fasthypernet-16k / checkpoint-1600 /trainer_state.json
aixk's picture
Upload folder using huggingface_hub
96be12c verified
Raw
History Blame Contribute Delete
28 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.005532120876841159,
"eval_steps": 500,
"global_step": 1600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 3.4575755480257244e-05,
"grad_norm": 38.12546920776367,
"learning_rate": 4.8e-07,
"loss": 9.3097,
"step": 10
},
{
"epoch": 6.915151096051449e-05,
"grad_norm": 31.766887664794922,
"learning_rate": 1.0799999999999998e-06,
"loss": 9.293,
"step": 20
},
{
"epoch": 0.00010372726644077173,
"grad_norm": 48.67866897583008,
"learning_rate": 1.6799999999999998e-06,
"loss": 9.2784,
"step": 30
},
{
"epoch": 0.00013830302192102898,
"grad_norm": 33.25461959838867,
"learning_rate": 2.2799999999999998e-06,
"loss": 9.2793,
"step": 40
},
{
"epoch": 0.00017287877740128623,
"grad_norm": 35.62735366821289,
"learning_rate": 2.8799999999999995e-06,
"loss": 9.2657,
"step": 50
},
{
"epoch": 0.00020745453288154345,
"grad_norm": 43.818328857421875,
"learning_rate": 3.4799999999999993e-06,
"loss": 9.2616,
"step": 60
},
{
"epoch": 0.0002420302883618007,
"grad_norm": 43.82963943481445,
"learning_rate": 4.079999999999999e-06,
"loss": 9.2633,
"step": 70
},
{
"epoch": 0.00027660604384205796,
"grad_norm": 57.38799285888672,
"learning_rate": 4.679999999999999e-06,
"loss": 9.2308,
"step": 80
},
{
"epoch": 0.0003111817993223152,
"grad_norm": 44.78755187988281,
"learning_rate": 5.28e-06,
"loss": 9.2083,
"step": 90
},
{
"epoch": 0.00034575755480257246,
"grad_norm": 40.440338134765625,
"learning_rate": 5.88e-06,
"loss": 9.2418,
"step": 100
},
{
"epoch": 0.00038033331028282965,
"grad_norm": 36.28009033203125,
"learning_rate": 6.48e-06,
"loss": 9.2208,
"step": 110
},
{
"epoch": 0.0004149090657630869,
"grad_norm": 42.1943473815918,
"learning_rate": 7.079999999999999e-06,
"loss": 9.2197,
"step": 120
},
{
"epoch": 0.00044948482124334416,
"grad_norm": 37.897483825683594,
"learning_rate": 7.68e-06,
"loss": 9.2271,
"step": 130
},
{
"epoch": 0.0004840605767236014,
"grad_norm": 33.77066421508789,
"learning_rate": 8.279999999999999e-06,
"loss": 9.2311,
"step": 140
},
{
"epoch": 0.0005186363322038587,
"grad_norm": 34.798553466796875,
"learning_rate": 8.88e-06,
"loss": 9.2079,
"step": 150
},
{
"epoch": 0.0005532120876841159,
"grad_norm": 39.236812591552734,
"learning_rate": 9.48e-06,
"loss": 9.2186,
"step": 160
},
{
"epoch": 0.0005877878431643732,
"grad_norm": 37.878868103027344,
"learning_rate": 1.0079999999999998e-05,
"loss": 9.2176,
"step": 170
},
{
"epoch": 0.0006223635986446304,
"grad_norm": 64.34676361083984,
"learning_rate": 1.068e-05,
"loss": 9.2086,
"step": 180
},
{
"epoch": 0.0006569393541248877,
"grad_norm": 35.60507583618164,
"learning_rate": 1.1279999999999999e-05,
"loss": 9.2045,
"step": 190
},
{
"epoch": 0.0006915151096051449,
"grad_norm": 45.516971588134766,
"learning_rate": 1.188e-05,
"loss": 9.2144,
"step": 200
},
{
"epoch": 0.0007260908650854022,
"grad_norm": 33.20392608642578,
"learning_rate": 1.2479999999999999e-05,
"loss": 9.1846,
"step": 210
},
{
"epoch": 0.0007606666205656593,
"grad_norm": 33.04153060913086,
"learning_rate": 1.3019999999999999e-05,
"loss": 9.1875,
"step": 220
},
{
"epoch": 0.0007952423760459166,
"grad_norm": 29.512754440307617,
"learning_rate": 1.362e-05,
"loss": 9.1849,
"step": 230
},
{
"epoch": 0.0008298181315261738,
"grad_norm": 24.111255645751953,
"learning_rate": 1.4219999999999998e-05,
"loss": 9.2074,
"step": 240
},
{
"epoch": 0.0008643938870064311,
"grad_norm": 36.84242630004883,
"learning_rate": 1.4819999999999999e-05,
"loss": 9.1818,
"step": 250
},
{
"epoch": 0.0008989696424866883,
"grad_norm": 22.629037857055664,
"learning_rate": 1.5419999999999998e-05,
"loss": 9.1991,
"step": 260
},
{
"epoch": 0.0009335453979669456,
"grad_norm": 28.257951736450195,
"learning_rate": 1.602e-05,
"loss": 9.1579,
"step": 270
},
{
"epoch": 0.0009681211534472028,
"grad_norm": 26.29328155517578,
"learning_rate": 1.6619999999999997e-05,
"loss": 9.1753,
"step": 280
},
{
"epoch": 0.00100269690892746,
"grad_norm": 30.55463981628418,
"learning_rate": 1.7219999999999998e-05,
"loss": 9.1663,
"step": 290
},
{
"epoch": 0.0010372726644077173,
"grad_norm": 30.852642059326172,
"learning_rate": 1.782e-05,
"loss": 9.1624,
"step": 300
},
{
"epoch": 0.0010718484198879745,
"grad_norm": 30.8385009765625,
"learning_rate": 1.842e-05,
"loss": 9.1754,
"step": 310
},
{
"epoch": 0.0011064241753682318,
"grad_norm": 16.846324920654297,
"learning_rate": 1.9019999999999997e-05,
"loss": 9.1774,
"step": 320
},
{
"epoch": 0.001140999930848489,
"grad_norm": 26.020891189575195,
"learning_rate": 1.962e-05,
"loss": 9.1461,
"step": 330
},
{
"epoch": 0.0011755756863287463,
"grad_norm": 24.087093353271484,
"learning_rate": 2.022e-05,
"loss": 9.1698,
"step": 340
},
{
"epoch": 0.0012101514418090035,
"grad_norm": 22.612030029296875,
"learning_rate": 2.082e-05,
"loss": 9.1564,
"step": 350
},
{
"epoch": 0.0012447271972892608,
"grad_norm": 24.23309898376465,
"learning_rate": 2.1419999999999998e-05,
"loss": 9.1407,
"step": 360
},
{
"epoch": 0.001279302952769518,
"grad_norm": 19.798912048339844,
"learning_rate": 2.202e-05,
"loss": 9.1916,
"step": 370
},
{
"epoch": 0.0013138787082497753,
"grad_norm": 17.621305465698242,
"learning_rate": 2.2619999999999997e-05,
"loss": 9.1724,
"step": 380
},
{
"epoch": 0.0013484544637300325,
"grad_norm": 19.028974533081055,
"learning_rate": 2.3219999999999998e-05,
"loss": 9.1416,
"step": 390
},
{
"epoch": 0.0013830302192102898,
"grad_norm": 15.816458702087402,
"learning_rate": 2.382e-05,
"loss": 9.1704,
"step": 400
},
{
"epoch": 0.001417605974690547,
"grad_norm": 19.060916900634766,
"learning_rate": 2.4419999999999997e-05,
"loss": 9.1527,
"step": 410
},
{
"epoch": 0.0014521817301708043,
"grad_norm": 23.854291915893555,
"learning_rate": 2.5019999999999998e-05,
"loss": 9.146,
"step": 420
},
{
"epoch": 0.0014867574856510615,
"grad_norm": 20.823823928833008,
"learning_rate": 2.562e-05,
"loss": 9.1528,
"step": 430
},
{
"epoch": 0.0015213332411313186,
"grad_norm": 20.012248992919922,
"learning_rate": 2.622e-05,
"loss": 9.133,
"step": 440
},
{
"epoch": 0.001555908996611576,
"grad_norm": 17.968521118164062,
"learning_rate": 2.6819999999999997e-05,
"loss": 9.1451,
"step": 450
},
{
"epoch": 0.0015904847520918331,
"grad_norm": 19.263395309448242,
"learning_rate": 2.7419999999999995e-05,
"loss": 9.1414,
"step": 460
},
{
"epoch": 0.0016250605075720905,
"grad_norm": 19.886655807495117,
"learning_rate": 2.8019999999999996e-05,
"loss": 9.1132,
"step": 470
},
{
"epoch": 0.0016596362630523476,
"grad_norm": 20.179471969604492,
"learning_rate": 2.8619999999999997e-05,
"loss": 9.1274,
"step": 480
},
{
"epoch": 0.001694212018532605,
"grad_norm": 16.05131721496582,
"learning_rate": 2.9219999999999998e-05,
"loss": 9.1407,
"step": 490
},
{
"epoch": 0.0017287877740128621,
"grad_norm": 16.787809371948242,
"learning_rate": 2.982e-05,
"loss": 9.1274,
"step": 500
},
{
"epoch": 0.0017633635294931195,
"grad_norm": 14.712662696838379,
"learning_rate": 3.042e-05,
"loss": 9.1694,
"step": 510
},
{
"epoch": 0.0017979392849733766,
"grad_norm": 14.417969703674316,
"learning_rate": 3.102e-05,
"loss": 9.1474,
"step": 520
},
{
"epoch": 0.001832515040453634,
"grad_norm": 29.00182342529297,
"learning_rate": 3.161999999999999e-05,
"loss": 9.1549,
"step": 530
},
{
"epoch": 0.0018670907959338911,
"grad_norm": 16.824987411499023,
"learning_rate": 3.221999999999999e-05,
"loss": 9.1384,
"step": 540
},
{
"epoch": 0.0019016665514141485,
"grad_norm": 18.04323387145996,
"learning_rate": 3.2819999999999994e-05,
"loss": 9.1522,
"step": 550
},
{
"epoch": 0.0019362423068944056,
"grad_norm": 14.293987274169922,
"learning_rate": 3.3419999999999995e-05,
"loss": 9.1301,
"step": 560
},
{
"epoch": 0.0019708180623746628,
"grad_norm": 19.81171226501465,
"learning_rate": 3.4019999999999996e-05,
"loss": 9.1233,
"step": 570
},
{
"epoch": 0.00200539381785492,
"grad_norm": 14.640559196472168,
"learning_rate": 3.462e-05,
"loss": 9.1311,
"step": 580
},
{
"epoch": 0.0020399695733351775,
"grad_norm": 18.38323974609375,
"learning_rate": 3.522e-05,
"loss": 9.1222,
"step": 590
},
{
"epoch": 0.0020745453288154346,
"grad_norm": 17.136762619018555,
"learning_rate": 3.582e-05,
"loss": 9.1199,
"step": 600
},
{
"epoch": 0.0021091210842956918,
"grad_norm": 15.868051528930664,
"learning_rate": 3.6419999999999994e-05,
"loss": 9.1615,
"step": 610
},
{
"epoch": 0.002143696839775949,
"grad_norm": 12.776758193969727,
"learning_rate": 3.7019999999999995e-05,
"loss": 9.1367,
"step": 620
},
{
"epoch": 0.0021782725952562065,
"grad_norm": 24.72905731201172,
"learning_rate": 3.762e-05,
"loss": 9.0972,
"step": 630
},
{
"epoch": 0.0022128483507364636,
"grad_norm": 10.394266128540039,
"learning_rate": 3.822e-05,
"loss": 9.1424,
"step": 640
},
{
"epoch": 0.002247424106216721,
"grad_norm": 23.022293090820312,
"learning_rate": 3.881999999999999e-05,
"loss": 9.1513,
"step": 650
},
{
"epoch": 0.002281999861696978,
"grad_norm": 13.62468147277832,
"learning_rate": 3.941999999999999e-05,
"loss": 9.1056,
"step": 660
},
{
"epoch": 0.0023165756171772355,
"grad_norm": 10.622007369995117,
"learning_rate": 4.001999999999999e-05,
"loss": 9.1389,
"step": 670
},
{
"epoch": 0.0023511513726574926,
"grad_norm": 12.142419815063477,
"learning_rate": 4.0619999999999994e-05,
"loss": 9.1279,
"step": 680
},
{
"epoch": 0.00238572712813775,
"grad_norm": 11.485246658325195,
"learning_rate": 4.1219999999999995e-05,
"loss": 9.1175,
"step": 690
},
{
"epoch": 0.002420302883618007,
"grad_norm": 10.780238151550293,
"learning_rate": 4.1819999999999996e-05,
"loss": 9.1417,
"step": 700
},
{
"epoch": 0.0024548786390982645,
"grad_norm": 11.59386157989502,
"learning_rate": 4.242e-05,
"loss": 9.1246,
"step": 710
},
{
"epoch": 0.0024894543945785217,
"grad_norm": 24.624088287353516,
"learning_rate": 4.302e-05,
"loss": 9.121,
"step": 720
},
{
"epoch": 0.002524030150058779,
"grad_norm": 11.020679473876953,
"learning_rate": 4.362e-05,
"loss": 9.1379,
"step": 730
},
{
"epoch": 0.002558605905539036,
"grad_norm": 14.969200134277344,
"learning_rate": 4.422e-05,
"loss": 9.1268,
"step": 740
},
{
"epoch": 0.002593181661019293,
"grad_norm": 14.996768951416016,
"learning_rate": 4.4819999999999994e-05,
"loss": 9.1071,
"step": 750
},
{
"epoch": 0.0026277574164995507,
"grad_norm": 16.976953506469727,
"learning_rate": 4.5419999999999995e-05,
"loss": 9.1073,
"step": 760
},
{
"epoch": 0.002662333171979808,
"grad_norm": 17.404489517211914,
"learning_rate": 4.6019999999999996e-05,
"loss": 9.0887,
"step": 770
},
{
"epoch": 0.002696908927460065,
"grad_norm": 11.283190727233887,
"learning_rate": 4.662e-05,
"loss": 9.1061,
"step": 780
},
{
"epoch": 0.002731484682940322,
"grad_norm": 30.573930740356445,
"learning_rate": 4.722e-05,
"loss": 9.0974,
"step": 790
},
{
"epoch": 0.0027660604384205797,
"grad_norm": 13.684494972229004,
"learning_rate": 4.781999999999999e-05,
"loss": 9.1181,
"step": 800
},
{
"epoch": 0.002800636193900837,
"grad_norm": 8.140137672424316,
"learning_rate": 4.8419999999999994e-05,
"loss": 9.1429,
"step": 810
},
{
"epoch": 0.002835211949381094,
"grad_norm": 11.793691635131836,
"learning_rate": 4.9019999999999995e-05,
"loss": 9.1238,
"step": 820
},
{
"epoch": 0.002869787704861351,
"grad_norm": 10.155129432678223,
"learning_rate": 4.956e-05,
"loss": 9.1099,
"step": 830
},
{
"epoch": 0.0029043634603416087,
"grad_norm": 18.905136108398438,
"learning_rate": 5.0159999999999994e-05,
"loss": 9.118,
"step": 840
},
{
"epoch": 0.002938939215821866,
"grad_norm": 14.191570281982422,
"learning_rate": 5.0759999999999995e-05,
"loss": 9.0971,
"step": 850
},
{
"epoch": 0.002973514971302123,
"grad_norm": 13.144892692565918,
"learning_rate": 5.1359999999999996e-05,
"loss": 9.1168,
"step": 860
},
{
"epoch": 0.00300809072678238,
"grad_norm": 17.27440643310547,
"learning_rate": 5.195999999999999e-05,
"loss": 9.1007,
"step": 870
},
{
"epoch": 0.0030426664822626372,
"grad_norm": 14.852960586547852,
"learning_rate": 5.255999999999999e-05,
"loss": 9.1023,
"step": 880
},
{
"epoch": 0.003077242237742895,
"grad_norm": 12.116819381713867,
"learning_rate": 5.315999999999999e-05,
"loss": 9.1209,
"step": 890
},
{
"epoch": 0.003111817993223152,
"grad_norm": 13.627985000610352,
"learning_rate": 5.3759999999999994e-05,
"loss": 9.1106,
"step": 900
},
{
"epoch": 0.003146393748703409,
"grad_norm": 16.725046157836914,
"learning_rate": 5.4359999999999995e-05,
"loss": 9.1138,
"step": 910
},
{
"epoch": 0.0031809695041836662,
"grad_norm": 19.617530822753906,
"learning_rate": 5.4959999999999996e-05,
"loss": 9.1102,
"step": 920
},
{
"epoch": 0.003215545259663924,
"grad_norm": 14.019784927368164,
"learning_rate": 5.5559999999999997e-05,
"loss": 9.1069,
"step": 930
},
{
"epoch": 0.003250121015144181,
"grad_norm": 15.217008590698242,
"learning_rate": 5.616e-05,
"loss": 9.0959,
"step": 940
},
{
"epoch": 0.003284696770624438,
"grad_norm": 23.325428009033203,
"learning_rate": 5.6699999999999996e-05,
"loss": 9.0977,
"step": 950
},
{
"epoch": 0.0033192725261046952,
"grad_norm": 11.030868530273438,
"learning_rate": 5.73e-05,
"loss": 9.1116,
"step": 960
},
{
"epoch": 0.003353848281584953,
"grad_norm": 13.278918266296387,
"learning_rate": 5.79e-05,
"loss": 9.0861,
"step": 970
},
{
"epoch": 0.00338842403706521,
"grad_norm": 13.311212539672852,
"learning_rate": 5.85e-05,
"loss": 9.088,
"step": 980
},
{
"epoch": 0.003422999792545467,
"grad_norm": 10.462119102478027,
"learning_rate": 5.91e-05,
"loss": 9.0808,
"step": 990
},
{
"epoch": 0.0034575755480257243,
"grad_norm": 68.89077758789062,
"learning_rate": 5.97e-05,
"loss": 9.0981,
"step": 1000
},
{
"epoch": 0.0034921513035059814,
"grad_norm": 12.376023292541504,
"learning_rate": 6.0299999999999995e-05,
"loss": 9.1087,
"step": 1010
},
{
"epoch": 0.003526727058986239,
"grad_norm": 11.641372680664062,
"learning_rate": 6.0899999999999996e-05,
"loss": 9.1097,
"step": 1020
},
{
"epoch": 0.003561302814466496,
"grad_norm": 17.008573532104492,
"learning_rate": 6.149999999999999e-05,
"loss": 9.118,
"step": 1030
},
{
"epoch": 0.0035958785699467533,
"grad_norm": 13.044571876525879,
"learning_rate": 6.209999999999999e-05,
"loss": 9.1045,
"step": 1040
},
{
"epoch": 0.0036304543254270104,
"grad_norm": 8.668659210205078,
"learning_rate": 6.269999999999999e-05,
"loss": 9.1138,
"step": 1050
},
{
"epoch": 0.003665030080907268,
"grad_norm": 13.858012199401855,
"learning_rate": 6.33e-05,
"loss": 9.1086,
"step": 1060
},
{
"epoch": 0.003699605836387525,
"grad_norm": 9.014534950256348,
"learning_rate": 6.39e-05,
"loss": 9.125,
"step": 1070
},
{
"epoch": 0.0037341815918677823,
"grad_norm": 9.205188751220703,
"learning_rate": 6.45e-05,
"loss": 9.1078,
"step": 1080
},
{
"epoch": 0.0037687573473480394,
"grad_norm": 16.21772003173828,
"learning_rate": 6.51e-05,
"loss": 9.0993,
"step": 1090
},
{
"epoch": 0.003803333102828297,
"grad_norm": 10.28536319732666,
"learning_rate": 6.57e-05,
"loss": 9.0929,
"step": 1100
},
{
"epoch": 0.003837908858308554,
"grad_norm": 11.15615177154541,
"learning_rate": 6.63e-05,
"loss": 9.0807,
"step": 1110
},
{
"epoch": 0.0038724846137888113,
"grad_norm": 8.5701322555542,
"learning_rate": 6.69e-05,
"loss": 9.0883,
"step": 1120
},
{
"epoch": 0.003907060369269069,
"grad_norm": 10.674054145812988,
"learning_rate": 6.75e-05,
"loss": 9.0875,
"step": 1130
},
{
"epoch": 0.0039416361247493256,
"grad_norm": 9.421123504638672,
"learning_rate": 6.81e-05,
"loss": 9.0875,
"step": 1140
},
{
"epoch": 0.003976211880229583,
"grad_norm": 8.26066780090332,
"learning_rate": 6.87e-05,
"loss": 9.0797,
"step": 1150
},
{
"epoch": 0.00401078763570984,
"grad_norm": 10.132837295532227,
"learning_rate": 6.93e-05,
"loss": 9.0935,
"step": 1160
},
{
"epoch": 0.004045363391190097,
"grad_norm": 10.218791007995605,
"learning_rate": 6.989999999999999e-05,
"loss": 9.0963,
"step": 1170
},
{
"epoch": 0.004079939146670355,
"grad_norm": 13.95479679107666,
"learning_rate": 7.049999999999999e-05,
"loss": 9.0497,
"step": 1180
},
{
"epoch": 0.004114514902150612,
"grad_norm": 8.425973892211914,
"learning_rate": 7.11e-05,
"loss": 9.0762,
"step": 1190
},
{
"epoch": 0.004149090657630869,
"grad_norm": 5.935906887054443,
"learning_rate": 7.17e-05,
"loss": 9.1127,
"step": 1200
},
{
"epoch": 0.004183666413111127,
"grad_norm": 7.9404168128967285,
"learning_rate": 7.23e-05,
"loss": 9.0896,
"step": 1210
},
{
"epoch": 0.0042182421685913836,
"grad_norm": 9.45516586303711,
"learning_rate": 7.29e-05,
"loss": 9.0694,
"step": 1220
},
{
"epoch": 0.004252817924071641,
"grad_norm": 10.396097183227539,
"learning_rate": 7.35e-05,
"loss": 9.0839,
"step": 1230
},
{
"epoch": 0.004287393679551898,
"grad_norm": 12.011253356933594,
"learning_rate": 7.41e-05,
"loss": 9.0603,
"step": 1240
},
{
"epoch": 0.004321969435032155,
"grad_norm": 8.560672760009766,
"learning_rate": 7.47e-05,
"loss": 9.086,
"step": 1250
},
{
"epoch": 0.004356545190512413,
"grad_norm": 8.79507827758789,
"learning_rate": 7.529999999999999e-05,
"loss": 9.0915,
"step": 1260
},
{
"epoch": 0.00439112094599267,
"grad_norm": 9.286111831665039,
"learning_rate": 7.589999999999999e-05,
"loss": 9.0783,
"step": 1270
},
{
"epoch": 0.004425696701472927,
"grad_norm": 9.932263374328613,
"learning_rate": 7.649999999999999e-05,
"loss": 9.0808,
"step": 1280
},
{
"epoch": 0.004460272456953184,
"grad_norm": 13.201107025146484,
"learning_rate": 7.709999999999999e-05,
"loss": 9.0991,
"step": 1290
},
{
"epoch": 0.004494848212433442,
"grad_norm": 8.324228286743164,
"learning_rate": 7.769999999999999e-05,
"loss": 9.0859,
"step": 1300
},
{
"epoch": 0.004529423967913699,
"grad_norm": 11.608538627624512,
"learning_rate": 7.829999999999999e-05,
"loss": 9.0831,
"step": 1310
},
{
"epoch": 0.004563999723393956,
"grad_norm": 9.669852256774902,
"learning_rate": 7.89e-05,
"loss": 9.0847,
"step": 1320
},
{
"epoch": 0.004598575478874213,
"grad_norm": 4.503360271453857,
"learning_rate": 7.95e-05,
"loss": 9.0789,
"step": 1330
},
{
"epoch": 0.004633151234354471,
"grad_norm": 8.09322452545166,
"learning_rate": 8.01e-05,
"loss": 9.096,
"step": 1340
},
{
"epoch": 0.004667726989834728,
"grad_norm": 7.773679256439209,
"learning_rate": 8.07e-05,
"loss": 9.0822,
"step": 1350
},
{
"epoch": 0.004702302745314985,
"grad_norm": 12.035070419311523,
"learning_rate": 8.13e-05,
"loss": 9.0535,
"step": 1360
},
{
"epoch": 0.004736878500795242,
"grad_norm": 6.7790422439575195,
"learning_rate": 8.19e-05,
"loss": 9.0923,
"step": 1370
},
{
"epoch": 0.0047714542562755,
"grad_norm": 49.749053955078125,
"learning_rate": 8.25e-05,
"loss": 9.0877,
"step": 1380
},
{
"epoch": 0.004806030011755757,
"grad_norm": 8.277181625366211,
"learning_rate": 8.31e-05,
"loss": 9.0854,
"step": 1390
},
{
"epoch": 0.004840605767236014,
"grad_norm": 17.89580726623535,
"learning_rate": 8.37e-05,
"loss": 9.0793,
"step": 1400
},
{
"epoch": 0.0048751815227162714,
"grad_norm": 9.948437690734863,
"learning_rate": 8.43e-05,
"loss": 9.0526,
"step": 1410
},
{
"epoch": 0.004909757278196529,
"grad_norm": 8.40725040435791,
"learning_rate": 8.489999999999999e-05,
"loss": 9.0885,
"step": 1420
},
{
"epoch": 0.004944333033676786,
"grad_norm": 10.181068420410156,
"learning_rate": 8.549999999999999e-05,
"loss": 9.0791,
"step": 1430
},
{
"epoch": 0.004978908789157043,
"grad_norm": 8.453405380249023,
"learning_rate": 8.609999999999999e-05,
"loss": 9.0557,
"step": 1440
},
{
"epoch": 0.0050134845446373,
"grad_norm": 9.927013397216797,
"learning_rate": 8.669999999999998e-05,
"loss": 9.1001,
"step": 1450
},
{
"epoch": 0.005048060300117558,
"grad_norm": 8.138837814331055,
"learning_rate": 8.729999999999998e-05,
"loss": 9.06,
"step": 1460
},
{
"epoch": 0.005082636055597815,
"grad_norm": 8.283705711364746,
"learning_rate": 8.789999999999998e-05,
"loss": 9.0746,
"step": 1470
},
{
"epoch": 0.005117211811078072,
"grad_norm": 8.518525123596191,
"learning_rate": 8.849999999999998e-05,
"loss": 9.0558,
"step": 1480
},
{
"epoch": 0.0051517875665583295,
"grad_norm": 9.632137298583984,
"learning_rate": 8.909999999999998e-05,
"loss": 9.0739,
"step": 1490
},
{
"epoch": 0.005186363322038586,
"grad_norm": 8.111174583435059,
"learning_rate": 8.969999999999998e-05,
"loss": 9.0722,
"step": 1500
},
{
"epoch": 0.005220939077518844,
"grad_norm": 12.209863662719727,
"learning_rate": 9.029999999999999e-05,
"loss": 9.0793,
"step": 1510
},
{
"epoch": 0.005255514832999101,
"grad_norm": 5.503786087036133,
"learning_rate": 9.089999999999999e-05,
"loss": 9.0934,
"step": 1520
},
{
"epoch": 0.005290090588479358,
"grad_norm": 10.64088249206543,
"learning_rate": 9.149999999999999e-05,
"loss": 9.0582,
"step": 1530
},
{
"epoch": 0.005324666343959616,
"grad_norm": 9.057332038879395,
"learning_rate": 9.209999999999999e-05,
"loss": 9.0736,
"step": 1540
},
{
"epoch": 0.005359242099439873,
"grad_norm": 6.388193130493164,
"learning_rate": 9.269999999999999e-05,
"loss": 9.0618,
"step": 1550
},
{
"epoch": 0.00539381785492013,
"grad_norm": 6.372312545776367,
"learning_rate": 9.329999999999999e-05,
"loss": 9.0672,
"step": 1560
},
{
"epoch": 0.0054283936104003875,
"grad_norm": 9.673222541809082,
"learning_rate": 9.389999999999999e-05,
"loss": 9.0556,
"step": 1570
},
{
"epoch": 0.005462969365880644,
"grad_norm": 8.361108779907227,
"learning_rate": 9.449999999999999e-05,
"loss": 9.0648,
"step": 1580
},
{
"epoch": 0.005497545121360902,
"grad_norm": 8.267026901245117,
"learning_rate": 9.51e-05,
"loss": 9.0594,
"step": 1590
},
{
"epoch": 0.005532120876841159,
"grad_norm": 10.614418983459473,
"learning_rate": 9.57e-05,
"loss": 9.0755,
"step": 1600
}
],
"logging_steps": 10,
"max_steps": 1000000,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"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": 5123186933760000.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}