Files
100M_low_1000_6910/checkpoint-10000/trainer_state.json
ModelHub XC 76e92de654 初始化项目,由ModelHub XC社区提供模型
Model: craa/100M_low_1000_6910
Source: Original Platform
2026-08-13 19:27:39 +08:00

1524 lines
37 KiB
JSON

{
"best_metric": 3.759232521057129,
"best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_low_1000_6910/checkpoint-10000",
"epoch": 1.0781671159029649,
"eval_steps": 1000,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.005390835579514825,
"grad_norm": 4.540576457977295,
"learning_rate": 0.000276,
"loss": 9.0234,
"step": 50
},
{
"epoch": 0.01078167115902965,
"grad_norm": 1.3573050498962402,
"learning_rate": 0.0005759999999999999,
"loss": 6.9351,
"step": 100
},
{
"epoch": 0.016172506738544475,
"grad_norm": 1.575430989265442,
"learning_rate": 0.000599702104695089,
"loss": 6.533,
"step": 150
},
{
"epoch": 0.0215633423180593,
"grad_norm": 1.919303297996521,
"learning_rate": 0.0005993783054506205,
"loss": 6.2466,
"step": 200
},
{
"epoch": 0.026954177897574125,
"grad_norm": 1.5167046785354614,
"learning_rate": 0.0005990545062061521,
"loss": 6.0902,
"step": 250
},
{
"epoch": 0.03234501347708895,
"grad_norm": 2.198723793029785,
"learning_rate": 0.0005987307069616836,
"loss": 5.9584,
"step": 300
},
{
"epoch": 0.03773584905660377,
"grad_norm": 1.1930028200149536,
"learning_rate": 0.0005984069077172153,
"loss": 5.8824,
"step": 350
},
{
"epoch": 0.0431266846361186,
"grad_norm": 1.5988584756851196,
"learning_rate": 0.0005980831084727469,
"loss": 5.8242,
"step": 400
},
{
"epoch": 0.04851752021563342,
"grad_norm": 1.7518678903579712,
"learning_rate": 0.0005977593092282784,
"loss": 5.7575,
"step": 450
},
{
"epoch": 0.05390835579514825,
"grad_norm": 1.542738914489746,
"learning_rate": 0.00059743550998381,
"loss": 5.65,
"step": 500
},
{
"epoch": 0.05929919137466307,
"grad_norm": 1.1888893842697144,
"learning_rate": 0.0005971117107393416,
"loss": 5.6048,
"step": 550
},
{
"epoch": 0.0646900269541779,
"grad_norm": 1.1427416801452637,
"learning_rate": 0.0005967879114948732,
"loss": 5.5294,
"step": 600
},
{
"epoch": 0.07008086253369272,
"grad_norm": 1.5147260427474976,
"learning_rate": 0.0005964641122504047,
"loss": 5.4641,
"step": 650
},
{
"epoch": 0.07547169811320754,
"grad_norm": 1.2342205047607422,
"learning_rate": 0.0005961403130059363,
"loss": 5.3932,
"step": 700
},
{
"epoch": 0.08086253369272237,
"grad_norm": 0.8510802388191223,
"learning_rate": 0.0005958165137614678,
"loss": 5.3301,
"step": 750
},
{
"epoch": 0.0862533692722372,
"grad_norm": 1.226401686668396,
"learning_rate": 0.0005954927145169995,
"loss": 5.2543,
"step": 800
},
{
"epoch": 0.09164420485175202,
"grad_norm": 1.0827720165252686,
"learning_rate": 0.0005951689152725309,
"loss": 5.2325,
"step": 850
},
{
"epoch": 0.09703504043126684,
"grad_norm": 1.071048378944397,
"learning_rate": 0.0005948451160280626,
"loss": 5.1791,
"step": 900
},
{
"epoch": 0.10242587601078167,
"grad_norm": 0.8528303503990173,
"learning_rate": 0.0005945213167835941,
"loss": 5.1133,
"step": 950
},
{
"epoch": 0.1078167115902965,
"grad_norm": 1.1848195791244507,
"learning_rate": 0.0005941975175391257,
"loss": 5.1184,
"step": 1000
},
{
"epoch": 0.1078167115902965,
"eval_accuracy": 0.22659494561977292,
"eval_loss": 5.030709743499756,
"eval_runtime": 186.5313,
"eval_samples_per_second": 96.558,
"eval_steps_per_second": 6.037,
"step": 1000
},
{
"epoch": 0.11320754716981132,
"grad_norm": 1.076619029045105,
"learning_rate": 0.0005938737182946572,
"loss": 5.052,
"step": 1050
},
{
"epoch": 0.11859838274932614,
"grad_norm": 1.5022995471954346,
"learning_rate": 0.0005935499190501888,
"loss": 5.0099,
"step": 1100
},
{
"epoch": 0.12398921832884097,
"grad_norm": 0.9214362502098083,
"learning_rate": 0.0005932261198057204,
"loss": 5.017,
"step": 1150
},
{
"epoch": 0.1293800539083558,
"grad_norm": 1.1205987930297852,
"learning_rate": 0.000592902320561252,
"loss": 4.9733,
"step": 1200
},
{
"epoch": 0.1347708894878706,
"grad_norm": 1.3065948486328125,
"learning_rate": 0.0005925785213167835,
"loss": 4.9494,
"step": 1250
},
{
"epoch": 0.14016172506738545,
"grad_norm": 1.0144490003585815,
"learning_rate": 0.0005922547220723151,
"loss": 4.9026,
"step": 1300
},
{
"epoch": 0.14555256064690028,
"grad_norm": 0.8630505204200745,
"learning_rate": 0.0005919309228278468,
"loss": 4.8527,
"step": 1350
},
{
"epoch": 0.1509433962264151,
"grad_norm": 0.9217053651809692,
"learning_rate": 0.0005916071235833783,
"loss": 4.8597,
"step": 1400
},
{
"epoch": 0.15633423180592992,
"grad_norm": 1.303423523902893,
"learning_rate": 0.0005912833243389097,
"loss": 4.816,
"step": 1450
},
{
"epoch": 0.16172506738544473,
"grad_norm": 0.9101192951202393,
"learning_rate": 0.0005909595250944414,
"loss": 4.8184,
"step": 1500
},
{
"epoch": 0.16711590296495957,
"grad_norm": 1.0290004014968872,
"learning_rate": 0.000590635725849973,
"loss": 4.7953,
"step": 1550
},
{
"epoch": 0.1725067385444744,
"grad_norm": 0.9769102334976196,
"learning_rate": 0.0005903119266055045,
"loss": 4.7756,
"step": 1600
},
{
"epoch": 0.1778975741239892,
"grad_norm": 1.380980134010315,
"learning_rate": 0.0005899881273610361,
"loss": 4.7469,
"step": 1650
},
{
"epoch": 0.18328840970350405,
"grad_norm": 1.0679956674575806,
"learning_rate": 0.0005896643281165677,
"loss": 4.728,
"step": 1700
},
{
"epoch": 0.18867924528301888,
"grad_norm": 0.9642156362533569,
"learning_rate": 0.0005893405288720993,
"loss": 4.6778,
"step": 1750
},
{
"epoch": 0.1940700808625337,
"grad_norm": 0.9266220927238464,
"learning_rate": 0.0005890167296276308,
"loss": 4.7051,
"step": 1800
},
{
"epoch": 0.19946091644204852,
"grad_norm": 1.5473623275756836,
"learning_rate": 0.0005886929303831624,
"loss": 4.6514,
"step": 1850
},
{
"epoch": 0.20485175202156333,
"grad_norm": 0.8920683860778809,
"learning_rate": 0.0005883691311386939,
"loss": 4.6281,
"step": 1900
},
{
"epoch": 0.21024258760107817,
"grad_norm": 0.6964493989944458,
"learning_rate": 0.0005880453318942256,
"loss": 4.6063,
"step": 1950
},
{
"epoch": 0.215633423180593,
"grad_norm": 0.9296350479125977,
"learning_rate": 0.0005877215326497571,
"loss": 4.5882,
"step": 2000
},
{
"epoch": 0.215633423180593,
"eval_accuracy": 0.2696204026654292,
"eval_loss": 4.511911392211914,
"eval_runtime": 186.0484,
"eval_samples_per_second": 96.808,
"eval_steps_per_second": 6.052,
"step": 2000
},
{
"epoch": 0.2210242587601078,
"grad_norm": 0.8285171389579773,
"learning_rate": 0.0005873977334052887,
"loss": 4.5739,
"step": 2050
},
{
"epoch": 0.22641509433962265,
"grad_norm": 1.025179386138916,
"learning_rate": 0.0005870739341608202,
"loss": 4.5558,
"step": 2100
},
{
"epoch": 0.23180592991913745,
"grad_norm": 0.9890318512916565,
"learning_rate": 0.0005867501349163519,
"loss": 4.53,
"step": 2150
},
{
"epoch": 0.2371967654986523,
"grad_norm": 1.0426075458526611,
"learning_rate": 0.0005864263356718833,
"loss": 4.5316,
"step": 2200
},
{
"epoch": 0.24258760107816713,
"grad_norm": 0.7025043964385986,
"learning_rate": 0.000586102536427415,
"loss": 4.5041,
"step": 2250
},
{
"epoch": 0.24797843665768193,
"grad_norm": 0.8847193717956543,
"learning_rate": 0.0005857787371829465,
"loss": 4.4971,
"step": 2300
},
{
"epoch": 0.25336927223719674,
"grad_norm": 0.9764680862426758,
"learning_rate": 0.0005854549379384781,
"loss": 4.4769,
"step": 2350
},
{
"epoch": 0.2587601078167116,
"grad_norm": 0.9064006805419922,
"learning_rate": 0.0005851311386940096,
"loss": 4.4604,
"step": 2400
},
{
"epoch": 0.2641509433962264,
"grad_norm": 0.8132376074790955,
"learning_rate": 0.0005848073394495412,
"loss": 4.4506,
"step": 2450
},
{
"epoch": 0.2695417789757412,
"grad_norm": 0.7548365592956543,
"learning_rate": 0.0005844835402050728,
"loss": 4.4258,
"step": 2500
},
{
"epoch": 0.2749326145552561,
"grad_norm": 0.824798047542572,
"learning_rate": 0.0005841597409606044,
"loss": 4.3996,
"step": 2550
},
{
"epoch": 0.2803234501347709,
"grad_norm": 0.9818605184555054,
"learning_rate": 0.000583835941716136,
"loss": 4.4099,
"step": 2600
},
{
"epoch": 0.2857142857142857,
"grad_norm": 0.8206386566162109,
"learning_rate": 0.0005835121424716675,
"loss": 4.3904,
"step": 2650
},
{
"epoch": 0.29110512129380056,
"grad_norm": 0.8314753770828247,
"learning_rate": 0.0005831883432271992,
"loss": 4.3802,
"step": 2700
},
{
"epoch": 0.29649595687331537,
"grad_norm": 0.8359651565551758,
"learning_rate": 0.0005828645439827307,
"loss": 4.3767,
"step": 2750
},
{
"epoch": 0.3018867924528302,
"grad_norm": 0.673989474773407,
"learning_rate": 0.0005825407447382622,
"loss": 4.3286,
"step": 2800
},
{
"epoch": 0.30727762803234504,
"grad_norm": 0.7831907868385315,
"learning_rate": 0.0005822169454937938,
"loss": 4.3547,
"step": 2850
},
{
"epoch": 0.31266846361185985,
"grad_norm": 0.8069139719009399,
"learning_rate": 0.0005818931462493254,
"loss": 4.3245,
"step": 2900
},
{
"epoch": 0.31805929919137466,
"grad_norm": 0.8398146033287048,
"learning_rate": 0.0005815693470048569,
"loss": 4.3271,
"step": 2950
},
{
"epoch": 0.32345013477088946,
"grad_norm": 0.7504966855049133,
"learning_rate": 0.0005812455477603885,
"loss": 4.3179,
"step": 3000
},
{
"epoch": 0.32345013477088946,
"eval_accuracy": 0.29801085898691393,
"eval_loss": 4.238383769989014,
"eval_runtime": 185.9371,
"eval_samples_per_second": 96.866,
"eval_steps_per_second": 6.056,
"step": 3000
},
{
"epoch": 0.3288409703504043,
"grad_norm": 0.7269638180732727,
"learning_rate": 0.0005809217485159201,
"loss": 4.3182,
"step": 3050
},
{
"epoch": 0.33423180592991913,
"grad_norm": 0.826326847076416,
"learning_rate": 0.0005805979492714517,
"loss": 4.2988,
"step": 3100
},
{
"epoch": 0.33962264150943394,
"grad_norm": 0.9023220539093018,
"learning_rate": 0.0005802741500269832,
"loss": 4.2764,
"step": 3150
},
{
"epoch": 0.3450134770889488,
"grad_norm": 0.7048584222793579,
"learning_rate": 0.0005799503507825148,
"loss": 4.2794,
"step": 3200
},
{
"epoch": 0.3504043126684636,
"grad_norm": 0.8540084362030029,
"learning_rate": 0.0005796265515380463,
"loss": 4.2674,
"step": 3250
},
{
"epoch": 0.3557951482479784,
"grad_norm": 0.9496080875396729,
"learning_rate": 0.000579302752293578,
"loss": 4.276,
"step": 3300
},
{
"epoch": 0.3611859838274933,
"grad_norm": 0.8834256529808044,
"learning_rate": 0.0005789789530491095,
"loss": 4.251,
"step": 3350
},
{
"epoch": 0.3665768194070081,
"grad_norm": 0.8843632340431213,
"learning_rate": 0.0005786551538046411,
"loss": 4.2396,
"step": 3400
},
{
"epoch": 0.3719676549865229,
"grad_norm": 0.6297584176063538,
"learning_rate": 0.0005783313545601726,
"loss": 4.2318,
"step": 3450
},
{
"epoch": 0.37735849056603776,
"grad_norm": 0.8427825570106506,
"learning_rate": 0.0005780075553157043,
"loss": 4.2304,
"step": 3500
},
{
"epoch": 0.38274932614555257,
"grad_norm": 0.6668996810913086,
"learning_rate": 0.0005776837560712357,
"loss": 4.2258,
"step": 3550
},
{
"epoch": 0.3881401617250674,
"grad_norm": 0.7428986430168152,
"learning_rate": 0.0005773599568267673,
"loss": 4.2192,
"step": 3600
},
{
"epoch": 0.3935309973045822,
"grad_norm": 0.7153974175453186,
"learning_rate": 0.0005770361575822989,
"loss": 4.205,
"step": 3650
},
{
"epoch": 0.39892183288409705,
"grad_norm": 0.7248367667198181,
"learning_rate": 0.0005767123583378305,
"loss": 4.1946,
"step": 3700
},
{
"epoch": 0.40431266846361186,
"grad_norm": 0.6670223474502563,
"learning_rate": 0.000576388559093362,
"loss": 4.1958,
"step": 3750
},
{
"epoch": 0.40970350404312667,
"grad_norm": 0.6554595232009888,
"learning_rate": 0.0005760647598488936,
"loss": 4.1611,
"step": 3800
},
{
"epoch": 0.41509433962264153,
"grad_norm": 0.6678674221038818,
"learning_rate": 0.0005757409606044253,
"loss": 4.1724,
"step": 3850
},
{
"epoch": 0.42048517520215634,
"grad_norm": 0.7696819305419922,
"learning_rate": 0.0005754171613599568,
"loss": 4.1913,
"step": 3900
},
{
"epoch": 0.42587601078167114,
"grad_norm": 1.3464781045913696,
"learning_rate": 0.0005750933621154884,
"loss": 4.1607,
"step": 3950
},
{
"epoch": 0.431266846361186,
"grad_norm": 0.7259769439697266,
"learning_rate": 0.0005747695628710199,
"loss": 4.1615,
"step": 4000
},
{
"epoch": 0.431266846361186,
"eval_accuracy": 0.3115824738980451,
"eval_loss": 4.0961432456970215,
"eval_runtime": 185.8978,
"eval_samples_per_second": 96.887,
"eval_steps_per_second": 6.057,
"step": 4000
},
{
"epoch": 0.4366576819407008,
"grad_norm": 0.7618716359138489,
"learning_rate": 0.0005744457636265515,
"loss": 4.1695,
"step": 4050
},
{
"epoch": 0.4420485175202156,
"grad_norm": 0.6841301321983337,
"learning_rate": 0.0005741219643820831,
"loss": 4.1634,
"step": 4100
},
{
"epoch": 0.4474393530997305,
"grad_norm": 0.591198205947876,
"learning_rate": 0.0005737981651376146,
"loss": 4.1427,
"step": 4150
},
{
"epoch": 0.4528301886792453,
"grad_norm": 0.6859827637672424,
"learning_rate": 0.0005734743658931462,
"loss": 4.1505,
"step": 4200
},
{
"epoch": 0.4582210242587601,
"grad_norm": 0.6032926440238953,
"learning_rate": 0.0005731505666486778,
"loss": 4.138,
"step": 4250
},
{
"epoch": 0.4636118598382749,
"grad_norm": 0.6650081872940063,
"learning_rate": 0.0005728267674042093,
"loss": 4.1412,
"step": 4300
},
{
"epoch": 0.46900269541778977,
"grad_norm": 0.8347264528274536,
"learning_rate": 0.0005725029681597409,
"loss": 4.1275,
"step": 4350
},
{
"epoch": 0.4743935309973046,
"grad_norm": 0.6658074855804443,
"learning_rate": 0.0005721791689152725,
"loss": 4.1125,
"step": 4400
},
{
"epoch": 0.4797843665768194,
"grad_norm": 0.6057379245758057,
"learning_rate": 0.0005718553696708041,
"loss": 4.1205,
"step": 4450
},
{
"epoch": 0.48517520215633425,
"grad_norm": 0.6311624050140381,
"learning_rate": 0.0005715315704263356,
"loss": 4.1149,
"step": 4500
},
{
"epoch": 0.49056603773584906,
"grad_norm": 0.7309545278549194,
"learning_rate": 0.0005712077711818672,
"loss": 4.125,
"step": 4550
},
{
"epoch": 0.49595687331536387,
"grad_norm": 0.8806862235069275,
"learning_rate": 0.0005708839719373987,
"loss": 4.1193,
"step": 4600
},
{
"epoch": 0.5013477088948787,
"grad_norm": 0.7140979170799255,
"learning_rate": 0.0005705601726929304,
"loss": 4.075,
"step": 4650
},
{
"epoch": 0.5067385444743935,
"grad_norm": 0.738926112651825,
"learning_rate": 0.0005702363734484619,
"loss": 4.0752,
"step": 4700
},
{
"epoch": 0.5121293800539084,
"grad_norm": 0.5916194319725037,
"learning_rate": 0.0005699125742039935,
"loss": 4.083,
"step": 4750
},
{
"epoch": 0.5175202156334232,
"grad_norm": 0.7595686912536621,
"learning_rate": 0.000569588774959525,
"loss": 4.0794,
"step": 4800
},
{
"epoch": 0.522911051212938,
"grad_norm": 0.6311149597167969,
"learning_rate": 0.0005692649757150567,
"loss": 4.0744,
"step": 4850
},
{
"epoch": 0.5283018867924528,
"grad_norm": 0.7534050941467285,
"learning_rate": 0.0005689411764705881,
"loss": 4.0754,
"step": 4900
},
{
"epoch": 0.5336927223719676,
"grad_norm": 0.7457841038703918,
"learning_rate": 0.0005686173772261197,
"loss": 4.0669,
"step": 4950
},
{
"epoch": 0.5390835579514824,
"grad_norm": 0.7841017842292786,
"learning_rate": 0.0005682935779816514,
"loss": 4.0664,
"step": 5000
},
{
"epoch": 0.5390835579514824,
"eval_accuracy": 0.32084295952647335,
"eval_loss": 3.998095750808716,
"eval_runtime": 185.8924,
"eval_samples_per_second": 96.889,
"eval_steps_per_second": 6.057,
"step": 5000
},
{
"epoch": 0.5444743935309974,
"grad_norm": 0.6891273856163025,
"learning_rate": 0.0005679697787371829,
"loss": 4.0733,
"step": 5050
},
{
"epoch": 0.5498652291105122,
"grad_norm": 0.5636075735092163,
"learning_rate": 0.0005676459794927145,
"loss": 4.0588,
"step": 5100
},
{
"epoch": 0.555256064690027,
"grad_norm": 0.7175034880638123,
"learning_rate": 0.000567322180248246,
"loss": 4.0517,
"step": 5150
},
{
"epoch": 0.5606469002695418,
"grad_norm": 0.7910956144332886,
"learning_rate": 0.0005669983810037777,
"loss": 4.0616,
"step": 5200
},
{
"epoch": 0.5660377358490566,
"grad_norm": 0.6039184331893921,
"learning_rate": 0.0005666745817593092,
"loss": 4.0371,
"step": 5250
},
{
"epoch": 0.5714285714285714,
"grad_norm": 0.5606033205986023,
"learning_rate": 0.0005663507825148408,
"loss": 4.039,
"step": 5300
},
{
"epoch": 0.5768194070080862,
"grad_norm": 0.7192058563232422,
"learning_rate": 0.0005660269832703723,
"loss": 4.0264,
"step": 5350
},
{
"epoch": 0.5822102425876011,
"grad_norm": 0.596314549446106,
"learning_rate": 0.0005657031840259039,
"loss": 4.0295,
"step": 5400
},
{
"epoch": 0.5876010781671159,
"grad_norm": 0.7095101475715637,
"learning_rate": 0.0005653793847814355,
"loss": 4.0314,
"step": 5450
},
{
"epoch": 0.5929919137466307,
"grad_norm": 0.688493013381958,
"learning_rate": 0.000565055585536967,
"loss": 4.0088,
"step": 5500
},
{
"epoch": 0.5983827493261455,
"grad_norm": 0.6665045619010925,
"learning_rate": 0.0005647317862924986,
"loss": 4.0332,
"step": 5550
},
{
"epoch": 0.6037735849056604,
"grad_norm": 0.6965456008911133,
"learning_rate": 0.0005644079870480302,
"loss": 3.9809,
"step": 5600
},
{
"epoch": 0.6091644204851752,
"grad_norm": 0.625882625579834,
"learning_rate": 0.0005640841878035617,
"loss": 4.0153,
"step": 5650
},
{
"epoch": 0.6145552560646901,
"grad_norm": 0.6585850715637207,
"learning_rate": 0.0005637603885590933,
"loss": 4.0181,
"step": 5700
},
{
"epoch": 0.6199460916442049,
"grad_norm": 0.6237135529518127,
"learning_rate": 0.0005634365893146248,
"loss": 4.0169,
"step": 5750
},
{
"epoch": 0.6253369272237197,
"grad_norm": 0.6779605150222778,
"learning_rate": 0.0005631127900701565,
"loss": 4.0086,
"step": 5800
},
{
"epoch": 0.6307277628032345,
"grad_norm": 0.6086890697479248,
"learning_rate": 0.000562788990825688,
"loss": 3.9814,
"step": 5850
},
{
"epoch": 0.6361185983827493,
"grad_norm": 0.6210436820983887,
"learning_rate": 0.0005624651915812196,
"loss": 3.9972,
"step": 5900
},
{
"epoch": 0.6415094339622641,
"grad_norm": 0.7014210820198059,
"learning_rate": 0.0005621413923367511,
"loss": 3.9997,
"step": 5950
},
{
"epoch": 0.6469002695417789,
"grad_norm": 0.6019284725189209,
"learning_rate": 0.0005618175930922828,
"loss": 4.0053,
"step": 6000
},
{
"epoch": 0.6469002695417789,
"eval_accuracy": 0.32752848036658616,
"eval_loss": 3.9227378368377686,
"eval_runtime": 185.8855,
"eval_samples_per_second": 96.893,
"eval_steps_per_second": 6.057,
"step": 6000
},
{
"epoch": 0.6522911051212938,
"grad_norm": 0.6057129502296448,
"learning_rate": 0.0005614937938478143,
"loss": 3.9961,
"step": 6050
},
{
"epoch": 0.6576819407008087,
"grad_norm": 0.6797792911529541,
"learning_rate": 0.0005611699946033459,
"loss": 3.9911,
"step": 6100
},
{
"epoch": 0.6630727762803235,
"grad_norm": 0.5967230200767517,
"learning_rate": 0.0005608461953588774,
"loss": 3.9835,
"step": 6150
},
{
"epoch": 0.6684636118598383,
"grad_norm": 0.6430239677429199,
"learning_rate": 0.000560522396114409,
"loss": 3.9776,
"step": 6200
},
{
"epoch": 0.6738544474393531,
"grad_norm": 0.6402201056480408,
"learning_rate": 0.0005601985968699405,
"loss": 3.9578,
"step": 6250
},
{
"epoch": 0.6792452830188679,
"grad_norm": 0.6837700009346008,
"learning_rate": 0.0005598747976254721,
"loss": 3.9743,
"step": 6300
},
{
"epoch": 0.6846361185983828,
"grad_norm": 0.5868032574653625,
"learning_rate": 0.0005595509983810038,
"loss": 3.9487,
"step": 6350
},
{
"epoch": 0.6900269541778976,
"grad_norm": 0.6631322503089905,
"learning_rate": 0.0005592271991365353,
"loss": 3.9597,
"step": 6400
},
{
"epoch": 0.6954177897574124,
"grad_norm": 0.6614171862602234,
"learning_rate": 0.0005589033998920669,
"loss": 3.976,
"step": 6450
},
{
"epoch": 0.7008086253369272,
"grad_norm": 0.6480840444564819,
"learning_rate": 0.0005585796006475984,
"loss": 3.9484,
"step": 6500
},
{
"epoch": 0.706199460916442,
"grad_norm": 0.6667515635490417,
"learning_rate": 0.0005582558014031301,
"loss": 3.9724,
"step": 6550
},
{
"epoch": 0.7115902964959568,
"grad_norm": 0.5956810116767883,
"learning_rate": 0.0005579320021586616,
"loss": 3.9531,
"step": 6600
},
{
"epoch": 0.7169811320754716,
"grad_norm": 0.7182953357696533,
"learning_rate": 0.0005576082029141932,
"loss": 3.9483,
"step": 6650
},
{
"epoch": 0.7223719676549866,
"grad_norm": 0.540218710899353,
"learning_rate": 0.0005572844036697247,
"loss": 3.9354,
"step": 6700
},
{
"epoch": 0.7277628032345014,
"grad_norm": 0.5549564957618713,
"learning_rate": 0.0005569606044252563,
"loss": 3.9416,
"step": 6750
},
{
"epoch": 0.7331536388140162,
"grad_norm": 0.955643355846405,
"learning_rate": 0.0005566368051807879,
"loss": 3.9594,
"step": 6800
},
{
"epoch": 0.738544474393531,
"grad_norm": 0.7693118453025818,
"learning_rate": 0.0005563130059363194,
"loss": 3.945,
"step": 6850
},
{
"epoch": 0.7439353099730458,
"grad_norm": 0.587993860244751,
"learning_rate": 0.000555989206691851,
"loss": 3.9435,
"step": 6900
},
{
"epoch": 0.7493261455525606,
"grad_norm": 0.5942090749740601,
"learning_rate": 0.0005556654074473826,
"loss": 3.9466,
"step": 6950
},
{
"epoch": 0.7547169811320755,
"grad_norm": 0.6959335207939148,
"learning_rate": 0.0005553416082029141,
"loss": 3.9311,
"step": 7000
},
{
"epoch": 0.7547169811320755,
"eval_accuracy": 0.3327442535932325,
"eval_loss": 3.868434190750122,
"eval_runtime": 185.8133,
"eval_samples_per_second": 96.931,
"eval_steps_per_second": 6.06,
"step": 7000
},
{
"epoch": 0.7601078167115903,
"grad_norm": 0.6941900253295898,
"learning_rate": 0.0005550178089584457,
"loss": 3.9224,
"step": 7050
},
{
"epoch": 0.7654986522911051,
"grad_norm": 0.5540989637374878,
"learning_rate": 0.0005546940097139772,
"loss": 3.9235,
"step": 7100
},
{
"epoch": 0.77088948787062,
"grad_norm": 0.6097570061683655,
"learning_rate": 0.0005543702104695089,
"loss": 3.9264,
"step": 7150
},
{
"epoch": 0.7762803234501348,
"grad_norm": 0.5658494830131531,
"learning_rate": 0.0005540464112250404,
"loss": 3.9229,
"step": 7200
},
{
"epoch": 0.7816711590296496,
"grad_norm": 0.6700815558433533,
"learning_rate": 0.000553722611980572,
"loss": 3.9357,
"step": 7250
},
{
"epoch": 0.7870619946091644,
"grad_norm": 0.5885155200958252,
"learning_rate": 0.0005533988127361035,
"loss": 3.9183,
"step": 7300
},
{
"epoch": 0.7924528301886793,
"grad_norm": 0.5735599398612976,
"learning_rate": 0.0005530750134916352,
"loss": 3.9369,
"step": 7350
},
{
"epoch": 0.7978436657681941,
"grad_norm": 0.6089330911636353,
"learning_rate": 0.0005527512142471668,
"loss": 3.9099,
"step": 7400
},
{
"epoch": 0.8032345013477089,
"grad_norm": 0.6232162714004517,
"learning_rate": 0.0005524274150026982,
"loss": 3.9112,
"step": 7450
},
{
"epoch": 0.8086253369272237,
"grad_norm": 0.606532096862793,
"learning_rate": 0.0005521036157582299,
"loss": 3.9004,
"step": 7500
},
{
"epoch": 0.8140161725067385,
"grad_norm": 0.7327796816825867,
"learning_rate": 0.0005517798165137614,
"loss": 3.9057,
"step": 7550
},
{
"epoch": 0.8194070080862533,
"grad_norm": 0.5931469202041626,
"learning_rate": 0.000551456017269293,
"loss": 3.881,
"step": 7600
},
{
"epoch": 0.8247978436657682,
"grad_norm": 0.6132405400276184,
"learning_rate": 0.0005511322180248245,
"loss": 3.8988,
"step": 7650
},
{
"epoch": 0.8301886792452831,
"grad_norm": 0.5921282172203064,
"learning_rate": 0.0005508084187803562,
"loss": 3.9097,
"step": 7700
},
{
"epoch": 0.8355795148247979,
"grad_norm": 0.6329267024993896,
"learning_rate": 0.0005504846195358877,
"loss": 3.9026,
"step": 7750
},
{
"epoch": 0.8409703504043127,
"grad_norm": 0.6082526445388794,
"learning_rate": 0.0005501608202914193,
"loss": 3.8919,
"step": 7800
},
{
"epoch": 0.8463611859838275,
"grad_norm": 0.5989628434181213,
"learning_rate": 0.0005498370210469508,
"loss": 3.895,
"step": 7850
},
{
"epoch": 0.8517520215633423,
"grad_norm": 0.592002809047699,
"learning_rate": 0.0005495132218024824,
"loss": 3.88,
"step": 7900
},
{
"epoch": 0.8571428571428571,
"grad_norm": 0.5714187026023865,
"learning_rate": 0.000549189422558014,
"loss": 3.8903,
"step": 7950
},
{
"epoch": 0.862533692722372,
"grad_norm": 0.6176041960716248,
"learning_rate": 0.0005488656233135456,
"loss": 3.8888,
"step": 8000
},
{
"epoch": 0.862533692722372,
"eval_accuracy": 0.3374015509765124,
"eval_loss": 3.820584535598755,
"eval_runtime": 185.9631,
"eval_samples_per_second": 96.853,
"eval_steps_per_second": 6.055,
"step": 8000
},
{
"epoch": 0.8679245283018868,
"grad_norm": 0.6291443109512329,
"learning_rate": 0.0005485418240690771,
"loss": 3.8699,
"step": 8050
},
{
"epoch": 0.8733153638814016,
"grad_norm": 0.6424714922904968,
"learning_rate": 0.0005482180248246087,
"loss": 3.8858,
"step": 8100
},
{
"epoch": 0.8787061994609164,
"grad_norm": 0.6014903783798218,
"learning_rate": 0.0005478942255801403,
"loss": 3.8827,
"step": 8150
},
{
"epoch": 0.8840970350404312,
"grad_norm": 0.5504648089408875,
"learning_rate": 0.0005475704263356718,
"loss": 3.8694,
"step": 8200
},
{
"epoch": 0.889487870619946,
"grad_norm": 0.5945785045623779,
"learning_rate": 0.0005472466270912034,
"loss": 3.8799,
"step": 8250
},
{
"epoch": 0.894878706199461,
"grad_norm": 0.5891640186309814,
"learning_rate": 0.000546922827846735,
"loss": 3.8689,
"step": 8300
},
{
"epoch": 0.9002695417789758,
"grad_norm": 0.5595161318778992,
"learning_rate": 0.0005465990286022665,
"loss": 3.8689,
"step": 8350
},
{
"epoch": 0.9056603773584906,
"grad_norm": 0.5921164155006409,
"learning_rate": 0.0005462752293577981,
"loss": 3.8853,
"step": 8400
},
{
"epoch": 0.9110512129380054,
"grad_norm": 0.5961281657218933,
"learning_rate": 0.0005459514301133296,
"loss": 3.8789,
"step": 8450
},
{
"epoch": 0.9164420485175202,
"grad_norm": 0.5500056147575378,
"learning_rate": 0.0005456276308688613,
"loss": 3.875,
"step": 8500
},
{
"epoch": 0.921832884097035,
"grad_norm": 0.5573350787162781,
"learning_rate": 0.0005453038316243929,
"loss": 3.868,
"step": 8550
},
{
"epoch": 0.9272237196765498,
"grad_norm": 0.5513932108879089,
"learning_rate": 0.0005449800323799244,
"loss": 3.8786,
"step": 8600
},
{
"epoch": 0.9326145552560647,
"grad_norm": 0.5722886323928833,
"learning_rate": 0.000544656233135456,
"loss": 3.8585,
"step": 8650
},
{
"epoch": 0.9380053908355795,
"grad_norm": 0.6074110269546509,
"learning_rate": 0.0005443324338909875,
"loss": 3.8613,
"step": 8700
},
{
"epoch": 0.9433962264150944,
"grad_norm": 0.5840967893600464,
"learning_rate": 0.0005440086346465192,
"loss": 3.8728,
"step": 8750
},
{
"epoch": 0.9487870619946092,
"grad_norm": 0.580897867679596,
"learning_rate": 0.0005436848354020506,
"loss": 3.8535,
"step": 8800
},
{
"epoch": 0.954177897574124,
"grad_norm": 0.5885988473892212,
"learning_rate": 0.0005433610361575823,
"loss": 3.8595,
"step": 8850
},
{
"epoch": 0.9595687331536388,
"grad_norm": 0.568071186542511,
"learning_rate": 0.0005430372369131138,
"loss": 3.8457,
"step": 8900
},
{
"epoch": 0.9649595687331537,
"grad_norm": 0.612762987613678,
"learning_rate": 0.0005427134376686454,
"loss": 3.857,
"step": 8950
},
{
"epoch": 0.9703504043126685,
"grad_norm": 0.6389222145080566,
"learning_rate": 0.0005423896384241769,
"loss": 3.8417,
"step": 9000
},
{
"epoch": 0.9703504043126685,
"eval_accuracy": 0.3407095967989121,
"eval_loss": 3.7826406955718994,
"eval_runtime": 185.8291,
"eval_samples_per_second": 96.922,
"eval_steps_per_second": 6.059,
"step": 9000
},
{
"epoch": 0.9757412398921833,
"grad_norm": 0.6433724761009216,
"learning_rate": 0.0005420658391797086,
"loss": 3.8386,
"step": 9050
},
{
"epoch": 0.9811320754716981,
"grad_norm": 0.616686999797821,
"learning_rate": 0.0005417420399352401,
"loss": 3.8486,
"step": 9100
},
{
"epoch": 0.9865229110512129,
"grad_norm": 0.6407685875892639,
"learning_rate": 0.0005414182406907717,
"loss": 3.8421,
"step": 9150
},
{
"epoch": 0.9919137466307277,
"grad_norm": 0.731094479560852,
"learning_rate": 0.0005410944414463032,
"loss": 3.8593,
"step": 9200
},
{
"epoch": 0.9973045822102425,
"grad_norm": 0.617189884185791,
"learning_rate": 0.0005407706422018348,
"loss": 3.8551,
"step": 9250
},
{
"epoch": 1.0026954177897573,
"grad_norm": 0.5850970149040222,
"learning_rate": 0.0005404468429573664,
"loss": 3.8018,
"step": 9300
},
{
"epoch": 1.0080862533692723,
"grad_norm": 0.6619550585746765,
"learning_rate": 0.000540123043712898,
"loss": 3.7742,
"step": 9350
},
{
"epoch": 1.013477088948787,
"grad_norm": 0.5740201473236084,
"learning_rate": 0.0005397992444684295,
"loss": 3.7547,
"step": 9400
},
{
"epoch": 1.0188679245283019,
"grad_norm": 0.6828320622444153,
"learning_rate": 0.0005394754452239611,
"loss": 3.7691,
"step": 9450
},
{
"epoch": 1.0242587601078168,
"grad_norm": 0.5637569427490234,
"learning_rate": 0.0005391516459794927,
"loss": 3.7799,
"step": 9500
},
{
"epoch": 1.0296495956873315,
"grad_norm": 0.5780271887779236,
"learning_rate": 0.0005388278467350242,
"loss": 3.7877,
"step": 9550
},
{
"epoch": 1.0350404312668464,
"grad_norm": 0.5565044283866882,
"learning_rate": 0.0005385040474905557,
"loss": 3.7821,
"step": 9600
},
{
"epoch": 1.0404312668463611,
"grad_norm": 0.6038862466812134,
"learning_rate": 0.0005381802482460874,
"loss": 3.7916,
"step": 9650
},
{
"epoch": 1.045822102425876,
"grad_norm": 0.5745878219604492,
"learning_rate": 0.000537856449001619,
"loss": 3.7901,
"step": 9700
},
{
"epoch": 1.0512129380053907,
"grad_norm": 0.5736984014511108,
"learning_rate": 0.0005375326497571505,
"loss": 3.763,
"step": 9750
},
{
"epoch": 1.0566037735849056,
"grad_norm": 0.6319218277931213,
"learning_rate": 0.000537208850512682,
"loss": 3.7697,
"step": 9800
},
{
"epoch": 1.0619946091644206,
"grad_norm": 0.5360982418060303,
"learning_rate": 0.0005368850512682137,
"loss": 3.7678,
"step": 9850
},
{
"epoch": 1.0673854447439353,
"grad_norm": 0.5352616906166077,
"learning_rate": 0.0005365612520237453,
"loss": 3.7801,
"step": 9900
},
{
"epoch": 1.0727762803234502,
"grad_norm": 0.6189696788787842,
"learning_rate": 0.0005362374527792768,
"loss": 3.7794,
"step": 9950
},
{
"epoch": 1.0781671159029649,
"grad_norm": 0.6562055945396423,
"learning_rate": 0.0005359136535348084,
"loss": 3.7812,
"step": 10000
},
{
"epoch": 1.0781671159029649,
"eval_accuracy": 0.34422147543885173,
"eval_loss": 3.759232521057129,
"eval_runtime": 185.6272,
"eval_samples_per_second": 97.028,
"eval_steps_per_second": 6.066,
"step": 10000
}
],
"logging_steps": 50,
"max_steps": 92750,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 10000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 8.36069179392e+16,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}