End of training
Browse files- all_results.json +8 -0
- test_results.json +8 -0
- trainer_state.json +2163 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.25,
|
| 3 |
+
"eval_accuracy": 0.7043010752688172,
|
| 4 |
+
"eval_loss": 0.8327615261077881,
|
| 5 |
+
"eval_runtime": 1683.6481,
|
| 6 |
+
"eval_samples_per_second": 0.442,
|
| 7 |
+
"eval_steps_per_second": 0.055
|
| 8 |
+
}
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.25,
|
| 3 |
+
"eval_accuracy": 0.7043010752688172,
|
| 4 |
+
"eval_loss": 0.8327615261077881,
|
| 5 |
+
"eval_runtime": 1683.6481,
|
| 6 |
+
"eval_samples_per_second": 0.442,
|
| 7 |
+
"eval_steps_per_second": 0.055
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2163 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7328859060402685,
|
| 3 |
+
"best_model_checkpoint": "vivit-b-16x2-kinetics400-finetuned-crema-d/checkpoint-2976",
|
| 4 |
+
"epoch": 3.248991935483871,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2976,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"grad_norm": 10.041251182556152,
|
| 14 |
+
"learning_rate": 1.6778523489932886e-06,
|
| 15 |
+
"loss": 2.0618,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01,
|
| 20 |
+
"grad_norm": 12.547879219055176,
|
| 21 |
+
"learning_rate": 3.3557046979865773e-06,
|
| 22 |
+
"loss": 1.9746,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.01,
|
| 27 |
+
"grad_norm": 9.306560516357422,
|
| 28 |
+
"learning_rate": 5.033557046979865e-06,
|
| 29 |
+
"loss": 1.8185,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.01,
|
| 34 |
+
"grad_norm": 7.815144062042236,
|
| 35 |
+
"learning_rate": 6.7114093959731546e-06,
|
| 36 |
+
"loss": 1.8203,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.02,
|
| 41 |
+
"grad_norm": 10.477940559387207,
|
| 42 |
+
"learning_rate": 8.389261744966444e-06,
|
| 43 |
+
"loss": 1.8039,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02,
|
| 48 |
+
"grad_norm": 8.257948875427246,
|
| 49 |
+
"learning_rate": 1.006711409395973e-05,
|
| 50 |
+
"loss": 1.7131,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02,
|
| 55 |
+
"grad_norm": 7.9578776359558105,
|
| 56 |
+
"learning_rate": 1.174496644295302e-05,
|
| 57 |
+
"loss": 1.701,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.03,
|
| 62 |
+
"grad_norm": 11.580317497253418,
|
| 63 |
+
"learning_rate": 1.3422818791946309e-05,
|
| 64 |
+
"loss": 1.5872,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.03,
|
| 69 |
+
"grad_norm": 12.274490356445312,
|
| 70 |
+
"learning_rate": 1.51006711409396e-05,
|
| 71 |
+
"loss": 1.5683,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.03,
|
| 76 |
+
"grad_norm": 9.367303848266602,
|
| 77 |
+
"learning_rate": 1.6778523489932888e-05,
|
| 78 |
+
"loss": 1.557,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.04,
|
| 83 |
+
"grad_norm": 7.54719877243042,
|
| 84 |
+
"learning_rate": 1.8456375838926178e-05,
|
| 85 |
+
"loss": 1.4923,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.04,
|
| 90 |
+
"grad_norm": 8.119046211242676,
|
| 91 |
+
"learning_rate": 2.013422818791946e-05,
|
| 92 |
+
"loss": 1.4376,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.04,
|
| 97 |
+
"grad_norm": 8.935946464538574,
|
| 98 |
+
"learning_rate": 2.181208053691275e-05,
|
| 99 |
+
"loss": 1.5429,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.05,
|
| 104 |
+
"grad_norm": 7.0674147605896,
|
| 105 |
+
"learning_rate": 2.348993288590604e-05,
|
| 106 |
+
"loss": 1.3362,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.05,
|
| 111 |
+
"grad_norm": 11.671470642089844,
|
| 112 |
+
"learning_rate": 2.516778523489933e-05,
|
| 113 |
+
"loss": 1.371,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.05,
|
| 118 |
+
"grad_norm": 11.36544132232666,
|
| 119 |
+
"learning_rate": 2.6845637583892618e-05,
|
| 120 |
+
"loss": 1.3582,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.06,
|
| 125 |
+
"grad_norm": 11.540151596069336,
|
| 126 |
+
"learning_rate": 2.8523489932885905e-05,
|
| 127 |
+
"loss": 1.3666,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.06,
|
| 132 |
+
"grad_norm": 12.664009094238281,
|
| 133 |
+
"learning_rate": 3.02013422818792e-05,
|
| 134 |
+
"loss": 1.5347,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.06,
|
| 139 |
+
"grad_norm": 8.551016807556152,
|
| 140 |
+
"learning_rate": 3.1879194630872485e-05,
|
| 141 |
+
"loss": 1.3923,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.07,
|
| 146 |
+
"grad_norm": 10.388668060302734,
|
| 147 |
+
"learning_rate": 3.3557046979865775e-05,
|
| 148 |
+
"loss": 1.3822,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.07,
|
| 153 |
+
"grad_norm": 10.217473983764648,
|
| 154 |
+
"learning_rate": 3.523489932885906e-05,
|
| 155 |
+
"loss": 1.2387,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.07,
|
| 160 |
+
"grad_norm": 10.830105781555176,
|
| 161 |
+
"learning_rate": 3.6912751677852356e-05,
|
| 162 |
+
"loss": 1.2035,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.08,
|
| 167 |
+
"grad_norm": 14.143545150756836,
|
| 168 |
+
"learning_rate": 3.859060402684564e-05,
|
| 169 |
+
"loss": 1.2766,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.08,
|
| 174 |
+
"grad_norm": 12.051302909851074,
|
| 175 |
+
"learning_rate": 4.026845637583892e-05,
|
| 176 |
+
"loss": 1.2969,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.08,
|
| 181 |
+
"grad_norm": 7.3714823722839355,
|
| 182 |
+
"learning_rate": 4.194630872483222e-05,
|
| 183 |
+
"loss": 1.1517,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.09,
|
| 188 |
+
"grad_norm": 9.341689109802246,
|
| 189 |
+
"learning_rate": 4.36241610738255e-05,
|
| 190 |
+
"loss": 1.2486,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.09,
|
| 195 |
+
"grad_norm": 9.874135971069336,
|
| 196 |
+
"learning_rate": 4.530201342281879e-05,
|
| 197 |
+
"loss": 1.1304,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.09,
|
| 202 |
+
"grad_norm": 10.551310539245605,
|
| 203 |
+
"learning_rate": 4.697986577181208e-05,
|
| 204 |
+
"loss": 1.1465,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.1,
|
| 209 |
+
"grad_norm": 14.167583465576172,
|
| 210 |
+
"learning_rate": 4.865771812080537e-05,
|
| 211 |
+
"loss": 1.2225,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.1,
|
| 216 |
+
"grad_norm": 7.046817302703857,
|
| 217 |
+
"learning_rate": 4.996265870052278e-05,
|
| 218 |
+
"loss": 1.1917,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.1,
|
| 223 |
+
"grad_norm": 8.558070182800293,
|
| 224 |
+
"learning_rate": 4.977595220313667e-05,
|
| 225 |
+
"loss": 1.3288,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.11,
|
| 230 |
+
"grad_norm": 7.99010705947876,
|
| 231 |
+
"learning_rate": 4.958924570575056e-05,
|
| 232 |
+
"loss": 1.1893,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.11,
|
| 237 |
+
"grad_norm": 9.63879108428955,
|
| 238 |
+
"learning_rate": 4.9402539208364454e-05,
|
| 239 |
+
"loss": 1.1212,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.11,
|
| 244 |
+
"grad_norm": 11.649455070495605,
|
| 245 |
+
"learning_rate": 4.9215832710978346e-05,
|
| 246 |
+
"loss": 1.0572,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.12,
|
| 251 |
+
"grad_norm": 13.81701946258545,
|
| 252 |
+
"learning_rate": 4.902912621359224e-05,
|
| 253 |
+
"loss": 1.1869,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.12,
|
| 258 |
+
"grad_norm": 13.85696029663086,
|
| 259 |
+
"learning_rate": 4.884241971620613e-05,
|
| 260 |
+
"loss": 1.5038,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.12,
|
| 265 |
+
"grad_norm": 8.543469429016113,
|
| 266 |
+
"learning_rate": 4.8655713218820016e-05,
|
| 267 |
+
"loss": 1.0457,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.13,
|
| 272 |
+
"grad_norm": 10.270995140075684,
|
| 273 |
+
"learning_rate": 4.846900672143391e-05,
|
| 274 |
+
"loss": 0.9543,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.13,
|
| 279 |
+
"grad_norm": 8.127942085266113,
|
| 280 |
+
"learning_rate": 4.82823002240478e-05,
|
| 281 |
+
"loss": 0.9389,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.13,
|
| 286 |
+
"grad_norm": 10.777777671813965,
|
| 287 |
+
"learning_rate": 4.809559372666169e-05,
|
| 288 |
+
"loss": 1.0655,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.14,
|
| 293 |
+
"grad_norm": 8.693421363830566,
|
| 294 |
+
"learning_rate": 4.790888722927558e-05,
|
| 295 |
+
"loss": 1.0275,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.14,
|
| 300 |
+
"grad_norm": 10.789031028747559,
|
| 301 |
+
"learning_rate": 4.772218073188947e-05,
|
| 302 |
+
"loss": 1.2218,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.14,
|
| 307 |
+
"grad_norm": 6.10308313369751,
|
| 308 |
+
"learning_rate": 4.753547423450336e-05,
|
| 309 |
+
"loss": 0.9795,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.15,
|
| 314 |
+
"grad_norm": 11.924713134765625,
|
| 315 |
+
"learning_rate": 4.7348767737117256e-05,
|
| 316 |
+
"loss": 1.0568,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.15,
|
| 321 |
+
"grad_norm": 5.780080795288086,
|
| 322 |
+
"learning_rate": 4.716206123973114e-05,
|
| 323 |
+
"loss": 0.9916,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.15,
|
| 328 |
+
"grad_norm": 7.771975994110107,
|
| 329 |
+
"learning_rate": 4.697535474234503e-05,
|
| 330 |
+
"loss": 0.9661,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.16,
|
| 335 |
+
"grad_norm": 9.780990600585938,
|
| 336 |
+
"learning_rate": 4.6788648244958926e-05,
|
| 337 |
+
"loss": 0.884,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.16,
|
| 342 |
+
"grad_norm": 6.3119354248046875,
|
| 343 |
+
"learning_rate": 4.660194174757282e-05,
|
| 344 |
+
"loss": 1.3266,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.16,
|
| 349 |
+
"grad_norm": 6.907146453857422,
|
| 350 |
+
"learning_rate": 4.6415235250186703e-05,
|
| 351 |
+
"loss": 1.0728,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.17,
|
| 356 |
+
"grad_norm": 6.732804775238037,
|
| 357 |
+
"learning_rate": 4.6228528752800596e-05,
|
| 358 |
+
"loss": 1.1793,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.17,
|
| 363 |
+
"grad_norm": 5.275418281555176,
|
| 364 |
+
"learning_rate": 4.604182225541449e-05,
|
| 365 |
+
"loss": 1.0455,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.17,
|
| 370 |
+
"grad_norm": 8.184633255004883,
|
| 371 |
+
"learning_rate": 4.585511575802838e-05,
|
| 372 |
+
"loss": 1.2988,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.18,
|
| 377 |
+
"grad_norm": 9.226140022277832,
|
| 378 |
+
"learning_rate": 4.566840926064227e-05,
|
| 379 |
+
"loss": 0.9656,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.18,
|
| 384 |
+
"grad_norm": 9.135764122009277,
|
| 385 |
+
"learning_rate": 4.5481702763256165e-05,
|
| 386 |
+
"loss": 0.9574,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.18,
|
| 391 |
+
"grad_norm": 9.493541717529297,
|
| 392 |
+
"learning_rate": 4.529499626587006e-05,
|
| 393 |
+
"loss": 1.177,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.19,
|
| 398 |
+
"grad_norm": 10.075041770935059,
|
| 399 |
+
"learning_rate": 4.510828976848395e-05,
|
| 400 |
+
"loss": 1.0754,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.19,
|
| 405 |
+
"grad_norm": 9.099665641784668,
|
| 406 |
+
"learning_rate": 4.492158327109784e-05,
|
| 407 |
+
"loss": 1.179,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.19,
|
| 412 |
+
"grad_norm": 18.290414810180664,
|
| 413 |
+
"learning_rate": 4.473487677371173e-05,
|
| 414 |
+
"loss": 1.14,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.2,
|
| 419 |
+
"grad_norm": 6.768624305725098,
|
| 420 |
+
"learning_rate": 4.454817027632562e-05,
|
| 421 |
+
"loss": 1.0689,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.2,
|
| 426 |
+
"grad_norm": 11.730013847351074,
|
| 427 |
+
"learning_rate": 4.436146377893951e-05,
|
| 428 |
+
"loss": 0.8808,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.2,
|
| 433 |
+
"grad_norm": 7.4615159034729,
|
| 434 |
+
"learning_rate": 4.4174757281553404e-05,
|
| 435 |
+
"loss": 1.093,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.21,
|
| 440 |
+
"grad_norm": 11.849823951721191,
|
| 441 |
+
"learning_rate": 4.398805078416729e-05,
|
| 442 |
+
"loss": 0.9596,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.21,
|
| 447 |
+
"grad_norm": 8.06859302520752,
|
| 448 |
+
"learning_rate": 4.380134428678118e-05,
|
| 449 |
+
"loss": 0.9528,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.22,
|
| 454 |
+
"grad_norm": 11.52597713470459,
|
| 455 |
+
"learning_rate": 4.3614637789395075e-05,
|
| 456 |
+
"loss": 0.8866,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.22,
|
| 461 |
+
"grad_norm": 12.404598236083984,
|
| 462 |
+
"learning_rate": 4.342793129200897e-05,
|
| 463 |
+
"loss": 0.7092,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.22,
|
| 468 |
+
"grad_norm": 6.062406539916992,
|
| 469 |
+
"learning_rate": 4.324122479462285e-05,
|
| 470 |
+
"loss": 0.7407,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.23,
|
| 475 |
+
"grad_norm": 11.590093612670898,
|
| 476 |
+
"learning_rate": 4.3054518297236745e-05,
|
| 477 |
+
"loss": 1.1617,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.23,
|
| 482 |
+
"grad_norm": 8.003750801086426,
|
| 483 |
+
"learning_rate": 4.286781179985064e-05,
|
| 484 |
+
"loss": 0.9215,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.23,
|
| 489 |
+
"grad_norm": 6.014645099639893,
|
| 490 |
+
"learning_rate": 4.268110530246453e-05,
|
| 491 |
+
"loss": 0.8719,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.24,
|
| 496 |
+
"grad_norm": 7.771167755126953,
|
| 497 |
+
"learning_rate": 4.2494398805078415e-05,
|
| 498 |
+
"loss": 0.865,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.24,
|
| 503 |
+
"grad_norm": 6.302040100097656,
|
| 504 |
+
"learning_rate": 4.230769230769231e-05,
|
| 505 |
+
"loss": 0.7946,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.24,
|
| 510 |
+
"grad_norm": 14.395988464355469,
|
| 511 |
+
"learning_rate": 4.21209858103062e-05,
|
| 512 |
+
"loss": 1.0984,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.25,
|
| 517 |
+
"grad_norm": 7.767952919006348,
|
| 518 |
+
"learning_rate": 4.193427931292009e-05,
|
| 519 |
+
"loss": 0.9101,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.25,
|
| 524 |
+
"grad_norm": 11.00684642791748,
|
| 525 |
+
"learning_rate": 4.1747572815533984e-05,
|
| 526 |
+
"loss": 0.9023,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.25,
|
| 531 |
+
"eval_accuracy": 0.6013422818791946,
|
| 532 |
+
"eval_loss": 1.0584642887115479,
|
| 533 |
+
"eval_runtime": 1869.6092,
|
| 534 |
+
"eval_samples_per_second": 0.398,
|
| 535 |
+
"eval_steps_per_second": 0.05,
|
| 536 |
+
"step": 745
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.0,
|
| 540 |
+
"grad_norm": 8.297401428222656,
|
| 541 |
+
"learning_rate": 4.1560866318147876e-05,
|
| 542 |
+
"loss": 0.6975,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.01,
|
| 547 |
+
"grad_norm": 8.592920303344727,
|
| 548 |
+
"learning_rate": 4.137415982076177e-05,
|
| 549 |
+
"loss": 0.6746,
|
| 550 |
+
"step": 760
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.01,
|
| 554 |
+
"grad_norm": 12.73000717163086,
|
| 555 |
+
"learning_rate": 4.118745332337566e-05,
|
| 556 |
+
"loss": 0.7792,
|
| 557 |
+
"step": 770
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.01,
|
| 561 |
+
"grad_norm": 12.524226188659668,
|
| 562 |
+
"learning_rate": 4.1000746825989546e-05,
|
| 563 |
+
"loss": 1.1009,
|
| 564 |
+
"step": 780
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.02,
|
| 568 |
+
"grad_norm": 14.503805160522461,
|
| 569 |
+
"learning_rate": 4.081404032860344e-05,
|
| 570 |
+
"loss": 0.8554,
|
| 571 |
+
"step": 790
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.02,
|
| 575 |
+
"grad_norm": 6.50175666809082,
|
| 576 |
+
"learning_rate": 4.062733383121733e-05,
|
| 577 |
+
"loss": 0.9718,
|
| 578 |
+
"step": 800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.02,
|
| 582 |
+
"grad_norm": 6.452315807342529,
|
| 583 |
+
"learning_rate": 4.0440627333831223e-05,
|
| 584 |
+
"loss": 0.9421,
|
| 585 |
+
"step": 810
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.03,
|
| 589 |
+
"grad_norm": 17.507164001464844,
|
| 590 |
+
"learning_rate": 4.025392083644511e-05,
|
| 591 |
+
"loss": 0.9593,
|
| 592 |
+
"step": 820
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.03,
|
| 596 |
+
"grad_norm": 6.719306945800781,
|
| 597 |
+
"learning_rate": 4.0067214339059e-05,
|
| 598 |
+
"loss": 0.8568,
|
| 599 |
+
"step": 830
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.03,
|
| 603 |
+
"grad_norm": 8.009960174560547,
|
| 604 |
+
"learning_rate": 3.9880507841672894e-05,
|
| 605 |
+
"loss": 0.796,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.04,
|
| 610 |
+
"grad_norm": 6.025060653686523,
|
| 611 |
+
"learning_rate": 3.9693801344286786e-05,
|
| 612 |
+
"loss": 0.8486,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.04,
|
| 617 |
+
"grad_norm": 8.915096282958984,
|
| 618 |
+
"learning_rate": 3.950709484690067e-05,
|
| 619 |
+
"loss": 0.8042,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.04,
|
| 624 |
+
"grad_norm": 12.27297592163086,
|
| 625 |
+
"learning_rate": 3.9320388349514564e-05,
|
| 626 |
+
"loss": 0.8384,
|
| 627 |
+
"step": 870
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.05,
|
| 631 |
+
"grad_norm": 5.073918342590332,
|
| 632 |
+
"learning_rate": 3.9133681852128456e-05,
|
| 633 |
+
"loss": 0.8608,
|
| 634 |
+
"step": 880
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.05,
|
| 638 |
+
"grad_norm": 10.088605880737305,
|
| 639 |
+
"learning_rate": 3.894697535474235e-05,
|
| 640 |
+
"loss": 0.8858,
|
| 641 |
+
"step": 890
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.05,
|
| 645 |
+
"grad_norm": 5.86115837097168,
|
| 646 |
+
"learning_rate": 3.8760268857356234e-05,
|
| 647 |
+
"loss": 0.6226,
|
| 648 |
+
"step": 900
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.06,
|
| 652 |
+
"grad_norm": 11.275440216064453,
|
| 653 |
+
"learning_rate": 3.8573562359970126e-05,
|
| 654 |
+
"loss": 0.8814,
|
| 655 |
+
"step": 910
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.06,
|
| 659 |
+
"grad_norm": 5.241692543029785,
|
| 660 |
+
"learning_rate": 3.838685586258402e-05,
|
| 661 |
+
"loss": 0.8163,
|
| 662 |
+
"step": 920
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.06,
|
| 666 |
+
"grad_norm": 8.397137641906738,
|
| 667 |
+
"learning_rate": 3.820014936519791e-05,
|
| 668 |
+
"loss": 0.7511,
|
| 669 |
+
"step": 930
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.07,
|
| 673 |
+
"grad_norm": 9.3834810256958,
|
| 674 |
+
"learning_rate": 3.8013442867811796e-05,
|
| 675 |
+
"loss": 0.7023,
|
| 676 |
+
"step": 940
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.07,
|
| 680 |
+
"grad_norm": 11.98423957824707,
|
| 681 |
+
"learning_rate": 3.782673637042569e-05,
|
| 682 |
+
"loss": 0.7885,
|
| 683 |
+
"step": 950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.07,
|
| 687 |
+
"grad_norm": 11.273078918457031,
|
| 688 |
+
"learning_rate": 3.764002987303958e-05,
|
| 689 |
+
"loss": 0.6545,
|
| 690 |
+
"step": 960
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.08,
|
| 694 |
+
"grad_norm": 14.190473556518555,
|
| 695 |
+
"learning_rate": 3.745332337565347e-05,
|
| 696 |
+
"loss": 1.0184,
|
| 697 |
+
"step": 970
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.08,
|
| 701 |
+
"grad_norm": 4.971963405609131,
|
| 702 |
+
"learning_rate": 3.7266616878267365e-05,
|
| 703 |
+
"loss": 0.7963,
|
| 704 |
+
"step": 980
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.08,
|
| 708 |
+
"grad_norm": 17.721282958984375,
|
| 709 |
+
"learning_rate": 3.707991038088126e-05,
|
| 710 |
+
"loss": 0.9323,
|
| 711 |
+
"step": 990
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.09,
|
| 715 |
+
"grad_norm": 10.512914657592773,
|
| 716 |
+
"learning_rate": 3.689320388349515e-05,
|
| 717 |
+
"loss": 0.9391,
|
| 718 |
+
"step": 1000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.09,
|
| 722 |
+
"grad_norm": 11.076910018920898,
|
| 723 |
+
"learning_rate": 3.670649738610904e-05,
|
| 724 |
+
"loss": 0.8941,
|
| 725 |
+
"step": 1010
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.09,
|
| 729 |
+
"grad_norm": 5.027378082275391,
|
| 730 |
+
"learning_rate": 3.651979088872293e-05,
|
| 731 |
+
"loss": 0.8346,
|
| 732 |
+
"step": 1020
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.1,
|
| 736 |
+
"grad_norm": 12.940291404724121,
|
| 737 |
+
"learning_rate": 3.633308439133682e-05,
|
| 738 |
+
"loss": 0.7896,
|
| 739 |
+
"step": 1030
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.1,
|
| 743 |
+
"grad_norm": 13.178154945373535,
|
| 744 |
+
"learning_rate": 3.614637789395071e-05,
|
| 745 |
+
"loss": 0.6817,
|
| 746 |
+
"step": 1040
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.1,
|
| 750 |
+
"grad_norm": 13.505011558532715,
|
| 751 |
+
"learning_rate": 3.5959671396564605e-05,
|
| 752 |
+
"loss": 0.7627,
|
| 753 |
+
"step": 1050
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.11,
|
| 757 |
+
"grad_norm": 4.09556245803833,
|
| 758 |
+
"learning_rate": 3.577296489917849e-05,
|
| 759 |
+
"loss": 0.8155,
|
| 760 |
+
"step": 1060
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.11,
|
| 764 |
+
"grad_norm": 12.600196838378906,
|
| 765 |
+
"learning_rate": 3.558625840179238e-05,
|
| 766 |
+
"loss": 0.89,
|
| 767 |
+
"step": 1070
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.11,
|
| 771 |
+
"grad_norm": 8.050426483154297,
|
| 772 |
+
"learning_rate": 3.5399551904406275e-05,
|
| 773 |
+
"loss": 0.8544,
|
| 774 |
+
"step": 1080
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.12,
|
| 778 |
+
"grad_norm": 13.491654396057129,
|
| 779 |
+
"learning_rate": 3.521284540702017e-05,
|
| 780 |
+
"loss": 0.6694,
|
| 781 |
+
"step": 1090
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.12,
|
| 785 |
+
"grad_norm": 6.23304557800293,
|
| 786 |
+
"learning_rate": 3.502613890963405e-05,
|
| 787 |
+
"loss": 0.8363,
|
| 788 |
+
"step": 1100
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.12,
|
| 792 |
+
"grad_norm": 12.755766868591309,
|
| 793 |
+
"learning_rate": 3.4839432412247945e-05,
|
| 794 |
+
"loss": 0.6294,
|
| 795 |
+
"step": 1110
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.13,
|
| 799 |
+
"grad_norm": 7.085630416870117,
|
| 800 |
+
"learning_rate": 3.465272591486184e-05,
|
| 801 |
+
"loss": 0.6517,
|
| 802 |
+
"step": 1120
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.13,
|
| 806 |
+
"grad_norm": 7.111999988555908,
|
| 807 |
+
"learning_rate": 3.446601941747573e-05,
|
| 808 |
+
"loss": 0.6923,
|
| 809 |
+
"step": 1130
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.13,
|
| 813 |
+
"grad_norm": 9.9430513381958,
|
| 814 |
+
"learning_rate": 3.427931292008962e-05,
|
| 815 |
+
"loss": 0.7989,
|
| 816 |
+
"step": 1140
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.14,
|
| 820 |
+
"grad_norm": 7.441895484924316,
|
| 821 |
+
"learning_rate": 3.409260642270351e-05,
|
| 822 |
+
"loss": 0.8628,
|
| 823 |
+
"step": 1150
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.14,
|
| 827 |
+
"grad_norm": 11.109057426452637,
|
| 828 |
+
"learning_rate": 3.39058999253174e-05,
|
| 829 |
+
"loss": 0.9556,
|
| 830 |
+
"step": 1160
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.14,
|
| 834 |
+
"grad_norm": 3.8683066368103027,
|
| 835 |
+
"learning_rate": 3.371919342793129e-05,
|
| 836 |
+
"loss": 0.7598,
|
| 837 |
+
"step": 1170
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.15,
|
| 841 |
+
"grad_norm": 9.229301452636719,
|
| 842 |
+
"learning_rate": 3.3532486930545184e-05,
|
| 843 |
+
"loss": 0.7694,
|
| 844 |
+
"step": 1180
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.15,
|
| 848 |
+
"grad_norm": 6.90978479385376,
|
| 849 |
+
"learning_rate": 3.334578043315908e-05,
|
| 850 |
+
"loss": 0.6816,
|
| 851 |
+
"step": 1190
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.15,
|
| 855 |
+
"grad_norm": 10.782751083374023,
|
| 856 |
+
"learning_rate": 3.315907393577297e-05,
|
| 857 |
+
"loss": 0.7015,
|
| 858 |
+
"step": 1200
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.16,
|
| 862 |
+
"grad_norm": 12.352110862731934,
|
| 863 |
+
"learning_rate": 3.297236743838686e-05,
|
| 864 |
+
"loss": 0.6476,
|
| 865 |
+
"step": 1210
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.16,
|
| 869 |
+
"grad_norm": 12.289131164550781,
|
| 870 |
+
"learning_rate": 3.2785660941000754e-05,
|
| 871 |
+
"loss": 0.6243,
|
| 872 |
+
"step": 1220
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.16,
|
| 876 |
+
"grad_norm": 10.164794921875,
|
| 877 |
+
"learning_rate": 3.259895444361464e-05,
|
| 878 |
+
"loss": 0.7136,
|
| 879 |
+
"step": 1230
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.17,
|
| 883 |
+
"grad_norm": 11.630318641662598,
|
| 884 |
+
"learning_rate": 3.241224794622853e-05,
|
| 885 |
+
"loss": 0.8095,
|
| 886 |
+
"step": 1240
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.17,
|
| 890 |
+
"grad_norm": 8.73193359375,
|
| 891 |
+
"learning_rate": 3.2225541448842424e-05,
|
| 892 |
+
"loss": 0.7918,
|
| 893 |
+
"step": 1250
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.17,
|
| 897 |
+
"grad_norm": 10.5596284866333,
|
| 898 |
+
"learning_rate": 3.2038834951456316e-05,
|
| 899 |
+
"loss": 0.9087,
|
| 900 |
+
"step": 1260
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.18,
|
| 904 |
+
"grad_norm": 10.88122844696045,
|
| 905 |
+
"learning_rate": 3.18521284540702e-05,
|
| 906 |
+
"loss": 0.5758,
|
| 907 |
+
"step": 1270
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.18,
|
| 911 |
+
"grad_norm": 7.157528877258301,
|
| 912 |
+
"learning_rate": 3.1665421956684094e-05,
|
| 913 |
+
"loss": 0.645,
|
| 914 |
+
"step": 1280
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.18,
|
| 918 |
+
"grad_norm": 6.619960784912109,
|
| 919 |
+
"learning_rate": 3.1478715459297986e-05,
|
| 920 |
+
"loss": 0.6412,
|
| 921 |
+
"step": 1290
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.19,
|
| 925 |
+
"grad_norm": 9.903483390808105,
|
| 926 |
+
"learning_rate": 3.129200896191188e-05,
|
| 927 |
+
"loss": 0.5866,
|
| 928 |
+
"step": 1300
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.19,
|
| 932 |
+
"grad_norm": 8.340232849121094,
|
| 933 |
+
"learning_rate": 3.1105302464525764e-05,
|
| 934 |
+
"loss": 0.6971,
|
| 935 |
+
"step": 1310
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.19,
|
| 939 |
+
"grad_norm": 4.8595170974731445,
|
| 940 |
+
"learning_rate": 3.0918595967139656e-05,
|
| 941 |
+
"loss": 0.6824,
|
| 942 |
+
"step": 1320
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.2,
|
| 946 |
+
"grad_norm": 13.451203346252441,
|
| 947 |
+
"learning_rate": 3.073188946975355e-05,
|
| 948 |
+
"loss": 0.5937,
|
| 949 |
+
"step": 1330
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.2,
|
| 953 |
+
"grad_norm": 8.88749885559082,
|
| 954 |
+
"learning_rate": 3.054518297236744e-05,
|
| 955 |
+
"loss": 0.7642,
|
| 956 |
+
"step": 1340
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.2,
|
| 960 |
+
"grad_norm": 10.660982131958008,
|
| 961 |
+
"learning_rate": 3.035847647498133e-05,
|
| 962 |
+
"loss": 0.7327,
|
| 963 |
+
"step": 1350
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.21,
|
| 967 |
+
"grad_norm": 12.592216491699219,
|
| 968 |
+
"learning_rate": 3.0171769977595222e-05,
|
| 969 |
+
"loss": 0.6769,
|
| 970 |
+
"step": 1360
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.21,
|
| 974 |
+
"grad_norm": 3.5256502628326416,
|
| 975 |
+
"learning_rate": 2.9985063480209115e-05,
|
| 976 |
+
"loss": 0.6646,
|
| 977 |
+
"step": 1370
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.21,
|
| 981 |
+
"grad_norm": 6.979560852050781,
|
| 982 |
+
"learning_rate": 2.9798356982823007e-05,
|
| 983 |
+
"loss": 0.7394,
|
| 984 |
+
"step": 1380
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.22,
|
| 988 |
+
"grad_norm": 8.0765962600708,
|
| 989 |
+
"learning_rate": 2.9611650485436892e-05,
|
| 990 |
+
"loss": 0.597,
|
| 991 |
+
"step": 1390
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.22,
|
| 995 |
+
"grad_norm": 6.306230545043945,
|
| 996 |
+
"learning_rate": 2.9424943988050785e-05,
|
| 997 |
+
"loss": 0.557,
|
| 998 |
+
"step": 1400
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.22,
|
| 1002 |
+
"grad_norm": 7.691712856292725,
|
| 1003 |
+
"learning_rate": 2.9238237490664677e-05,
|
| 1004 |
+
"loss": 0.778,
|
| 1005 |
+
"step": 1410
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.23,
|
| 1009 |
+
"grad_norm": 10.76540756225586,
|
| 1010 |
+
"learning_rate": 2.905153099327857e-05,
|
| 1011 |
+
"loss": 0.7148,
|
| 1012 |
+
"step": 1420
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.23,
|
| 1016 |
+
"grad_norm": 8.003518104553223,
|
| 1017 |
+
"learning_rate": 2.8864824495892455e-05,
|
| 1018 |
+
"loss": 0.799,
|
| 1019 |
+
"step": 1430
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.23,
|
| 1023 |
+
"grad_norm": 8.381891250610352,
|
| 1024 |
+
"learning_rate": 2.8678117998506347e-05,
|
| 1025 |
+
"loss": 0.7274,
|
| 1026 |
+
"step": 1440
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.24,
|
| 1030 |
+
"grad_norm": 11.402746200561523,
|
| 1031 |
+
"learning_rate": 2.849141150112024e-05,
|
| 1032 |
+
"loss": 0.767,
|
| 1033 |
+
"step": 1450
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.24,
|
| 1037 |
+
"grad_norm": 12.372224807739258,
|
| 1038 |
+
"learning_rate": 2.8304705003734132e-05,
|
| 1039 |
+
"loss": 0.4786,
|
| 1040 |
+
"step": 1460
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.24,
|
| 1044 |
+
"grad_norm": 6.330965518951416,
|
| 1045 |
+
"learning_rate": 2.811799850634802e-05,
|
| 1046 |
+
"loss": 0.6594,
|
| 1047 |
+
"step": 1470
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.25,
|
| 1051 |
+
"grad_norm": 10.601953506469727,
|
| 1052 |
+
"learning_rate": 2.7931292008961913e-05,
|
| 1053 |
+
"loss": 0.5358,
|
| 1054 |
+
"step": 1480
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.25,
|
| 1058 |
+
"grad_norm": 55.654109954833984,
|
| 1059 |
+
"learning_rate": 2.7744585511575805e-05,
|
| 1060 |
+
"loss": 0.9428,
|
| 1061 |
+
"step": 1490
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.25,
|
| 1065 |
+
"eval_accuracy": 0.6510067114093959,
|
| 1066 |
+
"eval_loss": 0.897102415561676,
|
| 1067 |
+
"eval_runtime": 1724.9592,
|
| 1068 |
+
"eval_samples_per_second": 0.432,
|
| 1069 |
+
"eval_steps_per_second": 0.054,
|
| 1070 |
+
"step": 1490
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.0,
|
| 1074 |
+
"grad_norm": 14.913216590881348,
|
| 1075 |
+
"learning_rate": 2.7557879014189698e-05,
|
| 1076 |
+
"loss": 0.6672,
|
| 1077 |
+
"step": 1500
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.01,
|
| 1081 |
+
"grad_norm": 8.555597305297852,
|
| 1082 |
+
"learning_rate": 2.7371172516803583e-05,
|
| 1083 |
+
"loss": 0.517,
|
| 1084 |
+
"step": 1510
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.01,
|
| 1088 |
+
"grad_norm": 13.102532386779785,
|
| 1089 |
+
"learning_rate": 2.7184466019417475e-05,
|
| 1090 |
+
"loss": 0.4576,
|
| 1091 |
+
"step": 1520
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.01,
|
| 1095 |
+
"grad_norm": 8.147948265075684,
|
| 1096 |
+
"learning_rate": 2.6997759522031368e-05,
|
| 1097 |
+
"loss": 0.5292,
|
| 1098 |
+
"step": 1530
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.02,
|
| 1102 |
+
"grad_norm": 9.257566452026367,
|
| 1103 |
+
"learning_rate": 2.681105302464526e-05,
|
| 1104 |
+
"loss": 0.5036,
|
| 1105 |
+
"step": 1540
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.02,
|
| 1109 |
+
"grad_norm": 11.560401916503906,
|
| 1110 |
+
"learning_rate": 2.662434652725915e-05,
|
| 1111 |
+
"loss": 0.6044,
|
| 1112 |
+
"step": 1550
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.02,
|
| 1116 |
+
"grad_norm": 11.875014305114746,
|
| 1117 |
+
"learning_rate": 2.643764002987304e-05,
|
| 1118 |
+
"loss": 0.5306,
|
| 1119 |
+
"step": 1560
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.03,
|
| 1123 |
+
"grad_norm": 13.541814804077148,
|
| 1124 |
+
"learning_rate": 2.6250933532486934e-05,
|
| 1125 |
+
"loss": 0.6254,
|
| 1126 |
+
"step": 1570
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.03,
|
| 1130 |
+
"grad_norm": 9.984253883361816,
|
| 1131 |
+
"learning_rate": 2.6064227035100826e-05,
|
| 1132 |
+
"loss": 0.5974,
|
| 1133 |
+
"step": 1580
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 2.03,
|
| 1137 |
+
"grad_norm": 10.722140312194824,
|
| 1138 |
+
"learning_rate": 2.587752053771471e-05,
|
| 1139 |
+
"loss": 0.7041,
|
| 1140 |
+
"step": 1590
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.04,
|
| 1144 |
+
"grad_norm": 10.557269096374512,
|
| 1145 |
+
"learning_rate": 2.5690814040328604e-05,
|
| 1146 |
+
"loss": 0.5909,
|
| 1147 |
+
"step": 1600
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 2.04,
|
| 1151 |
+
"grad_norm": 11.491299629211426,
|
| 1152 |
+
"learning_rate": 2.5504107542942496e-05,
|
| 1153 |
+
"loss": 0.4946,
|
| 1154 |
+
"step": 1610
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.04,
|
| 1158 |
+
"grad_norm": 8.835367202758789,
|
| 1159 |
+
"learning_rate": 2.531740104555639e-05,
|
| 1160 |
+
"loss": 0.6644,
|
| 1161 |
+
"step": 1620
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 2.05,
|
| 1165 |
+
"grad_norm": 11.745617866516113,
|
| 1166 |
+
"learning_rate": 2.5130694548170274e-05,
|
| 1167 |
+
"loss": 0.6128,
|
| 1168 |
+
"step": 1630
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 2.05,
|
| 1172 |
+
"grad_norm": 13.023053169250488,
|
| 1173 |
+
"learning_rate": 2.4943988050784166e-05,
|
| 1174 |
+
"loss": 0.6555,
|
| 1175 |
+
"step": 1640
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 2.05,
|
| 1179 |
+
"grad_norm": 6.082352161407471,
|
| 1180 |
+
"learning_rate": 2.475728155339806e-05,
|
| 1181 |
+
"loss": 0.5754,
|
| 1182 |
+
"step": 1650
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.06,
|
| 1186 |
+
"grad_norm": 4.05959415435791,
|
| 1187 |
+
"learning_rate": 2.4570575056011947e-05,
|
| 1188 |
+
"loss": 0.4325,
|
| 1189 |
+
"step": 1660
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 2.06,
|
| 1193 |
+
"grad_norm": 14.252350807189941,
|
| 1194 |
+
"learning_rate": 2.4383868558625843e-05,
|
| 1195 |
+
"loss": 0.4956,
|
| 1196 |
+
"step": 1670
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.06,
|
| 1200 |
+
"grad_norm": 6.615436553955078,
|
| 1201 |
+
"learning_rate": 2.4197162061239732e-05,
|
| 1202 |
+
"loss": 0.5474,
|
| 1203 |
+
"step": 1680
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.07,
|
| 1207 |
+
"grad_norm": 7.025275230407715,
|
| 1208 |
+
"learning_rate": 2.4010455563853624e-05,
|
| 1209 |
+
"loss": 0.6072,
|
| 1210 |
+
"step": 1690
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 2.07,
|
| 1214 |
+
"grad_norm": 15.101826667785645,
|
| 1215 |
+
"learning_rate": 2.3823749066467517e-05,
|
| 1216 |
+
"loss": 0.6705,
|
| 1217 |
+
"step": 1700
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 2.07,
|
| 1221 |
+
"grad_norm": 11.996399879455566,
|
| 1222 |
+
"learning_rate": 2.3637042569081406e-05,
|
| 1223 |
+
"loss": 0.5326,
|
| 1224 |
+
"step": 1710
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 2.08,
|
| 1228 |
+
"grad_norm": 11.746566772460938,
|
| 1229 |
+
"learning_rate": 2.3450336071695298e-05,
|
| 1230 |
+
"loss": 0.6512,
|
| 1231 |
+
"step": 1720
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 2.08,
|
| 1235 |
+
"grad_norm": 7.299741744995117,
|
| 1236 |
+
"learning_rate": 2.3263629574309187e-05,
|
| 1237 |
+
"loss": 0.4751,
|
| 1238 |
+
"step": 1730
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.08,
|
| 1242 |
+
"grad_norm": 10.497876167297363,
|
| 1243 |
+
"learning_rate": 2.307692307692308e-05,
|
| 1244 |
+
"loss": 0.7083,
|
| 1245 |
+
"step": 1740
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 2.09,
|
| 1249 |
+
"grad_norm": 6.633430480957031,
|
| 1250 |
+
"learning_rate": 2.2890216579536968e-05,
|
| 1251 |
+
"loss": 0.4854,
|
| 1252 |
+
"step": 1750
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 2.09,
|
| 1256 |
+
"grad_norm": 8.004325866699219,
|
| 1257 |
+
"learning_rate": 2.270351008215086e-05,
|
| 1258 |
+
"loss": 0.5795,
|
| 1259 |
+
"step": 1760
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 2.09,
|
| 1263 |
+
"grad_norm": 9.713132858276367,
|
| 1264 |
+
"learning_rate": 2.251680358476475e-05,
|
| 1265 |
+
"loss": 0.5735,
|
| 1266 |
+
"step": 1770
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 2.1,
|
| 1270 |
+
"grad_norm": 6.221426963806152,
|
| 1271 |
+
"learning_rate": 2.233009708737864e-05,
|
| 1272 |
+
"loss": 0.5405,
|
| 1273 |
+
"step": 1780
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 2.1,
|
| 1277 |
+
"grad_norm": 6.872443675994873,
|
| 1278 |
+
"learning_rate": 2.2143390589992534e-05,
|
| 1279 |
+
"loss": 0.7661,
|
| 1280 |
+
"step": 1790
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 2.1,
|
| 1284 |
+
"grad_norm": 6.894876956939697,
|
| 1285 |
+
"learning_rate": 2.1956684092606426e-05,
|
| 1286 |
+
"loss": 0.4521,
|
| 1287 |
+
"step": 1800
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 2.11,
|
| 1291 |
+
"grad_norm": 3.6292059421539307,
|
| 1292 |
+
"learning_rate": 2.1769977595220315e-05,
|
| 1293 |
+
"loss": 0.5084,
|
| 1294 |
+
"step": 1810
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 2.11,
|
| 1298 |
+
"grad_norm": 13.099087715148926,
|
| 1299 |
+
"learning_rate": 2.1583271097834207e-05,
|
| 1300 |
+
"loss": 0.644,
|
| 1301 |
+
"step": 1820
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 2.11,
|
| 1305 |
+
"grad_norm": 13.508889198303223,
|
| 1306 |
+
"learning_rate": 2.1396564600448096e-05,
|
| 1307 |
+
"loss": 0.6271,
|
| 1308 |
+
"step": 1830
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 2.12,
|
| 1312 |
+
"grad_norm": 9.162331581115723,
|
| 1313 |
+
"learning_rate": 2.120985810306199e-05,
|
| 1314 |
+
"loss": 0.5586,
|
| 1315 |
+
"step": 1840
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 2.12,
|
| 1319 |
+
"grad_norm": 9.340535163879395,
|
| 1320 |
+
"learning_rate": 2.1023151605675877e-05,
|
| 1321 |
+
"loss": 0.6222,
|
| 1322 |
+
"step": 1850
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 2.12,
|
| 1326 |
+
"grad_norm": 10.59788990020752,
|
| 1327 |
+
"learning_rate": 2.083644510828977e-05,
|
| 1328 |
+
"loss": 0.6439,
|
| 1329 |
+
"step": 1860
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 2.13,
|
| 1333 |
+
"grad_norm": 13.132092475891113,
|
| 1334 |
+
"learning_rate": 2.064973861090366e-05,
|
| 1335 |
+
"loss": 0.5401,
|
| 1336 |
+
"step": 1870
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 2.13,
|
| 1340 |
+
"grad_norm": 11.530851364135742,
|
| 1341 |
+
"learning_rate": 2.046303211351755e-05,
|
| 1342 |
+
"loss": 0.6462,
|
| 1343 |
+
"step": 1880
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 2.13,
|
| 1347 |
+
"grad_norm": 15.348837852478027,
|
| 1348 |
+
"learning_rate": 2.0276325616131443e-05,
|
| 1349 |
+
"loss": 0.6606,
|
| 1350 |
+
"step": 1890
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 2.14,
|
| 1354 |
+
"grad_norm": 16.934356689453125,
|
| 1355 |
+
"learning_rate": 2.0089619118745336e-05,
|
| 1356 |
+
"loss": 0.6352,
|
| 1357 |
+
"step": 1900
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 2.14,
|
| 1361 |
+
"grad_norm": 9.712738037109375,
|
| 1362 |
+
"learning_rate": 1.9902912621359225e-05,
|
| 1363 |
+
"loss": 0.8095,
|
| 1364 |
+
"step": 1910
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 2.14,
|
| 1368 |
+
"grad_norm": 10.83343505859375,
|
| 1369 |
+
"learning_rate": 1.9716206123973117e-05,
|
| 1370 |
+
"loss": 0.5243,
|
| 1371 |
+
"step": 1920
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 2.15,
|
| 1375 |
+
"grad_norm": 10.138250350952148,
|
| 1376 |
+
"learning_rate": 1.9529499626587006e-05,
|
| 1377 |
+
"loss": 0.4842,
|
| 1378 |
+
"step": 1930
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 2.15,
|
| 1382 |
+
"grad_norm": 9.77036190032959,
|
| 1383 |
+
"learning_rate": 1.9342793129200898e-05,
|
| 1384 |
+
"loss": 0.6791,
|
| 1385 |
+
"step": 1940
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 2.15,
|
| 1389 |
+
"grad_norm": 9.89406967163086,
|
| 1390 |
+
"learning_rate": 1.9156086631814787e-05,
|
| 1391 |
+
"loss": 0.5933,
|
| 1392 |
+
"step": 1950
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 2.16,
|
| 1396 |
+
"grad_norm": 7.667896747589111,
|
| 1397 |
+
"learning_rate": 1.896938013442868e-05,
|
| 1398 |
+
"loss": 0.558,
|
| 1399 |
+
"step": 1960
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 2.16,
|
| 1403 |
+
"grad_norm": 21.039913177490234,
|
| 1404 |
+
"learning_rate": 1.8782673637042568e-05,
|
| 1405 |
+
"loss": 0.5028,
|
| 1406 |
+
"step": 1970
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 2.16,
|
| 1410 |
+
"grad_norm": 3.437727689743042,
|
| 1411 |
+
"learning_rate": 1.859596713965646e-05,
|
| 1412 |
+
"loss": 0.4429,
|
| 1413 |
+
"step": 1980
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 2.17,
|
| 1417 |
+
"grad_norm": 12.366796493530273,
|
| 1418 |
+
"learning_rate": 1.8409260642270353e-05,
|
| 1419 |
+
"loss": 0.6008,
|
| 1420 |
+
"step": 1990
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 2.17,
|
| 1424 |
+
"grad_norm": 17.259078979492188,
|
| 1425 |
+
"learning_rate": 1.8222554144884245e-05,
|
| 1426 |
+
"loss": 0.5279,
|
| 1427 |
+
"step": 2000
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 2.17,
|
| 1431 |
+
"grad_norm": 8.58683967590332,
|
| 1432 |
+
"learning_rate": 1.8035847647498134e-05,
|
| 1433 |
+
"loss": 0.4038,
|
| 1434 |
+
"step": 2010
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 2.18,
|
| 1438 |
+
"grad_norm": 6.0911970138549805,
|
| 1439 |
+
"learning_rate": 1.7849141150112026e-05,
|
| 1440 |
+
"loss": 0.5495,
|
| 1441 |
+
"step": 2020
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 2.18,
|
| 1445 |
+
"grad_norm": 9.944643020629883,
|
| 1446 |
+
"learning_rate": 1.7662434652725915e-05,
|
| 1447 |
+
"loss": 0.5382,
|
| 1448 |
+
"step": 2030
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 2.18,
|
| 1452 |
+
"grad_norm": 13.859691619873047,
|
| 1453 |
+
"learning_rate": 1.7475728155339808e-05,
|
| 1454 |
+
"loss": 0.6927,
|
| 1455 |
+
"step": 2040
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 2.19,
|
| 1459 |
+
"grad_norm": 7.948896408081055,
|
| 1460 |
+
"learning_rate": 1.7289021657953697e-05,
|
| 1461 |
+
"loss": 0.3874,
|
| 1462 |
+
"step": 2050
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 2.19,
|
| 1466 |
+
"grad_norm": 6.207942485809326,
|
| 1467 |
+
"learning_rate": 1.710231516056759e-05,
|
| 1468 |
+
"loss": 0.4023,
|
| 1469 |
+
"step": 2060
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 2.19,
|
| 1473 |
+
"grad_norm": 5.350401401519775,
|
| 1474 |
+
"learning_rate": 1.6915608663181478e-05,
|
| 1475 |
+
"loss": 0.5397,
|
| 1476 |
+
"step": 2070
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 2.2,
|
| 1480 |
+
"grad_norm": 10.37016487121582,
|
| 1481 |
+
"learning_rate": 1.672890216579537e-05,
|
| 1482 |
+
"loss": 0.5941,
|
| 1483 |
+
"step": 2080
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 2.2,
|
| 1487 |
+
"grad_norm": 9.93877124786377,
|
| 1488 |
+
"learning_rate": 1.654219566840926e-05,
|
| 1489 |
+
"loss": 0.4433,
|
| 1490 |
+
"step": 2090
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 2.2,
|
| 1494 |
+
"grad_norm": 14.147184371948242,
|
| 1495 |
+
"learning_rate": 1.635548917102315e-05,
|
| 1496 |
+
"loss": 0.6687,
|
| 1497 |
+
"step": 2100
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 2.21,
|
| 1501 |
+
"grad_norm": 2.5294337272644043,
|
| 1502 |
+
"learning_rate": 1.6168782673637044e-05,
|
| 1503 |
+
"loss": 0.5144,
|
| 1504 |
+
"step": 2110
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 2.21,
|
| 1508 |
+
"grad_norm": 14.119583129882812,
|
| 1509 |
+
"learning_rate": 1.5982076176250936e-05,
|
| 1510 |
+
"loss": 0.4195,
|
| 1511 |
+
"step": 2120
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 2.22,
|
| 1515 |
+
"grad_norm": 5.246395587921143,
|
| 1516 |
+
"learning_rate": 1.5795369678864825e-05,
|
| 1517 |
+
"loss": 0.5291,
|
| 1518 |
+
"step": 2130
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 2.22,
|
| 1522 |
+
"grad_norm": 5.959507465362549,
|
| 1523 |
+
"learning_rate": 1.5608663181478717e-05,
|
| 1524 |
+
"loss": 0.4513,
|
| 1525 |
+
"step": 2140
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 2.22,
|
| 1529 |
+
"grad_norm": 12.30712604522705,
|
| 1530 |
+
"learning_rate": 1.5421956684092606e-05,
|
| 1531 |
+
"loss": 0.5737,
|
| 1532 |
+
"step": 2150
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 2.23,
|
| 1536 |
+
"grad_norm": 13.365155220031738,
|
| 1537 |
+
"learning_rate": 1.5235250186706498e-05,
|
| 1538 |
+
"loss": 0.6143,
|
| 1539 |
+
"step": 2160
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 2.23,
|
| 1543 |
+
"grad_norm": 15.408020973205566,
|
| 1544 |
+
"learning_rate": 1.5048543689320387e-05,
|
| 1545 |
+
"loss": 0.4525,
|
| 1546 |
+
"step": 2170
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 2.23,
|
| 1550 |
+
"grad_norm": 8.355168342590332,
|
| 1551 |
+
"learning_rate": 1.486183719193428e-05,
|
| 1552 |
+
"loss": 0.6192,
|
| 1553 |
+
"step": 2180
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 2.24,
|
| 1557 |
+
"grad_norm": 14.046072006225586,
|
| 1558 |
+
"learning_rate": 1.467513069454817e-05,
|
| 1559 |
+
"loss": 0.5334,
|
| 1560 |
+
"step": 2190
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 2.24,
|
| 1564 |
+
"grad_norm": 6.0648908615112305,
|
| 1565 |
+
"learning_rate": 1.4488424197162062e-05,
|
| 1566 |
+
"loss": 0.6275,
|
| 1567 |
+
"step": 2200
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 2.24,
|
| 1571 |
+
"grad_norm": 9.870409965515137,
|
| 1572 |
+
"learning_rate": 1.4301717699775951e-05,
|
| 1573 |
+
"loss": 0.5459,
|
| 1574 |
+
"step": 2210
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 2.25,
|
| 1578 |
+
"grad_norm": 10.847134590148926,
|
| 1579 |
+
"learning_rate": 1.4115011202389844e-05,
|
| 1580 |
+
"loss": 0.3645,
|
| 1581 |
+
"step": 2220
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 2.25,
|
| 1585 |
+
"grad_norm": 7.336729049682617,
|
| 1586 |
+
"learning_rate": 1.3928304705003734e-05,
|
| 1587 |
+
"loss": 0.4575,
|
| 1588 |
+
"step": 2230
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 2.25,
|
| 1592 |
+
"eval_accuracy": 0.7181208053691275,
|
| 1593 |
+
"eval_loss": 0.7427847981452942,
|
| 1594 |
+
"eval_runtime": 1683.7503,
|
| 1595 |
+
"eval_samples_per_second": 0.442,
|
| 1596 |
+
"eval_steps_per_second": 0.056,
|
| 1597 |
+
"step": 2235
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 3.0,
|
| 1601 |
+
"grad_norm": 8.262899398803711,
|
| 1602 |
+
"learning_rate": 1.3741598207617627e-05,
|
| 1603 |
+
"loss": 0.6098,
|
| 1604 |
+
"step": 2240
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 3.01,
|
| 1608 |
+
"grad_norm": 11.888318061828613,
|
| 1609 |
+
"learning_rate": 1.3554891710231516e-05,
|
| 1610 |
+
"loss": 0.6722,
|
| 1611 |
+
"step": 2250
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 3.01,
|
| 1615 |
+
"grad_norm": 4.9109320640563965,
|
| 1616 |
+
"learning_rate": 1.3368185212845408e-05,
|
| 1617 |
+
"loss": 0.302,
|
| 1618 |
+
"step": 2260
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 3.01,
|
| 1622 |
+
"grad_norm": 3.660041570663452,
|
| 1623 |
+
"learning_rate": 1.3181478715459297e-05,
|
| 1624 |
+
"loss": 0.4001,
|
| 1625 |
+
"step": 2270
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 3.02,
|
| 1629 |
+
"grad_norm": 5.666093349456787,
|
| 1630 |
+
"learning_rate": 1.2994772218073189e-05,
|
| 1631 |
+
"loss": 0.4406,
|
| 1632 |
+
"step": 2280
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 3.02,
|
| 1636 |
+
"grad_norm": 5.51093053817749,
|
| 1637 |
+
"learning_rate": 1.2808065720687081e-05,
|
| 1638 |
+
"loss": 0.434,
|
| 1639 |
+
"step": 2290
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 3.02,
|
| 1643 |
+
"grad_norm": 11.179670333862305,
|
| 1644 |
+
"learning_rate": 1.2621359223300972e-05,
|
| 1645 |
+
"loss": 0.4668,
|
| 1646 |
+
"step": 2300
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 3.03,
|
| 1650 |
+
"grad_norm": 8.03001880645752,
|
| 1651 |
+
"learning_rate": 1.2434652725914863e-05,
|
| 1652 |
+
"loss": 0.4569,
|
| 1653 |
+
"step": 2310
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 3.03,
|
| 1657 |
+
"grad_norm": 8.949047088623047,
|
| 1658 |
+
"learning_rate": 1.2247946228528753e-05,
|
| 1659 |
+
"loss": 0.4284,
|
| 1660 |
+
"step": 2320
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 3.03,
|
| 1664 |
+
"grad_norm": 10.540315628051758,
|
| 1665 |
+
"learning_rate": 1.2061239731142644e-05,
|
| 1666 |
+
"loss": 0.5279,
|
| 1667 |
+
"step": 2330
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 3.04,
|
| 1671 |
+
"grad_norm": 14.215402603149414,
|
| 1672 |
+
"learning_rate": 1.1874533233756534e-05,
|
| 1673 |
+
"loss": 0.4004,
|
| 1674 |
+
"step": 2340
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 3.04,
|
| 1678 |
+
"grad_norm": 18.047544479370117,
|
| 1679 |
+
"learning_rate": 1.1687826736370427e-05,
|
| 1680 |
+
"loss": 0.4778,
|
| 1681 |
+
"step": 2350
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 3.04,
|
| 1685 |
+
"grad_norm": 9.015625953674316,
|
| 1686 |
+
"learning_rate": 1.1501120238984317e-05,
|
| 1687 |
+
"loss": 0.5168,
|
| 1688 |
+
"step": 2360
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 3.05,
|
| 1692 |
+
"grad_norm": 11.635021209716797,
|
| 1693 |
+
"learning_rate": 1.1314413741598208e-05,
|
| 1694 |
+
"loss": 0.4268,
|
| 1695 |
+
"step": 2370
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 3.05,
|
| 1699 |
+
"grad_norm": 7.7300004959106445,
|
| 1700 |
+
"learning_rate": 1.1127707244212099e-05,
|
| 1701 |
+
"loss": 0.4186,
|
| 1702 |
+
"step": 2380
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 3.05,
|
| 1706 |
+
"grad_norm": 8.837210655212402,
|
| 1707 |
+
"learning_rate": 1.094100074682599e-05,
|
| 1708 |
+
"loss": 0.555,
|
| 1709 |
+
"step": 2390
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 3.06,
|
| 1713 |
+
"grad_norm": 6.158858299255371,
|
| 1714 |
+
"learning_rate": 1.0754294249439881e-05,
|
| 1715 |
+
"loss": 0.5783,
|
| 1716 |
+
"step": 2400
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 3.06,
|
| 1720 |
+
"grad_norm": 7.499691009521484,
|
| 1721 |
+
"learning_rate": 1.0567587752053772e-05,
|
| 1722 |
+
"loss": 0.4923,
|
| 1723 |
+
"step": 2410
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 3.06,
|
| 1727 |
+
"grad_norm": 12.26675796508789,
|
| 1728 |
+
"learning_rate": 1.0380881254667663e-05,
|
| 1729 |
+
"loss": 0.3842,
|
| 1730 |
+
"step": 2420
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 3.07,
|
| 1734 |
+
"grad_norm": 11.513700485229492,
|
| 1735 |
+
"learning_rate": 1.0194174757281553e-05,
|
| 1736 |
+
"loss": 0.3955,
|
| 1737 |
+
"step": 2430
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 3.07,
|
| 1741 |
+
"grad_norm": 5.970608234405518,
|
| 1742 |
+
"learning_rate": 1.0007468259895444e-05,
|
| 1743 |
+
"loss": 0.3256,
|
| 1744 |
+
"step": 2440
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 3.07,
|
| 1748 |
+
"grad_norm": 10.886767387390137,
|
| 1749 |
+
"learning_rate": 9.820761762509336e-06,
|
| 1750 |
+
"loss": 0.3808,
|
| 1751 |
+
"step": 2450
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 3.08,
|
| 1755 |
+
"grad_norm": 5.535740375518799,
|
| 1756 |
+
"learning_rate": 9.634055265123227e-06,
|
| 1757 |
+
"loss": 0.4004,
|
| 1758 |
+
"step": 2460
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 3.08,
|
| 1762 |
+
"grad_norm": 7.606982231140137,
|
| 1763 |
+
"learning_rate": 9.447348767737117e-06,
|
| 1764 |
+
"loss": 0.492,
|
| 1765 |
+
"step": 2470
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 3.08,
|
| 1769 |
+
"grad_norm": 4.145596981048584,
|
| 1770 |
+
"learning_rate": 9.260642270351008e-06,
|
| 1771 |
+
"loss": 0.3662,
|
| 1772 |
+
"step": 2480
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 3.09,
|
| 1776 |
+
"grad_norm": 5.0407490730285645,
|
| 1777 |
+
"learning_rate": 9.073935772964899e-06,
|
| 1778 |
+
"loss": 0.2492,
|
| 1779 |
+
"step": 2490
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 3.09,
|
| 1783 |
+
"grad_norm": 7.193727016448975,
|
| 1784 |
+
"learning_rate": 8.88722927557879e-06,
|
| 1785 |
+
"loss": 0.3642,
|
| 1786 |
+
"step": 2500
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 3.09,
|
| 1790 |
+
"grad_norm": 6.277304172515869,
|
| 1791 |
+
"learning_rate": 8.700522778192682e-06,
|
| 1792 |
+
"loss": 0.4605,
|
| 1793 |
+
"step": 2510
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 3.1,
|
| 1797 |
+
"grad_norm": 10.555391311645508,
|
| 1798 |
+
"learning_rate": 8.513816280806572e-06,
|
| 1799 |
+
"loss": 0.3938,
|
| 1800 |
+
"step": 2520
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 3.1,
|
| 1804 |
+
"grad_norm": 5.211946964263916,
|
| 1805 |
+
"learning_rate": 8.327109783420463e-06,
|
| 1806 |
+
"loss": 0.3817,
|
| 1807 |
+
"step": 2530
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 3.1,
|
| 1811 |
+
"grad_norm": 11.565357208251953,
|
| 1812 |
+
"learning_rate": 8.140403286034353e-06,
|
| 1813 |
+
"loss": 0.3402,
|
| 1814 |
+
"step": 2540
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 3.11,
|
| 1818 |
+
"grad_norm": 7.738372802734375,
|
| 1819 |
+
"learning_rate": 7.953696788648244e-06,
|
| 1820 |
+
"loss": 0.555,
|
| 1821 |
+
"step": 2550
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 3.11,
|
| 1825 |
+
"grad_norm": 5.373858451843262,
|
| 1826 |
+
"learning_rate": 7.766990291262136e-06,
|
| 1827 |
+
"loss": 0.4478,
|
| 1828 |
+
"step": 2560
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 3.11,
|
| 1832 |
+
"grad_norm": 11.583991050720215,
|
| 1833 |
+
"learning_rate": 7.580283793876028e-06,
|
| 1834 |
+
"loss": 0.3699,
|
| 1835 |
+
"step": 2570
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 3.12,
|
| 1839 |
+
"grad_norm": 7.290119171142578,
|
| 1840 |
+
"learning_rate": 7.393577296489919e-06,
|
| 1841 |
+
"loss": 0.4119,
|
| 1842 |
+
"step": 2580
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 3.12,
|
| 1846 |
+
"grad_norm": 13.199901580810547,
|
| 1847 |
+
"learning_rate": 7.20687079910381e-06,
|
| 1848 |
+
"loss": 0.3973,
|
| 1849 |
+
"step": 2590
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 3.12,
|
| 1853 |
+
"grad_norm": 5.801746368408203,
|
| 1854 |
+
"learning_rate": 7.0201643017177005e-06,
|
| 1855 |
+
"loss": 0.278,
|
| 1856 |
+
"step": 2600
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 3.13,
|
| 1860 |
+
"grad_norm": 7.40468692779541,
|
| 1861 |
+
"learning_rate": 6.833457804331592e-06,
|
| 1862 |
+
"loss": 0.4457,
|
| 1863 |
+
"step": 2610
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 3.13,
|
| 1867 |
+
"grad_norm": 7.056453227996826,
|
| 1868 |
+
"learning_rate": 6.6467513069454825e-06,
|
| 1869 |
+
"loss": 0.3933,
|
| 1870 |
+
"step": 2620
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 3.13,
|
| 1874 |
+
"grad_norm": 1.5562771558761597,
|
| 1875 |
+
"learning_rate": 6.460044809559373e-06,
|
| 1876 |
+
"loss": 0.3659,
|
| 1877 |
+
"step": 2630
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 3.14,
|
| 1881 |
+
"grad_norm": 12.124537467956543,
|
| 1882 |
+
"learning_rate": 6.273338312173265e-06,
|
| 1883 |
+
"loss": 0.417,
|
| 1884 |
+
"step": 2640
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 3.14,
|
| 1888 |
+
"grad_norm": 8.535222053527832,
|
| 1889 |
+
"learning_rate": 6.086631814787154e-06,
|
| 1890 |
+
"loss": 0.4438,
|
| 1891 |
+
"step": 2650
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 3.14,
|
| 1895 |
+
"grad_norm": 14.003257751464844,
|
| 1896 |
+
"learning_rate": 5.899925317401046e-06,
|
| 1897 |
+
"loss": 0.4157,
|
| 1898 |
+
"step": 2660
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 3.15,
|
| 1902 |
+
"grad_norm": 8.428786277770996,
|
| 1903 |
+
"learning_rate": 5.7132188200149364e-06,
|
| 1904 |
+
"loss": 0.3332,
|
| 1905 |
+
"step": 2670
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 3.15,
|
| 1909 |
+
"grad_norm": 7.349361896514893,
|
| 1910 |
+
"learning_rate": 5.526512322628828e-06,
|
| 1911 |
+
"loss": 0.354,
|
| 1912 |
+
"step": 2680
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 3.15,
|
| 1916 |
+
"grad_norm": 9.289456367492676,
|
| 1917 |
+
"learning_rate": 5.3398058252427185e-06,
|
| 1918 |
+
"loss": 0.4407,
|
| 1919 |
+
"step": 2690
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 3.16,
|
| 1923 |
+
"grad_norm": 6.27097225189209,
|
| 1924 |
+
"learning_rate": 5.153099327856609e-06,
|
| 1925 |
+
"loss": 0.2596,
|
| 1926 |
+
"step": 2700
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 3.16,
|
| 1930 |
+
"grad_norm": 15.719304084777832,
|
| 1931 |
+
"learning_rate": 4.966392830470501e-06,
|
| 1932 |
+
"loss": 0.5207,
|
| 1933 |
+
"step": 2710
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 3.16,
|
| 1937 |
+
"grad_norm": 7.534648895263672,
|
| 1938 |
+
"learning_rate": 4.779686333084392e-06,
|
| 1939 |
+
"loss": 0.3938,
|
| 1940 |
+
"step": 2720
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 3.17,
|
| 1944 |
+
"grad_norm": 8.981090545654297,
|
| 1945 |
+
"learning_rate": 4.592979835698283e-06,
|
| 1946 |
+
"loss": 0.248,
|
| 1947 |
+
"step": 2730
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 3.17,
|
| 1951 |
+
"grad_norm": 11.375850677490234,
|
| 1952 |
+
"learning_rate": 4.406273338312174e-06,
|
| 1953 |
+
"loss": 0.2799,
|
| 1954 |
+
"step": 2740
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 3.17,
|
| 1958 |
+
"grad_norm": 12.491207122802734,
|
| 1959 |
+
"learning_rate": 4.219566840926065e-06,
|
| 1960 |
+
"loss": 0.4839,
|
| 1961 |
+
"step": 2750
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 3.18,
|
| 1965 |
+
"grad_norm": 5.428218364715576,
|
| 1966 |
+
"learning_rate": 4.032860343539955e-06,
|
| 1967 |
+
"loss": 0.4289,
|
| 1968 |
+
"step": 2760
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 3.18,
|
| 1972 |
+
"grad_norm": 7.776615142822266,
|
| 1973 |
+
"learning_rate": 3.846153846153847e-06,
|
| 1974 |
+
"loss": 0.4563,
|
| 1975 |
+
"step": 2770
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 3.18,
|
| 1979 |
+
"grad_norm": 4.943689823150635,
|
| 1980 |
+
"learning_rate": 3.6594473487677374e-06,
|
| 1981 |
+
"loss": 0.3819,
|
| 1982 |
+
"step": 2780
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 3.19,
|
| 1986 |
+
"grad_norm": 10.917070388793945,
|
| 1987 |
+
"learning_rate": 3.4727408513816284e-06,
|
| 1988 |
+
"loss": 0.4038,
|
| 1989 |
+
"step": 2790
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 3.19,
|
| 1993 |
+
"grad_norm": 7.394994258880615,
|
| 1994 |
+
"learning_rate": 3.2860343539955195e-06,
|
| 1995 |
+
"loss": 0.4088,
|
| 1996 |
+
"step": 2800
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 3.19,
|
| 2000 |
+
"grad_norm": 6.575070381164551,
|
| 2001 |
+
"learning_rate": 3.09932785660941e-06,
|
| 2002 |
+
"loss": 0.6557,
|
| 2003 |
+
"step": 2810
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 3.2,
|
| 2007 |
+
"grad_norm": 14.510526657104492,
|
| 2008 |
+
"learning_rate": 2.912621359223301e-06,
|
| 2009 |
+
"loss": 0.2805,
|
| 2010 |
+
"step": 2820
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 3.2,
|
| 2014 |
+
"grad_norm": 12.12006950378418,
|
| 2015 |
+
"learning_rate": 2.725914861837192e-06,
|
| 2016 |
+
"loss": 0.4065,
|
| 2017 |
+
"step": 2830
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 3.2,
|
| 2021 |
+
"grad_norm": 5.202164173126221,
|
| 2022 |
+
"learning_rate": 2.539208364451083e-06,
|
| 2023 |
+
"loss": 0.3511,
|
| 2024 |
+
"step": 2840
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 3.21,
|
| 2028 |
+
"grad_norm": 13.489916801452637,
|
| 2029 |
+
"learning_rate": 2.352501867064974e-06,
|
| 2030 |
+
"loss": 0.3357,
|
| 2031 |
+
"step": 2850
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 3.21,
|
| 2035 |
+
"grad_norm": 12.779403686523438,
|
| 2036 |
+
"learning_rate": 2.165795369678865e-06,
|
| 2037 |
+
"loss": 0.367,
|
| 2038 |
+
"step": 2860
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 3.21,
|
| 2042 |
+
"grad_norm": 13.837797164916992,
|
| 2043 |
+
"learning_rate": 1.979088872292756e-06,
|
| 2044 |
+
"loss": 0.4565,
|
| 2045 |
+
"step": 2870
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 3.22,
|
| 2049 |
+
"grad_norm": 4.743872165679932,
|
| 2050 |
+
"learning_rate": 1.7923823749066467e-06,
|
| 2051 |
+
"loss": 0.3372,
|
| 2052 |
+
"step": 2880
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 3.22,
|
| 2056 |
+
"grad_norm": 8.704170227050781,
|
| 2057 |
+
"learning_rate": 1.6056758775205377e-06,
|
| 2058 |
+
"loss": 0.3423,
|
| 2059 |
+
"step": 2890
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 3.22,
|
| 2063 |
+
"grad_norm": 2.67067551612854,
|
| 2064 |
+
"learning_rate": 1.4189693801344288e-06,
|
| 2065 |
+
"loss": 0.3444,
|
| 2066 |
+
"step": 2900
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 3.23,
|
| 2070 |
+
"grad_norm": 11.893070220947266,
|
| 2071 |
+
"learning_rate": 1.2322628827483198e-06,
|
| 2072 |
+
"loss": 0.2581,
|
| 2073 |
+
"step": 2910
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 3.23,
|
| 2077 |
+
"grad_norm": 4.046337127685547,
|
| 2078 |
+
"learning_rate": 1.0455563853622106e-06,
|
| 2079 |
+
"loss": 0.3214,
|
| 2080 |
+
"step": 2920
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 3.23,
|
| 2084 |
+
"grad_norm": 11.184314727783203,
|
| 2085 |
+
"learning_rate": 8.588498879761017e-07,
|
| 2086 |
+
"loss": 0.3076,
|
| 2087 |
+
"step": 2930
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 3.24,
|
| 2091 |
+
"grad_norm": 12.178153991699219,
|
| 2092 |
+
"learning_rate": 6.721433905899926e-07,
|
| 2093 |
+
"loss": 0.2403,
|
| 2094 |
+
"step": 2940
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 3.24,
|
| 2098 |
+
"grad_norm": 1.377287745475769,
|
| 2099 |
+
"learning_rate": 4.854368932038835e-07,
|
| 2100 |
+
"loss": 0.2376,
|
| 2101 |
+
"step": 2950
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 3.24,
|
| 2105 |
+
"grad_norm": 14.243566513061523,
|
| 2106 |
+
"learning_rate": 2.987303958177745e-07,
|
| 2107 |
+
"loss": 0.3316,
|
| 2108 |
+
"step": 2960
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 3.25,
|
| 2112 |
+
"grad_norm": 4.011444091796875,
|
| 2113 |
+
"learning_rate": 1.1202389843166542e-07,
|
| 2114 |
+
"loss": 0.3262,
|
| 2115 |
+
"step": 2970
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 3.25,
|
| 2119 |
+
"eval_accuracy": 0.7328859060402685,
|
| 2120 |
+
"eval_loss": 0.7344102263450623,
|
| 2121 |
+
"eval_runtime": 1687.1353,
|
| 2122 |
+
"eval_samples_per_second": 0.442,
|
| 2123 |
+
"eval_steps_per_second": 0.056,
|
| 2124 |
+
"step": 2976
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 3.25,
|
| 2128 |
+
"step": 2976,
|
| 2129 |
+
"total_flos": 6.0945535892058145e+19,
|
| 2130 |
+
"train_loss": 0.7349759022234589,
|
| 2131 |
+
"train_runtime": 150955.4956,
|
| 2132 |
+
"train_samples_per_second": 0.158,
|
| 2133 |
+
"train_steps_per_second": 0.02
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 3.25,
|
| 2137 |
+
"eval_accuracy": 0.7043010752688172,
|
| 2138 |
+
"eval_loss": 0.8327615261077881,
|
| 2139 |
+
"eval_runtime": 1693.8047,
|
| 2140 |
+
"eval_samples_per_second": 0.439,
|
| 2141 |
+
"eval_steps_per_second": 0.055,
|
| 2142 |
+
"step": 2976
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 3.25,
|
| 2146 |
+
"eval_accuracy": 0.7043010752688172,
|
| 2147 |
+
"eval_loss": 0.8327615261077881,
|
| 2148 |
+
"eval_runtime": 1683.6481,
|
| 2149 |
+
"eval_samples_per_second": 0.442,
|
| 2150 |
+
"eval_steps_per_second": 0.055,
|
| 2151 |
+
"step": 2976
|
| 2152 |
+
}
|
| 2153 |
+
],
|
| 2154 |
+
"logging_steps": 10,
|
| 2155 |
+
"max_steps": 2976,
|
| 2156 |
+
"num_input_tokens_seen": 0,
|
| 2157 |
+
"num_train_epochs": 9223372036854775807,
|
| 2158 |
+
"save_steps": 500,
|
| 2159 |
+
"total_flos": 6.0945535892058145e+19,
|
| 2160 |
+
"train_batch_size": 8,
|
| 2161 |
+
"trial_name": null,
|
| 2162 |
+
"trial_params": null
|
| 2163 |
+
}
|