End of training
Browse files- README.md +2 -2
- all_results.json +8 -0
- test_results.json +8 -0
- trainer_state.json +1502 -0
README.md
CHANGED
|
@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
-
- Accuracy: 0.
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.5588
|
| 22 |
+
- Accuracy: 0.8973
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 19.040078125,
|
| 3 |
+
"eval_accuracy": 0.8973354231974922,
|
| 4 |
+
"eval_loss": 0.5587517619132996,
|
| 5 |
+
"eval_runtime": 352.6874,
|
| 6 |
+
"eval_samples_per_second": 3.618,
|
| 7 |
+
"eval_steps_per_second": 1.809
|
| 8 |
+
}
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 19.040078125,
|
| 3 |
+
"eval_accuracy": 0.8973354231974922,
|
| 4 |
+
"eval_loss": 0.5587517619132996,
|
| 5 |
+
"eval_runtime": 352.6874,
|
| 6 |
+
"eval_samples_per_second": 3.618,
|
| 7 |
+
"eval_steps_per_second": 1.809
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1502 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9929411764705882,
|
| 3 |
+
"best_model_checkpoint": "/media/cse/HDD/Shawon/shawon/MY DATA/VideoMAE_BdSLW60_FrameRateCorrected_withAug_100/checkpoint-13466",
|
| 4 |
+
"epoch": 19.040078125,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 17955,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.004464285714285714,
|
| 13 |
+
"grad_norm": 11.12140941619873,
|
| 14 |
+
"learning_rate": 2.2321428571428573e-06,
|
| 15 |
+
"loss": 4.1557,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.008928571428571428,
|
| 20 |
+
"grad_norm": 10.578296661376953,
|
| 21 |
+
"learning_rate": 4.464285714285715e-06,
|
| 22 |
+
"loss": 4.1159,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.013392857142857142,
|
| 27 |
+
"grad_norm": 9.035299301147461,
|
| 28 |
+
"learning_rate": 6.696428571428572e-06,
|
| 29 |
+
"loss": 4.0848,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.017857142857142856,
|
| 34 |
+
"grad_norm": 9.214325904846191,
|
| 35 |
+
"learning_rate": 8.92857142857143e-06,
|
| 36 |
+
"loss": 4.0703,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.022321428571428572,
|
| 41 |
+
"grad_norm": 8.834626197814941,
|
| 42 |
+
"learning_rate": 1.1160714285714287e-05,
|
| 43 |
+
"loss": 4.0688,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.026785714285714284,
|
| 48 |
+
"grad_norm": 10.655806541442871,
|
| 49 |
+
"learning_rate": 1.3392857142857144e-05,
|
| 50 |
+
"loss": 3.8577,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.03125,
|
| 55 |
+
"grad_norm": 11.894658088684082,
|
| 56 |
+
"learning_rate": 1.5625e-05,
|
| 57 |
+
"loss": 3.4927,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.03571428571428571,
|
| 62 |
+
"grad_norm": 13.875555992126465,
|
| 63 |
+
"learning_rate": 1.785714285714286e-05,
|
| 64 |
+
"loss": 3.0699,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.040044642857142855,
|
| 69 |
+
"eval_accuracy": 0.4752941176470588,
|
| 70 |
+
"eval_loss": 2.454066514968872,
|
| 71 |
+
"eval_runtime": 290.1198,
|
| 72 |
+
"eval_samples_per_second": 2.93,
|
| 73 |
+
"eval_steps_per_second": 1.465,
|
| 74 |
+
"step": 897
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.0001004464285714,
|
| 78 |
+
"grad_norm": 11.632246017456055,
|
| 79 |
+
"learning_rate": 2.0089285714285717e-05,
|
| 80 |
+
"loss": 2.5881,
|
| 81 |
+
"step": 900
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.0045647321428572,
|
| 85 |
+
"grad_norm": 15.39003849029541,
|
| 86 |
+
"learning_rate": 2.2321428571428575e-05,
|
| 87 |
+
"loss": 2.2052,
|
| 88 |
+
"step": 1000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.0090290178571428,
|
| 92 |
+
"grad_norm": 17.561227798461914,
|
| 93 |
+
"learning_rate": 2.455357142857143e-05,
|
| 94 |
+
"loss": 1.8017,
|
| 95 |
+
"step": 1100
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.0134933035714286,
|
| 99 |
+
"grad_norm": 16.368633270263672,
|
| 100 |
+
"learning_rate": 2.6785714285714288e-05,
|
| 101 |
+
"loss": 1.5213,
|
| 102 |
+
"step": 1200
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.0179575892857142,
|
| 106 |
+
"grad_norm": 18.419261932373047,
|
| 107 |
+
"learning_rate": 2.9017857142857146e-05,
|
| 108 |
+
"loss": 1.1462,
|
| 109 |
+
"step": 1300
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.022421875,
|
| 113 |
+
"grad_norm": 14.493526458740234,
|
| 114 |
+
"learning_rate": 3.125e-05,
|
| 115 |
+
"loss": 1.0545,
|
| 116 |
+
"step": 1400
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.0268861607142856,
|
| 120 |
+
"grad_norm": 15.404373168945312,
|
| 121 |
+
"learning_rate": 3.348214285714286e-05,
|
| 122 |
+
"loss": 0.7972,
|
| 123 |
+
"step": 1500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.0313504464285714,
|
| 127 |
+
"grad_norm": 7.37654972076416,
|
| 128 |
+
"learning_rate": 3.571428571428572e-05,
|
| 129 |
+
"loss": 0.6743,
|
| 130 |
+
"step": 1600
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.0358147321428572,
|
| 134 |
+
"grad_norm": 17.836456298828125,
|
| 135 |
+
"learning_rate": 3.794642857142857e-05,
|
| 136 |
+
"loss": 0.6366,
|
| 137 |
+
"step": 1700
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.0400558035714287,
|
| 141 |
+
"eval_accuracy": 0.84,
|
| 142 |
+
"eval_loss": 0.6831679344177246,
|
| 143 |
+
"eval_runtime": 295.5183,
|
| 144 |
+
"eval_samples_per_second": 2.876,
|
| 145 |
+
"eval_steps_per_second": 1.438,
|
| 146 |
+
"step": 1795
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 2.000200892857143,
|
| 150 |
+
"grad_norm": 27.33871078491211,
|
| 151 |
+
"learning_rate": 4.017857142857143e-05,
|
| 152 |
+
"loss": 0.6165,
|
| 153 |
+
"step": 1800
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.0046651785714285,
|
| 157 |
+
"grad_norm": 1.4543864727020264,
|
| 158 |
+
"learning_rate": 4.2410714285714285e-05,
|
| 159 |
+
"loss": 0.4179,
|
| 160 |
+
"step": 1900
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.0091294642857145,
|
| 164 |
+
"grad_norm": 7.2733659744262695,
|
| 165 |
+
"learning_rate": 4.464285714285715e-05,
|
| 166 |
+
"loss": 0.4156,
|
| 167 |
+
"step": 2000
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.01359375,
|
| 171 |
+
"grad_norm": 21.995115280151367,
|
| 172 |
+
"learning_rate": 4.6875e-05,
|
| 173 |
+
"loss": 0.3666,
|
| 174 |
+
"step": 2100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.0180580357142857,
|
| 178 |
+
"grad_norm": 19.265806198120117,
|
| 179 |
+
"learning_rate": 4.910714285714286e-05,
|
| 180 |
+
"loss": 0.3751,
|
| 181 |
+
"step": 2200
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.0225223214285712,
|
| 185 |
+
"grad_norm": 26.048490524291992,
|
| 186 |
+
"learning_rate": 4.985119047619048e-05,
|
| 187 |
+
"loss": 0.3401,
|
| 188 |
+
"step": 2300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.0269866071428573,
|
| 192 |
+
"grad_norm": 26.414731979370117,
|
| 193 |
+
"learning_rate": 4.960317460317461e-05,
|
| 194 |
+
"loss": 0.2955,
|
| 195 |
+
"step": 2400
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.031450892857143,
|
| 199 |
+
"grad_norm": 17.34372901916504,
|
| 200 |
+
"learning_rate": 4.9355158730158735e-05,
|
| 201 |
+
"loss": 0.2859,
|
| 202 |
+
"step": 2500
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.0359151785714285,
|
| 206 |
+
"grad_norm": 3.029252767562866,
|
| 207 |
+
"learning_rate": 4.910714285714286e-05,
|
| 208 |
+
"loss": 0.2253,
|
| 209 |
+
"step": 2600
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.0400669642857143,
|
| 213 |
+
"eval_accuracy": 0.9023529411764706,
|
| 214 |
+
"eval_loss": 0.3464316725730896,
|
| 215 |
+
"eval_runtime": 282.6757,
|
| 216 |
+
"eval_samples_per_second": 3.007,
|
| 217 |
+
"eval_steps_per_second": 1.503,
|
| 218 |
+
"step": 2693
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.000301339285714,
|
| 222 |
+
"grad_norm": 11.130131721496582,
|
| 223 |
+
"learning_rate": 4.8859126984126984e-05,
|
| 224 |
+
"loss": 0.232,
|
| 225 |
+
"step": 2700
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.004765625,
|
| 229 |
+
"grad_norm": 3.47011661529541,
|
| 230 |
+
"learning_rate": 4.8611111111111115e-05,
|
| 231 |
+
"loss": 0.1247,
|
| 232 |
+
"step": 2800
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.0092299107142857,
|
| 236 |
+
"grad_norm": 18.701496124267578,
|
| 237 |
+
"learning_rate": 4.836309523809524e-05,
|
| 238 |
+
"loss": 0.1293,
|
| 239 |
+
"step": 2900
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.0136941964285713,
|
| 243 |
+
"grad_norm": 0.7256734371185303,
|
| 244 |
+
"learning_rate": 4.811507936507937e-05,
|
| 245 |
+
"loss": 0.1291,
|
| 246 |
+
"step": 3000
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.0181584821428573,
|
| 250 |
+
"grad_norm": 24.983957290649414,
|
| 251 |
+
"learning_rate": 4.7867063492063496e-05,
|
| 252 |
+
"loss": 0.195,
|
| 253 |
+
"step": 3100
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.022622767857143,
|
| 257 |
+
"grad_norm": 0.1959875524044037,
|
| 258 |
+
"learning_rate": 4.761904761904762e-05,
|
| 259 |
+
"loss": 0.0969,
|
| 260 |
+
"step": 3200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.0270870535714285,
|
| 264 |
+
"grad_norm": 1.1051886081695557,
|
| 265 |
+
"learning_rate": 4.7371031746031745e-05,
|
| 266 |
+
"loss": 0.1691,
|
| 267 |
+
"step": 3300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 3.031551339285714,
|
| 271 |
+
"grad_norm": 0.48205551505088806,
|
| 272 |
+
"learning_rate": 4.7123015873015876e-05,
|
| 273 |
+
"loss": 0.1297,
|
| 274 |
+
"step": 3400
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 3.036015625,
|
| 278 |
+
"grad_norm": 0.8840370774269104,
|
| 279 |
+
"learning_rate": 4.6875e-05,
|
| 280 |
+
"loss": 0.1229,
|
| 281 |
+
"step": 3500
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 3.040078125,
|
| 285 |
+
"eval_accuracy": 0.9647058823529412,
|
| 286 |
+
"eval_loss": 0.14670781791210175,
|
| 287 |
+
"eval_runtime": 285.475,
|
| 288 |
+
"eval_samples_per_second": 2.977,
|
| 289 |
+
"eval_steps_per_second": 1.489,
|
| 290 |
+
"step": 3591
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 4.000401785714286,
|
| 294 |
+
"grad_norm": 0.21204273402690887,
|
| 295 |
+
"learning_rate": 4.662698412698413e-05,
|
| 296 |
+
"loss": 0.1337,
|
| 297 |
+
"step": 3600
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 4.004866071428571,
|
| 301 |
+
"grad_norm": 2.2111618518829346,
|
| 302 |
+
"learning_rate": 4.637896825396826e-05,
|
| 303 |
+
"loss": 0.0821,
|
| 304 |
+
"step": 3700
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 4.009330357142857,
|
| 308 |
+
"grad_norm": 2.208402395248413,
|
| 309 |
+
"learning_rate": 4.613095238095239e-05,
|
| 310 |
+
"loss": 0.098,
|
| 311 |
+
"step": 3800
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 4.0137946428571425,
|
| 315 |
+
"grad_norm": 3.035139560699463,
|
| 316 |
+
"learning_rate": 4.5882936507936506e-05,
|
| 317 |
+
"loss": 0.0828,
|
| 318 |
+
"step": 3900
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 4.018258928571429,
|
| 322 |
+
"grad_norm": 0.06664509326219559,
|
| 323 |
+
"learning_rate": 4.563492063492064e-05,
|
| 324 |
+
"loss": 0.0705,
|
| 325 |
+
"step": 4000
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 4.0227232142857146,
|
| 329 |
+
"grad_norm": 0.049911659210920334,
|
| 330 |
+
"learning_rate": 4.538690476190476e-05,
|
| 331 |
+
"loss": 0.0506,
|
| 332 |
+
"step": 4100
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 4.0271875,
|
| 336 |
+
"grad_norm": 6.9254374504089355,
|
| 337 |
+
"learning_rate": 4.5138888888888894e-05,
|
| 338 |
+
"loss": 0.0895,
|
| 339 |
+
"step": 4200
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 4.031651785714286,
|
| 343 |
+
"grad_norm": 0.6636308431625366,
|
| 344 |
+
"learning_rate": 4.489087301587302e-05,
|
| 345 |
+
"loss": 0.0762,
|
| 346 |
+
"step": 4300
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 4.036116071428571,
|
| 350 |
+
"grad_norm": 0.07036083936691284,
|
| 351 |
+
"learning_rate": 4.464285714285715e-05,
|
| 352 |
+
"loss": 0.1045,
|
| 353 |
+
"step": 4400
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 4.040044642857143,
|
| 357 |
+
"eval_accuracy": 0.9635294117647059,
|
| 358 |
+
"eval_loss": 0.1458999365568161,
|
| 359 |
+
"eval_runtime": 292.1403,
|
| 360 |
+
"eval_samples_per_second": 2.91,
|
| 361 |
+
"eval_steps_per_second": 1.455,
|
| 362 |
+
"step": 4488
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 5.000502232142857,
|
| 366 |
+
"grad_norm": 25.948030471801758,
|
| 367 |
+
"learning_rate": 4.439484126984127e-05,
|
| 368 |
+
"loss": 0.1201,
|
| 369 |
+
"step": 4500
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 5.0049665178571425,
|
| 373 |
+
"grad_norm": 4.851236343383789,
|
| 374 |
+
"learning_rate": 4.41468253968254e-05,
|
| 375 |
+
"loss": 0.0751,
|
| 376 |
+
"step": 4600
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 5.009430803571429,
|
| 380 |
+
"grad_norm": 2.069117307662964,
|
| 381 |
+
"learning_rate": 4.3898809523809523e-05,
|
| 382 |
+
"loss": 0.06,
|
| 383 |
+
"step": 4700
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 5.013895089285715,
|
| 387 |
+
"grad_norm": 0.02893979474902153,
|
| 388 |
+
"learning_rate": 4.3650793650793655e-05,
|
| 389 |
+
"loss": 0.0583,
|
| 390 |
+
"step": 4800
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 5.018359375,
|
| 394 |
+
"grad_norm": 38.84079360961914,
|
| 395 |
+
"learning_rate": 4.340277777777778e-05,
|
| 396 |
+
"loss": 0.0854,
|
| 397 |
+
"step": 4900
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 5.022823660714286,
|
| 401 |
+
"grad_norm": 0.01713498868048191,
|
| 402 |
+
"learning_rate": 4.315476190476191e-05,
|
| 403 |
+
"loss": 0.1064,
|
| 404 |
+
"step": 5000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 5.027287946428571,
|
| 408 |
+
"grad_norm": 2.2113935947418213,
|
| 409 |
+
"learning_rate": 4.290674603174603e-05,
|
| 410 |
+
"loss": 0.0534,
|
| 411 |
+
"step": 5100
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 5.031752232142857,
|
| 415 |
+
"grad_norm": 0.030846355482935905,
|
| 416 |
+
"learning_rate": 4.265873015873016e-05,
|
| 417 |
+
"loss": 0.0812,
|
| 418 |
+
"step": 5200
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 5.0362165178571425,
|
| 422 |
+
"grad_norm": 63.66303253173828,
|
| 423 |
+
"learning_rate": 4.2410714285714285e-05,
|
| 424 |
+
"loss": 0.0631,
|
| 425 |
+
"step": 5300
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 5.040055803571429,
|
| 429 |
+
"eval_accuracy": 0.971764705882353,
|
| 430 |
+
"eval_loss": 0.13126207888126373,
|
| 431 |
+
"eval_runtime": 282.9661,
|
| 432 |
+
"eval_samples_per_second": 3.004,
|
| 433 |
+
"eval_steps_per_second": 1.502,
|
| 434 |
+
"step": 5386
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 6.000602678571428,
|
| 438 |
+
"grad_norm": 0.01721133291721344,
|
| 439 |
+
"learning_rate": 4.2162698412698416e-05,
|
| 440 |
+
"loss": 0.1066,
|
| 441 |
+
"step": 5400
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 6.005066964285715,
|
| 445 |
+
"grad_norm": 0.06797400861978531,
|
| 446 |
+
"learning_rate": 4.191468253968254e-05,
|
| 447 |
+
"loss": 0.0751,
|
| 448 |
+
"step": 5500
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 6.00953125,
|
| 452 |
+
"grad_norm": 0.22653132677078247,
|
| 453 |
+
"learning_rate": 4.166666666666667e-05,
|
| 454 |
+
"loss": 0.0417,
|
| 455 |
+
"step": 5600
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 6.013995535714286,
|
| 459 |
+
"grad_norm": 0.07131924480199814,
|
| 460 |
+
"learning_rate": 4.14186507936508e-05,
|
| 461 |
+
"loss": 0.0158,
|
| 462 |
+
"step": 5700
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 6.018459821428571,
|
| 466 |
+
"grad_norm": 40.63113784790039,
|
| 467 |
+
"learning_rate": 4.117063492063492e-05,
|
| 468 |
+
"loss": 0.0522,
|
| 469 |
+
"step": 5800
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 6.022924107142857,
|
| 473 |
+
"grad_norm": 0.09443258494138718,
|
| 474 |
+
"learning_rate": 4.0922619047619046e-05,
|
| 475 |
+
"loss": 0.072,
|
| 476 |
+
"step": 5900
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 6.027388392857143,
|
| 480 |
+
"grad_norm": 0.5265907049179077,
|
| 481 |
+
"learning_rate": 4.067460317460318e-05,
|
| 482 |
+
"loss": 0.0318,
|
| 483 |
+
"step": 6000
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 6.031852678571428,
|
| 487 |
+
"grad_norm": 0.03210202232003212,
|
| 488 |
+
"learning_rate": 4.04265873015873e-05,
|
| 489 |
+
"loss": 0.0877,
|
| 490 |
+
"step": 6100
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 6.036316964285715,
|
| 494 |
+
"grad_norm": 0.34825244545936584,
|
| 495 |
+
"learning_rate": 4.017857142857143e-05,
|
| 496 |
+
"loss": 0.0736,
|
| 497 |
+
"step": 6200
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 6.040066964285714,
|
| 501 |
+
"eval_accuracy": 0.9635294117647059,
|
| 502 |
+
"eval_loss": 0.18067213892936707,
|
| 503 |
+
"eval_runtime": 285.3373,
|
| 504 |
+
"eval_samples_per_second": 2.979,
|
| 505 |
+
"eval_steps_per_second": 1.489,
|
| 506 |
+
"step": 6284
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 7.000703125,
|
| 510 |
+
"grad_norm": 0.006914912257343531,
|
| 511 |
+
"learning_rate": 3.993055555555556e-05,
|
| 512 |
+
"loss": 0.0283,
|
| 513 |
+
"step": 6300
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 7.005167410714286,
|
| 517 |
+
"grad_norm": 0.0338265560567379,
|
| 518 |
+
"learning_rate": 3.968253968253968e-05,
|
| 519 |
+
"loss": 0.0499,
|
| 520 |
+
"step": 6400
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 7.009631696428571,
|
| 524 |
+
"grad_norm": 10.877938270568848,
|
| 525 |
+
"learning_rate": 3.943452380952381e-05,
|
| 526 |
+
"loss": 0.0082,
|
| 527 |
+
"step": 6500
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 7.014095982142857,
|
| 531 |
+
"grad_norm": 0.10941223055124283,
|
| 532 |
+
"learning_rate": 3.918650793650794e-05,
|
| 533 |
+
"loss": 0.0657,
|
| 534 |
+
"step": 6600
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 7.018560267857143,
|
| 538 |
+
"grad_norm": 12.054357528686523,
|
| 539 |
+
"learning_rate": 3.893849206349206e-05,
|
| 540 |
+
"loss": 0.0609,
|
| 541 |
+
"step": 6700
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 7.023024553571428,
|
| 545 |
+
"grad_norm": 0.006210957653820515,
|
| 546 |
+
"learning_rate": 3.8690476190476195e-05,
|
| 547 |
+
"loss": 0.0486,
|
| 548 |
+
"step": 6800
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 7.027488839285715,
|
| 552 |
+
"grad_norm": 0.013958507217466831,
|
| 553 |
+
"learning_rate": 3.844246031746032e-05,
|
| 554 |
+
"loss": 0.0747,
|
| 555 |
+
"step": 6900
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 7.031953125,
|
| 559 |
+
"grad_norm": 14.515870094299316,
|
| 560 |
+
"learning_rate": 3.8194444444444444e-05,
|
| 561 |
+
"loss": 0.0343,
|
| 562 |
+
"step": 7000
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 7.036417410714286,
|
| 566 |
+
"grad_norm": 0.007723964750766754,
|
| 567 |
+
"learning_rate": 3.794642857142857e-05,
|
| 568 |
+
"loss": 0.0673,
|
| 569 |
+
"step": 7100
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 7.040078125,
|
| 573 |
+
"eval_accuracy": 0.9694117647058823,
|
| 574 |
+
"eval_loss": 0.14643678069114685,
|
| 575 |
+
"eval_runtime": 288.72,
|
| 576 |
+
"eval_samples_per_second": 2.944,
|
| 577 |
+
"eval_steps_per_second": 1.472,
|
| 578 |
+
"step": 7182
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 8.000803571428571,
|
| 582 |
+
"grad_norm": 45.418617248535156,
|
| 583 |
+
"learning_rate": 3.76984126984127e-05,
|
| 584 |
+
"loss": 0.0476,
|
| 585 |
+
"step": 7200
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 8.005267857142858,
|
| 589 |
+
"grad_norm": 0.008381331339478493,
|
| 590 |
+
"learning_rate": 3.7450396825396824e-05,
|
| 591 |
+
"loss": 0.0421,
|
| 592 |
+
"step": 7300
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 8.009732142857143,
|
| 596 |
+
"grad_norm": 0.7666055560112,
|
| 597 |
+
"learning_rate": 3.7202380952380956e-05,
|
| 598 |
+
"loss": 0.0832,
|
| 599 |
+
"step": 7400
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 8.014196428571429,
|
| 603 |
+
"grad_norm": 0.09307380765676498,
|
| 604 |
+
"learning_rate": 3.695436507936508e-05,
|
| 605 |
+
"loss": 0.0875,
|
| 606 |
+
"step": 7500
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 8.018660714285714,
|
| 610 |
+
"grad_norm": 0.012713397853076458,
|
| 611 |
+
"learning_rate": 3.6706349206349205e-05,
|
| 612 |
+
"loss": 0.0441,
|
| 613 |
+
"step": 7600
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 8.023125,
|
| 617 |
+
"grad_norm": 0.021006299182772636,
|
| 618 |
+
"learning_rate": 3.6458333333333336e-05,
|
| 619 |
+
"loss": 0.054,
|
| 620 |
+
"step": 7700
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 8.027589285714285,
|
| 624 |
+
"grad_norm": 0.1419028341770172,
|
| 625 |
+
"learning_rate": 3.621031746031746e-05,
|
| 626 |
+
"loss": 0.0608,
|
| 627 |
+
"step": 7800
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 8.032053571428571,
|
| 631 |
+
"grad_norm": 0.025018220767378807,
|
| 632 |
+
"learning_rate": 3.5962301587301586e-05,
|
| 633 |
+
"loss": 0.0479,
|
| 634 |
+
"step": 7900
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 8.036517857142858,
|
| 638 |
+
"grad_norm": 0.5912023186683655,
|
| 639 |
+
"learning_rate": 3.571428571428572e-05,
|
| 640 |
+
"loss": 0.0239,
|
| 641 |
+
"step": 8000
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 8.040044642857143,
|
| 645 |
+
"eval_accuracy": 0.9576470588235294,
|
| 646 |
+
"eval_loss": 0.193200945854187,
|
| 647 |
+
"eval_runtime": 279.9813,
|
| 648 |
+
"eval_samples_per_second": 3.036,
|
| 649 |
+
"eval_steps_per_second": 1.518,
|
| 650 |
+
"step": 8079
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 9.000904017857144,
|
| 654 |
+
"grad_norm": 0.0350213348865509,
|
| 655 |
+
"learning_rate": 3.546626984126984e-05,
|
| 656 |
+
"loss": 0.067,
|
| 657 |
+
"step": 8100
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 9.005368303571428,
|
| 661 |
+
"grad_norm": 2.537632465362549,
|
| 662 |
+
"learning_rate": 3.521825396825397e-05,
|
| 663 |
+
"loss": 0.0245,
|
| 664 |
+
"step": 8200
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 9.009832589285715,
|
| 668 |
+
"grad_norm": 2.564781665802002,
|
| 669 |
+
"learning_rate": 3.49702380952381e-05,
|
| 670 |
+
"loss": 0.0262,
|
| 671 |
+
"step": 8300
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 9.014296875,
|
| 675 |
+
"grad_norm": 0.00803827028721571,
|
| 676 |
+
"learning_rate": 3.472222222222222e-05,
|
| 677 |
+
"loss": 0.0559,
|
| 678 |
+
"step": 8400
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 9.018761160714286,
|
| 682 |
+
"grad_norm": 0.005816516932100058,
|
| 683 |
+
"learning_rate": 3.4474206349206354e-05,
|
| 684 |
+
"loss": 0.0519,
|
| 685 |
+
"step": 8500
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 9.02322544642857,
|
| 689 |
+
"grad_norm": 0.021420830860733986,
|
| 690 |
+
"learning_rate": 3.422619047619048e-05,
|
| 691 |
+
"loss": 0.032,
|
| 692 |
+
"step": 8600
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 9.027689732142857,
|
| 696 |
+
"grad_norm": 0.028336547315120697,
|
| 697 |
+
"learning_rate": 3.397817460317461e-05,
|
| 698 |
+
"loss": 0.0227,
|
| 699 |
+
"step": 8700
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 9.032154017857144,
|
| 703 |
+
"grad_norm": 0.02300655096769333,
|
| 704 |
+
"learning_rate": 3.3730158730158734e-05,
|
| 705 |
+
"loss": 0.0392,
|
| 706 |
+
"step": 8800
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 9.036618303571428,
|
| 710 |
+
"grad_norm": 0.05427232384681702,
|
| 711 |
+
"learning_rate": 3.348214285714286e-05,
|
| 712 |
+
"loss": 0.0868,
|
| 713 |
+
"step": 8900
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 9.040055803571429,
|
| 717 |
+
"eval_accuracy": 0.9882352941176471,
|
| 718 |
+
"eval_loss": 0.05633905157446861,
|
| 719 |
+
"eval_runtime": 285.433,
|
| 720 |
+
"eval_samples_per_second": 2.978,
|
| 721 |
+
"eval_steps_per_second": 1.489,
|
| 722 |
+
"step": 8977
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 10.001004464285714,
|
| 726 |
+
"grad_norm": 0.0491323284804821,
|
| 727 |
+
"learning_rate": 3.3234126984126983e-05,
|
| 728 |
+
"loss": 0.0618,
|
| 729 |
+
"step": 9000
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 10.00546875,
|
| 733 |
+
"grad_norm": 1.0003972053527832,
|
| 734 |
+
"learning_rate": 3.2986111111111115e-05,
|
| 735 |
+
"loss": 0.0202,
|
| 736 |
+
"step": 9100
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 10.009933035714285,
|
| 740 |
+
"grad_norm": 0.00252954987809062,
|
| 741 |
+
"learning_rate": 3.273809523809524e-05,
|
| 742 |
+
"loss": 0.0531,
|
| 743 |
+
"step": 9200
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 10.014397321428572,
|
| 747 |
+
"grad_norm": 9.270633697509766,
|
| 748 |
+
"learning_rate": 3.249007936507937e-05,
|
| 749 |
+
"loss": 0.035,
|
| 750 |
+
"step": 9300
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 10.018861607142858,
|
| 754 |
+
"grad_norm": 0.014138671569526196,
|
| 755 |
+
"learning_rate": 3.2242063492063495e-05,
|
| 756 |
+
"loss": 0.0392,
|
| 757 |
+
"step": 9400
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 10.023325892857143,
|
| 761 |
+
"grad_norm": 0.01277222577482462,
|
| 762 |
+
"learning_rate": 3.199404761904762e-05,
|
| 763 |
+
"loss": 0.059,
|
| 764 |
+
"step": 9500
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 10.02779017857143,
|
| 768 |
+
"grad_norm": 0.0034905134234577417,
|
| 769 |
+
"learning_rate": 3.1746031746031745e-05,
|
| 770 |
+
"loss": 0.0664,
|
| 771 |
+
"step": 9600
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 10.032254464285714,
|
| 775 |
+
"grad_norm": 0.0024051007349044085,
|
| 776 |
+
"learning_rate": 3.1498015873015876e-05,
|
| 777 |
+
"loss": 0.0286,
|
| 778 |
+
"step": 9700
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 10.03671875,
|
| 782 |
+
"grad_norm": 0.002095526549965143,
|
| 783 |
+
"learning_rate": 3.125e-05,
|
| 784 |
+
"loss": 0.0016,
|
| 785 |
+
"step": 9800
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 10.040066964285714,
|
| 789 |
+
"eval_accuracy": 0.9776470588235294,
|
| 790 |
+
"eval_loss": 0.08437661826610565,
|
| 791 |
+
"eval_runtime": 280.7764,
|
| 792 |
+
"eval_samples_per_second": 3.027,
|
| 793 |
+
"eval_steps_per_second": 1.514,
|
| 794 |
+
"step": 9875
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 11.001104910714286,
|
| 798 |
+
"grad_norm": 0.0019545548129826784,
|
| 799 |
+
"learning_rate": 3.100198412698413e-05,
|
| 800 |
+
"loss": 0.0109,
|
| 801 |
+
"step": 9900
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 11.00556919642857,
|
| 805 |
+
"grad_norm": 0.005866718012839556,
|
| 806 |
+
"learning_rate": 3.075396825396826e-05,
|
| 807 |
+
"loss": 0.0479,
|
| 808 |
+
"step": 10000
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 11.010033482142857,
|
| 812 |
+
"grad_norm": 0.012244959361851215,
|
| 813 |
+
"learning_rate": 3.0505952380952385e-05,
|
| 814 |
+
"loss": 0.0116,
|
| 815 |
+
"step": 10100
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 11.014497767857144,
|
| 819 |
+
"grad_norm": 0.004522031173110008,
|
| 820 |
+
"learning_rate": 3.0257936507936506e-05,
|
| 821 |
+
"loss": 0.025,
|
| 822 |
+
"step": 10200
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 11.018962053571428,
|
| 826 |
+
"grad_norm": 0.010159791447222233,
|
| 827 |
+
"learning_rate": 3.0009920634920634e-05,
|
| 828 |
+
"loss": 0.0036,
|
| 829 |
+
"step": 10300
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 11.023426339285715,
|
| 833 |
+
"grad_norm": 0.40824609994888306,
|
| 834 |
+
"learning_rate": 2.9761904761904762e-05,
|
| 835 |
+
"loss": 0.0933,
|
| 836 |
+
"step": 10400
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 11.027890625,
|
| 840 |
+
"grad_norm": 0.11058317124843597,
|
| 841 |
+
"learning_rate": 2.951388888888889e-05,
|
| 842 |
+
"loss": 0.0161,
|
| 843 |
+
"step": 10500
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 11.032354910714286,
|
| 847 |
+
"grad_norm": 1.2187433242797852,
|
| 848 |
+
"learning_rate": 2.9265873015873018e-05,
|
| 849 |
+
"loss": 0.0329,
|
| 850 |
+
"step": 10600
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 11.03681919642857,
|
| 854 |
+
"grad_norm": 0.020026879385113716,
|
| 855 |
+
"learning_rate": 2.9017857142857146e-05,
|
| 856 |
+
"loss": 0.0318,
|
| 857 |
+
"step": 10700
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 11.040078125,
|
| 861 |
+
"eval_accuracy": 0.9752941176470589,
|
| 862 |
+
"eval_loss": 0.11233757436275482,
|
| 863 |
+
"eval_runtime": 279.6949,
|
| 864 |
+
"eval_samples_per_second": 3.039,
|
| 865 |
+
"eval_steps_per_second": 1.52,
|
| 866 |
+
"step": 10773
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 12.001205357142856,
|
| 870 |
+
"grad_norm": 0.004233605694025755,
|
| 871 |
+
"learning_rate": 2.876984126984127e-05,
|
| 872 |
+
"loss": 0.0145,
|
| 873 |
+
"step": 10800
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 12.005669642857143,
|
| 877 |
+
"grad_norm": 0.0020020680967718363,
|
| 878 |
+
"learning_rate": 2.8521825396825395e-05,
|
| 879 |
+
"loss": 0.0022,
|
| 880 |
+
"step": 10900
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 12.01013392857143,
|
| 884 |
+
"grad_norm": 0.0010592287871986628,
|
| 885 |
+
"learning_rate": 2.8273809523809523e-05,
|
| 886 |
+
"loss": 0.0029,
|
| 887 |
+
"step": 11000
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 12.014598214285714,
|
| 891 |
+
"grad_norm": 0.01872986927628517,
|
| 892 |
+
"learning_rate": 2.802579365079365e-05,
|
| 893 |
+
"loss": 0.0352,
|
| 894 |
+
"step": 11100
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 12.0190625,
|
| 898 |
+
"grad_norm": 0.05156349390745163,
|
| 899 |
+
"learning_rate": 2.777777777777778e-05,
|
| 900 |
+
"loss": 0.0047,
|
| 901 |
+
"step": 11200
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 12.023526785714285,
|
| 905 |
+
"grad_norm": 0.00894691701978445,
|
| 906 |
+
"learning_rate": 2.7529761904761907e-05,
|
| 907 |
+
"loss": 0.0303,
|
| 908 |
+
"step": 11300
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 12.027991071428572,
|
| 912 |
+
"grad_norm": 0.004200028255581856,
|
| 913 |
+
"learning_rate": 2.7281746031746032e-05,
|
| 914 |
+
"loss": 0.0782,
|
| 915 |
+
"step": 11400
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 12.032455357142856,
|
| 919 |
+
"grad_norm": 0.008372528478503227,
|
| 920 |
+
"learning_rate": 2.703373015873016e-05,
|
| 921 |
+
"loss": 0.0154,
|
| 922 |
+
"step": 11500
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 12.036919642857143,
|
| 926 |
+
"grad_norm": 0.010021534748375416,
|
| 927 |
+
"learning_rate": 2.6785714285714288e-05,
|
| 928 |
+
"loss": 0.0144,
|
| 929 |
+
"step": 11600
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 12.040044642857143,
|
| 933 |
+
"eval_accuracy": 0.9894117647058823,
|
| 934 |
+
"eval_loss": 0.04987098649144173,
|
| 935 |
+
"eval_runtime": 331.781,
|
| 936 |
+
"eval_samples_per_second": 2.562,
|
| 937 |
+
"eval_steps_per_second": 1.281,
|
| 938 |
+
"step": 11670
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 13.001305803571428,
|
| 942 |
+
"grad_norm": 0.3831511437892914,
|
| 943 |
+
"learning_rate": 2.6537698412698416e-05,
|
| 944 |
+
"loss": 0.0175,
|
| 945 |
+
"step": 11700
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 13.005770089285715,
|
| 949 |
+
"grad_norm": 0.0010712681105360389,
|
| 950 |
+
"learning_rate": 2.628968253968254e-05,
|
| 951 |
+
"loss": 0.0281,
|
| 952 |
+
"step": 11800
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 13.010234375,
|
| 956 |
+
"grad_norm": 0.004961916245520115,
|
| 957 |
+
"learning_rate": 2.604166666666667e-05,
|
| 958 |
+
"loss": 0.0162,
|
| 959 |
+
"step": 11900
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 13.014698660714286,
|
| 963 |
+
"grad_norm": 0.3577312231063843,
|
| 964 |
+
"learning_rate": 2.5793650793650796e-05,
|
| 965 |
+
"loss": 0.0133,
|
| 966 |
+
"step": 12000
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 13.01916294642857,
|
| 970 |
+
"grad_norm": 0.0016846248181536794,
|
| 971 |
+
"learning_rate": 2.554563492063492e-05,
|
| 972 |
+
"loss": 0.0456,
|
| 973 |
+
"step": 12100
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 13.023627232142857,
|
| 977 |
+
"grad_norm": 0.005252454895526171,
|
| 978 |
+
"learning_rate": 2.529761904761905e-05,
|
| 979 |
+
"loss": 0.0043,
|
| 980 |
+
"step": 12200
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 13.028091517857144,
|
| 984 |
+
"grad_norm": 65.35294342041016,
|
| 985 |
+
"learning_rate": 2.5049603174603177e-05,
|
| 986 |
+
"loss": 0.0248,
|
| 987 |
+
"step": 12300
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 13.032555803571428,
|
| 991 |
+
"grad_norm": 0.0010413563577458262,
|
| 992 |
+
"learning_rate": 2.4801587301587305e-05,
|
| 993 |
+
"loss": 0.033,
|
| 994 |
+
"step": 12400
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 13.037020089285715,
|
| 998 |
+
"grad_norm": 28.086708068847656,
|
| 999 |
+
"learning_rate": 2.455357142857143e-05,
|
| 1000 |
+
"loss": 0.0028,
|
| 1001 |
+
"step": 12500
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 13.040055803571429,
|
| 1005 |
+
"eval_accuracy": 0.9870588235294118,
|
| 1006 |
+
"eval_loss": 0.08093971014022827,
|
| 1007 |
+
"eval_runtime": 287.2538,
|
| 1008 |
+
"eval_samples_per_second": 2.959,
|
| 1009 |
+
"eval_steps_per_second": 1.48,
|
| 1010 |
+
"step": 12568
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 14.00140625,
|
| 1014 |
+
"grad_norm": 0.011327456682920456,
|
| 1015 |
+
"learning_rate": 2.4305555555555558e-05,
|
| 1016 |
+
"loss": 0.0203,
|
| 1017 |
+
"step": 12600
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 14.005870535714285,
|
| 1021 |
+
"grad_norm": 0.006360394414514303,
|
| 1022 |
+
"learning_rate": 2.4057539682539686e-05,
|
| 1023 |
+
"loss": 0.0009,
|
| 1024 |
+
"step": 12700
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 14.010334821428572,
|
| 1028 |
+
"grad_norm": 1.3321506977081299,
|
| 1029 |
+
"learning_rate": 2.380952380952381e-05,
|
| 1030 |
+
"loss": 0.0186,
|
| 1031 |
+
"step": 12800
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 14.014799107142856,
|
| 1035 |
+
"grad_norm": 0.0009386364254169166,
|
| 1036 |
+
"learning_rate": 2.3561507936507938e-05,
|
| 1037 |
+
"loss": 0.0048,
|
| 1038 |
+
"step": 12900
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 14.019263392857143,
|
| 1042 |
+
"grad_norm": 0.0016534485621377826,
|
| 1043 |
+
"learning_rate": 2.3313492063492066e-05,
|
| 1044 |
+
"loss": 0.037,
|
| 1045 |
+
"step": 13000
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 14.02372767857143,
|
| 1049 |
+
"grad_norm": 0.001421699533239007,
|
| 1050 |
+
"learning_rate": 2.3065476190476194e-05,
|
| 1051 |
+
"loss": 0.0111,
|
| 1052 |
+
"step": 13100
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 14.028191964285714,
|
| 1056 |
+
"grad_norm": 0.0014466221909970045,
|
| 1057 |
+
"learning_rate": 2.281746031746032e-05,
|
| 1058 |
+
"loss": 0.0169,
|
| 1059 |
+
"step": 13200
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 14.03265625,
|
| 1063 |
+
"grad_norm": 0.0036468463949859142,
|
| 1064 |
+
"learning_rate": 2.2569444444444447e-05,
|
| 1065 |
+
"loss": 0.019,
|
| 1066 |
+
"step": 13300
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 14.037120535714285,
|
| 1070 |
+
"grad_norm": 0.0012320175301283598,
|
| 1071 |
+
"learning_rate": 2.2321428571428575e-05,
|
| 1072 |
+
"loss": 0.0074,
|
| 1073 |
+
"step": 13400
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 14.040066964285714,
|
| 1077 |
+
"eval_accuracy": 0.9929411764705882,
|
| 1078 |
+
"eval_loss": 0.045501772314310074,
|
| 1079 |
+
"eval_runtime": 285.3107,
|
| 1080 |
+
"eval_samples_per_second": 2.979,
|
| 1081 |
+
"eval_steps_per_second": 1.49,
|
| 1082 |
+
"step": 13466
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 15.00150669642857,
|
| 1086 |
+
"grad_norm": 0.0006422046571969986,
|
| 1087 |
+
"learning_rate": 2.20734126984127e-05,
|
| 1088 |
+
"loss": 0.0202,
|
| 1089 |
+
"step": 13500
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 15.005970982142857,
|
| 1093 |
+
"grad_norm": 0.0008420124650001526,
|
| 1094 |
+
"learning_rate": 2.1825396825396827e-05,
|
| 1095 |
+
"loss": 0.0116,
|
| 1096 |
+
"step": 13600
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 15.010435267857142,
|
| 1100 |
+
"grad_norm": 0.018089979887008667,
|
| 1101 |
+
"learning_rate": 2.1577380952380955e-05,
|
| 1102 |
+
"loss": 0.0099,
|
| 1103 |
+
"step": 13700
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 15.014899553571428,
|
| 1107 |
+
"grad_norm": 0.0031337908003479242,
|
| 1108 |
+
"learning_rate": 2.132936507936508e-05,
|
| 1109 |
+
"loss": 0.0566,
|
| 1110 |
+
"step": 13800
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 15.019363839285715,
|
| 1114 |
+
"grad_norm": 0.0016157528152689338,
|
| 1115 |
+
"learning_rate": 2.1081349206349208e-05,
|
| 1116 |
+
"loss": 0.0212,
|
| 1117 |
+
"step": 13900
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 15.023828125,
|
| 1121 |
+
"grad_norm": 0.01456926204264164,
|
| 1122 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 1123 |
+
"loss": 0.0003,
|
| 1124 |
+
"step": 14000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 15.028292410714286,
|
| 1128 |
+
"grad_norm": 0.001924099400639534,
|
| 1129 |
+
"learning_rate": 2.058531746031746e-05,
|
| 1130 |
+
"loss": 0.0149,
|
| 1131 |
+
"step": 14100
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 15.03275669642857,
|
| 1135 |
+
"grad_norm": 0.0008741599158383906,
|
| 1136 |
+
"learning_rate": 2.033730158730159e-05,
|
| 1137 |
+
"loss": 0.0168,
|
| 1138 |
+
"step": 14200
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 15.037220982142857,
|
| 1142 |
+
"grad_norm": 0.06954433768987656,
|
| 1143 |
+
"learning_rate": 2.0089285714285717e-05,
|
| 1144 |
+
"loss": 0.0002,
|
| 1145 |
+
"step": 14300
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 15.040078125,
|
| 1149 |
+
"eval_accuracy": 0.9905882352941177,
|
| 1150 |
+
"eval_loss": 0.058066971600055695,
|
| 1151 |
+
"eval_runtime": 289.1743,
|
| 1152 |
+
"eval_samples_per_second": 2.939,
|
| 1153 |
+
"eval_steps_per_second": 1.47,
|
| 1154 |
+
"step": 14364
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 16.001607142857143,
|
| 1158 |
+
"grad_norm": 0.0014486366417258978,
|
| 1159 |
+
"learning_rate": 1.984126984126984e-05,
|
| 1160 |
+
"loss": 0.0063,
|
| 1161 |
+
"step": 14400
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 16.006071428571428,
|
| 1165 |
+
"grad_norm": 0.0007301854784600437,
|
| 1166 |
+
"learning_rate": 1.959325396825397e-05,
|
| 1167 |
+
"loss": 0.0186,
|
| 1168 |
+
"step": 14500
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 16.010535714285716,
|
| 1172 |
+
"grad_norm": 0.003457231679931283,
|
| 1173 |
+
"learning_rate": 1.9345238095238097e-05,
|
| 1174 |
+
"loss": 0.0236,
|
| 1175 |
+
"step": 14600
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 16.015,
|
| 1179 |
+
"grad_norm": 0.005807195790112019,
|
| 1180 |
+
"learning_rate": 1.9097222222222222e-05,
|
| 1181 |
+
"loss": 0.0183,
|
| 1182 |
+
"step": 14700
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 16.019464285714285,
|
| 1186 |
+
"grad_norm": 0.002843959955498576,
|
| 1187 |
+
"learning_rate": 1.884920634920635e-05,
|
| 1188 |
+
"loss": 0.0242,
|
| 1189 |
+
"step": 14800
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 16.02392857142857,
|
| 1193 |
+
"grad_norm": 0.37613585591316223,
|
| 1194 |
+
"learning_rate": 1.8601190476190478e-05,
|
| 1195 |
+
"loss": 0.0101,
|
| 1196 |
+
"step": 14900
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 16.028392857142858,
|
| 1200 |
+
"grad_norm": 0.0005575509858317673,
|
| 1201 |
+
"learning_rate": 1.8353174603174602e-05,
|
| 1202 |
+
"loss": 0.0109,
|
| 1203 |
+
"step": 15000
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 16.032857142857143,
|
| 1207 |
+
"grad_norm": 0.0006386680179275572,
|
| 1208 |
+
"learning_rate": 1.810515873015873e-05,
|
| 1209 |
+
"loss": 0.0013,
|
| 1210 |
+
"step": 15100
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 16.037321428571428,
|
| 1214 |
+
"grad_norm": 0.0010088573908433318,
|
| 1215 |
+
"learning_rate": 1.785714285714286e-05,
|
| 1216 |
+
"loss": 0.0077,
|
| 1217 |
+
"step": 15200
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 16.040044642857143,
|
| 1221 |
+
"eval_accuracy": 0.9894117647058823,
|
| 1222 |
+
"eval_loss": 0.05021252483129501,
|
| 1223 |
+
"eval_runtime": 284.0094,
|
| 1224 |
+
"eval_samples_per_second": 2.993,
|
| 1225 |
+
"eval_steps_per_second": 1.496,
|
| 1226 |
+
"step": 15261
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 17.001707589285715,
|
| 1230 |
+
"grad_norm": 0.0006336846854537725,
|
| 1231 |
+
"learning_rate": 1.7609126984126986e-05,
|
| 1232 |
+
"loss": 0.0212,
|
| 1233 |
+
"step": 15300
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 17.006171875,
|
| 1237 |
+
"grad_norm": 0.0005883209523744881,
|
| 1238 |
+
"learning_rate": 1.736111111111111e-05,
|
| 1239 |
+
"loss": 0.0095,
|
| 1240 |
+
"step": 15400
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 17.010636160714284,
|
| 1244 |
+
"grad_norm": 0.0021267228294163942,
|
| 1245 |
+
"learning_rate": 1.711309523809524e-05,
|
| 1246 |
+
"loss": 0.0113,
|
| 1247 |
+
"step": 15500
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 17.015100446428573,
|
| 1251 |
+
"grad_norm": 0.0009908992797136307,
|
| 1252 |
+
"learning_rate": 1.6865079365079367e-05,
|
| 1253 |
+
"loss": 0.0118,
|
| 1254 |
+
"step": 15600
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 17.019564732142857,
|
| 1258 |
+
"grad_norm": 0.000644190120510757,
|
| 1259 |
+
"learning_rate": 1.6617063492063492e-05,
|
| 1260 |
+
"loss": 0.0007,
|
| 1261 |
+
"step": 15700
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 17.024029017857142,
|
| 1265 |
+
"grad_norm": 0.0005113797378726304,
|
| 1266 |
+
"learning_rate": 1.636904761904762e-05,
|
| 1267 |
+
"loss": 0.01,
|
| 1268 |
+
"step": 15800
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 17.02849330357143,
|
| 1272 |
+
"grad_norm": 0.0008760132477618754,
|
| 1273 |
+
"learning_rate": 1.6121031746031748e-05,
|
| 1274 |
+
"loss": 0.0026,
|
| 1275 |
+
"step": 15900
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 17.032957589285715,
|
| 1279 |
+
"grad_norm": 0.00030510194483213127,
|
| 1280 |
+
"learning_rate": 1.5873015873015872e-05,
|
| 1281 |
+
"loss": 0.0015,
|
| 1282 |
+
"step": 16000
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 17.037421875,
|
| 1286 |
+
"grad_norm": 0.0004963899846188724,
|
| 1287 |
+
"learning_rate": 1.5625e-05,
|
| 1288 |
+
"loss": 0.0005,
|
| 1289 |
+
"step": 16100
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 17.040055803571427,
|
| 1293 |
+
"eval_accuracy": 0.9929411764705882,
|
| 1294 |
+
"eval_loss": 0.04069099575281143,
|
| 1295 |
+
"eval_runtime": 285.9614,
|
| 1296 |
+
"eval_samples_per_second": 2.972,
|
| 1297 |
+
"eval_steps_per_second": 1.486,
|
| 1298 |
+
"step": 16159
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 18.001808035714287,
|
| 1302 |
+
"grad_norm": 0.0015891814837232232,
|
| 1303 |
+
"learning_rate": 1.537698412698413e-05,
|
| 1304 |
+
"loss": 0.0376,
|
| 1305 |
+
"step": 16200
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 18.006272321428572,
|
| 1309 |
+
"grad_norm": 0.008500500582158566,
|
| 1310 |
+
"learning_rate": 1.5128968253968253e-05,
|
| 1311 |
+
"loss": 0.0203,
|
| 1312 |
+
"step": 16300
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 18.010736607142857,
|
| 1316 |
+
"grad_norm": 0.0030595629941672087,
|
| 1317 |
+
"learning_rate": 1.4880952380952381e-05,
|
| 1318 |
+
"loss": 0.0042,
|
| 1319 |
+
"step": 16400
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 18.01520089285714,
|
| 1323 |
+
"grad_norm": 1.0810060501098633,
|
| 1324 |
+
"learning_rate": 1.4632936507936509e-05,
|
| 1325 |
+
"loss": 0.017,
|
| 1326 |
+
"step": 16500
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 18.01966517857143,
|
| 1330 |
+
"grad_norm": 0.0005325720412656665,
|
| 1331 |
+
"learning_rate": 1.4384920634920635e-05,
|
| 1332 |
+
"loss": 0.0036,
|
| 1333 |
+
"step": 16600
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 18.024129464285714,
|
| 1337 |
+
"grad_norm": 0.0014920306857675314,
|
| 1338 |
+
"learning_rate": 1.4136904761904762e-05,
|
| 1339 |
+
"loss": 0.0236,
|
| 1340 |
+
"step": 16700
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 18.02859375,
|
| 1344 |
+
"grad_norm": 0.00048302882350981236,
|
| 1345 |
+
"learning_rate": 1.388888888888889e-05,
|
| 1346 |
+
"loss": 0.0127,
|
| 1347 |
+
"step": 16800
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 18.033058035714287,
|
| 1351 |
+
"grad_norm": 0.002715888200327754,
|
| 1352 |
+
"learning_rate": 1.3640873015873016e-05,
|
| 1353 |
+
"loss": 0.0146,
|
| 1354 |
+
"step": 16900
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 18.037522321428572,
|
| 1358 |
+
"grad_norm": 0.0004213691863697022,
|
| 1359 |
+
"learning_rate": 1.3392857142857144e-05,
|
| 1360 |
+
"loss": 0.0004,
|
| 1361 |
+
"step": 17000
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 18.040066964285714,
|
| 1365 |
+
"eval_accuracy": 0.9905882352941177,
|
| 1366 |
+
"eval_loss": 0.05496314540505409,
|
| 1367 |
+
"eval_runtime": 318.6642,
|
| 1368 |
+
"eval_samples_per_second": 2.667,
|
| 1369 |
+
"eval_steps_per_second": 1.334,
|
| 1370 |
+
"step": 17057
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 19.001908482142856,
|
| 1374 |
+
"grad_norm": 0.00044045469257980585,
|
| 1375 |
+
"learning_rate": 1.314484126984127e-05,
|
| 1376 |
+
"loss": 0.0143,
|
| 1377 |
+
"step": 17100
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 19.006372767857144,
|
| 1381 |
+
"grad_norm": 0.0004946400295011699,
|
| 1382 |
+
"learning_rate": 1.2896825396825398e-05,
|
| 1383 |
+
"loss": 0.0002,
|
| 1384 |
+
"step": 17200
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 19.01083705357143,
|
| 1388 |
+
"grad_norm": 0.014897634275257587,
|
| 1389 |
+
"learning_rate": 1.2648809523809524e-05,
|
| 1390 |
+
"loss": 0.0011,
|
| 1391 |
+
"step": 17300
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 19.015301339285713,
|
| 1395 |
+
"grad_norm": 0.015875551849603653,
|
| 1396 |
+
"learning_rate": 1.2400793650793652e-05,
|
| 1397 |
+
"loss": 0.0007,
|
| 1398 |
+
"step": 17400
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 19.019765625,
|
| 1402 |
+
"grad_norm": 0.0004391854163259268,
|
| 1403 |
+
"learning_rate": 1.2152777777777779e-05,
|
| 1404 |
+
"loss": 0.0068,
|
| 1405 |
+
"step": 17500
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 19.024229910714286,
|
| 1409 |
+
"grad_norm": 0.00046034177648834884,
|
| 1410 |
+
"learning_rate": 1.1904761904761905e-05,
|
| 1411 |
+
"loss": 0.0001,
|
| 1412 |
+
"step": 17600
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 19.02869419642857,
|
| 1416 |
+
"grad_norm": 0.0017288514645770192,
|
| 1417 |
+
"learning_rate": 1.1656746031746033e-05,
|
| 1418 |
+
"loss": 0.0001,
|
| 1419 |
+
"step": 17700
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 19.033158482142856,
|
| 1423 |
+
"grad_norm": 0.0026627290062606335,
|
| 1424 |
+
"learning_rate": 1.140873015873016e-05,
|
| 1425 |
+
"loss": 0.0001,
|
| 1426 |
+
"step": 17800
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 19.037622767857144,
|
| 1430 |
+
"grad_norm": 0.0004681396530941129,
|
| 1431 |
+
"learning_rate": 1.1160714285714287e-05,
|
| 1432 |
+
"loss": 0.0001,
|
| 1433 |
+
"step": 17900
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 19.040078125,
|
| 1437 |
+
"eval_accuracy": 0.9929411764705882,
|
| 1438 |
+
"eval_loss": 0.05834496021270752,
|
| 1439 |
+
"eval_runtime": 239.7594,
|
| 1440 |
+
"eval_samples_per_second": 3.545,
|
| 1441 |
+
"eval_steps_per_second": 1.773,
|
| 1442 |
+
"step": 17955
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 19.040078125,
|
| 1446 |
+
"step": 17955,
|
| 1447 |
+
"total_flos": 1.7905236367909847e+20,
|
| 1448 |
+
"train_loss": 0.295465733557037,
|
| 1449 |
+
"train_runtime": 68646.4816,
|
| 1450 |
+
"train_samples_per_second": 2.61,
|
| 1451 |
+
"train_steps_per_second": 0.326
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 19.040078125,
|
| 1455 |
+
"eval_accuracy": 0.9929411764705882,
|
| 1456 |
+
"eval_loss": 0.04550177976489067,
|
| 1457 |
+
"eval_runtime": 230.0122,
|
| 1458 |
+
"eval_samples_per_second": 3.695,
|
| 1459 |
+
"eval_steps_per_second": 1.848,
|
| 1460 |
+
"step": 17955
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 19.040078125,
|
| 1464 |
+
"eval_accuracy": 0.8973354231974922,
|
| 1465 |
+
"eval_loss": 0.5587517619132996,
|
| 1466 |
+
"eval_runtime": 352.6874,
|
| 1467 |
+
"eval_samples_per_second": 3.618,
|
| 1468 |
+
"eval_steps_per_second": 1.809,
|
| 1469 |
+
"step": 17955
|
| 1470 |
+
}
|
| 1471 |
+
],
|
| 1472 |
+
"logging_steps": 100,
|
| 1473 |
+
"max_steps": 22400,
|
| 1474 |
+
"num_input_tokens_seen": 0,
|
| 1475 |
+
"num_train_epochs": 9223372036854775807,
|
| 1476 |
+
"save_steps": 500,
|
| 1477 |
+
"stateful_callbacks": {
|
| 1478 |
+
"EarlyStoppingCallback": {
|
| 1479 |
+
"args": {
|
| 1480 |
+
"early_stopping_patience": 5,
|
| 1481 |
+
"early_stopping_threshold": 0.0
|
| 1482 |
+
},
|
| 1483 |
+
"attributes": {
|
| 1484 |
+
"early_stopping_patience_counter": 0
|
| 1485 |
+
}
|
| 1486 |
+
},
|
| 1487 |
+
"TrainerControl": {
|
| 1488 |
+
"args": {
|
| 1489 |
+
"should_epoch_stop": false,
|
| 1490 |
+
"should_evaluate": false,
|
| 1491 |
+
"should_log": false,
|
| 1492 |
+
"should_save": true,
|
| 1493 |
+
"should_training_stop": true
|
| 1494 |
+
},
|
| 1495 |
+
"attributes": {}
|
| 1496 |
+
}
|
| 1497 |
+
},
|
| 1498 |
+
"total_flos": 1.7905236367909847e+20,
|
| 1499 |
+
"train_batch_size": 2,
|
| 1500 |
+
"trial_name": null,
|
| 1501 |
+
"trial_params": null
|
| 1502 |
+
}
|