Model save
Browse files- README.md +2 -2
- all_results.json +8 -0
- test_results.json +8 -0
- trainer_state.json +3625 -0
README.md
CHANGED
|
@@ -15,8 +15,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model was trained from scratch on an unknown dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss: 0.
|
| 19 |
-
- Accuracy: 0
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
|
|
| 15 |
|
| 16 |
This model was trained from scratch on an unknown dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 0.0001
|
| 19 |
+
- Accuracy: 1.0
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 49.00898876404494,
|
| 3 |
+
"eval_accuracy": 1.0,
|
| 4 |
+
"eval_loss": 0.00010065778042189777,
|
| 5 |
+
"eval_runtime": 340.718,
|
| 6 |
+
"eval_samples_per_second": 2.421,
|
| 7 |
+
"eval_steps_per_second": 0.153
|
| 8 |
+
}
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 49.00898876404494,
|
| 3 |
+
"eval_accuracy": 1.0,
|
| 4 |
+
"eval_loss": 0.00010065778042189777,
|
| 5 |
+
"eval_runtime": 340.718,
|
| 6 |
+
"eval_samples_per_second": 2.421,
|
| 7 |
+
"eval_steps_per_second": 0.153
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3625 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8771021992238034,
|
| 3 |
+
"best_model_checkpoint": "trained_FM_plus_minus_finetuned-finetuned/checkpoint-2160",
|
| 4 |
+
"epoch": 49.00898876404494,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4450,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0022471910112359553,
|
| 13 |
+
"grad_norm": 22.698049545288086,
|
| 14 |
+
"learning_rate": 1.1235955056179775e-06,
|
| 15 |
+
"loss": 1.6489,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0044943820224719105,
|
| 20 |
+
"grad_norm": 137.66839599609375,
|
| 21 |
+
"learning_rate": 2.247191011235955e-06,
|
| 22 |
+
"loss": 0.8257,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.006741573033707865,
|
| 27 |
+
"grad_norm": 0.09208225458860397,
|
| 28 |
+
"learning_rate": 3.3707865168539327e-06,
|
| 29 |
+
"loss": 0.5424,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.008988764044943821,
|
| 34 |
+
"grad_norm": 20.217649459838867,
|
| 35 |
+
"learning_rate": 4.49438202247191e-06,
|
| 36 |
+
"loss": 0.2567,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.011235955056179775,
|
| 41 |
+
"grad_norm": 6.477929592132568,
|
| 42 |
+
"learning_rate": 5.617977528089888e-06,
|
| 43 |
+
"loss": 0.1167,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.01348314606741573,
|
| 48 |
+
"grad_norm": 1.9699681997299194,
|
| 49 |
+
"learning_rate": 6.741573033707865e-06,
|
| 50 |
+
"loss": 0.1324,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.015730337078651686,
|
| 55 |
+
"grad_norm": 0.1321302354335785,
|
| 56 |
+
"learning_rate": 7.865168539325843e-06,
|
| 57 |
+
"loss": 0.0394,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.017977528089887642,
|
| 62 |
+
"grad_norm": 0.16645067930221558,
|
| 63 |
+
"learning_rate": 8.98876404494382e-06,
|
| 64 |
+
"loss": 0.0398,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.020224719101123594,
|
| 69 |
+
"grad_norm": 0.01523497886955738,
|
| 70 |
+
"learning_rate": 1.0112359550561798e-05,
|
| 71 |
+
"loss": 0.075,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.020224719101123594,
|
| 76 |
+
"eval_accuracy": 0.6364812419146184,
|
| 77 |
+
"eval_loss": 1.6105461120605469,
|
| 78 |
+
"eval_runtime": 3297.1136,
|
| 79 |
+
"eval_samples_per_second": 0.234,
|
| 80 |
+
"eval_steps_per_second": 0.015,
|
| 81 |
+
"step": 90
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.002247191011236,
|
| 85 |
+
"grad_norm": 1.9008361101150513,
|
| 86 |
+
"learning_rate": 1.1235955056179776e-05,
|
| 87 |
+
"loss": 0.0981,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.0044943820224719,
|
| 92 |
+
"grad_norm": 202.8159637451172,
|
| 93 |
+
"learning_rate": 1.2359550561797752e-05,
|
| 94 |
+
"loss": 0.0302,
|
| 95 |
+
"step": 110
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.006741573033708,
|
| 99 |
+
"grad_norm": 0.2720589339733124,
|
| 100 |
+
"learning_rate": 1.348314606741573e-05,
|
| 101 |
+
"loss": 0.0412,
|
| 102 |
+
"step": 120
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.0089887640449438,
|
| 106 |
+
"grad_norm": 0.04250582307577133,
|
| 107 |
+
"learning_rate": 1.4606741573033709e-05,
|
| 108 |
+
"loss": 0.0024,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.0112359550561798,
|
| 113 |
+
"grad_norm": 0.01215441431850195,
|
| 114 |
+
"learning_rate": 1.5730337078651687e-05,
|
| 115 |
+
"loss": 0.0763,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.0134831460674156,
|
| 120 |
+
"grad_norm": 27.981979370117188,
|
| 121 |
+
"learning_rate": 1.6853932584269665e-05,
|
| 122 |
+
"loss": 0.0823,
|
| 123 |
+
"step": 150
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.0157303370786517,
|
| 127 |
+
"grad_norm": 0.05339406058192253,
|
| 128 |
+
"learning_rate": 1.797752808988764e-05,
|
| 129 |
+
"loss": 0.0024,
|
| 130 |
+
"step": 160
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.0179775280898877,
|
| 134 |
+
"grad_norm": 0.10670967400074005,
|
| 135 |
+
"learning_rate": 1.9101123595505618e-05,
|
| 136 |
+
"loss": 0.0404,
|
| 137 |
+
"step": 170
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.0202247191011236,
|
| 141 |
+
"grad_norm": 0.025777198374271393,
|
| 142 |
+
"learning_rate": 2.0224719101123596e-05,
|
| 143 |
+
"loss": 0.0019,
|
| 144 |
+
"step": 180
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.0202247191011236,
|
| 148 |
+
"eval_accuracy": 0.6157826649417852,
|
| 149 |
+
"eval_loss": 2.255258083343506,
|
| 150 |
+
"eval_runtime": 502.2881,
|
| 151 |
+
"eval_samples_per_second": 1.539,
|
| 152 |
+
"eval_steps_per_second": 0.098,
|
| 153 |
+
"step": 180
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.002247191011236,
|
| 157 |
+
"grad_norm": 0.03325297310948372,
|
| 158 |
+
"learning_rate": 2.1348314606741574e-05,
|
| 159 |
+
"loss": 0.0011,
|
| 160 |
+
"step": 190
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.004494382022472,
|
| 164 |
+
"grad_norm": 0.018461693078279495,
|
| 165 |
+
"learning_rate": 2.2471910112359552e-05,
|
| 166 |
+
"loss": 0.0008,
|
| 167 |
+
"step": 200
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.006741573033708,
|
| 171 |
+
"grad_norm": 0.02512010931968689,
|
| 172 |
+
"learning_rate": 2.359550561797753e-05,
|
| 173 |
+
"loss": 0.0824,
|
| 174 |
+
"step": 210
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.0089887640449438,
|
| 178 |
+
"grad_norm": 0.06770236790180206,
|
| 179 |
+
"learning_rate": 2.4719101123595505e-05,
|
| 180 |
+
"loss": 0.0327,
|
| 181 |
+
"step": 220
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.0112359550561796,
|
| 185 |
+
"grad_norm": 0.0409482941031456,
|
| 186 |
+
"learning_rate": 2.5842696629213486e-05,
|
| 187 |
+
"loss": 0.0028,
|
| 188 |
+
"step": 230
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.013483146067416,
|
| 192 |
+
"grad_norm": 150.85275268554688,
|
| 193 |
+
"learning_rate": 2.696629213483146e-05,
|
| 194 |
+
"loss": 0.2355,
|
| 195 |
+
"step": 240
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.0157303370786517,
|
| 199 |
+
"grad_norm": 2.0705699920654297,
|
| 200 |
+
"learning_rate": 2.8089887640449443e-05,
|
| 201 |
+
"loss": 0.105,
|
| 202 |
+
"step": 250
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.0179775280898875,
|
| 206 |
+
"grad_norm": 0.41521573066711426,
|
| 207 |
+
"learning_rate": 2.9213483146067417e-05,
|
| 208 |
+
"loss": 0.2252,
|
| 209 |
+
"step": 260
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.020224719101124,
|
| 213 |
+
"grad_norm": 20.540042877197266,
|
| 214 |
+
"learning_rate": 3.0337078651685396e-05,
|
| 215 |
+
"loss": 0.1088,
|
| 216 |
+
"step": 270
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.020224719101124,
|
| 220 |
+
"eval_accuracy": 0.6144890038809832,
|
| 221 |
+
"eval_loss": 1.8407323360443115,
|
| 222 |
+
"eval_runtime": 375.6551,
|
| 223 |
+
"eval_samples_per_second": 2.058,
|
| 224 |
+
"eval_steps_per_second": 0.13,
|
| 225 |
+
"step": 270
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.002247191011236,
|
| 229 |
+
"grad_norm": 3.0002777576446533,
|
| 230 |
+
"learning_rate": 3.1460674157303374e-05,
|
| 231 |
+
"loss": 0.1083,
|
| 232 |
+
"step": 280
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.004494382022472,
|
| 236 |
+
"grad_norm": 0.09991969913244247,
|
| 237 |
+
"learning_rate": 3.258426966292135e-05,
|
| 238 |
+
"loss": 0.0214,
|
| 239 |
+
"step": 290
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.006741573033708,
|
| 243 |
+
"grad_norm": 0.0313391350209713,
|
| 244 |
+
"learning_rate": 3.370786516853933e-05,
|
| 245 |
+
"loss": 0.0536,
|
| 246 |
+
"step": 300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.0089887640449438,
|
| 250 |
+
"grad_norm": 1.4865280389785767,
|
| 251 |
+
"learning_rate": 3.483146067415731e-05,
|
| 252 |
+
"loss": 0.0764,
|
| 253 |
+
"step": 310
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.0112359550561796,
|
| 257 |
+
"grad_norm": 0.1041778028011322,
|
| 258 |
+
"learning_rate": 3.595505617977528e-05,
|
| 259 |
+
"loss": 0.0048,
|
| 260 |
+
"step": 320
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.013483146067416,
|
| 264 |
+
"grad_norm": 0.030931729823350906,
|
| 265 |
+
"learning_rate": 3.7078651685393264e-05,
|
| 266 |
+
"loss": 0.0029,
|
| 267 |
+
"step": 330
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 3.0157303370786517,
|
| 271 |
+
"grad_norm": 0.036244966089725494,
|
| 272 |
+
"learning_rate": 3.8202247191011236e-05,
|
| 273 |
+
"loss": 0.0355,
|
| 274 |
+
"step": 340
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 3.0179775280898875,
|
| 278 |
+
"grad_norm": 0.02242976799607277,
|
| 279 |
+
"learning_rate": 3.9325842696629214e-05,
|
| 280 |
+
"loss": 0.0017,
|
| 281 |
+
"step": 350
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 3.020224719101124,
|
| 285 |
+
"grad_norm": 0.018726468086242676,
|
| 286 |
+
"learning_rate": 4.044943820224719e-05,
|
| 287 |
+
"loss": 0.0012,
|
| 288 |
+
"step": 360
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 3.020224719101124,
|
| 292 |
+
"eval_accuracy": 0.777490297542044,
|
| 293 |
+
"eval_loss": 1.3938281536102295,
|
| 294 |
+
"eval_runtime": 358.0182,
|
| 295 |
+
"eval_samples_per_second": 2.159,
|
| 296 |
+
"eval_steps_per_second": 0.137,
|
| 297 |
+
"step": 360
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 4.002247191011236,
|
| 301 |
+
"grad_norm": 0.032676514238119125,
|
| 302 |
+
"learning_rate": 4.157303370786517e-05,
|
| 303 |
+
"loss": 0.0439,
|
| 304 |
+
"step": 370
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 4.004494382022472,
|
| 308 |
+
"grad_norm": 0.03491745889186859,
|
| 309 |
+
"learning_rate": 4.269662921348315e-05,
|
| 310 |
+
"loss": 0.0017,
|
| 311 |
+
"step": 380
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 4.0067415730337075,
|
| 315 |
+
"grad_norm": 0.021868839859962463,
|
| 316 |
+
"learning_rate": 4.3820224719101126e-05,
|
| 317 |
+
"loss": 0.0384,
|
| 318 |
+
"step": 390
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 4.008988764044944,
|
| 322 |
+
"grad_norm": 0.12130255252122879,
|
| 323 |
+
"learning_rate": 4.4943820224719104e-05,
|
| 324 |
+
"loss": 0.0342,
|
| 325 |
+
"step": 400
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 4.01123595505618,
|
| 329 |
+
"grad_norm": 0.060738105326890945,
|
| 330 |
+
"learning_rate": 4.606741573033708e-05,
|
| 331 |
+
"loss": 0.0045,
|
| 332 |
+
"step": 410
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 4.013483146067416,
|
| 336 |
+
"grad_norm": 0.02228277549147606,
|
| 337 |
+
"learning_rate": 4.719101123595506e-05,
|
| 338 |
+
"loss": 0.1898,
|
| 339 |
+
"step": 420
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 4.015730337078652,
|
| 343 |
+
"grad_norm": 5.419518947601318,
|
| 344 |
+
"learning_rate": 4.831460674157304e-05,
|
| 345 |
+
"loss": 0.3639,
|
| 346 |
+
"step": 430
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 4.0179775280898875,
|
| 350 |
+
"grad_norm": 107.74591827392578,
|
| 351 |
+
"learning_rate": 4.943820224719101e-05,
|
| 352 |
+
"loss": 0.1896,
|
| 353 |
+
"step": 440
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 4.020224719101123,
|
| 357 |
+
"grad_norm": 0.04789801314473152,
|
| 358 |
+
"learning_rate": 4.993757802746567e-05,
|
| 359 |
+
"loss": 0.0653,
|
| 360 |
+
"step": 450
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 4.020224719101123,
|
| 364 |
+
"eval_accuracy": 0.49805950840879687,
|
| 365 |
+
"eval_loss": 2.990910291671753,
|
| 366 |
+
"eval_runtime": 363.2571,
|
| 367 |
+
"eval_samples_per_second": 2.128,
|
| 368 |
+
"eval_steps_per_second": 0.135,
|
| 369 |
+
"step": 450
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 5.002247191011236,
|
| 373 |
+
"grad_norm": 0.021787848323583603,
|
| 374 |
+
"learning_rate": 4.9812734082397005e-05,
|
| 375 |
+
"loss": 0.0651,
|
| 376 |
+
"step": 460
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 5.004494382022472,
|
| 380 |
+
"grad_norm": 0.06374188512563705,
|
| 381 |
+
"learning_rate": 4.968789013732834e-05,
|
| 382 |
+
"loss": 0.0436,
|
| 383 |
+
"step": 470
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 5.0067415730337075,
|
| 387 |
+
"grad_norm": 0.046116895973682404,
|
| 388 |
+
"learning_rate": 4.956304619225968e-05,
|
| 389 |
+
"loss": 0.075,
|
| 390 |
+
"step": 480
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 5.008988764044944,
|
| 394 |
+
"grad_norm": 0.03146898001432419,
|
| 395 |
+
"learning_rate": 4.943820224719101e-05,
|
| 396 |
+
"loss": 0.1484,
|
| 397 |
+
"step": 490
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 5.01123595505618,
|
| 401 |
+
"grad_norm": 3.9672632217407227,
|
| 402 |
+
"learning_rate": 4.931335830212235e-05,
|
| 403 |
+
"loss": 0.5941,
|
| 404 |
+
"step": 500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 5.013483146067416,
|
| 408 |
+
"grad_norm": 0.04689766839146614,
|
| 409 |
+
"learning_rate": 4.9188514357053684e-05,
|
| 410 |
+
"loss": 0.0589,
|
| 411 |
+
"step": 510
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 5.015730337078652,
|
| 415 |
+
"grad_norm": 0.022010423243045807,
|
| 416 |
+
"learning_rate": 4.906367041198502e-05,
|
| 417 |
+
"loss": 0.0241,
|
| 418 |
+
"step": 520
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 5.0179775280898875,
|
| 422 |
+
"grad_norm": 0.10352443158626556,
|
| 423 |
+
"learning_rate": 4.893882646691635e-05,
|
| 424 |
+
"loss": 0.0418,
|
| 425 |
+
"step": 530
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 5.020224719101123,
|
| 429 |
+
"grad_norm": 0.12475921213626862,
|
| 430 |
+
"learning_rate": 4.8813982521847696e-05,
|
| 431 |
+
"loss": 0.0883,
|
| 432 |
+
"step": 540
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 5.020224719101123,
|
| 436 |
+
"eval_accuracy": 0.351875808538163,
|
| 437 |
+
"eval_loss": 4.658880710601807,
|
| 438 |
+
"eval_runtime": 363.3989,
|
| 439 |
+
"eval_samples_per_second": 2.127,
|
| 440 |
+
"eval_steps_per_second": 0.135,
|
| 441 |
+
"step": 540
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 6.002247191011236,
|
| 445 |
+
"grad_norm": 0.18845953047275543,
|
| 446 |
+
"learning_rate": 4.8689138576779034e-05,
|
| 447 |
+
"loss": 0.0521,
|
| 448 |
+
"step": 550
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 6.004494382022472,
|
| 452 |
+
"grad_norm": 0.06792347878217697,
|
| 453 |
+
"learning_rate": 4.8564294631710364e-05,
|
| 454 |
+
"loss": 0.0111,
|
| 455 |
+
"step": 560
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 6.0067415730337075,
|
| 459 |
+
"grad_norm": 0.07692062854766846,
|
| 460 |
+
"learning_rate": 4.84394506866417e-05,
|
| 461 |
+
"loss": 0.1613,
|
| 462 |
+
"step": 570
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 6.008988764044944,
|
| 466 |
+
"grad_norm": 0.12970906496047974,
|
| 467 |
+
"learning_rate": 4.831460674157304e-05,
|
| 468 |
+
"loss": 0.005,
|
| 469 |
+
"step": 580
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 6.01123595505618,
|
| 473 |
+
"grad_norm": 0.07438448816537857,
|
| 474 |
+
"learning_rate": 4.8189762796504376e-05,
|
| 475 |
+
"loss": 0.0346,
|
| 476 |
+
"step": 590
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 6.013483146067416,
|
| 480 |
+
"grad_norm": 0.037353940308094025,
|
| 481 |
+
"learning_rate": 4.8064918851435706e-05,
|
| 482 |
+
"loss": 0.0287,
|
| 483 |
+
"step": 600
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 6.015730337078652,
|
| 487 |
+
"grad_norm": 0.03469238057732582,
|
| 488 |
+
"learning_rate": 4.7940074906367044e-05,
|
| 489 |
+
"loss": 0.0321,
|
| 490 |
+
"step": 610
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 6.0179775280898875,
|
| 494 |
+
"grad_norm": 0.14126905798912048,
|
| 495 |
+
"learning_rate": 4.781523096129838e-05,
|
| 496 |
+
"loss": 0.0894,
|
| 497 |
+
"step": 620
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 6.020224719101123,
|
| 501 |
+
"grad_norm": 0.20986488461494446,
|
| 502 |
+
"learning_rate": 4.769038701622972e-05,
|
| 503 |
+
"loss": 0.0389,
|
| 504 |
+
"step": 630
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 6.020224719101123,
|
| 508 |
+
"eval_accuracy": 0.4579560155239327,
|
| 509 |
+
"eval_loss": 3.409106492996216,
|
| 510 |
+
"eval_runtime": 321.2302,
|
| 511 |
+
"eval_samples_per_second": 2.406,
|
| 512 |
+
"eval_steps_per_second": 0.153,
|
| 513 |
+
"step": 630
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 7.002247191011236,
|
| 517 |
+
"grad_norm": 0.05564555898308754,
|
| 518 |
+
"learning_rate": 4.756554307116105e-05,
|
| 519 |
+
"loss": 0.0455,
|
| 520 |
+
"step": 640
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 7.004494382022472,
|
| 524 |
+
"grad_norm": 0.07562582194805145,
|
| 525 |
+
"learning_rate": 4.7440699126092386e-05,
|
| 526 |
+
"loss": 0.005,
|
| 527 |
+
"step": 650
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 7.0067415730337075,
|
| 531 |
+
"grad_norm": 0.10327350348234177,
|
| 532 |
+
"learning_rate": 4.731585518102372e-05,
|
| 533 |
+
"loss": 0.0394,
|
| 534 |
+
"step": 660
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 7.008988764044944,
|
| 538 |
+
"grad_norm": 0.05257737264037132,
|
| 539 |
+
"learning_rate": 4.719101123595506e-05,
|
| 540 |
+
"loss": 0.0032,
|
| 541 |
+
"step": 670
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 7.01123595505618,
|
| 545 |
+
"grad_norm": 0.04399643838405609,
|
| 546 |
+
"learning_rate": 4.706616729088639e-05,
|
| 547 |
+
"loss": 0.0371,
|
| 548 |
+
"step": 680
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 7.013483146067416,
|
| 552 |
+
"grad_norm": 0.04171612858772278,
|
| 553 |
+
"learning_rate": 4.694132334581773e-05,
|
| 554 |
+
"loss": 0.0024,
|
| 555 |
+
"step": 690
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 7.015730337078652,
|
| 559 |
+
"grad_norm": 0.07221619784832001,
|
| 560 |
+
"learning_rate": 4.6816479400749066e-05,
|
| 561 |
+
"loss": 0.068,
|
| 562 |
+
"step": 700
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 7.0179775280898875,
|
| 566 |
+
"grad_norm": 0.08562880754470825,
|
| 567 |
+
"learning_rate": 4.6691635455680396e-05,
|
| 568 |
+
"loss": 0.0107,
|
| 569 |
+
"step": 710
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 7.020224719101123,
|
| 573 |
+
"grad_norm": 0.04621588811278343,
|
| 574 |
+
"learning_rate": 4.656679151061174e-05,
|
| 575 |
+
"loss": 0.0346,
|
| 576 |
+
"step": 720
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 7.020224719101123,
|
| 580 |
+
"eval_accuracy": 0.6222509702457956,
|
| 581 |
+
"eval_loss": 2.998917579650879,
|
| 582 |
+
"eval_runtime": 371.0049,
|
| 583 |
+
"eval_samples_per_second": 2.084,
|
| 584 |
+
"eval_steps_per_second": 0.132,
|
| 585 |
+
"step": 720
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 8.002247191011236,
|
| 589 |
+
"grad_norm": 0.14536446332931519,
|
| 590 |
+
"learning_rate": 4.644194756554308e-05,
|
| 591 |
+
"loss": 0.0404,
|
| 592 |
+
"step": 730
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 8.004494382022472,
|
| 596 |
+
"grad_norm": 0.06857080012559891,
|
| 597 |
+
"learning_rate": 4.6317103620474415e-05,
|
| 598 |
+
"loss": 0.0343,
|
| 599 |
+
"step": 740
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 8.006741573033707,
|
| 603 |
+
"grad_norm": 1.3145660161972046,
|
| 604 |
+
"learning_rate": 4.6192259675405745e-05,
|
| 605 |
+
"loss": 0.0458,
|
| 606 |
+
"step": 750
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 8.008988764044943,
|
| 610 |
+
"grad_norm": 4.400092124938965,
|
| 611 |
+
"learning_rate": 4.606741573033708e-05,
|
| 612 |
+
"loss": 0.1135,
|
| 613 |
+
"step": 760
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 8.01123595505618,
|
| 617 |
+
"grad_norm": 1.3112094402313232,
|
| 618 |
+
"learning_rate": 4.594257178526842e-05,
|
| 619 |
+
"loss": 0.0597,
|
| 620 |
+
"step": 770
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 8.013483146067415,
|
| 624 |
+
"grad_norm": 0.17381101846694946,
|
| 625 |
+
"learning_rate": 4.581772784019975e-05,
|
| 626 |
+
"loss": 0.0431,
|
| 627 |
+
"step": 780
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 8.015730337078653,
|
| 631 |
+
"grad_norm": 0.07657615840435028,
|
| 632 |
+
"learning_rate": 4.569288389513109e-05,
|
| 633 |
+
"loss": 0.0064,
|
| 634 |
+
"step": 790
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 8.017977528089888,
|
| 638 |
+
"grad_norm": 0.02938857115805149,
|
| 639 |
+
"learning_rate": 4.5568039950062425e-05,
|
| 640 |
+
"loss": 0.0031,
|
| 641 |
+
"step": 800
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 8.020224719101124,
|
| 645 |
+
"grad_norm": 0.03589854761958122,
|
| 646 |
+
"learning_rate": 4.544319600499376e-05,
|
| 647 |
+
"loss": 0.0014,
|
| 648 |
+
"step": 810
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 8.020224719101124,
|
| 652 |
+
"eval_accuracy": 0.3855109961190168,
|
| 653 |
+
"eval_loss": 4.314896583557129,
|
| 654 |
+
"eval_runtime": 379.8459,
|
| 655 |
+
"eval_samples_per_second": 2.035,
|
| 656 |
+
"eval_steps_per_second": 0.129,
|
| 657 |
+
"step": 810
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 9.002247191011236,
|
| 661 |
+
"grad_norm": 0.02095959149301052,
|
| 662 |
+
"learning_rate": 4.531835205992509e-05,
|
| 663 |
+
"loss": 0.0012,
|
| 664 |
+
"step": 820
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 9.004494382022472,
|
| 668 |
+
"grad_norm": 0.02048562467098236,
|
| 669 |
+
"learning_rate": 4.519350811485643e-05,
|
| 670 |
+
"loss": 0.0008,
|
| 671 |
+
"step": 830
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 9.006741573033707,
|
| 675 |
+
"grad_norm": 0.019248204305768013,
|
| 676 |
+
"learning_rate": 4.506866416978777e-05,
|
| 677 |
+
"loss": 0.0007,
|
| 678 |
+
"step": 840
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 9.008988764044943,
|
| 682 |
+
"grad_norm": 0.025466041639447212,
|
| 683 |
+
"learning_rate": 4.4943820224719104e-05,
|
| 684 |
+
"loss": 0.083,
|
| 685 |
+
"step": 850
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 9.01123595505618,
|
| 689 |
+
"grad_norm": 0.051293402910232544,
|
| 690 |
+
"learning_rate": 4.4818976279650435e-05,
|
| 691 |
+
"loss": 0.0357,
|
| 692 |
+
"step": 860
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 9.013483146067415,
|
| 696 |
+
"grad_norm": 0.05512285605072975,
|
| 697 |
+
"learning_rate": 4.469413233458177e-05,
|
| 698 |
+
"loss": 0.0645,
|
| 699 |
+
"step": 870
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 9.015730337078653,
|
| 703 |
+
"grad_norm": 0.26267877221107483,
|
| 704 |
+
"learning_rate": 4.456928838951311e-05,
|
| 705 |
+
"loss": 0.0821,
|
| 706 |
+
"step": 880
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 9.017977528089888,
|
| 710 |
+
"grad_norm": 0.14021214842796326,
|
| 711 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 712 |
+
"loss": 0.0082,
|
| 713 |
+
"step": 890
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 9.020224719101124,
|
| 717 |
+
"grad_norm": 0.08691973239183426,
|
| 718 |
+
"learning_rate": 4.431960049937578e-05,
|
| 719 |
+
"loss": 0.0049,
|
| 720 |
+
"step": 900
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 9.020224719101124,
|
| 724 |
+
"eval_accuracy": 0.4372574385510996,
|
| 725 |
+
"eval_loss": 2.7767715454101562,
|
| 726 |
+
"eval_runtime": 309.8617,
|
| 727 |
+
"eval_samples_per_second": 2.495,
|
| 728 |
+
"eval_steps_per_second": 0.158,
|
| 729 |
+
"step": 900
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 10.002247191011236,
|
| 733 |
+
"grad_norm": 0.03673604130744934,
|
| 734 |
+
"learning_rate": 4.419475655430712e-05,
|
| 735 |
+
"loss": 0.0354,
|
| 736 |
+
"step": 910
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 10.004494382022472,
|
| 740 |
+
"grad_norm": 0.021990135312080383,
|
| 741 |
+
"learning_rate": 4.406991260923846e-05,
|
| 742 |
+
"loss": 0.0017,
|
| 743 |
+
"step": 920
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 10.006741573033707,
|
| 747 |
+
"grad_norm": 0.08961662650108337,
|
| 748 |
+
"learning_rate": 4.394506866416979e-05,
|
| 749 |
+
"loss": 0.0677,
|
| 750 |
+
"step": 930
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 10.008988764044943,
|
| 754 |
+
"grad_norm": 0.15133880078792572,
|
| 755 |
+
"learning_rate": 4.3820224719101126e-05,
|
| 756 |
+
"loss": 0.0623,
|
| 757 |
+
"step": 940
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 10.01123595505618,
|
| 761 |
+
"grad_norm": 0.12693992257118225,
|
| 762 |
+
"learning_rate": 4.3695380774032463e-05,
|
| 763 |
+
"loss": 0.0073,
|
| 764 |
+
"step": 950
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 10.013483146067415,
|
| 768 |
+
"grad_norm": 0.07039490342140198,
|
| 769 |
+
"learning_rate": 4.35705368289638e-05,
|
| 770 |
+
"loss": 0.0041,
|
| 771 |
+
"step": 960
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 10.015730337078653,
|
| 775 |
+
"grad_norm": 0.054619334638118744,
|
| 776 |
+
"learning_rate": 4.344569288389513e-05,
|
| 777 |
+
"loss": 0.0329,
|
| 778 |
+
"step": 970
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 10.017977528089888,
|
| 782 |
+
"grad_norm": 0.054939307272434235,
|
| 783 |
+
"learning_rate": 4.332084893882647e-05,
|
| 784 |
+
"loss": 0.0028,
|
| 785 |
+
"step": 980
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 10.020224719101124,
|
| 789 |
+
"grad_norm": 0.03190053254365921,
|
| 790 |
+
"learning_rate": 4.3196004993757806e-05,
|
| 791 |
+
"loss": 0.0021,
|
| 792 |
+
"step": 990
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 10.020224719101124,
|
| 796 |
+
"eval_accuracy": 0.4333764553686934,
|
| 797 |
+
"eval_loss": 3.9372849464416504,
|
| 798 |
+
"eval_runtime": 307.7393,
|
| 799 |
+
"eval_samples_per_second": 2.512,
|
| 800 |
+
"eval_steps_per_second": 0.159,
|
| 801 |
+
"step": 990
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 11.002247191011236,
|
| 805 |
+
"grad_norm": 0.018473900854587555,
|
| 806 |
+
"learning_rate": 4.307116104868914e-05,
|
| 807 |
+
"loss": 0.0017,
|
| 808 |
+
"step": 1000
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 11.004494382022472,
|
| 812 |
+
"grad_norm": 0.031548287719488144,
|
| 813 |
+
"learning_rate": 4.2946317103620473e-05,
|
| 814 |
+
"loss": 0.0358,
|
| 815 |
+
"step": 1010
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 11.006741573033707,
|
| 819 |
+
"grad_norm": 0.0293214563280344,
|
| 820 |
+
"learning_rate": 4.282147315855181e-05,
|
| 821 |
+
"loss": 0.0347,
|
| 822 |
+
"step": 1020
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 11.008988764044943,
|
| 826 |
+
"grad_norm": 0.031038669869303703,
|
| 827 |
+
"learning_rate": 4.269662921348315e-05,
|
| 828 |
+
"loss": 0.0028,
|
| 829 |
+
"step": 1030
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 11.01123595505618,
|
| 833 |
+
"grad_norm": 0.03167790174484253,
|
| 834 |
+
"learning_rate": 4.2571785268414485e-05,
|
| 835 |
+
"loss": 0.0358,
|
| 836 |
+
"step": 1040
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 11.013483146067415,
|
| 840 |
+
"grad_norm": 0.045855436474084854,
|
| 841 |
+
"learning_rate": 4.2446941323345816e-05,
|
| 842 |
+
"loss": 0.0643,
|
| 843 |
+
"step": 1050
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 11.015730337078653,
|
| 847 |
+
"grad_norm": 0.08114203065633774,
|
| 848 |
+
"learning_rate": 4.232209737827715e-05,
|
| 849 |
+
"loss": 0.0333,
|
| 850 |
+
"step": 1060
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 11.017977528089888,
|
| 854 |
+
"grad_norm": 0.08527044206857681,
|
| 855 |
+
"learning_rate": 4.219725343320849e-05,
|
| 856 |
+
"loss": 0.0049,
|
| 857 |
+
"step": 1070
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 11.020224719101124,
|
| 861 |
+
"grad_norm": 0.016140511259436607,
|
| 862 |
+
"learning_rate": 4.207240948813983e-05,
|
| 863 |
+
"loss": 0.0033,
|
| 864 |
+
"step": 1080
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 11.020224719101124,
|
| 868 |
+
"eval_accuracy": 0.3984476067270375,
|
| 869 |
+
"eval_loss": 4.449057579040527,
|
| 870 |
+
"eval_runtime": 373.7261,
|
| 871 |
+
"eval_samples_per_second": 2.068,
|
| 872 |
+
"eval_steps_per_second": 0.131,
|
| 873 |
+
"step": 1080
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 12.002247191011236,
|
| 877 |
+
"grad_norm": 0.04579917713999748,
|
| 878 |
+
"learning_rate": 4.1947565543071165e-05,
|
| 879 |
+
"loss": 0.0389,
|
| 880 |
+
"step": 1090
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 12.004494382022472,
|
| 884 |
+
"grad_norm": 0.0433347187936306,
|
| 885 |
+
"learning_rate": 4.18227215980025e-05,
|
| 886 |
+
"loss": 0.0851,
|
| 887 |
+
"step": 1100
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 12.006741573033707,
|
| 891 |
+
"grad_norm": 0.04116983711719513,
|
| 892 |
+
"learning_rate": 4.169787765293384e-05,
|
| 893 |
+
"loss": 0.0702,
|
| 894 |
+
"step": 1110
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 12.008988764044943,
|
| 898 |
+
"grad_norm": 0.0757342129945755,
|
| 899 |
+
"learning_rate": 4.157303370786517e-05,
|
| 900 |
+
"loss": 0.0685,
|
| 901 |
+
"step": 1120
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 12.01123595505618,
|
| 905 |
+
"grad_norm": 0.12625212967395782,
|
| 906 |
+
"learning_rate": 4.144818976279651e-05,
|
| 907 |
+
"loss": 0.0345,
|
| 908 |
+
"step": 1130
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 12.013483146067415,
|
| 912 |
+
"grad_norm": 0.10289767384529114,
|
| 913 |
+
"learning_rate": 4.1323345817727844e-05,
|
| 914 |
+
"loss": 0.0311,
|
| 915 |
+
"step": 1140
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 12.015730337078653,
|
| 919 |
+
"grad_norm": 0.1303280144929886,
|
| 920 |
+
"learning_rate": 4.119850187265918e-05,
|
| 921 |
+
"loss": 0.0332,
|
| 922 |
+
"step": 1150
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 12.017977528089888,
|
| 926 |
+
"grad_norm": 0.03924856334924698,
|
| 927 |
+
"learning_rate": 4.107365792759051e-05,
|
| 928 |
+
"loss": 0.0045,
|
| 929 |
+
"step": 1160
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 12.020224719101124,
|
| 933 |
+
"grad_norm": 0.025983136147260666,
|
| 934 |
+
"learning_rate": 4.094881398252185e-05,
|
| 935 |
+
"loss": 0.0032,
|
| 936 |
+
"step": 1170
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 12.020224719101124,
|
| 940 |
+
"eval_accuracy": 0.517464424320828,
|
| 941 |
+
"eval_loss": 3.0959889888763428,
|
| 942 |
+
"eval_runtime": 356.3556,
|
| 943 |
+
"eval_samples_per_second": 2.169,
|
| 944 |
+
"eval_steps_per_second": 0.138,
|
| 945 |
+
"step": 1170
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 13.002247191011236,
|
| 949 |
+
"grad_norm": 0.035728052258491516,
|
| 950 |
+
"learning_rate": 4.082397003745319e-05,
|
| 951 |
+
"loss": 0.0018,
|
| 952 |
+
"step": 1180
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 13.004494382022472,
|
| 956 |
+
"grad_norm": 0.026223180815577507,
|
| 957 |
+
"learning_rate": 4.0699126092384524e-05,
|
| 958 |
+
"loss": 0.0018,
|
| 959 |
+
"step": 1190
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 13.006741573033707,
|
| 963 |
+
"grad_norm": 1.3215516805648804,
|
| 964 |
+
"learning_rate": 4.0574282147315855e-05,
|
| 965 |
+
"loss": 0.0708,
|
| 966 |
+
"step": 1200
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 13.008988764044943,
|
| 970 |
+
"grad_norm": 0.04870336502790451,
|
| 971 |
+
"learning_rate": 4.044943820224719e-05,
|
| 972 |
+
"loss": 0.0019,
|
| 973 |
+
"step": 1210
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 13.01123595505618,
|
| 977 |
+
"grad_norm": 0.040859926491975784,
|
| 978 |
+
"learning_rate": 4.032459425717853e-05,
|
| 979 |
+
"loss": 0.0025,
|
| 980 |
+
"step": 1220
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 13.013483146067415,
|
| 984 |
+
"grad_norm": 0.04700184240937233,
|
| 985 |
+
"learning_rate": 4.019975031210986e-05,
|
| 986 |
+
"loss": 0.0018,
|
| 987 |
+
"step": 1230
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 13.015730337078653,
|
| 991 |
+
"grad_norm": 0.038996122777462006,
|
| 992 |
+
"learning_rate": 4.00749063670412e-05,
|
| 993 |
+
"loss": 0.0689,
|
| 994 |
+
"step": 1240
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 13.017977528089888,
|
| 998 |
+
"grad_norm": 0.11987262219190598,
|
| 999 |
+
"learning_rate": 3.9950062421972534e-05,
|
| 1000 |
+
"loss": 0.0632,
|
| 1001 |
+
"step": 1250
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 13.020224719101124,
|
| 1005 |
+
"grad_norm": 0.19100521504878998,
|
| 1006 |
+
"learning_rate": 3.982521847690387e-05,
|
| 1007 |
+
"loss": 0.0063,
|
| 1008 |
+
"step": 1260
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 13.020224719101124,
|
| 1012 |
+
"eval_accuracy": 0.3868046571798189,
|
| 1013 |
+
"eval_loss": 4.379216194152832,
|
| 1014 |
+
"eval_runtime": 302.542,
|
| 1015 |
+
"eval_samples_per_second": 2.555,
|
| 1016 |
+
"eval_steps_per_second": 0.162,
|
| 1017 |
+
"step": 1260
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 14.002247191011236,
|
| 1021 |
+
"grad_norm": 1.3284196853637695,
|
| 1022 |
+
"learning_rate": 3.970037453183521e-05,
|
| 1023 |
+
"loss": 0.0542,
|
| 1024 |
+
"step": 1270
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 14.004494382022472,
|
| 1028 |
+
"grad_norm": 0.15026536583900452,
|
| 1029 |
+
"learning_rate": 3.9575530586766546e-05,
|
| 1030 |
+
"loss": 0.0663,
|
| 1031 |
+
"step": 1280
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 14.006741573033707,
|
| 1035 |
+
"grad_norm": 0.1195875033736229,
|
| 1036 |
+
"learning_rate": 3.945068664169788e-05,
|
| 1037 |
+
"loss": 0.0053,
|
| 1038 |
+
"step": 1290
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 14.008988764044943,
|
| 1042 |
+
"grad_norm": 0.06441330164670944,
|
| 1043 |
+
"learning_rate": 3.9325842696629214e-05,
|
| 1044 |
+
"loss": 0.0317,
|
| 1045 |
+
"step": 1300
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 14.01123595505618,
|
| 1049 |
+
"grad_norm": 0.3119406998157501,
|
| 1050 |
+
"learning_rate": 3.920099875156055e-05,
|
| 1051 |
+
"loss": 0.003,
|
| 1052 |
+
"step": 1310
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 14.013483146067415,
|
| 1056 |
+
"grad_norm": 0.05156432464718819,
|
| 1057 |
+
"learning_rate": 3.907615480649189e-05,
|
| 1058 |
+
"loss": 0.0306,
|
| 1059 |
+
"step": 1320
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 14.015730337078653,
|
| 1063 |
+
"grad_norm": 0.04788357391953468,
|
| 1064 |
+
"learning_rate": 3.8951310861423226e-05,
|
| 1065 |
+
"loss": 0.0326,
|
| 1066 |
+
"step": 1330
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 14.017977528089888,
|
| 1070 |
+
"grad_norm": 0.06591865420341492,
|
| 1071 |
+
"learning_rate": 3.8826466916354556e-05,
|
| 1072 |
+
"loss": 0.0025,
|
| 1073 |
+
"step": 1340
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 14.020224719101124,
|
| 1077 |
+
"grad_norm": 0.09529684484004974,
|
| 1078 |
+
"learning_rate": 3.870162297128589e-05,
|
| 1079 |
+
"loss": 0.0323,
|
| 1080 |
+
"step": 1350
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 14.020224719101124,
|
| 1084 |
+
"eval_accuracy": 0.3648124191461837,
|
| 1085 |
+
"eval_loss": 5.674530982971191,
|
| 1086 |
+
"eval_runtime": 314.2357,
|
| 1087 |
+
"eval_samples_per_second": 2.46,
|
| 1088 |
+
"eval_steps_per_second": 0.156,
|
| 1089 |
+
"step": 1350
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 15.002247191011236,
|
| 1093 |
+
"grad_norm": 0.08000009506940842,
|
| 1094 |
+
"learning_rate": 3.857677902621723e-05,
|
| 1095 |
+
"loss": 0.0348,
|
| 1096 |
+
"step": 1360
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 15.004494382022472,
|
| 1100 |
+
"grad_norm": 0.05557028576731682,
|
| 1101 |
+
"learning_rate": 3.845193508114857e-05,
|
| 1102 |
+
"loss": 0.0339,
|
| 1103 |
+
"step": 1370
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 15.006741573033707,
|
| 1107 |
+
"grad_norm": 0.05977383628487587,
|
| 1108 |
+
"learning_rate": 3.83270911360799e-05,
|
| 1109 |
+
"loss": 0.0033,
|
| 1110 |
+
"step": 1380
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 15.008988764044943,
|
| 1114 |
+
"grad_norm": 0.04291630908846855,
|
| 1115 |
+
"learning_rate": 3.8202247191011236e-05,
|
| 1116 |
+
"loss": 0.0029,
|
| 1117 |
+
"step": 1390
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 15.01123595505618,
|
| 1121 |
+
"grad_norm": 0.03595967963337898,
|
| 1122 |
+
"learning_rate": 3.807740324594257e-05,
|
| 1123 |
+
"loss": 0.0343,
|
| 1124 |
+
"step": 1400
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 15.013483146067415,
|
| 1128 |
+
"grad_norm": 0.03875007480382919,
|
| 1129 |
+
"learning_rate": 3.795255930087391e-05,
|
| 1130 |
+
"loss": 0.0027,
|
| 1131 |
+
"step": 1410
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 15.015730337078653,
|
| 1135 |
+
"grad_norm": 0.025117939338088036,
|
| 1136 |
+
"learning_rate": 3.782771535580524e-05,
|
| 1137 |
+
"loss": 0.0018,
|
| 1138 |
+
"step": 1420
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 15.017977528089888,
|
| 1142 |
+
"grad_norm": 1.3055503368377686,
|
| 1143 |
+
"learning_rate": 3.770287141073658e-05,
|
| 1144 |
+
"loss": 0.0363,
|
| 1145 |
+
"step": 1430
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 15.020224719101124,
|
| 1149 |
+
"grad_norm": 0.035218678414821625,
|
| 1150 |
+
"learning_rate": 3.7578027465667915e-05,
|
| 1151 |
+
"loss": 0.0673,
|
| 1152 |
+
"step": 1440
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 15.020224719101124,
|
| 1156 |
+
"eval_accuracy": 0.3648124191461837,
|
| 1157 |
+
"eval_loss": 5.768198490142822,
|
| 1158 |
+
"eval_runtime": 295.2377,
|
| 1159 |
+
"eval_samples_per_second": 2.618,
|
| 1160 |
+
"eval_steps_per_second": 0.166,
|
| 1161 |
+
"step": 1440
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 16.002247191011236,
|
| 1165 |
+
"grad_norm": 0.1669590324163437,
|
| 1166 |
+
"learning_rate": 3.745318352059925e-05,
|
| 1167 |
+
"loss": 0.0584,
|
| 1168 |
+
"step": 1450
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 16.00449438202247,
|
| 1172 |
+
"grad_norm": 0.29184722900390625,
|
| 1173 |
+
"learning_rate": 3.732833957553059e-05,
|
| 1174 |
+
"loss": 0.0539,
|
| 1175 |
+
"step": 1460
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 16.006741573033707,
|
| 1179 |
+
"grad_norm": 0.09675876796245575,
|
| 1180 |
+
"learning_rate": 3.720349563046193e-05,
|
| 1181 |
+
"loss": 0.0301,
|
| 1182 |
+
"step": 1470
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 16.008988764044943,
|
| 1186 |
+
"grad_norm": 0.11480775475502014,
|
| 1187 |
+
"learning_rate": 3.7078651685393264e-05,
|
| 1188 |
+
"loss": 0.0054,
|
| 1189 |
+
"step": 1480
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 16.01123595505618,
|
| 1193 |
+
"grad_norm": 0.09580578655004501,
|
| 1194 |
+
"learning_rate": 3.6953807740324595e-05,
|
| 1195 |
+
"loss": 0.0032,
|
| 1196 |
+
"step": 1490
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 16.013483146067415,
|
| 1200 |
+
"grad_norm": 0.14860224723815918,
|
| 1201 |
+
"learning_rate": 3.682896379525593e-05,
|
| 1202 |
+
"loss": 0.0631,
|
| 1203 |
+
"step": 1500
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 16.01573033707865,
|
| 1207 |
+
"grad_norm": 0.07166357338428497,
|
| 1208 |
+
"learning_rate": 3.670411985018727e-05,
|
| 1209 |
+
"loss": 0.0033,
|
| 1210 |
+
"step": 1510
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 16.017977528089887,
|
| 1214 |
+
"grad_norm": 0.03686222434043884,
|
| 1215 |
+
"learning_rate": 3.6579275905118607e-05,
|
| 1216 |
+
"loss": 0.0024,
|
| 1217 |
+
"step": 1520
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 16.020224719101122,
|
| 1221 |
+
"grad_norm": 0.023792656138539314,
|
| 1222 |
+
"learning_rate": 3.645443196004994e-05,
|
| 1223 |
+
"loss": 0.0018,
|
| 1224 |
+
"step": 1530
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 16.020224719101122,
|
| 1228 |
+
"eval_accuracy": 0.3648124191461837,
|
| 1229 |
+
"eval_loss": 4.825681209564209,
|
| 1230 |
+
"eval_runtime": 296.2416,
|
| 1231 |
+
"eval_samples_per_second": 2.609,
|
| 1232 |
+
"eval_steps_per_second": 0.165,
|
| 1233 |
+
"step": 1530
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 17.002247191011236,
|
| 1237 |
+
"grad_norm": 0.016941450536251068,
|
| 1238 |
+
"learning_rate": 3.6329588014981274e-05,
|
| 1239 |
+
"loss": 0.0682,
|
| 1240 |
+
"step": 1540
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 17.00449438202247,
|
| 1244 |
+
"grad_norm": 1.3070281744003296,
|
| 1245 |
+
"learning_rate": 3.620474406991261e-05,
|
| 1246 |
+
"loss": 0.0367,
|
| 1247 |
+
"step": 1550
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 17.006741573033707,
|
| 1251 |
+
"grad_norm": 0.10310333967208862,
|
| 1252 |
+
"learning_rate": 3.607990012484395e-05,
|
| 1253 |
+
"loss": 0.0347,
|
| 1254 |
+
"step": 1560
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 17.008988764044943,
|
| 1258 |
+
"grad_norm": 0.05226253718137741,
|
| 1259 |
+
"learning_rate": 3.595505617977528e-05,
|
| 1260 |
+
"loss": 0.0034,
|
| 1261 |
+
"step": 1570
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 17.01123595505618,
|
| 1265 |
+
"grad_norm": 0.06494501978158951,
|
| 1266 |
+
"learning_rate": 3.583021223470662e-05,
|
| 1267 |
+
"loss": 0.0029,
|
| 1268 |
+
"step": 1580
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 17.013483146067415,
|
| 1272 |
+
"grad_norm": 0.030287954956293106,
|
| 1273 |
+
"learning_rate": 3.5705368289637954e-05,
|
| 1274 |
+
"loss": 0.0022,
|
| 1275 |
+
"step": 1590
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 17.01573033707865,
|
| 1279 |
+
"grad_norm": 2.615795612335205,
|
| 1280 |
+
"learning_rate": 3.558052434456929e-05,
|
| 1281 |
+
"loss": 0.0691,
|
| 1282 |
+
"step": 1600
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 17.017977528089887,
|
| 1286 |
+
"grad_norm": 0.06764879822731018,
|
| 1287 |
+
"learning_rate": 3.545568039950062e-05,
|
| 1288 |
+
"loss": 0.0018,
|
| 1289 |
+
"step": 1610
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 17.020224719101122,
|
| 1293 |
+
"grad_norm": 0.011349334381520748,
|
| 1294 |
+
"learning_rate": 3.533083645443196e-05,
|
| 1295 |
+
"loss": 0.0355,
|
| 1296 |
+
"step": 1620
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 17.020224719101122,
|
| 1300 |
+
"eval_accuracy": 0.5058214747736093,
|
| 1301 |
+
"eval_loss": 4.13436222076416,
|
| 1302 |
+
"eval_runtime": 300.4098,
|
| 1303 |
+
"eval_samples_per_second": 2.573,
|
| 1304 |
+
"eval_steps_per_second": 0.163,
|
| 1305 |
+
"step": 1620
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 18.002247191011236,
|
| 1309 |
+
"grad_norm": 0.05169708654284477,
|
| 1310 |
+
"learning_rate": 3.52059925093633e-05,
|
| 1311 |
+
"loss": 0.0654,
|
| 1312 |
+
"step": 1630
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 18.00449438202247,
|
| 1316 |
+
"grad_norm": 0.09435762465000153,
|
| 1317 |
+
"learning_rate": 3.5081148564294633e-05,
|
| 1318 |
+
"loss": 0.061,
|
| 1319 |
+
"step": 1640
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 18.006741573033707,
|
| 1323 |
+
"grad_norm": 0.18586553633213043,
|
| 1324 |
+
"learning_rate": 3.495630461922597e-05,
|
| 1325 |
+
"loss": 0.0095,
|
| 1326 |
+
"step": 1650
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 18.008988764044943,
|
| 1330 |
+
"grad_norm": 0.11713388562202454,
|
| 1331 |
+
"learning_rate": 3.483146067415731e-05,
|
| 1332 |
+
"loss": 0.0049,
|
| 1333 |
+
"step": 1660
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 18.01123595505618,
|
| 1337 |
+
"grad_norm": 1.2770711183547974,
|
| 1338 |
+
"learning_rate": 3.4706616729088645e-05,
|
| 1339 |
+
"loss": 0.0323,
|
| 1340 |
+
"step": 1670
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 18.013483146067415,
|
| 1344 |
+
"grad_norm": 0.060533493757247925,
|
| 1345 |
+
"learning_rate": 3.4581772784019976e-05,
|
| 1346 |
+
"loss": 0.0031,
|
| 1347 |
+
"step": 1680
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 18.01573033707865,
|
| 1351 |
+
"grad_norm": 0.058554574847221375,
|
| 1352 |
+
"learning_rate": 3.445692883895131e-05,
|
| 1353 |
+
"loss": 0.0024,
|
| 1354 |
+
"step": 1690
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 18.017977528089887,
|
| 1358 |
+
"grad_norm": 0.03402779623866081,
|
| 1359 |
+
"learning_rate": 3.433208489388265e-05,
|
| 1360 |
+
"loss": 0.0021,
|
| 1361 |
+
"step": 1700
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 18.020224719101122,
|
| 1365 |
+
"grad_norm": 0.047420646995306015,
|
| 1366 |
+
"learning_rate": 3.420724094881399e-05,
|
| 1367 |
+
"loss": 0.0346,
|
| 1368 |
+
"step": 1710
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 18.020224719101122,
|
| 1372 |
+
"eval_accuracy": 0.5873221216041398,
|
| 1373 |
+
"eval_loss": 3.5245649814605713,
|
| 1374 |
+
"eval_runtime": 303.2193,
|
| 1375 |
+
"eval_samples_per_second": 2.549,
|
| 1376 |
+
"eval_steps_per_second": 0.162,
|
| 1377 |
+
"step": 1710
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 19.002247191011236,
|
| 1381 |
+
"grad_norm": 0.047677114605903625,
|
| 1382 |
+
"learning_rate": 3.408239700374532e-05,
|
| 1383 |
+
"loss": 0.002,
|
| 1384 |
+
"step": 1720
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 19.00449438202247,
|
| 1388 |
+
"grad_norm": 0.06761440634727478,
|
| 1389 |
+
"learning_rate": 3.3957553058676655e-05,
|
| 1390 |
+
"loss": 0.0691,
|
| 1391 |
+
"step": 1730
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 19.006741573033707,
|
| 1395 |
+
"grad_norm": 0.06686104089021683,
|
| 1396 |
+
"learning_rate": 3.383270911360799e-05,
|
| 1397 |
+
"loss": 0.0028,
|
| 1398 |
+
"step": 1740
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 19.008988764044943,
|
| 1402 |
+
"grad_norm": 0.04437251761555672,
|
| 1403 |
+
"learning_rate": 3.370786516853933e-05,
|
| 1404 |
+
"loss": 0.033,
|
| 1405 |
+
"step": 1750
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 19.01123595505618,
|
| 1409 |
+
"grad_norm": 0.07953111082315445,
|
| 1410 |
+
"learning_rate": 3.358302122347066e-05,
|
| 1411 |
+
"loss": 0.033,
|
| 1412 |
+
"step": 1760
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 19.013483146067415,
|
| 1416 |
+
"grad_norm": 0.060533471405506134,
|
| 1417 |
+
"learning_rate": 3.3458177278402e-05,
|
| 1418 |
+
"loss": 0.0315,
|
| 1419 |
+
"step": 1770
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 19.01573033707865,
|
| 1423 |
+
"grad_norm": 0.09098009020090103,
|
| 1424 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 1425 |
+
"loss": 0.0322,
|
| 1426 |
+
"step": 1780
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 19.017977528089887,
|
| 1430 |
+
"grad_norm": 0.08467002213001251,
|
| 1431 |
+
"learning_rate": 3.3208489388264665e-05,
|
| 1432 |
+
"loss": 0.0046,
|
| 1433 |
+
"step": 1790
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 19.020224719101122,
|
| 1437 |
+
"grad_norm": 0.08982325345277786,
|
| 1438 |
+
"learning_rate": 3.3083645443196e-05,
|
| 1439 |
+
"loss": 0.0032,
|
| 1440 |
+
"step": 1800
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 19.020224719101122,
|
| 1444 |
+
"eval_accuracy": 0.5071151358344114,
|
| 1445 |
+
"eval_loss": 4.05754280090332,
|
| 1446 |
+
"eval_runtime": 309.6055,
|
| 1447 |
+
"eval_samples_per_second": 2.497,
|
| 1448 |
+
"eval_steps_per_second": 0.158,
|
| 1449 |
+
"step": 1800
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 20.002247191011236,
|
| 1453 |
+
"grad_norm": 0.02631296031177044,
|
| 1454 |
+
"learning_rate": 3.295880149812734e-05,
|
| 1455 |
+
"loss": 0.0019,
|
| 1456 |
+
"step": 1810
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 20.00449438202247,
|
| 1460 |
+
"grad_norm": 0.021504152566194534,
|
| 1461 |
+
"learning_rate": 3.2833957553058684e-05,
|
| 1462 |
+
"loss": 0.0021,
|
| 1463 |
+
"step": 1820
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 20.006741573033707,
|
| 1467 |
+
"grad_norm": 0.06888879090547562,
|
| 1468 |
+
"learning_rate": 3.2709113607990015e-05,
|
| 1469 |
+
"loss": 0.0959,
|
| 1470 |
+
"step": 1830
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 20.008988764044943,
|
| 1474 |
+
"grad_norm": 0.06275144964456558,
|
| 1475 |
+
"learning_rate": 3.258426966292135e-05,
|
| 1476 |
+
"loss": 0.003,
|
| 1477 |
+
"step": 1840
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 20.01123595505618,
|
| 1481 |
+
"grad_norm": 0.08931156247854233,
|
| 1482 |
+
"learning_rate": 3.245942571785269e-05,
|
| 1483 |
+
"loss": 0.0317,
|
| 1484 |
+
"step": 1850
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 20.013483146067415,
|
| 1488 |
+
"grad_norm": 1.2402229309082031,
|
| 1489 |
+
"learning_rate": 3.233458177278402e-05,
|
| 1490 |
+
"loss": 0.0309,
|
| 1491 |
+
"step": 1860
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 20.01573033707865,
|
| 1495 |
+
"grad_norm": 0.09801881015300751,
|
| 1496 |
+
"learning_rate": 3.220973782771536e-05,
|
| 1497 |
+
"loss": 0.0044,
|
| 1498 |
+
"step": 1870
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 20.017977528089887,
|
| 1502 |
+
"grad_norm": 0.0487775057554245,
|
| 1503 |
+
"learning_rate": 3.2084893882646694e-05,
|
| 1504 |
+
"loss": 0.0033,
|
| 1505 |
+
"step": 1880
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 20.020224719101122,
|
| 1509 |
+
"grad_norm": 0.08219785243272781,
|
| 1510 |
+
"learning_rate": 3.196004993757803e-05,
|
| 1511 |
+
"loss": 0.0325,
|
| 1512 |
+
"step": 1890
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 20.020224719101122,
|
| 1516 |
+
"eval_accuracy": 0.4282018111254851,
|
| 1517 |
+
"eval_loss": 4.583955764770508,
|
| 1518 |
+
"eval_runtime": 316.2855,
|
| 1519 |
+
"eval_samples_per_second": 2.444,
|
| 1520 |
+
"eval_steps_per_second": 0.155,
|
| 1521 |
+
"step": 1890
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 21.002247191011236,
|
| 1525 |
+
"grad_norm": 0.06626549363136292,
|
| 1526 |
+
"learning_rate": 3.183520599250936e-05,
|
| 1527 |
+
"loss": 0.0027,
|
| 1528 |
+
"step": 1900
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 21.00449438202247,
|
| 1532 |
+
"grad_norm": 0.07979313284158707,
|
| 1533 |
+
"learning_rate": 3.17103620474407e-05,
|
| 1534 |
+
"loss": 0.0326,
|
| 1535 |
+
"step": 1910
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 21.006741573033707,
|
| 1539 |
+
"grad_norm": 0.038574449717998505,
|
| 1540 |
+
"learning_rate": 3.1585518102372036e-05,
|
| 1541 |
+
"loss": 0.0316,
|
| 1542 |
+
"step": 1920
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 21.008988764044943,
|
| 1546 |
+
"grad_norm": 0.049700915813446045,
|
| 1547 |
+
"learning_rate": 3.1460674157303374e-05,
|
| 1548 |
+
"loss": 0.0029,
|
| 1549 |
+
"step": 1930
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 21.01123595505618,
|
| 1553 |
+
"grad_norm": 0.045398663729429245,
|
| 1554 |
+
"learning_rate": 3.1335830212234704e-05,
|
| 1555 |
+
"loss": 0.0305,
|
| 1556 |
+
"step": 1940
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 21.013483146067415,
|
| 1560 |
+
"grad_norm": 0.05370797589421272,
|
| 1561 |
+
"learning_rate": 3.121098626716604e-05,
|
| 1562 |
+
"loss": 0.0338,
|
| 1563 |
+
"step": 1950
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 21.01573033707865,
|
| 1567 |
+
"grad_norm": 0.2418488711118698,
|
| 1568 |
+
"learning_rate": 3.108614232209738e-05,
|
| 1569 |
+
"loss": 0.0334,
|
| 1570 |
+
"step": 1960
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 21.017977528089887,
|
| 1574 |
+
"grad_norm": 0.06898712366819382,
|
| 1575 |
+
"learning_rate": 3.0961298377028716e-05,
|
| 1576 |
+
"loss": 0.0041,
|
| 1577 |
+
"step": 1970
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 21.020224719101122,
|
| 1581 |
+
"grad_norm": 0.0021224431693553925,
|
| 1582 |
+
"learning_rate": 3.0836454431960046e-05,
|
| 1583 |
+
"loss": 0.0413,
|
| 1584 |
+
"step": 1980
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 21.020224719101122,
|
| 1588 |
+
"eval_accuracy": 0.5705045278137129,
|
| 1589 |
+
"eval_loss": 3.6885340213775635,
|
| 1590 |
+
"eval_runtime": 310.6197,
|
| 1591 |
+
"eval_samples_per_second": 2.489,
|
| 1592 |
+
"eval_steps_per_second": 0.158,
|
| 1593 |
+
"step": 1980
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 22.002247191011236,
|
| 1597 |
+
"grad_norm": 1.2743548154830933,
|
| 1598 |
+
"learning_rate": 3.0711610486891384e-05,
|
| 1599 |
+
"loss": 0.0312,
|
| 1600 |
+
"step": 1990
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 22.00449438202247,
|
| 1604 |
+
"grad_norm": 0.05916481465101242,
|
| 1605 |
+
"learning_rate": 3.058676654182273e-05,
|
| 1606 |
+
"loss": 0.0324,
|
| 1607 |
+
"step": 2000
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 22.006741573033707,
|
| 1611 |
+
"grad_norm": 0.1024039015173912,
|
| 1612 |
+
"learning_rate": 3.046192259675406e-05,
|
| 1613 |
+
"loss": 0.0311,
|
| 1614 |
+
"step": 2010
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 22.008988764044943,
|
| 1618 |
+
"grad_norm": 0.03204580396413803,
|
| 1619 |
+
"learning_rate": 3.0337078651685396e-05,
|
| 1620 |
+
"loss": 0.0041,
|
| 1621 |
+
"step": 2020
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 22.01123595505618,
|
| 1625 |
+
"grad_norm": 0.09468382596969604,
|
| 1626 |
+
"learning_rate": 3.0212234706616733e-05,
|
| 1627 |
+
"loss": 0.0305,
|
| 1628 |
+
"step": 2030
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 22.013483146067415,
|
| 1632 |
+
"grad_norm": 0.11132004857063293,
|
| 1633 |
+
"learning_rate": 3.0087390761548067e-05,
|
| 1634 |
+
"loss": 0.0294,
|
| 1635 |
+
"step": 2040
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 22.01573033707865,
|
| 1639 |
+
"grad_norm": 0.026780493557453156,
|
| 1640 |
+
"learning_rate": 2.9962546816479404e-05,
|
| 1641 |
+
"loss": 0.0045,
|
| 1642 |
+
"step": 2050
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 22.017977528089887,
|
| 1646 |
+
"grad_norm": 0.006886694114655256,
|
| 1647 |
+
"learning_rate": 2.9837702871410738e-05,
|
| 1648 |
+
"loss": 0.0024,
|
| 1649 |
+
"step": 2060
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 22.020224719101122,
|
| 1653 |
+
"grad_norm": 0.08298806101083755,
|
| 1654 |
+
"learning_rate": 2.9712858926342075e-05,
|
| 1655 |
+
"loss": 0.0909,
|
| 1656 |
+
"step": 2070
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 22.020224719101122,
|
| 1660 |
+
"eval_accuracy": 0.8188874514877102,
|
| 1661 |
+
"eval_loss": 0.9124556183815002,
|
| 1662 |
+
"eval_runtime": 303.7511,
|
| 1663 |
+
"eval_samples_per_second": 2.545,
|
| 1664 |
+
"eval_steps_per_second": 0.161,
|
| 1665 |
+
"step": 2070
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 23.002247191011236,
|
| 1669 |
+
"grad_norm": 1.3124110698699951,
|
| 1670 |
+
"learning_rate": 2.958801498127341e-05,
|
| 1671 |
+
"loss": 0.0304,
|
| 1672 |
+
"step": 2080
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 23.00449438202247,
|
| 1676 |
+
"grad_norm": 0.1172858476638794,
|
| 1677 |
+
"learning_rate": 2.9463171036204746e-05,
|
| 1678 |
+
"loss": 0.0308,
|
| 1679 |
+
"step": 2090
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 23.006741573033707,
|
| 1683 |
+
"grad_norm": 0.11574792116880417,
|
| 1684 |
+
"learning_rate": 2.933832709113608e-05,
|
| 1685 |
+
"loss": 0.0301,
|
| 1686 |
+
"step": 2100
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 23.008988764044943,
|
| 1690 |
+
"grad_norm": 0.09672237932682037,
|
| 1691 |
+
"learning_rate": 2.9213483146067417e-05,
|
| 1692 |
+
"loss": 0.0041,
|
| 1693 |
+
"step": 2110
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 23.01123595505618,
|
| 1697 |
+
"grad_norm": 0.039515670388936996,
|
| 1698 |
+
"learning_rate": 2.908863920099875e-05,
|
| 1699 |
+
"loss": 0.0299,
|
| 1700 |
+
"step": 2120
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 23.013483146067415,
|
| 1704 |
+
"grad_norm": 0.08289401233196259,
|
| 1705 |
+
"learning_rate": 2.896379525593009e-05,
|
| 1706 |
+
"loss": 0.0333,
|
| 1707 |
+
"step": 2130
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 23.01573033707865,
|
| 1711 |
+
"grad_norm": 0.05611934885382652,
|
| 1712 |
+
"learning_rate": 2.8838951310861422e-05,
|
| 1713 |
+
"loss": 0.0037,
|
| 1714 |
+
"step": 2140
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 23.017977528089887,
|
| 1718 |
+
"grad_norm": 0.029819313436746597,
|
| 1719 |
+
"learning_rate": 2.871410736579276e-05,
|
| 1720 |
+
"loss": 0.002,
|
| 1721 |
+
"step": 2150
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 23.020224719101122,
|
| 1725 |
+
"grad_norm": 0.010641505010426044,
|
| 1726 |
+
"learning_rate": 2.8589263420724094e-05,
|
| 1727 |
+
"loss": 0.037,
|
| 1728 |
+
"step": 2160
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 23.020224719101122,
|
| 1732 |
+
"eval_accuracy": 0.8771021992238034,
|
| 1733 |
+
"eval_loss": 0.9978875517845154,
|
| 1734 |
+
"eval_runtime": 372.4855,
|
| 1735 |
+
"eval_samples_per_second": 2.075,
|
| 1736 |
+
"eval_steps_per_second": 0.132,
|
| 1737 |
+
"step": 2160
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 24.002247191011236,
|
| 1741 |
+
"grad_norm": 0.048041634261608124,
|
| 1742 |
+
"learning_rate": 2.846441947565543e-05,
|
| 1743 |
+
"loss": 0.0016,
|
| 1744 |
+
"step": 2170
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 24.00449438202247,
|
| 1748 |
+
"grad_norm": 0.08231844753026962,
|
| 1749 |
+
"learning_rate": 2.833957553058677e-05,
|
| 1750 |
+
"loss": 0.0345,
|
| 1751 |
+
"step": 2180
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 24.006741573033707,
|
| 1755 |
+
"grad_norm": 0.05691877752542496,
|
| 1756 |
+
"learning_rate": 2.8214731585518105e-05,
|
| 1757 |
+
"loss": 0.002,
|
| 1758 |
+
"step": 2190
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 24.008988764044943,
|
| 1762 |
+
"grad_norm": 0.04746109992265701,
|
| 1763 |
+
"learning_rate": 2.8089887640449443e-05,
|
| 1764 |
+
"loss": 0.0315,
|
| 1765 |
+
"step": 2200
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 24.01123595505618,
|
| 1769 |
+
"grad_norm": 0.08345949649810791,
|
| 1770 |
+
"learning_rate": 2.7965043695380777e-05,
|
| 1771 |
+
"loss": 0.0526,
|
| 1772 |
+
"step": 2210
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 24.013483146067415,
|
| 1776 |
+
"grad_norm": 0.012946105562150478,
|
| 1777 |
+
"learning_rate": 2.7840199750312114e-05,
|
| 1778 |
+
"loss": 0.0044,
|
| 1779 |
+
"step": 2220
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 24.01573033707865,
|
| 1783 |
+
"grad_norm": 0.10040261596441269,
|
| 1784 |
+
"learning_rate": 2.7715355805243448e-05,
|
| 1785 |
+
"loss": 0.0752,
|
| 1786 |
+
"step": 2230
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 24.017977528089887,
|
| 1790 |
+
"grad_norm": 0.01900775358080864,
|
| 1791 |
+
"learning_rate": 2.7590511860174785e-05,
|
| 1792 |
+
"loss": 0.0328,
|
| 1793 |
+
"step": 2240
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 24.020224719101122,
|
| 1797 |
+
"grad_norm": 0.05815144628286362,
|
| 1798 |
+
"learning_rate": 2.746566791510612e-05,
|
| 1799 |
+
"loss": 0.0037,
|
| 1800 |
+
"step": 2250
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 24.020224719101122,
|
| 1804 |
+
"eval_accuracy": 0.37128072445019406,
|
| 1805 |
+
"eval_loss": 3.880995273590088,
|
| 1806 |
+
"eval_runtime": 338.6791,
|
| 1807 |
+
"eval_samples_per_second": 2.282,
|
| 1808 |
+
"eval_steps_per_second": 0.145,
|
| 1809 |
+
"step": 2250
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 25.002247191011236,
|
| 1813 |
+
"grad_norm": 0.03393116593360901,
|
| 1814 |
+
"learning_rate": 2.7340823970037456e-05,
|
| 1815 |
+
"loss": 0.0025,
|
| 1816 |
+
"step": 2260
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 25.00449438202247,
|
| 1820 |
+
"grad_norm": 0.01504984125494957,
|
| 1821 |
+
"learning_rate": 2.721598002496879e-05,
|
| 1822 |
+
"loss": 0.0008,
|
| 1823 |
+
"step": 2270
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 25.006741573033707,
|
| 1827 |
+
"grad_norm": 0.008866808377206326,
|
| 1828 |
+
"learning_rate": 2.7091136079900127e-05,
|
| 1829 |
+
"loss": 0.0005,
|
| 1830 |
+
"step": 2280
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 25.008988764044943,
|
| 1834 |
+
"grad_norm": 0.006507043726742268,
|
| 1835 |
+
"learning_rate": 2.696629213483146e-05,
|
| 1836 |
+
"loss": 0.0004,
|
| 1837 |
+
"step": 2290
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 25.01123595505618,
|
| 1841 |
+
"grad_norm": 0.005518138408660889,
|
| 1842 |
+
"learning_rate": 2.6841448189762795e-05,
|
| 1843 |
+
"loss": 0.0002,
|
| 1844 |
+
"step": 2300
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 25.013483146067415,
|
| 1848 |
+
"grad_norm": 0.020528404042124748,
|
| 1849 |
+
"learning_rate": 2.6716604244694132e-05,
|
| 1850 |
+
"loss": 0.0827,
|
| 1851 |
+
"step": 2310
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 25.01573033707865,
|
| 1855 |
+
"grad_norm": 1.467294454574585,
|
| 1856 |
+
"learning_rate": 2.6591760299625466e-05,
|
| 1857 |
+
"loss": 0.0988,
|
| 1858 |
+
"step": 2320
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 25.017977528089887,
|
| 1862 |
+
"grad_norm": 0.19738095998764038,
|
| 1863 |
+
"learning_rate": 2.6466916354556804e-05,
|
| 1864 |
+
"loss": 0.0621,
|
| 1865 |
+
"step": 2330
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 25.020224719101122,
|
| 1869 |
+
"grad_norm": 0.08690626919269562,
|
| 1870 |
+
"learning_rate": 2.6342072409488137e-05,
|
| 1871 |
+
"loss": 0.0283,
|
| 1872 |
+
"step": 2340
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 25.020224719101122,
|
| 1876 |
+
"eval_accuracy": 0.3648124191461837,
|
| 1877 |
+
"eval_loss": 5.556036949157715,
|
| 1878 |
+
"eval_runtime": 296.3412,
|
| 1879 |
+
"eval_samples_per_second": 2.608,
|
| 1880 |
+
"eval_steps_per_second": 0.165,
|
| 1881 |
+
"step": 2340
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 26.002247191011236,
|
| 1885 |
+
"grad_norm": 0.14786496758460999,
|
| 1886 |
+
"learning_rate": 2.6217228464419475e-05,
|
| 1887 |
+
"loss": 0.029,
|
| 1888 |
+
"step": 2350
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 26.00449438202247,
|
| 1892 |
+
"grad_norm": 0.09694144874811172,
|
| 1893 |
+
"learning_rate": 2.6092384519350815e-05,
|
| 1894 |
+
"loss": 0.0556,
|
| 1895 |
+
"step": 2360
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 26.006741573033707,
|
| 1899 |
+
"grad_norm": 0.0992373526096344,
|
| 1900 |
+
"learning_rate": 2.5967540574282153e-05,
|
| 1901 |
+
"loss": 0.0066,
|
| 1902 |
+
"step": 2370
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 26.008988764044943,
|
| 1906 |
+
"grad_norm": 0.07103633135557175,
|
| 1907 |
+
"learning_rate": 2.5842696629213486e-05,
|
| 1908 |
+
"loss": 0.0048,
|
| 1909 |
+
"step": 2380
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 26.01123595505618,
|
| 1913 |
+
"grad_norm": 0.03303452953696251,
|
| 1914 |
+
"learning_rate": 2.571785268414482e-05,
|
| 1915 |
+
"loss": 0.0665,
|
| 1916 |
+
"step": 2390
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 26.013483146067415,
|
| 1920 |
+
"grad_norm": 0.11361945420503616,
|
| 1921 |
+
"learning_rate": 2.5593008739076158e-05,
|
| 1922 |
+
"loss": 0.0041,
|
| 1923 |
+
"step": 2400
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 26.01573033707865,
|
| 1927 |
+
"grad_norm": 0.05113370344042778,
|
| 1928 |
+
"learning_rate": 2.546816479400749e-05,
|
| 1929 |
+
"loss": 0.0305,
|
| 1930 |
+
"step": 2410
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 26.017977528089887,
|
| 1934 |
+
"grad_norm": 0.07407377660274506,
|
| 1935 |
+
"learning_rate": 2.534332084893883e-05,
|
| 1936 |
+
"loss": 0.0027,
|
| 1937 |
+
"step": 2420
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 26.020224719101122,
|
| 1941 |
+
"grad_norm": 0.04178846254944801,
|
| 1942 |
+
"learning_rate": 2.5218476903870163e-05,
|
| 1943 |
+
"loss": 0.0024,
|
| 1944 |
+
"step": 2430
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 26.020224719101122,
|
| 1948 |
+
"eval_accuracy": 0.3648124191461837,
|
| 1949 |
+
"eval_loss": 5.732365131378174,
|
| 1950 |
+
"eval_runtime": 293.1434,
|
| 1951 |
+
"eval_samples_per_second": 2.637,
|
| 1952 |
+
"eval_steps_per_second": 0.167,
|
| 1953 |
+
"step": 2430
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 27.002247191011236,
|
| 1957 |
+
"grad_norm": 0.023682553321123123,
|
| 1958 |
+
"learning_rate": 2.50936329588015e-05,
|
| 1959 |
+
"loss": 0.0309,
|
| 1960 |
+
"step": 2440
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 27.00449438202247,
|
| 1964 |
+
"grad_norm": 15.560580253601074,
|
| 1965 |
+
"learning_rate": 2.4968789013732834e-05,
|
| 1966 |
+
"loss": 0.072,
|
| 1967 |
+
"step": 2450
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 27.006741573033707,
|
| 1971 |
+
"grad_norm": 0.05713484063744545,
|
| 1972 |
+
"learning_rate": 2.484394506866417e-05,
|
| 1973 |
+
"loss": 0.0034,
|
| 1974 |
+
"step": 2460
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 27.008988764044943,
|
| 1978 |
+
"grad_norm": 0.08814993500709534,
|
| 1979 |
+
"learning_rate": 2.4719101123595505e-05,
|
| 1980 |
+
"loss": 0.0551,
|
| 1981 |
+
"step": 2470
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 27.01123595505618,
|
| 1985 |
+
"grad_norm": 0.05809677764773369,
|
| 1986 |
+
"learning_rate": 2.4594257178526842e-05,
|
| 1987 |
+
"loss": 0.0031,
|
| 1988 |
+
"step": 2480
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 27.013483146067415,
|
| 1992 |
+
"grad_norm": 0.028738778084516525,
|
| 1993 |
+
"learning_rate": 2.4469413233458176e-05,
|
| 1994 |
+
"loss": 0.0309,
|
| 1995 |
+
"step": 2490
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 27.01573033707865,
|
| 1999 |
+
"grad_norm": 0.052632275968790054,
|
| 2000 |
+
"learning_rate": 2.4344569288389517e-05,
|
| 2001 |
+
"loss": 0.0024,
|
| 2002 |
+
"step": 2500
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 27.017977528089887,
|
| 2006 |
+
"grad_norm": 0.041094765067100525,
|
| 2007 |
+
"learning_rate": 2.421972534332085e-05,
|
| 2008 |
+
"loss": 0.0014,
|
| 2009 |
+
"step": 2510
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 27.020224719101122,
|
| 2013 |
+
"grad_norm": 0.03725634515285492,
|
| 2014 |
+
"learning_rate": 2.4094881398252188e-05,
|
| 2015 |
+
"loss": 0.0518,
|
| 2016 |
+
"step": 2520
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 27.020224719101122,
|
| 2020 |
+
"eval_accuracy": 0.3648124191461837,
|
| 2021 |
+
"eval_loss": 5.63041353225708,
|
| 2022 |
+
"eval_runtime": 300.2011,
|
| 2023 |
+
"eval_samples_per_second": 2.575,
|
| 2024 |
+
"eval_steps_per_second": 0.163,
|
| 2025 |
+
"step": 2520
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 28.002247191011236,
|
| 2029 |
+
"grad_norm": 0.015194701962172985,
|
| 2030 |
+
"learning_rate": 2.3970037453183522e-05,
|
| 2031 |
+
"loss": 0.0454,
|
| 2032 |
+
"step": 2530
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 28.00449438202247,
|
| 2036 |
+
"grad_norm": 0.04806046932935715,
|
| 2037 |
+
"learning_rate": 2.384519350811486e-05,
|
| 2038 |
+
"loss": 0.0022,
|
| 2039 |
+
"step": 2540
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 28.006741573033707,
|
| 2043 |
+
"grad_norm": 0.02999090403318405,
|
| 2044 |
+
"learning_rate": 2.3720349563046193e-05,
|
| 2045 |
+
"loss": 0.0014,
|
| 2046 |
+
"step": 2550
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 28.008988764044943,
|
| 2050 |
+
"grad_norm": 0.02070489712059498,
|
| 2051 |
+
"learning_rate": 2.359550561797753e-05,
|
| 2052 |
+
"loss": 0.001,
|
| 2053 |
+
"step": 2560
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 28.01123595505618,
|
| 2057 |
+
"grad_norm": 0.0163896307349205,
|
| 2058 |
+
"learning_rate": 2.3470661672908864e-05,
|
| 2059 |
+
"loss": 0.0694,
|
| 2060 |
+
"step": 2570
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 28.013483146067415,
|
| 2064 |
+
"grad_norm": 0.18144291639328003,
|
| 2065 |
+
"learning_rate": 2.3345817727840198e-05,
|
| 2066 |
+
"loss": 0.0565,
|
| 2067 |
+
"step": 2580
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 28.01573033707865,
|
| 2071 |
+
"grad_norm": 0.14388130605220795,
|
| 2072 |
+
"learning_rate": 2.322097378277154e-05,
|
| 2073 |
+
"loss": 0.0178,
|
| 2074 |
+
"step": 2590
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 28.017977528089887,
|
| 2078 |
+
"grad_norm": 0.03906840458512306,
|
| 2079 |
+
"learning_rate": 2.3096129837702873e-05,
|
| 2080 |
+
"loss": 0.0034,
|
| 2081 |
+
"step": 2600
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 28.020224719101122,
|
| 2085 |
+
"grad_norm": 0.020977217704057693,
|
| 2086 |
+
"learning_rate": 2.297128589263421e-05,
|
| 2087 |
+
"loss": 0.0018,
|
| 2088 |
+
"step": 2610
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 28.020224719101122,
|
| 2092 |
+
"eval_accuracy": 0.4126778783958603,
|
| 2093 |
+
"eval_loss": 3.4627420902252197,
|
| 2094 |
+
"eval_runtime": 303.2398,
|
| 2095 |
+
"eval_samples_per_second": 2.549,
|
| 2096 |
+
"eval_steps_per_second": 0.162,
|
| 2097 |
+
"step": 2610
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 29.002247191011236,
|
| 2101 |
+
"grad_norm": 0.02470986731350422,
|
| 2102 |
+
"learning_rate": 2.2846441947565544e-05,
|
| 2103 |
+
"loss": 0.0363,
|
| 2104 |
+
"step": 2620
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 29.00449438202247,
|
| 2108 |
+
"grad_norm": 0.08453196287155151,
|
| 2109 |
+
"learning_rate": 2.272159800249688e-05,
|
| 2110 |
+
"loss": 0.033,
|
| 2111 |
+
"step": 2630
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 29.006741573033707,
|
| 2115 |
+
"grad_norm": 0.06280290335416794,
|
| 2116 |
+
"learning_rate": 2.2596754057428215e-05,
|
| 2117 |
+
"loss": 0.0297,
|
| 2118 |
+
"step": 2640
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 29.008988764044943,
|
| 2122 |
+
"grad_norm": 0.2928611636161804,
|
| 2123 |
+
"learning_rate": 2.2471910112359552e-05,
|
| 2124 |
+
"loss": 0.0138,
|
| 2125 |
+
"step": 2650
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 29.01123595505618,
|
| 2129 |
+
"grad_norm": 0.022598376497626305,
|
| 2130 |
+
"learning_rate": 2.2347066167290886e-05,
|
| 2131 |
+
"loss": 0.0025,
|
| 2132 |
+
"step": 2660
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 29.013483146067415,
|
| 2136 |
+
"grad_norm": 0.019650747999548912,
|
| 2137 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 2138 |
+
"loss": 0.0016,
|
| 2139 |
+
"step": 2670
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 29.01573033707865,
|
| 2143 |
+
"grad_norm": 0.036207448691129684,
|
| 2144 |
+
"learning_rate": 2.209737827715356e-05,
|
| 2145 |
+
"loss": 0.039,
|
| 2146 |
+
"step": 2680
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 29.017977528089887,
|
| 2150 |
+
"grad_norm": 0.08668278902769089,
|
| 2151 |
+
"learning_rate": 2.1972534332084894e-05,
|
| 2152 |
+
"loss": 0.0625,
|
| 2153 |
+
"step": 2690
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 29.020224719101122,
|
| 2157 |
+
"grad_norm": 0.06385768949985504,
|
| 2158 |
+
"learning_rate": 2.1847690387016232e-05,
|
| 2159 |
+
"loss": 0.0024,
|
| 2160 |
+
"step": 2700
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 29.020224719101122,
|
| 2164 |
+
"eval_accuracy": 0.3648124191461837,
|
| 2165 |
+
"eval_loss": 5.386835098266602,
|
| 2166 |
+
"eval_runtime": 362.8581,
|
| 2167 |
+
"eval_samples_per_second": 2.13,
|
| 2168 |
+
"eval_steps_per_second": 0.135,
|
| 2169 |
+
"step": 2700
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 30.002247191011236,
|
| 2173 |
+
"grad_norm": 0.016925379633903503,
|
| 2174 |
+
"learning_rate": 2.1722846441947566e-05,
|
| 2175 |
+
"loss": 0.0018,
|
| 2176 |
+
"step": 2710
|
| 2177 |
+
},
|
| 2178 |
+
{
|
| 2179 |
+
"epoch": 30.00449438202247,
|
| 2180 |
+
"grad_norm": 0.02950431779026985,
|
| 2181 |
+
"learning_rate": 2.1598002496878903e-05,
|
| 2182 |
+
"loss": 0.0365,
|
| 2183 |
+
"step": 2720
|
| 2184 |
+
},
|
| 2185 |
+
{
|
| 2186 |
+
"epoch": 30.006741573033707,
|
| 2187 |
+
"grad_norm": 0.025786850601434708,
|
| 2188 |
+
"learning_rate": 2.1473158551810237e-05,
|
| 2189 |
+
"loss": 0.0316,
|
| 2190 |
+
"step": 2730
|
| 2191 |
+
},
|
| 2192 |
+
{
|
| 2193 |
+
"epoch": 30.008988764044943,
|
| 2194 |
+
"grad_norm": 0.1407640129327774,
|
| 2195 |
+
"learning_rate": 2.1348314606741574e-05,
|
| 2196 |
+
"loss": 0.0091,
|
| 2197 |
+
"step": 2740
|
| 2198 |
+
},
|
| 2199 |
+
{
|
| 2200 |
+
"epoch": 30.01123595505618,
|
| 2201 |
+
"grad_norm": 0.07694732397794724,
|
| 2202 |
+
"learning_rate": 2.1223470661672908e-05,
|
| 2203 |
+
"loss": 0.0336,
|
| 2204 |
+
"step": 2750
|
| 2205 |
+
},
|
| 2206 |
+
{
|
| 2207 |
+
"epoch": 30.013483146067415,
|
| 2208 |
+
"grad_norm": 0.08740699291229248,
|
| 2209 |
+
"learning_rate": 2.1098626716604245e-05,
|
| 2210 |
+
"loss": 0.0328,
|
| 2211 |
+
"step": 2760
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 30.01573033707865,
|
| 2215 |
+
"grad_norm": 0.07603742182254791,
|
| 2216 |
+
"learning_rate": 2.0973782771535582e-05,
|
| 2217 |
+
"loss": 0.0301,
|
| 2218 |
+
"step": 2770
|
| 2219 |
+
},
|
| 2220 |
+
{
|
| 2221 |
+
"epoch": 30.017977528089887,
|
| 2222 |
+
"grad_norm": 0.02898871898651123,
|
| 2223 |
+
"learning_rate": 2.084893882646692e-05,
|
| 2224 |
+
"loss": 0.0019,
|
| 2225 |
+
"step": 2780
|
| 2226 |
+
},
|
| 2227 |
+
{
|
| 2228 |
+
"epoch": 30.020224719101122,
|
| 2229 |
+
"grad_norm": 0.03231247514486313,
|
| 2230 |
+
"learning_rate": 2.0724094881398254e-05,
|
| 2231 |
+
"loss": 0.0021,
|
| 2232 |
+
"step": 2790
|
| 2233 |
+
},
|
| 2234 |
+
{
|
| 2235 |
+
"epoch": 30.020224719101122,
|
| 2236 |
+
"eval_accuracy": 0.8771021992238034,
|
| 2237 |
+
"eval_loss": 0.8241329789161682,
|
| 2238 |
+
"eval_runtime": 354.869,
|
| 2239 |
+
"eval_samples_per_second": 2.178,
|
| 2240 |
+
"eval_steps_per_second": 0.138,
|
| 2241 |
+
"step": 2790
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 31.002247191011236,
|
| 2245 |
+
"grad_norm": 0.08487377315759659,
|
| 2246 |
+
"learning_rate": 2.059925093632959e-05,
|
| 2247 |
+
"loss": 0.0915,
|
| 2248 |
+
"step": 2800
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 31.00449438202247,
|
| 2252 |
+
"grad_norm": 0.30065643787384033,
|
| 2253 |
+
"learning_rate": 2.0474406991260925e-05,
|
| 2254 |
+
"loss": 0.0315,
|
| 2255 |
+
"step": 2810
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 31.006741573033707,
|
| 2259 |
+
"grad_norm": 0.15625542402267456,
|
| 2260 |
+
"learning_rate": 2.0349563046192262e-05,
|
| 2261 |
+
"loss": 0.0076,
|
| 2262 |
+
"step": 2820
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 31.008988764044943,
|
| 2266 |
+
"grad_norm": 0.049321502447128296,
|
| 2267 |
+
"learning_rate": 2.0224719101123596e-05,
|
| 2268 |
+
"loss": 0.0028,
|
| 2269 |
+
"step": 2830
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"epoch": 31.01123595505618,
|
| 2273 |
+
"grad_norm": 0.03509892523288727,
|
| 2274 |
+
"learning_rate": 2.009987515605493e-05,
|
| 2275 |
+
"loss": 0.034,
|
| 2276 |
+
"step": 2840
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 31.013483146067415,
|
| 2280 |
+
"grad_norm": 0.02723236382007599,
|
| 2281 |
+
"learning_rate": 1.9975031210986267e-05,
|
| 2282 |
+
"loss": 0.0018,
|
| 2283 |
+
"step": 2850
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 31.01573033707865,
|
| 2287 |
+
"grad_norm": 0.033657535910606384,
|
| 2288 |
+
"learning_rate": 1.9850187265917604e-05,
|
| 2289 |
+
"loss": 0.0016,
|
| 2290 |
+
"step": 2860
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 31.017977528089887,
|
| 2294 |
+
"grad_norm": 0.030874161049723625,
|
| 2295 |
+
"learning_rate": 1.972534332084894e-05,
|
| 2296 |
+
"loss": 0.029,
|
| 2297 |
+
"step": 2870
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 31.020224719101122,
|
| 2301 |
+
"grad_norm": 0.025281598791480064,
|
| 2302 |
+
"learning_rate": 1.9600499375780275e-05,
|
| 2303 |
+
"loss": 0.0015,
|
| 2304 |
+
"step": 2880
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 31.020224719101122,
|
| 2308 |
+
"eval_accuracy": 0.49805950840879687,
|
| 2309 |
+
"eval_loss": 3.789454221725464,
|
| 2310 |
+
"eval_runtime": 352.055,
|
| 2311 |
+
"eval_samples_per_second": 2.196,
|
| 2312 |
+
"eval_steps_per_second": 0.139,
|
| 2313 |
+
"step": 2880
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 32.002247191011236,
|
| 2317 |
+
"grad_norm": 0.02349044755101204,
|
| 2318 |
+
"learning_rate": 1.9475655430711613e-05,
|
| 2319 |
+
"loss": 0.0286,
|
| 2320 |
+
"step": 2890
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 32.00449438202247,
|
| 2324 |
+
"grad_norm": 3.3046371936798096,
|
| 2325 |
+
"learning_rate": 1.9350811485642947e-05,
|
| 2326 |
+
"loss": 0.0651,
|
| 2327 |
+
"step": 2900
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 32.00674157303371,
|
| 2331 |
+
"grad_norm": 1.4021590948104858,
|
| 2332 |
+
"learning_rate": 1.9225967540574284e-05,
|
| 2333 |
+
"loss": 0.045,
|
| 2334 |
+
"step": 2910
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 32.00898876404494,
|
| 2338 |
+
"grad_norm": 0.04723320156335831,
|
| 2339 |
+
"learning_rate": 1.9101123595505618e-05,
|
| 2340 |
+
"loss": 0.0022,
|
| 2341 |
+
"step": 2920
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 32.01123595505618,
|
| 2345 |
+
"grad_norm": 31.909048080444336,
|
| 2346 |
+
"learning_rate": 1.8976279650436955e-05,
|
| 2347 |
+
"loss": 0.0131,
|
| 2348 |
+
"step": 2930
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 32.013483146067415,
|
| 2352 |
+
"grad_norm": 0.03553476184606552,
|
| 2353 |
+
"learning_rate": 1.885143570536829e-05,
|
| 2354 |
+
"loss": 0.0015,
|
| 2355 |
+
"step": 2940
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 32.01573033707865,
|
| 2359 |
+
"grad_norm": 0.018578186631202698,
|
| 2360 |
+
"learning_rate": 1.8726591760299626e-05,
|
| 2361 |
+
"loss": 0.0597,
|
| 2362 |
+
"step": 2950
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 32.01797752808989,
|
| 2366 |
+
"grad_norm": 0.015782801434397697,
|
| 2367 |
+
"learning_rate": 1.8601747815230963e-05,
|
| 2368 |
+
"loss": 0.0008,
|
| 2369 |
+
"step": 2960
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 32.02022471910112,
|
| 2373 |
+
"grad_norm": 0.037042029201984406,
|
| 2374 |
+
"learning_rate": 1.8476903870162297e-05,
|
| 2375 |
+
"loss": 0.0344,
|
| 2376 |
+
"step": 2970
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 32.02022471910112,
|
| 2380 |
+
"eval_accuracy": 0.4514877102199224,
|
| 2381 |
+
"eval_loss": 2.8992700576782227,
|
| 2382 |
+
"eval_runtime": 297.1202,
|
| 2383 |
+
"eval_samples_per_second": 2.602,
|
| 2384 |
+
"eval_steps_per_second": 0.165,
|
| 2385 |
+
"step": 2970
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 33.002247191011236,
|
| 2389 |
+
"grad_norm": 0.01591086946427822,
|
| 2390 |
+
"learning_rate": 1.8352059925093635e-05,
|
| 2391 |
+
"loss": 0.0011,
|
| 2392 |
+
"step": 2980
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 33.00449438202247,
|
| 2396 |
+
"grad_norm": 0.022745437920093536,
|
| 2397 |
+
"learning_rate": 1.822721598002497e-05,
|
| 2398 |
+
"loss": 0.0024,
|
| 2399 |
+
"step": 2990
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 33.00674157303371,
|
| 2403 |
+
"grad_norm": 0.019866686314344406,
|
| 2404 |
+
"learning_rate": 1.8102372034956306e-05,
|
| 2405 |
+
"loss": 0.0007,
|
| 2406 |
+
"step": 3000
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 33.00898876404494,
|
| 2410 |
+
"grad_norm": 0.0220208577811718,
|
| 2411 |
+
"learning_rate": 1.797752808988764e-05,
|
| 2412 |
+
"loss": 0.073,
|
| 2413 |
+
"step": 3010
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 33.01123595505618,
|
| 2417 |
+
"grad_norm": 0.05724029242992401,
|
| 2418 |
+
"learning_rate": 1.7852684144818977e-05,
|
| 2419 |
+
"loss": 0.0327,
|
| 2420 |
+
"step": 3020
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 33.013483146067415,
|
| 2424 |
+
"grad_norm": 0.06116447225213051,
|
| 2425 |
+
"learning_rate": 1.772784019975031e-05,
|
| 2426 |
+
"loss": 0.0046,
|
| 2427 |
+
"step": 3030
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 33.01573033707865,
|
| 2431 |
+
"grad_norm": 1.6813271045684814,
|
| 2432 |
+
"learning_rate": 1.760299625468165e-05,
|
| 2433 |
+
"loss": 0.0632,
|
| 2434 |
+
"step": 3040
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 33.01797752808989,
|
| 2438 |
+
"grad_norm": 0.07180715352296829,
|
| 2439 |
+
"learning_rate": 1.7478152309612985e-05,
|
| 2440 |
+
"loss": 0.003,
|
| 2441 |
+
"step": 3050
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 33.02022471910112,
|
| 2445 |
+
"grad_norm": 0.04319235309958458,
|
| 2446 |
+
"learning_rate": 1.7353308364544323e-05,
|
| 2447 |
+
"loss": 0.0021,
|
| 2448 |
+
"step": 3060
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 33.02022471910112,
|
| 2452 |
+
"eval_accuracy": 0.37257438551099614,
|
| 2453 |
+
"eval_loss": 4.647999286651611,
|
| 2454 |
+
"eval_runtime": 303.1066,
|
| 2455 |
+
"eval_samples_per_second": 2.55,
|
| 2456 |
+
"eval_steps_per_second": 0.162,
|
| 2457 |
+
"step": 3060
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 34.002247191011236,
|
| 2461 |
+
"grad_norm": 0.013276039622724056,
|
| 2462 |
+
"learning_rate": 1.7228464419475657e-05,
|
| 2463 |
+
"loss": 0.0017,
|
| 2464 |
+
"step": 3070
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 34.00449438202247,
|
| 2468 |
+
"grad_norm": 0.04157795011997223,
|
| 2469 |
+
"learning_rate": 1.7103620474406994e-05,
|
| 2470 |
+
"loss": 0.029,
|
| 2471 |
+
"step": 3080
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 34.00674157303371,
|
| 2475 |
+
"grad_norm": 0.049296777695417404,
|
| 2476 |
+
"learning_rate": 1.6978776529338328e-05,
|
| 2477 |
+
"loss": 0.0014,
|
| 2478 |
+
"step": 3090
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 34.00898876404494,
|
| 2482 |
+
"grad_norm": 0.02109871432185173,
|
| 2483 |
+
"learning_rate": 1.6853932584269665e-05,
|
| 2484 |
+
"loss": 0.0184,
|
| 2485 |
+
"step": 3100
|
| 2486 |
+
},
|
| 2487 |
+
{
|
| 2488 |
+
"epoch": 34.01123595505618,
|
| 2489 |
+
"grad_norm": 0.04515732452273369,
|
| 2490 |
+
"learning_rate": 1.6729088639201e-05,
|
| 2491 |
+
"loss": 0.0318,
|
| 2492 |
+
"step": 3110
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"epoch": 34.013483146067415,
|
| 2496 |
+
"grad_norm": 0.0058238692581653595,
|
| 2497 |
+
"learning_rate": 1.6604244694132333e-05,
|
| 2498 |
+
"loss": 0.0385,
|
| 2499 |
+
"step": 3120
|
| 2500 |
+
},
|
| 2501 |
+
{
|
| 2502 |
+
"epoch": 34.01573033707865,
|
| 2503 |
+
"grad_norm": 0.01912079006433487,
|
| 2504 |
+
"learning_rate": 1.647940074906367e-05,
|
| 2505 |
+
"loss": 0.0007,
|
| 2506 |
+
"step": 3130
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 34.01797752808989,
|
| 2510 |
+
"grad_norm": 0.03223242983222008,
|
| 2511 |
+
"learning_rate": 1.6354556803995007e-05,
|
| 2512 |
+
"loss": 0.0351,
|
| 2513 |
+
"step": 3140
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 34.02022471910112,
|
| 2517 |
+
"grad_norm": 0.015201385132968426,
|
| 2518 |
+
"learning_rate": 1.6229712858926345e-05,
|
| 2519 |
+
"loss": 0.0103,
|
| 2520 |
+
"step": 3150
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 34.02022471910112,
|
| 2524 |
+
"eval_accuracy": 0.3648124191461837,
|
| 2525 |
+
"eval_loss": 6.121939659118652,
|
| 2526 |
+
"eval_runtime": 396.7515,
|
| 2527 |
+
"eval_samples_per_second": 1.948,
|
| 2528 |
+
"eval_steps_per_second": 0.124,
|
| 2529 |
+
"step": 3150
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 35.002247191011236,
|
| 2533 |
+
"grad_norm": 0.005879187490791082,
|
| 2534 |
+
"learning_rate": 1.610486891385768e-05,
|
| 2535 |
+
"loss": 0.0007,
|
| 2536 |
+
"step": 3160
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 35.00449438202247,
|
| 2540 |
+
"grad_norm": 0.011820907704532146,
|
| 2541 |
+
"learning_rate": 1.5980024968789016e-05,
|
| 2542 |
+
"loss": 0.0406,
|
| 2543 |
+
"step": 3170
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 35.00674157303371,
|
| 2547 |
+
"grad_norm": 0.014389347285032272,
|
| 2548 |
+
"learning_rate": 1.585518102372035e-05,
|
| 2549 |
+
"loss": 0.0009,
|
| 2550 |
+
"step": 3180
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 35.00898876404494,
|
| 2554 |
+
"grad_norm": 0.015713289380073547,
|
| 2555 |
+
"learning_rate": 1.5730337078651687e-05,
|
| 2556 |
+
"loss": 0.0458,
|
| 2557 |
+
"step": 3190
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 35.01123595505618,
|
| 2561 |
+
"grad_norm": 0.025565631687641144,
|
| 2562 |
+
"learning_rate": 1.560549313358302e-05,
|
| 2563 |
+
"loss": 0.0009,
|
| 2564 |
+
"step": 3200
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 35.013483146067415,
|
| 2568 |
+
"grad_norm": 0.02846636064350605,
|
| 2569 |
+
"learning_rate": 1.5480649188514358e-05,
|
| 2570 |
+
"loss": 0.0311,
|
| 2571 |
+
"step": 3210
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 35.01573033707865,
|
| 2575 |
+
"grad_norm": 0.02792159467935562,
|
| 2576 |
+
"learning_rate": 1.5355805243445692e-05,
|
| 2577 |
+
"loss": 0.0029,
|
| 2578 |
+
"step": 3220
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 35.01797752808989,
|
| 2582 |
+
"grad_norm": 0.013648627325892448,
|
| 2583 |
+
"learning_rate": 1.523096129837703e-05,
|
| 2584 |
+
"loss": 0.0007,
|
| 2585 |
+
"step": 3230
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 35.02022471910112,
|
| 2589 |
+
"grad_norm": 0.07899946719408035,
|
| 2590 |
+
"learning_rate": 1.5106117353308366e-05,
|
| 2591 |
+
"loss": 0.039,
|
| 2592 |
+
"step": 3240
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 35.02022471910112,
|
| 2596 |
+
"eval_accuracy": 0.3648124191461837,
|
| 2597 |
+
"eval_loss": 6.1177849769592285,
|
| 2598 |
+
"eval_runtime": 306.2905,
|
| 2599 |
+
"eval_samples_per_second": 2.524,
|
| 2600 |
+
"eval_steps_per_second": 0.16,
|
| 2601 |
+
"step": 3240
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 36.002247191011236,
|
| 2605 |
+
"grad_norm": 0.049049124121665955,
|
| 2606 |
+
"learning_rate": 1.4981273408239702e-05,
|
| 2607 |
+
"loss": 0.0016,
|
| 2608 |
+
"step": 3250
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 36.00449438202247,
|
| 2612 |
+
"grad_norm": 0.05747610330581665,
|
| 2613 |
+
"learning_rate": 1.4856429463171038e-05,
|
| 2614 |
+
"loss": 0.0299,
|
| 2615 |
+
"step": 3260
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 36.00674157303371,
|
| 2619 |
+
"grad_norm": 0.03523377701640129,
|
| 2620 |
+
"learning_rate": 1.4731585518102373e-05,
|
| 2621 |
+
"loss": 0.0015,
|
| 2622 |
+
"step": 3270
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 36.00898876404494,
|
| 2626 |
+
"grad_norm": 0.03630174323916435,
|
| 2627 |
+
"learning_rate": 1.4606741573033709e-05,
|
| 2628 |
+
"loss": 0.0677,
|
| 2629 |
+
"step": 3280
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 36.01123595505618,
|
| 2633 |
+
"grad_norm": 0.03166479244828224,
|
| 2634 |
+
"learning_rate": 1.4481897627965044e-05,
|
| 2635 |
+
"loss": 0.0041,
|
| 2636 |
+
"step": 3290
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 36.013483146067415,
|
| 2640 |
+
"grad_norm": 0.03983873128890991,
|
| 2641 |
+
"learning_rate": 1.435705368289638e-05,
|
| 2642 |
+
"loss": 0.0016,
|
| 2643 |
+
"step": 3300
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 36.01573033707865,
|
| 2647 |
+
"grad_norm": 0.015312553383409977,
|
| 2648 |
+
"learning_rate": 1.4232209737827715e-05,
|
| 2649 |
+
"loss": 0.0009,
|
| 2650 |
+
"step": 3310
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 36.01797752808989,
|
| 2654 |
+
"grad_norm": 0.009930821135640144,
|
| 2655 |
+
"learning_rate": 1.4107365792759053e-05,
|
| 2656 |
+
"loss": 0.014,
|
| 2657 |
+
"step": 3320
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 36.02022471910112,
|
| 2661 |
+
"grad_norm": 0.010043827816843987,
|
| 2662 |
+
"learning_rate": 1.3982521847690388e-05,
|
| 2663 |
+
"loss": 0.0363,
|
| 2664 |
+
"step": 3330
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 36.02022471910112,
|
| 2668 |
+
"eval_accuracy": 0.3648124191461837,
|
| 2669 |
+
"eval_loss": 5.580588340759277,
|
| 2670 |
+
"eval_runtime": 300.5478,
|
| 2671 |
+
"eval_samples_per_second": 2.572,
|
| 2672 |
+
"eval_steps_per_second": 0.163,
|
| 2673 |
+
"step": 3330
|
| 2674 |
+
},
|
| 2675 |
+
{
|
| 2676 |
+
"epoch": 37.002247191011236,
|
| 2677 |
+
"grad_norm": 0.04191559925675392,
|
| 2678 |
+
"learning_rate": 1.3857677902621724e-05,
|
| 2679 |
+
"loss": 0.0012,
|
| 2680 |
+
"step": 3340
|
| 2681 |
+
},
|
| 2682 |
+
{
|
| 2683 |
+
"epoch": 37.00449438202247,
|
| 2684 |
+
"grad_norm": 0.02907818742096424,
|
| 2685 |
+
"learning_rate": 1.373283395755306e-05,
|
| 2686 |
+
"loss": 0.0243,
|
| 2687 |
+
"step": 3350
|
| 2688 |
+
},
|
| 2689 |
+
{
|
| 2690 |
+
"epoch": 37.00674157303371,
|
| 2691 |
+
"grad_norm": 0.009355819784104824,
|
| 2692 |
+
"learning_rate": 1.3607990012484395e-05,
|
| 2693 |
+
"loss": 0.0356,
|
| 2694 |
+
"step": 3360
|
| 2695 |
+
},
|
| 2696 |
+
{
|
| 2697 |
+
"epoch": 37.00898876404494,
|
| 2698 |
+
"grad_norm": 0.025420306250452995,
|
| 2699 |
+
"learning_rate": 1.348314606741573e-05,
|
| 2700 |
+
"loss": 0.0596,
|
| 2701 |
+
"step": 3370
|
| 2702 |
+
},
|
| 2703 |
+
{
|
| 2704 |
+
"epoch": 37.01123595505618,
|
| 2705 |
+
"grad_norm": 0.016435761004686356,
|
| 2706 |
+
"learning_rate": 1.3358302122347066e-05,
|
| 2707 |
+
"loss": 0.0021,
|
| 2708 |
+
"step": 3380
|
| 2709 |
+
},
|
| 2710 |
+
{
|
| 2711 |
+
"epoch": 37.013483146067415,
|
| 2712 |
+
"grad_norm": 0.012210534885525703,
|
| 2713 |
+
"learning_rate": 1.3233458177278402e-05,
|
| 2714 |
+
"loss": 0.0008,
|
| 2715 |
+
"step": 3390
|
| 2716 |
+
},
|
| 2717 |
+
{
|
| 2718 |
+
"epoch": 37.01573033707865,
|
| 2719 |
+
"grad_norm": 0.02066761627793312,
|
| 2720 |
+
"learning_rate": 1.3108614232209737e-05,
|
| 2721 |
+
"loss": 0.0009,
|
| 2722 |
+
"step": 3400
|
| 2723 |
+
},
|
| 2724 |
+
{
|
| 2725 |
+
"epoch": 37.01797752808989,
|
| 2726 |
+
"grad_norm": 0.02623753249645233,
|
| 2727 |
+
"learning_rate": 1.2983770287141076e-05,
|
| 2728 |
+
"loss": 0.035,
|
| 2729 |
+
"step": 3410
|
| 2730 |
+
},
|
| 2731 |
+
{
|
| 2732 |
+
"epoch": 37.02022471910112,
|
| 2733 |
+
"grad_norm": 0.027239438146352768,
|
| 2734 |
+
"learning_rate": 1.285892634207241e-05,
|
| 2735 |
+
"loss": 0.001,
|
| 2736 |
+
"step": 3420
|
| 2737 |
+
},
|
| 2738 |
+
{
|
| 2739 |
+
"epoch": 37.02022471910112,
|
| 2740 |
+
"eval_accuracy": 0.6222509702457956,
|
| 2741 |
+
"eval_loss": 3.0043742656707764,
|
| 2742 |
+
"eval_runtime": 294.5119,
|
| 2743 |
+
"eval_samples_per_second": 2.625,
|
| 2744 |
+
"eval_steps_per_second": 0.166,
|
| 2745 |
+
"step": 3420
|
| 2746 |
+
},
|
| 2747 |
+
{
|
| 2748 |
+
"epoch": 38.002247191011236,
|
| 2749 |
+
"grad_norm": 0.02889459766447544,
|
| 2750 |
+
"learning_rate": 1.2734082397003746e-05,
|
| 2751 |
+
"loss": 0.0011,
|
| 2752 |
+
"step": 3430
|
| 2753 |
+
},
|
| 2754 |
+
{
|
| 2755 |
+
"epoch": 38.00449438202247,
|
| 2756 |
+
"grad_norm": 0.04096652567386627,
|
| 2757 |
+
"learning_rate": 1.2609238451935081e-05,
|
| 2758 |
+
"loss": 0.0251,
|
| 2759 |
+
"step": 3440
|
| 2760 |
+
},
|
| 2761 |
+
{
|
| 2762 |
+
"epoch": 38.00674157303371,
|
| 2763 |
+
"grad_norm": 0.04747606813907623,
|
| 2764 |
+
"learning_rate": 1.2484394506866417e-05,
|
| 2765 |
+
"loss": 0.0378,
|
| 2766 |
+
"step": 3450
|
| 2767 |
+
},
|
| 2768 |
+
{
|
| 2769 |
+
"epoch": 38.00898876404494,
|
| 2770 |
+
"grad_norm": 0.03352310508489609,
|
| 2771 |
+
"learning_rate": 1.2359550561797752e-05,
|
| 2772 |
+
"loss": 0.0022,
|
| 2773 |
+
"step": 3460
|
| 2774 |
+
},
|
| 2775 |
+
{
|
| 2776 |
+
"epoch": 38.01123595505618,
|
| 2777 |
+
"grad_norm": 0.06360885500907898,
|
| 2778 |
+
"learning_rate": 1.2234706616729088e-05,
|
| 2779 |
+
"loss": 0.0017,
|
| 2780 |
+
"step": 3470
|
| 2781 |
+
},
|
| 2782 |
+
{
|
| 2783 |
+
"epoch": 38.013483146067415,
|
| 2784 |
+
"grad_norm": 1.631386637687683,
|
| 2785 |
+
"learning_rate": 1.2109862671660425e-05,
|
| 2786 |
+
"loss": 0.032,
|
| 2787 |
+
"step": 3480
|
| 2788 |
+
},
|
| 2789 |
+
{
|
| 2790 |
+
"epoch": 38.01573033707865,
|
| 2791 |
+
"grad_norm": 1.7111048698425293,
|
| 2792 |
+
"learning_rate": 1.1985018726591761e-05,
|
| 2793 |
+
"loss": 0.0298,
|
| 2794 |
+
"step": 3490
|
| 2795 |
+
},
|
| 2796 |
+
{
|
| 2797 |
+
"epoch": 38.01797752808989,
|
| 2798 |
+
"grad_norm": 0.041028641164302826,
|
| 2799 |
+
"learning_rate": 1.1860174781523096e-05,
|
| 2800 |
+
"loss": 0.0092,
|
| 2801 |
+
"step": 3500
|
| 2802 |
+
},
|
| 2803 |
+
{
|
| 2804 |
+
"epoch": 38.02022471910112,
|
| 2805 |
+
"grad_norm": 0.016466397792100906,
|
| 2806 |
+
"learning_rate": 1.1735330836454432e-05,
|
| 2807 |
+
"loss": 0.0031,
|
| 2808 |
+
"step": 3510
|
| 2809 |
+
},
|
| 2810 |
+
{
|
| 2811 |
+
"epoch": 38.02022471910112,
|
| 2812 |
+
"eval_accuracy": 0.34023285899094435,
|
| 2813 |
+
"eval_loss": 5.574769020080566,
|
| 2814 |
+
"eval_runtime": 344.7805,
|
| 2815 |
+
"eval_samples_per_second": 2.242,
|
| 2816 |
+
"eval_steps_per_second": 0.142,
|
| 2817 |
+
"step": 3510
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 39.002247191011236,
|
| 2821 |
+
"grad_norm": 0.05005490779876709,
|
| 2822 |
+
"learning_rate": 1.161048689138577e-05,
|
| 2823 |
+
"loss": 0.0013,
|
| 2824 |
+
"step": 3520
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 39.00449438202247,
|
| 2828 |
+
"grad_norm": 0.02787802182137966,
|
| 2829 |
+
"learning_rate": 1.1485642946317105e-05,
|
| 2830 |
+
"loss": 0.0012,
|
| 2831 |
+
"step": 3530
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 39.00674157303371,
|
| 2835 |
+
"grad_norm": 0.00706561328843236,
|
| 2836 |
+
"learning_rate": 1.136079900124844e-05,
|
| 2837 |
+
"loss": 0.0006,
|
| 2838 |
+
"step": 3540
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 39.00898876404494,
|
| 2842 |
+
"grad_norm": 0.007546106819063425,
|
| 2843 |
+
"learning_rate": 1.1235955056179776e-05,
|
| 2844 |
+
"loss": 0.0004,
|
| 2845 |
+
"step": 3550
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 39.01123595505618,
|
| 2849 |
+
"grad_norm": 0.022885961458086967,
|
| 2850 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 2851 |
+
"loss": 0.0004,
|
| 2852 |
+
"step": 3560
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 39.013483146067415,
|
| 2856 |
+
"grad_norm": 0.010719657875597477,
|
| 2857 |
+
"learning_rate": 1.0986267166042447e-05,
|
| 2858 |
+
"loss": 0.025,
|
| 2859 |
+
"step": 3570
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 39.01573033707865,
|
| 2863 |
+
"grad_norm": 0.0037333795335143805,
|
| 2864 |
+
"learning_rate": 1.0861423220973783e-05,
|
| 2865 |
+
"loss": 0.0264,
|
| 2866 |
+
"step": 3580
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 39.01797752808989,
|
| 2870 |
+
"grad_norm": 0.015252028591930866,
|
| 2871 |
+
"learning_rate": 1.0736579275905118e-05,
|
| 2872 |
+
"loss": 0.0004,
|
| 2873 |
+
"step": 3590
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 39.02022471910112,
|
| 2877 |
+
"grad_norm": 0.03425499051809311,
|
| 2878 |
+
"learning_rate": 1.0611735330836454e-05,
|
| 2879 |
+
"loss": 0.0507,
|
| 2880 |
+
"step": 3600
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 39.02022471910112,
|
| 2884 |
+
"eval_accuracy": 0.3648124191461837,
|
| 2885 |
+
"eval_loss": 5.198662757873535,
|
| 2886 |
+
"eval_runtime": 300.7511,
|
| 2887 |
+
"eval_samples_per_second": 2.57,
|
| 2888 |
+
"eval_steps_per_second": 0.163,
|
| 2889 |
+
"step": 3600
|
| 2890 |
+
},
|
| 2891 |
+
{
|
| 2892 |
+
"epoch": 40.002247191011236,
|
| 2893 |
+
"grad_norm": 0.12242499738931656,
|
| 2894 |
+
"learning_rate": 1.0486891385767791e-05,
|
| 2895 |
+
"loss": 0.001,
|
| 2896 |
+
"step": 3610
|
| 2897 |
+
},
|
| 2898 |
+
{
|
| 2899 |
+
"epoch": 40.00449438202247,
|
| 2900 |
+
"grad_norm": 0.4021182060241699,
|
| 2901 |
+
"learning_rate": 1.0362047440699127e-05,
|
| 2902 |
+
"loss": 0.0391,
|
| 2903 |
+
"step": 3620
|
| 2904 |
+
},
|
| 2905 |
+
{
|
| 2906 |
+
"epoch": 40.00674157303371,
|
| 2907 |
+
"grad_norm": 0.005329201929271221,
|
| 2908 |
+
"learning_rate": 1.0237203495630462e-05,
|
| 2909 |
+
"loss": 0.0025,
|
| 2910 |
+
"step": 3630
|
| 2911 |
+
},
|
| 2912 |
+
{
|
| 2913 |
+
"epoch": 40.00898876404494,
|
| 2914 |
+
"grad_norm": 0.1311003714799881,
|
| 2915 |
+
"learning_rate": 1.0112359550561798e-05,
|
| 2916 |
+
"loss": 0.0008,
|
| 2917 |
+
"step": 3640
|
| 2918 |
+
},
|
| 2919 |
+
{
|
| 2920 |
+
"epoch": 40.01123595505618,
|
| 2921 |
+
"grad_norm": 0.004013449884951115,
|
| 2922 |
+
"learning_rate": 9.987515605493134e-06,
|
| 2923 |
+
"loss": 0.0007,
|
| 2924 |
+
"step": 3650
|
| 2925 |
+
},
|
| 2926 |
+
{
|
| 2927 |
+
"epoch": 40.013483146067415,
|
| 2928 |
+
"grad_norm": 3.0733349323272705,
|
| 2929 |
+
"learning_rate": 9.86267166042447e-06,
|
| 2930 |
+
"loss": 0.0373,
|
| 2931 |
+
"step": 3660
|
| 2932 |
+
},
|
| 2933 |
+
{
|
| 2934 |
+
"epoch": 40.01573033707865,
|
| 2935 |
+
"grad_norm": 0.02098255604505539,
|
| 2936 |
+
"learning_rate": 9.737827715355806e-06,
|
| 2937 |
+
"loss": 0.0005,
|
| 2938 |
+
"step": 3670
|
| 2939 |
+
},
|
| 2940 |
+
{
|
| 2941 |
+
"epoch": 40.01797752808989,
|
| 2942 |
+
"grad_norm": 0.010064397938549519,
|
| 2943 |
+
"learning_rate": 9.612983770287142e-06,
|
| 2944 |
+
"loss": 0.0007,
|
| 2945 |
+
"step": 3680
|
| 2946 |
+
},
|
| 2947 |
+
{
|
| 2948 |
+
"epoch": 40.02022471910112,
|
| 2949 |
+
"grad_norm": 0.03359581157565117,
|
| 2950 |
+
"learning_rate": 9.488139825218478e-06,
|
| 2951 |
+
"loss": 0.026,
|
| 2952 |
+
"step": 3690
|
| 2953 |
+
},
|
| 2954 |
+
{
|
| 2955 |
+
"epoch": 40.02022471910112,
|
| 2956 |
+
"eval_accuracy": 0.3648124191461837,
|
| 2957 |
+
"eval_loss": 5.082982540130615,
|
| 2958 |
+
"eval_runtime": 287.2511,
|
| 2959 |
+
"eval_samples_per_second": 2.691,
|
| 2960 |
+
"eval_steps_per_second": 0.171,
|
| 2961 |
+
"step": 3690
|
| 2962 |
+
},
|
| 2963 |
+
{
|
| 2964 |
+
"epoch": 41.002247191011236,
|
| 2965 |
+
"grad_norm": 0.0668310597538948,
|
| 2966 |
+
"learning_rate": 9.363295880149813e-06,
|
| 2967 |
+
"loss": 0.0015,
|
| 2968 |
+
"step": 3700
|
| 2969 |
+
},
|
| 2970 |
+
{
|
| 2971 |
+
"epoch": 41.00449438202247,
|
| 2972 |
+
"grad_norm": 0.02239208109676838,
|
| 2973 |
+
"learning_rate": 9.238451935081149e-06,
|
| 2974 |
+
"loss": 0.0008,
|
| 2975 |
+
"step": 3710
|
| 2976 |
+
},
|
| 2977 |
+
{
|
| 2978 |
+
"epoch": 41.00674157303371,
|
| 2979 |
+
"grad_norm": 0.014757172204554081,
|
| 2980 |
+
"learning_rate": 9.113607990012484e-06,
|
| 2981 |
+
"loss": 0.0004,
|
| 2982 |
+
"step": 3720
|
| 2983 |
+
},
|
| 2984 |
+
{
|
| 2985 |
+
"epoch": 41.00898876404494,
|
| 2986 |
+
"grad_norm": 0.014439449645578861,
|
| 2987 |
+
"learning_rate": 8.98876404494382e-06,
|
| 2988 |
+
"loss": 0.0258,
|
| 2989 |
+
"step": 3730
|
| 2990 |
+
},
|
| 2991 |
+
{
|
| 2992 |
+
"epoch": 41.01123595505618,
|
| 2993 |
+
"grad_norm": 0.02586790733039379,
|
| 2994 |
+
"learning_rate": 8.863920099875155e-06,
|
| 2995 |
+
"loss": 0.0008,
|
| 2996 |
+
"step": 3740
|
| 2997 |
+
},
|
| 2998 |
+
{
|
| 2999 |
+
"epoch": 41.013483146067415,
|
| 3000 |
+
"grad_norm": 0.01682235859334469,
|
| 3001 |
+
"learning_rate": 8.739076154806493e-06,
|
| 3002 |
+
"loss": 0.0007,
|
| 3003 |
+
"step": 3750
|
| 3004 |
+
},
|
| 3005 |
+
{
|
| 3006 |
+
"epoch": 41.01573033707865,
|
| 3007 |
+
"grad_norm": 2.420248508453369,
|
| 3008 |
+
"learning_rate": 8.614232209737828e-06,
|
| 3009 |
+
"loss": 0.0652,
|
| 3010 |
+
"step": 3760
|
| 3011 |
+
},
|
| 3012 |
+
{
|
| 3013 |
+
"epoch": 41.01797752808989,
|
| 3014 |
+
"grad_norm": 0.005543038249015808,
|
| 3015 |
+
"learning_rate": 8.489388264669164e-06,
|
| 3016 |
+
"loss": 0.001,
|
| 3017 |
+
"step": 3770
|
| 3018 |
+
},
|
| 3019 |
+
{
|
| 3020 |
+
"epoch": 41.02022471910112,
|
| 3021 |
+
"grad_norm": 0.05314994975924492,
|
| 3022 |
+
"learning_rate": 8.3645443196005e-06,
|
| 3023 |
+
"loss": 0.0037,
|
| 3024 |
+
"step": 3780
|
| 3025 |
+
},
|
| 3026 |
+
{
|
| 3027 |
+
"epoch": 41.02022471910112,
|
| 3028 |
+
"eval_accuracy": 0.36739974126778785,
|
| 3029 |
+
"eval_loss": 5.115584850311279,
|
| 3030 |
+
"eval_runtime": 303.1998,
|
| 3031 |
+
"eval_samples_per_second": 2.549,
|
| 3032 |
+
"eval_steps_per_second": 0.162,
|
| 3033 |
+
"step": 3780
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 42.002247191011236,
|
| 3037 |
+
"grad_norm": 0.009197683073580265,
|
| 3038 |
+
"learning_rate": 8.239700374531835e-06,
|
| 3039 |
+
"loss": 0.0007,
|
| 3040 |
+
"step": 3790
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 42.00449438202247,
|
| 3044 |
+
"grad_norm": 0.026024773716926575,
|
| 3045 |
+
"learning_rate": 8.114856429463172e-06,
|
| 3046 |
+
"loss": 0.0099,
|
| 3047 |
+
"step": 3800
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 42.00674157303371,
|
| 3051 |
+
"grad_norm": 0.006882408633828163,
|
| 3052 |
+
"learning_rate": 7.990012484394508e-06,
|
| 3053 |
+
"loss": 0.0005,
|
| 3054 |
+
"step": 3810
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 42.00898876404494,
|
| 3058 |
+
"grad_norm": 0.008539941161870956,
|
| 3059 |
+
"learning_rate": 7.865168539325843e-06,
|
| 3060 |
+
"loss": 0.0011,
|
| 3061 |
+
"step": 3820
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 42.01123595505618,
|
| 3065 |
+
"grad_norm": 0.010466672480106354,
|
| 3066 |
+
"learning_rate": 7.740324594257179e-06,
|
| 3067 |
+
"loss": 0.0012,
|
| 3068 |
+
"step": 3830
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 42.013483146067415,
|
| 3072 |
+
"grad_norm": 0.01654103398323059,
|
| 3073 |
+
"learning_rate": 7.615480649188515e-06,
|
| 3074 |
+
"loss": 0.0004,
|
| 3075 |
+
"step": 3840
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 42.01573033707865,
|
| 3079 |
+
"grad_norm": 0.00959163997322321,
|
| 3080 |
+
"learning_rate": 7.490636704119851e-06,
|
| 3081 |
+
"loss": 0.0006,
|
| 3082 |
+
"step": 3850
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 42.01797752808989,
|
| 3086 |
+
"grad_norm": 0.06230335310101509,
|
| 3087 |
+
"learning_rate": 7.3657927590511866e-06,
|
| 3088 |
+
"loss": 0.0005,
|
| 3089 |
+
"step": 3860
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 42.02022471910112,
|
| 3093 |
+
"grad_norm": 0.003797879908233881,
|
| 3094 |
+
"learning_rate": 7.240948813982522e-06,
|
| 3095 |
+
"loss": 0.0332,
|
| 3096 |
+
"step": 3870
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 42.02022471910112,
|
| 3100 |
+
"eval_accuracy": 0.3648124191461837,
|
| 3101 |
+
"eval_loss": 5.296960830688477,
|
| 3102 |
+
"eval_runtime": 299.3441,
|
| 3103 |
+
"eval_samples_per_second": 2.582,
|
| 3104 |
+
"eval_steps_per_second": 0.164,
|
| 3105 |
+
"step": 3870
|
| 3106 |
+
},
|
| 3107 |
+
{
|
| 3108 |
+
"epoch": 43.002247191011236,
|
| 3109 |
+
"grad_norm": 0.013983920216560364,
|
| 3110 |
+
"learning_rate": 7.116104868913858e-06,
|
| 3111 |
+
"loss": 0.0269,
|
| 3112 |
+
"step": 3880
|
| 3113 |
+
},
|
| 3114 |
+
{
|
| 3115 |
+
"epoch": 43.00449438202247,
|
| 3116 |
+
"grad_norm": 0.14841292798519135,
|
| 3117 |
+
"learning_rate": 6.991260923845194e-06,
|
| 3118 |
+
"loss": 0.0413,
|
| 3119 |
+
"step": 3890
|
| 3120 |
+
},
|
| 3121 |
+
{
|
| 3122 |
+
"epoch": 43.00674157303371,
|
| 3123 |
+
"grad_norm": 0.009824411943554878,
|
| 3124 |
+
"learning_rate": 6.86641697877653e-06,
|
| 3125 |
+
"loss": 0.0005,
|
| 3126 |
+
"step": 3900
|
| 3127 |
+
},
|
| 3128 |
+
{
|
| 3129 |
+
"epoch": 43.00898876404494,
|
| 3130 |
+
"grad_norm": 0.04055115953087807,
|
| 3131 |
+
"learning_rate": 6.741573033707865e-06,
|
| 3132 |
+
"loss": 0.0311,
|
| 3133 |
+
"step": 3910
|
| 3134 |
+
},
|
| 3135 |
+
{
|
| 3136 |
+
"epoch": 43.01123595505618,
|
| 3137 |
+
"grad_norm": 0.03246951103210449,
|
| 3138 |
+
"learning_rate": 6.616729088639201e-06,
|
| 3139 |
+
"loss": 0.0009,
|
| 3140 |
+
"step": 3920
|
| 3141 |
+
},
|
| 3142 |
+
{
|
| 3143 |
+
"epoch": 43.013483146067415,
|
| 3144 |
+
"grad_norm": 0.007103465497493744,
|
| 3145 |
+
"learning_rate": 6.491885143570538e-06,
|
| 3146 |
+
"loss": 0.0007,
|
| 3147 |
+
"step": 3930
|
| 3148 |
+
},
|
| 3149 |
+
{
|
| 3150 |
+
"epoch": 43.01573033707865,
|
| 3151 |
+
"grad_norm": 0.014741869643330574,
|
| 3152 |
+
"learning_rate": 6.367041198501873e-06,
|
| 3153 |
+
"loss": 0.0006,
|
| 3154 |
+
"step": 3940
|
| 3155 |
+
},
|
| 3156 |
+
{
|
| 3157 |
+
"epoch": 43.01797752808989,
|
| 3158 |
+
"grad_norm": 0.0032137033995240927,
|
| 3159 |
+
"learning_rate": 6.2421972534332085e-06,
|
| 3160 |
+
"loss": 0.0009,
|
| 3161 |
+
"step": 3950
|
| 3162 |
+
},
|
| 3163 |
+
{
|
| 3164 |
+
"epoch": 43.02022471910112,
|
| 3165 |
+
"grad_norm": 0.03307018056511879,
|
| 3166 |
+
"learning_rate": 6.117353308364544e-06,
|
| 3167 |
+
"loss": 0.0004,
|
| 3168 |
+
"step": 3960
|
| 3169 |
+
},
|
| 3170 |
+
{
|
| 3171 |
+
"epoch": 43.02022471910112,
|
| 3172 |
+
"eval_accuracy": 0.3648124191461837,
|
| 3173 |
+
"eval_loss": 5.138851165771484,
|
| 3174 |
+
"eval_runtime": 352.1276,
|
| 3175 |
+
"eval_samples_per_second": 2.195,
|
| 3176 |
+
"eval_steps_per_second": 0.139,
|
| 3177 |
+
"step": 3960
|
| 3178 |
+
},
|
| 3179 |
+
{
|
| 3180 |
+
"epoch": 44.002247191011236,
|
| 3181 |
+
"grad_norm": 0.027850504964590073,
|
| 3182 |
+
"learning_rate": 5.9925093632958805e-06,
|
| 3183 |
+
"loss": 0.0006,
|
| 3184 |
+
"step": 3970
|
| 3185 |
+
},
|
| 3186 |
+
{
|
| 3187 |
+
"epoch": 44.00449438202247,
|
| 3188 |
+
"grad_norm": 0.005263834726065397,
|
| 3189 |
+
"learning_rate": 5.867665418227216e-06,
|
| 3190 |
+
"loss": 0.0005,
|
| 3191 |
+
"step": 3980
|
| 3192 |
+
},
|
| 3193 |
+
{
|
| 3194 |
+
"epoch": 44.00674157303371,
|
| 3195 |
+
"grad_norm": 0.031321991235017776,
|
| 3196 |
+
"learning_rate": 5.7428214731585525e-06,
|
| 3197 |
+
"loss": 0.0004,
|
| 3198 |
+
"step": 3990
|
| 3199 |
+
},
|
| 3200 |
+
{
|
| 3201 |
+
"epoch": 44.00898876404494,
|
| 3202 |
+
"grad_norm": 0.036189004778862,
|
| 3203 |
+
"learning_rate": 5.617977528089888e-06,
|
| 3204 |
+
"loss": 0.0003,
|
| 3205 |
+
"step": 4000
|
| 3206 |
+
},
|
| 3207 |
+
{
|
| 3208 |
+
"epoch": 44.01123595505618,
|
| 3209 |
+
"grad_norm": 0.01437300629913807,
|
| 3210 |
+
"learning_rate": 5.493133583021224e-06,
|
| 3211 |
+
"loss": 0.0326,
|
| 3212 |
+
"step": 4010
|
| 3213 |
+
},
|
| 3214 |
+
{
|
| 3215 |
+
"epoch": 44.013483146067415,
|
| 3216 |
+
"grad_norm": 0.004481036216020584,
|
| 3217 |
+
"learning_rate": 5.368289637952559e-06,
|
| 3218 |
+
"loss": 0.0004,
|
| 3219 |
+
"step": 4020
|
| 3220 |
+
},
|
| 3221 |
+
{
|
| 3222 |
+
"epoch": 44.01573033707865,
|
| 3223 |
+
"grad_norm": 5.446296215057373,
|
| 3224 |
+
"learning_rate": 5.243445692883896e-06,
|
| 3225 |
+
"loss": 0.0245,
|
| 3226 |
+
"step": 4030
|
| 3227 |
+
},
|
| 3228 |
+
{
|
| 3229 |
+
"epoch": 44.01797752808989,
|
| 3230 |
+
"grad_norm": 0.03322074934840202,
|
| 3231 |
+
"learning_rate": 5.118601747815231e-06,
|
| 3232 |
+
"loss": 0.0005,
|
| 3233 |
+
"step": 4040
|
| 3234 |
+
},
|
| 3235 |
+
{
|
| 3236 |
+
"epoch": 44.02022471910112,
|
| 3237 |
+
"grad_norm": 0.01612257957458496,
|
| 3238 |
+
"learning_rate": 4.993757802746567e-06,
|
| 3239 |
+
"loss": 0.0005,
|
| 3240 |
+
"step": 4050
|
| 3241 |
+
},
|
| 3242 |
+
{
|
| 3243 |
+
"epoch": 44.02022471910112,
|
| 3244 |
+
"eval_accuracy": 0.3648124191461837,
|
| 3245 |
+
"eval_loss": 5.487203121185303,
|
| 3246 |
+
"eval_runtime": 346.1337,
|
| 3247 |
+
"eval_samples_per_second": 2.233,
|
| 3248 |
+
"eval_steps_per_second": 0.142,
|
| 3249 |
+
"step": 4050
|
| 3250 |
+
},
|
| 3251 |
+
{
|
| 3252 |
+
"epoch": 45.002247191011236,
|
| 3253 |
+
"grad_norm": 0.10249646008014679,
|
| 3254 |
+
"learning_rate": 4.868913857677903e-06,
|
| 3255 |
+
"loss": 0.0006,
|
| 3256 |
+
"step": 4060
|
| 3257 |
+
},
|
| 3258 |
+
{
|
| 3259 |
+
"epoch": 45.00449438202247,
|
| 3260 |
+
"grad_norm": 0.01965424045920372,
|
| 3261 |
+
"learning_rate": 4.744069912609239e-06,
|
| 3262 |
+
"loss": 0.0012,
|
| 3263 |
+
"step": 4070
|
| 3264 |
+
},
|
| 3265 |
+
{
|
| 3266 |
+
"epoch": 45.00674157303371,
|
| 3267 |
+
"grad_norm": 0.004558570217341185,
|
| 3268 |
+
"learning_rate": 4.619225967540574e-06,
|
| 3269 |
+
"loss": 0.0003,
|
| 3270 |
+
"step": 4080
|
| 3271 |
+
},
|
| 3272 |
+
{
|
| 3273 |
+
"epoch": 45.00898876404494,
|
| 3274 |
+
"grad_norm": 0.00845757033675909,
|
| 3275 |
+
"learning_rate": 4.49438202247191e-06,
|
| 3276 |
+
"loss": 0.0003,
|
| 3277 |
+
"step": 4090
|
| 3278 |
+
},
|
| 3279 |
+
{
|
| 3280 |
+
"epoch": 45.01123595505618,
|
| 3281 |
+
"grad_norm": 0.010018822737038136,
|
| 3282 |
+
"learning_rate": 4.369538077403246e-06,
|
| 3283 |
+
"loss": 0.0002,
|
| 3284 |
+
"step": 4100
|
| 3285 |
+
},
|
| 3286 |
+
{
|
| 3287 |
+
"epoch": 45.013483146067415,
|
| 3288 |
+
"grad_norm": 0.021851379424333572,
|
| 3289 |
+
"learning_rate": 4.244694132334582e-06,
|
| 3290 |
+
"loss": 0.0292,
|
| 3291 |
+
"step": 4110
|
| 3292 |
+
},
|
| 3293 |
+
{
|
| 3294 |
+
"epoch": 45.01573033707865,
|
| 3295 |
+
"grad_norm": 0.048974115401506424,
|
| 3296 |
+
"learning_rate": 4.1198501872659175e-06,
|
| 3297 |
+
"loss": 0.0036,
|
| 3298 |
+
"step": 4120
|
| 3299 |
+
},
|
| 3300 |
+
{
|
| 3301 |
+
"epoch": 45.01797752808989,
|
| 3302 |
+
"grad_norm": 0.004967077169567347,
|
| 3303 |
+
"learning_rate": 3.995006242197254e-06,
|
| 3304 |
+
"loss": 0.0005,
|
| 3305 |
+
"step": 4130
|
| 3306 |
+
},
|
| 3307 |
+
{
|
| 3308 |
+
"epoch": 45.02022471910112,
|
| 3309 |
+
"grad_norm": 0.002462703036144376,
|
| 3310 |
+
"learning_rate": 3.8701622971285895e-06,
|
| 3311 |
+
"loss": 0.0535,
|
| 3312 |
+
"step": 4140
|
| 3313 |
+
},
|
| 3314 |
+
{
|
| 3315 |
+
"epoch": 45.02022471910112,
|
| 3316 |
+
"eval_accuracy": 0.3648124191461837,
|
| 3317 |
+
"eval_loss": 5.3780927658081055,
|
| 3318 |
+
"eval_runtime": 289.1026,
|
| 3319 |
+
"eval_samples_per_second": 2.674,
|
| 3320 |
+
"eval_steps_per_second": 0.169,
|
| 3321 |
+
"step": 4140
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 46.002247191011236,
|
| 3325 |
+
"grad_norm": 0.00382815208286047,
|
| 3326 |
+
"learning_rate": 3.7453183520599255e-06,
|
| 3327 |
+
"loss": 0.0005,
|
| 3328 |
+
"step": 4150
|
| 3329 |
+
},
|
| 3330 |
+
{
|
| 3331 |
+
"epoch": 46.00449438202247,
|
| 3332 |
+
"grad_norm": 0.10452337563037872,
|
| 3333 |
+
"learning_rate": 3.620474406991261e-06,
|
| 3334 |
+
"loss": 0.0007,
|
| 3335 |
+
"step": 4160
|
| 3336 |
+
},
|
| 3337 |
+
{
|
| 3338 |
+
"epoch": 46.00674157303371,
|
| 3339 |
+
"grad_norm": 0.020702200010418892,
|
| 3340 |
+
"learning_rate": 3.495630461922597e-06,
|
| 3341 |
+
"loss": 0.0004,
|
| 3342 |
+
"step": 4170
|
| 3343 |
+
},
|
| 3344 |
+
{
|
| 3345 |
+
"epoch": 46.00898876404494,
|
| 3346 |
+
"grad_norm": 0.015017688274383545,
|
| 3347 |
+
"learning_rate": 3.3707865168539327e-06,
|
| 3348 |
+
"loss": 0.023,
|
| 3349 |
+
"step": 4180
|
| 3350 |
+
},
|
| 3351 |
+
{
|
| 3352 |
+
"epoch": 46.01123595505618,
|
| 3353 |
+
"grad_norm": 0.4489755928516388,
|
| 3354 |
+
"learning_rate": 3.245942571785269e-06,
|
| 3355 |
+
"loss": 0.0008,
|
| 3356 |
+
"step": 4190
|
| 3357 |
+
},
|
| 3358 |
+
{
|
| 3359 |
+
"epoch": 46.013483146067415,
|
| 3360 |
+
"grad_norm": 0.001848419662564993,
|
| 3361 |
+
"learning_rate": 3.1210986267166042e-06,
|
| 3362 |
+
"loss": 0.0004,
|
| 3363 |
+
"step": 4200
|
| 3364 |
+
},
|
| 3365 |
+
{
|
| 3366 |
+
"epoch": 46.01573033707865,
|
| 3367 |
+
"grad_norm": 0.005650599952787161,
|
| 3368 |
+
"learning_rate": 2.9962546816479402e-06,
|
| 3369 |
+
"loss": 0.0006,
|
| 3370 |
+
"step": 4210
|
| 3371 |
+
},
|
| 3372 |
+
{
|
| 3373 |
+
"epoch": 46.01797752808989,
|
| 3374 |
+
"grad_norm": 0.008271172642707825,
|
| 3375 |
+
"learning_rate": 2.8714107365792762e-06,
|
| 3376 |
+
"loss": 0.0469,
|
| 3377 |
+
"step": 4220
|
| 3378 |
+
},
|
| 3379 |
+
{
|
| 3380 |
+
"epoch": 46.02022471910112,
|
| 3381 |
+
"grad_norm": 0.014952776953577995,
|
| 3382 |
+
"learning_rate": 2.746566791510612e-06,
|
| 3383 |
+
"loss": 0.0007,
|
| 3384 |
+
"step": 4230
|
| 3385 |
+
},
|
| 3386 |
+
{
|
| 3387 |
+
"epoch": 46.02022471910112,
|
| 3388 |
+
"eval_accuracy": 0.3648124191461837,
|
| 3389 |
+
"eval_loss": 5.643406867980957,
|
| 3390 |
+
"eval_runtime": 350.3809,
|
| 3391 |
+
"eval_samples_per_second": 2.206,
|
| 3392 |
+
"eval_steps_per_second": 0.14,
|
| 3393 |
+
"step": 4230
|
| 3394 |
+
},
|
| 3395 |
+
{
|
| 3396 |
+
"epoch": 47.002247191011236,
|
| 3397 |
+
"grad_norm": 0.00561157800257206,
|
| 3398 |
+
"learning_rate": 2.621722846441948e-06,
|
| 3399 |
+
"loss": 0.0006,
|
| 3400 |
+
"step": 4240
|
| 3401 |
+
},
|
| 3402 |
+
{
|
| 3403 |
+
"epoch": 47.00449438202247,
|
| 3404 |
+
"grad_norm": 0.019782327115535736,
|
| 3405 |
+
"learning_rate": 2.4968789013732834e-06,
|
| 3406 |
+
"loss": 0.0003,
|
| 3407 |
+
"step": 4250
|
| 3408 |
+
},
|
| 3409 |
+
{
|
| 3410 |
+
"epoch": 47.00674157303371,
|
| 3411 |
+
"grad_norm": 0.01003081165254116,
|
| 3412 |
+
"learning_rate": 2.3720349563046194e-06,
|
| 3413 |
+
"loss": 0.0301,
|
| 3414 |
+
"step": 4260
|
| 3415 |
+
},
|
| 3416 |
+
{
|
| 3417 |
+
"epoch": 47.00898876404494,
|
| 3418 |
+
"grad_norm": 0.012754262425005436,
|
| 3419 |
+
"learning_rate": 2.247191011235955e-06,
|
| 3420 |
+
"loss": 0.0005,
|
| 3421 |
+
"step": 4270
|
| 3422 |
+
},
|
| 3423 |
+
{
|
| 3424 |
+
"epoch": 47.01123595505618,
|
| 3425 |
+
"grad_norm": 0.06103210523724556,
|
| 3426 |
+
"learning_rate": 2.122347066167291e-06,
|
| 3427 |
+
"loss": 0.0395,
|
| 3428 |
+
"step": 4280
|
| 3429 |
+
},
|
| 3430 |
+
{
|
| 3431 |
+
"epoch": 47.013483146067415,
|
| 3432 |
+
"grad_norm": 0.01997164450585842,
|
| 3433 |
+
"learning_rate": 1.997503121098627e-06,
|
| 3434 |
+
"loss": 0.0007,
|
| 3435 |
+
"step": 4290
|
| 3436 |
+
},
|
| 3437 |
+
{
|
| 3438 |
+
"epoch": 47.01573033707865,
|
| 3439 |
+
"grad_norm": 0.011755211278796196,
|
| 3440 |
+
"learning_rate": 1.8726591760299627e-06,
|
| 3441 |
+
"loss": 0.0003,
|
| 3442 |
+
"step": 4300
|
| 3443 |
+
},
|
| 3444 |
+
{
|
| 3445 |
+
"epoch": 47.01797752808989,
|
| 3446 |
+
"grad_norm": 0.011371107771992683,
|
| 3447 |
+
"learning_rate": 1.7478152309612985e-06,
|
| 3448 |
+
"loss": 0.0005,
|
| 3449 |
+
"step": 4310
|
| 3450 |
+
},
|
| 3451 |
+
{
|
| 3452 |
+
"epoch": 47.02022471910112,
|
| 3453 |
+
"grad_norm": 0.00812207069247961,
|
| 3454 |
+
"learning_rate": 1.6229712858926345e-06,
|
| 3455 |
+
"loss": 0.0197,
|
| 3456 |
+
"step": 4320
|
| 3457 |
+
},
|
| 3458 |
+
{
|
| 3459 |
+
"epoch": 47.02022471910112,
|
| 3460 |
+
"eval_accuracy": 0.3648124191461837,
|
| 3461 |
+
"eval_loss": 5.724650859832764,
|
| 3462 |
+
"eval_runtime": 316.4918,
|
| 3463 |
+
"eval_samples_per_second": 2.442,
|
| 3464 |
+
"eval_steps_per_second": 0.155,
|
| 3465 |
+
"step": 4320
|
| 3466 |
+
},
|
| 3467 |
+
{
|
| 3468 |
+
"epoch": 48.002247191011236,
|
| 3469 |
+
"grad_norm": 0.018794557079672813,
|
| 3470 |
+
"learning_rate": 1.4981273408239701e-06,
|
| 3471 |
+
"loss": 0.0004,
|
| 3472 |
+
"step": 4330
|
| 3473 |
+
},
|
| 3474 |
+
{
|
| 3475 |
+
"epoch": 48.00449438202247,
|
| 3476 |
+
"grad_norm": 0.012647592462599277,
|
| 3477 |
+
"learning_rate": 1.373283395755306e-06,
|
| 3478 |
+
"loss": 0.0007,
|
| 3479 |
+
"step": 4340
|
| 3480 |
+
},
|
| 3481 |
+
{
|
| 3482 |
+
"epoch": 48.00674157303371,
|
| 3483 |
+
"grad_norm": 2.9073069095611572,
|
| 3484 |
+
"learning_rate": 1.2484394506866417e-06,
|
| 3485 |
+
"loss": 0.0394,
|
| 3486 |
+
"step": 4350
|
| 3487 |
+
},
|
| 3488 |
+
{
|
| 3489 |
+
"epoch": 48.00898876404494,
|
| 3490 |
+
"grad_norm": 0.01212468184530735,
|
| 3491 |
+
"learning_rate": 1.1235955056179775e-06,
|
| 3492 |
+
"loss": 0.0009,
|
| 3493 |
+
"step": 4360
|
| 3494 |
+
},
|
| 3495 |
+
{
|
| 3496 |
+
"epoch": 48.01123595505618,
|
| 3497 |
+
"grad_norm": 0.013384783640503883,
|
| 3498 |
+
"learning_rate": 9.987515605493135e-07,
|
| 3499 |
+
"loss": 0.0005,
|
| 3500 |
+
"step": 4370
|
| 3501 |
+
},
|
| 3502 |
+
{
|
| 3503 |
+
"epoch": 48.013483146067415,
|
| 3504 |
+
"grad_norm": 0.02995358593761921,
|
| 3505 |
+
"learning_rate": 8.739076154806493e-07,
|
| 3506 |
+
"loss": 0.0006,
|
| 3507 |
+
"step": 4380
|
| 3508 |
+
},
|
| 3509 |
+
{
|
| 3510 |
+
"epoch": 48.01573033707865,
|
| 3511 |
+
"grad_norm": 0.009797775186598301,
|
| 3512 |
+
"learning_rate": 7.490636704119851e-07,
|
| 3513 |
+
"loss": 0.0009,
|
| 3514 |
+
"step": 4390
|
| 3515 |
+
},
|
| 3516 |
+
{
|
| 3517 |
+
"epoch": 48.01797752808989,
|
| 3518 |
+
"grad_norm": 0.00529641006141901,
|
| 3519 |
+
"learning_rate": 6.242197253433208e-07,
|
| 3520 |
+
"loss": 0.0006,
|
| 3521 |
+
"step": 4400
|
| 3522 |
+
},
|
| 3523 |
+
{
|
| 3524 |
+
"epoch": 48.02022471910112,
|
| 3525 |
+
"grad_norm": 0.047177501022815704,
|
| 3526 |
+
"learning_rate": 4.993757802746567e-07,
|
| 3527 |
+
"loss": 0.0008,
|
| 3528 |
+
"step": 4410
|
| 3529 |
+
},
|
| 3530 |
+
{
|
| 3531 |
+
"epoch": 48.02022471910112,
|
| 3532 |
+
"eval_accuracy": 0.3648124191461837,
|
| 3533 |
+
"eval_loss": 5.757874965667725,
|
| 3534 |
+
"eval_runtime": 341.4522,
|
| 3535 |
+
"eval_samples_per_second": 2.264,
|
| 3536 |
+
"eval_steps_per_second": 0.144,
|
| 3537 |
+
"step": 4410
|
| 3538 |
+
},
|
| 3539 |
+
{
|
| 3540 |
+
"epoch": 49.002247191011236,
|
| 3541 |
+
"grad_norm": 0.0071371011435985565,
|
| 3542 |
+
"learning_rate": 3.7453183520599253e-07,
|
| 3543 |
+
"loss": 0.0004,
|
| 3544 |
+
"step": 4420
|
| 3545 |
+
},
|
| 3546 |
+
{
|
| 3547 |
+
"epoch": 49.00449438202247,
|
| 3548 |
+
"grad_norm": 0.010511082597076893,
|
| 3549 |
+
"learning_rate": 2.4968789013732837e-07,
|
| 3550 |
+
"loss": 0.0127,
|
| 3551 |
+
"step": 4430
|
| 3552 |
+
},
|
| 3553 |
+
{
|
| 3554 |
+
"epoch": 49.00674157303371,
|
| 3555 |
+
"grad_norm": 0.006362362299114466,
|
| 3556 |
+
"learning_rate": 1.2484394506866418e-07,
|
| 3557 |
+
"loss": 0.0004,
|
| 3558 |
+
"step": 4440
|
| 3559 |
+
},
|
| 3560 |
+
{
|
| 3561 |
+
"epoch": 49.00898876404494,
|
| 3562 |
+
"grad_norm": 0.004377502948045731,
|
| 3563 |
+
"learning_rate": 0.0,
|
| 3564 |
+
"loss": 0.0007,
|
| 3565 |
+
"step": 4450
|
| 3566 |
+
},
|
| 3567 |
+
{
|
| 3568 |
+
"epoch": 49.00898876404494,
|
| 3569 |
+
"eval_accuracy": 0.3648124191461837,
|
| 3570 |
+
"eval_loss": 5.757900714874268,
|
| 3571 |
+
"eval_runtime": 297.0765,
|
| 3572 |
+
"eval_samples_per_second": 2.602,
|
| 3573 |
+
"eval_steps_per_second": 0.165,
|
| 3574 |
+
"step": 4450
|
| 3575 |
+
},
|
| 3576 |
+
{
|
| 3577 |
+
"epoch": 49.00898876404494,
|
| 3578 |
+
"step": 4450,
|
| 3579 |
+
"total_flos": 8.84145129459676e+19,
|
| 3580 |
+
"train_loss": 0.03518637677183731,
|
| 3581 |
+
"train_runtime": 60526.4933,
|
| 3582 |
+
"train_samples_per_second": 1.176,
|
| 3583 |
+
"train_steps_per_second": 0.074
|
| 3584 |
+
},
|
| 3585 |
+
{
|
| 3586 |
+
"epoch": 49.00898876404494,
|
| 3587 |
+
"eval_accuracy": 0.8771021992238034,
|
| 3588 |
+
"eval_loss": 0.9978875517845154,
|
| 3589 |
+
"eval_runtime": 293.2635,
|
| 3590 |
+
"eval_samples_per_second": 2.636,
|
| 3591 |
+
"eval_steps_per_second": 0.167,
|
| 3592 |
+
"step": 4450
|
| 3593 |
+
},
|
| 3594 |
+
{
|
| 3595 |
+
"epoch": 49.00898876404494,
|
| 3596 |
+
"eval_accuracy": 1.0,
|
| 3597 |
+
"eval_loss": 0.00010065778042189777,
|
| 3598 |
+
"eval_runtime": 340.718,
|
| 3599 |
+
"eval_samples_per_second": 2.421,
|
| 3600 |
+
"eval_steps_per_second": 0.153,
|
| 3601 |
+
"step": 4450
|
| 3602 |
+
}
|
| 3603 |
+
],
|
| 3604 |
+
"logging_steps": 10,
|
| 3605 |
+
"max_steps": 4450,
|
| 3606 |
+
"num_input_tokens_seen": 0,
|
| 3607 |
+
"num_train_epochs": 9223372036854775807,
|
| 3608 |
+
"save_steps": 500,
|
| 3609 |
+
"stateful_callbacks": {
|
| 3610 |
+
"TrainerControl": {
|
| 3611 |
+
"args": {
|
| 3612 |
+
"should_epoch_stop": false,
|
| 3613 |
+
"should_evaluate": false,
|
| 3614 |
+
"should_log": false,
|
| 3615 |
+
"should_save": true,
|
| 3616 |
+
"should_training_stop": true
|
| 3617 |
+
},
|
| 3618 |
+
"attributes": {}
|
| 3619 |
+
}
|
| 3620 |
+
},
|
| 3621 |
+
"total_flos": 8.84145129459676e+19,
|
| 3622 |
+
"train_batch_size": 16,
|
| 3623 |
+
"trial_name": null,
|
| 3624 |
+
"trial_params": null
|
| 3625 |
+
}
|