VibaoAI / trainer_state.json
Toilatop1sever's picture
checkpoint step 46000 (khẩn cấp - hết quota GPU)
532f820 verified
Raw
History Blame Contribute Delete
177 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.7732546933425852,
"eval_steps": 2000,
"global_step": 46000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0019274507536332446,
"grad_norm": 1.2952269315719604,
"learning_rate": 1.47e-05,
"loss": 9.958944091796875,
"step": 50
},
{
"epoch": 0.0038549015072664892,
"grad_norm": 1.0338493585586548,
"learning_rate": 2.97e-05,
"loss": 8.95384521484375,
"step": 100
},
{
"epoch": 0.005782352260899734,
"grad_norm": 0.6325502395629883,
"learning_rate": 4.4699999999999996e-05,
"loss": 7.8542041015625,
"step": 150
},
{
"epoch": 0.0077098030145329784,
"grad_norm": 1.2084871530532837,
"learning_rate": 5.97e-05,
"loss": 7.141786499023437,
"step": 200
},
{
"epoch": 0.009637253768166223,
"grad_norm": 0.7696707248687744,
"learning_rate": 7.47e-05,
"loss": 6.751356201171875,
"step": 250
},
{
"epoch": 0.011564704521799469,
"grad_norm": 0.6730088591575623,
"learning_rate": 8.969999999999998e-05,
"loss": 6.432244873046875,
"step": 300
},
{
"epoch": 0.013492155275432712,
"grad_norm": 0.6428735852241516,
"learning_rate": 0.00010469999999999998,
"loss": 6.174566040039062,
"step": 350
},
{
"epoch": 0.015419606029065957,
"grad_norm": 0.7519970536231995,
"learning_rate": 0.0001197,
"loss": 5.961318359375,
"step": 400
},
{
"epoch": 0.017347056782699204,
"grad_norm": 0.6962649822235107,
"learning_rate": 0.0001347,
"loss": 5.794595947265625,
"step": 450
},
{
"epoch": 0.019274507536332447,
"grad_norm": 0.6648260951042175,
"learning_rate": 0.00014969999999999998,
"loss": 5.688720703125,
"step": 500
},
{
"epoch": 0.02120195828996569,
"grad_norm": 0.7536967992782593,
"learning_rate": 0.0001647,
"loss": 5.573690185546875,
"step": 550
},
{
"epoch": 0.023129409043598937,
"grad_norm": 0.8137773275375366,
"learning_rate": 0.00017969999999999998,
"loss": 5.488543090820312,
"step": 600
},
{
"epoch": 0.02505685979723218,
"grad_norm": 0.666891872882843,
"learning_rate": 0.0001947,
"loss": 5.3848114013671875,
"step": 650
},
{
"epoch": 0.026984310550865424,
"grad_norm": 0.5968290567398071,
"learning_rate": 0.00020969999999999997,
"loss": 5.3349072265625,
"step": 700
},
{
"epoch": 0.02891176130449867,
"grad_norm": 0.6266096234321594,
"learning_rate": 0.0002247,
"loss": 5.26029296875,
"step": 750
},
{
"epoch": 0.030839212058131914,
"grad_norm": 0.5615338683128357,
"learning_rate": 0.0002397,
"loss": 5.207418212890625,
"step": 800
},
{
"epoch": 0.03276666281176516,
"grad_norm": 0.5597893595695496,
"learning_rate": 0.00025469999999999996,
"loss": 5.14937255859375,
"step": 850
},
{
"epoch": 0.03469411356539841,
"grad_norm": 0.6763906478881836,
"learning_rate": 0.0002697,
"loss": 5.0670947265625,
"step": 900
},
{
"epoch": 0.03662156431903165,
"grad_norm": 0.612106204032898,
"learning_rate": 0.0002847,
"loss": 5.014092712402344,
"step": 950
},
{
"epoch": 0.038549015072664894,
"grad_norm": 0.5845397710800171,
"learning_rate": 0.00029969999999999997,
"loss": 4.931963500976562,
"step": 1000
},
{
"epoch": 0.04047646582629814,
"grad_norm": 0.5691105723381042,
"learning_rate": 0.00029999925978027874,
"loss": 4.883682861328125,
"step": 1050
},
{
"epoch": 0.04240391657993138,
"grad_norm": 0.553705632686615,
"learning_rate": 0.0002999969783943782,
"loss": 4.806473999023438,
"step": 1100
},
{
"epoch": 0.04433136733356463,
"grad_norm": 0.5707918405532837,
"learning_rate": 0.0002999931555574465,
"loss": 4.74484619140625,
"step": 1150
},
{
"epoch": 0.046258818087197874,
"grad_norm": 0.5924341678619385,
"learning_rate": 0.00029998779130876926,
"loss": 4.702532958984375,
"step": 1200
},
{
"epoch": 0.04818626884083112,
"grad_norm": 0.5594196915626526,
"learning_rate": 0.00029998088570347235,
"loss": 4.645215454101563,
"step": 1250
},
{
"epoch": 0.05011371959446436,
"grad_norm": 0.5578745603561401,
"learning_rate": 0.00029997243881252166,
"loss": 4.573051147460937,
"step": 1300
},
{
"epoch": 0.052041170348097604,
"grad_norm": 0.582054853439331,
"learning_rate": 0.0002999624507227219,
"loss": 4.53611083984375,
"step": 1350
},
{
"epoch": 0.05396862110173085,
"grad_norm": 0.5916629433631897,
"learning_rate": 0.0002999509215367163,
"loss": 4.50299560546875,
"step": 1400
},
{
"epoch": 0.0558960718553641,
"grad_norm": 0.5494801998138428,
"learning_rate": 0.000299937851372985,
"loss": 4.456146850585937,
"step": 1450
},
{
"epoch": 0.05782352260899734,
"grad_norm": 0.6153786182403564,
"learning_rate": 0.00029992324036584415,
"loss": 4.419460144042969,
"step": 1500
},
{
"epoch": 0.059750973362630584,
"grad_norm": 0.5355115532875061,
"learning_rate": 0.0002999070886654445,
"loss": 4.38196044921875,
"step": 1550
},
{
"epoch": 0.06167842411626383,
"grad_norm": 0.5394632816314697,
"learning_rate": 0.0002998893964377698,
"loss": 4.343935546875,
"step": 1600
},
{
"epoch": 0.06360587486989708,
"grad_norm": 0.5479700565338135,
"learning_rate": 0.0002998701638646349,
"loss": 4.330123291015625,
"step": 1650
},
{
"epoch": 0.06553332562353031,
"grad_norm": 0.5667453408241272,
"learning_rate": 0.0002998493911436844,
"loss": 4.30684326171875,
"step": 1700
},
{
"epoch": 0.06746077637716356,
"grad_norm": 0.5475571751594543,
"learning_rate": 0.0002998270784883901,
"loss": 4.268912963867187,
"step": 1750
},
{
"epoch": 0.06938822713079681,
"grad_norm": 0.5723440647125244,
"learning_rate": 0.0002998032261280491,
"loss": 4.243427734375,
"step": 1800
},
{
"epoch": 0.07131567788443005,
"grad_norm": 0.5883432030677795,
"learning_rate": 0.00029977783430778116,
"loss": 4.210003051757813,
"step": 1850
},
{
"epoch": 0.0732431286380633,
"grad_norm": 0.5907633304595947,
"learning_rate": 0.0002997509032885267,
"loss": 4.182230224609375,
"step": 1900
},
{
"epoch": 0.07517057939169654,
"grad_norm": 0.5539486408233643,
"learning_rate": 0.0002997224333470435,
"loss": 4.142240600585938,
"step": 1950
},
{
"epoch": 0.07709803014532979,
"grad_norm": 0.5779871344566345,
"learning_rate": 0.00029969242477590434,
"loss": 4.125773315429687,
"step": 2000
},
{
"epoch": 0.07709803014532979,
"eval_loss": 4.046743392944336,
"eval_runtime": 71.4471,
"eval_samples_per_second": 116.758,
"eval_steps_per_second": 3.653,
"step": 2000
},
{
"epoch": 0.07902548089896304,
"grad_norm": 0.5690900087356567,
"learning_rate": 0.0002996608778834936,
"loss": 4.125428771972656,
"step": 2050
},
{
"epoch": 0.08095293165259627,
"grad_norm": 0.5566129684448242,
"learning_rate": 0.0002996277929940045,
"loss": 4.076891784667969,
"step": 2100
},
{
"epoch": 0.08288038240622952,
"grad_norm": 0.5267688632011414,
"learning_rate": 0.0002995931704474355,
"loss": 4.043207397460938,
"step": 2150
},
{
"epoch": 0.08480783315986276,
"grad_norm": 0.5993160605430603,
"learning_rate": 0.00029955701059958674,
"loss": 4.027578735351563,
"step": 2200
},
{
"epoch": 0.08673528391349601,
"grad_norm": 0.5521589517593384,
"learning_rate": 0.0002995193138220566,
"loss": 4.025603637695313,
"step": 2250
},
{
"epoch": 0.08866273466712926,
"grad_norm": 0.5826160907745361,
"learning_rate": 0.00029948008050223793,
"loss": 4.002682495117187,
"step": 2300
},
{
"epoch": 0.0905901854207625,
"grad_norm": 0.5729931592941284,
"learning_rate": 0.00029943931104331364,
"loss": 3.971737976074219,
"step": 2350
},
{
"epoch": 0.09251763617439575,
"grad_norm": 0.5670749545097351,
"learning_rate": 0.0002993970058642531,
"loss": 3.952825012207031,
"step": 2400
},
{
"epoch": 0.09444508692802898,
"grad_norm": 0.5358520746231079,
"learning_rate": 0.00029935316539980744,
"loss": 3.9312689208984377,
"step": 2450
},
{
"epoch": 0.09637253768166223,
"grad_norm": 0.5385018587112427,
"learning_rate": 0.00029930779010050527,
"loss": 3.9208279418945313,
"step": 2500
},
{
"epoch": 0.09829998843529549,
"grad_norm": 0.5354301929473877,
"learning_rate": 0.000299260880432648,
"loss": 3.8972625732421875,
"step": 2550
},
{
"epoch": 0.10022743918892872,
"grad_norm": 0.5218163728713989,
"learning_rate": 0.000299212436878305,
"loss": 3.9156268310546873,
"step": 2600
},
{
"epoch": 0.10215488994256197,
"grad_norm": 0.5363887548446655,
"learning_rate": 0.00029916245993530876,
"loss": 3.8841168212890627,
"step": 2650
},
{
"epoch": 0.10408234069619521,
"grad_norm": 0.5277280211448669,
"learning_rate": 0.0002991109501172498,
"loss": 3.8729959106445313,
"step": 2700
},
{
"epoch": 0.10600979144982846,
"grad_norm": 0.5406108498573303,
"learning_rate": 0.00029905790795347107,
"loss": 3.84861083984375,
"step": 2750
},
{
"epoch": 0.1079372422034617,
"grad_norm": 0.5141267776489258,
"learning_rate": 0.00029900333398906297,
"loss": 3.8415374755859375,
"step": 2800
},
{
"epoch": 0.10986469295709494,
"grad_norm": 0.5229519009590149,
"learning_rate": 0.0002989472287848573,
"loss": 3.8388543701171876,
"step": 2850
},
{
"epoch": 0.1117921437107282,
"grad_norm": 0.5372052192687988,
"learning_rate": 0.0002988895929174219,
"loss": 3.84042724609375,
"step": 2900
},
{
"epoch": 0.11371959446436143,
"grad_norm": 0.5206784605979919,
"learning_rate": 0.0002988304269790545,
"loss": 3.8165792846679687,
"step": 2950
},
{
"epoch": 0.11564704521799468,
"grad_norm": 0.5360506772994995,
"learning_rate": 0.0002987697315777767,
"loss": 3.788231201171875,
"step": 3000
},
{
"epoch": 0.11757449597162792,
"grad_norm": 0.5505446195602417,
"learning_rate": 0.00029870750733732764,
"loss": 3.8016390991210938,
"step": 3050
},
{
"epoch": 0.11950194672526117,
"grad_norm": 0.4984762966632843,
"learning_rate": 0.00029864375489715755,
"loss": 3.785032043457031,
"step": 3100
},
{
"epoch": 0.12142939747889442,
"grad_norm": 0.5347452759742737,
"learning_rate": 0.0002985784749124215,
"loss": 3.75969970703125,
"step": 3150
},
{
"epoch": 0.12335684823252765,
"grad_norm": 0.5266030430793762,
"learning_rate": 0.00029851166805397233,
"loss": 3.78552978515625,
"step": 3200
},
{
"epoch": 0.1252842989861609,
"grad_norm": 0.4952186346054077,
"learning_rate": 0.0002984433350083538,
"loss": 3.7509979248046874,
"step": 3250
},
{
"epoch": 0.12721174973979416,
"grad_norm": 0.5166611075401306,
"learning_rate": 0.0002983734764777938,
"loss": 3.7393206787109374,
"step": 3300
},
{
"epoch": 0.1291392004934274,
"grad_norm": 0.5206714868545532,
"learning_rate": 0.00029830209318019676,
"loss": 3.7568234252929686,
"step": 3350
},
{
"epoch": 0.13106665124706063,
"grad_norm": 0.5084478855133057,
"learning_rate": 0.00029822918584913643,
"loss": 3.7494546508789064,
"step": 3400
},
{
"epoch": 0.1329941020006939,
"grad_norm": 0.4988846778869629,
"learning_rate": 0.0002981547552338486,
"loss": 3.71193603515625,
"step": 3450
},
{
"epoch": 0.13492155275432713,
"grad_norm": 0.5246216058731079,
"learning_rate": 0.00029807880209922283,
"loss": 3.710958251953125,
"step": 3500
},
{
"epoch": 0.13684900350796037,
"grad_norm": 0.5130035877227783,
"learning_rate": 0.0002980013272257953,
"loss": 3.701152038574219,
"step": 3550
},
{
"epoch": 0.13877645426159363,
"grad_norm": 0.5510739088058472,
"learning_rate": 0.0002979223314097399,
"loss": 3.7176693725585936,
"step": 3600
},
{
"epoch": 0.14070390501522687,
"grad_norm": 0.5193265080451965,
"learning_rate": 0.0002978418154628612,
"loss": 3.6770294189453123,
"step": 3650
},
{
"epoch": 0.1426313557688601,
"grad_norm": 0.49664029479026794,
"learning_rate": 0.00029775978021258495,
"loss": 3.685413513183594,
"step": 3700
},
{
"epoch": 0.14455880652249334,
"grad_norm": 0.5101340413093567,
"learning_rate": 0.00029767622650195025,
"loss": 3.6848016357421876,
"step": 3750
},
{
"epoch": 0.1464862572761266,
"grad_norm": 0.511060893535614,
"learning_rate": 0.0002975911551896009,
"loss": 3.680214538574219,
"step": 3800
},
{
"epoch": 0.14841370802975984,
"grad_norm": 0.49506616592407227,
"learning_rate": 0.00029750456714977607,
"loss": 3.659295654296875,
"step": 3850
},
{
"epoch": 0.15034115878339308,
"grad_norm": 0.5068514347076416,
"learning_rate": 0.000297416463272302,
"loss": 3.665128173828125,
"step": 3900
},
{
"epoch": 0.15226860953702634,
"grad_norm": 0.5018253922462463,
"learning_rate": 0.00029732684446258216,
"loss": 3.661046142578125,
"step": 3950
},
{
"epoch": 0.15419606029065958,
"grad_norm": 0.4982050061225891,
"learning_rate": 0.0002972357116415887,
"loss": 3.648701171875,
"step": 4000
},
{
"epoch": 0.15419606029065958,
"eval_loss": 3.585422992706299,
"eval_runtime": 71.8171,
"eval_samples_per_second": 116.156,
"eval_steps_per_second": 3.634,
"step": 4000
},
{
"epoch": 0.1561235110442928,
"grad_norm": 0.5031542778015137,
"learning_rate": 0.00029714306574585227,
"loss": 3.64906982421875,
"step": 4050
},
{
"epoch": 0.15805096179792608,
"grad_norm": 0.5202347040176392,
"learning_rate": 0.0002970489077274528,
"loss": 3.6250360107421873,
"step": 4100
},
{
"epoch": 0.1599784125515593,
"grad_norm": 0.5083176493644714,
"learning_rate": 0.0002969532385540095,
"loss": 3.6470623779296876,
"step": 4150
},
{
"epoch": 0.16190586330519255,
"grad_norm": 0.5052973628044128,
"learning_rate": 0.00029685605920867124,
"loss": 3.6127923583984374,
"step": 4200
},
{
"epoch": 0.16383331405882579,
"grad_norm": 0.5176007747650146,
"learning_rate": 0.0002967573706901062,
"loss": 3.647228088378906,
"step": 4250
},
{
"epoch": 0.16576076481245905,
"grad_norm": 0.511567234992981,
"learning_rate": 0.0002966571740124915,
"loss": 3.6209234619140624,
"step": 4300
},
{
"epoch": 0.16768821556609229,
"grad_norm": 0.5003057718276978,
"learning_rate": 0.00029655547020550316,
"loss": 3.618309326171875,
"step": 4350
},
{
"epoch": 0.16961566631972552,
"grad_norm": 0.4762120842933655,
"learning_rate": 0.000296452260314305,
"loss": 3.6237115478515625,
"step": 4400
},
{
"epoch": 0.1715431170733588,
"grad_norm": 0.5058835744857788,
"learning_rate": 0.0002963475453995385,
"loss": 3.5987994384765627,
"step": 4450
},
{
"epoch": 0.17347056782699202,
"grad_norm": 0.5018859505653381,
"learning_rate": 0.00029624132653731134,
"loss": 3.595791015625,
"step": 4500
},
{
"epoch": 0.17539801858062526,
"grad_norm": 0.49935054779052734,
"learning_rate": 0.00029613360481918656,
"loss": 3.5763400268554686,
"step": 4550
},
{
"epoch": 0.17732546933425852,
"grad_norm": 0.48425784707069397,
"learning_rate": 0.00029602438135217167,
"loss": 3.5953665161132813,
"step": 4600
},
{
"epoch": 0.17925292008789176,
"grad_norm": 0.48799827694892883,
"learning_rate": 0.00029591365725870674,
"loss": 3.583829040527344,
"step": 4650
},
{
"epoch": 0.181180370841525,
"grad_norm": 0.4945991635322571,
"learning_rate": 0.0002958014336766531,
"loss": 3.5828253173828126,
"step": 4700
},
{
"epoch": 0.18310782159515823,
"grad_norm": 0.5197539925575256,
"learning_rate": 0.00029568771175928166,
"loss": 3.5751007080078123,
"step": 4750
},
{
"epoch": 0.1850352723487915,
"grad_norm": 0.48726120591163635,
"learning_rate": 0.0002955724926752612,
"loss": 3.5640985107421876,
"step": 4800
},
{
"epoch": 0.18696272310242473,
"grad_norm": 0.5021805763244629,
"learning_rate": 0.000295455777608646,
"loss": 3.571569519042969,
"step": 4850
},
{
"epoch": 0.18889017385605797,
"grad_norm": 0.5098811388015747,
"learning_rate": 0.00029533756775886386,
"loss": 3.5522198486328125,
"step": 4900
},
{
"epoch": 0.19081762460969123,
"grad_norm": 0.507034182548523,
"learning_rate": 0.000295217864340704,
"loss": 3.558240966796875,
"step": 4950
},
{
"epoch": 0.19274507536332447,
"grad_norm": 0.49225080013275146,
"learning_rate": 0.0002950966685843043,
"loss": 3.560546569824219,
"step": 5000
},
{
"epoch": 0.1946725261169577,
"grad_norm": 0.4917547404766083,
"learning_rate": 0.0002949739817351386,
"loss": 3.5673858642578127,
"step": 5050
},
{
"epoch": 0.19659997687059097,
"grad_norm": 0.4910099506378174,
"learning_rate": 0.000294849805054004,
"loss": 3.5580047607421874,
"step": 5100
},
{
"epoch": 0.1985274276242242,
"grad_norm": 0.47943800687789917,
"learning_rate": 0.0002947241398170082,
"loss": 3.5566455078125,
"step": 5150
},
{
"epoch": 0.20045487837785744,
"grad_norm": 0.5112025737762451,
"learning_rate": 0.00029459698731555586,
"loss": 3.5433697509765625,
"step": 5200
},
{
"epoch": 0.20238232913149068,
"grad_norm": 0.47489073872566223,
"learning_rate": 0.0002944683488563357,
"loss": 3.5471234130859375,
"step": 5250
},
{
"epoch": 0.20430977988512394,
"grad_norm": 0.4896199405193329,
"learning_rate": 0.0002943382257613068,
"loss": 3.5303497314453125,
"step": 5300
},
{
"epoch": 0.20623723063875718,
"grad_norm": 0.4875354468822479,
"learning_rate": 0.0002942066193676854,
"loss": 3.5377548217773436,
"step": 5350
},
{
"epoch": 0.20816468139239042,
"grad_norm": 0.4764191508293152,
"learning_rate": 0.00029407353102793083,
"loss": 3.52126220703125,
"step": 5400
},
{
"epoch": 0.21009213214602368,
"grad_norm": 0.4801150858402252,
"learning_rate": 0.00029393896210973166,
"loss": 3.521195373535156,
"step": 5450
},
{
"epoch": 0.21201958289965692,
"grad_norm": 0.47158384323120117,
"learning_rate": 0.00029380291399599176,
"loss": 3.5299456787109373,
"step": 5500
},
{
"epoch": 0.21394703365329015,
"grad_norm": 0.4970817267894745,
"learning_rate": 0.00029366538808481594,
"loss": 3.5156060791015626,
"step": 5550
},
{
"epoch": 0.2158744844069234,
"grad_norm": 0.505355179309845,
"learning_rate": 0.0002935263857894958,
"loss": 3.5102020263671876,
"step": 5600
},
{
"epoch": 0.21780193516055665,
"grad_norm": 0.484792560338974,
"learning_rate": 0.0002933859085384949,
"loss": 3.5256784057617185,
"step": 5650
},
{
"epoch": 0.2197293859141899,
"grad_norm": 0.507871150970459,
"learning_rate": 0.0002932439577754344,
"loss": 3.51783203125,
"step": 5700
},
{
"epoch": 0.22165683666782313,
"grad_norm": 0.48395925760269165,
"learning_rate": 0.00029310053495907804,
"loss": 3.499888610839844,
"step": 5750
},
{
"epoch": 0.2235842874214564,
"grad_norm": 0.48618054389953613,
"learning_rate": 0.00029295564156331714,
"loss": 3.503265380859375,
"step": 5800
},
{
"epoch": 0.22551173817508963,
"grad_norm": 0.4817364513874054,
"learning_rate": 0.0002928092790771555,
"loss": 3.51520751953125,
"step": 5850
},
{
"epoch": 0.22743918892872286,
"grad_norm": 0.483539879322052,
"learning_rate": 0.0002926614490046941,
"loss": 3.5028955078125,
"step": 5900
},
{
"epoch": 0.22936663968235613,
"grad_norm": 0.48083049058914185,
"learning_rate": 0.00029251215286511573,
"loss": 3.4965765380859377,
"step": 5950
},
{
"epoch": 0.23129409043598936,
"grad_norm": 0.5043441653251648,
"learning_rate": 0.00029236139219266917,
"loss": 3.479033203125,
"step": 6000
},
{
"epoch": 0.23129409043598936,
"eval_loss": 3.434314489364624,
"eval_runtime": 72.2122,
"eval_samples_per_second": 115.521,
"eval_steps_per_second": 3.614,
"step": 6000
},
{
"epoch": 0.2332215411896226,
"grad_norm": 0.4738307297229767,
"learning_rate": 0.00029220916853665363,
"loss": 3.489224853515625,
"step": 6050
},
{
"epoch": 0.23514899194325584,
"grad_norm": 0.48999547958374023,
"learning_rate": 0.0002920554834614027,
"loss": 3.4707180786132814,
"step": 6100
},
{
"epoch": 0.2370764426968891,
"grad_norm": 0.46922704577445984,
"learning_rate": 0.00029190033854626816,
"loss": 3.4679449462890624,
"step": 6150
},
{
"epoch": 0.23900389345052234,
"grad_norm": 0.49618691205978394,
"learning_rate": 0.0002917437353856043,
"loss": 3.49760498046875,
"step": 6200
},
{
"epoch": 0.24093134420415557,
"grad_norm": 0.48404061794281006,
"learning_rate": 0.00029158567558875076,
"loss": 3.4569793701171876,
"step": 6250
},
{
"epoch": 0.24285879495778884,
"grad_norm": 0.49926745891571045,
"learning_rate": 0.0002914261607800165,
"loss": 3.4843402099609375,
"step": 6300
},
{
"epoch": 0.24478624571142207,
"grad_norm": 0.48060157895088196,
"learning_rate": 0.000291265192598663,
"loss": 3.4751336669921873,
"step": 6350
},
{
"epoch": 0.2467136964650553,
"grad_norm": 0.49387532472610474,
"learning_rate": 0.00029110277269888736,
"loss": 3.458801574707031,
"step": 6400
},
{
"epoch": 0.24864114721868857,
"grad_norm": 0.4965198338031769,
"learning_rate": 0.0002909389027498054,
"loss": 3.4571484375,
"step": 6450
},
{
"epoch": 0.2505685979723218,
"grad_norm": 0.4690743088722229,
"learning_rate": 0.0002907735844354344,
"loss": 3.46658203125,
"step": 6500
},
{
"epoch": 0.25249604872595505,
"grad_norm": 0.4719405770301819,
"learning_rate": 0.0002906068194546757,
"loss": 3.4751458740234376,
"step": 6550
},
{
"epoch": 0.2544234994795883,
"grad_norm": 0.4720248281955719,
"learning_rate": 0.00029043860952129765,
"loss": 3.480400390625,
"step": 6600
},
{
"epoch": 0.2563509502332215,
"grad_norm": 0.4931813180446625,
"learning_rate": 0.00029026895636391747,
"loss": 3.4634954833984377,
"step": 6650
},
{
"epoch": 0.2582784009868548,
"grad_norm": 0.5063376426696777,
"learning_rate": 0.00029009786172598393,
"loss": 3.4451904296875,
"step": 6700
},
{
"epoch": 0.26020585174048805,
"grad_norm": 0.4933936893939972,
"learning_rate": 0.0002899253273657591,
"loss": 3.4367218017578125,
"step": 6750
},
{
"epoch": 0.26213330249412126,
"grad_norm": 0.48165440559387207,
"learning_rate": 0.0002897513550563004,
"loss": 3.4317822265625,
"step": 6800
},
{
"epoch": 0.2640607532477545,
"grad_norm": 0.48148947954177856,
"learning_rate": 0.00028957594658544266,
"loss": 3.4445721435546877,
"step": 6850
},
{
"epoch": 0.2659882040013878,
"grad_norm": 0.49338117241859436,
"learning_rate": 0.0002893991037557792,
"loss": 3.4329559326171877,
"step": 6900
},
{
"epoch": 0.267915654755021,
"grad_norm": 0.47753119468688965,
"learning_rate": 0.00028922082838464373,
"loss": 3.4400469970703127,
"step": 6950
},
{
"epoch": 0.26984310550865426,
"grad_norm": 0.4834132194519043,
"learning_rate": 0.0002890411223040915,
"loss": 3.4297528076171875,
"step": 7000
},
{
"epoch": 0.2717705562622875,
"grad_norm": 0.47897446155548096,
"learning_rate": 0.0002888599873608806,
"loss": 3.4394317626953126,
"step": 7050
},
{
"epoch": 0.27369800701592073,
"grad_norm": 0.4720662534236908,
"learning_rate": 0.00028867742541645276,
"loss": 3.435224609375,
"step": 7100
},
{
"epoch": 0.275625457769554,
"grad_norm": 0.484162300825119,
"learning_rate": 0.00028849343834691445,
"loss": 3.4345529174804685,
"step": 7150
},
{
"epoch": 0.27755290852318726,
"grad_norm": 0.4884200096130371,
"learning_rate": 0.00028830802804301743,
"loss": 3.435291442871094,
"step": 7200
},
{
"epoch": 0.27948035927682047,
"grad_norm": 0.5109637975692749,
"learning_rate": 0.0002881211964101395,
"loss": 3.4217279052734373,
"step": 7250
},
{
"epoch": 0.28140781003045373,
"grad_norm": 0.47823017835617065,
"learning_rate": 0.0002879329453682647,
"loss": 3.426511535644531,
"step": 7300
},
{
"epoch": 0.28333526078408694,
"grad_norm": 0.5128335952758789,
"learning_rate": 0.00028774327685196385,
"loss": 3.427406005859375,
"step": 7350
},
{
"epoch": 0.2852627115377202,
"grad_norm": 0.4827037453651428,
"learning_rate": 0.0002875521928103744,
"loss": 3.4155810546875,
"step": 7400
},
{
"epoch": 0.28719016229135347,
"grad_norm": 0.481082022190094,
"learning_rate": 0.0002873596952071804,
"loss": 3.4052203369140623,
"step": 7450
},
{
"epoch": 0.2891176130449867,
"grad_norm": 0.4779789447784424,
"learning_rate": 0.0002871657860205926,
"loss": 3.4142279052734374,
"step": 7500
},
{
"epoch": 0.29104506379861994,
"grad_norm": 0.49225643277168274,
"learning_rate": 0.00028697438741990324,
"loss": 3.4227734375,
"step": 7550
},
{
"epoch": 0.2929725145522532,
"grad_norm": 0.5048698782920837,
"learning_rate": 0.00028677768919106977,
"loss": 3.411636962890625,
"step": 7600
},
{
"epoch": 0.2948999653058864,
"grad_norm": 0.47102615237236023,
"learning_rate": 0.0002865795853598548,
"loss": 3.4300201416015623,
"step": 7650
},
{
"epoch": 0.2968274160595197,
"grad_norm": 0.5100015997886658,
"learning_rate": 0.00028638007796208176,
"loss": 3.4064337158203126,
"step": 7700
},
{
"epoch": 0.29875486681315294,
"grad_norm": 0.48272567987442017,
"learning_rate": 0.00028617916904799783,
"loss": 3.4024899291992186,
"step": 7750
},
{
"epoch": 0.30068231756678615,
"grad_norm": 0.4890754222869873,
"learning_rate": 0.000285976860682253,
"loss": 3.4252352905273438,
"step": 7800
},
{
"epoch": 0.3026097683204194,
"grad_norm": 0.48382148146629333,
"learning_rate": 0.0002857731549438787,
"loss": 3.4206988525390627,
"step": 7850
},
{
"epoch": 0.3045372190740527,
"grad_norm": 0.47125667333602905,
"learning_rate": 0.00028556805392626654,
"loss": 3.41896728515625,
"step": 7900
},
{
"epoch": 0.3064646698276859,
"grad_norm": 0.4767087697982788,
"learning_rate": 0.0002853615597371468,
"loss": 3.4015020751953124,
"step": 7950
},
{
"epoch": 0.30839212058131915,
"grad_norm": 0.4715645909309387,
"learning_rate": 0.0002851536744985669,
"loss": 3.40042236328125,
"step": 8000
},
{
"epoch": 0.30839212058131915,
"eval_loss": 3.342198371887207,
"eval_runtime": 71.6585,
"eval_samples_per_second": 116.413,
"eval_steps_per_second": 3.642,
"step": 8000
},
{
"epoch": 0.3103195713349524,
"grad_norm": 0.4898245930671692,
"learning_rate": 0.00028494440034686913,
"loss": 3.415509033203125,
"step": 8050
},
{
"epoch": 0.3122470220885856,
"grad_norm": 0.48199841380119324,
"learning_rate": 0.0002847337394326693,
"loss": 3.3922320556640626,
"step": 8100
},
{
"epoch": 0.3141744728422189,
"grad_norm": 0.48237958550453186,
"learning_rate": 0.0002845216939208343,
"loss": 3.393681640625,
"step": 8150
},
{
"epoch": 0.31610192359585215,
"grad_norm": 0.4685034453868866,
"learning_rate": 0.0002843082659904597,
"loss": 3.40257568359375,
"step": 8200
},
{
"epoch": 0.31802937434948536,
"grad_norm": 0.4959200620651245,
"learning_rate": 0.00028409345783484774,
"loss": 3.3937017822265627,
"step": 8250
},
{
"epoch": 0.3199568251031186,
"grad_norm": 0.48208650946617126,
"learning_rate": 0.00028387727166148463,
"loss": 3.3995901489257814,
"step": 8300
},
{
"epoch": 0.32188427585675183,
"grad_norm": 0.49182265996932983,
"learning_rate": 0.0002836597096920178,
"loss": 3.3908148193359375,
"step": 8350
},
{
"epoch": 0.3238117266103851,
"grad_norm": 0.4829288423061371,
"learning_rate": 0.000283440774162233,
"loss": 3.39423583984375,
"step": 8400
},
{
"epoch": 0.32573917736401836,
"grad_norm": 0.4829620122909546,
"learning_rate": 0.00028322046732203165,
"loss": 3.3739398193359373,
"step": 8450
},
{
"epoch": 0.32766662811765157,
"grad_norm": 0.49594053626060486,
"learning_rate": 0.00028299879143540735,
"loss": 3.378362731933594,
"step": 8500
},
{
"epoch": 0.32959407887128483,
"grad_norm": 0.489305704832077,
"learning_rate": 0.0002827757487804228,
"loss": 3.382164306640625,
"step": 8550
},
{
"epoch": 0.3315215296249181,
"grad_norm": 0.4979233145713806,
"learning_rate": 0.0002825513416491864,
"loss": 3.3737216186523438,
"step": 8600
},
{
"epoch": 0.3334489803785513,
"grad_norm": 0.4722122251987457,
"learning_rate": 0.00028232557234782863,
"loss": 3.3916912841796876,
"step": 8650
},
{
"epoch": 0.33537643113218457,
"grad_norm": 0.5052176713943481,
"learning_rate": 0.00028209844319647837,
"loss": 3.386777648925781,
"step": 8700
},
{
"epoch": 0.33730388188581784,
"grad_norm": 0.4875766932964325,
"learning_rate": 0.00028186995652923903,
"loss": 3.3826312255859374,
"step": 8750
},
{
"epoch": 0.33923133263945104,
"grad_norm": 0.4967861473560333,
"learning_rate": 0.0002816401146941647,
"loss": 3.3676068115234377,
"step": 8800
},
{
"epoch": 0.3411587833930843,
"grad_norm": 0.4899598956108093,
"learning_rate": 0.00028140892005323583,
"loss": 3.358997497558594,
"step": 8850
},
{
"epoch": 0.3430862341467176,
"grad_norm": 0.48528599739074707,
"learning_rate": 0.00028117637498233507,
"loss": 3.3772552490234373,
"step": 8900
},
{
"epoch": 0.3450136849003508,
"grad_norm": 0.4654850661754608,
"learning_rate": 0.0002809424818712228,
"loss": 3.3589202880859377,
"step": 8950
},
{
"epoch": 0.34694113565398405,
"grad_norm": 0.5098646283149719,
"learning_rate": 0.0002807072431235126,
"loss": 3.3637164306640623,
"step": 9000
},
{
"epoch": 0.3488685864076173,
"grad_norm": 0.4997151792049408,
"learning_rate": 0.00028047066115664663,
"loss": 3.3785467529296875,
"step": 9050
},
{
"epoch": 0.3507960371612505,
"grad_norm": 0.46549203991889954,
"learning_rate": 0.0002802327384018706,
"loss": 3.360137634277344,
"step": 9100
},
{
"epoch": 0.3527234879148838,
"grad_norm": 0.5423162579536438,
"learning_rate": 0.00027999347730420893,
"loss": 3.35434326171875,
"step": 9150
},
{
"epoch": 0.35465093866851705,
"grad_norm": 0.47321420907974243,
"learning_rate": 0.00027975288032243965,
"loss": 3.3585409545898437,
"step": 9200
},
{
"epoch": 0.35657838942215025,
"grad_norm": 0.4811253845691681,
"learning_rate": 0.0002795109499290689,
"loss": 3.367620849609375,
"step": 9250
},
{
"epoch": 0.3585058401757835,
"grad_norm": 0.4982966184616089,
"learning_rate": 0.0002792676886103059,
"loss": 3.3639886474609373,
"step": 9300
},
{
"epoch": 0.3604332909294167,
"grad_norm": 0.5057178735733032,
"learning_rate": 0.000279023098866037,
"loss": 3.360538330078125,
"step": 9350
},
{
"epoch": 0.36236074168305,
"grad_norm": 0.4778556823730469,
"learning_rate": 0.00027877718320980027,
"loss": 3.3363970947265624,
"step": 9400
},
{
"epoch": 0.36428819243668326,
"grad_norm": 0.4765235185623169,
"learning_rate": 0.00027852994416875963,
"loss": 3.3325067138671876,
"step": 9450
},
{
"epoch": 0.36621564319031646,
"grad_norm": 0.49356508255004883,
"learning_rate": 0.0002782813842836787,
"loss": 3.3406036376953123,
"step": 9500
},
{
"epoch": 0.36814309394394973,
"grad_norm": 0.47266724705696106,
"learning_rate": 0.0002780315061088948,
"loss": 3.351762390136719,
"step": 9550
},
{
"epoch": 0.370070544697583,
"grad_norm": 0.5001718401908875,
"learning_rate": 0.0002777853489676245,
"loss": 3.3632208251953126,
"step": 9600
},
{
"epoch": 0.3719979954512162,
"grad_norm": 0.4980491101741791,
"learning_rate": 0.0002775328681680337,
"loss": 3.3641986083984374,
"step": 9650
},
{
"epoch": 0.37392544620484947,
"grad_norm": 0.49997082352638245,
"learning_rate": 0.00027727907677090117,
"loss": 3.340047607421875,
"step": 9700
},
{
"epoch": 0.37585289695848273,
"grad_norm": 0.4940063953399658,
"learning_rate": 0.00027702397738432624,
"loss": 3.3303033447265626,
"step": 9750
},
{
"epoch": 0.37778034771211594,
"grad_norm": 0.48513808846473694,
"learning_rate": 0.00027676757262984987,
"loss": 3.3373089599609376,
"step": 9800
},
{
"epoch": 0.3797077984657492,
"grad_norm": 0.5037530660629272,
"learning_rate": 0.0002765098651424275,
"loss": 3.350380554199219,
"step": 9850
},
{
"epoch": 0.38163524921938247,
"grad_norm": 0.49754294753074646,
"learning_rate": 0.00027625085757040235,
"loss": 3.357091064453125,
"step": 9900
},
{
"epoch": 0.3835626999730157,
"grad_norm": 0.5108275413513184,
"learning_rate": 0.00027599055257547793,
"loss": 3.3248159790039065,
"step": 9950
},
{
"epoch": 0.38549015072664894,
"grad_norm": 0.5188531875610352,
"learning_rate": 0.0002757289528326909,
"loss": 3.3242608642578126,
"step": 10000
},
{
"epoch": 0.38549015072664894,
"eval_loss": 3.279729127883911,
"eval_runtime": 72.1979,
"eval_samples_per_second": 115.544,
"eval_steps_per_second": 3.615,
"step": 10000
},
{
"epoch": 0.3874176014802822,
"grad_norm": 0.485309898853302,
"learning_rate": 0.00027546606103038326,
"loss": 3.3307257080078125,
"step": 10050
},
{
"epoch": 0.3893450522339154,
"grad_norm": 0.48716261982917786,
"learning_rate": 0.00027520187987017513,
"loss": 3.354161682128906,
"step": 10100
},
{
"epoch": 0.3912725029875487,
"grad_norm": 0.4917289912700653,
"learning_rate": 0.00027493641206693655,
"loss": 3.3324749755859373,
"step": 10150
},
{
"epoch": 0.39319995374118194,
"grad_norm": 0.49136367440223694,
"learning_rate": 0.0002746696603487601,
"loss": 3.321602478027344,
"step": 10200
},
{
"epoch": 0.39512740449481515,
"grad_norm": 0.4706106185913086,
"learning_rate": 0.00027440162745693224,
"loss": 3.3309661865234377,
"step": 10250
},
{
"epoch": 0.3970548552484484,
"grad_norm": 0.49642741680145264,
"learning_rate": 0.00027413231614590574,
"loss": 3.3401861572265625,
"step": 10300
},
{
"epoch": 0.3989823060020816,
"grad_norm": 0.4948924481868744,
"learning_rate": 0.00027386172918327103,
"loss": 3.32413818359375,
"step": 10350
},
{
"epoch": 0.4009097567557149,
"grad_norm": 0.5285770297050476,
"learning_rate": 0.00027358986934972776,
"loss": 3.334150390625,
"step": 10400
},
{
"epoch": 0.40283720750934815,
"grad_norm": 0.4994802474975586,
"learning_rate": 0.00027331673943905633,
"loss": 3.3236483764648437,
"step": 10450
},
{
"epoch": 0.40476465826298136,
"grad_norm": 0.4856112599372864,
"learning_rate": 0.00027304234225808914,
"loss": 3.325201110839844,
"step": 10500
},
{
"epoch": 0.4066921090166146,
"grad_norm": 0.4792839288711548,
"learning_rate": 0.0002727666806266819,
"loss": 3.3177694702148437,
"step": 10550
},
{
"epoch": 0.4086195597702479,
"grad_norm": 0.49488285183906555,
"learning_rate": 0.0002724897573776842,
"loss": 3.3090228271484374,
"step": 10600
},
{
"epoch": 0.4105470105238811,
"grad_norm": 0.4831986427307129,
"learning_rate": 0.0002722115753569111,
"loss": 3.302528076171875,
"step": 10650
},
{
"epoch": 0.41247446127751436,
"grad_norm": 0.5022083520889282,
"learning_rate": 0.00027193213742311305,
"loss": 3.3182904052734377,
"step": 10700
},
{
"epoch": 0.4144019120311476,
"grad_norm": 0.48181504011154175,
"learning_rate": 0.0002716514464479474,
"loss": 3.3185986328125,
"step": 10750
},
{
"epoch": 0.41632936278478083,
"grad_norm": 0.4877314269542694,
"learning_rate": 0.000271369505315948,
"loss": 3.3362991333007814,
"step": 10800
},
{
"epoch": 0.4182568135384141,
"grad_norm": 0.5079516172409058,
"learning_rate": 0.0002710863169244963,
"loss": 3.3230828857421875,
"step": 10850
},
{
"epoch": 0.42018426429204736,
"grad_norm": 0.5222039222717285,
"learning_rate": 0.00027080188418379116,
"loss": 3.310635681152344,
"step": 10900
},
{
"epoch": 0.42211171504568057,
"grad_norm": 0.48713019490242004,
"learning_rate": 0.00027051621001681914,
"loss": 3.3143301391601563,
"step": 10950
},
{
"epoch": 0.42403916579931383,
"grad_norm": 0.5177494883537292,
"learning_rate": 0.0002702292973593242,
"loss": 3.305172119140625,
"step": 11000
},
{
"epoch": 0.4259666165529471,
"grad_norm": 0.5103081464767456,
"learning_rate": 0.00026994114915977784,
"loss": 3.29141845703125,
"step": 11050
},
{
"epoch": 0.4278940673065803,
"grad_norm": 0.48814478516578674,
"learning_rate": 0.00026965176837934874,
"loss": 3.3208197021484374,
"step": 11100
},
{
"epoch": 0.42982151806021357,
"grad_norm": 0.5066275596618652,
"learning_rate": 0.000269361157991872,
"loss": 3.320732421875,
"step": 11150
},
{
"epoch": 0.4317489688138468,
"grad_norm": 0.5084295868873596,
"learning_rate": 0.0002690693209838191,
"loss": 3.3152029418945315,
"step": 11200
},
{
"epoch": 0.43367641956748004,
"grad_norm": 0.5029125213623047,
"learning_rate": 0.00026877626035426677,
"loss": 3.302169189453125,
"step": 11250
},
{
"epoch": 0.4356038703211133,
"grad_norm": 0.5140436291694641,
"learning_rate": 0.00026848197911486643,
"loss": 3.3106985473632813,
"step": 11300
},
{
"epoch": 0.4375313210747465,
"grad_norm": 0.4871470630168915,
"learning_rate": 0.0002681864802898132,
"loss": 3.3046832275390625,
"step": 11350
},
{
"epoch": 0.4394587718283798,
"grad_norm": 0.510291576385498,
"learning_rate": 0.00026788976691581454,
"loss": 3.318800048828125,
"step": 11400
},
{
"epoch": 0.44138622258201304,
"grad_norm": 0.500678300857544,
"learning_rate": 0.00026759184204205963,
"loss": 3.290931396484375,
"step": 11450
},
{
"epoch": 0.44331367333564625,
"grad_norm": 0.5076199769973755,
"learning_rate": 0.0002672927087301874,
"loss": 3.2901202392578126,
"step": 11500
},
{
"epoch": 0.4452411240892795,
"grad_norm": 0.49749693274497986,
"learning_rate": 0.0002669923700542555,
"loss": 3.307938232421875,
"step": 11550
},
{
"epoch": 0.4471685748429128,
"grad_norm": 0.5069649815559387,
"learning_rate": 0.0002666908291007085,
"loss": 3.287379150390625,
"step": 11600
},
{
"epoch": 0.449096025596546,
"grad_norm": 0.49996042251586914,
"learning_rate": 0.00026638808896834625,
"loss": 3.294720458984375,
"step": 11650
},
{
"epoch": 0.45102347635017925,
"grad_norm": 0.5172936320304871,
"learning_rate": 0.000266084152768292,
"loss": 3.3005242919921876,
"step": 11700
},
{
"epoch": 0.4529509271038125,
"grad_norm": 0.4884229302406311,
"learning_rate": 0.0002657790236239604,
"loss": 3.3028387451171874,
"step": 11750
},
{
"epoch": 0.4548783778574457,
"grad_norm": 0.4896923005580902,
"learning_rate": 0.0002654727046710256,
"loss": 3.303028564453125,
"step": 11800
},
{
"epoch": 0.456805828611079,
"grad_norm": 0.49592748284339905,
"learning_rate": 0.00026516519905738883,
"loss": 3.285137939453125,
"step": 11850
},
{
"epoch": 0.45873327936471225,
"grad_norm": 0.4845772683620453,
"learning_rate": 0.0002648565099431459,
"loss": 3.28921630859375,
"step": 11900
},
{
"epoch": 0.46066073011834546,
"grad_norm": 0.5072834491729736,
"learning_rate": 0.0002645466405005552,
"loss": 3.3063189697265627,
"step": 11950
},
{
"epoch": 0.4625881808719787,
"grad_norm": 0.49562397599220276,
"learning_rate": 0.00026423559391400465,
"loss": 3.2722076416015624,
"step": 12000
},
{
"epoch": 0.4625881808719787,
"eval_loss": 3.23704195022583,
"eval_runtime": 63.8348,
"eval_samples_per_second": 130.681,
"eval_steps_per_second": 4.089,
"step": 12000
},
{
"epoch": 0.464515631625612,
"grad_norm": 0.485429972410202,
"learning_rate": 0.0002639233733799791,
"loss": 3.2845184326171877,
"step": 12050
},
{
"epoch": 0.4664430823792452,
"grad_norm": 0.5060269236564636,
"learning_rate": 0.00026361626138491734,
"loss": 3.289698181152344,
"step": 12100
},
{
"epoch": 0.46837053313287846,
"grad_norm": 0.5207314491271973,
"learning_rate": 0.00026330172591234794,
"loss": 3.283543701171875,
"step": 12150
},
{
"epoch": 0.4702979838865117,
"grad_norm": 0.4899747669696808,
"learning_rate": 0.00026298602608924214,
"loss": 3.2631646728515626,
"step": 12200
},
{
"epoch": 0.47222543464014494,
"grad_norm": 0.511677086353302,
"learning_rate": 0.0002626691651599042,
"loss": 3.2772393798828126,
"step": 12250
},
{
"epoch": 0.4741528853937782,
"grad_norm": 0.5158467888832092,
"learning_rate": 0.0002623511463805704,
"loss": 3.292477722167969,
"step": 12300
},
{
"epoch": 0.4760803361474114,
"grad_norm": 0.5188326239585876,
"learning_rate": 0.0002620319730193757,
"loss": 3.268270263671875,
"step": 12350
},
{
"epoch": 0.4780077869010447,
"grad_norm": 0.4904669523239136,
"learning_rate": 0.00026171164835632027,
"loss": 3.2848394775390624,
"step": 12400
},
{
"epoch": 0.47993523765467794,
"grad_norm": 0.525324821472168,
"learning_rate": 0.0002613901756832356,
"loss": 3.27462890625,
"step": 12450
},
{
"epoch": 0.48186268840831115,
"grad_norm": 0.5027859210968018,
"learning_rate": 0.0002610675583037508,
"loss": 3.2780300903320314,
"step": 12500
},
{
"epoch": 0.4837901391619444,
"grad_norm": 0.5029728412628174,
"learning_rate": 0.00026074379953325863,
"loss": 3.2874945068359374,
"step": 12550
},
{
"epoch": 0.4857175899155777,
"grad_norm": 0.5005213022232056,
"learning_rate": 0.00026041890269888136,
"loss": 3.2655682373046875,
"step": 12600
},
{
"epoch": 0.4876450406692109,
"grad_norm": 0.5063687562942505,
"learning_rate": 0.0002600928711394367,
"loss": 3.2774481201171874,
"step": 12650
},
{
"epoch": 0.48957249142284415,
"grad_norm": 0.4986209273338318,
"learning_rate": 0.00025976570820540325,
"loss": 3.289039306640625,
"step": 12700
},
{
"epoch": 0.4914999421764774,
"grad_norm": 0.5154808759689331,
"learning_rate": 0.00025943741725888646,
"loss": 3.270457458496094,
"step": 12750
},
{
"epoch": 0.4934273929301106,
"grad_norm": 0.5098702311515808,
"learning_rate": 0.00025910800167358366,
"loss": 3.2625686645507814,
"step": 12800
},
{
"epoch": 0.4953548436837439,
"grad_norm": 0.5277470946311951,
"learning_rate": 0.00025877746483474976,
"loss": 3.289451904296875,
"step": 12850
},
{
"epoch": 0.49728229443737715,
"grad_norm": 0.5058562159538269,
"learning_rate": 0.00025844581013916213,
"loss": 3.287037658691406,
"step": 12900
},
{
"epoch": 0.49920974519101036,
"grad_norm": 0.5200469493865967,
"learning_rate": 0.0002581130409950859,
"loss": 3.271123352050781,
"step": 12950
},
{
"epoch": 0.5011371959446436,
"grad_norm": 0.5302641987800598,
"learning_rate": 0.0002577791608222389,
"loss": 3.2679959106445313,
"step": 13000
},
{
"epoch": 0.5030646466982769,
"grad_norm": 0.5055278539657593,
"learning_rate": 0.00025744417305175653,
"loss": 3.2799420166015625,
"step": 13050
},
{
"epoch": 0.5049920974519101,
"grad_norm": 0.5066748261451721,
"learning_rate": 0.00025710808112615644,
"loss": 3.2776678466796874,
"step": 13100
},
{
"epoch": 0.5069195482055433,
"grad_norm": 0.5080742835998535,
"learning_rate": 0.00025677088849930307,
"loss": 3.2569970703125,
"step": 13150
},
{
"epoch": 0.5088469989591766,
"grad_norm": 0.49316340684890747,
"learning_rate": 0.0002564325986363725,
"loss": 3.2653009033203126,
"step": 13200
},
{
"epoch": 0.5107744497128098,
"grad_norm": 0.5022343993186951,
"learning_rate": 0.00025609321501381637,
"loss": 3.254708251953125,
"step": 13250
},
{
"epoch": 0.512701900466443,
"grad_norm": 0.5125695466995239,
"learning_rate": 0.00025575274111932654,
"loss": 3.2687692260742187,
"step": 13300
},
{
"epoch": 0.5146293512200764,
"grad_norm": 0.5127341747283936,
"learning_rate": 0.00025541118045179916,
"loss": 3.24674560546875,
"step": 13350
},
{
"epoch": 0.5165568019737096,
"grad_norm": 0.5302950143814087,
"learning_rate": 0.00025506853652129837,
"loss": 3.2523040771484375,
"step": 13400
},
{
"epoch": 0.5184842527273428,
"grad_norm": 0.4974203109741211,
"learning_rate": 0.0002547248128490209,
"loss": 3.2576162719726565,
"step": 13450
},
{
"epoch": 0.5204117034809761,
"grad_norm": 0.5059196352958679,
"learning_rate": 0.0002543800129672591,
"loss": 3.26962158203125,
"step": 13500
},
{
"epoch": 0.5223391542346093,
"grad_norm": 0.4979102313518524,
"learning_rate": 0.00025403414041936547,
"loss": 3.2441473388671875,
"step": 13550
},
{
"epoch": 0.5242666049882425,
"grad_norm": 0.5194156765937805,
"learning_rate": 0.00025368719875971534,
"loss": 3.2662139892578126,
"step": 13600
},
{
"epoch": 0.5261940557418758,
"grad_norm": 0.5145624279975891,
"learning_rate": 0.00025333919155367125,
"loss": 3.246636962890625,
"step": 13650
},
{
"epoch": 0.528121506495509,
"grad_norm": 0.5134981870651245,
"learning_rate": 0.0002529901223775455,
"loss": 3.2625299072265626,
"step": 13700
},
{
"epoch": 0.5300489572491422,
"grad_norm": 0.5281381607055664,
"learning_rate": 0.0002526399948185642,
"loss": 3.2776739501953127,
"step": 13750
},
{
"epoch": 0.5319764080027756,
"grad_norm": 0.5279331207275391,
"learning_rate": 0.00025228881247482953,
"loss": 3.252415771484375,
"step": 13800
},
{
"epoch": 0.5339038587564088,
"grad_norm": 0.5229082703590393,
"learning_rate": 0.0002519365789552835,
"loss": 3.258053894042969,
"step": 13850
},
{
"epoch": 0.535831309510042,
"grad_norm": 0.5027236342430115,
"learning_rate": 0.0002515832978796705,
"loss": 3.2474884033203124,
"step": 13900
},
{
"epoch": 0.5377587602636753,
"grad_norm": 0.5057716369628906,
"learning_rate": 0.00025122897287850024,
"loss": 3.244605712890625,
"step": 13950
},
{
"epoch": 0.5396862110173085,
"grad_norm": 0.519944965839386,
"learning_rate": 0.00025087360759301023,
"loss": 3.25479736328125,
"step": 14000
},
{
"epoch": 0.5396862110173085,
"eval_loss": 3.1955931186676025,
"eval_runtime": 63.6143,
"eval_samples_per_second": 131.134,
"eval_steps_per_second": 4.103,
"step": 14000
},
{
"epoch": 0.5416136617709417,
"grad_norm": 0.4972708821296692,
"learning_rate": 0.0002505172056751287,
"loss": 3.276119384765625,
"step": 14050
},
{
"epoch": 0.543541112524575,
"grad_norm": 0.5104284286499023,
"learning_rate": 0.00025015977078743677,
"loss": 3.229161376953125,
"step": 14100
},
{
"epoch": 0.5454685632782083,
"grad_norm": 0.5257759094238281,
"learning_rate": 0.00024980848595011453,
"loss": 3.2508782958984375,
"step": 14150
},
{
"epoch": 0.5473960140318415,
"grad_norm": 0.5083267688751221,
"learning_rate": 0.0002494490166290564,
"loss": 3.2490048217773437,
"step": 14200
},
{
"epoch": 0.5493234647854748,
"grad_norm": 0.5066636204719543,
"learning_rate": 0.000249088525315483,
"loss": 3.23043701171875,
"step": 14250
},
{
"epoch": 0.551250915539108,
"grad_norm": 0.5071622133255005,
"learning_rate": 0.0002487270157140004,
"loss": 3.2611257934570315,
"step": 14300
},
{
"epoch": 0.5531783662927412,
"grad_norm": 0.520742654800415,
"learning_rate": 0.00024836449153967896,
"loss": 3.2509039306640624,
"step": 14350
},
{
"epoch": 0.5551058170463745,
"grad_norm": 0.5163410305976868,
"learning_rate": 0.00024800095651801563,
"loss": 3.2470028686523436,
"step": 14400
},
{
"epoch": 0.5570332678000077,
"grad_norm": 0.5406373739242554,
"learning_rate": 0.00024763641438489513,
"loss": 3.233168640136719,
"step": 14450
},
{
"epoch": 0.5589607185536409,
"grad_norm": 0.6013800501823425,
"learning_rate": 0.0002472708688865519,
"loss": 3.246990966796875,
"step": 14500
},
{
"epoch": 0.5608881693072743,
"grad_norm": 0.5319772958755493,
"learning_rate": 0.00024690432377953155,
"loss": 3.244833984375,
"step": 14550
},
{
"epoch": 0.5628156200609075,
"grad_norm": 0.5161493420600891,
"learning_rate": 0.0002465367828306522,
"loss": 3.2414398193359375,
"step": 14600
},
{
"epoch": 0.5647430708145407,
"grad_norm": 0.5242125391960144,
"learning_rate": 0.00024616824981696573,
"loss": 3.2221527099609375,
"step": 14650
},
{
"epoch": 0.5666705215681739,
"grad_norm": 0.5306032299995422,
"learning_rate": 0.00024579872852571906,
"loss": 3.2450616455078123,
"step": 14700
},
{
"epoch": 0.5685979723218072,
"grad_norm": 0.5260001420974731,
"learning_rate": 0.0002454282227543152,
"loss": 3.226510009765625,
"step": 14750
},
{
"epoch": 0.5705254230754404,
"grad_norm": 0.5230858325958252,
"learning_rate": 0.00024505673631027415,
"loss": 3.2494967651367186,
"step": 14800
},
{
"epoch": 0.5724528738290736,
"grad_norm": 0.5204994678497314,
"learning_rate": 0.000244684273011194,
"loss": 3.245614013671875,
"step": 14850
},
{
"epoch": 0.5743803245827069,
"grad_norm": 0.5140077471733093,
"learning_rate": 0.00024431083668471134,
"loss": 3.2390606689453123,
"step": 14900
},
{
"epoch": 0.5763077753363401,
"grad_norm": 0.533420741558075,
"learning_rate": 0.00024393643116846225,
"loss": 3.240533752441406,
"step": 14950
},
{
"epoch": 0.5782352260899734,
"grad_norm": 0.515992283821106,
"learning_rate": 0.0002435610603100427,
"loss": 3.2578091430664062,
"step": 15000
},
{
"epoch": 0.5801626768436067,
"grad_norm": 0.5196648240089417,
"learning_rate": 0.000243184727966969,
"loss": 3.223043212890625,
"step": 15050
},
{
"epoch": 0.5820901275972399,
"grad_norm": 0.5403545498847961,
"learning_rate": 0.00024280743800663828,
"loss": 3.2343881225585935,
"step": 15100
},
{
"epoch": 0.5840175783508731,
"grad_norm": 0.5275980234146118,
"learning_rate": 0.00024242919430628865,
"loss": 3.239459228515625,
"step": 15150
},
{
"epoch": 0.5859450291045064,
"grad_norm": 0.5118796229362488,
"learning_rate": 0.00024205000075295933,
"loss": 3.2324664306640627,
"step": 15200
},
{
"epoch": 0.5878724798581396,
"grad_norm": 0.5176474452018738,
"learning_rate": 0.00024166986124345076,
"loss": 3.2280877685546874,
"step": 15250
},
{
"epoch": 0.5897999306117728,
"grad_norm": 0.5145363211631775,
"learning_rate": 0.00024128877968428457,
"loss": 3.233060302734375,
"step": 15300
},
{
"epoch": 0.5917273813654061,
"grad_norm": 0.5313825607299805,
"learning_rate": 0.0002409067599916634,
"loss": 3.228967590332031,
"step": 15350
},
{
"epoch": 0.5936548321190394,
"grad_norm": 0.5396929383277893,
"learning_rate": 0.0002405238060914307,
"loss": 3.2446685791015626,
"step": 15400
},
{
"epoch": 0.5955822828726726,
"grad_norm": 0.5130940079689026,
"learning_rate": 0.0002401399219190303,
"loss": 3.2330755615234374,
"step": 15450
},
{
"epoch": 0.5975097336263059,
"grad_norm": 0.5438942313194275,
"learning_rate": 0.0002397551114194659,
"loss": 3.24219970703125,
"step": 15500
},
{
"epoch": 0.5994371843799391,
"grad_norm": 0.524650514125824,
"learning_rate": 0.00023936937854726094,
"loss": 3.2422085571289063,
"step": 15550
},
{
"epoch": 0.6013646351335723,
"grad_norm": 0.5230799317359924,
"learning_rate": 0.00023898272726641741,
"loss": 3.223502197265625,
"step": 15600
},
{
"epoch": 0.6032920858872056,
"grad_norm": 0.5275322794914246,
"learning_rate": 0.00023859516155037546,
"loss": 3.225958251953125,
"step": 15650
},
{
"epoch": 0.6052195366408388,
"grad_norm": 0.5163736343383789,
"learning_rate": 0.00023820668538197248,
"loss": 3.20025390625,
"step": 15700
},
{
"epoch": 0.607146987394472,
"grad_norm": 0.5169758796691895,
"learning_rate": 0.00023781730275340216,
"loss": 3.2232843017578126,
"step": 15750
},
{
"epoch": 0.6090744381481054,
"grad_norm": 0.5409614443778992,
"learning_rate": 0.00023742701766617344,
"loss": 3.2230377197265625,
"step": 15800
},
{
"epoch": 0.6110018889017386,
"grad_norm": 0.5149165391921997,
"learning_rate": 0.0002370358341310695,
"loss": 3.2072076416015625,
"step": 15850
},
{
"epoch": 0.6129293396553718,
"grad_norm": 0.5395666360855103,
"learning_rate": 0.00023664375616810638,
"loss": 3.2132049560546876,
"step": 15900
},
{
"epoch": 0.6148567904090051,
"grad_norm": 0.5156650543212891,
"learning_rate": 0.00023625078780649178,
"loss": 3.2226068115234376,
"step": 15950
},
{
"epoch": 0.6167842411626383,
"grad_norm": 0.5319775938987732,
"learning_rate": 0.0002358569330845836,
"loss": 3.218251953125,
"step": 16000
},
{
"epoch": 0.6167842411626383,
"eval_loss": 3.1663146018981934,
"eval_runtime": 63.6923,
"eval_samples_per_second": 130.973,
"eval_steps_per_second": 4.098,
"step": 16000
},
{
"epoch": 0.6187116919162715,
"grad_norm": 0.5353730320930481,
"learning_rate": 0.0002354700994109862,
"loss": 3.2148919677734376,
"step": 16050
},
{
"epoch": 0.6206391426699048,
"grad_norm": 0.5472115874290466,
"learning_rate": 0.00023507450164527135,
"loss": 3.235498962402344,
"step": 16100
},
{
"epoch": 0.622566593423538,
"grad_norm": 0.5607327818870544,
"learning_rate": 0.0002346780296074233,
"loss": 3.2080108642578127,
"step": 16150
},
{
"epoch": 0.6244940441771712,
"grad_norm": 0.5322075486183167,
"learning_rate": 0.00023428068737180557,
"loss": 3.22909423828125,
"step": 16200
},
{
"epoch": 0.6264214949308046,
"grad_norm": 0.5252315998077393,
"learning_rate": 0.00023388247902172437,
"loss": 3.2227642822265623,
"step": 16250
},
{
"epoch": 0.6283489456844378,
"grad_norm": 0.5412209630012512,
"learning_rate": 0.00023348340864938672,
"loss": 3.2188629150390624,
"step": 16300
},
{
"epoch": 0.630276396438071,
"grad_norm": 0.5388709306716919,
"learning_rate": 0.0002330834803558581,
"loss": 3.2186550903320312,
"step": 16350
},
{
"epoch": 0.6322038471917043,
"grad_norm": 0.5240402817726135,
"learning_rate": 0.00023268269825102049,
"loss": 3.219837646484375,
"step": 16400
},
{
"epoch": 0.6341312979453375,
"grad_norm": 0.5506020784378052,
"learning_rate": 0.00023228106645353013,
"loss": 3.21667236328125,
"step": 16450
},
{
"epoch": 0.6360587486989707,
"grad_norm": 0.5462612509727478,
"learning_rate": 0.0002318785890907751,
"loss": 3.2001947021484374,
"step": 16500
},
{
"epoch": 0.637986199452604,
"grad_norm": 0.5411362051963806,
"learning_rate": 0.00023147527029883312,
"loss": 3.22234130859375,
"step": 16550
},
{
"epoch": 0.6399136502062373,
"grad_norm": 0.5390626192092896,
"learning_rate": 0.00023107111422242875,
"loss": 3.217271423339844,
"step": 16600
},
{
"epoch": 0.6418411009598705,
"grad_norm": 0.5203182697296143,
"learning_rate": 0.00023066612501489097,
"loss": 3.1918304443359373,
"step": 16650
},
{
"epoch": 0.6437685517135037,
"grad_norm": 0.5306735038757324,
"learning_rate": 0.0002302603068381105,
"loss": 3.2123394775390626,
"step": 16700
},
{
"epoch": 0.645696002467137,
"grad_norm": 0.5205233097076416,
"learning_rate": 0.00022985366386249703,
"loss": 3.1963800048828124,
"step": 16750
},
{
"epoch": 0.6476234532207702,
"grad_norm": 0.518959641456604,
"learning_rate": 0.00022944620026693627,
"loss": 3.19573486328125,
"step": 16800
},
{
"epoch": 0.6495509039744034,
"grad_norm": 0.551644504070282,
"learning_rate": 0.00022903792023874713,
"loss": 3.2187109375,
"step": 16850
},
{
"epoch": 0.6514783547280367,
"grad_norm": 0.5383501648902893,
"learning_rate": 0.00022862882797363852,
"loss": 3.209862060546875,
"step": 16900
},
{
"epoch": 0.6534058054816699,
"grad_norm": 0.5318331122398376,
"learning_rate": 0.00022821892767566648,
"loss": 3.2025387573242186,
"step": 16950
},
{
"epoch": 0.6553332562353031,
"grad_norm": 0.5309449434280396,
"learning_rate": 0.00022780822355719085,
"loss": 3.198634033203125,
"step": 17000
},
{
"epoch": 0.6572607069889365,
"grad_norm": 0.556352436542511,
"learning_rate": 0.00022739671983883184,
"loss": 3.2343719482421873,
"step": 17050
},
{
"epoch": 0.6591881577425697,
"grad_norm": 0.5366655588150024,
"learning_rate": 0.00022698442074942676,
"loss": 3.205378112792969,
"step": 17100
},
{
"epoch": 0.6611156084962029,
"grad_norm": 0.5222042202949524,
"learning_rate": 0.00022657133052598672,
"loss": 3.2009152221679686,
"step": 17150
},
{
"epoch": 0.6630430592498362,
"grad_norm": 0.541424572467804,
"learning_rate": 0.00022615745341365292,
"loss": 3.224151611328125,
"step": 17200
},
{
"epoch": 0.6649705100034694,
"grad_norm": 0.5297459959983826,
"learning_rate": 0.0002257427936656531,
"loss": 3.2087701416015624,
"step": 17250
},
{
"epoch": 0.6668979607571026,
"grad_norm": 0.5537073016166687,
"learning_rate": 0.0002253273555432577,
"loss": 3.20187255859375,
"step": 17300
},
{
"epoch": 0.6688254115107359,
"grad_norm": 0.5436522960662842,
"learning_rate": 0.00022491114331573615,
"loss": 3.218807373046875,
"step": 17350
},
{
"epoch": 0.6707528622643691,
"grad_norm": 0.5424281358718872,
"learning_rate": 0.00022449416126031307,
"loss": 3.193819885253906,
"step": 17400
},
{
"epoch": 0.6726803130180024,
"grad_norm": 0.5302148461341858,
"learning_rate": 0.0002240764136621243,
"loss": 3.2095501708984373,
"step": 17450
},
{
"epoch": 0.6746077637716357,
"grad_norm": 0.5466500520706177,
"learning_rate": 0.00022365790481417274,
"loss": 3.2022940063476564,
"step": 17500
},
{
"epoch": 0.6765352145252689,
"grad_norm": 0.5302472114562988,
"learning_rate": 0.00022323863901728432,
"loss": 3.202532958984375,
"step": 17550
},
{
"epoch": 0.6784626652789021,
"grad_norm": 0.533517062664032,
"learning_rate": 0.00022281862058006377,
"loss": 3.196273193359375,
"step": 17600
},
{
"epoch": 0.6803901160325354,
"grad_norm": 0.5435385704040527,
"learning_rate": 0.00022239785381885044,
"loss": 3.209317626953125,
"step": 17650
},
{
"epoch": 0.6823175667861686,
"grad_norm": 0.5550362467765808,
"learning_rate": 0.0002219763430576738,
"loss": 3.178984375,
"step": 17700
},
{
"epoch": 0.6842450175398018,
"grad_norm": 0.5522629618644714,
"learning_rate": 0.00022155409262820909,
"loss": 3.2081500244140626,
"step": 17750
},
{
"epoch": 0.6861724682934351,
"grad_norm": 0.5408194661140442,
"learning_rate": 0.00022113110686973277,
"loss": 3.2006549072265624,
"step": 17800
},
{
"epoch": 0.6880999190470684,
"grad_norm": 0.5404806733131409,
"learning_rate": 0.00022070739012907797,
"loss": 3.1964431762695313,
"step": 17850
},
{
"epoch": 0.6900273698007016,
"grad_norm": 0.5439943075180054,
"learning_rate": 0.0002202829467605898,
"loss": 3.204049072265625,
"step": 17900
},
{
"epoch": 0.6919548205543349,
"grad_norm": 0.5549587607383728,
"learning_rate": 0.00021985778112608057,
"loss": 3.1962115478515627,
"step": 17950
},
{
"epoch": 0.6938822713079681,
"grad_norm": 0.5523220896720886,
"learning_rate": 0.00021943189759478503,
"loss": 3.2010546875,
"step": 18000
},
{
"epoch": 0.6938822713079681,
"eval_loss": 3.140258550643921,
"eval_runtime": 70.1344,
"eval_samples_per_second": 118.943,
"eval_steps_per_second": 3.721,
"step": 18000
},
{
"epoch": 0.6958097220616013,
"grad_norm": 0.5462304353713989,
"learning_rate": 0.0002190053005433153,
"loss": 3.217257080078125,
"step": 18050
},
{
"epoch": 0.6977371728152346,
"grad_norm": 0.5415950417518616,
"learning_rate": 0.00021857799435561626,
"loss": 3.186925048828125,
"step": 18100
},
{
"epoch": 0.6996646235688678,
"grad_norm": 0.5496586561203003,
"learning_rate": 0.00021814998342291995,
"loss": 3.184859619140625,
"step": 18150
},
{
"epoch": 0.701592074322501,
"grad_norm": 0.5240462422370911,
"learning_rate": 0.00021772127214370109,
"loss": 3.1860537719726563,
"step": 18200
},
{
"epoch": 0.7035195250761344,
"grad_norm": 0.5483691096305847,
"learning_rate": 0.0002173004598597227,
"loss": 3.196702880859375,
"step": 18250
},
{
"epoch": 0.7054469758297676,
"grad_norm": 0.5493341684341431,
"learning_rate": 0.0002168703748988944,
"loss": 3.202568359375,
"step": 18300
},
{
"epoch": 0.7073744265834008,
"grad_norm": 0.5448978543281555,
"learning_rate": 0.00021643960274150096,
"loss": 3.1949456787109374,
"step": 18350
},
{
"epoch": 0.7093018773370341,
"grad_norm": 0.5448238849639893,
"learning_rate": 0.00021600814781439276,
"loss": 3.1895770263671874,
"step": 18400
},
{
"epoch": 0.7112293280906673,
"grad_norm": 0.5450513362884521,
"learning_rate": 0.00021557601455143667,
"loss": 3.1710748291015625,
"step": 18450
},
{
"epoch": 0.7131567788443005,
"grad_norm": 0.5654709935188293,
"learning_rate": 0.0002151432073934707,
"loss": 3.1629629516601563,
"step": 18500
},
{
"epoch": 0.7150842295979337,
"grad_norm": 0.533403754234314,
"learning_rate": 0.0002147097307882579,
"loss": 3.178038330078125,
"step": 18550
},
{
"epoch": 0.717011680351567,
"grad_norm": 0.546596884727478,
"learning_rate": 0.00021427558919044114,
"loss": 3.1864089965820312,
"step": 18600
},
{
"epoch": 0.7189391311052002,
"grad_norm": 0.537943959236145,
"learning_rate": 0.00021384078706149706,
"loss": 3.172017822265625,
"step": 18650
},
{
"epoch": 0.7208665818588335,
"grad_norm": 0.5493637323379517,
"learning_rate": 0.00021340532886969025,
"loss": 3.1849749755859373,
"step": 18700
},
{
"epoch": 0.7227940326124668,
"grad_norm": 0.5344792604446411,
"learning_rate": 0.0002129692190900274,
"loss": 3.183013916015625,
"step": 18750
},
{
"epoch": 0.7247214833661,
"grad_norm": 0.5494260787963867,
"learning_rate": 0.00021253246220421122,
"loss": 3.1716021728515624,
"step": 18800
},
{
"epoch": 0.7266489341197332,
"grad_norm": 0.5650742053985596,
"learning_rate": 0.00021209506270059448,
"loss": 3.163990478515625,
"step": 18850
},
{
"epoch": 0.7285763848733665,
"grad_norm": 0.5641888976097107,
"learning_rate": 0.00021165702507413385,
"loss": 3.1848876953125,
"step": 18900
},
{
"epoch": 0.7305038356269997,
"grad_norm": 0.5492480397224426,
"learning_rate": 0.00021121835382634365,
"loss": 3.180323486328125,
"step": 18950
},
{
"epoch": 0.7324312863806329,
"grad_norm": 0.5651627779006958,
"learning_rate": 0.00021077905346524975,
"loss": 3.1726800537109376,
"step": 19000
},
{
"epoch": 0.7343587371342662,
"grad_norm": 0.5526878833770752,
"learning_rate": 0.00021033912850534296,
"loss": 3.161160888671875,
"step": 19050
},
{
"epoch": 0.7362861878878995,
"grad_norm": 0.5558029413223267,
"learning_rate": 0.00020989858346753292,
"loss": 3.1849957275390626,
"step": 19100
},
{
"epoch": 0.7382136386415327,
"grad_norm": 0.5392846465110779,
"learning_rate": 0.0002094574228791016,
"loss": 3.173941650390625,
"step": 19150
},
{
"epoch": 0.740141089395166,
"grad_norm": 0.5676681995391846,
"learning_rate": 0.0002090156512736565,
"loss": 3.1795391845703125,
"step": 19200
},
{
"epoch": 0.7420685401487992,
"grad_norm": 0.5614210963249207,
"learning_rate": 0.00020857327319108446,
"loss": 3.1860287475585936,
"step": 19250
},
{
"epoch": 0.7439959909024324,
"grad_norm": 0.5414462685585022,
"learning_rate": 0.0002081302931775047,
"loss": 3.178353576660156,
"step": 19300
},
{
"epoch": 0.7459234416560657,
"grad_norm": 0.5587648749351501,
"learning_rate": 0.00020768671578522225,
"loss": 3.155638427734375,
"step": 19350
},
{
"epoch": 0.7478508924096989,
"grad_norm": 0.5393547415733337,
"learning_rate": 0.0002072425455726812,
"loss": 3.1642330932617186,
"step": 19400
},
{
"epoch": 0.7497783431633321,
"grad_norm": 0.5882778167724609,
"learning_rate": 0.00020679778710441758,
"loss": 3.1766650390625,
"step": 19450
},
{
"epoch": 0.7517057939169655,
"grad_norm": 0.5372825860977173,
"learning_rate": 0.0002063524449510129,
"loss": 3.1782821655273437,
"step": 19500
},
{
"epoch": 0.7536332446705987,
"grad_norm": 0.5670105814933777,
"learning_rate": 0.00020590652368904672,
"loss": 3.17004150390625,
"step": 19550
},
{
"epoch": 0.7555606954242319,
"grad_norm": 0.534831166267395,
"learning_rate": 0.00020546002790105004,
"loss": 3.1736834716796873,
"step": 19600
},
{
"epoch": 0.7574881461778652,
"grad_norm": 0.5483341813087463,
"learning_rate": 0.00020501296217545787,
"loss": 3.1846588134765623,
"step": 19650
},
{
"epoch": 0.7594155969314984,
"grad_norm": 0.5427250862121582,
"learning_rate": 0.00020456533110656226,
"loss": 3.158591003417969,
"step": 19700
},
{
"epoch": 0.7613430476851316,
"grad_norm": 0.5472317934036255,
"learning_rate": 0.00020411713929446504,
"loss": 3.177190246582031,
"step": 19750
},
{
"epoch": 0.7632704984387649,
"grad_norm": 0.5449865460395813,
"learning_rate": 0.0002036683913450305,
"loss": 3.1783905029296875,
"step": 19800
},
{
"epoch": 0.7651979491923981,
"grad_norm": 0.581668496131897,
"learning_rate": 0.00020321909186983816,
"loss": 3.1630877685546874,
"step": 19850
},
{
"epoch": 0.7671253999460313,
"grad_norm": 0.5557700395584106,
"learning_rate": 0.00020276924548613524,
"loss": 3.181783447265625,
"step": 19900
},
{
"epoch": 0.7690528506996647,
"grad_norm": 0.5492613911628723,
"learning_rate": 0.00020231885681678942,
"loss": 3.1710394287109374,
"step": 19950
},
{
"epoch": 0.7709803014532979,
"grad_norm": 0.545788586139679,
"learning_rate": 0.00020186793049024103,
"loss": 3.1749493408203127,
"step": 20000
},
{
"epoch": 0.7709803014532979,
"eval_loss": 3.1167919635772705,
"eval_runtime": 70.0002,
"eval_samples_per_second": 119.171,
"eval_steps_per_second": 3.729,
"step": 20000
},
{
"epoch": 0.7729077522069311,
"grad_norm": 0.5734344124794006,
"learning_rate": 0.0002014164711404559,
"loss": 3.1556051635742186,
"step": 20050
},
{
"epoch": 0.7748352029605644,
"grad_norm": 0.5648406147956848,
"learning_rate": 0.00020096448340687723,
"loss": 3.1638214111328127,
"step": 20100
},
{
"epoch": 0.7767626537141976,
"grad_norm": 0.5655689239501953,
"learning_rate": 0.00020051197193437834,
"loss": 3.155677490234375,
"step": 20150
},
{
"epoch": 0.7786901044678308,
"grad_norm": 0.55703204870224,
"learning_rate": 0.00020005894137321473,
"loss": 3.187183837890625,
"step": 20200
},
{
"epoch": 0.7806175552214641,
"grad_norm": 0.566350519657135,
"learning_rate": 0.00019960539637897638,
"loss": 3.1689105224609375,
"step": 20250
},
{
"epoch": 0.7825450059750974,
"grad_norm": 0.5704211592674255,
"learning_rate": 0.00019915134161253983,
"loss": 3.1697052001953123,
"step": 20300
},
{
"epoch": 0.7844724567287306,
"grad_norm": 0.5612184405326843,
"learning_rate": 0.00019869678174002037,
"loss": 3.1789315795898436,
"step": 20350
},
{
"epoch": 0.7863999074823639,
"grad_norm": 0.5699100494384766,
"learning_rate": 0.0001982417214327239,
"loss": 3.160164794921875,
"step": 20400
},
{
"epoch": 0.7883273582359971,
"grad_norm": 0.5764229893684387,
"learning_rate": 0.00019778616536709923,
"loss": 3.1545220947265626,
"step": 20450
},
{
"epoch": 0.7902548089896303,
"grad_norm": 0.5840250253677368,
"learning_rate": 0.0001973301182246898,
"loss": 3.1714324951171875,
"step": 20500
},
{
"epoch": 0.7921822597432635,
"grad_norm": 0.5466107130050659,
"learning_rate": 0.0001968827200990245,
"loss": 3.144857177734375,
"step": 20550
},
{
"epoch": 0.7941097104968968,
"grad_norm": 0.5752685070037842,
"learning_rate": 0.0001964257144557763,
"loss": 3.1514706420898437,
"step": 20600
},
{
"epoch": 0.79603716125053,
"grad_norm": 0.5478172898292542,
"learning_rate": 0.0001959682317164813,
"loss": 3.1699078369140623,
"step": 20650
},
{
"epoch": 0.7979646120041632,
"grad_norm": 0.5521666407585144,
"learning_rate": 0.0001955102765824825,
"loss": 3.152471008300781,
"step": 20700
},
{
"epoch": 0.7998920627577966,
"grad_norm": 0.5669108033180237,
"learning_rate": 0.00019505185375997747,
"loss": 3.1632684326171874,
"step": 20750
},
{
"epoch": 0.8018195135114298,
"grad_norm": 0.561879575252533,
"learning_rate": 0.00019459296795997,
"loss": 3.16443115234375,
"step": 20800
},
{
"epoch": 0.803746964265063,
"grad_norm": 0.5720276832580566,
"learning_rate": 0.00019413362389822167,
"loss": 3.156514892578125,
"step": 20850
},
{
"epoch": 0.8056744150186963,
"grad_norm": 0.5463959574699402,
"learning_rate": 0.00019367382629520338,
"loss": 3.143111267089844,
"step": 20900
},
{
"epoch": 0.8076018657723295,
"grad_norm": 0.565447986125946,
"learning_rate": 0.00019321357987604694,
"loss": 3.1429367065429688,
"step": 20950
},
{
"epoch": 0.8095293165259627,
"grad_norm": 0.5889939069747925,
"learning_rate": 0.0001927528893704964,
"loss": 3.1348284912109374,
"step": 21000
},
{
"epoch": 0.811456767279596,
"grad_norm": 0.5574634671211243,
"learning_rate": 0.00019229175951285954,
"loss": 3.1616162109375,
"step": 21050
},
{
"epoch": 0.8133842180332292,
"grad_norm": 0.5584666132926941,
"learning_rate": 0.00019183019504195902,
"loss": 3.154088134765625,
"step": 21100
},
{
"epoch": 0.8153116687868625,
"grad_norm": 0.5621055364608765,
"learning_rate": 0.00019136820070108404,
"loss": 3.1489691162109374,
"step": 21150
},
{
"epoch": 0.8172391195404958,
"grad_norm": 0.5915434956550598,
"learning_rate": 0.00019090578123794115,
"loss": 3.152672119140625,
"step": 21200
},
{
"epoch": 0.819166570294129,
"grad_norm": 0.5811152458190918,
"learning_rate": 0.0001904429414046059,
"loss": 3.1727081298828126,
"step": 21250
},
{
"epoch": 0.8210940210477622,
"grad_norm": 0.5779080390930176,
"learning_rate": 0.00018997968595747358,
"loss": 3.1502005004882814,
"step": 21300
},
{
"epoch": 0.8230214718013955,
"grad_norm": 0.5683503746986389,
"learning_rate": 0.00018951601965721076,
"loss": 3.1443341064453123,
"step": 21350
},
{
"epoch": 0.8249489225550287,
"grad_norm": 0.563933789730072,
"learning_rate": 0.00018905194726870602,
"loss": 3.133697509765625,
"step": 21400
},
{
"epoch": 0.8268763733086619,
"grad_norm": 0.5592430830001831,
"learning_rate": 0.0001885874735610212,
"loss": 3.161279296875,
"step": 21450
},
{
"epoch": 0.8288038240622952,
"grad_norm": 0.5582791566848755,
"learning_rate": 0.00018812260330734227,
"loss": 3.1484185791015626,
"step": 21500
},
{
"epoch": 0.8307312748159285,
"grad_norm": 0.5692592859268188,
"learning_rate": 0.00018765734128493034,
"loss": 3.1613739013671873,
"step": 21550
},
{
"epoch": 0.8326587255695617,
"grad_norm": 0.5692372918128967,
"learning_rate": 0.00018719169227507257,
"loss": 3.1404452514648438,
"step": 21600
},
{
"epoch": 0.834586176323195,
"grad_norm": 0.5661136507987976,
"learning_rate": 0.00018672566106303295,
"loss": 3.1603948974609377,
"step": 21650
},
{
"epoch": 0.8365136270768282,
"grad_norm": 0.562249481678009,
"learning_rate": 0.0001862592524380033,
"loss": 3.1410427856445313,
"step": 21700
},
{
"epoch": 0.8384410778304614,
"grad_norm": 0.5753993988037109,
"learning_rate": 0.00018579247119305384,
"loss": 3.144683837890625,
"step": 21750
},
{
"epoch": 0.8403685285840947,
"grad_norm": 0.558670163154602,
"learning_rate": 0.00018532532212508409,
"loss": 3.1535107421875,
"step": 21800
},
{
"epoch": 0.8422959793377279,
"grad_norm": 0.5808515548706055,
"learning_rate": 0.00018485781003477352,
"loss": 3.161909484863281,
"step": 21850
},
{
"epoch": 0.8442234300913611,
"grad_norm": 0.5838216543197632,
"learning_rate": 0.00018438993972653213,
"loss": 3.1452462768554685,
"step": 21900
},
{
"epoch": 0.8461508808449945,
"grad_norm": 0.5656279921531677,
"learning_rate": 0.00018392171600845137,
"loss": 3.138179931640625,
"step": 21950
},
{
"epoch": 0.8480783315986277,
"grad_norm": 0.5472198724746704,
"learning_rate": 0.00018345314369225428,
"loss": 3.14174560546875,
"step": 22000
},
{
"epoch": 0.8480783315986277,
"eval_loss": 3.092756748199463,
"eval_runtime": 69.6645,
"eval_samples_per_second": 119.745,
"eval_steps_per_second": 3.747,
"step": 22000
},
{
"epoch": 0.8500057823522609,
"grad_norm": 0.5661357641220093,
"learning_rate": 0.00018298422759324638,
"loss": 3.1307247924804686,
"step": 22050
},
{
"epoch": 0.8519332331058942,
"grad_norm": 0.5698456168174744,
"learning_rate": 0.00018251497253026614,
"loss": 3.1494296264648436,
"step": 22100
},
{
"epoch": 0.8538606838595274,
"grad_norm": 0.5618436932563782,
"learning_rate": 0.00018204538332563546,
"loss": 3.13890380859375,
"step": 22150
},
{
"epoch": 0.8557881346131606,
"grad_norm": 0.5600304007530212,
"learning_rate": 0.0001815754648051098,
"loss": 3.1329217529296876,
"step": 22200
},
{
"epoch": 0.8577155853667939,
"grad_norm": 0.5475085377693176,
"learning_rate": 0.00018110522179782913,
"loss": 3.133092041015625,
"step": 22250
},
{
"epoch": 0.8596430361204271,
"grad_norm": 0.5875421166419983,
"learning_rate": 0.00018063465913626786,
"loss": 3.1294207763671875,
"step": 22300
},
{
"epoch": 0.8615704868740603,
"grad_norm": 0.5569576621055603,
"learning_rate": 0.00018016378165618553,
"loss": 3.1345751953125,
"step": 22350
},
{
"epoch": 0.8634979376276936,
"grad_norm": 0.5917547941207886,
"learning_rate": 0.00017969259419657667,
"loss": 3.15324951171875,
"step": 22400
},
{
"epoch": 0.8654253883813269,
"grad_norm": 0.5652753114700317,
"learning_rate": 0.0001792211015996215,
"loss": 3.1455862426757815,
"step": 22450
},
{
"epoch": 0.8673528391349601,
"grad_norm": 0.5713505148887634,
"learning_rate": 0.00017874930871063596,
"loss": 3.1215853881835938,
"step": 22500
},
{
"epoch": 0.8692802898885933,
"grad_norm": 0.5742377638816833,
"learning_rate": 0.00017827722037802194,
"loss": 3.1541537475585937,
"step": 22550
},
{
"epoch": 0.8712077406422266,
"grad_norm": 0.5754672884941101,
"learning_rate": 0.00017780484145321756,
"loss": 3.138553466796875,
"step": 22600
},
{
"epoch": 0.8731351913958598,
"grad_norm": 0.5959605574607849,
"learning_rate": 0.00017733217679064697,
"loss": 3.12017822265625,
"step": 22650
},
{
"epoch": 0.875062642149493,
"grad_norm": 0.5881737470626831,
"learning_rate": 0.00017685923124767102,
"loss": 3.130048828125,
"step": 22700
},
{
"epoch": 0.8769900929031264,
"grad_norm": 0.5836317539215088,
"learning_rate": 0.00017638600968453692,
"loss": 3.133804931640625,
"step": 22750
},
{
"epoch": 0.8789175436567596,
"grad_norm": 0.577653706073761,
"learning_rate": 0.00017591251696432838,
"loss": 3.133560791015625,
"step": 22800
},
{
"epoch": 0.8808449944103928,
"grad_norm": 0.5812749862670898,
"learning_rate": 0.00017543875795291566,
"loss": 3.1333111572265624,
"step": 22850
},
{
"epoch": 0.8827724451640261,
"grad_norm": 0.5724690556526184,
"learning_rate": 0.00017496473751890568,
"loss": 3.121082763671875,
"step": 22900
},
{
"epoch": 0.8846998959176593,
"grad_norm": 0.5756459832191467,
"learning_rate": 0.00017449046053359178,
"loss": 3.1173480224609373,
"step": 22950
},
{
"epoch": 0.8866273466712925,
"grad_norm": 0.5805308818817139,
"learning_rate": 0.00017402542487905876,
"loss": 3.1197103881835937,
"step": 23000
},
{
"epoch": 0.8885547974249258,
"grad_norm": 0.5663387775421143,
"learning_rate": 0.00017355065430372369,
"loss": 3.143828125,
"step": 23050
},
{
"epoch": 0.890482248178559,
"grad_norm": 0.5675112009048462,
"learning_rate": 0.00017307564170897767,
"loss": 3.1355856323242186,
"step": 23100
},
{
"epoch": 0.8924096989321922,
"grad_norm": 0.5664778351783752,
"learning_rate": 0.00017260039197630997,
"loss": 3.125141906738281,
"step": 23150
},
{
"epoch": 0.8943371496858256,
"grad_norm": 0.578191339969635,
"learning_rate": 0.000172124909989647,
"loss": 3.1383987426757813,
"step": 23200
},
{
"epoch": 0.8962646004394588,
"grad_norm": 0.5867224931716919,
"learning_rate": 0.00017164920063530178,
"loss": 3.1277456665039063,
"step": 23250
},
{
"epoch": 0.898192051193092,
"grad_norm": 0.6003968715667725,
"learning_rate": 0.00017117326880192402,
"loss": 3.134222106933594,
"step": 23300
},
{
"epoch": 0.9001195019467253,
"grad_norm": 0.5659716129302979,
"learning_rate": 0.00017069711938044968,
"loss": 3.1279827880859377,
"step": 23350
},
{
"epoch": 0.9020469527003585,
"grad_norm": 0.5737743973731995,
"learning_rate": 0.00017022075726405065,
"loss": 3.114278564453125,
"step": 23400
},
{
"epoch": 0.9039744034539917,
"grad_norm": 0.5945197939872742,
"learning_rate": 0.00016974418734808482,
"loss": 3.119969482421875,
"step": 23450
},
{
"epoch": 0.905901854207625,
"grad_norm": 0.5680420398712158,
"learning_rate": 0.00016926741453004543,
"loss": 3.124642333984375,
"step": 23500
},
{
"epoch": 0.9078293049612582,
"grad_norm": 0.5707167387008667,
"learning_rate": 0.00016879044370951085,
"loss": 3.1347332763671876,
"step": 23550
},
{
"epoch": 0.9097567557148915,
"grad_norm": 0.5724088549613953,
"learning_rate": 0.00016831327978809418,
"loss": 3.1314816284179687,
"step": 23600
},
{
"epoch": 0.9116842064685248,
"grad_norm": 0.5946201682090759,
"learning_rate": 0.00016783592766939307,
"loss": 3.1320989990234374,
"step": 23650
},
{
"epoch": 0.913611657222158,
"grad_norm": 0.5980028510093689,
"learning_rate": 0.000167358392258939,
"loss": 3.1286688232421875,
"step": 23700
},
{
"epoch": 0.9155391079757912,
"grad_norm": 0.5889047384262085,
"learning_rate": 0.00016688067846414728,
"loss": 3.131365966796875,
"step": 23750
},
{
"epoch": 0.9174665587294245,
"grad_norm": 0.5753546953201294,
"learning_rate": 0.00016640279119426628,
"loss": 3.1349615478515624,
"step": 23800
},
{
"epoch": 0.9193940094830577,
"grad_norm": 0.5793065428733826,
"learning_rate": 0.000165924735360327,
"loss": 3.1140676879882814,
"step": 23850
},
{
"epoch": 0.9213214602366909,
"grad_norm": 0.5712112784385681,
"learning_rate": 0.0001654465158750929,
"loss": 3.125685119628906,
"step": 23900
},
{
"epoch": 0.9232489109903242,
"grad_norm": 0.5831735134124756,
"learning_rate": 0.000164968137653009,
"loss": 3.1148748779296875,
"step": 23950
},
{
"epoch": 0.9251763617439575,
"grad_norm": 0.5710425972938538,
"learning_rate": 0.00016448960561015177,
"loss": 3.1261447143554686,
"step": 24000
},
{
"epoch": 0.9251763617439575,
"eval_loss": 3.0723135471343994,
"eval_runtime": 69.7806,
"eval_samples_per_second": 119.546,
"eval_steps_per_second": 3.74,
"step": 24000
},
{
"epoch": 0.9271038124975907,
"grad_norm": 0.5861913561820984,
"learning_rate": 0.00016401092466417826,
"loss": 3.121595458984375,
"step": 24050
},
{
"epoch": 0.929031263251224,
"grad_norm": 0.5746450424194336,
"learning_rate": 0.00016353209973427588,
"loss": 3.1021661376953125,
"step": 24100
},
{
"epoch": 0.9309587140048572,
"grad_norm": 0.6018072366714478,
"learning_rate": 0.00016305313574111163,
"loss": 3.1265057373046874,
"step": 24150
},
{
"epoch": 0.9328861647584904,
"grad_norm": 0.5745505094528198,
"learning_rate": 0.00016257403760678156,
"loss": 3.1126507568359374,
"step": 24200
},
{
"epoch": 0.9348136155121237,
"grad_norm": 0.5755428671836853,
"learning_rate": 0.00016209481025476026,
"loss": 3.1199249267578124,
"step": 24250
},
{
"epoch": 0.9367410662657569,
"grad_norm": 0.6154507398605347,
"learning_rate": 0.0001616154586098503,
"loss": 3.1147506713867186,
"step": 24300
},
{
"epoch": 0.9386685170193901,
"grad_norm": 0.6009291410446167,
"learning_rate": 0.00016113598759813147,
"loss": 3.1118508911132814,
"step": 24350
},
{
"epoch": 0.9405959677730233,
"grad_norm": 0.62513267993927,
"learning_rate": 0.00016065640214691027,
"loss": 3.1212445068359376,
"step": 24400
},
{
"epoch": 0.9425234185266567,
"grad_norm": 0.5766476988792419,
"learning_rate": 0.00016017670718466922,
"loss": 3.105079345703125,
"step": 24450
},
{
"epoch": 0.9444508692802899,
"grad_norm": 0.5826324820518494,
"learning_rate": 0.00015969690764101627,
"loss": 3.1244757080078127,
"step": 24500
},
{
"epoch": 0.9463783200339231,
"grad_norm": 0.576753556728363,
"learning_rate": 0.00015921700844663403,
"loss": 3.121556396484375,
"step": 24550
},
{
"epoch": 0.9483057707875564,
"grad_norm": 0.6016496419906616,
"learning_rate": 0.0001587370145332293,
"loss": 3.1150927734375,
"step": 24600
},
{
"epoch": 0.9502332215411896,
"grad_norm": 0.5723993182182312,
"learning_rate": 0.00015825693083348214,
"loss": 3.11465087890625,
"step": 24650
},
{
"epoch": 0.9521606722948228,
"grad_norm": 0.5783793330192566,
"learning_rate": 0.00015777676228099545,
"loss": 3.1239398193359373,
"step": 24700
},
{
"epoch": 0.9540881230484561,
"grad_norm": 0.5800228714942932,
"learning_rate": 0.00015729651381024394,
"loss": 3.094305419921875,
"step": 24750
},
{
"epoch": 0.9560155738020893,
"grad_norm": 0.5899364352226257,
"learning_rate": 0.00015681619035652367,
"loss": 3.1132583618164062,
"step": 24800
},
{
"epoch": 0.9579430245557226,
"grad_norm": 0.5847409963607788,
"learning_rate": 0.00015633579685590143,
"loss": 3.1175335693359374,
"step": 24850
},
{
"epoch": 0.9598704753093559,
"grad_norm": 0.6097803711891174,
"learning_rate": 0.00015585533824516359,
"loss": 3.0956768798828125,
"step": 24900
},
{
"epoch": 0.9617979260629891,
"grad_norm": 0.5999714732170105,
"learning_rate": 0.00015537481946176579,
"loss": 3.1311724853515623,
"step": 24950
},
{
"epoch": 0.9637253768166223,
"grad_norm": 0.5835265517234802,
"learning_rate": 0.000154894245443782,
"loss": 3.110315856933594,
"step": 25000
},
{
"epoch": 0.9656528275702556,
"grad_norm": 0.5831000208854675,
"learning_rate": 0.0001544136211298538,
"loss": 3.1094940185546873,
"step": 25050
},
{
"epoch": 0.9675802783238888,
"grad_norm": 0.5802441835403442,
"learning_rate": 0.00015393295145913968,
"loss": 3.1400872802734376,
"step": 25100
},
{
"epoch": 0.969507729077522,
"grad_norm": 0.5813797116279602,
"learning_rate": 0.00015345224137126417,
"loss": 3.089359130859375,
"step": 25150
},
{
"epoch": 0.9714351798311553,
"grad_norm": 0.5961971879005432,
"learning_rate": 0.0001529714958062672,
"loss": 3.1206634521484373,
"step": 25200
},
{
"epoch": 0.9733626305847886,
"grad_norm": 0.5971590876579285,
"learning_rate": 0.00015249071970455322,
"loss": 3.1062298583984376,
"step": 25250
},
{
"epoch": 0.9752900813384218,
"grad_norm": 0.6121759414672852,
"learning_rate": 0.00015200991800684064,
"loss": 3.1030987548828124,
"step": 25300
},
{
"epoch": 0.9772175320920551,
"grad_norm": 0.5767819881439209,
"learning_rate": 0.00015152909565411074,
"loss": 3.107874755859375,
"step": 25350
},
{
"epoch": 0.9791449828456883,
"grad_norm": 0.5813241004943848,
"learning_rate": 0.00015104825758755717,
"loss": 3.0926480102539062,
"step": 25400
},
{
"epoch": 0.9810724335993215,
"grad_norm": 0.5876830220222473,
"learning_rate": 0.000150567408748535,
"loss": 3.102254638671875,
"step": 25450
},
{
"epoch": 0.9829998843529548,
"grad_norm": 0.6042032837867737,
"learning_rate": 0.00015008655407851003,
"loss": 3.104966735839844,
"step": 25500
},
{
"epoch": 0.984927335106588,
"grad_norm": 0.6113137006759644,
"learning_rate": 0.00014960569851900794,
"loss": 3.116665954589844,
"step": 25550
},
{
"epoch": 0.9868547858602212,
"grad_norm": 0.5809758305549622,
"learning_rate": 0.0001491248470115637,
"loss": 3.10265380859375,
"step": 25600
},
{
"epoch": 0.9887822366138546,
"grad_norm": 0.5843464136123657,
"learning_rate": 0.0001486440044976704,
"loss": 3.1052166748046877,
"step": 25650
},
{
"epoch": 0.9907096873674878,
"grad_norm": 0.5919933915138245,
"learning_rate": 0.000148163175918729,
"loss": 3.0959814453125,
"step": 25700
},
{
"epoch": 0.992637138121121,
"grad_norm": 0.6088231801986694,
"learning_rate": 0.00014768236621599693,
"loss": 3.1056008911132813,
"step": 25750
},
{
"epoch": 0.9945645888747543,
"grad_norm": 0.596182107925415,
"learning_rate": 0.0001472015803305379,
"loss": 3.1014004516601563,
"step": 25800
},
{
"epoch": 0.9964920396283875,
"grad_norm": 0.5897939205169678,
"learning_rate": 0.00014672082320317076,
"loss": 3.108145751953125,
"step": 25850
},
{
"epoch": 0.9984194903820207,
"grad_norm": 0.6012491583824158,
"learning_rate": 0.00014624009977441877,
"loss": 3.12437255859375,
"step": 25900
},
{
"epoch": 1.000346941135654,
"grad_norm": 0.6032795906066895,
"learning_rate": 0.000145759414984459,
"loss": 3.113802795410156,
"step": 25950
},
{
"epoch": 1.0022743918892871,
"grad_norm": 0.5917011499404907,
"learning_rate": 0.0001452787737730714,
"loss": 3.0812371826171874,
"step": 26000
},
{
"epoch": 1.0022743918892871,
"eval_loss": 3.0529704093933105,
"eval_runtime": 70.2028,
"eval_samples_per_second": 118.827,
"eval_steps_per_second": 3.718,
"step": 26000
},
{
"epoch": 1.0042018426429205,
"grad_norm": 0.5938454866409302,
"learning_rate": 0.00014479818107958803,
"loss": 3.08142333984375,
"step": 26050
},
{
"epoch": 1.0061292933965538,
"grad_norm": 0.6021835207939148,
"learning_rate": 0.00014431764184284242,
"loss": 3.0975140380859374,
"step": 26100
},
{
"epoch": 1.0080567441501869,
"grad_norm": 0.5974454283714294,
"learning_rate": 0.00014383716100111878,
"loss": 3.08563720703125,
"step": 26150
},
{
"epoch": 1.0099841949038202,
"grad_norm": 0.6034616231918335,
"learning_rate": 0.00014335674349210106,
"loss": 3.084225769042969,
"step": 26200
},
{
"epoch": 1.0119116456574535,
"grad_norm": 0.5942801833152771,
"learning_rate": 0.0001428763942528225,
"loss": 3.0716229248046876,
"step": 26250
},
{
"epoch": 1.0138390964110866,
"grad_norm": 0.6265310645103455,
"learning_rate": 0.00014239611821961483,
"loss": 3.0716766357421874,
"step": 26300
},
{
"epoch": 1.01576654716472,
"grad_norm": 0.596889317035675,
"learning_rate": 0.00014191592032805728,
"loss": 3.08568115234375,
"step": 26350
},
{
"epoch": 1.0176939979183532,
"grad_norm": 0.5878603458404541,
"learning_rate": 0.00014143580551292622,
"loss": 3.0839080810546875,
"step": 26400
},
{
"epoch": 1.0196214486719863,
"grad_norm": 0.6099098920822144,
"learning_rate": 0.0001409557787081442,
"loss": 3.0745391845703125,
"step": 26450
},
{
"epoch": 1.0215488994256197,
"grad_norm": 0.5866267681121826,
"learning_rate": 0.00014047584484672932,
"loss": 3.0808059692382814,
"step": 26500
},
{
"epoch": 1.023476350179253,
"grad_norm": 0.6030349135398865,
"learning_rate": 0.0001400056045893893,
"loss": 3.100005798339844,
"step": 26550
},
{
"epoch": 1.025403800932886,
"grad_norm": 0.614894449710846,
"learning_rate": 0.00013952586930544457,
"loss": 3.0686688232421875,
"step": 26600
},
{
"epoch": 1.0273312516865194,
"grad_norm": 0.5909005403518677,
"learning_rate": 0.00013904624165939856,
"loss": 3.072357177734375,
"step": 26650
},
{
"epoch": 1.0292587024401527,
"grad_norm": 0.6144493818283081,
"learning_rate": 0.00013856672658016728,
"loss": 3.083033447265625,
"step": 26700
},
{
"epoch": 1.0311861531937858,
"grad_norm": 0.5897349119186401,
"learning_rate": 0.00013808732899551015,
"loss": 3.078177795410156,
"step": 26750
},
{
"epoch": 1.0331136039474191,
"grad_norm": 0.5989056825637817,
"learning_rate": 0.0001376080538319791,
"loss": 3.10062255859375,
"step": 26800
},
{
"epoch": 1.0350410547010525,
"grad_norm": 0.6002673506736755,
"learning_rate": 0.000137128906014868,
"loss": 3.0752838134765623,
"step": 26850
},
{
"epoch": 1.0369685054546856,
"grad_norm": 0.5887547731399536,
"learning_rate": 0.000136649890468162,
"loss": 3.0672976684570314,
"step": 26900
},
{
"epoch": 1.0388959562083189,
"grad_norm": 0.6285821199417114,
"learning_rate": 0.00013617101211448697,
"loss": 3.080418701171875,
"step": 26950
},
{
"epoch": 1.0408234069619522,
"grad_norm": 0.6109331846237183,
"learning_rate": 0.00013569227587505894,
"loss": 3.0784490966796874,
"step": 27000
},
{
"epoch": 1.0427508577155853,
"grad_norm": 0.6052854061126709,
"learning_rate": 0.00013521368666963345,
"loss": 3.0815579223632814,
"step": 27050
},
{
"epoch": 1.0446783084692186,
"grad_norm": 0.6137946844100952,
"learning_rate": 0.00013473524941645516,
"loss": 3.06681396484375,
"step": 27100
},
{
"epoch": 1.046605759222852,
"grad_norm": 0.6042759418487549,
"learning_rate": 0.00013425696903220699,
"loss": 3.090870056152344,
"step": 27150
},
{
"epoch": 1.048533209976485,
"grad_norm": 0.6349405646324158,
"learning_rate": 0.0001337788504319599,
"loss": 3.0747454833984373,
"step": 27200
},
{
"epoch": 1.0504606607301183,
"grad_norm": 0.6088934540748596,
"learning_rate": 0.0001333008985291223,
"loss": 3.07860107421875,
"step": 27250
},
{
"epoch": 1.0523881114837517,
"grad_norm": 0.6098368167877197,
"learning_rate": 0.00013282311823538948,
"loss": 3.080094909667969,
"step": 27300
},
{
"epoch": 1.0543155622373848,
"grad_norm": 0.5968213081359863,
"learning_rate": 0.0001323455144606932,
"loss": 3.0572119140625,
"step": 27350
},
{
"epoch": 1.056243012991018,
"grad_norm": 0.5875847339630127,
"learning_rate": 0.00013186809211315116,
"loss": 3.0876007080078125,
"step": 27400
},
{
"epoch": 1.0581704637446514,
"grad_norm": 0.6149488091468811,
"learning_rate": 0.0001313908560990167,
"loss": 3.078609619140625,
"step": 27450
},
{
"epoch": 1.0600979144982845,
"grad_norm": 0.5971936583518982,
"learning_rate": 0.00013091381132262825,
"loss": 3.082308654785156,
"step": 27500
},
{
"epoch": 1.0620253652519178,
"grad_norm": 0.6102308034896851,
"learning_rate": 0.00013043696268635896,
"loss": 3.063638916015625,
"step": 27550
},
{
"epoch": 1.0639528160055511,
"grad_norm": 0.6145762205123901,
"learning_rate": 0.00012996031509056641,
"loss": 3.058980712890625,
"step": 27600
},
{
"epoch": 1.0658802667591842,
"grad_norm": 0.6291123032569885,
"learning_rate": 0.00012948387343354205,
"loss": 3.07666015625,
"step": 27650
},
{
"epoch": 1.0678077175128176,
"grad_norm": 0.6137670874595642,
"learning_rate": 0.00012900764261146108,
"loss": 3.073582763671875,
"step": 27700
},
{
"epoch": 1.0697351682664509,
"grad_norm": 0.622054934501648,
"learning_rate": 0.00012853162751833206,
"loss": 3.0610140991210937,
"step": 27750
},
{
"epoch": 1.071662619020084,
"grad_norm": 0.6032930016517639,
"learning_rate": 0.00012805583304594656,
"loss": 3.0665289306640626,
"step": 27800
},
{
"epoch": 1.0735900697737173,
"grad_norm": 0.6125069856643677,
"learning_rate": 0.00012758026408382894,
"loss": 3.067874755859375,
"step": 27850
},
{
"epoch": 1.0755175205273506,
"grad_norm": 0.6138522624969482,
"learning_rate": 0.0001271049255191861,
"loss": 3.07188720703125,
"step": 27900
},
{
"epoch": 1.0774449712809837,
"grad_norm": 0.6235440969467163,
"learning_rate": 0.00012662982223685725,
"loss": 3.089921875,
"step": 27950
},
{
"epoch": 1.079372422034617,
"grad_norm": 0.5986319184303284,
"learning_rate": 0.00012615495911926368,
"loss": 3.070625,
"step": 28000
},
{
"epoch": 1.079372422034617,
"eval_loss": 3.035137176513672,
"eval_runtime": 69.8207,
"eval_samples_per_second": 119.478,
"eval_steps_per_second": 3.738,
"step": 28000
},
{
"epoch": 1.0812998727882503,
"grad_norm": 0.6178694367408752,
"learning_rate": 0.00012568034104635872,
"loss": 3.070972900390625,
"step": 28050
},
{
"epoch": 1.0832273235418834,
"grad_norm": 0.6282202005386353,
"learning_rate": 0.00012520597289557728,
"loss": 3.0785171508789064,
"step": 28100
},
{
"epoch": 1.0851547742955168,
"grad_norm": 0.6131503582000732,
"learning_rate": 0.00012473185954178613,
"loss": 3.084732666015625,
"step": 28150
},
{
"epoch": 1.08708222504915,
"grad_norm": 0.5968056321144104,
"learning_rate": 0.0001242580058572335,
"loss": 3.0688177490234376,
"step": 28200
},
{
"epoch": 1.0890096758027832,
"grad_norm": 0.5882805585861206,
"learning_rate": 0.00012378441671149923,
"loss": 3.06754638671875,
"step": 28250
},
{
"epoch": 1.0909371265564165,
"grad_norm": 0.6154142618179321,
"learning_rate": 0.0001233110969714445,
"loss": 3.06905029296875,
"step": 28300
},
{
"epoch": 1.0928645773100498,
"grad_norm": 0.6151012182235718,
"learning_rate": 0.0001228380515011619,
"loss": 3.0826080322265623,
"step": 28350
},
{
"epoch": 1.094792028063683,
"grad_norm": 0.5990608334541321,
"learning_rate": 0.0001223652851619256,
"loss": 3.0492529296875,
"step": 28400
},
{
"epoch": 1.0967194788173162,
"grad_norm": 0.6095677018165588,
"learning_rate": 0.00012189280281214126,
"loss": 3.05683837890625,
"step": 28450
},
{
"epoch": 1.0986469295709496,
"grad_norm": 0.6118326783180237,
"learning_rate": 0.00012142060930729608,
"loss": 3.0574200439453123,
"step": 28500
},
{
"epoch": 1.1005743803245827,
"grad_norm": 0.5937634110450745,
"learning_rate": 0.00012094870949990876,
"loss": 3.0692398071289064,
"step": 28550
},
{
"epoch": 1.102501831078216,
"grad_norm": 0.6214609742164612,
"learning_rate": 0.00012047710823948003,
"loss": 3.0681939697265626,
"step": 28600
},
{
"epoch": 1.1044292818318493,
"grad_norm": 0.6386919021606445,
"learning_rate": 0.00012000581037244252,
"loss": 3.0709185791015625,
"step": 28650
},
{
"epoch": 1.1063567325854824,
"grad_norm": 0.6218888759613037,
"learning_rate": 0.000119534820742111,
"loss": 3.06106201171875,
"step": 28700
},
{
"epoch": 1.1082841833391157,
"grad_norm": 0.6106704473495483,
"learning_rate": 0.00011906414418863266,
"loss": 3.0639297485351564,
"step": 28750
},
{
"epoch": 1.110211634092749,
"grad_norm": 0.6272237300872803,
"learning_rate": 0.00011859378554893731,
"loss": 3.062796630859375,
"step": 28800
},
{
"epoch": 1.1121390848463821,
"grad_norm": 0.6092466711997986,
"learning_rate": 0.00011812374965668775,
"loss": 3.0817913818359375,
"step": 28850
},
{
"epoch": 1.1140665356000155,
"grad_norm": 0.6068085432052612,
"learning_rate": 0.00011765404134222998,
"loss": 3.0612603759765626,
"step": 28900
},
{
"epoch": 1.1159939863536485,
"grad_norm": 0.6144910454750061,
"learning_rate": 0.00011718466543254374,
"loss": 3.0641851806640625,
"step": 28950
},
{
"epoch": 1.1179214371072819,
"grad_norm": 0.6144693493843079,
"learning_rate": 0.00011671562675119265,
"loss": 3.0634112548828125,
"step": 29000
},
{
"epoch": 1.1198488878609152,
"grad_norm": 0.60772705078125,
"learning_rate": 0.0001162469301182749,
"loss": 3.06512451171875,
"step": 29050
},
{
"epoch": 1.1217763386145485,
"grad_norm": 0.6088334918022156,
"learning_rate": 0.00011578794391531792,
"loss": 3.0491845703125,
"step": 29100
},
{
"epoch": 1.1237037893681816,
"grad_norm": 0.6137677431106567,
"learning_rate": 0.00011531993874474651,
"loss": 3.065159912109375,
"step": 29150
},
{
"epoch": 1.125631240121815,
"grad_norm": 0.6335411071777344,
"learning_rate": 0.00011485228996546217,
"loss": 3.0766531372070314,
"step": 29200
},
{
"epoch": 1.127558690875448,
"grad_norm": 0.6167498230934143,
"learning_rate": 0.00011438500238327962,
"loss": 3.0707281494140624,
"step": 29250
},
{
"epoch": 1.1294861416290813,
"grad_norm": 0.6054856181144714,
"learning_rate": 0.00011391808080030178,
"loss": 3.055472412109375,
"step": 29300
},
{
"epoch": 1.1314135923827147,
"grad_norm": 0.6131047010421753,
"learning_rate": 0.00011345153001487035,
"loss": 3.0802825927734374,
"step": 29350
},
{
"epoch": 1.133341043136348,
"grad_norm": 0.6018222570419312,
"learning_rate": 0.00011298535482151649,
"loss": 3.0450469970703127,
"step": 29400
},
{
"epoch": 1.135268493889981,
"grad_norm": 0.6303772330284119,
"learning_rate": 0.00011251956001091166,
"loss": 3.0607635498046877,
"step": 29450
},
{
"epoch": 1.1371959446436144,
"grad_norm": 0.6043986678123474,
"learning_rate": 0.0001120541503698181,
"loss": 3.0429705810546874,
"step": 29500
},
{
"epoch": 1.1391233953972475,
"grad_norm": 0.6038321256637573,
"learning_rate": 0.00011158913068104002,
"loss": 3.042392578125,
"step": 29550
},
{
"epoch": 1.1410508461508808,
"grad_norm": 0.6067884564399719,
"learning_rate": 0.00011112450572337422,
"loss": 3.049588928222656,
"step": 29600
},
{
"epoch": 1.1429782969045141,
"grad_norm": 0.6415013074874878,
"learning_rate": 0.00011066028027156099,
"loss": 3.0401870727539064,
"step": 29650
},
{
"epoch": 1.1449057476581475,
"grad_norm": 0.612087607383728,
"learning_rate": 0.00011019645909623515,
"loss": 3.064275817871094,
"step": 29700
},
{
"epoch": 1.1468331984117806,
"grad_norm": 0.6156501173973083,
"learning_rate": 0.00010973304696387687,
"loss": 3.0658770751953126,
"step": 29750
},
{
"epoch": 1.1487606491654139,
"grad_norm": 0.6180119514465332,
"learning_rate": 0.00010927004863676279,
"loss": 3.058834228515625,
"step": 29800
},
{
"epoch": 1.150688099919047,
"grad_norm": 0.624697208404541,
"learning_rate": 0.00010880746887291714,
"loss": 3.0575326538085936,
"step": 29850
},
{
"epoch": 1.1526155506726803,
"grad_norm": 0.606814980506897,
"learning_rate": 0.00010834531242606277,
"loss": 3.038275146484375,
"step": 29900
},
{
"epoch": 1.1545430014263136,
"grad_norm": 0.6243532299995422,
"learning_rate": 0.00010788358404557206,
"loss": 3.05741455078125,
"step": 29950
},
{
"epoch": 1.1564704521799467,
"grad_norm": 0.6085441708564758,
"learning_rate": 0.00010742228847641858,
"loss": 3.059033203125,
"step": 30000
},
{
"epoch": 1.1564704521799467,
"eval_loss": 3.0178170204162598,
"eval_runtime": 70.2212,
"eval_samples_per_second": 118.796,
"eval_steps_per_second": 3.717,
"step": 30000
},
{
"epoch": 1.15839790293358,
"grad_norm": 0.6142354607582092,
"learning_rate": 0.00010696143045912805,
"loss": 3.042823486328125,
"step": 30050
},
{
"epoch": 1.1603253536872133,
"grad_norm": 0.6206157207489014,
"learning_rate": 0.00010650101472972966,
"loss": 3.049760437011719,
"step": 30100
},
{
"epoch": 1.1622528044408464,
"grad_norm": 0.6237283945083618,
"learning_rate": 0.00010604104601970736,
"loss": 3.032750244140625,
"step": 30150
},
{
"epoch": 1.1641802551944798,
"grad_norm": 0.623376190662384,
"learning_rate": 0.00010558152905595132,
"loss": 3.0694830322265627,
"step": 30200
},
{
"epoch": 1.166107705948113,
"grad_norm": 0.6116712093353271,
"learning_rate": 0.00010512246856070932,
"loss": 3.0331646728515627,
"step": 30250
},
{
"epoch": 1.1680351567017462,
"grad_norm": 0.6372489333152771,
"learning_rate": 0.00010466386925153822,
"loss": 3.066712646484375,
"step": 30300
},
{
"epoch": 1.1699626074553795,
"grad_norm": 0.619139552116394,
"learning_rate": 0.00010420573584125552,
"loss": 3.0659710693359377,
"step": 30350
},
{
"epoch": 1.1718900582090128,
"grad_norm": 0.6283072233200073,
"learning_rate": 0.00010374807303789075,
"loss": 3.0614788818359373,
"step": 30400
},
{
"epoch": 1.173817508962646,
"grad_norm": 0.6019864082336426,
"learning_rate": 0.0001032908855446374,
"loss": 3.0419598388671876,
"step": 30450
},
{
"epoch": 1.1757449597162792,
"grad_norm": 0.6169721484184265,
"learning_rate": 0.00010283417805980427,
"loss": 3.048824462890625,
"step": 30500
},
{
"epoch": 1.1776724104699126,
"grad_norm": 0.6106935143470764,
"learning_rate": 0.00010237795527676744,
"loss": 3.0280535888671873,
"step": 30550
},
{
"epoch": 1.1795998612235457,
"grad_norm": 0.6264392733573914,
"learning_rate": 0.00010192222188392186,
"loss": 3.0526617431640624,
"step": 30600
},
{
"epoch": 1.181527311977179,
"grad_norm": 0.6348525285720825,
"learning_rate": 0.00010146698256463327,
"loss": 3.0232196044921875,
"step": 30650
},
{
"epoch": 1.1834547627308123,
"grad_norm": 0.6170756816864014,
"learning_rate": 0.00010101224199719003,
"loss": 3.0458843994140623,
"step": 30700
},
{
"epoch": 1.1853822134844454,
"grad_norm": 0.6128750443458557,
"learning_rate": 0.00010055800485475502,
"loss": 3.049295654296875,
"step": 30750
},
{
"epoch": 1.1873096642380787,
"grad_norm": 0.6313346028327942,
"learning_rate": 0.0001001042758053178,
"loss": 3.0549395751953123,
"step": 30800
},
{
"epoch": 1.189237114991712,
"grad_norm": 0.6283858418464661,
"learning_rate": 9.965105951164622e-05,
"loss": 3.0375518798828125,
"step": 30850
},
{
"epoch": 1.1911645657453451,
"grad_norm": 0.6274822354316711,
"learning_rate": 9.919836063123898e-05,
"loss": 3.0390936279296876,
"step": 30900
},
{
"epoch": 1.1930920164989784,
"grad_norm": 0.6340188384056091,
"learning_rate": 9.87461838162775e-05,
"loss": 3.035181884765625,
"step": 30950
},
{
"epoch": 1.1950194672526118,
"grad_norm": 0.6270986795425415,
"learning_rate": 9.829453371357818e-05,
"loss": 3.045999755859375,
"step": 31000
},
{
"epoch": 1.1969469180062449,
"grad_norm": 0.6205368638038635,
"learning_rate": 9.784341496454464e-05,
"loss": 3.0430706787109374,
"step": 31050
},
{
"epoch": 1.1988743687598782,
"grad_norm": 0.6337893605232239,
"learning_rate": 9.739283220511998e-05,
"loss": 3.051441955566406,
"step": 31100
},
{
"epoch": 1.2008018195135115,
"grad_norm": 0.6329116821289062,
"learning_rate": 9.694279006573921e-05,
"loss": 3.0241876220703126,
"step": 31150
},
{
"epoch": 1.2027292702671446,
"grad_norm": 0.6341269016265869,
"learning_rate": 9.649329317128166e-05,
"loss": 3.0431637573242187,
"step": 31200
},
{
"epoch": 1.204656721020778,
"grad_norm": 0.6145036220550537,
"learning_rate": 9.604434614102337e-05,
"loss": 3.05436767578125,
"step": 31250
},
{
"epoch": 1.2065841717744112,
"grad_norm": 0.6161719560623169,
"learning_rate": 9.559595358858968e-05,
"loss": 3.0441265869140626,
"step": 31300
},
{
"epoch": 1.2085116225280443,
"grad_norm": 0.612169623374939,
"learning_rate": 9.51481201219078e-05,
"loss": 3.0354547119140625,
"step": 31350
},
{
"epoch": 1.2104390732816777,
"grad_norm": 0.6361745595932007,
"learning_rate": 9.470085034315952e-05,
"loss": 3.0442941284179685,
"step": 31400
},
{
"epoch": 1.212366524035311,
"grad_norm": 0.6232249140739441,
"learning_rate": 9.425414884873387e-05,
"loss": 3.03318603515625,
"step": 31450
},
{
"epoch": 1.214293974788944,
"grad_norm": 0.6282437443733215,
"learning_rate": 9.38080202291798e-05,
"loss": 3.03639404296875,
"step": 31500
},
{
"epoch": 1.2162214255425774,
"grad_norm": 0.617189347743988,
"learning_rate": 9.33624690691592e-05,
"loss": 3.043210144042969,
"step": 31550
},
{
"epoch": 1.2181488762962107,
"grad_norm": 0.6475757360458374,
"learning_rate": 9.291749994739954e-05,
"loss": 3.0463693237304685,
"step": 31600
},
{
"epoch": 1.2200763270498438,
"grad_norm": 0.6404930949211121,
"learning_rate": 9.247311743664706e-05,
"loss": 3.058622131347656,
"step": 31650
},
{
"epoch": 1.2220037778034771,
"grad_norm": 0.6376821994781494,
"learning_rate": 9.202932610361962e-05,
"loss": 3.0478814697265624,
"step": 31700
},
{
"epoch": 1.2239312285571105,
"grad_norm": 0.624325692653656,
"learning_rate": 9.158613050895976e-05,
"loss": 3.048443603515625,
"step": 31750
},
{
"epoch": 1.2258586793107435,
"grad_norm": 0.623665988445282,
"learning_rate": 9.114353520718795e-05,
"loss": 3.0423553466796873,
"step": 31800
},
{
"epoch": 1.2277861300643769,
"grad_norm": 0.6296160221099854,
"learning_rate": 9.070154474665571e-05,
"loss": 3.0418792724609376,
"step": 31850
},
{
"epoch": 1.2297135808180102,
"grad_norm": 0.6253572106361389,
"learning_rate": 9.02601636694989e-05,
"loss": 3.045480041503906,
"step": 31900
},
{
"epoch": 1.2316410315716433,
"grad_norm": 0.626477062702179,
"learning_rate": 8.981939651159098e-05,
"loss": 3.0366168212890625,
"step": 31950
},
{
"epoch": 1.2335684823252766,
"grad_norm": 0.6054511070251465,
"learning_rate": 8.937924780249646e-05,
"loss": 3.02014404296875,
"step": 32000
},
{
"epoch": 1.2335684823252766,
"eval_loss": 3.000340223312378,
"eval_runtime": 69.7252,
"eval_samples_per_second": 119.641,
"eval_steps_per_second": 3.743,
"step": 32000
},
{
"epoch": 1.23549593307891,
"grad_norm": 0.6225705742835999,
"learning_rate": 8.893972206542434e-05,
"loss": 3.0374151611328126,
"step": 32050
},
{
"epoch": 1.237423383832543,
"grad_norm": 0.6371407508850098,
"learning_rate": 8.850082381718161e-05,
"loss": 3.0510772705078124,
"step": 32100
},
{
"epoch": 1.2393508345861763,
"grad_norm": 0.6352553963661194,
"learning_rate": 8.806255756812687e-05,
"loss": 3.0414532470703124,
"step": 32150
},
{
"epoch": 1.2412782853398097,
"grad_norm": 0.6342623829841614,
"learning_rate": 8.762492782212387e-05,
"loss": 3.038360900878906,
"step": 32200
},
{
"epoch": 1.2432057360934428,
"grad_norm": 0.6284528970718384,
"learning_rate": 8.718793907649539e-05,
"loss": 3.0395050048828125,
"step": 32250
},
{
"epoch": 1.245133186847076,
"grad_norm": 0.6224302649497986,
"learning_rate": 8.675159582197693e-05,
"loss": 3.0444467163085935,
"step": 32300
},
{
"epoch": 1.2470606376007094,
"grad_norm": 0.6323182582855225,
"learning_rate": 8.631590254267045e-05,
"loss": 3.027712097167969,
"step": 32350
},
{
"epoch": 1.2489880883543425,
"grad_norm": 0.6297076344490051,
"learning_rate": 8.588086371599865e-05,
"loss": 3.027608642578125,
"step": 32400
},
{
"epoch": 1.2509155391079758,
"grad_norm": 0.6271635890007019,
"learning_rate": 8.544648381265843e-05,
"loss": 3.034465637207031,
"step": 32450
},
{
"epoch": 1.252842989861609,
"grad_norm": 0.6575161218643188,
"learning_rate": 8.501276729657551e-05,
"loss": 3.0370046997070315,
"step": 32500
},
{
"epoch": 1.2547704406152422,
"grad_norm": 0.6541010141372681,
"learning_rate": 8.457971862485806e-05,
"loss": 3.032591552734375,
"step": 32550
},
{
"epoch": 1.2566978913688756,
"grad_norm": 0.6241432428359985,
"learning_rate": 8.414734224775125e-05,
"loss": 3.0212411499023437,
"step": 32600
},
{
"epoch": 1.2586253421225089,
"grad_norm": 0.6301848888397217,
"learning_rate": 8.371564260859133e-05,
"loss": 3.048199462890625,
"step": 32650
},
{
"epoch": 1.260552792876142,
"grad_norm": 0.6241071224212646,
"learning_rate": 8.328462414375996e-05,
"loss": 3.037410888671875,
"step": 32700
},
{
"epoch": 1.2624802436297753,
"grad_norm": 0.6445004940032959,
"learning_rate": 8.285429128263887e-05,
"loss": 3.02450927734375,
"step": 32750
},
{
"epoch": 1.2644076943834084,
"grad_norm": 0.6178706884384155,
"learning_rate": 8.242464844756388e-05,
"loss": 3.0374530029296873,
"step": 32800
},
{
"epoch": 1.2663351451370417,
"grad_norm": 0.6305875778198242,
"learning_rate": 8.199570005378e-05,
"loss": 3.0251800537109377,
"step": 32850
},
{
"epoch": 1.268262595890675,
"grad_norm": 0.6229849457740784,
"learning_rate": 8.156745050939554e-05,
"loss": 3.037336730957031,
"step": 32900
},
{
"epoch": 1.2701900466443083,
"grad_norm": 0.6180618405342102,
"learning_rate": 8.11399042153373e-05,
"loss": 3.03130615234375,
"step": 32950
},
{
"epoch": 1.2721174973979414,
"grad_norm": 0.6285006403923035,
"learning_rate": 8.071306556530482e-05,
"loss": 3.0460336303710935,
"step": 33000
},
{
"epoch": 1.2740449481515748,
"grad_norm": 0.6295358538627625,
"learning_rate": 8.028693894572583e-05,
"loss": 3.0395074462890626,
"step": 33050
},
{
"epoch": 1.2759723989052079,
"grad_norm": 0.6418010592460632,
"learning_rate": 7.987002989080831e-05,
"loss": 3.03727783203125,
"step": 33100
},
{
"epoch": 1.2778998496588412,
"grad_norm": 0.6400101184844971,
"learning_rate": 7.944532600368323e-05,
"loss": 3.033727111816406,
"step": 33150
},
{
"epoch": 1.2798273004124745,
"grad_norm": 0.6185802221298218,
"learning_rate": 7.902134717499691e-05,
"loss": 3.025660095214844,
"step": 33200
},
{
"epoch": 1.2817547511661078,
"grad_norm": 0.6498960852622986,
"learning_rate": 7.859809776178795e-05,
"loss": 3.0384698486328126,
"step": 33250
},
{
"epoch": 1.283682201919741,
"grad_norm": 0.614106297492981,
"learning_rate": 7.817558211359877e-05,
"loss": 3.031085205078125,
"step": 33300
},
{
"epoch": 1.2856096526733742,
"grad_norm": 0.6262015700340271,
"learning_rate": 7.775380457243153e-05,
"loss": 3.0491497802734373,
"step": 33350
},
{
"epoch": 1.2875371034270073,
"grad_norm": 0.6320862770080566,
"learning_rate": 7.733276947270293e-05,
"loss": 3.0254330444335937,
"step": 33400
},
{
"epoch": 1.2894645541806407,
"grad_norm": 0.6323364973068237,
"learning_rate": 7.691248114120011e-05,
"loss": 3.01587890625,
"step": 33450
},
{
"epoch": 1.291392004934274,
"grad_norm": 0.6382852792739868,
"learning_rate": 7.649294389703601e-05,
"loss": 3.031526794433594,
"step": 33500
},
{
"epoch": 1.2933194556879073,
"grad_norm": 0.6332157850265503,
"learning_rate": 7.607416205160487e-05,
"loss": 3.024804992675781,
"step": 33550
},
{
"epoch": 1.2952469064415404,
"grad_norm": 0.6224969029426575,
"learning_rate": 7.565613990853808e-05,
"loss": 3.03927490234375,
"step": 33600
},
{
"epoch": 1.2971743571951737,
"grad_norm": 0.649642288684845,
"learning_rate": 7.52388817636599e-05,
"loss": 3.0306216430664064,
"step": 33650
},
{
"epoch": 1.2991018079488068,
"grad_norm": 0.6452252268791199,
"learning_rate": 7.482239190494347e-05,
"loss": 3.029102783203125,
"step": 33700
},
{
"epoch": 1.3010292587024401,
"grad_norm": 0.636181652545929,
"learning_rate": 7.440667461246637e-05,
"loss": 3.034852294921875,
"step": 33750
},
{
"epoch": 1.3029567094560734,
"grad_norm": 0.6497048735618591,
"learning_rate": 7.399173415836709e-05,
"loss": 3.0421319580078126,
"step": 33800
},
{
"epoch": 1.3048841602097068,
"grad_norm": 0.6105928421020508,
"learning_rate": 7.35775748068007e-05,
"loss": 3.034879455566406,
"step": 33850
},
{
"epoch": 1.3068116109633399,
"grad_norm": 0.6290647387504578,
"learning_rate": 7.316420081389545e-05,
"loss": 3.0401675415039064,
"step": 33900
},
{
"epoch": 1.3087390617169732,
"grad_norm": 0.6315209865570068,
"learning_rate": 7.275161642770858e-05,
"loss": 3.0372125244140626,
"step": 33950
},
{
"epoch": 1.3106665124706063,
"grad_norm": 0.6306831240653992,
"learning_rate": 7.233982588818312e-05,
"loss": 3.0102716064453126,
"step": 34000
},
{
"epoch": 1.3106665124706063,
"eval_loss": 2.9862301349639893,
"eval_runtime": 70.31,
"eval_samples_per_second": 118.646,
"eval_steps_per_second": 3.712,
"step": 34000
},
{
"epoch": 1.3125939632242396,
"grad_norm": 0.6495873928070068,
"learning_rate": 7.192883342710393e-05,
"loss": 3.02664306640625,
"step": 34050
},
{
"epoch": 1.314521413977873,
"grad_norm": 0.6229225397109985,
"learning_rate": 7.151864326805443e-05,
"loss": 3.0202597045898436,
"step": 34100
},
{
"epoch": 1.3164488647315062,
"grad_norm": 0.6285218000411987,
"learning_rate": 7.11092596263732e-05,
"loss": 3.0182623291015624,
"step": 34150
},
{
"epoch": 1.3183763154851393,
"grad_norm": 0.6604203581809998,
"learning_rate": 7.070068670911051e-05,
"loss": 3.0358001708984377,
"step": 34200
},
{
"epoch": 1.3203037662387727,
"grad_norm": 0.6439961194992065,
"learning_rate": 7.029292871498534e-05,
"loss": 3.0391226196289063,
"step": 34250
},
{
"epoch": 1.3222312169924058,
"grad_norm": 0.6501875519752502,
"learning_rate": 6.988598983434184e-05,
"loss": 3.0246237182617186,
"step": 34300
},
{
"epoch": 1.324158667746039,
"grad_norm": 0.6348546743392944,
"learning_rate": 6.947987424910676e-05,
"loss": 3.021990966796875,
"step": 34350
},
{
"epoch": 1.3260861184996724,
"grad_norm": 0.6358281373977661,
"learning_rate": 6.907458613274596e-05,
"loss": 3.011370849609375,
"step": 34400
},
{
"epoch": 1.3280135692533057,
"grad_norm": 0.6379692554473877,
"learning_rate": 6.867012965022208e-05,
"loss": 3.014393310546875,
"step": 34450
},
{
"epoch": 1.3299410200069388,
"grad_norm": 0.6277499794960022,
"learning_rate": 6.826650895795105e-05,
"loss": 3.02208740234375,
"step": 34500
},
{
"epoch": 1.3318684707605721,
"grad_norm": 0.6472629904747009,
"learning_rate": 6.78637282037601e-05,
"loss": 3.016767578125,
"step": 34550
},
{
"epoch": 1.3337959215142052,
"grad_norm": 0.6619810461997986,
"learning_rate": 6.746179152684463e-05,
"loss": 3.0127658081054687,
"step": 34600
},
{
"epoch": 1.3357233722678385,
"grad_norm": 0.6245758533477783,
"learning_rate": 6.706070305772587e-05,
"loss": 3.0284637451171874,
"step": 34650
},
{
"epoch": 1.3376508230214719,
"grad_norm": 0.6226040720939636,
"learning_rate": 6.66604669182085e-05,
"loss": 3.021361083984375,
"step": 34700
},
{
"epoch": 1.339578273775105,
"grad_norm": 0.6320768594741821,
"learning_rate": 6.626108722133796e-05,
"loss": 3.0464654541015626,
"step": 34750
},
{
"epoch": 1.3415057245287383,
"grad_norm": 0.6307546496391296,
"learning_rate": 6.586256807135871e-05,
"loss": 3.005638427734375,
"step": 34800
},
{
"epoch": 1.3434331752823716,
"grad_norm": 0.6351662278175354,
"learning_rate": 6.54649135636715e-05,
"loss": 3.0298114013671875,
"step": 34850
},
{
"epoch": 1.3453606260360047,
"grad_norm": 0.6542770266532898,
"learning_rate": 6.506812778479176e-05,
"loss": 3.0092144775390626,
"step": 34900
},
{
"epoch": 1.347288076789638,
"grad_norm": 0.6357635855674744,
"learning_rate": 6.467221481230722e-05,
"loss": 3.0104840087890623,
"step": 34950
},
{
"epoch": 1.3492155275432713,
"grad_norm": 0.6285920739173889,
"learning_rate": 6.427717871483623e-05,
"loss": 3.020973815917969,
"step": 35000
},
{
"epoch": 1.3511429782969044,
"grad_norm": 0.6299211978912354,
"learning_rate": 6.388302355198585e-05,
"loss": 3.0234722900390625,
"step": 35050
},
{
"epoch": 1.3530704290505378,
"grad_norm": 0.6482565999031067,
"learning_rate": 6.348975337431034e-05,
"loss": 3.0237777709960936,
"step": 35100
},
{
"epoch": 1.354997879804171,
"grad_norm": 0.6380200982093811,
"learning_rate": 6.310521110773279e-05,
"loss": 3.006343994140625,
"step": 35150
},
{
"epoch": 1.3569253305578042,
"grad_norm": 0.62126225233078,
"learning_rate": 6.271370511501279e-05,
"loss": 3.016763916015625,
"step": 35200
},
{
"epoch": 1.3588527813114375,
"grad_norm": 0.6541768312454224,
"learning_rate": 6.232309612402369e-05,
"loss": 3.0403656005859374,
"step": 35250
},
{
"epoch": 1.3607802320650708,
"grad_norm": 0.6393908858299255,
"learning_rate": 6.193338814887732e-05,
"loss": 2.999876708984375,
"step": 35300
},
{
"epoch": 1.362707682818704,
"grad_norm": 0.620763897895813,
"learning_rate": 6.154458519442599e-05,
"loss": 3.0015032958984373,
"step": 35350
},
{
"epoch": 1.3646351335723372,
"grad_norm": 0.6346343159675598,
"learning_rate": 6.115669125622172e-05,
"loss": 3.0304629516601564,
"step": 35400
},
{
"epoch": 1.3665625843259706,
"grad_norm": 0.6651772856712341,
"learning_rate": 6.0769710320474844e-05,
"loss": 3.01797607421875,
"step": 35450
},
{
"epoch": 1.3684900350796037,
"grad_norm": 0.6441340446472168,
"learning_rate": 6.0383646364013136e-05,
"loss": 3.00494873046875,
"step": 35500
},
{
"epoch": 1.370417485833237,
"grad_norm": 0.6288766860961914,
"learning_rate": 5.999850335424122e-05,
"loss": 3.0138446044921876,
"step": 35550
},
{
"epoch": 1.3723449365868703,
"grad_norm": 0.6625560522079468,
"learning_rate": 5.961428524909929e-05,
"loss": 3.015384521484375,
"step": 35600
},
{
"epoch": 1.3742723873405034,
"grad_norm": 0.6305920481681824,
"learning_rate": 5.9230995997022955e-05,
"loss": 3.0071783447265625,
"step": 35650
},
{
"epoch": 1.3761998380941367,
"grad_norm": 0.633285641670227,
"learning_rate": 5.884863953690221e-05,
"loss": 3.0072540283203124,
"step": 35700
},
{
"epoch": 1.3781272888477698,
"grad_norm": 0.6546490788459778,
"learning_rate": 5.846721979804138e-05,
"loss": 3.006717529296875,
"step": 35750
},
{
"epoch": 1.3800547396014031,
"grad_norm": 0.6545742154121399,
"learning_rate": 5.8086740700118325e-05,
"loss": 3.020059814453125,
"step": 35800
},
{
"epoch": 1.3819821903550364,
"grad_norm": 0.6293085813522339,
"learning_rate": 5.770720615314454e-05,
"loss": 2.995181884765625,
"step": 35850
},
{
"epoch": 1.3839096411086698,
"grad_norm": 0.6314324736595154,
"learning_rate": 5.732862005742469e-05,
"loss": 2.993491516113281,
"step": 35900
},
{
"epoch": 1.3858370918623029,
"grad_norm": 0.6504847407341003,
"learning_rate": 5.6950986303516644e-05,
"loss": 3.017044982910156,
"step": 35950
},
{
"epoch": 1.3877645426159362,
"grad_norm": 0.6358339786529541,
"learning_rate": 5.657430877219147e-05,
"loss": 3.028511657714844,
"step": 36000
},
{
"epoch": 1.3877645426159362,
"eval_loss": 2.9710333347320557,
"eval_runtime": 70.1501,
"eval_samples_per_second": 118.916,
"eval_steps_per_second": 3.721,
"step": 36000
},
{
"epoch": 1.3896919933695693,
"grad_norm": 0.6393367648124695,
"learning_rate": 5.619859133439364e-05,
"loss": 3.012779541015625,
"step": 36050
},
{
"epoch": 1.3916194441232026,
"grad_norm": 0.6493411064147949,
"learning_rate": 5.582383785120122e-05,
"loss": 3.000853271484375,
"step": 36100
},
{
"epoch": 1.393546894876836,
"grad_norm": 0.6275473237037659,
"learning_rate": 5.545005217378594e-05,
"loss": 3.0173858642578124,
"step": 36150
},
{
"epoch": 1.3954743456304692,
"grad_norm": 0.6341407895088196,
"learning_rate": 5.507723814337411e-05,
"loss": 3.012436828613281,
"step": 36200
},
{
"epoch": 1.3974017963841023,
"grad_norm": 0.648978590965271,
"learning_rate": 5.470539959120657e-05,
"loss": 3.022398681640625,
"step": 36250
},
{
"epoch": 1.3993292471377357,
"grad_norm": 0.6560739278793335,
"learning_rate": 5.4334540338499894e-05,
"loss": 3.0263824462890625,
"step": 36300
},
{
"epoch": 1.4012566978913688,
"grad_norm": 0.638169527053833,
"learning_rate": 5.396466419640663e-05,
"loss": 3.0015203857421877,
"step": 36350
},
{
"epoch": 1.403184148645002,
"grad_norm": 0.6476456522941589,
"learning_rate": 5.359577496597643e-05,
"loss": 2.9965289306640623,
"step": 36400
},
{
"epoch": 1.4051115993986354,
"grad_norm": 0.6480238437652588,
"learning_rate": 5.322787643811684e-05,
"loss": 2.9867437744140624,
"step": 36450
},
{
"epoch": 1.4070390501522687,
"grad_norm": 0.6348977088928223,
"learning_rate": 5.2860972393554544e-05,
"loss": 3.0067373657226564,
"step": 36500
},
{
"epoch": 1.4089665009059018,
"grad_norm": 0.6375760436058044,
"learning_rate": 5.2495066602796206e-05,
"loss": 3.0192449951171874,
"step": 36550
},
{
"epoch": 1.4108939516595351,
"grad_norm": 0.6387436985969543,
"learning_rate": 5.2130162826090024e-05,
"loss": 2.99778076171875,
"step": 36600
},
{
"epoch": 1.4128214024131682,
"grad_norm": 0.6538995504379272,
"learning_rate": 5.1766264813386814e-05,
"loss": 3.021131896972656,
"step": 36650
},
{
"epoch": 1.4147488531668015,
"grad_norm": 0.6433481574058533,
"learning_rate": 5.140337630430179e-05,
"loss": 3.0046710205078124,
"step": 36700
},
{
"epoch": 1.4166763039204349,
"grad_norm": 0.6791552901268005,
"learning_rate": 5.104150102807572e-05,
"loss": 3.012547912597656,
"step": 36750
},
{
"epoch": 1.4186037546740682,
"grad_norm": 0.6463810205459595,
"learning_rate": 5.06806427035371e-05,
"loss": 2.9944467163085937,
"step": 36800
},
{
"epoch": 1.4205312054277013,
"grad_norm": 0.637865424156189,
"learning_rate": 5.0320805039063494e-05,
"loss": 2.9921234130859373,
"step": 36850
},
{
"epoch": 1.4224586561813346,
"grad_norm": 0.6582602858543396,
"learning_rate": 4.9961991732543586e-05,
"loss": 3.0204217529296873,
"step": 36900
},
{
"epoch": 1.4243861069349677,
"grad_norm": 0.650175154209137,
"learning_rate": 4.960420647133935e-05,
"loss": 2.9906246948242186,
"step": 36950
},
{
"epoch": 1.426313557688601,
"grad_norm": 0.6676084995269775,
"learning_rate": 4.924745293224784e-05,
"loss": 3.010247802734375,
"step": 37000
},
{
"epoch": 1.4282410084422343,
"grad_norm": 0.646117627620697,
"learning_rate": 4.889173478146374e-05,
"loss": 3.0143060302734375,
"step": 37050
},
{
"epoch": 1.4301684591958677,
"grad_norm": 0.6552669405937195,
"learning_rate": 4.853705567454132e-05,
"loss": 3.0090032958984376,
"step": 37100
},
{
"epoch": 1.4320959099495008,
"grad_norm": 0.6410604119300842,
"learning_rate": 4.818341925635729e-05,
"loss": 2.989464111328125,
"step": 37150
},
{
"epoch": 1.434023360703134,
"grad_norm": 0.660908579826355,
"learning_rate": 4.783082916107286e-05,
"loss": 2.9977349853515625,
"step": 37200
},
{
"epoch": 1.4359508114567672,
"grad_norm": 0.6467908620834351,
"learning_rate": 4.7479289012096926e-05,
"loss": 2.9914288330078125,
"step": 37250
},
{
"epoch": 1.4378782622104005,
"grad_norm": 0.6461449265480042,
"learning_rate": 4.712880242204836e-05,
"loss": 3.003518371582031,
"step": 37300
},
{
"epoch": 1.4398057129640338,
"grad_norm": 0.6480746865272522,
"learning_rate": 4.678635119788662e-05,
"loss": 2.9931671142578127,
"step": 37350
},
{
"epoch": 1.4417331637176671,
"grad_norm": 0.6509934663772583,
"learning_rate": 4.6437961270051164e-05,
"loss": 3.00406005859375,
"step": 37400
},
{
"epoch": 1.4436606144713002,
"grad_norm": 0.6493057608604431,
"learning_rate": 4.609063560239681e-05,
"loss": 3.0092340087890626,
"step": 37450
},
{
"epoch": 1.4455880652249335,
"grad_norm": 0.6554242372512817,
"learning_rate": 4.574437776423222e-05,
"loss": 2.9999981689453126,
"step": 37500
},
{
"epoch": 1.4475155159785666,
"grad_norm": 0.6676389575004578,
"learning_rate": 4.539919131389231e-05,
"loss": 3.0145263671875,
"step": 37550
},
{
"epoch": 1.4494429667322,
"grad_norm": 0.6601196527481079,
"learning_rate": 4.5055079798702024e-05,
"loss": 3.0149322509765626,
"step": 37600
},
{
"epoch": 1.4513704174858333,
"grad_norm": 0.6483643054962158,
"learning_rate": 4.4712046754939535e-05,
"loss": 3.006389465332031,
"step": 37650
},
{
"epoch": 1.4532978682394666,
"grad_norm": 0.6562464833259583,
"learning_rate": 4.437009570780019e-05,
"loss": 3.009471435546875,
"step": 37700
},
{
"epoch": 1.4552253189930997,
"grad_norm": 0.668907105922699,
"learning_rate": 4.402923017136001e-05,
"loss": 3.0093399047851563,
"step": 37750
},
{
"epoch": 1.457152769746733,
"grad_norm": 0.6720010638237,
"learning_rate": 4.368945364853981e-05,
"loss": 3.0017431640625,
"step": 37800
},
{
"epoch": 1.4590802205003661,
"grad_norm": 0.6314327716827393,
"learning_rate": 4.3350769631068975e-05,
"loss": 3.0013021850585937,
"step": 37850
},
{
"epoch": 1.4610076712539994,
"grad_norm": 0.6467392444610596,
"learning_rate": 4.301318159944994e-05,
"loss": 3.0018572998046875,
"step": 37900
},
{
"epoch": 1.4629351220076328,
"grad_norm": 0.6472886204719543,
"learning_rate": 4.267669302292195e-05,
"loss": 2.9911578369140623,
"step": 37950
},
{
"epoch": 1.464862572761266,
"grad_norm": 0.6552727818489075,
"learning_rate": 4.234130735942583e-05,
"loss": 3.0146478271484374,
"step": 38000
},
{
"epoch": 1.464862572761266,
"eval_loss": 2.9608397483825684,
"eval_runtime": 69.9925,
"eval_samples_per_second": 119.184,
"eval_steps_per_second": 3.729,
"step": 38000
},
{
"epoch": 1.4667900235148992,
"grad_norm": 0.6628807783126831,
"learning_rate": 4.2007028055568095e-05,
"loss": 2.987430419921875,
"step": 38050
},
{
"epoch": 1.4687174742685325,
"grad_norm": 0.6167513132095337,
"learning_rate": 4.1673858546585915e-05,
"loss": 2.983268127441406,
"step": 38100
},
{
"epoch": 1.4706449250221656,
"grad_norm": 0.6317017078399658,
"learning_rate": 4.1341802256311346e-05,
"loss": 2.98488525390625,
"step": 38150
},
{
"epoch": 1.472572375775799,
"grad_norm": 0.6541942954063416,
"learning_rate": 4.101086259713666e-05,
"loss": 3.001531677246094,
"step": 38200
},
{
"epoch": 1.4744998265294322,
"grad_norm": 0.6744600534439087,
"learning_rate": 4.0681042969978845e-05,
"loss": 3.006551513671875,
"step": 38250
},
{
"epoch": 1.4764272772830656,
"grad_norm": 0.6821916699409485,
"learning_rate": 4.0352346764244796e-05,
"loss": 3.007157287597656,
"step": 38300
},
{
"epoch": 1.4783547280366987,
"grad_norm": 0.6707517504692078,
"learning_rate": 4.002477735779671e-05,
"loss": 2.987646179199219,
"step": 38350
},
{
"epoch": 1.480282178790332,
"grad_norm": 0.6475400924682617,
"learning_rate": 3.969833811691697e-05,
"loss": 2.986300354003906,
"step": 38400
},
{
"epoch": 1.482209629543965,
"grad_norm": 0.6588340401649475,
"learning_rate": 3.937303239627392e-05,
"loss": 3.013193359375,
"step": 38450
},
{
"epoch": 1.4841370802975984,
"grad_norm": 0.6350312829017639,
"learning_rate": 3.9048863538887095e-05,
"loss": 3.002423095703125,
"step": 38500
},
{
"epoch": 1.4860645310512317,
"grad_norm": 0.667890191078186,
"learning_rate": 3.872583487609314e-05,
"loss": 3.0143673706054686,
"step": 38550
},
{
"epoch": 1.4879919818048648,
"grad_norm": 0.6580042839050293,
"learning_rate": 3.840394972751125e-05,
"loss": 2.9950054931640624,
"step": 38600
},
{
"epoch": 1.4899194325584981,
"grad_norm": 0.648325502872467,
"learning_rate": 3.8083211401009485e-05,
"loss": 3.026676025390625,
"step": 38650
},
{
"epoch": 1.4918468833121314,
"grad_norm": 0.6553570628166199,
"learning_rate": 3.776362319267034e-05,
"loss": 2.999617614746094,
"step": 38700
},
{
"epoch": 1.4937743340657645,
"grad_norm": 0.6672315001487732,
"learning_rate": 3.7445188386757066e-05,
"loss": 3.0035385131835937,
"step": 38750
},
{
"epoch": 1.4957017848193979,
"grad_norm": 0.6690788269042969,
"learning_rate": 3.71279102556801e-05,
"loss": 3.002785949707031,
"step": 38800
},
{
"epoch": 1.4976292355730312,
"grad_norm": 0.6553841829299927,
"learning_rate": 3.681179205996303e-05,
"loss": 3.0008990478515627,
"step": 38850
},
{
"epoch": 1.4995566863266643,
"grad_norm": 0.6610223054885864,
"learning_rate": 3.649683704820951e-05,
"loss": 3.0096612548828126,
"step": 38900
},
{
"epoch": 1.5014841370802976,
"grad_norm": 0.6567665338516235,
"learning_rate": 3.61830484570695e-05,
"loss": 2.9983200073242187,
"step": 38950
},
{
"epoch": 1.5034115878339307,
"grad_norm": 0.6621602773666382,
"learning_rate": 3.5870429511206394e-05,
"loss": 2.9971014404296876,
"step": 39000
},
{
"epoch": 1.505339038587564,
"grad_norm": 0.6436423659324646,
"learning_rate": 3.5558983423263424e-05,
"loss": 2.9980072021484374,
"step": 39050
},
{
"epoch": 1.5072664893411973,
"grad_norm": 0.6546348333358765,
"learning_rate": 3.5248713393831156e-05,
"loss": 2.9966961669921877,
"step": 39100
},
{
"epoch": 1.5091939400948307,
"grad_norm": 0.6690251231193542,
"learning_rate": 3.493962261141403e-05,
"loss": 3.0205093383789063,
"step": 39150
},
{
"epoch": 1.511121390848464,
"grad_norm": 0.6529703140258789,
"learning_rate": 3.4631714252398194e-05,
"loss": 2.974471435546875,
"step": 39200
},
{
"epoch": 1.513048841602097,
"grad_norm": 0.6409212350845337,
"learning_rate": 3.432499148101838e-05,
"loss": 2.9807284545898436,
"step": 39250
},
{
"epoch": 1.5149762923557302,
"grad_norm": 0.6577807068824768,
"learning_rate": 3.401945744932566e-05,
"loss": 2.9837249755859374,
"step": 39300
},
{
"epoch": 1.5169037431093635,
"grad_norm": 0.6554986834526062,
"learning_rate": 3.3715115297154905e-05,
"loss": 2.9847296142578124,
"step": 39350
},
{
"epoch": 1.5188311938629968,
"grad_norm": 0.6742623448371887,
"learning_rate": 3.3411968152092684e-05,
"loss": 3.002864990234375,
"step": 39400
},
{
"epoch": 1.5207586446166301,
"grad_norm": 0.8173781633377075,
"learning_rate": 3.311604634820763e-05,
"loss": 2.990213623046875,
"step": 39450
},
{
"epoch": 1.5226860953702634,
"grad_norm": 0.6421291828155518,
"learning_rate": 3.281527449612996e-05,
"loss": 2.9789743041992187,
"step": 39500
},
{
"epoch": 1.5246135461238965,
"grad_norm": 0.6436623334884644,
"learning_rate": 3.251570689841721e-05,
"loss": 2.9856884765625,
"step": 39550
},
{
"epoch": 1.5265409968775296,
"grad_norm": 0.667201042175293,
"learning_rate": 3.221734663359004e-05,
"loss": 3.000289306640625,
"step": 39600
},
{
"epoch": 1.528468447631163,
"grad_norm": 0.6464906930923462,
"learning_rate": 3.192019676776173e-05,
"loss": 2.9970965576171875,
"step": 39650
},
{
"epoch": 1.5303958983847963,
"grad_norm": 0.6583192944526672,
"learning_rate": 3.162426035460691e-05,
"loss": 2.986309814453125,
"step": 39700
},
{
"epoch": 1.5323233491384296,
"grad_norm": 0.6383770108222961,
"learning_rate": 3.132954043533014e-05,
"loss": 2.9810650634765623,
"step": 39750
},
{
"epoch": 1.534250799892063,
"grad_norm": 0.6760772466659546,
"learning_rate": 3.103604003863454e-05,
"loss": 2.9895010375976563,
"step": 39800
},
{
"epoch": 1.536178250645696,
"grad_norm": 0.6493309736251831,
"learning_rate": 3.0743762180690855e-05,
"loss": 3.0052420043945314,
"step": 39850
},
{
"epoch": 1.5381057013993291,
"grad_norm": 0.6615433096885681,
"learning_rate": 3.0452709865106217e-05,
"loss": 2.995009765625,
"step": 39900
},
{
"epoch": 1.5400331521529624,
"grad_norm": 0.6336745023727417,
"learning_rate": 3.016288608289358e-05,
"loss": 3.0002447509765626,
"step": 39950
},
{
"epoch": 1.5419606029065958,
"grad_norm": 0.6437167525291443,
"learning_rate": 2.9874293812440624e-05,
"loss": 2.98669677734375,
"step": 40000
},
{
"epoch": 1.5419606029065958,
"eval_loss": 2.9493415355682373,
"eval_runtime": 69.758,
"eval_samples_per_second": 119.585,
"eval_steps_per_second": 3.742,
"step": 40000
},
{
"epoch": 1.543888053660229,
"grad_norm": 0.6454316973686218,
"learning_rate": 2.9586936019479567e-05,
"loss": 2.9892059326171876,
"step": 40050
},
{
"epoch": 1.5458155044138622,
"grad_norm": 0.6593596935272217,
"learning_rate": 2.9300815657056247e-05,
"loss": 2.975858154296875,
"step": 40100
},
{
"epoch": 1.5477429551674955,
"grad_norm": 0.6588073968887329,
"learning_rate": 2.901593566550008e-05,
"loss": 3.001131591796875,
"step": 40150
},
{
"epoch": 1.5496704059211286,
"grad_norm": 0.6525108814239502,
"learning_rate": 2.8732298972393804e-05,
"loss": 2.9824761962890625,
"step": 40200
},
{
"epoch": 1.551597856674762,
"grad_norm": 0.6474674940109253,
"learning_rate": 2.8449908492543217e-05,
"loss": 2.9728018188476564,
"step": 40250
},
{
"epoch": 1.5535253074283952,
"grad_norm": 0.6582866311073303,
"learning_rate": 2.8168767127947495e-05,
"loss": 2.9858349609375,
"step": 40300
},
{
"epoch": 1.5554527581820285,
"grad_norm": 0.6579475402832031,
"learning_rate": 2.7888877767769065e-05,
"loss": 3.0010595703125,
"step": 40350
},
{
"epoch": 1.5573802089356616,
"grad_norm": 0.6484814286231995,
"learning_rate": 2.761024328830426e-05,
"loss": 2.9938113403320314,
"step": 40400
},
{
"epoch": 1.559307659689295,
"grad_norm": 0.6481432318687439,
"learning_rate": 2.733286655295334e-05,
"loss": 3.0031854248046876,
"step": 40450
},
{
"epoch": 1.561235110442928,
"grad_norm": 0.6666094064712524,
"learning_rate": 2.7056750412191573e-05,
"loss": 3.0000149536132814,
"step": 40500
},
{
"epoch": 1.5631625611965614,
"grad_norm": 0.6500440239906311,
"learning_rate": 2.6781897703539356e-05,
"loss": 2.9727203369140627,
"step": 40550
},
{
"epoch": 1.5650900119501947,
"grad_norm": 0.6589999198913574,
"learning_rate": 2.6508311251533642e-05,
"loss": 2.9960476684570314,
"step": 40600
},
{
"epoch": 1.567017462703828,
"grad_norm": 0.646747887134552,
"learning_rate": 2.6235993867698462e-05,
"loss": 2.9949468994140624,
"step": 40650
},
{
"epoch": 1.5689449134574611,
"grad_norm": 0.6587992310523987,
"learning_rate": 2.5964948350516312e-05,
"loss": 2.97785888671875,
"step": 40700
},
{
"epoch": 1.5708723642110944,
"grad_norm": 0.6446172595024109,
"learning_rate": 2.5695177485399194e-05,
"loss": 2.9886749267578123,
"step": 40750
},
{
"epoch": 1.5727998149647275,
"grad_norm": 0.6597557663917542,
"learning_rate": 2.5426684044660217e-05,
"loss": 2.9851254272460936,
"step": 40800
},
{
"epoch": 1.5747272657183609,
"grad_norm": 0.6633270382881165,
"learning_rate": 2.5159470787484813e-05,
"loss": 2.976287841796875,
"step": 40850
},
{
"epoch": 1.5766547164719942,
"grad_norm": 0.6537542939186096,
"learning_rate": 2.4893540459902723e-05,
"loss": 2.9737039184570313,
"step": 40900
},
{
"epoch": 1.5785821672256275,
"grad_norm": 0.6411861777305603,
"learning_rate": 2.462889579475949e-05,
"loss": 2.996640625,
"step": 40950
},
{
"epoch": 1.5805096179792606,
"grad_norm": 0.6503924131393433,
"learning_rate": 2.4365539511688452e-05,
"loss": 2.9758453369140625,
"step": 41000
},
{
"epoch": 1.582437068732894,
"grad_norm": 0.6585896611213684,
"learning_rate": 2.410347431708297e-05,
"loss": 2.9821368408203126,
"step": 41050
},
{
"epoch": 1.584364519486527,
"grad_norm": 0.6517696380615234,
"learning_rate": 2.3842702904068334e-05,
"loss": 2.9787860107421875,
"step": 41100
},
{
"epoch": 1.5862919702401603,
"grad_norm": 0.6550498604774475,
"learning_rate": 2.358322795247433e-05,
"loss": 2.9960858154296877,
"step": 41150
},
{
"epoch": 1.5882194209937937,
"grad_norm": 0.6607086658477783,
"learning_rate": 2.332505212880748e-05,
"loss": 2.994189453125,
"step": 41200
},
{
"epoch": 1.590146871747427,
"grad_norm": 0.6481448411941528,
"learning_rate": 2.306817808622391e-05,
"loss": 2.950994873046875,
"step": 41250
},
{
"epoch": 1.59207432250106,
"grad_norm": 0.6590821743011475,
"learning_rate": 2.281260846450178e-05,
"loss": 2.965792236328125,
"step": 41300
},
{
"epoch": 1.5940017732546934,
"grad_norm": 0.6617052555084229,
"learning_rate": 2.2558345890014445e-05,
"loss": 2.9904522705078125,
"step": 41350
},
{
"epoch": 1.5959292240083265,
"grad_norm": 0.6477006673812866,
"learning_rate": 2.23053929757032e-05,
"loss": 2.961939697265625,
"step": 41400
},
{
"epoch": 1.5978566747619598,
"grad_norm": 0.660005509853363,
"learning_rate": 2.205375232105061e-05,
"loss": 2.9788470458984375,
"step": 41450
},
{
"epoch": 1.5997841255155931,
"grad_norm": 0.6647515892982483,
"learning_rate": 2.1808420126085168e-05,
"loss": 2.9819390869140623,
"step": 41500
},
{
"epoch": 1.6017115762692264,
"grad_norm": 0.6566334962844849,
"learning_rate": 2.1559385361743416e-05,
"loss": 2.993948974609375,
"step": 41550
},
{
"epoch": 1.6036390270228595,
"grad_norm": 0.6431276202201843,
"learning_rate": 2.1311670523442844e-05,
"loss": 2.9922021484375,
"step": 41600
},
{
"epoch": 1.6055664777764929,
"grad_norm": 0.6472877264022827,
"learning_rate": 2.1065278156836606e-05,
"loss": 2.9915176391601563,
"step": 41650
},
{
"epoch": 1.607493928530126,
"grad_norm": 0.6416468024253845,
"learning_rate": 2.0820210793987585e-05,
"loss": 2.97776123046875,
"step": 41700
},
{
"epoch": 1.6094213792837593,
"grad_norm": 0.650649905204773,
"learning_rate": 2.0576470953342013e-05,
"loss": 2.9841278076171873,
"step": 41750
},
{
"epoch": 1.6113488300373926,
"grad_norm": 0.650781512260437,
"learning_rate": 2.033406113970395e-05,
"loss": 2.9824905395507812,
"step": 41800
},
{
"epoch": 1.613276280791026,
"grad_norm": 0.6646984815597534,
"learning_rate": 2.0092983844209233e-05,
"loss": 2.9951324462890625,
"step": 41850
},
{
"epoch": 1.615203731544659,
"grad_norm": 0.6532541513442993,
"learning_rate": 1.9853241544300157e-05,
"loss": 2.975092468261719,
"step": 41900
},
{
"epoch": 1.6171311822982921,
"grad_norm": 0.6527358889579773,
"learning_rate": 1.961483670369967e-05,
"loss": 2.9765081787109375,
"step": 41950
},
{
"epoch": 1.6190586330519254,
"grad_norm": 0.657174825668335,
"learning_rate": 1.9377771772386458e-05,
"loss": 2.9948556518554685,
"step": 42000
},
{
"epoch": 1.6190586330519254,
"eval_loss": 2.940765142440796,
"eval_runtime": 62.2773,
"eval_samples_per_second": 133.949,
"eval_steps_per_second": 4.191,
"step": 42000
},
{
"epoch": 1.6209860838055588,
"grad_norm": 0.7008171081542969,
"learning_rate": 1.9142049186569437e-05,
"loss": 3.0023468017578123,
"step": 42050
},
{
"epoch": 1.622913534559192,
"grad_norm": 0.6562812328338623,
"learning_rate": 1.8907671368662924e-05,
"loss": 2.9704852294921875,
"step": 42100
},
{
"epoch": 1.6248409853128254,
"grad_norm": 0.6428791880607605,
"learning_rate": 1.8674640727261596e-05,
"loss": 2.9770135498046875,
"step": 42150
},
{
"epoch": 1.6267684360664585,
"grad_norm": 0.644219160079956,
"learning_rate": 1.8442959657115886e-05,
"loss": 2.9960116577148437,
"step": 42200
},
{
"epoch": 1.6286958868200916,
"grad_norm": 0.6446249485015869,
"learning_rate": 1.8212630539107216e-05,
"loss": 2.9935546875,
"step": 42250
},
{
"epoch": 1.630623337573725,
"grad_norm": 0.6473619937896729,
"learning_rate": 1.7983655740223674e-05,
"loss": 2.9875433349609377,
"step": 42300
},
{
"epoch": 1.6325507883273582,
"grad_norm": 0.6668522953987122,
"learning_rate": 1.7756037613535568e-05,
"loss": 2.9762387084960937,
"step": 42350
},
{
"epoch": 1.6344782390809915,
"grad_norm": 0.6523205637931824,
"learning_rate": 1.752977849817132e-05,
"loss": 2.9738616943359375,
"step": 42400
},
{
"epoch": 1.6364056898346249,
"grad_norm": 0.6620166301727295,
"learning_rate": 1.7304880719293446e-05,
"loss": 2.9900347900390627,
"step": 42450
},
{
"epoch": 1.638333140588258,
"grad_norm": 0.6480960845947266,
"learning_rate": 1.708134658807453e-05,
"loss": 2.984730529785156,
"step": 42500
},
{
"epoch": 1.640260591341891,
"grad_norm": 0.6719658970832825,
"learning_rate": 1.6859178401673702e-05,
"loss": 2.9981591796875,
"step": 42550
},
{
"epoch": 1.6421880420955244,
"grad_norm": 0.6721029281616211,
"learning_rate": 1.663837844321275e-05,
"loss": 2.9902972412109374,
"step": 42600
},
{
"epoch": 1.6441154928491577,
"grad_norm": 0.6641767024993896,
"learning_rate": 1.641894898175294e-05,
"loss": 2.9818768310546875,
"step": 42650
},
{
"epoch": 1.646042943602791,
"grad_norm": 0.6644138097763062,
"learning_rate": 1.6200892272271434e-05,
"loss": 2.97240478515625,
"step": 42700
},
{
"epoch": 1.6479703943564243,
"grad_norm": 0.6787228584289551,
"learning_rate": 1.5984210555638356e-05,
"loss": 2.9776239013671875,
"step": 42750
},
{
"epoch": 1.6498978451100574,
"grad_norm": 0.6702457666397095,
"learning_rate": 1.5768906058593532e-05,
"loss": 2.970464782714844,
"step": 42800
},
{
"epoch": 1.6518252958636905,
"grad_norm": 0.651032567024231,
"learning_rate": 1.5554980993723866e-05,
"loss": 2.980189208984375,
"step": 42850
},
{
"epoch": 1.6537527466173239,
"grad_norm": 0.6449264883995056,
"learning_rate": 1.5342437559440368e-05,
"loss": 2.9849993896484377,
"step": 42900
},
{
"epoch": 1.6556801973709572,
"grad_norm": 0.6449978351593018,
"learning_rate": 1.5131277939955622e-05,
"loss": 2.9706134033203124,
"step": 42950
},
{
"epoch": 1.6576076481245905,
"grad_norm": 0.659964382648468,
"learning_rate": 1.4921504305261494e-05,
"loss": 2.9645953369140625,
"step": 43000
},
{
"epoch": 1.6595350988782238,
"grad_norm": 0.6383562684059143,
"learning_rate": 1.47131188111066e-05,
"loss": 2.9672946166992187,
"step": 43050
},
{
"epoch": 1.661462549631857,
"grad_norm": 0.6401688456535339,
"learning_rate": 1.4506123598974362e-05,
"loss": 3.0041815185546876,
"step": 43100
},
{
"epoch": 1.66339000038549,
"grad_norm": 0.65959233045578,
"learning_rate": 1.4300520796060782e-05,
"loss": 2.969549255371094,
"step": 43150
},
{
"epoch": 1.6653174511391233,
"grad_norm": 0.6700117588043213,
"learning_rate": 1.4096312515252844e-05,
"loss": 2.987606201171875,
"step": 43200
},
{
"epoch": 1.6672449018927566,
"grad_norm": 0.6429051160812378,
"learning_rate": 1.3893500855106543e-05,
"loss": 2.97798095703125,
"step": 43250
},
{
"epoch": 1.66917235264639,
"grad_norm": 0.6753007173538208,
"learning_rate": 1.3692087899825538e-05,
"loss": 2.9788525390625,
"step": 43300
},
{
"epoch": 1.6710998034000233,
"grad_norm": 0.6652014851570129,
"learning_rate": 1.349207571923947e-05,
"loss": 2.9761001586914064,
"step": 43350
},
{
"epoch": 1.6730272541536564,
"grad_norm": 0.6653535962104797,
"learning_rate": 1.3293466368783045e-05,
"loss": 2.98156982421875,
"step": 43400
},
{
"epoch": 1.6749547049072895,
"grad_norm": 0.6538166999816895,
"learning_rate": 1.3096261889474585e-05,
"loss": 2.9844979858398437,
"step": 43450
},
{
"epoch": 1.6768821556609228,
"grad_norm": 0.6545195579528809,
"learning_rate": 1.2900464307895275e-05,
"loss": 2.97701171875,
"step": 43500
},
{
"epoch": 1.6788096064145561,
"grad_norm": 0.6603434085845947,
"learning_rate": 1.2706075636168168e-05,
"loss": 2.9773919677734373,
"step": 43550
},
{
"epoch": 1.6807370571681894,
"grad_norm": 0.660017728805542,
"learning_rate": 1.2513097871937627e-05,
"loss": 2.9722406005859376,
"step": 43600
},
{
"epoch": 1.6826645079218228,
"grad_norm": 0.6577938199043274,
"learning_rate": 1.2321532998348805e-05,
"loss": 2.97467041015625,
"step": 43650
},
{
"epoch": 1.6845919586754559,
"grad_norm": 0.6334798336029053,
"learning_rate": 1.2131382984027128e-05,
"loss": 2.968843994140625,
"step": 43700
},
{
"epoch": 1.686519409429089,
"grad_norm": 0.6506395936012268,
"learning_rate": 1.1942649783058255e-05,
"loss": 2.9857269287109376,
"step": 43750
},
{
"epoch": 1.6884468601827223,
"grad_norm": 0.665668249130249,
"learning_rate": 1.1759067707676727e-05,
"loss": 2.9876959228515627,
"step": 43800
},
{
"epoch": 1.6903743109363556,
"grad_norm": 0.6575157642364502,
"learning_rate": 1.1573145505083858e-05,
"loss": 2.9751861572265623,
"step": 43850
},
{
"epoch": 1.692301761689989,
"grad_norm": 0.6600083708763123,
"learning_rate": 1.138864585259755e-05,
"loss": 2.9563943481445314,
"step": 43900
},
{
"epoch": 1.694229212443622,
"grad_norm": 0.6317420601844788,
"learning_rate": 1.120557064623721e-05,
"loss": 2.9699560546875,
"step": 43950
},
{
"epoch": 1.6961566631972553,
"grad_norm": 0.6581457853317261,
"learning_rate": 1.1023921767383826e-05,
"loss": 2.971065673828125,
"step": 44000
},
{
"epoch": 1.6961566631972553,
"eval_loss": 2.9355552196502686,
"eval_runtime": 62.5346,
"eval_samples_per_second": 133.398,
"eval_steps_per_second": 4.174,
"step": 44000
},
{
"epoch": 1.6980841139508884,
"grad_norm": 0.6822545528411865,
"learning_rate": 1.0843701082760703e-05,
"loss": 2.981640625,
"step": 44050
},
{
"epoch": 1.7000115647045217,
"grad_norm": 0.6611153483390808,
"learning_rate": 1.0664910444414177e-05,
"loss": 2.9805462646484373,
"step": 44100
},
{
"epoch": 1.701939015458155,
"grad_norm": 0.6585414409637451,
"learning_rate": 1.0487551689694751e-05,
"loss": 2.982611083984375,
"step": 44150
},
{
"epoch": 1.7038664662117884,
"grad_norm": 0.659685492515564,
"learning_rate": 1.0311626641237952e-05,
"loss": 2.96666259765625,
"step": 44200
},
{
"epoch": 1.7057939169654215,
"grad_norm": 0.6496697664260864,
"learning_rate": 1.0137137106945925e-05,
"loss": 2.979638977050781,
"step": 44250
},
{
"epoch": 1.7077213677190548,
"grad_norm": 0.662509024143219,
"learning_rate": 9.964084879968581e-06,
"loss": 2.9772076416015625,
"step": 44300
},
{
"epoch": 1.709648818472688,
"grad_norm": 0.667682409286499,
"learning_rate": 9.79247173868528e-06,
"loss": 2.9707763671875,
"step": 44350
},
{
"epoch": 1.7115762692263212,
"grad_norm": 0.6531767249107361,
"learning_rate": 9.625688760868028e-06,
"loss": 2.9765960693359377,
"step": 44400
},
{
"epoch": 1.7135037199799545,
"grad_norm": 0.6421325206756592,
"learning_rate": 9.456930197933959e-06,
"loss": 2.992750244140625,
"step": 44450
},
{
"epoch": 1.7154311707335879,
"grad_norm": 0.6376025080680847,
"learning_rate": 9.289615932493011e-06,
"loss": 2.96099609375,
"step": 44500
},
{
"epoch": 1.717358621487221,
"grad_norm": 0.6602917909622192,
"learning_rate": 9.123747683958094e-06,
"loss": 2.976321105957031,
"step": 44550
},
{
"epoch": 1.7192860722408543,
"grad_norm": 0.6535906791687012,
"learning_rate": 8.95932715688214e-06,
"loss": 2.961824951171875,
"step": 44600
},
{
"epoch": 1.7212135229944874,
"grad_norm": 0.6507993936538696,
"learning_rate": 8.7963560409404e-06,
"loss": 2.9769668579101562,
"step": 44650
},
{
"epoch": 1.7231409737481207,
"grad_norm": 0.6706106662750244,
"learning_rate": 8.63483601091332e-06,
"loss": 2.9771478271484373,
"step": 44700
},
{
"epoch": 1.725068424501754,
"grad_norm": 0.6625539064407349,
"learning_rate": 8.474768726669045e-06,
"loss": 2.9715240478515623,
"step": 44750
},
{
"epoch": 1.7269958752553873,
"grad_norm": 0.6580246686935425,
"learning_rate": 8.316155833146627e-06,
"loss": 2.9936322021484374,
"step": 44800
},
{
"epoch": 1.7289233260090204,
"grad_norm": 0.6569129228591919,
"learning_rate": 8.158998960338904e-06,
"loss": 2.9744677734375,
"step": 44850
},
{
"epoch": 1.7308507767626538,
"grad_norm": 0.6563003659248352,
"learning_rate": 8.003299723275952e-06,
"loss": 2.97982666015625,
"step": 44900
},
{
"epoch": 1.7327782275162869,
"grad_norm": 0.6553057432174683,
"learning_rate": 7.849059722008282e-06,
"loss": 2.9949554443359374,
"step": 44950
},
{
"epoch": 1.7347056782699202,
"grad_norm": 0.6569401025772095,
"learning_rate": 7.696280541590617e-06,
"loss": 2.9625689697265627,
"step": 45000
},
{
"epoch": 1.7366331290235535,
"grad_norm": 0.6830917000770569,
"learning_rate": 7.544963752065347e-06,
"loss": 2.9723980712890623,
"step": 45050
},
{
"epoch": 1.7385605797771868,
"grad_norm": 0.6623573899269104,
"learning_rate": 7.395110908446678e-06,
"loss": 2.974309997558594,
"step": 45100
},
{
"epoch": 1.74048803053082,
"grad_norm": 0.6592599153518677,
"learning_rate": 7.246723550704358e-06,
"loss": 2.9534552001953127,
"step": 45150
},
{
"epoch": 1.7424154812844532,
"grad_norm": 0.6647738814353943,
"learning_rate": 7.0998032037481385e-06,
"loss": 2.970636901855469,
"step": 45200
},
{
"epoch": 1.7443429320380863,
"grad_norm": 0.6631134748458862,
"learning_rate": 6.95435137741186e-06,
"loss": 2.971834411621094,
"step": 45250
},
{
"epoch": 1.7462703827917196,
"grad_norm": 0.6524649858474731,
"learning_rate": 6.810369566438101e-06,
"loss": 2.9704510498046877,
"step": 45300
},
{
"epoch": 1.748197833545353,
"grad_norm": 0.662055253982544,
"learning_rate": 6.667859250462743e-06,
"loss": 2.983234558105469,
"step": 45350
},
{
"epoch": 1.7501252842989863,
"grad_norm": 0.6571261882781982,
"learning_rate": 6.526821893999762e-06,
"loss": 2.9623980712890625,
"step": 45400
},
{
"epoch": 1.7520527350526194,
"grad_norm": 0.6639049649238586,
"learning_rate": 6.387258946426244e-06,
"loss": 2.970660400390625,
"step": 45450
},
{
"epoch": 1.7539801858062527,
"grad_norm": 0.6736387610435486,
"learning_rate": 6.249171841967376e-06,
"loss": 2.974384765625,
"step": 45500
},
{
"epoch": 1.7559076365598858,
"grad_norm": 0.6472794413566589,
"learning_rate": 6.112561999681831e-06,
"loss": 2.983541259765625,
"step": 45550
},
{
"epoch": 1.7578350873135191,
"grad_norm": 0.6505047678947449,
"learning_rate": 5.977430823447071e-06,
"loss": 2.968780517578125,
"step": 45600
},
{
"epoch": 1.7597625380671524,
"grad_norm": 0.6508398652076721,
"learning_rate": 5.8437797019450506e-06,
"loss": 2.9603814697265625,
"step": 45650
},
{
"epoch": 1.7616899888207858,
"grad_norm": 0.6736019253730774,
"learning_rate": 5.711610008647804e-06,
"loss": 2.9674127197265623,
"step": 45700
},
{
"epoch": 1.7636174395744189,
"grad_norm": 0.6542773842811584,
"learning_rate": 5.58092310180343e-06,
"loss": 2.975087890625,
"step": 45750
},
{
"epoch": 1.765544890328052,
"grad_norm": 0.656532883644104,
"learning_rate": 5.451720324422099e-06,
"loss": 2.9654754638671874,
"step": 45800
},
{
"epoch": 1.7674723410816853,
"grad_norm": 0.6590448021888733,
"learning_rate": 5.324003004262261e-06,
"loss": 2.9947637939453124,
"step": 45850
},
{
"epoch": 1.7693997918353186,
"grad_norm": 0.6381296515464783,
"learning_rate": 5.197772453816984e-06,
"loss": 2.9542913818359375,
"step": 45900
},
{
"epoch": 1.771327242588952,
"grad_norm": 0.6407474279403687,
"learning_rate": 5.0730299703004515e-06,
"loss": 2.9807843017578124,
"step": 45950
},
{
"epoch": 1.7732546933425852,
"grad_norm": 0.6471443772315979,
"learning_rate": 4.94977683563475e-06,
"loss": 2.97317626953125,
"step": 46000
},
{
"epoch": 1.7732546933425852,
"eval_loss": 2.930370807647705,
"eval_runtime": 62.496,
"eval_samples_per_second": 133.481,
"eval_steps_per_second": 4.176,
"step": 46000
}
],
"logging_steps": 50,
"max_steps": 50000,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 2000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 7.69243742208e+17,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}