Commit
•
18b86e4
1
Parent(s):
5cb7f79
End of training
Browse files- README.md +3 -3
- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2316 -0
README.md
CHANGED
|
@@ -23,7 +23,7 @@ model-index:
|
|
| 23 |
metrics:
|
| 24 |
- name: Accuracy
|
| 25 |
type: accuracy
|
| 26 |
-
value: 0.
|
| 27 |
---
|
| 28 |
|
| 29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -33,8 +33,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 33 |
|
| 34 |
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
-
- Loss: 0.
|
| 37 |
-
- Accuracy: 0.
|
| 38 |
|
| 39 |
## Model description
|
| 40 |
|
|
|
|
| 23 |
metrics:
|
| 24 |
- name: Accuracy
|
| 25 |
type: accuracy
|
| 26 |
+
value: 0.8892018779342723
|
| 27 |
---
|
| 28 |
|
| 29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 33 |
|
| 34 |
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
+
- Loss: 0.3631
|
| 37 |
+
- Accuracy: 0.8892
|
| 38 |
|
| 39 |
## Model description
|
| 40 |
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.984,
|
| 3 |
+
"eval_accuracy": 0.8892018779342723,
|
| 4 |
+
"eval_loss": 0.3630879521369934,
|
| 5 |
+
"eval_runtime": 16.4219,
|
| 6 |
+
"eval_samples_per_second": 64.852,
|
| 7 |
+
"eval_steps_per_second": 2.07,
|
| 8 |
+
"total_flos": 9.926487761391452e+18,
|
| 9 |
+
"train_loss": 0.15716634974934351,
|
| 10 |
+
"train_runtime": 142469.0271,
|
| 11 |
+
"train_samples_per_second": 2.808,
|
| 12 |
+
"train_steps_per_second": 0.022
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.984,
|
| 3 |
+
"eval_accuracy": 0.8892018779342723,
|
| 4 |
+
"eval_loss": 0.3630879521369934,
|
| 5 |
+
"eval_runtime": 16.4219,
|
| 6 |
+
"eval_samples_per_second": 64.852,
|
| 7 |
+
"eval_steps_per_second": 2.07
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.984,
|
| 3 |
+
"total_flos": 9.926487761391452e+18,
|
| 4 |
+
"train_loss": 0.15716634974934351,
|
| 5 |
+
"train_runtime": 142469.0271,
|
| 6 |
+
"train_samples_per_second": 2.808,
|
| 7 |
+
"train_steps_per_second": 0.022
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2316 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8892018779342723,
|
| 3 |
+
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-leukemia-08-2024.v1.1\\checkpoint-2187",
|
| 4 |
+
"epoch": 9.984,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3120,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.032,
|
| 13 |
+
"grad_norm": 2.98797607421875,
|
| 14 |
+
"learning_rate": 1.6025641025641025e-06,
|
| 15 |
+
"loss": 0.712,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.064,
|
| 20 |
+
"grad_norm": 6.277791976928711,
|
| 21 |
+
"learning_rate": 3.205128205128205e-06,
|
| 22 |
+
"loss": 0.6981,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.096,
|
| 27 |
+
"grad_norm": 5.27135705947876,
|
| 28 |
+
"learning_rate": 4.807692307692308e-06,
|
| 29 |
+
"loss": 0.6783,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.128,
|
| 34 |
+
"grad_norm": 6.831840991973877,
|
| 35 |
+
"learning_rate": 6.41025641025641e-06,
|
| 36 |
+
"loss": 0.6498,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.16,
|
| 41 |
+
"grad_norm": 8.679888725280762,
|
| 42 |
+
"learning_rate": 8.012820512820515e-06,
|
| 43 |
+
"loss": 0.6079,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.192,
|
| 48 |
+
"grad_norm": 11.260509490966797,
|
| 49 |
+
"learning_rate": 9.615384615384616e-06,
|
| 50 |
+
"loss": 0.579,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.224,
|
| 55 |
+
"grad_norm": 17.14796257019043,
|
| 56 |
+
"learning_rate": 1.1217948717948719e-05,
|
| 57 |
+
"loss": 0.5673,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.256,
|
| 62 |
+
"grad_norm": 10.239635467529297,
|
| 63 |
+
"learning_rate": 1.282051282051282e-05,
|
| 64 |
+
"loss": 0.5263,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.288,
|
| 69 |
+
"grad_norm": 7.6730055809021,
|
| 70 |
+
"learning_rate": 1.4423076923076923e-05,
|
| 71 |
+
"loss": 0.5371,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.32,
|
| 76 |
+
"grad_norm": 7.861423492431641,
|
| 77 |
+
"learning_rate": 1.602564102564103e-05,
|
| 78 |
+
"loss": 0.5144,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.352,
|
| 83 |
+
"grad_norm": 9.051122665405273,
|
| 84 |
+
"learning_rate": 1.762820512820513e-05,
|
| 85 |
+
"loss": 0.5046,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.384,
|
| 90 |
+
"grad_norm": 14.384810447692871,
|
| 91 |
+
"learning_rate": 1.923076923076923e-05,
|
| 92 |
+
"loss": 0.4784,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.416,
|
| 97 |
+
"grad_norm": 12.508864402770996,
|
| 98 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 99 |
+
"loss": 0.4626,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.448,
|
| 104 |
+
"grad_norm": 14.830451011657715,
|
| 105 |
+
"learning_rate": 2.2435897435897437e-05,
|
| 106 |
+
"loss": 0.4459,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.48,
|
| 111 |
+
"grad_norm": 52.42909622192383,
|
| 112 |
+
"learning_rate": 2.4038461538461542e-05,
|
| 113 |
+
"loss": 0.4163,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.512,
|
| 118 |
+
"grad_norm": 32.466487884521484,
|
| 119 |
+
"learning_rate": 2.564102564102564e-05,
|
| 120 |
+
"loss": 0.4801,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.544,
|
| 125 |
+
"grad_norm": 30.3638973236084,
|
| 126 |
+
"learning_rate": 2.724358974358974e-05,
|
| 127 |
+
"loss": 0.4896,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.576,
|
| 132 |
+
"grad_norm": 27.437076568603516,
|
| 133 |
+
"learning_rate": 2.8846153846153845e-05,
|
| 134 |
+
"loss": 0.4598,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.608,
|
| 139 |
+
"grad_norm": 11.979604721069336,
|
| 140 |
+
"learning_rate": 3.0448717948717947e-05,
|
| 141 |
+
"loss": 0.421,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.64,
|
| 146 |
+
"grad_norm": 68.23617553710938,
|
| 147 |
+
"learning_rate": 3.205128205128206e-05,
|
| 148 |
+
"loss": 0.4518,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.672,
|
| 153 |
+
"grad_norm": 41.465309143066406,
|
| 154 |
+
"learning_rate": 3.365384615384616e-05,
|
| 155 |
+
"loss": 0.4403,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.704,
|
| 160 |
+
"grad_norm": 14.963610649108887,
|
| 161 |
+
"learning_rate": 3.525641025641026e-05,
|
| 162 |
+
"loss": 0.4282,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.736,
|
| 167 |
+
"grad_norm": 16.748451232910156,
|
| 168 |
+
"learning_rate": 3.685897435897436e-05,
|
| 169 |
+
"loss": 0.3553,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.768,
|
| 174 |
+
"grad_norm": 47.92767333984375,
|
| 175 |
+
"learning_rate": 3.846153846153846e-05,
|
| 176 |
+
"loss": 0.4068,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.8,
|
| 181 |
+
"grad_norm": 32.47563552856445,
|
| 182 |
+
"learning_rate": 4.006410256410257e-05,
|
| 183 |
+
"loss": 0.4356,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.832,
|
| 188 |
+
"grad_norm": 32.39358901977539,
|
| 189 |
+
"learning_rate": 4.166666666666667e-05,
|
| 190 |
+
"loss": 0.3873,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.864,
|
| 195 |
+
"grad_norm": 17.5948429107666,
|
| 196 |
+
"learning_rate": 4.326923076923077e-05,
|
| 197 |
+
"loss": 0.3428,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.896,
|
| 202 |
+
"grad_norm": 34.92612075805664,
|
| 203 |
+
"learning_rate": 4.4871794871794874e-05,
|
| 204 |
+
"loss": 0.4007,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.928,
|
| 209 |
+
"grad_norm": 12.984992980957031,
|
| 210 |
+
"learning_rate": 4.6474358974358976e-05,
|
| 211 |
+
"loss": 0.4148,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.96,
|
| 216 |
+
"grad_norm": 62.389808654785156,
|
| 217 |
+
"learning_rate": 4.8076923076923084e-05,
|
| 218 |
+
"loss": 0.3258,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.992,
|
| 223 |
+
"grad_norm": 20.061573028564453,
|
| 224 |
+
"learning_rate": 4.9679487179487185e-05,
|
| 225 |
+
"loss": 0.32,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9984,
|
| 230 |
+
"eval_accuracy": 0.6910798122065728,
|
| 231 |
+
"eval_loss": 0.9902069568634033,
|
| 232 |
+
"eval_runtime": 15.0205,
|
| 233 |
+
"eval_samples_per_second": 70.903,
|
| 234 |
+
"eval_steps_per_second": 2.264,
|
| 235 |
+
"step": 312
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.024,
|
| 239 |
+
"grad_norm": 95.87650299072266,
|
| 240 |
+
"learning_rate": 4.985754985754986e-05,
|
| 241 |
+
"loss": 0.4095,
|
| 242 |
+
"step": 320
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.056,
|
| 246 |
+
"grad_norm": 15.716072082519531,
|
| 247 |
+
"learning_rate": 4.9679487179487185e-05,
|
| 248 |
+
"loss": 0.3434,
|
| 249 |
+
"step": 330
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.088,
|
| 253 |
+
"grad_norm": 24.56471061706543,
|
| 254 |
+
"learning_rate": 4.95014245014245e-05,
|
| 255 |
+
"loss": 0.3115,
|
| 256 |
+
"step": 340
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.12,
|
| 260 |
+
"grad_norm": 27.441110610961914,
|
| 261 |
+
"learning_rate": 4.932336182336182e-05,
|
| 262 |
+
"loss": 0.3818,
|
| 263 |
+
"step": 350
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.152,
|
| 267 |
+
"grad_norm": 19.63530921936035,
|
| 268 |
+
"learning_rate": 4.9145299145299147e-05,
|
| 269 |
+
"loss": 0.4001,
|
| 270 |
+
"step": 360
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.184,
|
| 274 |
+
"grad_norm": 59.710182189941406,
|
| 275 |
+
"learning_rate": 4.896723646723647e-05,
|
| 276 |
+
"loss": 0.3836,
|
| 277 |
+
"step": 370
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.216,
|
| 281 |
+
"grad_norm": 11.739165306091309,
|
| 282 |
+
"learning_rate": 4.878917378917379e-05,
|
| 283 |
+
"loss": 0.3505,
|
| 284 |
+
"step": 380
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.248,
|
| 288 |
+
"grad_norm": 9.130122184753418,
|
| 289 |
+
"learning_rate": 4.8611111111111115e-05,
|
| 290 |
+
"loss": 0.2724,
|
| 291 |
+
"step": 390
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.28,
|
| 295 |
+
"grad_norm": 35.71299362182617,
|
| 296 |
+
"learning_rate": 4.8433048433048433e-05,
|
| 297 |
+
"loss": 0.2992,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.312,
|
| 302 |
+
"grad_norm": 23.9375,
|
| 303 |
+
"learning_rate": 4.825498575498576e-05,
|
| 304 |
+
"loss": 0.3046,
|
| 305 |
+
"step": 410
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.3439999999999999,
|
| 309 |
+
"grad_norm": 25.04848289489746,
|
| 310 |
+
"learning_rate": 4.8076923076923084e-05,
|
| 311 |
+
"loss": 0.315,
|
| 312 |
+
"step": 420
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.376,
|
| 316 |
+
"grad_norm": 18.383527755737305,
|
| 317 |
+
"learning_rate": 4.78988603988604e-05,
|
| 318 |
+
"loss": 0.3109,
|
| 319 |
+
"step": 430
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.408,
|
| 323 |
+
"grad_norm": 13.666149139404297,
|
| 324 |
+
"learning_rate": 4.772079772079772e-05,
|
| 325 |
+
"loss": 0.2999,
|
| 326 |
+
"step": 440
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.44,
|
| 330 |
+
"grad_norm": 9.764185905456543,
|
| 331 |
+
"learning_rate": 4.7542735042735045e-05,
|
| 332 |
+
"loss": 0.2615,
|
| 333 |
+
"step": 450
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.472,
|
| 337 |
+
"grad_norm": 15.155381202697754,
|
| 338 |
+
"learning_rate": 4.736467236467237e-05,
|
| 339 |
+
"loss": 0.261,
|
| 340 |
+
"step": 460
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.504,
|
| 344 |
+
"grad_norm": 14.417835235595703,
|
| 345 |
+
"learning_rate": 4.718660968660969e-05,
|
| 346 |
+
"loss": 0.2596,
|
| 347 |
+
"step": 470
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.536,
|
| 351 |
+
"grad_norm": 26.80813217163086,
|
| 352 |
+
"learning_rate": 4.700854700854701e-05,
|
| 353 |
+
"loss": 0.3011,
|
| 354 |
+
"step": 480
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.568,
|
| 358 |
+
"grad_norm": 29.056140899658203,
|
| 359 |
+
"learning_rate": 4.683048433048433e-05,
|
| 360 |
+
"loss": 0.2741,
|
| 361 |
+
"step": 490
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.6,
|
| 365 |
+
"grad_norm": 8.897234916687012,
|
| 366 |
+
"learning_rate": 4.665242165242166e-05,
|
| 367 |
+
"loss": 0.2906,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.6320000000000001,
|
| 372 |
+
"grad_norm": 15.488897323608398,
|
| 373 |
+
"learning_rate": 4.6474358974358976e-05,
|
| 374 |
+
"loss": 0.3148,
|
| 375 |
+
"step": 510
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.6640000000000001,
|
| 379 |
+
"grad_norm": 45.335018157958984,
|
| 380 |
+
"learning_rate": 4.62962962962963e-05,
|
| 381 |
+
"loss": 0.3026,
|
| 382 |
+
"step": 520
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.696,
|
| 386 |
+
"grad_norm": 8.628056526184082,
|
| 387 |
+
"learning_rate": 4.611823361823362e-05,
|
| 388 |
+
"loss": 0.2719,
|
| 389 |
+
"step": 530
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.728,
|
| 393 |
+
"grad_norm": 11.930350303649902,
|
| 394 |
+
"learning_rate": 4.594017094017094e-05,
|
| 395 |
+
"loss": 0.2831,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.76,
|
| 400 |
+
"grad_norm": 7.246833324432373,
|
| 401 |
+
"learning_rate": 4.576210826210827e-05,
|
| 402 |
+
"loss": 0.2292,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.792,
|
| 407 |
+
"grad_norm": 31.572784423828125,
|
| 408 |
+
"learning_rate": 4.558404558404559e-05,
|
| 409 |
+
"loss": 0.2698,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.8239999999999998,
|
| 414 |
+
"grad_norm": 9.255568504333496,
|
| 415 |
+
"learning_rate": 4.5405982905982906e-05,
|
| 416 |
+
"loss": 0.2469,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.8559999999999999,
|
| 421 |
+
"grad_norm": 13.890963554382324,
|
| 422 |
+
"learning_rate": 4.522792022792023e-05,
|
| 423 |
+
"loss": 0.2614,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.888,
|
| 428 |
+
"grad_norm": 13.776374816894531,
|
| 429 |
+
"learning_rate": 4.504985754985755e-05,
|
| 430 |
+
"loss": 0.264,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.92,
|
| 435 |
+
"grad_norm": 7.997173309326172,
|
| 436 |
+
"learning_rate": 4.4871794871794874e-05,
|
| 437 |
+
"loss": 0.2548,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.952,
|
| 442 |
+
"grad_norm": 44.6745719909668,
|
| 443 |
+
"learning_rate": 4.46937321937322e-05,
|
| 444 |
+
"loss": 0.2476,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.984,
|
| 449 |
+
"grad_norm": 13.378036499023438,
|
| 450 |
+
"learning_rate": 4.451566951566952e-05,
|
| 451 |
+
"loss": 0.2625,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.0,
|
| 456 |
+
"eval_accuracy": 0.7690140845070422,
|
| 457 |
+
"eval_loss": 0.5525702238082886,
|
| 458 |
+
"eval_runtime": 14.9955,
|
| 459 |
+
"eval_samples_per_second": 71.021,
|
| 460 |
+
"eval_steps_per_second": 2.267,
|
| 461 |
+
"step": 625
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.016,
|
| 465 |
+
"grad_norm": 7.805148601531982,
|
| 466 |
+
"learning_rate": 4.4337606837606836e-05,
|
| 467 |
+
"loss": 0.2232,
|
| 468 |
+
"step": 630
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 2.048,
|
| 472 |
+
"grad_norm": 10.30744457244873,
|
| 473 |
+
"learning_rate": 4.415954415954416e-05,
|
| 474 |
+
"loss": 0.2577,
|
| 475 |
+
"step": 640
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 2.08,
|
| 479 |
+
"grad_norm": 15.62030029296875,
|
| 480 |
+
"learning_rate": 4.3981481481481486e-05,
|
| 481 |
+
"loss": 0.2273,
|
| 482 |
+
"step": 650
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.112,
|
| 486 |
+
"grad_norm": 28.10418128967285,
|
| 487 |
+
"learning_rate": 4.3803418803418805e-05,
|
| 488 |
+
"loss": 0.243,
|
| 489 |
+
"step": 660
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.144,
|
| 493 |
+
"grad_norm": 9.574631690979004,
|
| 494 |
+
"learning_rate": 4.362535612535612e-05,
|
| 495 |
+
"loss": 0.2286,
|
| 496 |
+
"step": 670
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 2.176,
|
| 500 |
+
"grad_norm": 14.983197212219238,
|
| 501 |
+
"learning_rate": 4.344729344729345e-05,
|
| 502 |
+
"loss": 0.2173,
|
| 503 |
+
"step": 680
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.208,
|
| 507 |
+
"grad_norm": 19.074600219726562,
|
| 508 |
+
"learning_rate": 4.326923076923077e-05,
|
| 509 |
+
"loss": 0.2339,
|
| 510 |
+
"step": 690
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 2.24,
|
| 514 |
+
"grad_norm": 27.23953628540039,
|
| 515 |
+
"learning_rate": 4.309116809116809e-05,
|
| 516 |
+
"loss": 0.2008,
|
| 517 |
+
"step": 700
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.2720000000000002,
|
| 521 |
+
"grad_norm": 13.12559986114502,
|
| 522 |
+
"learning_rate": 4.291310541310542e-05,
|
| 523 |
+
"loss": 0.1996,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.304,
|
| 528 |
+
"grad_norm": 22.73211097717285,
|
| 529 |
+
"learning_rate": 4.2735042735042735e-05,
|
| 530 |
+
"loss": 0.1986,
|
| 531 |
+
"step": 720
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.336,
|
| 535 |
+
"grad_norm": 12.57758903503418,
|
| 536 |
+
"learning_rate": 4.255698005698006e-05,
|
| 537 |
+
"loss": 0.2283,
|
| 538 |
+
"step": 730
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.368,
|
| 542 |
+
"grad_norm": 61.22172927856445,
|
| 543 |
+
"learning_rate": 4.2378917378917385e-05,
|
| 544 |
+
"loss": 0.2175,
|
| 545 |
+
"step": 740
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.4,
|
| 549 |
+
"grad_norm": 14.797411918640137,
|
| 550 |
+
"learning_rate": 4.2200854700854704e-05,
|
| 551 |
+
"loss": 0.1758,
|
| 552 |
+
"step": 750
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.432,
|
| 556 |
+
"grad_norm": 9.504317283630371,
|
| 557 |
+
"learning_rate": 4.202279202279202e-05,
|
| 558 |
+
"loss": 0.2087,
|
| 559 |
+
"step": 760
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 2.464,
|
| 563 |
+
"grad_norm": 8.436298370361328,
|
| 564 |
+
"learning_rate": 4.184472934472935e-05,
|
| 565 |
+
"loss": 0.2318,
|
| 566 |
+
"step": 770
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.496,
|
| 570 |
+
"grad_norm": 31.693523406982422,
|
| 571 |
+
"learning_rate": 4.166666666666667e-05,
|
| 572 |
+
"loss": 0.1909,
|
| 573 |
+
"step": 780
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.528,
|
| 577 |
+
"grad_norm": 14.50406265258789,
|
| 578 |
+
"learning_rate": 4.148860398860399e-05,
|
| 579 |
+
"loss": 0.2157,
|
| 580 |
+
"step": 790
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.56,
|
| 584 |
+
"grad_norm": 25.511077880859375,
|
| 585 |
+
"learning_rate": 4.131054131054131e-05,
|
| 586 |
+
"loss": 0.2237,
|
| 587 |
+
"step": 800
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.592,
|
| 591 |
+
"grad_norm": 6.311196804046631,
|
| 592 |
+
"learning_rate": 4.1132478632478634e-05,
|
| 593 |
+
"loss": 0.1922,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.624,
|
| 598 |
+
"grad_norm": 37.99801254272461,
|
| 599 |
+
"learning_rate": 4.095441595441596e-05,
|
| 600 |
+
"loss": 0.1835,
|
| 601 |
+
"step": 820
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.656,
|
| 605 |
+
"grad_norm": 48.04703140258789,
|
| 606 |
+
"learning_rate": 4.077635327635328e-05,
|
| 607 |
+
"loss": 0.191,
|
| 608 |
+
"step": 830
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.6879999999999997,
|
| 612 |
+
"grad_norm": 52.908992767333984,
|
| 613 |
+
"learning_rate": 4.05982905982906e-05,
|
| 614 |
+
"loss": 0.2908,
|
| 615 |
+
"step": 840
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.7199999999999998,
|
| 619 |
+
"grad_norm": 8.987102508544922,
|
| 620 |
+
"learning_rate": 4.042022792022792e-05,
|
| 621 |
+
"loss": 0.216,
|
| 622 |
+
"step": 850
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 2.752,
|
| 626 |
+
"grad_norm": 44.67879104614258,
|
| 627 |
+
"learning_rate": 4.024216524216524e-05,
|
| 628 |
+
"loss": 0.1991,
|
| 629 |
+
"step": 860
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 2.784,
|
| 633 |
+
"grad_norm": 23.197927474975586,
|
| 634 |
+
"learning_rate": 4.006410256410257e-05,
|
| 635 |
+
"loss": 0.1751,
|
| 636 |
+
"step": 870
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 2.816,
|
| 640 |
+
"grad_norm": 29.43416404724121,
|
| 641 |
+
"learning_rate": 3.988603988603989e-05,
|
| 642 |
+
"loss": 0.2064,
|
| 643 |
+
"step": 880
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 2.848,
|
| 647 |
+
"grad_norm": 10.251945495605469,
|
| 648 |
+
"learning_rate": 3.970797720797721e-05,
|
| 649 |
+
"loss": 0.178,
|
| 650 |
+
"step": 890
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.88,
|
| 654 |
+
"grad_norm": 29.398643493652344,
|
| 655 |
+
"learning_rate": 3.952991452991453e-05,
|
| 656 |
+
"loss": 0.1813,
|
| 657 |
+
"step": 900
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.912,
|
| 661 |
+
"grad_norm": 37.31789016723633,
|
| 662 |
+
"learning_rate": 3.935185185185186e-05,
|
| 663 |
+
"loss": 0.1778,
|
| 664 |
+
"step": 910
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 2.944,
|
| 668 |
+
"grad_norm": 21.298931121826172,
|
| 669 |
+
"learning_rate": 3.9173789173789176e-05,
|
| 670 |
+
"loss": 0.2124,
|
| 671 |
+
"step": 920
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 2.976,
|
| 675 |
+
"grad_norm": 10.21948528289795,
|
| 676 |
+
"learning_rate": 3.89957264957265e-05,
|
| 677 |
+
"loss": 0.1584,
|
| 678 |
+
"step": 930
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.9984,
|
| 682 |
+
"eval_accuracy": 0.8018779342723005,
|
| 683 |
+
"eval_loss": 0.560497522354126,
|
| 684 |
+
"eval_runtime": 14.76,
|
| 685 |
+
"eval_samples_per_second": 72.155,
|
| 686 |
+
"eval_steps_per_second": 2.304,
|
| 687 |
+
"step": 937
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.008,
|
| 691 |
+
"grad_norm": 9.502830505371094,
|
| 692 |
+
"learning_rate": 3.881766381766382e-05,
|
| 693 |
+
"loss": 0.1569,
|
| 694 |
+
"step": 940
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 3.04,
|
| 698 |
+
"grad_norm": 23.640422821044922,
|
| 699 |
+
"learning_rate": 3.863960113960114e-05,
|
| 700 |
+
"loss": 0.2172,
|
| 701 |
+
"step": 950
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.072,
|
| 705 |
+
"grad_norm": 7.1912312507629395,
|
| 706 |
+
"learning_rate": 3.846153846153846e-05,
|
| 707 |
+
"loss": 0.1954,
|
| 708 |
+
"step": 960
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.104,
|
| 712 |
+
"grad_norm": 61.15785598754883,
|
| 713 |
+
"learning_rate": 3.828347578347579e-05,
|
| 714 |
+
"loss": 0.2364,
|
| 715 |
+
"step": 970
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 3.136,
|
| 719 |
+
"grad_norm": 11.975686073303223,
|
| 720 |
+
"learning_rate": 3.8105413105413106e-05,
|
| 721 |
+
"loss": 0.1856,
|
| 722 |
+
"step": 980
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.168,
|
| 726 |
+
"grad_norm": 13.5460844039917,
|
| 727 |
+
"learning_rate": 3.7927350427350425e-05,
|
| 728 |
+
"loss": 0.1608,
|
| 729 |
+
"step": 990
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.2,
|
| 733 |
+
"grad_norm": 13.04128646850586,
|
| 734 |
+
"learning_rate": 3.774928774928775e-05,
|
| 735 |
+
"loss": 0.1485,
|
| 736 |
+
"step": 1000
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 3.232,
|
| 740 |
+
"grad_norm": 17.87329864501953,
|
| 741 |
+
"learning_rate": 3.7571225071225075e-05,
|
| 742 |
+
"loss": 0.1782,
|
| 743 |
+
"step": 1010
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.2640000000000002,
|
| 747 |
+
"grad_norm": 21.913515090942383,
|
| 748 |
+
"learning_rate": 3.739316239316239e-05,
|
| 749 |
+
"loss": 0.1636,
|
| 750 |
+
"step": 1020
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.296,
|
| 754 |
+
"grad_norm": 53.9290771484375,
|
| 755 |
+
"learning_rate": 3.721509971509972e-05,
|
| 756 |
+
"loss": 0.1869,
|
| 757 |
+
"step": 1030
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.328,
|
| 761 |
+
"grad_norm": 12.648904800415039,
|
| 762 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 763 |
+
"loss": 0.1374,
|
| 764 |
+
"step": 1040
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.36,
|
| 768 |
+
"grad_norm": 30.469200134277344,
|
| 769 |
+
"learning_rate": 3.685897435897436e-05,
|
| 770 |
+
"loss": 0.1677,
|
| 771 |
+
"step": 1050
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.392,
|
| 775 |
+
"grad_norm": 53.974571228027344,
|
| 776 |
+
"learning_rate": 3.668091168091169e-05,
|
| 777 |
+
"loss": 0.1297,
|
| 778 |
+
"step": 1060
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.424,
|
| 782 |
+
"grad_norm": 13.038790702819824,
|
| 783 |
+
"learning_rate": 3.6502849002849005e-05,
|
| 784 |
+
"loss": 0.1773,
|
| 785 |
+
"step": 1070
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.456,
|
| 789 |
+
"grad_norm": 10.653145790100098,
|
| 790 |
+
"learning_rate": 3.6324786324786323e-05,
|
| 791 |
+
"loss": 0.1672,
|
| 792 |
+
"step": 1080
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.488,
|
| 796 |
+
"grad_norm": 19.27906608581543,
|
| 797 |
+
"learning_rate": 3.614672364672365e-05,
|
| 798 |
+
"loss": 0.1587,
|
| 799 |
+
"step": 1090
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.52,
|
| 803 |
+
"grad_norm": 9.866512298583984,
|
| 804 |
+
"learning_rate": 3.5968660968660974e-05,
|
| 805 |
+
"loss": 0.1566,
|
| 806 |
+
"step": 1100
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.552,
|
| 810 |
+
"grad_norm": 24.01140785217285,
|
| 811 |
+
"learning_rate": 3.579059829059829e-05,
|
| 812 |
+
"loss": 0.1546,
|
| 813 |
+
"step": 1110
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.584,
|
| 817 |
+
"grad_norm": 33.78230285644531,
|
| 818 |
+
"learning_rate": 3.561253561253561e-05,
|
| 819 |
+
"loss": 0.1769,
|
| 820 |
+
"step": 1120
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.616,
|
| 824 |
+
"grad_norm": 29.797826766967773,
|
| 825 |
+
"learning_rate": 3.5434472934472935e-05,
|
| 826 |
+
"loss": 0.1477,
|
| 827 |
+
"step": 1130
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.648,
|
| 831 |
+
"grad_norm": 11.370080947875977,
|
| 832 |
+
"learning_rate": 3.525641025641026e-05,
|
| 833 |
+
"loss": 0.1749,
|
| 834 |
+
"step": 1140
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.68,
|
| 838 |
+
"grad_norm": 6.187923908233643,
|
| 839 |
+
"learning_rate": 3.507834757834758e-05,
|
| 840 |
+
"loss": 0.1187,
|
| 841 |
+
"step": 1150
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.7119999999999997,
|
| 845 |
+
"grad_norm": 57.243919372558594,
|
| 846 |
+
"learning_rate": 3.4900284900284904e-05,
|
| 847 |
+
"loss": 0.1455,
|
| 848 |
+
"step": 1160
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.7439999999999998,
|
| 852 |
+
"grad_norm": 48.05120849609375,
|
| 853 |
+
"learning_rate": 3.472222222222222e-05,
|
| 854 |
+
"loss": 0.1792,
|
| 855 |
+
"step": 1170
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.776,
|
| 859 |
+
"grad_norm": 18.508798599243164,
|
| 860 |
+
"learning_rate": 3.454415954415954e-05,
|
| 861 |
+
"loss": 0.1977,
|
| 862 |
+
"step": 1180
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.808,
|
| 866 |
+
"grad_norm": 11.206692695617676,
|
| 867 |
+
"learning_rate": 3.436609686609687e-05,
|
| 868 |
+
"loss": 0.1752,
|
| 869 |
+
"step": 1190
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.84,
|
| 873 |
+
"grad_norm": 8.796914100646973,
|
| 874 |
+
"learning_rate": 3.418803418803419e-05,
|
| 875 |
+
"loss": 0.134,
|
| 876 |
+
"step": 1200
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.872,
|
| 880 |
+
"grad_norm": 28.23493003845215,
|
| 881 |
+
"learning_rate": 3.400997150997151e-05,
|
| 882 |
+
"loss": 0.1466,
|
| 883 |
+
"step": 1210
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 3.904,
|
| 887 |
+
"grad_norm": 24.789331436157227,
|
| 888 |
+
"learning_rate": 3.3831908831908834e-05,
|
| 889 |
+
"loss": 0.1674,
|
| 890 |
+
"step": 1220
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 3.936,
|
| 894 |
+
"grad_norm": 29.97929573059082,
|
| 895 |
+
"learning_rate": 3.365384615384616e-05,
|
| 896 |
+
"loss": 0.1626,
|
| 897 |
+
"step": 1230
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 3.968,
|
| 901 |
+
"grad_norm": 12.732221603393555,
|
| 902 |
+
"learning_rate": 3.347578347578348e-05,
|
| 903 |
+
"loss": 0.1162,
|
| 904 |
+
"step": 1240
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 4.0,
|
| 908 |
+
"grad_norm": 23.768869400024414,
|
| 909 |
+
"learning_rate": 3.32977207977208e-05,
|
| 910 |
+
"loss": 0.1382,
|
| 911 |
+
"step": 1250
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 4.0,
|
| 915 |
+
"eval_accuracy": 0.8497652582159625,
|
| 916 |
+
"eval_loss": 0.429149866104126,
|
| 917 |
+
"eval_runtime": 14.7304,
|
| 918 |
+
"eval_samples_per_second": 72.299,
|
| 919 |
+
"eval_steps_per_second": 2.308,
|
| 920 |
+
"step": 1250
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 4.032,
|
| 924 |
+
"grad_norm": 26.3099308013916,
|
| 925 |
+
"learning_rate": 3.311965811965812e-05,
|
| 926 |
+
"loss": 0.1279,
|
| 927 |
+
"step": 1260
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 4.064,
|
| 931 |
+
"grad_norm": 26.10764503479004,
|
| 932 |
+
"learning_rate": 3.294159544159544e-05,
|
| 933 |
+
"loss": 0.1868,
|
| 934 |
+
"step": 1270
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 4.096,
|
| 938 |
+
"grad_norm": 42.908058166503906,
|
| 939 |
+
"learning_rate": 3.2763532763532764e-05,
|
| 940 |
+
"loss": 0.1102,
|
| 941 |
+
"step": 1280
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 4.128,
|
| 945 |
+
"grad_norm": 18.12682342529297,
|
| 946 |
+
"learning_rate": 3.258547008547009e-05,
|
| 947 |
+
"loss": 0.1372,
|
| 948 |
+
"step": 1290
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 4.16,
|
| 952 |
+
"grad_norm": 32.7211799621582,
|
| 953 |
+
"learning_rate": 3.240740740740741e-05,
|
| 954 |
+
"loss": 0.1342,
|
| 955 |
+
"step": 1300
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 4.192,
|
| 959 |
+
"grad_norm": 43.664955139160156,
|
| 960 |
+
"learning_rate": 3.2229344729344726e-05,
|
| 961 |
+
"loss": 0.1205,
|
| 962 |
+
"step": 1310
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 4.224,
|
| 966 |
+
"grad_norm": 21.014827728271484,
|
| 967 |
+
"learning_rate": 3.205128205128206e-05,
|
| 968 |
+
"loss": 0.1328,
|
| 969 |
+
"step": 1320
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 4.256,
|
| 973 |
+
"grad_norm": 9.477145195007324,
|
| 974 |
+
"learning_rate": 3.1873219373219376e-05,
|
| 975 |
+
"loss": 0.12,
|
| 976 |
+
"step": 1330
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 4.288,
|
| 980 |
+
"grad_norm": 11.2942533493042,
|
| 981 |
+
"learning_rate": 3.1695156695156695e-05,
|
| 982 |
+
"loss": 0.1033,
|
| 983 |
+
"step": 1340
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 4.32,
|
| 987 |
+
"grad_norm": 19.64799690246582,
|
| 988 |
+
"learning_rate": 3.151709401709402e-05,
|
| 989 |
+
"loss": 0.1157,
|
| 990 |
+
"step": 1350
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 4.352,
|
| 994 |
+
"grad_norm": 26.556772232055664,
|
| 995 |
+
"learning_rate": 3.133903133903134e-05,
|
| 996 |
+
"loss": 0.1053,
|
| 997 |
+
"step": 1360
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 4.384,
|
| 1001 |
+
"grad_norm": 17.614032745361328,
|
| 1002 |
+
"learning_rate": 3.116096866096866e-05,
|
| 1003 |
+
"loss": 0.1266,
|
| 1004 |
+
"step": 1370
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 4.416,
|
| 1008 |
+
"grad_norm": 6.941921710968018,
|
| 1009 |
+
"learning_rate": 3.098290598290599e-05,
|
| 1010 |
+
"loss": 0.1177,
|
| 1011 |
+
"step": 1380
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 4.448,
|
| 1015 |
+
"grad_norm": 12.218695640563965,
|
| 1016 |
+
"learning_rate": 3.080484330484331e-05,
|
| 1017 |
+
"loss": 0.1209,
|
| 1018 |
+
"step": 1390
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 4.48,
|
| 1022 |
+
"grad_norm": 43.25284194946289,
|
| 1023 |
+
"learning_rate": 3.0626780626780625e-05,
|
| 1024 |
+
"loss": 0.1346,
|
| 1025 |
+
"step": 1400
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 4.5120000000000005,
|
| 1029 |
+
"grad_norm": 13.644841194152832,
|
| 1030 |
+
"learning_rate": 3.0448717948717947e-05,
|
| 1031 |
+
"loss": 0.1156,
|
| 1032 |
+
"step": 1410
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 4.5440000000000005,
|
| 1036 |
+
"grad_norm": 66.71968078613281,
|
| 1037 |
+
"learning_rate": 3.0270655270655275e-05,
|
| 1038 |
+
"loss": 0.1197,
|
| 1039 |
+
"step": 1420
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 4.576,
|
| 1043 |
+
"grad_norm": 59.87635040283203,
|
| 1044 |
+
"learning_rate": 3.0092592592592593e-05,
|
| 1045 |
+
"loss": 0.1105,
|
| 1046 |
+
"step": 1430
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 4.608,
|
| 1050 |
+
"grad_norm": 10.875226020812988,
|
| 1051 |
+
"learning_rate": 2.9914529914529915e-05,
|
| 1052 |
+
"loss": 0.0932,
|
| 1053 |
+
"step": 1440
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 4.64,
|
| 1057 |
+
"grad_norm": 21.233959197998047,
|
| 1058 |
+
"learning_rate": 2.9736467236467237e-05,
|
| 1059 |
+
"loss": 0.0914,
|
| 1060 |
+
"step": 1450
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 4.672,
|
| 1064 |
+
"grad_norm": 25.02607536315918,
|
| 1065 |
+
"learning_rate": 2.9558404558404562e-05,
|
| 1066 |
+
"loss": 0.1175,
|
| 1067 |
+
"step": 1460
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 4.704,
|
| 1071 |
+
"grad_norm": 9.968689918518066,
|
| 1072 |
+
"learning_rate": 2.9380341880341884e-05,
|
| 1073 |
+
"loss": 0.1111,
|
| 1074 |
+
"step": 1470
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 4.736,
|
| 1078 |
+
"grad_norm": 19.770673751831055,
|
| 1079 |
+
"learning_rate": 2.9202279202279202e-05,
|
| 1080 |
+
"loss": 0.1627,
|
| 1081 |
+
"step": 1480
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 4.768,
|
| 1085 |
+
"grad_norm": 16.976701736450195,
|
| 1086 |
+
"learning_rate": 2.9024216524216524e-05,
|
| 1087 |
+
"loss": 0.1355,
|
| 1088 |
+
"step": 1490
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 4.8,
|
| 1092 |
+
"grad_norm": 32.00942611694336,
|
| 1093 |
+
"learning_rate": 2.8846153846153845e-05,
|
| 1094 |
+
"loss": 0.1339,
|
| 1095 |
+
"step": 1500
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 4.832,
|
| 1099 |
+
"grad_norm": 12.106451034545898,
|
| 1100 |
+
"learning_rate": 2.866809116809117e-05,
|
| 1101 |
+
"loss": 0.1192,
|
| 1102 |
+
"step": 1510
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 4.864,
|
| 1106 |
+
"grad_norm": 20.953189849853516,
|
| 1107 |
+
"learning_rate": 2.8490028490028492e-05,
|
| 1108 |
+
"loss": 0.1295,
|
| 1109 |
+
"step": 1520
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 4.896,
|
| 1113 |
+
"grad_norm": 18.536453247070312,
|
| 1114 |
+
"learning_rate": 2.8311965811965814e-05,
|
| 1115 |
+
"loss": 0.101,
|
| 1116 |
+
"step": 1530
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 4.928,
|
| 1120 |
+
"grad_norm": 14.893009185791016,
|
| 1121 |
+
"learning_rate": 2.8133903133903132e-05,
|
| 1122 |
+
"loss": 0.112,
|
| 1123 |
+
"step": 1540
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 4.96,
|
| 1127 |
+
"grad_norm": 10.388225555419922,
|
| 1128 |
+
"learning_rate": 2.795584045584046e-05,
|
| 1129 |
+
"loss": 0.1166,
|
| 1130 |
+
"step": 1550
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 4.992,
|
| 1134 |
+
"grad_norm": 15.933597564697266,
|
| 1135 |
+
"learning_rate": 2.777777777777778e-05,
|
| 1136 |
+
"loss": 0.1058,
|
| 1137 |
+
"step": 1560
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 4.9984,
|
| 1141 |
+
"eval_accuracy": 0.8666666666666667,
|
| 1142 |
+
"eval_loss": 0.39110323786735535,
|
| 1143 |
+
"eval_runtime": 14.8044,
|
| 1144 |
+
"eval_samples_per_second": 71.938,
|
| 1145 |
+
"eval_steps_per_second": 2.297,
|
| 1146 |
+
"step": 1562
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 5.024,
|
| 1150 |
+
"grad_norm": 21.4670352935791,
|
| 1151 |
+
"learning_rate": 2.75997150997151e-05,
|
| 1152 |
+
"loss": 0.1265,
|
| 1153 |
+
"step": 1570
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 5.056,
|
| 1157 |
+
"grad_norm": 22.757436752319336,
|
| 1158 |
+
"learning_rate": 2.7421652421652423e-05,
|
| 1159 |
+
"loss": 0.1,
|
| 1160 |
+
"step": 1580
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 5.088,
|
| 1164 |
+
"grad_norm": 11.63531494140625,
|
| 1165 |
+
"learning_rate": 2.724358974358974e-05,
|
| 1166 |
+
"loss": 0.1181,
|
| 1167 |
+
"step": 1590
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 5.12,
|
| 1171 |
+
"grad_norm": 11.879034996032715,
|
| 1172 |
+
"learning_rate": 2.706552706552707e-05,
|
| 1173 |
+
"loss": 0.1001,
|
| 1174 |
+
"step": 1600
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 5.152,
|
| 1178 |
+
"grad_norm": 13.50586223602295,
|
| 1179 |
+
"learning_rate": 2.688746438746439e-05,
|
| 1180 |
+
"loss": 0.098,
|
| 1181 |
+
"step": 1610
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 5.184,
|
| 1185 |
+
"grad_norm": 6.154415130615234,
|
| 1186 |
+
"learning_rate": 2.670940170940171e-05,
|
| 1187 |
+
"loss": 0.1005,
|
| 1188 |
+
"step": 1620
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 5.216,
|
| 1192 |
+
"grad_norm": 8.229863166809082,
|
| 1193 |
+
"learning_rate": 2.653133903133903e-05,
|
| 1194 |
+
"loss": 0.1044,
|
| 1195 |
+
"step": 1630
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 5.248,
|
| 1199 |
+
"grad_norm": 18.829721450805664,
|
| 1200 |
+
"learning_rate": 2.6353276353276356e-05,
|
| 1201 |
+
"loss": 0.1118,
|
| 1202 |
+
"step": 1640
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 5.28,
|
| 1206 |
+
"grad_norm": 17.462610244750977,
|
| 1207 |
+
"learning_rate": 2.6175213675213678e-05,
|
| 1208 |
+
"loss": 0.0809,
|
| 1209 |
+
"step": 1650
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 5.312,
|
| 1213 |
+
"grad_norm": 15.725268363952637,
|
| 1214 |
+
"learning_rate": 2.5997150997151e-05,
|
| 1215 |
+
"loss": 0.109,
|
| 1216 |
+
"step": 1660
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 5.344,
|
| 1220 |
+
"grad_norm": 12.376166343688965,
|
| 1221 |
+
"learning_rate": 2.5819088319088318e-05,
|
| 1222 |
+
"loss": 0.0775,
|
| 1223 |
+
"step": 1670
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 5.376,
|
| 1227 |
+
"grad_norm": 9.472005844116211,
|
| 1228 |
+
"learning_rate": 2.564102564102564e-05,
|
| 1229 |
+
"loss": 0.0958,
|
| 1230 |
+
"step": 1680
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 5.408,
|
| 1234 |
+
"grad_norm": 16.398996353149414,
|
| 1235 |
+
"learning_rate": 2.5462962962962965e-05,
|
| 1236 |
+
"loss": 0.0818,
|
| 1237 |
+
"step": 1690
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 5.44,
|
| 1241 |
+
"grad_norm": 15.111603736877441,
|
| 1242 |
+
"learning_rate": 2.5284900284900286e-05,
|
| 1243 |
+
"loss": 0.0945,
|
| 1244 |
+
"step": 1700
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 5.4719999999999995,
|
| 1248 |
+
"grad_norm": 13.802181243896484,
|
| 1249 |
+
"learning_rate": 2.5106837606837608e-05,
|
| 1250 |
+
"loss": 0.0694,
|
| 1251 |
+
"step": 1710
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 5.504,
|
| 1255 |
+
"grad_norm": 6.6055216789245605,
|
| 1256 |
+
"learning_rate": 2.492877492877493e-05,
|
| 1257 |
+
"loss": 0.0886,
|
| 1258 |
+
"step": 1720
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 5.536,
|
| 1262 |
+
"grad_norm": 17.351755142211914,
|
| 1263 |
+
"learning_rate": 2.475071225071225e-05,
|
| 1264 |
+
"loss": 0.0939,
|
| 1265 |
+
"step": 1730
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 5.568,
|
| 1269 |
+
"grad_norm": 14.629916191101074,
|
| 1270 |
+
"learning_rate": 2.4572649572649573e-05,
|
| 1271 |
+
"loss": 0.0749,
|
| 1272 |
+
"step": 1740
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 5.6,
|
| 1276 |
+
"grad_norm": 13.987975120544434,
|
| 1277 |
+
"learning_rate": 2.4394586894586895e-05,
|
| 1278 |
+
"loss": 0.0696,
|
| 1279 |
+
"step": 1750
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 5.632,
|
| 1283 |
+
"grad_norm": 19.911739349365234,
|
| 1284 |
+
"learning_rate": 2.4216524216524217e-05,
|
| 1285 |
+
"loss": 0.0829,
|
| 1286 |
+
"step": 1760
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 5.664,
|
| 1290 |
+
"grad_norm": 11.428616523742676,
|
| 1291 |
+
"learning_rate": 2.4038461538461542e-05,
|
| 1292 |
+
"loss": 0.0816,
|
| 1293 |
+
"step": 1770
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 5.696,
|
| 1297 |
+
"grad_norm": 24.872535705566406,
|
| 1298 |
+
"learning_rate": 2.386039886039886e-05,
|
| 1299 |
+
"loss": 0.0876,
|
| 1300 |
+
"step": 1780
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 5.728,
|
| 1304 |
+
"grad_norm": 6.562127113342285,
|
| 1305 |
+
"learning_rate": 2.3682336182336185e-05,
|
| 1306 |
+
"loss": 0.0869,
|
| 1307 |
+
"step": 1790
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 5.76,
|
| 1311 |
+
"grad_norm": 14.26649284362793,
|
| 1312 |
+
"learning_rate": 2.3504273504273504e-05,
|
| 1313 |
+
"loss": 0.1098,
|
| 1314 |
+
"step": 1800
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 5.792,
|
| 1318 |
+
"grad_norm": 10.148120880126953,
|
| 1319 |
+
"learning_rate": 2.332621082621083e-05,
|
| 1320 |
+
"loss": 0.0725,
|
| 1321 |
+
"step": 1810
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 5.824,
|
| 1325 |
+
"grad_norm": 31.81497573852539,
|
| 1326 |
+
"learning_rate": 2.314814814814815e-05,
|
| 1327 |
+
"loss": 0.0599,
|
| 1328 |
+
"step": 1820
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 5.856,
|
| 1332 |
+
"grad_norm": 20.058345794677734,
|
| 1333 |
+
"learning_rate": 2.297008547008547e-05,
|
| 1334 |
+
"loss": 0.0987,
|
| 1335 |
+
"step": 1830
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 5.888,
|
| 1339 |
+
"grad_norm": 23.388731002807617,
|
| 1340 |
+
"learning_rate": 2.2792022792022794e-05,
|
| 1341 |
+
"loss": 0.0993,
|
| 1342 |
+
"step": 1840
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 5.92,
|
| 1346 |
+
"grad_norm": 12.417990684509277,
|
| 1347 |
+
"learning_rate": 2.2613960113960116e-05,
|
| 1348 |
+
"loss": 0.0997,
|
| 1349 |
+
"step": 1850
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 5.952,
|
| 1353 |
+
"grad_norm": 18.029085159301758,
|
| 1354 |
+
"learning_rate": 2.2435897435897437e-05,
|
| 1355 |
+
"loss": 0.0875,
|
| 1356 |
+
"step": 1860
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 5.984,
|
| 1360 |
+
"grad_norm": 8.003488540649414,
|
| 1361 |
+
"learning_rate": 2.225783475783476e-05,
|
| 1362 |
+
"loss": 0.0703,
|
| 1363 |
+
"step": 1870
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 6.0,
|
| 1367 |
+
"eval_accuracy": 0.8028169014084507,
|
| 1368 |
+
"eval_loss": 0.8593423366546631,
|
| 1369 |
+
"eval_runtime": 26.4195,
|
| 1370 |
+
"eval_samples_per_second": 40.311,
|
| 1371 |
+
"eval_steps_per_second": 1.287,
|
| 1372 |
+
"step": 1875
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 6.016,
|
| 1376 |
+
"grad_norm": 13.387948036193848,
|
| 1377 |
+
"learning_rate": 2.207977207977208e-05,
|
| 1378 |
+
"loss": 0.0723,
|
| 1379 |
+
"step": 1880
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 6.048,
|
| 1383 |
+
"grad_norm": 3.795748472213745,
|
| 1384 |
+
"learning_rate": 2.1901709401709402e-05,
|
| 1385 |
+
"loss": 0.067,
|
| 1386 |
+
"step": 1890
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 6.08,
|
| 1390 |
+
"grad_norm": 15.893839836120605,
|
| 1391 |
+
"learning_rate": 2.1723646723646724e-05,
|
| 1392 |
+
"loss": 0.0718,
|
| 1393 |
+
"step": 1900
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 6.112,
|
| 1397 |
+
"grad_norm": 6.81917142868042,
|
| 1398 |
+
"learning_rate": 2.1545584045584046e-05,
|
| 1399 |
+
"loss": 0.0738,
|
| 1400 |
+
"step": 1910
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 6.144,
|
| 1404 |
+
"grad_norm": 16.609046936035156,
|
| 1405 |
+
"learning_rate": 2.1367521367521368e-05,
|
| 1406 |
+
"loss": 0.0698,
|
| 1407 |
+
"step": 1920
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 6.176,
|
| 1411 |
+
"grad_norm": 21.796428680419922,
|
| 1412 |
+
"learning_rate": 2.1189458689458693e-05,
|
| 1413 |
+
"loss": 0.0681,
|
| 1414 |
+
"step": 1930
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 6.208,
|
| 1418 |
+
"grad_norm": 9.367144584655762,
|
| 1419 |
+
"learning_rate": 2.101139601139601e-05,
|
| 1420 |
+
"loss": 0.0776,
|
| 1421 |
+
"step": 1940
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 6.24,
|
| 1425 |
+
"grad_norm": 45.68296432495117,
|
| 1426 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 1427 |
+
"loss": 0.075,
|
| 1428 |
+
"step": 1950
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 6.272,
|
| 1432 |
+
"grad_norm": 22.519546508789062,
|
| 1433 |
+
"learning_rate": 2.0655270655270654e-05,
|
| 1434 |
+
"loss": 0.1017,
|
| 1435 |
+
"step": 1960
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 6.304,
|
| 1439 |
+
"grad_norm": 26.589696884155273,
|
| 1440 |
+
"learning_rate": 2.047720797720798e-05,
|
| 1441 |
+
"loss": 0.0699,
|
| 1442 |
+
"step": 1970
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 6.336,
|
| 1446 |
+
"grad_norm": 10.323519706726074,
|
| 1447 |
+
"learning_rate": 2.02991452991453e-05,
|
| 1448 |
+
"loss": 0.0628,
|
| 1449 |
+
"step": 1980
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 6.368,
|
| 1453 |
+
"grad_norm": 19.11711311340332,
|
| 1454 |
+
"learning_rate": 2.012108262108262e-05,
|
| 1455 |
+
"loss": 0.0812,
|
| 1456 |
+
"step": 1990
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 6.4,
|
| 1460 |
+
"grad_norm": 33.67527770996094,
|
| 1461 |
+
"learning_rate": 1.9943019943019945e-05,
|
| 1462 |
+
"loss": 0.085,
|
| 1463 |
+
"step": 2000
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 6.432,
|
| 1467 |
+
"grad_norm": 11.990752220153809,
|
| 1468 |
+
"learning_rate": 1.9764957264957266e-05,
|
| 1469 |
+
"loss": 0.0702,
|
| 1470 |
+
"step": 2010
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 6.464,
|
| 1474 |
+
"grad_norm": 30.36664581298828,
|
| 1475 |
+
"learning_rate": 1.9586894586894588e-05,
|
| 1476 |
+
"loss": 0.0628,
|
| 1477 |
+
"step": 2020
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 6.496,
|
| 1481 |
+
"grad_norm": 21.322391510009766,
|
| 1482 |
+
"learning_rate": 1.940883190883191e-05,
|
| 1483 |
+
"loss": 0.0719,
|
| 1484 |
+
"step": 2030
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 6.5280000000000005,
|
| 1488 |
+
"grad_norm": 29.107898712158203,
|
| 1489 |
+
"learning_rate": 1.923076923076923e-05,
|
| 1490 |
+
"loss": 0.0525,
|
| 1491 |
+
"step": 2040
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 6.5600000000000005,
|
| 1495 |
+
"grad_norm": 17.11159324645996,
|
| 1496 |
+
"learning_rate": 1.9052706552706553e-05,
|
| 1497 |
+
"loss": 0.0687,
|
| 1498 |
+
"step": 2050
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 6.592,
|
| 1502 |
+
"grad_norm": 12.946314811706543,
|
| 1503 |
+
"learning_rate": 1.8874643874643875e-05,
|
| 1504 |
+
"loss": 0.0867,
|
| 1505 |
+
"step": 2060
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 6.624,
|
| 1509 |
+
"grad_norm": 7.321907997131348,
|
| 1510 |
+
"learning_rate": 1.8696581196581197e-05,
|
| 1511 |
+
"loss": 0.0661,
|
| 1512 |
+
"step": 2070
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 6.656,
|
| 1516 |
+
"grad_norm": 34.47880172729492,
|
| 1517 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 1518 |
+
"loss": 0.0594,
|
| 1519 |
+
"step": 2080
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 6.688,
|
| 1523 |
+
"grad_norm": 10.502516746520996,
|
| 1524 |
+
"learning_rate": 1.8340455840455843e-05,
|
| 1525 |
+
"loss": 0.0545,
|
| 1526 |
+
"step": 2090
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 6.72,
|
| 1530 |
+
"grad_norm": 24.073810577392578,
|
| 1531 |
+
"learning_rate": 1.8162393162393162e-05,
|
| 1532 |
+
"loss": 0.0592,
|
| 1533 |
+
"step": 2100
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 6.752,
|
| 1537 |
+
"grad_norm": 25.363536834716797,
|
| 1538 |
+
"learning_rate": 1.7984330484330487e-05,
|
| 1539 |
+
"loss": 0.0767,
|
| 1540 |
+
"step": 2110
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 6.784,
|
| 1544 |
+
"grad_norm": 15.268250465393066,
|
| 1545 |
+
"learning_rate": 1.7806267806267805e-05,
|
| 1546 |
+
"loss": 0.069,
|
| 1547 |
+
"step": 2120
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 6.816,
|
| 1551 |
+
"grad_norm": 8.48227596282959,
|
| 1552 |
+
"learning_rate": 1.762820512820513e-05,
|
| 1553 |
+
"loss": 0.0625,
|
| 1554 |
+
"step": 2130
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 6.848,
|
| 1558 |
+
"grad_norm": 10.944808006286621,
|
| 1559 |
+
"learning_rate": 1.7450142450142452e-05,
|
| 1560 |
+
"loss": 0.0729,
|
| 1561 |
+
"step": 2140
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 6.88,
|
| 1565 |
+
"grad_norm": 12.689846992492676,
|
| 1566 |
+
"learning_rate": 1.727207977207977e-05,
|
| 1567 |
+
"loss": 0.0634,
|
| 1568 |
+
"step": 2150
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 6.912,
|
| 1572 |
+
"grad_norm": 6.600920677185059,
|
| 1573 |
+
"learning_rate": 1.7094017094017095e-05,
|
| 1574 |
+
"loss": 0.0758,
|
| 1575 |
+
"step": 2160
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 6.944,
|
| 1579 |
+
"grad_norm": 19.84024429321289,
|
| 1580 |
+
"learning_rate": 1.6915954415954417e-05,
|
| 1581 |
+
"loss": 0.0565,
|
| 1582 |
+
"step": 2170
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 6.976,
|
| 1586 |
+
"grad_norm": 8.632323265075684,
|
| 1587 |
+
"learning_rate": 1.673789173789174e-05,
|
| 1588 |
+
"loss": 0.0671,
|
| 1589 |
+
"step": 2180
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 6.9984,
|
| 1593 |
+
"eval_accuracy": 0.8892018779342723,
|
| 1594 |
+
"eval_loss": 0.3630879521369934,
|
| 1595 |
+
"eval_runtime": 14.8354,
|
| 1596 |
+
"eval_samples_per_second": 71.788,
|
| 1597 |
+
"eval_steps_per_second": 2.292,
|
| 1598 |
+
"step": 2187
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.008,
|
| 1602 |
+
"grad_norm": 7.642409324645996,
|
| 1603 |
+
"learning_rate": 1.655982905982906e-05,
|
| 1604 |
+
"loss": 0.0489,
|
| 1605 |
+
"step": 2190
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.04,
|
| 1609 |
+
"grad_norm": 8.677549362182617,
|
| 1610 |
+
"learning_rate": 1.6381766381766382e-05,
|
| 1611 |
+
"loss": 0.0596,
|
| 1612 |
+
"step": 2200
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 7.072,
|
| 1616 |
+
"grad_norm": 20.24688148498535,
|
| 1617 |
+
"learning_rate": 1.6203703703703704e-05,
|
| 1618 |
+
"loss": 0.0546,
|
| 1619 |
+
"step": 2210
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.104,
|
| 1623 |
+
"grad_norm": 19.509143829345703,
|
| 1624 |
+
"learning_rate": 1.602564102564103e-05,
|
| 1625 |
+
"loss": 0.0581,
|
| 1626 |
+
"step": 2220
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.136,
|
| 1630 |
+
"grad_norm": 26.461017608642578,
|
| 1631 |
+
"learning_rate": 1.5847578347578347e-05,
|
| 1632 |
+
"loss": 0.0857,
|
| 1633 |
+
"step": 2230
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 7.168,
|
| 1637 |
+
"grad_norm": 14.442258834838867,
|
| 1638 |
+
"learning_rate": 1.566951566951567e-05,
|
| 1639 |
+
"loss": 0.0688,
|
| 1640 |
+
"step": 2240
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 7.2,
|
| 1644 |
+
"grad_norm": 8.91716194152832,
|
| 1645 |
+
"learning_rate": 1.5491452991452994e-05,
|
| 1646 |
+
"loss": 0.079,
|
| 1647 |
+
"step": 2250
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 7.232,
|
| 1651 |
+
"grad_norm": 15.394043922424316,
|
| 1652 |
+
"learning_rate": 1.5313390313390312e-05,
|
| 1653 |
+
"loss": 0.0561,
|
| 1654 |
+
"step": 2260
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 7.264,
|
| 1658 |
+
"grad_norm": 19.662343978881836,
|
| 1659 |
+
"learning_rate": 1.5135327635327638e-05,
|
| 1660 |
+
"loss": 0.0581,
|
| 1661 |
+
"step": 2270
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 7.296,
|
| 1665 |
+
"grad_norm": 8.76469898223877,
|
| 1666 |
+
"learning_rate": 1.4957264957264958e-05,
|
| 1667 |
+
"loss": 0.0439,
|
| 1668 |
+
"step": 2280
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 7.328,
|
| 1672 |
+
"grad_norm": 10.02730941772461,
|
| 1673 |
+
"learning_rate": 1.4779202279202281e-05,
|
| 1674 |
+
"loss": 0.0547,
|
| 1675 |
+
"step": 2290
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 7.36,
|
| 1679 |
+
"grad_norm": 21.142120361328125,
|
| 1680 |
+
"learning_rate": 1.4601139601139601e-05,
|
| 1681 |
+
"loss": 0.0618,
|
| 1682 |
+
"step": 2300
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 7.392,
|
| 1686 |
+
"grad_norm": 8.849529266357422,
|
| 1687 |
+
"learning_rate": 1.4423076923076923e-05,
|
| 1688 |
+
"loss": 0.0601,
|
| 1689 |
+
"step": 2310
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 7.424,
|
| 1693 |
+
"grad_norm": 13.313969612121582,
|
| 1694 |
+
"learning_rate": 1.4245014245014246e-05,
|
| 1695 |
+
"loss": 0.0578,
|
| 1696 |
+
"step": 2320
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 7.456,
|
| 1700 |
+
"grad_norm": 17.729047775268555,
|
| 1701 |
+
"learning_rate": 1.4066951566951566e-05,
|
| 1702 |
+
"loss": 0.0593,
|
| 1703 |
+
"step": 2330
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 7.4879999999999995,
|
| 1707 |
+
"grad_norm": 13.839546203613281,
|
| 1708 |
+
"learning_rate": 1.388888888888889e-05,
|
| 1709 |
+
"loss": 0.0566,
|
| 1710 |
+
"step": 2340
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 7.52,
|
| 1714 |
+
"grad_norm": 6.299365997314453,
|
| 1715 |
+
"learning_rate": 1.3710826210826211e-05,
|
| 1716 |
+
"loss": 0.0469,
|
| 1717 |
+
"step": 2350
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 7.552,
|
| 1721 |
+
"grad_norm": 12.264620780944824,
|
| 1722 |
+
"learning_rate": 1.3532763532763535e-05,
|
| 1723 |
+
"loss": 0.0476,
|
| 1724 |
+
"step": 2360
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 7.584,
|
| 1728 |
+
"grad_norm": 13.175020217895508,
|
| 1729 |
+
"learning_rate": 1.3354700854700855e-05,
|
| 1730 |
+
"loss": 0.0483,
|
| 1731 |
+
"step": 2370
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 7.616,
|
| 1735 |
+
"grad_norm": 8.695143699645996,
|
| 1736 |
+
"learning_rate": 1.3176638176638178e-05,
|
| 1737 |
+
"loss": 0.0676,
|
| 1738 |
+
"step": 2380
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 7.648,
|
| 1742 |
+
"grad_norm": 7.434255599975586,
|
| 1743 |
+
"learning_rate": 1.29985754985755e-05,
|
| 1744 |
+
"loss": 0.0514,
|
| 1745 |
+
"step": 2390
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 7.68,
|
| 1749 |
+
"grad_norm": 11.564408302307129,
|
| 1750 |
+
"learning_rate": 1.282051282051282e-05,
|
| 1751 |
+
"loss": 0.0496,
|
| 1752 |
+
"step": 2400
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 7.712,
|
| 1756 |
+
"grad_norm": 15.317635536193848,
|
| 1757 |
+
"learning_rate": 1.2642450142450143e-05,
|
| 1758 |
+
"loss": 0.0723,
|
| 1759 |
+
"step": 2410
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 7.744,
|
| 1763 |
+
"grad_norm": 5.177554130554199,
|
| 1764 |
+
"learning_rate": 1.2464387464387465e-05,
|
| 1765 |
+
"loss": 0.0508,
|
| 1766 |
+
"step": 2420
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 7.776,
|
| 1770 |
+
"grad_norm": 21.86553382873535,
|
| 1771 |
+
"learning_rate": 1.2286324786324787e-05,
|
| 1772 |
+
"loss": 0.0531,
|
| 1773 |
+
"step": 2430
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 7.808,
|
| 1777 |
+
"grad_norm": 5.066462516784668,
|
| 1778 |
+
"learning_rate": 1.2108262108262108e-05,
|
| 1779 |
+
"loss": 0.0601,
|
| 1780 |
+
"step": 2440
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 7.84,
|
| 1784 |
+
"grad_norm": 3.860797643661499,
|
| 1785 |
+
"learning_rate": 1.193019943019943e-05,
|
| 1786 |
+
"loss": 0.0529,
|
| 1787 |
+
"step": 2450
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 7.872,
|
| 1791 |
+
"grad_norm": 27.842437744140625,
|
| 1792 |
+
"learning_rate": 1.1752136752136752e-05,
|
| 1793 |
+
"loss": 0.0736,
|
| 1794 |
+
"step": 2460
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 7.904,
|
| 1798 |
+
"grad_norm": 26.736413955688477,
|
| 1799 |
+
"learning_rate": 1.1574074074074075e-05,
|
| 1800 |
+
"loss": 0.0589,
|
| 1801 |
+
"step": 2470
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 7.936,
|
| 1805 |
+
"grad_norm": 25.008304595947266,
|
| 1806 |
+
"learning_rate": 1.1396011396011397e-05,
|
| 1807 |
+
"loss": 0.0549,
|
| 1808 |
+
"step": 2480
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 7.968,
|
| 1812 |
+
"grad_norm": 4.6437506675720215,
|
| 1813 |
+
"learning_rate": 1.1217948717948719e-05,
|
| 1814 |
+
"loss": 0.0533,
|
| 1815 |
+
"step": 2490
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 8.0,
|
| 1819 |
+
"grad_norm": 26.05324935913086,
|
| 1820 |
+
"learning_rate": 1.103988603988604e-05,
|
| 1821 |
+
"loss": 0.0457,
|
| 1822 |
+
"step": 2500
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 8.0,
|
| 1826 |
+
"eval_accuracy": 0.8338028169014085,
|
| 1827 |
+
"eval_loss": 0.7270519137382507,
|
| 1828 |
+
"eval_runtime": 16.0951,
|
| 1829 |
+
"eval_samples_per_second": 66.169,
|
| 1830 |
+
"eval_steps_per_second": 2.112,
|
| 1831 |
+
"step": 2500
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 8.032,
|
| 1835 |
+
"grad_norm": 6.876763820648193,
|
| 1836 |
+
"learning_rate": 1.0861823361823362e-05,
|
| 1837 |
+
"loss": 0.0392,
|
| 1838 |
+
"step": 2510
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 8.064,
|
| 1842 |
+
"grad_norm": 9.069485664367676,
|
| 1843 |
+
"learning_rate": 1.0683760683760684e-05,
|
| 1844 |
+
"loss": 0.059,
|
| 1845 |
+
"step": 2520
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 8.096,
|
| 1849 |
+
"grad_norm": 14.368523597717285,
|
| 1850 |
+
"learning_rate": 1.0505698005698005e-05,
|
| 1851 |
+
"loss": 0.0455,
|
| 1852 |
+
"step": 2530
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 8.128,
|
| 1856 |
+
"grad_norm": 10.847954750061035,
|
| 1857 |
+
"learning_rate": 1.0327635327635327e-05,
|
| 1858 |
+
"loss": 0.0368,
|
| 1859 |
+
"step": 2540
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 8.16,
|
| 1863 |
+
"grad_norm": 7.805965900421143,
|
| 1864 |
+
"learning_rate": 1.014957264957265e-05,
|
| 1865 |
+
"loss": 0.0374,
|
| 1866 |
+
"step": 2550
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 8.192,
|
| 1870 |
+
"grad_norm": 37.152957916259766,
|
| 1871 |
+
"learning_rate": 9.971509971509972e-06,
|
| 1872 |
+
"loss": 0.0434,
|
| 1873 |
+
"step": 2560
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 8.224,
|
| 1877 |
+
"grad_norm": 32.26515197753906,
|
| 1878 |
+
"learning_rate": 9.793447293447294e-06,
|
| 1879 |
+
"loss": 0.0538,
|
| 1880 |
+
"step": 2570
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 8.256,
|
| 1884 |
+
"grad_norm": 9.871603012084961,
|
| 1885 |
+
"learning_rate": 9.615384615384616e-06,
|
| 1886 |
+
"loss": 0.0443,
|
| 1887 |
+
"step": 2580
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 8.288,
|
| 1891 |
+
"grad_norm": 18.939363479614258,
|
| 1892 |
+
"learning_rate": 9.437321937321937e-06,
|
| 1893 |
+
"loss": 0.0444,
|
| 1894 |
+
"step": 2590
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 8.32,
|
| 1898 |
+
"grad_norm": 8.318560600280762,
|
| 1899 |
+
"learning_rate": 9.259259259259259e-06,
|
| 1900 |
+
"loss": 0.0367,
|
| 1901 |
+
"step": 2600
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 8.352,
|
| 1905 |
+
"grad_norm": 10.871614456176758,
|
| 1906 |
+
"learning_rate": 9.081196581196581e-06,
|
| 1907 |
+
"loss": 0.0493,
|
| 1908 |
+
"step": 2610
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 8.384,
|
| 1912 |
+
"grad_norm": 5.46032190322876,
|
| 1913 |
+
"learning_rate": 8.903133903133903e-06,
|
| 1914 |
+
"loss": 0.0413,
|
| 1915 |
+
"step": 2620
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 8.416,
|
| 1919 |
+
"grad_norm": 1.5709623098373413,
|
| 1920 |
+
"learning_rate": 8.725071225071226e-06,
|
| 1921 |
+
"loss": 0.0658,
|
| 1922 |
+
"step": 2630
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 8.448,
|
| 1926 |
+
"grad_norm": 26.417600631713867,
|
| 1927 |
+
"learning_rate": 8.547008547008548e-06,
|
| 1928 |
+
"loss": 0.0681,
|
| 1929 |
+
"step": 2640
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 8.48,
|
| 1933 |
+
"grad_norm": 9.628387451171875,
|
| 1934 |
+
"learning_rate": 8.36894586894587e-06,
|
| 1935 |
+
"loss": 0.0399,
|
| 1936 |
+
"step": 2650
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 8.512,
|
| 1940 |
+
"grad_norm": 2.8144702911376953,
|
| 1941 |
+
"learning_rate": 8.190883190883191e-06,
|
| 1942 |
+
"loss": 0.0435,
|
| 1943 |
+
"step": 2660
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 8.544,
|
| 1947 |
+
"grad_norm": 13.873322486877441,
|
| 1948 |
+
"learning_rate": 8.012820512820515e-06,
|
| 1949 |
+
"loss": 0.0501,
|
| 1950 |
+
"step": 2670
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 8.576,
|
| 1954 |
+
"grad_norm": 4.96248722076416,
|
| 1955 |
+
"learning_rate": 7.834757834757835e-06,
|
| 1956 |
+
"loss": 0.0423,
|
| 1957 |
+
"step": 2680
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 8.608,
|
| 1961 |
+
"grad_norm": 12.236059188842773,
|
| 1962 |
+
"learning_rate": 7.656695156695156e-06,
|
| 1963 |
+
"loss": 0.0488,
|
| 1964 |
+
"step": 2690
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 8.64,
|
| 1968 |
+
"grad_norm": 6.332564353942871,
|
| 1969 |
+
"learning_rate": 7.478632478632479e-06,
|
| 1970 |
+
"loss": 0.0437,
|
| 1971 |
+
"step": 2700
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 8.672,
|
| 1975 |
+
"grad_norm": 7.448380470275879,
|
| 1976 |
+
"learning_rate": 7.3005698005698005e-06,
|
| 1977 |
+
"loss": 0.035,
|
| 1978 |
+
"step": 2710
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 8.704,
|
| 1982 |
+
"grad_norm": 7.228295803070068,
|
| 1983 |
+
"learning_rate": 7.122507122507123e-06,
|
| 1984 |
+
"loss": 0.0457,
|
| 1985 |
+
"step": 2720
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 8.736,
|
| 1989 |
+
"grad_norm": 5.306349277496338,
|
| 1990 |
+
"learning_rate": 6.944444444444445e-06,
|
| 1991 |
+
"loss": 0.0496,
|
| 1992 |
+
"step": 2730
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 8.768,
|
| 1996 |
+
"grad_norm": 15.959365844726562,
|
| 1997 |
+
"learning_rate": 6.766381766381767e-06,
|
| 1998 |
+
"loss": 0.04,
|
| 1999 |
+
"step": 2740
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 8.8,
|
| 2003 |
+
"grad_norm": 13.594867706298828,
|
| 2004 |
+
"learning_rate": 6.588319088319089e-06,
|
| 2005 |
+
"loss": 0.0466,
|
| 2006 |
+
"step": 2750
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 8.832,
|
| 2010 |
+
"grad_norm": 7.143582820892334,
|
| 2011 |
+
"learning_rate": 6.41025641025641e-06,
|
| 2012 |
+
"loss": 0.0367,
|
| 2013 |
+
"step": 2760
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 8.864,
|
| 2017 |
+
"grad_norm": 10.158649444580078,
|
| 2018 |
+
"learning_rate": 6.2321937321937325e-06,
|
| 2019 |
+
"loss": 0.0442,
|
| 2020 |
+
"step": 2770
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 8.896,
|
| 2024 |
+
"grad_norm": 4.369603633880615,
|
| 2025 |
+
"learning_rate": 6.054131054131054e-06,
|
| 2026 |
+
"loss": 0.0352,
|
| 2027 |
+
"step": 2780
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 8.928,
|
| 2031 |
+
"grad_norm": 11.626177787780762,
|
| 2032 |
+
"learning_rate": 5.876068376068376e-06,
|
| 2033 |
+
"loss": 0.0414,
|
| 2034 |
+
"step": 2790
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 8.96,
|
| 2038 |
+
"grad_norm": 8.576985359191895,
|
| 2039 |
+
"learning_rate": 5.6980056980056985e-06,
|
| 2040 |
+
"loss": 0.0445,
|
| 2041 |
+
"step": 2800
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 8.992,
|
| 2045 |
+
"grad_norm": 11.782782554626465,
|
| 2046 |
+
"learning_rate": 5.51994301994302e-06,
|
| 2047 |
+
"loss": 0.0396,
|
| 2048 |
+
"step": 2810
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 8.9984,
|
| 2052 |
+
"eval_accuracy": 0.8826291079812206,
|
| 2053 |
+
"eval_loss": 0.46553394198417664,
|
| 2054 |
+
"eval_runtime": 16.3312,
|
| 2055 |
+
"eval_samples_per_second": 65.213,
|
| 2056 |
+
"eval_steps_per_second": 2.082,
|
| 2057 |
+
"step": 2812
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 9.024,
|
| 2061 |
+
"grad_norm": 9.517723083496094,
|
| 2062 |
+
"learning_rate": 5.341880341880342e-06,
|
| 2063 |
+
"loss": 0.0281,
|
| 2064 |
+
"step": 2820
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 9.056,
|
| 2068 |
+
"grad_norm": 11.273005485534668,
|
| 2069 |
+
"learning_rate": 5.163817663817664e-06,
|
| 2070 |
+
"loss": 0.0395,
|
| 2071 |
+
"step": 2830
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 9.088,
|
| 2075 |
+
"grad_norm": 12.864770889282227,
|
| 2076 |
+
"learning_rate": 4.985754985754986e-06,
|
| 2077 |
+
"loss": 0.0399,
|
| 2078 |
+
"step": 2840
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 9.12,
|
| 2082 |
+
"grad_norm": 11.933670997619629,
|
| 2083 |
+
"learning_rate": 4.807692307692308e-06,
|
| 2084 |
+
"loss": 0.0381,
|
| 2085 |
+
"step": 2850
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 9.152,
|
| 2089 |
+
"grad_norm": 21.274028778076172,
|
| 2090 |
+
"learning_rate": 4.6296296296296296e-06,
|
| 2091 |
+
"loss": 0.03,
|
| 2092 |
+
"step": 2860
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 9.184,
|
| 2096 |
+
"grad_norm": 6.597903251647949,
|
| 2097 |
+
"learning_rate": 4.451566951566951e-06,
|
| 2098 |
+
"loss": 0.042,
|
| 2099 |
+
"step": 2870
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 9.216,
|
| 2103 |
+
"grad_norm": 1.0557407140731812,
|
| 2104 |
+
"learning_rate": 4.273504273504274e-06,
|
| 2105 |
+
"loss": 0.0268,
|
| 2106 |
+
"step": 2880
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 9.248,
|
| 2110 |
+
"grad_norm": 11.865984916687012,
|
| 2111 |
+
"learning_rate": 4.0954415954415956e-06,
|
| 2112 |
+
"loss": 0.0301,
|
| 2113 |
+
"step": 2890
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 9.28,
|
| 2117 |
+
"grad_norm": 25.20126724243164,
|
| 2118 |
+
"learning_rate": 3.917378917378917e-06,
|
| 2119 |
+
"loss": 0.0383,
|
| 2120 |
+
"step": 2900
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 9.312,
|
| 2124 |
+
"grad_norm": 11.134941101074219,
|
| 2125 |
+
"learning_rate": 3.7393162393162394e-06,
|
| 2126 |
+
"loss": 0.0375,
|
| 2127 |
+
"step": 2910
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 9.344,
|
| 2131 |
+
"grad_norm": 2.754523992538452,
|
| 2132 |
+
"learning_rate": 3.5612535612535615e-06,
|
| 2133 |
+
"loss": 0.043,
|
| 2134 |
+
"step": 2920
|
| 2135 |
+
},
|
| 2136 |
+
{
|
| 2137 |
+
"epoch": 9.376,
|
| 2138 |
+
"grad_norm": 12.286087989807129,
|
| 2139 |
+
"learning_rate": 3.3831908831908837e-06,
|
| 2140 |
+
"loss": 0.035,
|
| 2141 |
+
"step": 2930
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"epoch": 9.408,
|
| 2145 |
+
"grad_norm": 3.8736507892608643,
|
| 2146 |
+
"learning_rate": 3.205128205128205e-06,
|
| 2147 |
+
"loss": 0.0391,
|
| 2148 |
+
"step": 2940
|
| 2149 |
+
},
|
| 2150 |
+
{
|
| 2151 |
+
"epoch": 9.44,
|
| 2152 |
+
"grad_norm": 22.06711769104004,
|
| 2153 |
+
"learning_rate": 3.027065527065527e-06,
|
| 2154 |
+
"loss": 0.0377,
|
| 2155 |
+
"step": 2950
|
| 2156 |
+
},
|
| 2157 |
+
{
|
| 2158 |
+
"epoch": 9.472,
|
| 2159 |
+
"grad_norm": 8.435704231262207,
|
| 2160 |
+
"learning_rate": 2.8490028490028492e-06,
|
| 2161 |
+
"loss": 0.0301,
|
| 2162 |
+
"step": 2960
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 9.504,
|
| 2166 |
+
"grad_norm": 27.129562377929688,
|
| 2167 |
+
"learning_rate": 2.670940170940171e-06,
|
| 2168 |
+
"loss": 0.039,
|
| 2169 |
+
"step": 2970
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 9.536,
|
| 2173 |
+
"grad_norm": 4.00566291809082,
|
| 2174 |
+
"learning_rate": 2.492877492877493e-06,
|
| 2175 |
+
"loss": 0.0261,
|
| 2176 |
+
"step": 2980
|
| 2177 |
+
},
|
| 2178 |
+
{
|
| 2179 |
+
"epoch": 9.568,
|
| 2180 |
+
"grad_norm": 18.572261810302734,
|
| 2181 |
+
"learning_rate": 2.3148148148148148e-06,
|
| 2182 |
+
"loss": 0.036,
|
| 2183 |
+
"step": 2990
|
| 2184 |
+
},
|
| 2185 |
+
{
|
| 2186 |
+
"epoch": 9.6,
|
| 2187 |
+
"grad_norm": 6.907130241394043,
|
| 2188 |
+
"learning_rate": 2.136752136752137e-06,
|
| 2189 |
+
"loss": 0.0455,
|
| 2190 |
+
"step": 3000
|
| 2191 |
+
},
|
| 2192 |
+
{
|
| 2193 |
+
"epoch": 9.632,
|
| 2194 |
+
"grad_norm": 3.8767590522766113,
|
| 2195 |
+
"learning_rate": 1.9586894586894586e-06,
|
| 2196 |
+
"loss": 0.0456,
|
| 2197 |
+
"step": 3010
|
| 2198 |
+
},
|
| 2199 |
+
{
|
| 2200 |
+
"epoch": 9.664,
|
| 2201 |
+
"grad_norm": 12.119085311889648,
|
| 2202 |
+
"learning_rate": 1.7806267806267808e-06,
|
| 2203 |
+
"loss": 0.037,
|
| 2204 |
+
"step": 3020
|
| 2205 |
+
},
|
| 2206 |
+
{
|
| 2207 |
+
"epoch": 9.696,
|
| 2208 |
+
"grad_norm": 7.650444507598877,
|
| 2209 |
+
"learning_rate": 1.6025641025641025e-06,
|
| 2210 |
+
"loss": 0.0363,
|
| 2211 |
+
"step": 3030
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 9.728,
|
| 2215 |
+
"grad_norm": 8.467825889587402,
|
| 2216 |
+
"learning_rate": 1.4245014245014246e-06,
|
| 2217 |
+
"loss": 0.0198,
|
| 2218 |
+
"step": 3040
|
| 2219 |
+
},
|
| 2220 |
+
{
|
| 2221 |
+
"epoch": 9.76,
|
| 2222 |
+
"grad_norm": 19.763225555419922,
|
| 2223 |
+
"learning_rate": 1.2464387464387465e-06,
|
| 2224 |
+
"loss": 0.0408,
|
| 2225 |
+
"step": 3050
|
| 2226 |
+
},
|
| 2227 |
+
{
|
| 2228 |
+
"epoch": 9.792,
|
| 2229 |
+
"grad_norm": 25.160770416259766,
|
| 2230 |
+
"learning_rate": 1.0683760683760685e-06,
|
| 2231 |
+
"loss": 0.0392,
|
| 2232 |
+
"step": 3060
|
| 2233 |
+
},
|
| 2234 |
+
{
|
| 2235 |
+
"epoch": 9.824,
|
| 2236 |
+
"grad_norm": 7.482017517089844,
|
| 2237 |
+
"learning_rate": 8.903133903133904e-07,
|
| 2238 |
+
"loss": 0.0283,
|
| 2239 |
+
"step": 3070
|
| 2240 |
+
},
|
| 2241 |
+
{
|
| 2242 |
+
"epoch": 9.856,
|
| 2243 |
+
"grad_norm": 7.559390544891357,
|
| 2244 |
+
"learning_rate": 7.122507122507123e-07,
|
| 2245 |
+
"loss": 0.0463,
|
| 2246 |
+
"step": 3080
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 9.888,
|
| 2250 |
+
"grad_norm": 10.798991203308105,
|
| 2251 |
+
"learning_rate": 5.341880341880342e-07,
|
| 2252 |
+
"loss": 0.0298,
|
| 2253 |
+
"step": 3090
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 9.92,
|
| 2257 |
+
"grad_norm": 3.9386239051818848,
|
| 2258 |
+
"learning_rate": 3.5612535612535615e-07,
|
| 2259 |
+
"loss": 0.0386,
|
| 2260 |
+
"step": 3100
|
| 2261 |
+
},
|
| 2262 |
+
{
|
| 2263 |
+
"epoch": 9.952,
|
| 2264 |
+
"grad_norm": 15.766891479492188,
|
| 2265 |
+
"learning_rate": 1.7806267806267808e-07,
|
| 2266 |
+
"loss": 0.0277,
|
| 2267 |
+
"step": 3110
|
| 2268 |
+
},
|
| 2269 |
+
{
|
| 2270 |
+
"epoch": 9.984,
|
| 2271 |
+
"grad_norm": 3.3344645500183105,
|
| 2272 |
+
"learning_rate": 0.0,
|
| 2273 |
+
"loss": 0.0293,
|
| 2274 |
+
"step": 3120
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 9.984,
|
| 2278 |
+
"eval_accuracy": 0.8610328638497653,
|
| 2279 |
+
"eval_loss": 0.7580540180206299,
|
| 2280 |
+
"eval_runtime": 16.5093,
|
| 2281 |
+
"eval_samples_per_second": 64.509,
|
| 2282 |
+
"eval_steps_per_second": 2.059,
|
| 2283 |
+
"step": 3120
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 9.984,
|
| 2287 |
+
"step": 3120,
|
| 2288 |
+
"total_flos": 9.926487761391452e+18,
|
| 2289 |
+
"train_loss": 0.15716634974934351,
|
| 2290 |
+
"train_runtime": 142469.0271,
|
| 2291 |
+
"train_samples_per_second": 2.808,
|
| 2292 |
+
"train_steps_per_second": 0.022
|
| 2293 |
+
}
|
| 2294 |
+
],
|
| 2295 |
+
"logging_steps": 10,
|
| 2296 |
+
"max_steps": 3120,
|
| 2297 |
+
"num_input_tokens_seen": 0,
|
| 2298 |
+
"num_train_epochs": 10,
|
| 2299 |
+
"save_steps": 500,
|
| 2300 |
+
"stateful_callbacks": {
|
| 2301 |
+
"TrainerControl": {
|
| 2302 |
+
"args": {
|
| 2303 |
+
"should_epoch_stop": false,
|
| 2304 |
+
"should_evaluate": false,
|
| 2305 |
+
"should_log": false,
|
| 2306 |
+
"should_save": true,
|
| 2307 |
+
"should_training_stop": true
|
| 2308 |
+
},
|
| 2309 |
+
"attributes": {}
|
| 2310 |
+
}
|
| 2311 |
+
},
|
| 2312 |
+
"total_flos": 9.926487761391452e+18,
|
| 2313 |
+
"train_batch_size": 32,
|
| 2314 |
+
"trial_name": null,
|
| 2315 |
+
"trial_params": null
|
| 2316 |
+
}
|