End of training
Browse files- all_results.json +8 -0
- test_results.json +8 -0
- trainer_state.json +1901 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.16,
|
| 3 |
+
"eval_accuracy": 0.9895833333333334,
|
| 4 |
+
"eval_loss": 0.11018910259008408,
|
| 5 |
+
"eval_runtime": 269.2698,
|
| 6 |
+
"eval_samples_per_second": 0.357,
|
| 7 |
+
"eval_steps_per_second": 0.178
|
| 8 |
+
}
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.16,
|
| 3 |
+
"eval_accuracy": 0.9895833333333334,
|
| 4 |
+
"eval_loss": 0.11018910259008408,
|
| 5 |
+
"eval_runtime": 269.2698,
|
| 6 |
+
"eval_samples_per_second": 0.357,
|
| 7 |
+
"eval_steps_per_second": 0.178
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1901 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9421052631578948,
|
| 3 |
+
"best_model_checkpoint": "google/vivit-b-16x2-kinetics400-finetuned-my-dataset-6-epochs\\checkpoint-2574",
|
| 4 |
+
"epoch": 5.164724164724165,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2574,
|
| 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": 19.156469345092773,
|
| 14 |
+
"learning_rate": 1.937984496124031e-06,
|
| 15 |
+
"loss": 1.6746,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01,
|
| 20 |
+
"grad_norm": 20.032257080078125,
|
| 21 |
+
"learning_rate": 3.875968992248062e-06,
|
| 22 |
+
"loss": 1.8523,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.01,
|
| 27 |
+
"grad_norm": 19.687211990356445,
|
| 28 |
+
"learning_rate": 5.8139534883720935e-06,
|
| 29 |
+
"loss": 1.7049,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.02,
|
| 34 |
+
"grad_norm": 21.64842414855957,
|
| 35 |
+
"learning_rate": 7.751937984496124e-06,
|
| 36 |
+
"loss": 1.8233,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.02,
|
| 41 |
+
"grad_norm": 18.463632583618164,
|
| 42 |
+
"learning_rate": 9.689922480620156e-06,
|
| 43 |
+
"loss": 1.5469,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02,
|
| 48 |
+
"grad_norm": 21.310239791870117,
|
| 49 |
+
"learning_rate": 1.1627906976744187e-05,
|
| 50 |
+
"loss": 1.8318,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.03,
|
| 55 |
+
"grad_norm": 16.913654327392578,
|
| 56 |
+
"learning_rate": 1.3565891472868217e-05,
|
| 57 |
+
"loss": 1.3699,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.03,
|
| 62 |
+
"grad_norm": 17.656085968017578,
|
| 63 |
+
"learning_rate": 1.5503875968992248e-05,
|
| 64 |
+
"loss": 1.3487,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.03,
|
| 69 |
+
"grad_norm": 19.021942138671875,
|
| 70 |
+
"learning_rate": 1.744186046511628e-05,
|
| 71 |
+
"loss": 1.5383,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.04,
|
| 76 |
+
"grad_norm": 16.88703727722168,
|
| 77 |
+
"learning_rate": 1.937984496124031e-05,
|
| 78 |
+
"loss": 1.3378,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.04,
|
| 83 |
+
"grad_norm": 28.135297775268555,
|
| 84 |
+
"learning_rate": 2.131782945736434e-05,
|
| 85 |
+
"loss": 1.3925,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.05,
|
| 90 |
+
"grad_norm": 27.902355194091797,
|
| 91 |
+
"learning_rate": 2.3255813953488374e-05,
|
| 92 |
+
"loss": 1.2723,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.05,
|
| 97 |
+
"grad_norm": 25.579036712646484,
|
| 98 |
+
"learning_rate": 2.5193798449612404e-05,
|
| 99 |
+
"loss": 1.453,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.05,
|
| 104 |
+
"grad_norm": 11.281214714050293,
|
| 105 |
+
"learning_rate": 2.7131782945736434e-05,
|
| 106 |
+
"loss": 1.1327,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.06,
|
| 111 |
+
"grad_norm": 31.06259536743164,
|
| 112 |
+
"learning_rate": 2.9069767441860467e-05,
|
| 113 |
+
"loss": 1.0795,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.06,
|
| 118 |
+
"grad_norm": 9.809606552124023,
|
| 119 |
+
"learning_rate": 3.1007751937984497e-05,
|
| 120 |
+
"loss": 1.0496,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.07,
|
| 125 |
+
"grad_norm": 6.943488121032715,
|
| 126 |
+
"learning_rate": 3.294573643410852e-05,
|
| 127 |
+
"loss": 0.7182,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.07,
|
| 132 |
+
"grad_norm": 21.161239624023438,
|
| 133 |
+
"learning_rate": 3.488372093023256e-05,
|
| 134 |
+
"loss": 0.9181,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.07,
|
| 139 |
+
"grad_norm": 22.49970817565918,
|
| 140 |
+
"learning_rate": 3.682170542635659e-05,
|
| 141 |
+
"loss": 0.6194,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.08,
|
| 146 |
+
"grad_norm": 29.908571243286133,
|
| 147 |
+
"learning_rate": 3.875968992248062e-05,
|
| 148 |
+
"loss": 1.3104,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.08,
|
| 153 |
+
"grad_norm": 5.580902099609375,
|
| 154 |
+
"learning_rate": 4.0697674418604655e-05,
|
| 155 |
+
"loss": 0.74,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.09,
|
| 160 |
+
"grad_norm": 7.390064239501953,
|
| 161 |
+
"learning_rate": 4.263565891472868e-05,
|
| 162 |
+
"loss": 0.785,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.09,
|
| 167 |
+
"grad_norm": 17.90110206604004,
|
| 168 |
+
"learning_rate": 4.4573643410852715e-05,
|
| 169 |
+
"loss": 1.0281,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.09,
|
| 174 |
+
"grad_norm": 8.22541618347168,
|
| 175 |
+
"learning_rate": 4.651162790697675e-05,
|
| 176 |
+
"loss": 0.6409,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.1,
|
| 181 |
+
"grad_norm": 15.333436012268066,
|
| 182 |
+
"learning_rate": 4.8449612403100775e-05,
|
| 183 |
+
"loss": 0.2936,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.1,
|
| 188 |
+
"grad_norm": 28.581796646118164,
|
| 189 |
+
"learning_rate": 4.9956822107081176e-05,
|
| 190 |
+
"loss": 0.7362,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.1,
|
| 195 |
+
"grad_norm": 20.49916648864746,
|
| 196 |
+
"learning_rate": 4.974093264248705e-05,
|
| 197 |
+
"loss": 1.1056,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.11,
|
| 202 |
+
"grad_norm": 18.066146850585938,
|
| 203 |
+
"learning_rate": 4.952504317789292e-05,
|
| 204 |
+
"loss": 1.1734,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.11,
|
| 209 |
+
"grad_norm": 7.937985420227051,
|
| 210 |
+
"learning_rate": 4.9309153713298795e-05,
|
| 211 |
+
"loss": 0.8946,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.12,
|
| 216 |
+
"grad_norm": 18.581632614135742,
|
| 217 |
+
"learning_rate": 4.909326424870467e-05,
|
| 218 |
+
"loss": 0.8767,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.12,
|
| 223 |
+
"grad_norm": 10.53328800201416,
|
| 224 |
+
"learning_rate": 4.887737478411054e-05,
|
| 225 |
+
"loss": 0.72,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.12,
|
| 230 |
+
"grad_norm": 30.599201202392578,
|
| 231 |
+
"learning_rate": 4.866148531951641e-05,
|
| 232 |
+
"loss": 0.1731,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.13,
|
| 237 |
+
"grad_norm": 0.21214953064918518,
|
| 238 |
+
"learning_rate": 4.844559585492228e-05,
|
| 239 |
+
"loss": 0.3763,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.13,
|
| 244 |
+
"grad_norm": 40.29164505004883,
|
| 245 |
+
"learning_rate": 4.822970639032816e-05,
|
| 246 |
+
"loss": 1.725,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.14,
|
| 251 |
+
"grad_norm": 0.04869541898369789,
|
| 252 |
+
"learning_rate": 4.8013816925734026e-05,
|
| 253 |
+
"loss": 0.1703,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.14,
|
| 258 |
+
"grad_norm": 2.7087643146514893,
|
| 259 |
+
"learning_rate": 4.7797927461139894e-05,
|
| 260 |
+
"loss": 0.4218,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.14,
|
| 265 |
+
"grad_norm": 15.27876091003418,
|
| 266 |
+
"learning_rate": 4.758203799654577e-05,
|
| 267 |
+
"loss": 0.3047,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.15,
|
| 272 |
+
"grad_norm": 4.537576198577881,
|
| 273 |
+
"learning_rate": 4.7366148531951644e-05,
|
| 274 |
+
"loss": 0.1536,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.15,
|
| 279 |
+
"grad_norm": 16.916975021362305,
|
| 280 |
+
"learning_rate": 4.715025906735751e-05,
|
| 281 |
+
"loss": 0.2277,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.16,
|
| 286 |
+
"grad_norm": 3.467759370803833,
|
| 287 |
+
"learning_rate": 4.693436960276339e-05,
|
| 288 |
+
"loss": 0.7564,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.16,
|
| 293 |
+
"grad_norm": 0.6760605573654175,
|
| 294 |
+
"learning_rate": 4.6718480138169256e-05,
|
| 295 |
+
"loss": 0.5586,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.16,
|
| 300 |
+
"grad_norm": 48.665504455566406,
|
| 301 |
+
"learning_rate": 4.650259067357513e-05,
|
| 302 |
+
"loss": 0.5656,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.17,
|
| 307 |
+
"grad_norm": 0.1229218989610672,
|
| 308 |
+
"learning_rate": 4.628670120898101e-05,
|
| 309 |
+
"loss": 1.1371,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.17,
|
| 314 |
+
"eval_accuracy": 0.7789473684210526,
|
| 315 |
+
"eval_loss": 0.7348644137382507,
|
| 316 |
+
"eval_runtime": 514.0489,
|
| 317 |
+
"eval_samples_per_second": 0.37,
|
| 318 |
+
"eval_steps_per_second": 0.185,
|
| 319 |
+
"step": 430
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.0,
|
| 323 |
+
"grad_norm": 10.519988059997559,
|
| 324 |
+
"learning_rate": 4.6070811744386875e-05,
|
| 325 |
+
"loss": 0.0473,
|
| 326 |
+
"step": 440
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.01,
|
| 330 |
+
"grad_norm": 0.7223495841026306,
|
| 331 |
+
"learning_rate": 4.585492227979275e-05,
|
| 332 |
+
"loss": 0.234,
|
| 333 |
+
"step": 450
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.01,
|
| 337 |
+
"grad_norm": 2.6696724891662598,
|
| 338 |
+
"learning_rate": 4.563903281519862e-05,
|
| 339 |
+
"loss": 0.0174,
|
| 340 |
+
"step": 460
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.02,
|
| 344 |
+
"grad_norm": 36.59164810180664,
|
| 345 |
+
"learning_rate": 4.5423143350604494e-05,
|
| 346 |
+
"loss": 0.6521,
|
| 347 |
+
"step": 470
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.02,
|
| 351 |
+
"grad_norm": 0.5735955834388733,
|
| 352 |
+
"learning_rate": 4.520725388601036e-05,
|
| 353 |
+
"loss": 0.0282,
|
| 354 |
+
"step": 480
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.02,
|
| 358 |
+
"grad_norm": 0.06086433678865433,
|
| 359 |
+
"learning_rate": 4.499136442141624e-05,
|
| 360 |
+
"loss": 0.2758,
|
| 361 |
+
"step": 490
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.03,
|
| 365 |
+
"grad_norm": 52.46869659423828,
|
| 366 |
+
"learning_rate": 4.477547495682211e-05,
|
| 367 |
+
"loss": 0.7029,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.03,
|
| 372 |
+
"grad_norm": 32.279354095458984,
|
| 373 |
+
"learning_rate": 4.455958549222798e-05,
|
| 374 |
+
"loss": 0.3286,
|
| 375 |
+
"step": 510
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.03,
|
| 379 |
+
"grad_norm": 0.4026491045951843,
|
| 380 |
+
"learning_rate": 4.434369602763385e-05,
|
| 381 |
+
"loss": 0.1499,
|
| 382 |
+
"step": 520
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.04,
|
| 386 |
+
"grad_norm": 0.010530718602240086,
|
| 387 |
+
"learning_rate": 4.4127806563039725e-05,
|
| 388 |
+
"loss": 0.3031,
|
| 389 |
+
"step": 530
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.04,
|
| 393 |
+
"grad_norm": 0.0508076511323452,
|
| 394 |
+
"learning_rate": 4.39119170984456e-05,
|
| 395 |
+
"loss": 0.638,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.05,
|
| 400 |
+
"grad_norm": 0.019481591880321503,
|
| 401 |
+
"learning_rate": 4.3696027633851475e-05,
|
| 402 |
+
"loss": 0.1141,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.05,
|
| 407 |
+
"grad_norm": 1.2322841882705688,
|
| 408 |
+
"learning_rate": 4.3480138169257344e-05,
|
| 409 |
+
"loss": 0.831,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.05,
|
| 414 |
+
"grad_norm": 0.04800877720117569,
|
| 415 |
+
"learning_rate": 4.326424870466321e-05,
|
| 416 |
+
"loss": 0.0428,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.06,
|
| 421 |
+
"grad_norm": 72.7511978149414,
|
| 422 |
+
"learning_rate": 4.304835924006909e-05,
|
| 423 |
+
"loss": 0.2349,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.06,
|
| 428 |
+
"grad_norm": 0.14265328645706177,
|
| 429 |
+
"learning_rate": 4.283246977547496e-05,
|
| 430 |
+
"loss": 0.263,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.07,
|
| 435 |
+
"grad_norm": 1.0183712244033813,
|
| 436 |
+
"learning_rate": 4.261658031088083e-05,
|
| 437 |
+
"loss": 0.4732,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.07,
|
| 442 |
+
"grad_norm": 1.8658190965652466,
|
| 443 |
+
"learning_rate": 4.2400690846286706e-05,
|
| 444 |
+
"loss": 0.3766,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.07,
|
| 449 |
+
"grad_norm": 0.16890831291675568,
|
| 450 |
+
"learning_rate": 4.2184801381692574e-05,
|
| 451 |
+
"loss": 1.5658,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.08,
|
| 456 |
+
"grad_norm": 0.25139570236206055,
|
| 457 |
+
"learning_rate": 4.196891191709845e-05,
|
| 458 |
+
"loss": 1.0975,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.08,
|
| 463 |
+
"grad_norm": 19.84381103515625,
|
| 464 |
+
"learning_rate": 4.175302245250432e-05,
|
| 465 |
+
"loss": 0.0726,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.09,
|
| 470 |
+
"grad_norm": 0.07212476432323456,
|
| 471 |
+
"learning_rate": 4.153713298791019e-05,
|
| 472 |
+
"loss": 0.1175,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.09,
|
| 477 |
+
"grad_norm": 0.2093745470046997,
|
| 478 |
+
"learning_rate": 4.132124352331607e-05,
|
| 479 |
+
"loss": 0.2994,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.09,
|
| 484 |
+
"grad_norm": 0.0611826628446579,
|
| 485 |
+
"learning_rate": 4.110535405872194e-05,
|
| 486 |
+
"loss": 0.1013,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.1,
|
| 491 |
+
"grad_norm": 1.0494344234466553,
|
| 492 |
+
"learning_rate": 4.0889464594127805e-05,
|
| 493 |
+
"loss": 0.2448,
|
| 494 |
+
"step": 680
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.1,
|
| 498 |
+
"grad_norm": 0.3157791197299957,
|
| 499 |
+
"learning_rate": 4.067357512953368e-05,
|
| 500 |
+
"loss": 0.1015,
|
| 501 |
+
"step": 690
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.1,
|
| 505 |
+
"grad_norm": 4.036553859710693,
|
| 506 |
+
"learning_rate": 4.0457685664939555e-05,
|
| 507 |
+
"loss": 0.0134,
|
| 508 |
+
"step": 700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.11,
|
| 512 |
+
"grad_norm": 0.4784717857837677,
|
| 513 |
+
"learning_rate": 4.024179620034543e-05,
|
| 514 |
+
"loss": 0.2529,
|
| 515 |
+
"step": 710
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.11,
|
| 519 |
+
"grad_norm": 0.03620552271604538,
|
| 520 |
+
"learning_rate": 4.002590673575129e-05,
|
| 521 |
+
"loss": 0.2876,
|
| 522 |
+
"step": 720
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.12,
|
| 526 |
+
"grad_norm": 0.013772057369351387,
|
| 527 |
+
"learning_rate": 3.981001727115717e-05,
|
| 528 |
+
"loss": 0.2814,
|
| 529 |
+
"step": 730
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.12,
|
| 533 |
+
"grad_norm": 0.01539964322000742,
|
| 534 |
+
"learning_rate": 3.959412780656304e-05,
|
| 535 |
+
"loss": 1.0851,
|
| 536 |
+
"step": 740
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.12,
|
| 540 |
+
"grad_norm": 0.030828053131699562,
|
| 541 |
+
"learning_rate": 3.937823834196892e-05,
|
| 542 |
+
"loss": 0.1881,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.13,
|
| 547 |
+
"grad_norm": 0.06151743233203888,
|
| 548 |
+
"learning_rate": 3.9162348877374786e-05,
|
| 549 |
+
"loss": 0.0725,
|
| 550 |
+
"step": 760
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.13,
|
| 554 |
+
"grad_norm": 0.05447068437933922,
|
| 555 |
+
"learning_rate": 3.8946459412780655e-05,
|
| 556 |
+
"loss": 0.0167,
|
| 557 |
+
"step": 770
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.14,
|
| 561 |
+
"grad_norm": 0.1398928165435791,
|
| 562 |
+
"learning_rate": 3.873056994818653e-05,
|
| 563 |
+
"loss": 0.4225,
|
| 564 |
+
"step": 780
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.14,
|
| 568 |
+
"grad_norm": 0.024751029908657074,
|
| 569 |
+
"learning_rate": 3.8514680483592405e-05,
|
| 570 |
+
"loss": 0.0037,
|
| 571 |
+
"step": 790
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.14,
|
| 575 |
+
"grad_norm": 0.7933443188667297,
|
| 576 |
+
"learning_rate": 3.8298791018998273e-05,
|
| 577 |
+
"loss": 0.0055,
|
| 578 |
+
"step": 800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.15,
|
| 582 |
+
"grad_norm": 7.173922538757324,
|
| 583 |
+
"learning_rate": 3.808290155440415e-05,
|
| 584 |
+
"loss": 0.7573,
|
| 585 |
+
"step": 810
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.15,
|
| 589 |
+
"grad_norm": 85.79830169677734,
|
| 590 |
+
"learning_rate": 3.786701208981002e-05,
|
| 591 |
+
"loss": 0.3346,
|
| 592 |
+
"step": 820
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.16,
|
| 596 |
+
"grad_norm": 0.06877858936786652,
|
| 597 |
+
"learning_rate": 3.765112262521589e-05,
|
| 598 |
+
"loss": 0.0852,
|
| 599 |
+
"step": 830
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.16,
|
| 603 |
+
"grad_norm": 0.08467548340559006,
|
| 604 |
+
"learning_rate": 3.743523316062176e-05,
|
| 605 |
+
"loss": 0.1547,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.16,
|
| 610 |
+
"grad_norm": 0.3360983431339264,
|
| 611 |
+
"learning_rate": 3.7219343696027636e-05,
|
| 612 |
+
"loss": 0.1827,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.17,
|
| 617 |
+
"grad_norm": 0.09804761409759521,
|
| 618 |
+
"learning_rate": 3.700345423143351e-05,
|
| 619 |
+
"loss": 0.5346,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.17,
|
| 624 |
+
"eval_accuracy": 0.8894736842105263,
|
| 625 |
+
"eval_loss": 0.2911826968193054,
|
| 626 |
+
"eval_runtime": 518.0915,
|
| 627 |
+
"eval_samples_per_second": 0.367,
|
| 628 |
+
"eval_steps_per_second": 0.183,
|
| 629 |
+
"step": 860
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 2.0,
|
| 633 |
+
"grad_norm": 0.2422347515821457,
|
| 634 |
+
"learning_rate": 3.678756476683938e-05,
|
| 635 |
+
"loss": 0.1598,
|
| 636 |
+
"step": 870
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 2.01,
|
| 640 |
+
"grad_norm": 0.09047175943851471,
|
| 641 |
+
"learning_rate": 3.657167530224525e-05,
|
| 642 |
+
"loss": 0.6796,
|
| 643 |
+
"step": 880
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 2.01,
|
| 647 |
+
"grad_norm": 0.006813277490437031,
|
| 648 |
+
"learning_rate": 3.635578583765112e-05,
|
| 649 |
+
"loss": 0.4698,
|
| 650 |
+
"step": 890
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.02,
|
| 654 |
+
"grad_norm": 0.07180597633123398,
|
| 655 |
+
"learning_rate": 3.6139896373057e-05,
|
| 656 |
+
"loss": 0.0101,
|
| 657 |
+
"step": 900
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.02,
|
| 661 |
+
"grad_norm": 0.0530867837369442,
|
| 662 |
+
"learning_rate": 3.592400690846287e-05,
|
| 663 |
+
"loss": 0.0263,
|
| 664 |
+
"step": 910
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 2.02,
|
| 668 |
+
"grad_norm": 0.03178895264863968,
|
| 669 |
+
"learning_rate": 3.5708117443868735e-05,
|
| 670 |
+
"loss": 0.0015,
|
| 671 |
+
"step": 920
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 2.03,
|
| 675 |
+
"grad_norm": 0.2046024203300476,
|
| 676 |
+
"learning_rate": 3.549222797927461e-05,
|
| 677 |
+
"loss": 0.2336,
|
| 678 |
+
"step": 930
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.03,
|
| 682 |
+
"grad_norm": 0.01445363275706768,
|
| 683 |
+
"learning_rate": 3.5276338514680485e-05,
|
| 684 |
+
"loss": 0.5172,
|
| 685 |
+
"step": 940
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 2.03,
|
| 689 |
+
"grad_norm": 0.005609246902167797,
|
| 690 |
+
"learning_rate": 3.506044905008636e-05,
|
| 691 |
+
"loss": 0.0465,
|
| 692 |
+
"step": 950
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.04,
|
| 696 |
+
"grad_norm": 2.6773831844329834,
|
| 697 |
+
"learning_rate": 3.484455958549223e-05,
|
| 698 |
+
"loss": 0.093,
|
| 699 |
+
"step": 960
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 2.04,
|
| 703 |
+
"grad_norm": 19.465383529663086,
|
| 704 |
+
"learning_rate": 3.4628670120898104e-05,
|
| 705 |
+
"loss": 0.1096,
|
| 706 |
+
"step": 970
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.05,
|
| 710 |
+
"grad_norm": 0.005580793134868145,
|
| 711 |
+
"learning_rate": 3.441278065630397e-05,
|
| 712 |
+
"loss": 0.1617,
|
| 713 |
+
"step": 980
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 2.05,
|
| 717 |
+
"grad_norm": 0.017572615295648575,
|
| 718 |
+
"learning_rate": 3.419689119170985e-05,
|
| 719 |
+
"loss": 0.1416,
|
| 720 |
+
"step": 990
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.05,
|
| 724 |
+
"grad_norm": 0.004967101849615574,
|
| 725 |
+
"learning_rate": 3.3981001727115716e-05,
|
| 726 |
+
"loss": 0.0009,
|
| 727 |
+
"step": 1000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 2.06,
|
| 731 |
+
"grad_norm": 0.029761947691440582,
|
| 732 |
+
"learning_rate": 3.376511226252159e-05,
|
| 733 |
+
"loss": 0.1359,
|
| 734 |
+
"step": 1010
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 2.06,
|
| 738 |
+
"grad_norm": 8.259181022644043,
|
| 739 |
+
"learning_rate": 3.3549222797927467e-05,
|
| 740 |
+
"loss": 0.0199,
|
| 741 |
+
"step": 1020
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 2.07,
|
| 745 |
+
"grad_norm": 0.014490657486021519,
|
| 746 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 747 |
+
"loss": 0.1351,
|
| 748 |
+
"step": 1030
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 2.07,
|
| 752 |
+
"grad_norm": 0.006075919605791569,
|
| 753 |
+
"learning_rate": 3.31174438687392e-05,
|
| 754 |
+
"loss": 0.2135,
|
| 755 |
+
"step": 1040
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 2.07,
|
| 759 |
+
"grad_norm": 0.005090653896331787,
|
| 760 |
+
"learning_rate": 3.290155440414508e-05,
|
| 761 |
+
"loss": 0.0575,
|
| 762 |
+
"step": 1050
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 2.08,
|
| 766 |
+
"grad_norm": 19.904199600219727,
|
| 767 |
+
"learning_rate": 3.2685664939550954e-05,
|
| 768 |
+
"loss": 0.0234,
|
| 769 |
+
"step": 1060
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.08,
|
| 773 |
+
"grad_norm": 0.022977175191044807,
|
| 774 |
+
"learning_rate": 3.246977547495683e-05,
|
| 775 |
+
"loss": 0.0154,
|
| 776 |
+
"step": 1070
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.09,
|
| 780 |
+
"grad_norm": 0.015029653906822205,
|
| 781 |
+
"learning_rate": 3.225388601036269e-05,
|
| 782 |
+
"loss": 0.2719,
|
| 783 |
+
"step": 1080
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 2.09,
|
| 787 |
+
"grad_norm": 0.010515735484659672,
|
| 788 |
+
"learning_rate": 3.2037996545768566e-05,
|
| 789 |
+
"loss": 0.0067,
|
| 790 |
+
"step": 1090
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 2.09,
|
| 794 |
+
"grad_norm": 0.0053952718153595924,
|
| 795 |
+
"learning_rate": 3.182210708117444e-05,
|
| 796 |
+
"loss": 0.0114,
|
| 797 |
+
"step": 1100
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 2.1,
|
| 801 |
+
"grad_norm": 53.84306335449219,
|
| 802 |
+
"learning_rate": 3.1606217616580316e-05,
|
| 803 |
+
"loss": 0.4551,
|
| 804 |
+
"step": 1110
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 2.1,
|
| 808 |
+
"grad_norm": 1.077244758605957,
|
| 809 |
+
"learning_rate": 3.1390328151986184e-05,
|
| 810 |
+
"loss": 0.2994,
|
| 811 |
+
"step": 1120
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 2.1,
|
| 815 |
+
"grad_norm": 22.892969131469727,
|
| 816 |
+
"learning_rate": 3.117443868739205e-05,
|
| 817 |
+
"loss": 0.1593,
|
| 818 |
+
"step": 1130
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.11,
|
| 822 |
+
"grad_norm": 0.013651028275489807,
|
| 823 |
+
"learning_rate": 3.095854922279793e-05,
|
| 824 |
+
"loss": 0.0082,
|
| 825 |
+
"step": 1140
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 2.11,
|
| 829 |
+
"grad_norm": 0.008860527537763119,
|
| 830 |
+
"learning_rate": 3.07426597582038e-05,
|
| 831 |
+
"loss": 0.2133,
|
| 832 |
+
"step": 1150
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.12,
|
| 836 |
+
"grad_norm": 0.06896065920591354,
|
| 837 |
+
"learning_rate": 3.052677029360967e-05,
|
| 838 |
+
"loss": 0.0106,
|
| 839 |
+
"step": 1160
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 2.12,
|
| 843 |
+
"grad_norm": 0.025299793109297752,
|
| 844 |
+
"learning_rate": 3.0310880829015547e-05,
|
| 845 |
+
"loss": 0.0008,
|
| 846 |
+
"step": 1170
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.12,
|
| 850 |
+
"grad_norm": 0.005163596011698246,
|
| 851 |
+
"learning_rate": 3.0094991364421415e-05,
|
| 852 |
+
"loss": 0.0056,
|
| 853 |
+
"step": 1180
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 2.13,
|
| 857 |
+
"grad_norm": 0.5263453125953674,
|
| 858 |
+
"learning_rate": 2.9879101899827287e-05,
|
| 859 |
+
"loss": 0.0034,
|
| 860 |
+
"step": 1190
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.13,
|
| 864 |
+
"grad_norm": 0.005412234924733639,
|
| 865 |
+
"learning_rate": 2.9663212435233162e-05,
|
| 866 |
+
"loss": 0.0011,
|
| 867 |
+
"step": 1200
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 2.14,
|
| 871 |
+
"grad_norm": 0.004280978813767433,
|
| 872 |
+
"learning_rate": 2.9447322970639034e-05,
|
| 873 |
+
"loss": 0.315,
|
| 874 |
+
"step": 1210
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 2.14,
|
| 878 |
+
"grad_norm": 1.1846489906311035,
|
| 879 |
+
"learning_rate": 2.923143350604491e-05,
|
| 880 |
+
"loss": 0.0016,
|
| 881 |
+
"step": 1220
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 2.14,
|
| 885 |
+
"grad_norm": 3.384139060974121,
|
| 886 |
+
"learning_rate": 2.9015544041450778e-05,
|
| 887 |
+
"loss": 0.004,
|
| 888 |
+
"step": 1230
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 2.15,
|
| 892 |
+
"grad_norm": 0.003910099621862173,
|
| 893 |
+
"learning_rate": 2.879965457685665e-05,
|
| 894 |
+
"loss": 0.3247,
|
| 895 |
+
"step": 1240
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 2.15,
|
| 899 |
+
"grad_norm": 55.08097839355469,
|
| 900 |
+
"learning_rate": 2.858376511226252e-05,
|
| 901 |
+
"loss": 0.1251,
|
| 902 |
+
"step": 1250
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.16,
|
| 906 |
+
"grad_norm": 0.01313870307058096,
|
| 907 |
+
"learning_rate": 2.8367875647668396e-05,
|
| 908 |
+
"loss": 0.0334,
|
| 909 |
+
"step": 1260
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 2.16,
|
| 913 |
+
"grad_norm": 4.602208614349365,
|
| 914 |
+
"learning_rate": 2.8151986183074268e-05,
|
| 915 |
+
"loss": 0.0044,
|
| 916 |
+
"step": 1270
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 2.16,
|
| 920 |
+
"grad_norm": 0.5221287608146667,
|
| 921 |
+
"learning_rate": 2.7936096718480137e-05,
|
| 922 |
+
"loss": 0.0129,
|
| 923 |
+
"step": 1280
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.17,
|
| 927 |
+
"grad_norm": 0.005316704977303743,
|
| 928 |
+
"learning_rate": 2.7720207253886012e-05,
|
| 929 |
+
"loss": 0.0004,
|
| 930 |
+
"step": 1290
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 2.17,
|
| 934 |
+
"eval_accuracy": 0.9210526315789473,
|
| 935 |
+
"eval_loss": 0.31835657358169556,
|
| 936 |
+
"eval_runtime": 20895.2371,
|
| 937 |
+
"eval_samples_per_second": 0.009,
|
| 938 |
+
"eval_steps_per_second": 0.005,
|
| 939 |
+
"step": 1290
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 3.0,
|
| 943 |
+
"grad_norm": 0.23466819524765015,
|
| 944 |
+
"learning_rate": 2.7504317789291884e-05,
|
| 945 |
+
"loss": 0.0007,
|
| 946 |
+
"step": 1300
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 3.01,
|
| 950 |
+
"grad_norm": 0.08376140892505646,
|
| 951 |
+
"learning_rate": 2.7288428324697755e-05,
|
| 952 |
+
"loss": 0.0011,
|
| 953 |
+
"step": 1310
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 3.01,
|
| 957 |
+
"grad_norm": 0.0067549836821854115,
|
| 958 |
+
"learning_rate": 2.707253886010363e-05,
|
| 959 |
+
"loss": 0.0004,
|
| 960 |
+
"step": 1320
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 3.02,
|
| 964 |
+
"grad_norm": 0.00525926286354661,
|
| 965 |
+
"learning_rate": 2.68566493955095e-05,
|
| 966 |
+
"loss": 0.0407,
|
| 967 |
+
"step": 1330
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 3.02,
|
| 971 |
+
"grad_norm": 0.003173103556036949,
|
| 972 |
+
"learning_rate": 2.664075993091537e-05,
|
| 973 |
+
"loss": 0.0091,
|
| 974 |
+
"step": 1340
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 3.02,
|
| 978 |
+
"grad_norm": 0.1465524435043335,
|
| 979 |
+
"learning_rate": 2.6424870466321246e-05,
|
| 980 |
+
"loss": 0.0004,
|
| 981 |
+
"step": 1350
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 3.03,
|
| 985 |
+
"grad_norm": 0.01814081333577633,
|
| 986 |
+
"learning_rate": 2.6208981001727118e-05,
|
| 987 |
+
"loss": 0.0003,
|
| 988 |
+
"step": 1360
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 3.03,
|
| 992 |
+
"grad_norm": 0.16582980751991272,
|
| 993 |
+
"learning_rate": 2.599309153713299e-05,
|
| 994 |
+
"loss": 0.1912,
|
| 995 |
+
"step": 1370
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 3.03,
|
| 999 |
+
"grad_norm": 114.0838394165039,
|
| 1000 |
+
"learning_rate": 2.5777202072538865e-05,
|
| 1001 |
+
"loss": 0.1365,
|
| 1002 |
+
"step": 1380
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 3.04,
|
| 1006 |
+
"grad_norm": 0.00724332919344306,
|
| 1007 |
+
"learning_rate": 2.5561312607944733e-05,
|
| 1008 |
+
"loss": 0.1497,
|
| 1009 |
+
"step": 1390
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 3.04,
|
| 1013 |
+
"grad_norm": 0.0034754930529743433,
|
| 1014 |
+
"learning_rate": 2.5345423143350605e-05,
|
| 1015 |
+
"loss": 0.0057,
|
| 1016 |
+
"step": 1400
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 3.05,
|
| 1020 |
+
"grad_norm": 0.0051283338107168674,
|
| 1021 |
+
"learning_rate": 2.5129533678756477e-05,
|
| 1022 |
+
"loss": 0.0017,
|
| 1023 |
+
"step": 1410
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 3.05,
|
| 1027 |
+
"grad_norm": 0.001818397780880332,
|
| 1028 |
+
"learning_rate": 2.491364421416235e-05,
|
| 1029 |
+
"loss": 0.0156,
|
| 1030 |
+
"step": 1420
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 3.05,
|
| 1034 |
+
"grad_norm": 0.01026181410998106,
|
| 1035 |
+
"learning_rate": 2.4697754749568224e-05,
|
| 1036 |
+
"loss": 0.0003,
|
| 1037 |
+
"step": 1430
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 3.06,
|
| 1041 |
+
"grad_norm": 0.013726489618420601,
|
| 1042 |
+
"learning_rate": 2.4481865284974096e-05,
|
| 1043 |
+
"loss": 0.0002,
|
| 1044 |
+
"step": 1440
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 3.06,
|
| 1048 |
+
"grad_norm": 0.004174862988293171,
|
| 1049 |
+
"learning_rate": 2.4265975820379967e-05,
|
| 1050 |
+
"loss": 0.0482,
|
| 1051 |
+
"step": 1450
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 3.07,
|
| 1055 |
+
"grad_norm": 0.013808720745146275,
|
| 1056 |
+
"learning_rate": 2.405008635578584e-05,
|
| 1057 |
+
"loss": 0.1185,
|
| 1058 |
+
"step": 1460
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 3.07,
|
| 1062 |
+
"grad_norm": 0.07046998292207718,
|
| 1063 |
+
"learning_rate": 2.383419689119171e-05,
|
| 1064 |
+
"loss": 0.0119,
|
| 1065 |
+
"step": 1470
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 3.07,
|
| 1069 |
+
"grad_norm": 0.0045805503614246845,
|
| 1070 |
+
"learning_rate": 2.3618307426597583e-05,
|
| 1071 |
+
"loss": 0.0296,
|
| 1072 |
+
"step": 1480
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 3.08,
|
| 1076 |
+
"grad_norm": 0.06411267071962357,
|
| 1077 |
+
"learning_rate": 2.3402417962003455e-05,
|
| 1078 |
+
"loss": 0.2067,
|
| 1079 |
+
"step": 1490
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 3.08,
|
| 1083 |
+
"grad_norm": 0.046669330447912216,
|
| 1084 |
+
"learning_rate": 2.3186528497409326e-05,
|
| 1085 |
+
"loss": 0.0761,
|
| 1086 |
+
"step": 1500
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 3.09,
|
| 1090 |
+
"grad_norm": 41.745513916015625,
|
| 1091 |
+
"learning_rate": 2.29706390328152e-05,
|
| 1092 |
+
"loss": 0.45,
|
| 1093 |
+
"step": 1510
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 3.09,
|
| 1097 |
+
"grad_norm": 0.0023008265998214483,
|
| 1098 |
+
"learning_rate": 2.2754749568221073e-05,
|
| 1099 |
+
"loss": 0.0003,
|
| 1100 |
+
"step": 1520
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 3.09,
|
| 1104 |
+
"grad_norm": 0.008280325680971146,
|
| 1105 |
+
"learning_rate": 2.2538860103626945e-05,
|
| 1106 |
+
"loss": 0.0002,
|
| 1107 |
+
"step": 1530
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 3.1,
|
| 1111 |
+
"grad_norm": 0.06091548874974251,
|
| 1112 |
+
"learning_rate": 2.2322970639032817e-05,
|
| 1113 |
+
"loss": 0.0003,
|
| 1114 |
+
"step": 1540
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 3.1,
|
| 1118 |
+
"grad_norm": 0.009856736287474632,
|
| 1119 |
+
"learning_rate": 2.210708117443869e-05,
|
| 1120 |
+
"loss": 0.0005,
|
| 1121 |
+
"step": 1550
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 3.1,
|
| 1125 |
+
"grad_norm": 0.0037303988356143236,
|
| 1126 |
+
"learning_rate": 2.189119170984456e-05,
|
| 1127 |
+
"loss": 0.1055,
|
| 1128 |
+
"step": 1560
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 3.11,
|
| 1132 |
+
"grad_norm": 0.3009835183620453,
|
| 1133 |
+
"learning_rate": 2.1675302245250432e-05,
|
| 1134 |
+
"loss": 0.0005,
|
| 1135 |
+
"step": 1570
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 3.11,
|
| 1139 |
+
"grad_norm": 0.005187860690057278,
|
| 1140 |
+
"learning_rate": 2.1459412780656304e-05,
|
| 1141 |
+
"loss": 0.0024,
|
| 1142 |
+
"step": 1580
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 3.12,
|
| 1146 |
+
"grad_norm": 0.0036395310889929533,
|
| 1147 |
+
"learning_rate": 2.124352331606218e-05,
|
| 1148 |
+
"loss": 0.288,
|
| 1149 |
+
"step": 1590
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 3.12,
|
| 1153 |
+
"grad_norm": 0.006734526250511408,
|
| 1154 |
+
"learning_rate": 2.1027633851468048e-05,
|
| 1155 |
+
"loss": 0.0004,
|
| 1156 |
+
"step": 1600
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 3.12,
|
| 1160 |
+
"grad_norm": 0.003171958727762103,
|
| 1161 |
+
"learning_rate": 2.0811744386873923e-05,
|
| 1162 |
+
"loss": 0.1013,
|
| 1163 |
+
"step": 1610
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"epoch": 3.13,
|
| 1167 |
+
"grad_norm": 0.0022121332585811615,
|
| 1168 |
+
"learning_rate": 2.0595854922279795e-05,
|
| 1169 |
+
"loss": 0.0716,
|
| 1170 |
+
"step": 1620
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 3.13,
|
| 1174 |
+
"grad_norm": 0.04961927980184555,
|
| 1175 |
+
"learning_rate": 2.0379965457685666e-05,
|
| 1176 |
+
"loss": 0.014,
|
| 1177 |
+
"step": 1630
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 3.14,
|
| 1181 |
+
"grad_norm": 0.042126741260290146,
|
| 1182 |
+
"learning_rate": 2.0164075993091538e-05,
|
| 1183 |
+
"loss": 0.1356,
|
| 1184 |
+
"step": 1640
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 3.14,
|
| 1188 |
+
"grad_norm": 0.6265960931777954,
|
| 1189 |
+
"learning_rate": 1.994818652849741e-05,
|
| 1190 |
+
"loss": 0.0009,
|
| 1191 |
+
"step": 1650
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 3.14,
|
| 1195 |
+
"grad_norm": 2.0805187225341797,
|
| 1196 |
+
"learning_rate": 1.9732297063903282e-05,
|
| 1197 |
+
"loss": 0.0439,
|
| 1198 |
+
"step": 1660
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 3.15,
|
| 1202 |
+
"grad_norm": 0.0070223091170191765,
|
| 1203 |
+
"learning_rate": 1.9516407599309157e-05,
|
| 1204 |
+
"loss": 0.015,
|
| 1205 |
+
"step": 1670
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 3.15,
|
| 1209 |
+
"grad_norm": 0.008443433791399002,
|
| 1210 |
+
"learning_rate": 1.9300518134715025e-05,
|
| 1211 |
+
"loss": 0.0016,
|
| 1212 |
+
"step": 1680
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 3.16,
|
| 1216 |
+
"grad_norm": 0.008806809782981873,
|
| 1217 |
+
"learning_rate": 1.90846286701209e-05,
|
| 1218 |
+
"loss": 0.047,
|
| 1219 |
+
"step": 1690
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 3.16,
|
| 1223 |
+
"grad_norm": 0.008072856813669205,
|
| 1224 |
+
"learning_rate": 1.8868739205526772e-05,
|
| 1225 |
+
"loss": 0.0005,
|
| 1226 |
+
"step": 1700
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 3.16,
|
| 1230 |
+
"grad_norm": 0.061166293919086456,
|
| 1231 |
+
"learning_rate": 1.8652849740932644e-05,
|
| 1232 |
+
"loss": 0.0056,
|
| 1233 |
+
"step": 1710
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 3.17,
|
| 1237 |
+
"grad_norm": 0.026545794680714607,
|
| 1238 |
+
"learning_rate": 1.8436960276338516e-05,
|
| 1239 |
+
"loss": 0.0006,
|
| 1240 |
+
"step": 1720
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 3.17,
|
| 1244 |
+
"eval_accuracy": 0.9368421052631579,
|
| 1245 |
+
"eval_loss": 0.357727587223053,
|
| 1246 |
+
"eval_runtime": 523.7452,
|
| 1247 |
+
"eval_samples_per_second": 0.363,
|
| 1248 |
+
"eval_steps_per_second": 0.181,
|
| 1249 |
+
"step": 1720
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 4.0,
|
| 1253 |
+
"grad_norm": 0.20675712823867798,
|
| 1254 |
+
"learning_rate": 1.8221070811744388e-05,
|
| 1255 |
+
"loss": 0.0005,
|
| 1256 |
+
"step": 1730
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 4.01,
|
| 1260 |
+
"grad_norm": 0.007863791659474373,
|
| 1261 |
+
"learning_rate": 1.800518134715026e-05,
|
| 1262 |
+
"loss": 0.0046,
|
| 1263 |
+
"step": 1740
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 4.01,
|
| 1267 |
+
"grad_norm": 0.00224371743388474,
|
| 1268 |
+
"learning_rate": 1.7789291882556135e-05,
|
| 1269 |
+
"loss": 0.0002,
|
| 1270 |
+
"step": 1750
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 4.02,
|
| 1274 |
+
"grad_norm": 0.00742675457149744,
|
| 1275 |
+
"learning_rate": 1.7573402417962003e-05,
|
| 1276 |
+
"loss": 0.0021,
|
| 1277 |
+
"step": 1760
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 4.02,
|
| 1281 |
+
"grad_norm": 0.02301434613764286,
|
| 1282 |
+
"learning_rate": 1.735751295336788e-05,
|
| 1283 |
+
"loss": 0.0002,
|
| 1284 |
+
"step": 1770
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 4.02,
|
| 1288 |
+
"grad_norm": 0.03734876960515976,
|
| 1289 |
+
"learning_rate": 1.7141623488773747e-05,
|
| 1290 |
+
"loss": 0.0001,
|
| 1291 |
+
"step": 1780
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 4.03,
|
| 1295 |
+
"grad_norm": 0.0020719885360449553,
|
| 1296 |
+
"learning_rate": 1.6925734024179622e-05,
|
| 1297 |
+
"loss": 0.0052,
|
| 1298 |
+
"step": 1790
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 4.03,
|
| 1302 |
+
"grad_norm": 0.021191716194152832,
|
| 1303 |
+
"learning_rate": 1.6709844559585494e-05,
|
| 1304 |
+
"loss": 0.2831,
|
| 1305 |
+
"step": 1800
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 4.03,
|
| 1309 |
+
"grad_norm": 0.033936869353055954,
|
| 1310 |
+
"learning_rate": 1.6493955094991366e-05,
|
| 1311 |
+
"loss": 0.0004,
|
| 1312 |
+
"step": 1810
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 4.04,
|
| 1316 |
+
"grad_norm": 0.3438062071800232,
|
| 1317 |
+
"learning_rate": 1.6278065630397237e-05,
|
| 1318 |
+
"loss": 0.0005,
|
| 1319 |
+
"step": 1820
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 4.04,
|
| 1323 |
+
"grad_norm": 0.0040876902639865875,
|
| 1324 |
+
"learning_rate": 1.606217616580311e-05,
|
| 1325 |
+
"loss": 0.0001,
|
| 1326 |
+
"step": 1830
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 4.05,
|
| 1330 |
+
"grad_norm": 0.0035211730282753706,
|
| 1331 |
+
"learning_rate": 1.584628670120898e-05,
|
| 1332 |
+
"loss": 0.0002,
|
| 1333 |
+
"step": 1840
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 4.05,
|
| 1337 |
+
"grad_norm": 0.004084211308509111,
|
| 1338 |
+
"learning_rate": 1.5630397236614856e-05,
|
| 1339 |
+
"loss": 0.0004,
|
| 1340 |
+
"step": 1850
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 4.05,
|
| 1344 |
+
"grad_norm": 0.0032758773304522038,
|
| 1345 |
+
"learning_rate": 1.5414507772020725e-05,
|
| 1346 |
+
"loss": 0.0365,
|
| 1347 |
+
"step": 1860
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 4.06,
|
| 1351 |
+
"grad_norm": 14.158347129821777,
|
| 1352 |
+
"learning_rate": 1.5198618307426598e-05,
|
| 1353 |
+
"loss": 0.0074,
|
| 1354 |
+
"step": 1870
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 4.06,
|
| 1358 |
+
"grad_norm": 0.002404096070677042,
|
| 1359 |
+
"learning_rate": 1.498272884283247e-05,
|
| 1360 |
+
"loss": 0.0001,
|
| 1361 |
+
"step": 1880
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 4.07,
|
| 1365 |
+
"grad_norm": 0.02108081988990307,
|
| 1366 |
+
"learning_rate": 1.4766839378238342e-05,
|
| 1367 |
+
"loss": 0.0002,
|
| 1368 |
+
"step": 1890
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 4.07,
|
| 1372 |
+
"grad_norm": 0.12371557205915451,
|
| 1373 |
+
"learning_rate": 1.4550949913644215e-05,
|
| 1374 |
+
"loss": 0.038,
|
| 1375 |
+
"step": 1900
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 4.07,
|
| 1379 |
+
"grad_norm": 0.0032504727132618427,
|
| 1380 |
+
"learning_rate": 1.4335060449050087e-05,
|
| 1381 |
+
"loss": 0.0007,
|
| 1382 |
+
"step": 1910
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 4.08,
|
| 1386 |
+
"grad_norm": 71.4847183227539,
|
| 1387 |
+
"learning_rate": 1.4119170984455959e-05,
|
| 1388 |
+
"loss": 0.0218,
|
| 1389 |
+
"step": 1920
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 4.08,
|
| 1393 |
+
"grad_norm": 0.007250852882862091,
|
| 1394 |
+
"learning_rate": 1.3903281519861832e-05,
|
| 1395 |
+
"loss": 0.0002,
|
| 1396 |
+
"step": 1930
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 4.09,
|
| 1400 |
+
"grad_norm": 0.0018006942700594664,
|
| 1401 |
+
"learning_rate": 1.3687392055267704e-05,
|
| 1402 |
+
"loss": 0.0007,
|
| 1403 |
+
"step": 1940
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 4.09,
|
| 1407 |
+
"grad_norm": 0.0018308752914890647,
|
| 1408 |
+
"learning_rate": 1.3471502590673576e-05,
|
| 1409 |
+
"loss": 0.0001,
|
| 1410 |
+
"step": 1950
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 4.09,
|
| 1414 |
+
"grad_norm": 0.004195887595415115,
|
| 1415 |
+
"learning_rate": 1.3255613126079448e-05,
|
| 1416 |
+
"loss": 0.0001,
|
| 1417 |
+
"step": 1960
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 4.1,
|
| 1421 |
+
"grad_norm": 0.005353459622710943,
|
| 1422 |
+
"learning_rate": 1.303972366148532e-05,
|
| 1423 |
+
"loss": 0.0002,
|
| 1424 |
+
"step": 1970
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 4.1,
|
| 1428 |
+
"grad_norm": 0.00350914616137743,
|
| 1429 |
+
"learning_rate": 1.2823834196891193e-05,
|
| 1430 |
+
"loss": 0.0002,
|
| 1431 |
+
"step": 1980
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 4.1,
|
| 1435 |
+
"grad_norm": 0.003935214597731829,
|
| 1436 |
+
"learning_rate": 1.2607944732297065e-05,
|
| 1437 |
+
"loss": 0.0002,
|
| 1438 |
+
"step": 1990
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 4.11,
|
| 1442 |
+
"grad_norm": 1.447480320930481,
|
| 1443 |
+
"learning_rate": 1.2392055267702936e-05,
|
| 1444 |
+
"loss": 0.1091,
|
| 1445 |
+
"step": 2000
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 4.11,
|
| 1449 |
+
"grad_norm": 0.0019440186442807317,
|
| 1450 |
+
"learning_rate": 1.2176165803108808e-05,
|
| 1451 |
+
"loss": 0.0005,
|
| 1452 |
+
"step": 2010
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 4.12,
|
| 1456 |
+
"grad_norm": 0.007630598731338978,
|
| 1457 |
+
"learning_rate": 1.1960276338514682e-05,
|
| 1458 |
+
"loss": 0.0001,
|
| 1459 |
+
"step": 2020
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 4.12,
|
| 1463 |
+
"grad_norm": 0.0019236513180658221,
|
| 1464 |
+
"learning_rate": 1.1744386873920554e-05,
|
| 1465 |
+
"loss": 0.0003,
|
| 1466 |
+
"step": 2030
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 4.12,
|
| 1470 |
+
"grad_norm": 0.005540070589631796,
|
| 1471 |
+
"learning_rate": 1.1528497409326425e-05,
|
| 1472 |
+
"loss": 0.0054,
|
| 1473 |
+
"step": 2040
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 4.13,
|
| 1477 |
+
"grad_norm": 0.016037726774811745,
|
| 1478 |
+
"learning_rate": 1.1312607944732297e-05,
|
| 1479 |
+
"loss": 0.0002,
|
| 1480 |
+
"step": 2050
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 4.13,
|
| 1484 |
+
"grad_norm": 0.006410875823348761,
|
| 1485 |
+
"learning_rate": 1.109671848013817e-05,
|
| 1486 |
+
"loss": 0.0006,
|
| 1487 |
+
"step": 2060
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 4.14,
|
| 1491 |
+
"grad_norm": 0.03595797345042229,
|
| 1492 |
+
"learning_rate": 1.0880829015544042e-05,
|
| 1493 |
+
"loss": 0.0009,
|
| 1494 |
+
"step": 2070
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 4.14,
|
| 1498 |
+
"grad_norm": 1.9901659488677979,
|
| 1499 |
+
"learning_rate": 1.0664939550949914e-05,
|
| 1500 |
+
"loss": 0.0027,
|
| 1501 |
+
"step": 2080
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 4.14,
|
| 1505 |
+
"grad_norm": 0.00291555211879313,
|
| 1506 |
+
"learning_rate": 1.0449050086355788e-05,
|
| 1507 |
+
"loss": 0.0045,
|
| 1508 |
+
"step": 2090
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 4.15,
|
| 1512 |
+
"grad_norm": 0.0020950380712747574,
|
| 1513 |
+
"learning_rate": 1.023316062176166e-05,
|
| 1514 |
+
"loss": 0.0167,
|
| 1515 |
+
"step": 2100
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 4.15,
|
| 1519 |
+
"grad_norm": 0.0015124623896554112,
|
| 1520 |
+
"learning_rate": 1.0017271157167531e-05,
|
| 1521 |
+
"loss": 0.0002,
|
| 1522 |
+
"step": 2110
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 4.16,
|
| 1526 |
+
"grad_norm": 0.001781588769517839,
|
| 1527 |
+
"learning_rate": 9.801381692573403e-06,
|
| 1528 |
+
"loss": 0.0004,
|
| 1529 |
+
"step": 2120
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 4.16,
|
| 1533 |
+
"grad_norm": 0.006017952226102352,
|
| 1534 |
+
"learning_rate": 9.585492227979275e-06,
|
| 1535 |
+
"loss": 0.0003,
|
| 1536 |
+
"step": 2130
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 4.16,
|
| 1540 |
+
"grad_norm": 0.017710313200950623,
|
| 1541 |
+
"learning_rate": 9.369602763385148e-06,
|
| 1542 |
+
"loss": 0.0007,
|
| 1543 |
+
"step": 2140
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 4.17,
|
| 1547 |
+
"grad_norm": 0.002331692725419998,
|
| 1548 |
+
"learning_rate": 9.15371329879102e-06,
|
| 1549 |
+
"loss": 0.0382,
|
| 1550 |
+
"step": 2150
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 4.17,
|
| 1554 |
+
"eval_accuracy": 0.9315789473684211,
|
| 1555 |
+
"eval_loss": 0.34909680485725403,
|
| 1556 |
+
"eval_runtime": 564.551,
|
| 1557 |
+
"eval_samples_per_second": 0.337,
|
| 1558 |
+
"eval_steps_per_second": 0.168,
|
| 1559 |
+
"step": 2150
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 5.0,
|
| 1563 |
+
"grad_norm": 0.002333582378923893,
|
| 1564 |
+
"learning_rate": 8.937823834196892e-06,
|
| 1565 |
+
"loss": 0.004,
|
| 1566 |
+
"step": 2160
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 5.01,
|
| 1570 |
+
"grad_norm": 0.0014494940405711532,
|
| 1571 |
+
"learning_rate": 8.721934369602764e-06,
|
| 1572 |
+
"loss": 0.0003,
|
| 1573 |
+
"step": 2170
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 5.01,
|
| 1577 |
+
"grad_norm": 0.016125734895467758,
|
| 1578 |
+
"learning_rate": 8.506044905008637e-06,
|
| 1579 |
+
"loss": 0.0001,
|
| 1580 |
+
"step": 2180
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 5.02,
|
| 1584 |
+
"grad_norm": 0.004009012598544359,
|
| 1585 |
+
"learning_rate": 8.290155440414509e-06,
|
| 1586 |
+
"loss": 0.0001,
|
| 1587 |
+
"step": 2190
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 5.02,
|
| 1591 |
+
"grad_norm": 0.009335234761238098,
|
| 1592 |
+
"learning_rate": 8.074265975820381e-06,
|
| 1593 |
+
"loss": 0.0118,
|
| 1594 |
+
"step": 2200
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 5.02,
|
| 1598 |
+
"grad_norm": 0.01937965489923954,
|
| 1599 |
+
"learning_rate": 7.858376511226253e-06,
|
| 1600 |
+
"loss": 0.0019,
|
| 1601 |
+
"step": 2210
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 5.03,
|
| 1605 |
+
"grad_norm": 0.0025466824881732464,
|
| 1606 |
+
"learning_rate": 7.642487046632124e-06,
|
| 1607 |
+
"loss": 0.0001,
|
| 1608 |
+
"step": 2220
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 5.03,
|
| 1612 |
+
"grad_norm": 0.0036932919174432755,
|
| 1613 |
+
"learning_rate": 7.426597582037997e-06,
|
| 1614 |
+
"loss": 0.0001,
|
| 1615 |
+
"step": 2230
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 5.03,
|
| 1619 |
+
"grad_norm": 0.003243980696424842,
|
| 1620 |
+
"learning_rate": 7.210708117443869e-06,
|
| 1621 |
+
"loss": 0.0001,
|
| 1622 |
+
"step": 2240
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 5.04,
|
| 1626 |
+
"grad_norm": 0.002856618259102106,
|
| 1627 |
+
"learning_rate": 6.994818652849741e-06,
|
| 1628 |
+
"loss": 0.0001,
|
| 1629 |
+
"step": 2250
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 5.04,
|
| 1633 |
+
"grad_norm": 0.0019520550267770886,
|
| 1634 |
+
"learning_rate": 6.778929188255613e-06,
|
| 1635 |
+
"loss": 0.0001,
|
| 1636 |
+
"step": 2260
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 5.05,
|
| 1640 |
+
"grad_norm": 0.003286935854703188,
|
| 1641 |
+
"learning_rate": 6.563039723661486e-06,
|
| 1642 |
+
"loss": 0.0001,
|
| 1643 |
+
"step": 2270
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 5.05,
|
| 1647 |
+
"grad_norm": 0.001508223474957049,
|
| 1648 |
+
"learning_rate": 6.347150259067358e-06,
|
| 1649 |
+
"loss": 0.0056,
|
| 1650 |
+
"step": 2280
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 5.05,
|
| 1654 |
+
"grad_norm": 0.05285673215985298,
|
| 1655 |
+
"learning_rate": 6.13126079447323e-06,
|
| 1656 |
+
"loss": 0.0054,
|
| 1657 |
+
"step": 2290
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 5.06,
|
| 1661 |
+
"grad_norm": 0.004176164045929909,
|
| 1662 |
+
"learning_rate": 5.915371329879102e-06,
|
| 1663 |
+
"loss": 0.0589,
|
| 1664 |
+
"step": 2300
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 5.06,
|
| 1668 |
+
"grad_norm": 0.0049232905730605125,
|
| 1669 |
+
"learning_rate": 5.699481865284975e-06,
|
| 1670 |
+
"loss": 0.0001,
|
| 1671 |
+
"step": 2310
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 5.07,
|
| 1675 |
+
"grad_norm": 29.9996337890625,
|
| 1676 |
+
"learning_rate": 5.483592400690847e-06,
|
| 1677 |
+
"loss": 0.0541,
|
| 1678 |
+
"step": 2320
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 5.07,
|
| 1682 |
+
"grad_norm": 0.006466562859714031,
|
| 1683 |
+
"learning_rate": 5.267702936096719e-06,
|
| 1684 |
+
"loss": 0.0002,
|
| 1685 |
+
"step": 2330
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 5.07,
|
| 1689 |
+
"grad_norm": 0.002998851239681244,
|
| 1690 |
+
"learning_rate": 5.051813471502591e-06,
|
| 1691 |
+
"loss": 0.0001,
|
| 1692 |
+
"step": 2340
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 5.08,
|
| 1696 |
+
"grad_norm": 0.001166225876659155,
|
| 1697 |
+
"learning_rate": 4.835924006908464e-06,
|
| 1698 |
+
"loss": 0.0002,
|
| 1699 |
+
"step": 2350
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 5.08,
|
| 1703 |
+
"grad_norm": 0.0021310816518962383,
|
| 1704 |
+
"learning_rate": 4.620034542314336e-06,
|
| 1705 |
+
"loss": 0.0009,
|
| 1706 |
+
"step": 2360
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 5.09,
|
| 1710 |
+
"grad_norm": 0.004912310279905796,
|
| 1711 |
+
"learning_rate": 4.404145077720207e-06,
|
| 1712 |
+
"loss": 0.0002,
|
| 1713 |
+
"step": 2370
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 5.09,
|
| 1717 |
+
"grad_norm": 0.0024258983321487904,
|
| 1718 |
+
"learning_rate": 4.18825561312608e-06,
|
| 1719 |
+
"loss": 0.0014,
|
| 1720 |
+
"step": 2380
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 5.09,
|
| 1724 |
+
"grad_norm": 0.0011308868415653706,
|
| 1725 |
+
"learning_rate": 3.972366148531952e-06,
|
| 1726 |
+
"loss": 0.0001,
|
| 1727 |
+
"step": 2390
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 5.1,
|
| 1731 |
+
"grad_norm": 0.0015520398737862706,
|
| 1732 |
+
"learning_rate": 3.756476683937824e-06,
|
| 1733 |
+
"loss": 0.0001,
|
| 1734 |
+
"step": 2400
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 5.1,
|
| 1738 |
+
"grad_norm": 0.0016552858287468553,
|
| 1739 |
+
"learning_rate": 3.5405872193436963e-06,
|
| 1740 |
+
"loss": 0.0002,
|
| 1741 |
+
"step": 2410
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 5.1,
|
| 1745 |
+
"grad_norm": 0.003429860807955265,
|
| 1746 |
+
"learning_rate": 3.3246977547495685e-06,
|
| 1747 |
+
"loss": 0.0005,
|
| 1748 |
+
"step": 2420
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 5.11,
|
| 1752 |
+
"grad_norm": 3.591295003890991,
|
| 1753 |
+
"learning_rate": 3.1088082901554407e-06,
|
| 1754 |
+
"loss": 0.0029,
|
| 1755 |
+
"step": 2430
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 5.11,
|
| 1759 |
+
"grad_norm": 0.002391704125329852,
|
| 1760 |
+
"learning_rate": 2.892918825561313e-06,
|
| 1761 |
+
"loss": 0.0416,
|
| 1762 |
+
"step": 2440
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 5.12,
|
| 1766 |
+
"grad_norm": 0.0021623370703309774,
|
| 1767 |
+
"learning_rate": 2.677029360967185e-06,
|
| 1768 |
+
"loss": 0.0001,
|
| 1769 |
+
"step": 2450
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 5.12,
|
| 1773 |
+
"grad_norm": 0.006243993528187275,
|
| 1774 |
+
"learning_rate": 2.4611398963730574e-06,
|
| 1775 |
+
"loss": 0.0001,
|
| 1776 |
+
"step": 2460
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 5.12,
|
| 1780 |
+
"grad_norm": 0.004033067263662815,
|
| 1781 |
+
"learning_rate": 2.2452504317789296e-06,
|
| 1782 |
+
"loss": 0.0047,
|
| 1783 |
+
"step": 2470
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 5.13,
|
| 1787 |
+
"grad_norm": 0.002611899748444557,
|
| 1788 |
+
"learning_rate": 2.029360967184802e-06,
|
| 1789 |
+
"loss": 0.0001,
|
| 1790 |
+
"step": 2480
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 5.13,
|
| 1794 |
+
"grad_norm": 0.0029090410098433495,
|
| 1795 |
+
"learning_rate": 1.8134715025906736e-06,
|
| 1796 |
+
"loss": 0.0002,
|
| 1797 |
+
"step": 2490
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 5.14,
|
| 1801 |
+
"grad_norm": 0.007737706881016493,
|
| 1802 |
+
"learning_rate": 1.5975820379965458e-06,
|
| 1803 |
+
"loss": 0.0001,
|
| 1804 |
+
"step": 2500
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 5.14,
|
| 1808 |
+
"grad_norm": 0.0028760808054357767,
|
| 1809 |
+
"learning_rate": 1.381692573402418e-06,
|
| 1810 |
+
"loss": 0.0005,
|
| 1811 |
+
"step": 2510
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 5.14,
|
| 1815 |
+
"grad_norm": 0.2445225566625595,
|
| 1816 |
+
"learning_rate": 1.1658031088082903e-06,
|
| 1817 |
+
"loss": 0.0011,
|
| 1818 |
+
"step": 2520
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 5.15,
|
| 1822 |
+
"grad_norm": 0.001734812743961811,
|
| 1823 |
+
"learning_rate": 9.499136442141624e-07,
|
| 1824 |
+
"loss": 0.0004,
|
| 1825 |
+
"step": 2530
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 5.15,
|
| 1829 |
+
"grad_norm": 0.0022820383310317993,
|
| 1830 |
+
"learning_rate": 7.340241796200346e-07,
|
| 1831 |
+
"loss": 0.0001,
|
| 1832 |
+
"step": 2540
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 5.16,
|
| 1836 |
+
"grad_norm": 0.002981896046549082,
|
| 1837 |
+
"learning_rate": 5.181347150259068e-07,
|
| 1838 |
+
"loss": 0.0002,
|
| 1839 |
+
"step": 2550
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 5.16,
|
| 1843 |
+
"grad_norm": 31.39242935180664,
|
| 1844 |
+
"learning_rate": 3.02245250431779e-07,
|
| 1845 |
+
"loss": 0.037,
|
| 1846 |
+
"step": 2560
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 5.16,
|
| 1850 |
+
"grad_norm": 0.0018530270317569375,
|
| 1851 |
+
"learning_rate": 8.635578583765113e-08,
|
| 1852 |
+
"loss": 0.0002,
|
| 1853 |
+
"step": 2570
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 5.16,
|
| 1857 |
+
"eval_accuracy": 0.9421052631578948,
|
| 1858 |
+
"eval_loss": 0.3020946681499481,
|
| 1859 |
+
"eval_runtime": 526.5435,
|
| 1860 |
+
"eval_samples_per_second": 0.361,
|
| 1861 |
+
"eval_steps_per_second": 0.18,
|
| 1862 |
+
"step": 2574
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 5.16,
|
| 1866 |
+
"step": 2574,
|
| 1867 |
+
"total_flos": 1.3176952627228508e+19,
|
| 1868 |
+
"train_loss": 0.2533098812526059,
|
| 1869 |
+
"train_runtime": 42827.5494,
|
| 1870 |
+
"train_samples_per_second": 0.12,
|
| 1871 |
+
"train_steps_per_second": 0.06
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 5.16,
|
| 1875 |
+
"eval_accuracy": 0.9895833333333334,
|
| 1876 |
+
"eval_loss": 0.11018911004066467,
|
| 1877 |
+
"eval_runtime": 273.5625,
|
| 1878 |
+
"eval_samples_per_second": 0.351,
|
| 1879 |
+
"eval_steps_per_second": 0.175,
|
| 1880 |
+
"step": 2574
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 5.16,
|
| 1884 |
+
"eval_accuracy": 0.9895833333333334,
|
| 1885 |
+
"eval_loss": 0.11018910259008408,
|
| 1886 |
+
"eval_runtime": 269.2698,
|
| 1887 |
+
"eval_samples_per_second": 0.357,
|
| 1888 |
+
"eval_steps_per_second": 0.178,
|
| 1889 |
+
"step": 2574
|
| 1890 |
+
}
|
| 1891 |
+
],
|
| 1892 |
+
"logging_steps": 10,
|
| 1893 |
+
"max_steps": 2574,
|
| 1894 |
+
"num_input_tokens_seen": 0,
|
| 1895 |
+
"num_train_epochs": 9223372036854775807,
|
| 1896 |
+
"save_steps": 500,
|
| 1897 |
+
"total_flos": 1.3176952627228508e+19,
|
| 1898 |
+
"train_batch_size": 2,
|
| 1899 |
+
"trial_name": null,
|
| 1900 |
+
"trial_params": null
|
| 1901 |
+
}
|