End of training
Browse files- all_results.json +8 -0
- test_results.json +8 -0
- trainer_state.json +2960 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 49.02,
|
| 3 |
+
"eval_accuracy": 0.8372093023255814,
|
| 4 |
+
"eval_loss": 0.8805379867553711,
|
| 5 |
+
"eval_runtime": 9.0595,
|
| 6 |
+
"eval_samples_per_second": 4.746,
|
| 7 |
+
"eval_steps_per_second": 1.214
|
| 8 |
+
}
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 49.02,
|
| 3 |
+
"eval_accuracy": 0.8372093023255814,
|
| 4 |
+
"eval_loss": 0.8805379867553711,
|
| 5 |
+
"eval_runtime": 9.0595,
|
| 6 |
+
"eval_samples_per_second": 4.746,
|
| 7 |
+
"eval_steps_per_second": 1.214
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2960 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.75,
|
| 3 |
+
"best_model_checkpoint": "MAE-CT-CPC-Dicotomized-n0-m1-v8-1/checkpoint-980",
|
| 4 |
+
"epoch": 49.02,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.002857142857142857,
|
| 13 |
+
"grad_norm": 3.5134518146514893,
|
| 14 |
+
"learning_rate": 2.8571428571428575e-07,
|
| 15 |
+
"loss": 0.2572,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.005714285714285714,
|
| 20 |
+
"grad_norm": 19.781003952026367,
|
| 21 |
+
"learning_rate": 5.714285714285715e-07,
|
| 22 |
+
"loss": 0.2784,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.008571428571428572,
|
| 27 |
+
"grad_norm": 27.620603561401367,
|
| 28 |
+
"learning_rate": 8.571428571428572e-07,
|
| 29 |
+
"loss": 0.4606,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.011428571428571429,
|
| 34 |
+
"grad_norm": 5.682175159454346,
|
| 35 |
+
"learning_rate": 1.142857142857143e-06,
|
| 36 |
+
"loss": 0.1624,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.014285714285714285,
|
| 41 |
+
"grad_norm": 57.06011199951172,
|
| 42 |
+
"learning_rate": 1.4285714285714286e-06,
|
| 43 |
+
"loss": 0.5168,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.017142857142857144,
|
| 48 |
+
"grad_norm": 12.413666725158691,
|
| 49 |
+
"learning_rate": 1.7142857142857145e-06,
|
| 50 |
+
"loss": 0.3112,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02,
|
| 55 |
+
"grad_norm": 40.50763702392578,
|
| 56 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 57 |
+
"loss": 0.3313,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.02,
|
| 62 |
+
"eval_accuracy": 0.6875,
|
| 63 |
+
"eval_loss": 0.8278725147247314,
|
| 64 |
+
"eval_runtime": 7.2014,
|
| 65 |
+
"eval_samples_per_second": 4.444,
|
| 66 |
+
"eval_steps_per_second": 1.111,
|
| 67 |
+
"step": 70
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 1.002857142857143,
|
| 71 |
+
"grad_norm": 2.85071063041687,
|
| 72 |
+
"learning_rate": 2.285714285714286e-06,
|
| 73 |
+
"loss": 0.2007,
|
| 74 |
+
"step": 80
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.0057142857142858,
|
| 78 |
+
"grad_norm": 3.0488057136535645,
|
| 79 |
+
"learning_rate": 2.571428571428571e-06,
|
| 80 |
+
"loss": 0.4132,
|
| 81 |
+
"step": 90
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.0085714285714287,
|
| 85 |
+
"grad_norm": 1.0658628940582275,
|
| 86 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 87 |
+
"loss": 0.2365,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.0114285714285713,
|
| 92 |
+
"grad_norm": 32.61592483520508,
|
| 93 |
+
"learning_rate": 3.142857142857143e-06,
|
| 94 |
+
"loss": 0.2773,
|
| 95 |
+
"step": 110
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.0142857142857142,
|
| 99 |
+
"grad_norm": 0.24897827208042145,
|
| 100 |
+
"learning_rate": 3.428571428571429e-06,
|
| 101 |
+
"loss": 0.2216,
|
| 102 |
+
"step": 120
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.0171428571428571,
|
| 106 |
+
"grad_norm": 40.215003967285156,
|
| 107 |
+
"learning_rate": 3.7142857142857146e-06,
|
| 108 |
+
"loss": 0.6413,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.02,
|
| 113 |
+
"grad_norm": 21.867782592773438,
|
| 114 |
+
"learning_rate": 4.000000000000001e-06,
|
| 115 |
+
"loss": 0.4602,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.02,
|
| 120 |
+
"eval_accuracy": 0.65625,
|
| 121 |
+
"eval_loss": 0.7750873565673828,
|
| 122 |
+
"eval_runtime": 6.0768,
|
| 123 |
+
"eval_samples_per_second": 5.266,
|
| 124 |
+
"eval_steps_per_second": 1.316,
|
| 125 |
+
"step": 140
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 2.0028571428571427,
|
| 129 |
+
"grad_norm": 39.138465881347656,
|
| 130 |
+
"learning_rate": 4.2857142857142855e-06,
|
| 131 |
+
"loss": 0.1423,
|
| 132 |
+
"step": 150
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 2.005714285714286,
|
| 136 |
+
"grad_norm": 0.8473556041717529,
|
| 137 |
+
"learning_rate": 4.571428571428572e-06,
|
| 138 |
+
"loss": 0.1991,
|
| 139 |
+
"step": 160
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 2.0085714285714285,
|
| 143 |
+
"grad_norm": 31.097688674926758,
|
| 144 |
+
"learning_rate": 4.857142857142858e-06,
|
| 145 |
+
"loss": 0.327,
|
| 146 |
+
"step": 170
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 2.0114285714285716,
|
| 150 |
+
"grad_norm": 0.581165611743927,
|
| 151 |
+
"learning_rate": 5.142857142857142e-06,
|
| 152 |
+
"loss": 0.332,
|
| 153 |
+
"step": 180
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.0142857142857142,
|
| 157 |
+
"grad_norm": 54.0560188293457,
|
| 158 |
+
"learning_rate": 5.428571428571429e-06,
|
| 159 |
+
"loss": 0.3274,
|
| 160 |
+
"step": 190
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.0171428571428573,
|
| 164 |
+
"grad_norm": 8.153368949890137,
|
| 165 |
+
"learning_rate": 5.7142857142857145e-06,
|
| 166 |
+
"loss": 0.1702,
|
| 167 |
+
"step": 200
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.02,
|
| 171 |
+
"grad_norm": 82.87539672851562,
|
| 172 |
+
"learning_rate": 6e-06,
|
| 173 |
+
"loss": 0.413,
|
| 174 |
+
"step": 210
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.02,
|
| 178 |
+
"eval_accuracy": 0.65625,
|
| 179 |
+
"eval_loss": 0.9297134280204773,
|
| 180 |
+
"eval_runtime": 6.3778,
|
| 181 |
+
"eval_samples_per_second": 5.017,
|
| 182 |
+
"eval_steps_per_second": 1.254,
|
| 183 |
+
"step": 210
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.0028571428571427,
|
| 187 |
+
"grad_norm": 6.549503803253174,
|
| 188 |
+
"learning_rate": 6.285714285714286e-06,
|
| 189 |
+
"loss": 0.126,
|
| 190 |
+
"step": 220
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.005714285714286,
|
| 194 |
+
"grad_norm": 0.07536959648132324,
|
| 195 |
+
"learning_rate": 6.571428571428572e-06,
|
| 196 |
+
"loss": 0.2436,
|
| 197 |
+
"step": 230
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.0085714285714285,
|
| 201 |
+
"grad_norm": 75.6919174194336,
|
| 202 |
+
"learning_rate": 6.857142857142858e-06,
|
| 203 |
+
"loss": 0.3704,
|
| 204 |
+
"step": 240
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.0114285714285716,
|
| 208 |
+
"grad_norm": 6.202108860015869,
|
| 209 |
+
"learning_rate": 7.1428571428571436e-06,
|
| 210 |
+
"loss": 0.2685,
|
| 211 |
+
"step": 250
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.0142857142857142,
|
| 215 |
+
"grad_norm": 5.261022567749023,
|
| 216 |
+
"learning_rate": 7.428571428571429e-06,
|
| 217 |
+
"loss": 0.0579,
|
| 218 |
+
"step": 260
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.0171428571428573,
|
| 222 |
+
"grad_norm": 62.4367790222168,
|
| 223 |
+
"learning_rate": 7.714285714285716e-06,
|
| 224 |
+
"loss": 0.4266,
|
| 225 |
+
"step": 270
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.02,
|
| 229 |
+
"grad_norm": 0.3603902757167816,
|
| 230 |
+
"learning_rate": 8.000000000000001e-06,
|
| 231 |
+
"loss": 0.2196,
|
| 232 |
+
"step": 280
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.02,
|
| 236 |
+
"eval_accuracy": 0.59375,
|
| 237 |
+
"eval_loss": 1.307774543762207,
|
| 238 |
+
"eval_runtime": 6.077,
|
| 239 |
+
"eval_samples_per_second": 5.266,
|
| 240 |
+
"eval_steps_per_second": 1.316,
|
| 241 |
+
"step": 280
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 4.002857142857143,
|
| 245 |
+
"grad_norm": 82.43415832519531,
|
| 246 |
+
"learning_rate": 8.285714285714287e-06,
|
| 247 |
+
"loss": 0.1117,
|
| 248 |
+
"step": 290
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 4.005714285714285,
|
| 252 |
+
"grad_norm": 1.1366825103759766,
|
| 253 |
+
"learning_rate": 8.571428571428571e-06,
|
| 254 |
+
"loss": 0.1042,
|
| 255 |
+
"step": 300
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 4.008571428571429,
|
| 259 |
+
"grad_norm": 0.5070340037345886,
|
| 260 |
+
"learning_rate": 8.857142857142858e-06,
|
| 261 |
+
"loss": 0.3576,
|
| 262 |
+
"step": 310
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 4.011428571428572,
|
| 266 |
+
"grad_norm": 0.18760715425014496,
|
| 267 |
+
"learning_rate": 9.142857142857144e-06,
|
| 268 |
+
"loss": 0.2198,
|
| 269 |
+
"step": 320
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 4.014285714285714,
|
| 273 |
+
"grad_norm": 0.5741114020347595,
|
| 274 |
+
"learning_rate": 9.42857142857143e-06,
|
| 275 |
+
"loss": 0.111,
|
| 276 |
+
"step": 330
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 4.017142857142857,
|
| 280 |
+
"grad_norm": 0.07025701552629471,
|
| 281 |
+
"learning_rate": 9.714285714285715e-06,
|
| 282 |
+
"loss": 0.149,
|
| 283 |
+
"step": 340
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 4.02,
|
| 287 |
+
"grad_norm": 62.01303482055664,
|
| 288 |
+
"learning_rate": 1e-05,
|
| 289 |
+
"loss": 0.4898,
|
| 290 |
+
"step": 350
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 4.02,
|
| 294 |
+
"eval_accuracy": 0.625,
|
| 295 |
+
"eval_loss": 1.286909580230713,
|
| 296 |
+
"eval_runtime": 6.1054,
|
| 297 |
+
"eval_samples_per_second": 5.241,
|
| 298 |
+
"eval_steps_per_second": 1.31,
|
| 299 |
+
"step": 350
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 5.002857142857143,
|
| 303 |
+
"grad_norm": 59.823509216308594,
|
| 304 |
+
"learning_rate": 9.968253968253969e-06,
|
| 305 |
+
"loss": 0.2052,
|
| 306 |
+
"step": 360
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 5.005714285714285,
|
| 310 |
+
"grad_norm": 0.020512551069259644,
|
| 311 |
+
"learning_rate": 9.936507936507937e-06,
|
| 312 |
+
"loss": 0.2319,
|
| 313 |
+
"step": 370
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 5.008571428571429,
|
| 317 |
+
"grad_norm": 0.013519791886210442,
|
| 318 |
+
"learning_rate": 9.904761904761906e-06,
|
| 319 |
+
"loss": 0.2941,
|
| 320 |
+
"step": 380
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 5.011428571428572,
|
| 324 |
+
"grad_norm": 103.2642822265625,
|
| 325 |
+
"learning_rate": 9.873015873015874e-06,
|
| 326 |
+
"loss": 0.123,
|
| 327 |
+
"step": 390
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 5.014285714285714,
|
| 331 |
+
"grad_norm": 0.5531640648841858,
|
| 332 |
+
"learning_rate": 9.841269841269842e-06,
|
| 333 |
+
"loss": 0.3053,
|
| 334 |
+
"step": 400
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 5.017142857142857,
|
| 338 |
+
"grad_norm": 8.770979881286621,
|
| 339 |
+
"learning_rate": 9.80952380952381e-06,
|
| 340 |
+
"loss": 0.3701,
|
| 341 |
+
"step": 410
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 5.02,
|
| 345 |
+
"grad_norm": 5.519145965576172,
|
| 346 |
+
"learning_rate": 9.777777777777779e-06,
|
| 347 |
+
"loss": 0.1867,
|
| 348 |
+
"step": 420
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 5.02,
|
| 352 |
+
"eval_accuracy": 0.625,
|
| 353 |
+
"eval_loss": 1.5883848667144775,
|
| 354 |
+
"eval_runtime": 6.5433,
|
| 355 |
+
"eval_samples_per_second": 4.891,
|
| 356 |
+
"eval_steps_per_second": 1.223,
|
| 357 |
+
"step": 420
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 6.002857142857143,
|
| 361 |
+
"grad_norm": 114.1990966796875,
|
| 362 |
+
"learning_rate": 9.746031746031747e-06,
|
| 363 |
+
"loss": 0.1602,
|
| 364 |
+
"step": 430
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 6.005714285714285,
|
| 368 |
+
"grad_norm": 0.034505173563957214,
|
| 369 |
+
"learning_rate": 9.714285714285715e-06,
|
| 370 |
+
"loss": 0.003,
|
| 371 |
+
"step": 440
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 6.008571428571429,
|
| 375 |
+
"grad_norm": 3.3536415100097656,
|
| 376 |
+
"learning_rate": 9.682539682539683e-06,
|
| 377 |
+
"loss": 0.158,
|
| 378 |
+
"step": 450
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 6.011428571428572,
|
| 382 |
+
"grad_norm": 0.010583457536995411,
|
| 383 |
+
"learning_rate": 9.650793650793652e-06,
|
| 384 |
+
"loss": 0.3164,
|
| 385 |
+
"step": 460
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 6.014285714285714,
|
| 389 |
+
"grad_norm": 109.80305480957031,
|
| 390 |
+
"learning_rate": 9.61904761904762e-06,
|
| 391 |
+
"loss": 0.1235,
|
| 392 |
+
"step": 470
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 6.017142857142857,
|
| 396 |
+
"grad_norm": 0.5373286008834839,
|
| 397 |
+
"learning_rate": 9.587301587301588e-06,
|
| 398 |
+
"loss": 0.3013,
|
| 399 |
+
"step": 480
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 6.02,
|
| 403 |
+
"grad_norm": 9.764594078063965,
|
| 404 |
+
"learning_rate": 9.555555555555556e-06,
|
| 405 |
+
"loss": 0.0394,
|
| 406 |
+
"step": 490
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 6.02,
|
| 410 |
+
"eval_accuracy": 0.625,
|
| 411 |
+
"eval_loss": 1.7786798477172852,
|
| 412 |
+
"eval_runtime": 6.3022,
|
| 413 |
+
"eval_samples_per_second": 5.078,
|
| 414 |
+
"eval_steps_per_second": 1.269,
|
| 415 |
+
"step": 490
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 7.002857142857143,
|
| 419 |
+
"grad_norm": 0.10722041130065918,
|
| 420 |
+
"learning_rate": 9.523809523809525e-06,
|
| 421 |
+
"loss": 0.0015,
|
| 422 |
+
"step": 500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 7.005714285714285,
|
| 426 |
+
"grad_norm": 0.04226737469434738,
|
| 427 |
+
"learning_rate": 9.492063492063493e-06,
|
| 428 |
+
"loss": 0.201,
|
| 429 |
+
"step": 510
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 7.008571428571429,
|
| 433 |
+
"grad_norm": 1.7212629318237305,
|
| 434 |
+
"learning_rate": 9.460317460317461e-06,
|
| 435 |
+
"loss": 0.0604,
|
| 436 |
+
"step": 520
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 7.011428571428572,
|
| 440 |
+
"grad_norm": 0.01773512363433838,
|
| 441 |
+
"learning_rate": 9.42857142857143e-06,
|
| 442 |
+
"loss": 0.0661,
|
| 443 |
+
"step": 530
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 7.014285714285714,
|
| 447 |
+
"grad_norm": 0.024557197466492653,
|
| 448 |
+
"learning_rate": 9.396825396825398e-06,
|
| 449 |
+
"loss": 0.0466,
|
| 450 |
+
"step": 540
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 7.017142857142857,
|
| 454 |
+
"grad_norm": 0.01698165014386177,
|
| 455 |
+
"learning_rate": 9.365079365079366e-06,
|
| 456 |
+
"loss": 0.0009,
|
| 457 |
+
"step": 550
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 7.02,
|
| 461 |
+
"grad_norm": 0.40697580575942993,
|
| 462 |
+
"learning_rate": 9.333333333333334e-06,
|
| 463 |
+
"loss": 0.1218,
|
| 464 |
+
"step": 560
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 7.02,
|
| 468 |
+
"eval_accuracy": 0.59375,
|
| 469 |
+
"eval_loss": 1.79988694190979,
|
| 470 |
+
"eval_runtime": 5.9173,
|
| 471 |
+
"eval_samples_per_second": 5.408,
|
| 472 |
+
"eval_steps_per_second": 1.352,
|
| 473 |
+
"step": 560
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 8.002857142857144,
|
| 477 |
+
"grad_norm": 0.09231330454349518,
|
| 478 |
+
"learning_rate": 9.301587301587303e-06,
|
| 479 |
+
"loss": 0.5704,
|
| 480 |
+
"step": 570
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 8.005714285714285,
|
| 484 |
+
"grad_norm": 0.06326180696487427,
|
| 485 |
+
"learning_rate": 9.26984126984127e-06,
|
| 486 |
+
"loss": 0.1072,
|
| 487 |
+
"step": 580
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 8.008571428571429,
|
| 491 |
+
"grad_norm": 0.077267587184906,
|
| 492 |
+
"learning_rate": 9.238095238095239e-06,
|
| 493 |
+
"loss": 0.0308,
|
| 494 |
+
"step": 590
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 8.01142857142857,
|
| 498 |
+
"grad_norm": 0.05788983032107353,
|
| 499 |
+
"learning_rate": 9.206349206349207e-06,
|
| 500 |
+
"loss": 0.1882,
|
| 501 |
+
"step": 600
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 8.014285714285714,
|
| 505 |
+
"grad_norm": 154.5283203125,
|
| 506 |
+
"learning_rate": 9.174603174603176e-06,
|
| 507 |
+
"loss": 0.4305,
|
| 508 |
+
"step": 610
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 8.017142857142858,
|
| 512 |
+
"grad_norm": 0.03793035447597504,
|
| 513 |
+
"learning_rate": 9.142857142857144e-06,
|
| 514 |
+
"loss": 0.3163,
|
| 515 |
+
"step": 620
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 8.02,
|
| 519 |
+
"grad_norm": 1.2432268857955933,
|
| 520 |
+
"learning_rate": 9.111111111111112e-06,
|
| 521 |
+
"loss": 0.192,
|
| 522 |
+
"step": 630
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 8.02,
|
| 526 |
+
"eval_accuracy": 0.625,
|
| 527 |
+
"eval_loss": 1.7294597625732422,
|
| 528 |
+
"eval_runtime": 6.4051,
|
| 529 |
+
"eval_samples_per_second": 4.996,
|
| 530 |
+
"eval_steps_per_second": 1.249,
|
| 531 |
+
"step": 630
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 9.002857142857144,
|
| 535 |
+
"grad_norm": 33.93547439575195,
|
| 536 |
+
"learning_rate": 9.07936507936508e-06,
|
| 537 |
+
"loss": 0.3914,
|
| 538 |
+
"step": 640
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 9.005714285714285,
|
| 542 |
+
"grad_norm": 266.6748962402344,
|
| 543 |
+
"learning_rate": 9.047619047619049e-06,
|
| 544 |
+
"loss": 0.0957,
|
| 545 |
+
"step": 650
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 9.008571428571429,
|
| 549 |
+
"grad_norm": 1.426464557647705,
|
| 550 |
+
"learning_rate": 9.015873015873017e-06,
|
| 551 |
+
"loss": 0.0762,
|
| 552 |
+
"step": 660
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 9.01142857142857,
|
| 556 |
+
"grad_norm": 28.881240844726562,
|
| 557 |
+
"learning_rate": 8.984126984126985e-06,
|
| 558 |
+
"loss": 0.5325,
|
| 559 |
+
"step": 670
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 9.014285714285714,
|
| 563 |
+
"grad_norm": 0.008092758245766163,
|
| 564 |
+
"learning_rate": 8.952380952380953e-06,
|
| 565 |
+
"loss": 0.3857,
|
| 566 |
+
"step": 680
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 9.017142857142858,
|
| 570 |
+
"grad_norm": 53.07158279418945,
|
| 571 |
+
"learning_rate": 8.920634920634922e-06,
|
| 572 |
+
"loss": 0.6686,
|
| 573 |
+
"step": 690
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 9.02,
|
| 577 |
+
"grad_norm": 147.34286499023438,
|
| 578 |
+
"learning_rate": 8.888888888888888e-06,
|
| 579 |
+
"loss": 0.3974,
|
| 580 |
+
"step": 700
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 9.02,
|
| 584 |
+
"eval_accuracy": 0.65625,
|
| 585 |
+
"eval_loss": 1.4708929061889648,
|
| 586 |
+
"eval_runtime": 6.0018,
|
| 587 |
+
"eval_samples_per_second": 5.332,
|
| 588 |
+
"eval_steps_per_second": 1.333,
|
| 589 |
+
"step": 700
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 10.002857142857144,
|
| 593 |
+
"grad_norm": 0.08053329586982727,
|
| 594 |
+
"learning_rate": 8.857142857142858e-06,
|
| 595 |
+
"loss": 0.2338,
|
| 596 |
+
"step": 710
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 10.005714285714285,
|
| 600 |
+
"grad_norm": 0.04598785564303398,
|
| 601 |
+
"learning_rate": 8.825396825396827e-06,
|
| 602 |
+
"loss": 0.2478,
|
| 603 |
+
"step": 720
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 10.008571428571429,
|
| 607 |
+
"grad_norm": 0.032458722591400146,
|
| 608 |
+
"learning_rate": 8.793650793650795e-06,
|
| 609 |
+
"loss": 0.3256,
|
| 610 |
+
"step": 730
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 10.01142857142857,
|
| 614 |
+
"grad_norm": 3.2461020946502686,
|
| 615 |
+
"learning_rate": 8.761904761904763e-06,
|
| 616 |
+
"loss": 0.542,
|
| 617 |
+
"step": 740
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 10.014285714285714,
|
| 621 |
+
"grad_norm": 38.98529815673828,
|
| 622 |
+
"learning_rate": 8.730158730158731e-06,
|
| 623 |
+
"loss": 0.3005,
|
| 624 |
+
"step": 750
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 10.017142857142858,
|
| 628 |
+
"grad_norm": 5.08363151550293,
|
| 629 |
+
"learning_rate": 8.6984126984127e-06,
|
| 630 |
+
"loss": 0.0276,
|
| 631 |
+
"step": 760
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 10.02,
|
| 635 |
+
"grad_norm": 1.1978472471237183,
|
| 636 |
+
"learning_rate": 8.666666666666668e-06,
|
| 637 |
+
"loss": 0.4529,
|
| 638 |
+
"step": 770
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 10.02,
|
| 642 |
+
"eval_accuracy": 0.625,
|
| 643 |
+
"eval_loss": 1.6416354179382324,
|
| 644 |
+
"eval_runtime": 6.0866,
|
| 645 |
+
"eval_samples_per_second": 5.257,
|
| 646 |
+
"eval_steps_per_second": 1.314,
|
| 647 |
+
"step": 770
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 11.002857142857144,
|
| 651 |
+
"grad_norm": 54.74907684326172,
|
| 652 |
+
"learning_rate": 8.634920634920636e-06,
|
| 653 |
+
"loss": 0.1156,
|
| 654 |
+
"step": 780
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 11.005714285714285,
|
| 658 |
+
"grad_norm": 5.403316974639893,
|
| 659 |
+
"learning_rate": 8.603174603174604e-06,
|
| 660 |
+
"loss": 0.0258,
|
| 661 |
+
"step": 790
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 11.008571428571429,
|
| 665 |
+
"grad_norm": 0.43036800622940063,
|
| 666 |
+
"learning_rate": 8.571428571428571e-06,
|
| 667 |
+
"loss": 0.1555,
|
| 668 |
+
"step": 800
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 11.01142857142857,
|
| 672 |
+
"grad_norm": 176.08836364746094,
|
| 673 |
+
"learning_rate": 8.53968253968254e-06,
|
| 674 |
+
"loss": 0.5625,
|
| 675 |
+
"step": 810
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 11.014285714285714,
|
| 679 |
+
"grad_norm": 52.73113250732422,
|
| 680 |
+
"learning_rate": 8.507936507936509e-06,
|
| 681 |
+
"loss": 0.1161,
|
| 682 |
+
"step": 820
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 11.017142857142858,
|
| 686 |
+
"grad_norm": 0.23291529715061188,
|
| 687 |
+
"learning_rate": 8.476190476190477e-06,
|
| 688 |
+
"loss": 0.0022,
|
| 689 |
+
"step": 830
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 11.02,
|
| 693 |
+
"grad_norm": 0.01016359031200409,
|
| 694 |
+
"learning_rate": 8.444444444444446e-06,
|
| 695 |
+
"loss": 0.0831,
|
| 696 |
+
"step": 840
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 11.02,
|
| 700 |
+
"eval_accuracy": 0.6875,
|
| 701 |
+
"eval_loss": 2.043224334716797,
|
| 702 |
+
"eval_runtime": 5.9603,
|
| 703 |
+
"eval_samples_per_second": 5.369,
|
| 704 |
+
"eval_steps_per_second": 1.342,
|
| 705 |
+
"step": 840
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 12.002857142857144,
|
| 709 |
+
"grad_norm": 0.025461383163928986,
|
| 710 |
+
"learning_rate": 8.412698412698414e-06,
|
| 711 |
+
"loss": 0.2046,
|
| 712 |
+
"step": 850
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 12.005714285714285,
|
| 716 |
+
"grad_norm": 192.4024658203125,
|
| 717 |
+
"learning_rate": 8.380952380952382e-06,
|
| 718 |
+
"loss": 0.3202,
|
| 719 |
+
"step": 860
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 12.008571428571429,
|
| 723 |
+
"grad_norm": 0.008105052635073662,
|
| 724 |
+
"learning_rate": 8.34920634920635e-06,
|
| 725 |
+
"loss": 0.2158,
|
| 726 |
+
"step": 870
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 12.01142857142857,
|
| 730 |
+
"grad_norm": 0.009196307510137558,
|
| 731 |
+
"learning_rate": 8.317460317460319e-06,
|
| 732 |
+
"loss": 0.1244,
|
| 733 |
+
"step": 880
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 12.014285714285714,
|
| 737 |
+
"grad_norm": 192.51708984375,
|
| 738 |
+
"learning_rate": 8.285714285714287e-06,
|
| 739 |
+
"loss": 0.1151,
|
| 740 |
+
"step": 890
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 12.017142857142858,
|
| 744 |
+
"grad_norm": 1.3545656204223633,
|
| 745 |
+
"learning_rate": 8.253968253968254e-06,
|
| 746 |
+
"loss": 0.1114,
|
| 747 |
+
"step": 900
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 12.02,
|
| 751 |
+
"grad_norm": 0.9690192341804504,
|
| 752 |
+
"learning_rate": 8.222222222222222e-06,
|
| 753 |
+
"loss": 0.0104,
|
| 754 |
+
"step": 910
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 12.02,
|
| 758 |
+
"eval_accuracy": 0.59375,
|
| 759 |
+
"eval_loss": 2.1283388137817383,
|
| 760 |
+
"eval_runtime": 6.3439,
|
| 761 |
+
"eval_samples_per_second": 5.044,
|
| 762 |
+
"eval_steps_per_second": 1.261,
|
| 763 |
+
"step": 910
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 13.002857142857144,
|
| 767 |
+
"grad_norm": 0.026668542996048927,
|
| 768 |
+
"learning_rate": 8.190476190476192e-06,
|
| 769 |
+
"loss": 0.0327,
|
| 770 |
+
"step": 920
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 13.005714285714285,
|
| 774 |
+
"grad_norm": 0.01906030997633934,
|
| 775 |
+
"learning_rate": 8.15873015873016e-06,
|
| 776 |
+
"loss": 0.1063,
|
| 777 |
+
"step": 930
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 13.008571428571429,
|
| 781 |
+
"grad_norm": 0.13048304617404938,
|
| 782 |
+
"learning_rate": 8.126984126984128e-06,
|
| 783 |
+
"loss": 0.1734,
|
| 784 |
+
"step": 940
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 13.01142857142857,
|
| 788 |
+
"grad_norm": 0.012808074243366718,
|
| 789 |
+
"learning_rate": 8.095238095238097e-06,
|
| 790 |
+
"loss": 0.087,
|
| 791 |
+
"step": 950
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 13.014285714285714,
|
| 795 |
+
"grad_norm": 219.556640625,
|
| 796 |
+
"learning_rate": 8.063492063492065e-06,
|
| 797 |
+
"loss": 0.2768,
|
| 798 |
+
"step": 960
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 13.017142857142858,
|
| 802 |
+
"grad_norm": 0.009385952726006508,
|
| 803 |
+
"learning_rate": 8.031746031746033e-06,
|
| 804 |
+
"loss": 0.005,
|
| 805 |
+
"step": 970
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 13.02,
|
| 809 |
+
"grad_norm": 21.89446449279785,
|
| 810 |
+
"learning_rate": 8.000000000000001e-06,
|
| 811 |
+
"loss": 0.0287,
|
| 812 |
+
"step": 980
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 13.02,
|
| 816 |
+
"eval_accuracy": 0.75,
|
| 817 |
+
"eval_loss": 1.853172779083252,
|
| 818 |
+
"eval_runtime": 5.936,
|
| 819 |
+
"eval_samples_per_second": 5.391,
|
| 820 |
+
"eval_steps_per_second": 1.348,
|
| 821 |
+
"step": 980
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 14.002857142857144,
|
| 825 |
+
"grad_norm": 0.0067256903275847435,
|
| 826 |
+
"learning_rate": 7.968253968253968e-06,
|
| 827 |
+
"loss": 0.0056,
|
| 828 |
+
"step": 990
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 14.005714285714285,
|
| 832 |
+
"grad_norm": 309.82196044921875,
|
| 833 |
+
"learning_rate": 7.936507936507936e-06,
|
| 834 |
+
"loss": 0.1169,
|
| 835 |
+
"step": 1000
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 14.008571428571429,
|
| 839 |
+
"grad_norm": 214.8688201904297,
|
| 840 |
+
"learning_rate": 7.904761904761904e-06,
|
| 841 |
+
"loss": 0.1285,
|
| 842 |
+
"step": 1010
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 14.01142857142857,
|
| 846 |
+
"grad_norm": 0.005147908348590136,
|
| 847 |
+
"learning_rate": 7.873015873015873e-06,
|
| 848 |
+
"loss": 0.0166,
|
| 849 |
+
"step": 1020
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 14.014285714285714,
|
| 853 |
+
"grad_norm": 0.013515499420464039,
|
| 854 |
+
"learning_rate": 7.841269841269843e-06,
|
| 855 |
+
"loss": 0.0549,
|
| 856 |
+
"step": 1030
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 14.017142857142858,
|
| 860 |
+
"grad_norm": 0.005485304165631533,
|
| 861 |
+
"learning_rate": 7.809523809523811e-06,
|
| 862 |
+
"loss": 0.0002,
|
| 863 |
+
"step": 1040
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 14.02,
|
| 867 |
+
"grad_norm": 0.026508044451475143,
|
| 868 |
+
"learning_rate": 7.77777777777778e-06,
|
| 869 |
+
"loss": 0.0426,
|
| 870 |
+
"step": 1050
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 14.02,
|
| 874 |
+
"eval_accuracy": 0.6875,
|
| 875 |
+
"eval_loss": 2.092482089996338,
|
| 876 |
+
"eval_runtime": 6.4209,
|
| 877 |
+
"eval_samples_per_second": 4.984,
|
| 878 |
+
"eval_steps_per_second": 1.246,
|
| 879 |
+
"step": 1050
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 15.002857142857144,
|
| 883 |
+
"grad_norm": 0.49312490224838257,
|
| 884 |
+
"learning_rate": 7.746031746031747e-06,
|
| 885 |
+
"loss": 0.0009,
|
| 886 |
+
"step": 1060
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 15.005714285714285,
|
| 890 |
+
"grad_norm": 0.02295481227338314,
|
| 891 |
+
"learning_rate": 7.714285714285716e-06,
|
| 892 |
+
"loss": 0.0036,
|
| 893 |
+
"step": 1070
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 15.008571428571429,
|
| 897 |
+
"grad_norm": 0.009320501238107681,
|
| 898 |
+
"learning_rate": 7.682539682539684e-06,
|
| 899 |
+
"loss": 0.1635,
|
| 900 |
+
"step": 1080
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 15.01142857142857,
|
| 904 |
+
"grad_norm": 0.01872088387608528,
|
| 905 |
+
"learning_rate": 7.65079365079365e-06,
|
| 906 |
+
"loss": 0.4035,
|
| 907 |
+
"step": 1090
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 15.014285714285714,
|
| 911 |
+
"grad_norm": 0.007027905434370041,
|
| 912 |
+
"learning_rate": 7.61904761904762e-06,
|
| 913 |
+
"loss": 0.0004,
|
| 914 |
+
"step": 1100
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 15.017142857142858,
|
| 918 |
+
"grad_norm": 0.006057267542928457,
|
| 919 |
+
"learning_rate": 7.587301587301588e-06,
|
| 920 |
+
"loss": 0.3872,
|
| 921 |
+
"step": 1110
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 15.02,
|
| 925 |
+
"grad_norm": 3.802086591720581,
|
| 926 |
+
"learning_rate": 7.555555555555556e-06,
|
| 927 |
+
"loss": 0.0145,
|
| 928 |
+
"step": 1120
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 15.02,
|
| 932 |
+
"eval_accuracy": 0.625,
|
| 933 |
+
"eval_loss": 2.351853370666504,
|
| 934 |
+
"eval_runtime": 5.7934,
|
| 935 |
+
"eval_samples_per_second": 5.524,
|
| 936 |
+
"eval_steps_per_second": 1.381,
|
| 937 |
+
"step": 1120
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 16.002857142857142,
|
| 941 |
+
"grad_norm": 92.55747985839844,
|
| 942 |
+
"learning_rate": 7.523809523809524e-06,
|
| 943 |
+
"loss": 0.1877,
|
| 944 |
+
"step": 1130
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 16.005714285714287,
|
| 948 |
+
"grad_norm": 0.005619277711957693,
|
| 949 |
+
"learning_rate": 7.492063492063493e-06,
|
| 950 |
+
"loss": 0.001,
|
| 951 |
+
"step": 1140
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 16.00857142857143,
|
| 955 |
+
"grad_norm": 196.1888885498047,
|
| 956 |
+
"learning_rate": 7.460317460317461e-06,
|
| 957 |
+
"loss": 0.1907,
|
| 958 |
+
"step": 1150
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 16.01142857142857,
|
| 962 |
+
"grad_norm": 0.005707439035177231,
|
| 963 |
+
"learning_rate": 7.428571428571429e-06,
|
| 964 |
+
"loss": 0.1524,
|
| 965 |
+
"step": 1160
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 16.014285714285716,
|
| 969 |
+
"grad_norm": 0.018987352028489113,
|
| 970 |
+
"learning_rate": 7.3968253968253975e-06,
|
| 971 |
+
"loss": 0.2278,
|
| 972 |
+
"step": 1170
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 16.017142857142858,
|
| 976 |
+
"grad_norm": 0.029227713122963905,
|
| 977 |
+
"learning_rate": 7.3650793650793666e-06,
|
| 978 |
+
"loss": 0.0281,
|
| 979 |
+
"step": 1180
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 16.02,
|
| 983 |
+
"grad_norm": 0.006308805197477341,
|
| 984 |
+
"learning_rate": 7.333333333333333e-06,
|
| 985 |
+
"loss": 0.0024,
|
| 986 |
+
"step": 1190
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 16.02,
|
| 990 |
+
"eval_accuracy": 0.625,
|
| 991 |
+
"eval_loss": 2.070525646209717,
|
| 992 |
+
"eval_runtime": 5.8174,
|
| 993 |
+
"eval_samples_per_second": 5.501,
|
| 994 |
+
"eval_steps_per_second": 1.375,
|
| 995 |
+
"step": 1190
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 17.002857142857142,
|
| 999 |
+
"grad_norm": 0.0030912691727280617,
|
| 1000 |
+
"learning_rate": 7.301587301587301e-06,
|
| 1001 |
+
"loss": 0.1571,
|
| 1002 |
+
"step": 1200
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 17.005714285714287,
|
| 1006 |
+
"grad_norm": 0.024552060291171074,
|
| 1007 |
+
"learning_rate": 7.2698412698412705e-06,
|
| 1008 |
+
"loss": 0.0002,
|
| 1009 |
+
"step": 1210
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 17.00857142857143,
|
| 1013 |
+
"grad_norm": 0.01592804305255413,
|
| 1014 |
+
"learning_rate": 7.238095238095239e-06,
|
| 1015 |
+
"loss": 0.0002,
|
| 1016 |
+
"step": 1220
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 17.01142857142857,
|
| 1020 |
+
"grad_norm": 0.039011064916849136,
|
| 1021 |
+
"learning_rate": 7.206349206349207e-06,
|
| 1022 |
+
"loss": 0.0003,
|
| 1023 |
+
"step": 1230
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 17.014285714285716,
|
| 1027 |
+
"grad_norm": 0.0051954942755401134,
|
| 1028 |
+
"learning_rate": 7.174603174603175e-06,
|
| 1029 |
+
"loss": 0.0028,
|
| 1030 |
+
"step": 1240
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 17.017142857142858,
|
| 1034 |
+
"grad_norm": 0.014356645755469799,
|
| 1035 |
+
"learning_rate": 7.1428571428571436e-06,
|
| 1036 |
+
"loss": 0.0001,
|
| 1037 |
+
"step": 1250
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 17.02,
|
| 1041 |
+
"grad_norm": 0.04629682004451752,
|
| 1042 |
+
"learning_rate": 7.111111111111112e-06,
|
| 1043 |
+
"loss": 0.0176,
|
| 1044 |
+
"step": 1260
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 17.02,
|
| 1048 |
+
"eval_accuracy": 0.625,
|
| 1049 |
+
"eval_loss": 2.360275983810425,
|
| 1050 |
+
"eval_runtime": 6.3101,
|
| 1051 |
+
"eval_samples_per_second": 5.071,
|
| 1052 |
+
"eval_steps_per_second": 1.268,
|
| 1053 |
+
"step": 1260
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 18.002857142857142,
|
| 1057 |
+
"grad_norm": 0.002282822737470269,
|
| 1058 |
+
"learning_rate": 7.07936507936508e-06,
|
| 1059 |
+
"loss": 0.0003,
|
| 1060 |
+
"step": 1270
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 18.005714285714287,
|
| 1064 |
+
"grad_norm": 0.007674033287912607,
|
| 1065 |
+
"learning_rate": 7.047619047619048e-06,
|
| 1066 |
+
"loss": 0.005,
|
| 1067 |
+
"step": 1280
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 18.00857142857143,
|
| 1071 |
+
"grad_norm": 107.018310546875,
|
| 1072 |
+
"learning_rate": 7.015873015873016e-06,
|
| 1073 |
+
"loss": 0.223,
|
| 1074 |
+
"step": 1290
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 18.01142857142857,
|
| 1078 |
+
"grad_norm": 0.027542075142264366,
|
| 1079 |
+
"learning_rate": 6.984126984126984e-06,
|
| 1080 |
+
"loss": 0.2571,
|
| 1081 |
+
"step": 1300
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 18.014285714285716,
|
| 1085 |
+
"grad_norm": 0.013514521531760693,
|
| 1086 |
+
"learning_rate": 6.952380952380952e-06,
|
| 1087 |
+
"loss": 0.1242,
|
| 1088 |
+
"step": 1310
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 18.017142857142858,
|
| 1092 |
+
"grad_norm": 0.004058144986629486,
|
| 1093 |
+
"learning_rate": 6.920634920634921e-06,
|
| 1094 |
+
"loss": 0.0026,
|
| 1095 |
+
"step": 1320
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 18.02,
|
| 1099 |
+
"grad_norm": 0.052907612174749374,
|
| 1100 |
+
"learning_rate": 6.88888888888889e-06,
|
| 1101 |
+
"loss": 0.0023,
|
| 1102 |
+
"step": 1330
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 18.02,
|
| 1106 |
+
"eval_accuracy": 0.65625,
|
| 1107 |
+
"eval_loss": 2.3409135341644287,
|
| 1108 |
+
"eval_runtime": 5.6456,
|
| 1109 |
+
"eval_samples_per_second": 5.668,
|
| 1110 |
+
"eval_steps_per_second": 1.417,
|
| 1111 |
+
"step": 1330
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 19.002857142857142,
|
| 1115 |
+
"grad_norm": 0.003937090281397104,
|
| 1116 |
+
"learning_rate": 6.857142857142858e-06,
|
| 1117 |
+
"loss": 0.0083,
|
| 1118 |
+
"step": 1340
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 19.005714285714287,
|
| 1122 |
+
"grad_norm": 0.17072078585624695,
|
| 1123 |
+
"learning_rate": 6.825396825396826e-06,
|
| 1124 |
+
"loss": 0.0016,
|
| 1125 |
+
"step": 1350
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 19.00857142857143,
|
| 1129 |
+
"grad_norm": 0.0025130421854555607,
|
| 1130 |
+
"learning_rate": 6.7936507936507944e-06,
|
| 1131 |
+
"loss": 0.0018,
|
| 1132 |
+
"step": 1360
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 19.01142857142857,
|
| 1136 |
+
"grad_norm": 0.0014831767184659839,
|
| 1137 |
+
"learning_rate": 6.761904761904763e-06,
|
| 1138 |
+
"loss": 0.007,
|
| 1139 |
+
"step": 1370
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 19.014285714285716,
|
| 1143 |
+
"grad_norm": 48.3748893737793,
|
| 1144 |
+
"learning_rate": 6.730158730158731e-06,
|
| 1145 |
+
"loss": 0.2262,
|
| 1146 |
+
"step": 1380
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 19.017142857142858,
|
| 1150 |
+
"grad_norm": 0.0189586840569973,
|
| 1151 |
+
"learning_rate": 6.698412698412698e-06,
|
| 1152 |
+
"loss": 0.1803,
|
| 1153 |
+
"step": 1390
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 19.02,
|
| 1157 |
+
"grad_norm": 0.003652828047052026,
|
| 1158 |
+
"learning_rate": 6.666666666666667e-06,
|
| 1159 |
+
"loss": 0.1111,
|
| 1160 |
+
"step": 1400
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 19.02,
|
| 1164 |
+
"eval_accuracy": 0.625,
|
| 1165 |
+
"eval_loss": 2.3460936546325684,
|
| 1166 |
+
"eval_runtime": 5.6345,
|
| 1167 |
+
"eval_samples_per_second": 5.679,
|
| 1168 |
+
"eval_steps_per_second": 1.42,
|
| 1169 |
+
"step": 1400
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 20.002857142857142,
|
| 1173 |
+
"grad_norm": 0.033134013414382935,
|
| 1174 |
+
"learning_rate": 6.634920634920635e-06,
|
| 1175 |
+
"loss": 0.2262,
|
| 1176 |
+
"step": 1410
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 20.005714285714287,
|
| 1180 |
+
"grad_norm": 0.015590427443385124,
|
| 1181 |
+
"learning_rate": 6.603174603174603e-06,
|
| 1182 |
+
"loss": 0.0606,
|
| 1183 |
+
"step": 1420
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 20.00857142857143,
|
| 1187 |
+
"grad_norm": 0.008228067308664322,
|
| 1188 |
+
"learning_rate": 6.571428571428572e-06,
|
| 1189 |
+
"loss": 0.0001,
|
| 1190 |
+
"step": 1430
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 20.01142857142857,
|
| 1194 |
+
"grad_norm": 0.007313844282180071,
|
| 1195 |
+
"learning_rate": 6.5396825396825405e-06,
|
| 1196 |
+
"loss": 0.0812,
|
| 1197 |
+
"step": 1440
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 20.014285714285716,
|
| 1201 |
+
"grad_norm": 0.03062896430492401,
|
| 1202 |
+
"learning_rate": 6.507936507936509e-06,
|
| 1203 |
+
"loss": 0.3162,
|
| 1204 |
+
"step": 1450
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 20.017142857142858,
|
| 1208 |
+
"grad_norm": 0.00798776838928461,
|
| 1209 |
+
"learning_rate": 6.476190476190477e-06,
|
| 1210 |
+
"loss": 0.0034,
|
| 1211 |
+
"step": 1460
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 20.02,
|
| 1215 |
+
"grad_norm": 0.0768931582570076,
|
| 1216 |
+
"learning_rate": 6.444444444444445e-06,
|
| 1217 |
+
"loss": 0.0002,
|
| 1218 |
+
"step": 1470
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 20.02,
|
| 1222 |
+
"eval_accuracy": 0.6875,
|
| 1223 |
+
"eval_loss": 2.0998449325561523,
|
| 1224 |
+
"eval_runtime": 6.206,
|
| 1225 |
+
"eval_samples_per_second": 5.156,
|
| 1226 |
+
"eval_steps_per_second": 1.289,
|
| 1227 |
+
"step": 1470
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 21.002857142857142,
|
| 1231 |
+
"grad_norm": 0.08019956201314926,
|
| 1232 |
+
"learning_rate": 6.412698412698414e-06,
|
| 1233 |
+
"loss": 0.0007,
|
| 1234 |
+
"step": 1480
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 21.005714285714287,
|
| 1238 |
+
"grad_norm": 0.008561481721699238,
|
| 1239 |
+
"learning_rate": 6.380952380952381e-06,
|
| 1240 |
+
"loss": 0.0003,
|
| 1241 |
+
"step": 1490
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 21.00857142857143,
|
| 1245 |
+
"grad_norm": 0.17695729434490204,
|
| 1246 |
+
"learning_rate": 6.349206349206349e-06,
|
| 1247 |
+
"loss": 0.0085,
|
| 1248 |
+
"step": 1500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 21.01142857142857,
|
| 1252 |
+
"grad_norm": 0.0022817952558398247,
|
| 1253 |
+
"learning_rate": 6.3174603174603175e-06,
|
| 1254 |
+
"loss": 0.0001,
|
| 1255 |
+
"step": 1510
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 21.014285714285716,
|
| 1259 |
+
"grad_norm": 0.004223952069878578,
|
| 1260 |
+
"learning_rate": 6.285714285714286e-06,
|
| 1261 |
+
"loss": 0.0008,
|
| 1262 |
+
"step": 1520
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 21.017142857142858,
|
| 1266 |
+
"grad_norm": 0.004527420271188021,
|
| 1267 |
+
"learning_rate": 6.253968253968254e-06,
|
| 1268 |
+
"loss": 0.0002,
|
| 1269 |
+
"step": 1530
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 21.02,
|
| 1273 |
+
"grad_norm": 0.020056281238794327,
|
| 1274 |
+
"learning_rate": 6.222222222222223e-06,
|
| 1275 |
+
"loss": 0.0432,
|
| 1276 |
+
"step": 1540
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 21.02,
|
| 1280 |
+
"eval_accuracy": 0.59375,
|
| 1281 |
+
"eval_loss": 2.2702555656433105,
|
| 1282 |
+
"eval_runtime": 5.806,
|
| 1283 |
+
"eval_samples_per_second": 5.512,
|
| 1284 |
+
"eval_steps_per_second": 1.378,
|
| 1285 |
+
"step": 1540
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 22.002857142857142,
|
| 1289 |
+
"grad_norm": 0.002663273364305496,
|
| 1290 |
+
"learning_rate": 6.1904761904761914e-06,
|
| 1291 |
+
"loss": 0.1748,
|
| 1292 |
+
"step": 1550
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 22.005714285714287,
|
| 1296 |
+
"grad_norm": 0.002945307642221451,
|
| 1297 |
+
"learning_rate": 6.15873015873016e-06,
|
| 1298 |
+
"loss": 0.0364,
|
| 1299 |
+
"step": 1560
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 22.00857142857143,
|
| 1303 |
+
"grad_norm": 6.704668045043945,
|
| 1304 |
+
"learning_rate": 6.126984126984128e-06,
|
| 1305 |
+
"loss": 0.0104,
|
| 1306 |
+
"step": 1570
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 22.01142857142857,
|
| 1310 |
+
"grad_norm": 0.0020908783189952374,
|
| 1311 |
+
"learning_rate": 6.095238095238096e-06,
|
| 1312 |
+
"loss": 0.0893,
|
| 1313 |
+
"step": 1580
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 22.014285714285716,
|
| 1317 |
+
"grad_norm": 0.004042261280119419,
|
| 1318 |
+
"learning_rate": 6.063492063492064e-06,
|
| 1319 |
+
"loss": 0.1432,
|
| 1320 |
+
"step": 1590
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 22.017142857142858,
|
| 1324 |
+
"grad_norm": 0.026853501796722412,
|
| 1325 |
+
"learning_rate": 6.031746031746032e-06,
|
| 1326 |
+
"loss": 0.0002,
|
| 1327 |
+
"step": 1600
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 22.02,
|
| 1331 |
+
"grad_norm": 0.002750640967860818,
|
| 1332 |
+
"learning_rate": 6e-06,
|
| 1333 |
+
"loss": 0.006,
|
| 1334 |
+
"step": 1610
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 22.02,
|
| 1338 |
+
"eval_accuracy": 0.65625,
|
| 1339 |
+
"eval_loss": 2.3306009769439697,
|
| 1340 |
+
"eval_runtime": 5.6486,
|
| 1341 |
+
"eval_samples_per_second": 5.665,
|
| 1342 |
+
"eval_steps_per_second": 1.416,
|
| 1343 |
+
"step": 1610
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 23.002857142857142,
|
| 1347 |
+
"grad_norm": 0.0037402939051389694,
|
| 1348 |
+
"learning_rate": 5.968253968253968e-06,
|
| 1349 |
+
"loss": 0.0001,
|
| 1350 |
+
"step": 1620
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 23.005714285714287,
|
| 1354 |
+
"grad_norm": 0.002422990510240197,
|
| 1355 |
+
"learning_rate": 5.936507936507937e-06,
|
| 1356 |
+
"loss": 0.0014,
|
| 1357 |
+
"step": 1630
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 23.00857142857143,
|
| 1361 |
+
"grad_norm": 0.004334081895649433,
|
| 1362 |
+
"learning_rate": 5.904761904761905e-06,
|
| 1363 |
+
"loss": 0.0052,
|
| 1364 |
+
"step": 1640
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 23.01142857142857,
|
| 1368 |
+
"grad_norm": 0.0010784378973767161,
|
| 1369 |
+
"learning_rate": 5.873015873015874e-06,
|
| 1370 |
+
"loss": 0.1291,
|
| 1371 |
+
"step": 1650
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 23.014285714285716,
|
| 1375 |
+
"grad_norm": 0.0017067866865545511,
|
| 1376 |
+
"learning_rate": 5.841269841269842e-06,
|
| 1377 |
+
"loss": 0.0001,
|
| 1378 |
+
"step": 1660
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 23.017142857142858,
|
| 1382 |
+
"grad_norm": 0.08872966468334198,
|
| 1383 |
+
"learning_rate": 5.8095238095238106e-06,
|
| 1384 |
+
"loss": 0.0001,
|
| 1385 |
+
"step": 1670
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 23.02,
|
| 1389 |
+
"grad_norm": 0.001540614292025566,
|
| 1390 |
+
"learning_rate": 5.777777777777778e-06,
|
| 1391 |
+
"loss": 0.0014,
|
| 1392 |
+
"step": 1680
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 23.02,
|
| 1396 |
+
"eval_accuracy": 0.625,
|
| 1397 |
+
"eval_loss": 2.446324348449707,
|
| 1398 |
+
"eval_runtime": 6.0676,
|
| 1399 |
+
"eval_samples_per_second": 5.274,
|
| 1400 |
+
"eval_steps_per_second": 1.318,
|
| 1401 |
+
"step": 1680
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 24.002857142857142,
|
| 1405 |
+
"grad_norm": 0.005013082176446915,
|
| 1406 |
+
"learning_rate": 5.746031746031746e-06,
|
| 1407 |
+
"loss": 0.0001,
|
| 1408 |
+
"step": 1690
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 24.005714285714287,
|
| 1412 |
+
"grad_norm": 0.09714154154062271,
|
| 1413 |
+
"learning_rate": 5.7142857142857145e-06,
|
| 1414 |
+
"loss": 0.0016,
|
| 1415 |
+
"step": 1700
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 24.00857142857143,
|
| 1419 |
+
"grad_norm": 0.0058003757148981094,
|
| 1420 |
+
"learning_rate": 5.682539682539683e-06,
|
| 1421 |
+
"loss": 0.0001,
|
| 1422 |
+
"step": 1710
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 24.01142857142857,
|
| 1426 |
+
"grad_norm": 0.0029599149711430073,
|
| 1427 |
+
"learning_rate": 5.650793650793651e-06,
|
| 1428 |
+
"loss": 0.0024,
|
| 1429 |
+
"step": 1720
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 24.014285714285716,
|
| 1433 |
+
"grad_norm": 0.001937561435624957,
|
| 1434 |
+
"learning_rate": 5.619047619047619e-06,
|
| 1435 |
+
"loss": 0.1056,
|
| 1436 |
+
"step": 1730
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 24.017142857142858,
|
| 1440 |
+
"grad_norm": 0.02229410596191883,
|
| 1441 |
+
"learning_rate": 5.5873015873015876e-06,
|
| 1442 |
+
"loss": 0.0001,
|
| 1443 |
+
"step": 1740
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 24.02,
|
| 1447 |
+
"grad_norm": 0.004641120787709951,
|
| 1448 |
+
"learning_rate": 5.555555555555557e-06,
|
| 1449 |
+
"loss": 0.0173,
|
| 1450 |
+
"step": 1750
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 24.02,
|
| 1454 |
+
"eval_accuracy": 0.65625,
|
| 1455 |
+
"eval_loss": 2.556457042694092,
|
| 1456 |
+
"eval_runtime": 5.7748,
|
| 1457 |
+
"eval_samples_per_second": 5.541,
|
| 1458 |
+
"eval_steps_per_second": 1.385,
|
| 1459 |
+
"step": 1750
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 25.002857142857142,
|
| 1463 |
+
"grad_norm": 0.0013144315453246236,
|
| 1464 |
+
"learning_rate": 5.523809523809525e-06,
|
| 1465 |
+
"loss": 0.0001,
|
| 1466 |
+
"step": 1760
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 25.005714285714287,
|
| 1470 |
+
"grad_norm": 0.00479417759925127,
|
| 1471 |
+
"learning_rate": 5.492063492063493e-06,
|
| 1472 |
+
"loss": 0.071,
|
| 1473 |
+
"step": 1770
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 25.00857142857143,
|
| 1477 |
+
"grad_norm": 0.0011662091128528118,
|
| 1478 |
+
"learning_rate": 5.460317460317461e-06,
|
| 1479 |
+
"loss": 0.0001,
|
| 1480 |
+
"step": 1780
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 25.01142857142857,
|
| 1484 |
+
"grad_norm": 0.003162315348163247,
|
| 1485 |
+
"learning_rate": 5.428571428571429e-06,
|
| 1486 |
+
"loss": 0.0001,
|
| 1487 |
+
"step": 1790
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 25.014285714285716,
|
| 1491 |
+
"grad_norm": 0.004806616809219122,
|
| 1492 |
+
"learning_rate": 5.396825396825397e-06,
|
| 1493 |
+
"loss": 0.0,
|
| 1494 |
+
"step": 1800
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 25.017142857142858,
|
| 1498 |
+
"grad_norm": 0.010089408606290817,
|
| 1499 |
+
"learning_rate": 5.365079365079365e-06,
|
| 1500 |
+
"loss": 0.0001,
|
| 1501 |
+
"step": 1810
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 25.02,
|
| 1505 |
+
"grad_norm": 0.0010941866785287857,
|
| 1506 |
+
"learning_rate": 5.333333333333334e-06,
|
| 1507 |
+
"loss": 0.0001,
|
| 1508 |
+
"step": 1820
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 25.02,
|
| 1512 |
+
"eval_accuracy": 0.625,
|
| 1513 |
+
"eval_loss": 2.7107367515563965,
|
| 1514 |
+
"eval_runtime": 6.0469,
|
| 1515 |
+
"eval_samples_per_second": 5.292,
|
| 1516 |
+
"eval_steps_per_second": 1.323,
|
| 1517 |
+
"step": 1820
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 26.002857142857142,
|
| 1521 |
+
"grad_norm": 0.003195547964423895,
|
| 1522 |
+
"learning_rate": 5.301587301587302e-06,
|
| 1523 |
+
"loss": 0.1521,
|
| 1524 |
+
"step": 1830
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 26.005714285714287,
|
| 1528 |
+
"grad_norm": 0.009643026627600193,
|
| 1529 |
+
"learning_rate": 5.26984126984127e-06,
|
| 1530 |
+
"loss": 0.0005,
|
| 1531 |
+
"step": 1840
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 26.00857142857143,
|
| 1535 |
+
"grad_norm": 0.0033331862650811672,
|
| 1536 |
+
"learning_rate": 5.2380952380952384e-06,
|
| 1537 |
+
"loss": 0.0001,
|
| 1538 |
+
"step": 1850
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 26.01142857142857,
|
| 1542 |
+
"grad_norm": 0.0029240387957543135,
|
| 1543 |
+
"learning_rate": 5.2063492063492076e-06,
|
| 1544 |
+
"loss": 0.0001,
|
| 1545 |
+
"step": 1860
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 26.014285714285716,
|
| 1549 |
+
"grad_norm": 0.007532194256782532,
|
| 1550 |
+
"learning_rate": 5.174603174603176e-06,
|
| 1551 |
+
"loss": 0.0001,
|
| 1552 |
+
"step": 1870
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 26.017142857142858,
|
| 1556 |
+
"grad_norm": 0.00938828382641077,
|
| 1557 |
+
"learning_rate": 5.142857142857142e-06,
|
| 1558 |
+
"loss": 0.0001,
|
| 1559 |
+
"step": 1880
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 26.02,
|
| 1563 |
+
"grad_norm": 0.0021244913805276155,
|
| 1564 |
+
"learning_rate": 5.1111111111111115e-06,
|
| 1565 |
+
"loss": 0.0001,
|
| 1566 |
+
"step": 1890
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 26.02,
|
| 1570 |
+
"eval_accuracy": 0.59375,
|
| 1571 |
+
"eval_loss": 3.092498302459717,
|
| 1572 |
+
"eval_runtime": 6.514,
|
| 1573 |
+
"eval_samples_per_second": 4.912,
|
| 1574 |
+
"eval_steps_per_second": 1.228,
|
| 1575 |
+
"step": 1890
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 27.002857142857142,
|
| 1579 |
+
"grad_norm": 0.0038645656313747168,
|
| 1580 |
+
"learning_rate": 5.07936507936508e-06,
|
| 1581 |
+
"loss": 0.0001,
|
| 1582 |
+
"step": 1900
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 27.005714285714287,
|
| 1586 |
+
"grad_norm": 0.0035917272325605154,
|
| 1587 |
+
"learning_rate": 5.047619047619048e-06,
|
| 1588 |
+
"loss": 0.0001,
|
| 1589 |
+
"step": 1910
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 27.00857142857143,
|
| 1593 |
+
"grad_norm": 0.002653130330145359,
|
| 1594 |
+
"learning_rate": 5.015873015873016e-06,
|
| 1595 |
+
"loss": 0.0001,
|
| 1596 |
+
"step": 1920
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 27.01142857142857,
|
| 1600 |
+
"grad_norm": 0.001524316263385117,
|
| 1601 |
+
"learning_rate": 4.9841269841269845e-06,
|
| 1602 |
+
"loss": 0.0001,
|
| 1603 |
+
"step": 1930
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 27.014285714285716,
|
| 1607 |
+
"grad_norm": 0.00874999351799488,
|
| 1608 |
+
"learning_rate": 4.952380952380953e-06,
|
| 1609 |
+
"loss": 0.0001,
|
| 1610 |
+
"step": 1940
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 27.017142857142858,
|
| 1614 |
+
"grad_norm": 0.0011307639069855213,
|
| 1615 |
+
"learning_rate": 4.920634920634921e-06,
|
| 1616 |
+
"loss": 0.0001,
|
| 1617 |
+
"step": 1950
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 27.02,
|
| 1621 |
+
"grad_norm": 0.0028375948313623667,
|
| 1622 |
+
"learning_rate": 4.888888888888889e-06,
|
| 1623 |
+
"loss": 0.1792,
|
| 1624 |
+
"step": 1960
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 27.02,
|
| 1628 |
+
"eval_accuracy": 0.625,
|
| 1629 |
+
"eval_loss": 2.536710500717163,
|
| 1630 |
+
"eval_runtime": 5.634,
|
| 1631 |
+
"eval_samples_per_second": 5.68,
|
| 1632 |
+
"eval_steps_per_second": 1.42,
|
| 1633 |
+
"step": 1960
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 28.002857142857142,
|
| 1637 |
+
"grad_norm": 0.00281306984834373,
|
| 1638 |
+
"learning_rate": 4.857142857142858e-06,
|
| 1639 |
+
"loss": 0.0049,
|
| 1640 |
+
"step": 1970
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 28.005714285714287,
|
| 1644 |
+
"grad_norm": 0.001993451965972781,
|
| 1645 |
+
"learning_rate": 4.825396825396826e-06,
|
| 1646 |
+
"loss": 0.0001,
|
| 1647 |
+
"step": 1980
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 28.00857142857143,
|
| 1651 |
+
"grad_norm": 0.0021242112852633,
|
| 1652 |
+
"learning_rate": 4.793650793650794e-06,
|
| 1653 |
+
"loss": 0.0,
|
| 1654 |
+
"step": 1990
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 28.01142857142857,
|
| 1658 |
+
"grad_norm": 0.0019546120893210173,
|
| 1659 |
+
"learning_rate": 4.761904761904762e-06,
|
| 1660 |
+
"loss": 0.0001,
|
| 1661 |
+
"step": 2000
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 28.014285714285716,
|
| 1665 |
+
"grad_norm": 0.0021625503432005644,
|
| 1666 |
+
"learning_rate": 4.730158730158731e-06,
|
| 1667 |
+
"loss": 0.0144,
|
| 1668 |
+
"step": 2010
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 28.017142857142858,
|
| 1672 |
+
"grad_norm": 0.001239254605025053,
|
| 1673 |
+
"learning_rate": 4.698412698412699e-06,
|
| 1674 |
+
"loss": 0.0001,
|
| 1675 |
+
"step": 2020
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 28.02,
|
| 1679 |
+
"grad_norm": 0.0012608050601556897,
|
| 1680 |
+
"learning_rate": 4.666666666666667e-06,
|
| 1681 |
+
"loss": 0.0004,
|
| 1682 |
+
"step": 2030
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 28.02,
|
| 1686 |
+
"eval_accuracy": 0.5625,
|
| 1687 |
+
"eval_loss": 2.9952054023742676,
|
| 1688 |
+
"eval_runtime": 6.2717,
|
| 1689 |
+
"eval_samples_per_second": 5.102,
|
| 1690 |
+
"eval_steps_per_second": 1.276,
|
| 1691 |
+
"step": 2030
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 29.002857142857142,
|
| 1695 |
+
"grad_norm": 0.0023157112300395966,
|
| 1696 |
+
"learning_rate": 4.634920634920635e-06,
|
| 1697 |
+
"loss": 0.0001,
|
| 1698 |
+
"step": 2040
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 29.005714285714287,
|
| 1702 |
+
"grad_norm": 0.0012295986525714397,
|
| 1703 |
+
"learning_rate": 4.603174603174604e-06,
|
| 1704 |
+
"loss": 0.0001,
|
| 1705 |
+
"step": 2050
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 29.00857142857143,
|
| 1709 |
+
"grad_norm": 0.005791760981082916,
|
| 1710 |
+
"learning_rate": 4.571428571428572e-06,
|
| 1711 |
+
"loss": 0.0001,
|
| 1712 |
+
"step": 2060
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 29.01142857142857,
|
| 1716 |
+
"grad_norm": 0.003394305007532239,
|
| 1717 |
+
"learning_rate": 4.53968253968254e-06,
|
| 1718 |
+
"loss": 0.1151,
|
| 1719 |
+
"step": 2070
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 29.014285714285716,
|
| 1723 |
+
"grad_norm": 0.0017611499642953277,
|
| 1724 |
+
"learning_rate": 4.5079365079365085e-06,
|
| 1725 |
+
"loss": 0.0001,
|
| 1726 |
+
"step": 2080
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 29.017142857142858,
|
| 1730 |
+
"grad_norm": 0.0028206913266330957,
|
| 1731 |
+
"learning_rate": 4.476190476190477e-06,
|
| 1732 |
+
"loss": 0.0001,
|
| 1733 |
+
"step": 2090
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 29.02,
|
| 1737 |
+
"grad_norm": 0.0028997170738875866,
|
| 1738 |
+
"learning_rate": 4.444444444444444e-06,
|
| 1739 |
+
"loss": 0.0001,
|
| 1740 |
+
"step": 2100
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 29.02,
|
| 1744 |
+
"eval_accuracy": 0.625,
|
| 1745 |
+
"eval_loss": 2.7962350845336914,
|
| 1746 |
+
"eval_runtime": 6.6512,
|
| 1747 |
+
"eval_samples_per_second": 4.811,
|
| 1748 |
+
"eval_steps_per_second": 1.203,
|
| 1749 |
+
"step": 2100
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 30.002857142857142,
|
| 1753 |
+
"grad_norm": 0.004685261752456427,
|
| 1754 |
+
"learning_rate": 4.412698412698413e-06,
|
| 1755 |
+
"loss": 0.0001,
|
| 1756 |
+
"step": 2110
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 30.005714285714287,
|
| 1760 |
+
"grad_norm": 0.0024346013087779284,
|
| 1761 |
+
"learning_rate": 4.3809523809523815e-06,
|
| 1762 |
+
"loss": 0.0,
|
| 1763 |
+
"step": 2120
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 30.00857142857143,
|
| 1767 |
+
"grad_norm": 0.0026598083786666393,
|
| 1768 |
+
"learning_rate": 4.34920634920635e-06,
|
| 1769 |
+
"loss": 0.0,
|
| 1770 |
+
"step": 2130
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 30.01142857142857,
|
| 1774 |
+
"grad_norm": 0.002897858852520585,
|
| 1775 |
+
"learning_rate": 4.317460317460318e-06,
|
| 1776 |
+
"loss": 0.0001,
|
| 1777 |
+
"step": 2140
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 30.014285714285716,
|
| 1781 |
+
"grad_norm": 0.01607625186443329,
|
| 1782 |
+
"learning_rate": 4.2857142857142855e-06,
|
| 1783 |
+
"loss": 0.0001,
|
| 1784 |
+
"step": 2150
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 30.017142857142858,
|
| 1788 |
+
"grad_norm": 0.002586986403912306,
|
| 1789 |
+
"learning_rate": 4.2539682539682546e-06,
|
| 1790 |
+
"loss": 0.0,
|
| 1791 |
+
"step": 2160
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 30.02,
|
| 1795 |
+
"grad_norm": 0.0013718613190576434,
|
| 1796 |
+
"learning_rate": 4.222222222222223e-06,
|
| 1797 |
+
"loss": 0.0,
|
| 1798 |
+
"step": 2170
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 30.02,
|
| 1802 |
+
"eval_accuracy": 0.59375,
|
| 1803 |
+
"eval_loss": 3.108156204223633,
|
| 1804 |
+
"eval_runtime": 5.9931,
|
| 1805 |
+
"eval_samples_per_second": 5.34,
|
| 1806 |
+
"eval_steps_per_second": 1.335,
|
| 1807 |
+
"step": 2170
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 31.002857142857142,
|
| 1811 |
+
"grad_norm": 0.004130430519580841,
|
| 1812 |
+
"learning_rate": 4.190476190476191e-06,
|
| 1813 |
+
"loss": 0.0001,
|
| 1814 |
+
"step": 2180
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 31.005714285714287,
|
| 1818 |
+
"grad_norm": 0.0036993578542023897,
|
| 1819 |
+
"learning_rate": 4.158730158730159e-06,
|
| 1820 |
+
"loss": 0.0001,
|
| 1821 |
+
"step": 2190
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 31.00857142857143,
|
| 1825 |
+
"grad_norm": 0.0024625733494758606,
|
| 1826 |
+
"learning_rate": 4.126984126984127e-06,
|
| 1827 |
+
"loss": 0.0008,
|
| 1828 |
+
"step": 2200
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 31.01142857142857,
|
| 1832 |
+
"grad_norm": 0.007389547768980265,
|
| 1833 |
+
"learning_rate": 4.095238095238096e-06,
|
| 1834 |
+
"loss": 0.1028,
|
| 1835 |
+
"step": 2210
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 31.014285714285716,
|
| 1839 |
+
"grad_norm": 0.001320167793892324,
|
| 1840 |
+
"learning_rate": 4.063492063492064e-06,
|
| 1841 |
+
"loss": 0.0004,
|
| 1842 |
+
"step": 2220
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 31.017142857142858,
|
| 1846 |
+
"grad_norm": 466.27154541015625,
|
| 1847 |
+
"learning_rate": 4.031746031746032e-06,
|
| 1848 |
+
"loss": 0.069,
|
| 1849 |
+
"step": 2230
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 31.02,
|
| 1853 |
+
"grad_norm": 0.004771101288497448,
|
| 1854 |
+
"learning_rate": 4.000000000000001e-06,
|
| 1855 |
+
"loss": 0.0001,
|
| 1856 |
+
"step": 2240
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 31.02,
|
| 1860 |
+
"eval_accuracy": 0.625,
|
| 1861 |
+
"eval_loss": 2.686725616455078,
|
| 1862 |
+
"eval_runtime": 7.1944,
|
| 1863 |
+
"eval_samples_per_second": 4.448,
|
| 1864 |
+
"eval_steps_per_second": 1.112,
|
| 1865 |
+
"step": 2240
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 32.002857142857145,
|
| 1869 |
+
"grad_norm": 0.004328535404056311,
|
| 1870 |
+
"learning_rate": 3.968253968253968e-06,
|
| 1871 |
+
"loss": 0.0001,
|
| 1872 |
+
"step": 2250
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 32.005714285714284,
|
| 1876 |
+
"grad_norm": 0.007427292410284281,
|
| 1877 |
+
"learning_rate": 3.936507936507936e-06,
|
| 1878 |
+
"loss": 0.0001,
|
| 1879 |
+
"step": 2260
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 32.00857142857143,
|
| 1883 |
+
"grad_norm": 0.001498078927397728,
|
| 1884 |
+
"learning_rate": 3.9047619047619055e-06,
|
| 1885 |
+
"loss": 0.0024,
|
| 1886 |
+
"step": 2270
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 32.011428571428574,
|
| 1890 |
+
"grad_norm": 0.0010804787743836641,
|
| 1891 |
+
"learning_rate": 3.873015873015874e-06,
|
| 1892 |
+
"loss": 0.0,
|
| 1893 |
+
"step": 2280
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 32.01428571428571,
|
| 1897 |
+
"grad_norm": 0.07372881472110748,
|
| 1898 |
+
"learning_rate": 3.841269841269842e-06,
|
| 1899 |
+
"loss": 0.0284,
|
| 1900 |
+
"step": 2290
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 32.01714285714286,
|
| 1904 |
+
"grad_norm": 0.002815135521814227,
|
| 1905 |
+
"learning_rate": 3.80952380952381e-06,
|
| 1906 |
+
"loss": 0.0,
|
| 1907 |
+
"step": 2300
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 32.02,
|
| 1911 |
+
"grad_norm": 0.0011799400672316551,
|
| 1912 |
+
"learning_rate": 3.777777777777778e-06,
|
| 1913 |
+
"loss": 0.0,
|
| 1914 |
+
"step": 2310
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 32.02,
|
| 1918 |
+
"eval_accuracy": 0.625,
|
| 1919 |
+
"eval_loss": 2.6702141761779785,
|
| 1920 |
+
"eval_runtime": 7.3014,
|
| 1921 |
+
"eval_samples_per_second": 4.383,
|
| 1922 |
+
"eval_steps_per_second": 1.096,
|
| 1923 |
+
"step": 2310
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 33.002857142857145,
|
| 1927 |
+
"grad_norm": 0.004017932340502739,
|
| 1928 |
+
"learning_rate": 3.7460317460317463e-06,
|
| 1929 |
+
"loss": 0.0046,
|
| 1930 |
+
"step": 2320
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 33.005714285714284,
|
| 1934 |
+
"grad_norm": 0.002392595401033759,
|
| 1935 |
+
"learning_rate": 3.7142857142857146e-06,
|
| 1936 |
+
"loss": 0.0,
|
| 1937 |
+
"step": 2330
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 33.00857142857143,
|
| 1941 |
+
"grad_norm": 0.0008441254030913115,
|
| 1942 |
+
"learning_rate": 3.6825396825396833e-06,
|
| 1943 |
+
"loss": 0.0,
|
| 1944 |
+
"step": 2340
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 33.011428571428574,
|
| 1948 |
+
"grad_norm": 0.01818065159022808,
|
| 1949 |
+
"learning_rate": 3.6507936507936507e-06,
|
| 1950 |
+
"loss": 0.0,
|
| 1951 |
+
"step": 2350
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 33.01428571428571,
|
| 1955 |
+
"grad_norm": 0.0024400062393397093,
|
| 1956 |
+
"learning_rate": 3.6190476190476194e-06,
|
| 1957 |
+
"loss": 0.0002,
|
| 1958 |
+
"step": 2360
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 33.01714285714286,
|
| 1962 |
+
"grad_norm": 0.001259080832824111,
|
| 1963 |
+
"learning_rate": 3.5873015873015877e-06,
|
| 1964 |
+
"loss": 0.0244,
|
| 1965 |
+
"step": 2370
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 33.02,
|
| 1969 |
+
"grad_norm": 0.0022309625055640936,
|
| 1970 |
+
"learning_rate": 3.555555555555556e-06,
|
| 1971 |
+
"loss": 0.0,
|
| 1972 |
+
"step": 2380
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 33.02,
|
| 1976 |
+
"eval_accuracy": 0.59375,
|
| 1977 |
+
"eval_loss": 3.3402888774871826,
|
| 1978 |
+
"eval_runtime": 7.3496,
|
| 1979 |
+
"eval_samples_per_second": 4.354,
|
| 1980 |
+
"eval_steps_per_second": 1.088,
|
| 1981 |
+
"step": 2380
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 34.002857142857145,
|
| 1985 |
+
"grad_norm": 0.001507407519966364,
|
| 1986 |
+
"learning_rate": 3.523809523809524e-06,
|
| 1987 |
+
"loss": 0.1484,
|
| 1988 |
+
"step": 2390
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 34.005714285714284,
|
| 1992 |
+
"grad_norm": 0.001335164182819426,
|
| 1993 |
+
"learning_rate": 3.492063492063492e-06,
|
| 1994 |
+
"loss": 0.0,
|
| 1995 |
+
"step": 2400
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 34.00857142857143,
|
| 1999 |
+
"grad_norm": 0.0010363466572016478,
|
| 2000 |
+
"learning_rate": 3.4603174603174607e-06,
|
| 2001 |
+
"loss": 0.0,
|
| 2002 |
+
"step": 2410
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 34.011428571428574,
|
| 2006 |
+
"grad_norm": 0.0027341984678059816,
|
| 2007 |
+
"learning_rate": 3.428571428571429e-06,
|
| 2008 |
+
"loss": 0.0,
|
| 2009 |
+
"step": 2420
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 34.01428571428571,
|
| 2013 |
+
"grad_norm": 0.007578617427498102,
|
| 2014 |
+
"learning_rate": 3.3968253968253972e-06,
|
| 2015 |
+
"loss": 0.0001,
|
| 2016 |
+
"step": 2430
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 34.01714285714286,
|
| 2020 |
+
"grad_norm": 0.001968264114111662,
|
| 2021 |
+
"learning_rate": 3.3650793650793655e-06,
|
| 2022 |
+
"loss": 0.0,
|
| 2023 |
+
"step": 2440
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 34.02,
|
| 2027 |
+
"grad_norm": 0.0030165906064212322,
|
| 2028 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 2029 |
+
"loss": 0.0,
|
| 2030 |
+
"step": 2450
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 34.02,
|
| 2034 |
+
"eval_accuracy": 0.625,
|
| 2035 |
+
"eval_loss": 3.053858757019043,
|
| 2036 |
+
"eval_runtime": 8.5887,
|
| 2037 |
+
"eval_samples_per_second": 3.726,
|
| 2038 |
+
"eval_steps_per_second": 0.931,
|
| 2039 |
+
"step": 2450
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 35.002857142857145,
|
| 2043 |
+
"grad_norm": 0.0015692274319007993,
|
| 2044 |
+
"learning_rate": 3.3015873015873016e-06,
|
| 2045 |
+
"loss": 0.0,
|
| 2046 |
+
"step": 2460
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 35.005714285714284,
|
| 2050 |
+
"grad_norm": 0.0013480527559295297,
|
| 2051 |
+
"learning_rate": 3.2698412698412703e-06,
|
| 2052 |
+
"loss": 0.0,
|
| 2053 |
+
"step": 2470
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 35.00857142857143,
|
| 2057 |
+
"grad_norm": 0.003836671821773052,
|
| 2058 |
+
"learning_rate": 3.2380952380952385e-06,
|
| 2059 |
+
"loss": 0.1318,
|
| 2060 |
+
"step": 2480
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 35.011428571428574,
|
| 2064 |
+
"grad_norm": 0.0024338432122021914,
|
| 2065 |
+
"learning_rate": 3.206349206349207e-06,
|
| 2066 |
+
"loss": 0.0,
|
| 2067 |
+
"step": 2490
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 35.01428571428571,
|
| 2071 |
+
"grad_norm": 0.0123060941696167,
|
| 2072 |
+
"learning_rate": 3.1746031746031746e-06,
|
| 2073 |
+
"loss": 0.0001,
|
| 2074 |
+
"step": 2500
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 35.01714285714286,
|
| 2078 |
+
"grad_norm": 0.001270064152777195,
|
| 2079 |
+
"learning_rate": 3.142857142857143e-06,
|
| 2080 |
+
"loss": 0.0,
|
| 2081 |
+
"step": 2510
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 35.02,
|
| 2085 |
+
"grad_norm": 0.0015140828909352422,
|
| 2086 |
+
"learning_rate": 3.1111111111111116e-06,
|
| 2087 |
+
"loss": 0.2031,
|
| 2088 |
+
"step": 2520
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 35.02,
|
| 2092 |
+
"eval_accuracy": 0.6875,
|
| 2093 |
+
"eval_loss": 2.73325252532959,
|
| 2094 |
+
"eval_runtime": 8.3821,
|
| 2095 |
+
"eval_samples_per_second": 3.818,
|
| 2096 |
+
"eval_steps_per_second": 0.954,
|
| 2097 |
+
"step": 2520
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 36.002857142857145,
|
| 2101 |
+
"grad_norm": 0.0005715539446100593,
|
| 2102 |
+
"learning_rate": 3.07936507936508e-06,
|
| 2103 |
+
"loss": 0.0,
|
| 2104 |
+
"step": 2530
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 36.005714285714284,
|
| 2108 |
+
"grad_norm": 0.003236456075683236,
|
| 2109 |
+
"learning_rate": 3.047619047619048e-06,
|
| 2110 |
+
"loss": 0.0,
|
| 2111 |
+
"step": 2540
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 36.00857142857143,
|
| 2115 |
+
"grad_norm": 0.0009249376598745584,
|
| 2116 |
+
"learning_rate": 3.015873015873016e-06,
|
| 2117 |
+
"loss": 0.2215,
|
| 2118 |
+
"step": 2550
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 36.011428571428574,
|
| 2122 |
+
"grad_norm": 0.0008299104520119727,
|
| 2123 |
+
"learning_rate": 2.984126984126984e-06,
|
| 2124 |
+
"loss": 0.031,
|
| 2125 |
+
"step": 2560
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 36.01428571428571,
|
| 2129 |
+
"grad_norm": 0.0019169561564922333,
|
| 2130 |
+
"learning_rate": 2.9523809523809525e-06,
|
| 2131 |
+
"loss": 0.2959,
|
| 2132 |
+
"step": 2570
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 36.01714285714286,
|
| 2136 |
+
"grad_norm": 0.016660558059811592,
|
| 2137 |
+
"learning_rate": 2.920634920634921e-06,
|
| 2138 |
+
"loss": 0.0,
|
| 2139 |
+
"step": 2580
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 36.02,
|
| 2143 |
+
"grad_norm": 0.0024716516491025686,
|
| 2144 |
+
"learning_rate": 2.888888888888889e-06,
|
| 2145 |
+
"loss": 0.0,
|
| 2146 |
+
"step": 2590
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 36.02,
|
| 2150 |
+
"eval_accuracy": 0.59375,
|
| 2151 |
+
"eval_loss": 2.9419589042663574,
|
| 2152 |
+
"eval_runtime": 7.6689,
|
| 2153 |
+
"eval_samples_per_second": 4.173,
|
| 2154 |
+
"eval_steps_per_second": 1.043,
|
| 2155 |
+
"step": 2590
|
| 2156 |
+
},
|
| 2157 |
+
{
|
| 2158 |
+
"epoch": 37.002857142857145,
|
| 2159 |
+
"grad_norm": 0.0009388060425408185,
|
| 2160 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 2161 |
+
"loss": 0.0,
|
| 2162 |
+
"step": 2600
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 37.005714285714284,
|
| 2166 |
+
"grad_norm": 0.0014066151343286037,
|
| 2167 |
+
"learning_rate": 2.8253968253968255e-06,
|
| 2168 |
+
"loss": 0.0049,
|
| 2169 |
+
"step": 2610
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 37.00857142857143,
|
| 2173 |
+
"grad_norm": 0.0009500043233856559,
|
| 2174 |
+
"learning_rate": 2.7936507936507938e-06,
|
| 2175 |
+
"loss": 0.0,
|
| 2176 |
+
"step": 2620
|
| 2177 |
+
},
|
| 2178 |
+
{
|
| 2179 |
+
"epoch": 37.011428571428574,
|
| 2180 |
+
"grad_norm": 0.0007660359842702746,
|
| 2181 |
+
"learning_rate": 2.7619047619047625e-06,
|
| 2182 |
+
"loss": 0.0002,
|
| 2183 |
+
"step": 2630
|
| 2184 |
+
},
|
| 2185 |
+
{
|
| 2186 |
+
"epoch": 37.01428571428571,
|
| 2187 |
+
"grad_norm": 0.005225882399827242,
|
| 2188 |
+
"learning_rate": 2.7301587301587303e-06,
|
| 2189 |
+
"loss": 0.0,
|
| 2190 |
+
"step": 2640
|
| 2191 |
+
},
|
| 2192 |
+
{
|
| 2193 |
+
"epoch": 37.01714285714286,
|
| 2194 |
+
"grad_norm": 0.0036217437591403723,
|
| 2195 |
+
"learning_rate": 2.6984126984126986e-06,
|
| 2196 |
+
"loss": 0.0001,
|
| 2197 |
+
"step": 2650
|
| 2198 |
+
},
|
| 2199 |
+
{
|
| 2200 |
+
"epoch": 37.02,
|
| 2201 |
+
"grad_norm": 0.0021994500420987606,
|
| 2202 |
+
"learning_rate": 2.666666666666667e-06,
|
| 2203 |
+
"loss": 0.0,
|
| 2204 |
+
"step": 2660
|
| 2205 |
+
},
|
| 2206 |
+
{
|
| 2207 |
+
"epoch": 37.02,
|
| 2208 |
+
"eval_accuracy": 0.65625,
|
| 2209 |
+
"eval_loss": 2.919912338256836,
|
| 2210 |
+
"eval_runtime": 7.4701,
|
| 2211 |
+
"eval_samples_per_second": 4.284,
|
| 2212 |
+
"eval_steps_per_second": 1.071,
|
| 2213 |
+
"step": 2660
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 38.002857142857145,
|
| 2217 |
+
"grad_norm": 0.0009795749792829156,
|
| 2218 |
+
"learning_rate": 2.634920634920635e-06,
|
| 2219 |
+
"loss": 0.0,
|
| 2220 |
+
"step": 2670
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 38.005714285714284,
|
| 2224 |
+
"grad_norm": 0.0015403656288981438,
|
| 2225 |
+
"learning_rate": 2.6031746031746038e-06,
|
| 2226 |
+
"loss": 0.0007,
|
| 2227 |
+
"step": 2680
|
| 2228 |
+
},
|
| 2229 |
+
{
|
| 2230 |
+
"epoch": 38.00857142857143,
|
| 2231 |
+
"grad_norm": 0.007823006249964237,
|
| 2232 |
+
"learning_rate": 2.571428571428571e-06,
|
| 2233 |
+
"loss": 0.0701,
|
| 2234 |
+
"step": 2690
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 38.011428571428574,
|
| 2238 |
+
"grad_norm": 0.12522433698177338,
|
| 2239 |
+
"learning_rate": 2.53968253968254e-06,
|
| 2240 |
+
"loss": 0.0643,
|
| 2241 |
+
"step": 2700
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 38.01428571428571,
|
| 2245 |
+
"grad_norm": 0.0010730504291132092,
|
| 2246 |
+
"learning_rate": 2.507936507936508e-06,
|
| 2247 |
+
"loss": 0.0001,
|
| 2248 |
+
"step": 2710
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 38.01714285714286,
|
| 2252 |
+
"grad_norm": 0.001265152939595282,
|
| 2253 |
+
"learning_rate": 2.4761904761904764e-06,
|
| 2254 |
+
"loss": 0.2465,
|
| 2255 |
+
"step": 2720
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 38.02,
|
| 2259 |
+
"grad_norm": 0.45712485909461975,
|
| 2260 |
+
"learning_rate": 2.4444444444444447e-06,
|
| 2261 |
+
"loss": 0.0018,
|
| 2262 |
+
"step": 2730
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 38.02,
|
| 2266 |
+
"eval_accuracy": 0.59375,
|
| 2267 |
+
"eval_loss": 2.8933959007263184,
|
| 2268 |
+
"eval_runtime": 7.7323,
|
| 2269 |
+
"eval_samples_per_second": 4.138,
|
| 2270 |
+
"eval_steps_per_second": 1.035,
|
| 2271 |
+
"step": 2730
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 39.002857142857145,
|
| 2275 |
+
"grad_norm": 0.0031093251891434193,
|
| 2276 |
+
"learning_rate": 2.412698412698413e-06,
|
| 2277 |
+
"loss": 0.0042,
|
| 2278 |
+
"step": 2740
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 39.005714285714284,
|
| 2282 |
+
"grad_norm": 0.004323534667491913,
|
| 2283 |
+
"learning_rate": 2.380952380952381e-06,
|
| 2284 |
+
"loss": 0.0,
|
| 2285 |
+
"step": 2750
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 39.00857142857143,
|
| 2289 |
+
"grad_norm": 0.0008992227958515286,
|
| 2290 |
+
"learning_rate": 2.3492063492063494e-06,
|
| 2291 |
+
"loss": 0.0,
|
| 2292 |
+
"step": 2760
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 39.011428571428574,
|
| 2296 |
+
"grad_norm": 0.0029326954390853643,
|
| 2297 |
+
"learning_rate": 2.3174603174603177e-06,
|
| 2298 |
+
"loss": 0.0,
|
| 2299 |
+
"step": 2770
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 39.01428571428571,
|
| 2303 |
+
"grad_norm": 0.0016185936983674765,
|
| 2304 |
+
"learning_rate": 2.285714285714286e-06,
|
| 2305 |
+
"loss": 0.0,
|
| 2306 |
+
"step": 2780
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 39.01714285714286,
|
| 2310 |
+
"grad_norm": 0.005735237151384354,
|
| 2311 |
+
"learning_rate": 2.2539682539682542e-06,
|
| 2312 |
+
"loss": 0.0,
|
| 2313 |
+
"step": 2790
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 39.02,
|
| 2317 |
+
"grad_norm": 0.0014396852347999811,
|
| 2318 |
+
"learning_rate": 2.222222222222222e-06,
|
| 2319 |
+
"loss": 0.0,
|
| 2320 |
+
"step": 2800
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 39.02,
|
| 2324 |
+
"eval_accuracy": 0.625,
|
| 2325 |
+
"eval_loss": 2.832045078277588,
|
| 2326 |
+
"eval_runtime": 7.5616,
|
| 2327 |
+
"eval_samples_per_second": 4.232,
|
| 2328 |
+
"eval_steps_per_second": 1.058,
|
| 2329 |
+
"step": 2800
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 40.002857142857145,
|
| 2333 |
+
"grad_norm": 0.0041121807880699635,
|
| 2334 |
+
"learning_rate": 2.1904761904761908e-06,
|
| 2335 |
+
"loss": 0.0,
|
| 2336 |
+
"step": 2810
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 40.005714285714284,
|
| 2340 |
+
"grad_norm": 0.0013577585341408849,
|
| 2341 |
+
"learning_rate": 2.158730158730159e-06,
|
| 2342 |
+
"loss": 0.0,
|
| 2343 |
+
"step": 2820
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 40.00857142857143,
|
| 2347 |
+
"grad_norm": 0.0017459145747125149,
|
| 2348 |
+
"learning_rate": 2.1269841269841273e-06,
|
| 2349 |
+
"loss": 0.0,
|
| 2350 |
+
"step": 2830
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 40.011428571428574,
|
| 2354 |
+
"grad_norm": 0.0019927776884287596,
|
| 2355 |
+
"learning_rate": 2.0952380952380955e-06,
|
| 2356 |
+
"loss": 0.0,
|
| 2357 |
+
"step": 2840
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 40.01428571428571,
|
| 2361 |
+
"grad_norm": 0.0009335738723166287,
|
| 2362 |
+
"learning_rate": 2.0634920634920634e-06,
|
| 2363 |
+
"loss": 0.0,
|
| 2364 |
+
"step": 2850
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 40.01714285714286,
|
| 2368 |
+
"grad_norm": 0.0010714407544583082,
|
| 2369 |
+
"learning_rate": 2.031746031746032e-06,
|
| 2370 |
+
"loss": 0.0,
|
| 2371 |
+
"step": 2860
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 40.02,
|
| 2375 |
+
"grad_norm": 0.0007551051676273346,
|
| 2376 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 2377 |
+
"loss": 0.0,
|
| 2378 |
+
"step": 2870
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 40.02,
|
| 2382 |
+
"eval_accuracy": 0.625,
|
| 2383 |
+
"eval_loss": 2.8381476402282715,
|
| 2384 |
+
"eval_runtime": 7.9752,
|
| 2385 |
+
"eval_samples_per_second": 4.012,
|
| 2386 |
+
"eval_steps_per_second": 1.003,
|
| 2387 |
+
"step": 2870
|
| 2388 |
+
},
|
| 2389 |
+
{
|
| 2390 |
+
"epoch": 41.002857142857145,
|
| 2391 |
+
"grad_norm": 0.001309627783484757,
|
| 2392 |
+
"learning_rate": 1.968253968253968e-06,
|
| 2393 |
+
"loss": 0.0001,
|
| 2394 |
+
"step": 2880
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 41.005714285714284,
|
| 2398 |
+
"grad_norm": 0.0011593608651310205,
|
| 2399 |
+
"learning_rate": 1.936507936507937e-06,
|
| 2400 |
+
"loss": 0.0,
|
| 2401 |
+
"step": 2890
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 41.00857142857143,
|
| 2405 |
+
"grad_norm": 0.0005718428292311728,
|
| 2406 |
+
"learning_rate": 1.904761904761905e-06,
|
| 2407 |
+
"loss": 0.0,
|
| 2408 |
+
"step": 2900
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 41.011428571428574,
|
| 2412 |
+
"grad_norm": 0.0008164609898813069,
|
| 2413 |
+
"learning_rate": 1.8730158730158732e-06,
|
| 2414 |
+
"loss": 0.0,
|
| 2415 |
+
"step": 2910
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 41.01428571428571,
|
| 2419 |
+
"grad_norm": 0.0068624867126345634,
|
| 2420 |
+
"learning_rate": 1.8412698412698416e-06,
|
| 2421 |
+
"loss": 0.0,
|
| 2422 |
+
"step": 2920
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 41.01714285714286,
|
| 2426 |
+
"grad_norm": 0.002447180449962616,
|
| 2427 |
+
"learning_rate": 1.8095238095238097e-06,
|
| 2428 |
+
"loss": 0.0,
|
| 2429 |
+
"step": 2930
|
| 2430 |
+
},
|
| 2431 |
+
{
|
| 2432 |
+
"epoch": 41.02,
|
| 2433 |
+
"grad_norm": 0.001345404889434576,
|
| 2434 |
+
"learning_rate": 1.777777777777778e-06,
|
| 2435 |
+
"loss": 0.0,
|
| 2436 |
+
"step": 2940
|
| 2437 |
+
},
|
| 2438 |
+
{
|
| 2439 |
+
"epoch": 41.02,
|
| 2440 |
+
"eval_accuracy": 0.625,
|
| 2441 |
+
"eval_loss": 2.8438665866851807,
|
| 2442 |
+
"eval_runtime": 6.6318,
|
| 2443 |
+
"eval_samples_per_second": 4.825,
|
| 2444 |
+
"eval_steps_per_second": 1.206,
|
| 2445 |
+
"step": 2940
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 42.002857142857145,
|
| 2449 |
+
"grad_norm": 0.004871055483818054,
|
| 2450 |
+
"learning_rate": 1.746031746031746e-06,
|
| 2451 |
+
"loss": 0.0,
|
| 2452 |
+
"step": 2950
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 42.005714285714284,
|
| 2456 |
+
"grad_norm": 0.0018843573052436113,
|
| 2457 |
+
"learning_rate": 1.7142857142857145e-06,
|
| 2458 |
+
"loss": 0.0,
|
| 2459 |
+
"step": 2960
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 42.00857142857143,
|
| 2463 |
+
"grad_norm": 0.0010428063105791807,
|
| 2464 |
+
"learning_rate": 1.6825396825396827e-06,
|
| 2465 |
+
"loss": 0.0001,
|
| 2466 |
+
"step": 2970
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 42.011428571428574,
|
| 2470 |
+
"grad_norm": 0.0007727840566076338,
|
| 2471 |
+
"learning_rate": 1.6507936507936508e-06,
|
| 2472 |
+
"loss": 0.0,
|
| 2473 |
+
"step": 2980
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 42.01428571428571,
|
| 2477 |
+
"grad_norm": 0.0022347676567733288,
|
| 2478 |
+
"learning_rate": 1.6190476190476193e-06,
|
| 2479 |
+
"loss": 0.0,
|
| 2480 |
+
"step": 2990
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 42.01714285714286,
|
| 2484 |
+
"grad_norm": 0.0014822926605120301,
|
| 2485 |
+
"learning_rate": 1.5873015873015873e-06,
|
| 2486 |
+
"loss": 0.0,
|
| 2487 |
+
"step": 3000
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 42.02,
|
| 2491 |
+
"grad_norm": 0.0008702924242243171,
|
| 2492 |
+
"learning_rate": 1.5555555555555558e-06,
|
| 2493 |
+
"loss": 0.0,
|
| 2494 |
+
"step": 3010
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 42.02,
|
| 2498 |
+
"eval_accuracy": 0.625,
|
| 2499 |
+
"eval_loss": 2.8524856567382812,
|
| 2500 |
+
"eval_runtime": 6.573,
|
| 2501 |
+
"eval_samples_per_second": 4.868,
|
| 2502 |
+
"eval_steps_per_second": 1.217,
|
| 2503 |
+
"step": 3010
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 43.002857142857145,
|
| 2507 |
+
"grad_norm": 0.001486993976868689,
|
| 2508 |
+
"learning_rate": 1.523809523809524e-06,
|
| 2509 |
+
"loss": 0.0,
|
| 2510 |
+
"step": 3020
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 43.005714285714284,
|
| 2514 |
+
"grad_norm": 0.0009380248957313597,
|
| 2515 |
+
"learning_rate": 1.492063492063492e-06,
|
| 2516 |
+
"loss": 0.0,
|
| 2517 |
+
"step": 3030
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 43.00857142857143,
|
| 2521 |
+
"grad_norm": 0.004946484230458736,
|
| 2522 |
+
"learning_rate": 1.4603174603174606e-06,
|
| 2523 |
+
"loss": 0.0,
|
| 2524 |
+
"step": 3040
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 43.011428571428574,
|
| 2528 |
+
"grad_norm": 0.0016762095037847757,
|
| 2529 |
+
"learning_rate": 1.4285714285714286e-06,
|
| 2530 |
+
"loss": 0.0002,
|
| 2531 |
+
"step": 3050
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 43.01428571428571,
|
| 2535 |
+
"grad_norm": 0.001995038939639926,
|
| 2536 |
+
"learning_rate": 1.3968253968253969e-06,
|
| 2537 |
+
"loss": 0.0,
|
| 2538 |
+
"step": 3060
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 43.01714285714286,
|
| 2542 |
+
"grad_norm": 0.0014240635791793466,
|
| 2543 |
+
"learning_rate": 1.3650793650793652e-06,
|
| 2544 |
+
"loss": 0.0001,
|
| 2545 |
+
"step": 3070
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 43.02,
|
| 2549 |
+
"grad_norm": 0.0009568611858412623,
|
| 2550 |
+
"learning_rate": 1.3333333333333334e-06,
|
| 2551 |
+
"loss": 0.0,
|
| 2552 |
+
"step": 3080
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 43.02,
|
| 2556 |
+
"eval_accuracy": 0.59375,
|
| 2557 |
+
"eval_loss": 2.6921167373657227,
|
| 2558 |
+
"eval_runtime": 7.5296,
|
| 2559 |
+
"eval_samples_per_second": 4.25,
|
| 2560 |
+
"eval_steps_per_second": 1.062,
|
| 2561 |
+
"step": 3080
|
| 2562 |
+
},
|
| 2563 |
+
{
|
| 2564 |
+
"epoch": 44.002857142857145,
|
| 2565 |
+
"grad_norm": 0.01802194118499756,
|
| 2566 |
+
"learning_rate": 1.3015873015873019e-06,
|
| 2567 |
+
"loss": 0.0,
|
| 2568 |
+
"step": 3090
|
| 2569 |
+
},
|
| 2570 |
+
{
|
| 2571 |
+
"epoch": 44.005714285714284,
|
| 2572 |
+
"grad_norm": 0.000703324971254915,
|
| 2573 |
+
"learning_rate": 1.26984126984127e-06,
|
| 2574 |
+
"loss": 0.0,
|
| 2575 |
+
"step": 3100
|
| 2576 |
+
},
|
| 2577 |
+
{
|
| 2578 |
+
"epoch": 44.00857142857143,
|
| 2579 |
+
"grad_norm": 0.0015024804743006825,
|
| 2580 |
+
"learning_rate": 1.2380952380952382e-06,
|
| 2581 |
+
"loss": 0.0,
|
| 2582 |
+
"step": 3110
|
| 2583 |
+
},
|
| 2584 |
+
{
|
| 2585 |
+
"epoch": 44.011428571428574,
|
| 2586 |
+
"grad_norm": 0.0023550696205347776,
|
| 2587 |
+
"learning_rate": 1.2063492063492065e-06,
|
| 2588 |
+
"loss": 0.0,
|
| 2589 |
+
"step": 3120
|
| 2590 |
+
},
|
| 2591 |
+
{
|
| 2592 |
+
"epoch": 44.01428571428571,
|
| 2593 |
+
"grad_norm": 0.0009432346560060978,
|
| 2594 |
+
"learning_rate": 1.1746031746031747e-06,
|
| 2595 |
+
"loss": 0.0001,
|
| 2596 |
+
"step": 3130
|
| 2597 |
+
},
|
| 2598 |
+
{
|
| 2599 |
+
"epoch": 44.01714285714286,
|
| 2600 |
+
"grad_norm": 0.0014145496534183621,
|
| 2601 |
+
"learning_rate": 1.142857142857143e-06,
|
| 2602 |
+
"loss": 0.0,
|
| 2603 |
+
"step": 3140
|
| 2604 |
+
},
|
| 2605 |
+
{
|
| 2606 |
+
"epoch": 44.02,
|
| 2607 |
+
"grad_norm": 0.0011549839982762933,
|
| 2608 |
+
"learning_rate": 1.111111111111111e-06,
|
| 2609 |
+
"loss": 0.0,
|
| 2610 |
+
"step": 3150
|
| 2611 |
+
},
|
| 2612 |
+
{
|
| 2613 |
+
"epoch": 44.02,
|
| 2614 |
+
"eval_accuracy": 0.625,
|
| 2615 |
+
"eval_loss": 2.9579904079437256,
|
| 2616 |
+
"eval_runtime": 7.1267,
|
| 2617 |
+
"eval_samples_per_second": 4.49,
|
| 2618 |
+
"eval_steps_per_second": 1.123,
|
| 2619 |
+
"step": 3150
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 45.002857142857145,
|
| 2623 |
+
"grad_norm": 0.004226373508572578,
|
| 2624 |
+
"learning_rate": 1.0793650793650795e-06,
|
| 2625 |
+
"loss": 0.0003,
|
| 2626 |
+
"step": 3160
|
| 2627 |
+
},
|
| 2628 |
+
{
|
| 2629 |
+
"epoch": 45.005714285714284,
|
| 2630 |
+
"grad_norm": 0.0005224282504059374,
|
| 2631 |
+
"learning_rate": 1.0476190476190478e-06,
|
| 2632 |
+
"loss": 0.0,
|
| 2633 |
+
"step": 3170
|
| 2634 |
+
},
|
| 2635 |
+
{
|
| 2636 |
+
"epoch": 45.00857142857143,
|
| 2637 |
+
"grad_norm": 0.0016394226113334298,
|
| 2638 |
+
"learning_rate": 1.015873015873016e-06,
|
| 2639 |
+
"loss": 0.0,
|
| 2640 |
+
"step": 3180
|
| 2641 |
+
},
|
| 2642 |
+
{
|
| 2643 |
+
"epoch": 45.011428571428574,
|
| 2644 |
+
"grad_norm": 0.004058617167174816,
|
| 2645 |
+
"learning_rate": 9.84126984126984e-07,
|
| 2646 |
+
"loss": 0.0,
|
| 2647 |
+
"step": 3190
|
| 2648 |
+
},
|
| 2649 |
+
{
|
| 2650 |
+
"epoch": 45.01428571428571,
|
| 2651 |
+
"grad_norm": 0.0013102118391543627,
|
| 2652 |
+
"learning_rate": 9.523809523809525e-07,
|
| 2653 |
+
"loss": 0.0,
|
| 2654 |
+
"step": 3200
|
| 2655 |
+
},
|
| 2656 |
+
{
|
| 2657 |
+
"epoch": 45.01714285714286,
|
| 2658 |
+
"grad_norm": 0.0019316988764330745,
|
| 2659 |
+
"learning_rate": 9.206349206349208e-07,
|
| 2660 |
+
"loss": 0.0,
|
| 2661 |
+
"step": 3210
|
| 2662 |
+
},
|
| 2663 |
+
{
|
| 2664 |
+
"epoch": 45.02,
|
| 2665 |
+
"grad_norm": 0.000915765471290797,
|
| 2666 |
+
"learning_rate": 8.88888888888889e-07,
|
| 2667 |
+
"loss": 0.0,
|
| 2668 |
+
"step": 3220
|
| 2669 |
+
},
|
| 2670 |
+
{
|
| 2671 |
+
"epoch": 45.02,
|
| 2672 |
+
"eval_accuracy": 0.59375,
|
| 2673 |
+
"eval_loss": 3.029210090637207,
|
| 2674 |
+
"eval_runtime": 6.7888,
|
| 2675 |
+
"eval_samples_per_second": 4.714,
|
| 2676 |
+
"eval_steps_per_second": 1.178,
|
| 2677 |
+
"step": 3220
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 46.002857142857145,
|
| 2681 |
+
"grad_norm": 0.001245733117684722,
|
| 2682 |
+
"learning_rate": 8.571428571428572e-07,
|
| 2683 |
+
"loss": 0.068,
|
| 2684 |
+
"step": 3230
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 46.005714285714284,
|
| 2688 |
+
"grad_norm": 0.0010560519294813275,
|
| 2689 |
+
"learning_rate": 8.253968253968254e-07,
|
| 2690 |
+
"loss": 0.0,
|
| 2691 |
+
"step": 3240
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 46.00857142857143,
|
| 2695 |
+
"grad_norm": 0.0011368553387001157,
|
| 2696 |
+
"learning_rate": 7.936507936507937e-07,
|
| 2697 |
+
"loss": 0.0,
|
| 2698 |
+
"step": 3250
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 46.011428571428574,
|
| 2702 |
+
"grad_norm": 0.00201380904763937,
|
| 2703 |
+
"learning_rate": 7.61904761904762e-07,
|
| 2704 |
+
"loss": 0.0,
|
| 2705 |
+
"step": 3260
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 46.01428571428571,
|
| 2709 |
+
"grad_norm": 0.0008550912025384605,
|
| 2710 |
+
"learning_rate": 7.301587301587303e-07,
|
| 2711 |
+
"loss": 0.0,
|
| 2712 |
+
"step": 3270
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 46.01714285714286,
|
| 2716 |
+
"grad_norm": 0.0008424161933362484,
|
| 2717 |
+
"learning_rate": 6.984126984126984e-07,
|
| 2718 |
+
"loss": 0.0,
|
| 2719 |
+
"step": 3280
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 46.02,
|
| 2723 |
+
"grad_norm": 0.0006465838523581624,
|
| 2724 |
+
"learning_rate": 6.666666666666667e-07,
|
| 2725 |
+
"loss": 0.0,
|
| 2726 |
+
"step": 3290
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 46.02,
|
| 2730 |
+
"eval_accuracy": 0.625,
|
| 2731 |
+
"eval_loss": 3.1119141578674316,
|
| 2732 |
+
"eval_runtime": 7.4394,
|
| 2733 |
+
"eval_samples_per_second": 4.301,
|
| 2734 |
+
"eval_steps_per_second": 1.075,
|
| 2735 |
+
"step": 3290
|
| 2736 |
+
},
|
| 2737 |
+
{
|
| 2738 |
+
"epoch": 47.002857142857145,
|
| 2739 |
+
"grad_norm": 0.0010397338774055243,
|
| 2740 |
+
"learning_rate": 6.34920634920635e-07,
|
| 2741 |
+
"loss": 0.0,
|
| 2742 |
+
"step": 3300
|
| 2743 |
+
},
|
| 2744 |
+
{
|
| 2745 |
+
"epoch": 47.005714285714284,
|
| 2746 |
+
"grad_norm": 0.0018721016822382808,
|
| 2747 |
+
"learning_rate": 6.031746031746032e-07,
|
| 2748 |
+
"loss": 0.0,
|
| 2749 |
+
"step": 3310
|
| 2750 |
+
},
|
| 2751 |
+
{
|
| 2752 |
+
"epoch": 47.00857142857143,
|
| 2753 |
+
"grad_norm": 0.0009640470379963517,
|
| 2754 |
+
"learning_rate": 5.714285714285715e-07,
|
| 2755 |
+
"loss": 0.0,
|
| 2756 |
+
"step": 3320
|
| 2757 |
+
},
|
| 2758 |
+
{
|
| 2759 |
+
"epoch": 47.011428571428574,
|
| 2760 |
+
"grad_norm": 0.0024156668223440647,
|
| 2761 |
+
"learning_rate": 5.396825396825398e-07,
|
| 2762 |
+
"loss": 0.0,
|
| 2763 |
+
"step": 3330
|
| 2764 |
+
},
|
| 2765 |
+
{
|
| 2766 |
+
"epoch": 47.01428571428571,
|
| 2767 |
+
"grad_norm": 0.0012575440341606736,
|
| 2768 |
+
"learning_rate": 5.07936507936508e-07,
|
| 2769 |
+
"loss": 0.0,
|
| 2770 |
+
"step": 3340
|
| 2771 |
+
},
|
| 2772 |
+
{
|
| 2773 |
+
"epoch": 47.01714285714286,
|
| 2774 |
+
"grad_norm": 0.0007373109692707658,
|
| 2775 |
+
"learning_rate": 4.7619047619047623e-07,
|
| 2776 |
+
"loss": 0.0,
|
| 2777 |
+
"step": 3350
|
| 2778 |
+
},
|
| 2779 |
+
{
|
| 2780 |
+
"epoch": 47.02,
|
| 2781 |
+
"grad_norm": 0.0050420379266142845,
|
| 2782 |
+
"learning_rate": 4.444444444444445e-07,
|
| 2783 |
+
"loss": 0.0,
|
| 2784 |
+
"step": 3360
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"epoch": 47.02,
|
| 2788 |
+
"eval_accuracy": 0.625,
|
| 2789 |
+
"eval_loss": 3.1150264739990234,
|
| 2790 |
+
"eval_runtime": 7.5053,
|
| 2791 |
+
"eval_samples_per_second": 4.264,
|
| 2792 |
+
"eval_steps_per_second": 1.066,
|
| 2793 |
+
"step": 3360
|
| 2794 |
+
},
|
| 2795 |
+
{
|
| 2796 |
+
"epoch": 48.002857142857145,
|
| 2797 |
+
"grad_norm": 0.001978852553293109,
|
| 2798 |
+
"learning_rate": 4.126984126984127e-07,
|
| 2799 |
+
"loss": 0.0,
|
| 2800 |
+
"step": 3370
|
| 2801 |
+
},
|
| 2802 |
+
{
|
| 2803 |
+
"epoch": 48.005714285714284,
|
| 2804 |
+
"grad_norm": 0.0011000612284988165,
|
| 2805 |
+
"learning_rate": 3.80952380952381e-07,
|
| 2806 |
+
"loss": 0.0007,
|
| 2807 |
+
"step": 3380
|
| 2808 |
+
},
|
| 2809 |
+
{
|
| 2810 |
+
"epoch": 48.00857142857143,
|
| 2811 |
+
"grad_norm": 0.0011735723819583654,
|
| 2812 |
+
"learning_rate": 3.492063492063492e-07,
|
| 2813 |
+
"loss": 0.0,
|
| 2814 |
+
"step": 3390
|
| 2815 |
+
},
|
| 2816 |
+
{
|
| 2817 |
+
"epoch": 48.011428571428574,
|
| 2818 |
+
"grad_norm": 0.0010873243445530534,
|
| 2819 |
+
"learning_rate": 3.174603174603175e-07,
|
| 2820 |
+
"loss": 0.0,
|
| 2821 |
+
"step": 3400
|
| 2822 |
+
},
|
| 2823 |
+
{
|
| 2824 |
+
"epoch": 48.01428571428571,
|
| 2825 |
+
"grad_norm": 0.0011251752730458975,
|
| 2826 |
+
"learning_rate": 2.8571428571428575e-07,
|
| 2827 |
+
"loss": 0.0,
|
| 2828 |
+
"step": 3410
|
| 2829 |
+
},
|
| 2830 |
+
{
|
| 2831 |
+
"epoch": 48.01714285714286,
|
| 2832 |
+
"grad_norm": 0.0013193455524742603,
|
| 2833 |
+
"learning_rate": 2.53968253968254e-07,
|
| 2834 |
+
"loss": 0.0,
|
| 2835 |
+
"step": 3420
|
| 2836 |
+
},
|
| 2837 |
+
{
|
| 2838 |
+
"epoch": 48.02,
|
| 2839 |
+
"grad_norm": 0.0006919482257217169,
|
| 2840 |
+
"learning_rate": 2.2222222222222224e-07,
|
| 2841 |
+
"loss": 0.0,
|
| 2842 |
+
"step": 3430
|
| 2843 |
+
},
|
| 2844 |
+
{
|
| 2845 |
+
"epoch": 48.02,
|
| 2846 |
+
"eval_accuracy": 0.625,
|
| 2847 |
+
"eval_loss": 3.112720489501953,
|
| 2848 |
+
"eval_runtime": 7.7277,
|
| 2849 |
+
"eval_samples_per_second": 4.141,
|
| 2850 |
+
"eval_steps_per_second": 1.035,
|
| 2851 |
+
"step": 3430
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 49.002857142857145,
|
| 2855 |
+
"grad_norm": 0.0010557180503383279,
|
| 2856 |
+
"learning_rate": 1.904761904761905e-07,
|
| 2857 |
+
"loss": 0.0,
|
| 2858 |
+
"step": 3440
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 49.005714285714284,
|
| 2862 |
+
"grad_norm": 0.006352465599775314,
|
| 2863 |
+
"learning_rate": 1.5873015873015874e-07,
|
| 2864 |
+
"loss": 0.0,
|
| 2865 |
+
"step": 3450
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 49.00857142857143,
|
| 2869 |
+
"grad_norm": 0.0013192709302529693,
|
| 2870 |
+
"learning_rate": 1.26984126984127e-07,
|
| 2871 |
+
"loss": 0.0,
|
| 2872 |
+
"step": 3460
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 49.011428571428574,
|
| 2876 |
+
"grad_norm": 0.0010165185667574406,
|
| 2877 |
+
"learning_rate": 9.523809523809525e-08,
|
| 2878 |
+
"loss": 0.0,
|
| 2879 |
+
"step": 3470
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 49.01428571428571,
|
| 2883 |
+
"grad_norm": 0.000668003223836422,
|
| 2884 |
+
"learning_rate": 6.34920634920635e-08,
|
| 2885 |
+
"loss": 0.0,
|
| 2886 |
+
"step": 3480
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 49.01714285714286,
|
| 2890 |
+
"grad_norm": 0.0007290692301467061,
|
| 2891 |
+
"learning_rate": 3.174603174603175e-08,
|
| 2892 |
+
"loss": 0.0,
|
| 2893 |
+
"step": 3490
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 49.02,
|
| 2897 |
+
"grad_norm": 0.0005605400656349957,
|
| 2898 |
+
"learning_rate": 0.0,
|
| 2899 |
+
"loss": 0.0,
|
| 2900 |
+
"step": 3500
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 49.02,
|
| 2904 |
+
"eval_accuracy": 0.625,
|
| 2905 |
+
"eval_loss": 3.1125903129577637,
|
| 2906 |
+
"eval_runtime": 7.757,
|
| 2907 |
+
"eval_samples_per_second": 4.125,
|
| 2908 |
+
"eval_steps_per_second": 1.031,
|
| 2909 |
+
"step": 3500
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 49.02,
|
| 2913 |
+
"step": 3500,
|
| 2914 |
+
"total_flos": 6.147436841415475e+19,
|
| 2915 |
+
"train_loss": 0.08523796891186586,
|
| 2916 |
+
"train_runtime": 6238.6099,
|
| 2917 |
+
"train_samples_per_second": 2.244,
|
| 2918 |
+
"train_steps_per_second": 0.561
|
| 2919 |
+
},
|
| 2920 |
+
{
|
| 2921 |
+
"epoch": 49.02,
|
| 2922 |
+
"eval_accuracy": 0.8372093023255814,
|
| 2923 |
+
"eval_loss": 0.8805378675460815,
|
| 2924 |
+
"eval_runtime": 9.9977,
|
| 2925 |
+
"eval_samples_per_second": 4.301,
|
| 2926 |
+
"eval_steps_per_second": 1.1,
|
| 2927 |
+
"step": 3500
|
| 2928 |
+
},
|
| 2929 |
+
{
|
| 2930 |
+
"epoch": 49.02,
|
| 2931 |
+
"eval_accuracy": 0.8372093023255814,
|
| 2932 |
+
"eval_loss": 0.8805379867553711,
|
| 2933 |
+
"eval_runtime": 9.0595,
|
| 2934 |
+
"eval_samples_per_second": 4.746,
|
| 2935 |
+
"eval_steps_per_second": 1.214,
|
| 2936 |
+
"step": 3500
|
| 2937 |
+
}
|
| 2938 |
+
],
|
| 2939 |
+
"logging_steps": 10,
|
| 2940 |
+
"max_steps": 3500,
|
| 2941 |
+
"num_input_tokens_seen": 0,
|
| 2942 |
+
"num_train_epochs": 9223372036854775807,
|
| 2943 |
+
"save_steps": 500,
|
| 2944 |
+
"stateful_callbacks": {
|
| 2945 |
+
"TrainerControl": {
|
| 2946 |
+
"args": {
|
| 2947 |
+
"should_epoch_stop": false,
|
| 2948 |
+
"should_evaluate": false,
|
| 2949 |
+
"should_log": false,
|
| 2950 |
+
"should_save": true,
|
| 2951 |
+
"should_training_stop": true
|
| 2952 |
+
},
|
| 2953 |
+
"attributes": {}
|
| 2954 |
+
}
|
| 2955 |
+
},
|
| 2956 |
+
"total_flos": 6.147436841415475e+19,
|
| 2957 |
+
"train_batch_size": 4,
|
| 2958 |
+
"trial_name": null,
|
| 2959 |
+
"trial_params": null
|
| 2960 |
+
}
|