zilo-sft-qlora / trainer_state.json
todenthal's picture
Initial commit of the model
3394e52
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9696969696969697,
"eval_steps": 500,
"global_step": 147,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.020202020202020204,
"grad_norm": 2.578125,
"learning_rate": 1.3333333333333334e-06,
"loss": 1.5448,
"step": 1
},
{
"epoch": 0.10101010101010101,
"grad_norm": 2.546875,
"learning_rate": 6.666666666666667e-06,
"loss": 1.5503,
"step": 5
},
{
"epoch": 0.20202020202020202,
"grad_norm": 2.390625,
"learning_rate": 1.3333333333333333e-05,
"loss": 1.469,
"step": 10
},
{
"epoch": 0.30303030303030304,
"grad_norm": 1.65625,
"learning_rate": 2e-05,
"loss": 1.2465,
"step": 15
},
{
"epoch": 0.40404040404040403,
"grad_norm": 1.203125,
"learning_rate": 1.9929278846732883e-05,
"loss": 1.0305,
"step": 20
},
{
"epoch": 0.5050505050505051,
"grad_norm": 0.953125,
"learning_rate": 1.9718115683235418e-05,
"loss": 0.9205,
"step": 25
},
{
"epoch": 0.6060606060606061,
"grad_norm": 0.609375,
"learning_rate": 1.936949724999762e-05,
"loss": 0.8521,
"step": 30
},
{
"epoch": 0.7070707070707071,
"grad_norm": 0.48046875,
"learning_rate": 1.8888354486549238e-05,
"loss": 0.8354,
"step": 35
},
{
"epoch": 0.8080808080808081,
"grad_norm": 0.478515625,
"learning_rate": 1.8281492787113707e-05,
"loss": 0.7916,
"step": 40
},
{
"epoch": 0.9090909090909091,
"grad_norm": 0.46484375,
"learning_rate": 1.7557495743542586e-05,
"loss": 0.7588,
"step": 45
},
{
"epoch": 0.98989898989899,
"eval_loss": 0.7386298179626465,
"eval_runtime": 17.3037,
"eval_samples_per_second": 2.485,
"eval_steps_per_second": 0.347,
"step": 49
},
{
"epoch": 1.0101010101010102,
"grad_norm": 0.5390625,
"learning_rate": 1.6726603737012527e-05,
"loss": 0.7461,
"step": 50
},
{
"epoch": 1.1111111111111112,
"grad_norm": 0.5078125,
"learning_rate": 1.5800569095711983e-05,
"loss": 0.715,
"step": 55
},
{
"epoch": 1.2121212121212122,
"grad_norm": 0.3046875,
"learning_rate": 1.479248986720057e-05,
"loss": 0.6996,
"step": 60
},
{
"epoch": 1.3131313131313131,
"grad_norm": 0.28125,
"learning_rate": 1.3716624556603275e-05,
"loss": 0.6809,
"step": 65
},
{
"epoch": 1.4141414141414141,
"grad_norm": 0.318359375,
"learning_rate": 1.2588190451025209e-05,
"loss": 0.6585,
"step": 70
},
{
"epoch": 1.5151515151515151,
"grad_norm": 0.2255859375,
"learning_rate": 1.1423148382732854e-05,
"loss": 0.6599,
"step": 75
},
{
"epoch": 1.6161616161616161,
"grad_norm": 0.2431640625,
"learning_rate": 1.0237976975461074e-05,
"loss": 0.646,
"step": 80
},
{
"epoch": 1.7171717171717171,
"grad_norm": 0.203125,
"learning_rate": 9.049439566958176e-06,
"loss": 0.6349,
"step": 85
},
{
"epoch": 1.8181818181818183,
"grad_norm": 0.2392578125,
"learning_rate": 7.874347104470234e-06,
"loss": 0.6352,
"step": 90
},
{
"epoch": 1.9191919191919191,
"grad_norm": 0.2060546875,
"learning_rate": 6.729320366825785e-06,
"loss": 0.6201,
"step": 95
},
{
"epoch": 2.0,
"eval_loss": 0.6306869983673096,
"eval_runtime": 17.31,
"eval_samples_per_second": 2.484,
"eval_steps_per_second": 0.347,
"step": 99
},
{
"epoch": 2.0202020202020203,
"grad_norm": 0.2001953125,
"learning_rate": 5.630554876306407e-06,
"loss": 0.6233,
"step": 100
},
{
"epoch": 2.121212121212121,
"grad_norm": 0.2021484375,
"learning_rate": 4.593591825444028e-06,
"loss": 0.6138,
"step": 105
},
{
"epoch": 2.2222222222222223,
"grad_norm": 0.2275390625,
"learning_rate": 3.633098258809119e-06,
"loss": 0.6218,
"step": 110
},
{
"epoch": 2.323232323232323,
"grad_norm": 0.2021484375,
"learning_rate": 2.7626596189492983e-06,
"loss": 0.6201,
"step": 115
},
{
"epoch": 2.4242424242424243,
"grad_norm": 0.19921875,
"learning_rate": 1.994587590756397e-06,
"loss": 0.6164,
"step": 120
},
{
"epoch": 2.525252525252525,
"grad_norm": 0.2041015625,
"learning_rate": 1.339745962155613e-06,
"loss": 0.6157,
"step": 125
},
{
"epoch": 2.6262626262626263,
"grad_norm": 0.216796875,
"learning_rate": 8.073969641833446e-07,
"loss": 0.6151,
"step": 130
},
{
"epoch": 2.7272727272727275,
"grad_norm": 0.267578125,
"learning_rate": 4.0507026385502747e-07,
"loss": 0.6097,
"step": 135
},
{
"epoch": 2.8282828282828283,
"grad_norm": 0.208984375,
"learning_rate": 1.3845646281813508e-07,
"loss": 0.617,
"step": 140
},
{
"epoch": 2.929292929292929,
"grad_norm": 0.1943359375,
"learning_rate": 1.1326608169920373e-08,
"loss": 0.6087,
"step": 145
},
{
"epoch": 2.9696969696969697,
"eval_loss": 0.6245681643486023,
"eval_runtime": 17.3067,
"eval_samples_per_second": 2.485,
"eval_steps_per_second": 0.347,
"step": 147
},
{
"epoch": 2.9696969696969697,
"step": 147,
"total_flos": 2.0566538541072384e+17,
"train_loss": 0.7671454966473742,
"train_runtime": 1686.4021,
"train_samples_per_second": 0.699,
"train_steps_per_second": 0.087
}
],
"logging_steps": 5,
"max_steps": 147,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.0566538541072384e+17,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}