Invalid JSON:Unexpected token 'I', ..."ad_norm": Infinity,
"... is not valid JSON
{
"best_global_step": 100000,
"best_metric": 3.2741730213165283,
"best_model_checkpoint": "/scratch/cl5625/exceptions/models/high_2000_8397/checkpoint-100000",
"epoch": 32.29312015503876,
"eval_steps": 1000,
"global_step": 120000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.013458225667527994,
"grad_norm": 2.913999080657959,
"learning_rate": 0.000294,
"loss": 8.4436,
"step": 50
},
{
"epoch": 0.026916451335055987,
"grad_norm": 2.57014799118042,
"learning_rate": 0.0005939999999999999,
"loss": 6.7256,
"step": 100
},
{
"epoch": 0.04037467700258398,
"grad_norm": 1.4274643659591675,
"learning_rate": 0.0005998020735155513,
"loss": 6.2787,
"step": 150
},
{
"epoch": 0.053832902670111975,
"grad_norm": 1.2983430624008179,
"learning_rate": 0.0005996001077150935,
"loss": 6.0173,
"step": 200
},
{
"epoch": 0.06729112833763996,
"grad_norm": 1.371605396270752,
"learning_rate": 0.0005993981419146358,
"loss": 5.8389,
"step": 250
},
{
"epoch": 0.08074935400516796,
"grad_norm": 1.4190587997436523,
"learning_rate": 0.0005991961761141779,
"loss": 5.7187,
"step": 300
},
{
"epoch": 0.09420757967269595,
"grad_norm": 1.192716360092163,
"learning_rate": 0.0005989942103137202,
"loss": 5.5788,
"step": 350
},
{
"epoch": 0.10766580534022395,
"grad_norm": 1.3442068099975586,
"learning_rate": 0.0005987922445132624,
"loss": 5.4524,
"step": 400
},
{
"epoch": 0.12112403100775193,
"grad_norm": 1.0547122955322266,
"learning_rate": 0.0005985902787128047,
"loss": 5.327,
"step": 450
},
{
"epoch": 0.13458225667527993,
"grad_norm": 0.9016146063804626,
"learning_rate": 0.0005983883129123468,
"loss": 5.2192,
"step": 500
},
{
"epoch": 0.14804048234280792,
"grad_norm": 1.7201956510543823,
"learning_rate": 0.000598186347111889,
"loss": 5.1144,
"step": 550
},
{
"epoch": 0.16149870801033592,
"grad_norm": 1.1931437253952026,
"learning_rate": 0.0005979843813114312,
"loss": 5.0328,
"step": 600
},
{
"epoch": 0.1749569336778639,
"grad_norm": 1.468700647354126,
"learning_rate": 0.0005977824155109734,
"loss": 4.9485,
"step": 650
},
{
"epoch": 0.1884151593453919,
"grad_norm": 1.1093205213546753,
"learning_rate": 0.0005975804497105157,
"loss": 4.8785,
"step": 700
},
{
"epoch": 0.2018733850129199,
"grad_norm": 0.9616383910179138,
"learning_rate": 0.0005973784839100578,
"loss": 4.83,
"step": 750
},
{
"epoch": 0.2153316106804479,
"grad_norm": 0.8918668627738953,
"learning_rate": 0.0005971765181096001,
"loss": 4.7774,
"step": 800
},
{
"epoch": 0.2287898363479759,
"grad_norm": 1.216214895248413,
"learning_rate": 0.0005969745523091422,
"loss": 4.7205,
"step": 850
},
{
"epoch": 0.24224806201550386,
"grad_norm": 1.0829071998596191,
"learning_rate": 0.0005967725865086845,
"loss": 4.667,
"step": 900
},
{
"epoch": 0.2557062876830319,
"grad_norm": 0.8637637495994568,
"learning_rate": 0.0005965706207082267,
"loss": 4.62,
"step": 950
},
{
"epoch": 0.26916451335055985,
"grad_norm": 0.7872169613838196,
"learning_rate": 0.0005963686549077689,
"loss": 4.569,
"step": 1000
},
{
"epoch": 0.26916451335055985,
"eval_accuracy": 0.26840325627663747,
"eval_loss": 4.504179954528809,
"eval_runtime": 185.6371,
"eval_samples_per_second": 97.028,
"eval_steps_per_second": 6.066,
"step": 1000
},
{
"epoch": 0.2826227390180879,
"grad_norm": 0.9128012657165527,
"learning_rate": 0.0005961666891073111,
"loss": 4.5257,
"step": 1050
},
{
"epoch": 0.29608096468561584,
"grad_norm": 1.097090721130371,
"learning_rate": 0.0005959647233068533,
"loss": 4.4947,
"step": 1100
},
{
"epoch": 0.30953919035314387,
"grad_norm": 1.179639220237732,
"learning_rate": 0.0005957627575063955,
"loss": 4.4512,
"step": 1150
},
{
"epoch": 0.32299741602067183,
"grad_norm": 0.9106621146202087,
"learning_rate": 0.0005955607917059378,
"loss": 4.4088,
"step": 1200
},
{
"epoch": 0.3364556416881998,
"grad_norm": 0.7705498337745667,
"learning_rate": 0.00059535882590548,
"loss": 4.3808,
"step": 1250
},
{
"epoch": 0.3499138673557278,
"grad_norm": 1.0174599885940552,
"learning_rate": 0.0005951568601050221,
"loss": 4.3655,
"step": 1300
},
{
"epoch": 0.3633720930232558,
"grad_norm": 0.776151180267334,
"learning_rate": 0.0005949548943045644,
"loss": 4.3143,
"step": 1350
},
{
"epoch": 0.3768303186907838,
"grad_norm": 0.6459659934043884,
"learning_rate": 0.0005947529285041066,
"loss": 4.2846,
"step": 1400
},
{
"epoch": 0.3902885443583118,
"grad_norm": 0.6657201647758484,
"learning_rate": 0.0005945509627036488,
"loss": 4.2581,
"step": 1450
},
{
"epoch": 0.4037467700258398,
"grad_norm": 1.0717779397964478,
"learning_rate": 0.000594348996903191,
"loss": 4.2406,
"step": 1500
},
{
"epoch": 0.4172049956933678,
"grad_norm": 0.7116280198097229,
"learning_rate": 0.0005941470311027332,
"loss": 4.2383,
"step": 1550
},
{
"epoch": 0.4306632213608958,
"grad_norm": 0.6161189079284668,
"learning_rate": 0.0005939450653022754,
"loss": 4.2077,
"step": 1600
},
{
"epoch": 0.44412144702842377,
"grad_norm": 0.8404054641723633,
"learning_rate": 0.0005937430995018177,
"loss": 4.1763,
"step": 1650
},
{
"epoch": 0.4575796726959518,
"grad_norm": 0.6853687763214111,
"learning_rate": 0.0005935411337013598,
"loss": 4.1794,
"step": 1700
},
{
"epoch": 0.47103789836347976,
"grad_norm": 0.6380472183227539,
"learning_rate": 0.0005933391679009021,
"loss": 4.16,
"step": 1750
},
{
"epoch": 0.4844961240310077,
"grad_norm": 0.6678838133811951,
"learning_rate": 0.0005931372021004442,
"loss": 4.1278,
"step": 1800
},
{
"epoch": 0.49795434969853575,
"grad_norm": 0.7158453464508057,
"learning_rate": 0.0005929352362999865,
"loss": 4.1155,
"step": 1850
},
{
"epoch": 0.5114125753660638,
"grad_norm": 0.7045041918754578,
"learning_rate": 0.0005927332704995287,
"loss": 4.1073,
"step": 1900
},
{
"epoch": 0.5248708010335917,
"grad_norm": 0.584650993347168,
"learning_rate": 0.0005925313046990709,
"loss": 4.1065,
"step": 1950
},
{
"epoch": 0.5383290267011197,
"grad_norm": 0.7326251268386841,
"learning_rate": 0.0005923293388986131,
"loss": 4.0844,
"step": 2000
},
{
"epoch": 0.5383290267011197,
"eval_accuracy": 0.31736604820530606,
"eval_loss": 4.022258281707764,
"eval_runtime": 186.6439,
"eval_samples_per_second": 96.505,
"eval_steps_per_second": 6.033,
"step": 2000
},
{
"epoch": 0.5517872523686477,
"grad_norm": 0.7888703346252441,
"learning_rate": 0.0005921273730981553,
"loss": 4.0745,
"step": 2050
},
{
"epoch": 0.5652454780361758,
"grad_norm": 0.7352796196937561,
"learning_rate": 0.0005919254072976975,
"loss": 4.0406,
"step": 2100
},
{
"epoch": 0.5787037037037037,
"grad_norm": 0.6670855283737183,
"learning_rate": 0.0005917234414972398,
"loss": 4.0465,
"step": 2150
},
{
"epoch": 0.5921619293712317,
"grad_norm": 0.635707437992096,
"learning_rate": 0.000591521475696782,
"loss": 4.0324,
"step": 2200
},
{
"epoch": 0.6056201550387597,
"grad_norm": 0.5971513986587524,
"learning_rate": 0.0005913195098963241,
"loss": 4.0291,
"step": 2250
},
{
"epoch": 0.6190783807062877,
"grad_norm": 0.5715792775154114,
"learning_rate": 0.0005911175440958664,
"loss": 4.0216,
"step": 2300
},
{
"epoch": 0.6325366063738157,
"grad_norm": 0.664419412612915,
"learning_rate": 0.0005909155782954086,
"loss": 3.988,
"step": 2350
},
{
"epoch": 0.6459948320413437,
"grad_norm": 0.6859356760978699,
"learning_rate": 0.0005907136124949508,
"loss": 3.9883,
"step": 2400
},
{
"epoch": 0.6594530577088716,
"grad_norm": 0.6854323744773865,
"learning_rate": 0.000590511646694493,
"loss": 3.983,
"step": 2450
},
{
"epoch": 0.6729112833763996,
"grad_norm": 0.5573345422744751,
"learning_rate": 0.0005903096808940352,
"loss": 3.9829,
"step": 2500
},
{
"epoch": 0.6863695090439277,
"grad_norm": 0.5315501689910889,
"learning_rate": 0.0005901077150935774,
"loss": 3.9566,
"step": 2550
},
{
"epoch": 0.6998277347114557,
"grad_norm": 0.5556917190551758,
"learning_rate": 0.0005899057492931197,
"loss": 3.9523,
"step": 2600
},
{
"epoch": 0.7132859603789836,
"grad_norm": 0.5229679346084595,
"learning_rate": 0.0005897037834926618,
"loss": 3.951,
"step": 2650
},
{
"epoch": 0.7267441860465116,
"grad_norm": 0.5850522518157959,
"learning_rate": 0.0005895018176922041,
"loss": 3.9428,
"step": 2700
},
{
"epoch": 0.7402024117140397,
"grad_norm": 0.5948578119277954,
"learning_rate": 0.0005892998518917462,
"loss": 3.927,
"step": 2750
},
{
"epoch": 0.7536606373815676,
"grad_norm": 0.5638688802719116,
"learning_rate": 0.0005890978860912885,
"loss": 3.9215,
"step": 2800
},
{
"epoch": 0.7671188630490956,
"grad_norm": 0.5502722859382629,
"learning_rate": 0.0005888959202908307,
"loss": 3.9149,
"step": 2850
},
{
"epoch": 0.7805770887166236,
"grad_norm": 0.6695708632469177,
"learning_rate": 0.0005886939544903729,
"loss": 3.9088,
"step": 2900
},
{
"epoch": 0.7940353143841516,
"grad_norm": 0.46178966760635376,
"learning_rate": 0.0005884919886899151,
"loss": 3.9,
"step": 2950
},
{
"epoch": 0.8074935400516796,
"grad_norm": 0.5293999910354614,
"learning_rate": 0.0005882900228894573,
"loss": 3.8998,
"step": 3000
},
{
"epoch": 0.8074935400516796,
"eval_accuracy": 0.3341444907569774,
"eval_loss": 3.835969924926758,
"eval_runtime": 183.5002,
"eval_samples_per_second": 98.158,
"eval_steps_per_second": 6.136,
"step": 3000
},
{
"epoch": 0.8209517657192076,
"grad_norm": 0.5343823432922363,
"learning_rate": 0.0005880880570889996,
"loss": 3.8891,
"step": 3050
},
{
"epoch": 0.8344099913867356,
"grad_norm": 0.5775337815284729,
"learning_rate": 0.0005878860912885418,
"loss": 3.8889,
"step": 3100
},
{
"epoch": 0.8478682170542635,
"grad_norm": 0.46231314539909363,
"learning_rate": 0.000587684125488084,
"loss": 3.8704,
"step": 3150
},
{
"epoch": 0.8613264427217916,
"grad_norm": 0.47632086277008057,
"learning_rate": 0.0005874821596876261,
"loss": 3.8705,
"step": 3200
},
{
"epoch": 0.8747846683893196,
"grad_norm": 0.44586512446403503,
"learning_rate": 0.0005872801938871684,
"loss": 3.8572,
"step": 3250
},
{
"epoch": 0.8882428940568475,
"grad_norm": 0.46257761120796204,
"learning_rate": 0.0005870782280867106,
"loss": 3.855,
"step": 3300
},
{
"epoch": 0.9017011197243755,
"grad_norm": 0.47898146510124207,
"learning_rate": 0.0005868762622862528,
"loss": 3.8568,
"step": 3350
},
{
"epoch": 0.9151593453919036,
"grad_norm": 0.4588397741317749,
"learning_rate": 0.000586674296485795,
"loss": 3.8413,
"step": 3400
},
{
"epoch": 0.9286175710594315,
"grad_norm": 0.5241335034370422,
"learning_rate": 0.0005864723306853373,
"loss": 3.8342,
"step": 3450
},
{
"epoch": 0.9420757967269595,
"grad_norm": 0.4803313910961151,
"learning_rate": 0.0005862703648848794,
"loss": 3.8278,
"step": 3500
},
{
"epoch": 0.9555340223944875,
"grad_norm": 0.5260579586029053,
"learning_rate": 0.0005860683990844217,
"loss": 3.833,
"step": 3550
},
{
"epoch": 0.9689922480620154,
"grad_norm": 0.4580845832824707,
"learning_rate": 0.0005858664332839638,
"loss": 3.8272,
"step": 3600
},
{
"epoch": 0.9824504737295435,
"grad_norm": 0.47201746702194214,
"learning_rate": 0.0005856644674835061,
"loss": 3.8088,
"step": 3650
},
{
"epoch": 0.9959086993970715,
"grad_norm": 0.4708392918109894,
"learning_rate": 0.0005854625016830483,
"loss": 3.8142,
"step": 3700
},
{
"epoch": 1.009151593453919,
"grad_norm": 0.4994507133960724,
"learning_rate": 0.0005852605358825905,
"loss": 3.7753,
"step": 3750
},
{
"epoch": 1.022609819121447,
"grad_norm": 0.4856177270412445,
"learning_rate": 0.0005850585700821327,
"loss": 3.7344,
"step": 3800
},
{
"epoch": 1.036068044788975,
"grad_norm": 0.4874962568283081,
"learning_rate": 0.0005848566042816749,
"loss": 3.7375,
"step": 3850
},
{
"epoch": 1.049526270456503,
"grad_norm": 0.47349369525909424,
"learning_rate": 0.0005846546384812171,
"loss": 3.7266,
"step": 3900
},
{
"epoch": 1.062984496124031,
"grad_norm": 0.4920184314250946,
"learning_rate": 0.0005844526726807594,
"loss": 3.7349,
"step": 3950
},
{
"epoch": 1.076442721791559,
"grad_norm": 0.427706241607666,
"learning_rate": 0.0005842507068803016,
"loss": 3.7377,
"step": 4000
},
{
"epoch": 1.076442721791559,
"eval_accuracy": 0.3441772673403641,
"eval_loss": 3.732637405395508,
"eval_runtime": 183.7058,
"eval_samples_per_second": 98.048,
"eval_steps_per_second": 6.129,
"step": 4000
},
{
"epoch": 1.089900947459087,
"grad_norm": 0.4717354476451874,
"learning_rate": 0.0005840487410798437,
"loss": 3.7301,
"step": 4050
},
{
"epoch": 1.103359173126615,
"grad_norm": 0.4636068344116211,
"learning_rate": 0.000583846775279386,
"loss": 3.736,
"step": 4100
},
{
"epoch": 1.116817398794143,
"grad_norm": 0.43443241715431213,
"learning_rate": 0.0005836448094789282,
"loss": 3.7398,
"step": 4150
},
{
"epoch": 1.130275624461671,
"grad_norm": 0.4374564290046692,
"learning_rate": 0.0005834428436784704,
"loss": 3.7323,
"step": 4200
},
{
"epoch": 1.143733850129199,
"grad_norm": 0.4285607933998108,
"learning_rate": 0.0005832408778780126,
"loss": 3.7224,
"step": 4250
},
{
"epoch": 1.157192075796727,
"grad_norm": 0.5262153744697571,
"learning_rate": 0.0005830389120775548,
"loss": 3.7272,
"step": 4300
},
{
"epoch": 1.1706503014642549,
"grad_norm": 0.41815322637557983,
"learning_rate": 0.000582836946277097,
"loss": 3.7274,
"step": 4350
},
{
"epoch": 1.1841085271317828,
"grad_norm": 0.4538526237010956,
"learning_rate": 0.0005826349804766393,
"loss": 3.7265,
"step": 4400
},
{
"epoch": 1.197566752799311,
"grad_norm": 0.4141399562358856,
"learning_rate": 0.0005824330146761814,
"loss": 3.7141,
"step": 4450
},
{
"epoch": 1.211024978466839,
"grad_norm": 0.42039555311203003,
"learning_rate": 0.0005822310488757237,
"loss": 3.7048,
"step": 4500
},
{
"epoch": 1.224483204134367,
"grad_norm": 0.4190801978111267,
"learning_rate": 0.0005820290830752658,
"loss": 3.7014,
"step": 4550
},
{
"epoch": 1.237941429801895,
"grad_norm": 0.43000900745391846,
"learning_rate": 0.0005818271172748081,
"loss": 3.6964,
"step": 4600
},
{
"epoch": 1.251399655469423,
"grad_norm": 0.4709755480289459,
"learning_rate": 0.0005816251514743503,
"loss": 3.6989,
"step": 4650
},
{
"epoch": 1.2648578811369509,
"grad_norm": 0.43646031618118286,
"learning_rate": 0.0005814231856738925,
"loss": 3.6967,
"step": 4700
},
{
"epoch": 1.2783161068044788,
"grad_norm": 0.46974101662635803,
"learning_rate": 0.0005812212198734347,
"loss": 3.6999,
"step": 4750
},
{
"epoch": 1.291774332472007,
"grad_norm": 0.45160266757011414,
"learning_rate": 0.0005810192540729769,
"loss": 3.6927,
"step": 4800
},
{
"epoch": 1.3052325581395348,
"grad_norm": 0.42115554213523865,
"learning_rate": 0.0005808172882725192,
"loss": 3.7006,
"step": 4850
},
{
"epoch": 1.318690783807063,
"grad_norm": 0.4420790672302246,
"learning_rate": 0.0005806153224720614,
"loss": 3.6923,
"step": 4900
},
{
"epoch": 1.332149009474591,
"grad_norm": 0.4141576886177063,
"learning_rate": 0.0005804133566716036,
"loss": 3.6851,
"step": 4950
},
{
"epoch": 1.3456072351421189,
"grad_norm": 0.40737617015838623,
"learning_rate": 0.0005802113908711457,
"loss": 3.6862,
"step": 5000
},
{
"epoch": 1.3456072351421189,
"eval_accuracy": 0.35089142572053506,
"eval_loss": 3.664952516555786,
"eval_runtime": 184.0369,
"eval_samples_per_second": 97.872,
"eval_steps_per_second": 6.118,
"step": 5000
},
{
"epoch": 1.3590654608096469,
"grad_norm": 0.42510783672332764,
"learning_rate": 0.000580009425070688,
"loss": 3.6741,
"step": 5050
},
{
"epoch": 1.3725236864771748,
"grad_norm": 0.37819960713386536,
"learning_rate": 0.0005798074592702302,
"loss": 3.6908,
"step": 5100
},
{
"epoch": 1.3859819121447028,
"grad_norm": 0.45490550994873047,
"learning_rate": 0.0005796054934697724,
"loss": 3.6831,
"step": 5150
},
{
"epoch": 1.3994401378122308,
"grad_norm": 0.41686487197875977,
"learning_rate": 0.0005794035276693146,
"loss": 3.6754,
"step": 5200
},
{
"epoch": 1.412898363479759,
"grad_norm": 0.3555591404438019,
"learning_rate": 0.0005792015618688568,
"loss": 3.6794,
"step": 5250
},
{
"epoch": 1.4263565891472867,
"grad_norm": 0.4676069915294647,
"learning_rate": 0.000578999596068399,
"loss": 3.6623,
"step": 5300
},
{
"epoch": 1.4398148148148149,
"grad_norm": 0.4197990894317627,
"learning_rate": 0.0005787976302679413,
"loss": 3.6771,
"step": 5350
},
{
"epoch": 1.4532730404823428,
"grad_norm": 0.399293452501297,
"learning_rate": 0.0005785956644674834,
"loss": 3.6711,
"step": 5400
},
{
"epoch": 1.4667312661498708,
"grad_norm": 0.4178318977355957,
"learning_rate": 0.0005783936986670257,
"loss": 3.6648,
"step": 5450
},
{
"epoch": 1.4801894918173988,
"grad_norm": 0.42317429184913635,
"learning_rate": 0.0005781917328665678,
"loss": 3.6587,
"step": 5500
},
{
"epoch": 1.4936477174849268,
"grad_norm": 0.3713272213935852,
"learning_rate": 0.0005779897670661101,
"loss": 3.6516,
"step": 5550
},
{
"epoch": 1.507105943152455,
"grad_norm": 0.36793479323387146,
"learning_rate": 0.0005777878012656523,
"loss": 3.6562,
"step": 5600
},
{
"epoch": 1.5205641688199827,
"grad_norm": 0.4037824273109436,
"learning_rate": 0.0005775858354651945,
"loss": 3.6604,
"step": 5650
},
{
"epoch": 1.5340223944875109,
"grad_norm": 0.48168888688087463,
"learning_rate": 0.0005773838696647367,
"loss": 3.6491,
"step": 5700
},
{
"epoch": 1.5474806201550386,
"grad_norm": 0.3919402062892914,
"learning_rate": 0.000577181903864279,
"loss": 3.6503,
"step": 5750
},
{
"epoch": 1.5609388458225668,
"grad_norm": 0.4166146516799927,
"learning_rate": 0.0005769799380638212,
"loss": 3.6417,
"step": 5800
},
{
"epoch": 1.5743970714900948,
"grad_norm": 0.4657573103904724,
"learning_rate": 0.0005767779722633633,
"loss": 3.641,
"step": 5850
},
{
"epoch": 1.5878552971576227,
"grad_norm": 0.42925915122032166,
"learning_rate": 0.0005765760064629056,
"loss": 3.6397,
"step": 5900
},
{
"epoch": 1.6013135228251507,
"grad_norm": 0.3777695596218109,
"learning_rate": 0.0005763740406624477,
"loss": 3.6299,
"step": 5950
},
{
"epoch": 1.6147717484926787,
"grad_norm": 0.37473684549331665,
"learning_rate": 0.00057617207486199,
"loss": 3.6327,
"step": 6000
},
{
"epoch": 1.6147717484926787,
"eval_accuracy": 0.3564730492783024,
"eval_loss": 3.609661817550659,
"eval_runtime": 184.2829,
"eval_samples_per_second": 97.741,
"eval_steps_per_second": 6.11,
"step": 6000
},
{
"epoch": 1.6282299741602069,
"grad_norm": 0.39562341570854187,
"learning_rate": 0.0005759701090615322,
"loss": 3.6378,
"step": 6050
},
{
"epoch": 1.6416881998277346,
"grad_norm": 0.38754016160964966,
"learning_rate": 0.0005757681432610744,
"loss": 3.6255,
"step": 6100
},
{
"epoch": 1.6551464254952628,
"grad_norm": 0.39801645278930664,
"learning_rate": 0.0005755661774606166,
"loss": 3.6319,
"step": 6150
},
{
"epoch": 1.6686046511627906,
"grad_norm": 0.42124342918395996,
"learning_rate": 0.0005753642116601588,
"loss": 3.6243,
"step": 6200
},
{
"epoch": 1.6820628768303187,
"grad_norm": 0.38019752502441406,
"learning_rate": 0.000575162245859701,
"loss": 3.6336,
"step": 6250
},
{
"epoch": 1.6955211024978467,
"grad_norm": 0.4245755970478058,
"learning_rate": 0.0005749602800592433,
"loss": 3.6165,
"step": 6300
},
{
"epoch": 1.7089793281653747,
"grad_norm": 0.38058555126190186,
"learning_rate": 0.0005747583142587854,
"loss": 3.6209,
"step": 6350
},
{
"epoch": 1.7224375538329026,
"grad_norm": 0.3985741138458252,
"learning_rate": 0.0005745563484583277,
"loss": 3.6316,
"step": 6400
},
{
"epoch": 1.7358957795004306,
"grad_norm": 0.38295063376426697,
"learning_rate": 0.0005743543826578698,
"loss": 3.6254,
"step": 6450
},
{
"epoch": 1.7493540051679588,
"grad_norm": 0.3692377805709839,
"learning_rate": 0.0005741524168574121,
"loss": 3.6074,
"step": 6500
},
{
"epoch": 1.7628122308354865,
"grad_norm": 0.37884026765823364,
"learning_rate": 0.0005739504510569543,
"loss": 3.6256,
"step": 6550
},
{
"epoch": 1.7762704565030147,
"grad_norm": 0.40681174397468567,
"learning_rate": 0.0005737484852564966,
"loss": 3.6121,
"step": 6600
},
{
"epoch": 1.7897286821705425,
"grad_norm": 0.38766586780548096,
"learning_rate": 0.0005735465194560387,
"loss": 3.6114,
"step": 6650
},
{
"epoch": 1.8031869078380707,
"grad_norm": 0.36811351776123047,
"learning_rate": 0.000573344553655581,
"loss": 3.6241,
"step": 6700
},
{
"epoch": 1.8166451335055986,
"grad_norm": 0.35865598917007446,
"learning_rate": 0.0005731425878551232,
"loss": 3.6103,
"step": 6750
},
{
"epoch": 1.8301033591731266,
"grad_norm": 0.38588201999664307,
"learning_rate": 0.0005729406220546653,
"loss": 3.602,
"step": 6800
},
{
"epoch": 1.8435615848406546,
"grad_norm": 0.3562093675136566,
"learning_rate": 0.0005727386562542076,
"loss": 3.6094,
"step": 6850
},
{
"epoch": 1.8570198105081825,
"grad_norm": 0.3887605667114258,
"learning_rate": 0.0005725366904537497,
"loss": 3.5926,
"step": 6900
},
{
"epoch": 1.8704780361757107,
"grad_norm": 0.3471638858318329,
"learning_rate": 0.000572334724653292,
"loss": 3.6032,
"step": 6950
},
{
"epoch": 1.8839362618432385,
"grad_norm": 0.4244278073310852,
"learning_rate": 0.0005721327588528342,
"loss": 3.6095,
"step": 7000
},
{
"epoch": 1.8839362618432385,
"eval_accuracy": 0.3600449233018388,
"eval_loss": 3.566401720046997,
"eval_runtime": 184.446,
"eval_samples_per_second": 97.655,
"eval_steps_per_second": 6.105,
"step": 7000
},
{
"epoch": 1.8973944875107667,
"grad_norm": 0.404874324798584,
"learning_rate": 0.0005719307930523764,
"loss": 3.5949,
"step": 7050
},
{
"epoch": 1.9108527131782944,
"grad_norm": 0.4324790835380554,
"learning_rate": 0.0005717288272519186,
"loss": 3.5988,
"step": 7100
},
{
"epoch": 1.9243109388458226,
"grad_norm": 0.3776759207248688,
"learning_rate": 0.0005715268614514608,
"loss": 3.5972,
"step": 7150
},
{
"epoch": 1.9377691645133506,
"grad_norm": 0.4079386591911316,
"learning_rate": 0.000571324895651003,
"loss": 3.5963,
"step": 7200
},
{
"epoch": 1.9512273901808785,
"grad_norm": 0.3583625853061676,
"learning_rate": 0.0005711229298505453,
"loss": 3.5889,
"step": 7250
},
{
"epoch": 1.9646856158484065,
"grad_norm": 0.3998764753341675,
"learning_rate": 0.0005709209640500874,
"loss": 3.5879,
"step": 7300
},
{
"epoch": 1.9781438415159345,
"grad_norm": 0.3784210979938507,
"learning_rate": 0.0005707189982496297,
"loss": 3.5836,
"step": 7350
},
{
"epoch": 1.9916020671834627,
"grad_norm": 0.34722816944122314,
"learning_rate": 0.0005705170324491719,
"loss": 3.5783,
"step": 7400
},
{
"epoch": 2.00484496124031,
"grad_norm": 0.37250837683677673,
"learning_rate": 0.0005703150666487141,
"loss": 3.5462,
"step": 7450
},
{
"epoch": 2.018303186907838,
"grad_norm": 0.35703244805336,
"learning_rate": 0.0005701131008482563,
"loss": 3.4976,
"step": 7500
},
{
"epoch": 2.0317614125753662,
"grad_norm": 0.4143078029155731,
"learning_rate": 0.0005699111350477986,
"loss": 3.5049,
"step": 7550
},
{
"epoch": 2.045219638242894,
"grad_norm": 0.39301028847694397,
"learning_rate": 0.0005697091692473408,
"loss": 3.4977,
"step": 7600
},
{
"epoch": 2.058677863910422,
"grad_norm": 0.3683079183101654,
"learning_rate": 0.000569507203446883,
"loss": 3.4922,
"step": 7650
},
{
"epoch": 2.07213608957795,
"grad_norm": 0.4449765384197235,
"learning_rate": 0.0005693052376464252,
"loss": 3.5022,
"step": 7700
},
{
"epoch": 2.085594315245478,
"grad_norm": 0.37594074010849,
"learning_rate": 0.0005691032718459673,
"loss": 3.4973,
"step": 7750
},
{
"epoch": 2.099052540913006,
"grad_norm": 0.4027359187602997,
"learning_rate": 0.0005689013060455096,
"loss": 3.5087,
"step": 7800
},
{
"epoch": 2.112510766580534,
"grad_norm": 0.3720776438713074,
"learning_rate": 0.0005686993402450518,
"loss": 3.5034,
"step": 7850
},
{
"epoch": 2.125968992248062,
"grad_norm": 0.3649153411388397,
"learning_rate": 0.000568497374444594,
"loss": 3.5076,
"step": 7900
},
{
"epoch": 2.13942721791559,
"grad_norm": 0.3859637379646301,
"learning_rate": 0.0005682954086441362,
"loss": 3.5163,
"step": 7950
},
{
"epoch": 2.152885443583118,
"grad_norm": 0.3887077867984772,
"learning_rate": 0.0005680934428436784,
"loss": 3.4977,
"step": 8000
},
{
"epoch": 2.152885443583118,
"eval_accuracy": 0.3636901339528812,
"eval_loss": 3.537388324737549,
"eval_runtime": 184.6441,
"eval_samples_per_second": 97.55,
"eval_steps_per_second": 6.098,
"step": 8000
},
{
"epoch": 2.166343669250646,
"grad_norm": 0.38032057881355286,
"learning_rate": 0.0005678914770432206,
"loss": 3.4899,
"step": 8050
},
{
"epoch": 2.179801894918174,
"grad_norm": 0.38354671001434326,
"learning_rate": 0.0005676895112427629,
"loss": 3.5082,
"step": 8100
},
{
"epoch": 2.193260120585702,
"grad_norm": 0.3909951448440552,
"learning_rate": 0.000567487545442305,
"loss": 3.4917,
"step": 8150
},
{
"epoch": 2.20671834625323,
"grad_norm": 0.3544139862060547,
"learning_rate": 0.0005672855796418473,
"loss": 3.5021,
"step": 8200
},
{
"epoch": 2.220176571920758,
"grad_norm": 0.3458568751811981,
"learning_rate": 0.0005670836138413894,
"loss": 3.4959,
"step": 8250
},
{
"epoch": 2.233634797588286,
"grad_norm": 0.40227895975112915,
"learning_rate": 0.0005668816480409317,
"loss": 3.5013,
"step": 8300
},
{
"epoch": 2.2470930232558137,
"grad_norm": 0.32855820655822754,
"learning_rate": 0.0005666796822404739,
"loss": 3.5069,
"step": 8350
},
{
"epoch": 2.260551248923342,
"grad_norm": 0.39815717935562134,
"learning_rate": 0.0005664777164400162,
"loss": 3.5031,
"step": 8400
},
{
"epoch": 2.27400947459087,
"grad_norm": 0.39607518911361694,
"learning_rate": 0.0005662757506395583,
"loss": 3.5061,
"step": 8450
},
{
"epoch": 2.287467700258398,
"grad_norm": 0.3551819920539856,
"learning_rate": 0.0005660737848391006,
"loss": 3.5097,
"step": 8500
},
{
"epoch": 2.300925925925926,
"grad_norm": 0.38442036509513855,
"learning_rate": 0.0005658718190386428,
"loss": 3.5027,
"step": 8550
},
{
"epoch": 2.314384151593454,
"grad_norm": 0.37170907855033875,
"learning_rate": 0.000565669853238185,
"loss": 3.5,
"step": 8600
},
{
"epoch": 2.327842377260982,
"grad_norm": 0.3632299602031708,
"learning_rate": 0.0005654678874377272,
"loss": 3.4993,
"step": 8650
},
{
"epoch": 2.3413006029285097,
"grad_norm": 0.3846384286880493,
"learning_rate": 0.0005652659216372693,
"loss": 3.5076,
"step": 8700
},
{
"epoch": 2.354758828596038,
"grad_norm": 0.3948732614517212,
"learning_rate": 0.0005650639558368116,
"loss": 3.5045,
"step": 8750
},
{
"epoch": 2.3682170542635657,
"grad_norm": 0.37992361187934875,
"learning_rate": 0.0005648619900363538,
"loss": 3.5031,
"step": 8800
},
{
"epoch": 2.381675279931094,
"grad_norm": 0.3956155776977539,
"learning_rate": 0.000564660024235896,
"loss": 3.497,
"step": 8850
},
{
"epoch": 2.395133505598622,
"grad_norm": 0.35251104831695557,
"learning_rate": 0.0005644580584354382,
"loss": 3.4921,
"step": 8900
},
{
"epoch": 2.40859173126615,
"grad_norm": 0.3741273283958435,
"learning_rate": 0.0005642560926349804,
"loss": 3.4909,
"step": 8950
},
{
"epoch": 2.422049956933678,
"grad_norm": 0.36765018105506897,
"learning_rate": 0.0005640541268345226,
"loss": 3.5021,
"step": 9000
},
{
"epoch": 2.422049956933678,
"eval_accuracy": 0.3669376971125577,
"eval_loss": 3.5090339183807373,
"eval_runtime": 184.4548,
"eval_samples_per_second": 97.65,
"eval_steps_per_second": 6.104,
"step": 9000
},
{
"epoch": 2.4355081826012057,
"grad_norm": 0.350831538438797,
"learning_rate": 0.0005638521610340649,
"loss": 3.4964,
"step": 9050
},
{
"epoch": 2.448966408268734,
"grad_norm": 0.34787774085998535,
"learning_rate": 0.000563650195233607,
"loss": 3.4902,
"step": 9100
},
{
"epoch": 2.4624246339362617,
"grad_norm": 0.3351658880710602,
"learning_rate": 0.0005634482294331493,
"loss": 3.4943,
"step": 9150
},
{
"epoch": 2.47588285960379,
"grad_norm": 0.358770489692688,
"learning_rate": 0.0005632462636326914,
"loss": 3.4964,
"step": 9200
},
{
"epoch": 2.4893410852713176,
"grad_norm": 0.35480979084968567,
"learning_rate": 0.0005630442978322337,
"loss": 3.4875,
"step": 9250
},
{
"epoch": 2.502799310938846,
"grad_norm": 0.36890196800231934,
"learning_rate": 0.0005628423320317759,
"loss": 3.491,
"step": 9300
},
{
"epoch": 2.516257536606374,
"grad_norm": 0.365732342004776,
"learning_rate": 0.0005626403662313182,
"loss": 3.4825,
"step": 9350
},
{
"epoch": 2.5297157622739017,
"grad_norm": 0.45305827260017395,
"learning_rate": 0.0005624384004308603,
"loss": 3.4911,
"step": 9400
},
{
"epoch": 2.54317398794143,
"grad_norm": 0.38888585567474365,
"learning_rate": 0.0005622364346304026,
"loss": 3.4886,
"step": 9450
},
{
"epoch": 2.5566322136089576,
"grad_norm": 0.360650897026062,
"learning_rate": 0.0005620344688299448,
"loss": 3.5015,
"step": 9500
},
{
"epoch": 2.570090439276486,
"grad_norm": 0.3577306866645813,
"learning_rate": 0.0005618325030294869,
"loss": 3.4923,
"step": 9550
},
{
"epoch": 2.583548664944014,
"grad_norm": 0.3855704665184021,
"learning_rate": 0.0005616305372290292,
"loss": 3.4905,
"step": 9600
},
{
"epoch": 2.5970068906115418,
"grad_norm": 0.36281871795654297,
"learning_rate": 0.0005614285714285713,
"loss": 3.4984,
"step": 9650
},
{
"epoch": 2.6104651162790695,
"grad_norm": 0.374353289604187,
"learning_rate": 0.0005612266056281136,
"loss": 3.4994,
"step": 9700
},
{
"epoch": 2.6239233419465977,
"grad_norm": 0.3648017644882202,
"learning_rate": 0.0005610246398276558,
"loss": 3.4749,
"step": 9750
},
{
"epoch": 2.637381567614126,
"grad_norm": 0.36185312271118164,
"learning_rate": 0.000560822674027198,
"loss": 3.4736,
"step": 9800
},
{
"epoch": 2.6508397932816536,
"grad_norm": 0.36867278814315796,
"learning_rate": 0.0005606207082267402,
"loss": 3.4916,
"step": 9850
},
{
"epoch": 2.664298018949182,
"grad_norm": 0.373800665140152,
"learning_rate": 0.0005604187424262824,
"loss": 3.4803,
"step": 9900
},
{
"epoch": 2.6777562446167096,
"grad_norm": 0.33762213587760925,
"learning_rate": 0.0005602167766258246,
"loss": 3.4809,
"step": 9950
},
{
"epoch": 2.6912144702842378,
"grad_norm": 0.3567734956741333,
"learning_rate": 0.0005600148108253669,
"loss": 3.4823,
"step": 10000
},
{
"epoch": 2.6912144702842378,
"eval_accuracy": 0.3692814270807937,
"eval_loss": 3.482372522354126,
"eval_runtime": 184.6865,
"eval_samples_per_second": 97.527,
"eval_steps_per_second": 6.097,
"step": 10000
},
{
"epoch": 2.704672695951766,
"grad_norm": 0.41365304589271545,
"learning_rate": 0.000559812845024909,
"loss": 3.4815,
"step": 10050
},
{
"epoch": 2.7181309216192937,
"grad_norm": 0.3500403165817261,
"learning_rate": 0.0005596108792244513,
"loss": 3.484,
"step": 10100
},
{
"epoch": 2.7315891472868215,
"grad_norm": 0.340261310338974,
"learning_rate": 0.0005594089134239935,
"loss": 3.4844,
"step": 10150
},
{
"epoch": 2.7450473729543496,
"grad_norm": 0.3168796896934509,
"learning_rate": 0.0005592069476235358,
"loss": 3.4746,
"step": 10200
},
{
"epoch": 2.758505598621878,
"grad_norm": 0.3759358823299408,
"learning_rate": 0.0005590049818230779,
"loss": 3.4831,
"step": 10250
},
{
"epoch": 2.7719638242894056,
"grad_norm": 0.3580891788005829,
"learning_rate": 0.0005588030160226202,
"loss": 3.4733,
"step": 10300
},
{
"epoch": 2.7854220499569338,
"grad_norm": 0.3765806555747986,
"learning_rate": 0.0005586010502221623,
"loss": 3.484,
"step": 10350
},
{
"epoch": 2.7988802756244615,
"grad_norm": 0.36185717582702637,
"learning_rate": 0.0005583990844217045,
"loss": 3.476,
"step": 10400
},
{
"epoch": 2.8123385012919897,
"grad_norm": 0.34202250838279724,
"learning_rate": 0.0005581971186212468,
"loss": 3.4767,
"step": 10450
},
{
"epoch": 2.825796726959518,
"grad_norm": 0.36029788851737976,
"learning_rate": 0.0005579951528207889,
"loss": 3.4784,
"step": 10500
},
{
"epoch": 2.8392549526270456,
"grad_norm": 0.3801499903202057,
"learning_rate": 0.0005577931870203312,
"loss": 3.4652,
"step": 10550
},
{
"epoch": 2.8527131782945734,
"grad_norm": 0.3782118558883667,
"learning_rate": 0.0005575912212198733,
"loss": 3.4861,
"step": 10600
},
{
"epoch": 2.8661714039621016,
"grad_norm": 0.3364644944667816,
"learning_rate": 0.0005573892554194156,
"loss": 3.4768,
"step": 10650
},
{
"epoch": 2.8796296296296298,
"grad_norm": 0.3544559180736542,
"learning_rate": 0.0005571872896189578,
"loss": 3.4771,
"step": 10700
},
{
"epoch": 2.8930878552971575,
"grad_norm": 0.35688602924346924,
"learning_rate": 0.0005569853238185,
"loss": 3.4643,
"step": 10750
},
{
"epoch": 2.9065460809646857,
"grad_norm": 0.35698455572128296,
"learning_rate": 0.0005567833580180422,
"loss": 3.474,
"step": 10800
},
{
"epoch": 2.9200043066322134,
"grad_norm": 0.36227646470069885,
"learning_rate": 0.0005565813922175844,
"loss": 3.4596,
"step": 10850
},
{
"epoch": 2.9334625322997416,
"grad_norm": 0.3645004332065582,
"learning_rate": 0.0005563794264171266,
"loss": 3.4718,
"step": 10900
},
{
"epoch": 2.94692075796727,
"grad_norm": 0.40124833583831787,
"learning_rate": 0.0005561774606166689,
"loss": 3.4653,
"step": 10950
},
{
"epoch": 2.9603789836347976,
"grad_norm": 0.3600238561630249,
"learning_rate": 0.000555975494816211,
"loss": 3.4569,
"step": 11000
},
{
"epoch": 2.9603789836347976,
"eval_accuracy": 0.37177976152449793,
"eval_loss": 3.4617209434509277,
"eval_runtime": 183.9456,
"eval_samples_per_second": 97.92,
"eval_steps_per_second": 6.121,
"step": 11000
},
{
"epoch": 2.9738372093023253,
"grad_norm": 0.3510929346084595,
"learning_rate": 0.0005557735290157534,
"loss": 3.4682,
"step": 11050
},
{
"epoch": 2.9872954349698535,
"grad_norm": 0.3392506241798401,
"learning_rate": 0.0005555715632152955,
"loss": 3.4766,
"step": 11100
},
{
"epoch": 3.000538329026701,
"grad_norm": 0.37191227078437805,
"learning_rate": 0.0005553695974148378,
"loss": 3.4687,
"step": 11150
},
{
"epoch": 3.013996554694229,
"grad_norm": 0.3616611957550049,
"learning_rate": 0.0005551676316143799,
"loss": 3.3737,
"step": 11200
},
{
"epoch": 3.027454780361757,
"grad_norm": 0.3481079041957855,
"learning_rate": 0.0005549656658139222,
"loss": 3.3667,
"step": 11250
},
{
"epoch": 3.0409130060292853,
"grad_norm": 0.39582017064094543,
"learning_rate": 0.0005547637000134644,
"loss": 3.385,
"step": 11300
},
{
"epoch": 3.054371231696813,
"grad_norm": 0.36397379636764526,
"learning_rate": 0.0005545617342130065,
"loss": 3.3817,
"step": 11350
},
{
"epoch": 3.067829457364341,
"grad_norm": 0.3545301556587219,
"learning_rate": 0.0005543597684125488,
"loss": 3.3771,
"step": 11400
},
{
"epoch": 3.081287683031869,
"grad_norm": 0.3880403935909271,
"learning_rate": 0.0005541578026120909,
"loss": 3.3727,
"step": 11450
},
{
"epoch": 3.094745908699397,
"grad_norm": 0.40306705236434937,
"learning_rate": 0.0005539558368116332,
"loss": 3.3931,
"step": 11500
},
{
"epoch": 3.108204134366925,
"grad_norm": 0.34913113713264465,
"learning_rate": 0.0005537538710111754,
"loss": 3.3755,
"step": 11550
},
{
"epoch": 3.121662360034453,
"grad_norm": 0.34303605556488037,
"learning_rate": 0.0005535519052107176,
"loss": 3.3921,
"step": 11600
},
{
"epoch": 3.135120585701981,
"grad_norm": 0.33962705731391907,
"learning_rate": 0.0005533499394102598,
"loss": 3.3733,
"step": 11650
},
{
"epoch": 3.148578811369509,
"grad_norm": 0.3546980917453766,
"learning_rate": 0.000553147973609802,
"loss": 3.3936,
"step": 11700
},
{
"epoch": 3.162037037037037,
"grad_norm": 0.35574012994766235,
"learning_rate": 0.0005529460078093442,
"loss": 3.3943,
"step": 11750
},
{
"epoch": 3.175495262704565,
"grad_norm": 0.3347347676753998,
"learning_rate": 0.0005527440420088865,
"loss": 3.3944,
"step": 11800
},
{
"epoch": 3.188953488372093,
"grad_norm": 0.3656306564807892,
"learning_rate": 0.0005525420762084286,
"loss": 3.388,
"step": 11850
},
{
"epoch": 3.202411714039621,
"grad_norm": 0.37094923853874207,
"learning_rate": 0.0005523401104079709,
"loss": 3.3907,
"step": 11900
},
{
"epoch": 3.215869939707149,
"grad_norm": 0.35495519638061523,
"learning_rate": 0.0005521381446075131,
"loss": 3.392,
"step": 11950
},
{
"epoch": 3.229328165374677,
"grad_norm": 0.322475403547287,
"learning_rate": 0.0005519361788070554,
"loss": 3.3867,
"step": 12000
},
{
"epoch": 3.229328165374677,
"eval_accuracy": 0.3730108390448985,
"eval_loss": 3.4523074626922607,
"eval_runtime": 184.3276,
"eval_samples_per_second": 97.717,
"eval_steps_per_second": 6.109,
"step": 12000
},
{
"epoch": 3.242786391042205,
"grad_norm": 0.3748001456260681,
"learning_rate": 0.0005517342130065975,
"loss": 3.4042,
"step": 12050
},
{
"epoch": 3.2562446167097328,
"grad_norm": 0.4192953407764435,
"learning_rate": 0.0005515322472061398,
"loss": 3.4013,
"step": 12100
},
{
"epoch": 3.269702842377261,
"grad_norm": 0.3523455560207367,
"learning_rate": 0.0005513302814056819,
"loss": 3.3801,
"step": 12150
},
{
"epoch": 3.283161068044789,
"grad_norm": 0.34908753633499146,
"learning_rate": 0.0005511283156052242,
"loss": 3.3911,
"step": 12200
},
{
"epoch": 3.296619293712317,
"grad_norm": 0.34149667620658875,
"learning_rate": 0.0005509263498047664,
"loss": 3.399,
"step": 12250
},
{
"epoch": 3.310077519379845,
"grad_norm": 0.3445163071155548,
"learning_rate": 0.0005507243840043085,
"loss": 3.3944,
"step": 12300
},
{
"epoch": 3.323535745047373,
"grad_norm": 0.3592802882194519,
"learning_rate": 0.0005505224182038508,
"loss": 3.3998,
"step": 12350
},
{
"epoch": 3.336993970714901,
"grad_norm": 0.3553261458873749,
"learning_rate": 0.0005503204524033929,
"loss": 3.4035,
"step": 12400
},
{
"epoch": 3.3504521963824287,
"grad_norm": 0.40120527148246765,
"learning_rate": 0.0005501184866029352,
"loss": 3.3995,
"step": 12450
},
{
"epoch": 3.363910422049957,
"grad_norm": 0.3470523953437805,
"learning_rate": 0.0005499165208024774,
"loss": 3.3988,
"step": 12500
},
{
"epoch": 3.3773686477174847,
"grad_norm": 0.3485461473464966,
"learning_rate": 0.0005497145550020196,
"loss": 3.3977,
"step": 12550
},
{
"epoch": 3.390826873385013,
"grad_norm": 0.3667030930519104,
"learning_rate": 0.0005495125892015618,
"loss": 3.4014,
"step": 12600
},
{
"epoch": 3.404285099052541,
"grad_norm": 0.35710710287094116,
"learning_rate": 0.000549310623401104,
"loss": 3.4017,
"step": 12650
},
{
"epoch": 3.417743324720069,
"grad_norm": 0.3693983256816864,
"learning_rate": 0.0005491086576006462,
"loss": 3.3934,
"step": 12700
},
{
"epoch": 3.431201550387597,
"grad_norm": 0.36102887988090515,
"learning_rate": 0.0005489066918001885,
"loss": 3.4037,
"step": 12750
},
{
"epoch": 3.4446597760551247,
"grad_norm": 0.38098475337028503,
"learning_rate": 0.0005487047259997306,
"loss": 3.416,
"step": 12800
},
{
"epoch": 3.458118001722653,
"grad_norm": 0.33953848481178284,
"learning_rate": 0.0005485027601992729,
"loss": 3.4006,
"step": 12850
},
{
"epoch": 3.471576227390181,
"grad_norm": 0.37008365988731384,
"learning_rate": 0.0005483007943988151,
"loss": 3.3858,
"step": 12900
},
{
"epoch": 3.485034453057709,
"grad_norm": 0.3219403028488159,
"learning_rate": 0.0005480988285983574,
"loss": 3.3811,
"step": 12950
},
{
"epoch": 3.4984926787252366,
"grad_norm": 0.33084872364997864,
"learning_rate": 0.0005478968627978995,
"loss": 3.4035,
"step": 13000
},
{
"epoch": 3.4984926787252366,
"eval_accuracy": 0.3747549470172744,
"eval_loss": 3.4349913597106934,
"eval_runtime": 184.9263,
"eval_samples_per_second": 97.401,
"eval_steps_per_second": 6.089,
"step": 13000
},
{
"epoch": 3.511950904392765,
"grad_norm": 0.38038837909698486,
"learning_rate": 0.0005476948969974418,
"loss": 3.394,
"step": 13050
},
{
"epoch": 3.525409130060293,
"grad_norm": 0.357761949300766,
"learning_rate": 0.0005474929311969839,
"loss": 3.4,
"step": 13100
},
{
"epoch": 3.5388673557278207,
"grad_norm": 0.3668404817581177,
"learning_rate": 0.0005472909653965261,
"loss": 3.4033,
"step": 13150
},
{
"epoch": 3.552325581395349,
"grad_norm": 0.3330177366733551,
"learning_rate": 0.0005470889995960684,
"loss": 3.3979,
"step": 13200
},
{
"epoch": 3.5657838070628767,
"grad_norm": 0.33386555314064026,
"learning_rate": 0.0005468870337956105,
"loss": 3.3952,
"step": 13250
},
{
"epoch": 3.579242032730405,
"grad_norm": 0.349822998046875,
"learning_rate": 0.0005466850679951528,
"loss": 3.4001,
"step": 13300
},
{
"epoch": 3.592700258397933,
"grad_norm": 0.3559633195400238,
"learning_rate": 0.0005464831021946949,
"loss": 3.3951,
"step": 13350
},
{
"epoch": 3.606158484065461,
"grad_norm": 0.3450447618961334,
"learning_rate": 0.0005462811363942372,
"loss": 3.3942,
"step": 13400
},
{
"epoch": 3.6196167097329885,
"grad_norm": 0.36225467920303345,
"learning_rate": 0.0005460791705937794,
"loss": 3.3874,
"step": 13450
},
{
"epoch": 3.6330749354005167,
"grad_norm": 0.3331061899662018,
"learning_rate": 0.0005458772047933216,
"loss": 3.3931,
"step": 13500
},
{
"epoch": 3.646533161068045,
"grad_norm": 0.34889641404151917,
"learning_rate": 0.0005456752389928638,
"loss": 3.4022,
"step": 13550
},
{
"epoch": 3.6599913867355727,
"grad_norm": 0.33603474497795105,
"learning_rate": 0.000545473273192406,
"loss": 3.4002,
"step": 13600
},
{
"epoch": 3.673449612403101,
"grad_norm": 0.3359422981739044,
"learning_rate": 0.0005452713073919482,
"loss": 3.4101,
"step": 13650
},
{
"epoch": 3.6869078380706286,
"grad_norm": 0.36187952756881714,
"learning_rate": 0.0005450693415914905,
"loss": 3.3975,
"step": 13700
},
{
"epoch": 3.700366063738157,
"grad_norm": 0.3648735582828522,
"learning_rate": 0.0005448673757910327,
"loss": 3.3876,
"step": 13750
},
{
"epoch": 3.713824289405685,
"grad_norm": 0.35344839096069336,
"learning_rate": 0.0005446654099905749,
"loss": 3.3996,
"step": 13800
},
{
"epoch": 3.7272825150732127,
"grad_norm": 0.3489483892917633,
"learning_rate": 0.0005444634441901171,
"loss": 3.3892,
"step": 13850
},
{
"epoch": 3.7407407407407405,
"grad_norm": 0.32150471210479736,
"learning_rate": 0.0005442614783896594,
"loss": 3.3913,
"step": 13900
},
{
"epoch": 3.7541989664082687,
"grad_norm": 0.35157185792922974,
"learning_rate": 0.0005440595125892015,
"loss": 3.4083,
"step": 13950
},
{
"epoch": 3.767657192075797,
"grad_norm": 0.35811540484428406,
"learning_rate": 0.0005438575467887438,
"loss": 3.4025,
"step": 14000
},
{
"epoch": 3.767657192075797,
"eval_accuracy": 0.3760738111969711,
"eval_loss": 3.4210078716278076,
"eval_runtime": 185.6104,
"eval_samples_per_second": 97.042,
"eval_steps_per_second": 6.066,
"step": 14000
},
{
"epoch": 3.7811154177433246,
"grad_norm": 0.36421844363212585,
"learning_rate": 0.0005436555809882859,
"loss": 3.4043,
"step": 14050
},
{
"epoch": 3.794573643410853,
"grad_norm": 0.3407555818557739,
"learning_rate": 0.0005434536151878281,
"loss": 3.3928,
"step": 14100
},
{
"epoch": 3.8080318690783805,
"grad_norm": 0.3166675865650177,
"learning_rate": 0.0005432516493873704,
"loss": 3.3838,
"step": 14150
},
{
"epoch": 3.8214900947459087,
"grad_norm": 0.3488039970397949,
"learning_rate": 0.0005430496835869125,
"loss": 3.3983,
"step": 14200
},
{
"epoch": 3.834948320413437,
"grad_norm": 0.3746275305747986,
"learning_rate": 0.0005428477177864548,
"loss": 3.3853,
"step": 14250
},
{
"epoch": 3.8484065460809647,
"grad_norm": 0.33796748518943787,
"learning_rate": 0.0005426457519859969,
"loss": 3.3917,
"step": 14300
},
{
"epoch": 3.8618647717484924,
"grad_norm": 0.3440247178077698,
"learning_rate": 0.0005424437861855392,
"loss": 3.394,
"step": 14350
},
{
"epoch": 3.8753229974160206,
"grad_norm": 0.3565462827682495,
"learning_rate": 0.0005422418203850814,
"loss": 3.4043,
"step": 14400
},
{
"epoch": 3.888781223083549,
"grad_norm": 0.349168598651886,
"learning_rate": 0.0005420398545846236,
"loss": 3.4037,
"step": 14450
},
{
"epoch": 3.9022394487510765,
"grad_norm": 0.3481910526752472,
"learning_rate": 0.0005418378887841658,
"loss": 3.3835,
"step": 14500
},
{
"epoch": 3.9156976744186047,
"grad_norm": 0.3512874245643616,
"learning_rate": 0.000541635922983708,
"loss": 3.3848,
"step": 14550
},
{
"epoch": 3.9291559000861325,
"grad_norm": 0.35648831725120544,
"learning_rate": 0.0005414339571832503,
"loss": 3.3943,
"step": 14600
},
{
"epoch": 3.9426141257536607,
"grad_norm": 0.3414505422115326,
"learning_rate": 0.0005412319913827925,
"loss": 3.4044,
"step": 14650
},
{
"epoch": 3.956072351421189,
"grad_norm": 0.35054147243499756,
"learning_rate": 0.0005410300255823347,
"loss": 3.4052,
"step": 14700
},
{
"epoch": 3.9695305770887166,
"grad_norm": 0.32000890374183655,
"learning_rate": 0.0005408280597818769,
"loss": 3.389,
"step": 14750
},
{
"epoch": 3.9829888027562443,
"grad_norm": 0.3527786433696747,
"learning_rate": 0.0005406260939814191,
"loss": 3.3926,
"step": 14800
},
{
"epoch": 3.9964470284237725,
"grad_norm": 0.317381888628006,
"learning_rate": 0.0005404241281809614,
"loss": 3.3859,
"step": 14850
},
{
"epoch": 4.00968992248062,
"grad_norm": 0.3511665165424347,
"learning_rate": 0.0005402221623805035,
"loss": 3.324,
"step": 14900
},
{
"epoch": 4.023148148148148,
"grad_norm": 0.3269699513912201,
"learning_rate": 0.0005400201965800457,
"loss": 3.3004,
"step": 14950
},
{
"epoch": 4.036606373815676,
"grad_norm": 0.32972702383995056,
"learning_rate": 0.000539818230779588,
"loss": 3.3059,
"step": 15000
},
{
"epoch": 4.036606373815676,
"eval_accuracy": 0.3776203991859308,
"eval_loss": 3.411360740661621,
"eval_runtime": 185.0413,
"eval_samples_per_second": 97.34,
"eval_steps_per_second": 6.085,
"step": 15000
},
{
"epoch": 4.050064599483204,
"grad_norm": 0.35312792658805847,
"learning_rate": 0.0005396162649791301,
"loss": 3.2921,
"step": 15050
},
{
"epoch": 4.0635228251507325,
"grad_norm": 0.3108118176460266,
"learning_rate": 0.0005394142991786724,
"loss": 3.299,
"step": 15100
},
{
"epoch": 4.07698105081826,
"grad_norm": 0.3434188961982727,
"learning_rate": 0.0005392123333782145,
"loss": 3.3026,
"step": 15150
},
{
"epoch": 4.090439276485788,
"grad_norm": 0.36485904455184937,
"learning_rate": 0.0005390103675777568,
"loss": 3.3008,
"step": 15200
},
{
"epoch": 4.103897502153316,
"grad_norm": 0.32428795099258423,
"learning_rate": 0.000538808401777299,
"loss": 3.3073,
"step": 15250
},
{
"epoch": 4.117355727820844,
"grad_norm": 0.3441103398799896,
"learning_rate": 0.0005386064359768412,
"loss": 3.3166,
"step": 15300
},
{
"epoch": 4.1308139534883725,
"grad_norm": 0.3301752507686615,
"learning_rate": 0.0005384044701763834,
"loss": 3.3143,
"step": 15350
},
{
"epoch": 4.1442721791559,
"grad_norm": 0.3428856134414673,
"learning_rate": 0.0005382025043759256,
"loss": 3.3188,
"step": 15400
},
{
"epoch": 4.157730404823428,
"grad_norm": 0.3465999364852905,
"learning_rate": 0.0005380005385754678,
"loss": 3.3004,
"step": 15450
},
{
"epoch": 4.171188630490956,
"grad_norm": 0.35543176531791687,
"learning_rate": 0.0005377985727750101,
"loss": 3.3301,
"step": 15500
},
{
"epoch": 4.184646856158484,
"grad_norm": 0.32972484827041626,
"learning_rate": 0.0005375966069745523,
"loss": 3.3249,
"step": 15550
},
{
"epoch": 4.198105081826012,
"grad_norm": 0.34662866592407227,
"learning_rate": 0.0005373946411740945,
"loss": 3.3283,
"step": 15600
},
{
"epoch": 4.21156330749354,
"grad_norm": 0.3478318750858307,
"learning_rate": 0.0005371926753736367,
"loss": 3.3261,
"step": 15650
},
{
"epoch": 4.225021533161068,
"grad_norm": 0.36225807666778564,
"learning_rate": 0.000536990709573179,
"loss": 3.3188,
"step": 15700
},
{
"epoch": 4.238479758828596,
"grad_norm": 0.33634069561958313,
"learning_rate": 0.0005367887437727211,
"loss": 3.3267,
"step": 15750
},
{
"epoch": 4.251937984496124,
"grad_norm": 0.36333349347114563,
"learning_rate": 0.0005365867779722634,
"loss": 3.3165,
"step": 15800
},
{
"epoch": 4.265396210163652,
"grad_norm": 0.333570271730423,
"learning_rate": 0.0005363848121718055,
"loss": 3.3293,
"step": 15850
},
{
"epoch": 4.27885443583118,
"grad_norm": 0.34312504529953003,
"learning_rate": 0.0005361828463713477,
"loss": 3.3186,
"step": 15900
},
{
"epoch": 4.292312661498708,
"grad_norm": 0.3460608720779419,
"learning_rate": 0.00053598088057089,
"loss": 3.3305,
"step": 15950
},
{
"epoch": 4.305770887166236,
"grad_norm": 0.3470321297645569,
"learning_rate": 0.0005357789147704321,
"loss": 3.3324,
"step": 16000
},
{
"epoch": 4.305770887166236,
"eval_accuracy": 0.3784762104672119,
"eval_loss": 3.404461622238159,
"eval_runtime": 184.8832,
"eval_samples_per_second": 97.424,
"eval_steps_per_second": 6.09,
"step": 16000
},
{
"epoch": 4.319229112833764,
"grad_norm": 0.3456474840641022,
"learning_rate": 0.0005355769489699744,
"loss": 3.3251,
"step": 16050
},
{
"epoch": 4.332687338501292,
"grad_norm": 0.3597697615623474,
"learning_rate": 0.0005353749831695165,
"loss": 3.3326,
"step": 16100
},
{
"epoch": 4.34614556416882,
"grad_norm": 0.3345887362957001,
"learning_rate": 0.0005351730173690588,
"loss": 3.3342,
"step": 16150
},
{
"epoch": 4.359603789836348,
"grad_norm": 0.3175329267978668,
"learning_rate": 0.000534971051568601,
"loss": 3.3307,
"step": 16200
},
{
"epoch": 4.373062015503876,
"grad_norm": 0.33834436535835266,
"learning_rate": 0.0005347690857681432,
"loss": 3.3305,
"step": 16250
},
{
"epoch": 4.386520241171404,
"grad_norm": 0.36735835671424866,
"learning_rate": 0.0005345671199676854,
"loss": 3.3305,
"step": 16300
},
{
"epoch": 4.399978466838932,
"grad_norm": 0.35801461338996887,
"learning_rate": 0.0005343651541672276,
"loss": 3.3335,
"step": 16350
},
{
"epoch": 4.41343669250646,
"grad_norm": 0.32037919759750366,
"learning_rate": 0.0005341631883667699,
"loss": 3.335,
"step": 16400
},
{
"epoch": 4.426894918173988,
"grad_norm": 0.3480973541736603,
"learning_rate": 0.0005339612225663121,
"loss": 3.3349,
"step": 16450
},
{
"epoch": 4.440353143841516,
"grad_norm": 0.37662386894226074,
"learning_rate": 0.0005337592567658543,
"loss": 3.3379,
"step": 16500
},
{
"epoch": 4.453811369509044,
"grad_norm": 0.33138397336006165,
"learning_rate": 0.0005335572909653965,
"loss": 3.3306,
"step": 16550
},
{
"epoch": 4.467269595176572,
"grad_norm": 0.3491424322128296,
"learning_rate": 0.0005333553251649387,
"loss": 3.3316,
"step": 16600
},
{
"epoch": 4.4807278208441,
"grad_norm": 0.3313773274421692,
"learning_rate": 0.000533153359364481,
"loss": 3.3288,
"step": 16650
},
{
"epoch": 4.4941860465116275,
"grad_norm": 0.3446807265281677,
"learning_rate": 0.0005329513935640231,
"loss": 3.3302,
"step": 16700
},
{
"epoch": 4.507644272179156,
"grad_norm": 0.32654106616973877,
"learning_rate": 0.0005327494277635654,
"loss": 3.3379,
"step": 16750
},
{
"epoch": 4.521102497846684,
"grad_norm": 0.34293171763420105,
"learning_rate": 0.0005325474619631075,
"loss": 3.3308,
"step": 16800
},
{
"epoch": 4.534560723514212,
"grad_norm": 0.33090731501579285,
"learning_rate": 0.0005323454961626497,
"loss": 3.3406,
"step": 16850
},
{
"epoch": 4.54801894918174,
"grad_norm": 0.3575376272201538,
"learning_rate": 0.000532143530362192,
"loss": 3.3398,
"step": 16900
},
{
"epoch": 4.5614771748492675,
"grad_norm": 0.33871203660964966,
"learning_rate": 0.0005319415645617341,
"loss": 3.3326,
"step": 16950
},
{
"epoch": 4.574935400516796,
"grad_norm": 0.36322900652885437,
"learning_rate": 0.0005317395987612764,
"loss": 3.3339,
"step": 17000
},
{
"epoch": 4.574935400516796,
"eval_accuracy": 0.3797959438217531,
"eval_loss": 3.3930764198303223,
"eval_runtime": 184.9905,
"eval_samples_per_second": 97.367,
"eval_steps_per_second": 6.087,
"step": 17000
},
{
"epoch": 4.588393626184324,
"grad_norm": 0.34915104508399963,
"learning_rate": 0.0005315376329608185,
"loss": 3.3277,
"step": 17050
},
{
"epoch": 4.601851851851852,
"grad_norm": 0.346442312002182,
"learning_rate": 0.0005313356671603608,
"loss": 3.3315,
"step": 17100
},
{
"epoch": 4.61531007751938,
"grad_norm": 0.342703253030777,
"learning_rate": 0.000531133701359903,
"loss": 3.3384,
"step": 17150
},
{
"epoch": 4.628768303186908,
"grad_norm": 0.36335626244544983,
"learning_rate": 0.0005309317355594452,
"loss": 3.3264,
"step": 17200
},
{
"epoch": 4.642226528854436,
"grad_norm": 0.3631543517112732,
"learning_rate": 0.0005307297697589874,
"loss": 3.3375,
"step": 17250
},
{
"epoch": 4.655684754521964,
"grad_norm": 0.33104273676872253,
"learning_rate": 0.0005305278039585297,
"loss": 3.3387,
"step": 17300
},
{
"epoch": 4.669142980189492,
"grad_norm": 0.33900973200798035,
"learning_rate": 0.0005303258381580719,
"loss": 3.3404,
"step": 17350
},
{
"epoch": 4.682601205857019,
"grad_norm": 0.3157927989959717,
"learning_rate": 0.0005301238723576141,
"loss": 3.3416,
"step": 17400
},
{
"epoch": 4.696059431524548,
"grad_norm": 0.3543723523616791,
"learning_rate": 0.0005299219065571563,
"loss": 3.3305,
"step": 17450
},
{
"epoch": 4.709517657192076,
"grad_norm": 0.32342445850372314,
"learning_rate": 0.0005297199407566985,
"loss": 3.3366,
"step": 17500
},
{
"epoch": 4.722975882859604,
"grad_norm": 0.33042553067207336,
"learning_rate": 0.0005295179749562407,
"loss": 3.3386,
"step": 17550
},
{
"epoch": 4.736434108527131,
"grad_norm": 0.3297114074230194,
"learning_rate": 0.000529316009155783,
"loss": 3.3348,
"step": 17600
},
{
"epoch": 4.7498923341946595,
"grad_norm": 0.33898666501045227,
"learning_rate": 0.0005291140433553251,
"loss": 3.3298,
"step": 17650
},
{
"epoch": 4.763350559862188,
"grad_norm": 0.3274131417274475,
"learning_rate": 0.0005289120775548673,
"loss": 3.3439,
"step": 17700
},
{
"epoch": 4.776808785529716,
"grad_norm": 0.3546232581138611,
"learning_rate": 0.0005287101117544095,
"loss": 3.3372,
"step": 17750
},
{
"epoch": 4.790267011197244,
"grad_norm": 0.3214088976383209,
"learning_rate": 0.0005285081459539517,
"loss": 3.3273,
"step": 17800
},
{
"epoch": 4.803725236864771,
"grad_norm": 0.33602869510650635,
"learning_rate": 0.000528306180153494,
"loss": 3.3263,
"step": 17850
},
{
"epoch": 4.8171834625323,
"grad_norm": 0.34128642082214355,
"learning_rate": 0.0005281042143530361,
"loss": 3.3391,
"step": 17900
},
{
"epoch": 4.830641688199828,
"grad_norm": 0.3308497965335846,
"learning_rate": 0.0005279022485525784,
"loss": 3.3376,
"step": 17950
},
{
"epoch": 4.844099913867356,
"grad_norm": 0.3582690358161926,
"learning_rate": 0.0005277002827521205,
"loss": 3.3349,
"step": 18000
},
{
"epoch": 4.844099913867356,
"eval_accuracy": 0.3811932510311673,
"eval_loss": 3.3793694972991943,
"eval_runtime": 185.0814,
"eval_samples_per_second": 97.319,
"eval_steps_per_second": 6.084,
"step": 18000
},
{
"epoch": 4.857558139534884,
"grad_norm": 0.3303445279598236,
"learning_rate": 0.0005274983169516628,
"loss": 3.3383,
"step": 18050
},
{
"epoch": 4.871016365202411,
"grad_norm": 0.3253082036972046,
"learning_rate": 0.000527296351151205,
"loss": 3.3224,
"step": 18100
},
{
"epoch": 4.88447459086994,
"grad_norm": 0.34937795996665955,
"learning_rate": 0.0005270943853507472,
"loss": 3.3298,
"step": 18150
},
{
"epoch": 4.897932816537468,
"grad_norm": 0.32902005314826965,
"learning_rate": 0.0005268924195502894,
"loss": 3.3471,
"step": 18200
},
{
"epoch": 4.911391042204996,
"grad_norm": 0.333818644285202,
"learning_rate": 0.0005266904537498317,
"loss": 3.3296,
"step": 18250
},
{
"epoch": 4.924849267872523,
"grad_norm": 0.3378560245037079,
"learning_rate": 0.0005264884879493739,
"loss": 3.3393,
"step": 18300
},
{
"epoch": 4.9383074935400515,
"grad_norm": 0.33791688084602356,
"learning_rate": 0.0005262865221489161,
"loss": 3.337,
"step": 18350
},
{
"epoch": 4.95176571920758,
"grad_norm": 0.3007248640060425,
"learning_rate": 0.0005260845563484583,
"loss": 3.334,
"step": 18400
},
{
"epoch": 4.965223944875108,
"grad_norm": 0.33719944953918457,
"learning_rate": 0.0005258825905480005,
"loss": 3.3277,
"step": 18450
},
{
"epoch": 4.978682170542635,
"grad_norm": 0.32557839155197144,
"learning_rate": 0.0005256806247475427,
"loss": 3.3373,
"step": 18500
},
{
"epoch": 4.992140396210163,
"grad_norm": 0.33758941292762756,
"learning_rate": 0.000525478658947085,
"loss": 3.3396,
"step": 18550
},
{
"epoch": 5.0053832902670115,
"grad_norm": 0.3397212326526642,
"learning_rate": 0.0005252766931466271,
"loss": 3.2986,
"step": 18600
},
{
"epoch": 5.018841515934539,
"grad_norm": 0.35023489594459534,
"learning_rate": 0.0005250747273461693,
"loss": 3.2393,
"step": 18650
},
{
"epoch": 5.032299741602067,
"grad_norm": 0.3497323989868164,
"learning_rate": 0.0005248727615457115,
"loss": 3.2437,
"step": 18700
},
{
"epoch": 5.045757967269595,
"grad_norm": 0.32975849509239197,
"learning_rate": 0.0005246707957452537,
"loss": 3.251,
"step": 18750
},
{
"epoch": 5.059216192937123,
"grad_norm": 0.33672475814819336,
"learning_rate": 0.000524468829944796,
"loss": 3.2435,
"step": 18800
},
{
"epoch": 5.0726744186046515,
"grad_norm": 0.33457061648368835,
"learning_rate": 0.0005242668641443381,
"loss": 3.248,
"step": 18850
},
{
"epoch": 5.086132644272179,
"grad_norm": 0.33664408326148987,
"learning_rate": 0.0005240648983438804,
"loss": 3.2423,
"step": 18900
},
{
"epoch": 5.099590869939707,
"grad_norm": 0.32727932929992676,
"learning_rate": 0.0005238629325434225,
"loss": 3.2531,
"step": 18950
},
{
"epoch": 5.113049095607235,
"grad_norm": 0.3277779817581177,
"learning_rate": 0.0005236609667429648,
"loss": 3.2506,
"step": 19000
},
{
"epoch": 5.113049095607235,
"eval_accuracy": 0.3816402241949594,
"eval_loss": 3.3798582553863525,
"eval_runtime": 185.1863,
"eval_samples_per_second": 97.264,
"eval_steps_per_second": 6.08,
"step": 19000
},
{
"epoch": 5.126507321274763,
"grad_norm": 0.3330926299095154,
"learning_rate": 0.000523459000942507,
"loss": 3.2464,
"step": 19050
},
{
"epoch": 5.139965546942291,
"grad_norm": 0.36706265807151794,
"learning_rate": 0.0005232570351420493,
"loss": 3.27,
"step": 19100
},
{
"epoch": 5.153423772609819,
"grad_norm": 0.3373613655567169,
"learning_rate": 0.0005230550693415915,
"loss": 3.2607,
"step": 19150
},
{
"epoch": 5.166881998277347,
"grad_norm": 0.3493236005306244,
"learning_rate": 0.0005228531035411337,
"loss": 3.2714,
"step": 19200
},
{
"epoch": 5.180340223944875,
"grad_norm": 0.33827143907546997,
"learning_rate": 0.0005226511377406759,
"loss": 3.2675,
"step": 19250
},
{
"epoch": 5.1937984496124034,
"grad_norm": 0.3532494306564331,
"learning_rate": 0.0005224491719402181,
"loss": 3.2647,
"step": 19300
},
{
"epoch": 5.207256675279931,
"grad_norm": 0.3496478199958801,
"learning_rate": 0.0005222472061397603,
"loss": 3.269,
"step": 19350
},
{
"epoch": 5.220714900947459,
"grad_norm": 0.33100876212120056,
"learning_rate": 0.0005220452403393026,
"loss": 3.2773,
"step": 19400
},
{
"epoch": 5.234173126614987,
"grad_norm": 0.3529794216156006,
"learning_rate": 0.0005218432745388447,
"loss": 3.2684,
"step": 19450
},
{
"epoch": 5.247631352282515,
"grad_norm": 0.37019971013069153,
"learning_rate": 0.000521641308738387,
"loss": 3.2801,
"step": 19500
},
{
"epoch": 5.2610895779500435,
"grad_norm": 0.4020121693611145,
"learning_rate": 0.0005214393429379291,
"loss": 3.265,
"step": 19550
},
{
"epoch": 5.274547803617571,
"grad_norm": 0.34098562598228455,
"learning_rate": 0.0005212373771374713,
"loss": 3.2825,
"step": 19600
},
{
"epoch": 5.288006029285099,
"grad_norm": 0.3309837281703949,
"learning_rate": 0.0005210354113370136,
"loss": 3.2791,
"step": 19650
},
{
"epoch": 5.301464254952627,
"grad_norm": 0.3449201285839081,
"learning_rate": 0.0005208334455365557,
"loss": 3.2743,
"step": 19700
},
{
"epoch": 5.314922480620155,
"grad_norm": 0.3398899734020233,
"learning_rate": 0.000520631479736098,
"loss": 3.271,
"step": 19750
},
{
"epoch": 5.328380706287683,
"grad_norm": 0.3175956606864929,
"learning_rate": 0.0005204295139356401,
"loss": 3.2828,
"step": 19800
},
{
"epoch": 5.341838931955211,
"grad_norm": 0.33574339747428894,
"learning_rate": 0.0005202275481351824,
"loss": 3.2703,
"step": 19850
},
{
"epoch": 5.355297157622739,
"grad_norm": 0.3375524878501892,
"learning_rate": 0.0005200255823347246,
"loss": 3.2748,
"step": 19900
},
{
"epoch": 5.368755383290267,
"grad_norm": 0.3423841595649719,
"learning_rate": 0.0005198236165342669,
"loss": 3.2742,
"step": 19950
},
{
"epoch": 5.3822136089577945,
"grad_norm": 0.3087540566921234,
"learning_rate": 0.000519621650733809,
"loss": 3.2804,
"step": 20000
},
{
"epoch": 5.3822136089577945,
"eval_accuracy": 0.38213684897174444,
"eval_loss": 3.372101306915283,
"eval_runtime": 185.191,
"eval_samples_per_second": 97.262,
"eval_steps_per_second": 6.08,
"step": 20000
},
{
"epoch": 5.395671834625323,
"grad_norm": 0.32726848125457764,
"learning_rate": 0.0005194196849333513,
"loss": 3.2896,
"step": 20050
},
{
"epoch": 5.409130060292851,
"grad_norm": 0.33420437574386597,
"learning_rate": 0.0005192177191328935,
"loss": 3.2797,
"step": 20100
},
{
"epoch": 5.422588285960379,
"grad_norm": 0.3604329228401184,
"learning_rate": 0.0005190157533324357,
"loss": 3.2816,
"step": 20150
},
{
"epoch": 5.436046511627907,
"grad_norm": 0.3442542850971222,
"learning_rate": 0.0005188137875319779,
"loss": 3.2896,
"step": 20200
},
{
"epoch": 5.449504737295435,
"grad_norm": 0.32473087310791016,
"learning_rate": 0.00051861182173152,
"loss": 3.2847,
"step": 20250
},
{
"epoch": 5.462962962962963,
"grad_norm": 0.312188059091568,
"learning_rate": 0.0005184098559310623,
"loss": 3.2826,
"step": 20300
},
{
"epoch": 5.476421188630491,
"grad_norm": 0.3342075049877167,
"learning_rate": 0.0005182078901306046,
"loss": 3.2834,
"step": 20350
},
{
"epoch": 5.489879414298019,
"grad_norm": 0.3341984748840332,
"learning_rate": 0.0005180059243301467,
"loss": 3.2788,
"step": 20400
},
{
"epoch": 5.503337639965547,
"grad_norm": 0.3368922472000122,
"learning_rate": 0.000517803958529689,
"loss": 3.2837,
"step": 20450
},
{
"epoch": 5.516795865633075,
"grad_norm": 0.3083626627922058,
"learning_rate": 0.0005176019927292311,
"loss": 3.2782,
"step": 20500
},
{
"epoch": 5.530254091300603,
"grad_norm": 0.3793262839317322,
"learning_rate": 0.0005174000269287733,
"loss": 3.2717,
"step": 20550
},
{
"epoch": 5.543712316968131,
"grad_norm": 0.34878599643707275,
"learning_rate": 0.0005171980611283156,
"loss": 3.2918,
"step": 20600
},
{
"epoch": 5.557170542635659,
"grad_norm": 0.32627832889556885,
"learning_rate": 0.0005169960953278577,
"loss": 3.2817,
"step": 20650
},
{
"epoch": 5.5706287683031865,
"grad_norm": 0.341691255569458,
"learning_rate": 0.0005167941295274,
"loss": 3.2901,
"step": 20700
},
{
"epoch": 5.584086993970715,
"grad_norm": 0.30279743671417236,
"learning_rate": 0.0005165921637269421,
"loss": 3.2882,
"step": 20750
},
{
"epoch": 5.597545219638243,
"grad_norm": 0.32872816920280457,
"learning_rate": 0.0005163901979264844,
"loss": 3.2781,
"step": 20800
},
{
"epoch": 5.611003445305771,
"grad_norm": 0.34876811504364014,
"learning_rate": 0.0005161882321260266,
"loss": 3.2935,
"step": 20850
},
{
"epoch": 5.624461670973298,
"grad_norm": 0.32410216331481934,
"learning_rate": 0.0005159862663255689,
"loss": 3.2839,
"step": 20900
},
{
"epoch": 5.637919896640827,
"grad_norm": 0.3129251003265381,
"learning_rate": 0.000515784300525111,
"loss": 3.2871,
"step": 20950
},
{
"epoch": 5.651378122308355,
"grad_norm": 0.36038488149642944,
"learning_rate": 0.0005155823347246533,
"loss": 3.278,
"step": 21000
},
{
"epoch": 5.651378122308355,
"eval_accuracy": 0.38312434024196956,
"eval_loss": 3.3651108741760254,
"eval_runtime": 184.035,
"eval_samples_per_second": 97.873,
"eval_steps_per_second": 6.118,
"step": 21000
},
{
"epoch": 5.664836347975883,
"grad_norm": 0.3260107636451721,
"learning_rate": 0.0005153803689241955,
"loss": 3.2833,
"step": 21050
},
{
"epoch": 5.678294573643411,
"grad_norm": 0.32443967461586,
"learning_rate": 0.0005151784031237377,
"loss": 3.3009,
"step": 21100
},
{
"epoch": 5.6917527993109385,
"grad_norm": 0.32225948572158813,
"learning_rate": 0.0005149764373232799,
"loss": 3.2917,
"step": 21150
},
{
"epoch": 5.705211024978467,
"grad_norm": 0.3265635669231415,
"learning_rate": 0.000514774471522822,
"loss": 3.2792,
"step": 21200
},
{
"epoch": 5.718669250645995,
"grad_norm": 0.35437294840812683,
"learning_rate": 0.0005145725057223643,
"loss": 3.2879,
"step": 21250
},
{
"epoch": 5.732127476313523,
"grad_norm": 0.34943774342536926,
"learning_rate": 0.0005143705399219066,
"loss": 3.284,
"step": 21300
},
{
"epoch": 5.745585701981051,
"grad_norm": 0.3317306935787201,
"learning_rate": 0.0005141685741214487,
"loss": 3.2902,
"step": 21350
},
{
"epoch": 5.7590439276485785,
"grad_norm": 0.3153773844242096,
"learning_rate": 0.0005139666083209909,
"loss": 3.2947,
"step": 21400
},
{
"epoch": 5.772502153316107,
"grad_norm": 0.34252089262008667,
"learning_rate": 0.0005137646425205331,
"loss": 3.2983,
"step": 21450
},
{
"epoch": 5.785960378983635,
"grad_norm": 0.3145025372505188,
"learning_rate": 0.0005135626767200753,
"loss": 3.2888,
"step": 21500
},
{
"epoch": 5.799418604651163,
"grad_norm": 0.31910642981529236,
"learning_rate": 0.0005133607109196176,
"loss": 3.2989,
"step": 21550
},
{
"epoch": 5.81287683031869,
"grad_norm": 0.3340320587158203,
"learning_rate": 0.0005131587451191597,
"loss": 3.2954,
"step": 21600
},
{
"epoch": 5.826335055986219,
"grad_norm": 0.32419732213020325,
"learning_rate": 0.000512956779318702,
"loss": 3.2923,
"step": 21650
},
{
"epoch": 5.839793281653747,
"grad_norm": 0.3374021649360657,
"learning_rate": 0.0005127548135182441,
"loss": 3.298,
"step": 21700
},
{
"epoch": 5.853251507321275,
"grad_norm": 0.31075945496559143,
"learning_rate": 0.0005125528477177865,
"loss": 3.2927,
"step": 21750
},
{
"epoch": 5.866709732988802,
"grad_norm": 0.32570937275886536,
"learning_rate": 0.0005123508819173286,
"loss": 3.2908,
"step": 21800
},
{
"epoch": 5.8801679586563305,
"grad_norm": 0.3305903673171997,
"learning_rate": 0.0005121489161168709,
"loss": 3.2959,
"step": 21850
},
{
"epoch": 5.893626184323859,
"grad_norm": 0.33639341592788696,
"learning_rate": 0.000511946950316413,
"loss": 3.3047,
"step": 21900
},
{
"epoch": 5.907084409991387,
"grad_norm": 0.3326457738876343,
"learning_rate": 0.0005117449845159553,
"loss": 3.287,
"step": 21950
},
{
"epoch": 5.920542635658915,
"grad_norm": 0.31374940276145935,
"learning_rate": 0.0005115430187154975,
"loss": 3.2908,
"step": 22000
},
{
"epoch": 5.920542635658915,
"eval_accuracy": 0.3842338419309936,
"eval_loss": 3.3543901443481445,
"eval_runtime": 190.7033,
"eval_samples_per_second": 94.45,
"eval_steps_per_second": 5.904,
"step": 22000
},
{
"epoch": 5.934000861326442,
"grad_norm": 0.32681697607040405,
"learning_rate": 0.0005113410529150397,
"loss": 3.3061,
"step": 22050
},
{
"epoch": 5.9474590869939705,
"grad_norm": 0.3471546769142151,
"learning_rate": 0.0005111390871145819,
"loss": 3.2961,
"step": 22100
},
{
"epoch": 5.960917312661499,
"grad_norm": 0.30677521228790283,
"learning_rate": 0.000510937121314124,
"loss": 3.2998,
"step": 22150
},
{
"epoch": 5.974375538329027,
"grad_norm": 0.33545923233032227,
"learning_rate": 0.0005107351555136663,
"loss": 3.2842,
"step": 22200
},
{
"epoch": 5.987833763996555,
"grad_norm": 0.3202589750289917,
"learning_rate": 0.0005105331897132085,
"loss": 3.2898,
"step": 22250
},
{
"epoch": 6.001076658053402,
"grad_norm": 0.37194105982780457,
"learning_rate": 0.0005103312239127507,
"loss": 3.2943,
"step": 22300
},
{
"epoch": 6.0145348837209305,
"grad_norm": 0.31743311882019043,
"learning_rate": 0.0005101292581122929,
"loss": 3.1983,
"step": 22350
},
{
"epoch": 6.027993109388458,
"grad_norm": 0.3532339930534363,
"learning_rate": 0.0005099272923118351,
"loss": 3.1929,
"step": 22400
},
{
"epoch": 6.041451335055986,
"grad_norm": 0.3264458477497101,
"learning_rate": 0.0005097253265113773,
"loss": 3.2006,
"step": 22450
},
{
"epoch": 6.054909560723514,
"grad_norm": 0.3228222131729126,
"learning_rate": 0.0005095233607109196,
"loss": 3.2009,
"step": 22500
},
{
"epoch": 6.068367786391042,
"grad_norm": 0.32845258712768555,
"learning_rate": 0.0005093213949104617,
"loss": 3.2067,
"step": 22550
},
{
"epoch": 6.0818260120585705,
"grad_norm": 0.3545193076133728,
"learning_rate": 0.000509119429110004,
"loss": 3.2049,
"step": 22600
},
{
"epoch": 6.095284237726098,
"grad_norm": 0.3232004940509796,
"learning_rate": 0.0005089174633095462,
"loss": 3.2102,
"step": 22650
},
{
"epoch": 6.108742463393626,
"grad_norm": 0.34131884574890137,
"learning_rate": 0.0005087154975090885,
"loss": 3.2187,
"step": 22700
},
{
"epoch": 6.122200689061154,
"grad_norm": 0.334764689207077,
"learning_rate": 0.0005085135317086306,
"loss": 3.2092,
"step": 22750
},
{
"epoch": 6.135658914728682,
"grad_norm": 0.32335901260375977,
"learning_rate": 0.0005083115659081729,
"loss": 3.221,
"step": 22800
},
{
"epoch": 6.149117140396211,
"grad_norm": 0.3423883318901062,
"learning_rate": 0.0005081096001077151,
"loss": 3.2185,
"step": 22850
},
{
"epoch": 6.162575366063738,
"grad_norm": 0.32779791951179504,
"learning_rate": 0.0005079076343072573,
"loss": 3.2163,
"step": 22900
},
{
"epoch": 6.176033591731266,
"grad_norm": 0.33753520250320435,
"learning_rate": 0.0005077056685067995,
"loss": 3.2257,
"step": 22950
},
{
"epoch": 6.189491817398794,
"grad_norm": 0.33796757459640503,
"learning_rate": 0.0005075037027063417,
"loss": 3.2329,
"step": 23000
},
{
"epoch": 6.189491817398794,
"eval_accuracy": 0.3843496594790253,
"eval_loss": 3.3589394092559814,
"eval_runtime": 184.1796,
"eval_samples_per_second": 97.796,
"eval_steps_per_second": 6.114,
"step": 23000
},
{
"epoch": 6.2029500430663225,
"grad_norm": 0.35933730006217957,
"learning_rate": 0.0005073017369058839,
"loss": 3.2308,
"step": 23050
},
{
"epoch": 6.21640826873385,
"grad_norm": 0.3183475732803345,
"learning_rate": 0.0005070997711054262,
"loss": 3.2298,
"step": 23100
},
{
"epoch": 6.229866494401378,
"grad_norm": 0.3436269462108612,
"learning_rate": 0.0005068978053049683,
"loss": 3.2352,
"step": 23150
},
{
"epoch": 6.243324720068906,
"grad_norm": 0.3466615378856659,
"learning_rate": 0.0005066958395045105,
"loss": 3.2333,
"step": 23200
},
{
"epoch": 6.256782945736434,
"grad_norm": 0.33215463161468506,
"learning_rate": 0.0005064938737040527,
"loss": 3.2334,
"step": 23250
},
{
"epoch": 6.270241171403962,
"grad_norm": 0.34930655360221863,
"learning_rate": 0.0005062919079035949,
"loss": 3.214,
"step": 23300
},
{
"epoch": 6.28369939707149,
"grad_norm": 0.33633023500442505,
"learning_rate": 0.0005060899421031372,
"loss": 3.2329,
"step": 23350
},
{
"epoch": 6.297157622739018,
"grad_norm": 0.3429878056049347,
"learning_rate": 0.0005058879763026793,
"loss": 3.2422,
"step": 23400
},
{
"epoch": 6.310615848406546,
"grad_norm": 0.33559584617614746,
"learning_rate": 0.0005056860105022216,
"loss": 3.2331,
"step": 23450
},
{
"epoch": 6.324074074074074,
"grad_norm": 0.327995628118515,
"learning_rate": 0.0005054840447017637,
"loss": 3.2353,
"step": 23500
},
{
"epoch": 6.337532299741602,
"grad_norm": 0.3223600685596466,
"learning_rate": 0.0005052820789013061,
"loss": 3.2351,
"step": 23550
},
{
"epoch": 6.35099052540913,
"grad_norm": 0.33200380206108093,
"learning_rate": 0.0005050801131008482,
"loss": 3.2385,
"step": 23600
},
{
"epoch": 6.364448751076658,
"grad_norm": 0.32991448044776917,
"learning_rate": 0.0005048781473003905,
"loss": 3.2366,
"step": 23650
},
{
"epoch": 6.377906976744186,
"grad_norm": 0.3292626440525055,
"learning_rate": 0.0005046761814999326,
"loss": 3.2419,
"step": 23700
},
{
"epoch": 6.3913652024117145,
"grad_norm": 0.32095879316329956,
"learning_rate": 0.0005044742156994749,
"loss": 3.2405,
"step": 23750
},
{
"epoch": 6.404823428079242,
"grad_norm": 0.3349224030971527,
"learning_rate": 0.0005042722498990171,
"loss": 3.2426,
"step": 23800
},
{
"epoch": 6.41828165374677,
"grad_norm": 0.34159955382347107,
"learning_rate": 0.0005040702840985593,
"loss": 3.2423,
"step": 23850
},
{
"epoch": 6.431739879414298,
"grad_norm": 0.33553841710090637,
"learning_rate": 0.0005038683182981015,
"loss": 3.2437,
"step": 23900
},
{
"epoch": 6.445198105081826,
"grad_norm": 0.3332814872264862,
"learning_rate": 0.0005036663524976436,
"loss": 3.2517,
"step": 23950
},
{
"epoch": 6.458656330749354,
"grad_norm": 0.3452901244163513,
"learning_rate": 0.0005034643866971859,
"loss": 3.244,
"step": 24000
},
{
"epoch": 6.458656330749354,
"eval_accuracy": 0.3850031703152454,
"eval_loss": 3.35088849067688,
"eval_runtime": 210.2355,
"eval_samples_per_second": 85.675,
"eval_steps_per_second": 5.356,
"step": 24000
},
{
"epoch": 6.472114556416882,
"grad_norm": 0.32716378569602966,
"learning_rate": 0.0005032624208967281,
"loss": 3.2462,
"step": 24050
},
{
"epoch": 6.48557278208441,
"grad_norm": 0.31713059544563293,
"learning_rate": 0.0005030604550962703,
"loss": 3.2398,
"step": 24100
},
{
"epoch": 6.499031007751938,
"grad_norm": 0.36761173605918884,
"learning_rate": 0.0005028584892958125,
"loss": 3.2529,
"step": 24150
},
{
"epoch": 6.5124892334194655,
"grad_norm": 0.3432551324367523,
"learning_rate": 0.0005026565234953547,
"loss": 3.2589,
"step": 24200
},
{
"epoch": 6.525947459086994,
"grad_norm": 0.3205644488334656,
"learning_rate": 0.0005024545576948969,
"loss": 3.2506,
"step": 24250
},
{
"epoch": 6.539405684754522,
"grad_norm": 0.3178030252456665,
"learning_rate": 0.0005022525918944392,
"loss": 3.244,
"step": 24300
},
{
"epoch": 6.55286391042205,
"grad_norm": 0.3202732801437378,
"learning_rate": 0.0005020506260939813,
"loss": 3.2571,
"step": 24350
},
{
"epoch": 6.566322136089578,
"grad_norm": 0.3263641893863678,
"learning_rate": 0.0005018486602935236,
"loss": 3.2576,
"step": 24400
},
{
"epoch": 6.579780361757106,
"grad_norm": 0.3225899338722229,
"learning_rate": 0.0005016466944930658,
"loss": 3.2348,
"step": 24450
},
{
"epoch": 6.593238587424634,
"grad_norm": 0.3206603527069092,
"learning_rate": 0.0005014447286926081,
"loss": 3.2524,
"step": 24500
},
{
"epoch": 6.606696813092162,
"grad_norm": 0.3224721848964691,
"learning_rate": 0.0005012427628921502,
"loss": 3.2371,
"step": 24550
},
{
"epoch": 6.62015503875969,
"grad_norm": 0.35857126116752625,
"learning_rate": 0.0005010407970916925,
"loss": 3.247,
"step": 24600
},
{
"epoch": 6.633613264427218,
"grad_norm": 0.352993905544281,
"learning_rate": 0.0005008388312912346,
"loss": 3.2538,
"step": 24650
},
{
"epoch": 6.647071490094746,
"grad_norm": 0.3263927102088928,
"learning_rate": 0.0005006368654907769,
"loss": 3.2538,
"step": 24700
},
{
"epoch": 6.660529715762274,
"grad_norm": 0.3047545552253723,
"learning_rate": 0.0005004348996903191,
"loss": 3.2616,
"step": 24750
},
{
"epoch": 6.673987941429802,
"grad_norm": 0.3201688230037689,
"learning_rate": 0.0005002329338898613,
"loss": 3.2486,
"step": 24800
},
{
"epoch": 6.68744616709733,
"grad_norm": 0.34142646193504333,
"learning_rate": 0.0005000309680894035,
"loss": 3.242,
"step": 24850
},
{
"epoch": 6.7009043927648575,
"grad_norm": 0.3450940251350403,
"learning_rate": 0.0004998290022889456,
"loss": 3.257,
"step": 24900
},
{
"epoch": 6.714362618432386,
"grad_norm": 0.3261885643005371,
"learning_rate": 0.0004996270364884879,
"loss": 3.2464,
"step": 24950
},
{
"epoch": 6.727820844099914,
"grad_norm": 0.34189534187316895,
"learning_rate": 0.0004994250706880301,
"loss": 3.2578,
"step": 25000
},
{
"epoch": 6.727820844099914,
"eval_accuracy": 0.38570307335879145,
"eval_loss": 3.3450446128845215,
"eval_runtime": 207.9016,
"eval_samples_per_second": 86.637,
"eval_steps_per_second": 5.416,
"step": 25000
},
{
"epoch": 6.741279069767442,
"grad_norm": 0.36942005157470703,
"learning_rate": 0.0004992231048875723,
"loss": 3.2594,
"step": 25050
},
{
"epoch": 6.754737295434969,
"grad_norm": 0.34088629484176636,
"learning_rate": 0.0004990211390871145,
"loss": 3.2477,
"step": 25100
},
{
"epoch": 6.768195521102498,
"grad_norm": 0.3602665662765503,
"learning_rate": 0.0004988191732866567,
"loss": 3.2514,
"step": 25150
},
{
"epoch": 6.781653746770026,
"grad_norm": 0.33099210262298584,
"learning_rate": 0.0004986172074861989,
"loss": 3.2535,
"step": 25200
},
{
"epoch": 6.795111972437554,
"grad_norm": 0.3461016118526459,
"learning_rate": 0.0004984152416857412,
"loss": 3.2556,
"step": 25250
},
{
"epoch": 6.808570198105082,
"grad_norm": 0.3348025977611542,
"learning_rate": 0.0004982132758852834,
"loss": 3.2517,
"step": 25300
},
{
"epoch": 6.822028423772609,
"grad_norm": 0.3723352253437042,
"learning_rate": 0.0004980113100848256,
"loss": 3.265,
"step": 25350
},
{
"epoch": 6.835486649440138,
"grad_norm": 0.312146931886673,
"learning_rate": 0.0004978093442843678,
"loss": 3.2713,
"step": 25400
},
{
"epoch": 6.848944875107666,
"grad_norm": 0.3408113121986389,
"learning_rate": 0.0004976073784839101,
"loss": 3.256,
"step": 25450
},
{
"epoch": 6.862403100775194,
"grad_norm": 0.3340238630771637,
"learning_rate": 0.0004974054126834522,
"loss": 3.2592,
"step": 25500
},
{
"epoch": 6.875861326442722,
"grad_norm": 0.31821152567863464,
"learning_rate": 0.0004972034468829945,
"loss": 3.2595,
"step": 25550
},
{
"epoch": 6.8893195521102495,
"grad_norm": 0.35574498772621155,
"learning_rate": 0.0004970014810825366,
"loss": 3.2451,
"step": 25600
},
{
"epoch": 6.902777777777778,
"grad_norm": 0.3287249505519867,
"learning_rate": 0.0004967995152820789,
"loss": 3.2443,
"step": 25650
},
{
"epoch": 6.916236003445306,
"grad_norm": 0.3408069312572479,
"learning_rate": 0.0004965975494816211,
"loss": 3.2611,
"step": 25700
},
{
"epoch": 6.929694229112834,
"grad_norm": 0.3223821818828583,
"learning_rate": 0.0004963955836811633,
"loss": 3.2572,
"step": 25750
},
{
"epoch": 6.943152454780362,
"grad_norm": 0.33784809708595276,
"learning_rate": 0.0004961936178807055,
"loss": 3.2551,
"step": 25800
},
{
"epoch": 6.9566106804478895,
"grad_norm": 0.33710768818855286,
"learning_rate": 0.0004959916520802476,
"loss": 3.2555,
"step": 25850
},
{
"epoch": 6.970068906115418,
"grad_norm": 0.32541221380233765,
"learning_rate": 0.0004957896862797899,
"loss": 3.2704,
"step": 25900
},
{
"epoch": 6.983527131782946,
"grad_norm": 0.3266059160232544,
"learning_rate": 0.0004955877204793321,
"loss": 3.2463,
"step": 25950
},
{
"epoch": 6.996985357450473,
"grad_norm": 0.33027753233909607,
"learning_rate": 0.0004953857546788743,
"loss": 3.2524,
"step": 26000
},
{
"epoch": 6.996985357450473,
"eval_accuracy": 0.386393632772759,
"eval_loss": 3.336153030395508,
"eval_runtime": 184.8554,
"eval_samples_per_second": 97.438,
"eval_steps_per_second": 6.091,
"step": 26000
},
{
"epoch": 7.010228251507321,
"grad_norm": 0.34838128089904785,
"learning_rate": 0.0004951837888784165,
"loss": 3.1776,
"step": 26050
},
{
"epoch": 7.0236864771748495,
"grad_norm": 0.3692553639411926,
"learning_rate": 0.0004949818230779587,
"loss": 3.1535,
"step": 26100
},
{
"epoch": 7.037144702842378,
"grad_norm": 0.3452013432979584,
"learning_rate": 0.0004947798572775009,
"loss": 3.1758,
"step": 26150
},
{
"epoch": 7.050602928509905,
"grad_norm": 0.33391502499580383,
"learning_rate": 0.0004945778914770432,
"loss": 3.1686,
"step": 26200
},
{
"epoch": 7.064061154177433,
"grad_norm": 0.3253699243068695,
"learning_rate": 0.0004943759256765854,
"loss": 3.1726,
"step": 26250
},
{
"epoch": 7.077519379844961,
"grad_norm": 0.3186137080192566,
"learning_rate": 0.0004941739598761276,
"loss": 3.1728,
"step": 26300
},
{
"epoch": 7.09097760551249,
"grad_norm": 0.33141475915908813,
"learning_rate": 0.0004939719940756698,
"loss": 3.1771,
"step": 26350
},
{
"epoch": 7.104435831180017,
"grad_norm": 0.36418527364730835,
"learning_rate": 0.0004937700282752121,
"loss": 3.1887,
"step": 26400
},
{
"epoch": 7.117894056847545,
"grad_norm": 0.3267836272716522,
"learning_rate": 0.0004935680624747542,
"loss": 3.1857,
"step": 26450
},
{
"epoch": 7.131352282515073,
"grad_norm": 0.338020920753479,
"learning_rate": 0.0004933660966742965,
"loss": 3.1869,
"step": 26500
},
{
"epoch": 7.144810508182601,
"grad_norm": 0.36454349756240845,
"learning_rate": 0.0004931641308738386,
"loss": 3.1822,
"step": 26550
},
{
"epoch": 7.15826873385013,
"grad_norm": 0.3372531533241272,
"learning_rate": 0.0004929621650733809,
"loss": 3.1775,
"step": 26600
},
{
"epoch": 7.171726959517657,
"grad_norm": 0.35561153292655945,
"learning_rate": 0.0004927601992729231,
"loss": 3.1858,
"step": 26650
},
{
"epoch": 7.185185185185185,
"grad_norm": 0.32375407218933105,
"learning_rate": 0.0004925582334724652,
"loss": 3.1866,
"step": 26700
},
{
"epoch": 7.198643410852713,
"grad_norm": 0.3519543409347534,
"learning_rate": 0.0004923562676720075,
"loss": 3.1967,
"step": 26750
},
{
"epoch": 7.2121016365202415,
"grad_norm": 0.3225705325603485,
"learning_rate": 0.0004921543018715497,
"loss": 3.1941,
"step": 26800
},
{
"epoch": 7.225559862187769,
"grad_norm": 0.33228105306625366,
"learning_rate": 0.0004919523360710919,
"loss": 3.1905,
"step": 26850
},
{
"epoch": 7.239018087855297,
"grad_norm": 0.3495282530784607,
"learning_rate": 0.0004917503702706341,
"loss": 3.1953,
"step": 26900
},
{
"epoch": 7.252476313522825,
"grad_norm": 0.3218580484390259,
"learning_rate": 0.0004915484044701763,
"loss": 3.2045,
"step": 26950
},
{
"epoch": 7.265934539190353,
"grad_norm": 0.3397384285926819,
"learning_rate": 0.0004913464386697185,
"loss": 3.2003,
"step": 27000
},
{
"epoch": 7.265934539190353,
"eval_accuracy": 0.3861059358992244,
"eval_loss": 3.3435847759246826,
"eval_runtime": 184.0316,
"eval_samples_per_second": 97.875,
"eval_steps_per_second": 6.119,
"step": 27000
},
{
"epoch": 7.279392764857882,
"grad_norm": 0.3547893464565277,
"learning_rate": 0.0004911444728692608,
"loss": 3.1919,
"step": 27050
},
{
"epoch": 7.292850990525409,
"grad_norm": 0.33063820004463196,
"learning_rate": 0.000490942507068803,
"loss": 3.2085,
"step": 27100
},
{
"epoch": 7.306309216192937,
"grad_norm": 0.3458538055419922,
"learning_rate": 0.0004907405412683452,
"loss": 3.2009,
"step": 27150
},
{
"epoch": 7.319767441860465,
"grad_norm": 0.37300506234169006,
"learning_rate": 0.0004905385754678874,
"loss": 3.2119,
"step": 27200
},
{
"epoch": 7.333225667527993,
"grad_norm": 0.3209126591682434,
"learning_rate": 0.0004903366096674297,
"loss": 3.2041,
"step": 27250
},
{
"epoch": 7.346683893195521,
"grad_norm": 0.33929765224456787,
"learning_rate": 0.0004901346438669718,
"loss": 3.2028,
"step": 27300
},
{
"epoch": 7.360142118863049,
"grad_norm": 0.34647515416145325,
"learning_rate": 0.0004899326780665141,
"loss": 3.196,
"step": 27350
},
{
"epoch": 7.373600344530577,
"grad_norm": 0.3496509790420532,
"learning_rate": 0.0004897307122660562,
"loss": 3.2065,
"step": 27400
},
{
"epoch": 7.387058570198105,
"grad_norm": 0.3370739221572876,
"learning_rate": 0.0004895287464655985,
"loss": 3.1989,
"step": 27450
},
{
"epoch": 7.4005167958656335,
"grad_norm": 0.3540850579738617,
"learning_rate": 0.0004893267806651407,
"loss": 3.2069,
"step": 27500
},
{
"epoch": 7.413975021533161,
"grad_norm": 0.3450012505054474,
"learning_rate": 0.0004891248148646829,
"loss": 3.1975,
"step": 27550
},
{
"epoch": 7.427433247200689,
"grad_norm": 0.3474440574645996,
"learning_rate": 0.0004889228490642251,
"loss": 3.2067,
"step": 27600
},
{
"epoch": 7.440891472868217,
"grad_norm": 0.3327823579311371,
"learning_rate": 0.0004887208832637672,
"loss": 3.2106,
"step": 27650
},
{
"epoch": 7.454349698535745,
"grad_norm": 0.32230469584465027,
"learning_rate": 0.0004885189174633095,
"loss": 3.2057,
"step": 27700
},
{
"epoch": 7.467807924203273,
"grad_norm": 0.32911375164985657,
"learning_rate": 0.0004883169516628517,
"loss": 3.2025,
"step": 27750
},
{
"epoch": 7.481266149870801,
"grad_norm": 0.31828540563583374,
"learning_rate": 0.0004881149858623939,
"loss": 3.2186,
"step": 27800
},
{
"epoch": 7.494724375538329,
"grad_norm": 0.3589164912700653,
"learning_rate": 0.00048791302006193614,
"loss": 3.2118,
"step": 27850
},
{
"epoch": 7.508182601205857,
"grad_norm": 0.34432271122932434,
"learning_rate": 0.00048771105426147833,
"loss": 3.2092,
"step": 27900
},
{
"epoch": 7.521640826873385,
"grad_norm": 0.32864126563072205,
"learning_rate": 0.00048750908846102053,
"loss": 3.2127,
"step": 27950
},
{
"epoch": 7.535099052540913,
"grad_norm": 0.32947754859924316,
"learning_rate": 0.0004873071226605628,
"loss": 3.2254,
"step": 28000
},
{
"epoch": 7.535099052540913,
"eval_accuracy": 0.38689862335742253,
"eval_loss": 3.3337244987487793,
"eval_runtime": 185.797,
"eval_samples_per_second": 96.945,
"eval_steps_per_second": 6.06,
"step": 28000
},
{
"epoch": 7.548557278208441,
"grad_norm": 0.3368169665336609,
"learning_rate": 0.00048710515686010503,
"loss": 3.2183,
"step": 28050
},
{
"epoch": 7.562015503875969,
"grad_norm": 0.3416260778903961,
"learning_rate": 0.0004869031910596472,
"loss": 3.213,
"step": 28100
},
{
"epoch": 7.575473729543497,
"grad_norm": 0.32012367248535156,
"learning_rate": 0.0004867012252591894,
"loss": 3.2193,
"step": 28150
},
{
"epoch": 7.588931955211025,
"grad_norm": 0.34071072936058044,
"learning_rate": 0.0004864992594587316,
"loss": 3.2124,
"step": 28200
},
{
"epoch": 7.602390180878553,
"grad_norm": 0.3275602161884308,
"learning_rate": 0.0004862972936582738,
"loss": 3.2261,
"step": 28250
},
{
"epoch": 7.615848406546081,
"grad_norm": 0.39443784952163696,
"learning_rate": 0.00048609532785781606,
"loss": 3.2217,
"step": 28300
},
{
"epoch": 7.629306632213609,
"grad_norm": 0.32139527797698975,
"learning_rate": 0.00048589336205735826,
"loss": 3.2222,
"step": 28350
},
{
"epoch": 7.6427648578811365,
"grad_norm": 0.35414424538612366,
"learning_rate": 0.00048569139625690046,
"loss": 3.2162,
"step": 28400
},
{
"epoch": 7.656223083548665,
"grad_norm": 0.3450945019721985,
"learning_rate": 0.00048548943045644265,
"loss": 3.2188,
"step": 28450
},
{
"epoch": 7.669681309216193,
"grad_norm": 0.36079463362693787,
"learning_rate": 0.00048528746465598485,
"loss": 3.2222,
"step": 28500
},
{
"epoch": 7.683139534883721,
"grad_norm": 0.3425680100917816,
"learning_rate": 0.0004850854988555271,
"loss": 3.2185,
"step": 28550
},
{
"epoch": 7.696597760551249,
"grad_norm": 0.32520195841789246,
"learning_rate": 0.0004848835330550693,
"loss": 3.226,
"step": 28600
},
{
"epoch": 7.7100559862187765,
"grad_norm": 0.3207383155822754,
"learning_rate": 0.0004846815672546115,
"loss": 3.234,
"step": 28650
},
{
"epoch": 7.723514211886305,
"grad_norm": 0.31726786494255066,
"learning_rate": 0.0004844796014541537,
"loss": 3.2278,
"step": 28700
},
{
"epoch": 7.736972437553833,
"grad_norm": 0.33718428015708923,
"learning_rate": 0.0004842776356536959,
"loss": 3.221,
"step": 28750
},
{
"epoch": 7.750430663221361,
"grad_norm": 0.335921972990036,
"learning_rate": 0.00048407566985323813,
"loss": 3.228,
"step": 28800
},
{
"epoch": 7.763888888888889,
"grad_norm": 0.3463897407054901,
"learning_rate": 0.0004838737040527804,
"loss": 3.2252,
"step": 28850
},
{
"epoch": 7.777347114556417,
"grad_norm": 0.3304252624511719,
"learning_rate": 0.0004836717382523226,
"loss": 3.2208,
"step": 28900
},
{
"epoch": 7.790805340223945,
"grad_norm": 0.3352733254432678,
"learning_rate": 0.0004834697724518648,
"loss": 3.2211,
"step": 28950
},
{
"epoch": 7.804263565891473,
"grad_norm": 0.34633034467697144,
"learning_rate": 0.000483267806651407,
"loss": 3.233,
"step": 29000
},
{
"epoch": 7.804263565891473,
"eval_accuracy": 0.38776812414250467,
"eval_loss": 3.3259787559509277,
"eval_runtime": 186.6484,
"eval_samples_per_second": 96.502,
"eval_steps_per_second": 6.033,
"step": 29000
},
{
"epoch": 7.817721791559001,
"grad_norm": 0.3276371955871582,
"learning_rate": 0.0004830658408509492,
"loss": 3.227,
"step": 29050
},
{
"epoch": 7.831180017226529,
"grad_norm": 0.35462358593940735,
"learning_rate": 0.0004828638750504914,
"loss": 3.2217,
"step": 29100
},
{
"epoch": 7.844638242894057,
"grad_norm": 0.3029215335845947,
"learning_rate": 0.0004826619092500336,
"loss": 3.2117,
"step": 29150
},
{
"epoch": 7.858096468561585,
"grad_norm": 0.35368239879608154,
"learning_rate": 0.0004824599434495758,
"loss": 3.2306,
"step": 29200
},
{
"epoch": 7.871554694229113,
"grad_norm": 0.3346642553806305,
"learning_rate": 0.00048225797764911806,
"loss": 3.2237,
"step": 29250
},
{
"epoch": 7.885012919896641,
"grad_norm": 0.32475826144218445,
"learning_rate": 0.00048205601184866026,
"loss": 3.2253,
"step": 29300
},
{
"epoch": 7.8984711455641685,
"grad_norm": 0.3455280065536499,
"learning_rate": 0.00048185404604820245,
"loss": 3.2261,
"step": 29350
},
{
"epoch": 7.911929371231697,
"grad_norm": 0.3063681423664093,
"learning_rate": 0.00048165208024774465,
"loss": 3.2363,
"step": 29400
},
{
"epoch": 7.925387596899225,
"grad_norm": 0.33367931842803955,
"learning_rate": 0.00048145011444728685,
"loss": 3.224,
"step": 29450
},
{
"epoch": 7.938845822566753,
"grad_norm": 0.3382367789745331,
"learning_rate": 0.0004812481486468291,
"loss": 3.2307,
"step": 29500
},
{
"epoch": 7.95230404823428,
"grad_norm": 0.2993413507938385,
"learning_rate": 0.0004810461828463713,
"loss": 3.236,
"step": 29550
},
{
"epoch": 7.965762273901809,
"grad_norm": 0.32578542828559875,
"learning_rate": 0.0004808442170459135,
"loss": 3.2261,
"step": 29600
},
{
"epoch": 7.979220499569337,
"grad_norm": 0.31778058409690857,
"learning_rate": 0.0004806422512454557,
"loss": 3.2387,
"step": 29650
},
{
"epoch": 7.992678725236865,
"grad_norm": 0.3182278275489807,
"learning_rate": 0.00048044028544499793,
"loss": 3.2351,
"step": 29700
},
{
"epoch": 8.005921619293712,
"grad_norm": 0.32739055156707764,
"learning_rate": 0.0004802383196445402,
"loss": 3.1853,
"step": 29750
},
{
"epoch": 8.01937984496124,
"grad_norm": 0.3315098285675049,
"learning_rate": 0.0004800363538440824,
"loss": 3.1305,
"step": 29800
},
{
"epoch": 8.032838070628769,
"grad_norm": 0.33604395389556885,
"learning_rate": 0.0004798343880436246,
"loss": 3.1391,
"step": 29850
},
{
"epoch": 8.046296296296296,
"grad_norm": 0.32399725914001465,
"learning_rate": 0.00047963242224316683,
"loss": 3.1262,
"step": 29900
},
{
"epoch": 8.059754521963825,
"grad_norm": 0.3137926161289215,
"learning_rate": 0.000479430456442709,
"loss": 3.1302,
"step": 29950
},
{
"epoch": 8.073212747631352,
"grad_norm": 0.3311665952205658,
"learning_rate": 0.0004792284906422512,
"loss": 3.1498,
"step": 30000
},
{
"epoch": 8.073212747631352,
"eval_accuracy": 0.387525841654596,
"eval_loss": 3.336094617843628,
"eval_runtime": 185.8271,
"eval_samples_per_second": 96.929,
"eval_steps_per_second": 6.059,
"step": 30000
},
{
"epoch": 8.08667097329888,
"grad_norm": 0.34065452218055725,
"learning_rate": 0.0004790265248417934,
"loss": 3.1511,
"step": 30050
},
{
"epoch": 8.100129198966409,
"grad_norm": 0.3404686748981476,
"learning_rate": 0.0004788245590413356,
"loss": 3.1546,
"step": 30100
},
{
"epoch": 8.113587424633936,
"grad_norm": 0.33900272846221924,
"learning_rate": 0.00047862259324087786,
"loss": 3.1439,
"step": 30150
},
{
"epoch": 8.127045650301465,
"grad_norm": 0.3401391804218292,
"learning_rate": 0.00047842062744042006,
"loss": 3.15,
"step": 30200
},
{
"epoch": 8.140503875968992,
"grad_norm": 0.3531796932220459,
"learning_rate": 0.00047821866163996225,
"loss": 3.148,
"step": 30250
},
{
"epoch": 8.15396210163652,
"grad_norm": 0.37316933274269104,
"learning_rate": 0.00047801669583950445,
"loss": 3.1508,
"step": 30300
},
{
"epoch": 8.167420327304049,
"grad_norm": 0.34742501378059387,
"learning_rate": 0.00047781473003904665,
"loss": 3.1604,
"step": 30350
},
{
"epoch": 8.180878552971576,
"grad_norm": 0.350009560585022,
"learning_rate": 0.0004776127642385889,
"loss": 3.1539,
"step": 30400
},
{
"epoch": 8.194336778639105,
"grad_norm": 0.32917118072509766,
"learning_rate": 0.0004774107984381311,
"loss": 3.1672,
"step": 30450
},
{
"epoch": 8.207795004306632,
"grad_norm": 0.343580961227417,
"learning_rate": 0.0004772088326376733,
"loss": 3.1731,
"step": 30500
},
{
"epoch": 8.22125322997416,
"grad_norm": 0.33712247014045715,
"learning_rate": 0.0004770068668372155,
"loss": 3.162,
"step": 30550
},
{
"epoch": 8.234711455641689,
"grad_norm": 0.3233252763748169,
"learning_rate": 0.0004768049010367577,
"loss": 3.1722,
"step": 30600
},
{
"epoch": 8.248169681309216,
"grad_norm": 0.3444722592830658,
"learning_rate": 0.0004766029352363,
"loss": 3.174,
"step": 30650
},
{
"epoch": 8.261627906976745,
"grad_norm": 0.3427727520465851,
"learning_rate": 0.0004764009694358422,
"loss": 3.1821,
"step": 30700
},
{
"epoch": 8.275086132644272,
"grad_norm": 0.35526159405708313,
"learning_rate": 0.0004761990036353844,
"loss": 3.1806,
"step": 30750
},
{
"epoch": 8.2885443583118,
"grad_norm": 0.3580595850944519,
"learning_rate": 0.0004759970378349266,
"loss": 3.1672,
"step": 30800
},
{
"epoch": 8.302002583979329,
"grad_norm": 0.3232937455177307,
"learning_rate": 0.0004757950720344688,
"loss": 3.1725,
"step": 30850
},
{
"epoch": 8.315460809646856,
"grad_norm": 0.35559311509132385,
"learning_rate": 0.000475593106234011,
"loss": 3.1735,
"step": 30900
},
{
"epoch": 8.328919035314383,
"grad_norm": 0.343546599149704,
"learning_rate": 0.0004753911404335532,
"loss": 3.1692,
"step": 30950
},
{
"epoch": 8.342377260981912,
"grad_norm": 0.34287354350090027,
"learning_rate": 0.0004751891746330954,
"loss": 3.1748,
"step": 31000
},
{
"epoch": 8.342377260981912,
"eval_accuracy": 0.38795521402778665,
"eval_loss": 3.331664800643921,
"eval_runtime": 185.9984,
"eval_samples_per_second": 96.84,
"eval_steps_per_second": 6.054,
"step": 31000
},
{
"epoch": 8.35583548664944,
"grad_norm": 0.33396196365356445,
"learning_rate": 0.0004749872088326376,
"loss": 3.1837,
"step": 31050
},
{
"epoch": 8.369293712316969,
"grad_norm": 0.3586781919002533,
"learning_rate": 0.00047478524303217986,
"loss": 3.1816,
"step": 31100
},
{
"epoch": 8.382751937984496,
"grad_norm": 0.32934367656707764,
"learning_rate": 0.00047458327723172206,
"loss": 3.179,
"step": 31150
},
{
"epoch": 8.396210163652023,
"grad_norm": 0.3491109013557434,
"learning_rate": 0.00047438131143126425,
"loss": 3.1809,
"step": 31200
},
{
"epoch": 8.409668389319553,
"grad_norm": 0.3457385003566742,
"learning_rate": 0.00047417934563080645,
"loss": 3.1748,
"step": 31250
},
{
"epoch": 8.42312661498708,
"grad_norm": 0.3532573878765106,
"learning_rate": 0.00047397737983034864,
"loss": 3.1719,
"step": 31300
},
{
"epoch": 8.436584840654609,
"grad_norm": 0.37766119837760925,
"learning_rate": 0.0004737754140298909,
"loss": 3.1846,
"step": 31350
},
{
"epoch": 8.450043066322136,
"grad_norm": 0.3715192973613739,
"learning_rate": 0.0004735734482294331,
"loss": 3.1964,
"step": 31400
},
{
"epoch": 8.463501291989663,
"grad_norm": 0.3497709631919861,
"learning_rate": 0.0004733714824289753,
"loss": 3.1848,
"step": 31450
},
{
"epoch": 8.476959517657193,
"grad_norm": 0.3450281620025635,
"learning_rate": 0.0004731695166285175,
"loss": 3.2005,
"step": 31500
},
{
"epoch": 8.49041774332472,
"grad_norm": 0.34188246726989746,
"learning_rate": 0.0004729675508280598,
"loss": 3.1848,
"step": 31550
},
{
"epoch": 8.503875968992247,
"grad_norm": 0.3407573103904724,
"learning_rate": 0.000472765585027602,
"loss": 3.1898,
"step": 31600
},
{
"epoch": 8.517334194659776,
"grad_norm": 0.32599061727523804,
"learning_rate": 0.0004725636192271442,
"loss": 3.1896,
"step": 31650
},
{
"epoch": 8.530792420327304,
"grad_norm": 0.35123708844184875,
"learning_rate": 0.0004723616534266864,
"loss": 3.1975,
"step": 31700
},
{
"epoch": 8.544250645994833,
"grad_norm": 0.3338778018951416,
"learning_rate": 0.0004721596876262286,
"loss": 3.1891,
"step": 31750
},
{
"epoch": 8.55770887166236,
"grad_norm": 0.3650515079498291,
"learning_rate": 0.0004719577218257708,
"loss": 3.1983,
"step": 31800
},
{
"epoch": 8.571167097329887,
"grad_norm": 0.3237103819847107,
"learning_rate": 0.000471755756025313,
"loss": 3.1913,
"step": 31850
},
{
"epoch": 8.584625322997416,
"grad_norm": 0.3291459381580353,
"learning_rate": 0.0004715537902248552,
"loss": 3.1934,
"step": 31900
},
{
"epoch": 8.598083548664944,
"grad_norm": 0.3320617079734802,
"learning_rate": 0.0004713518244243974,
"loss": 3.1903,
"step": 31950
},
{
"epoch": 8.611541774332473,
"grad_norm": 0.3336716890335083,
"learning_rate": 0.00047114985862393966,
"loss": 3.198,
"step": 32000
},
{
"epoch": 8.611541774332473,
"eval_accuracy": 0.3888428588377481,
"eval_loss": 3.32208514213562,
"eval_runtime": 186.3237,
"eval_samples_per_second": 96.67,
"eval_steps_per_second": 6.043,
"step": 32000
},
{
"epoch": 8.625,
"grad_norm": 0.3506399989128113,
"learning_rate": 0.00047094789282348186,
"loss": 3.1827,
"step": 32050
},
{
"epoch": 8.638458225667527,
"grad_norm": 0.35203826427459717,
"learning_rate": 0.00047074592702302405,
"loss": 3.1782,
"step": 32100
},
{
"epoch": 8.651916451335056,
"grad_norm": 0.3547818958759308,
"learning_rate": 0.00047054396122256625,
"loss": 3.1977,
"step": 32150
},
{
"epoch": 8.665374677002584,
"grad_norm": 0.3448457419872284,
"learning_rate": 0.00047034199542210844,
"loss": 3.1872,
"step": 32200
},
{
"epoch": 8.678832902670113,
"grad_norm": 0.33776548504829407,
"learning_rate": 0.0004701400296216507,
"loss": 3.1992,
"step": 32250
},
{
"epoch": 8.69229112833764,
"grad_norm": 0.33755213022232056,
"learning_rate": 0.0004699380638211929,
"loss": 3.1936,
"step": 32300
},
{
"epoch": 8.705749354005167,
"grad_norm": 0.3232575058937073,
"learning_rate": 0.0004697360980207351,
"loss": 3.1984,
"step": 32350
},
{
"epoch": 8.719207579672696,
"grad_norm": 0.33216333389282227,
"learning_rate": 0.0004695341322202773,
"loss": 3.1866,
"step": 32400
},
{
"epoch": 8.732665805340224,
"grad_norm": 0.3267657458782196,
"learning_rate": 0.0004693321664198196,
"loss": 3.1969,
"step": 32450
},
{
"epoch": 8.746124031007753,
"grad_norm": 0.34185025095939636,
"learning_rate": 0.0004691302006193618,
"loss": 3.2037,
"step": 32500
},
{
"epoch": 8.75958225667528,
"grad_norm": 0.3482416272163391,
"learning_rate": 0.000468928234818904,
"loss": 3.1938,
"step": 32550
},
{
"epoch": 8.773040482342807,
"grad_norm": 0.31263479590415955,
"learning_rate": 0.0004687262690184462,
"loss": 3.1997,
"step": 32600
},
{
"epoch": 8.786498708010337,
"grad_norm": 0.34348827600479126,
"learning_rate": 0.00046852430321798837,
"loss": 3.1997,
"step": 32650
},
{
"epoch": 8.799956933677864,
"grad_norm": 0.34785979986190796,
"learning_rate": 0.0004683223374175306,
"loss": 3.203,
"step": 32700
},
{
"epoch": 8.813415159345391,
"grad_norm": 0.3745204210281372,
"learning_rate": 0.0004681203716170728,
"loss": 3.204,
"step": 32750
},
{
"epoch": 8.82687338501292,
"grad_norm": 0.34430521726608276,
"learning_rate": 0.000467918405816615,
"loss": 3.1966,
"step": 32800
},
{
"epoch": 8.840331610680447,
"grad_norm": 0.32021716237068176,
"learning_rate": 0.0004677164400161572,
"loss": 3.2131,
"step": 32850
},
{
"epoch": 8.853789836347977,
"grad_norm": 0.3275514543056488,
"learning_rate": 0.0004675144742156994,
"loss": 3.203,
"step": 32900
},
{
"epoch": 8.867248062015504,
"grad_norm": 0.33682364225387573,
"learning_rate": 0.00046731250841524166,
"loss": 3.1923,
"step": 32950
},
{
"epoch": 8.880706287683031,
"grad_norm": 0.33345380425453186,
"learning_rate": 0.00046711054261478385,
"loss": 3.1958,
"step": 33000
},
{
"epoch": 8.880706287683031,
"eval_accuracy": 0.38926853721784954,
"eval_loss": 3.3148586750030518,
"eval_runtime": 186.0932,
"eval_samples_per_second": 96.79,
"eval_steps_per_second": 6.051,
"step": 33000
},
{
"epoch": 8.89416451335056,
"grad_norm": 0.337838351726532,
"learning_rate": 0.00046690857681432605,
"loss": 3.2054,
"step": 33050
},
{
"epoch": 8.907622739018088,
"grad_norm": 0.3299519717693329,
"learning_rate": 0.00046670661101386825,
"loss": 3.194,
"step": 33100
},
{
"epoch": 8.921080964685617,
"grad_norm": 0.3324587941169739,
"learning_rate": 0.00046650464521341044,
"loss": 3.1969,
"step": 33150
},
{
"epoch": 8.934539190353144,
"grad_norm": 0.3407862186431885,
"learning_rate": 0.0004663026794129527,
"loss": 3.2005,
"step": 33200
},
{
"epoch": 8.947997416020671,
"grad_norm": 0.3523876965045929,
"learning_rate": 0.0004661007136124949,
"loss": 3.1857,
"step": 33250
},
{
"epoch": 8.9614556416882,
"grad_norm": 0.3353176712989807,
"learning_rate": 0.00046589874781203714,
"loss": 3.2032,
"step": 33300
},
{
"epoch": 8.974913867355728,
"grad_norm": 0.3416292369365692,
"learning_rate": 0.0004656967820115794,
"loss": 3.2035,
"step": 33350
},
{
"epoch": 8.988372093023255,
"grad_norm": 0.3281981945037842,
"learning_rate": 0.0004654948162111216,
"loss": 3.2024,
"step": 33400
},
{
"epoch": 9.001614987080103,
"grad_norm": 0.3397214710712433,
"learning_rate": 0.0004652928504106638,
"loss": 3.2031,
"step": 33450
},
{
"epoch": 9.015073212747632,
"grad_norm": 0.36114996671676636,
"learning_rate": 0.000465090884610206,
"loss": 3.0945,
"step": 33500
},
{
"epoch": 9.02853143841516,
"grad_norm": 0.33979684114456177,
"learning_rate": 0.00046488891880974817,
"loss": 3.1044,
"step": 33550
},
{
"epoch": 9.041989664082687,
"grad_norm": 0.3235820531845093,
"learning_rate": 0.0004646869530092904,
"loss": 3.1085,
"step": 33600
},
{
"epoch": 9.055447889750216,
"grad_norm": 0.34147909283638,
"learning_rate": 0.0004644849872088326,
"loss": 3.1189,
"step": 33650
},
{
"epoch": 9.068906115417743,
"grad_norm": 0.36441078782081604,
"learning_rate": 0.0004642830214083748,
"loss": 3.133,
"step": 33700
},
{
"epoch": 9.082364341085272,
"grad_norm": 0.3297235369682312,
"learning_rate": 0.000464081055607917,
"loss": 3.112,
"step": 33750
},
{
"epoch": 9.0958225667528,
"grad_norm": 0.32468757033348083,
"learning_rate": 0.0004638790898074592,
"loss": 3.1274,
"step": 33800
},
{
"epoch": 9.109280792420327,
"grad_norm": 0.36266252398490906,
"learning_rate": 0.00046367712400700146,
"loss": 3.1284,
"step": 33850
},
{
"epoch": 9.122739018087856,
"grad_norm": 0.3813181221485138,
"learning_rate": 0.00046347515820654365,
"loss": 3.1313,
"step": 33900
},
{
"epoch": 9.136197243755383,
"grad_norm": 0.3511427938938141,
"learning_rate": 0.00046327319240608585,
"loss": 3.1277,
"step": 33950
},
{
"epoch": 9.14965546942291,
"grad_norm": 0.337711364030838,
"learning_rate": 0.00046307122660562805,
"loss": 3.1271,
"step": 34000
},
{
"epoch": 9.14965546942291,
"eval_accuracy": 0.3890759063429338,
"eval_loss": 3.3241565227508545,
"eval_runtime": 186.1149,
"eval_samples_per_second": 96.779,
"eval_steps_per_second": 6.05,
"step": 34000
},
{
"epoch": 9.16311369509044,
"grad_norm": 0.3225274980068207,
"learning_rate": 0.00046286926080517024,
"loss": 3.1298,
"step": 34050
},
{
"epoch": 9.176571920757967,
"grad_norm": 0.3847256600856781,
"learning_rate": 0.0004626672950047125,
"loss": 3.1399,
"step": 34100
},
{
"epoch": 9.190030146425496,
"grad_norm": 0.33026322722435,
"learning_rate": 0.0004624653292042547,
"loss": 3.1359,
"step": 34150
},
{
"epoch": 9.203488372093023,
"grad_norm": 0.34631940722465515,
"learning_rate": 0.00046226336340379694,
"loss": 3.1266,
"step": 34200
},
{
"epoch": 9.21694659776055,
"grad_norm": 0.3768419027328491,
"learning_rate": 0.00046206139760333913,
"loss": 3.1406,
"step": 34250
},
{
"epoch": 9.23040482342808,
"grad_norm": 0.34046241641044617,
"learning_rate": 0.0004618594318028814,
"loss": 3.1383,
"step": 34300
},
{
"epoch": 9.243863049095607,
"grad_norm": 0.32925283908843994,
"learning_rate": 0.0004616574660024236,
"loss": 3.1369,
"step": 34350
},
{
"epoch": 9.257321274763136,
"grad_norm": 0.3488571345806122,
"learning_rate": 0.0004614555002019658,
"loss": 3.1561,
"step": 34400
},
{
"epoch": 9.270779500430663,
"grad_norm": 0.35901886224746704,
"learning_rate": 0.000461253534401508,
"loss": 3.1345,
"step": 34450
},
{
"epoch": 9.28423772609819,
"grad_norm": 0.34168362617492676,
"learning_rate": 0.00046105156860105017,
"loss": 3.1428,
"step": 34500
},
{
"epoch": 9.29769595176572,
"grad_norm": 0.34041690826416016,
"learning_rate": 0.0004608496028005924,
"loss": 3.152,
"step": 34550
},
{
"epoch": 9.311154177433247,
"grad_norm": 0.34896326065063477,
"learning_rate": 0.0004606476370001346,
"loss": 3.1456,
"step": 34600
},
{
"epoch": 9.324612403100776,
"grad_norm": 0.32929185032844543,
"learning_rate": 0.0004604456711996768,
"loss": 3.1539,
"step": 34650
},
{
"epoch": 9.338070628768303,
"grad_norm": 0.3246285319328308,
"learning_rate": 0.000460243705399219,
"loss": 3.148,
"step": 34700
},
{
"epoch": 9.35152885443583,
"grad_norm": 0.32944419980049133,
"learning_rate": 0.0004600417395987612,
"loss": 3.1459,
"step": 34750
},
{
"epoch": 9.36498708010336,
"grad_norm": 0.35956427454948425,
"learning_rate": 0.00045983977379830345,
"loss": 3.1624,
"step": 34800
},
{
"epoch": 9.378445305770887,
"grad_norm": 0.33783426880836487,
"learning_rate": 0.00045963780799784565,
"loss": 3.1519,
"step": 34850
},
{
"epoch": 9.391903531438416,
"grad_norm": 0.3285314738750458,
"learning_rate": 0.00045943584219738785,
"loss": 3.1479,
"step": 34900
},
{
"epoch": 9.405361757105943,
"grad_norm": 0.365739107131958,
"learning_rate": 0.00045923387639693004,
"loss": 3.1474,
"step": 34950
},
{
"epoch": 9.41881998277347,
"grad_norm": 0.32455775141716003,
"learning_rate": 0.00045903191059647224,
"loss": 3.1447,
"step": 35000
},
{
"epoch": 9.41881998277347,
"eval_accuracy": 0.38960827593519953,
"eval_loss": 3.319317102432251,
"eval_runtime": 186.1186,
"eval_samples_per_second": 96.777,
"eval_steps_per_second": 6.05,
"step": 35000
},
{
"epoch": 9.432278208441,
"grad_norm": 0.35738861560821533,
"learning_rate": 0.0004588299447960145,
"loss": 3.164,
"step": 35050
},
{
"epoch": 9.445736434108527,
"grad_norm": 0.34225067496299744,
"learning_rate": 0.00045862797899555674,
"loss": 3.1624,
"step": 35100
},
{
"epoch": 9.459194659776054,
"grad_norm": 0.33486807346343994,
"learning_rate": 0.00045842601319509894,
"loss": 3.1688,
"step": 35150
},
{
"epoch": 9.472652885443583,
"grad_norm": 0.3351769745349884,
"learning_rate": 0.00045822404739464113,
"loss": 3.156,
"step": 35200
},
{
"epoch": 9.48611111111111,
"grad_norm": 0.32922759652137756,
"learning_rate": 0.0004580220815941834,
"loss": 3.1547,
"step": 35250
},
{
"epoch": 9.49956933677864,
"grad_norm": 0.33052223920822144,
"learning_rate": 0.0004578201157937256,
"loss": 3.1597,
"step": 35300
},
{
"epoch": 9.513027562446167,
"grad_norm": 0.30773890018463135,
"learning_rate": 0.0004576181499932678,
"loss": 3.1651,
"step": 35350
},
{
"epoch": 9.526485788113694,
"grad_norm": 0.3447849154472351,
"learning_rate": 0.00045741618419280997,
"loss": 3.1633,
"step": 35400
},
{
"epoch": 9.539944013781223,
"grad_norm": 0.3342150151729584,
"learning_rate": 0.0004572142183923522,
"loss": 3.1675,
"step": 35450
},
{
"epoch": 9.55340223944875,
"grad_norm": 0.32008886337280273,
"learning_rate": 0.0004570122525918944,
"loss": 3.1699,
"step": 35500
},
{
"epoch": 9.56686046511628,
"grad_norm": 0.32356926798820496,
"learning_rate": 0.0004568102867914366,
"loss": 3.1685,
"step": 35550
},
{
"epoch": 9.580318690783807,
"grad_norm": 0.3588160574436188,
"learning_rate": 0.0004566083209909788,
"loss": 3.1694,
"step": 35600
},
{
"epoch": 9.593776916451334,
"grad_norm": 0.34635359048843384,
"learning_rate": 0.000456406355190521,
"loss": 3.1525,
"step": 35650
},
{
"epoch": 9.607235142118864,
"grad_norm": 0.3160857856273651,
"learning_rate": 0.00045620438939006326,
"loss": 3.1657,
"step": 35700
},
{
"epoch": 9.62069336778639,
"grad_norm": 0.35644182562828064,
"learning_rate": 0.00045600242358960545,
"loss": 3.1695,
"step": 35750
},
{
"epoch": 9.634151593453918,
"grad_norm": 0.34612566232681274,
"learning_rate": 0.00045580045778914765,
"loss": 3.1743,
"step": 35800
},
{
"epoch": 9.647609819121447,
"grad_norm": 0.33747079968452454,
"learning_rate": 0.00045559849198868984,
"loss": 3.1688,
"step": 35850
},
{
"epoch": 9.661068044788975,
"grad_norm": 0.37145566940307617,
"learning_rate": 0.00045539652618823204,
"loss": 3.1804,
"step": 35900
},
{
"epoch": 9.674526270456504,
"grad_norm": 0.34744930267333984,
"learning_rate": 0.0004551945603877743,
"loss": 3.1871,
"step": 35950
},
{
"epoch": 9.687984496124031,
"grad_norm": 0.33580830693244934,
"learning_rate": 0.00045499259458731654,
"loss": 3.1872,
"step": 36000
},
{
"epoch": 9.687984496124031,
"eval_accuracy": 0.3898796757803995,
"eval_loss": 3.312854528427124,
"eval_runtime": 185.9837,
"eval_samples_per_second": 96.847,
"eval_steps_per_second": 6.054,
"step": 36000
},
{
"epoch": 9.701442721791558,
"grad_norm": 0.34832096099853516,
"learning_rate": 0.00045479062878685874,
"loss": 3.1804,
"step": 36050
},
{
"epoch": 9.714900947459087,
"grad_norm": 0.3381385803222656,
"learning_rate": 0.00045458866298640093,
"loss": 3.1817,
"step": 36100
},
{
"epoch": 9.728359173126615,
"grad_norm": 0.33040767908096313,
"learning_rate": 0.0004543866971859432,
"loss": 3.173,
"step": 36150
},
{
"epoch": 9.741817398794144,
"grad_norm": 0.3448199927806854,
"learning_rate": 0.0004541847313854854,
"loss": 3.1741,
"step": 36200
},
{
"epoch": 9.755275624461671,
"grad_norm": 0.35476458072662354,
"learning_rate": 0.0004539827655850276,
"loss": 3.1775,
"step": 36250
},
{
"epoch": 9.768733850129198,
"grad_norm": 0.3371541500091553,
"learning_rate": 0.00045378079978456977,
"loss": 3.1848,
"step": 36300
},
{
"epoch": 9.782192075796727,
"grad_norm": 0.3349960744380951,
"learning_rate": 0.00045357883398411197,
"loss": 3.1779,
"step": 36350
},
{
"epoch": 9.795650301464255,
"grad_norm": 0.330491840839386,
"learning_rate": 0.0004533768681836542,
"loss": 3.18,
"step": 36400
},
{
"epoch": 9.809108527131784,
"grad_norm": 0.3703135550022125,
"learning_rate": 0.0004531749023831964,
"loss": 3.1764,
"step": 36450
},
{
"epoch": 9.822566752799311,
"grad_norm": 0.3322940766811371,
"learning_rate": 0.0004529729365827386,
"loss": 3.1845,
"step": 36500
},
{
"epoch": 9.836024978466838,
"grad_norm": 0.34261733293533325,
"learning_rate": 0.0004527709707822808,
"loss": 3.1719,
"step": 36550
},
{
"epoch": 9.849483204134367,
"grad_norm": 0.35383984446525574,
"learning_rate": 0.000452569004981823,
"loss": 3.1692,
"step": 36600
},
{
"epoch": 9.862941429801895,
"grad_norm": 0.34045153856277466,
"learning_rate": 0.00045236703918136525,
"loss": 3.1725,
"step": 36650
},
{
"epoch": 9.876399655469424,
"grad_norm": 0.3356339633464813,
"learning_rate": 0.00045216507338090745,
"loss": 3.18,
"step": 36700
},
{
"epoch": 9.889857881136951,
"grad_norm": 0.32461047172546387,
"learning_rate": 0.00045196310758044964,
"loss": 3.1829,
"step": 36750
},
{
"epoch": 9.903316106804478,
"grad_norm": 0.3401801288127899,
"learning_rate": 0.00045176114177999184,
"loss": 3.1876,
"step": 36800
},
{
"epoch": 9.916774332472007,
"grad_norm": 0.3416830599308014,
"learning_rate": 0.00045155917597953404,
"loss": 3.1587,
"step": 36850
},
{
"epoch": 9.930232558139535,
"grad_norm": 0.3350129723548889,
"learning_rate": 0.00045135721017907634,
"loss": 3.1857,
"step": 36900
},
{
"epoch": 9.943690783807062,
"grad_norm": 0.35851678252220154,
"learning_rate": 0.00045115524437861854,
"loss": 3.1834,
"step": 36950
},
{
"epoch": 9.957149009474591,
"grad_norm": 0.33560651540756226,
"learning_rate": 0.00045095327857816073,
"loss": 3.1771,
"step": 37000
},
{
"epoch": 9.957149009474591,
"eval_accuracy": 0.39045669923030224,
"eval_loss": 3.3063430786132812,
"eval_runtime": 184.1549,
"eval_samples_per_second": 97.809,
"eval_steps_per_second": 6.114,
"step": 37000
},
{
"epoch": 9.970607235142118,
"grad_norm": 0.3233015835285187,
"learning_rate": 0.00045075131277770293,
"loss": 3.1889,
"step": 37050
},
{
"epoch": 9.984065460809648,
"grad_norm": 0.3450278639793396,
"learning_rate": 0.0004505493469772452,
"loss": 3.1855,
"step": 37100
},
{
"epoch": 9.997523686477175,
"grad_norm": 0.3550490140914917,
"learning_rate": 0.0004503473811767874,
"loss": 3.1896,
"step": 37150
},
{
"epoch": 10.010766580534023,
"grad_norm": 0.3696052134037018,
"learning_rate": 0.00045014541537632957,
"loss": 3.1127,
"step": 37200
},
{
"epoch": 10.02422480620155,
"grad_norm": 0.32492905855178833,
"learning_rate": 0.00044994344957587177,
"loss": 3.0756,
"step": 37250
},
{
"epoch": 10.037683031869078,
"grad_norm": 0.3414837419986725,
"learning_rate": 0.00044974148377541396,
"loss": 3.0937,
"step": 37300
},
{
"epoch": 10.051141257536607,
"grad_norm": 0.35179856419563293,
"learning_rate": 0.0004495395179749562,
"loss": 3.0899,
"step": 37350
},
{
"epoch": 10.064599483204134,
"grad_norm": 0.338301420211792,
"learning_rate": 0.0004493375521744984,
"loss": 3.0917,
"step": 37400
},
{
"epoch": 10.078057708871663,
"grad_norm": 0.34596017003059387,
"learning_rate": 0.0004491355863740406,
"loss": 3.0969,
"step": 37450
},
{
"epoch": 10.09151593453919,
"grad_norm": 0.3326232433319092,
"learning_rate": 0.0004489336205735828,
"loss": 3.0979,
"step": 37500
},
{
"epoch": 10.104974160206718,
"grad_norm": 0.35589876770973206,
"learning_rate": 0.00044873165477312505,
"loss": 3.1084,
"step": 37550
},
{
"epoch": 10.118432385874247,
"grad_norm": 0.3320654630661011,
"learning_rate": 0.00044852968897266725,
"loss": 3.1078,
"step": 37600
},
{
"epoch": 10.131890611541774,
"grad_norm": 0.35532307624816895,
"learning_rate": 0.00044832772317220945,
"loss": 3.1036,
"step": 37650
},
{
"epoch": 10.145348837209303,
"grad_norm": 0.3208889663219452,
"learning_rate": 0.00044812575737175164,
"loss": 3.1026,
"step": 37700
},
{
"epoch": 10.15880706287683,
"grad_norm": 0.33862733840942383,
"learning_rate": 0.00044792379157129384,
"loss": 3.0999,
"step": 37750
},
{
"epoch": 10.172265288544358,
"grad_norm": 0.35630491375923157,
"learning_rate": 0.00044772182577083614,
"loss": 3.1054,
"step": 37800
},
{
"epoch": 10.185723514211887,
"grad_norm": 0.3559173345565796,
"learning_rate": 0.00044751985997037834,
"loss": 3.1161,
"step": 37850
},
{
"epoch": 10.199181739879414,
"grad_norm": 0.3298300802707672,
"learning_rate": 0.00044731789416992053,
"loss": 3.1111,
"step": 37900
},
{
"epoch": 10.212639965546943,
"grad_norm": 0.3417648375034332,
"learning_rate": 0.00044711592836946273,
"loss": 3.1149,
"step": 37950
},
{
"epoch": 10.22609819121447,
"grad_norm": 0.34908634424209595,
"learning_rate": 0.000446913962569005,
"loss": 3.1086,
"step": 38000
},
{
"epoch": 10.22609819121447,
"eval_accuracy": 0.3897027986995406,
"eval_loss": 3.320788860321045,
"eval_runtime": 183.9879,
"eval_samples_per_second": 97.898,
"eval_steps_per_second": 6.12,
"step": 38000
},
{
"epoch": 10.239556416881998,
"grad_norm": 0.32685866951942444,
"learning_rate": 0.0004467119967685472,
"loss": 3.1134,
"step": 38050
},
{
"epoch": 10.253014642549527,
"grad_norm": 0.36496973037719727,
"learning_rate": 0.00044651003096808937,
"loss": 3.1224,
"step": 38100
},
{
"epoch": 10.266472868217054,
"grad_norm": 0.35888203978538513,
"learning_rate": 0.00044630806516763157,
"loss": 3.1315,
"step": 38150
},
{
"epoch": 10.279931093884581,
"grad_norm": 0.342044860124588,
"learning_rate": 0.00044610609936717376,
"loss": 3.1257,
"step": 38200
},
{
"epoch": 10.29338931955211,
"grad_norm": 0.35341212153434753,
"learning_rate": 0.000445904133566716,
"loss": 3.1227,
"step": 38250
},
{
"epoch": 10.306847545219638,
"grad_norm": 0.3511538803577423,
"learning_rate": 0.0004457021677662582,
"loss": 3.1289,
"step": 38300
},
{
"epoch": 10.320305770887167,
"grad_norm": 0.3421810269355774,
"learning_rate": 0.0004455002019658004,
"loss": 3.1389,
"step": 38350
},
{
"epoch": 10.333763996554694,
"grad_norm": 0.3299587368965149,
"learning_rate": 0.0004452982361653426,
"loss": 3.1353,
"step": 38400
},
{
"epoch": 10.347222222222221,
"grad_norm": 0.33756786584854126,
"learning_rate": 0.0004450962703648848,
"loss": 3.1287,
"step": 38450
},
{
"epoch": 10.36068044788975,
"grad_norm": 0.3481440246105194,
"learning_rate": 0.00044489430456442705,
"loss": 3.1339,
"step": 38500
},
{
"epoch": 10.374138673557278,
"grad_norm": 0.3543851673603058,
"learning_rate": 0.00044469233876396925,
"loss": 3.1407,
"step": 38550
},
{
"epoch": 10.387596899224807,
"grad_norm": 0.3405233323574066,
"learning_rate": 0.00044449037296351144,
"loss": 3.1386,
"step": 38600
},
{
"epoch": 10.401055124892334,
"grad_norm": 0.35309383273124695,
"learning_rate": 0.0004442884071630537,
"loss": 3.1386,
"step": 38650
},
{
"epoch": 10.414513350559861,
"grad_norm": 0.3596074879169464,
"learning_rate": 0.00044408644136259594,
"loss": 3.1371,
"step": 38700
},
{
"epoch": 10.42797157622739,
"grad_norm": 0.34392717480659485,
"learning_rate": 0.00044388447556213814,
"loss": 3.1289,
"step": 38750
},
{
"epoch": 10.441429801894918,
"grad_norm": 0.33129942417144775,
"learning_rate": 0.00044368250976168033,
"loss": 3.1412,
"step": 38800
},
{
"epoch": 10.454888027562447,
"grad_norm": 0.3527941405773163,
"learning_rate": 0.00044348054396122253,
"loss": 3.1311,
"step": 38850
},
{
"epoch": 10.468346253229974,
"grad_norm": 0.35372987389564514,
"learning_rate": 0.0004432785781607647,
"loss": 3.1466,
"step": 38900
},
{
"epoch": 10.481804478897502,
"grad_norm": 0.3324158489704132,
"learning_rate": 0.000443076612360307,
"loss": 3.1428,
"step": 38950
},
{
"epoch": 10.49526270456503,
"grad_norm": 0.3677845299243927,
"learning_rate": 0.0004428746465598492,
"loss": 3.139,
"step": 39000
},
{
"epoch": 10.49526270456503,
"eval_accuracy": 0.39043649091516724,
"eval_loss": 3.3111002445220947,
"eval_runtime": 184.3459,
"eval_samples_per_second": 97.708,
"eval_steps_per_second": 6.108,
"step": 39000
},
{
"epoch": 10.508720930232558,
"grad_norm": 0.33336344361305237,
"learning_rate": 0.00044267268075939137,
"loss": 3.1474,
"step": 39050
},
{
"epoch": 10.522179155900087,
"grad_norm": 0.3398861587047577,
"learning_rate": 0.00044247071495893357,
"loss": 3.1327,
"step": 39100
},
{
"epoch": 10.535637381567614,
"grad_norm": 0.35154464840888977,
"learning_rate": 0.00044226874915847576,
"loss": 3.1372,
"step": 39150
},
{
"epoch": 10.549095607235142,
"grad_norm": 0.33072924613952637,
"learning_rate": 0.000442066783358018,
"loss": 3.1478,
"step": 39200
},
{
"epoch": 10.56255383290267,
"grad_norm": 0.36191022396087646,
"learning_rate": 0.0004418648175575602,
"loss": 3.1442,
"step": 39250
},
{
"epoch": 10.576012058570198,
"grad_norm": 0.34505319595336914,
"learning_rate": 0.0004416628517571024,
"loss": 3.1442,
"step": 39300
},
{
"epoch": 10.589470284237725,
"grad_norm": 0.3552291691303253,
"learning_rate": 0.0004414608859566446,
"loss": 3.1517,
"step": 39350
},
{
"epoch": 10.602928509905254,
"grad_norm": 0.32551783323287964,
"learning_rate": 0.00044125892015618685,
"loss": 3.1511,
"step": 39400
},
{
"epoch": 10.616386735572782,
"grad_norm": 0.32012292742729187,
"learning_rate": 0.00044105695435572905,
"loss": 3.151,
"step": 39450
},
{
"epoch": 10.62984496124031,
"grad_norm": 0.3490573465824127,
"learning_rate": 0.00044085498855527124,
"loss": 3.1482,
"step": 39500
},
{
"epoch": 10.643303186907838,
"grad_norm": 0.32286253571510315,
"learning_rate": 0.0004406530227548135,
"loss": 3.1519,
"step": 39550
},
{
"epoch": 10.656761412575365,
"grad_norm": 0.34155648946762085,
"learning_rate": 0.00044045105695435574,
"loss": 3.1481,
"step": 39600
},
{
"epoch": 10.670219638242894,
"grad_norm": 0.35249364376068115,
"learning_rate": 0.00044024909115389794,
"loss": 3.1475,
"step": 39650
},
{
"epoch": 10.683677863910422,
"grad_norm": 0.3633196949958801,
"learning_rate": 0.00044004712535344014,
"loss": 3.1552,
"step": 39700
},
{
"epoch": 10.69713608957795,
"grad_norm": 0.3480837643146515,
"learning_rate": 0.00043984515955298233,
"loss": 3.155,
"step": 39750
},
{
"epoch": 10.710594315245478,
"grad_norm": 0.33564871549606323,
"learning_rate": 0.00043964319375252453,
"loss": 3.1654,
"step": 39800
},
{
"epoch": 10.724052540913005,
"grad_norm": 0.3439793288707733,
"learning_rate": 0.0004394412279520668,
"loss": 3.1521,
"step": 39850
},
{
"epoch": 10.737510766580534,
"grad_norm": 0.3309365510940552,
"learning_rate": 0.000439239262151609,
"loss": 3.1516,
"step": 39900
},
{
"epoch": 10.750968992248062,
"grad_norm": 0.34511929750442505,
"learning_rate": 0.00043903729635115117,
"loss": 3.1459,
"step": 39950
},
{
"epoch": 10.764427217915589,
"grad_norm": 0.35131409764289856,
"learning_rate": 0.00043883533055069337,
"loss": 3.1597,
"step": 40000
},
{
"epoch": 10.764427217915589,
"eval_accuracy": 0.391130744322224,
"eval_loss": 3.305757522583008,
"eval_runtime": 184.235,
"eval_samples_per_second": 97.766,
"eval_steps_per_second": 6.112,
"step": 40000
},
{
"epoch": 10.777885443583118,
"grad_norm": 0.35120439529418945,
"learning_rate": 0.00043863336475023556,
"loss": 3.1469,
"step": 40050
},
{
"epoch": 10.791343669250645,
"grad_norm": 0.3553171753883362,
"learning_rate": 0.0004384313989497778,
"loss": 3.1557,
"step": 40100
},
{
"epoch": 10.804801894918175,
"grad_norm": 0.3421008586883545,
"learning_rate": 0.00043822943314932,
"loss": 3.1617,
"step": 40150
},
{
"epoch": 10.818260120585702,
"grad_norm": 0.3476274311542511,
"learning_rate": 0.0004380274673488622,
"loss": 3.1654,
"step": 40200
},
{
"epoch": 10.83171834625323,
"grad_norm": 0.32448360323905945,
"learning_rate": 0.0004378255015484044,
"loss": 3.1654,
"step": 40250
},
{
"epoch": 10.845176571920758,
"grad_norm": 0.3648766875267029,
"learning_rate": 0.0004376235357479466,
"loss": 3.1702,
"step": 40300
},
{
"epoch": 10.858634797588286,
"grad_norm": 0.3453064560890198,
"learning_rate": 0.00043742156994748885,
"loss": 3.1548,
"step": 40350
},
{
"epoch": 10.872093023255815,
"grad_norm": 0.3510722219944,
"learning_rate": 0.00043721960414703104,
"loss": 3.1634,
"step": 40400
},
{
"epoch": 10.885551248923342,
"grad_norm": 0.37653017044067383,
"learning_rate": 0.0004370176383465733,
"loss": 3.1574,
"step": 40450
},
{
"epoch": 10.89900947459087,
"grad_norm": 0.3574938476085663,
"learning_rate": 0.0004368156725461155,
"loss": 3.1662,
"step": 40500
},
{
"epoch": 10.912467700258398,
"grad_norm": 0.34568560123443604,
"learning_rate": 0.00043661370674565774,
"loss": 3.1594,
"step": 40550
},
{
"epoch": 10.925925925925926,
"grad_norm": 0.3241981565952301,
"learning_rate": 0.00043641174094519994,
"loss": 3.164,
"step": 40600
},
{
"epoch": 10.939384151593455,
"grad_norm": 0.33587074279785156,
"learning_rate": 0.00043620977514474213,
"loss": 3.1729,
"step": 40650
},
{
"epoch": 10.952842377260982,
"grad_norm": 0.35898181796073914,
"learning_rate": 0.00043600780934428433,
"loss": 3.1596,
"step": 40700
},
{
"epoch": 10.96630060292851,
"grad_norm": 0.329121857881546,
"learning_rate": 0.0004358058435438265,
"loss": 3.1592,
"step": 40750
},
{
"epoch": 10.979758828596038,
"grad_norm": 0.3552490472793579,
"learning_rate": 0.0004356038777433688,
"loss": 3.1662,
"step": 40800
},
{
"epoch": 10.993217054263566,
"grad_norm": 0.3337413966655731,
"learning_rate": 0.00043540191194291097,
"loss": 3.162,
"step": 40850
},
{
"epoch": 11.006459948320414,
"grad_norm": 0.3448871374130249,
"learning_rate": 0.00043519994614245317,
"loss": 3.1149,
"step": 40900
},
{
"epoch": 11.019918173987941,
"grad_norm": 0.35687804222106934,
"learning_rate": 0.00043499798034199536,
"loss": 3.0439,
"step": 40950
},
{
"epoch": 11.03337639965547,
"grad_norm": 0.3468148410320282,
"learning_rate": 0.00043479601454153756,
"loss": 3.0624,
"step": 41000
},
{
"epoch": 11.03337639965547,
"eval_accuracy": 0.3910536050547732,
"eval_loss": 3.3113489151000977,
"eval_runtime": 184.1594,
"eval_samples_per_second": 97.807,
"eval_steps_per_second": 6.114,
"step": 41000
},
{
"epoch": 11.046834625322997,
"grad_norm": 0.3479900062084198,
"learning_rate": 0.0004345940487410798,
"loss": 3.0661,
"step": 41050
},
{
"epoch": 11.060292850990525,
"grad_norm": 0.3558938205242157,
"learning_rate": 0.000434392082940622,
"loss": 3.0705,
"step": 41100
},
{
"epoch": 11.073751076658054,
"grad_norm": 0.35417845845222473,
"learning_rate": 0.0004341901171401642,
"loss": 3.0766,
"step": 41150
},
{
"epoch": 11.087209302325581,
"grad_norm": 0.3405758738517761,
"learning_rate": 0.0004339881513397064,
"loss": 3.0856,
"step": 41200
},
{
"epoch": 11.10066752799311,
"grad_norm": 0.3564693331718445,
"learning_rate": 0.0004337861855392486,
"loss": 3.0749,
"step": 41250
},
{
"epoch": 11.114125753660637,
"grad_norm": 0.33654284477233887,
"learning_rate": 0.00043358421973879084,
"loss": 3.0759,
"step": 41300
},
{
"epoch": 11.127583979328165,
"grad_norm": 0.3569105565547943,
"learning_rate": 0.0004333822539383331,
"loss": 3.0908,
"step": 41350
},
{
"epoch": 11.141042204995694,
"grad_norm": 0.3643410801887512,
"learning_rate": 0.0004331802881378753,
"loss": 3.0932,
"step": 41400
},
{
"epoch": 11.154500430663221,
"grad_norm": 0.3264082670211792,
"learning_rate": 0.00043297832233741754,
"loss": 3.0895,
"step": 41450
},
{
"epoch": 11.16795865633075,
"grad_norm": 0.3422051668167114,
"learning_rate": 0.00043277635653695974,
"loss": 3.0891,
"step": 41500
},
{
"epoch": 11.181416881998278,
"grad_norm": 0.36176085472106934,
"learning_rate": 0.00043257439073650193,
"loss": 3.0945,
"step": 41550
},
{
"epoch": 11.194875107665805,
"grad_norm": 0.34829625487327576,
"learning_rate": 0.00043237242493604413,
"loss": 3.1002,
"step": 41600
},
{
"epoch": 11.208333333333334,
"grad_norm": 0.35498711466789246,
"learning_rate": 0.0004321704591355863,
"loss": 3.0914,
"step": 41650
},
{
"epoch": 11.221791559000861,
"grad_norm": 0.34146609902381897,
"learning_rate": 0.0004319684933351286,
"loss": 3.104,
"step": 41700
},
{
"epoch": 11.235249784668389,
"grad_norm": 0.3370722532272339,
"learning_rate": 0.00043176652753467077,
"loss": 3.1059,
"step": 41750
},
{
"epoch": 11.248708010335918,
"grad_norm": 0.33574923872947693,
"learning_rate": 0.00043156456173421297,
"loss": 3.101,
"step": 41800
},
{
"epoch": 11.262166236003445,
"grad_norm": 0.3323531746864319,
"learning_rate": 0.00043136259593375516,
"loss": 3.0991,
"step": 41850
},
{
"epoch": 11.275624461670974,
"grad_norm": 0.3613124191761017,
"learning_rate": 0.00043116063013329736,
"loss": 3.1122,
"step": 41900
},
{
"epoch": 11.289082687338501,
"grad_norm": 0.332539439201355,
"learning_rate": 0.0004309586643328396,
"loss": 3.1042,
"step": 41950
},
{
"epoch": 11.302540913006029,
"grad_norm": 0.3468951880931854,
"learning_rate": 0.0004307566985323818,
"loss": 3.113,
"step": 42000
},
{
"epoch": 11.302540913006029,
"eval_accuracy": 0.3912112516421972,
"eval_loss": 3.310168504714966,
"eval_runtime": 184.0259,
"eval_samples_per_second": 97.878,
"eval_steps_per_second": 6.119,
"step": 42000
},
{
"epoch": 11.315999138673558,
"grad_norm": 0.3509575426578522,
"learning_rate": 0.000430554732731924,
"loss": 3.1084,
"step": 42050
},
{
"epoch": 11.329457364341085,
"grad_norm": 0.3336736559867859,
"learning_rate": 0.0004303527669314662,
"loss": 3.1153,
"step": 42100
},
{
"epoch": 11.342915590008614,
"grad_norm": 0.3402518928050995,
"learning_rate": 0.0004301508011310084,
"loss": 3.1167,
"step": 42150
},
{
"epoch": 11.356373815676141,
"grad_norm": 0.3378124535083771,
"learning_rate": 0.00042994883533055065,
"loss": 3.1258,
"step": 42200
},
{
"epoch": 11.369832041343669,
"grad_norm": 0.349943608045578,
"learning_rate": 0.0004297468695300929,
"loss": 3.1119,
"step": 42250
},
{
"epoch": 11.383290267011198,
"grad_norm": 0.34309375286102295,
"learning_rate": 0.0004295449037296351,
"loss": 3.1074,
"step": 42300
},
{
"epoch": 11.396748492678725,
"grad_norm": 0.34346091747283936,
"learning_rate": 0.0004293429379291773,
"loss": 3.1067,
"step": 42350
},
{
"epoch": 11.410206718346252,
"grad_norm": 0.33718785643577576,
"learning_rate": 0.00042914097212871954,
"loss": 3.1042,
"step": 42400
},
{
"epoch": 11.423664944013781,
"grad_norm": 0.34948456287384033,
"learning_rate": 0.00042893900632826173,
"loss": 3.1212,
"step": 42450
},
{
"epoch": 11.437123169681309,
"grad_norm": 0.35313427448272705,
"learning_rate": 0.00042873704052780393,
"loss": 3.1103,
"step": 42500
},
{
"epoch": 11.450581395348838,
"grad_norm": 0.3484950065612793,
"learning_rate": 0.0004285350747273461,
"loss": 3.1196,
"step": 42550
},
{
"epoch": 11.464039621016365,
"grad_norm": 0.3654516935348511,
"learning_rate": 0.0004283331089268883,
"loss": 3.1259,
"step": 42600
},
{
"epoch": 11.477497846683892,
"grad_norm": 0.3586653172969818,
"learning_rate": 0.00042813114312643057,
"loss": 3.1296,
"step": 42650
},
{
"epoch": 11.490956072351421,
"grad_norm": 0.3398750424385071,
"learning_rate": 0.00042792917732597277,
"loss": 3.1235,
"step": 42700
},
{
"epoch": 11.504414298018949,
"grad_norm": 0.3511350154876709,
"learning_rate": 0.00042772721152551496,
"loss": 3.1213,
"step": 42750
},
{
"epoch": 11.517872523686478,
"grad_norm": 0.3482672870159149,
"learning_rate": 0.00042752524572505716,
"loss": 3.1218,
"step": 42800
},
{
"epoch": 11.531330749354005,
"grad_norm": 0.3323778808116913,
"learning_rate": 0.00042732327992459936,
"loss": 3.1257,
"step": 42850
},
{
"epoch": 11.544788975021532,
"grad_norm": 0.33932504057884216,
"learning_rate": 0.0004271213141241416,
"loss": 3.1266,
"step": 42900
},
{
"epoch": 11.558247200689062,
"grad_norm": 0.341813325881958,
"learning_rate": 0.0004269193483236838,
"loss": 3.1403,
"step": 42950
},
{
"epoch": 11.571705426356589,
"grad_norm": 0.3711701035499573,
"learning_rate": 0.000426717382523226,
"loss": 3.1245,
"step": 43000
},
{
"epoch": 11.571705426356589,
"eval_accuracy": 0.39164866388269964,
"eval_loss": 3.3033370971679688,
"eval_runtime": 184.4759,
"eval_samples_per_second": 97.639,
"eval_steps_per_second": 6.104,
"step": 43000
},
{
"epoch": 11.585163652024118,
"grad_norm": 0.34867534041404724,
"learning_rate": 0.0004265154167227682,
"loss": 3.1279,
"step": 43050
},
{
"epoch": 11.598621877691645,
"grad_norm": 0.37046706676483154,
"learning_rate": 0.0004263134509223105,
"loss": 3.1235,
"step": 43100
},
{
"epoch": 11.612080103359173,
"grad_norm": 0.36348801851272583,
"learning_rate": 0.0004261114851218527,
"loss": 3.1412,
"step": 43150
},
{
"epoch": 11.625538329026702,
"grad_norm": 0.3507877290248871,
"learning_rate": 0.0004259095193213949,
"loss": 3.118,
"step": 43200
},
{
"epoch": 11.638996554694229,
"grad_norm": 0.33848443627357483,
"learning_rate": 0.0004257075535209371,
"loss": 3.1363,
"step": 43250
},
{
"epoch": 11.652454780361758,
"grad_norm": 0.360700398683548,
"learning_rate": 0.00042550558772047934,
"loss": 3.1376,
"step": 43300
},
{
"epoch": 11.665913006029285,
"grad_norm": 0.3454159200191498,
"learning_rate": 0.00042530362192002153,
"loss": 3.137,
"step": 43350
},
{
"epoch": 11.679371231696813,
"grad_norm": 0.34485337138175964,
"learning_rate": 0.00042510165611956373,
"loss": 3.141,
"step": 43400
},
{
"epoch": 11.692829457364342,
"grad_norm": 0.3429340720176697,
"learning_rate": 0.0004248996903191059,
"loss": 3.1344,
"step": 43450
},
{
"epoch": 11.706287683031869,
"grad_norm": 0.36432963609695435,
"learning_rate": 0.0004246977245186481,
"loss": 3.1307,
"step": 43500
},
{
"epoch": 11.719745908699396,
"grad_norm": 0.332146555185318,
"learning_rate": 0.0004244957587181904,
"loss": 3.1308,
"step": 43550
},
{
"epoch": 11.733204134366925,
"grad_norm": 0.33760833740234375,
"learning_rate": 0.00042429379291773257,
"loss": 3.1369,
"step": 43600
},
{
"epoch": 11.746662360034453,
"grad_norm": 0.36386963725090027,
"learning_rate": 0.00042409182711727477,
"loss": 3.1408,
"step": 43650
},
{
"epoch": 11.760120585701982,
"grad_norm": 0.35179609060287476,
"learning_rate": 0.00042388986131681696,
"loss": 3.1428,
"step": 43700
},
{
"epoch": 11.773578811369509,
"grad_norm": 0.3373825252056122,
"learning_rate": 0.00042368789551635916,
"loss": 3.1336,
"step": 43750
},
{
"epoch": 11.787037037037036,
"grad_norm": 0.3600446879863739,
"learning_rate": 0.0004234859297159014,
"loss": 3.137,
"step": 43800
},
{
"epoch": 11.800495262704565,
"grad_norm": 0.3524215817451477,
"learning_rate": 0.0004232839639154436,
"loss": 3.1293,
"step": 43850
},
{
"epoch": 11.813953488372093,
"grad_norm": 0.380283921957016,
"learning_rate": 0.0004230819981149858,
"loss": 3.1335,
"step": 43900
},
{
"epoch": 11.827411714039622,
"grad_norm": 0.34472236037254333,
"learning_rate": 0.000422880032314528,
"loss": 3.1474,
"step": 43950
},
{
"epoch": 11.840869939707149,
"grad_norm": 0.3843957483768463,
"learning_rate": 0.0004226780665140703,
"loss": 3.143,
"step": 44000
},
{
"epoch": 11.840869939707149,
"eval_accuracy": 0.39227034119023935,
"eval_loss": 3.296403408050537,
"eval_runtime": 183.9659,
"eval_samples_per_second": 97.909,
"eval_steps_per_second": 6.121,
"step": 44000
},
{
"epoch": 11.854328165374676,
"grad_norm": 0.3647637367248535,
"learning_rate": 0.0004224761007136125,
"loss": 3.1438,
"step": 44050
},
{
"epoch": 11.867786391042205,
"grad_norm": 0.3520395755767822,
"learning_rate": 0.0004222741349131547,
"loss": 3.1418,
"step": 44100
},
{
"epoch": 11.881244616709733,
"grad_norm": 0.353656142950058,
"learning_rate": 0.0004220721691126969,
"loss": 3.145,
"step": 44150
},
{
"epoch": 11.89470284237726,
"grad_norm": 0.35087084770202637,
"learning_rate": 0.0004218702033122391,
"loss": 3.1493,
"step": 44200
},
{
"epoch": 11.90816106804479,
"grad_norm": 0.3734379708766937,
"learning_rate": 0.00042166823751178134,
"loss": 3.1414,
"step": 44250
},
{
"epoch": 11.921619293712316,
"grad_norm": 0.37449437379837036,
"learning_rate": 0.00042146627171132353,
"loss": 3.1394,
"step": 44300
},
{
"epoch": 11.935077519379846,
"grad_norm": 0.3344019651412964,
"learning_rate": 0.00042126430591086573,
"loss": 3.1409,
"step": 44350
},
{
"epoch": 11.948535745047373,
"grad_norm": 0.34413352608680725,
"learning_rate": 0.0004210623401104079,
"loss": 3.145,
"step": 44400
},
{
"epoch": 11.9619939707149,
"grad_norm": 0.33789893984794617,
"learning_rate": 0.0004208603743099501,
"loss": 3.1506,
"step": 44450
},
{
"epoch": 11.97545219638243,
"grad_norm": 0.3334933817386627,
"learning_rate": 0.00042065840850949237,
"loss": 3.1528,
"step": 44500
},
{
"epoch": 11.988910422049956,
"grad_norm": 0.33397191762924194,
"learning_rate": 0.00042045644270903457,
"loss": 3.1525,
"step": 44550
},
{
"epoch": 12.002153316106805,
"grad_norm": 0.3530355393886566,
"learning_rate": 0.00042025447690857676,
"loss": 3.1187,
"step": 44600
},
{
"epoch": 12.015611541774332,
"grad_norm": 0.36705562472343445,
"learning_rate": 0.00042005251110811896,
"loss": 3.0554,
"step": 44650
},
{
"epoch": 12.029069767441861,
"grad_norm": 0.33843058347702026,
"learning_rate": 0.00041985054530766115,
"loss": 3.0531,
"step": 44700
},
{
"epoch": 12.042527993109388,
"grad_norm": 0.35231149196624756,
"learning_rate": 0.0004196485795072034,
"loss": 3.0547,
"step": 44750
},
{
"epoch": 12.055986218776916,
"grad_norm": 0.36499878764152527,
"learning_rate": 0.0004194466137067456,
"loss": 3.049,
"step": 44800
},
{
"epoch": 12.069444444444445,
"grad_norm": 0.353890597820282,
"learning_rate": 0.0004192446479062878,
"loss": 3.0625,
"step": 44850
},
{
"epoch": 12.082902670111972,
"grad_norm": 0.35161858797073364,
"learning_rate": 0.00041904268210583005,
"loss": 3.0611,
"step": 44900
},
{
"epoch": 12.096360895779501,
"grad_norm": 0.37081584334373474,
"learning_rate": 0.0004188407163053723,
"loss": 3.0695,
"step": 44950
},
{
"epoch": 12.109819121447028,
"grad_norm": 0.34427958726882935,
"learning_rate": 0.0004186387505049145,
"loss": 3.0648,
"step": 45000
},
{
"epoch": 12.109819121447028,
"eval_accuracy": 0.3916719143097904,
"eval_loss": 3.307880163192749,
"eval_runtime": 183.7859,
"eval_samples_per_second": 98.005,
"eval_steps_per_second": 6.127,
"step": 45000
},
{
"epoch": 12.123277347114556,
"grad_norm": 0.3758573532104492,
"learning_rate": 0.0004184367847044567,
"loss": 3.0672,
"step": 45050
},
{
"epoch": 12.136735572782085,
"grad_norm": 0.3592199981212616,
"learning_rate": 0.0004182348189039989,
"loss": 3.0681,
"step": 45100
},
{
"epoch": 12.150193798449612,
"grad_norm": 0.35564643144607544,
"learning_rate": 0.00041803285310354114,
"loss": 3.0757,
"step": 45150
},
{
"epoch": 12.163652024117141,
"grad_norm": 0.3589390516281128,
"learning_rate": 0.00041783088730308333,
"loss": 3.0875,
"step": 45200
},
{
"epoch": 12.177110249784668,
"grad_norm": 0.33536219596862793,
"learning_rate": 0.00041762892150262553,
"loss": 3.0783,
"step": 45250
},
{
"epoch": 12.190568475452196,
"grad_norm": 0.35109660029411316,
"learning_rate": 0.0004174269557021677,
"loss": 3.0745,
"step": 45300
},
{
"epoch": 12.204026701119725,
"grad_norm": 0.3556494116783142,
"learning_rate": 0.0004172249899017099,
"loss": 3.0809,
"step": 45350
},
{
"epoch": 12.217484926787252,
"grad_norm": 0.3466927409172058,
"learning_rate": 0.00041702302410125217,
"loss": 3.0877,
"step": 45400
},
{
"epoch": 12.230943152454781,
"grad_norm": 0.34194013476371765,
"learning_rate": 0.00041682105830079437,
"loss": 3.0852,
"step": 45450
},
{
"epoch": 12.244401378122308,
"grad_norm": 0.3396938443183899,
"learning_rate": 0.00041661909250033656,
"loss": 3.089,
"step": 45500
},
{
"epoch": 12.257859603789836,
"grad_norm": 0.3699965476989746,
"learning_rate": 0.00041641712669987876,
"loss": 3.0754,
"step": 45550
},
{
"epoch": 12.271317829457365,
"grad_norm": 0.36619871854782104,
"learning_rate": 0.00041621516089942096,
"loss": 3.0827,
"step": 45600
},
{
"epoch": 12.284776055124892,
"grad_norm": 0.35216283798217773,
"learning_rate": 0.0004160131950989632,
"loss": 3.0976,
"step": 45650
},
{
"epoch": 12.298234280792421,
"grad_norm": 0.3299254775047302,
"learning_rate": 0.0004158112292985054,
"loss": 3.0862,
"step": 45700
},
{
"epoch": 12.311692506459949,
"grad_norm": 0.3577413856983185,
"learning_rate": 0.0004156092634980476,
"loss": 3.0911,
"step": 45750
},
{
"epoch": 12.325150732127476,
"grad_norm": 0.3620416522026062,
"learning_rate": 0.00041540729769758985,
"loss": 3.0936,
"step": 45800
},
{
"epoch": 12.338608957795005,
"grad_norm": 0.3336474895477295,
"learning_rate": 0.0004152053318971321,
"loss": 3.1046,
"step": 45850
},
{
"epoch": 12.352067183462532,
"grad_norm": 0.3471616804599762,
"learning_rate": 0.0004150033660966743,
"loss": 3.0941,
"step": 45900
},
{
"epoch": 12.36552540913006,
"grad_norm": 0.35189089179039,
"learning_rate": 0.0004148014002962165,
"loss": 3.1027,
"step": 45950
},
{
"epoch": 12.378983634797589,
"grad_norm": 0.351576030254364,
"learning_rate": 0.0004145994344957587,
"loss": 3.0998,
"step": 46000
},
{
"epoch": 12.378983634797589,
"eval_accuracy": 0.39206771480461167,
"eval_loss": 3.3011374473571777,
"eval_runtime": 183.8406,
"eval_samples_per_second": 97.976,
"eval_steps_per_second": 6.125,
"step": 46000
},
{
"epoch": 12.392441860465116,
"grad_norm": 0.3573558032512665,
"learning_rate": 0.0004143974686953009,
"loss": 3.1038,
"step": 46050
},
{
"epoch": 12.405900086132645,
"grad_norm": 0.3651130497455597,
"learning_rate": 0.00041419550289484313,
"loss": 3.0936,
"step": 46100
},
{
"epoch": 12.419358311800172,
"grad_norm": 0.3354813754558563,
"learning_rate": 0.00041399353709438533,
"loss": 3.1008,
"step": 46150
},
{
"epoch": 12.4328165374677,
"grad_norm": 0.3393497169017792,
"learning_rate": 0.0004137915712939275,
"loss": 3.1074,
"step": 46200
},
{
"epoch": 12.446274763135229,
"grad_norm": 0.33397138118743896,
"learning_rate": 0.0004135896054934697,
"loss": 3.1051,
"step": 46250
},
{
"epoch": 12.459732988802756,
"grad_norm": 0.34602561593055725,
"learning_rate": 0.0004133876396930119,
"loss": 3.1112,
"step": 46300
},
{
"epoch": 12.473191214470285,
"grad_norm": 0.38050827383995056,
"learning_rate": 0.00041318567389255417,
"loss": 3.0961,
"step": 46350
},
{
"epoch": 12.486649440137812,
"grad_norm": 0.38827836513519287,
"learning_rate": 0.00041298370809209636,
"loss": 3.1052,
"step": 46400
},
{
"epoch": 12.50010766580534,
"grad_norm": 0.3689483404159546,
"learning_rate": 0.00041278174229163856,
"loss": 3.1011,
"step": 46450
},
{
"epoch": 12.513565891472869,
"grad_norm": 0.34539344906806946,
"learning_rate": 0.00041257977649118076,
"loss": 3.1089,
"step": 46500
},
{
"epoch": 12.527024117140396,
"grad_norm": 0.3866020739078522,
"learning_rate": 0.00041237781069072295,
"loss": 3.1012,
"step": 46550
},
{
"epoch": 12.540482342807923,
"grad_norm": 0.3452168405056,
"learning_rate": 0.0004121758448902652,
"loss": 3.1037,
"step": 46600
},
{
"epoch": 12.553940568475452,
"grad_norm": 0.3894827961921692,
"learning_rate": 0.0004119738790898074,
"loss": 3.1067,
"step": 46650
},
{
"epoch": 12.56739879414298,
"grad_norm": 0.38671883940696716,
"learning_rate": 0.00041177191328934965,
"loss": 3.1124,
"step": 46700
},
{
"epoch": 12.580857019810509,
"grad_norm": 0.3547767698764801,
"learning_rate": 0.00041156994748889185,
"loss": 3.1027,
"step": 46750
},
{
"epoch": 12.594315245478036,
"grad_norm": 0.36329013109207153,
"learning_rate": 0.0004113679816884341,
"loss": 3.1206,
"step": 46800
},
{
"epoch": 12.607773471145563,
"grad_norm": 0.36048048734664917,
"learning_rate": 0.0004111660158879763,
"loss": 3.1224,
"step": 46850
},
{
"epoch": 12.621231696813092,
"grad_norm": 0.3339656591415405,
"learning_rate": 0.0004109640500875185,
"loss": 3.1161,
"step": 46900
},
{
"epoch": 12.63468992248062,
"grad_norm": 0.3573682904243469,
"learning_rate": 0.0004107620842870607,
"loss": 3.1049,
"step": 46950
},
{
"epoch": 12.648148148148149,
"grad_norm": 0.35174158215522766,
"learning_rate": 0.0004105601184866029,
"loss": 3.1062,
"step": 47000
},
{
"epoch": 12.648148148148149,
"eval_accuracy": 0.39254891172790657,
"eval_loss": 3.297520637512207,
"eval_runtime": 183.9722,
"eval_samples_per_second": 97.906,
"eval_steps_per_second": 6.12,
"step": 47000
},
{
"epoch": 12.661606373815676,
"grad_norm": 0.3673131465911865,
"learning_rate": 0.00041035815268614513,
"loss": 3.1182,
"step": 47050
},
{
"epoch": 12.675064599483203,
"grad_norm": 0.3440827429294586,
"learning_rate": 0.0004101561868856873,
"loss": 3.1132,
"step": 47100
},
{
"epoch": 12.688522825150732,
"grad_norm": 0.36985883116722107,
"learning_rate": 0.0004099542210852295,
"loss": 3.1174,
"step": 47150
},
{
"epoch": 12.70198105081826,
"grad_norm": 0.3606552481651306,
"learning_rate": 0.0004097522552847717,
"loss": 3.1067,
"step": 47200
},
{
"epoch": 12.715439276485789,
"grad_norm": 0.3608570694923401,
"learning_rate": 0.00040955028948431397,
"loss": 3.1114,
"step": 47250
},
{
"epoch": 12.728897502153316,
"grad_norm": 0.34414955973625183,
"learning_rate": 0.00040934832368385616,
"loss": 3.1153,
"step": 47300
},
{
"epoch": 12.742355727820843,
"grad_norm": 0.35615408420562744,
"learning_rate": 0.00040914635788339836,
"loss": 3.1251,
"step": 47350
},
{
"epoch": 12.755813953488373,
"grad_norm": 0.34484538435935974,
"learning_rate": 0.00040894439208294056,
"loss": 3.1217,
"step": 47400
},
{
"epoch": 12.7692721791559,
"grad_norm": 0.3897562026977539,
"learning_rate": 0.00040874242628248275,
"loss": 3.1241,
"step": 47450
},
{
"epoch": 12.782730404823429,
"grad_norm": 0.3534804582595825,
"learning_rate": 0.000408540460482025,
"loss": 3.1166,
"step": 47500
},
{
"epoch": 12.796188630490956,
"grad_norm": 0.3697199821472168,
"learning_rate": 0.00040833849468156725,
"loss": 3.1204,
"step": 47550
},
{
"epoch": 12.809646856158484,
"grad_norm": 0.3332458436489105,
"learning_rate": 0.00040813652888110945,
"loss": 3.114,
"step": 47600
},
{
"epoch": 12.823105081826013,
"grad_norm": 0.34950539469718933,
"learning_rate": 0.00040793456308065165,
"loss": 3.121,
"step": 47650
},
{
"epoch": 12.83656330749354,
"grad_norm": 0.33315515518188477,
"learning_rate": 0.0004077325972801939,
"loss": 3.1128,
"step": 47700
},
{
"epoch": 12.850021533161069,
"grad_norm": 0.37423792481422424,
"learning_rate": 0.0004075306314797361,
"loss": 3.1272,
"step": 47750
},
{
"epoch": 12.863479758828596,
"grad_norm": 0.3385739326477051,
"learning_rate": 0.0004073286656792783,
"loss": 3.129,
"step": 47800
},
{
"epoch": 12.876937984496124,
"grad_norm": 0.355631947517395,
"learning_rate": 0.0004071266998788205,
"loss": 3.1298,
"step": 47850
},
{
"epoch": 12.890396210163653,
"grad_norm": 0.3601670563220978,
"learning_rate": 0.0004069247340783627,
"loss": 3.1134,
"step": 47900
},
{
"epoch": 12.90385443583118,
"grad_norm": 0.36563220620155334,
"learning_rate": 0.00040672276827790493,
"loss": 3.132,
"step": 47950
},
{
"epoch": 12.917312661498707,
"grad_norm": 0.32263684272766113,
"learning_rate": 0.0004065208024774471,
"loss": 3.1273,
"step": 48000
},
{
"epoch": 12.917312661498707,
"eval_accuracy": 0.3929590536076623,
"eval_loss": 3.2909340858459473,
"eval_runtime": 183.8937,
"eval_samples_per_second": 97.948,
"eval_steps_per_second": 6.123,
"step": 48000
},
{
"epoch": 12.930770887166236,
"grad_norm": 0.3488883376121521,
"learning_rate": 0.0004063188366769893,
"loss": 3.1309,
"step": 48050
},
{
"epoch": 12.944229112833764,
"grad_norm": 0.35998743772506714,
"learning_rate": 0.0004061168708765315,
"loss": 3.1254,
"step": 48100
},
{
"epoch": 12.957687338501293,
"grad_norm": 0.3694109320640564,
"learning_rate": 0.0004059149050760737,
"loss": 3.1271,
"step": 48150
},
{
"epoch": 12.97114556416882,
"grad_norm": 0.3622800409793854,
"learning_rate": 0.00040571293927561597,
"loss": 3.1377,
"step": 48200
},
{
"epoch": 12.984603789836347,
"grad_norm": 0.3408743441104889,
"learning_rate": 0.00040551097347515816,
"loss": 3.1271,
"step": 48250
},
{
"epoch": 12.998062015503876,
"grad_norm": 0.34108561277389526,
"learning_rate": 0.00040530900767470036,
"loss": 3.1202,
"step": 48300
},
{
"epoch": 13.011304909560723,
"grad_norm": 0.365139901638031,
"learning_rate": 0.00040510704187424255,
"loss": 3.0547,
"step": 48350
},
{
"epoch": 13.024763135228252,
"grad_norm": 0.34027576446533203,
"learning_rate": 0.00040490507607378475,
"loss": 3.0415,
"step": 48400
},
{
"epoch": 13.038221360895779,
"grad_norm": 0.37130045890808105,
"learning_rate": 0.00040470311027332705,
"loss": 3.0304,
"step": 48450
},
{
"epoch": 13.051679586563308,
"grad_norm": 0.350969135761261,
"learning_rate": 0.00040450114447286925,
"loss": 3.0364,
"step": 48500
},
{
"epoch": 13.065137812230835,
"grad_norm": 0.3762068748474121,
"learning_rate": 0.00040429917867241145,
"loss": 3.0457,
"step": 48550
},
{
"epoch": 13.078596037898363,
"grad_norm": 0.3403458893299103,
"learning_rate": 0.00040409721287195364,
"loss": 3.0448,
"step": 48600
},
{
"epoch": 13.092054263565892,
"grad_norm": 0.34854280948638916,
"learning_rate": 0.0004038952470714959,
"loss": 3.0489,
"step": 48650
},
{
"epoch": 13.10551248923342,
"grad_norm": 0.3887944519519806,
"learning_rate": 0.0004036932812710381,
"loss": 3.0439,
"step": 48700
},
{
"epoch": 13.118970714900948,
"grad_norm": 0.3447207808494568,
"learning_rate": 0.0004034913154705803,
"loss": 3.0501,
"step": 48750
},
{
"epoch": 13.132428940568476,
"grad_norm": 0.3509175181388855,
"learning_rate": 0.0004032893496701225,
"loss": 3.0661,
"step": 48800
},
{
"epoch": 13.145887166236003,
"grad_norm": 0.3901561498641968,
"learning_rate": 0.0004030873838696647,
"loss": 3.0592,
"step": 48850
},
{
"epoch": 13.159345391903532,
"grad_norm": 0.36739397048950195,
"learning_rate": 0.00040288541806920693,
"loss": 3.0648,
"step": 48900
},
{
"epoch": 13.17280361757106,
"grad_norm": 0.3569977879524231,
"learning_rate": 0.0004026834522687491,
"loss": 3.0625,
"step": 48950
},
{
"epoch": 13.186261843238588,
"grad_norm": 0.38752639293670654,
"learning_rate": 0.0004024814864682913,
"loss": 3.0699,
"step": 49000
},
{
"epoch": 13.186261843238588,
"eval_accuracy": 0.39263387356895796,
"eval_loss": 3.303571939468384,
"eval_runtime": 183.8613,
"eval_samples_per_second": 97.965,
"eval_steps_per_second": 6.124,
"step": 49000
},
{
"epoch": 13.199720068906116,
"grad_norm": 0.34880366921424866,
"learning_rate": 0.0004022795206678335,
"loss": 3.0544,
"step": 49050
},
{
"epoch": 13.213178294573643,
"grad_norm": 0.38485971093177795,
"learning_rate": 0.00040207755486737577,
"loss": 3.0681,
"step": 49100
},
{
"epoch": 13.226636520241172,
"grad_norm": 0.3616187870502472,
"learning_rate": 0.00040187558906691796,
"loss": 3.0767,
"step": 49150
},
{
"epoch": 13.2400947459087,
"grad_norm": 0.35451483726501465,
"learning_rate": 0.00040167362326646016,
"loss": 3.073,
"step": 49200
},
{
"epoch": 13.253552971576227,
"grad_norm": 0.35461002588272095,
"learning_rate": 0.00040147165746600235,
"loss": 3.0683,
"step": 49250
},
{
"epoch": 13.267011197243756,
"grad_norm": 0.37109312415122986,
"learning_rate": 0.00040126969166554455,
"loss": 3.0666,
"step": 49300
},
{
"epoch": 13.280469422911283,
"grad_norm": 0.3642289340496063,
"learning_rate": 0.00040106772586508686,
"loss": 3.0707,
"step": 49350
},
{
"epoch": 13.293927648578812,
"grad_norm": 0.3483692705631256,
"learning_rate": 0.00040086576006462905,
"loss": 3.0662,
"step": 49400
},
{
"epoch": 13.30738587424634,
"grad_norm": 0.3468538224697113,
"learning_rate": 0.00040066379426417125,
"loss": 3.0761,
"step": 49450
},
{
"epoch": 13.320844099913867,
"grad_norm": 0.3627040386199951,
"learning_rate": 0.00040046182846371344,
"loss": 3.073,
"step": 49500
},
{
"epoch": 13.334302325581396,
"grad_norm": 0.3752869963645935,
"learning_rate": 0.0004002598626632557,
"loss": 3.0722,
"step": 49550
},
{
"epoch": 13.347760551248923,
"grad_norm": 0.3828841745853424,
"learning_rate": 0.0004000578968627979,
"loss": 3.0857,
"step": 49600
},
{
"epoch": 13.361218776916452,
"grad_norm": 0.374611496925354,
"learning_rate": 0.0003998559310623401,
"loss": 3.0783,
"step": 49650
},
{
"epoch": 13.37467700258398,
"grad_norm": 0.3917146921157837,
"learning_rate": 0.0003996539652618823,
"loss": 3.0901,
"step": 49700
},
{
"epoch": 13.388135228251507,
"grad_norm": 0.3606676757335663,
"learning_rate": 0.0003994519994614245,
"loss": 3.0853,
"step": 49750
},
{
"epoch": 13.401593453919036,
"grad_norm": 0.35497403144836426,
"learning_rate": 0.00039925003366096673,
"loss": 3.0961,
"step": 49800
},
{
"epoch": 13.415051679586563,
"grad_norm": 0.40613147616386414,
"learning_rate": 0.0003990480678605089,
"loss": 3.0928,
"step": 49850
},
{
"epoch": 13.428509905254092,
"grad_norm": 0.38314661383628845,
"learning_rate": 0.0003988461020600511,
"loss": 3.0763,
"step": 49900
},
{
"epoch": 13.44196813092162,
"grad_norm": 0.3707086145877838,
"learning_rate": 0.0003986441362595933,
"loss": 3.0874,
"step": 49950
},
{
"epoch": 13.455426356589147,
"grad_norm": 0.3763015866279602,
"learning_rate": 0.0003984421704591355,
"loss": 3.0718,
"step": 50000
},
{
"epoch": 13.455426356589147,
"eval_accuracy": 0.39262105324000135,
"eval_loss": 3.300663471221924,
"eval_runtime": 183.9615,
"eval_samples_per_second": 97.912,
"eval_steps_per_second": 6.121,
"step": 50000
},
{
"epoch": 13.468884582256676,
"grad_norm": 0.3490633964538574,
"learning_rate": 0.00039824020465867776,
"loss": 3.0966,
"step": 50050
},
{
"epoch": 13.482342807924203,
"grad_norm": 0.3602553904056549,
"learning_rate": 0.00039803823885821996,
"loss": 3.0911,
"step": 50100
},
{
"epoch": 13.49580103359173,
"grad_norm": 0.37203365564346313,
"learning_rate": 0.00039783627305776216,
"loss": 3.0932,
"step": 50150
},
{
"epoch": 13.50925925925926,
"grad_norm": 0.36460578441619873,
"learning_rate": 0.00039763430725730435,
"loss": 3.0864,
"step": 50200
},
{
"epoch": 13.522717484926787,
"grad_norm": 0.36982154846191406,
"learning_rate": 0.00039743234145684666,
"loss": 3.0933,
"step": 50250
},
{
"epoch": 13.536175710594316,
"grad_norm": 0.38517898321151733,
"learning_rate": 0.00039723037565638885,
"loss": 3.096,
"step": 50300
},
{
"epoch": 13.549633936261843,
"grad_norm": 0.3811021149158478,
"learning_rate": 0.00039702840985593105,
"loss": 3.0918,
"step": 50350
},
{
"epoch": 13.56309216192937,
"grad_norm": 0.38058093190193176,
"learning_rate": 0.00039682644405547324,
"loss": 3.0945,
"step": 50400
},
{
"epoch": 13.5765503875969,
"grad_norm": 0.3621980547904968,
"learning_rate": 0.00039662447825501544,
"loss": 3.0978,
"step": 50450
},
{
"epoch": 13.590008613264427,
"grad_norm": 0.3358357846736908,
"learning_rate": 0.0003964225124545577,
"loss": 3.0968,
"step": 50500
},
{
"epoch": 13.603466838931956,
"grad_norm": 0.36395880579948425,
"learning_rate": 0.0003962205466540999,
"loss": 3.094,
"step": 50550
},
{
"epoch": 13.616925064599483,
"grad_norm": 0.36760833859443665,
"learning_rate": 0.0003960185808536421,
"loss": 3.0976,
"step": 50600
},
{
"epoch": 13.63038329026701,
"grad_norm": 0.3504054546356201,
"learning_rate": 0.0003958166150531843,
"loss": 3.0889,
"step": 50650
},
{
"epoch": 13.64384151593454,
"grad_norm": 0.3391190767288208,
"learning_rate": 0.0003956146492527265,
"loss": 3.0961,
"step": 50700
},
{
"epoch": 13.657299741602067,
"grad_norm": 0.3838180899620056,
"learning_rate": 0.0003954126834522687,
"loss": 3.1037,
"step": 50750
},
{
"epoch": 13.670757967269594,
"grad_norm": 0.341469407081604,
"learning_rate": 0.0003952107176518109,
"loss": 3.0979,
"step": 50800
},
{
"epoch": 13.684216192937123,
"grad_norm": 0.37517330050468445,
"learning_rate": 0.0003950087518513531,
"loss": 3.1003,
"step": 50850
},
{
"epoch": 13.69767441860465,
"grad_norm": 0.3640974462032318,
"learning_rate": 0.0003948067860508953,
"loss": 3.1033,
"step": 50900
},
{
"epoch": 13.71113264427218,
"grad_norm": 0.34508809447288513,
"learning_rate": 0.0003946048202504375,
"loss": 3.091,
"step": 50950
},
{
"epoch": 13.724590869939707,
"grad_norm": 0.3852170407772064,
"learning_rate": 0.00039440285444997976,
"loss": 3.0956,
"step": 51000
},
{
"epoch": 13.724590869939707,
"eval_accuracy": 0.3932995528530012,
"eval_loss": 3.2934370040893555,
"eval_runtime": 183.9981,
"eval_samples_per_second": 97.892,
"eval_steps_per_second": 6.12,
"step": 51000
},
{
"epoch": 13.738049095607234,
"grad_norm": 0.33660662174224854,
"learning_rate": 0.00039420088864952196,
"loss": 3.1005,
"step": 51050
},
{
"epoch": 13.751507321274763,
"grad_norm": 0.3775493800640106,
"learning_rate": 0.00039399892284906415,
"loss": 3.1003,
"step": 51100
},
{
"epoch": 13.76496554694229,
"grad_norm": 0.4018406867980957,
"learning_rate": 0.00039379695704860646,
"loss": 3.1038,
"step": 51150
},
{
"epoch": 13.77842377260982,
"grad_norm": 0.34795209765434265,
"learning_rate": 0.00039359499124814865,
"loss": 3.1119,
"step": 51200
},
{
"epoch": 13.791881998277347,
"grad_norm": 0.3505890369415283,
"learning_rate": 0.00039339302544769085,
"loss": 3.1075,
"step": 51250
},
{
"epoch": 13.805340223944874,
"grad_norm": 0.32561638951301575,
"learning_rate": 0.00039319105964723305,
"loss": 3.1021,
"step": 51300
},
{
"epoch": 13.818798449612403,
"grad_norm": 0.33596768975257874,
"learning_rate": 0.00039298909384677524,
"loss": 3.112,
"step": 51350
},
{
"epoch": 13.83225667527993,
"grad_norm": 0.357796311378479,
"learning_rate": 0.0003927871280463175,
"loss": 3.1032,
"step": 51400
},
{
"epoch": 13.84571490094746,
"grad_norm": 0.3593471646308899,
"learning_rate": 0.0003925851622458597,
"loss": 3.1012,
"step": 51450
},
{
"epoch": 13.859173126614987,
"grad_norm": 0.40703701972961426,
"learning_rate": 0.0003923831964454019,
"loss": 3.1029,
"step": 51500
},
{
"epoch": 13.872631352282514,
"grad_norm": 0.3672133982181549,
"learning_rate": 0.0003921812306449441,
"loss": 3.1098,
"step": 51550
},
{
"epoch": 13.886089577950044,
"grad_norm": 0.3748154640197754,
"learning_rate": 0.0003919792648444863,
"loss": 3.1061,
"step": 51600
},
{
"epoch": 13.89954780361757,
"grad_norm": 0.3705759346485138,
"learning_rate": 0.0003917772990440285,
"loss": 3.1021,
"step": 51650
},
{
"epoch": 13.9130060292851,
"grad_norm": 0.342912882566452,
"learning_rate": 0.0003915753332435707,
"loss": 3.109,
"step": 51700
},
{
"epoch": 13.926464254952627,
"grad_norm": 0.3617236614227295,
"learning_rate": 0.0003913733674431129,
"loss": 3.1163,
"step": 51750
},
{
"epoch": 13.939922480620154,
"grad_norm": 0.3600302040576935,
"learning_rate": 0.0003911714016426551,
"loss": 3.109,
"step": 51800
},
{
"epoch": 13.953380706287684,
"grad_norm": 0.36543717980384827,
"learning_rate": 0.0003909694358421973,
"loss": 3.1125,
"step": 51850
},
{
"epoch": 13.96683893195521,
"grad_norm": 0.3439505994319916,
"learning_rate": 0.00039076747004173956,
"loss": 3.1121,
"step": 51900
},
{
"epoch": 13.98029715762274,
"grad_norm": 0.3901199400424957,
"learning_rate": 0.00039056550424128176,
"loss": 3.1179,
"step": 51950
},
{
"epoch": 13.993755383290267,
"grad_norm": 0.3661282956600189,
"learning_rate": 0.00039036353844082395,
"loss": 3.1265,
"step": 52000
},
{
"epoch": 13.993755383290267,
"eval_accuracy": 0.39374685195736003,
"eval_loss": 3.284869909286499,
"eval_runtime": 183.911,
"eval_samples_per_second": 97.939,
"eval_steps_per_second": 6.123,
"step": 52000
},
{
"epoch": 14.006998277347115,
"grad_norm": 0.3538947105407715,
"learning_rate": 0.0003901615726403662,
"loss": 3.0522,
"step": 52050
},
{
"epoch": 14.020456503014643,
"grad_norm": 0.34528523683547974,
"learning_rate": 0.00038995960683990845,
"loss": 3.0139,
"step": 52100
},
{
"epoch": 14.03391472868217,
"grad_norm": 0.3903304636478424,
"learning_rate": 0.00038975764103945065,
"loss": 3.0158,
"step": 52150
},
{
"epoch": 14.047372954349699,
"grad_norm": 0.3508067727088928,
"learning_rate": 0.00038955567523899285,
"loss": 3.0219,
"step": 52200
},
{
"epoch": 14.060831180017226,
"grad_norm": 0.3663598299026489,
"learning_rate": 0.00038935370943853504,
"loss": 3.0237,
"step": 52250
},
{
"epoch": 14.074289405684755,
"grad_norm": 0.3504553437232971,
"learning_rate": 0.00038915174363807724,
"loss": 3.0318,
"step": 52300
},
{
"epoch": 14.087747631352283,
"grad_norm": 0.3819941580295563,
"learning_rate": 0.0003889497778376195,
"loss": 3.0417,
"step": 52350
},
{
"epoch": 14.10120585701981,
"grad_norm": 0.3762374222278595,
"learning_rate": 0.0003887478120371617,
"loss": 3.0363,
"step": 52400
},
{
"epoch": 14.114664082687339,
"grad_norm": 0.37316834926605225,
"learning_rate": 0.0003885458462367039,
"loss": 3.0483,
"step": 52450
},
{
"epoch": 14.128122308354866,
"grad_norm": 0.3446917235851288,
"learning_rate": 0.0003883438804362461,
"loss": 3.0346,
"step": 52500
},
{
"epoch": 14.141580534022394,
"grad_norm": 0.34739944338798523,
"learning_rate": 0.0003881419146357883,
"loss": 3.0393,
"step": 52550
},
{
"epoch": 14.155038759689923,
"grad_norm": 0.37395405769348145,
"learning_rate": 0.0003879399488353305,
"loss": 3.043,
"step": 52600
},
{
"epoch": 14.16849698535745,
"grad_norm": 0.3630558252334595,
"learning_rate": 0.0003877379830348727,
"loss": 3.039,
"step": 52650
},
{
"epoch": 14.18195521102498,
"grad_norm": 0.3915100395679474,
"learning_rate": 0.0003875360172344149,
"loss": 3.0492,
"step": 52700
},
{
"epoch": 14.195413436692506,
"grad_norm": 0.3782561123371124,
"learning_rate": 0.0003873340514339571,
"loss": 3.048,
"step": 52750
},
{
"epoch": 14.208871662360034,
"grad_norm": 0.3983249366283417,
"learning_rate": 0.0003871320856334993,
"loss": 3.0516,
"step": 52800
},
{
"epoch": 14.222329888027563,
"grad_norm": 0.35647591948509216,
"learning_rate": 0.00038693011983304156,
"loss": 3.0449,
"step": 52850
},
{
"epoch": 14.23578811369509,
"grad_norm": 0.35614535212516785,
"learning_rate": 0.0003867281540325838,
"loss": 3.0592,
"step": 52900
},
{
"epoch": 14.24924633936262,
"grad_norm": 0.3985713720321655,
"learning_rate": 0.000386526188232126,
"loss": 3.0614,
"step": 52950
},
{
"epoch": 14.262704565030146,
"grad_norm": 0.3639850616455078,
"learning_rate": 0.00038632422243166825,
"loss": 3.0532,
"step": 53000
},
{
"epoch": 14.262704565030146,
"eval_accuracy": 0.3930079446926663,
"eval_loss": 3.3003249168395996,
"eval_runtime": 184.0533,
"eval_samples_per_second": 97.863,
"eval_steps_per_second": 6.118,
"step": 53000
},
{
"epoch": 14.276162790697674,
"grad_norm": 0.36184415221214294,
"learning_rate": 0.00038612225663121045,
"loss": 3.0586,
"step": 53050
},
{
"epoch": 14.289621016365203,
"grad_norm": 0.3654453754425049,
"learning_rate": 0.00038592029083075265,
"loss": 3.0622,
"step": 53100
},
{
"epoch": 14.30307924203273,
"grad_norm": 0.36275461316108704,
"learning_rate": 0.00038571832503029484,
"loss": 3.063,
"step": 53150
},
{
"epoch": 14.31653746770026,
"grad_norm": 0.3564082980155945,
"learning_rate": 0.00038551635922983704,
"loss": 3.0644,
"step": 53200
},
{
"epoch": 14.329995693367787,
"grad_norm": Infinity,
"learning_rate": 0.0003853143934293793,
"loss": 3.0645,
"step": 53250
},
{
"epoch": 14.343453919035314,
"grad_norm": 0.3679307997226715,
"learning_rate": 0.0003851124276289215,
"loss": 3.0655,
"step": 53300
},
{
"epoch": 14.356912144702843,
"grad_norm": 0.3395625948905945,
"learning_rate": 0.0003849104618284637,
"loss": 3.0609,
"step": 53350
},
{
"epoch": 14.37037037037037,
"grad_norm": 0.3531416654586792,
"learning_rate": 0.0003847084960280059,
"loss": 3.0669,
"step": 53400
},
{
"epoch": 14.383828596037898,
"grad_norm": 0.36635661125183105,
"learning_rate": 0.0003845065302275481,
"loss": 3.062,
"step": 53450
},
{
"epoch": 14.397286821705427,
"grad_norm": 0.36860954761505127,
"learning_rate": 0.0003843045644270903,
"loss": 3.0786,
"step": 53500
},
{
"epoch": 14.410745047372954,
"grad_norm": 0.36681869626045227,
"learning_rate": 0.0003841025986266325,
"loss": 3.0548,
"step": 53550
},
{
"epoch": 14.424203273040483,
"grad_norm": 0.34156012535095215,
"learning_rate": 0.0003839006328261747,
"loss": 3.0776,
"step": 53600
},
{
"epoch": 14.43766149870801,
"grad_norm": 0.39001914858818054,
"learning_rate": 0.0003836986670257169,
"loss": 3.0654,
"step": 53650
},
{
"epoch": 14.451119724375538,
"grad_norm": 0.3754884898662567,
"learning_rate": 0.0003834967012252591,
"loss": 3.0707,
"step": 53700
},
{
"epoch": 14.464577950043067,
"grad_norm": 0.36852484941482544,
"learning_rate": 0.00038329473542480136,
"loss": 3.071,
"step": 53750
},
{
"epoch": 14.478036175710594,
"grad_norm": 0.3700765371322632,
"learning_rate": 0.0003830927696243436,
"loss": 3.0744,
"step": 53800
},
{
"epoch": 14.491494401378123,
"grad_norm": 0.3541344404220581,
"learning_rate": 0.0003828908038238858,
"loss": 3.0862,
"step": 53850
},
{
"epoch": 14.50495262704565,
"grad_norm": 0.3627966344356537,
"learning_rate": 0.000382688838023428,
"loss": 3.0773,
"step": 53900
},
{
"epoch": 14.518410852713178,
"grad_norm": 0.368941992521286,
"learning_rate": 0.00038248687222297025,
"loss": 3.081,
"step": 53950
},
{
"epoch": 14.531869078380707,
"grad_norm": 0.3832266926765442,
"learning_rate": 0.00038228490642251245,
"loss": 3.0728,
"step": 54000
},
{
"epoch": 14.531869078380707,
"eval_accuracy": 0.3934765385807157,
"eval_loss": 3.294849157333374,
"eval_runtime": 183.8284,
"eval_samples_per_second": 97.983,
"eval_steps_per_second": 6.125,
"step": 54000
},
{
"epoch": 14.545327304048234,
"grad_norm": 0.3720203936100006,
"learning_rate": 0.00038208294062205464,
"loss": 3.0701,
"step": 54050
},
{
"epoch": 14.558785529715763,
"grad_norm": 0.3722517788410187,
"learning_rate": 0.00038188097482159684,
"loss": 3.0767,
"step": 54100
},
{
"epoch": 14.57224375538329,
"grad_norm": 0.3482821583747864,
"learning_rate": 0.00038167900902113904,
"loss": 3.0919,
"step": 54150
},
{
"epoch": 14.585701981050818,
"grad_norm": 0.3424507677555084,
"learning_rate": 0.0003814770432206813,
"loss": 3.0853,
"step": 54200
},
{
"epoch": 14.599160206718347,
"grad_norm": 0.39158084988594055,
"learning_rate": 0.0003812750774202235,
"loss": 3.0824,
"step": 54250
},
{
"epoch": 14.612618432385874,
"grad_norm": 0.36076387763023376,
"learning_rate": 0.0003810731116197657,
"loss": 3.084,
"step": 54300
},
{
"epoch": 14.626076658053403,
"grad_norm": 0.3608371317386627,
"learning_rate": 0.0003808711458193079,
"loss": 3.0772,
"step": 54350
},
{
"epoch": 14.63953488372093,
"grad_norm": 0.36772799491882324,
"learning_rate": 0.00038066918001885007,
"loss": 3.0876,
"step": 54400
},
{
"epoch": 14.652993109388458,
"grad_norm": 0.3680662512779236,
"learning_rate": 0.0003804672142183923,
"loss": 3.0945,
"step": 54450
},
{
"epoch": 14.666451335055987,
"grad_norm": 0.3548770248889923,
"learning_rate": 0.0003802652484179345,
"loss": 3.0962,
"step": 54500
},
{
"epoch": 14.679909560723514,
"grad_norm": 0.37391746044158936,
"learning_rate": 0.0003800632826174767,
"loss": 3.0828,
"step": 54550
},
{
"epoch": 14.693367786391041,
"grad_norm": 0.3779378831386566,
"learning_rate": 0.0003798613168170189,
"loss": 3.0861,
"step": 54600
},
{
"epoch": 14.70682601205857,
"grad_norm": 0.3715170919895172,
"learning_rate": 0.0003796593510165611,
"loss": 3.0835,
"step": 54650
},
{
"epoch": 14.720284237726098,
"grad_norm": 0.3467620313167572,
"learning_rate": 0.0003794573852161034,
"loss": 3.0864,
"step": 54700
},
{
"epoch": 14.733742463393627,
"grad_norm": 0.43162673711776733,
"learning_rate": 0.0003792554194156456,
"loss": 3.0866,
"step": 54750
},
{
"epoch": 14.747200689061154,
"grad_norm": 0.35143256187438965,
"learning_rate": 0.0003790534536151878,
"loss": 3.0972,
"step": 54800
},
{
"epoch": 14.760658914728682,
"grad_norm": 0.3636065721511841,
"learning_rate": 0.00037885148781473005,
"loss": 3.0703,
"step": 54850
},
{
"epoch": 14.77411714039621,
"grad_norm": 0.3672044277191162,
"learning_rate": 0.00037864952201427225,
"loss": 3.092,
"step": 54900
},
{
"epoch": 14.787575366063738,
"grad_norm": 0.35274073481559753,
"learning_rate": 0.00037844755621381444,
"loss": 3.0911,
"step": 54950
},
{
"epoch": 14.801033591731267,
"grad_norm": 0.3666404187679291,
"learning_rate": 0.00037824559041335664,
"loss": 3.092,
"step": 55000
},
{
"epoch": 14.801033591731267,
"eval_accuracy": 0.39371078120131264,
"eval_loss": 3.2903215885162354,
"eval_runtime": 183.9011,
"eval_samples_per_second": 97.944,
"eval_steps_per_second": 6.123,
"step": 55000
},
{
"epoch": 14.814491817398794,
"grad_norm": 0.358341246843338,
"learning_rate": 0.00037804362461289884,
"loss": 3.0803,
"step": 55050
},
{
"epoch": 14.827950043066322,
"grad_norm": 0.37972983717918396,
"learning_rate": 0.0003778416588124411,
"loss": 3.0977,
"step": 55100
},
{
"epoch": 14.84140826873385,
"grad_norm": 0.35581347346305847,
"learning_rate": 0.0003776396930119833,
"loss": 3.092,
"step": 55150
},
{
"epoch": 14.854866494401378,
"grad_norm": 0.3848021924495697,
"learning_rate": 0.0003774377272115255,
"loss": 3.1046,
"step": 55200
},
{
"epoch": 14.868324720068905,
"grad_norm": 0.37682121992111206,
"learning_rate": 0.0003772357614110677,
"loss": 3.0902,
"step": 55250
},
{
"epoch": 14.881782945736434,
"grad_norm": 0.3632367253303528,
"learning_rate": 0.00037703379561060987,
"loss": 3.0952,
"step": 55300
},
{
"epoch": 14.895241171403962,
"grad_norm": 0.36668041348457336,
"learning_rate": 0.0003768318298101521,
"loss": 3.0885,
"step": 55350
},
{
"epoch": 14.90869939707149,
"grad_norm": 0.38183483481407166,
"learning_rate": 0.0003766298640096943,
"loss": 3.0962,
"step": 55400
},
{
"epoch": 14.922157622739018,
"grad_norm": 0.375584214925766,
"learning_rate": 0.0003764278982092365,
"loss": 3.093,
"step": 55450
},
{
"epoch": 14.935615848406545,
"grad_norm": 0.3452996611595154,
"learning_rate": 0.0003762259324087787,
"loss": 3.0908,
"step": 55500
},
{
"epoch": 14.949074074074074,
"grad_norm": 0.3545491695404053,
"learning_rate": 0.0003760239666083209,
"loss": 3.0923,
"step": 55550
},
{
"epoch": 14.962532299741602,
"grad_norm": 0.4003560543060303,
"learning_rate": 0.0003758220008078632,
"loss": 3.106,
"step": 55600
},
{
"epoch": 14.97599052540913,
"grad_norm": 0.3790423274040222,
"learning_rate": 0.0003756200350074054,
"loss": 3.0942,
"step": 55650
},
{
"epoch": 14.989448751076658,
"grad_norm": 0.3614756166934967,
"learning_rate": 0.0003754180692069476,
"loss": 3.1007,
"step": 55700
},
{
"epoch": 15.002691645133506,
"grad_norm": 0.3859131634235382,
"learning_rate": 0.0003752161034064898,
"loss": 3.0874,
"step": 55750
},
{
"epoch": 15.016149870801033,
"grad_norm": 0.3801725506782532,
"learning_rate": 0.00037501413760603205,
"loss": 2.9953,
"step": 55800
},
{
"epoch": 15.02960809646856,
"grad_norm": 0.37565118074417114,
"learning_rate": 0.00037481217180557425,
"loss": 3.0091,
"step": 55850
},
{
"epoch": 15.04306632213609,
"grad_norm": 0.3631201982498169,
"learning_rate": 0.00037461020600511644,
"loss": 3.0079,
"step": 55900
},
{
"epoch": 15.056524547803617,
"grad_norm": 0.3828172981739044,
"learning_rate": 0.00037440824020465864,
"loss": 3.0151,
"step": 55950
},
{
"epoch": 15.069982773471146,
"grad_norm": 0.38245272636413574,
"learning_rate": 0.00037420627440420083,
"loss": 3.0072,
"step": 56000
},
{
"epoch": 15.069982773471146,
"eval_accuracy": 0.3935517222047663,
"eval_loss": 3.3011186122894287,
"eval_runtime": 184.0554,
"eval_samples_per_second": 97.862,
"eval_steps_per_second": 6.118,
"step": 56000
},
{
"epoch": 15.083440999138674,
"grad_norm": 0.4072425961494446,
"learning_rate": 0.0003740043086037431,
"loss": 3.0186,
"step": 56050
},
{
"epoch": 15.0968992248062,
"grad_norm": 0.3522871732711792,
"learning_rate": 0.0003738023428032853,
"loss": 3.0268,
"step": 56100
},
{
"epoch": 15.11035745047373,
"grad_norm": 0.3603980839252472,
"learning_rate": 0.0003736003770028275,
"loss": 3.0261,
"step": 56150
},
{
"epoch": 15.123815676141257,
"grad_norm": 0.3941093385219574,
"learning_rate": 0.00037339841120236967,
"loss": 3.0134,
"step": 56200
},
{
"epoch": 15.137273901808786,
"grad_norm": 0.3810867965221405,
"learning_rate": 0.00037319644540191187,
"loss": 3.019,
"step": 56250
},
{
"epoch": 15.150732127476314,
"grad_norm": 0.36587831377983093,
"learning_rate": 0.0003729944796014541,
"loss": 3.0323,
"step": 56300
},
{
"epoch": 15.164190353143841,
"grad_norm": 0.35182157158851624,
"learning_rate": 0.0003727925138009963,
"loss": 3.0326,
"step": 56350
},
{
"epoch": 15.17764857881137,
"grad_norm": 0.3745343089103699,
"learning_rate": 0.0003725905480005385,
"loss": 3.0297,
"step": 56400
},
{
"epoch": 15.191106804478897,
"grad_norm": 0.37251755595207214,
"learning_rate": 0.0003723885822000807,
"loss": 3.0349,
"step": 56450
},
{
"epoch": 15.204565030146426,
"grad_norm": 0.3710574507713318,
"learning_rate": 0.000372186616399623,
"loss": 3.0404,
"step": 56500
},
{
"epoch": 15.218023255813954,
"grad_norm": 0.35331496596336365,
"learning_rate": 0.0003719846505991652,
"loss": 3.0329,
"step": 56550
},
{
"epoch": 15.231481481481481,
"grad_norm": 0.37161630392074585,
"learning_rate": 0.0003717826847987074,
"loss": 3.0316,
"step": 56600
},
{
"epoch": 15.24493970714901,
"grad_norm": 0.4004305601119995,
"learning_rate": 0.0003715807189982496,
"loss": 3.046,
"step": 56650
},
{
"epoch": 15.258397932816537,
"grad_norm": 0.3815003037452698,
"learning_rate": 0.00037137875319779185,
"loss": 3.0403,
"step": 56700
},
{
"epoch": 15.271856158484065,
"grad_norm": 0.36729151010513306,
"learning_rate": 0.00037117678739733405,
"loss": 3.0468,
"step": 56750
},
{
"epoch": 15.285314384151594,
"grad_norm": 0.36916160583496094,
"learning_rate": 0.00037097482159687624,
"loss": 3.0357,
"step": 56800
},
{
"epoch": 15.298772609819121,
"grad_norm": 0.36987948417663574,
"learning_rate": 0.00037077285579641844,
"loss": 3.0373,
"step": 56850
},
{
"epoch": 15.31223083548665,
"grad_norm": 0.369785338640213,
"learning_rate": 0.00037057088999596063,
"loss": 3.0428,
"step": 56900
},
{
"epoch": 15.325689061154177,
"grad_norm": 0.38476186990737915,
"learning_rate": 0.0003703689241955029,
"loss": 3.0608,
"step": 56950
},
{
"epoch": 15.339147286821705,
"grad_norm": 0.3720690608024597,
"learning_rate": 0.0003701669583950451,
"loss": 3.05,
"step": 57000
},
{
"epoch": 15.339147286821705,
"eval_accuracy": 0.3939683828958559,
"eval_loss": 3.293480157852173,
"eval_runtime": 183.8593,
"eval_samples_per_second": 97.966,
"eval_steps_per_second": 6.124,
"step": 57000
},
{
"epoch": 15.352605512489234,
"grad_norm": 0.3989238440990448,
"learning_rate": 0.0003699649925945873,
"loss": 3.0592,
"step": 57050
},
{
"epoch": 15.366063738156761,
"grad_norm": 0.3753744959831238,
"learning_rate": 0.0003697630267941295,
"loss": 3.0392,
"step": 57100
},
{
"epoch": 15.37952196382429,
"grad_norm": 0.37204575538635254,
"learning_rate": 0.00036956106099367167,
"loss": 3.0542,
"step": 57150
},
{
"epoch": 15.392980189491817,
"grad_norm": 0.3679342269897461,
"learning_rate": 0.0003693590951932139,
"loss": 3.0593,
"step": 57200
},
{
"epoch": 15.406438415159345,
"grad_norm": 0.38045749068260193,
"learning_rate": 0.0003691571293927561,
"loss": 3.0625,
"step": 57250
},
{
"epoch": 15.419896640826874,
"grad_norm": 0.35899248719215393,
"learning_rate": 0.0003689551635922983,
"loss": 3.0481,
"step": 57300
},
{
"epoch": 15.433354866494401,
"grad_norm": 0.378963440656662,
"learning_rate": 0.00036875319779184056,
"loss": 3.057,
"step": 57350
},
{
"epoch": 15.44681309216193,
"grad_norm": 0.3703141510486603,
"learning_rate": 0.0003685512319913828,
"loss": 3.0626,
"step": 57400
},
{
"epoch": 15.460271317829458,
"grad_norm": 0.371365487575531,
"learning_rate": 0.000368349266190925,
"loss": 3.0621,
"step": 57450
},
{
"epoch": 15.473729543496985,
"grad_norm": 0.38545140624046326,
"learning_rate": 0.0003681473003904672,
"loss": 3.0624,
"step": 57500
},
{
"epoch": 15.487187769164514,
"grad_norm": 0.3766907751560211,
"learning_rate": 0.0003679453345900094,
"loss": 3.0699,
"step": 57550
},
{
"epoch": 15.500645994832041,
"grad_norm": 0.38133764266967773,
"learning_rate": 0.0003677433687895516,
"loss": 3.0671,
"step": 57600
},
{
"epoch": 15.514104220499568,
"grad_norm": 0.3568243086338043,
"learning_rate": 0.00036754140298909385,
"loss": 3.0575,
"step": 57650
},
{
"epoch": 15.527562446167098,
"grad_norm": 0.3961988687515259,
"learning_rate": 0.00036733943718863604,
"loss": 3.0551,
"step": 57700
},
{
"epoch": 15.541020671834625,
"grad_norm": 0.3775664269924164,
"learning_rate": 0.00036713747138817824,
"loss": 3.0636,
"step": 57750
},
{
"epoch": 15.554478897502154,
"grad_norm": 0.37928569316864014,
"learning_rate": 0.00036693550558772044,
"loss": 3.0636,
"step": 57800
},
{
"epoch": 15.567937123169681,
"grad_norm": 0.38878610730171204,
"learning_rate": 0.00036673353978726263,
"loss": 3.0568,
"step": 57850
},
{
"epoch": 15.581395348837209,
"grad_norm": 0.37552499771118164,
"learning_rate": 0.0003665315739868049,
"loss": 3.0703,
"step": 57900
},
{
"epoch": 15.594853574504738,
"grad_norm": 0.3700394034385681,
"learning_rate": 0.0003663296081863471,
"loss": 3.0738,
"step": 57950
},
{
"epoch": 15.608311800172265,
"grad_norm": 0.37880730628967285,
"learning_rate": 0.0003661276423858893,
"loss": 3.0773,
"step": 58000
},
{
"epoch": 15.608311800172265,
"eval_accuracy": 0.393973597944923,
"eval_loss": 3.289884567260742,
"eval_runtime": 184.019,
"eval_samples_per_second": 97.881,
"eval_steps_per_second": 6.119,
"step": 58000
},
{
"epoch": 15.621770025839794,
"grad_norm": 0.3650750517845154,
"learning_rate": 0.00036592567658543147,
"loss": 3.0671,
"step": 58050
},
{
"epoch": 15.635228251507321,
"grad_norm": 0.36111393570899963,
"learning_rate": 0.00036572371078497367,
"loss": 3.0655,
"step": 58100
},
{
"epoch": 15.648686477174849,
"grad_norm": 0.37520965933799744,
"learning_rate": 0.0003655217449845159,
"loss": 3.0723,
"step": 58150
},
{
"epoch": 15.662144702842378,
"grad_norm": 0.3664579689502716,
"learning_rate": 0.0003653197791840581,
"loss": 3.0657,
"step": 58200
},
{
"epoch": 15.675602928509905,
"grad_norm": 0.36714988946914673,
"learning_rate": 0.00036511781338360036,
"loss": 3.0735,
"step": 58250
},
{
"epoch": 15.689061154177434,
"grad_norm": 0.3452919125556946,
"learning_rate": 0.00036491584758314256,
"loss": 3.0768,
"step": 58300
},
{
"epoch": 15.702519379844961,
"grad_norm": 0.37615737318992615,
"learning_rate": 0.0003647138817826848,
"loss": 3.0755,
"step": 58350
},
{
"epoch": 15.715977605512489,
"grad_norm": 0.3382650315761566,
"learning_rate": 0.000364511915982227,
"loss": 3.0768,
"step": 58400
},
{
"epoch": 15.729435831180018,
"grad_norm": 0.414098858833313,
"learning_rate": 0.0003643099501817692,
"loss": 3.0782,
"step": 58450
},
{
"epoch": 15.742894056847545,
"grad_norm": 0.9801768064498901,
"learning_rate": 0.0003641079843813114,
"loss": 3.0874,
"step": 58500
},
{
"epoch": 15.756352282515074,
"grad_norm": 0.36925193667411804,
"learning_rate": 0.0003639060185808536,
"loss": 3.0657,
"step": 58550
},
{
"epoch": 15.769810508182601,
"grad_norm": 0.3678208291530609,
"learning_rate": 0.00036370405278039584,
"loss": 3.0671,
"step": 58600
},
{
"epoch": 15.783268733850129,
"grad_norm": 0.3620169460773468,
"learning_rate": 0.00036350208697993804,
"loss": 3.0776,
"step": 58650
},
{
"epoch": 15.796726959517658,
"grad_norm": 0.35823333263397217,
"learning_rate": 0.00036330012117948024,
"loss": 3.0841,
"step": 58700
},
{
"epoch": 15.810185185185185,
"grad_norm": 0.39870336651802063,
"learning_rate": 0.00036309815537902243,
"loss": 3.0826,
"step": 58750
},
{
"epoch": 15.823643410852712,
"grad_norm": 0.38010063767433167,
"learning_rate": 0.0003628961895785647,
"loss": 3.0861,
"step": 58800
},
{
"epoch": 15.837101636520241,
"grad_norm": 0.3682940900325775,
"learning_rate": 0.0003626942237781069,
"loss": 3.0915,
"step": 58850
},
{
"epoch": 15.850559862187769,
"grad_norm": 0.3713534474372864,
"learning_rate": 0.0003624922579776491,
"loss": 3.0794,
"step": 58900
},
{
"epoch": 15.864018087855298,
"grad_norm": 0.351123183965683,
"learning_rate": 0.00036229029217719127,
"loss": 3.0801,
"step": 58950
},
{
"epoch": 15.877476313522825,
"grad_norm": 0.3966992199420929,
"learning_rate": 0.00036208832637673347,
"loss": 3.0813,
"step": 59000
},
{
"epoch": 15.877476313522825,
"eval_accuracy": 0.39422435488756574,
"eval_loss": 3.284731149673462,
"eval_runtime": 183.9514,
"eval_samples_per_second": 97.917,
"eval_steps_per_second": 6.121,
"step": 59000
},
{
"epoch": 15.890934539190352,
"grad_norm": 0.35703474283218384,
"learning_rate": 0.0003618863605762757,
"loss": 3.0714,
"step": 59050
},
{
"epoch": 15.904392764857882,
"grad_norm": 0.34707483649253845,
"learning_rate": 0.0003616843947758179,
"loss": 3.0883,
"step": 59100
},
{
"epoch": 15.917850990525409,
"grad_norm": 0.3808785080909729,
"learning_rate": 0.00036148242897536016,
"loss": 3.0823,
"step": 59150
},
{
"epoch": 15.931309216192938,
"grad_norm": 0.3602835237979889,
"learning_rate": 0.00036128046317490236,
"loss": 3.0809,
"step": 59200
},
{
"epoch": 15.944767441860465,
"grad_norm": 0.35113275051116943,
"learning_rate": 0.0003610784973744446,
"loss": 3.0765,
"step": 59250
},
{
"epoch": 15.958225667527993,
"grad_norm": 0.37724050879478455,
"learning_rate": 0.0003608765315739868,
"loss": 3.0856,
"step": 59300
},
{
"epoch": 15.971683893195522,
"grad_norm": 0.36171412467956543,
"learning_rate": 0.000360674565773529,
"loss": 3.0791,
"step": 59350
},
{
"epoch": 15.985142118863049,
"grad_norm": 0.39010947942733765,
"learning_rate": 0.0003604725999730712,
"loss": 3.0902,
"step": 59400
},
{
"epoch": 15.998600344530576,
"grad_norm": 0.3492063581943512,
"learning_rate": 0.0003602706341726134,
"loss": 3.0879,
"step": 59450
},
{
"epoch": 16.011843238587424,
"grad_norm": 0.3660726845264435,
"learning_rate": 0.00036006866837215564,
"loss": 3.0072,
"step": 59500
},
{
"epoch": 16.02530146425495,
"grad_norm": 0.3649909198284149,
"learning_rate": 0.00035986670257169784,
"loss": 2.9826,
"step": 59550
},
{
"epoch": 16.03875968992248,
"grad_norm": 0.37757793068885803,
"learning_rate": 0.00035966473677124004,
"loss": 2.992,
"step": 59600
},
{
"epoch": 16.05221791559001,
"grad_norm": 0.4068199396133423,
"learning_rate": 0.00035946277097078223,
"loss": 3.0007,
"step": 59650
},
{
"epoch": 16.065676141257537,
"grad_norm": 0.3982296288013458,
"learning_rate": 0.00035926080517032443,
"loss": 2.9933,
"step": 59700
},
{
"epoch": 16.079134366925064,
"grad_norm": 0.40942931175231934,
"learning_rate": 0.0003590588393698667,
"loss": 3.0125,
"step": 59750
},
{
"epoch": 16.09259259259259,
"grad_norm": 0.4063355326652527,
"learning_rate": 0.0003588568735694089,
"loss": 2.9953,
"step": 59800
},
{
"epoch": 16.10605081826012,
"grad_norm": 0.3793976604938507,
"learning_rate": 0.00035865490776895107,
"loss": 3.0102,
"step": 59850
},
{
"epoch": 16.11950904392765,
"grad_norm": 0.3756101429462433,
"learning_rate": 0.00035845294196849327,
"loss": 3.0007,
"step": 59900
},
{
"epoch": 16.132967269595177,
"grad_norm": 0.36124300956726074,
"learning_rate": 0.00035825097616803546,
"loss": 3.0135,
"step": 59950
},
{
"epoch": 16.146425495262704,
"grad_norm": 0.37558791041374207,
"learning_rate": 0.0003580490103675777,
"loss": 3.01,
"step": 60000
},
{
"epoch": 16.146425495262704,
"eval_accuracy": 0.39374978542246025,
"eval_loss": 3.2961766719818115,
"eval_runtime": 183.916,
"eval_samples_per_second": 97.936,
"eval_steps_per_second": 6.122,
"step": 60000
},
{
"epoch": 16.15988372093023,
"grad_norm": 0.363165020942688,
"learning_rate": 0.00035784704456711996,
"loss": 3.0097,
"step": 60050
},
{
"epoch": 16.17334194659776,
"grad_norm": 0.3869576156139374,
"learning_rate": 0.00035764507876666216,
"loss": 3.0086,
"step": 60100
},
{
"epoch": 16.18680017226529,
"grad_norm": 0.3944200277328491,
"learning_rate": 0.00035744311296620436,
"loss": 3.0258,
"step": 60150
},
{
"epoch": 16.200258397932817,
"grad_norm": 0.3678097128868103,
"learning_rate": 0.0003572411471657466,
"loss": 3.0165,
"step": 60200
},
{
"epoch": 16.213716623600344,
"grad_norm": 0.3999050557613373,
"learning_rate": 0.0003570391813652888,
"loss": 3.0125,
"step": 60250
},
{
"epoch": 16.227174849267872,
"grad_norm": 0.360252320766449,
"learning_rate": 0.000356837215564831,
"loss": 3.0254,
"step": 60300
},
{
"epoch": 16.2406330749354,
"grad_norm": 0.3781474530696869,
"learning_rate": 0.0003566352497643732,
"loss": 3.0309,
"step": 60350
},
{
"epoch": 16.25409130060293,
"grad_norm": 0.3796384036540985,
"learning_rate": 0.0003564332839639154,
"loss": 3.0241,
"step": 60400
},
{
"epoch": 16.267549526270457,
"grad_norm": 0.3623600900173187,
"learning_rate": 0.00035623131816345764,
"loss": 3.0308,
"step": 60450
},
{
"epoch": 16.281007751937985,
"grad_norm": 0.37352216243743896,
"learning_rate": 0.00035602935236299984,
"loss": 3.0387,
"step": 60500
},
{
"epoch": 16.294465977605512,
"grad_norm": 0.389754980802536,
"learning_rate": 0.00035582738656254203,
"loss": 3.0379,
"step": 60550
},
{
"epoch": 16.30792420327304,
"grad_norm": 0.39191320538520813,
"learning_rate": 0.00035562542076208423,
"loss": 3.0275,
"step": 60600
},
{
"epoch": 16.32138242894057,
"grad_norm": 0.36948418617248535,
"learning_rate": 0.0003554234549616264,
"loss": 3.0353,
"step": 60650
},
{
"epoch": 16.334840654608097,
"grad_norm": 0.35763826966285706,
"learning_rate": 0.0003552214891611687,
"loss": 3.0315,
"step": 60700
},
{
"epoch": 16.348298880275625,
"grad_norm": 0.35640308260917664,
"learning_rate": 0.00035501952336071087,
"loss": 3.038,
"step": 60750
},
{
"epoch": 16.361757105943152,
"grad_norm": 0.3724249303340912,
"learning_rate": 0.00035481755756025307,
"loss": 3.0335,
"step": 60800
},
{
"epoch": 16.37521533161068,
"grad_norm": 0.3601083755493164,
"learning_rate": 0.00035461559175979526,
"loss": 3.0341,
"step": 60850
},
{
"epoch": 16.38867355727821,
"grad_norm": 0.3982009291648865,
"learning_rate": 0.0003544136259593375,
"loss": 3.0482,
"step": 60900
},
{
"epoch": 16.402131782945737,
"grad_norm": 0.3764951527118683,
"learning_rate": 0.00035421166015887977,
"loss": 3.0446,
"step": 60950
},
{
"epoch": 16.415590008613265,
"grad_norm": 0.36200639605522156,
"learning_rate": 0.00035400969435842196,
"loss": 3.0502,
"step": 61000
},
{
"epoch": 16.415590008613265,
"eval_accuracy": 0.3944675065503189,
"eval_loss": 3.291944742202759,
"eval_runtime": 184.0405,
"eval_samples_per_second": 97.87,
"eval_steps_per_second": 6.118,
"step": 61000
},
{
"epoch": 16.429048234280792,
"grad_norm": 0.3781476318836212,
"learning_rate": 0.00035380772855796416,
"loss": 3.0506,
"step": 61050
},
{
"epoch": 16.44250645994832,
"grad_norm": 0.398057222366333,
"learning_rate": 0.0003536057627575064,
"loss": 3.0475,
"step": 61100
},
{
"epoch": 16.45596468561585,
"grad_norm": 0.3827942907810211,
"learning_rate": 0.0003534037969570486,
"loss": 3.0572,
"step": 61150
},
{
"epoch": 16.469422911283377,
"grad_norm": 0.37938353419303894,
"learning_rate": 0.0003532018311565908,
"loss": 3.0439,
"step": 61200
},
{
"epoch": 16.482881136950905,
"grad_norm": 0.4024749994277954,
"learning_rate": 0.000352999865356133,
"loss": 3.0497,
"step": 61250
},
{
"epoch": 16.496339362618432,
"grad_norm": 0.3803747296333313,
"learning_rate": 0.0003527978995556752,
"loss": 3.0499,
"step": 61300
},
{
"epoch": 16.50979758828596,
"grad_norm": 0.3896123766899109,
"learning_rate": 0.00035259593375521744,
"loss": 3.0289,
"step": 61350
},
{
"epoch": 16.52325581395349,
"grad_norm": 0.3559419512748718,
"learning_rate": 0.00035239396795475964,
"loss": 3.0582,
"step": 61400
},
{
"epoch": 16.536714039621017,
"grad_norm": 0.3737967312335968,
"learning_rate": 0.00035219200215430183,
"loss": 3.0608,
"step": 61450
},
{
"epoch": 16.550172265288545,
"grad_norm": 0.359353631734848,
"learning_rate": 0.00035199003635384403,
"loss": 3.0523,
"step": 61500
},
{
"epoch": 16.563630490956072,
"grad_norm": 0.366377592086792,
"learning_rate": 0.0003517880705533862,
"loss": 3.0551,
"step": 61550
},
{
"epoch": 16.5770887166236,
"grad_norm": 0.3800708055496216,
"learning_rate": 0.0003515861047529285,
"loss": 3.0629,
"step": 61600
},
{
"epoch": 16.590546942291127,
"grad_norm": 0.38061776757240295,
"learning_rate": 0.0003513841389524707,
"loss": 3.0551,
"step": 61650
},
{
"epoch": 16.604005167958658,
"grad_norm": 0.3898010849952698,
"learning_rate": 0.00035118217315201287,
"loss": 3.0575,
"step": 61700
},
{
"epoch": 16.617463393626185,
"grad_norm": 0.37334635853767395,
"learning_rate": 0.00035098020735155507,
"loss": 3.0627,
"step": 61750
},
{
"epoch": 16.630921619293712,
"grad_norm": 0.37698668241500854,
"learning_rate": 0.00035077824155109737,
"loss": 3.0591,
"step": 61800
},
{
"epoch": 16.64437984496124,
"grad_norm": 0.37113550305366516,
"learning_rate": 0.00035057627575063957,
"loss": 3.0566,
"step": 61850
},
{
"epoch": 16.657838070628767,
"grad_norm": 0.3867282569408417,
"learning_rate": 0.00035037430995018176,
"loss": 3.0682,
"step": 61900
},
{
"epoch": 16.671296296296298,
"grad_norm": 0.376988023519516,
"learning_rate": 0.00035017234414972396,
"loss": 3.0715,
"step": 61950
},
{
"epoch": 16.684754521963825,
"grad_norm": 0.41075727343559265,
"learning_rate": 0.00034997037834926615,
"loss": 3.0535,
"step": 62000
},
{
"epoch": 16.684754521963825,
"eval_accuracy": 0.3946495986802449,
"eval_loss": 3.2840731143951416,
"eval_runtime": 183.9525,
"eval_samples_per_second": 97.917,
"eval_steps_per_second": 6.121,
"step": 62000
},
{
"epoch": 16.698212747631352,
"grad_norm": 0.36573725938796997,
"learning_rate": 0.0003497684125488084,
"loss": 3.0583,
"step": 62050
},
{
"epoch": 16.71167097329888,
"grad_norm": 0.35161828994750977,
"learning_rate": 0.0003495664467483506,
"loss": 3.0646,
"step": 62100
},
{
"epoch": 16.725129198966407,
"grad_norm": 0.41279423236846924,
"learning_rate": 0.0003493644809478928,
"loss": 3.0535,
"step": 62150
},
{
"epoch": 16.738587424633938,
"grad_norm": 0.38668304681777954,
"learning_rate": 0.000349162515147435,
"loss": 3.0568,
"step": 62200
},
{
"epoch": 16.752045650301465,
"grad_norm": 0.37586507201194763,
"learning_rate": 0.0003489605493469772,
"loss": 3.0578,
"step": 62250
},
{
"epoch": 16.765503875968992,
"grad_norm": 0.37224096059799194,
"learning_rate": 0.00034875858354651944,
"loss": 3.0685,
"step": 62300
},
{
"epoch": 16.77896210163652,
"grad_norm": 0.37761619687080383,
"learning_rate": 0.00034855661774606164,
"loss": 3.0553,
"step": 62350
},
{
"epoch": 16.792420327304047,
"grad_norm": 0.37128373980522156,
"learning_rate": 0.00034835465194560383,
"loss": 3.0686,
"step": 62400
},
{
"epoch": 16.805878552971578,
"grad_norm": 0.3824944794178009,
"learning_rate": 0.00034815268614514603,
"loss": 3.0669,
"step": 62450
},
{
"epoch": 16.819336778639105,
"grad_norm": 0.3694535195827484,
"learning_rate": 0.0003479507203446882,
"loss": 3.0681,
"step": 62500
},
{
"epoch": 16.832795004306632,
"grad_norm": 0.3706541061401367,
"learning_rate": 0.0003477487545442305,
"loss": 3.0663,
"step": 62550
},
{
"epoch": 16.84625322997416,
"grad_norm": 0.3763303756713867,
"learning_rate": 0.00034754678874377267,
"loss": 3.062,
"step": 62600
},
{
"epoch": 16.859711455641687,
"grad_norm": 0.34424299001693726,
"learning_rate": 0.00034734482294331487,
"loss": 3.0755,
"step": 62650
},
{
"epoch": 16.873169681309218,
"grad_norm": 0.3574504852294922,
"learning_rate": 0.00034714285714285717,
"loss": 3.0661,
"step": 62700
},
{
"epoch": 16.886627906976745,
"grad_norm": 0.36437368392944336,
"learning_rate": 0.00034694089134239937,
"loss": 3.0721,
"step": 62750
},
{
"epoch": 16.900086132644272,
"grad_norm": 0.3977968394756317,
"learning_rate": 0.00034673892554194156,
"loss": 3.0691,
"step": 62800
},
{
"epoch": 16.9135443583118,
"grad_norm": 0.3810684382915497,
"learning_rate": 0.00034653695974148376,
"loss": 3.0787,
"step": 62850
},
{
"epoch": 16.927002583979327,
"grad_norm": 0.3766857981681824,
"learning_rate": 0.00034633499394102595,
"loss": 3.0729,
"step": 62900
},
{
"epoch": 16.940460809646858,
"grad_norm": 0.3923269510269165,
"learning_rate": 0.0003461330281405682,
"loss": 3.0763,
"step": 62950
},
{
"epoch": 16.953919035314385,
"grad_norm": 0.3806127607822418,
"learning_rate": 0.0003459310623401104,
"loss": 3.0682,
"step": 63000
},
{
"epoch": 16.953919035314385,
"eval_accuracy": 0.3952091299864018,
"eval_loss": 3.2774946689605713,
"eval_runtime": 183.8424,
"eval_samples_per_second": 97.975,
"eval_steps_per_second": 6.125,
"step": 63000
},
{
"epoch": 16.967377260981912,
"grad_norm": 0.3795175850391388,
"learning_rate": 0.0003457290965396526,
"loss": 3.0703,
"step": 63050
},
{
"epoch": 16.98083548664944,
"grad_norm": 0.4223317503929138,
"learning_rate": 0.0003455271307391948,
"loss": 3.0675,
"step": 63100
},
{
"epoch": 16.994293712316967,
"grad_norm": 0.36941879987716675,
"learning_rate": 0.000345325164938737,
"loss": 3.0743,
"step": 63150
},
{
"epoch": 17.007536606373815,
"grad_norm": 0.36872872710227966,
"learning_rate": 0.00034512319913827924,
"loss": 3.0187,
"step": 63200
},
{
"epoch": 17.020994832041342,
"grad_norm": 0.3980470597743988,
"learning_rate": 0.00034492123333782144,
"loss": 2.9664,
"step": 63250
},
{
"epoch": 17.034453057708873,
"grad_norm": 0.3716992735862732,
"learning_rate": 0.00034471926753736363,
"loss": 2.981,
"step": 63300
},
{
"epoch": 17.0479112833764,
"grad_norm": 0.3717067241668701,
"learning_rate": 0.00034451730173690583,
"loss": 2.9743,
"step": 63350
},
{
"epoch": 17.061369509043928,
"grad_norm": 0.3738197088241577,
"learning_rate": 0.000344315335936448,
"loss": 2.9815,
"step": 63400
},
{
"epoch": 17.074827734711455,
"grad_norm": 0.3692030608654022,
"learning_rate": 0.0003441133701359903,
"loss": 2.9969,
"step": 63450
},
{
"epoch": 17.088285960378983,
"grad_norm": 0.36867818236351013,
"learning_rate": 0.00034391140433553247,
"loss": 2.988,
"step": 63500
},
{
"epoch": 17.101744186046513,
"grad_norm": 0.3985981345176697,
"learning_rate": 0.00034370943853507467,
"loss": 2.9825,
"step": 63550
},
{
"epoch": 17.11520241171404,
"grad_norm": 0.38838744163513184,
"learning_rate": 0.0003435074727346169,
"loss": 2.9961,
"step": 63600
},
{
"epoch": 17.128660637381568,
"grad_norm": 0.38319239020347595,
"learning_rate": 0.00034330550693415917,
"loss": 3.0088,
"step": 63650
},
{
"epoch": 17.142118863049095,
"grad_norm": 0.37000760436058044,
"learning_rate": 0.00034310354113370136,
"loss": 3.0018,
"step": 63700
},
{
"epoch": 17.155577088716623,
"grad_norm": 0.3671315014362335,
"learning_rate": 0.00034290157533324356,
"loss": 2.9976,
"step": 63750
},
{
"epoch": 17.16903531438415,
"grad_norm": 0.39094942808151245,
"learning_rate": 0.00034269960953278576,
"loss": 2.986,
"step": 63800
},
{
"epoch": 17.18249354005168,
"grad_norm": 0.39503708481788635,
"learning_rate": 0.00034249764373232795,
"loss": 3.0112,
"step": 63850
},
{
"epoch": 17.195951765719208,
"grad_norm": 0.37030139565467834,
"learning_rate": 0.0003422956779318702,
"loss": 3.0169,
"step": 63900
},
{
"epoch": 17.209409991386735,
"grad_norm": 0.3955477476119995,
"learning_rate": 0.0003420937121314124,
"loss": 3.0137,
"step": 63950
},
{
"epoch": 17.222868217054263,
"grad_norm": 0.3910418450832367,
"learning_rate": 0.0003418917463309546,
"loss": 3.015,
"step": 64000
},
{
"epoch": 17.222868217054263,
"eval_accuracy": 0.3940875684964101,
"eval_loss": 3.2983345985412598,
"eval_runtime": 183.8849,
"eval_samples_per_second": 97.953,
"eval_steps_per_second": 6.123,
"step": 64000
},
{
"epoch": 17.23632644272179,
"grad_norm": 0.39123421907424927,
"learning_rate": 0.0003416897805304968,
"loss": 3.0074,
"step": 64050
},
{
"epoch": 17.24978466838932,
"grad_norm": 0.3913165032863617,
"learning_rate": 0.000341487814730039,
"loss": 3.0229,
"step": 64100
},
{
"epoch": 17.263242894056848,
"grad_norm": 0.3819050192832947,
"learning_rate": 0.00034128584892958124,
"loss": 3.0252,
"step": 64150
},
{
"epoch": 17.276701119724375,
"grad_norm": 0.3827211856842041,
"learning_rate": 0.00034108388312912343,
"loss": 3.0314,
"step": 64200
},
{
"epoch": 17.290159345391903,
"grad_norm": 0.38685205578804016,
"learning_rate": 0.00034088191732866563,
"loss": 3.0134,
"step": 64250
},
{
"epoch": 17.30361757105943,
"grad_norm": 0.39279136061668396,
"learning_rate": 0.0003406799515282078,
"loss": 3.016,
"step": 64300
},
{
"epoch": 17.31707579672696,
"grad_norm": 0.3748781979084015,
"learning_rate": 0.00034047798572775,
"loss": 3.0114,
"step": 64350
},
{
"epoch": 17.330534022394488,
"grad_norm": 0.3888700306415558,
"learning_rate": 0.00034027601992729227,
"loss": 3.034,
"step": 64400
},
{
"epoch": 17.343992248062015,
"grad_norm": 0.3873020112514496,
"learning_rate": 0.00034007405412683447,
"loss": 3.017,
"step": 64450
},
{
"epoch": 17.357450473729543,
"grad_norm": 0.3965982496738434,
"learning_rate": 0.0003398720883263767,
"loss": 3.0257,
"step": 64500
},
{
"epoch": 17.37090869939707,
"grad_norm": 0.4169600307941437,
"learning_rate": 0.00033967012252591897,
"loss": 3.0326,
"step": 64550
},
{
"epoch": 17.3843669250646,
"grad_norm": 0.40470919013023376,
"learning_rate": 0.00033946815672546116,
"loss": 3.0308,
"step": 64600
},
{
"epoch": 17.397825150732128,
"grad_norm": 0.38358378410339355,
"learning_rate": 0.00033926619092500336,
"loss": 3.0386,
"step": 64650
},
{
"epoch": 17.411283376399656,
"grad_norm": 0.39426061511039734,
"learning_rate": 0.00033906422512454556,
"loss": 3.0273,
"step": 64700
},
{
"epoch": 17.424741602067183,
"grad_norm": 0.37751317024230957,
"learning_rate": 0.00033886225932408775,
"loss": 3.0319,
"step": 64750
},
{
"epoch": 17.43819982773471,
"grad_norm": 0.3839034140110016,
"learning_rate": 0.00033866029352363,
"loss": 3.0353,
"step": 64800
},
{
"epoch": 17.45165805340224,
"grad_norm": 0.3945179283618927,
"learning_rate": 0.0003384583277231722,
"loss": 3.0466,
"step": 64850
},
{
"epoch": 17.46511627906977,
"grad_norm": 0.37897995114326477,
"learning_rate": 0.0003382563619227144,
"loss": 3.0371,
"step": 64900
},
{
"epoch": 17.478574504737296,
"grad_norm": 0.3820478618144989,
"learning_rate": 0.0003380543961222566,
"loss": 3.0258,
"step": 64950
},
{
"epoch": 17.492032730404823,
"grad_norm": 0.38667598366737366,
"learning_rate": 0.0003378524303217988,
"loss": 3.0393,
"step": 65000
},
{
"epoch": 17.492032730404823,
"eval_accuracy": 0.39442622074520445,
"eval_loss": 3.2931580543518066,
"eval_runtime": 184.2243,
"eval_samples_per_second": 97.772,
"eval_steps_per_second": 6.112,
"step": 65000
},
{
"epoch": 17.50549095607235,
"grad_norm": 0.3721429109573364,
"learning_rate": 0.00033765046452134104,
"loss": 3.0374,
"step": 65050
},
{
"epoch": 17.51894918173988,
"grad_norm": 0.3983703553676605,
"learning_rate": 0.00033744849872088323,
"loss": 3.047,
"step": 65100
},
{
"epoch": 17.53240740740741,
"grad_norm": 0.39016565680503845,
"learning_rate": 0.00033724653292042543,
"loss": 3.0417,
"step": 65150
},
{
"epoch": 17.545865633074936,
"grad_norm": 0.39062273502349854,
"learning_rate": 0.0003370445671199676,
"loss": 3.046,
"step": 65200
},
{
"epoch": 17.559323858742463,
"grad_norm": 0.39542749524116516,
"learning_rate": 0.0003368426013195098,
"loss": 3.037,
"step": 65250
},
{
"epoch": 17.57278208440999,
"grad_norm": 0.365821897983551,
"learning_rate": 0.00033664063551905207,
"loss": 3.0416,
"step": 65300
},
{
"epoch": 17.58624031007752,
"grad_norm": 0.36592647433280945,
"learning_rate": 0.00033643866971859427,
"loss": 3.0483,
"step": 65350
},
{
"epoch": 17.59969853574505,
"grad_norm": 0.3573816120624542,
"learning_rate": 0.0003362367039181365,
"loss": 3.0396,
"step": 65400
},
{
"epoch": 17.613156761412576,
"grad_norm": 0.38235795497894287,
"learning_rate": 0.0003360347381176787,
"loss": 3.0398,
"step": 65450
},
{
"epoch": 17.626614987080103,
"grad_norm": 0.3708433508872986,
"learning_rate": 0.00033583277231722097,
"loss": 3.0481,
"step": 65500
},
{
"epoch": 17.64007321274763,
"grad_norm": 0.3956705331802368,
"learning_rate": 0.00033563080651676316,
"loss": 3.0416,
"step": 65550
},
{
"epoch": 17.653531438415158,
"grad_norm": 0.37540948390960693,
"learning_rate": 0.00033542884071630536,
"loss": 3.051,
"step": 65600
},
{
"epoch": 17.66698966408269,
"grad_norm": 0.3630600869655609,
"learning_rate": 0.00033522687491584755,
"loss": 3.0505,
"step": 65650
},
{
"epoch": 17.680447889750216,
"grad_norm": 0.3790877163410187,
"learning_rate": 0.00033502490911538975,
"loss": 3.0468,
"step": 65700
},
{
"epoch": 17.693906115417743,
"grad_norm": 0.38710546493530273,
"learning_rate": 0.000334822943314932,
"loss": 3.0674,
"step": 65750
},
{
"epoch": 17.70736434108527,
"grad_norm": 0.36554524302482605,
"learning_rate": 0.0003346209775144742,
"loss": 3.0573,
"step": 65800
},
{
"epoch": 17.720822566752798,
"grad_norm": 0.37317368388175964,
"learning_rate": 0.0003344190117140164,
"loss": 3.0476,
"step": 65850
},
{
"epoch": 17.73428079242033,
"grad_norm": 0.3835221529006958,
"learning_rate": 0.0003342170459135586,
"loss": 3.041,
"step": 65900
},
{
"epoch": 17.747739018087856,
"grad_norm": 0.36117589473724365,
"learning_rate": 0.0003340150801131008,
"loss": 3.0539,
"step": 65950
},
{
"epoch": 17.761197243755383,
"grad_norm": 0.40491634607315063,
"learning_rate": 0.00033381311431264303,
"loss": 3.0468,
"step": 66000
},
{
"epoch": 17.761197243755383,
"eval_accuracy": 0.39504822399331085,
"eval_loss": 3.2837345600128174,
"eval_runtime": 189.1963,
"eval_samples_per_second": 95.203,
"eval_steps_per_second": 5.951,
"step": 66000
},
{
"epoch": 17.77465546942291,
"grad_norm": 0.38448449969291687,
"learning_rate": 0.00033361114851218523,
"loss": 3.0548,
"step": 66050
},
{
"epoch": 17.788113695090438,
"grad_norm": 0.4005850851535797,
"learning_rate": 0.0003334091827117274,
"loss": 3.0596,
"step": 66100
},
{
"epoch": 17.80157192075797,
"grad_norm": 0.3795042037963867,
"learning_rate": 0.0003332072169112696,
"loss": 3.0468,
"step": 66150
},
{
"epoch": 17.815030146425496,
"grad_norm": 0.3866879940032959,
"learning_rate": 0.0003330052511108118,
"loss": 3.0582,
"step": 66200
},
{
"epoch": 17.828488372093023,
"grad_norm": 0.3612496852874756,
"learning_rate": 0.0003328032853103541,
"loss": 3.0498,
"step": 66250
},
{
"epoch": 17.84194659776055,
"grad_norm": 0.4045388698577881,
"learning_rate": 0.0003326013195098963,
"loss": 3.0557,
"step": 66300
},
{
"epoch": 17.855404823428078,
"grad_norm": 0.3705078065395355,
"learning_rate": 0.0003323993537094385,
"loss": 3.055,
"step": 66350
},
{
"epoch": 17.86886304909561,
"grad_norm": 0.3949183225631714,
"learning_rate": 0.00033219738790898077,
"loss": 3.0385,
"step": 66400
},
{
"epoch": 17.882321274763136,
"grad_norm": 0.41610458493232727,
"learning_rate": 0.00033199542210852296,
"loss": 3.0587,
"step": 66450
},
{
"epoch": 17.895779500430663,
"grad_norm": 0.3761899769306183,
"learning_rate": 0.00033179345630806516,
"loss": 3.0463,
"step": 66500
},
{
"epoch": 17.90923772609819,
"grad_norm": 0.38714057207107544,
"learning_rate": 0.00033159149050760735,
"loss": 3.0572,
"step": 66550
},
{
"epoch": 17.922695951765718,
"grad_norm": 0.37656745314598083,
"learning_rate": 0.00033138952470714955,
"loss": 3.0635,
"step": 66600
},
{
"epoch": 17.93615417743325,
"grad_norm": 0.3685174882411957,
"learning_rate": 0.0003311875589066918,
"loss": 3.0578,
"step": 66650
},
{
"epoch": 17.949612403100776,
"grad_norm": 0.40085816383361816,
"learning_rate": 0.000330985593106234,
"loss": 3.0552,
"step": 66700
},
{
"epoch": 17.963070628768303,
"grad_norm": 0.4063645899295807,
"learning_rate": 0.0003307836273057762,
"loss": 3.0573,
"step": 66750
},
{
"epoch": 17.97652885443583,
"grad_norm": 0.3877021074295044,
"learning_rate": 0.0003305816615053184,
"loss": 3.0481,
"step": 66800
},
{
"epoch": 17.989987080103358,
"grad_norm": 0.4049699902534485,
"learning_rate": 0.0003303796957048606,
"loss": 3.0685,
"step": 66850
},
{
"epoch": 18.003229974160206,
"grad_norm": 0.3703552484512329,
"learning_rate": 0.00033017772990440284,
"loss": 3.0453,
"step": 66900
},
{
"epoch": 18.016688199827733,
"grad_norm": 0.37052857875823975,
"learning_rate": 0.00032997576410394503,
"loss": 2.9735,
"step": 66950
},
{
"epoch": 18.030146425495264,
"grad_norm": 0.3851967453956604,
"learning_rate": 0.00032977379830348723,
"loss": 2.9703,
"step": 67000
},
{
"epoch": 18.030146425495264,
"eval_accuracy": 0.3951228643830836,
"eval_loss": 3.2897040843963623,
"eval_runtime": 188.8123,
"eval_samples_per_second": 95.396,
"eval_steps_per_second": 5.964,
"step": 67000
},
{
"epoch": 18.04360465116279,
"grad_norm": 0.4044719934463501,
"learning_rate": 0.0003295718325030294,
"loss": 2.9702,
"step": 67050
},
{
"epoch": 18.05706287683032,
"grad_norm": 0.40578025579452515,
"learning_rate": 0.0003293698667025716,
"loss": 2.9705,
"step": 67100
},
{
"epoch": 18.070521102497846,
"grad_norm": 0.3843235373497009,
"learning_rate": 0.0003291679009021139,
"loss": 2.9806,
"step": 67150
},
{
"epoch": 18.083979328165373,
"grad_norm": 0.3752574622631073,
"learning_rate": 0.0003289659351016561,
"loss": 2.9821,
"step": 67200
},
{
"epoch": 18.097437553832904,
"grad_norm": 0.3655632734298706,
"learning_rate": 0.0003287639693011983,
"loss": 2.9836,
"step": 67250
},
{
"epoch": 18.11089577950043,
"grad_norm": 0.3748213052749634,
"learning_rate": 0.0003285620035007405,
"loss": 2.9859,
"step": 67300
},
{
"epoch": 18.12435400516796,
"grad_norm": 0.3862340450286865,
"learning_rate": 0.00032836003770028276,
"loss": 2.9814,
"step": 67350
},
{
"epoch": 18.137812230835486,
"grad_norm": 0.3841283917427063,
"learning_rate": 0.00032815807189982496,
"loss": 2.9981,
"step": 67400
},
{
"epoch": 18.151270456503013,
"grad_norm": 0.3780440092086792,
"learning_rate": 0.00032795610609936715,
"loss": 2.9911,
"step": 67450
},
{
"epoch": 18.164728682170544,
"grad_norm": 0.41674312949180603,
"learning_rate": 0.00032775414029890935,
"loss": 2.9984,
"step": 67500
},
{
"epoch": 18.17818690783807,
"grad_norm": 0.39418357610702515,
"learning_rate": 0.00032755217449845155,
"loss": 3.0043,
"step": 67550
},
{
"epoch": 18.1916451335056,
"grad_norm": 0.3886651396751404,
"learning_rate": 0.0003273502086979938,
"loss": 2.9981,
"step": 67600
},
{
"epoch": 18.205103359173126,
"grad_norm": 0.3845861256122589,
"learning_rate": 0.000327148242897536,
"loss": 2.9928,
"step": 67650
},
{
"epoch": 18.218561584840653,
"grad_norm": 0.4107899069786072,
"learning_rate": 0.0003269462770970782,
"loss": 2.9944,
"step": 67700
},
{
"epoch": 18.232019810508184,
"grad_norm": 0.3802962303161621,
"learning_rate": 0.0003267443112966204,
"loss": 3.0084,
"step": 67750
},
{
"epoch": 18.24547803617571,
"grad_norm": 0.3840334713459015,
"learning_rate": 0.0003265423454961626,
"loss": 3.0029,
"step": 67800
},
{
"epoch": 18.25893626184324,
"grad_norm": 0.41139617562294006,
"learning_rate": 0.00032634037969570483,
"loss": 3.0128,
"step": 67850
},
{
"epoch": 18.272394487510766,
"grad_norm": 0.38927018642425537,
"learning_rate": 0.00032613841389524703,
"loss": 3.0058,
"step": 67900
},
{
"epoch": 18.285852713178294,
"grad_norm": 0.38476383686065674,
"learning_rate": 0.0003259364480947892,
"loss": 3.0019,
"step": 67950
},
{
"epoch": 18.29931093884582,
"grad_norm": 0.40857797861099243,
"learning_rate": 0.0003257344822943314,
"loss": 3.0132,
"step": 68000
},
{
"epoch": 18.29931093884582,
"eval_accuracy": 0.3949620670368482,
"eval_loss": 3.2903034687042236,
"eval_runtime": 184.0784,
"eval_samples_per_second": 97.85,
"eval_steps_per_second": 6.117,
"step": 68000
},
{
"epoch": 18.31276916451335,
"grad_norm": 0.37042713165283203,
"learning_rate": 0.0003255325164938737,
"loss": 3.0022,
"step": 68050
},
{
"epoch": 18.32622739018088,
"grad_norm": 0.3854128122329712,
"learning_rate": 0.0003253305506934159,
"loss": 3.0162,
"step": 68100
},
{
"epoch": 18.339685615848406,
"grad_norm": 0.40771451592445374,
"learning_rate": 0.0003251285848929581,
"loss": 3.0076,
"step": 68150
},
{
"epoch": 18.353143841515934,
"grad_norm": 0.4190777838230133,
"learning_rate": 0.0003249266190925003,
"loss": 3.0179,
"step": 68200
},
{
"epoch": 18.36660206718346,
"grad_norm": 0.39956730604171753,
"learning_rate": 0.0003247246532920425,
"loss": 3.0172,
"step": 68250
},
{
"epoch": 18.38006029285099,
"grad_norm": 0.4219856560230255,
"learning_rate": 0.00032452268749158476,
"loss": 3.0133,
"step": 68300
},
{
"epoch": 18.39351851851852,
"grad_norm": 0.4254962205886841,
"learning_rate": 0.00032432072169112696,
"loss": 3.0112,
"step": 68350
},
{
"epoch": 18.406976744186046,
"grad_norm": 0.3884270191192627,
"learning_rate": 0.00032411875589066915,
"loss": 3.0209,
"step": 68400
},
{
"epoch": 18.420434969853574,
"grad_norm": 0.3815394341945648,
"learning_rate": 0.00032391679009021135,
"loss": 3.0197,
"step": 68450
},
{
"epoch": 18.4338931955211,
"grad_norm": 0.3894241750240326,
"learning_rate": 0.0003237148242897536,
"loss": 3.0165,
"step": 68500
},
{
"epoch": 18.447351421188632,
"grad_norm": 0.3976706862449646,
"learning_rate": 0.0003235128584892958,
"loss": 3.0156,
"step": 68550
},
{
"epoch": 18.46080964685616,
"grad_norm": 0.3741625249385834,
"learning_rate": 0.000323310892688838,
"loss": 3.0187,
"step": 68600
},
{
"epoch": 18.474267872523686,
"grad_norm": 0.3954469561576843,
"learning_rate": 0.0003231089268883802,
"loss": 3.0181,
"step": 68650
},
{
"epoch": 18.487726098191214,
"grad_norm": 0.408069908618927,
"learning_rate": 0.0003229069610879224,
"loss": 3.0055,
"step": 68700
},
{
"epoch": 18.50118432385874,
"grad_norm": 0.3852391839027405,
"learning_rate": 0.00032270499528746463,
"loss": 3.0274,
"step": 68750
},
{
"epoch": 18.514642549526272,
"grad_norm": 0.39901864528656006,
"learning_rate": 0.00032250302948700683,
"loss": 3.0206,
"step": 68800
},
{
"epoch": 18.5281007751938,
"grad_norm": 0.3986130356788635,
"learning_rate": 0.000322301063686549,
"loss": 3.0362,
"step": 68850
},
{
"epoch": 18.541559000861326,
"grad_norm": 0.3933941721916199,
"learning_rate": 0.0003220990978860912,
"loss": 3.0329,
"step": 68900
},
{
"epoch": 18.555017226528854,
"grad_norm": 0.37979447841644287,
"learning_rate": 0.0003218971320856335,
"loss": 3.0289,
"step": 68950
},
{
"epoch": 18.56847545219638,
"grad_norm": 0.4128454327583313,
"learning_rate": 0.0003216951662851757,
"loss": 3.0333,
"step": 69000
},
{
"epoch": 18.56847545219638,
"eval_accuracy": 0.3952685598163955,
"eval_loss": 3.2874090671539307,
"eval_runtime": 189.6716,
"eval_samples_per_second": 94.964,
"eval_steps_per_second": 5.937,
"step": 69000
},
{
"epoch": 18.581933677863912,
"grad_norm": 0.4021677076816559,
"learning_rate": 0.0003214932004847179,
"loss": 3.0352,
"step": 69050
},
{
"epoch": 18.59539190353144,
"grad_norm": 0.3812282681465149,
"learning_rate": 0.0003212912346842601,
"loss": 3.0263,
"step": 69100
},
{
"epoch": 18.608850129198967,
"grad_norm": 0.3537648320198059,
"learning_rate": 0.0003210892688838023,
"loss": 3.0324,
"step": 69150
},
{
"epoch": 18.622308354866494,
"grad_norm": 0.40709200501441956,
"learning_rate": 0.00032088730308334456,
"loss": 3.0286,
"step": 69200
},
{
"epoch": 18.63576658053402,
"grad_norm": 0.38711681962013245,
"learning_rate": 0.00032068533728288676,
"loss": 3.0252,
"step": 69250
},
{
"epoch": 18.649224806201552,
"grad_norm": 0.402924507856369,
"learning_rate": 0.00032048337148242895,
"loss": 3.0387,
"step": 69300
},
{
"epoch": 18.66268303186908,
"grad_norm": 0.37442925572395325,
"learning_rate": 0.00032028140568197115,
"loss": 3.0399,
"step": 69350
},
{
"epoch": 18.676141257536607,
"grad_norm": 0.39951178431510925,
"learning_rate": 0.00032007943988151334,
"loss": 3.0346,
"step": 69400
},
{
"epoch": 18.689599483204134,
"grad_norm": 0.387456476688385,
"learning_rate": 0.0003198774740810556,
"loss": 3.041,
"step": 69450
},
{
"epoch": 18.70305770887166,
"grad_norm": 0.41782769560813904,
"learning_rate": 0.0003196755082805978,
"loss": 3.0331,
"step": 69500
},
{
"epoch": 18.716515934539192,
"grad_norm": 0.3862665295600891,
"learning_rate": 0.00031947354248014,
"loss": 3.0371,
"step": 69550
},
{
"epoch": 18.72997416020672,
"grad_norm": 0.3616723418235779,
"learning_rate": 0.0003192715766796822,
"loss": 3.044,
"step": 69600
},
{
"epoch": 18.743432385874247,
"grad_norm": 0.3773272633552551,
"learning_rate": 0.0003190696108792244,
"loss": 3.0479,
"step": 69650
},
{
"epoch": 18.756890611541774,
"grad_norm": 0.40163275599479675,
"learning_rate": 0.00031886764507876663,
"loss": 3.0389,
"step": 69700
},
{
"epoch": 18.7703488372093,
"grad_norm": 0.40031328797340393,
"learning_rate": 0.0003186656792783088,
"loss": 3.0345,
"step": 69750
},
{
"epoch": 18.783807062876832,
"grad_norm": 0.3898553252220154,
"learning_rate": 0.000318463713477851,
"loss": 3.041,
"step": 69800
},
{
"epoch": 18.79726528854436,
"grad_norm": 0.38567641377449036,
"learning_rate": 0.00031826174767739327,
"loss": 3.0446,
"step": 69850
},
{
"epoch": 18.810723514211887,
"grad_norm": 0.3848070800304413,
"learning_rate": 0.0003180597818769355,
"loss": 3.0428,
"step": 69900
},
{
"epoch": 18.824181739879414,
"grad_norm": 0.3871477544307709,
"learning_rate": 0.0003178578160764777,
"loss": 3.0273,
"step": 69950
},
{
"epoch": 18.83763996554694,
"grad_norm": 0.3792720437049866,
"learning_rate": 0.0003176558502760199,
"loss": 3.0468,
"step": 70000
},
{
"epoch": 18.83763996554694,
"eval_accuracy": 0.3952458526235825,
"eval_loss": 3.2808475494384766,
"eval_runtime": 189.2525,
"eval_samples_per_second": 95.174,
"eval_steps_per_second": 5.95,
"step": 70000
},
{
"epoch": 18.85109819121447,
"grad_norm": 0.39831456542015076,
"learning_rate": 0.0003174538844755621,
"loss": 3.05,
"step": 70050
},
{
"epoch": 18.864556416882,
"grad_norm": 0.3866911828517914,
"learning_rate": 0.0003172519186751043,
"loss": 3.0484,
"step": 70100
},
{
"epoch": 18.878014642549527,
"grad_norm": 0.40169939398765564,
"learning_rate": 0.00031704995287464656,
"loss": 3.0408,
"step": 70150
},
{
"epoch": 18.891472868217054,
"grad_norm": 0.38556337356567383,
"learning_rate": 0.00031684798707418875,
"loss": 3.0421,
"step": 70200
},
{
"epoch": 18.90493109388458,
"grad_norm": 0.38593757152557373,
"learning_rate": 0.00031664602127373095,
"loss": 3.0546,
"step": 70250
},
{
"epoch": 18.91838931955211,
"grad_norm": 0.4056928753852844,
"learning_rate": 0.00031644405547327315,
"loss": 3.0465,
"step": 70300
},
{
"epoch": 18.93184754521964,
"grad_norm": 0.38051751255989075,
"learning_rate": 0.00031624208967281534,
"loss": 3.041,
"step": 70350
},
{
"epoch": 18.945305770887167,
"grad_norm": 0.38520440459251404,
"learning_rate": 0.0003160401238723576,
"loss": 3.0434,
"step": 70400
},
{
"epoch": 18.958763996554694,
"grad_norm": 0.3821350336074829,
"learning_rate": 0.0003158381580718998,
"loss": 3.0566,
"step": 70450
},
{
"epoch": 18.97222222222222,
"grad_norm": 0.39325737953186035,
"learning_rate": 0.000315636192271442,
"loss": 3.0419,
"step": 70500
},
{
"epoch": 18.98568044788975,
"grad_norm": 0.3835277855396271,
"learning_rate": 0.0003154342264709842,
"loss": 3.036,
"step": 70550
},
{
"epoch": 18.99913867355728,
"grad_norm": 0.4259049892425537,
"learning_rate": 0.00031523226067052643,
"loss": 3.0468,
"step": 70600
},
{
"epoch": 19.012381567614124,
"grad_norm": 0.4106285870075226,
"learning_rate": 0.0003150302948700686,
"loss": 2.951,
"step": 70650
},
{
"epoch": 19.025839793281655,
"grad_norm": 0.39034339785575867,
"learning_rate": 0.0003148283290696108,
"loss": 2.949,
"step": 70700
},
{
"epoch": 19.039298018949182,
"grad_norm": 0.40479207038879395,
"learning_rate": 0.0003146263632691531,
"loss": 2.9668,
"step": 70750
},
{
"epoch": 19.05275624461671,
"grad_norm": 0.39626309275627136,
"learning_rate": 0.0003144243974686953,
"loss": 2.9687,
"step": 70800
},
{
"epoch": 19.066214470284237,
"grad_norm": 0.3860742747783661,
"learning_rate": 0.0003142224316682375,
"loss": 2.9639,
"step": 70850
},
{
"epoch": 19.079672695951764,
"grad_norm": 0.41022372245788574,
"learning_rate": 0.0003140204658677797,
"loss": 2.9624,
"step": 70900
},
{
"epoch": 19.093130921619295,
"grad_norm": 0.4118083119392395,
"learning_rate": 0.0003138185000673219,
"loss": 2.9673,
"step": 70950
},
{
"epoch": 19.106589147286822,
"grad_norm": 0.38944822549819946,
"learning_rate": 0.0003136165342668641,
"loss": 2.9668,
"step": 71000
},
{
"epoch": 19.106589147286822,
"eval_accuracy": 0.39475639853926475,
"eval_loss": 3.296081304550171,
"eval_runtime": 184.2628,
"eval_samples_per_second": 97.752,
"eval_steps_per_second": 6.111,
"step": 71000
},
{
"epoch": 19.12004737295435,
"grad_norm": 0.37801867723464966,
"learning_rate": 0.00031341456846640636,
"loss": 2.9702,
"step": 71050
},
{
"epoch": 19.133505598621877,
"grad_norm": 0.4007479250431061,
"learning_rate": 0.00031321260266594855,
"loss": 2.9778,
"step": 71100
},
{
"epoch": 19.146963824289404,
"grad_norm": 0.40579596161842346,
"learning_rate": 0.00031301063686549075,
"loss": 2.9767,
"step": 71150
},
{
"epoch": 19.160422049956935,
"grad_norm": 0.4116387963294983,
"learning_rate": 0.00031280867106503295,
"loss": 2.983,
"step": 71200
},
{
"epoch": 19.173880275624462,
"grad_norm": 0.392718642950058,
"learning_rate": 0.00031260670526457514,
"loss": 2.9764,
"step": 71250
},
{
"epoch": 19.18733850129199,
"grad_norm": 0.42457208037376404,
"learning_rate": 0.0003124047394641174,
"loss": 2.996,
"step": 71300
},
{
"epoch": 19.200796726959517,
"grad_norm": 0.4037551283836365,
"learning_rate": 0.0003122027736636596,
"loss": 3.0011,
"step": 71350
},
{
"epoch": 19.214254952627044,
"grad_norm": 0.39706146717071533,
"learning_rate": 0.0003120008078632018,
"loss": 2.9931,
"step": 71400
},
{
"epoch": 19.227713178294575,
"grad_norm": 0.38429006934165955,
"learning_rate": 0.000311798842062744,
"loss": 2.9832,
"step": 71450
},
{
"epoch": 19.241171403962102,
"grad_norm": 0.39253419637680054,
"learning_rate": 0.0003115968762622862,
"loss": 2.988,
"step": 71500
},
{
"epoch": 19.25462962962963,
"grad_norm": 0.4121507704257965,
"learning_rate": 0.00031139491046182843,
"loss": 2.9826,
"step": 71550
},
{
"epoch": 19.268087855297157,
"grad_norm": 0.3723185956478119,
"learning_rate": 0.0003111929446613707,
"loss": 2.9906,
"step": 71600
},
{
"epoch": 19.281546080964684,
"grad_norm": 0.37958472967147827,
"learning_rate": 0.0003109909788609129,
"loss": 2.9938,
"step": 71650
},
{
"epoch": 19.295004306632215,
"grad_norm": 0.4000052511692047,
"learning_rate": 0.00031078901306045507,
"loss": 3.0022,
"step": 71700
},
{
"epoch": 19.308462532299743,
"grad_norm": 0.3927767276763916,
"learning_rate": 0.0003105870472599973,
"loss": 2.9989,
"step": 71750
},
{
"epoch": 19.32192075796727,
"grad_norm": 0.4024776518344879,
"learning_rate": 0.0003103850814595395,
"loss": 2.9991,
"step": 71800
},
{
"epoch": 19.335378983634797,
"grad_norm": 0.4009549915790558,
"learning_rate": 0.0003101831156590817,
"loss": 2.9873,
"step": 71850
},
{
"epoch": 19.348837209302324,
"grad_norm": 0.3862859904766083,
"learning_rate": 0.0003099811498586239,
"loss": 3.0074,
"step": 71900
},
{
"epoch": 19.362295434969855,
"grad_norm": 0.41279205679893494,
"learning_rate": 0.0003097791840581661,
"loss": 3.0057,
"step": 71950
},
{
"epoch": 19.375753660637383,
"grad_norm": 0.40375789999961853,
"learning_rate": 0.00030957721825770836,
"loss": 3.0052,
"step": 72000
},
{
"epoch": 19.375753660637383,
"eval_accuracy": 0.39542479399469715,
"eval_loss": 3.2887797355651855,
"eval_runtime": 190.3996,
"eval_samples_per_second": 94.601,
"eval_steps_per_second": 5.914,
"step": 72000
},
{
"epoch": 19.38921188630491,
"grad_norm": 0.40637972950935364,
"learning_rate": 0.00030937525245725055,
"loss": 3.006,
"step": 72050
},
{
"epoch": 19.402670111972437,
"grad_norm": 0.38699838519096375,
"learning_rate": 0.00030917328665679275,
"loss": 3.0048,
"step": 72100
},
{
"epoch": 19.416128337639964,
"grad_norm": 0.37975725531578064,
"learning_rate": 0.00030897132085633494,
"loss": 3.0084,
"step": 72150
},
{
"epoch": 19.429586563307495,
"grad_norm": 0.38831770420074463,
"learning_rate": 0.00030876935505587714,
"loss": 3.0069,
"step": 72200
},
{
"epoch": 19.443044788975023,
"grad_norm": 0.4179569482803345,
"learning_rate": 0.0003085673892554194,
"loss": 3.0175,
"step": 72250
},
{
"epoch": 19.45650301464255,
"grad_norm": 0.3928871154785156,
"learning_rate": 0.0003083654234549616,
"loss": 3.0115,
"step": 72300
},
{
"epoch": 19.469961240310077,
"grad_norm": 0.3936832845211029,
"learning_rate": 0.0003081634576545038,
"loss": 3.0044,
"step": 72350
},
{
"epoch": 19.483419465977605,
"grad_norm": 0.39544516801834106,
"learning_rate": 0.000307961491854046,
"loss": 3.0151,
"step": 72400
},
{
"epoch": 19.496877691645132,
"grad_norm": 0.39217159152030945,
"learning_rate": 0.00030775952605358823,
"loss": 3.0229,
"step": 72450
},
{
"epoch": 19.510335917312663,
"grad_norm": 0.4028678238391876,
"learning_rate": 0.0003075575602531305,
"loss": 3.0184,
"step": 72500
},
{
"epoch": 19.52379414298019,
"grad_norm": 0.4130101501941681,
"learning_rate": 0.0003073555944526727,
"loss": 3.0208,
"step": 72550
},
{
"epoch": 19.537252368647717,
"grad_norm": 0.3764054775238037,
"learning_rate": 0.00030715362865221487,
"loss": 3.0171,
"step": 72600
},
{
"epoch": 19.550710594315245,
"grad_norm": 0.39863958954811096,
"learning_rate": 0.0003069516628517571,
"loss": 3.017,
"step": 72650
},
{
"epoch": 19.564168819982772,
"grad_norm": 0.4149794280529022,
"learning_rate": 0.0003067496970512993,
"loss": 3.0175,
"step": 72700
},
{
"epoch": 19.577627045650303,
"grad_norm": 0.3763502240180969,
"learning_rate": 0.0003065477312508415,
"loss": 3.0147,
"step": 72750
},
{
"epoch": 19.59108527131783,
"grad_norm": 0.4097903072834015,
"learning_rate": 0.0003063457654503837,
"loss": 3.019,
"step": 72800
},
{
"epoch": 19.604543496985357,
"grad_norm": 0.40566208958625793,
"learning_rate": 0.0003061437996499259,
"loss": 3.0179,
"step": 72850
},
{
"epoch": 19.618001722652885,
"grad_norm": 0.3858266770839691,
"learning_rate": 0.00030594183384946816,
"loss": 3.0164,
"step": 72900
},
{
"epoch": 19.631459948320412,
"grad_norm": 0.39712074398994446,
"learning_rate": 0.00030573986804901035,
"loss": 3.0249,
"step": 72950
},
{
"epoch": 19.644918173987943,
"grad_norm": 0.4112600088119507,
"learning_rate": 0.00030553790224855255,
"loss": 3.03,
"step": 73000
},
{
"epoch": 19.644918173987943,
"eval_accuracy": 0.39553322355655046,
"eval_loss": 3.2810277938842773,
"eval_runtime": 189.544,
"eval_samples_per_second": 95.028,
"eval_steps_per_second": 5.941,
"step": 73000
},
{
"epoch": 19.65837639965547,
"grad_norm": 0.3868493437767029,
"learning_rate": 0.00030533593644809474,
"loss": 3.0143,
"step": 73050
},
{
"epoch": 19.671834625322997,
"grad_norm": 0.4045863151550293,
"learning_rate": 0.00030513397064763694,
"loss": 3.0275,
"step": 73100
},
{
"epoch": 19.685292850990525,
"grad_norm": 0.3915933072566986,
"learning_rate": 0.0003049320048471792,
"loss": 3.0216,
"step": 73150
},
{
"epoch": 19.698751076658052,
"grad_norm": 0.35868579149246216,
"learning_rate": 0.0003047300390467214,
"loss": 3.0273,
"step": 73200
},
{
"epoch": 19.712209302325583,
"grad_norm": 0.4197313189506531,
"learning_rate": 0.0003045280732462636,
"loss": 3.015,
"step": 73250
},
{
"epoch": 19.72566752799311,
"grad_norm": 0.407588928937912,
"learning_rate": 0.0003043261074458058,
"loss": 3.033,
"step": 73300
},
{
"epoch": 19.739125753660637,
"grad_norm": 0.4119780957698822,
"learning_rate": 0.000304124141645348,
"loss": 3.0217,
"step": 73350
},
{
"epoch": 19.752583979328165,
"grad_norm": 0.3930842876434326,
"learning_rate": 0.0003039221758448903,
"loss": 3.0327,
"step": 73400
},
{
"epoch": 19.766042204995692,
"grad_norm": 0.3857138752937317,
"learning_rate": 0.0003037202100444325,
"loss": 3.0267,
"step": 73450
},
{
"epoch": 19.779500430663223,
"grad_norm": 0.3695453405380249,
"learning_rate": 0.00030351824424397467,
"loss": 3.0283,
"step": 73500
},
{
"epoch": 19.79295865633075,
"grad_norm": 0.3961401879787445,
"learning_rate": 0.00030331627844351687,
"loss": 3.0212,
"step": 73550
},
{
"epoch": 19.806416881998278,
"grad_norm": 0.3860207796096802,
"learning_rate": 0.0003031143126430591,
"loss": 3.0291,
"step": 73600
},
{
"epoch": 19.819875107665805,
"grad_norm": 0.4158506691455841,
"learning_rate": 0.0003029123468426013,
"loss": 3.0322,
"step": 73650
},
{
"epoch": 19.833333333333332,
"grad_norm": 0.41216781735420227,
"learning_rate": 0.0003027103810421435,
"loss": 3.0453,
"step": 73700
},
{
"epoch": 19.846791559000863,
"grad_norm": 0.3952142894268036,
"learning_rate": 0.0003025084152416857,
"loss": 3.03,
"step": 73750
},
{
"epoch": 19.86024978466839,
"grad_norm": 0.43454983830451965,
"learning_rate": 0.0003023064494412279,
"loss": 3.04,
"step": 73800
},
{
"epoch": 19.873708010335918,
"grad_norm": 0.39687493443489075,
"learning_rate": 0.00030210448364077015,
"loss": 3.0341,
"step": 73850
},
{
"epoch": 19.887166236003445,
"grad_norm": 0.39751407504081726,
"learning_rate": 0.00030190251784031235,
"loss": 3.0331,
"step": 73900
},
{
"epoch": 19.900624461670972,
"grad_norm": 0.3789171874523163,
"learning_rate": 0.00030170055203985454,
"loss": 3.0375,
"step": 73950
},
{
"epoch": 19.9140826873385,
"grad_norm": 0.408931165933609,
"learning_rate": 0.00030149858623939674,
"loss": 3.0357,
"step": 74000
},
{
"epoch": 19.9140826873385,
"eval_accuracy": 0.3961871689801928,
"eval_loss": 3.2747573852539062,
"eval_runtime": 183.7521,
"eval_samples_per_second": 98.023,
"eval_steps_per_second": 6.128,
"step": 74000
},
{
"epoch": 19.92754091300603,
"grad_norm": 0.3790091574192047,
"learning_rate": 0.00030129662043893894,
"loss": 3.0349,
"step": 74050
},
{
"epoch": 19.940999138673558,
"grad_norm": 0.39011651277542114,
"learning_rate": 0.0003010946546384812,
"loss": 3.0433,
"step": 74100
},
{
"epoch": 19.954457364341085,
"grad_norm": 0.3978932201862335,
"learning_rate": 0.0003008926888380234,
"loss": 3.0369,
"step": 74150
},
{
"epoch": 19.967915590008612,
"grad_norm": 0.42351245880126953,
"learning_rate": 0.0003006907230375656,
"loss": 3.0408,
"step": 74200
},
{
"epoch": 19.98137381567614,
"grad_norm": 0.42332905530929565,
"learning_rate": 0.0003004887572371078,
"loss": 3.0381,
"step": 74250
},
{
"epoch": 19.99483204134367,
"grad_norm": 0.3873363137245178,
"learning_rate": 0.0003002867914366501,
"loss": 3.0329,
"step": 74300
},
{
"epoch": 20.00807493540052,
"grad_norm": 0.4089195728302002,
"learning_rate": 0.0003000848256361923,
"loss": 2.9764,
"step": 74350
},
{
"epoch": 20.021533161068046,
"grad_norm": 0.39422905445098877,
"learning_rate": 0.00029988285983573447,
"loss": 2.9574,
"step": 74400
},
{
"epoch": 20.034991386735573,
"grad_norm": 0.39730948209762573,
"learning_rate": 0.00029968089403527667,
"loss": 2.9511,
"step": 74450
},
{
"epoch": 20.0484496124031,
"grad_norm": 0.3876565396785736,
"learning_rate": 0.0002994789282348189,
"loss": 2.9479,
"step": 74500
},
{
"epoch": 20.061907838070628,
"grad_norm": 0.41247814893722534,
"learning_rate": 0.0002992769624343611,
"loss": 2.9596,
"step": 74550
},
{
"epoch": 20.075366063738155,
"grad_norm": 0.4031682312488556,
"learning_rate": 0.0002990749966339033,
"loss": 2.9519,
"step": 74600
},
{
"epoch": 20.088824289405686,
"grad_norm": 0.39940112829208374,
"learning_rate": 0.0002988730308334455,
"loss": 2.9628,
"step": 74650
},
{
"epoch": 20.102282515073213,
"grad_norm": 0.39713823795318604,
"learning_rate": 0.0002986710650329877,
"loss": 2.9619,
"step": 74700
},
{
"epoch": 20.11574074074074,
"grad_norm": 0.39015573263168335,
"learning_rate": 0.00029846909923252995,
"loss": 2.9656,
"step": 74750
},
{
"epoch": 20.129198966408268,
"grad_norm": 0.4017931818962097,
"learning_rate": 0.00029826713343207215,
"loss": 2.9643,
"step": 74800
},
{
"epoch": 20.142657192075795,
"grad_norm": 0.4000353217124939,
"learning_rate": 0.00029806516763161435,
"loss": 2.9692,
"step": 74850
},
{
"epoch": 20.156115417743326,
"grad_norm": 0.42363348603248596,
"learning_rate": 0.00029786320183115654,
"loss": 2.9703,
"step": 74900
},
{
"epoch": 20.169573643410853,
"grad_norm": 0.4110814332962036,
"learning_rate": 0.0002976612360306988,
"loss": 2.9761,
"step": 74950
},
{
"epoch": 20.18303186907838,
"grad_norm": 0.39522799849510193,
"learning_rate": 0.000297459270230241,
"loss": 2.9628,
"step": 75000
},
{
"epoch": 20.18303186907838,
"eval_accuracy": 0.39536188746532536,
"eval_loss": 3.2912895679473877,
"eval_runtime": 184.019,
"eval_samples_per_second": 97.881,
"eval_steps_per_second": 6.119,
"step": 75000
},
{
"epoch": 20.196490094745908,
"grad_norm": 0.39896243810653687,
"learning_rate": 0.0002972573044297832,
"loss": 2.9689,
"step": 75050
},
{
"epoch": 20.209948320413435,
"grad_norm": 0.4028981924057007,
"learning_rate": 0.00029705533862932543,
"loss": 2.979,
"step": 75100
},
{
"epoch": 20.223406546080966,
"grad_norm": 0.3958054482936859,
"learning_rate": 0.00029685337282886763,
"loss": 2.9803,
"step": 75150
},
{
"epoch": 20.236864771748493,
"grad_norm": 0.37057286500930786,
"learning_rate": 0.0002966514070284098,
"loss": 2.9796,
"step": 75200
},
{
"epoch": 20.25032299741602,
"grad_norm": 0.40401262044906616,
"learning_rate": 0.000296449441227952,
"loss": 2.9936,
"step": 75250
},
{
"epoch": 20.263781223083548,
"grad_norm": 0.40256842970848083,
"learning_rate": 0.0002962474754274942,
"loss": 2.9796,
"step": 75300
},
{
"epoch": 20.277239448751075,
"grad_norm": 0.39651206135749817,
"learning_rate": 0.00029604550962703647,
"loss": 2.9734,
"step": 75350
},
{
"epoch": 20.290697674418606,
"grad_norm": 0.3988285958766937,
"learning_rate": 0.00029584354382657867,
"loss": 2.9734,
"step": 75400
},
{
"epoch": 20.304155900086133,
"grad_norm": 0.4057372510433197,
"learning_rate": 0.0002956415780261209,
"loss": 2.9858,
"step": 75450
},
{
"epoch": 20.31761412575366,
"grad_norm": 0.42367351055145264,
"learning_rate": 0.0002954396122256631,
"loss": 2.9924,
"step": 75500
},
{
"epoch": 20.331072351421188,
"grad_norm": 0.4053604006767273,
"learning_rate": 0.0002952376464252053,
"loss": 2.9919,
"step": 75550
},
{
"epoch": 20.344530577088715,
"grad_norm": 0.4110828936100006,
"learning_rate": 0.0002950356806247475,
"loss": 2.9851,
"step": 75600
},
{
"epoch": 20.357988802756246,
"grad_norm": 0.39539217948913574,
"learning_rate": 0.0002948337148242897,
"loss": 2.9911,
"step": 75650
},
{
"epoch": 20.371447028423773,
"grad_norm": 0.3663056790828705,
"learning_rate": 0.00029463174902383195,
"loss": 2.9918,
"step": 75700
},
{
"epoch": 20.3849052540913,
"grad_norm": 0.4023444950580597,
"learning_rate": 0.00029442978322337415,
"loss": 2.9905,
"step": 75750
},
{
"epoch": 20.398363479758828,
"grad_norm": 0.3879607617855072,
"learning_rate": 0.00029422781742291634,
"loss": 2.9939,
"step": 75800
},
{
"epoch": 20.411821705426355,
"grad_norm": 0.4148692488670349,
"learning_rate": 0.0002940258516224586,
"loss": 3.012,
"step": 75850
},
{
"epoch": 20.425279931093886,
"grad_norm": 0.39173731207847595,
"learning_rate": 0.0002938238858220008,
"loss": 3.0053,
"step": 75900
},
{
"epoch": 20.438738156761413,
"grad_norm": 0.39822375774383545,
"learning_rate": 0.000293621920021543,
"loss": 3.0039,
"step": 75950
},
{
"epoch": 20.45219638242894,
"grad_norm": 0.4079001843929291,
"learning_rate": 0.0002934199542210852,
"loss": 3.0068,
"step": 76000
},
{
"epoch": 20.45219638242894,
"eval_accuracy": 0.39590316609974735,
"eval_loss": 3.284907341003418,
"eval_runtime": 184.136,
"eval_samples_per_second": 97.819,
"eval_steps_per_second": 6.115,
"step": 76000
},
{
"epoch": 20.465654608096468,
"grad_norm": 0.399150013923645,
"learning_rate": 0.00029321798842062743,
"loss": 2.9923,
"step": 76050
},
{
"epoch": 20.479112833763995,
"grad_norm": 0.41484132409095764,
"learning_rate": 0.00029301602262016963,
"loss": 3.0037,
"step": 76100
},
{
"epoch": 20.492571059431526,
"grad_norm": 0.38788777589797974,
"learning_rate": 0.0002928140568197118,
"loss": 2.9952,
"step": 76150
},
{
"epoch": 20.506029285099054,
"grad_norm": 0.4266553223133087,
"learning_rate": 0.000292612091019254,
"loss": 3.0068,
"step": 76200
},
{
"epoch": 20.51948751076658,
"grad_norm": 0.42803579568862915,
"learning_rate": 0.0002924101252187962,
"loss": 3.0048,
"step": 76250
},
{
"epoch": 20.532945736434108,
"grad_norm": 0.3889942765235901,
"learning_rate": 0.00029220815941833847,
"loss": 2.9994,
"step": 76300
},
{
"epoch": 20.546403962101635,
"grad_norm": 0.4029588997364044,
"learning_rate": 0.0002920061936178807,
"loss": 3.0083,
"step": 76350
},
{
"epoch": 20.559862187769163,
"grad_norm": 0.39647406339645386,
"learning_rate": 0.0002918042278174229,
"loss": 3.0154,
"step": 76400
},
{
"epoch": 20.573320413436694,
"grad_norm": 0.39515092968940735,
"learning_rate": 0.0002916022620169651,
"loss": 3.011,
"step": 76450
},
{
"epoch": 20.58677863910422,
"grad_norm": 0.39511623978614807,
"learning_rate": 0.0002914002962165073,
"loss": 3.008,
"step": 76500
},
{
"epoch": 20.600236864771748,
"grad_norm": 0.3756105899810791,
"learning_rate": 0.0002911983304160495,
"loss": 3.0044,
"step": 76550
},
{
"epoch": 20.613695090439276,
"grad_norm": 0.4061303734779358,
"learning_rate": 0.00029099636461559175,
"loss": 3.0088,
"step": 76600
},
{
"epoch": 20.627153316106803,
"grad_norm": 0.39304038882255554,
"learning_rate": 0.00029079439881513395,
"loss": 3.0113,
"step": 76650
},
{
"epoch": 20.640611541774334,
"grad_norm": 0.41393136978149414,
"learning_rate": 0.00029059243301467614,
"loss": 3.0058,
"step": 76700
},
{
"epoch": 20.65406976744186,
"grad_norm": 0.42535823583602905,
"learning_rate": 0.0002903904672142184,
"loss": 3.0027,
"step": 76750
},
{
"epoch": 20.66752799310939,
"grad_norm": 0.3962944447994232,
"learning_rate": 0.0002901885014137606,
"loss": 3.0206,
"step": 76800
},
{
"epoch": 20.680986218776916,
"grad_norm": 0.40562066435813904,
"learning_rate": 0.0002899865356133028,
"loss": 3.0187,
"step": 76850
},
{
"epoch": 20.694444444444443,
"grad_norm": 0.38790571689605713,
"learning_rate": 0.000289784569812845,
"loss": 3.0059,
"step": 76900
},
{
"epoch": 20.707902670111974,
"grad_norm": 0.3984473645687103,
"learning_rate": 0.00028958260401238723,
"loss": 3.0283,
"step": 76950
},
{
"epoch": 20.7213608957795,
"grad_norm": 0.4049811363220215,
"learning_rate": 0.00028938063821192943,
"loss": 3.0089,
"step": 77000
},
{
"epoch": 20.7213608957795,
"eval_accuracy": 0.39621270099125044,
"eval_loss": 3.2795207500457764,
"eval_runtime": 183.997,
"eval_samples_per_second": 97.893,
"eval_steps_per_second": 6.12,
"step": 77000
},
{
"epoch": 20.73481912144703,
"grad_norm": 0.4167289733886719,
"learning_rate": 0.0002891786724114716,
"loss": 3.0002,
"step": 77050
},
{
"epoch": 20.748277347114556,
"grad_norm": 0.41908499598503113,
"learning_rate": 0.0002889767066110138,
"loss": 3.0249,
"step": 77100
},
{
"epoch": 20.761735572782083,
"grad_norm": 0.4360521137714386,
"learning_rate": 0.000288774740810556,
"loss": 3.0182,
"step": 77150
},
{
"epoch": 20.775193798449614,
"grad_norm": 0.4084163010120392,
"learning_rate": 0.00028857277501009827,
"loss": 3.0184,
"step": 77200
},
{
"epoch": 20.78865202411714,
"grad_norm": 0.4210131764411926,
"learning_rate": 0.00028837080920964046,
"loss": 3.0239,
"step": 77250
},
{
"epoch": 20.80211024978467,
"grad_norm": 0.390126496553421,
"learning_rate": 0.0002881688434091827,
"loss": 3.0166,
"step": 77300
},
{
"epoch": 20.815568475452196,
"grad_norm": 0.406694233417511,
"learning_rate": 0.0002879668776087249,
"loss": 3.013,
"step": 77350
},
{
"epoch": 20.829026701119723,
"grad_norm": 0.4124731421470642,
"learning_rate": 0.0002877649118082671,
"loss": 3.0141,
"step": 77400
},
{
"epoch": 20.842484926787254,
"grad_norm": 0.4077942967414856,
"learning_rate": 0.0002875629460078093,
"loss": 3.0179,
"step": 77450
},
{
"epoch": 20.85594315245478,
"grad_norm": 0.3900335729122162,
"learning_rate": 0.0002873609802073515,
"loss": 3.0131,
"step": 77500
},
{
"epoch": 20.86940137812231,
"grad_norm": 0.412984162569046,
"learning_rate": 0.00028715901440689375,
"loss": 3.0147,
"step": 77550
},
{
"epoch": 20.882859603789836,
"grad_norm": 0.385636568069458,
"learning_rate": 0.00028695704860643594,
"loss": 3.0211,
"step": 77600
},
{
"epoch": 20.896317829457363,
"grad_norm": 0.42929261922836304,
"learning_rate": 0.0002867550828059782,
"loss": 3.0306,
"step": 77650
},
{
"epoch": 20.909776055124894,
"grad_norm": 0.3917275369167328,
"learning_rate": 0.0002865531170055204,
"loss": 3.0331,
"step": 77700
},
{
"epoch": 20.92323428079242,
"grad_norm": 0.40508773922920227,
"learning_rate": 0.0002863511512050626,
"loss": 3.0255,
"step": 77750
},
{
"epoch": 20.93669250645995,
"grad_norm": 0.39373040199279785,
"learning_rate": 0.0002861491854046048,
"loss": 3.0327,
"step": 77800
},
{
"epoch": 20.950150732127476,
"grad_norm": 0.4229165315628052,
"learning_rate": 0.000285947219604147,
"loss": 3.014,
"step": 77850
},
{
"epoch": 20.963608957795003,
"grad_norm": 0.39980220794677734,
"learning_rate": 0.00028574525380368923,
"loss": 3.0261,
"step": 77900
},
{
"epoch": 20.977067183462534,
"grad_norm": 0.3871513605117798,
"learning_rate": 0.0002855432880032314,
"loss": 3.0257,
"step": 77950
},
{
"epoch": 20.99052540913006,
"grad_norm": 0.40497952699661255,
"learning_rate": 0.0002853413222027736,
"loss": 3.0351,
"step": 78000
},
{
"epoch": 20.99052540913006,
"eval_accuracy": 0.3964473781992696,
"eval_loss": 3.2759013175964355,
"eval_runtime": 183.6955,
"eval_samples_per_second": 98.054,
"eval_steps_per_second": 6.13,
"step": 78000
},
{
"epoch": 21.00376830318691,
"grad_norm": 0.4050183594226837,
"learning_rate": 0.00028513935640231587,
"loss": 3.0079,
"step": 78050
},
{
"epoch": 21.017226528854437,
"grad_norm": 0.42034977674484253,
"learning_rate": 0.00028493739060185807,
"loss": 2.9429,
"step": 78100
},
{
"epoch": 21.030684754521964,
"grad_norm": 0.394689679145813,
"learning_rate": 0.00028473542480140026,
"loss": 2.9501,
"step": 78150
},
{
"epoch": 21.04414298018949,
"grad_norm": 0.40265870094299316,
"learning_rate": 0.0002845334590009425,
"loss": 2.9452,
"step": 78200
},
{
"epoch": 21.05760120585702,
"grad_norm": 0.41180652379989624,
"learning_rate": 0.0002843314932004847,
"loss": 2.9436,
"step": 78250
},
{
"epoch": 21.07105943152455,
"grad_norm": 0.40233322978019714,
"learning_rate": 0.0002841295274000269,
"loss": 2.9538,
"step": 78300
},
{
"epoch": 21.084517657192077,
"grad_norm": 0.4016159474849701,
"learning_rate": 0.0002839275615995691,
"loss": 2.9539,
"step": 78350
},
{
"epoch": 21.097975882859604,
"grad_norm": 0.39117708802223206,
"learning_rate": 0.0002837255957991113,
"loss": 2.9506,
"step": 78400
},
{
"epoch": 21.11143410852713,
"grad_norm": 0.4153161644935608,
"learning_rate": 0.00028352362999865355,
"loss": 2.9579,
"step": 78450
},
{
"epoch": 21.12489233419466,
"grad_norm": 0.41663727164268494,
"learning_rate": 0.00028332166419819574,
"loss": 2.9574,
"step": 78500
},
{
"epoch": 21.13835055986219,
"grad_norm": 0.4258229434490204,
"learning_rate": 0.000283119698397738,
"loss": 2.9621,
"step": 78550
},
{
"epoch": 21.151808785529717,
"grad_norm": 0.41690391302108765,
"learning_rate": 0.0002829177325972802,
"loss": 2.958,
"step": 78600
},
{
"epoch": 21.165267011197244,
"grad_norm": 0.44034937024116516,
"learning_rate": 0.0002827157667968224,
"loss": 2.9651,
"step": 78650
},
{
"epoch": 21.17872523686477,
"grad_norm": 0.43333423137664795,
"learning_rate": 0.0002825138009963646,
"loss": 2.9736,
"step": 78700
},
{
"epoch": 21.1921834625323,
"grad_norm": 0.41911640763282776,
"learning_rate": 0.0002823118351959068,
"loss": 2.9529,
"step": 78750
},
{
"epoch": 21.205641688199826,
"grad_norm": 0.4228711426258087,
"learning_rate": 0.00028210986939544903,
"loss": 2.9702,
"step": 78800
},
{
"epoch": 21.219099913867357,
"grad_norm": 0.4386906623840332,
"learning_rate": 0.0002819079035949912,
"loss": 2.9698,
"step": 78850
},
{
"epoch": 21.232558139534884,
"grad_norm": 0.4098826050758362,
"learning_rate": 0.0002817059377945334,
"loss": 2.9756,
"step": 78900
},
{
"epoch": 21.24601636520241,
"grad_norm": 0.3815784454345703,
"learning_rate": 0.00028150397199407567,
"loss": 2.9596,
"step": 78950
},
{
"epoch": 21.25947459086994,
"grad_norm": 0.4362355172634125,
"learning_rate": 0.00028130200619361787,
"loss": 2.9646,
"step": 79000
},
{
"epoch": 21.25947459086994,
"eval_accuracy": 0.39573280783022236,
"eval_loss": 3.2918295860290527,
"eval_runtime": 184.4614,
"eval_samples_per_second": 97.646,
"eval_steps_per_second": 6.104,
"step": 79000
},
{
"epoch": 21.272932816537466,
"grad_norm": 0.3898933529853821,
"learning_rate": 0.00028110004039316006,
"loss": 2.9613,
"step": 79050
},
{
"epoch": 21.286391042204997,
"grad_norm": 0.4078889787197113,
"learning_rate": 0.00028089807459270226,
"loss": 2.973,
"step": 79100
},
{
"epoch": 21.299849267872524,
"grad_norm": 0.4291805624961853,
"learning_rate": 0.0002806961087922445,
"loss": 2.9902,
"step": 79150
},
{
"epoch": 21.31330749354005,
"grad_norm": 0.39538031816482544,
"learning_rate": 0.0002804941429917867,
"loss": 2.979,
"step": 79200
},
{
"epoch": 21.32676571920758,
"grad_norm": 0.43110764026641846,
"learning_rate": 0.0002802921771913289,
"loss": 2.977,
"step": 79250
},
{
"epoch": 21.340223944875106,
"grad_norm": 0.40676701068878174,
"learning_rate": 0.0002800902113908711,
"loss": 2.9725,
"step": 79300
},
{
"epoch": 21.353682170542637,
"grad_norm": 0.4107375741004944,
"learning_rate": 0.0002798882455904133,
"loss": 2.983,
"step": 79350
},
{
"epoch": 21.367140396210164,
"grad_norm": 0.40966796875,
"learning_rate": 0.00027968627978995555,
"loss": 2.9871,
"step": 79400
},
{
"epoch": 21.38059862187769,
"grad_norm": 0.41336268186569214,
"learning_rate": 0.00027948431398949774,
"loss": 2.9836,
"step": 79450
},
{
"epoch": 21.39405684754522,
"grad_norm": 0.4066007733345032,
"learning_rate": 0.00027928234818904,
"loss": 2.983,
"step": 79500
},
{
"epoch": 21.407515073212746,
"grad_norm": 0.41753581166267395,
"learning_rate": 0.0002790803823885822,
"loss": 2.9916,
"step": 79550
},
{
"epoch": 21.420973298880277,
"grad_norm": 0.3889058530330658,
"learning_rate": 0.0002788784165881244,
"loss": 2.9861,
"step": 79600
},
{
"epoch": 21.434431524547804,
"grad_norm": 0.4066503345966339,
"learning_rate": 0.0002786764507876666,
"loss": 2.9815,
"step": 79650
},
{
"epoch": 21.44788975021533,
"grad_norm": 0.40962690114974976,
"learning_rate": 0.0002784744849872088,
"loss": 2.9917,
"step": 79700
},
{
"epoch": 21.46134797588286,
"grad_norm": 0.40320873260498047,
"learning_rate": 0.000278272519186751,
"loss": 2.993,
"step": 79750
},
{
"epoch": 21.474806201550386,
"grad_norm": 0.3946399390697479,
"learning_rate": 0.0002780705533862932,
"loss": 2.9864,
"step": 79800
},
{
"epoch": 21.488264427217917,
"grad_norm": 0.4374764859676361,
"learning_rate": 0.0002778685875858355,
"loss": 2.9883,
"step": 79850
},
{
"epoch": 21.501722652885444,
"grad_norm": 0.39441385865211487,
"learning_rate": 0.00027766662178537767,
"loss": 3.0073,
"step": 79900
},
{
"epoch": 21.51518087855297,
"grad_norm": 0.44837531447410583,
"learning_rate": 0.00027746465598491987,
"loss": 2.9916,
"step": 79950
},
{
"epoch": 21.5286391042205,
"grad_norm": 0.39739325642585754,
"learning_rate": 0.00027726269018446206,
"loss": 2.9889,
"step": 80000
},
{
"epoch": 21.5286391042205,
"eval_accuracy": 0.3961463177625006,
"eval_loss": 3.285086154937744,
"eval_runtime": 184.1174,
"eval_samples_per_second": 97.829,
"eval_steps_per_second": 6.116,
"step": 80000
},
{
"epoch": 21.542097329888026,
"grad_norm": 0.42453229427337646,
"learning_rate": 0.00027706072438400426,
"loss": 2.9329,
"step": 80050
},
{
"epoch": 21.555555555555557,
"grad_norm": 0.4305717349052429,
"learning_rate": 0.0002768587585835465,
"loss": 2.9363,
"step": 80100
},
{
"epoch": 21.569013781223084,
"grad_norm": 0.4032965302467346,
"learning_rate": 0.0002766567927830887,
"loss": 2.9397,
"step": 80150
},
{
"epoch": 21.58247200689061,
"grad_norm": 0.408806174993515,
"learning_rate": 0.0002764548269826309,
"loss": 2.9397,
"step": 80200
},
{
"epoch": 21.59593023255814,
"grad_norm": 0.43218156695365906,
"learning_rate": 0.0002762528611821731,
"loss": 2.9552,
"step": 80250
},
{
"epoch": 21.609388458225666,
"grad_norm": 0.40385672450065613,
"learning_rate": 0.00027605089538171535,
"loss": 2.956,
"step": 80300
},
{
"epoch": 21.622846683893197,
"grad_norm": 0.42708295583724976,
"learning_rate": 0.00027584892958125754,
"loss": 2.9497,
"step": 80350
},
{
"epoch": 21.636304909560724,
"grad_norm": 0.40945789217948914,
"learning_rate": 0.0002756469637807998,
"loss": 2.9269,
"step": 80400
},
{
"epoch": 21.649763135228252,
"grad_norm": 0.40183472633361816,
"learning_rate": 0.000275444997980342,
"loss": 2.9572,
"step": 80450
},
{
"epoch": 21.66322136089578,
"grad_norm": 0.428378164768219,
"learning_rate": 0.0002752430321798842,
"loss": 2.9607,
"step": 80500
},
{
"epoch": 21.676679586563306,
"grad_norm": 0.4280439019203186,
"learning_rate": 0.0002750410663794264,
"loss": 2.9494,
"step": 80550
},
{
"epoch": 21.690137812230837,
"grad_norm": 0.43740445375442505,
"learning_rate": 0.0002748391005789686,
"loss": 2.9702,
"step": 80600
},
{
"epoch": 21.703596037898365,
"grad_norm": 0.4336613714694977,
"learning_rate": 0.00027463713477851083,
"loss": 2.9567,
"step": 80650
},
{
"epoch": 21.717054263565892,
"grad_norm": 0.42132559418678284,
"learning_rate": 0.000274435168978053,
"loss": 2.9667,
"step": 80700
},
{
"epoch": 21.73051248923342,
"grad_norm": 0.440794438123703,
"learning_rate": 0.0002742332031775953,
"loss": 2.9661,
"step": 80750
},
{
"epoch": 21.743970714900946,
"grad_norm": 0.39131981134414673,
"learning_rate": 0.00027403123737713747,
"loss": 2.9664,
"step": 80800
},
{
"epoch": 21.757428940568474,
"grad_norm": 0.43297865986824036,
"learning_rate": 0.00027382927157667967,
"loss": 2.9666,
"step": 80850
},
{
"epoch": 21.770887166236005,
"grad_norm": 0.40193209052085876,
"learning_rate": 0.00027362730577622186,
"loss": 2.9673,
"step": 80900
},
{
"epoch": 21.784345391903532,
"grad_norm": 0.4124595522880554,
"learning_rate": 0.00027342533997576406,
"loss": 2.971,
"step": 80950
},
{
"epoch": 21.79780361757106,
"grad_norm": 0.4073331952095032,
"learning_rate": 0.0002732233741753063,
"loss": 2.9748,
"step": 81000
},
{
"epoch": 21.79780361757106,
"eval_accuracy": 0.3957925636007828,
"eval_loss": 3.28918194770813,
"eval_runtime": 188.2992,
"eval_samples_per_second": 95.656,
"eval_steps_per_second": 5.98,
"step": 81000
},
{
"epoch": 21.811261843238587,
"grad_norm": 0.4031127989292145,
"learning_rate": 0.0002730214083748485,
"loss": 2.9708,
"step": 81050
},
{
"epoch": 21.824720068906114,
"grad_norm": 0.40961411595344543,
"learning_rate": 0.0002728194425743907,
"loss": 2.978,
"step": 81100
},
{
"epoch": 21.838178294573645,
"grad_norm": 0.42932042479515076,
"learning_rate": 0.0002726174767739329,
"loss": 2.9786,
"step": 81150
},
{
"epoch": 21.851636520241172,
"grad_norm": 0.4209270775318146,
"learning_rate": 0.00027241551097347515,
"loss": 2.9597,
"step": 81200
},
{
"epoch": 21.8650947459087,
"grad_norm": 0.3944014310836792,
"learning_rate": 0.00027221354517301734,
"loss": 2.9844,
"step": 81250
},
{
"epoch": 21.878552971576227,
"grad_norm": 0.40551164746284485,
"learning_rate": 0.00027201157937255954,
"loss": 2.9744,
"step": 81300
},
{
"epoch": 21.892011197243754,
"grad_norm": 0.40660858154296875,
"learning_rate": 0.0002718096135721018,
"loss": 2.9779,
"step": 81350
},
{
"epoch": 21.905469422911285,
"grad_norm": 0.4533492624759674,
"learning_rate": 0.000271607647771644,
"loss": 2.9897,
"step": 81400
},
{
"epoch": 21.918927648578812,
"grad_norm": 0.41591590642929077,
"learning_rate": 0.0002714056819711862,
"loss": 2.9734,
"step": 81450
},
{
"epoch": 21.93238587424634,
"grad_norm": 0.4329775273799896,
"learning_rate": 0.0002712037161707284,
"loss": 2.9853,
"step": 81500
},
{
"epoch": 21.945844099913867,
"grad_norm": 0.4260086119174957,
"learning_rate": 0.0002710017503702706,
"loss": 2.9748,
"step": 81550
},
{
"epoch": 21.959302325581394,
"grad_norm": 0.4102717339992523,
"learning_rate": 0.0002707997845698128,
"loss": 2.9875,
"step": 81600
},
{
"epoch": 21.972760551248925,
"grad_norm": 0.43519127368927,
"learning_rate": 0.000270597818769355,
"loss": 2.9921,
"step": 81650
},
{
"epoch": 21.986218776916452,
"grad_norm": 0.4127667546272278,
"learning_rate": 0.00027039585296889727,
"loss": 2.9809,
"step": 81700
},
{
"epoch": 21.99967700258398,
"grad_norm": 0.4082183241844177,
"learning_rate": 0.00027019388716843947,
"loss": 2.9917,
"step": 81750
},
{
"epoch": 22.01318906115418,
"grad_norm": 0.38447314500808716,
"learning_rate": 0.00026999192136798166,
"loss": 2.9948,
"step": 81800
},
{
"epoch": 22.026647286821706,
"grad_norm": 0.3957959711551666,
"learning_rate": 0.00026978995556752386,
"loss": 2.9274,
"step": 81850
},
{
"epoch": 22.040105512489234,
"grad_norm": 0.39288565516471863,
"learning_rate": 0.00026958798976706606,
"loss": 2.9345,
"step": 81900
},
{
"epoch": 22.05356373815676,
"grad_norm": 0.41844412684440613,
"learning_rate": 0.0002693860239666083,
"loss": 2.9465,
"step": 81950
},
{
"epoch": 22.06702196382429,
"grad_norm": 0.43770354986190796,
"learning_rate": 0.0002691840581661505,
"loss": 2.9453,
"step": 82000
},
{
"epoch": 22.06702196382429,
"eval_accuracy": 0.39549508851024734,
"eval_loss": 3.295212984085083,
"eval_runtime": 187.0052,
"eval_samples_per_second": 96.318,
"eval_steps_per_second": 6.021,
"step": 82000
},
{
"epoch": 22.080480189491816,
"grad_norm": 0.44374948740005493,
"learning_rate": 0.0002689820923656927,
"loss": 2.9405,
"step": 82050
},
{
"epoch": 22.093938415159347,
"grad_norm": 0.4129500985145569,
"learning_rate": 0.00026878012656523495,
"loss": 2.945,
"step": 82100
},
{
"epoch": 22.107396640826874,
"grad_norm": 0.434146910905838,
"learning_rate": 0.00026857816076477714,
"loss": 2.9551,
"step": 82150
},
{
"epoch": 22.1208548664944,
"grad_norm": 0.4277571737766266,
"learning_rate": 0.00026837619496431934,
"loss": 2.955,
"step": 82200
},
{
"epoch": 22.13431309216193,
"grad_norm": 0.41571956872940063,
"learning_rate": 0.0002681742291638616,
"loss": 2.965,
"step": 82250
},
{
"epoch": 22.147771317829456,
"grad_norm": 0.4457704424858093,
"learning_rate": 0.0002679722633634038,
"loss": 2.964,
"step": 82300
},
{
"epoch": 22.161229543496987,
"grad_norm": 0.41916701197624207,
"learning_rate": 0.000267770297562946,
"loss": 2.944,
"step": 82350
},
{
"epoch": 22.174687769164514,
"grad_norm": 0.43160828948020935,
"learning_rate": 0.0002675683317624882,
"loss": 2.9556,
"step": 82400
},
{
"epoch": 22.18814599483204,
"grad_norm": 0.456165075302124,
"learning_rate": 0.0002673663659620304,
"loss": 2.9579,
"step": 82450
},
{
"epoch": 22.20160422049957,
"grad_norm": 0.4302372336387634,
"learning_rate": 0.0002671644001615726,
"loss": 2.9569,
"step": 82500
},
{
"epoch": 22.215062446167096,
"grad_norm": 0.4164350628852844,
"learning_rate": 0.0002669624343611148,
"loss": 2.96,
"step": 82550
},
{
"epoch": 22.228520671834627,
"grad_norm": 0.41148215532302856,
"learning_rate": 0.00026676046856065707,
"loss": 2.9727,
"step": 82600
},
{
"epoch": 22.241978897502154,
"grad_norm": 0.41684970259666443,
"learning_rate": 0.00026655850276019927,
"loss": 2.9621,
"step": 82650
},
{
"epoch": 22.25543712316968,
"grad_norm": 0.43197721242904663,
"learning_rate": 0.00026635653695974146,
"loss": 2.97,
"step": 82700
},
{
"epoch": 22.26889534883721,
"grad_norm": 0.42877650260925293,
"learning_rate": 0.00026615457115928366,
"loss": 2.9657,
"step": 82750
},
{
"epoch": 22.282353574504736,
"grad_norm": 0.4146842360496521,
"learning_rate": 0.00026595260535882586,
"loss": 2.9772,
"step": 82800
},
{
"epoch": 22.295811800172267,
"grad_norm": 0.4308156371116638,
"learning_rate": 0.0002657506395583681,
"loss": 2.973,
"step": 82850
},
{
"epoch": 22.309270025839794,
"grad_norm": 0.4283556640148163,
"learning_rate": 0.0002655486737579103,
"loss": 2.9732,
"step": 82900
},
{
"epoch": 22.32272825150732,
"grad_norm": 0.4151584208011627,
"learning_rate": 0.00026534670795745255,
"loss": 2.9841,
"step": 82950
},
{
"epoch": 22.33618647717485,
"grad_norm": 0.419960618019104,
"learning_rate": 0.00026514474215699475,
"loss": 2.9769,
"step": 83000
},
{
"epoch": 22.33618647717485,
"eval_accuracy": 0.39613262825869944,
"eval_loss": 3.289202928543091,
"eval_runtime": 184.9311,
"eval_samples_per_second": 97.398,
"eval_steps_per_second": 6.089,
"step": 83000
},
{
"epoch": 22.349644702842376,
"grad_norm": 0.4388822913169861,
"learning_rate": 0.00026494277635653694,
"loss": 2.9617,
"step": 83050
},
{
"epoch": 22.363102928509907,
"grad_norm": 0.4281676411628723,
"learning_rate": 0.00026474081055607914,
"loss": 3.0034,
"step": 83100
},
{
"epoch": 22.376561154177434,
"grad_norm": 0.4084988534450531,
"learning_rate": 0.00026453884475562134,
"loss": 2.9767,
"step": 83150
},
{
"epoch": 22.39001937984496,
"grad_norm": 0.44110220670700073,
"learning_rate": 0.0002643368789551636,
"loss": 2.9756,
"step": 83200
},
{
"epoch": 22.40347760551249,
"grad_norm": 0.43258190155029297,
"learning_rate": 0.0002641349131547058,
"loss": 2.9762,
"step": 83250
},
{
"epoch": 22.416935831180016,
"grad_norm": 0.4122479557991028,
"learning_rate": 0.000263932947354248,
"loss": 2.9764,
"step": 83300
},
{
"epoch": 22.430394056847547,
"grad_norm": 0.43426617980003357,
"learning_rate": 0.0002637309815537902,
"loss": 2.985,
"step": 83350
},
{
"epoch": 22.443852282515074,
"grad_norm": 0.4156150817871094,
"learning_rate": 0.0002635290157533324,
"loss": 2.977,
"step": 83400
},
{
"epoch": 22.4573105081826,
"grad_norm": 0.40508025884628296,
"learning_rate": 0.0002633270499528746,
"loss": 2.9736,
"step": 83450
},
{
"epoch": 22.47076873385013,
"grad_norm": 0.4699075520038605,
"learning_rate": 0.0002631250841524168,
"loss": 2.9745,
"step": 83500
},
{
"epoch": 22.484226959517656,
"grad_norm": 0.44109007716178894,
"learning_rate": 0.00026292311835195907,
"loss": 2.9859,
"step": 83550
},
{
"epoch": 22.497685185185187,
"grad_norm": 0.42453140020370483,
"learning_rate": 0.00026272115255150126,
"loss": 2.9854,
"step": 83600
},
{
"epoch": 22.511143410852714,
"grad_norm": 0.4197845160961151,
"learning_rate": 0.00026251918675104346,
"loss": 2.9847,
"step": 83650
},
{
"epoch": 22.52460163652024,
"grad_norm": 0.42948266863822937,
"learning_rate": 0.00026231722095058566,
"loss": 2.9777,
"step": 83700
},
{
"epoch": 22.53805986218777,
"grad_norm": 0.4153449535369873,
"learning_rate": 0.00026211525515012785,
"loss": 2.9847,
"step": 83750
},
{
"epoch": 22.551518087855296,
"grad_norm": 0.4096290171146393,
"learning_rate": 0.0002619132893496701,
"loss": 2.9847,
"step": 83800
},
{
"epoch": 22.564976313522827,
"grad_norm": 0.406325101852417,
"learning_rate": 0.0002617113235492123,
"loss": 2.9865,
"step": 83850
},
{
"epoch": 22.578434539190354,
"grad_norm": 0.40534356236457825,
"learning_rate": 0.00026150935774875455,
"loss": 2.9926,
"step": 83900
},
{
"epoch": 22.59189276485788,
"grad_norm": 0.4309810698032379,
"learning_rate": 0.00026130739194829675,
"loss": 2.9961,
"step": 83950
},
{
"epoch": 22.60535099052541,
"grad_norm": 0.4095143675804138,
"learning_rate": 0.00026110542614783894,
"loss": 2.9852,
"step": 84000
},
{
"epoch": 22.60535099052541,
"eval_accuracy": 0.39617358812324727,
"eval_loss": 3.2858405113220215,
"eval_runtime": 184.2446,
"eval_samples_per_second": 97.761,
"eval_steps_per_second": 6.111,
"step": 84000
},
{
"epoch": 22.618809216192936,
"grad_norm": 0.4406158924102783,
"learning_rate": 0.00026090346034738114,
"loss": 2.9936,
"step": 84050
},
{
"epoch": 22.632267441860463,
"grad_norm": 0.4202785789966583,
"learning_rate": 0.0002607014945469234,
"loss": 2.9884,
"step": 84100
},
{
"epoch": 22.645725667527994,
"grad_norm": 0.4082145094871521,
"learning_rate": 0.0002604995287464656,
"loss": 3.0034,
"step": 84150
},
{
"epoch": 22.65918389319552,
"grad_norm": 0.4127369523048401,
"learning_rate": 0.0002602975629460078,
"loss": 2.991,
"step": 84200
},
{
"epoch": 22.67264211886305,
"grad_norm": 0.4268158972263336,
"learning_rate": 0.00026009559714555,
"loss": 2.998,
"step": 84250
},
{
"epoch": 22.686100344530576,
"grad_norm": 0.42113062739372253,
"learning_rate": 0.0002598936313450922,
"loss": 2.9931,
"step": 84300
},
{
"epoch": 22.699558570198104,
"grad_norm": 0.4232160747051239,
"learning_rate": 0.0002596916655446344,
"loss": 2.9946,
"step": 84350
},
{
"epoch": 22.713016795865634,
"grad_norm": 0.4746277928352356,
"learning_rate": 0.0002594896997441766,
"loss": 3.0051,
"step": 84400
},
{
"epoch": 22.72647502153316,
"grad_norm": 0.42749130725860596,
"learning_rate": 0.00025928773394371887,
"loss": 2.9987,
"step": 84450
},
{
"epoch": 22.73993324720069,
"grad_norm": 0.4101299047470093,
"learning_rate": 0.00025908576814326107,
"loss": 2.9997,
"step": 84500
},
{
"epoch": 22.753391472868216,
"grad_norm": 0.43315228819847107,
"learning_rate": 0.00025888380234280326,
"loss": 2.9871,
"step": 84550
},
{
"epoch": 22.766849698535744,
"grad_norm": 0.4385271966457367,
"learning_rate": 0.00025868183654234546,
"loss": 3.0094,
"step": 84600
},
{
"epoch": 22.780307924203274,
"grad_norm": 0.4066489040851593,
"learning_rate": 0.00025847987074188765,
"loss": 2.9988,
"step": 84650
},
{
"epoch": 22.7937661498708,
"grad_norm": 0.433079332113266,
"learning_rate": 0.0002582779049414299,
"loss": 2.9995,
"step": 84700
},
{
"epoch": 22.80722437553833,
"grad_norm": 0.4584760367870331,
"learning_rate": 0.0002580759391409721,
"loss": 3.0,
"step": 84750
},
{
"epoch": 22.820682601205856,
"grad_norm": 0.42474591732025146,
"learning_rate": 0.00025787397334051435,
"loss": 3.0108,
"step": 84800
},
{
"epoch": 22.834140826873384,
"grad_norm": 0.42554858326911926,
"learning_rate": 0.00025767200754005655,
"loss": 2.9813,
"step": 84850
},
{
"epoch": 22.847599052540914,
"grad_norm": 0.4180266559123993,
"learning_rate": 0.00025747004173959874,
"loss": 3.0009,
"step": 84900
},
{
"epoch": 22.861057278208442,
"grad_norm": 0.41329315304756165,
"learning_rate": 0.00025726807593914094,
"loss": 3.0092,
"step": 84950
},
{
"epoch": 22.87451550387597,
"grad_norm": 0.42833954095840454,
"learning_rate": 0.00025706611013868313,
"loss": 2.9832,
"step": 85000
},
{
"epoch": 22.87451550387597,
"eval_accuracy": 0.39637284645635246,
"eval_loss": 3.2803292274475098,
"eval_runtime": 184.2113,
"eval_samples_per_second": 97.779,
"eval_steps_per_second": 6.113,
"step": 85000
},
{
"epoch": 22.887973729543496,
"grad_norm": 0.4507509171962738,
"learning_rate": 0.0002568641443382254,
"loss": 3.0046,
"step": 85050
},
{
"epoch": 22.901431955211024,
"grad_norm": 0.41896918416023254,
"learning_rate": 0.0002566621785377676,
"loss": 3.0002,
"step": 85100
},
{
"epoch": 22.914890180878555,
"grad_norm": 0.43956953287124634,
"learning_rate": 0.0002564602127373098,
"loss": 2.9913,
"step": 85150
},
{
"epoch": 22.928348406546082,
"grad_norm": 0.4055449664592743,
"learning_rate": 0.00025625824693685203,
"loss": 3.0041,
"step": 85200
},
{
"epoch": 22.94180663221361,
"grad_norm": 0.4386661648750305,
"learning_rate": 0.0002560562811363942,
"loss": 3.0053,
"step": 85250
},
{
"epoch": 22.955264857881136,
"grad_norm": 0.39694416522979736,
"learning_rate": 0.0002558543153359364,
"loss": 3.0117,
"step": 85300
},
{
"epoch": 22.968723083548664,
"grad_norm": 0.42682603001594543,
"learning_rate": 0.0002556523495354786,
"loss": 2.9956,
"step": 85350
},
{
"epoch": 22.982181309216195,
"grad_norm": 0.46057724952697754,
"learning_rate": 0.00025545038373502087,
"loss": 3.0002,
"step": 85400
},
{
"epoch": 22.995639534883722,
"grad_norm": 0.44713103771209717,
"learning_rate": 0.00025524841793456306,
"loss": 3.0148,
"step": 85450
},
{
"epoch": 23.00888242894057,
"grad_norm": 0.42075687646865845,
"learning_rate": 0.00025504645213410526,
"loss": 2.9539,
"step": 85500
},
{
"epoch": 23.022340654608097,
"grad_norm": 0.44748881459236145,
"learning_rate": 0.00025484448633364745,
"loss": 2.9246,
"step": 85550
},
{
"epoch": 23.035798880275625,
"grad_norm": 0.4051612615585327,
"learning_rate": 0.00025464252053318965,
"loss": 2.9287,
"step": 85600
},
{
"epoch": 23.049257105943152,
"grad_norm": 0.42038244009017944,
"learning_rate": 0.0002544405547327319,
"loss": 2.9351,
"step": 85650
},
{
"epoch": 23.06271533161068,
"grad_norm": 0.4450347125530243,
"learning_rate": 0.0002542385889322741,
"loss": 2.9205,
"step": 85700
},
{
"epoch": 23.07617355727821,
"grad_norm": 0.400465190410614,
"learning_rate": 0.00025403662313181635,
"loss": 2.9305,
"step": 85750
},
{
"epoch": 23.089631782945737,
"grad_norm": 0.43276166915893555,
"learning_rate": 0.00025383465733135854,
"loss": 2.9302,
"step": 85800
},
{
"epoch": 23.103090008613265,
"grad_norm": 0.41941651701927185,
"learning_rate": 0.00025363269153090074,
"loss": 2.9367,
"step": 85850
},
{
"epoch": 23.116548234280792,
"grad_norm": 0.42673546075820923,
"learning_rate": 0.00025343072573044294,
"loss": 2.9403,
"step": 85900
},
{
"epoch": 23.13000645994832,
"grad_norm": 0.43960341811180115,
"learning_rate": 0.00025322875992998513,
"loss": 2.9334,
"step": 85950
},
{
"epoch": 23.14346468561585,
"grad_norm": 0.4396515488624573,
"learning_rate": 0.0002530267941295274,
"loss": 2.9363,
"step": 86000
},
{
"epoch": 23.14346468561585,
"eval_accuracy": 0.39621889386201764,
"eval_loss": 3.290853977203369,
"eval_runtime": 186.6339,
"eval_samples_per_second": 96.51,
"eval_steps_per_second": 6.033,
"step": 86000
},
{
"epoch": 23.156922911283377,
"grad_norm": 0.4387992322444916,
"learning_rate": 0.0002528248283290696,
"loss": 2.9435,
"step": 86050
},
{
"epoch": 23.170381136950905,
"grad_norm": 0.41315823793411255,
"learning_rate": 0.00025262286252861183,
"loss": 2.9381,
"step": 86100
},
{
"epoch": 23.183839362618432,
"grad_norm": 0.41604095697402954,
"learning_rate": 0.000252420896728154,
"loss": 2.9471,
"step": 86150
},
{
"epoch": 23.19729758828596,
"grad_norm": 0.41397571563720703,
"learning_rate": 0.0002522189309276962,
"loss": 2.948,
"step": 86200
},
{
"epoch": 23.210755813953487,
"grad_norm": 0.4200591742992401,
"learning_rate": 0.0002520169651272384,
"loss": 2.9545,
"step": 86250
},
{
"epoch": 23.224214039621017,
"grad_norm": 0.46514901518821716,
"learning_rate": 0.00025181499932678067,
"loss": 2.9564,
"step": 86300
},
{
"epoch": 23.237672265288545,
"grad_norm": 0.4545646905899048,
"learning_rate": 0.00025161303352632286,
"loss": 2.9484,
"step": 86350
},
{
"epoch": 23.251130490956072,
"grad_norm": 0.4002546966075897,
"learning_rate": 0.00025141106772586506,
"loss": 2.9538,
"step": 86400
},
{
"epoch": 23.2645887166236,
"grad_norm": 0.4405435025691986,
"learning_rate": 0.00025120910192540726,
"loss": 2.9565,
"step": 86450
},
{
"epoch": 23.278046942291127,
"grad_norm": 0.43282967805862427,
"learning_rate": 0.00025100713612494945,
"loss": 2.9634,
"step": 86500
},
{
"epoch": 23.291505167958658,
"grad_norm": 0.4273679852485657,
"learning_rate": 0.0002508051703244917,
"loss": 2.9652,
"step": 86550
},
{
"epoch": 23.304963393626185,
"grad_norm": 0.4182887673377991,
"learning_rate": 0.0002506032045240339,
"loss": 2.9558,
"step": 86600
},
{
"epoch": 23.318421619293712,
"grad_norm": 0.4365881681442261,
"learning_rate": 0.00025040123872357615,
"loss": 2.9625,
"step": 86650
},
{
"epoch": 23.33187984496124,
"grad_norm": 0.44422706961631775,
"learning_rate": 0.00025019927292311834,
"loss": 2.9655,
"step": 86700
},
{
"epoch": 23.345338070628767,
"grad_norm": 0.4393934905529022,
"learning_rate": 0.00024999730712266054,
"loss": 2.9558,
"step": 86750
},
{
"epoch": 23.358796296296298,
"grad_norm": 0.4141547679901123,
"learning_rate": 0.00024979534132220274,
"loss": 2.9621,
"step": 86800
},
{
"epoch": 23.372254521963825,
"grad_norm": 0.47132158279418945,
"learning_rate": 0.00024959337552174493,
"loss": 2.9621,
"step": 86850
},
{
"epoch": 23.385712747631352,
"grad_norm": 0.4127795994281769,
"learning_rate": 0.0002493914097212872,
"loss": 2.9646,
"step": 86900
},
{
"epoch": 23.39917097329888,
"grad_norm": 0.4307352602481842,
"learning_rate": 0.0002491894439208294,
"loss": 2.9545,
"step": 86950
},
{
"epoch": 23.412629198966407,
"grad_norm": 0.41616472601890564,
"learning_rate": 0.00024898747812037163,
"loss": 2.9663,
"step": 87000
},
{
"epoch": 23.412629198966407,
"eval_accuracy": 0.3963653498233185,
"eval_loss": 3.2865498065948486,
"eval_runtime": 186.8168,
"eval_samples_per_second": 96.415,
"eval_steps_per_second": 6.027,
"step": 87000
},
{
"epoch": 23.426087424633938,
"grad_norm": 0.4423924386501312,
"learning_rate": 0.0002487855123199138,
"loss": 2.9568,
"step": 87050
},
{
"epoch": 23.439545650301465,
"grad_norm": 0.4344804584980011,
"learning_rate": 0.000248583546519456,
"loss": 2.9651,
"step": 87100
},
{
"epoch": 23.453003875968992,
"grad_norm": 0.45271238684654236,
"learning_rate": 0.0002483815807189982,
"loss": 2.9647,
"step": 87150
},
{
"epoch": 23.46646210163652,
"grad_norm": 0.43850594758987427,
"learning_rate": 0.0002481796149185404,
"loss": 2.9678,
"step": 87200
},
{
"epoch": 23.479920327304047,
"grad_norm": 0.4547707438468933,
"learning_rate": 0.00024797764911808266,
"loss": 2.973,
"step": 87250
},
{
"epoch": 23.493378552971578,
"grad_norm": 0.4392283856868744,
"learning_rate": 0.00024777568331762486,
"loss": 2.9769,
"step": 87300
},
{
"epoch": 23.506836778639105,
"grad_norm": 0.4168759882450104,
"learning_rate": 0.00024757371751716706,
"loss": 2.9727,
"step": 87350
},
{
"epoch": 23.520295004306632,
"grad_norm": 0.42110675573349,
"learning_rate": 0.0002473717517167093,
"loss": 2.9756,
"step": 87400
},
{
"epoch": 23.53375322997416,
"grad_norm": 0.4109584093093872,
"learning_rate": 0.0002471697859162515,
"loss": 2.9736,
"step": 87450
},
{
"epoch": 23.547211455641687,
"grad_norm": 0.43613022565841675,
"learning_rate": 0.0002469678201157937,
"loss": 2.978,
"step": 87500
},
{
"epoch": 23.560669681309218,
"grad_norm": 0.41340798139572144,
"learning_rate": 0.0002467658543153359,
"loss": 2.9725,
"step": 87550
},
{
"epoch": 23.574127906976745,
"grad_norm": 0.4117782711982727,
"learning_rate": 0.00024656388851487815,
"loss": 2.973,
"step": 87600
},
{
"epoch": 23.587586132644272,
"grad_norm": 0.4211580157279968,
"learning_rate": 0.00024636192271442034,
"loss": 2.977,
"step": 87650
},
{
"epoch": 23.6010443583118,
"grad_norm": 0.4073871374130249,
"learning_rate": 0.00024615995691396254,
"loss": 2.972,
"step": 87700
},
{
"epoch": 23.614502583979327,
"grad_norm": 0.43755197525024414,
"learning_rate": 0.00024595799111350473,
"loss": 2.9778,
"step": 87750
},
{
"epoch": 23.627960809646858,
"grad_norm": 0.43448543548583984,
"learning_rate": 0.00024575602531304693,
"loss": 2.9928,
"step": 87800
},
{
"epoch": 23.641419035314385,
"grad_norm": 0.42906805872917175,
"learning_rate": 0.0002455540595125892,
"loss": 2.982,
"step": 87850
},
{
"epoch": 23.654877260981912,
"grad_norm": 0.46099740266799927,
"learning_rate": 0.00024535209371213143,
"loss": 2.978,
"step": 87900
},
{
"epoch": 23.66833548664944,
"grad_norm": 0.4394725561141968,
"learning_rate": 0.0002451501279116736,
"loss": 2.9826,
"step": 87950
},
{
"epoch": 23.681793712316967,
"grad_norm": 0.42354512214660645,
"learning_rate": 0.0002449481621112158,
"loss": 2.9933,
"step": 88000
},
{
"epoch": 23.681793712316967,
"eval_accuracy": 0.39634068698710534,
"eval_loss": 3.283106803894043,
"eval_runtime": 187.6295,
"eval_samples_per_second": 95.998,
"eval_steps_per_second": 6.001,
"step": 88000
},
{
"epoch": 23.695251937984494,
"grad_norm": 0.4249362647533417,
"learning_rate": 0.000244746196310758,
"loss": 2.9775,
"step": 88050
},
{
"epoch": 23.708710163652025,
"grad_norm": 0.4240496754646301,
"learning_rate": 0.0002445442305103002,
"loss": 2.9888,
"step": 88100
},
{
"epoch": 23.722168389319553,
"grad_norm": 0.44241058826446533,
"learning_rate": 0.00024434226470984246,
"loss": 2.9669,
"step": 88150
},
{
"epoch": 23.73562661498708,
"grad_norm": 0.41765931248664856,
"learning_rate": 0.00024414029890938463,
"loss": 2.982,
"step": 88200
},
{
"epoch": 23.749084840654607,
"grad_norm": 0.43778789043426514,
"learning_rate": 0.00024393833310892686,
"loss": 2.986,
"step": 88250
},
{
"epoch": 23.762543066322134,
"grad_norm": 0.4122593104839325,
"learning_rate": 0.0002437363673084691,
"loss": 2.9838,
"step": 88300
},
{
"epoch": 23.776001291989665,
"grad_norm": 0.4394925534725189,
"learning_rate": 0.0002435344015080113,
"loss": 2.9811,
"step": 88350
},
{
"epoch": 23.789459517657193,
"grad_norm": 0.4328765571117401,
"learning_rate": 0.0002433324357075535,
"loss": 2.991,
"step": 88400
},
{
"epoch": 23.80291774332472,
"grad_norm": 0.42027604579925537,
"learning_rate": 0.00024313046990709572,
"loss": 2.9966,
"step": 88450
},
{
"epoch": 23.816375968992247,
"grad_norm": 0.44125255942344666,
"learning_rate": 0.00024292850410663792,
"loss": 2.9819,
"step": 88500
},
{
"epoch": 23.829834194659774,
"grad_norm": 0.4372418224811554,
"learning_rate": 0.00024272653830618014,
"loss": 2.9924,
"step": 88550
},
{
"epoch": 23.843292420327305,
"grad_norm": 0.4453098475933075,
"learning_rate": 0.00024252457250572234,
"loss": 2.9865,
"step": 88600
},
{
"epoch": 23.856750645994833,
"grad_norm": 0.42281657457351685,
"learning_rate": 0.00024232260670526453,
"loss": 2.995,
"step": 88650
},
{
"epoch": 23.87020887166236,
"grad_norm": 0.4352130591869354,
"learning_rate": 0.00024212064090480676,
"loss": 2.99,
"step": 88700
},
{
"epoch": 23.883667097329887,
"grad_norm": 0.4067487120628357,
"learning_rate": 0.00024191867510434898,
"loss": 2.9879,
"step": 88750
},
{
"epoch": 23.897125322997415,
"grad_norm": 0.42826223373413086,
"learning_rate": 0.0002417167093038912,
"loss": 2.9809,
"step": 88800
},
{
"epoch": 23.910583548664945,
"grad_norm": 0.44276687502861023,
"learning_rate": 0.0002415147435034334,
"loss": 2.9942,
"step": 88850
},
{
"epoch": 23.924041774332473,
"grad_norm": 0.4136715531349182,
"learning_rate": 0.00024131277770297562,
"loss": 2.992,
"step": 88900
},
{
"epoch": 23.9375,
"grad_norm": 0.4275543987751007,
"learning_rate": 0.00024111081190251782,
"loss": 2.9999,
"step": 88950
},
{
"epoch": 23.950958225667527,
"grad_norm": 0.40405896306037903,
"learning_rate": 0.00024090884610206002,
"loss": 2.9898,
"step": 89000
},
{
"epoch": 23.950958225667527,
"eval_accuracy": 0.3971029533257454,
"eval_loss": 3.2765560150146484,
"eval_runtime": 187.9412,
"eval_samples_per_second": 95.838,
"eval_steps_per_second": 5.991,
"step": 89000
},
{
"epoch": 23.964416451335055,
"grad_norm": 0.45323434472084045,
"learning_rate": 0.00024070688030160224,
"loss": 2.9921,
"step": 89050
},
{
"epoch": 23.977874677002585,
"grad_norm": 0.42327436804771423,
"learning_rate": 0.00024050491450114443,
"loss": 3.0075,
"step": 89100
},
{
"epoch": 23.991332902670113,
"grad_norm": 0.4318790137767792,
"learning_rate": 0.00024030294870068666,
"loss": 2.9982,
"step": 89150
},
{
"epoch": 24.00457579672696,
"grad_norm": 0.452729195356369,
"learning_rate": 0.00024010098290022888,
"loss": 2.9709,
"step": 89200
},
{
"epoch": 24.018034022394488,
"grad_norm": 0.4201224446296692,
"learning_rate": 0.0002398990170997711,
"loss": 2.9096,
"step": 89250
},
{
"epoch": 24.031492248062015,
"grad_norm": 0.42149022221565247,
"learning_rate": 0.0002396970512993133,
"loss": 2.9137,
"step": 89300
},
{
"epoch": 24.044950473729543,
"grad_norm": 0.43318113684654236,
"learning_rate": 0.00023949508549885552,
"loss": 2.9129,
"step": 89350
},
{
"epoch": 24.05840869939707,
"grad_norm": 0.44143423438072205,
"learning_rate": 0.00023929311969839772,
"loss": 2.9172,
"step": 89400
},
{
"epoch": 24.0718669250646,
"grad_norm": 0.4444623291492462,
"learning_rate": 0.00023909115389793992,
"loss": 2.9116,
"step": 89450
},
{
"epoch": 24.085325150732128,
"grad_norm": 0.4420110583305359,
"learning_rate": 0.00023888918809748214,
"loss": 2.9245,
"step": 89500
},
{
"epoch": 24.098783376399656,
"grad_norm": 0.4289356470108032,
"learning_rate": 0.00023868722229702433,
"loss": 2.9165,
"step": 89550
},
{
"epoch": 24.112241602067183,
"grad_norm": 0.45878368616104126,
"learning_rate": 0.00023848525649656656,
"loss": 2.9269,
"step": 89600
},
{
"epoch": 24.12569982773471,
"grad_norm": 0.4211243987083435,
"learning_rate": 0.00023828329069610878,
"loss": 2.9261,
"step": 89650
},
{
"epoch": 24.13915805340224,
"grad_norm": 0.45009878277778625,
"learning_rate": 0.000238081324895651,
"loss": 2.933,
"step": 89700
},
{
"epoch": 24.15261627906977,
"grad_norm": 0.4429611265659332,
"learning_rate": 0.0002378793590951932,
"loss": 2.9334,
"step": 89750
},
{
"epoch": 24.166074504737296,
"grad_norm": 0.43477439880371094,
"learning_rate": 0.0002376773932947354,
"loss": 2.9337,
"step": 89800
},
{
"epoch": 24.179532730404823,
"grad_norm": 0.4752548933029175,
"learning_rate": 0.00023747542749427762,
"loss": 2.9455,
"step": 89850
},
{
"epoch": 24.19299095607235,
"grad_norm": 0.45708489418029785,
"learning_rate": 0.00023727346169381982,
"loss": 2.9402,
"step": 89900
},
{
"epoch": 24.20644918173988,
"grad_norm": 0.425158828496933,
"learning_rate": 0.00023707149589336204,
"loss": 2.9452,
"step": 89950
},
{
"epoch": 24.21990740740741,
"grad_norm": 0.431226521730423,
"learning_rate": 0.00023686953009290424,
"loss": 2.9358,
"step": 90000
},
{
"epoch": 24.21990740740741,
"eval_accuracy": 0.3965300584563542,
"eval_loss": 3.2868492603302,
"eval_runtime": 183.5971,
"eval_samples_per_second": 98.106,
"eval_steps_per_second": 6.133,
"step": 90000
},
{
"epoch": 24.233365633074936,
"grad_norm": 0.4395337402820587,
"learning_rate": 0.00023666756429244643,
"loss": 2.943,
"step": 90050
},
{
"epoch": 24.246823858742463,
"grad_norm": 0.40836101770401,
"learning_rate": 0.00023646559849198868,
"loss": 2.9445,
"step": 90100
},
{
"epoch": 24.26028208440999,
"grad_norm": 0.4602183401584625,
"learning_rate": 0.00023626363269153088,
"loss": 2.945,
"step": 90150
},
{
"epoch": 24.27374031007752,
"grad_norm": 0.4670216143131256,
"learning_rate": 0.0002360616668910731,
"loss": 2.9451,
"step": 90200
},
{
"epoch": 24.28719853574505,
"grad_norm": 0.44003626704216003,
"learning_rate": 0.0002358597010906153,
"loss": 2.9481,
"step": 90250
},
{
"epoch": 24.300656761412576,
"grad_norm": 0.4452160894870758,
"learning_rate": 0.00023565773529015752,
"loss": 2.9469,
"step": 90300
},
{
"epoch": 24.314114987080103,
"grad_norm": 0.45113688707351685,
"learning_rate": 0.00023545576948969972,
"loss": 2.9458,
"step": 90350
},
{
"epoch": 24.32757321274763,
"grad_norm": 0.44268161058425903,
"learning_rate": 0.00023525380368924194,
"loss": 2.9528,
"step": 90400
},
{
"epoch": 24.341031438415158,
"grad_norm": 0.41929733753204346,
"learning_rate": 0.00023505183788878414,
"loss": 2.9475,
"step": 90450
},
{
"epoch": 24.35448966408269,
"grad_norm": 0.42923370003700256,
"learning_rate": 0.00023484987208832633,
"loss": 2.9589,
"step": 90500
},
{
"epoch": 24.367947889750216,
"grad_norm": 0.43205559253692627,
"learning_rate": 0.00023464790628786858,
"loss": 2.9562,
"step": 90550
},
{
"epoch": 24.381406115417743,
"grad_norm": 0.4304930567741394,
"learning_rate": 0.00023444594048741078,
"loss": 2.9503,
"step": 90600
},
{
"epoch": 24.39486434108527,
"grad_norm": 0.4301570653915405,
"learning_rate": 0.000234243974686953,
"loss": 2.9656,
"step": 90650
},
{
"epoch": 24.408322566752798,
"grad_norm": 0.42494460940361023,
"learning_rate": 0.0002340420088864952,
"loss": 2.9501,
"step": 90700
},
{
"epoch": 24.42178079242033,
"grad_norm": 0.44443538784980774,
"learning_rate": 0.00023384004308603742,
"loss": 2.9579,
"step": 90750
},
{
"epoch": 24.435239018087856,
"grad_norm": 0.41315990686416626,
"learning_rate": 0.00023363807728557962,
"loss": 2.9534,
"step": 90800
},
{
"epoch": 24.448697243755383,
"grad_norm": 0.4209133982658386,
"learning_rate": 0.0002334361114851218,
"loss": 2.9546,
"step": 90850
},
{
"epoch": 24.46215546942291,
"grad_norm": 0.43274033069610596,
"learning_rate": 0.00023323414568466404,
"loss": 2.9719,
"step": 90900
},
{
"epoch": 24.475613695090438,
"grad_norm": 0.42735084891319275,
"learning_rate": 0.00023303217988420623,
"loss": 2.9464,
"step": 90950
},
{
"epoch": 24.48907192075797,
"grad_norm": 0.3951086699962616,
"learning_rate": 0.00023283021408374848,
"loss": 2.9564,
"step": 91000
},
{
"epoch": 24.48907192075797,
"eval_accuracy": 0.3966551109871088,
"eval_loss": 3.284184455871582,
"eval_runtime": 182.7703,
"eval_samples_per_second": 98.55,
"eval_steps_per_second": 6.161,
"step": 91000
},
{
"epoch": 24.502530146425496,
"grad_norm": 0.4607436954975128,
"learning_rate": 0.00023262824828329068,
"loss": 2.9641,
"step": 91050
},
{
"epoch": 24.515988372093023,
"grad_norm": 0.45036646723747253,
"learning_rate": 0.0002324262824828329,
"loss": 2.9667,
"step": 91100
},
{
"epoch": 24.52944659776055,
"grad_norm": 0.4311272203922272,
"learning_rate": 0.0002322243166823751,
"loss": 2.9692,
"step": 91150
},
{
"epoch": 24.542904823428078,
"grad_norm": 0.4353450536727905,
"learning_rate": 0.0002320223508819173,
"loss": 2.9553,
"step": 91200
},
{
"epoch": 24.55636304909561,
"grad_norm": 0.43171682953834534,
"learning_rate": 0.00023182038508145952,
"loss": 2.9581,
"step": 91250
},
{
"epoch": 24.569821274763136,
"grad_norm": 0.4325876235961914,
"learning_rate": 0.0002316184192810017,
"loss": 2.9578,
"step": 91300
},
{
"epoch": 24.583279500430663,
"grad_norm": 0.4323505759239197,
"learning_rate": 0.00023141645348054394,
"loss": 2.9723,
"step": 91350
},
{
"epoch": 24.59673772609819,
"grad_norm": 0.4744899868965149,
"learning_rate": 0.00023121448768008613,
"loss": 2.9752,
"step": 91400
},
{
"epoch": 24.610195951765718,
"grad_norm": 0.4508267641067505,
"learning_rate": 0.00023101252187962838,
"loss": 2.9733,
"step": 91450
},
{
"epoch": 24.62365417743325,
"grad_norm": 0.41971635818481445,
"learning_rate": 0.00023081055607917058,
"loss": 2.9681,
"step": 91500
},
{
"epoch": 24.637112403100776,
"grad_norm": 0.4390505850315094,
"learning_rate": 0.0002306085902787128,
"loss": 2.9751,
"step": 91550
},
{
"epoch": 24.650570628768303,
"grad_norm": 0.46359091997146606,
"learning_rate": 0.000230406624478255,
"loss": 2.9661,
"step": 91600
},
{
"epoch": 24.66402885443583,
"grad_norm": 0.4584541320800781,
"learning_rate": 0.0002302046586777972,
"loss": 2.9598,
"step": 91650
},
{
"epoch": 24.677487080103358,
"grad_norm": 0.45301374793052673,
"learning_rate": 0.00023000269287733942,
"loss": 2.9762,
"step": 91700
},
{
"epoch": 24.69094530577089,
"grad_norm": 0.42839041352272034,
"learning_rate": 0.00022980072707688161,
"loss": 2.9657,
"step": 91750
},
{
"epoch": 24.704403531438416,
"grad_norm": 0.4463431239128113,
"learning_rate": 0.00022959876127642384,
"loss": 2.9762,
"step": 91800
},
{
"epoch": 24.717861757105943,
"grad_norm": 0.4341801106929779,
"learning_rate": 0.00022939679547596603,
"loss": 2.9696,
"step": 91850
},
{
"epoch": 24.73131998277347,
"grad_norm": 0.42408186197280884,
"learning_rate": 0.00022919482967550828,
"loss": 2.9741,
"step": 91900
},
{
"epoch": 24.744778208440998,
"grad_norm": 0.46379968523979187,
"learning_rate": 0.00022899286387505048,
"loss": 2.9831,
"step": 91950
},
{
"epoch": 24.75823643410853,
"grad_norm": 0.4441967308521271,
"learning_rate": 0.00022879089807459268,
"loss": 2.985,
"step": 92000
},
{
"epoch": 24.75823643410853,
"eval_accuracy": 0.3971473898896713,
"eval_loss": 3.278430461883545,
"eval_runtime": 214.8332,
"eval_samples_per_second": 83.842,
"eval_steps_per_second": 5.241,
"step": 92000
},
{
"epoch": 24.771694659776056,
"grad_norm": 0.4211469292640686,
"learning_rate": 0.0002285889322741349,
"loss": 2.9677,
"step": 92050
},
{
"epoch": 24.785152885443583,
"grad_norm": 0.42314404249191284,
"learning_rate": 0.0002283869664736771,
"loss": 2.9743,
"step": 92100
},
{
"epoch": 24.79861111111111,
"grad_norm": 0.4325233995914459,
"learning_rate": 0.00022818500067321932,
"loss": 2.967,
"step": 92150
},
{
"epoch": 24.812069336778638,
"grad_norm": 0.45030513405799866,
"learning_rate": 0.00022798303487276151,
"loss": 2.9838,
"step": 92200
},
{
"epoch": 24.82552756244617,
"grad_norm": 0.4559642970561981,
"learning_rate": 0.0002277810690723037,
"loss": 2.9691,
"step": 92250
},
{
"epoch": 24.838985788113696,
"grad_norm": 0.4240502119064331,
"learning_rate": 0.00022757910327184596,
"loss": 2.9725,
"step": 92300
},
{
"epoch": 24.852444013781223,
"grad_norm": 0.42013439536094666,
"learning_rate": 0.00022737713747138818,
"loss": 2.9772,
"step": 92350
},
{
"epoch": 24.86590223944875,
"grad_norm": 0.42543137073516846,
"learning_rate": 0.00022717517167093038,
"loss": 2.9784,
"step": 92400
},
{
"epoch": 24.879360465116278,
"grad_norm": 0.4418637156486511,
"learning_rate": 0.00022697320587047258,
"loss": 2.9745,
"step": 92450
},
{
"epoch": 24.892818690783805,
"grad_norm": 0.431985080242157,
"learning_rate": 0.0002267712400700148,
"loss": 2.9792,
"step": 92500
},
{
"epoch": 24.906276916451336,
"grad_norm": 0.42724892497062683,
"learning_rate": 0.000226569274269557,
"loss": 2.9803,
"step": 92550
},
{
"epoch": 24.919735142118864,
"grad_norm": 0.4235786199569702,
"learning_rate": 0.00022636730846909922,
"loss": 2.9854,
"step": 92600
},
{
"epoch": 24.93319336778639,
"grad_norm": 0.41270336508750916,
"learning_rate": 0.00022616534266864141,
"loss": 2.9741,
"step": 92650
},
{
"epoch": 24.946651593453918,
"grad_norm": 0.42850589752197266,
"learning_rate": 0.0002259633768681836,
"loss": 2.9845,
"step": 92700
},
{
"epoch": 24.960109819121445,
"grad_norm": 0.4491378962993622,
"learning_rate": 0.00022576141106772586,
"loss": 2.983,
"step": 92750
},
{
"epoch": 24.973568044788976,
"grad_norm": 0.4432944655418396,
"learning_rate": 0.00022555944526726806,
"loss": 2.9883,
"step": 92800
},
{
"epoch": 24.987026270456504,
"grad_norm": 0.43878373503685,
"learning_rate": 0.00022535747946681028,
"loss": 2.9847,
"step": 92850
},
{
"epoch": 25.00026916451335,
"grad_norm": 0.4839756488800049,
"learning_rate": 0.00022515551366635248,
"loss": 2.9812,
"step": 92900
},
{
"epoch": 25.01372739018088,
"grad_norm": 0.45553848147392273,
"learning_rate": 0.0002249535478658947,
"loss": 2.9025,
"step": 92950
},
{
"epoch": 25.027185615848406,
"grad_norm": 0.4480719268321991,
"learning_rate": 0.0002247515820654369,
"loss": 2.9033,
"step": 93000
},
{
"epoch": 25.027185615848406,
"eval_accuracy": 0.39642836499954587,
"eval_loss": 3.287465810775757,
"eval_runtime": 182.9652,
"eval_samples_per_second": 98.445,
"eval_steps_per_second": 6.154,
"step": 93000
},
{
"epoch": 25.040643841515934,
"grad_norm": 0.44877710938453674,
"learning_rate": 0.0002245496162649791,
"loss": 2.9082,
"step": 93050
},
{
"epoch": 25.05410206718346,
"grad_norm": 0.44390127062797546,
"learning_rate": 0.00022434765046452131,
"loss": 2.907,
"step": 93100
},
{
"epoch": 25.06756029285099,
"grad_norm": 0.4346546530723572,
"learning_rate": 0.0002241456846640635,
"loss": 2.9031,
"step": 93150
},
{
"epoch": 25.08101851851852,
"grad_norm": 0.4312244951725006,
"learning_rate": 0.00022394371886360576,
"loss": 2.9066,
"step": 93200
},
{
"epoch": 25.094476744186046,
"grad_norm": 0.43850862979888916,
"learning_rate": 0.00022374175306314796,
"loss": 2.9064,
"step": 93250
},
{
"epoch": 25.107934969853574,
"grad_norm": 0.46429237723350525,
"learning_rate": 0.00022353978726269018,
"loss": 2.914,
"step": 93300
},
{
"epoch": 25.1213931955211,
"grad_norm": 0.44453561305999756,
"learning_rate": 0.00022333782146223238,
"loss": 2.915,
"step": 93350
},
{
"epoch": 25.134851421188632,
"grad_norm": 0.43875470757484436,
"learning_rate": 0.0002231358556617746,
"loss": 2.9203,
"step": 93400
},
{
"epoch": 25.14830964685616,
"grad_norm": 0.42957693338394165,
"learning_rate": 0.0002229338898613168,
"loss": 2.9164,
"step": 93450
},
{
"epoch": 25.161767872523686,
"grad_norm": 0.43665623664855957,
"learning_rate": 0.000222731924060859,
"loss": 2.9255,
"step": 93500
},
{
"epoch": 25.175226098191214,
"grad_norm": 0.42491114139556885,
"learning_rate": 0.00022252995826040122,
"loss": 2.922,
"step": 93550
},
{
"epoch": 25.18868432385874,
"grad_norm": 0.45817166566848755,
"learning_rate": 0.0002223279924599434,
"loss": 2.9396,
"step": 93600
},
{
"epoch": 25.202142549526272,
"grad_norm": 0.48348522186279297,
"learning_rate": 0.00022212602665948566,
"loss": 2.9259,
"step": 93650
},
{
"epoch": 25.2156007751938,
"grad_norm": 0.45023709535598755,
"learning_rate": 0.00022192406085902786,
"loss": 2.9211,
"step": 93700
},
{
"epoch": 25.229059000861326,
"grad_norm": 0.4651678502559662,
"learning_rate": 0.00022172209505857008,
"loss": 2.9247,
"step": 93750
},
{
"epoch": 25.242517226528854,
"grad_norm": 0.4332248270511627,
"learning_rate": 0.00022152012925811228,
"loss": 2.9311,
"step": 93800
},
{
"epoch": 25.25597545219638,
"grad_norm": 0.44667574763298035,
"learning_rate": 0.00022131816345765447,
"loss": 2.936,
"step": 93850
},
{
"epoch": 25.269433677863912,
"grad_norm": 0.44624167680740356,
"learning_rate": 0.0002211161976571967,
"loss": 2.9473,
"step": 93900
},
{
"epoch": 25.28289190353144,
"grad_norm": 0.45082974433898926,
"learning_rate": 0.0002209142318567389,
"loss": 2.9413,
"step": 93950
},
{
"epoch": 25.296350129198967,
"grad_norm": 0.45720174908638,
"learning_rate": 0.00022071226605628112,
"loss": 2.9328,
"step": 94000
},
{
"epoch": 25.296350129198967,
"eval_accuracy": 0.39647671285027203,
"eval_loss": 3.2913010120391846,
"eval_runtime": 182.9541,
"eval_samples_per_second": 98.451,
"eval_steps_per_second": 6.155,
"step": 94000
},
{
"epoch": 25.309808354866494,
"grad_norm": 0.43020617961883545,
"learning_rate": 0.0002205103002558233,
"loss": 2.9343,
"step": 94050
},
{
"epoch": 25.32326658053402,
"grad_norm": 0.45349353551864624,
"learning_rate": 0.00022030833445536556,
"loss": 2.9422,
"step": 94100
},
{
"epoch": 25.336724806201552,
"grad_norm": 0.47829487919807434,
"learning_rate": 0.00022010636865490776,
"loss": 2.9334,
"step": 94150
},
{
"epoch": 25.35018303186908,
"grad_norm": 0.44192102551460266,
"learning_rate": 0.00021990440285444998,
"loss": 2.9473,
"step": 94200
},
{
"epoch": 25.363641257536607,
"grad_norm": 0.4501858055591583,
"learning_rate": 0.00021970243705399218,
"loss": 2.9399,
"step": 94250
},
{
"epoch": 25.377099483204134,
"grad_norm": 0.43409064412117004,
"learning_rate": 0.00021950047125353437,
"loss": 2.9386,
"step": 94300
},
{
"epoch": 25.39055770887166,
"grad_norm": 0.46234411001205444,
"learning_rate": 0.0002192985054530766,
"loss": 2.9472,
"step": 94350
},
{
"epoch": 25.404015934539192,
"grad_norm": 0.48046693205833435,
"learning_rate": 0.0002190965396526188,
"loss": 2.9519,
"step": 94400
},
{
"epoch": 25.41747416020672,
"grad_norm": 0.44661945104599,
"learning_rate": 0.00021889457385216102,
"loss": 2.94,
"step": 94450
},
{
"epoch": 25.430932385874247,
"grad_norm": 0.44060006737709045,
"learning_rate": 0.0002186926080517032,
"loss": 2.9422,
"step": 94500
},
{
"epoch": 25.444390611541774,
"grad_norm": 0.4293282926082611,
"learning_rate": 0.00021849064225124546,
"loss": 2.9414,
"step": 94550
},
{
"epoch": 25.4578488372093,
"grad_norm": 0.43087467551231384,
"learning_rate": 0.00021828867645078766,
"loss": 2.9452,
"step": 94600
},
{
"epoch": 25.471307062876832,
"grad_norm": 0.4379226267337799,
"learning_rate": 0.00021808671065032985,
"loss": 2.9603,
"step": 94650
},
{
"epoch": 25.48476528854436,
"grad_norm": 0.45308583974838257,
"learning_rate": 0.00021788474484987208,
"loss": 2.9455,
"step": 94700
},
{
"epoch": 25.498223514211887,
"grad_norm": 0.4435552954673767,
"learning_rate": 0.00021768277904941427,
"loss": 2.9515,
"step": 94750
},
{
"epoch": 25.511681739879414,
"grad_norm": 0.46348798274993896,
"learning_rate": 0.0002174808132489565,
"loss": 2.9564,
"step": 94800
},
{
"epoch": 25.52513996554694,
"grad_norm": 0.42802226543426514,
"learning_rate": 0.0002172788474484987,
"loss": 2.9541,
"step": 94850
},
{
"epoch": 25.53859819121447,
"grad_norm": 0.46017369627952576,
"learning_rate": 0.0002170768816480409,
"loss": 2.9664,
"step": 94900
},
{
"epoch": 25.552056416882,
"grad_norm": 0.443142294883728,
"learning_rate": 0.0002168749158475831,
"loss": 2.9441,
"step": 94950
},
{
"epoch": 25.565514642549527,
"grad_norm": 0.45258423686027527,
"learning_rate": 0.00021667295004712534,
"loss": 2.9515,
"step": 95000
},
{
"epoch": 25.565514642549527,
"eval_accuracy": 0.39699017788966956,
"eval_loss": 3.279651165008545,
"eval_runtime": 187.5815,
"eval_samples_per_second": 96.022,
"eval_steps_per_second": 6.003,
"step": 95000
},
{
"epoch": 25.578972868217054,
"grad_norm": 0.4450850784778595,
"learning_rate": 0.00021647098424666756,
"loss": 2.9651,
"step": 95050
},
{
"epoch": 25.59243109388458,
"grad_norm": 0.4583265483379364,
"learning_rate": 0.00021626901844620976,
"loss": 2.9529,
"step": 95100
},
{
"epoch": 25.60588931955211,
"grad_norm": 0.4328291118144989,
"learning_rate": 0.00021606705264575198,
"loss": 2.9595,
"step": 95150
},
{
"epoch": 25.61934754521964,
"grad_norm": 0.43307721614837646,
"learning_rate": 0.00021586508684529417,
"loss": 2.9587,
"step": 95200
},
{
"epoch": 25.632805770887167,
"grad_norm": 0.47918468713760376,
"learning_rate": 0.0002156631210448364,
"loss": 2.9634,
"step": 95250
},
{
"epoch": 25.646263996554694,
"grad_norm": 0.4483031630516052,
"learning_rate": 0.0002154611552443786,
"loss": 2.9624,
"step": 95300
},
{
"epoch": 25.65972222222222,
"grad_norm": 0.453348845243454,
"learning_rate": 0.0002152591894439208,
"loss": 2.965,
"step": 95350
},
{
"epoch": 25.67318044788975,
"grad_norm": 0.4389757215976715,
"learning_rate": 0.000215057223643463,
"loss": 2.9607,
"step": 95400
},
{
"epoch": 25.68663867355728,
"grad_norm": 0.4491554796695709,
"learning_rate": 0.00021485525784300524,
"loss": 2.9643,
"step": 95450
},
{
"epoch": 25.700096899224807,
"grad_norm": 0.43662896752357483,
"learning_rate": 0.00021465329204254746,
"loss": 2.9584,
"step": 95500
},
{
"epoch": 25.713555124892334,
"grad_norm": 0.45692935585975647,
"learning_rate": 0.00021445132624208966,
"loss": 2.9557,
"step": 95550
},
{
"epoch": 25.72701335055986,
"grad_norm": 0.45086410641670227,
"learning_rate": 0.00021424936044163188,
"loss": 2.9632,
"step": 95600
},
{
"epoch": 25.74047157622739,
"grad_norm": 0.4729593098163605,
"learning_rate": 0.00021404739464117407,
"loss": 2.9728,
"step": 95650
},
{
"epoch": 25.75392980189492,
"grad_norm": 0.45664075016975403,
"learning_rate": 0.00021384542884071627,
"loss": 2.9578,
"step": 95700
},
{
"epoch": 25.767388027562447,
"grad_norm": 0.44564682245254517,
"learning_rate": 0.0002136434630402585,
"loss": 2.9686,
"step": 95750
},
{
"epoch": 25.780846253229974,
"grad_norm": 0.44380345940589905,
"learning_rate": 0.0002134414972398007,
"loss": 2.9594,
"step": 95800
},
{
"epoch": 25.7943044788975,
"grad_norm": 0.4766307473182678,
"learning_rate": 0.0002132395314393429,
"loss": 2.9633,
"step": 95850
},
{
"epoch": 25.80776270456503,
"grad_norm": 0.4366249740123749,
"learning_rate": 0.00021303756563888514,
"loss": 2.9668,
"step": 95900
},
{
"epoch": 25.82122093023256,
"grad_norm": 0.4472011923789978,
"learning_rate": 0.00021283559983842736,
"loss": 2.9644,
"step": 95950
},
{
"epoch": 25.834679155900087,
"grad_norm": 0.4363434910774231,
"learning_rate": 0.00021263363403796956,
"loss": 2.9658,
"step": 96000
},
{
"epoch": 25.834679155900087,
"eval_accuracy": 0.3974370424066061,
"eval_loss": 3.2766849994659424,
"eval_runtime": 185.1496,
"eval_samples_per_second": 97.284,
"eval_steps_per_second": 6.082,
"step": 96000
},
{
"epoch": 25.848137381567614,
"grad_norm": 0.41911518573760986,
"learning_rate": 0.00021243166823751175,
"loss": 2.9746,
"step": 96050
},
{
"epoch": 25.86159560723514,
"grad_norm": 0.4263755679130554,
"learning_rate": 0.00021222970243705398,
"loss": 2.9638,
"step": 96100
},
{
"epoch": 25.87505383290267,
"grad_norm": 0.42852187156677246,
"learning_rate": 0.00021202773663659617,
"loss": 2.9743,
"step": 96150
},
{
"epoch": 25.8885120585702,
"grad_norm": 0.4651477336883545,
"learning_rate": 0.0002118257708361384,
"loss": 2.9721,
"step": 96200
},
{
"epoch": 25.901970284237727,
"grad_norm": 0.49060264229774475,
"learning_rate": 0.0002116238050356806,
"loss": 2.9727,
"step": 96250
},
{
"epoch": 25.915428509905254,
"grad_norm": 0.4756370186805725,
"learning_rate": 0.00021142183923522281,
"loss": 2.9752,
"step": 96300
},
{
"epoch": 25.92888673557278,
"grad_norm": 0.44004926085472107,
"learning_rate": 0.00021121987343476504,
"loss": 2.9743,
"step": 96350
},
{
"epoch": 25.94234496124031,
"grad_norm": 0.44824323058128357,
"learning_rate": 0.00021101790763430726,
"loss": 2.9768,
"step": 96400
},
{
"epoch": 25.955803186907836,
"grad_norm": 0.43313372135162354,
"learning_rate": 0.00021081594183384946,
"loss": 2.9785,
"step": 96450
},
{
"epoch": 25.969261412575367,
"grad_norm": 0.4766993820667267,
"learning_rate": 0.00021061397603339165,
"loss": 2.9616,
"step": 96500
},
{
"epoch": 25.982719638242894,
"grad_norm": 0.45998021960258484,
"learning_rate": 0.00021041201023293388,
"loss": 2.9753,
"step": 96550
},
{
"epoch": 25.99617786391042,
"grad_norm": 0.42615368962287903,
"learning_rate": 0.00021021004443247607,
"loss": 2.973,
"step": 96600
},
{
"epoch": 26.00942075796727,
"grad_norm": 0.4443589448928833,
"learning_rate": 0.0002100080786320183,
"loss": 2.9167,
"step": 96650
},
{
"epoch": 26.022878983634797,
"grad_norm": 0.4575954079627991,
"learning_rate": 0.0002098061128315605,
"loss": 2.8917,
"step": 96700
},
{
"epoch": 26.036337209302324,
"grad_norm": 0.4684363007545471,
"learning_rate": 0.00020960414703110274,
"loss": 2.892,
"step": 96750
},
{
"epoch": 26.049795434969855,
"grad_norm": 0.4875715374946594,
"learning_rate": 0.00020940218123064494,
"loss": 2.8959,
"step": 96800
},
{
"epoch": 26.063253660637383,
"grad_norm": 0.44224879145622253,
"learning_rate": 0.00020920021543018713,
"loss": 2.9101,
"step": 96850
},
{
"epoch": 26.07671188630491,
"grad_norm": 0.4343826174736023,
"learning_rate": 0.00020899824962972936,
"loss": 2.9055,
"step": 96900
},
{
"epoch": 26.090170111972437,
"grad_norm": 0.4552612006664276,
"learning_rate": 0.00020879628382927155,
"loss": 2.9004,
"step": 96950
},
{
"epoch": 26.103628337639964,
"grad_norm": 0.4400579333305359,
"learning_rate": 0.00020859431802881378,
"loss": 2.9157,
"step": 97000
},
{
"epoch": 26.103628337639964,
"eval_accuracy": 0.39695834436098915,
"eval_loss": 3.2879819869995117,
"eval_runtime": 187.8091,
"eval_samples_per_second": 95.906,
"eval_steps_per_second": 5.995,
"step": 97000
},
{
"epoch": 26.117086563307492,
"grad_norm": 0.4384458363056183,
"learning_rate": 0.00020839235222835597,
"loss": 2.9095,
"step": 97050
},
{
"epoch": 26.130544788975023,
"grad_norm": 0.4629943370819092,
"learning_rate": 0.00020819038642789817,
"loss": 2.9151,
"step": 97100
},
{
"epoch": 26.14400301464255,
"grad_norm": 0.46021533012390137,
"learning_rate": 0.0002079884206274404,
"loss": 2.9115,
"step": 97150
},
{
"epoch": 26.157461240310077,
"grad_norm": 0.4656979739665985,
"learning_rate": 0.00020778645482698264,
"loss": 2.9047,
"step": 97200
},
{
"epoch": 26.170919465977605,
"grad_norm": 0.42537960410118103,
"learning_rate": 0.00020758448902652484,
"loss": 2.9146,
"step": 97250
},
{
"epoch": 26.184377691645132,
"grad_norm": 0.4642103910446167,
"learning_rate": 0.00020738252322606703,
"loss": 2.9202,
"step": 97300
},
{
"epoch": 26.197835917312663,
"grad_norm": 0.4689278304576874,
"learning_rate": 0.00020718055742560926,
"loss": 2.9222,
"step": 97350
},
{
"epoch": 26.21129414298019,
"grad_norm": 0.4738834500312805,
"learning_rate": 0.00020697859162515145,
"loss": 2.9204,
"step": 97400
},
{
"epoch": 26.224752368647717,
"grad_norm": 0.46898141503334045,
"learning_rate": 0.00020677662582469368,
"loss": 2.9228,
"step": 97450
},
{
"epoch": 26.238210594315245,
"grad_norm": 0.45319777727127075,
"learning_rate": 0.00020657466002423587,
"loss": 2.9298,
"step": 97500
},
{
"epoch": 26.251668819982772,
"grad_norm": 0.4632851183414459,
"learning_rate": 0.00020637269422377807,
"loss": 2.9168,
"step": 97550
},
{
"epoch": 26.265127045650303,
"grad_norm": 0.44312527775764465,
"learning_rate": 0.0002061707284233203,
"loss": 2.9144,
"step": 97600
},
{
"epoch": 26.27858527131783,
"grad_norm": 0.47308871150016785,
"learning_rate": 0.00020596876262286251,
"loss": 2.9226,
"step": 97650
},
{
"epoch": 26.292043496985357,
"grad_norm": 0.4948664605617523,
"learning_rate": 0.00020576679682240474,
"loss": 2.9275,
"step": 97700
},
{
"epoch": 26.305501722652885,
"grad_norm": 0.4614150822162628,
"learning_rate": 0.00020556483102194693,
"loss": 2.9365,
"step": 97750
},
{
"epoch": 26.318959948320412,
"grad_norm": 0.5010912418365479,
"learning_rate": 0.00020536286522148916,
"loss": 2.9239,
"step": 97800
},
{
"epoch": 26.332418173987943,
"grad_norm": 0.47301313281059265,
"learning_rate": 0.00020516089942103135,
"loss": 2.9332,
"step": 97850
},
{
"epoch": 26.34587639965547,
"grad_norm": 0.45448145270347595,
"learning_rate": 0.00020495893362057355,
"loss": 2.9243,
"step": 97900
},
{
"epoch": 26.359334625322997,
"grad_norm": 0.4401485323905945,
"learning_rate": 0.00020475696782011577,
"loss": 2.9357,
"step": 97950
},
{
"epoch": 26.372792850990525,
"grad_norm": 0.48110631108283997,
"learning_rate": 0.00020455500201965797,
"loss": 2.9406,
"step": 98000
},
{
"epoch": 26.372792850990525,
"eval_accuracy": 0.39708796005967756,
"eval_loss": 3.284863233566284,
"eval_runtime": 182.6664,
"eval_samples_per_second": 98.606,
"eval_steps_per_second": 6.164,
"step": 98000
},
{
"epoch": 26.386251076658052,
"grad_norm": 0.4605270326137543,
"learning_rate": 0.0002043530362192002,
"loss": 2.9359,
"step": 98050
},
{
"epoch": 26.399709302325583,
"grad_norm": 0.469814658164978,
"learning_rate": 0.00020415107041874242,
"loss": 2.9346,
"step": 98100
},
{
"epoch": 26.41316752799311,
"grad_norm": 0.43314987421035767,
"learning_rate": 0.00020394910461828464,
"loss": 2.9304,
"step": 98150
},
{
"epoch": 26.426625753660637,
"grad_norm": 0.43660035729408264,
"learning_rate": 0.00020374713881782683,
"loss": 2.9253,
"step": 98200
},
{
"epoch": 26.440083979328165,
"grad_norm": 0.45597589015960693,
"learning_rate": 0.00020354517301736906,
"loss": 2.9352,
"step": 98250
},
{
"epoch": 26.453542204995692,
"grad_norm": 0.4542783796787262,
"learning_rate": 0.00020334320721691125,
"loss": 2.9485,
"step": 98300
},
{
"epoch": 26.467000430663223,
"grad_norm": 0.46545565128326416,
"learning_rate": 0.00020314124141645345,
"loss": 2.9324,
"step": 98350
},
{
"epoch": 26.48045865633075,
"grad_norm": 0.4858553111553192,
"learning_rate": 0.00020293927561599567,
"loss": 2.9359,
"step": 98400
},
{
"epoch": 26.493916881998278,
"grad_norm": 0.46775391697883606,
"learning_rate": 0.00020273730981553787,
"loss": 2.9458,
"step": 98450
},
{
"epoch": 26.507375107665805,
"grad_norm": 0.4647384583950043,
"learning_rate": 0.0002025353440150801,
"loss": 2.9622,
"step": 98500
},
{
"epoch": 26.520833333333332,
"grad_norm": 0.4390053153038025,
"learning_rate": 0.00020233337821462232,
"loss": 2.9472,
"step": 98550
},
{
"epoch": 26.534291559000863,
"grad_norm": 0.4600248336791992,
"learning_rate": 0.00020213141241416454,
"loss": 2.9416,
"step": 98600
},
{
"epoch": 26.54774978466839,
"grad_norm": 0.4381559193134308,
"learning_rate": 0.00020192944661370674,
"loss": 2.9305,
"step": 98650
},
{
"epoch": 26.561208010335918,
"grad_norm": 0.4481687545776367,
"learning_rate": 0.00020172748081324893,
"loss": 2.948,
"step": 98700
},
{
"epoch": 26.574666236003445,
"grad_norm": 0.45131444931030273,
"learning_rate": 0.00020152551501279115,
"loss": 2.9451,
"step": 98750
},
{
"epoch": 26.588124461670972,
"grad_norm": 0.45956605672836304,
"learning_rate": 0.00020132354921233335,
"loss": 2.9518,
"step": 98800
},
{
"epoch": 26.6015826873385,
"grad_norm": 0.4756692051887512,
"learning_rate": 0.00020112158341187557,
"loss": 2.9396,
"step": 98850
},
{
"epoch": 26.61504091300603,
"grad_norm": 0.46689143776893616,
"learning_rate": 0.00020091961761141777,
"loss": 2.9514,
"step": 98900
},
{
"epoch": 26.628499138673558,
"grad_norm": 0.4691162109375,
"learning_rate": 0.00020071765181095997,
"loss": 2.9541,
"step": 98950
},
{
"epoch": 26.641957364341085,
"grad_norm": 0.44409865140914917,
"learning_rate": 0.00020051568601050222,
"loss": 2.9521,
"step": 99000
},
{
"epoch": 26.641957364341085,
"eval_accuracy": 0.39710273603203433,
"eval_loss": 3.2821266651153564,
"eval_runtime": 182.6774,
"eval_samples_per_second": 98.6,
"eval_steps_per_second": 6.164,
"step": 99000
},
{
"epoch": 26.655415590008612,
"grad_norm": 0.4674365818500519,
"learning_rate": 0.00020031372021004444,
"loss": 2.9515,
"step": 99050
},
{
"epoch": 26.66887381567614,
"grad_norm": 0.46960291266441345,
"learning_rate": 0.00020011175440958664,
"loss": 2.9506,
"step": 99100
},
{
"epoch": 26.68233204134367,
"grad_norm": 0.4617456793785095,
"learning_rate": 0.00019990978860912883,
"loss": 2.9521,
"step": 99150
},
{
"epoch": 26.695790267011198,
"grad_norm": 0.46739956736564636,
"learning_rate": 0.00019970782280867105,
"loss": 2.9495,
"step": 99200
},
{
"epoch": 26.709248492678725,
"grad_norm": 0.4411316514015198,
"learning_rate": 0.00019950585700821325,
"loss": 2.9531,
"step": 99250
},
{
"epoch": 26.722706718346252,
"grad_norm": 0.43900373578071594,
"learning_rate": 0.00019930389120775547,
"loss": 2.949,
"step": 99300
},
{
"epoch": 26.73616494401378,
"grad_norm": 0.46046778559684753,
"learning_rate": 0.00019910192540729767,
"loss": 2.9484,
"step": 99350
},
{
"epoch": 26.74962316968131,
"grad_norm": 0.43259429931640625,
"learning_rate": 0.00019889995960683987,
"loss": 2.954,
"step": 99400
},
{
"epoch": 26.763081395348838,
"grad_norm": 0.4836384356021881,
"learning_rate": 0.00019869799380638212,
"loss": 2.9558,
"step": 99450
},
{
"epoch": 26.776539621016365,
"grad_norm": 0.4613456428050995,
"learning_rate": 0.0001984960280059243,
"loss": 2.9546,
"step": 99500
},
{
"epoch": 26.789997846683892,
"grad_norm": 0.45613229274749756,
"learning_rate": 0.00019829406220546654,
"loss": 2.9516,
"step": 99550
},
{
"epoch": 26.80345607235142,
"grad_norm": 0.46392738819122314,
"learning_rate": 0.00019809209640500873,
"loss": 2.9625,
"step": 99600
},
{
"epoch": 26.81691429801895,
"grad_norm": 0.4266089200973511,
"learning_rate": 0.00019789013060455096,
"loss": 2.9599,
"step": 99650
},
{
"epoch": 26.830372523686478,
"grad_norm": 0.4583572447299957,
"learning_rate": 0.00019768816480409315,
"loss": 2.9557,
"step": 99700
},
{
"epoch": 26.843830749354005,
"grad_norm": 0.46834680438041687,
"learning_rate": 0.00019748619900363535,
"loss": 2.9516,
"step": 99750
},
{
"epoch": 26.857288975021532,
"grad_norm": 0.46385806798934937,
"learning_rate": 0.00019728423320317757,
"loss": 2.957,
"step": 99800
},
{
"epoch": 26.87074720068906,
"grad_norm": 0.4585137367248535,
"learning_rate": 0.00019708226740271977,
"loss": 2.9487,
"step": 99850
},
{
"epoch": 26.88420542635659,
"grad_norm": 0.45682016015052795,
"learning_rate": 0.00019688030160226202,
"loss": 2.9568,
"step": 99900
},
{
"epoch": 26.897663652024118,
"grad_norm": 0.47088906168937683,
"learning_rate": 0.0001966783358018042,
"loss": 2.9625,
"step": 99950
},
{
"epoch": 26.911121877691645,
"grad_norm": 0.4769439697265625,
"learning_rate": 0.00019647637000134644,
"loss": 2.9665,
"step": 100000
},
{
"epoch": 26.911121877691645,
"eval_accuracy": 0.3978143729359814,
"eval_loss": 3.2741730213165283,
"eval_runtime": 182.7773,
"eval_samples_per_second": 98.546,
"eval_steps_per_second": 6.161,
"step": 100000
},
{
"epoch": 26.924580103359173,
"grad_norm": 0.4691717028617859,
"learning_rate": 0.00019627440420088863,
"loss": 2.9644,
"step": 100050
},
{
"epoch": 26.9380383290267,
"grad_norm": 0.43463969230651855,
"learning_rate": 0.00019607243840043086,
"loss": 2.9516,
"step": 100100
},
{
"epoch": 26.95149655469423,
"grad_norm": 0.43889477849006653,
"learning_rate": 0.00019587047259997305,
"loss": 2.9751,
"step": 100150
},
{
"epoch": 26.964954780361758,
"grad_norm": 0.4451788067817688,
"learning_rate": 0.00019566850679951525,
"loss": 2.9538,
"step": 100200
},
{
"epoch": 26.978413006029285,
"grad_norm": 0.4419739246368408,
"learning_rate": 0.00019546654099905747,
"loss": 2.958,
"step": 100250
},
{
"epoch": 26.991871231696813,
"grad_norm": 0.4447169899940491,
"learning_rate": 0.00019526457519859967,
"loss": 2.9587,
"step": 100300
},
{
"epoch": 27.00511412575366,
"grad_norm": 0.4693147838115692,
"learning_rate": 0.00019506260939814192,
"loss": 2.9407,
"step": 100350
},
{
"epoch": 27.018572351421188,
"grad_norm": 0.4691426753997803,
"learning_rate": 0.0001948606435976841,
"loss": 2.8858,
"step": 100400
},
{
"epoch": 27.032030577088715,
"grad_norm": 0.44919729232788086,
"learning_rate": 0.00019465867779722634,
"loss": 2.8837,
"step": 100450
},
{
"epoch": 27.045488802756246,
"grad_norm": 0.45154300332069397,
"learning_rate": 0.00019445671199676853,
"loss": 2.8874,
"step": 100500
},
{
"epoch": 27.058947028423773,
"grad_norm": 0.44890469312667847,
"learning_rate": 0.00019425474619631073,
"loss": 2.9001,
"step": 100550
},
{
"epoch": 27.0724052540913,
"grad_norm": 0.43744274973869324,
"learning_rate": 0.00019405278039585295,
"loss": 2.8969,
"step": 100600
},
{
"epoch": 27.085863479758828,
"grad_norm": 0.47108370065689087,
"learning_rate": 0.00019385081459539515,
"loss": 2.8997,
"step": 100650
},
{
"epoch": 27.099321705426355,
"grad_norm": 0.46258291602134705,
"learning_rate": 0.00019364884879493737,
"loss": 2.8984,
"step": 100700
},
{
"epoch": 27.112779931093886,
"grad_norm": 0.4900825321674347,
"learning_rate": 0.00019344688299447957,
"loss": 2.896,
"step": 100750
},
{
"epoch": 27.126238156761413,
"grad_norm": 0.4664347469806671,
"learning_rate": 0.00019324491719402182,
"loss": 2.9012,
"step": 100800
},
{
"epoch": 27.13969638242894,
"grad_norm": 0.45574066042900085,
"learning_rate": 0.00019304295139356401,
"loss": 2.9063,
"step": 100850
},
{
"epoch": 27.153154608096468,
"grad_norm": 0.4748891592025757,
"learning_rate": 0.0001928409855931062,
"loss": 2.8963,
"step": 100900
},
{
"epoch": 27.166612833763995,
"grad_norm": 0.4680997133255005,
"learning_rate": 0.00019263901979264843,
"loss": 2.9056,
"step": 100950
},
{
"epoch": 27.180071059431526,
"grad_norm": 0.4529612958431244,
"learning_rate": 0.00019243705399219063,
"loss": 2.9038,
"step": 101000
},
{
"epoch": 27.180071059431526,
"eval_accuracy": 0.3971947599186974,
"eval_loss": 3.2878904342651367,
"eval_runtime": 182.555,
"eval_samples_per_second": 98.666,
"eval_steps_per_second": 6.168,
"step": 101000
},
{
"epoch": 27.193529285099054,
"grad_norm": 0.4609749913215637,
"learning_rate": 0.00019223508819173285,
"loss": 2.9105,
"step": 101050
},
{
"epoch": 27.20698751076658,
"grad_norm": 0.43284550309181213,
"learning_rate": 0.00019203312239127505,
"loss": 2.9105,
"step": 101100
},
{
"epoch": 27.220445736434108,
"grad_norm": 0.46304851770401,
"learning_rate": 0.00019183115659081727,
"loss": 2.9086,
"step": 101150
},
{
"epoch": 27.233903962101635,
"grad_norm": 0.4563630223274231,
"learning_rate": 0.00019162919079035947,
"loss": 2.9084,
"step": 101200
},
{
"epoch": 27.247362187769163,
"grad_norm": 0.4489329159259796,
"learning_rate": 0.00019142722498990172,
"loss": 2.9166,
"step": 101250
},
{
"epoch": 27.260820413436694,
"grad_norm": 0.4651649594306946,
"learning_rate": 0.00019122525918944391,
"loss": 2.9121,
"step": 101300
},
{
"epoch": 27.27427863910422,
"grad_norm": 0.4722488820552826,
"learning_rate": 0.0001910232933889861,
"loss": 2.9111,
"step": 101350
},
{
"epoch": 27.287736864771748,
"grad_norm": 0.4390490651130676,
"learning_rate": 0.00019082132758852833,
"loss": 2.915,
"step": 101400
},
{
"epoch": 27.301195090439276,
"grad_norm": 0.48112547397613525,
"learning_rate": 0.00019061936178807053,
"loss": 2.916,
"step": 101450
},
{
"epoch": 27.314653316106803,
"grad_norm": 0.48611268401145935,
"learning_rate": 0.00019041739598761275,
"loss": 2.9214,
"step": 101500
},
{
"epoch": 27.328111541774334,
"grad_norm": 0.497575581073761,
"learning_rate": 0.00019021543018715495,
"loss": 2.9108,
"step": 101550
},
{
"epoch": 27.34156976744186,
"grad_norm": 0.470059871673584,
"learning_rate": 0.00019001346438669714,
"loss": 2.9208,
"step": 101600
},
{
"epoch": 27.35502799310939,
"grad_norm": 0.45641663670539856,
"learning_rate": 0.0001898114985862394,
"loss": 2.9236,
"step": 101650
},
{
"epoch": 27.368486218776916,
"grad_norm": 0.48994702100753784,
"learning_rate": 0.0001896095327857816,
"loss": 2.9093,
"step": 101700
},
{
"epoch": 27.381944444444443,
"grad_norm": 0.4846106171607971,
"learning_rate": 0.00018940756698532381,
"loss": 2.9305,
"step": 101750
},
{
"epoch": 27.395402670111974,
"grad_norm": 0.44658008217811584,
"learning_rate": 0.000189205601184866,
"loss": 2.9285,
"step": 101800
},
{
"epoch": 27.4088608957795,
"grad_norm": 0.45827382802963257,
"learning_rate": 0.00018900363538440823,
"loss": 2.9349,
"step": 101850
},
{
"epoch": 27.42231912144703,
"grad_norm": 0.4490291178226471,
"learning_rate": 0.00018880166958395043,
"loss": 2.9377,
"step": 101900
},
{
"epoch": 27.435777347114556,
"grad_norm": 0.4456426203250885,
"learning_rate": 0.00018859970378349263,
"loss": 2.9267,
"step": 101950
},
{
"epoch": 27.449235572782083,
"grad_norm": 0.47101685404777527,
"learning_rate": 0.00018839773798303485,
"loss": 2.9357,
"step": 102000
},
{
"epoch": 27.449235572782083,
"eval_accuracy": 0.3972929766761276,
"eval_loss": 3.2838637828826904,
"eval_runtime": 182.4934,
"eval_samples_per_second": 98.699,
"eval_steps_per_second": 6.17,
"step": 102000
},
{
"epoch": 27.462693798449614,
"grad_norm": 0.46900230646133423,
"learning_rate": 0.00018819577218257705,
"loss": 2.9282,
"step": 102050
},
{
"epoch": 27.47615202411714,
"grad_norm": 0.46709170937538147,
"learning_rate": 0.0001879938063821193,
"loss": 2.9214,
"step": 102100
},
{
"epoch": 27.48961024978467,
"grad_norm": 0.4618961215019226,
"learning_rate": 0.0001877918405816615,
"loss": 2.939,
"step": 102150
},
{
"epoch": 27.503068475452196,
"grad_norm": 0.44597670435905457,
"learning_rate": 0.00018758987478120371,
"loss": 2.9453,
"step": 102200
},
{
"epoch": 27.516526701119723,
"grad_norm": 0.4340479075908661,
"learning_rate": 0.0001873879089807459,
"loss": 2.9348,
"step": 102250
},
{
"epoch": 27.529984926787254,
"grad_norm": 0.48661714792251587,
"learning_rate": 0.00018718594318028813,
"loss": 2.9314,
"step": 102300
},
{
"epoch": 27.54344315245478,
"grad_norm": 0.45866867899894714,
"learning_rate": 0.00018698397737983033,
"loss": 2.9382,
"step": 102350
},
{
"epoch": 27.55690137812231,
"grad_norm": 0.47287243604660034,
"learning_rate": 0.00018678201157937253,
"loss": 2.9387,
"step": 102400
},
{
"epoch": 27.570359603789836,
"grad_norm": 0.4486309885978699,
"learning_rate": 0.00018658004577891475,
"loss": 2.9296,
"step": 102450
},
{
"epoch": 27.583817829457363,
"grad_norm": 0.49782565236091614,
"learning_rate": 0.00018637807997845695,
"loss": 2.9425,
"step": 102500
},
{
"epoch": 27.597276055124894,
"grad_norm": 0.46115875244140625,
"learning_rate": 0.0001861761141779992,
"loss": 2.9287,
"step": 102550
},
{
"epoch": 27.61073428079242,
"grad_norm": 0.47809669375419617,
"learning_rate": 0.0001859741483775414,
"loss": 2.9384,
"step": 102600
},
{
"epoch": 27.62419250645995,
"grad_norm": 0.46067842841148376,
"learning_rate": 0.00018577218257708362,
"loss": 2.945,
"step": 102650
},
{
"epoch": 27.637650732127476,
"grad_norm": 0.486005038022995,
"learning_rate": 0.0001855702167766258,
"loss": 2.9357,
"step": 102700
},
{
"epoch": 27.651108957795003,
"grad_norm": 0.475208044052124,
"learning_rate": 0.000185368250976168,
"loss": 2.9475,
"step": 102750
},
{
"epoch": 27.664567183462534,
"grad_norm": 0.46157121658325195,
"learning_rate": 0.00018516628517571023,
"loss": 2.9366,
"step": 102800
},
{
"epoch": 27.67802540913006,
"grad_norm": 0.4684220254421234,
"learning_rate": 0.00018496431937525243,
"loss": 2.939,
"step": 102850
},
{
"epoch": 27.69148363479759,
"grad_norm": 0.4938398003578186,
"learning_rate": 0.00018476235357479465,
"loss": 2.94,
"step": 102900
},
{
"epoch": 27.704941860465116,
"grad_norm": 0.4568376839160919,
"learning_rate": 0.00018456038777433685,
"loss": 2.9464,
"step": 102950
},
{
"epoch": 27.718400086132643,
"grad_norm": 0.45346930623054504,
"learning_rate": 0.0001843584219738791,
"loss": 2.9344,
"step": 103000
},
{
"epoch": 27.718400086132643,
"eval_accuracy": 0.3976646575690136,
"eval_loss": 3.28147554397583,
"eval_runtime": 182.5512,
"eval_samples_per_second": 98.668,
"eval_steps_per_second": 6.168,
"step": 103000
},
{
"epoch": 27.731858311800174,
"grad_norm": 0.49002352356910706,
"learning_rate": 0.0001841564561734213,
"loss": 2.9437,
"step": 103050
},
{
"epoch": 27.7453165374677,
"grad_norm": 0.43377572298049927,
"learning_rate": 0.00018395449037296352,
"loss": 2.9371,
"step": 103100
},
{
"epoch": 27.75877476313523,
"grad_norm": 0.46353358030319214,
"learning_rate": 0.0001837525245725057,
"loss": 2.9401,
"step": 103150
},
{
"epoch": 27.772232988802756,
"grad_norm": 0.47084367275238037,
"learning_rate": 0.0001835505587720479,
"loss": 2.9465,
"step": 103200
},
{
"epoch": 27.785691214470283,
"grad_norm": 0.46358805894851685,
"learning_rate": 0.00018334859297159013,
"loss": 2.9479,
"step": 103250
},
{
"epoch": 27.79914944013781,
"grad_norm": 0.46683570742607117,
"learning_rate": 0.00018314662717113233,
"loss": 2.9471,
"step": 103300
},
{
"epoch": 27.81260766580534,
"grad_norm": 0.45332661271095276,
"learning_rate": 0.00018294466137067455,
"loss": 2.9458,
"step": 103350
},
{
"epoch": 27.82606589147287,
"grad_norm": 0.466243714094162,
"learning_rate": 0.00018274269557021675,
"loss": 2.9526,
"step": 103400
},
{
"epoch": 27.839524117140396,
"grad_norm": 0.46517500281333923,
"learning_rate": 0.000182540729769759,
"loss": 2.9347,
"step": 103450
},
{
"epoch": 27.852982342807923,
"grad_norm": 0.4544712007045746,
"learning_rate": 0.0001823387639693012,
"loss": 2.9546,
"step": 103500
},
{
"epoch": 27.86644056847545,
"grad_norm": 0.4736934304237366,
"learning_rate": 0.0001821367981688434,
"loss": 2.9469,
"step": 103550
},
{
"epoch": 27.87989879414298,
"grad_norm": 0.4843897819519043,
"learning_rate": 0.0001819348323683856,
"loss": 2.9614,
"step": 103600
},
{
"epoch": 27.89335701981051,
"grad_norm": 0.4619082808494568,
"learning_rate": 0.0001817328665679278,
"loss": 2.9482,
"step": 103650
},
{
"epoch": 27.906815245478036,
"grad_norm": 0.478361576795578,
"learning_rate": 0.00018153090076747003,
"loss": 2.9454,
"step": 103700
},
{
"epoch": 27.920273471145563,
"grad_norm": 0.4733973741531372,
"learning_rate": 0.00018132893496701223,
"loss": 2.9371,
"step": 103750
},
{
"epoch": 27.93373169681309,
"grad_norm": 0.4832918047904968,
"learning_rate": 0.00018112696916655442,
"loss": 2.951,
"step": 103800
},
{
"epoch": 27.94718992248062,
"grad_norm": 0.4726840853691101,
"learning_rate": 0.00018092500336609665,
"loss": 2.9467,
"step": 103850
},
{
"epoch": 27.96064814814815,
"grad_norm": 0.5040254592895508,
"learning_rate": 0.0001807230375656389,
"loss": 2.9438,
"step": 103900
},
{
"epoch": 27.974106373815676,
"grad_norm": 0.49846261739730835,
"learning_rate": 0.0001805210717651811,
"loss": 2.9544,
"step": 103950
},
{
"epoch": 27.987564599483203,
"grad_norm": 0.46601295471191406,
"learning_rate": 0.0001803191059647233,
"loss": 2.9578,
"step": 104000
},
{
"epoch": 27.987564599483203,
"eval_accuracy": 0.39764944700923455,
"eval_loss": 3.2772696018218994,
"eval_runtime": 182.572,
"eval_samples_per_second": 98.657,
"eval_steps_per_second": 6.167,
"step": 104000
},
{
"epoch": 28.00080749354005,
"grad_norm": 0.46461671590805054,
"learning_rate": 0.0001801171401642655,
"loss": 2.9472,
"step": 104050
},
{
"epoch": 28.01426571920758,
"grad_norm": 0.4493894875049591,
"learning_rate": 0.0001799151743638077,
"loss": 2.874,
"step": 104100
},
{
"epoch": 28.027723944875106,
"grad_norm": 0.4868014454841614,
"learning_rate": 0.00017971320856334993,
"loss": 2.88,
"step": 104150
},
{
"epoch": 28.041182170542637,
"grad_norm": 0.48048654198646545,
"learning_rate": 0.00017951124276289213,
"loss": 2.877,
"step": 104200
},
{
"epoch": 28.054640396210164,
"grad_norm": 0.46957871317863464,
"learning_rate": 0.00017930927696243432,
"loss": 2.8798,
"step": 104250
},
{
"epoch": 28.06809862187769,
"grad_norm": 0.46928441524505615,
"learning_rate": 0.00017910731116197655,
"loss": 2.8845,
"step": 104300
},
{
"epoch": 28.08155684754522,
"grad_norm": 0.47827425599098206,
"learning_rate": 0.00017890534536151877,
"loss": 2.8832,
"step": 104350
},
{
"epoch": 28.095015073212746,
"grad_norm": 0.4577322006225586,
"learning_rate": 0.000178703379561061,
"loss": 2.8905,
"step": 104400
},
{
"epoch": 28.108473298880277,
"grad_norm": 0.44855794310569763,
"learning_rate": 0.0001785014137606032,
"loss": 2.9001,
"step": 104450
},
{
"epoch": 28.121931524547804,
"grad_norm": 0.47200289368629456,
"learning_rate": 0.0001782994479601454,
"loss": 2.8942,
"step": 104500
},
{
"epoch": 28.13538975021533,
"grad_norm": 0.4787461757659912,
"learning_rate": 0.0001780974821596876,
"loss": 2.8928,
"step": 104550
},
{
"epoch": 28.14884797588286,
"grad_norm": 0.4635203778743744,
"learning_rate": 0.0001778955163592298,
"loss": 2.8906,
"step": 104600
},
{
"epoch": 28.162306201550386,
"grad_norm": 0.5049405694007874,
"learning_rate": 0.00017769355055877203,
"loss": 2.9104,
"step": 104650
},
{
"epoch": 28.175764427217917,
"grad_norm": 0.5026819109916687,
"learning_rate": 0.00017749158475831422,
"loss": 2.8953,
"step": 104700
},
{
"epoch": 28.189222652885444,
"grad_norm": 0.4447897672653198,
"learning_rate": 0.00017728961895785645,
"loss": 2.9041,
"step": 104750
},
{
"epoch": 28.20268087855297,
"grad_norm": 0.49646249413490295,
"learning_rate": 0.00017708765315739867,
"loss": 2.9079,
"step": 104800
},
{
"epoch": 28.2161391042205,
"grad_norm": 0.4800196588039398,
"learning_rate": 0.0001768856873569409,
"loss": 2.906,
"step": 104850
},
{
"epoch": 28.229597329888026,
"grad_norm": 0.4657552242279053,
"learning_rate": 0.0001766837215564831,
"loss": 2.9031,
"step": 104900
},
{
"epoch": 28.243055555555557,
"grad_norm": 0.47703903913497925,
"learning_rate": 0.0001764817557560253,
"loss": 2.9075,
"step": 104950
},
{
"epoch": 28.256513781223084,
"grad_norm": 0.5011478066444397,
"learning_rate": 0.0001762797899555675,
"loss": 2.9156,
"step": 105000
},
{
"epoch": 28.256513781223084,
"eval_accuracy": 0.3972138817652767,
"eval_loss": 3.289583444595337,
"eval_runtime": 182.5564,
"eval_samples_per_second": 98.665,
"eval_steps_per_second": 6.168,
"step": 105000
},
{
"epoch": 28.26997200689061,
"grad_norm": 0.44775184988975525,
"learning_rate": 0.0001760778241551097,
"loss": 2.9132,
"step": 105050
},
{
"epoch": 28.28343023255814,
"grad_norm": 0.44015103578567505,
"learning_rate": 0.00017587585835465193,
"loss": 2.9034,
"step": 105100
},
{
"epoch": 28.296888458225666,
"grad_norm": 0.4723224937915802,
"learning_rate": 0.00017567389255419412,
"loss": 2.9026,
"step": 105150
},
{
"epoch": 28.310346683893197,
"grad_norm": 0.45084238052368164,
"learning_rate": 0.00017547192675373635,
"loss": 2.9011,
"step": 105200
},
{
"epoch": 28.323804909560724,
"grad_norm": 0.4948933720588684,
"learning_rate": 0.00017526996095327857,
"loss": 2.9072,
"step": 105250
},
{
"epoch": 28.337263135228252,
"grad_norm": 0.4854006767272949,
"learning_rate": 0.0001750679951528208,
"loss": 2.9107,
"step": 105300
},
{
"epoch": 28.35072136089578,
"grad_norm": 0.46420085430145264,
"learning_rate": 0.000174866029352363,
"loss": 2.9143,
"step": 105350
},
{
"epoch": 28.364179586563306,
"grad_norm": 0.5050031542778015,
"learning_rate": 0.0001746640635519052,
"loss": 2.9083,
"step": 105400
},
{
"epoch": 28.377637812230837,
"grad_norm": 0.4668348729610443,
"learning_rate": 0.0001744620977514474,
"loss": 2.9165,
"step": 105450
},
{
"epoch": 28.391096037898365,
"grad_norm": 0.48908141255378723,
"learning_rate": 0.0001742601319509896,
"loss": 2.9196,
"step": 105500
},
{
"epoch": 28.404554263565892,
"grad_norm": 0.4991946518421173,
"learning_rate": 0.00017405816615053183,
"loss": 2.9313,
"step": 105550
},
{
"epoch": 28.41801248923342,
"grad_norm": 0.4851292669773102,
"learning_rate": 0.00017385620035007403,
"loss": 2.9182,
"step": 105600
},
{
"epoch": 28.431470714900946,
"grad_norm": 0.46164289116859436,
"learning_rate": 0.00017365423454961622,
"loss": 2.9236,
"step": 105650
},
{
"epoch": 28.444928940568474,
"grad_norm": 0.4600871503353119,
"learning_rate": 0.00017345226874915847,
"loss": 2.922,
"step": 105700
},
{
"epoch": 28.458387166236005,
"grad_norm": 0.47667500376701355,
"learning_rate": 0.00017325030294870067,
"loss": 2.9164,
"step": 105750
},
{
"epoch": 28.471845391903532,
"grad_norm": 0.4906790256500244,
"learning_rate": 0.0001730483371482429,
"loss": 2.9074,
"step": 105800
},
{
"epoch": 28.48530361757106,
"grad_norm": 0.47919005155563354,
"learning_rate": 0.0001728463713477851,
"loss": 2.9145,
"step": 105850
},
{
"epoch": 28.498761843238587,
"grad_norm": 0.4737989008426666,
"learning_rate": 0.0001726444055473273,
"loss": 2.9271,
"step": 105900
},
{
"epoch": 28.512220068906114,
"grad_norm": 0.4676021635532379,
"learning_rate": 0.0001724424397468695,
"loss": 2.9208,
"step": 105950
},
{
"epoch": 28.525678294573645,
"grad_norm": 0.47199440002441406,
"learning_rate": 0.00017224047394641173,
"loss": 2.9258,
"step": 106000
},
{
"epoch": 28.525678294573645,
"eval_accuracy": 0.397545580615315,
"eval_loss": 3.282310962677002,
"eval_runtime": 182.511,
"eval_samples_per_second": 98.69,
"eval_steps_per_second": 6.169,
"step": 106000
},
{
"epoch": 28.539136520241172,
"grad_norm": 0.49489060044288635,
"learning_rate": 0.00017203850814595393,
"loss": 2.922,
"step": 106050
},
{
"epoch": 28.5525947459087,
"grad_norm": 0.4765784740447998,
"learning_rate": 0.00017183654234549612,
"loss": 2.9186,
"step": 106100
},
{
"epoch": 28.566052971576227,
"grad_norm": 0.49360308051109314,
"learning_rate": 0.00017163457654503837,
"loss": 2.92,
"step": 106150
},
{
"epoch": 28.579511197243754,
"grad_norm": 0.46456533670425415,
"learning_rate": 0.00017143261074458057,
"loss": 2.9261,
"step": 106200
},
{
"epoch": 28.592969422911285,
"grad_norm": 0.4661301374435425,
"learning_rate": 0.0001712306449441228,
"loss": 2.917,
"step": 106250
},
{
"epoch": 28.606427648578812,
"grad_norm": 0.4966243803501129,
"learning_rate": 0.000171028679143665,
"loss": 2.9365,
"step": 106300
},
{
"epoch": 28.61988587424634,
"grad_norm": 0.5036020278930664,
"learning_rate": 0.0001708267133432072,
"loss": 2.9322,
"step": 106350
},
{
"epoch": 28.633344099913867,
"grad_norm": 0.46409788727760315,
"learning_rate": 0.0001706247475427494,
"loss": 2.935,
"step": 106400
},
{
"epoch": 28.646802325581394,
"grad_norm": 0.4711965024471283,
"learning_rate": 0.0001704227817422916,
"loss": 2.9208,
"step": 106450
},
{
"epoch": 28.660260551248925,
"grad_norm": 0.4660088121891022,
"learning_rate": 0.00017022081594183383,
"loss": 2.9387,
"step": 106500
},
{
"epoch": 28.673718776916452,
"grad_norm": 0.4687451720237732,
"learning_rate": 0.00017001885014137605,
"loss": 2.9202,
"step": 106550
},
{
"epoch": 28.68717700258398,
"grad_norm": 0.4957084655761719,
"learning_rate": 0.00016981688434091827,
"loss": 2.947,
"step": 106600
},
{
"epoch": 28.700635228251507,
"grad_norm": 0.46830615401268005,
"learning_rate": 0.00016961491854046047,
"loss": 2.9203,
"step": 106650
},
{
"epoch": 28.714093453919034,
"grad_norm": 0.47599518299102783,
"learning_rate": 0.0001694129527400027,
"loss": 2.9317,
"step": 106700
},
{
"epoch": 28.727551679586565,
"grad_norm": 0.4689963459968567,
"learning_rate": 0.0001692109869395449,
"loss": 2.9317,
"step": 106750
},
{
"epoch": 28.741009905254092,
"grad_norm": 0.4767916202545166,
"learning_rate": 0.00016900902113908708,
"loss": 2.9167,
"step": 106800
},
{
"epoch": 28.75446813092162,
"grad_norm": 0.4934927821159363,
"learning_rate": 0.0001688070553386293,
"loss": 2.9323,
"step": 106850
},
{
"epoch": 28.767926356589147,
"grad_norm": 0.45270422101020813,
"learning_rate": 0.0001686050895381715,
"loss": 2.9384,
"step": 106900
},
{
"epoch": 28.781384582256674,
"grad_norm": 0.482998788356781,
"learning_rate": 0.00016840312373771373,
"loss": 2.9311,
"step": 106950
},
{
"epoch": 28.794842807924205,
"grad_norm": 0.4722956418991089,
"learning_rate": 0.00016820115793725595,
"loss": 2.9385,
"step": 107000
},
{
"epoch": 28.794842807924205,
"eval_accuracy": 0.39784131735616135,
"eval_loss": 3.2763619422912598,
"eval_runtime": 187.5601,
"eval_samples_per_second": 96.033,
"eval_steps_per_second": 6.003,
"step": 107000
},
{
"epoch": 28.808301033591732,
"grad_norm": 0.46862295269966125,
"learning_rate": 0.00016799919213679817,
"loss": 2.9462,
"step": 107050
},
{
"epoch": 28.82175925925926,
"grad_norm": 0.469465970993042,
"learning_rate": 0.00016779722633634037,
"loss": 2.9372,
"step": 107100
},
{
"epoch": 28.835217484926787,
"grad_norm": 0.4718657433986664,
"learning_rate": 0.0001675952605358826,
"loss": 2.9376,
"step": 107150
},
{
"epoch": 28.848675710594314,
"grad_norm": 0.46478015184402466,
"learning_rate": 0.0001673932947354248,
"loss": 2.9324,
"step": 107200
},
{
"epoch": 28.86213393626184,
"grad_norm": 0.4923340082168579,
"learning_rate": 0.00016719132893496698,
"loss": 2.9352,
"step": 107250
},
{
"epoch": 28.875592161929372,
"grad_norm": 0.4849088490009308,
"learning_rate": 0.0001669893631345092,
"loss": 2.942,
"step": 107300
},
{
"epoch": 28.8890503875969,
"grad_norm": 0.4664568603038788,
"learning_rate": 0.0001667873973340514,
"loss": 2.9399,
"step": 107350
},
{
"epoch": 28.902508613264427,
"grad_norm": 0.46912139654159546,
"learning_rate": 0.00016658543153359363,
"loss": 2.9367,
"step": 107400
},
{
"epoch": 28.915966838931954,
"grad_norm": 0.48870810866355896,
"learning_rate": 0.00016638346573313585,
"loss": 2.9364,
"step": 107450
},
{
"epoch": 28.92942506459948,
"grad_norm": 0.4678412079811096,
"learning_rate": 0.00016618149993267807,
"loss": 2.9308,
"step": 107500
},
{
"epoch": 28.942883290267012,
"grad_norm": 0.45937883853912354,
"learning_rate": 0.00016597953413222027,
"loss": 2.9418,
"step": 107550
},
{
"epoch": 28.95634151593454,
"grad_norm": 0.46293509006500244,
"learning_rate": 0.00016577756833176247,
"loss": 2.9301,
"step": 107600
},
{
"epoch": 28.969799741602067,
"grad_norm": 0.48103538155555725,
"learning_rate": 0.0001655756025313047,
"loss": 2.9396,
"step": 107650
},
{
"epoch": 28.983257967269594,
"grad_norm": 0.45742693543434143,
"learning_rate": 0.00016537363673084688,
"loss": 2.9419,
"step": 107700
},
{
"epoch": 28.99671619293712,
"grad_norm": 0.48283737897872925,
"learning_rate": 0.0001651716709303891,
"loss": 2.9393,
"step": 107750
},
{
"epoch": 29.00995908699397,
"grad_norm": 0.48249244689941406,
"learning_rate": 0.0001649697051299313,
"loss": 2.9,
"step": 107800
},
{
"epoch": 29.023417312661497,
"grad_norm": 0.4821341037750244,
"learning_rate": 0.0001647677393294735,
"loss": 2.8744,
"step": 107850
},
{
"epoch": 29.036875538329028,
"grad_norm": 0.4741370677947998,
"learning_rate": 0.00016456577352901575,
"loss": 2.8748,
"step": 107900
},
{
"epoch": 29.050333763996555,
"grad_norm": 0.49252137541770935,
"learning_rate": 0.00016436380772855797,
"loss": 2.8865,
"step": 107950
},
{
"epoch": 29.063791989664082,
"grad_norm": 0.4660206735134125,
"learning_rate": 0.00016416184192810017,
"loss": 2.8663,
"step": 108000
},
{
"epoch": 29.063791989664082,
"eval_accuracy": 0.39770083697191655,
"eval_loss": 3.287839651107788,
"eval_runtime": 186.878,
"eval_samples_per_second": 96.384,
"eval_steps_per_second": 6.025,
"step": 108000
},
{
"epoch": 29.07725021533161,
"grad_norm": 0.48520275950431824,
"learning_rate": 0.00016395987612764237,
"loss": 2.8748,
"step": 108050
},
{
"epoch": 29.090708440999137,
"grad_norm": 0.49605780839920044,
"learning_rate": 0.0001637579103271846,
"loss": 2.8828,
"step": 108100
},
{
"epoch": 29.104166666666668,
"grad_norm": 0.4688926935195923,
"learning_rate": 0.00016355594452672679,
"loss": 2.8756,
"step": 108150
},
{
"epoch": 29.117624892334195,
"grad_norm": 0.4537731409072876,
"learning_rate": 0.000163353978726269,
"loss": 2.8861,
"step": 108200
},
{
"epoch": 29.131083118001722,
"grad_norm": 0.49110716581344604,
"learning_rate": 0.0001631520129258112,
"loss": 2.884,
"step": 108250
},
{
"epoch": 29.14454134366925,
"grad_norm": 0.49029913544654846,
"learning_rate": 0.0001629500471253534,
"loss": 2.8902,
"step": 108300
},
{
"epoch": 29.157999569336777,
"grad_norm": 0.506089985370636,
"learning_rate": 0.00016274808132489565,
"loss": 2.8854,
"step": 108350
},
{
"epoch": 29.171457795004308,
"grad_norm": 0.5150735378265381,
"learning_rate": 0.00016254611552443785,
"loss": 2.8867,
"step": 108400
},
{
"epoch": 29.184916020671835,
"grad_norm": 0.4747299253940582,
"learning_rate": 0.00016234414972398007,
"loss": 2.8845,
"step": 108450
},
{
"epoch": 29.198374246339363,
"grad_norm": 0.4814194142818451,
"learning_rate": 0.00016214218392352227,
"loss": 2.8986,
"step": 108500
},
{
"epoch": 29.21183247200689,
"grad_norm": 0.4664464592933655,
"learning_rate": 0.0001619402181230645,
"loss": 2.9037,
"step": 108550
},
{
"epoch": 29.225290697674417,
"grad_norm": 0.5049362778663635,
"learning_rate": 0.00016173825232260669,
"loss": 2.8973,
"step": 108600
},
{
"epoch": 29.238748923341948,
"grad_norm": 0.4829135239124298,
"learning_rate": 0.00016153628652214888,
"loss": 2.8926,
"step": 108650
},
{
"epoch": 29.252207149009475,
"grad_norm": 0.45376384258270264,
"learning_rate": 0.0001613343207216911,
"loss": 2.8905,
"step": 108700
},
{
"epoch": 29.265665374677003,
"grad_norm": 0.5056002140045166,
"learning_rate": 0.0001611323549212333,
"loss": 2.8993,
"step": 108750
},
{
"epoch": 29.27912360034453,
"grad_norm": 0.4627412259578705,
"learning_rate": 0.00016093038912077555,
"loss": 2.8979,
"step": 108800
},
{
"epoch": 29.292581826012057,
"grad_norm": 0.48572027683258057,
"learning_rate": 0.00016072842332031775,
"loss": 2.9072,
"step": 108850
},
{
"epoch": 29.306040051679588,
"grad_norm": 0.4790059030056,
"learning_rate": 0.00016052645751985997,
"loss": 2.8884,
"step": 108900
},
{
"epoch": 29.319498277347115,
"grad_norm": 0.4833512604236603,
"learning_rate": 0.00016032449171940217,
"loss": 2.9017,
"step": 108950
},
{
"epoch": 29.332956503014643,
"grad_norm": 0.49676379561424255,
"learning_rate": 0.0001601225259189444,
"loss": 2.9124,
"step": 109000
},
{
"epoch": 29.332956503014643,
"eval_accuracy": 0.3977577679242323,
"eval_loss": 3.285517454147339,
"eval_runtime": 186.7918,
"eval_samples_per_second": 96.428,
"eval_steps_per_second": 6.028,
"step": 109000
},
{
"epoch": 29.34641472868217,
"grad_norm": 0.46966609358787537,
"learning_rate": 0.00015992056011848659,
"loss": 2.9049,
"step": 109050
},
{
"epoch": 29.359872954349697,
"grad_norm": 0.4910942316055298,
"learning_rate": 0.00015971859431802878,
"loss": 2.8899,
"step": 109100
},
{
"epoch": 29.373331180017228,
"grad_norm": 0.489801824092865,
"learning_rate": 0.000159516628517571,
"loss": 2.9026,
"step": 109150
},
{
"epoch": 29.386789405684755,
"grad_norm": 0.7795575857162476,
"learning_rate": 0.0001593146627171132,
"loss": 2.9124,
"step": 109200
},
{
"epoch": 29.400247631352283,
"grad_norm": 0.48971161246299744,
"learning_rate": 0.00015911269691665545,
"loss": 2.9079,
"step": 109250
},
{
"epoch": 29.41370585701981,
"grad_norm": 0.46526631712913513,
"learning_rate": 0.00015891073111619765,
"loss": 2.8997,
"step": 109300
},
{
"epoch": 29.427164082687337,
"grad_norm": 0.48566874861717224,
"learning_rate": 0.00015870876531573987,
"loss": 2.9212,
"step": 109350
},
{
"epoch": 29.440622308354868,
"grad_norm": 0.4755590260028839,
"learning_rate": 0.00015850679951528207,
"loss": 2.908,
"step": 109400
},
{
"epoch": 29.454080534022395,
"grad_norm": 0.4979674816131592,
"learning_rate": 0.00015830483371482426,
"loss": 2.9114,
"step": 109450
},
{
"epoch": 29.467538759689923,
"grad_norm": 0.47721847891807556,
"learning_rate": 0.00015810286791436649,
"loss": 2.9074,
"step": 109500
},
{
"epoch": 29.48099698535745,
"grad_norm": 0.4785979092121124,
"learning_rate": 0.00015790090211390868,
"loss": 2.9206,
"step": 109550
},
{
"epoch": 29.494455211024977,
"grad_norm": 0.465573251247406,
"learning_rate": 0.0001576989363134509,
"loss": 2.9084,
"step": 109600
},
{
"epoch": 29.507913436692505,
"grad_norm": 0.502953827381134,
"learning_rate": 0.0001574969705129931,
"loss": 2.9027,
"step": 109650
},
{
"epoch": 29.521371662360036,
"grad_norm": 0.4925035238265991,
"learning_rate": 0.00015729500471253535,
"loss": 2.9009,
"step": 109700
},
{
"epoch": 29.534829888027563,
"grad_norm": 0.48377102613449097,
"learning_rate": 0.00015709303891207755,
"loss": 2.9149,
"step": 109750
},
{
"epoch": 29.54828811369509,
"grad_norm": 0.49283069372177124,
"learning_rate": 0.00015689107311161977,
"loss": 2.9203,
"step": 109800
},
{
"epoch": 29.561746339362617,
"grad_norm": 0.4812520742416382,
"learning_rate": 0.00015668910731116197,
"loss": 2.9072,
"step": 109850
},
{
"epoch": 29.575204565030145,
"grad_norm": 0.4918520152568817,
"learning_rate": 0.00015648714151070416,
"loss": 2.9186,
"step": 109900
},
{
"epoch": 29.588662790697676,
"grad_norm": 0.46839213371276855,
"learning_rate": 0.0001562851757102464,
"loss": 2.9166,
"step": 109950
},
{
"epoch": 29.602121016365203,
"grad_norm": 0.48223790526390076,
"learning_rate": 0.00015608320990978858,
"loss": 2.9185,
"step": 110000
},
{
"epoch": 29.602121016365203,
"eval_accuracy": 0.39786011326217396,
"eval_loss": 3.280489206314087,
"eval_runtime": 186.0675,
"eval_samples_per_second": 96.804,
"eval_steps_per_second": 6.052,
"step": 110000
},
{
"epoch": 29.61557924203273,
"grad_norm": 0.4872387945652008,
"learning_rate": 0.0001558812441093308,
"loss": 2.9188,
"step": 110050
},
{
"epoch": 29.629037467700257,
"grad_norm": 0.5209236741065979,
"learning_rate": 0.000155679278308873,
"loss": 2.9155,
"step": 110100
},
{
"epoch": 29.642495693367785,
"grad_norm": 0.49196842312812805,
"learning_rate": 0.00015547731250841525,
"loss": 2.9113,
"step": 110150
},
{
"epoch": 29.655953919035316,
"grad_norm": 0.4668411612510681,
"learning_rate": 0.00015527534670795745,
"loss": 2.9194,
"step": 110200
},
{
"epoch": 29.669412144702843,
"grad_norm": 0.502856969833374,
"learning_rate": 0.00015507338090749964,
"loss": 2.915,
"step": 110250
},
{
"epoch": 29.68287037037037,
"grad_norm": 0.47303318977355957,
"learning_rate": 0.00015487141510704187,
"loss": 2.9272,
"step": 110300
},
{
"epoch": 29.696328596037898,
"grad_norm": 0.5249906778335571,
"learning_rate": 0.00015466944930658406,
"loss": 2.9217,
"step": 110350
},
{
"epoch": 29.709786821705425,
"grad_norm": 0.49294140934944153,
"learning_rate": 0.0001544674835061263,
"loss": 2.9152,
"step": 110400
},
{
"epoch": 29.723245047372956,
"grad_norm": 0.4729110300540924,
"learning_rate": 0.00015426551770566848,
"loss": 2.9122,
"step": 110450
},
{
"epoch": 29.736703273040483,
"grad_norm": 0.4749280512332916,
"learning_rate": 0.00015406355190521068,
"loss": 2.907,
"step": 110500
},
{
"epoch": 29.75016149870801,
"grad_norm": 0.47957757115364075,
"learning_rate": 0.0001538615861047529,
"loss": 2.9292,
"step": 110550
},
{
"epoch": 29.763619724375538,
"grad_norm": 0.48484504222869873,
"learning_rate": 0.00015365962030429513,
"loss": 2.9307,
"step": 110600
},
{
"epoch": 29.777077950043065,
"grad_norm": 0.500281572341919,
"learning_rate": 0.00015345765450383735,
"loss": 2.9245,
"step": 110650
},
{
"epoch": 29.790536175710596,
"grad_norm": 0.4570538401603699,
"learning_rate": 0.00015325568870337955,
"loss": 2.9293,
"step": 110700
},
{
"epoch": 29.803994401378123,
"grad_norm": 0.5007463693618774,
"learning_rate": 0.00015305372290292177,
"loss": 2.9221,
"step": 110750
},
{
"epoch": 29.81745262704565,
"grad_norm": 0.48854824900627136,
"learning_rate": 0.00015285175710246396,
"loss": 2.9168,
"step": 110800
},
{
"epoch": 29.830910852713178,
"grad_norm": 0.48463812470436096,
"learning_rate": 0.0001526497913020062,
"loss": 2.9295,
"step": 110850
},
{
"epoch": 29.844369078380705,
"grad_norm": 0.5036105513572693,
"learning_rate": 0.00015244782550154838,
"loss": 2.9289,
"step": 110900
},
{
"epoch": 29.857827304048236,
"grad_norm": 0.490828275680542,
"learning_rate": 0.00015224585970109058,
"loss": 2.9352,
"step": 110950
},
{
"epoch": 29.871285529715763,
"grad_norm": 0.47763335704803467,
"learning_rate": 0.00015204389390063283,
"loss": 2.9229,
"step": 111000
},
{
"epoch": 29.871285529715763,
"eval_accuracy": 0.39816628010115457,
"eval_loss": 3.2758564949035645,
"eval_runtime": 186.0842,
"eval_samples_per_second": 96.795,
"eval_steps_per_second": 6.051,
"step": 111000
},
{
"epoch": 29.88474375538329,
"grad_norm": 0.49680614471435547,
"learning_rate": 0.00015184192810017503,
"loss": 2.9208,
"step": 111050
},
{
"epoch": 29.898201981050818,
"grad_norm": 0.48516368865966797,
"learning_rate": 0.00015163996229971725,
"loss": 2.9322,
"step": 111100
},
{
"epoch": 29.911660206718345,
"grad_norm": 0.481813907623291,
"learning_rate": 0.00015143799649925945,
"loss": 2.9293,
"step": 111150
},
{
"epoch": 29.925118432385876,
"grad_norm": 0.4639509618282318,
"learning_rate": 0.00015123603069880167,
"loss": 2.9229,
"step": 111200
},
{
"epoch": 29.938576658053403,
"grad_norm": 0.5075692534446716,
"learning_rate": 0.00015103406489834386,
"loss": 2.9337,
"step": 111250
},
{
"epoch": 29.95203488372093,
"grad_norm": 0.48359689116477966,
"learning_rate": 0.00015083209909788606,
"loss": 2.9307,
"step": 111300
},
{
"epoch": 29.965493109388458,
"grad_norm": 0.5042138695716858,
"learning_rate": 0.00015063013329742828,
"loss": 2.9382,
"step": 111350
},
{
"epoch": 29.978951335055985,
"grad_norm": 0.481554239988327,
"learning_rate": 0.00015042816749697048,
"loss": 2.9281,
"step": 111400
},
{
"epoch": 29.992409560723516,
"grad_norm": 0.4682770371437073,
"learning_rate": 0.00015022620169651273,
"loss": 2.9381,
"step": 111450
},
{
"epoch": 30.00565245478036,
"grad_norm": 0.47752559185028076,
"learning_rate": 0.00015002423589605493,
"loss": 2.9019,
"step": 111500
},
{
"epoch": 30.01911068044789,
"grad_norm": 0.47453469038009644,
"learning_rate": 0.00014982227009559712,
"loss": 2.8578,
"step": 111550
},
{
"epoch": 30.03256890611542,
"grad_norm": 0.4740888476371765,
"learning_rate": 0.00014962030429513935,
"loss": 2.8593,
"step": 111600
},
{
"epoch": 30.046027131782946,
"grad_norm": 0.4668754041194916,
"learning_rate": 0.00014941833849468154,
"loss": 2.8571,
"step": 111650
},
{
"epoch": 30.059485357450473,
"grad_norm": 0.4677064120769501,
"learning_rate": 0.00014921637269422377,
"loss": 2.8622,
"step": 111700
},
{
"epoch": 30.072943583118,
"grad_norm": 0.478365957736969,
"learning_rate": 0.00014901440689376596,
"loss": 2.864,
"step": 111750
},
{
"epoch": 30.08640180878553,
"grad_norm": 0.46888357400894165,
"learning_rate": 0.00014881244109330818,
"loss": 2.8654,
"step": 111800
},
{
"epoch": 30.09986003445306,
"grad_norm": 0.46399155259132385,
"learning_rate": 0.0001486104752928504,
"loss": 2.8733,
"step": 111850
},
{
"epoch": 30.113318260120586,
"grad_norm": 0.4954034984111786,
"learning_rate": 0.0001484085094923926,
"loss": 2.8637,
"step": 111900
},
{
"epoch": 30.126776485788113,
"grad_norm": 0.5087417960166931,
"learning_rate": 0.00014820654369193483,
"loss": 2.8815,
"step": 111950
},
{
"epoch": 30.14023471145564,
"grad_norm": 0.4896148443222046,
"learning_rate": 0.00014800457789147702,
"loss": 2.8683,
"step": 112000
},
{
"epoch": 30.14023471145564,
"eval_accuracy": 0.3974803925019763,
"eval_loss": 3.289569139480591,
"eval_runtime": 186.0631,
"eval_samples_per_second": 96.806,
"eval_steps_per_second": 6.052,
"step": 112000
},
{
"epoch": 30.153692937123168,
"grad_norm": 0.491746187210083,
"learning_rate": 0.00014780261209101925,
"loss": 2.8838,
"step": 112050
},
{
"epoch": 30.1671511627907,
"grad_norm": 0.48746776580810547,
"learning_rate": 0.00014760064629056144,
"loss": 2.8786,
"step": 112100
},
{
"epoch": 30.180609388458226,
"grad_norm": 0.5001115798950195,
"learning_rate": 0.00014739868049010367,
"loss": 2.8733,
"step": 112150
},
{
"epoch": 30.194067614125753,
"grad_norm": 0.49509507417678833,
"learning_rate": 0.00014719671468964586,
"loss": 2.8865,
"step": 112200
},
{
"epoch": 30.20752583979328,
"grad_norm": 0.5134970545768738,
"learning_rate": 0.00014699474888918808,
"loss": 2.8814,
"step": 112250
},
{
"epoch": 30.220984065460808,
"grad_norm": 0.47717851400375366,
"learning_rate": 0.0001467927830887303,
"loss": 2.8865,
"step": 112300
},
{
"epoch": 30.23444229112834,
"grad_norm": 0.4734904170036316,
"learning_rate": 0.0001465908172882725,
"loss": 2.8806,
"step": 112350
},
{
"epoch": 30.247900516795866,
"grad_norm": 0.5131638050079346,
"learning_rate": 0.0001463888514878147,
"loss": 2.8897,
"step": 112400
},
{
"epoch": 30.261358742463393,
"grad_norm": 0.49894216656684875,
"learning_rate": 0.00014618688568735692,
"loss": 2.8925,
"step": 112450
},
{
"epoch": 30.27481696813092,
"grad_norm": 0.5135336518287659,
"learning_rate": 0.00014598491988689915,
"loss": 2.8844,
"step": 112500
},
{
"epoch": 30.288275193798448,
"grad_norm": 0.48499059677124023,
"learning_rate": 0.00014578295408644134,
"loss": 2.8952,
"step": 112550
},
{
"epoch": 30.30173341946598,
"grad_norm": 0.48516878485679626,
"learning_rate": 0.00014558098828598357,
"loss": 2.8949,
"step": 112600
},
{
"epoch": 30.315191645133506,
"grad_norm": 0.5160548090934753,
"learning_rate": 0.00014537902248552576,
"loss": 2.8986,
"step": 112650
},
{
"epoch": 30.328649870801033,
"grad_norm": 0.463051438331604,
"learning_rate": 0.00014517705668506799,
"loss": 2.8848,
"step": 112700
},
{
"epoch": 30.34210809646856,
"grad_norm": 0.5012433528900146,
"learning_rate": 0.0001449750908846102,
"loss": 2.8885,
"step": 112750
},
{
"epoch": 30.355566322136088,
"grad_norm": 0.4897739291191101,
"learning_rate": 0.0001447731250841524,
"loss": 2.8898,
"step": 112800
},
{
"epoch": 30.36902454780362,
"grad_norm": 0.49385741353034973,
"learning_rate": 0.0001445711592836946,
"loss": 2.8853,
"step": 112850
},
{
"epoch": 30.382482773471146,
"grad_norm": 0.47586560249328613,
"learning_rate": 0.00014436919348323682,
"loss": 2.905,
"step": 112900
},
{
"epoch": 30.395940999138674,
"grad_norm": 0.5367991924285889,
"learning_rate": 0.00014416722768277905,
"loss": 2.885,
"step": 112950
},
{
"epoch": 30.4093992248062,
"grad_norm": 0.49246811866760254,
"learning_rate": 0.00014396526188232124,
"loss": 2.89,
"step": 113000
},
{
"epoch": 30.4093992248062,
"eval_accuracy": 0.3974334570603725,
"eval_loss": 3.2866954803466797,
"eval_runtime": 185.9041,
"eval_samples_per_second": 96.889,
"eval_steps_per_second": 6.057,
"step": 113000
},
{
"epoch": 30.422857450473728,
"grad_norm": 0.5175289511680603,
"learning_rate": 0.00014376329608186347,
"loss": 2.9148,
"step": 113050
},
{
"epoch": 30.43631567614126,
"grad_norm": 0.5106175541877747,
"learning_rate": 0.00014356133028140566,
"loss": 2.8992,
"step": 113100
},
{
"epoch": 30.449773901808786,
"grad_norm": 0.49530425667762756,
"learning_rate": 0.00014335936448094789,
"loss": 2.9004,
"step": 113150
},
{
"epoch": 30.463232127476314,
"grad_norm": 0.5244776010513306,
"learning_rate": 0.00014315739868049008,
"loss": 2.8983,
"step": 113200
},
{
"epoch": 30.47669035314384,
"grad_norm": 0.4911174476146698,
"learning_rate": 0.0001429554328800323,
"loss": 2.9019,
"step": 113250
},
{
"epoch": 30.490148578811368,
"grad_norm": 0.48649704456329346,
"learning_rate": 0.0001427534670795745,
"loss": 2.9027,
"step": 113300
},
{
"epoch": 30.5036068044789,
"grad_norm": 0.4809096157550812,
"learning_rate": 0.00014255150127911672,
"loss": 2.9037,
"step": 113350
},
{
"epoch": 30.517065030146426,
"grad_norm": 0.5086159706115723,
"learning_rate": 0.00014234953547865895,
"loss": 2.8907,
"step": 113400
},
{
"epoch": 30.530523255813954,
"grad_norm": 0.49828189611434937,
"learning_rate": 0.00014214756967820114,
"loss": 2.9053,
"step": 113450
},
{
"epoch": 30.54398148148148,
"grad_norm": 0.4803487956523895,
"learning_rate": 0.00014194560387774334,
"loss": 2.9091,
"step": 113500
},
{
"epoch": 30.55743970714901,
"grad_norm": 0.5035832524299622,
"learning_rate": 0.00014174363807728556,
"loss": 2.9033,
"step": 113550
},
{
"epoch": 30.57089793281654,
"grad_norm": 0.5142761468887329,
"learning_rate": 0.00014154167227682779,
"loss": 2.8921,
"step": 113600
},
{
"epoch": 30.584356158484066,
"grad_norm": 0.46249037981033325,
"learning_rate": 0.00014133970647636998,
"loss": 2.9063,
"step": 113650
},
{
"epoch": 30.597814384151594,
"grad_norm": 0.4805637300014496,
"learning_rate": 0.0001411377406759122,
"loss": 2.9011,
"step": 113700
},
{
"epoch": 30.61127260981912,
"grad_norm": 0.4967007339000702,
"learning_rate": 0.0001409357748754544,
"loss": 2.9185,
"step": 113750
},
{
"epoch": 30.62473083548665,
"grad_norm": 0.49374496936798096,
"learning_rate": 0.00014073380907499662,
"loss": 2.9065,
"step": 113800
},
{
"epoch": 30.63818906115418,
"grad_norm": 0.5127381086349487,
"learning_rate": 0.00014053184327453885,
"loss": 2.9112,
"step": 113850
},
{
"epoch": 30.651647286821706,
"grad_norm": 0.4961664080619812,
"learning_rate": 0.00014032987747408104,
"loss": 2.9109,
"step": 113900
},
{
"epoch": 30.665105512489234,
"grad_norm": 0.4843076765537262,
"learning_rate": 0.00014012791167362324,
"loss": 2.9191,
"step": 113950
},
{
"epoch": 30.67856373815676,
"grad_norm": 0.4942310154438019,
"learning_rate": 0.00013992594587316546,
"loss": 2.9145,
"step": 114000
},
{
"epoch": 30.67856373815676,
"eval_accuracy": 0.39781676316680376,
"eval_loss": 3.2825369834899902,
"eval_runtime": 186.2197,
"eval_samples_per_second": 96.724,
"eval_steps_per_second": 6.047,
"step": 114000
},
{
"epoch": 30.69202196382429,
"grad_norm": 0.5027679800987244,
"learning_rate": 0.0001397239800727077,
"loss": 2.9171,
"step": 114050
},
{
"epoch": 30.705480189491816,
"grad_norm": 0.5086604952812195,
"learning_rate": 0.00013952201427224988,
"loss": 2.9179,
"step": 114100
},
{
"epoch": 30.718938415159347,
"grad_norm": 0.497111976146698,
"learning_rate": 0.0001393200484717921,
"loss": 2.9044,
"step": 114150
},
{
"epoch": 30.732396640826874,
"grad_norm": 0.4975685775279999,
"learning_rate": 0.0001391180826713343,
"loss": 2.9177,
"step": 114200
},
{
"epoch": 30.7458548664944,
"grad_norm": 0.4827300012111664,
"learning_rate": 0.00013891611687087653,
"loss": 2.9161,
"step": 114250
},
{
"epoch": 30.75931309216193,
"grad_norm": 0.4842374324798584,
"learning_rate": 0.00013871415107041872,
"loss": 2.9154,
"step": 114300
},
{
"epoch": 30.772771317829456,
"grad_norm": 0.48095107078552246,
"learning_rate": 0.00013851218526996094,
"loss": 2.9112,
"step": 114350
},
{
"epoch": 30.786229543496987,
"grad_norm": 0.487163782119751,
"learning_rate": 0.00013831021946950314,
"loss": 2.9244,
"step": 114400
},
{
"epoch": 30.799687769164514,
"grad_norm": 0.49638232588768005,
"learning_rate": 0.00013810825366904536,
"loss": 2.9157,
"step": 114450
},
{
"epoch": 30.81314599483204,
"grad_norm": 0.5074690580368042,
"learning_rate": 0.0001379062878685876,
"loss": 2.9111,
"step": 114500
},
{
"epoch": 30.82660422049957,
"grad_norm": 0.4984564185142517,
"learning_rate": 0.00013770432206812978,
"loss": 2.9123,
"step": 114550
},
{
"epoch": 30.840062446167096,
"grad_norm": 0.4880812168121338,
"learning_rate": 0.00013750235626767198,
"loss": 2.9248,
"step": 114600
},
{
"epoch": 30.853520671834627,
"grad_norm": 0.5138365030288696,
"learning_rate": 0.0001373003904672142,
"loss": 2.9122,
"step": 114650
},
{
"epoch": 30.866978897502154,
"grad_norm": 0.48817211389541626,
"learning_rate": 0.00013709842466675643,
"loss": 2.9181,
"step": 114700
},
{
"epoch": 30.88043712316968,
"grad_norm": 0.48954612016677856,
"learning_rate": 0.00013689645886629862,
"loss": 2.9186,
"step": 114750
},
{
"epoch": 30.89389534883721,
"grad_norm": 0.48388442397117615,
"learning_rate": 0.00013669449306584084,
"loss": 2.9266,
"step": 114800
},
{
"epoch": 30.907353574504736,
"grad_norm": 0.4795021414756775,
"learning_rate": 0.00013649252726538304,
"loss": 2.9177,
"step": 114850
},
{
"epoch": 30.920811800172267,
"grad_norm": 0.4757089912891388,
"learning_rate": 0.00013629056146492526,
"loss": 2.9244,
"step": 114900
},
{
"epoch": 30.934270025839794,
"grad_norm": 0.497738242149353,
"learning_rate": 0.0001360885956644675,
"loss": 2.9185,
"step": 114950
},
{
"epoch": 30.94772825150732,
"grad_norm": 0.498090535402298,
"learning_rate": 0.00013588662986400968,
"loss": 2.9325,
"step": 115000
},
{
"epoch": 30.94772825150732,
"eval_accuracy": 0.39826525738657376,
"eval_loss": 3.277034282684326,
"eval_runtime": 184.7264,
"eval_samples_per_second": 97.506,
"eval_steps_per_second": 6.096,
"step": 115000
},
{
"epoch": 30.96118647717485,
"grad_norm": 0.5387504696846008,
"learning_rate": 0.00013568466406355188,
"loss": 2.916,
"step": 115050
},
{
"epoch": 30.974644702842376,
"grad_norm": 0.49516457319259644,
"learning_rate": 0.0001354826982630941,
"loss": 2.919,
"step": 115100
},
{
"epoch": 30.988102928509907,
"grad_norm": 0.47475531697273254,
"learning_rate": 0.00013528073246263633,
"loss": 2.9093,
"step": 115150
},
{
"epoch": 31.00134582256675,
"grad_norm": 0.476597398519516,
"learning_rate": 0.00013507876666217852,
"loss": 2.9159,
"step": 115200
},
{
"epoch": 31.014804048234282,
"grad_norm": 0.49474477767944336,
"learning_rate": 0.00013487680086172075,
"loss": 2.8627,
"step": 115250
},
{
"epoch": 31.02826227390181,
"grad_norm": 0.4991047978401184,
"learning_rate": 0.00013467483506126294,
"loss": 2.8499,
"step": 115300
},
{
"epoch": 31.041720499569337,
"grad_norm": 0.5250011086463928,
"learning_rate": 0.00013447286926080516,
"loss": 2.8637,
"step": 115350
},
{
"epoch": 31.055178725236864,
"grad_norm": 0.48069050908088684,
"learning_rate": 0.00013427090346034736,
"loss": 2.8583,
"step": 115400
},
{
"epoch": 31.06863695090439,
"grad_norm": 0.49658381938934326,
"learning_rate": 0.00013406893765988958,
"loss": 2.8617,
"step": 115450
},
{
"epoch": 31.082095176571922,
"grad_norm": 0.49932461977005005,
"learning_rate": 0.00013386697185943178,
"loss": 2.8644,
"step": 115500
},
{
"epoch": 31.09555340223945,
"grad_norm": 0.49038806557655334,
"learning_rate": 0.000133665006058974,
"loss": 2.872,
"step": 115550
},
{
"epoch": 31.109011627906977,
"grad_norm": 0.46469274163246155,
"learning_rate": 0.00013346304025851623,
"loss": 2.8664,
"step": 115600
},
{
"epoch": 31.122469853574504,
"grad_norm": 0.503940761089325,
"learning_rate": 0.00013326107445805842,
"loss": 2.8732,
"step": 115650
},
{
"epoch": 31.13592807924203,
"grad_norm": 0.49561211466789246,
"learning_rate": 0.00013305910865760065,
"loss": 2.8661,
"step": 115700
},
{
"epoch": 31.149386304909562,
"grad_norm": 0.4893884062767029,
"learning_rate": 0.00013285714285714284,
"loss": 2.8591,
"step": 115750
},
{
"epoch": 31.16284453057709,
"grad_norm": 0.5082573890686035,
"learning_rate": 0.00013265517705668506,
"loss": 2.8542,
"step": 115800
},
{
"epoch": 31.176302756244617,
"grad_norm": 0.4964279234409332,
"learning_rate": 0.00013245321125622726,
"loss": 2.8785,
"step": 115850
},
{
"epoch": 31.189760981912144,
"grad_norm": 0.48469647765159607,
"learning_rate": 0.00013225124545576948,
"loss": 2.8716,
"step": 115900
},
{
"epoch": 31.20321920757967,
"grad_norm": 0.5207317471504211,
"learning_rate": 0.00013204927965531168,
"loss": 2.8698,
"step": 115950
},
{
"epoch": 31.216677433247202,
"grad_norm": 0.5387663841247559,
"learning_rate": 0.0001318473138548539,
"loss": 2.8755,
"step": 116000
},
{
"epoch": 31.216677433247202,
"eval_accuracy": 0.39777623788967825,
"eval_loss": 3.2870469093322754,
"eval_runtime": 184.8146,
"eval_samples_per_second": 97.46,
"eval_steps_per_second": 6.093,
"step": 116000
},
{
"epoch": 31.23013565891473,
"grad_norm": 0.5226250290870667,
"learning_rate": 0.00013164534805439613,
"loss": 2.875,
"step": 116050
},
{
"epoch": 31.243593884582257,
"grad_norm": 0.49980151653289795,
"learning_rate": 0.00013144338225393832,
"loss": 2.8832,
"step": 116100
},
{
"epoch": 31.257052110249784,
"grad_norm": 0.49096962809562683,
"learning_rate": 0.00013124141645348052,
"loss": 2.8718,
"step": 116150
},
{
"epoch": 31.27051033591731,
"grad_norm": 0.5504794120788574,
"learning_rate": 0.00013103945065302274,
"loss": 2.8813,
"step": 116200
},
{
"epoch": 31.283968561584842,
"grad_norm": 0.5013906955718994,
"learning_rate": 0.00013083748485256497,
"loss": 2.8819,
"step": 116250
},
{
"epoch": 31.29742678725237,
"grad_norm": 0.508328378200531,
"learning_rate": 0.00013063551905210716,
"loss": 2.878,
"step": 116300
},
{
"epoch": 31.310885012919897,
"grad_norm": 0.4899929165840149,
"learning_rate": 0.00013043355325164938,
"loss": 2.8779,
"step": 116350
},
{
"epoch": 31.324343238587424,
"grad_norm": 0.5202932953834534,
"learning_rate": 0.00013023158745119158,
"loss": 2.8792,
"step": 116400
},
{
"epoch": 31.33780146425495,
"grad_norm": 0.49529343843460083,
"learning_rate": 0.00013002962165073378,
"loss": 2.8858,
"step": 116450
},
{
"epoch": 31.35125968992248,
"grad_norm": 0.5189668536186218,
"learning_rate": 0.000129827655850276,
"loss": 2.8812,
"step": 116500
},
{
"epoch": 31.36471791559001,
"grad_norm": 0.5060351490974426,
"learning_rate": 0.00012962569004981822,
"loss": 2.8864,
"step": 116550
},
{
"epoch": 31.378176141257537,
"grad_norm": 0.5147632956504822,
"learning_rate": 0.00012942372424936042,
"loss": 2.8876,
"step": 116600
},
{
"epoch": 31.391634366925064,
"grad_norm": 0.5026252865791321,
"learning_rate": 0.00012922175844890264,
"loss": 2.8867,
"step": 116650
},
{
"epoch": 31.40509259259259,
"grad_norm": 0.5080686807632446,
"learning_rate": 0.00012901979264844487,
"loss": 2.8923,
"step": 116700
},
{
"epoch": 31.41855081826012,
"grad_norm": 0.5131353139877319,
"learning_rate": 0.00012881782684798706,
"loss": 2.8778,
"step": 116750
},
{
"epoch": 31.43200904392765,
"grad_norm": 0.5074872374534607,
"learning_rate": 0.00012861586104752928,
"loss": 2.8855,
"step": 116800
},
{
"epoch": 31.445467269595177,
"grad_norm": 0.5141700506210327,
"learning_rate": 0.00012841389524707148,
"loss": 2.895,
"step": 116850
},
{
"epoch": 31.458925495262704,
"grad_norm": 0.4880288243293762,
"learning_rate": 0.0001282119294466137,
"loss": 2.9006,
"step": 116900
},
{
"epoch": 31.47238372093023,
"grad_norm": 0.5205245018005371,
"learning_rate": 0.0001280099636461559,
"loss": 2.8948,
"step": 116950
},
{
"epoch": 31.48584194659776,
"grad_norm": 0.4987119436264038,
"learning_rate": 0.00012780799784569812,
"loss": 2.8941,
"step": 117000
},
{
"epoch": 31.48584194659776,
"eval_accuracy": 0.39796560935892705,
"eval_loss": 3.2823922634124756,
"eval_runtime": 183.6286,
"eval_samples_per_second": 98.089,
"eval_steps_per_second": 6.132,
"step": 117000
},
{
"epoch": 31.49930017226529,
"grad_norm": 0.4958367645740509,
"learning_rate": 0.00012760603204524032,
"loss": 2.8822,
"step": 117050
},
{
"epoch": 31.512758397932817,
"grad_norm": 0.4825536608695984,
"learning_rate": 0.00012740406624478254,
"loss": 2.8946,
"step": 117100
},
{
"epoch": 31.526216623600344,
"grad_norm": 0.507009744644165,
"learning_rate": 0.00012720210044432477,
"loss": 2.8953,
"step": 117150
},
{
"epoch": 31.539674849267872,
"grad_norm": 0.5003730654716492,
"learning_rate": 0.00012700013464386696,
"loss": 2.8973,
"step": 117200
},
{
"epoch": 31.5531330749354,
"grad_norm": 0.4873824715614319,
"learning_rate": 0.00012679816884340916,
"loss": 2.8955,
"step": 117250
},
{
"epoch": 31.56659130060293,
"grad_norm": 0.5300673842430115,
"learning_rate": 0.00012659620304295138,
"loss": 2.8836,
"step": 117300
},
{
"epoch": 31.580049526270457,
"grad_norm": 0.4678681790828705,
"learning_rate": 0.0001263942372424936,
"loss": 2.8932,
"step": 117350
},
{
"epoch": 31.593507751937985,
"grad_norm": 0.49590036273002625,
"learning_rate": 0.0001261922714420358,
"loss": 2.9003,
"step": 117400
},
{
"epoch": 31.606965977605512,
"grad_norm": 0.5043445825576782,
"learning_rate": 0.00012599030564157802,
"loss": 2.899,
"step": 117450
},
{
"epoch": 31.62042420327304,
"grad_norm": 0.4818818271160126,
"learning_rate": 0.00012578833984112022,
"loss": 2.8976,
"step": 117500
},
{
"epoch": 31.63388242894057,
"grad_norm": 0.4927334487438202,
"learning_rate": 0.00012558637404066244,
"loss": 2.9034,
"step": 117550
},
{
"epoch": 31.647340654608097,
"grad_norm": 0.490279883146286,
"learning_rate": 0.00012538440824020467,
"loss": 2.9005,
"step": 117600
},
{
"epoch": 31.660798880275625,
"grad_norm": 0.5263948440551758,
"learning_rate": 0.00012518244243974686,
"loss": 2.899,
"step": 117650
},
{
"epoch": 31.674257105943152,
"grad_norm": 0.490869402885437,
"learning_rate": 0.00012498047663928906,
"loss": 2.9017,
"step": 117700
},
{
"epoch": 31.68771533161068,
"grad_norm": 0.5171840786933899,
"learning_rate": 0.00012477851083883128,
"loss": 2.8941,
"step": 117750
},
{
"epoch": 31.70117355727821,
"grad_norm": 0.48714497685432434,
"learning_rate": 0.0001245765450383735,
"loss": 2.907,
"step": 117800
},
{
"epoch": 31.714631782945737,
"grad_norm": 0.5022086501121521,
"learning_rate": 0.0001243745792379157,
"loss": 2.9019,
"step": 117850
},
{
"epoch": 31.728090008613265,
"grad_norm": 0.49601811170578003,
"learning_rate": 0.00012417261343745792,
"loss": 2.9042,
"step": 117900
},
{
"epoch": 31.741548234280792,
"grad_norm": 0.505938708782196,
"learning_rate": 0.00012397064763700012,
"loss": 2.9157,
"step": 117950
},
{
"epoch": 31.75500645994832,
"grad_norm": 0.54503333568573,
"learning_rate": 0.00012376868183654232,
"loss": 2.9035,
"step": 118000
},
{
"epoch": 31.75500645994832,
"eval_accuracy": 0.39847592363951323,
"eval_loss": 3.2784652709960938,
"eval_runtime": 183.7323,
"eval_samples_per_second": 98.034,
"eval_steps_per_second": 6.128,
"step": 118000
},
{
"epoch": 31.768464685615847,
"grad_norm": 0.5179860591888428,
"learning_rate": 0.00012356671603608454,
"loss": 2.8986,
"step": 118050
},
{
"epoch": 31.781922911283377,
"grad_norm": 0.48850664496421814,
"learning_rate": 0.00012336475023562676,
"loss": 2.9016,
"step": 118100
},
{
"epoch": 31.795381136950905,
"grad_norm": 0.4846647083759308,
"learning_rate": 0.00012316278443516896,
"loss": 2.9058,
"step": 118150
},
{
"epoch": 31.808839362618432,
"grad_norm": 0.5397312641143799,
"learning_rate": 0.00012296081863471118,
"loss": 2.9001,
"step": 118200
},
{
"epoch": 31.82229758828596,
"grad_norm": 0.503955602645874,
"learning_rate": 0.0001227588528342534,
"loss": 2.9031,
"step": 118250
},
{
"epoch": 31.83575581395349,
"grad_norm": 0.5033538937568665,
"learning_rate": 0.0001225568870337956,
"loss": 2.9098,
"step": 118300
},
{
"epoch": 31.849214039621017,
"grad_norm": 0.5270665884017944,
"learning_rate": 0.0001223549212333378,
"loss": 2.9136,
"step": 118350
},
{
"epoch": 31.862672265288545,
"grad_norm": 0.5030663013458252,
"learning_rate": 0.00012215295543288002,
"loss": 2.9095,
"step": 118400
},
{
"epoch": 31.876130490956072,
"grad_norm": 0.4942440092563629,
"learning_rate": 0.00012195098963242223,
"loss": 2.9153,
"step": 118450
},
{
"epoch": 31.8895887166236,
"grad_norm": 0.4918186664581299,
"learning_rate": 0.00012174902383196445,
"loss": 2.8934,
"step": 118500
},
{
"epoch": 31.903046942291127,
"grad_norm": 0.5160676836967468,
"learning_rate": 0.00012154705803150665,
"loss": 2.9054,
"step": 118550
},
{
"epoch": 31.916505167958658,
"grad_norm": 0.5123493671417236,
"learning_rate": 0.00012134509223104886,
"loss": 2.9041,
"step": 118600
},
{
"epoch": 31.929963393626185,
"grad_norm": 0.5015071630477905,
"learning_rate": 0.00012114312643059107,
"loss": 2.9036,
"step": 118650
},
{
"epoch": 31.943421619293712,
"grad_norm": 0.5266225337982178,
"learning_rate": 0.00012094116063013329,
"loss": 2.9007,
"step": 118700
},
{
"epoch": 31.95687984496124,
"grad_norm": 0.5361496210098267,
"learning_rate": 0.0001207391948296755,
"loss": 2.9064,
"step": 118750
},
{
"epoch": 31.970338070628767,
"grad_norm": 0.5224943161010742,
"learning_rate": 0.00012053722902921771,
"loss": 2.9031,
"step": 118800
},
{
"epoch": 31.983796296296298,
"grad_norm": 0.5123460292816162,
"learning_rate": 0.00012033526322875992,
"loss": 2.9105,
"step": 118850
},
{
"epoch": 31.997254521963825,
"grad_norm": 0.5165979862213135,
"learning_rate": 0.00012013329742830212,
"loss": 2.9058,
"step": 118900
},
{
"epoch": 32.01049741602067,
"grad_norm": 0.4943908751010895,
"learning_rate": 0.00011993133162784434,
"loss": 2.8516,
"step": 118950
},
{
"epoch": 32.0239556416882,
"grad_norm": 0.5239517688751221,
"learning_rate": 0.00011972936582738655,
"loss": 2.8517,
"step": 119000
},
{
"epoch": 32.0239556416882,
"eval_accuracy": 0.3977709141937556,
"eval_loss": 3.286245346069336,
"eval_runtime": 184.0407,
"eval_samples_per_second": 97.87,
"eval_steps_per_second": 6.118,
"step": 119000
},
{
"epoch": 32.03741386735573,
"grad_norm": 0.5282511115074158,
"learning_rate": 0.00011952740002692876,
"loss": 2.8517,
"step": 119050
},
{
"epoch": 32.05087209302326,
"grad_norm": 0.525146484375,
"learning_rate": 0.00011932543422647097,
"loss": 2.8526,
"step": 119100
},
{
"epoch": 32.064330318690786,
"grad_norm": 0.5040988326072693,
"learning_rate": 0.00011912346842601319,
"loss": 2.8471,
"step": 119150
},
{
"epoch": 32.07778854435831,
"grad_norm": 0.4914143681526184,
"learning_rate": 0.0001189215026255554,
"loss": 2.8566,
"step": 119200
},
{
"epoch": 32.09124677002584,
"grad_norm": 0.5200188755989075,
"learning_rate": 0.00011871953682509761,
"loss": 2.85,
"step": 119250
},
{
"epoch": 32.10470499569337,
"grad_norm": 0.5235844850540161,
"learning_rate": 0.00011851757102463981,
"loss": 2.855,
"step": 119300
},
{
"epoch": 32.118163221360895,
"grad_norm": 0.4853653013706207,
"learning_rate": 0.00011831560522418203,
"loss": 2.8606,
"step": 119350
},
{
"epoch": 32.13162144702842,
"grad_norm": 0.4894345700740814,
"learning_rate": 0.00011811363942372424,
"loss": 2.861,
"step": 119400
},
{
"epoch": 32.14507967269595,
"grad_norm": 0.4907407760620117,
"learning_rate": 0.00011791167362326645,
"loss": 2.8569,
"step": 119450
},
{
"epoch": 32.15853789836348,
"grad_norm": 0.5211261510848999,
"learning_rate": 0.00011770970782280866,
"loss": 2.8519,
"step": 119500
},
{
"epoch": 32.17199612403101,
"grad_norm": 0.5244186520576477,
"learning_rate": 0.00011750774202235087,
"loss": 2.8627,
"step": 119550
},
{
"epoch": 32.18545434969854,
"grad_norm": 0.4988206624984741,
"learning_rate": 0.00011730577622189309,
"loss": 2.8713,
"step": 119600
},
{
"epoch": 32.198912575366066,
"grad_norm": 0.491857647895813,
"learning_rate": 0.00011710381042143529,
"loss": 2.8714,
"step": 119650
},
{
"epoch": 32.21237080103359,
"grad_norm": 0.4960365891456604,
"learning_rate": 0.0001169018446209775,
"loss": 2.8707,
"step": 119700
},
{
"epoch": 32.22582902670112,
"grad_norm": 0.5037326812744141,
"learning_rate": 0.00011669987882051971,
"loss": 2.872,
"step": 119750
},
{
"epoch": 32.23928725236865,
"grad_norm": 0.5272979736328125,
"learning_rate": 0.00011649791302006193,
"loss": 2.8718,
"step": 119800
},
{
"epoch": 32.252745478036175,
"grad_norm": 0.4927959144115448,
"learning_rate": 0.00011629594721960414,
"loss": 2.8644,
"step": 119850
},
{
"epoch": 32.2662037037037,
"grad_norm": 0.5483260750770569,
"learning_rate": 0.00011609398141914635,
"loss": 2.873,
"step": 119900
},
{
"epoch": 32.27966192937123,
"grad_norm": 0.5078279376029968,
"learning_rate": 0.00011589201561868856,
"loss": 2.8625,
"step": 119950
},
{
"epoch": 32.29312015503876,
"grad_norm": 0.5347286462783813,
"learning_rate": 0.00011569004981823076,
"loss": 2.8663,
"step": 120000
},
{
"epoch": 32.29312015503876,
"eval_accuracy": 0.3981518300693645,
"eval_loss": 3.2857956886291504,
"eval_runtime": 183.7799,
"eval_samples_per_second": 98.009,
"eval_steps_per_second": 6.127,
"step": 120000
},
{
"epoch": 32.29312015503876,
"step": 120000,
"total_flos": 2.507889284481024e+18,
"train_loss": 0.9776502927144368,
"train_runtime": 76851.2629,
"train_samples_per_second": 154.697,
"train_steps_per_second": 1.934
}
],
"logging_steps": 50,
"max_steps": 148640,
"num_input_tokens_seen": 0,
"num_train_epochs": 40,
"save_steps": 10000,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 20,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 20
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.507889284481024e+18,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}
{
"best_global_step": 100000,
"best_metric": 3.2741730213165283,
"best_model_checkpoint": "/scratch/cl5625/exceptions/models/high_2000_8397/checkpoint-100000",
"epoch": 32.29312015503876,
"eval_steps": 1000,
"global_step": 120000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.013458225667527994,
"grad_norm": 2.913999080657959,
"learning_rate": 0.000294,
"loss": 8.4436,
"step": 50
},
{
"epoch": 0.026916451335055987,
"grad_norm": 2.57014799118042,
"learning_rate": 0.0005939999999999999,
"loss": 6.7256,
"step": 100
},
{
"epoch": 0.04037467700258398,
"grad_norm": 1.4274643659591675,
"learning_rate": 0.0005998020735155513,
"loss": 6.2787,
"step": 150
},
{
"epoch": 0.053832902670111975,
"grad_norm": 1.2983430624008179,
"learning_rate": 0.0005996001077150935,
"loss": 6.0173,
"step": 200
},
{
"epoch": 0.06729112833763996,
"grad_norm": 1.371605396270752,
"learning_rate": 0.0005993981419146358,
"loss": 5.8389,
"step": 250
},
{
"epoch": 0.08074935400516796,
"grad_norm": 1.4190587997436523,
"learning_rate": 0.0005991961761141779,
"loss": 5.7187,
"step": 300
},
{
"epoch": 0.09420757967269595,
"grad_norm": 1.192716360092163,
"learning_rate": 0.0005989942103137202,
"loss": 5.5788,
"step": 350
},
{
"epoch": 0.10766580534022395,
"grad_norm": 1.3442068099975586,
"learning_rate": 0.0005987922445132624,
"loss": 5.4524,
"step": 400
},
{
"epoch": 0.12112403100775193,
"grad_norm": 1.0547122955322266,
"learning_rate": 0.0005985902787128047,
"loss": 5.327,
"step": 450
},
{
"epoch": 0.13458225667527993,
"grad_norm": 0.9016146063804626,
"learning_rate": 0.0005983883129123468,
"loss": 5.2192,
"step": 500
},
{
"epoch": 0.14804048234280792,
"grad_norm": 1.7201956510543823,
"learning_rate": 0.000598186347111889,
"loss": 5.1144,
"step": 550
},
{
"epoch": 0.16149870801033592,
"grad_norm": 1.1931437253952026,
"learning_rate": 0.0005979843813114312,
"loss": 5.0328,
"step": 600
},
{
"epoch": 0.1749569336778639,
"grad_norm": 1.468700647354126,
"learning_rate": 0.0005977824155109734,
"loss": 4.9485,