End of training
Browse files- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +103 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.91,
|
| 3 |
+
"total_flos": 6.869074650501612e+17,
|
| 4 |
+
"train_loss": 1.3921312093734741,
|
| 5 |
+
"train_runtime": 216.3192,
|
| 6 |
+
"train_samples_per_second": 41.605,
|
| 7 |
+
"train_steps_per_second": 0.185
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.91,
|
| 3 |
+
"total_flos": 6.869074650501612e+17,
|
| 4 |
+
"train_loss": 1.3921312093734741,
|
| 5 |
+
"train_runtime": 216.3192,
|
| 6 |
+
"train_samples_per_second": 41.605,
|
| 7 |
+
"train_steps_per_second": 0.185
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,103 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.695,
|
| 3 |
+
"best_model_checkpoint": "chord-final-model/checkpoint-40",
|
| 4 |
+
"epoch": 4.912280701754386,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 40,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.98,
|
| 13 |
+
"eval_accuracy": 0.4,
|
| 14 |
+
"eval_loss": 1.5147000551223755,
|
| 15 |
+
"eval_runtime": 2.3038,
|
| 16 |
+
"eval_samples_per_second": 86.815,
|
| 17 |
+
"eval_steps_per_second": 3.039,
|
| 18 |
+
"step": 8
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 1.23,
|
| 22 |
+
"grad_norm": 2.2252283096313477,
|
| 23 |
+
"learning_rate": 8.333333333333334e-06,
|
| 24 |
+
"loss": 1.6354,
|
| 25 |
+
"step": 10
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 1.96,
|
| 29 |
+
"eval_accuracy": 0.585,
|
| 30 |
+
"eval_loss": 1.3898262977600098,
|
| 31 |
+
"eval_runtime": 2.2865,
|
| 32 |
+
"eval_samples_per_second": 87.469,
|
| 33 |
+
"eval_steps_per_second": 3.061,
|
| 34 |
+
"step": 16
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 2.46,
|
| 38 |
+
"grad_norm": 2.2747433185577393,
|
| 39 |
+
"learning_rate": 5.555555555555557e-06,
|
| 40 |
+
"loss": 1.4272,
|
| 41 |
+
"step": 20
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 2.95,
|
| 45 |
+
"eval_accuracy": 0.67,
|
| 46 |
+
"eval_loss": 1.296023964881897,
|
| 47 |
+
"eval_runtime": 2.2818,
|
| 48 |
+
"eval_samples_per_second": 87.65,
|
| 49 |
+
"eval_steps_per_second": 3.068,
|
| 50 |
+
"step": 24
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 3.68,
|
| 54 |
+
"grad_norm": 2.199195623397827,
|
| 55 |
+
"learning_rate": 2.7777777777777783e-06,
|
| 56 |
+
"loss": 1.2884,
|
| 57 |
+
"step": 30
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 3.93,
|
| 61 |
+
"eval_accuracy": 0.67,
|
| 62 |
+
"eval_loss": 1.2369245290756226,
|
| 63 |
+
"eval_runtime": 2.2809,
|
| 64 |
+
"eval_samples_per_second": 87.684,
|
| 65 |
+
"eval_steps_per_second": 3.069,
|
| 66 |
+
"step": 32
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 4.91,
|
| 70 |
+
"grad_norm": 2.335061550140381,
|
| 71 |
+
"learning_rate": 0.0,
|
| 72 |
+
"loss": 1.2176,
|
| 73 |
+
"step": 40
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 4.91,
|
| 77 |
+
"eval_accuracy": 0.695,
|
| 78 |
+
"eval_loss": 1.213279128074646,
|
| 79 |
+
"eval_runtime": 2.345,
|
| 80 |
+
"eval_samples_per_second": 85.287,
|
| 81 |
+
"eval_steps_per_second": 2.985,
|
| 82 |
+
"step": 40
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 4.91,
|
| 86 |
+
"step": 40,
|
| 87 |
+
"total_flos": 6.869074650501612e+17,
|
| 88 |
+
"train_loss": 1.3921312093734741,
|
| 89 |
+
"train_runtime": 216.3192,
|
| 90 |
+
"train_samples_per_second": 41.605,
|
| 91 |
+
"train_steps_per_second": 0.185
|
| 92 |
+
}
|
| 93 |
+
],
|
| 94 |
+
"logging_steps": 10,
|
| 95 |
+
"max_steps": 40,
|
| 96 |
+
"num_input_tokens_seen": 0,
|
| 97 |
+
"num_train_epochs": 5,
|
| 98 |
+
"save_steps": 500,
|
| 99 |
+
"total_flos": 6.869074650501612e+17,
|
| 100 |
+
"train_batch_size": 32,
|
| 101 |
+
"trial_name": null,
|
| 102 |
+
"trial_params": null
|
| 103 |
+
}
|