Llama-2-8k-2m-rethink / trainer_state.json
ccore's picture
Upload folder using huggingface_hub
93f1edb
raw
history blame contribute delete
No virus
32.4 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1528,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 0.001,
"loss": 10.3529,
"step": 5
},
{
"epoch": 0.01,
"learning_rate": 0.001,
"loss": 10.0747,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 0.001,
"loss": 9.7304,
"step": 15
},
{
"epoch": 0.01,
"learning_rate": 0.001,
"loss": 9.4139,
"step": 20
},
{
"epoch": 0.02,
"learning_rate": 0.001,
"loss": 9.0418,
"step": 25
},
{
"epoch": 0.02,
"learning_rate": 0.001,
"loss": 8.7655,
"step": 30
},
{
"epoch": 0.02,
"learning_rate": 0.001,
"loss": 8.3599,
"step": 35
},
{
"epoch": 0.03,
"learning_rate": 0.001,
"loss": 8.0603,
"step": 40
},
{
"epoch": 0.03,
"learning_rate": 0.001,
"loss": 7.7776,
"step": 45
},
{
"epoch": 0.03,
"learning_rate": 0.001,
"loss": 7.5678,
"step": 50
},
{
"epoch": 0.04,
"learning_rate": 0.001,
"loss": 7.5338,
"step": 55
},
{
"epoch": 0.04,
"learning_rate": 0.001,
"loss": 7.2249,
"step": 60
},
{
"epoch": 0.04,
"learning_rate": 0.001,
"loss": 7.0286,
"step": 65
},
{
"epoch": 0.05,
"learning_rate": 0.001,
"loss": 6.9518,
"step": 70
},
{
"epoch": 0.05,
"learning_rate": 0.001,
"loss": 6.7623,
"step": 75
},
{
"epoch": 0.05,
"learning_rate": 0.001,
"loss": 6.7361,
"step": 80
},
{
"epoch": 0.06,
"learning_rate": 0.001,
"loss": 6.9201,
"step": 85
},
{
"epoch": 0.06,
"learning_rate": 0.001,
"loss": 6.8491,
"step": 90
},
{
"epoch": 0.06,
"learning_rate": 0.001,
"loss": 6.7565,
"step": 95
},
{
"epoch": 0.07,
"learning_rate": 0.001,
"loss": 6.6537,
"step": 100
},
{
"epoch": 0.07,
"learning_rate": 0.001,
"loss": 6.8522,
"step": 105
},
{
"epoch": 0.07,
"learning_rate": 0.001,
"loss": 6.6052,
"step": 110
},
{
"epoch": 0.08,
"learning_rate": 0.001,
"loss": 6.4885,
"step": 115
},
{
"epoch": 0.08,
"learning_rate": 0.001,
"loss": 6.5162,
"step": 120
},
{
"epoch": 0.08,
"learning_rate": 0.001,
"loss": 6.4759,
"step": 125
},
{
"epoch": 0.09,
"learning_rate": 0.001,
"loss": 6.3678,
"step": 130
},
{
"epoch": 0.09,
"learning_rate": 0.001,
"loss": 6.6435,
"step": 135
},
{
"epoch": 0.09,
"learning_rate": 0.001,
"loss": 6.4821,
"step": 140
},
{
"epoch": 0.09,
"learning_rate": 0.001,
"loss": 6.3594,
"step": 145
},
{
"epoch": 0.1,
"learning_rate": 0.001,
"loss": 6.1634,
"step": 150
},
{
"epoch": 0.1,
"learning_rate": 0.001,
"loss": 6.377,
"step": 155
},
{
"epoch": 0.1,
"learning_rate": 0.001,
"loss": 6.368,
"step": 160
},
{
"epoch": 0.11,
"learning_rate": 0.001,
"loss": 6.1571,
"step": 165
},
{
"epoch": 0.11,
"learning_rate": 0.001,
"loss": 6.122,
"step": 170
},
{
"epoch": 0.11,
"learning_rate": 0.001,
"loss": 6.2277,
"step": 175
},
{
"epoch": 0.12,
"learning_rate": 0.001,
"loss": 5.9614,
"step": 180
},
{
"epoch": 0.12,
"learning_rate": 0.001,
"loss": 5.9394,
"step": 185
},
{
"epoch": 0.12,
"learning_rate": 0.001,
"loss": 5.8261,
"step": 190
},
{
"epoch": 0.13,
"learning_rate": 0.001,
"loss": 6.2232,
"step": 195
},
{
"epoch": 0.13,
"learning_rate": 0.001,
"loss": 6.0747,
"step": 200
},
{
"epoch": 0.13,
"learning_rate": 0.001,
"loss": 5.9842,
"step": 205
},
{
"epoch": 0.14,
"learning_rate": 0.001,
"loss": 6.2139,
"step": 210
},
{
"epoch": 0.14,
"learning_rate": 0.001,
"loss": 6.0711,
"step": 215
},
{
"epoch": 0.14,
"learning_rate": 0.001,
"loss": 5.9545,
"step": 220
},
{
"epoch": 0.15,
"learning_rate": 0.001,
"loss": 5.7681,
"step": 225
},
{
"epoch": 0.15,
"learning_rate": 0.001,
"loss": 5.8673,
"step": 230
},
{
"epoch": 0.15,
"learning_rate": 0.001,
"loss": 5.6984,
"step": 235
},
{
"epoch": 0.16,
"learning_rate": 0.001,
"loss": 5.6021,
"step": 240
},
{
"epoch": 0.16,
"learning_rate": 0.001,
"loss": 5.8764,
"step": 245
},
{
"epoch": 0.16,
"learning_rate": 0.001,
"loss": 5.6526,
"step": 250
},
{
"epoch": 0.17,
"learning_rate": 0.001,
"loss": 5.7407,
"step": 255
},
{
"epoch": 0.17,
"learning_rate": 0.001,
"loss": 5.6711,
"step": 260
},
{
"epoch": 0.17,
"learning_rate": 0.001,
"loss": 5.6802,
"step": 265
},
{
"epoch": 0.18,
"learning_rate": 0.001,
"loss": 5.7385,
"step": 270
},
{
"epoch": 0.18,
"learning_rate": 0.001,
"loss": 5.8663,
"step": 275
},
{
"epoch": 0.18,
"learning_rate": 0.001,
"loss": 5.5553,
"step": 280
},
{
"epoch": 0.19,
"learning_rate": 0.001,
"loss": 5.6528,
"step": 285
},
{
"epoch": 0.19,
"learning_rate": 0.001,
"loss": 5.8758,
"step": 290
},
{
"epoch": 0.19,
"learning_rate": 0.001,
"loss": 5.513,
"step": 295
},
{
"epoch": 0.2,
"learning_rate": 0.001,
"loss": 5.3608,
"step": 300
},
{
"epoch": 0.2,
"learning_rate": 0.001,
"loss": 5.7339,
"step": 305
},
{
"epoch": 0.2,
"learning_rate": 0.001,
"loss": 5.5918,
"step": 310
},
{
"epoch": 0.21,
"learning_rate": 0.001,
"loss": 5.7015,
"step": 315
},
{
"epoch": 0.21,
"learning_rate": 0.001,
"loss": 5.5825,
"step": 320
},
{
"epoch": 0.21,
"learning_rate": 0.001,
"loss": 5.3423,
"step": 325
},
{
"epoch": 0.22,
"learning_rate": 0.001,
"loss": 5.2627,
"step": 330
},
{
"epoch": 0.22,
"learning_rate": 0.001,
"loss": 5.4692,
"step": 335
},
{
"epoch": 0.22,
"learning_rate": 0.001,
"loss": 5.474,
"step": 340
},
{
"epoch": 0.23,
"learning_rate": 0.001,
"loss": 5.3762,
"step": 345
},
{
"epoch": 0.23,
"learning_rate": 0.001,
"loss": 5.2908,
"step": 350
},
{
"epoch": 0.23,
"learning_rate": 0.001,
"loss": 5.4183,
"step": 355
},
{
"epoch": 0.24,
"learning_rate": 0.001,
"loss": 5.2073,
"step": 360
},
{
"epoch": 0.24,
"learning_rate": 0.001,
"loss": 5.4598,
"step": 365
},
{
"epoch": 0.24,
"learning_rate": 0.001,
"loss": 5.4487,
"step": 370
},
{
"epoch": 0.25,
"learning_rate": 0.001,
"loss": 5.2604,
"step": 375
},
{
"epoch": 0.25,
"learning_rate": 0.001,
"loss": 5.1575,
"step": 380
},
{
"epoch": 0.25,
"learning_rate": 0.001,
"loss": 5.2809,
"step": 385
},
{
"epoch": 0.26,
"learning_rate": 0.001,
"loss": 5.1923,
"step": 390
},
{
"epoch": 0.26,
"learning_rate": 0.001,
"loss": 5.3494,
"step": 395
},
{
"epoch": 0.26,
"learning_rate": 0.001,
"loss": 5.2338,
"step": 400
},
{
"epoch": 0.27,
"learning_rate": 0.001,
"loss": 5.2143,
"step": 405
},
{
"epoch": 0.27,
"learning_rate": 0.001,
"loss": 5.3615,
"step": 410
},
{
"epoch": 0.27,
"learning_rate": 0.001,
"loss": 5.4357,
"step": 415
},
{
"epoch": 0.27,
"learning_rate": 0.001,
"loss": 5.1693,
"step": 420
},
{
"epoch": 0.28,
"learning_rate": 0.001,
"loss": 5.1072,
"step": 425
},
{
"epoch": 0.28,
"learning_rate": 0.001,
"loss": 5.2298,
"step": 430
},
{
"epoch": 0.28,
"learning_rate": 0.001,
"loss": 4.9306,
"step": 435
},
{
"epoch": 0.29,
"learning_rate": 0.001,
"loss": 5.321,
"step": 440
},
{
"epoch": 0.29,
"learning_rate": 0.001,
"loss": 4.9341,
"step": 445
},
{
"epoch": 0.29,
"learning_rate": 0.001,
"loss": 5.0898,
"step": 450
},
{
"epoch": 0.3,
"learning_rate": 0.001,
"loss": 5.1687,
"step": 455
},
{
"epoch": 0.3,
"learning_rate": 0.001,
"loss": 4.8764,
"step": 460
},
{
"epoch": 0.3,
"learning_rate": 0.001,
"loss": 5.1905,
"step": 465
},
{
"epoch": 0.31,
"learning_rate": 0.001,
"loss": 5.0816,
"step": 470
},
{
"epoch": 0.31,
"learning_rate": 0.001,
"loss": 5.2568,
"step": 475
},
{
"epoch": 0.31,
"learning_rate": 0.001,
"loss": 5.2742,
"step": 480
},
{
"epoch": 0.32,
"learning_rate": 0.001,
"loss": 5.0711,
"step": 485
},
{
"epoch": 0.32,
"learning_rate": 0.001,
"loss": 5.0671,
"step": 490
},
{
"epoch": 0.32,
"learning_rate": 0.001,
"loss": 5.1016,
"step": 495
},
{
"epoch": 0.33,
"learning_rate": 0.001,
"loss": 5.3311,
"step": 500
},
{
"epoch": 0.33,
"learning_rate": 0.001,
"loss": 5.2327,
"step": 505
},
{
"epoch": 0.33,
"learning_rate": 0.001,
"loss": 5.3222,
"step": 510
},
{
"epoch": 0.34,
"learning_rate": 0.001,
"loss": 4.9495,
"step": 515
},
{
"epoch": 0.34,
"learning_rate": 0.001,
"loss": 4.957,
"step": 520
},
{
"epoch": 0.34,
"learning_rate": 0.001,
"loss": 5.3251,
"step": 525
},
{
"epoch": 0.35,
"learning_rate": 0.001,
"loss": 4.6873,
"step": 530
},
{
"epoch": 0.35,
"learning_rate": 0.001,
"loss": 5.289,
"step": 535
},
{
"epoch": 0.35,
"learning_rate": 0.001,
"loss": 5.0009,
"step": 540
},
{
"epoch": 0.36,
"learning_rate": 0.001,
"loss": 4.8415,
"step": 545
},
{
"epoch": 0.36,
"learning_rate": 0.001,
"loss": 5.2063,
"step": 550
},
{
"epoch": 0.36,
"learning_rate": 0.001,
"loss": 4.8384,
"step": 555
},
{
"epoch": 0.37,
"learning_rate": 0.001,
"loss": 4.3815,
"step": 560
},
{
"epoch": 0.37,
"learning_rate": 0.001,
"loss": 4.9327,
"step": 565
},
{
"epoch": 0.37,
"learning_rate": 0.001,
"loss": 4.8049,
"step": 570
},
{
"epoch": 0.38,
"learning_rate": 0.001,
"loss": 4.9983,
"step": 575
},
{
"epoch": 0.38,
"learning_rate": 0.001,
"loss": 4.9048,
"step": 580
},
{
"epoch": 0.38,
"learning_rate": 0.001,
"loss": 4.8615,
"step": 585
},
{
"epoch": 0.39,
"learning_rate": 0.001,
"loss": 4.8885,
"step": 590
},
{
"epoch": 0.39,
"learning_rate": 0.001,
"loss": 4.9189,
"step": 595
},
{
"epoch": 0.39,
"learning_rate": 0.001,
"loss": 4.537,
"step": 600
},
{
"epoch": 0.4,
"learning_rate": 0.001,
"loss": 4.5421,
"step": 605
},
{
"epoch": 0.4,
"learning_rate": 0.001,
"loss": 4.8229,
"step": 610
},
{
"epoch": 0.4,
"learning_rate": 0.001,
"loss": 4.9116,
"step": 615
},
{
"epoch": 0.41,
"learning_rate": 0.001,
"loss": 4.9846,
"step": 620
},
{
"epoch": 0.41,
"learning_rate": 0.001,
"loss": 4.7512,
"step": 625
},
{
"epoch": 0.41,
"learning_rate": 0.001,
"loss": 4.9718,
"step": 630
},
{
"epoch": 0.42,
"learning_rate": 0.001,
"loss": 4.9162,
"step": 635
},
{
"epoch": 0.42,
"learning_rate": 0.001,
"loss": 4.5091,
"step": 640
},
{
"epoch": 0.42,
"learning_rate": 0.001,
"loss": 4.8638,
"step": 645
},
{
"epoch": 0.43,
"learning_rate": 0.001,
"loss": 4.5735,
"step": 650
},
{
"epoch": 0.43,
"learning_rate": 0.001,
"loss": 4.8966,
"step": 655
},
{
"epoch": 0.43,
"learning_rate": 0.001,
"loss": 4.9169,
"step": 660
},
{
"epoch": 0.44,
"learning_rate": 0.001,
"loss": 4.8778,
"step": 665
},
{
"epoch": 0.44,
"learning_rate": 0.001,
"loss": 4.8185,
"step": 670
},
{
"epoch": 0.44,
"learning_rate": 0.001,
"loss": 4.6256,
"step": 675
},
{
"epoch": 0.45,
"learning_rate": 0.001,
"loss": 4.7773,
"step": 680
},
{
"epoch": 0.45,
"learning_rate": 0.001,
"loss": 4.6482,
"step": 685
},
{
"epoch": 0.45,
"learning_rate": 0.001,
"loss": 4.3789,
"step": 690
},
{
"epoch": 0.45,
"learning_rate": 0.001,
"loss": 4.849,
"step": 695
},
{
"epoch": 0.46,
"learning_rate": 0.001,
"loss": 4.7127,
"step": 700
},
{
"epoch": 0.46,
"learning_rate": 0.001,
"loss": 4.6414,
"step": 705
},
{
"epoch": 0.46,
"learning_rate": 0.001,
"loss": 4.512,
"step": 710
},
{
"epoch": 0.47,
"learning_rate": 0.001,
"loss": 4.5855,
"step": 715
},
{
"epoch": 0.47,
"learning_rate": 0.001,
"loss": 4.5004,
"step": 720
},
{
"epoch": 0.47,
"learning_rate": 0.001,
"loss": 4.8402,
"step": 725
},
{
"epoch": 0.48,
"learning_rate": 0.001,
"loss": 4.7665,
"step": 730
},
{
"epoch": 0.48,
"learning_rate": 0.001,
"loss": 4.7962,
"step": 735
},
{
"epoch": 0.48,
"learning_rate": 0.001,
"loss": 4.7931,
"step": 740
},
{
"epoch": 0.49,
"learning_rate": 0.001,
"loss": 4.7293,
"step": 745
},
{
"epoch": 0.49,
"learning_rate": 0.001,
"loss": 4.6017,
"step": 750
},
{
"epoch": 0.49,
"learning_rate": 0.001,
"loss": 4.7313,
"step": 755
},
{
"epoch": 0.5,
"learning_rate": 0.001,
"loss": 4.58,
"step": 760
},
{
"epoch": 0.5,
"learning_rate": 0.001,
"loss": 4.6188,
"step": 765
},
{
"epoch": 0.5,
"learning_rate": 0.001,
"loss": 4.5284,
"step": 770
},
{
"epoch": 0.51,
"learning_rate": 0.001,
"loss": 4.5003,
"step": 775
},
{
"epoch": 0.51,
"learning_rate": 0.001,
"loss": 4.5868,
"step": 780
},
{
"epoch": 0.51,
"learning_rate": 0.001,
"loss": 4.5261,
"step": 785
},
{
"epoch": 0.52,
"learning_rate": 0.001,
"loss": 4.6488,
"step": 790
},
{
"epoch": 0.52,
"learning_rate": 0.001,
"loss": 4.2441,
"step": 795
},
{
"epoch": 0.52,
"learning_rate": 0.001,
"loss": 4.9201,
"step": 800
},
{
"epoch": 0.53,
"learning_rate": 0.001,
"loss": 4.5616,
"step": 805
},
{
"epoch": 0.53,
"learning_rate": 0.001,
"loss": 4.4575,
"step": 810
},
{
"epoch": 0.53,
"learning_rate": 0.001,
"loss": 4.4801,
"step": 815
},
{
"epoch": 0.54,
"learning_rate": 0.001,
"loss": 4.4761,
"step": 820
},
{
"epoch": 0.54,
"learning_rate": 0.001,
"loss": 4.7145,
"step": 825
},
{
"epoch": 0.54,
"learning_rate": 0.001,
"loss": 4.6759,
"step": 830
},
{
"epoch": 0.55,
"learning_rate": 0.001,
"loss": 4.3177,
"step": 835
},
{
"epoch": 0.55,
"learning_rate": 0.001,
"loss": 4.4118,
"step": 840
},
{
"epoch": 0.55,
"learning_rate": 0.001,
"loss": 4.4915,
"step": 845
},
{
"epoch": 0.56,
"learning_rate": 0.001,
"loss": 4.6066,
"step": 850
},
{
"epoch": 0.56,
"learning_rate": 0.001,
"loss": 4.6189,
"step": 855
},
{
"epoch": 0.56,
"learning_rate": 0.001,
"loss": 4.4977,
"step": 860
},
{
"epoch": 0.57,
"learning_rate": 0.001,
"loss": 4.2084,
"step": 865
},
{
"epoch": 0.57,
"learning_rate": 0.001,
"loss": 4.4593,
"step": 870
},
{
"epoch": 0.57,
"learning_rate": 0.001,
"loss": 4.4324,
"step": 875
},
{
"epoch": 0.58,
"learning_rate": 0.001,
"loss": 4.5696,
"step": 880
},
{
"epoch": 0.58,
"learning_rate": 0.001,
"loss": 4.2641,
"step": 885
},
{
"epoch": 0.58,
"learning_rate": 0.001,
"loss": 4.2248,
"step": 890
},
{
"epoch": 0.59,
"learning_rate": 0.001,
"loss": 4.577,
"step": 895
},
{
"epoch": 0.59,
"learning_rate": 0.001,
"loss": 4.3748,
"step": 900
},
{
"epoch": 0.59,
"learning_rate": 0.001,
"loss": 4.4647,
"step": 905
},
{
"epoch": 0.6,
"learning_rate": 0.001,
"loss": 4.3836,
"step": 910
},
{
"epoch": 0.6,
"learning_rate": 0.001,
"loss": 4.7384,
"step": 915
},
{
"epoch": 0.6,
"learning_rate": 0.001,
"loss": 4.3023,
"step": 920
},
{
"epoch": 0.61,
"learning_rate": 0.001,
"loss": 4.2472,
"step": 925
},
{
"epoch": 0.61,
"learning_rate": 0.001,
"loss": 4.6622,
"step": 930
},
{
"epoch": 0.61,
"learning_rate": 0.001,
"loss": 4.5409,
"step": 935
},
{
"epoch": 0.62,
"learning_rate": 0.001,
"loss": 3.9355,
"step": 940
},
{
"epoch": 0.62,
"learning_rate": 0.001,
"loss": 4.4457,
"step": 945
},
{
"epoch": 0.62,
"learning_rate": 0.001,
"loss": 4.4636,
"step": 950
},
{
"epoch": 0.62,
"learning_rate": 0.001,
"loss": 3.7582,
"step": 955
},
{
"epoch": 0.63,
"learning_rate": 0.001,
"loss": 4.4637,
"step": 960
},
{
"epoch": 0.63,
"learning_rate": 0.001,
"loss": 4.4201,
"step": 965
},
{
"epoch": 0.63,
"learning_rate": 0.001,
"loss": 3.9001,
"step": 970
},
{
"epoch": 0.64,
"learning_rate": 0.001,
"loss": 4.1943,
"step": 975
},
{
"epoch": 0.64,
"learning_rate": 0.001,
"loss": 4.1346,
"step": 980
},
{
"epoch": 0.64,
"learning_rate": 0.001,
"loss": 4.057,
"step": 985
},
{
"epoch": 0.65,
"learning_rate": 0.001,
"loss": 4.1786,
"step": 990
},
{
"epoch": 0.65,
"learning_rate": 0.001,
"loss": 4.0282,
"step": 995
},
{
"epoch": 0.65,
"learning_rate": 0.001,
"loss": 4.2299,
"step": 1000
},
{
"epoch": 0.66,
"learning_rate": 0.001,
"loss": 4.4361,
"step": 1005
},
{
"epoch": 0.66,
"learning_rate": 0.001,
"loss": 4.5289,
"step": 1010
},
{
"epoch": 0.66,
"learning_rate": 0.001,
"loss": 3.9963,
"step": 1015
},
{
"epoch": 0.67,
"learning_rate": 0.001,
"loss": 4.295,
"step": 1020
},
{
"epoch": 0.67,
"learning_rate": 0.001,
"loss": 4.2905,
"step": 1025
},
{
"epoch": 0.67,
"learning_rate": 0.001,
"loss": 4.3483,
"step": 1030
},
{
"epoch": 0.68,
"learning_rate": 0.001,
"loss": 4.43,
"step": 1035
},
{
"epoch": 0.68,
"learning_rate": 0.001,
"loss": 4.2837,
"step": 1040
},
{
"epoch": 0.68,
"learning_rate": 0.001,
"loss": 4.4342,
"step": 1045
},
{
"epoch": 0.69,
"learning_rate": 0.001,
"loss": 4.091,
"step": 1050
},
{
"epoch": 0.69,
"learning_rate": 0.001,
"loss": 4.2579,
"step": 1055
},
{
"epoch": 0.69,
"learning_rate": 0.001,
"loss": 4.3192,
"step": 1060
},
{
"epoch": 0.7,
"learning_rate": 0.001,
"loss": 4.3401,
"step": 1065
},
{
"epoch": 0.7,
"learning_rate": 0.001,
"loss": 4.2295,
"step": 1070
},
{
"epoch": 0.7,
"learning_rate": 0.001,
"loss": 4.4993,
"step": 1075
},
{
"epoch": 0.71,
"learning_rate": 0.001,
"loss": 4.0614,
"step": 1080
},
{
"epoch": 0.71,
"learning_rate": 0.001,
"loss": 4.1098,
"step": 1085
},
{
"epoch": 0.71,
"learning_rate": 0.001,
"loss": 3.8367,
"step": 1090
},
{
"epoch": 0.72,
"learning_rate": 0.001,
"loss": 4.1385,
"step": 1095
},
{
"epoch": 0.72,
"learning_rate": 0.001,
"loss": 4.4547,
"step": 1100
},
{
"epoch": 0.72,
"learning_rate": 0.001,
"loss": 4.366,
"step": 1105
},
{
"epoch": 0.73,
"learning_rate": 0.001,
"loss": 4.1179,
"step": 1110
},
{
"epoch": 0.73,
"learning_rate": 0.001,
"loss": 4.3339,
"step": 1115
},
{
"epoch": 0.73,
"learning_rate": 0.001,
"loss": 4.1327,
"step": 1120
},
{
"epoch": 0.74,
"learning_rate": 0.001,
"loss": 4.3913,
"step": 1125
},
{
"epoch": 0.74,
"learning_rate": 0.001,
"loss": 4.3716,
"step": 1130
},
{
"epoch": 0.74,
"learning_rate": 0.001,
"loss": 4.5199,
"step": 1135
},
{
"epoch": 0.75,
"learning_rate": 0.001,
"loss": 4.5701,
"step": 1140
},
{
"epoch": 0.75,
"learning_rate": 0.001,
"loss": 4.0489,
"step": 1145
},
{
"epoch": 0.75,
"learning_rate": 0.001,
"loss": 4.0723,
"step": 1150
},
{
"epoch": 0.76,
"learning_rate": 0.001,
"loss": 4.2067,
"step": 1155
},
{
"epoch": 0.76,
"learning_rate": 0.001,
"loss": 4.082,
"step": 1160
},
{
"epoch": 0.76,
"learning_rate": 0.001,
"loss": 3.9242,
"step": 1165
},
{
"epoch": 0.77,
"learning_rate": 0.001,
"loss": 4.4565,
"step": 1170
},
{
"epoch": 0.77,
"learning_rate": 0.001,
"loss": 4.2539,
"step": 1175
},
{
"epoch": 0.77,
"learning_rate": 0.001,
"loss": 4.1224,
"step": 1180
},
{
"epoch": 0.78,
"learning_rate": 0.001,
"loss": 4.1788,
"step": 1185
},
{
"epoch": 0.78,
"learning_rate": 0.001,
"loss": 4.3302,
"step": 1190
},
{
"epoch": 0.78,
"learning_rate": 0.001,
"loss": 3.599,
"step": 1195
},
{
"epoch": 0.79,
"learning_rate": 0.001,
"loss": 4.1497,
"step": 1200
},
{
"epoch": 0.79,
"learning_rate": 0.001,
"loss": 4.4196,
"step": 1205
},
{
"epoch": 0.79,
"learning_rate": 0.001,
"loss": 4.1635,
"step": 1210
},
{
"epoch": 0.8,
"learning_rate": 0.001,
"loss": 3.43,
"step": 1215
},
{
"epoch": 0.8,
"learning_rate": 0.001,
"loss": 4.0208,
"step": 1220
},
{
"epoch": 0.8,
"learning_rate": 0.001,
"loss": 4.1878,
"step": 1225
},
{
"epoch": 0.8,
"learning_rate": 0.001,
"loss": 4.3068,
"step": 1230
},
{
"epoch": 0.81,
"learning_rate": 0.001,
"loss": 4.1217,
"step": 1235
},
{
"epoch": 0.81,
"learning_rate": 0.001,
"loss": 4.1326,
"step": 1240
},
{
"epoch": 0.81,
"learning_rate": 0.001,
"loss": 4.0908,
"step": 1245
},
{
"epoch": 0.82,
"learning_rate": 0.001,
"loss": 4.5034,
"step": 1250
},
{
"epoch": 0.82,
"learning_rate": 0.001,
"loss": 3.8997,
"step": 1255
},
{
"epoch": 0.82,
"learning_rate": 0.001,
"loss": 3.6657,
"step": 1260
},
{
"epoch": 0.83,
"learning_rate": 0.001,
"loss": 4.0684,
"step": 1265
},
{
"epoch": 0.83,
"learning_rate": 0.001,
"loss": 3.8496,
"step": 1270
},
{
"epoch": 0.83,
"learning_rate": 0.001,
"loss": 4.2138,
"step": 1275
},
{
"epoch": 0.84,
"learning_rate": 0.001,
"loss": 3.8833,
"step": 1280
},
{
"epoch": 0.84,
"learning_rate": 0.001,
"loss": 4.0407,
"step": 1285
},
{
"epoch": 0.84,
"learning_rate": 0.001,
"loss": 4.0523,
"step": 1290
},
{
"epoch": 0.85,
"learning_rate": 0.001,
"loss": 4.1253,
"step": 1295
},
{
"epoch": 0.85,
"learning_rate": 0.001,
"loss": 4.1417,
"step": 1300
},
{
"epoch": 0.85,
"learning_rate": 0.001,
"loss": 3.6037,
"step": 1305
},
{
"epoch": 0.86,
"learning_rate": 0.001,
"loss": 4.0606,
"step": 1310
},
{
"epoch": 0.86,
"learning_rate": 0.001,
"loss": 3.778,
"step": 1315
},
{
"epoch": 0.86,
"learning_rate": 0.001,
"loss": 4.2186,
"step": 1320
},
{
"epoch": 0.87,
"learning_rate": 0.001,
"loss": 4.3195,
"step": 1325
},
{
"epoch": 0.87,
"learning_rate": 0.001,
"loss": 3.8675,
"step": 1330
},
{
"epoch": 0.87,
"learning_rate": 0.001,
"loss": 4.1566,
"step": 1335
},
{
"epoch": 0.88,
"learning_rate": 0.001,
"loss": 3.9631,
"step": 1340
},
{
"epoch": 0.88,
"learning_rate": 0.001,
"loss": 4.3259,
"step": 1345
},
{
"epoch": 0.88,
"learning_rate": 0.001,
"loss": 3.5373,
"step": 1350
},
{
"epoch": 0.89,
"learning_rate": 0.001,
"loss": 4.205,
"step": 1355
},
{
"epoch": 0.89,
"learning_rate": 0.001,
"loss": 4.2473,
"step": 1360
},
{
"epoch": 0.89,
"learning_rate": 0.001,
"loss": 3.954,
"step": 1365
},
{
"epoch": 0.9,
"learning_rate": 0.001,
"loss": 3.839,
"step": 1370
},
{
"epoch": 0.9,
"learning_rate": 0.001,
"loss": 3.7552,
"step": 1375
},
{
"epoch": 0.9,
"learning_rate": 0.001,
"loss": 3.9832,
"step": 1380
},
{
"epoch": 0.91,
"learning_rate": 0.001,
"loss": 4.249,
"step": 1385
},
{
"epoch": 0.91,
"learning_rate": 0.001,
"loss": 3.7355,
"step": 1390
},
{
"epoch": 0.91,
"learning_rate": 0.001,
"loss": 3.6361,
"step": 1395
},
{
"epoch": 0.92,
"learning_rate": 0.001,
"loss": 4.0132,
"step": 1400
},
{
"epoch": 0.92,
"learning_rate": 0.001,
"loss": 4.1217,
"step": 1405
},
{
"epoch": 0.92,
"learning_rate": 0.001,
"loss": 3.7764,
"step": 1410
},
{
"epoch": 0.93,
"learning_rate": 0.001,
"loss": 4.0255,
"step": 1415
},
{
"epoch": 0.93,
"learning_rate": 0.001,
"loss": 4.0114,
"step": 1420
},
{
"epoch": 0.93,
"learning_rate": 0.001,
"loss": 4.2254,
"step": 1425
},
{
"epoch": 0.94,
"learning_rate": 0.001,
"loss": 3.6828,
"step": 1430
},
{
"epoch": 0.94,
"learning_rate": 0.001,
"loss": 4.2974,
"step": 1435
},
{
"epoch": 0.94,
"learning_rate": 0.001,
"loss": 3.5061,
"step": 1440
},
{
"epoch": 0.95,
"learning_rate": 0.001,
"loss": 4.1774,
"step": 1445
},
{
"epoch": 0.95,
"learning_rate": 0.001,
"loss": 3.7006,
"step": 1450
},
{
"epoch": 0.95,
"learning_rate": 0.001,
"loss": 3.7333,
"step": 1455
},
{
"epoch": 0.96,
"learning_rate": 0.001,
"loss": 3.9734,
"step": 1460
},
{
"epoch": 0.96,
"learning_rate": 0.001,
"loss": 4.1223,
"step": 1465
},
{
"epoch": 0.96,
"learning_rate": 0.001,
"loss": 3.7076,
"step": 1470
},
{
"epoch": 0.97,
"learning_rate": 0.001,
"loss": 3.7552,
"step": 1475
},
{
"epoch": 0.97,
"learning_rate": 0.001,
"loss": 4.1927,
"step": 1480
},
{
"epoch": 0.97,
"learning_rate": 0.001,
"loss": 3.7616,
"step": 1485
},
{
"epoch": 0.98,
"learning_rate": 0.001,
"loss": 3.9175,
"step": 1490
},
{
"epoch": 0.98,
"learning_rate": 0.001,
"loss": 3.978,
"step": 1495
},
{
"epoch": 0.98,
"learning_rate": 0.001,
"loss": 3.6113,
"step": 1500
},
{
"epoch": 0.98,
"learning_rate": 0.001,
"loss": 3.9397,
"step": 1505
},
{
"epoch": 0.99,
"learning_rate": 0.001,
"loss": 3.8551,
"step": 1510
},
{
"epoch": 0.99,
"learning_rate": 0.001,
"loss": 3.8985,
"step": 1515
},
{
"epoch": 0.99,
"learning_rate": 0.001,
"loss": 3.8025,
"step": 1520
},
{
"epoch": 1.0,
"learning_rate": 0.001,
"loss": 3.7443,
"step": 1525
},
{
"epoch": 1.0,
"step": 1528,
"total_flos": 117360914595840.0,
"train_loss": 4.918852043401508,
"train_runtime": 510.3139,
"train_samples_per_second": 2.994,
"train_steps_per_second": 2.994
}
],
"logging_steps": 5,
"max_steps": 1528,
"num_train_epochs": 1,
"save_steps": 1000,
"total_flos": 117360914595840.0,
"trial_name": null,
"trial_params": null
}