Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +111 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,111 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.0291192531585693,
|
| 3 |
+
"best_model_checkpoint": "savedEpoch_nccl/checkpoint-91656",
|
| 4 |
+
"epoch": 6.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 91656,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.0,
|
| 13 |
+
"grad_norm": 0.6372342109680176,
|
| 14 |
+
"learning_rate": 9.882461854198269e-05,
|
| 15 |
+
"loss": 1.1441,
|
| 16 |
+
"step": 15276
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_loss": 1.0554957389831543,
|
| 21 |
+
"eval_runtime": 57.5344,
|
| 22 |
+
"eval_samples_per_second": 12743.952,
|
| 23 |
+
"eval_steps_per_second": 33.198,
|
| 24 |
+
"step": 15276
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 2.0,
|
| 28 |
+
"grad_norm": 0.6399237513542175,
|
| 29 |
+
"learning_rate": 9.541848844472206e-05,
|
| 30 |
+
"loss": 1.0486,
|
| 31 |
+
"step": 30552
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 2.0,
|
| 35 |
+
"eval_loss": 1.0447489023208618,
|
| 36 |
+
"eval_runtime": 240.72,
|
| 37 |
+
"eval_samples_per_second": 3045.924,
|
| 38 |
+
"eval_steps_per_second": 7.935,
|
| 39 |
+
"step": 30552
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 3.0,
|
| 43 |
+
"grad_norm": 0.7907009124755859,
|
| 44 |
+
"learning_rate": 9.201235834746142e-05,
|
| 45 |
+
"loss": 1.0407,
|
| 46 |
+
"step": 45828
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 3.0,
|
| 50 |
+
"eval_loss": 1.0373780727386475,
|
| 51 |
+
"eval_runtime": 56.906,
|
| 52 |
+
"eval_samples_per_second": 12884.672,
|
| 53 |
+
"eval_steps_per_second": 33.564,
|
| 54 |
+
"step": 45828
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 4.0,
|
| 58 |
+
"grad_norm": 0.5956621766090393,
|
| 59 |
+
"learning_rate": 8.860622825020078e-05,
|
| 60 |
+
"loss": 1.0376,
|
| 61 |
+
"step": 61104
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 4.0,
|
| 65 |
+
"eval_loss": 1.0327318906784058,
|
| 66 |
+
"eval_runtime": 57.0015,
|
| 67 |
+
"eval_samples_per_second": 12863.086,
|
| 68 |
+
"eval_steps_per_second": 33.508,
|
| 69 |
+
"step": 61104
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 5.0,
|
| 73 |
+
"grad_norm": 0.6499078273773193,
|
| 74 |
+
"learning_rate": 8.519965200321228e-05,
|
| 75 |
+
"loss": 1.037,
|
| 76 |
+
"step": 76380
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 5.0,
|
| 80 |
+
"eval_loss": 1.0331288576126099,
|
| 81 |
+
"eval_runtime": 57.0146,
|
| 82 |
+
"eval_samples_per_second": 12860.131,
|
| 83 |
+
"eval_steps_per_second": 33.5,
|
| 84 |
+
"step": 76380
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 6.0,
|
| 88 |
+
"grad_norm": 0.5220733284950256,
|
| 89 |
+
"learning_rate": 8.179329883108772e-05,
|
| 90 |
+
"loss": 1.0344,
|
| 91 |
+
"step": 91656
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 6.0,
|
| 95 |
+
"eval_loss": 1.0291192531585693,
|
| 96 |
+
"eval_runtime": 56.9287,
|
| 97 |
+
"eval_samples_per_second": 12879.541,
|
| 98 |
+
"eval_steps_per_second": 33.551,
|
| 99 |
+
"step": 91656
|
| 100 |
+
}
|
| 101 |
+
],
|
| 102 |
+
"logging_steps": 500,
|
| 103 |
+
"max_steps": 458280,
|
| 104 |
+
"num_input_tokens_seen": 0,
|
| 105 |
+
"num_train_epochs": 30,
|
| 106 |
+
"save_steps": 500,
|
| 107 |
+
"total_flos": 8.139053392031908e+17,
|
| 108 |
+
"train_batch_size": 96,
|
| 109 |
+
"trial_name": null,
|
| 110 |
+
"trial_params": null
|
| 111 |
+
}
|