Edison
commited on
Commit
·
692cf65
1
Parent(s):
dacb8a1
End of training
Browse files- all_results.json +18 -0
- eval_results.json +13 -0
- train_results.json +8 -0
- trainer_state.json +82 -0
all_results.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.9258982035928144,
|
| 4 |
+
"eval_f1": 0.9438457175269428,
|
| 5 |
+
"eval_loss": 0.22528469562530518,
|
| 6 |
+
"eval_precision": 0.9486887115165337,
|
| 7 |
+
"eval_recall": 0.9390519187358917,
|
| 8 |
+
"eval_roc_auc": 0.9195259593679458,
|
| 9 |
+
"eval_runtime": 1.9774,
|
| 10 |
+
"eval_samples": 1336,
|
| 11 |
+
"eval_samples_per_second": 675.629,
|
| 12 |
+
"eval_steps_per_second": 5.563,
|
| 13 |
+
"train_loss": 0.19893105635567318,
|
| 14 |
+
"train_runtime": 118.1813,
|
| 15 |
+
"train_samples": 5342,
|
| 16 |
+
"train_samples_per_second": 135.605,
|
| 17 |
+
"train_steps_per_second": 1.066
|
| 18 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.9258982035928144,
|
| 4 |
+
"eval_f1": 0.9438457175269428,
|
| 5 |
+
"eval_loss": 0.22528469562530518,
|
| 6 |
+
"eval_precision": 0.9486887115165337,
|
| 7 |
+
"eval_recall": 0.9390519187358917,
|
| 8 |
+
"eval_roc_auc": 0.9195259593679458,
|
| 9 |
+
"eval_runtime": 1.9774,
|
| 10 |
+
"eval_samples": 1336,
|
| 11 |
+
"eval_samples_per_second": 675.629,
|
| 12 |
+
"eval_steps_per_second": 5.563
|
| 13 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 0.19893105635567318,
|
| 4 |
+
"train_runtime": 118.1813,
|
| 5 |
+
"train_samples": 5342,
|
| 6 |
+
"train_samples_per_second": 135.605,
|
| 7 |
+
"train_steps_per_second": 1.066
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"global_step": 126,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 1.0,
|
| 12 |
+
"learning_rate": 6.825396825396825e-05,
|
| 13 |
+
"loss": 0.3612,
|
| 14 |
+
"step": 42
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.0,
|
| 18 |
+
"eval_accuracy": 0.905688622754491,
|
| 19 |
+
"eval_f1": 0.9257075471698112,
|
| 20 |
+
"eval_loss": 0.23578055202960968,
|
| 21 |
+
"eval_precision": 0.9691358024691358,
|
| 22 |
+
"eval_recall": 0.8860045146726863,
|
| 23 |
+
"eval_roc_auc": 0.9152244795585653,
|
| 24 |
+
"eval_runtime": 1.9731,
|
| 25 |
+
"eval_samples_per_second": 677.118,
|
| 26 |
+
"eval_steps_per_second": 5.575,
|
| 27 |
+
"step": 42
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 2.0,
|
| 31 |
+
"learning_rate": 3.492063492063492e-05,
|
| 32 |
+
"loss": 0.1616,
|
| 33 |
+
"step": 84
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 2.0,
|
| 37 |
+
"eval_accuracy": 0.9296407185628742,
|
| 38 |
+
"eval_f1": 0.9468325791855204,
|
| 39 |
+
"eval_loss": 0.19503971934318542,
|
| 40 |
+
"eval_precision": 0.9489795918367347,
|
| 41 |
+
"eval_recall": 0.9446952595936795,
|
| 42 |
+
"eval_roc_auc": 0.9223476297968398,
|
| 43 |
+
"eval_runtime": 1.9788,
|
| 44 |
+
"eval_samples_per_second": 675.158,
|
| 45 |
+
"eval_steps_per_second": 5.559,
|
| 46 |
+
"step": 84
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 3.0,
|
| 50 |
+
"learning_rate": 1.5873015873015873e-06,
|
| 51 |
+
"loss": 0.074,
|
| 52 |
+
"step": 126
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 3.0,
|
| 56 |
+
"eval_accuracy": 0.9258982035928144,
|
| 57 |
+
"eval_f1": 0.9438457175269428,
|
| 58 |
+
"eval_loss": 0.22528469562530518,
|
| 59 |
+
"eval_precision": 0.9486887115165337,
|
| 60 |
+
"eval_recall": 0.9390519187358917,
|
| 61 |
+
"eval_roc_auc": 0.9195259593679458,
|
| 62 |
+
"eval_runtime": 1.9809,
|
| 63 |
+
"eval_samples_per_second": 674.439,
|
| 64 |
+
"eval_steps_per_second": 5.553,
|
| 65 |
+
"step": 126
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 3.0,
|
| 69 |
+
"step": 126,
|
| 70 |
+
"total_flos": 2108308886599680.0,
|
| 71 |
+
"train_loss": 0.19893105635567318,
|
| 72 |
+
"train_runtime": 118.1813,
|
| 73 |
+
"train_samples_per_second": 135.605,
|
| 74 |
+
"train_steps_per_second": 1.066
|
| 75 |
+
}
|
| 76 |
+
],
|
| 77 |
+
"max_steps": 126,
|
| 78 |
+
"num_train_epochs": 3,
|
| 79 |
+
"total_flos": 2108308886599680.0,
|
| 80 |
+
"trial_name": null,
|
| 81 |
+
"trial_params": null
|
| 82 |
+
}
|