154302670624,
"grad_norm": 0.2678869962692261,
"learning_rate": 7.551980218055879e-05,
"loss": 1.4674576759338378,
"step": 420
},
{
"epoch": 0.1801610852055956,
"grad_norm": 0.334695965051651,
"learning_rate": 7.539268367975257e-05,
"loss": 1.4947544097900392,
"step": 425
},
{
"epoch": 0.18228062738448494,
"grad_norm": 0.2693050503730774,
"learning_rate": 7.526389701203475e-05,
"loss": 1.4793996810913086,
"step": 430
},
{
"epoch": 0.18440016956337432,
"grad_norm": 0.25020599365234375,
"learning_rate": 7.513344824752025e-05,
"loss": 1.4669428825378419,
"step": 435
},
{
"epoch": 0.18651971174226367,
"grad_norm": 0.28943920135498047,
"learning_rate": 7.500134353466379e-05,
"loss": 1.5114886283874511,
"step": 440
},
{
"epoch": 0.18863925392115302,
"grad_norm": 0.30052438378334045,
"learning_rate": 7.486758909997004e-05,
"loss": 1.4719146728515624,
"step": 445
},
{
"epoch": 0.1907587961000424,
"grad_norm": 0.3031732738018036,
"learning_rate": 7.473219124770023e-05,
"loss": 1.4437793731689452,
"step": 450
},
{
"epoch": 0.19287833827893175,
"grad_norm": 0.2890752851963043,
"learning_rate": 7.459515635957488e-05,
"loss": 1.5025915145874023,
"step": 455
},
{
"epoch": 0.1949978804578211,
"grad_norm": 0.27070459723472595,
"learning_rate": 7.445649089447314e-05,
"loss": 1.496159267425537,
"step": 460
},
{
"epoch": 0.19711742263671048,
"grad_norm": 0.29470834136009216,
"learning_rate": 7.431620138812829e-05,
"loss": 1.4885208129882812,
"step": 465
},
{
"epoch": 0.19923696481559983,
"grad_norm": 0.27974218130111694,
"learning_rate": 7.417429445281972e-05,
"loss": 1.5112152099609375,
"step": 470
},
{
"epoch": 0.2013565069944892,
"grad_norm": 0.27625009417533875,
"learning_rate": 7.403077677706125e-05,
"loss": 1.4725955963134765,
"step": 475
},
{
"epoch": 0.20347604917337855,
"grad_norm": 0.30430784821510315,
"learning_rate": 7.388565512528588e-05,
"loss": 1.456763458251953,
"step": 480
},
{
"epoch": 0.2055955913522679,
"grad_norm": 0.2560711205005646,
"learning_rate": 7.373893633752698e-05,
"loss": 1.4968692779541015,
"step": 485
},
{
"epoch": 0.20771513353115728,
"grad_norm": 0.2583972215652466,
"learning_rate": 7.359062732909593e-05,
"loss": 1.4387727737426759,
"step": 490
},
{
"epoch": 0.20983467571004663,
"grad_norm": 0.3071255683898926,
"learning_rate": 7.344073509025608e-05,
"loss": 1.4839001655578614,
"step": 495
},
{
"epoch": 0.21195421788893598,
"grad_norm": 0.2775452136993408,
"learning_rate": 7.328926668589338e-05,
"loss": 1.4580822944641114,
"step": 500
},
{
"epoch": 0.21407376006782536,
"grad_norm": 0.2661127746105194,
"learning_rate": 7.313622925518332e-05,
"loss": 1.4488059997558593,
"step": 505
},
{
"epoch": 0.2161933022467147,
"grad_norm": 0.2531096637248993,
"learning_rate": 7.29816300112545e-05,
"loss": 1.430902862548828,
"step": 510
},
{
"epoch": 0.21831284442560406,
"grad_norm": 0.2797805368900299,
"learning_rate": 7.282547624084856e-05,
"loss": 1.4120213508605957,
"step": 515
},
{
"epoch": 0.22043238660449344,
"grad_norm": 0.23670950531959534,
"learning_rate": 7.266777530397688e-05,
"loss": 1.4558536529541015,
"step": 520
},
{
"epoch": 0.22255192878338279,
"grad_norm": 0.2692963778972626,
"learning_rate": 7.25085346335735e-05,
"loss": 1.4960713386535645,
"step": 525
},
{
"epoch": 0.22467147096227216,
"grad_norm": 0.31609517335891724,
"learning_rate": 7.234776173514496e-05,
"loss": 1.4226519584655761,
"step": 530
},
{
"epoch": 0.2267910131411615,
"grad_norm": 0.2578374445438385,
"learning_rate": 7.218546418641637e-05,
"loss": 1.468405818939209,
"step": 535
},
{
"epoch": 0.22891055532005086,
"grad_norm": 0.2876923084259033,
"learning_rate": 7.202164963697442e-05,
"loss": 1.486027717590332,
"step": 540
},
{
"epoch": 0.23103009749894024,
"grad_norm": 0.3240324854850769,
"learning_rate": 7.185632580790669e-05,
"loss": 1.4940210342407227,
"step": 545
},
{
"epoch": 0.2331496396778296,
"grad_norm": 0.31703120470046997,
"learning_rate": 7.16895004914378e-05,
"loss": 1.5034875869750977,
"step": 550
},
{
"epoch": 0.23526918185671894,
"grad_norm": 0.28383785486221313,
"learning_rate": 7.152118155056215e-05,
"loss": 1.479176902770996,
"step": 555
},
{
"epoch": 0.23738872403560832,
"grad_norm": 0.270242840051651,
"learning_rate": 7.135137691867325e-05,
"loss": 1.4717596054077149,
"step": 560
},
{
"epoch": 0.23950826621449767,
"grad_norm": 0.3007745146751404,
"learning_rate": 7.118009459918988e-05,
"loss": 1.4467639923095703,
"step": 565
},
{
"epoch": 0.24162780839338702,
"grad_norm": 0.2557879388332367,
"learning_rate": 7.100734266517878e-05,
"loss": 1.4492422103881837,
"step": 570
},
{
"epoch": 0.2437473505722764,
"grad_norm": 0.2721036672592163,
"learning_rate": 7.083312925897417e-05,
"loss": 1.45943603515625,
"step": 575
},
{
"epoch": 0.24586689275116574,
"grad_norm": 0.32169830799102783,
"learning_rate": 7.065746259179405e-05,
"loss": 1.411389446258545,
"step": 580
},
{
"epoch": 0.24798643493005512,
"grad_norm": 0.28373900055885315,
"learning_rate": 7.048035094335305e-05,
"loss": 1.4764095306396485,
"step": 585
},
{
"epoch": 0.25010597710894444,
"grad_norm": 0.25566622614860535,
"learning_rate": 7.030180266147227e-05,
"loss": 1.5153303146362305,
"step": 590
},
{
"epoch": 0.2522255192878338,
"grad_norm": 0.27002280950546265,
"learning_rate": 7.012182616168579e-05,
"loss": 1.4313461303710937,
"step": 595
},
{
"epoch": 0.2543450614667232,
"grad_norm": 0.3017001152038574,
"learning_rate": 6.994042992684406e-05,
"loss": 1.4548938751220704,
"step": 600
},
{
"epoch": 0.2564646036456125,
"grad_norm": 0.271880567073822,
"learning_rate": 6.975762250671399e-05,
"loss": 1.4693748474121093,
"step": 605
},
{
"epoch": 0.2585841458245019,
"grad_norm": 0.2892245650291443,
"learning_rate": 6.957341251757607e-05,
"loss": 1.4723827362060546,
"step": 610
},
{
"epoch": 0.2607036880033913,
"grad_norm": 0.2887561321258545,
"learning_rate": 6.938780864181821e-05,
"loss": 1.490568256378174,
"step": 615
},
{
"epoch": 0.26282323018228065,
"grad_norm": 0.32353076338768005,
"learning_rate": 6.920081962752651e-05,
"loss": 1.4510770797729493,
"step": 620
},
{
"epoch": 0.26494277236117,
"grad_norm": 0.2901272773742676,
"learning_rate": 6.901245428807295e-05,
"loss": 1.4940450668334961,
"step": 625
},
{
"epoch": 0.26706231454005935,
"grad_norm": 0.28859612345695496,
"learning_rate": 6.882272150169995e-05,
"loss": 1.42889347076416,
"step": 630
},
{
"epoch": 0.26918185671894873,
"grad_norm": 0.23918817937374115,
"learning_rate": 6.8631630211102e-05,
"loss": 1.454274559020996,
"step": 635
},
{
"epoch": 0.27130139889783805,
"grad_norm": 0.3091767132282257,
"learning_rate": 6.843918942300407e-05,
"loss": 1.4363265991210938,
"step": 640
},
{
"epoch": 0.27342094107672743,
"grad_norm": 0.23447060585021973,
"learning_rate": 6.824540820773715e-05,
"loss": 1.4451414108276368,
"step": 645
},
{
"epoch": 0.2755404832556168,
"grad_norm": 0.29321345686912537,
"learning_rate": 6.805029569881068e-05,
"loss": 1.4987229347229003,
"step": 650
},
{
"epoch": 0.27766002543450613,
"grad_norm": 0.28928396105766296,
"learning_rate": 6.785386109248212e-05,
"loss": 1.4561105728149415,
"step": 655
},
{
"epoch": 0.2797795676133955,
"grad_norm": 0.28260499238967896,
"learning_rate": 6.765611364732352e-05,
"loss": 1.4290050506591796,
"step": 660
},
{
"epoch": 0.2818991097922849,
"grad_norm": 0.2832860052585602,
"learning_rate": 6.7457062683785e-05,
"loss": 1.4748251914978028,
"step": 665
},
{
"epoch": 0.2840186519711742,
"grad_norm": 0.24946916103363037,
"learning_rate": 6.725671758375566e-05,
"loss": 1.4290283203125,
"step": 670
},
{
"epoch": 0.2861381941500636,
"grad_norm": 0.31382250785827637,
"learning_rate": 6.705508779012116e-05,
"loss": 1.4489578247070312,
"step": 675
},
{
"epoch": 0.28825773632895296,
"grad_norm": 0.2788618206977844,
"learning_rate": 6.685218280631882e-05,
"loss": 1.5023835182189942,
"step": 680
},
{
"epoch": 0.2903772785078423,
"grad_norm": 0.25866934657096863,
"learning_rate": 6.664801219588959e-05,
"loss": 1.4318917274475098,
"step": 685
},
{
"epoch": 0.29249682068673166,
"grad_norm": 0.302611380815506,
"learning_rate": 6.644258558202737e-05,
"loss": 1.3967741012573243,
"step": 690
},
{
"epoch": 0.29461636286562104,
"grad_norm": 0.27050140500068665,
"learning_rate": 6.623591264712533e-05,
"loss": 1.4011164665222169,
"step": 695
},
{
"epoch": 0.29673590504451036,
"grad_norm": 0.29683297872543335,
"learning_rate": 6.602800313231965e-05,
"loss": 1.429396915435791,
"step": 700
},
{
"epoch": 0.29885544722339974,
"grad_norm": 0.2876051366329193,
"learning_rate": 6.581886683703039e-05,
"loss": 1.4962058067321777,
"step": 705
},
{
"epoch": 0.3009749894022891,
"grad_norm": 0.3241434395313263,
"learning_rate": 6.560851361849952e-05,
"loss": 1.4170503616333008,
"step": 710
},
{
"epoch": 0.30309453158117844,
"grad_norm": 0.26529553532600403,
"learning_rate": 6.539695339132641e-05,
"loss": 1.4393165588378907,
"step": 715
},
{
"epoch": 0.3052140737600678,
"grad_norm": 0.2935298681259155,
"learning_rate": 6.518419612700045e-05,
"loss": 1.392915916442871,
"step": 720
},
{
"epoch": 0.3073336159389572,
"grad_norm": 0.31954842805862427,
"learning_rate": 6.497025185343117e-05,
"loss": 1.4277592658996583,
"step": 725
},
{
"epoch": 0.30945315811784657,
"grad_norm": 0.2968384325504303,
"learning_rate": 6.475513065447546e-05,
"loss": 1.4118597030639648,
"step": 730
},
{
"epoch": 0.3115727002967359,
"grad_norm": 0.25474441051483154,
"learning_rate": 6.453884266946239e-05,
"loss": 1.4357147216796875,
"step": 735
},
{
"epoch": 0.31369224247562527,
"grad_norm": 0.29025760293006897,
"learning_rate": 6.432139809271524e-05,
"loss": 1.4395485877990724,
"step": 740
},
{
"epoch": 0.31581178465451465,
"grad_norm": 0.2717384994029999,
"learning_rate": 6.410280717307109e-05,
"loss": 1.4146919250488281,
"step": 745
},
{
"epoch": 0.31793132683340397,
"grad_norm": 0.2808854281902313,
"learning_rate": 6.38830802133977e-05,
"loss": 1.3758182525634766,
"step": 750
},
{
"epoch": 0.32005086901229335,
"grad_norm": 0.2576456665992737,
"learning_rate": 6.366222757010789e-05,
"loss": 1.4187177658081054,
"step": 755
},
{
"epoch": 0.3221704111911827,
"grad_norm": 0.29566165804862976,
"learning_rate": 6.344025965267145e-05,
"loss": 1.4072937965393066,
"step": 760
},
{
"epoch": 0.32428995337007205,
"grad_norm": 0.32720693945884705,
"learning_rate": 6.321718692312454e-05,
"loss": 1.4214115142822266,
"step": 765
},
{
"epoch": 0.3264094955489614,
"grad_norm": 0.2695600390434265,
"learning_rate": 6.299301989557648e-05,
"loss": 1.4314722061157226,
"step": 770
},
{
"epoch": 0.3285290377278508,
"grad_norm": 0.28979578614234924,
"learning_rate": 6.276776913571428e-05,
"loss": 1.4325143814086914,
"step": 775
},
{
"epoch": 0.3306485799067401,
"grad_norm": 0.2972823977470398,
"learning_rate": 6.254144526030463e-05,
"loss": 1.4565296173095703,
"step": 780
},
{
"epoch": 0.3327681220856295,
"grad_norm": 0.30112603306770325,
"learning_rate": 6.231405893669342e-05,
"loss": 1.3919147491455077,
"step": 785
},
{
"epoch": 0.3348876642645189,
"grad_norm": 0.2823403775691986,
"learning_rate": 6.208562088230312e-05,
"loss": 1.41244535446167,
"step": 790
},
{
"epoch": 0.3370072064434082,
"grad_norm": 0.28726106882095337,
"learning_rate": 6.18561418641274e-05,
"loss": 1.3777569770812987,
"step": 795
},
{
"epoch": 0.3391267486222976,
"grad_norm": 0.2762209177017212,
"learning_rate": 6.162563269822391e-05,
"loss": 1.384923553466797,
"step": 800
},
{
"epoch": 0.34124629080118696,
"grad_norm": 0.3245485723018646,
"learning_rate": 6.139410424920428e-05,
"loss": 1.4152031898498536,
"step": 805
},
{
"epoch": 0.3433658329800763,
"grad_norm": 0.2730148434638977,
"learning_rate": 6.116156742972211e-05,
"loss": 1.3835166931152343,
"step": 810
},
{
"epoch": 0.34548537515896566,
"grad_norm": 0.29458382725715637,
"learning_rate": 6.0928033199958696e-05,
"loss": 1.427520751953125,
"step": 815
},
{
"epoch": 0.34760491733785503,
"grad_norm": 0.3228289783000946,
"learning_rate": 6.0693512567106275e-05,
"loss": 1.4304326057434082,
"step": 820
},
{
"epoch": 0.34972445951674436,
"grad_norm": 0.2773505747318268,
"learning_rate": 6.04580165848494e-05,
"loss": 1.3744176864624023,
"step": 825
},
{
"epoch": 0.35184400169563373,
"grad_norm": 0.28608816862106323,
"learning_rate": 6.022155635284384e-05,
"loss": 1.3878042221069335,
"step": 830
},
{
"epoch": 0.3539635438745231,
"grad_norm": 0.3341328203678131,
"learning_rate": 5.9984143016193405e-05,
"loss": 1.3884491920471191,
"step": 835
},
{
"epoch": 0.3560830860534125,
"grad_norm": 0.2521774172782898,
"learning_rate": 5.974578776492474e-05,
"loss": 1.399277114868164,
"step": 840
},
{
"epoch": 0.3582026282323018,
"grad_norm": 0.3582310378551483,
"learning_rate": 5.950650183345983e-05,
"loss": 1.435608196258545,
"step": 845
},
{
"epoch": 0.3603221704111912,
"grad_norm": 0.25554847717285156,
"learning_rate": 5.9266296500086485e-05,
"loss": 1.3810482025146484,
"step": 850
},
{
"epoch": 0.36244171259008057,
"grad_norm": 0.3398644030094147,
"learning_rate": 5.902518308642679e-05,
"loss": 1.3792922973632813,
"step": 855
},
{
"epoch": 0.3645612547689699,
"grad_norm": 0.31561702489852905,
"learning_rate": 5.8783172956903454e-05,
"loss": 1.3731876373291017,
"step": 860
},
{
"epoch": 0.36668079694785927,
"grad_norm": 0.27747732400894165,
"learning_rate": 5.8540277518204216e-05,
"loss": 1.44711332321167,
"step": 865
},
{
"epoch": 0.36880033912674864,
"grad_norm": 0.2459808588027954,
"learning_rate": 5.829650821874417e-05,
"loss": 1.4043840408325194,
"step": 870
},
{
"epoch": 0.37091988130563797,
"grad_norm": 0.3043918013572693,
"learning_rate": 5.8051876548126154e-05,
"loss": 1.3533967971801757,
"step": 875
},
{
"epoch": 0.37303942348452734,
"grad_norm": 0.2619969844818115,
"learning_rate": 5.7806394036599236e-05,
"loss": 1.356023406982422,
"step": 880
},
{
"epoch": 0.3751589656634167,
"grad_norm": 0.3271007537841797,
"learning_rate": 5.75600722545153e-05,
"loss": 1.422224521636963,
"step": 885
},
{
"epoch": 0.37727850784230604,
"grad_norm": 0.2727625370025635,
"learning_rate": 5.731292281178359e-05,
"loss": 1.4609024047851562,
"step": 890
},
{
"epoch": 0.3793980500211954,
"grad_norm": 0.2901216745376587,
"learning_rate": 5.706495735732361e-05,
"loss": 1.3732920646667481,
"step": 895
},
{
"epoch": 0.3815175922000848,
"grad_norm": 0.266664057970047,
"learning_rate": 5.6816187578516006e-05,
"loss": 1.3978240013122558,
"step": 900
},
{
"epoch": 0.3836371343789741,
"grad_norm": 0.3094821274280548,
"learning_rate": 5.656662520065173e-05,
"loss": 1.402490997314453,
"step": 905
},
{
"epoch": 0.3857566765578635,
"grad_norm": 0.27117303013801575,
"learning_rate": 5.631628198637936e-05,
"loss": 1.426488494873047,
"step": 910
},
{
"epoch": 0.3878762187367529,
"grad_norm": 0.3420483469963074,
"learning_rate": 5.606516973515076e-05,
"loss": 1.3942727088928222,
"step": 915
},
{
"epoch": 0.3899957609156422,
"grad_norm": 0.25338223576545715,
"learning_rate": 5.581330028266486e-05,
"loss": 1.3662307739257813,
"step": 920
},
{
"epoch": 0.3921153030945316,
"grad_norm": 0.325119286775589,
"learning_rate": 5.5560685500309847e-05,
"loss": 1.391487979888916,
"step": 925
},
{
"epoch": 0.39423484527342095,
"grad_norm": 0.3021058142185211,
"learning_rate": 5.5307337294603595e-05,
"loss": 1.4478245735168458,
"step": 930
},
{
"epoch": 0.3963543874523103,
"grad_norm": 0.28974759578704834,
"learning_rate": 5.5053267606632515e-05,
"loss": 1.394073486328125,
"step": 935
},
{
"epoch": 0.39847392963119965,
"grad_norm": 0.28786420822143555,
"learning_rate": 5.479848841148871e-05,
"loss": 1.3515398025512695,
"step": 940
},
{
"epoch": 0.40059347181008903,
"grad_norm": 0.2820453345775604,
"learning_rate": 5.454301171770554e-05,
"loss": 1.4101629257202148,
"step": 945
},
{
"epoch": 0.4027130139889784,
"grad_norm": 0.2964695692062378,
"learning_rate": 5.428684956669166e-05,
"loss": 1.3510038375854492,
"step": 950
},
{
"epoch": 0.40483255616786773,
"grad_norm": 0.36414000391960144,
"learning_rate": 5.403001403216344e-05,
"loss": 1.351788330078125,
"step": 955
},
{
"epoch": 0.4069520983467571,
"grad_norm": 0.3478855788707733,
"learning_rate": 5.37725172195759e-05,
"loss": 1.3816933631896973,
"step": 960
},
{
"epoch": 0.4090716405256465,
"grad_norm": 0.29232123494148254,
"learning_rate": 5.351437126555215e-05,
"loss": 1.3767603874206542,
"step": 965
},
{
"epoch": 0.4111911827045358,
"grad_norm": 0.26996973156929016,
"learning_rate": 5.325558833731134e-05,
"loss": 1.4302228927612304,
"step": 970
},
{
"epoch": 0.4133107248834252,
"grad_norm": 0.2792315185070038,
"learning_rate": 5.2996180632095204e-05,
"loss": 1.4092483520507812,
"step": 975
},
{
"epoch": 0.41543026706231456,
"grad_norm": 0.30166417360305786,
"learning_rate": 5.2736160376593155e-05,
"loss": 1.3743454933166503,
"step": 980
},
{
"epoch": 0.4175498092412039,
"grad_norm": 0.30971014499664307,
"learning_rate": 5.247553982636596e-05,
"loss": 1.3428321838378907,
"step": 985
},
{
"epoch": 0.41966935142009326,
"grad_norm": 0.27801430225372314,
"learning_rate": 5.22143312652682e-05,
"loss": 1.374162197113037,
"step": 990
},
{
"epoch": 0.42178889359898264,
"grad_norm": 0.28014230728149414,
"learning_rate": 5.195254700486917e-05,
"loss": 1.3973739624023438,
"step": 995
},
{
"epoch": 0.42390843577787196,
"grad_norm": 0.2701701521873474,
"learning_rate": 5.1690199383872695e-05,
"loss": 1.3959041595458985,
"step": 1000
},
{
"epoch": 0.42602797795676134,
"grad_norm": 0.25473693013191223,
"learning_rate": 5.1427300767535525e-05,
"loss": 1.3902218818664551,
"step": 1005
},
{
"epoch": 0.4281475201356507,
"grad_norm": 0.3194516897201538,
"learning_rate": 5.116386354708454e-05,
"loss": 1.4032052040100098,
"step": 1010
},
{
"epoch": 0.43026706231454004,
"grad_norm": 0.2776491641998291,
"learning_rate": 5.089990013913265e-05,
"loss": 1.3790559768676758,
"step": 1015
},
{
"epoch": 0.4323866044934294,
"grad_norm": 0.2968991994857788,
"learning_rate": 5.063542298509368e-05,
"loss": 1.363190269470215,
"step": 1020
},
{
"epoch": 0.4345061466723188,
"grad_norm": 0.297929972410202,
"learning_rate": 5.037044455059587e-05,
"loss": 1.4066566467285155,
"step": 1025
},
{
"epoch": 0.4366256888512081,
"grad_norm": 0.27815088629722595,
"learning_rate": 5.010497732489437e-05,
"loss": 1.2938894271850585,
"step": 1030
},
{
"epoch": 0.4387452310300975,
"grad_norm": 0.2646707594394684,
"learning_rate": 4.9839033820282554e-05,
"loss": 1.3316048622131347,
"step": 1035
},
{
"epoch": 0.44086477320898687,
"grad_norm": 0.2931884825229645,
"learning_rate": 4.9572626571502316e-05,
"loss": 1.3887158393859864,
"step": 1040
},
{
"epoch": 0.4429843153878762,
"grad_norm": 0.3342426121234894,
"learning_rate": 4.930576813515325e-05,
"loss": 1.428691291809082,
"step": 1045
},
{
"epoch": 0.44510385756676557,
"grad_norm": 0.27253708243370056,
"learning_rate": 4.903847108910083e-05,
"loss": 1.36983642578125,
"step": 1050
},
{
"epoch": 0.44722339974565495,
"grad_norm": 0.31104961037635803,
"learning_rate": 4.877074803188352e-05,
"loss": 1.3992633819580078,
"step": 1055
},
{
"epoch": 0.4493429419245443,
"grad_norm": 0.34496423602104187,
"learning_rate": 4.8502611582119065e-05,
"loss": 1.3749829292297364,
"step": 1060
},
{
"epoch": 0.45146248410343365,
"grad_norm": 0.31395459175109863,
"learning_rate": 4.823407437790967e-05,
"loss": 1.361133861541748,
"step": 1065
},
{
"epoch": 0.453582026282323,
"grad_norm": 0.23204976320266724,
"learning_rate": 4.79651490762463e-05,
"loss": 1.3968187332153321,
"step": 1070
},
{
"epoch": 0.4557015684612124,
"grad_norm": 0.2741721570491791,
"learning_rate": 4.769584835241221e-05,
"loss": 1.3845906257629395,
"step": 1075
},
{
"epoch": 0.4578211106401017,
"grad_norm": 0.2999100089073181,
"learning_rate": 4.742618489938542e-05,
"loss": 1.3473276138305663,
"step": 1080
},
{
"epoch": 0.4599406528189911,
"grad_norm": 0.29869338870048523,
"learning_rate": 4.7156171427240534e-05,
"loss": 1.336467170715332,
"step": 1085
},
{
"epoch": 0.4620601949978805,
"grad_norm": 0.29905396699905396,
"learning_rate": 4.688582066254965e-05,
"loss": 1.3615987777709961,
"step": 1090
},
{
"epoch": 0.4641797371767698,
"grad_norm": 0.31158316135406494,
"learning_rate": 4.6615145347782455e-05,
"loss": 1.3145706176757812,
"step": 1095
},
{
"epoch": 0.4662992793556592,
"grad_norm": 0.294842928647995,
"learning_rate": 4.634415824070574e-05,
"loss": 1.3890361785888672,
"step": 1100
},
{
"epoch": 0.46841882153454856,
"grad_norm": 0.35659804940223694,
"learning_rate": 4.607287211378202e-05,
"loss": 1.3405324935913085,
"step": 1105
},
{
"epoch": 0.4705383637134379,
"grad_norm": 0.2612285017967224,
"learning_rate": 4.5801299753567546e-05,
"loss": 1.4677175521850585,
"step": 1110
},
{
"epoch": 0.47265790589232726,
"grad_norm": 0.2813534736633301,
"learning_rate": 4.55294539601096e-05,
"loss": 1.3577984809875487,
"step": 1115
},
{
"epoch": 0.47477744807121663,
"grad_norm": 0.3052712678909302,
"learning_rate": 4.525734754634327e-05,
"loss": 1.4065069198608398,
"step": 1120
},
{
"epoch": 0.47689699025010596,
"grad_norm": 0.2793770432472229,
"learning_rate": 4.498499333748744e-05,
"loss": 1.3577356338500977,
"step": 1125
},
{
"epoch": 0.47901653242899533,
"grad_norm": 0.27925431728363037,
"learning_rate": 4.471240417044039e-05,
"loss": 1.348134708404541,
"step": 1130
},
{
"epoch": 0.4811360746078847,
"grad_norm": 0.31317630410194397,
"learning_rate": 4.443959289317468e-05,
"loss": 1.381777572631836,
"step": 1135
},
{
"epoch": 0.48325561678677403,
"grad_norm": 0.2854962646961212,
"learning_rate": 4.4166572364131605e-05,
"loss": 1.3869144439697265,
"step": 1140
},
{
"epoch": 0.4853751589656634,
"grad_norm": 0.30988067388534546,
"learning_rate": 4.389335545161517e-05,
"loss": 1.3498130798339845,
"step": 1145
},
{
"epoch": 0.4874947011445528,
"grad_norm": 0.29037973284721375,
"learning_rate": 4.361995503318552e-05,
"loss": 1.3699258804321288,
"step": 1150
},
{
"epoch": 0.4896142433234421,
"grad_norm": 0.3077887296676636,
"learning_rate": 4.3346383995052025e-05,
"loss": 1.343306541442871,
"step": 1155
},
{
"epoch": 0.4917337855023315,
"grad_norm": 0.3252895176410675,
"learning_rate": 4.307265523146588e-05,
"loss": 1.3722208976745605,
"step": 1160
},
{
"epoch": 0.49385332768122087,
"grad_norm": 0.29235875606536865,
"learning_rate": 4.279878164411236e-05,
"loss": 1.3436902999877929,
"step": 1165
},
{
"epoch": 0.49597286986011024,
"grad_norm": 0.3031516671180725,
"learning_rate": 4.252477614150274e-05,
"loss": 1.3670766830444336,
"step": 1170
},
{
"epoch": 0.49809241203899957,
"grad_norm": 0.28785112500190735,
"learning_rate": 4.2250651638365884e-05,
"loss": 1.3856138229370116,
"step": 1175
},
{
"epoch": 0.5002119542178889,
"grad_norm": 0.3161895275115967,
"learning_rate": 4.1976421055039494e-05,
"loss": 1.4238411903381347,
"step": 1180
},
{
"epoch": 0.5023314963967783,
"grad_norm": 0.2956555187702179,
"learning_rate": 4.170209731686118e-05,
"loss": 1.3821171760559081,
"step": 1185
},
{
"epoch": 0.5044510385756676,
"grad_norm": 0.27351105213165283,
"learning_rate": 4.142769335355923e-05,
"loss": 1.4207469940185546,
"step": 1190
},
{
"epoch": 0.506570580754557,
"grad_norm": 0.3182237148284912,
"learning_rate": 4.115322209864316e-05,
"loss": 1.3701305389404297,
"step": 1195
},
{
"epoch": 0.5086901229334464,
"grad_norm": 0.3289029002189636,
"learning_rate": 4.0878696488794194e-05,
"loss": 1.28782320022583,
"step": 1200
},
{
"epoch": 0.5108096651123357,
"grad_norm": 0.3152224123477936,
"learning_rate": 4.060412946325544e-05,
"loss": 1.3477300643920898,
"step": 1205
},
{
"epoch": 0.512929207291225,
"grad_norm": 0.3382745683193207,
"learning_rate": 4.032953396322207e-05,
"loss": 1.3643617630004883,
"step": 1210
},
{
"epoch": 0.5150487494701145,
"grad_norm": 0.28936922550201416,
"learning_rate": 4.0054922931231344e-05,
"loss": 1.3420391082763672,
"step": 1215
},
{
"epoch": 0.5171682916490038,
"grad_norm": 0.27600735425949097,
"learning_rate": 3.978030931055258e-05,
"loss": 1.3704567909240724,
"step": 1220
},
{
"epoch": 0.5192878338278932,
"grad_norm": 0.2981545329093933,
"learning_rate": 3.950570604457715e-05,
"loss": 1.3226594924926758,
"step": 1225
},
{
"epoch": 0.5214073760067826,
"grad_norm": 0.3156023323535919,
"learning_rate": 3.92311260762083e-05,
"loss": 1.3496877670288085,
"step": 1230
},
{
"epoch": 0.5235269181856719,
"grad_norm": 0.3057446777820587,
"learning_rate": 3.895658234725127e-05,
"loss": 1.3594446182250977,
"step": 1235
},
{
"epoch": 0.5256464603645613,
"grad_norm": 0.26598092913627625,
"learning_rate": 3.8682087797803173e-05,
"loss": 1.369096565246582,
"step": 1240
},
{
"epoch": 0.5277660025434506,
"grad_norm": 0.266901433467865,
"learning_rate": 3.8407655365643156e-05,
"loss": 1.3782513618469239,
"step": 1245
},
{
"epoch": 0.52988554472234,
"grad_norm": 0.3055073022842407,
"learning_rate": 3.813329798562259e-05,
"loss": 1.364052677154541,
"step": 1250
},
{
"epoch": 0.5320050869012294,
"grad_norm": 0.2441306859254837,
"learning_rate": 3.7859028589055397e-05,
"loss": 1.3302003860473632,
"step": 1255
},
{
"epoch": 0.5341246290801187,
"grad_norm": 0.29582616686820984,
"learning_rate": 3.758486010310856e-05,
"loss": 1.3652389526367188,
"step": 1260
},
{
"epoch": 0.536244171259008,
"grad_norm": 0.30144402384757996,
"learning_rate": 3.731080545019285e-05,
"loss": 1.3447776794433595,
"step": 1265
},
{
"epoch": 0.5383637134378975,
"grad_norm": 0.296194851398468,
"learning_rate": 3.703687754735372e-05,
"loss": 1.359937572479248,
"step": 1270
},
{
"epoch": 0.5404832556167868,
"grad_norm": 0.27526044845581055,
"learning_rate": 3.676308930566251e-05,
"loss": 1.3508113861083983,
"step": 1275
},
{
"epoch": 0.5426027977956761,
"grad_norm": 0.2665107548236847,
"learning_rate": 3.6489453629607886e-05,
"loss": 1.3498497009277344,
"step": 1280
},
{
"epoch": 0.5447223399745655,
"grad_norm": 0.2724160850048065,
"learning_rate": 3.6215983416487625e-05,
"loss": 1.3505146026611328,
"step": 1285
},
{
"epoch": 0.5468418821534549,
"grad_norm": 0.29234686493873596,
"learning_rate": 3.5942691555800746e-05,
"loss": 1.3415228843688964,
"step": 1290
},
{
"epoch": 0.5489614243323442,
"grad_norm": 0.29788339138031006,
"learning_rate": 3.566959092863991e-05,
"loss": 1.359758186340332,
"step": 1295
},
{
"epoch": 0.5510809665112336,
"grad_norm": 0.3022080957889557,
"learning_rate": 3.5396694407084403e-05,
"loss": 1.3911295890808106,
"step": 1300
},
{
"epoch": 0.5532005086901229,
"grad_norm": 0.30012717843055725,
"learning_rate": 3.5124014853593356e-05,
"loss": 1.3562230110168456,
"step": 1305
},
{
"epoch": 0.5553200508690123,
"grad_norm": 0.2808501124382019,
"learning_rate": 3.485156512039953e-05,
"loss": 1.3265674591064454,
"step": 1310
},
{
"epoch": 0.5574395930479017,
"grad_norm": 0.26187118887901306,
"learning_rate": 3.4579358048903544e-05,
"loss": 1.3436233520507812,
"step": 1315
},
{
"epoch": 0.559559135226791,
"grad_norm": 0.29779013991355896,
"learning_rate": 3.4307406469068604e-05,
"loss": 1.3720013618469238,
"step": 1320
},
{
"epoch": 0.5616786774056803,
"grad_norm": 0.31964024901390076,
"learning_rate": 3.403572319881583e-05,
"loss": 1.345313549041748,
"step": 1325
},
{
"epoch": 0.5637982195845698,
"grad_norm": 0.3027617335319519,
"learning_rate": 3.376432104342004e-05,
"loss": 1.3682847023010254,
"step": 1330
},
{
"epoch": 0.5659177617634591,
"grad_norm": 0.30285483598709106,
"learning_rate": 3.349321279490628e-05,
"loss": 1.3469213485717773,
"step": 1335
},
{
"epoch": 0.5680373039423484,
"grad_norm": 0.272352397441864,
"learning_rate": 3.322241123144684e-05,
"loss": 1.3357921600341798,
"step": 1340
},
{
"epoch": 0.5701568461212378,
"grad_norm": 0.32597053050994873,
"learning_rate": 3.295192911675896e-05,
"loss": 1.3577280044555664,
"step": 1345
},
{
"epoch": 0.5722763883001272,
"grad_norm": 0.2904679477214813,
"learning_rate": 3.268177919950331e-05,
"loss": 1.3203404426574707,
"step": 1350
},
{
"epoch": 0.5743959304790165,
"grad_norm": 0.26609060168266296,
"learning_rate": 3.241197421268306e-05,
"loss": 1.3391555786132812,
"step": 1355
},
{
"epoch": 0.5765154726579059,
"grad_norm": 0.3589499890804291,
"learning_rate": 3.214252687304369e-05,
"loss": 1.356784725189209,
"step": 1360
},
{
"epoch": 0.5786350148367952,
"grad_norm": 0.26654329895973206,
"learning_rate": 3.187344988047376e-05,
"loss": 1.3379940032958983,
"step": 1365
},
{
"epoch": 0.5807545570156846,
"grad_norm": 0.2604674994945526,
"learning_rate": 3.160475591740611e-05,
"loss": 1.3727352142333984,
"step": 1370
},
{
"epoch": 0.582874099194574,
"grad_norm": 0.2784073054790497,
"learning_rate": 3.133645764822028e-05,
"loss": 1.2988357543945312,
"step": 1375
},
{
"epoch": 0.5849936413734633,
"grad_norm": 0.32018858194351196,
"learning_rate": 3.106856771864553e-05,
"loss": 1.3487788200378419,
"step": 1380
},
{
"epoch": 0.5871131835523526,
"grad_norm": 0.2879828214645386,
"learning_rate": 3.0801098755164784e-05,
"loss": 1.3587177276611329,
"step": 1385
},
{
"epoch": 0.5892327257312421,
"grad_norm": 0.2925914227962494,
"learning_rate": 3.0534063364419536e-05,
"loss": 1.3518965721130372,
"step": 1390
},
{
"epoch": 0.5913522679101314,
"grad_norm": 0.27614936232566833,
"learning_rate": 3.026747413261567e-05,
"loss": 1.3254805564880372,
"step": 1395
},
{
"epoch": 0.5934718100890207,
"grad_norm": 0.3269197940826416,
"learning_rate": 3.0001343624930176e-05,
"loss": 1.3298006057739258,
"step": 1400
},
{
"epoch": 0.5955913522679102,
"grad_norm": 0.31079626083374023,
"learning_rate": 2.973568438491901e-05,
"loss": 1.3350279808044434,
"step": 1405
},
{
"epoch": 0.5977108944467995,
"grad_norm": 0.3243544101715088,
"learning_rate": 2.947050893392576e-05,
"loss": 1.342665672302246,
"step": 1410
},
{
"epoch": 0.5998304366256888,
"grad_norm": 0.2529478073120117,
"learning_rate": 2.9205829770491588e-05,
"loss": 1.3809905052185059,
"step": 1415
},
{
"epoch": 0.6019499788045782,
"grad_norm": 0.28786149621009827,
"learning_rate": 2.894165936976607e-05,
"loss": 1.3337972640991211,
"step": 1420
},
{
"epoch": 0.6040695209834676,
"grad_norm": 0.258808970451355,
"learning_rate": 2.867801018291919e-05,
"loss": 1.3615898132324218,
"step": 1425
},
{
"epoch": 0.6061890631623569,
"grad_norm": 0.33244943618774414,
"learning_rate": 2.8414894636554545e-05,
"loss": 1.2915332794189454,
"step": 1430
},
{
"epoch": 0.6083086053412463,
"grad_norm": 0.2811191976070404,
"learning_rate": 2.815232513212359e-05,
"loss": 1.3436386108398437,
"step": 1435
},
{
"epoch": 0.6104281475201356,
"grad_norm": 0.27492520213127136,
"learning_rate": 2.789031404534112e-05,
"loss": 1.3245553970336914,
"step": 1440
},
{
"epoch": 0.612547689699025,
"grad_norm": 0.3087509274482727,
"learning_rate": 2.762887372560199e-05,
"loss": 1.3431172370910645,
"step": 1445
},
{
"epoch": 0.6146672318779144,
"grad_norm": 0.27661657333374023,
"learning_rate": 2.736801649539901e-05,
"loss": 1.3166393280029296,
"step": 1450
},
{
"epoch": 0.6167867740568037,
"grad_norm": 0.3516507148742676,
"learning_rate": 2.710775464974219e-05,
"loss": 1.3689906120300293,
"step": 1455
},
{
"epoch": 0.6189063162356931,
"grad_norm": 0.2980262339115143,
"learning_rate": 2.6848100455579244e-05,
"loss": 1.3488425254821776,
"step": 1460
},
{
"epoch": 0.6210258584145825,
"grad_norm": 0.31659355759620667,
"learning_rate": 2.658906615121734e-05,
"loss": 1.3151927947998048,
"step": 1465
},
{
"epoch": 0.6231454005934718,
"grad_norm": 0.29856571555137634,
"learning_rate": 2.633066394574634e-05,
"loss": 1.3405999183654784,
"step": 1470
},
{
"epoch": 0.6252649427723612,
"grad_norm": 0.2728654444217682,
"learning_rate": 2.607290601846334e-05,
"loss": 1.3735544204711914,
"step": 1475
},
{
"epoch": 0.6273844849512505,
"grad_norm": 0.2929301857948303,
"learning_rate": 2.5815804518298575e-05,
"loss": 1.3423578262329101,
"step": 1480
},
{
"epoch": 0.6295040271301399,
"grad_norm": 0.3109055161476135,
"learning_rate": 2.555937156324287e-05,
"loss": 1.3435787200927733,
"step": 1485
},
{
"epoch": 0.6316235693090293,
"grad_norm": 0.3500500023365021,
"learning_rate": 2.5303619239776418e-05,
"loss": 1.3023492813110351,
"step": 1490
},
{
"epoch": 0.6337431114879186,
"grad_norm": 0.2653202712535858,
"learning_rate": 2.5048559602299146e-05,
"loss": 1.3653303146362306,
"step": 1495
},
{
"epoch": 0.6358626536668079,
"grad_norm": 0.31638070940971375,
"learning_rate": 2.4794204672562552e-05,
"loss": 1.345554256439209,
"step": 1500
},
{
"epoch": 0.6379821958456974,
"grad_norm": 0.26279744505882263,
"learning_rate": 2.4540566439103056e-05,
"loss": 1.3190049171447753,
"step": 1505
},
{
"epoch": 0.6401017380245867,
"grad_norm": 0.2678429186344147,
"learning_rate": 2.4287656856676974e-05,
"loss": 1.297449493408203,
"step": 1510
},
{
"epoch": 0.642221280203476,
"grad_norm": 0.3409959077835083,
"learning_rate": 2.4035487845697036e-05,
"loss": 1.3357473373413087,
"step": 1515
},
{
"epoch": 0.6443408223823655,
"grad_norm": 0.2577531635761261,
"learning_rate": 2.378407129167053e-05,
"loss": 1.321412944793701,
"step": 1520
},
{
"epoch": 0.6464603645612548,
"grad_norm": 0.3435700237751007,
"learning_rate": 2.3533419044639137e-05,
"loss": 1.319340133666992,
"step": 1525
},
{
"epoch": 0.6485799067401441,
"grad_norm": 0.3360212743282318,
"learning_rate": 2.328354291862039e-05,
"loss": 1.3458290100097656,
"step": 1530
},
{
"epoch": 0.6506994489190335,
"grad_norm": 0.2735581696033478,
"learning_rate": 2.303445469105076e-05,
"loss": 1.3498548507690429,
"step": 1535
},
{
"epoch": 0.6528189910979229,
"grad_norm": 0.28868424892425537,
"learning_rate": 2.2786166102230726e-05,
"loss": 1.2990142822265625,
"step": 1540
},
{
"epoch": 0.6549385332768122,
"grad_norm": 0.3868809640407562,
"learning_rate": 2.2538688854771232e-05,
"loss": 1.3235331535339356,
"step": 1545
},
{
"epoch": 0.6570580754557016,
"grad_norm": 0.33511966466903687,
"learning_rate": 2.229203461304221e-05,
"loss": 1.3248214721679688,
"step": 1550
},
{
"epoch": 0.6591776176345909,
"grad_norm": 0.2927248477935791,
"learning_rate": 2.204621500262282e-05,
"loss": 1.3123723030090333,
"step": 1555
},
{
"epoch": 0.6612971598134803,
"grad_norm": 0.33650854229927063,
"learning_rate": 2.180124160975341e-05,
"loss": 1.3422239303588868,
"step": 1560
},
{
"epoch": 0.6634167019923697,
"grad_norm": 0.3284781873226166,
"learning_rate": 2.155712598078949e-05,
"loss": 1.2769356727600099,
"step": 1565
},
{
"epoch": 0.665536244171259,
"grad_norm": 0.3051789402961731,
"learning_rate": 2.1313879621657482e-05,
"loss": 1.3215795516967774,
"step": 1570
},
{
"epoch": 0.6676557863501483,
"grad_norm": 0.3087153136730194,
"learning_rate": 2.1071513997312474e-05,
"loss": 1.355457878112793,
"step": 1575
},
{
"epoch": 0.6697753285290378,
"grad_norm": 0.31778043508529663,
"learning_rate": 2.0830040531197744e-05,
"loss": 1.3082546234130858,
"step": 1580
},
{
"epoch": 0.6718948707079271,
"grad_norm": 0.3050025403499603,
"learning_rate": 2.0589470604706384e-05,
"loss": 1.3769618034362794,
"step": 1585
},
{
"epoch": 0.6740144128868164,
"grad_norm": 0.36851558089256287,
"learning_rate": 2.0349815556644917e-05,
"loss": 1.3241981506347655,
"step": 1590
},
{
"epoch": 0.6761339550657058,
"grad_norm": 0.2913540005683899,
"learning_rate": 2.0111086682698753e-05,
"loss": 1.3253866195678712,
"step": 1595
},
{
"epoch": 0.6782534972445952,
"grad_norm": 0.2919890284538269,
"learning_rate": 1.987329523489985e-05,
"loss": 1.279546356201172,
"step": 1600
},
{
"epoch": 0.6803730394234845,
"grad_norm": 0.31500008702278137,
"learning_rate": 1.9636452421096412e-05,
"loss": 1.3314085960388184,
"step": 1605
},
{
"epoch": 0.6824925816023739,
"grad_norm": 0.3655550479888916,
"learning_rate": 1.9400569404424538e-05,
"loss": 1.3648877143859863,
"step": 1610
},
{
"epoch": 0.6846121237812632,
"grad_norm": 0.3227079212665558,
"learning_rate": 1.9165657302782102e-05,
"loss": 1.3001755714416503,
"step": 1615
},
{
"epoch": 0.6867316659601526,
"grad_norm": 0.38053417205810547,
"learning_rate": 1.89317271883048e-05,
"loss": 1.3457537651062013,
"step": 1620
},
{
"epoch": 0.688851208139042,
"grad_norm": 0.2881540060043335,
"learning_rate": 1.8698790086844178e-05,
"loss": 1.3005136489868163,
"step": 1625
},
{
"epoch": 0.6909707503179313,
"grad_norm": 0.3191037178039551,
"learning_rate": 1.8466856977447992e-05,
"loss": 1.3812469482421874,
"step": 1630
},
{
"epoch": 0.6930902924968206,
"grad_norm": 0.2870635986328125,
"learning_rate": 1.8235938791842805e-05,
"loss": 1.301462745666504,
"step": 1635
},
{
"epoch": 0.6952098346757101,
"grad_norm": 0.35562899708747864,
"learning_rate": 1.800604641391861e-05,
"loss": 1.2857603073120116,
"step": 1640
},
{
"epoch": 0.6973293768545994,
"grad_norm": 0.26940205693244934,
"learning_rate": 1.7777190679215923e-05,
"loss": 1.2914587020874024,
"step": 1645
},
{
"epoch": 0.6994489190334887,
"grad_norm": 0.29495933651924133,
"learning_rate": 1.7549382374415043e-05,
"loss": 1.3379831314086914,
"step": 1650
},
{
"epoch": 0.7015684612123781,
"grad_norm": 0.3363819420337677,
"learning_rate": 1.732263223682769e-05,
"loss": 1.3093337059020995,
"step": 1655
},
{
"epoch": 0.7036880033912675,
"grad_norm": 0.30879881978034973,
"learning_rate": 1.7096950953890842e-05,
"loss": 1.3341976165771485,
"step": 1660
},
{
"epoch": 0.7058075455701568,
"grad_norm": 0.3222876191139221,
"learning_rate": 1.6872349162663037e-05,
"loss": 1.3249503135681153,
"step": 1665
},
{
"epoch": 0.7079270877490462,
"grad_norm": 0.2765469253063202,
"learning_rate": 1.6648837449323067e-05,
"loss": 1.3475645065307618,
"step": 1670
},
{
"epoch": 0.7100466299279355,
"grad_norm": 0.36354443430900574,
"learning_rate": 1.6426426348670935e-05,
"loss": 1.279947566986084,
"step": 1675
},
{
"epoch": 0.712166172106825,
"grad_norm": 0.26925837993621826,
"learning_rate": 1.6205126343631363e-05,
"loss": 1.2720775604248047,
"step": 1680
},
{
"epoch": 0.7142857142857143,
"grad_norm": 0.3512694537639618,
"learning_rate": 1.5984947864759702e-05,
"loss": 1.308497428894043,
"step": 1685
},
{
"epoch": 0.7164052564646036,
"grad_norm": 0.29668059945106506,
"learning_rate": 1.5765901289750297e-05,
"loss": 1.3075239181518554,
"step": 1690
},
{
"epoch": 0.7185247986434931,
"grad_norm": 0.34925898909568787,
"learning_rate": 1.5547996942947333e-05,
"loss": 1.3636594772338868,
"step": 1695
},
{
"epoch": 0.7206443408223824,
"grad_norm": 0.270380437374115,
"learning_rate": 1.5331245094858277e-05,
"loss": 1.3095337867736816,
"step": 1700
},
{
"epoch": 0.7227638830012717,
"grad_norm": 0.2901579737663269,
"learning_rate": 1.511565596166973e-05,
"loss": 1.3417037010192872,
"step": 1705
},
{
"epoch": 0.7248834251801611,
"grad_norm": 0.34968939423561096,
"learning_rate": 1.4901239704765917e-05,
"loss": 1.3205747604370117,
"step": 1710
},
{
"epoch": 0.7270029673590505,
"grad_norm": 0.27065205574035645,
"learning_rate": 1.4688006430249825e-05,
"loss": 1.335763645172119,
"step": 1715
},
{
"epoch": 0.7291225095379398,
"grad_norm": 0.34851571917533875,
"learning_rate": 1.447596618846677e-05,
"loss": 1.287162971496582,
"step": 1720
},
{
"epoch": 0.7312420517168292,
"grad_norm": 0.3229304552078247,
"learning_rate": 1.4265128973530731e-05,
"loss": 1.3325061798095703,
"step": 1725
},
{
"epoch": 0.7333615938957185,
"grad_norm": 0.33570748567581177,
"learning_rate": 1.4055504722853339e-05,
"loss": 1.3185012817382813,
"step": 1730
},
{
"epoch": 0.7354811360746079,
"grad_norm": 0.29018643498420715,
"learning_rate": 1.384710331667543e-05,
"loss": 1.3453382492065429,
"step": 1735
},
{
"epoch": 0.7376006782534973,
"grad_norm": 0.30483657121658325,
"learning_rate": 1.3639934577601395e-05,
"loss": 1.3572277069091796,
"step": 1740
},
{
"epoch": 0.7397202204323866,
"grad_norm": 0.28786394000053406,
"learning_rate": 1.3434008270136185e-05,
"loss": 1.3255316734313964,
"step": 1745
},
{
"epoch": 0.7418397626112759,
"grad_norm": 0.3231492340564728,
"learning_rate": 1.3229334100225147e-05,
"loss": 1.3233895301818848,
"step": 1750
},
{
"epoch": 0.7439593047901654,
"grad_norm": 0.27897658944129944,
"learning_rate": 1.3025921714796455e-05,
"loss": 1.3510261535644532,
"step": 1755
},
{
"epoch": 0.7460788469690547,
"grad_norm": 0.2828425467014313,
"learning_rate": 1.2823780701306468e-05,
"loss": 1.357948112487793,
"step": 1760
},
{
"epoch": 0.748198389147944,
"grad_norm": 0.29348132014274597,
"learning_rate": 1.2622920587287904e-05,
"loss": 1.2803407669067384,
"step": 1765
},
{
"epoch": 0.7503179313268334,
"grad_norm": 0.2772983908653259,
"learning_rate": 1.2423350839900615e-05,
"loss": 1.3109886169433593,
"step": 1770
},
{
"epoch": 0.7524374735057228,
"grad_norm": 0.26257404685020447,
"learning_rate": 1.2225080865485582e-05,
"loss": 1.2990903854370117,
"step": 1775
},
{
"epoch": 0.7545570156846121,
"grad_norm": 0.24683360755443573,
"learning_rate": 1.2028120009121382e-05,
"loss": 1.319027042388916,
"step": 1780
},
{
"epoch": 0.7566765578635015,
"grad_norm": 0.2640559673309326,
"learning_rate": 1.1832477554183828e-05,
"loss": 1.3019865989685058,
"step": 1785
},
{
"epoch": 0.7587961000423908,
"grad_norm": 0.2786967158317566,
"learning_rate": 1.1638162721908369e-05,
"loss": 1.3104122161865235,
"step": 1790
},
{
"epoch": 0.7609156422212802,
"grad_norm": 0.3318904936313629,
"learning_rate": 1.1445184670955522e-05,
"loss": 1.3181925773620606,
"step": 1795
},
{
"epoch": 0.7630351844001696,
"grad_norm": 0.28481072187423706,
"learning_rate": 1.1253552496979111e-05,
"loss": 1.3262067794799806,
"step": 1800
},
{
"epoch": 0.7651547265790589,
"grad_norm": 0.292675256729126,
"learning_rate": 1.1063275232197617e-05,
"loss": 1.3193824768066407,
"step": 1805
},
{
"epoch": 0.7672742687579482,
"grad_norm": 0.48738452792167664,
"learning_rate": 1.0874361844968471e-05,
"loss": 1.254288673400879,
"step": 1810
},
{
"epoch": 0.7693938109368377,
"grad_norm": 0.2857394516468048,
"learning_rate": 1.0686821239365304e-05,
"loss": 1.3502561569213867,
"step": 1815
},
{
"epoch": 0.771513353115727,
"grad_norm": 0.2841188311576843,
"learning_rate": 1.050066225475829e-05,
"loss": 1.3571770668029786,
"step": 1820
},
{
"epoch": 0.7736328952946163,
"grad_norm": 0.25839078426361084,
"learning_rate": 1.0315893665397553e-05,
"loss": 1.2797527313232422,
"step": 1825
},
{
"epoch": 0.7757524374735058,
"grad_norm": 0.3031015992164612,
"learning_rate": 1.0132524179999549e-05,
"loss": 1.286862277984619,
"step": 1830
},
{
"epoch": 0.7778719796523951,
"grad_norm": 0.27194085717201233,
"learning_rate": 9.950562441336617e-06,
"loss": 1.3276911735534669,
"step": 1835
},
{
"epoch": 0.7799915218312844,
"grad_norm": 0.26917746663093567,
"learning_rate": 9.770017025829675e-06,
"loss": 1.325631046295166,
"step": 1840
},
{
"epoch": 0.7821110640101738,
"grad_norm": 0.3251597285270691,
"learning_rate": 9.590896443143891e-06,
"loss": 1.3492204666137695,
"step": 1845
},
{
"epoch": 0.7842306061890632,
"grad_norm": 0.2875700294971466,
"learning_rate": 9.413209135787662e-06,
"loss": 1.3610807418823243,
"step": 1850
},
{
"epoch": 0.7863501483679525,
"grad_norm": 0.3361499607563019,
"learning_rate": 9.23696347871469e-06,
"loss": 1.247593879699707,
"step": 1855
},
{
"epoch": 0.7884696905468419,
"grad_norm": 0.28728315234184265,
"learning_rate": 9.062167778929214e-06,
"loss": 1.327476692199707,
"step": 1860
},
{
"epoch": 0.7905892327257312,
"grad_norm": 0.2754060924053192,
"learning_rate": 8.888830275094476e-06,
"loss": 1.2668123245239258,
"step": 1865
},
{
"epoch": 0.7927087749046206,
"grad_norm": 0.2830967605113983,
"learning_rate": 8.716959137144462e-06,
"loss": 1.3329549789428712,
"step": 1870
},
{
"epoch": 0.79482831708351,
"grad_norm": 0.33522310853004456,
"learning_rate": 8.546562465898756e-06,
"loss": 1.2876480102539063,
"step": 1875
},
{
"epoch": 0.7969478592623993,
"grad_norm": 0.3227211534976959,
"learning_rate": 8.37764829268076e-06,
"loss": 1.2873859405517578,
"step": 1880
},
{
"epoch": 0.7990674014412886,
"grad_norm": 0.2690141797065735,
"learning_rate": 8.210224578939146e-06,
"loss": 1.281691551208496,
"step": 1885
},
{
"epoch": 0.8011869436201781,
"grad_norm": 0.3446494936943054,
"learning_rate": 8.044299215872629e-06,
"loss": 1.3154401779174805,
"step": 1890
},
{
"epoch": 0.8033064857990674,
"grad_norm": 0.28599879145622253,
"learning_rate": 7.879880024057995e-06,
"loss": 1.2795836448669433,
"step": 1895
},
{
"epoch": 0.8054260279779568,
"grad_norm": 0.37984657287597656,
"learning_rate": 7.716974753081494e-06,
"loss": 1.249599075317383,
"step": 1900
},
{
"epoch": 0.8075455701568461,
"grad_norm": 0.3835977613925934,
"learning_rate": 7.555591081173634e-06,
"loss": 1.3208285331726075,
"step": 1905
},
{
"epoch": 0.8096651123357355,
"grad_norm": 0.28392940759658813,
"learning_rate": 7.3957366148472085e-06,
"loss": 1.3550309181213378,
"step": 1910
},
{
"epoch": 0.8117846545146249,
"grad_norm": 0.2897721230983734,
"learning_rate": 7.237418888538812e-06,
"loss": 1.2971899032592773,
"step": 1915
},
{
"epoch": 0.8139041966935142,
"grad_norm": 0.39264678955078125,
"learning_rate": 7.080645364253747e-06,
"loss": 1.361825180053711,
"step": 1920
},
{
"epoch": 0.8160237388724035,
"grad_norm": 0.32630112767219543,
"learning_rate": 6.925423431214251e-06,
"loss": 1.3256078720092774,
"step": 1925
},
{
"epoch": 0.818143281051293,
"grad_norm": 0.3046013116836548,
"learning_rate": 6.771760405511263e-06,
"loss": 1.2435759544372558,
"step": 1930
},
{
"epoch": 0.8202628232301823,
"grad_norm": 0.30348455905914307,
"learning_rate": 6.6196635297596146e-06,
"loss": 1.2791961669921874,
"step": 1935
},
{
"epoch": 0.8223823654090716,
"grad_norm": 0.3114020824432373,
"learning_rate": 6.469139972756604e-06,
"loss": 1.3734837532043458,
"step": 1940
},
{
"epoch": 0.824501907587961,
"grad_norm": 0.33188411593437195,
"learning_rate": 6.3201968291441365e-06,
"loss": 1.3239216804504395,
"step": 1945
},
{
"epoch": 0.8266214497668504,
"grad_norm": 0.32971006631851196,
"learning_rate": 6.172841119074369e-06,
"loss": 1.3304693222045898,
"step": 1950
},
{
"epoch": 0.8287409919457397,
"grad_norm": 0.26540112495422363,
"learning_rate": 6.027079787878763e-06,
"loss": 1.3449939727783202,
"step": 1955
},
{
"epoch": 0.8308605341246291,
"grad_norm": 0.44900065660476685,
"learning_rate": 5.882919705740766e-06,
"loss": 1.3049022674560546,
"step": 1960
},
{
"epoch": 0.8329800763035184,
"grad_norm": 0.2551844120025635,
"learning_rate": 5.740367667371991e-06,
"loss": 1.3034229278564453,
"step": 1965
},
{
"epoch": 0.8350996184824078,
"grad_norm": 0.3436892628669739,
"learning_rate": 5.5994303916919776e-06,
"loss": 1.3763395309448243,
"step": 1970
},
{
"epoch": 0.8372191606612972,
"grad_norm": 0.2672841548919678,
"learning_rate": 5.460114521511486e-06,
"loss": 1.3012068748474122,
"step": 1975
},
{
"epoch": 0.8393387028401865,
"grad_norm": 0.28498658537864685,
"learning_rate": 5.322426623219392e-06,
"loss": 1.2495477676391602,
"step": 1980
},
{
"epoch": 0.8414582450190758,
"grad_norm": 0.23883037269115448,
"learning_rate": 5.186373186473237e-06,
"loss": 1.2991844177246095,
"step": 1985
},
{
"epoch": 0.8435777871979653,
"grad_norm": 0.2611273229122162,
"learning_rate": 5.051960623893308e-06,
"loss": 1.2900188446044922,
"step": 1990
},
{
"epoch": 0.8456973293768546,
"grad_norm": 0.2825237810611725,
"learning_rate": 4.919195270760386e-06,
"loss": 1.2969111442565917,
"step": 1995
},
{
"epoch": 0.8478168715557439,
"grad_norm": 0.31668326258659363,
"learning_rate": 4.788083384717212e-06,
"loss": 1.2768826484680176,
"step": 2000
},
{
"epoch": 0.8499364137346334,
"grad_norm": 0.29829099774360657,
"learning_rate": 4.658631145473451e-06,
"loss": 1.3164323806762694,
"step": 2005
},
{
"epoch": 0.8520559559135227,
"grad_norm": 0.2907348573207855,
"learning_rate": 4.530844654514495e-06,
"loss": 1.335688591003418,
"step": 2010
},
{
"epoch": 0.854175498092412,
"grad_norm": 0.3484869599342346,
"learning_rate": 4.404729934813836e-06,
"loss": 1.3329750061035157,
"step": 2015
},
{
"epoch": 0.8562950402713014,
"grad_norm": 0.36669397354125977,
"learning_rate": 4.280292930549217e-06,
"loss": 1.2887075424194336,
"step": 2020
},
{
"epoch": 0.8584145824501908,
"grad_norm": 0.3710174858570099,
"learning_rate": 4.157539506822446e-06,
"loss": 1.3391199111938477,
"step": 2025
},
{
"epoch": 0.8605341246290801,
"grad_norm": 0.29658403992652893,
"learning_rate": 4.036475449382966e-06,
"loss": 1.3254976272583008,
"step": 2030
},
{
"epoch": 0.8626536668079695,
"grad_norm": 0.3214455246925354,
"learning_rate": 3.917106464355147e-06,
"loss": 1.250115966796875,
"step": 2035
},
{
"epoch": 0.8647732089868588,
"grad_norm": 0.32465845346450806,
"learning_rate": 3.799438177969328e-06,
"loss": 1.3117265701293945,
"step": 2040
},
{
"epoch": 0.8668927511657482,
"grad_norm": 0.2953072190284729,
"learning_rate": 3.68347613629668e-06,
"loss": 1.3151223182678222,
"step": 2045
},
{
"epoch": 0.8690122933446376,
"grad_norm": 0.2593151330947876,
"learning_rate": 3.5692258049877303e-06,
"loss": 1.2428454399108886,
"step": 2050
},
{
"epoch": 0.8711318355235269,
"grad_norm": 0.3140961527824402,
"learning_rate": 3.4566925690148145e-06,
"loss": 1.3111312866210938,
"step": 2055
},
{
"epoch": 0.8732513777024162,
"grad_norm": 0.3819583058357239,
"learning_rate": 3.3458817324182456e-06,
"loss": 1.3091193199157716,
"step": 2060
},
{
"epoch": 0.8753709198813057,
"grad_norm": 0.34935933351516724,
"learning_rate": 3.236798518056303e-06,
"loss": 1.3158819198608398,
"step": 2065
},
{
"epoch": 0.877490462060195,
"grad_norm": 0.28201058506965637,
"learning_rate": 3.12944806735906e-06,
"loss": 1.3029821395874024,
"step": 2070
},
{
"epoch": 0.8796100042390843,
"grad_norm": 0.27294135093688965,
"learning_rate": 3.0238354400861005e-06,
"loss": 1.2982366561889649,
"step": 2075
},
{
"epoch": 0.8817295464179737,
"grad_norm": 0.27961230278015137,
"learning_rate": 2.919965614087983e-06,
"loss": 1.3389122009277343,
"step": 2080
},
{
"epoch": 0.8838490885968631,
"grad_norm": 0.27565574645996094,
"learning_rate": 2.817843485071623e-06,
"loss": 1.2884050369262696,
"step": 2085
},
{
"epoch": 0.8859686307757524,
"grad_norm": 0.2931448817253113,
"learning_rate": 2.717473866369593e-06,
"loss": 1.2905622482299806,
"step": 2090
},
{
"epoch": 0.8880881729546418,
"grad_norm": 0.2940874397754669,
"learning_rate": 2.6188614887132023e-06,
"loss": 1.328693962097168,
"step": 2095
},
{
"epoch": 0.8902077151335311,
"grad_norm": 0.3092103898525238,
"learning_rate": 2.5220110000095366e-06,
"loss": 1.3134818077087402,
"step": 2100
},
{
"epoch": 0.8923272573124205,
"grad_norm": 0.27005988359451294,
"learning_rate": 2.4269269651223936e-06,
"loss": 1.2883139610290528,
"step": 2105
},
{
"epoch": 0.8944467994913099,
"grad_norm": 0.32885122299194336,
"learning_rate": 2.333613865657145e-06,
"loss": 1.3413027763366698,
"step": 2110
},
{
"epoch": 0.8965663416701992,
"grad_norm": 0.36202335357666016,
"learning_rate": 2.242076099749464e-06,
"loss": 1.2506613731384277,
"step": 2115
},
{
"epoch": 0.8986858838490887,
"grad_norm": 0.33259114623069763,
"learning_rate": 2.1523179818580563e-06,
"loss": 1.3003890991210938,
"step": 2120
},
{
"epoch": 0.900805426027978,
"grad_norm": 0.30203136801719666,
"learning_rate": 2.064343742561308e-06,
"loss": 1.2926207542419434,
"step": 2125
},
{
"epoch": 0.9029249682068673,
"grad_norm": 0.38430875539779663,
"learning_rate": 1.9781575283578647e-06,
"loss": 1.313484001159668,
"step": 2130
},
{
"epoch": 0.9050445103857567,
"grad_norm": 0.31734201312065125,
"learning_rate": 1.8937634014711958e-06,
"loss": 1.267928123474121,
"step": 2135
},
{
"epoch": 0.907164052564646,
"grad_norm": 0.25555816292762756,
"learning_rate": 1.8111653396581762e-06,
"loss": 1.246608352661133,
"step": 2140
},
{
"epoch": 0.9092835947435354,
"grad_norm": 0.32622987031936646,
"learning_rate": 1.7303672360215262e-06,
"loss": 1.2804512977600098,
"step": 2145
},
{
"epoch": 0.9114031369224248,
"grad_norm": 0.3292995095252991,
"learning_rate": 1.6513728988263754e-06,
"loss": 1.287643814086914,
"step": 2150
},
{
"epoch": 0.9135226791013141,
"grad_norm": 0.3014987111091614,
"learning_rate": 1.574186051320754e-06,
"loss": 1.3262786865234375,
"step": 2155
},
{
"epoch": 0.9156422212802034,
"grad_norm": 0.3536871671676636,
"learning_rate": 1.4988103315600832e-06,
"loss": 1.2743340492248536,
"step": 2160
},
{
"epoch": 0.9177617634590929,
"grad_norm": 0.3734804689884186,
"learning_rate": 1.4252492922357219e-06,
"loss": 1.2381834983825684,
"step": 2165
},
{
"epoch": 0.9198813056379822,
"grad_norm": 0.3409561216831207,
"learning_rate": 1.353506400507536e-06,
"loss": 1.312345314025879,
"step": 2170
},
{
"epoch": 0.9220008478168715,
"grad_norm": 0.31126758456230164,
"learning_rate": 1.2835850378404247e-06,
"loss": 1.2810562133789063,
"step": 2175
},
{
"epoch": 0.924120389995761,
"grad_norm": 0.3201841413974762,
"learning_rate": 1.215488499844999e-06,
"loss": 1.3295242309570312,
"step": 2180
},
{
"epoch": 0.9262399321746503,
"grad_norm": 0.31913650035858154,
"learning_rate": 1.1492199961222127e-06,
"loss": 1.3159330368041993,
"step": 2185
},
{
"epoch": 0.9283594743535396,
"grad_norm": 0.27366265654563904,
"learning_rate": 1.0847826501121062e-06,
"loss": 1.3026064872741698,
"step": 2190
},
{
"epoch": 0.930479016532429,
"grad_norm": 0.3409899175167084,
"learning_rate": 1.0221794989465761e-06,
"loss": 1.3022947311401367,
"step": 2195
},
{
"epoch": 0.9325985587113184,
"grad_norm": 0.36054855585098267,
"learning_rate": 9.614134933062201e-07,
"loss": 1.3410823822021485,
"step": 2200
},
{
"epoch": 0.9347181008902077,
"grad_norm": 0.2941698431968689,
"learning_rate": 9.024874972812969e-07,
"loss": 1.3064786911010742,
"step": 2205
},
{
"epoch": 0.9368376430690971,
"grad_norm": 0.3120056986808777,
"learning_rate": 8.45404288236673e-07,
"loss": 1.3244601249694825,
"step": 2210
},
{
"epoch": 0.9389571852479864,
"grad_norm": 0.3773142099380493,
"learning_rate": 7.90166556680978e-07,
"loss": 1.308661651611328,
"step": 2215
},
{
"epoch": 0.9410767274268758,
"grad_norm": 0.31075984239578247,
"learning_rate": 7.367769061397534e-07,
"loss": 1.27926607131958,
"step": 2220
},
{
"epoch": 0.9431962696057652,
"grad_norm": 0.3036041557788849,
"learning_rate": 6.852378530327608e-07,
"loss": 1.3526771545410157,
"step": 2225
},
{
"epoch": 0.9453158117846545,
"grad_norm": 0.3305091857910156,
"learning_rate": 6.355518265553561e-07,
"loss": 1.3071530342102051,
"step": 2230
},
{
"epoch": 0.9474353539635438,
"grad_norm": 0.33323293924331665,
"learning_rate": 5.877211685640127e-07,
"loss": 1.2693525314331056,
"step": 2235
},
{
"epoch": 0.9495548961424333,
"grad_norm": 0.2906723916530609,
"learning_rate": 5.41748133465938e-07,
"loss": 1.306605339050293,
"step": 2240
},
{
"epoch": 0.9516744383213226,
"grad_norm": 0.3504180312156677,
"learning_rate": 4.976348881128034e-07,
"loss": 1.3358493804931642,
"step": 2245
},
{
"epoch": 0.9537939805002119,
"grad_norm": 0.33575376868247986,
"learning_rate": 4.553835116986349e-07,
"loss": 1.2731388092041016,
"step": 2250
},
{
"epoch": 0.9559135226791013,
"grad_norm": 0.29073020815849304,
"learning_rate": 4.1499599566178883e-07,
"loss": 1.3242993354797363,
"step": 2255
},
{
"epoch": 0.9580330648579907,
"grad_norm": 0.2980799376964569,
"learning_rate": 3.7647424359111615e-07,
"loss": 1.3340805053710938,
"step": 2260
},
{
"epoch": 0.96015260703688,
"grad_norm": 0.2866630256175995,
"learning_rate": 3.39820071136221e-07,
"loss": 1.3176158905029296,
"step": 2265
},
{
"epoch": 0.9622721492157694,
"grad_norm": 0.3219427466392517,
"learning_rate": 3.050352059218975e-07,
"loss": 1.3003140449523927,
"step": 2270
},
{
"epoch": 0.9643916913946587,
"grad_norm": 0.31648269295692444,
"learning_rate": 2.7212128746668454e-07,
"loss": 1.278752326965332,
"step": 2275
},
{
"epoch": 0.9665112335735481,
"grad_norm": 0.3372032940387726,
"learning_rate": 2.410798671056025e-07,
"loss": 1.3539626121520996,
"step": 2280
},
{
"epoch": 0.9686307757524375,
"grad_norm": 0.4131452143192291,
"learning_rate": 2.1191240791703426e-07,
"loss": 1.2815422058105468,
"step": 2285
},
{
"epoch": 0.9707503179313268,
"grad_norm": 0.27749812602996826,
"learning_rate": 1.8462028465375815e-07,
"loss": 1.3147613525390625,
"step": 2290
},
{
"epoch": 0.9728698601102161,
"grad_norm": 0.2816482484340668,
"learning_rate": 1.5920478367816405e-07,
"loss": 1.3334857940673828,
"step": 2295
},
{
"epoch": 0.9749894022891056,
"grad_norm": 0.32452377676963806,
"learning_rate": 1.3566710290160878e-07,
"loss": 1.3083940505981446,
"step": 2300
},
{
"epoch": 0.9771089444679949,
"grad_norm": 0.34873369336128235,
"learning_rate": 1.1400835172795444e-07,
"loss": 1.3483657836914062,
"step": 2305
},
{
"epoch": 0.9792284866468842,
"grad_norm": 0.40136268734931946,
"learning_rate": 9.42295510013036e-08,
"loss": 1.2984387397766113,
"step": 2310
},
{
"epoch": 0.9813480288257737,
"grad_norm": 0.34510400891304016,
"learning_rate": 7.633163295785118e-08,
"loss": 1.3370220184326171,
"step": 2315
},
{
"epoch": 0.983467571004663,
"grad_norm": 0.31063640117645264,
"learning_rate": 6.031544118196398e-08,
"loss": 1.3359445571899413,
"step": 2320
},
{
"epoch": 0.9855871131835523,
"grad_norm": 0.286815881729126,
"learning_rate": 4.618173056641695e-08,
"loss": 1.299731731414795,
"step": 2325
},
{
"epoch": 0.9877066553624417,
"grad_norm": 0.2510542571544647,
"learning_rate": 3.393116727681278e-08,
"loss": 1.2871533393859864,
"step": 2330
},
{
"epoch": 0.989826197541331,
"grad_norm": 0.29293227195739746,
"learning_rate": 2.3564328720175887e-08,
"loss": 1.3614470481872558,
"step": 2335
},
{
"epoch": 0.9919457397202205,
"grad_norm": 0.3270072638988495,
"learning_rate": 1.5081703517751956e-08,
"loss": 1.3502976417541503,
"step": 2340
},
{
"epoch": 0.9940652818991098,
"grad_norm": 0.3049931824207306,
"learning_rate": 8.483691481968592e-09,
"loss": 1.328630256652832,
"step": 2345
},
{
"epoch": 0.9961848240779991,
"grad_norm": 0.2985376715660095,
"learning_rate": 3.770603597583744e-09,
"loss": 1.3159720420837402,
"step": 2350
},
{
"epoch": 0.9983043662568886,
"grad_norm": 0.3223288059234619,
"learning_rate": 9.426620070396298e-10,
"loss": 1.3115079879760743,
"step": 2355
}
],
"logging_steps": 5,
"max_steps": 2359,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.7440308287438848e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}