Commit
•
69e26b6
1
Parent(s):
e189a7a
End of training
Browse files- README.md +5 -5
- all_results.json +16 -0
- test_results.json +11 -0
- train_results.json +8 -0
- trainer_state.json +524 -0
README.md
CHANGED
|
@@ -18,11 +18,11 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [vinai/phobert-base-v2](https://huggingface.co/vinai/phobert-base-v2) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
-
- Accuracy: 0.
|
| 23 |
-
- F1 Score: 0.
|
| 24 |
-
- Recall: 0.
|
| 25 |
-
- Precision: 0.
|
| 26 |
|
| 27 |
## Model description
|
| 28 |
|
|
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [vinai/phobert-base-v2](https://huggingface.co/vinai/phobert-base-v2) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.3163
|
| 22 |
+
- Accuracy: 0.9317
|
| 23 |
+
- F1 Score: 0.8875
|
| 24 |
+
- Recall: 0.8777
|
| 25 |
+
- Precision: 0.9015
|
| 26 |
|
| 27 |
## Model description
|
| 28 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 27.37,
|
| 3 |
+
"eval_accuracy": 0.9317057076872319,
|
| 4 |
+
"eval_f1_score": 0.8875182540327614,
|
| 5 |
+
"eval_loss": 0.3163454234600067,
|
| 6 |
+
"eval_precision": 0.9014604415235737,
|
| 7 |
+
"eval_recall": 0.8777053165164634,
|
| 8 |
+
"eval_runtime": 5.1648,
|
| 9 |
+
"eval_samples_per_second": 586.855,
|
| 10 |
+
"eval_steps_per_second": 9.294,
|
| 11 |
+
"total_flos": 5414987099379456.0,
|
| 12 |
+
"train_loss": 0.2901941172893231,
|
| 13 |
+
"train_runtime": 2142.8815,
|
| 14 |
+
"train_samples_per_second": 238.931,
|
| 15 |
+
"train_steps_per_second": 1.867
|
| 16 |
+
}
|
test_results.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 27.37,
|
| 3 |
+
"eval_accuracy": 0.9317057076872319,
|
| 4 |
+
"eval_f1_score": 0.8875182540327614,
|
| 5 |
+
"eval_loss": 0.3163454234600067,
|
| 6 |
+
"eval_precision": 0.9014604415235737,
|
| 7 |
+
"eval_recall": 0.8777053165164634,
|
| 8 |
+
"eval_runtime": 5.1648,
|
| 9 |
+
"eval_samples_per_second": 586.855,
|
| 10 |
+
"eval_steps_per_second": 9.294
|
| 11 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 27.37,
|
| 3 |
+
"total_flos": 5414987099379456.0,
|
| 4 |
+
"train_loss": 0.2901941172893231,
|
| 5 |
+
"train_runtime": 2142.8815,
|
| 6 |
+
"train_samples_per_second": 238.931,
|
| 7 |
+
"train_steps_per_second": 1.867
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,524 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8875182540327614,
|
| 3 |
+
"best_model_checkpoint": "cls_comment-phobert-base-v2-v2.2/checkpoint-2100",
|
| 4 |
+
"epoch": 27.36842105263158,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 2600,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.05,
|
| 13 |
+
"grad_norm": 1.7893420457839966,
|
| 14 |
+
"learning_rate": 2.5e-06,
|
| 15 |
+
"loss": 1.663,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.05,
|
| 20 |
+
"eval_accuracy": 0.5041240514681623,
|
| 21 |
+
"eval_f1_score": 0.11172040652189807,
|
| 22 |
+
"eval_loss": 1.4870415925979614,
|
| 23 |
+
"eval_precision": 0.08402067524469371,
|
| 24 |
+
"eval_recall": 0.16666666666666666,
|
| 25 |
+
"eval_runtime": 5.0768,
|
| 26 |
+
"eval_samples_per_second": 597.028,
|
| 27 |
+
"eval_steps_per_second": 9.455,
|
| 28 |
+
"step": 100
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 2.11,
|
| 32 |
+
"grad_norm": 2.4307305812835693,
|
| 33 |
+
"learning_rate": 5e-06,
|
| 34 |
+
"loss": 1.294,
|
| 35 |
+
"step": 200
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 2.11,
|
| 39 |
+
"eval_accuracy": 0.6974595842956121,
|
| 40 |
+
"eval_f1_score": 0.39113317224801986,
|
| 41 |
+
"eval_loss": 0.9955642223358154,
|
| 42 |
+
"eval_precision": 0.49018024013865347,
|
| 43 |
+
"eval_recall": 0.3899919974680958,
|
| 44 |
+
"eval_runtime": 5.1701,
|
| 45 |
+
"eval_samples_per_second": 586.259,
|
| 46 |
+
"eval_steps_per_second": 9.284,
|
| 47 |
+
"step": 200
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 3.16,
|
| 51 |
+
"grad_norm": 3.8080785274505615,
|
| 52 |
+
"learning_rate": 7.500000000000001e-06,
|
| 53 |
+
"loss": 0.898,
|
| 54 |
+
"step": 300
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 3.16,
|
| 58 |
+
"eval_accuracy": 0.8231606730451996,
|
| 59 |
+
"eval_f1_score": 0.5499484227307703,
|
| 60 |
+
"eval_loss": 0.6778652667999268,
|
| 61 |
+
"eval_precision": 0.535367810981718,
|
| 62 |
+
"eval_recall": 0.5697240354177526,
|
| 63 |
+
"eval_runtime": 5.2797,
|
| 64 |
+
"eval_samples_per_second": 574.084,
|
| 65 |
+
"eval_steps_per_second": 9.091,
|
| 66 |
+
"step": 300
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 4.21,
|
| 70 |
+
"grad_norm": 4.932918548583984,
|
| 71 |
+
"learning_rate": 1e-05,
|
| 72 |
+
"loss": 0.6411,
|
| 73 |
+
"step": 400
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 4.21,
|
| 77 |
+
"eval_accuracy": 0.8568129330254042,
|
| 78 |
+
"eval_f1_score": 0.5739966477156212,
|
| 79 |
+
"eval_loss": 0.5163660645484924,
|
| 80 |
+
"eval_precision": 0.5613296135162277,
|
| 81 |
+
"eval_recall": 0.5895128268802295,
|
| 82 |
+
"eval_runtime": 5.1719,
|
| 83 |
+
"eval_samples_per_second": 586.047,
|
| 84 |
+
"eval_steps_per_second": 9.281,
|
| 85 |
+
"step": 400
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 5.26,
|
| 89 |
+
"grad_norm": 3.761213541030884,
|
| 90 |
+
"learning_rate": 9.722222222222223e-06,
|
| 91 |
+
"loss": 0.5031,
|
| 92 |
+
"step": 500
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 5.26,
|
| 96 |
+
"eval_accuracy": 0.8937644341801386,
|
| 97 |
+
"eval_f1_score": 0.7180655496783576,
|
| 98 |
+
"eval_loss": 0.4105552136898041,
|
| 99 |
+
"eval_precision": 0.7319397223704595,
|
| 100 |
+
"eval_recall": 0.7114016304859992,
|
| 101 |
+
"eval_runtime": 5.1958,
|
| 102 |
+
"eval_samples_per_second": 583.36,
|
| 103 |
+
"eval_steps_per_second": 9.238,
|
| 104 |
+
"step": 500
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 6.32,
|
| 108 |
+
"grad_norm": 5.329339504241943,
|
| 109 |
+
"learning_rate": 9.444444444444445e-06,
|
| 110 |
+
"loss": 0.38,
|
| 111 |
+
"step": 600
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 6.32,
|
| 115 |
+
"eval_accuracy": 0.9096007918178819,
|
| 116 |
+
"eval_f1_score": 0.8325922485904789,
|
| 117 |
+
"eval_loss": 0.3474419116973877,
|
| 118 |
+
"eval_precision": 0.8738654107630431,
|
| 119 |
+
"eval_recall": 0.8144519804951473,
|
| 120 |
+
"eval_runtime": 5.1729,
|
| 121 |
+
"eval_samples_per_second": 585.934,
|
| 122 |
+
"eval_steps_per_second": 9.279,
|
| 123 |
+
"step": 600
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 7.37,
|
| 127 |
+
"grad_norm": 5.044618129730225,
|
| 128 |
+
"learning_rate": 9.166666666666666e-06,
|
| 129 |
+
"loss": 0.2927,
|
| 130 |
+
"step": 700
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 7.37,
|
| 134 |
+
"eval_accuracy": 0.9142197294622237,
|
| 135 |
+
"eval_f1_score": 0.8598244657605173,
|
| 136 |
+
"eval_loss": 0.3109701871871948,
|
| 137 |
+
"eval_precision": 0.8809676600022426,
|
| 138 |
+
"eval_recall": 0.8455076124995076,
|
| 139 |
+
"eval_runtime": 6.0257,
|
| 140 |
+
"eval_samples_per_second": 503.009,
|
| 141 |
+
"eval_steps_per_second": 7.966,
|
| 142 |
+
"step": 700
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 8.42,
|
| 146 |
+
"grad_norm": 4.748721599578857,
|
| 147 |
+
"learning_rate": 8.888888888888888e-06,
|
| 148 |
+
"loss": 0.2532,
|
| 149 |
+
"step": 800
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 8.42,
|
| 153 |
+
"eval_accuracy": 0.9188386671065655,
|
| 154 |
+
"eval_f1_score": 0.8702202017241888,
|
| 155 |
+
"eval_loss": 0.3045769929885864,
|
| 156 |
+
"eval_precision": 0.8880655151226781,
|
| 157 |
+
"eval_recall": 0.85512094950349,
|
| 158 |
+
"eval_runtime": 5.1742,
|
| 159 |
+
"eval_samples_per_second": 585.795,
|
| 160 |
+
"eval_steps_per_second": 9.277,
|
| 161 |
+
"step": 800
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 9.47,
|
| 165 |
+
"grad_norm": 4.976996898651123,
|
| 166 |
+
"learning_rate": 8.611111111111112e-06,
|
| 167 |
+
"loss": 0.2049,
|
| 168 |
+
"step": 900
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 9.47,
|
| 172 |
+
"eval_accuracy": 0.9218079841636424,
|
| 173 |
+
"eval_f1_score": 0.8688677581871923,
|
| 174 |
+
"eval_loss": 0.28507497906684875,
|
| 175 |
+
"eval_precision": 0.890238230846033,
|
| 176 |
+
"eval_recall": 0.8539333664213857,
|
| 177 |
+
"eval_runtime": 5.1661,
|
| 178 |
+
"eval_samples_per_second": 586.706,
|
| 179 |
+
"eval_steps_per_second": 9.291,
|
| 180 |
+
"step": 900
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 10.53,
|
| 184 |
+
"grad_norm": 5.230645656585693,
|
| 185 |
+
"learning_rate": 8.333333333333334e-06,
|
| 186 |
+
"loss": 0.1785,
|
| 187 |
+
"step": 1000
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 10.53,
|
| 191 |
+
"eval_accuracy": 0.9251072253381722,
|
| 192 |
+
"eval_f1_score": 0.8768747602558463,
|
| 193 |
+
"eval_loss": 0.2801537811756134,
|
| 194 |
+
"eval_precision": 0.9044956542270928,
|
| 195 |
+
"eval_recall": 0.8561455297131971,
|
| 196 |
+
"eval_runtime": 5.2425,
|
| 197 |
+
"eval_samples_per_second": 578.155,
|
| 198 |
+
"eval_steps_per_second": 9.156,
|
| 199 |
+
"step": 1000
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 11.58,
|
| 203 |
+
"grad_norm": 5.533747673034668,
|
| 204 |
+
"learning_rate": 8.055555555555557e-06,
|
| 205 |
+
"loss": 0.1511,
|
| 206 |
+
"step": 1100
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 11.58,
|
| 210 |
+
"eval_accuracy": 0.9231276806334543,
|
| 211 |
+
"eval_f1_score": 0.8744252567641876,
|
| 212 |
+
"eval_loss": 0.28748995065689087,
|
| 213 |
+
"eval_precision": 0.8770259775941726,
|
| 214 |
+
"eval_recall": 0.8748474059729406,
|
| 215 |
+
"eval_runtime": 5.2089,
|
| 216 |
+
"eval_samples_per_second": 581.891,
|
| 217 |
+
"eval_steps_per_second": 9.215,
|
| 218 |
+
"step": 1100
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 12.63,
|
| 222 |
+
"grad_norm": 5.897303104400635,
|
| 223 |
+
"learning_rate": 7.77777777777778e-06,
|
| 224 |
+
"loss": 0.1392,
|
| 225 |
+
"step": 1200
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 12.63,
|
| 229 |
+
"eval_accuracy": 0.9264269218079841,
|
| 230 |
+
"eval_f1_score": 0.8774620123424183,
|
| 231 |
+
"eval_loss": 0.2811359465122223,
|
| 232 |
+
"eval_precision": 0.9005448213619175,
|
| 233 |
+
"eval_recall": 0.8596724578451488,
|
| 234 |
+
"eval_runtime": 5.1974,
|
| 235 |
+
"eval_samples_per_second": 583.175,
|
| 236 |
+
"eval_steps_per_second": 9.235,
|
| 237 |
+
"step": 1200
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 13.68,
|
| 241 |
+
"grad_norm": 3.5150811672210693,
|
| 242 |
+
"learning_rate": 7.500000000000001e-06,
|
| 243 |
+
"loss": 0.1166,
|
| 244 |
+
"step": 1300
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 13.68,
|
| 248 |
+
"eval_accuracy": 0.9247773012207192,
|
| 249 |
+
"eval_f1_score": 0.875133246925271,
|
| 250 |
+
"eval_loss": 0.27574026584625244,
|
| 251 |
+
"eval_precision": 0.8785572499470087,
|
| 252 |
+
"eval_recall": 0.8746042908646392,
|
| 253 |
+
"eval_runtime": 5.198,
|
| 254 |
+
"eval_samples_per_second": 583.104,
|
| 255 |
+
"eval_steps_per_second": 9.234,
|
| 256 |
+
"step": 1300
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 14.74,
|
| 260 |
+
"grad_norm": 5.184170722961426,
|
| 261 |
+
"learning_rate": 7.222222222222223e-06,
|
| 262 |
+
"loss": 0.1087,
|
| 263 |
+
"step": 1400
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 14.74,
|
| 267 |
+
"eval_accuracy": 0.9257670735730782,
|
| 268 |
+
"eval_f1_score": 0.8803669741747352,
|
| 269 |
+
"eval_loss": 0.272718608379364,
|
| 270 |
+
"eval_precision": 0.8857591375966419,
|
| 271 |
+
"eval_recall": 0.8761427115927698,
|
| 272 |
+
"eval_runtime": 5.2326,
|
| 273 |
+
"eval_samples_per_second": 579.256,
|
| 274 |
+
"eval_steps_per_second": 9.173,
|
| 275 |
+
"step": 1400
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 15.79,
|
| 279 |
+
"grad_norm": 2.6620497703552246,
|
| 280 |
+
"learning_rate": 6.944444444444445e-06,
|
| 281 |
+
"loss": 0.0918,
|
| 282 |
+
"step": 1500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 15.79,
|
| 286 |
+
"eval_accuracy": 0.9284064665127021,
|
| 287 |
+
"eval_f1_score": 0.8829680164086238,
|
| 288 |
+
"eval_loss": 0.286248117685318,
|
| 289 |
+
"eval_precision": 0.8987862948246464,
|
| 290 |
+
"eval_recall": 0.8711650603773539,
|
| 291 |
+
"eval_runtime": 5.184,
|
| 292 |
+
"eval_samples_per_second": 584.68,
|
| 293 |
+
"eval_steps_per_second": 9.259,
|
| 294 |
+
"step": 1500
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 16.84,
|
| 298 |
+
"grad_norm": 1.70383882522583,
|
| 299 |
+
"learning_rate": 6.666666666666667e-06,
|
| 300 |
+
"loss": 0.0824,
|
| 301 |
+
"step": 1600
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 16.84,
|
| 305 |
+
"eval_accuracy": 0.929066314747608,
|
| 306 |
+
"eval_f1_score": 0.8832576833606276,
|
| 307 |
+
"eval_loss": 0.2915453612804413,
|
| 308 |
+
"eval_precision": 0.9008601442216951,
|
| 309 |
+
"eval_recall": 0.868938029158805,
|
| 310 |
+
"eval_runtime": 5.1614,
|
| 311 |
+
"eval_samples_per_second": 587.245,
|
| 312 |
+
"eval_steps_per_second": 9.3,
|
| 313 |
+
"step": 1600
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 17.89,
|
| 317 |
+
"grad_norm": 7.806049823760986,
|
| 318 |
+
"learning_rate": 6.3888888888888885e-06,
|
| 319 |
+
"loss": 0.0745,
|
| 320 |
+
"step": 1700
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 17.89,
|
| 324 |
+
"eval_accuracy": 0.929066314747608,
|
| 325 |
+
"eval_f1_score": 0.8796971687211418,
|
| 326 |
+
"eval_loss": 0.2994127869606018,
|
| 327 |
+
"eval_precision": 0.8847125167911893,
|
| 328 |
+
"eval_recall": 0.8796360078693516,
|
| 329 |
+
"eval_runtime": 5.1775,
|
| 330 |
+
"eval_samples_per_second": 585.421,
|
| 331 |
+
"eval_steps_per_second": 9.271,
|
| 332 |
+
"step": 1700
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 18.95,
|
| 336 |
+
"grad_norm": 9.288145065307617,
|
| 337 |
+
"learning_rate": 6.111111111111112e-06,
|
| 338 |
+
"loss": 0.0743,
|
| 339 |
+
"step": 1800
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 18.95,
|
| 343 |
+
"eval_accuracy": 0.9254371494556252,
|
| 344 |
+
"eval_f1_score": 0.8782515232850638,
|
| 345 |
+
"eval_loss": 0.3092294931411743,
|
| 346 |
+
"eval_precision": 0.8909710349230974,
|
| 347 |
+
"eval_recall": 0.8686312796266131,
|
| 348 |
+
"eval_runtime": 5.2161,
|
| 349 |
+
"eval_samples_per_second": 581.085,
|
| 350 |
+
"eval_steps_per_second": 9.202,
|
| 351 |
+
"step": 1800
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 20.0,
|
| 355 |
+
"grad_norm": 0.29120656847953796,
|
| 356 |
+
"learning_rate": 5.833333333333334e-06,
|
| 357 |
+
"loss": 0.0636,
|
| 358 |
+
"step": 1900
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 20.0,
|
| 362 |
+
"eval_accuracy": 0.929066314747608,
|
| 363 |
+
"eval_f1_score": 0.8810930807565414,
|
| 364 |
+
"eval_loss": 0.3142439126968384,
|
| 365 |
+
"eval_precision": 0.8915600488637722,
|
| 366 |
+
"eval_recall": 0.8743221351307482,
|
| 367 |
+
"eval_runtime": 5.1856,
|
| 368 |
+
"eval_samples_per_second": 584.505,
|
| 369 |
+
"eval_steps_per_second": 9.256,
|
| 370 |
+
"step": 1900
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 21.05,
|
| 374 |
+
"grad_norm": 3.0424692630767822,
|
| 375 |
+
"learning_rate": 5.555555555555557e-06,
|
| 376 |
+
"loss": 0.0605,
|
| 377 |
+
"step": 2000
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 21.05,
|
| 381 |
+
"eval_accuracy": 0.929066314747608,
|
| 382 |
+
"eval_f1_score": 0.8823037485745876,
|
| 383 |
+
"eval_loss": 0.30989953875541687,
|
| 384 |
+
"eval_precision": 0.8974412044321375,
|
| 385 |
+
"eval_recall": 0.8700202850817763,
|
| 386 |
+
"eval_runtime": 5.2587,
|
| 387 |
+
"eval_samples_per_second": 576.383,
|
| 388 |
+
"eval_steps_per_second": 9.128,
|
| 389 |
+
"step": 2000
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 22.11,
|
| 393 |
+
"grad_norm": 0.18823295831680298,
|
| 394 |
+
"learning_rate": 5.2777777777777785e-06,
|
| 395 |
+
"loss": 0.0501,
|
| 396 |
+
"step": 2100
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 22.11,
|
| 400 |
+
"eval_accuracy": 0.9317057076872319,
|
| 401 |
+
"eval_f1_score": 0.8875182540327614,
|
| 402 |
+
"eval_loss": 0.3163454234600067,
|
| 403 |
+
"eval_precision": 0.9014604415235737,
|
| 404 |
+
"eval_recall": 0.8777053165164634,
|
| 405 |
+
"eval_runtime": 5.1733,
|
| 406 |
+
"eval_samples_per_second": 585.895,
|
| 407 |
+
"eval_steps_per_second": 9.278,
|
| 408 |
+
"step": 2100
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 23.16,
|
| 412 |
+
"grad_norm": 0.22511476278305054,
|
| 413 |
+
"learning_rate": 5e-06,
|
| 414 |
+
"loss": 0.0519,
|
| 415 |
+
"step": 2200
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 23.16,
|
| 419 |
+
"eval_accuracy": 0.929726162982514,
|
| 420 |
+
"eval_f1_score": 0.883696747826972,
|
| 421 |
+
"eval_loss": 0.3289995491504669,
|
| 422 |
+
"eval_precision": 0.9010520871368185,
|
| 423 |
+
"eval_recall": 0.8692223543133784,
|
| 424 |
+
"eval_runtime": 5.2183,
|
| 425 |
+
"eval_samples_per_second": 580.844,
|
| 426 |
+
"eval_steps_per_second": 9.198,
|
| 427 |
+
"step": 2200
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 24.21,
|
| 431 |
+
"grad_norm": 8.104951858520508,
|
| 432 |
+
"learning_rate": 4.722222222222222e-06,
|
| 433 |
+
"loss": 0.0464,
|
| 434 |
+
"step": 2300
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 24.21,
|
| 438 |
+
"eval_accuracy": 0.9274166941603431,
|
| 439 |
+
"eval_f1_score": 0.8804834286728226,
|
| 440 |
+
"eval_loss": 0.34061628580093384,
|
| 441 |
+
"eval_precision": 0.8871850911314553,
|
| 442 |
+
"eval_recall": 0.8771793667200236,
|
| 443 |
+
"eval_runtime": 5.1789,
|
| 444 |
+
"eval_samples_per_second": 585.258,
|
| 445 |
+
"eval_steps_per_second": 9.268,
|
| 446 |
+
"step": 2300
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 25.26,
|
| 450 |
+
"grad_norm": 2.2458958625793457,
|
| 451 |
+
"learning_rate": 4.444444444444444e-06,
|
| 452 |
+
"loss": 0.0432,
|
| 453 |
+
"step": 2400
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 25.26,
|
| 457 |
+
"eval_accuracy": 0.9284064665127021,
|
| 458 |
+
"eval_f1_score": 0.8809944439022924,
|
| 459 |
+
"eval_loss": 0.3304508328437805,
|
| 460 |
+
"eval_precision": 0.8875581681302606,
|
| 461 |
+
"eval_recall": 0.8774637238579275,
|
| 462 |
+
"eval_runtime": 5.2785,
|
| 463 |
+
"eval_samples_per_second": 574.218,
|
| 464 |
+
"eval_steps_per_second": 9.094,
|
| 465 |
+
"step": 2400
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 26.32,
|
| 469 |
+
"grad_norm": 8.245588302612305,
|
| 470 |
+
"learning_rate": 4.166666666666667e-06,
|
| 471 |
+
"loss": 0.0404,
|
| 472 |
+
"step": 2500
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 26.32,
|
| 476 |
+
"eval_accuracy": 0.9293962388650611,
|
| 477 |
+
"eval_f1_score": 0.8825762014276358,
|
| 478 |
+
"eval_loss": 0.3377806544303894,
|
| 479 |
+
"eval_precision": 0.890131895776857,
|
| 480 |
+
"eval_recall": 0.8785119382363503,
|
| 481 |
+
"eval_runtime": 5.1758,
|
| 482 |
+
"eval_samples_per_second": 585.606,
|
| 483 |
+
"eval_steps_per_second": 9.274,
|
| 484 |
+
"step": 2500
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 27.37,
|
| 488 |
+
"grad_norm": 5.161190032958984,
|
| 489 |
+
"learning_rate": 3.88888888888889e-06,
|
| 490 |
+
"loss": 0.0416,
|
| 491 |
+
"step": 2600
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 27.37,
|
| 495 |
+
"eval_accuracy": 0.9284064665127021,
|
| 496 |
+
"eval_f1_score": 0.8829688694414259,
|
| 497 |
+
"eval_loss": 0.343578040599823,
|
| 498 |
+
"eval_precision": 0.897684321914049,
|
| 499 |
+
"eval_recall": 0.8726474609617948,
|
| 500 |
+
"eval_runtime": 5.2526,
|
| 501 |
+
"eval_samples_per_second": 577.048,
|
| 502 |
+
"eval_steps_per_second": 9.138,
|
| 503 |
+
"step": 2600
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 27.37,
|
| 507 |
+
"step": 2600,
|
| 508 |
+
"total_flos": 5414987099379456.0,
|
| 509 |
+
"train_loss": 0.2901941172893231,
|
| 510 |
+
"train_runtime": 2142.8815,
|
| 511 |
+
"train_samples_per_second": 238.931,
|
| 512 |
+
"train_steps_per_second": 1.867
|
| 513 |
+
}
|
| 514 |
+
],
|
| 515 |
+
"logging_steps": 100,
|
| 516 |
+
"max_steps": 4000,
|
| 517 |
+
"num_input_tokens_seen": 0,
|
| 518 |
+
"num_train_epochs": 43,
|
| 519 |
+
"save_steps": 100,
|
| 520 |
+
"total_flos": 5414987099379456.0,
|
| 521 |
+
"train_batch_size": 64,
|
| 522 |
+
"trial_name": null,
|
| 523 |
+
"trial_params": null
|
| 524 |
+
}
|