ZenAI-v2 / last-checkpoint /trainer_state.json
kmnis's picture
Training in progress, step 2000, checkpoint
2ae4579
raw
history blame
21.6 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.8431703204047217,
"eval_steps": 500,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 4.1470233384070705e-06,
"loss": 1.4705,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 5.4552635448851985e-06,
"loss": 1.5012,
"step": 20
},
{
"epoch": 0.01,
"learning_rate": 6.289174995846284e-06,
"loss": 1.385,
"step": 30
},
{
"epoch": 0.02,
"learning_rate": 6.865549773769684e-06,
"loss": 1.2729,
"step": 40
},
{
"epoch": 0.02,
"learning_rate": 7.306472495116047e-06,
"loss": 1.2978,
"step": 50
},
{
"epoch": 0.03,
"learning_rate": 7.663646266610644e-06,
"loss": 1.4354,
"step": 60
},
{
"epoch": 0.03,
"learning_rate": 7.963863644365277e-06,
"loss": 1.4443,
"step": 70
},
{
"epoch": 0.03,
"learning_rate": 8.222815896602431e-06,
"loss": 1.5108,
"step": 80
},
{
"epoch": 0.04,
"learning_rate": 8.450488154497406e-06,
"loss": 1.3964,
"step": 90
},
{
"epoch": 0.04,
"learning_rate": 8.653629372258186e-06,
"loss": 1.2203,
"step": 100
},
{
"epoch": 0.05,
"learning_rate": 8.837015420566862e-06,
"loss": 1.3174,
"step": 110
},
{
"epoch": 0.05,
"learning_rate": 9.004150364012388e-06,
"loss": 1.1764,
"step": 120
},
{
"epoch": 0.05,
"learning_rate": 9.157681445346895e-06,
"loss": 1.1009,
"step": 130
},
{
"epoch": 0.06,
"learning_rate": 9.299657503090295e-06,
"loss": 1.0422,
"step": 140
},
{
"epoch": 0.06,
"learning_rate": 9.43169675964485e-06,
"loss": 1.0693,
"step": 150
},
{
"epoch": 0.07,
"learning_rate": 9.555099634066188e-06,
"loss": 0.9296,
"step": 160
},
{
"epoch": 0.07,
"learning_rate": 9.670926871527948e-06,
"loss": 1.0445,
"step": 170
},
{
"epoch": 0.08,
"learning_rate": 9.780055047302923e-06,
"loss": 1.008,
"step": 180
},
{
"epoch": 0.08,
"learning_rate": 9.883216879176116e-06,
"loss": 0.9062,
"step": 190
},
{
"epoch": 0.08,
"learning_rate": 9.98103107994822e-06,
"loss": 1.0707,
"step": 200
},
{
"epoch": 0.09,
"learning_rate": 1e-05,
"loss": 1.0204,
"step": 210
},
{
"epoch": 0.09,
"learning_rate": 1e-05,
"loss": 0.8036,
"step": 220
},
{
"epoch": 0.1,
"learning_rate": 1e-05,
"loss": 0.9934,
"step": 230
},
{
"epoch": 0.1,
"learning_rate": 1e-05,
"loss": 0.9144,
"step": 240
},
{
"epoch": 0.11,
"learning_rate": 1e-05,
"loss": 0.8351,
"step": 250
},
{
"epoch": 0.11,
"learning_rate": 1e-05,
"loss": 0.8797,
"step": 260
},
{
"epoch": 0.11,
"learning_rate": 1e-05,
"loss": 0.7679,
"step": 270
},
{
"epoch": 0.12,
"learning_rate": 1e-05,
"loss": 0.9388,
"step": 280
},
{
"epoch": 0.12,
"learning_rate": 1e-05,
"loss": 0.9333,
"step": 290
},
{
"epoch": 0.13,
"learning_rate": 1e-05,
"loss": 0.8698,
"step": 300
},
{
"epoch": 0.13,
"learning_rate": 1e-05,
"loss": 0.7842,
"step": 310
},
{
"epoch": 0.13,
"learning_rate": 1e-05,
"loss": 0.8371,
"step": 320
},
{
"epoch": 0.14,
"learning_rate": 1e-05,
"loss": 0.7694,
"step": 330
},
{
"epoch": 0.14,
"learning_rate": 1e-05,
"loss": 0.8426,
"step": 340
},
{
"epoch": 0.15,
"learning_rate": 1e-05,
"loss": 0.7972,
"step": 350
},
{
"epoch": 0.15,
"learning_rate": 1e-05,
"loss": 0.8183,
"step": 360
},
{
"epoch": 0.16,
"learning_rate": 1e-05,
"loss": 0.7462,
"step": 370
},
{
"epoch": 0.16,
"learning_rate": 1e-05,
"loss": 0.7919,
"step": 380
},
{
"epoch": 0.16,
"learning_rate": 1e-05,
"loss": 0.8173,
"step": 390
},
{
"epoch": 0.17,
"learning_rate": 1e-05,
"loss": 0.7654,
"step": 400
},
{
"epoch": 0.17,
"learning_rate": 1e-05,
"loss": 0.7601,
"step": 410
},
{
"epoch": 0.18,
"learning_rate": 1e-05,
"loss": 0.788,
"step": 420
},
{
"epoch": 0.18,
"learning_rate": 1e-05,
"loss": 0.7488,
"step": 430
},
{
"epoch": 0.19,
"learning_rate": 1e-05,
"loss": 0.7721,
"step": 440
},
{
"epoch": 0.19,
"learning_rate": 1e-05,
"loss": 0.8289,
"step": 450
},
{
"epoch": 0.19,
"learning_rate": 1e-05,
"loss": 0.7367,
"step": 460
},
{
"epoch": 0.2,
"learning_rate": 1e-05,
"loss": 0.8422,
"step": 470
},
{
"epoch": 0.2,
"learning_rate": 1e-05,
"loss": 0.7383,
"step": 480
},
{
"epoch": 0.21,
"learning_rate": 1e-05,
"loss": 0.7729,
"step": 490
},
{
"epoch": 0.21,
"learning_rate": 1e-05,
"loss": 0.7537,
"step": 500
},
{
"epoch": 0.22,
"learning_rate": 1e-05,
"loss": 0.7956,
"step": 510
},
{
"epoch": 0.22,
"learning_rate": 1e-05,
"loss": 0.8438,
"step": 520
},
{
"epoch": 0.22,
"learning_rate": 1e-05,
"loss": 0.7865,
"step": 530
},
{
"epoch": 0.23,
"learning_rate": 1e-05,
"loss": 0.7995,
"step": 540
},
{
"epoch": 0.23,
"learning_rate": 1e-05,
"loss": 0.6958,
"step": 550
},
{
"epoch": 0.24,
"learning_rate": 1e-05,
"loss": 0.805,
"step": 560
},
{
"epoch": 0.24,
"learning_rate": 1e-05,
"loss": 0.7969,
"step": 570
},
{
"epoch": 0.24,
"learning_rate": 1e-05,
"loss": 0.7798,
"step": 580
},
{
"epoch": 0.25,
"learning_rate": 1e-05,
"loss": 0.7764,
"step": 590
},
{
"epoch": 0.25,
"learning_rate": 1e-05,
"loss": 0.7392,
"step": 600
},
{
"epoch": 0.26,
"learning_rate": 1e-05,
"loss": 0.7266,
"step": 610
},
{
"epoch": 0.26,
"learning_rate": 1e-05,
"loss": 0.7475,
"step": 620
},
{
"epoch": 0.27,
"learning_rate": 1e-05,
"loss": 0.7397,
"step": 630
},
{
"epoch": 0.27,
"learning_rate": 1e-05,
"loss": 0.72,
"step": 640
},
{
"epoch": 0.27,
"learning_rate": 1e-05,
"loss": 0.7729,
"step": 650
},
{
"epoch": 0.28,
"learning_rate": 1e-05,
"loss": 0.7378,
"step": 660
},
{
"epoch": 0.28,
"learning_rate": 1e-05,
"loss": 0.7306,
"step": 670
},
{
"epoch": 0.29,
"learning_rate": 1e-05,
"loss": 0.8145,
"step": 680
},
{
"epoch": 0.29,
"learning_rate": 1e-05,
"loss": 0.69,
"step": 690
},
{
"epoch": 0.3,
"learning_rate": 1e-05,
"loss": 0.7128,
"step": 700
},
{
"epoch": 0.3,
"learning_rate": 1e-05,
"loss": 0.7634,
"step": 710
},
{
"epoch": 0.3,
"learning_rate": 1e-05,
"loss": 0.7451,
"step": 720
},
{
"epoch": 0.31,
"learning_rate": 1e-05,
"loss": 0.7208,
"step": 730
},
{
"epoch": 0.31,
"learning_rate": 1e-05,
"loss": 0.8274,
"step": 740
},
{
"epoch": 0.32,
"learning_rate": 1e-05,
"loss": 0.6745,
"step": 750
},
{
"epoch": 0.32,
"learning_rate": 1e-05,
"loss": 0.6813,
"step": 760
},
{
"epoch": 0.32,
"learning_rate": 1e-05,
"loss": 0.7841,
"step": 770
},
{
"epoch": 0.33,
"learning_rate": 1e-05,
"loss": 0.7459,
"step": 780
},
{
"epoch": 0.33,
"learning_rate": 1e-05,
"loss": 0.7265,
"step": 790
},
{
"epoch": 0.34,
"learning_rate": 1e-05,
"loss": 0.6608,
"step": 800
},
{
"epoch": 0.34,
"learning_rate": 1e-05,
"loss": 0.6852,
"step": 810
},
{
"epoch": 0.35,
"learning_rate": 1e-05,
"loss": 0.6859,
"step": 820
},
{
"epoch": 0.35,
"learning_rate": 1e-05,
"loss": 0.7223,
"step": 830
},
{
"epoch": 0.35,
"learning_rate": 1e-05,
"loss": 0.7669,
"step": 840
},
{
"epoch": 0.36,
"learning_rate": 1e-05,
"loss": 0.7634,
"step": 850
},
{
"epoch": 0.36,
"learning_rate": 1e-05,
"loss": 0.7344,
"step": 860
},
{
"epoch": 0.37,
"learning_rate": 1e-05,
"loss": 0.7292,
"step": 870
},
{
"epoch": 0.37,
"learning_rate": 1e-05,
"loss": 0.7836,
"step": 880
},
{
"epoch": 0.38,
"learning_rate": 1e-05,
"loss": 0.6891,
"step": 890
},
{
"epoch": 0.38,
"learning_rate": 1e-05,
"loss": 0.7374,
"step": 900
},
{
"epoch": 0.38,
"learning_rate": 1e-05,
"loss": 0.749,
"step": 910
},
{
"epoch": 0.39,
"learning_rate": 1e-05,
"loss": 0.7338,
"step": 920
},
{
"epoch": 0.39,
"learning_rate": 1e-05,
"loss": 0.7597,
"step": 930
},
{
"epoch": 0.4,
"learning_rate": 1e-05,
"loss": 0.6914,
"step": 940
},
{
"epoch": 0.4,
"learning_rate": 1e-05,
"loss": 0.6781,
"step": 950
},
{
"epoch": 0.4,
"learning_rate": 1e-05,
"loss": 0.7396,
"step": 960
},
{
"epoch": 0.41,
"learning_rate": 1e-05,
"loss": 0.7296,
"step": 970
},
{
"epoch": 0.41,
"learning_rate": 1e-05,
"loss": 0.7038,
"step": 980
},
{
"epoch": 0.42,
"learning_rate": 1e-05,
"loss": 0.7363,
"step": 990
},
{
"epoch": 0.42,
"learning_rate": 1e-05,
"loss": 0.7154,
"step": 1000
},
{
"epoch": 0.43,
"learning_rate": 1e-05,
"loss": 0.7956,
"step": 1010
},
{
"epoch": 0.43,
"learning_rate": 1e-05,
"loss": 0.7585,
"step": 1020
},
{
"epoch": 0.43,
"learning_rate": 1e-05,
"loss": 0.6719,
"step": 1030
},
{
"epoch": 0.44,
"learning_rate": 1e-05,
"loss": 0.7522,
"step": 1040
},
{
"epoch": 0.44,
"learning_rate": 1e-05,
"loss": 0.6662,
"step": 1050
},
{
"epoch": 0.45,
"learning_rate": 1e-05,
"loss": 0.7183,
"step": 1060
},
{
"epoch": 0.45,
"learning_rate": 1e-05,
"loss": 0.7223,
"step": 1070
},
{
"epoch": 0.46,
"learning_rate": 1e-05,
"loss": 0.7465,
"step": 1080
},
{
"epoch": 0.46,
"learning_rate": 1e-05,
"loss": 0.7484,
"step": 1090
},
{
"epoch": 0.46,
"learning_rate": 1e-05,
"loss": 0.6895,
"step": 1100
},
{
"epoch": 0.47,
"learning_rate": 1e-05,
"loss": 0.6901,
"step": 1110
},
{
"epoch": 0.47,
"learning_rate": 1e-05,
"loss": 0.722,
"step": 1120
},
{
"epoch": 0.48,
"learning_rate": 1e-05,
"loss": 0.7205,
"step": 1130
},
{
"epoch": 0.48,
"learning_rate": 1e-05,
"loss": 0.7023,
"step": 1140
},
{
"epoch": 0.48,
"learning_rate": 1e-05,
"loss": 0.7362,
"step": 1150
},
{
"epoch": 0.49,
"learning_rate": 1e-05,
"loss": 0.7017,
"step": 1160
},
{
"epoch": 0.49,
"learning_rate": 1e-05,
"loss": 0.7371,
"step": 1170
},
{
"epoch": 0.5,
"learning_rate": 1e-05,
"loss": 0.6835,
"step": 1180
},
{
"epoch": 0.5,
"learning_rate": 1e-05,
"loss": 0.6932,
"step": 1190
},
{
"epoch": 0.51,
"learning_rate": 1e-05,
"loss": 0.709,
"step": 1200
},
{
"epoch": 0.51,
"learning_rate": 1e-05,
"loss": 0.7183,
"step": 1210
},
{
"epoch": 0.51,
"learning_rate": 1e-05,
"loss": 0.6728,
"step": 1220
},
{
"epoch": 0.52,
"learning_rate": 1e-05,
"loss": 0.7082,
"step": 1230
},
{
"epoch": 0.52,
"learning_rate": 1e-05,
"loss": 0.7386,
"step": 1240
},
{
"epoch": 0.53,
"learning_rate": 1e-05,
"loss": 0.7759,
"step": 1250
},
{
"epoch": 0.53,
"learning_rate": 1e-05,
"loss": 0.7074,
"step": 1260
},
{
"epoch": 0.54,
"learning_rate": 1e-05,
"loss": 0.7434,
"step": 1270
},
{
"epoch": 0.54,
"learning_rate": 1e-05,
"loss": 0.7182,
"step": 1280
},
{
"epoch": 0.54,
"learning_rate": 1e-05,
"loss": 0.7363,
"step": 1290
},
{
"epoch": 0.55,
"learning_rate": 1e-05,
"loss": 0.7298,
"step": 1300
},
{
"epoch": 0.55,
"learning_rate": 1e-05,
"loss": 0.7573,
"step": 1310
},
{
"epoch": 0.56,
"learning_rate": 1e-05,
"loss": 0.6634,
"step": 1320
},
{
"epoch": 0.56,
"learning_rate": 1e-05,
"loss": 0.7064,
"step": 1330
},
{
"epoch": 0.56,
"learning_rate": 1e-05,
"loss": 0.7203,
"step": 1340
},
{
"epoch": 0.57,
"learning_rate": 1e-05,
"loss": 0.7016,
"step": 1350
},
{
"epoch": 0.57,
"learning_rate": 1e-05,
"loss": 0.7063,
"step": 1360
},
{
"epoch": 0.58,
"learning_rate": 1e-05,
"loss": 0.7035,
"step": 1370
},
{
"epoch": 0.58,
"learning_rate": 1e-05,
"loss": 0.7432,
"step": 1380
},
{
"epoch": 0.59,
"learning_rate": 1e-05,
"loss": 0.6509,
"step": 1390
},
{
"epoch": 0.59,
"learning_rate": 1e-05,
"loss": 0.7226,
"step": 1400
},
{
"epoch": 0.59,
"learning_rate": 1e-05,
"loss": 0.6995,
"step": 1410
},
{
"epoch": 0.6,
"learning_rate": 1e-05,
"loss": 0.6947,
"step": 1420
},
{
"epoch": 0.6,
"learning_rate": 1e-05,
"loss": 0.7199,
"step": 1430
},
{
"epoch": 0.61,
"learning_rate": 1e-05,
"loss": 0.677,
"step": 1440
},
{
"epoch": 0.61,
"learning_rate": 1e-05,
"loss": 0.6938,
"step": 1450
},
{
"epoch": 0.62,
"learning_rate": 1e-05,
"loss": 0.7324,
"step": 1460
},
{
"epoch": 0.62,
"learning_rate": 1e-05,
"loss": 0.6413,
"step": 1470
},
{
"epoch": 0.62,
"learning_rate": 1e-05,
"loss": 0.7271,
"step": 1480
},
{
"epoch": 0.63,
"learning_rate": 1e-05,
"loss": 0.685,
"step": 1490
},
{
"epoch": 0.63,
"learning_rate": 1e-05,
"loss": 0.7034,
"step": 1500
},
{
"epoch": 0.64,
"learning_rate": 1e-05,
"loss": 0.654,
"step": 1510
},
{
"epoch": 0.64,
"learning_rate": 1e-05,
"loss": 0.6833,
"step": 1520
},
{
"epoch": 0.65,
"learning_rate": 1e-05,
"loss": 0.7105,
"step": 1530
},
{
"epoch": 0.65,
"learning_rate": 1e-05,
"loss": 0.7379,
"step": 1540
},
{
"epoch": 0.65,
"learning_rate": 1e-05,
"loss": 0.7178,
"step": 1550
},
{
"epoch": 0.66,
"learning_rate": 1e-05,
"loss": 0.7208,
"step": 1560
},
{
"epoch": 0.66,
"learning_rate": 1e-05,
"loss": 0.7744,
"step": 1570
},
{
"epoch": 0.67,
"learning_rate": 1e-05,
"loss": 0.6935,
"step": 1580
},
{
"epoch": 0.67,
"learning_rate": 1e-05,
"loss": 0.6516,
"step": 1590
},
{
"epoch": 0.67,
"learning_rate": 1e-05,
"loss": 0.6761,
"step": 1600
},
{
"epoch": 0.68,
"learning_rate": 1e-05,
"loss": 0.7512,
"step": 1610
},
{
"epoch": 0.68,
"learning_rate": 1e-05,
"loss": 0.6614,
"step": 1620
},
{
"epoch": 0.69,
"learning_rate": 1e-05,
"loss": 0.6641,
"step": 1630
},
{
"epoch": 0.69,
"learning_rate": 1e-05,
"loss": 0.6886,
"step": 1640
},
{
"epoch": 0.7,
"learning_rate": 1e-05,
"loss": 0.7974,
"step": 1650
},
{
"epoch": 0.7,
"learning_rate": 1e-05,
"loss": 0.677,
"step": 1660
},
{
"epoch": 0.7,
"learning_rate": 1e-05,
"loss": 0.6576,
"step": 1670
},
{
"epoch": 0.71,
"learning_rate": 1e-05,
"loss": 0.6805,
"step": 1680
},
{
"epoch": 0.71,
"learning_rate": 1e-05,
"loss": 0.7206,
"step": 1690
},
{
"epoch": 0.72,
"learning_rate": 1e-05,
"loss": 0.7076,
"step": 1700
},
{
"epoch": 0.72,
"learning_rate": 1e-05,
"loss": 0.716,
"step": 1710
},
{
"epoch": 0.73,
"learning_rate": 1e-05,
"loss": 0.6832,
"step": 1720
},
{
"epoch": 0.73,
"learning_rate": 1e-05,
"loss": 0.7273,
"step": 1730
},
{
"epoch": 0.73,
"learning_rate": 1e-05,
"loss": 0.6822,
"step": 1740
},
{
"epoch": 0.74,
"learning_rate": 1e-05,
"loss": 0.7383,
"step": 1750
},
{
"epoch": 0.74,
"learning_rate": 1e-05,
"loss": 0.7376,
"step": 1760
},
{
"epoch": 0.75,
"learning_rate": 1e-05,
"loss": 0.7059,
"step": 1770
},
{
"epoch": 0.75,
"learning_rate": 1e-05,
"loss": 0.6817,
"step": 1780
},
{
"epoch": 0.75,
"learning_rate": 1e-05,
"loss": 0.691,
"step": 1790
},
{
"epoch": 0.76,
"learning_rate": 1e-05,
"loss": 0.7114,
"step": 1800
},
{
"epoch": 0.76,
"learning_rate": 1e-05,
"loss": 0.6577,
"step": 1810
},
{
"epoch": 0.77,
"learning_rate": 1e-05,
"loss": 0.6957,
"step": 1820
},
{
"epoch": 0.77,
"learning_rate": 1e-05,
"loss": 0.7366,
"step": 1830
},
{
"epoch": 0.78,
"learning_rate": 1e-05,
"loss": 0.7042,
"step": 1840
},
{
"epoch": 0.78,
"learning_rate": 1e-05,
"loss": 0.6852,
"step": 1850
},
{
"epoch": 0.78,
"learning_rate": 1e-05,
"loss": 0.7562,
"step": 1860
},
{
"epoch": 0.79,
"learning_rate": 1e-05,
"loss": 0.7761,
"step": 1870
},
{
"epoch": 0.79,
"learning_rate": 1e-05,
"loss": 0.7314,
"step": 1880
},
{
"epoch": 0.8,
"learning_rate": 1e-05,
"loss": 0.6932,
"step": 1890
},
{
"epoch": 0.8,
"learning_rate": 1e-05,
"loss": 0.7072,
"step": 1900
},
{
"epoch": 0.81,
"learning_rate": 1e-05,
"loss": 0.7314,
"step": 1910
},
{
"epoch": 0.81,
"learning_rate": 1e-05,
"loss": 0.6293,
"step": 1920
},
{
"epoch": 0.81,
"learning_rate": 1e-05,
"loss": 0.6994,
"step": 1930
},
{
"epoch": 0.82,
"learning_rate": 1e-05,
"loss": 0.6681,
"step": 1940
},
{
"epoch": 0.82,
"learning_rate": 1e-05,
"loss": 0.6923,
"step": 1950
},
{
"epoch": 0.83,
"learning_rate": 1e-05,
"loss": 0.7328,
"step": 1960
},
{
"epoch": 0.83,
"learning_rate": 1e-05,
"loss": 0.6874,
"step": 1970
},
{
"epoch": 0.83,
"learning_rate": 1e-05,
"loss": 0.682,
"step": 1980
},
{
"epoch": 0.84,
"learning_rate": 1e-05,
"loss": 0.6647,
"step": 1990
},
{
"epoch": 0.84,
"learning_rate": 1e-05,
"loss": 0.6565,
"step": 2000
}
],
"logging_steps": 10,
"max_steps": 5000,
"num_train_epochs": 3,
"save_steps": 500,
"total_flos": 502812786032640.0,
"trial_name": null,
"trial_params": null
}