mrm8488's picture
Upload the model
0ab1b15
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.2910379888528719,
"global_step": 1950,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 6.000000000000001e-07,
"loss": 2.9477,
"step": 10
},
{
"epoch": 0.0,
"learning_rate": 1.2000000000000002e-06,
"loss": 2.9437,
"step": 20
},
{
"epoch": 0.0,
"learning_rate": 1.8e-06,
"loss": 2.8391,
"step": 30
},
{
"epoch": 0.01,
"learning_rate": 2.4000000000000003e-06,
"loss": 2.7878,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 3e-06,
"loss": 2.7033,
"step": 50
},
{
"epoch": 0.01,
"learning_rate": 3.6e-06,
"loss": 2.5839,
"step": 60
},
{
"epoch": 0.01,
"learning_rate": 4.2000000000000004e-06,
"loss": 2.5077,
"step": 70
},
{
"epoch": 0.01,
"learning_rate": 4.800000000000001e-06,
"loss": 2.3937,
"step": 80
},
{
"epoch": 0.01,
"learning_rate": 5.4e-06,
"loss": 2.2994,
"step": 90
},
{
"epoch": 0.01,
"learning_rate": 6e-06,
"loss": 2.3544,
"step": 100
},
{
"epoch": 0.02,
"learning_rate": 6.6e-06,
"loss": 2.2563,
"step": 110
},
{
"epoch": 0.02,
"learning_rate": 7.2e-06,
"loss": 2.1624,
"step": 120
},
{
"epoch": 0.02,
"learning_rate": 7.8e-06,
"loss": 2.1635,
"step": 130
},
{
"epoch": 0.02,
"learning_rate": 8.400000000000001e-06,
"loss": 2.1196,
"step": 140
},
{
"epoch": 0.02,
"learning_rate": 9e-06,
"loss": 2.0701,
"step": 150
},
{
"epoch": 0.02,
"learning_rate": 9.600000000000001e-06,
"loss": 2.0222,
"step": 160
},
{
"epoch": 0.03,
"learning_rate": 1.02e-05,
"loss": 2.0375,
"step": 170
},
{
"epoch": 0.03,
"learning_rate": 1.08e-05,
"loss": 1.9524,
"step": 180
},
{
"epoch": 0.03,
"learning_rate": 1.1400000000000001e-05,
"loss": 1.9589,
"step": 190
},
{
"epoch": 0.03,
"learning_rate": 1.2e-05,
"loss": 1.9512,
"step": 200
},
{
"epoch": 0.03,
"learning_rate": 1.26e-05,
"loss": 1.8844,
"step": 210
},
{
"epoch": 0.03,
"learning_rate": 1.32e-05,
"loss": 1.8968,
"step": 220
},
{
"epoch": 0.03,
"learning_rate": 1.3800000000000002e-05,
"loss": 1.8407,
"step": 230
},
{
"epoch": 0.04,
"learning_rate": 1.44e-05,
"loss": 1.8488,
"step": 240
},
{
"epoch": 0.04,
"learning_rate": 1.5e-05,
"loss": 1.8461,
"step": 250
},
{
"epoch": 0.04,
"learning_rate": 1.56e-05,
"loss": 1.8473,
"step": 260
},
{
"epoch": 0.04,
"learning_rate": 1.62e-05,
"loss": 1.8165,
"step": 270
},
{
"epoch": 0.04,
"learning_rate": 1.6800000000000002e-05,
"loss": 1.8255,
"step": 280
},
{
"epoch": 0.04,
"learning_rate": 1.74e-05,
"loss": 1.8499,
"step": 290
},
{
"epoch": 0.04,
"learning_rate": 1.8e-05,
"loss": 1.8384,
"step": 300
},
{
"epoch": 0.05,
"learning_rate": 1.86e-05,
"loss": 1.8069,
"step": 310
},
{
"epoch": 0.05,
"learning_rate": 1.9200000000000003e-05,
"loss": 1.7542,
"step": 320
},
{
"epoch": 0.05,
"learning_rate": 1.98e-05,
"loss": 1.8121,
"step": 330
},
{
"epoch": 0.05,
"learning_rate": 2.04e-05,
"loss": 1.7528,
"step": 340
},
{
"epoch": 0.05,
"learning_rate": 2.1e-05,
"loss": 1.7386,
"step": 350
},
{
"epoch": 0.05,
"learning_rate": 2.16e-05,
"loss": 1.7411,
"step": 360
},
{
"epoch": 0.06,
"learning_rate": 2.22e-05,
"loss": 1.7301,
"step": 370
},
{
"epoch": 0.06,
"learning_rate": 2.2800000000000002e-05,
"loss": 1.7389,
"step": 380
},
{
"epoch": 0.06,
"learning_rate": 2.3400000000000003e-05,
"loss": 1.7339,
"step": 390
},
{
"epoch": 0.06,
"learning_rate": 2.4e-05,
"loss": 1.7242,
"step": 400
},
{
"epoch": 0.06,
"learning_rate": 2.4599999999999998e-05,
"loss": 1.7381,
"step": 410
},
{
"epoch": 0.06,
"learning_rate": 2.52e-05,
"loss": 1.7049,
"step": 420
},
{
"epoch": 0.06,
"learning_rate": 2.58e-05,
"loss": 1.699,
"step": 430
},
{
"epoch": 0.07,
"learning_rate": 2.64e-05,
"loss": 1.7059,
"step": 440
},
{
"epoch": 0.07,
"learning_rate": 2.7000000000000002e-05,
"loss": 1.7285,
"step": 450
},
{
"epoch": 0.07,
"learning_rate": 2.7600000000000003e-05,
"loss": 1.6645,
"step": 460
},
{
"epoch": 0.07,
"learning_rate": 2.8199999999999998e-05,
"loss": 1.6943,
"step": 470
},
{
"epoch": 0.07,
"learning_rate": 2.88e-05,
"loss": 1.6647,
"step": 480
},
{
"epoch": 0.07,
"learning_rate": 2.94e-05,
"loss": 1.6977,
"step": 490
},
{
"epoch": 0.07,
"learning_rate": 3e-05,
"loss": 1.672,
"step": 500
},
{
"epoch": 0.08,
"learning_rate": 2.9880000000000002e-05,
"loss": 1.6515,
"step": 510
},
{
"epoch": 0.08,
"learning_rate": 2.976e-05,
"loss": 1.7002,
"step": 520
},
{
"epoch": 0.08,
"learning_rate": 2.964e-05,
"loss": 1.6703,
"step": 530
},
{
"epoch": 0.08,
"learning_rate": 2.9520000000000002e-05,
"loss": 1.6593,
"step": 540
},
{
"epoch": 0.08,
"learning_rate": 2.94e-05,
"loss": 1.6481,
"step": 550
},
{
"epoch": 0.08,
"learning_rate": 2.928e-05,
"loss": 1.618,
"step": 560
},
{
"epoch": 0.09,
"learning_rate": 2.916e-05,
"loss": 1.6649,
"step": 570
},
{
"epoch": 0.09,
"learning_rate": 2.904e-05,
"loss": 1.6145,
"step": 580
},
{
"epoch": 0.09,
"learning_rate": 2.892e-05,
"loss": 1.6316,
"step": 590
},
{
"epoch": 0.09,
"learning_rate": 2.88e-05,
"loss": 1.6236,
"step": 600
},
{
"epoch": 0.09,
"learning_rate": 2.868e-05,
"loss": 1.6178,
"step": 610
},
{
"epoch": 0.09,
"learning_rate": 2.856e-05,
"loss": 1.5701,
"step": 620
},
{
"epoch": 0.09,
"learning_rate": 2.844e-05,
"loss": 1.6125,
"step": 630
},
{
"epoch": 0.1,
"learning_rate": 2.832e-05,
"loss": 1.5994,
"step": 640
},
{
"epoch": 0.1,
"learning_rate": 2.8199999999999998e-05,
"loss": 1.5694,
"step": 650
},
{
"epoch": 0.1,
"learning_rate": 2.8080000000000002e-05,
"loss": 1.5835,
"step": 660
},
{
"epoch": 0.1,
"learning_rate": 2.7960000000000003e-05,
"loss": 1.5758,
"step": 670
},
{
"epoch": 0.1,
"learning_rate": 2.784e-05,
"loss": 1.6216,
"step": 680
},
{
"epoch": 0.1,
"learning_rate": 2.7720000000000002e-05,
"loss": 1.5372,
"step": 690
},
{
"epoch": 0.1,
"learning_rate": 2.7600000000000003e-05,
"loss": 1.5672,
"step": 700
},
{
"epoch": 0.11,
"learning_rate": 2.748e-05,
"loss": 1.5815,
"step": 710
},
{
"epoch": 0.11,
"learning_rate": 2.7360000000000002e-05,
"loss": 1.5739,
"step": 720
},
{
"epoch": 0.11,
"learning_rate": 2.724e-05,
"loss": 1.5802,
"step": 730
},
{
"epoch": 0.11,
"learning_rate": 2.712e-05,
"loss": 1.5669,
"step": 740
},
{
"epoch": 0.11,
"learning_rate": 2.7000000000000002e-05,
"loss": 1.5571,
"step": 750
},
{
"epoch": 0.11,
"learning_rate": 2.688e-05,
"loss": 1.5661,
"step": 760
},
{
"epoch": 0.11,
"learning_rate": 2.676e-05,
"loss": 1.5643,
"step": 770
},
{
"epoch": 0.12,
"learning_rate": 2.6640000000000002e-05,
"loss": 1.5446,
"step": 780
},
{
"epoch": 0.12,
"learning_rate": 2.652e-05,
"loss": 1.5343,
"step": 790
},
{
"epoch": 0.12,
"learning_rate": 2.64e-05,
"loss": 1.5592,
"step": 800
},
{
"epoch": 0.12,
"learning_rate": 2.628e-05,
"loss": 1.5131,
"step": 810
},
{
"epoch": 0.12,
"learning_rate": 2.616e-05,
"loss": 1.5568,
"step": 820
},
{
"epoch": 0.12,
"learning_rate": 2.604e-05,
"loss": 1.5287,
"step": 830
},
{
"epoch": 0.13,
"learning_rate": 2.592e-05,
"loss": 1.5364,
"step": 840
},
{
"epoch": 0.13,
"learning_rate": 2.58e-05,
"loss": 1.5537,
"step": 850
},
{
"epoch": 0.13,
"learning_rate": 2.568e-05,
"loss": 1.5066,
"step": 860
},
{
"epoch": 0.13,
"learning_rate": 2.556e-05,
"loss": 1.5276,
"step": 870
},
{
"epoch": 0.13,
"learning_rate": 2.544e-05,
"loss": 1.5165,
"step": 880
},
{
"epoch": 0.13,
"learning_rate": 2.5319999999999998e-05,
"loss": 1.518,
"step": 890
},
{
"epoch": 0.13,
"learning_rate": 2.52e-05,
"loss": 1.5238,
"step": 900
},
{
"epoch": 0.14,
"learning_rate": 2.508e-05,
"loss": 1.5033,
"step": 910
},
{
"epoch": 0.14,
"learning_rate": 2.4959999999999998e-05,
"loss": 1.5505,
"step": 920
},
{
"epoch": 0.14,
"learning_rate": 2.484e-05,
"loss": 1.5228,
"step": 930
},
{
"epoch": 0.14,
"learning_rate": 2.472e-05,
"loss": 1.5207,
"step": 940
},
{
"epoch": 0.14,
"learning_rate": 2.4599999999999998e-05,
"loss": 1.5547,
"step": 950
},
{
"epoch": 0.14,
"learning_rate": 2.448e-05,
"loss": 1.5234,
"step": 960
},
{
"epoch": 0.14,
"learning_rate": 2.4360000000000004e-05,
"loss": 1.5325,
"step": 970
},
{
"epoch": 0.15,
"learning_rate": 2.4240000000000002e-05,
"loss": 1.5049,
"step": 980
},
{
"epoch": 0.15,
"learning_rate": 2.4120000000000003e-05,
"loss": 1.5138,
"step": 990
},
{
"epoch": 0.15,
"learning_rate": 2.4e-05,
"loss": 1.503,
"step": 1000
},
{
"epoch": 0.15,
"learning_rate": 2.3880000000000002e-05,
"loss": 1.5241,
"step": 1010
},
{
"epoch": 0.15,
"learning_rate": 2.3760000000000003e-05,
"loss": 1.5353,
"step": 1020
},
{
"epoch": 0.15,
"learning_rate": 2.364e-05,
"loss": 1.489,
"step": 1030
},
{
"epoch": 0.16,
"learning_rate": 2.3520000000000002e-05,
"loss": 1.4958,
"step": 1040
},
{
"epoch": 0.16,
"learning_rate": 2.3400000000000003e-05,
"loss": 1.5068,
"step": 1050
},
{
"epoch": 0.16,
"learning_rate": 2.328e-05,
"loss": 1.4988,
"step": 1060
},
{
"epoch": 0.16,
"learning_rate": 2.3160000000000002e-05,
"loss": 1.5311,
"step": 1070
},
{
"epoch": 0.16,
"learning_rate": 2.304e-05,
"loss": 1.4811,
"step": 1080
},
{
"epoch": 0.16,
"learning_rate": 2.292e-05,
"loss": 1.5139,
"step": 1090
},
{
"epoch": 0.16,
"learning_rate": 2.2800000000000002e-05,
"loss": 1.5403,
"step": 1100
},
{
"epoch": 0.17,
"learning_rate": 2.268e-05,
"loss": 1.4928,
"step": 1110
},
{
"epoch": 0.17,
"learning_rate": 2.256e-05,
"loss": 1.4635,
"step": 1120
},
{
"epoch": 0.17,
"learning_rate": 2.2440000000000002e-05,
"loss": 1.5088,
"step": 1130
},
{
"epoch": 0.17,
"learning_rate": 2.232e-05,
"loss": 1.5081,
"step": 1140
},
{
"epoch": 0.17,
"learning_rate": 2.22e-05,
"loss": 1.4923,
"step": 1150
},
{
"epoch": 0.17,
"learning_rate": 2.208e-05,
"loss": 1.4836,
"step": 1160
},
{
"epoch": 0.17,
"learning_rate": 2.196e-05,
"loss": 1.4861,
"step": 1170
},
{
"epoch": 0.18,
"learning_rate": 2.184e-05,
"loss": 1.4594,
"step": 1180
},
{
"epoch": 0.18,
"learning_rate": 2.172e-05,
"loss": 1.4895,
"step": 1190
},
{
"epoch": 0.18,
"learning_rate": 2.16e-05,
"loss": 1.4775,
"step": 1200
},
{
"epoch": 0.18,
"learning_rate": 2.148e-05,
"loss": 1.4682,
"step": 1210
},
{
"epoch": 0.18,
"learning_rate": 2.136e-05,
"loss": 1.5022,
"step": 1220
},
{
"epoch": 0.18,
"learning_rate": 2.124e-05,
"loss": 1.4594,
"step": 1230
},
{
"epoch": 0.19,
"learning_rate": 2.1119999999999998e-05,
"loss": 1.453,
"step": 1240
},
{
"epoch": 0.19,
"learning_rate": 2.1e-05,
"loss": 1.4665,
"step": 1250
},
{
"epoch": 0.19,
"learning_rate": 2.088e-05,
"loss": 1.489,
"step": 1260
},
{
"epoch": 0.19,
"learning_rate": 2.0759999999999998e-05,
"loss": 1.4959,
"step": 1270
},
{
"epoch": 0.19,
"learning_rate": 2.064e-05,
"loss": 1.4831,
"step": 1280
},
{
"epoch": 0.19,
"learning_rate": 2.0520000000000003e-05,
"loss": 1.4488,
"step": 1290
},
{
"epoch": 0.19,
"learning_rate": 2.04e-05,
"loss": 1.4885,
"step": 1300
},
{
"epoch": 0.2,
"learning_rate": 2.0280000000000002e-05,
"loss": 1.4756,
"step": 1310
},
{
"epoch": 0.2,
"learning_rate": 2.016e-05,
"loss": 1.4813,
"step": 1320
},
{
"epoch": 0.2,
"learning_rate": 2.004e-05,
"loss": 1.4878,
"step": 1330
},
{
"epoch": 0.2,
"learning_rate": 1.9920000000000002e-05,
"loss": 1.4756,
"step": 1340
},
{
"epoch": 0.2,
"learning_rate": 1.98e-05,
"loss": 1.4759,
"step": 1350
},
{
"epoch": 0.2,
"learning_rate": 1.968e-05,
"loss": 1.4541,
"step": 1360
},
{
"epoch": 0.2,
"learning_rate": 1.9560000000000002e-05,
"loss": 1.4533,
"step": 1370
},
{
"epoch": 0.21,
"learning_rate": 1.944e-05,
"loss": 1.4842,
"step": 1380
},
{
"epoch": 0.21,
"learning_rate": 1.932e-05,
"loss": 1.4585,
"step": 1390
},
{
"epoch": 0.21,
"learning_rate": 1.9200000000000003e-05,
"loss": 1.4781,
"step": 1400
},
{
"epoch": 0.21,
"learning_rate": 1.908e-05,
"loss": 1.4637,
"step": 1410
},
{
"epoch": 0.21,
"learning_rate": 1.896e-05,
"loss": 1.4691,
"step": 1420
},
{
"epoch": 0.21,
"learning_rate": 1.884e-05,
"loss": 1.4782,
"step": 1430
},
{
"epoch": 0.21,
"learning_rate": 1.872e-05,
"loss": 1.4689,
"step": 1440
},
{
"epoch": 0.22,
"learning_rate": 1.86e-05,
"loss": 1.4603,
"step": 1450
},
{
"epoch": 0.22,
"learning_rate": 1.848e-05,
"loss": 1.4666,
"step": 1460
},
{
"epoch": 0.22,
"learning_rate": 1.836e-05,
"loss": 1.4694,
"step": 1470
},
{
"epoch": 0.22,
"learning_rate": 1.824e-05,
"loss": 1.4499,
"step": 1480
},
{
"epoch": 0.22,
"learning_rate": 1.812e-05,
"loss": 1.4794,
"step": 1490
},
{
"epoch": 0.22,
"learning_rate": 1.8e-05,
"loss": 1.475,
"step": 1500
},
{
"epoch": 0.23,
"learning_rate": 1.7879999999999998e-05,
"loss": 1.4431,
"step": 1510
},
{
"epoch": 0.23,
"learning_rate": 1.776e-05,
"loss": 1.4492,
"step": 1520
},
{
"epoch": 0.23,
"learning_rate": 1.764e-05,
"loss": 1.4656,
"step": 1530
},
{
"epoch": 0.23,
"learning_rate": 1.7519999999999998e-05,
"loss": 1.4524,
"step": 1540
},
{
"epoch": 0.23,
"learning_rate": 1.74e-05,
"loss": 1.4611,
"step": 1550
},
{
"epoch": 0.23,
"learning_rate": 1.728e-05,
"loss": 1.4602,
"step": 1560
},
{
"epoch": 0.23,
"learning_rate": 1.716e-05,
"loss": 1.4603,
"step": 1570
},
{
"epoch": 0.24,
"learning_rate": 1.704e-05,
"loss": 1.4334,
"step": 1580
},
{
"epoch": 0.24,
"learning_rate": 1.6919999999999997e-05,
"loss": 1.4761,
"step": 1590
},
{
"epoch": 0.24,
"learning_rate": 1.6800000000000002e-05,
"loss": 1.4193,
"step": 1600
},
{
"epoch": 0.24,
"learning_rate": 1.6680000000000003e-05,
"loss": 1.4209,
"step": 1610
},
{
"epoch": 0.24,
"learning_rate": 1.656e-05,
"loss": 1.4473,
"step": 1620
},
{
"epoch": 0.24,
"learning_rate": 1.6440000000000002e-05,
"loss": 1.4451,
"step": 1630
},
{
"epoch": 0.24,
"learning_rate": 1.6320000000000003e-05,
"loss": 1.4428,
"step": 1640
},
{
"epoch": 0.25,
"learning_rate": 1.62e-05,
"loss": 1.444,
"step": 1650
},
{
"epoch": 0.25,
"learning_rate": 1.6080000000000002e-05,
"loss": 1.4677,
"step": 1660
},
{
"epoch": 0.25,
"learning_rate": 1.596e-05,
"loss": 1.4483,
"step": 1670
},
{
"epoch": 0.25,
"learning_rate": 1.584e-05,
"loss": 1.4536,
"step": 1680
},
{
"epoch": 0.25,
"learning_rate": 1.5720000000000002e-05,
"loss": 1.4515,
"step": 1690
},
{
"epoch": 0.25,
"learning_rate": 1.56e-05,
"loss": 1.436,
"step": 1700
},
{
"epoch": 0.26,
"learning_rate": 1.548e-05,
"loss": 1.4277,
"step": 1710
},
{
"epoch": 0.26,
"learning_rate": 1.5360000000000002e-05,
"loss": 1.4434,
"step": 1720
},
{
"epoch": 0.26,
"learning_rate": 1.524e-05,
"loss": 1.4312,
"step": 1730
},
{
"epoch": 0.26,
"learning_rate": 1.5120000000000001e-05,
"loss": 1.4457,
"step": 1740
},
{
"epoch": 0.26,
"learning_rate": 1.5e-05,
"loss": 1.4404,
"step": 1750
},
{
"epoch": 0.26,
"learning_rate": 1.488e-05,
"loss": 1.4684,
"step": 1760
},
{
"epoch": 0.26,
"learning_rate": 1.4760000000000001e-05,
"loss": 1.4449,
"step": 1770
},
{
"epoch": 0.27,
"learning_rate": 1.464e-05,
"loss": 1.4354,
"step": 1780
},
{
"epoch": 0.27,
"learning_rate": 1.452e-05,
"loss": 1.4303,
"step": 1790
},
{
"epoch": 0.27,
"learning_rate": 1.44e-05,
"loss": 1.4215,
"step": 1800
},
{
"epoch": 0.27,
"learning_rate": 1.428e-05,
"loss": 1.4385,
"step": 1810
},
{
"epoch": 0.27,
"learning_rate": 1.416e-05,
"loss": 1.4239,
"step": 1820
},
{
"epoch": 0.27,
"learning_rate": 1.4040000000000001e-05,
"loss": 1.4431,
"step": 1830
},
{
"epoch": 0.27,
"learning_rate": 1.392e-05,
"loss": 1.4354,
"step": 1840
},
{
"epoch": 0.28,
"learning_rate": 1.3800000000000002e-05,
"loss": 1.4428,
"step": 1850
},
{
"epoch": 0.28,
"learning_rate": 1.3680000000000001e-05,
"loss": 1.4062,
"step": 1860
},
{
"epoch": 0.28,
"learning_rate": 1.356e-05,
"loss": 1.4423,
"step": 1870
},
{
"epoch": 0.28,
"learning_rate": 1.344e-05,
"loss": 1.4218,
"step": 1880
},
{
"epoch": 0.28,
"learning_rate": 1.3320000000000001e-05,
"loss": 1.4131,
"step": 1890
},
{
"epoch": 0.28,
"learning_rate": 1.32e-05,
"loss": 1.4536,
"step": 1900
},
{
"epoch": 0.29,
"learning_rate": 1.308e-05,
"loss": 1.3985,
"step": 1910
},
{
"epoch": 0.29,
"learning_rate": 1.296e-05,
"loss": 1.4204,
"step": 1920
},
{
"epoch": 0.29,
"learning_rate": 1.284e-05,
"loss": 1.4266,
"step": 1930
},
{
"epoch": 0.29,
"learning_rate": 1.272e-05,
"loss": 1.4255,
"step": 1940
},
{
"epoch": 0.29,
"learning_rate": 1.26e-05,
"loss": 1.4357,
"step": 1950
}
],
"max_steps": 3000,
"num_train_epochs": 1,
"total_flos": 3.28059806810112e+17,
"trial_name": null,
"trial_params": null
}