384 lines
8.0 KiB
JSON
384 lines
8.0 KiB
JSON
|
|
{
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 0.0005,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 50,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 1e-05,
|
||
|
|
"grad_norm": 8.875,
|
||
|
|
"learning_rate": 9.999999997532599e-06,
|
||
|
|
"loss": 1.6459,
|
||
|
|
"step": 1
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2e-05,
|
||
|
|
"grad_norm": 4.40625,
|
||
|
|
"learning_rate": 9.999999990130395e-06,
|
||
|
|
"loss": 1.6742,
|
||
|
|
"step": 2
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3e-05,
|
||
|
|
"grad_norm": 4.3125,
|
||
|
|
"learning_rate": 9.99999997779339e-06,
|
||
|
|
"loss": 1.6223,
|
||
|
|
"step": 3
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4e-05,
|
||
|
|
"grad_norm": 4.0625,
|
||
|
|
"learning_rate": 9.999999960521582e-06,
|
||
|
|
"loss": 1.5398,
|
||
|
|
"step": 4
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 5e-05,
|
||
|
|
"grad_norm": 3.375,
|
||
|
|
"learning_rate": 9.999999938314972e-06,
|
||
|
|
"loss": 1.5666,
|
||
|
|
"step": 5
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 6e-05,
|
||
|
|
"grad_norm": 1.859375,
|
||
|
|
"learning_rate": 9.999999911173561e-06,
|
||
|
|
"loss": 1.5981,
|
||
|
|
"step": 6
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 7e-05,
|
||
|
|
"grad_norm": 1.734375,
|
||
|
|
"learning_rate": 9.999999879097347e-06,
|
||
|
|
"loss": 1.644,
|
||
|
|
"step": 7
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 8e-05,
|
||
|
|
"grad_norm": 1.796875,
|
||
|
|
"learning_rate": 9.999999842086332e-06,
|
||
|
|
"loss": 1.6331,
|
||
|
|
"step": 8
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 9e-05,
|
||
|
|
"grad_norm": 1.3125,
|
||
|
|
"learning_rate": 9.999999800140514e-06,
|
||
|
|
"loss": 1.626,
|
||
|
|
"step": 9
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0001,
|
||
|
|
"grad_norm": 1.9296875,
|
||
|
|
"learning_rate": 9.999999753259893e-06,
|
||
|
|
"loss": 1.5778,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00011,
|
||
|
|
"grad_norm": 1.34375,
|
||
|
|
"learning_rate": 9.99999970144447e-06,
|
||
|
|
"loss": 1.6286,
|
||
|
|
"step": 11
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00012,
|
||
|
|
"grad_norm": 1.203125,
|
||
|
|
"learning_rate": 9.999999644694247e-06,
|
||
|
|
"loss": 1.5614,
|
||
|
|
"step": 12
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00013,
|
||
|
|
"grad_norm": 1.015625,
|
||
|
|
"learning_rate": 9.999999583009221e-06,
|
||
|
|
"loss": 1.6447,
|
||
|
|
"step": 13
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00014,
|
||
|
|
"grad_norm": 1.3359375,
|
||
|
|
"learning_rate": 9.999999516389394e-06,
|
||
|
|
"loss": 1.5258,
|
||
|
|
"step": 14
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00015,
|
||
|
|
"grad_norm": 1.25,
|
||
|
|
"learning_rate": 9.999999444834763e-06,
|
||
|
|
"loss": 1.6336,
|
||
|
|
"step": 15
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00016,
|
||
|
|
"grad_norm": 1.5546875,
|
||
|
|
"learning_rate": 9.999999368345333e-06,
|
||
|
|
"loss": 1.6073,
|
||
|
|
"step": 16
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00017,
|
||
|
|
"grad_norm": 1.34375,
|
||
|
|
"learning_rate": 9.999999286921101e-06,
|
||
|
|
"loss": 1.5919,
|
||
|
|
"step": 17
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00018,
|
||
|
|
"grad_norm": 0.96875,
|
||
|
|
"learning_rate": 9.999999200562065e-06,
|
||
|
|
"loss": 1.543,
|
||
|
|
"step": 18
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00019,
|
||
|
|
"grad_norm": 1.6875,
|
||
|
|
"learning_rate": 9.99999910926823e-06,
|
||
|
|
"loss": 1.6101,
|
||
|
|
"step": 19
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0002,
|
||
|
|
"grad_norm": 1.578125,
|
||
|
|
"learning_rate": 9.999999013039593e-06,
|
||
|
|
"loss": 1.5796,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00021,
|
||
|
|
"grad_norm": 2.578125,
|
||
|
|
"learning_rate": 9.999998911876154e-06,
|
||
|
|
"loss": 1.5748,
|
||
|
|
"step": 21
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00022,
|
||
|
|
"grad_norm": 1.203125,
|
||
|
|
"learning_rate": 9.999998805777915e-06,
|
||
|
|
"loss": 1.5479,
|
||
|
|
"step": 22
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00023,
|
||
|
|
"grad_norm": 1.4921875,
|
||
|
|
"learning_rate": 9.999998694744875e-06,
|
||
|
|
"loss": 1.5318,
|
||
|
|
"step": 23
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00024,
|
||
|
|
"grad_norm": 1.125,
|
||
|
|
"learning_rate": 9.999998578777036e-06,
|
||
|
|
"loss": 1.6259,
|
||
|
|
"step": 24
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00025,
|
||
|
|
"grad_norm": 2.21875,
|
||
|
|
"learning_rate": 9.999998457874392e-06,
|
||
|
|
"loss": 1.5525,
|
||
|
|
"step": 25
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00026,
|
||
|
|
"grad_norm": 3.234375,
|
||
|
|
"learning_rate": 9.99999833203695e-06,
|
||
|
|
"loss": 1.5576,
|
||
|
|
"step": 26
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00027,
|
||
|
|
"grad_norm": 2.046875,
|
||
|
|
"learning_rate": 9.999998201264707e-06,
|
||
|
|
"loss": 1.3934,
|
||
|
|
"step": 27
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00028,
|
||
|
|
"grad_norm": 3.15625,
|
||
|
|
"learning_rate": 9.999998065557664e-06,
|
||
|
|
"loss": 1.5423,
|
||
|
|
"step": 28
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00029,
|
||
|
|
"grad_norm": 1.2734375,
|
||
|
|
"learning_rate": 9.999997924915818e-06,
|
||
|
|
"loss": 1.5679,
|
||
|
|
"step": 29
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0003,
|
||
|
|
"grad_norm": 1.625,
|
||
|
|
"learning_rate": 9.999997779339175e-06,
|
||
|
|
"loss": 1.5329,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00031,
|
||
|
|
"grad_norm": 1.2421875,
|
||
|
|
"learning_rate": 9.999997628827732e-06,
|
||
|
|
"loss": 1.4603,
|
||
|
|
"step": 31
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00032,
|
||
|
|
"grad_norm": 1.46875,
|
||
|
|
"learning_rate": 9.999997473381487e-06,
|
||
|
|
"loss": 1.5774,
|
||
|
|
"step": 32
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00033,
|
||
|
|
"grad_norm": 2.0625,
|
||
|
|
"learning_rate": 9.999997313000444e-06,
|
||
|
|
"loss": 1.5522,
|
||
|
|
"step": 33
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00034,
|
||
|
|
"grad_norm": 1.421875,
|
||
|
|
"learning_rate": 9.9999971476846e-06,
|
||
|
|
"loss": 1.5964,
|
||
|
|
"step": 34
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00035,
|
||
|
|
"grad_norm": 1.59375,
|
||
|
|
"learning_rate": 9.999996977433957e-06,
|
||
|
|
"loss": 1.6129,
|
||
|
|
"step": 35
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00036,
|
||
|
|
"grad_norm": 1.78125,
|
||
|
|
"learning_rate": 9.999996802248514e-06,
|
||
|
|
"loss": 1.548,
|
||
|
|
"step": 36
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00037,
|
||
|
|
"grad_norm": 0.94140625,
|
||
|
|
"learning_rate": 9.999996622128274e-06,
|
||
|
|
"loss": 1.5662,
|
||
|
|
"step": 37
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00038,
|
||
|
|
"grad_norm": 4.84375,
|
||
|
|
"learning_rate": 9.999996437073236e-06,
|
||
|
|
"loss": 1.6197,
|
||
|
|
"step": 38
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00039,
|
||
|
|
"grad_norm": 3.234375,
|
||
|
|
"learning_rate": 9.999996247083397e-06,
|
||
|
|
"loss": 1.5308,
|
||
|
|
"step": 39
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0004,
|
||
|
|
"grad_norm": 1.375,
|
||
|
|
"learning_rate": 9.99999605215876e-06,
|
||
|
|
"loss": 1.5846,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00041,
|
||
|
|
"grad_norm": 1.34375,
|
||
|
|
"learning_rate": 9.999995852299324e-06,
|
||
|
|
"loss": 1.4274,
|
||
|
|
"step": 41
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00042,
|
||
|
|
"grad_norm": 1.15625,
|
||
|
|
"learning_rate": 9.999995647505092e-06,
|
||
|
|
"loss": 1.4986,
|
||
|
|
"step": 42
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00043,
|
||
|
|
"grad_norm": 0.94140625,
|
||
|
|
"learning_rate": 9.99999543777606e-06,
|
||
|
|
"loss": 1.5135,
|
||
|
|
"step": 43
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00044,
|
||
|
|
"grad_norm": 0.8125,
|
||
|
|
"learning_rate": 9.999995223112231e-06,
|
||
|
|
"loss": 1.5472,
|
||
|
|
"step": 44
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00045,
|
||
|
|
"grad_norm": 1.5703125,
|
||
|
|
"learning_rate": 9.999995003513605e-06,
|
||
|
|
"loss": 1.5635,
|
||
|
|
"step": 45
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00046,
|
||
|
|
"grad_norm": 3.34375,
|
||
|
|
"learning_rate": 9.999994778980182e-06,
|
||
|
|
"loss": 1.5506,
|
||
|
|
"step": 46
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00047,
|
||
|
|
"grad_norm": 5.4375,
|
||
|
|
"learning_rate": 9.99999454951196e-06,
|
||
|
|
"loss": 1.5071,
|
||
|
|
"step": 47
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00048,
|
||
|
|
"grad_norm": 1.578125,
|
||
|
|
"learning_rate": 9.999994315108943e-06,
|
||
|
|
"loss": 1.5532,
|
||
|
|
"step": 48
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00049,
|
||
|
|
"grad_norm": 1.828125,
|
||
|
|
"learning_rate": 9.999994075771128e-06,
|
||
|
|
"loss": 1.6061,
|
||
|
|
"step": 49
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0005,
|
||
|
|
"grad_norm": 1.4453125,
|
||
|
|
"learning_rate": 9.999993831498517e-06,
|
||
|
|
"loss": 1.5629,
|
||
|
|
"step": 50
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 1.0,
|
||
|
|
"max_steps": 100000,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 9223372036854775807,
|
||
|
|
"save_steps": 50,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": false
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 1.1527564296192e+18,
|
||
|
|
"train_batch_size": 1,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|