End of training
Browse files- all_results.json +8 -0
- test_results.json +8 -0
- trainer_state.json +3275 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 49.00759493670886,
|
| 3 |
+
"eval_accuracy": 0.2564102564102564,
|
| 4 |
+
"eval_loss": 2.5595290660858154,
|
| 5 |
+
"eval_runtime": 6.9771,
|
| 6 |
+
"eval_samples_per_second": 5.59,
|
| 7 |
+
"eval_steps_per_second": 1.433
|
| 8 |
+
}
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 49.00759493670886,
|
| 3 |
+
"eval_accuracy": 0.2564102564102564,
|
| 4 |
+
"eval_loss": 2.5595290660858154,
|
| 5 |
+
"eval_runtime": 6.9771,
|
| 6 |
+
"eval_samples_per_second": 5.59,
|
| 7 |
+
"eval_steps_per_second": 1.433
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3275 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.46808510638297873,
|
| 3 |
+
"best_model_checkpoint": "MAE-CT-CPC-Dicotomized-v6-tricot/checkpoint-1200",
|
| 4 |
+
"epoch": 49.00759493670886,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3950,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.002531645569620253,
|
| 13 |
+
"grad_norm": 3.760711669921875,
|
| 14 |
+
"learning_rate": 2.5316455696202533e-07,
|
| 15 |
+
"loss": 1.0986,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.005063291139240506,
|
| 20 |
+
"grad_norm": 3.1336684226989746,
|
| 21 |
+
"learning_rate": 5.063291139240507e-07,
|
| 22 |
+
"loss": 1.1068,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.007594936708860759,
|
| 27 |
+
"grad_norm": 4.340556621551514,
|
| 28 |
+
"learning_rate": 7.59493670886076e-07,
|
| 29 |
+
"loss": 1.1121,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.010126582278481013,
|
| 34 |
+
"grad_norm": 1.9803290367126465,
|
| 35 |
+
"learning_rate": 1.0126582278481013e-06,
|
| 36 |
+
"loss": 1.0948,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.012658227848101266,
|
| 41 |
+
"grad_norm": 4.55110502243042,
|
| 42 |
+
"learning_rate": 1.2658227848101267e-06,
|
| 43 |
+
"loss": 1.0829,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.015189873417721518,
|
| 48 |
+
"grad_norm": 3.38923716545105,
|
| 49 |
+
"learning_rate": 1.518987341772152e-06,
|
| 50 |
+
"loss": 1.1132,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.017721518987341773,
|
| 55 |
+
"grad_norm": 2.0994908809661865,
|
| 56 |
+
"learning_rate": 1.7721518987341774e-06,
|
| 57 |
+
"loss": 1.108,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.020253164556962026,
|
| 62 |
+
"grad_norm": 6.034689903259277,
|
| 63 |
+
"learning_rate": 2.0253164556962026e-06,
|
| 64 |
+
"loss": 1.1005,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.020253164556962026,
|
| 69 |
+
"eval_accuracy": 0.3191489361702128,
|
| 70 |
+
"eval_loss": 1.1025245189666748,
|
| 71 |
+
"eval_runtime": 10.0087,
|
| 72 |
+
"eval_samples_per_second": 4.696,
|
| 73 |
+
"eval_steps_per_second": 1.199,
|
| 74 |
+
"step": 80
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.0025316455696203,
|
| 78 |
+
"grad_norm": 3.7084062099456787,
|
| 79 |
+
"learning_rate": 2.278481012658228e-06,
|
| 80 |
+
"loss": 1.0858,
|
| 81 |
+
"step": 90
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.0050632911392405,
|
| 85 |
+
"grad_norm": 1.3401702642440796,
|
| 86 |
+
"learning_rate": 2.5316455696202535e-06,
|
| 87 |
+
"loss": 1.0927,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.0075949367088608,
|
| 92 |
+
"grad_norm": 3.43015193939209,
|
| 93 |
+
"learning_rate": 2.7848101265822785e-06,
|
| 94 |
+
"loss": 1.0919,
|
| 95 |
+
"step": 110
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.010126582278481,
|
| 99 |
+
"grad_norm": 2.784862518310547,
|
| 100 |
+
"learning_rate": 3.037974683544304e-06,
|
| 101 |
+
"loss": 1.0827,
|
| 102 |
+
"step": 120
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.0126582278481013,
|
| 106 |
+
"grad_norm": 2.417646884918213,
|
| 107 |
+
"learning_rate": 3.2911392405063294e-06,
|
| 108 |
+
"loss": 1.1063,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.0151898734177216,
|
| 113 |
+
"grad_norm": 6.169160842895508,
|
| 114 |
+
"learning_rate": 3.544303797468355e-06,
|
| 115 |
+
"loss": 1.0809,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.0177215189873419,
|
| 120 |
+
"grad_norm": 9.149710655212402,
|
| 121 |
+
"learning_rate": 3.7974683544303802e-06,
|
| 122 |
+
"loss": 1.1522,
|
| 123 |
+
"step": 150
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.0202531645569621,
|
| 127 |
+
"grad_norm": 9.3113431930542,
|
| 128 |
+
"learning_rate": 4.050632911392405e-06,
|
| 129 |
+
"loss": 1.1073,
|
| 130 |
+
"step": 160
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.0202531645569621,
|
| 134 |
+
"eval_accuracy": 0.40425531914893614,
|
| 135 |
+
"eval_loss": 1.1014573574066162,
|
| 136 |
+
"eval_runtime": 9.1101,
|
| 137 |
+
"eval_samples_per_second": 5.159,
|
| 138 |
+
"eval_steps_per_second": 1.317,
|
| 139 |
+
"step": 160
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 2.0025316455696203,
|
| 143 |
+
"grad_norm": 5.42681360244751,
|
| 144 |
+
"learning_rate": 4.303797468354431e-06,
|
| 145 |
+
"loss": 1.0812,
|
| 146 |
+
"step": 170
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 2.0050632911392405,
|
| 150 |
+
"grad_norm": 4.614086151123047,
|
| 151 |
+
"learning_rate": 4.556962025316456e-06,
|
| 152 |
+
"loss": 1.1353,
|
| 153 |
+
"step": 180
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.007594936708861,
|
| 157 |
+
"grad_norm": 4.795217990875244,
|
| 158 |
+
"learning_rate": 4.8101265822784815e-06,
|
| 159 |
+
"loss": 1.0904,
|
| 160 |
+
"step": 190
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.010126582278481,
|
| 164 |
+
"grad_norm": 4.968494892120361,
|
| 165 |
+
"learning_rate": 5.063291139240507e-06,
|
| 166 |
+
"loss": 1.0834,
|
| 167 |
+
"step": 200
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.0126582278481013,
|
| 171 |
+
"grad_norm": 7.959131717681885,
|
| 172 |
+
"learning_rate": 5.3164556962025316e-06,
|
| 173 |
+
"loss": 1.0648,
|
| 174 |
+
"step": 210
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.0151898734177216,
|
| 178 |
+
"grad_norm": 10.18262767791748,
|
| 179 |
+
"learning_rate": 5.569620253164557e-06,
|
| 180 |
+
"loss": 1.0537,
|
| 181 |
+
"step": 220
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.017721518987342,
|
| 185 |
+
"grad_norm": 7.758669853210449,
|
| 186 |
+
"learning_rate": 5.8227848101265824e-06,
|
| 187 |
+
"loss": 1.0765,
|
| 188 |
+
"step": 230
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.020253164556962,
|
| 192 |
+
"grad_norm": 11.847743034362793,
|
| 193 |
+
"learning_rate": 6.075949367088608e-06,
|
| 194 |
+
"loss": 1.117,
|
| 195 |
+
"step": 240
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.020253164556962,
|
| 199 |
+
"eval_accuracy": 0.23404255319148937,
|
| 200 |
+
"eval_loss": 1.1485627889633179,
|
| 201 |
+
"eval_runtime": 9.2773,
|
| 202 |
+
"eval_samples_per_second": 5.066,
|
| 203 |
+
"eval_steps_per_second": 1.293,
|
| 204 |
+
"step": 240
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.0025316455696203,
|
| 208 |
+
"grad_norm": 8.35399341583252,
|
| 209 |
+
"learning_rate": 6.329113924050634e-06,
|
| 210 |
+
"loss": 1.0378,
|
| 211 |
+
"step": 250
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.0050632911392405,
|
| 215 |
+
"grad_norm": 7.146502494812012,
|
| 216 |
+
"learning_rate": 6.582278481012659e-06,
|
| 217 |
+
"loss": 1.0584,
|
| 218 |
+
"step": 260
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.007594936708861,
|
| 222 |
+
"grad_norm": 12.241239547729492,
|
| 223 |
+
"learning_rate": 6.835443037974684e-06,
|
| 224 |
+
"loss": 1.0281,
|
| 225 |
+
"step": 270
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.010126582278481,
|
| 229 |
+
"grad_norm": 14.954208374023438,
|
| 230 |
+
"learning_rate": 7.08860759493671e-06,
|
| 231 |
+
"loss": 1.0181,
|
| 232 |
+
"step": 280
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.0126582278481013,
|
| 236 |
+
"grad_norm": 16.156879425048828,
|
| 237 |
+
"learning_rate": 7.341772151898735e-06,
|
| 238 |
+
"loss": 1.019,
|
| 239 |
+
"step": 290
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.0151898734177216,
|
| 243 |
+
"grad_norm": 11.172712326049805,
|
| 244 |
+
"learning_rate": 7.5949367088607605e-06,
|
| 245 |
+
"loss": 1.0673,
|
| 246 |
+
"step": 300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.017721518987342,
|
| 250 |
+
"grad_norm": 6.405689239501953,
|
| 251 |
+
"learning_rate": 7.848101265822786e-06,
|
| 252 |
+
"loss": 1.0382,
|
| 253 |
+
"step": 310
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.020253164556962,
|
| 257 |
+
"grad_norm": 21.32834243774414,
|
| 258 |
+
"learning_rate": 8.10126582278481e-06,
|
| 259 |
+
"loss": 1.0481,
|
| 260 |
+
"step": 320
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.020253164556962,
|
| 264 |
+
"eval_accuracy": 0.1702127659574468,
|
| 265 |
+
"eval_loss": 1.1603105068206787,
|
| 266 |
+
"eval_runtime": 9.6154,
|
| 267 |
+
"eval_samples_per_second": 4.888,
|
| 268 |
+
"eval_steps_per_second": 1.248,
|
| 269 |
+
"step": 320
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 4.00253164556962,
|
| 273 |
+
"grad_norm": 12.187129020690918,
|
| 274 |
+
"learning_rate": 8.354430379746837e-06,
|
| 275 |
+
"loss": 0.9641,
|
| 276 |
+
"step": 330
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 4.0050632911392405,
|
| 280 |
+
"grad_norm": 13.029256820678711,
|
| 281 |
+
"learning_rate": 8.607594936708861e-06,
|
| 282 |
+
"loss": 0.8872,
|
| 283 |
+
"step": 340
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 4.007594936708861,
|
| 287 |
+
"grad_norm": 12.334785461425781,
|
| 288 |
+
"learning_rate": 8.860759493670886e-06,
|
| 289 |
+
"loss": 0.9901,
|
| 290 |
+
"step": 350
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 4.010126582278481,
|
| 294 |
+
"grad_norm": 19.241605758666992,
|
| 295 |
+
"learning_rate": 9.113924050632912e-06,
|
| 296 |
+
"loss": 0.9087,
|
| 297 |
+
"step": 360
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 4.012658227848101,
|
| 301 |
+
"grad_norm": 14.55397891998291,
|
| 302 |
+
"learning_rate": 9.367088607594937e-06,
|
| 303 |
+
"loss": 0.9091,
|
| 304 |
+
"step": 370
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 4.015189873417722,
|
| 308 |
+
"grad_norm": 14.195683479309082,
|
| 309 |
+
"learning_rate": 9.620253164556963e-06,
|
| 310 |
+
"loss": 0.9591,
|
| 311 |
+
"step": 380
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 4.017721518987342,
|
| 315 |
+
"grad_norm": 11.787992477416992,
|
| 316 |
+
"learning_rate": 9.87341772151899e-06,
|
| 317 |
+
"loss": 0.9352,
|
| 318 |
+
"step": 390
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 4.020253164556962,
|
| 322 |
+
"grad_norm": 37.00648498535156,
|
| 323 |
+
"learning_rate": 9.985935302391e-06,
|
| 324 |
+
"loss": 1.0614,
|
| 325 |
+
"step": 400
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 4.020253164556962,
|
| 329 |
+
"eval_accuracy": 0.2978723404255319,
|
| 330 |
+
"eval_loss": 1.2154021263122559,
|
| 331 |
+
"eval_runtime": 9.1769,
|
| 332 |
+
"eval_samples_per_second": 5.122,
|
| 333 |
+
"eval_steps_per_second": 1.308,
|
| 334 |
+
"step": 400
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 5.00253164556962,
|
| 338 |
+
"grad_norm": 12.617149353027344,
|
| 339 |
+
"learning_rate": 9.957805907172996e-06,
|
| 340 |
+
"loss": 0.806,
|
| 341 |
+
"step": 410
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 5.0050632911392405,
|
| 345 |
+
"grad_norm": 19.428220748901367,
|
| 346 |
+
"learning_rate": 9.929676511954994e-06,
|
| 347 |
+
"loss": 1.0,
|
| 348 |
+
"step": 420
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 5.007594936708861,
|
| 352 |
+
"grad_norm": 23.193632125854492,
|
| 353 |
+
"learning_rate": 9.901547116736992e-06,
|
| 354 |
+
"loss": 1.0473,
|
| 355 |
+
"step": 430
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 5.010126582278481,
|
| 359 |
+
"grad_norm": 15.946794509887695,
|
| 360 |
+
"learning_rate": 9.87341772151899e-06,
|
| 361 |
+
"loss": 0.8784,
|
| 362 |
+
"step": 440
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 5.012658227848101,
|
| 366 |
+
"grad_norm": 26.185508728027344,
|
| 367 |
+
"learning_rate": 9.845288326300985e-06,
|
| 368 |
+
"loss": 0.9104,
|
| 369 |
+
"step": 450
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 5.015189873417722,
|
| 373 |
+
"grad_norm": 17.916744232177734,
|
| 374 |
+
"learning_rate": 9.817158931082983e-06,
|
| 375 |
+
"loss": 0.8883,
|
| 376 |
+
"step": 460
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 5.017721518987342,
|
| 380 |
+
"grad_norm": 32.958045959472656,
|
| 381 |
+
"learning_rate": 9.78902953586498e-06,
|
| 382 |
+
"loss": 0.9555,
|
| 383 |
+
"step": 470
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 5.020253164556962,
|
| 387 |
+
"grad_norm": 20.871667861938477,
|
| 388 |
+
"learning_rate": 9.760900140646977e-06,
|
| 389 |
+
"loss": 0.9155,
|
| 390 |
+
"step": 480
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 5.020253164556962,
|
| 394 |
+
"eval_accuracy": 0.3829787234042553,
|
| 395 |
+
"eval_loss": 1.123763918876648,
|
| 396 |
+
"eval_runtime": 9.1226,
|
| 397 |
+
"eval_samples_per_second": 5.152,
|
| 398 |
+
"eval_steps_per_second": 1.315,
|
| 399 |
+
"step": 480
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 6.00253164556962,
|
| 403 |
+
"grad_norm": 16.76120948791504,
|
| 404 |
+
"learning_rate": 9.732770745428974e-06,
|
| 405 |
+
"loss": 0.8061,
|
| 406 |
+
"step": 490
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 6.0050632911392405,
|
| 410 |
+
"grad_norm": 16.275365829467773,
|
| 411 |
+
"learning_rate": 9.704641350210972e-06,
|
| 412 |
+
"loss": 0.7271,
|
| 413 |
+
"step": 500
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 6.007594936708861,
|
| 417 |
+
"grad_norm": 20.02506446838379,
|
| 418 |
+
"learning_rate": 9.67651195499297e-06,
|
| 419 |
+
"loss": 0.808,
|
| 420 |
+
"step": 510
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 6.010126582278481,
|
| 424 |
+
"grad_norm": 30.183837890625,
|
| 425 |
+
"learning_rate": 9.648382559774965e-06,
|
| 426 |
+
"loss": 0.8235,
|
| 427 |
+
"step": 520
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 6.012658227848101,
|
| 431 |
+
"grad_norm": 16.959373474121094,
|
| 432 |
+
"learning_rate": 9.620253164556963e-06,
|
| 433 |
+
"loss": 0.8137,
|
| 434 |
+
"step": 530
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 6.015189873417722,
|
| 438 |
+
"grad_norm": 29.162246704101562,
|
| 439 |
+
"learning_rate": 9.59212376933896e-06,
|
| 440 |
+
"loss": 0.7164,
|
| 441 |
+
"step": 540
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 6.017721518987342,
|
| 445 |
+
"grad_norm": 65.3390884399414,
|
| 446 |
+
"learning_rate": 9.563994374120957e-06,
|
| 447 |
+
"loss": 0.8364,
|
| 448 |
+
"step": 550
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 6.020253164556962,
|
| 452 |
+
"grad_norm": 36.35063552856445,
|
| 453 |
+
"learning_rate": 9.535864978902954e-06,
|
| 454 |
+
"loss": 0.9109,
|
| 455 |
+
"step": 560
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 6.020253164556962,
|
| 459 |
+
"eval_accuracy": 0.44680851063829785,
|
| 460 |
+
"eval_loss": 1.1734052896499634,
|
| 461 |
+
"eval_runtime": 9.3875,
|
| 462 |
+
"eval_samples_per_second": 5.007,
|
| 463 |
+
"eval_steps_per_second": 1.278,
|
| 464 |
+
"step": 560
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 7.00253164556962,
|
| 468 |
+
"grad_norm": 15.767574310302734,
|
| 469 |
+
"learning_rate": 9.507735583684952e-06,
|
| 470 |
+
"loss": 0.8023,
|
| 471 |
+
"step": 570
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 7.0050632911392405,
|
| 475 |
+
"grad_norm": 28.233354568481445,
|
| 476 |
+
"learning_rate": 9.47960618846695e-06,
|
| 477 |
+
"loss": 0.687,
|
| 478 |
+
"step": 580
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 7.007594936708861,
|
| 482 |
+
"grad_norm": 34.78061294555664,
|
| 483 |
+
"learning_rate": 9.451476793248946e-06,
|
| 484 |
+
"loss": 0.6199,
|
| 485 |
+
"step": 590
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 7.010126582278481,
|
| 489 |
+
"grad_norm": 18.762407302856445,
|
| 490 |
+
"learning_rate": 9.423347398030943e-06,
|
| 491 |
+
"loss": 0.7075,
|
| 492 |
+
"step": 600
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 7.012658227848101,
|
| 496 |
+
"grad_norm": 70.0352554321289,
|
| 497 |
+
"learning_rate": 9.395218002812941e-06,
|
| 498 |
+
"loss": 0.7117,
|
| 499 |
+
"step": 610
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 7.015189873417722,
|
| 503 |
+
"grad_norm": 23.976449966430664,
|
| 504 |
+
"learning_rate": 9.367088607594937e-06,
|
| 505 |
+
"loss": 0.7338,
|
| 506 |
+
"step": 620
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 7.017721518987342,
|
| 510 |
+
"grad_norm": 32.20431900024414,
|
| 511 |
+
"learning_rate": 9.338959212376934e-06,
|
| 512 |
+
"loss": 0.625,
|
| 513 |
+
"step": 630
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 7.020253164556962,
|
| 517 |
+
"grad_norm": 32.83122253417969,
|
| 518 |
+
"learning_rate": 9.310829817158932e-06,
|
| 519 |
+
"loss": 0.6969,
|
| 520 |
+
"step": 640
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 7.020253164556962,
|
| 524 |
+
"eval_accuracy": 0.44680851063829785,
|
| 525 |
+
"eval_loss": 1.2587645053863525,
|
| 526 |
+
"eval_runtime": 8.5327,
|
| 527 |
+
"eval_samples_per_second": 5.508,
|
| 528 |
+
"eval_steps_per_second": 1.406,
|
| 529 |
+
"step": 640
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 8.00253164556962,
|
| 533 |
+
"grad_norm": 24.469850540161133,
|
| 534 |
+
"learning_rate": 9.28270042194093e-06,
|
| 535 |
+
"loss": 0.4605,
|
| 536 |
+
"step": 650
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 8.00506329113924,
|
| 540 |
+
"grad_norm": 30.739809036254883,
|
| 541 |
+
"learning_rate": 9.254571026722926e-06,
|
| 542 |
+
"loss": 0.6949,
|
| 543 |
+
"step": 660
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 8.00759493670886,
|
| 547 |
+
"grad_norm": 34.987892150878906,
|
| 548 |
+
"learning_rate": 9.226441631504923e-06,
|
| 549 |
+
"loss": 0.6473,
|
| 550 |
+
"step": 670
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 8.010126582278481,
|
| 554 |
+
"grad_norm": 13.710770606994629,
|
| 555 |
+
"learning_rate": 9.198312236286921e-06,
|
| 556 |
+
"loss": 0.5177,
|
| 557 |
+
"step": 680
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 8.012658227848101,
|
| 561 |
+
"grad_norm": 19.927839279174805,
|
| 562 |
+
"learning_rate": 9.170182841068917e-06,
|
| 563 |
+
"loss": 0.6864,
|
| 564 |
+
"step": 690
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 8.015189873417722,
|
| 568 |
+
"grad_norm": 10.516642570495605,
|
| 569 |
+
"learning_rate": 9.142053445850915e-06,
|
| 570 |
+
"loss": 0.5454,
|
| 571 |
+
"step": 700
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 8.017721518987342,
|
| 575 |
+
"grad_norm": 23.24897575378418,
|
| 576 |
+
"learning_rate": 9.113924050632912e-06,
|
| 577 |
+
"loss": 0.6143,
|
| 578 |
+
"step": 710
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 8.020253164556962,
|
| 582 |
+
"grad_norm": 16.84145164489746,
|
| 583 |
+
"learning_rate": 9.085794655414908e-06,
|
| 584 |
+
"loss": 0.6381,
|
| 585 |
+
"step": 720
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 8.020253164556962,
|
| 589 |
+
"eval_accuracy": 0.425531914893617,
|
| 590 |
+
"eval_loss": 1.2711387872695923,
|
| 591 |
+
"eval_runtime": 8.5388,
|
| 592 |
+
"eval_samples_per_second": 5.504,
|
| 593 |
+
"eval_steps_per_second": 1.405,
|
| 594 |
+
"step": 720
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 9.00253164556962,
|
| 598 |
+
"grad_norm": 14.482467651367188,
|
| 599 |
+
"learning_rate": 9.057665260196906e-06,
|
| 600 |
+
"loss": 0.5143,
|
| 601 |
+
"step": 730
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 9.00506329113924,
|
| 605 |
+
"grad_norm": 28.448471069335938,
|
| 606 |
+
"learning_rate": 9.029535864978903e-06,
|
| 607 |
+
"loss": 0.5399,
|
| 608 |
+
"step": 740
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 9.00759493670886,
|
| 612 |
+
"grad_norm": 24.899137496948242,
|
| 613 |
+
"learning_rate": 9.001406469760901e-06,
|
| 614 |
+
"loss": 0.4783,
|
| 615 |
+
"step": 750
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 9.010126582278481,
|
| 619 |
+
"grad_norm": 32.908470153808594,
|
| 620 |
+
"learning_rate": 8.973277074542897e-06,
|
| 621 |
+
"loss": 0.6049,
|
| 622 |
+
"step": 760
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 9.012658227848101,
|
| 626 |
+
"grad_norm": 30.97938346862793,
|
| 627 |
+
"learning_rate": 8.945147679324895e-06,
|
| 628 |
+
"loss": 0.5292,
|
| 629 |
+
"step": 770
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 9.015189873417722,
|
| 633 |
+
"grad_norm": 15.302186965942383,
|
| 634 |
+
"learning_rate": 8.917018284106892e-06,
|
| 635 |
+
"loss": 0.5043,
|
| 636 |
+
"step": 780
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 9.017721518987342,
|
| 640 |
+
"grad_norm": 6.016586780548096,
|
| 641 |
+
"learning_rate": 8.888888888888888e-06,
|
| 642 |
+
"loss": 0.5436,
|
| 643 |
+
"step": 790
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 9.020253164556962,
|
| 647 |
+
"grad_norm": 38.37864685058594,
|
| 648 |
+
"learning_rate": 8.860759493670886e-06,
|
| 649 |
+
"loss": 0.5455,
|
| 650 |
+
"step": 800
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 9.020253164556962,
|
| 654 |
+
"eval_accuracy": 0.3829787234042553,
|
| 655 |
+
"eval_loss": 1.397441029548645,
|
| 656 |
+
"eval_runtime": 9.0173,
|
| 657 |
+
"eval_samples_per_second": 5.212,
|
| 658 |
+
"eval_steps_per_second": 1.331,
|
| 659 |
+
"step": 800
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 10.00253164556962,
|
| 663 |
+
"grad_norm": 37.03520965576172,
|
| 664 |
+
"learning_rate": 8.832630098452884e-06,
|
| 665 |
+
"loss": 0.2854,
|
| 666 |
+
"step": 810
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 10.00506329113924,
|
| 670 |
+
"grad_norm": 28.204374313354492,
|
| 671 |
+
"learning_rate": 8.804500703234881e-06,
|
| 672 |
+
"loss": 0.3685,
|
| 673 |
+
"step": 820
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 10.00759493670886,
|
| 677 |
+
"grad_norm": 49.104583740234375,
|
| 678 |
+
"learning_rate": 8.776371308016879e-06,
|
| 679 |
+
"loss": 0.505,
|
| 680 |
+
"step": 830
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 10.010126582278481,
|
| 684 |
+
"grad_norm": 43.10760498046875,
|
| 685 |
+
"learning_rate": 8.748241912798877e-06,
|
| 686 |
+
"loss": 0.4673,
|
| 687 |
+
"step": 840
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 10.012658227848101,
|
| 691 |
+
"grad_norm": 42.76717758178711,
|
| 692 |
+
"learning_rate": 8.720112517580872e-06,
|
| 693 |
+
"loss": 0.5466,
|
| 694 |
+
"step": 850
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 10.015189873417722,
|
| 698 |
+
"grad_norm": 15.837285995483398,
|
| 699 |
+
"learning_rate": 8.69198312236287e-06,
|
| 700 |
+
"loss": 0.4993,
|
| 701 |
+
"step": 860
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 10.017721518987342,
|
| 705 |
+
"grad_norm": 29.81525993347168,
|
| 706 |
+
"learning_rate": 8.663853727144868e-06,
|
| 707 |
+
"loss": 0.6739,
|
| 708 |
+
"step": 870
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 10.020253164556962,
|
| 712 |
+
"grad_norm": 3.9737327098846436,
|
| 713 |
+
"learning_rate": 8.635724331926865e-06,
|
| 714 |
+
"loss": 0.4878,
|
| 715 |
+
"step": 880
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 10.020253164556962,
|
| 719 |
+
"eval_accuracy": 0.44680851063829785,
|
| 720 |
+
"eval_loss": 1.2367063760757446,
|
| 721 |
+
"eval_runtime": 8.5434,
|
| 722 |
+
"eval_samples_per_second": 5.501,
|
| 723 |
+
"eval_steps_per_second": 1.405,
|
| 724 |
+
"step": 880
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 11.00253164556962,
|
| 728 |
+
"grad_norm": 5.285787582397461,
|
| 729 |
+
"learning_rate": 8.607594936708861e-06,
|
| 730 |
+
"loss": 0.2708,
|
| 731 |
+
"step": 890
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 11.00506329113924,
|
| 735 |
+
"grad_norm": 24.084455490112305,
|
| 736 |
+
"learning_rate": 8.579465541490859e-06,
|
| 737 |
+
"loss": 0.3374,
|
| 738 |
+
"step": 900
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 11.00759493670886,
|
| 742 |
+
"grad_norm": 44.67960739135742,
|
| 743 |
+
"learning_rate": 8.551336146272857e-06,
|
| 744 |
+
"loss": 0.344,
|
| 745 |
+
"step": 910
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 11.010126582278481,
|
| 749 |
+
"grad_norm": 44.98979949951172,
|
| 750 |
+
"learning_rate": 8.523206751054853e-06,
|
| 751 |
+
"loss": 0.4461,
|
| 752 |
+
"step": 920
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 11.012658227848101,
|
| 756 |
+
"grad_norm": 47.225440979003906,
|
| 757 |
+
"learning_rate": 8.49507735583685e-06,
|
| 758 |
+
"loss": 0.3232,
|
| 759 |
+
"step": 930
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 11.015189873417722,
|
| 763 |
+
"grad_norm": 54.807865142822266,
|
| 764 |
+
"learning_rate": 8.466947960618848e-06,
|
| 765 |
+
"loss": 0.4236,
|
| 766 |
+
"step": 940
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 11.017721518987342,
|
| 770 |
+
"grad_norm": 59.33114242553711,
|
| 771 |
+
"learning_rate": 8.438818565400846e-06,
|
| 772 |
+
"loss": 0.4856,
|
| 773 |
+
"step": 950
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 11.020253164556962,
|
| 777 |
+
"grad_norm": 73.03024291992188,
|
| 778 |
+
"learning_rate": 8.410689170182841e-06,
|
| 779 |
+
"loss": 0.3125,
|
| 780 |
+
"step": 960
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 11.020253164556962,
|
| 784 |
+
"eval_accuracy": 0.44680851063829785,
|
| 785 |
+
"eval_loss": 1.5499666929244995,
|
| 786 |
+
"eval_runtime": 8.5355,
|
| 787 |
+
"eval_samples_per_second": 5.506,
|
| 788 |
+
"eval_steps_per_second": 1.406,
|
| 789 |
+
"step": 960
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 12.00253164556962,
|
| 793 |
+
"grad_norm": 2.4662981033325195,
|
| 794 |
+
"learning_rate": 8.382559774964839e-06,
|
| 795 |
+
"loss": 0.2143,
|
| 796 |
+
"step": 970
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 12.00506329113924,
|
| 800 |
+
"grad_norm": 15.445357322692871,
|
| 801 |
+
"learning_rate": 8.354430379746837e-06,
|
| 802 |
+
"loss": 0.1185,
|
| 803 |
+
"step": 980
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 12.00759493670886,
|
| 807 |
+
"grad_norm": 49.94412612915039,
|
| 808 |
+
"learning_rate": 8.326300984528833e-06,
|
| 809 |
+
"loss": 0.5804,
|
| 810 |
+
"step": 990
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 12.010126582278481,
|
| 814 |
+
"grad_norm": 74.07793426513672,
|
| 815 |
+
"learning_rate": 8.29817158931083e-06,
|
| 816 |
+
"loss": 0.5126,
|
| 817 |
+
"step": 1000
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 12.012658227848101,
|
| 821 |
+
"grad_norm": 41.05870819091797,
|
| 822 |
+
"learning_rate": 8.270042194092828e-06,
|
| 823 |
+
"loss": 0.3563,
|
| 824 |
+
"step": 1010
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 12.015189873417722,
|
| 828 |
+
"grad_norm": 38.70441818237305,
|
| 829 |
+
"learning_rate": 8.241912798874826e-06,
|
| 830 |
+
"loss": 0.3129,
|
| 831 |
+
"step": 1020
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 12.017721518987342,
|
| 835 |
+
"grad_norm": 58.30833435058594,
|
| 836 |
+
"learning_rate": 8.213783403656822e-06,
|
| 837 |
+
"loss": 0.5785,
|
| 838 |
+
"step": 1030
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 12.020253164556962,
|
| 842 |
+
"grad_norm": 21.522768020629883,
|
| 843 |
+
"learning_rate": 8.18565400843882e-06,
|
| 844 |
+
"loss": 0.5886,
|
| 845 |
+
"step": 1040
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 12.020253164556962,
|
| 849 |
+
"eval_accuracy": 0.3191489361702128,
|
| 850 |
+
"eval_loss": 1.7876856327056885,
|
| 851 |
+
"eval_runtime": 8.4637,
|
| 852 |
+
"eval_samples_per_second": 5.553,
|
| 853 |
+
"eval_steps_per_second": 1.418,
|
| 854 |
+
"step": 1040
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 13.00253164556962,
|
| 858 |
+
"grad_norm": 21.94713020324707,
|
| 859 |
+
"learning_rate": 8.157524613220817e-06,
|
| 860 |
+
"loss": 0.5098,
|
| 861 |
+
"step": 1050
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 13.00506329113924,
|
| 865 |
+
"grad_norm": 50.94339370727539,
|
| 866 |
+
"learning_rate": 8.129395218002813e-06,
|
| 867 |
+
"loss": 0.4418,
|
| 868 |
+
"step": 1060
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 13.00759493670886,
|
| 872 |
+
"grad_norm": 48.228641510009766,
|
| 873 |
+
"learning_rate": 8.10126582278481e-06,
|
| 874 |
+
"loss": 0.2643,
|
| 875 |
+
"step": 1070
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 13.010126582278481,
|
| 879 |
+
"grad_norm": 4.573431968688965,
|
| 880 |
+
"learning_rate": 8.073136427566808e-06,
|
| 881 |
+
"loss": 0.237,
|
| 882 |
+
"step": 1080
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 13.012658227848101,
|
| 886 |
+
"grad_norm": 10.239933013916016,
|
| 887 |
+
"learning_rate": 8.045007032348806e-06,
|
| 888 |
+
"loss": 0.3282,
|
| 889 |
+
"step": 1090
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 13.015189873417722,
|
| 893 |
+
"grad_norm": 50.27717971801758,
|
| 894 |
+
"learning_rate": 8.016877637130802e-06,
|
| 895 |
+
"loss": 0.6337,
|
| 896 |
+
"step": 1100
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 13.017721518987342,
|
| 900 |
+
"grad_norm": 0.98967045545578,
|
| 901 |
+
"learning_rate": 7.9887482419128e-06,
|
| 902 |
+
"loss": 0.2943,
|
| 903 |
+
"step": 1110
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 13.020253164556962,
|
| 907 |
+
"grad_norm": 22.69672393798828,
|
| 908 |
+
"learning_rate": 7.960618846694797e-06,
|
| 909 |
+
"loss": 0.1826,
|
| 910 |
+
"step": 1120
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 13.020253164556962,
|
| 914 |
+
"eval_accuracy": 0.3404255319148936,
|
| 915 |
+
"eval_loss": 1.812400460243225,
|
| 916 |
+
"eval_runtime": 8.4503,
|
| 917 |
+
"eval_samples_per_second": 5.562,
|
| 918 |
+
"eval_steps_per_second": 1.42,
|
| 919 |
+
"step": 1120
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 14.00253164556962,
|
| 923 |
+
"grad_norm": 6.671234130859375,
|
| 924 |
+
"learning_rate": 7.932489451476793e-06,
|
| 925 |
+
"loss": 0.2216,
|
| 926 |
+
"step": 1130
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 14.00506329113924,
|
| 930 |
+
"grad_norm": 9.171055793762207,
|
| 931 |
+
"learning_rate": 7.90436005625879e-06,
|
| 932 |
+
"loss": 0.322,
|
| 933 |
+
"step": 1140
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 14.00759493670886,
|
| 937 |
+
"grad_norm": 12.581557273864746,
|
| 938 |
+
"learning_rate": 7.876230661040788e-06,
|
| 939 |
+
"loss": 0.6145,
|
| 940 |
+
"step": 1150
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 14.010126582278481,
|
| 944 |
+
"grad_norm": 7.152590274810791,
|
| 945 |
+
"learning_rate": 7.848101265822786e-06,
|
| 946 |
+
"loss": 0.1229,
|
| 947 |
+
"step": 1160
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 14.012658227848101,
|
| 951 |
+
"grad_norm": 0.8645901083946228,
|
| 952 |
+
"learning_rate": 7.819971870604782e-06,
|
| 953 |
+
"loss": 0.1611,
|
| 954 |
+
"step": 1170
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 14.015189873417722,
|
| 958 |
+
"grad_norm": 1.11472749710083,
|
| 959 |
+
"learning_rate": 7.79184247538678e-06,
|
| 960 |
+
"loss": 0.4917,
|
| 961 |
+
"step": 1180
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 14.017721518987342,
|
| 965 |
+
"grad_norm": 107.85749816894531,
|
| 966 |
+
"learning_rate": 7.763713080168777e-06,
|
| 967 |
+
"loss": 0.9024,
|
| 968 |
+
"step": 1190
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 14.020253164556962,
|
| 972 |
+
"grad_norm": 14.994379997253418,
|
| 973 |
+
"learning_rate": 7.735583684950773e-06,
|
| 974 |
+
"loss": 0.3447,
|
| 975 |
+
"step": 1200
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 14.020253164556962,
|
| 979 |
+
"eval_accuracy": 0.46808510638297873,
|
| 980 |
+
"eval_loss": 1.9852231740951538,
|
| 981 |
+
"eval_runtime": 8.5469,
|
| 982 |
+
"eval_samples_per_second": 5.499,
|
| 983 |
+
"eval_steps_per_second": 1.404,
|
| 984 |
+
"step": 1200
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 15.00253164556962,
|
| 988 |
+
"grad_norm": 24.349227905273438,
|
| 989 |
+
"learning_rate": 7.70745428973277e-06,
|
| 990 |
+
"loss": 0.2401,
|
| 991 |
+
"step": 1210
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 15.00506329113924,
|
| 995 |
+
"grad_norm": 21.898780822753906,
|
| 996 |
+
"learning_rate": 7.679324894514768e-06,
|
| 997 |
+
"loss": 0.0448,
|
| 998 |
+
"step": 1220
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 15.00759493670886,
|
| 1002 |
+
"grad_norm": 22.42329978942871,
|
| 1003 |
+
"learning_rate": 7.651195499296766e-06,
|
| 1004 |
+
"loss": 0.0714,
|
| 1005 |
+
"step": 1230
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 15.010126582278481,
|
| 1009 |
+
"grad_norm": 47.88456726074219,
|
| 1010 |
+
"learning_rate": 7.623066104078764e-06,
|
| 1011 |
+
"loss": 0.0964,
|
| 1012 |
+
"step": 1240
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 15.012658227848101,
|
| 1016 |
+
"grad_norm": 35.635562896728516,
|
| 1017 |
+
"learning_rate": 7.5949367088607605e-06,
|
| 1018 |
+
"loss": 0.2036,
|
| 1019 |
+
"step": 1250
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 15.015189873417722,
|
| 1023 |
+
"grad_norm": 4.004649639129639,
|
| 1024 |
+
"learning_rate": 7.566807313642758e-06,
|
| 1025 |
+
"loss": 0.2016,
|
| 1026 |
+
"step": 1260
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 15.017721518987342,
|
| 1030 |
+
"grad_norm": 90.96910095214844,
|
| 1031 |
+
"learning_rate": 7.538677918424755e-06,
|
| 1032 |
+
"loss": 0.443,
|
| 1033 |
+
"step": 1270
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 15.020253164556962,
|
| 1037 |
+
"grad_norm": 2.9360158443450928,
|
| 1038 |
+
"learning_rate": 7.510548523206752e-06,
|
| 1039 |
+
"loss": 0.2065,
|
| 1040 |
+
"step": 1280
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 15.020253164556962,
|
| 1044 |
+
"eval_accuracy": 0.40425531914893614,
|
| 1045 |
+
"eval_loss": 2.3935444355010986,
|
| 1046 |
+
"eval_runtime": 8.466,
|
| 1047 |
+
"eval_samples_per_second": 5.552,
|
| 1048 |
+
"eval_steps_per_second": 1.417,
|
| 1049 |
+
"step": 1280
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 16.00253164556962,
|
| 1053 |
+
"grad_norm": 1.5907315015792847,
|
| 1054 |
+
"learning_rate": 7.482419127988749e-06,
|
| 1055 |
+
"loss": 0.2055,
|
| 1056 |
+
"step": 1290
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 16.00506329113924,
|
| 1060 |
+
"grad_norm": 75.98424530029297,
|
| 1061 |
+
"learning_rate": 7.454289732770746e-06,
|
| 1062 |
+
"loss": 0.1082,
|
| 1063 |
+
"step": 1300
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 16.00759493670886,
|
| 1067 |
+
"grad_norm": 0.1283876597881317,
|
| 1068 |
+
"learning_rate": 7.426160337552744e-06,
|
| 1069 |
+
"loss": 0.0128,
|
| 1070 |
+
"step": 1310
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 16.01012658227848,
|
| 1074 |
+
"grad_norm": 0.6492541432380676,
|
| 1075 |
+
"learning_rate": 7.398030942334741e-06,
|
| 1076 |
+
"loss": 0.1411,
|
| 1077 |
+
"step": 1320
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 16.0126582278481,
|
| 1081 |
+
"grad_norm": 0.14707793295383453,
|
| 1082 |
+
"learning_rate": 7.369901547116738e-06,
|
| 1083 |
+
"loss": 0.119,
|
| 1084 |
+
"step": 1330
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 16.01518987341772,
|
| 1088 |
+
"grad_norm": 1.1961164474487305,
|
| 1089 |
+
"learning_rate": 7.341772151898735e-06,
|
| 1090 |
+
"loss": 0.2281,
|
| 1091 |
+
"step": 1340
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 16.01772151898734,
|
| 1095 |
+
"grad_norm": 3.3282575607299805,
|
| 1096 |
+
"learning_rate": 7.313642756680732e-06,
|
| 1097 |
+
"loss": 0.0599,
|
| 1098 |
+
"step": 1350
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 16.020253164556962,
|
| 1102 |
+
"grad_norm": 225.71522521972656,
|
| 1103 |
+
"learning_rate": 7.2855133614627295e-06,
|
| 1104 |
+
"loss": 0.3104,
|
| 1105 |
+
"step": 1360
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 16.020253164556962,
|
| 1109 |
+
"eval_accuracy": 0.3191489361702128,
|
| 1110 |
+
"eval_loss": 2.998093366622925,
|
| 1111 |
+
"eval_runtime": 8.4852,
|
| 1112 |
+
"eval_samples_per_second": 5.539,
|
| 1113 |
+
"eval_steps_per_second": 1.414,
|
| 1114 |
+
"step": 1360
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 17.00253164556962,
|
| 1118 |
+
"grad_norm": 0.06949834525585175,
|
| 1119 |
+
"learning_rate": 7.257383966244726e-06,
|
| 1120 |
+
"loss": 0.1566,
|
| 1121 |
+
"step": 1370
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 17.00506329113924,
|
| 1125 |
+
"grad_norm": 1.2171714305877686,
|
| 1126 |
+
"learning_rate": 7.229254571026724e-06,
|
| 1127 |
+
"loss": 0.1073,
|
| 1128 |
+
"step": 1380
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 17.00759493670886,
|
| 1132 |
+
"grad_norm": 0.09082494676113129,
|
| 1133 |
+
"learning_rate": 7.201125175808721e-06,
|
| 1134 |
+
"loss": 0.2472,
|
| 1135 |
+
"step": 1390
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 17.01012658227848,
|
| 1139 |
+
"grad_norm": 4.602968215942383,
|
| 1140 |
+
"learning_rate": 7.172995780590718e-06,
|
| 1141 |
+
"loss": 0.1145,
|
| 1142 |
+
"step": 1400
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 17.0126582278481,
|
| 1146 |
+
"grad_norm": 125.78803253173828,
|
| 1147 |
+
"learning_rate": 7.144866385372715e-06,
|
| 1148 |
+
"loss": 0.1733,
|
| 1149 |
+
"step": 1410
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 17.01518987341772,
|
| 1153 |
+
"grad_norm": 19.27991485595703,
|
| 1154 |
+
"learning_rate": 7.116736990154712e-06,
|
| 1155 |
+
"loss": 0.1231,
|
| 1156 |
+
"step": 1420
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 17.01772151898734,
|
| 1160 |
+
"grad_norm": 28.62449073791504,
|
| 1161 |
+
"learning_rate": 7.08860759493671e-06,
|
| 1162 |
+
"loss": 0.1575,
|
| 1163 |
+
"step": 1430
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"epoch": 17.020253164556962,
|
| 1167 |
+
"grad_norm": 238.7103729248047,
|
| 1168 |
+
"learning_rate": 7.060478199718706e-06,
|
| 1169 |
+
"loss": 0.3517,
|
| 1170 |
+
"step": 1440
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 17.020253164556962,
|
| 1174 |
+
"eval_accuracy": 0.3829787234042553,
|
| 1175 |
+
"eval_loss": 2.552189350128174,
|
| 1176 |
+
"eval_runtime": 8.4896,
|
| 1177 |
+
"eval_samples_per_second": 5.536,
|
| 1178 |
+
"eval_steps_per_second": 1.413,
|
| 1179 |
+
"step": 1440
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 18.00253164556962,
|
| 1183 |
+
"grad_norm": 0.39606913924217224,
|
| 1184 |
+
"learning_rate": 7.032348804500704e-06,
|
| 1185 |
+
"loss": 0.1785,
|
| 1186 |
+
"step": 1450
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 18.00506329113924,
|
| 1190 |
+
"grad_norm": 0.8556973338127136,
|
| 1191 |
+
"learning_rate": 7.004219409282701e-06,
|
| 1192 |
+
"loss": 0.1719,
|
| 1193 |
+
"step": 1460
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 18.00759493670886,
|
| 1197 |
+
"grad_norm": 0.09971759468317032,
|
| 1198 |
+
"learning_rate": 6.9760900140646985e-06,
|
| 1199 |
+
"loss": 0.193,
|
| 1200 |
+
"step": 1470
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 18.01012658227848,
|
| 1204 |
+
"grad_norm": 0.11218991130590439,
|
| 1205 |
+
"learning_rate": 6.947960618846695e-06,
|
| 1206 |
+
"loss": 0.0563,
|
| 1207 |
+
"step": 1480
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 18.0126582278481,
|
| 1211 |
+
"grad_norm": 3.4581549167633057,
|
| 1212 |
+
"learning_rate": 6.919831223628692e-06,
|
| 1213 |
+
"loss": 0.0881,
|
| 1214 |
+
"step": 1490
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 18.01518987341772,
|
| 1218 |
+
"grad_norm": 13.419587135314941,
|
| 1219 |
+
"learning_rate": 6.89170182841069e-06,
|
| 1220 |
+
"loss": 0.0535,
|
| 1221 |
+
"step": 1500
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 18.01772151898734,
|
| 1225 |
+
"grad_norm": 73.61225891113281,
|
| 1226 |
+
"learning_rate": 6.8635724331926865e-06,
|
| 1227 |
+
"loss": 0.009,
|
| 1228 |
+
"step": 1510
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 18.020253164556962,
|
| 1232 |
+
"grad_norm": 0.08218306303024292,
|
| 1233 |
+
"learning_rate": 6.835443037974684e-06,
|
| 1234 |
+
"loss": 0.0988,
|
| 1235 |
+
"step": 1520
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 18.020253164556962,
|
| 1239 |
+
"eval_accuracy": 0.44680851063829785,
|
| 1240 |
+
"eval_loss": 3.1462602615356445,
|
| 1241 |
+
"eval_runtime": 8.465,
|
| 1242 |
+
"eval_samples_per_second": 5.552,
|
| 1243 |
+
"eval_steps_per_second": 1.418,
|
| 1244 |
+
"step": 1520
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 19.00253164556962,
|
| 1248 |
+
"grad_norm": 2.85400390625,
|
| 1249 |
+
"learning_rate": 6.807313642756681e-06,
|
| 1250 |
+
"loss": 0.1308,
|
| 1251 |
+
"step": 1530
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 19.00506329113924,
|
| 1255 |
+
"grad_norm": 0.41226041316986084,
|
| 1256 |
+
"learning_rate": 6.779184247538679e-06,
|
| 1257 |
+
"loss": 0.0499,
|
| 1258 |
+
"step": 1540
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 19.00759493670886,
|
| 1262 |
+
"grad_norm": 1.642881989479065,
|
| 1263 |
+
"learning_rate": 6.751054852320675e-06,
|
| 1264 |
+
"loss": 0.0122,
|
| 1265 |
+
"step": 1550
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 19.01012658227848,
|
| 1269 |
+
"grad_norm": 0.10481233894824982,
|
| 1270 |
+
"learning_rate": 6.722925457102672e-06,
|
| 1271 |
+
"loss": 0.0591,
|
| 1272 |
+
"step": 1560
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 19.0126582278481,
|
| 1276 |
+
"grad_norm": 5.857264995574951,
|
| 1277 |
+
"learning_rate": 6.69479606188467e-06,
|
| 1278 |
+
"loss": 0.1052,
|
| 1279 |
+
"step": 1570
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 19.01518987341772,
|
| 1283 |
+
"grad_norm": 0.14555718004703522,
|
| 1284 |
+
"learning_rate": 6.666666666666667e-06,
|
| 1285 |
+
"loss": 0.0016,
|
| 1286 |
+
"step": 1580
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 19.01772151898734,
|
| 1290 |
+
"grad_norm": 0.03223154693841934,
|
| 1291 |
+
"learning_rate": 6.638537271448664e-06,
|
| 1292 |
+
"loss": 0.0201,
|
| 1293 |
+
"step": 1590
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 19.020253164556962,
|
| 1297 |
+
"grad_norm": 0.4471926689147949,
|
| 1298 |
+
"learning_rate": 6.610407876230661e-06,
|
| 1299 |
+
"loss": 0.0532,
|
| 1300 |
+
"step": 1600
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 19.020253164556962,
|
| 1304 |
+
"eval_accuracy": 0.44680851063829785,
|
| 1305 |
+
"eval_loss": 2.8538002967834473,
|
| 1306 |
+
"eval_runtime": 8.449,
|
| 1307 |
+
"eval_samples_per_second": 5.563,
|
| 1308 |
+
"eval_steps_per_second": 1.42,
|
| 1309 |
+
"step": 1600
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 20.00253164556962,
|
| 1313 |
+
"grad_norm": 3.7759382724761963,
|
| 1314 |
+
"learning_rate": 6.582278481012659e-06,
|
| 1315 |
+
"loss": 0.0043,
|
| 1316 |
+
"step": 1610
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 20.00506329113924,
|
| 1320 |
+
"grad_norm": 0.15632715821266174,
|
| 1321 |
+
"learning_rate": 6.5541490857946555e-06,
|
| 1322 |
+
"loss": 0.1257,
|
| 1323 |
+
"step": 1620
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 20.00759493670886,
|
| 1327 |
+
"grad_norm": 0.18101413547992706,
|
| 1328 |
+
"learning_rate": 6.526019690576652e-06,
|
| 1329 |
+
"loss": 0.0017,
|
| 1330 |
+
"step": 1630
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 20.01012658227848,
|
| 1334 |
+
"grad_norm": 0.02908501587808132,
|
| 1335 |
+
"learning_rate": 6.49789029535865e-06,
|
| 1336 |
+
"loss": 0.1291,
|
| 1337 |
+
"step": 1640
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 20.0126582278481,
|
| 1341 |
+
"grad_norm": 0.18558746576309204,
|
| 1342 |
+
"learning_rate": 6.4697609001406485e-06,
|
| 1343 |
+
"loss": 0.0534,
|
| 1344 |
+
"step": 1650
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 20.01518987341772,
|
| 1348 |
+
"grad_norm": 75.53803253173828,
|
| 1349 |
+
"learning_rate": 6.441631504922645e-06,
|
| 1350 |
+
"loss": 0.0682,
|
| 1351 |
+
"step": 1660
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 20.01772151898734,
|
| 1355 |
+
"grad_norm": 1.0015976428985596,
|
| 1356 |
+
"learning_rate": 6.413502109704642e-06,
|
| 1357 |
+
"loss": 0.1744,
|
| 1358 |
+
"step": 1670
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 20.020253164556962,
|
| 1362 |
+
"grad_norm": 356.8922119140625,
|
| 1363 |
+
"learning_rate": 6.38537271448664e-06,
|
| 1364 |
+
"loss": 0.1791,
|
| 1365 |
+
"step": 1680
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 20.020253164556962,
|
| 1369 |
+
"eval_accuracy": 0.44680851063829785,
|
| 1370 |
+
"eval_loss": 3.0305848121643066,
|
| 1371 |
+
"eval_runtime": 8.4694,
|
| 1372 |
+
"eval_samples_per_second": 5.549,
|
| 1373 |
+
"eval_steps_per_second": 1.417,
|
| 1374 |
+
"step": 1680
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 21.00253164556962,
|
| 1378 |
+
"grad_norm": 0.7954501509666443,
|
| 1379 |
+
"learning_rate": 6.3572433192686365e-06,
|
| 1380 |
+
"loss": 0.0365,
|
| 1381 |
+
"step": 1690
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 21.00506329113924,
|
| 1385 |
+
"grad_norm": 0.20515747368335724,
|
| 1386 |
+
"learning_rate": 6.329113924050634e-06,
|
| 1387 |
+
"loss": 0.0012,
|
| 1388 |
+
"step": 1700
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 21.00759493670886,
|
| 1392 |
+
"grad_norm": 0.3028893768787384,
|
| 1393 |
+
"learning_rate": 6.300984528832631e-06,
|
| 1394 |
+
"loss": 0.0024,
|
| 1395 |
+
"step": 1710
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 21.01012658227848,
|
| 1399 |
+
"grad_norm": 0.11144322901964188,
|
| 1400 |
+
"learning_rate": 6.272855133614629e-06,
|
| 1401 |
+
"loss": 0.0196,
|
| 1402 |
+
"step": 1720
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 21.0126582278481,
|
| 1406 |
+
"grad_norm": 0.19480203092098236,
|
| 1407 |
+
"learning_rate": 6.244725738396625e-06,
|
| 1408 |
+
"loss": 0.0049,
|
| 1409 |
+
"step": 1730
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 21.01518987341772,
|
| 1413 |
+
"grad_norm": 38.44449234008789,
|
| 1414 |
+
"learning_rate": 6.216596343178622e-06,
|
| 1415 |
+
"loss": 0.2805,
|
| 1416 |
+
"step": 1740
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 21.01772151898734,
|
| 1420 |
+
"grad_norm": 4.373863697052002,
|
| 1421 |
+
"learning_rate": 6.18846694796062e-06,
|
| 1422 |
+
"loss": 0.2709,
|
| 1423 |
+
"step": 1750
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 21.020253164556962,
|
| 1427 |
+
"grad_norm": 0.9946948885917664,
|
| 1428 |
+
"learning_rate": 6.160337552742617e-06,
|
| 1429 |
+
"loss": 0.1584,
|
| 1430 |
+
"step": 1760
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 21.020253164556962,
|
| 1434 |
+
"eval_accuracy": 0.3829787234042553,
|
| 1435 |
+
"eval_loss": 3.484677314758301,
|
| 1436 |
+
"eval_runtime": 8.4268,
|
| 1437 |
+
"eval_samples_per_second": 5.577,
|
| 1438 |
+
"eval_steps_per_second": 1.424,
|
| 1439 |
+
"step": 1760
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 22.00253164556962,
|
| 1443 |
+
"grad_norm": 0.14784257113933563,
|
| 1444 |
+
"learning_rate": 6.132208157524614e-06,
|
| 1445 |
+
"loss": 0.0099,
|
| 1446 |
+
"step": 1770
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 22.00506329113924,
|
| 1450 |
+
"grad_norm": 0.6995251774787903,
|
| 1451 |
+
"learning_rate": 6.104078762306611e-06,
|
| 1452 |
+
"loss": 0.0016,
|
| 1453 |
+
"step": 1780
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 22.00759493670886,
|
| 1457 |
+
"grad_norm": 0.10788624733686447,
|
| 1458 |
+
"learning_rate": 6.075949367088608e-06,
|
| 1459 |
+
"loss": 0.023,
|
| 1460 |
+
"step": 1790
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 22.01012658227848,
|
| 1464 |
+
"grad_norm": 135.8372344970703,
|
| 1465 |
+
"learning_rate": 6.0478199718706055e-06,
|
| 1466 |
+
"loss": 0.0438,
|
| 1467 |
+
"step": 1800
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 22.0126582278481,
|
| 1471 |
+
"grad_norm": 0.03643108159303665,
|
| 1472 |
+
"learning_rate": 6.019690576652602e-06,
|
| 1473 |
+
"loss": 0.0042,
|
| 1474 |
+
"step": 1810
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 22.01518987341772,
|
| 1478 |
+
"grad_norm": 0.17580589652061462,
|
| 1479 |
+
"learning_rate": 5.9915611814346e-06,
|
| 1480 |
+
"loss": 0.1806,
|
| 1481 |
+
"step": 1820
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 22.01772151898734,
|
| 1485 |
+
"grad_norm": 166.19268798828125,
|
| 1486 |
+
"learning_rate": 5.963431786216597e-06,
|
| 1487 |
+
"loss": 0.0667,
|
| 1488 |
+
"step": 1830
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 22.020253164556962,
|
| 1492 |
+
"grad_norm": 0.03489532321691513,
|
| 1493 |
+
"learning_rate": 5.935302390998594e-06,
|
| 1494 |
+
"loss": 0.016,
|
| 1495 |
+
"step": 1840
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 22.020253164556962,
|
| 1499 |
+
"eval_accuracy": 0.3191489361702128,
|
| 1500 |
+
"eval_loss": 3.412083625793457,
|
| 1501 |
+
"eval_runtime": 8.4496,
|
| 1502 |
+
"eval_samples_per_second": 5.562,
|
| 1503 |
+
"eval_steps_per_second": 1.42,
|
| 1504 |
+
"step": 1840
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 23.00253164556962,
|
| 1508 |
+
"grad_norm": 0.021203240379691124,
|
| 1509 |
+
"learning_rate": 5.907172995780591e-06,
|
| 1510 |
+
"loss": 0.0019,
|
| 1511 |
+
"step": 1850
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 23.00506329113924,
|
| 1515 |
+
"grad_norm": 0.10056442022323608,
|
| 1516 |
+
"learning_rate": 5.879043600562588e-06,
|
| 1517 |
+
"loss": 0.003,
|
| 1518 |
+
"step": 1860
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 23.00759493670886,
|
| 1522 |
+
"grad_norm": 0.012614885345101357,
|
| 1523 |
+
"learning_rate": 5.850914205344586e-06,
|
| 1524 |
+
"loss": 0.0013,
|
| 1525 |
+
"step": 1870
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 23.01012658227848,
|
| 1529 |
+
"grad_norm": 0.06593940407037735,
|
| 1530 |
+
"learning_rate": 5.8227848101265824e-06,
|
| 1531 |
+
"loss": 0.0012,
|
| 1532 |
+
"step": 1880
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 23.0126582278481,
|
| 1536 |
+
"grad_norm": 0.04878053069114685,
|
| 1537 |
+
"learning_rate": 5.79465541490858e-06,
|
| 1538 |
+
"loss": 0.0843,
|
| 1539 |
+
"step": 1890
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 23.01518987341772,
|
| 1543 |
+
"grad_norm": 0.35371971130371094,
|
| 1544 |
+
"learning_rate": 5.766526019690577e-06,
|
| 1545 |
+
"loss": 0.0015,
|
| 1546 |
+
"step": 1900
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 23.01772151898734,
|
| 1550 |
+
"grad_norm": 0.020635871216654778,
|
| 1551 |
+
"learning_rate": 5.7383966244725745e-06,
|
| 1552 |
+
"loss": 0.0011,
|
| 1553 |
+
"step": 1910
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 23.020253164556962,
|
| 1557 |
+
"grad_norm": 1.8640389442443848,
|
| 1558 |
+
"learning_rate": 5.710267229254571e-06,
|
| 1559 |
+
"loss": 0.0012,
|
| 1560 |
+
"step": 1920
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 23.020253164556962,
|
| 1564 |
+
"eval_accuracy": 0.3617021276595745,
|
| 1565 |
+
"eval_loss": 3.8550264835357666,
|
| 1566 |
+
"eval_runtime": 8.4387,
|
| 1567 |
+
"eval_samples_per_second": 5.57,
|
| 1568 |
+
"eval_steps_per_second": 1.422,
|
| 1569 |
+
"step": 1920
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 24.00253164556962,
|
| 1573 |
+
"grad_norm": 1.7900668382644653,
|
| 1574 |
+
"learning_rate": 5.682137834036568e-06,
|
| 1575 |
+
"loss": 0.0025,
|
| 1576 |
+
"step": 1930
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 24.00506329113924,
|
| 1580 |
+
"grad_norm": 149.13136291503906,
|
| 1581 |
+
"learning_rate": 5.654008438818566e-06,
|
| 1582 |
+
"loss": 0.1809,
|
| 1583 |
+
"step": 1940
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 24.00759493670886,
|
| 1587 |
+
"grad_norm": 0.01757533848285675,
|
| 1588 |
+
"learning_rate": 5.6258790436005626e-06,
|
| 1589 |
+
"loss": 0.0556,
|
| 1590 |
+
"step": 1950
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 24.01012658227848,
|
| 1594 |
+
"grad_norm": 0.0452081672847271,
|
| 1595 |
+
"learning_rate": 5.59774964838256e-06,
|
| 1596 |
+
"loss": 0.0616,
|
| 1597 |
+
"step": 1960
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 24.0126582278481,
|
| 1601 |
+
"grad_norm": 0.09417957812547684,
|
| 1602 |
+
"learning_rate": 5.569620253164557e-06,
|
| 1603 |
+
"loss": 0.0008,
|
| 1604 |
+
"step": 1970
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 24.01518987341772,
|
| 1608 |
+
"grad_norm": 0.030936162918806076,
|
| 1609 |
+
"learning_rate": 5.541490857946555e-06,
|
| 1610 |
+
"loss": 0.102,
|
| 1611 |
+
"step": 1980
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 24.01772151898734,
|
| 1615 |
+
"grad_norm": 0.019063590094447136,
|
| 1616 |
+
"learning_rate": 5.5133614627285514e-06,
|
| 1617 |
+
"loss": 0.0157,
|
| 1618 |
+
"step": 1990
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 24.020253164556962,
|
| 1622 |
+
"grad_norm": 0.015954723581671715,
|
| 1623 |
+
"learning_rate": 5.485232067510548e-06,
|
| 1624 |
+
"loss": 0.0005,
|
| 1625 |
+
"step": 2000
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 24.020253164556962,
|
| 1629 |
+
"eval_accuracy": 0.40425531914893614,
|
| 1630 |
+
"eval_loss": 3.905510425567627,
|
| 1631 |
+
"eval_runtime": 8.4672,
|
| 1632 |
+
"eval_samples_per_second": 5.551,
|
| 1633 |
+
"eval_steps_per_second": 1.417,
|
| 1634 |
+
"step": 2000
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 25.00253164556962,
|
| 1638 |
+
"grad_norm": 0.01930183358490467,
|
| 1639 |
+
"learning_rate": 5.457102672292546e-06,
|
| 1640 |
+
"loss": 0.001,
|
| 1641 |
+
"step": 2010
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 25.00506329113924,
|
| 1645 |
+
"grad_norm": 104.06456756591797,
|
| 1646 |
+
"learning_rate": 5.428973277074543e-06,
|
| 1647 |
+
"loss": 0.1745,
|
| 1648 |
+
"step": 2020
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 25.00759493670886,
|
| 1652 |
+
"grad_norm": 0.13353954255580902,
|
| 1653 |
+
"learning_rate": 5.40084388185654e-06,
|
| 1654 |
+
"loss": 0.0016,
|
| 1655 |
+
"step": 2030
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 25.01012658227848,
|
| 1659 |
+
"grad_norm": 148.02767944335938,
|
| 1660 |
+
"learning_rate": 5.372714486638537e-06,
|
| 1661 |
+
"loss": 0.0277,
|
| 1662 |
+
"step": 2040
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 25.0126582278481,
|
| 1666 |
+
"grad_norm": 0.013294242322444916,
|
| 1667 |
+
"learning_rate": 5.344585091420535e-06,
|
| 1668 |
+
"loss": 0.1552,
|
| 1669 |
+
"step": 2050
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 25.01518987341772,
|
| 1673 |
+
"grad_norm": 0.028865955770015717,
|
| 1674 |
+
"learning_rate": 5.3164556962025316e-06,
|
| 1675 |
+
"loss": 0.0006,
|
| 1676 |
+
"step": 2060
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 25.01772151898734,
|
| 1680 |
+
"grad_norm": 0.0064802165143191814,
|
| 1681 |
+
"learning_rate": 5.28832630098453e-06,
|
| 1682 |
+
"loss": 0.0007,
|
| 1683 |
+
"step": 2070
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 25.020253164556962,
|
| 1687 |
+
"grad_norm": 0.07768774032592773,
|
| 1688 |
+
"learning_rate": 5.260196905766527e-06,
|
| 1689 |
+
"loss": 0.0023,
|
| 1690 |
+
"step": 2080
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 25.020253164556962,
|
| 1694 |
+
"eval_accuracy": 0.425531914893617,
|
| 1695 |
+
"eval_loss": 4.050126552581787,
|
| 1696 |
+
"eval_runtime": 8.4727,
|
| 1697 |
+
"eval_samples_per_second": 5.547,
|
| 1698 |
+
"eval_steps_per_second": 1.416,
|
| 1699 |
+
"step": 2080
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 26.00253164556962,
|
| 1703 |
+
"grad_norm": 0.029080139473080635,
|
| 1704 |
+
"learning_rate": 5.2320675105485245e-06,
|
| 1705 |
+
"loss": 0.0005,
|
| 1706 |
+
"step": 2090
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 26.00506329113924,
|
| 1710 |
+
"grad_norm": 0.03519600257277489,
|
| 1711 |
+
"learning_rate": 5.203938115330521e-06,
|
| 1712 |
+
"loss": 0.0013,
|
| 1713 |
+
"step": 2100
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 26.00759493670886,
|
| 1717 |
+
"grad_norm": 0.020978575572371483,
|
| 1718 |
+
"learning_rate": 5.175808720112518e-06,
|
| 1719 |
+
"loss": 0.0182,
|
| 1720 |
+
"step": 2110
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 26.01012658227848,
|
| 1724 |
+
"grad_norm": 0.025504613295197487,
|
| 1725 |
+
"learning_rate": 5.147679324894516e-06,
|
| 1726 |
+
"loss": 0.001,
|
| 1727 |
+
"step": 2120
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 26.0126582278481,
|
| 1731 |
+
"grad_norm": 0.04360447824001312,
|
| 1732 |
+
"learning_rate": 5.1195499296765125e-06,
|
| 1733 |
+
"loss": 0.0006,
|
| 1734 |
+
"step": 2130
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 26.01518987341772,
|
| 1738 |
+
"grad_norm": 189.72640991210938,
|
| 1739 |
+
"learning_rate": 5.09142053445851e-06,
|
| 1740 |
+
"loss": 0.0226,
|
| 1741 |
+
"step": 2140
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 26.01772151898734,
|
| 1745 |
+
"grad_norm": 0.011660604737699032,
|
| 1746 |
+
"learning_rate": 5.063291139240507e-06,
|
| 1747 |
+
"loss": 0.0007,
|
| 1748 |
+
"step": 2150
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 26.020253164556962,
|
| 1752 |
+
"grad_norm": 0.008466072380542755,
|
| 1753 |
+
"learning_rate": 5.035161744022505e-06,
|
| 1754 |
+
"loss": 0.0009,
|
| 1755 |
+
"step": 2160
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 26.020253164556962,
|
| 1759 |
+
"eval_accuracy": 0.3404255319148936,
|
| 1760 |
+
"eval_loss": 4.20005989074707,
|
| 1761 |
+
"eval_runtime": 8.5061,
|
| 1762 |
+
"eval_samples_per_second": 5.525,
|
| 1763 |
+
"eval_steps_per_second": 1.411,
|
| 1764 |
+
"step": 2160
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 27.00253164556962,
|
| 1768 |
+
"grad_norm": 0.01523934118449688,
|
| 1769 |
+
"learning_rate": 5.007032348804501e-06,
|
| 1770 |
+
"loss": 0.0009,
|
| 1771 |
+
"step": 2170
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 27.00506329113924,
|
| 1775 |
+
"grad_norm": 549.9719848632812,
|
| 1776 |
+
"learning_rate": 4.978902953586498e-06,
|
| 1777 |
+
"loss": 0.099,
|
| 1778 |
+
"step": 2180
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 27.00759493670886,
|
| 1782 |
+
"grad_norm": 33.501705169677734,
|
| 1783 |
+
"learning_rate": 4.950773558368496e-06,
|
| 1784 |
+
"loss": 0.0368,
|
| 1785 |
+
"step": 2190
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 27.01012658227848,
|
| 1789 |
+
"grad_norm": 0.01092901173979044,
|
| 1790 |
+
"learning_rate": 4.922644163150493e-06,
|
| 1791 |
+
"loss": 0.0184,
|
| 1792 |
+
"step": 2200
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 27.0126582278481,
|
| 1796 |
+
"grad_norm": 0.4083053469657898,
|
| 1797 |
+
"learning_rate": 4.89451476793249e-06,
|
| 1798 |
+
"loss": 0.0218,
|
| 1799 |
+
"step": 2210
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 27.01518987341772,
|
| 1803 |
+
"grad_norm": 0.025765903294086456,
|
| 1804 |
+
"learning_rate": 4.866385372714487e-06,
|
| 1805 |
+
"loss": 0.0142,
|
| 1806 |
+
"step": 2220
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 27.01772151898734,
|
| 1810 |
+
"grad_norm": 0.07521297037601471,
|
| 1811 |
+
"learning_rate": 4.838255977496485e-06,
|
| 1812 |
+
"loss": 0.0005,
|
| 1813 |
+
"step": 2230
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 27.020253164556962,
|
| 1817 |
+
"grad_norm": 0.6407820582389832,
|
| 1818 |
+
"learning_rate": 4.8101265822784815e-06,
|
| 1819 |
+
"loss": 0.0004,
|
| 1820 |
+
"step": 2240
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 27.020253164556962,
|
| 1824 |
+
"eval_accuracy": 0.3829787234042553,
|
| 1825 |
+
"eval_loss": 4.0130414962768555,
|
| 1826 |
+
"eval_runtime": 8.4723,
|
| 1827 |
+
"eval_samples_per_second": 5.547,
|
| 1828 |
+
"eval_steps_per_second": 1.416,
|
| 1829 |
+
"step": 2240
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 28.00253164556962,
|
| 1833 |
+
"grad_norm": 0.0875251367688179,
|
| 1834 |
+
"learning_rate": 4.781997187060478e-06,
|
| 1835 |
+
"loss": 0.0005,
|
| 1836 |
+
"step": 2250
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 28.00506329113924,
|
| 1840 |
+
"grad_norm": 0.022297605872154236,
|
| 1841 |
+
"learning_rate": 4.753867791842476e-06,
|
| 1842 |
+
"loss": 0.0032,
|
| 1843 |
+
"step": 2260
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 28.00759493670886,
|
| 1847 |
+
"grad_norm": 0.032631952315568924,
|
| 1848 |
+
"learning_rate": 4.725738396624473e-06,
|
| 1849 |
+
"loss": 0.0005,
|
| 1850 |
+
"step": 2270
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 28.01012658227848,
|
| 1854 |
+
"grad_norm": 0.01031473744660616,
|
| 1855 |
+
"learning_rate": 4.6976090014064704e-06,
|
| 1856 |
+
"loss": 0.0005,
|
| 1857 |
+
"step": 2280
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 28.0126582278481,
|
| 1861 |
+
"grad_norm": 0.016337795183062553,
|
| 1862 |
+
"learning_rate": 4.669479606188467e-06,
|
| 1863 |
+
"loss": 0.0008,
|
| 1864 |
+
"step": 2290
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 28.01518987341772,
|
| 1868 |
+
"grad_norm": 0.006589374039322138,
|
| 1869 |
+
"learning_rate": 4.641350210970465e-06,
|
| 1870 |
+
"loss": 0.0005,
|
| 1871 |
+
"step": 2300
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 28.01772151898734,
|
| 1875 |
+
"grad_norm": 0.08726366609334946,
|
| 1876 |
+
"learning_rate": 4.613220815752462e-06,
|
| 1877 |
+
"loss": 0.0008,
|
| 1878 |
+
"step": 2310
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 28.020253164556962,
|
| 1882 |
+
"grad_norm": 0.019069937989115715,
|
| 1883 |
+
"learning_rate": 4.5850914205344585e-06,
|
| 1884 |
+
"loss": 0.0162,
|
| 1885 |
+
"step": 2320
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 28.020253164556962,
|
| 1889 |
+
"eval_accuracy": 0.40425531914893614,
|
| 1890 |
+
"eval_loss": 4.046835422515869,
|
| 1891 |
+
"eval_runtime": 8.4543,
|
| 1892 |
+
"eval_samples_per_second": 5.559,
|
| 1893 |
+
"eval_steps_per_second": 1.419,
|
| 1894 |
+
"step": 2320
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 29.00253164556962,
|
| 1898 |
+
"grad_norm": 0.021134885028004646,
|
| 1899 |
+
"learning_rate": 4.556962025316456e-06,
|
| 1900 |
+
"loss": 0.0003,
|
| 1901 |
+
"step": 2330
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 29.00506329113924,
|
| 1905 |
+
"grad_norm": 0.01198391243815422,
|
| 1906 |
+
"learning_rate": 4.528832630098453e-06,
|
| 1907 |
+
"loss": 0.0202,
|
| 1908 |
+
"step": 2340
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 29.00759493670886,
|
| 1912 |
+
"grad_norm": 0.0206964910030365,
|
| 1913 |
+
"learning_rate": 4.5007032348804506e-06,
|
| 1914 |
+
"loss": 0.0003,
|
| 1915 |
+
"step": 2350
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 29.01012658227848,
|
| 1919 |
+
"grad_norm": 0.012637944892048836,
|
| 1920 |
+
"learning_rate": 4.472573839662447e-06,
|
| 1921 |
+
"loss": 0.0353,
|
| 1922 |
+
"step": 2360
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 29.0126582278481,
|
| 1926 |
+
"grad_norm": 0.0505475290119648,
|
| 1927 |
+
"learning_rate": 4.444444444444444e-06,
|
| 1928 |
+
"loss": 0.0007,
|
| 1929 |
+
"step": 2370
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 29.01518987341772,
|
| 1933 |
+
"grad_norm": 0.008409290574491024,
|
| 1934 |
+
"learning_rate": 4.416315049226442e-06,
|
| 1935 |
+
"loss": 0.0006,
|
| 1936 |
+
"step": 2380
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 29.01772151898734,
|
| 1940 |
+
"grad_norm": 1.1566442251205444,
|
| 1941 |
+
"learning_rate": 4.3881856540084394e-06,
|
| 1942 |
+
"loss": 0.0004,
|
| 1943 |
+
"step": 2390
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 29.020253164556962,
|
| 1947 |
+
"grad_norm": 0.012755032628774643,
|
| 1948 |
+
"learning_rate": 4.360056258790436e-06,
|
| 1949 |
+
"loss": 0.0073,
|
| 1950 |
+
"step": 2400
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 29.020253164556962,
|
| 1954 |
+
"eval_accuracy": 0.425531914893617,
|
| 1955 |
+
"eval_loss": 4.191919803619385,
|
| 1956 |
+
"eval_runtime": 8.4177,
|
| 1957 |
+
"eval_samples_per_second": 5.584,
|
| 1958 |
+
"eval_steps_per_second": 1.426,
|
| 1959 |
+
"step": 2400
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 30.00253164556962,
|
| 1963 |
+
"grad_norm": 0.009344339370727539,
|
| 1964 |
+
"learning_rate": 4.331926863572434e-06,
|
| 1965 |
+
"loss": 0.0004,
|
| 1966 |
+
"step": 2410
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 30.00506329113924,
|
| 1970 |
+
"grad_norm": 0.007806244306266308,
|
| 1971 |
+
"learning_rate": 4.303797468354431e-06,
|
| 1972 |
+
"loss": 0.0004,
|
| 1973 |
+
"step": 2420
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 30.00759493670886,
|
| 1977 |
+
"grad_norm": 0.006953861564397812,
|
| 1978 |
+
"learning_rate": 4.275668073136428e-06,
|
| 1979 |
+
"loss": 0.0009,
|
| 1980 |
+
"step": 2430
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 30.01012658227848,
|
| 1984 |
+
"grad_norm": 0.011007839813828468,
|
| 1985 |
+
"learning_rate": 4.247538677918425e-06,
|
| 1986 |
+
"loss": 0.0002,
|
| 1987 |
+
"step": 2440
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 30.0126582278481,
|
| 1991 |
+
"grad_norm": 0.007067366968840361,
|
| 1992 |
+
"learning_rate": 4.219409282700423e-06,
|
| 1993 |
+
"loss": 0.0007,
|
| 1994 |
+
"step": 2450
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 30.01518987341772,
|
| 1998 |
+
"grad_norm": 0.0060151745565235615,
|
| 1999 |
+
"learning_rate": 4.1912798874824196e-06,
|
| 2000 |
+
"loss": 0.0004,
|
| 2001 |
+
"step": 2460
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 30.01772151898734,
|
| 2005 |
+
"grad_norm": 0.006391868926584721,
|
| 2006 |
+
"learning_rate": 4.163150492264416e-06,
|
| 2007 |
+
"loss": 0.0006,
|
| 2008 |
+
"step": 2470
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 30.020253164556962,
|
| 2012 |
+
"grad_norm": 0.018146460875868797,
|
| 2013 |
+
"learning_rate": 4.135021097046414e-06,
|
| 2014 |
+
"loss": 0.0012,
|
| 2015 |
+
"step": 2480
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 30.020253164556962,
|
| 2019 |
+
"eval_accuracy": 0.425531914893617,
|
| 2020 |
+
"eval_loss": 4.000429630279541,
|
| 2021 |
+
"eval_runtime": 8.4099,
|
| 2022 |
+
"eval_samples_per_second": 5.589,
|
| 2023 |
+
"eval_steps_per_second": 1.427,
|
| 2024 |
+
"step": 2480
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 31.00253164556962,
|
| 2028 |
+
"grad_norm": 0.004094243515282869,
|
| 2029 |
+
"learning_rate": 4.106891701828411e-06,
|
| 2030 |
+
"loss": 0.0004,
|
| 2031 |
+
"step": 2490
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 31.00506329113924,
|
| 2035 |
+
"grad_norm": 0.013013644143939018,
|
| 2036 |
+
"learning_rate": 4.0787623066104084e-06,
|
| 2037 |
+
"loss": 0.0412,
|
| 2038 |
+
"step": 2500
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 31.00759493670886,
|
| 2042 |
+
"grad_norm": 0.004616581369191408,
|
| 2043 |
+
"learning_rate": 4.050632911392405e-06,
|
| 2044 |
+
"loss": 0.0006,
|
| 2045 |
+
"step": 2510
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 31.01012658227848,
|
| 2049 |
+
"grad_norm": 0.09256187826395035,
|
| 2050 |
+
"learning_rate": 4.022503516174403e-06,
|
| 2051 |
+
"loss": 0.0006,
|
| 2052 |
+
"step": 2520
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 31.0126582278481,
|
| 2056 |
+
"grad_norm": 0.013411607593297958,
|
| 2057 |
+
"learning_rate": 3.9943741209564e-06,
|
| 2058 |
+
"loss": 0.0004,
|
| 2059 |
+
"step": 2530
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 31.01518987341772,
|
| 2063 |
+
"grad_norm": 0.011695249006152153,
|
| 2064 |
+
"learning_rate": 3.9662447257383965e-06,
|
| 2065 |
+
"loss": 0.0007,
|
| 2066 |
+
"step": 2540
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 31.01772151898734,
|
| 2070 |
+
"grad_norm": 0.04082596302032471,
|
| 2071 |
+
"learning_rate": 3.938115330520394e-06,
|
| 2072 |
+
"loss": 0.0003,
|
| 2073 |
+
"step": 2550
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 31.020253164556962,
|
| 2077 |
+
"grad_norm": 0.01295529119670391,
|
| 2078 |
+
"learning_rate": 3.909985935302391e-06,
|
| 2079 |
+
"loss": 0.0125,
|
| 2080 |
+
"step": 2560
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 31.020253164556962,
|
| 2084 |
+
"eval_accuracy": 0.3829787234042553,
|
| 2085 |
+
"eval_loss": 4.115077972412109,
|
| 2086 |
+
"eval_runtime": 8.4877,
|
| 2087 |
+
"eval_samples_per_second": 5.537,
|
| 2088 |
+
"eval_steps_per_second": 1.414,
|
| 2089 |
+
"step": 2560
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 32.00253164556962,
|
| 2093 |
+
"grad_norm": 34.59695816040039,
|
| 2094 |
+
"learning_rate": 3.8818565400843886e-06,
|
| 2095 |
+
"loss": 0.004,
|
| 2096 |
+
"step": 2570
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 32.00506329113924,
|
| 2100 |
+
"grad_norm": 0.03401665389537811,
|
| 2101 |
+
"learning_rate": 3.853727144866385e-06,
|
| 2102 |
+
"loss": 0.0927,
|
| 2103 |
+
"step": 2580
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 32.00759493670886,
|
| 2107 |
+
"grad_norm": 0.06631805002689362,
|
| 2108 |
+
"learning_rate": 3.825597749648383e-06,
|
| 2109 |
+
"loss": 0.0003,
|
| 2110 |
+
"step": 2590
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 32.01012658227848,
|
| 2114 |
+
"grad_norm": 0.006998293101787567,
|
| 2115 |
+
"learning_rate": 3.7974683544303802e-06,
|
| 2116 |
+
"loss": 0.0006,
|
| 2117 |
+
"step": 2600
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 32.0126582278481,
|
| 2121 |
+
"grad_norm": 0.006872397381812334,
|
| 2122 |
+
"learning_rate": 3.7693389592123775e-06,
|
| 2123 |
+
"loss": 0.0003,
|
| 2124 |
+
"step": 2610
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 32.01518987341772,
|
| 2128 |
+
"grad_norm": 0.09216930717229843,
|
| 2129 |
+
"learning_rate": 3.7412095639943747e-06,
|
| 2130 |
+
"loss": 0.0005,
|
| 2131 |
+
"step": 2620
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 32.017721518987344,
|
| 2135 |
+
"grad_norm": 0.008273018524050713,
|
| 2136 |
+
"learning_rate": 3.713080168776372e-06,
|
| 2137 |
+
"loss": 0.0003,
|
| 2138 |
+
"step": 2630
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 32.02025316455696,
|
| 2142 |
+
"grad_norm": 0.01438277680426836,
|
| 2143 |
+
"learning_rate": 3.684950773558369e-06,
|
| 2144 |
+
"loss": 0.0005,
|
| 2145 |
+
"step": 2640
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 32.02025316455696,
|
| 2149 |
+
"eval_accuracy": 0.3829787234042553,
|
| 2150 |
+
"eval_loss": 4.328236103057861,
|
| 2151 |
+
"eval_runtime": 8.4295,
|
| 2152 |
+
"eval_samples_per_second": 5.576,
|
| 2153 |
+
"eval_steps_per_second": 1.424,
|
| 2154 |
+
"step": 2640
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 33.00253164556962,
|
| 2158 |
+
"grad_norm": 0.007496337406337261,
|
| 2159 |
+
"learning_rate": 3.656821378340366e-06,
|
| 2160 |
+
"loss": 0.0003,
|
| 2161 |
+
"step": 2650
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 33.00506329113924,
|
| 2165 |
+
"grad_norm": 0.024606458842754364,
|
| 2166 |
+
"learning_rate": 3.628691983122363e-06,
|
| 2167 |
+
"loss": 0.0004,
|
| 2168 |
+
"step": 2660
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 33.00759493670886,
|
| 2172 |
+
"grad_norm": 116.44566345214844,
|
| 2173 |
+
"learning_rate": 3.6005625879043604e-06,
|
| 2174 |
+
"loss": 0.1643,
|
| 2175 |
+
"step": 2670
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 33.01012658227848,
|
| 2179 |
+
"grad_norm": 0.23663385212421417,
|
| 2180 |
+
"learning_rate": 3.5724331926863576e-06,
|
| 2181 |
+
"loss": 0.0004,
|
| 2182 |
+
"step": 2680
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 33.0126582278481,
|
| 2186 |
+
"grad_norm": 0.20308883488178253,
|
| 2187 |
+
"learning_rate": 3.544303797468355e-06,
|
| 2188 |
+
"loss": 0.0037,
|
| 2189 |
+
"step": 2690
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 33.01518987341772,
|
| 2193 |
+
"grad_norm": 0.006437554955482483,
|
| 2194 |
+
"learning_rate": 3.516174402250352e-06,
|
| 2195 |
+
"loss": 0.0003,
|
| 2196 |
+
"step": 2700
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 33.017721518987344,
|
| 2200 |
+
"grad_norm": 0.01716802455484867,
|
| 2201 |
+
"learning_rate": 3.4880450070323492e-06,
|
| 2202 |
+
"loss": 0.0005,
|
| 2203 |
+
"step": 2710
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 33.02025316455696,
|
| 2207 |
+
"grad_norm": 0.009229468181729317,
|
| 2208 |
+
"learning_rate": 3.459915611814346e-06,
|
| 2209 |
+
"loss": 0.0098,
|
| 2210 |
+
"step": 2720
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 33.02025316455696,
|
| 2214 |
+
"eval_accuracy": 0.3829787234042553,
|
| 2215 |
+
"eval_loss": 4.468872547149658,
|
| 2216 |
+
"eval_runtime": 8.4658,
|
| 2217 |
+
"eval_samples_per_second": 5.552,
|
| 2218 |
+
"eval_steps_per_second": 1.417,
|
| 2219 |
+
"step": 2720
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 34.00253164556962,
|
| 2223 |
+
"grad_norm": 0.024370867758989334,
|
| 2224 |
+
"learning_rate": 3.4317862165963433e-06,
|
| 2225 |
+
"loss": 0.0006,
|
| 2226 |
+
"step": 2730
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 34.00506329113924,
|
| 2230 |
+
"grad_norm": 0.009673592634499073,
|
| 2231 |
+
"learning_rate": 3.4036568213783405e-06,
|
| 2232 |
+
"loss": 0.0003,
|
| 2233 |
+
"step": 2740
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 34.00759493670886,
|
| 2237 |
+
"grad_norm": 0.02182767540216446,
|
| 2238 |
+
"learning_rate": 3.3755274261603377e-06,
|
| 2239 |
+
"loss": 0.0018,
|
| 2240 |
+
"step": 2750
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 34.01012658227848,
|
| 2244 |
+
"grad_norm": 0.014191425405442715,
|
| 2245 |
+
"learning_rate": 3.347398030942335e-06,
|
| 2246 |
+
"loss": 0.0003,
|
| 2247 |
+
"step": 2760
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 34.0126582278481,
|
| 2251 |
+
"grad_norm": 0.008674095384776592,
|
| 2252 |
+
"learning_rate": 3.319268635724332e-06,
|
| 2253 |
+
"loss": 0.0003,
|
| 2254 |
+
"step": 2770
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 34.01518987341772,
|
| 2258 |
+
"grad_norm": 0.011231840588152409,
|
| 2259 |
+
"learning_rate": 3.2911392405063294e-06,
|
| 2260 |
+
"loss": 0.0002,
|
| 2261 |
+
"step": 2780
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 34.017721518987344,
|
| 2265 |
+
"grad_norm": 0.008253734558820724,
|
| 2266 |
+
"learning_rate": 3.263009845288326e-06,
|
| 2267 |
+
"loss": 0.0002,
|
| 2268 |
+
"step": 2790
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 34.02025316455696,
|
| 2272 |
+
"grad_norm": 0.00744025781750679,
|
| 2273 |
+
"learning_rate": 3.2348804500703242e-06,
|
| 2274 |
+
"loss": 0.0036,
|
| 2275 |
+
"step": 2800
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 34.02025316455696,
|
| 2279 |
+
"eval_accuracy": 0.40425531914893614,
|
| 2280 |
+
"eval_loss": 4.335427284240723,
|
| 2281 |
+
"eval_runtime": 8.4069,
|
| 2282 |
+
"eval_samples_per_second": 5.591,
|
| 2283 |
+
"eval_steps_per_second": 1.427,
|
| 2284 |
+
"step": 2800
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 35.00253164556962,
|
| 2288 |
+
"grad_norm": 0.011312518268823624,
|
| 2289 |
+
"learning_rate": 3.206751054852321e-06,
|
| 2290 |
+
"loss": 0.0003,
|
| 2291 |
+
"step": 2810
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 35.00506329113924,
|
| 2295 |
+
"grad_norm": 0.004388440866023302,
|
| 2296 |
+
"learning_rate": 3.1786216596343183e-06,
|
| 2297 |
+
"loss": 0.0002,
|
| 2298 |
+
"step": 2820
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 35.00759493670886,
|
| 2302 |
+
"grad_norm": 0.23399221897125244,
|
| 2303 |
+
"learning_rate": 3.1504922644163155e-06,
|
| 2304 |
+
"loss": 0.0003,
|
| 2305 |
+
"step": 2830
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 35.01012658227848,
|
| 2309 |
+
"grad_norm": 0.02679286152124405,
|
| 2310 |
+
"learning_rate": 3.1223628691983127e-06,
|
| 2311 |
+
"loss": 0.0002,
|
| 2312 |
+
"step": 2840
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 35.0126582278481,
|
| 2316 |
+
"grad_norm": 0.007838438265025616,
|
| 2317 |
+
"learning_rate": 3.09423347398031e-06,
|
| 2318 |
+
"loss": 0.0003,
|
| 2319 |
+
"step": 2850
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 35.01518987341772,
|
| 2323 |
+
"grad_norm": 0.007448969408869743,
|
| 2324 |
+
"learning_rate": 3.066104078762307e-06,
|
| 2325 |
+
"loss": 0.0002,
|
| 2326 |
+
"step": 2860
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 35.017721518987344,
|
| 2330 |
+
"grad_norm": 0.009503643028438091,
|
| 2331 |
+
"learning_rate": 3.037974683544304e-06,
|
| 2332 |
+
"loss": 0.0002,
|
| 2333 |
+
"step": 2870
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 35.02025316455696,
|
| 2337 |
+
"grad_norm": 0.004561970010399818,
|
| 2338 |
+
"learning_rate": 3.009845288326301e-06,
|
| 2339 |
+
"loss": 0.0002,
|
| 2340 |
+
"step": 2880
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 35.02025316455696,
|
| 2344 |
+
"eval_accuracy": 0.3617021276595745,
|
| 2345 |
+
"eval_loss": 4.460489273071289,
|
| 2346 |
+
"eval_runtime": 8.4509,
|
| 2347 |
+
"eval_samples_per_second": 5.562,
|
| 2348 |
+
"eval_steps_per_second": 1.42,
|
| 2349 |
+
"step": 2880
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 36.00253164556962,
|
| 2353 |
+
"grad_norm": 0.008112301118671894,
|
| 2354 |
+
"learning_rate": 2.9817158931082984e-06,
|
| 2355 |
+
"loss": 0.0396,
|
| 2356 |
+
"step": 2890
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 36.00506329113924,
|
| 2360 |
+
"grad_norm": 0.0911855548620224,
|
| 2361 |
+
"learning_rate": 2.9535864978902956e-06,
|
| 2362 |
+
"loss": 0.0004,
|
| 2363 |
+
"step": 2900
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 36.00759493670886,
|
| 2367 |
+
"grad_norm": 0.11559173464775085,
|
| 2368 |
+
"learning_rate": 2.925457102672293e-06,
|
| 2369 |
+
"loss": 0.0004,
|
| 2370 |
+
"step": 2910
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 36.01012658227848,
|
| 2374 |
+
"grad_norm": 0.014976758509874344,
|
| 2375 |
+
"learning_rate": 2.89732770745429e-06,
|
| 2376 |
+
"loss": 0.0002,
|
| 2377 |
+
"step": 2920
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 36.0126582278481,
|
| 2381 |
+
"grad_norm": 0.05852414667606354,
|
| 2382 |
+
"learning_rate": 2.8691983122362873e-06,
|
| 2383 |
+
"loss": 0.0003,
|
| 2384 |
+
"step": 2930
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 36.01518987341772,
|
| 2388 |
+
"grad_norm": 0.01464304979890585,
|
| 2389 |
+
"learning_rate": 2.841068917018284e-06,
|
| 2390 |
+
"loss": 0.0003,
|
| 2391 |
+
"step": 2940
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 36.017721518987344,
|
| 2395 |
+
"grad_norm": 0.008967959322035313,
|
| 2396 |
+
"learning_rate": 2.8129395218002813e-06,
|
| 2397 |
+
"loss": 0.0002,
|
| 2398 |
+
"step": 2950
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 36.02025316455696,
|
| 2402 |
+
"grad_norm": 0.005782134830951691,
|
| 2403 |
+
"learning_rate": 2.7848101265822785e-06,
|
| 2404 |
+
"loss": 0.0002,
|
| 2405 |
+
"step": 2960
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 36.02025316455696,
|
| 2409 |
+
"eval_accuracy": 0.425531914893617,
|
| 2410 |
+
"eval_loss": 4.158644199371338,
|
| 2411 |
+
"eval_runtime": 8.4415,
|
| 2412 |
+
"eval_samples_per_second": 5.568,
|
| 2413 |
+
"eval_steps_per_second": 1.422,
|
| 2414 |
+
"step": 2960
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"epoch": 37.00253164556962,
|
| 2418 |
+
"grad_norm": 0.011933619156479836,
|
| 2419 |
+
"learning_rate": 2.7566807313642757e-06,
|
| 2420 |
+
"loss": 0.0006,
|
| 2421 |
+
"step": 2970
|
| 2422 |
+
},
|
| 2423 |
+
{
|
| 2424 |
+
"epoch": 37.00506329113924,
|
| 2425 |
+
"grad_norm": 0.025380389764904976,
|
| 2426 |
+
"learning_rate": 2.728551336146273e-06,
|
| 2427 |
+
"loss": 0.0002,
|
| 2428 |
+
"step": 2980
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 37.00759493670886,
|
| 2432 |
+
"grad_norm": 0.006639150902628899,
|
| 2433 |
+
"learning_rate": 2.70042194092827e-06,
|
| 2434 |
+
"loss": 0.0175,
|
| 2435 |
+
"step": 2990
|
| 2436 |
+
},
|
| 2437 |
+
{
|
| 2438 |
+
"epoch": 37.01012658227848,
|
| 2439 |
+
"grad_norm": 0.005557660013437271,
|
| 2440 |
+
"learning_rate": 2.6722925457102674e-06,
|
| 2441 |
+
"loss": 0.0004,
|
| 2442 |
+
"step": 3000
|
| 2443 |
+
},
|
| 2444 |
+
{
|
| 2445 |
+
"epoch": 37.0126582278481,
|
| 2446 |
+
"grad_norm": 0.013791055418550968,
|
| 2447 |
+
"learning_rate": 2.644163150492265e-06,
|
| 2448 |
+
"loss": 0.0003,
|
| 2449 |
+
"step": 3010
|
| 2450 |
+
},
|
| 2451 |
+
{
|
| 2452 |
+
"epoch": 37.01518987341772,
|
| 2453 |
+
"grad_norm": 0.007814054377377033,
|
| 2454 |
+
"learning_rate": 2.6160337552742622e-06,
|
| 2455 |
+
"loss": 0.0653,
|
| 2456 |
+
"step": 3020
|
| 2457 |
+
},
|
| 2458 |
+
{
|
| 2459 |
+
"epoch": 37.017721518987344,
|
| 2460 |
+
"grad_norm": 0.0073919533751904964,
|
| 2461 |
+
"learning_rate": 2.587904360056259e-06,
|
| 2462 |
+
"loss": 0.0003,
|
| 2463 |
+
"step": 3030
|
| 2464 |
+
},
|
| 2465 |
+
{
|
| 2466 |
+
"epoch": 37.02025316455696,
|
| 2467 |
+
"grad_norm": 0.006769287399947643,
|
| 2468 |
+
"learning_rate": 2.5597749648382563e-06,
|
| 2469 |
+
"loss": 0.0002,
|
| 2470 |
+
"step": 3040
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 37.02025316455696,
|
| 2474 |
+
"eval_accuracy": 0.40425531914893614,
|
| 2475 |
+
"eval_loss": 4.257404804229736,
|
| 2476 |
+
"eval_runtime": 8.5782,
|
| 2477 |
+
"eval_samples_per_second": 5.479,
|
| 2478 |
+
"eval_steps_per_second": 1.399,
|
| 2479 |
+
"step": 3040
|
| 2480 |
+
},
|
| 2481 |
+
{
|
| 2482 |
+
"epoch": 38.00253164556962,
|
| 2483 |
+
"grad_norm": 0.012128209695219994,
|
| 2484 |
+
"learning_rate": 2.5316455696202535e-06,
|
| 2485 |
+
"loss": 0.0002,
|
| 2486 |
+
"step": 3050
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 38.00506329113924,
|
| 2490 |
+
"grad_norm": 0.004270182456821203,
|
| 2491 |
+
"learning_rate": 2.5035161744022507e-06,
|
| 2492 |
+
"loss": 0.0002,
|
| 2493 |
+
"step": 3060
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 38.00759493670886,
|
| 2497 |
+
"grad_norm": 0.02535262331366539,
|
| 2498 |
+
"learning_rate": 2.475386779184248e-06,
|
| 2499 |
+
"loss": 0.0002,
|
| 2500 |
+
"step": 3070
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 38.01012658227848,
|
| 2504 |
+
"grad_norm": 0.00449047377333045,
|
| 2505 |
+
"learning_rate": 2.447257383966245e-06,
|
| 2506 |
+
"loss": 0.0004,
|
| 2507 |
+
"step": 3080
|
| 2508 |
+
},
|
| 2509 |
+
{
|
| 2510 |
+
"epoch": 38.0126582278481,
|
| 2511 |
+
"grad_norm": 0.010649596340954304,
|
| 2512 |
+
"learning_rate": 2.4191279887482424e-06,
|
| 2513 |
+
"loss": 0.0002,
|
| 2514 |
+
"step": 3090
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 38.01518987341772,
|
| 2518 |
+
"grad_norm": 0.0074311415664851665,
|
| 2519 |
+
"learning_rate": 2.390998593530239e-06,
|
| 2520 |
+
"loss": 0.0004,
|
| 2521 |
+
"step": 3100
|
| 2522 |
+
},
|
| 2523 |
+
{
|
| 2524 |
+
"epoch": 38.017721518987344,
|
| 2525 |
+
"grad_norm": 0.012918173335492611,
|
| 2526 |
+
"learning_rate": 2.3628691983122364e-06,
|
| 2527 |
+
"loss": 0.0002,
|
| 2528 |
+
"step": 3110
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 38.02025316455696,
|
| 2532 |
+
"grad_norm": 0.13585978746414185,
|
| 2533 |
+
"learning_rate": 2.3347398030942336e-06,
|
| 2534 |
+
"loss": 0.0002,
|
| 2535 |
+
"step": 3120
|
| 2536 |
+
},
|
| 2537 |
+
{
|
| 2538 |
+
"epoch": 38.02025316455696,
|
| 2539 |
+
"eval_accuracy": 0.3829787234042553,
|
| 2540 |
+
"eval_loss": 4.639108180999756,
|
| 2541 |
+
"eval_runtime": 8.4222,
|
| 2542 |
+
"eval_samples_per_second": 5.581,
|
| 2543 |
+
"eval_steps_per_second": 1.425,
|
| 2544 |
+
"step": 3120
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 39.00253164556962,
|
| 2548 |
+
"grad_norm": 0.006208473350852728,
|
| 2549 |
+
"learning_rate": 2.306610407876231e-06,
|
| 2550 |
+
"loss": 0.0005,
|
| 2551 |
+
"step": 3130
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 39.00506329113924,
|
| 2555 |
+
"grad_norm": 0.008609656244516373,
|
| 2556 |
+
"learning_rate": 2.278481012658228e-06,
|
| 2557 |
+
"loss": 0.0218,
|
| 2558 |
+
"step": 3140
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 39.00759493670886,
|
| 2562 |
+
"grad_norm": 0.015961425378918648,
|
| 2563 |
+
"learning_rate": 2.2503516174402253e-06,
|
| 2564 |
+
"loss": 0.0003,
|
| 2565 |
+
"step": 3150
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 39.01012658227848,
|
| 2569 |
+
"grad_norm": 0.0073599363677203655,
|
| 2570 |
+
"learning_rate": 2.222222222222222e-06,
|
| 2571 |
+
"loss": 0.0002,
|
| 2572 |
+
"step": 3160
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 39.0126582278481,
|
| 2576 |
+
"grad_norm": 0.0034563420340418816,
|
| 2577 |
+
"learning_rate": 2.1940928270042197e-06,
|
| 2578 |
+
"loss": 0.0001,
|
| 2579 |
+
"step": 3170
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 39.01518987341772,
|
| 2583 |
+
"grad_norm": 0.007735641673207283,
|
| 2584 |
+
"learning_rate": 2.165963431786217e-06,
|
| 2585 |
+
"loss": 0.0002,
|
| 2586 |
+
"step": 3180
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 39.017721518987344,
|
| 2590 |
+
"grad_norm": 0.007485373876988888,
|
| 2591 |
+
"learning_rate": 2.137834036568214e-06,
|
| 2592 |
+
"loss": 0.0001,
|
| 2593 |
+
"step": 3190
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 39.02025316455696,
|
| 2597 |
+
"grad_norm": 0.002605201443657279,
|
| 2598 |
+
"learning_rate": 2.1097046413502114e-06,
|
| 2599 |
+
"loss": 0.0008,
|
| 2600 |
+
"step": 3200
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 39.02025316455696,
|
| 2604 |
+
"eval_accuracy": 0.3617021276595745,
|
| 2605 |
+
"eval_loss": 4.552647113800049,
|
| 2606 |
+
"eval_runtime": 8.4611,
|
| 2607 |
+
"eval_samples_per_second": 5.555,
|
| 2608 |
+
"eval_steps_per_second": 1.418,
|
| 2609 |
+
"step": 3200
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 40.00253164556962,
|
| 2613 |
+
"grad_norm": 0.007657783105969429,
|
| 2614 |
+
"learning_rate": 2.081575246132208e-06,
|
| 2615 |
+
"loss": 0.0002,
|
| 2616 |
+
"step": 3210
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 40.00506329113924,
|
| 2620 |
+
"grad_norm": 0.006229284685105085,
|
| 2621 |
+
"learning_rate": 2.0534458509142054e-06,
|
| 2622 |
+
"loss": 0.0002,
|
| 2623 |
+
"step": 3220
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 40.00759493670886,
|
| 2627 |
+
"grad_norm": 0.0027491943910717964,
|
| 2628 |
+
"learning_rate": 2.0253164556962026e-06,
|
| 2629 |
+
"loss": 0.0002,
|
| 2630 |
+
"step": 3230
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 40.01012658227848,
|
| 2634 |
+
"grad_norm": 0.009845227934420109,
|
| 2635 |
+
"learning_rate": 1.9971870604782e-06,
|
| 2636 |
+
"loss": 0.0002,
|
| 2637 |
+
"step": 3240
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 40.0126582278481,
|
| 2641 |
+
"grad_norm": 0.00487924600020051,
|
| 2642 |
+
"learning_rate": 1.969057665260197e-06,
|
| 2643 |
+
"loss": 0.0002,
|
| 2644 |
+
"step": 3250
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 40.01518987341772,
|
| 2648 |
+
"grad_norm": 0.4248621463775635,
|
| 2649 |
+
"learning_rate": 1.9409282700421943e-06,
|
| 2650 |
+
"loss": 0.0002,
|
| 2651 |
+
"step": 3260
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 40.017721518987344,
|
| 2655 |
+
"grad_norm": 0.02595403790473938,
|
| 2656 |
+
"learning_rate": 1.9127988748241915e-06,
|
| 2657 |
+
"loss": 0.0002,
|
| 2658 |
+
"step": 3270
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 40.02025316455696,
|
| 2662 |
+
"grad_norm": 0.012371093034744263,
|
| 2663 |
+
"learning_rate": 1.8846694796061887e-06,
|
| 2664 |
+
"loss": 0.0001,
|
| 2665 |
+
"step": 3280
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 40.02025316455696,
|
| 2669 |
+
"eval_accuracy": 0.3829787234042553,
|
| 2670 |
+
"eval_loss": 4.565773963928223,
|
| 2671 |
+
"eval_runtime": 8.374,
|
| 2672 |
+
"eval_samples_per_second": 5.613,
|
| 2673 |
+
"eval_steps_per_second": 1.433,
|
| 2674 |
+
"step": 3280
|
| 2675 |
+
},
|
| 2676 |
+
{
|
| 2677 |
+
"epoch": 41.00253164556962,
|
| 2678 |
+
"grad_norm": 0.012245562858879566,
|
| 2679 |
+
"learning_rate": 1.856540084388186e-06,
|
| 2680 |
+
"loss": 0.0002,
|
| 2681 |
+
"step": 3290
|
| 2682 |
+
},
|
| 2683 |
+
{
|
| 2684 |
+
"epoch": 41.00506329113924,
|
| 2685 |
+
"grad_norm": 0.00565865496173501,
|
| 2686 |
+
"learning_rate": 1.828410689170183e-06,
|
| 2687 |
+
"loss": 0.0001,
|
| 2688 |
+
"step": 3300
|
| 2689 |
+
},
|
| 2690 |
+
{
|
| 2691 |
+
"epoch": 41.00759493670886,
|
| 2692 |
+
"grad_norm": 0.003963225055485964,
|
| 2693 |
+
"learning_rate": 1.8002812939521802e-06,
|
| 2694 |
+
"loss": 0.0002,
|
| 2695 |
+
"step": 3310
|
| 2696 |
+
},
|
| 2697 |
+
{
|
| 2698 |
+
"epoch": 41.01012658227848,
|
| 2699 |
+
"grad_norm": 0.0046040392480790615,
|
| 2700 |
+
"learning_rate": 1.7721518987341774e-06,
|
| 2701 |
+
"loss": 0.0001,
|
| 2702 |
+
"step": 3320
|
| 2703 |
+
},
|
| 2704 |
+
{
|
| 2705 |
+
"epoch": 41.0126582278481,
|
| 2706 |
+
"grad_norm": 0.00872375350445509,
|
| 2707 |
+
"learning_rate": 1.7440225035161746e-06,
|
| 2708 |
+
"loss": 0.0002,
|
| 2709 |
+
"step": 3330
|
| 2710 |
+
},
|
| 2711 |
+
{
|
| 2712 |
+
"epoch": 41.01518987341772,
|
| 2713 |
+
"grad_norm": 0.005013366229832172,
|
| 2714 |
+
"learning_rate": 1.7158931082981716e-06,
|
| 2715 |
+
"loss": 0.0002,
|
| 2716 |
+
"step": 3340
|
| 2717 |
+
},
|
| 2718 |
+
{
|
| 2719 |
+
"epoch": 41.017721518987344,
|
| 2720 |
+
"grad_norm": 0.0043080090545117855,
|
| 2721 |
+
"learning_rate": 1.6877637130801689e-06,
|
| 2722 |
+
"loss": 0.0002,
|
| 2723 |
+
"step": 3350
|
| 2724 |
+
},
|
| 2725 |
+
{
|
| 2726 |
+
"epoch": 41.02025316455696,
|
| 2727 |
+
"grad_norm": 0.0092382300645113,
|
| 2728 |
+
"learning_rate": 1.659634317862166e-06,
|
| 2729 |
+
"loss": 0.0001,
|
| 2730 |
+
"step": 3360
|
| 2731 |
+
},
|
| 2732 |
+
{
|
| 2733 |
+
"epoch": 41.02025316455696,
|
| 2734 |
+
"eval_accuracy": 0.3829787234042553,
|
| 2735 |
+
"eval_loss": 4.571171760559082,
|
| 2736 |
+
"eval_runtime": 8.4128,
|
| 2737 |
+
"eval_samples_per_second": 5.587,
|
| 2738 |
+
"eval_steps_per_second": 1.426,
|
| 2739 |
+
"step": 3360
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 42.00253164556962,
|
| 2743 |
+
"grad_norm": 0.029707299545407295,
|
| 2744 |
+
"learning_rate": 1.631504922644163e-06,
|
| 2745 |
+
"loss": 0.0002,
|
| 2746 |
+
"step": 3370
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 42.00506329113924,
|
| 2750 |
+
"grad_norm": 0.004554817918688059,
|
| 2751 |
+
"learning_rate": 1.6033755274261605e-06,
|
| 2752 |
+
"loss": 0.0002,
|
| 2753 |
+
"step": 3380
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 42.00759493670886,
|
| 2757 |
+
"grad_norm": 0.0048960368148982525,
|
| 2758 |
+
"learning_rate": 1.5752461322081577e-06,
|
| 2759 |
+
"loss": 0.0002,
|
| 2760 |
+
"step": 3390
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 42.01012658227848,
|
| 2764 |
+
"grad_norm": 0.004947884473949671,
|
| 2765 |
+
"learning_rate": 1.547116736990155e-06,
|
| 2766 |
+
"loss": 0.0002,
|
| 2767 |
+
"step": 3400
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 42.0126582278481,
|
| 2771 |
+
"grad_norm": 0.005734297446906567,
|
| 2772 |
+
"learning_rate": 1.518987341772152e-06,
|
| 2773 |
+
"loss": 0.0001,
|
| 2774 |
+
"step": 3410
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 42.01518987341772,
|
| 2778 |
+
"grad_norm": 0.00551837170496583,
|
| 2779 |
+
"learning_rate": 1.4908579465541492e-06,
|
| 2780 |
+
"loss": 0.0002,
|
| 2781 |
+
"step": 3420
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 42.017721518987344,
|
| 2785 |
+
"grad_norm": 0.03257923945784569,
|
| 2786 |
+
"learning_rate": 1.4627285513361464e-06,
|
| 2787 |
+
"loss": 0.0002,
|
| 2788 |
+
"step": 3430
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 42.02025316455696,
|
| 2792 |
+
"grad_norm": 0.00785852875560522,
|
| 2793 |
+
"learning_rate": 1.4345991561181436e-06,
|
| 2794 |
+
"loss": 0.0001,
|
| 2795 |
+
"step": 3440
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 42.02025316455696,
|
| 2799 |
+
"eval_accuracy": 0.3829787234042553,
|
| 2800 |
+
"eval_loss": 4.601936340332031,
|
| 2801 |
+
"eval_runtime": 8.4662,
|
| 2802 |
+
"eval_samples_per_second": 5.552,
|
| 2803 |
+
"eval_steps_per_second": 1.417,
|
| 2804 |
+
"step": 3440
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 43.00253164556962,
|
| 2808 |
+
"grad_norm": 0.004956628195941448,
|
| 2809 |
+
"learning_rate": 1.4064697609001406e-06,
|
| 2810 |
+
"loss": 0.0002,
|
| 2811 |
+
"step": 3450
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 43.00506329113924,
|
| 2815 |
+
"grad_norm": 0.002330609830096364,
|
| 2816 |
+
"learning_rate": 1.3783403656821379e-06,
|
| 2817 |
+
"loss": 0.0002,
|
| 2818 |
+
"step": 3460
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 43.00759493670886,
|
| 2822 |
+
"grad_norm": 0.004357906058430672,
|
| 2823 |
+
"learning_rate": 1.350210970464135e-06,
|
| 2824 |
+
"loss": 0.0002,
|
| 2825 |
+
"step": 3470
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 43.01012658227848,
|
| 2829 |
+
"grad_norm": 0.004682646133005619,
|
| 2830 |
+
"learning_rate": 1.3220815752461325e-06,
|
| 2831 |
+
"loss": 0.0001,
|
| 2832 |
+
"step": 3480
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 43.0126582278481,
|
| 2836 |
+
"grad_norm": 0.0037931317929178476,
|
| 2837 |
+
"learning_rate": 1.2939521800281295e-06,
|
| 2838 |
+
"loss": 0.0002,
|
| 2839 |
+
"step": 3490
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 43.01518987341772,
|
| 2843 |
+
"grad_norm": 0.004410360474139452,
|
| 2844 |
+
"learning_rate": 1.2658227848101267e-06,
|
| 2845 |
+
"loss": 0.0001,
|
| 2846 |
+
"step": 3500
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 43.017721518987344,
|
| 2850 |
+
"grad_norm": 0.0027343304827809334,
|
| 2851 |
+
"learning_rate": 1.237693389592124e-06,
|
| 2852 |
+
"loss": 0.0002,
|
| 2853 |
+
"step": 3510
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 43.02025316455696,
|
| 2857 |
+
"grad_norm": 0.007019107695668936,
|
| 2858 |
+
"learning_rate": 1.2095639943741212e-06,
|
| 2859 |
+
"loss": 0.0002,
|
| 2860 |
+
"step": 3520
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 43.02025316455696,
|
| 2864 |
+
"eval_accuracy": 0.40425531914893614,
|
| 2865 |
+
"eval_loss": 4.591475486755371,
|
| 2866 |
+
"eval_runtime": 8.4358,
|
| 2867 |
+
"eval_samples_per_second": 5.572,
|
| 2868 |
+
"eval_steps_per_second": 1.423,
|
| 2869 |
+
"step": 3520
|
| 2870 |
+
},
|
| 2871 |
+
{
|
| 2872 |
+
"epoch": 44.00253164556962,
|
| 2873 |
+
"grad_norm": 0.010241471230983734,
|
| 2874 |
+
"learning_rate": 1.1814345991561182e-06,
|
| 2875 |
+
"loss": 0.0001,
|
| 2876 |
+
"step": 3530
|
| 2877 |
+
},
|
| 2878 |
+
{
|
| 2879 |
+
"epoch": 44.00506329113924,
|
| 2880 |
+
"grad_norm": 0.005635945592075586,
|
| 2881 |
+
"learning_rate": 1.1533052039381154e-06,
|
| 2882 |
+
"loss": 0.0002,
|
| 2883 |
+
"step": 3540
|
| 2884 |
+
},
|
| 2885 |
+
{
|
| 2886 |
+
"epoch": 44.00759493670886,
|
| 2887 |
+
"grad_norm": 0.005905819591134787,
|
| 2888 |
+
"learning_rate": 1.1251758087201126e-06,
|
| 2889 |
+
"loss": 0.0002,
|
| 2890 |
+
"step": 3550
|
| 2891 |
+
},
|
| 2892 |
+
{
|
| 2893 |
+
"epoch": 44.01012658227848,
|
| 2894 |
+
"grad_norm": 0.0035794128198176622,
|
| 2895 |
+
"learning_rate": 1.0970464135021099e-06,
|
| 2896 |
+
"loss": 0.0001,
|
| 2897 |
+
"step": 3560
|
| 2898 |
+
},
|
| 2899 |
+
{
|
| 2900 |
+
"epoch": 44.0126582278481,
|
| 2901 |
+
"grad_norm": 0.00724047003313899,
|
| 2902 |
+
"learning_rate": 1.068917018284107e-06,
|
| 2903 |
+
"loss": 0.0001,
|
| 2904 |
+
"step": 3570
|
| 2905 |
+
},
|
| 2906 |
+
{
|
| 2907 |
+
"epoch": 44.01518987341772,
|
| 2908 |
+
"grad_norm": 0.0037932819686830044,
|
| 2909 |
+
"learning_rate": 1.040787623066104e-06,
|
| 2910 |
+
"loss": 0.0001,
|
| 2911 |
+
"step": 3580
|
| 2912 |
+
},
|
| 2913 |
+
{
|
| 2914 |
+
"epoch": 44.017721518987344,
|
| 2915 |
+
"grad_norm": 0.003542226040735841,
|
| 2916 |
+
"learning_rate": 1.0126582278481013e-06,
|
| 2917 |
+
"loss": 0.0004,
|
| 2918 |
+
"step": 3590
|
| 2919 |
+
},
|
| 2920 |
+
{
|
| 2921 |
+
"epoch": 44.02025316455696,
|
| 2922 |
+
"grad_norm": 0.0041389502584934235,
|
| 2923 |
+
"learning_rate": 9.845288326300985e-07,
|
| 2924 |
+
"loss": 0.0001,
|
| 2925 |
+
"step": 3600
|
| 2926 |
+
},
|
| 2927 |
+
{
|
| 2928 |
+
"epoch": 44.02025316455696,
|
| 2929 |
+
"eval_accuracy": 0.3829787234042553,
|
| 2930 |
+
"eval_loss": 4.686789035797119,
|
| 2931 |
+
"eval_runtime": 8.3951,
|
| 2932 |
+
"eval_samples_per_second": 5.599,
|
| 2933 |
+
"eval_steps_per_second": 1.429,
|
| 2934 |
+
"step": 3600
|
| 2935 |
+
},
|
| 2936 |
+
{
|
| 2937 |
+
"epoch": 45.00253164556962,
|
| 2938 |
+
"grad_norm": 0.008083141408860683,
|
| 2939 |
+
"learning_rate": 9.563994374120958e-07,
|
| 2940 |
+
"loss": 0.0001,
|
| 2941 |
+
"step": 3610
|
| 2942 |
+
},
|
| 2943 |
+
{
|
| 2944 |
+
"epoch": 45.00506329113924,
|
| 2945 |
+
"grad_norm": 0.017192961648106575,
|
| 2946 |
+
"learning_rate": 9.28270042194093e-07,
|
| 2947 |
+
"loss": 0.0001,
|
| 2948 |
+
"step": 3620
|
| 2949 |
+
},
|
| 2950 |
+
{
|
| 2951 |
+
"epoch": 45.00759493670886,
|
| 2952 |
+
"grad_norm": 0.015804950147867203,
|
| 2953 |
+
"learning_rate": 9.001406469760901e-07,
|
| 2954 |
+
"loss": 0.0001,
|
| 2955 |
+
"step": 3630
|
| 2956 |
+
},
|
| 2957 |
+
{
|
| 2958 |
+
"epoch": 45.01012658227848,
|
| 2959 |
+
"grad_norm": 0.014944680966436863,
|
| 2960 |
+
"learning_rate": 8.720112517580873e-07,
|
| 2961 |
+
"loss": 0.0002,
|
| 2962 |
+
"step": 3640
|
| 2963 |
+
},
|
| 2964 |
+
{
|
| 2965 |
+
"epoch": 45.0126582278481,
|
| 2966 |
+
"grad_norm": 0.003325532888993621,
|
| 2967 |
+
"learning_rate": 8.438818565400844e-07,
|
| 2968 |
+
"loss": 0.0001,
|
| 2969 |
+
"step": 3650
|
| 2970 |
+
},
|
| 2971 |
+
{
|
| 2972 |
+
"epoch": 45.01518987341772,
|
| 2973 |
+
"grad_norm": 0.0040122633799910545,
|
| 2974 |
+
"learning_rate": 8.157524613220815e-07,
|
| 2975 |
+
"loss": 0.0002,
|
| 2976 |
+
"step": 3660
|
| 2977 |
+
},
|
| 2978 |
+
{
|
| 2979 |
+
"epoch": 45.017721518987344,
|
| 2980 |
+
"grad_norm": 0.01050250418484211,
|
| 2981 |
+
"learning_rate": 7.876230661040789e-07,
|
| 2982 |
+
"loss": 0.0002,
|
| 2983 |
+
"step": 3670
|
| 2984 |
+
},
|
| 2985 |
+
{
|
| 2986 |
+
"epoch": 45.02025316455696,
|
| 2987 |
+
"grad_norm": 0.006878922693431377,
|
| 2988 |
+
"learning_rate": 7.59493670886076e-07,
|
| 2989 |
+
"loss": 0.0001,
|
| 2990 |
+
"step": 3680
|
| 2991 |
+
},
|
| 2992 |
+
{
|
| 2993 |
+
"epoch": 45.02025316455696,
|
| 2994 |
+
"eval_accuracy": 0.3829787234042553,
|
| 2995 |
+
"eval_loss": 4.6618781089782715,
|
| 2996 |
+
"eval_runtime": 8.4159,
|
| 2997 |
+
"eval_samples_per_second": 5.585,
|
| 2998 |
+
"eval_steps_per_second": 1.426,
|
| 2999 |
+
"step": 3680
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 46.00253164556962,
|
| 3003 |
+
"grad_norm": 0.009057370945811272,
|
| 3004 |
+
"learning_rate": 7.313642756680732e-07,
|
| 3005 |
+
"loss": 0.0002,
|
| 3006 |
+
"step": 3690
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 46.00506329113924,
|
| 3010 |
+
"grad_norm": 0.003986147698014975,
|
| 3011 |
+
"learning_rate": 7.032348804500703e-07,
|
| 3012 |
+
"loss": 0.0105,
|
| 3013 |
+
"step": 3700
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 46.00759493670886,
|
| 3017 |
+
"grad_norm": 0.01812109723687172,
|
| 3018 |
+
"learning_rate": 6.751054852320675e-07,
|
| 3019 |
+
"loss": 0.0001,
|
| 3020 |
+
"step": 3710
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 46.01012658227848,
|
| 3024 |
+
"grad_norm": 0.016234632581472397,
|
| 3025 |
+
"learning_rate": 6.469760900140648e-07,
|
| 3026 |
+
"loss": 0.0001,
|
| 3027 |
+
"step": 3720
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 46.0126582278481,
|
| 3031 |
+
"grad_norm": 0.0052651516161859035,
|
| 3032 |
+
"learning_rate": 6.18846694796062e-07,
|
| 3033 |
+
"loss": 0.0216,
|
| 3034 |
+
"step": 3730
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 46.01518987341772,
|
| 3038 |
+
"grad_norm": 0.009317484684288502,
|
| 3039 |
+
"learning_rate": 5.907172995780591e-07,
|
| 3040 |
+
"loss": 0.0001,
|
| 3041 |
+
"step": 3740
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 46.017721518987344,
|
| 3045 |
+
"grad_norm": 0.0024894699454307556,
|
| 3046 |
+
"learning_rate": 5.625879043600563e-07,
|
| 3047 |
+
"loss": 0.0003,
|
| 3048 |
+
"step": 3750
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 46.02025316455696,
|
| 3052 |
+
"grad_norm": 0.011036810465157032,
|
| 3053 |
+
"learning_rate": 5.344585091420535e-07,
|
| 3054 |
+
"loss": 0.0002,
|
| 3055 |
+
"step": 3760
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 46.02025316455696,
|
| 3059 |
+
"eval_accuracy": 0.3617021276595745,
|
| 3060 |
+
"eval_loss": 4.714182376861572,
|
| 3061 |
+
"eval_runtime": 8.4053,
|
| 3062 |
+
"eval_samples_per_second": 5.592,
|
| 3063 |
+
"eval_steps_per_second": 1.428,
|
| 3064 |
+
"step": 3760
|
| 3065 |
+
},
|
| 3066 |
+
{
|
| 3067 |
+
"epoch": 47.00253164556962,
|
| 3068 |
+
"grad_norm": 0.0035392444115132093,
|
| 3069 |
+
"learning_rate": 5.063291139240507e-07,
|
| 3070 |
+
"loss": 0.0002,
|
| 3071 |
+
"step": 3770
|
| 3072 |
+
},
|
| 3073 |
+
{
|
| 3074 |
+
"epoch": 47.00506329113924,
|
| 3075 |
+
"grad_norm": 0.008757932111620903,
|
| 3076 |
+
"learning_rate": 4.781997187060479e-07,
|
| 3077 |
+
"loss": 0.0005,
|
| 3078 |
+
"step": 3780
|
| 3079 |
+
},
|
| 3080 |
+
{
|
| 3081 |
+
"epoch": 47.00759493670886,
|
| 3082 |
+
"grad_norm": 0.003641449846327305,
|
| 3083 |
+
"learning_rate": 4.5007032348804504e-07,
|
| 3084 |
+
"loss": 0.0001,
|
| 3085 |
+
"step": 3790
|
| 3086 |
+
},
|
| 3087 |
+
{
|
| 3088 |
+
"epoch": 47.01012658227848,
|
| 3089 |
+
"grad_norm": 0.021657196804881096,
|
| 3090 |
+
"learning_rate": 4.219409282700422e-07,
|
| 3091 |
+
"loss": 0.0002,
|
| 3092 |
+
"step": 3800
|
| 3093 |
+
},
|
| 3094 |
+
{
|
| 3095 |
+
"epoch": 47.0126582278481,
|
| 3096 |
+
"grad_norm": 0.0035243236925452948,
|
| 3097 |
+
"learning_rate": 3.9381153305203943e-07,
|
| 3098 |
+
"loss": 0.0001,
|
| 3099 |
+
"step": 3810
|
| 3100 |
+
},
|
| 3101 |
+
{
|
| 3102 |
+
"epoch": 47.01518987341772,
|
| 3103 |
+
"grad_norm": 0.005966987926512957,
|
| 3104 |
+
"learning_rate": 3.656821378340366e-07,
|
| 3105 |
+
"loss": 0.0001,
|
| 3106 |
+
"step": 3820
|
| 3107 |
+
},
|
| 3108 |
+
{
|
| 3109 |
+
"epoch": 47.017721518987344,
|
| 3110 |
+
"grad_norm": 0.004663312342017889,
|
| 3111 |
+
"learning_rate": 3.3755274261603377e-07,
|
| 3112 |
+
"loss": 0.0001,
|
| 3113 |
+
"step": 3830
|
| 3114 |
+
},
|
| 3115 |
+
{
|
| 3116 |
+
"epoch": 47.02025316455696,
|
| 3117 |
+
"grad_norm": 0.005123648792505264,
|
| 3118 |
+
"learning_rate": 3.09423347398031e-07,
|
| 3119 |
+
"loss": 0.0002,
|
| 3120 |
+
"step": 3840
|
| 3121 |
+
},
|
| 3122 |
+
{
|
| 3123 |
+
"epoch": 47.02025316455696,
|
| 3124 |
+
"eval_accuracy": 0.3617021276595745,
|
| 3125 |
+
"eval_loss": 4.652493000030518,
|
| 3126 |
+
"eval_runtime": 8.5396,
|
| 3127 |
+
"eval_samples_per_second": 5.504,
|
| 3128 |
+
"eval_steps_per_second": 1.405,
|
| 3129 |
+
"step": 3840
|
| 3130 |
+
},
|
| 3131 |
+
{
|
| 3132 |
+
"epoch": 48.00253164556962,
|
| 3133 |
+
"grad_norm": 0.03579903393983841,
|
| 3134 |
+
"learning_rate": 2.8129395218002816e-07,
|
| 3135 |
+
"loss": 0.0002,
|
| 3136 |
+
"step": 3850
|
| 3137 |
+
},
|
| 3138 |
+
{
|
| 3139 |
+
"epoch": 48.00506329113924,
|
| 3140 |
+
"grad_norm": 0.0028870755340903997,
|
| 3141 |
+
"learning_rate": 2.5316455696202533e-07,
|
| 3142 |
+
"loss": 0.0001,
|
| 3143 |
+
"step": 3860
|
| 3144 |
+
},
|
| 3145 |
+
{
|
| 3146 |
+
"epoch": 48.00759493670886,
|
| 3147 |
+
"grad_norm": 0.010382084175944328,
|
| 3148 |
+
"learning_rate": 2.2503516174402252e-07,
|
| 3149 |
+
"loss": 0.0002,
|
| 3150 |
+
"step": 3870
|
| 3151 |
+
},
|
| 3152 |
+
{
|
| 3153 |
+
"epoch": 48.01012658227848,
|
| 3154 |
+
"grad_norm": 0.008079590275883675,
|
| 3155 |
+
"learning_rate": 1.9690576652601972e-07,
|
| 3156 |
+
"loss": 0.0001,
|
| 3157 |
+
"step": 3880
|
| 3158 |
+
},
|
| 3159 |
+
{
|
| 3160 |
+
"epoch": 48.0126582278481,
|
| 3161 |
+
"grad_norm": 0.007244328502565622,
|
| 3162 |
+
"learning_rate": 1.6877637130801689e-07,
|
| 3163 |
+
"loss": 0.0001,
|
| 3164 |
+
"step": 3890
|
| 3165 |
+
},
|
| 3166 |
+
{
|
| 3167 |
+
"epoch": 48.01518987341772,
|
| 3168 |
+
"grad_norm": 0.0033432315103709698,
|
| 3169 |
+
"learning_rate": 1.4064697609001408e-07,
|
| 3170 |
+
"loss": 0.0001,
|
| 3171 |
+
"step": 3900
|
| 3172 |
+
},
|
| 3173 |
+
{
|
| 3174 |
+
"epoch": 48.017721518987344,
|
| 3175 |
+
"grad_norm": 0.13150210678577423,
|
| 3176 |
+
"learning_rate": 1.1251758087201126e-07,
|
| 3177 |
+
"loss": 0.0002,
|
| 3178 |
+
"step": 3910
|
| 3179 |
+
},
|
| 3180 |
+
{
|
| 3181 |
+
"epoch": 48.02025316455696,
|
| 3182 |
+
"grad_norm": 0.00498367240652442,
|
| 3183 |
+
"learning_rate": 8.438818565400844e-08,
|
| 3184 |
+
"loss": 0.0001,
|
| 3185 |
+
"step": 3920
|
| 3186 |
+
},
|
| 3187 |
+
{
|
| 3188 |
+
"epoch": 48.02025316455696,
|
| 3189 |
+
"eval_accuracy": 0.3617021276595745,
|
| 3190 |
+
"eval_loss": 4.668392181396484,
|
| 3191 |
+
"eval_runtime": 8.367,
|
| 3192 |
+
"eval_samples_per_second": 5.617,
|
| 3193 |
+
"eval_steps_per_second": 1.434,
|
| 3194 |
+
"step": 3920
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 49.00253164556962,
|
| 3198 |
+
"grad_norm": 0.005051782354712486,
|
| 3199 |
+
"learning_rate": 5.625879043600563e-08,
|
| 3200 |
+
"loss": 0.0002,
|
| 3201 |
+
"step": 3930
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 49.00506329113924,
|
| 3205 |
+
"grad_norm": 0.02646501734852791,
|
| 3206 |
+
"learning_rate": 2.8129395218002815e-08,
|
| 3207 |
+
"loss": 0.0002,
|
| 3208 |
+
"step": 3940
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 49.00759493670886,
|
| 3212 |
+
"grad_norm": 0.00457751052454114,
|
| 3213 |
+
"learning_rate": 0.0,
|
| 3214 |
+
"loss": 0.0001,
|
| 3215 |
+
"step": 3950
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 49.00759493670886,
|
| 3219 |
+
"eval_accuracy": 0.3617021276595745,
|
| 3220 |
+
"eval_loss": 4.666772842407227,
|
| 3221 |
+
"eval_runtime": 9.2282,
|
| 3222 |
+
"eval_samples_per_second": 5.093,
|
| 3223 |
+
"eval_steps_per_second": 1.3,
|
| 3224 |
+
"step": 3950
|
| 3225 |
+
},
|
| 3226 |
+
{
|
| 3227 |
+
"epoch": 49.00759493670886,
|
| 3228 |
+
"step": 3950,
|
| 3229 |
+
"total_flos": 6.89481277801104e+19,
|
| 3230 |
+
"train_loss": 0.24269407705082616,
|
| 3231 |
+
"train_runtime": 6713.1285,
|
| 3232 |
+
"train_samples_per_second": 2.354,
|
| 3233 |
+
"train_steps_per_second": 0.588
|
| 3234 |
+
},
|
| 3235 |
+
{
|
| 3236 |
+
"epoch": 49.00759493670886,
|
| 3237 |
+
"eval_accuracy": 0.2564102564102564,
|
| 3238 |
+
"eval_loss": 2.559528350830078,
|
| 3239 |
+
"eval_runtime": 7.861,
|
| 3240 |
+
"eval_samples_per_second": 4.961,
|
| 3241 |
+
"eval_steps_per_second": 1.272,
|
| 3242 |
+
"step": 3950
|
| 3243 |
+
},
|
| 3244 |
+
{
|
| 3245 |
+
"epoch": 49.00759493670886,
|
| 3246 |
+
"eval_accuracy": 0.2564102564102564,
|
| 3247 |
+
"eval_loss": 2.5595290660858154,
|
| 3248 |
+
"eval_runtime": 6.9771,
|
| 3249 |
+
"eval_samples_per_second": 5.59,
|
| 3250 |
+
"eval_steps_per_second": 1.433,
|
| 3251 |
+
"step": 3950
|
| 3252 |
+
}
|
| 3253 |
+
],
|
| 3254 |
+
"logging_steps": 10,
|
| 3255 |
+
"max_steps": 3950,
|
| 3256 |
+
"num_input_tokens_seen": 0,
|
| 3257 |
+
"num_train_epochs": 9223372036854775807,
|
| 3258 |
+
"save_steps": 500,
|
| 3259 |
+
"stateful_callbacks": {
|
| 3260 |
+
"TrainerControl": {
|
| 3261 |
+
"args": {
|
| 3262 |
+
"should_epoch_stop": false,
|
| 3263 |
+
"should_evaluate": false,
|
| 3264 |
+
"should_log": false,
|
| 3265 |
+
"should_save": true,
|
| 3266 |
+
"should_training_stop": true
|
| 3267 |
+
},
|
| 3268 |
+
"attributes": {}
|
| 3269 |
+
}
|
| 3270 |
+
},
|
| 3271 |
+
"total_flos": 6.89481277801104e+19,
|
| 3272 |
+
"train_batch_size": 4,
|
| 3273 |
+
"trial_name": null,
|
| 3274 |
+
"trial_params": null
|
| 3275 |
+
}
|