ZenAI-v2 / last-checkpoint /trainer_state.json
kmnis's picture
Training in progress, step 3500, checkpoint
bcafc6f
raw
history blame
37.3 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.4755480607082632,
"eval_steps": 500,
"global_step": 3500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 4.345879896760937e-06,
"loss": 1.4454,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 5.6541201032390644e-06,
"loss": 1.4588,
"step": 20
},
{
"epoch": 0.01,
"learning_rate": 6.355406060132516e-06,
"loss": 1.3719,
"step": 30
},
{
"epoch": 0.02,
"learning_rate": 6.914575690124303e-06,
"loss": 1.256,
"step": 40
},
{
"epoch": 0.02,
"learning_rate": 7.345389165780057e-06,
"loss": 1.2855,
"step": 50
},
{
"epoch": 0.03,
"learning_rate": 7.663646266610644e-06,
"loss": 1.3919,
"step": 60
},
{
"epoch": 0.03,
"learning_rate": 7.963863644365277e-06,
"loss": 1.404,
"step": 70
},
{
"epoch": 0.03,
"learning_rate": 8.222815896602431e-06,
"loss": 1.4293,
"step": 80
},
{
"epoch": 0.04,
"learning_rate": 8.450488154497406e-06,
"loss": 1.3077,
"step": 90
},
{
"epoch": 0.04,
"learning_rate": 8.653629372258186e-06,
"loss": 1.1747,
"step": 100
},
{
"epoch": 0.05,
"learning_rate": 8.837015420566862e-06,
"loss": 1.2659,
"step": 110
},
{
"epoch": 0.05,
"learning_rate": 9.004150364012388e-06,
"loss": 1.137,
"step": 120
},
{
"epoch": 0.05,
"learning_rate": 9.157681445346895e-06,
"loss": 1.0709,
"step": 130
},
{
"epoch": 0.06,
"learning_rate": 9.299657503090295e-06,
"loss": 1.021,
"step": 140
},
{
"epoch": 0.06,
"learning_rate": 9.43169675964485e-06,
"loss": 1.0298,
"step": 150
},
{
"epoch": 0.07,
"learning_rate": 9.555099634066188e-06,
"loss": 0.9256,
"step": 160
},
{
"epoch": 0.07,
"learning_rate": 9.670926871527948e-06,
"loss": 0.9835,
"step": 170
},
{
"epoch": 0.08,
"learning_rate": 9.780055047302923e-06,
"loss": 0.9704,
"step": 180
},
{
"epoch": 0.08,
"learning_rate": 9.883216879176116e-06,
"loss": 0.8905,
"step": 190
},
{
"epoch": 0.08,
"learning_rate": 9.9714746460726e-06,
"loss": 0.9926,
"step": 200
},
{
"epoch": 0.09,
"learning_rate": 1e-05,
"loss": 0.9701,
"step": 210
},
{
"epoch": 0.09,
"learning_rate": 1e-05,
"loss": 0.8021,
"step": 220
},
{
"epoch": 0.1,
"learning_rate": 1e-05,
"loss": 0.9571,
"step": 230
},
{
"epoch": 0.1,
"learning_rate": 1e-05,
"loss": 0.8937,
"step": 240
},
{
"epoch": 0.11,
"learning_rate": 1e-05,
"loss": 0.7971,
"step": 250
},
{
"epoch": 0.11,
"learning_rate": 1e-05,
"loss": 0.8399,
"step": 260
},
{
"epoch": 0.11,
"learning_rate": 1e-05,
"loss": 0.7557,
"step": 270
},
{
"epoch": 0.12,
"learning_rate": 1e-05,
"loss": 0.871,
"step": 280
},
{
"epoch": 0.12,
"learning_rate": 1e-05,
"loss": 0.9115,
"step": 290
},
{
"epoch": 0.13,
"learning_rate": 1e-05,
"loss": 0.8509,
"step": 300
},
{
"epoch": 0.13,
"learning_rate": 1e-05,
"loss": 0.7635,
"step": 310
},
{
"epoch": 0.13,
"learning_rate": 1e-05,
"loss": 0.8126,
"step": 320
},
{
"epoch": 0.14,
"learning_rate": 1e-05,
"loss": 0.7692,
"step": 330
},
{
"epoch": 0.14,
"learning_rate": 1e-05,
"loss": 0.8264,
"step": 340
},
{
"epoch": 0.15,
"learning_rate": 1e-05,
"loss": 0.8006,
"step": 350
},
{
"epoch": 0.15,
"learning_rate": 1e-05,
"loss": 0.7863,
"step": 360
},
{
"epoch": 0.16,
"learning_rate": 1e-05,
"loss": 0.7427,
"step": 370
},
{
"epoch": 0.16,
"learning_rate": 1e-05,
"loss": 0.7672,
"step": 380
},
{
"epoch": 0.16,
"learning_rate": 1e-05,
"loss": 0.7908,
"step": 390
},
{
"epoch": 0.17,
"learning_rate": 1e-05,
"loss": 0.758,
"step": 400
},
{
"epoch": 0.17,
"learning_rate": 1e-05,
"loss": 0.7599,
"step": 410
},
{
"epoch": 0.18,
"learning_rate": 1e-05,
"loss": 0.7805,
"step": 420
},
{
"epoch": 0.18,
"learning_rate": 1e-05,
"loss": 0.7353,
"step": 430
},
{
"epoch": 0.19,
"learning_rate": 1e-05,
"loss": 0.7634,
"step": 440
},
{
"epoch": 0.19,
"learning_rate": 1e-05,
"loss": 0.822,
"step": 450
},
{
"epoch": 0.19,
"learning_rate": 1e-05,
"loss": 0.7312,
"step": 460
},
{
"epoch": 0.2,
"learning_rate": 1e-05,
"loss": 0.827,
"step": 470
},
{
"epoch": 0.2,
"learning_rate": 1e-05,
"loss": 0.7295,
"step": 480
},
{
"epoch": 0.21,
"learning_rate": 1e-05,
"loss": 0.7707,
"step": 490
},
{
"epoch": 0.21,
"learning_rate": 1e-05,
"loss": 0.7604,
"step": 500
},
{
"epoch": 0.22,
"learning_rate": 1e-05,
"loss": 0.7913,
"step": 510
},
{
"epoch": 0.22,
"learning_rate": 1e-05,
"loss": 0.8093,
"step": 520
},
{
"epoch": 0.22,
"learning_rate": 1e-05,
"loss": 0.7815,
"step": 530
},
{
"epoch": 0.23,
"learning_rate": 1e-05,
"loss": 0.7916,
"step": 540
},
{
"epoch": 0.23,
"learning_rate": 1e-05,
"loss": 0.7065,
"step": 550
},
{
"epoch": 0.24,
"learning_rate": 1e-05,
"loss": 0.8119,
"step": 560
},
{
"epoch": 0.24,
"learning_rate": 1e-05,
"loss": 0.7954,
"step": 570
},
{
"epoch": 0.24,
"learning_rate": 1e-05,
"loss": 0.7627,
"step": 580
},
{
"epoch": 0.25,
"learning_rate": 1e-05,
"loss": 0.7579,
"step": 590
},
{
"epoch": 0.25,
"learning_rate": 1e-05,
"loss": 0.7452,
"step": 600
},
{
"epoch": 0.26,
"learning_rate": 1e-05,
"loss": 0.7241,
"step": 610
},
{
"epoch": 0.26,
"learning_rate": 1e-05,
"loss": 0.7471,
"step": 620
},
{
"epoch": 0.27,
"learning_rate": 1e-05,
"loss": 0.7383,
"step": 630
},
{
"epoch": 0.27,
"learning_rate": 1e-05,
"loss": 0.7276,
"step": 640
},
{
"epoch": 0.27,
"learning_rate": 1e-05,
"loss": 0.7763,
"step": 650
},
{
"epoch": 0.28,
"learning_rate": 1e-05,
"loss": 0.7354,
"step": 660
},
{
"epoch": 0.28,
"learning_rate": 1e-05,
"loss": 0.7331,
"step": 670
},
{
"epoch": 0.29,
"learning_rate": 1e-05,
"loss": 0.809,
"step": 680
},
{
"epoch": 0.29,
"learning_rate": 1e-05,
"loss": 0.6893,
"step": 690
},
{
"epoch": 0.3,
"learning_rate": 1e-05,
"loss": 0.7099,
"step": 700
},
{
"epoch": 0.3,
"learning_rate": 1e-05,
"loss": 0.7721,
"step": 710
},
{
"epoch": 0.3,
"learning_rate": 1e-05,
"loss": 0.7445,
"step": 720
},
{
"epoch": 0.31,
"learning_rate": 1e-05,
"loss": 0.6955,
"step": 730
},
{
"epoch": 0.31,
"learning_rate": 1e-05,
"loss": 0.8126,
"step": 740
},
{
"epoch": 0.32,
"learning_rate": 1e-05,
"loss": 0.6801,
"step": 750
},
{
"epoch": 0.32,
"learning_rate": 1e-05,
"loss": 0.6793,
"step": 760
},
{
"epoch": 0.32,
"learning_rate": 1e-05,
"loss": 0.7701,
"step": 770
},
{
"epoch": 0.33,
"learning_rate": 1e-05,
"loss": 0.7438,
"step": 780
},
{
"epoch": 0.33,
"learning_rate": 1e-05,
"loss": 0.7271,
"step": 790
},
{
"epoch": 0.34,
"learning_rate": 1e-05,
"loss": 0.6587,
"step": 800
},
{
"epoch": 0.34,
"learning_rate": 1e-05,
"loss": 0.6863,
"step": 810
},
{
"epoch": 0.35,
"learning_rate": 1e-05,
"loss": 0.6853,
"step": 820
},
{
"epoch": 0.35,
"learning_rate": 1e-05,
"loss": 0.7209,
"step": 830
},
{
"epoch": 0.35,
"learning_rate": 1e-05,
"loss": 0.767,
"step": 840
},
{
"epoch": 0.36,
"learning_rate": 1e-05,
"loss": 0.7447,
"step": 850
},
{
"epoch": 0.36,
"learning_rate": 1e-05,
"loss": 0.7326,
"step": 860
},
{
"epoch": 0.37,
"learning_rate": 1e-05,
"loss": 0.729,
"step": 870
},
{
"epoch": 0.37,
"learning_rate": 1e-05,
"loss": 0.7824,
"step": 880
},
{
"epoch": 0.38,
"learning_rate": 1e-05,
"loss": 0.6864,
"step": 890
},
{
"epoch": 0.38,
"learning_rate": 1e-05,
"loss": 0.7131,
"step": 900
},
{
"epoch": 0.38,
"learning_rate": 1e-05,
"loss": 0.7237,
"step": 910
},
{
"epoch": 0.39,
"learning_rate": 1e-05,
"loss": 0.7222,
"step": 920
},
{
"epoch": 0.39,
"learning_rate": 1e-05,
"loss": 0.7473,
"step": 930
},
{
"epoch": 0.4,
"learning_rate": 1e-05,
"loss": 0.6746,
"step": 940
},
{
"epoch": 0.4,
"learning_rate": 1e-05,
"loss": 0.6792,
"step": 950
},
{
"epoch": 0.4,
"learning_rate": 1e-05,
"loss": 0.7368,
"step": 960
},
{
"epoch": 0.41,
"learning_rate": 1e-05,
"loss": 0.7109,
"step": 970
},
{
"epoch": 0.41,
"learning_rate": 1e-05,
"loss": 0.6923,
"step": 980
},
{
"epoch": 0.42,
"learning_rate": 1e-05,
"loss": 0.7358,
"step": 990
},
{
"epoch": 0.42,
"learning_rate": 1e-05,
"loss": 0.715,
"step": 1000
},
{
"epoch": 0.43,
"learning_rate": 1e-05,
"loss": 0.7782,
"step": 1010
},
{
"epoch": 0.43,
"learning_rate": 1e-05,
"loss": 0.756,
"step": 1020
},
{
"epoch": 0.43,
"learning_rate": 1e-05,
"loss": 0.6691,
"step": 1030
},
{
"epoch": 0.44,
"learning_rate": 1e-05,
"loss": 0.7408,
"step": 1040
},
{
"epoch": 0.44,
"learning_rate": 1e-05,
"loss": 0.6667,
"step": 1050
},
{
"epoch": 0.45,
"learning_rate": 1e-05,
"loss": 0.7215,
"step": 1060
},
{
"epoch": 0.45,
"learning_rate": 1e-05,
"loss": 0.7156,
"step": 1070
},
{
"epoch": 0.46,
"learning_rate": 1e-05,
"loss": 0.7457,
"step": 1080
},
{
"epoch": 0.46,
"learning_rate": 1e-05,
"loss": 0.7344,
"step": 1090
},
{
"epoch": 0.46,
"learning_rate": 1e-05,
"loss": 0.6902,
"step": 1100
},
{
"epoch": 0.47,
"learning_rate": 1e-05,
"loss": 0.6864,
"step": 1110
},
{
"epoch": 0.47,
"learning_rate": 1e-05,
"loss": 0.7097,
"step": 1120
},
{
"epoch": 0.48,
"learning_rate": 1e-05,
"loss": 0.7181,
"step": 1130
},
{
"epoch": 0.48,
"learning_rate": 1e-05,
"loss": 0.7039,
"step": 1140
},
{
"epoch": 0.48,
"learning_rate": 1e-05,
"loss": 0.7255,
"step": 1150
},
{
"epoch": 0.49,
"learning_rate": 1e-05,
"loss": 0.7,
"step": 1160
},
{
"epoch": 0.49,
"learning_rate": 1e-05,
"loss": 0.7237,
"step": 1170
},
{
"epoch": 0.5,
"learning_rate": 1e-05,
"loss": 0.6806,
"step": 1180
},
{
"epoch": 0.5,
"learning_rate": 1e-05,
"loss": 0.6932,
"step": 1190
},
{
"epoch": 0.51,
"learning_rate": 1e-05,
"loss": 0.6879,
"step": 1200
},
{
"epoch": 0.51,
"learning_rate": 1e-05,
"loss": 0.7186,
"step": 1210
},
{
"epoch": 0.51,
"learning_rate": 1e-05,
"loss": 0.672,
"step": 1220
},
{
"epoch": 0.52,
"learning_rate": 1e-05,
"loss": 0.7086,
"step": 1230
},
{
"epoch": 0.52,
"learning_rate": 1e-05,
"loss": 0.7354,
"step": 1240
},
{
"epoch": 0.53,
"learning_rate": 1e-05,
"loss": 0.776,
"step": 1250
},
{
"epoch": 0.53,
"learning_rate": 1e-05,
"loss": 0.7083,
"step": 1260
},
{
"epoch": 0.54,
"learning_rate": 1e-05,
"loss": 0.7331,
"step": 1270
},
{
"epoch": 0.54,
"learning_rate": 1e-05,
"loss": 0.7114,
"step": 1280
},
{
"epoch": 0.54,
"learning_rate": 1e-05,
"loss": 0.7341,
"step": 1290
},
{
"epoch": 0.55,
"learning_rate": 1e-05,
"loss": 0.7306,
"step": 1300
},
{
"epoch": 0.55,
"learning_rate": 1e-05,
"loss": 0.7518,
"step": 1310
},
{
"epoch": 0.56,
"learning_rate": 1e-05,
"loss": 0.6534,
"step": 1320
},
{
"epoch": 0.56,
"learning_rate": 1e-05,
"loss": 0.7034,
"step": 1330
},
{
"epoch": 0.56,
"learning_rate": 1e-05,
"loss": 0.7164,
"step": 1340
},
{
"epoch": 0.57,
"learning_rate": 1e-05,
"loss": 0.7005,
"step": 1350
},
{
"epoch": 0.57,
"learning_rate": 1e-05,
"loss": 0.7034,
"step": 1360
},
{
"epoch": 0.58,
"learning_rate": 1e-05,
"loss": 0.7044,
"step": 1370
},
{
"epoch": 0.58,
"learning_rate": 1e-05,
"loss": 0.7411,
"step": 1380
},
{
"epoch": 0.59,
"learning_rate": 1e-05,
"loss": 0.6513,
"step": 1390
},
{
"epoch": 0.59,
"learning_rate": 1e-05,
"loss": 0.7152,
"step": 1400
},
{
"epoch": 0.59,
"learning_rate": 1e-05,
"loss": 0.6982,
"step": 1410
},
{
"epoch": 0.6,
"learning_rate": 1e-05,
"loss": 0.6936,
"step": 1420
},
{
"epoch": 0.6,
"learning_rate": 1e-05,
"loss": 0.7053,
"step": 1430
},
{
"epoch": 0.61,
"learning_rate": 1e-05,
"loss": 0.6756,
"step": 1440
},
{
"epoch": 0.61,
"learning_rate": 1e-05,
"loss": 0.6939,
"step": 1450
},
{
"epoch": 0.62,
"learning_rate": 1e-05,
"loss": 0.7322,
"step": 1460
},
{
"epoch": 0.62,
"learning_rate": 1e-05,
"loss": 0.6409,
"step": 1470
},
{
"epoch": 0.62,
"learning_rate": 1e-05,
"loss": 0.7243,
"step": 1480
},
{
"epoch": 0.63,
"learning_rate": 1e-05,
"loss": 0.6697,
"step": 1490
},
{
"epoch": 0.63,
"learning_rate": 1e-05,
"loss": 0.705,
"step": 1500
},
{
"epoch": 0.64,
"learning_rate": 1e-05,
"loss": 0.6431,
"step": 1510
},
{
"epoch": 0.64,
"learning_rate": 1e-05,
"loss": 0.6615,
"step": 1520
},
{
"epoch": 0.65,
"learning_rate": 1e-05,
"loss": 0.7101,
"step": 1530
},
{
"epoch": 0.65,
"learning_rate": 1e-05,
"loss": 0.7265,
"step": 1540
},
{
"epoch": 0.65,
"learning_rate": 1e-05,
"loss": 0.7188,
"step": 1550
},
{
"epoch": 0.66,
"learning_rate": 1e-05,
"loss": 0.7148,
"step": 1560
},
{
"epoch": 0.66,
"learning_rate": 1e-05,
"loss": 0.7713,
"step": 1570
},
{
"epoch": 0.67,
"learning_rate": 1e-05,
"loss": 0.6934,
"step": 1580
},
{
"epoch": 0.67,
"learning_rate": 1e-05,
"loss": 0.6369,
"step": 1590
},
{
"epoch": 0.67,
"learning_rate": 1e-05,
"loss": 0.673,
"step": 1600
},
{
"epoch": 0.68,
"learning_rate": 1e-05,
"loss": 0.7517,
"step": 1610
},
{
"epoch": 0.68,
"learning_rate": 1e-05,
"loss": 0.6622,
"step": 1620
},
{
"epoch": 0.69,
"learning_rate": 1e-05,
"loss": 0.6657,
"step": 1630
},
{
"epoch": 0.69,
"learning_rate": 1e-05,
"loss": 0.6895,
"step": 1640
},
{
"epoch": 0.7,
"learning_rate": 1e-05,
"loss": 0.7961,
"step": 1650
},
{
"epoch": 0.7,
"learning_rate": 1e-05,
"loss": 0.677,
"step": 1660
},
{
"epoch": 0.7,
"learning_rate": 1e-05,
"loss": 0.6579,
"step": 1670
},
{
"epoch": 0.71,
"learning_rate": 1e-05,
"loss": 0.6797,
"step": 1680
},
{
"epoch": 0.71,
"learning_rate": 1e-05,
"loss": 0.7199,
"step": 1690
},
{
"epoch": 0.72,
"learning_rate": 1e-05,
"loss": 0.7065,
"step": 1700
},
{
"epoch": 0.72,
"learning_rate": 1e-05,
"loss": 0.7162,
"step": 1710
},
{
"epoch": 0.73,
"learning_rate": 1e-05,
"loss": 0.6851,
"step": 1720
},
{
"epoch": 0.73,
"learning_rate": 1e-05,
"loss": 0.7251,
"step": 1730
},
{
"epoch": 0.73,
"learning_rate": 1e-05,
"loss": 0.6808,
"step": 1740
},
{
"epoch": 0.74,
"learning_rate": 1e-05,
"loss": 0.7358,
"step": 1750
},
{
"epoch": 0.74,
"learning_rate": 1e-05,
"loss": 0.7403,
"step": 1760
},
{
"epoch": 0.75,
"learning_rate": 1e-05,
"loss": 0.7051,
"step": 1770
},
{
"epoch": 0.75,
"learning_rate": 1e-05,
"loss": 0.6662,
"step": 1780
},
{
"epoch": 0.75,
"learning_rate": 1e-05,
"loss": 0.6922,
"step": 1790
},
{
"epoch": 0.76,
"learning_rate": 1e-05,
"loss": 0.7086,
"step": 1800
},
{
"epoch": 0.76,
"learning_rate": 1e-05,
"loss": 0.6558,
"step": 1810
},
{
"epoch": 0.77,
"learning_rate": 1e-05,
"loss": 0.6953,
"step": 1820
},
{
"epoch": 0.77,
"learning_rate": 1e-05,
"loss": 0.7355,
"step": 1830
},
{
"epoch": 0.78,
"learning_rate": 1e-05,
"loss": 0.7042,
"step": 1840
},
{
"epoch": 0.78,
"learning_rate": 1e-05,
"loss": 0.6836,
"step": 1850
},
{
"epoch": 0.78,
"learning_rate": 1e-05,
"loss": 0.7482,
"step": 1860
},
{
"epoch": 0.79,
"learning_rate": 1e-05,
"loss": 0.7727,
"step": 1870
},
{
"epoch": 0.79,
"learning_rate": 1e-05,
"loss": 0.7302,
"step": 1880
},
{
"epoch": 0.8,
"learning_rate": 1e-05,
"loss": 0.686,
"step": 1890
},
{
"epoch": 0.8,
"learning_rate": 1e-05,
"loss": 0.7062,
"step": 1900
},
{
"epoch": 0.81,
"learning_rate": 1e-05,
"loss": 0.7315,
"step": 1910
},
{
"epoch": 0.81,
"learning_rate": 1e-05,
"loss": 0.6259,
"step": 1920
},
{
"epoch": 0.81,
"learning_rate": 1e-05,
"loss": 0.6999,
"step": 1930
},
{
"epoch": 0.82,
"learning_rate": 1e-05,
"loss": 0.6675,
"step": 1940
},
{
"epoch": 0.82,
"learning_rate": 1e-05,
"loss": 0.6953,
"step": 1950
},
{
"epoch": 0.83,
"learning_rate": 1e-05,
"loss": 0.7274,
"step": 1960
},
{
"epoch": 0.83,
"learning_rate": 1e-05,
"loss": 0.6864,
"step": 1970
},
{
"epoch": 0.83,
"learning_rate": 1e-05,
"loss": 0.6822,
"step": 1980
},
{
"epoch": 0.84,
"learning_rate": 1e-05,
"loss": 0.6641,
"step": 1990
},
{
"epoch": 0.84,
"learning_rate": 1e-05,
"loss": 0.6513,
"step": 2000
},
{
"epoch": 0.85,
"learning_rate": 1e-05,
"loss": 0.6249,
"step": 2010
},
{
"epoch": 0.85,
"learning_rate": 1e-05,
"loss": 0.6665,
"step": 2020
},
{
"epoch": 0.86,
"learning_rate": 1e-05,
"loss": 0.702,
"step": 2030
},
{
"epoch": 0.86,
"learning_rate": 1e-05,
"loss": 0.7431,
"step": 2040
},
{
"epoch": 0.86,
"learning_rate": 1e-05,
"loss": 0.6716,
"step": 2050
},
{
"epoch": 0.87,
"learning_rate": 1e-05,
"loss": 0.6754,
"step": 2060
},
{
"epoch": 0.87,
"learning_rate": 1e-05,
"loss": 0.7649,
"step": 2070
},
{
"epoch": 0.88,
"learning_rate": 1e-05,
"loss": 0.62,
"step": 2080
},
{
"epoch": 0.88,
"learning_rate": 1e-05,
"loss": 0.6863,
"step": 2090
},
{
"epoch": 0.89,
"learning_rate": 1e-05,
"loss": 0.6828,
"step": 2100
},
{
"epoch": 0.89,
"learning_rate": 1e-05,
"loss": 0.6259,
"step": 2110
},
{
"epoch": 0.89,
"learning_rate": 1e-05,
"loss": 0.76,
"step": 2120
},
{
"epoch": 0.9,
"learning_rate": 1e-05,
"loss": 0.74,
"step": 2130
},
{
"epoch": 0.9,
"learning_rate": 1e-05,
"loss": 0.68,
"step": 2140
},
{
"epoch": 0.91,
"learning_rate": 1e-05,
"loss": 0.6328,
"step": 2150
},
{
"epoch": 0.91,
"learning_rate": 1e-05,
"loss": 0.6892,
"step": 2160
},
{
"epoch": 0.91,
"learning_rate": 1e-05,
"loss": 0.7415,
"step": 2170
},
{
"epoch": 0.92,
"learning_rate": 1e-05,
"loss": 0.6976,
"step": 2180
},
{
"epoch": 0.92,
"learning_rate": 1e-05,
"loss": 0.6841,
"step": 2190
},
{
"epoch": 0.93,
"learning_rate": 1e-05,
"loss": 0.68,
"step": 2200
},
{
"epoch": 0.93,
"learning_rate": 1e-05,
"loss": 0.7259,
"step": 2210
},
{
"epoch": 0.94,
"learning_rate": 1e-05,
"loss": 0.7159,
"step": 2220
},
{
"epoch": 0.94,
"learning_rate": 1e-05,
"loss": 0.6852,
"step": 2230
},
{
"epoch": 0.94,
"learning_rate": 1e-05,
"loss": 0.7414,
"step": 2240
},
{
"epoch": 0.95,
"learning_rate": 1e-05,
"loss": 0.6784,
"step": 2250
},
{
"epoch": 0.95,
"learning_rate": 1e-05,
"loss": 0.599,
"step": 2260
},
{
"epoch": 0.96,
"learning_rate": 1e-05,
"loss": 0.7278,
"step": 2270
},
{
"epoch": 0.96,
"learning_rate": 1e-05,
"loss": 0.7143,
"step": 2280
},
{
"epoch": 0.97,
"learning_rate": 1e-05,
"loss": 0.6473,
"step": 2290
},
{
"epoch": 0.97,
"learning_rate": 1e-05,
"loss": 0.7026,
"step": 2300
},
{
"epoch": 0.97,
"learning_rate": 1e-05,
"loss": 0.6816,
"step": 2310
},
{
"epoch": 0.98,
"learning_rate": 1e-05,
"loss": 0.7041,
"step": 2320
},
{
"epoch": 0.98,
"learning_rate": 1e-05,
"loss": 0.6185,
"step": 2330
},
{
"epoch": 0.99,
"learning_rate": 1e-05,
"loss": 0.6164,
"step": 2340
},
{
"epoch": 0.99,
"learning_rate": 1e-05,
"loss": 0.6526,
"step": 2350
},
{
"epoch": 0.99,
"learning_rate": 1e-05,
"loss": 0.7141,
"step": 2360
},
{
"epoch": 1.0,
"learning_rate": 1e-05,
"loss": 0.6393,
"step": 2370
},
{
"epoch": 1.0,
"learning_rate": 1e-05,
"loss": 0.6799,
"step": 2380
},
{
"epoch": 1.01,
"learning_rate": 1e-05,
"loss": 0.6859,
"step": 2390
},
{
"epoch": 1.01,
"learning_rate": 1e-05,
"loss": 0.661,
"step": 2400
},
{
"epoch": 1.02,
"learning_rate": 1e-05,
"loss": 0.7197,
"step": 2410
},
{
"epoch": 1.02,
"learning_rate": 1e-05,
"loss": 0.7034,
"step": 2420
},
{
"epoch": 1.02,
"learning_rate": 1e-05,
"loss": 0.6855,
"step": 2430
},
{
"epoch": 1.03,
"learning_rate": 1e-05,
"loss": 0.6689,
"step": 2440
},
{
"epoch": 1.03,
"learning_rate": 1e-05,
"loss": 0.6702,
"step": 2450
},
{
"epoch": 1.04,
"learning_rate": 1e-05,
"loss": 0.6505,
"step": 2460
},
{
"epoch": 1.04,
"learning_rate": 1e-05,
"loss": 0.6594,
"step": 2470
},
{
"epoch": 1.05,
"learning_rate": 1e-05,
"loss": 0.6523,
"step": 2480
},
{
"epoch": 1.05,
"learning_rate": 1e-05,
"loss": 0.6744,
"step": 2490
},
{
"epoch": 1.05,
"learning_rate": 1e-05,
"loss": 0.6425,
"step": 2500
},
{
"epoch": 1.06,
"learning_rate": 1e-05,
"loss": 0.6704,
"step": 2510
},
{
"epoch": 1.06,
"learning_rate": 1e-05,
"loss": 0.6431,
"step": 2520
},
{
"epoch": 1.07,
"learning_rate": 1e-05,
"loss": 0.702,
"step": 2530
},
{
"epoch": 1.07,
"learning_rate": 1e-05,
"loss": 0.7384,
"step": 2540
},
{
"epoch": 1.08,
"learning_rate": 1e-05,
"loss": 0.6324,
"step": 2550
},
{
"epoch": 1.08,
"learning_rate": 1e-05,
"loss": 0.647,
"step": 2560
},
{
"epoch": 1.08,
"learning_rate": 1e-05,
"loss": 0.6693,
"step": 2570
},
{
"epoch": 1.09,
"learning_rate": 1e-05,
"loss": 0.6746,
"step": 2580
},
{
"epoch": 1.09,
"learning_rate": 1e-05,
"loss": 0.6975,
"step": 2590
},
{
"epoch": 1.1,
"learning_rate": 1e-05,
"loss": 0.594,
"step": 2600
},
{
"epoch": 1.1,
"learning_rate": 1e-05,
"loss": 0.7621,
"step": 2610
},
{
"epoch": 1.1,
"learning_rate": 1e-05,
"loss": 0.6325,
"step": 2620
},
{
"epoch": 1.11,
"learning_rate": 1e-05,
"loss": 0.6287,
"step": 2630
},
{
"epoch": 1.11,
"learning_rate": 1e-05,
"loss": 0.7122,
"step": 2640
},
{
"epoch": 1.12,
"learning_rate": 1e-05,
"loss": 0.6411,
"step": 2650
},
{
"epoch": 1.12,
"learning_rate": 1e-05,
"loss": 0.7214,
"step": 2660
},
{
"epoch": 1.13,
"learning_rate": 1e-05,
"loss": 0.6269,
"step": 2670
},
{
"epoch": 1.13,
"learning_rate": 1e-05,
"loss": 0.6836,
"step": 2680
},
{
"epoch": 1.13,
"learning_rate": 1e-05,
"loss": 0.6308,
"step": 2690
},
{
"epoch": 1.14,
"learning_rate": 1e-05,
"loss": 0.6759,
"step": 2700
},
{
"epoch": 1.14,
"learning_rate": 1e-05,
"loss": 0.6376,
"step": 2710
},
{
"epoch": 1.15,
"learning_rate": 1e-05,
"loss": 0.7418,
"step": 2720
},
{
"epoch": 1.15,
"learning_rate": 1e-05,
"loss": 0.6286,
"step": 2730
},
{
"epoch": 1.16,
"learning_rate": 1e-05,
"loss": 0.722,
"step": 2740
},
{
"epoch": 1.16,
"learning_rate": 1e-05,
"loss": 0.6689,
"step": 2750
},
{
"epoch": 1.16,
"learning_rate": 1e-05,
"loss": 0.598,
"step": 2760
},
{
"epoch": 1.17,
"learning_rate": 1e-05,
"loss": 0.6646,
"step": 2770
},
{
"epoch": 1.17,
"learning_rate": 1e-05,
"loss": 0.6716,
"step": 2780
},
{
"epoch": 1.18,
"learning_rate": 1e-05,
"loss": 0.6279,
"step": 2790
},
{
"epoch": 1.18,
"learning_rate": 1e-05,
"loss": 0.61,
"step": 2800
},
{
"epoch": 1.18,
"learning_rate": 1e-05,
"loss": 0.6543,
"step": 2810
},
{
"epoch": 1.19,
"learning_rate": 1e-05,
"loss": 0.6929,
"step": 2820
},
{
"epoch": 1.19,
"learning_rate": 1e-05,
"loss": 0.6731,
"step": 2830
},
{
"epoch": 1.2,
"learning_rate": 1e-05,
"loss": 0.6931,
"step": 2840
},
{
"epoch": 1.2,
"learning_rate": 1e-05,
"loss": 0.7082,
"step": 2850
},
{
"epoch": 1.21,
"learning_rate": 1e-05,
"loss": 0.6964,
"step": 2860
},
{
"epoch": 1.21,
"learning_rate": 1e-05,
"loss": 0.654,
"step": 2870
},
{
"epoch": 1.21,
"learning_rate": 1e-05,
"loss": 0.6697,
"step": 2880
},
{
"epoch": 1.22,
"learning_rate": 1e-05,
"loss": 0.6724,
"step": 2890
},
{
"epoch": 1.22,
"learning_rate": 1e-05,
"loss": 0.6244,
"step": 2900
},
{
"epoch": 1.23,
"learning_rate": 1e-05,
"loss": 0.6244,
"step": 2910
},
{
"epoch": 1.23,
"learning_rate": 1e-05,
"loss": 0.6343,
"step": 2920
},
{
"epoch": 1.24,
"learning_rate": 1e-05,
"loss": 0.6054,
"step": 2930
},
{
"epoch": 1.24,
"learning_rate": 1e-05,
"loss": 0.6593,
"step": 2940
},
{
"epoch": 1.24,
"learning_rate": 1e-05,
"loss": 0.6636,
"step": 2950
},
{
"epoch": 1.25,
"learning_rate": 1e-05,
"loss": 0.6098,
"step": 2960
},
{
"epoch": 1.25,
"learning_rate": 1e-05,
"loss": 0.6938,
"step": 2970
},
{
"epoch": 1.26,
"learning_rate": 1e-05,
"loss": 0.7027,
"step": 2980
},
{
"epoch": 1.26,
"learning_rate": 1e-05,
"loss": 0.6503,
"step": 2990
},
{
"epoch": 1.26,
"learning_rate": 1e-05,
"loss": 0.6872,
"step": 3000
},
{
"epoch": 1.27,
"learning_rate": 1e-05,
"loss": 0.6234,
"step": 3010
},
{
"epoch": 1.27,
"learning_rate": 1e-05,
"loss": 0.6219,
"step": 3020
},
{
"epoch": 1.28,
"learning_rate": 1e-05,
"loss": 0.5698,
"step": 3030
},
{
"epoch": 1.28,
"learning_rate": 1e-05,
"loss": 0.6414,
"step": 3040
},
{
"epoch": 1.29,
"learning_rate": 1e-05,
"loss": 0.659,
"step": 3050
},
{
"epoch": 1.29,
"learning_rate": 1e-05,
"loss": 0.7126,
"step": 3060
},
{
"epoch": 1.29,
"learning_rate": 1e-05,
"loss": 0.6448,
"step": 3070
},
{
"epoch": 1.3,
"learning_rate": 1e-05,
"loss": 0.629,
"step": 3080
},
{
"epoch": 1.3,
"learning_rate": 1e-05,
"loss": 0.6906,
"step": 3090
},
{
"epoch": 1.31,
"learning_rate": 1e-05,
"loss": 0.6332,
"step": 3100
},
{
"epoch": 1.31,
"learning_rate": 1e-05,
"loss": 0.6317,
"step": 3110
},
{
"epoch": 1.32,
"learning_rate": 1e-05,
"loss": 0.6817,
"step": 3120
},
{
"epoch": 1.32,
"learning_rate": 1e-05,
"loss": 0.6523,
"step": 3130
},
{
"epoch": 1.32,
"learning_rate": 1e-05,
"loss": 0.6224,
"step": 3140
},
{
"epoch": 1.33,
"learning_rate": 1e-05,
"loss": 0.6497,
"step": 3150
},
{
"epoch": 1.33,
"learning_rate": 1e-05,
"loss": 0.65,
"step": 3160
},
{
"epoch": 1.34,
"learning_rate": 1e-05,
"loss": 0.6525,
"step": 3170
},
{
"epoch": 1.34,
"learning_rate": 1e-05,
"loss": 0.6569,
"step": 3180
},
{
"epoch": 1.34,
"learning_rate": 1e-05,
"loss": 0.6612,
"step": 3190
},
{
"epoch": 1.35,
"learning_rate": 1e-05,
"loss": 0.6478,
"step": 3200
},
{
"epoch": 1.35,
"learning_rate": 1e-05,
"loss": 0.6334,
"step": 3210
},
{
"epoch": 1.36,
"learning_rate": 1e-05,
"loss": 0.6665,
"step": 3220
},
{
"epoch": 1.36,
"learning_rate": 1e-05,
"loss": 0.6804,
"step": 3230
},
{
"epoch": 1.37,
"learning_rate": 1e-05,
"loss": 0.6878,
"step": 3240
},
{
"epoch": 1.37,
"learning_rate": 1e-05,
"loss": 0.6877,
"step": 3250
},
{
"epoch": 1.37,
"learning_rate": 1e-05,
"loss": 0.6811,
"step": 3260
},
{
"epoch": 1.38,
"learning_rate": 1e-05,
"loss": 0.6677,
"step": 3270
},
{
"epoch": 1.38,
"learning_rate": 1e-05,
"loss": 0.8179,
"step": 3280
},
{
"epoch": 1.39,
"learning_rate": 1e-05,
"loss": 0.6174,
"step": 3290
},
{
"epoch": 1.39,
"learning_rate": 1e-05,
"loss": 0.7416,
"step": 3300
},
{
"epoch": 1.4,
"learning_rate": 1e-05,
"loss": 0.6779,
"step": 3310
},
{
"epoch": 1.4,
"learning_rate": 1e-05,
"loss": 0.6456,
"step": 3320
},
{
"epoch": 1.4,
"learning_rate": 1e-05,
"loss": 0.6693,
"step": 3330
},
{
"epoch": 1.41,
"learning_rate": 1e-05,
"loss": 0.6122,
"step": 3340
},
{
"epoch": 1.41,
"learning_rate": 1e-05,
"loss": 0.6766,
"step": 3350
},
{
"epoch": 1.42,
"learning_rate": 1e-05,
"loss": 0.6921,
"step": 3360
},
{
"epoch": 1.42,
"learning_rate": 1e-05,
"loss": 0.6365,
"step": 3370
},
{
"epoch": 1.42,
"learning_rate": 1e-05,
"loss": 0.6456,
"step": 3380
},
{
"epoch": 1.43,
"learning_rate": 1e-05,
"loss": 0.665,
"step": 3390
},
{
"epoch": 1.43,
"learning_rate": 1e-05,
"loss": 0.7139,
"step": 3400
},
{
"epoch": 1.44,
"learning_rate": 1e-05,
"loss": 0.6901,
"step": 3410
},
{
"epoch": 1.44,
"learning_rate": 1e-05,
"loss": 0.5972,
"step": 3420
},
{
"epoch": 1.45,
"learning_rate": 1e-05,
"loss": 0.6747,
"step": 3430
},
{
"epoch": 1.45,
"learning_rate": 1e-05,
"loss": 0.6729,
"step": 3440
},
{
"epoch": 1.45,
"learning_rate": 1e-05,
"loss": 0.6376,
"step": 3450
},
{
"epoch": 1.46,
"learning_rate": 1e-05,
"loss": 0.677,
"step": 3460
},
{
"epoch": 1.46,
"learning_rate": 1e-05,
"loss": 0.6607,
"step": 3470
},
{
"epoch": 1.47,
"learning_rate": 1e-05,
"loss": 0.7486,
"step": 3480
},
{
"epoch": 1.47,
"learning_rate": 1e-05,
"loss": 0.6687,
"step": 3490
},
{
"epoch": 1.48,
"learning_rate": 1e-05,
"loss": 0.6098,
"step": 3500
}
],
"logging_steps": 10,
"max_steps": 10000,
"num_train_epochs": 5,
"save_steps": 500,
"total_flos": 879985321574400.0,
"trial_name": null,
"trial_params": null
}