fasthypernet-9 / checkpoint-1700 /trainer_state.json
aixk's picture
Upload folder using huggingface_hub
7155359 verified
Raw
History Blame Contribute Delete
29.3 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0455367101265336,
"eval_steps": 500,
"global_step": 1700,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006153609476558594,
"grad_norm": 25468.8046875,
"learning_rate": 1.9999999999999996e-07,
"loss": 8.7082,
"step": 10
},
{
"epoch": 0.012307218953117188,
"grad_norm": 1228.4501953125,
"learning_rate": 1.2e-06,
"loss": 8.0776,
"step": 20
},
{
"epoch": 0.018460828429675783,
"grad_norm": 355.1081237792969,
"learning_rate": 2.1999999999999997e-06,
"loss": 7.2372,
"step": 30
},
{
"epoch": 0.024614437906234375,
"grad_norm": 255.58433532714844,
"learning_rate": 3.1999999999999994e-06,
"loss": 6.9725,
"step": 40
},
{
"epoch": 0.03076804738279297,
"grad_norm": 253.23995971679688,
"learning_rate": 4.2e-06,
"loss": 6.8585,
"step": 50
},
{
"epoch": 0.036921656859351566,
"grad_norm": 241.82064819335938,
"learning_rate": 5.199999999999999e-06,
"loss": 6.7367,
"step": 60
},
{
"epoch": 0.043075266335910155,
"grad_norm": 229.0261993408203,
"learning_rate": 6.199999999999999e-06,
"loss": 6.6534,
"step": 70
},
{
"epoch": 0.04922887581246875,
"grad_norm": 321.6871337890625,
"learning_rate": 7.2e-06,
"loss": 6.6305,
"step": 80
},
{
"epoch": 0.055382485289027346,
"grad_norm": 465.00543212890625,
"learning_rate": 8.2e-06,
"loss": 6.6253,
"step": 90
},
{
"epoch": 0.06153609476558594,
"grad_norm": 566.6929321289062,
"learning_rate": 9.199999999999998e-06,
"loss": 6.5724,
"step": 100
},
{
"epoch": 0.06768970424214453,
"grad_norm": 432.230712890625,
"learning_rate": 1.02e-05,
"loss": 6.5328,
"step": 110
},
{
"epoch": 0.07384331371870313,
"grad_norm": 289.2374267578125,
"learning_rate": 1.12e-05,
"loss": 6.5547,
"step": 120
},
{
"epoch": 0.07999692319526172,
"grad_norm": 595.7386474609375,
"learning_rate": 1.2199999999999998e-05,
"loss": 6.5651,
"step": 130
},
{
"epoch": 0.08615053267182031,
"grad_norm": 861.4376220703125,
"learning_rate": 1.3199999999999997e-05,
"loss": 6.5337,
"step": 140
},
{
"epoch": 0.09230414214837891,
"grad_norm": 689.42919921875,
"learning_rate": 1.4199999999999998e-05,
"loss": 6.4927,
"step": 150
},
{
"epoch": 0.0984577516249375,
"grad_norm": 380.9496765136719,
"learning_rate": 1.5199999999999998e-05,
"loss": 6.4812,
"step": 160
},
{
"epoch": 0.1046113611014961,
"grad_norm": 221.21005249023438,
"learning_rate": 1.6199999999999997e-05,
"loss": 6.4508,
"step": 170
},
{
"epoch": 0.11076497057805469,
"grad_norm": 199.30874633789062,
"learning_rate": 1.7199999999999998e-05,
"loss": 6.4642,
"step": 180
},
{
"epoch": 0.11691858005461328,
"grad_norm": 252.44459533691406,
"learning_rate": 1.82e-05,
"loss": 6.45,
"step": 190
},
{
"epoch": 0.12307218953117188,
"grad_norm": 121.32320404052734,
"learning_rate": 1.92e-05,
"loss": 6.4115,
"step": 200
},
{
"epoch": 0.12922579900773049,
"grad_norm": 207.04344177246094,
"learning_rate": 2.0199999999999996e-05,
"loss": 6.4532,
"step": 210
},
{
"epoch": 0.13537940848428906,
"grad_norm": 296.6274719238281,
"learning_rate": 2.12e-05,
"loss": 6.3985,
"step": 220
},
{
"epoch": 0.14153301796084766,
"grad_norm": 140.3836212158203,
"learning_rate": 2.2199999999999998e-05,
"loss": 6.4152,
"step": 230
},
{
"epoch": 0.14768662743740626,
"grad_norm": 168.14443969726562,
"learning_rate": 2.3199999999999998e-05,
"loss": 6.3686,
"step": 240
},
{
"epoch": 0.15384023691396484,
"grad_norm": 141.07810974121094,
"learning_rate": 2.42e-05,
"loss": 6.3809,
"step": 250
},
{
"epoch": 0.15999384639052344,
"grad_norm": 117.08052062988281,
"learning_rate": 2.52e-05,
"loss": 6.3423,
"step": 260
},
{
"epoch": 0.16614745586708204,
"grad_norm": 122.97789764404297,
"learning_rate": 2.6199999999999996e-05,
"loss": 6.3486,
"step": 270
},
{
"epoch": 0.17230106534364062,
"grad_norm": 73.85701751708984,
"learning_rate": 2.72e-05,
"loss": 6.329,
"step": 280
},
{
"epoch": 0.17845467482019922,
"grad_norm": 78.00408935546875,
"learning_rate": 2.8199999999999998e-05,
"loss": 6.3096,
"step": 290
},
{
"epoch": 0.18460828429675782,
"grad_norm": 150.3341827392578,
"learning_rate": 2.9199999999999995e-05,
"loss": 6.3063,
"step": 300
},
{
"epoch": 0.1907618937733164,
"grad_norm": 61.75702667236328,
"learning_rate": 3.02e-05,
"loss": 6.3004,
"step": 310
},
{
"epoch": 0.196915503249875,
"grad_norm": 83.836181640625,
"learning_rate": 3.119999999999999e-05,
"loss": 6.2534,
"step": 320
},
{
"epoch": 0.2030691127264336,
"grad_norm": 190.2839813232422,
"learning_rate": 3.22e-05,
"loss": 6.2603,
"step": 330
},
{
"epoch": 0.2092227222029922,
"grad_norm": 52.196563720703125,
"learning_rate": 3.3199999999999994e-05,
"loss": 6.2439,
"step": 340
},
{
"epoch": 0.21537633167955078,
"grad_norm": 65.71736907958984,
"learning_rate": 3.42e-05,
"loss": 6.227,
"step": 350
},
{
"epoch": 0.22152994115610938,
"grad_norm": 66.71468353271484,
"learning_rate": 3.5199999999999995e-05,
"loss": 6.2164,
"step": 360
},
{
"epoch": 0.22768355063266799,
"grad_norm": 37.26243209838867,
"learning_rate": 3.62e-05,
"loss": 6.212,
"step": 370
},
{
"epoch": 0.23383716010922656,
"grad_norm": 190.98501586914062,
"learning_rate": 3.7199999999999996e-05,
"loss": 6.1833,
"step": 380
},
{
"epoch": 0.23999076958578516,
"grad_norm": 51.51064682006836,
"learning_rate": 3.8199999999999993e-05,
"loss": 6.1796,
"step": 390
},
{
"epoch": 0.24614437906234377,
"grad_norm": 71.14685821533203,
"learning_rate": 3.919999999999999e-05,
"loss": 6.1614,
"step": 400
},
{
"epoch": 0.25229798853890234,
"grad_norm": 47.48910140991211,
"learning_rate": 4.02e-05,
"loss": 6.1277,
"step": 410
},
{
"epoch": 0.25845159801546097,
"grad_norm": 49.60618209838867,
"learning_rate": 4.12e-05,
"loss": 6.107,
"step": 420
},
{
"epoch": 0.26460520749201955,
"grad_norm": 38.03694534301758,
"learning_rate": 4.2199999999999996e-05,
"loss": 6.0939,
"step": 430
},
{
"epoch": 0.2707588169685781,
"grad_norm": 38.81101989746094,
"learning_rate": 4.319999999999999e-05,
"loss": 6.118,
"step": 440
},
{
"epoch": 0.27691242644513675,
"grad_norm": 45.36625671386719,
"learning_rate": 4.42e-05,
"loss": 6.0641,
"step": 450
},
{
"epoch": 0.2830660359216953,
"grad_norm": 44.09037780761719,
"learning_rate": 4.5199999999999994e-05,
"loss": 6.0417,
"step": 460
},
{
"epoch": 0.2892196453982539,
"grad_norm": 39.48929214477539,
"learning_rate": 4.62e-05,
"loss": 6.0371,
"step": 470
},
{
"epoch": 0.29537325487481253,
"grad_norm": 44.747039794921875,
"learning_rate": 4.7199999999999995e-05,
"loss": 6.0157,
"step": 480
},
{
"epoch": 0.3015268643513711,
"grad_norm": 51.372249603271484,
"learning_rate": 4.82e-05,
"loss": 6.0378,
"step": 490
},
{
"epoch": 0.3076804738279297,
"grad_norm": 46.31863021850586,
"learning_rate": 4.9199999999999997e-05,
"loss": 5.9815,
"step": 500
},
{
"epoch": 0.3138340833044883,
"grad_norm": 74.35610961914062,
"learning_rate": 5.0199999999999994e-05,
"loss": 6.0108,
"step": 510
},
{
"epoch": 0.3199876927810469,
"grad_norm": 31.986116409301758,
"learning_rate": 5.119999999999999e-05,
"loss": 5.9452,
"step": 520
},
{
"epoch": 0.32614130225760546,
"grad_norm": 62.82748794555664,
"learning_rate": 5.2199999999999995e-05,
"loss": 5.9396,
"step": 530
},
{
"epoch": 0.3322949117341641,
"grad_norm": 36.623085021972656,
"learning_rate": 5.32e-05,
"loss": 5.9078,
"step": 540
},
{
"epoch": 0.33844852121072266,
"grad_norm": 66.77005004882812,
"learning_rate": 5.4199999999999996e-05,
"loss": 5.9127,
"step": 550
},
{
"epoch": 0.34460213068728124,
"grad_norm": 42.2696533203125,
"learning_rate": 5.519999999999999e-05,
"loss": 5.8751,
"step": 560
},
{
"epoch": 0.35075574016383987,
"grad_norm": 36.03275680541992,
"learning_rate": 5.619999999999999e-05,
"loss": 5.8704,
"step": 570
},
{
"epoch": 0.35690934964039844,
"grad_norm": 34.99599838256836,
"learning_rate": 5.72e-05,
"loss": 5.8357,
"step": 580
},
{
"epoch": 0.363062959116957,
"grad_norm": 37.03520584106445,
"learning_rate": 5.82e-05,
"loss": 5.8248,
"step": 590
},
{
"epoch": 0.36921656859351565,
"grad_norm": 39.70176696777344,
"learning_rate": 5.9199999999999996e-05,
"loss": 5.782,
"step": 600
},
{
"epoch": 0.3753701780700742,
"grad_norm": 39.41380310058594,
"learning_rate": 6.019999999999999e-05,
"loss": 5.7875,
"step": 610
},
{
"epoch": 0.3815237875466328,
"grad_norm": 36.758148193359375,
"learning_rate": 6.12e-05,
"loss": 5.7705,
"step": 620
},
{
"epoch": 0.38767739702319143,
"grad_norm": 47.42232894897461,
"learning_rate": 6.22e-05,
"loss": 5.7891,
"step": 630
},
{
"epoch": 0.39383100649975,
"grad_norm": 41.998783111572266,
"learning_rate": 6.319999999999999e-05,
"loss": 5.735,
"step": 640
},
{
"epoch": 0.3999846159763086,
"grad_norm": 35.05141830444336,
"learning_rate": 6.419999999999999e-05,
"loss": 5.7239,
"step": 650
},
{
"epoch": 0.4061382254528672,
"grad_norm": 38.990840911865234,
"learning_rate": 6.519999999999999e-05,
"loss": 5.7106,
"step": 660
},
{
"epoch": 0.4122918349294258,
"grad_norm": 38.231101989746094,
"learning_rate": 6.62e-05,
"loss": 5.6812,
"step": 670
},
{
"epoch": 0.4184454444059844,
"grad_norm": 35.416473388671875,
"learning_rate": 6.72e-05,
"loss": 5.6728,
"step": 680
},
{
"epoch": 0.424599053882543,
"grad_norm": 45.61351776123047,
"learning_rate": 6.819999999999999e-05,
"loss": 5.6484,
"step": 690
},
{
"epoch": 0.43075266335910156,
"grad_norm": 41.97138595581055,
"learning_rate": 6.919999999999999e-05,
"loss": 5.6279,
"step": 700
},
{
"epoch": 0.4369062728356602,
"grad_norm": 35.231204986572266,
"learning_rate": 7.02e-05,
"loss": 5.6131,
"step": 710
},
{
"epoch": 0.44305988231221877,
"grad_norm": 43.76620864868164,
"learning_rate": 7.12e-05,
"loss": 5.576,
"step": 720
},
{
"epoch": 0.44921349178877734,
"grad_norm": 34.20133972167969,
"learning_rate": 7.219999999999999e-05,
"loss": 5.5689,
"step": 730
},
{
"epoch": 0.45536710126533597,
"grad_norm": 37.153480529785156,
"learning_rate": 7.319999999999999e-05,
"loss": 5.5499,
"step": 740
},
{
"epoch": 0.46152071074189455,
"grad_norm": 55.655784606933594,
"learning_rate": 7.419999999999999e-05,
"loss": 5.5353,
"step": 750
},
{
"epoch": 0.4676743202184531,
"grad_norm": 33.66733169555664,
"learning_rate": 7.519999999999998e-05,
"loss": 5.5018,
"step": 760
},
{
"epoch": 0.47382792969501175,
"grad_norm": 33.522857666015625,
"learning_rate": 7.62e-05,
"loss": 5.483,
"step": 770
},
{
"epoch": 0.4799815391715703,
"grad_norm": 35.31947326660156,
"learning_rate": 7.72e-05,
"loss": 5.4719,
"step": 780
},
{
"epoch": 0.4861351486481289,
"grad_norm": 34.84073257446289,
"learning_rate": 7.819999999999999e-05,
"loss": 5.4676,
"step": 790
},
{
"epoch": 0.49228875812468753,
"grad_norm": 31.39910316467285,
"learning_rate": 7.92e-05,
"loss": 5.4421,
"step": 800
},
{
"epoch": 0.4984423676012461,
"grad_norm": 33.14794158935547,
"learning_rate": 8.019999999999998e-05,
"loss": 5.3994,
"step": 810
},
{
"epoch": 0.5045959770778047,
"grad_norm": 36.15346145629883,
"learning_rate": 8.12e-05,
"loss": 5.3823,
"step": 820
},
{
"epoch": 0.5107495865543633,
"grad_norm": 32.183895111083984,
"learning_rate": 8.22e-05,
"loss": 5.3428,
"step": 830
},
{
"epoch": 0.5169031960309219,
"grad_norm": 32.388916015625,
"learning_rate": 8.319999999999999e-05,
"loss": 5.3514,
"step": 840
},
{
"epoch": 0.5230568055074805,
"grad_norm": 31.176891326904297,
"learning_rate": 8.42e-05,
"loss": 5.3256,
"step": 850
},
{
"epoch": 0.5292104149840391,
"grad_norm": 33.613014221191406,
"learning_rate": 8.519999999999998e-05,
"loss": 5.3117,
"step": 860
},
{
"epoch": 0.5353640244605977,
"grad_norm": 30.59576416015625,
"learning_rate": 8.62e-05,
"loss": 5.2776,
"step": 870
},
{
"epoch": 0.5415176339371562,
"grad_norm": 34.864505767822266,
"learning_rate": 8.719999999999999e-05,
"loss": 5.2651,
"step": 880
},
{
"epoch": 0.5476712434137149,
"grad_norm": 33.03851318359375,
"learning_rate": 8.819999999999999e-05,
"loss": 5.2467,
"step": 890
},
{
"epoch": 0.5538248528902735,
"grad_norm": 36.94324493408203,
"learning_rate": 8.92e-05,
"loss": 5.2287,
"step": 900
},
{
"epoch": 0.559978462366832,
"grad_norm": 32.03374481201172,
"learning_rate": 9.02e-05,
"loss": 5.197,
"step": 910
},
{
"epoch": 0.5661320718433906,
"grad_norm": 31.21794891357422,
"learning_rate": 9.12e-05,
"loss": 5.1565,
"step": 920
},
{
"epoch": 0.5722856813199493,
"grad_norm": 29.664560317993164,
"learning_rate": 9.219999999999999e-05,
"loss": 5.1325,
"step": 930
},
{
"epoch": 0.5784392907965078,
"grad_norm": 29.547014236450195,
"learning_rate": 9.319999999999999e-05,
"loss": 5.1322,
"step": 940
},
{
"epoch": 0.5845929002730664,
"grad_norm": 30.9229679107666,
"learning_rate": 9.419999999999999e-05,
"loss": 5.1176,
"step": 950
},
{
"epoch": 0.5907465097496251,
"grad_norm": 32.148712158203125,
"learning_rate": 9.52e-05,
"loss": 5.0631,
"step": 960
},
{
"epoch": 0.5969001192261836,
"grad_norm": 30.888355255126953,
"learning_rate": 9.62e-05,
"loss": 5.0937,
"step": 970
},
{
"epoch": 0.6030537287027422,
"grad_norm": 37.31905746459961,
"learning_rate": 9.719999999999999e-05,
"loss": 5.0572,
"step": 980
},
{
"epoch": 0.6092073381793008,
"grad_norm": 27.809349060058594,
"learning_rate": 9.819999999999999e-05,
"loss": 5.0144,
"step": 990
},
{
"epoch": 0.6153609476558594,
"grad_norm": 34.41085433959961,
"learning_rate": 9.919999999999999e-05,
"loss": 5.0256,
"step": 1000
},
{
"epoch": 0.621514557132418,
"grad_norm": 29.413028717041016,
"learning_rate": 0.0001002,
"loss": 4.9648,
"step": 1010
},
{
"epoch": 0.6276681666089766,
"grad_norm": 30.73712158203125,
"learning_rate": 0.0001012,
"loss": 4.9744,
"step": 1020
},
{
"epoch": 0.6338217760855351,
"grad_norm": 30.328628540039062,
"learning_rate": 0.00010219999999999999,
"loss": 4.9165,
"step": 1030
},
{
"epoch": 0.6399753855620938,
"grad_norm": 29.132736206054688,
"learning_rate": 0.00010319999999999999,
"loss": 4.9574,
"step": 1040
},
{
"epoch": 0.6461289950386524,
"grad_norm": 28.395519256591797,
"learning_rate": 0.00010419999999999998,
"loss": 4.8759,
"step": 1050
},
{
"epoch": 0.6522826045152109,
"grad_norm": 26.395824432373047,
"learning_rate": 0.0001052,
"loss": 4.9002,
"step": 1060
},
{
"epoch": 0.6584362139917695,
"grad_norm": 27.33197593688965,
"learning_rate": 0.00010619999999999998,
"loss": 4.8743,
"step": 1070
},
{
"epoch": 0.6645898234683282,
"grad_norm": 25.773096084594727,
"learning_rate": 0.00010719999999999999,
"loss": 4.8093,
"step": 1080
},
{
"epoch": 0.6707434329448867,
"grad_norm": 27.487308502197266,
"learning_rate": 0.0001082,
"loss": 4.8178,
"step": 1090
},
{
"epoch": 0.6768970424214453,
"grad_norm": 26.95354652404785,
"learning_rate": 0.00010919999999999998,
"loss": 4.7463,
"step": 1100
},
{
"epoch": 0.683050651898004,
"grad_norm": 30.955183029174805,
"learning_rate": 0.0001102,
"loss": 4.7438,
"step": 1110
},
{
"epoch": 0.6892042613745625,
"grad_norm": 29.22182846069336,
"learning_rate": 0.00011119999999999998,
"loss": 4.739,
"step": 1120
},
{
"epoch": 0.6953578708511211,
"grad_norm": 25.642513275146484,
"learning_rate": 0.00011219999999999999,
"loss": 4.6881,
"step": 1130
},
{
"epoch": 0.7015114803276797,
"grad_norm": 28.86591148376465,
"learning_rate": 0.0001132,
"loss": 4.6853,
"step": 1140
},
{
"epoch": 0.7076650898042383,
"grad_norm": 28.62666130065918,
"learning_rate": 0.00011419999999999998,
"loss": 4.6572,
"step": 1150
},
{
"epoch": 0.7138186992807969,
"grad_norm": 27.118507385253906,
"learning_rate": 0.0001152,
"loss": 4.6118,
"step": 1160
},
{
"epoch": 0.7199723087573555,
"grad_norm": 28.83325958251953,
"learning_rate": 0.00011619999999999998,
"loss": 4.6325,
"step": 1170
},
{
"epoch": 0.726125918233914,
"grad_norm": 27.323766708374023,
"learning_rate": 0.00011719999999999999,
"loss": 4.5598,
"step": 1180
},
{
"epoch": 0.7322795277104727,
"grad_norm": 28.03806495666504,
"learning_rate": 0.0001182,
"loss": 4.5404,
"step": 1190
},
{
"epoch": 0.7384331371870313,
"grad_norm": 30.81092071533203,
"learning_rate": 0.00011919999999999998,
"loss": 4.5604,
"step": 1200
},
{
"epoch": 0.7445867466635898,
"grad_norm": 26.44489288330078,
"learning_rate": 0.0001202,
"loss": 4.5447,
"step": 1210
},
{
"epoch": 0.7507403561401484,
"grad_norm": 27.878007888793945,
"learning_rate": 0.00012119999999999999,
"loss": 4.5259,
"step": 1220
},
{
"epoch": 0.7568939656167071,
"grad_norm": 29.378477096557617,
"learning_rate": 0.0001222,
"loss": 4.4411,
"step": 1230
},
{
"epoch": 0.7630475750932656,
"grad_norm": 25.017864227294922,
"learning_rate": 0.00012319999999999999,
"loss": 4.4634,
"step": 1240
},
{
"epoch": 0.7692011845698242,
"grad_norm": 26.963115692138672,
"learning_rate": 0.00012419999999999998,
"loss": 4.4299,
"step": 1250
},
{
"epoch": 0.7753547940463829,
"grad_norm": 28.014739990234375,
"learning_rate": 0.00012519999999999998,
"loss": 4.4148,
"step": 1260
},
{
"epoch": 0.7815084035229414,
"grad_norm": 27.028656005859375,
"learning_rate": 0.0001262,
"loss": 4.4078,
"step": 1270
},
{
"epoch": 0.7876620129995,
"grad_norm": 28.67365264892578,
"learning_rate": 0.00012719999999999997,
"loss": 4.3355,
"step": 1280
},
{
"epoch": 0.7938156224760586,
"grad_norm": 26.66254997253418,
"learning_rate": 0.0001282,
"loss": 4.3166,
"step": 1290
},
{
"epoch": 0.7999692319526172,
"grad_norm": 27.120641708374023,
"learning_rate": 0.00012919999999999997,
"loss": 4.2582,
"step": 1300
},
{
"epoch": 0.8061228414291758,
"grad_norm": 24.28302001953125,
"learning_rate": 0.0001302,
"loss": 4.2967,
"step": 1310
},
{
"epoch": 0.8122764509057344,
"grad_norm": 27.492734909057617,
"learning_rate": 0.0001312,
"loss": 4.263,
"step": 1320
},
{
"epoch": 0.8184300603822929,
"grad_norm": 22.861173629760742,
"learning_rate": 0.0001322,
"loss": 4.2387,
"step": 1330
},
{
"epoch": 0.8245836698588516,
"grad_norm": 24.373743057250977,
"learning_rate": 0.00013319999999999999,
"loss": 4.2242,
"step": 1340
},
{
"epoch": 0.8307372793354102,
"grad_norm": 28.591249465942383,
"learning_rate": 0.0001342,
"loss": 4.1666,
"step": 1350
},
{
"epoch": 0.8368908888119688,
"grad_norm": 24.015104293823242,
"learning_rate": 0.00013519999999999998,
"loss": 4.1729,
"step": 1360
},
{
"epoch": 0.8430444982885273,
"grad_norm": 28.374052047729492,
"learning_rate": 0.0001362,
"loss": 4.1558,
"step": 1370
},
{
"epoch": 0.849198107765086,
"grad_norm": 24.058826446533203,
"learning_rate": 0.00013719999999999997,
"loss": 4.086,
"step": 1380
},
{
"epoch": 0.8553517172416446,
"grad_norm": 24.95467185974121,
"learning_rate": 0.0001382,
"loss": 4.0687,
"step": 1390
},
{
"epoch": 0.8615053267182031,
"grad_norm": 26.37350082397461,
"learning_rate": 0.0001392,
"loss": 4.0863,
"step": 1400
},
{
"epoch": 0.8676589361947618,
"grad_norm": 25.355846405029297,
"learning_rate": 0.0001402,
"loss": 4.0348,
"step": 1410
},
{
"epoch": 0.8738125456713204,
"grad_norm": 24.209243774414062,
"learning_rate": 0.0001412,
"loss": 4.0416,
"step": 1420
},
{
"epoch": 0.8799661551478789,
"grad_norm": 27.465925216674805,
"learning_rate": 0.0001422,
"loss": 3.9939,
"step": 1430
},
{
"epoch": 0.8861197646244375,
"grad_norm": 25.085203170776367,
"learning_rate": 0.00014319999999999998,
"loss": 4.0192,
"step": 1440
},
{
"epoch": 0.8922733741009962,
"grad_norm": 23.539377212524414,
"learning_rate": 0.00014419999999999998,
"loss": 3.9236,
"step": 1450
},
{
"epoch": 0.8984269835775547,
"grad_norm": 26.392433166503906,
"learning_rate": 0.00014519999999999998,
"loss": 3.9303,
"step": 1460
},
{
"epoch": 0.9045805930541133,
"grad_norm": 23.23287582397461,
"learning_rate": 0.0001462,
"loss": 3.9163,
"step": 1470
},
{
"epoch": 0.9107342025306719,
"grad_norm": 24.316932678222656,
"learning_rate": 0.00014719999999999997,
"loss": 3.868,
"step": 1480
},
{
"epoch": 0.9168878120072305,
"grad_norm": 23.149797439575195,
"learning_rate": 0.0001482,
"loss": 3.8912,
"step": 1490
},
{
"epoch": 0.9230414214837891,
"grad_norm": 25.656810760498047,
"learning_rate": 0.0001492,
"loss": 3.7883,
"step": 1500
},
{
"epoch": 0.9291950309603477,
"grad_norm": 24.43971824645996,
"learning_rate": 0.0001502,
"loss": 3.8087,
"step": 1510
},
{
"epoch": 0.9353486404369062,
"grad_norm": 24.23967170715332,
"learning_rate": 0.0001512,
"loss": 3.7772,
"step": 1520
},
{
"epoch": 0.9415022499134649,
"grad_norm": 33.623390197753906,
"learning_rate": 0.00015219999999999999,
"loss": 3.7577,
"step": 1530
},
{
"epoch": 0.9476558593900235,
"grad_norm": 21.980667114257812,
"learning_rate": 0.0001532,
"loss": 3.7378,
"step": 1540
},
{
"epoch": 0.953809468866582,
"grad_norm": 22.828100204467773,
"learning_rate": 0.00015419999999999998,
"loss": 3.701,
"step": 1550
},
{
"epoch": 0.9599630783431407,
"grad_norm": 22.391563415527344,
"learning_rate": 0.00015519999999999998,
"loss": 3.6866,
"step": 1560
},
{
"epoch": 0.9661166878196993,
"grad_norm": 21.836578369140625,
"learning_rate": 0.0001562,
"loss": 3.6618,
"step": 1570
},
{
"epoch": 0.9722702972962578,
"grad_norm": 24.34965705871582,
"learning_rate": 0.0001572,
"loss": 3.6196,
"step": 1580
},
{
"epoch": 0.9784239067728164,
"grad_norm": 23.68498420715332,
"learning_rate": 0.00015819999999999997,
"loss": 3.5995,
"step": 1590
},
{
"epoch": 0.9845775162493751,
"grad_norm": 23.469165802001953,
"learning_rate": 0.00015919999999999997,
"loss": 3.602,
"step": 1600
},
{
"epoch": 0.9907311257259336,
"grad_norm": 22.707569122314453,
"learning_rate": 0.0001602,
"loss": 3.548,
"step": 1610
},
{
"epoch": 0.9968847352024922,
"grad_norm": 23.006725311279297,
"learning_rate": 0.0001612,
"loss": 3.5996,
"step": 1620
},
{
"epoch": 1.0024614437906234,
"grad_norm": 24.067211151123047,
"learning_rate": 0.00016219999999999999,
"loss": 3.5286,
"step": 1630
},
{
"epoch": 1.008615053267182,
"grad_norm": 23.17717933654785,
"learning_rate": 0.0001632,
"loss": 3.4575,
"step": 1640
},
{
"epoch": 1.0147686627437407,
"grad_norm": 21.80594825744629,
"learning_rate": 0.00016419999999999998,
"loss": 3.4402,
"step": 1650
},
{
"epoch": 1.0209222722202993,
"grad_norm": 23.56287384033203,
"learning_rate": 0.00016519999999999998,
"loss": 3.418,
"step": 1660
},
{
"epoch": 1.0270758816968577,
"grad_norm": 22.383155822753906,
"learning_rate": 0.0001662,
"loss": 3.4202,
"step": 1670
},
{
"epoch": 1.0332294911734163,
"grad_norm": 22.10567855834961,
"learning_rate": 0.0001672,
"loss": 3.3831,
"step": 1680
},
{
"epoch": 1.039383100649975,
"grad_norm": 21.35976219177246,
"learning_rate": 0.00016819999999999997,
"loss": 3.3813,
"step": 1690
},
{
"epoch": 1.0455367101265336,
"grad_norm": 22.603801727294922,
"learning_rate": 0.00016919999999999997,
"loss": 3.3525,
"step": 1700
}
],
"logging_steps": 10,
"max_steps": 100000,
"num_input_tokens_seen": 0,
"num_train_epochs": 62,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.2868101986304e+17,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}