End of training
Browse files- all_results.json +8 -0
- test_results.json +8 -0
- trainer_state.json +2785 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 49.00492307692308,
|
| 3 |
+
"eval_accuracy": 0.8,
|
| 4 |
+
"eval_loss": 0.9770981073379517,
|
| 5 |
+
"eval_runtime": 13.7323,
|
| 6 |
+
"eval_samples_per_second": 3.277,
|
| 7 |
+
"eval_steps_per_second": 0.874
|
| 8 |
+
}
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 49.00492307692308,
|
| 3 |
+
"eval_accuracy": 0.8,
|
| 4 |
+
"eval_loss": 0.9770981073379517,
|
| 5 |
+
"eval_runtime": 13.7323,
|
| 6 |
+
"eval_samples_per_second": 3.277,
|
| 7 |
+
"eval_steps_per_second": 0.874
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2785 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8,
|
| 3 |
+
"best_model_checkpoint": "CTMAE-P2-V3-3G-S2/checkpoint-1452",
|
| 4 |
+
"epoch": 49.00492307692308,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3250,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.003076923076923077,
|
| 13 |
+
"grad_norm": 1.4794609546661377,
|
| 14 |
+
"learning_rate": 3.0769230769230774e-07,
|
| 15 |
+
"loss": 0.7229,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.006153846153846154,
|
| 20 |
+
"grad_norm": 3.5014188289642334,
|
| 21 |
+
"learning_rate": 6.153846153846155e-07,
|
| 22 |
+
"loss": 0.7135,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.009230769230769232,
|
| 27 |
+
"grad_norm": 2.8026018142700195,
|
| 28 |
+
"learning_rate": 9.230769230769232e-07,
|
| 29 |
+
"loss": 0.7085,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.012307692307692308,
|
| 34 |
+
"grad_norm": 2.06451678276062,
|
| 35 |
+
"learning_rate": 1.230769230769231e-06,
|
| 36 |
+
"loss": 0.686,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.015384615384615385,
|
| 41 |
+
"grad_norm": 2.966813802719116,
|
| 42 |
+
"learning_rate": 1.5384615384615387e-06,
|
| 43 |
+
"loss": 0.6659,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.018461538461538463,
|
| 48 |
+
"grad_norm": 3.384611129760742,
|
| 49 |
+
"learning_rate": 1.8461538461538465e-06,
|
| 50 |
+
"loss": 0.6262,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.020307692307692308,
|
| 55 |
+
"eval_accuracy": 0.4,
|
| 56 |
+
"eval_loss": 0.8422228693962097,
|
| 57 |
+
"eval_runtime": 16.0183,
|
| 58 |
+
"eval_samples_per_second": 2.809,
|
| 59 |
+
"eval_steps_per_second": 0.749,
|
| 60 |
+
"step": 66
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.0012307692307691,
|
| 64 |
+
"grad_norm": 2.9542248249053955,
|
| 65 |
+
"learning_rate": 2.153846153846154e-06,
|
| 66 |
+
"loss": 0.6232,
|
| 67 |
+
"step": 70
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 1.0043076923076923,
|
| 71 |
+
"grad_norm": 5.368092060089111,
|
| 72 |
+
"learning_rate": 2.461538461538462e-06,
|
| 73 |
+
"loss": 0.7361,
|
| 74 |
+
"step": 80
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.0073846153846153,
|
| 78 |
+
"grad_norm": 6.5317206382751465,
|
| 79 |
+
"learning_rate": 2.7692307692307697e-06,
|
| 80 |
+
"loss": 0.6857,
|
| 81 |
+
"step": 90
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.0104615384615385,
|
| 85 |
+
"grad_norm": 8.286185264587402,
|
| 86 |
+
"learning_rate": 3.0769230769230774e-06,
|
| 87 |
+
"loss": 0.6384,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.0135384615384615,
|
| 92 |
+
"grad_norm": 11.786270141601562,
|
| 93 |
+
"learning_rate": 3.384615384615385e-06,
|
| 94 |
+
"loss": 0.5941,
|
| 95 |
+
"step": 110
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.0166153846153847,
|
| 99 |
+
"grad_norm": 4.752856254577637,
|
| 100 |
+
"learning_rate": 3.692307692307693e-06,
|
| 101 |
+
"loss": 0.5692,
|
| 102 |
+
"step": 120
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.0196923076923077,
|
| 106 |
+
"grad_norm": 2.9264302253723145,
|
| 107 |
+
"learning_rate": 4.000000000000001e-06,
|
| 108 |
+
"loss": 0.5961,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.0203076923076924,
|
| 113 |
+
"eval_accuracy": 0.4,
|
| 114 |
+
"eval_loss": 0.917664647102356,
|
| 115 |
+
"eval_runtime": 14.9958,
|
| 116 |
+
"eval_samples_per_second": 3.001,
|
| 117 |
+
"eval_steps_per_second": 0.8,
|
| 118 |
+
"step": 132
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.0024615384615383,
|
| 122 |
+
"grad_norm": 5.976866722106934,
|
| 123 |
+
"learning_rate": 4.307692307692308e-06,
|
| 124 |
+
"loss": 0.7066,
|
| 125 |
+
"step": 140
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 2.0055384615384617,
|
| 129 |
+
"grad_norm": 3.7721126079559326,
|
| 130 |
+
"learning_rate": 4.615384615384616e-06,
|
| 131 |
+
"loss": 0.6393,
|
| 132 |
+
"step": 150
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 2.0086153846153847,
|
| 136 |
+
"grad_norm": 3.9207699298858643,
|
| 137 |
+
"learning_rate": 4.923076923076924e-06,
|
| 138 |
+
"loss": 0.54,
|
| 139 |
+
"step": 160
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 2.0116923076923077,
|
| 143 |
+
"grad_norm": 21.63241958618164,
|
| 144 |
+
"learning_rate": 5.230769230769232e-06,
|
| 145 |
+
"loss": 0.6078,
|
| 146 |
+
"step": 170
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 2.0147692307692306,
|
| 150 |
+
"grad_norm": 7.049718379974365,
|
| 151 |
+
"learning_rate": 5.538461538461539e-06,
|
| 152 |
+
"loss": 0.6306,
|
| 153 |
+
"step": 180
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.017846153846154,
|
| 157 |
+
"grad_norm": 4.662971019744873,
|
| 158 |
+
"learning_rate": 5.846153846153847e-06,
|
| 159 |
+
"loss": 0.6691,
|
| 160 |
+
"step": 190
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.0203076923076924,
|
| 164 |
+
"eval_accuracy": 0.4,
|
| 165 |
+
"eval_loss": 0.86167311668396,
|
| 166 |
+
"eval_runtime": 15.0869,
|
| 167 |
+
"eval_samples_per_second": 2.983,
|
| 168 |
+
"eval_steps_per_second": 0.795,
|
| 169 |
+
"step": 198
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 3.0006153846153847,
|
| 173 |
+
"grad_norm": 6.2993035316467285,
|
| 174 |
+
"learning_rate": 6.153846153846155e-06,
|
| 175 |
+
"loss": 0.6775,
|
| 176 |
+
"step": 200
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.0036923076923077,
|
| 180 |
+
"grad_norm": 5.143687725067139,
|
| 181 |
+
"learning_rate": 6.461538461538463e-06,
|
| 182 |
+
"loss": 0.542,
|
| 183 |
+
"step": 210
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.0067692307692306,
|
| 187 |
+
"grad_norm": 5.491978645324707,
|
| 188 |
+
"learning_rate": 6.76923076923077e-06,
|
| 189 |
+
"loss": 0.5436,
|
| 190 |
+
"step": 220
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.009846153846154,
|
| 194 |
+
"grad_norm": 5.525275707244873,
|
| 195 |
+
"learning_rate": 7.076923076923078e-06,
|
| 196 |
+
"loss": 0.5095,
|
| 197 |
+
"step": 230
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.012923076923077,
|
| 201 |
+
"grad_norm": 5.9389777183532715,
|
| 202 |
+
"learning_rate": 7.384615384615386e-06,
|
| 203 |
+
"loss": 0.685,
|
| 204 |
+
"step": 240
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.016,
|
| 208 |
+
"grad_norm": 8.1178617477417,
|
| 209 |
+
"learning_rate": 7.692307692307694e-06,
|
| 210 |
+
"loss": 0.5544,
|
| 211 |
+
"step": 250
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.019076923076923,
|
| 215 |
+
"grad_norm": 7.101919174194336,
|
| 216 |
+
"learning_rate": 8.000000000000001e-06,
|
| 217 |
+
"loss": 0.6242,
|
| 218 |
+
"step": 260
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.0203076923076924,
|
| 222 |
+
"eval_accuracy": 0.4,
|
| 223 |
+
"eval_loss": 0.9505811333656311,
|
| 224 |
+
"eval_runtime": 14.1232,
|
| 225 |
+
"eval_samples_per_second": 3.186,
|
| 226 |
+
"eval_steps_per_second": 0.85,
|
| 227 |
+
"step": 264
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 4.001846153846154,
|
| 231 |
+
"grad_norm": 2.8857269287109375,
|
| 232 |
+
"learning_rate": 8.307692307692309e-06,
|
| 233 |
+
"loss": 0.3628,
|
| 234 |
+
"step": 270
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.004923076923077,
|
| 238 |
+
"grad_norm": 6.900577068328857,
|
| 239 |
+
"learning_rate": 8.615384615384617e-06,
|
| 240 |
+
"loss": 0.9504,
|
| 241 |
+
"step": 280
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 4.008,
|
| 245 |
+
"grad_norm": 4.167863368988037,
|
| 246 |
+
"learning_rate": 8.923076923076925e-06,
|
| 247 |
+
"loss": 0.7059,
|
| 248 |
+
"step": 290
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 4.011076923076923,
|
| 252 |
+
"grad_norm": 3.5212466716766357,
|
| 253 |
+
"learning_rate": 9.230769230769232e-06,
|
| 254 |
+
"loss": 0.6493,
|
| 255 |
+
"step": 300
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 4.014153846153846,
|
| 259 |
+
"grad_norm": 12.61021614074707,
|
| 260 |
+
"learning_rate": 9.53846153846154e-06,
|
| 261 |
+
"loss": 0.5847,
|
| 262 |
+
"step": 310
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 4.017230769230769,
|
| 266 |
+
"grad_norm": 4.4552764892578125,
|
| 267 |
+
"learning_rate": 9.846153846153848e-06,
|
| 268 |
+
"loss": 0.5272,
|
| 269 |
+
"step": 320
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 4.020307692307692,
|
| 273 |
+
"grad_norm": 2.0534627437591553,
|
| 274 |
+
"learning_rate": 9.982905982905984e-06,
|
| 275 |
+
"loss": 0.631,
|
| 276 |
+
"step": 330
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 4.020307692307692,
|
| 280 |
+
"eval_accuracy": 0.4,
|
| 281 |
+
"eval_loss": 0.7678373456001282,
|
| 282 |
+
"eval_runtime": 14.2654,
|
| 283 |
+
"eval_samples_per_second": 3.154,
|
| 284 |
+
"eval_steps_per_second": 0.841,
|
| 285 |
+
"step": 330
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 5.003076923076923,
|
| 289 |
+
"grad_norm": 9.449999809265137,
|
| 290 |
+
"learning_rate": 9.94871794871795e-06,
|
| 291 |
+
"loss": 0.4265,
|
| 292 |
+
"step": 340
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.006153846153846,
|
| 296 |
+
"grad_norm": 14.458154678344727,
|
| 297 |
+
"learning_rate": 9.914529914529915e-06,
|
| 298 |
+
"loss": 0.6222,
|
| 299 |
+
"step": 350
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 5.009230769230769,
|
| 303 |
+
"grad_norm": 10.737525939941406,
|
| 304 |
+
"learning_rate": 9.880341880341882e-06,
|
| 305 |
+
"loss": 0.4503,
|
| 306 |
+
"step": 360
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 5.012307692307692,
|
| 310 |
+
"grad_norm": 17.312559127807617,
|
| 311 |
+
"learning_rate": 9.846153846153848e-06,
|
| 312 |
+
"loss": 0.4326,
|
| 313 |
+
"step": 370
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 5.015384615384615,
|
| 317 |
+
"grad_norm": 52.598114013671875,
|
| 318 |
+
"learning_rate": 9.811965811965812e-06,
|
| 319 |
+
"loss": 0.5235,
|
| 320 |
+
"step": 380
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 5.018461538461539,
|
| 324 |
+
"grad_norm": 9.938750267028809,
|
| 325 |
+
"learning_rate": 9.777777777777779e-06,
|
| 326 |
+
"loss": 0.6747,
|
| 327 |
+
"step": 390
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 5.020307692307692,
|
| 331 |
+
"eval_accuracy": 0.6,
|
| 332 |
+
"eval_loss": 0.663204550743103,
|
| 333 |
+
"eval_runtime": 14.4114,
|
| 334 |
+
"eval_samples_per_second": 3.123,
|
| 335 |
+
"eval_steps_per_second": 0.833,
|
| 336 |
+
"step": 396
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 6.001230769230769,
|
| 340 |
+
"grad_norm": 39.090858459472656,
|
| 341 |
+
"learning_rate": 9.743589743589744e-06,
|
| 342 |
+
"loss": 0.4702,
|
| 343 |
+
"step": 400
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 6.004307692307692,
|
| 347 |
+
"grad_norm": 54.20093536376953,
|
| 348 |
+
"learning_rate": 9.70940170940171e-06,
|
| 349 |
+
"loss": 0.5863,
|
| 350 |
+
"step": 410
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 6.007384615384615,
|
| 354 |
+
"grad_norm": 19.63410186767578,
|
| 355 |
+
"learning_rate": 9.675213675213677e-06,
|
| 356 |
+
"loss": 0.383,
|
| 357 |
+
"step": 420
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 6.010461538461539,
|
| 361 |
+
"grad_norm": 29.095314025878906,
|
| 362 |
+
"learning_rate": 9.641025641025642e-06,
|
| 363 |
+
"loss": 0.7057,
|
| 364 |
+
"step": 430
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 6.013538461538461,
|
| 368 |
+
"grad_norm": 18.241823196411133,
|
| 369 |
+
"learning_rate": 9.606837606837607e-06,
|
| 370 |
+
"loss": 0.4765,
|
| 371 |
+
"step": 440
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 6.016615384615385,
|
| 375 |
+
"grad_norm": 35.08736038208008,
|
| 376 |
+
"learning_rate": 9.572649572649575e-06,
|
| 377 |
+
"loss": 0.487,
|
| 378 |
+
"step": 450
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 6.019692307692308,
|
| 382 |
+
"grad_norm": 6.303005218505859,
|
| 383 |
+
"learning_rate": 9.53846153846154e-06,
|
| 384 |
+
"loss": 0.6345,
|
| 385 |
+
"step": 460
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 6.020307692307692,
|
| 389 |
+
"eval_accuracy": 0.5555555555555556,
|
| 390 |
+
"eval_loss": 0.7090754508972168,
|
| 391 |
+
"eval_runtime": 15.0395,
|
| 392 |
+
"eval_samples_per_second": 2.992,
|
| 393 |
+
"eval_steps_per_second": 0.798,
|
| 394 |
+
"step": 462
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 7.002461538461539,
|
| 398 |
+
"grad_norm": 13.644148826599121,
|
| 399 |
+
"learning_rate": 9.504273504273504e-06,
|
| 400 |
+
"loss": 0.4836,
|
| 401 |
+
"step": 470
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 7.005538461538461,
|
| 405 |
+
"grad_norm": 12.159551620483398,
|
| 406 |
+
"learning_rate": 9.470085470085471e-06,
|
| 407 |
+
"loss": 0.5005,
|
| 408 |
+
"step": 480
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.008615384615385,
|
| 412 |
+
"grad_norm": 2.893991470336914,
|
| 413 |
+
"learning_rate": 9.435897435897436e-06,
|
| 414 |
+
"loss": 0.2963,
|
| 415 |
+
"step": 490
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 7.011692307692308,
|
| 419 |
+
"grad_norm": 13.471207618713379,
|
| 420 |
+
"learning_rate": 9.401709401709402e-06,
|
| 421 |
+
"loss": 0.5479,
|
| 422 |
+
"step": 500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 7.014769230769231,
|
| 426 |
+
"grad_norm": 14.097221374511719,
|
| 427 |
+
"learning_rate": 9.367521367521369e-06,
|
| 428 |
+
"loss": 0.5449,
|
| 429 |
+
"step": 510
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 7.017846153846154,
|
| 433 |
+
"grad_norm": 62.047916412353516,
|
| 434 |
+
"learning_rate": 9.333333333333334e-06,
|
| 435 |
+
"loss": 0.6357,
|
| 436 |
+
"step": 520
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 7.020307692307692,
|
| 440 |
+
"eval_accuracy": 0.5111111111111111,
|
| 441 |
+
"eval_loss": 1.096928358078003,
|
| 442 |
+
"eval_runtime": 13.707,
|
| 443 |
+
"eval_samples_per_second": 3.283,
|
| 444 |
+
"eval_steps_per_second": 0.875,
|
| 445 |
+
"step": 528
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 8.000615384615385,
|
| 449 |
+
"grad_norm": 4.9958038330078125,
|
| 450 |
+
"learning_rate": 9.2991452991453e-06,
|
| 451 |
+
"loss": 0.509,
|
| 452 |
+
"step": 530
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 8.003692307692308,
|
| 456 |
+
"grad_norm": 6.135051727294922,
|
| 457 |
+
"learning_rate": 9.264957264957267e-06,
|
| 458 |
+
"loss": 0.3708,
|
| 459 |
+
"step": 540
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 8.006769230769232,
|
| 463 |
+
"grad_norm": 28.816497802734375,
|
| 464 |
+
"learning_rate": 9.230769230769232e-06,
|
| 465 |
+
"loss": 1.3963,
|
| 466 |
+
"step": 550
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 8.009846153846153,
|
| 470 |
+
"grad_norm": 16.59303092956543,
|
| 471 |
+
"learning_rate": 9.196581196581196e-06,
|
| 472 |
+
"loss": 0.6271,
|
| 473 |
+
"step": 560
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 8.012923076923077,
|
| 477 |
+
"grad_norm": 11.57801628112793,
|
| 478 |
+
"learning_rate": 9.162393162393163e-06,
|
| 479 |
+
"loss": 0.4178,
|
| 480 |
+
"step": 570
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 8.016,
|
| 484 |
+
"grad_norm": 14.430103302001953,
|
| 485 |
+
"learning_rate": 9.128205128205129e-06,
|
| 486 |
+
"loss": 0.5318,
|
| 487 |
+
"step": 580
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 8.019076923076923,
|
| 491 |
+
"grad_norm": 12.58498477935791,
|
| 492 |
+
"learning_rate": 9.094017094017094e-06,
|
| 493 |
+
"loss": 0.4717,
|
| 494 |
+
"step": 590
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 8.020307692307693,
|
| 498 |
+
"eval_accuracy": 0.6444444444444445,
|
| 499 |
+
"eval_loss": 0.6718806624412537,
|
| 500 |
+
"eval_runtime": 13.729,
|
| 501 |
+
"eval_samples_per_second": 3.278,
|
| 502 |
+
"eval_steps_per_second": 0.874,
|
| 503 |
+
"step": 594
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 9.001846153846154,
|
| 507 |
+
"grad_norm": 9.91074275970459,
|
| 508 |
+
"learning_rate": 9.059829059829061e-06,
|
| 509 |
+
"loss": 0.3508,
|
| 510 |
+
"step": 600
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 9.004923076923077,
|
| 514 |
+
"grad_norm": 7.515194892883301,
|
| 515 |
+
"learning_rate": 9.025641025641027e-06,
|
| 516 |
+
"loss": 0.6135,
|
| 517 |
+
"step": 610
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 9.008,
|
| 521 |
+
"grad_norm": 17.925878524780273,
|
| 522 |
+
"learning_rate": 8.991452991452992e-06,
|
| 523 |
+
"loss": 0.3836,
|
| 524 |
+
"step": 620
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 9.011076923076923,
|
| 528 |
+
"grad_norm": 7.74843692779541,
|
| 529 |
+
"learning_rate": 8.957264957264959e-06,
|
| 530 |
+
"loss": 0.4139,
|
| 531 |
+
"step": 630
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 9.014153846153846,
|
| 535 |
+
"grad_norm": 4.378147125244141,
|
| 536 |
+
"learning_rate": 8.923076923076925e-06,
|
| 537 |
+
"loss": 0.2696,
|
| 538 |
+
"step": 640
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 9.01723076923077,
|
| 542 |
+
"grad_norm": 7.040340900421143,
|
| 543 |
+
"learning_rate": 8.888888888888888e-06,
|
| 544 |
+
"loss": 0.4402,
|
| 545 |
+
"step": 650
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 9.020307692307693,
|
| 549 |
+
"grad_norm": 18.362342834472656,
|
| 550 |
+
"learning_rate": 8.854700854700855e-06,
|
| 551 |
+
"loss": 0.827,
|
| 552 |
+
"step": 660
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 9.020307692307693,
|
| 556 |
+
"eval_accuracy": 0.7777777777777778,
|
| 557 |
+
"eval_loss": 0.5489006638526917,
|
| 558 |
+
"eval_runtime": 13.758,
|
| 559 |
+
"eval_samples_per_second": 3.271,
|
| 560 |
+
"eval_steps_per_second": 0.872,
|
| 561 |
+
"step": 660
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 10.003076923076923,
|
| 565 |
+
"grad_norm": 6.362973690032959,
|
| 566 |
+
"learning_rate": 8.820512820512821e-06,
|
| 567 |
+
"loss": 0.5009,
|
| 568 |
+
"step": 670
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 10.006153846153847,
|
| 572 |
+
"grad_norm": 7.7518510818481445,
|
| 573 |
+
"learning_rate": 8.786324786324786e-06,
|
| 574 |
+
"loss": 0.2564,
|
| 575 |
+
"step": 680
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 10.009230769230768,
|
| 579 |
+
"grad_norm": 9.082799911499023,
|
| 580 |
+
"learning_rate": 8.752136752136753e-06,
|
| 581 |
+
"loss": 0.3174,
|
| 582 |
+
"step": 690
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 10.012307692307692,
|
| 586 |
+
"grad_norm": 15.28465461730957,
|
| 587 |
+
"learning_rate": 8.717948717948719e-06,
|
| 588 |
+
"loss": 0.5038,
|
| 589 |
+
"step": 700
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 10.015384615384615,
|
| 593 |
+
"grad_norm": 39.25768280029297,
|
| 594 |
+
"learning_rate": 8.683760683760684e-06,
|
| 595 |
+
"loss": 0.5722,
|
| 596 |
+
"step": 710
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 10.018461538461539,
|
| 600 |
+
"grad_norm": 15.29712963104248,
|
| 601 |
+
"learning_rate": 8.649572649572651e-06,
|
| 602 |
+
"loss": 0.3107,
|
| 603 |
+
"step": 720
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 10.020307692307693,
|
| 607 |
+
"eval_accuracy": 0.6888888888888889,
|
| 608 |
+
"eval_loss": 0.7149955630302429,
|
| 609 |
+
"eval_runtime": 11.7609,
|
| 610 |
+
"eval_samples_per_second": 3.826,
|
| 611 |
+
"eval_steps_per_second": 1.02,
|
| 612 |
+
"step": 726
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 11.00123076923077,
|
| 616 |
+
"grad_norm": 4.504512786865234,
|
| 617 |
+
"learning_rate": 8.615384615384617e-06,
|
| 618 |
+
"loss": 0.4243,
|
| 619 |
+
"step": 730
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 11.004307692307693,
|
| 623 |
+
"grad_norm": 6.083981037139893,
|
| 624 |
+
"learning_rate": 8.58119658119658e-06,
|
| 625 |
+
"loss": 0.2924,
|
| 626 |
+
"step": 740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 11.007384615384616,
|
| 630 |
+
"grad_norm": 1.0353838205337524,
|
| 631 |
+
"learning_rate": 8.547008547008548e-06,
|
| 632 |
+
"loss": 0.5983,
|
| 633 |
+
"step": 750
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 11.010461538461538,
|
| 637 |
+
"grad_norm": 26.63617706298828,
|
| 638 |
+
"learning_rate": 8.512820512820513e-06,
|
| 639 |
+
"loss": 0.4385,
|
| 640 |
+
"step": 760
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 11.013538461538461,
|
| 644 |
+
"grad_norm": 73.77381134033203,
|
| 645 |
+
"learning_rate": 8.478632478632479e-06,
|
| 646 |
+
"loss": 0.5767,
|
| 647 |
+
"step": 770
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 11.016615384615385,
|
| 651 |
+
"grad_norm": 5.014594078063965,
|
| 652 |
+
"learning_rate": 8.444444444444446e-06,
|
| 653 |
+
"loss": 0.3618,
|
| 654 |
+
"step": 780
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 11.019692307692308,
|
| 658 |
+
"grad_norm": 10.796072006225586,
|
| 659 |
+
"learning_rate": 8.410256410256411e-06,
|
| 660 |
+
"loss": 0.3366,
|
| 661 |
+
"step": 790
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 11.020307692307693,
|
| 665 |
+
"eval_accuracy": 0.7111111111111111,
|
| 666 |
+
"eval_loss": 0.7248274087905884,
|
| 667 |
+
"eval_runtime": 13.7343,
|
| 668 |
+
"eval_samples_per_second": 3.276,
|
| 669 |
+
"eval_steps_per_second": 0.874,
|
| 670 |
+
"step": 792
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 12.002461538461539,
|
| 674 |
+
"grad_norm": 74.85016632080078,
|
| 675 |
+
"learning_rate": 8.376068376068377e-06,
|
| 676 |
+
"loss": 0.1946,
|
| 677 |
+
"step": 800
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 12.005538461538462,
|
| 681 |
+
"grad_norm": 26.65357780456543,
|
| 682 |
+
"learning_rate": 8.341880341880344e-06,
|
| 683 |
+
"loss": 0.2482,
|
| 684 |
+
"step": 810
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 12.008615384615384,
|
| 688 |
+
"grad_norm": 30.542070388793945,
|
| 689 |
+
"learning_rate": 8.307692307692309e-06,
|
| 690 |
+
"loss": 0.4699,
|
| 691 |
+
"step": 820
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 12.011692307692307,
|
| 695 |
+
"grad_norm": 1.3302332162857056,
|
| 696 |
+
"learning_rate": 8.273504273504273e-06,
|
| 697 |
+
"loss": 0.1524,
|
| 698 |
+
"step": 830
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 12.01476923076923,
|
| 702 |
+
"grad_norm": 26.43408966064453,
|
| 703 |
+
"learning_rate": 8.23931623931624e-06,
|
| 704 |
+
"loss": 0.4394,
|
| 705 |
+
"step": 840
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 12.017846153846154,
|
| 709 |
+
"grad_norm": 42.91722869873047,
|
| 710 |
+
"learning_rate": 8.205128205128205e-06,
|
| 711 |
+
"loss": 0.8919,
|
| 712 |
+
"step": 850
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 12.020307692307693,
|
| 716 |
+
"eval_accuracy": 0.7777777777777778,
|
| 717 |
+
"eval_loss": 0.6666911840438843,
|
| 718 |
+
"eval_runtime": 13.732,
|
| 719 |
+
"eval_samples_per_second": 3.277,
|
| 720 |
+
"eval_steps_per_second": 0.874,
|
| 721 |
+
"step": 858
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 13.000615384615385,
|
| 725 |
+
"grad_norm": 31.700401306152344,
|
| 726 |
+
"learning_rate": 8.17094017094017e-06,
|
| 727 |
+
"loss": 0.6633,
|
| 728 |
+
"step": 860
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 13.003692307692308,
|
| 732 |
+
"grad_norm": 0.5030388236045837,
|
| 733 |
+
"learning_rate": 8.136752136752138e-06,
|
| 734 |
+
"loss": 0.2475,
|
| 735 |
+
"step": 870
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 13.006769230769232,
|
| 739 |
+
"grad_norm": 42.06875991821289,
|
| 740 |
+
"learning_rate": 8.102564102564103e-06,
|
| 741 |
+
"loss": 0.8717,
|
| 742 |
+
"step": 880
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 13.009846153846153,
|
| 746 |
+
"grad_norm": 33.55354690551758,
|
| 747 |
+
"learning_rate": 8.068376068376069e-06,
|
| 748 |
+
"loss": 0.5353,
|
| 749 |
+
"step": 890
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 13.012923076923077,
|
| 753 |
+
"grad_norm": 36.84174728393555,
|
| 754 |
+
"learning_rate": 8.034188034188036e-06,
|
| 755 |
+
"loss": 0.3346,
|
| 756 |
+
"step": 900
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 13.016,
|
| 760 |
+
"grad_norm": 149.687255859375,
|
| 761 |
+
"learning_rate": 8.000000000000001e-06,
|
| 762 |
+
"loss": 0.7291,
|
| 763 |
+
"step": 910
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 13.019076923076923,
|
| 767 |
+
"grad_norm": 0.7351716160774231,
|
| 768 |
+
"learning_rate": 7.965811965811967e-06,
|
| 769 |
+
"loss": 0.4823,
|
| 770 |
+
"step": 920
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 13.020307692307693,
|
| 774 |
+
"eval_accuracy": 0.4222222222222222,
|
| 775 |
+
"eval_loss": 2.105036497116089,
|
| 776 |
+
"eval_runtime": 15.1948,
|
| 777 |
+
"eval_samples_per_second": 2.962,
|
| 778 |
+
"eval_steps_per_second": 0.79,
|
| 779 |
+
"step": 924
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 14.001846153846154,
|
| 783 |
+
"grad_norm": 35.5843391418457,
|
| 784 |
+
"learning_rate": 7.931623931623932e-06,
|
| 785 |
+
"loss": 0.5195,
|
| 786 |
+
"step": 930
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 14.004923076923077,
|
| 790 |
+
"grad_norm": 2.578026056289673,
|
| 791 |
+
"learning_rate": 7.897435897435898e-06,
|
| 792 |
+
"loss": 0.1636,
|
| 793 |
+
"step": 940
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 14.008,
|
| 797 |
+
"grad_norm": 21.901315689086914,
|
| 798 |
+
"learning_rate": 7.863247863247863e-06,
|
| 799 |
+
"loss": 0.384,
|
| 800 |
+
"step": 950
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 14.011076923076923,
|
| 804 |
+
"grad_norm": 51.293582916259766,
|
| 805 |
+
"learning_rate": 7.82905982905983e-06,
|
| 806 |
+
"loss": 0.5037,
|
| 807 |
+
"step": 960
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 14.014153846153846,
|
| 811 |
+
"grad_norm": 30.774192810058594,
|
| 812 |
+
"learning_rate": 7.794871794871796e-06,
|
| 813 |
+
"loss": 0.8682,
|
| 814 |
+
"step": 970
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 14.01723076923077,
|
| 818 |
+
"grad_norm": 31.879384994506836,
|
| 819 |
+
"learning_rate": 7.760683760683761e-06,
|
| 820 |
+
"loss": 0.3589,
|
| 821 |
+
"step": 980
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 14.020307692307693,
|
| 825 |
+
"grad_norm": 22.549419403076172,
|
| 826 |
+
"learning_rate": 7.726495726495728e-06,
|
| 827 |
+
"loss": 0.3742,
|
| 828 |
+
"step": 990
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 14.020307692307693,
|
| 832 |
+
"eval_accuracy": 0.5777777777777777,
|
| 833 |
+
"eval_loss": 1.0016950368881226,
|
| 834 |
+
"eval_runtime": 13.7811,
|
| 835 |
+
"eval_samples_per_second": 3.265,
|
| 836 |
+
"eval_steps_per_second": 0.871,
|
| 837 |
+
"step": 990
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 15.003076923076923,
|
| 841 |
+
"grad_norm": 0.20596639811992645,
|
| 842 |
+
"learning_rate": 7.692307692307694e-06,
|
| 843 |
+
"loss": 0.0428,
|
| 844 |
+
"step": 1000
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 15.006153846153847,
|
| 848 |
+
"grad_norm": 12.73334789276123,
|
| 849 |
+
"learning_rate": 7.658119658119659e-06,
|
| 850 |
+
"loss": 0.4527,
|
| 851 |
+
"step": 1010
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 15.009230769230768,
|
| 855 |
+
"grad_norm": 0.39726388454437256,
|
| 856 |
+
"learning_rate": 7.6239316239316244e-06,
|
| 857 |
+
"loss": 0.6545,
|
| 858 |
+
"step": 1020
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 15.012307692307692,
|
| 862 |
+
"grad_norm": 47.79050064086914,
|
| 863 |
+
"learning_rate": 7.58974358974359e-06,
|
| 864 |
+
"loss": 0.2677,
|
| 865 |
+
"step": 1030
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 15.015384615384615,
|
| 869 |
+
"grad_norm": 49.48329544067383,
|
| 870 |
+
"learning_rate": 7.555555555555556e-06,
|
| 871 |
+
"loss": 0.4924,
|
| 872 |
+
"step": 1040
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 15.018461538461539,
|
| 876 |
+
"grad_norm": 23.630064010620117,
|
| 877 |
+
"learning_rate": 7.521367521367522e-06,
|
| 878 |
+
"loss": 0.3399,
|
| 879 |
+
"step": 1050
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 15.020307692307693,
|
| 883 |
+
"eval_accuracy": 0.5555555555555556,
|
| 884 |
+
"eval_loss": 1.5679446458816528,
|
| 885 |
+
"eval_runtime": 13.7728,
|
| 886 |
+
"eval_samples_per_second": 3.267,
|
| 887 |
+
"eval_steps_per_second": 0.871,
|
| 888 |
+
"step": 1056
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 16.00123076923077,
|
| 892 |
+
"grad_norm": 1.3588582277297974,
|
| 893 |
+
"learning_rate": 7.487179487179488e-06,
|
| 894 |
+
"loss": 0.1339,
|
| 895 |
+
"step": 1060
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 16.00430769230769,
|
| 899 |
+
"grad_norm": 10.130599021911621,
|
| 900 |
+
"learning_rate": 7.452991452991454e-06,
|
| 901 |
+
"loss": 0.5679,
|
| 902 |
+
"step": 1070
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 16.007384615384616,
|
| 906 |
+
"grad_norm": 1.314917802810669,
|
| 907 |
+
"learning_rate": 7.4188034188034196e-06,
|
| 908 |
+
"loss": 0.3619,
|
| 909 |
+
"step": 1080
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 16.010461538461538,
|
| 913 |
+
"grad_norm": 24.52656364440918,
|
| 914 |
+
"learning_rate": 7.384615384615386e-06,
|
| 915 |
+
"loss": 0.1272,
|
| 916 |
+
"step": 1090
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 16.013538461538463,
|
| 920 |
+
"grad_norm": 33.17477798461914,
|
| 921 |
+
"learning_rate": 7.350427350427351e-06,
|
| 922 |
+
"loss": 0.26,
|
| 923 |
+
"step": 1100
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 16.016615384615385,
|
| 927 |
+
"grad_norm": 30.977619171142578,
|
| 928 |
+
"learning_rate": 7.316239316239317e-06,
|
| 929 |
+
"loss": 0.7437,
|
| 930 |
+
"step": 1110
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 16.019692307692306,
|
| 934 |
+
"grad_norm": 1.8153678178787231,
|
| 935 |
+
"learning_rate": 7.282051282051282e-06,
|
| 936 |
+
"loss": 0.6571,
|
| 937 |
+
"step": 1120
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 16.020307692307693,
|
| 941 |
+
"eval_accuracy": 0.6,
|
| 942 |
+
"eval_loss": 1.352131962776184,
|
| 943 |
+
"eval_runtime": 13.7467,
|
| 944 |
+
"eval_samples_per_second": 3.274,
|
| 945 |
+
"eval_steps_per_second": 0.873,
|
| 946 |
+
"step": 1122
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 17.00246153846154,
|
| 950 |
+
"grad_norm": 47.651878356933594,
|
| 951 |
+
"learning_rate": 7.247863247863248e-06,
|
| 952 |
+
"loss": 0.1583,
|
| 953 |
+
"step": 1130
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 17.00553846153846,
|
| 957 |
+
"grad_norm": 7.4866766929626465,
|
| 958 |
+
"learning_rate": 7.213675213675214e-06,
|
| 959 |
+
"loss": 0.3367,
|
| 960 |
+
"step": 1140
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 17.008615384615386,
|
| 964 |
+
"grad_norm": 61.95333480834961,
|
| 965 |
+
"learning_rate": 7.17948717948718e-06,
|
| 966 |
+
"loss": 0.5305,
|
| 967 |
+
"step": 1150
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 17.011692307692307,
|
| 971 |
+
"grad_norm": 0.09725925326347351,
|
| 972 |
+
"learning_rate": 7.145299145299146e-06,
|
| 973 |
+
"loss": 0.2062,
|
| 974 |
+
"step": 1160
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 17.014769230769232,
|
| 978 |
+
"grad_norm": 27.84596824645996,
|
| 979 |
+
"learning_rate": 7.111111111111112e-06,
|
| 980 |
+
"loss": 0.5889,
|
| 981 |
+
"step": 1170
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 17.017846153846154,
|
| 985 |
+
"grad_norm": 0.5110335350036621,
|
| 986 |
+
"learning_rate": 7.076923076923078e-06,
|
| 987 |
+
"loss": 0.2434,
|
| 988 |
+
"step": 1180
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 17.020307692307693,
|
| 992 |
+
"eval_accuracy": 0.7777777777777778,
|
| 993 |
+
"eval_loss": 0.7811753749847412,
|
| 994 |
+
"eval_runtime": 13.7851,
|
| 995 |
+
"eval_samples_per_second": 3.264,
|
| 996 |
+
"eval_steps_per_second": 0.871,
|
| 997 |
+
"step": 1188
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 18.000615384615383,
|
| 1001 |
+
"grad_norm": 4.819025993347168,
|
| 1002 |
+
"learning_rate": 7.0427350427350435e-06,
|
| 1003 |
+
"loss": 0.5103,
|
| 1004 |
+
"step": 1190
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 18.003692307692308,
|
| 1008 |
+
"grad_norm": 0.03198942542076111,
|
| 1009 |
+
"learning_rate": 7.008547008547009e-06,
|
| 1010 |
+
"loss": 0.2102,
|
| 1011 |
+
"step": 1200
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 18.00676923076923,
|
| 1015 |
+
"grad_norm": 53.40505599975586,
|
| 1016 |
+
"learning_rate": 6.974358974358974e-06,
|
| 1017 |
+
"loss": 0.7028,
|
| 1018 |
+
"step": 1210
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 18.009846153846155,
|
| 1022 |
+
"grad_norm": 0.3131262958049774,
|
| 1023 |
+
"learning_rate": 6.940170940170941e-06,
|
| 1024 |
+
"loss": 0.103,
|
| 1025 |
+
"step": 1220
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 18.012923076923077,
|
| 1029 |
+
"grad_norm": 0.027441751211881638,
|
| 1030 |
+
"learning_rate": 6.905982905982906e-06,
|
| 1031 |
+
"loss": 0.3159,
|
| 1032 |
+
"step": 1230
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 18.016,
|
| 1036 |
+
"grad_norm": 0.5056867599487305,
|
| 1037 |
+
"learning_rate": 6.871794871794872e-06,
|
| 1038 |
+
"loss": 0.1821,
|
| 1039 |
+
"step": 1240
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 18.019076923076923,
|
| 1043 |
+
"grad_norm": 0.06143081188201904,
|
| 1044 |
+
"learning_rate": 6.837606837606839e-06,
|
| 1045 |
+
"loss": 0.5967,
|
| 1046 |
+
"step": 1250
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 18.020307692307693,
|
| 1050 |
+
"eval_accuracy": 0.6888888888888889,
|
| 1051 |
+
"eval_loss": 0.9575199484825134,
|
| 1052 |
+
"eval_runtime": 13.783,
|
| 1053 |
+
"eval_samples_per_second": 3.265,
|
| 1054 |
+
"eval_steps_per_second": 0.871,
|
| 1055 |
+
"step": 1254
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 19.001846153846152,
|
| 1059 |
+
"grad_norm": 68.97899627685547,
|
| 1060 |
+
"learning_rate": 6.803418803418804e-06,
|
| 1061 |
+
"loss": 0.5041,
|
| 1062 |
+
"step": 1260
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 19.004923076923077,
|
| 1066 |
+
"grad_norm": 0.17927193641662598,
|
| 1067 |
+
"learning_rate": 6.76923076923077e-06,
|
| 1068 |
+
"loss": 0.152,
|
| 1069 |
+
"step": 1270
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 19.008,
|
| 1073 |
+
"grad_norm": 63.69683074951172,
|
| 1074 |
+
"learning_rate": 6.735042735042736e-06,
|
| 1075 |
+
"loss": 0.2777,
|
| 1076 |
+
"step": 1280
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 19.011076923076924,
|
| 1080 |
+
"grad_norm": 0.1938479095697403,
|
| 1081 |
+
"learning_rate": 6.700854700854701e-06,
|
| 1082 |
+
"loss": 0.5759,
|
| 1083 |
+
"step": 1290
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 19.014153846153846,
|
| 1087 |
+
"grad_norm": 4.396932125091553,
|
| 1088 |
+
"learning_rate": 6.666666666666667e-06,
|
| 1089 |
+
"loss": 0.0564,
|
| 1090 |
+
"step": 1300
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 19.017230769230768,
|
| 1094 |
+
"grad_norm": 0.019336098805069923,
|
| 1095 |
+
"learning_rate": 6.632478632478633e-06,
|
| 1096 |
+
"loss": 0.4901,
|
| 1097 |
+
"step": 1310
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 19.020307692307693,
|
| 1101 |
+
"grad_norm": 0.24376414716243744,
|
| 1102 |
+
"learning_rate": 6.598290598290598e-06,
|
| 1103 |
+
"loss": 0.1982,
|
| 1104 |
+
"step": 1320
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 19.020307692307693,
|
| 1108 |
+
"eval_accuracy": 0.6666666666666666,
|
| 1109 |
+
"eval_loss": 1.1720720529556274,
|
| 1110 |
+
"eval_runtime": 13.8989,
|
| 1111 |
+
"eval_samples_per_second": 3.238,
|
| 1112 |
+
"eval_steps_per_second": 0.863,
|
| 1113 |
+
"step": 1320
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 20.00307692307692,
|
| 1117 |
+
"grad_norm": 6.922430992126465,
|
| 1118 |
+
"learning_rate": 6.564102564102565e-06,
|
| 1119 |
+
"loss": 0.144,
|
| 1120 |
+
"step": 1330
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 20.006153846153847,
|
| 1124 |
+
"grad_norm": 14.44478702545166,
|
| 1125 |
+
"learning_rate": 6.529914529914531e-06,
|
| 1126 |
+
"loss": 0.1007,
|
| 1127 |
+
"step": 1340
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 20.00923076923077,
|
| 1131 |
+
"grad_norm": 95.26171112060547,
|
| 1132 |
+
"learning_rate": 6.495726495726496e-06,
|
| 1133 |
+
"loss": 0.4298,
|
| 1134 |
+
"step": 1350
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 20.012307692307694,
|
| 1138 |
+
"grad_norm": 110.86443328857422,
|
| 1139 |
+
"learning_rate": 6.461538461538463e-06,
|
| 1140 |
+
"loss": 0.1362,
|
| 1141 |
+
"step": 1360
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 20.015384615384615,
|
| 1145 |
+
"grad_norm": 0.13159264624118805,
|
| 1146 |
+
"learning_rate": 6.427350427350428e-06,
|
| 1147 |
+
"loss": 0.086,
|
| 1148 |
+
"step": 1370
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 20.018461538461537,
|
| 1152 |
+
"grad_norm": 116.35437774658203,
|
| 1153 |
+
"learning_rate": 6.3931623931623935e-06,
|
| 1154 |
+
"loss": 0.2631,
|
| 1155 |
+
"step": 1380
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 20.020307692307693,
|
| 1159 |
+
"eval_accuracy": 0.4666666666666667,
|
| 1160 |
+
"eval_loss": 2.5732531547546387,
|
| 1161 |
+
"eval_runtime": 15.2291,
|
| 1162 |
+
"eval_samples_per_second": 2.955,
|
| 1163 |
+
"eval_steps_per_second": 0.788,
|
| 1164 |
+
"step": 1386
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 21.00123076923077,
|
| 1168 |
+
"grad_norm": 0.08772929012775421,
|
| 1169 |
+
"learning_rate": 6.358974358974359e-06,
|
| 1170 |
+
"loss": 0.3654,
|
| 1171 |
+
"step": 1390
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 21.00430769230769,
|
| 1175 |
+
"grad_norm": 0.5012686848640442,
|
| 1176 |
+
"learning_rate": 6.324786324786325e-06,
|
| 1177 |
+
"loss": 0.3005,
|
| 1178 |
+
"step": 1400
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 21.007384615384616,
|
| 1182 |
+
"grad_norm": 0.3065190017223358,
|
| 1183 |
+
"learning_rate": 6.290598290598291e-06,
|
| 1184 |
+
"loss": 0.2305,
|
| 1185 |
+
"step": 1410
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 21.010461538461538,
|
| 1189 |
+
"grad_norm": 0.326849102973938,
|
| 1190 |
+
"learning_rate": 6.256410256410257e-06,
|
| 1191 |
+
"loss": 0.2465,
|
| 1192 |
+
"step": 1420
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 21.013538461538463,
|
| 1196 |
+
"grad_norm": 58.84406661987305,
|
| 1197 |
+
"learning_rate": 6.222222222222223e-06,
|
| 1198 |
+
"loss": 0.1797,
|
| 1199 |
+
"step": 1430
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 21.016615384615385,
|
| 1203 |
+
"grad_norm": 0.0164014995098114,
|
| 1204 |
+
"learning_rate": 6.188034188034189e-06,
|
| 1205 |
+
"loss": 0.1588,
|
| 1206 |
+
"step": 1440
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 21.019692307692306,
|
| 1210 |
+
"grad_norm": 34.90568923950195,
|
| 1211 |
+
"learning_rate": 6.153846153846155e-06,
|
| 1212 |
+
"loss": 0.3235,
|
| 1213 |
+
"step": 1450
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 21.020307692307693,
|
| 1217 |
+
"eval_accuracy": 0.8,
|
| 1218 |
+
"eval_loss": 0.9770981669425964,
|
| 1219 |
+
"eval_runtime": 13.7436,
|
| 1220 |
+
"eval_samples_per_second": 3.274,
|
| 1221 |
+
"eval_steps_per_second": 0.873,
|
| 1222 |
+
"step": 1452
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 22.00246153846154,
|
| 1226 |
+
"grad_norm": 0.055868543684482574,
|
| 1227 |
+
"learning_rate": 6.11965811965812e-06,
|
| 1228 |
+
"loss": 0.0287,
|
| 1229 |
+
"step": 1460
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 22.00553846153846,
|
| 1233 |
+
"grad_norm": 205.71531677246094,
|
| 1234 |
+
"learning_rate": 6.085470085470086e-06,
|
| 1235 |
+
"loss": 0.3236,
|
| 1236 |
+
"step": 1470
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 22.008615384615386,
|
| 1240 |
+
"grad_norm": 0.009756018407642841,
|
| 1241 |
+
"learning_rate": 6.051282051282051e-06,
|
| 1242 |
+
"loss": 0.311,
|
| 1243 |
+
"step": 1480
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 22.011692307692307,
|
| 1247 |
+
"grad_norm": 125.73477172851562,
|
| 1248 |
+
"learning_rate": 6.0170940170940174e-06,
|
| 1249 |
+
"loss": 0.6165,
|
| 1250 |
+
"step": 1490
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 22.014769230769232,
|
| 1254 |
+
"grad_norm": 0.32701388001441956,
|
| 1255 |
+
"learning_rate": 5.982905982905983e-06,
|
| 1256 |
+
"loss": 0.0298,
|
| 1257 |
+
"step": 1500
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 22.017846153846154,
|
| 1261 |
+
"grad_norm": 87.97734069824219,
|
| 1262 |
+
"learning_rate": 5.948717948717949e-06,
|
| 1263 |
+
"loss": 0.1786,
|
| 1264 |
+
"step": 1510
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 22.020307692307693,
|
| 1268 |
+
"eval_accuracy": 0.7111111111111111,
|
| 1269 |
+
"eval_loss": 1.1978042125701904,
|
| 1270 |
+
"eval_runtime": 13.8304,
|
| 1271 |
+
"eval_samples_per_second": 3.254,
|
| 1272 |
+
"eval_steps_per_second": 0.868,
|
| 1273 |
+
"step": 1518
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 23.000615384615383,
|
| 1277 |
+
"grad_norm": 104.00849151611328,
|
| 1278 |
+
"learning_rate": 5.914529914529915e-06,
|
| 1279 |
+
"loss": 0.1849,
|
| 1280 |
+
"step": 1520
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 23.003692307692308,
|
| 1284 |
+
"grad_norm": 5.975285053253174,
|
| 1285 |
+
"learning_rate": 5.880341880341881e-06,
|
| 1286 |
+
"loss": 0.0709,
|
| 1287 |
+
"step": 1530
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 23.00676923076923,
|
| 1291 |
+
"grad_norm": 0.09302225708961487,
|
| 1292 |
+
"learning_rate": 5.846153846153847e-06,
|
| 1293 |
+
"loss": 0.2343,
|
| 1294 |
+
"step": 1540
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 23.009846153846155,
|
| 1298 |
+
"grad_norm": 113.04926300048828,
|
| 1299 |
+
"learning_rate": 5.8119658119658126e-06,
|
| 1300 |
+
"loss": 0.506,
|
| 1301 |
+
"step": 1550
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 23.012923076923077,
|
| 1305 |
+
"grad_norm": 9.404973983764648,
|
| 1306 |
+
"learning_rate": 5.777777777777778e-06,
|
| 1307 |
+
"loss": 0.3141,
|
| 1308 |
+
"step": 1560
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 23.016,
|
| 1312 |
+
"grad_norm": 0.007651908788830042,
|
| 1313 |
+
"learning_rate": 5.743589743589743e-06,
|
| 1314 |
+
"loss": 0.7863,
|
| 1315 |
+
"step": 1570
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 23.019076923076923,
|
| 1319 |
+
"grad_norm": 3.120847225189209,
|
| 1320 |
+
"learning_rate": 5.70940170940171e-06,
|
| 1321 |
+
"loss": 0.1352,
|
| 1322 |
+
"step": 1580
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 23.020307692307693,
|
| 1326 |
+
"eval_accuracy": 0.7777777777777778,
|
| 1327 |
+
"eval_loss": 0.86919766664505,
|
| 1328 |
+
"eval_runtime": 15.1646,
|
| 1329 |
+
"eval_samples_per_second": 2.967,
|
| 1330 |
+
"eval_steps_per_second": 0.791,
|
| 1331 |
+
"step": 1584
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 24.001846153846152,
|
| 1335 |
+
"grad_norm": 0.05145508795976639,
|
| 1336 |
+
"learning_rate": 5.675213675213675e-06,
|
| 1337 |
+
"loss": 0.3145,
|
| 1338 |
+
"step": 1590
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 24.004923076923077,
|
| 1342 |
+
"grad_norm": 44.031471252441406,
|
| 1343 |
+
"learning_rate": 5.641025641025641e-06,
|
| 1344 |
+
"loss": 0.4755,
|
| 1345 |
+
"step": 1600
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 24.008,
|
| 1349 |
+
"grad_norm": 1.2100696563720703,
|
| 1350 |
+
"learning_rate": 5.606837606837608e-06,
|
| 1351 |
+
"loss": 0.2257,
|
| 1352 |
+
"step": 1610
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 24.011076923076924,
|
| 1356 |
+
"grad_norm": 38.097190856933594,
|
| 1357 |
+
"learning_rate": 5.572649572649573e-06,
|
| 1358 |
+
"loss": 0.1551,
|
| 1359 |
+
"step": 1620
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 24.014153846153846,
|
| 1363 |
+
"grad_norm": 4.676908016204834,
|
| 1364 |
+
"learning_rate": 5.538461538461539e-06,
|
| 1365 |
+
"loss": 0.1355,
|
| 1366 |
+
"step": 1630
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 24.017230769230768,
|
| 1370 |
+
"grad_norm": 45.96378707885742,
|
| 1371 |
+
"learning_rate": 5.504273504273505e-06,
|
| 1372 |
+
"loss": 0.1754,
|
| 1373 |
+
"step": 1640
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 24.020307692307693,
|
| 1377 |
+
"grad_norm": 0.0224474985152483,
|
| 1378 |
+
"learning_rate": 5.470085470085471e-06,
|
| 1379 |
+
"loss": 0.1709,
|
| 1380 |
+
"step": 1650
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 24.020307692307693,
|
| 1384 |
+
"eval_accuracy": 0.7555555555555555,
|
| 1385 |
+
"eval_loss": 1.1424057483673096,
|
| 1386 |
+
"eval_runtime": 13.7879,
|
| 1387 |
+
"eval_samples_per_second": 3.264,
|
| 1388 |
+
"eval_steps_per_second": 0.87,
|
| 1389 |
+
"step": 1650
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 25.00307692307692,
|
| 1393 |
+
"grad_norm": 0.019953293725848198,
|
| 1394 |
+
"learning_rate": 5.435897435897436e-06,
|
| 1395 |
+
"loss": 0.0806,
|
| 1396 |
+
"step": 1660
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 25.006153846153847,
|
| 1400 |
+
"grad_norm": 0.046231500804424286,
|
| 1401 |
+
"learning_rate": 5.401709401709402e-06,
|
| 1402 |
+
"loss": 0.0549,
|
| 1403 |
+
"step": 1670
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 25.00923076923077,
|
| 1407 |
+
"grad_norm": 0.0693158209323883,
|
| 1408 |
+
"learning_rate": 5.367521367521367e-06,
|
| 1409 |
+
"loss": 0.3095,
|
| 1410 |
+
"step": 1680
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 25.012307692307694,
|
| 1414 |
+
"grad_norm": 20.43744468688965,
|
| 1415 |
+
"learning_rate": 5.333333333333334e-06,
|
| 1416 |
+
"loss": 0.1346,
|
| 1417 |
+
"step": 1690
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 25.015384615384615,
|
| 1421 |
+
"grad_norm": 0.02532690390944481,
|
| 1422 |
+
"learning_rate": 5.2991452991453e-06,
|
| 1423 |
+
"loss": 0.3013,
|
| 1424 |
+
"step": 1700
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 25.018461538461537,
|
| 1428 |
+
"grad_norm": 0.1781366467475891,
|
| 1429 |
+
"learning_rate": 5.264957264957265e-06,
|
| 1430 |
+
"loss": 0.0873,
|
| 1431 |
+
"step": 1710
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 25.020307692307693,
|
| 1435 |
+
"eval_accuracy": 0.6222222222222222,
|
| 1436 |
+
"eval_loss": 1.8760292530059814,
|
| 1437 |
+
"eval_runtime": 13.807,
|
| 1438 |
+
"eval_samples_per_second": 3.259,
|
| 1439 |
+
"eval_steps_per_second": 0.869,
|
| 1440 |
+
"step": 1716
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 26.00123076923077,
|
| 1444 |
+
"grad_norm": 2.7792463302612305,
|
| 1445 |
+
"learning_rate": 5.230769230769232e-06,
|
| 1446 |
+
"loss": 0.1224,
|
| 1447 |
+
"step": 1720
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 26.00430769230769,
|
| 1451 |
+
"grad_norm": 87.59558868408203,
|
| 1452 |
+
"learning_rate": 5.196581196581197e-06,
|
| 1453 |
+
"loss": 0.2843,
|
| 1454 |
+
"step": 1730
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 26.007384615384616,
|
| 1458 |
+
"grad_norm": 8.613319396972656,
|
| 1459 |
+
"learning_rate": 5.162393162393163e-06,
|
| 1460 |
+
"loss": 0.0209,
|
| 1461 |
+
"step": 1740
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 26.010461538461538,
|
| 1465 |
+
"grad_norm": 0.06284106522798538,
|
| 1466 |
+
"learning_rate": 5.128205128205128e-06,
|
| 1467 |
+
"loss": 0.0257,
|
| 1468 |
+
"step": 1750
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 26.013538461538463,
|
| 1472 |
+
"grad_norm": 9.368754386901855,
|
| 1473 |
+
"learning_rate": 5.094017094017094e-06,
|
| 1474 |
+
"loss": 0.1877,
|
| 1475 |
+
"step": 1760
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 26.016615384615385,
|
| 1479 |
+
"grad_norm": 0.07090014964342117,
|
| 1480 |
+
"learning_rate": 5.05982905982906e-06,
|
| 1481 |
+
"loss": 0.1936,
|
| 1482 |
+
"step": 1770
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 26.019692307692306,
|
| 1486 |
+
"grad_norm": 22.935270309448242,
|
| 1487 |
+
"learning_rate": 5.025641025641026e-06,
|
| 1488 |
+
"loss": 0.1418,
|
| 1489 |
+
"step": 1780
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 26.020307692307693,
|
| 1493 |
+
"eval_accuracy": 0.8,
|
| 1494 |
+
"eval_loss": 1.0963565111160278,
|
| 1495 |
+
"eval_runtime": 14.0498,
|
| 1496 |
+
"eval_samples_per_second": 3.203,
|
| 1497 |
+
"eval_steps_per_second": 0.854,
|
| 1498 |
+
"step": 1782
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 27.00246153846154,
|
| 1502 |
+
"grad_norm": 0.096647247672081,
|
| 1503 |
+
"learning_rate": 4.991452991452992e-06,
|
| 1504 |
+
"loss": 0.0147,
|
| 1505 |
+
"step": 1790
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 27.00553846153846,
|
| 1509 |
+
"grad_norm": 0.10912120342254639,
|
| 1510 |
+
"learning_rate": 4.957264957264958e-06,
|
| 1511 |
+
"loss": 0.1364,
|
| 1512 |
+
"step": 1800
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 27.008615384615386,
|
| 1516 |
+
"grad_norm": 97.78022766113281,
|
| 1517 |
+
"learning_rate": 4.923076923076924e-06,
|
| 1518 |
+
"loss": 0.2783,
|
| 1519 |
+
"step": 1810
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 27.011692307692307,
|
| 1523 |
+
"grad_norm": 168.70310974121094,
|
| 1524 |
+
"learning_rate": 4.888888888888889e-06,
|
| 1525 |
+
"loss": 0.2077,
|
| 1526 |
+
"step": 1820
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 27.014769230769232,
|
| 1530 |
+
"grad_norm": 0.3277463912963867,
|
| 1531 |
+
"learning_rate": 4.854700854700855e-06,
|
| 1532 |
+
"loss": 0.003,
|
| 1533 |
+
"step": 1830
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 27.017846153846154,
|
| 1537 |
+
"grad_norm": 0.03846672177314758,
|
| 1538 |
+
"learning_rate": 4.820512820512821e-06,
|
| 1539 |
+
"loss": 0.0075,
|
| 1540 |
+
"step": 1840
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 27.020307692307693,
|
| 1544 |
+
"eval_accuracy": 0.6222222222222222,
|
| 1545 |
+
"eval_loss": 1.9129718542099,
|
| 1546 |
+
"eval_runtime": 15.1161,
|
| 1547 |
+
"eval_samples_per_second": 2.977,
|
| 1548 |
+
"eval_steps_per_second": 0.794,
|
| 1549 |
+
"step": 1848
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 28.000615384615383,
|
| 1553 |
+
"grad_norm": 9.841035842895508,
|
| 1554 |
+
"learning_rate": 4.786324786324787e-06,
|
| 1555 |
+
"loss": 0.0969,
|
| 1556 |
+
"step": 1850
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 28.003692307692308,
|
| 1560 |
+
"grad_norm": 37.528804779052734,
|
| 1561 |
+
"learning_rate": 4.752136752136752e-06,
|
| 1562 |
+
"loss": 0.2131,
|
| 1563 |
+
"step": 1860
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 28.00676923076923,
|
| 1567 |
+
"grad_norm": 0.033336490392684937,
|
| 1568 |
+
"learning_rate": 4.717948717948718e-06,
|
| 1569 |
+
"loss": 0.1671,
|
| 1570 |
+
"step": 1870
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 28.009846153846155,
|
| 1574 |
+
"grad_norm": 17.54317855834961,
|
| 1575 |
+
"learning_rate": 4.6837606837606844e-06,
|
| 1576 |
+
"loss": 0.2178,
|
| 1577 |
+
"step": 1880
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 28.012923076923077,
|
| 1581 |
+
"grad_norm": 0.035124845802783966,
|
| 1582 |
+
"learning_rate": 4.64957264957265e-06,
|
| 1583 |
+
"loss": 0.0193,
|
| 1584 |
+
"step": 1890
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 28.016,
|
| 1588 |
+
"grad_norm": 0.01775071956217289,
|
| 1589 |
+
"learning_rate": 4.615384615384616e-06,
|
| 1590 |
+
"loss": 0.1719,
|
| 1591 |
+
"step": 1900
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 28.019076923076923,
|
| 1595 |
+
"grad_norm": 18.559314727783203,
|
| 1596 |
+
"learning_rate": 4.581196581196582e-06,
|
| 1597 |
+
"loss": 0.4534,
|
| 1598 |
+
"step": 1910
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 28.020307692307693,
|
| 1602 |
+
"eval_accuracy": 0.7777777777777778,
|
| 1603 |
+
"eval_loss": 1.1176104545593262,
|
| 1604 |
+
"eval_runtime": 13.7703,
|
| 1605 |
+
"eval_samples_per_second": 3.268,
|
| 1606 |
+
"eval_steps_per_second": 0.871,
|
| 1607 |
+
"step": 1914
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 29.001846153846152,
|
| 1611 |
+
"grad_norm": 0.03852153941988945,
|
| 1612 |
+
"learning_rate": 4.547008547008547e-06,
|
| 1613 |
+
"loss": 0.0281,
|
| 1614 |
+
"step": 1920
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 29.004923076923077,
|
| 1618 |
+
"grad_norm": 0.040109530091285706,
|
| 1619 |
+
"learning_rate": 4.512820512820513e-06,
|
| 1620 |
+
"loss": 0.0877,
|
| 1621 |
+
"step": 1930
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 29.008,
|
| 1625 |
+
"grad_norm": 0.011689484119415283,
|
| 1626 |
+
"learning_rate": 4.4786324786324796e-06,
|
| 1627 |
+
"loss": 0.0611,
|
| 1628 |
+
"step": 1940
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 29.011076923076924,
|
| 1632 |
+
"grad_norm": 0.053006622940301895,
|
| 1633 |
+
"learning_rate": 4.444444444444444e-06,
|
| 1634 |
+
"loss": 0.1979,
|
| 1635 |
+
"step": 1950
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 29.014153846153846,
|
| 1639 |
+
"grad_norm": 0.04539922997355461,
|
| 1640 |
+
"learning_rate": 4.4102564102564104e-06,
|
| 1641 |
+
"loss": 0.218,
|
| 1642 |
+
"step": 1960
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 29.017230769230768,
|
| 1646 |
+
"grad_norm": 0.6297591328620911,
|
| 1647 |
+
"learning_rate": 4.376068376068377e-06,
|
| 1648 |
+
"loss": 0.3363,
|
| 1649 |
+
"step": 1970
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 29.020307692307693,
|
| 1653 |
+
"grad_norm": 0.01002031471580267,
|
| 1654 |
+
"learning_rate": 4.341880341880342e-06,
|
| 1655 |
+
"loss": 0.0019,
|
| 1656 |
+
"step": 1980
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 29.020307692307693,
|
| 1660 |
+
"eval_accuracy": 0.6222222222222222,
|
| 1661 |
+
"eval_loss": 2.0683937072753906,
|
| 1662 |
+
"eval_runtime": 13.7766,
|
| 1663 |
+
"eval_samples_per_second": 3.266,
|
| 1664 |
+
"eval_steps_per_second": 0.871,
|
| 1665 |
+
"step": 1980
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 30.00307692307692,
|
| 1669 |
+
"grad_norm": 8.234089851379395,
|
| 1670 |
+
"learning_rate": 4.307692307692308e-06,
|
| 1671 |
+
"loss": 0.1411,
|
| 1672 |
+
"step": 1990
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 30.006153846153847,
|
| 1676 |
+
"grad_norm": 0.18931716680526733,
|
| 1677 |
+
"learning_rate": 4.273504273504274e-06,
|
| 1678 |
+
"loss": 0.1407,
|
| 1679 |
+
"step": 2000
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 30.00923076923077,
|
| 1683 |
+
"grad_norm": 0.020825600251555443,
|
| 1684 |
+
"learning_rate": 4.239316239316239e-06,
|
| 1685 |
+
"loss": 0.0226,
|
| 1686 |
+
"step": 2010
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 30.012307692307694,
|
| 1690 |
+
"grad_norm": 0.09039674699306488,
|
| 1691 |
+
"learning_rate": 4.2051282051282055e-06,
|
| 1692 |
+
"loss": 0.0008,
|
| 1693 |
+
"step": 2020
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 30.015384615384615,
|
| 1697 |
+
"grad_norm": 0.036115147173404694,
|
| 1698 |
+
"learning_rate": 4.170940170940172e-06,
|
| 1699 |
+
"loss": 0.0254,
|
| 1700 |
+
"step": 2030
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 30.018461538461537,
|
| 1704 |
+
"grad_norm": 18.284534454345703,
|
| 1705 |
+
"learning_rate": 4.136752136752136e-06,
|
| 1706 |
+
"loss": 0.4098,
|
| 1707 |
+
"step": 2040
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 30.020307692307693,
|
| 1711 |
+
"eval_accuracy": 0.6666666666666666,
|
| 1712 |
+
"eval_loss": 1.9198042154312134,
|
| 1713 |
+
"eval_runtime": 13.8755,
|
| 1714 |
+
"eval_samples_per_second": 3.243,
|
| 1715 |
+
"eval_steps_per_second": 0.865,
|
| 1716 |
+
"step": 2046
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 31.00123076923077,
|
| 1720 |
+
"grad_norm": 0.01804875209927559,
|
| 1721 |
+
"learning_rate": 4.102564102564103e-06,
|
| 1722 |
+
"loss": 0.1426,
|
| 1723 |
+
"step": 2050
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 31.00430769230769,
|
| 1727 |
+
"grad_norm": 0.08219186961650848,
|
| 1728 |
+
"learning_rate": 4.068376068376069e-06,
|
| 1729 |
+
"loss": 0.0015,
|
| 1730 |
+
"step": 2060
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 31.007384615384616,
|
| 1734 |
+
"grad_norm": 0.007714936509728432,
|
| 1735 |
+
"learning_rate": 4.034188034188034e-06,
|
| 1736 |
+
"loss": 0.0082,
|
| 1737 |
+
"step": 2070
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 31.010461538461538,
|
| 1741 |
+
"grad_norm": 32.443336486816406,
|
| 1742 |
+
"learning_rate": 4.000000000000001e-06,
|
| 1743 |
+
"loss": 0.0092,
|
| 1744 |
+
"step": 2080
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 31.013538461538463,
|
| 1748 |
+
"grad_norm": 183.87896728515625,
|
| 1749 |
+
"learning_rate": 3.965811965811966e-06,
|
| 1750 |
+
"loss": 0.2114,
|
| 1751 |
+
"step": 2090
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 31.016615384615385,
|
| 1755 |
+
"grad_norm": 0.09029895812273026,
|
| 1756 |
+
"learning_rate": 3.9316239316239315e-06,
|
| 1757 |
+
"loss": 0.0216,
|
| 1758 |
+
"step": 2100
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 31.019692307692306,
|
| 1762 |
+
"grad_norm": 0.007797298952937126,
|
| 1763 |
+
"learning_rate": 3.897435897435898e-06,
|
| 1764 |
+
"loss": 0.0006,
|
| 1765 |
+
"step": 2110
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 31.020307692307693,
|
| 1769 |
+
"eval_accuracy": 0.7333333333333333,
|
| 1770 |
+
"eval_loss": 1.2723807096481323,
|
| 1771 |
+
"eval_runtime": 14.8246,
|
| 1772 |
+
"eval_samples_per_second": 3.035,
|
| 1773 |
+
"eval_steps_per_second": 0.809,
|
| 1774 |
+
"step": 2112
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 32.00246153846154,
|
| 1778 |
+
"grad_norm": 1.2352715730667114,
|
| 1779 |
+
"learning_rate": 3.863247863247864e-06,
|
| 1780 |
+
"loss": 0.2567,
|
| 1781 |
+
"step": 2120
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 32.005538461538464,
|
| 1785 |
+
"grad_norm": 1.0585654973983765,
|
| 1786 |
+
"learning_rate": 3.8290598290598295e-06,
|
| 1787 |
+
"loss": 0.06,
|
| 1788 |
+
"step": 2130
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 32.00861538461538,
|
| 1792 |
+
"grad_norm": 0.025132305920124054,
|
| 1793 |
+
"learning_rate": 3.794871794871795e-06,
|
| 1794 |
+
"loss": 0.1982,
|
| 1795 |
+
"step": 2140
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 32.01169230769231,
|
| 1799 |
+
"grad_norm": 15.088484764099121,
|
| 1800 |
+
"learning_rate": 3.760683760683761e-06,
|
| 1801 |
+
"loss": 0.3051,
|
| 1802 |
+
"step": 2150
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 32.01476923076923,
|
| 1806 |
+
"grad_norm": 0.003967353142797947,
|
| 1807 |
+
"learning_rate": 3.726495726495727e-06,
|
| 1808 |
+
"loss": 0.0105,
|
| 1809 |
+
"step": 2160
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 32.01784615384615,
|
| 1813 |
+
"grad_norm": 0.07517848908901215,
|
| 1814 |
+
"learning_rate": 3.692307692307693e-06,
|
| 1815 |
+
"loss": 0.1891,
|
| 1816 |
+
"step": 2170
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 32.02030769230769,
|
| 1820 |
+
"eval_accuracy": 0.6444444444444445,
|
| 1821 |
+
"eval_loss": 1.821338176727295,
|
| 1822 |
+
"eval_runtime": 14.7875,
|
| 1823 |
+
"eval_samples_per_second": 3.043,
|
| 1824 |
+
"eval_steps_per_second": 0.811,
|
| 1825 |
+
"step": 2178
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 33.00061538461539,
|
| 1829 |
+
"grad_norm": 0.453707754611969,
|
| 1830 |
+
"learning_rate": 3.6581196581196584e-06,
|
| 1831 |
+
"loss": 0.251,
|
| 1832 |
+
"step": 2180
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 33.003692307692305,
|
| 1836 |
+
"grad_norm": 0.01656286045908928,
|
| 1837 |
+
"learning_rate": 3.623931623931624e-06,
|
| 1838 |
+
"loss": 0.1222,
|
| 1839 |
+
"step": 2190
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 33.00676923076923,
|
| 1843 |
+
"grad_norm": 0.21087659895420074,
|
| 1844 |
+
"learning_rate": 3.58974358974359e-06,
|
| 1845 |
+
"loss": 0.0307,
|
| 1846 |
+
"step": 2200
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 33.009846153846155,
|
| 1850 |
+
"grad_norm": 0.017987869679927826,
|
| 1851 |
+
"learning_rate": 3.555555555555556e-06,
|
| 1852 |
+
"loss": 0.1697,
|
| 1853 |
+
"step": 2210
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 33.01292307692308,
|
| 1857 |
+
"grad_norm": 0.5001537203788757,
|
| 1858 |
+
"learning_rate": 3.5213675213675218e-06,
|
| 1859 |
+
"loss": 0.0074,
|
| 1860 |
+
"step": 2220
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 33.016,
|
| 1864 |
+
"grad_norm": 0.023721082136034966,
|
| 1865 |
+
"learning_rate": 3.487179487179487e-06,
|
| 1866 |
+
"loss": 0.0803,
|
| 1867 |
+
"step": 2230
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 33.01907692307692,
|
| 1871 |
+
"grad_norm": 0.03775130957365036,
|
| 1872 |
+
"learning_rate": 3.452991452991453e-06,
|
| 1873 |
+
"loss": 0.0604,
|
| 1874 |
+
"step": 2240
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 33.02030769230769,
|
| 1878 |
+
"eval_accuracy": 0.5555555555555556,
|
| 1879 |
+
"eval_loss": 2.6845359802246094,
|
| 1880 |
+
"eval_runtime": 14.7376,
|
| 1881 |
+
"eval_samples_per_second": 3.053,
|
| 1882 |
+
"eval_steps_per_second": 0.814,
|
| 1883 |
+
"step": 2244
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 34.00184615384615,
|
| 1887 |
+
"grad_norm": 0.006655918899923563,
|
| 1888 |
+
"learning_rate": 3.4188034188034193e-06,
|
| 1889 |
+
"loss": 0.2833,
|
| 1890 |
+
"step": 2250
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 34.00492307692308,
|
| 1894 |
+
"grad_norm": 0.013796042650938034,
|
| 1895 |
+
"learning_rate": 3.384615384615385e-06,
|
| 1896 |
+
"loss": 0.0196,
|
| 1897 |
+
"step": 2260
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 34.008,
|
| 1901 |
+
"grad_norm": 0.05266215279698372,
|
| 1902 |
+
"learning_rate": 3.3504273504273506e-06,
|
| 1903 |
+
"loss": 0.0876,
|
| 1904 |
+
"step": 2270
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 34.01107692307692,
|
| 1908 |
+
"grad_norm": 0.45490553975105286,
|
| 1909 |
+
"learning_rate": 3.3162393162393165e-06,
|
| 1910 |
+
"loss": 0.0007,
|
| 1911 |
+
"step": 2280
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 34.014153846153846,
|
| 1915 |
+
"grad_norm": 0.12290433049201965,
|
| 1916 |
+
"learning_rate": 3.2820512820512823e-06,
|
| 1917 |
+
"loss": 0.1642,
|
| 1918 |
+
"step": 2290
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 34.01723076923077,
|
| 1922 |
+
"grad_norm": 1.6173644065856934,
|
| 1923 |
+
"learning_rate": 3.247863247863248e-06,
|
| 1924 |
+
"loss": 0.1504,
|
| 1925 |
+
"step": 2300
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 34.02030769230769,
|
| 1929 |
+
"grad_norm": 0.007919330149888992,
|
| 1930 |
+
"learning_rate": 3.213675213675214e-06,
|
| 1931 |
+
"loss": 0.104,
|
| 1932 |
+
"step": 2310
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 34.02030769230769,
|
| 1936 |
+
"eval_accuracy": 0.6222222222222222,
|
| 1937 |
+
"eval_loss": 2.7467894554138184,
|
| 1938 |
+
"eval_runtime": 14.7559,
|
| 1939 |
+
"eval_samples_per_second": 3.05,
|
| 1940 |
+
"eval_steps_per_second": 0.813,
|
| 1941 |
+
"step": 2310
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 35.003076923076925,
|
| 1945 |
+
"grad_norm": 0.0022740329150110483,
|
| 1946 |
+
"learning_rate": 3.1794871794871795e-06,
|
| 1947 |
+
"loss": 0.0003,
|
| 1948 |
+
"step": 2320
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 35.00615384615384,
|
| 1952 |
+
"grad_norm": 0.16805915534496307,
|
| 1953 |
+
"learning_rate": 3.1452991452991453e-06,
|
| 1954 |
+
"loss": 0.001,
|
| 1955 |
+
"step": 2330
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 35.00923076923077,
|
| 1959 |
+
"grad_norm": 0.01777742989361286,
|
| 1960 |
+
"learning_rate": 3.1111111111111116e-06,
|
| 1961 |
+
"loss": 0.039,
|
| 1962 |
+
"step": 2340
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 35.012307692307694,
|
| 1966 |
+
"grad_norm": 6.031890869140625,
|
| 1967 |
+
"learning_rate": 3.0769230769230774e-06,
|
| 1968 |
+
"loss": 0.0026,
|
| 1969 |
+
"step": 2350
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 35.01538461538462,
|
| 1973 |
+
"grad_norm": 17.243139266967773,
|
| 1974 |
+
"learning_rate": 3.042735042735043e-06,
|
| 1975 |
+
"loss": 0.0728,
|
| 1976 |
+
"step": 2360
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 35.01846153846154,
|
| 1980 |
+
"grad_norm": 0.01590224727988243,
|
| 1981 |
+
"learning_rate": 3.0085470085470087e-06,
|
| 1982 |
+
"loss": 0.028,
|
| 1983 |
+
"step": 2370
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 35.02030769230769,
|
| 1987 |
+
"eval_accuracy": 0.7333333333333333,
|
| 1988 |
+
"eval_loss": 1.4457882642745972,
|
| 1989 |
+
"eval_runtime": 13.767,
|
| 1990 |
+
"eval_samples_per_second": 3.269,
|
| 1991 |
+
"eval_steps_per_second": 0.872,
|
| 1992 |
+
"step": 2376
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 36.001230769230766,
|
| 1996 |
+
"grad_norm": 256.8249816894531,
|
| 1997 |
+
"learning_rate": 2.9743589743589746e-06,
|
| 1998 |
+
"loss": 0.1384,
|
| 1999 |
+
"step": 2380
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 36.00430769230769,
|
| 2003 |
+
"grad_norm": 0.012335916049778461,
|
| 2004 |
+
"learning_rate": 2.9401709401709404e-06,
|
| 2005 |
+
"loss": 0.3509,
|
| 2006 |
+
"step": 2390
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 36.007384615384616,
|
| 2010 |
+
"grad_norm": 0.01209163572639227,
|
| 2011 |
+
"learning_rate": 2.9059829059829063e-06,
|
| 2012 |
+
"loss": 0.0437,
|
| 2013 |
+
"step": 2400
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 36.01046153846154,
|
| 2017 |
+
"grad_norm": 0.15365378558635712,
|
| 2018 |
+
"learning_rate": 2.8717948717948717e-06,
|
| 2019 |
+
"loss": 0.1917,
|
| 2020 |
+
"step": 2410
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 36.01353846153846,
|
| 2024 |
+
"grad_norm": 0.006477895192801952,
|
| 2025 |
+
"learning_rate": 2.8376068376068376e-06,
|
| 2026 |
+
"loss": 0.1432,
|
| 2027 |
+
"step": 2420
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 36.016615384615385,
|
| 2031 |
+
"grad_norm": 0.006202262826263905,
|
| 2032 |
+
"learning_rate": 2.803418803418804e-06,
|
| 2033 |
+
"loss": 0.2241,
|
| 2034 |
+
"step": 2430
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 36.01969230769231,
|
| 2038 |
+
"grad_norm": 0.0072620464488863945,
|
| 2039 |
+
"learning_rate": 2.7692307692307697e-06,
|
| 2040 |
+
"loss": 0.2479,
|
| 2041 |
+
"step": 2440
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 36.02030769230769,
|
| 2045 |
+
"eval_accuracy": 0.6222222222222222,
|
| 2046 |
+
"eval_loss": 2.161320447921753,
|
| 2047 |
+
"eval_runtime": 13.7692,
|
| 2048 |
+
"eval_samples_per_second": 3.268,
|
| 2049 |
+
"eval_steps_per_second": 0.872,
|
| 2050 |
+
"step": 2442
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 37.00246153846154,
|
| 2054 |
+
"grad_norm": 0.007183571346104145,
|
| 2055 |
+
"learning_rate": 2.7350427350427355e-06,
|
| 2056 |
+
"loss": 0.0003,
|
| 2057 |
+
"step": 2450
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 37.005538461538464,
|
| 2061 |
+
"grad_norm": 0.7165050506591797,
|
| 2062 |
+
"learning_rate": 2.700854700854701e-06,
|
| 2063 |
+
"loss": 0.0026,
|
| 2064 |
+
"step": 2460
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 37.00861538461538,
|
| 2068 |
+
"grad_norm": 0.08769255876541138,
|
| 2069 |
+
"learning_rate": 2.666666666666667e-06,
|
| 2070 |
+
"loss": 0.0422,
|
| 2071 |
+
"step": 2470
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 37.01169230769231,
|
| 2075 |
+
"grad_norm": 0.17536067962646484,
|
| 2076 |
+
"learning_rate": 2.6324786324786327e-06,
|
| 2077 |
+
"loss": 0.0004,
|
| 2078 |
+
"step": 2480
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 37.01476923076923,
|
| 2082 |
+
"grad_norm": 0.6829420924186707,
|
| 2083 |
+
"learning_rate": 2.5982905982905985e-06,
|
| 2084 |
+
"loss": 0.0898,
|
| 2085 |
+
"step": 2490
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 37.01784615384615,
|
| 2089 |
+
"grad_norm": 0.5209939479827881,
|
| 2090 |
+
"learning_rate": 2.564102564102564e-06,
|
| 2091 |
+
"loss": 0.0967,
|
| 2092 |
+
"step": 2500
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 37.02030769230769,
|
| 2096 |
+
"eval_accuracy": 0.7777777777777778,
|
| 2097 |
+
"eval_loss": 1.3894845247268677,
|
| 2098 |
+
"eval_runtime": 13.7479,
|
| 2099 |
+
"eval_samples_per_second": 3.273,
|
| 2100 |
+
"eval_steps_per_second": 0.873,
|
| 2101 |
+
"step": 2508
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 38.00061538461539,
|
| 2105 |
+
"grad_norm": 75.50067901611328,
|
| 2106 |
+
"learning_rate": 2.52991452991453e-06,
|
| 2107 |
+
"loss": 0.0198,
|
| 2108 |
+
"step": 2510
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 38.003692307692305,
|
| 2112 |
+
"grad_norm": 0.04112862050533295,
|
| 2113 |
+
"learning_rate": 2.495726495726496e-06,
|
| 2114 |
+
"loss": 0.2984,
|
| 2115 |
+
"step": 2520
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 38.00676923076923,
|
| 2119 |
+
"grad_norm": 0.005224079359322786,
|
| 2120 |
+
"learning_rate": 2.461538461538462e-06,
|
| 2121 |
+
"loss": 0.0146,
|
| 2122 |
+
"step": 2530
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 38.009846153846155,
|
| 2126 |
+
"grad_norm": 0.00474022189155221,
|
| 2127 |
+
"learning_rate": 2.4273504273504274e-06,
|
| 2128 |
+
"loss": 0.0487,
|
| 2129 |
+
"step": 2540
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 38.01292307692308,
|
| 2133 |
+
"grad_norm": 1.4753004312515259,
|
| 2134 |
+
"learning_rate": 2.3931623931623937e-06,
|
| 2135 |
+
"loss": 0.0005,
|
| 2136 |
+
"step": 2550
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 38.016,
|
| 2140 |
+
"grad_norm": 0.0049695721827447414,
|
| 2141 |
+
"learning_rate": 2.358974358974359e-06,
|
| 2142 |
+
"loss": 0.0004,
|
| 2143 |
+
"step": 2560
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 38.01907692307692,
|
| 2147 |
+
"grad_norm": 117.37168884277344,
|
| 2148 |
+
"learning_rate": 2.324786324786325e-06,
|
| 2149 |
+
"loss": 0.0668,
|
| 2150 |
+
"step": 2570
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 38.02030769230769,
|
| 2154 |
+
"eval_accuracy": 0.6666666666666666,
|
| 2155 |
+
"eval_loss": 2.014683246612549,
|
| 2156 |
+
"eval_runtime": 13.7937,
|
| 2157 |
+
"eval_samples_per_second": 3.262,
|
| 2158 |
+
"eval_steps_per_second": 0.87,
|
| 2159 |
+
"step": 2574
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 39.00184615384615,
|
| 2163 |
+
"grad_norm": 0.007441969588398933,
|
| 2164 |
+
"learning_rate": 2.290598290598291e-06,
|
| 2165 |
+
"loss": 0.062,
|
| 2166 |
+
"step": 2580
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 39.00492307692308,
|
| 2170 |
+
"grad_norm": 16.319353103637695,
|
| 2171 |
+
"learning_rate": 2.2564102564102566e-06,
|
| 2172 |
+
"loss": 0.0194,
|
| 2173 |
+
"step": 2590
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 39.008,
|
| 2177 |
+
"grad_norm": 180.9895477294922,
|
| 2178 |
+
"learning_rate": 2.222222222222222e-06,
|
| 2179 |
+
"loss": 0.1649,
|
| 2180 |
+
"step": 2600
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 39.01107692307692,
|
| 2184 |
+
"grad_norm": 0.0043008350767195225,
|
| 2185 |
+
"learning_rate": 2.1880341880341884e-06,
|
| 2186 |
+
"loss": 0.0003,
|
| 2187 |
+
"step": 2610
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 39.014153846153846,
|
| 2191 |
+
"grad_norm": 0.0688568577170372,
|
| 2192 |
+
"learning_rate": 2.153846153846154e-06,
|
| 2193 |
+
"loss": 0.0141,
|
| 2194 |
+
"step": 2620
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 39.01723076923077,
|
| 2198 |
+
"grad_norm": 1.9923820495605469,
|
| 2199 |
+
"learning_rate": 2.1196581196581196e-06,
|
| 2200 |
+
"loss": 0.0107,
|
| 2201 |
+
"step": 2630
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 39.02030769230769,
|
| 2205 |
+
"grad_norm": 0.002165107289329171,
|
| 2206 |
+
"learning_rate": 2.085470085470086e-06,
|
| 2207 |
+
"loss": 0.0004,
|
| 2208 |
+
"step": 2640
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 39.02030769230769,
|
| 2212 |
+
"eval_accuracy": 0.6888888888888889,
|
| 2213 |
+
"eval_loss": 1.5766425132751465,
|
| 2214 |
+
"eval_runtime": 13.7894,
|
| 2215 |
+
"eval_samples_per_second": 3.263,
|
| 2216 |
+
"eval_steps_per_second": 0.87,
|
| 2217 |
+
"step": 2640
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 40.003076923076925,
|
| 2221 |
+
"grad_norm": 0.022886795923113823,
|
| 2222 |
+
"learning_rate": 2.0512820512820513e-06,
|
| 2223 |
+
"loss": 0.0006,
|
| 2224 |
+
"step": 2650
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 40.00615384615384,
|
| 2228 |
+
"grad_norm": 0.007079885341227055,
|
| 2229 |
+
"learning_rate": 2.017094017094017e-06,
|
| 2230 |
+
"loss": 0.0003,
|
| 2231 |
+
"step": 2660
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 40.00923076923077,
|
| 2235 |
+
"grad_norm": 0.005172636825591326,
|
| 2236 |
+
"learning_rate": 1.982905982905983e-06,
|
| 2237 |
+
"loss": 0.1478,
|
| 2238 |
+
"step": 2670
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 40.012307692307694,
|
| 2242 |
+
"grad_norm": 0.003111331257969141,
|
| 2243 |
+
"learning_rate": 1.948717948717949e-06,
|
| 2244 |
+
"loss": 0.154,
|
| 2245 |
+
"step": 2680
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 40.01538461538462,
|
| 2249 |
+
"grad_norm": 0.0919618308544159,
|
| 2250 |
+
"learning_rate": 1.9145299145299148e-06,
|
| 2251 |
+
"loss": 0.0004,
|
| 2252 |
+
"step": 2690
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 40.01846153846154,
|
| 2256 |
+
"grad_norm": 0.004653541836887598,
|
| 2257 |
+
"learning_rate": 1.8803418803418804e-06,
|
| 2258 |
+
"loss": 0.0027,
|
| 2259 |
+
"step": 2700
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 40.02030769230769,
|
| 2263 |
+
"eval_accuracy": 0.6444444444444445,
|
| 2264 |
+
"eval_loss": 2.35329532623291,
|
| 2265 |
+
"eval_runtime": 13.9233,
|
| 2266 |
+
"eval_samples_per_second": 3.232,
|
| 2267 |
+
"eval_steps_per_second": 0.862,
|
| 2268 |
+
"step": 2706
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 41.001230769230766,
|
| 2272 |
+
"grad_norm": 0.007990103214979172,
|
| 2273 |
+
"learning_rate": 1.8461538461538465e-06,
|
| 2274 |
+
"loss": 0.166,
|
| 2275 |
+
"step": 2710
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 41.00430769230769,
|
| 2279 |
+
"grad_norm": 0.31218963861465454,
|
| 2280 |
+
"learning_rate": 1.811965811965812e-06,
|
| 2281 |
+
"loss": 0.0003,
|
| 2282 |
+
"step": 2720
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 41.007384615384616,
|
| 2286 |
+
"grad_norm": 0.020261919125914574,
|
| 2287 |
+
"learning_rate": 1.777777777777778e-06,
|
| 2288 |
+
"loss": 0.005,
|
| 2289 |
+
"step": 2730
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 41.01046153846154,
|
| 2293 |
+
"grad_norm": 0.08753982931375504,
|
| 2294 |
+
"learning_rate": 1.7435897435897436e-06,
|
| 2295 |
+
"loss": 0.0001,
|
| 2296 |
+
"step": 2740
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 41.01353846153846,
|
| 2300 |
+
"grad_norm": 0.0024725000839680433,
|
| 2301 |
+
"learning_rate": 1.7094017094017097e-06,
|
| 2302 |
+
"loss": 0.0003,
|
| 2303 |
+
"step": 2750
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 41.016615384615385,
|
| 2307 |
+
"grad_norm": 0.44845858216285706,
|
| 2308 |
+
"learning_rate": 1.6752136752136753e-06,
|
| 2309 |
+
"loss": 0.0003,
|
| 2310 |
+
"step": 2760
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 41.01969230769231,
|
| 2314 |
+
"grad_norm": 0.10153026133775711,
|
| 2315 |
+
"learning_rate": 1.6410256410256412e-06,
|
| 2316 |
+
"loss": 0.1436,
|
| 2317 |
+
"step": 2770
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 41.02030769230769,
|
| 2321 |
+
"eval_accuracy": 0.6444444444444445,
|
| 2322 |
+
"eval_loss": 2.149592161178589,
|
| 2323 |
+
"eval_runtime": 14.0747,
|
| 2324 |
+
"eval_samples_per_second": 3.197,
|
| 2325 |
+
"eval_steps_per_second": 0.853,
|
| 2326 |
+
"step": 2772
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 42.00246153846154,
|
| 2330 |
+
"grad_norm": 0.0065581537783145905,
|
| 2331 |
+
"learning_rate": 1.606837606837607e-06,
|
| 2332 |
+
"loss": 0.1022,
|
| 2333 |
+
"step": 2780
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 42.005538461538464,
|
| 2337 |
+
"grad_norm": 0.010586395859718323,
|
| 2338 |
+
"learning_rate": 1.5726495726495727e-06,
|
| 2339 |
+
"loss": 0.0034,
|
| 2340 |
+
"step": 2790
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 42.00861538461538,
|
| 2344 |
+
"grad_norm": 0.012179357931017876,
|
| 2345 |
+
"learning_rate": 1.5384615384615387e-06,
|
| 2346 |
+
"loss": 0.0002,
|
| 2347 |
+
"step": 2800
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 42.01169230769231,
|
| 2351 |
+
"grad_norm": 0.0030303443782031536,
|
| 2352 |
+
"learning_rate": 1.5042735042735044e-06,
|
| 2353 |
+
"loss": 0.164,
|
| 2354 |
+
"step": 2810
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 42.01476923076923,
|
| 2358 |
+
"grad_norm": 0.09926377236843109,
|
| 2359 |
+
"learning_rate": 1.4700854700854702e-06,
|
| 2360 |
+
"loss": 0.0204,
|
| 2361 |
+
"step": 2820
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 42.01784615384615,
|
| 2365 |
+
"grad_norm": 0.00442868797108531,
|
| 2366 |
+
"learning_rate": 1.4358974358974359e-06,
|
| 2367 |
+
"loss": 0.0327,
|
| 2368 |
+
"step": 2830
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 42.02030769230769,
|
| 2372 |
+
"eval_accuracy": 0.6444444444444445,
|
| 2373 |
+
"eval_loss": 2.2866263389587402,
|
| 2374 |
+
"eval_runtime": 14.4549,
|
| 2375 |
+
"eval_samples_per_second": 3.113,
|
| 2376 |
+
"eval_steps_per_second": 0.83,
|
| 2377 |
+
"step": 2838
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 43.00061538461539,
|
| 2381 |
+
"grad_norm": 0.010174613445997238,
|
| 2382 |
+
"learning_rate": 1.401709401709402e-06,
|
| 2383 |
+
"loss": 0.0003,
|
| 2384 |
+
"step": 2840
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 43.003692307692305,
|
| 2388 |
+
"grad_norm": 0.0034826870542019606,
|
| 2389 |
+
"learning_rate": 1.3675213675213678e-06,
|
| 2390 |
+
"loss": 0.0011,
|
| 2391 |
+
"step": 2850
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 43.00676923076923,
|
| 2395 |
+
"grad_norm": 0.006616006139665842,
|
| 2396 |
+
"learning_rate": 1.3333333333333334e-06,
|
| 2397 |
+
"loss": 0.1652,
|
| 2398 |
+
"step": 2860
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 43.009846153846155,
|
| 2402 |
+
"grad_norm": 0.006677310448139906,
|
| 2403 |
+
"learning_rate": 1.2991452991452993e-06,
|
| 2404 |
+
"loss": 0.0002,
|
| 2405 |
+
"step": 2870
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 43.01292307692308,
|
| 2409 |
+
"grad_norm": 0.024410521611571312,
|
| 2410 |
+
"learning_rate": 1.264957264957265e-06,
|
| 2411 |
+
"loss": 0.0002,
|
| 2412 |
+
"step": 2880
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 43.016,
|
| 2416 |
+
"grad_norm": 0.016464335843920708,
|
| 2417 |
+
"learning_rate": 1.230769230769231e-06,
|
| 2418 |
+
"loss": 0.0058,
|
| 2419 |
+
"step": 2890
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 43.01907692307692,
|
| 2423 |
+
"grad_norm": 0.005353263579308987,
|
| 2424 |
+
"learning_rate": 1.1965811965811968e-06,
|
| 2425 |
+
"loss": 0.0349,
|
| 2426 |
+
"step": 2900
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 43.02030769230769,
|
| 2430 |
+
"eval_accuracy": 0.6666666666666666,
|
| 2431 |
+
"eval_loss": 2.2496182918548584,
|
| 2432 |
+
"eval_runtime": 13.6552,
|
| 2433 |
+
"eval_samples_per_second": 3.295,
|
| 2434 |
+
"eval_steps_per_second": 0.879,
|
| 2435 |
+
"step": 2904
|
| 2436 |
+
},
|
| 2437 |
+
{
|
| 2438 |
+
"epoch": 44.00184615384615,
|
| 2439 |
+
"grad_norm": 0.002462292555719614,
|
| 2440 |
+
"learning_rate": 1.1623931623931625e-06,
|
| 2441 |
+
"loss": 0.0001,
|
| 2442 |
+
"step": 2910
|
| 2443 |
+
},
|
| 2444 |
+
{
|
| 2445 |
+
"epoch": 44.00492307692308,
|
| 2446 |
+
"grad_norm": 0.04071797430515289,
|
| 2447 |
+
"learning_rate": 1.1282051282051283e-06,
|
| 2448 |
+
"loss": 0.1643,
|
| 2449 |
+
"step": 2920
|
| 2450 |
+
},
|
| 2451 |
+
{
|
| 2452 |
+
"epoch": 44.008,
|
| 2453 |
+
"grad_norm": 0.12705393135547638,
|
| 2454 |
+
"learning_rate": 1.0940170940170942e-06,
|
| 2455 |
+
"loss": 0.0002,
|
| 2456 |
+
"step": 2930
|
| 2457 |
+
},
|
| 2458 |
+
{
|
| 2459 |
+
"epoch": 44.01107692307692,
|
| 2460 |
+
"grad_norm": 0.027374783530831337,
|
| 2461 |
+
"learning_rate": 1.0598290598290598e-06,
|
| 2462 |
+
"loss": 0.0002,
|
| 2463 |
+
"step": 2940
|
| 2464 |
+
},
|
| 2465 |
+
{
|
| 2466 |
+
"epoch": 44.014153846153846,
|
| 2467 |
+
"grad_norm": 0.0212483499199152,
|
| 2468 |
+
"learning_rate": 1.0256410256410257e-06,
|
| 2469 |
+
"loss": 0.1404,
|
| 2470 |
+
"step": 2950
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 44.01723076923077,
|
| 2474 |
+
"grad_norm": 0.01933380588889122,
|
| 2475 |
+
"learning_rate": 9.914529914529915e-07,
|
| 2476 |
+
"loss": 0.0002,
|
| 2477 |
+
"step": 2960
|
| 2478 |
+
},
|
| 2479 |
+
{
|
| 2480 |
+
"epoch": 44.02030769230769,
|
| 2481 |
+
"grad_norm": 0.005917856004089117,
|
| 2482 |
+
"learning_rate": 9.572649572649574e-07,
|
| 2483 |
+
"loss": 0.1178,
|
| 2484 |
+
"step": 2970
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 44.02030769230769,
|
| 2488 |
+
"eval_accuracy": 0.6888888888888889,
|
| 2489 |
+
"eval_loss": 1.8929468393325806,
|
| 2490 |
+
"eval_runtime": 14.1233,
|
| 2491 |
+
"eval_samples_per_second": 3.186,
|
| 2492 |
+
"eval_steps_per_second": 0.85,
|
| 2493 |
+
"step": 2970
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 45.003076923076925,
|
| 2497 |
+
"grad_norm": 0.0025133194867521524,
|
| 2498 |
+
"learning_rate": 9.230769230769232e-07,
|
| 2499 |
+
"loss": 0.0023,
|
| 2500 |
+
"step": 2980
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 45.00615384615384,
|
| 2504 |
+
"grad_norm": 0.3587754964828491,
|
| 2505 |
+
"learning_rate": 8.88888888888889e-07,
|
| 2506 |
+
"loss": 0.0003,
|
| 2507 |
+
"step": 2990
|
| 2508 |
+
},
|
| 2509 |
+
{
|
| 2510 |
+
"epoch": 45.00923076923077,
|
| 2511 |
+
"grad_norm": 0.00992894358932972,
|
| 2512 |
+
"learning_rate": 8.547008547008548e-07,
|
| 2513 |
+
"loss": 0.0003,
|
| 2514 |
+
"step": 3000
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 45.012307692307694,
|
| 2518 |
+
"grad_norm": 0.007983594201505184,
|
| 2519 |
+
"learning_rate": 8.205128205128206e-07,
|
| 2520 |
+
"loss": 0.0006,
|
| 2521 |
+
"step": 3010
|
| 2522 |
+
},
|
| 2523 |
+
{
|
| 2524 |
+
"epoch": 45.01538461538462,
|
| 2525 |
+
"grad_norm": 0.07788557559251785,
|
| 2526 |
+
"learning_rate": 7.863247863247863e-07,
|
| 2527 |
+
"loss": 0.0003,
|
| 2528 |
+
"step": 3020
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 45.01846153846154,
|
| 2532 |
+
"grad_norm": 0.006400303915143013,
|
| 2533 |
+
"learning_rate": 7.521367521367522e-07,
|
| 2534 |
+
"loss": 0.0001,
|
| 2535 |
+
"step": 3030
|
| 2536 |
+
},
|
| 2537 |
+
{
|
| 2538 |
+
"epoch": 45.02030769230769,
|
| 2539 |
+
"eval_accuracy": 0.7111111111111111,
|
| 2540 |
+
"eval_loss": 1.903012990951538,
|
| 2541 |
+
"eval_runtime": 13.7223,
|
| 2542 |
+
"eval_samples_per_second": 3.279,
|
| 2543 |
+
"eval_steps_per_second": 0.874,
|
| 2544 |
+
"step": 3036
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 46.001230769230766,
|
| 2548 |
+
"grad_norm": 0.01581370085477829,
|
| 2549 |
+
"learning_rate": 7.179487179487179e-07,
|
| 2550 |
+
"loss": 0.0004,
|
| 2551 |
+
"step": 3040
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 46.00430769230769,
|
| 2555 |
+
"grad_norm": 0.006594088859856129,
|
| 2556 |
+
"learning_rate": 6.837606837606839e-07,
|
| 2557 |
+
"loss": 0.0007,
|
| 2558 |
+
"step": 3050
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 46.007384615384616,
|
| 2562 |
+
"grad_norm": 0.00645817257463932,
|
| 2563 |
+
"learning_rate": 6.495726495726496e-07,
|
| 2564 |
+
"loss": 0.1504,
|
| 2565 |
+
"step": 3060
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 46.01046153846154,
|
| 2569 |
+
"grad_norm": 0.002733448054641485,
|
| 2570 |
+
"learning_rate": 6.153846153846155e-07,
|
| 2571 |
+
"loss": 0.0001,
|
| 2572 |
+
"step": 3070
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 46.01353846153846,
|
| 2576 |
+
"grad_norm": 0.0036937277764081955,
|
| 2577 |
+
"learning_rate": 5.811965811965812e-07,
|
| 2578 |
+
"loss": 0.0001,
|
| 2579 |
+
"step": 3080
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 46.016615384615385,
|
| 2583 |
+
"grad_norm": 0.14585720002651215,
|
| 2584 |
+
"learning_rate": 5.470085470085471e-07,
|
| 2585 |
+
"loss": 0.1164,
|
| 2586 |
+
"step": 3090
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 46.01969230769231,
|
| 2590 |
+
"grad_norm": 0.0056580170057713985,
|
| 2591 |
+
"learning_rate": 5.128205128205128e-07,
|
| 2592 |
+
"loss": 0.1289,
|
| 2593 |
+
"step": 3100
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 46.02030769230769,
|
| 2597 |
+
"eval_accuracy": 0.6888888888888889,
|
| 2598 |
+
"eval_loss": 1.9212490320205688,
|
| 2599 |
+
"eval_runtime": 13.6271,
|
| 2600 |
+
"eval_samples_per_second": 3.302,
|
| 2601 |
+
"eval_steps_per_second": 0.881,
|
| 2602 |
+
"step": 3102
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 47.00246153846154,
|
| 2606 |
+
"grad_norm": 0.0057002464309334755,
|
| 2607 |
+
"learning_rate": 4.786324786324787e-07,
|
| 2608 |
+
"loss": 0.0002,
|
| 2609 |
+
"step": 3110
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 47.005538461538464,
|
| 2613 |
+
"grad_norm": 37.331138610839844,
|
| 2614 |
+
"learning_rate": 4.444444444444445e-07,
|
| 2615 |
+
"loss": 0.0038,
|
| 2616 |
+
"step": 3120
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 47.00861538461538,
|
| 2620 |
+
"grad_norm": 0.003734712256118655,
|
| 2621 |
+
"learning_rate": 4.102564102564103e-07,
|
| 2622 |
+
"loss": 0.0119,
|
| 2623 |
+
"step": 3130
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 47.01169230769231,
|
| 2627 |
+
"grad_norm": 0.005541326943784952,
|
| 2628 |
+
"learning_rate": 3.760683760683761e-07,
|
| 2629 |
+
"loss": 0.0003,
|
| 2630 |
+
"step": 3140
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 47.01476923076923,
|
| 2634 |
+
"grad_norm": 0.012346943840384483,
|
| 2635 |
+
"learning_rate": 3.4188034188034194e-07,
|
| 2636 |
+
"loss": 0.0001,
|
| 2637 |
+
"step": 3150
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 47.01784615384615,
|
| 2641 |
+
"grad_norm": 0.05111232027411461,
|
| 2642 |
+
"learning_rate": 3.0769230769230774e-07,
|
| 2643 |
+
"loss": 0.137,
|
| 2644 |
+
"step": 3160
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 47.02030769230769,
|
| 2648 |
+
"eval_accuracy": 0.7555555555555555,
|
| 2649 |
+
"eval_loss": 1.6076961755752563,
|
| 2650 |
+
"eval_runtime": 13.6997,
|
| 2651 |
+
"eval_samples_per_second": 3.285,
|
| 2652 |
+
"eval_steps_per_second": 0.876,
|
| 2653 |
+
"step": 3168
|
| 2654 |
+
},
|
| 2655 |
+
{
|
| 2656 |
+
"epoch": 48.00061538461539,
|
| 2657 |
+
"grad_norm": 0.005067374091595411,
|
| 2658 |
+
"learning_rate": 2.7350427350427354e-07,
|
| 2659 |
+
"loss": 0.0002,
|
| 2660 |
+
"step": 3170
|
| 2661 |
+
},
|
| 2662 |
+
{
|
| 2663 |
+
"epoch": 48.003692307692305,
|
| 2664 |
+
"grad_norm": 0.5271658301353455,
|
| 2665 |
+
"learning_rate": 2.3931623931623934e-07,
|
| 2666 |
+
"loss": 0.0002,
|
| 2667 |
+
"step": 3180
|
| 2668 |
+
},
|
| 2669 |
+
{
|
| 2670 |
+
"epoch": 48.00676923076923,
|
| 2671 |
+
"grad_norm": 0.01560221053659916,
|
| 2672 |
+
"learning_rate": 2.0512820512820514e-07,
|
| 2673 |
+
"loss": 0.0459,
|
| 2674 |
+
"step": 3190
|
| 2675 |
+
},
|
| 2676 |
+
{
|
| 2677 |
+
"epoch": 48.009846153846155,
|
| 2678 |
+
"grad_norm": 0.007082493510097265,
|
| 2679 |
+
"learning_rate": 1.7094017094017097e-07,
|
| 2680 |
+
"loss": 0.0005,
|
| 2681 |
+
"step": 3200
|
| 2682 |
+
},
|
| 2683 |
+
{
|
| 2684 |
+
"epoch": 48.01292307692308,
|
| 2685 |
+
"grad_norm": 0.0056843808852136135,
|
| 2686 |
+
"learning_rate": 1.3675213675213677e-07,
|
| 2687 |
+
"loss": 0.0009,
|
| 2688 |
+
"step": 3210
|
| 2689 |
+
},
|
| 2690 |
+
{
|
| 2691 |
+
"epoch": 48.016,
|
| 2692 |
+
"grad_norm": 0.020851995795965195,
|
| 2693 |
+
"learning_rate": 1.0256410256410257e-07,
|
| 2694 |
+
"loss": 0.0197,
|
| 2695 |
+
"step": 3220
|
| 2696 |
+
},
|
| 2697 |
+
{
|
| 2698 |
+
"epoch": 48.01907692307692,
|
| 2699 |
+
"grad_norm": 0.0018140689935535192,
|
| 2700 |
+
"learning_rate": 6.837606837606839e-08,
|
| 2701 |
+
"loss": 0.0004,
|
| 2702 |
+
"step": 3230
|
| 2703 |
+
},
|
| 2704 |
+
{
|
| 2705 |
+
"epoch": 48.02030769230769,
|
| 2706 |
+
"eval_accuracy": 0.7333333333333333,
|
| 2707 |
+
"eval_loss": 1.7398090362548828,
|
| 2708 |
+
"eval_runtime": 13.7259,
|
| 2709 |
+
"eval_samples_per_second": 3.278,
|
| 2710 |
+
"eval_steps_per_second": 0.874,
|
| 2711 |
+
"step": 3234
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 49.00184615384615,
|
| 2715 |
+
"grad_norm": 0.007604257669299841,
|
| 2716 |
+
"learning_rate": 3.418803418803419e-08,
|
| 2717 |
+
"loss": 0.0001,
|
| 2718 |
+
"step": 3240
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 49.00492307692308,
|
| 2722 |
+
"grad_norm": 0.003877155715599656,
|
| 2723 |
+
"learning_rate": 0.0,
|
| 2724 |
+
"loss": 0.0001,
|
| 2725 |
+
"step": 3250
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 49.00492307692308,
|
| 2729 |
+
"eval_accuracy": 0.7333333333333333,
|
| 2730 |
+
"eval_loss": 1.7422600984573364,
|
| 2731 |
+
"eval_runtime": 16.804,
|
| 2732 |
+
"eval_samples_per_second": 2.678,
|
| 2733 |
+
"eval_steps_per_second": 0.714,
|
| 2734 |
+
"step": 3250
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 49.00492307692308,
|
| 2738 |
+
"step": 3250,
|
| 2739 |
+
"total_flos": 5.643786123050936e+19,
|
| 2740 |
+
"train_loss": 0.2673717604703628,
|
| 2741 |
+
"train_runtime": 7905.9537,
|
| 2742 |
+
"train_samples_per_second": 1.644,
|
| 2743 |
+
"train_steps_per_second": 0.411
|
| 2744 |
+
},
|
| 2745 |
+
{
|
| 2746 |
+
"epoch": 49.00492307692308,
|
| 2747 |
+
"eval_accuracy": 0.8,
|
| 2748 |
+
"eval_loss": 0.9770982265472412,
|
| 2749 |
+
"eval_runtime": 13.7887,
|
| 2750 |
+
"eval_samples_per_second": 3.264,
|
| 2751 |
+
"eval_steps_per_second": 0.87,
|
| 2752 |
+
"step": 3250
|
| 2753 |
+
},
|
| 2754 |
+
{
|
| 2755 |
+
"epoch": 49.00492307692308,
|
| 2756 |
+
"eval_accuracy": 0.8,
|
| 2757 |
+
"eval_loss": 0.9770981073379517,
|
| 2758 |
+
"eval_runtime": 13.7323,
|
| 2759 |
+
"eval_samples_per_second": 3.277,
|
| 2760 |
+
"eval_steps_per_second": 0.874,
|
| 2761 |
+
"step": 3250
|
| 2762 |
+
}
|
| 2763 |
+
],
|
| 2764 |
+
"logging_steps": 10,
|
| 2765 |
+
"max_steps": 3250,
|
| 2766 |
+
"num_input_tokens_seen": 0,
|
| 2767 |
+
"num_train_epochs": 9223372036854775807,
|
| 2768 |
+
"save_steps": 500,
|
| 2769 |
+
"stateful_callbacks": {
|
| 2770 |
+
"TrainerControl": {
|
| 2771 |
+
"args": {
|
| 2772 |
+
"should_epoch_stop": false,
|
| 2773 |
+
"should_evaluate": false,
|
| 2774 |
+
"should_log": false,
|
| 2775 |
+
"should_save": true,
|
| 2776 |
+
"should_training_stop": true
|
| 2777 |
+
},
|
| 2778 |
+
"attributes": {}
|
| 2779 |
+
}
|
| 2780 |
+
},
|
| 2781 |
+
"total_flos": 5.643786123050936e+19,
|
| 2782 |
+
"train_batch_size": 4,
|
| 2783 |
+
"trial_name": null,
|
| 2784 |
+
"trial_params": null
|
| 2785 |
+
}
|